diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e80e8628d5a369f7402fb0b4d6c569a64f1ac9c9 --- /dev/null +++ b/README.md @@ -0,0 +1,155 @@ +--- +library_name: peft +license: apache-2.0 +base_model: openlm-research/open_llama_3b_v2 +tags: +- generated_from_trainer +model-index: +- name: outputs/lora-out + results: [] +--- + + + +[Built with Axolotl](https://github.com/axolotl-ai-cloud/axolotl) +
See axolotl config + +axolotl version: `0.5.0` +```yaml +base_model: openlm-research/open_llama_3b_v2 +model_type: LlamaForCausalLM +tokenizer_type: LlamaTokenizer +load_in_8bit: true +load_in_4bit: false +strict: false +push_dataset_to_hub: +datasets: + - path: vicgalle/alpaca-gpt4 + type: alpaca +dataset_prepared_path: +val_set_size: 0.02 +adapter: lora +lora_model_dir: +sequence_len: 1024 +sample_packing: true +lora_r: 8 +lora_alpha: 16 +lora_dropout: 0.0 +lora_target_modules: + - gate_proj + - down_proj + - up_proj + - q_proj + - v_proj + - k_proj + - o_proj +lora_fan_in_fan_out: +wandb_project: +wandb_entity: +wandb_watch: +wandb_name: +wandb_log_model: +output_dir: ./outputs/lora-out +gradient_accumulation_steps: 1 +micro_batch_size: 2 +num_epochs: 4 +optimizer: adamw_bnb_8bit +torchdistx_path: +lr_scheduler: cosine +learning_rate: 0.0002 +train_on_inputs: false +group_by_length: false +bf16: false +fp16: true +tf32: false +gradient_checkpointing: true +early_stopping_patience: +resume_from_checkpoint: +local_rank: +logging_steps: 1 +xformers_attention: +flash_attention: true +gptq_groupsize: +s2_attention: +gptq_model_v1: +warmup_steps: 20 +evals_per_epoch: 4 +saves_per_epoch: 1 +debug: +deepspeed: +weight_decay: 0.1 +fsdp: +fsdp_config: +special_tokens: + bos_token: "" + eos_token: "" + unk_token: "" + +``` + +

+ +# outputs/lora-out + +This model is a fine-tuned version of [openlm-research/open_llama_3b_v2](https://huggingface.co./openlm-research/open_llama_3b_v2) on the None dataset. +It achieves the following results on the evaluation set: +- Loss: 1.1770 + +## Model description + +More information needed + +## Intended uses & limitations + +More information needed + +## Training and evaluation data + +More information needed + +## Training procedure + +### Training hyperparameters + +The following hyperparameters were used during training: +- learning_rate: 0.0002 +- train_batch_size: 2 +- eval_batch_size: 2 +- seed: 42 +- optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments +- lr_scheduler_type: cosine +- lr_scheduler_warmup_steps: 20 +- num_epochs: 4 +- mixed_precision_training: Native AMP + +### Training results + +| Training Loss | Epoch | Step | Validation Loss | +|:-------------:|:------:|:-----:|:---------------:| +| 1.1997 | 0.0002 | 1 | 1.3698 | +| 1.1468 | 0.25 | 1404 | 1.1159 | +| 1.2207 | 0.5 | 2808 | 1.1072 | +| 0.9448 | 0.75 | 4212 | 1.0982 | +| 1.0709 | 1.0 | 5616 | 1.0931 | +| 0.9592 | 1.2498 | 7020 | 1.1051 | +| 1.1133 | 1.4998 | 8424 | 1.1058 | +| 0.884 | 1.7498 | 9828 | 1.1018 | +| 0.9117 | 1.9998 | 11232 | 1.0963 | +| 0.9594 | 2.2496 | 12636 | 1.1336 | +| 0.9034 | 2.4996 | 14040 | 1.1338 | +| 0.6645 | 2.7496 | 15444 | 1.1326 | +| 0.8913 | 2.9996 | 16848 | 1.1309 | +| 0.9476 | 3.2495 | 18252 | 1.1752 | +| 0.9015 | 3.4995 | 19656 | 1.1762 | +| 0.6284 | 3.7495 | 21060 | 1.1768 | +| 0.7522 | 3.9995 | 22464 | 1.1770 | + + +### Framework versions + +- PEFT 0.13.2 +- Transformers 4.46.1 +- Pytorch 2.3.1+cu121 +- Datasets 3.0.1 +- Tokenizers 0.20.3 \ No newline at end of file diff --git a/adapter_config.json b/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6f20a570fc808390da3f2e001093ac1e56c1da --- /dev/null +++ b/adapter_config.json @@ -0,0 +1,34 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "openlm-research/open_llama_3b_v2", + "bias": "none", + "fan_in_fan_out": null, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.0, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "v_proj", + "down_proj", + "o_proj", + "q_proj", + "up_proj", + "k_proj" + ], + "task_type": "CAUSAL_LM", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/adapter_model.bin b/adapter_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..25f077981ec7f2b43cb2ae0684498280c7a98296 --- /dev/null +++ b/adapter_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:65b03db891ccca4d33155d35967b2a3c2b857232e53784271e82e9ce4f562160 +size 50982842 diff --git a/checkpoint-11232/README.md b/checkpoint-11232/README.md new file mode 100644 index 0000000000000000000000000000000000000000..719b4726992f7d0707a4253e9123dec35e4de390 --- /dev/null +++ b/checkpoint-11232/README.md @@ -0,0 +1,202 @@ +--- +base_model: openlm-research/open_llama_3b_v2 +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.13.2 \ No newline at end of file diff --git a/checkpoint-11232/adapter_config.json b/checkpoint-11232/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6f20a570fc808390da3f2e001093ac1e56c1da --- /dev/null +++ b/checkpoint-11232/adapter_config.json @@ -0,0 +1,34 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "openlm-research/open_llama_3b_v2", + "bias": "none", + "fan_in_fan_out": null, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.0, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "v_proj", + "down_proj", + "o_proj", + "q_proj", + "up_proj", + "k_proj" + ], + "task_type": "CAUSAL_LM", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/checkpoint-11232/adapter_model.safetensors b/checkpoint-11232/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5bb3ac9263fab3b5f0b1cc6d25d305fcb5ad958f --- /dev/null +++ b/checkpoint-11232/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91227646d5faac43c528b816cc789555bb955e55a2da268fccb375f6d47f80e2 +size 50899792 diff --git a/checkpoint-11232/optimizer.pt b/checkpoint-11232/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..7819e6803809f44ad4ced2362a126ff855f71d35 --- /dev/null +++ b/checkpoint-11232/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:36ee80f0c39d96999aa9d2716aa41db37fefa4acc0f0bba7d635fb957ebd2168 +size 26231684 diff --git a/checkpoint-11232/rng_state.pth b/checkpoint-11232/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..87571ba044576778d1d6e555eff20ea04c20bbab --- /dev/null +++ b/checkpoint-11232/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5edb34d031c0c2b447f3eaadb401a4c1e7e7e6d8c096e28b7092e01a8bd48c92 +size 14244 diff --git a/checkpoint-11232/scheduler.pt b/checkpoint-11232/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3104cf43abddc114b9a13947bb72f96cafb7034 --- /dev/null +++ b/checkpoint-11232/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ba5b5a15573ac131ab58c742274eeb95853472637445e0705a6a1a1a1234b7c0 +size 1064 diff --git a/checkpoint-11232/special_tokens_map.json b/checkpoint-11232/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/checkpoint-11232/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoint-11232/tokenizer.model b/checkpoint-11232/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..98866ff8ae3631f331c57923c921a0c9ad22b97d --- /dev/null +++ b/checkpoint-11232/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b289e85fa20fd375d8b33dc12f77616f18abc6359804471d1fafcb425fecb8 +size 511574 diff --git a/checkpoint-11232/tokenizer_config.json b/checkpoint-11232/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/checkpoint-11232/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/checkpoint-11232/trainer_state.json b/checkpoint-11232/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..cc8f7a4e47df175615f8a3cbe6d24ae81dd759a6 --- /dev/null +++ b/checkpoint-11232/trainer_state.json @@ -0,0 +1,78729 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.9998219373219372, + "eval_steps": 1404, + "global_step": 11232, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00017806267806267807, + "grad_norm": 0.2854898273944855, + "learning_rate": 1e-05, + "loss": 1.1997, + "step": 1 + }, + { + "epoch": 0.00017806267806267807, + "eval_loss": 1.3698358535766602, + "eval_runtime": 24.1591, + "eval_samples_per_second": 43.089, + "eval_steps_per_second": 21.565, + "step": 1 + }, + { + "epoch": 0.00035612535612535614, + "grad_norm": 0.3508087396621704, + "learning_rate": 2e-05, + "loss": 1.4134, + "step": 2 + }, + { + "epoch": 0.0005341880341880342, + "grad_norm": 0.27050870656967163, + "learning_rate": 3e-05, + "loss": 1.3447, + "step": 3 + }, + { + "epoch": 0.0007122507122507123, + "grad_norm": 0.27706292271614075, + "learning_rate": 4e-05, + "loss": 1.0354, + "step": 4 + }, + { + "epoch": 0.0008903133903133903, + "grad_norm": 0.30398961901664734, + "learning_rate": 5e-05, + "loss": 1.1441, + "step": 5 + }, + { + "epoch": 0.0010683760683760685, + "grad_norm": 0.3103881776332855, + "learning_rate": 6e-05, + "loss": 1.341, + "step": 6 + }, + { + "epoch": 0.0012464387464387464, + "grad_norm": 0.5191189646720886, + "learning_rate": 7e-05, + "loss": 1.3457, + "step": 7 + }, + { + "epoch": 0.0014245014245014246, + "grad_norm": 0.4449467360973358, + "learning_rate": 8e-05, + "loss": 1.5051, + "step": 8 + }, + { + "epoch": 0.0016025641025641025, + "grad_norm": 0.3914581537246704, + "learning_rate": 9e-05, + "loss": 1.5525, + "step": 9 + }, + { + "epoch": 0.0017806267806267807, + "grad_norm": 0.37746086716651917, + "learning_rate": 0.0001, + "loss": 1.3266, + "step": 10 + }, + { + "epoch": 0.001958689458689459, + "grad_norm": 0.35226109623908997, + "learning_rate": 0.00011000000000000002, + "loss": 1.5416, + "step": 11 + }, + { + "epoch": 0.002136752136752137, + "grad_norm": 0.3343672454357147, + "learning_rate": 0.00012, + "loss": 1.3221, + "step": 12 + }, + { + "epoch": 0.0023148148148148147, + "grad_norm": 0.47298333048820496, + "learning_rate": 0.00013000000000000002, + "loss": 1.2999, + "step": 13 + }, + { + "epoch": 0.002492877492877493, + "grad_norm": 0.377814918756485, + "learning_rate": 0.00014, + "loss": 1.1688, + "step": 14 + }, + { + "epoch": 0.002670940170940171, + "grad_norm": 0.46344801783561707, + "learning_rate": 0.00015000000000000001, + "loss": 1.3565, + "step": 15 + }, + { + "epoch": 0.002849002849002849, + "grad_norm": 0.49615249037742615, + "learning_rate": 0.00016, + "loss": 1.5692, + "step": 16 + }, + { + "epoch": 0.003027065527065527, + "grad_norm": 0.5109946131706238, + "learning_rate": 0.00017, + "loss": 1.2991, + "step": 17 + }, + { + "epoch": 0.003205128205128205, + "grad_norm": 0.5125070214271545, + "learning_rate": 0.00018, + "loss": 1.3309, + "step": 18 + }, + { + "epoch": 0.003383190883190883, + "grad_norm": 0.4517767131328583, + "learning_rate": 0.00019, + "loss": 1.357, + "step": 19 + }, + { + "epoch": 0.0035612535612535613, + "grad_norm": 0.47267794609069824, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 20 + }, + { + "epoch": 0.0037393162393162395, + "grad_norm": 0.46823424100875854, + "learning_rate": 0.00019999999902035388, + "loss": 1.1195, + "step": 21 + }, + { + "epoch": 0.003917378917378918, + "grad_norm": 0.440036803483963, + "learning_rate": 0.00019999999608141548, + "loss": 1.2822, + "step": 22 + }, + { + "epoch": 0.004095441595441595, + "grad_norm": 0.371101975440979, + "learning_rate": 0.00019999999118318492, + "loss": 1.132, + "step": 23 + }, + { + "epoch": 0.004273504273504274, + "grad_norm": 0.44691094756126404, + "learning_rate": 0.00019999998432566226, + "loss": 1.2968, + "step": 24 + }, + { + "epoch": 0.004451566951566952, + "grad_norm": 0.5462725162506104, + "learning_rate": 0.0001999999755088476, + "loss": 1.1714, + "step": 25 + }, + { + "epoch": 0.004629629629629629, + "grad_norm": 0.39860013127326965, + "learning_rate": 0.0001999999647327412, + "loss": 1.0407, + "step": 26 + }, + { + "epoch": 0.004807692307692308, + "grad_norm": 0.5031934380531311, + "learning_rate": 0.0001999999519973432, + "loss": 1.2773, + "step": 27 + }, + { + "epoch": 0.004985754985754986, + "grad_norm": 0.42162764072418213, + "learning_rate": 0.0001999999373026539, + "loss": 1.2824, + "step": 28 + }, + { + "epoch": 0.005163817663817663, + "grad_norm": 0.40964868664741516, + "learning_rate": 0.00019999992064867353, + "loss": 1.226, + "step": 29 + }, + { + "epoch": 0.005341880341880342, + "grad_norm": 0.41650915145874023, + "learning_rate": 0.00019999990203540245, + "loss": 1.2677, + "step": 30 + }, + { + "epoch": 0.00551994301994302, + "grad_norm": 0.40052226185798645, + "learning_rate": 0.00019999988146284103, + "loss": 0.9443, + "step": 31 + }, + { + "epoch": 0.005698005698005698, + "grad_norm": 0.5198387503623962, + "learning_rate": 0.00019999985893098964, + "loss": 1.3043, + "step": 32 + }, + { + "epoch": 0.005876068376068376, + "grad_norm": 0.50941002368927, + "learning_rate": 0.00019999983443984878, + "loss": 1.2002, + "step": 33 + }, + { + "epoch": 0.006054131054131054, + "grad_norm": 0.30082932114601135, + "learning_rate": 0.00019999980798941888, + "loss": 0.9904, + "step": 34 + }, + { + "epoch": 0.006232193732193732, + "grad_norm": 0.4228935241699219, + "learning_rate": 0.00019999977957970048, + "loss": 1.1137, + "step": 35 + }, + { + "epoch": 0.00641025641025641, + "grad_norm": 0.41294750571250916, + "learning_rate": 0.0001999997492106941, + "loss": 1.3385, + "step": 36 + }, + { + "epoch": 0.006588319088319089, + "grad_norm": 0.4415493905544281, + "learning_rate": 0.00019999971688240041, + "loss": 1.1695, + "step": 37 + }, + { + "epoch": 0.006766381766381766, + "grad_norm": 0.3726460933685303, + "learning_rate": 0.00019999968259482, + "loss": 1.1734, + "step": 38 + }, + { + "epoch": 0.006944444444444444, + "grad_norm": 0.3969627320766449, + "learning_rate": 0.0001999996463479535, + "loss": 1.1209, + "step": 39 + }, + { + "epoch": 0.007122507122507123, + "grad_norm": 0.3779667913913727, + "learning_rate": 0.0001999996081418017, + "loss": 1.1635, + "step": 40 + }, + { + "epoch": 0.0073005698005698, + "grad_norm": 0.3933636546134949, + "learning_rate": 0.0001999995679763653, + "loss": 1.1514, + "step": 41 + }, + { + "epoch": 0.007478632478632479, + "grad_norm": 0.3567957282066345, + "learning_rate": 0.00019999952585164507, + "loss": 1.2488, + "step": 42 + }, + { + "epoch": 0.007656695156695157, + "grad_norm": 0.32506081461906433, + "learning_rate": 0.00019999948176764186, + "loss": 1.149, + "step": 43 + }, + { + "epoch": 0.007834757834757835, + "grad_norm": 0.46588361263275146, + "learning_rate": 0.0001999994357243566, + "loss": 1.4263, + "step": 44 + }, + { + "epoch": 0.008012820512820512, + "grad_norm": 0.5070307850837708, + "learning_rate": 0.00019999938772179005, + "loss": 1.0698, + "step": 45 + }, + { + "epoch": 0.00819088319088319, + "grad_norm": 0.38199326395988464, + "learning_rate": 0.00019999933775994327, + "loss": 0.9907, + "step": 46 + }, + { + "epoch": 0.00836894586894587, + "grad_norm": 0.43684661388397217, + "learning_rate": 0.0001999992858388172, + "loss": 1.2905, + "step": 47 + }, + { + "epoch": 0.008547008547008548, + "grad_norm": 0.44482162594795227, + "learning_rate": 0.00019999923195841284, + "loss": 1.2153, + "step": 48 + }, + { + "epoch": 0.008725071225071225, + "grad_norm": 0.4259667694568634, + "learning_rate": 0.0001999991761187313, + "loss": 1.1582, + "step": 49 + }, + { + "epoch": 0.008903133903133903, + "grad_norm": 0.41649091243743896, + "learning_rate": 0.00019999911831977357, + "loss": 1.0185, + "step": 50 + }, + { + "epoch": 0.009081196581196582, + "grad_norm": 0.4179716110229492, + "learning_rate": 0.0001999990585615409, + "loss": 1.3579, + "step": 51 + }, + { + "epoch": 0.009259259259259259, + "grad_norm": 0.3372558355331421, + "learning_rate": 0.00019999899684403438, + "loss": 1.0638, + "step": 52 + }, + { + "epoch": 0.009437321937321937, + "grad_norm": 0.41294020414352417, + "learning_rate": 0.00019999893316725525, + "loss": 1.1932, + "step": 53 + }, + { + "epoch": 0.009615384615384616, + "grad_norm": 0.4407919645309448, + "learning_rate": 0.00019999886753120473, + "loss": 1.4129, + "step": 54 + }, + { + "epoch": 0.009793447293447293, + "grad_norm": 0.47948843240737915, + "learning_rate": 0.00019999879993588414, + "loss": 1.2424, + "step": 55 + }, + { + "epoch": 0.009971509971509971, + "grad_norm": 0.3535355031490326, + "learning_rate": 0.00019999873038129484, + "loss": 1.0145, + "step": 56 + }, + { + "epoch": 0.01014957264957265, + "grad_norm": 0.5067078471183777, + "learning_rate": 0.00019999865886743813, + "loss": 1.4708, + "step": 57 + }, + { + "epoch": 0.010327635327635327, + "grad_norm": 0.42862898111343384, + "learning_rate": 0.0001999985853943154, + "loss": 1.0399, + "step": 58 + }, + { + "epoch": 0.010505698005698005, + "grad_norm": 0.4769059419631958, + "learning_rate": 0.00019999850996192816, + "loss": 1.1258, + "step": 59 + }, + { + "epoch": 0.010683760683760684, + "grad_norm": 0.4065442383289337, + "learning_rate": 0.0001999984325702778, + "loss": 1.2077, + "step": 60 + }, + { + "epoch": 0.010861823361823363, + "grad_norm": 0.5318329930305481, + "learning_rate": 0.0001999983532193659, + "loss": 1.2298, + "step": 61 + }, + { + "epoch": 0.01103988603988604, + "grad_norm": 0.4777173101902008, + "learning_rate": 0.000199998271909194, + "loss": 1.3195, + "step": 62 + }, + { + "epoch": 0.011217948717948718, + "grad_norm": 0.37553808093070984, + "learning_rate": 0.0001999981886397637, + "loss": 1.1188, + "step": 63 + }, + { + "epoch": 0.011396011396011397, + "grad_norm": 0.3920556902885437, + "learning_rate": 0.0001999981034110766, + "loss": 1.1448, + "step": 64 + }, + { + "epoch": 0.011574074074074073, + "grad_norm": 0.454272598028183, + "learning_rate": 0.0001999980162231344, + "loss": 1.0812, + "step": 65 + }, + { + "epoch": 0.011752136752136752, + "grad_norm": 0.4354456663131714, + "learning_rate": 0.00019999792707593882, + "loss": 1.1174, + "step": 66 + }, + { + "epoch": 0.01193019943019943, + "grad_norm": 0.5030252933502197, + "learning_rate": 0.00019999783596949156, + "loss": 1.2925, + "step": 67 + }, + { + "epoch": 0.012108262108262107, + "grad_norm": 0.5141571164131165, + "learning_rate": 0.00019999774290379446, + "loss": 1.6193, + "step": 68 + }, + { + "epoch": 0.012286324786324786, + "grad_norm": 0.417298287153244, + "learning_rate": 0.0001999976478788493, + "loss": 1.1875, + "step": 69 + }, + { + "epoch": 0.012464387464387465, + "grad_norm": 0.4642415940761566, + "learning_rate": 0.00019999755089465795, + "loss": 1.4138, + "step": 70 + }, + { + "epoch": 0.012642450142450143, + "grad_norm": 0.43184754252433777, + "learning_rate": 0.0001999974519512223, + "loss": 1.0697, + "step": 71 + }, + { + "epoch": 0.01282051282051282, + "grad_norm": 0.46698349714279175, + "learning_rate": 0.00019999735104854436, + "loss": 0.709, + "step": 72 + }, + { + "epoch": 0.012998575498575499, + "grad_norm": 0.37253814935684204, + "learning_rate": 0.000199997248186626, + "loss": 1.2084, + "step": 73 + }, + { + "epoch": 0.013176638176638177, + "grad_norm": 0.3851388692855835, + "learning_rate": 0.0001999971433654693, + "loss": 1.0548, + "step": 74 + }, + { + "epoch": 0.013354700854700854, + "grad_norm": 0.4434688985347748, + "learning_rate": 0.00019999703658507635, + "loss": 1.4084, + "step": 75 + }, + { + "epoch": 0.013532763532763533, + "grad_norm": 0.43164482712745667, + "learning_rate": 0.00019999692784544913, + "loss": 1.4872, + "step": 76 + }, + { + "epoch": 0.013710826210826211, + "grad_norm": 0.4224303364753723, + "learning_rate": 0.00019999681714658984, + "loss": 1.2221, + "step": 77 + }, + { + "epoch": 0.013888888888888888, + "grad_norm": 0.35588955879211426, + "learning_rate": 0.00019999670448850069, + "loss": 0.84, + "step": 78 + }, + { + "epoch": 0.014066951566951567, + "grad_norm": 0.3970590829849243, + "learning_rate": 0.0001999965898711838, + "loss": 1.1886, + "step": 79 + }, + { + "epoch": 0.014245014245014245, + "grad_norm": 0.4331924319267273, + "learning_rate": 0.00019999647329464146, + "loss": 1.179, + "step": 80 + }, + { + "epoch": 0.014423076923076924, + "grad_norm": 0.4226946234703064, + "learning_rate": 0.00019999635475887598, + "loss": 1.1496, + "step": 81 + }, + { + "epoch": 0.0146011396011396, + "grad_norm": 0.381592720746994, + "learning_rate": 0.00019999623426388962, + "loss": 1.1774, + "step": 82 + }, + { + "epoch": 0.01477920227920228, + "grad_norm": 0.4190855622291565, + "learning_rate": 0.00019999611180968478, + "loss": 1.1491, + "step": 83 + }, + { + "epoch": 0.014957264957264958, + "grad_norm": 0.3904292583465576, + "learning_rate": 0.00019999598739626389, + "loss": 1.1275, + "step": 84 + }, + { + "epoch": 0.015135327635327635, + "grad_norm": 0.4515478014945984, + "learning_rate": 0.0001999958610236293, + "loss": 1.2404, + "step": 85 + }, + { + "epoch": 0.015313390313390313, + "grad_norm": 0.48341724276542664, + "learning_rate": 0.00019999573269178359, + "loss": 1.3572, + "step": 86 + }, + { + "epoch": 0.015491452991452992, + "grad_norm": 0.42150333523750305, + "learning_rate": 0.00019999560240072914, + "loss": 1.0203, + "step": 87 + }, + { + "epoch": 0.01566951566951567, + "grad_norm": 0.45445525646209717, + "learning_rate": 0.00019999547015046867, + "loss": 1.0677, + "step": 88 + }, + { + "epoch": 0.01584757834757835, + "grad_norm": 0.3581015467643738, + "learning_rate": 0.00019999533594100463, + "loss": 1.0693, + "step": 89 + }, + { + "epoch": 0.016025641025641024, + "grad_norm": 0.4430878758430481, + "learning_rate": 0.00019999519977233971, + "loss": 1.1591, + "step": 90 + }, + { + "epoch": 0.016203703703703703, + "grad_norm": 0.3940352201461792, + "learning_rate": 0.0001999950616444766, + "loss": 1.1325, + "step": 91 + }, + { + "epoch": 0.01638176638176638, + "grad_norm": 0.4521673321723938, + "learning_rate": 0.00019999492155741794, + "loss": 1.3288, + "step": 92 + }, + { + "epoch": 0.01655982905982906, + "grad_norm": 0.3988296687602997, + "learning_rate": 0.00019999477951116658, + "loss": 1.0023, + "step": 93 + }, + { + "epoch": 0.01673789173789174, + "grad_norm": 0.38709723949432373, + "learning_rate": 0.00019999463550572516, + "loss": 1.2623, + "step": 94 + }, + { + "epoch": 0.016915954415954417, + "grad_norm": 0.35376182198524475, + "learning_rate": 0.00019999448954109662, + "loss": 1.0643, + "step": 95 + }, + { + "epoch": 0.017094017094017096, + "grad_norm": 0.49547120928764343, + "learning_rate": 0.00019999434161728377, + "loss": 1.2121, + "step": 96 + }, + { + "epoch": 0.01727207977207977, + "grad_norm": 0.49593672156333923, + "learning_rate": 0.00019999419173428952, + "loss": 1.1635, + "step": 97 + }, + { + "epoch": 0.01745014245014245, + "grad_norm": 0.4146541953086853, + "learning_rate": 0.0001999940398921168, + "loss": 1.1452, + "step": 98 + }, + { + "epoch": 0.017628205128205128, + "grad_norm": 0.5177254676818848, + "learning_rate": 0.00019999388609076858, + "loss": 1.2178, + "step": 99 + }, + { + "epoch": 0.017806267806267807, + "grad_norm": 0.4012768864631653, + "learning_rate": 0.0001999937303302479, + "loss": 0.9222, + "step": 100 + }, + { + "epoch": 0.017984330484330485, + "grad_norm": 0.4597131907939911, + "learning_rate": 0.00019999357261055777, + "loss": 0.979, + "step": 101 + }, + { + "epoch": 0.018162393162393164, + "grad_norm": 0.6190966963768005, + "learning_rate": 0.00019999341293170132, + "loss": 1.3909, + "step": 102 + }, + { + "epoch": 0.01834045584045584, + "grad_norm": 0.4576462209224701, + "learning_rate": 0.00019999325129368164, + "loss": 1.073, + "step": 103 + }, + { + "epoch": 0.018518518518518517, + "grad_norm": 0.4036749005317688, + "learning_rate": 0.00019999308769650192, + "loss": 1.1354, + "step": 104 + }, + { + "epoch": 0.018696581196581196, + "grad_norm": 0.4722452759742737, + "learning_rate": 0.00019999292214016538, + "loss": 1.2039, + "step": 105 + }, + { + "epoch": 0.018874643874643875, + "grad_norm": 0.5338274240493774, + "learning_rate": 0.00019999275462467527, + "loss": 1.225, + "step": 106 + }, + { + "epoch": 0.019052706552706553, + "grad_norm": 0.4301491677761078, + "learning_rate": 0.00019999258515003484, + "loss": 1.0601, + "step": 107 + }, + { + "epoch": 0.019230769230769232, + "grad_norm": 0.33271175622940063, + "learning_rate": 0.0001999924137162474, + "loss": 0.8441, + "step": 108 + }, + { + "epoch": 0.01940883190883191, + "grad_norm": 0.4648784399032593, + "learning_rate": 0.0001999922403233163, + "loss": 1.2038, + "step": 109 + }, + { + "epoch": 0.019586894586894586, + "grad_norm": 0.37915176153182983, + "learning_rate": 0.00019999206497124504, + "loss": 1.0923, + "step": 110 + }, + { + "epoch": 0.019764957264957264, + "grad_norm": 0.3865506052970886, + "learning_rate": 0.00019999188766003695, + "loss": 0.9535, + "step": 111 + }, + { + "epoch": 0.019943019943019943, + "grad_norm": 0.35739636421203613, + "learning_rate": 0.0001999917083896955, + "loss": 1.2688, + "step": 112 + }, + { + "epoch": 0.02012108262108262, + "grad_norm": 0.3943796157836914, + "learning_rate": 0.0001999915271602243, + "loss": 1.1097, + "step": 113 + }, + { + "epoch": 0.0202991452991453, + "grad_norm": 0.44758161902427673, + "learning_rate": 0.0001999913439716268, + "loss": 1.2698, + "step": 114 + }, + { + "epoch": 0.02047720797720798, + "grad_norm": 0.3749747574329376, + "learning_rate": 0.00019999115882390664, + "loss": 1.1091, + "step": 115 + }, + { + "epoch": 0.020655270655270654, + "grad_norm": 0.3479487895965576, + "learning_rate": 0.00019999097171706745, + "loss": 1.0049, + "step": 116 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 0.4491243064403534, + "learning_rate": 0.00019999078265111285, + "loss": 1.1857, + "step": 117 + }, + { + "epoch": 0.02101139601139601, + "grad_norm": 0.345289021730423, + "learning_rate": 0.00019999059162604662, + "loss": 1.1397, + "step": 118 + }, + { + "epoch": 0.02118945868945869, + "grad_norm": 0.5467649698257446, + "learning_rate": 0.00019999039864187243, + "loss": 1.2196, + "step": 119 + }, + { + "epoch": 0.021367521367521368, + "grad_norm": 0.36446481943130493, + "learning_rate": 0.00019999020369859409, + "loss": 0.796, + "step": 120 + }, + { + "epoch": 0.021545584045584047, + "grad_norm": 0.4225841760635376, + "learning_rate": 0.00019999000679621543, + "loss": 0.9684, + "step": 121 + }, + { + "epoch": 0.021723646723646725, + "grad_norm": 0.4205594062805176, + "learning_rate": 0.0001999898079347403, + "loss": 1.2762, + "step": 122 + }, + { + "epoch": 0.0219017094017094, + "grad_norm": 0.43773892521858215, + "learning_rate": 0.00019998960711417257, + "loss": 1.117, + "step": 123 + }, + { + "epoch": 0.02207977207977208, + "grad_norm": 0.41279685497283936, + "learning_rate": 0.00019998940433451623, + "loss": 1.1502, + "step": 124 + }, + { + "epoch": 0.022257834757834757, + "grad_norm": 0.4090803563594818, + "learning_rate": 0.0001999891995957752, + "loss": 1.2591, + "step": 125 + }, + { + "epoch": 0.022435897435897436, + "grad_norm": 0.6000410914421082, + "learning_rate": 0.0001999889928979535, + "loss": 1.4321, + "step": 126 + }, + { + "epoch": 0.022613960113960115, + "grad_norm": 0.524264395236969, + "learning_rate": 0.00019998878424105524, + "loss": 1.1849, + "step": 127 + }, + { + "epoch": 0.022792022792022793, + "grad_norm": 0.4581047296524048, + "learning_rate": 0.00019998857362508443, + "loss": 1.0598, + "step": 128 + }, + { + "epoch": 0.022970085470085472, + "grad_norm": 0.42663446068763733, + "learning_rate": 0.00019998836105004526, + "loss": 1.1909, + "step": 129 + }, + { + "epoch": 0.023148148148148147, + "grad_norm": 0.45709118247032166, + "learning_rate": 0.00019998814651594183, + "loss": 1.2104, + "step": 130 + }, + { + "epoch": 0.023326210826210825, + "grad_norm": 0.39528369903564453, + "learning_rate": 0.0001999879300227784, + "loss": 1.3073, + "step": 131 + }, + { + "epoch": 0.023504273504273504, + "grad_norm": 0.46896448731422424, + "learning_rate": 0.00019998771157055914, + "loss": 1.3202, + "step": 132 + }, + { + "epoch": 0.023682336182336183, + "grad_norm": 0.4386129677295685, + "learning_rate": 0.00019998749115928842, + "loss": 1.2196, + "step": 133 + }, + { + "epoch": 0.02386039886039886, + "grad_norm": 0.45920488238334656, + "learning_rate": 0.00019998726878897051, + "loss": 1.3668, + "step": 134 + }, + { + "epoch": 0.02403846153846154, + "grad_norm": 0.4115797281265259, + "learning_rate": 0.0001999870444596098, + "loss": 1.1052, + "step": 135 + }, + { + "epoch": 0.024216524216524215, + "grad_norm": 0.3860839903354645, + "learning_rate": 0.0001999868181712106, + "loss": 1.0344, + "step": 136 + }, + { + "epoch": 0.024394586894586893, + "grad_norm": 0.42514732480049133, + "learning_rate": 0.00019998658992377742, + "loss": 1.1979, + "step": 137 + }, + { + "epoch": 0.024572649572649572, + "grad_norm": 0.36001840233802795, + "learning_rate": 0.00019998635971731475, + "loss": 1.4536, + "step": 138 + }, + { + "epoch": 0.02475071225071225, + "grad_norm": 0.3739112317562103, + "learning_rate": 0.00019998612755182707, + "loss": 1.0097, + "step": 139 + }, + { + "epoch": 0.02492877492877493, + "grad_norm": 0.37545472383499146, + "learning_rate": 0.00019998589342731888, + "loss": 0.829, + "step": 140 + }, + { + "epoch": 0.025106837606837608, + "grad_norm": 0.38660728931427, + "learning_rate": 0.0001999856573437948, + "loss": 1.1324, + "step": 141 + }, + { + "epoch": 0.025284900284900286, + "grad_norm": 0.3741356432437897, + "learning_rate": 0.00019998541930125953, + "loss": 1.0934, + "step": 142 + }, + { + "epoch": 0.02546296296296296, + "grad_norm": 0.41900336742401123, + "learning_rate": 0.00019998517929971764, + "loss": 1.0336, + "step": 143 + }, + { + "epoch": 0.02564102564102564, + "grad_norm": 0.4167572259902954, + "learning_rate": 0.00019998493733917384, + "loss": 1.2571, + "step": 144 + }, + { + "epoch": 0.02581908831908832, + "grad_norm": 0.39437636733055115, + "learning_rate": 0.0001999846934196329, + "loss": 1.2283, + "step": 145 + }, + { + "epoch": 0.025997150997150997, + "grad_norm": 0.39129480719566345, + "learning_rate": 0.00019998444754109964, + "loss": 0.9893, + "step": 146 + }, + { + "epoch": 0.026175213675213676, + "grad_norm": 0.45533549785614014, + "learning_rate": 0.0001999841997035788, + "loss": 1.0793, + "step": 147 + }, + { + "epoch": 0.026353276353276354, + "grad_norm": 0.3741768002510071, + "learning_rate": 0.00019998394990707524, + "loss": 1.2179, + "step": 148 + }, + { + "epoch": 0.026531339031339033, + "grad_norm": 0.4066533148288727, + "learning_rate": 0.0001999836981515939, + "loss": 1.1443, + "step": 149 + }, + { + "epoch": 0.026709401709401708, + "grad_norm": 0.4851688742637634, + "learning_rate": 0.0001999834444371397, + "loss": 1.1668, + "step": 150 + }, + { + "epoch": 0.026887464387464387, + "grad_norm": 0.428091436624527, + "learning_rate": 0.0001999831887637176, + "loss": 1.2676, + "step": 151 + }, + { + "epoch": 0.027065527065527065, + "grad_norm": 0.4024655222892761, + "learning_rate": 0.0001999829311313326, + "loss": 1.3115, + "step": 152 + }, + { + "epoch": 0.027243589743589744, + "grad_norm": 0.43983033299446106, + "learning_rate": 0.00019998267153998976, + "loss": 1.1019, + "step": 153 + }, + { + "epoch": 0.027421652421652423, + "grad_norm": 0.4317505359649658, + "learning_rate": 0.0001999824099896942, + "loss": 1.3129, + "step": 154 + }, + { + "epoch": 0.0275997150997151, + "grad_norm": 0.43107882142066956, + "learning_rate": 0.000199982146480451, + "loss": 1.2134, + "step": 155 + }, + { + "epoch": 0.027777777777777776, + "grad_norm": 0.3939448297023773, + "learning_rate": 0.00019998188101226532, + "loss": 1.0321, + "step": 156 + }, + { + "epoch": 0.027955840455840455, + "grad_norm": 0.4641847610473633, + "learning_rate": 0.00019998161358514237, + "loss": 1.2369, + "step": 157 + }, + { + "epoch": 0.028133903133903133, + "grad_norm": 0.3538529872894287, + "learning_rate": 0.0001999813441990874, + "loss": 1.2061, + "step": 158 + }, + { + "epoch": 0.028311965811965812, + "grad_norm": 0.3277950584888458, + "learning_rate": 0.0001999810728541057, + "loss": 0.9419, + "step": 159 + }, + { + "epoch": 0.02849002849002849, + "grad_norm": 0.424710750579834, + "learning_rate": 0.00019998079955020254, + "loss": 1.3302, + "step": 160 + }, + { + "epoch": 0.02866809116809117, + "grad_norm": 0.4120834469795227, + "learning_rate": 0.00019998052428738333, + "loss": 1.079, + "step": 161 + }, + { + "epoch": 0.028846153846153848, + "grad_norm": 0.45811930298805237, + "learning_rate": 0.00019998024706565346, + "loss": 1.1259, + "step": 162 + }, + { + "epoch": 0.029024216524216523, + "grad_norm": 0.3873266875743866, + "learning_rate": 0.0001999799678850183, + "loss": 1.2124, + "step": 163 + }, + { + "epoch": 0.0292022792022792, + "grad_norm": 0.5806412696838379, + "learning_rate": 0.00019997968674548337, + "loss": 1.3467, + "step": 164 + }, + { + "epoch": 0.02938034188034188, + "grad_norm": 0.3906802833080292, + "learning_rate": 0.00019997940364705418, + "loss": 1.1438, + "step": 165 + }, + { + "epoch": 0.02955840455840456, + "grad_norm": 0.45201995968818665, + "learning_rate": 0.00019997911858973626, + "loss": 1.1469, + "step": 166 + }, + { + "epoch": 0.029736467236467237, + "grad_norm": 0.4965892732143402, + "learning_rate": 0.0001999788315735352, + "loss": 1.0829, + "step": 167 + }, + { + "epoch": 0.029914529914529916, + "grad_norm": 0.32578057050704956, + "learning_rate": 0.0001999785425984566, + "loss": 1.0432, + "step": 168 + }, + { + "epoch": 0.03009259259259259, + "grad_norm": 0.4146028161048889, + "learning_rate": 0.00019997825166450617, + "loss": 1.1657, + "step": 169 + }, + { + "epoch": 0.03027065527065527, + "grad_norm": 0.4342964291572571, + "learning_rate": 0.0001999779587716896, + "loss": 1.2038, + "step": 170 + }, + { + "epoch": 0.030448717948717948, + "grad_norm": 0.40128546953201294, + "learning_rate": 0.00019997766392001258, + "loss": 1.3044, + "step": 171 + }, + { + "epoch": 0.030626780626780627, + "grad_norm": 0.4357539117336273, + "learning_rate": 0.00019997736710948094, + "loss": 1.2143, + "step": 172 + }, + { + "epoch": 0.030804843304843305, + "grad_norm": 0.4821035861968994, + "learning_rate": 0.00019997706834010045, + "loss": 1.0469, + "step": 173 + }, + { + "epoch": 0.030982905982905984, + "grad_norm": 0.3966675102710724, + "learning_rate": 0.000199976767611877, + "loss": 1.2122, + "step": 174 + }, + { + "epoch": 0.031160968660968662, + "grad_norm": 0.4265064299106598, + "learning_rate": 0.00019997646492481648, + "loss": 1.0871, + "step": 175 + }, + { + "epoch": 0.03133903133903134, + "grad_norm": 0.3445652723312378, + "learning_rate": 0.00019997616027892485, + "loss": 1.0412, + "step": 176 + }, + { + "epoch": 0.031517094017094016, + "grad_norm": 0.47187718749046326, + "learning_rate": 0.000199975853674208, + "loss": 1.0822, + "step": 177 + }, + { + "epoch": 0.0316951566951567, + "grad_norm": 0.37751707434654236, + "learning_rate": 0.000199975545110672, + "loss": 1.1439, + "step": 178 + }, + { + "epoch": 0.03187321937321937, + "grad_norm": 0.38792455196380615, + "learning_rate": 0.00019997523458832286, + "loss": 0.8604, + "step": 179 + }, + { + "epoch": 0.03205128205128205, + "grad_norm": 0.35199594497680664, + "learning_rate": 0.00019997492210716667, + "loss": 1.0819, + "step": 180 + }, + { + "epoch": 0.03222934472934473, + "grad_norm": 0.4828922748565674, + "learning_rate": 0.00019997460766720958, + "loss": 1.1879, + "step": 181 + }, + { + "epoch": 0.032407407407407406, + "grad_norm": 0.46153363585472107, + "learning_rate": 0.00019997429126845774, + "loss": 1.1592, + "step": 182 + }, + { + "epoch": 0.03258547008547009, + "grad_norm": 0.4844890832901001, + "learning_rate": 0.0001999739729109173, + "loss": 1.1334, + "step": 183 + }, + { + "epoch": 0.03276353276353276, + "grad_norm": 0.414617121219635, + "learning_rate": 0.00019997365259459457, + "loss": 1.0547, + "step": 184 + }, + { + "epoch": 0.032941595441595445, + "grad_norm": 0.46544626355171204, + "learning_rate": 0.00019997333031949581, + "loss": 1.4067, + "step": 185 + }, + { + "epoch": 0.03311965811965812, + "grad_norm": 0.48489415645599365, + "learning_rate": 0.0001999730060856273, + "loss": 1.4027, + "step": 186 + }, + { + "epoch": 0.033297720797720795, + "grad_norm": 0.3963346481323242, + "learning_rate": 0.0001999726798929954, + "loss": 1.1327, + "step": 187 + }, + { + "epoch": 0.03347578347578348, + "grad_norm": 0.3809385895729065, + "learning_rate": 0.00019997235174160652, + "loss": 1.3475, + "step": 188 + }, + { + "epoch": 0.03365384615384615, + "grad_norm": 0.3866960406303406, + "learning_rate": 0.0001999720216314671, + "loss": 1.1576, + "step": 189 + }, + { + "epoch": 0.033831908831908834, + "grad_norm": 0.34976935386657715, + "learning_rate": 0.00019997168956258356, + "loss": 0.9361, + "step": 190 + }, + { + "epoch": 0.03400997150997151, + "grad_norm": 0.38681939244270325, + "learning_rate": 0.00019997135553496243, + "loss": 1.1796, + "step": 191 + }, + { + "epoch": 0.03418803418803419, + "grad_norm": 0.41905197501182556, + "learning_rate": 0.0001999710195486103, + "loss": 1.1714, + "step": 192 + }, + { + "epoch": 0.03436609686609687, + "grad_norm": 0.42356589436531067, + "learning_rate": 0.0001999706816035337, + "loss": 1.0022, + "step": 193 + }, + { + "epoch": 0.03454415954415954, + "grad_norm": 0.3929740786552429, + "learning_rate": 0.00019997034169973925, + "loss": 1.3769, + "step": 194 + }, + { + "epoch": 0.034722222222222224, + "grad_norm": 0.4325186312198639, + "learning_rate": 0.00019996999983723366, + "loss": 1.3057, + "step": 195 + }, + { + "epoch": 0.0349002849002849, + "grad_norm": 0.3954029381275177, + "learning_rate": 0.00019996965601602355, + "loss": 1.1958, + "step": 196 + }, + { + "epoch": 0.03507834757834758, + "grad_norm": 0.34454262256622314, + "learning_rate": 0.00019996931023611572, + "loss": 1.0972, + "step": 197 + }, + { + "epoch": 0.035256410256410256, + "grad_norm": 0.48900291323661804, + "learning_rate": 0.0001999689624975169, + "loss": 1.213, + "step": 198 + }, + { + "epoch": 0.03543447293447293, + "grad_norm": 0.35214388370513916, + "learning_rate": 0.00019996861280023397, + "loss": 1.0285, + "step": 199 + }, + { + "epoch": 0.03561253561253561, + "grad_norm": 0.49393126368522644, + "learning_rate": 0.00019996826114427373, + "loss": 1.2313, + "step": 200 + }, + { + "epoch": 0.03579059829059829, + "grad_norm": 0.3994458019733429, + "learning_rate": 0.00019996790752964305, + "loss": 1.0474, + "step": 201 + }, + { + "epoch": 0.03596866096866097, + "grad_norm": 0.5387318730354309, + "learning_rate": 0.0001999675519563489, + "loss": 1.3067, + "step": 202 + }, + { + "epoch": 0.036146723646723646, + "grad_norm": 0.4976751208305359, + "learning_rate": 0.00019996719442439824, + "loss": 1.2593, + "step": 203 + }, + { + "epoch": 0.03632478632478633, + "grad_norm": 0.47052907943725586, + "learning_rate": 0.0001999668349337981, + "loss": 1.1036, + "step": 204 + }, + { + "epoch": 0.036502849002849, + "grad_norm": 0.39616644382476807, + "learning_rate": 0.00019996647348455543, + "loss": 1.0481, + "step": 205 + }, + { + "epoch": 0.03668091168091168, + "grad_norm": 0.42987677454948425, + "learning_rate": 0.00019996611007667742, + "loss": 1.0923, + "step": 206 + }, + { + "epoch": 0.03685897435897436, + "grad_norm": 0.47065848112106323, + "learning_rate": 0.00019996574471017113, + "loss": 1.1403, + "step": 207 + }, + { + "epoch": 0.037037037037037035, + "grad_norm": 0.4363015592098236, + "learning_rate": 0.00019996537738504373, + "loss": 1.253, + "step": 208 + }, + { + "epoch": 0.03721509971509972, + "grad_norm": 0.4038296937942505, + "learning_rate": 0.00019996500810130243, + "loss": 1.1679, + "step": 209 + }, + { + "epoch": 0.03739316239316239, + "grad_norm": 0.5038532018661499, + "learning_rate": 0.00019996463685895445, + "loss": 1.1182, + "step": 210 + }, + { + "epoch": 0.037571225071225074, + "grad_norm": 0.37740692496299744, + "learning_rate": 0.00019996426365800706, + "loss": 1.0465, + "step": 211 + }, + { + "epoch": 0.03774928774928775, + "grad_norm": 0.47794604301452637, + "learning_rate": 0.00019996388849846759, + "loss": 1.2836, + "step": 212 + }, + { + "epoch": 0.037927350427350424, + "grad_norm": 0.38460609316825867, + "learning_rate": 0.0001999635113803434, + "loss": 1.2099, + "step": 213 + }, + { + "epoch": 0.038105413105413107, + "grad_norm": 0.42016157507896423, + "learning_rate": 0.0001999631323036418, + "loss": 1.152, + "step": 214 + }, + { + "epoch": 0.03828347578347578, + "grad_norm": 0.4024946391582489, + "learning_rate": 0.00019996275126837033, + "loss": 1.1534, + "step": 215 + }, + { + "epoch": 0.038461538461538464, + "grad_norm": 0.4573793411254883, + "learning_rate": 0.00019996236827453642, + "loss": 1.2019, + "step": 216 + }, + { + "epoch": 0.03863960113960114, + "grad_norm": 0.3642503321170807, + "learning_rate": 0.0001999619833221475, + "loss": 1.0541, + "step": 217 + }, + { + "epoch": 0.03881766381766382, + "grad_norm": 0.38492897152900696, + "learning_rate": 0.0001999615964112112, + "loss": 1.1269, + "step": 218 + }, + { + "epoch": 0.038995726495726496, + "grad_norm": 0.427219420671463, + "learning_rate": 0.0001999612075417351, + "loss": 1.1126, + "step": 219 + }, + { + "epoch": 0.03917378917378917, + "grad_norm": 0.40781742334365845, + "learning_rate": 0.00019996081671372676, + "loss": 1.2207, + "step": 220 + }, + { + "epoch": 0.03935185185185185, + "grad_norm": 0.39229512214660645, + "learning_rate": 0.00019996042392719386, + "loss": 1.0403, + "step": 221 + }, + { + "epoch": 0.03952991452991453, + "grad_norm": 0.42038577795028687, + "learning_rate": 0.0001999600291821441, + "loss": 1.2157, + "step": 222 + }, + { + "epoch": 0.03970797720797721, + "grad_norm": 0.3963491916656494, + "learning_rate": 0.00019995963247858525, + "loss": 1.0532, + "step": 223 + }, + { + "epoch": 0.039886039886039885, + "grad_norm": 0.4389874041080475, + "learning_rate": 0.00019995923381652502, + "loss": 1.4279, + "step": 224 + }, + { + "epoch": 0.04006410256410257, + "grad_norm": 0.357312947511673, + "learning_rate": 0.00019995883319597123, + "loss": 0.9871, + "step": 225 + }, + { + "epoch": 0.04024216524216524, + "grad_norm": 0.3644427955150604, + "learning_rate": 0.00019995843061693181, + "loss": 1.0879, + "step": 226 + }, + { + "epoch": 0.04042022792022792, + "grad_norm": 0.4074651002883911, + "learning_rate": 0.00019995802607941453, + "loss": 1.2138, + "step": 227 + }, + { + "epoch": 0.0405982905982906, + "grad_norm": 0.40709465742111206, + "learning_rate": 0.0001999576195834274, + "loss": 1.1905, + "step": 228 + }, + { + "epoch": 0.040776353276353275, + "grad_norm": 0.4280182719230652, + "learning_rate": 0.00019995721112897838, + "loss": 1.2331, + "step": 229 + }, + { + "epoch": 0.04095441595441596, + "grad_norm": 0.37846076488494873, + "learning_rate": 0.00019995680071607544, + "loss": 1.078, + "step": 230 + }, + { + "epoch": 0.04113247863247863, + "grad_norm": 0.3877260088920593, + "learning_rate": 0.0001999563883447266, + "loss": 1.0309, + "step": 231 + }, + { + "epoch": 0.04131054131054131, + "grad_norm": 0.42886826395988464, + "learning_rate": 0.00019995597401494, + "loss": 1.0403, + "step": 232 + }, + { + "epoch": 0.04148860398860399, + "grad_norm": 0.4316534101963043, + "learning_rate": 0.00019995555772672372, + "loss": 1.2418, + "step": 233 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 0.45768865942955017, + "learning_rate": 0.00019995513948008593, + "loss": 1.233, + "step": 234 + }, + { + "epoch": 0.041844729344729346, + "grad_norm": 0.5647913813591003, + "learning_rate": 0.00019995471927503481, + "loss": 1.1346, + "step": 235 + }, + { + "epoch": 0.04202279202279202, + "grad_norm": 0.3797492980957031, + "learning_rate": 0.00019995429711157863, + "loss": 1.1574, + "step": 236 + }, + { + "epoch": 0.042200854700854704, + "grad_norm": 0.4392767548561096, + "learning_rate": 0.00019995387298972562, + "loss": 0.8988, + "step": 237 + }, + { + "epoch": 0.04237891737891738, + "grad_norm": 0.37331557273864746, + "learning_rate": 0.0001999534469094841, + "loss": 1.0439, + "step": 238 + }, + { + "epoch": 0.042556980056980054, + "grad_norm": 0.3785935938358307, + "learning_rate": 0.00019995301887086245, + "loss": 0.9839, + "step": 239 + }, + { + "epoch": 0.042735042735042736, + "grad_norm": 0.4351862668991089, + "learning_rate": 0.00019995258887386898, + "loss": 1.2653, + "step": 240 + }, + { + "epoch": 0.04291310541310541, + "grad_norm": 0.399475634098053, + "learning_rate": 0.0001999521569185122, + "loss": 0.9877, + "step": 241 + }, + { + "epoch": 0.04309116809116809, + "grad_norm": 0.42332810163497925, + "learning_rate": 0.00019995172300480053, + "loss": 1.2403, + "step": 242 + }, + { + "epoch": 0.04326923076923077, + "grad_norm": 0.4397708475589752, + "learning_rate": 0.00019995128713274247, + "loss": 0.9316, + "step": 243 + }, + { + "epoch": 0.04344729344729345, + "grad_norm": 0.3614110052585602, + "learning_rate": 0.00019995084930234658, + "loss": 1.1088, + "step": 244 + }, + { + "epoch": 0.043625356125356125, + "grad_norm": 0.39433717727661133, + "learning_rate": 0.0001999504095136214, + "loss": 1.2002, + "step": 245 + }, + { + "epoch": 0.0438034188034188, + "grad_norm": 0.33088216185569763, + "learning_rate": 0.0001999499677665756, + "loss": 0.8796, + "step": 246 + }, + { + "epoch": 0.04398148148148148, + "grad_norm": 0.5239143967628479, + "learning_rate": 0.00019994952406121784, + "loss": 1.2808, + "step": 247 + }, + { + "epoch": 0.04415954415954416, + "grad_norm": 0.42156723141670227, + "learning_rate": 0.00019994907839755675, + "loss": 1.1775, + "step": 248 + }, + { + "epoch": 0.04433760683760684, + "grad_norm": 0.42569902539253235, + "learning_rate": 0.0001999486307756011, + "loss": 1.001, + "step": 249 + }, + { + "epoch": 0.044515669515669515, + "grad_norm": 0.38241544365882874, + "learning_rate": 0.00019994818119535964, + "loss": 1.1064, + "step": 250 + }, + { + "epoch": 0.0446937321937322, + "grad_norm": 0.4185071885585785, + "learning_rate": 0.0001999477296568412, + "loss": 1.2109, + "step": 251 + }, + { + "epoch": 0.04487179487179487, + "grad_norm": 0.4189644157886505, + "learning_rate": 0.00019994727616005464, + "loss": 1.2902, + "step": 252 + }, + { + "epoch": 0.04504985754985755, + "grad_norm": 0.34671884775161743, + "learning_rate": 0.0001999468207050088, + "loss": 0.9429, + "step": 253 + }, + { + "epoch": 0.04522792022792023, + "grad_norm": 0.42391687631607056, + "learning_rate": 0.00019994636329171266, + "loss": 0.7179, + "step": 254 + }, + { + "epoch": 0.045405982905982904, + "grad_norm": 0.3803195655345917, + "learning_rate": 0.00019994590392017513, + "loss": 1.0318, + "step": 255 + }, + { + "epoch": 0.045584045584045586, + "grad_norm": 0.3389956057071686, + "learning_rate": 0.00019994544259040525, + "loss": 1.0485, + "step": 256 + }, + { + "epoch": 0.04576210826210826, + "grad_norm": 0.4927038550376892, + "learning_rate": 0.000199944979302412, + "loss": 1.3426, + "step": 257 + }, + { + "epoch": 0.045940170940170943, + "grad_norm": 0.33200421929359436, + "learning_rate": 0.00019994451405620453, + "loss": 1.0071, + "step": 258 + }, + { + "epoch": 0.04611823361823362, + "grad_norm": 0.38028615713119507, + "learning_rate": 0.00019994404685179195, + "loss": 1.0985, + "step": 259 + }, + { + "epoch": 0.046296296296296294, + "grad_norm": 0.3752151429653168, + "learning_rate": 0.00019994357768918333, + "loss": 0.9209, + "step": 260 + }, + { + "epoch": 0.046474358974358976, + "grad_norm": 0.43030866980552673, + "learning_rate": 0.00019994310656838796, + "loss": 0.9921, + "step": 261 + }, + { + "epoch": 0.04665242165242165, + "grad_norm": 0.4402460753917694, + "learning_rate": 0.00019994263348941502, + "loss": 1.1051, + "step": 262 + }, + { + "epoch": 0.04683048433048433, + "grad_norm": 0.43012720346450806, + "learning_rate": 0.0001999421584522738, + "loss": 1.1839, + "step": 263 + }, + { + "epoch": 0.04700854700854701, + "grad_norm": 0.4195305407047272, + "learning_rate": 0.0001999416814569736, + "loss": 1.1749, + "step": 264 + }, + { + "epoch": 0.04718660968660968, + "grad_norm": 0.45623287558555603, + "learning_rate": 0.00019994120250352372, + "loss": 1.2433, + "step": 265 + }, + { + "epoch": 0.047364672364672365, + "grad_norm": 0.4736156761646271, + "learning_rate": 0.00019994072159193363, + "loss": 1.2882, + "step": 266 + }, + { + "epoch": 0.04754273504273504, + "grad_norm": 0.36698561906814575, + "learning_rate": 0.0001999402387222127, + "loss": 1.1486, + "step": 267 + }, + { + "epoch": 0.04772079772079772, + "grad_norm": 0.3854144215583801, + "learning_rate": 0.00019993975389437038, + "loss": 0.8115, + "step": 268 + }, + { + "epoch": 0.0478988603988604, + "grad_norm": 0.41512808203697205, + "learning_rate": 0.0001999392671084162, + "loss": 1.0959, + "step": 269 + }, + { + "epoch": 0.04807692307692308, + "grad_norm": 0.3869563341140747, + "learning_rate": 0.0001999387783643597, + "loss": 1.087, + "step": 270 + }, + { + "epoch": 0.048254985754985755, + "grad_norm": 0.4649744927883148, + "learning_rate": 0.00019993828766221044, + "loss": 1.0011, + "step": 271 + }, + { + "epoch": 0.04843304843304843, + "grad_norm": 0.40331923961639404, + "learning_rate": 0.00019993779500197803, + "loss": 1.1463, + "step": 272 + }, + { + "epoch": 0.04861111111111111, + "grad_norm": 0.3826279938220978, + "learning_rate": 0.0001999373003836721, + "loss": 1.1491, + "step": 273 + }, + { + "epoch": 0.04878917378917379, + "grad_norm": 0.3967166543006897, + "learning_rate": 0.00019993680380730243, + "loss": 1.1462, + "step": 274 + }, + { + "epoch": 0.04896723646723647, + "grad_norm": 0.4298507869243622, + "learning_rate": 0.00019993630527287865, + "loss": 1.2471, + "step": 275 + }, + { + "epoch": 0.049145299145299144, + "grad_norm": 0.41486215591430664, + "learning_rate": 0.0001999358047804106, + "loss": 1.287, + "step": 276 + }, + { + "epoch": 0.049323361823361826, + "grad_norm": 0.3914124369621277, + "learning_rate": 0.00019993530232990803, + "loss": 1.0935, + "step": 277 + }, + { + "epoch": 0.0495014245014245, + "grad_norm": 0.39888378977775574, + "learning_rate": 0.00019993479792138082, + "loss": 1.2347, + "step": 278 + }, + { + "epoch": 0.049679487179487176, + "grad_norm": 0.3911665678024292, + "learning_rate": 0.00019993429155483884, + "loss": 1.0917, + "step": 279 + }, + { + "epoch": 0.04985754985754986, + "grad_norm": 0.42871445417404175, + "learning_rate": 0.00019993378323029197, + "loss": 1.0277, + "step": 280 + }, + { + "epoch": 0.050035612535612534, + "grad_norm": 0.35397860407829285, + "learning_rate": 0.00019993327294775027, + "loss": 0.9549, + "step": 281 + }, + { + "epoch": 0.050213675213675216, + "grad_norm": 0.4528059959411621, + "learning_rate": 0.00019993276070722364, + "loss": 1.2338, + "step": 282 + }, + { + "epoch": 0.05039173789173789, + "grad_norm": 0.354735791683197, + "learning_rate": 0.00019993224650872218, + "loss": 1.1892, + "step": 283 + }, + { + "epoch": 0.05056980056980057, + "grad_norm": 0.44407567381858826, + "learning_rate": 0.00019993173035225592, + "loss": 1.1621, + "step": 284 + }, + { + "epoch": 0.05074786324786325, + "grad_norm": 0.4177244305610657, + "learning_rate": 0.000199931212237835, + "loss": 1.1184, + "step": 285 + }, + { + "epoch": 0.05092592592592592, + "grad_norm": 0.5627759695053101, + "learning_rate": 0.0001999306921654696, + "loss": 1.0755, + "step": 286 + }, + { + "epoch": 0.051103988603988605, + "grad_norm": 0.46767523884773254, + "learning_rate": 0.00019993017013516986, + "loss": 1.2654, + "step": 287 + }, + { + "epoch": 0.05128205128205128, + "grad_norm": 0.4163128733634949, + "learning_rate": 0.000199929646146946, + "loss": 1.1307, + "step": 288 + }, + { + "epoch": 0.05146011396011396, + "grad_norm": 0.36954161524772644, + "learning_rate": 0.00019992912020080832, + "loss": 0.8274, + "step": 289 + }, + { + "epoch": 0.05163817663817664, + "grad_norm": 0.4770594835281372, + "learning_rate": 0.00019992859229676712, + "loss": 1.2235, + "step": 290 + }, + { + "epoch": 0.05181623931623932, + "grad_norm": 0.4174608290195465, + "learning_rate": 0.00019992806243483274, + "loss": 1.2893, + "step": 291 + }, + { + "epoch": 0.051994301994301995, + "grad_norm": 0.3794898986816406, + "learning_rate": 0.00019992753061501555, + "loss": 1.104, + "step": 292 + }, + { + "epoch": 0.05217236467236467, + "grad_norm": 0.3912592828273773, + "learning_rate": 0.000199926996837326, + "loss": 1.0043, + "step": 293 + }, + { + "epoch": 0.05235042735042735, + "grad_norm": 0.39641159772872925, + "learning_rate": 0.00019992646110177448, + "loss": 1.083, + "step": 294 + }, + { + "epoch": 0.05252849002849003, + "grad_norm": 0.3518857955932617, + "learning_rate": 0.00019992592340837157, + "loss": 0.9275, + "step": 295 + }, + { + "epoch": 0.05270655270655271, + "grad_norm": 0.3955721855163574, + "learning_rate": 0.00019992538375712777, + "loss": 1.0153, + "step": 296 + }, + { + "epoch": 0.052884615384615384, + "grad_norm": 0.3837333023548126, + "learning_rate": 0.00019992484214805364, + "loss": 1.1664, + "step": 297 + }, + { + "epoch": 0.053062678062678066, + "grad_norm": 0.39400920271873474, + "learning_rate": 0.0001999242985811598, + "loss": 1.0532, + "step": 298 + }, + { + "epoch": 0.05324074074074074, + "grad_norm": 0.39258649945259094, + "learning_rate": 0.00019992375305645692, + "loss": 1.0081, + "step": 299 + }, + { + "epoch": 0.053418803418803416, + "grad_norm": 0.49768248200416565, + "learning_rate": 0.00019992320557395566, + "loss": 1.2553, + "step": 300 + }, + { + "epoch": 0.0535968660968661, + "grad_norm": 0.364776074886322, + "learning_rate": 0.00019992265613366677, + "loss": 1.0582, + "step": 301 + }, + { + "epoch": 0.053774928774928774, + "grad_norm": 0.47317907214164734, + "learning_rate": 0.00019992210473560097, + "loss": 1.3114, + "step": 302 + }, + { + "epoch": 0.053952991452991456, + "grad_norm": 0.3706119656562805, + "learning_rate": 0.00019992155137976917, + "loss": 0.9554, + "step": 303 + }, + { + "epoch": 0.05413105413105413, + "grad_norm": 0.42809563875198364, + "learning_rate": 0.0001999209960661821, + "loss": 1.306, + "step": 304 + }, + { + "epoch": 0.054309116809116806, + "grad_norm": 0.4514487385749817, + "learning_rate": 0.00019992043879485066, + "loss": 1.0147, + "step": 305 + }, + { + "epoch": 0.05448717948717949, + "grad_norm": 0.36672836542129517, + "learning_rate": 0.0001999198795657858, + "loss": 1.1392, + "step": 306 + }, + { + "epoch": 0.05466524216524216, + "grad_norm": 0.4206554889678955, + "learning_rate": 0.00019991931837899847, + "loss": 1.2405, + "step": 307 + }, + { + "epoch": 0.054843304843304845, + "grad_norm": 0.46168261766433716, + "learning_rate": 0.00019991875523449966, + "loss": 1.2707, + "step": 308 + }, + { + "epoch": 0.05502136752136752, + "grad_norm": 0.39503365755081177, + "learning_rate": 0.00019991819013230039, + "loss": 1.0776, + "step": 309 + }, + { + "epoch": 0.0551994301994302, + "grad_norm": 0.35244834423065186, + "learning_rate": 0.00019991762307241178, + "loss": 1.0864, + "step": 310 + }, + { + "epoch": 0.05537749287749288, + "grad_norm": 0.3865319490432739, + "learning_rate": 0.0001999170540548449, + "loss": 1.3659, + "step": 311 + }, + { + "epoch": 0.05555555555555555, + "grad_norm": 0.3666876554489136, + "learning_rate": 0.0001999164830796109, + "loss": 0.9884, + "step": 312 + }, + { + "epoch": 0.055733618233618235, + "grad_norm": 0.4278281629085541, + "learning_rate": 0.00019991591014672096, + "loss": 1.1522, + "step": 313 + }, + { + "epoch": 0.05591168091168091, + "grad_norm": 0.4172627031803131, + "learning_rate": 0.0001999153352561863, + "loss": 1.2527, + "step": 314 + }, + { + "epoch": 0.05608974358974359, + "grad_norm": 0.38872212171554565, + "learning_rate": 0.00019991475840801823, + "loss": 1.2985, + "step": 315 + }, + { + "epoch": 0.05626780626780627, + "grad_norm": 0.4160458445549011, + "learning_rate": 0.00019991417960222804, + "loss": 1.1347, + "step": 316 + }, + { + "epoch": 0.05644586894586895, + "grad_norm": 0.5169723033905029, + "learning_rate": 0.00019991359883882705, + "loss": 1.0819, + "step": 317 + }, + { + "epoch": 0.056623931623931624, + "grad_norm": 0.42306259274482727, + "learning_rate": 0.0001999130161178266, + "loss": 1.3139, + "step": 318 + }, + { + "epoch": 0.0568019943019943, + "grad_norm": 0.41975873708724976, + "learning_rate": 0.00019991243143923816, + "loss": 1.2277, + "step": 319 + }, + { + "epoch": 0.05698005698005698, + "grad_norm": 0.3873472511768341, + "learning_rate": 0.00019991184480307324, + "loss": 1.156, + "step": 320 + }, + { + "epoch": 0.057158119658119656, + "grad_norm": 0.43656104803085327, + "learning_rate": 0.0001999112562093432, + "loss": 1.2344, + "step": 321 + }, + { + "epoch": 0.05733618233618234, + "grad_norm": 0.3738791048526764, + "learning_rate": 0.00019991066565805968, + "loss": 0.9573, + "step": 322 + }, + { + "epoch": 0.05751424501424501, + "grad_norm": 0.3838156461715698, + "learning_rate": 0.00019991007314923418, + "loss": 0.9274, + "step": 323 + }, + { + "epoch": 0.057692307692307696, + "grad_norm": 0.4564770758152008, + "learning_rate": 0.00019990947868287837, + "loss": 1.0756, + "step": 324 + }, + { + "epoch": 0.05787037037037037, + "grad_norm": 0.4560079872608185, + "learning_rate": 0.00019990888225900386, + "loss": 1.1508, + "step": 325 + }, + { + "epoch": 0.058048433048433046, + "grad_norm": 0.44356057047843933, + "learning_rate": 0.00019990828387762236, + "loss": 1.2323, + "step": 326 + }, + { + "epoch": 0.05822649572649573, + "grad_norm": 0.46390119194984436, + "learning_rate": 0.00019990768353874553, + "loss": 1.0031, + "step": 327 + }, + { + "epoch": 0.0584045584045584, + "grad_norm": 0.4502357244491577, + "learning_rate": 0.00019990708124238525, + "loss": 1.3454, + "step": 328 + }, + { + "epoch": 0.058582621082621085, + "grad_norm": 0.3979945182800293, + "learning_rate": 0.0001999064769885532, + "loss": 1.2833, + "step": 329 + }, + { + "epoch": 0.05876068376068376, + "grad_norm": 0.3899286687374115, + "learning_rate": 0.00019990587077726128, + "loss": 1.0175, + "step": 330 + }, + { + "epoch": 0.05893874643874644, + "grad_norm": 0.41422948241233826, + "learning_rate": 0.00019990526260852139, + "loss": 1.1151, + "step": 331 + }, + { + "epoch": 0.05911680911680912, + "grad_norm": 0.4266608953475952, + "learning_rate": 0.0001999046524823454, + "loss": 1.1119, + "step": 332 + }, + { + "epoch": 0.05929487179487179, + "grad_norm": 0.46563324332237244, + "learning_rate": 0.00019990404039874524, + "loss": 1.2358, + "step": 333 + }, + { + "epoch": 0.059472934472934474, + "grad_norm": 0.4404347240924835, + "learning_rate": 0.00019990342635773297, + "loss": 1.1748, + "step": 334 + }, + { + "epoch": 0.05965099715099715, + "grad_norm": 0.5133237838745117, + "learning_rate": 0.00019990281035932062, + "loss": 1.1649, + "step": 335 + }, + { + "epoch": 0.05982905982905983, + "grad_norm": 0.3593895435333252, + "learning_rate": 0.00019990219240352018, + "loss": 1.0318, + "step": 336 + }, + { + "epoch": 0.06000712250712251, + "grad_norm": 0.40554583072662354, + "learning_rate": 0.00019990157249034384, + "loss": 1.1202, + "step": 337 + }, + { + "epoch": 0.06018518518518518, + "grad_norm": 0.3770706057548523, + "learning_rate": 0.00019990095061980372, + "loss": 0.9908, + "step": 338 + }, + { + "epoch": 0.060363247863247864, + "grad_norm": 0.39676955342292786, + "learning_rate": 0.000199900326791912, + "loss": 0.8176, + "step": 339 + }, + { + "epoch": 0.06054131054131054, + "grad_norm": 0.41448578238487244, + "learning_rate": 0.00019989970100668086, + "loss": 1.2877, + "step": 340 + }, + { + "epoch": 0.06071937321937322, + "grad_norm": 0.4200015068054199, + "learning_rate": 0.00019989907326412265, + "loss": 1.2293, + "step": 341 + }, + { + "epoch": 0.060897435897435896, + "grad_norm": 0.47350621223449707, + "learning_rate": 0.0001998984435642496, + "loss": 1.2331, + "step": 342 + }, + { + "epoch": 0.06107549857549858, + "grad_norm": 0.47050634026527405, + "learning_rate": 0.00019989781190707406, + "loss": 0.8888, + "step": 343 + }, + { + "epoch": 0.06125356125356125, + "grad_norm": 0.4994896948337555, + "learning_rate": 0.00019989717829260842, + "loss": 1.0921, + "step": 344 + }, + { + "epoch": 0.06143162393162393, + "grad_norm": 0.36340200901031494, + "learning_rate": 0.0001998965427208651, + "loss": 0.9777, + "step": 345 + }, + { + "epoch": 0.06160968660968661, + "grad_norm": 0.3538152873516083, + "learning_rate": 0.00019989590519185654, + "loss": 1.0055, + "step": 346 + }, + { + "epoch": 0.061787749287749286, + "grad_norm": 0.5388944149017334, + "learning_rate": 0.00019989526570559526, + "loss": 1.1001, + "step": 347 + }, + { + "epoch": 0.06196581196581197, + "grad_norm": 0.4411574602127075, + "learning_rate": 0.00019989462426209373, + "loss": 1.0038, + "step": 348 + }, + { + "epoch": 0.06214387464387464, + "grad_norm": 0.3930876851081848, + "learning_rate": 0.00019989398086136455, + "loss": 1.1534, + "step": 349 + }, + { + "epoch": 0.062321937321937325, + "grad_norm": 0.47357070446014404, + "learning_rate": 0.00019989333550342033, + "loss": 1.2687, + "step": 350 + }, + { + "epoch": 0.0625, + "grad_norm": 0.40302303433418274, + "learning_rate": 0.00019989268818827372, + "loss": 1.1894, + "step": 351 + }, + { + "epoch": 0.06267806267806268, + "grad_norm": 0.4470510184764862, + "learning_rate": 0.00019989203891593738, + "loss": 1.2207, + "step": 352 + }, + { + "epoch": 0.06285612535612535, + "grad_norm": 0.42235100269317627, + "learning_rate": 0.00019989138768642406, + "loss": 1.2086, + "step": 353 + }, + { + "epoch": 0.06303418803418803, + "grad_norm": 0.38305309414863586, + "learning_rate": 0.0001998907344997465, + "loss": 1.0473, + "step": 354 + }, + { + "epoch": 0.06321225071225071, + "grad_norm": 0.3893027901649475, + "learning_rate": 0.0001998900793559175, + "loss": 1.1746, + "step": 355 + }, + { + "epoch": 0.0633903133903134, + "grad_norm": 0.41206735372543335, + "learning_rate": 0.0001998894222549499, + "loss": 1.188, + "step": 356 + }, + { + "epoch": 0.06356837606837606, + "grad_norm": 0.3700513243675232, + "learning_rate": 0.00019988876319685658, + "loss": 0.9862, + "step": 357 + }, + { + "epoch": 0.06374643874643875, + "grad_norm": 0.3708794116973877, + "learning_rate": 0.0001998881021816504, + "loss": 1.2003, + "step": 358 + }, + { + "epoch": 0.06392450142450143, + "grad_norm": 0.4058014154434204, + "learning_rate": 0.00019988743920934442, + "loss": 1.2311, + "step": 359 + }, + { + "epoch": 0.0641025641025641, + "grad_norm": 0.39134132862091064, + "learning_rate": 0.00019988677427995155, + "loss": 1.001, + "step": 360 + }, + { + "epoch": 0.06428062678062678, + "grad_norm": 0.3853437602519989, + "learning_rate": 0.00019988610739348484, + "loss": 1.0725, + "step": 361 + }, + { + "epoch": 0.06445868945868946, + "grad_norm": 0.47114330530166626, + "learning_rate": 0.00019988543854995735, + "loss": 1.2196, + "step": 362 + }, + { + "epoch": 0.06463675213675214, + "grad_norm": 0.40465688705444336, + "learning_rate": 0.00019988476774938216, + "loss": 1.1869, + "step": 363 + }, + { + "epoch": 0.06481481481481481, + "grad_norm": 0.40301886200904846, + "learning_rate": 0.00019988409499177245, + "loss": 1.1765, + "step": 364 + }, + { + "epoch": 0.0649928774928775, + "grad_norm": 0.43443185091018677, + "learning_rate": 0.0001998834202771414, + "loss": 1.2022, + "step": 365 + }, + { + "epoch": 0.06517094017094018, + "grad_norm": 0.4712986350059509, + "learning_rate": 0.00019988274360550217, + "loss": 1.156, + "step": 366 + }, + { + "epoch": 0.06534900284900284, + "grad_norm": 0.4524450898170471, + "learning_rate": 0.00019988206497686815, + "loss": 1.2917, + "step": 367 + }, + { + "epoch": 0.06552706552706553, + "grad_norm": 0.40302205085754395, + "learning_rate": 0.0001998813843912525, + "loss": 0.9993, + "step": 368 + }, + { + "epoch": 0.06570512820512821, + "grad_norm": 0.39435216784477234, + "learning_rate": 0.00019988070184866864, + "loss": 1.0914, + "step": 369 + }, + { + "epoch": 0.06588319088319089, + "grad_norm": 0.39267390966415405, + "learning_rate": 0.00019988001734912988, + "loss": 1.3138, + "step": 370 + }, + { + "epoch": 0.06606125356125356, + "grad_norm": 0.38351675868034363, + "learning_rate": 0.00019987933089264968, + "loss": 1.0997, + "step": 371 + }, + { + "epoch": 0.06623931623931624, + "grad_norm": 0.3294839859008789, + "learning_rate": 0.00019987864247924145, + "loss": 0.9656, + "step": 372 + }, + { + "epoch": 0.06641737891737892, + "grad_norm": 0.45333364605903625, + "learning_rate": 0.00019987795210891872, + "loss": 1.095, + "step": 373 + }, + { + "epoch": 0.06659544159544159, + "grad_norm": 0.4362282454967499, + "learning_rate": 0.00019987725978169501, + "loss": 1.2103, + "step": 374 + }, + { + "epoch": 0.06677350427350427, + "grad_norm": 0.41314780712127686, + "learning_rate": 0.00019987656549758385, + "loss": 1.2115, + "step": 375 + }, + { + "epoch": 0.06695156695156695, + "grad_norm": 0.4230864644050598, + "learning_rate": 0.00019987586925659888, + "loss": 1.17, + "step": 376 + }, + { + "epoch": 0.06712962962962964, + "grad_norm": 0.4703855812549591, + "learning_rate": 0.00019987517105875372, + "loss": 1.367, + "step": 377 + }, + { + "epoch": 0.0673076923076923, + "grad_norm": 0.4671297073364258, + "learning_rate": 0.00019987447090406206, + "loss": 1.2543, + "step": 378 + }, + { + "epoch": 0.06748575498575499, + "grad_norm": 0.43746981024742126, + "learning_rate": 0.0001998737687925376, + "loss": 1.214, + "step": 379 + }, + { + "epoch": 0.06766381766381767, + "grad_norm": 0.40889596939086914, + "learning_rate": 0.00019987306472419412, + "loss": 1.0496, + "step": 380 + }, + { + "epoch": 0.06784188034188034, + "grad_norm": 0.3677358627319336, + "learning_rate": 0.0001998723586990454, + "loss": 1.1242, + "step": 381 + }, + { + "epoch": 0.06801994301994302, + "grad_norm": 0.3892628848552704, + "learning_rate": 0.00019987165071710527, + "loss": 1.0246, + "step": 382 + }, + { + "epoch": 0.0681980056980057, + "grad_norm": 0.4281293749809265, + "learning_rate": 0.00019987094077838764, + "loss": 1.2817, + "step": 383 + }, + { + "epoch": 0.06837606837606838, + "grad_norm": 0.45030340552330017, + "learning_rate": 0.00019987022888290636, + "loss": 1.159, + "step": 384 + }, + { + "epoch": 0.06855413105413105, + "grad_norm": 0.6327905058860779, + "learning_rate": 0.00019986951503067545, + "loss": 0.9577, + "step": 385 + }, + { + "epoch": 0.06873219373219373, + "grad_norm": 0.40339627861976624, + "learning_rate": 0.0001998687992217088, + "loss": 1.138, + "step": 386 + }, + { + "epoch": 0.06891025641025642, + "grad_norm": 0.4018291234970093, + "learning_rate": 0.00019986808145602052, + "loss": 0.9109, + "step": 387 + }, + { + "epoch": 0.06908831908831908, + "grad_norm": 0.41566264629364014, + "learning_rate": 0.00019986736173362464, + "loss": 1.1516, + "step": 388 + }, + { + "epoch": 0.06926638176638177, + "grad_norm": 0.3569067418575287, + "learning_rate": 0.00019986664005453527, + "loss": 1.2329, + "step": 389 + }, + { + "epoch": 0.06944444444444445, + "grad_norm": 0.3959648907184601, + "learning_rate": 0.0001998659164187665, + "loss": 1.1041, + "step": 390 + }, + { + "epoch": 0.06962250712250712, + "grad_norm": 0.42853206396102905, + "learning_rate": 0.00019986519082633257, + "loss": 1.0859, + "step": 391 + }, + { + "epoch": 0.0698005698005698, + "grad_norm": 0.42005518078804016, + "learning_rate": 0.0001998644632772477, + "loss": 1.2017, + "step": 392 + }, + { + "epoch": 0.06997863247863248, + "grad_norm": 0.4296947419643402, + "learning_rate": 0.00019986373377152612, + "loss": 1.1464, + "step": 393 + }, + { + "epoch": 0.07015669515669516, + "grad_norm": 0.394747793674469, + "learning_rate": 0.0001998630023091821, + "loss": 1.0316, + "step": 394 + }, + { + "epoch": 0.07033475783475783, + "grad_norm": 0.3779357969760895, + "learning_rate": 0.00019986226889023002, + "loss": 1.1081, + "step": 395 + }, + { + "epoch": 0.07051282051282051, + "grad_norm": 0.4271804690361023, + "learning_rate": 0.00019986153351468424, + "loss": 0.985, + "step": 396 + }, + { + "epoch": 0.0706908831908832, + "grad_norm": 0.49412235617637634, + "learning_rate": 0.00019986079618255912, + "loss": 1.2606, + "step": 397 + }, + { + "epoch": 0.07086894586894586, + "grad_norm": 0.43657439947128296, + "learning_rate": 0.00019986005689386915, + "loss": 1.2266, + "step": 398 + }, + { + "epoch": 0.07104700854700854, + "grad_norm": 0.4060729444026947, + "learning_rate": 0.0001998593156486288, + "loss": 1.1787, + "step": 399 + }, + { + "epoch": 0.07122507122507123, + "grad_norm": 0.387046217918396, + "learning_rate": 0.00019985857244685264, + "loss": 0.9411, + "step": 400 + }, + { + "epoch": 0.07140313390313391, + "grad_norm": 0.4243999123573303, + "learning_rate": 0.00019985782728855516, + "loss": 1.2024, + "step": 401 + }, + { + "epoch": 0.07158119658119658, + "grad_norm": 0.43113812804222107, + "learning_rate": 0.000199857080173751, + "loss": 1.1246, + "step": 402 + }, + { + "epoch": 0.07175925925925926, + "grad_norm": 0.4653271436691284, + "learning_rate": 0.0001998563311024548, + "loss": 1.2343, + "step": 403 + }, + { + "epoch": 0.07193732193732194, + "grad_norm": 0.43260812759399414, + "learning_rate": 0.0001998555800746812, + "loss": 0.9543, + "step": 404 + }, + { + "epoch": 0.07211538461538461, + "grad_norm": 0.4635484516620636, + "learning_rate": 0.00019985482709044495, + "loss": 1.1091, + "step": 405 + }, + { + "epoch": 0.07229344729344729, + "grad_norm": 0.38362643122673035, + "learning_rate": 0.00019985407214976076, + "loss": 1.2584, + "step": 406 + }, + { + "epoch": 0.07247150997150997, + "grad_norm": 0.4068310558795929, + "learning_rate": 0.00019985331525264351, + "loss": 1.1944, + "step": 407 + }, + { + "epoch": 0.07264957264957266, + "grad_norm": 0.43909943103790283, + "learning_rate": 0.00019985255639910795, + "loss": 1.3748, + "step": 408 + }, + { + "epoch": 0.07282763532763532, + "grad_norm": 0.48674601316452026, + "learning_rate": 0.000199851795589169, + "loss": 1.2684, + "step": 409 + }, + { + "epoch": 0.073005698005698, + "grad_norm": 0.4218580722808838, + "learning_rate": 0.0001998510328228415, + "loss": 1.168, + "step": 410 + }, + { + "epoch": 0.07318376068376069, + "grad_norm": 0.4688236117362976, + "learning_rate": 0.00019985026810014046, + "loss": 1.3088, + "step": 411 + }, + { + "epoch": 0.07336182336182336, + "grad_norm": 0.3863612711429596, + "learning_rate": 0.00019984950142108083, + "loss": 1.0261, + "step": 412 + }, + { + "epoch": 0.07353988603988604, + "grad_norm": 0.4177640378475189, + "learning_rate": 0.00019984873278567765, + "loss": 1.1985, + "step": 413 + }, + { + "epoch": 0.07371794871794872, + "grad_norm": 0.4645586311817169, + "learning_rate": 0.00019984796219394592, + "loss": 1.2463, + "step": 414 + }, + { + "epoch": 0.0738960113960114, + "grad_norm": 0.5051766633987427, + "learning_rate": 0.00019984718964590083, + "loss": 1.3031, + "step": 415 + }, + { + "epoch": 0.07407407407407407, + "grad_norm": 0.4200040400028229, + "learning_rate": 0.0001998464151415575, + "loss": 1.0842, + "step": 416 + }, + { + "epoch": 0.07425213675213675, + "grad_norm": 0.34211036562919617, + "learning_rate": 0.000199845638680931, + "loss": 0.9659, + "step": 417 + }, + { + "epoch": 0.07443019943019943, + "grad_norm": 0.3553323447704315, + "learning_rate": 0.00019984486026403668, + "loss": 1.0102, + "step": 418 + }, + { + "epoch": 0.0746082621082621, + "grad_norm": 0.4967300295829773, + "learning_rate": 0.00019984407989088974, + "loss": 1.3125, + "step": 419 + }, + { + "epoch": 0.07478632478632478, + "grad_norm": 0.41649797558784485, + "learning_rate": 0.00019984329756150544, + "loss": 1.3092, + "step": 420 + }, + { + "epoch": 0.07496438746438747, + "grad_norm": 0.43825802206993103, + "learning_rate": 0.00019984251327589912, + "loss": 1.3678, + "step": 421 + }, + { + "epoch": 0.07514245014245015, + "grad_norm": 0.363394170999527, + "learning_rate": 0.00019984172703408617, + "loss": 1.305, + "step": 422 + }, + { + "epoch": 0.07532051282051282, + "grad_norm": 0.411563903093338, + "learning_rate": 0.000199840938836082, + "loss": 1.4248, + "step": 423 + }, + { + "epoch": 0.0754985754985755, + "grad_norm": 0.40548190474510193, + "learning_rate": 0.000199840148681902, + "loss": 1.1081, + "step": 424 + }, + { + "epoch": 0.07567663817663818, + "grad_norm": 0.3781099021434784, + "learning_rate": 0.00019983935657156171, + "loss": 1.185, + "step": 425 + }, + { + "epoch": 0.07585470085470085, + "grad_norm": 0.46597573161125183, + "learning_rate": 0.00019983856250507662, + "loss": 1.119, + "step": 426 + }, + { + "epoch": 0.07603276353276353, + "grad_norm": 0.3988197147846222, + "learning_rate": 0.00019983776648246232, + "loss": 1.206, + "step": 427 + }, + { + "epoch": 0.07621082621082621, + "grad_norm": 0.41210901737213135, + "learning_rate": 0.00019983696850373433, + "loss": 1.1843, + "step": 428 + }, + { + "epoch": 0.0763888888888889, + "grad_norm": 0.41870948672294617, + "learning_rate": 0.00019983616856890837, + "loss": 1.2248, + "step": 429 + }, + { + "epoch": 0.07656695156695156, + "grad_norm": 0.4320056140422821, + "learning_rate": 0.00019983536667800007, + "loss": 0.9743, + "step": 430 + }, + { + "epoch": 0.07674501424501425, + "grad_norm": 0.48455503582954407, + "learning_rate": 0.00019983456283102517, + "loss": 1.0438, + "step": 431 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 0.38712427020072937, + "learning_rate": 0.00019983375702799935, + "loss": 1.2041, + "step": 432 + }, + { + "epoch": 0.0771011396011396, + "grad_norm": 0.3578857481479645, + "learning_rate": 0.0001998329492689385, + "loss": 1.1623, + "step": 433 + }, + { + "epoch": 0.07727920227920228, + "grad_norm": 0.43065932393074036, + "learning_rate": 0.00019983213955385834, + "loss": 1.3033, + "step": 434 + }, + { + "epoch": 0.07745726495726496, + "grad_norm": 0.4882095754146576, + "learning_rate": 0.00019983132788277484, + "loss": 1.1635, + "step": 435 + }, + { + "epoch": 0.07763532763532764, + "grad_norm": 0.3429015874862671, + "learning_rate": 0.00019983051425570382, + "loss": 0.7289, + "step": 436 + }, + { + "epoch": 0.07781339031339031, + "grad_norm": 0.4320310056209564, + "learning_rate": 0.00019982969867266128, + "loss": 1.3685, + "step": 437 + }, + { + "epoch": 0.07799145299145299, + "grad_norm": 0.39891982078552246, + "learning_rate": 0.00019982888113366314, + "loss": 1.0444, + "step": 438 + }, + { + "epoch": 0.07816951566951567, + "grad_norm": 0.3675695061683655, + "learning_rate": 0.00019982806163872547, + "loss": 1.0527, + "step": 439 + }, + { + "epoch": 0.07834757834757834, + "grad_norm": 0.42824694514274597, + "learning_rate": 0.0001998272401878643, + "loss": 1.166, + "step": 440 + }, + { + "epoch": 0.07852564102564102, + "grad_norm": 0.3721694350242615, + "learning_rate": 0.00019982641678109575, + "loss": 1.1328, + "step": 441 + }, + { + "epoch": 0.0787037037037037, + "grad_norm": 0.33899208903312683, + "learning_rate": 0.00019982559141843592, + "loss": 1.016, + "step": 442 + }, + { + "epoch": 0.07888176638176639, + "grad_norm": 0.4029340147972107, + "learning_rate": 0.000199824764099901, + "loss": 1.0076, + "step": 443 + }, + { + "epoch": 0.07905982905982906, + "grad_norm": 0.4169132113456726, + "learning_rate": 0.0001998239348255072, + "loss": 1.208, + "step": 444 + }, + { + "epoch": 0.07923789173789174, + "grad_norm": 0.3865824043750763, + "learning_rate": 0.00019982310359527075, + "loss": 1.067, + "step": 445 + }, + { + "epoch": 0.07941595441595442, + "grad_norm": 0.4218919277191162, + "learning_rate": 0.00019982227040920796, + "loss": 1.195, + "step": 446 + }, + { + "epoch": 0.07959401709401709, + "grad_norm": 0.40504586696624756, + "learning_rate": 0.00019982143526733512, + "loss": 1.0188, + "step": 447 + }, + { + "epoch": 0.07977207977207977, + "grad_norm": 0.38330578804016113, + "learning_rate": 0.00019982059816966863, + "loss": 1.0484, + "step": 448 + }, + { + "epoch": 0.07995014245014245, + "grad_norm": 0.43731689453125, + "learning_rate": 0.00019981975911622488, + "loss": 1.074, + "step": 449 + }, + { + "epoch": 0.08012820512820513, + "grad_norm": 0.40858447551727295, + "learning_rate": 0.00019981891810702033, + "loss": 1.0008, + "step": 450 + }, + { + "epoch": 0.0803062678062678, + "grad_norm": 0.4031754732131958, + "learning_rate": 0.00019981807514207143, + "loss": 1.2179, + "step": 451 + }, + { + "epoch": 0.08048433048433049, + "grad_norm": 0.41920867562294006, + "learning_rate": 0.00019981723022139466, + "loss": 1.1406, + "step": 452 + }, + { + "epoch": 0.08066239316239317, + "grad_norm": 0.40305474400520325, + "learning_rate": 0.00019981638334500668, + "loss": 1.098, + "step": 453 + }, + { + "epoch": 0.08084045584045584, + "grad_norm": 0.4564182460308075, + "learning_rate": 0.00019981553451292396, + "loss": 1.419, + "step": 454 + }, + { + "epoch": 0.08101851851851852, + "grad_norm": 0.3832945227622986, + "learning_rate": 0.00019981468372516322, + "loss": 1.0919, + "step": 455 + }, + { + "epoch": 0.0811965811965812, + "grad_norm": 0.43062624335289, + "learning_rate": 0.0001998138309817411, + "loss": 1.0458, + "step": 456 + }, + { + "epoch": 0.08137464387464387, + "grad_norm": 0.3871173560619354, + "learning_rate": 0.0001998129762826743, + "loss": 1.1391, + "step": 457 + }, + { + "epoch": 0.08155270655270655, + "grad_norm": 0.43423157930374146, + "learning_rate": 0.0001998121196279796, + "loss": 1.1132, + "step": 458 + }, + { + "epoch": 0.08173076923076923, + "grad_norm": 0.4341012239456177, + "learning_rate": 0.00019981126101767372, + "loss": 1.113, + "step": 459 + }, + { + "epoch": 0.08190883190883191, + "grad_norm": 0.36748576164245605, + "learning_rate": 0.00019981040045177352, + "loss": 0.8108, + "step": 460 + }, + { + "epoch": 0.08208689458689458, + "grad_norm": 0.43133220076560974, + "learning_rate": 0.00019980953793029586, + "loss": 1.1861, + "step": 461 + }, + { + "epoch": 0.08226495726495726, + "grad_norm": 0.37204909324645996, + "learning_rate": 0.00019980867345325767, + "loss": 0.9222, + "step": 462 + }, + { + "epoch": 0.08244301994301995, + "grad_norm": 0.43370047211647034, + "learning_rate": 0.00019980780702067582, + "loss": 1.2984, + "step": 463 + }, + { + "epoch": 0.08262108262108261, + "grad_norm": 0.4991510808467865, + "learning_rate": 0.00019980693863256736, + "loss": 1.2222, + "step": 464 + }, + { + "epoch": 0.0827991452991453, + "grad_norm": 0.44318175315856934, + "learning_rate": 0.00019980606828894927, + "loss": 1.2262, + "step": 465 + }, + { + "epoch": 0.08297720797720798, + "grad_norm": 0.380231648683548, + "learning_rate": 0.0001998051959898386, + "loss": 1.0274, + "step": 466 + }, + { + "epoch": 0.08315527065527066, + "grad_norm": 0.39519667625427246, + "learning_rate": 0.0001998043217352524, + "loss": 1.2499, + "step": 467 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.457499235868454, + "learning_rate": 0.0001998034455252079, + "loss": 1.0751, + "step": 468 + }, + { + "epoch": 0.08351139601139601, + "grad_norm": 0.368522584438324, + "learning_rate": 0.00019980256735972215, + "loss": 1.0776, + "step": 469 + }, + { + "epoch": 0.08368945868945869, + "grad_norm": 0.3768427073955536, + "learning_rate": 0.00019980168723881243, + "loss": 1.2198, + "step": 470 + }, + { + "epoch": 0.08386752136752136, + "grad_norm": 0.37045565247535706, + "learning_rate": 0.000199800805162496, + "loss": 1.1816, + "step": 471 + }, + { + "epoch": 0.08404558404558404, + "grad_norm": 0.4219281077384949, + "learning_rate": 0.0001997999211307901, + "loss": 1.0515, + "step": 472 + }, + { + "epoch": 0.08422364672364673, + "grad_norm": 0.3815271258354187, + "learning_rate": 0.00019979903514371207, + "loss": 1.1709, + "step": 473 + }, + { + "epoch": 0.08440170940170941, + "grad_norm": 0.4566493630409241, + "learning_rate": 0.00019979814720127924, + "loss": 1.3063, + "step": 474 + }, + { + "epoch": 0.08457977207977208, + "grad_norm": 0.4043879806995392, + "learning_rate": 0.000199797257303509, + "loss": 1.0549, + "step": 475 + }, + { + "epoch": 0.08475783475783476, + "grad_norm": 0.3897830545902252, + "learning_rate": 0.00019979636545041886, + "loss": 1.1483, + "step": 476 + }, + { + "epoch": 0.08493589743589744, + "grad_norm": 0.36097025871276855, + "learning_rate": 0.00019979547164202622, + "loss": 1.1196, + "step": 477 + }, + { + "epoch": 0.08511396011396011, + "grad_norm": 0.3766986131668091, + "learning_rate": 0.00019979457587834863, + "loss": 1.0131, + "step": 478 + }, + { + "epoch": 0.08529202279202279, + "grad_norm": 0.39460286498069763, + "learning_rate": 0.00019979367815940364, + "loss": 1.1729, + "step": 479 + }, + { + "epoch": 0.08547008547008547, + "grad_norm": 0.4137469232082367, + "learning_rate": 0.00019979277848520885, + "loss": 1.2569, + "step": 480 + }, + { + "epoch": 0.08564814814814815, + "grad_norm": 0.464688777923584, + "learning_rate": 0.00019979187685578183, + "loss": 1.2064, + "step": 481 + }, + { + "epoch": 0.08582621082621082, + "grad_norm": 0.4245518147945404, + "learning_rate": 0.0001997909732711403, + "loss": 0.9812, + "step": 482 + }, + { + "epoch": 0.0860042735042735, + "grad_norm": 0.43368837237358093, + "learning_rate": 0.00019979006773130197, + "loss": 1.2822, + "step": 483 + }, + { + "epoch": 0.08618233618233619, + "grad_norm": 0.4232824444770813, + "learning_rate": 0.00019978916023628452, + "loss": 1.1446, + "step": 484 + }, + { + "epoch": 0.08636039886039885, + "grad_norm": 0.4183506369590759, + "learning_rate": 0.00019978825078610578, + "loss": 1.2605, + "step": 485 + }, + { + "epoch": 0.08653846153846154, + "grad_norm": 0.4391268491744995, + "learning_rate": 0.00019978733938078356, + "loss": 1.2165, + "step": 486 + }, + { + "epoch": 0.08671652421652422, + "grad_norm": 0.4139612317085266, + "learning_rate": 0.0001997864260203357, + "loss": 0.9389, + "step": 487 + }, + { + "epoch": 0.0868945868945869, + "grad_norm": 0.4058656096458435, + "learning_rate": 0.00019978551070478013, + "loss": 1.0652, + "step": 488 + }, + { + "epoch": 0.08707264957264957, + "grad_norm": 0.42333099246025085, + "learning_rate": 0.00019978459343413473, + "loss": 1.119, + "step": 489 + }, + { + "epoch": 0.08725071225071225, + "grad_norm": 0.4573031961917877, + "learning_rate": 0.00019978367420841754, + "loss": 1.1546, + "step": 490 + }, + { + "epoch": 0.08742877492877493, + "grad_norm": 0.4161617159843445, + "learning_rate": 0.00019978275302764655, + "loss": 1.0836, + "step": 491 + }, + { + "epoch": 0.0876068376068376, + "grad_norm": 0.422145277261734, + "learning_rate": 0.00019978182989183977, + "loss": 1.1908, + "step": 492 + }, + { + "epoch": 0.08778490028490028, + "grad_norm": 0.4588126838207245, + "learning_rate": 0.00019978090480101532, + "loss": 1.1758, + "step": 493 + }, + { + "epoch": 0.08796296296296297, + "grad_norm": 0.4425722062587738, + "learning_rate": 0.00019977997775519132, + "loss": 1.088, + "step": 494 + }, + { + "epoch": 0.08814102564102565, + "grad_norm": 0.37860307097435, + "learning_rate": 0.00019977904875438594, + "loss": 1.1532, + "step": 495 + }, + { + "epoch": 0.08831908831908832, + "grad_norm": 0.40435823798179626, + "learning_rate": 0.00019977811779861733, + "loss": 1.1271, + "step": 496 + }, + { + "epoch": 0.088497150997151, + "grad_norm": 0.42578884959220886, + "learning_rate": 0.0001997771848879038, + "loss": 0.9889, + "step": 497 + }, + { + "epoch": 0.08867521367521368, + "grad_norm": 0.3439478874206543, + "learning_rate": 0.00019977625002226361, + "loss": 1.1273, + "step": 498 + }, + { + "epoch": 0.08885327635327635, + "grad_norm": 0.362341970205307, + "learning_rate": 0.00019977531320171504, + "loss": 1.0214, + "step": 499 + }, + { + "epoch": 0.08903133903133903, + "grad_norm": 0.4305768609046936, + "learning_rate": 0.0001997743744262765, + "loss": 1.2648, + "step": 500 + }, + { + "epoch": 0.08920940170940171, + "grad_norm": 0.35900023579597473, + "learning_rate": 0.00019977343369596636, + "loss": 1.0274, + "step": 501 + }, + { + "epoch": 0.0893874643874644, + "grad_norm": 0.4950818717479706, + "learning_rate": 0.00019977249101080306, + "loss": 1.1483, + "step": 502 + }, + { + "epoch": 0.08956552706552706, + "grad_norm": 0.3800346553325653, + "learning_rate": 0.00019977154637080503, + "loss": 1.0636, + "step": 503 + }, + { + "epoch": 0.08974358974358974, + "grad_norm": 0.46202352643013, + "learning_rate": 0.0001997705997759908, + "loss": 1.1544, + "step": 504 + }, + { + "epoch": 0.08992165242165243, + "grad_norm": 0.36818403005599976, + "learning_rate": 0.00019976965122637895, + "loss": 0.9824, + "step": 505 + }, + { + "epoch": 0.0900997150997151, + "grad_norm": 0.40248095989227295, + "learning_rate": 0.00019976870072198805, + "loss": 1.1002, + "step": 506 + }, + { + "epoch": 0.09027777777777778, + "grad_norm": 0.3841850459575653, + "learning_rate": 0.00019976774826283667, + "loss": 1.2433, + "step": 507 + }, + { + "epoch": 0.09045584045584046, + "grad_norm": 0.46892330050468445, + "learning_rate": 0.0001997667938489435, + "loss": 1.3194, + "step": 508 + }, + { + "epoch": 0.09063390313390314, + "grad_norm": 0.39059561491012573, + "learning_rate": 0.0001997658374803273, + "loss": 1.1778, + "step": 509 + }, + { + "epoch": 0.09081196581196581, + "grad_norm": 0.3793235421180725, + "learning_rate": 0.00019976487915700672, + "loss": 1.0659, + "step": 510 + }, + { + "epoch": 0.09099002849002849, + "grad_norm": 0.39067742228507996, + "learning_rate": 0.00019976391887900058, + "loss": 1.107, + "step": 511 + }, + { + "epoch": 0.09116809116809117, + "grad_norm": 0.40121713280677795, + "learning_rate": 0.00019976295664632772, + "loss": 1.102, + "step": 512 + }, + { + "epoch": 0.09134615384615384, + "grad_norm": 0.49830010533332825, + "learning_rate": 0.00019976199245900697, + "loss": 1.1701, + "step": 513 + }, + { + "epoch": 0.09152421652421652, + "grad_norm": 0.4536968171596527, + "learning_rate": 0.0001997610263170572, + "loss": 1.1067, + "step": 514 + }, + { + "epoch": 0.0917022792022792, + "grad_norm": 0.3832971453666687, + "learning_rate": 0.00019976005822049735, + "loss": 1.0991, + "step": 515 + }, + { + "epoch": 0.09188034188034189, + "grad_norm": 0.4093509614467621, + "learning_rate": 0.0001997590881693464, + "loss": 1.0565, + "step": 516 + }, + { + "epoch": 0.09205840455840456, + "grad_norm": 0.46073687076568604, + "learning_rate": 0.0001997581161636233, + "loss": 1.0057, + "step": 517 + }, + { + "epoch": 0.09223646723646724, + "grad_norm": 0.5001922845840454, + "learning_rate": 0.0001997571422033472, + "loss": 1.2639, + "step": 518 + }, + { + "epoch": 0.09241452991452992, + "grad_norm": 0.4620618224143982, + "learning_rate": 0.00019975616628853713, + "loss": 1.0966, + "step": 519 + }, + { + "epoch": 0.09259259259259259, + "grad_norm": 0.3788183927536011, + "learning_rate": 0.0001997551884192122, + "loss": 0.9783, + "step": 520 + }, + { + "epoch": 0.09277065527065527, + "grad_norm": 0.45589539408683777, + "learning_rate": 0.00019975420859539154, + "loss": 1.2194, + "step": 521 + }, + { + "epoch": 0.09294871794871795, + "grad_norm": 0.40747523307800293, + "learning_rate": 0.00019975322681709443, + "loss": 1.0349, + "step": 522 + }, + { + "epoch": 0.09312678062678063, + "grad_norm": 0.5045142769813538, + "learning_rate": 0.00019975224308434002, + "loss": 1.1373, + "step": 523 + }, + { + "epoch": 0.0933048433048433, + "grad_norm": 0.40352702140808105, + "learning_rate": 0.00019975125739714767, + "loss": 1.1236, + "step": 524 + }, + { + "epoch": 0.09348290598290598, + "grad_norm": 0.4301735758781433, + "learning_rate": 0.0001997502697555366, + "loss": 1.2932, + "step": 525 + }, + { + "epoch": 0.09366096866096867, + "grad_norm": 0.36800238490104675, + "learning_rate": 0.00019974928015952624, + "loss": 1.0734, + "step": 526 + }, + { + "epoch": 0.09383903133903133, + "grad_norm": 0.4027230143547058, + "learning_rate": 0.00019974828860913594, + "loss": 1.2776, + "step": 527 + }, + { + "epoch": 0.09401709401709402, + "grad_norm": 0.42497140169143677, + "learning_rate": 0.0001997472951043851, + "loss": 1.248, + "step": 528 + }, + { + "epoch": 0.0941951566951567, + "grad_norm": 0.3888593018054962, + "learning_rate": 0.00019974629964529325, + "loss": 1.0231, + "step": 529 + }, + { + "epoch": 0.09437321937321937, + "grad_norm": 0.3761361241340637, + "learning_rate": 0.00019974530223187986, + "loss": 1.0216, + "step": 530 + }, + { + "epoch": 0.09455128205128205, + "grad_norm": 0.42192980647087097, + "learning_rate": 0.00019974430286416448, + "loss": 1.0731, + "step": 531 + }, + { + "epoch": 0.09472934472934473, + "grad_norm": 0.44244512915611267, + "learning_rate": 0.00019974330154216667, + "loss": 1.2793, + "step": 532 + }, + { + "epoch": 0.09490740740740741, + "grad_norm": 0.378252774477005, + "learning_rate": 0.0001997422982659061, + "loss": 1.0462, + "step": 533 + }, + { + "epoch": 0.09508547008547008, + "grad_norm": 0.45589110255241394, + "learning_rate": 0.00019974129303540236, + "loss": 1.1884, + "step": 534 + }, + { + "epoch": 0.09526353276353276, + "grad_norm": 0.33930808305740356, + "learning_rate": 0.0001997402858506752, + "loss": 0.8381, + "step": 535 + }, + { + "epoch": 0.09544159544159544, + "grad_norm": 0.45408427715301514, + "learning_rate": 0.0001997392767117443, + "loss": 1.2379, + "step": 536 + }, + { + "epoch": 0.09561965811965811, + "grad_norm": 0.44125741720199585, + "learning_rate": 0.0001997382656186295, + "loss": 1.1941, + "step": 537 + }, + { + "epoch": 0.0957977207977208, + "grad_norm": 0.4075697660446167, + "learning_rate": 0.00019973725257135054, + "loss": 1.0142, + "step": 538 + }, + { + "epoch": 0.09597578347578348, + "grad_norm": 0.4258415102958679, + "learning_rate": 0.00019973623756992733, + "loss": 1.0447, + "step": 539 + }, + { + "epoch": 0.09615384615384616, + "grad_norm": 0.2738485038280487, + "learning_rate": 0.0001997352206143797, + "loss": 0.5521, + "step": 540 + }, + { + "epoch": 0.09633190883190883, + "grad_norm": 0.38815587759017944, + "learning_rate": 0.00019973420170472762, + "loss": 1.1052, + "step": 541 + }, + { + "epoch": 0.09650997150997151, + "grad_norm": 0.3909834027290344, + "learning_rate": 0.00019973318084099106, + "loss": 1.0494, + "step": 542 + }, + { + "epoch": 0.09668803418803419, + "grad_norm": 0.4517597258090973, + "learning_rate": 0.00019973215802318996, + "loss": 1.0611, + "step": 543 + }, + { + "epoch": 0.09686609686609686, + "grad_norm": 0.48659002780914307, + "learning_rate": 0.00019973113325134442, + "loss": 0.9967, + "step": 544 + }, + { + "epoch": 0.09704415954415954, + "grad_norm": 0.4039791524410248, + "learning_rate": 0.0001997301065254745, + "loss": 1.251, + "step": 545 + }, + { + "epoch": 0.09722222222222222, + "grad_norm": 0.3985383212566376, + "learning_rate": 0.0001997290778456003, + "loss": 1.2263, + "step": 546 + }, + { + "epoch": 0.0974002849002849, + "grad_norm": 0.4540637731552124, + "learning_rate": 0.00019972804721174199, + "loss": 1.2084, + "step": 547 + }, + { + "epoch": 0.09757834757834757, + "grad_norm": 0.36867982149124146, + "learning_rate": 0.00019972701462391977, + "loss": 0.9704, + "step": 548 + }, + { + "epoch": 0.09775641025641026, + "grad_norm": 0.40199780464172363, + "learning_rate": 0.00019972598008215385, + "loss": 1.1121, + "step": 549 + }, + { + "epoch": 0.09793447293447294, + "grad_norm": 0.42728984355926514, + "learning_rate": 0.00019972494358646455, + "loss": 1.1606, + "step": 550 + }, + { + "epoch": 0.0981125356125356, + "grad_norm": 0.4212374687194824, + "learning_rate": 0.0001997239051368721, + "loss": 1.3093, + "step": 551 + }, + { + "epoch": 0.09829059829059829, + "grad_norm": 0.3972226083278656, + "learning_rate": 0.0001997228647333969, + "loss": 1.1218, + "step": 552 + }, + { + "epoch": 0.09846866096866097, + "grad_norm": 0.43649932742118835, + "learning_rate": 0.00019972182237605935, + "loss": 1.2532, + "step": 553 + }, + { + "epoch": 0.09864672364672365, + "grad_norm": 0.3812280595302582, + "learning_rate": 0.0001997207780648798, + "loss": 1.0409, + "step": 554 + }, + { + "epoch": 0.09882478632478632, + "grad_norm": 0.41684821248054504, + "learning_rate": 0.00019971973179987878, + "loss": 0.9569, + "step": 555 + }, + { + "epoch": 0.099002849002849, + "grad_norm": 0.38081470131874084, + "learning_rate": 0.00019971868358107674, + "loss": 1.1615, + "step": 556 + }, + { + "epoch": 0.09918091168091168, + "grad_norm": 0.3702073097229004, + "learning_rate": 0.0001997176334084943, + "loss": 1.3907, + "step": 557 + }, + { + "epoch": 0.09935897435897435, + "grad_norm": 0.3625728189945221, + "learning_rate": 0.00019971658128215193, + "loss": 1.1897, + "step": 558 + }, + { + "epoch": 0.09953703703703703, + "grad_norm": 0.3815405070781708, + "learning_rate": 0.0001997155272020703, + "loss": 1.1473, + "step": 559 + }, + { + "epoch": 0.09971509971509972, + "grad_norm": 0.48664286732673645, + "learning_rate": 0.00019971447116827004, + "loss": 1.2462, + "step": 560 + }, + { + "epoch": 0.0998931623931624, + "grad_norm": 0.3708696663379669, + "learning_rate": 0.0001997134131807719, + "loss": 1.0979, + "step": 561 + }, + { + "epoch": 0.10007122507122507, + "grad_norm": 0.44511324167251587, + "learning_rate": 0.00019971235323959654, + "loss": 1.2313, + "step": 562 + }, + { + "epoch": 0.10024928774928775, + "grad_norm": 0.3687448799610138, + "learning_rate": 0.00019971129134476473, + "loss": 1.1526, + "step": 563 + }, + { + "epoch": 0.10042735042735043, + "grad_norm": 0.4506866931915283, + "learning_rate": 0.00019971022749629735, + "loss": 1.0003, + "step": 564 + }, + { + "epoch": 0.1006054131054131, + "grad_norm": 0.41910406947135925, + "learning_rate": 0.00019970916169421515, + "loss": 1.013, + "step": 565 + }, + { + "epoch": 0.10078347578347578, + "grad_norm": 0.39728936553001404, + "learning_rate": 0.0001997080939385391, + "loss": 1.0501, + "step": 566 + }, + { + "epoch": 0.10096153846153846, + "grad_norm": 0.41415902972221375, + "learning_rate": 0.00019970702422929005, + "loss": 1.0791, + "step": 567 + }, + { + "epoch": 0.10113960113960115, + "grad_norm": 0.45630788803100586, + "learning_rate": 0.00019970595256648896, + "loss": 1.2884, + "step": 568 + }, + { + "epoch": 0.10131766381766381, + "grad_norm": 0.4371698796749115, + "learning_rate": 0.00019970487895015686, + "loss": 1.0684, + "step": 569 + }, + { + "epoch": 0.1014957264957265, + "grad_norm": 0.4350591003894806, + "learning_rate": 0.00019970380338031477, + "loss": 1.2415, + "step": 570 + }, + { + "epoch": 0.10167378917378918, + "grad_norm": 0.4232708215713501, + "learning_rate": 0.00019970272585698382, + "loss": 1.2656, + "step": 571 + }, + { + "epoch": 0.10185185185185185, + "grad_norm": 0.3917689919471741, + "learning_rate": 0.00019970164638018502, + "loss": 1.0178, + "step": 572 + }, + { + "epoch": 0.10202991452991453, + "grad_norm": 0.4262804388999939, + "learning_rate": 0.0001997005649499396, + "loss": 1.1805, + "step": 573 + }, + { + "epoch": 0.10220797720797721, + "grad_norm": 0.5217884182929993, + "learning_rate": 0.0001996994815662687, + "loss": 1.2392, + "step": 574 + }, + { + "epoch": 0.10238603988603989, + "grad_norm": 0.4273875057697296, + "learning_rate": 0.00019969839622919358, + "loss": 1.0844, + "step": 575 + }, + { + "epoch": 0.10256410256410256, + "grad_norm": 0.41588085889816284, + "learning_rate": 0.00019969730893873547, + "loss": 1.2437, + "step": 576 + }, + { + "epoch": 0.10274216524216524, + "grad_norm": 0.41617709398269653, + "learning_rate": 0.0001996962196949157, + "loss": 0.9519, + "step": 577 + }, + { + "epoch": 0.10292022792022792, + "grad_norm": 0.4832979142665863, + "learning_rate": 0.00019969512849775565, + "loss": 1.1889, + "step": 578 + }, + { + "epoch": 0.10309829059829059, + "grad_norm": 0.3936060965061188, + "learning_rate": 0.0001996940353472766, + "loss": 0.9888, + "step": 579 + }, + { + "epoch": 0.10327635327635327, + "grad_norm": 0.4147680997848511, + "learning_rate": 0.00019969294024350004, + "loss": 1.0733, + "step": 580 + }, + { + "epoch": 0.10345441595441596, + "grad_norm": 0.37791356444358826, + "learning_rate": 0.00019969184318644742, + "loss": 1.212, + "step": 581 + }, + { + "epoch": 0.10363247863247864, + "grad_norm": 0.44297221302986145, + "learning_rate": 0.00019969074417614023, + "loss": 1.0535, + "step": 582 + }, + { + "epoch": 0.10381054131054131, + "grad_norm": 0.4032835066318512, + "learning_rate": 0.0001996896432126, + "loss": 1.1869, + "step": 583 + }, + { + "epoch": 0.10398860398860399, + "grad_norm": 0.49271953105926514, + "learning_rate": 0.00019968854029584827, + "loss": 1.1661, + "step": 584 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 0.362699031829834, + "learning_rate": 0.0001996874354259067, + "loss": 0.868, + "step": 585 + }, + { + "epoch": 0.10434472934472934, + "grad_norm": 0.401795357465744, + "learning_rate": 0.0001996863286027969, + "loss": 1.1045, + "step": 586 + }, + { + "epoch": 0.10452279202279202, + "grad_norm": 0.45380479097366333, + "learning_rate": 0.00019968521982654058, + "loss": 0.8503, + "step": 587 + }, + { + "epoch": 0.1047008547008547, + "grad_norm": 0.49759066104888916, + "learning_rate": 0.00019968410909715947, + "loss": 1.4073, + "step": 588 + }, + { + "epoch": 0.10487891737891739, + "grad_norm": 0.4421198070049286, + "learning_rate": 0.0001996829964146753, + "loss": 1.1512, + "step": 589 + }, + { + "epoch": 0.10505698005698005, + "grad_norm": 0.46675658226013184, + "learning_rate": 0.00019968188177910988, + "loss": 1.0132, + "step": 590 + }, + { + "epoch": 0.10523504273504274, + "grad_norm": 0.5710657238960266, + "learning_rate": 0.00019968076519048507, + "loss": 1.267, + "step": 591 + }, + { + "epoch": 0.10541310541310542, + "grad_norm": 0.4655563235282898, + "learning_rate": 0.00019967964664882276, + "loss": 1.1204, + "step": 592 + }, + { + "epoch": 0.10559116809116809, + "grad_norm": 0.3895256519317627, + "learning_rate": 0.00019967852615414478, + "loss": 1.0814, + "step": 593 + }, + { + "epoch": 0.10576923076923077, + "grad_norm": 0.424216091632843, + "learning_rate": 0.00019967740370647322, + "loss": 1.1663, + "step": 594 + }, + { + "epoch": 0.10594729344729345, + "grad_norm": 0.3978985846042633, + "learning_rate": 0.00019967627930582996, + "loss": 0.909, + "step": 595 + }, + { + "epoch": 0.10612535612535613, + "grad_norm": 0.47064995765686035, + "learning_rate": 0.00019967515295223705, + "loss": 1.2351, + "step": 596 + }, + { + "epoch": 0.1063034188034188, + "grad_norm": 0.42449644207954407, + "learning_rate": 0.0001996740246457166, + "loss": 0.9739, + "step": 597 + }, + { + "epoch": 0.10648148148148148, + "grad_norm": 0.39033401012420654, + "learning_rate": 0.00019967289438629066, + "loss": 1.0933, + "step": 598 + }, + { + "epoch": 0.10665954415954416, + "grad_norm": 0.4398612678050995, + "learning_rate": 0.00019967176217398143, + "loss": 1.2479, + "step": 599 + }, + { + "epoch": 0.10683760683760683, + "grad_norm": 0.3946632742881775, + "learning_rate": 0.00019967062800881107, + "loss": 1.0417, + "step": 600 + }, + { + "epoch": 0.10701566951566951, + "grad_norm": 0.5083445906639099, + "learning_rate": 0.0001996694918908018, + "loss": 1.1109, + "step": 601 + }, + { + "epoch": 0.1071937321937322, + "grad_norm": 0.477724552154541, + "learning_rate": 0.00019966835381997585, + "loss": 1.2891, + "step": 602 + }, + { + "epoch": 0.10737179487179487, + "grad_norm": 0.4110167920589447, + "learning_rate": 0.0001996672137963556, + "loss": 1.0555, + "step": 603 + }, + { + "epoch": 0.10754985754985755, + "grad_norm": 0.44078320264816284, + "learning_rate": 0.00019966607181996334, + "loss": 0.9188, + "step": 604 + }, + { + "epoch": 0.10772792022792023, + "grad_norm": 0.41251105070114136, + "learning_rate": 0.00019966492789082142, + "loss": 1.2592, + "step": 605 + }, + { + "epoch": 0.10790598290598291, + "grad_norm": 0.37701505422592163, + "learning_rate": 0.00019966378200895227, + "loss": 1.0233, + "step": 606 + }, + { + "epoch": 0.10808404558404558, + "grad_norm": 0.44624966382980347, + "learning_rate": 0.00019966263417437835, + "loss": 1.2273, + "step": 607 + }, + { + "epoch": 0.10826210826210826, + "grad_norm": 0.3618549108505249, + "learning_rate": 0.00019966148438712214, + "loss": 0.9101, + "step": 608 + }, + { + "epoch": 0.10844017094017094, + "grad_norm": 0.384574294090271, + "learning_rate": 0.00019966033264720616, + "loss": 1.1769, + "step": 609 + }, + { + "epoch": 0.10861823361823361, + "grad_norm": 0.50872403383255, + "learning_rate": 0.000199659178954653, + "loss": 1.1213, + "step": 610 + }, + { + "epoch": 0.1087962962962963, + "grad_norm": 0.39736685156822205, + "learning_rate": 0.00019965802330948527, + "loss": 1.275, + "step": 611 + }, + { + "epoch": 0.10897435897435898, + "grad_norm": 0.484660267829895, + "learning_rate": 0.00019965686571172557, + "loss": 1.1671, + "step": 612 + }, + { + "epoch": 0.10915242165242166, + "grad_norm": 0.41420218348503113, + "learning_rate": 0.0001996557061613966, + "loss": 0.9541, + "step": 613 + }, + { + "epoch": 0.10933048433048433, + "grad_norm": 0.4057196080684662, + "learning_rate": 0.00019965454465852112, + "loss": 1.0145, + "step": 614 + }, + { + "epoch": 0.10950854700854701, + "grad_norm": 0.4559510052204132, + "learning_rate": 0.00019965338120312182, + "loss": 1.0889, + "step": 615 + }, + { + "epoch": 0.10968660968660969, + "grad_norm": 0.40960055589675903, + "learning_rate": 0.00019965221579522154, + "loss": 1.1447, + "step": 616 + }, + { + "epoch": 0.10986467236467236, + "grad_norm": 0.4701732099056244, + "learning_rate": 0.0001996510484348431, + "loss": 1.2871, + "step": 617 + }, + { + "epoch": 0.11004273504273504, + "grad_norm": 0.38420796394348145, + "learning_rate": 0.0001996498791220094, + "loss": 1.058, + "step": 618 + }, + { + "epoch": 0.11022079772079772, + "grad_norm": 0.4014730453491211, + "learning_rate": 0.00019964870785674327, + "loss": 1.023, + "step": 619 + }, + { + "epoch": 0.1103988603988604, + "grad_norm": 0.38846179842948914, + "learning_rate": 0.00019964753463906773, + "loss": 0.9834, + "step": 620 + }, + { + "epoch": 0.11057692307692307, + "grad_norm": 0.5120236277580261, + "learning_rate": 0.00019964635946900577, + "loss": 1.2347, + "step": 621 + }, + { + "epoch": 0.11075498575498575, + "grad_norm": 0.40483301877975464, + "learning_rate": 0.00019964518234658038, + "loss": 1.131, + "step": 622 + }, + { + "epoch": 0.11093304843304844, + "grad_norm": 0.445782870054245, + "learning_rate": 0.00019964400327181464, + "loss": 0.9349, + "step": 623 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 0.490460604429245, + "learning_rate": 0.00019964282224473165, + "loss": 1.0257, + "step": 624 + }, + { + "epoch": 0.11128917378917379, + "grad_norm": 0.37585243582725525, + "learning_rate": 0.00019964163926535454, + "loss": 0.9724, + "step": 625 + }, + { + "epoch": 0.11146723646723647, + "grad_norm": 0.4160473346710205, + "learning_rate": 0.00019964045433370651, + "loss": 0.874, + "step": 626 + }, + { + "epoch": 0.11164529914529915, + "grad_norm": 0.442425012588501, + "learning_rate": 0.00019963926744981074, + "loss": 1.064, + "step": 627 + }, + { + "epoch": 0.11182336182336182, + "grad_norm": 0.4451471269130707, + "learning_rate": 0.00019963807861369054, + "loss": 1.2343, + "step": 628 + }, + { + "epoch": 0.1120014245014245, + "grad_norm": 0.5018183588981628, + "learning_rate": 0.00019963688782536913, + "loss": 1.1226, + "step": 629 + }, + { + "epoch": 0.11217948717948718, + "grad_norm": 0.43723925948143005, + "learning_rate": 0.0001996356950848699, + "loss": 1.0178, + "step": 630 + }, + { + "epoch": 0.11235754985754985, + "grad_norm": 0.4794611930847168, + "learning_rate": 0.0001996345003922162, + "loss": 0.9695, + "step": 631 + }, + { + "epoch": 0.11253561253561253, + "grad_norm": 0.5021790266036987, + "learning_rate": 0.00019963330374743143, + "loss": 1.1748, + "step": 632 + }, + { + "epoch": 0.11271367521367522, + "grad_norm": 0.47228625416755676, + "learning_rate": 0.00019963210515053906, + "loss": 1.2138, + "step": 633 + }, + { + "epoch": 0.1128917378917379, + "grad_norm": 0.4261155128479004, + "learning_rate": 0.00019963090460156256, + "loss": 0.9428, + "step": 634 + }, + { + "epoch": 0.11306980056980057, + "grad_norm": 0.3279525339603424, + "learning_rate": 0.00019962970210052542, + "loss": 0.7803, + "step": 635 + }, + { + "epoch": 0.11324786324786325, + "grad_norm": 0.5106086730957031, + "learning_rate": 0.00019962849764745125, + "loss": 1.113, + "step": 636 + }, + { + "epoch": 0.11342592592592593, + "grad_norm": 0.38272222876548767, + "learning_rate": 0.00019962729124236363, + "loss": 0.896, + "step": 637 + }, + { + "epoch": 0.1136039886039886, + "grad_norm": 0.39532098174095154, + "learning_rate": 0.0001996260828852862, + "loss": 0.9308, + "step": 638 + }, + { + "epoch": 0.11378205128205128, + "grad_norm": 0.44947221875190735, + "learning_rate": 0.00019962487257624262, + "loss": 1.207, + "step": 639 + }, + { + "epoch": 0.11396011396011396, + "grad_norm": 0.40684598684310913, + "learning_rate": 0.00019962366031525664, + "loss": 1.11, + "step": 640 + }, + { + "epoch": 0.11413817663817664, + "grad_norm": 0.4296625852584839, + "learning_rate": 0.00019962244610235194, + "loss": 1.2784, + "step": 641 + }, + { + "epoch": 0.11431623931623931, + "grad_norm": 0.4560794532299042, + "learning_rate": 0.0001996212299375524, + "loss": 1.1191, + "step": 642 + }, + { + "epoch": 0.114494301994302, + "grad_norm": 0.40246087312698364, + "learning_rate": 0.00019962001182088177, + "loss": 1.1401, + "step": 643 + }, + { + "epoch": 0.11467236467236468, + "grad_norm": 0.3938910663127899, + "learning_rate": 0.000199618791752364, + "loss": 1.0959, + "step": 644 + }, + { + "epoch": 0.11485042735042734, + "grad_norm": 0.4123380184173584, + "learning_rate": 0.00019961756973202287, + "loss": 1.2824, + "step": 645 + }, + { + "epoch": 0.11502849002849003, + "grad_norm": 0.41085442900657654, + "learning_rate": 0.00019961634575988243, + "loss": 1.1137, + "step": 646 + }, + { + "epoch": 0.11520655270655271, + "grad_norm": 0.38276201486587524, + "learning_rate": 0.0001996151198359667, + "loss": 1.0747, + "step": 647 + }, + { + "epoch": 0.11538461538461539, + "grad_norm": 0.49269407987594604, + "learning_rate": 0.00019961389196029953, + "loss": 1.1731, + "step": 648 + }, + { + "epoch": 0.11556267806267806, + "grad_norm": 0.5152469277381897, + "learning_rate": 0.00019961266213290512, + "loss": 1.3574, + "step": 649 + }, + { + "epoch": 0.11574074074074074, + "grad_norm": 0.4835714101791382, + "learning_rate": 0.0001996114303538075, + "loss": 1.2859, + "step": 650 + }, + { + "epoch": 0.11591880341880342, + "grad_norm": 0.4284524917602539, + "learning_rate": 0.00019961019662303087, + "loss": 1.1103, + "step": 651 + }, + { + "epoch": 0.11609686609686609, + "grad_norm": 0.3933276832103729, + "learning_rate": 0.00019960896094059933, + "loss": 1.2647, + "step": 652 + }, + { + "epoch": 0.11627492877492877, + "grad_norm": 0.33749741315841675, + "learning_rate": 0.00019960772330653712, + "loss": 0.819, + "step": 653 + }, + { + "epoch": 0.11645299145299146, + "grad_norm": 0.48122069239616394, + "learning_rate": 0.00019960648372086852, + "loss": 1.2781, + "step": 654 + }, + { + "epoch": 0.11663105413105414, + "grad_norm": 0.4681607186794281, + "learning_rate": 0.00019960524218361775, + "loss": 0.9723, + "step": 655 + }, + { + "epoch": 0.1168091168091168, + "grad_norm": 0.3974960148334503, + "learning_rate": 0.0001996039986948092, + "loss": 1.0302, + "step": 656 + }, + { + "epoch": 0.11698717948717949, + "grad_norm": 0.43180662393569946, + "learning_rate": 0.0001996027532544672, + "loss": 1.3265, + "step": 657 + }, + { + "epoch": 0.11716524216524217, + "grad_norm": 0.4481917917728424, + "learning_rate": 0.00019960150586261613, + "loss": 1.136, + "step": 658 + }, + { + "epoch": 0.11734330484330484, + "grad_norm": 0.43428945541381836, + "learning_rate": 0.00019960025651928045, + "loss": 1.2412, + "step": 659 + }, + { + "epoch": 0.11752136752136752, + "grad_norm": 0.36211395263671875, + "learning_rate": 0.00019959900522448467, + "loss": 0.9563, + "step": 660 + }, + { + "epoch": 0.1176994301994302, + "grad_norm": 0.43585848808288574, + "learning_rate": 0.0001995977519782533, + "loss": 1.1677, + "step": 661 + }, + { + "epoch": 0.11787749287749288, + "grad_norm": 0.4232597351074219, + "learning_rate": 0.00019959649678061086, + "loss": 1.1187, + "step": 662 + }, + { + "epoch": 0.11805555555555555, + "grad_norm": 0.3304753303527832, + "learning_rate": 0.00019959523963158194, + "loss": 0.8473, + "step": 663 + }, + { + "epoch": 0.11823361823361823, + "grad_norm": 0.37600061297416687, + "learning_rate": 0.0001995939805311912, + "loss": 1.1227, + "step": 664 + }, + { + "epoch": 0.11841168091168092, + "grad_norm": 0.33417847752571106, + "learning_rate": 0.0001995927194794633, + "loss": 1.0315, + "step": 665 + }, + { + "epoch": 0.11858974358974358, + "grad_norm": 0.46799129247665405, + "learning_rate": 0.00019959145647642298, + "loss": 1.135, + "step": 666 + }, + { + "epoch": 0.11876780626780627, + "grad_norm": 0.4141576886177063, + "learning_rate": 0.0001995901915220949, + "loss": 1.0956, + "step": 667 + }, + { + "epoch": 0.11894586894586895, + "grad_norm": 0.3824596405029297, + "learning_rate": 0.0001995889246165039, + "loss": 1.1782, + "step": 668 + }, + { + "epoch": 0.11912393162393162, + "grad_norm": 0.4087786376476288, + "learning_rate": 0.00019958765575967484, + "loss": 0.9704, + "step": 669 + }, + { + "epoch": 0.1193019943019943, + "grad_norm": 0.5161317586898804, + "learning_rate": 0.00019958638495163252, + "loss": 1.2207, + "step": 670 + }, + { + "epoch": 0.11948005698005698, + "grad_norm": 0.4782274067401886, + "learning_rate": 0.0001995851121924019, + "loss": 1.1257, + "step": 671 + }, + { + "epoch": 0.11965811965811966, + "grad_norm": 0.40617331862449646, + "learning_rate": 0.00019958383748200782, + "loss": 1.1153, + "step": 672 + }, + { + "epoch": 0.11983618233618233, + "grad_norm": 0.40149980783462524, + "learning_rate": 0.00019958256082047533, + "loss": 0.9785, + "step": 673 + }, + { + "epoch": 0.12001424501424501, + "grad_norm": 0.4378886818885803, + "learning_rate": 0.00019958128220782942, + "loss": 1.1355, + "step": 674 + }, + { + "epoch": 0.1201923076923077, + "grad_norm": 0.4449596703052521, + "learning_rate": 0.0001995800016440952, + "loss": 1.0325, + "step": 675 + }, + { + "epoch": 0.12037037037037036, + "grad_norm": 0.4268079698085785, + "learning_rate": 0.00019957871912929765, + "loss": 1.1901, + "step": 676 + }, + { + "epoch": 0.12054843304843305, + "grad_norm": 0.4250091016292572, + "learning_rate": 0.00019957743466346198, + "loss": 1.0084, + "step": 677 + }, + { + "epoch": 0.12072649572649573, + "grad_norm": 0.40724286437034607, + "learning_rate": 0.0001995761482466133, + "loss": 1.0866, + "step": 678 + }, + { + "epoch": 0.12090455840455841, + "grad_norm": 0.42478349804878235, + "learning_rate": 0.00019957485987877688, + "loss": 1.1909, + "step": 679 + }, + { + "epoch": 0.12108262108262108, + "grad_norm": 0.371362566947937, + "learning_rate": 0.0001995735695599779, + "loss": 1.083, + "step": 680 + }, + { + "epoch": 0.12126068376068376, + "grad_norm": 0.4715283513069153, + "learning_rate": 0.0001995722772902417, + "loss": 1.2942, + "step": 681 + }, + { + "epoch": 0.12143874643874644, + "grad_norm": 0.3611983060836792, + "learning_rate": 0.00019957098306959355, + "loss": 0.9878, + "step": 682 + }, + { + "epoch": 0.12161680911680911, + "grad_norm": 0.4764883816242218, + "learning_rate": 0.00019956968689805883, + "loss": 1.0082, + "step": 683 + }, + { + "epoch": 0.12179487179487179, + "grad_norm": 0.33170604705810547, + "learning_rate": 0.00019956838877566293, + "loss": 0.8529, + "step": 684 + }, + { + "epoch": 0.12197293447293447, + "grad_norm": 0.46896886825561523, + "learning_rate": 0.00019956708870243133, + "loss": 1.0745, + "step": 685 + }, + { + "epoch": 0.12215099715099716, + "grad_norm": 0.4120674431324005, + "learning_rate": 0.00019956578667838941, + "loss": 1.1828, + "step": 686 + }, + { + "epoch": 0.12232905982905982, + "grad_norm": 0.45671191811561584, + "learning_rate": 0.00019956448270356275, + "loss": 1.3484, + "step": 687 + }, + { + "epoch": 0.1225071225071225, + "grad_norm": 0.4023838937282562, + "learning_rate": 0.00019956317677797687, + "loss": 0.9623, + "step": 688 + }, + { + "epoch": 0.12268518518518519, + "grad_norm": 0.5205856561660767, + "learning_rate": 0.00019956186890165737, + "loss": 1.2221, + "step": 689 + }, + { + "epoch": 0.12286324786324786, + "grad_norm": 0.43956050276756287, + "learning_rate": 0.00019956055907462987, + "loss": 1.1051, + "step": 690 + }, + { + "epoch": 0.12304131054131054, + "grad_norm": 0.4341758191585541, + "learning_rate": 0.00019955924729692003, + "loss": 0.8972, + "step": 691 + }, + { + "epoch": 0.12321937321937322, + "grad_norm": 0.42025020718574524, + "learning_rate": 0.00019955793356855357, + "loss": 1.1137, + "step": 692 + }, + { + "epoch": 0.1233974358974359, + "grad_norm": 0.44375079870224, + "learning_rate": 0.0001995566178895562, + "loss": 1.2783, + "step": 693 + }, + { + "epoch": 0.12357549857549857, + "grad_norm": 0.4703320264816284, + "learning_rate": 0.00019955530025995372, + "loss": 1.1991, + "step": 694 + }, + { + "epoch": 0.12375356125356125, + "grad_norm": 0.43781620264053345, + "learning_rate": 0.00019955398067977195, + "loss": 1.2316, + "step": 695 + }, + { + "epoch": 0.12393162393162394, + "grad_norm": 0.4362877607345581, + "learning_rate": 0.0001995526591490367, + "loss": 1.1374, + "step": 696 + }, + { + "epoch": 0.1241096866096866, + "grad_norm": 0.4434499442577362, + "learning_rate": 0.00019955133566777392, + "loss": 1.1034, + "step": 697 + }, + { + "epoch": 0.12428774928774929, + "grad_norm": 0.46613508462905884, + "learning_rate": 0.00019955001023600955, + "loss": 1.2252, + "step": 698 + }, + { + "epoch": 0.12446581196581197, + "grad_norm": 0.46226736903190613, + "learning_rate": 0.00019954868285376945, + "loss": 1.0296, + "step": 699 + }, + { + "epoch": 0.12464387464387465, + "grad_norm": 0.4460904002189636, + "learning_rate": 0.00019954735352107977, + "loss": 1.0553, + "step": 700 + }, + { + "epoch": 0.12482193732193732, + "grad_norm": 0.36708924174308777, + "learning_rate": 0.00019954602223796648, + "loss": 0.9384, + "step": 701 + }, + { + "epoch": 0.125, + "grad_norm": 0.3780093491077423, + "learning_rate": 0.00019954468900445566, + "loss": 0.9062, + "step": 702 + }, + { + "epoch": 0.12517806267806267, + "grad_norm": 0.41797417402267456, + "learning_rate": 0.00019954335382057345, + "loss": 1.0344, + "step": 703 + }, + { + "epoch": 0.12535612535612536, + "grad_norm": 0.43710798025131226, + "learning_rate": 0.00019954201668634597, + "loss": 1.1324, + "step": 704 + }, + { + "epoch": 0.12553418803418803, + "grad_norm": 0.4732789695262909, + "learning_rate": 0.00019954067760179952, + "loss": 1.1419, + "step": 705 + }, + { + "epoch": 0.1257122507122507, + "grad_norm": 0.43248575925827026, + "learning_rate": 0.00019953933656696022, + "loss": 1.5112, + "step": 706 + }, + { + "epoch": 0.1258903133903134, + "grad_norm": 0.4074753522872925, + "learning_rate": 0.00019953799358185442, + "loss": 0.9751, + "step": 707 + }, + { + "epoch": 0.12606837606837606, + "grad_norm": 0.4586823880672455, + "learning_rate": 0.0001995366486465084, + "loss": 1.267, + "step": 708 + }, + { + "epoch": 0.12624643874643873, + "grad_norm": 0.4716857075691223, + "learning_rate": 0.0001995353017609485, + "loss": 1.1636, + "step": 709 + }, + { + "epoch": 0.12642450142450143, + "grad_norm": 0.5214398503303528, + "learning_rate": 0.00019953395292520115, + "loss": 1.2317, + "step": 710 + }, + { + "epoch": 0.1266025641025641, + "grad_norm": 0.42961129546165466, + "learning_rate": 0.00019953260213929276, + "loss": 1.0271, + "step": 711 + }, + { + "epoch": 0.1267806267806268, + "grad_norm": 0.4764653444290161, + "learning_rate": 0.00019953124940324979, + "loss": 1.1747, + "step": 712 + }, + { + "epoch": 0.12695868945868946, + "grad_norm": 0.4420304000377655, + "learning_rate": 0.00019952989471709874, + "loss": 0.9783, + "step": 713 + }, + { + "epoch": 0.12713675213675213, + "grad_norm": 0.44114625453948975, + "learning_rate": 0.00019952853808086616, + "loss": 1.1953, + "step": 714 + }, + { + "epoch": 0.12731481481481483, + "grad_norm": 0.501923143863678, + "learning_rate": 0.0001995271794945786, + "loss": 0.9886, + "step": 715 + }, + { + "epoch": 0.1274928774928775, + "grad_norm": 0.42266538739204407, + "learning_rate": 0.00019952581895826276, + "loss": 1.2033, + "step": 716 + }, + { + "epoch": 0.12767094017094016, + "grad_norm": 0.37770554423332214, + "learning_rate": 0.00019952445647194523, + "loss": 1.0164, + "step": 717 + }, + { + "epoch": 0.12784900284900286, + "grad_norm": 0.369266152381897, + "learning_rate": 0.00019952309203565268, + "loss": 0.9186, + "step": 718 + }, + { + "epoch": 0.12802706552706553, + "grad_norm": 0.40446221828460693, + "learning_rate": 0.00019952172564941193, + "loss": 1.1576, + "step": 719 + }, + { + "epoch": 0.1282051282051282, + "grad_norm": 0.504172146320343, + "learning_rate": 0.00019952035731324967, + "loss": 1.2695, + "step": 720 + }, + { + "epoch": 0.1283831908831909, + "grad_norm": 0.37284108996391296, + "learning_rate": 0.0001995189870271928, + "loss": 1.0288, + "step": 721 + }, + { + "epoch": 0.12856125356125356, + "grad_norm": 0.41811618208885193, + "learning_rate": 0.00019951761479126805, + "loss": 1.2241, + "step": 722 + }, + { + "epoch": 0.12873931623931623, + "grad_norm": 0.44706249237060547, + "learning_rate": 0.0001995162406055024, + "loss": 1.0831, + "step": 723 + }, + { + "epoch": 0.12891737891737892, + "grad_norm": 0.426572322845459, + "learning_rate": 0.00019951486446992273, + "loss": 1.0047, + "step": 724 + }, + { + "epoch": 0.1290954415954416, + "grad_norm": 0.4446277618408203, + "learning_rate": 0.00019951348638455602, + "loss": 1.0827, + "step": 725 + }, + { + "epoch": 0.12927350427350429, + "grad_norm": 0.3934919834136963, + "learning_rate": 0.00019951210634942926, + "loss": 0.9808, + "step": 726 + }, + { + "epoch": 0.12945156695156695, + "grad_norm": 0.4316558241844177, + "learning_rate": 0.0001995107243645695, + "loss": 1.3341, + "step": 727 + }, + { + "epoch": 0.12962962962962962, + "grad_norm": 0.43074217438697815, + "learning_rate": 0.00019950934043000382, + "loss": 1.007, + "step": 728 + }, + { + "epoch": 0.12980769230769232, + "grad_norm": 0.5212171673774719, + "learning_rate": 0.0001995079545457593, + "loss": 1.1822, + "step": 729 + }, + { + "epoch": 0.129985754985755, + "grad_norm": 0.3749600946903229, + "learning_rate": 0.00019950656671186313, + "loss": 0.9657, + "step": 730 + }, + { + "epoch": 0.13016381766381765, + "grad_norm": 0.36626043915748596, + "learning_rate": 0.00019950517692834252, + "loss": 1.1274, + "step": 731 + }, + { + "epoch": 0.13034188034188035, + "grad_norm": 0.4635467529296875, + "learning_rate": 0.00019950378519522467, + "loss": 1.2305, + "step": 732 + }, + { + "epoch": 0.13051994301994302, + "grad_norm": 0.4077455699443817, + "learning_rate": 0.00019950239151253683, + "loss": 0.9485, + "step": 733 + }, + { + "epoch": 0.1306980056980057, + "grad_norm": 0.4222758114337921, + "learning_rate": 0.0001995009958803063, + "loss": 1.0376, + "step": 734 + }, + { + "epoch": 0.13087606837606838, + "grad_norm": 0.4330402612686157, + "learning_rate": 0.0001994995982985605, + "loss": 1.1774, + "step": 735 + }, + { + "epoch": 0.13105413105413105, + "grad_norm": 0.42275673151016235, + "learning_rate": 0.00019949819876732673, + "loss": 1.1238, + "step": 736 + }, + { + "epoch": 0.13123219373219372, + "grad_norm": 0.45576968789100647, + "learning_rate": 0.00019949679728663246, + "loss": 1.0428, + "step": 737 + }, + { + "epoch": 0.13141025641025642, + "grad_norm": 0.5508752465248108, + "learning_rate": 0.00019949539385650514, + "loss": 1.3221, + "step": 738 + }, + { + "epoch": 0.13158831908831908, + "grad_norm": 0.4115872383117676, + "learning_rate": 0.00019949398847697225, + "loss": 1.0301, + "step": 739 + }, + { + "epoch": 0.13176638176638178, + "grad_norm": 0.4662442207336426, + "learning_rate": 0.00019949258114806132, + "loss": 1.3263, + "step": 740 + }, + { + "epoch": 0.13194444444444445, + "grad_norm": 0.6077266931533813, + "learning_rate": 0.00019949117186979999, + "loss": 1.0269, + "step": 741 + }, + { + "epoch": 0.13212250712250712, + "grad_norm": 0.47039318084716797, + "learning_rate": 0.00019948976064221579, + "loss": 1.3782, + "step": 742 + }, + { + "epoch": 0.1323005698005698, + "grad_norm": 0.4773450493812561, + "learning_rate": 0.0001994883474653364, + "loss": 1.289, + "step": 743 + }, + { + "epoch": 0.13247863247863248, + "grad_norm": 0.40180155634880066, + "learning_rate": 0.00019948693233918952, + "loss": 0.8691, + "step": 744 + }, + { + "epoch": 0.13265669515669515, + "grad_norm": 0.45216289162635803, + "learning_rate": 0.00019948551526380288, + "loss": 1.071, + "step": 745 + }, + { + "epoch": 0.13283475783475784, + "grad_norm": 0.4289272427558899, + "learning_rate": 0.0001994840962392042, + "loss": 1.0422, + "step": 746 + }, + { + "epoch": 0.1330128205128205, + "grad_norm": 0.4617730379104614, + "learning_rate": 0.00019948267526542134, + "loss": 1.0835, + "step": 747 + }, + { + "epoch": 0.13319088319088318, + "grad_norm": 0.42710617184638977, + "learning_rate": 0.00019948125234248208, + "loss": 1.0535, + "step": 748 + }, + { + "epoch": 0.13336894586894588, + "grad_norm": 0.43433234095573425, + "learning_rate": 0.0001994798274704144, + "loss": 0.9313, + "step": 749 + }, + { + "epoch": 0.13354700854700854, + "grad_norm": 0.46270284056663513, + "learning_rate": 0.0001994784006492461, + "loss": 1.0903, + "step": 750 + }, + { + "epoch": 0.1337250712250712, + "grad_norm": 0.5319814682006836, + "learning_rate": 0.00019947697187900517, + "loss": 1.2329, + "step": 751 + }, + { + "epoch": 0.1339031339031339, + "grad_norm": 0.3511372208595276, + "learning_rate": 0.00019947554115971967, + "loss": 0.7116, + "step": 752 + }, + { + "epoch": 0.13408119658119658, + "grad_norm": 0.4103890359401703, + "learning_rate": 0.00019947410849141756, + "loss": 1.1527, + "step": 753 + }, + { + "epoch": 0.13425925925925927, + "grad_norm": 0.5390757322311401, + "learning_rate": 0.00019947267387412695, + "loss": 1.1682, + "step": 754 + }, + { + "epoch": 0.13443732193732194, + "grad_norm": 0.29939723014831543, + "learning_rate": 0.0001994712373078759, + "loss": 0.5848, + "step": 755 + }, + { + "epoch": 0.1346153846153846, + "grad_norm": 0.4605920612812042, + "learning_rate": 0.0001994697987926926, + "loss": 0.9448, + "step": 756 + }, + { + "epoch": 0.1347934472934473, + "grad_norm": 0.426213800907135, + "learning_rate": 0.00019946835832860527, + "loss": 1.0487, + "step": 757 + }, + { + "epoch": 0.13497150997150997, + "grad_norm": 0.4209515154361725, + "learning_rate": 0.00019946691591564203, + "loss": 1.0951, + "step": 758 + }, + { + "epoch": 0.13514957264957264, + "grad_norm": 0.39555591344833374, + "learning_rate": 0.0001994654715538312, + "loss": 0.8754, + "step": 759 + }, + { + "epoch": 0.13532763532763534, + "grad_norm": 0.4065483510494232, + "learning_rate": 0.0001994640252432011, + "loss": 0.9451, + "step": 760 + }, + { + "epoch": 0.135505698005698, + "grad_norm": 0.4489104151725769, + "learning_rate": 0.00019946257698378003, + "loss": 1.2031, + "step": 761 + }, + { + "epoch": 0.13568376068376067, + "grad_norm": 0.39928409457206726, + "learning_rate": 0.0001994611267755964, + "loss": 1.1124, + "step": 762 + }, + { + "epoch": 0.13586182336182337, + "grad_norm": 0.4145409166812897, + "learning_rate": 0.00019945967461867858, + "loss": 1.083, + "step": 763 + }, + { + "epoch": 0.13603988603988604, + "grad_norm": 0.43508613109588623, + "learning_rate": 0.00019945822051305507, + "loss": 1.1119, + "step": 764 + }, + { + "epoch": 0.1362179487179487, + "grad_norm": 0.5186598300933838, + "learning_rate": 0.0001994567644587543, + "loss": 1.3256, + "step": 765 + }, + { + "epoch": 0.1363960113960114, + "grad_norm": 0.4615778625011444, + "learning_rate": 0.00019945530645580487, + "loss": 1.3906, + "step": 766 + }, + { + "epoch": 0.13657407407407407, + "grad_norm": 0.4838152527809143, + "learning_rate": 0.00019945384650423532, + "loss": 0.8169, + "step": 767 + }, + { + "epoch": 0.13675213675213677, + "grad_norm": 0.49253368377685547, + "learning_rate": 0.0001994523846040742, + "loss": 1.1613, + "step": 768 + }, + { + "epoch": 0.13693019943019943, + "grad_norm": 0.4697009325027466, + "learning_rate": 0.00019945092075535024, + "loss": 1.1722, + "step": 769 + }, + { + "epoch": 0.1371082621082621, + "grad_norm": 0.47162383794784546, + "learning_rate": 0.00019944945495809204, + "loss": 1.054, + "step": 770 + }, + { + "epoch": 0.1372863247863248, + "grad_norm": 0.4653547704219818, + "learning_rate": 0.00019944798721232835, + "loss": 1.1791, + "step": 771 + }, + { + "epoch": 0.13746438746438747, + "grad_norm": 0.4244011640548706, + "learning_rate": 0.000199446517518088, + "loss": 1.1557, + "step": 772 + }, + { + "epoch": 0.13764245014245013, + "grad_norm": 0.43812859058380127, + "learning_rate": 0.00019944504587539967, + "loss": 1.1567, + "step": 773 + }, + { + "epoch": 0.13782051282051283, + "grad_norm": 0.3984275162220001, + "learning_rate": 0.00019944357228429227, + "loss": 1.0715, + "step": 774 + }, + { + "epoch": 0.1379985754985755, + "grad_norm": 0.3794248104095459, + "learning_rate": 0.0001994420967447946, + "loss": 0.9377, + "step": 775 + }, + { + "epoch": 0.13817663817663817, + "grad_norm": 0.4214578866958618, + "learning_rate": 0.00019944061925693566, + "loss": 1.0112, + "step": 776 + }, + { + "epoch": 0.13835470085470086, + "grad_norm": 0.4738999605178833, + "learning_rate": 0.00019943913982074435, + "loss": 0.8718, + "step": 777 + }, + { + "epoch": 0.13853276353276353, + "grad_norm": 0.43455326557159424, + "learning_rate": 0.00019943765843624965, + "loss": 1.1343, + "step": 778 + }, + { + "epoch": 0.1387108262108262, + "grad_norm": 0.44973456859588623, + "learning_rate": 0.00019943617510348062, + "loss": 1.0487, + "step": 779 + }, + { + "epoch": 0.1388888888888889, + "grad_norm": 0.4216597080230713, + "learning_rate": 0.00019943468982246628, + "loss": 1.0765, + "step": 780 + }, + { + "epoch": 0.13906695156695156, + "grad_norm": 0.5089883208274841, + "learning_rate": 0.00019943320259323578, + "loss": 1.3137, + "step": 781 + }, + { + "epoch": 0.13924501424501423, + "grad_norm": 0.4358222782611847, + "learning_rate": 0.00019943171341581822, + "loss": 1.1891, + "step": 782 + }, + { + "epoch": 0.13942307692307693, + "grad_norm": 0.40918609499931335, + "learning_rate": 0.00019943022229024275, + "loss": 1.279, + "step": 783 + }, + { + "epoch": 0.1396011396011396, + "grad_norm": 0.4614863395690918, + "learning_rate": 0.00019942872921653866, + "loss": 1.2477, + "step": 784 + }, + { + "epoch": 0.1397792022792023, + "grad_norm": 0.4141528904438019, + "learning_rate": 0.00019942723419473515, + "loss": 0.9622, + "step": 785 + }, + { + "epoch": 0.13995726495726496, + "grad_norm": 0.536139726638794, + "learning_rate": 0.00019942573722486154, + "loss": 1.2127, + "step": 786 + }, + { + "epoch": 0.14013532763532763, + "grad_norm": 0.4968845546245575, + "learning_rate": 0.0001994242383069471, + "loss": 1.2965, + "step": 787 + }, + { + "epoch": 0.14031339031339032, + "grad_norm": 0.3897174894809723, + "learning_rate": 0.00019942273744102132, + "loss": 0.9907, + "step": 788 + }, + { + "epoch": 0.140491452991453, + "grad_norm": 0.466307669878006, + "learning_rate": 0.0001994212346271135, + "loss": 1.2021, + "step": 789 + }, + { + "epoch": 0.14066951566951566, + "grad_norm": 0.49283576011657715, + "learning_rate": 0.0001994197298652531, + "loss": 1.0969, + "step": 790 + }, + { + "epoch": 0.14084757834757836, + "grad_norm": 0.4686102271080017, + "learning_rate": 0.00019941822315546964, + "loss": 1.0125, + "step": 791 + }, + { + "epoch": 0.14102564102564102, + "grad_norm": 0.4389997124671936, + "learning_rate": 0.0001994167144977926, + "loss": 1.1294, + "step": 792 + }, + { + "epoch": 0.1412037037037037, + "grad_norm": 0.38539355993270874, + "learning_rate": 0.00019941520389225162, + "loss": 1.1231, + "step": 793 + }, + { + "epoch": 0.1413817663817664, + "grad_norm": 0.4860847592353821, + "learning_rate": 0.00019941369133887618, + "loss": 1.2268, + "step": 794 + }, + { + "epoch": 0.14155982905982906, + "grad_norm": 0.4567467272281647, + "learning_rate": 0.00019941217683769598, + "loss": 1.1482, + "step": 795 + }, + { + "epoch": 0.14173789173789172, + "grad_norm": 0.5549420714378357, + "learning_rate": 0.00019941066038874067, + "loss": 1.1899, + "step": 796 + }, + { + "epoch": 0.14191595441595442, + "grad_norm": 0.3950003385543823, + "learning_rate": 0.00019940914199204, + "loss": 0.96, + "step": 797 + }, + { + "epoch": 0.1420940170940171, + "grad_norm": 0.43845999240875244, + "learning_rate": 0.00019940762164762373, + "loss": 1.0338, + "step": 798 + }, + { + "epoch": 0.14227207977207978, + "grad_norm": 0.468537300825119, + "learning_rate": 0.00019940609935552157, + "loss": 1.2416, + "step": 799 + }, + { + "epoch": 0.14245014245014245, + "grad_norm": 0.4292038679122925, + "learning_rate": 0.0001994045751157634, + "loss": 1.1397, + "step": 800 + }, + { + "epoch": 0.14262820512820512, + "grad_norm": 0.3800995647907257, + "learning_rate": 0.00019940304892837908, + "loss": 0.939, + "step": 801 + }, + { + "epoch": 0.14280626780626782, + "grad_norm": 0.38004353642463684, + "learning_rate": 0.00019940152079339852, + "loss": 1.0485, + "step": 802 + }, + { + "epoch": 0.14298433048433049, + "grad_norm": 0.4658142924308777, + "learning_rate": 0.00019939999071085163, + "loss": 1.1561, + "step": 803 + }, + { + "epoch": 0.14316239316239315, + "grad_norm": 0.4235048294067383, + "learning_rate": 0.0001993984586807684, + "loss": 1.0516, + "step": 804 + }, + { + "epoch": 0.14334045584045585, + "grad_norm": 0.42925819754600525, + "learning_rate": 0.00019939692470317887, + "loss": 1.2238, + "step": 805 + }, + { + "epoch": 0.14351851851851852, + "grad_norm": 0.43701639771461487, + "learning_rate": 0.00019939538877811308, + "loss": 1.0129, + "step": 806 + }, + { + "epoch": 0.14369658119658119, + "grad_norm": 0.42786353826522827, + "learning_rate": 0.00019939385090560113, + "loss": 1.1355, + "step": 807 + }, + { + "epoch": 0.14387464387464388, + "grad_norm": 0.371218740940094, + "learning_rate": 0.00019939231108567312, + "loss": 0.9712, + "step": 808 + }, + { + "epoch": 0.14405270655270655, + "grad_norm": 0.4834294617176056, + "learning_rate": 0.00019939076931835926, + "loss": 1.1375, + "step": 809 + }, + { + "epoch": 0.14423076923076922, + "grad_norm": 0.4700150191783905, + "learning_rate": 0.00019938922560368974, + "loss": 1.1943, + "step": 810 + }, + { + "epoch": 0.14440883190883191, + "grad_norm": 0.4430996775627136, + "learning_rate": 0.0001993876799416948, + "loss": 1.1976, + "step": 811 + }, + { + "epoch": 0.14458689458689458, + "grad_norm": 0.4161672592163086, + "learning_rate": 0.00019938613233240476, + "loss": 1.0291, + "step": 812 + }, + { + "epoch": 0.14476495726495728, + "grad_norm": 0.39838850498199463, + "learning_rate": 0.0001993845827758499, + "loss": 1.2103, + "step": 813 + }, + { + "epoch": 0.14494301994301995, + "grad_norm": 0.429198294878006, + "learning_rate": 0.00019938303127206057, + "loss": 0.9971, + "step": 814 + }, + { + "epoch": 0.14512108262108261, + "grad_norm": 0.4589254856109619, + "learning_rate": 0.00019938147782106719, + "loss": 1.2392, + "step": 815 + }, + { + "epoch": 0.1452991452991453, + "grad_norm": 0.42506635189056396, + "learning_rate": 0.00019937992242290023, + "loss": 1.0827, + "step": 816 + }, + { + "epoch": 0.14547720797720798, + "grad_norm": 0.3778113126754761, + "learning_rate": 0.00019937836507759012, + "loss": 1.021, + "step": 817 + }, + { + "epoch": 0.14565527065527065, + "grad_norm": 0.43071216344833374, + "learning_rate": 0.0001993768057851674, + "loss": 1.273, + "step": 818 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 0.4944681227207184, + "learning_rate": 0.00019937524454566262, + "loss": 1.3037, + "step": 819 + }, + { + "epoch": 0.146011396011396, + "grad_norm": 0.4438824951648712, + "learning_rate": 0.00019937368135910632, + "loss": 1.1383, + "step": 820 + }, + { + "epoch": 0.14618945868945868, + "grad_norm": 0.400215744972229, + "learning_rate": 0.0001993721162255292, + "loss": 1.0669, + "step": 821 + }, + { + "epoch": 0.14636752136752137, + "grad_norm": 0.4341452121734619, + "learning_rate": 0.00019937054914496185, + "loss": 1.1431, + "step": 822 + }, + { + "epoch": 0.14654558404558404, + "grad_norm": 0.3941744267940521, + "learning_rate": 0.00019936898011743503, + "loss": 1.1593, + "step": 823 + }, + { + "epoch": 0.1467236467236467, + "grad_norm": 0.4318541884422302, + "learning_rate": 0.00019936740914297947, + "loss": 1.2814, + "step": 824 + }, + { + "epoch": 0.1469017094017094, + "grad_norm": 0.44488632678985596, + "learning_rate": 0.00019936583622162595, + "loss": 1.1054, + "step": 825 + }, + { + "epoch": 0.14707977207977208, + "grad_norm": 0.38701096177101135, + "learning_rate": 0.00019936426135340528, + "loss": 1.1086, + "step": 826 + }, + { + "epoch": 0.14725783475783477, + "grad_norm": 0.45794424414634705, + "learning_rate": 0.0001993626845383483, + "loss": 1.2395, + "step": 827 + }, + { + "epoch": 0.14743589743589744, + "grad_norm": 0.49237680435180664, + "learning_rate": 0.00019936110577648596, + "loss": 1.3483, + "step": 828 + }, + { + "epoch": 0.1476139601139601, + "grad_norm": 0.481666624546051, + "learning_rate": 0.00019935952506784914, + "loss": 1.1848, + "step": 829 + }, + { + "epoch": 0.1477920227920228, + "grad_norm": 0.4015209376811981, + "learning_rate": 0.00019935794241246883, + "loss": 1.0624, + "step": 830 + }, + { + "epoch": 0.14797008547008547, + "grad_norm": 0.47975999116897583, + "learning_rate": 0.00019935635781037606, + "loss": 1.1595, + "step": 831 + }, + { + "epoch": 0.14814814814814814, + "grad_norm": 0.4440356492996216, + "learning_rate": 0.00019935477126160181, + "loss": 1.1325, + "step": 832 + }, + { + "epoch": 0.14832621082621084, + "grad_norm": 0.4167410731315613, + "learning_rate": 0.00019935318276617723, + "loss": 1.0662, + "step": 833 + }, + { + "epoch": 0.1485042735042735, + "grad_norm": 0.4107447862625122, + "learning_rate": 0.0001993515923241334, + "loss": 0.8816, + "step": 834 + }, + { + "epoch": 0.14868233618233617, + "grad_norm": 0.4020158648490906, + "learning_rate": 0.00019934999993550154, + "loss": 0.9797, + "step": 835 + }, + { + "epoch": 0.14886039886039887, + "grad_norm": 0.4186473786830902, + "learning_rate": 0.0001993484056003128, + "loss": 1.1243, + "step": 836 + }, + { + "epoch": 0.14903846153846154, + "grad_norm": 0.5534794926643372, + "learning_rate": 0.00019934680931859842, + "loss": 1.1189, + "step": 837 + }, + { + "epoch": 0.1492165242165242, + "grad_norm": 0.37901270389556885, + "learning_rate": 0.0001993452110903897, + "loss": 0.9241, + "step": 838 + }, + { + "epoch": 0.1493945868945869, + "grad_norm": 0.41773587465286255, + "learning_rate": 0.00019934361091571793, + "loss": 0.9467, + "step": 839 + }, + { + "epoch": 0.14957264957264957, + "grad_norm": 0.4962073564529419, + "learning_rate": 0.00019934200879461448, + "loss": 1.2423, + "step": 840 + }, + { + "epoch": 0.14975071225071226, + "grad_norm": 0.38565897941589355, + "learning_rate": 0.00019934040472711074, + "loss": 1.1545, + "step": 841 + }, + { + "epoch": 0.14992877492877493, + "grad_norm": 0.4295346736907959, + "learning_rate": 0.0001993387987132381, + "loss": 1.2482, + "step": 842 + }, + { + "epoch": 0.1501068376068376, + "grad_norm": 0.4279189705848694, + "learning_rate": 0.0001993371907530281, + "loss": 1.1135, + "step": 843 + }, + { + "epoch": 0.1502849002849003, + "grad_norm": 0.44649168848991394, + "learning_rate": 0.0001993355808465122, + "loss": 1.0734, + "step": 844 + }, + { + "epoch": 0.15046296296296297, + "grad_norm": 0.453707218170166, + "learning_rate": 0.0001993339689937219, + "loss": 1.0992, + "step": 845 + }, + { + "epoch": 0.15064102564102563, + "grad_norm": 0.5113263726234436, + "learning_rate": 0.00019933235519468886, + "loss": 1.1792, + "step": 846 + }, + { + "epoch": 0.15081908831908833, + "grad_norm": 0.5822970271110535, + "learning_rate": 0.00019933073944944466, + "loss": 1.367, + "step": 847 + }, + { + "epoch": 0.150997150997151, + "grad_norm": 0.3946528732776642, + "learning_rate": 0.00019932912175802097, + "loss": 0.9781, + "step": 848 + }, + { + "epoch": 0.15117521367521367, + "grad_norm": 0.5429860949516296, + "learning_rate": 0.00019932750212044945, + "loss": 0.9783, + "step": 849 + }, + { + "epoch": 0.15135327635327636, + "grad_norm": 0.45847952365875244, + "learning_rate": 0.0001993258805367619, + "loss": 1.1352, + "step": 850 + }, + { + "epoch": 0.15153133903133903, + "grad_norm": 0.42770692706108093, + "learning_rate": 0.00019932425700699004, + "loss": 1.2365, + "step": 851 + }, + { + "epoch": 0.1517094017094017, + "grad_norm": 0.41845405101776123, + "learning_rate": 0.00019932263153116565, + "loss": 1.2642, + "step": 852 + }, + { + "epoch": 0.1518874643874644, + "grad_norm": 0.4641731083393097, + "learning_rate": 0.00019932100410932066, + "loss": 1.2009, + "step": 853 + }, + { + "epoch": 0.15206552706552706, + "grad_norm": 0.4128672778606415, + "learning_rate": 0.00019931937474148689, + "loss": 1.1981, + "step": 854 + }, + { + "epoch": 0.15224358974358973, + "grad_norm": 0.4730764925479889, + "learning_rate": 0.00019931774342769632, + "loss": 1.2145, + "step": 855 + }, + { + "epoch": 0.15242165242165243, + "grad_norm": 0.36611825227737427, + "learning_rate": 0.00019931611016798089, + "loss": 0.8504, + "step": 856 + }, + { + "epoch": 0.1525997150997151, + "grad_norm": 0.40944692492485046, + "learning_rate": 0.00019931447496237254, + "loss": 1.2853, + "step": 857 + }, + { + "epoch": 0.1527777777777778, + "grad_norm": 0.4521993398666382, + "learning_rate": 0.0001993128378109034, + "loss": 1.0198, + "step": 858 + }, + { + "epoch": 0.15295584045584046, + "grad_norm": 0.42113015055656433, + "learning_rate": 0.0001993111987136055, + "loss": 1.1284, + "step": 859 + }, + { + "epoch": 0.15313390313390313, + "grad_norm": 0.4117624759674072, + "learning_rate": 0.00019930955767051098, + "loss": 1.0445, + "step": 860 + }, + { + "epoch": 0.15331196581196582, + "grad_norm": 0.4807964265346527, + "learning_rate": 0.00019930791468165197, + "loss": 1.1378, + "step": 861 + }, + { + "epoch": 0.1534900284900285, + "grad_norm": 0.4186483323574066, + "learning_rate": 0.00019930626974706063, + "loss": 1.1636, + "step": 862 + }, + { + "epoch": 0.15366809116809116, + "grad_norm": 0.3764737844467163, + "learning_rate": 0.00019930462286676926, + "loss": 0.9523, + "step": 863 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.4283556044101715, + "learning_rate": 0.00019930297404081008, + "loss": 1.1008, + "step": 864 + }, + { + "epoch": 0.15402421652421652, + "grad_norm": 0.4485796093940735, + "learning_rate": 0.00019930132326921541, + "loss": 1.0834, + "step": 865 + }, + { + "epoch": 0.1542022792022792, + "grad_norm": 0.3882720172405243, + "learning_rate": 0.0001992996705520176, + "loss": 1.1086, + "step": 866 + }, + { + "epoch": 0.1543803418803419, + "grad_norm": 0.44698455929756165, + "learning_rate": 0.00019929801588924902, + "loss": 1.1437, + "step": 867 + }, + { + "epoch": 0.15455840455840456, + "grad_norm": 0.46978411078453064, + "learning_rate": 0.00019929635928094208, + "loss": 1.091, + "step": 868 + }, + { + "epoch": 0.15473646723646722, + "grad_norm": 0.4717854857444763, + "learning_rate": 0.00019929470072712927, + "loss": 1.1959, + "step": 869 + }, + { + "epoch": 0.15491452991452992, + "grad_norm": 0.4324854016304016, + "learning_rate": 0.00019929304022784305, + "loss": 1.2062, + "step": 870 + }, + { + "epoch": 0.1550925925925926, + "grad_norm": 0.3948180675506592, + "learning_rate": 0.00019929137778311597, + "loss": 1.1101, + "step": 871 + }, + { + "epoch": 0.15527065527065528, + "grad_norm": 0.40345287322998047, + "learning_rate": 0.0001992897133929806, + "loss": 0.8894, + "step": 872 + }, + { + "epoch": 0.15544871794871795, + "grad_norm": 0.44931963086128235, + "learning_rate": 0.00019928804705746957, + "loss": 0.9389, + "step": 873 + }, + { + "epoch": 0.15562678062678062, + "grad_norm": 0.529196560382843, + "learning_rate": 0.0001992863787766155, + "loss": 1.3362, + "step": 874 + }, + { + "epoch": 0.15580484330484332, + "grad_norm": 0.41218671202659607, + "learning_rate": 0.0001992847085504511, + "loss": 1.0727, + "step": 875 + }, + { + "epoch": 0.15598290598290598, + "grad_norm": 0.44074541330337524, + "learning_rate": 0.00019928303637900907, + "loss": 1.1091, + "step": 876 + }, + { + "epoch": 0.15616096866096865, + "grad_norm": 0.5264310240745544, + "learning_rate": 0.00019928136226232218, + "loss": 1.201, + "step": 877 + }, + { + "epoch": 0.15633903133903135, + "grad_norm": 0.4255099594593048, + "learning_rate": 0.00019927968620042324, + "loss": 1.2514, + "step": 878 + }, + { + "epoch": 0.15651709401709402, + "grad_norm": 0.4030280113220215, + "learning_rate": 0.0001992780081933451, + "loss": 1.0422, + "step": 879 + }, + { + "epoch": 0.15669515669515668, + "grad_norm": 0.5270203948020935, + "learning_rate": 0.00019927632824112058, + "loss": 1.2476, + "step": 880 + }, + { + "epoch": 0.15687321937321938, + "grad_norm": 0.37767237424850464, + "learning_rate": 0.00019927464634378268, + "loss": 1.0768, + "step": 881 + }, + { + "epoch": 0.15705128205128205, + "grad_norm": 0.4535936415195465, + "learning_rate": 0.0001992729625013643, + "loss": 1.2097, + "step": 882 + }, + { + "epoch": 0.15722934472934472, + "grad_norm": 0.4282119870185852, + "learning_rate": 0.00019927127671389843, + "loss": 1.0904, + "step": 883 + }, + { + "epoch": 0.1574074074074074, + "grad_norm": 0.3924157917499542, + "learning_rate": 0.0001992695889814181, + "loss": 0.9692, + "step": 884 + }, + { + "epoch": 0.15758547008547008, + "grad_norm": 0.525075376033783, + "learning_rate": 0.0001992678993039564, + "loss": 1.0292, + "step": 885 + }, + { + "epoch": 0.15776353276353278, + "grad_norm": 0.4388505518436432, + "learning_rate": 0.00019926620768154644, + "loss": 1.1944, + "step": 886 + }, + { + "epoch": 0.15794159544159544, + "grad_norm": 0.4362235963344574, + "learning_rate": 0.00019926451411422132, + "loss": 0.97, + "step": 887 + }, + { + "epoch": 0.1581196581196581, + "grad_norm": 0.4265296459197998, + "learning_rate": 0.0001992628186020143, + "loss": 0.9196, + "step": 888 + }, + { + "epoch": 0.1582977207977208, + "grad_norm": 0.4019876718521118, + "learning_rate": 0.0001992611211449585, + "loss": 1.1368, + "step": 889 + }, + { + "epoch": 0.15847578347578348, + "grad_norm": 0.5003397464752197, + "learning_rate": 0.00019925942174308726, + "loss": 1.2582, + "step": 890 + }, + { + "epoch": 0.15865384615384615, + "grad_norm": 0.4774404466152191, + "learning_rate": 0.00019925772039643382, + "loss": 1.2277, + "step": 891 + }, + { + "epoch": 0.15883190883190884, + "grad_norm": 0.4590449333190918, + "learning_rate": 0.00019925601710503153, + "loss": 1.1679, + "step": 892 + }, + { + "epoch": 0.1590099715099715, + "grad_norm": 0.4221442639827728, + "learning_rate": 0.0001992543118689138, + "loss": 1.1626, + "step": 893 + }, + { + "epoch": 0.15918803418803418, + "grad_norm": 0.47613003849983215, + "learning_rate": 0.00019925260468811403, + "loss": 1.1509, + "step": 894 + }, + { + "epoch": 0.15936609686609687, + "grad_norm": 0.41706812381744385, + "learning_rate": 0.0001992508955626656, + "loss": 1.0366, + "step": 895 + }, + { + "epoch": 0.15954415954415954, + "grad_norm": 0.5064654350280762, + "learning_rate": 0.00019924918449260205, + "loss": 1.0729, + "step": 896 + }, + { + "epoch": 0.1597222222222222, + "grad_norm": 0.5019610524177551, + "learning_rate": 0.00019924747147795696, + "loss": 1.0642, + "step": 897 + }, + { + "epoch": 0.1599002849002849, + "grad_norm": 0.4345671534538269, + "learning_rate": 0.00019924575651876378, + "loss": 1.1747, + "step": 898 + }, + { + "epoch": 0.16007834757834757, + "grad_norm": 0.4397568702697754, + "learning_rate": 0.0001992440396150562, + "loss": 1.282, + "step": 899 + }, + { + "epoch": 0.16025641025641027, + "grad_norm": 0.520187497138977, + "learning_rate": 0.0001992423207668678, + "loss": 0.976, + "step": 900 + }, + { + "epoch": 0.16043447293447294, + "grad_norm": 0.39329993724823, + "learning_rate": 0.0001992405999742323, + "loss": 0.9829, + "step": 901 + }, + { + "epoch": 0.1606125356125356, + "grad_norm": 0.42361345887184143, + "learning_rate": 0.00019923887723718339, + "loss": 1.139, + "step": 902 + }, + { + "epoch": 0.1607905982905983, + "grad_norm": 0.3846314251422882, + "learning_rate": 0.00019923715255575482, + "loss": 0.8262, + "step": 903 + }, + { + "epoch": 0.16096866096866097, + "grad_norm": 0.39258381724357605, + "learning_rate": 0.0001992354259299804, + "loss": 0.9638, + "step": 904 + }, + { + "epoch": 0.16114672364672364, + "grad_norm": 0.4000850319862366, + "learning_rate": 0.00019923369735989397, + "loss": 0.91, + "step": 905 + }, + { + "epoch": 0.16132478632478633, + "grad_norm": 0.46303513646125793, + "learning_rate": 0.00019923196684552936, + "loss": 1.1447, + "step": 906 + }, + { + "epoch": 0.161502849002849, + "grad_norm": 0.38437438011169434, + "learning_rate": 0.0001992302343869205, + "loss": 1.0212, + "step": 907 + }, + { + "epoch": 0.16168091168091167, + "grad_norm": 0.44585472345352173, + "learning_rate": 0.00019922849998410135, + "loss": 1.1964, + "step": 908 + }, + { + "epoch": 0.16185897435897437, + "grad_norm": 0.41959813237190247, + "learning_rate": 0.00019922676363710583, + "loss": 0.9925, + "step": 909 + }, + { + "epoch": 0.16203703703703703, + "grad_norm": 0.47442761063575745, + "learning_rate": 0.00019922502534596803, + "loss": 0.9237, + "step": 910 + }, + { + "epoch": 0.1622150997150997, + "grad_norm": 0.5065128207206726, + "learning_rate": 0.00019922328511072198, + "loss": 1.2573, + "step": 911 + }, + { + "epoch": 0.1623931623931624, + "grad_norm": 0.4739879369735718, + "learning_rate": 0.0001992215429314018, + "loss": 1.4416, + "step": 912 + }, + { + "epoch": 0.16257122507122507, + "grad_norm": 0.48763832449913025, + "learning_rate": 0.00019921979880804157, + "loss": 1.0408, + "step": 913 + }, + { + "epoch": 0.16274928774928774, + "grad_norm": 0.4841614067554474, + "learning_rate": 0.0001992180527406755, + "loss": 1.1826, + "step": 914 + }, + { + "epoch": 0.16292735042735043, + "grad_norm": 0.49433308839797974, + "learning_rate": 0.0001992163047293378, + "loss": 1.3552, + "step": 915 + }, + { + "epoch": 0.1631054131054131, + "grad_norm": 0.4985002875328064, + "learning_rate": 0.0001992145547740627, + "loss": 1.2639, + "step": 916 + }, + { + "epoch": 0.1632834757834758, + "grad_norm": 0.40348032116889954, + "learning_rate": 0.00019921280287488448, + "loss": 1.1731, + "step": 917 + }, + { + "epoch": 0.16346153846153846, + "grad_norm": 0.5166002511978149, + "learning_rate": 0.0001992110490318375, + "loss": 1.0692, + "step": 918 + }, + { + "epoch": 0.16363960113960113, + "grad_norm": 0.44233468174934387, + "learning_rate": 0.00019920929324495615, + "loss": 1.0488, + "step": 919 + }, + { + "epoch": 0.16381766381766383, + "grad_norm": 0.43709903955459595, + "learning_rate": 0.00019920753551427476, + "loss": 0.8884, + "step": 920 + }, + { + "epoch": 0.1639957264957265, + "grad_norm": 0.4054167568683624, + "learning_rate": 0.00019920577583982778, + "loss": 0.9872, + "step": 921 + }, + { + "epoch": 0.16417378917378916, + "grad_norm": 0.4657362997531891, + "learning_rate": 0.0001992040142216497, + "loss": 1.4402, + "step": 922 + }, + { + "epoch": 0.16435185185185186, + "grad_norm": 0.42550426721572876, + "learning_rate": 0.0001992022506597751, + "loss": 1.0456, + "step": 923 + }, + { + "epoch": 0.16452991452991453, + "grad_norm": 0.49346762895584106, + "learning_rate": 0.00019920048515423842, + "loss": 1.527, + "step": 924 + }, + { + "epoch": 0.1647079772079772, + "grad_norm": 0.3970337510108948, + "learning_rate": 0.0001991987177050743, + "loss": 1.0363, + "step": 925 + }, + { + "epoch": 0.1648860398860399, + "grad_norm": 0.4027378559112549, + "learning_rate": 0.0001991969483123174, + "loss": 0.8416, + "step": 926 + }, + { + "epoch": 0.16506410256410256, + "grad_norm": 0.4181644916534424, + "learning_rate": 0.00019919517697600237, + "loss": 1.2253, + "step": 927 + }, + { + "epoch": 0.16524216524216523, + "grad_norm": 0.43686383962631226, + "learning_rate": 0.0001991934036961639, + "loss": 1.0808, + "step": 928 + }, + { + "epoch": 0.16542022792022792, + "grad_norm": 0.4242876172065735, + "learning_rate": 0.0001991916284728367, + "loss": 0.9483, + "step": 929 + }, + { + "epoch": 0.1655982905982906, + "grad_norm": 0.3690609037876129, + "learning_rate": 0.00019918985130605563, + "loss": 0.9495, + "step": 930 + }, + { + "epoch": 0.1657763532763533, + "grad_norm": 0.42184555530548096, + "learning_rate": 0.00019918807219585546, + "loss": 1.0966, + "step": 931 + }, + { + "epoch": 0.16595441595441596, + "grad_norm": 0.4342746138572693, + "learning_rate": 0.00019918629114227106, + "loss": 1.0875, + "step": 932 + }, + { + "epoch": 0.16613247863247863, + "grad_norm": 0.4191494286060333, + "learning_rate": 0.00019918450814533737, + "loss": 1.0777, + "step": 933 + }, + { + "epoch": 0.16631054131054132, + "grad_norm": 0.37124550342559814, + "learning_rate": 0.00019918272320508922, + "loss": 1.0131, + "step": 934 + }, + { + "epoch": 0.166488603988604, + "grad_norm": 0.4475722014904022, + "learning_rate": 0.00019918093632156168, + "loss": 1.1185, + "step": 935 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.4629058241844177, + "learning_rate": 0.0001991791474947897, + "loss": 1.0353, + "step": 936 + }, + { + "epoch": 0.16684472934472935, + "grad_norm": 0.48192909359931946, + "learning_rate": 0.00019917735672480834, + "loss": 1.1628, + "step": 937 + }, + { + "epoch": 0.16702279202279202, + "grad_norm": 0.5542252063751221, + "learning_rate": 0.00019917556401165273, + "loss": 1.3133, + "step": 938 + }, + { + "epoch": 0.1672008547008547, + "grad_norm": 0.4172651171684265, + "learning_rate": 0.00019917376935535796, + "loss": 1.1733, + "step": 939 + }, + { + "epoch": 0.16737891737891739, + "grad_norm": 0.4424920380115509, + "learning_rate": 0.0001991719727559592, + "loss": 1.0262, + "step": 940 + }, + { + "epoch": 0.16755698005698005, + "grad_norm": 0.4551742970943451, + "learning_rate": 0.00019917017421349162, + "loss": 1.0883, + "step": 941 + }, + { + "epoch": 0.16773504273504272, + "grad_norm": 0.45929640531539917, + "learning_rate": 0.00019916837372799048, + "loss": 1.1836, + "step": 942 + }, + { + "epoch": 0.16791310541310542, + "grad_norm": 0.4609353542327881, + "learning_rate": 0.0001991665712994911, + "loss": 1.0682, + "step": 943 + }, + { + "epoch": 0.16809116809116809, + "grad_norm": 0.42617303133010864, + "learning_rate": 0.00019916476692802873, + "loss": 1.074, + "step": 944 + }, + { + "epoch": 0.16826923076923078, + "grad_norm": 0.41919493675231934, + "learning_rate": 0.00019916296061363875, + "loss": 1.0969, + "step": 945 + }, + { + "epoch": 0.16844729344729345, + "grad_norm": 0.450979083776474, + "learning_rate": 0.00019916115235635656, + "loss": 1.1686, + "step": 946 + }, + { + "epoch": 0.16862535612535612, + "grad_norm": 0.42166751623153687, + "learning_rate": 0.00019915934215621758, + "loss": 0.9273, + "step": 947 + }, + { + "epoch": 0.16880341880341881, + "grad_norm": 0.4404160976409912, + "learning_rate": 0.00019915753001325729, + "loss": 1.1663, + "step": 948 + }, + { + "epoch": 0.16898148148148148, + "grad_norm": 0.42025226354599, + "learning_rate": 0.0001991557159275111, + "loss": 0.9433, + "step": 949 + }, + { + "epoch": 0.16915954415954415, + "grad_norm": 0.4277796745300293, + "learning_rate": 0.00019915389989901474, + "loss": 0.8475, + "step": 950 + }, + { + "epoch": 0.16933760683760685, + "grad_norm": 0.5162755250930786, + "learning_rate": 0.00019915208192780365, + "loss": 1.1155, + "step": 951 + }, + { + "epoch": 0.16951566951566951, + "grad_norm": 0.4214856028556824, + "learning_rate": 0.00019915026201391346, + "loss": 1.173, + "step": 952 + }, + { + "epoch": 0.16969373219373218, + "grad_norm": 0.4713292419910431, + "learning_rate": 0.00019914844015737985, + "loss": 1.1615, + "step": 953 + }, + { + "epoch": 0.16987179487179488, + "grad_norm": 0.461179256439209, + "learning_rate": 0.00019914661635823854, + "loss": 1.1169, + "step": 954 + }, + { + "epoch": 0.17004985754985755, + "grad_norm": 0.46200552582740784, + "learning_rate": 0.00019914479061652527, + "loss": 1.0274, + "step": 955 + }, + { + "epoch": 0.17022792022792022, + "grad_norm": 0.40968334674835205, + "learning_rate": 0.00019914296293227572, + "loss": 1.066, + "step": 956 + }, + { + "epoch": 0.1704059829059829, + "grad_norm": 0.40877434611320496, + "learning_rate": 0.0001991411333055258, + "loss": 1.1595, + "step": 957 + }, + { + "epoch": 0.17058404558404558, + "grad_norm": 0.42940187454223633, + "learning_rate": 0.00019913930173631132, + "loss": 1.0364, + "step": 958 + }, + { + "epoch": 0.17076210826210828, + "grad_norm": 0.49648910760879517, + "learning_rate": 0.00019913746822466819, + "loss": 1.0763, + "step": 959 + }, + { + "epoch": 0.17094017094017094, + "grad_norm": 0.4353426396846771, + "learning_rate": 0.00019913563277063228, + "loss": 0.9698, + "step": 960 + }, + { + "epoch": 0.1711182336182336, + "grad_norm": 0.45079681277275085, + "learning_rate": 0.00019913379537423958, + "loss": 1.2244, + "step": 961 + }, + { + "epoch": 0.1712962962962963, + "grad_norm": 0.4276828467845917, + "learning_rate": 0.00019913195603552607, + "loss": 0.9976, + "step": 962 + }, + { + "epoch": 0.17147435897435898, + "grad_norm": 0.41122403740882874, + "learning_rate": 0.00019913011475452785, + "loss": 1.0077, + "step": 963 + }, + { + "epoch": 0.17165242165242164, + "grad_norm": 0.43170276284217834, + "learning_rate": 0.00019912827153128096, + "loss": 1.1402, + "step": 964 + }, + { + "epoch": 0.17183048433048434, + "grad_norm": 0.37950268387794495, + "learning_rate": 0.0001991264263658215, + "loss": 0.9818, + "step": 965 + }, + { + "epoch": 0.172008547008547, + "grad_norm": 0.477333128452301, + "learning_rate": 0.00019912457925818562, + "loss": 1.1756, + "step": 966 + }, + { + "epoch": 0.17218660968660968, + "grad_norm": 0.4326401352882385, + "learning_rate": 0.00019912273020840954, + "loss": 1.3718, + "step": 967 + }, + { + "epoch": 0.17236467236467237, + "grad_norm": 0.37711042165756226, + "learning_rate": 0.00019912087921652945, + "loss": 0.9011, + "step": 968 + }, + { + "epoch": 0.17254273504273504, + "grad_norm": 0.50013667345047, + "learning_rate": 0.00019911902628258162, + "loss": 1.1163, + "step": 969 + }, + { + "epoch": 0.1727207977207977, + "grad_norm": 0.41913339495658875, + "learning_rate": 0.0001991171714066024, + "loss": 1.2614, + "step": 970 + }, + { + "epoch": 0.1728988603988604, + "grad_norm": 0.4075855612754822, + "learning_rate": 0.00019911531458862813, + "loss": 0.8984, + "step": 971 + }, + { + "epoch": 0.17307692307692307, + "grad_norm": 0.40277954936027527, + "learning_rate": 0.00019911345582869513, + "loss": 1.0851, + "step": 972 + }, + { + "epoch": 0.17325498575498577, + "grad_norm": 0.4312847852706909, + "learning_rate": 0.00019911159512683987, + "loss": 1.1273, + "step": 973 + }, + { + "epoch": 0.17343304843304844, + "grad_norm": 0.40303611755371094, + "learning_rate": 0.0001991097324830988, + "loss": 0.9645, + "step": 974 + }, + { + "epoch": 0.1736111111111111, + "grad_norm": 0.45560577511787415, + "learning_rate": 0.00019910786789750838, + "loss": 1.0864, + "step": 975 + }, + { + "epoch": 0.1737891737891738, + "grad_norm": 0.43775680661201477, + "learning_rate": 0.00019910600137010517, + "loss": 1.028, + "step": 976 + }, + { + "epoch": 0.17396723646723647, + "grad_norm": 0.3917224407196045, + "learning_rate": 0.00019910413290092572, + "loss": 1.0491, + "step": 977 + }, + { + "epoch": 0.17414529914529914, + "grad_norm": 0.4068751037120819, + "learning_rate": 0.0001991022624900067, + "loss": 1.0476, + "step": 978 + }, + { + "epoch": 0.17432336182336183, + "grad_norm": 0.4463370144367218, + "learning_rate": 0.0001991003901373847, + "loss": 1.0612, + "step": 979 + }, + { + "epoch": 0.1745014245014245, + "grad_norm": 0.46949052810668945, + "learning_rate": 0.0001990985158430964, + "loss": 1.3099, + "step": 980 + }, + { + "epoch": 0.17467948717948717, + "grad_norm": 0.4250012934207916, + "learning_rate": 0.00019909663960717856, + "loss": 0.9903, + "step": 981 + }, + { + "epoch": 0.17485754985754987, + "grad_norm": 0.5293903946876526, + "learning_rate": 0.0001990947614296679, + "loss": 0.9908, + "step": 982 + }, + { + "epoch": 0.17503561253561253, + "grad_norm": 0.3838284909725189, + "learning_rate": 0.0001990928813106013, + "loss": 0.716, + "step": 983 + }, + { + "epoch": 0.1752136752136752, + "grad_norm": 0.4597751200199127, + "learning_rate": 0.0001990909992500155, + "loss": 1.0126, + "step": 984 + }, + { + "epoch": 0.1753917378917379, + "grad_norm": 0.4844081699848175, + "learning_rate": 0.0001990891152479474, + "loss": 1.1043, + "step": 985 + }, + { + "epoch": 0.17556980056980057, + "grad_norm": 0.4763399660587311, + "learning_rate": 0.00019908722930443392, + "loss": 1.019, + "step": 986 + }, + { + "epoch": 0.17574786324786323, + "grad_norm": 0.4670077860355377, + "learning_rate": 0.00019908534141951204, + "loss": 1.1382, + "step": 987 + }, + { + "epoch": 0.17592592592592593, + "grad_norm": 0.39372730255126953, + "learning_rate": 0.00019908345159321873, + "loss": 1.1219, + "step": 988 + }, + { + "epoch": 0.1761039886039886, + "grad_norm": 0.41869843006134033, + "learning_rate": 0.00019908155982559098, + "loss": 0.9461, + "step": 989 + }, + { + "epoch": 0.1762820512820513, + "grad_norm": 0.4398406147956848, + "learning_rate": 0.00019907966611666593, + "loss": 1.1328, + "step": 990 + }, + { + "epoch": 0.17646011396011396, + "grad_norm": 0.4315733015537262, + "learning_rate": 0.0001990777704664806, + "loss": 1.0974, + "step": 991 + }, + { + "epoch": 0.17663817663817663, + "grad_norm": 0.42859575152397156, + "learning_rate": 0.00019907587287507222, + "loss": 1.2637, + "step": 992 + }, + { + "epoch": 0.17681623931623933, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.0001990739733424779, + "loss": 1.0699, + "step": 993 + }, + { + "epoch": 0.176994301994302, + "grad_norm": 0.4443826973438263, + "learning_rate": 0.00019907207186873488, + "loss": 1.0547, + "step": 994 + }, + { + "epoch": 0.17717236467236466, + "grad_norm": 0.4108099937438965, + "learning_rate": 0.00019907016845388043, + "loss": 1.1401, + "step": 995 + }, + { + "epoch": 0.17735042735042736, + "grad_norm": 0.4474675953388214, + "learning_rate": 0.00019906826309795182, + "loss": 1.0712, + "step": 996 + }, + { + "epoch": 0.17752849002849003, + "grad_norm": 0.4149756133556366, + "learning_rate": 0.00019906635580098638, + "loss": 0.9585, + "step": 997 + }, + { + "epoch": 0.1777065527065527, + "grad_norm": 0.4875968098640442, + "learning_rate": 0.00019906444656302152, + "loss": 1.0659, + "step": 998 + }, + { + "epoch": 0.1778846153846154, + "grad_norm": 0.5494784116744995, + "learning_rate": 0.0001990625353840946, + "loss": 1.2858, + "step": 999 + }, + { + "epoch": 0.17806267806267806, + "grad_norm": 0.425062358379364, + "learning_rate": 0.0001990606222642431, + "loss": 1.1826, + "step": 1000 + }, + { + "epoch": 0.17824074074074073, + "grad_norm": 0.3890725374221802, + "learning_rate": 0.00019905870720350445, + "loss": 0.9568, + "step": 1001 + }, + { + "epoch": 0.17841880341880342, + "grad_norm": 0.3884070813655853, + "learning_rate": 0.00019905679020191624, + "loss": 0.9674, + "step": 1002 + }, + { + "epoch": 0.1785968660968661, + "grad_norm": 0.49496129155158997, + "learning_rate": 0.00019905487125951597, + "loss": 0.9143, + "step": 1003 + }, + { + "epoch": 0.1787749287749288, + "grad_norm": 0.43448135256767273, + "learning_rate": 0.00019905295037634128, + "loss": 1.2677, + "step": 1004 + }, + { + "epoch": 0.17895299145299146, + "grad_norm": 0.47327905893325806, + "learning_rate": 0.00019905102755242982, + "loss": 0.9089, + "step": 1005 + }, + { + "epoch": 0.17913105413105412, + "grad_norm": 0.4962378442287445, + "learning_rate": 0.00019904910278781922, + "loss": 1.1748, + "step": 1006 + }, + { + "epoch": 0.17930911680911682, + "grad_norm": 0.4343934655189514, + "learning_rate": 0.0001990471760825472, + "loss": 1.2176, + "step": 1007 + }, + { + "epoch": 0.1794871794871795, + "grad_norm": 0.4695793092250824, + "learning_rate": 0.0001990452474366515, + "loss": 1.1822, + "step": 1008 + }, + { + "epoch": 0.17966524216524216, + "grad_norm": 0.4156060516834259, + "learning_rate": 0.00019904331685016995, + "loss": 0.8231, + "step": 1009 + }, + { + "epoch": 0.17984330484330485, + "grad_norm": 0.5068191885948181, + "learning_rate": 0.00019904138432314035, + "loss": 1.1363, + "step": 1010 + }, + { + "epoch": 0.18002136752136752, + "grad_norm": 0.5189786553382874, + "learning_rate": 0.00019903944985560058, + "loss": 1.3131, + "step": 1011 + }, + { + "epoch": 0.1801994301994302, + "grad_norm": 0.5126828551292419, + "learning_rate": 0.00019903751344758848, + "loss": 1.0305, + "step": 1012 + }, + { + "epoch": 0.18037749287749288, + "grad_norm": 0.41045933961868286, + "learning_rate": 0.00019903557509914205, + "loss": 1.2726, + "step": 1013 + }, + { + "epoch": 0.18055555555555555, + "grad_norm": 0.4141713082790375, + "learning_rate": 0.0001990336348102993, + "loss": 0.9606, + "step": 1014 + }, + { + "epoch": 0.18073361823361822, + "grad_norm": 0.42652079463005066, + "learning_rate": 0.00019903169258109812, + "loss": 1.0235, + "step": 1015 + }, + { + "epoch": 0.18091168091168092, + "grad_norm": 0.42098379135131836, + "learning_rate": 0.0001990297484115767, + "loss": 1.0602, + "step": 1016 + }, + { + "epoch": 0.18108974358974358, + "grad_norm": 0.49920013546943665, + "learning_rate": 0.0001990278023017731, + "loss": 1.3322, + "step": 1017 + }, + { + "epoch": 0.18126780626780628, + "grad_norm": 0.412304550409317, + "learning_rate": 0.00019902585425172537, + "loss": 1.1011, + "step": 1018 + }, + { + "epoch": 0.18144586894586895, + "grad_norm": 0.44226935505867004, + "learning_rate": 0.00019902390426147177, + "loss": 0.9777, + "step": 1019 + }, + { + "epoch": 0.18162393162393162, + "grad_norm": 0.4685269594192505, + "learning_rate": 0.00019902195233105046, + "loss": 1.3587, + "step": 1020 + }, + { + "epoch": 0.1818019943019943, + "grad_norm": 0.4500584304332733, + "learning_rate": 0.00019901999846049968, + "loss": 0.9888, + "step": 1021 + }, + { + "epoch": 0.18198005698005698, + "grad_norm": 0.48566994071006775, + "learning_rate": 0.00019901804264985774, + "loss": 1.2364, + "step": 1022 + }, + { + "epoch": 0.18215811965811965, + "grad_norm": 0.4063156247138977, + "learning_rate": 0.00019901608489916294, + "loss": 1.2224, + "step": 1023 + }, + { + "epoch": 0.18233618233618235, + "grad_norm": 0.471276193857193, + "learning_rate": 0.00019901412520845367, + "loss": 0.9926, + "step": 1024 + }, + { + "epoch": 0.182514245014245, + "grad_norm": 0.5165421366691589, + "learning_rate": 0.00019901216357776829, + "loss": 0.9595, + "step": 1025 + }, + { + "epoch": 0.18269230769230768, + "grad_norm": 0.4746754467487335, + "learning_rate": 0.0001990102000071452, + "loss": 1.2057, + "step": 1026 + }, + { + "epoch": 0.18287037037037038, + "grad_norm": 0.44803035259246826, + "learning_rate": 0.00019900823449662297, + "loss": 1.2114, + "step": 1027 + }, + { + "epoch": 0.18304843304843305, + "grad_norm": 0.47256240248680115, + "learning_rate": 0.00019900626704624005, + "loss": 1.112, + "step": 1028 + }, + { + "epoch": 0.18322649572649571, + "grad_norm": 0.4253387153148651, + "learning_rate": 0.000199004297656035, + "loss": 0.9899, + "step": 1029 + }, + { + "epoch": 0.1834045584045584, + "grad_norm": 0.44958099722862244, + "learning_rate": 0.00019900232632604636, + "loss": 1.1445, + "step": 1030 + }, + { + "epoch": 0.18358262108262108, + "grad_norm": 0.5296537280082703, + "learning_rate": 0.00019900035305631285, + "loss": 1.2502, + "step": 1031 + }, + { + "epoch": 0.18376068376068377, + "grad_norm": 0.5057148933410645, + "learning_rate": 0.00019899837784687302, + "loss": 1.1426, + "step": 1032 + }, + { + "epoch": 0.18393874643874644, + "grad_norm": 0.41463762521743774, + "learning_rate": 0.00019899640069776566, + "loss": 1.1854, + "step": 1033 + }, + { + "epoch": 0.1841168091168091, + "grad_norm": 0.45800045132637024, + "learning_rate": 0.00019899442160902945, + "loss": 1.2438, + "step": 1034 + }, + { + "epoch": 0.1842948717948718, + "grad_norm": 0.43450453877449036, + "learning_rate": 0.00019899244058070324, + "loss": 1.0598, + "step": 1035 + }, + { + "epoch": 0.18447293447293447, + "grad_norm": 0.4141148626804352, + "learning_rate": 0.00019899045761282577, + "loss": 1.0465, + "step": 1036 + }, + { + "epoch": 0.18465099715099714, + "grad_norm": 0.3938458263874054, + "learning_rate": 0.0001989884727054359, + "loss": 1.0142, + "step": 1037 + }, + { + "epoch": 0.18482905982905984, + "grad_norm": 0.43898263573646545, + "learning_rate": 0.00019898648585857257, + "loss": 0.9212, + "step": 1038 + }, + { + "epoch": 0.1850071225071225, + "grad_norm": 0.4425487816333771, + "learning_rate": 0.00019898449707227465, + "loss": 1.2987, + "step": 1039 + }, + { + "epoch": 0.18518518518518517, + "grad_norm": 0.4537975490093231, + "learning_rate": 0.00019898250634658115, + "loss": 1.2023, + "step": 1040 + }, + { + "epoch": 0.18536324786324787, + "grad_norm": 0.4107198119163513, + "learning_rate": 0.00019898051368153104, + "loss": 0.8443, + "step": 1041 + }, + { + "epoch": 0.18554131054131054, + "grad_norm": 0.4389404058456421, + "learning_rate": 0.0001989785190771634, + "loss": 1.0502, + "step": 1042 + }, + { + "epoch": 0.1857193732193732, + "grad_norm": 0.4288824796676636, + "learning_rate": 0.00019897652253351726, + "loss": 1.01, + "step": 1043 + }, + { + "epoch": 0.1858974358974359, + "grad_norm": 0.50815349817276, + "learning_rate": 0.00019897452405063178, + "loss": 1.0308, + "step": 1044 + }, + { + "epoch": 0.18607549857549857, + "grad_norm": 0.45252710580825806, + "learning_rate": 0.0001989725236285461, + "loss": 1.0967, + "step": 1045 + }, + { + "epoch": 0.18625356125356127, + "grad_norm": 0.45049402117729187, + "learning_rate": 0.00019897052126729943, + "loss": 1.0141, + "step": 1046 + }, + { + "epoch": 0.18643162393162394, + "grad_norm": 0.49637508392333984, + "learning_rate": 0.00019896851696693098, + "loss": 1.0997, + "step": 1047 + }, + { + "epoch": 0.1866096866096866, + "grad_norm": 0.4465886056423187, + "learning_rate": 0.00019896651072748005, + "loss": 1.1415, + "step": 1048 + }, + { + "epoch": 0.1867877492877493, + "grad_norm": 0.5309500694274902, + "learning_rate": 0.00019896450254898592, + "loss": 1.1028, + "step": 1049 + }, + { + "epoch": 0.18696581196581197, + "grad_norm": 0.3516653776168823, + "learning_rate": 0.00019896249243148793, + "loss": 0.9841, + "step": 1050 + }, + { + "epoch": 0.18714387464387464, + "grad_norm": 0.4529176950454712, + "learning_rate": 0.0001989604803750255, + "loss": 1.1335, + "step": 1051 + }, + { + "epoch": 0.18732193732193733, + "grad_norm": 0.47694942355155945, + "learning_rate": 0.000198958466379638, + "loss": 1.2383, + "step": 1052 + }, + { + "epoch": 0.1875, + "grad_norm": 0.5524206757545471, + "learning_rate": 0.0001989564504453649, + "loss": 1.3668, + "step": 1053 + }, + { + "epoch": 0.18767806267806267, + "grad_norm": 0.39203691482543945, + "learning_rate": 0.00019895443257224576, + "loss": 1.2203, + "step": 1054 + }, + { + "epoch": 0.18785612535612536, + "grad_norm": 0.4164120852947235, + "learning_rate": 0.00019895241276032005, + "loss": 0.8954, + "step": 1055 + }, + { + "epoch": 0.18803418803418803, + "grad_norm": 0.41217970848083496, + "learning_rate": 0.0001989503910096274, + "loss": 1.0238, + "step": 1056 + }, + { + "epoch": 0.1882122507122507, + "grad_norm": 0.44038307666778564, + "learning_rate": 0.00019894836732020735, + "loss": 0.8159, + "step": 1057 + }, + { + "epoch": 0.1883903133903134, + "grad_norm": 0.45780670642852783, + "learning_rate": 0.0001989463416920996, + "loss": 1.2864, + "step": 1058 + }, + { + "epoch": 0.18856837606837606, + "grad_norm": 0.5197559595108032, + "learning_rate": 0.00019894431412534384, + "loss": 1.0756, + "step": 1059 + }, + { + "epoch": 0.18874643874643873, + "grad_norm": 0.43283385038375854, + "learning_rate": 0.00019894228461997979, + "loss": 1.0642, + "step": 1060 + }, + { + "epoch": 0.18892450142450143, + "grad_norm": 0.4657376706600189, + "learning_rate": 0.00019894025317604717, + "loss": 1.1159, + "step": 1061 + }, + { + "epoch": 0.1891025641025641, + "grad_norm": 0.4474908113479614, + "learning_rate": 0.00019893821979358588, + "loss": 1.2006, + "step": 1062 + }, + { + "epoch": 0.1892806267806268, + "grad_norm": 0.43878164887428284, + "learning_rate": 0.00019893618447263566, + "loss": 1.1599, + "step": 1063 + }, + { + "epoch": 0.18945868945868946, + "grad_norm": 0.4598735272884369, + "learning_rate": 0.00019893414721323645, + "loss": 1.3346, + "step": 1064 + }, + { + "epoch": 0.18963675213675213, + "grad_norm": 0.3947420120239258, + "learning_rate": 0.00019893210801542812, + "loss": 1.1201, + "step": 1065 + }, + { + "epoch": 0.18981481481481483, + "grad_norm": 0.3401558995246887, + "learning_rate": 0.00019893006687925064, + "loss": 0.7568, + "step": 1066 + }, + { + "epoch": 0.1899928774928775, + "grad_norm": 0.4400341808795929, + "learning_rate": 0.00019892802380474405, + "loss": 1.1706, + "step": 1067 + }, + { + "epoch": 0.19017094017094016, + "grad_norm": 0.42394164204597473, + "learning_rate": 0.00019892597879194829, + "loss": 1.0163, + "step": 1068 + }, + { + "epoch": 0.19034900284900286, + "grad_norm": 0.42904096841812134, + "learning_rate": 0.00019892393184090353, + "loss": 0.9193, + "step": 1069 + }, + { + "epoch": 0.19052706552706553, + "grad_norm": 0.497601181268692, + "learning_rate": 0.00019892188295164977, + "loss": 1.0377, + "step": 1070 + }, + { + "epoch": 0.1907051282051282, + "grad_norm": 0.4536020755767822, + "learning_rate": 0.00019891983212422723, + "loss": 1.0946, + "step": 1071 + }, + { + "epoch": 0.1908831908831909, + "grad_norm": 0.44916942715644836, + "learning_rate": 0.00019891777935867607, + "loss": 1.0563, + "step": 1072 + }, + { + "epoch": 0.19106125356125356, + "grad_norm": 0.4256889820098877, + "learning_rate": 0.0001989157246550365, + "loss": 1.0988, + "step": 1073 + }, + { + "epoch": 0.19123931623931623, + "grad_norm": 0.5559163689613342, + "learning_rate": 0.0001989136680133488, + "loss": 0.9155, + "step": 1074 + }, + { + "epoch": 0.19141737891737892, + "grad_norm": 0.391804963350296, + "learning_rate": 0.00019891160943365322, + "loss": 0.9314, + "step": 1075 + }, + { + "epoch": 0.1915954415954416, + "grad_norm": 0.4535716474056244, + "learning_rate": 0.00019890954891599015, + "loss": 1.0768, + "step": 1076 + }, + { + "epoch": 0.19177350427350429, + "grad_norm": 0.46770521998405457, + "learning_rate": 0.00019890748646039991, + "loss": 0.8406, + "step": 1077 + }, + { + "epoch": 0.19195156695156695, + "grad_norm": 0.4875394403934479, + "learning_rate": 0.00019890542206692295, + "loss": 1.1055, + "step": 1078 + }, + { + "epoch": 0.19212962962962962, + "grad_norm": 0.5072727203369141, + "learning_rate": 0.0001989033557355997, + "loss": 1.3093, + "step": 1079 + }, + { + "epoch": 0.19230769230769232, + "grad_norm": 0.4419287443161011, + "learning_rate": 0.00019890128746647068, + "loss": 1.1916, + "step": 1080 + }, + { + "epoch": 0.192485754985755, + "grad_norm": 0.45803651213645935, + "learning_rate": 0.00019889921725957637, + "loss": 1.2579, + "step": 1081 + }, + { + "epoch": 0.19266381766381765, + "grad_norm": 0.4832262098789215, + "learning_rate": 0.0001988971451149573, + "loss": 1.3217, + "step": 1082 + }, + { + "epoch": 0.19284188034188035, + "grad_norm": 0.4819786250591278, + "learning_rate": 0.00019889507103265416, + "loss": 1.0979, + "step": 1083 + }, + { + "epoch": 0.19301994301994302, + "grad_norm": 0.49360713362693787, + "learning_rate": 0.0001988929950127075, + "loss": 1.0987, + "step": 1084 + }, + { + "epoch": 0.1931980056980057, + "grad_norm": 0.44209200143814087, + "learning_rate": 0.00019889091705515806, + "loss": 1.2616, + "step": 1085 + }, + { + "epoch": 0.19337606837606838, + "grad_norm": 0.41626206040382385, + "learning_rate": 0.00019888883716004654, + "loss": 1.0922, + "step": 1086 + }, + { + "epoch": 0.19355413105413105, + "grad_norm": 0.4916635751724243, + "learning_rate": 0.00019888675532741366, + "loss": 0.9331, + "step": 1087 + }, + { + "epoch": 0.19373219373219372, + "grad_norm": 0.4493125379085541, + "learning_rate": 0.00019888467155730025, + "loss": 1.1261, + "step": 1088 + }, + { + "epoch": 0.19391025641025642, + "grad_norm": 0.3755671977996826, + "learning_rate": 0.00019888258584974708, + "loss": 0.9821, + "step": 1089 + }, + { + "epoch": 0.19408831908831908, + "grad_norm": 0.41917556524276733, + "learning_rate": 0.00019888049820479507, + "loss": 1.251, + "step": 1090 + }, + { + "epoch": 0.19426638176638178, + "grad_norm": 0.46184420585632324, + "learning_rate": 0.0001988784086224851, + "loss": 1.1731, + "step": 1091 + }, + { + "epoch": 0.19444444444444445, + "grad_norm": 0.4783691465854645, + "learning_rate": 0.00019887631710285812, + "loss": 1.1635, + "step": 1092 + }, + { + "epoch": 0.19462250712250712, + "grad_norm": 0.4710482060909271, + "learning_rate": 0.00019887422364595512, + "loss": 1.0229, + "step": 1093 + }, + { + "epoch": 0.1948005698005698, + "grad_norm": 0.4738706648349762, + "learning_rate": 0.00019887212825181707, + "loss": 1.128, + "step": 1094 + }, + { + "epoch": 0.19497863247863248, + "grad_norm": 0.45665010809898376, + "learning_rate": 0.00019887003092048508, + "loss": 1.0425, + "step": 1095 + }, + { + "epoch": 0.19515669515669515, + "grad_norm": 0.42740485072135925, + "learning_rate": 0.0001988679316520002, + "loss": 1.0738, + "step": 1096 + }, + { + "epoch": 0.19533475783475784, + "grad_norm": 0.5977092385292053, + "learning_rate": 0.0001988658304464036, + "loss": 1.2687, + "step": 1097 + }, + { + "epoch": 0.1955128205128205, + "grad_norm": 0.4411074221134186, + "learning_rate": 0.0001988637273037364, + "loss": 1.287, + "step": 1098 + }, + { + "epoch": 0.19569088319088318, + "grad_norm": 0.4409518539905548, + "learning_rate": 0.00019886162222403986, + "loss": 1.0515, + "step": 1099 + }, + { + "epoch": 0.19586894586894588, + "grad_norm": 0.4926736652851105, + "learning_rate": 0.0001988595152073552, + "loss": 1.1388, + "step": 1100 + }, + { + "epoch": 0.19604700854700854, + "grad_norm": 0.4607115387916565, + "learning_rate": 0.00019885740625372368, + "loss": 0.9803, + "step": 1101 + }, + { + "epoch": 0.1962250712250712, + "grad_norm": 0.4725342094898224, + "learning_rate": 0.0001988552953631867, + "loss": 1.199, + "step": 1102 + }, + { + "epoch": 0.1964031339031339, + "grad_norm": 0.48014503717422485, + "learning_rate": 0.00019885318253578548, + "loss": 1.1868, + "step": 1103 + }, + { + "epoch": 0.19658119658119658, + "grad_norm": 0.3872644603252411, + "learning_rate": 0.00019885106777156155, + "loss": 0.9182, + "step": 1104 + }, + { + "epoch": 0.19675925925925927, + "grad_norm": 0.4737720787525177, + "learning_rate": 0.00019884895107055627, + "loss": 1.1513, + "step": 1105 + }, + { + "epoch": 0.19693732193732194, + "grad_norm": 0.4144562780857086, + "learning_rate": 0.00019884683243281116, + "loss": 1.1711, + "step": 1106 + }, + { + "epoch": 0.1971153846153846, + "grad_norm": 0.4672079384326935, + "learning_rate": 0.00019884471185836769, + "loss": 1.0386, + "step": 1107 + }, + { + "epoch": 0.1972934472934473, + "grad_norm": 0.4558824598789215, + "learning_rate": 0.0001988425893472674, + "loss": 1.0535, + "step": 1108 + }, + { + "epoch": 0.19747150997150997, + "grad_norm": 0.5149834752082825, + "learning_rate": 0.00019884046489955192, + "loss": 1.0296, + "step": 1109 + }, + { + "epoch": 0.19764957264957264, + "grad_norm": 0.43444496393203735, + "learning_rate": 0.00019883833851526287, + "loss": 1.1475, + "step": 1110 + }, + { + "epoch": 0.19782763532763534, + "grad_norm": 0.46062374114990234, + "learning_rate": 0.00019883621019444188, + "loss": 1.183, + "step": 1111 + }, + { + "epoch": 0.198005698005698, + "grad_norm": 0.4893282949924469, + "learning_rate": 0.00019883407993713065, + "loss": 1.3733, + "step": 1112 + }, + { + "epoch": 0.19818376068376067, + "grad_norm": 0.5434843897819519, + "learning_rate": 0.00019883194774337096, + "loss": 1.2505, + "step": 1113 + }, + { + "epoch": 0.19836182336182337, + "grad_norm": 0.4698035418987274, + "learning_rate": 0.00019882981361320456, + "loss": 1.0152, + "step": 1114 + }, + { + "epoch": 0.19853988603988604, + "grad_norm": 0.4582163989543915, + "learning_rate": 0.00019882767754667325, + "loss": 1.1718, + "step": 1115 + }, + { + "epoch": 0.1987179487179487, + "grad_norm": 0.48744696378707886, + "learning_rate": 0.0001988255395438189, + "loss": 1.2923, + "step": 1116 + }, + { + "epoch": 0.1988960113960114, + "grad_norm": 0.4172030985355377, + "learning_rate": 0.0001988233996046834, + "loss": 0.8098, + "step": 1117 + }, + { + "epoch": 0.19907407407407407, + "grad_norm": 0.4556557834148407, + "learning_rate": 0.00019882125772930867, + "loss": 0.9654, + "step": 1118 + }, + { + "epoch": 0.19925213675213677, + "grad_norm": 0.4363219141960144, + "learning_rate": 0.00019881911391773666, + "loss": 1.0333, + "step": 1119 + }, + { + "epoch": 0.19943019943019943, + "grad_norm": 0.4336536228656769, + "learning_rate": 0.0001988169681700094, + "loss": 1.091, + "step": 1120 + }, + { + "epoch": 0.1996082621082621, + "grad_norm": 0.42073166370391846, + "learning_rate": 0.00019881482048616893, + "loss": 0.9687, + "step": 1121 + }, + { + "epoch": 0.1997863247863248, + "grad_norm": 0.4330587685108185, + "learning_rate": 0.00019881267086625733, + "loss": 1.0512, + "step": 1122 + }, + { + "epoch": 0.19996438746438747, + "grad_norm": 0.4602276682853699, + "learning_rate": 0.0001988105193103167, + "loss": 1.1806, + "step": 1123 + }, + { + "epoch": 0.20014245014245013, + "grad_norm": 0.4271257817745209, + "learning_rate": 0.0001988083658183892, + "loss": 1.1079, + "step": 1124 + }, + { + "epoch": 0.20032051282051283, + "grad_norm": 0.35446426272392273, + "learning_rate": 0.00019880621039051707, + "loss": 0.6769, + "step": 1125 + }, + { + "epoch": 0.2004985754985755, + "grad_norm": 0.413753479719162, + "learning_rate": 0.00019880405302674244, + "loss": 1.1088, + "step": 1126 + }, + { + "epoch": 0.20067663817663817, + "grad_norm": 0.4423675835132599, + "learning_rate": 0.00019880189372710767, + "loss": 1.1371, + "step": 1127 + }, + { + "epoch": 0.20085470085470086, + "grad_norm": 0.41865605115890503, + "learning_rate": 0.00019879973249165502, + "loss": 1.0027, + "step": 1128 + }, + { + "epoch": 0.20103276353276353, + "grad_norm": 0.4109594225883484, + "learning_rate": 0.00019879756932042686, + "loss": 0.8734, + "step": 1129 + }, + { + "epoch": 0.2012108262108262, + "grad_norm": 0.42326363921165466, + "learning_rate": 0.00019879540421346555, + "loss": 0.9722, + "step": 1130 + }, + { + "epoch": 0.2013888888888889, + "grad_norm": 0.4601542055606842, + "learning_rate": 0.00019879323717081354, + "loss": 1.1251, + "step": 1131 + }, + { + "epoch": 0.20156695156695156, + "grad_norm": 0.4704367518424988, + "learning_rate": 0.00019879106819251327, + "loss": 0.9457, + "step": 1132 + }, + { + "epoch": 0.20174501424501423, + "grad_norm": 0.465023934841156, + "learning_rate": 0.00019878889727860724, + "loss": 0.9633, + "step": 1133 + }, + { + "epoch": 0.20192307692307693, + "grad_norm": 0.4572450518608093, + "learning_rate": 0.00019878672442913796, + "loss": 1.1965, + "step": 1134 + }, + { + "epoch": 0.2021011396011396, + "grad_norm": 0.4323410391807556, + "learning_rate": 0.00019878454964414807, + "loss": 1.1296, + "step": 1135 + }, + { + "epoch": 0.2022792022792023, + "grad_norm": 0.4513751268386841, + "learning_rate": 0.00019878237292368013, + "loss": 1.0571, + "step": 1136 + }, + { + "epoch": 0.20245726495726496, + "grad_norm": 0.45504096150398254, + "learning_rate": 0.00019878019426777677, + "loss": 1.0316, + "step": 1137 + }, + { + "epoch": 0.20263532763532763, + "grad_norm": 0.45715275406837463, + "learning_rate": 0.0001987780136764807, + "loss": 1.0528, + "step": 1138 + }, + { + "epoch": 0.20281339031339032, + "grad_norm": 0.4934465289115906, + "learning_rate": 0.00019877583114983466, + "loss": 1.3238, + "step": 1139 + }, + { + "epoch": 0.202991452991453, + "grad_norm": 0.4304082989692688, + "learning_rate": 0.0001987736466878814, + "loss": 1.1774, + "step": 1140 + }, + { + "epoch": 0.20316951566951566, + "grad_norm": 0.49721968173980713, + "learning_rate": 0.00019877146029066372, + "loss": 1.1767, + "step": 1141 + }, + { + "epoch": 0.20334757834757836, + "grad_norm": 0.3629468083381653, + "learning_rate": 0.00019876927195822445, + "loss": 0.8588, + "step": 1142 + }, + { + "epoch": 0.20352564102564102, + "grad_norm": 0.49310383200645447, + "learning_rate": 0.00019876708169060648, + "loss": 1.0588, + "step": 1143 + }, + { + "epoch": 0.2037037037037037, + "grad_norm": 0.4270328879356384, + "learning_rate": 0.00019876488948785271, + "loss": 1.1523, + "step": 1144 + }, + { + "epoch": 0.2038817663817664, + "grad_norm": 0.4559730887413025, + "learning_rate": 0.0001987626953500061, + "loss": 1.1736, + "step": 1145 + }, + { + "epoch": 0.20405982905982906, + "grad_norm": 0.5335259437561035, + "learning_rate": 0.00019876049927710962, + "loss": 0.991, + "step": 1146 + }, + { + "epoch": 0.20423789173789172, + "grad_norm": 0.43500083684921265, + "learning_rate": 0.0001987583012692063, + "loss": 1.0631, + "step": 1147 + }, + { + "epoch": 0.20441595441595442, + "grad_norm": 0.4135417938232422, + "learning_rate": 0.00019875610132633927, + "loss": 1.0896, + "step": 1148 + }, + { + "epoch": 0.2045940170940171, + "grad_norm": 0.4078896641731262, + "learning_rate": 0.00019875389944855153, + "loss": 1.0395, + "step": 1149 + }, + { + "epoch": 0.20477207977207978, + "grad_norm": 0.46612194180488586, + "learning_rate": 0.00019875169563588632, + "loss": 1.0541, + "step": 1150 + }, + { + "epoch": 0.20495014245014245, + "grad_norm": 0.5093224048614502, + "learning_rate": 0.00019874948988838674, + "loss": 1.1486, + "step": 1151 + }, + { + "epoch": 0.20512820512820512, + "grad_norm": 0.5079755187034607, + "learning_rate": 0.00019874728220609607, + "loss": 1.2614, + "step": 1152 + }, + { + "epoch": 0.20530626780626782, + "grad_norm": 0.43663498759269714, + "learning_rate": 0.0001987450725890575, + "loss": 1.0683, + "step": 1153 + }, + { + "epoch": 0.20548433048433049, + "grad_norm": 0.5029327273368835, + "learning_rate": 0.00019874286103731435, + "loss": 1.1934, + "step": 1154 + }, + { + "epoch": 0.20566239316239315, + "grad_norm": 0.48770397901535034, + "learning_rate": 0.00019874064755090999, + "loss": 1.1634, + "step": 1155 + }, + { + "epoch": 0.20584045584045585, + "grad_norm": 0.46826690435409546, + "learning_rate": 0.00019873843212988776, + "loss": 1.0621, + "step": 1156 + }, + { + "epoch": 0.20601851851851852, + "grad_norm": 0.4810047149658203, + "learning_rate": 0.00019873621477429105, + "loss": 1.0879, + "step": 1157 + }, + { + "epoch": 0.20619658119658119, + "grad_norm": 0.4769522249698639, + "learning_rate": 0.00019873399548416335, + "loss": 1.1365, + "step": 1158 + }, + { + "epoch": 0.20637464387464388, + "grad_norm": 0.4221782982349396, + "learning_rate": 0.00019873177425954806, + "loss": 1.1168, + "step": 1159 + }, + { + "epoch": 0.20655270655270655, + "grad_norm": 0.4084923565387726, + "learning_rate": 0.00019872955110048876, + "loss": 1.2364, + "step": 1160 + }, + { + "epoch": 0.20673076923076922, + "grad_norm": 0.4781704545021057, + "learning_rate": 0.00019872732600702904, + "loss": 1.19, + "step": 1161 + }, + { + "epoch": 0.20690883190883191, + "grad_norm": 0.3984242081642151, + "learning_rate": 0.0001987250989792124, + "loss": 1.0568, + "step": 1162 + }, + { + "epoch": 0.20708689458689458, + "grad_norm": 0.4601972997188568, + "learning_rate": 0.00019872287001708257, + "loss": 1.1625, + "step": 1163 + }, + { + "epoch": 0.20726495726495728, + "grad_norm": 0.4853581190109253, + "learning_rate": 0.00019872063912068316, + "loss": 1.2304, + "step": 1164 + }, + { + "epoch": 0.20744301994301995, + "grad_norm": 0.41779839992523193, + "learning_rate": 0.0001987184062900579, + "loss": 0.9807, + "step": 1165 + }, + { + "epoch": 0.20762108262108261, + "grad_norm": 0.4945356249809265, + "learning_rate": 0.00019871617152525056, + "loss": 1.1861, + "step": 1166 + }, + { + "epoch": 0.2077991452991453, + "grad_norm": 0.47432294487953186, + "learning_rate": 0.00019871393482630487, + "loss": 1.1448, + "step": 1167 + }, + { + "epoch": 0.20797720797720798, + "grad_norm": 0.44647398591041565, + "learning_rate": 0.00019871169619326473, + "loss": 1.096, + "step": 1168 + }, + { + "epoch": 0.20815527065527065, + "grad_norm": 0.4643072783946991, + "learning_rate": 0.00019870945562617393, + "loss": 1.1561, + "step": 1169 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 0.4544340968132019, + "learning_rate": 0.0001987072131250764, + "loss": 1.0764, + "step": 1170 + }, + { + "epoch": 0.208511396011396, + "grad_norm": 0.6036561727523804, + "learning_rate": 0.00019870496869001607, + "loss": 1.3961, + "step": 1171 + }, + { + "epoch": 0.20868945868945868, + "grad_norm": 0.41348758339881897, + "learning_rate": 0.00019870272232103695, + "loss": 1.2219, + "step": 1172 + }, + { + "epoch": 0.20886752136752137, + "grad_norm": 0.4184056222438812, + "learning_rate": 0.000198700474018183, + "loss": 1.1115, + "step": 1173 + }, + { + "epoch": 0.20904558404558404, + "grad_norm": 0.41920599341392517, + "learning_rate": 0.0001986982237814983, + "loss": 0.9207, + "step": 1174 + }, + { + "epoch": 0.2092236467236467, + "grad_norm": 0.4710249602794647, + "learning_rate": 0.00019869597161102694, + "loss": 1.1342, + "step": 1175 + }, + { + "epoch": 0.2094017094017094, + "grad_norm": 0.46897777915000916, + "learning_rate": 0.000198693717506813, + "loss": 0.983, + "step": 1176 + }, + { + "epoch": 0.20957977207977208, + "grad_norm": 0.4817039370536804, + "learning_rate": 0.00019869146146890074, + "loss": 1.0923, + "step": 1177 + }, + { + "epoch": 0.20975783475783477, + "grad_norm": 0.4806751012802124, + "learning_rate": 0.00019868920349733427, + "loss": 1.2296, + "step": 1178 + }, + { + "epoch": 0.20993589743589744, + "grad_norm": 0.44182994961738586, + "learning_rate": 0.0001986869435921579, + "loss": 1.1856, + "step": 1179 + }, + { + "epoch": 0.2101139601139601, + "grad_norm": 0.4282805621623993, + "learning_rate": 0.00019868468175341584, + "loss": 1.0046, + "step": 1180 + }, + { + "epoch": 0.2102920227920228, + "grad_norm": 0.5011838674545288, + "learning_rate": 0.00019868241798115242, + "loss": 1.2401, + "step": 1181 + }, + { + "epoch": 0.21047008547008547, + "grad_norm": 0.4282447397708893, + "learning_rate": 0.00019868015227541208, + "loss": 0.9338, + "step": 1182 + }, + { + "epoch": 0.21064814814814814, + "grad_norm": 0.4348810911178589, + "learning_rate": 0.00019867788463623912, + "loss": 0.926, + "step": 1183 + }, + { + "epoch": 0.21082621082621084, + "grad_norm": 0.41518425941467285, + "learning_rate": 0.00019867561506367799, + "loss": 1.2723, + "step": 1184 + }, + { + "epoch": 0.2110042735042735, + "grad_norm": 0.47346001863479614, + "learning_rate": 0.00019867334355777315, + "loss": 1.1931, + "step": 1185 + }, + { + "epoch": 0.21118233618233617, + "grad_norm": 0.4071715474128723, + "learning_rate": 0.00019867107011856914, + "loss": 0.9619, + "step": 1186 + }, + { + "epoch": 0.21136039886039887, + "grad_norm": 0.4803447425365448, + "learning_rate": 0.00019866879474611046, + "loss": 1.2, + "step": 1187 + }, + { + "epoch": 0.21153846153846154, + "grad_norm": 0.4827699661254883, + "learning_rate": 0.00019866651744044172, + "loss": 1.0938, + "step": 1188 + }, + { + "epoch": 0.2117165242165242, + "grad_norm": 0.4528424143791199, + "learning_rate": 0.00019866423820160756, + "loss": 0.9721, + "step": 1189 + }, + { + "epoch": 0.2118945868945869, + "grad_norm": 0.43566834926605225, + "learning_rate": 0.0001986619570296526, + "loss": 1.0352, + "step": 1190 + }, + { + "epoch": 0.21207264957264957, + "grad_norm": 0.4516540467739105, + "learning_rate": 0.0001986596739246215, + "loss": 1.1333, + "step": 1191 + }, + { + "epoch": 0.21225071225071226, + "grad_norm": 0.4456641376018524, + "learning_rate": 0.00019865738888655908, + "loss": 1.2813, + "step": 1192 + }, + { + "epoch": 0.21242877492877493, + "grad_norm": 0.47048309445381165, + "learning_rate": 0.00019865510191551008, + "loss": 1.1067, + "step": 1193 + }, + { + "epoch": 0.2126068376068376, + "grad_norm": 0.4604061543941498, + "learning_rate": 0.00019865281301151928, + "loss": 0.925, + "step": 1194 + }, + { + "epoch": 0.2127849002849003, + "grad_norm": 0.49341437220573425, + "learning_rate": 0.00019865052217463153, + "loss": 1.2319, + "step": 1195 + }, + { + "epoch": 0.21296296296296297, + "grad_norm": 0.5099014639854431, + "learning_rate": 0.00019864822940489173, + "loss": 1.139, + "step": 1196 + }, + { + "epoch": 0.21314102564102563, + "grad_norm": 0.41396936774253845, + "learning_rate": 0.0001986459347023448, + "loss": 1.0594, + "step": 1197 + }, + { + "epoch": 0.21331908831908833, + "grad_norm": 0.46071869134902954, + "learning_rate": 0.0001986436380670357, + "loss": 1.0815, + "step": 1198 + }, + { + "epoch": 0.213497150997151, + "grad_norm": 0.507882297039032, + "learning_rate": 0.00019864133949900942, + "loss": 1.3841, + "step": 1199 + }, + { + "epoch": 0.21367521367521367, + "grad_norm": 0.45680439472198486, + "learning_rate": 0.00019863903899831103, + "loss": 1.0945, + "step": 1200 + }, + { + "epoch": 0.21385327635327636, + "grad_norm": 0.44277429580688477, + "learning_rate": 0.00019863673656498555, + "loss": 1.1655, + "step": 1201 + }, + { + "epoch": 0.21403133903133903, + "grad_norm": 0.43890756368637085, + "learning_rate": 0.00019863443219907812, + "loss": 1.1186, + "step": 1202 + }, + { + "epoch": 0.2142094017094017, + "grad_norm": 0.3910178542137146, + "learning_rate": 0.0001986321259006339, + "loss": 1.0817, + "step": 1203 + }, + { + "epoch": 0.2143874643874644, + "grad_norm": 0.3803878128528595, + "learning_rate": 0.00019862981766969803, + "loss": 0.8022, + "step": 1204 + }, + { + "epoch": 0.21456552706552706, + "grad_norm": 0.4495108425617218, + "learning_rate": 0.0001986275075063158, + "loss": 1.2212, + "step": 1205 + }, + { + "epoch": 0.21474358974358973, + "grad_norm": 0.5211976766586304, + "learning_rate": 0.00019862519541053244, + "loss": 1.2771, + "step": 1206 + }, + { + "epoch": 0.21492165242165243, + "grad_norm": 0.4313061535358429, + "learning_rate": 0.00019862288138239325, + "loss": 1.1205, + "step": 1207 + }, + { + "epoch": 0.2150997150997151, + "grad_norm": 0.47110888361930847, + "learning_rate": 0.00019862056542194355, + "loss": 1.1835, + "step": 1208 + }, + { + "epoch": 0.2152777777777778, + "grad_norm": 0.5129403471946716, + "learning_rate": 0.00019861824752922876, + "loss": 1.1655, + "step": 1209 + }, + { + "epoch": 0.21545584045584046, + "grad_norm": 0.4353938102722168, + "learning_rate": 0.00019861592770429427, + "loss": 1.2794, + "step": 1210 + }, + { + "epoch": 0.21563390313390313, + "grad_norm": 0.48590636253356934, + "learning_rate": 0.0001986136059471855, + "loss": 1.2003, + "step": 1211 + }, + { + "epoch": 0.21581196581196582, + "grad_norm": 0.4738406836986542, + "learning_rate": 0.00019861128225794804, + "loss": 1.2271, + "step": 1212 + }, + { + "epoch": 0.2159900284900285, + "grad_norm": 0.45983126759529114, + "learning_rate": 0.0001986089566366273, + "loss": 1.1896, + "step": 1213 + }, + { + "epoch": 0.21616809116809116, + "grad_norm": 0.37296006083488464, + "learning_rate": 0.00019860662908326892, + "loss": 1.079, + "step": 1214 + }, + { + "epoch": 0.21634615384615385, + "grad_norm": 0.4442676305770874, + "learning_rate": 0.00019860429959791845, + "loss": 1.1754, + "step": 1215 + }, + { + "epoch": 0.21652421652421652, + "grad_norm": 0.4950128495693207, + "learning_rate": 0.0001986019681806216, + "loss": 1.1571, + "step": 1216 + }, + { + "epoch": 0.2167022792022792, + "grad_norm": 0.4374556541442871, + "learning_rate": 0.000198599634831424, + "loss": 1.1003, + "step": 1217 + }, + { + "epoch": 0.2168803418803419, + "grad_norm": 0.47301414608955383, + "learning_rate": 0.00019859729955037136, + "loss": 1.1426, + "step": 1218 + }, + { + "epoch": 0.21705840455840456, + "grad_norm": 0.41213178634643555, + "learning_rate": 0.00019859496233750947, + "loss": 1.0659, + "step": 1219 + }, + { + "epoch": 0.21723646723646722, + "grad_norm": 0.41601964831352234, + "learning_rate": 0.0001985926231928841, + "loss": 1.0248, + "step": 1220 + }, + { + "epoch": 0.21741452991452992, + "grad_norm": 0.46328839659690857, + "learning_rate": 0.0001985902821165411, + "loss": 1.0405, + "step": 1221 + }, + { + "epoch": 0.2175925925925926, + "grad_norm": 0.43287959694862366, + "learning_rate": 0.0001985879391085263, + "loss": 0.9202, + "step": 1222 + }, + { + "epoch": 0.21777065527065528, + "grad_norm": 0.4770444631576538, + "learning_rate": 0.00019858559416888568, + "loss": 1.0911, + "step": 1223 + }, + { + "epoch": 0.21794871794871795, + "grad_norm": 0.4756585955619812, + "learning_rate": 0.00019858324729766507, + "loss": 1.1566, + "step": 1224 + }, + { + "epoch": 0.21812678062678062, + "grad_norm": 0.4337233006954193, + "learning_rate": 0.00019858089849491054, + "loss": 0.9084, + "step": 1225 + }, + { + "epoch": 0.21830484330484332, + "grad_norm": 0.5165579319000244, + "learning_rate": 0.00019857854776066813, + "loss": 1.4154, + "step": 1226 + }, + { + "epoch": 0.21848290598290598, + "grad_norm": 0.4280378520488739, + "learning_rate": 0.00019857619509498382, + "loss": 1.1291, + "step": 1227 + }, + { + "epoch": 0.21866096866096865, + "grad_norm": 0.5375089049339294, + "learning_rate": 0.00019857384049790376, + "loss": 1.2985, + "step": 1228 + }, + { + "epoch": 0.21883903133903135, + "grad_norm": 0.4708811640739441, + "learning_rate": 0.00019857148396947401, + "loss": 1.0589, + "step": 1229 + }, + { + "epoch": 0.21901709401709402, + "grad_norm": 0.4744570255279541, + "learning_rate": 0.00019856912550974084, + "loss": 1.1269, + "step": 1230 + }, + { + "epoch": 0.21919515669515668, + "grad_norm": 0.5355265736579895, + "learning_rate": 0.00019856676511875043, + "loss": 1.1441, + "step": 1231 + }, + { + "epoch": 0.21937321937321938, + "grad_norm": 0.42718183994293213, + "learning_rate": 0.00019856440279654897, + "loss": 1.0244, + "step": 1232 + }, + { + "epoch": 0.21955128205128205, + "grad_norm": 0.5162127614021301, + "learning_rate": 0.00019856203854318283, + "loss": 1.2674, + "step": 1233 + }, + { + "epoch": 0.21972934472934472, + "grad_norm": 0.5180695652961731, + "learning_rate": 0.00019855967235869827, + "loss": 1.2472, + "step": 1234 + }, + { + "epoch": 0.2199074074074074, + "grad_norm": 0.4290023744106293, + "learning_rate": 0.00019855730424314167, + "loss": 1.0502, + "step": 1235 + }, + { + "epoch": 0.22008547008547008, + "grad_norm": 0.4418254792690277, + "learning_rate": 0.00019855493419655945, + "loss": 1.0589, + "step": 1236 + }, + { + "epoch": 0.22026353276353278, + "grad_norm": 0.4074663817882538, + "learning_rate": 0.000198552562218998, + "loss": 0.9197, + "step": 1237 + }, + { + "epoch": 0.22044159544159544, + "grad_norm": 0.4526660740375519, + "learning_rate": 0.00019855018831050383, + "loss": 1.2578, + "step": 1238 + }, + { + "epoch": 0.2206196581196581, + "grad_norm": 0.4747827649116516, + "learning_rate": 0.00019854781247112343, + "loss": 1.0841, + "step": 1239 + }, + { + "epoch": 0.2207977207977208, + "grad_norm": 0.41567128896713257, + "learning_rate": 0.00019854543470090334, + "loss": 1.0737, + "step": 1240 + }, + { + "epoch": 0.22097578347578348, + "grad_norm": 0.4793100953102112, + "learning_rate": 0.00019854305499989022, + "loss": 1.1972, + "step": 1241 + }, + { + "epoch": 0.22115384615384615, + "grad_norm": 0.41755473613739014, + "learning_rate": 0.00019854067336813058, + "loss": 1.2529, + "step": 1242 + }, + { + "epoch": 0.22133190883190884, + "grad_norm": 0.40421152114868164, + "learning_rate": 0.0001985382898056712, + "loss": 1.0549, + "step": 1243 + }, + { + "epoch": 0.2215099715099715, + "grad_norm": 0.45779645442962646, + "learning_rate": 0.0001985359043125587, + "loss": 1.1586, + "step": 1244 + }, + { + "epoch": 0.22168803418803418, + "grad_norm": 0.4380546808242798, + "learning_rate": 0.00019853351688883987, + "loss": 1.1024, + "step": 1245 + }, + { + "epoch": 0.22186609686609687, + "grad_norm": 0.39917269349098206, + "learning_rate": 0.00019853112753456142, + "loss": 0.9823, + "step": 1246 + }, + { + "epoch": 0.22204415954415954, + "grad_norm": 0.4228038489818573, + "learning_rate": 0.00019852873624977022, + "loss": 1.1684, + "step": 1247 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 0.4462146759033203, + "learning_rate": 0.00019852634303451315, + "loss": 0.9027, + "step": 1248 + }, + { + "epoch": 0.2224002849002849, + "grad_norm": 0.5682163834571838, + "learning_rate": 0.000198523947888837, + "loss": 1.141, + "step": 1249 + }, + { + "epoch": 0.22257834757834757, + "grad_norm": 0.44866830110549927, + "learning_rate": 0.0001985215508127888, + "loss": 1.0759, + "step": 1250 + }, + { + "epoch": 0.22275641025641027, + "grad_norm": 0.4034106135368347, + "learning_rate": 0.00019851915180641548, + "loss": 1.0675, + "step": 1251 + }, + { + "epoch": 0.22293447293447294, + "grad_norm": 0.4780726432800293, + "learning_rate": 0.00019851675086976397, + "loss": 1.0283, + "step": 1252 + }, + { + "epoch": 0.2231125356125356, + "grad_norm": 0.48892372846603394, + "learning_rate": 0.00019851434800288145, + "loss": 1.1159, + "step": 1253 + }, + { + "epoch": 0.2232905982905983, + "grad_norm": 0.42629215121269226, + "learning_rate": 0.0001985119432058149, + "loss": 1.0292, + "step": 1254 + }, + { + "epoch": 0.22346866096866097, + "grad_norm": 0.4496444761753082, + "learning_rate": 0.00019850953647861146, + "loss": 1.0252, + "step": 1255 + }, + { + "epoch": 0.22364672364672364, + "grad_norm": 0.4371408224105835, + "learning_rate": 0.00019850712782131828, + "loss": 1.1104, + "step": 1256 + }, + { + "epoch": 0.22382478632478633, + "grad_norm": 0.4910794496536255, + "learning_rate": 0.00019850471723398258, + "loss": 1.1928, + "step": 1257 + }, + { + "epoch": 0.224002849002849, + "grad_norm": 0.41235068440437317, + "learning_rate": 0.00019850230471665157, + "loss": 1.1261, + "step": 1258 + }, + { + "epoch": 0.22418091168091167, + "grad_norm": 0.4507700502872467, + "learning_rate": 0.0001984998902693725, + "loss": 1.0602, + "step": 1259 + }, + { + "epoch": 0.22435897435897437, + "grad_norm": 0.4654198884963989, + "learning_rate": 0.00019849747389219272, + "loss": 1.1258, + "step": 1260 + }, + { + "epoch": 0.22453703703703703, + "grad_norm": 0.439807653427124, + "learning_rate": 0.00019849505558515952, + "loss": 1.2312, + "step": 1261 + }, + { + "epoch": 0.2247150997150997, + "grad_norm": 0.4309258759021759, + "learning_rate": 0.00019849263534832035, + "loss": 1.0083, + "step": 1262 + }, + { + "epoch": 0.2248931623931624, + "grad_norm": 0.4920141100883484, + "learning_rate": 0.00019849021318172255, + "loss": 1.0254, + "step": 1263 + }, + { + "epoch": 0.22507122507122507, + "grad_norm": 0.5333457589149475, + "learning_rate": 0.00019848778908541367, + "loss": 1.3017, + "step": 1264 + }, + { + "epoch": 0.22524928774928774, + "grad_norm": 0.4096757769584656, + "learning_rate": 0.0001984853630594411, + "loss": 0.9531, + "step": 1265 + }, + { + "epoch": 0.22542735042735043, + "grad_norm": 0.5744075775146484, + "learning_rate": 0.00019848293510385244, + "loss": 1.1414, + "step": 1266 + }, + { + "epoch": 0.2256054131054131, + "grad_norm": 0.44707193970680237, + "learning_rate": 0.00019848050521869529, + "loss": 1.1926, + "step": 1267 + }, + { + "epoch": 0.2257834757834758, + "grad_norm": 0.4162999391555786, + "learning_rate": 0.00019847807340401716, + "loss": 1.1354, + "step": 1268 + }, + { + "epoch": 0.22596153846153846, + "grad_norm": 0.4273204207420349, + "learning_rate": 0.0001984756396598658, + "loss": 0.9956, + "step": 1269 + }, + { + "epoch": 0.22613960113960113, + "grad_norm": 0.5670466423034668, + "learning_rate": 0.00019847320398628878, + "loss": 1.2384, + "step": 1270 + }, + { + "epoch": 0.22631766381766383, + "grad_norm": 0.424544095993042, + "learning_rate": 0.00019847076638333395, + "loss": 0.9963, + "step": 1271 + }, + { + "epoch": 0.2264957264957265, + "grad_norm": 0.3716120719909668, + "learning_rate": 0.000198468326851049, + "loss": 0.865, + "step": 1272 + }, + { + "epoch": 0.22667378917378916, + "grad_norm": 0.4472847282886505, + "learning_rate": 0.00019846588538948172, + "loss": 1.174, + "step": 1273 + }, + { + "epoch": 0.22685185185185186, + "grad_norm": 0.4599195718765259, + "learning_rate": 0.00019846344199867994, + "loss": 1.289, + "step": 1274 + }, + { + "epoch": 0.22702991452991453, + "grad_norm": 0.4303213357925415, + "learning_rate": 0.0001984609966786916, + "loss": 1.1606, + "step": 1275 + }, + { + "epoch": 0.2272079772079772, + "grad_norm": 0.44893527030944824, + "learning_rate": 0.00019845854942956455, + "loss": 1.1043, + "step": 1276 + }, + { + "epoch": 0.2273860398860399, + "grad_norm": 0.40033379197120667, + "learning_rate": 0.00019845610025134676, + "loss": 1.1434, + "step": 1277 + }, + { + "epoch": 0.22756410256410256, + "grad_norm": 0.4385402202606201, + "learning_rate": 0.00019845364914408616, + "loss": 0.9943, + "step": 1278 + }, + { + "epoch": 0.22774216524216523, + "grad_norm": 0.42123618721961975, + "learning_rate": 0.0001984511961078309, + "loss": 1.0911, + "step": 1279 + }, + { + "epoch": 0.22792022792022792, + "grad_norm": 0.5558577179908752, + "learning_rate": 0.00019844874114262893, + "loss": 1.3893, + "step": 1280 + }, + { + "epoch": 0.2280982905982906, + "grad_norm": 0.3996453583240509, + "learning_rate": 0.00019844628424852835, + "loss": 0.8951, + "step": 1281 + }, + { + "epoch": 0.2282763532763533, + "grad_norm": 0.3943425714969635, + "learning_rate": 0.0001984438254255774, + "loss": 1.0595, + "step": 1282 + }, + { + "epoch": 0.22845441595441596, + "grad_norm": 0.4429021179676056, + "learning_rate": 0.00019844136467382414, + "loss": 1.0853, + "step": 1283 + }, + { + "epoch": 0.22863247863247863, + "grad_norm": 0.4515686631202698, + "learning_rate": 0.00019843890199331687, + "loss": 1.0829, + "step": 1284 + }, + { + "epoch": 0.22881054131054132, + "grad_norm": 0.5157768726348877, + "learning_rate": 0.00019843643738410378, + "loss": 1.334, + "step": 1285 + }, + { + "epoch": 0.228988603988604, + "grad_norm": 0.45833173394203186, + "learning_rate": 0.0001984339708462332, + "loss": 1.1353, + "step": 1286 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 0.46610337495803833, + "learning_rate": 0.00019843150237975344, + "loss": 1.1338, + "step": 1287 + }, + { + "epoch": 0.22934472934472935, + "grad_norm": 0.5076978802680969, + "learning_rate": 0.00019842903198471286, + "loss": 1.1811, + "step": 1288 + }, + { + "epoch": 0.22952279202279202, + "grad_norm": 0.4297824800014496, + "learning_rate": 0.00019842655966115986, + "loss": 1.1799, + "step": 1289 + }, + { + "epoch": 0.2297008547008547, + "grad_norm": 0.5304586291313171, + "learning_rate": 0.0001984240854091429, + "loss": 1.1315, + "step": 1290 + }, + { + "epoch": 0.22987891737891739, + "grad_norm": 0.45359212160110474, + "learning_rate": 0.00019842160922871042, + "loss": 1.1037, + "step": 1291 + }, + { + "epoch": 0.23005698005698005, + "grad_norm": 0.4416881203651428, + "learning_rate": 0.00019841913111991096, + "loss": 1.122, + "step": 1292 + }, + { + "epoch": 0.23023504273504272, + "grad_norm": 0.46682995557785034, + "learning_rate": 0.0001984166510827931, + "loss": 0.9808, + "step": 1293 + }, + { + "epoch": 0.23041310541310542, + "grad_norm": 0.44172337651252747, + "learning_rate": 0.00019841416911740538, + "loss": 0.9167, + "step": 1294 + }, + { + "epoch": 0.23059116809116809, + "grad_norm": 0.40562742948532104, + "learning_rate": 0.0001984116852237965, + "loss": 0.9547, + "step": 1295 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.4040384888648987, + "learning_rate": 0.00019840919940201503, + "loss": 1.1039, + "step": 1296 + }, + { + "epoch": 0.23094729344729345, + "grad_norm": 0.5094077587127686, + "learning_rate": 0.00019840671165210973, + "loss": 1.2283, + "step": 1297 + }, + { + "epoch": 0.23112535612535612, + "grad_norm": 0.48553213477134705, + "learning_rate": 0.00019840422197412938, + "loss": 1.0927, + "step": 1298 + }, + { + "epoch": 0.23130341880341881, + "grad_norm": 0.5197509527206421, + "learning_rate": 0.00019840173036812266, + "loss": 1.2154, + "step": 1299 + }, + { + "epoch": 0.23148148148148148, + "grad_norm": 0.42069005966186523, + "learning_rate": 0.0001983992368341385, + "loss": 1.0076, + "step": 1300 + }, + { + "epoch": 0.23165954415954415, + "grad_norm": 0.475204735994339, + "learning_rate": 0.00019839674137222567, + "loss": 1.1682, + "step": 1301 + }, + { + "epoch": 0.23183760683760685, + "grad_norm": 0.55730140209198, + "learning_rate": 0.0001983942439824331, + "loss": 1.2948, + "step": 1302 + }, + { + "epoch": 0.23201566951566951, + "grad_norm": 0.4533313512802124, + "learning_rate": 0.00019839174466480973, + "loss": 1.2691, + "step": 1303 + }, + { + "epoch": 0.23219373219373218, + "grad_norm": 0.4733520746231079, + "learning_rate": 0.0001983892434194045, + "loss": 1.2232, + "step": 1304 + }, + { + "epoch": 0.23237179487179488, + "grad_norm": 0.5085756182670593, + "learning_rate": 0.00019838674024626643, + "loss": 1.1347, + "step": 1305 + }, + { + "epoch": 0.23254985754985755, + "grad_norm": 0.4679976999759674, + "learning_rate": 0.00019838423514544456, + "loss": 1.0018, + "step": 1306 + }, + { + "epoch": 0.23272792022792022, + "grad_norm": 0.4234481751918793, + "learning_rate": 0.00019838172811698795, + "loss": 1.0472, + "step": 1307 + }, + { + "epoch": 0.2329059829059829, + "grad_norm": 0.5749204158782959, + "learning_rate": 0.00019837921916094579, + "loss": 1.2239, + "step": 1308 + }, + { + "epoch": 0.23308404558404558, + "grad_norm": 0.46715882420539856, + "learning_rate": 0.0001983767082773672, + "loss": 1.1924, + "step": 1309 + }, + { + "epoch": 0.23326210826210828, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.00019837419546630137, + "loss": 1.1086, + "step": 1310 + }, + { + "epoch": 0.23344017094017094, + "grad_norm": 0.4419243037700653, + "learning_rate": 0.0001983716807277975, + "loss": 1.1911, + "step": 1311 + }, + { + "epoch": 0.2336182336182336, + "grad_norm": 0.5107570290565491, + "learning_rate": 0.00019836916406190493, + "loss": 1.1071, + "step": 1312 + }, + { + "epoch": 0.2337962962962963, + "grad_norm": 0.5295659303665161, + "learning_rate": 0.00019836664546867293, + "loss": 1.2905, + "step": 1313 + }, + { + "epoch": 0.23397435897435898, + "grad_norm": 0.4844837784767151, + "learning_rate": 0.00019836412494815084, + "loss": 1.3507, + "step": 1314 + }, + { + "epoch": 0.23415242165242164, + "grad_norm": 0.6166049242019653, + "learning_rate": 0.00019836160250038808, + "loss": 1.2822, + "step": 1315 + }, + { + "epoch": 0.23433048433048434, + "grad_norm": 0.3229198753833771, + "learning_rate": 0.00019835907812543402, + "loss": 0.4959, + "step": 1316 + }, + { + "epoch": 0.234508547008547, + "grad_norm": 0.5788772702217102, + "learning_rate": 0.00019835655182333815, + "loss": 1.0832, + "step": 1317 + }, + { + "epoch": 0.23468660968660968, + "grad_norm": 0.525705099105835, + "learning_rate": 0.00019835402359414997, + "loss": 1.0968, + "step": 1318 + }, + { + "epoch": 0.23486467236467237, + "grad_norm": 0.5007779002189636, + "learning_rate": 0.000198351493437919, + "loss": 1.2788, + "step": 1319 + }, + { + "epoch": 0.23504273504273504, + "grad_norm": 0.4276871383190155, + "learning_rate": 0.00019834896135469484, + "loss": 1.0419, + "step": 1320 + }, + { + "epoch": 0.2352207977207977, + "grad_norm": 0.5359070301055908, + "learning_rate": 0.00019834642734452708, + "loss": 1.1308, + "step": 1321 + }, + { + "epoch": 0.2353988603988604, + "grad_norm": 0.4854908883571625, + "learning_rate": 0.0001983438914074654, + "loss": 1.1211, + "step": 1322 + }, + { + "epoch": 0.23557692307692307, + "grad_norm": 0.4913707375526428, + "learning_rate": 0.0001983413535435594, + "loss": 1.2392, + "step": 1323 + }, + { + "epoch": 0.23575498575498577, + "grad_norm": 0.46755748987197876, + "learning_rate": 0.0001983388137528589, + "loss": 0.9348, + "step": 1324 + }, + { + "epoch": 0.23593304843304844, + "grad_norm": 0.4592570960521698, + "learning_rate": 0.0001983362720354136, + "loss": 1.1339, + "step": 1325 + }, + { + "epoch": 0.2361111111111111, + "grad_norm": 0.5121711492538452, + "learning_rate": 0.00019833372839127335, + "loss": 1.2973, + "step": 1326 + }, + { + "epoch": 0.2362891737891738, + "grad_norm": 0.4809017479419708, + "learning_rate": 0.000198331182820488, + "loss": 0.9849, + "step": 1327 + }, + { + "epoch": 0.23646723646723647, + "grad_norm": 0.42340895533561707, + "learning_rate": 0.00019832863532310733, + "loss": 1.0731, + "step": 1328 + }, + { + "epoch": 0.23664529914529914, + "grad_norm": 0.5388045310974121, + "learning_rate": 0.00019832608589918135, + "loss": 1.0729, + "step": 1329 + }, + { + "epoch": 0.23682336182336183, + "grad_norm": 0.43075770139694214, + "learning_rate": 0.00019832353454875992, + "loss": 1.1684, + "step": 1330 + }, + { + "epoch": 0.2370014245014245, + "grad_norm": 0.554927408695221, + "learning_rate": 0.00019832098127189313, + "loss": 1.0842, + "step": 1331 + }, + { + "epoch": 0.23717948717948717, + "grad_norm": 0.5359260439872742, + "learning_rate": 0.0001983184260686309, + "loss": 1.2399, + "step": 1332 + }, + { + "epoch": 0.23735754985754987, + "grad_norm": 0.5141251087188721, + "learning_rate": 0.0001983158689390234, + "loss": 1.3752, + "step": 1333 + }, + { + "epoch": 0.23753561253561253, + "grad_norm": 0.4578750431537628, + "learning_rate": 0.00019831330988312067, + "loss": 1.0965, + "step": 1334 + }, + { + "epoch": 0.2377136752136752, + "grad_norm": 0.47974497079849243, + "learning_rate": 0.00019831074890097286, + "loss": 1.3379, + "step": 1335 + }, + { + "epoch": 0.2378917378917379, + "grad_norm": 0.4618176817893982, + "learning_rate": 0.00019830818599263014, + "loss": 1.274, + "step": 1336 + }, + { + "epoch": 0.23806980056980057, + "grad_norm": 0.4279816448688507, + "learning_rate": 0.00019830562115814276, + "loss": 0.996, + "step": 1337 + }, + { + "epoch": 0.23824786324786323, + "grad_norm": 0.4255026876926422, + "learning_rate": 0.0001983030543975609, + "loss": 0.969, + "step": 1338 + }, + { + "epoch": 0.23842592592592593, + "grad_norm": 0.4551412761211395, + "learning_rate": 0.00019830048571093493, + "loss": 1.0204, + "step": 1339 + }, + { + "epoch": 0.2386039886039886, + "grad_norm": 0.4747903048992157, + "learning_rate": 0.00019829791509831513, + "loss": 1.1816, + "step": 1340 + }, + { + "epoch": 0.2387820512820513, + "grad_norm": 0.47187140583992004, + "learning_rate": 0.00019829534255975188, + "loss": 1.1205, + "step": 1341 + }, + { + "epoch": 0.23896011396011396, + "grad_norm": 0.49332180619239807, + "learning_rate": 0.0001982927680952956, + "loss": 1.2657, + "step": 1342 + }, + { + "epoch": 0.23913817663817663, + "grad_norm": 0.5162837505340576, + "learning_rate": 0.0001982901917049967, + "loss": 1.2247, + "step": 1343 + }, + { + "epoch": 0.23931623931623933, + "grad_norm": 0.43407055735588074, + "learning_rate": 0.0001982876133889057, + "loss": 1.0038, + "step": 1344 + }, + { + "epoch": 0.239494301994302, + "grad_norm": 0.5132251977920532, + "learning_rate": 0.00019828503314707306, + "loss": 1.0678, + "step": 1345 + }, + { + "epoch": 0.23967236467236466, + "grad_norm": 0.46295464038848877, + "learning_rate": 0.00019828245097954937, + "loss": 1.1802, + "step": 1346 + }, + { + "epoch": 0.23985042735042736, + "grad_norm": 0.4682658314704895, + "learning_rate": 0.00019827986688638523, + "loss": 1.0249, + "step": 1347 + }, + { + "epoch": 0.24002849002849003, + "grad_norm": 0.49990561604499817, + "learning_rate": 0.00019827728086763125, + "loss": 1.0691, + "step": 1348 + }, + { + "epoch": 0.2402065527065527, + "grad_norm": 0.39090847969055176, + "learning_rate": 0.00019827469292333806, + "loss": 0.8367, + "step": 1349 + }, + { + "epoch": 0.2403846153846154, + "grad_norm": 0.5023905634880066, + "learning_rate": 0.00019827210305355645, + "loss": 1.0675, + "step": 1350 + }, + { + "epoch": 0.24056267806267806, + "grad_norm": 0.4744076430797577, + "learning_rate": 0.00019826951125833715, + "loss": 1.3166, + "step": 1351 + }, + { + "epoch": 0.24074074074074073, + "grad_norm": 0.44914689660072327, + "learning_rate": 0.00019826691753773088, + "loss": 0.9818, + "step": 1352 + }, + { + "epoch": 0.24091880341880342, + "grad_norm": 0.44391971826553345, + "learning_rate": 0.00019826432189178853, + "loss": 1.0448, + "step": 1353 + }, + { + "epoch": 0.2410968660968661, + "grad_norm": 0.46102839708328247, + "learning_rate": 0.00019826172432056086, + "loss": 0.9952, + "step": 1354 + }, + { + "epoch": 0.2412749287749288, + "grad_norm": 0.4796878695487976, + "learning_rate": 0.00019825912482409884, + "loss": 1.0977, + "step": 1355 + }, + { + "epoch": 0.24145299145299146, + "grad_norm": 0.5003768801689148, + "learning_rate": 0.0001982565234024534, + "loss": 1.3149, + "step": 1356 + }, + { + "epoch": 0.24163105413105412, + "grad_norm": 0.43475663661956787, + "learning_rate": 0.00019825392005567551, + "loss": 1.0527, + "step": 1357 + }, + { + "epoch": 0.24180911680911682, + "grad_norm": 0.46120527386665344, + "learning_rate": 0.00019825131478381613, + "loss": 1.2333, + "step": 1358 + }, + { + "epoch": 0.2419871794871795, + "grad_norm": 0.43748101592063904, + "learning_rate": 0.00019824870758692638, + "loss": 0.9788, + "step": 1359 + }, + { + "epoch": 0.24216524216524216, + "grad_norm": 0.5275192856788635, + "learning_rate": 0.00019824609846505727, + "loss": 1.1473, + "step": 1360 + }, + { + "epoch": 0.24234330484330485, + "grad_norm": 0.346463143825531, + "learning_rate": 0.00019824348741825993, + "loss": 0.6824, + "step": 1361 + }, + { + "epoch": 0.24252136752136752, + "grad_norm": 0.5004115700721741, + "learning_rate": 0.00019824087444658556, + "loss": 1.1853, + "step": 1362 + }, + { + "epoch": 0.2426994301994302, + "grad_norm": 0.42746666073799133, + "learning_rate": 0.00019823825955008533, + "loss": 0.9355, + "step": 1363 + }, + { + "epoch": 0.24287749287749288, + "grad_norm": 0.4099743068218231, + "learning_rate": 0.00019823564272881047, + "loss": 1.0753, + "step": 1364 + }, + { + "epoch": 0.24305555555555555, + "grad_norm": 0.5262967944145203, + "learning_rate": 0.00019823302398281226, + "loss": 1.2324, + "step": 1365 + }, + { + "epoch": 0.24323361823361822, + "grad_norm": 0.436069518327713, + "learning_rate": 0.000198230403312142, + "loss": 1.1887, + "step": 1366 + }, + { + "epoch": 0.24341168091168092, + "grad_norm": 0.38252368569374084, + "learning_rate": 0.00019822778071685107, + "loss": 1.0211, + "step": 1367 + }, + { + "epoch": 0.24358974358974358, + "grad_norm": 0.48024141788482666, + "learning_rate": 0.00019822515619699081, + "loss": 1.065, + "step": 1368 + }, + { + "epoch": 0.24376780626780628, + "grad_norm": 0.47421589493751526, + "learning_rate": 0.00019822252975261267, + "loss": 1.0433, + "step": 1369 + }, + { + "epoch": 0.24394586894586895, + "grad_norm": 0.46094807982444763, + "learning_rate": 0.00019821990138376808, + "loss": 1.1427, + "step": 1370 + }, + { + "epoch": 0.24412393162393162, + "grad_norm": 0.5093680620193481, + "learning_rate": 0.00019821727109050856, + "loss": 1.1086, + "step": 1371 + }, + { + "epoch": 0.2443019943019943, + "grad_norm": 0.41084879636764526, + "learning_rate": 0.00019821463887288566, + "loss": 1.0068, + "step": 1372 + }, + { + "epoch": 0.24448005698005698, + "grad_norm": 0.4991084635257721, + "learning_rate": 0.0001982120047309509, + "loss": 1.1884, + "step": 1373 + }, + { + "epoch": 0.24465811965811965, + "grad_norm": 0.39198383688926697, + "learning_rate": 0.00019820936866475595, + "loss": 0.9776, + "step": 1374 + }, + { + "epoch": 0.24483618233618235, + "grad_norm": 0.4517424702644348, + "learning_rate": 0.00019820673067435244, + "loss": 1.1491, + "step": 1375 + }, + { + "epoch": 0.245014245014245, + "grad_norm": 0.45881983637809753, + "learning_rate": 0.00019820409075979202, + "loss": 1.1198, + "step": 1376 + }, + { + "epoch": 0.24519230769230768, + "grad_norm": 0.4498792290687561, + "learning_rate": 0.00019820144892112646, + "loss": 1.0897, + "step": 1377 + }, + { + "epoch": 0.24537037037037038, + "grad_norm": 0.4128037393093109, + "learning_rate": 0.00019819880515840752, + "loss": 0.9415, + "step": 1378 + }, + { + "epoch": 0.24554843304843305, + "grad_norm": 0.4340885281562805, + "learning_rate": 0.00019819615947168698, + "loss": 1.201, + "step": 1379 + }, + { + "epoch": 0.24572649572649571, + "grad_norm": 0.43814027309417725, + "learning_rate": 0.00019819351186101667, + "loss": 1.1039, + "step": 1380 + }, + { + "epoch": 0.2459045584045584, + "grad_norm": 0.40115082263946533, + "learning_rate": 0.00019819086232644845, + "loss": 1.2599, + "step": 1381 + }, + { + "epoch": 0.24608262108262108, + "grad_norm": 0.4947351813316345, + "learning_rate": 0.00019818821086803426, + "loss": 1.252, + "step": 1382 + }, + { + "epoch": 0.24626068376068377, + "grad_norm": 0.45179441571235657, + "learning_rate": 0.0001981855574858261, + "loss": 1.1323, + "step": 1383 + }, + { + "epoch": 0.24643874643874644, + "grad_norm": 0.47159844636917114, + "learning_rate": 0.00019818290217987587, + "loss": 1.2053, + "step": 1384 + }, + { + "epoch": 0.2466168091168091, + "grad_norm": 0.4358448386192322, + "learning_rate": 0.0001981802449502356, + "loss": 1.1174, + "step": 1385 + }, + { + "epoch": 0.2467948717948718, + "grad_norm": 0.4588233530521393, + "learning_rate": 0.00019817758579695745, + "loss": 1.1098, + "step": 1386 + }, + { + "epoch": 0.24697293447293447, + "grad_norm": 0.4955112636089325, + "learning_rate": 0.00019817492472009338, + "loss": 1.258, + "step": 1387 + }, + { + "epoch": 0.24715099715099714, + "grad_norm": 0.4226941764354706, + "learning_rate": 0.00019817226171969565, + "loss": 1.0976, + "step": 1388 + }, + { + "epoch": 0.24732905982905984, + "grad_norm": 0.4076840579509735, + "learning_rate": 0.00019816959679581637, + "loss": 1.0121, + "step": 1389 + }, + { + "epoch": 0.2475071225071225, + "grad_norm": 0.4395063519477844, + "learning_rate": 0.0001981669299485078, + "loss": 1.3153, + "step": 1390 + }, + { + "epoch": 0.24768518518518517, + "grad_norm": 0.41010400652885437, + "learning_rate": 0.0001981642611778221, + "loss": 1.0717, + "step": 1391 + }, + { + "epoch": 0.24786324786324787, + "grad_norm": 0.43459352850914, + "learning_rate": 0.00019816159048381167, + "loss": 1.1077, + "step": 1392 + }, + { + "epoch": 0.24804131054131054, + "grad_norm": 0.46291449666023254, + "learning_rate": 0.00019815891786652875, + "loss": 1.0257, + "step": 1393 + }, + { + "epoch": 0.2482193732193732, + "grad_norm": 0.46408146619796753, + "learning_rate": 0.00019815624332602578, + "loss": 0.7899, + "step": 1394 + }, + { + "epoch": 0.2483974358974359, + "grad_norm": 0.4763357937335968, + "learning_rate": 0.00019815356686235508, + "loss": 0.9857, + "step": 1395 + }, + { + "epoch": 0.24857549857549857, + "grad_norm": 0.4766457676887512, + "learning_rate": 0.00019815088847556918, + "loss": 1.0589, + "step": 1396 + }, + { + "epoch": 0.24875356125356127, + "grad_norm": 0.4486583173274994, + "learning_rate": 0.0001981482081657205, + "loss": 1.2572, + "step": 1397 + }, + { + "epoch": 0.24893162393162394, + "grad_norm": 0.468878835439682, + "learning_rate": 0.00019814552593286155, + "loss": 1.101, + "step": 1398 + }, + { + "epoch": 0.2491096866096866, + "grad_norm": 0.4230278730392456, + "learning_rate": 0.0001981428417770449, + "loss": 0.9457, + "step": 1399 + }, + { + "epoch": 0.2492877492877493, + "grad_norm": 0.45630761981010437, + "learning_rate": 0.00019814015569832315, + "loss": 1.0665, + "step": 1400 + }, + { + "epoch": 0.24946581196581197, + "grad_norm": 0.5780113935470581, + "learning_rate": 0.00019813746769674893, + "loss": 1.1064, + "step": 1401 + }, + { + "epoch": 0.24964387464387464, + "grad_norm": 0.4343436658382416, + "learning_rate": 0.0001981347777723749, + "loss": 1.1132, + "step": 1402 + }, + { + "epoch": 0.24982193732193733, + "grad_norm": 0.4879056513309479, + "learning_rate": 0.0001981320859252537, + "loss": 1.1301, + "step": 1403 + }, + { + "epoch": 0.25, + "grad_norm": 0.5248328447341919, + "learning_rate": 0.00019812939215543818, + "loss": 1.1468, + "step": 1404 + }, + { + "epoch": 0.25, + "eval_loss": 1.115895390510559, + "eval_runtime": 25.0474, + "eval_samples_per_second": 41.561, + "eval_steps_per_second": 20.801, + "step": 1404 + }, + { + "epoch": 0.2501780626780627, + "grad_norm": 0.5076769590377808, + "learning_rate": 0.00019812669646298106, + "loss": 1.1428, + "step": 1405 + }, + { + "epoch": 0.25035612535612534, + "grad_norm": 0.5510252714157104, + "learning_rate": 0.00019812399884793514, + "loss": 1.3383, + "step": 1406 + }, + { + "epoch": 0.25053418803418803, + "grad_norm": 0.48918986320495605, + "learning_rate": 0.0001981212993103533, + "loss": 1.1507, + "step": 1407 + }, + { + "epoch": 0.25071225071225073, + "grad_norm": 0.4678935110569, + "learning_rate": 0.00019811859785028846, + "loss": 1.13, + "step": 1408 + }, + { + "epoch": 0.25089031339031337, + "grad_norm": 0.5155254602432251, + "learning_rate": 0.0001981158944677935, + "loss": 1.1194, + "step": 1409 + }, + { + "epoch": 0.25106837606837606, + "grad_norm": 0.4533839523792267, + "learning_rate": 0.00019811318916292142, + "loss": 0.9464, + "step": 1410 + }, + { + "epoch": 0.25124643874643876, + "grad_norm": 0.5142433047294617, + "learning_rate": 0.00019811048193572517, + "loss": 1.0837, + "step": 1411 + }, + { + "epoch": 0.2514245014245014, + "grad_norm": 0.4330446124076843, + "learning_rate": 0.00019810777278625788, + "loss": 0.9117, + "step": 1412 + }, + { + "epoch": 0.2516025641025641, + "grad_norm": 0.44806256890296936, + "learning_rate": 0.00019810506171457254, + "loss": 1.1643, + "step": 1413 + }, + { + "epoch": 0.2517806267806268, + "grad_norm": 0.43526285886764526, + "learning_rate": 0.00019810234872072235, + "loss": 0.9776, + "step": 1414 + }, + { + "epoch": 0.25195868945868943, + "grad_norm": 0.47394511103630066, + "learning_rate": 0.00019809963380476039, + "loss": 1.0935, + "step": 1415 + }, + { + "epoch": 0.25213675213675213, + "grad_norm": 0.48961278796195984, + "learning_rate": 0.00019809691696673993, + "loss": 1.179, + "step": 1416 + }, + { + "epoch": 0.2523148148148148, + "grad_norm": 0.43153589963912964, + "learning_rate": 0.00019809419820671412, + "loss": 0.906, + "step": 1417 + }, + { + "epoch": 0.25249287749287747, + "grad_norm": 0.41187527775764465, + "learning_rate": 0.00019809147752473632, + "loss": 0.899, + "step": 1418 + }, + { + "epoch": 0.25267094017094016, + "grad_norm": 0.5003183484077454, + "learning_rate": 0.00019808875492085973, + "loss": 1.0606, + "step": 1419 + }, + { + "epoch": 0.25284900284900286, + "grad_norm": 0.4430316984653473, + "learning_rate": 0.00019808603039513778, + "loss": 0.9167, + "step": 1420 + }, + { + "epoch": 0.25302706552706555, + "grad_norm": 0.4577699601650238, + "learning_rate": 0.00019808330394762382, + "loss": 1.1184, + "step": 1421 + }, + { + "epoch": 0.2532051282051282, + "grad_norm": 0.42656826972961426, + "learning_rate": 0.0001980805755783713, + "loss": 0.9335, + "step": 1422 + }, + { + "epoch": 0.2533831908831909, + "grad_norm": 0.40980881452560425, + "learning_rate": 0.0001980778452874336, + "loss": 0.9756, + "step": 1423 + }, + { + "epoch": 0.2535612535612536, + "grad_norm": 0.5752090811729431, + "learning_rate": 0.00019807511307486423, + "loss": 1.1694, + "step": 1424 + }, + { + "epoch": 0.2537393162393162, + "grad_norm": 0.5000349283218384, + "learning_rate": 0.00019807237894071681, + "loss": 0.9515, + "step": 1425 + }, + { + "epoch": 0.2539173789173789, + "grad_norm": 0.5159069299697876, + "learning_rate": 0.00019806964288504483, + "loss": 1.4014, + "step": 1426 + }, + { + "epoch": 0.2540954415954416, + "grad_norm": 0.5377941131591797, + "learning_rate": 0.00019806690490790194, + "loss": 1.2832, + "step": 1427 + }, + { + "epoch": 0.25427350427350426, + "grad_norm": 0.4565938711166382, + "learning_rate": 0.00019806416500934174, + "loss": 1.0629, + "step": 1428 + }, + { + "epoch": 0.25445156695156695, + "grad_norm": 0.49867144227027893, + "learning_rate": 0.00019806142318941797, + "loss": 1.2011, + "step": 1429 + }, + { + "epoch": 0.25462962962962965, + "grad_norm": 0.5111994743347168, + "learning_rate": 0.00019805867944818427, + "loss": 0.8925, + "step": 1430 + }, + { + "epoch": 0.2548076923076923, + "grad_norm": 0.5204268097877502, + "learning_rate": 0.00019805593378569448, + "loss": 1.2956, + "step": 1431 + }, + { + "epoch": 0.254985754985755, + "grad_norm": 0.3889026939868927, + "learning_rate": 0.00019805318620200234, + "loss": 1.0355, + "step": 1432 + }, + { + "epoch": 0.2551638176638177, + "grad_norm": 0.46825656294822693, + "learning_rate": 0.00019805043669716174, + "loss": 1.0444, + "step": 1433 + }, + { + "epoch": 0.2553418803418803, + "grad_norm": 0.4509420394897461, + "learning_rate": 0.00019804768527122648, + "loss": 1.0423, + "step": 1434 + }, + { + "epoch": 0.255519943019943, + "grad_norm": 0.4514774978160858, + "learning_rate": 0.0001980449319242505, + "loss": 1.1588, + "step": 1435 + }, + { + "epoch": 0.2556980056980057, + "grad_norm": 0.43019044399261475, + "learning_rate": 0.0001980421766562878, + "loss": 0.9939, + "step": 1436 + }, + { + "epoch": 0.25587606837606836, + "grad_norm": 0.5056091547012329, + "learning_rate": 0.00019803941946739228, + "loss": 1.1238, + "step": 1437 + }, + { + "epoch": 0.25605413105413105, + "grad_norm": 0.48664605617523193, + "learning_rate": 0.000198036660357618, + "loss": 1.0702, + "step": 1438 + }, + { + "epoch": 0.25623219373219375, + "grad_norm": 0.4500972032546997, + "learning_rate": 0.000198033899327019, + "loss": 0.9365, + "step": 1439 + }, + { + "epoch": 0.2564102564102564, + "grad_norm": 0.4800589382648468, + "learning_rate": 0.0001980311363756494, + "loss": 1.1159, + "step": 1440 + }, + { + "epoch": 0.2565883190883191, + "grad_norm": 0.3486495316028595, + "learning_rate": 0.0001980283715035633, + "loss": 0.6029, + "step": 1441 + }, + { + "epoch": 0.2567663817663818, + "grad_norm": 0.46258702874183655, + "learning_rate": 0.00019802560471081493, + "loss": 1.025, + "step": 1442 + }, + { + "epoch": 0.2569444444444444, + "grad_norm": 0.4846673607826233, + "learning_rate": 0.00019802283599745844, + "loss": 1.1105, + "step": 1443 + }, + { + "epoch": 0.2571225071225071, + "grad_norm": 0.4586990475654602, + "learning_rate": 0.00019802006536354813, + "loss": 0.9897, + "step": 1444 + }, + { + "epoch": 0.2573005698005698, + "grad_norm": 0.5177786350250244, + "learning_rate": 0.00019801729280913825, + "loss": 1.2558, + "step": 1445 + }, + { + "epoch": 0.25747863247863245, + "grad_norm": 0.43213751912117004, + "learning_rate": 0.00019801451833428312, + "loss": 1.0961, + "step": 1446 + }, + { + "epoch": 0.25765669515669515, + "grad_norm": 0.42974478006362915, + "learning_rate": 0.00019801174193903714, + "loss": 1.0659, + "step": 1447 + }, + { + "epoch": 0.25783475783475784, + "grad_norm": 0.4424504339694977, + "learning_rate": 0.00019800896362345464, + "loss": 0.9805, + "step": 1448 + }, + { + "epoch": 0.25801282051282054, + "grad_norm": 0.4734833836555481, + "learning_rate": 0.0001980061833875901, + "loss": 1.255, + "step": 1449 + }, + { + "epoch": 0.2581908831908832, + "grad_norm": 0.41024845838546753, + "learning_rate": 0.000198003401231498, + "loss": 1.0908, + "step": 1450 + }, + { + "epoch": 0.2583689458689459, + "grad_norm": 0.43603816628456116, + "learning_rate": 0.00019800061715523283, + "loss": 1.0611, + "step": 1451 + }, + { + "epoch": 0.25854700854700857, + "grad_norm": 0.4871339499950409, + "learning_rate": 0.00019799783115884915, + "loss": 1.1851, + "step": 1452 + }, + { + "epoch": 0.2587250712250712, + "grad_norm": 0.49758270382881165, + "learning_rate": 0.00019799504324240157, + "loss": 1.1936, + "step": 1453 + }, + { + "epoch": 0.2589031339031339, + "grad_norm": 0.4201010763645172, + "learning_rate": 0.00019799225340594466, + "loss": 1.1567, + "step": 1454 + }, + { + "epoch": 0.2590811965811966, + "grad_norm": 0.4200313091278076, + "learning_rate": 0.00019798946164953309, + "loss": 0.9666, + "step": 1455 + }, + { + "epoch": 0.25925925925925924, + "grad_norm": 0.43001702427864075, + "learning_rate": 0.0001979866679732216, + "loss": 1.0104, + "step": 1456 + }, + { + "epoch": 0.25943732193732194, + "grad_norm": 0.46733465790748596, + "learning_rate": 0.0001979838723770649, + "loss": 1.0927, + "step": 1457 + }, + { + "epoch": 0.25961538461538464, + "grad_norm": 0.4513280391693115, + "learning_rate": 0.00019798107486111773, + "loss": 1.0282, + "step": 1458 + }, + { + "epoch": 0.2597934472934473, + "grad_norm": 0.40411749482154846, + "learning_rate": 0.00019797827542543495, + "loss": 1.0789, + "step": 1459 + }, + { + "epoch": 0.25997150997151, + "grad_norm": 0.4359099268913269, + "learning_rate": 0.0001979754740700714, + "loss": 1.0616, + "step": 1460 + }, + { + "epoch": 0.26014957264957267, + "grad_norm": 0.4979047477245331, + "learning_rate": 0.00019797267079508198, + "loss": 1.2948, + "step": 1461 + }, + { + "epoch": 0.2603276353276353, + "grad_norm": 0.44698619842529297, + "learning_rate": 0.0001979698656005216, + "loss": 0.9198, + "step": 1462 + }, + { + "epoch": 0.260505698005698, + "grad_norm": 0.48437631130218506, + "learning_rate": 0.00019796705848644516, + "loss": 1.3207, + "step": 1463 + }, + { + "epoch": 0.2606837606837607, + "grad_norm": 0.4382587671279907, + "learning_rate": 0.00019796424945290778, + "loss": 1.1315, + "step": 1464 + }, + { + "epoch": 0.26086182336182334, + "grad_norm": 0.4565944969654083, + "learning_rate": 0.0001979614384999644, + "loss": 1.1893, + "step": 1465 + }, + { + "epoch": 0.26103988603988604, + "grad_norm": 0.4705163836479187, + "learning_rate": 0.00019795862562767017, + "loss": 1.1132, + "step": 1466 + }, + { + "epoch": 0.26121794871794873, + "grad_norm": 0.525184690952301, + "learning_rate": 0.00019795581083608012, + "loss": 1.2111, + "step": 1467 + }, + { + "epoch": 0.2613960113960114, + "grad_norm": 0.45215457677841187, + "learning_rate": 0.00019795299412524945, + "loss": 1.1851, + "step": 1468 + }, + { + "epoch": 0.26157407407407407, + "grad_norm": 0.4336663484573364, + "learning_rate": 0.00019795017549523335, + "loss": 1.0147, + "step": 1469 + }, + { + "epoch": 0.26175213675213677, + "grad_norm": 0.5327649712562561, + "learning_rate": 0.00019794735494608703, + "loss": 1.1743, + "step": 1470 + }, + { + "epoch": 0.2619301994301994, + "grad_norm": 0.49972307682037354, + "learning_rate": 0.00019794453247786578, + "loss": 1.1624, + "step": 1471 + }, + { + "epoch": 0.2621082621082621, + "grad_norm": 0.43475785851478577, + "learning_rate": 0.00019794170809062485, + "loss": 0.9888, + "step": 1472 + }, + { + "epoch": 0.2622863247863248, + "grad_norm": 0.428838849067688, + "learning_rate": 0.0001979388817844196, + "loss": 0.9154, + "step": 1473 + }, + { + "epoch": 0.26246438746438744, + "grad_norm": 0.508568286895752, + "learning_rate": 0.00019793605355930544, + "loss": 1.1679, + "step": 1474 + }, + { + "epoch": 0.26264245014245013, + "grad_norm": 0.47791770100593567, + "learning_rate": 0.00019793322341533776, + "loss": 1.1375, + "step": 1475 + }, + { + "epoch": 0.26282051282051283, + "grad_norm": 0.41909220814704895, + "learning_rate": 0.00019793039135257196, + "loss": 1.0235, + "step": 1476 + }, + { + "epoch": 0.26299857549857547, + "grad_norm": 0.5564408302307129, + "learning_rate": 0.00019792755737106361, + "loss": 1.0756, + "step": 1477 + }, + { + "epoch": 0.26317663817663817, + "grad_norm": 0.42813625931739807, + "learning_rate": 0.0001979247214708682, + "loss": 0.8213, + "step": 1478 + }, + { + "epoch": 0.26335470085470086, + "grad_norm": 0.44495970010757446, + "learning_rate": 0.00019792188365204126, + "loss": 0.9654, + "step": 1479 + }, + { + "epoch": 0.26353276353276356, + "grad_norm": 0.47473424673080444, + "learning_rate": 0.00019791904391463846, + "loss": 1.1643, + "step": 1480 + }, + { + "epoch": 0.2637108262108262, + "grad_norm": 0.40189051628112793, + "learning_rate": 0.0001979162022587154, + "loss": 0.8687, + "step": 1481 + }, + { + "epoch": 0.2638888888888889, + "grad_norm": 0.44629937410354614, + "learning_rate": 0.00019791335868432776, + "loss": 1.0284, + "step": 1482 + }, + { + "epoch": 0.2640669515669516, + "grad_norm": 0.511275053024292, + "learning_rate": 0.00019791051319153124, + "loss": 1.2217, + "step": 1483 + }, + { + "epoch": 0.26424501424501423, + "grad_norm": 0.5136445164680481, + "learning_rate": 0.00019790766578038163, + "loss": 1.1129, + "step": 1484 + }, + { + "epoch": 0.2644230769230769, + "grad_norm": 0.4450451135635376, + "learning_rate": 0.00019790481645093469, + "loss": 0.9912, + "step": 1485 + }, + { + "epoch": 0.2646011396011396, + "grad_norm": 0.39455199241638184, + "learning_rate": 0.00019790196520324621, + "loss": 1.0887, + "step": 1486 + }, + { + "epoch": 0.26477920227920226, + "grad_norm": 0.4444045126438141, + "learning_rate": 0.00019789911203737216, + "loss": 1.1559, + "step": 1487 + }, + { + "epoch": 0.26495726495726496, + "grad_norm": 0.4769677221775055, + "learning_rate": 0.0001978962569533683, + "loss": 1.147, + "step": 1488 + }, + { + "epoch": 0.26513532763532766, + "grad_norm": 0.40226617455482483, + "learning_rate": 0.0001978933999512907, + "loss": 1.0966, + "step": 1489 + }, + { + "epoch": 0.2653133903133903, + "grad_norm": 0.4640974700450897, + "learning_rate": 0.00019789054103119526, + "loss": 1.1002, + "step": 1490 + }, + { + "epoch": 0.265491452991453, + "grad_norm": 0.48251107335090637, + "learning_rate": 0.00019788768019313806, + "loss": 1.07, + "step": 1491 + }, + { + "epoch": 0.2656695156695157, + "grad_norm": 0.4836949408054352, + "learning_rate": 0.00019788481743717506, + "loss": 1.2992, + "step": 1492 + }, + { + "epoch": 0.26584757834757833, + "grad_norm": 0.4253857135772705, + "learning_rate": 0.00019788195276336244, + "loss": 1.1326, + "step": 1493 + }, + { + "epoch": 0.266025641025641, + "grad_norm": 0.5161862373352051, + "learning_rate": 0.0001978790861717563, + "loss": 1.2131, + "step": 1494 + }, + { + "epoch": 0.2662037037037037, + "grad_norm": 0.5223346948623657, + "learning_rate": 0.00019787621766241274, + "loss": 1.0933, + "step": 1495 + }, + { + "epoch": 0.26638176638176636, + "grad_norm": 0.37622541189193726, + "learning_rate": 0.000197873347235388, + "loss": 0.8919, + "step": 1496 + }, + { + "epoch": 0.26655982905982906, + "grad_norm": 0.4425419569015503, + "learning_rate": 0.0001978704748907384, + "loss": 1.0411, + "step": 1497 + }, + { + "epoch": 0.26673789173789175, + "grad_norm": 0.4536985456943512, + "learning_rate": 0.00019786760062852015, + "loss": 1.2747, + "step": 1498 + }, + { + "epoch": 0.2669159544159544, + "grad_norm": 0.4998049736022949, + "learning_rate": 0.00019786472444878955, + "loss": 1.3214, + "step": 1499 + }, + { + "epoch": 0.2670940170940171, + "grad_norm": 0.42104312777519226, + "learning_rate": 0.00019786184635160295, + "loss": 0.7878, + "step": 1500 + }, + { + "epoch": 0.2672720797720798, + "grad_norm": 0.5354288220405579, + "learning_rate": 0.00019785896633701678, + "loss": 1.0642, + "step": 1501 + }, + { + "epoch": 0.2674501424501424, + "grad_norm": 0.4681485891342163, + "learning_rate": 0.00019785608440508744, + "loss": 1.1737, + "step": 1502 + }, + { + "epoch": 0.2676282051282051, + "grad_norm": 0.49107062816619873, + "learning_rate": 0.0001978532005558714, + "loss": 1.1507, + "step": 1503 + }, + { + "epoch": 0.2678062678062678, + "grad_norm": 0.4173283576965332, + "learning_rate": 0.0001978503147894252, + "loss": 1.0538, + "step": 1504 + }, + { + "epoch": 0.26798433048433046, + "grad_norm": 0.49354055523872375, + "learning_rate": 0.0001978474271058053, + "loss": 1.1043, + "step": 1505 + }, + { + "epoch": 0.26816239316239315, + "grad_norm": 0.5787215232849121, + "learning_rate": 0.00019784453750506834, + "loss": 0.9245, + "step": 1506 + }, + { + "epoch": 0.26834045584045585, + "grad_norm": 0.48982590436935425, + "learning_rate": 0.00019784164598727095, + "loss": 1.2007, + "step": 1507 + }, + { + "epoch": 0.26851851851851855, + "grad_norm": 0.4971007704734802, + "learning_rate": 0.00019783875255246973, + "loss": 1.1174, + "step": 1508 + }, + { + "epoch": 0.2686965811965812, + "grad_norm": 0.5200340151786804, + "learning_rate": 0.00019783585720072142, + "loss": 1.1967, + "step": 1509 + }, + { + "epoch": 0.2688746438746439, + "grad_norm": 0.47911885380744934, + "learning_rate": 0.00019783295993208271, + "loss": 1.162, + "step": 1510 + }, + { + "epoch": 0.2690527065527066, + "grad_norm": 0.4764275848865509, + "learning_rate": 0.00019783006074661037, + "loss": 1.1358, + "step": 1511 + }, + { + "epoch": 0.2692307692307692, + "grad_norm": 0.478545606136322, + "learning_rate": 0.00019782715964436124, + "loss": 1.0096, + "step": 1512 + }, + { + "epoch": 0.2694088319088319, + "grad_norm": 0.5512787699699402, + "learning_rate": 0.00019782425662539212, + "loss": 1.1799, + "step": 1513 + }, + { + "epoch": 0.2695868945868946, + "grad_norm": 0.5495108962059021, + "learning_rate": 0.00019782135168975988, + "loss": 1.0959, + "step": 1514 + }, + { + "epoch": 0.26976495726495725, + "grad_norm": 0.42052868008613586, + "learning_rate": 0.0001978184448375215, + "loss": 1.1872, + "step": 1515 + }, + { + "epoch": 0.26994301994301995, + "grad_norm": 0.4994426965713501, + "learning_rate": 0.0001978155360687339, + "loss": 1.0568, + "step": 1516 + }, + { + "epoch": 0.27012108262108264, + "grad_norm": 0.459577351808548, + "learning_rate": 0.00019781262538345402, + "loss": 1.0315, + "step": 1517 + }, + { + "epoch": 0.2702991452991453, + "grad_norm": 0.4792841374874115, + "learning_rate": 0.00019780971278173895, + "loss": 1.2055, + "step": 1518 + }, + { + "epoch": 0.270477207977208, + "grad_norm": 0.5017708539962769, + "learning_rate": 0.00019780679826364575, + "loss": 1.157, + "step": 1519 + }, + { + "epoch": 0.2706552706552707, + "grad_norm": 0.5197349786758423, + "learning_rate": 0.00019780388182923152, + "loss": 0.9101, + "step": 1520 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 0.4226742684841156, + "learning_rate": 0.00019780096347855338, + "loss": 1.0525, + "step": 1521 + }, + { + "epoch": 0.271011396011396, + "grad_norm": 0.5058164596557617, + "learning_rate": 0.00019779804321166852, + "loss": 0.931, + "step": 1522 + }, + { + "epoch": 0.2711894586894587, + "grad_norm": 0.44492244720458984, + "learning_rate": 0.00019779512102863418, + "loss": 1.0641, + "step": 1523 + }, + { + "epoch": 0.27136752136752135, + "grad_norm": 0.5348989963531494, + "learning_rate": 0.00019779219692950758, + "loss": 1.1692, + "step": 1524 + }, + { + "epoch": 0.27154558404558404, + "grad_norm": 0.4631774425506592, + "learning_rate": 0.00019778927091434602, + "loss": 1.0876, + "step": 1525 + }, + { + "epoch": 0.27172364672364674, + "grad_norm": 0.45957499742507935, + "learning_rate": 0.00019778634298320684, + "loss": 0.9527, + "step": 1526 + }, + { + "epoch": 0.2719017094017094, + "grad_norm": 0.4506755769252777, + "learning_rate": 0.00019778341313614743, + "loss": 1.086, + "step": 1527 + }, + { + "epoch": 0.2720797720797721, + "grad_norm": 0.4900587797164917, + "learning_rate": 0.00019778048137322513, + "loss": 0.9911, + "step": 1528 + }, + { + "epoch": 0.27225783475783477, + "grad_norm": 0.478127658367157, + "learning_rate": 0.00019777754769449745, + "loss": 1.2083, + "step": 1529 + }, + { + "epoch": 0.2724358974358974, + "grad_norm": 0.47220897674560547, + "learning_rate": 0.00019777461210002183, + "loss": 1.0313, + "step": 1530 + }, + { + "epoch": 0.2726139601139601, + "grad_norm": 0.4526277184486389, + "learning_rate": 0.0001977716745898558, + "loss": 1.2648, + "step": 1531 + }, + { + "epoch": 0.2727920227920228, + "grad_norm": 0.42907601594924927, + "learning_rate": 0.00019776873516405688, + "loss": 0.8645, + "step": 1532 + }, + { + "epoch": 0.27297008547008544, + "grad_norm": 0.43440163135528564, + "learning_rate": 0.00019776579382268272, + "loss": 0.9702, + "step": 1533 + }, + { + "epoch": 0.27314814814814814, + "grad_norm": 0.48213550448417664, + "learning_rate": 0.0001977628505657909, + "loss": 0.998, + "step": 1534 + }, + { + "epoch": 0.27332621082621084, + "grad_norm": 0.43385565280914307, + "learning_rate": 0.00019775990539343914, + "loss": 1.0575, + "step": 1535 + }, + { + "epoch": 0.27350427350427353, + "grad_norm": 0.45706847310066223, + "learning_rate": 0.00019775695830568507, + "loss": 1.3024, + "step": 1536 + }, + { + "epoch": 0.27368233618233617, + "grad_norm": 0.45769137144088745, + "learning_rate": 0.00019775400930258652, + "loss": 1.0987, + "step": 1537 + }, + { + "epoch": 0.27386039886039887, + "grad_norm": 0.44682395458221436, + "learning_rate": 0.00019775105838420117, + "loss": 1.1327, + "step": 1538 + }, + { + "epoch": 0.27403846153846156, + "grad_norm": 0.5923072099685669, + "learning_rate": 0.00019774810555058694, + "loss": 1.4766, + "step": 1539 + }, + { + "epoch": 0.2742165242165242, + "grad_norm": 0.4327206015586853, + "learning_rate": 0.0001977451508018016, + "loss": 1.1175, + "step": 1540 + }, + { + "epoch": 0.2743945868945869, + "grad_norm": 0.48036691546440125, + "learning_rate": 0.00019774219413790315, + "loss": 1.1189, + "step": 1541 + }, + { + "epoch": 0.2745726495726496, + "grad_norm": 0.41371914744377136, + "learning_rate": 0.00019773923555894935, + "loss": 1.1366, + "step": 1542 + }, + { + "epoch": 0.27475071225071224, + "grad_norm": 0.4452378749847412, + "learning_rate": 0.00019773627506499832, + "loss": 0.9517, + "step": 1543 + }, + { + "epoch": 0.27492877492877493, + "grad_norm": 0.469098299741745, + "learning_rate": 0.00019773331265610802, + "loss": 1.0848, + "step": 1544 + }, + { + "epoch": 0.27510683760683763, + "grad_norm": 0.5390294790267944, + "learning_rate": 0.00019773034833233646, + "loss": 0.8589, + "step": 1545 + }, + { + "epoch": 0.27528490028490027, + "grad_norm": 0.5368238091468811, + "learning_rate": 0.00019772738209374174, + "loss": 1.2954, + "step": 1546 + }, + { + "epoch": 0.27546296296296297, + "grad_norm": 0.4705318510532379, + "learning_rate": 0.00019772441394038198, + "loss": 1.2252, + "step": 1547 + }, + { + "epoch": 0.27564102564102566, + "grad_norm": 0.4682813286781311, + "learning_rate": 0.00019772144387231533, + "loss": 1.0855, + "step": 1548 + }, + { + "epoch": 0.2758190883190883, + "grad_norm": 0.46876460313796997, + "learning_rate": 0.0001977184718896, + "loss": 1.1959, + "step": 1549 + }, + { + "epoch": 0.275997150997151, + "grad_norm": 0.4172806441783905, + "learning_rate": 0.00019771549799229416, + "loss": 1.2166, + "step": 1550 + }, + { + "epoch": 0.2761752136752137, + "grad_norm": 0.5088075399398804, + "learning_rate": 0.0001977125221804562, + "loss": 1.1285, + "step": 1551 + }, + { + "epoch": 0.27635327635327633, + "grad_norm": 0.4728628396987915, + "learning_rate": 0.0001977095444541443, + "loss": 1.2985, + "step": 1552 + }, + { + "epoch": 0.27653133903133903, + "grad_norm": 0.4431236684322357, + "learning_rate": 0.00019770656481341684, + "loss": 1.1298, + "step": 1553 + }, + { + "epoch": 0.2767094017094017, + "grad_norm": 0.474065363407135, + "learning_rate": 0.00019770358325833223, + "loss": 1.1915, + "step": 1554 + }, + { + "epoch": 0.27688746438746437, + "grad_norm": 0.45718875527381897, + "learning_rate": 0.00019770059978894885, + "loss": 1.0626, + "step": 1555 + }, + { + "epoch": 0.27706552706552706, + "grad_norm": 0.49300211668014526, + "learning_rate": 0.00019769761440532522, + "loss": 1.0134, + "step": 1556 + }, + { + "epoch": 0.27724358974358976, + "grad_norm": 0.4389498829841614, + "learning_rate": 0.00019769462710751974, + "loss": 1.0292, + "step": 1557 + }, + { + "epoch": 0.2774216524216524, + "grad_norm": 0.47330448031425476, + "learning_rate": 0.000197691637895591, + "loss": 1.1273, + "step": 1558 + }, + { + "epoch": 0.2775997150997151, + "grad_norm": 0.5322058200836182, + "learning_rate": 0.00019768864676959755, + "loss": 1.059, + "step": 1559 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 0.4714536964893341, + "learning_rate": 0.000197685653729598, + "loss": 1.1987, + "step": 1560 + }, + { + "epoch": 0.27795584045584043, + "grad_norm": 0.48687809705734253, + "learning_rate": 0.00019768265877565097, + "loss": 1.3206, + "step": 1561 + }, + { + "epoch": 0.2781339031339031, + "grad_norm": 0.46066713333129883, + "learning_rate": 0.00019767966190781518, + "loss": 1.0845, + "step": 1562 + }, + { + "epoch": 0.2783119658119658, + "grad_norm": 0.44372090697288513, + "learning_rate": 0.00019767666312614935, + "loss": 1.0942, + "step": 1563 + }, + { + "epoch": 0.27849002849002846, + "grad_norm": 0.4615907073020935, + "learning_rate": 0.00019767366243071216, + "loss": 1.071, + "step": 1564 + }, + { + "epoch": 0.27866809116809116, + "grad_norm": 0.502097487449646, + "learning_rate": 0.0001976706598215625, + "loss": 1.1164, + "step": 1565 + }, + { + "epoch": 0.27884615384615385, + "grad_norm": 0.4371815621852875, + "learning_rate": 0.00019766765529875913, + "loss": 1.0252, + "step": 1566 + }, + { + "epoch": 0.27902421652421655, + "grad_norm": 0.43035808205604553, + "learning_rate": 0.00019766464886236093, + "loss": 1.073, + "step": 1567 + }, + { + "epoch": 0.2792022792022792, + "grad_norm": 0.49721601605415344, + "learning_rate": 0.00019766164051242683, + "loss": 1.0316, + "step": 1568 + }, + { + "epoch": 0.2793803418803419, + "grad_norm": 0.44866231083869934, + "learning_rate": 0.00019765863024901576, + "loss": 1.0951, + "step": 1569 + }, + { + "epoch": 0.2795584045584046, + "grad_norm": 0.46318337321281433, + "learning_rate": 0.0001976556180721867, + "loss": 0.9836, + "step": 1570 + }, + { + "epoch": 0.2797364672364672, + "grad_norm": 0.4227696657180786, + "learning_rate": 0.00019765260398199868, + "loss": 1.0414, + "step": 1571 + }, + { + "epoch": 0.2799145299145299, + "grad_norm": 0.6062980890274048, + "learning_rate": 0.00019764958797851073, + "loss": 1.137, + "step": 1572 + }, + { + "epoch": 0.2800925925925926, + "grad_norm": 0.4856833219528198, + "learning_rate": 0.00019764657006178196, + "loss": 1.1361, + "step": 1573 + }, + { + "epoch": 0.28027065527065526, + "grad_norm": 0.45612895488739014, + "learning_rate": 0.00019764355023187146, + "loss": 1.0005, + "step": 1574 + }, + { + "epoch": 0.28044871794871795, + "grad_norm": 0.4143696129322052, + "learning_rate": 0.00019764052848883845, + "loss": 1.051, + "step": 1575 + }, + { + "epoch": 0.28062678062678065, + "grad_norm": 0.4532071352005005, + "learning_rate": 0.00019763750483274212, + "loss": 1.0595, + "step": 1576 + }, + { + "epoch": 0.2808048433048433, + "grad_norm": 0.4940357208251953, + "learning_rate": 0.0001976344792636417, + "loss": 1.0983, + "step": 1577 + }, + { + "epoch": 0.280982905982906, + "grad_norm": 0.44405099749565125, + "learning_rate": 0.0001976314517815965, + "loss": 1.0846, + "step": 1578 + }, + { + "epoch": 0.2811609686609687, + "grad_norm": 0.5508625507354736, + "learning_rate": 0.00019762842238666578, + "loss": 1.1722, + "step": 1579 + }, + { + "epoch": 0.2813390313390313, + "grad_norm": 0.5241084694862366, + "learning_rate": 0.00019762539107890894, + "loss": 1.351, + "step": 1580 + }, + { + "epoch": 0.281517094017094, + "grad_norm": 0.5307353734970093, + "learning_rate": 0.00019762235785838537, + "loss": 1.1868, + "step": 1581 + }, + { + "epoch": 0.2816951566951567, + "grad_norm": 0.45697924494743347, + "learning_rate": 0.00019761932272515447, + "loss": 1.1982, + "step": 1582 + }, + { + "epoch": 0.28187321937321935, + "grad_norm": 0.412483811378479, + "learning_rate": 0.00019761628567927574, + "loss": 1.0433, + "step": 1583 + }, + { + "epoch": 0.28205128205128205, + "grad_norm": 0.4614165425300598, + "learning_rate": 0.00019761324672080868, + "loss": 1.104, + "step": 1584 + }, + { + "epoch": 0.28222934472934474, + "grad_norm": 0.47644901275634766, + "learning_rate": 0.00019761020584981284, + "loss": 1.1037, + "step": 1585 + }, + { + "epoch": 0.2824074074074074, + "grad_norm": 0.4985184669494629, + "learning_rate": 0.00019760716306634773, + "loss": 1.2213, + "step": 1586 + }, + { + "epoch": 0.2825854700854701, + "grad_norm": 0.508301317691803, + "learning_rate": 0.00019760411837047305, + "loss": 1.1315, + "step": 1587 + }, + { + "epoch": 0.2827635327635328, + "grad_norm": 0.5346587300300598, + "learning_rate": 0.00019760107176224845, + "loss": 1.2281, + "step": 1588 + }, + { + "epoch": 0.2829415954415954, + "grad_norm": 0.5106825232505798, + "learning_rate": 0.00019759802324173357, + "loss": 1.2904, + "step": 1589 + }, + { + "epoch": 0.2831196581196581, + "grad_norm": 0.46458688378334045, + "learning_rate": 0.00019759497280898817, + "loss": 1.0861, + "step": 1590 + }, + { + "epoch": 0.2832977207977208, + "grad_norm": 0.49115365743637085, + "learning_rate": 0.00019759192046407201, + "loss": 1.0529, + "step": 1591 + }, + { + "epoch": 0.28347578347578345, + "grad_norm": 0.5114167332649231, + "learning_rate": 0.0001975888662070449, + "loss": 1.2555, + "step": 1592 + }, + { + "epoch": 0.28365384615384615, + "grad_norm": 0.45844775438308716, + "learning_rate": 0.0001975858100379667, + "loss": 1.0662, + "step": 1593 + }, + { + "epoch": 0.28383190883190884, + "grad_norm": 0.4684161841869354, + "learning_rate": 0.00019758275195689727, + "loss": 1.0537, + "step": 1594 + }, + { + "epoch": 0.28400997150997154, + "grad_norm": 0.4816220998764038, + "learning_rate": 0.0001975796919638965, + "loss": 1.126, + "step": 1595 + }, + { + "epoch": 0.2841880341880342, + "grad_norm": 0.46578118205070496, + "learning_rate": 0.0001975766300590244, + "loss": 0.9651, + "step": 1596 + }, + { + "epoch": 0.2843660968660969, + "grad_norm": 0.4181675612926483, + "learning_rate": 0.0001975735662423409, + "loss": 1.0888, + "step": 1597 + }, + { + "epoch": 0.28454415954415957, + "grad_norm": 0.49417954683303833, + "learning_rate": 0.00019757050051390609, + "loss": 1.1878, + "step": 1598 + }, + { + "epoch": 0.2847222222222222, + "grad_norm": 0.47264960408210754, + "learning_rate": 0.00019756743287377998, + "loss": 1.027, + "step": 1599 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 0.47686338424682617, + "learning_rate": 0.0001975643633220227, + "loss": 1.1307, + "step": 1600 + }, + { + "epoch": 0.2850783475783476, + "grad_norm": 0.5571266412734985, + "learning_rate": 0.00019756129185869443, + "loss": 0.984, + "step": 1601 + }, + { + "epoch": 0.28525641025641024, + "grad_norm": 0.46942809224128723, + "learning_rate": 0.00019755821848385527, + "loss": 1.0397, + "step": 1602 + }, + { + "epoch": 0.28543447293447294, + "grad_norm": 0.6325890421867371, + "learning_rate": 0.00019755514319756551, + "loss": 1.0918, + "step": 1603 + }, + { + "epoch": 0.28561253561253563, + "grad_norm": 0.5297608375549316, + "learning_rate": 0.00019755206599988533, + "loss": 0.9911, + "step": 1604 + }, + { + "epoch": 0.2857905982905983, + "grad_norm": 0.4736945331096649, + "learning_rate": 0.00019754898689087512, + "loss": 1.0786, + "step": 1605 + }, + { + "epoch": 0.28596866096866097, + "grad_norm": 0.5048685669898987, + "learning_rate": 0.00019754590587059512, + "loss": 0.9834, + "step": 1606 + }, + { + "epoch": 0.28614672364672367, + "grad_norm": 0.3823149502277374, + "learning_rate": 0.00019754282293910574, + "loss": 0.8341, + "step": 1607 + }, + { + "epoch": 0.2863247863247863, + "grad_norm": 0.44071945548057556, + "learning_rate": 0.00019753973809646738, + "loss": 1.131, + "step": 1608 + }, + { + "epoch": 0.286502849002849, + "grad_norm": 0.44182759523391724, + "learning_rate": 0.00019753665134274043, + "loss": 1.0321, + "step": 1609 + }, + { + "epoch": 0.2866809116809117, + "grad_norm": 0.4486250877380371, + "learning_rate": 0.00019753356267798546, + "loss": 0.9941, + "step": 1610 + }, + { + "epoch": 0.28685897435897434, + "grad_norm": 0.42796584963798523, + "learning_rate": 0.00019753047210226292, + "loss": 1.0235, + "step": 1611 + }, + { + "epoch": 0.28703703703703703, + "grad_norm": 0.47294023633003235, + "learning_rate": 0.00019752737961563336, + "loss": 1.11, + "step": 1612 + }, + { + "epoch": 0.28721509971509973, + "grad_norm": 0.44550734758377075, + "learning_rate": 0.00019752428521815742, + "loss": 1.0849, + "step": 1613 + }, + { + "epoch": 0.28739316239316237, + "grad_norm": 0.44189929962158203, + "learning_rate": 0.0001975211889098957, + "loss": 0.8904, + "step": 1614 + }, + { + "epoch": 0.28757122507122507, + "grad_norm": 0.5302733182907104, + "learning_rate": 0.00019751809069090885, + "loss": 1.2348, + "step": 1615 + }, + { + "epoch": 0.28774928774928776, + "grad_norm": 0.5951390862464905, + "learning_rate": 0.00019751499056125762, + "loss": 1.3035, + "step": 1616 + }, + { + "epoch": 0.2879273504273504, + "grad_norm": 0.5431534647941589, + "learning_rate": 0.0001975118885210027, + "loss": 1.0016, + "step": 1617 + }, + { + "epoch": 0.2881054131054131, + "grad_norm": 0.47301986813545227, + "learning_rate": 0.00019750878457020489, + "loss": 1.2245, + "step": 1618 + }, + { + "epoch": 0.2882834757834758, + "grad_norm": 0.44785359501838684, + "learning_rate": 0.00019750567870892497, + "loss": 1.122, + "step": 1619 + }, + { + "epoch": 0.28846153846153844, + "grad_norm": 0.49494361877441406, + "learning_rate": 0.00019750257093722383, + "loss": 0.9421, + "step": 1620 + }, + { + "epoch": 0.28863960113960113, + "grad_norm": 0.4484521150588989, + "learning_rate": 0.00019749946125516242, + "loss": 1.2146, + "step": 1621 + }, + { + "epoch": 0.28881766381766383, + "grad_norm": 0.4635269343852997, + "learning_rate": 0.00019749634966280156, + "loss": 0.976, + "step": 1622 + }, + { + "epoch": 0.28899572649572647, + "grad_norm": 0.5532249808311462, + "learning_rate": 0.00019749323616020226, + "loss": 1.1818, + "step": 1623 + }, + { + "epoch": 0.28917378917378916, + "grad_norm": 0.4730629622936249, + "learning_rate": 0.00019749012074742552, + "loss": 1.0321, + "step": 1624 + }, + { + "epoch": 0.28935185185185186, + "grad_norm": 0.47437289357185364, + "learning_rate": 0.0001974870034245324, + "loss": 1.1572, + "step": 1625 + }, + { + "epoch": 0.28952991452991456, + "grad_norm": 0.4796304404735565, + "learning_rate": 0.00019748388419158394, + "loss": 1.1667, + "step": 1626 + }, + { + "epoch": 0.2897079772079772, + "grad_norm": 0.42686304450035095, + "learning_rate": 0.0001974807630486413, + "loss": 0.9824, + "step": 1627 + }, + { + "epoch": 0.2898860398860399, + "grad_norm": 0.4444865584373474, + "learning_rate": 0.00019747763999576558, + "loss": 1.2789, + "step": 1628 + }, + { + "epoch": 0.2900641025641026, + "grad_norm": 0.5039985179901123, + "learning_rate": 0.000197474515033018, + "loss": 1.1488, + "step": 1629 + }, + { + "epoch": 0.29024216524216523, + "grad_norm": 0.581479549407959, + "learning_rate": 0.00019747138816045978, + "loss": 1.1232, + "step": 1630 + }, + { + "epoch": 0.2904202279202279, + "grad_norm": 0.5415821075439453, + "learning_rate": 0.00019746825937815222, + "loss": 1.2326, + "step": 1631 + }, + { + "epoch": 0.2905982905982906, + "grad_norm": 0.45528364181518555, + "learning_rate": 0.00019746512868615656, + "loss": 1.0246, + "step": 1632 + }, + { + "epoch": 0.29077635327635326, + "grad_norm": 0.5255574584007263, + "learning_rate": 0.00019746199608453418, + "loss": 1.0592, + "step": 1633 + }, + { + "epoch": 0.29095441595441596, + "grad_norm": 0.5064096450805664, + "learning_rate": 0.00019745886157334646, + "loss": 1.3439, + "step": 1634 + }, + { + "epoch": 0.29113247863247865, + "grad_norm": 0.500848650932312, + "learning_rate": 0.00019745572515265475, + "loss": 1.1212, + "step": 1635 + }, + { + "epoch": 0.2913105413105413, + "grad_norm": 0.5229088068008423, + "learning_rate": 0.00019745258682252062, + "loss": 1.1019, + "step": 1636 + }, + { + "epoch": 0.291488603988604, + "grad_norm": 0.4494398832321167, + "learning_rate": 0.00019744944658300545, + "loss": 1.1298, + "step": 1637 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 0.48383277654647827, + "learning_rate": 0.00019744630443417082, + "loss": 1.206, + "step": 1638 + }, + { + "epoch": 0.2918447293447293, + "grad_norm": 0.4870131313800812, + "learning_rate": 0.00019744316037607828, + "loss": 1.2096, + "step": 1639 + }, + { + "epoch": 0.292022792022792, + "grad_norm": 0.4153090715408325, + "learning_rate": 0.00019744001440878944, + "loss": 1.0478, + "step": 1640 + }, + { + "epoch": 0.2922008547008547, + "grad_norm": 0.4262249171733856, + "learning_rate": 0.0001974368665323659, + "loss": 1.0393, + "step": 1641 + }, + { + "epoch": 0.29237891737891736, + "grad_norm": 0.46131134033203125, + "learning_rate": 0.00019743371674686938, + "loss": 1.0908, + "step": 1642 + }, + { + "epoch": 0.29255698005698005, + "grad_norm": 0.44877463579177856, + "learning_rate": 0.0001974305650523616, + "loss": 1.1906, + "step": 1643 + }, + { + "epoch": 0.29273504273504275, + "grad_norm": 0.5199326276779175, + "learning_rate": 0.00019742741144890432, + "loss": 1.1147, + "step": 1644 + }, + { + "epoch": 0.2929131054131054, + "grad_norm": 0.48142504692077637, + "learning_rate": 0.00019742425593655924, + "loss": 1.1951, + "step": 1645 + }, + { + "epoch": 0.2930911680911681, + "grad_norm": 0.5672988891601562, + "learning_rate": 0.0001974210985153883, + "loss": 1.1817, + "step": 1646 + }, + { + "epoch": 0.2932692307692308, + "grad_norm": 0.38135233521461487, + "learning_rate": 0.00019741793918545326, + "loss": 0.8567, + "step": 1647 + }, + { + "epoch": 0.2934472934472934, + "grad_norm": 0.6153588891029358, + "learning_rate": 0.0001974147779468161, + "loss": 1.0593, + "step": 1648 + }, + { + "epoch": 0.2936253561253561, + "grad_norm": 0.38935527205467224, + "learning_rate": 0.0001974116147995387, + "loss": 0.9907, + "step": 1649 + }, + { + "epoch": 0.2938034188034188, + "grad_norm": 0.467351496219635, + "learning_rate": 0.0001974084497436831, + "loss": 1.091, + "step": 1650 + }, + { + "epoch": 0.29398148148148145, + "grad_norm": 0.45613420009613037, + "learning_rate": 0.00019740528277931128, + "loss": 0.6789, + "step": 1651 + }, + { + "epoch": 0.29415954415954415, + "grad_norm": 0.4045158326625824, + "learning_rate": 0.00019740211390648524, + "loss": 1.0727, + "step": 1652 + }, + { + "epoch": 0.29433760683760685, + "grad_norm": 0.5122803449630737, + "learning_rate": 0.00019739894312526714, + "loss": 1.2297, + "step": 1653 + }, + { + "epoch": 0.29451566951566954, + "grad_norm": 0.44304123520851135, + "learning_rate": 0.00019739577043571908, + "loss": 0.9562, + "step": 1654 + }, + { + "epoch": 0.2946937321937322, + "grad_norm": 0.6070618629455566, + "learning_rate": 0.00019739259583790322, + "loss": 1.2745, + "step": 1655 + }, + { + "epoch": 0.2948717948717949, + "grad_norm": 0.48815637826919556, + "learning_rate": 0.00019738941933188176, + "loss": 1.0574, + "step": 1656 + }, + { + "epoch": 0.2950498575498576, + "grad_norm": 0.5067802667617798, + "learning_rate": 0.00019738624091771693, + "loss": 1.1874, + "step": 1657 + }, + { + "epoch": 0.2952279202279202, + "grad_norm": 0.4956928491592407, + "learning_rate": 0.000197383060595471, + "loss": 1.1085, + "step": 1658 + }, + { + "epoch": 0.2954059829059829, + "grad_norm": 0.46313008666038513, + "learning_rate": 0.00019737987836520633, + "loss": 1.0548, + "step": 1659 + }, + { + "epoch": 0.2955840455840456, + "grad_norm": 0.49944064021110535, + "learning_rate": 0.0001973766942269852, + "loss": 1.1485, + "step": 1660 + }, + { + "epoch": 0.29576210826210825, + "grad_norm": 0.4743517339229584, + "learning_rate": 0.00019737350818087003, + "loss": 0.9279, + "step": 1661 + }, + { + "epoch": 0.29594017094017094, + "grad_norm": 0.45935431122779846, + "learning_rate": 0.00019737032022692326, + "loss": 0.9574, + "step": 1662 + }, + { + "epoch": 0.29611823361823364, + "grad_norm": 0.4550873637199402, + "learning_rate": 0.00019736713036520734, + "loss": 1.1642, + "step": 1663 + }, + { + "epoch": 0.2962962962962963, + "grad_norm": 0.45252951979637146, + "learning_rate": 0.00019736393859578474, + "loss": 1.0113, + "step": 1664 + }, + { + "epoch": 0.296474358974359, + "grad_norm": 0.5147238969802856, + "learning_rate": 0.00019736074491871804, + "loss": 1.1604, + "step": 1665 + }, + { + "epoch": 0.29665242165242167, + "grad_norm": 0.5122934579849243, + "learning_rate": 0.00019735754933406977, + "loss": 0.9525, + "step": 1666 + }, + { + "epoch": 0.2968304843304843, + "grad_norm": 0.438620001077652, + "learning_rate": 0.00019735435184190257, + "loss": 1.0728, + "step": 1667 + }, + { + "epoch": 0.297008547008547, + "grad_norm": 0.41970670223236084, + "learning_rate": 0.00019735115244227908, + "loss": 0.9782, + "step": 1668 + }, + { + "epoch": 0.2971866096866097, + "grad_norm": 0.5447152256965637, + "learning_rate": 0.000197347951135262, + "loss": 1.0633, + "step": 1669 + }, + { + "epoch": 0.29736467236467234, + "grad_norm": 0.4846996068954468, + "learning_rate": 0.00019734474792091407, + "loss": 0.9019, + "step": 1670 + }, + { + "epoch": 0.29754273504273504, + "grad_norm": 0.4721437990665436, + "learning_rate": 0.00019734154279929796, + "loss": 1.1793, + "step": 1671 + }, + { + "epoch": 0.29772079772079774, + "grad_norm": 0.4659852385520935, + "learning_rate": 0.00019733833577047655, + "loss": 1.1503, + "step": 1672 + }, + { + "epoch": 0.2978988603988604, + "grad_norm": 0.3733183443546295, + "learning_rate": 0.00019733512683451268, + "loss": 0.7763, + "step": 1673 + }, + { + "epoch": 0.2980769230769231, + "grad_norm": 0.4898292124271393, + "learning_rate": 0.0001973319159914692, + "loss": 1.3146, + "step": 1674 + }, + { + "epoch": 0.29825498575498577, + "grad_norm": 0.41774725914001465, + "learning_rate": 0.00019732870324140899, + "loss": 1.2069, + "step": 1675 + }, + { + "epoch": 0.2984330484330484, + "grad_norm": 0.4607912003993988, + "learning_rate": 0.000197325488584395, + "loss": 1.2255, + "step": 1676 + }, + { + "epoch": 0.2986111111111111, + "grad_norm": 0.4692424237728119, + "learning_rate": 0.00019732227202049025, + "loss": 1.0793, + "step": 1677 + }, + { + "epoch": 0.2987891737891738, + "grad_norm": 0.5925022959709167, + "learning_rate": 0.00019731905354975778, + "loss": 1.0297, + "step": 1678 + }, + { + "epoch": 0.29896723646723644, + "grad_norm": 0.44047990441322327, + "learning_rate": 0.00019731583317226056, + "loss": 1.0982, + "step": 1679 + }, + { + "epoch": 0.29914529914529914, + "grad_norm": 0.5863066911697388, + "learning_rate": 0.0001973126108880618, + "loss": 1.0284, + "step": 1680 + }, + { + "epoch": 0.29932336182336183, + "grad_norm": 0.48962152004241943, + "learning_rate": 0.00019730938669722457, + "loss": 1.1861, + "step": 1681 + }, + { + "epoch": 0.29950142450142453, + "grad_norm": 0.5445577502250671, + "learning_rate": 0.00019730616059981205, + "loss": 1.2574, + "step": 1682 + }, + { + "epoch": 0.29967948717948717, + "grad_norm": 0.49327564239501953, + "learning_rate": 0.00019730293259588743, + "loss": 0.9578, + "step": 1683 + }, + { + "epoch": 0.29985754985754987, + "grad_norm": 0.4252840578556061, + "learning_rate": 0.00019729970268551398, + "loss": 1.0083, + "step": 1684 + }, + { + "epoch": 0.30003561253561256, + "grad_norm": 0.5140926241874695, + "learning_rate": 0.000197296470868755, + "loss": 1.3263, + "step": 1685 + }, + { + "epoch": 0.3002136752136752, + "grad_norm": 0.5143948197364807, + "learning_rate": 0.00019729323714567375, + "loss": 1.0424, + "step": 1686 + }, + { + "epoch": 0.3003917378917379, + "grad_norm": 0.3811354339122772, + "learning_rate": 0.00019729000151633367, + "loss": 0.6319, + "step": 1687 + }, + { + "epoch": 0.3005698005698006, + "grad_norm": 0.5249716639518738, + "learning_rate": 0.0001972867639807981, + "loss": 1.0173, + "step": 1688 + }, + { + "epoch": 0.30074786324786323, + "grad_norm": 0.41832098364830017, + "learning_rate": 0.00019728352453913048, + "loss": 1.0503, + "step": 1689 + }, + { + "epoch": 0.30092592592592593, + "grad_norm": 0.5961149334907532, + "learning_rate": 0.00019728028319139428, + "loss": 1.1843, + "step": 1690 + }, + { + "epoch": 0.3011039886039886, + "grad_norm": 0.44083690643310547, + "learning_rate": 0.00019727703993765303, + "loss": 1.1311, + "step": 1691 + }, + { + "epoch": 0.30128205128205127, + "grad_norm": 0.4368111491203308, + "learning_rate": 0.00019727379477797022, + "loss": 0.9463, + "step": 1692 + }, + { + "epoch": 0.30146011396011396, + "grad_norm": 0.5289376974105835, + "learning_rate": 0.00019727054771240954, + "loss": 0.9836, + "step": 1693 + }, + { + "epoch": 0.30163817663817666, + "grad_norm": 0.4132843613624573, + "learning_rate": 0.00019726729874103448, + "loss": 1.1052, + "step": 1694 + }, + { + "epoch": 0.3018162393162393, + "grad_norm": 0.4919086992740631, + "learning_rate": 0.00019726404786390877, + "loss": 1.2219, + "step": 1695 + }, + { + "epoch": 0.301994301994302, + "grad_norm": 0.42561691999435425, + "learning_rate": 0.0001972607950810961, + "loss": 1.0756, + "step": 1696 + }, + { + "epoch": 0.3021723646723647, + "grad_norm": 0.5030396580696106, + "learning_rate": 0.0001972575403926602, + "loss": 1.2207, + "step": 1697 + }, + { + "epoch": 0.30235042735042733, + "grad_norm": 0.4779801666736603, + "learning_rate": 0.0001972542837986648, + "loss": 1.194, + "step": 1698 + }, + { + "epoch": 0.30252849002849, + "grad_norm": 0.45395568013191223, + "learning_rate": 0.00019725102529917377, + "loss": 1.0775, + "step": 1699 + }, + { + "epoch": 0.3027065527065527, + "grad_norm": 0.6540699005126953, + "learning_rate": 0.0001972477648942509, + "loss": 1.181, + "step": 1700 + }, + { + "epoch": 0.30288461538461536, + "grad_norm": 0.46281275153160095, + "learning_rate": 0.00019724450258396008, + "loss": 0.629, + "step": 1701 + }, + { + "epoch": 0.30306267806267806, + "grad_norm": 0.3452845811843872, + "learning_rate": 0.00019724123836836527, + "loss": 0.51, + "step": 1702 + }, + { + "epoch": 0.30324074074074076, + "grad_norm": 0.4507991671562195, + "learning_rate": 0.00019723797224753038, + "loss": 1.0258, + "step": 1703 + }, + { + "epoch": 0.3034188034188034, + "grad_norm": 0.5385412573814392, + "learning_rate": 0.0001972347042215194, + "loss": 1.0232, + "step": 1704 + }, + { + "epoch": 0.3035968660968661, + "grad_norm": 0.4460466504096985, + "learning_rate": 0.00019723143429039642, + "loss": 1.1307, + "step": 1705 + }, + { + "epoch": 0.3037749287749288, + "grad_norm": 0.5229718685150146, + "learning_rate": 0.00019722816245422545, + "loss": 1.0964, + "step": 1706 + }, + { + "epoch": 0.30395299145299143, + "grad_norm": 0.4776979088783264, + "learning_rate": 0.00019722488871307058, + "loss": 1.2678, + "step": 1707 + }, + { + "epoch": 0.3041310541310541, + "grad_norm": 0.5371831655502319, + "learning_rate": 0.00019722161306699601, + "loss": 1.2808, + "step": 1708 + }, + { + "epoch": 0.3043091168091168, + "grad_norm": 0.45322108268737793, + "learning_rate": 0.0001972183355160659, + "loss": 1.0775, + "step": 1709 + }, + { + "epoch": 0.30448717948717946, + "grad_norm": 0.5036569833755493, + "learning_rate": 0.00019721505606034448, + "loss": 1.1859, + "step": 1710 + }, + { + "epoch": 0.30466524216524216, + "grad_norm": 0.5425969958305359, + "learning_rate": 0.00019721177469989593, + "loss": 1.0173, + "step": 1711 + }, + { + "epoch": 0.30484330484330485, + "grad_norm": 0.5638980269432068, + "learning_rate": 0.00019720849143478462, + "loss": 1.182, + "step": 1712 + }, + { + "epoch": 0.30502136752136755, + "grad_norm": 0.5160546898841858, + "learning_rate": 0.00019720520626507486, + "loss": 0.9853, + "step": 1713 + }, + { + "epoch": 0.3051994301994302, + "grad_norm": 0.5079004168510437, + "learning_rate": 0.000197201919190831, + "loss": 1.3154, + "step": 1714 + }, + { + "epoch": 0.3053774928774929, + "grad_norm": 0.4590355455875397, + "learning_rate": 0.00019719863021211745, + "loss": 1.007, + "step": 1715 + }, + { + "epoch": 0.3055555555555556, + "grad_norm": 0.49656423926353455, + "learning_rate": 0.00019719533932899865, + "loss": 1.2187, + "step": 1716 + }, + { + "epoch": 0.3057336182336182, + "grad_norm": 0.46426209807395935, + "learning_rate": 0.0001971920465415391, + "loss": 1.3007, + "step": 1717 + }, + { + "epoch": 0.3059116809116809, + "grad_norm": 0.5211917757987976, + "learning_rate": 0.00019718875184980328, + "loss": 1.2256, + "step": 1718 + }, + { + "epoch": 0.3060897435897436, + "grad_norm": 0.42953309416770935, + "learning_rate": 0.00019718545525385578, + "loss": 1.2838, + "step": 1719 + }, + { + "epoch": 0.30626780626780625, + "grad_norm": 0.4893105924129486, + "learning_rate": 0.00019718215675376116, + "loss": 1.052, + "step": 1720 + }, + { + "epoch": 0.30644586894586895, + "grad_norm": 0.4833602011203766, + "learning_rate": 0.00019717885634958405, + "loss": 1.069, + "step": 1721 + }, + { + "epoch": 0.30662393162393164, + "grad_norm": 0.502176821231842, + "learning_rate": 0.0001971755540413891, + "loss": 1.1659, + "step": 1722 + }, + { + "epoch": 0.3068019943019943, + "grad_norm": 0.4648856818675995, + "learning_rate": 0.00019717224982924108, + "loss": 1.1873, + "step": 1723 + }, + { + "epoch": 0.306980056980057, + "grad_norm": 0.405429869890213, + "learning_rate": 0.00019716894371320465, + "loss": 0.99, + "step": 1724 + }, + { + "epoch": 0.3071581196581197, + "grad_norm": 0.4306945204734802, + "learning_rate": 0.00019716563569334463, + "loss": 0.8751, + "step": 1725 + }, + { + "epoch": 0.3073361823361823, + "grad_norm": 0.49424824118614197, + "learning_rate": 0.00019716232576972583, + "loss": 0.9205, + "step": 1726 + }, + { + "epoch": 0.307514245014245, + "grad_norm": 0.5044034123420715, + "learning_rate": 0.00019715901394241306, + "loss": 1.2042, + "step": 1727 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 0.512180507183075, + "learning_rate": 0.00019715570021147126, + "loss": 1.1644, + "step": 1728 + }, + { + "epoch": 0.30787037037037035, + "grad_norm": 0.4377981126308441, + "learning_rate": 0.00019715238457696538, + "loss": 1.1625, + "step": 1729 + }, + { + "epoch": 0.30804843304843305, + "grad_norm": 0.49107855558395386, + "learning_rate": 0.00019714906703896027, + "loss": 1.1037, + "step": 1730 + }, + { + "epoch": 0.30822649572649574, + "grad_norm": 0.47342559695243835, + "learning_rate": 0.00019714574759752105, + "loss": 1.3186, + "step": 1731 + }, + { + "epoch": 0.3084045584045584, + "grad_norm": 0.487177312374115, + "learning_rate": 0.0001971424262527127, + "loss": 1.1196, + "step": 1732 + }, + { + "epoch": 0.3085826210826211, + "grad_norm": 0.5290025472640991, + "learning_rate": 0.0001971391030046003, + "loss": 1.2103, + "step": 1733 + }, + { + "epoch": 0.3087606837606838, + "grad_norm": 0.4587760269641876, + "learning_rate": 0.00019713577785324896, + "loss": 1.1017, + "step": 1734 + }, + { + "epoch": 0.3089387464387464, + "grad_norm": 0.45323294401168823, + "learning_rate": 0.00019713245079872388, + "loss": 1.0, + "step": 1735 + }, + { + "epoch": 0.3091168091168091, + "grad_norm": 0.43414804339408875, + "learning_rate": 0.00019712912184109013, + "loss": 1.0341, + "step": 1736 + }, + { + "epoch": 0.3092948717948718, + "grad_norm": 0.49604663252830505, + "learning_rate": 0.00019712579098041304, + "loss": 0.9437, + "step": 1737 + }, + { + "epoch": 0.30947293447293445, + "grad_norm": 0.48580703139305115, + "learning_rate": 0.00019712245821675785, + "loss": 1.2622, + "step": 1738 + }, + { + "epoch": 0.30965099715099714, + "grad_norm": 0.45333603024482727, + "learning_rate": 0.00019711912355018982, + "loss": 1.2063, + "step": 1739 + }, + { + "epoch": 0.30982905982905984, + "grad_norm": 0.5990764498710632, + "learning_rate": 0.00019711578698077432, + "loss": 1.5097, + "step": 1740 + }, + { + "epoch": 0.31000712250712253, + "grad_norm": 0.4386102259159088, + "learning_rate": 0.0001971124485085767, + "loss": 1.1283, + "step": 1741 + }, + { + "epoch": 0.3101851851851852, + "grad_norm": 0.4476035237312317, + "learning_rate": 0.00019710910813366242, + "loss": 0.8922, + "step": 1742 + }, + { + "epoch": 0.31036324786324787, + "grad_norm": 0.5276228785514832, + "learning_rate": 0.00019710576585609685, + "loss": 1.2373, + "step": 1743 + }, + { + "epoch": 0.31054131054131057, + "grad_norm": 0.4885637164115906, + "learning_rate": 0.00019710242167594557, + "loss": 1.0881, + "step": 1744 + }, + { + "epoch": 0.3107193732193732, + "grad_norm": 0.421132355928421, + "learning_rate": 0.000197099075593274, + "loss": 1.0544, + "step": 1745 + }, + { + "epoch": 0.3108974358974359, + "grad_norm": 0.5257927179336548, + "learning_rate": 0.00019709572760814777, + "loss": 1.265, + "step": 1746 + }, + { + "epoch": 0.3110754985754986, + "grad_norm": 0.5164850950241089, + "learning_rate": 0.00019709237772063247, + "loss": 0.9593, + "step": 1747 + }, + { + "epoch": 0.31125356125356124, + "grad_norm": 0.5176383256912231, + "learning_rate": 0.00019708902593079374, + "loss": 1.0194, + "step": 1748 + }, + { + "epoch": 0.31143162393162394, + "grad_norm": 0.4620790481567383, + "learning_rate": 0.00019708567223869716, + "loss": 0.9241, + "step": 1749 + }, + { + "epoch": 0.31160968660968663, + "grad_norm": 0.48307979106903076, + "learning_rate": 0.00019708231664440854, + "loss": 1.2314, + "step": 1750 + }, + { + "epoch": 0.31178774928774927, + "grad_norm": 0.4931468069553375, + "learning_rate": 0.00019707895914799364, + "loss": 1.2065, + "step": 1751 + }, + { + "epoch": 0.31196581196581197, + "grad_norm": 0.5035979747772217, + "learning_rate": 0.00019707559974951818, + "loss": 1.1867, + "step": 1752 + }, + { + "epoch": 0.31214387464387466, + "grad_norm": 0.47543632984161377, + "learning_rate": 0.00019707223844904795, + "loss": 1.0603, + "step": 1753 + }, + { + "epoch": 0.3123219373219373, + "grad_norm": 0.49929797649383545, + "learning_rate": 0.00019706887524664892, + "loss": 1.0597, + "step": 1754 + }, + { + "epoch": 0.3125, + "grad_norm": 0.5075222253799438, + "learning_rate": 0.00019706551014238687, + "loss": 1.1398, + "step": 1755 + }, + { + "epoch": 0.3126780626780627, + "grad_norm": 0.5096884369850159, + "learning_rate": 0.00019706214313632784, + "loss": 1.1382, + "step": 1756 + }, + { + "epoch": 0.31285612535612534, + "grad_norm": 0.4629988372325897, + "learning_rate": 0.0001970587742285377, + "loss": 1.0009, + "step": 1757 + }, + { + "epoch": 0.31303418803418803, + "grad_norm": 0.5244084596633911, + "learning_rate": 0.00019705540341908253, + "loss": 1.047, + "step": 1758 + }, + { + "epoch": 0.31321225071225073, + "grad_norm": 0.5136716961860657, + "learning_rate": 0.00019705203070802832, + "loss": 1.29, + "step": 1759 + }, + { + "epoch": 0.31339031339031337, + "grad_norm": 0.43991541862487793, + "learning_rate": 0.0001970486560954412, + "loss": 0.9605, + "step": 1760 + }, + { + "epoch": 0.31356837606837606, + "grad_norm": 0.4633477032184601, + "learning_rate": 0.00019704527958138725, + "loss": 1.1507, + "step": 1761 + }, + { + "epoch": 0.31374643874643876, + "grad_norm": 0.4419999420642853, + "learning_rate": 0.00019704190116593266, + "loss": 0.9262, + "step": 1762 + }, + { + "epoch": 0.3139245014245014, + "grad_norm": 0.49359434843063354, + "learning_rate": 0.00019703852084914357, + "loss": 0.9348, + "step": 1763 + }, + { + "epoch": 0.3141025641025641, + "grad_norm": 0.5072139501571655, + "learning_rate": 0.00019703513863108627, + "loss": 1.1592, + "step": 1764 + }, + { + "epoch": 0.3142806267806268, + "grad_norm": 0.45969831943511963, + "learning_rate": 0.00019703175451182698, + "loss": 1.1519, + "step": 1765 + }, + { + "epoch": 0.31445868945868943, + "grad_norm": 0.5148758292198181, + "learning_rate": 0.00019702836849143208, + "loss": 1.1673, + "step": 1766 + }, + { + "epoch": 0.31463675213675213, + "grad_norm": 0.43033209443092346, + "learning_rate": 0.0001970249805699678, + "loss": 0.9256, + "step": 1767 + }, + { + "epoch": 0.3148148148148148, + "grad_norm": 0.48143425583839417, + "learning_rate": 0.00019702159074750058, + "loss": 1.08, + "step": 1768 + }, + { + "epoch": 0.31499287749287747, + "grad_norm": 0.4780619740486145, + "learning_rate": 0.00019701819902409685, + "loss": 1.1198, + "step": 1769 + }, + { + "epoch": 0.31517094017094016, + "grad_norm": 0.4662075936794281, + "learning_rate": 0.00019701480539982305, + "loss": 0.8424, + "step": 1770 + }, + { + "epoch": 0.31534900284900286, + "grad_norm": 0.503901481628418, + "learning_rate": 0.00019701140987474566, + "loss": 1.1026, + "step": 1771 + }, + { + "epoch": 0.31552706552706555, + "grad_norm": 0.5197132229804993, + "learning_rate": 0.00019700801244893124, + "loss": 1.2148, + "step": 1772 + }, + { + "epoch": 0.3157051282051282, + "grad_norm": 0.4746309220790863, + "learning_rate": 0.00019700461312244634, + "loss": 1.0906, + "step": 1773 + }, + { + "epoch": 0.3158831908831909, + "grad_norm": 0.5277339816093445, + "learning_rate": 0.00019700121189535752, + "loss": 1.0588, + "step": 1774 + }, + { + "epoch": 0.3160612535612536, + "grad_norm": 0.436002254486084, + "learning_rate": 0.00019699780876773147, + "loss": 1.0341, + "step": 1775 + }, + { + "epoch": 0.3162393162393162, + "grad_norm": 0.5171145796775818, + "learning_rate": 0.00019699440373963486, + "loss": 1.282, + "step": 1776 + }, + { + "epoch": 0.3164173789173789, + "grad_norm": 0.38382846117019653, + "learning_rate": 0.00019699099681113436, + "loss": 0.8908, + "step": 1777 + }, + { + "epoch": 0.3165954415954416, + "grad_norm": 0.4621630609035492, + "learning_rate": 0.0001969875879822968, + "loss": 1.1074, + "step": 1778 + }, + { + "epoch": 0.31677350427350426, + "grad_norm": 0.5543130040168762, + "learning_rate": 0.00019698417725318892, + "loss": 0.9682, + "step": 1779 + }, + { + "epoch": 0.31695156695156695, + "grad_norm": 0.49534836411476135, + "learning_rate": 0.00019698076462387753, + "loss": 1.107, + "step": 1780 + }, + { + "epoch": 0.31712962962962965, + "grad_norm": 0.48844948410987854, + "learning_rate": 0.00019697735009442956, + "loss": 1.1295, + "step": 1781 + }, + { + "epoch": 0.3173076923076923, + "grad_norm": 0.5070686936378479, + "learning_rate": 0.00019697393366491185, + "loss": 1.083, + "step": 1782 + }, + { + "epoch": 0.317485754985755, + "grad_norm": 0.47817620635032654, + "learning_rate": 0.00019697051533539134, + "loss": 1.3014, + "step": 1783 + }, + { + "epoch": 0.3176638176638177, + "grad_norm": 0.538488507270813, + "learning_rate": 0.00019696709510593502, + "loss": 1.0354, + "step": 1784 + }, + { + "epoch": 0.3178418803418803, + "grad_norm": 0.5141439437866211, + "learning_rate": 0.0001969636729766099, + "loss": 1.2912, + "step": 1785 + }, + { + "epoch": 0.318019943019943, + "grad_norm": 0.5009665489196777, + "learning_rate": 0.00019696024894748306, + "loss": 0.9014, + "step": 1786 + }, + { + "epoch": 0.3181980056980057, + "grad_norm": 0.46199744939804077, + "learning_rate": 0.00019695682301862155, + "loss": 1.0532, + "step": 1787 + }, + { + "epoch": 0.31837606837606836, + "grad_norm": 0.4649423062801361, + "learning_rate": 0.0001969533951900925, + "loss": 0.8608, + "step": 1788 + }, + { + "epoch": 0.31855413105413105, + "grad_norm": 0.516909658908844, + "learning_rate": 0.0001969499654619631, + "loss": 1.1385, + "step": 1789 + }, + { + "epoch": 0.31873219373219375, + "grad_norm": 0.46016669273376465, + "learning_rate": 0.00019694653383430048, + "loss": 0.9168, + "step": 1790 + }, + { + "epoch": 0.3189102564102564, + "grad_norm": 0.4794938564300537, + "learning_rate": 0.00019694310030717193, + "loss": 1.0244, + "step": 1791 + }, + { + "epoch": 0.3190883190883191, + "grad_norm": 0.46577662229537964, + "learning_rate": 0.00019693966488064471, + "loss": 1.0954, + "step": 1792 + }, + { + "epoch": 0.3192663817663818, + "grad_norm": 0.4866746962070465, + "learning_rate": 0.00019693622755478614, + "loss": 1.2925, + "step": 1793 + }, + { + "epoch": 0.3194444444444444, + "grad_norm": 0.4841702878475189, + "learning_rate": 0.00019693278832966357, + "loss": 1.119, + "step": 1794 + }, + { + "epoch": 0.3196225071225071, + "grad_norm": 0.4835243821144104, + "learning_rate": 0.00019692934720534435, + "loss": 1.1702, + "step": 1795 + }, + { + "epoch": 0.3198005698005698, + "grad_norm": 0.5200608968734741, + "learning_rate": 0.00019692590418189594, + "loss": 1.1989, + "step": 1796 + }, + { + "epoch": 0.31997863247863245, + "grad_norm": 0.5147821307182312, + "learning_rate": 0.00019692245925938577, + "loss": 1.1417, + "step": 1797 + }, + { + "epoch": 0.32015669515669515, + "grad_norm": 0.5145614743232727, + "learning_rate": 0.00019691901243788136, + "loss": 1.0571, + "step": 1798 + }, + { + "epoch": 0.32033475783475784, + "grad_norm": 0.5416026711463928, + "learning_rate": 0.00019691556371745022, + "loss": 1.188, + "step": 1799 + }, + { + "epoch": 0.32051282051282054, + "grad_norm": 0.5140644311904907, + "learning_rate": 0.00019691211309815995, + "loss": 1.1795, + "step": 1800 + }, + { + "epoch": 0.3206908831908832, + "grad_norm": 0.44219106435775757, + "learning_rate": 0.00019690866058007817, + "loss": 0.9215, + "step": 1801 + }, + { + "epoch": 0.3208689458689459, + "grad_norm": 0.49523603916168213, + "learning_rate": 0.00019690520616327245, + "loss": 1.1117, + "step": 1802 + }, + { + "epoch": 0.32104700854700857, + "grad_norm": 0.5818293690681458, + "learning_rate": 0.0001969017498478105, + "loss": 1.16, + "step": 1803 + }, + { + "epoch": 0.3212250712250712, + "grad_norm": 0.5175749659538269, + "learning_rate": 0.0001968982916337601, + "loss": 1.1999, + "step": 1804 + }, + { + "epoch": 0.3214031339031339, + "grad_norm": 0.49916017055511475, + "learning_rate": 0.00019689483152118898, + "loss": 0.9505, + "step": 1805 + }, + { + "epoch": 0.3215811965811966, + "grad_norm": 0.46849536895751953, + "learning_rate": 0.00019689136951016488, + "loss": 0.9627, + "step": 1806 + }, + { + "epoch": 0.32175925925925924, + "grad_norm": 0.4226818382740021, + "learning_rate": 0.00019688790560075568, + "loss": 1.037, + "step": 1807 + }, + { + "epoch": 0.32193732193732194, + "grad_norm": 0.4697103798389435, + "learning_rate": 0.00019688443979302923, + "loss": 1.1431, + "step": 1808 + }, + { + "epoch": 0.32211538461538464, + "grad_norm": 0.4999365508556366, + "learning_rate": 0.00019688097208705343, + "loss": 1.171, + "step": 1809 + }, + { + "epoch": 0.3222934472934473, + "grad_norm": 0.5229731798171997, + "learning_rate": 0.00019687750248289625, + "loss": 1.3395, + "step": 1810 + }, + { + "epoch": 0.32247150997151, + "grad_norm": 0.512525737285614, + "learning_rate": 0.00019687403098062566, + "loss": 1.1438, + "step": 1811 + }, + { + "epoch": 0.32264957264957267, + "grad_norm": 0.4558548927307129, + "learning_rate": 0.00019687055758030967, + "loss": 1.0012, + "step": 1812 + }, + { + "epoch": 0.3228276353276353, + "grad_norm": 0.45195743441581726, + "learning_rate": 0.00019686708228201636, + "loss": 1.0222, + "step": 1813 + }, + { + "epoch": 0.323005698005698, + "grad_norm": 0.5023126602172852, + "learning_rate": 0.00019686360508581373, + "loss": 1.2128, + "step": 1814 + }, + { + "epoch": 0.3231837606837607, + "grad_norm": 0.46516045928001404, + "learning_rate": 0.00019686012599177003, + "loss": 0.989, + "step": 1815 + }, + { + "epoch": 0.32336182336182334, + "grad_norm": 0.4142672121524811, + "learning_rate": 0.00019685664499995338, + "loss": 1.0144, + "step": 1816 + }, + { + "epoch": 0.32353988603988604, + "grad_norm": 0.4511009752750397, + "learning_rate": 0.0001968531621104319, + "loss": 0.885, + "step": 1817 + }, + { + "epoch": 0.32371794871794873, + "grad_norm": 0.49583545327186584, + "learning_rate": 0.00019684967732327396, + "loss": 1.0986, + "step": 1818 + }, + { + "epoch": 0.3238960113960114, + "grad_norm": 0.5872161388397217, + "learning_rate": 0.0001968461906385478, + "loss": 1.1482, + "step": 1819 + }, + { + "epoch": 0.32407407407407407, + "grad_norm": 0.4509563148021698, + "learning_rate": 0.00019684270205632168, + "loss": 1.0578, + "step": 1820 + }, + { + "epoch": 0.32425213675213677, + "grad_norm": 0.501345157623291, + "learning_rate": 0.00019683921157666402, + "loss": 1.1792, + "step": 1821 + }, + { + "epoch": 0.3244301994301994, + "grad_norm": 0.48257577419281006, + "learning_rate": 0.00019683571919964314, + "loss": 1.0448, + "step": 1822 + }, + { + "epoch": 0.3246082621082621, + "grad_norm": 0.5399422645568848, + "learning_rate": 0.00019683222492532752, + "loss": 1.0579, + "step": 1823 + }, + { + "epoch": 0.3247863247863248, + "grad_norm": 0.4382506012916565, + "learning_rate": 0.0001968287287537856, + "loss": 1.0246, + "step": 1824 + }, + { + "epoch": 0.32496438746438744, + "grad_norm": 0.49247491359710693, + "learning_rate": 0.00019682523068508586, + "loss": 1.318, + "step": 1825 + }, + { + "epoch": 0.32514245014245013, + "grad_norm": 0.49067625403404236, + "learning_rate": 0.0001968217307192969, + "loss": 1.1028, + "step": 1826 + }, + { + "epoch": 0.32532051282051283, + "grad_norm": 0.4832286238670349, + "learning_rate": 0.00019681822885648723, + "loss": 1.0996, + "step": 1827 + }, + { + "epoch": 0.32549857549857547, + "grad_norm": 0.47144386172294617, + "learning_rate": 0.0001968147250967255, + "loss": 1.0707, + "step": 1828 + }, + { + "epoch": 0.32567663817663817, + "grad_norm": 0.46299225091934204, + "learning_rate": 0.0001968112194400803, + "loss": 1.0461, + "step": 1829 + }, + { + "epoch": 0.32585470085470086, + "grad_norm": 0.4880816340446472, + "learning_rate": 0.00019680771188662044, + "loss": 1.1198, + "step": 1830 + }, + { + "epoch": 0.32603276353276356, + "grad_norm": 0.43837276101112366, + "learning_rate": 0.00019680420243641452, + "loss": 1.0599, + "step": 1831 + }, + { + "epoch": 0.3262108262108262, + "grad_norm": 0.453168660402298, + "learning_rate": 0.0001968006910895314, + "loss": 1.0327, + "step": 1832 + }, + { + "epoch": 0.3263888888888889, + "grad_norm": 0.45183828473091125, + "learning_rate": 0.00019679717784603975, + "loss": 1.1381, + "step": 1833 + }, + { + "epoch": 0.3265669515669516, + "grad_norm": 0.5326765775680542, + "learning_rate": 0.00019679366270600852, + "loss": 1.3169, + "step": 1834 + }, + { + "epoch": 0.32674501424501423, + "grad_norm": 0.47468429803848267, + "learning_rate": 0.00019679014566950653, + "loss": 1.1816, + "step": 1835 + }, + { + "epoch": 0.3269230769230769, + "grad_norm": 0.5096879005432129, + "learning_rate": 0.0001967866267366027, + "loss": 1.1162, + "step": 1836 + }, + { + "epoch": 0.3271011396011396, + "grad_norm": 0.491514652967453, + "learning_rate": 0.00019678310590736598, + "loss": 1.2793, + "step": 1837 + }, + { + "epoch": 0.32727920227920226, + "grad_norm": 0.601439356803894, + "learning_rate": 0.00019677958318186533, + "loss": 0.9851, + "step": 1838 + }, + { + "epoch": 0.32745726495726496, + "grad_norm": 0.45270970463752747, + "learning_rate": 0.0001967760585601698, + "loss": 1.0042, + "step": 1839 + }, + { + "epoch": 0.32763532763532766, + "grad_norm": 0.48864325881004333, + "learning_rate": 0.00019677253204234847, + "loss": 1.0835, + "step": 1840 + }, + { + "epoch": 0.3278133903133903, + "grad_norm": 0.5855685472488403, + "learning_rate": 0.00019676900362847037, + "loss": 1.193, + "step": 1841 + }, + { + "epoch": 0.327991452991453, + "grad_norm": 0.7181013822555542, + "learning_rate": 0.00019676547331860466, + "loss": 1.2028, + "step": 1842 + }, + { + "epoch": 0.3281695156695157, + "grad_norm": 0.4517378807067871, + "learning_rate": 0.00019676194111282054, + "loss": 1.013, + "step": 1843 + }, + { + "epoch": 0.32834757834757833, + "grad_norm": 0.5477756857872009, + "learning_rate": 0.00019675840701118718, + "loss": 1.2311, + "step": 1844 + }, + { + "epoch": 0.328525641025641, + "grad_norm": 0.5194997191429138, + "learning_rate": 0.00019675487101377382, + "loss": 1.0953, + "step": 1845 + }, + { + "epoch": 0.3287037037037037, + "grad_norm": 0.44454067945480347, + "learning_rate": 0.00019675133312064977, + "loss": 0.8505, + "step": 1846 + }, + { + "epoch": 0.32888176638176636, + "grad_norm": 0.3938713073730469, + "learning_rate": 0.00019674779333188428, + "loss": 0.8525, + "step": 1847 + }, + { + "epoch": 0.32905982905982906, + "grad_norm": 0.4927884340286255, + "learning_rate": 0.00019674425164754682, + "loss": 1.2477, + "step": 1848 + }, + { + "epoch": 0.32923789173789175, + "grad_norm": 0.4516635239124298, + "learning_rate": 0.0001967407080677067, + "loss": 0.8333, + "step": 1849 + }, + { + "epoch": 0.3294159544159544, + "grad_norm": 0.47105780243873596, + "learning_rate": 0.00019673716259243336, + "loss": 1.0989, + "step": 1850 + }, + { + "epoch": 0.3295940170940171, + "grad_norm": 0.5192127823829651, + "learning_rate": 0.00019673361522179627, + "loss": 1.1164, + "step": 1851 + }, + { + "epoch": 0.3297720797720798, + "grad_norm": 0.5222696661949158, + "learning_rate": 0.00019673006595586495, + "loss": 1.3191, + "step": 1852 + }, + { + "epoch": 0.3299501424501424, + "grad_norm": 0.6046679019927979, + "learning_rate": 0.0001967265147947089, + "loss": 0.9782, + "step": 1853 + }, + { + "epoch": 0.3301282051282051, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.00019672296173839775, + "loss": 1.2247, + "step": 1854 + }, + { + "epoch": 0.3303062678062678, + "grad_norm": 0.5435982346534729, + "learning_rate": 0.00019671940678700107, + "loss": 1.1647, + "step": 1855 + }, + { + "epoch": 0.33048433048433046, + "grad_norm": 0.46878984570503235, + "learning_rate": 0.00019671584994058856, + "loss": 1.132, + "step": 1856 + }, + { + "epoch": 0.33066239316239315, + "grad_norm": 0.5336877107620239, + "learning_rate": 0.00019671229119922986, + "loss": 1.0583, + "step": 1857 + }, + { + "epoch": 0.33084045584045585, + "grad_norm": 0.4811093807220459, + "learning_rate": 0.0001967087305629947, + "loss": 1.0089, + "step": 1858 + }, + { + "epoch": 0.33101851851851855, + "grad_norm": 0.5140184760093689, + "learning_rate": 0.0001967051680319529, + "loss": 1.2335, + "step": 1859 + }, + { + "epoch": 0.3311965811965812, + "grad_norm": 0.5855883955955505, + "learning_rate": 0.00019670160360617418, + "loss": 1.1107, + "step": 1860 + }, + { + "epoch": 0.3313746438746439, + "grad_norm": 0.5081531405448914, + "learning_rate": 0.00019669803728572844, + "loss": 1.0669, + "step": 1861 + }, + { + "epoch": 0.3315527065527066, + "grad_norm": 0.48749417066574097, + "learning_rate": 0.0001966944690706855, + "loss": 1.1465, + "step": 1862 + }, + { + "epoch": 0.3317307692307692, + "grad_norm": 0.5175687670707703, + "learning_rate": 0.00019669089896111536, + "loss": 1.254, + "step": 1863 + }, + { + "epoch": 0.3319088319088319, + "grad_norm": 0.4198860824108124, + "learning_rate": 0.0001966873269570879, + "loss": 0.9811, + "step": 1864 + }, + { + "epoch": 0.3320868945868946, + "grad_norm": 0.5220273733139038, + "learning_rate": 0.0001966837530586731, + "loss": 1.277, + "step": 1865 + }, + { + "epoch": 0.33226495726495725, + "grad_norm": 0.551954448223114, + "learning_rate": 0.00019668017726594101, + "loss": 1.0627, + "step": 1866 + }, + { + "epoch": 0.33244301994301995, + "grad_norm": 0.5289301872253418, + "learning_rate": 0.00019667659957896166, + "loss": 1.4525, + "step": 1867 + }, + { + "epoch": 0.33262108262108264, + "grad_norm": 0.5190161466598511, + "learning_rate": 0.00019667301999780522, + "loss": 1.1064, + "step": 1868 + }, + { + "epoch": 0.3327991452991453, + "grad_norm": 0.437637060880661, + "learning_rate": 0.00019666943852254172, + "loss": 1.1304, + "step": 1869 + }, + { + "epoch": 0.332977207977208, + "grad_norm": 0.4801286458969116, + "learning_rate": 0.00019666585515324138, + "loss": 1.032, + "step": 1870 + }, + { + "epoch": 0.3331552706552707, + "grad_norm": 0.5041908621788025, + "learning_rate": 0.00019666226988997445, + "loss": 1.2611, + "step": 1871 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.4529375731945038, + "learning_rate": 0.00019665868273281115, + "loss": 1.1346, + "step": 1872 + }, + { + "epoch": 0.333511396011396, + "grad_norm": 0.4797019064426422, + "learning_rate": 0.00019665509368182172, + "loss": 1.1716, + "step": 1873 + }, + { + "epoch": 0.3336894586894587, + "grad_norm": 0.5505055785179138, + "learning_rate": 0.00019665150273707652, + "loss": 0.9729, + "step": 1874 + }, + { + "epoch": 0.33386752136752135, + "grad_norm": 0.4228051006793976, + "learning_rate": 0.00019664790989864592, + "loss": 0.9023, + "step": 1875 + }, + { + "epoch": 0.33404558404558404, + "grad_norm": 0.4926959276199341, + "learning_rate": 0.00019664431516660028, + "loss": 1.0999, + "step": 1876 + }, + { + "epoch": 0.33422364672364674, + "grad_norm": 0.4273219704627991, + "learning_rate": 0.00019664071854101005, + "loss": 1.1039, + "step": 1877 + }, + { + "epoch": 0.3344017094017094, + "grad_norm": 0.48438936471939087, + "learning_rate": 0.00019663712002194566, + "loss": 1.1308, + "step": 1878 + }, + { + "epoch": 0.3345797720797721, + "grad_norm": 0.5102053284645081, + "learning_rate": 0.0001966335196094777, + "loss": 1.0618, + "step": 1879 + }, + { + "epoch": 0.33475783475783477, + "grad_norm": 0.4357300400733948, + "learning_rate": 0.00019662991730367663, + "loss": 1.0521, + "step": 1880 + }, + { + "epoch": 0.3349358974358974, + "grad_norm": 0.5052695870399475, + "learning_rate": 0.00019662631310461308, + "loss": 0.9579, + "step": 1881 + }, + { + "epoch": 0.3351139601139601, + "grad_norm": 0.4889117181301117, + "learning_rate": 0.00019662270701235762, + "loss": 1.0304, + "step": 1882 + }, + { + "epoch": 0.3352920227920228, + "grad_norm": 0.4671195149421692, + "learning_rate": 0.000196619099026981, + "loss": 1.2228, + "step": 1883 + }, + { + "epoch": 0.33547008547008544, + "grad_norm": 0.4700174331665039, + "learning_rate": 0.0001966154891485538, + "loss": 0.9634, + "step": 1884 + }, + { + "epoch": 0.33564814814814814, + "grad_norm": 0.488817423582077, + "learning_rate": 0.00019661187737714676, + "loss": 1.2499, + "step": 1885 + }, + { + "epoch": 0.33582621082621084, + "grad_norm": 0.5336169600486755, + "learning_rate": 0.00019660826371283073, + "loss": 1.251, + "step": 1886 + }, + { + "epoch": 0.33600427350427353, + "grad_norm": 0.5054540038108826, + "learning_rate": 0.00019660464815567642, + "loss": 1.221, + "step": 1887 + }, + { + "epoch": 0.33618233618233617, + "grad_norm": 0.5078747868537903, + "learning_rate": 0.00019660103070575472, + "loss": 0.9792, + "step": 1888 + }, + { + "epoch": 0.33636039886039887, + "grad_norm": 0.498571515083313, + "learning_rate": 0.0001965974113631365, + "loss": 1.1682, + "step": 1889 + }, + { + "epoch": 0.33653846153846156, + "grad_norm": 0.49969518184661865, + "learning_rate": 0.00019659379012789264, + "loss": 1.0012, + "step": 1890 + }, + { + "epoch": 0.3367165242165242, + "grad_norm": 0.4238094687461853, + "learning_rate": 0.00019659016700009416, + "loss": 1.0455, + "step": 1891 + }, + { + "epoch": 0.3368945868945869, + "grad_norm": 0.5139104723930359, + "learning_rate": 0.000196586541979812, + "loss": 0.9979, + "step": 1892 + }, + { + "epoch": 0.3370726495726496, + "grad_norm": 0.5446547269821167, + "learning_rate": 0.00019658291506711715, + "loss": 0.9271, + "step": 1893 + }, + { + "epoch": 0.33725071225071224, + "grad_norm": 0.5284572839736938, + "learning_rate": 0.00019657928626208077, + "loss": 1.0356, + "step": 1894 + }, + { + "epoch": 0.33742877492877493, + "grad_norm": 0.49936217069625854, + "learning_rate": 0.00019657565556477387, + "loss": 0.9785, + "step": 1895 + }, + { + "epoch": 0.33760683760683763, + "grad_norm": 0.4678729772567749, + "learning_rate": 0.00019657202297526763, + "loss": 1.2135, + "step": 1896 + }, + { + "epoch": 0.33778490028490027, + "grad_norm": 0.46844249963760376, + "learning_rate": 0.0001965683884936332, + "loss": 0.9369, + "step": 1897 + }, + { + "epoch": 0.33796296296296297, + "grad_norm": 0.4307389557361603, + "learning_rate": 0.0001965647521199418, + "loss": 0.9301, + "step": 1898 + }, + { + "epoch": 0.33814102564102566, + "grad_norm": 0.48227834701538086, + "learning_rate": 0.00019656111385426468, + "loss": 1.3169, + "step": 1899 + }, + { + "epoch": 0.3383190883190883, + "grad_norm": 0.45860713720321655, + "learning_rate": 0.00019655747369667315, + "loss": 0.9835, + "step": 1900 + }, + { + "epoch": 0.338497150997151, + "grad_norm": 0.5522414445877075, + "learning_rate": 0.00019655383164723846, + "loss": 1.363, + "step": 1901 + }, + { + "epoch": 0.3386752136752137, + "grad_norm": 0.5283710360527039, + "learning_rate": 0.000196550187706032, + "loss": 1.1499, + "step": 1902 + }, + { + "epoch": 0.33885327635327633, + "grad_norm": 0.4419134259223938, + "learning_rate": 0.00019654654187312525, + "loss": 1.2039, + "step": 1903 + }, + { + "epoch": 0.33903133903133903, + "grad_norm": 0.49066096544265747, + "learning_rate": 0.00019654289414858952, + "loss": 0.9707, + "step": 1904 + }, + { + "epoch": 0.3392094017094017, + "grad_norm": 0.4619338810443878, + "learning_rate": 0.00019653924453249633, + "loss": 1.0849, + "step": 1905 + }, + { + "epoch": 0.33938746438746437, + "grad_norm": 0.5191119313240051, + "learning_rate": 0.0001965355930249172, + "loss": 1.1387, + "step": 1906 + }, + { + "epoch": 0.33956552706552706, + "grad_norm": 0.5245711207389832, + "learning_rate": 0.00019653193962592368, + "loss": 1.3435, + "step": 1907 + }, + { + "epoch": 0.33974358974358976, + "grad_norm": 0.49562904238700867, + "learning_rate": 0.0001965282843355873, + "loss": 1.2781, + "step": 1908 + }, + { + "epoch": 0.3399216524216524, + "grad_norm": 0.4661353826522827, + "learning_rate": 0.0001965246271539797, + "loss": 0.9317, + "step": 1909 + }, + { + "epoch": 0.3400997150997151, + "grad_norm": 0.4723222851753235, + "learning_rate": 0.00019652096808117254, + "loss": 1.0733, + "step": 1910 + }, + { + "epoch": 0.3402777777777778, + "grad_norm": 0.4358505308628082, + "learning_rate": 0.00019651730711723754, + "loss": 1.1461, + "step": 1911 + }, + { + "epoch": 0.34045584045584043, + "grad_norm": 0.462422251701355, + "learning_rate": 0.00019651364426224638, + "loss": 1.0914, + "step": 1912 + }, + { + "epoch": 0.3406339031339031, + "grad_norm": 0.47952914237976074, + "learning_rate": 0.0001965099795162709, + "loss": 1.0392, + "step": 1913 + }, + { + "epoch": 0.3408119658119658, + "grad_norm": 0.5036373734474182, + "learning_rate": 0.00019650631287938282, + "loss": 1.4002, + "step": 1914 + }, + { + "epoch": 0.34099002849002846, + "grad_norm": 0.5130090713500977, + "learning_rate": 0.000196502644351654, + "loss": 1.3499, + "step": 1915 + }, + { + "epoch": 0.34116809116809116, + "grad_norm": 0.4426332414150238, + "learning_rate": 0.00019649897393315635, + "loss": 1.0726, + "step": 1916 + }, + { + "epoch": 0.34134615384615385, + "grad_norm": 0.5580727458000183, + "learning_rate": 0.00019649530162396176, + "loss": 1.1164, + "step": 1917 + }, + { + "epoch": 0.34152421652421655, + "grad_norm": 0.545001745223999, + "learning_rate": 0.00019649162742414218, + "loss": 0.962, + "step": 1918 + }, + { + "epoch": 0.3417022792022792, + "grad_norm": 0.5225808024406433, + "learning_rate": 0.00019648795133376962, + "loss": 1.1415, + "step": 1919 + }, + { + "epoch": 0.3418803418803419, + "grad_norm": 0.48210129141807556, + "learning_rate": 0.0001964842733529161, + "loss": 1.1188, + "step": 1920 + }, + { + "epoch": 0.3420584045584046, + "grad_norm": 0.4515395164489746, + "learning_rate": 0.00019648059348165365, + "loss": 1.0828, + "step": 1921 + }, + { + "epoch": 0.3422364672364672, + "grad_norm": 0.5802633166313171, + "learning_rate": 0.0001964769117200544, + "loss": 1.3137, + "step": 1922 + }, + { + "epoch": 0.3424145299145299, + "grad_norm": 0.4432032108306885, + "learning_rate": 0.00019647322806819046, + "loss": 1.0523, + "step": 1923 + }, + { + "epoch": 0.3425925925925926, + "grad_norm": 0.4697614908218384, + "learning_rate": 0.00019646954252613402, + "loss": 0.8426, + "step": 1924 + }, + { + "epoch": 0.34277065527065526, + "grad_norm": 0.4610968232154846, + "learning_rate": 0.0001964658550939573, + "loss": 0.9826, + "step": 1925 + }, + { + "epoch": 0.34294871794871795, + "grad_norm": 0.5278257727622986, + "learning_rate": 0.00019646216577173258, + "loss": 1.1064, + "step": 1926 + }, + { + "epoch": 0.34312678062678065, + "grad_norm": 0.5686144232749939, + "learning_rate": 0.00019645847455953205, + "loss": 0.9138, + "step": 1927 + }, + { + "epoch": 0.3433048433048433, + "grad_norm": 0.42894792556762695, + "learning_rate": 0.0001964547814574281, + "loss": 1.0461, + "step": 1928 + }, + { + "epoch": 0.343482905982906, + "grad_norm": 0.5567317605018616, + "learning_rate": 0.0001964510864654931, + "loss": 0.8787, + "step": 1929 + }, + { + "epoch": 0.3436609686609687, + "grad_norm": 0.5015586614608765, + "learning_rate": 0.0001964473895837994, + "loss": 1.1406, + "step": 1930 + }, + { + "epoch": 0.3438390313390313, + "grad_norm": 0.47391530871391296, + "learning_rate": 0.00019644369081241948, + "loss": 1.0685, + "step": 1931 + }, + { + "epoch": 0.344017094017094, + "grad_norm": 0.546037495136261, + "learning_rate": 0.00019643999015142574, + "loss": 1.2349, + "step": 1932 + }, + { + "epoch": 0.3441951566951567, + "grad_norm": 0.4724953770637512, + "learning_rate": 0.00019643628760089078, + "loss": 1.0621, + "step": 1933 + }, + { + "epoch": 0.34437321937321935, + "grad_norm": 0.5644593834877014, + "learning_rate": 0.00019643258316088703, + "loss": 1.2559, + "step": 1934 + }, + { + "epoch": 0.34455128205128205, + "grad_norm": 0.500815749168396, + "learning_rate": 0.00019642887683148718, + "loss": 1.0439, + "step": 1935 + }, + { + "epoch": 0.34472934472934474, + "grad_norm": 0.4932316541671753, + "learning_rate": 0.0001964251686127638, + "loss": 1.0404, + "step": 1936 + }, + { + "epoch": 0.3449074074074074, + "grad_norm": 0.48494651913642883, + "learning_rate": 0.00019642145850478954, + "loss": 0.9951, + "step": 1937 + }, + { + "epoch": 0.3450854700854701, + "grad_norm": 0.5191963315010071, + "learning_rate": 0.00019641774650763706, + "loss": 1.1258, + "step": 1938 + }, + { + "epoch": 0.3452635327635328, + "grad_norm": 0.4439312815666199, + "learning_rate": 0.00019641403262137918, + "loss": 1.1158, + "step": 1939 + }, + { + "epoch": 0.3454415954415954, + "grad_norm": 0.4829137921333313, + "learning_rate": 0.0001964103168460886, + "loss": 1.0531, + "step": 1940 + }, + { + "epoch": 0.3456196581196581, + "grad_norm": 0.49433329701423645, + "learning_rate": 0.00019640659918183811, + "loss": 1.1295, + "step": 1941 + }, + { + "epoch": 0.3457977207977208, + "grad_norm": 0.5351347923278809, + "learning_rate": 0.00019640287962870062, + "loss": 1.2379, + "step": 1942 + }, + { + "epoch": 0.34597578347578345, + "grad_norm": 0.4845680892467499, + "learning_rate": 0.00019639915818674895, + "loss": 1.0197, + "step": 1943 + }, + { + "epoch": 0.34615384615384615, + "grad_norm": 0.5312514901161194, + "learning_rate": 0.00019639543485605604, + "loss": 0.9734, + "step": 1944 + }, + { + "epoch": 0.34633190883190884, + "grad_norm": 0.4571874737739563, + "learning_rate": 0.00019639170963669478, + "loss": 1.1012, + "step": 1945 + }, + { + "epoch": 0.34650997150997154, + "grad_norm": 0.4449031949043274, + "learning_rate": 0.00019638798252873824, + "loss": 1.1393, + "step": 1946 + }, + { + "epoch": 0.3466880341880342, + "grad_norm": 0.47470834851264954, + "learning_rate": 0.0001963842535322594, + "loss": 0.981, + "step": 1947 + }, + { + "epoch": 0.3468660968660969, + "grad_norm": 0.5386981964111328, + "learning_rate": 0.00019638052264733132, + "loss": 1.1247, + "step": 1948 + }, + { + "epoch": 0.34704415954415957, + "grad_norm": 0.535589873790741, + "learning_rate": 0.00019637678987402714, + "loss": 1.3157, + "step": 1949 + }, + { + "epoch": 0.3472222222222222, + "grad_norm": 0.49338245391845703, + "learning_rate": 0.00019637305521242, + "loss": 1.1066, + "step": 1950 + }, + { + "epoch": 0.3474002849002849, + "grad_norm": 0.4247688353061676, + "learning_rate": 0.00019636931866258298, + "loss": 1.0039, + "step": 1951 + }, + { + "epoch": 0.3475783475783476, + "grad_norm": 0.5351517200469971, + "learning_rate": 0.00019636558022458934, + "loss": 1.0344, + "step": 1952 + }, + { + "epoch": 0.34775641025641024, + "grad_norm": 0.4633362889289856, + "learning_rate": 0.00019636183989851238, + "loss": 1.1383, + "step": 1953 + }, + { + "epoch": 0.34793447293447294, + "grad_norm": 0.553709089756012, + "learning_rate": 0.00019635809768442535, + "loss": 1.0389, + "step": 1954 + }, + { + "epoch": 0.34811253561253563, + "grad_norm": 0.479374498128891, + "learning_rate": 0.00019635435358240154, + "loss": 1.1774, + "step": 1955 + }, + { + "epoch": 0.3482905982905983, + "grad_norm": 0.5274081230163574, + "learning_rate": 0.0001963506075925143, + "loss": 1.1809, + "step": 1956 + }, + { + "epoch": 0.34846866096866097, + "grad_norm": 0.45398542284965515, + "learning_rate": 0.0001963468597148371, + "loss": 1.0502, + "step": 1957 + }, + { + "epoch": 0.34864672364672367, + "grad_norm": 0.48201611638069153, + "learning_rate": 0.00019634310994944332, + "loss": 1.0557, + "step": 1958 + }, + { + "epoch": 0.3488247863247863, + "grad_norm": 0.6407544016838074, + "learning_rate": 0.00019633935829640642, + "loss": 1.2138, + "step": 1959 + }, + { + "epoch": 0.349002849002849, + "grad_norm": 0.5385687351226807, + "learning_rate": 0.00019633560475579995, + "loss": 1.3496, + "step": 1960 + }, + { + "epoch": 0.3491809116809117, + "grad_norm": 0.5260964035987854, + "learning_rate": 0.0001963318493276974, + "loss": 1.0253, + "step": 1961 + }, + { + "epoch": 0.34935897435897434, + "grad_norm": 0.48478585481643677, + "learning_rate": 0.00019632809201217238, + "loss": 1.137, + "step": 1962 + }, + { + "epoch": 0.34953703703703703, + "grad_norm": 0.620033860206604, + "learning_rate": 0.0001963243328092985, + "loss": 1.3445, + "step": 1963 + }, + { + "epoch": 0.34971509971509973, + "grad_norm": 0.5149700045585632, + "learning_rate": 0.00019632057171914942, + "loss": 1.1042, + "step": 1964 + }, + { + "epoch": 0.34989316239316237, + "grad_norm": 0.42695048451423645, + "learning_rate": 0.0001963168087417988, + "loss": 0.8789, + "step": 1965 + }, + { + "epoch": 0.35007122507122507, + "grad_norm": 0.5281283855438232, + "learning_rate": 0.00019631304387732044, + "loss": 1.1155, + "step": 1966 + }, + { + "epoch": 0.35024928774928776, + "grad_norm": 0.4994089901447296, + "learning_rate": 0.00019630927712578804, + "loss": 1.1226, + "step": 1967 + }, + { + "epoch": 0.3504273504273504, + "grad_norm": 0.4433288276195526, + "learning_rate": 0.0001963055084872754, + "loss": 1.0262, + "step": 1968 + }, + { + "epoch": 0.3506054131054131, + "grad_norm": 0.46541857719421387, + "learning_rate": 0.0001963017379618564, + "loss": 1.1438, + "step": 1969 + }, + { + "epoch": 0.3507834757834758, + "grad_norm": 0.5097604393959045, + "learning_rate": 0.00019629796554960488, + "loss": 0.9641, + "step": 1970 + }, + { + "epoch": 0.35096153846153844, + "grad_norm": 0.49461981654167175, + "learning_rate": 0.00019629419125059478, + "loss": 1.1765, + "step": 1971 + }, + { + "epoch": 0.35113960113960113, + "grad_norm": 0.4763339161872864, + "learning_rate": 0.00019629041506490005, + "loss": 1.0527, + "step": 1972 + }, + { + "epoch": 0.35131766381766383, + "grad_norm": 0.4528443217277527, + "learning_rate": 0.00019628663699259463, + "loss": 1.1409, + "step": 1973 + }, + { + "epoch": 0.35149572649572647, + "grad_norm": 0.4436309039592743, + "learning_rate": 0.00019628285703375258, + "loss": 1.0459, + "step": 1974 + }, + { + "epoch": 0.35167378917378916, + "grad_norm": 0.5146129727363586, + "learning_rate": 0.00019627907518844797, + "loss": 1.2527, + "step": 1975 + }, + { + "epoch": 0.35185185185185186, + "grad_norm": 0.5202171802520752, + "learning_rate": 0.0001962752914567549, + "loss": 1.226, + "step": 1976 + }, + { + "epoch": 0.35202991452991456, + "grad_norm": 0.5267411470413208, + "learning_rate": 0.00019627150583874747, + "loss": 1.0898, + "step": 1977 + }, + { + "epoch": 0.3522079772079772, + "grad_norm": 0.546840250492096, + "learning_rate": 0.00019626771833449987, + "loss": 1.1716, + "step": 1978 + }, + { + "epoch": 0.3523860398860399, + "grad_norm": 0.5525290966033936, + "learning_rate": 0.0001962639289440863, + "loss": 1.1762, + "step": 1979 + }, + { + "epoch": 0.3525641025641026, + "grad_norm": 0.48967215418815613, + "learning_rate": 0.000196260137667581, + "loss": 1.1884, + "step": 1980 + }, + { + "epoch": 0.35274216524216523, + "grad_norm": 0.5908235907554626, + "learning_rate": 0.0001962563445050583, + "loss": 1.1887, + "step": 1981 + }, + { + "epoch": 0.3529202279202279, + "grad_norm": 0.46708086133003235, + "learning_rate": 0.00019625254945659245, + "loss": 0.8842, + "step": 1982 + }, + { + "epoch": 0.3530982905982906, + "grad_norm": 0.41652458906173706, + "learning_rate": 0.00019624875252225788, + "loss": 1.0268, + "step": 1983 + }, + { + "epoch": 0.35327635327635326, + "grad_norm": 0.5084529519081116, + "learning_rate": 0.00019624495370212892, + "loss": 1.0547, + "step": 1984 + }, + { + "epoch": 0.35345441595441596, + "grad_norm": 0.5667507648468018, + "learning_rate": 0.00019624115299628003, + "loss": 1.0656, + "step": 1985 + }, + { + "epoch": 0.35363247863247865, + "grad_norm": 0.5022873282432556, + "learning_rate": 0.00019623735040478568, + "loss": 1.0627, + "step": 1986 + }, + { + "epoch": 0.3538105413105413, + "grad_norm": 0.48342058062553406, + "learning_rate": 0.00019623354592772035, + "loss": 1.0976, + "step": 1987 + }, + { + "epoch": 0.353988603988604, + "grad_norm": 0.48117366433143616, + "learning_rate": 0.0001962297395651586, + "loss": 1.0515, + "step": 1988 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 0.492564857006073, + "learning_rate": 0.000196225931317175, + "loss": 1.1957, + "step": 1989 + }, + { + "epoch": 0.3543447293447293, + "grad_norm": 0.4756208658218384, + "learning_rate": 0.00019622212118384417, + "loss": 1.007, + "step": 1990 + }, + { + "epoch": 0.354522792022792, + "grad_norm": 0.581930935382843, + "learning_rate": 0.00019621830916524076, + "loss": 1.232, + "step": 1991 + }, + { + "epoch": 0.3547008547008547, + "grad_norm": 0.480064332485199, + "learning_rate": 0.00019621449526143947, + "loss": 1.2693, + "step": 1992 + }, + { + "epoch": 0.35487891737891736, + "grad_norm": 0.5679123401641846, + "learning_rate": 0.000196210679472515, + "loss": 1.2985, + "step": 1993 + }, + { + "epoch": 0.35505698005698005, + "grad_norm": 0.43757280707359314, + "learning_rate": 0.00019620686179854213, + "loss": 1.1387, + "step": 1994 + }, + { + "epoch": 0.35523504273504275, + "grad_norm": 0.4950634837150574, + "learning_rate": 0.00019620304223959566, + "loss": 1.1809, + "step": 1995 + }, + { + "epoch": 0.3554131054131054, + "grad_norm": 0.5574113726615906, + "learning_rate": 0.00019619922079575043, + "loss": 1.2434, + "step": 1996 + }, + { + "epoch": 0.3555911680911681, + "grad_norm": 0.5154930949211121, + "learning_rate": 0.00019619539746708128, + "loss": 1.1747, + "step": 1997 + }, + { + "epoch": 0.3557692307692308, + "grad_norm": 0.4377825856208801, + "learning_rate": 0.00019619157225366315, + "loss": 0.9547, + "step": 1998 + }, + { + "epoch": 0.3559472934472934, + "grad_norm": 0.530714213848114, + "learning_rate": 0.00019618774515557097, + "loss": 1.2057, + "step": 1999 + }, + { + "epoch": 0.3561253561253561, + "grad_norm": 0.5703464150428772, + "learning_rate": 0.00019618391617287978, + "loss": 1.3068, + "step": 2000 + }, + { + "epoch": 0.3563034188034188, + "grad_norm": 0.4862228333950043, + "learning_rate": 0.0001961800853056645, + "loss": 1.0077, + "step": 2001 + }, + { + "epoch": 0.35648148148148145, + "grad_norm": 0.5575395822525024, + "learning_rate": 0.00019617625255400028, + "loss": 1.03, + "step": 2002 + }, + { + "epoch": 0.35665954415954415, + "grad_norm": 0.4826279580593109, + "learning_rate": 0.0001961724179179622, + "loss": 1.268, + "step": 2003 + }, + { + "epoch": 0.35683760683760685, + "grad_norm": 0.49423274397850037, + "learning_rate": 0.00019616858139762534, + "loss": 1.1305, + "step": 2004 + }, + { + "epoch": 0.35701566951566954, + "grad_norm": 0.5208541750907898, + "learning_rate": 0.00019616474299306491, + "loss": 1.1651, + "step": 2005 + }, + { + "epoch": 0.3571937321937322, + "grad_norm": 0.5324164032936096, + "learning_rate": 0.0001961609027043561, + "loss": 1.1406, + "step": 2006 + }, + { + "epoch": 0.3573717948717949, + "grad_norm": 0.45385462045669556, + "learning_rate": 0.00019615706053157416, + "loss": 1.0716, + "step": 2007 + }, + { + "epoch": 0.3575498575498576, + "grad_norm": 0.5016173720359802, + "learning_rate": 0.00019615321647479438, + "loss": 1.0878, + "step": 2008 + }, + { + "epoch": 0.3577279202279202, + "grad_norm": 0.5073097348213196, + "learning_rate": 0.00019614937053409205, + "loss": 1.237, + "step": 2009 + }, + { + "epoch": 0.3579059829059829, + "grad_norm": 0.48880141973495483, + "learning_rate": 0.00019614552270954256, + "loss": 0.8794, + "step": 2010 + }, + { + "epoch": 0.3580840455840456, + "grad_norm": 0.43902209401130676, + "learning_rate": 0.00019614167300122126, + "loss": 0.912, + "step": 2011 + }, + { + "epoch": 0.35826210826210825, + "grad_norm": 0.42809322476387024, + "learning_rate": 0.0001961378214092036, + "loss": 0.7804, + "step": 2012 + }, + { + "epoch": 0.35844017094017094, + "grad_norm": 0.4464281499385834, + "learning_rate": 0.00019613396793356503, + "loss": 1.0004, + "step": 2013 + }, + { + "epoch": 0.35861823361823364, + "grad_norm": 0.49085676670074463, + "learning_rate": 0.00019613011257438109, + "loss": 1.1087, + "step": 2014 + }, + { + "epoch": 0.3587962962962963, + "grad_norm": 0.4997732937335968, + "learning_rate": 0.00019612625533172725, + "loss": 0.9591, + "step": 2015 + }, + { + "epoch": 0.358974358974359, + "grad_norm": 0.48442545533180237, + "learning_rate": 0.00019612239620567912, + "loss": 0.9744, + "step": 2016 + }, + { + "epoch": 0.35915242165242167, + "grad_norm": 0.4989205002784729, + "learning_rate": 0.00019611853519631233, + "loss": 0.9844, + "step": 2017 + }, + { + "epoch": 0.3593304843304843, + "grad_norm": 0.6107521653175354, + "learning_rate": 0.00019611467230370248, + "loss": 1.147, + "step": 2018 + }, + { + "epoch": 0.359508547008547, + "grad_norm": 0.5594844818115234, + "learning_rate": 0.00019611080752792535, + "loss": 1.3195, + "step": 2019 + }, + { + "epoch": 0.3596866096866097, + "grad_norm": 0.4786946475505829, + "learning_rate": 0.00019610694086905656, + "loss": 1.2108, + "step": 2020 + }, + { + "epoch": 0.35986467236467234, + "grad_norm": 0.5186030268669128, + "learning_rate": 0.0001961030723271719, + "loss": 1.0008, + "step": 2021 + }, + { + "epoch": 0.36004273504273504, + "grad_norm": 0.4520573318004608, + "learning_rate": 0.0001960992019023472, + "loss": 1.1307, + "step": 2022 + }, + { + "epoch": 0.36022079772079774, + "grad_norm": 0.4983210563659668, + "learning_rate": 0.00019609532959465823, + "loss": 1.1486, + "step": 2023 + }, + { + "epoch": 0.3603988603988604, + "grad_norm": 0.6209200024604797, + "learning_rate": 0.00019609145540418094, + "loss": 1.2566, + "step": 2024 + }, + { + "epoch": 0.3605769230769231, + "grad_norm": 0.47047603130340576, + "learning_rate": 0.00019608757933099117, + "loss": 1.1588, + "step": 2025 + }, + { + "epoch": 0.36075498575498577, + "grad_norm": 0.5147389769554138, + "learning_rate": 0.0001960837013751649, + "loss": 1.2113, + "step": 2026 + }, + { + "epoch": 0.3609330484330484, + "grad_norm": 0.45826098322868347, + "learning_rate": 0.00019607982153677808, + "loss": 1.13, + "step": 2027 + }, + { + "epoch": 0.3611111111111111, + "grad_norm": 0.5699561834335327, + "learning_rate": 0.00019607593981590675, + "loss": 1.2476, + "step": 2028 + }, + { + "epoch": 0.3612891737891738, + "grad_norm": 0.5349239110946655, + "learning_rate": 0.000196072056212627, + "loss": 1.2295, + "step": 2029 + }, + { + "epoch": 0.36146723646723644, + "grad_norm": 0.6212165355682373, + "learning_rate": 0.00019606817072701484, + "loss": 1.1965, + "step": 2030 + }, + { + "epoch": 0.36164529914529914, + "grad_norm": 0.4870990216732025, + "learning_rate": 0.00019606428335914645, + "loss": 1.4464, + "step": 2031 + }, + { + "epoch": 0.36182336182336183, + "grad_norm": 0.42427828907966614, + "learning_rate": 0.00019606039410909797, + "loss": 1.1546, + "step": 2032 + }, + { + "epoch": 0.36200142450142453, + "grad_norm": 0.5081788301467896, + "learning_rate": 0.0001960565029769456, + "loss": 1.1867, + "step": 2033 + }, + { + "epoch": 0.36217948717948717, + "grad_norm": 0.4813104271888733, + "learning_rate": 0.00019605260996276565, + "loss": 1.3726, + "step": 2034 + }, + { + "epoch": 0.36235754985754987, + "grad_norm": 0.4648851156234741, + "learning_rate": 0.0001960487150666343, + "loss": 1.2434, + "step": 2035 + }, + { + "epoch": 0.36253561253561256, + "grad_norm": 0.484161913394928, + "learning_rate": 0.00019604481828862792, + "loss": 1.1309, + "step": 2036 + }, + { + "epoch": 0.3627136752136752, + "grad_norm": 0.4929439127445221, + "learning_rate": 0.00019604091962882283, + "loss": 1.1007, + "step": 2037 + }, + { + "epoch": 0.3628917378917379, + "grad_norm": 0.45599642395973206, + "learning_rate": 0.00019603701908729544, + "loss": 1.2628, + "step": 2038 + }, + { + "epoch": 0.3630698005698006, + "grad_norm": 0.45295149087905884, + "learning_rate": 0.00019603311666412213, + "loss": 0.9808, + "step": 2039 + }, + { + "epoch": 0.36324786324786323, + "grad_norm": 0.48681163787841797, + "learning_rate": 0.00019602921235937942, + "loss": 1.0574, + "step": 2040 + }, + { + "epoch": 0.36342592592592593, + "grad_norm": 0.41232365369796753, + "learning_rate": 0.00019602530617314378, + "loss": 1.0454, + "step": 2041 + }, + { + "epoch": 0.3636039886039886, + "grad_norm": 0.46214723587036133, + "learning_rate": 0.00019602139810549174, + "loss": 0.9985, + "step": 2042 + }, + { + "epoch": 0.36378205128205127, + "grad_norm": 0.44307878613471985, + "learning_rate": 0.00019601748815649989, + "loss": 0.9683, + "step": 2043 + }, + { + "epoch": 0.36396011396011396, + "grad_norm": 0.4809451401233673, + "learning_rate": 0.00019601357632624477, + "loss": 1.028, + "step": 2044 + }, + { + "epoch": 0.36413817663817666, + "grad_norm": 0.4638497531414032, + "learning_rate": 0.0001960096626148031, + "loss": 0.9851, + "step": 2045 + }, + { + "epoch": 0.3643162393162393, + "grad_norm": 0.5942164063453674, + "learning_rate": 0.00019600574702225153, + "loss": 1.1606, + "step": 2046 + }, + { + "epoch": 0.364494301994302, + "grad_norm": 0.5171293616294861, + "learning_rate": 0.00019600182954866675, + "loss": 1.2335, + "step": 2047 + }, + { + "epoch": 0.3646723646723647, + "grad_norm": 0.5294404625892639, + "learning_rate": 0.00019599791019412558, + "loss": 1.0966, + "step": 2048 + }, + { + "epoch": 0.36485042735042733, + "grad_norm": 0.46117448806762695, + "learning_rate": 0.00019599398895870477, + "loss": 1.0565, + "step": 2049 + }, + { + "epoch": 0.36502849002849, + "grad_norm": 0.5385118126869202, + "learning_rate": 0.00019599006584248118, + "loss": 1.0076, + "step": 2050 + }, + { + "epoch": 0.3652065527065527, + "grad_norm": 0.4915166199207306, + "learning_rate": 0.00019598614084553165, + "loss": 0.9686, + "step": 2051 + }, + { + "epoch": 0.36538461538461536, + "grad_norm": 0.46769094467163086, + "learning_rate": 0.00019598221396793303, + "loss": 1.1217, + "step": 2052 + }, + { + "epoch": 0.36556267806267806, + "grad_norm": 0.5440493822097778, + "learning_rate": 0.00019597828520976236, + "loss": 1.2344, + "step": 2053 + }, + { + "epoch": 0.36574074074074076, + "grad_norm": 0.616727352142334, + "learning_rate": 0.00019597435457109657, + "loss": 1.2953, + "step": 2054 + }, + { + "epoch": 0.3659188034188034, + "grad_norm": 0.4859183430671692, + "learning_rate": 0.00019597042205201265, + "loss": 1.16, + "step": 2055 + }, + { + "epoch": 0.3660968660968661, + "grad_norm": 0.47056329250335693, + "learning_rate": 0.0001959664876525877, + "loss": 0.9982, + "step": 2056 + }, + { + "epoch": 0.3662749287749288, + "grad_norm": 0.48347967863082886, + "learning_rate": 0.00019596255137289875, + "loss": 1.0966, + "step": 2057 + }, + { + "epoch": 0.36645299145299143, + "grad_norm": 0.5068454742431641, + "learning_rate": 0.00019595861321302296, + "loss": 1.2891, + "step": 2058 + }, + { + "epoch": 0.3666310541310541, + "grad_norm": 0.5702359080314636, + "learning_rate": 0.00019595467317303747, + "loss": 1.1394, + "step": 2059 + }, + { + "epoch": 0.3668091168091168, + "grad_norm": 0.5028812885284424, + "learning_rate": 0.0001959507312530195, + "loss": 1.2324, + "step": 2060 + }, + { + "epoch": 0.36698717948717946, + "grad_norm": 0.4672880172729492, + "learning_rate": 0.00019594678745304628, + "loss": 1.0581, + "step": 2061 + }, + { + "epoch": 0.36716524216524216, + "grad_norm": 0.5233900547027588, + "learning_rate": 0.00019594284177319504, + "loss": 1.138, + "step": 2062 + }, + { + "epoch": 0.36734330484330485, + "grad_norm": 0.46871712803840637, + "learning_rate": 0.00019593889421354316, + "loss": 1.2159, + "step": 2063 + }, + { + "epoch": 0.36752136752136755, + "grad_norm": 0.5180533528327942, + "learning_rate": 0.00019593494477416793, + "loss": 1.1116, + "step": 2064 + }, + { + "epoch": 0.3676994301994302, + "grad_norm": 0.5398494005203247, + "learning_rate": 0.0001959309934551467, + "loss": 1.2038, + "step": 2065 + }, + { + "epoch": 0.3678774928774929, + "grad_norm": 0.4850373864173889, + "learning_rate": 0.000195927040256557, + "loss": 1.4315, + "step": 2066 + }, + { + "epoch": 0.3680555555555556, + "grad_norm": 0.49190905690193176, + "learning_rate": 0.0001959230851784762, + "loss": 0.9993, + "step": 2067 + }, + { + "epoch": 0.3682336182336182, + "grad_norm": 0.4546903073787689, + "learning_rate": 0.00019591912822098178, + "loss": 1.0979, + "step": 2068 + }, + { + "epoch": 0.3684116809116809, + "grad_norm": 0.4726468622684479, + "learning_rate": 0.00019591516938415133, + "loss": 1.1629, + "step": 2069 + }, + { + "epoch": 0.3685897435897436, + "grad_norm": 0.47856009006500244, + "learning_rate": 0.00019591120866806235, + "loss": 1.2048, + "step": 2070 + }, + { + "epoch": 0.36876780626780625, + "grad_norm": 0.46847718954086304, + "learning_rate": 0.0001959072460727925, + "loss": 1.0958, + "step": 2071 + }, + { + "epoch": 0.36894586894586895, + "grad_norm": 0.47164350748062134, + "learning_rate": 0.0001959032815984194, + "loss": 1.1912, + "step": 2072 + }, + { + "epoch": 0.36912393162393164, + "grad_norm": 0.4838213324546814, + "learning_rate": 0.0001958993152450207, + "loss": 1.1466, + "step": 2073 + }, + { + "epoch": 0.3693019943019943, + "grad_norm": 0.47234636545181274, + "learning_rate": 0.00019589534701267412, + "loss": 0.9475, + "step": 2074 + }, + { + "epoch": 0.369480056980057, + "grad_norm": 0.4913126826286316, + "learning_rate": 0.00019589137690145746, + "loss": 1.1571, + "step": 2075 + }, + { + "epoch": 0.3696581196581197, + "grad_norm": 0.4696233570575714, + "learning_rate": 0.00019588740491144842, + "loss": 0.9797, + "step": 2076 + }, + { + "epoch": 0.3698361823361823, + "grad_norm": 0.46146106719970703, + "learning_rate": 0.00019588343104272492, + "loss": 1.027, + "step": 2077 + }, + { + "epoch": 0.370014245014245, + "grad_norm": 0.4920627176761627, + "learning_rate": 0.00019587945529536474, + "loss": 1.1008, + "step": 2078 + }, + { + "epoch": 0.3701923076923077, + "grad_norm": 0.4854249954223633, + "learning_rate": 0.0001958754776694458, + "loss": 1.0759, + "step": 2079 + }, + { + "epoch": 0.37037037037037035, + "grad_norm": 0.4884897768497467, + "learning_rate": 0.00019587149816504608, + "loss": 1.1403, + "step": 2080 + }, + { + "epoch": 0.37054843304843305, + "grad_norm": 0.5062584280967712, + "learning_rate": 0.00019586751678224345, + "loss": 1.0185, + "step": 2081 + }, + { + "epoch": 0.37072649572649574, + "grad_norm": 0.44697675108909607, + "learning_rate": 0.000195863533521116, + "loss": 1.0462, + "step": 2082 + }, + { + "epoch": 0.3709045584045584, + "grad_norm": 0.5122885704040527, + "learning_rate": 0.00019585954838174176, + "loss": 1.108, + "step": 2083 + }, + { + "epoch": 0.3710826210826211, + "grad_norm": 0.486650288105011, + "learning_rate": 0.0001958555613641988, + "loss": 1.126, + "step": 2084 + }, + { + "epoch": 0.3712606837606838, + "grad_norm": 0.5296297669410706, + "learning_rate": 0.00019585157246856523, + "loss": 1.1757, + "step": 2085 + }, + { + "epoch": 0.3714387464387464, + "grad_norm": 0.4935721457004547, + "learning_rate": 0.0001958475816949192, + "loss": 1.1654, + "step": 2086 + }, + { + "epoch": 0.3716168091168091, + "grad_norm": 0.6226509213447571, + "learning_rate": 0.00019584358904333891, + "loss": 1.1981, + "step": 2087 + }, + { + "epoch": 0.3717948717948718, + "grad_norm": 0.44094228744506836, + "learning_rate": 0.0001958395945139026, + "loss": 0.8468, + "step": 2088 + }, + { + "epoch": 0.37197293447293445, + "grad_norm": 0.5335884690284729, + "learning_rate": 0.00019583559810668858, + "loss": 1.1597, + "step": 2089 + }, + { + "epoch": 0.37215099715099714, + "grad_norm": 0.4585414528846741, + "learning_rate": 0.000195831599821775, + "loss": 0.9343, + "step": 2090 + }, + { + "epoch": 0.37232905982905984, + "grad_norm": 0.533087432384491, + "learning_rate": 0.00019582759965924035, + "loss": 1.1209, + "step": 2091 + }, + { + "epoch": 0.37250712250712253, + "grad_norm": 0.5302683711051941, + "learning_rate": 0.00019582359761916295, + "loss": 1.236, + "step": 2092 + }, + { + "epoch": 0.3726851851851852, + "grad_norm": 0.4522508382797241, + "learning_rate": 0.00019581959370162122, + "loss": 1.0196, + "step": 2093 + }, + { + "epoch": 0.37286324786324787, + "grad_norm": 0.52391517162323, + "learning_rate": 0.00019581558790669358, + "loss": 1.0077, + "step": 2094 + }, + { + "epoch": 0.37304131054131057, + "grad_norm": 0.47144797444343567, + "learning_rate": 0.00019581158023445854, + "loss": 1.0956, + "step": 2095 + }, + { + "epoch": 0.3732193732193732, + "grad_norm": 0.4486723244190216, + "learning_rate": 0.00019580757068499459, + "loss": 0.8697, + "step": 2096 + }, + { + "epoch": 0.3733974358974359, + "grad_norm": 0.4626580476760864, + "learning_rate": 0.00019580355925838034, + "loss": 0.8489, + "step": 2097 + }, + { + "epoch": 0.3735754985754986, + "grad_norm": 0.5647920370101929, + "learning_rate": 0.00019579954595469438, + "loss": 1.1458, + "step": 2098 + }, + { + "epoch": 0.37375356125356124, + "grad_norm": 0.4734349846839905, + "learning_rate": 0.00019579553077401528, + "loss": 1.1036, + "step": 2099 + }, + { + "epoch": 0.37393162393162394, + "grad_norm": 0.5624295473098755, + "learning_rate": 0.00019579151371642176, + "loss": 0.9793, + "step": 2100 + }, + { + "epoch": 0.37410968660968663, + "grad_norm": 0.47507283091545105, + "learning_rate": 0.00019578749478199256, + "loss": 1.0371, + "step": 2101 + }, + { + "epoch": 0.37428774928774927, + "grad_norm": 0.550865113735199, + "learning_rate": 0.00019578347397080633, + "loss": 1.046, + "step": 2102 + }, + { + "epoch": 0.37446581196581197, + "grad_norm": 0.5249403715133667, + "learning_rate": 0.00019577945128294193, + "loss": 1.3185, + "step": 2103 + }, + { + "epoch": 0.37464387464387466, + "grad_norm": 0.4921024739742279, + "learning_rate": 0.00019577542671847815, + "loss": 1.0758, + "step": 2104 + }, + { + "epoch": 0.3748219373219373, + "grad_norm": 0.5351784825325012, + "learning_rate": 0.00019577140027749384, + "loss": 1.067, + "step": 2105 + }, + { + "epoch": 0.375, + "grad_norm": 0.44420507550239563, + "learning_rate": 0.00019576737196006787, + "loss": 1.1065, + "step": 2106 + }, + { + "epoch": 0.3751780626780627, + "grad_norm": 0.531384289264679, + "learning_rate": 0.0001957633417662792, + "loss": 1.1634, + "step": 2107 + }, + { + "epoch": 0.37535612535612534, + "grad_norm": 0.5167618989944458, + "learning_rate": 0.00019575930969620677, + "loss": 1.1646, + "step": 2108 + }, + { + "epoch": 0.37553418803418803, + "grad_norm": 0.41487228870391846, + "learning_rate": 0.0001957552757499296, + "loss": 0.793, + "step": 2109 + }, + { + "epoch": 0.37571225071225073, + "grad_norm": 0.5110787153244019, + "learning_rate": 0.00019575123992752672, + "loss": 1.1752, + "step": 2110 + }, + { + "epoch": 0.37589031339031337, + "grad_norm": 0.4422051012516022, + "learning_rate": 0.00019574720222907717, + "loss": 1.0102, + "step": 2111 + }, + { + "epoch": 0.37606837606837606, + "grad_norm": 0.4757538139820099, + "learning_rate": 0.0001957431626546601, + "loss": 1.0467, + "step": 2112 + }, + { + "epoch": 0.37624643874643876, + "grad_norm": 0.4736764430999756, + "learning_rate": 0.00019573912120435466, + "loss": 1.3048, + "step": 2113 + }, + { + "epoch": 0.3764245014245014, + "grad_norm": 0.49894335865974426, + "learning_rate": 0.00019573507787824004, + "loss": 1.0502, + "step": 2114 + }, + { + "epoch": 0.3766025641025641, + "grad_norm": 0.48120981454849243, + "learning_rate": 0.00019573103267639543, + "loss": 1.2405, + "step": 2115 + }, + { + "epoch": 0.3767806267806268, + "grad_norm": 0.4826737642288208, + "learning_rate": 0.0001957269855989001, + "loss": 1.1189, + "step": 2116 + }, + { + "epoch": 0.37695868945868943, + "grad_norm": 0.4736921489238739, + "learning_rate": 0.0001957229366458333, + "loss": 1.2862, + "step": 2117 + }, + { + "epoch": 0.37713675213675213, + "grad_norm": 0.3895208537578583, + "learning_rate": 0.00019571888581727446, + "loss": 1.0573, + "step": 2118 + }, + { + "epoch": 0.3773148148148148, + "grad_norm": 0.5107510089874268, + "learning_rate": 0.00019571483311330284, + "loss": 1.2913, + "step": 2119 + }, + { + "epoch": 0.37749287749287747, + "grad_norm": 0.4543241262435913, + "learning_rate": 0.00019571077853399794, + "loss": 0.949, + "step": 2120 + }, + { + "epoch": 0.37767094017094016, + "grad_norm": 0.46897491812705994, + "learning_rate": 0.00019570672207943913, + "loss": 1.2235, + "step": 2121 + }, + { + "epoch": 0.37784900284900286, + "grad_norm": 0.4812130630016327, + "learning_rate": 0.0001957026637497059, + "loss": 0.8857, + "step": 2122 + }, + { + "epoch": 0.37802706552706555, + "grad_norm": 0.47452476620674133, + "learning_rate": 0.00019569860354487782, + "loss": 1.0549, + "step": 2123 + }, + { + "epoch": 0.3782051282051282, + "grad_norm": 0.49879950284957886, + "learning_rate": 0.00019569454146503438, + "loss": 1.0475, + "step": 2124 + }, + { + "epoch": 0.3783831908831909, + "grad_norm": 0.4246445894241333, + "learning_rate": 0.00019569047751025518, + "loss": 0.8788, + "step": 2125 + }, + { + "epoch": 0.3785612535612536, + "grad_norm": 0.4868565499782562, + "learning_rate": 0.00019568641168061986, + "loss": 1.1801, + "step": 2126 + }, + { + "epoch": 0.3787393162393162, + "grad_norm": 0.46723654866218567, + "learning_rate": 0.0001956823439762081, + "loss": 1.1661, + "step": 2127 + }, + { + "epoch": 0.3789173789173789, + "grad_norm": 0.4989059269428253, + "learning_rate": 0.00019567827439709954, + "loss": 1.3037, + "step": 2128 + }, + { + "epoch": 0.3790954415954416, + "grad_norm": 0.441307932138443, + "learning_rate": 0.00019567420294337395, + "loss": 1.0197, + "step": 2129 + }, + { + "epoch": 0.37927350427350426, + "grad_norm": 0.5200160145759583, + "learning_rate": 0.0001956701296151111, + "loss": 1.3366, + "step": 2130 + }, + { + "epoch": 0.37945156695156695, + "grad_norm": 0.43610256910324097, + "learning_rate": 0.00019566605441239082, + "loss": 1.0148, + "step": 2131 + }, + { + "epoch": 0.37962962962962965, + "grad_norm": 0.4160982370376587, + "learning_rate": 0.00019566197733529293, + "loss": 1.0758, + "step": 2132 + }, + { + "epoch": 0.3798076923076923, + "grad_norm": 0.5007950663566589, + "learning_rate": 0.00019565789838389726, + "loss": 1.1937, + "step": 2133 + }, + { + "epoch": 0.379985754985755, + "grad_norm": 0.4991525113582611, + "learning_rate": 0.00019565381755828385, + "loss": 1.1788, + "step": 2134 + }, + { + "epoch": 0.3801638176638177, + "grad_norm": 0.6313113570213318, + "learning_rate": 0.00019564973485853258, + "loss": 1.1241, + "step": 2135 + }, + { + "epoch": 0.3803418803418803, + "grad_norm": 0.49736538529396057, + "learning_rate": 0.0001956456502847234, + "loss": 1.0299, + "step": 2136 + }, + { + "epoch": 0.380519943019943, + "grad_norm": 0.4384380578994751, + "learning_rate": 0.00019564156383693643, + "loss": 1.132, + "step": 2137 + }, + { + "epoch": 0.3806980056980057, + "grad_norm": 0.4696183502674103, + "learning_rate": 0.00019563747551525168, + "loss": 1.1145, + "step": 2138 + }, + { + "epoch": 0.38087606837606836, + "grad_norm": 0.42039749026298523, + "learning_rate": 0.0001956333853197493, + "loss": 0.9549, + "step": 2139 + }, + { + "epoch": 0.38105413105413105, + "grad_norm": 0.5547221899032593, + "learning_rate": 0.00019562929325050936, + "loss": 1.0476, + "step": 2140 + }, + { + "epoch": 0.38123219373219375, + "grad_norm": 0.4803301692008972, + "learning_rate": 0.0001956251993076121, + "loss": 1.1285, + "step": 2141 + }, + { + "epoch": 0.3814102564102564, + "grad_norm": 0.609501838684082, + "learning_rate": 0.00019562110349113766, + "loss": 1.2375, + "step": 2142 + }, + { + "epoch": 0.3815883190883191, + "grad_norm": 0.5134759545326233, + "learning_rate": 0.00019561700580116639, + "loss": 1.0895, + "step": 2143 + }, + { + "epoch": 0.3817663817663818, + "grad_norm": 0.5086711049079895, + "learning_rate": 0.00019561290623777846, + "loss": 1.1139, + "step": 2144 + }, + { + "epoch": 0.3819444444444444, + "grad_norm": 0.5371596813201904, + "learning_rate": 0.00019560880480105428, + "loss": 0.9302, + "step": 2145 + }, + { + "epoch": 0.3821225071225071, + "grad_norm": 0.4966319799423218, + "learning_rate": 0.00019560470149107418, + "loss": 1.2485, + "step": 2146 + }, + { + "epoch": 0.3823005698005698, + "grad_norm": 0.5296950340270996, + "learning_rate": 0.00019560059630791855, + "loss": 1.4449, + "step": 2147 + }, + { + "epoch": 0.38247863247863245, + "grad_norm": 0.5564194321632385, + "learning_rate": 0.00019559648925166783, + "loss": 1.0817, + "step": 2148 + }, + { + "epoch": 0.38265669515669515, + "grad_norm": 0.5763841867446899, + "learning_rate": 0.0001955923803224025, + "loss": 1.1915, + "step": 2149 + }, + { + "epoch": 0.38283475783475784, + "grad_norm": 0.4782295823097229, + "learning_rate": 0.00019558826952020304, + "loss": 1.1317, + "step": 2150 + }, + { + "epoch": 0.38301282051282054, + "grad_norm": 0.4876856207847595, + "learning_rate": 0.00019558415684515002, + "loss": 1.2113, + "step": 2151 + }, + { + "epoch": 0.3831908831908832, + "grad_norm": 0.4894421398639679, + "learning_rate": 0.00019558004229732398, + "loss": 1.0761, + "step": 2152 + }, + { + "epoch": 0.3833689458689459, + "grad_norm": 0.47914227843284607, + "learning_rate": 0.0001955759258768056, + "loss": 1.0869, + "step": 2153 + }, + { + "epoch": 0.38354700854700857, + "grad_norm": 0.43933629989624023, + "learning_rate": 0.00019557180758367543, + "loss": 1.0581, + "step": 2154 + }, + { + "epoch": 0.3837250712250712, + "grad_norm": 0.4078103005886078, + "learning_rate": 0.00019556768741801428, + "loss": 1.065, + "step": 2155 + }, + { + "epoch": 0.3839031339031339, + "grad_norm": 0.5112793445587158, + "learning_rate": 0.00019556356537990278, + "loss": 1.2023, + "step": 2156 + }, + { + "epoch": 0.3840811965811966, + "grad_norm": 0.4699678122997284, + "learning_rate": 0.00019555944146942177, + "loss": 1.2459, + "step": 2157 + }, + { + "epoch": 0.38425925925925924, + "grad_norm": 0.4723528027534485, + "learning_rate": 0.00019555531568665198, + "loss": 1.2204, + "step": 2158 + }, + { + "epoch": 0.38443732193732194, + "grad_norm": 0.4648225009441376, + "learning_rate": 0.00019555118803167432, + "loss": 1.1355, + "step": 2159 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 0.49861815571784973, + "learning_rate": 0.00019554705850456961, + "loss": 1.1301, + "step": 2160 + }, + { + "epoch": 0.3847934472934473, + "grad_norm": 0.4076344966888428, + "learning_rate": 0.00019554292710541874, + "loss": 0.8997, + "step": 2161 + }, + { + "epoch": 0.38497150997151, + "grad_norm": 0.5510796308517456, + "learning_rate": 0.00019553879383430272, + "loss": 1.0594, + "step": 2162 + }, + { + "epoch": 0.38514957264957267, + "grad_norm": 0.55793696641922, + "learning_rate": 0.00019553465869130249, + "loss": 1.1284, + "step": 2163 + }, + { + "epoch": 0.3853276353276353, + "grad_norm": 0.5096491575241089, + "learning_rate": 0.00019553052167649906, + "loss": 1.0419, + "step": 2164 + }, + { + "epoch": 0.385505698005698, + "grad_norm": 0.49077361822128296, + "learning_rate": 0.0001955263827899735, + "loss": 1.1632, + "step": 2165 + }, + { + "epoch": 0.3856837606837607, + "grad_norm": 0.5546894073486328, + "learning_rate": 0.00019552224203180693, + "loss": 1.1487, + "step": 2166 + }, + { + "epoch": 0.38586182336182334, + "grad_norm": 0.4930037260055542, + "learning_rate": 0.00019551809940208047, + "loss": 1.2668, + "step": 2167 + }, + { + "epoch": 0.38603988603988604, + "grad_norm": 0.5600671172142029, + "learning_rate": 0.00019551395490087525, + "loss": 1.3988, + "step": 2168 + }, + { + "epoch": 0.38621794871794873, + "grad_norm": 0.45897629857063293, + "learning_rate": 0.0001955098085282725, + "loss": 0.7792, + "step": 2169 + }, + { + "epoch": 0.3863960113960114, + "grad_norm": 0.46138936281204224, + "learning_rate": 0.00019550566028435346, + "loss": 1.1749, + "step": 2170 + }, + { + "epoch": 0.38657407407407407, + "grad_norm": 0.5136167407035828, + "learning_rate": 0.0001955015101691994, + "loss": 1.0153, + "step": 2171 + }, + { + "epoch": 0.38675213675213677, + "grad_norm": 0.4886440336704254, + "learning_rate": 0.00019549735818289165, + "loss": 1.0006, + "step": 2172 + }, + { + "epoch": 0.3869301994301994, + "grad_norm": 0.4339776635169983, + "learning_rate": 0.00019549320432551154, + "loss": 1.0109, + "step": 2173 + }, + { + "epoch": 0.3871082621082621, + "grad_norm": 0.48729443550109863, + "learning_rate": 0.00019548904859714044, + "loss": 1.2016, + "step": 2174 + }, + { + "epoch": 0.3872863247863248, + "grad_norm": 0.5128757357597351, + "learning_rate": 0.0001954848909978598, + "loss": 1.085, + "step": 2175 + }, + { + "epoch": 0.38746438746438744, + "grad_norm": 0.49636292457580566, + "learning_rate": 0.0001954807315277511, + "loss": 1.0671, + "step": 2176 + }, + { + "epoch": 0.38764245014245013, + "grad_norm": 0.4946988821029663, + "learning_rate": 0.00019547657018689578, + "loss": 1.2091, + "step": 2177 + }, + { + "epoch": 0.38782051282051283, + "grad_norm": 0.49004554748535156, + "learning_rate": 0.00019547240697537544, + "loss": 1.0241, + "step": 2178 + }, + { + "epoch": 0.38799857549857547, + "grad_norm": 0.48750075697898865, + "learning_rate": 0.00019546824189327157, + "loss": 1.1082, + "step": 2179 + }, + { + "epoch": 0.38817663817663817, + "grad_norm": 0.47726166248321533, + "learning_rate": 0.00019546407494066585, + "loss": 1.1275, + "step": 2180 + }, + { + "epoch": 0.38835470085470086, + "grad_norm": 0.5253444910049438, + "learning_rate": 0.00019545990611763986, + "loss": 1.0164, + "step": 2181 + }, + { + "epoch": 0.38853276353276356, + "grad_norm": 0.4470371603965759, + "learning_rate": 0.00019545573542427533, + "loss": 1.0138, + "step": 2182 + }, + { + "epoch": 0.3887108262108262, + "grad_norm": 0.6645087599754333, + "learning_rate": 0.00019545156286065397, + "loss": 1.0884, + "step": 2183 + }, + { + "epoch": 0.3888888888888889, + "grad_norm": 0.498775839805603, + "learning_rate": 0.0001954473884268575, + "loss": 1.1035, + "step": 2184 + }, + { + "epoch": 0.3890669515669516, + "grad_norm": 0.5830566883087158, + "learning_rate": 0.00019544321212296772, + "loss": 1.1665, + "step": 2185 + }, + { + "epoch": 0.38924501424501423, + "grad_norm": 0.48162809014320374, + "learning_rate": 0.00019543903394906646, + "loss": 1.1035, + "step": 2186 + }, + { + "epoch": 0.3894230769230769, + "grad_norm": 0.46334075927734375, + "learning_rate": 0.0001954348539052356, + "loss": 0.9764, + "step": 2187 + }, + { + "epoch": 0.3896011396011396, + "grad_norm": 0.6343515515327454, + "learning_rate": 0.00019543067199155704, + "loss": 0.9474, + "step": 2188 + }, + { + "epoch": 0.38977920227920226, + "grad_norm": 0.4867806136608124, + "learning_rate": 0.0001954264882081127, + "loss": 1.1161, + "step": 2189 + }, + { + "epoch": 0.38995726495726496, + "grad_norm": 0.49305734038352966, + "learning_rate": 0.00019542230255498454, + "loss": 1.1825, + "step": 2190 + }, + { + "epoch": 0.39013532763532766, + "grad_norm": 0.518465518951416, + "learning_rate": 0.00019541811503225457, + "loss": 1.0695, + "step": 2191 + }, + { + "epoch": 0.3903133903133903, + "grad_norm": 0.4892457127571106, + "learning_rate": 0.00019541392564000488, + "loss": 1.3113, + "step": 2192 + }, + { + "epoch": 0.390491452991453, + "grad_norm": 0.5150920152664185, + "learning_rate": 0.00019540973437831753, + "loss": 1.0735, + "step": 2193 + }, + { + "epoch": 0.3906695156695157, + "grad_norm": 0.5414708256721497, + "learning_rate": 0.00019540554124727462, + "loss": 1.0773, + "step": 2194 + }, + { + "epoch": 0.39084757834757833, + "grad_norm": 0.49826398491859436, + "learning_rate": 0.0001954013462469583, + "loss": 1.0542, + "step": 2195 + }, + { + "epoch": 0.391025641025641, + "grad_norm": 0.5203596949577332, + "learning_rate": 0.0001953971493774508, + "loss": 1.178, + "step": 2196 + }, + { + "epoch": 0.3912037037037037, + "grad_norm": 0.45095738768577576, + "learning_rate": 0.00019539295063883432, + "loss": 1.1254, + "step": 2197 + }, + { + "epoch": 0.39138176638176636, + "grad_norm": 0.4938857853412628, + "learning_rate": 0.00019538875003119113, + "loss": 1.1061, + "step": 2198 + }, + { + "epoch": 0.39155982905982906, + "grad_norm": 0.5260919332504272, + "learning_rate": 0.00019538454755460354, + "loss": 1.3292, + "step": 2199 + }, + { + "epoch": 0.39173789173789175, + "grad_norm": 0.46527108550071716, + "learning_rate": 0.00019538034320915388, + "loss": 1.2074, + "step": 2200 + }, + { + "epoch": 0.3919159544159544, + "grad_norm": 0.5608304738998413, + "learning_rate": 0.00019537613699492453, + "loss": 1.0385, + "step": 2201 + }, + { + "epoch": 0.3920940170940171, + "grad_norm": 0.5056684613227844, + "learning_rate": 0.00019537192891199792, + "loss": 1.1513, + "step": 2202 + }, + { + "epoch": 0.3922720797720798, + "grad_norm": 0.3764426112174988, + "learning_rate": 0.00019536771896045644, + "loss": 0.8966, + "step": 2203 + }, + { + "epoch": 0.3924501424501424, + "grad_norm": 0.4983638823032379, + "learning_rate": 0.0001953635071403827, + "loss": 1.097, + "step": 2204 + }, + { + "epoch": 0.3926282051282051, + "grad_norm": 0.5733919739723206, + "learning_rate": 0.00019535929345185904, + "loss": 1.4992, + "step": 2205 + }, + { + "epoch": 0.3928062678062678, + "grad_norm": 0.632064163684845, + "learning_rate": 0.00019535507789496817, + "loss": 1.0611, + "step": 2206 + }, + { + "epoch": 0.39298433048433046, + "grad_norm": 0.409978449344635, + "learning_rate": 0.00019535086046979262, + "loss": 0.7172, + "step": 2207 + }, + { + "epoch": 0.39316239316239315, + "grad_norm": 0.40910813212394714, + "learning_rate": 0.00019534664117641502, + "loss": 0.8803, + "step": 2208 + }, + { + "epoch": 0.39334045584045585, + "grad_norm": 0.4696179926395416, + "learning_rate": 0.00019534242001491807, + "loss": 1.1551, + "step": 2209 + }, + { + "epoch": 0.39351851851851855, + "grad_norm": 0.538425862789154, + "learning_rate": 0.00019533819698538444, + "loss": 1.1296, + "step": 2210 + }, + { + "epoch": 0.3936965811965812, + "grad_norm": 0.5913630723953247, + "learning_rate": 0.00019533397208789692, + "loss": 0.9757, + "step": 2211 + }, + { + "epoch": 0.3938746438746439, + "grad_norm": 0.5649870038032532, + "learning_rate": 0.00019532974532253822, + "loss": 0.9976, + "step": 2212 + }, + { + "epoch": 0.3940527065527066, + "grad_norm": 0.5012063980102539, + "learning_rate": 0.00019532551668939121, + "loss": 0.9969, + "step": 2213 + }, + { + "epoch": 0.3942307692307692, + "grad_norm": 0.5098594427108765, + "learning_rate": 0.00019532128618853872, + "loss": 1.1229, + "step": 2214 + }, + { + "epoch": 0.3944088319088319, + "grad_norm": 0.4753342568874359, + "learning_rate": 0.0001953170538200636, + "loss": 1.0808, + "step": 2215 + }, + { + "epoch": 0.3945868945868946, + "grad_norm": 0.4770098626613617, + "learning_rate": 0.00019531281958404888, + "loss": 1.0656, + "step": 2216 + }, + { + "epoch": 0.39476495726495725, + "grad_norm": 0.6007979512214661, + "learning_rate": 0.00019530858348057746, + "loss": 1.0093, + "step": 2217 + }, + { + "epoch": 0.39494301994301995, + "grad_norm": 0.4501650929450989, + "learning_rate": 0.00019530434550973227, + "loss": 0.8557, + "step": 2218 + }, + { + "epoch": 0.39512108262108264, + "grad_norm": 0.5123980641365051, + "learning_rate": 0.00019530010567159645, + "loss": 0.9833, + "step": 2219 + }, + { + "epoch": 0.3952991452991453, + "grad_norm": 0.4623969495296478, + "learning_rate": 0.000195295863966253, + "loss": 0.913, + "step": 2220 + }, + { + "epoch": 0.395477207977208, + "grad_norm": 0.4341880679130554, + "learning_rate": 0.0001952916203937851, + "loss": 1.0234, + "step": 2221 + }, + { + "epoch": 0.3956552706552707, + "grad_norm": 0.5935006141662598, + "learning_rate": 0.00019528737495427581, + "loss": 1.061, + "step": 2222 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 0.44835174083709717, + "learning_rate": 0.00019528312764780837, + "loss": 1.1567, + "step": 2223 + }, + { + "epoch": 0.396011396011396, + "grad_norm": 0.5476976633071899, + "learning_rate": 0.00019527887847446595, + "loss": 1.2304, + "step": 2224 + }, + { + "epoch": 0.3961894586894587, + "grad_norm": 0.4487939774990082, + "learning_rate": 0.00019527462743433187, + "loss": 1.1813, + "step": 2225 + }, + { + "epoch": 0.39636752136752135, + "grad_norm": 0.4053241014480591, + "learning_rate": 0.00019527037452748936, + "loss": 0.7899, + "step": 2226 + }, + { + "epoch": 0.39654558404558404, + "grad_norm": 0.534570574760437, + "learning_rate": 0.00019526611975402176, + "loss": 1.0681, + "step": 2227 + }, + { + "epoch": 0.39672364672364674, + "grad_norm": 0.46096158027648926, + "learning_rate": 0.00019526186311401246, + "loss": 0.9234, + "step": 2228 + }, + { + "epoch": 0.3969017094017094, + "grad_norm": 0.47363516688346863, + "learning_rate": 0.00019525760460754483, + "loss": 1.0197, + "step": 2229 + }, + { + "epoch": 0.3970797720797721, + "grad_norm": 0.46317258477211, + "learning_rate": 0.00019525334423470234, + "loss": 1.2103, + "step": 2230 + }, + { + "epoch": 0.39725783475783477, + "grad_norm": 0.4924237132072449, + "learning_rate": 0.0001952490819955684, + "loss": 1.3299, + "step": 2231 + }, + { + "epoch": 0.3974358974358974, + "grad_norm": 0.5419978499412537, + "learning_rate": 0.0001952448178902266, + "loss": 1.2526, + "step": 2232 + }, + { + "epoch": 0.3976139601139601, + "grad_norm": 0.5003267526626587, + "learning_rate": 0.00019524055191876043, + "loss": 1.1073, + "step": 2233 + }, + { + "epoch": 0.3977920227920228, + "grad_norm": 0.621789276599884, + "learning_rate": 0.00019523628408125347, + "loss": 1.3409, + "step": 2234 + }, + { + "epoch": 0.39797008547008544, + "grad_norm": 0.44235602021217346, + "learning_rate": 0.0001952320143777894, + "loss": 0.9799, + "step": 2235 + }, + { + "epoch": 0.39814814814814814, + "grad_norm": 0.49954718351364136, + "learning_rate": 0.0001952277428084518, + "loss": 1.2227, + "step": 2236 + }, + { + "epoch": 0.39832621082621084, + "grad_norm": 0.5113739967346191, + "learning_rate": 0.00019522346937332443, + "loss": 1.1644, + "step": 2237 + }, + { + "epoch": 0.39850427350427353, + "grad_norm": 0.5026139616966248, + "learning_rate": 0.00019521919407249096, + "loss": 1.0823, + "step": 2238 + }, + { + "epoch": 0.39868233618233617, + "grad_norm": 0.4943205714225769, + "learning_rate": 0.0001952149169060352, + "loss": 1.0961, + "step": 2239 + }, + { + "epoch": 0.39886039886039887, + "grad_norm": 0.4680631458759308, + "learning_rate": 0.00019521063787404094, + "loss": 0.9787, + "step": 2240 + }, + { + "epoch": 0.39903846153846156, + "grad_norm": 0.5511566400527954, + "learning_rate": 0.00019520635697659202, + "loss": 1.2543, + "step": 2241 + }, + { + "epoch": 0.3992165242165242, + "grad_norm": 0.5494263172149658, + "learning_rate": 0.00019520207421377229, + "loss": 1.1978, + "step": 2242 + }, + { + "epoch": 0.3993945868945869, + "grad_norm": 0.4850340485572815, + "learning_rate": 0.00019519778958566568, + "loss": 0.8531, + "step": 2243 + }, + { + "epoch": 0.3995726495726496, + "grad_norm": 0.47168150544166565, + "learning_rate": 0.00019519350309235613, + "loss": 1.0746, + "step": 2244 + }, + { + "epoch": 0.39975071225071224, + "grad_norm": 0.571133553981781, + "learning_rate": 0.00019518921473392765, + "loss": 1.2984, + "step": 2245 + }, + { + "epoch": 0.39992877492877493, + "grad_norm": 0.4636089503765106, + "learning_rate": 0.00019518492451046427, + "loss": 1.019, + "step": 2246 + }, + { + "epoch": 0.40010683760683763, + "grad_norm": 0.4573518931865692, + "learning_rate": 0.00019518063242205, + "loss": 1.1042, + "step": 2247 + }, + { + "epoch": 0.40028490028490027, + "grad_norm": 0.49098989367485046, + "learning_rate": 0.00019517633846876894, + "loss": 1.1224, + "step": 2248 + }, + { + "epoch": 0.40046296296296297, + "grad_norm": 0.5475491881370544, + "learning_rate": 0.00019517204265070523, + "loss": 1.0984, + "step": 2249 + }, + { + "epoch": 0.40064102564102566, + "grad_norm": 0.45498281717300415, + "learning_rate": 0.00019516774496794307, + "loss": 0.8883, + "step": 2250 + }, + { + "epoch": 0.4008190883190883, + "grad_norm": 0.4908423125743866, + "learning_rate": 0.00019516344542056666, + "loss": 1.328, + "step": 2251 + }, + { + "epoch": 0.400997150997151, + "grad_norm": 0.5474920272827148, + "learning_rate": 0.0001951591440086602, + "loss": 1.3825, + "step": 2252 + }, + { + "epoch": 0.4011752136752137, + "grad_norm": 0.5165615081787109, + "learning_rate": 0.000195154840732308, + "loss": 1.33, + "step": 2253 + }, + { + "epoch": 0.40135327635327633, + "grad_norm": 0.5185585021972656, + "learning_rate": 0.00019515053559159435, + "loss": 1.1689, + "step": 2254 + }, + { + "epoch": 0.40153133903133903, + "grad_norm": 0.5468854904174805, + "learning_rate": 0.00019514622858660363, + "loss": 1.2708, + "step": 2255 + }, + { + "epoch": 0.4017094017094017, + "grad_norm": 0.47556906938552856, + "learning_rate": 0.0001951419197174202, + "loss": 1.0488, + "step": 2256 + }, + { + "epoch": 0.40188746438746437, + "grad_norm": 0.5521323084831238, + "learning_rate": 0.0001951376089841285, + "loss": 1.0868, + "step": 2257 + }, + { + "epoch": 0.40206552706552706, + "grad_norm": 0.6029638051986694, + "learning_rate": 0.00019513329638681296, + "loss": 1.1735, + "step": 2258 + }, + { + "epoch": 0.40224358974358976, + "grad_norm": 0.4897766411304474, + "learning_rate": 0.00019512898192555812, + "loss": 1.1687, + "step": 2259 + }, + { + "epoch": 0.4024216524216524, + "grad_norm": 0.45527184009552, + "learning_rate": 0.00019512466560044848, + "loss": 1.0352, + "step": 2260 + }, + { + "epoch": 0.4025997150997151, + "grad_norm": 0.5025625824928284, + "learning_rate": 0.00019512034741156863, + "loss": 1.2503, + "step": 2261 + }, + { + "epoch": 0.4027777777777778, + "grad_norm": 0.46415451169013977, + "learning_rate": 0.00019511602735900317, + "loss": 1.032, + "step": 2262 + }, + { + "epoch": 0.40295584045584043, + "grad_norm": 0.4812934398651123, + "learning_rate": 0.00019511170544283678, + "loss": 1.0523, + "step": 2263 + }, + { + "epoch": 0.4031339031339031, + "grad_norm": 0.49937039613723755, + "learning_rate": 0.00019510738166315404, + "loss": 1.2238, + "step": 2264 + }, + { + "epoch": 0.4033119658119658, + "grad_norm": 0.5428698062896729, + "learning_rate": 0.00019510305602003975, + "loss": 1.0361, + "step": 2265 + }, + { + "epoch": 0.40349002849002846, + "grad_norm": 0.44836854934692383, + "learning_rate": 0.0001950987285135786, + "loss": 1.169, + "step": 2266 + }, + { + "epoch": 0.40366809116809116, + "grad_norm": 0.5071489214897156, + "learning_rate": 0.00019509439914385549, + "loss": 1.1567, + "step": 2267 + }, + { + "epoch": 0.40384615384615385, + "grad_norm": 0.5204613208770752, + "learning_rate": 0.00019509006791095513, + "loss": 0.9949, + "step": 2268 + }, + { + "epoch": 0.40402421652421655, + "grad_norm": 0.4583234488964081, + "learning_rate": 0.00019508573481496238, + "loss": 0.9051, + "step": 2269 + }, + { + "epoch": 0.4042022792022792, + "grad_norm": 0.5436791181564331, + "learning_rate": 0.00019508139985596222, + "loss": 1.3239, + "step": 2270 + }, + { + "epoch": 0.4043803418803419, + "grad_norm": 0.48774269223213196, + "learning_rate": 0.00019507706303403954, + "loss": 1.2102, + "step": 2271 + }, + { + "epoch": 0.4045584045584046, + "grad_norm": 0.4742540717124939, + "learning_rate": 0.00019507272434927933, + "loss": 1.1137, + "step": 2272 + }, + { + "epoch": 0.4047364672364672, + "grad_norm": 0.531148374080658, + "learning_rate": 0.00019506838380176658, + "loss": 1.3162, + "step": 2273 + }, + { + "epoch": 0.4049145299145299, + "grad_norm": 0.5002314448356628, + "learning_rate": 0.0001950640413915863, + "loss": 1.0743, + "step": 2274 + }, + { + "epoch": 0.4050925925925926, + "grad_norm": 0.39826446771621704, + "learning_rate": 0.00019505969711882366, + "loss": 0.7698, + "step": 2275 + }, + { + "epoch": 0.40527065527065526, + "grad_norm": 0.5177471041679382, + "learning_rate": 0.00019505535098356371, + "loss": 1.1821, + "step": 2276 + }, + { + "epoch": 0.40544871794871795, + "grad_norm": 0.467241108417511, + "learning_rate": 0.00019505100298589158, + "loss": 0.8036, + "step": 2277 + }, + { + "epoch": 0.40562678062678065, + "grad_norm": 0.43711844086647034, + "learning_rate": 0.00019504665312589255, + "loss": 0.8667, + "step": 2278 + }, + { + "epoch": 0.4058048433048433, + "grad_norm": 0.4929116368293762, + "learning_rate": 0.00019504230140365177, + "loss": 1.1279, + "step": 2279 + }, + { + "epoch": 0.405982905982906, + "grad_norm": 0.5279183983802795, + "learning_rate": 0.00019503794781925452, + "loss": 1.1318, + "step": 2280 + }, + { + "epoch": 0.4061609686609687, + "grad_norm": 0.549217939376831, + "learning_rate": 0.00019503359237278608, + "loss": 1.2007, + "step": 2281 + }, + { + "epoch": 0.4063390313390313, + "grad_norm": 0.5485880374908447, + "learning_rate": 0.00019502923506433187, + "loss": 1.1079, + "step": 2282 + }, + { + "epoch": 0.406517094017094, + "grad_norm": 0.48379644751548767, + "learning_rate": 0.0001950248758939772, + "loss": 0.9978, + "step": 2283 + }, + { + "epoch": 0.4066951566951567, + "grad_norm": 0.5943657755851746, + "learning_rate": 0.00019502051486180744, + "loss": 1.0466, + "step": 2284 + }, + { + "epoch": 0.40687321937321935, + "grad_norm": 0.5721273422241211, + "learning_rate": 0.00019501615196790812, + "loss": 1.2674, + "step": 2285 + }, + { + "epoch": 0.40705128205128205, + "grad_norm": 0.47624221444129944, + "learning_rate": 0.00019501178721236464, + "loss": 1.089, + "step": 2286 + }, + { + "epoch": 0.40722934472934474, + "grad_norm": 0.5091297030448914, + "learning_rate": 0.0001950074205952626, + "loss": 1.2035, + "step": 2287 + }, + { + "epoch": 0.4074074074074074, + "grad_norm": 0.45206236839294434, + "learning_rate": 0.0001950030521166875, + "loss": 0.9188, + "step": 2288 + }, + { + "epoch": 0.4075854700854701, + "grad_norm": 0.5563844442367554, + "learning_rate": 0.00019499868177672497, + "loss": 1.3444, + "step": 2289 + }, + { + "epoch": 0.4077635327635328, + "grad_norm": 0.4971138536930084, + "learning_rate": 0.00019499430957546055, + "loss": 1.1615, + "step": 2290 + }, + { + "epoch": 0.4079415954415954, + "grad_norm": 0.49355944991111755, + "learning_rate": 0.00019498993551298, + "loss": 1.1528, + "step": 2291 + }, + { + "epoch": 0.4081196581196581, + "grad_norm": 0.534705638885498, + "learning_rate": 0.000194985559589369, + "loss": 1.197, + "step": 2292 + }, + { + "epoch": 0.4082977207977208, + "grad_norm": 0.5113020539283752, + "learning_rate": 0.0001949811818047133, + "loss": 1.109, + "step": 2293 + }, + { + "epoch": 0.40847578347578345, + "grad_norm": 0.4823366701602936, + "learning_rate": 0.00019497680215909858, + "loss": 1.168, + "step": 2294 + }, + { + "epoch": 0.40865384615384615, + "grad_norm": 0.500792920589447, + "learning_rate": 0.00019497242065261077, + "loss": 1.1567, + "step": 2295 + }, + { + "epoch": 0.40883190883190884, + "grad_norm": 0.5047918558120728, + "learning_rate": 0.00019496803728533566, + "loss": 1.0515, + "step": 2296 + }, + { + "epoch": 0.40900997150997154, + "grad_norm": 0.474624365568161, + "learning_rate": 0.00019496365205735913, + "loss": 1.1747, + "step": 2297 + }, + { + "epoch": 0.4091880341880342, + "grad_norm": 0.5522183179855347, + "learning_rate": 0.0001949592649687671, + "loss": 1.1506, + "step": 2298 + }, + { + "epoch": 0.4093660968660969, + "grad_norm": 0.4526083767414093, + "learning_rate": 0.00019495487601964553, + "loss": 0.9968, + "step": 2299 + }, + { + "epoch": 0.40954415954415957, + "grad_norm": 0.545845091342926, + "learning_rate": 0.00019495048521008044, + "loss": 1.146, + "step": 2300 + }, + { + "epoch": 0.4097222222222222, + "grad_norm": 0.5475544333457947, + "learning_rate": 0.00019494609254015784, + "loss": 1.0101, + "step": 2301 + }, + { + "epoch": 0.4099002849002849, + "grad_norm": 0.43419042229652405, + "learning_rate": 0.00019494169800996373, + "loss": 1.065, + "step": 2302 + }, + { + "epoch": 0.4100783475783476, + "grad_norm": 0.44998374581336975, + "learning_rate": 0.00019493730161958435, + "loss": 0.9948, + "step": 2303 + }, + { + "epoch": 0.41025641025641024, + "grad_norm": 0.5401661992073059, + "learning_rate": 0.0001949329033691057, + "loss": 1.0473, + "step": 2304 + }, + { + "epoch": 0.41043447293447294, + "grad_norm": 0.48064103722572327, + "learning_rate": 0.00019492850325861404, + "loss": 1.0486, + "step": 2305 + }, + { + "epoch": 0.41061253561253563, + "grad_norm": 0.5398300290107727, + "learning_rate": 0.00019492410128819557, + "loss": 1.0314, + "step": 2306 + }, + { + "epoch": 0.4107905982905983, + "grad_norm": 0.4771125912666321, + "learning_rate": 0.0001949196974579365, + "loss": 0.9855, + "step": 2307 + }, + { + "epoch": 0.41096866096866097, + "grad_norm": 0.5375809669494629, + "learning_rate": 0.00019491529176792315, + "loss": 1.0777, + "step": 2308 + }, + { + "epoch": 0.41114672364672367, + "grad_norm": 0.48424094915390015, + "learning_rate": 0.00019491088421824183, + "loss": 1.0751, + "step": 2309 + }, + { + "epoch": 0.4113247863247863, + "grad_norm": 0.5054880380630493, + "learning_rate": 0.00019490647480897887, + "loss": 1.2457, + "step": 2310 + }, + { + "epoch": 0.411502849002849, + "grad_norm": 0.47118356823921204, + "learning_rate": 0.0001949020635402207, + "loss": 1.0445, + "step": 2311 + }, + { + "epoch": 0.4116809116809117, + "grad_norm": 0.47171851992607117, + "learning_rate": 0.00019489765041205375, + "loss": 1.0062, + "step": 2312 + }, + { + "epoch": 0.41185897435897434, + "grad_norm": 0.5703238844871521, + "learning_rate": 0.00019489323542456447, + "loss": 1.5639, + "step": 2313 + }, + { + "epoch": 0.41203703703703703, + "grad_norm": 0.5045075416564941, + "learning_rate": 0.00019488881857783935, + "loss": 1.1665, + "step": 2314 + }, + { + "epoch": 0.41221509971509973, + "grad_norm": 0.46835362911224365, + "learning_rate": 0.00019488439987196495, + "loss": 1.2078, + "step": 2315 + }, + { + "epoch": 0.41239316239316237, + "grad_norm": 0.5187196731567383, + "learning_rate": 0.00019487997930702785, + "loss": 1.1049, + "step": 2316 + }, + { + "epoch": 0.41257122507122507, + "grad_norm": 0.5190554857254028, + "learning_rate": 0.00019487555688311463, + "loss": 1.331, + "step": 2317 + }, + { + "epoch": 0.41274928774928776, + "grad_norm": 0.7394969463348389, + "learning_rate": 0.00019487113260031197, + "loss": 0.9646, + "step": 2318 + }, + { + "epoch": 0.4129273504273504, + "grad_norm": 0.532982349395752, + "learning_rate": 0.00019486670645870656, + "loss": 1.166, + "step": 2319 + }, + { + "epoch": 0.4131054131054131, + "grad_norm": 0.48659515380859375, + "learning_rate": 0.00019486227845838509, + "loss": 1.0016, + "step": 2320 + }, + { + "epoch": 0.4132834757834758, + "grad_norm": 0.5364453196525574, + "learning_rate": 0.00019485784859943434, + "loss": 1.3877, + "step": 2321 + }, + { + "epoch": 0.41346153846153844, + "grad_norm": 0.49788740277290344, + "learning_rate": 0.0001948534168819411, + "loss": 1.2949, + "step": 2322 + }, + { + "epoch": 0.41363960113960113, + "grad_norm": 0.5125377774238586, + "learning_rate": 0.00019484898330599217, + "loss": 0.9769, + "step": 2323 + }, + { + "epoch": 0.41381766381766383, + "grad_norm": 0.5434861779212952, + "learning_rate": 0.00019484454787167447, + "loss": 1.254, + "step": 2324 + }, + { + "epoch": 0.41399572649572647, + "grad_norm": 0.5324583053588867, + "learning_rate": 0.00019484011057907487, + "loss": 0.9788, + "step": 2325 + }, + { + "epoch": 0.41417378917378916, + "grad_norm": 0.4806961715221405, + "learning_rate": 0.00019483567142828033, + "loss": 1.0089, + "step": 2326 + }, + { + "epoch": 0.41435185185185186, + "grad_norm": 0.5152947306632996, + "learning_rate": 0.0001948312304193778, + "loss": 1.15, + "step": 2327 + }, + { + "epoch": 0.41452991452991456, + "grad_norm": 0.6030138731002808, + "learning_rate": 0.0001948267875524543, + "loss": 1.196, + "step": 2328 + }, + { + "epoch": 0.4147079772079772, + "grad_norm": 0.4504946768283844, + "learning_rate": 0.0001948223428275969, + "loss": 0.8742, + "step": 2329 + }, + { + "epoch": 0.4148860398860399, + "grad_norm": 0.5195745825767517, + "learning_rate": 0.00019481789624489263, + "loss": 1.0104, + "step": 2330 + }, + { + "epoch": 0.4150641025641026, + "grad_norm": 0.5269250869750977, + "learning_rate": 0.0001948134478044287, + "loss": 1.2284, + "step": 2331 + }, + { + "epoch": 0.41524216524216523, + "grad_norm": 0.5302315354347229, + "learning_rate": 0.00019480899750629218, + "loss": 1.1374, + "step": 2332 + }, + { + "epoch": 0.4154202279202279, + "grad_norm": 0.5501471161842346, + "learning_rate": 0.0001948045453505703, + "loss": 1.214, + "step": 2333 + }, + { + "epoch": 0.4155982905982906, + "grad_norm": 0.4674588739871979, + "learning_rate": 0.0001948000913373503, + "loss": 1.0568, + "step": 2334 + }, + { + "epoch": 0.41577635327635326, + "grad_norm": 0.5262266993522644, + "learning_rate": 0.0001947956354667195, + "loss": 1.111, + "step": 2335 + }, + { + "epoch": 0.41595441595441596, + "grad_norm": 0.4549071788787842, + "learning_rate": 0.00019479117773876507, + "loss": 1.2655, + "step": 2336 + }, + { + "epoch": 0.41613247863247865, + "grad_norm": 0.48897311091423035, + "learning_rate": 0.00019478671815357447, + "loss": 1.0543, + "step": 2337 + }, + { + "epoch": 0.4163105413105413, + "grad_norm": 0.5544867515563965, + "learning_rate": 0.000194782256711235, + "loss": 1.2276, + "step": 2338 + }, + { + "epoch": 0.416488603988604, + "grad_norm": 0.5050773024559021, + "learning_rate": 0.0001947777934118341, + "loss": 0.9781, + "step": 2339 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 0.4831899106502533, + "learning_rate": 0.00019477332825545925, + "loss": 1.0213, + "step": 2340 + }, + { + "epoch": 0.4168447293447293, + "grad_norm": 0.5392552614212036, + "learning_rate": 0.0001947688612421979, + "loss": 1.3251, + "step": 2341 + }, + { + "epoch": 0.417022792022792, + "grad_norm": 0.5003608465194702, + "learning_rate": 0.00019476439237213754, + "loss": 1.0714, + "step": 2342 + }, + { + "epoch": 0.4172008547008547, + "grad_norm": 0.5016986727714539, + "learning_rate": 0.00019475992164536582, + "loss": 1.0656, + "step": 2343 + }, + { + "epoch": 0.41737891737891736, + "grad_norm": 0.5139234066009521, + "learning_rate": 0.00019475544906197024, + "loss": 1.1317, + "step": 2344 + }, + { + "epoch": 0.41755698005698005, + "grad_norm": 0.582478940486908, + "learning_rate": 0.00019475097462203847, + "loss": 1.4209, + "step": 2345 + }, + { + "epoch": 0.41773504273504275, + "grad_norm": 0.5248767137527466, + "learning_rate": 0.00019474649832565823, + "loss": 1.2965, + "step": 2346 + }, + { + "epoch": 0.4179131054131054, + "grad_norm": 0.4977390170097351, + "learning_rate": 0.00019474202017291713, + "loss": 1.3319, + "step": 2347 + }, + { + "epoch": 0.4180911680911681, + "grad_norm": 0.4868984818458557, + "learning_rate": 0.00019473754016390298, + "loss": 1.0595, + "step": 2348 + }, + { + "epoch": 0.4182692307692308, + "grad_norm": 0.5965346693992615, + "learning_rate": 0.00019473305829870353, + "loss": 1.2289, + "step": 2349 + }, + { + "epoch": 0.4184472934472934, + "grad_norm": 0.46590209007263184, + "learning_rate": 0.0001947285745774066, + "loss": 1.0468, + "step": 2350 + }, + { + "epoch": 0.4186253561253561, + "grad_norm": 0.497811883687973, + "learning_rate": 0.0001947240890001, + "loss": 1.1247, + "step": 2351 + }, + { + "epoch": 0.4188034188034188, + "grad_norm": 0.5348289012908936, + "learning_rate": 0.0001947196015668717, + "loss": 0.9496, + "step": 2352 + }, + { + "epoch": 0.41898148148148145, + "grad_norm": 0.5086174607276917, + "learning_rate": 0.0001947151122778095, + "loss": 0.8869, + "step": 2353 + }, + { + "epoch": 0.41915954415954415, + "grad_norm": 0.4844677150249481, + "learning_rate": 0.00019471062113300146, + "loss": 0.847, + "step": 2354 + }, + { + "epoch": 0.41933760683760685, + "grad_norm": 0.5395866632461548, + "learning_rate": 0.00019470612813253556, + "loss": 0.9684, + "step": 2355 + }, + { + "epoch": 0.41951566951566954, + "grad_norm": 0.479403018951416, + "learning_rate": 0.0001947016332764998, + "loss": 1.0532, + "step": 2356 + }, + { + "epoch": 0.4196937321937322, + "grad_norm": 0.5499961376190186, + "learning_rate": 0.00019469713656498227, + "loss": 1.2565, + "step": 2357 + }, + { + "epoch": 0.4198717948717949, + "grad_norm": 0.5865352153778076, + "learning_rate": 0.00019469263799807104, + "loss": 1.1349, + "step": 2358 + }, + { + "epoch": 0.4200498575498576, + "grad_norm": 0.4454309046268463, + "learning_rate": 0.00019468813757585432, + "loss": 0.9631, + "step": 2359 + }, + { + "epoch": 0.4202279202279202, + "grad_norm": 0.48426875472068787, + "learning_rate": 0.00019468363529842023, + "loss": 0.9795, + "step": 2360 + }, + { + "epoch": 0.4204059829059829, + "grad_norm": 0.47428226470947266, + "learning_rate": 0.00019467913116585697, + "loss": 0.9316, + "step": 2361 + }, + { + "epoch": 0.4205840455840456, + "grad_norm": 0.5193758010864258, + "learning_rate": 0.00019467462517825282, + "loss": 1.235, + "step": 2362 + }, + { + "epoch": 0.42076210826210825, + "grad_norm": 0.49845513701438904, + "learning_rate": 0.00019467011733569607, + "loss": 1.2413, + "step": 2363 + }, + { + "epoch": 0.42094017094017094, + "grad_norm": 0.45483845472335815, + "learning_rate": 0.00019466560763827502, + "loss": 1.2817, + "step": 2364 + }, + { + "epoch": 0.42111823361823364, + "grad_norm": 0.43345287442207336, + "learning_rate": 0.00019466109608607806, + "loss": 0.8568, + "step": 2365 + }, + { + "epoch": 0.4212962962962963, + "grad_norm": 0.4467088282108307, + "learning_rate": 0.00019465658267919352, + "loss": 1.1408, + "step": 2366 + }, + { + "epoch": 0.421474358974359, + "grad_norm": 0.6705610156059265, + "learning_rate": 0.00019465206741770992, + "loss": 1.445, + "step": 2367 + }, + { + "epoch": 0.42165242165242167, + "grad_norm": 0.5037859678268433, + "learning_rate": 0.00019464755030171565, + "loss": 0.8682, + "step": 2368 + }, + { + "epoch": 0.4218304843304843, + "grad_norm": 0.49576324224472046, + "learning_rate": 0.00019464303133129928, + "loss": 0.8387, + "step": 2369 + }, + { + "epoch": 0.422008547008547, + "grad_norm": 0.5222806334495544, + "learning_rate": 0.00019463851050654927, + "loss": 1.1443, + "step": 2370 + }, + { + "epoch": 0.4221866096866097, + "grad_norm": 0.4966863989830017, + "learning_rate": 0.00019463398782755426, + "loss": 1.1555, + "step": 2371 + }, + { + "epoch": 0.42236467236467234, + "grad_norm": 0.6140168309211731, + "learning_rate": 0.00019462946329440285, + "loss": 1.2264, + "step": 2372 + }, + { + "epoch": 0.42254273504273504, + "grad_norm": 0.4906651973724365, + "learning_rate": 0.0001946249369071837, + "loss": 1.2459, + "step": 2373 + }, + { + "epoch": 0.42272079772079774, + "grad_norm": 0.5956700444221497, + "learning_rate": 0.00019462040866598544, + "loss": 1.1521, + "step": 2374 + }, + { + "epoch": 0.4228988603988604, + "grad_norm": 0.46044886112213135, + "learning_rate": 0.00019461587857089687, + "loss": 1.2084, + "step": 2375 + }, + { + "epoch": 0.4230769230769231, + "grad_norm": 0.5109430551528931, + "learning_rate": 0.00019461134662200668, + "loss": 1.2684, + "step": 2376 + }, + { + "epoch": 0.42325498575498577, + "grad_norm": 0.4373733103275299, + "learning_rate": 0.0001946068128194037, + "loss": 1.0451, + "step": 2377 + }, + { + "epoch": 0.4234330484330484, + "grad_norm": 0.553817868232727, + "learning_rate": 0.00019460227716317673, + "loss": 1.1052, + "step": 2378 + }, + { + "epoch": 0.4236111111111111, + "grad_norm": 0.5742647647857666, + "learning_rate": 0.00019459773965341468, + "loss": 1.1647, + "step": 2379 + }, + { + "epoch": 0.4237891737891738, + "grad_norm": 0.5461940169334412, + "learning_rate": 0.00019459320029020642, + "loss": 1.0953, + "step": 2380 + }, + { + "epoch": 0.42396723646723644, + "grad_norm": 0.5837802290916443, + "learning_rate": 0.0001945886590736409, + "loss": 1.1303, + "step": 2381 + }, + { + "epoch": 0.42414529914529914, + "grad_norm": 0.5316985249519348, + "learning_rate": 0.0001945841160038071, + "loss": 1.1204, + "step": 2382 + }, + { + "epoch": 0.42432336182336183, + "grad_norm": 0.5846191048622131, + "learning_rate": 0.00019457957108079404, + "loss": 1.2622, + "step": 2383 + }, + { + "epoch": 0.42450142450142453, + "grad_norm": 0.43266957998275757, + "learning_rate": 0.00019457502430469075, + "loss": 0.9834, + "step": 2384 + }, + { + "epoch": 0.42467948717948717, + "grad_norm": 0.514081597328186, + "learning_rate": 0.00019457047567558632, + "loss": 0.8413, + "step": 2385 + }, + { + "epoch": 0.42485754985754987, + "grad_norm": 0.4831700325012207, + "learning_rate": 0.00019456592519356987, + "loss": 0.9244, + "step": 2386 + }, + { + "epoch": 0.42503561253561256, + "grad_norm": 0.5612850785255432, + "learning_rate": 0.00019456137285873057, + "loss": 0.9438, + "step": 2387 + }, + { + "epoch": 0.4252136752136752, + "grad_norm": 0.5197352766990662, + "learning_rate": 0.00019455681867115758, + "loss": 1.1095, + "step": 2388 + }, + { + "epoch": 0.4253917378917379, + "grad_norm": 0.5045261979103088, + "learning_rate": 0.00019455226263094018, + "loss": 1.0007, + "step": 2389 + }, + { + "epoch": 0.4255698005698006, + "grad_norm": 0.5167570114135742, + "learning_rate": 0.00019454770473816758, + "loss": 1.1335, + "step": 2390 + }, + { + "epoch": 0.42574786324786323, + "grad_norm": 0.49262070655822754, + "learning_rate": 0.00019454314499292913, + "loss": 1.0436, + "step": 2391 + }, + { + "epoch": 0.42592592592592593, + "grad_norm": 0.4489207863807678, + "learning_rate": 0.00019453858339531417, + "loss": 1.0138, + "step": 2392 + }, + { + "epoch": 0.4261039886039886, + "grad_norm": 0.6024920344352722, + "learning_rate": 0.00019453401994541203, + "loss": 1.1921, + "step": 2393 + }, + { + "epoch": 0.42628205128205127, + "grad_norm": 0.46807861328125, + "learning_rate": 0.00019452945464331215, + "loss": 1.0947, + "step": 2394 + }, + { + "epoch": 0.42646011396011396, + "grad_norm": 0.48776543140411377, + "learning_rate": 0.00019452488748910397, + "loss": 1.0029, + "step": 2395 + }, + { + "epoch": 0.42663817663817666, + "grad_norm": 0.4798663556575775, + "learning_rate": 0.000194520318482877, + "loss": 0.7863, + "step": 2396 + }, + { + "epoch": 0.4268162393162393, + "grad_norm": 0.5067816972732544, + "learning_rate": 0.0001945157476247207, + "loss": 1.0049, + "step": 2397 + }, + { + "epoch": 0.426994301994302, + "grad_norm": 0.5179638266563416, + "learning_rate": 0.00019451117491472468, + "loss": 1.1851, + "step": 2398 + }, + { + "epoch": 0.4271723646723647, + "grad_norm": 0.4782430827617645, + "learning_rate": 0.00019450660035297854, + "loss": 1.125, + "step": 2399 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 0.560077965259552, + "learning_rate": 0.00019450202393957186, + "loss": 1.1843, + "step": 2400 + }, + { + "epoch": 0.42752849002849, + "grad_norm": 0.5247970223426819, + "learning_rate": 0.00019449744567459436, + "loss": 1.1576, + "step": 2401 + }, + { + "epoch": 0.4277065527065527, + "grad_norm": 0.6414062976837158, + "learning_rate": 0.00019449286555813568, + "loss": 1.1833, + "step": 2402 + }, + { + "epoch": 0.42788461538461536, + "grad_norm": 0.5006586909294128, + "learning_rate": 0.00019448828359028563, + "loss": 1.1778, + "step": 2403 + }, + { + "epoch": 0.42806267806267806, + "grad_norm": 0.4946450889110565, + "learning_rate": 0.0001944836997711339, + "loss": 1.1611, + "step": 2404 + }, + { + "epoch": 0.42824074074074076, + "grad_norm": 0.4601200222969055, + "learning_rate": 0.00019447911410077037, + "loss": 1.2456, + "step": 2405 + }, + { + "epoch": 0.4284188034188034, + "grad_norm": 0.4653947651386261, + "learning_rate": 0.00019447452657928485, + "loss": 1.0941, + "step": 2406 + }, + { + "epoch": 0.4285968660968661, + "grad_norm": 0.5015713572502136, + "learning_rate": 0.00019446993720676726, + "loss": 1.3113, + "step": 2407 + }, + { + "epoch": 0.4287749287749288, + "grad_norm": 0.5803143978118896, + "learning_rate": 0.0001944653459833075, + "loss": 1.0568, + "step": 2408 + }, + { + "epoch": 0.42895299145299143, + "grad_norm": 0.5259647965431213, + "learning_rate": 0.0001944607529089955, + "loss": 1.1243, + "step": 2409 + }, + { + "epoch": 0.4291310541310541, + "grad_norm": 0.5150414109230042, + "learning_rate": 0.00019445615798392124, + "loss": 1.0676, + "step": 2410 + }, + { + "epoch": 0.4293091168091168, + "grad_norm": 0.5848649740219116, + "learning_rate": 0.0001944515612081748, + "loss": 1.0671, + "step": 2411 + }, + { + "epoch": 0.42948717948717946, + "grad_norm": 0.5696990489959717, + "learning_rate": 0.00019444696258184626, + "loss": 1.3323, + "step": 2412 + }, + { + "epoch": 0.42966524216524216, + "grad_norm": 0.49822330474853516, + "learning_rate": 0.00019444236210502567, + "loss": 1.1004, + "step": 2413 + }, + { + "epoch": 0.42984330484330485, + "grad_norm": 0.4683490991592407, + "learning_rate": 0.00019443775977780317, + "loss": 0.9768, + "step": 2414 + }, + { + "epoch": 0.43002136752136755, + "grad_norm": 0.5703811049461365, + "learning_rate": 0.00019443315560026893, + "loss": 1.154, + "step": 2415 + }, + { + "epoch": 0.4301994301994302, + "grad_norm": 0.5121861100196838, + "learning_rate": 0.0001944285495725132, + "loss": 1.1388, + "step": 2416 + }, + { + "epoch": 0.4303774928774929, + "grad_norm": 0.4864094853401184, + "learning_rate": 0.00019442394169462619, + "loss": 0.9214, + "step": 2417 + }, + { + "epoch": 0.4305555555555556, + "grad_norm": 0.5234864354133606, + "learning_rate": 0.0001944193319666982, + "loss": 1.2787, + "step": 2418 + }, + { + "epoch": 0.4307336182336182, + "grad_norm": 0.5137650370597839, + "learning_rate": 0.00019441472038881955, + "loss": 1.1406, + "step": 2419 + }, + { + "epoch": 0.4309116809116809, + "grad_norm": 0.49687784910202026, + "learning_rate": 0.00019441010696108054, + "loss": 0.93, + "step": 2420 + }, + { + "epoch": 0.4310897435897436, + "grad_norm": 0.5078722834587097, + "learning_rate": 0.00019440549168357163, + "loss": 1.1417, + "step": 2421 + }, + { + "epoch": 0.43126780626780625, + "grad_norm": 0.4483391046524048, + "learning_rate": 0.00019440087455638324, + "loss": 0.9016, + "step": 2422 + }, + { + "epoch": 0.43144586894586895, + "grad_norm": 0.5963045954704285, + "learning_rate": 0.00019439625557960576, + "loss": 1.1567, + "step": 2423 + }, + { + "epoch": 0.43162393162393164, + "grad_norm": 0.5534471273422241, + "learning_rate": 0.0001943916347533298, + "loss": 1.1409, + "step": 2424 + }, + { + "epoch": 0.4318019943019943, + "grad_norm": 0.6400241851806641, + "learning_rate": 0.0001943870120776458, + "loss": 1.2041, + "step": 2425 + }, + { + "epoch": 0.431980056980057, + "grad_norm": 0.4599420726299286, + "learning_rate": 0.0001943823875526444, + "loss": 1.023, + "step": 2426 + }, + { + "epoch": 0.4321581196581197, + "grad_norm": 0.4799708425998688, + "learning_rate": 0.00019437776117841614, + "loss": 1.0872, + "step": 2427 + }, + { + "epoch": 0.4323361823361823, + "grad_norm": 0.5138532519340515, + "learning_rate": 0.00019437313295505172, + "loss": 1.1175, + "step": 2428 + }, + { + "epoch": 0.432514245014245, + "grad_norm": 0.538223147392273, + "learning_rate": 0.00019436850288264183, + "loss": 1.1203, + "step": 2429 + }, + { + "epoch": 0.4326923076923077, + "grad_norm": 0.458044171333313, + "learning_rate": 0.00019436387096127713, + "loss": 1.0383, + "step": 2430 + }, + { + "epoch": 0.43287037037037035, + "grad_norm": 0.5928303599357605, + "learning_rate": 0.00019435923719104842, + "loss": 1.1191, + "step": 2431 + }, + { + "epoch": 0.43304843304843305, + "grad_norm": 0.5818437933921814, + "learning_rate": 0.00019435460157204645, + "loss": 1.0352, + "step": 2432 + }, + { + "epoch": 0.43322649572649574, + "grad_norm": 0.487341046333313, + "learning_rate": 0.0001943499641043621, + "loss": 1.2608, + "step": 2433 + }, + { + "epoch": 0.4334045584045584, + "grad_norm": 0.4737292230129242, + "learning_rate": 0.0001943453247880862, + "loss": 1.0084, + "step": 2434 + }, + { + "epoch": 0.4335826210826211, + "grad_norm": 0.4251207709312439, + "learning_rate": 0.0001943406836233096, + "loss": 0.9163, + "step": 2435 + }, + { + "epoch": 0.4337606837606838, + "grad_norm": 0.49468478560447693, + "learning_rate": 0.00019433604061012331, + "loss": 1.0293, + "step": 2436 + }, + { + "epoch": 0.4339387464387464, + "grad_norm": 0.47120022773742676, + "learning_rate": 0.00019433139574861826, + "loss": 1.0097, + "step": 2437 + }, + { + "epoch": 0.4341168091168091, + "grad_norm": 0.5060358047485352, + "learning_rate": 0.00019432674903888548, + "loss": 1.0683, + "step": 2438 + }, + { + "epoch": 0.4342948717948718, + "grad_norm": 0.5455917119979858, + "learning_rate": 0.00019432210048101598, + "loss": 0.8886, + "step": 2439 + }, + { + "epoch": 0.43447293447293445, + "grad_norm": 0.7960546612739563, + "learning_rate": 0.00019431745007510086, + "loss": 0.8648, + "step": 2440 + }, + { + "epoch": 0.43465099715099714, + "grad_norm": 0.5069689154624939, + "learning_rate": 0.00019431279782123126, + "loss": 1.1315, + "step": 2441 + }, + { + "epoch": 0.43482905982905984, + "grad_norm": 0.5597776174545288, + "learning_rate": 0.0001943081437194983, + "loss": 1.2281, + "step": 2442 + }, + { + "epoch": 0.43500712250712253, + "grad_norm": 0.4527420997619629, + "learning_rate": 0.00019430348776999315, + "loss": 0.7576, + "step": 2443 + }, + { + "epoch": 0.4351851851851852, + "grad_norm": 0.5625936388969421, + "learning_rate": 0.00019429882997280706, + "loss": 1.0302, + "step": 2444 + }, + { + "epoch": 0.43536324786324787, + "grad_norm": 0.5173513293266296, + "learning_rate": 0.0001942941703280313, + "loss": 1.2255, + "step": 2445 + }, + { + "epoch": 0.43554131054131057, + "grad_norm": 0.45889151096343994, + "learning_rate": 0.00019428950883575714, + "loss": 0.9322, + "step": 2446 + }, + { + "epoch": 0.4357193732193732, + "grad_norm": 0.5288477540016174, + "learning_rate": 0.00019428484549607593, + "loss": 1.0572, + "step": 2447 + }, + { + "epoch": 0.4358974358974359, + "grad_norm": 0.48328033089637756, + "learning_rate": 0.00019428018030907902, + "loss": 1.1213, + "step": 2448 + }, + { + "epoch": 0.4360754985754986, + "grad_norm": 0.5146737098693848, + "learning_rate": 0.00019427551327485786, + "loss": 0.9633, + "step": 2449 + }, + { + "epoch": 0.43625356125356124, + "grad_norm": 0.5138360261917114, + "learning_rate": 0.00019427084439350382, + "loss": 1.0561, + "step": 2450 + }, + { + "epoch": 0.43643162393162394, + "grad_norm": 0.5192533135414124, + "learning_rate": 0.00019426617366510843, + "loss": 1.1704, + "step": 2451 + }, + { + "epoch": 0.43660968660968663, + "grad_norm": 0.4819495379924774, + "learning_rate": 0.00019426150108976318, + "loss": 1.0958, + "step": 2452 + }, + { + "epoch": 0.43678774928774927, + "grad_norm": 0.4626680910587311, + "learning_rate": 0.00019425682666755965, + "loss": 1.1872, + "step": 2453 + }, + { + "epoch": 0.43696581196581197, + "grad_norm": 0.5773931741714478, + "learning_rate": 0.00019425215039858937, + "loss": 1.0722, + "step": 2454 + }, + { + "epoch": 0.43714387464387466, + "grad_norm": 0.5003872513771057, + "learning_rate": 0.00019424747228294402, + "loss": 1.0561, + "step": 2455 + }, + { + "epoch": 0.4373219373219373, + "grad_norm": 0.47370314598083496, + "learning_rate": 0.0001942427923207152, + "loss": 1.1619, + "step": 2456 + }, + { + "epoch": 0.4375, + "grad_norm": 0.466421514749527, + "learning_rate": 0.00019423811051199466, + "loss": 1.1311, + "step": 2457 + }, + { + "epoch": 0.4376780626780627, + "grad_norm": 0.44564682245254517, + "learning_rate": 0.00019423342685687413, + "loss": 1.1889, + "step": 2458 + }, + { + "epoch": 0.43785612535612534, + "grad_norm": 0.40986698865890503, + "learning_rate": 0.00019422874135544533, + "loss": 0.7312, + "step": 2459 + }, + { + "epoch": 0.43803418803418803, + "grad_norm": 0.4714358448982239, + "learning_rate": 0.0001942240540078001, + "loss": 0.9273, + "step": 2460 + }, + { + "epoch": 0.43821225071225073, + "grad_norm": 0.5298398733139038, + "learning_rate": 0.00019421936481403025, + "loss": 1.3377, + "step": 2461 + }, + { + "epoch": 0.43839031339031337, + "grad_norm": 0.6326695680618286, + "learning_rate": 0.0001942146737742277, + "loss": 1.0258, + "step": 2462 + }, + { + "epoch": 0.43856837606837606, + "grad_norm": 0.5087653994560242, + "learning_rate": 0.00019420998088848427, + "loss": 1.0007, + "step": 2463 + }, + { + "epoch": 0.43874643874643876, + "grad_norm": 0.4895429313182831, + "learning_rate": 0.00019420528615689202, + "loss": 1.0032, + "step": 2464 + }, + { + "epoch": 0.4389245014245014, + "grad_norm": 0.5029937028884888, + "learning_rate": 0.00019420058957954285, + "loss": 1.2877, + "step": 2465 + }, + { + "epoch": 0.4391025641025641, + "grad_norm": 0.4953192174434662, + "learning_rate": 0.00019419589115652884, + "loss": 1.0759, + "step": 2466 + }, + { + "epoch": 0.4392806267806268, + "grad_norm": 0.5081778168678284, + "learning_rate": 0.000194191190887942, + "loss": 0.8816, + "step": 2467 + }, + { + "epoch": 0.43945868945868943, + "grad_norm": 0.5065913200378418, + "learning_rate": 0.00019418648877387446, + "loss": 1.0362, + "step": 2468 + }, + { + "epoch": 0.43963675213675213, + "grad_norm": 0.540600061416626, + "learning_rate": 0.00019418178481441832, + "loss": 1.0911, + "step": 2469 + }, + { + "epoch": 0.4398148148148148, + "grad_norm": 0.5122954845428467, + "learning_rate": 0.00019417707900966572, + "loss": 0.9866, + "step": 2470 + }, + { + "epoch": 0.43999287749287747, + "grad_norm": 0.5380190014839172, + "learning_rate": 0.00019417237135970893, + "loss": 1.2775, + "step": 2471 + }, + { + "epoch": 0.44017094017094016, + "grad_norm": 1.2977570295333862, + "learning_rate": 0.00019416766186464016, + "loss": 1.3993, + "step": 2472 + }, + { + "epoch": 0.44034900284900286, + "grad_norm": 0.48105308413505554, + "learning_rate": 0.00019416295052455165, + "loss": 0.9369, + "step": 2473 + }, + { + "epoch": 0.44052706552706555, + "grad_norm": 0.4742157459259033, + "learning_rate": 0.00019415823733953574, + "loss": 1.101, + "step": 2474 + }, + { + "epoch": 0.4407051282051282, + "grad_norm": 0.4958631694316864, + "learning_rate": 0.00019415352230968473, + "loss": 0.9906, + "step": 2475 + }, + { + "epoch": 0.4408831908831909, + "grad_norm": 0.5808146595954895, + "learning_rate": 0.00019414880543509107, + "loss": 1.2315, + "step": 2476 + }, + { + "epoch": 0.4410612535612536, + "grad_norm": 0.4294755160808563, + "learning_rate": 0.00019414408671584714, + "loss": 0.8275, + "step": 2477 + }, + { + "epoch": 0.4412393162393162, + "grad_norm": 0.5346055626869202, + "learning_rate": 0.0001941393661520454, + "loss": 1.2432, + "step": 2478 + }, + { + "epoch": 0.4414173789173789, + "grad_norm": 0.5827590227127075, + "learning_rate": 0.00019413464374377833, + "loss": 1.3204, + "step": 2479 + }, + { + "epoch": 0.4415954415954416, + "grad_norm": 0.45688143372535706, + "learning_rate": 0.00019412991949113847, + "loss": 0.9307, + "step": 2480 + }, + { + "epoch": 0.44177350427350426, + "grad_norm": 0.512999415397644, + "learning_rate": 0.0001941251933942184, + "loss": 1.2808, + "step": 2481 + }, + { + "epoch": 0.44195156695156695, + "grad_norm": 0.4546334445476532, + "learning_rate": 0.00019412046545311064, + "loss": 1.0156, + "step": 2482 + }, + { + "epoch": 0.44212962962962965, + "grad_norm": 0.48552581667900085, + "learning_rate": 0.00019411573566790793, + "loss": 1.3798, + "step": 2483 + }, + { + "epoch": 0.4423076923076923, + "grad_norm": 0.511970579624176, + "learning_rate": 0.00019411100403870287, + "loss": 1.065, + "step": 2484 + }, + { + "epoch": 0.442485754985755, + "grad_norm": 0.6367824077606201, + "learning_rate": 0.00019410627056558815, + "loss": 1.3242, + "step": 2485 + }, + { + "epoch": 0.4426638176638177, + "grad_norm": 0.48913368582725525, + "learning_rate": 0.00019410153524865659, + "loss": 0.9761, + "step": 2486 + }, + { + "epoch": 0.4428418803418803, + "grad_norm": 0.5077710151672363, + "learning_rate": 0.0001940967980880009, + "loss": 1.1023, + "step": 2487 + }, + { + "epoch": 0.443019943019943, + "grad_norm": 0.4956335723400116, + "learning_rate": 0.00019409205908371395, + "loss": 1.1788, + "step": 2488 + }, + { + "epoch": 0.4431980056980057, + "grad_norm": 0.4726616442203522, + "learning_rate": 0.00019408731823588853, + "loss": 1.1445, + "step": 2489 + }, + { + "epoch": 0.44337606837606836, + "grad_norm": 0.5676438212394714, + "learning_rate": 0.00019408257554461757, + "loss": 1.0344, + "step": 2490 + }, + { + "epoch": 0.44355413105413105, + "grad_norm": 0.537656843662262, + "learning_rate": 0.000194077831009994, + "loss": 0.9876, + "step": 2491 + }, + { + "epoch": 0.44373219373219375, + "grad_norm": 0.517905592918396, + "learning_rate": 0.00019407308463211074, + "loss": 1.1389, + "step": 2492 + }, + { + "epoch": 0.4439102564102564, + "grad_norm": 0.49227026104927063, + "learning_rate": 0.0001940683364110608, + "loss": 1.0351, + "step": 2493 + }, + { + "epoch": 0.4440883190883191, + "grad_norm": 0.5131173729896545, + "learning_rate": 0.00019406358634693725, + "loss": 1.0351, + "step": 2494 + }, + { + "epoch": 0.4442663817663818, + "grad_norm": 0.5064495205879211, + "learning_rate": 0.0001940588344398331, + "loss": 1.0248, + "step": 2495 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 0.44107526540756226, + "learning_rate": 0.00019405408068984148, + "loss": 0.8068, + "step": 2496 + }, + { + "epoch": 0.4446225071225071, + "grad_norm": 0.6711848378181458, + "learning_rate": 0.00019404932509705554, + "loss": 1.059, + "step": 2497 + }, + { + "epoch": 0.4448005698005698, + "grad_norm": 0.5862596035003662, + "learning_rate": 0.00019404456766156845, + "loss": 1.2012, + "step": 2498 + }, + { + "epoch": 0.44497863247863245, + "grad_norm": 0.5528512001037598, + "learning_rate": 0.0001940398083834734, + "loss": 1.1121, + "step": 2499 + }, + { + "epoch": 0.44515669515669515, + "grad_norm": 0.5326655507087708, + "learning_rate": 0.0001940350472628637, + "loss": 1.166, + "step": 2500 + }, + { + "epoch": 0.44533475783475784, + "grad_norm": 0.5384873747825623, + "learning_rate": 0.00019403028429983252, + "loss": 1.4111, + "step": 2501 + }, + { + "epoch": 0.44551282051282054, + "grad_norm": 0.5142310857772827, + "learning_rate": 0.0001940255194944733, + "loss": 1.3353, + "step": 2502 + }, + { + "epoch": 0.4456908831908832, + "grad_norm": 0.49124231934547424, + "learning_rate": 0.0001940207528468793, + "loss": 1.1443, + "step": 2503 + }, + { + "epoch": 0.4458689458689459, + "grad_norm": 0.509713888168335, + "learning_rate": 0.000194015984357144, + "loss": 1.1857, + "step": 2504 + }, + { + "epoch": 0.44604700854700857, + "grad_norm": 0.5211275219917297, + "learning_rate": 0.00019401121402536078, + "loss": 0.9911, + "step": 2505 + }, + { + "epoch": 0.4462250712250712, + "grad_norm": 0.480340838432312, + "learning_rate": 0.00019400644185162312, + "loss": 1.1018, + "step": 2506 + }, + { + "epoch": 0.4464031339031339, + "grad_norm": 0.4212559163570404, + "learning_rate": 0.00019400166783602448, + "loss": 0.7501, + "step": 2507 + }, + { + "epoch": 0.4465811965811966, + "grad_norm": 0.5110511183738708, + "learning_rate": 0.00019399689197865846, + "loss": 1.1244, + "step": 2508 + }, + { + "epoch": 0.44675925925925924, + "grad_norm": 0.5604230165481567, + "learning_rate": 0.0001939921142796186, + "loss": 1.1066, + "step": 2509 + }, + { + "epoch": 0.44693732193732194, + "grad_norm": 0.5578675270080566, + "learning_rate": 0.0001939873347389985, + "loss": 1.0514, + "step": 2510 + }, + { + "epoch": 0.44711538461538464, + "grad_norm": 0.520908772945404, + "learning_rate": 0.00019398255335689184, + "loss": 1.1217, + "step": 2511 + }, + { + "epoch": 0.4472934472934473, + "grad_norm": 0.4405131936073303, + "learning_rate": 0.00019397777013339224, + "loss": 1.043, + "step": 2512 + }, + { + "epoch": 0.44747150997151, + "grad_norm": 0.5217751860618591, + "learning_rate": 0.0001939729850685935, + "loss": 1.1301, + "step": 2513 + }, + { + "epoch": 0.44764957264957267, + "grad_norm": 0.6151493191719055, + "learning_rate": 0.00019396819816258932, + "loss": 1.3498, + "step": 2514 + }, + { + "epoch": 0.4478276353276353, + "grad_norm": 0.5622836947441101, + "learning_rate": 0.0001939634094154735, + "loss": 1.146, + "step": 2515 + }, + { + "epoch": 0.448005698005698, + "grad_norm": 0.4671688973903656, + "learning_rate": 0.00019395861882733984, + "loss": 0.9456, + "step": 2516 + }, + { + "epoch": 0.4481837606837607, + "grad_norm": 0.453951358795166, + "learning_rate": 0.00019395382639828223, + "loss": 1.0042, + "step": 2517 + }, + { + "epoch": 0.44836182336182334, + "grad_norm": 0.5150699615478516, + "learning_rate": 0.0001939490321283946, + "loss": 1.166, + "step": 2518 + }, + { + "epoch": 0.44853988603988604, + "grad_norm": 0.5718298554420471, + "learning_rate": 0.0001939442360177708, + "loss": 1.2033, + "step": 2519 + }, + { + "epoch": 0.44871794871794873, + "grad_norm": 0.5306782126426697, + "learning_rate": 0.00019393943806650488, + "loss": 1.0765, + "step": 2520 + }, + { + "epoch": 0.4488960113960114, + "grad_norm": 0.47633033990859985, + "learning_rate": 0.0001939346382746908, + "loss": 0.9957, + "step": 2521 + }, + { + "epoch": 0.44907407407407407, + "grad_norm": 0.496441513299942, + "learning_rate": 0.00019392983664242262, + "loss": 1.2016, + "step": 2522 + }, + { + "epoch": 0.44925213675213677, + "grad_norm": 0.45956477522850037, + "learning_rate": 0.00019392503316979442, + "loss": 1.026, + "step": 2523 + }, + { + "epoch": 0.4494301994301994, + "grad_norm": 0.5400575995445251, + "learning_rate": 0.0001939202278569003, + "loss": 1.0785, + "step": 2524 + }, + { + "epoch": 0.4496082621082621, + "grad_norm": 0.4847868084907532, + "learning_rate": 0.00019391542070383442, + "loss": 1.013, + "step": 2525 + }, + { + "epoch": 0.4497863247863248, + "grad_norm": 0.4694063663482666, + "learning_rate": 0.00019391061171069094, + "loss": 0.8793, + "step": 2526 + }, + { + "epoch": 0.44996438746438744, + "grad_norm": 0.5158169269561768, + "learning_rate": 0.00019390580087756413, + "loss": 0.9602, + "step": 2527 + }, + { + "epoch": 0.45014245014245013, + "grad_norm": 0.5404585003852844, + "learning_rate": 0.00019390098820454822, + "loss": 1.2247, + "step": 2528 + }, + { + "epoch": 0.45032051282051283, + "grad_norm": 0.5302738547325134, + "learning_rate": 0.00019389617369173752, + "loss": 0.918, + "step": 2529 + }, + { + "epoch": 0.45049857549857547, + "grad_norm": 0.5065485835075378, + "learning_rate": 0.00019389135733922634, + "loss": 1.0934, + "step": 2530 + }, + { + "epoch": 0.45067663817663817, + "grad_norm": 0.5491471886634827, + "learning_rate": 0.00019388653914710903, + "loss": 1.0736, + "step": 2531 + }, + { + "epoch": 0.45085470085470086, + "grad_norm": 0.4850206971168518, + "learning_rate": 0.00019388171911548005, + "loss": 1.2401, + "step": 2532 + }, + { + "epoch": 0.45103276353276356, + "grad_norm": 0.5419789552688599, + "learning_rate": 0.0001938768972444338, + "loss": 1.269, + "step": 2533 + }, + { + "epoch": 0.4512108262108262, + "grad_norm": 0.4209023714065552, + "learning_rate": 0.00019387207353406476, + "loss": 1.0544, + "step": 2534 + }, + { + "epoch": 0.4513888888888889, + "grad_norm": 0.578588604927063, + "learning_rate": 0.00019386724798446743, + "loss": 1.0564, + "step": 2535 + }, + { + "epoch": 0.4515669515669516, + "grad_norm": 0.5277524590492249, + "learning_rate": 0.00019386242059573638, + "loss": 1.1497, + "step": 2536 + }, + { + "epoch": 0.45174501424501423, + "grad_norm": 0.5536073446273804, + "learning_rate": 0.0001938575913679662, + "loss": 1.2213, + "step": 2537 + }, + { + "epoch": 0.4519230769230769, + "grad_norm": 0.5572254657745361, + "learning_rate": 0.00019385276030125143, + "loss": 1.0231, + "step": 2538 + }, + { + "epoch": 0.4521011396011396, + "grad_norm": 0.493847131729126, + "learning_rate": 0.00019384792739568686, + "loss": 0.9385, + "step": 2539 + }, + { + "epoch": 0.45227920227920226, + "grad_norm": 0.4641396403312683, + "learning_rate": 0.00019384309265136707, + "loss": 0.9332, + "step": 2540 + }, + { + "epoch": 0.45245726495726496, + "grad_norm": 0.5439442992210388, + "learning_rate": 0.00019383825606838681, + "loss": 1.317, + "step": 2541 + }, + { + "epoch": 0.45263532763532766, + "grad_norm": 0.7050970792770386, + "learning_rate": 0.00019383341764684086, + "loss": 0.9508, + "step": 2542 + }, + { + "epoch": 0.4528133903133903, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.000193828577386824, + "loss": 1.2704, + "step": 2543 + }, + { + "epoch": 0.452991452991453, + "grad_norm": 0.47641924023628235, + "learning_rate": 0.0001938237352884311, + "loss": 1.0101, + "step": 2544 + }, + { + "epoch": 0.4531695156695157, + "grad_norm": 0.5223637819290161, + "learning_rate": 0.000193818891351757, + "loss": 1.0548, + "step": 2545 + }, + { + "epoch": 0.45334757834757833, + "grad_norm": 0.49065667390823364, + "learning_rate": 0.0001938140455768966, + "loss": 1.0927, + "step": 2546 + }, + { + "epoch": 0.453525641025641, + "grad_norm": 0.4808312654495239, + "learning_rate": 0.0001938091979639449, + "loss": 1.0599, + "step": 2547 + }, + { + "epoch": 0.4537037037037037, + "grad_norm": 0.5157489776611328, + "learning_rate": 0.0001938043485129968, + "loss": 1.2596, + "step": 2548 + }, + { + "epoch": 0.45388176638176636, + "grad_norm": 0.5983387231826782, + "learning_rate": 0.0001937994972241474, + "loss": 1.2276, + "step": 2549 + }, + { + "epoch": 0.45405982905982906, + "grad_norm": 0.49776506423950195, + "learning_rate": 0.00019379464409749163, + "loss": 1.3666, + "step": 2550 + }, + { + "epoch": 0.45423789173789175, + "grad_norm": 0.4693490266799927, + "learning_rate": 0.00019378978913312471, + "loss": 1.087, + "step": 2551 + }, + { + "epoch": 0.4544159544159544, + "grad_norm": 0.4754335880279541, + "learning_rate": 0.00019378493233114167, + "loss": 1.1282, + "step": 2552 + }, + { + "epoch": 0.4545940170940171, + "grad_norm": 0.5852862000465393, + "learning_rate": 0.00019378007369163776, + "loss": 1.1113, + "step": 2553 + }, + { + "epoch": 0.4547720797720798, + "grad_norm": 0.47442635893821716, + "learning_rate": 0.00019377521321470805, + "loss": 0.983, + "step": 2554 + }, + { + "epoch": 0.4549501424501424, + "grad_norm": 0.47432273626327515, + "learning_rate": 0.00019377035090044787, + "loss": 1.0169, + "step": 2555 + }, + { + "epoch": 0.4551282051282051, + "grad_norm": 0.4929196834564209, + "learning_rate": 0.00019376548674895246, + "loss": 1.0182, + "step": 2556 + }, + { + "epoch": 0.4553062678062678, + "grad_norm": 0.5433184504508972, + "learning_rate": 0.00019376062076031708, + "loss": 1.1339, + "step": 2557 + }, + { + "epoch": 0.45548433048433046, + "grad_norm": 0.47430408000946045, + "learning_rate": 0.00019375575293463715, + "loss": 1.1589, + "step": 2558 + }, + { + "epoch": 0.45566239316239315, + "grad_norm": 0.46641045808792114, + "learning_rate": 0.000193750883272008, + "loss": 1.029, + "step": 2559 + }, + { + "epoch": 0.45584045584045585, + "grad_norm": 0.44476228952407837, + "learning_rate": 0.00019374601177252502, + "loss": 0.8494, + "step": 2560 + }, + { + "epoch": 0.45601851851851855, + "grad_norm": 0.4886183440685272, + "learning_rate": 0.00019374113843628366, + "loss": 1.1374, + "step": 2561 + }, + { + "epoch": 0.4561965811965812, + "grad_norm": 0.4786703288555145, + "learning_rate": 0.00019373626326337946, + "loss": 1.2861, + "step": 2562 + }, + { + "epoch": 0.4563746438746439, + "grad_norm": 0.5752716660499573, + "learning_rate": 0.0001937313862539079, + "loss": 1.2365, + "step": 2563 + }, + { + "epoch": 0.4565527065527066, + "grad_norm": 0.519176185131073, + "learning_rate": 0.00019372650740796452, + "loss": 1.2264, + "step": 2564 + }, + { + "epoch": 0.4567307692307692, + "grad_norm": 0.5927292704582214, + "learning_rate": 0.00019372162672564493, + "loss": 0.8979, + "step": 2565 + }, + { + "epoch": 0.4569088319088319, + "grad_norm": 0.5467435121536255, + "learning_rate": 0.00019371674420704478, + "loss": 1.1016, + "step": 2566 + }, + { + "epoch": 0.4570868945868946, + "grad_norm": 0.49593284726142883, + "learning_rate": 0.00019371185985225968, + "loss": 0.982, + "step": 2567 + }, + { + "epoch": 0.45726495726495725, + "grad_norm": 0.5696587562561035, + "learning_rate": 0.00019370697366138538, + "loss": 0.979, + "step": 2568 + }, + { + "epoch": 0.45744301994301995, + "grad_norm": 0.4455752968788147, + "learning_rate": 0.00019370208563451757, + "loss": 0.8832, + "step": 2569 + }, + { + "epoch": 0.45762108262108264, + "grad_norm": 0.5072923302650452, + "learning_rate": 0.00019369719577175203, + "loss": 1.1046, + "step": 2570 + }, + { + "epoch": 0.4577991452991453, + "grad_norm": 0.45119982957839966, + "learning_rate": 0.0001936923040731846, + "loss": 1.0083, + "step": 2571 + }, + { + "epoch": 0.457977207977208, + "grad_norm": 0.5062251091003418, + "learning_rate": 0.00019368741053891108, + "loss": 1.2771, + "step": 2572 + }, + { + "epoch": 0.4581552706552707, + "grad_norm": 0.5511104464530945, + "learning_rate": 0.0001936825151690274, + "loss": 1.0039, + "step": 2573 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 0.4721006453037262, + "learning_rate": 0.0001936776179636294, + "loss": 1.3246, + "step": 2574 + }, + { + "epoch": 0.458511396011396, + "grad_norm": 0.5021488666534424, + "learning_rate": 0.0001936727189228131, + "loss": 1.1733, + "step": 2575 + }, + { + "epoch": 0.4586894586894587, + "grad_norm": 0.5755292177200317, + "learning_rate": 0.0001936678180466745, + "loss": 1.2241, + "step": 2576 + }, + { + "epoch": 0.45886752136752135, + "grad_norm": 0.4501610994338989, + "learning_rate": 0.00019366291533530952, + "loss": 1.0503, + "step": 2577 + }, + { + "epoch": 0.45904558404558404, + "grad_norm": 0.4067458212375641, + "learning_rate": 0.00019365801078881432, + "loss": 0.8259, + "step": 2578 + }, + { + "epoch": 0.45922364672364674, + "grad_norm": 0.539730429649353, + "learning_rate": 0.0001936531044072849, + "loss": 1.1964, + "step": 2579 + }, + { + "epoch": 0.4594017094017094, + "grad_norm": 0.5624797344207764, + "learning_rate": 0.0001936481961908175, + "loss": 1.2059, + "step": 2580 + }, + { + "epoch": 0.4595797720797721, + "grad_norm": 0.43679240345954895, + "learning_rate": 0.00019364328613950824, + "loss": 1.1371, + "step": 2581 + }, + { + "epoch": 0.45975783475783477, + "grad_norm": 0.5214769244194031, + "learning_rate": 0.00019363837425345328, + "loss": 1.109, + "step": 2582 + }, + { + "epoch": 0.4599358974358974, + "grad_norm": 0.4522894024848938, + "learning_rate": 0.00019363346053274892, + "loss": 1.0532, + "step": 2583 + }, + { + "epoch": 0.4601139601139601, + "grad_norm": 0.44980281591415405, + "learning_rate": 0.0001936285449774914, + "loss": 0.9352, + "step": 2584 + }, + { + "epoch": 0.4602920227920228, + "grad_norm": 0.5697414875030518, + "learning_rate": 0.00019362362758777705, + "loss": 1.2171, + "step": 2585 + }, + { + "epoch": 0.46047008547008544, + "grad_norm": 0.4636315107345581, + "learning_rate": 0.00019361870836370217, + "loss": 1.0662, + "step": 2586 + }, + { + "epoch": 0.46064814814814814, + "grad_norm": 0.5144017338752747, + "learning_rate": 0.00019361378730536321, + "loss": 1.0681, + "step": 2587 + }, + { + "epoch": 0.46082621082621084, + "grad_norm": 0.5007636547088623, + "learning_rate": 0.00019360886441285654, + "loss": 1.2058, + "step": 2588 + }, + { + "epoch": 0.46100427350427353, + "grad_norm": 0.5024117231369019, + "learning_rate": 0.00019360393968627864, + "loss": 1.065, + "step": 2589 + }, + { + "epoch": 0.46118233618233617, + "grad_norm": 0.48105588555336, + "learning_rate": 0.00019359901312572596, + "loss": 1.0887, + "step": 2590 + }, + { + "epoch": 0.46136039886039887, + "grad_norm": 0.5381982326507568, + "learning_rate": 0.00019359408473129506, + "loss": 1.2754, + "step": 2591 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 0.5051333904266357, + "learning_rate": 0.0001935891545030825, + "loss": 0.9334, + "step": 2592 + }, + { + "epoch": 0.4617165242165242, + "grad_norm": 0.43818601965904236, + "learning_rate": 0.0001935842224411849, + "loss": 1.0967, + "step": 2593 + }, + { + "epoch": 0.4618945868945869, + "grad_norm": 0.4727257490158081, + "learning_rate": 0.0001935792885456988, + "loss": 0.8136, + "step": 2594 + }, + { + "epoch": 0.4620726495726496, + "grad_norm": 0.5505291223526001, + "learning_rate": 0.00019357435281672098, + "loss": 1.3113, + "step": 2595 + }, + { + "epoch": 0.46225071225071224, + "grad_norm": 0.4705682396888733, + "learning_rate": 0.0001935694152543481, + "loss": 0.9863, + "step": 2596 + }, + { + "epoch": 0.46242877492877493, + "grad_norm": 0.49653419852256775, + "learning_rate": 0.0001935644758586769, + "loss": 1.035, + "step": 2597 + }, + { + "epoch": 0.46260683760683763, + "grad_norm": 0.4788367748260498, + "learning_rate": 0.00019355953462980415, + "loss": 1.1253, + "step": 2598 + }, + { + "epoch": 0.46278490028490027, + "grad_norm": 0.5295125842094421, + "learning_rate": 0.00019355459156782668, + "loss": 1.0853, + "step": 2599 + }, + { + "epoch": 0.46296296296296297, + "grad_norm": 0.4878056049346924, + "learning_rate": 0.00019354964667284133, + "loss": 1.1381, + "step": 2600 + }, + { + "epoch": 0.46314102564102566, + "grad_norm": 0.5442031025886536, + "learning_rate": 0.00019354469994494497, + "loss": 1.1349, + "step": 2601 + }, + { + "epoch": 0.4633190883190883, + "grad_norm": 0.4845225214958191, + "learning_rate": 0.00019353975138423457, + "loss": 1.0538, + "step": 2602 + }, + { + "epoch": 0.463497150997151, + "grad_norm": 0.4957871437072754, + "learning_rate": 0.00019353480099080703, + "loss": 1.2765, + "step": 2603 + }, + { + "epoch": 0.4636752136752137, + "grad_norm": 0.5414339303970337, + "learning_rate": 0.00019352984876475936, + "loss": 1.1015, + "step": 2604 + }, + { + "epoch": 0.46385327635327633, + "grad_norm": 0.5171043872833252, + "learning_rate": 0.0001935248947061886, + "loss": 0.9995, + "step": 2605 + }, + { + "epoch": 0.46403133903133903, + "grad_norm": 0.46040529012680054, + "learning_rate": 0.0001935199388151918, + "loss": 1.1126, + "step": 2606 + }, + { + "epoch": 0.4642094017094017, + "grad_norm": 0.5327033400535583, + "learning_rate": 0.00019351498109186613, + "loss": 1.1983, + "step": 2607 + }, + { + "epoch": 0.46438746438746437, + "grad_norm": 0.4451361298561096, + "learning_rate": 0.0001935100215363086, + "loss": 0.9689, + "step": 2608 + }, + { + "epoch": 0.46456552706552706, + "grad_norm": 0.5462809801101685, + "learning_rate": 0.00019350506014861646, + "loss": 1.036, + "step": 2609 + }, + { + "epoch": 0.46474358974358976, + "grad_norm": 0.4907000958919525, + "learning_rate": 0.00019350009692888694, + "loss": 1.0724, + "step": 2610 + }, + { + "epoch": 0.4649216524216524, + "grad_norm": 0.47523510456085205, + "learning_rate": 0.00019349513187721723, + "loss": 0.9214, + "step": 2611 + }, + { + "epoch": 0.4650997150997151, + "grad_norm": 0.539732813835144, + "learning_rate": 0.0001934901649937046, + "loss": 1.1166, + "step": 2612 + }, + { + "epoch": 0.4652777777777778, + "grad_norm": 0.4827860891819, + "learning_rate": 0.00019348519627844643, + "loss": 1.1613, + "step": 2613 + }, + { + "epoch": 0.46545584045584043, + "grad_norm": 0.5385223031044006, + "learning_rate": 0.00019348022573154, + "loss": 1.0105, + "step": 2614 + }, + { + "epoch": 0.4656339031339031, + "grad_norm": 0.4629383087158203, + "learning_rate": 0.0001934752533530828, + "loss": 1.0298, + "step": 2615 + }, + { + "epoch": 0.4658119658119658, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00019347027914317212, + "loss": 1.3158, + "step": 2616 + }, + { + "epoch": 0.46599002849002846, + "grad_norm": 0.5954698324203491, + "learning_rate": 0.00019346530310190553, + "loss": 1.1882, + "step": 2617 + }, + { + "epoch": 0.46616809116809116, + "grad_norm": 0.49185171723365784, + "learning_rate": 0.00019346032522938046, + "loss": 1.0977, + "step": 2618 + }, + { + "epoch": 0.46634615384615385, + "grad_norm": 0.5145422220230103, + "learning_rate": 0.0001934553455256945, + "loss": 0.9948, + "step": 2619 + }, + { + "epoch": 0.46652421652421655, + "grad_norm": 0.6809412837028503, + "learning_rate": 0.00019345036399094517, + "loss": 1.5798, + "step": 2620 + }, + { + "epoch": 0.4667022792022792, + "grad_norm": 0.4606841206550598, + "learning_rate": 0.00019344538062523005, + "loss": 0.7357, + "step": 2621 + }, + { + "epoch": 0.4668803418803419, + "grad_norm": 0.49036628007888794, + "learning_rate": 0.00019344039542864685, + "loss": 1.1518, + "step": 2622 + }, + { + "epoch": 0.4670584045584046, + "grad_norm": 0.47904539108276367, + "learning_rate": 0.0001934354084012932, + "loss": 0.9929, + "step": 2623 + }, + { + "epoch": 0.4672364672364672, + "grad_norm": 0.5224666595458984, + "learning_rate": 0.0001934304195432668, + "loss": 1.2544, + "step": 2624 + }, + { + "epoch": 0.4674145299145299, + "grad_norm": 0.4902483820915222, + "learning_rate": 0.00019342542885466543, + "loss": 1.0301, + "step": 2625 + }, + { + "epoch": 0.4675925925925926, + "grad_norm": 0.46824702620506287, + "learning_rate": 0.00019342043633558683, + "loss": 0.9364, + "step": 2626 + }, + { + "epoch": 0.46777065527065526, + "grad_norm": 0.46272051334381104, + "learning_rate": 0.00019341544198612888, + "loss": 1.056, + "step": 2627 + }, + { + "epoch": 0.46794871794871795, + "grad_norm": 0.6216606497764587, + "learning_rate": 0.0001934104458063894, + "loss": 1.0825, + "step": 2628 + }, + { + "epoch": 0.46812678062678065, + "grad_norm": 0.5024014115333557, + "learning_rate": 0.00019340544779646623, + "loss": 1.1832, + "step": 2629 + }, + { + "epoch": 0.4683048433048433, + "grad_norm": 0.5547130107879639, + "learning_rate": 0.00019340044795645737, + "loss": 1.1335, + "step": 2630 + }, + { + "epoch": 0.468482905982906, + "grad_norm": 0.5439161658287048, + "learning_rate": 0.0001933954462864608, + "loss": 1.0229, + "step": 2631 + }, + { + "epoch": 0.4686609686609687, + "grad_norm": 0.4782990515232086, + "learning_rate": 0.0001933904427865744, + "loss": 1.2318, + "step": 2632 + }, + { + "epoch": 0.4688390313390313, + "grad_norm": 0.5872140526771545, + "learning_rate": 0.00019338543745689633, + "loss": 1.0132, + "step": 2633 + }, + { + "epoch": 0.469017094017094, + "grad_norm": 0.44163307547569275, + "learning_rate": 0.00019338043029752458, + "loss": 1.0091, + "step": 2634 + }, + { + "epoch": 0.4691951566951567, + "grad_norm": 0.541081428527832, + "learning_rate": 0.0001933754213085573, + "loss": 1.2155, + "step": 2635 + }, + { + "epoch": 0.46937321937321935, + "grad_norm": 0.4761527478694916, + "learning_rate": 0.00019337041049009255, + "loss": 1.1138, + "step": 2636 + }, + { + "epoch": 0.46955128205128205, + "grad_norm": 0.46414369344711304, + "learning_rate": 0.0001933653978422286, + "loss": 0.9903, + "step": 2637 + }, + { + "epoch": 0.46972934472934474, + "grad_norm": 0.5337086915969849, + "learning_rate": 0.00019336038336506363, + "loss": 1.2873, + "step": 2638 + }, + { + "epoch": 0.4699074074074074, + "grad_norm": 0.5065379738807678, + "learning_rate": 0.00019335536705869592, + "loss": 1.1436, + "step": 2639 + }, + { + "epoch": 0.4700854700854701, + "grad_norm": 0.5539217591285706, + "learning_rate": 0.0001933503489232237, + "loss": 1.2881, + "step": 2640 + }, + { + "epoch": 0.4702635327635328, + "grad_norm": 0.48303213715553284, + "learning_rate": 0.0001933453289587453, + "loss": 1.0209, + "step": 2641 + }, + { + "epoch": 0.4704415954415954, + "grad_norm": 0.6986871957778931, + "learning_rate": 0.00019334030716535908, + "loss": 1.1979, + "step": 2642 + }, + { + "epoch": 0.4706196581196581, + "grad_norm": 0.46137234568595886, + "learning_rate": 0.00019333528354316347, + "loss": 1.0682, + "step": 2643 + }, + { + "epoch": 0.4707977207977208, + "grad_norm": 0.4726654291152954, + "learning_rate": 0.00019333025809225684, + "loss": 1.1712, + "step": 2644 + }, + { + "epoch": 0.47097578347578345, + "grad_norm": 0.46188637614250183, + "learning_rate": 0.0001933252308127377, + "loss": 1.0183, + "step": 2645 + }, + { + "epoch": 0.47115384615384615, + "grad_norm": 0.5323259830474854, + "learning_rate": 0.0001933202017047045, + "loss": 0.935, + "step": 2646 + }, + { + "epoch": 0.47133190883190884, + "grad_norm": 0.5004189014434814, + "learning_rate": 0.00019331517076825582, + "loss": 1.1331, + "step": 2647 + }, + { + "epoch": 0.47150997150997154, + "grad_norm": 0.5443634986877441, + "learning_rate": 0.0001933101380034902, + "loss": 1.0514, + "step": 2648 + }, + { + "epoch": 0.4716880341880342, + "grad_norm": 0.504180371761322, + "learning_rate": 0.0001933051034105063, + "loss": 1.3099, + "step": 2649 + }, + { + "epoch": 0.4718660968660969, + "grad_norm": 0.5092344284057617, + "learning_rate": 0.0001933000669894027, + "loss": 1.0716, + "step": 2650 + }, + { + "epoch": 0.47204415954415957, + "grad_norm": 0.5236422419548035, + "learning_rate": 0.0001932950287402781, + "loss": 1.0981, + "step": 2651 + }, + { + "epoch": 0.4722222222222222, + "grad_norm": 0.6228063702583313, + "learning_rate": 0.0001932899886632312, + "loss": 1.3398, + "step": 2652 + }, + { + "epoch": 0.4724002849002849, + "grad_norm": 0.5112748146057129, + "learning_rate": 0.00019328494675836078, + "loss": 1.0151, + "step": 2653 + }, + { + "epoch": 0.4725783475783476, + "grad_norm": 0.5554201602935791, + "learning_rate": 0.00019327990302576563, + "loss": 1.404, + "step": 2654 + }, + { + "epoch": 0.47275641025641024, + "grad_norm": 0.5050725340843201, + "learning_rate": 0.0001932748574655445, + "loss": 0.951, + "step": 2655 + }, + { + "epoch": 0.47293447293447294, + "grad_norm": 0.5161749720573425, + "learning_rate": 0.00019326981007779636, + "loss": 1.2425, + "step": 2656 + }, + { + "epoch": 0.47311253561253563, + "grad_norm": 0.4865442216396332, + "learning_rate": 0.00019326476086262002, + "loss": 1.1175, + "step": 2657 + }, + { + "epoch": 0.4732905982905983, + "grad_norm": 0.5276186466217041, + "learning_rate": 0.0001932597098201144, + "loss": 1.3687, + "step": 2658 + }, + { + "epoch": 0.47346866096866097, + "grad_norm": 0.509139358997345, + "learning_rate": 0.00019325465695037855, + "loss": 1.0546, + "step": 2659 + }, + { + "epoch": 0.47364672364672367, + "grad_norm": 0.49815434217453003, + "learning_rate": 0.00019324960225351138, + "loss": 1.0807, + "step": 2660 + }, + { + "epoch": 0.4738247863247863, + "grad_norm": 0.5059618353843689, + "learning_rate": 0.00019324454572961197, + "loss": 1.0827, + "step": 2661 + }, + { + "epoch": 0.474002849002849, + "grad_norm": 0.5698565244674683, + "learning_rate": 0.00019323948737877942, + "loss": 1.2019, + "step": 2662 + }, + { + "epoch": 0.4741809116809117, + "grad_norm": 0.49661511182785034, + "learning_rate": 0.00019323442720111276, + "loss": 1.1447, + "step": 2663 + }, + { + "epoch": 0.47435897435897434, + "grad_norm": 0.46442747116088867, + "learning_rate": 0.0001932293651967112, + "loss": 0.8796, + "step": 2664 + }, + { + "epoch": 0.47453703703703703, + "grad_norm": 0.48306044936180115, + "learning_rate": 0.00019322430136567388, + "loss": 1.1358, + "step": 2665 + }, + { + "epoch": 0.47471509971509973, + "grad_norm": 0.5677350759506226, + "learning_rate": 0.00019321923570810005, + "loss": 1.1026, + "step": 2666 + }, + { + "epoch": 0.47489316239316237, + "grad_norm": 0.3700144588947296, + "learning_rate": 0.0001932141682240889, + "loss": 0.7514, + "step": 2667 + }, + { + "epoch": 0.47507122507122507, + "grad_norm": 0.6003054976463318, + "learning_rate": 0.0001932090989137398, + "loss": 1.1591, + "step": 2668 + }, + { + "epoch": 0.47524928774928776, + "grad_norm": 0.520298421382904, + "learning_rate": 0.00019320402777715204, + "loss": 1.339, + "step": 2669 + }, + { + "epoch": 0.4754273504273504, + "grad_norm": 0.46453598141670227, + "learning_rate": 0.00019319895481442493, + "loss": 0.9879, + "step": 2670 + }, + { + "epoch": 0.4756054131054131, + "grad_norm": 0.5247363448143005, + "learning_rate": 0.00019319388002565793, + "loss": 0.9862, + "step": 2671 + }, + { + "epoch": 0.4757834757834758, + "grad_norm": 0.5498613715171814, + "learning_rate": 0.00019318880341095046, + "loss": 1.2224, + "step": 2672 + }, + { + "epoch": 0.47596153846153844, + "grad_norm": 0.565838098526001, + "learning_rate": 0.00019318372497040192, + "loss": 1.0712, + "step": 2673 + }, + { + "epoch": 0.47613960113960113, + "grad_norm": 0.5797489881515503, + "learning_rate": 0.00019317864470411191, + "loss": 1.0176, + "step": 2674 + }, + { + "epoch": 0.47631766381766383, + "grad_norm": 0.5114326477050781, + "learning_rate": 0.0001931735626121799, + "loss": 1.1027, + "step": 2675 + }, + { + "epoch": 0.47649572649572647, + "grad_norm": 0.5396515727043152, + "learning_rate": 0.00019316847869470547, + "loss": 1.1782, + "step": 2676 + }, + { + "epoch": 0.47667378917378916, + "grad_norm": 0.4812076985836029, + "learning_rate": 0.00019316339295178824, + "loss": 1.1196, + "step": 2677 + }, + { + "epoch": 0.47685185185185186, + "grad_norm": 0.4875647723674774, + "learning_rate": 0.00019315830538352787, + "loss": 1.1407, + "step": 2678 + }, + { + "epoch": 0.47702991452991456, + "grad_norm": 0.5036377906799316, + "learning_rate": 0.00019315321599002404, + "loss": 0.9842, + "step": 2679 + }, + { + "epoch": 0.4772079772079772, + "grad_norm": 0.5054177641868591, + "learning_rate": 0.00019314812477137645, + "loss": 0.8196, + "step": 2680 + }, + { + "epoch": 0.4773860398860399, + "grad_norm": 0.5050665736198425, + "learning_rate": 0.00019314303172768483, + "loss": 0.8463, + "step": 2681 + }, + { + "epoch": 0.4775641025641026, + "grad_norm": 0.5179004669189453, + "learning_rate": 0.000193137936859049, + "loss": 1.2485, + "step": 2682 + }, + { + "epoch": 0.47774216524216523, + "grad_norm": 0.44986143708229065, + "learning_rate": 0.00019313284016556876, + "loss": 0.9855, + "step": 2683 + }, + { + "epoch": 0.4779202279202279, + "grad_norm": 0.5594347715377808, + "learning_rate": 0.00019312774164734398, + "loss": 1.0987, + "step": 2684 + }, + { + "epoch": 0.4780982905982906, + "grad_norm": 0.4837244749069214, + "learning_rate": 0.0001931226413044746, + "loss": 1.1119, + "step": 2685 + }, + { + "epoch": 0.47827635327635326, + "grad_norm": 0.489145427942276, + "learning_rate": 0.0001931175391370605, + "loss": 1.1962, + "step": 2686 + }, + { + "epoch": 0.47845441595441596, + "grad_norm": 0.503568708896637, + "learning_rate": 0.00019311243514520164, + "loss": 0.9668, + "step": 2687 + }, + { + "epoch": 0.47863247863247865, + "grad_norm": 0.5401005744934082, + "learning_rate": 0.00019310732932899805, + "loss": 1.3072, + "step": 2688 + }, + { + "epoch": 0.4788105413105413, + "grad_norm": 0.526523768901825, + "learning_rate": 0.00019310222168854971, + "loss": 1.1387, + "step": 2689 + }, + { + "epoch": 0.478988603988604, + "grad_norm": 0.5223183631896973, + "learning_rate": 0.00019309711222395678, + "loss": 1.1391, + "step": 2690 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 0.5840879082679749, + "learning_rate": 0.00019309200093531933, + "loss": 1.1543, + "step": 2691 + }, + { + "epoch": 0.4793447293447293, + "grad_norm": 0.5173699259757996, + "learning_rate": 0.00019308688782273753, + "loss": 1.1889, + "step": 2692 + }, + { + "epoch": 0.479522792022792, + "grad_norm": 0.5417894124984741, + "learning_rate": 0.00019308177288631146, + "loss": 1.299, + "step": 2693 + }, + { + "epoch": 0.4797008547008547, + "grad_norm": 0.4890797734260559, + "learning_rate": 0.0001930766561261415, + "loss": 1.1516, + "step": 2694 + }, + { + "epoch": 0.47987891737891736, + "grad_norm": 0.5422119498252869, + "learning_rate": 0.00019307153754232772, + "loss": 1.0301, + "step": 2695 + }, + { + "epoch": 0.48005698005698005, + "grad_norm": 0.5838702917098999, + "learning_rate": 0.00019306641713497057, + "loss": 1.265, + "step": 2696 + }, + { + "epoch": 0.48023504273504275, + "grad_norm": 0.5020943284034729, + "learning_rate": 0.00019306129490417027, + "loss": 1.1119, + "step": 2697 + }, + { + "epoch": 0.4804131054131054, + "grad_norm": 0.412993460893631, + "learning_rate": 0.00019305617085002723, + "loss": 0.8083, + "step": 2698 + }, + { + "epoch": 0.4805911680911681, + "grad_norm": 0.6270101070404053, + "learning_rate": 0.00019305104497264184, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 0.4807692307692308, + "grad_norm": 0.45256730914115906, + "learning_rate": 0.0001930459172721145, + "loss": 1.0368, + "step": 2700 + }, + { + "epoch": 0.4809472934472934, + "grad_norm": 0.5351749658584595, + "learning_rate": 0.0001930407877485457, + "loss": 1.135, + "step": 2701 + }, + { + "epoch": 0.4811253561253561, + "grad_norm": 0.49324163794517517, + "learning_rate": 0.00019303565640203593, + "loss": 0.9383, + "step": 2702 + }, + { + "epoch": 0.4813034188034188, + "grad_norm": 0.5434361100196838, + "learning_rate": 0.00019303052323268576, + "loss": 1.2605, + "step": 2703 + }, + { + "epoch": 0.48148148148148145, + "grad_norm": 0.5858064889907837, + "learning_rate": 0.00019302538824059572, + "loss": 1.0846, + "step": 2704 + }, + { + "epoch": 0.48165954415954415, + "grad_norm": 0.5753700733184814, + "learning_rate": 0.00019302025142586647, + "loss": 1.0371, + "step": 2705 + }, + { + "epoch": 0.48183760683760685, + "grad_norm": 0.43102699518203735, + "learning_rate": 0.00019301511278859858, + "loss": 0.9189, + "step": 2706 + }, + { + "epoch": 0.48201566951566954, + "grad_norm": 0.4731025993824005, + "learning_rate": 0.0001930099723288928, + "loss": 1.1291, + "step": 2707 + }, + { + "epoch": 0.4821937321937322, + "grad_norm": 0.5685615539550781, + "learning_rate": 0.00019300483004684987, + "loss": 1.1006, + "step": 2708 + }, + { + "epoch": 0.4823717948717949, + "grad_norm": 0.4368155896663666, + "learning_rate": 0.00019299968594257044, + "loss": 0.9959, + "step": 2709 + }, + { + "epoch": 0.4825498575498576, + "grad_norm": 0.5594738125801086, + "learning_rate": 0.00019299454001615537, + "loss": 1.0826, + "step": 2710 + }, + { + "epoch": 0.4827279202279202, + "grad_norm": 0.48876598477363586, + "learning_rate": 0.00019298939226770548, + "loss": 1.1556, + "step": 2711 + }, + { + "epoch": 0.4829059829059829, + "grad_norm": 0.548039436340332, + "learning_rate": 0.00019298424269732157, + "loss": 1.158, + "step": 2712 + }, + { + "epoch": 0.4830840455840456, + "grad_norm": 0.4957645535469055, + "learning_rate": 0.00019297909130510464, + "loss": 0.9824, + "step": 2713 + }, + { + "epoch": 0.48326210826210825, + "grad_norm": 0.5197011232376099, + "learning_rate": 0.00019297393809115555, + "loss": 1.1074, + "step": 2714 + }, + { + "epoch": 0.48344017094017094, + "grad_norm": 0.5742064118385315, + "learning_rate": 0.00019296878305557526, + "loss": 1.0431, + "step": 2715 + }, + { + "epoch": 0.48361823361823364, + "grad_norm": 0.5698413252830505, + "learning_rate": 0.0001929636261984648, + "loss": 1.0713, + "step": 2716 + }, + { + "epoch": 0.4837962962962963, + "grad_norm": 0.48126333951950073, + "learning_rate": 0.0001929584675199252, + "loss": 0.9274, + "step": 2717 + }, + { + "epoch": 0.483974358974359, + "grad_norm": 0.49299830198287964, + "learning_rate": 0.00019295330702005754, + "loss": 0.9392, + "step": 2718 + }, + { + "epoch": 0.48415242165242167, + "grad_norm": 0.4780774414539337, + "learning_rate": 0.0001929481446989629, + "loss": 1.1459, + "step": 2719 + }, + { + "epoch": 0.4843304843304843, + "grad_norm": 0.5462654829025269, + "learning_rate": 0.00019294298055674248, + "loss": 1.0635, + "step": 2720 + }, + { + "epoch": 0.484508547008547, + "grad_norm": 0.5371061563491821, + "learning_rate": 0.00019293781459349743, + "loss": 1.3578, + "step": 2721 + }, + { + "epoch": 0.4846866096866097, + "grad_norm": 0.46308520436286926, + "learning_rate": 0.00019293264680932893, + "loss": 0.9001, + "step": 2722 + }, + { + "epoch": 0.48486467236467234, + "grad_norm": 0.5149807929992676, + "learning_rate": 0.0001929274772043383, + "loss": 0.6908, + "step": 2723 + }, + { + "epoch": 0.48504273504273504, + "grad_norm": 0.5435031056404114, + "learning_rate": 0.00019292230577862678, + "loss": 1.2143, + "step": 2724 + }, + { + "epoch": 0.48522079772079774, + "grad_norm": 0.44217726588249207, + "learning_rate": 0.00019291713253229568, + "loss": 0.9303, + "step": 2725 + }, + { + "epoch": 0.4853988603988604, + "grad_norm": 0.6120226383209229, + "learning_rate": 0.00019291195746544643, + "loss": 1.3801, + "step": 2726 + }, + { + "epoch": 0.4855769230769231, + "grad_norm": 0.5014316439628601, + "learning_rate": 0.00019290678057818037, + "loss": 1.0631, + "step": 2727 + }, + { + "epoch": 0.48575498575498577, + "grad_norm": 0.5667829513549805, + "learning_rate": 0.00019290160187059895, + "loss": 1.3166, + "step": 2728 + }, + { + "epoch": 0.4859330484330484, + "grad_norm": 0.5011509656906128, + "learning_rate": 0.0001928964213428036, + "loss": 1.1887, + "step": 2729 + }, + { + "epoch": 0.4861111111111111, + "grad_norm": 0.48317405581474304, + "learning_rate": 0.00019289123899489586, + "loss": 1.1125, + "step": 2730 + }, + { + "epoch": 0.4862891737891738, + "grad_norm": 0.4669005870819092, + "learning_rate": 0.00019288605482697726, + "loss": 1.0091, + "step": 2731 + }, + { + "epoch": 0.48646723646723644, + "grad_norm": 0.4330739974975586, + "learning_rate": 0.00019288086883914937, + "loss": 0.9789, + "step": 2732 + }, + { + "epoch": 0.48664529914529914, + "grad_norm": 0.48482781648635864, + "learning_rate": 0.0001928756810315138, + "loss": 1.1922, + "step": 2733 + }, + { + "epoch": 0.48682336182336183, + "grad_norm": 0.5781838297843933, + "learning_rate": 0.0001928704914041722, + "loss": 1.1793, + "step": 2734 + }, + { + "epoch": 0.48700142450142453, + "grad_norm": 0.5955413579940796, + "learning_rate": 0.00019286529995722623, + "loss": 1.1001, + "step": 2735 + }, + { + "epoch": 0.48717948717948717, + "grad_norm": 0.49204322695732117, + "learning_rate": 0.00019286010669077763, + "loss": 0.9219, + "step": 2736 + }, + { + "epoch": 0.48735754985754987, + "grad_norm": 0.5853500962257385, + "learning_rate": 0.00019285491160492813, + "loss": 1.1133, + "step": 2737 + }, + { + "epoch": 0.48753561253561256, + "grad_norm": 0.5555846095085144, + "learning_rate": 0.0001928497146997795, + "loss": 1.0915, + "step": 2738 + }, + { + "epoch": 0.4877136752136752, + "grad_norm": 0.5166759490966797, + "learning_rate": 0.00019284451597543364, + "loss": 0.9349, + "step": 2739 + }, + { + "epoch": 0.4878917378917379, + "grad_norm": 0.47816506028175354, + "learning_rate": 0.00019283931543199234, + "loss": 0.8978, + "step": 2740 + }, + { + "epoch": 0.4880698005698006, + "grad_norm": 0.5632442831993103, + "learning_rate": 0.0001928341130695575, + "loss": 1.0491, + "step": 2741 + }, + { + "epoch": 0.48824786324786323, + "grad_norm": 0.6532769799232483, + "learning_rate": 0.00019282890888823107, + "loss": 1.2779, + "step": 2742 + }, + { + "epoch": 0.48842592592592593, + "grad_norm": 0.5733640789985657, + "learning_rate": 0.000192823702888115, + "loss": 1.4127, + "step": 2743 + }, + { + "epoch": 0.4886039886039886, + "grad_norm": 0.5701746344566345, + "learning_rate": 0.00019281849506931132, + "loss": 1.138, + "step": 2744 + }, + { + "epoch": 0.48878205128205127, + "grad_norm": 0.5227449536323547, + "learning_rate": 0.000192813285431922, + "loss": 1.1831, + "step": 2745 + }, + { + "epoch": 0.48896011396011396, + "grad_norm": 0.48457080125808716, + "learning_rate": 0.00019280807397604915, + "loss": 1.2468, + "step": 2746 + }, + { + "epoch": 0.48913817663817666, + "grad_norm": 0.4596176743507385, + "learning_rate": 0.0001928028607017949, + "loss": 1.1098, + "step": 2747 + }, + { + "epoch": 0.4893162393162393, + "grad_norm": 0.5204966068267822, + "learning_rate": 0.00019279764560926142, + "loss": 1.1501, + "step": 2748 + }, + { + "epoch": 0.489494301994302, + "grad_norm": 0.5179490447044373, + "learning_rate": 0.0001927924286985508, + "loss": 1.2601, + "step": 2749 + }, + { + "epoch": 0.4896723646723647, + "grad_norm": 0.4563423693180084, + "learning_rate": 0.00019278720996976533, + "loss": 1.081, + "step": 2750 + }, + { + "epoch": 0.48985042735042733, + "grad_norm": 0.4906339943408966, + "learning_rate": 0.00019278198942300717, + "loss": 1.157, + "step": 2751 + }, + { + "epoch": 0.49002849002849, + "grad_norm": 0.42241403460502625, + "learning_rate": 0.00019277676705837873, + "loss": 1.0333, + "step": 2752 + }, + { + "epoch": 0.4902065527065527, + "grad_norm": 0.6310175657272339, + "learning_rate": 0.00019277154287598226, + "loss": 1.1225, + "step": 2753 + }, + { + "epoch": 0.49038461538461536, + "grad_norm": 0.5109034776687622, + "learning_rate": 0.0001927663168759201, + "loss": 1.1619, + "step": 2754 + }, + { + "epoch": 0.49056267806267806, + "grad_norm": 0.4809598922729492, + "learning_rate": 0.00019276108905829465, + "loss": 1.0423, + "step": 2755 + }, + { + "epoch": 0.49074074074074076, + "grad_norm": 0.557502806186676, + "learning_rate": 0.00019275585942320837, + "loss": 0.8783, + "step": 2756 + }, + { + "epoch": 0.4909188034188034, + "grad_norm": 0.5434393882751465, + "learning_rate": 0.0001927506279707637, + "loss": 1.1701, + "step": 2757 + }, + { + "epoch": 0.4910968660968661, + "grad_norm": 0.49278944730758667, + "learning_rate": 0.00019274539470106317, + "loss": 1.0447, + "step": 2758 + }, + { + "epoch": 0.4912749287749288, + "grad_norm": 0.5634264349937439, + "learning_rate": 0.00019274015961420927, + "loss": 1.0639, + "step": 2759 + }, + { + "epoch": 0.49145299145299143, + "grad_norm": 0.5632645487785339, + "learning_rate": 0.00019273492271030464, + "loss": 0.9223, + "step": 2760 + }, + { + "epoch": 0.4916310541310541, + "grad_norm": 0.5949172377586365, + "learning_rate": 0.00019272968398945177, + "loss": 0.894, + "step": 2761 + }, + { + "epoch": 0.4918091168091168, + "grad_norm": 0.5375374555587769, + "learning_rate": 0.00019272444345175342, + "loss": 1.0311, + "step": 2762 + }, + { + "epoch": 0.49198717948717946, + "grad_norm": 0.5211305022239685, + "learning_rate": 0.00019271920109731222, + "loss": 1.1531, + "step": 2763 + }, + { + "epoch": 0.49216524216524216, + "grad_norm": 0.44022253155708313, + "learning_rate": 0.00019271395692623084, + "loss": 0.9147, + "step": 2764 + }, + { + "epoch": 0.49234330484330485, + "grad_norm": 0.4682174623012543, + "learning_rate": 0.0001927087109386121, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 0.49252136752136755, + "grad_norm": 0.4971517324447632, + "learning_rate": 0.0001927034631345588, + "loss": 1.1017, + "step": 2766 + }, + { + "epoch": 0.4926994301994302, + "grad_norm": 0.5015294551849365, + "learning_rate": 0.00019269821351417364, + "loss": 1.1093, + "step": 2767 + }, + { + "epoch": 0.4928774928774929, + "grad_norm": 0.5512694716453552, + "learning_rate": 0.00019269296207755958, + "loss": 0.9657, + "step": 2768 + }, + { + "epoch": 0.4930555555555556, + "grad_norm": 0.4914868474006653, + "learning_rate": 0.00019268770882481948, + "loss": 1.0379, + "step": 2769 + }, + { + "epoch": 0.4932336182336182, + "grad_norm": 0.567337691783905, + "learning_rate": 0.00019268245375605626, + "loss": 1.004, + "step": 2770 + }, + { + "epoch": 0.4934116809116809, + "grad_norm": 0.518489420413971, + "learning_rate": 0.0001926771968713729, + "loss": 1.0734, + "step": 2771 + }, + { + "epoch": 0.4935897435897436, + "grad_norm": 0.567742109298706, + "learning_rate": 0.00019267193817087237, + "loss": 1.1276, + "step": 2772 + }, + { + "epoch": 0.49376780626780625, + "grad_norm": 0.5287964344024658, + "learning_rate": 0.00019266667765465773, + "loss": 1.1429, + "step": 2773 + }, + { + "epoch": 0.49394586894586895, + "grad_norm": 0.5302085876464844, + "learning_rate": 0.00019266141532283207, + "loss": 1.0934, + "step": 2774 + }, + { + "epoch": 0.49412393162393164, + "grad_norm": 0.5569987297058105, + "learning_rate": 0.00019265615117549842, + "loss": 1.1453, + "step": 2775 + }, + { + "epoch": 0.4943019943019943, + "grad_norm": 0.519695520401001, + "learning_rate": 0.00019265088521275997, + "loss": 1.1255, + "step": 2776 + }, + { + "epoch": 0.494480056980057, + "grad_norm": 0.5073211193084717, + "learning_rate": 0.0001926456174347199, + "loss": 1.0609, + "step": 2777 + }, + { + "epoch": 0.4946581196581197, + "grad_norm": 0.45028239488601685, + "learning_rate": 0.00019264034784148142, + "loss": 0.9098, + "step": 2778 + }, + { + "epoch": 0.4948361823361823, + "grad_norm": 0.6641215682029724, + "learning_rate": 0.00019263507643314776, + "loss": 0.8903, + "step": 2779 + }, + { + "epoch": 0.495014245014245, + "grad_norm": 0.5281413793563843, + "learning_rate": 0.00019262980320982224, + "loss": 1.2906, + "step": 2780 + }, + { + "epoch": 0.4951923076923077, + "grad_norm": 0.6256437301635742, + "learning_rate": 0.0001926245281716081, + "loss": 1.4142, + "step": 2781 + }, + { + "epoch": 0.49537037037037035, + "grad_norm": 0.5422517657279968, + "learning_rate": 0.00019261925131860877, + "loss": 1.1606, + "step": 2782 + }, + { + "epoch": 0.49554843304843305, + "grad_norm": 0.46938949823379517, + "learning_rate": 0.0001926139726509276, + "loss": 1.0333, + "step": 2783 + }, + { + "epoch": 0.49572649572649574, + "grad_norm": 0.5799683928489685, + "learning_rate": 0.000192608692168668, + "loss": 1.0333, + "step": 2784 + }, + { + "epoch": 0.4959045584045584, + "grad_norm": 0.5231602787971497, + "learning_rate": 0.0001926034098719335, + "loss": 1.1847, + "step": 2785 + }, + { + "epoch": 0.4960826210826211, + "grad_norm": 0.477845698595047, + "learning_rate": 0.00019259812576082752, + "loss": 1.0746, + "step": 2786 + }, + { + "epoch": 0.4962606837606838, + "grad_norm": 0.5490350723266602, + "learning_rate": 0.00019259283983545365, + "loss": 1.2462, + "step": 2787 + }, + { + "epoch": 0.4964387464387464, + "grad_norm": 0.5788847208023071, + "learning_rate": 0.0001925875520959154, + "loss": 1.3485, + "step": 2788 + }, + { + "epoch": 0.4966168091168091, + "grad_norm": 0.46184736490249634, + "learning_rate": 0.00019258226254231643, + "loss": 0.8673, + "step": 2789 + }, + { + "epoch": 0.4967948717948718, + "grad_norm": 0.4890633225440979, + "learning_rate": 0.0001925769711747603, + "loss": 0.9474, + "step": 2790 + }, + { + "epoch": 0.49697293447293445, + "grad_norm": 0.5719282627105713, + "learning_rate": 0.00019257167799335078, + "loss": 1.2532, + "step": 2791 + }, + { + "epoch": 0.49715099715099714, + "grad_norm": 0.5385584235191345, + "learning_rate": 0.0001925663829981915, + "loss": 1.1326, + "step": 2792 + }, + { + "epoch": 0.49732905982905984, + "grad_norm": 0.5339545011520386, + "learning_rate": 0.00019256108618938625, + "loss": 1.1362, + "step": 2793 + }, + { + "epoch": 0.49750712250712253, + "grad_norm": 0.5017803907394409, + "learning_rate": 0.00019255578756703878, + "loss": 1.0449, + "step": 2794 + }, + { + "epoch": 0.4976851851851852, + "grad_norm": 0.6004226803779602, + "learning_rate": 0.00019255048713125294, + "loss": 0.9346, + "step": 2795 + }, + { + "epoch": 0.49786324786324787, + "grad_norm": 0.44581490755081177, + "learning_rate": 0.00019254518488213255, + "loss": 1.038, + "step": 2796 + }, + { + "epoch": 0.49804131054131057, + "grad_norm": 0.5180951356887817, + "learning_rate": 0.00019253988081978151, + "loss": 1.0479, + "step": 2797 + }, + { + "epoch": 0.4982193732193732, + "grad_norm": 0.53944993019104, + "learning_rate": 0.00019253457494430376, + "loss": 1.2598, + "step": 2798 + }, + { + "epoch": 0.4983974358974359, + "grad_norm": 0.5633010268211365, + "learning_rate": 0.00019252926725580322, + "loss": 1.205, + "step": 2799 + }, + { + "epoch": 0.4985754985754986, + "grad_norm": 0.6653175950050354, + "learning_rate": 0.0001925239577543839, + "loss": 1.2383, + "step": 2800 + }, + { + "epoch": 0.49875356125356124, + "grad_norm": 0.5083333849906921, + "learning_rate": 0.00019251864644014984, + "loss": 1.0649, + "step": 2801 + }, + { + "epoch": 0.49893162393162394, + "grad_norm": 0.4842020571231842, + "learning_rate": 0.00019251333331320506, + "loss": 1.1991, + "step": 2802 + }, + { + "epoch": 0.49910968660968663, + "grad_norm": 0.47987112402915955, + "learning_rate": 0.00019250801837365373, + "loss": 1.1686, + "step": 2803 + }, + { + "epoch": 0.49928774928774927, + "grad_norm": 0.5316333770751953, + "learning_rate": 0.00019250270162159992, + "loss": 1.1759, + "step": 2804 + }, + { + "epoch": 0.49946581196581197, + "grad_norm": 0.5015079379081726, + "learning_rate": 0.00019249738305714787, + "loss": 0.9424, + "step": 2805 + }, + { + "epoch": 0.49964387464387466, + "grad_norm": 0.6488274931907654, + "learning_rate": 0.00019249206268040172, + "loss": 1.066, + "step": 2806 + }, + { + "epoch": 0.4998219373219373, + "grad_norm": 0.40364864468574524, + "learning_rate": 0.00019248674049146574, + "loss": 0.6998, + "step": 2807 + }, + { + "epoch": 0.5, + "grad_norm": 0.5535672903060913, + "learning_rate": 0.00019248141649044423, + "loss": 1.2207, + "step": 2808 + }, + { + "epoch": 0.5, + "eval_loss": 1.1072274446487427, + "eval_runtime": 28.6913, + "eval_samples_per_second": 36.283, + "eval_steps_per_second": 18.159, + "step": 2808 + }, + { + "epoch": 0.5001780626780626, + "grad_norm": 0.4834389090538025, + "learning_rate": 0.00019247609067744143, + "loss": 1.1686, + "step": 2809 + }, + { + "epoch": 0.5003561253561254, + "grad_norm": 0.5007249712944031, + "learning_rate": 0.00019247076305256176, + "loss": 1.1343, + "step": 2810 + }, + { + "epoch": 0.500534188034188, + "grad_norm": 0.4773348271846771, + "learning_rate": 0.00019246543361590957, + "loss": 0.9324, + "step": 2811 + }, + { + "epoch": 0.5007122507122507, + "grad_norm": 0.47324609756469727, + "learning_rate": 0.0001924601023675893, + "loss": 1.0223, + "step": 2812 + }, + { + "epoch": 0.5008903133903134, + "grad_norm": 0.5583845973014832, + "learning_rate": 0.00019245476930770537, + "loss": 1.1328, + "step": 2813 + }, + { + "epoch": 0.5010683760683761, + "grad_norm": 0.4814579486846924, + "learning_rate": 0.00019244943443636232, + "loss": 1.0528, + "step": 2814 + }, + { + "epoch": 0.5012464387464387, + "grad_norm": 0.4996104836463928, + "learning_rate": 0.00019244409775366465, + "loss": 1.2482, + "step": 2815 + }, + { + "epoch": 0.5014245014245015, + "grad_norm": 0.47870904207229614, + "learning_rate": 0.0001924387592597169, + "loss": 0.9452, + "step": 2816 + }, + { + "epoch": 0.5016025641025641, + "grad_norm": 0.5617441534996033, + "learning_rate": 0.0001924334189546237, + "loss": 1.378, + "step": 2817 + }, + { + "epoch": 0.5017806267806267, + "grad_norm": 0.4872083365917206, + "learning_rate": 0.00019242807683848967, + "loss": 1.1571, + "step": 2818 + }, + { + "epoch": 0.5019586894586895, + "grad_norm": 0.5147804021835327, + "learning_rate": 0.00019242273291141947, + "loss": 1.1086, + "step": 2819 + }, + { + "epoch": 0.5021367521367521, + "grad_norm": 0.4698995351791382, + "learning_rate": 0.00019241738717351784, + "loss": 1.1579, + "step": 2820 + }, + { + "epoch": 0.5023148148148148, + "grad_norm": 0.5158926844596863, + "learning_rate": 0.00019241203962488946, + "loss": 1.2763, + "step": 2821 + }, + { + "epoch": 0.5024928774928775, + "grad_norm": 0.5218976736068726, + "learning_rate": 0.00019240669026563914, + "loss": 1.0633, + "step": 2822 + }, + { + "epoch": 0.5026709401709402, + "grad_norm": 0.5511452555656433, + "learning_rate": 0.0001924013390958717, + "loss": 0.9939, + "step": 2823 + }, + { + "epoch": 0.5028490028490028, + "grad_norm": 0.5227555632591248, + "learning_rate": 0.00019239598611569191, + "loss": 1.2478, + "step": 2824 + }, + { + "epoch": 0.5030270655270656, + "grad_norm": 0.5444719791412354, + "learning_rate": 0.00019239063132520475, + "loss": 1.1574, + "step": 2825 + }, + { + "epoch": 0.5032051282051282, + "grad_norm": 0.4752781093120575, + "learning_rate": 0.0001923852747245151, + "loss": 0.9034, + "step": 2826 + }, + { + "epoch": 0.5033831908831908, + "grad_norm": 0.5286496877670288, + "learning_rate": 0.00019237991631372792, + "loss": 1.1391, + "step": 2827 + }, + { + "epoch": 0.5035612535612536, + "grad_norm": 0.5009933710098267, + "learning_rate": 0.00019237455609294815, + "loss": 1.2178, + "step": 2828 + }, + { + "epoch": 0.5037393162393162, + "grad_norm": 0.5012276768684387, + "learning_rate": 0.00019236919406228085, + "loss": 0.9877, + "step": 2829 + }, + { + "epoch": 0.5039173789173789, + "grad_norm": 0.576508104801178, + "learning_rate": 0.00019236383022183106, + "loss": 1.1299, + "step": 2830 + }, + { + "epoch": 0.5040954415954416, + "grad_norm": 0.4716590642929077, + "learning_rate": 0.0001923584645717039, + "loss": 1.0451, + "step": 2831 + }, + { + "epoch": 0.5042735042735043, + "grad_norm": 0.5817418098449707, + "learning_rate": 0.00019235309711200448, + "loss": 1.0911, + "step": 2832 + }, + { + "epoch": 0.5044515669515669, + "grad_norm": 0.5695745944976807, + "learning_rate": 0.000192347727842838, + "loss": 1.0229, + "step": 2833 + }, + { + "epoch": 0.5046296296296297, + "grad_norm": 0.49127066135406494, + "learning_rate": 0.00019234235676430958, + "loss": 1.1377, + "step": 2834 + }, + { + "epoch": 0.5048076923076923, + "grad_norm": 0.5426172614097595, + "learning_rate": 0.00019233698387652453, + "loss": 1.2427, + "step": 2835 + }, + { + "epoch": 0.5049857549857549, + "grad_norm": 0.5342385172843933, + "learning_rate": 0.0001923316091795881, + "loss": 1.1427, + "step": 2836 + }, + { + "epoch": 0.5051638176638177, + "grad_norm": 0.5480486750602722, + "learning_rate": 0.00019232623267360558, + "loss": 1.0647, + "step": 2837 + }, + { + "epoch": 0.5053418803418803, + "grad_norm": 0.4584530293941498, + "learning_rate": 0.00019232085435868235, + "loss": 1.0461, + "step": 2838 + }, + { + "epoch": 0.5055199430199431, + "grad_norm": 0.5992119908332825, + "learning_rate": 0.00019231547423492371, + "loss": 1.1456, + "step": 2839 + }, + { + "epoch": 0.5056980056980057, + "grad_norm": 0.514018177986145, + "learning_rate": 0.00019231009230243515, + "loss": 1.2559, + "step": 2840 + }, + { + "epoch": 0.5058760683760684, + "grad_norm": 0.5392283797264099, + "learning_rate": 0.0001923047085613221, + "loss": 1.044, + "step": 2841 + }, + { + "epoch": 0.5060541310541311, + "grad_norm": 0.4486566483974457, + "learning_rate": 0.00019229932301169, + "loss": 1.0679, + "step": 2842 + }, + { + "epoch": 0.5062321937321937, + "grad_norm": 0.4523460566997528, + "learning_rate": 0.00019229393565364442, + "loss": 1.1651, + "step": 2843 + }, + { + "epoch": 0.5064102564102564, + "grad_norm": 0.6032688021659851, + "learning_rate": 0.0001922885464872909, + "loss": 1.15, + "step": 2844 + }, + { + "epoch": 0.5065883190883191, + "grad_norm": 0.5883688926696777, + "learning_rate": 0.000192283155512735, + "loss": 1.2179, + "step": 2845 + }, + { + "epoch": 0.5067663817663818, + "grad_norm": 0.5534378886222839, + "learning_rate": 0.00019227776273008238, + "loss": 1.0387, + "step": 2846 + }, + { + "epoch": 0.5069444444444444, + "grad_norm": 0.5899033546447754, + "learning_rate": 0.00019227236813943872, + "loss": 1.0812, + "step": 2847 + }, + { + "epoch": 0.5071225071225072, + "grad_norm": 0.5718855261802673, + "learning_rate": 0.00019226697174090965, + "loss": 1.1375, + "step": 2848 + }, + { + "epoch": 0.5073005698005698, + "grad_norm": 0.5080967545509338, + "learning_rate": 0.00019226157353460094, + "loss": 1.1421, + "step": 2849 + }, + { + "epoch": 0.5074786324786325, + "grad_norm": 0.5253677368164062, + "learning_rate": 0.0001922561735206184, + "loss": 1.0166, + "step": 2850 + }, + { + "epoch": 0.5076566951566952, + "grad_norm": 0.47797444462776184, + "learning_rate": 0.00019225077169906772, + "loss": 1.0504, + "step": 2851 + }, + { + "epoch": 0.5078347578347578, + "grad_norm": 0.4911690652370453, + "learning_rate": 0.0001922453680700548, + "loss": 1.0629, + "step": 2852 + }, + { + "epoch": 0.5080128205128205, + "grad_norm": 0.49678200483322144, + "learning_rate": 0.00019223996263368557, + "loss": 1.1672, + "step": 2853 + }, + { + "epoch": 0.5081908831908832, + "grad_norm": 0.5451810359954834, + "learning_rate": 0.00019223455539006586, + "loss": 1.3031, + "step": 2854 + }, + { + "epoch": 0.5083689458689459, + "grad_norm": 0.5708984136581421, + "learning_rate": 0.00019222914633930166, + "loss": 1.0986, + "step": 2855 + }, + { + "epoch": 0.5085470085470085, + "grad_norm": 0.47232356667518616, + "learning_rate": 0.00019222373548149888, + "loss": 1.0449, + "step": 2856 + }, + { + "epoch": 0.5087250712250713, + "grad_norm": 0.6027610898017883, + "learning_rate": 0.0001922183228167636, + "loss": 0.862, + "step": 2857 + }, + { + "epoch": 0.5089031339031339, + "grad_norm": 0.5211802124977112, + "learning_rate": 0.00019221290834520188, + "loss": 1.1048, + "step": 2858 + }, + { + "epoch": 0.5090811965811965, + "grad_norm": 0.45101237297058105, + "learning_rate": 0.00019220749206691972, + "loss": 1.0046, + "step": 2859 + }, + { + "epoch": 0.5092592592592593, + "grad_norm": 0.5526158213615417, + "learning_rate": 0.00019220207398202335, + "loss": 1.2275, + "step": 2860 + }, + { + "epoch": 0.5094373219373219, + "grad_norm": 0.48322010040283203, + "learning_rate": 0.00019219665409061885, + "loss": 0.9974, + "step": 2861 + }, + { + "epoch": 0.5096153846153846, + "grad_norm": 0.4775219261646271, + "learning_rate": 0.00019219123239281244, + "loss": 1.1852, + "step": 2862 + }, + { + "epoch": 0.5097934472934473, + "grad_norm": 0.46184200048446655, + "learning_rate": 0.00019218580888871034, + "loss": 0.9393, + "step": 2863 + }, + { + "epoch": 0.50997150997151, + "grad_norm": 0.47495174407958984, + "learning_rate": 0.00019218038357841883, + "loss": 0.9631, + "step": 2864 + }, + { + "epoch": 0.5101495726495726, + "grad_norm": 0.48600029945373535, + "learning_rate": 0.00019217495646204418, + "loss": 1.0498, + "step": 2865 + }, + { + "epoch": 0.5103276353276354, + "grad_norm": 0.5801547169685364, + "learning_rate": 0.00019216952753969274, + "loss": 1.2181, + "step": 2866 + }, + { + "epoch": 0.510505698005698, + "grad_norm": 0.5082106590270996, + "learning_rate": 0.00019216409681147085, + "loss": 1.2009, + "step": 2867 + }, + { + "epoch": 0.5106837606837606, + "grad_norm": 0.4184330701828003, + "learning_rate": 0.00019215866427748493, + "loss": 0.8462, + "step": 2868 + }, + { + "epoch": 0.5108618233618234, + "grad_norm": 0.518099844455719, + "learning_rate": 0.00019215322993784147, + "loss": 1.2091, + "step": 2869 + }, + { + "epoch": 0.511039886039886, + "grad_norm": 0.569464921951294, + "learning_rate": 0.0001921477937926469, + "loss": 1.0264, + "step": 2870 + }, + { + "epoch": 0.5112179487179487, + "grad_norm": 0.526767909526825, + "learning_rate": 0.00019214235584200768, + "loss": 1.1192, + "step": 2871 + }, + { + "epoch": 0.5113960113960114, + "grad_norm": 0.6511057019233704, + "learning_rate": 0.00019213691608603047, + "loss": 1.3193, + "step": 2872 + }, + { + "epoch": 0.5115740740740741, + "grad_norm": 0.48536401987075806, + "learning_rate": 0.00019213147452482173, + "loss": 1.1671, + "step": 2873 + }, + { + "epoch": 0.5117521367521367, + "grad_norm": 0.7972469329833984, + "learning_rate": 0.00019212603115848818, + "loss": 1.1393, + "step": 2874 + }, + { + "epoch": 0.5119301994301995, + "grad_norm": 0.5543264746665955, + "learning_rate": 0.00019212058598713642, + "loss": 1.1436, + "step": 2875 + }, + { + "epoch": 0.5121082621082621, + "grad_norm": 0.49688720703125, + "learning_rate": 0.0001921151390108731, + "loss": 1.0897, + "step": 2876 + }, + { + "epoch": 0.5122863247863247, + "grad_norm": 0.4928736090660095, + "learning_rate": 0.000192109690229805, + "loss": 1.2426, + "step": 2877 + }, + { + "epoch": 0.5124643874643875, + "grad_norm": 0.4917896091938019, + "learning_rate": 0.0001921042396440389, + "loss": 1.0047, + "step": 2878 + }, + { + "epoch": 0.5126424501424501, + "grad_norm": 0.5485204458236694, + "learning_rate": 0.00019209878725368152, + "loss": 1.2615, + "step": 2879 + }, + { + "epoch": 0.5128205128205128, + "grad_norm": 0.5229470133781433, + "learning_rate": 0.0001920933330588397, + "loss": 1.3249, + "step": 2880 + }, + { + "epoch": 0.5129985754985755, + "grad_norm": 0.4783077538013458, + "learning_rate": 0.00019208787705962037, + "loss": 1.2004, + "step": 2881 + }, + { + "epoch": 0.5131766381766382, + "grad_norm": 0.5106910467147827, + "learning_rate": 0.00019208241925613035, + "loss": 1.1745, + "step": 2882 + }, + { + "epoch": 0.5133547008547008, + "grad_norm": 0.5308730006217957, + "learning_rate": 0.00019207695964847666, + "loss": 0.9706, + "step": 2883 + }, + { + "epoch": 0.5135327635327636, + "grad_norm": 0.5489775538444519, + "learning_rate": 0.00019207149823676617, + "loss": 1.0073, + "step": 2884 + }, + { + "epoch": 0.5137108262108262, + "grad_norm": 0.4992835521697998, + "learning_rate": 0.00019206603502110596, + "loss": 1.1053, + "step": 2885 + }, + { + "epoch": 0.5138888888888888, + "grad_norm": 0.5304922461509705, + "learning_rate": 0.00019206057000160302, + "loss": 1.0565, + "step": 2886 + }, + { + "epoch": 0.5140669515669516, + "grad_norm": 0.46411609649658203, + "learning_rate": 0.00019205510317836448, + "loss": 0.9202, + "step": 2887 + }, + { + "epoch": 0.5142450142450142, + "grad_norm": 0.5236835479736328, + "learning_rate": 0.0001920496345514974, + "loss": 0.9075, + "step": 2888 + }, + { + "epoch": 0.5144230769230769, + "grad_norm": 0.4416964054107666, + "learning_rate": 0.00019204416412110895, + "loss": 0.9225, + "step": 2889 + }, + { + "epoch": 0.5146011396011396, + "grad_norm": 0.5470940470695496, + "learning_rate": 0.00019203869188730633, + "loss": 1.2195, + "step": 2890 + }, + { + "epoch": 0.5147792022792023, + "grad_norm": 0.5380414128303528, + "learning_rate": 0.0001920332178501967, + "loss": 1.0731, + "step": 2891 + }, + { + "epoch": 0.5149572649572649, + "grad_norm": 0.4405716359615326, + "learning_rate": 0.00019202774200988737, + "loss": 0.8739, + "step": 2892 + }, + { + "epoch": 0.5151353276353277, + "grad_norm": 0.5222984552383423, + "learning_rate": 0.0001920222643664856, + "loss": 1.1806, + "step": 2893 + }, + { + "epoch": 0.5153133903133903, + "grad_norm": 0.48545539379119873, + "learning_rate": 0.0001920167849200987, + "loss": 0.9939, + "step": 2894 + }, + { + "epoch": 0.5154914529914529, + "grad_norm": 0.45078009366989136, + "learning_rate": 0.0001920113036708341, + "loss": 1.0085, + "step": 2895 + }, + { + "epoch": 0.5156695156695157, + "grad_norm": 0.5029830932617188, + "learning_rate": 0.00019200582061879913, + "loss": 1.1095, + "step": 2896 + }, + { + "epoch": 0.5158475783475783, + "grad_norm": 0.5316143035888672, + "learning_rate": 0.00019200033576410118, + "loss": 0.9883, + "step": 2897 + }, + { + "epoch": 0.5160256410256411, + "grad_norm": 0.5282100439071655, + "learning_rate": 0.0001919948491068478, + "loss": 1.1441, + "step": 2898 + }, + { + "epoch": 0.5162037037037037, + "grad_norm": 0.5145367980003357, + "learning_rate": 0.00019198936064714647, + "loss": 1.1999, + "step": 2899 + }, + { + "epoch": 0.5163817663817664, + "grad_norm": 0.5385651588439941, + "learning_rate": 0.00019198387038510468, + "loss": 1.1831, + "step": 2900 + }, + { + "epoch": 0.5165598290598291, + "grad_norm": 0.4971916377544403, + "learning_rate": 0.00019197837832083002, + "loss": 1.2518, + "step": 2901 + }, + { + "epoch": 0.5167378917378918, + "grad_norm": 0.5253807306289673, + "learning_rate": 0.00019197288445443016, + "loss": 1.0788, + "step": 2902 + }, + { + "epoch": 0.5169159544159544, + "grad_norm": 0.49724945425987244, + "learning_rate": 0.00019196738878601263, + "loss": 1.0985, + "step": 2903 + }, + { + "epoch": 0.5170940170940171, + "grad_norm": 0.5327325463294983, + "learning_rate": 0.0001919618913156852, + "loss": 1.2862, + "step": 2904 + }, + { + "epoch": 0.5172720797720798, + "grad_norm": 0.639999270439148, + "learning_rate": 0.00019195639204355554, + "loss": 1.2052, + "step": 2905 + }, + { + "epoch": 0.5174501424501424, + "grad_norm": 0.4630785584449768, + "learning_rate": 0.0001919508909697314, + "loss": 1.1157, + "step": 2906 + }, + { + "epoch": 0.5176282051282052, + "grad_norm": 0.513949990272522, + "learning_rate": 0.00019194538809432055, + "loss": 1.0047, + "step": 2907 + }, + { + "epoch": 0.5178062678062678, + "grad_norm": 0.488034725189209, + "learning_rate": 0.0001919398834174308, + "loss": 0.9008, + "step": 2908 + }, + { + "epoch": 0.5179843304843305, + "grad_norm": 0.4892788529396057, + "learning_rate": 0.00019193437693917006, + "loss": 1.1024, + "step": 2909 + }, + { + "epoch": 0.5181623931623932, + "grad_norm": 0.5503842830657959, + "learning_rate": 0.00019192886865964618, + "loss": 1.2283, + "step": 2910 + }, + { + "epoch": 0.5183404558404558, + "grad_norm": 0.48885393142700195, + "learning_rate": 0.00019192335857896707, + "loss": 0.9522, + "step": 2911 + }, + { + "epoch": 0.5185185185185185, + "grad_norm": 0.5479527115821838, + "learning_rate": 0.00019191784669724072, + "loss": 1.1616, + "step": 2912 + }, + { + "epoch": 0.5186965811965812, + "grad_norm": 0.42701148986816406, + "learning_rate": 0.00019191233301457506, + "loss": 0.8434, + "step": 2913 + }, + { + "epoch": 0.5188746438746439, + "grad_norm": 0.4273422658443451, + "learning_rate": 0.00019190681753107822, + "loss": 0.8316, + "step": 2914 + }, + { + "epoch": 0.5190527065527065, + "grad_norm": 0.5047736763954163, + "learning_rate": 0.00019190130024685818, + "loss": 1.171, + "step": 2915 + }, + { + "epoch": 0.5192307692307693, + "grad_norm": 0.5221177935600281, + "learning_rate": 0.00019189578116202307, + "loss": 1.0256, + "step": 2916 + }, + { + "epoch": 0.5194088319088319, + "grad_norm": 0.4782322943210602, + "learning_rate": 0.00019189026027668105, + "loss": 0.8598, + "step": 2917 + }, + { + "epoch": 0.5195868945868946, + "grad_norm": 0.5627185702323914, + "learning_rate": 0.00019188473759094022, + "loss": 1.1825, + "step": 2918 + }, + { + "epoch": 0.5197649572649573, + "grad_norm": 0.5036423206329346, + "learning_rate": 0.00019187921310490888, + "loss": 1.0881, + "step": 2919 + }, + { + "epoch": 0.51994301994302, + "grad_norm": 0.4271143972873688, + "learning_rate": 0.0001918736868186952, + "loss": 0.9265, + "step": 2920 + }, + { + "epoch": 0.5201210826210826, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00019186815873240747, + "loss": 1.196, + "step": 2921 + }, + { + "epoch": 0.5202991452991453, + "grad_norm": 0.5494198203086853, + "learning_rate": 0.00019186262884615402, + "loss": 1.1207, + "step": 2922 + }, + { + "epoch": 0.520477207977208, + "grad_norm": 0.5305119752883911, + "learning_rate": 0.0001918570971600432, + "loss": 1.0393, + "step": 2923 + }, + { + "epoch": 0.5206552706552706, + "grad_norm": 0.46713170409202576, + "learning_rate": 0.00019185156367418333, + "loss": 0.9583, + "step": 2924 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 0.597776472568512, + "learning_rate": 0.00019184602838868292, + "loss": 1.2978, + "step": 2925 + }, + { + "epoch": 0.521011396011396, + "grad_norm": 0.520976722240448, + "learning_rate": 0.00019184049130365036, + "loss": 1.0515, + "step": 2926 + }, + { + "epoch": 0.5211894586894587, + "grad_norm": 0.5266290307044983, + "learning_rate": 0.00019183495241919415, + "loss": 1.0437, + "step": 2927 + }, + { + "epoch": 0.5213675213675214, + "grad_norm": 0.50911545753479, + "learning_rate": 0.00019182941173542285, + "loss": 0.9977, + "step": 2928 + }, + { + "epoch": 0.521545584045584, + "grad_norm": 0.4924670457839966, + "learning_rate": 0.00019182386925244496, + "loss": 0.9309, + "step": 2929 + }, + { + "epoch": 0.5217236467236467, + "grad_norm": 0.4979301393032074, + "learning_rate": 0.00019181832497036912, + "loss": 0.87, + "step": 2930 + }, + { + "epoch": 0.5219017094017094, + "grad_norm": 0.6307916045188904, + "learning_rate": 0.0001918127788893039, + "loss": 1.2159, + "step": 2931 + }, + { + "epoch": 0.5220797720797721, + "grad_norm": 0.4915660619735718, + "learning_rate": 0.00019180723100935802, + "loss": 1.0828, + "step": 2932 + }, + { + "epoch": 0.5222578347578347, + "grad_norm": 0.4312742352485657, + "learning_rate": 0.00019180168133064017, + "loss": 1.0496, + "step": 2933 + }, + { + "epoch": 0.5224358974358975, + "grad_norm": 0.6006124019622803, + "learning_rate": 0.00019179612985325908, + "loss": 1.0751, + "step": 2934 + }, + { + "epoch": 0.5226139601139601, + "grad_norm": 0.5332220196723938, + "learning_rate": 0.0001917905765773235, + "loss": 1.2601, + "step": 2935 + }, + { + "epoch": 0.5227920227920227, + "grad_norm": 0.4877954423427582, + "learning_rate": 0.00019178502150294223, + "loss": 1.2279, + "step": 2936 + }, + { + "epoch": 0.5229700854700855, + "grad_norm": 0.5975968837738037, + "learning_rate": 0.00019177946463022418, + "loss": 1.3371, + "step": 2937 + }, + { + "epoch": 0.5231481481481481, + "grad_norm": 0.5363923907279968, + "learning_rate": 0.00019177390595927815, + "loss": 1.0705, + "step": 2938 + }, + { + "epoch": 0.5233262108262108, + "grad_norm": 0.4314909875392914, + "learning_rate": 0.0001917683454902131, + "loss": 0.9172, + "step": 2939 + }, + { + "epoch": 0.5235042735042735, + "grad_norm": 0.46187883615493774, + "learning_rate": 0.0001917627832231379, + "loss": 1.1201, + "step": 2940 + }, + { + "epoch": 0.5236823361823362, + "grad_norm": 0.4648260772228241, + "learning_rate": 0.00019175721915816162, + "loss": 1.1307, + "step": 2941 + }, + { + "epoch": 0.5238603988603988, + "grad_norm": 0.4427165687084198, + "learning_rate": 0.00019175165329539325, + "loss": 0.9459, + "step": 2942 + }, + { + "epoch": 0.5240384615384616, + "grad_norm": 0.4645056128501892, + "learning_rate": 0.0001917460856349418, + "loss": 0.9176, + "step": 2943 + }, + { + "epoch": 0.5242165242165242, + "grad_norm": 0.4939568042755127, + "learning_rate": 0.0001917405161769164, + "loss": 1.1056, + "step": 2944 + }, + { + "epoch": 0.5243945868945868, + "grad_norm": 0.6057310104370117, + "learning_rate": 0.00019173494492142617, + "loss": 1.2714, + "step": 2945 + }, + { + "epoch": 0.5245726495726496, + "grad_norm": 0.5038546323776245, + "learning_rate": 0.00019172937186858025, + "loss": 0.911, + "step": 2946 + }, + { + "epoch": 0.5247507122507122, + "grad_norm": 0.5521321296691895, + "learning_rate": 0.00019172379701848784, + "loss": 1.0781, + "step": 2947 + }, + { + "epoch": 0.5249287749287749, + "grad_norm": 0.516979455947876, + "learning_rate": 0.00019171822037125817, + "loss": 1.1051, + "step": 2948 + }, + { + "epoch": 0.5251068376068376, + "grad_norm": 0.5443150997161865, + "learning_rate": 0.0001917126419270005, + "loss": 1.0802, + "step": 2949 + }, + { + "epoch": 0.5252849002849003, + "grad_norm": 0.5373311042785645, + "learning_rate": 0.00019170706168582412, + "loss": 0.9313, + "step": 2950 + }, + { + "epoch": 0.5254629629629629, + "grad_norm": 0.7511917948722839, + "learning_rate": 0.0001917014796478384, + "loss": 1.1958, + "step": 2951 + }, + { + "epoch": 0.5256410256410257, + "grad_norm": 0.49893468618392944, + "learning_rate": 0.00019169589581315263, + "loss": 0.9387, + "step": 2952 + }, + { + "epoch": 0.5258190883190883, + "grad_norm": 0.48010289669036865, + "learning_rate": 0.00019169031018187628, + "loss": 1.2459, + "step": 2953 + }, + { + "epoch": 0.5259971509971509, + "grad_norm": 0.48768678307533264, + "learning_rate": 0.0001916847227541188, + "loss": 1.0127, + "step": 2954 + }, + { + "epoch": 0.5261752136752137, + "grad_norm": 0.5973068475723267, + "learning_rate": 0.00019167913352998963, + "loss": 1.1685, + "step": 2955 + }, + { + "epoch": 0.5263532763532763, + "grad_norm": 0.5567806959152222, + "learning_rate": 0.00019167354250959826, + "loss": 1.142, + "step": 2956 + }, + { + "epoch": 0.5265313390313391, + "grad_norm": 0.47819700837135315, + "learning_rate": 0.00019166794969305428, + "loss": 0.712, + "step": 2957 + }, + { + "epoch": 0.5267094017094017, + "grad_norm": 0.5191744565963745, + "learning_rate": 0.00019166235508046725, + "loss": 1.2208, + "step": 2958 + }, + { + "epoch": 0.5268874643874644, + "grad_norm": 0.4987856149673462, + "learning_rate": 0.00019165675867194675, + "loss": 1.0466, + "step": 2959 + }, + { + "epoch": 0.5270655270655271, + "grad_norm": 0.5017665028572083, + "learning_rate": 0.0001916511604676025, + "loss": 1.1236, + "step": 2960 + }, + { + "epoch": 0.5272435897435898, + "grad_norm": 0.5115348696708679, + "learning_rate": 0.00019164556046754415, + "loss": 1.1497, + "step": 2961 + }, + { + "epoch": 0.5274216524216524, + "grad_norm": 0.4934345781803131, + "learning_rate": 0.0001916399586718814, + "loss": 1.0183, + "step": 2962 + }, + { + "epoch": 0.5275997150997151, + "grad_norm": 0.5033719539642334, + "learning_rate": 0.00019163435508072404, + "loss": 1.0256, + "step": 2963 + }, + { + "epoch": 0.5277777777777778, + "grad_norm": 0.5325372219085693, + "learning_rate": 0.00019162874969418184, + "loss": 1.1384, + "step": 2964 + }, + { + "epoch": 0.5279558404558404, + "grad_norm": 0.4901772141456604, + "learning_rate": 0.00019162314251236465, + "loss": 1.0831, + "step": 2965 + }, + { + "epoch": 0.5281339031339032, + "grad_norm": 0.4743805229663849, + "learning_rate": 0.0001916175335353823, + "loss": 1.1894, + "step": 2966 + }, + { + "epoch": 0.5283119658119658, + "grad_norm": 0.5439450740814209, + "learning_rate": 0.00019161192276334466, + "loss": 1.2066, + "step": 2967 + }, + { + "epoch": 0.5284900284900285, + "grad_norm": 0.5123090744018555, + "learning_rate": 0.00019160631019636174, + "loss": 1.1829, + "step": 2968 + }, + { + "epoch": 0.5286680911680912, + "grad_norm": 0.5995343923568726, + "learning_rate": 0.00019160069583454346, + "loss": 1.4872, + "step": 2969 + }, + { + "epoch": 0.5288461538461539, + "grad_norm": 0.4596657156944275, + "learning_rate": 0.00019159507967799985, + "loss": 0.8948, + "step": 2970 + }, + { + "epoch": 0.5290242165242165, + "grad_norm": 0.5533682107925415, + "learning_rate": 0.0001915894617268409, + "loss": 1.1779, + "step": 2971 + }, + { + "epoch": 0.5292022792022792, + "grad_norm": 0.3860718309879303, + "learning_rate": 0.00019158384198117673, + "loss": 0.6424, + "step": 2972 + }, + { + "epoch": 0.5293803418803419, + "grad_norm": 0.47424063086509705, + "learning_rate": 0.0001915782204411174, + "loss": 1.1592, + "step": 2973 + }, + { + "epoch": 0.5295584045584045, + "grad_norm": 0.5050228834152222, + "learning_rate": 0.00019157259710677309, + "loss": 1.1971, + "step": 2974 + }, + { + "epoch": 0.5297364672364673, + "grad_norm": 0.6080113649368286, + "learning_rate": 0.00019156697197825396, + "loss": 1.1511, + "step": 2975 + }, + { + "epoch": 0.5299145299145299, + "grad_norm": 0.4805932641029358, + "learning_rate": 0.00019156134505567024, + "loss": 1.1033, + "step": 2976 + }, + { + "epoch": 0.5300925925925926, + "grad_norm": 0.4835345447063446, + "learning_rate": 0.00019155571633913215, + "loss": 1.1832, + "step": 2977 + }, + { + "epoch": 0.5302706552706553, + "grad_norm": 0.5183725953102112, + "learning_rate": 0.00019155008582875, + "loss": 0.9221, + "step": 2978 + }, + { + "epoch": 0.530448717948718, + "grad_norm": 0.48015761375427246, + "learning_rate": 0.00019154445352463412, + "loss": 1.045, + "step": 2979 + }, + { + "epoch": 0.5306267806267806, + "grad_norm": 0.4670043885707855, + "learning_rate": 0.0001915388194268948, + "loss": 0.9025, + "step": 2980 + }, + { + "epoch": 0.5308048433048433, + "grad_norm": 0.5048824548721313, + "learning_rate": 0.0001915331835356425, + "loss": 1.0681, + "step": 2981 + }, + { + "epoch": 0.530982905982906, + "grad_norm": 0.4785633981227875, + "learning_rate": 0.00019152754585098758, + "loss": 1.0097, + "step": 2982 + }, + { + "epoch": 0.5311609686609686, + "grad_norm": 0.4829573333263397, + "learning_rate": 0.00019152190637304056, + "loss": 1.0856, + "step": 2983 + }, + { + "epoch": 0.5313390313390314, + "grad_norm": 0.5425563454627991, + "learning_rate": 0.00019151626510191189, + "loss": 1.2313, + "step": 2984 + }, + { + "epoch": 0.531517094017094, + "grad_norm": 0.5532251596450806, + "learning_rate": 0.0001915106220377121, + "loss": 1.0328, + "step": 2985 + }, + { + "epoch": 0.5316951566951567, + "grad_norm": 0.47016972303390503, + "learning_rate": 0.0001915049771805518, + "loss": 1.2003, + "step": 2986 + }, + { + "epoch": 0.5318732193732194, + "grad_norm": 0.5241743326187134, + "learning_rate": 0.00019149933053054153, + "loss": 1.046, + "step": 2987 + }, + { + "epoch": 0.532051282051282, + "grad_norm": 0.5043526887893677, + "learning_rate": 0.00019149368208779197, + "loss": 1.0022, + "step": 2988 + }, + { + "epoch": 0.5322293447293447, + "grad_norm": 0.5563312768936157, + "learning_rate": 0.00019148803185241374, + "loss": 1.1017, + "step": 2989 + }, + { + "epoch": 0.5324074074074074, + "grad_norm": 0.5414231419563293, + "learning_rate": 0.00019148237982451763, + "loss": 0.9649, + "step": 2990 + }, + { + "epoch": 0.5325854700854701, + "grad_norm": 0.5452231764793396, + "learning_rate": 0.0001914767260042143, + "loss": 1.2281, + "step": 2991 + }, + { + "epoch": 0.5327635327635327, + "grad_norm": 0.5500698685646057, + "learning_rate": 0.00019147107039161454, + "loss": 1.2865, + "step": 2992 + }, + { + "epoch": 0.5329415954415955, + "grad_norm": 0.49747416377067566, + "learning_rate": 0.00019146541298682918, + "loss": 1.1296, + "step": 2993 + }, + { + "epoch": 0.5331196581196581, + "grad_norm": 0.5684167742729187, + "learning_rate": 0.00019145975378996903, + "loss": 1.0685, + "step": 2994 + }, + { + "epoch": 0.5332977207977208, + "grad_norm": 0.5411235690116882, + "learning_rate": 0.00019145409280114502, + "loss": 1.1372, + "step": 2995 + }, + { + "epoch": 0.5334757834757835, + "grad_norm": 0.5006675720214844, + "learning_rate": 0.00019144843002046806, + "loss": 1.0688, + "step": 2996 + }, + { + "epoch": 0.5336538461538461, + "grad_norm": 0.4591315686702728, + "learning_rate": 0.00019144276544804908, + "loss": 1.1071, + "step": 2997 + }, + { + "epoch": 0.5338319088319088, + "grad_norm": 0.5615306496620178, + "learning_rate": 0.000191437099083999, + "loss": 1.1033, + "step": 2998 + }, + { + "epoch": 0.5340099715099715, + "grad_norm": 0.4986817240715027, + "learning_rate": 0.00019143143092842897, + "loss": 1.176, + "step": 2999 + }, + { + "epoch": 0.5341880341880342, + "grad_norm": 0.5017120242118835, + "learning_rate": 0.00019142576098144995, + "loss": 1.0174, + "step": 3000 + }, + { + "epoch": 0.5343660968660968, + "grad_norm": 0.508298397064209, + "learning_rate": 0.0001914200892431731, + "loss": 1.164, + "step": 3001 + }, + { + "epoch": 0.5345441595441596, + "grad_norm": 0.48068809509277344, + "learning_rate": 0.0001914144157137095, + "loss": 0.7959, + "step": 3002 + }, + { + "epoch": 0.5347222222222222, + "grad_norm": 0.6347028017044067, + "learning_rate": 0.0001914087403931703, + "loss": 1.1727, + "step": 3003 + }, + { + "epoch": 0.5349002849002849, + "grad_norm": 0.5558401942253113, + "learning_rate": 0.00019140306328166676, + "loss": 1.2282, + "step": 3004 + }, + { + "epoch": 0.5350783475783476, + "grad_norm": 0.5093596577644348, + "learning_rate": 0.00019139738437931004, + "loss": 1.3258, + "step": 3005 + }, + { + "epoch": 0.5352564102564102, + "grad_norm": 0.4653106927871704, + "learning_rate": 0.0001913917036862114, + "loss": 1.1062, + "step": 3006 + }, + { + "epoch": 0.5354344729344729, + "grad_norm": 0.48085781931877136, + "learning_rate": 0.00019138602120248222, + "loss": 0.9019, + "step": 3007 + }, + { + "epoch": 0.5356125356125356, + "grad_norm": 0.5174745321273804, + "learning_rate": 0.0001913803369282338, + "loss": 1.044, + "step": 3008 + }, + { + "epoch": 0.5357905982905983, + "grad_norm": 0.5359669327735901, + "learning_rate": 0.00019137465086357746, + "loss": 1.0723, + "step": 3009 + }, + { + "epoch": 0.5359686609686609, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.00019136896300862467, + "loss": 1.2192, + "step": 3010 + }, + { + "epoch": 0.5361467236467237, + "grad_norm": 0.4905693829059601, + "learning_rate": 0.00019136327336348688, + "loss": 1.2372, + "step": 3011 + }, + { + "epoch": 0.5363247863247863, + "grad_norm": 0.5741264820098877, + "learning_rate": 0.0001913575819282755, + "loss": 1.1703, + "step": 3012 + }, + { + "epoch": 0.5365028490028491, + "grad_norm": 0.577033281326294, + "learning_rate": 0.0001913518887031021, + "loss": 1.1555, + "step": 3013 + }, + { + "epoch": 0.5366809116809117, + "grad_norm": 0.46795153617858887, + "learning_rate": 0.00019134619368807822, + "loss": 0.8583, + "step": 3014 + }, + { + "epoch": 0.5368589743589743, + "grad_norm": 0.5973345637321472, + "learning_rate": 0.0001913404968833154, + "loss": 1.1509, + "step": 3015 + }, + { + "epoch": 0.5370370370370371, + "grad_norm": 0.62020343542099, + "learning_rate": 0.00019133479828892531, + "loss": 1.0781, + "step": 3016 + }, + { + "epoch": 0.5372150997150997, + "grad_norm": 0.5342286229133606, + "learning_rate": 0.00019132909790501958, + "loss": 1.1556, + "step": 3017 + }, + { + "epoch": 0.5373931623931624, + "grad_norm": 0.49612846970558167, + "learning_rate": 0.0001913233957317099, + "loss": 0.9027, + "step": 3018 + }, + { + "epoch": 0.5375712250712251, + "grad_norm": 0.5403908491134644, + "learning_rate": 0.00019131769176910796, + "loss": 1.1125, + "step": 3019 + }, + { + "epoch": 0.5377492877492878, + "grad_norm": 0.4952050447463989, + "learning_rate": 0.0001913119860173256, + "loss": 1.2329, + "step": 3020 + }, + { + "epoch": 0.5379273504273504, + "grad_norm": 0.5877819657325745, + "learning_rate": 0.0001913062784764745, + "loss": 1.2855, + "step": 3021 + }, + { + "epoch": 0.5381054131054132, + "grad_norm": 0.49312907457351685, + "learning_rate": 0.00019130056914666655, + "loss": 1.0212, + "step": 3022 + }, + { + "epoch": 0.5382834757834758, + "grad_norm": 0.45544490218162537, + "learning_rate": 0.00019129485802801366, + "loss": 0.9748, + "step": 3023 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.5535242557525635, + "learning_rate": 0.00019128914512062762, + "loss": 1.2134, + "step": 3024 + }, + { + "epoch": 0.5386396011396012, + "grad_norm": 0.45369696617126465, + "learning_rate": 0.00019128343042462044, + "loss": 0.9964, + "step": 3025 + }, + { + "epoch": 0.5388176638176638, + "grad_norm": 0.6240725517272949, + "learning_rate": 0.00019127771394010406, + "loss": 1.425, + "step": 3026 + }, + { + "epoch": 0.5389957264957265, + "grad_norm": 0.4859573245048523, + "learning_rate": 0.0001912719956671905, + "loss": 1.087, + "step": 3027 + }, + { + "epoch": 0.5391737891737892, + "grad_norm": 0.47529762983322144, + "learning_rate": 0.0001912662756059918, + "loss": 0.9517, + "step": 3028 + }, + { + "epoch": 0.5393518518518519, + "grad_norm": 0.5317288637161255, + "learning_rate": 0.00019126055375661997, + "loss": 1.0945, + "step": 3029 + }, + { + "epoch": 0.5395299145299145, + "grad_norm": 0.55974280834198, + "learning_rate": 0.00019125483011918722, + "loss": 1.0794, + "step": 3030 + }, + { + "epoch": 0.5397079772079773, + "grad_norm": 0.48579123616218567, + "learning_rate": 0.0001912491046938056, + "loss": 1.1421, + "step": 3031 + }, + { + "epoch": 0.5398860398860399, + "grad_norm": 0.4917181134223938, + "learning_rate": 0.00019124337748058733, + "loss": 0.9708, + "step": 3032 + }, + { + "epoch": 0.5400641025641025, + "grad_norm": 0.525291383266449, + "learning_rate": 0.00019123764847964466, + "loss": 1.064, + "step": 3033 + }, + { + "epoch": 0.5402421652421653, + "grad_norm": 0.5733301639556885, + "learning_rate": 0.00019123191769108977, + "loss": 1.2142, + "step": 3034 + }, + { + "epoch": 0.5404202279202279, + "grad_norm": 0.5400987863540649, + "learning_rate": 0.00019122618511503494, + "loss": 1.1309, + "step": 3035 + }, + { + "epoch": 0.5405982905982906, + "grad_norm": 0.6261051893234253, + "learning_rate": 0.00019122045075159257, + "loss": 1.2112, + "step": 3036 + }, + { + "epoch": 0.5407763532763533, + "grad_norm": 0.5483576059341431, + "learning_rate": 0.0001912147146008749, + "loss": 1.2705, + "step": 3037 + }, + { + "epoch": 0.540954415954416, + "grad_norm": 0.5442137122154236, + "learning_rate": 0.00019120897666299443, + "loss": 1.2512, + "step": 3038 + }, + { + "epoch": 0.5411324786324786, + "grad_norm": 0.5680811405181885, + "learning_rate": 0.00019120323693806355, + "loss": 1.392, + "step": 3039 + }, + { + "epoch": 0.5413105413105413, + "grad_norm": 0.5237287878990173, + "learning_rate": 0.00019119749542619466, + "loss": 1.1599, + "step": 3040 + }, + { + "epoch": 0.541488603988604, + "grad_norm": 0.48119300603866577, + "learning_rate": 0.00019119175212750032, + "loss": 1.0976, + "step": 3041 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 0.507033109664917, + "learning_rate": 0.00019118600704209302, + "loss": 1.0181, + "step": 3042 + }, + { + "epoch": 0.5418447293447294, + "grad_norm": 0.484672874212265, + "learning_rate": 0.00019118026017008531, + "loss": 1.1636, + "step": 3043 + }, + { + "epoch": 0.542022792022792, + "grad_norm": 0.4923502206802368, + "learning_rate": 0.00019117451151158985, + "loss": 1.0388, + "step": 3044 + }, + { + "epoch": 0.5422008547008547, + "grad_norm": 0.4882057309150696, + "learning_rate": 0.00019116876106671922, + "loss": 1.131, + "step": 3045 + }, + { + "epoch": 0.5423789173789174, + "grad_norm": 0.6068355441093445, + "learning_rate": 0.0001911630088355861, + "loss": 1.3218, + "step": 3046 + }, + { + "epoch": 0.54255698005698, + "grad_norm": 0.5012881755828857, + "learning_rate": 0.0001911572548183032, + "loss": 1.0514, + "step": 3047 + }, + { + "epoch": 0.5427350427350427, + "grad_norm": 0.49849793314933777, + "learning_rate": 0.00019115149901498328, + "loss": 1.0003, + "step": 3048 + }, + { + "epoch": 0.5429131054131054, + "grad_norm": 0.4934251010417938, + "learning_rate": 0.00019114574142573904, + "loss": 1.0319, + "step": 3049 + }, + { + "epoch": 0.5430911680911681, + "grad_norm": 0.4947762191295624, + "learning_rate": 0.00019113998205068334, + "loss": 1.0906, + "step": 3050 + }, + { + "epoch": 0.5432692307692307, + "grad_norm": 0.5449416041374207, + "learning_rate": 0.00019113422088992907, + "loss": 0.9093, + "step": 3051 + }, + { + "epoch": 0.5434472934472935, + "grad_norm": 0.49395284056663513, + "learning_rate": 0.00019112845794358902, + "loss": 1.0071, + "step": 3052 + }, + { + "epoch": 0.5436253561253561, + "grad_norm": 0.5478728413581848, + "learning_rate": 0.00019112269321177613, + "loss": 1.2124, + "step": 3053 + }, + { + "epoch": 0.5438034188034188, + "grad_norm": 0.6205173134803772, + "learning_rate": 0.0001911169266946034, + "loss": 1.021, + "step": 3054 + }, + { + "epoch": 0.5439814814814815, + "grad_norm": 0.4777783751487732, + "learning_rate": 0.00019111115839218372, + "loss": 0.9192, + "step": 3055 + }, + { + "epoch": 0.5441595441595442, + "grad_norm": 0.5541689991950989, + "learning_rate": 0.00019110538830463018, + "loss": 1.1248, + "step": 3056 + }, + { + "epoch": 0.5443376068376068, + "grad_norm": 0.4750942289829254, + "learning_rate": 0.0001910996164320558, + "loss": 1.3147, + "step": 3057 + }, + { + "epoch": 0.5445156695156695, + "grad_norm": 0.6283948421478271, + "learning_rate": 0.0001910938427745737, + "loss": 1.0919, + "step": 3058 + }, + { + "epoch": 0.5446937321937322, + "grad_norm": 0.552725076675415, + "learning_rate": 0.00019108806733229698, + "loss": 1.3807, + "step": 3059 + }, + { + "epoch": 0.5448717948717948, + "grad_norm": 0.4832848310470581, + "learning_rate": 0.0001910822901053388, + "loss": 1.0705, + "step": 3060 + }, + { + "epoch": 0.5450498575498576, + "grad_norm": 0.6468375325202942, + "learning_rate": 0.00019107651109381233, + "loss": 1.0766, + "step": 3061 + }, + { + "epoch": 0.5452279202279202, + "grad_norm": 0.5464920401573181, + "learning_rate": 0.00019107073029783083, + "loss": 1.0453, + "step": 3062 + }, + { + "epoch": 0.5454059829059829, + "grad_norm": 0.5321210026741028, + "learning_rate": 0.0001910649477175076, + "loss": 1.2326, + "step": 3063 + }, + { + "epoch": 0.5455840455840456, + "grad_norm": 0.5572962164878845, + "learning_rate": 0.00019105916335295582, + "loss": 1.0673, + "step": 3064 + }, + { + "epoch": 0.5457621082621082, + "grad_norm": 0.5239177942276001, + "learning_rate": 0.00019105337720428894, + "loss": 1.04, + "step": 3065 + }, + { + "epoch": 0.5459401709401709, + "grad_norm": 0.5633319616317749, + "learning_rate": 0.00019104758927162023, + "loss": 0.9606, + "step": 3066 + }, + { + "epoch": 0.5461182336182336, + "grad_norm": 0.5317914485931396, + "learning_rate": 0.0001910417995550632, + "loss": 1.0651, + "step": 3067 + }, + { + "epoch": 0.5462962962962963, + "grad_norm": 0.5126453638076782, + "learning_rate": 0.00019103600805473118, + "loss": 1.0316, + "step": 3068 + }, + { + "epoch": 0.5464743589743589, + "grad_norm": 0.5262107253074646, + "learning_rate": 0.00019103021477073773, + "loss": 1.0752, + "step": 3069 + }, + { + "epoch": 0.5466524216524217, + "grad_norm": 0.5384877324104309, + "learning_rate": 0.0001910244197031963, + "loss": 1.1731, + "step": 3070 + }, + { + "epoch": 0.5468304843304843, + "grad_norm": 0.5126553773880005, + "learning_rate": 0.00019101862285222048, + "loss": 1.2229, + "step": 3071 + }, + { + "epoch": 0.5470085470085471, + "grad_norm": 0.4841194450855255, + "learning_rate": 0.0001910128242179238, + "loss": 0.9955, + "step": 3072 + }, + { + "epoch": 0.5471866096866097, + "grad_norm": 0.526546061038971, + "learning_rate": 0.00019100702380041987, + "loss": 1.2436, + "step": 3073 + }, + { + "epoch": 0.5473646723646723, + "grad_norm": 0.5085833072662354, + "learning_rate": 0.0001910012215998224, + "loss": 1.011, + "step": 3074 + }, + { + "epoch": 0.5475427350427351, + "grad_norm": 0.5149994492530823, + "learning_rate": 0.000190995417616245, + "loss": 0.8632, + "step": 3075 + }, + { + "epoch": 0.5477207977207977, + "grad_norm": 0.48079630732536316, + "learning_rate": 0.00019098961184980145, + "loss": 1.1115, + "step": 3076 + }, + { + "epoch": 0.5478988603988604, + "grad_norm": 0.5769477486610413, + "learning_rate": 0.00019098380430060546, + "loss": 0.9544, + "step": 3077 + }, + { + "epoch": 0.5480769230769231, + "grad_norm": 0.5260093808174133, + "learning_rate": 0.0001909779949687708, + "loss": 1.2354, + "step": 3078 + }, + { + "epoch": 0.5482549857549858, + "grad_norm": 0.5518734455108643, + "learning_rate": 0.00019097218385441135, + "loss": 1.1944, + "step": 3079 + }, + { + "epoch": 0.5484330484330484, + "grad_norm": 0.5436808466911316, + "learning_rate": 0.00019096637095764095, + "loss": 1.0717, + "step": 3080 + }, + { + "epoch": 0.5486111111111112, + "grad_norm": 0.4749584197998047, + "learning_rate": 0.00019096055627857344, + "loss": 1.0417, + "step": 3081 + }, + { + "epoch": 0.5487891737891738, + "grad_norm": 0.5485591292381287, + "learning_rate": 0.0001909547398173228, + "loss": 1.2515, + "step": 3082 + }, + { + "epoch": 0.5489672364672364, + "grad_norm": 0.5751016736030579, + "learning_rate": 0.00019094892157400296, + "loss": 1.2112, + "step": 3083 + }, + { + "epoch": 0.5491452991452992, + "grad_norm": 0.5404475331306458, + "learning_rate": 0.00019094310154872795, + "loss": 0.4334, + "step": 3084 + }, + { + "epoch": 0.5493233618233618, + "grad_norm": 0.5198020935058594, + "learning_rate": 0.00019093727974161178, + "loss": 0.9759, + "step": 3085 + }, + { + "epoch": 0.5495014245014245, + "grad_norm": 0.4893439710140228, + "learning_rate": 0.0001909314561527685, + "loss": 1.1287, + "step": 3086 + }, + { + "epoch": 0.5496794871794872, + "grad_norm": 0.5675956606864929, + "learning_rate": 0.00019092563078231228, + "loss": 1.234, + "step": 3087 + }, + { + "epoch": 0.5498575498575499, + "grad_norm": 0.5539132356643677, + "learning_rate": 0.00019091980363035714, + "loss": 1.2378, + "step": 3088 + }, + { + "epoch": 0.5500356125356125, + "grad_norm": 0.5194353461265564, + "learning_rate": 0.00019091397469701735, + "loss": 1.1338, + "step": 3089 + }, + { + "epoch": 0.5502136752136753, + "grad_norm": 0.5143756866455078, + "learning_rate": 0.0001909081439824071, + "loss": 0.9118, + "step": 3090 + }, + { + "epoch": 0.5503917378917379, + "grad_norm": 0.5624327659606934, + "learning_rate": 0.0001909023114866406, + "loss": 1.035, + "step": 3091 + }, + { + "epoch": 0.5505698005698005, + "grad_norm": 0.5285067558288574, + "learning_rate": 0.0001908964772098321, + "loss": 1.0451, + "step": 3092 + }, + { + "epoch": 0.5507478632478633, + "grad_norm": 0.5730587244033813, + "learning_rate": 0.000190890641152096, + "loss": 1.0672, + "step": 3093 + }, + { + "epoch": 0.5509259259259259, + "grad_norm": 0.5822951197624207, + "learning_rate": 0.0001908848033135466, + "loss": 1.1791, + "step": 3094 + }, + { + "epoch": 0.5511039886039886, + "grad_norm": 0.596161961555481, + "learning_rate": 0.00019087896369429826, + "loss": 1.0954, + "step": 3095 + }, + { + "epoch": 0.5512820512820513, + "grad_norm": 0.5138190984725952, + "learning_rate": 0.00019087312229446542, + "loss": 0.896, + "step": 3096 + }, + { + "epoch": 0.551460113960114, + "grad_norm": 0.5061872601509094, + "learning_rate": 0.0001908672791141625, + "loss": 1.1017, + "step": 3097 + }, + { + "epoch": 0.5516381766381766, + "grad_norm": 0.5189547538757324, + "learning_rate": 0.00019086143415350404, + "loss": 1.2906, + "step": 3098 + }, + { + "epoch": 0.5518162393162394, + "grad_norm": 0.5640039443969727, + "learning_rate": 0.00019085558741260448, + "loss": 1.1001, + "step": 3099 + }, + { + "epoch": 0.551994301994302, + "grad_norm": 0.453867107629776, + "learning_rate": 0.00019084973889157844, + "loss": 0.9731, + "step": 3100 + }, + { + "epoch": 0.5521723646723646, + "grad_norm": 0.5431303977966309, + "learning_rate": 0.0001908438885905405, + "loss": 1.3511, + "step": 3101 + }, + { + "epoch": 0.5523504273504274, + "grad_norm": 0.47693368792533875, + "learning_rate": 0.00019083803650960527, + "loss": 1.0426, + "step": 3102 + }, + { + "epoch": 0.55252849002849, + "grad_norm": 0.4663422703742981, + "learning_rate": 0.00019083218264888743, + "loss": 1.05, + "step": 3103 + }, + { + "epoch": 0.5527065527065527, + "grad_norm": 0.561354398727417, + "learning_rate": 0.00019082632700850164, + "loss": 0.9608, + "step": 3104 + }, + { + "epoch": 0.5528846153846154, + "grad_norm": 0.4981916844844818, + "learning_rate": 0.00019082046958856266, + "loss": 1.1935, + "step": 3105 + }, + { + "epoch": 0.5530626780626781, + "grad_norm": 0.5301326513290405, + "learning_rate": 0.0001908146103891852, + "loss": 1.0646, + "step": 3106 + }, + { + "epoch": 0.5532407407407407, + "grad_norm": 0.5023610591888428, + "learning_rate": 0.00019080874941048416, + "loss": 1.127, + "step": 3107 + }, + { + "epoch": 0.5534188034188035, + "grad_norm": 0.5172514319419861, + "learning_rate": 0.00019080288665257426, + "loss": 1.0435, + "step": 3108 + }, + { + "epoch": 0.5535968660968661, + "grad_norm": 0.6340598464012146, + "learning_rate": 0.00019079702211557048, + "loss": 1.3528, + "step": 3109 + }, + { + "epoch": 0.5537749287749287, + "grad_norm": 0.46882256865501404, + "learning_rate": 0.0001907911557995876, + "loss": 1.1361, + "step": 3110 + }, + { + "epoch": 0.5539529914529915, + "grad_norm": 0.6401382088661194, + "learning_rate": 0.00019078528770474068, + "loss": 1.2415, + "step": 3111 + }, + { + "epoch": 0.5541310541310541, + "grad_norm": 0.5141328573226929, + "learning_rate": 0.00019077941783114463, + "loss": 1.0505, + "step": 3112 + }, + { + "epoch": 0.5543091168091168, + "grad_norm": 0.522318959236145, + "learning_rate": 0.00019077354617891444, + "loss": 1.0964, + "step": 3113 + }, + { + "epoch": 0.5544871794871795, + "grad_norm": 0.539551854133606, + "learning_rate": 0.00019076767274816517, + "loss": 1.0735, + "step": 3114 + }, + { + "epoch": 0.5546652421652422, + "grad_norm": 0.495320200920105, + "learning_rate": 0.00019076179753901195, + "loss": 0.9754, + "step": 3115 + }, + { + "epoch": 0.5548433048433048, + "grad_norm": 0.5499199628829956, + "learning_rate": 0.00019075592055156984, + "loss": 1.0043, + "step": 3116 + }, + { + "epoch": 0.5550213675213675, + "grad_norm": 0.5352509617805481, + "learning_rate": 0.00019075004178595396, + "loss": 1.1701, + "step": 3117 + }, + { + "epoch": 0.5551994301994302, + "grad_norm": 0.5392300486564636, + "learning_rate": 0.00019074416124227953, + "loss": 1.1612, + "step": 3118 + }, + { + "epoch": 0.5553774928774928, + "grad_norm": 0.5195050835609436, + "learning_rate": 0.0001907382789206618, + "loss": 1.0934, + "step": 3119 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 0.5276884436607361, + "learning_rate": 0.000190732394821216, + "loss": 0.9011, + "step": 3120 + }, + { + "epoch": 0.5557336182336182, + "grad_norm": 0.6115903258323669, + "learning_rate": 0.00019072650894405734, + "loss": 1.3065, + "step": 3121 + }, + { + "epoch": 0.5559116809116809, + "grad_norm": 0.5752483010292053, + "learning_rate": 0.00019072062128930127, + "loss": 1.0063, + "step": 3122 + }, + { + "epoch": 0.5560897435897436, + "grad_norm": 0.5508273243904114, + "learning_rate": 0.00019071473185706302, + "loss": 1.2598, + "step": 3123 + }, + { + "epoch": 0.5562678062678063, + "grad_norm": 0.49712198972702026, + "learning_rate": 0.00019070884064745808, + "loss": 0.924, + "step": 3124 + }, + { + "epoch": 0.5564458689458689, + "grad_norm": 0.572849452495575, + "learning_rate": 0.00019070294766060185, + "loss": 0.9683, + "step": 3125 + }, + { + "epoch": 0.5566239316239316, + "grad_norm": 0.4807920753955841, + "learning_rate": 0.00019069705289660976, + "loss": 1.0998, + "step": 3126 + }, + { + "epoch": 0.5568019943019943, + "grad_norm": 0.5543031096458435, + "learning_rate": 0.0001906911563555973, + "loss": 1.0878, + "step": 3127 + }, + { + "epoch": 0.5569800569800569, + "grad_norm": 0.5710418820381165, + "learning_rate": 0.00019068525803768007, + "loss": 1.0381, + "step": 3128 + }, + { + "epoch": 0.5571581196581197, + "grad_norm": 0.5169163346290588, + "learning_rate": 0.00019067935794297357, + "loss": 1.1149, + "step": 3129 + }, + { + "epoch": 0.5573361823361823, + "grad_norm": 0.6474376916885376, + "learning_rate": 0.00019067345607159345, + "loss": 0.9828, + "step": 3130 + }, + { + "epoch": 0.5575142450142451, + "grad_norm": 0.5029847621917725, + "learning_rate": 0.0001906675524236553, + "loss": 0.797, + "step": 3131 + }, + { + "epoch": 0.5576923076923077, + "grad_norm": 0.5681431293487549, + "learning_rate": 0.00019066164699927478, + "loss": 1.1565, + "step": 3132 + }, + { + "epoch": 0.5578703703703703, + "grad_norm": 0.5654549598693848, + "learning_rate": 0.00019065573979856764, + "loss": 1.2488, + "step": 3133 + }, + { + "epoch": 0.5580484330484331, + "grad_norm": 0.47653043270111084, + "learning_rate": 0.0001906498308216496, + "loss": 1.0428, + "step": 3134 + }, + { + "epoch": 0.5582264957264957, + "grad_norm": 0.5068467259407043, + "learning_rate": 0.00019064392006863643, + "loss": 0.9659, + "step": 3135 + }, + { + "epoch": 0.5584045584045584, + "grad_norm": 0.7076661586761475, + "learning_rate": 0.00019063800753964393, + "loss": 1.1289, + "step": 3136 + }, + { + "epoch": 0.5585826210826211, + "grad_norm": 0.551456868648529, + "learning_rate": 0.000190632093234788, + "loss": 1.1925, + "step": 3137 + }, + { + "epoch": 0.5587606837606838, + "grad_norm": 0.518276035785675, + "learning_rate": 0.00019062617715418442, + "loss": 0.8681, + "step": 3138 + }, + { + "epoch": 0.5589387464387464, + "grad_norm": 0.5272278785705566, + "learning_rate": 0.0001906202592979492, + "loss": 1.0865, + "step": 3139 + }, + { + "epoch": 0.5591168091168092, + "grad_norm": 0.5344942212104797, + "learning_rate": 0.00019061433966619822, + "loss": 1.1647, + "step": 3140 + }, + { + "epoch": 0.5592948717948718, + "grad_norm": 0.5833460092544556, + "learning_rate": 0.00019060841825904753, + "loss": 1.3403, + "step": 3141 + }, + { + "epoch": 0.5594729344729344, + "grad_norm": 0.5707054734230042, + "learning_rate": 0.00019060249507661306, + "loss": 1.1236, + "step": 3142 + }, + { + "epoch": 0.5596509971509972, + "grad_norm": 0.5446065664291382, + "learning_rate": 0.00019059657011901094, + "loss": 1.017, + "step": 3143 + }, + { + "epoch": 0.5598290598290598, + "grad_norm": 0.5285109281539917, + "learning_rate": 0.0001905906433863572, + "loss": 1.3186, + "step": 3144 + }, + { + "epoch": 0.5600071225071225, + "grad_norm": 0.5308659672737122, + "learning_rate": 0.00019058471487876802, + "loss": 0.8464, + "step": 3145 + }, + { + "epoch": 0.5601851851851852, + "grad_norm": 0.5218054056167603, + "learning_rate": 0.00019057878459635948, + "loss": 1.0219, + "step": 3146 + }, + { + "epoch": 0.5603632478632479, + "grad_norm": 0.45067787170410156, + "learning_rate": 0.00019057285253924785, + "loss": 1.0364, + "step": 3147 + }, + { + "epoch": 0.5605413105413105, + "grad_norm": 0.4856041669845581, + "learning_rate": 0.0001905669187075493, + "loss": 1.1928, + "step": 3148 + }, + { + "epoch": 0.5607193732193733, + "grad_norm": 0.506912112236023, + "learning_rate": 0.00019056098310138016, + "loss": 1.119, + "step": 3149 + }, + { + "epoch": 0.5608974358974359, + "grad_norm": 0.49049463868141174, + "learning_rate": 0.00019055504572085662, + "loss": 1.2165, + "step": 3150 + }, + { + "epoch": 0.5610754985754985, + "grad_norm": 0.5250293612480164, + "learning_rate": 0.0001905491065660951, + "loss": 1.1427, + "step": 3151 + }, + { + "epoch": 0.5612535612535613, + "grad_norm": 0.43438446521759033, + "learning_rate": 0.00019054316563721195, + "loss": 0.884, + "step": 3152 + }, + { + "epoch": 0.5614316239316239, + "grad_norm": 0.5386807918548584, + "learning_rate": 0.00019053722293432354, + "loss": 1.1494, + "step": 3153 + }, + { + "epoch": 0.5616096866096866, + "grad_norm": 0.5403809547424316, + "learning_rate": 0.00019053127845754632, + "loss": 1.1743, + "step": 3154 + }, + { + "epoch": 0.5617877492877493, + "grad_norm": 0.4759823977947235, + "learning_rate": 0.00019052533220699678, + "loss": 1.0716, + "step": 3155 + }, + { + "epoch": 0.561965811965812, + "grad_norm": 0.45332327485084534, + "learning_rate": 0.0001905193841827914, + "loss": 0.8405, + "step": 3156 + }, + { + "epoch": 0.5621438746438746, + "grad_norm": 0.5617053508758545, + "learning_rate": 0.00019051343438504671, + "loss": 1.0422, + "step": 3157 + }, + { + "epoch": 0.5623219373219374, + "grad_norm": 0.5088049173355103, + "learning_rate": 0.00019050748281387931, + "loss": 1.0067, + "step": 3158 + }, + { + "epoch": 0.5625, + "grad_norm": 0.5174484848976135, + "learning_rate": 0.00019050152946940578, + "loss": 1.0623, + "step": 3159 + }, + { + "epoch": 0.5626780626780626, + "grad_norm": 0.6093568801879883, + "learning_rate": 0.0001904955743517428, + "loss": 1.24, + "step": 3160 + }, + { + "epoch": 0.5628561253561254, + "grad_norm": 0.49063584208488464, + "learning_rate": 0.00019048961746100703, + "loss": 0.8563, + "step": 3161 + }, + { + "epoch": 0.563034188034188, + "grad_norm": 0.583940863609314, + "learning_rate": 0.00019048365879731517, + "loss": 1.0695, + "step": 3162 + }, + { + "epoch": 0.5632122507122507, + "grad_norm": 0.4943268597126007, + "learning_rate": 0.000190477698360784, + "loss": 0.8606, + "step": 3163 + }, + { + "epoch": 0.5633903133903134, + "grad_norm": 0.5050932168960571, + "learning_rate": 0.00019047173615153028, + "loss": 1.1591, + "step": 3164 + }, + { + "epoch": 0.5635683760683761, + "grad_norm": 0.5445677638053894, + "learning_rate": 0.0001904657721696708, + "loss": 1.262, + "step": 3165 + }, + { + "epoch": 0.5637464387464387, + "grad_norm": 0.5445297360420227, + "learning_rate": 0.00019045980641532246, + "loss": 1.223, + "step": 3166 + }, + { + "epoch": 0.5639245014245015, + "grad_norm": 0.5098413228988647, + "learning_rate": 0.00019045383888860213, + "loss": 1.0829, + "step": 3167 + }, + { + "epoch": 0.5641025641025641, + "grad_norm": 0.484998881816864, + "learning_rate": 0.0001904478695896267, + "loss": 1.0711, + "step": 3168 + }, + { + "epoch": 0.5642806267806267, + "grad_norm": 0.5515334010124207, + "learning_rate": 0.0001904418985185132, + "loss": 1.1583, + "step": 3169 + }, + { + "epoch": 0.5644586894586895, + "grad_norm": 0.545460045337677, + "learning_rate": 0.00019043592567537853, + "loss": 1.2321, + "step": 3170 + }, + { + "epoch": 0.5646367521367521, + "grad_norm": 0.5463964343070984, + "learning_rate": 0.0001904299510603398, + "loss": 1.1019, + "step": 3171 + }, + { + "epoch": 0.5648148148148148, + "grad_norm": 0.5619220733642578, + "learning_rate": 0.000190423974673514, + "loss": 1.1001, + "step": 3172 + }, + { + "epoch": 0.5649928774928775, + "grad_norm": 0.4448916018009186, + "learning_rate": 0.00019041799651501825, + "loss": 1.057, + "step": 3173 + }, + { + "epoch": 0.5651709401709402, + "grad_norm": 0.6073006987571716, + "learning_rate": 0.00019041201658496975, + "loss": 1.0306, + "step": 3174 + }, + { + "epoch": 0.5653490028490028, + "grad_norm": 0.5342072248458862, + "learning_rate": 0.0001904060348834855, + "loss": 0.9231, + "step": 3175 + }, + { + "epoch": 0.5655270655270656, + "grad_norm": 0.4505697786808014, + "learning_rate": 0.0001904000514106829, + "loss": 1.1134, + "step": 3176 + }, + { + "epoch": 0.5657051282051282, + "grad_norm": 0.5627852082252502, + "learning_rate": 0.00019039406616667902, + "loss": 1.2138, + "step": 3177 + }, + { + "epoch": 0.5658831908831908, + "grad_norm": 0.499734103679657, + "learning_rate": 0.0001903880791515912, + "loss": 1.1074, + "step": 3178 + }, + { + "epoch": 0.5660612535612536, + "grad_norm": 0.4768189489841461, + "learning_rate": 0.00019038209036553676, + "loss": 0.9442, + "step": 3179 + }, + { + "epoch": 0.5662393162393162, + "grad_norm": 0.5265373587608337, + "learning_rate": 0.00019037609980863298, + "loss": 1.0907, + "step": 3180 + }, + { + "epoch": 0.5664173789173789, + "grad_norm": 0.5506128072738647, + "learning_rate": 0.00019037010748099728, + "loss": 1.2541, + "step": 3181 + }, + { + "epoch": 0.5665954415954416, + "grad_norm": 0.44860872626304626, + "learning_rate": 0.00019036411338274703, + "loss": 0.893, + "step": 3182 + }, + { + "epoch": 0.5667735042735043, + "grad_norm": 0.4901522994041443, + "learning_rate": 0.00019035811751399973, + "loss": 1.0469, + "step": 3183 + }, + { + "epoch": 0.5669515669515669, + "grad_norm": 0.500868022441864, + "learning_rate": 0.0001903521198748728, + "loss": 1.0527, + "step": 3184 + }, + { + "epoch": 0.5671296296296297, + "grad_norm": 0.5508102774620056, + "learning_rate": 0.00019034612046548376, + "loss": 1.283, + "step": 3185 + }, + { + "epoch": 0.5673076923076923, + "grad_norm": 0.5079495906829834, + "learning_rate": 0.0001903401192859502, + "loss": 1.0808, + "step": 3186 + }, + { + "epoch": 0.5674857549857549, + "grad_norm": 0.5758788585662842, + "learning_rate": 0.00019033411633638964, + "loss": 1.1301, + "step": 3187 + }, + { + "epoch": 0.5676638176638177, + "grad_norm": 0.46557924151420593, + "learning_rate": 0.00019032811161691972, + "loss": 1.0205, + "step": 3188 + }, + { + "epoch": 0.5678418803418803, + "grad_norm": 0.5665056109428406, + "learning_rate": 0.0001903221051276581, + "loss": 1.1926, + "step": 3189 + }, + { + "epoch": 0.5680199430199431, + "grad_norm": 0.5948992967605591, + "learning_rate": 0.00019031609686872246, + "loss": 1.2724, + "step": 3190 + }, + { + "epoch": 0.5681980056980057, + "grad_norm": 0.6189367771148682, + "learning_rate": 0.00019031008684023055, + "loss": 1.2762, + "step": 3191 + }, + { + "epoch": 0.5683760683760684, + "grad_norm": 0.49511992931365967, + "learning_rate": 0.00019030407504230006, + "loss": 1.0117, + "step": 3192 + }, + { + "epoch": 0.5685541310541311, + "grad_norm": 0.5358837842941284, + "learning_rate": 0.00019029806147504878, + "loss": 0.944, + "step": 3193 + }, + { + "epoch": 0.5687321937321937, + "grad_norm": 0.458636999130249, + "learning_rate": 0.00019029204613859463, + "loss": 0.8174, + "step": 3194 + }, + { + "epoch": 0.5689102564102564, + "grad_norm": 0.5168304443359375, + "learning_rate": 0.00019028602903305535, + "loss": 1.1533, + "step": 3195 + }, + { + "epoch": 0.5690883190883191, + "grad_norm": 0.5334134697914124, + "learning_rate": 0.00019028001015854892, + "loss": 1.1868, + "step": 3196 + }, + { + "epoch": 0.5692663817663818, + "grad_norm": 0.5649123191833496, + "learning_rate": 0.0001902739895151932, + "loss": 0.9876, + "step": 3197 + }, + { + "epoch": 0.5694444444444444, + "grad_norm": 0.5647651553153992, + "learning_rate": 0.0001902679671031062, + "loss": 1.0805, + "step": 3198 + }, + { + "epoch": 0.5696225071225072, + "grad_norm": 0.5251876711845398, + "learning_rate": 0.00019026194292240587, + "loss": 1.2335, + "step": 3199 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 0.5268014669418335, + "learning_rate": 0.0001902559169732103, + "loss": 1.19, + "step": 3200 + }, + { + "epoch": 0.5699786324786325, + "grad_norm": 0.5301041007041931, + "learning_rate": 0.00019024988925563752, + "loss": 1.1173, + "step": 3201 + }, + { + "epoch": 0.5701566951566952, + "grad_norm": 0.4531562030315399, + "learning_rate": 0.00019024385976980566, + "loss": 0.7576, + "step": 3202 + }, + { + "epoch": 0.5703347578347578, + "grad_norm": 0.5779716372489929, + "learning_rate": 0.00019023782851583282, + "loss": 1.1719, + "step": 3203 + }, + { + "epoch": 0.5705128205128205, + "grad_norm": 0.4886093735694885, + "learning_rate": 0.00019023179549383716, + "loss": 1.085, + "step": 3204 + }, + { + "epoch": 0.5706908831908832, + "grad_norm": 0.510117769241333, + "learning_rate": 0.0001902257607039369, + "loss": 0.8931, + "step": 3205 + }, + { + "epoch": 0.5708689458689459, + "grad_norm": 0.5195479393005371, + "learning_rate": 0.00019021972414625036, + "loss": 0.9922, + "step": 3206 + }, + { + "epoch": 0.5710470085470085, + "grad_norm": 0.5791407227516174, + "learning_rate": 0.00019021368582089568, + "loss": 1.112, + "step": 3207 + }, + { + "epoch": 0.5712250712250713, + "grad_norm": 0.5056005716323853, + "learning_rate": 0.00019020764572799122, + "loss": 0.8474, + "step": 3208 + }, + { + "epoch": 0.5714031339031339, + "grad_norm": 0.5060068964958191, + "learning_rate": 0.00019020160386765537, + "loss": 1.071, + "step": 3209 + }, + { + "epoch": 0.5715811965811965, + "grad_norm": 0.5396568775177002, + "learning_rate": 0.00019019556024000648, + "loss": 1.0436, + "step": 3210 + }, + { + "epoch": 0.5717592592592593, + "grad_norm": 0.6552190780639648, + "learning_rate": 0.0001901895148451629, + "loss": 0.9869, + "step": 3211 + }, + { + "epoch": 0.5719373219373219, + "grad_norm": 0.5177004337310791, + "learning_rate": 0.00019018346768324314, + "loss": 1.0193, + "step": 3212 + }, + { + "epoch": 0.5721153846153846, + "grad_norm": 0.5192117094993591, + "learning_rate": 0.0001901774187543657, + "loss": 1.1263, + "step": 3213 + }, + { + "epoch": 0.5722934472934473, + "grad_norm": 0.4857729971408844, + "learning_rate": 0.00019017136805864906, + "loss": 0.9808, + "step": 3214 + }, + { + "epoch": 0.57247150997151, + "grad_norm": 0.5800918936729431, + "learning_rate": 0.00019016531559621177, + "loss": 1.2334, + "step": 3215 + }, + { + "epoch": 0.5726495726495726, + "grad_norm": 0.4812086522579193, + "learning_rate": 0.00019015926136717242, + "loss": 1.2409, + "step": 3216 + }, + { + "epoch": 0.5728276353276354, + "grad_norm": 0.5128398537635803, + "learning_rate": 0.00019015320537164963, + "loss": 0.9036, + "step": 3217 + }, + { + "epoch": 0.573005698005698, + "grad_norm": 0.4761141538619995, + "learning_rate": 0.00019014714760976205, + "loss": 1.1058, + "step": 3218 + }, + { + "epoch": 0.5731837606837606, + "grad_norm": 0.5850459933280945, + "learning_rate": 0.0001901410880816284, + "loss": 1.1011, + "step": 3219 + }, + { + "epoch": 0.5733618233618234, + "grad_norm": 0.5648714303970337, + "learning_rate": 0.00019013502678736738, + "loss": 1.0479, + "step": 3220 + }, + { + "epoch": 0.573539886039886, + "grad_norm": 0.5835902094841003, + "learning_rate": 0.00019012896372709774, + "loss": 1.0555, + "step": 3221 + }, + { + "epoch": 0.5737179487179487, + "grad_norm": 0.5155113935470581, + "learning_rate": 0.00019012289890093828, + "loss": 0.9488, + "step": 3222 + }, + { + "epoch": 0.5738960113960114, + "grad_norm": 0.5064889788627625, + "learning_rate": 0.00019011683230900784, + "loss": 0.9144, + "step": 3223 + }, + { + "epoch": 0.5740740740740741, + "grad_norm": 0.53825843334198, + "learning_rate": 0.00019011076395142527, + "loss": 1.0713, + "step": 3224 + }, + { + "epoch": 0.5742521367521367, + "grad_norm": 0.5341386198997498, + "learning_rate": 0.00019010469382830947, + "loss": 1.1438, + "step": 3225 + }, + { + "epoch": 0.5744301994301995, + "grad_norm": 0.5300050973892212, + "learning_rate": 0.00019009862193977936, + "loss": 1.0114, + "step": 3226 + }, + { + "epoch": 0.5746082621082621, + "grad_norm": 0.6033682823181152, + "learning_rate": 0.0001900925482859539, + "loss": 1.0458, + "step": 3227 + }, + { + "epoch": 0.5747863247863247, + "grad_norm": 0.5108983516693115, + "learning_rate": 0.00019008647286695215, + "loss": 1.1211, + "step": 3228 + }, + { + "epoch": 0.5749643874643875, + "grad_norm": 0.5263782739639282, + "learning_rate": 0.00019008039568289308, + "loss": 0.8647, + "step": 3229 + }, + { + "epoch": 0.5751424501424501, + "grad_norm": 0.47119566798210144, + "learning_rate": 0.0001900743167338958, + "loss": 1.019, + "step": 3230 + }, + { + "epoch": 0.5753205128205128, + "grad_norm": 0.56391841173172, + "learning_rate": 0.00019006823602007937, + "loss": 0.9791, + "step": 3231 + }, + { + "epoch": 0.5754985754985755, + "grad_norm": 0.5364985466003418, + "learning_rate": 0.000190062153541563, + "loss": 1.1355, + "step": 3232 + }, + { + "epoch": 0.5756766381766382, + "grad_norm": 0.5098565220832825, + "learning_rate": 0.00019005606929846578, + "loss": 0.987, + "step": 3233 + }, + { + "epoch": 0.5758547008547008, + "grad_norm": 0.6640968918800354, + "learning_rate": 0.00019004998329090692, + "loss": 1.1165, + "step": 3234 + }, + { + "epoch": 0.5760327635327636, + "grad_norm": 0.5044721961021423, + "learning_rate": 0.00019004389551900578, + "loss": 0.8643, + "step": 3235 + }, + { + "epoch": 0.5762108262108262, + "grad_norm": 0.4822785258293152, + "learning_rate": 0.00019003780598288153, + "loss": 1.0735, + "step": 3236 + }, + { + "epoch": 0.5763888888888888, + "grad_norm": 0.505261242389679, + "learning_rate": 0.00019003171468265348, + "loss": 1.0001, + "step": 3237 + }, + { + "epoch": 0.5765669515669516, + "grad_norm": 0.5020412802696228, + "learning_rate": 0.00019002562161844102, + "loss": 0.9601, + "step": 3238 + }, + { + "epoch": 0.5767450142450142, + "grad_norm": 0.4920475482940674, + "learning_rate": 0.00019001952679036354, + "loss": 1.0111, + "step": 3239 + }, + { + "epoch": 0.5769230769230769, + "grad_norm": 0.5638813376426697, + "learning_rate": 0.00019001343019854042, + "loss": 1.1456, + "step": 3240 + }, + { + "epoch": 0.5771011396011396, + "grad_norm": 0.5519235134124756, + "learning_rate": 0.0001900073318430911, + "loss": 0.9258, + "step": 3241 + }, + { + "epoch": 0.5772792022792023, + "grad_norm": 0.5207770466804504, + "learning_rate": 0.0001900012317241351, + "loss": 0.9859, + "step": 3242 + }, + { + "epoch": 0.5774572649572649, + "grad_norm": 0.5493707656860352, + "learning_rate": 0.00018999512984179195, + "loss": 1.1183, + "step": 3243 + }, + { + "epoch": 0.5776353276353277, + "grad_norm": 0.4504764676094055, + "learning_rate": 0.00018998902619618116, + "loss": 0.9363, + "step": 3244 + }, + { + "epoch": 0.5778133903133903, + "grad_norm": 0.5232836604118347, + "learning_rate": 0.00018998292078742233, + "loss": 1.1887, + "step": 3245 + }, + { + "epoch": 0.5779914529914529, + "grad_norm": 0.5715088248252869, + "learning_rate": 0.0001899768136156351, + "loss": 1.4524, + "step": 3246 + }, + { + "epoch": 0.5781695156695157, + "grad_norm": 0.59555584192276, + "learning_rate": 0.0001899707046809391, + "loss": 1.0922, + "step": 3247 + }, + { + "epoch": 0.5783475783475783, + "grad_norm": 0.4500894546508789, + "learning_rate": 0.00018996459398345404, + "loss": 1.0087, + "step": 3248 + }, + { + "epoch": 0.5785256410256411, + "grad_norm": 0.49126625061035156, + "learning_rate": 0.00018995848152329967, + "loss": 1.1512, + "step": 3249 + }, + { + "epoch": 0.5787037037037037, + "grad_norm": 0.4096335172653198, + "learning_rate": 0.00018995236730059574, + "loss": 0.7633, + "step": 3250 + }, + { + "epoch": 0.5788817663817664, + "grad_norm": 0.5364313721656799, + "learning_rate": 0.00018994625131546199, + "loss": 1.295, + "step": 3251 + }, + { + "epoch": 0.5790598290598291, + "grad_norm": 0.4897502660751343, + "learning_rate": 0.00018994013356801834, + "loss": 1.2197, + "step": 3252 + }, + { + "epoch": 0.5792378917378918, + "grad_norm": 0.5101368427276611, + "learning_rate": 0.00018993401405838456, + "loss": 1.1129, + "step": 3253 + }, + { + "epoch": 0.5794159544159544, + "grad_norm": 0.5426377654075623, + "learning_rate": 0.00018992789278668063, + "loss": 1.188, + "step": 3254 + }, + { + "epoch": 0.5795940170940171, + "grad_norm": 0.5066362023353577, + "learning_rate": 0.00018992176975302644, + "loss": 1.2802, + "step": 3255 + }, + { + "epoch": 0.5797720797720798, + "grad_norm": 0.5418947339057922, + "learning_rate": 0.00018991564495754196, + "loss": 1.1675, + "step": 3256 + }, + { + "epoch": 0.5799501424501424, + "grad_norm": 0.5139963626861572, + "learning_rate": 0.0001899095184003472, + "loss": 0.9717, + "step": 3257 + }, + { + "epoch": 0.5801282051282052, + "grad_norm": 0.5167285799980164, + "learning_rate": 0.00018990339008156219, + "loss": 1.1529, + "step": 3258 + }, + { + "epoch": 0.5803062678062678, + "grad_norm": 0.53471440076828, + "learning_rate": 0.00018989726000130704, + "loss": 1.0711, + "step": 3259 + }, + { + "epoch": 0.5804843304843305, + "grad_norm": 0.49875229597091675, + "learning_rate": 0.0001898911281597018, + "loss": 1.1095, + "step": 3260 + }, + { + "epoch": 0.5806623931623932, + "grad_norm": 0.4473155438899994, + "learning_rate": 0.00018988499455686663, + "loss": 0.836, + "step": 3261 + }, + { + "epoch": 0.5808404558404558, + "grad_norm": 0.6181996464729309, + "learning_rate": 0.00018987885919292174, + "loss": 1.2787, + "step": 3262 + }, + { + "epoch": 0.5810185185185185, + "grad_norm": 0.4996899664402008, + "learning_rate": 0.00018987272206798733, + "loss": 1.2132, + "step": 3263 + }, + { + "epoch": 0.5811965811965812, + "grad_norm": 0.49979713559150696, + "learning_rate": 0.00018986658318218358, + "loss": 0.8388, + "step": 3264 + }, + { + "epoch": 0.5813746438746439, + "grad_norm": 0.5288876295089722, + "learning_rate": 0.00018986044253563084, + "loss": 1.1871, + "step": 3265 + }, + { + "epoch": 0.5815527065527065, + "grad_norm": 0.534063458442688, + "learning_rate": 0.00018985430012844937, + "loss": 0.96, + "step": 3266 + }, + { + "epoch": 0.5817307692307693, + "grad_norm": 0.5081285834312439, + "learning_rate": 0.00018984815596075953, + "loss": 1.1577, + "step": 3267 + }, + { + "epoch": 0.5819088319088319, + "grad_norm": 0.5648202896118164, + "learning_rate": 0.00018984201003268176, + "loss": 1.2235, + "step": 3268 + }, + { + "epoch": 0.5820868945868946, + "grad_norm": 0.495061993598938, + "learning_rate": 0.00018983586234433642, + "loss": 1.056, + "step": 3269 + }, + { + "epoch": 0.5822649572649573, + "grad_norm": 0.47149857878685, + "learning_rate": 0.000189829712895844, + "loss": 1.0844, + "step": 3270 + }, + { + "epoch": 0.58244301994302, + "grad_norm": 0.6107062697410583, + "learning_rate": 0.00018982356168732492, + "loss": 0.9868, + "step": 3271 + }, + { + "epoch": 0.5826210826210826, + "grad_norm": 0.7355940341949463, + "learning_rate": 0.00018981740871889974, + "loss": 1.1448, + "step": 3272 + }, + { + "epoch": 0.5827991452991453, + "grad_norm": 0.5950441956520081, + "learning_rate": 0.00018981125399068907, + "loss": 0.9618, + "step": 3273 + }, + { + "epoch": 0.582977207977208, + "grad_norm": 0.47607290744781494, + "learning_rate": 0.0001898050975028134, + "loss": 0.957, + "step": 3274 + }, + { + "epoch": 0.5831552706552706, + "grad_norm": 0.541164755821228, + "learning_rate": 0.00018979893925539338, + "loss": 1.1426, + "step": 3275 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 0.5240640044212341, + "learning_rate": 0.00018979277924854974, + "loss": 1.1421, + "step": 3276 + }, + { + "epoch": 0.583511396011396, + "grad_norm": 0.48155727982521057, + "learning_rate": 0.00018978661748240307, + "loss": 1.0069, + "step": 3277 + }, + { + "epoch": 0.5836894586894587, + "grad_norm": 0.5559938549995422, + "learning_rate": 0.00018978045395707418, + "loss": 1.1227, + "step": 3278 + }, + { + "epoch": 0.5838675213675214, + "grad_norm": 0.5244291424751282, + "learning_rate": 0.0001897742886726838, + "loss": 1.1103, + "step": 3279 + }, + { + "epoch": 0.584045584045584, + "grad_norm": 0.5277758240699768, + "learning_rate": 0.00018976812162935268, + "loss": 1.2125, + "step": 3280 + }, + { + "epoch": 0.5842236467236467, + "grad_norm": 0.5415039658546448, + "learning_rate": 0.00018976195282720173, + "loss": 1.146, + "step": 3281 + }, + { + "epoch": 0.5844017094017094, + "grad_norm": 0.5152051448822021, + "learning_rate": 0.00018975578226635177, + "loss": 1.0092, + "step": 3282 + }, + { + "epoch": 0.5845797720797721, + "grad_norm": 0.5489452481269836, + "learning_rate": 0.00018974960994692371, + "loss": 1.2425, + "step": 3283 + }, + { + "epoch": 0.5847578347578347, + "grad_norm": 0.491274356842041, + "learning_rate": 0.00018974343586903848, + "loss": 0.9559, + "step": 3284 + }, + { + "epoch": 0.5849358974358975, + "grad_norm": 0.5783739686012268, + "learning_rate": 0.00018973726003281707, + "loss": 1.1971, + "step": 3285 + }, + { + "epoch": 0.5851139601139601, + "grad_norm": 0.5056472420692444, + "learning_rate": 0.00018973108243838045, + "loss": 1.0313, + "step": 3286 + }, + { + "epoch": 0.5852920227920227, + "grad_norm": 0.4939729571342468, + "learning_rate": 0.00018972490308584962, + "loss": 1.1061, + "step": 3287 + }, + { + "epoch": 0.5854700854700855, + "grad_norm": 0.4889580011367798, + "learning_rate": 0.00018971872197534576, + "loss": 0.9157, + "step": 3288 + }, + { + "epoch": 0.5856481481481481, + "grad_norm": 0.40889349579811096, + "learning_rate": 0.00018971253910698993, + "loss": 0.8083, + "step": 3289 + }, + { + "epoch": 0.5858262108262108, + "grad_norm": 0.5221503973007202, + "learning_rate": 0.00018970635448090322, + "loss": 0.9995, + "step": 3290 + }, + { + "epoch": 0.5860042735042735, + "grad_norm": 0.47060561180114746, + "learning_rate": 0.00018970016809720687, + "loss": 0.9738, + "step": 3291 + }, + { + "epoch": 0.5861823361823362, + "grad_norm": 0.6083170771598816, + "learning_rate": 0.000189693979956022, + "loss": 1.188, + "step": 3292 + }, + { + "epoch": 0.5863603988603988, + "grad_norm": 0.4696751534938812, + "learning_rate": 0.00018968779005746998, + "loss": 1.089, + "step": 3293 + }, + { + "epoch": 0.5865384615384616, + "grad_norm": 0.5081014633178711, + "learning_rate": 0.00018968159840167202, + "loss": 1.1869, + "step": 3294 + }, + { + "epoch": 0.5867165242165242, + "grad_norm": 0.48042431473731995, + "learning_rate": 0.0001896754049887494, + "loss": 0.964, + "step": 3295 + }, + { + "epoch": 0.5868945868945868, + "grad_norm": 0.5075193643569946, + "learning_rate": 0.00018966920981882353, + "loss": 1.1884, + "step": 3296 + }, + { + "epoch": 0.5870726495726496, + "grad_norm": 0.5734842419624329, + "learning_rate": 0.00018966301289201576, + "loss": 1.1475, + "step": 3297 + }, + { + "epoch": 0.5872507122507122, + "grad_norm": 0.5525311231613159, + "learning_rate": 0.00018965681420844753, + "loss": 1.241, + "step": 3298 + }, + { + "epoch": 0.5874287749287749, + "grad_norm": 0.48142680525779724, + "learning_rate": 0.00018965061376824025, + "loss": 1.0871, + "step": 3299 + }, + { + "epoch": 0.5876068376068376, + "grad_norm": 0.5360350608825684, + "learning_rate": 0.00018964441157151544, + "loss": 1.1895, + "step": 3300 + }, + { + "epoch": 0.5877849002849003, + "grad_norm": 0.5207685232162476, + "learning_rate": 0.00018963820761839457, + "loss": 0.9323, + "step": 3301 + }, + { + "epoch": 0.5879629629629629, + "grad_norm": 0.453620970249176, + "learning_rate": 0.00018963200190899926, + "loss": 0.802, + "step": 3302 + }, + { + "epoch": 0.5881410256410257, + "grad_norm": 0.5198796391487122, + "learning_rate": 0.00018962579444345106, + "loss": 1.0243, + "step": 3303 + }, + { + "epoch": 0.5883190883190883, + "grad_norm": 0.5597525835037231, + "learning_rate": 0.0001896195852218716, + "loss": 0.9351, + "step": 3304 + }, + { + "epoch": 0.5884971509971509, + "grad_norm": 0.5738299489021301, + "learning_rate": 0.00018961337424438254, + "loss": 1.3737, + "step": 3305 + }, + { + "epoch": 0.5886752136752137, + "grad_norm": 0.5569949150085449, + "learning_rate": 0.00018960716151110554, + "loss": 1.0469, + "step": 3306 + }, + { + "epoch": 0.5888532763532763, + "grad_norm": 0.5088010430335999, + "learning_rate": 0.00018960094702216238, + "loss": 1.0982, + "step": 3307 + }, + { + "epoch": 0.5890313390313391, + "grad_norm": 0.5127636790275574, + "learning_rate": 0.0001895947307776748, + "loss": 0.9986, + "step": 3308 + }, + { + "epoch": 0.5892094017094017, + "grad_norm": 0.5160682797431946, + "learning_rate": 0.00018958851277776456, + "loss": 1.0219, + "step": 3309 + }, + { + "epoch": 0.5893874643874644, + "grad_norm": 0.5380711555480957, + "learning_rate": 0.00018958229302255356, + "loss": 1.118, + "step": 3310 + }, + { + "epoch": 0.5895655270655271, + "grad_norm": 0.5571228861808777, + "learning_rate": 0.0001895760715121636, + "loss": 1.0302, + "step": 3311 + }, + { + "epoch": 0.5897435897435898, + "grad_norm": 0.542266309261322, + "learning_rate": 0.00018956984824671657, + "loss": 1.0372, + "step": 3312 + }, + { + "epoch": 0.5899216524216524, + "grad_norm": 0.48350459337234497, + "learning_rate": 0.00018956362322633446, + "loss": 1.2, + "step": 3313 + }, + { + "epoch": 0.5900997150997151, + "grad_norm": 0.5001645088195801, + "learning_rate": 0.0001895573964511392, + "loss": 0.9749, + "step": 3314 + }, + { + "epoch": 0.5902777777777778, + "grad_norm": 0.5227531790733337, + "learning_rate": 0.00018955116792125276, + "loss": 1.025, + "step": 3315 + }, + { + "epoch": 0.5904558404558404, + "grad_norm": 0.522251546382904, + "learning_rate": 0.00018954493763679727, + "loss": 1.0821, + "step": 3316 + }, + { + "epoch": 0.5906339031339032, + "grad_norm": 0.5423251390457153, + "learning_rate": 0.00018953870559789467, + "loss": 1.0961, + "step": 3317 + }, + { + "epoch": 0.5908119658119658, + "grad_norm": 0.5615720748901367, + "learning_rate": 0.0001895324718046672, + "loss": 1.1209, + "step": 3318 + }, + { + "epoch": 0.5909900284900285, + "grad_norm": 0.44746771454811096, + "learning_rate": 0.00018952623625723692, + "loss": 0.9935, + "step": 3319 + }, + { + "epoch": 0.5911680911680912, + "grad_norm": 0.5993229150772095, + "learning_rate": 0.00018951999895572597, + "loss": 1.1409, + "step": 3320 + }, + { + "epoch": 0.5913461538461539, + "grad_norm": 0.4969801902770996, + "learning_rate": 0.00018951375990025666, + "loss": 1.1568, + "step": 3321 + }, + { + "epoch": 0.5915242165242165, + "grad_norm": 0.6001267433166504, + "learning_rate": 0.00018950751909095116, + "loss": 1.1135, + "step": 3322 + }, + { + "epoch": 0.5917022792022792, + "grad_norm": 0.5386021733283997, + "learning_rate": 0.00018950127652793172, + "loss": 0.947, + "step": 3323 + }, + { + "epoch": 0.5918803418803419, + "grad_norm": 0.49043843150138855, + "learning_rate": 0.00018949503221132074, + "loss": 0.9581, + "step": 3324 + }, + { + "epoch": 0.5920584045584045, + "grad_norm": 0.5241141319274902, + "learning_rate": 0.00018948878614124048, + "loss": 1.0797, + "step": 3325 + }, + { + "epoch": 0.5922364672364673, + "grad_norm": 0.5755026340484619, + "learning_rate": 0.00018948253831781338, + "loss": 1.1046, + "step": 3326 + }, + { + "epoch": 0.5924145299145299, + "grad_norm": 0.5004449486732483, + "learning_rate": 0.00018947628874116179, + "loss": 1.1416, + "step": 3327 + }, + { + "epoch": 0.5925925925925926, + "grad_norm": 0.53347247838974, + "learning_rate": 0.00018947003741140821, + "loss": 1.2718, + "step": 3328 + }, + { + "epoch": 0.5927706552706553, + "grad_norm": 0.6473469138145447, + "learning_rate": 0.0001894637843286751, + "loss": 1.2255, + "step": 3329 + }, + { + "epoch": 0.592948717948718, + "grad_norm": 0.4750518798828125, + "learning_rate": 0.00018945752949308498, + "loss": 1.0537, + "step": 3330 + }, + { + "epoch": 0.5931267806267806, + "grad_norm": 0.5636306405067444, + "learning_rate": 0.00018945127290476043, + "loss": 0.9906, + "step": 3331 + }, + { + "epoch": 0.5933048433048433, + "grad_norm": 0.4871736466884613, + "learning_rate": 0.00018944501456382397, + "loss": 1.0549, + "step": 3332 + }, + { + "epoch": 0.593482905982906, + "grad_norm": 0.5554637312889099, + "learning_rate": 0.0001894387544703983, + "loss": 1.1587, + "step": 3333 + }, + { + "epoch": 0.5936609686609686, + "grad_norm": 0.5385799407958984, + "learning_rate": 0.000189432492624606, + "loss": 0.9565, + "step": 3334 + }, + { + "epoch": 0.5938390313390314, + "grad_norm": 0.4996553063392639, + "learning_rate": 0.00018942622902656976, + "loss": 1.0456, + "step": 3335 + }, + { + "epoch": 0.594017094017094, + "grad_norm": 0.46810707449913025, + "learning_rate": 0.00018941996367641237, + "loss": 1.119, + "step": 3336 + }, + { + "epoch": 0.5941951566951567, + "grad_norm": 0.5672653913497925, + "learning_rate": 0.0001894136965742565, + "loss": 1.1317, + "step": 3337 + }, + { + "epoch": 0.5943732193732194, + "grad_norm": 0.4790053367614746, + "learning_rate": 0.00018940742772022504, + "loss": 1.0967, + "step": 3338 + }, + { + "epoch": 0.594551282051282, + "grad_norm": 0.5935906171798706, + "learning_rate": 0.00018940115711444072, + "loss": 1.3044, + "step": 3339 + }, + { + "epoch": 0.5947293447293447, + "grad_norm": 0.4790516793727875, + "learning_rate": 0.00018939488475702647, + "loss": 1.074, + "step": 3340 + }, + { + "epoch": 0.5949074074074074, + "grad_norm": 0.474588006734848, + "learning_rate": 0.00018938861064810516, + "loss": 1.1476, + "step": 3341 + }, + { + "epoch": 0.5950854700854701, + "grad_norm": 0.4908665120601654, + "learning_rate": 0.0001893823347877997, + "loss": 1.216, + "step": 3342 + }, + { + "epoch": 0.5952635327635327, + "grad_norm": 0.531650960445404, + "learning_rate": 0.00018937605717623307, + "loss": 1.1057, + "step": 3343 + }, + { + "epoch": 0.5954415954415955, + "grad_norm": 0.5581082105636597, + "learning_rate": 0.00018936977781352823, + "loss": 0.7972, + "step": 3344 + }, + { + "epoch": 0.5956196581196581, + "grad_norm": 0.42370662093162537, + "learning_rate": 0.00018936349669980827, + "loss": 0.8888, + "step": 3345 + }, + { + "epoch": 0.5957977207977208, + "grad_norm": 0.5817318558692932, + "learning_rate": 0.00018935721383519624, + "loss": 1.2801, + "step": 3346 + }, + { + "epoch": 0.5959757834757835, + "grad_norm": 0.4766376316547394, + "learning_rate": 0.00018935092921981524, + "loss": 1.0918, + "step": 3347 + }, + { + "epoch": 0.5961538461538461, + "grad_norm": 0.5567346811294556, + "learning_rate": 0.00018934464285378836, + "loss": 1.0269, + "step": 3348 + }, + { + "epoch": 0.5963319088319088, + "grad_norm": 0.5285565257072449, + "learning_rate": 0.0001893383547372388, + "loss": 1.1887, + "step": 3349 + }, + { + "epoch": 0.5965099715099715, + "grad_norm": 0.49052694439888, + "learning_rate": 0.00018933206487028979, + "loss": 1.0773, + "step": 3350 + }, + { + "epoch": 0.5966880341880342, + "grad_norm": 0.6175199151039124, + "learning_rate": 0.0001893257732530645, + "loss": 1.0192, + "step": 3351 + }, + { + "epoch": 0.5968660968660968, + "grad_norm": 0.56049644947052, + "learning_rate": 0.00018931947988568628, + "loss": 0.9516, + "step": 3352 + }, + { + "epoch": 0.5970441595441596, + "grad_norm": 0.47873660922050476, + "learning_rate": 0.00018931318476827838, + "loss": 0.8174, + "step": 3353 + }, + { + "epoch": 0.5972222222222222, + "grad_norm": 0.4748854339122772, + "learning_rate": 0.00018930688790096416, + "loss": 1.0238, + "step": 3354 + }, + { + "epoch": 0.5974002849002849, + "grad_norm": 0.5382232666015625, + "learning_rate": 0.00018930058928386698, + "loss": 1.0815, + "step": 3355 + }, + { + "epoch": 0.5975783475783476, + "grad_norm": 0.5038299560546875, + "learning_rate": 0.00018929428891711027, + "loss": 1.0472, + "step": 3356 + }, + { + "epoch": 0.5977564102564102, + "grad_norm": 0.5185908079147339, + "learning_rate": 0.00018928798680081744, + "loss": 1.0435, + "step": 3357 + }, + { + "epoch": 0.5979344729344729, + "grad_norm": 0.5169877409934998, + "learning_rate": 0.00018928168293511202, + "loss": 1.0437, + "step": 3358 + }, + { + "epoch": 0.5981125356125356, + "grad_norm": 0.5218369960784912, + "learning_rate": 0.00018927537732011749, + "loss": 1.082, + "step": 3359 + }, + { + "epoch": 0.5982905982905983, + "grad_norm": 0.5358219742774963, + "learning_rate": 0.0001892690699559574, + "loss": 1.2523, + "step": 3360 + }, + { + "epoch": 0.5984686609686609, + "grad_norm": 0.47716647386550903, + "learning_rate": 0.0001892627608427553, + "loss": 1.2069, + "step": 3361 + }, + { + "epoch": 0.5986467236467237, + "grad_norm": 0.5484169125556946, + "learning_rate": 0.00018925644998063482, + "loss": 1.2016, + "step": 3362 + }, + { + "epoch": 0.5988247863247863, + "grad_norm": 0.46814846992492676, + "learning_rate": 0.00018925013736971965, + "loss": 0.7989, + "step": 3363 + }, + { + "epoch": 0.5990028490028491, + "grad_norm": 0.5391258001327515, + "learning_rate": 0.0001892438230101334, + "loss": 1.224, + "step": 3364 + }, + { + "epoch": 0.5991809116809117, + "grad_norm": 0.5248384475708008, + "learning_rate": 0.00018923750690199987, + "loss": 1.1532, + "step": 3365 + }, + { + "epoch": 0.5993589743589743, + "grad_norm": 0.5074637532234192, + "learning_rate": 0.00018923118904544273, + "loss": 1.0968, + "step": 3366 + }, + { + "epoch": 0.5995370370370371, + "grad_norm": 0.5260029435157776, + "learning_rate": 0.00018922486944058581, + "loss": 1.1311, + "step": 3367 + }, + { + "epoch": 0.5997150997150997, + "grad_norm": 0.48497965931892395, + "learning_rate": 0.00018921854808755294, + "loss": 1.1208, + "step": 3368 + }, + { + "epoch": 0.5998931623931624, + "grad_norm": 0.5108651518821716, + "learning_rate": 0.00018921222498646792, + "loss": 1.147, + "step": 3369 + }, + { + "epoch": 0.6000712250712251, + "grad_norm": 0.5243437886238098, + "learning_rate": 0.00018920590013745471, + "loss": 0.9614, + "step": 3370 + }, + { + "epoch": 0.6002492877492878, + "grad_norm": 0.47022634744644165, + "learning_rate": 0.00018919957354063719, + "loss": 1.0579, + "step": 3371 + }, + { + "epoch": 0.6004273504273504, + "grad_norm": 0.6461413502693176, + "learning_rate": 0.00018919324519613931, + "loss": 1.2126, + "step": 3372 + }, + { + "epoch": 0.6006054131054132, + "grad_norm": 0.4654616713523865, + "learning_rate": 0.00018918691510408508, + "loss": 1.1476, + "step": 3373 + }, + { + "epoch": 0.6007834757834758, + "grad_norm": 0.48571303486824036, + "learning_rate": 0.00018918058326459854, + "loss": 1.2093, + "step": 3374 + }, + { + "epoch": 0.6009615384615384, + "grad_norm": 0.5255016684532166, + "learning_rate": 0.00018917424967780368, + "loss": 1.1538, + "step": 3375 + }, + { + "epoch": 0.6011396011396012, + "grad_norm": 0.5059894323348999, + "learning_rate": 0.00018916791434382468, + "loss": 1.0556, + "step": 3376 + }, + { + "epoch": 0.6013176638176638, + "grad_norm": 0.4581229090690613, + "learning_rate": 0.00018916157726278561, + "loss": 1.1468, + "step": 3377 + }, + { + "epoch": 0.6014957264957265, + "grad_norm": 0.5701818466186523, + "learning_rate": 0.00018915523843481067, + "loss": 1.3641, + "step": 3378 + }, + { + "epoch": 0.6016737891737892, + "grad_norm": 0.5007243752479553, + "learning_rate": 0.00018914889786002403, + "loss": 1.2705, + "step": 3379 + }, + { + "epoch": 0.6018518518518519, + "grad_norm": 0.5192995071411133, + "learning_rate": 0.0001891425555385499, + "loss": 0.9922, + "step": 3380 + }, + { + "epoch": 0.6020299145299145, + "grad_norm": 0.5880612134933472, + "learning_rate": 0.00018913621147051258, + "loss": 0.8783, + "step": 3381 + }, + { + "epoch": 0.6022079772079773, + "grad_norm": 0.5161563158035278, + "learning_rate": 0.0001891298656560364, + "loss": 0.9634, + "step": 3382 + }, + { + "epoch": 0.6023860398860399, + "grad_norm": 0.48450782895088196, + "learning_rate": 0.00018912351809524563, + "loss": 0.809, + "step": 3383 + }, + { + "epoch": 0.6025641025641025, + "grad_norm": 0.621537983417511, + "learning_rate": 0.00018911716878826465, + "loss": 1.2031, + "step": 3384 + }, + { + "epoch": 0.6027421652421653, + "grad_norm": 0.6014544367790222, + "learning_rate": 0.00018911081773521787, + "loss": 1.1552, + "step": 3385 + }, + { + "epoch": 0.6029202279202279, + "grad_norm": 0.49995481967926025, + "learning_rate": 0.00018910446493622976, + "loss": 0.8569, + "step": 3386 + }, + { + "epoch": 0.6030982905982906, + "grad_norm": 0.5157307386398315, + "learning_rate": 0.00018909811039142472, + "loss": 0.9515, + "step": 3387 + }, + { + "epoch": 0.6032763532763533, + "grad_norm": 0.5164140462875366, + "learning_rate": 0.0001890917541009273, + "loss": 0.9803, + "step": 3388 + }, + { + "epoch": 0.603454415954416, + "grad_norm": 0.5555596947669983, + "learning_rate": 0.00018908539606486206, + "loss": 1.2994, + "step": 3389 + }, + { + "epoch": 0.6036324786324786, + "grad_norm": 0.605697512626648, + "learning_rate": 0.00018907903628335353, + "loss": 1.2865, + "step": 3390 + }, + { + "epoch": 0.6038105413105413, + "grad_norm": 0.5700713992118835, + "learning_rate": 0.0001890726747565263, + "loss": 1.2493, + "step": 3391 + }, + { + "epoch": 0.603988603988604, + "grad_norm": 0.5516746044158936, + "learning_rate": 0.0001890663114845051, + "loss": 1.2743, + "step": 3392 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 0.5233162641525269, + "learning_rate": 0.0001890599464674145, + "loss": 0.9237, + "step": 3393 + }, + { + "epoch": 0.6043447293447294, + "grad_norm": 0.5709942579269409, + "learning_rate": 0.00018905357970537925, + "loss": 0.9922, + "step": 3394 + }, + { + "epoch": 0.604522792022792, + "grad_norm": 0.48403796553611755, + "learning_rate": 0.0001890472111985241, + "loss": 1.1255, + "step": 3395 + }, + { + "epoch": 0.6047008547008547, + "grad_norm": 0.628718376159668, + "learning_rate": 0.00018904084094697386, + "loss": 1.1458, + "step": 3396 + }, + { + "epoch": 0.6048789173789174, + "grad_norm": 0.46822869777679443, + "learning_rate": 0.00018903446895085328, + "loss": 0.8727, + "step": 3397 + }, + { + "epoch": 0.60505698005698, + "grad_norm": 0.505584180355072, + "learning_rate": 0.00018902809521028724, + "loss": 1.1595, + "step": 3398 + }, + { + "epoch": 0.6052350427350427, + "grad_norm": 0.4494974911212921, + "learning_rate": 0.00018902171972540058, + "loss": 0.6685, + "step": 3399 + }, + { + "epoch": 0.6054131054131054, + "grad_norm": 0.5101519227027893, + "learning_rate": 0.0001890153424963183, + "loss": 0.9313, + "step": 3400 + }, + { + "epoch": 0.6055911680911681, + "grad_norm": 0.5081079602241516, + "learning_rate": 0.00018900896352316528, + "loss": 1.2588, + "step": 3401 + }, + { + "epoch": 0.6057692307692307, + "grad_norm": 0.5784309506416321, + "learning_rate": 0.00018900258280606653, + "loss": 1.2077, + "step": 3402 + }, + { + "epoch": 0.6059472934472935, + "grad_norm": 0.4506312608718872, + "learning_rate": 0.00018899620034514705, + "loss": 1.05, + "step": 3403 + }, + { + "epoch": 0.6061253561253561, + "grad_norm": 0.5243048071861267, + "learning_rate": 0.0001889898161405319, + "loss": 1.2295, + "step": 3404 + }, + { + "epoch": 0.6063034188034188, + "grad_norm": 0.5447196364402771, + "learning_rate": 0.00018898343019234615, + "loss": 1.1476, + "step": 3405 + }, + { + "epoch": 0.6064814814814815, + "grad_norm": 0.46813663840293884, + "learning_rate": 0.00018897704250071492, + "loss": 1.2113, + "step": 3406 + }, + { + "epoch": 0.6066595441595442, + "grad_norm": 0.5340631604194641, + "learning_rate": 0.00018897065306576342, + "loss": 1.1656, + "step": 3407 + }, + { + "epoch": 0.6068376068376068, + "grad_norm": 0.513708233833313, + "learning_rate": 0.00018896426188761675, + "loss": 1.1616, + "step": 3408 + }, + { + "epoch": 0.6070156695156695, + "grad_norm": 0.594601035118103, + "learning_rate": 0.00018895786896640023, + "loss": 1.2564, + "step": 3409 + }, + { + "epoch": 0.6071937321937322, + "grad_norm": 0.45067599415779114, + "learning_rate": 0.000188951474302239, + "loss": 1.0107, + "step": 3410 + }, + { + "epoch": 0.6073717948717948, + "grad_norm": 0.5394250750541687, + "learning_rate": 0.00018894507789525843, + "loss": 1.4081, + "step": 3411 + }, + { + "epoch": 0.6075498575498576, + "grad_norm": 0.5612049102783203, + "learning_rate": 0.00018893867974558383, + "loss": 1.1015, + "step": 3412 + }, + { + "epoch": 0.6077279202279202, + "grad_norm": 0.4794061779975891, + "learning_rate": 0.00018893227985334056, + "loss": 1.2103, + "step": 3413 + }, + { + "epoch": 0.6079059829059829, + "grad_norm": 0.6060562133789062, + "learning_rate": 0.00018892587821865402, + "loss": 1.3693, + "step": 3414 + }, + { + "epoch": 0.6080840455840456, + "grad_norm": 0.44624534249305725, + "learning_rate": 0.00018891947484164963, + "loss": 0.8209, + "step": 3415 + }, + { + "epoch": 0.6082621082621082, + "grad_norm": 0.49297213554382324, + "learning_rate": 0.0001889130697224528, + "loss": 1.2027, + "step": 3416 + }, + { + "epoch": 0.6084401709401709, + "grad_norm": 0.4431746304035187, + "learning_rate": 0.0001889066628611891, + "loss": 1.0347, + "step": 3417 + }, + { + "epoch": 0.6086182336182336, + "grad_norm": 0.5425933599472046, + "learning_rate": 0.00018890025425798404, + "loss": 1.0556, + "step": 3418 + }, + { + "epoch": 0.6087962962962963, + "grad_norm": 0.5502763390541077, + "learning_rate": 0.00018889384391296315, + "loss": 1.2362, + "step": 3419 + }, + { + "epoch": 0.6089743589743589, + "grad_norm": 0.5442292094230652, + "learning_rate": 0.00018888743182625203, + "loss": 1.1306, + "step": 3420 + }, + { + "epoch": 0.6091524216524217, + "grad_norm": 0.4651123583316803, + "learning_rate": 0.00018888101799797636, + "loss": 0.9305, + "step": 3421 + }, + { + "epoch": 0.6093304843304843, + "grad_norm": 0.4713892340660095, + "learning_rate": 0.00018887460242826177, + "loss": 1.0789, + "step": 3422 + }, + { + "epoch": 0.6095085470085471, + "grad_norm": 0.5283244848251343, + "learning_rate": 0.00018886818511723398, + "loss": 1.345, + "step": 3423 + }, + { + "epoch": 0.6096866096866097, + "grad_norm": 0.5527324080467224, + "learning_rate": 0.0001888617660650187, + "loss": 1.1297, + "step": 3424 + }, + { + "epoch": 0.6098646723646723, + "grad_norm": 0.5412901043891907, + "learning_rate": 0.00018885534527174168, + "loss": 1.1213, + "step": 3425 + }, + { + "epoch": 0.6100427350427351, + "grad_norm": 0.5295354127883911, + "learning_rate": 0.00018884892273752878, + "loss": 1.1217, + "step": 3426 + }, + { + "epoch": 0.6102207977207977, + "grad_norm": 0.461900532245636, + "learning_rate": 0.0001888424984625058, + "loss": 0.827, + "step": 3427 + }, + { + "epoch": 0.6103988603988604, + "grad_norm": 0.4922671616077423, + "learning_rate": 0.00018883607244679865, + "loss": 1.2216, + "step": 3428 + }, + { + "epoch": 0.6105769230769231, + "grad_norm": 0.5080927014350891, + "learning_rate": 0.00018882964469053317, + "loss": 1.2446, + "step": 3429 + }, + { + "epoch": 0.6107549857549858, + "grad_norm": 0.5523943901062012, + "learning_rate": 0.00018882321519383534, + "loss": 1.3346, + "step": 3430 + }, + { + "epoch": 0.6109330484330484, + "grad_norm": 0.5105271935462952, + "learning_rate": 0.0001888167839568311, + "loss": 1.1311, + "step": 3431 + }, + { + "epoch": 0.6111111111111112, + "grad_norm": 0.5635872483253479, + "learning_rate": 0.0001888103509796465, + "loss": 1.1875, + "step": 3432 + }, + { + "epoch": 0.6112891737891738, + "grad_norm": 0.4619547426700592, + "learning_rate": 0.00018880391626240755, + "loss": 0.9176, + "step": 3433 + }, + { + "epoch": 0.6114672364672364, + "grad_norm": 0.5896356105804443, + "learning_rate": 0.00018879747980524034, + "loss": 1.0251, + "step": 3434 + }, + { + "epoch": 0.6116452991452992, + "grad_norm": 0.49062737822532654, + "learning_rate": 0.000188791041608271, + "loss": 1.1598, + "step": 3435 + }, + { + "epoch": 0.6118233618233618, + "grad_norm": 0.45717164874076843, + "learning_rate": 0.00018878460167162558, + "loss": 0.8647, + "step": 3436 + }, + { + "epoch": 0.6120014245014245, + "grad_norm": 0.5903525352478027, + "learning_rate": 0.00018877815999543038, + "loss": 0.9671, + "step": 3437 + }, + { + "epoch": 0.6121794871794872, + "grad_norm": 0.5315384268760681, + "learning_rate": 0.00018877171657981153, + "loss": 1.1759, + "step": 3438 + }, + { + "epoch": 0.6123575498575499, + "grad_norm": 0.5650150775909424, + "learning_rate": 0.0001887652714248953, + "loss": 1.0128, + "step": 3439 + }, + { + "epoch": 0.6125356125356125, + "grad_norm": 0.49841752648353577, + "learning_rate": 0.000188758824530808, + "loss": 1.1259, + "step": 3440 + }, + { + "epoch": 0.6127136752136753, + "grad_norm": 0.4985620975494385, + "learning_rate": 0.00018875237589767593, + "loss": 1.0158, + "step": 3441 + }, + { + "epoch": 0.6128917378917379, + "grad_norm": 0.45266565680503845, + "learning_rate": 0.00018874592552562536, + "loss": 0.93, + "step": 3442 + }, + { + "epoch": 0.6130698005698005, + "grad_norm": 0.5696130990982056, + "learning_rate": 0.00018873947341478274, + "loss": 1.1432, + "step": 3443 + }, + { + "epoch": 0.6132478632478633, + "grad_norm": 0.5211645364761353, + "learning_rate": 0.00018873301956527451, + "loss": 1.1317, + "step": 3444 + }, + { + "epoch": 0.6134259259259259, + "grad_norm": 0.4991866946220398, + "learning_rate": 0.00018872656397722707, + "loss": 1.0362, + "step": 3445 + }, + { + "epoch": 0.6136039886039886, + "grad_norm": 0.5109508037567139, + "learning_rate": 0.00018872010665076694, + "loss": 1.2728, + "step": 3446 + }, + { + "epoch": 0.6137820512820513, + "grad_norm": 0.5838373899459839, + "learning_rate": 0.00018871364758602058, + "loss": 1.1131, + "step": 3447 + }, + { + "epoch": 0.613960113960114, + "grad_norm": 0.5139824151992798, + "learning_rate": 0.00018870718678311462, + "loss": 1.238, + "step": 3448 + }, + { + "epoch": 0.6141381766381766, + "grad_norm": 0.4852082431316376, + "learning_rate": 0.00018870072424217562, + "loss": 1.0677, + "step": 3449 + }, + { + "epoch": 0.6143162393162394, + "grad_norm": 0.5312315225601196, + "learning_rate": 0.00018869425996333018, + "loss": 1.178, + "step": 3450 + }, + { + "epoch": 0.614494301994302, + "grad_norm": 0.6343565583229065, + "learning_rate": 0.00018868779394670492, + "loss": 0.8839, + "step": 3451 + }, + { + "epoch": 0.6146723646723646, + "grad_norm": 0.6029773950576782, + "learning_rate": 0.00018868132619242662, + "loss": 1.1188, + "step": 3452 + }, + { + "epoch": 0.6148504273504274, + "grad_norm": 0.5246016383171082, + "learning_rate": 0.00018867485670062193, + "loss": 1.0797, + "step": 3453 + }, + { + "epoch": 0.61502849002849, + "grad_norm": 0.49307698011398315, + "learning_rate": 0.00018866838547141763, + "loss": 0.9749, + "step": 3454 + }, + { + "epoch": 0.6152065527065527, + "grad_norm": 0.5232903361320496, + "learning_rate": 0.00018866191250494052, + "loss": 1.0785, + "step": 3455 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 0.5545645356178284, + "learning_rate": 0.0001886554378013174, + "loss": 1.0496, + "step": 3456 + }, + { + "epoch": 0.6155626780626781, + "grad_norm": 0.493945837020874, + "learning_rate": 0.00018864896136067515, + "loss": 0.9248, + "step": 3457 + }, + { + "epoch": 0.6157407407407407, + "grad_norm": 0.5223548412322998, + "learning_rate": 0.00018864248318314065, + "loss": 1.0617, + "step": 3458 + }, + { + "epoch": 0.6159188034188035, + "grad_norm": 0.5666514039039612, + "learning_rate": 0.00018863600326884082, + "loss": 0.9981, + "step": 3459 + }, + { + "epoch": 0.6160968660968661, + "grad_norm": 0.4648127257823944, + "learning_rate": 0.00018862952161790265, + "loss": 0.917, + "step": 3460 + }, + { + "epoch": 0.6162749287749287, + "grad_norm": 0.590326189994812, + "learning_rate": 0.0001886230382304531, + "loss": 1.044, + "step": 3461 + }, + { + "epoch": 0.6164529914529915, + "grad_norm": 0.5511625409126282, + "learning_rate": 0.00018861655310661925, + "loss": 1.0988, + "step": 3462 + }, + { + "epoch": 0.6166310541310541, + "grad_norm": 0.567182183265686, + "learning_rate": 0.0001886100662465281, + "loss": 1.3017, + "step": 3463 + }, + { + "epoch": 0.6168091168091168, + "grad_norm": 0.5708897709846497, + "learning_rate": 0.0001886035776503068, + "loss": 0.9123, + "step": 3464 + }, + { + "epoch": 0.6169871794871795, + "grad_norm": 0.4945180416107178, + "learning_rate": 0.0001885970873180824, + "loss": 1.1645, + "step": 3465 + }, + { + "epoch": 0.6171652421652422, + "grad_norm": 0.4713336229324341, + "learning_rate": 0.00018859059524998215, + "loss": 1.0546, + "step": 3466 + }, + { + "epoch": 0.6173433048433048, + "grad_norm": 0.532859206199646, + "learning_rate": 0.0001885841014461332, + "loss": 1.0795, + "step": 3467 + }, + { + "epoch": 0.6175213675213675, + "grad_norm": 0.5165733695030212, + "learning_rate": 0.00018857760590666284, + "loss": 1.1284, + "step": 3468 + }, + { + "epoch": 0.6176994301994302, + "grad_norm": 0.48623126745224, + "learning_rate": 0.00018857110863169826, + "loss": 0.8618, + "step": 3469 + }, + { + "epoch": 0.6178774928774928, + "grad_norm": 0.628559947013855, + "learning_rate": 0.0001885646096213668, + "loss": 1.1089, + "step": 3470 + }, + { + "epoch": 0.6180555555555556, + "grad_norm": 0.503545880317688, + "learning_rate": 0.0001885581088757958, + "loss": 1.2311, + "step": 3471 + }, + { + "epoch": 0.6182336182336182, + "grad_norm": 0.6172101497650146, + "learning_rate": 0.00018855160639511264, + "loss": 1.2651, + "step": 3472 + }, + { + "epoch": 0.6184116809116809, + "grad_norm": 0.49572527408599854, + "learning_rate": 0.00018854510217944465, + "loss": 1.1026, + "step": 3473 + }, + { + "epoch": 0.6185897435897436, + "grad_norm": 0.5373549461364746, + "learning_rate": 0.00018853859622891938, + "loss": 1.2562, + "step": 3474 + }, + { + "epoch": 0.6187678062678063, + "grad_norm": 0.5272396206855774, + "learning_rate": 0.0001885320885436642, + "loss": 1.1763, + "step": 3475 + }, + { + "epoch": 0.6189458689458689, + "grad_norm": 0.46584269404411316, + "learning_rate": 0.00018852557912380665, + "loss": 1.1762, + "step": 3476 + }, + { + "epoch": 0.6191239316239316, + "grad_norm": 0.4798245131969452, + "learning_rate": 0.0001885190679694743, + "loss": 0.9229, + "step": 3477 + }, + { + "epoch": 0.6193019943019943, + "grad_norm": 0.5221366286277771, + "learning_rate": 0.0001885125550807947, + "loss": 1.1078, + "step": 3478 + }, + { + "epoch": 0.6194800569800569, + "grad_norm": 0.5051897168159485, + "learning_rate": 0.0001885060404578954, + "loss": 1.0055, + "step": 3479 + }, + { + "epoch": 0.6196581196581197, + "grad_norm": 0.492662250995636, + "learning_rate": 0.00018849952410090413, + "loss": 1.1172, + "step": 3480 + }, + { + "epoch": 0.6198361823361823, + "grad_norm": 0.4906775951385498, + "learning_rate": 0.00018849300600994853, + "loss": 1.1223, + "step": 3481 + }, + { + "epoch": 0.6200142450142451, + "grad_norm": 0.5032641291618347, + "learning_rate": 0.0001884864861851563, + "loss": 0.9541, + "step": 3482 + }, + { + "epoch": 0.6201923076923077, + "grad_norm": 0.5262296795845032, + "learning_rate": 0.00018847996462665521, + "loss": 1.021, + "step": 3483 + }, + { + "epoch": 0.6203703703703703, + "grad_norm": 0.5253522992134094, + "learning_rate": 0.00018847344133457295, + "loss": 0.9075, + "step": 3484 + }, + { + "epoch": 0.6205484330484331, + "grad_norm": 0.4204299747943878, + "learning_rate": 0.00018846691630903744, + "loss": 0.895, + "step": 3485 + }, + { + "epoch": 0.6207264957264957, + "grad_norm": 0.557604193687439, + "learning_rate": 0.0001884603895501765, + "loss": 1.1758, + "step": 3486 + }, + { + "epoch": 0.6209045584045584, + "grad_norm": 0.5981321930885315, + "learning_rate": 0.00018845386105811795, + "loss": 1.1087, + "step": 3487 + }, + { + "epoch": 0.6210826210826211, + "grad_norm": 0.5285581946372986, + "learning_rate": 0.00018844733083298975, + "loss": 1.0692, + "step": 3488 + }, + { + "epoch": 0.6212606837606838, + "grad_norm": 0.5403170585632324, + "learning_rate": 0.00018844079887491986, + "loss": 1.1998, + "step": 3489 + }, + { + "epoch": 0.6214387464387464, + "grad_norm": 0.5471615791320801, + "learning_rate": 0.0001884342651840362, + "loss": 0.9556, + "step": 3490 + }, + { + "epoch": 0.6216168091168092, + "grad_norm": 0.6126871705055237, + "learning_rate": 0.00018842772976046686, + "loss": 1.2629, + "step": 3491 + }, + { + "epoch": 0.6217948717948718, + "grad_norm": 0.45669353008270264, + "learning_rate": 0.00018842119260433982, + "loss": 1.0203, + "step": 3492 + }, + { + "epoch": 0.6219729344729344, + "grad_norm": 0.4998520612716675, + "learning_rate": 0.0001884146537157832, + "loss": 1.0271, + "step": 3493 + }, + { + "epoch": 0.6221509971509972, + "grad_norm": 0.5820242166519165, + "learning_rate": 0.00018840811309492507, + "loss": 1.0321, + "step": 3494 + }, + { + "epoch": 0.6223290598290598, + "grad_norm": 0.581676185131073, + "learning_rate": 0.00018840157074189367, + "loss": 0.9219, + "step": 3495 + }, + { + "epoch": 0.6225071225071225, + "grad_norm": 0.6044120788574219, + "learning_rate": 0.0001883950266568171, + "loss": 1.1621, + "step": 3496 + }, + { + "epoch": 0.6226851851851852, + "grad_norm": 0.5448858737945557, + "learning_rate": 0.0001883884808398236, + "loss": 1.0686, + "step": 3497 + }, + { + "epoch": 0.6228632478632479, + "grad_norm": 0.4921551048755646, + "learning_rate": 0.00018838193329104143, + "loss": 1.2259, + "step": 3498 + }, + { + "epoch": 0.6230413105413105, + "grad_norm": 0.5374335646629333, + "learning_rate": 0.00018837538401059888, + "loss": 1.2608, + "step": 3499 + }, + { + "epoch": 0.6232193732193733, + "grad_norm": 0.5123008489608765, + "learning_rate": 0.0001883688329986243, + "loss": 0.8682, + "step": 3500 + }, + { + "epoch": 0.6233974358974359, + "grad_norm": 0.566145122051239, + "learning_rate": 0.00018836228025524595, + "loss": 1.1807, + "step": 3501 + }, + { + "epoch": 0.6235754985754985, + "grad_norm": 0.6658587455749512, + "learning_rate": 0.00018835572578059233, + "loss": 1.1641, + "step": 3502 + }, + { + "epoch": 0.6237535612535613, + "grad_norm": 0.4992465078830719, + "learning_rate": 0.00018834916957479177, + "loss": 0.9125, + "step": 3503 + }, + { + "epoch": 0.6239316239316239, + "grad_norm": 0.5081812739372253, + "learning_rate": 0.00018834261163797278, + "loss": 1.0939, + "step": 3504 + }, + { + "epoch": 0.6241096866096866, + "grad_norm": 0.5168607234954834, + "learning_rate": 0.0001883360519702638, + "loss": 1.2382, + "step": 3505 + }, + { + "epoch": 0.6242877492877493, + "grad_norm": 0.5517697334289551, + "learning_rate": 0.00018832949057179344, + "loss": 1.206, + "step": 3506 + }, + { + "epoch": 0.624465811965812, + "grad_norm": 0.4505497217178345, + "learning_rate": 0.00018832292744269013, + "loss": 0.8485, + "step": 3507 + }, + { + "epoch": 0.6246438746438746, + "grad_norm": 0.5230690240859985, + "learning_rate": 0.0001883163625830826, + "loss": 1.1701, + "step": 3508 + }, + { + "epoch": 0.6248219373219374, + "grad_norm": 0.5062205195426941, + "learning_rate": 0.00018830979599309937, + "loss": 1.0602, + "step": 3509 + }, + { + "epoch": 0.625, + "grad_norm": 0.49922460317611694, + "learning_rate": 0.00018830322767286913, + "loss": 1.1937, + "step": 3510 + }, + { + "epoch": 0.6251780626780626, + "grad_norm": 0.4637366831302643, + "learning_rate": 0.0001882966576225206, + "loss": 1.038, + "step": 3511 + }, + { + "epoch": 0.6253561253561254, + "grad_norm": 0.5330080389976501, + "learning_rate": 0.00018829008584218246, + "loss": 0.9308, + "step": 3512 + }, + { + "epoch": 0.625534188034188, + "grad_norm": 0.5443428754806519, + "learning_rate": 0.0001882835123319835, + "loss": 1.0006, + "step": 3513 + }, + { + "epoch": 0.6257122507122507, + "grad_norm": 0.5534018874168396, + "learning_rate": 0.00018827693709205253, + "loss": 1.2383, + "step": 3514 + }, + { + "epoch": 0.6258903133903134, + "grad_norm": 0.49207547307014465, + "learning_rate": 0.00018827036012251832, + "loss": 0.9804, + "step": 3515 + }, + { + "epoch": 0.6260683760683761, + "grad_norm": 0.4900086224079132, + "learning_rate": 0.0001882637814235098, + "loss": 1.012, + "step": 3516 + }, + { + "epoch": 0.6262464387464387, + "grad_norm": 0.5267475247383118, + "learning_rate": 0.00018825720099515585, + "loss": 1.1104, + "step": 3517 + }, + { + "epoch": 0.6264245014245015, + "grad_norm": 0.5711902379989624, + "learning_rate": 0.00018825061883758534, + "loss": 1.0616, + "step": 3518 + }, + { + "epoch": 0.6266025641025641, + "grad_norm": 0.5007771849632263, + "learning_rate": 0.0001882440349509273, + "loss": 0.9578, + "step": 3519 + }, + { + "epoch": 0.6267806267806267, + "grad_norm": 0.5657192468643188, + "learning_rate": 0.00018823744933531075, + "loss": 1.2768, + "step": 3520 + }, + { + "epoch": 0.6269586894586895, + "grad_norm": 0.6077173352241516, + "learning_rate": 0.00018823086199086462, + "loss": 1.147, + "step": 3521 + }, + { + "epoch": 0.6271367521367521, + "grad_norm": 0.5114718079566956, + "learning_rate": 0.000188224272917718, + "loss": 1.1176, + "step": 3522 + }, + { + "epoch": 0.6273148148148148, + "grad_norm": 0.4831676185131073, + "learning_rate": 0.0001882176821160001, + "loss": 0.8021, + "step": 3523 + }, + { + "epoch": 0.6274928774928775, + "grad_norm": 0.6327390670776367, + "learning_rate": 0.00018821108958583994, + "loss": 0.9449, + "step": 3524 + }, + { + "epoch": 0.6276709401709402, + "grad_norm": 0.5541796684265137, + "learning_rate": 0.00018820449532736672, + "loss": 1.2018, + "step": 3525 + }, + { + "epoch": 0.6278490028490028, + "grad_norm": 0.5224639773368835, + "learning_rate": 0.00018819789934070968, + "loss": 1.0138, + "step": 3526 + }, + { + "epoch": 0.6280270655270656, + "grad_norm": 0.49359360337257385, + "learning_rate": 0.00018819130162599798, + "loss": 1.0768, + "step": 3527 + }, + { + "epoch": 0.6282051282051282, + "grad_norm": 0.5525050759315491, + "learning_rate": 0.00018818470218336092, + "loss": 1.0883, + "step": 3528 + }, + { + "epoch": 0.6283831908831908, + "grad_norm": 0.5563427209854126, + "learning_rate": 0.00018817810101292787, + "loss": 1.1491, + "step": 3529 + }, + { + "epoch": 0.6285612535612536, + "grad_norm": 0.49363306164741516, + "learning_rate": 0.00018817149811482803, + "loss": 1.1409, + "step": 3530 + }, + { + "epoch": 0.6287393162393162, + "grad_norm": 0.5102340579032898, + "learning_rate": 0.00018816489348919086, + "loss": 1.1914, + "step": 3531 + }, + { + "epoch": 0.6289173789173789, + "grad_norm": 0.5173332691192627, + "learning_rate": 0.00018815828713614576, + "loss": 0.9308, + "step": 3532 + }, + { + "epoch": 0.6290954415954416, + "grad_norm": 0.5093010067939758, + "learning_rate": 0.00018815167905582216, + "loss": 0.9429, + "step": 3533 + }, + { + "epoch": 0.6292735042735043, + "grad_norm": 0.5453153848648071, + "learning_rate": 0.00018814506924834954, + "loss": 1.0147, + "step": 3534 + }, + { + "epoch": 0.6294515669515669, + "grad_norm": 0.5850773453712463, + "learning_rate": 0.00018813845771385737, + "loss": 1.3372, + "step": 3535 + }, + { + "epoch": 0.6296296296296297, + "grad_norm": 0.5095621943473816, + "learning_rate": 0.00018813184445247525, + "loss": 1.0515, + "step": 3536 + }, + { + "epoch": 0.6298076923076923, + "grad_norm": 0.6216054558753967, + "learning_rate": 0.00018812522946433266, + "loss": 0.8703, + "step": 3537 + }, + { + "epoch": 0.6299857549857549, + "grad_norm": 0.4945531189441681, + "learning_rate": 0.00018811861274955932, + "loss": 1.1485, + "step": 3538 + }, + { + "epoch": 0.6301638176638177, + "grad_norm": 0.47882601618766785, + "learning_rate": 0.00018811199430828477, + "loss": 1.1107, + "step": 3539 + }, + { + "epoch": 0.6303418803418803, + "grad_norm": 0.5005326867103577, + "learning_rate": 0.00018810537414063876, + "loss": 1.0237, + "step": 3540 + }, + { + "epoch": 0.6305199430199431, + "grad_norm": 0.5382370352745056, + "learning_rate": 0.00018809875224675093, + "loss": 0.9965, + "step": 3541 + }, + { + "epoch": 0.6306980056980057, + "grad_norm": 0.47002625465393066, + "learning_rate": 0.0001880921286267511, + "loss": 1.065, + "step": 3542 + }, + { + "epoch": 0.6308760683760684, + "grad_norm": 0.4519105851650238, + "learning_rate": 0.00018808550328076897, + "loss": 0.9312, + "step": 3543 + }, + { + "epoch": 0.6310541310541311, + "grad_norm": 0.45360881090164185, + "learning_rate": 0.0001880788762089344, + "loss": 1.0739, + "step": 3544 + }, + { + "epoch": 0.6312321937321937, + "grad_norm": 0.5578218698501587, + "learning_rate": 0.00018807224741137723, + "loss": 1.2478, + "step": 3545 + }, + { + "epoch": 0.6314102564102564, + "grad_norm": 0.4838615655899048, + "learning_rate": 0.0001880656168882273, + "loss": 1.0221, + "step": 3546 + }, + { + "epoch": 0.6315883190883191, + "grad_norm": 0.5733556747436523, + "learning_rate": 0.0001880589846396146, + "loss": 1.1249, + "step": 3547 + }, + { + "epoch": 0.6317663817663818, + "grad_norm": 0.4939686954021454, + "learning_rate": 0.00018805235066566894, + "loss": 0.8559, + "step": 3548 + }, + { + "epoch": 0.6319444444444444, + "grad_norm": 0.5072234869003296, + "learning_rate": 0.00018804571496652044, + "loss": 1.0842, + "step": 3549 + }, + { + "epoch": 0.6321225071225072, + "grad_norm": 0.4640493392944336, + "learning_rate": 0.00018803907754229903, + "loss": 1.0728, + "step": 3550 + }, + { + "epoch": 0.6323005698005698, + "grad_norm": 0.5314788818359375, + "learning_rate": 0.00018803243839313481, + "loss": 1.0752, + "step": 3551 + }, + { + "epoch": 0.6324786324786325, + "grad_norm": 0.5511462092399597, + "learning_rate": 0.0001880257975191578, + "loss": 1.0238, + "step": 3552 + }, + { + "epoch": 0.6326566951566952, + "grad_norm": 0.4980711042881012, + "learning_rate": 0.00018801915492049816, + "loss": 1.0981, + "step": 3553 + }, + { + "epoch": 0.6328347578347578, + "grad_norm": 0.7746123671531677, + "learning_rate": 0.00018801251059728604, + "loss": 1.0968, + "step": 3554 + }, + { + "epoch": 0.6330128205128205, + "grad_norm": 0.5006106495857239, + "learning_rate": 0.00018800586454965155, + "loss": 1.1802, + "step": 3555 + }, + { + "epoch": 0.6331908831908832, + "grad_norm": 0.49427780508995056, + "learning_rate": 0.000187999216777725, + "loss": 1.1257, + "step": 3556 + }, + { + "epoch": 0.6333689458689459, + "grad_norm": 0.5484146475791931, + "learning_rate": 0.00018799256728163662, + "loss": 1.1344, + "step": 3557 + }, + { + "epoch": 0.6335470085470085, + "grad_norm": 0.5007877349853516, + "learning_rate": 0.00018798591606151662, + "loss": 1.1328, + "step": 3558 + }, + { + "epoch": 0.6337250712250713, + "grad_norm": 0.5068148970603943, + "learning_rate": 0.00018797926311749544, + "loss": 0.976, + "step": 3559 + }, + { + "epoch": 0.6339031339031339, + "grad_norm": 0.44936859607696533, + "learning_rate": 0.00018797260844970334, + "loss": 0.9735, + "step": 3560 + }, + { + "epoch": 0.6340811965811965, + "grad_norm": 0.4592931866645813, + "learning_rate": 0.0001879659520582707, + "loss": 1.1306, + "step": 3561 + }, + { + "epoch": 0.6342592592592593, + "grad_norm": 0.4664020836353302, + "learning_rate": 0.00018795929394332795, + "loss": 1.0577, + "step": 3562 + }, + { + "epoch": 0.6344373219373219, + "grad_norm": 0.5638116002082825, + "learning_rate": 0.00018795263410500556, + "loss": 1.1747, + "step": 3563 + }, + { + "epoch": 0.6346153846153846, + "grad_norm": 0.524736225605011, + "learning_rate": 0.00018794597254343401, + "loss": 0.8964, + "step": 3564 + }, + { + "epoch": 0.6347934472934473, + "grad_norm": 0.4645404517650604, + "learning_rate": 0.00018793930925874386, + "loss": 0.8673, + "step": 3565 + }, + { + "epoch": 0.63497150997151, + "grad_norm": 0.4800064265727997, + "learning_rate": 0.00018793264425106558, + "loss": 1.0334, + "step": 3566 + }, + { + "epoch": 0.6351495726495726, + "grad_norm": 0.6202501058578491, + "learning_rate": 0.0001879259775205298, + "loss": 1.1061, + "step": 3567 + }, + { + "epoch": 0.6353276353276354, + "grad_norm": 0.503383457660675, + "learning_rate": 0.00018791930906726718, + "loss": 0.8545, + "step": 3568 + }, + { + "epoch": 0.635505698005698, + "grad_norm": 0.5256780982017517, + "learning_rate": 0.00018791263889140832, + "loss": 1.0785, + "step": 3569 + }, + { + "epoch": 0.6356837606837606, + "grad_norm": 0.47562023997306824, + "learning_rate": 0.00018790596699308392, + "loss": 1.0041, + "step": 3570 + }, + { + "epoch": 0.6358618233618234, + "grad_norm": 0.5103238224983215, + "learning_rate": 0.00018789929337242469, + "loss": 1.1488, + "step": 3571 + }, + { + "epoch": 0.636039886039886, + "grad_norm": 0.5023695826530457, + "learning_rate": 0.0001878926180295614, + "loss": 1.0696, + "step": 3572 + }, + { + "epoch": 0.6362179487179487, + "grad_norm": 0.5302290916442871, + "learning_rate": 0.00018788594096462487, + "loss": 1.0554, + "step": 3573 + }, + { + "epoch": 0.6363960113960114, + "grad_norm": 0.4798361361026764, + "learning_rate": 0.00018787926217774588, + "loss": 0.8872, + "step": 3574 + }, + { + "epoch": 0.6365740740740741, + "grad_norm": 0.5529209971427917, + "learning_rate": 0.00018787258166905527, + "loss": 1.0976, + "step": 3575 + }, + { + "epoch": 0.6367521367521367, + "grad_norm": 0.49757125973701477, + "learning_rate": 0.00018786589943868402, + "loss": 1.0049, + "step": 3576 + }, + { + "epoch": 0.6369301994301995, + "grad_norm": 0.5497848391532898, + "learning_rate": 0.00018785921548676295, + "loss": 1.2272, + "step": 3577 + }, + { + "epoch": 0.6371082621082621, + "grad_norm": 0.5061752200126648, + "learning_rate": 0.0001878525298134231, + "loss": 1.0307, + "step": 3578 + }, + { + "epoch": 0.6372863247863247, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00018784584241879538, + "loss": 1.1064, + "step": 3579 + }, + { + "epoch": 0.6374643874643875, + "grad_norm": 0.48312774300575256, + "learning_rate": 0.0001878391533030109, + "loss": 1.078, + "step": 3580 + }, + { + "epoch": 0.6376424501424501, + "grad_norm": 0.5059898495674133, + "learning_rate": 0.00018783246246620067, + "loss": 1.0922, + "step": 3581 + }, + { + "epoch": 0.6378205128205128, + "grad_norm": 0.5144124031066895, + "learning_rate": 0.00018782576990849581, + "loss": 1.0909, + "step": 3582 + }, + { + "epoch": 0.6379985754985755, + "grad_norm": 0.5535032153129578, + "learning_rate": 0.0001878190756300274, + "loss": 1.2579, + "step": 3583 + }, + { + "epoch": 0.6381766381766382, + "grad_norm": 0.49145692586898804, + "learning_rate": 0.00018781237963092667, + "loss": 1.0823, + "step": 3584 + }, + { + "epoch": 0.6383547008547008, + "grad_norm": 0.5245576500892639, + "learning_rate": 0.00018780568191132472, + "loss": 0.9595, + "step": 3585 + }, + { + "epoch": 0.6385327635327636, + "grad_norm": 0.5026637315750122, + "learning_rate": 0.00018779898247135287, + "loss": 1.153, + "step": 3586 + }, + { + "epoch": 0.6387108262108262, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018779228131114234, + "loss": 1.0661, + "step": 3587 + }, + { + "epoch": 0.6388888888888888, + "grad_norm": 0.517387330532074, + "learning_rate": 0.00018778557843082444, + "loss": 1.0113, + "step": 3588 + }, + { + "epoch": 0.6390669515669516, + "grad_norm": 0.5149948000907898, + "learning_rate": 0.00018777887383053047, + "loss": 0.9483, + "step": 3589 + }, + { + "epoch": 0.6392450142450142, + "grad_norm": 0.4854544997215271, + "learning_rate": 0.00018777216751039185, + "loss": 1.22, + "step": 3590 + }, + { + "epoch": 0.6394230769230769, + "grad_norm": 0.5317271947860718, + "learning_rate": 0.0001877654594705399, + "loss": 1.2483, + "step": 3591 + }, + { + "epoch": 0.6396011396011396, + "grad_norm": 0.4554755687713623, + "learning_rate": 0.0001877587497111061, + "loss": 0.9864, + "step": 3592 + }, + { + "epoch": 0.6397792022792023, + "grad_norm": 0.4833736717700958, + "learning_rate": 0.0001877520382322219, + "loss": 0.8895, + "step": 3593 + }, + { + "epoch": 0.6399572649572649, + "grad_norm": 0.5018072724342346, + "learning_rate": 0.00018774532503401878, + "loss": 1.2523, + "step": 3594 + }, + { + "epoch": 0.6401353276353277, + "grad_norm": 0.4478762447834015, + "learning_rate": 0.00018773861011662832, + "loss": 0.8833, + "step": 3595 + }, + { + "epoch": 0.6403133903133903, + "grad_norm": 0.5686985850334167, + "learning_rate": 0.00018773189348018205, + "loss": 0.9934, + "step": 3596 + }, + { + "epoch": 0.6404914529914529, + "grad_norm": 0.5144175291061401, + "learning_rate": 0.00018772517512481157, + "loss": 0.8149, + "step": 3597 + }, + { + "epoch": 0.6406695156695157, + "grad_norm": 0.5359936356544495, + "learning_rate": 0.00018771845505064852, + "loss": 1.1822, + "step": 3598 + }, + { + "epoch": 0.6408475783475783, + "grad_norm": 0.532573938369751, + "learning_rate": 0.00018771173325782457, + "loss": 1.0361, + "step": 3599 + }, + { + "epoch": 0.6410256410256411, + "grad_norm": 0.46121537685394287, + "learning_rate": 0.00018770500974647138, + "loss": 1.0792, + "step": 3600 + }, + { + "epoch": 0.6412037037037037, + "grad_norm": 0.4804821312427521, + "learning_rate": 0.00018769828451672076, + "loss": 1.1119, + "step": 3601 + }, + { + "epoch": 0.6413817663817664, + "grad_norm": 0.4955114722251892, + "learning_rate": 0.00018769155756870443, + "loss": 0.9312, + "step": 3602 + }, + { + "epoch": 0.6415598290598291, + "grad_norm": 0.4987298250198364, + "learning_rate": 0.00018768482890255415, + "loss": 1.2326, + "step": 3603 + }, + { + "epoch": 0.6417378917378918, + "grad_norm": 0.47216179966926575, + "learning_rate": 0.0001876780985184018, + "loss": 1.0114, + "step": 3604 + }, + { + "epoch": 0.6419159544159544, + "grad_norm": 0.5891931653022766, + "learning_rate": 0.0001876713664163793, + "loss": 1.2963, + "step": 3605 + }, + { + "epoch": 0.6420940170940171, + "grad_norm": 0.4645081162452698, + "learning_rate": 0.00018766463259661846, + "loss": 1.0874, + "step": 3606 + }, + { + "epoch": 0.6422720797720798, + "grad_norm": 0.5275476574897766, + "learning_rate": 0.00018765789705925125, + "loss": 0.9453, + "step": 3607 + }, + { + "epoch": 0.6424501424501424, + "grad_norm": 0.5884957313537598, + "learning_rate": 0.00018765115980440964, + "loss": 1.0796, + "step": 3608 + }, + { + "epoch": 0.6426282051282052, + "grad_norm": 0.4843178987503052, + "learning_rate": 0.00018764442083222567, + "loss": 1.1657, + "step": 3609 + }, + { + "epoch": 0.6428062678062678, + "grad_norm": 0.5188381671905518, + "learning_rate": 0.00018763768014283126, + "loss": 1.1109, + "step": 3610 + }, + { + "epoch": 0.6429843304843305, + "grad_norm": 0.4101468324661255, + "learning_rate": 0.00018763093773635863, + "loss": 0.895, + "step": 3611 + }, + { + "epoch": 0.6431623931623932, + "grad_norm": 0.4552084505558014, + "learning_rate": 0.00018762419361293979, + "loss": 0.9418, + "step": 3612 + }, + { + "epoch": 0.6433404558404558, + "grad_norm": 0.5924661159515381, + "learning_rate": 0.0001876174477727069, + "loss": 1.2562, + "step": 3613 + }, + { + "epoch": 0.6435185185185185, + "grad_norm": 0.5072348713874817, + "learning_rate": 0.00018761070021579212, + "loss": 1.1501, + "step": 3614 + }, + { + "epoch": 0.6436965811965812, + "grad_norm": 0.5312697887420654, + "learning_rate": 0.0001876039509423277, + "loss": 1.0751, + "step": 3615 + }, + { + "epoch": 0.6438746438746439, + "grad_norm": 0.6046462059020996, + "learning_rate": 0.0001875971999524458, + "loss": 1.0927, + "step": 3616 + }, + { + "epoch": 0.6440527065527065, + "grad_norm": 0.4992375373840332, + "learning_rate": 0.00018759044724627876, + "loss": 0.96, + "step": 3617 + }, + { + "epoch": 0.6442307692307693, + "grad_norm": 0.4983134865760803, + "learning_rate": 0.00018758369282395886, + "loss": 1.0599, + "step": 3618 + }, + { + "epoch": 0.6444088319088319, + "grad_norm": 0.5655683279037476, + "learning_rate": 0.00018757693668561843, + "loss": 1.2372, + "step": 3619 + }, + { + "epoch": 0.6445868945868946, + "grad_norm": 0.4968827962875366, + "learning_rate": 0.00018757017883138985, + "loss": 1.1639, + "step": 3620 + }, + { + "epoch": 0.6447649572649573, + "grad_norm": 0.5831420421600342, + "learning_rate": 0.00018756341926140553, + "loss": 0.9002, + "step": 3621 + }, + { + "epoch": 0.64494301994302, + "grad_norm": 0.4828467071056366, + "learning_rate": 0.0001875566579757979, + "loss": 0.9201, + "step": 3622 + }, + { + "epoch": 0.6451210826210826, + "grad_norm": 0.5067087411880493, + "learning_rate": 0.00018754989497469943, + "loss": 0.9874, + "step": 3623 + }, + { + "epoch": 0.6452991452991453, + "grad_norm": 0.5182318091392517, + "learning_rate": 0.00018754313025824267, + "loss": 1.1291, + "step": 3624 + }, + { + "epoch": 0.645477207977208, + "grad_norm": 0.472200483083725, + "learning_rate": 0.0001875363638265601, + "loss": 1.0286, + "step": 3625 + }, + { + "epoch": 0.6456552706552706, + "grad_norm": 0.4597308039665222, + "learning_rate": 0.0001875295956797843, + "loss": 0.7517, + "step": 3626 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 0.5358221530914307, + "learning_rate": 0.00018752282581804798, + "loss": 1.2264, + "step": 3627 + }, + { + "epoch": 0.646011396011396, + "grad_norm": 0.5268992781639099, + "learning_rate": 0.00018751605424148363, + "loss": 1.0801, + "step": 3628 + }, + { + "epoch": 0.6461894586894587, + "grad_norm": 0.5917379260063171, + "learning_rate": 0.00018750928095022403, + "loss": 0.9538, + "step": 3629 + }, + { + "epoch": 0.6463675213675214, + "grad_norm": 0.44506707787513733, + "learning_rate": 0.00018750250594440183, + "loss": 0.9818, + "step": 3630 + }, + { + "epoch": 0.646545584045584, + "grad_norm": 0.5578880906105042, + "learning_rate": 0.00018749572922414982, + "loss": 0.9958, + "step": 3631 + }, + { + "epoch": 0.6467236467236467, + "grad_norm": 0.5155318975448608, + "learning_rate": 0.00018748895078960076, + "loss": 1.2888, + "step": 3632 + }, + { + "epoch": 0.6469017094017094, + "grad_norm": 0.5117297768592834, + "learning_rate": 0.0001874821706408874, + "loss": 1.0452, + "step": 3633 + }, + { + "epoch": 0.6470797720797721, + "grad_norm": 0.5169841647148132, + "learning_rate": 0.00018747538877814267, + "loss": 1.1649, + "step": 3634 + }, + { + "epoch": 0.6472578347578347, + "grad_norm": 0.5001181960105896, + "learning_rate": 0.00018746860520149942, + "loss": 1.1472, + "step": 3635 + }, + { + "epoch": 0.6474358974358975, + "grad_norm": 0.6289856433868408, + "learning_rate": 0.00018746181991109056, + "loss": 1.0351, + "step": 3636 + }, + { + "epoch": 0.6476139601139601, + "grad_norm": 0.5490612983703613, + "learning_rate": 0.00018745503290704897, + "loss": 0.8938, + "step": 3637 + }, + { + "epoch": 0.6477920227920227, + "grad_norm": 0.47378283739089966, + "learning_rate": 0.00018744824418950775, + "loss": 0.937, + "step": 3638 + }, + { + "epoch": 0.6479700854700855, + "grad_norm": 0.6079059839248657, + "learning_rate": 0.0001874414537585998, + "loss": 1.0486, + "step": 3639 + }, + { + "epoch": 0.6481481481481481, + "grad_norm": 0.5351769924163818, + "learning_rate": 0.00018743466161445823, + "loss": 1.0316, + "step": 3640 + }, + { + "epoch": 0.6483262108262108, + "grad_norm": 0.5516425967216492, + "learning_rate": 0.0001874278677572161, + "loss": 1.1552, + "step": 3641 + }, + { + "epoch": 0.6485042735042735, + "grad_norm": 0.5027523636817932, + "learning_rate": 0.0001874210721870065, + "loss": 1.0491, + "step": 3642 + }, + { + "epoch": 0.6486823361823362, + "grad_norm": 0.5596168041229248, + "learning_rate": 0.00018741427490396258, + "loss": 1.0256, + "step": 3643 + }, + { + "epoch": 0.6488603988603988, + "grad_norm": 0.5601046681404114, + "learning_rate": 0.00018740747590821751, + "loss": 1.1604, + "step": 3644 + }, + { + "epoch": 0.6490384615384616, + "grad_norm": 0.49749523401260376, + "learning_rate": 0.0001874006751999046, + "loss": 1.0532, + "step": 3645 + }, + { + "epoch": 0.6492165242165242, + "grad_norm": 0.6226113438606262, + "learning_rate": 0.00018739387277915697, + "loss": 1.1402, + "step": 3646 + }, + { + "epoch": 0.6493945868945868, + "grad_norm": 0.6142009496688843, + "learning_rate": 0.00018738706864610794, + "loss": 1.2437, + "step": 3647 + }, + { + "epoch": 0.6495726495726496, + "grad_norm": 0.48814916610717773, + "learning_rate": 0.00018738026280089084, + "loss": 0.8429, + "step": 3648 + }, + { + "epoch": 0.6497507122507122, + "grad_norm": 0.5717982053756714, + "learning_rate": 0.00018737345524363902, + "loss": 1.1095, + "step": 3649 + }, + { + "epoch": 0.6499287749287749, + "grad_norm": 0.5150009989738464, + "learning_rate": 0.00018736664597448582, + "loss": 1.199, + "step": 3650 + }, + { + "epoch": 0.6501068376068376, + "grad_norm": 0.58461594581604, + "learning_rate": 0.00018735983499356472, + "loss": 1.0704, + "step": 3651 + }, + { + "epoch": 0.6502849002849003, + "grad_norm": 0.5108643770217896, + "learning_rate": 0.0001873530223010091, + "loss": 1.2039, + "step": 3652 + }, + { + "epoch": 0.6504629629629629, + "grad_norm": 0.513306736946106, + "learning_rate": 0.00018734620789695247, + "loss": 1.1448, + "step": 3653 + }, + { + "epoch": 0.6506410256410257, + "grad_norm": 0.5139986872673035, + "learning_rate": 0.00018733939178152835, + "loss": 1.0023, + "step": 3654 + }, + { + "epoch": 0.6508190883190883, + "grad_norm": 0.5187703967094421, + "learning_rate": 0.00018733257395487027, + "loss": 1.1304, + "step": 3655 + }, + { + "epoch": 0.6509971509971509, + "grad_norm": 0.5470501184463501, + "learning_rate": 0.00018732575441711183, + "loss": 1.0272, + "step": 3656 + }, + { + "epoch": 0.6511752136752137, + "grad_norm": 0.537309467792511, + "learning_rate": 0.00018731893316838665, + "loss": 1.0806, + "step": 3657 + }, + { + "epoch": 0.6513532763532763, + "grad_norm": 0.5187864899635315, + "learning_rate": 0.00018731211020882836, + "loss": 1.0154, + "step": 3658 + }, + { + "epoch": 0.6515313390313391, + "grad_norm": 0.48373252153396606, + "learning_rate": 0.00018730528553857062, + "loss": 1.0135, + "step": 3659 + }, + { + "epoch": 0.6517094017094017, + "grad_norm": 0.5645000338554382, + "learning_rate": 0.00018729845915774716, + "loss": 0.8924, + "step": 3660 + }, + { + "epoch": 0.6518874643874644, + "grad_norm": 0.5722129940986633, + "learning_rate": 0.00018729163106649178, + "loss": 1.2416, + "step": 3661 + }, + { + "epoch": 0.6520655270655271, + "grad_norm": 0.5904877185821533, + "learning_rate": 0.00018728480126493823, + "loss": 0.9792, + "step": 3662 + }, + { + "epoch": 0.6522435897435898, + "grad_norm": 0.5224713087081909, + "learning_rate": 0.00018727796975322026, + "loss": 1.079, + "step": 3663 + }, + { + "epoch": 0.6524216524216524, + "grad_norm": 0.5667217969894409, + "learning_rate": 0.00018727113653147184, + "loss": 1.1397, + "step": 3664 + }, + { + "epoch": 0.6525997150997151, + "grad_norm": 0.5274622440338135, + "learning_rate": 0.00018726430159982677, + "loss": 1.0569, + "step": 3665 + }, + { + "epoch": 0.6527777777777778, + "grad_norm": 0.5745310187339783, + "learning_rate": 0.00018725746495841896, + "loss": 1.2129, + "step": 3666 + }, + { + "epoch": 0.6529558404558404, + "grad_norm": 0.6123398542404175, + "learning_rate": 0.0001872506266073824, + "loss": 1.186, + "step": 3667 + }, + { + "epoch": 0.6531339031339032, + "grad_norm": 0.4983387291431427, + "learning_rate": 0.00018724378654685106, + "loss": 1.1957, + "step": 3668 + }, + { + "epoch": 0.6533119658119658, + "grad_norm": 0.5584192276000977, + "learning_rate": 0.00018723694477695897, + "loss": 1.0939, + "step": 3669 + }, + { + "epoch": 0.6534900284900285, + "grad_norm": 0.5318745374679565, + "learning_rate": 0.00018723010129784016, + "loss": 1.1869, + "step": 3670 + }, + { + "epoch": 0.6536680911680912, + "grad_norm": 0.4607617259025574, + "learning_rate": 0.0001872232561096287, + "loss": 0.8447, + "step": 3671 + }, + { + "epoch": 0.6538461538461539, + "grad_norm": 0.5312213897705078, + "learning_rate": 0.00018721640921245874, + "loss": 1.0623, + "step": 3672 + }, + { + "epoch": 0.6540242165242165, + "grad_norm": 0.5099136233329773, + "learning_rate": 0.0001872095606064644, + "loss": 0.7174, + "step": 3673 + }, + { + "epoch": 0.6542022792022792, + "grad_norm": 0.6894404888153076, + "learning_rate": 0.0001872027102917799, + "loss": 1.0251, + "step": 3674 + }, + { + "epoch": 0.6543803418803419, + "grad_norm": 0.5758535861968994, + "learning_rate": 0.00018719585826853944, + "loss": 1.1655, + "step": 3675 + }, + { + "epoch": 0.6545584045584045, + "grad_norm": 0.521824061870575, + "learning_rate": 0.0001871890045368773, + "loss": 1.1653, + "step": 3676 + }, + { + "epoch": 0.6547364672364673, + "grad_norm": 0.5370712280273438, + "learning_rate": 0.00018718214909692771, + "loss": 1.3152, + "step": 3677 + }, + { + "epoch": 0.6549145299145299, + "grad_norm": 0.4459827244281769, + "learning_rate": 0.000187175291948825, + "loss": 1.0953, + "step": 3678 + }, + { + "epoch": 0.6550925925925926, + "grad_norm": 0.44131460785865784, + "learning_rate": 0.00018716843309270353, + "loss": 0.8568, + "step": 3679 + }, + { + "epoch": 0.6552706552706553, + "grad_norm": 0.5529624819755554, + "learning_rate": 0.00018716157252869772, + "loss": 1.2085, + "step": 3680 + }, + { + "epoch": 0.655448717948718, + "grad_norm": 0.44604751467704773, + "learning_rate": 0.00018715471025694194, + "loss": 0.9605, + "step": 3681 + }, + { + "epoch": 0.6556267806267806, + "grad_norm": 0.4662449359893799, + "learning_rate": 0.0001871478462775707, + "loss": 1.2092, + "step": 3682 + }, + { + "epoch": 0.6558048433048433, + "grad_norm": 0.42632922530174255, + "learning_rate": 0.0001871409805907184, + "loss": 0.9141, + "step": 3683 + }, + { + "epoch": 0.655982905982906, + "grad_norm": 0.534009575843811, + "learning_rate": 0.00018713411319651958, + "loss": 1.0147, + "step": 3684 + }, + { + "epoch": 0.6561609686609686, + "grad_norm": 0.5433241724967957, + "learning_rate": 0.00018712724409510888, + "loss": 1.1998, + "step": 3685 + }, + { + "epoch": 0.6563390313390314, + "grad_norm": 0.4771319627761841, + "learning_rate": 0.0001871203732866208, + "loss": 1.0384, + "step": 3686 + }, + { + "epoch": 0.656517094017094, + "grad_norm": 0.507641077041626, + "learning_rate": 0.00018711350077119, + "loss": 0.9608, + "step": 3687 + }, + { + "epoch": 0.6566951566951567, + "grad_norm": 0.5069413185119629, + "learning_rate": 0.00018710662654895108, + "loss": 1.055, + "step": 3688 + }, + { + "epoch": 0.6568732193732194, + "grad_norm": 0.512340247631073, + "learning_rate": 0.00018709975062003876, + "loss": 0.9506, + "step": 3689 + }, + { + "epoch": 0.657051282051282, + "grad_norm": 0.5156390070915222, + "learning_rate": 0.00018709287298458778, + "loss": 1.0089, + "step": 3690 + }, + { + "epoch": 0.6572293447293447, + "grad_norm": 0.5101696252822876, + "learning_rate": 0.0001870859936427329, + "loss": 1.0441, + "step": 3691 + }, + { + "epoch": 0.6574074074074074, + "grad_norm": 0.4394689202308655, + "learning_rate": 0.00018707911259460884, + "loss": 0.9124, + "step": 3692 + }, + { + "epoch": 0.6575854700854701, + "grad_norm": 0.4842554032802582, + "learning_rate": 0.00018707222984035043, + "loss": 1.0051, + "step": 3693 + }, + { + "epoch": 0.6577635327635327, + "grad_norm": 0.6418108344078064, + "learning_rate": 0.00018706534538009262, + "loss": 1.1165, + "step": 3694 + }, + { + "epoch": 0.6579415954415955, + "grad_norm": 0.5596832036972046, + "learning_rate": 0.00018705845921397022, + "loss": 1.1127, + "step": 3695 + }, + { + "epoch": 0.6581196581196581, + "grad_norm": 0.6692909002304077, + "learning_rate": 0.00018705157134211813, + "loss": 1.2403, + "step": 3696 + }, + { + "epoch": 0.6582977207977208, + "grad_norm": 0.5046468377113342, + "learning_rate": 0.00018704468176467134, + "loss": 1.1016, + "step": 3697 + }, + { + "epoch": 0.6584757834757835, + "grad_norm": 0.6723586320877075, + "learning_rate": 0.00018703779048176485, + "loss": 1.1777, + "step": 3698 + }, + { + "epoch": 0.6586538461538461, + "grad_norm": 0.5269754528999329, + "learning_rate": 0.00018703089749353365, + "loss": 1.1441, + "step": 3699 + }, + { + "epoch": 0.6588319088319088, + "grad_norm": 0.5303323268890381, + "learning_rate": 0.0001870240028001128, + "loss": 1.07, + "step": 3700 + }, + { + "epoch": 0.6590099715099715, + "grad_norm": 0.4795511066913605, + "learning_rate": 0.00018701710640163738, + "loss": 1.0189, + "step": 3701 + }, + { + "epoch": 0.6591880341880342, + "grad_norm": 0.514659583568573, + "learning_rate": 0.00018701020829824255, + "loss": 1.0792, + "step": 3702 + }, + { + "epoch": 0.6593660968660968, + "grad_norm": 0.5407463312149048, + "learning_rate": 0.0001870033084900634, + "loss": 0.9346, + "step": 3703 + }, + { + "epoch": 0.6595441595441596, + "grad_norm": 0.5358424186706543, + "learning_rate": 0.0001869964069772352, + "loss": 1.1242, + "step": 3704 + }, + { + "epoch": 0.6597222222222222, + "grad_norm": 0.470825731754303, + "learning_rate": 0.00018698950375989307, + "loss": 0.9952, + "step": 3705 + }, + { + "epoch": 0.6599002849002849, + "grad_norm": 0.5711592435836792, + "learning_rate": 0.00018698259883817236, + "loss": 1.1678, + "step": 3706 + }, + { + "epoch": 0.6600783475783476, + "grad_norm": 0.5298995971679688, + "learning_rate": 0.00018697569221220832, + "loss": 0.869, + "step": 3707 + }, + { + "epoch": 0.6602564102564102, + "grad_norm": 0.5453875064849854, + "learning_rate": 0.00018696878388213626, + "loss": 0.9706, + "step": 3708 + }, + { + "epoch": 0.6604344729344729, + "grad_norm": 0.6219926476478577, + "learning_rate": 0.00018696187384809154, + "loss": 1.1902, + "step": 3709 + }, + { + "epoch": 0.6606125356125356, + "grad_norm": 0.5972491502761841, + "learning_rate": 0.00018695496211020953, + "loss": 1.2054, + "step": 3710 + }, + { + "epoch": 0.6607905982905983, + "grad_norm": 0.5048904418945312, + "learning_rate": 0.0001869480486686257, + "loss": 1.0405, + "step": 3711 + }, + { + "epoch": 0.6609686609686609, + "grad_norm": 0.5474200248718262, + "learning_rate": 0.00018694113352347546, + "loss": 1.09, + "step": 3712 + }, + { + "epoch": 0.6611467236467237, + "grad_norm": 0.5073318481445312, + "learning_rate": 0.00018693421667489432, + "loss": 1.0698, + "step": 3713 + }, + { + "epoch": 0.6613247863247863, + "grad_norm": 0.5693208575248718, + "learning_rate": 0.0001869272981230178, + "loss": 0.9664, + "step": 3714 + }, + { + "epoch": 0.6615028490028491, + "grad_norm": 0.5678503513336182, + "learning_rate": 0.00018692037786798143, + "loss": 1.0895, + "step": 3715 + }, + { + "epoch": 0.6616809116809117, + "grad_norm": 0.4950976073741913, + "learning_rate": 0.00018691345590992082, + "loss": 0.9584, + "step": 3716 + }, + { + "epoch": 0.6618589743589743, + "grad_norm": 0.4944666624069214, + "learning_rate": 0.0001869065322489716, + "loss": 0.8607, + "step": 3717 + }, + { + "epoch": 0.6620370370370371, + "grad_norm": 0.5197804570198059, + "learning_rate": 0.0001868996068852694, + "loss": 1.2335, + "step": 3718 + }, + { + "epoch": 0.6622150997150997, + "grad_norm": 0.6550365686416626, + "learning_rate": 0.00018689267981894994, + "loss": 1.0441, + "step": 3719 + }, + { + "epoch": 0.6623931623931624, + "grad_norm": 0.5331503748893738, + "learning_rate": 0.00018688575105014888, + "loss": 1.1696, + "step": 3720 + }, + { + "epoch": 0.6625712250712251, + "grad_norm": 0.47304239869117737, + "learning_rate": 0.00018687882057900207, + "loss": 0.9695, + "step": 3721 + }, + { + "epoch": 0.6627492877492878, + "grad_norm": 0.5653772354125977, + "learning_rate": 0.00018687188840564524, + "loss": 1.2082, + "step": 3722 + }, + { + "epoch": 0.6629273504273504, + "grad_norm": 0.5323491096496582, + "learning_rate": 0.00018686495453021417, + "loss": 0.9106, + "step": 3723 + }, + { + "epoch": 0.6631054131054132, + "grad_norm": 0.5612817406654358, + "learning_rate": 0.00018685801895284483, + "loss": 1.1302, + "step": 3724 + }, + { + "epoch": 0.6632834757834758, + "grad_norm": 0.4562164545059204, + "learning_rate": 0.000186851081673673, + "loss": 0.8886, + "step": 3725 + }, + { + "epoch": 0.6634615384615384, + "grad_norm": 0.5006430745124817, + "learning_rate": 0.00018684414269283463, + "loss": 0.9128, + "step": 3726 + }, + { + "epoch": 0.6636396011396012, + "grad_norm": 0.5305442810058594, + "learning_rate": 0.0001868372020104657, + "loss": 1.1766, + "step": 3727 + }, + { + "epoch": 0.6638176638176638, + "grad_norm": 0.6129274368286133, + "learning_rate": 0.0001868302596267022, + "loss": 1.04, + "step": 3728 + }, + { + "epoch": 0.6639957264957265, + "grad_norm": 0.5530399084091187, + "learning_rate": 0.00018682331554168013, + "loss": 1.4114, + "step": 3729 + }, + { + "epoch": 0.6641737891737892, + "grad_norm": 0.5397193431854248, + "learning_rate": 0.00018681636975553557, + "loss": 1.1945, + "step": 3730 + }, + { + "epoch": 0.6643518518518519, + "grad_norm": 0.5510205030441284, + "learning_rate": 0.00018680942226840456, + "loss": 1.0489, + "step": 3731 + }, + { + "epoch": 0.6645299145299145, + "grad_norm": 0.5519221425056458, + "learning_rate": 0.00018680247308042324, + "loss": 1.1633, + "step": 3732 + }, + { + "epoch": 0.6647079772079773, + "grad_norm": 0.4848768711090088, + "learning_rate": 0.00018679552219172784, + "loss": 0.8716, + "step": 3733 + }, + { + "epoch": 0.6648860398860399, + "grad_norm": 0.5490246415138245, + "learning_rate": 0.0001867885696024544, + "loss": 1.1347, + "step": 3734 + }, + { + "epoch": 0.6650641025641025, + "grad_norm": 0.5281458497047424, + "learning_rate": 0.00018678161531273928, + "loss": 1.0987, + "step": 3735 + }, + { + "epoch": 0.6652421652421653, + "grad_norm": 0.5313079953193665, + "learning_rate": 0.00018677465932271867, + "loss": 0.9705, + "step": 3736 + }, + { + "epoch": 0.6654202279202279, + "grad_norm": 0.5425750017166138, + "learning_rate": 0.0001867677016325289, + "loss": 1.1847, + "step": 3737 + }, + { + "epoch": 0.6655982905982906, + "grad_norm": 0.5796298980712891, + "learning_rate": 0.0001867607422423062, + "loss": 1.2639, + "step": 3738 + }, + { + "epoch": 0.6657763532763533, + "grad_norm": 0.49738675355911255, + "learning_rate": 0.00018675378115218702, + "loss": 1.0536, + "step": 3739 + }, + { + "epoch": 0.665954415954416, + "grad_norm": 0.665250301361084, + "learning_rate": 0.0001867468183623077, + "loss": 1.2836, + "step": 3740 + }, + { + "epoch": 0.6661324786324786, + "grad_norm": 0.5184717178344727, + "learning_rate": 0.00018673985387280469, + "loss": 1.0497, + "step": 3741 + }, + { + "epoch": 0.6663105413105413, + "grad_norm": 0.5129656791687012, + "learning_rate": 0.00018673288768381442, + "loss": 1.2041, + "step": 3742 + }, + { + "epoch": 0.666488603988604, + "grad_norm": 0.5308768153190613, + "learning_rate": 0.00018672591979547337, + "loss": 1.2092, + "step": 3743 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.5059141516685486, + "learning_rate": 0.00018671895020791812, + "loss": 1.1929, + "step": 3744 + }, + { + "epoch": 0.6668447293447294, + "grad_norm": 0.5237857103347778, + "learning_rate": 0.00018671197892128517, + "loss": 1.2538, + "step": 3745 + }, + { + "epoch": 0.667022792022792, + "grad_norm": 0.450000137090683, + "learning_rate": 0.0001867050059357111, + "loss": 0.7138, + "step": 3746 + }, + { + "epoch": 0.6672008547008547, + "grad_norm": 0.5413795709609985, + "learning_rate": 0.00018669803125133258, + "loss": 1.1383, + "step": 3747 + }, + { + "epoch": 0.6673789173789174, + "grad_norm": 0.4657825529575348, + "learning_rate": 0.00018669105486828622, + "loss": 1.0518, + "step": 3748 + }, + { + "epoch": 0.66755698005698, + "grad_norm": 0.6198551654815674, + "learning_rate": 0.00018668407678670875, + "loss": 1.2697, + "step": 3749 + }, + { + "epoch": 0.6677350427350427, + "grad_norm": 0.5112186074256897, + "learning_rate": 0.00018667709700673685, + "loss": 0.9907, + "step": 3750 + }, + { + "epoch": 0.6679131054131054, + "grad_norm": 0.5446593761444092, + "learning_rate": 0.00018667011552850728, + "loss": 1.0708, + "step": 3751 + }, + { + "epoch": 0.6680911680911681, + "grad_norm": 0.5673866271972656, + "learning_rate": 0.00018666313235215682, + "loss": 1.05, + "step": 3752 + }, + { + "epoch": 0.6682692307692307, + "grad_norm": 0.4821988046169281, + "learning_rate": 0.00018665614747782235, + "loss": 1.0543, + "step": 3753 + }, + { + "epoch": 0.6684472934472935, + "grad_norm": 0.5158842206001282, + "learning_rate": 0.00018664916090564067, + "loss": 1.0331, + "step": 3754 + }, + { + "epoch": 0.6686253561253561, + "grad_norm": 0.45486921072006226, + "learning_rate": 0.00018664217263574865, + "loss": 0.9262, + "step": 3755 + }, + { + "epoch": 0.6688034188034188, + "grad_norm": 0.46193036437034607, + "learning_rate": 0.00018663518266828327, + "loss": 0.9858, + "step": 3756 + }, + { + "epoch": 0.6689814814814815, + "grad_norm": 0.5144094824790955, + "learning_rate": 0.00018662819100338148, + "loss": 1.0302, + "step": 3757 + }, + { + "epoch": 0.6691595441595442, + "grad_norm": 0.5246134400367737, + "learning_rate": 0.0001866211976411802, + "loss": 1.064, + "step": 3758 + }, + { + "epoch": 0.6693376068376068, + "grad_norm": 0.4853166937828064, + "learning_rate": 0.0001866142025818165, + "loss": 0.9481, + "step": 3759 + }, + { + "epoch": 0.6695156695156695, + "grad_norm": 0.5029586553573608, + "learning_rate": 0.00018660720582542743, + "loss": 0.9443, + "step": 3760 + }, + { + "epoch": 0.6696937321937322, + "grad_norm": 0.5373172163963318, + "learning_rate": 0.0001866002073721501, + "loss": 1.1401, + "step": 3761 + }, + { + "epoch": 0.6698717948717948, + "grad_norm": 0.6236287951469421, + "learning_rate": 0.00018659320722212158, + "loss": 1.1255, + "step": 3762 + }, + { + "epoch": 0.6700498575498576, + "grad_norm": 0.5470684766769409, + "learning_rate": 0.00018658620537547903, + "loss": 1.0622, + "step": 3763 + }, + { + "epoch": 0.6702279202279202, + "grad_norm": 0.63177090883255, + "learning_rate": 0.00018657920183235964, + "loss": 0.9736, + "step": 3764 + }, + { + "epoch": 0.6704059829059829, + "grad_norm": 0.5456309914588928, + "learning_rate": 0.00018657219659290068, + "loss": 1.027, + "step": 3765 + }, + { + "epoch": 0.6705840455840456, + "grad_norm": 0.4816138744354248, + "learning_rate": 0.00018656518965723935, + "loss": 0.7801, + "step": 3766 + }, + { + "epoch": 0.6707621082621082, + "grad_norm": 0.4811640679836273, + "learning_rate": 0.00018655818102551294, + "loss": 1.0535, + "step": 3767 + }, + { + "epoch": 0.6709401709401709, + "grad_norm": 0.4677673280239105, + "learning_rate": 0.00018655117069785884, + "loss": 1.1043, + "step": 3768 + }, + { + "epoch": 0.6711182336182336, + "grad_norm": 0.5628635883331299, + "learning_rate": 0.0001865441586744143, + "loss": 1.0392, + "step": 3769 + }, + { + "epoch": 0.6712962962962963, + "grad_norm": 0.5484504103660583, + "learning_rate": 0.00018653714495531673, + "loss": 1.1533, + "step": 3770 + }, + { + "epoch": 0.6714743589743589, + "grad_norm": 0.5830571055412292, + "learning_rate": 0.0001865301295407036, + "loss": 1.2479, + "step": 3771 + }, + { + "epoch": 0.6716524216524217, + "grad_norm": 0.5516841411590576, + "learning_rate": 0.00018652311243071235, + "loss": 1.2152, + "step": 3772 + }, + { + "epoch": 0.6718304843304843, + "grad_norm": 0.6360766291618347, + "learning_rate": 0.0001865160936254804, + "loss": 1.0752, + "step": 3773 + }, + { + "epoch": 0.6720085470085471, + "grad_norm": 0.6038610935211182, + "learning_rate": 0.00018650907312514533, + "loss": 1.2425, + "step": 3774 + }, + { + "epoch": 0.6721866096866097, + "grad_norm": 0.49572908878326416, + "learning_rate": 0.0001865020509298447, + "loss": 1.0057, + "step": 3775 + }, + { + "epoch": 0.6723646723646723, + "grad_norm": 0.4551616311073303, + "learning_rate": 0.00018649502703971607, + "loss": 1.0763, + "step": 3776 + }, + { + "epoch": 0.6725427350427351, + "grad_norm": 0.6621482372283936, + "learning_rate": 0.00018648800145489706, + "loss": 1.0306, + "step": 3777 + }, + { + "epoch": 0.6727207977207977, + "grad_norm": 0.5523806810379028, + "learning_rate": 0.0001864809741755253, + "loss": 0.9906, + "step": 3778 + }, + { + "epoch": 0.6728988603988604, + "grad_norm": 0.5527048110961914, + "learning_rate": 0.00018647394520173856, + "loss": 1.0734, + "step": 3779 + }, + { + "epoch": 0.6730769230769231, + "grad_norm": 0.573573887348175, + "learning_rate": 0.00018646691453367444, + "loss": 1.1409, + "step": 3780 + }, + { + "epoch": 0.6732549857549858, + "grad_norm": 0.6273239254951477, + "learning_rate": 0.00018645988217147079, + "loss": 0.9682, + "step": 3781 + }, + { + "epoch": 0.6734330484330484, + "grad_norm": 0.4917762279510498, + "learning_rate": 0.00018645284811526534, + "loss": 0.9681, + "step": 3782 + }, + { + "epoch": 0.6736111111111112, + "grad_norm": 0.4901154339313507, + "learning_rate": 0.0001864458123651959, + "loss": 1.1828, + "step": 3783 + }, + { + "epoch": 0.6737891737891738, + "grad_norm": 0.6292546391487122, + "learning_rate": 0.00018643877492140036, + "loss": 1.1987, + "step": 3784 + }, + { + "epoch": 0.6739672364672364, + "grad_norm": 0.5334137678146362, + "learning_rate": 0.0001864317357840166, + "loss": 1.0347, + "step": 3785 + }, + { + "epoch": 0.6741452991452992, + "grad_norm": 0.6064338684082031, + "learning_rate": 0.0001864246949531825, + "loss": 1.4154, + "step": 3786 + }, + { + "epoch": 0.6743233618233618, + "grad_norm": 0.5442034602165222, + "learning_rate": 0.000186417652429036, + "loss": 1.2604, + "step": 3787 + }, + { + "epoch": 0.6745014245014245, + "grad_norm": 0.490858793258667, + "learning_rate": 0.00018641060821171518, + "loss": 1.1511, + "step": 3788 + }, + { + "epoch": 0.6746794871794872, + "grad_norm": 0.571116030216217, + "learning_rate": 0.00018640356230135798, + "loss": 1.1479, + "step": 3789 + }, + { + "epoch": 0.6748575498575499, + "grad_norm": 0.4857785105705261, + "learning_rate": 0.00018639651469810247, + "loss": 0.9, + "step": 3790 + }, + { + "epoch": 0.6750356125356125, + "grad_norm": 0.5320703983306885, + "learning_rate": 0.0001863894654020867, + "loss": 1.2284, + "step": 3791 + }, + { + "epoch": 0.6752136752136753, + "grad_norm": 0.5586925745010376, + "learning_rate": 0.0001863824144134488, + "loss": 1.1183, + "step": 3792 + }, + { + "epoch": 0.6753917378917379, + "grad_norm": 0.47740885615348816, + "learning_rate": 0.000186375361732327, + "loss": 1.1512, + "step": 3793 + }, + { + "epoch": 0.6755698005698005, + "grad_norm": 0.5867732167243958, + "learning_rate": 0.00018636830735885935, + "loss": 1.1903, + "step": 3794 + }, + { + "epoch": 0.6757478632478633, + "grad_norm": 0.5013887882232666, + "learning_rate": 0.0001863612512931842, + "loss": 0.8581, + "step": 3795 + }, + { + "epoch": 0.6759259259259259, + "grad_norm": 0.6026871204376221, + "learning_rate": 0.0001863541935354397, + "loss": 0.9581, + "step": 3796 + }, + { + "epoch": 0.6761039886039886, + "grad_norm": 0.5238468647003174, + "learning_rate": 0.00018634713408576415, + "loss": 1.0949, + "step": 3797 + }, + { + "epoch": 0.6762820512820513, + "grad_norm": 0.5128598213195801, + "learning_rate": 0.00018634007294429585, + "loss": 0.8992, + "step": 3798 + }, + { + "epoch": 0.676460113960114, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018633301011117324, + "loss": 1.0793, + "step": 3799 + }, + { + "epoch": 0.6766381766381766, + "grad_norm": 0.592566728591919, + "learning_rate": 0.00018632594558653457, + "loss": 1.3242, + "step": 3800 + }, + { + "epoch": 0.6768162393162394, + "grad_norm": 0.4953067898750305, + "learning_rate": 0.0001863188793705184, + "loss": 0.9925, + "step": 3801 + }, + { + "epoch": 0.676994301994302, + "grad_norm": 0.4989747107028961, + "learning_rate": 0.00018631181146326305, + "loss": 1.0677, + "step": 3802 + }, + { + "epoch": 0.6771723646723646, + "grad_norm": 0.5375261902809143, + "learning_rate": 0.00018630474186490705, + "loss": 1.0556, + "step": 3803 + }, + { + "epoch": 0.6773504273504274, + "grad_norm": 0.6512624025344849, + "learning_rate": 0.00018629767057558894, + "loss": 1.2041, + "step": 3804 + }, + { + "epoch": 0.67752849002849, + "grad_norm": 0.5428260564804077, + "learning_rate": 0.00018629059759544723, + "loss": 0.9645, + "step": 3805 + }, + { + "epoch": 0.6777065527065527, + "grad_norm": 0.5598662495613098, + "learning_rate": 0.00018628352292462052, + "loss": 1.1683, + "step": 3806 + }, + { + "epoch": 0.6778846153846154, + "grad_norm": 0.49351340532302856, + "learning_rate": 0.0001862764465632474, + "loss": 1.1622, + "step": 3807 + }, + { + "epoch": 0.6780626780626781, + "grad_norm": 0.4796701669692993, + "learning_rate": 0.00018626936851146657, + "loss": 1.0017, + "step": 3808 + }, + { + "epoch": 0.6782407407407407, + "grad_norm": 0.444533109664917, + "learning_rate": 0.00018626228876941664, + "loss": 0.9145, + "step": 3809 + }, + { + "epoch": 0.6784188034188035, + "grad_norm": 0.5197392702102661, + "learning_rate": 0.00018625520733723635, + "loss": 1.283, + "step": 3810 + }, + { + "epoch": 0.6785968660968661, + "grad_norm": 0.48785829544067383, + "learning_rate": 0.00018624812421506447, + "loss": 1.1084, + "step": 3811 + }, + { + "epoch": 0.6787749287749287, + "grad_norm": 0.5083680152893066, + "learning_rate": 0.00018624103940303974, + "loss": 0.9071, + "step": 3812 + }, + { + "epoch": 0.6789529914529915, + "grad_norm": 0.553819477558136, + "learning_rate": 0.00018623395290130103, + "loss": 0.9986, + "step": 3813 + }, + { + "epoch": 0.6791310541310541, + "grad_norm": 0.5347508788108826, + "learning_rate": 0.00018622686470998713, + "loss": 1.0148, + "step": 3814 + }, + { + "epoch": 0.6793091168091168, + "grad_norm": 0.5080769062042236, + "learning_rate": 0.00018621977482923693, + "loss": 1.0169, + "step": 3815 + }, + { + "epoch": 0.6794871794871795, + "grad_norm": 0.5444077849388123, + "learning_rate": 0.00018621268325918938, + "loss": 1.172, + "step": 3816 + }, + { + "epoch": 0.6796652421652422, + "grad_norm": 0.521946132183075, + "learning_rate": 0.00018620558999998335, + "loss": 1.0247, + "step": 3817 + }, + { + "epoch": 0.6798433048433048, + "grad_norm": 0.5257413983345032, + "learning_rate": 0.00018619849505175786, + "loss": 1.1574, + "step": 3818 + }, + { + "epoch": 0.6800213675213675, + "grad_norm": 0.5473007559776306, + "learning_rate": 0.00018619139841465193, + "loss": 1.1254, + "step": 3819 + }, + { + "epoch": 0.6801994301994302, + "grad_norm": 0.5479872226715088, + "learning_rate": 0.00018618430008880463, + "loss": 1.0196, + "step": 3820 + }, + { + "epoch": 0.6803774928774928, + "grad_norm": 0.5918973088264465, + "learning_rate": 0.00018617720007435497, + "loss": 1.082, + "step": 3821 + }, + { + "epoch": 0.6805555555555556, + "grad_norm": 0.5411791801452637, + "learning_rate": 0.0001861700983714421, + "loss": 0.7723, + "step": 3822 + }, + { + "epoch": 0.6807336182336182, + "grad_norm": 0.5466326475143433, + "learning_rate": 0.00018616299498020516, + "loss": 1.0979, + "step": 3823 + }, + { + "epoch": 0.6809116809116809, + "grad_norm": 0.5405182838439941, + "learning_rate": 0.00018615588990078332, + "loss": 0.8891, + "step": 3824 + }, + { + "epoch": 0.6810897435897436, + "grad_norm": 0.5415780544281006, + "learning_rate": 0.00018614878313331579, + "loss": 1.0927, + "step": 3825 + }, + { + "epoch": 0.6812678062678063, + "grad_norm": 0.5284909605979919, + "learning_rate": 0.00018614167467794182, + "loss": 1.0684, + "step": 3826 + }, + { + "epoch": 0.6814458689458689, + "grad_norm": 0.4873995780944824, + "learning_rate": 0.00018613456453480062, + "loss": 1.1653, + "step": 3827 + }, + { + "epoch": 0.6816239316239316, + "grad_norm": 0.5506551265716553, + "learning_rate": 0.0001861274527040316, + "loss": 0.9876, + "step": 3828 + }, + { + "epoch": 0.6818019943019943, + "grad_norm": 0.5031297206878662, + "learning_rate": 0.0001861203391857741, + "loss": 1.067, + "step": 3829 + }, + { + "epoch": 0.6819800569800569, + "grad_norm": 0.622346043586731, + "learning_rate": 0.0001861132239801674, + "loss": 1.1514, + "step": 3830 + }, + { + "epoch": 0.6821581196581197, + "grad_norm": 0.47706183791160583, + "learning_rate": 0.000186106107087351, + "loss": 0.9857, + "step": 3831 + }, + { + "epoch": 0.6823361823361823, + "grad_norm": 0.5082845091819763, + "learning_rate": 0.00018609898850746424, + "loss": 1.123, + "step": 3832 + }, + { + "epoch": 0.6825142450142451, + "grad_norm": 0.5119805932044983, + "learning_rate": 0.00018609186824064671, + "loss": 1.1386, + "step": 3833 + }, + { + "epoch": 0.6826923076923077, + "grad_norm": 0.5247541069984436, + "learning_rate": 0.00018608474628703788, + "loss": 0.9433, + "step": 3834 + }, + { + "epoch": 0.6828703703703703, + "grad_norm": 0.4618282616138458, + "learning_rate": 0.00018607762264677722, + "loss": 0.8727, + "step": 3835 + }, + { + "epoch": 0.6830484330484331, + "grad_norm": 0.6014040112495422, + "learning_rate": 0.00018607049732000436, + "loss": 1.1823, + "step": 3836 + }, + { + "epoch": 0.6832264957264957, + "grad_norm": 0.6489043831825256, + "learning_rate": 0.00018606337030685892, + "loss": 1.1466, + "step": 3837 + }, + { + "epoch": 0.6834045584045584, + "grad_norm": 0.5527763366699219, + "learning_rate": 0.00018605624160748053, + "loss": 1.3015, + "step": 3838 + }, + { + "epoch": 0.6835826210826211, + "grad_norm": 0.5628284215927124, + "learning_rate": 0.0001860491112220088, + "loss": 1.1504, + "step": 3839 + }, + { + "epoch": 0.6837606837606838, + "grad_norm": 0.5414566993713379, + "learning_rate": 0.00018604197915058355, + "loss": 1.0155, + "step": 3840 + }, + { + "epoch": 0.6839387464387464, + "grad_norm": 0.5378929376602173, + "learning_rate": 0.00018603484539334443, + "loss": 0.8917, + "step": 3841 + }, + { + "epoch": 0.6841168091168092, + "grad_norm": 0.5953748822212219, + "learning_rate": 0.00018602770995043125, + "loss": 1.1971, + "step": 3842 + }, + { + "epoch": 0.6842948717948718, + "grad_norm": 0.511813759803772, + "learning_rate": 0.00018602057282198376, + "loss": 1.1345, + "step": 3843 + }, + { + "epoch": 0.6844729344729344, + "grad_norm": 0.5145484209060669, + "learning_rate": 0.00018601343400814185, + "loss": 1.0786, + "step": 3844 + }, + { + "epoch": 0.6846509971509972, + "grad_norm": 0.5199604034423828, + "learning_rate": 0.00018600629350904542, + "loss": 1.2063, + "step": 3845 + }, + { + "epoch": 0.6848290598290598, + "grad_norm": 0.5653825998306274, + "learning_rate": 0.0001859991513248343, + "loss": 1.0314, + "step": 3846 + }, + { + "epoch": 0.6850071225071225, + "grad_norm": 0.5660843849182129, + "learning_rate": 0.00018599200745564843, + "loss": 1.2754, + "step": 3847 + }, + { + "epoch": 0.6851851851851852, + "grad_norm": 0.5225719809532166, + "learning_rate": 0.00018598486190162788, + "loss": 1.0837, + "step": 3848 + }, + { + "epoch": 0.6853632478632479, + "grad_norm": 0.5011669397354126, + "learning_rate": 0.00018597771466291252, + "loss": 1.1, + "step": 3849 + }, + { + "epoch": 0.6855413105413105, + "grad_norm": 0.5923115015029907, + "learning_rate": 0.00018597056573964245, + "loss": 1.1875, + "step": 3850 + }, + { + "epoch": 0.6857193732193733, + "grad_norm": 0.5666482448577881, + "learning_rate": 0.00018596341513195776, + "loss": 1.1663, + "step": 3851 + }, + { + "epoch": 0.6858974358974359, + "grad_norm": 0.5396790504455566, + "learning_rate": 0.0001859562628399985, + "loss": 1.1179, + "step": 3852 + }, + { + "epoch": 0.6860754985754985, + "grad_norm": 0.5709532499313354, + "learning_rate": 0.00018594910886390485, + "loss": 1.0369, + "step": 3853 + }, + { + "epoch": 0.6862535612535613, + "grad_norm": 0.45524322986602783, + "learning_rate": 0.00018594195320381692, + "loss": 1.0171, + "step": 3854 + }, + { + "epoch": 0.6864316239316239, + "grad_norm": 0.6130724549293518, + "learning_rate": 0.00018593479585987498, + "loss": 1.1944, + "step": 3855 + }, + { + "epoch": 0.6866096866096866, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.0001859276368322192, + "loss": 1.2567, + "step": 3856 + }, + { + "epoch": 0.6867877492877493, + "grad_norm": 0.49919846653938293, + "learning_rate": 0.00018592047612098992, + "loss": 0.9459, + "step": 3857 + }, + { + "epoch": 0.686965811965812, + "grad_norm": 0.5776857733726501, + "learning_rate": 0.00018591331372632734, + "loss": 1.2456, + "step": 3858 + }, + { + "epoch": 0.6871438746438746, + "grad_norm": 0.4740692377090454, + "learning_rate": 0.00018590614964837188, + "loss": 1.0401, + "step": 3859 + }, + { + "epoch": 0.6873219373219374, + "grad_norm": 0.5015742182731628, + "learning_rate": 0.00018589898388726389, + "loss": 1.2052, + "step": 3860 + }, + { + "epoch": 0.6875, + "grad_norm": 0.4819730818271637, + "learning_rate": 0.0001858918164431437, + "loss": 1.007, + "step": 3861 + }, + { + "epoch": 0.6876780626780626, + "grad_norm": 0.5510426163673401, + "learning_rate": 0.00018588464731615184, + "loss": 1.0123, + "step": 3862 + }, + { + "epoch": 0.6878561253561254, + "grad_norm": 0.4950829744338989, + "learning_rate": 0.00018587747650642867, + "loss": 1.033, + "step": 3863 + }, + { + "epoch": 0.688034188034188, + "grad_norm": 0.5278680920600891, + "learning_rate": 0.0001858703040141148, + "loss": 1.0912, + "step": 3864 + }, + { + "epoch": 0.6882122507122507, + "grad_norm": 0.6359158158302307, + "learning_rate": 0.00018586312983935068, + "loss": 1.2868, + "step": 3865 + }, + { + "epoch": 0.6883903133903134, + "grad_norm": 0.5098239183425903, + "learning_rate": 0.0001858559539822769, + "loss": 0.8364, + "step": 3866 + }, + { + "epoch": 0.6885683760683761, + "grad_norm": 0.5651038289070129, + "learning_rate": 0.000185848776443034, + "loss": 1.1983, + "step": 3867 + }, + { + "epoch": 0.6887464387464387, + "grad_norm": 0.5305678248405457, + "learning_rate": 0.00018584159722176272, + "loss": 1.32, + "step": 3868 + }, + { + "epoch": 0.6889245014245015, + "grad_norm": 0.5481845140457153, + "learning_rate": 0.00018583441631860368, + "loss": 1.013, + "step": 3869 + }, + { + "epoch": 0.6891025641025641, + "grad_norm": 0.5214795470237732, + "learning_rate": 0.00018582723373369753, + "loss": 1.172, + "step": 3870 + }, + { + "epoch": 0.6892806267806267, + "grad_norm": 0.6282780766487122, + "learning_rate": 0.00018582004946718502, + "loss": 1.7304, + "step": 3871 + }, + { + "epoch": 0.6894586894586895, + "grad_norm": 0.5266988277435303, + "learning_rate": 0.0001858128635192069, + "loss": 1.1418, + "step": 3872 + }, + { + "epoch": 0.6896367521367521, + "grad_norm": 0.4761001467704773, + "learning_rate": 0.000185805675889904, + "loss": 0.8585, + "step": 3873 + }, + { + "epoch": 0.6898148148148148, + "grad_norm": 0.528779923915863, + "learning_rate": 0.00018579848657941715, + "loss": 1.0036, + "step": 3874 + }, + { + "epoch": 0.6899928774928775, + "grad_norm": 0.5427684783935547, + "learning_rate": 0.00018579129558788716, + "loss": 0.9769, + "step": 3875 + }, + { + "epoch": 0.6901709401709402, + "grad_norm": 0.6229544281959534, + "learning_rate": 0.00018578410291545495, + "loss": 1.2848, + "step": 3876 + }, + { + "epoch": 0.6903490028490028, + "grad_norm": 0.6602693200111389, + "learning_rate": 0.00018577690856226147, + "loss": 1.2713, + "step": 3877 + }, + { + "epoch": 0.6905270655270656, + "grad_norm": 0.45884042978286743, + "learning_rate": 0.0001857697125284476, + "loss": 0.9143, + "step": 3878 + }, + { + "epoch": 0.6907051282051282, + "grad_norm": 0.4956444203853607, + "learning_rate": 0.00018576251481415443, + "loss": 0.9646, + "step": 3879 + }, + { + "epoch": 0.6908831908831908, + "grad_norm": 0.473561555147171, + "learning_rate": 0.00018575531541952292, + "loss": 0.843, + "step": 3880 + }, + { + "epoch": 0.6910612535612536, + "grad_norm": 0.4676312506198883, + "learning_rate": 0.00018574811434469415, + "loss": 0.9464, + "step": 3881 + }, + { + "epoch": 0.6912393162393162, + "grad_norm": 0.5452045202255249, + "learning_rate": 0.00018574091158980922, + "loss": 0.985, + "step": 3882 + }, + { + "epoch": 0.6914173789173789, + "grad_norm": 0.6274946331977844, + "learning_rate": 0.0001857337071550092, + "loss": 1.0357, + "step": 3883 + }, + { + "epoch": 0.6915954415954416, + "grad_norm": 0.5533788800239563, + "learning_rate": 0.00018572650104043531, + "loss": 1.2636, + "step": 3884 + }, + { + "epoch": 0.6917735042735043, + "grad_norm": 0.48312318325042725, + "learning_rate": 0.00018571929324622872, + "loss": 1.2402, + "step": 3885 + }, + { + "epoch": 0.6919515669515669, + "grad_norm": 0.6087453961372375, + "learning_rate": 0.00018571208377253062, + "loss": 1.2961, + "step": 3886 + }, + { + "epoch": 0.6921296296296297, + "grad_norm": 0.49156486988067627, + "learning_rate": 0.00018570487261948234, + "loss": 0.9585, + "step": 3887 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 0.5200015902519226, + "learning_rate": 0.0001856976597872251, + "loss": 0.9274, + "step": 3888 + }, + { + "epoch": 0.6924857549857549, + "grad_norm": 0.5185118913650513, + "learning_rate": 0.0001856904452759002, + "loss": 1.0015, + "step": 3889 + }, + { + "epoch": 0.6926638176638177, + "grad_norm": 0.5859049558639526, + "learning_rate": 0.00018568322908564904, + "loss": 1.0959, + "step": 3890 + }, + { + "epoch": 0.6928418803418803, + "grad_norm": 0.5882301926612854, + "learning_rate": 0.00018567601121661302, + "loss": 1.3214, + "step": 3891 + }, + { + "epoch": 0.6930199430199431, + "grad_norm": 0.6475503444671631, + "learning_rate": 0.0001856687916689335, + "loss": 1.3265, + "step": 3892 + }, + { + "epoch": 0.6931980056980057, + "grad_norm": 0.46175432205200195, + "learning_rate": 0.000185661570442752, + "loss": 0.8547, + "step": 3893 + }, + { + "epoch": 0.6933760683760684, + "grad_norm": 0.5362716913223267, + "learning_rate": 0.00018565434753820998, + "loss": 0.974, + "step": 3894 + }, + { + "epoch": 0.6935541310541311, + "grad_norm": 0.4317963719367981, + "learning_rate": 0.00018564712295544896, + "loss": 0.7653, + "step": 3895 + }, + { + "epoch": 0.6937321937321937, + "grad_norm": 0.5679717659950256, + "learning_rate": 0.00018563989669461047, + "loss": 1.0691, + "step": 3896 + }, + { + "epoch": 0.6939102564102564, + "grad_norm": 0.5058363676071167, + "learning_rate": 0.00018563266875583608, + "loss": 1.0665, + "step": 3897 + }, + { + "epoch": 0.6940883190883191, + "grad_norm": 0.5365496277809143, + "learning_rate": 0.00018562543913926746, + "loss": 0.9963, + "step": 3898 + }, + { + "epoch": 0.6942663817663818, + "grad_norm": 0.49945300817489624, + "learning_rate": 0.0001856182078450462, + "loss": 0.8668, + "step": 3899 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 0.5869430899620056, + "learning_rate": 0.00018561097487331405, + "loss": 1.1942, + "step": 3900 + }, + { + "epoch": 0.6946225071225072, + "grad_norm": 0.5188950300216675, + "learning_rate": 0.0001856037402242127, + "loss": 0.9493, + "step": 3901 + }, + { + "epoch": 0.6948005698005698, + "grad_norm": 0.510788083076477, + "learning_rate": 0.00018559650389788384, + "loss": 0.9989, + "step": 3902 + }, + { + "epoch": 0.6949786324786325, + "grad_norm": 0.5360601544380188, + "learning_rate": 0.0001855892658944693, + "loss": 1.2766, + "step": 3903 + }, + { + "epoch": 0.6951566951566952, + "grad_norm": 0.522502601146698, + "learning_rate": 0.00018558202621411093, + "loss": 0.8774, + "step": 3904 + }, + { + "epoch": 0.6953347578347578, + "grad_norm": 0.5330635905265808, + "learning_rate": 0.00018557478485695052, + "loss": 0.972, + "step": 3905 + }, + { + "epoch": 0.6955128205128205, + "grad_norm": 0.5387479066848755, + "learning_rate": 0.00018556754182312996, + "loss": 1.0574, + "step": 3906 + }, + { + "epoch": 0.6956908831908832, + "grad_norm": 0.5357984900474548, + "learning_rate": 0.00018556029711279116, + "loss": 1.396, + "step": 3907 + }, + { + "epoch": 0.6958689458689459, + "grad_norm": 0.5647178292274475, + "learning_rate": 0.00018555305072607612, + "loss": 1.3304, + "step": 3908 + }, + { + "epoch": 0.6960470085470085, + "grad_norm": 0.46460914611816406, + "learning_rate": 0.00018554580266312673, + "loss": 0.9574, + "step": 3909 + }, + { + "epoch": 0.6962250712250713, + "grad_norm": 0.6206206679344177, + "learning_rate": 0.00018553855292408503, + "loss": 1.1637, + "step": 3910 + }, + { + "epoch": 0.6964031339031339, + "grad_norm": 0.5899842977523804, + "learning_rate": 0.00018553130150909312, + "loss": 1.1067, + "step": 3911 + }, + { + "epoch": 0.6965811965811965, + "grad_norm": 0.47294262051582336, + "learning_rate": 0.000185524048418293, + "loss": 1.1516, + "step": 3912 + }, + { + "epoch": 0.6967592592592593, + "grad_norm": 0.5791197419166565, + "learning_rate": 0.00018551679365182684, + "loss": 1.0007, + "step": 3913 + }, + { + "epoch": 0.6969373219373219, + "grad_norm": 0.5678651332855225, + "learning_rate": 0.00018550953720983672, + "loss": 1.2698, + "step": 3914 + }, + { + "epoch": 0.6971153846153846, + "grad_norm": 0.6509683728218079, + "learning_rate": 0.0001855022790924649, + "loss": 1.0354, + "step": 3915 + }, + { + "epoch": 0.6972934472934473, + "grad_norm": 0.5176648497581482, + "learning_rate": 0.0001854950192998535, + "loss": 1.1243, + "step": 3916 + }, + { + "epoch": 0.69747150997151, + "grad_norm": 0.520631730556488, + "learning_rate": 0.00018548775783214477, + "loss": 1.1371, + "step": 3917 + }, + { + "epoch": 0.6976495726495726, + "grad_norm": 0.5408333539962769, + "learning_rate": 0.00018548049468948108, + "loss": 1.1185, + "step": 3918 + }, + { + "epoch": 0.6978276353276354, + "grad_norm": 0.5423790216445923, + "learning_rate": 0.00018547322987200461, + "loss": 1.1539, + "step": 3919 + }, + { + "epoch": 0.698005698005698, + "grad_norm": 0.5422113537788391, + "learning_rate": 0.0001854659633798578, + "loss": 1.171, + "step": 3920 + }, + { + "epoch": 0.6981837606837606, + "grad_norm": 0.5113416314125061, + "learning_rate": 0.00018545869521318292, + "loss": 1.0597, + "step": 3921 + }, + { + "epoch": 0.6983618233618234, + "grad_norm": 0.49901214241981506, + "learning_rate": 0.00018545142537212248, + "loss": 1.1043, + "step": 3922 + }, + { + "epoch": 0.698539886039886, + "grad_norm": 0.6606622338294983, + "learning_rate": 0.00018544415385681885, + "loss": 1.1797, + "step": 3923 + }, + { + "epoch": 0.6987179487179487, + "grad_norm": 0.4786234498023987, + "learning_rate": 0.00018543688066741454, + "loss": 0.9532, + "step": 3924 + }, + { + "epoch": 0.6988960113960114, + "grad_norm": 0.5900700688362122, + "learning_rate": 0.00018542960580405203, + "loss": 1.1171, + "step": 3925 + }, + { + "epoch": 0.6990740740740741, + "grad_norm": 0.53485506772995, + "learning_rate": 0.00018542232926687383, + "loss": 1.1535, + "step": 3926 + }, + { + "epoch": 0.6992521367521367, + "grad_norm": 0.5269177556037903, + "learning_rate": 0.00018541505105602255, + "loss": 1.0287, + "step": 3927 + }, + { + "epoch": 0.6994301994301995, + "grad_norm": 0.5185505151748657, + "learning_rate": 0.0001854077711716408, + "loss": 1.2526, + "step": 3928 + }, + { + "epoch": 0.6996082621082621, + "grad_norm": 0.5615512132644653, + "learning_rate": 0.00018540048961387115, + "loss": 1.0189, + "step": 3929 + }, + { + "epoch": 0.6997863247863247, + "grad_norm": 0.4492493271827698, + "learning_rate": 0.00018539320638285637, + "loss": 0.8917, + "step": 3930 + }, + { + "epoch": 0.6999643874643875, + "grad_norm": 0.5062302947044373, + "learning_rate": 0.00018538592147873906, + "loss": 1.053, + "step": 3931 + }, + { + "epoch": 0.7001424501424501, + "grad_norm": 0.5508798956871033, + "learning_rate": 0.000185378634901662, + "loss": 0.9638, + "step": 3932 + }, + { + "epoch": 0.7003205128205128, + "grad_norm": 0.463980108499527, + "learning_rate": 0.00018537134665176793, + "loss": 1.0945, + "step": 3933 + }, + { + "epoch": 0.7004985754985755, + "grad_norm": 0.5027088522911072, + "learning_rate": 0.0001853640567291997, + "loss": 1.1745, + "step": 3934 + }, + { + "epoch": 0.7006766381766382, + "grad_norm": 0.5006551146507263, + "learning_rate": 0.00018535676513410009, + "loss": 0.8521, + "step": 3935 + }, + { + "epoch": 0.7008547008547008, + "grad_norm": 0.5870724320411682, + "learning_rate": 0.000185349471866612, + "loss": 0.9197, + "step": 3936 + }, + { + "epoch": 0.7010327635327636, + "grad_norm": 0.5030696392059326, + "learning_rate": 0.00018534217692687825, + "loss": 1.1049, + "step": 3937 + }, + { + "epoch": 0.7012108262108262, + "grad_norm": 0.5212681889533997, + "learning_rate": 0.00018533488031504186, + "loss": 1.3397, + "step": 3938 + }, + { + "epoch": 0.7013888888888888, + "grad_norm": 0.5649709105491638, + "learning_rate": 0.0001853275820312458, + "loss": 1.1994, + "step": 3939 + }, + { + "epoch": 0.7015669515669516, + "grad_norm": 0.4892779290676117, + "learning_rate": 0.00018532028207563297, + "loss": 1.1511, + "step": 3940 + }, + { + "epoch": 0.7017450142450142, + "grad_norm": 0.4929407835006714, + "learning_rate": 0.00018531298044834643, + "loss": 1.0792, + "step": 3941 + }, + { + "epoch": 0.7019230769230769, + "grad_norm": 0.5645940899848938, + "learning_rate": 0.00018530567714952932, + "loss": 1.0937, + "step": 3942 + }, + { + "epoch": 0.7021011396011396, + "grad_norm": 0.5471178293228149, + "learning_rate": 0.00018529837217932466, + "loss": 1.193, + "step": 3943 + }, + { + "epoch": 0.7022792022792023, + "grad_norm": 0.576627790927887, + "learning_rate": 0.00018529106553787558, + "loss": 1.1032, + "step": 3944 + }, + { + "epoch": 0.7024572649572649, + "grad_norm": 0.5015735626220703, + "learning_rate": 0.00018528375722532526, + "loss": 1.066, + "step": 3945 + }, + { + "epoch": 0.7026353276353277, + "grad_norm": 0.5315404534339905, + "learning_rate": 0.00018527644724181683, + "loss": 1.2059, + "step": 3946 + }, + { + "epoch": 0.7028133903133903, + "grad_norm": 0.5516065955162048, + "learning_rate": 0.0001852691355874936, + "loss": 1.161, + "step": 3947 + }, + { + "epoch": 0.7029914529914529, + "grad_norm": 0.5026212930679321, + "learning_rate": 0.0001852618222624988, + "loss": 1.2616, + "step": 3948 + }, + { + "epoch": 0.7031695156695157, + "grad_norm": 0.49874603748321533, + "learning_rate": 0.0001852545072669757, + "loss": 0.805, + "step": 3949 + }, + { + "epoch": 0.7033475783475783, + "grad_norm": 0.47698748111724854, + "learning_rate": 0.00018524719060106763, + "loss": 1.2321, + "step": 3950 + }, + { + "epoch": 0.7035256410256411, + "grad_norm": 0.5201322436332703, + "learning_rate": 0.00018523987226491792, + "loss": 1.1577, + "step": 3951 + }, + { + "epoch": 0.7037037037037037, + "grad_norm": 0.5506543517112732, + "learning_rate": 0.00018523255225867002, + "loss": 1.2289, + "step": 3952 + }, + { + "epoch": 0.7038817663817664, + "grad_norm": 0.5691256523132324, + "learning_rate": 0.0001852252305824673, + "loss": 1.1945, + "step": 3953 + }, + { + "epoch": 0.7040598290598291, + "grad_norm": 0.5324838757514954, + "learning_rate": 0.00018521790723645322, + "loss": 1.1037, + "step": 3954 + }, + { + "epoch": 0.7042378917378918, + "grad_norm": 0.5238786339759827, + "learning_rate": 0.00018521058222077127, + "loss": 1.2075, + "step": 3955 + }, + { + "epoch": 0.7044159544159544, + "grad_norm": 0.4936453402042389, + "learning_rate": 0.00018520325553556498, + "loss": 1.0537, + "step": 3956 + }, + { + "epoch": 0.7045940170940171, + "grad_norm": 0.6198282837867737, + "learning_rate": 0.00018519592718097791, + "loss": 1.0728, + "step": 3957 + }, + { + "epoch": 0.7047720797720798, + "grad_norm": 0.44729140400886536, + "learning_rate": 0.0001851885971571536, + "loss": 0.8432, + "step": 3958 + }, + { + "epoch": 0.7049501424501424, + "grad_norm": 0.5884211659431458, + "learning_rate": 0.00018518126546423572, + "loss": 0.9515, + "step": 3959 + }, + { + "epoch": 0.7051282051282052, + "grad_norm": 0.5293807983398438, + "learning_rate": 0.00018517393210236788, + "loss": 1.1178, + "step": 3960 + }, + { + "epoch": 0.7053062678062678, + "grad_norm": 0.6036825180053711, + "learning_rate": 0.00018516659707169374, + "loss": 1.0408, + "step": 3961 + }, + { + "epoch": 0.7054843304843305, + "grad_norm": 0.5157122015953064, + "learning_rate": 0.0001851592603723571, + "loss": 1.2136, + "step": 3962 + }, + { + "epoch": 0.7056623931623932, + "grad_norm": 0.5354781150817871, + "learning_rate": 0.00018515192200450163, + "loss": 0.7165, + "step": 3963 + }, + { + "epoch": 0.7058404558404558, + "grad_norm": 0.6073734760284424, + "learning_rate": 0.00018514458196827111, + "loss": 1.3079, + "step": 3964 + }, + { + "epoch": 0.7060185185185185, + "grad_norm": 0.4324839413166046, + "learning_rate": 0.0001851372402638094, + "loss": 0.7903, + "step": 3965 + }, + { + "epoch": 0.7061965811965812, + "grad_norm": 0.6530333161354065, + "learning_rate": 0.00018512989689126034, + "loss": 1.3179, + "step": 3966 + }, + { + "epoch": 0.7063746438746439, + "grad_norm": 0.5500404238700867, + "learning_rate": 0.00018512255185076782, + "loss": 1.0624, + "step": 3967 + }, + { + "epoch": 0.7065527065527065, + "grad_norm": 0.6277863383293152, + "learning_rate": 0.00018511520514247567, + "loss": 1.1056, + "step": 3968 + }, + { + "epoch": 0.7067307692307693, + "grad_norm": 0.580544650554657, + "learning_rate": 0.0001851078567665279, + "loss": 0.9849, + "step": 3969 + }, + { + "epoch": 0.7069088319088319, + "grad_norm": 0.4880999028682709, + "learning_rate": 0.00018510050672306848, + "loss": 1.0185, + "step": 3970 + }, + { + "epoch": 0.7070868945868946, + "grad_norm": 0.4919959306716919, + "learning_rate": 0.0001850931550122414, + "loss": 1.0334, + "step": 3971 + }, + { + "epoch": 0.7072649572649573, + "grad_norm": 0.6001213192939758, + "learning_rate": 0.0001850858016341907, + "loss": 1.0729, + "step": 3972 + }, + { + "epoch": 0.70744301994302, + "grad_norm": 0.538690447807312, + "learning_rate": 0.00018507844658906052, + "loss": 1.0733, + "step": 3973 + }, + { + "epoch": 0.7076210826210826, + "grad_norm": 0.5427643656730652, + "learning_rate": 0.00018507108987699487, + "loss": 1.1207, + "step": 3974 + }, + { + "epoch": 0.7077991452991453, + "grad_norm": 0.43014347553253174, + "learning_rate": 0.00018506373149813795, + "loss": 0.7958, + "step": 3975 + }, + { + "epoch": 0.707977207977208, + "grad_norm": 0.56591796875, + "learning_rate": 0.00018505637145263394, + "loss": 1.2199, + "step": 3976 + }, + { + "epoch": 0.7081552706552706, + "grad_norm": 0.59147047996521, + "learning_rate": 0.000185049009740627, + "loss": 1.2354, + "step": 3977 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 0.5078346133232117, + "learning_rate": 0.00018504164636226137, + "loss": 0.976, + "step": 3978 + }, + { + "epoch": 0.708511396011396, + "grad_norm": 0.533302366733551, + "learning_rate": 0.00018503428131768135, + "loss": 0.9653, + "step": 3979 + }, + { + "epoch": 0.7086894586894587, + "grad_norm": 0.4985341727733612, + "learning_rate": 0.00018502691460703122, + "loss": 1.1485, + "step": 3980 + }, + { + "epoch": 0.7088675213675214, + "grad_norm": 0.5143141150474548, + "learning_rate": 0.00018501954623045532, + "loss": 1.148, + "step": 3981 + }, + { + "epoch": 0.709045584045584, + "grad_norm": 0.507189154624939, + "learning_rate": 0.00018501217618809804, + "loss": 0.9306, + "step": 3982 + }, + { + "epoch": 0.7092236467236467, + "grad_norm": 0.5246604084968567, + "learning_rate": 0.00018500480448010377, + "loss": 0.9116, + "step": 3983 + }, + { + "epoch": 0.7094017094017094, + "grad_norm": 0.5321049094200134, + "learning_rate": 0.00018499743110661693, + "loss": 0.9607, + "step": 3984 + }, + { + "epoch": 0.7095797720797721, + "grad_norm": 0.62645423412323, + "learning_rate": 0.000184990056067782, + "loss": 1.5834, + "step": 3985 + }, + { + "epoch": 0.7097578347578347, + "grad_norm": 0.486557275056839, + "learning_rate": 0.0001849826793637435, + "loss": 1.0598, + "step": 3986 + }, + { + "epoch": 0.7099358974358975, + "grad_norm": 0.5122783184051514, + "learning_rate": 0.0001849753009946459, + "loss": 1.2213, + "step": 3987 + }, + { + "epoch": 0.7101139601139601, + "grad_norm": 0.4864068627357483, + "learning_rate": 0.0001849679209606338, + "loss": 1.2708, + "step": 3988 + }, + { + "epoch": 0.7102920227920227, + "grad_norm": 0.5860990881919861, + "learning_rate": 0.00018496053926185183, + "loss": 1.2421, + "step": 3989 + }, + { + "epoch": 0.7104700854700855, + "grad_norm": 0.471194326877594, + "learning_rate": 0.00018495315589844453, + "loss": 0.879, + "step": 3990 + }, + { + "epoch": 0.7106481481481481, + "grad_norm": 0.5626323819160461, + "learning_rate": 0.00018494577087055662, + "loss": 1.1297, + "step": 3991 + }, + { + "epoch": 0.7108262108262108, + "grad_norm": 0.4706762135028839, + "learning_rate": 0.0001849383841783328, + "loss": 1.0444, + "step": 3992 + }, + { + "epoch": 0.7110042735042735, + "grad_norm": 0.5776444673538208, + "learning_rate": 0.00018493099582191783, + "loss": 1.1773, + "step": 3993 + }, + { + "epoch": 0.7111823361823362, + "grad_norm": 0.5493253469467163, + "learning_rate": 0.00018492360580145637, + "loss": 1.0354, + "step": 3994 + }, + { + "epoch": 0.7113603988603988, + "grad_norm": 0.5328514575958252, + "learning_rate": 0.0001849162141170933, + "loss": 0.9251, + "step": 3995 + }, + { + "epoch": 0.7115384615384616, + "grad_norm": 0.5814893841743469, + "learning_rate": 0.0001849088207689734, + "loss": 1.1066, + "step": 3996 + }, + { + "epoch": 0.7117165242165242, + "grad_norm": 0.5476071834564209, + "learning_rate": 0.00018490142575724154, + "loss": 1.1613, + "step": 3997 + }, + { + "epoch": 0.7118945868945868, + "grad_norm": 0.5216463208198547, + "learning_rate": 0.00018489402908204258, + "loss": 1.2574, + "step": 3998 + }, + { + "epoch": 0.7120726495726496, + "grad_norm": 0.5110020637512207, + "learning_rate": 0.00018488663074352153, + "loss": 1.0663, + "step": 3999 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 0.448090523481369, + "learning_rate": 0.00018487923074182326, + "loss": 0.6687, + "step": 4000 + }, + { + "epoch": 0.7124287749287749, + "grad_norm": 0.4980565011501312, + "learning_rate": 0.00018487182907709279, + "loss": 1.2365, + "step": 4001 + }, + { + "epoch": 0.7126068376068376, + "grad_norm": 0.485831081867218, + "learning_rate": 0.00018486442574947511, + "loss": 1.0941, + "step": 4002 + }, + { + "epoch": 0.7127849002849003, + "grad_norm": 0.4955040216445923, + "learning_rate": 0.00018485702075911534, + "loss": 1.248, + "step": 4003 + }, + { + "epoch": 0.7129629629629629, + "grad_norm": 0.5168375968933105, + "learning_rate": 0.00018484961410615845, + "loss": 1.1118, + "step": 4004 + }, + { + "epoch": 0.7131410256410257, + "grad_norm": 0.5255687832832336, + "learning_rate": 0.00018484220579074968, + "loss": 1.0558, + "step": 4005 + }, + { + "epoch": 0.7133190883190883, + "grad_norm": 0.5502219796180725, + "learning_rate": 0.00018483479581303416, + "loss": 1.1604, + "step": 4006 + }, + { + "epoch": 0.7134971509971509, + "grad_norm": 0.5155881643295288, + "learning_rate": 0.000184827384173157, + "loss": 0.8246, + "step": 4007 + }, + { + "epoch": 0.7136752136752137, + "grad_norm": 0.5321542024612427, + "learning_rate": 0.0001848199708712635, + "loss": 1.2058, + "step": 4008 + }, + { + "epoch": 0.7138532763532763, + "grad_norm": 0.4929848313331604, + "learning_rate": 0.00018481255590749884, + "loss": 1.4023, + "step": 4009 + }, + { + "epoch": 0.7140313390313391, + "grad_norm": 0.5070937871932983, + "learning_rate": 0.00018480513928200836, + "loss": 1.0561, + "step": 4010 + }, + { + "epoch": 0.7142094017094017, + "grad_norm": 0.5750083327293396, + "learning_rate": 0.00018479772099493728, + "loss": 1.0276, + "step": 4011 + }, + { + "epoch": 0.7143874643874644, + "grad_norm": 0.5265933275222778, + "learning_rate": 0.00018479030104643108, + "loss": 1.0295, + "step": 4012 + }, + { + "epoch": 0.7145655270655271, + "grad_norm": 0.526830792427063, + "learning_rate": 0.00018478287943663504, + "loss": 1.0157, + "step": 4013 + }, + { + "epoch": 0.7147435897435898, + "grad_norm": 0.5344091653823853, + "learning_rate": 0.00018477545616569458, + "loss": 1.1997, + "step": 4014 + }, + { + "epoch": 0.7149216524216524, + "grad_norm": 0.4935445189476013, + "learning_rate": 0.0001847680312337552, + "loss": 1.1858, + "step": 4015 + }, + { + "epoch": 0.7150997150997151, + "grad_norm": 0.5291212797164917, + "learning_rate": 0.0001847606046409623, + "loss": 0.926, + "step": 4016 + }, + { + "epoch": 0.7152777777777778, + "grad_norm": 0.559050977230072, + "learning_rate": 0.00018475317638746142, + "loss": 1.0947, + "step": 4017 + }, + { + "epoch": 0.7154558404558404, + "grad_norm": 0.4566570222377777, + "learning_rate": 0.00018474574647339814, + "loss": 1.0334, + "step": 4018 + }, + { + "epoch": 0.7156339031339032, + "grad_norm": 0.5156155824661255, + "learning_rate": 0.000184738314898918, + "loss": 1.0076, + "step": 4019 + }, + { + "epoch": 0.7158119658119658, + "grad_norm": 0.5008716583251953, + "learning_rate": 0.00018473088166416662, + "loss": 1.0378, + "step": 4020 + }, + { + "epoch": 0.7159900284900285, + "grad_norm": 0.49556368589401245, + "learning_rate": 0.0001847234467692896, + "loss": 1.15, + "step": 4021 + }, + { + "epoch": 0.7161680911680912, + "grad_norm": 0.5464680790901184, + "learning_rate": 0.00018471601021443265, + "loss": 1.2975, + "step": 4022 + }, + { + "epoch": 0.7163461538461539, + "grad_norm": 0.6291980147361755, + "learning_rate": 0.00018470857199974144, + "loss": 1.05, + "step": 4023 + }, + { + "epoch": 0.7165242165242165, + "grad_norm": 0.5566631555557251, + "learning_rate": 0.00018470113212536176, + "loss": 1.1296, + "step": 4024 + }, + { + "epoch": 0.7167022792022792, + "grad_norm": 0.5569562911987305, + "learning_rate": 0.00018469369059143933, + "loss": 1.2484, + "step": 4025 + }, + { + "epoch": 0.7168803418803419, + "grad_norm": 0.5804716944694519, + "learning_rate": 0.00018468624739812, + "loss": 1.0547, + "step": 4026 + }, + { + "epoch": 0.7170584045584045, + "grad_norm": 0.6316802501678467, + "learning_rate": 0.00018467880254554952, + "loss": 1.1188, + "step": 4027 + }, + { + "epoch": 0.7172364672364673, + "grad_norm": 0.6131419539451599, + "learning_rate": 0.00018467135603387385, + "loss": 1.1662, + "step": 4028 + }, + { + "epoch": 0.7174145299145299, + "grad_norm": 0.4703124761581421, + "learning_rate": 0.00018466390786323883, + "loss": 1.038, + "step": 4029 + }, + { + "epoch": 0.7175925925925926, + "grad_norm": 0.5718469023704529, + "learning_rate": 0.0001846564580337904, + "loss": 1.0786, + "step": 4030 + }, + { + "epoch": 0.7177706552706553, + "grad_norm": 0.5227612853050232, + "learning_rate": 0.00018464900654567457, + "loss": 1.0561, + "step": 4031 + }, + { + "epoch": 0.717948717948718, + "grad_norm": 0.5800358057022095, + "learning_rate": 0.00018464155339903727, + "loss": 1.0944, + "step": 4032 + }, + { + "epoch": 0.7181267806267806, + "grad_norm": 0.5562314987182617, + "learning_rate": 0.00018463409859402455, + "loss": 0.8573, + "step": 4033 + }, + { + "epoch": 0.7183048433048433, + "grad_norm": 0.6420153379440308, + "learning_rate": 0.0001846266421307825, + "loss": 1.088, + "step": 4034 + }, + { + "epoch": 0.718482905982906, + "grad_norm": 0.4745902717113495, + "learning_rate": 0.00018461918400945718, + "loss": 1.1679, + "step": 4035 + }, + { + "epoch": 0.7186609686609686, + "grad_norm": 0.5070300102233887, + "learning_rate": 0.00018461172423019475, + "loss": 1.1984, + "step": 4036 + }, + { + "epoch": 0.7188390313390314, + "grad_norm": 0.5339375138282776, + "learning_rate": 0.00018460426279314133, + "loss": 1.3038, + "step": 4037 + }, + { + "epoch": 0.719017094017094, + "grad_norm": 0.5947147607803345, + "learning_rate": 0.00018459679969844313, + "loss": 1.0103, + "step": 4038 + }, + { + "epoch": 0.7191951566951567, + "grad_norm": 0.5493791699409485, + "learning_rate": 0.00018458933494624642, + "loss": 1.1001, + "step": 4039 + }, + { + "epoch": 0.7193732193732194, + "grad_norm": 0.5700310468673706, + "learning_rate": 0.00018458186853669736, + "loss": 0.9006, + "step": 4040 + }, + { + "epoch": 0.719551282051282, + "grad_norm": 0.60371994972229, + "learning_rate": 0.0001845744004699423, + "loss": 1.3001, + "step": 4041 + }, + { + "epoch": 0.7197293447293447, + "grad_norm": 0.5469261407852173, + "learning_rate": 0.00018456693074612757, + "loss": 1.1745, + "step": 4042 + }, + { + "epoch": 0.7199074074074074, + "grad_norm": 0.5179165601730347, + "learning_rate": 0.00018455945936539947, + "loss": 0.9883, + "step": 4043 + }, + { + "epoch": 0.7200854700854701, + "grad_norm": 0.5396696329116821, + "learning_rate": 0.00018455198632790447, + "loss": 1.1277, + "step": 4044 + }, + { + "epoch": 0.7202635327635327, + "grad_norm": 0.4559909403324127, + "learning_rate": 0.00018454451163378888, + "loss": 0.9644, + "step": 4045 + }, + { + "epoch": 0.7204415954415955, + "grad_norm": 0.49863892793655396, + "learning_rate": 0.00018453703528319927, + "loss": 1.1276, + "step": 4046 + }, + { + "epoch": 0.7206196581196581, + "grad_norm": 0.4790710508823395, + "learning_rate": 0.000184529557276282, + "loss": 0.9443, + "step": 4047 + }, + { + "epoch": 0.7207977207977208, + "grad_norm": 0.541999876499176, + "learning_rate": 0.0001845220776131837, + "loss": 1.0681, + "step": 4048 + }, + { + "epoch": 0.7209757834757835, + "grad_norm": 0.5119109153747559, + "learning_rate": 0.00018451459629405088, + "loss": 1.2078, + "step": 4049 + }, + { + "epoch": 0.7211538461538461, + "grad_norm": 0.6141307353973389, + "learning_rate": 0.00018450711331903006, + "loss": 1.1071, + "step": 4050 + }, + { + "epoch": 0.7213319088319088, + "grad_norm": 0.48679864406585693, + "learning_rate": 0.00018449962868826795, + "loss": 0.9713, + "step": 4051 + }, + { + "epoch": 0.7215099715099715, + "grad_norm": 0.5548661947250366, + "learning_rate": 0.0001844921424019111, + "loss": 1.2099, + "step": 4052 + }, + { + "epoch": 0.7216880341880342, + "grad_norm": 0.5000107884407043, + "learning_rate": 0.00018448465446010626, + "loss": 1.0184, + "step": 4053 + }, + { + "epoch": 0.7218660968660968, + "grad_norm": 0.6131454110145569, + "learning_rate": 0.00018447716486300013, + "loss": 1.2581, + "step": 4054 + }, + { + "epoch": 0.7220441595441596, + "grad_norm": 0.5145987868309021, + "learning_rate": 0.0001844696736107394, + "loss": 1.1646, + "step": 4055 + }, + { + "epoch": 0.7222222222222222, + "grad_norm": 0.4361337125301361, + "learning_rate": 0.00018446218070347094, + "loss": 0.8239, + "step": 4056 + }, + { + "epoch": 0.7224002849002849, + "grad_norm": 0.5549173355102539, + "learning_rate": 0.00018445468614134146, + "loss": 1.1935, + "step": 4057 + }, + { + "epoch": 0.7225783475783476, + "grad_norm": 0.5569297671318054, + "learning_rate": 0.00018444718992449789, + "loss": 1.0137, + "step": 4058 + }, + { + "epoch": 0.7227564102564102, + "grad_norm": 0.44866305589675903, + "learning_rate": 0.00018443969205308704, + "loss": 0.987, + "step": 4059 + }, + { + "epoch": 0.7229344729344729, + "grad_norm": 0.5142943263053894, + "learning_rate": 0.0001844321925272558, + "loss": 1.0837, + "step": 4060 + }, + { + "epoch": 0.7231125356125356, + "grad_norm": 0.4922119379043579, + "learning_rate": 0.0001844246913471512, + "loss": 0.8477, + "step": 4061 + }, + { + "epoch": 0.7232905982905983, + "grad_norm": 0.5245375633239746, + "learning_rate": 0.0001844171885129201, + "loss": 0.9985, + "step": 4062 + }, + { + "epoch": 0.7234686609686609, + "grad_norm": 0.45562678575515747, + "learning_rate": 0.00018440968402470956, + "loss": 0.8678, + "step": 4063 + }, + { + "epoch": 0.7236467236467237, + "grad_norm": 0.5388376712799072, + "learning_rate": 0.0001844021778826666, + "loss": 1.0586, + "step": 4064 + }, + { + "epoch": 0.7238247863247863, + "grad_norm": 0.48945263028144836, + "learning_rate": 0.00018439467008693833, + "loss": 1.0547, + "step": 4065 + }, + { + "epoch": 0.7240028490028491, + "grad_norm": 0.5202330350875854, + "learning_rate": 0.00018438716063767178, + "loss": 1.3142, + "step": 4066 + }, + { + "epoch": 0.7241809116809117, + "grad_norm": 0.5432567000389099, + "learning_rate": 0.00018437964953501413, + "loss": 1.0192, + "step": 4067 + }, + { + "epoch": 0.7243589743589743, + "grad_norm": 0.5220325589179993, + "learning_rate": 0.00018437213677911253, + "loss": 1.0904, + "step": 4068 + }, + { + "epoch": 0.7245370370370371, + "grad_norm": 0.45711690187454224, + "learning_rate": 0.00018436462237011417, + "loss": 1.0417, + "step": 4069 + }, + { + "epoch": 0.7247150997150997, + "grad_norm": 0.560778021812439, + "learning_rate": 0.0001843571063081663, + "loss": 1.2316, + "step": 4070 + }, + { + "epoch": 0.7248931623931624, + "grad_norm": 0.591533362865448, + "learning_rate": 0.0001843495885934162, + "loss": 1.0294, + "step": 4071 + }, + { + "epoch": 0.7250712250712251, + "grad_norm": 0.5550443530082703, + "learning_rate": 0.00018434206922601106, + "loss": 1.0162, + "step": 4072 + }, + { + "epoch": 0.7252492877492878, + "grad_norm": 0.5744053721427917, + "learning_rate": 0.00018433454820609833, + "loss": 1.2774, + "step": 4073 + }, + { + "epoch": 0.7254273504273504, + "grad_norm": 0.6210703253746033, + "learning_rate": 0.0001843270255338253, + "loss": 1.2526, + "step": 4074 + }, + { + "epoch": 0.7256054131054132, + "grad_norm": 0.49684277176856995, + "learning_rate": 0.0001843195012093394, + "loss": 1.0786, + "step": 4075 + }, + { + "epoch": 0.7257834757834758, + "grad_norm": 0.5851606130599976, + "learning_rate": 0.00018431197523278802, + "loss": 1.14, + "step": 4076 + }, + { + "epoch": 0.7259615384615384, + "grad_norm": 0.5494425296783447, + "learning_rate": 0.00018430444760431862, + "loss": 1.211, + "step": 4077 + }, + { + "epoch": 0.7261396011396012, + "grad_norm": 0.5247658491134644, + "learning_rate": 0.00018429691832407867, + "loss": 0.8031, + "step": 4078 + }, + { + "epoch": 0.7263176638176638, + "grad_norm": 0.5012249946594238, + "learning_rate": 0.00018428938739221574, + "loss": 1.1258, + "step": 4079 + }, + { + "epoch": 0.7264957264957265, + "grad_norm": 0.5226427912712097, + "learning_rate": 0.0001842818548088774, + "loss": 1.0029, + "step": 4080 + }, + { + "epoch": 0.7266737891737892, + "grad_norm": 0.45008543133735657, + "learning_rate": 0.00018427432057421114, + "loss": 1.0681, + "step": 4081 + }, + { + "epoch": 0.7268518518518519, + "grad_norm": 0.5127285122871399, + "learning_rate": 0.00018426678468836467, + "loss": 1.1069, + "step": 4082 + }, + { + "epoch": 0.7270299145299145, + "grad_norm": 0.5406150221824646, + "learning_rate": 0.0001842592471514856, + "loss": 1.052, + "step": 4083 + }, + { + "epoch": 0.7272079772079773, + "grad_norm": 0.5001157522201538, + "learning_rate": 0.0001842517079637216, + "loss": 0.9157, + "step": 4084 + }, + { + "epoch": 0.7273860398860399, + "grad_norm": 0.6169779300689697, + "learning_rate": 0.00018424416712522042, + "loss": 1.3133, + "step": 4085 + }, + { + "epoch": 0.7275641025641025, + "grad_norm": 0.4891316890716553, + "learning_rate": 0.00018423662463612974, + "loss": 0.9505, + "step": 4086 + }, + { + "epoch": 0.7277421652421653, + "grad_norm": 0.5883708596229553, + "learning_rate": 0.00018422908049659743, + "loss": 1.2797, + "step": 4087 + }, + { + "epoch": 0.7279202279202279, + "grad_norm": 0.6679072976112366, + "learning_rate": 0.00018422153470677125, + "loss": 1.1096, + "step": 4088 + }, + { + "epoch": 0.7280982905982906, + "grad_norm": 0.5178479552268982, + "learning_rate": 0.00018421398726679904, + "loss": 1.0299, + "step": 4089 + }, + { + "epoch": 0.7282763532763533, + "grad_norm": 0.6343900561332703, + "learning_rate": 0.0001842064381768287, + "loss": 1.2983, + "step": 4090 + }, + { + "epoch": 0.728454415954416, + "grad_norm": 0.43816515803337097, + "learning_rate": 0.0001841988874370081, + "loss": 0.9452, + "step": 4091 + }, + { + "epoch": 0.7286324786324786, + "grad_norm": 0.579790472984314, + "learning_rate": 0.00018419133504748528, + "loss": 1.1037, + "step": 4092 + }, + { + "epoch": 0.7288105413105413, + "grad_norm": 0.571374773979187, + "learning_rate": 0.00018418378100840807, + "loss": 1.1655, + "step": 4093 + }, + { + "epoch": 0.728988603988604, + "grad_norm": 0.5163514018058777, + "learning_rate": 0.0001841762253199246, + "loss": 1.1579, + "step": 4094 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 0.6553022265434265, + "learning_rate": 0.0001841686679821828, + "loss": 0.9664, + "step": 4095 + }, + { + "epoch": 0.7293447293447294, + "grad_norm": 0.5072969198226929, + "learning_rate": 0.00018416110899533084, + "loss": 0.9416, + "step": 4096 + }, + { + "epoch": 0.729522792022792, + "grad_norm": 0.5103251338005066, + "learning_rate": 0.00018415354835951675, + "loss": 1.0715, + "step": 4097 + }, + { + "epoch": 0.7297008547008547, + "grad_norm": 0.49752289056777954, + "learning_rate": 0.00018414598607488874, + "loss": 1.1848, + "step": 4098 + }, + { + "epoch": 0.7298789173789174, + "grad_norm": 0.5361882448196411, + "learning_rate": 0.00018413842214159488, + "loss": 1.1035, + "step": 4099 + }, + { + "epoch": 0.73005698005698, + "grad_norm": 0.5167670249938965, + "learning_rate": 0.00018413085655978343, + "loss": 1.0015, + "step": 4100 + }, + { + "epoch": 0.7302350427350427, + "grad_norm": 0.5930629372596741, + "learning_rate": 0.00018412328932960263, + "loss": 0.9766, + "step": 4101 + }, + { + "epoch": 0.7304131054131054, + "grad_norm": 0.5234778523445129, + "learning_rate": 0.00018411572045120073, + "loss": 1.0317, + "step": 4102 + }, + { + "epoch": 0.7305911680911681, + "grad_norm": 0.5361374020576477, + "learning_rate": 0.000184108149924726, + "loss": 1.1228, + "step": 4103 + }, + { + "epoch": 0.7307692307692307, + "grad_norm": 0.5845770239830017, + "learning_rate": 0.0001841005777503268, + "loss": 0.9541, + "step": 4104 + }, + { + "epoch": 0.7309472934472935, + "grad_norm": 0.49320483207702637, + "learning_rate": 0.0001840930039281515, + "loss": 0.9445, + "step": 4105 + }, + { + "epoch": 0.7311253561253561, + "grad_norm": 0.5391250252723694, + "learning_rate": 0.00018408542845834845, + "loss": 1.1983, + "step": 4106 + }, + { + "epoch": 0.7313034188034188, + "grad_norm": 0.4890393316745758, + "learning_rate": 0.00018407785134106613, + "loss": 0.8353, + "step": 4107 + }, + { + "epoch": 0.7314814814814815, + "grad_norm": 0.5839747190475464, + "learning_rate": 0.00018407027257645296, + "loss": 1.4074, + "step": 4108 + }, + { + "epoch": 0.7316595441595442, + "grad_norm": 0.5957708358764648, + "learning_rate": 0.0001840626921646574, + "loss": 1.1032, + "step": 4109 + }, + { + "epoch": 0.7318376068376068, + "grad_norm": 0.5029017925262451, + "learning_rate": 0.00018405511010582805, + "loss": 1.095, + "step": 4110 + }, + { + "epoch": 0.7320156695156695, + "grad_norm": 0.6054347157478333, + "learning_rate": 0.00018404752640011345, + "loss": 1.0366, + "step": 4111 + }, + { + "epoch": 0.7321937321937322, + "grad_norm": 0.5476830005645752, + "learning_rate": 0.00018403994104766212, + "loss": 1.0976, + "step": 4112 + }, + { + "epoch": 0.7323717948717948, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00018403235404862277, + "loss": 1.0809, + "step": 4113 + }, + { + "epoch": 0.7325498575498576, + "grad_norm": 0.5119251012802124, + "learning_rate": 0.00018402476540314394, + "loss": 1.0176, + "step": 4114 + }, + { + "epoch": 0.7327279202279202, + "grad_norm": 0.5825830698013306, + "learning_rate": 0.00018401717511137445, + "loss": 1.2357, + "step": 4115 + }, + { + "epoch": 0.7329059829059829, + "grad_norm": 0.5702941417694092, + "learning_rate": 0.0001840095831734629, + "loss": 1.1549, + "step": 4116 + }, + { + "epoch": 0.7330840455840456, + "grad_norm": 0.5660699605941772, + "learning_rate": 0.00018400198958955807, + "loss": 1.1778, + "step": 4117 + }, + { + "epoch": 0.7332621082621082, + "grad_norm": 0.5241161584854126, + "learning_rate": 0.0001839943943598088, + "loss": 0.8587, + "step": 4118 + }, + { + "epoch": 0.7334401709401709, + "grad_norm": 0.581194281578064, + "learning_rate": 0.0001839867974843638, + "loss": 1.2169, + "step": 4119 + }, + { + "epoch": 0.7336182336182336, + "grad_norm": 0.4342379570007324, + "learning_rate": 0.00018397919896337198, + "loss": 0.9182, + "step": 4120 + }, + { + "epoch": 0.7337962962962963, + "grad_norm": 0.5708567500114441, + "learning_rate": 0.00018397159879698224, + "loss": 1.1781, + "step": 4121 + }, + { + "epoch": 0.7339743589743589, + "grad_norm": 0.5827265977859497, + "learning_rate": 0.00018396399698534344, + "loss": 1.2905, + "step": 4122 + }, + { + "epoch": 0.7341524216524217, + "grad_norm": 0.5274056792259216, + "learning_rate": 0.00018395639352860457, + "loss": 1.1786, + "step": 4123 + }, + { + "epoch": 0.7343304843304843, + "grad_norm": 0.5094266533851624, + "learning_rate": 0.00018394878842691452, + "loss": 1.2016, + "step": 4124 + }, + { + "epoch": 0.7345085470085471, + "grad_norm": 0.48779475688934326, + "learning_rate": 0.0001839411816804224, + "loss": 1.0562, + "step": 4125 + }, + { + "epoch": 0.7346866096866097, + "grad_norm": 0.5805709958076477, + "learning_rate": 0.00018393357328927716, + "loss": 1.1705, + "step": 4126 + }, + { + "epoch": 0.7348646723646723, + "grad_norm": 0.4910700023174286, + "learning_rate": 0.00018392596325362791, + "loss": 1.0682, + "step": 4127 + }, + { + "epoch": 0.7350427350427351, + "grad_norm": 0.5297428369522095, + "learning_rate": 0.0001839183515736238, + "loss": 0.9505, + "step": 4128 + }, + { + "epoch": 0.7352207977207977, + "grad_norm": 0.45442086458206177, + "learning_rate": 0.00018391073824941385, + "loss": 0.9548, + "step": 4129 + }, + { + "epoch": 0.7353988603988604, + "grad_norm": 0.49299946427345276, + "learning_rate": 0.00018390312328114733, + "loss": 1.0868, + "step": 4130 + }, + { + "epoch": 0.7355769230769231, + "grad_norm": 0.4839940369129181, + "learning_rate": 0.0001838955066689734, + "loss": 0.9565, + "step": 4131 + }, + { + "epoch": 0.7357549857549858, + "grad_norm": 0.48600608110427856, + "learning_rate": 0.00018388788841304128, + "loss": 1.2353, + "step": 4132 + }, + { + "epoch": 0.7359330484330484, + "grad_norm": 0.4893583357334137, + "learning_rate": 0.0001838802685135003, + "loss": 0.9595, + "step": 4133 + }, + { + "epoch": 0.7361111111111112, + "grad_norm": 0.4587398171424866, + "learning_rate": 0.00018387264697049963, + "loss": 1.1222, + "step": 4134 + }, + { + "epoch": 0.7362891737891738, + "grad_norm": 0.5361055731773376, + "learning_rate": 0.00018386502378418872, + "loss": 1.3304, + "step": 4135 + }, + { + "epoch": 0.7364672364672364, + "grad_norm": 0.5556629300117493, + "learning_rate": 0.00018385739895471686, + "loss": 1.0358, + "step": 4136 + }, + { + "epoch": 0.7366452991452992, + "grad_norm": 0.45555856823921204, + "learning_rate": 0.00018384977248223346, + "loss": 1.0081, + "step": 4137 + }, + { + "epoch": 0.7368233618233618, + "grad_norm": 0.5606052875518799, + "learning_rate": 0.00018384214436688797, + "loss": 0.9367, + "step": 4138 + }, + { + "epoch": 0.7370014245014245, + "grad_norm": 0.5428356528282166, + "learning_rate": 0.00018383451460882982, + "loss": 1.1391, + "step": 4139 + }, + { + "epoch": 0.7371794871794872, + "grad_norm": 0.4891330897808075, + "learning_rate": 0.00018382688320820853, + "loss": 0.9805, + "step": 4140 + }, + { + "epoch": 0.7373575498575499, + "grad_norm": 0.5407996773719788, + "learning_rate": 0.0001838192501651736, + "loss": 1.0532, + "step": 4141 + }, + { + "epoch": 0.7375356125356125, + "grad_norm": 0.5241971611976624, + "learning_rate": 0.00018381161547987454, + "loss": 0.9509, + "step": 4142 + }, + { + "epoch": 0.7377136752136753, + "grad_norm": 0.5370210409164429, + "learning_rate": 0.000183803979152461, + "loss": 1.2342, + "step": 4143 + }, + { + "epoch": 0.7378917378917379, + "grad_norm": 0.5470060706138611, + "learning_rate": 0.00018379634118308259, + "loss": 0.9621, + "step": 4144 + }, + { + "epoch": 0.7380698005698005, + "grad_norm": 0.546313464641571, + "learning_rate": 0.00018378870157188893, + "loss": 1.1253, + "step": 4145 + }, + { + "epoch": 0.7382478632478633, + "grad_norm": 0.502027153968811, + "learning_rate": 0.00018378106031902974, + "loss": 1.1919, + "step": 4146 + }, + { + "epoch": 0.7384259259259259, + "grad_norm": 0.5282283425331116, + "learning_rate": 0.0001837734174246547, + "loss": 1.0088, + "step": 4147 + }, + { + "epoch": 0.7386039886039886, + "grad_norm": 0.5152897238731384, + "learning_rate": 0.00018376577288891355, + "loss": 1.0813, + "step": 4148 + }, + { + "epoch": 0.7387820512820513, + "grad_norm": 0.5002804398536682, + "learning_rate": 0.0001837581267119561, + "loss": 0.9797, + "step": 4149 + }, + { + "epoch": 0.738960113960114, + "grad_norm": 0.5698176026344299, + "learning_rate": 0.00018375047889393215, + "loss": 1.1099, + "step": 4150 + }, + { + "epoch": 0.7391381766381766, + "grad_norm": 0.5384604930877686, + "learning_rate": 0.00018374282943499156, + "loss": 1.1944, + "step": 4151 + }, + { + "epoch": 0.7393162393162394, + "grad_norm": 0.5483044385910034, + "learning_rate": 0.00018373517833528418, + "loss": 1.1734, + "step": 4152 + }, + { + "epoch": 0.739494301994302, + "grad_norm": 0.4824066162109375, + "learning_rate": 0.0001837275255949599, + "loss": 0.9515, + "step": 4153 + }, + { + "epoch": 0.7396723646723646, + "grad_norm": 0.45413634181022644, + "learning_rate": 0.00018371987121416873, + "loss": 0.7534, + "step": 4154 + }, + { + "epoch": 0.7398504273504274, + "grad_norm": 0.5874246954917908, + "learning_rate": 0.00018371221519306055, + "loss": 0.9464, + "step": 4155 + }, + { + "epoch": 0.74002849002849, + "grad_norm": 0.5219913125038147, + "learning_rate": 0.00018370455753178544, + "loss": 1.0494, + "step": 4156 + }, + { + "epoch": 0.7402065527065527, + "grad_norm": 0.5937709212303162, + "learning_rate": 0.00018369689823049341, + "loss": 1.0529, + "step": 4157 + }, + { + "epoch": 0.7403846153846154, + "grad_norm": 0.5204295516014099, + "learning_rate": 0.00018368923728933449, + "loss": 1.0602, + "step": 4158 + }, + { + "epoch": 0.7405626780626781, + "grad_norm": 0.5422890186309814, + "learning_rate": 0.00018368157470845885, + "loss": 0.9261, + "step": 4159 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.6163852214813232, + "learning_rate": 0.00018367391048801655, + "loss": 1.2771, + "step": 4160 + }, + { + "epoch": 0.7409188034188035, + "grad_norm": 0.5070751309394836, + "learning_rate": 0.00018366624462815785, + "loss": 1.0401, + "step": 4161 + }, + { + "epoch": 0.7410968660968661, + "grad_norm": 0.4477100968360901, + "learning_rate": 0.00018365857712903283, + "loss": 1.1463, + "step": 4162 + }, + { + "epoch": 0.7412749287749287, + "grad_norm": 0.5421462655067444, + "learning_rate": 0.0001836509079907918, + "loss": 0.9373, + "step": 4163 + }, + { + "epoch": 0.7414529914529915, + "grad_norm": 0.6162141561508179, + "learning_rate": 0.000183643237213585, + "loss": 1.1827, + "step": 4164 + }, + { + "epoch": 0.7416310541310541, + "grad_norm": 0.5653836131095886, + "learning_rate": 0.00018363556479756272, + "loss": 1.0689, + "step": 4165 + }, + { + "epoch": 0.7418091168091168, + "grad_norm": 0.57053542137146, + "learning_rate": 0.00018362789074287527, + "loss": 1.0289, + "step": 4166 + }, + { + "epoch": 0.7419871794871795, + "grad_norm": 0.5603055953979492, + "learning_rate": 0.00018362021504967304, + "loss": 1.1926, + "step": 4167 + }, + { + "epoch": 0.7421652421652422, + "grad_norm": 0.5460166335105896, + "learning_rate": 0.0001836125377181064, + "loss": 1.1488, + "step": 4168 + }, + { + "epoch": 0.7423433048433048, + "grad_norm": 0.5097107887268066, + "learning_rate": 0.00018360485874832579, + "loss": 1.0781, + "step": 4169 + }, + { + "epoch": 0.7425213675213675, + "grad_norm": 0.6280624270439148, + "learning_rate": 0.00018359717814048164, + "loss": 1.3625, + "step": 4170 + }, + { + "epoch": 0.7426994301994302, + "grad_norm": 0.4528210759162903, + "learning_rate": 0.0001835894958947244, + "loss": 0.8417, + "step": 4171 + }, + { + "epoch": 0.7428774928774928, + "grad_norm": 0.48735132813453674, + "learning_rate": 0.00018358181201120468, + "loss": 0.9544, + "step": 4172 + }, + { + "epoch": 0.7430555555555556, + "grad_norm": 0.48388174176216125, + "learning_rate": 0.00018357412649007296, + "loss": 1.0663, + "step": 4173 + }, + { + "epoch": 0.7432336182336182, + "grad_norm": 0.5435357689857483, + "learning_rate": 0.00018356643933147986, + "loss": 1.2074, + "step": 4174 + }, + { + "epoch": 0.7434116809116809, + "grad_norm": 0.49890074133872986, + "learning_rate": 0.00018355875053557594, + "loss": 1.1322, + "step": 4175 + }, + { + "epoch": 0.7435897435897436, + "grad_norm": 0.5680708885192871, + "learning_rate": 0.0001835510601025119, + "loss": 1.1964, + "step": 4176 + }, + { + "epoch": 0.7437678062678063, + "grad_norm": 0.5002360939979553, + "learning_rate": 0.00018354336803243842, + "loss": 1.1396, + "step": 4177 + }, + { + "epoch": 0.7439458689458689, + "grad_norm": 0.5202965140342712, + "learning_rate": 0.00018353567432550616, + "loss": 1.1498, + "step": 4178 + }, + { + "epoch": 0.7441239316239316, + "grad_norm": 0.514492928981781, + "learning_rate": 0.00018352797898186588, + "loss": 1.0959, + "step": 4179 + }, + { + "epoch": 0.7443019943019943, + "grad_norm": 0.6395383477210999, + "learning_rate": 0.0001835202820016684, + "loss": 1.2867, + "step": 4180 + }, + { + "epoch": 0.7444800569800569, + "grad_norm": 0.5489062070846558, + "learning_rate": 0.00018351258338506447, + "loss": 1.1638, + "step": 4181 + }, + { + "epoch": 0.7446581196581197, + "grad_norm": 0.5705671906471252, + "learning_rate": 0.00018350488313220498, + "loss": 0.9493, + "step": 4182 + }, + { + "epoch": 0.7448361823361823, + "grad_norm": 0.5404297709465027, + "learning_rate": 0.00018349718124324076, + "loss": 0.9876, + "step": 4183 + }, + { + "epoch": 0.7450142450142451, + "grad_norm": 0.5841003060340881, + "learning_rate": 0.0001834894777183227, + "loss": 1.1225, + "step": 4184 + }, + { + "epoch": 0.7451923076923077, + "grad_norm": 0.49774688482284546, + "learning_rate": 0.00018348177255760178, + "loss": 1.1442, + "step": 4185 + }, + { + "epoch": 0.7453703703703703, + "grad_norm": 0.5212422609329224, + "learning_rate": 0.00018347406576122894, + "loss": 1.101, + "step": 4186 + }, + { + "epoch": 0.7455484330484331, + "grad_norm": 0.615024983882904, + "learning_rate": 0.00018346635732935517, + "loss": 1.4188, + "step": 4187 + }, + { + "epoch": 0.7457264957264957, + "grad_norm": 0.46818843483924866, + "learning_rate": 0.00018345864726213154, + "loss": 1.0071, + "step": 4188 + }, + { + "epoch": 0.7459045584045584, + "grad_norm": 0.4921121895313263, + "learning_rate": 0.00018345093555970906, + "loss": 1.015, + "step": 4189 + }, + { + "epoch": 0.7460826210826211, + "grad_norm": 0.5042136311531067, + "learning_rate": 0.00018344322222223889, + "loss": 0.9974, + "step": 4190 + }, + { + "epoch": 0.7462606837606838, + "grad_norm": 0.5872490406036377, + "learning_rate": 0.0001834355072498721, + "loss": 1.3166, + "step": 4191 + }, + { + "epoch": 0.7464387464387464, + "grad_norm": 0.559117317199707, + "learning_rate": 0.00018342779064275984, + "loss": 1.2227, + "step": 4192 + }, + { + "epoch": 0.7466168091168092, + "grad_norm": 0.5269635319709778, + "learning_rate": 0.00018342007240105336, + "loss": 1.0281, + "step": 4193 + }, + { + "epoch": 0.7467948717948718, + "grad_norm": 0.4608335793018341, + "learning_rate": 0.00018341235252490387, + "loss": 0.98, + "step": 4194 + }, + { + "epoch": 0.7469729344729344, + "grad_norm": 0.5818259119987488, + "learning_rate": 0.00018340463101446255, + "loss": 1.1544, + "step": 4195 + }, + { + "epoch": 0.7471509971509972, + "grad_norm": 0.5577529668807983, + "learning_rate": 0.00018339690786988079, + "loss": 1.3059, + "step": 4196 + }, + { + "epoch": 0.7473290598290598, + "grad_norm": 0.5430468320846558, + "learning_rate": 0.00018338918309130983, + "loss": 1.2766, + "step": 4197 + }, + { + "epoch": 0.7475071225071225, + "grad_norm": 0.4941701591014862, + "learning_rate": 0.0001833814566789011, + "loss": 1.193, + "step": 4198 + }, + { + "epoch": 0.7476851851851852, + "grad_norm": 0.5471884608268738, + "learning_rate": 0.00018337372863280589, + "loss": 1.2261, + "step": 4199 + }, + { + "epoch": 0.7478632478632479, + "grad_norm": 0.4641438126564026, + "learning_rate": 0.0001833659989531757, + "loss": 0.7953, + "step": 4200 + }, + { + "epoch": 0.7480413105413105, + "grad_norm": 0.5244714617729187, + "learning_rate": 0.0001833582676401619, + "loss": 0.9344, + "step": 4201 + }, + { + "epoch": 0.7482193732193733, + "grad_norm": 0.5964360237121582, + "learning_rate": 0.00018335053469391603, + "loss": 1.2072, + "step": 4202 + }, + { + "epoch": 0.7483974358974359, + "grad_norm": 0.4929158091545105, + "learning_rate": 0.00018334280011458954, + "loss": 1.2183, + "step": 4203 + }, + { + "epoch": 0.7485754985754985, + "grad_norm": 0.46221864223480225, + "learning_rate": 0.00018333506390233405, + "loss": 1.1957, + "step": 4204 + }, + { + "epoch": 0.7487535612535613, + "grad_norm": 0.6301732659339905, + "learning_rate": 0.0001833273260573011, + "loss": 1.0582, + "step": 4205 + }, + { + "epoch": 0.7489316239316239, + "grad_norm": 0.5606021881103516, + "learning_rate": 0.0001833195865796423, + "loss": 1.4034, + "step": 4206 + }, + { + "epoch": 0.7491096866096866, + "grad_norm": 0.44856077432632446, + "learning_rate": 0.00018331184546950926, + "loss": 0.8421, + "step": 4207 + }, + { + "epoch": 0.7492877492877493, + "grad_norm": 0.5487226247787476, + "learning_rate": 0.00018330410272705366, + "loss": 1.238, + "step": 4208 + }, + { + "epoch": 0.749465811965812, + "grad_norm": 0.6043636798858643, + "learning_rate": 0.00018329635835242724, + "loss": 1.1215, + "step": 4209 + }, + { + "epoch": 0.7496438746438746, + "grad_norm": 0.5145319104194641, + "learning_rate": 0.00018328861234578173, + "loss": 1.1002, + "step": 4210 + }, + { + "epoch": 0.7498219373219374, + "grad_norm": 0.5667078495025635, + "learning_rate": 0.00018328086470726884, + "loss": 1.2994, + "step": 4211 + }, + { + "epoch": 0.75, + "grad_norm": 0.5117634534835815, + "learning_rate": 0.00018327311543704043, + "loss": 0.9448, + "step": 4212 + }, + { + "epoch": 0.75, + "eval_loss": 1.0982474088668823, + "eval_runtime": 24.6617, + "eval_samples_per_second": 42.211, + "eval_steps_per_second": 21.126, + "step": 4212 + }, + { + "epoch": 0.7501780626780626, + "grad_norm": 0.5451585054397583, + "learning_rate": 0.00018326536453524826, + "loss": 0.9023, + "step": 4213 + }, + { + "epoch": 0.7503561253561254, + "grad_norm": 0.6585208773612976, + "learning_rate": 0.0001832576120020443, + "loss": 1.2798, + "step": 4214 + }, + { + "epoch": 0.750534188034188, + "grad_norm": 0.6444812417030334, + "learning_rate": 0.00018324985783758037, + "loss": 1.3999, + "step": 4215 + }, + { + "epoch": 0.7507122507122507, + "grad_norm": 0.6178330779075623, + "learning_rate": 0.0001832421020420084, + "loss": 1.1846, + "step": 4216 + }, + { + "epoch": 0.7508903133903134, + "grad_norm": 0.509969174861908, + "learning_rate": 0.00018323434461548036, + "loss": 1.1831, + "step": 4217 + }, + { + "epoch": 0.7510683760683761, + "grad_norm": 0.5558911561965942, + "learning_rate": 0.00018322658555814826, + "loss": 1.1599, + "step": 4218 + }, + { + "epoch": 0.7512464387464387, + "grad_norm": 0.5714917778968811, + "learning_rate": 0.0001832188248701641, + "loss": 0.9702, + "step": 4219 + }, + { + "epoch": 0.7514245014245015, + "grad_norm": 0.6136442422866821, + "learning_rate": 0.00018321106255167995, + "loss": 0.9376, + "step": 4220 + }, + { + "epoch": 0.7516025641025641, + "grad_norm": 0.5832077264785767, + "learning_rate": 0.00018320329860284785, + "loss": 1.2564, + "step": 4221 + }, + { + "epoch": 0.7517806267806267, + "grad_norm": 0.45330923795700073, + "learning_rate": 0.00018319553302381997, + "loss": 0.9321, + "step": 4222 + }, + { + "epoch": 0.7519586894586895, + "grad_norm": 0.5278468132019043, + "learning_rate": 0.00018318776581474847, + "loss": 1.1334, + "step": 4223 + }, + { + "epoch": 0.7521367521367521, + "grad_norm": 0.49267473816871643, + "learning_rate": 0.00018317999697578549, + "loss": 1.1577, + "step": 4224 + }, + { + "epoch": 0.7523148148148148, + "grad_norm": 0.5372124314308167, + "learning_rate": 0.00018317222650708325, + "loss": 1.037, + "step": 4225 + }, + { + "epoch": 0.7524928774928775, + "grad_norm": 0.5879829525947571, + "learning_rate": 0.000183164454408794, + "loss": 1.1312, + "step": 4226 + }, + { + "epoch": 0.7526709401709402, + "grad_norm": 0.5363932251930237, + "learning_rate": 0.00018315668068107004, + "loss": 1.174, + "step": 4227 + }, + { + "epoch": 0.7528490028490028, + "grad_norm": 0.5585991740226746, + "learning_rate": 0.00018314890532406366, + "loss": 1.2106, + "step": 4228 + }, + { + "epoch": 0.7530270655270656, + "grad_norm": 0.49395787715911865, + "learning_rate": 0.0001831411283379272, + "loss": 1.1163, + "step": 4229 + }, + { + "epoch": 0.7532051282051282, + "grad_norm": 0.5081066489219666, + "learning_rate": 0.00018313334972281306, + "loss": 1.184, + "step": 4230 + }, + { + "epoch": 0.7533831908831908, + "grad_norm": 0.40304034948349, + "learning_rate": 0.0001831255694788736, + "loss": 0.7548, + "step": 4231 + }, + { + "epoch": 0.7535612535612536, + "grad_norm": 0.4999815821647644, + "learning_rate": 0.0001831177876062613, + "loss": 1.0092, + "step": 4232 + }, + { + "epoch": 0.7537393162393162, + "grad_norm": 0.48917025327682495, + "learning_rate": 0.00018311000410512862, + "loss": 1.0354, + "step": 4233 + }, + { + "epoch": 0.7539173789173789, + "grad_norm": 0.475606769323349, + "learning_rate": 0.00018310221897562806, + "loss": 0.8728, + "step": 4234 + }, + { + "epoch": 0.7540954415954416, + "grad_norm": 0.630439817905426, + "learning_rate": 0.00018309443221791214, + "loss": 1.1436, + "step": 4235 + }, + { + "epoch": 0.7542735042735043, + "grad_norm": 0.524740993976593, + "learning_rate": 0.00018308664383213344, + "loss": 1.0487, + "step": 4236 + }, + { + "epoch": 0.7544515669515669, + "grad_norm": 0.4734523892402649, + "learning_rate": 0.0001830788538184445, + "loss": 1.0681, + "step": 4237 + }, + { + "epoch": 0.7546296296296297, + "grad_norm": 0.5767266750335693, + "learning_rate": 0.00018307106217699807, + "loss": 1.0599, + "step": 4238 + }, + { + "epoch": 0.7548076923076923, + "grad_norm": 0.6276642084121704, + "learning_rate": 0.0001830632689079467, + "loss": 1.2837, + "step": 4239 + }, + { + "epoch": 0.7549857549857549, + "grad_norm": 0.5539988279342651, + "learning_rate": 0.00018305547401144316, + "loss": 0.9072, + "step": 4240 + }, + { + "epoch": 0.7551638176638177, + "grad_norm": 0.4551292061805725, + "learning_rate": 0.00018304767748764014, + "loss": 1.0204, + "step": 4241 + }, + { + "epoch": 0.7553418803418803, + "grad_norm": 0.47344550490379333, + "learning_rate": 0.00018303987933669034, + "loss": 1.0473, + "step": 4242 + }, + { + "epoch": 0.7555199430199431, + "grad_norm": 0.6050213575363159, + "learning_rate": 0.00018303207955874665, + "loss": 1.1552, + "step": 4243 + }, + { + "epoch": 0.7556980056980057, + "grad_norm": 0.48943889141082764, + "learning_rate": 0.00018302427815396186, + "loss": 1.0002, + "step": 4244 + }, + { + "epoch": 0.7558760683760684, + "grad_norm": 0.5664682984352112, + "learning_rate": 0.00018301647512248878, + "loss": 1.1865, + "step": 4245 + }, + { + "epoch": 0.7560541310541311, + "grad_norm": 0.5702242255210876, + "learning_rate": 0.00018300867046448034, + "loss": 1.3029, + "step": 4246 + }, + { + "epoch": 0.7562321937321937, + "grad_norm": 0.593207836151123, + "learning_rate": 0.00018300086418008942, + "loss": 1.109, + "step": 4247 + }, + { + "epoch": 0.7564102564102564, + "grad_norm": 0.5887887477874756, + "learning_rate": 0.000182993056269469, + "loss": 1.3022, + "step": 4248 + }, + { + "epoch": 0.7565883190883191, + "grad_norm": 0.5277966260910034, + "learning_rate": 0.00018298524673277203, + "loss": 1.1738, + "step": 4249 + }, + { + "epoch": 0.7567663817663818, + "grad_norm": 0.589347779750824, + "learning_rate": 0.00018297743557015155, + "loss": 1.0185, + "step": 4250 + }, + { + "epoch": 0.7569444444444444, + "grad_norm": 0.49920859932899475, + "learning_rate": 0.0001829696227817606, + "loss": 1.118, + "step": 4251 + }, + { + "epoch": 0.7571225071225072, + "grad_norm": 0.502565324306488, + "learning_rate": 0.0001829618083677522, + "loss": 1.1856, + "step": 4252 + }, + { + "epoch": 0.7573005698005698, + "grad_norm": 0.49814435839653015, + "learning_rate": 0.00018295399232827955, + "loss": 1.0432, + "step": 4253 + }, + { + "epoch": 0.7574786324786325, + "grad_norm": 0.5087502598762512, + "learning_rate": 0.00018294617466349574, + "loss": 1.2325, + "step": 4254 + }, + { + "epoch": 0.7576566951566952, + "grad_norm": 0.5107288956642151, + "learning_rate": 0.00018293835537355394, + "loss": 1.0487, + "step": 4255 + }, + { + "epoch": 0.7578347578347578, + "grad_norm": 0.524725615978241, + "learning_rate": 0.00018293053445860732, + "loss": 1.1821, + "step": 4256 + }, + { + "epoch": 0.7580128205128205, + "grad_norm": 0.5234082937240601, + "learning_rate": 0.0001829227119188092, + "loss": 0.8896, + "step": 4257 + }, + { + "epoch": 0.7581908831908832, + "grad_norm": 0.5102918744087219, + "learning_rate": 0.00018291488775431275, + "loss": 1.0246, + "step": 4258 + }, + { + "epoch": 0.7583689458689459, + "grad_norm": 0.5552714467048645, + "learning_rate": 0.00018290706196527135, + "loss": 1.0193, + "step": 4259 + }, + { + "epoch": 0.7585470085470085, + "grad_norm": 0.5395022630691528, + "learning_rate": 0.00018289923455183825, + "loss": 1.3203, + "step": 4260 + }, + { + "epoch": 0.7587250712250713, + "grad_norm": 0.7474865913391113, + "learning_rate": 0.00018289140551416692, + "loss": 1.182, + "step": 4261 + }, + { + "epoch": 0.7589031339031339, + "grad_norm": 0.4892016649246216, + "learning_rate": 0.00018288357485241066, + "loss": 0.968, + "step": 4262 + }, + { + "epoch": 0.7590811965811965, + "grad_norm": 0.4627816081047058, + "learning_rate": 0.00018287574256672291, + "loss": 0.6895, + "step": 4263 + }, + { + "epoch": 0.7592592592592593, + "grad_norm": 0.6221280097961426, + "learning_rate": 0.00018286790865725715, + "loss": 0.9691, + "step": 4264 + }, + { + "epoch": 0.7594373219373219, + "grad_norm": 0.5542295575141907, + "learning_rate": 0.0001828600731241669, + "loss": 0.9996, + "step": 4265 + }, + { + "epoch": 0.7596153846153846, + "grad_norm": 0.5570770502090454, + "learning_rate": 0.00018285223596760562, + "loss": 1.1996, + "step": 4266 + }, + { + "epoch": 0.7597934472934473, + "grad_norm": 0.5495262742042542, + "learning_rate": 0.00018284439718772687, + "loss": 1.1572, + "step": 4267 + }, + { + "epoch": 0.75997150997151, + "grad_norm": 0.5006741881370544, + "learning_rate": 0.00018283655678468427, + "loss": 1.1215, + "step": 4268 + }, + { + "epoch": 0.7601495726495726, + "grad_norm": 0.4682157635688782, + "learning_rate": 0.00018282871475863144, + "loss": 1.0547, + "step": 4269 + }, + { + "epoch": 0.7603276353276354, + "grad_norm": 0.6275840997695923, + "learning_rate": 0.00018282087110972197, + "loss": 1.3855, + "step": 4270 + }, + { + "epoch": 0.760505698005698, + "grad_norm": 0.5341474413871765, + "learning_rate": 0.0001828130258381096, + "loss": 1.2024, + "step": 4271 + }, + { + "epoch": 0.7606837606837606, + "grad_norm": 0.4330833852291107, + "learning_rate": 0.000182805178943948, + "loss": 1.0508, + "step": 4272 + }, + { + "epoch": 0.7608618233618234, + "grad_norm": 0.6276537179946899, + "learning_rate": 0.00018279733042739094, + "loss": 1.1635, + "step": 4273 + }, + { + "epoch": 0.761039886039886, + "grad_norm": 0.5370199084281921, + "learning_rate": 0.00018278948028859217, + "loss": 1.0579, + "step": 4274 + }, + { + "epoch": 0.7612179487179487, + "grad_norm": 0.524959921836853, + "learning_rate": 0.00018278162852770552, + "loss": 1.0972, + "step": 4275 + }, + { + "epoch": 0.7613960113960114, + "grad_norm": 0.5029389262199402, + "learning_rate": 0.00018277377514488486, + "loss": 0.959, + "step": 4276 + }, + { + "epoch": 0.7615740740740741, + "grad_norm": 0.49772894382476807, + "learning_rate": 0.00018276592014028397, + "loss": 1.2773, + "step": 4277 + }, + { + "epoch": 0.7617521367521367, + "grad_norm": 0.5195719003677368, + "learning_rate": 0.00018275806351405685, + "loss": 1.0676, + "step": 4278 + }, + { + "epoch": 0.7619301994301995, + "grad_norm": 0.5167942643165588, + "learning_rate": 0.00018275020526635735, + "loss": 1.0615, + "step": 4279 + }, + { + "epoch": 0.7621082621082621, + "grad_norm": 0.4958035945892334, + "learning_rate": 0.0001827423453973395, + "loss": 0.9605, + "step": 4280 + }, + { + "epoch": 0.7622863247863247, + "grad_norm": 0.6256808042526245, + "learning_rate": 0.00018273448390715728, + "loss": 1.2526, + "step": 4281 + }, + { + "epoch": 0.7624643874643875, + "grad_norm": 0.5062580108642578, + "learning_rate": 0.0001827266207959647, + "loss": 1.0604, + "step": 4282 + }, + { + "epoch": 0.7626424501424501, + "grad_norm": 0.5080778002738953, + "learning_rate": 0.00018271875606391583, + "loss": 1.1246, + "step": 4283 + }, + { + "epoch": 0.7628205128205128, + "grad_norm": 0.5069389939308167, + "learning_rate": 0.00018271088971116479, + "loss": 1.3158, + "step": 4284 + }, + { + "epoch": 0.7629985754985755, + "grad_norm": 0.7280121445655823, + "learning_rate": 0.00018270302173786567, + "loss": 1.2066, + "step": 4285 + }, + { + "epoch": 0.7631766381766382, + "grad_norm": 0.6523470282554626, + "learning_rate": 0.00018269515214417267, + "loss": 1.3236, + "step": 4286 + }, + { + "epoch": 0.7633547008547008, + "grad_norm": 0.5799322724342346, + "learning_rate": 0.00018268728093023988, + "loss": 0.9786, + "step": 4287 + }, + { + "epoch": 0.7635327635327636, + "grad_norm": 0.46675166487693787, + "learning_rate": 0.00018267940809622163, + "loss": 0.8131, + "step": 4288 + }, + { + "epoch": 0.7637108262108262, + "grad_norm": 0.5566182732582092, + "learning_rate": 0.00018267153364227214, + "loss": 1.0565, + "step": 4289 + }, + { + "epoch": 0.7638888888888888, + "grad_norm": 0.532028079032898, + "learning_rate": 0.00018266365756854566, + "loss": 0.952, + "step": 4290 + }, + { + "epoch": 0.7640669515669516, + "grad_norm": 0.5082666873931885, + "learning_rate": 0.00018265577987519653, + "loss": 1.0704, + "step": 4291 + }, + { + "epoch": 0.7642450142450142, + "grad_norm": 0.5223562717437744, + "learning_rate": 0.00018264790056237912, + "loss": 1.1161, + "step": 4292 + }, + { + "epoch": 0.7644230769230769, + "grad_norm": 0.48472318053245544, + "learning_rate": 0.00018264001963024778, + "loss": 0.8784, + "step": 4293 + }, + { + "epoch": 0.7646011396011396, + "grad_norm": 0.5901281833648682, + "learning_rate": 0.0001826321370789569, + "loss": 1.1031, + "step": 4294 + }, + { + "epoch": 0.7647792022792023, + "grad_norm": 0.570350706577301, + "learning_rate": 0.000182624252908661, + "loss": 0.9047, + "step": 4295 + }, + { + "epoch": 0.7649572649572649, + "grad_norm": 0.568373441696167, + "learning_rate": 0.00018261636711951445, + "loss": 1.0106, + "step": 4296 + }, + { + "epoch": 0.7651353276353277, + "grad_norm": 0.6175880432128906, + "learning_rate": 0.00018260847971167182, + "loss": 1.3531, + "step": 4297 + }, + { + "epoch": 0.7653133903133903, + "grad_norm": 0.5682594776153564, + "learning_rate": 0.00018260059068528762, + "loss": 1.1261, + "step": 4298 + }, + { + "epoch": 0.7654914529914529, + "grad_norm": 0.5050225257873535, + "learning_rate": 0.00018259270004051644, + "loss": 1.0921, + "step": 4299 + }, + { + "epoch": 0.7656695156695157, + "grad_norm": 0.5416565537452698, + "learning_rate": 0.0001825848077775129, + "loss": 1.0881, + "step": 4300 + }, + { + "epoch": 0.7658475783475783, + "grad_norm": 0.5418867468833923, + "learning_rate": 0.0001825769138964316, + "loss": 1.2069, + "step": 4301 + }, + { + "epoch": 0.7660256410256411, + "grad_norm": 0.5447866320610046, + "learning_rate": 0.00018256901839742718, + "loss": 1.1827, + "step": 4302 + }, + { + "epoch": 0.7662037037037037, + "grad_norm": 0.5482802987098694, + "learning_rate": 0.00018256112128065439, + "loss": 1.0492, + "step": 4303 + }, + { + "epoch": 0.7663817663817664, + "grad_norm": 0.5059601664543152, + "learning_rate": 0.0001825532225462679, + "loss": 1.0996, + "step": 4304 + }, + { + "epoch": 0.7665598290598291, + "grad_norm": 0.5153701901435852, + "learning_rate": 0.00018254532219442258, + "loss": 1.3237, + "step": 4305 + }, + { + "epoch": 0.7667378917378918, + "grad_norm": 0.5370768904685974, + "learning_rate": 0.0001825374202252731, + "loss": 0.9925, + "step": 4306 + }, + { + "epoch": 0.7669159544159544, + "grad_norm": 0.4516580402851105, + "learning_rate": 0.00018252951663897432, + "loss": 1.0749, + "step": 4307 + }, + { + "epoch": 0.7670940170940171, + "grad_norm": 0.5565171837806702, + "learning_rate": 0.0001825216114356811, + "loss": 1.1617, + "step": 4308 + }, + { + "epoch": 0.7672720797720798, + "grad_norm": 0.5212662220001221, + "learning_rate": 0.00018251370461554834, + "loss": 1.1108, + "step": 4309 + }, + { + "epoch": 0.7674501424501424, + "grad_norm": 0.49061715602874756, + "learning_rate": 0.00018250579617873095, + "loss": 1.0881, + "step": 4310 + }, + { + "epoch": 0.7676282051282052, + "grad_norm": 0.5535751581192017, + "learning_rate": 0.00018249788612538387, + "loss": 0.9341, + "step": 4311 + }, + { + "epoch": 0.7678062678062678, + "grad_norm": 0.5425209403038025, + "learning_rate": 0.00018248997445566208, + "loss": 1.1858, + "step": 4312 + }, + { + "epoch": 0.7679843304843305, + "grad_norm": 0.6224395036697388, + "learning_rate": 0.0001824820611697206, + "loss": 1.0836, + "step": 4313 + }, + { + "epoch": 0.7681623931623932, + "grad_norm": 0.4895690977573395, + "learning_rate": 0.00018247414626771445, + "loss": 0.8598, + "step": 4314 + }, + { + "epoch": 0.7683404558404558, + "grad_norm": 0.5279615521430969, + "learning_rate": 0.00018246622974979877, + "loss": 1.1742, + "step": 4315 + }, + { + "epoch": 0.7685185185185185, + "grad_norm": 0.45300471782684326, + "learning_rate": 0.0001824583116161286, + "loss": 0.8872, + "step": 4316 + }, + { + "epoch": 0.7686965811965812, + "grad_norm": 0.6499692797660828, + "learning_rate": 0.00018245039186685916, + "loss": 1.2495, + "step": 4317 + }, + { + "epoch": 0.7688746438746439, + "grad_norm": 0.48151278495788574, + "learning_rate": 0.00018244247050214552, + "loss": 1.2382, + "step": 4318 + }, + { + "epoch": 0.7690527065527065, + "grad_norm": 0.6597028374671936, + "learning_rate": 0.0001824345475221429, + "loss": 1.3453, + "step": 4319 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.4536992609500885, + "learning_rate": 0.0001824266229270066, + "loss": 1.1141, + "step": 4320 + }, + { + "epoch": 0.7694088319088319, + "grad_norm": 0.5489405393600464, + "learning_rate": 0.00018241869671689184, + "loss": 1.0333, + "step": 4321 + }, + { + "epoch": 0.7695868945868946, + "grad_norm": 0.5741586089134216, + "learning_rate": 0.00018241076889195394, + "loss": 0.9939, + "step": 4322 + }, + { + "epoch": 0.7697649572649573, + "grad_norm": 0.47170960903167725, + "learning_rate": 0.00018240283945234823, + "loss": 0.9878, + "step": 4323 + }, + { + "epoch": 0.76994301994302, + "grad_norm": 0.4729093313217163, + "learning_rate": 0.00018239490839823004, + "loss": 1.0087, + "step": 4324 + }, + { + "epoch": 0.7701210826210826, + "grad_norm": 0.49869823455810547, + "learning_rate": 0.0001823869757297548, + "loss": 1.169, + "step": 4325 + }, + { + "epoch": 0.7702991452991453, + "grad_norm": 0.5118468403816223, + "learning_rate": 0.0001823790414470779, + "loss": 1.1092, + "step": 4326 + }, + { + "epoch": 0.770477207977208, + "grad_norm": 0.5076048970222473, + "learning_rate": 0.0001823711055503548, + "loss": 1.1028, + "step": 4327 + }, + { + "epoch": 0.7706552706552706, + "grad_norm": 0.5661569237709045, + "learning_rate": 0.00018236316803974098, + "loss": 1.1114, + "step": 4328 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 0.5542354583740234, + "learning_rate": 0.000182355228915392, + "loss": 1.0931, + "step": 4329 + }, + { + "epoch": 0.771011396011396, + "grad_norm": 0.5476680994033813, + "learning_rate": 0.0001823472881774634, + "loss": 1.036, + "step": 4330 + }, + { + "epoch": 0.7711894586894587, + "grad_norm": 0.5449798703193665, + "learning_rate": 0.00018233934582611073, + "loss": 1.0682, + "step": 4331 + }, + { + "epoch": 0.7713675213675214, + "grad_norm": 0.61089026927948, + "learning_rate": 0.00018233140186148963, + "loss": 1.0748, + "step": 4332 + }, + { + "epoch": 0.771545584045584, + "grad_norm": 0.5015206336975098, + "learning_rate": 0.00018232345628375576, + "loss": 1.2032, + "step": 4333 + }, + { + "epoch": 0.7717236467236467, + "grad_norm": 0.579289972782135, + "learning_rate": 0.00018231550909306475, + "loss": 1.0764, + "step": 4334 + }, + { + "epoch": 0.7719017094017094, + "grad_norm": 0.5889299511909485, + "learning_rate": 0.00018230756028957235, + "loss": 1.1768, + "step": 4335 + }, + { + "epoch": 0.7720797720797721, + "grad_norm": 0.5328249335289001, + "learning_rate": 0.00018229960987343428, + "loss": 1.0055, + "step": 4336 + }, + { + "epoch": 0.7722578347578347, + "grad_norm": 0.5766382217407227, + "learning_rate": 0.0001822916578448063, + "loss": 0.9923, + "step": 4337 + }, + { + "epoch": 0.7724358974358975, + "grad_norm": 0.6448187828063965, + "learning_rate": 0.00018228370420384423, + "loss": 1.1135, + "step": 4338 + }, + { + "epoch": 0.7726139601139601, + "grad_norm": 0.5505210757255554, + "learning_rate": 0.00018227574895070394, + "loss": 1.2048, + "step": 4339 + }, + { + "epoch": 0.7727920227920227, + "grad_norm": 0.6278925538063049, + "learning_rate": 0.00018226779208554126, + "loss": 1.1045, + "step": 4340 + }, + { + "epoch": 0.7729700854700855, + "grad_norm": 0.5345009565353394, + "learning_rate": 0.00018225983360851207, + "loss": 1.0102, + "step": 4341 + }, + { + "epoch": 0.7731481481481481, + "grad_norm": 0.566633403301239, + "learning_rate": 0.00018225187351977233, + "loss": 1.0038, + "step": 4342 + }, + { + "epoch": 0.7733262108262108, + "grad_norm": 0.5066078901290894, + "learning_rate": 0.000182243911819478, + "loss": 1.0339, + "step": 4343 + }, + { + "epoch": 0.7735042735042735, + "grad_norm": 0.5614920258522034, + "learning_rate": 0.00018223594850778503, + "loss": 1.1021, + "step": 4344 + }, + { + "epoch": 0.7736823361823362, + "grad_norm": 0.7747337818145752, + "learning_rate": 0.0001822279835848495, + "loss": 1.1129, + "step": 4345 + }, + { + "epoch": 0.7738603988603988, + "grad_norm": 0.7066529989242554, + "learning_rate": 0.00018222001705082744, + "loss": 1.3234, + "step": 4346 + }, + { + "epoch": 0.7740384615384616, + "grad_norm": 0.6340884566307068, + "learning_rate": 0.00018221204890587497, + "loss": 1.0726, + "step": 4347 + }, + { + "epoch": 0.7742165242165242, + "grad_norm": 0.5401145815849304, + "learning_rate": 0.00018220407915014818, + "loss": 0.9904, + "step": 4348 + }, + { + "epoch": 0.7743945868945868, + "grad_norm": 0.5069159269332886, + "learning_rate": 0.00018219610778380315, + "loss": 1.0654, + "step": 4349 + }, + { + "epoch": 0.7745726495726496, + "grad_norm": 0.5422839522361755, + "learning_rate": 0.00018218813480699623, + "loss": 1.1741, + "step": 4350 + }, + { + "epoch": 0.7747507122507122, + "grad_norm": 0.5550300478935242, + "learning_rate": 0.0001821801602198835, + "loss": 1.0033, + "step": 4351 + }, + { + "epoch": 0.7749287749287749, + "grad_norm": 0.5987736582756042, + "learning_rate": 0.00018217218402262123, + "loss": 0.935, + "step": 4352 + }, + { + "epoch": 0.7751068376068376, + "grad_norm": 0.6137008666992188, + "learning_rate": 0.00018216420621536573, + "loss": 1.17, + "step": 4353 + }, + { + "epoch": 0.7752849002849003, + "grad_norm": 0.47124359011650085, + "learning_rate": 0.0001821562267982733, + "loss": 0.8316, + "step": 4354 + }, + { + "epoch": 0.7754629629629629, + "grad_norm": 0.5057868361473083, + "learning_rate": 0.00018214824577150024, + "loss": 1.0246, + "step": 4355 + }, + { + "epoch": 0.7756410256410257, + "grad_norm": 0.604055643081665, + "learning_rate": 0.00018214026313520299, + "loss": 1.1272, + "step": 4356 + }, + { + "epoch": 0.7758190883190883, + "grad_norm": 0.6690384149551392, + "learning_rate": 0.0001821322788895379, + "loss": 1.0464, + "step": 4357 + }, + { + "epoch": 0.7759971509971509, + "grad_norm": 0.5458958745002747, + "learning_rate": 0.0001821242930346614, + "loss": 1.1712, + "step": 4358 + }, + { + "epoch": 0.7761752136752137, + "grad_norm": 0.6448663473129272, + "learning_rate": 0.00018211630557073, + "loss": 1.1125, + "step": 4359 + }, + { + "epoch": 0.7763532763532763, + "grad_norm": 0.49889448285102844, + "learning_rate": 0.00018210831649790018, + "loss": 1.097, + "step": 4360 + }, + { + "epoch": 0.7765313390313391, + "grad_norm": 0.5118046998977661, + "learning_rate": 0.00018210032581632843, + "loss": 1.009, + "step": 4361 + }, + { + "epoch": 0.7767094017094017, + "grad_norm": 0.5450068116188049, + "learning_rate": 0.00018209233352617135, + "loss": 1.1138, + "step": 4362 + }, + { + "epoch": 0.7768874643874644, + "grad_norm": 0.6147481203079224, + "learning_rate": 0.00018208433962758558, + "loss": 1.212, + "step": 4363 + }, + { + "epoch": 0.7770655270655271, + "grad_norm": 0.554176926612854, + "learning_rate": 0.00018207634412072764, + "loss": 1.1271, + "step": 4364 + }, + { + "epoch": 0.7772435897435898, + "grad_norm": 0.5872851014137268, + "learning_rate": 0.00018206834700575426, + "loss": 1.2793, + "step": 4365 + }, + { + "epoch": 0.7774216524216524, + "grad_norm": 0.5135685205459595, + "learning_rate": 0.00018206034828282207, + "loss": 0.9642, + "step": 4366 + }, + { + "epoch": 0.7775997150997151, + "grad_norm": 0.5699490308761597, + "learning_rate": 0.00018205234795208786, + "loss": 0.9086, + "step": 4367 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 0.5908057689666748, + "learning_rate": 0.00018204434601370832, + "loss": 1.1973, + "step": 4368 + }, + { + "epoch": 0.7779558404558404, + "grad_norm": 0.5777581334114075, + "learning_rate": 0.00018203634246784025, + "loss": 1.0447, + "step": 4369 + }, + { + "epoch": 0.7781339031339032, + "grad_norm": 0.4822927713394165, + "learning_rate": 0.00018202833731464048, + "loss": 0.814, + "step": 4370 + }, + { + "epoch": 0.7783119658119658, + "grad_norm": 0.5343610644340515, + "learning_rate": 0.0001820203305542658, + "loss": 1.2785, + "step": 4371 + }, + { + "epoch": 0.7784900284900285, + "grad_norm": 0.5462222695350647, + "learning_rate": 0.00018201232218687316, + "loss": 1.1785, + "step": 4372 + }, + { + "epoch": 0.7786680911680912, + "grad_norm": 0.5177609324455261, + "learning_rate": 0.00018200431221261943, + "loss": 1.111, + "step": 4373 + }, + { + "epoch": 0.7788461538461539, + "grad_norm": 0.5324625968933105, + "learning_rate": 0.00018199630063166157, + "loss": 1.0738, + "step": 4374 + }, + { + "epoch": 0.7790242165242165, + "grad_norm": 0.6392876505851746, + "learning_rate": 0.0001819882874441565, + "loss": 1.1758, + "step": 4375 + }, + { + "epoch": 0.7792022792022792, + "grad_norm": 0.49964696168899536, + "learning_rate": 0.00018198027265026127, + "loss": 1.0556, + "step": 4376 + }, + { + "epoch": 0.7793803418803419, + "grad_norm": 0.6090660691261292, + "learning_rate": 0.00018197225625013287, + "loss": 1.0102, + "step": 4377 + }, + { + "epoch": 0.7795584045584045, + "grad_norm": 0.5242345929145813, + "learning_rate": 0.00018196423824392842, + "loss": 0.8335, + "step": 4378 + }, + { + "epoch": 0.7797364672364673, + "grad_norm": 0.5265036225318909, + "learning_rate": 0.00018195621863180498, + "loss": 1.0781, + "step": 4379 + }, + { + "epoch": 0.7799145299145299, + "grad_norm": 0.5115378499031067, + "learning_rate": 0.0001819481974139197, + "loss": 1.1658, + "step": 4380 + }, + { + "epoch": 0.7800925925925926, + "grad_norm": 0.6489549875259399, + "learning_rate": 0.00018194017459042972, + "loss": 1.0572, + "step": 4381 + }, + { + "epoch": 0.7802706552706553, + "grad_norm": 0.5800202488899231, + "learning_rate": 0.0001819321501614922, + "loss": 0.9593, + "step": 4382 + }, + { + "epoch": 0.780448717948718, + "grad_norm": 0.5608528256416321, + "learning_rate": 0.00018192412412726443, + "loss": 1.0324, + "step": 4383 + }, + { + "epoch": 0.7806267806267806, + "grad_norm": 0.5596401691436768, + "learning_rate": 0.00018191609648790362, + "loss": 1.071, + "step": 4384 + }, + { + "epoch": 0.7808048433048433, + "grad_norm": 0.5712903141975403, + "learning_rate": 0.00018190806724356707, + "loss": 0.9011, + "step": 4385 + }, + { + "epoch": 0.780982905982906, + "grad_norm": 0.5079438090324402, + "learning_rate": 0.0001819000363944121, + "loss": 1.1194, + "step": 4386 + }, + { + "epoch": 0.7811609686609686, + "grad_norm": 0.5785079598426819, + "learning_rate": 0.00018189200394059602, + "loss": 1.1703, + "step": 4387 + }, + { + "epoch": 0.7813390313390314, + "grad_norm": 0.6901816129684448, + "learning_rate": 0.00018188396988227625, + "loss": 1.6689, + "step": 4388 + }, + { + "epoch": 0.781517094017094, + "grad_norm": 0.48107922077178955, + "learning_rate": 0.00018187593421961022, + "loss": 1.0116, + "step": 4389 + }, + { + "epoch": 0.7816951566951567, + "grad_norm": 0.5843084454536438, + "learning_rate": 0.0001818678969527553, + "loss": 1.1172, + "step": 4390 + }, + { + "epoch": 0.7818732193732194, + "grad_norm": 0.479034423828125, + "learning_rate": 0.00018185985808186902, + "loss": 0.811, + "step": 4391 + }, + { + "epoch": 0.782051282051282, + "grad_norm": 0.5864158272743225, + "learning_rate": 0.00018185181760710888, + "loss": 0.9522, + "step": 4392 + }, + { + "epoch": 0.7822293447293447, + "grad_norm": 0.4824625551700592, + "learning_rate": 0.00018184377552863242, + "loss": 0.9039, + "step": 4393 + }, + { + "epoch": 0.7824074074074074, + "grad_norm": 0.580102801322937, + "learning_rate": 0.00018183573184659717, + "loss": 1.2382, + "step": 4394 + }, + { + "epoch": 0.7825854700854701, + "grad_norm": 0.5300056338310242, + "learning_rate": 0.00018182768656116073, + "loss": 1.2268, + "step": 4395 + }, + { + "epoch": 0.7827635327635327, + "grad_norm": 0.5548123121261597, + "learning_rate": 0.00018181963967248078, + "loss": 1.0628, + "step": 4396 + }, + { + "epoch": 0.7829415954415955, + "grad_norm": 0.5485070943832397, + "learning_rate": 0.00018181159118071496, + "loss": 0.9628, + "step": 4397 + }, + { + "epoch": 0.7831196581196581, + "grad_norm": 0.47405415773391724, + "learning_rate": 0.00018180354108602095, + "loss": 1.1413, + "step": 4398 + }, + { + "epoch": 0.7832977207977208, + "grad_norm": 0.5545752644538879, + "learning_rate": 0.0001817954893885565, + "loss": 1.3807, + "step": 4399 + }, + { + "epoch": 0.7834757834757835, + "grad_norm": 0.5339497327804565, + "learning_rate": 0.00018178743608847933, + "loss": 0.9978, + "step": 4400 + }, + { + "epoch": 0.7836538461538461, + "grad_norm": 0.5006352663040161, + "learning_rate": 0.00018177938118594725, + "loss": 0.8873, + "step": 4401 + }, + { + "epoch": 0.7838319088319088, + "grad_norm": 0.4845179319381714, + "learning_rate": 0.00018177132468111812, + "loss": 0.8866, + "step": 4402 + }, + { + "epoch": 0.7840099715099715, + "grad_norm": 0.5240967869758606, + "learning_rate": 0.0001817632665741497, + "loss": 1.0347, + "step": 4403 + }, + { + "epoch": 0.7841880341880342, + "grad_norm": 0.5311884880065918, + "learning_rate": 0.00018175520686519993, + "loss": 1.2065, + "step": 4404 + }, + { + "epoch": 0.7843660968660968, + "grad_norm": 0.5562815070152283, + "learning_rate": 0.00018174714555442673, + "loss": 1.1272, + "step": 4405 + }, + { + "epoch": 0.7845441595441596, + "grad_norm": 0.5524366497993469, + "learning_rate": 0.00018173908264198802, + "loss": 1.2337, + "step": 4406 + }, + { + "epoch": 0.7847222222222222, + "grad_norm": 0.5612216591835022, + "learning_rate": 0.0001817310181280418, + "loss": 1.1809, + "step": 4407 + }, + { + "epoch": 0.7849002849002849, + "grad_norm": 0.5315343737602234, + "learning_rate": 0.000181722952012746, + "loss": 1.0491, + "step": 4408 + }, + { + "epoch": 0.7850783475783476, + "grad_norm": 0.5233435034751892, + "learning_rate": 0.00018171488429625878, + "loss": 1.0457, + "step": 4409 + }, + { + "epoch": 0.7852564102564102, + "grad_norm": 0.7809093594551086, + "learning_rate": 0.00018170681497873813, + "loss": 1.1578, + "step": 4410 + }, + { + "epoch": 0.7854344729344729, + "grad_norm": 0.49659839272499084, + "learning_rate": 0.00018169874406034217, + "loss": 1.0815, + "step": 4411 + }, + { + "epoch": 0.7856125356125356, + "grad_norm": 0.5020765066146851, + "learning_rate": 0.00018169067154122904, + "loss": 1.1985, + "step": 4412 + }, + { + "epoch": 0.7857905982905983, + "grad_norm": 0.6408432126045227, + "learning_rate": 0.0001816825974215569, + "loss": 1.2272, + "step": 4413 + }, + { + "epoch": 0.7859686609686609, + "grad_norm": 0.5062605142593384, + "learning_rate": 0.00018167452170148396, + "loss": 0.9663, + "step": 4414 + }, + { + "epoch": 0.7861467236467237, + "grad_norm": 0.5100119113922119, + "learning_rate": 0.0001816664443811684, + "loss": 1.0256, + "step": 4415 + }, + { + "epoch": 0.7863247863247863, + "grad_norm": 0.5277643799781799, + "learning_rate": 0.00018165836546076854, + "loss": 1.2885, + "step": 4416 + }, + { + "epoch": 0.7865028490028491, + "grad_norm": 0.5568150281906128, + "learning_rate": 0.0001816502849404426, + "loss": 1.2673, + "step": 4417 + }, + { + "epoch": 0.7866809116809117, + "grad_norm": 0.5061392188072205, + "learning_rate": 0.00018164220282034896, + "loss": 1.072, + "step": 4418 + }, + { + "epoch": 0.7868589743589743, + "grad_norm": 0.5383077263832092, + "learning_rate": 0.00018163411910064597, + "loss": 1.0621, + "step": 4419 + }, + { + "epoch": 0.7870370370370371, + "grad_norm": 0.5167948007583618, + "learning_rate": 0.00018162603378149198, + "loss": 1.099, + "step": 4420 + }, + { + "epoch": 0.7872150997150997, + "grad_norm": 0.5084534287452698, + "learning_rate": 0.0001816179468630454, + "loss": 1.3984, + "step": 4421 + }, + { + "epoch": 0.7873931623931624, + "grad_norm": 0.608762264251709, + "learning_rate": 0.00018160985834546475, + "loss": 1.3553, + "step": 4422 + }, + { + "epoch": 0.7875712250712251, + "grad_norm": 0.4900866746902466, + "learning_rate": 0.00018160176822890842, + "loss": 1.0009, + "step": 4423 + }, + { + "epoch": 0.7877492877492878, + "grad_norm": 0.5928917527198792, + "learning_rate": 0.00018159367651353496, + "loss": 1.0523, + "step": 4424 + }, + { + "epoch": 0.7879273504273504, + "grad_norm": 0.624422013759613, + "learning_rate": 0.0001815855831995029, + "loss": 1.0519, + "step": 4425 + }, + { + "epoch": 0.7881054131054132, + "grad_norm": 0.5140150785446167, + "learning_rate": 0.00018157748828697082, + "loss": 1.048, + "step": 4426 + }, + { + "epoch": 0.7882834757834758, + "grad_norm": 0.47006943821907043, + "learning_rate": 0.00018156939177609732, + "loss": 1.0067, + "step": 4427 + }, + { + "epoch": 0.7884615384615384, + "grad_norm": 0.5178864002227783, + "learning_rate": 0.00018156129366704105, + "loss": 1.0583, + "step": 4428 + }, + { + "epoch": 0.7886396011396012, + "grad_norm": 0.5279985666275024, + "learning_rate": 0.00018155319395996066, + "loss": 1.3023, + "step": 4429 + }, + { + "epoch": 0.7888176638176638, + "grad_norm": 0.5238787531852722, + "learning_rate": 0.00018154509265501482, + "loss": 1.0851, + "step": 4430 + }, + { + "epoch": 0.7889957264957265, + "grad_norm": 0.5914917588233948, + "learning_rate": 0.00018153698975236228, + "loss": 0.9291, + "step": 4431 + }, + { + "epoch": 0.7891737891737892, + "grad_norm": 0.5046082735061646, + "learning_rate": 0.00018152888525216183, + "loss": 0.9951, + "step": 4432 + }, + { + "epoch": 0.7893518518518519, + "grad_norm": 0.5042256116867065, + "learning_rate": 0.00018152077915457225, + "loss": 1.0243, + "step": 4433 + }, + { + "epoch": 0.7895299145299145, + "grad_norm": 0.5950339436531067, + "learning_rate": 0.0001815126714597523, + "loss": 0.9803, + "step": 4434 + }, + { + "epoch": 0.7897079772079773, + "grad_norm": 0.5163764953613281, + "learning_rate": 0.0001815045621678609, + "loss": 1.0353, + "step": 4435 + }, + { + "epoch": 0.7898860398860399, + "grad_norm": 0.5166211128234863, + "learning_rate": 0.00018149645127905691, + "loss": 0.9649, + "step": 4436 + }, + { + "epoch": 0.7900641025641025, + "grad_norm": 0.5239769220352173, + "learning_rate": 0.00018148833879349927, + "loss": 0.9747, + "step": 4437 + }, + { + "epoch": 0.7902421652421653, + "grad_norm": 0.5803237557411194, + "learning_rate": 0.00018148022471134692, + "loss": 1.315, + "step": 4438 + }, + { + "epoch": 0.7904202279202279, + "grad_norm": 0.5141370296478271, + "learning_rate": 0.00018147210903275877, + "loss": 1.0547, + "step": 4439 + }, + { + "epoch": 0.7905982905982906, + "grad_norm": 0.545788586139679, + "learning_rate": 0.00018146399175789394, + "loss": 1.0797, + "step": 4440 + }, + { + "epoch": 0.7907763532763533, + "grad_norm": 0.5273314714431763, + "learning_rate": 0.0001814558728869114, + "loss": 0.7928, + "step": 4441 + }, + { + "epoch": 0.790954415954416, + "grad_norm": 0.4614652693271637, + "learning_rate": 0.00018144775241997024, + "loss": 0.8826, + "step": 4442 + }, + { + "epoch": 0.7911324786324786, + "grad_norm": 0.6203590631484985, + "learning_rate": 0.00018143963035722958, + "loss": 1.2891, + "step": 4443 + }, + { + "epoch": 0.7913105413105413, + "grad_norm": 0.4870408773422241, + "learning_rate": 0.0001814315066988485, + "loss": 1.0717, + "step": 4444 + }, + { + "epoch": 0.791488603988604, + "grad_norm": 0.6468982696533203, + "learning_rate": 0.00018142338144498625, + "loss": 1.3398, + "step": 4445 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 0.4727918207645416, + "learning_rate": 0.00018141525459580197, + "loss": 1.0195, + "step": 4446 + }, + { + "epoch": 0.7918447293447294, + "grad_norm": 0.5080479979515076, + "learning_rate": 0.0001814071261514549, + "loss": 1.0163, + "step": 4447 + }, + { + "epoch": 0.792022792022792, + "grad_norm": 0.5380908250808716, + "learning_rate": 0.0001813989961121043, + "loss": 1.1673, + "step": 4448 + }, + { + "epoch": 0.7922008547008547, + "grad_norm": 0.5020384192466736, + "learning_rate": 0.00018139086447790945, + "loss": 0.8591, + "step": 4449 + }, + { + "epoch": 0.7923789173789174, + "grad_norm": 0.5279949903488159, + "learning_rate": 0.0001813827312490297, + "loss": 1.1221, + "step": 4450 + }, + { + "epoch": 0.79255698005698, + "grad_norm": 0.6739233732223511, + "learning_rate": 0.00018137459642562437, + "loss": 1.2704, + "step": 4451 + }, + { + "epoch": 0.7927350427350427, + "grad_norm": 0.5112259984016418, + "learning_rate": 0.00018136646000785288, + "loss": 1.1161, + "step": 4452 + }, + { + "epoch": 0.7929131054131054, + "grad_norm": 0.5244031548500061, + "learning_rate": 0.00018135832199587463, + "loss": 0.7866, + "step": 4453 + }, + { + "epoch": 0.7930911680911681, + "grad_norm": 0.5803347229957581, + "learning_rate": 0.0001813501823898491, + "loss": 0.994, + "step": 4454 + }, + { + "epoch": 0.7932692307692307, + "grad_norm": 0.6191152930259705, + "learning_rate": 0.00018134204118993568, + "loss": 1.0725, + "step": 4455 + }, + { + "epoch": 0.7934472934472935, + "grad_norm": 0.549735963344574, + "learning_rate": 0.00018133389839629396, + "loss": 0.9915, + "step": 4456 + }, + { + "epoch": 0.7936253561253561, + "grad_norm": 0.4940381646156311, + "learning_rate": 0.00018132575400908347, + "loss": 1.1815, + "step": 4457 + }, + { + "epoch": 0.7938034188034188, + "grad_norm": 0.5009099245071411, + "learning_rate": 0.00018131760802846377, + "loss": 1.0833, + "step": 4458 + }, + { + "epoch": 0.7939814814814815, + "grad_norm": 0.595853865146637, + "learning_rate": 0.00018130946045459445, + "loss": 1.2774, + "step": 4459 + }, + { + "epoch": 0.7941595441595442, + "grad_norm": 0.534794807434082, + "learning_rate": 0.00018130131128763513, + "loss": 1.0891, + "step": 4460 + }, + { + "epoch": 0.7943376068376068, + "grad_norm": 0.5828582048416138, + "learning_rate": 0.00018129316052774557, + "loss": 1.0786, + "step": 4461 + }, + { + "epoch": 0.7945156695156695, + "grad_norm": 0.4750654697418213, + "learning_rate": 0.00018128500817508533, + "loss": 1.0818, + "step": 4462 + }, + { + "epoch": 0.7946937321937322, + "grad_norm": 0.5626576542854309, + "learning_rate": 0.00018127685422981426, + "loss": 1.0807, + "step": 4463 + }, + { + "epoch": 0.7948717948717948, + "grad_norm": 0.6434760093688965, + "learning_rate": 0.00018126869869209203, + "loss": 1.0908, + "step": 4464 + }, + { + "epoch": 0.7950498575498576, + "grad_norm": 0.5577414631843567, + "learning_rate": 0.00018126054156207853, + "loss": 1.0281, + "step": 4465 + }, + { + "epoch": 0.7952279202279202, + "grad_norm": 0.5001249313354492, + "learning_rate": 0.00018125238283993347, + "loss": 0.9083, + "step": 4466 + }, + { + "epoch": 0.7954059829059829, + "grad_norm": 0.5298314690589905, + "learning_rate": 0.00018124422252581676, + "loss": 0.971, + "step": 4467 + }, + { + "epoch": 0.7955840455840456, + "grad_norm": 0.4872737228870392, + "learning_rate": 0.00018123606061988832, + "loss": 1.0515, + "step": 4468 + }, + { + "epoch": 0.7957621082621082, + "grad_norm": 0.5895398259162903, + "learning_rate": 0.00018122789712230798, + "loss": 1.0771, + "step": 4469 + }, + { + "epoch": 0.7959401709401709, + "grad_norm": 0.5212514996528625, + "learning_rate": 0.00018121973203323577, + "loss": 1.0365, + "step": 4470 + }, + { + "epoch": 0.7961182336182336, + "grad_norm": 0.4679451584815979, + "learning_rate": 0.0001812115653528316, + "loss": 0.9445, + "step": 4471 + }, + { + "epoch": 0.7962962962962963, + "grad_norm": 0.5852653980255127, + "learning_rate": 0.00018120339708125552, + "loss": 1.1781, + "step": 4472 + }, + { + "epoch": 0.7964743589743589, + "grad_norm": 0.6081342697143555, + "learning_rate": 0.00018119522721866756, + "loss": 1.3881, + "step": 4473 + }, + { + "epoch": 0.7966524216524217, + "grad_norm": 0.5254155993461609, + "learning_rate": 0.00018118705576522777, + "loss": 1.2198, + "step": 4474 + }, + { + "epoch": 0.7968304843304843, + "grad_norm": 0.5959419012069702, + "learning_rate": 0.00018117888272109632, + "loss": 1.0922, + "step": 4475 + }, + { + "epoch": 0.7970085470085471, + "grad_norm": 0.6243147253990173, + "learning_rate": 0.0001811707080864333, + "loss": 1.1782, + "step": 4476 + }, + { + "epoch": 0.7971866096866097, + "grad_norm": 0.5336906909942627, + "learning_rate": 0.0001811625318613988, + "loss": 1.167, + "step": 4477 + }, + { + "epoch": 0.7973646723646723, + "grad_norm": 0.5287907719612122, + "learning_rate": 0.00018115435404615315, + "loss": 0.9923, + "step": 4478 + }, + { + "epoch": 0.7975427350427351, + "grad_norm": 0.48941442370414734, + "learning_rate": 0.0001811461746408565, + "loss": 0.863, + "step": 4479 + }, + { + "epoch": 0.7977207977207977, + "grad_norm": 0.48465651273727417, + "learning_rate": 0.0001811379936456691, + "loss": 1.147, + "step": 4480 + }, + { + "epoch": 0.7978988603988604, + "grad_norm": 0.5676067471504211, + "learning_rate": 0.0001811298110607513, + "loss": 1.3121, + "step": 4481 + }, + { + "epoch": 0.7980769230769231, + "grad_norm": 0.4894018769264221, + "learning_rate": 0.00018112162688626337, + "loss": 1.1831, + "step": 4482 + }, + { + "epoch": 0.7982549857549858, + "grad_norm": 0.5626382827758789, + "learning_rate": 0.0001811134411223657, + "loss": 1.1977, + "step": 4483 + }, + { + "epoch": 0.7984330484330484, + "grad_norm": 0.564119815826416, + "learning_rate": 0.00018110525376921862, + "loss": 1.2686, + "step": 4484 + }, + { + "epoch": 0.7986111111111112, + "grad_norm": 0.6385740041732788, + "learning_rate": 0.00018109706482698256, + "loss": 1.2418, + "step": 4485 + }, + { + "epoch": 0.7987891737891738, + "grad_norm": 0.5550164580345154, + "learning_rate": 0.00018108887429581802, + "loss": 1.081, + "step": 4486 + }, + { + "epoch": 0.7989672364672364, + "grad_norm": 0.5583973526954651, + "learning_rate": 0.00018108068217588544, + "loss": 1.1757, + "step": 4487 + }, + { + "epoch": 0.7991452991452992, + "grad_norm": 0.5533342957496643, + "learning_rate": 0.00018107248846734527, + "loss": 1.1947, + "step": 4488 + }, + { + "epoch": 0.7993233618233618, + "grad_norm": 0.5291479229927063, + "learning_rate": 0.00018106429317035815, + "loss": 1.2769, + "step": 4489 + }, + { + "epoch": 0.7995014245014245, + "grad_norm": 0.4680160582065582, + "learning_rate": 0.00018105609628508458, + "loss": 0.7059, + "step": 4490 + }, + { + "epoch": 0.7996794871794872, + "grad_norm": 0.5364881157875061, + "learning_rate": 0.00018104789781168517, + "loss": 1.0566, + "step": 4491 + }, + { + "epoch": 0.7998575498575499, + "grad_norm": 0.5917307734489441, + "learning_rate": 0.0001810396977503206, + "loss": 1.2263, + "step": 4492 + }, + { + "epoch": 0.8000356125356125, + "grad_norm": 0.6013199090957642, + "learning_rate": 0.0001810314961011515, + "loss": 1.2053, + "step": 4493 + }, + { + "epoch": 0.8002136752136753, + "grad_norm": 0.6005663275718689, + "learning_rate": 0.0001810232928643385, + "loss": 1.2241, + "step": 4494 + }, + { + "epoch": 0.8003917378917379, + "grad_norm": 0.49207603931427, + "learning_rate": 0.00018101508804004246, + "loss": 1.0661, + "step": 4495 + }, + { + "epoch": 0.8005698005698005, + "grad_norm": 0.4834063947200775, + "learning_rate": 0.00018100688162842401, + "loss": 1.1745, + "step": 4496 + }, + { + "epoch": 0.8007478632478633, + "grad_norm": 0.5347156524658203, + "learning_rate": 0.000180998673629644, + "loss": 1.0679, + "step": 4497 + }, + { + "epoch": 0.8009259259259259, + "grad_norm": 0.5815600156784058, + "learning_rate": 0.00018099046404386327, + "loss": 1.2652, + "step": 4498 + }, + { + "epoch": 0.8011039886039886, + "grad_norm": 0.5291135311126709, + "learning_rate": 0.00018098225287124263, + "loss": 1.2072, + "step": 4499 + }, + { + "epoch": 0.8012820512820513, + "grad_norm": 0.5779497027397156, + "learning_rate": 0.000180974040111943, + "loss": 1.3277, + "step": 4500 + }, + { + "epoch": 0.801460113960114, + "grad_norm": 0.44566696882247925, + "learning_rate": 0.0001809658257661252, + "loss": 0.7702, + "step": 4501 + }, + { + "epoch": 0.8016381766381766, + "grad_norm": 0.5407577753067017, + "learning_rate": 0.00018095760983395027, + "loss": 1.2894, + "step": 4502 + }, + { + "epoch": 0.8018162393162394, + "grad_norm": 0.4771903455257416, + "learning_rate": 0.00018094939231557916, + "loss": 1.045, + "step": 4503 + }, + { + "epoch": 0.801994301994302, + "grad_norm": 0.5970945358276367, + "learning_rate": 0.00018094117321117286, + "loss": 1.2059, + "step": 4504 + }, + { + "epoch": 0.8021723646723646, + "grad_norm": 0.4959338903427124, + "learning_rate": 0.0001809329525208924, + "loss": 1.155, + "step": 4505 + }, + { + "epoch": 0.8023504273504274, + "grad_norm": 0.5142548084259033, + "learning_rate": 0.00018092473024489887, + "loss": 0.9413, + "step": 4506 + }, + { + "epoch": 0.80252849002849, + "grad_norm": 0.5336433053016663, + "learning_rate": 0.00018091650638335334, + "loss": 1.0699, + "step": 4507 + }, + { + "epoch": 0.8027065527065527, + "grad_norm": 0.47770628333091736, + "learning_rate": 0.00018090828093641698, + "loss": 1.1515, + "step": 4508 + }, + { + "epoch": 0.8028846153846154, + "grad_norm": 0.5443438291549683, + "learning_rate": 0.00018090005390425091, + "loss": 1.189, + "step": 4509 + }, + { + "epoch": 0.8030626780626781, + "grad_norm": 0.523179829120636, + "learning_rate": 0.00018089182528701632, + "loss": 1.1272, + "step": 4510 + }, + { + "epoch": 0.8032407407407407, + "grad_norm": 0.49628451466560364, + "learning_rate": 0.00018088359508487448, + "loss": 0.9754, + "step": 4511 + }, + { + "epoch": 0.8034188034188035, + "grad_norm": 0.5933086276054382, + "learning_rate": 0.00018087536329798663, + "loss": 1.2111, + "step": 4512 + }, + { + "epoch": 0.8035968660968661, + "grad_norm": 0.4565310776233673, + "learning_rate": 0.00018086712992651402, + "loss": 0.7729, + "step": 4513 + }, + { + "epoch": 0.8037749287749287, + "grad_norm": 0.5013461112976074, + "learning_rate": 0.00018085889497061798, + "loss": 1.2178, + "step": 4514 + }, + { + "epoch": 0.8039529914529915, + "grad_norm": 0.5170024633407593, + "learning_rate": 0.00018085065843045987, + "loss": 0.9181, + "step": 4515 + }, + { + "epoch": 0.8041310541310541, + "grad_norm": 0.583363950252533, + "learning_rate": 0.00018084242030620104, + "loss": 1.1542, + "step": 4516 + }, + { + "epoch": 0.8043091168091168, + "grad_norm": 0.46835777163505554, + "learning_rate": 0.00018083418059800297, + "loss": 0.8954, + "step": 4517 + }, + { + "epoch": 0.8044871794871795, + "grad_norm": 0.5145657062530518, + "learning_rate": 0.000180825939306027, + "loss": 1.0417, + "step": 4518 + }, + { + "epoch": 0.8046652421652422, + "grad_norm": 0.47216105461120605, + "learning_rate": 0.00018081769643043467, + "loss": 0.9516, + "step": 4519 + }, + { + "epoch": 0.8048433048433048, + "grad_norm": 0.5059915781021118, + "learning_rate": 0.0001808094519713875, + "loss": 1.1643, + "step": 4520 + }, + { + "epoch": 0.8050213675213675, + "grad_norm": 0.5406439900398254, + "learning_rate": 0.00018080120592904692, + "loss": 1.2038, + "step": 4521 + }, + { + "epoch": 0.8051994301994302, + "grad_norm": 0.6123420000076294, + "learning_rate": 0.0001807929583035746, + "loss": 1.4004, + "step": 4522 + }, + { + "epoch": 0.8053774928774928, + "grad_norm": 0.49699845910072327, + "learning_rate": 0.00018078470909513208, + "loss": 1.0347, + "step": 4523 + }, + { + "epoch": 0.8055555555555556, + "grad_norm": 0.5369421243667603, + "learning_rate": 0.000180776458303881, + "loss": 1.0418, + "step": 4524 + }, + { + "epoch": 0.8057336182336182, + "grad_norm": 0.5407396554946899, + "learning_rate": 0.00018076820592998301, + "loss": 0.9546, + "step": 4525 + }, + { + "epoch": 0.8059116809116809, + "grad_norm": 0.5749752521514893, + "learning_rate": 0.00018075995197359984, + "loss": 1.1438, + "step": 4526 + }, + { + "epoch": 0.8060897435897436, + "grad_norm": 0.5523102283477783, + "learning_rate": 0.00018075169643489317, + "loss": 1.1312, + "step": 4527 + }, + { + "epoch": 0.8062678062678063, + "grad_norm": 0.5767508149147034, + "learning_rate": 0.00018074343931402472, + "loss": 1.1951, + "step": 4528 + }, + { + "epoch": 0.8064458689458689, + "grad_norm": 0.5262924432754517, + "learning_rate": 0.00018073518061115633, + "loss": 1.1985, + "step": 4529 + }, + { + "epoch": 0.8066239316239316, + "grad_norm": 0.4742378294467926, + "learning_rate": 0.0001807269203264498, + "loss": 1.0126, + "step": 4530 + }, + { + "epoch": 0.8068019943019943, + "grad_norm": 0.5190158486366272, + "learning_rate": 0.00018071865846006692, + "loss": 0.9985, + "step": 4531 + }, + { + "epoch": 0.8069800569800569, + "grad_norm": 0.5910618305206299, + "learning_rate": 0.00018071039501216964, + "loss": 1.2776, + "step": 4532 + }, + { + "epoch": 0.8071581196581197, + "grad_norm": 0.5363098382949829, + "learning_rate": 0.00018070212998291983, + "loss": 1.3346, + "step": 4533 + }, + { + "epoch": 0.8073361823361823, + "grad_norm": 0.47711408138275146, + "learning_rate": 0.0001806938633724794, + "loss": 1.04, + "step": 4534 + }, + { + "epoch": 0.8075142450142451, + "grad_norm": 0.5092964172363281, + "learning_rate": 0.0001806855951810104, + "loss": 1.1409, + "step": 4535 + }, + { + "epoch": 0.8076923076923077, + "grad_norm": 0.5828777551651001, + "learning_rate": 0.00018067732540867472, + "loss": 1.3048, + "step": 4536 + }, + { + "epoch": 0.8078703703703703, + "grad_norm": 0.5779826045036316, + "learning_rate": 0.00018066905405563445, + "loss": 1.1599, + "step": 4537 + }, + { + "epoch": 0.8080484330484331, + "grad_norm": 0.49908435344696045, + "learning_rate": 0.00018066078112205167, + "loss": 1.1502, + "step": 4538 + }, + { + "epoch": 0.8082264957264957, + "grad_norm": 0.4772704839706421, + "learning_rate": 0.0001806525066080884, + "loss": 0.7925, + "step": 4539 + }, + { + "epoch": 0.8084045584045584, + "grad_norm": 0.4298383295536041, + "learning_rate": 0.00018064423051390683, + "loss": 0.7322, + "step": 4540 + }, + { + "epoch": 0.8085826210826211, + "grad_norm": 0.49349579215049744, + "learning_rate": 0.0001806359528396691, + "loss": 1.0021, + "step": 4541 + }, + { + "epoch": 0.8087606837606838, + "grad_norm": 0.4698609411716461, + "learning_rate": 0.00018062767358553735, + "loss": 0.9751, + "step": 4542 + }, + { + "epoch": 0.8089387464387464, + "grad_norm": 0.4949014186859131, + "learning_rate": 0.00018061939275167385, + "loss": 0.9553, + "step": 4543 + }, + { + "epoch": 0.8091168091168092, + "grad_norm": 0.5604463815689087, + "learning_rate": 0.0001806111103382408, + "loss": 0.9894, + "step": 4544 + }, + { + "epoch": 0.8092948717948718, + "grad_norm": 0.5761561989784241, + "learning_rate": 0.00018060282634540053, + "loss": 1.258, + "step": 4545 + }, + { + "epoch": 0.8094729344729344, + "grad_norm": 0.5239115357398987, + "learning_rate": 0.00018059454077331527, + "loss": 0.9189, + "step": 4546 + }, + { + "epoch": 0.8096509971509972, + "grad_norm": 0.47902220487594604, + "learning_rate": 0.00018058625362214742, + "loss": 1.0389, + "step": 4547 + }, + { + "epoch": 0.8098290598290598, + "grad_norm": 0.6274173259735107, + "learning_rate": 0.00018057796489205936, + "loss": 1.3368, + "step": 4548 + }, + { + "epoch": 0.8100071225071225, + "grad_norm": 0.5789401531219482, + "learning_rate": 0.00018056967458321345, + "loss": 1.1473, + "step": 4549 + }, + { + "epoch": 0.8101851851851852, + "grad_norm": 0.5850043296813965, + "learning_rate": 0.0001805613826957721, + "loss": 1.2224, + "step": 4550 + }, + { + "epoch": 0.8103632478632479, + "grad_norm": 0.6310738921165466, + "learning_rate": 0.00018055308922989788, + "loss": 1.0707, + "step": 4551 + }, + { + "epoch": 0.8105413105413105, + "grad_norm": 0.5198429822921753, + "learning_rate": 0.00018054479418575317, + "loss": 0.8984, + "step": 4552 + }, + { + "epoch": 0.8107193732193733, + "grad_norm": 0.5757743120193481, + "learning_rate": 0.00018053649756350054, + "loss": 1.2007, + "step": 4553 + }, + { + "epoch": 0.8108974358974359, + "grad_norm": 0.5109567642211914, + "learning_rate": 0.0001805281993633025, + "loss": 1.0696, + "step": 4554 + }, + { + "epoch": 0.8110754985754985, + "grad_norm": 0.5030225515365601, + "learning_rate": 0.00018051989958532173, + "loss": 0.9667, + "step": 4555 + }, + { + "epoch": 0.8112535612535613, + "grad_norm": 0.5291743874549866, + "learning_rate": 0.00018051159822972079, + "loss": 1.0219, + "step": 4556 + }, + { + "epoch": 0.8114316239316239, + "grad_norm": 0.5874896049499512, + "learning_rate": 0.00018050329529666233, + "loss": 0.8589, + "step": 4557 + }, + { + "epoch": 0.8116096866096866, + "grad_norm": 0.673284113407135, + "learning_rate": 0.000180494990786309, + "loss": 1.1902, + "step": 4558 + }, + { + "epoch": 0.8117877492877493, + "grad_norm": 0.4742524027824402, + "learning_rate": 0.00018048668469882354, + "loss": 1.0578, + "step": 4559 + }, + { + "epoch": 0.811965811965812, + "grad_norm": 0.5519167184829712, + "learning_rate": 0.0001804783770343687, + "loss": 1.083, + "step": 4560 + }, + { + "epoch": 0.8121438746438746, + "grad_norm": 0.5669941306114197, + "learning_rate": 0.00018047006779310727, + "loss": 1.0784, + "step": 4561 + }, + { + "epoch": 0.8123219373219374, + "grad_norm": 0.512759804725647, + "learning_rate": 0.000180461756975202, + "loss": 1.0361, + "step": 4562 + }, + { + "epoch": 0.8125, + "grad_norm": 0.5721749067306519, + "learning_rate": 0.00018045344458081575, + "loss": 1.0246, + "step": 4563 + }, + { + "epoch": 0.8126780626780626, + "grad_norm": 0.566430389881134, + "learning_rate": 0.00018044513061011137, + "loss": 1.1452, + "step": 4564 + }, + { + "epoch": 0.8128561253561254, + "grad_norm": 0.49391916394233704, + "learning_rate": 0.00018043681506325177, + "loss": 0.89, + "step": 4565 + }, + { + "epoch": 0.813034188034188, + "grad_norm": 0.5379437804222107, + "learning_rate": 0.00018042849794039988, + "loss": 1.1289, + "step": 4566 + }, + { + "epoch": 0.8132122507122507, + "grad_norm": 0.5667982697486877, + "learning_rate": 0.00018042017924171865, + "loss": 1.1596, + "step": 4567 + }, + { + "epoch": 0.8133903133903134, + "grad_norm": 0.6214209794998169, + "learning_rate": 0.00018041185896737109, + "loss": 1.0622, + "step": 4568 + }, + { + "epoch": 0.8135683760683761, + "grad_norm": 0.5442491173744202, + "learning_rate": 0.00018040353711752015, + "loss": 1.0536, + "step": 4569 + }, + { + "epoch": 0.8137464387464387, + "grad_norm": 0.5266172885894775, + "learning_rate": 0.00018039521369232894, + "loss": 1.0576, + "step": 4570 + }, + { + "epoch": 0.8139245014245015, + "grad_norm": 0.6057912111282349, + "learning_rate": 0.00018038688869196053, + "loss": 1.3067, + "step": 4571 + }, + { + "epoch": 0.8141025641025641, + "grad_norm": 0.489869087934494, + "learning_rate": 0.00018037856211657803, + "loss": 1.0279, + "step": 4572 + }, + { + "epoch": 0.8142806267806267, + "grad_norm": 0.5497978329658508, + "learning_rate": 0.00018037023396634457, + "loss": 1.1568, + "step": 4573 + }, + { + "epoch": 0.8144586894586895, + "grad_norm": 0.5243251919746399, + "learning_rate": 0.0001803619042414233, + "loss": 0.9767, + "step": 4574 + }, + { + "epoch": 0.8146367521367521, + "grad_norm": 0.503032922744751, + "learning_rate": 0.0001803535729419775, + "loss": 1.065, + "step": 4575 + }, + { + "epoch": 0.8148148148148148, + "grad_norm": 0.49955418705940247, + "learning_rate": 0.00018034524006817034, + "loss": 1.2752, + "step": 4576 + }, + { + "epoch": 0.8149928774928775, + "grad_norm": 0.5746406316757202, + "learning_rate": 0.00018033690562016508, + "loss": 1.098, + "step": 4577 + }, + { + "epoch": 0.8151709401709402, + "grad_norm": 0.5224192142486572, + "learning_rate": 0.00018032856959812507, + "loss": 1.1284, + "step": 4578 + }, + { + "epoch": 0.8153490028490028, + "grad_norm": 0.5484535694122314, + "learning_rate": 0.00018032023200221362, + "loss": 0.9182, + "step": 4579 + }, + { + "epoch": 0.8155270655270656, + "grad_norm": 0.5003355741500854, + "learning_rate": 0.00018031189283259405, + "loss": 1.136, + "step": 4580 + }, + { + "epoch": 0.8157051282051282, + "grad_norm": 0.5395768284797668, + "learning_rate": 0.00018030355208942977, + "loss": 1.2349, + "step": 4581 + }, + { + "epoch": 0.8158831908831908, + "grad_norm": 0.561966598033905, + "learning_rate": 0.0001802952097728842, + "loss": 0.999, + "step": 4582 + }, + { + "epoch": 0.8160612535612536, + "grad_norm": 0.4886479675769806, + "learning_rate": 0.00018028686588312083, + "loss": 0.9165, + "step": 4583 + }, + { + "epoch": 0.8162393162393162, + "grad_norm": 0.4769509732723236, + "learning_rate": 0.00018027852042030307, + "loss": 1.1377, + "step": 4584 + }, + { + "epoch": 0.8164173789173789, + "grad_norm": 0.4723633825778961, + "learning_rate": 0.00018027017338459448, + "loss": 1.0274, + "step": 4585 + }, + { + "epoch": 0.8165954415954416, + "grad_norm": 0.5773285627365112, + "learning_rate": 0.00018026182477615859, + "loss": 1.1468, + "step": 4586 + }, + { + "epoch": 0.8167735042735043, + "grad_norm": 0.5529203414916992, + "learning_rate": 0.00018025347459515895, + "loss": 1.0815, + "step": 4587 + }, + { + "epoch": 0.8169515669515669, + "grad_norm": 0.5449469685554504, + "learning_rate": 0.00018024512284175922, + "loss": 1.1637, + "step": 4588 + }, + { + "epoch": 0.8171296296296297, + "grad_norm": 0.5155341625213623, + "learning_rate": 0.00018023676951612298, + "loss": 1.1842, + "step": 4589 + }, + { + "epoch": 0.8173076923076923, + "grad_norm": 0.5569564700126648, + "learning_rate": 0.00018022841461841393, + "loss": 0.9254, + "step": 4590 + }, + { + "epoch": 0.8174857549857549, + "grad_norm": 0.45203131437301636, + "learning_rate": 0.00018022005814879573, + "loss": 0.9561, + "step": 4591 + }, + { + "epoch": 0.8176638176638177, + "grad_norm": 0.5735056400299072, + "learning_rate": 0.00018021170010743218, + "loss": 1.1402, + "step": 4592 + }, + { + "epoch": 0.8178418803418803, + "grad_norm": 0.6075260043144226, + "learning_rate": 0.00018020334049448697, + "loss": 0.8601, + "step": 4593 + }, + { + "epoch": 0.8180199430199431, + "grad_norm": 0.522682785987854, + "learning_rate": 0.0001801949793101239, + "loss": 1.0088, + "step": 4594 + }, + { + "epoch": 0.8181980056980057, + "grad_norm": 0.5648437142372131, + "learning_rate": 0.00018018661655450682, + "loss": 0.8359, + "step": 4595 + }, + { + "epoch": 0.8183760683760684, + "grad_norm": 0.5406472086906433, + "learning_rate": 0.00018017825222779954, + "loss": 1.1553, + "step": 4596 + }, + { + "epoch": 0.8185541310541311, + "grad_norm": 0.4917788803577423, + "learning_rate": 0.000180169886330166, + "loss": 1.2198, + "step": 4597 + }, + { + "epoch": 0.8187321937321937, + "grad_norm": 0.6293069124221802, + "learning_rate": 0.00018016151886177004, + "loss": 1.0245, + "step": 4598 + }, + { + "epoch": 0.8189102564102564, + "grad_norm": 0.47277843952178955, + "learning_rate": 0.00018015314982277564, + "loss": 1.1141, + "step": 4599 + }, + { + "epoch": 0.8190883190883191, + "grad_norm": 0.6132395267486572, + "learning_rate": 0.0001801447792133468, + "loss": 1.1227, + "step": 4600 + }, + { + "epoch": 0.8192663817663818, + "grad_norm": 0.46839597821235657, + "learning_rate": 0.00018013640703364747, + "loss": 0.9239, + "step": 4601 + }, + { + "epoch": 0.8194444444444444, + "grad_norm": 0.5055009722709656, + "learning_rate": 0.00018012803328384171, + "loss": 0.8486, + "step": 4602 + }, + { + "epoch": 0.8196225071225072, + "grad_norm": 0.5094841718673706, + "learning_rate": 0.00018011965796409362, + "loss": 0.9969, + "step": 4603 + }, + { + "epoch": 0.8198005698005698, + "grad_norm": 0.6177363395690918, + "learning_rate": 0.00018011128107456726, + "loss": 1.242, + "step": 4604 + }, + { + "epoch": 0.8199786324786325, + "grad_norm": 0.5280042290687561, + "learning_rate": 0.00018010290261542676, + "loss": 1.1569, + "step": 4605 + }, + { + "epoch": 0.8201566951566952, + "grad_norm": 0.5259367227554321, + "learning_rate": 0.00018009452258683625, + "loss": 0.9993, + "step": 4606 + }, + { + "epoch": 0.8203347578347578, + "grad_norm": 0.464469850063324, + "learning_rate": 0.00018008614098896, + "loss": 1.0288, + "step": 4607 + }, + { + "epoch": 0.8205128205128205, + "grad_norm": 0.6136324405670166, + "learning_rate": 0.00018007775782196214, + "loss": 1.1541, + "step": 4608 + }, + { + "epoch": 0.8206908831908832, + "grad_norm": 0.5376590490341187, + "learning_rate": 0.000180069373086007, + "loss": 1.0624, + "step": 4609 + }, + { + "epoch": 0.8208689458689459, + "grad_norm": 0.662916362285614, + "learning_rate": 0.0001800609867812588, + "loss": 1.1502, + "step": 4610 + }, + { + "epoch": 0.8210470085470085, + "grad_norm": 0.5153383612632751, + "learning_rate": 0.00018005259890788188, + "loss": 0.9789, + "step": 4611 + }, + { + "epoch": 0.8212250712250713, + "grad_norm": 0.5042359232902527, + "learning_rate": 0.00018004420946604057, + "loss": 0.9585, + "step": 4612 + }, + { + "epoch": 0.8214031339031339, + "grad_norm": 0.5395993590354919, + "learning_rate": 0.00018003581845589927, + "loss": 1.159, + "step": 4613 + }, + { + "epoch": 0.8215811965811965, + "grad_norm": 0.5561928749084473, + "learning_rate": 0.00018002742587762237, + "loss": 1.1604, + "step": 4614 + }, + { + "epoch": 0.8217592592592593, + "grad_norm": 0.5602710843086243, + "learning_rate": 0.00018001903173137432, + "loss": 0.9922, + "step": 4615 + }, + { + "epoch": 0.8219373219373219, + "grad_norm": 0.5529088377952576, + "learning_rate": 0.00018001063601731955, + "loss": 1.0943, + "step": 4616 + }, + { + "epoch": 0.8221153846153846, + "grad_norm": 0.5156456828117371, + "learning_rate": 0.00018000223873562254, + "loss": 1.1399, + "step": 4617 + }, + { + "epoch": 0.8222934472934473, + "grad_norm": 0.4868306517601013, + "learning_rate": 0.0001799938398864479, + "loss": 1.0692, + "step": 4618 + }, + { + "epoch": 0.82247150997151, + "grad_norm": 0.5372915267944336, + "learning_rate": 0.0001799854394699601, + "loss": 1.2675, + "step": 4619 + }, + { + "epoch": 0.8226495726495726, + "grad_norm": 0.6101839542388916, + "learning_rate": 0.0001799770374863238, + "loss": 0.9586, + "step": 4620 + }, + { + "epoch": 0.8228276353276354, + "grad_norm": 0.5034586787223816, + "learning_rate": 0.00017996863393570357, + "loss": 1.0885, + "step": 4621 + }, + { + "epoch": 0.823005698005698, + "grad_norm": 0.5608823299407959, + "learning_rate": 0.0001799602288182641, + "loss": 1.0002, + "step": 4622 + }, + { + "epoch": 0.8231837606837606, + "grad_norm": 0.5700048208236694, + "learning_rate": 0.00017995182213417, + "loss": 1.1484, + "step": 4623 + }, + { + "epoch": 0.8233618233618234, + "grad_norm": 0.5283229351043701, + "learning_rate": 0.00017994341388358608, + "loss": 1.0744, + "step": 4624 + }, + { + "epoch": 0.823539886039886, + "grad_norm": 0.5215758681297302, + "learning_rate": 0.00017993500406667703, + "loss": 1.2686, + "step": 4625 + }, + { + "epoch": 0.8237179487179487, + "grad_norm": 0.528883159160614, + "learning_rate": 0.0001799265926836076, + "loss": 1.1393, + "step": 4626 + }, + { + "epoch": 0.8238960113960114, + "grad_norm": 0.5589834451675415, + "learning_rate": 0.00017991817973454265, + "loss": 1.1744, + "step": 4627 + }, + { + "epoch": 0.8240740740740741, + "grad_norm": 0.49817174673080444, + "learning_rate": 0.00017990976521964697, + "loss": 1.0544, + "step": 4628 + }, + { + "epoch": 0.8242521367521367, + "grad_norm": 0.613961398601532, + "learning_rate": 0.00017990134913908542, + "loss": 1.0951, + "step": 4629 + }, + { + "epoch": 0.8244301994301995, + "grad_norm": 0.47278255224227905, + "learning_rate": 0.00017989293149302295, + "loss": 0.9742, + "step": 4630 + }, + { + "epoch": 0.8246082621082621, + "grad_norm": 0.49807092547416687, + "learning_rate": 0.00017988451228162443, + "loss": 1.0985, + "step": 4631 + }, + { + "epoch": 0.8247863247863247, + "grad_norm": 0.5624374747276306, + "learning_rate": 0.00017987609150505485, + "loss": 1.2446, + "step": 4632 + }, + { + "epoch": 0.8249643874643875, + "grad_norm": 0.4863535761833191, + "learning_rate": 0.00017986766916347916, + "loss": 1.0239, + "step": 4633 + }, + { + "epoch": 0.8251424501424501, + "grad_norm": 0.679585874080658, + "learning_rate": 0.00017985924525706245, + "loss": 1.1698, + "step": 4634 + }, + { + "epoch": 0.8253205128205128, + "grad_norm": 0.5545455813407898, + "learning_rate": 0.00017985081978596967, + "loss": 1.0926, + "step": 4635 + }, + { + "epoch": 0.8254985754985755, + "grad_norm": 0.5303109288215637, + "learning_rate": 0.000179842392750366, + "loss": 1.0978, + "step": 4636 + }, + { + "epoch": 0.8256766381766382, + "grad_norm": 0.6053299307823181, + "learning_rate": 0.00017983396415041644, + "loss": 1.0596, + "step": 4637 + }, + { + "epoch": 0.8258547008547008, + "grad_norm": 0.5241885185241699, + "learning_rate": 0.00017982553398628625, + "loss": 0.8541, + "step": 4638 + }, + { + "epoch": 0.8260327635327636, + "grad_norm": 0.5934443473815918, + "learning_rate": 0.00017981710225814052, + "loss": 1.145, + "step": 4639 + }, + { + "epoch": 0.8262108262108262, + "grad_norm": 0.5341619849205017, + "learning_rate": 0.00017980866896614447, + "loss": 1.0745, + "step": 4640 + }, + { + "epoch": 0.8263888888888888, + "grad_norm": 0.6732913851737976, + "learning_rate": 0.00017980023411046336, + "loss": 1.0775, + "step": 4641 + }, + { + "epoch": 0.8265669515669516, + "grad_norm": 0.5134359002113342, + "learning_rate": 0.0001797917976912624, + "loss": 1.0298, + "step": 4642 + }, + { + "epoch": 0.8267450142450142, + "grad_norm": 0.5234783887863159, + "learning_rate": 0.00017978335970870698, + "loss": 1.1069, + "step": 4643 + }, + { + "epoch": 0.8269230769230769, + "grad_norm": 0.4776439964771271, + "learning_rate": 0.00017977492016296232, + "loss": 0.6367, + "step": 4644 + }, + { + "epoch": 0.8271011396011396, + "grad_norm": 0.53763347864151, + "learning_rate": 0.0001797664790541938, + "loss": 1.1356, + "step": 4645 + }, + { + "epoch": 0.8272792022792023, + "grad_norm": 0.5082212686538696, + "learning_rate": 0.00017975803638256682, + "loss": 0.7873, + "step": 4646 + }, + { + "epoch": 0.8274572649572649, + "grad_norm": 0.5156424641609192, + "learning_rate": 0.00017974959214824685, + "loss": 1.084, + "step": 4647 + }, + { + "epoch": 0.8276353276353277, + "grad_norm": 0.5275198817253113, + "learning_rate": 0.00017974114635139926, + "loss": 1.1219, + "step": 4648 + }, + { + "epoch": 0.8278133903133903, + "grad_norm": 0.5548223257064819, + "learning_rate": 0.00017973269899218956, + "loss": 1.0808, + "step": 4649 + }, + { + "epoch": 0.8279914529914529, + "grad_norm": 0.535347580909729, + "learning_rate": 0.00017972425007078323, + "loss": 1.1211, + "step": 4650 + }, + { + "epoch": 0.8281695156695157, + "grad_norm": 0.5299580693244934, + "learning_rate": 0.00017971579958734587, + "loss": 0.9911, + "step": 4651 + }, + { + "epoch": 0.8283475783475783, + "grad_norm": 0.4863550066947937, + "learning_rate": 0.000179707347542043, + "loss": 0.9122, + "step": 4652 + }, + { + "epoch": 0.8285256410256411, + "grad_norm": 0.5284972190856934, + "learning_rate": 0.00017969889393504022, + "loss": 1.0424, + "step": 4653 + }, + { + "epoch": 0.8287037037037037, + "grad_norm": 0.5305661559104919, + "learning_rate": 0.00017969043876650317, + "loss": 1.1122, + "step": 4654 + }, + { + "epoch": 0.8288817663817664, + "grad_norm": 0.5645657777786255, + "learning_rate": 0.00017968198203659755, + "loss": 1.2195, + "step": 4655 + }, + { + "epoch": 0.8290598290598291, + "grad_norm": 0.521649181842804, + "learning_rate": 0.000179673523745489, + "loss": 1.2684, + "step": 4656 + }, + { + "epoch": 0.8292378917378918, + "grad_norm": 0.5984422564506531, + "learning_rate": 0.00017966506389334322, + "loss": 0.9894, + "step": 4657 + }, + { + "epoch": 0.8294159544159544, + "grad_norm": 0.5318729281425476, + "learning_rate": 0.00017965660248032603, + "loss": 1.2929, + "step": 4658 + }, + { + "epoch": 0.8295940170940171, + "grad_norm": 0.4666081368923187, + "learning_rate": 0.0001796481395066032, + "loss": 0.9646, + "step": 4659 + }, + { + "epoch": 0.8297720797720798, + "grad_norm": 0.5780388116836548, + "learning_rate": 0.00017963967497234054, + "loss": 1.1043, + "step": 4660 + }, + { + "epoch": 0.8299501424501424, + "grad_norm": 0.44089245796203613, + "learning_rate": 0.00017963120887770387, + "loss": 0.8932, + "step": 4661 + }, + { + "epoch": 0.8301282051282052, + "grad_norm": 0.5198349356651306, + "learning_rate": 0.0001796227412228591, + "loss": 0.9378, + "step": 4662 + }, + { + "epoch": 0.8303062678062678, + "grad_norm": 0.5298343896865845, + "learning_rate": 0.00017961427200797206, + "loss": 1.0272, + "step": 4663 + }, + { + "epoch": 0.8304843304843305, + "grad_norm": 0.5087099671363831, + "learning_rate": 0.0001796058012332088, + "loss": 0.989, + "step": 4664 + }, + { + "epoch": 0.8306623931623932, + "grad_norm": 0.504228949546814, + "learning_rate": 0.0001795973288987352, + "loss": 1.0134, + "step": 4665 + }, + { + "epoch": 0.8308404558404558, + "grad_norm": 0.6788033843040466, + "learning_rate": 0.00017958885500471728, + "loss": 0.8856, + "step": 4666 + }, + { + "epoch": 0.8310185185185185, + "grad_norm": 0.5166172385215759, + "learning_rate": 0.00017958037955132113, + "loss": 0.8711, + "step": 4667 + }, + { + "epoch": 0.8311965811965812, + "grad_norm": 0.5712400078773499, + "learning_rate": 0.00017957190253871272, + "loss": 1.0418, + "step": 4668 + }, + { + "epoch": 0.8313746438746439, + "grad_norm": 0.5531231164932251, + "learning_rate": 0.0001795634239670582, + "loss": 0.9021, + "step": 4669 + }, + { + "epoch": 0.8315527065527065, + "grad_norm": 0.6165615916252136, + "learning_rate": 0.00017955494383652365, + "loss": 1.0927, + "step": 4670 + }, + { + "epoch": 0.8317307692307693, + "grad_norm": 0.5920368432998657, + "learning_rate": 0.00017954646214727525, + "loss": 1.231, + "step": 4671 + }, + { + "epoch": 0.8319088319088319, + "grad_norm": 0.5037244558334351, + "learning_rate": 0.00017953797889947915, + "loss": 0.85, + "step": 4672 + }, + { + "epoch": 0.8320868945868946, + "grad_norm": 0.5618211627006531, + "learning_rate": 0.0001795294940933016, + "loss": 1.145, + "step": 4673 + }, + { + "epoch": 0.8322649572649573, + "grad_norm": 0.6275593042373657, + "learning_rate": 0.00017952100772890877, + "loss": 0.9061, + "step": 4674 + }, + { + "epoch": 0.83244301994302, + "grad_norm": 0.5376096367835999, + "learning_rate": 0.00017951251980646702, + "loss": 1.1948, + "step": 4675 + }, + { + "epoch": 0.8326210826210826, + "grad_norm": 0.5162268877029419, + "learning_rate": 0.0001795040303261426, + "loss": 1.2158, + "step": 4676 + }, + { + "epoch": 0.8327991452991453, + "grad_norm": 0.5730512142181396, + "learning_rate": 0.0001794955392881019, + "loss": 0.9962, + "step": 4677 + }, + { + "epoch": 0.832977207977208, + "grad_norm": 0.5128712058067322, + "learning_rate": 0.00017948704669251122, + "loss": 1.2797, + "step": 4678 + }, + { + "epoch": 0.8331552706552706, + "grad_norm": 0.5173979997634888, + "learning_rate": 0.00017947855253953697, + "loss": 1.1093, + "step": 4679 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 0.504646897315979, + "learning_rate": 0.0001794700568293456, + "loss": 1.3171, + "step": 4680 + }, + { + "epoch": 0.833511396011396, + "grad_norm": 0.5638105869293213, + "learning_rate": 0.00017946155956210356, + "loss": 0.9224, + "step": 4681 + }, + { + "epoch": 0.8336894586894587, + "grad_norm": 0.5289680361747742, + "learning_rate": 0.00017945306073797733, + "loss": 0.8919, + "step": 4682 + }, + { + "epoch": 0.8338675213675214, + "grad_norm": 0.5224629044532776, + "learning_rate": 0.0001794445603571334, + "loss": 1.0345, + "step": 4683 + }, + { + "epoch": 0.834045584045584, + "grad_norm": 0.5342282056808472, + "learning_rate": 0.00017943605841973836, + "loss": 1.2305, + "step": 4684 + }, + { + "epoch": 0.8342236467236467, + "grad_norm": 0.6118032336235046, + "learning_rate": 0.00017942755492595874, + "loss": 1.0316, + "step": 4685 + }, + { + "epoch": 0.8344017094017094, + "grad_norm": 0.49112311005592346, + "learning_rate": 0.00017941904987596121, + "loss": 0.9809, + "step": 4686 + }, + { + "epoch": 0.8345797720797721, + "grad_norm": 0.5044063925743103, + "learning_rate": 0.0001794105432699124, + "loss": 0.834, + "step": 4687 + }, + { + "epoch": 0.8347578347578347, + "grad_norm": 0.4849987328052521, + "learning_rate": 0.00017940203510797892, + "loss": 0.9971, + "step": 4688 + }, + { + "epoch": 0.8349358974358975, + "grad_norm": 0.5539469122886658, + "learning_rate": 0.00017939352539032748, + "loss": 1.1599, + "step": 4689 + }, + { + "epoch": 0.8351139601139601, + "grad_norm": 0.5474258065223694, + "learning_rate": 0.00017938501411712485, + "loss": 1.25, + "step": 4690 + }, + { + "epoch": 0.8352920227920227, + "grad_norm": 0.4880213737487793, + "learning_rate": 0.0001793765012885378, + "loss": 1.1471, + "step": 4691 + }, + { + "epoch": 0.8354700854700855, + "grad_norm": 0.5602759718894958, + "learning_rate": 0.00017936798690473309, + "loss": 1.0723, + "step": 4692 + }, + { + "epoch": 0.8356481481481481, + "grad_norm": 0.627775251865387, + "learning_rate": 0.00017935947096587755, + "loss": 1.3768, + "step": 4693 + }, + { + "epoch": 0.8358262108262108, + "grad_norm": 0.5324847102165222, + "learning_rate": 0.00017935095347213804, + "loss": 0.9945, + "step": 4694 + }, + { + "epoch": 0.8360042735042735, + "grad_norm": 0.5244048237800598, + "learning_rate": 0.0001793424344236814, + "loss": 1.1725, + "step": 4695 + }, + { + "epoch": 0.8361823361823362, + "grad_norm": 0.5420708656311035, + "learning_rate": 0.00017933391382067462, + "loss": 1.1267, + "step": 4696 + }, + { + "epoch": 0.8363603988603988, + "grad_norm": 0.5285456776618958, + "learning_rate": 0.00017932539166328458, + "loss": 1.0368, + "step": 4697 + }, + { + "epoch": 0.8365384615384616, + "grad_norm": 0.5330373048782349, + "learning_rate": 0.00017931686795167825, + "loss": 1.1082, + "step": 4698 + }, + { + "epoch": 0.8367165242165242, + "grad_norm": 0.5516682267189026, + "learning_rate": 0.0001793083426860227, + "loss": 1.1833, + "step": 4699 + }, + { + "epoch": 0.8368945868945868, + "grad_norm": 0.5229935646057129, + "learning_rate": 0.0001792998158664849, + "loss": 0.8527, + "step": 4700 + }, + { + "epoch": 0.8370726495726496, + "grad_norm": 0.4821490943431854, + "learning_rate": 0.00017929128749323195, + "loss": 1.1201, + "step": 4701 + }, + { + "epoch": 0.8372507122507122, + "grad_norm": 0.6276404857635498, + "learning_rate": 0.0001792827575664309, + "loss": 1.0986, + "step": 4702 + }, + { + "epoch": 0.8374287749287749, + "grad_norm": 0.5681334733963013, + "learning_rate": 0.00017927422608624897, + "loss": 1.3821, + "step": 4703 + }, + { + "epoch": 0.8376068376068376, + "grad_norm": 0.5257087349891663, + "learning_rate": 0.00017926569305285324, + "loss": 1.1033, + "step": 4704 + }, + { + "epoch": 0.8377849002849003, + "grad_norm": 0.5665168166160583, + "learning_rate": 0.0001792571584664109, + "loss": 1.104, + "step": 4705 + }, + { + "epoch": 0.8379629629629629, + "grad_norm": 0.5202076435089111, + "learning_rate": 0.00017924862232708918, + "loss": 1.052, + "step": 4706 + }, + { + "epoch": 0.8381410256410257, + "grad_norm": 0.5103010535240173, + "learning_rate": 0.00017924008463505534, + "loss": 1.1348, + "step": 4707 + }, + { + "epoch": 0.8383190883190883, + "grad_norm": 0.6811865568161011, + "learning_rate": 0.00017923154539047667, + "loss": 1.2804, + "step": 4708 + }, + { + "epoch": 0.8384971509971509, + "grad_norm": 0.46808311343193054, + "learning_rate": 0.00017922300459352042, + "loss": 0.9302, + "step": 4709 + }, + { + "epoch": 0.8386752136752137, + "grad_norm": 0.47713059186935425, + "learning_rate": 0.00017921446224435398, + "loss": 0.78, + "step": 4710 + }, + { + "epoch": 0.8388532763532763, + "grad_norm": 0.7579890489578247, + "learning_rate": 0.0001792059183431447, + "loss": 1.4776, + "step": 4711 + }, + { + "epoch": 0.8390313390313391, + "grad_norm": 0.6009423136711121, + "learning_rate": 0.00017919737289006, + "loss": 1.2679, + "step": 4712 + }, + { + "epoch": 0.8392094017094017, + "grad_norm": 0.56390780210495, + "learning_rate": 0.00017918882588526729, + "loss": 1.0402, + "step": 4713 + }, + { + "epoch": 0.8393874643874644, + "grad_norm": 0.5698862075805664, + "learning_rate": 0.00017918027732893404, + "loss": 1.2336, + "step": 4714 + }, + { + "epoch": 0.8395655270655271, + "grad_norm": 0.5016305446624756, + "learning_rate": 0.0001791717272212277, + "loss": 1.0373, + "step": 4715 + }, + { + "epoch": 0.8397435897435898, + "grad_norm": 0.5886971950531006, + "learning_rate": 0.0001791631755623159, + "loss": 1.1062, + "step": 4716 + }, + { + "epoch": 0.8399216524216524, + "grad_norm": 0.647833526134491, + "learning_rate": 0.00017915462235236607, + "loss": 1.0464, + "step": 4717 + }, + { + "epoch": 0.8400997150997151, + "grad_norm": 0.4961194396018982, + "learning_rate": 0.00017914606759154587, + "loss": 1.0763, + "step": 4718 + }, + { + "epoch": 0.8402777777777778, + "grad_norm": 0.47041359543800354, + "learning_rate": 0.00017913751128002288, + "loss": 1.0685, + "step": 4719 + }, + { + "epoch": 0.8404558404558404, + "grad_norm": 0.5752858519554138, + "learning_rate": 0.00017912895341796475, + "loss": 1.0577, + "step": 4720 + }, + { + "epoch": 0.8406339031339032, + "grad_norm": 0.5233224034309387, + "learning_rate": 0.00017912039400553914, + "loss": 1.1484, + "step": 4721 + }, + { + "epoch": 0.8408119658119658, + "grad_norm": 0.5327485203742981, + "learning_rate": 0.00017911183304291378, + "loss": 1.0028, + "step": 4722 + }, + { + "epoch": 0.8409900284900285, + "grad_norm": 0.5320752263069153, + "learning_rate": 0.00017910327053025638, + "loss": 1.1247, + "step": 4723 + }, + { + "epoch": 0.8411680911680912, + "grad_norm": 0.529617965221405, + "learning_rate": 0.00017909470646773477, + "loss": 1.1698, + "step": 4724 + }, + { + "epoch": 0.8413461538461539, + "grad_norm": 0.5055609345436096, + "learning_rate": 0.00017908614085551664, + "loss": 1.0925, + "step": 4725 + }, + { + "epoch": 0.8415242165242165, + "grad_norm": 0.5356255769729614, + "learning_rate": 0.00017907757369376985, + "loss": 1.0354, + "step": 4726 + }, + { + "epoch": 0.8417022792022792, + "grad_norm": 0.582834780216217, + "learning_rate": 0.00017906900498266233, + "loss": 1.1248, + "step": 4727 + }, + { + "epoch": 0.8418803418803419, + "grad_norm": 0.5750834941864014, + "learning_rate": 0.00017906043472236188, + "loss": 1.0119, + "step": 4728 + }, + { + "epoch": 0.8420584045584045, + "grad_norm": 0.5923320055007935, + "learning_rate": 0.00017905186291303644, + "loss": 1.0662, + "step": 4729 + }, + { + "epoch": 0.8422364672364673, + "grad_norm": 0.4767811894416809, + "learning_rate": 0.00017904328955485396, + "loss": 1.0911, + "step": 4730 + }, + { + "epoch": 0.8424145299145299, + "grad_norm": 0.5294556021690369, + "learning_rate": 0.00017903471464798245, + "loss": 1.2861, + "step": 4731 + }, + { + "epoch": 0.8425925925925926, + "grad_norm": 0.599117636680603, + "learning_rate": 0.00017902613819258985, + "loss": 1.1707, + "step": 4732 + }, + { + "epoch": 0.8427706552706553, + "grad_norm": 0.5912977457046509, + "learning_rate": 0.00017901756018884424, + "loss": 1.1884, + "step": 4733 + }, + { + "epoch": 0.842948717948718, + "grad_norm": 0.587676465511322, + "learning_rate": 0.0001790089806369137, + "loss": 1.1054, + "step": 4734 + }, + { + "epoch": 0.8431267806267806, + "grad_norm": 0.6271800398826599, + "learning_rate": 0.0001790003995369663, + "loss": 1.2094, + "step": 4735 + }, + { + "epoch": 0.8433048433048433, + "grad_norm": 0.47198590636253357, + "learning_rate": 0.00017899181688917017, + "loss": 0.9561, + "step": 4736 + }, + { + "epoch": 0.843482905982906, + "grad_norm": 0.690732479095459, + "learning_rate": 0.00017898323269369351, + "loss": 1.1629, + "step": 4737 + }, + { + "epoch": 0.8436609686609686, + "grad_norm": 0.4926888048648834, + "learning_rate": 0.00017897464695070445, + "loss": 1.1097, + "step": 4738 + }, + { + "epoch": 0.8438390313390314, + "grad_norm": 0.7071278691291809, + "learning_rate": 0.00017896605966037128, + "loss": 1.195, + "step": 4739 + }, + { + "epoch": 0.844017094017094, + "grad_norm": 0.5650486350059509, + "learning_rate": 0.00017895747082286216, + "loss": 1.0107, + "step": 4740 + }, + { + "epoch": 0.8441951566951567, + "grad_norm": 0.5291931629180908, + "learning_rate": 0.00017894888043834545, + "loss": 1.0104, + "step": 4741 + }, + { + "epoch": 0.8443732193732194, + "grad_norm": 0.5751241445541382, + "learning_rate": 0.00017894028850698942, + "loss": 1.2482, + "step": 4742 + }, + { + "epoch": 0.844551282051282, + "grad_norm": 0.5833632349967957, + "learning_rate": 0.0001789316950289624, + "loss": 1.0552, + "step": 4743 + }, + { + "epoch": 0.8447293447293447, + "grad_norm": 0.543729841709137, + "learning_rate": 0.00017892310000443282, + "loss": 1.1453, + "step": 4744 + }, + { + "epoch": 0.8449074074074074, + "grad_norm": 0.5674204230308533, + "learning_rate": 0.00017891450343356902, + "loss": 1.0757, + "step": 4745 + }, + { + "epoch": 0.8450854700854701, + "grad_norm": 0.5161892771720886, + "learning_rate": 0.00017890590531653946, + "loss": 1.1163, + "step": 4746 + }, + { + "epoch": 0.8452635327635327, + "grad_norm": 0.49907612800598145, + "learning_rate": 0.00017889730565351258, + "loss": 1.0356, + "step": 4747 + }, + { + "epoch": 0.8454415954415955, + "grad_norm": 0.4994732439517975, + "learning_rate": 0.00017888870444465692, + "loss": 1.026, + "step": 4748 + }, + { + "epoch": 0.8456196581196581, + "grad_norm": 0.6397520303726196, + "learning_rate": 0.00017888010169014095, + "loss": 0.957, + "step": 4749 + }, + { + "epoch": 0.8457977207977208, + "grad_norm": 0.5379729270935059, + "learning_rate": 0.00017887149739013327, + "loss": 1.1664, + "step": 4750 + }, + { + "epoch": 0.8459757834757835, + "grad_norm": 0.4487382769584656, + "learning_rate": 0.00017886289154480246, + "loss": 0.9377, + "step": 4751 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.5645943880081177, + "learning_rate": 0.00017885428415431707, + "loss": 1.273, + "step": 4752 + }, + { + "epoch": 0.8463319088319088, + "grad_norm": 0.5535289645195007, + "learning_rate": 0.00017884567521884577, + "loss": 1.1779, + "step": 4753 + }, + { + "epoch": 0.8465099715099715, + "grad_norm": 0.5039721131324768, + "learning_rate": 0.0001788370647385573, + "loss": 1.0237, + "step": 4754 + }, + { + "epoch": 0.8466880341880342, + "grad_norm": 0.4543854892253876, + "learning_rate": 0.00017882845271362032, + "loss": 0.8149, + "step": 4755 + }, + { + "epoch": 0.8468660968660968, + "grad_norm": 0.5095639824867249, + "learning_rate": 0.00017881983914420352, + "loss": 1.0141, + "step": 4756 + }, + { + "epoch": 0.8470441595441596, + "grad_norm": 0.5341798663139343, + "learning_rate": 0.00017881122403047575, + "loss": 1.1885, + "step": 4757 + }, + { + "epoch": 0.8472222222222222, + "grad_norm": 0.5595062971115112, + "learning_rate": 0.00017880260737260573, + "loss": 0.8939, + "step": 4758 + }, + { + "epoch": 0.8474002849002849, + "grad_norm": 0.5355880260467529, + "learning_rate": 0.00017879398917076232, + "loss": 1.2434, + "step": 4759 + }, + { + "epoch": 0.8475783475783476, + "grad_norm": 0.49477261304855347, + "learning_rate": 0.0001787853694251144, + "loss": 0.979, + "step": 4760 + }, + { + "epoch": 0.8477564102564102, + "grad_norm": 0.5154359340667725, + "learning_rate": 0.00017877674813583078, + "loss": 1.0957, + "step": 4761 + }, + { + "epoch": 0.8479344729344729, + "grad_norm": 0.5651070475578308, + "learning_rate": 0.00017876812530308046, + "loss": 1.1884, + "step": 4762 + }, + { + "epoch": 0.8481125356125356, + "grad_norm": 0.537277340888977, + "learning_rate": 0.00017875950092703232, + "loss": 1.0272, + "step": 4763 + }, + { + "epoch": 0.8482905982905983, + "grad_norm": 0.5259691476821899, + "learning_rate": 0.00017875087500785538, + "loss": 1.1493, + "step": 4764 + }, + { + "epoch": 0.8484686609686609, + "grad_norm": 0.5491300225257874, + "learning_rate": 0.00017874224754571867, + "loss": 0.8316, + "step": 4765 + }, + { + "epoch": 0.8486467236467237, + "grad_norm": 0.5493744611740112, + "learning_rate": 0.00017873361854079116, + "loss": 1.2328, + "step": 4766 + }, + { + "epoch": 0.8488247863247863, + "grad_norm": 0.571002185344696, + "learning_rate": 0.00017872498799324197, + "loss": 1.1384, + "step": 4767 + }, + { + "epoch": 0.8490028490028491, + "grad_norm": 0.538152813911438, + "learning_rate": 0.00017871635590324013, + "loss": 1.0581, + "step": 4768 + }, + { + "epoch": 0.8491809116809117, + "grad_norm": 0.5214923620223999, + "learning_rate": 0.00017870772227095486, + "loss": 1.0612, + "step": 4769 + }, + { + "epoch": 0.8493589743589743, + "grad_norm": 0.5714883804321289, + "learning_rate": 0.0001786990870965553, + "loss": 0.9076, + "step": 4770 + }, + { + "epoch": 0.8495370370370371, + "grad_norm": 0.4181775450706482, + "learning_rate": 0.00017869045038021054, + "loss": 0.8366, + "step": 4771 + }, + { + "epoch": 0.8497150997150997, + "grad_norm": 0.6266027688980103, + "learning_rate": 0.00017868181212208993, + "loss": 1.2047, + "step": 4772 + }, + { + "epoch": 0.8498931623931624, + "grad_norm": 0.5423732399940491, + "learning_rate": 0.0001786731723223626, + "loss": 1.3878, + "step": 4773 + }, + { + "epoch": 0.8500712250712251, + "grad_norm": 0.5512300133705139, + "learning_rate": 0.00017866453098119793, + "loss": 1.1132, + "step": 4774 + }, + { + "epoch": 0.8502492877492878, + "grad_norm": 0.5767185688018799, + "learning_rate": 0.00017865588809876519, + "loss": 0.97, + "step": 4775 + }, + { + "epoch": 0.8504273504273504, + "grad_norm": 0.5305790305137634, + "learning_rate": 0.00017864724367523368, + "loss": 1.1158, + "step": 4776 + }, + { + "epoch": 0.8506054131054132, + "grad_norm": 0.49702391028404236, + "learning_rate": 0.00017863859771077284, + "loss": 0.9669, + "step": 4777 + }, + { + "epoch": 0.8507834757834758, + "grad_norm": 0.5490063428878784, + "learning_rate": 0.00017862995020555205, + "loss": 1.0646, + "step": 4778 + }, + { + "epoch": 0.8509615384615384, + "grad_norm": 0.5308689475059509, + "learning_rate": 0.00017862130115974068, + "loss": 0.8922, + "step": 4779 + }, + { + "epoch": 0.8511396011396012, + "grad_norm": 0.5412983894348145, + "learning_rate": 0.00017861265057350826, + "loss": 1.1444, + "step": 4780 + }, + { + "epoch": 0.8513176638176638, + "grad_norm": 0.5857377052307129, + "learning_rate": 0.00017860399844702425, + "loss": 1.1643, + "step": 4781 + }, + { + "epoch": 0.8514957264957265, + "grad_norm": 0.599273681640625, + "learning_rate": 0.00017859534478045815, + "loss": 1.169, + "step": 4782 + }, + { + "epoch": 0.8516737891737892, + "grad_norm": 0.5677087903022766, + "learning_rate": 0.00017858668957397957, + "loss": 1.0793, + "step": 4783 + }, + { + "epoch": 0.8518518518518519, + "grad_norm": 0.5648362636566162, + "learning_rate": 0.00017857803282775807, + "loss": 1.1932, + "step": 4784 + }, + { + "epoch": 0.8520299145299145, + "grad_norm": 0.5138826966285706, + "learning_rate": 0.00017856937454196323, + "loss": 1.0011, + "step": 4785 + }, + { + "epoch": 0.8522079772079773, + "grad_norm": 0.5951429009437561, + "learning_rate": 0.0001785607147167647, + "loss": 1.3198, + "step": 4786 + }, + { + "epoch": 0.8523860398860399, + "grad_norm": 0.5341953039169312, + "learning_rate": 0.00017855205335233216, + "loss": 0.9094, + "step": 4787 + }, + { + "epoch": 0.8525641025641025, + "grad_norm": 0.5193579196929932, + "learning_rate": 0.00017854339044883535, + "loss": 0.892, + "step": 4788 + }, + { + "epoch": 0.8527421652421653, + "grad_norm": 0.5053097009658813, + "learning_rate": 0.00017853472600644392, + "loss": 1.0589, + "step": 4789 + }, + { + "epoch": 0.8529202279202279, + "grad_norm": 0.5819617509841919, + "learning_rate": 0.0001785260600253277, + "loss": 1.2646, + "step": 4790 + }, + { + "epoch": 0.8530982905982906, + "grad_norm": 0.5327470302581787, + "learning_rate": 0.00017851739250565645, + "loss": 1.056, + "step": 4791 + }, + { + "epoch": 0.8532763532763533, + "grad_norm": 0.5131269097328186, + "learning_rate": 0.0001785087234476, + "loss": 1.1192, + "step": 4792 + }, + { + "epoch": 0.853454415954416, + "grad_norm": 0.4698086977005005, + "learning_rate": 0.00017850005285132821, + "loss": 0.9849, + "step": 4793 + }, + { + "epoch": 0.8536324786324786, + "grad_norm": 0.5503947734832764, + "learning_rate": 0.00017849138071701092, + "loss": 1.1139, + "step": 4794 + }, + { + "epoch": 0.8538105413105413, + "grad_norm": 0.5120903849601746, + "learning_rate": 0.0001784827070448181, + "loss": 0.9801, + "step": 4795 + }, + { + "epoch": 0.853988603988604, + "grad_norm": 0.47650405764579773, + "learning_rate": 0.00017847403183491968, + "loss": 1.0268, + "step": 4796 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 0.5773387551307678, + "learning_rate": 0.0001784653550874856, + "loss": 1.0336, + "step": 4797 + }, + { + "epoch": 0.8543447293447294, + "grad_norm": 0.545531153678894, + "learning_rate": 0.00017845667680268593, + "loss": 1.0532, + "step": 4798 + }, + { + "epoch": 0.854522792022792, + "grad_norm": 0.533161461353302, + "learning_rate": 0.0001784479969806906, + "loss": 1.1964, + "step": 4799 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 0.5880789160728455, + "learning_rate": 0.00017843931562166977, + "loss": 1.1588, + "step": 4800 + }, + { + "epoch": 0.8548789173789174, + "grad_norm": 0.5381524562835693, + "learning_rate": 0.00017843063272579346, + "loss": 1.1533, + "step": 4801 + }, + { + "epoch": 0.85505698005698, + "grad_norm": 0.6280176639556885, + "learning_rate": 0.00017842194829323187, + "loss": 1.0084, + "step": 4802 + }, + { + "epoch": 0.8552350427350427, + "grad_norm": 0.5098552703857422, + "learning_rate": 0.0001784132623241551, + "loss": 1.0804, + "step": 4803 + }, + { + "epoch": 0.8554131054131054, + "grad_norm": 0.5406526923179626, + "learning_rate": 0.00017840457481873328, + "loss": 1.2571, + "step": 4804 + }, + { + "epoch": 0.8555911680911681, + "grad_norm": 0.5859003663063049, + "learning_rate": 0.00017839588577713678, + "loss": 1.2462, + "step": 4805 + }, + { + "epoch": 0.8557692307692307, + "grad_norm": 0.6209002137184143, + "learning_rate": 0.00017838719519953572, + "loss": 1.307, + "step": 4806 + }, + { + "epoch": 0.8559472934472935, + "grad_norm": 0.525753915309906, + "learning_rate": 0.00017837850308610037, + "loss": 1.2957, + "step": 4807 + }, + { + "epoch": 0.8561253561253561, + "grad_norm": 0.5096195340156555, + "learning_rate": 0.0001783698094370011, + "loss": 1.1433, + "step": 4808 + }, + { + "epoch": 0.8563034188034188, + "grad_norm": 0.5873076915740967, + "learning_rate": 0.0001783611142524082, + "loss": 1.2271, + "step": 4809 + }, + { + "epoch": 0.8564814814814815, + "grad_norm": 0.5093944668769836, + "learning_rate": 0.0001783524175324921, + "loss": 0.8788, + "step": 4810 + }, + { + "epoch": 0.8566595441595442, + "grad_norm": 0.5485084652900696, + "learning_rate": 0.00017834371927742307, + "loss": 1.256, + "step": 4811 + }, + { + "epoch": 0.8568376068376068, + "grad_norm": 0.5808873772621155, + "learning_rate": 0.00017833501948737163, + "loss": 0.9287, + "step": 4812 + }, + { + "epoch": 0.8570156695156695, + "grad_norm": 0.5113978385925293, + "learning_rate": 0.00017832631816250822, + "loss": 1.0372, + "step": 4813 + }, + { + "epoch": 0.8571937321937322, + "grad_norm": 0.5877016186714172, + "learning_rate": 0.0001783176153030033, + "loss": 1.3023, + "step": 4814 + }, + { + "epoch": 0.8573717948717948, + "grad_norm": 0.534328043460846, + "learning_rate": 0.00017830891090902742, + "loss": 1.1023, + "step": 4815 + }, + { + "epoch": 0.8575498575498576, + "grad_norm": 0.5781638026237488, + "learning_rate": 0.0001783002049807511, + "loss": 0.9562, + "step": 4816 + }, + { + "epoch": 0.8577279202279202, + "grad_norm": 0.5760263204574585, + "learning_rate": 0.00017829149751834487, + "loss": 0.8733, + "step": 4817 + }, + { + "epoch": 0.8579059829059829, + "grad_norm": 0.3887255787849426, + "learning_rate": 0.00017828278852197944, + "loss": 0.5949, + "step": 4818 + }, + { + "epoch": 0.8580840455840456, + "grad_norm": 0.47814446687698364, + "learning_rate": 0.00017827407799182537, + "loss": 1.0698, + "step": 4819 + }, + { + "epoch": 0.8582621082621082, + "grad_norm": 0.5520272254943848, + "learning_rate": 0.00017826536592805334, + "loss": 1.1314, + "step": 4820 + }, + { + "epoch": 0.8584401709401709, + "grad_norm": 0.5285319685935974, + "learning_rate": 0.00017825665233083405, + "loss": 1.1618, + "step": 4821 + }, + { + "epoch": 0.8586182336182336, + "grad_norm": 0.6080102324485779, + "learning_rate": 0.0001782479372003382, + "loss": 1.3817, + "step": 4822 + }, + { + "epoch": 0.8587962962962963, + "grad_norm": 0.7474410533905029, + "learning_rate": 0.00017823922053673662, + "loss": 1.1321, + "step": 4823 + }, + { + "epoch": 0.8589743589743589, + "grad_norm": 0.559283435344696, + "learning_rate": 0.0001782305023402, + "loss": 1.1894, + "step": 4824 + }, + { + "epoch": 0.8591524216524217, + "grad_norm": 0.5620571374893188, + "learning_rate": 0.00017822178261089918, + "loss": 1.134, + "step": 4825 + }, + { + "epoch": 0.8593304843304843, + "grad_norm": 0.5553044676780701, + "learning_rate": 0.00017821306134900504, + "loss": 1.3222, + "step": 4826 + }, + { + "epoch": 0.8595085470085471, + "grad_norm": 0.6177778244018555, + "learning_rate": 0.00017820433855468846, + "loss": 1.2545, + "step": 4827 + }, + { + "epoch": 0.8596866096866097, + "grad_norm": 0.656233012676239, + "learning_rate": 0.0001781956142281203, + "loss": 1.1346, + "step": 4828 + }, + { + "epoch": 0.8598646723646723, + "grad_norm": 0.6710973381996155, + "learning_rate": 0.0001781868883694715, + "loss": 1.1361, + "step": 4829 + }, + { + "epoch": 0.8600427350427351, + "grad_norm": 0.5093601942062378, + "learning_rate": 0.0001781781609789131, + "loss": 1.0509, + "step": 4830 + }, + { + "epoch": 0.8602207977207977, + "grad_norm": 0.5707578063011169, + "learning_rate": 0.00017816943205661598, + "loss": 1.0964, + "step": 4831 + }, + { + "epoch": 0.8603988603988604, + "grad_norm": 0.6159597635269165, + "learning_rate": 0.00017816070160275125, + "loss": 1.0322, + "step": 4832 + }, + { + "epoch": 0.8605769230769231, + "grad_norm": 0.5430580377578735, + "learning_rate": 0.0001781519696174899, + "loss": 1.2464, + "step": 4833 + }, + { + "epoch": 0.8607549857549858, + "grad_norm": 0.48104700446128845, + "learning_rate": 0.0001781432361010031, + "loss": 1.1031, + "step": 4834 + }, + { + "epoch": 0.8609330484330484, + "grad_norm": 0.5304946303367615, + "learning_rate": 0.0001781345010534619, + "loss": 1.0281, + "step": 4835 + }, + { + "epoch": 0.8611111111111112, + "grad_norm": 0.5230711698532104, + "learning_rate": 0.00017812576447503742, + "loss": 0.9499, + "step": 4836 + }, + { + "epoch": 0.8612891737891738, + "grad_norm": 0.5363606214523315, + "learning_rate": 0.00017811702636590093, + "loss": 1.1358, + "step": 4837 + }, + { + "epoch": 0.8614672364672364, + "grad_norm": 0.5880044102668762, + "learning_rate": 0.00017810828672622358, + "loss": 1.1765, + "step": 4838 + }, + { + "epoch": 0.8616452991452992, + "grad_norm": 0.5194395184516907, + "learning_rate": 0.0001780995455561766, + "loss": 1.1622, + "step": 4839 + }, + { + "epoch": 0.8618233618233618, + "grad_norm": 0.5114264488220215, + "learning_rate": 0.00017809080285593126, + "loss": 1.0081, + "step": 4840 + }, + { + "epoch": 0.8620014245014245, + "grad_norm": 0.6174240112304688, + "learning_rate": 0.00017808205862565886, + "loss": 1.0745, + "step": 4841 + }, + { + "epoch": 0.8621794871794872, + "grad_norm": 0.5662630200386047, + "learning_rate": 0.0001780733128655307, + "loss": 1.3369, + "step": 4842 + }, + { + "epoch": 0.8623575498575499, + "grad_norm": 0.5917882919311523, + "learning_rate": 0.00017806456557571817, + "loss": 1.1631, + "step": 4843 + }, + { + "epoch": 0.8625356125356125, + "grad_norm": 0.5305736660957336, + "learning_rate": 0.00017805581675639265, + "loss": 0.9875, + "step": 4844 + }, + { + "epoch": 0.8627136752136753, + "grad_norm": 0.5181219577789307, + "learning_rate": 0.00017804706640772556, + "loss": 0.9918, + "step": 4845 + }, + { + "epoch": 0.8628917378917379, + "grad_norm": 0.5467997789382935, + "learning_rate": 0.00017803831452988832, + "loss": 1.1395, + "step": 4846 + }, + { + "epoch": 0.8630698005698005, + "grad_norm": 0.5494031310081482, + "learning_rate": 0.00017802956112305241, + "loss": 1.0312, + "step": 4847 + }, + { + "epoch": 0.8632478632478633, + "grad_norm": 0.5804065465927124, + "learning_rate": 0.00017802080618738931, + "loss": 1.1555, + "step": 4848 + }, + { + "epoch": 0.8634259259259259, + "grad_norm": 0.5424801111221313, + "learning_rate": 0.00017801204972307067, + "loss": 1.0215, + "step": 4849 + }, + { + "epoch": 0.8636039886039886, + "grad_norm": 0.5321891903877258, + "learning_rate": 0.0001780032917302679, + "loss": 1.0187, + "step": 4850 + }, + { + "epoch": 0.8637820512820513, + "grad_norm": 0.5543400049209595, + "learning_rate": 0.0001779945322091527, + "loss": 1.1972, + "step": 4851 + }, + { + "epoch": 0.863960113960114, + "grad_norm": 0.566649317741394, + "learning_rate": 0.00017798577115989668, + "loss": 1.0758, + "step": 4852 + }, + { + "epoch": 0.8641381766381766, + "grad_norm": 0.5538444519042969, + "learning_rate": 0.00017797700858267145, + "loss": 1.1338, + "step": 4853 + }, + { + "epoch": 0.8643162393162394, + "grad_norm": 0.5641313791275024, + "learning_rate": 0.0001779682444776487, + "loss": 1.256, + "step": 4854 + }, + { + "epoch": 0.864494301994302, + "grad_norm": 0.6377350091934204, + "learning_rate": 0.00017795947884500016, + "loss": 1.144, + "step": 4855 + }, + { + "epoch": 0.8646723646723646, + "grad_norm": 0.5581876039505005, + "learning_rate": 0.0001779507116848976, + "loss": 1.3163, + "step": 4856 + }, + { + "epoch": 0.8648504273504274, + "grad_norm": 0.5416772365570068, + "learning_rate": 0.0001779419429975128, + "loss": 1.0219, + "step": 4857 + }, + { + "epoch": 0.86502849002849, + "grad_norm": 0.5450608730316162, + "learning_rate": 0.0001779331727830175, + "loss": 1.0093, + "step": 4858 + }, + { + "epoch": 0.8652065527065527, + "grad_norm": 0.5151242017745972, + "learning_rate": 0.00017792440104158358, + "loss": 1.067, + "step": 4859 + }, + { + "epoch": 0.8653846153846154, + "grad_norm": 0.5225046873092651, + "learning_rate": 0.0001779156277733829, + "loss": 1.0432, + "step": 4860 + }, + { + "epoch": 0.8655626780626781, + "grad_norm": 0.5168602466583252, + "learning_rate": 0.00017790685297858737, + "loss": 0.9665, + "step": 4861 + }, + { + "epoch": 0.8657407407407407, + "grad_norm": 0.5749059319496155, + "learning_rate": 0.00017789807665736889, + "loss": 1.1607, + "step": 4862 + }, + { + "epoch": 0.8659188034188035, + "grad_norm": 0.45656394958496094, + "learning_rate": 0.00017788929880989938, + "loss": 0.8362, + "step": 4863 + }, + { + "epoch": 0.8660968660968661, + "grad_norm": 0.5090615749359131, + "learning_rate": 0.00017788051943635086, + "loss": 0.9553, + "step": 4864 + }, + { + "epoch": 0.8662749287749287, + "grad_norm": 0.5381240248680115, + "learning_rate": 0.0001778717385368954, + "loss": 1.1391, + "step": 4865 + }, + { + "epoch": 0.8664529914529915, + "grad_norm": 0.522720456123352, + "learning_rate": 0.00017786295611170493, + "loss": 1.1869, + "step": 4866 + }, + { + "epoch": 0.8666310541310541, + "grad_norm": 0.530986487865448, + "learning_rate": 0.0001778541721609516, + "loss": 1.1046, + "step": 4867 + }, + { + "epoch": 0.8668091168091168, + "grad_norm": 0.5065864324569702, + "learning_rate": 0.0001778453866848075, + "loss": 1.008, + "step": 4868 + }, + { + "epoch": 0.8669871794871795, + "grad_norm": 0.5541394352912903, + "learning_rate": 0.00017783659968344476, + "loss": 1.0004, + "step": 4869 + }, + { + "epoch": 0.8671652421652422, + "grad_norm": 0.5059576630592346, + "learning_rate": 0.00017782781115703556, + "loss": 1.128, + "step": 4870 + }, + { + "epoch": 0.8673433048433048, + "grad_norm": 0.5052187442779541, + "learning_rate": 0.00017781902110575203, + "loss": 0.8544, + "step": 4871 + }, + { + "epoch": 0.8675213675213675, + "grad_norm": 0.5383397340774536, + "learning_rate": 0.00017781022952976646, + "loss": 1.1411, + "step": 4872 + }, + { + "epoch": 0.8676994301994302, + "grad_norm": 0.4760429859161377, + "learning_rate": 0.00017780143642925106, + "loss": 0.8246, + "step": 4873 + }, + { + "epoch": 0.8678774928774928, + "grad_norm": 0.5480535626411438, + "learning_rate": 0.00017779264180437817, + "loss": 1.013, + "step": 4874 + }, + { + "epoch": 0.8680555555555556, + "grad_norm": 0.5303317904472351, + "learning_rate": 0.00017778384565532004, + "loss": 1.0201, + "step": 4875 + }, + { + "epoch": 0.8682336182336182, + "grad_norm": 0.5365355014801025, + "learning_rate": 0.00017777504798224903, + "loss": 1.1107, + "step": 4876 + }, + { + "epoch": 0.8684116809116809, + "grad_norm": 0.5173360705375671, + "learning_rate": 0.00017776624878533754, + "loss": 1.0808, + "step": 4877 + }, + { + "epoch": 0.8685897435897436, + "grad_norm": 0.5088842511177063, + "learning_rate": 0.00017775744806475792, + "loss": 0.995, + "step": 4878 + }, + { + "epoch": 0.8687678062678063, + "grad_norm": 0.5796698927879333, + "learning_rate": 0.00017774864582068264, + "loss": 1.1485, + "step": 4879 + }, + { + "epoch": 0.8689458689458689, + "grad_norm": 0.5719375610351562, + "learning_rate": 0.00017773984205328417, + "loss": 1.0133, + "step": 4880 + }, + { + "epoch": 0.8691239316239316, + "grad_norm": 0.6396418213844299, + "learning_rate": 0.00017773103676273498, + "loss": 1.0932, + "step": 4881 + }, + { + "epoch": 0.8693019943019943, + "grad_norm": 0.5602468252182007, + "learning_rate": 0.00017772222994920763, + "loss": 0.9702, + "step": 4882 + }, + { + "epoch": 0.8694800569800569, + "grad_norm": 0.5167748332023621, + "learning_rate": 0.00017771342161287457, + "loss": 1.0528, + "step": 4883 + }, + { + "epoch": 0.8696581196581197, + "grad_norm": 0.5572916865348816, + "learning_rate": 0.00017770461175390848, + "loss": 1.1341, + "step": 4884 + }, + { + "epoch": 0.8698361823361823, + "grad_norm": 0.6666276454925537, + "learning_rate": 0.00017769580037248195, + "loss": 1.1948, + "step": 4885 + }, + { + "epoch": 0.8700142450142451, + "grad_norm": 0.5348601937294006, + "learning_rate": 0.0001776869874687676, + "loss": 1.0562, + "step": 4886 + }, + { + "epoch": 0.8701923076923077, + "grad_norm": 0.5449648499488831, + "learning_rate": 0.00017767817304293812, + "loss": 0.988, + "step": 4887 + }, + { + "epoch": 0.8703703703703703, + "grad_norm": 0.5995045304298401, + "learning_rate": 0.0001776693570951662, + "loss": 1.2526, + "step": 4888 + }, + { + "epoch": 0.8705484330484331, + "grad_norm": 0.6575320959091187, + "learning_rate": 0.00017766053962562457, + "loss": 1.1717, + "step": 4889 + }, + { + "epoch": 0.8707264957264957, + "grad_norm": 0.5882139801979065, + "learning_rate": 0.00017765172063448597, + "loss": 1.238, + "step": 4890 + }, + { + "epoch": 0.8709045584045584, + "grad_norm": 0.5908389091491699, + "learning_rate": 0.00017764290012192325, + "loss": 1.0606, + "step": 4891 + }, + { + "epoch": 0.8710826210826211, + "grad_norm": 0.6169339418411255, + "learning_rate": 0.00017763407808810917, + "loss": 1.1456, + "step": 4892 + }, + { + "epoch": 0.8712606837606838, + "grad_norm": 0.5916035771369934, + "learning_rate": 0.0001776252545332166, + "loss": 1.0026, + "step": 4893 + }, + { + "epoch": 0.8714387464387464, + "grad_norm": 0.539995551109314, + "learning_rate": 0.00017761642945741843, + "loss": 1.2397, + "step": 4894 + }, + { + "epoch": 0.8716168091168092, + "grad_norm": 0.5346137881278992, + "learning_rate": 0.00017760760286088755, + "loss": 1.1232, + "step": 4895 + }, + { + "epoch": 0.8717948717948718, + "grad_norm": 0.570202112197876, + "learning_rate": 0.00017759877474379692, + "loss": 1.0708, + "step": 4896 + }, + { + "epoch": 0.8719729344729344, + "grad_norm": 0.5023398399353027, + "learning_rate": 0.00017758994510631948, + "loss": 1.1056, + "step": 4897 + }, + { + "epoch": 0.8721509971509972, + "grad_norm": 0.5447137951850891, + "learning_rate": 0.00017758111394862826, + "loss": 0.8776, + "step": 4898 + }, + { + "epoch": 0.8723290598290598, + "grad_norm": 0.5193906426429749, + "learning_rate": 0.00017757228127089625, + "loss": 0.9959, + "step": 4899 + }, + { + "epoch": 0.8725071225071225, + "grad_norm": 0.5958787798881531, + "learning_rate": 0.00017756344707329656, + "loss": 1.092, + "step": 4900 + }, + { + "epoch": 0.8726851851851852, + "grad_norm": 0.521045982837677, + "learning_rate": 0.00017755461135600221, + "loss": 0.9864, + "step": 4901 + }, + { + "epoch": 0.8728632478632479, + "grad_norm": 0.5257635116577148, + "learning_rate": 0.00017754577411918638, + "loss": 1.216, + "step": 4902 + }, + { + "epoch": 0.8730413105413105, + "grad_norm": 0.5425964593887329, + "learning_rate": 0.0001775369353630222, + "loss": 1.1432, + "step": 4903 + }, + { + "epoch": 0.8732193732193733, + "grad_norm": 0.47995322942733765, + "learning_rate": 0.00017752809508768286, + "loss": 1.0227, + "step": 4904 + }, + { + "epoch": 0.8733974358974359, + "grad_norm": 0.5747429728507996, + "learning_rate": 0.0001775192532933415, + "loss": 0.9984, + "step": 4905 + }, + { + "epoch": 0.8735754985754985, + "grad_norm": 0.5745723247528076, + "learning_rate": 0.00017751040998017142, + "loss": 1.2559, + "step": 4906 + }, + { + "epoch": 0.8737535612535613, + "grad_norm": 0.6114141941070557, + "learning_rate": 0.0001775015651483459, + "loss": 1.3224, + "step": 4907 + }, + { + "epoch": 0.8739316239316239, + "grad_norm": 0.4757187068462372, + "learning_rate": 0.00017749271879803817, + "loss": 1.0352, + "step": 4908 + }, + { + "epoch": 0.8741096866096866, + "grad_norm": 0.48644450306892395, + "learning_rate": 0.0001774838709294216, + "loss": 1.0876, + "step": 4909 + }, + { + "epoch": 0.8742877492877493, + "grad_norm": 0.5652037262916565, + "learning_rate": 0.00017747502154266955, + "loss": 0.9189, + "step": 4910 + }, + { + "epoch": 0.874465811965812, + "grad_norm": 0.5289644002914429, + "learning_rate": 0.00017746617063795538, + "loss": 0.9431, + "step": 4911 + }, + { + "epoch": 0.8746438746438746, + "grad_norm": 0.594656229019165, + "learning_rate": 0.00017745731821545253, + "loss": 1.2408, + "step": 4912 + }, + { + "epoch": 0.8748219373219374, + "grad_norm": 0.5693240165710449, + "learning_rate": 0.0001774484642753344, + "loss": 1.347, + "step": 4913 + }, + { + "epoch": 0.875, + "grad_norm": 0.5291008949279785, + "learning_rate": 0.00017743960881777456, + "loss": 1.161, + "step": 4914 + }, + { + "epoch": 0.8751780626780626, + "grad_norm": 0.5958300232887268, + "learning_rate": 0.00017743075184294642, + "loss": 1.2058, + "step": 4915 + }, + { + "epoch": 0.8753561253561254, + "grad_norm": 0.513884425163269, + "learning_rate": 0.00017742189335102354, + "loss": 1.0952, + "step": 4916 + }, + { + "epoch": 0.875534188034188, + "grad_norm": 0.5860681533813477, + "learning_rate": 0.00017741303334217948, + "loss": 1.1801, + "step": 4917 + }, + { + "epoch": 0.8757122507122507, + "grad_norm": 0.47962820529937744, + "learning_rate": 0.00017740417181658788, + "loss": 1.0785, + "step": 4918 + }, + { + "epoch": 0.8758903133903134, + "grad_norm": 0.5110440254211426, + "learning_rate": 0.00017739530877442227, + "loss": 1.1385, + "step": 4919 + }, + { + "epoch": 0.8760683760683761, + "grad_norm": 0.5106285214424133, + "learning_rate": 0.00017738644421585643, + "loss": 1.1204, + "step": 4920 + }, + { + "epoch": 0.8762464387464387, + "grad_norm": 0.5709205865859985, + "learning_rate": 0.00017737757814106393, + "loss": 1.0108, + "step": 4921 + }, + { + "epoch": 0.8764245014245015, + "grad_norm": 0.5850250124931335, + "learning_rate": 0.0001773687105502185, + "loss": 1.0059, + "step": 4922 + }, + { + "epoch": 0.8766025641025641, + "grad_norm": 0.5194727778434753, + "learning_rate": 0.00017735984144349396, + "loss": 0.9466, + "step": 4923 + }, + { + "epoch": 0.8767806267806267, + "grad_norm": 0.5246787667274475, + "learning_rate": 0.000177350970821064, + "loss": 1.1336, + "step": 4924 + }, + { + "epoch": 0.8769586894586895, + "grad_norm": 0.5798323154449463, + "learning_rate": 0.00017734209868310244, + "loss": 1.1641, + "step": 4925 + }, + { + "epoch": 0.8771367521367521, + "grad_norm": 0.5188565850257874, + "learning_rate": 0.00017733322502978314, + "loss": 0.9959, + "step": 4926 + }, + { + "epoch": 0.8773148148148148, + "grad_norm": 0.5969653725624084, + "learning_rate": 0.00017732434986127995, + "loss": 1.2162, + "step": 4927 + }, + { + "epoch": 0.8774928774928775, + "grad_norm": 0.5520089268684387, + "learning_rate": 0.00017731547317776674, + "loss": 1.0163, + "step": 4928 + }, + { + "epoch": 0.8776709401709402, + "grad_norm": 0.48789507150650024, + "learning_rate": 0.00017730659497941745, + "loss": 0.9757, + "step": 4929 + }, + { + "epoch": 0.8778490028490028, + "grad_norm": 0.6034960746765137, + "learning_rate": 0.000177297715266406, + "loss": 1.1278, + "step": 4930 + }, + { + "epoch": 0.8780270655270656, + "grad_norm": 0.53016597032547, + "learning_rate": 0.00017728883403890638, + "loss": 1.0637, + "step": 4931 + }, + { + "epoch": 0.8782051282051282, + "grad_norm": 0.5073726177215576, + "learning_rate": 0.00017727995129709266, + "loss": 1.1491, + "step": 4932 + }, + { + "epoch": 0.8783831908831908, + "grad_norm": 0.540605366230011, + "learning_rate": 0.00017727106704113878, + "loss": 1.0133, + "step": 4933 + }, + { + "epoch": 0.8785612535612536, + "grad_norm": 0.5346775054931641, + "learning_rate": 0.0001772621812712189, + "loss": 1.1781, + "step": 4934 + }, + { + "epoch": 0.8787393162393162, + "grad_norm": 0.5659036040306091, + "learning_rate": 0.00017725329398750702, + "loss": 1.1023, + "step": 4935 + }, + { + "epoch": 0.8789173789173789, + "grad_norm": 0.591063380241394, + "learning_rate": 0.00017724440519017738, + "loss": 1.0298, + "step": 4936 + }, + { + "epoch": 0.8790954415954416, + "grad_norm": 0.5173781514167786, + "learning_rate": 0.0001772355148794041, + "loss": 1.0483, + "step": 4937 + }, + { + "epoch": 0.8792735042735043, + "grad_norm": 0.5405352711677551, + "learning_rate": 0.0001772266230553613, + "loss": 1.0716, + "step": 4938 + }, + { + "epoch": 0.8794515669515669, + "grad_norm": 0.518442690372467, + "learning_rate": 0.00017721772971822323, + "loss": 1.1373, + "step": 4939 + }, + { + "epoch": 0.8796296296296297, + "grad_norm": 0.533673107624054, + "learning_rate": 0.0001772088348681642, + "loss": 1.0489, + "step": 4940 + }, + { + "epoch": 0.8798076923076923, + "grad_norm": 0.46117857098579407, + "learning_rate": 0.0001771999385053584, + "loss": 1.0297, + "step": 4941 + }, + { + "epoch": 0.8799857549857549, + "grad_norm": 0.4687997102737427, + "learning_rate": 0.0001771910406299802, + "loss": 1.071, + "step": 4942 + }, + { + "epoch": 0.8801638176638177, + "grad_norm": 0.5064153075218201, + "learning_rate": 0.0001771821412422039, + "loss": 0.9518, + "step": 4943 + }, + { + "epoch": 0.8803418803418803, + "grad_norm": 0.6561978459358215, + "learning_rate": 0.00017717324034220385, + "loss": 1.11, + "step": 4944 + }, + { + "epoch": 0.8805199430199431, + "grad_norm": 0.5551498532295227, + "learning_rate": 0.00017716433793015454, + "loss": 0.9719, + "step": 4945 + }, + { + "epoch": 0.8806980056980057, + "grad_norm": 0.47059500217437744, + "learning_rate": 0.00017715543400623025, + "loss": 0.8891, + "step": 4946 + }, + { + "epoch": 0.8808760683760684, + "grad_norm": 0.5035740733146667, + "learning_rate": 0.00017714652857060554, + "loss": 0.9671, + "step": 4947 + }, + { + "epoch": 0.8810541310541311, + "grad_norm": 0.4599960446357727, + "learning_rate": 0.00017713762162345487, + "loss": 0.9588, + "step": 4948 + }, + { + "epoch": 0.8812321937321937, + "grad_norm": 0.5087231397628784, + "learning_rate": 0.0001771287131649527, + "loss": 1.1433, + "step": 4949 + }, + { + "epoch": 0.8814102564102564, + "grad_norm": 0.5609854459762573, + "learning_rate": 0.00017711980319527366, + "loss": 1.2022, + "step": 4950 + }, + { + "epoch": 0.8815883190883191, + "grad_norm": 0.49460700154304504, + "learning_rate": 0.00017711089171459227, + "loss": 1.019, + "step": 4951 + }, + { + "epoch": 0.8817663817663818, + "grad_norm": 0.5047259330749512, + "learning_rate": 0.00017710197872308314, + "loss": 0.8301, + "step": 4952 + }, + { + "epoch": 0.8819444444444444, + "grad_norm": 0.5784406065940857, + "learning_rate": 0.0001770930642209209, + "loss": 0.9336, + "step": 4953 + }, + { + "epoch": 0.8821225071225072, + "grad_norm": 0.5037121772766113, + "learning_rate": 0.00017708414820828022, + "loss": 1.0199, + "step": 4954 + }, + { + "epoch": 0.8823005698005698, + "grad_norm": 0.5683804750442505, + "learning_rate": 0.00017707523068533575, + "loss": 0.9758, + "step": 4955 + }, + { + "epoch": 0.8824786324786325, + "grad_norm": 0.5167922973632812, + "learning_rate": 0.0001770663116522623, + "loss": 1.0389, + "step": 4956 + }, + { + "epoch": 0.8826566951566952, + "grad_norm": 0.5813606381416321, + "learning_rate": 0.0001770573911092345, + "loss": 1.3998, + "step": 4957 + }, + { + "epoch": 0.8828347578347578, + "grad_norm": 0.5280475616455078, + "learning_rate": 0.00017704846905642723, + "loss": 1.0545, + "step": 4958 + }, + { + "epoch": 0.8830128205128205, + "grad_norm": 0.5421732068061829, + "learning_rate": 0.00017703954549401528, + "loss": 0.899, + "step": 4959 + }, + { + "epoch": 0.8831908831908832, + "grad_norm": 0.5177720189094543, + "learning_rate": 0.00017703062042217344, + "loss": 0.975, + "step": 4960 + }, + { + "epoch": 0.8833689458689459, + "grad_norm": 0.639327883720398, + "learning_rate": 0.00017702169384107666, + "loss": 1.1936, + "step": 4961 + }, + { + "epoch": 0.8835470085470085, + "grad_norm": 0.5201572179794312, + "learning_rate": 0.00017701276575089975, + "loss": 0.9891, + "step": 4962 + }, + { + "epoch": 0.8837250712250713, + "grad_norm": 0.5304145216941833, + "learning_rate": 0.00017700383615181767, + "loss": 1.0569, + "step": 4963 + }, + { + "epoch": 0.8839031339031339, + "grad_norm": 0.6068132519721985, + "learning_rate": 0.00017699490504400538, + "loss": 1.2653, + "step": 4964 + }, + { + "epoch": 0.8840811965811965, + "grad_norm": 0.597895085811615, + "learning_rate": 0.00017698597242763787, + "loss": 1.2577, + "step": 4965 + }, + { + "epoch": 0.8842592592592593, + "grad_norm": 0.5356902480125427, + "learning_rate": 0.00017697703830289017, + "loss": 1.1056, + "step": 4966 + }, + { + "epoch": 0.8844373219373219, + "grad_norm": 0.5429540872573853, + "learning_rate": 0.0001769681026699373, + "loss": 1.0951, + "step": 4967 + }, + { + "epoch": 0.8846153846153846, + "grad_norm": 0.5789309144020081, + "learning_rate": 0.00017695916552895436, + "loss": 1.0786, + "step": 4968 + }, + { + "epoch": 0.8847934472934473, + "grad_norm": 0.5621341466903687, + "learning_rate": 0.0001769502268801164, + "loss": 1.0645, + "step": 4969 + }, + { + "epoch": 0.88497150997151, + "grad_norm": 0.5879453420639038, + "learning_rate": 0.00017694128672359865, + "loss": 1.2171, + "step": 4970 + }, + { + "epoch": 0.8851495726495726, + "grad_norm": 0.5005951523780823, + "learning_rate": 0.0001769323450595762, + "loss": 1.0725, + "step": 4971 + }, + { + "epoch": 0.8853276353276354, + "grad_norm": 0.5439660549163818, + "learning_rate": 0.00017692340188822425, + "loss": 1.162, + "step": 4972 + }, + { + "epoch": 0.885505698005698, + "grad_norm": 0.6309837698936462, + "learning_rate": 0.00017691445720971802, + "loss": 1.2861, + "step": 4973 + }, + { + "epoch": 0.8856837606837606, + "grad_norm": 0.4997463822364807, + "learning_rate": 0.00017690551102423282, + "loss": 1.1887, + "step": 4974 + }, + { + "epoch": 0.8858618233618234, + "grad_norm": 0.5430852174758911, + "learning_rate": 0.00017689656333194385, + "loss": 1.1231, + "step": 4975 + }, + { + "epoch": 0.886039886039886, + "grad_norm": 0.5414215922355652, + "learning_rate": 0.00017688761413302644, + "loss": 1.2345, + "step": 4976 + }, + { + "epoch": 0.8862179487179487, + "grad_norm": 0.5594443082809448, + "learning_rate": 0.00017687866342765601, + "loss": 1.0775, + "step": 4977 + }, + { + "epoch": 0.8863960113960114, + "grad_norm": 0.5827134847640991, + "learning_rate": 0.00017686971121600787, + "loss": 1.0609, + "step": 4978 + }, + { + "epoch": 0.8865740740740741, + "grad_norm": 0.5075414776802063, + "learning_rate": 0.00017686075749825738, + "loss": 0.796, + "step": 4979 + }, + { + "epoch": 0.8867521367521367, + "grad_norm": 0.6007544994354248, + "learning_rate": 0.00017685180227458003, + "loss": 1.1716, + "step": 4980 + }, + { + "epoch": 0.8869301994301995, + "grad_norm": 0.6458030343055725, + "learning_rate": 0.00017684284554515128, + "loss": 1.1945, + "step": 4981 + }, + { + "epoch": 0.8871082621082621, + "grad_norm": 0.5519212484359741, + "learning_rate": 0.00017683388731014657, + "loss": 1.2571, + "step": 4982 + }, + { + "epoch": 0.8872863247863247, + "grad_norm": 0.5079960227012634, + "learning_rate": 0.00017682492756974146, + "loss": 1.1186, + "step": 4983 + }, + { + "epoch": 0.8874643874643875, + "grad_norm": 0.63576740026474, + "learning_rate": 0.00017681596632411147, + "loss": 1.389, + "step": 4984 + }, + { + "epoch": 0.8876424501424501, + "grad_norm": 0.43325698375701904, + "learning_rate": 0.0001768070035734322, + "loss": 0.7757, + "step": 4985 + }, + { + "epoch": 0.8878205128205128, + "grad_norm": 0.49492064118385315, + "learning_rate": 0.00017679803931787923, + "loss": 1.0096, + "step": 4986 + }, + { + "epoch": 0.8879985754985755, + "grad_norm": 0.5561224222183228, + "learning_rate": 0.00017678907355762825, + "loss": 0.952, + "step": 4987 + }, + { + "epoch": 0.8881766381766382, + "grad_norm": 0.5392457246780396, + "learning_rate": 0.00017678010629285486, + "loss": 1.0442, + "step": 4988 + }, + { + "epoch": 0.8883547008547008, + "grad_norm": 0.4659234881401062, + "learning_rate": 0.00017677113752373482, + "loss": 0.8668, + "step": 4989 + }, + { + "epoch": 0.8885327635327636, + "grad_norm": 0.5139175057411194, + "learning_rate": 0.0001767621672504438, + "loss": 0.8386, + "step": 4990 + }, + { + "epoch": 0.8887108262108262, + "grad_norm": 0.5395823121070862, + "learning_rate": 0.00017675319547315755, + "loss": 0.9754, + "step": 4991 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 0.4751867949962616, + "learning_rate": 0.0001767442221920519, + "loss": 0.8775, + "step": 4992 + }, + { + "epoch": 0.8890669515669516, + "grad_norm": 0.5728281736373901, + "learning_rate": 0.00017673524740730265, + "loss": 1.2807, + "step": 4993 + }, + { + "epoch": 0.8892450142450142, + "grad_norm": 0.5545622110366821, + "learning_rate": 0.00017672627111908558, + "loss": 1.0039, + "step": 4994 + }, + { + "epoch": 0.8894230769230769, + "grad_norm": 0.5127374529838562, + "learning_rate": 0.00017671729332757665, + "loss": 1.0505, + "step": 4995 + }, + { + "epoch": 0.8896011396011396, + "grad_norm": 0.5238714218139648, + "learning_rate": 0.00017670831403295175, + "loss": 1.1775, + "step": 4996 + }, + { + "epoch": 0.8897792022792023, + "grad_norm": 0.5610160827636719, + "learning_rate": 0.00017669933323538674, + "loss": 1.0555, + "step": 4997 + }, + { + "epoch": 0.8899572649572649, + "grad_norm": 0.5481634736061096, + "learning_rate": 0.00017669035093505762, + "loss": 1.0802, + "step": 4998 + }, + { + "epoch": 0.8901353276353277, + "grad_norm": 0.4725174307823181, + "learning_rate": 0.0001766813671321404, + "loss": 0.9611, + "step": 4999 + }, + { + "epoch": 0.8903133903133903, + "grad_norm": 0.5184635519981384, + "learning_rate": 0.0001766723818268111, + "loss": 1.1659, + "step": 5000 + }, + { + "epoch": 0.8904914529914529, + "grad_norm": 0.5503578186035156, + "learning_rate": 0.00017666339501924575, + "loss": 1.2165, + "step": 5001 + }, + { + "epoch": 0.8906695156695157, + "grad_norm": 0.5299594402313232, + "learning_rate": 0.0001766544067096204, + "loss": 1.0196, + "step": 5002 + }, + { + "epoch": 0.8908475783475783, + "grad_norm": 0.5673944354057312, + "learning_rate": 0.00017664541689811118, + "loss": 1.2058, + "step": 5003 + }, + { + "epoch": 0.8910256410256411, + "grad_norm": 0.6057320833206177, + "learning_rate": 0.00017663642558489426, + "loss": 1.0136, + "step": 5004 + }, + { + "epoch": 0.8912037037037037, + "grad_norm": 0.4767026901245117, + "learning_rate": 0.00017662743277014578, + "loss": 0.8522, + "step": 5005 + }, + { + "epoch": 0.8913817663817664, + "grad_norm": 0.5346270203590393, + "learning_rate": 0.00017661843845404192, + "loss": 1.1568, + "step": 5006 + }, + { + "epoch": 0.8915598290598291, + "grad_norm": 0.5365738868713379, + "learning_rate": 0.00017660944263675891, + "loss": 1.0488, + "step": 5007 + }, + { + "epoch": 0.8917378917378918, + "grad_norm": 0.5536269545555115, + "learning_rate": 0.00017660044531847305, + "loss": 1.1216, + "step": 5008 + }, + { + "epoch": 0.8919159544159544, + "grad_norm": 0.6325978636741638, + "learning_rate": 0.00017659144649936055, + "loss": 1.2843, + "step": 5009 + }, + { + "epoch": 0.8920940170940171, + "grad_norm": 0.5890641212463379, + "learning_rate": 0.00017658244617959777, + "loss": 1.1976, + "step": 5010 + }, + { + "epoch": 0.8922720797720798, + "grad_norm": 0.604870080947876, + "learning_rate": 0.00017657344435936107, + "loss": 1.2881, + "step": 5011 + }, + { + "epoch": 0.8924501424501424, + "grad_norm": 0.49805206060409546, + "learning_rate": 0.00017656444103882676, + "loss": 0.8998, + "step": 5012 + }, + { + "epoch": 0.8926282051282052, + "grad_norm": 0.506926953792572, + "learning_rate": 0.0001765554362181713, + "loss": 1.0731, + "step": 5013 + }, + { + "epoch": 0.8928062678062678, + "grad_norm": 0.5353260636329651, + "learning_rate": 0.0001765464298975711, + "loss": 1.0676, + "step": 5014 + }, + { + "epoch": 0.8929843304843305, + "grad_norm": 0.5641853213310242, + "learning_rate": 0.0001765374220772026, + "loss": 0.9606, + "step": 5015 + }, + { + "epoch": 0.8931623931623932, + "grad_norm": 0.5049327611923218, + "learning_rate": 0.00017652841275724233, + "loss": 1.009, + "step": 5016 + }, + { + "epoch": 0.8933404558404558, + "grad_norm": 0.6255155205726624, + "learning_rate": 0.0001765194019378668, + "loss": 1.138, + "step": 5017 + }, + { + "epoch": 0.8935185185185185, + "grad_norm": 0.5816851854324341, + "learning_rate": 0.00017651038961925247, + "loss": 1.3398, + "step": 5018 + }, + { + "epoch": 0.8936965811965812, + "grad_norm": 0.5188020467758179, + "learning_rate": 0.00017650137580157605, + "loss": 1.0126, + "step": 5019 + }, + { + "epoch": 0.8938746438746439, + "grad_norm": 0.5231554508209229, + "learning_rate": 0.00017649236048501406, + "loss": 1.0328, + "step": 5020 + }, + { + "epoch": 0.8940527065527065, + "grad_norm": 0.7638634443283081, + "learning_rate": 0.0001764833436697432, + "loss": 1.3016, + "step": 5021 + }, + { + "epoch": 0.8942307692307693, + "grad_norm": 0.5354094505310059, + "learning_rate": 0.00017647432535594008, + "loss": 1.0646, + "step": 5022 + }, + { + "epoch": 0.8944088319088319, + "grad_norm": 0.6938086748123169, + "learning_rate": 0.0001764653055437814, + "loss": 1.2051, + "step": 5023 + }, + { + "epoch": 0.8945868945868946, + "grad_norm": 0.5546849370002747, + "learning_rate": 0.00017645628423344393, + "loss": 1.0671, + "step": 5024 + }, + { + "epoch": 0.8947649572649573, + "grad_norm": 0.49294665455818176, + "learning_rate": 0.0001764472614251044, + "loss": 1.0328, + "step": 5025 + }, + { + "epoch": 0.89494301994302, + "grad_norm": 0.5965796113014221, + "learning_rate": 0.00017643823711893956, + "loss": 1.0741, + "step": 5026 + }, + { + "epoch": 0.8951210826210826, + "grad_norm": 0.4846448302268982, + "learning_rate": 0.00017642921131512626, + "loss": 1.0409, + "step": 5027 + }, + { + "epoch": 0.8952991452991453, + "grad_norm": 0.5767390131950378, + "learning_rate": 0.00017642018401384135, + "loss": 1.018, + "step": 5028 + }, + { + "epoch": 0.895477207977208, + "grad_norm": 0.503027617931366, + "learning_rate": 0.00017641115521526167, + "loss": 1.0002, + "step": 5029 + }, + { + "epoch": 0.8956552706552706, + "grad_norm": 0.6668619513511658, + "learning_rate": 0.00017640212491956412, + "loss": 1.2154, + "step": 5030 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 0.5544148683547974, + "learning_rate": 0.00017639309312692566, + "loss": 1.2701, + "step": 5031 + }, + { + "epoch": 0.896011396011396, + "grad_norm": 0.6026872992515564, + "learning_rate": 0.00017638405983752323, + "loss": 0.9335, + "step": 5032 + }, + { + "epoch": 0.8961894586894587, + "grad_norm": 0.6288694143295288, + "learning_rate": 0.00017637502505153384, + "loss": 0.9075, + "step": 5033 + }, + { + "epoch": 0.8963675213675214, + "grad_norm": 0.4890204966068268, + "learning_rate": 0.00017636598876913446, + "loss": 0.8492, + "step": 5034 + }, + { + "epoch": 0.896545584045584, + "grad_norm": 0.5746598243713379, + "learning_rate": 0.00017635695099050218, + "loss": 1.1557, + "step": 5035 + }, + { + "epoch": 0.8967236467236467, + "grad_norm": 0.5165683031082153, + "learning_rate": 0.00017634791171581405, + "loss": 1.0899, + "step": 5036 + }, + { + "epoch": 0.8969017094017094, + "grad_norm": 0.4621037244796753, + "learning_rate": 0.0001763388709452472, + "loss": 1.0457, + "step": 5037 + }, + { + "epoch": 0.8970797720797721, + "grad_norm": 0.532358705997467, + "learning_rate": 0.00017632982867897876, + "loss": 1.139, + "step": 5038 + }, + { + "epoch": 0.8972578347578347, + "grad_norm": 0.5794399976730347, + "learning_rate": 0.00017632078491718587, + "loss": 1.031, + "step": 5039 + }, + { + "epoch": 0.8974358974358975, + "grad_norm": 0.5031905174255371, + "learning_rate": 0.00017631173966004576, + "loss": 0.9508, + "step": 5040 + }, + { + "epoch": 0.8976139601139601, + "grad_norm": 0.6528840065002441, + "learning_rate": 0.00017630269290773564, + "loss": 0.9974, + "step": 5041 + }, + { + "epoch": 0.8977920227920227, + "grad_norm": 0.6007558703422546, + "learning_rate": 0.00017629364466043273, + "loss": 1.0993, + "step": 5042 + }, + { + "epoch": 0.8979700854700855, + "grad_norm": 0.5104095339775085, + "learning_rate": 0.00017628459491831437, + "loss": 0.9175, + "step": 5043 + }, + { + "epoch": 0.8981481481481481, + "grad_norm": 0.5285516977310181, + "learning_rate": 0.00017627554368155782, + "loss": 0.998, + "step": 5044 + }, + { + "epoch": 0.8983262108262108, + "grad_norm": 0.5629046559333801, + "learning_rate": 0.00017626649095034045, + "loss": 1.2021, + "step": 5045 + }, + { + "epoch": 0.8985042735042735, + "grad_norm": 0.57548987865448, + "learning_rate": 0.00017625743672483962, + "loss": 1.2076, + "step": 5046 + }, + { + "epoch": 0.8986823361823362, + "grad_norm": 0.4883024990558624, + "learning_rate": 0.0001762483810052327, + "loss": 0.9761, + "step": 5047 + }, + { + "epoch": 0.8988603988603988, + "grad_norm": 0.6378034949302673, + "learning_rate": 0.0001762393237916972, + "loss": 1.2266, + "step": 5048 + }, + { + "epoch": 0.8990384615384616, + "grad_norm": 0.5201624035835266, + "learning_rate": 0.0001762302650844105, + "loss": 1.247, + "step": 5049 + }, + { + "epoch": 0.8992165242165242, + "grad_norm": 0.5438048243522644, + "learning_rate": 0.0001762212048835501, + "loss": 0.993, + "step": 5050 + }, + { + "epoch": 0.8993945868945868, + "grad_norm": 0.5928253531455994, + "learning_rate": 0.00017621214318929354, + "loss": 1.0469, + "step": 5051 + }, + { + "epoch": 0.8995726495726496, + "grad_norm": 0.6437996625900269, + "learning_rate": 0.00017620308000181831, + "loss": 1.3136, + "step": 5052 + }, + { + "epoch": 0.8997507122507122, + "grad_norm": 0.5961456298828125, + "learning_rate": 0.00017619401532130208, + "loss": 1.1495, + "step": 5053 + }, + { + "epoch": 0.8999287749287749, + "grad_norm": 0.497388631105423, + "learning_rate": 0.0001761849491479224, + "loss": 0.7783, + "step": 5054 + }, + { + "epoch": 0.9001068376068376, + "grad_norm": 0.5984451174736023, + "learning_rate": 0.00017617588148185687, + "loss": 1.3115, + "step": 5055 + }, + { + "epoch": 0.9002849002849003, + "grad_norm": 0.549163818359375, + "learning_rate": 0.0001761668123232832, + "loss": 1.1649, + "step": 5056 + }, + { + "epoch": 0.9004629629629629, + "grad_norm": 0.5831968188285828, + "learning_rate": 0.00017615774167237903, + "loss": 1.1749, + "step": 5057 + }, + { + "epoch": 0.9006410256410257, + "grad_norm": 0.5111076235771179, + "learning_rate": 0.00017614866952932214, + "loss": 0.8936, + "step": 5058 + }, + { + "epoch": 0.9008190883190883, + "grad_norm": 0.5740947723388672, + "learning_rate": 0.00017613959589429028, + "loss": 1.2606, + "step": 5059 + }, + { + "epoch": 0.9009971509971509, + "grad_norm": 0.5881099700927734, + "learning_rate": 0.0001761305207674612, + "loss": 1.3682, + "step": 5060 + }, + { + "epoch": 0.9011752136752137, + "grad_norm": 0.5007091760635376, + "learning_rate": 0.00017612144414901268, + "loss": 0.7788, + "step": 5061 + }, + { + "epoch": 0.9013532763532763, + "grad_norm": 0.5127760171890259, + "learning_rate": 0.00017611236603912262, + "loss": 1.0519, + "step": 5062 + }, + { + "epoch": 0.9015313390313391, + "grad_norm": 0.6185184121131897, + "learning_rate": 0.00017610328643796882, + "loss": 1.1672, + "step": 5063 + }, + { + "epoch": 0.9017094017094017, + "grad_norm": 0.49707287549972534, + "learning_rate": 0.00017609420534572926, + "loss": 1.1865, + "step": 5064 + }, + { + "epoch": 0.9018874643874644, + "grad_norm": 0.5667552351951599, + "learning_rate": 0.0001760851227625818, + "loss": 1.1388, + "step": 5065 + }, + { + "epoch": 0.9020655270655271, + "grad_norm": 0.50298011302948, + "learning_rate": 0.00017607603868870442, + "loss": 0.9552, + "step": 5066 + }, + { + "epoch": 0.9022435897435898, + "grad_norm": 0.5709219574928284, + "learning_rate": 0.0001760669531242751, + "loss": 1.2636, + "step": 5067 + }, + { + "epoch": 0.9024216524216524, + "grad_norm": 0.4943496286869049, + "learning_rate": 0.0001760578660694718, + "loss": 0.8951, + "step": 5068 + }, + { + "epoch": 0.9025997150997151, + "grad_norm": 0.5475931167602539, + "learning_rate": 0.00017604877752447267, + "loss": 1.1442, + "step": 5069 + }, + { + "epoch": 0.9027777777777778, + "grad_norm": 0.5280239582061768, + "learning_rate": 0.0001760396874894557, + "loss": 0.9537, + "step": 5070 + }, + { + "epoch": 0.9029558404558404, + "grad_norm": 0.5480797290802002, + "learning_rate": 0.000176030595964599, + "loss": 1.1557, + "step": 5071 + }, + { + "epoch": 0.9031339031339032, + "grad_norm": 0.5232734680175781, + "learning_rate": 0.00017602150295008073, + "loss": 1.0219, + "step": 5072 + }, + { + "epoch": 0.9033119658119658, + "grad_norm": 0.5448359251022339, + "learning_rate": 0.000176012408446079, + "loss": 1.1964, + "step": 5073 + }, + { + "epoch": 0.9034900284900285, + "grad_norm": 0.4841914474964142, + "learning_rate": 0.00017600331245277206, + "loss": 1.0667, + "step": 5074 + }, + { + "epoch": 0.9036680911680912, + "grad_norm": 0.5407083630561829, + "learning_rate": 0.0001759942149703381, + "loss": 1.1895, + "step": 5075 + }, + { + "epoch": 0.9038461538461539, + "grad_norm": 0.5140416026115417, + "learning_rate": 0.00017598511599895534, + "loss": 0.9402, + "step": 5076 + }, + { + "epoch": 0.9040242165242165, + "grad_norm": 0.6333765983581543, + "learning_rate": 0.00017597601553880207, + "loss": 1.239, + "step": 5077 + }, + { + "epoch": 0.9042022792022792, + "grad_norm": 0.4996028244495392, + "learning_rate": 0.00017596691359005664, + "loss": 1.0259, + "step": 5078 + }, + { + "epoch": 0.9043803418803419, + "grad_norm": 0.591892421245575, + "learning_rate": 0.00017595781015289732, + "loss": 1.2148, + "step": 5079 + }, + { + "epoch": 0.9045584045584045, + "grad_norm": 0.736499011516571, + "learning_rate": 0.0001759487052275025, + "loss": 1.1373, + "step": 5080 + }, + { + "epoch": 0.9047364672364673, + "grad_norm": 0.5951572060585022, + "learning_rate": 0.00017593959881405057, + "loss": 1.1833, + "step": 5081 + }, + { + "epoch": 0.9049145299145299, + "grad_norm": 0.5092006325721741, + "learning_rate": 0.00017593049091271996, + "loss": 0.8841, + "step": 5082 + }, + { + "epoch": 0.9050925925925926, + "grad_norm": 0.5679013729095459, + "learning_rate": 0.0001759213815236891, + "loss": 1.1056, + "step": 5083 + }, + { + "epoch": 0.9052706552706553, + "grad_norm": 0.5708174109458923, + "learning_rate": 0.0001759122706471365, + "loss": 1.1952, + "step": 5084 + }, + { + "epoch": 0.905448717948718, + "grad_norm": 0.5726733803749084, + "learning_rate": 0.00017590315828324067, + "loss": 1.1013, + "step": 5085 + }, + { + "epoch": 0.9056267806267806, + "grad_norm": 0.5821273326873779, + "learning_rate": 0.00017589404443218008, + "loss": 1.2323, + "step": 5086 + }, + { + "epoch": 0.9058048433048433, + "grad_norm": 0.5811445713043213, + "learning_rate": 0.00017588492909413337, + "loss": 1.2241, + "step": 5087 + }, + { + "epoch": 0.905982905982906, + "grad_norm": 0.5377545952796936, + "learning_rate": 0.0001758758122692791, + "loss": 0.9777, + "step": 5088 + }, + { + "epoch": 0.9061609686609686, + "grad_norm": 0.5985640287399292, + "learning_rate": 0.0001758666939577959, + "loss": 0.9737, + "step": 5089 + }, + { + "epoch": 0.9063390313390314, + "grad_norm": 0.6038222908973694, + "learning_rate": 0.00017585757415986247, + "loss": 1.2116, + "step": 5090 + }, + { + "epoch": 0.906517094017094, + "grad_norm": 0.6752246022224426, + "learning_rate": 0.00017584845287565743, + "loss": 1.1975, + "step": 5091 + }, + { + "epoch": 0.9066951566951567, + "grad_norm": 0.5400625467300415, + "learning_rate": 0.0001758393301053595, + "loss": 0.9669, + "step": 5092 + }, + { + "epoch": 0.9068732193732194, + "grad_norm": 0.5637784004211426, + "learning_rate": 0.00017583020584914746, + "loss": 1.2672, + "step": 5093 + }, + { + "epoch": 0.907051282051282, + "grad_norm": 0.4825877249240875, + "learning_rate": 0.00017582108010720006, + "loss": 0.9719, + "step": 5094 + }, + { + "epoch": 0.9072293447293447, + "grad_norm": 0.49902790784835815, + "learning_rate": 0.00017581195287969613, + "loss": 0.7941, + "step": 5095 + }, + { + "epoch": 0.9074074074074074, + "grad_norm": 0.5991541743278503, + "learning_rate": 0.0001758028241668144, + "loss": 1.049, + "step": 5096 + }, + { + "epoch": 0.9075854700854701, + "grad_norm": 0.5788859724998474, + "learning_rate": 0.00017579369396873384, + "loss": 1.0318, + "step": 5097 + }, + { + "epoch": 0.9077635327635327, + "grad_norm": 0.5914160013198853, + "learning_rate": 0.0001757845622856333, + "loss": 1.1007, + "step": 5098 + }, + { + "epoch": 0.9079415954415955, + "grad_norm": 0.5361711382865906, + "learning_rate": 0.00017577542911769166, + "loss": 1.0694, + "step": 5099 + }, + { + "epoch": 0.9081196581196581, + "grad_norm": 0.5752849578857422, + "learning_rate": 0.00017576629446508792, + "loss": 1.1184, + "step": 5100 + }, + { + "epoch": 0.9082977207977208, + "grad_norm": 0.6042249798774719, + "learning_rate": 0.000175757158328001, + "loss": 1.2808, + "step": 5101 + }, + { + "epoch": 0.9084757834757835, + "grad_norm": 0.508352518081665, + "learning_rate": 0.00017574802070661, + "loss": 1.0038, + "step": 5102 + }, + { + "epoch": 0.9086538461538461, + "grad_norm": 0.5667358040809631, + "learning_rate": 0.00017573888160109385, + "loss": 1.0208, + "step": 5103 + }, + { + "epoch": 0.9088319088319088, + "grad_norm": 0.653619647026062, + "learning_rate": 0.00017572974101163165, + "loss": 1.2053, + "step": 5104 + }, + { + "epoch": 0.9090099715099715, + "grad_norm": 0.5069597363471985, + "learning_rate": 0.00017572059893840246, + "loss": 0.8634, + "step": 5105 + }, + { + "epoch": 0.9091880341880342, + "grad_norm": 0.6160602569580078, + "learning_rate": 0.00017571145538158547, + "loss": 1.2626, + "step": 5106 + }, + { + "epoch": 0.9093660968660968, + "grad_norm": 0.6335833668708801, + "learning_rate": 0.00017570231034135978, + "loss": 1.3381, + "step": 5107 + }, + { + "epoch": 0.9095441595441596, + "grad_norm": 0.5140398740768433, + "learning_rate": 0.00017569316381790454, + "loss": 1.1258, + "step": 5108 + }, + { + "epoch": 0.9097222222222222, + "grad_norm": 0.5682975649833679, + "learning_rate": 0.00017568401581139905, + "loss": 1.3367, + "step": 5109 + }, + { + "epoch": 0.9099002849002849, + "grad_norm": 0.49765729904174805, + "learning_rate": 0.00017567486632202246, + "loss": 1.1891, + "step": 5110 + }, + { + "epoch": 0.9100783475783476, + "grad_norm": 0.5139224529266357, + "learning_rate": 0.00017566571534995406, + "loss": 0.9768, + "step": 5111 + }, + { + "epoch": 0.9102564102564102, + "grad_norm": 0.5510922074317932, + "learning_rate": 0.00017565656289537316, + "loss": 1.1552, + "step": 5112 + }, + { + "epoch": 0.9104344729344729, + "grad_norm": 0.6243364810943604, + "learning_rate": 0.00017564740895845908, + "loss": 1.1341, + "step": 5113 + }, + { + "epoch": 0.9106125356125356, + "grad_norm": 0.5334977507591248, + "learning_rate": 0.00017563825353939116, + "loss": 1.0894, + "step": 5114 + }, + { + "epoch": 0.9107905982905983, + "grad_norm": 0.5195826292037964, + "learning_rate": 0.00017562909663834878, + "loss": 1.1011, + "step": 5115 + }, + { + "epoch": 0.9109686609686609, + "grad_norm": 0.5298168063163757, + "learning_rate": 0.00017561993825551138, + "loss": 1.0079, + "step": 5116 + }, + { + "epoch": 0.9111467236467237, + "grad_norm": 0.5858965516090393, + "learning_rate": 0.00017561077839105835, + "loss": 1.2746, + "step": 5117 + }, + { + "epoch": 0.9113247863247863, + "grad_norm": 0.5572476387023926, + "learning_rate": 0.0001756016170451692, + "loss": 0.8169, + "step": 5118 + }, + { + "epoch": 0.9115028490028491, + "grad_norm": 0.5247095823287964, + "learning_rate": 0.0001755924542180234, + "loss": 1.1206, + "step": 5119 + }, + { + "epoch": 0.9116809116809117, + "grad_norm": 0.5605118274688721, + "learning_rate": 0.0001755832899098005, + "loss": 1.371, + "step": 5120 + }, + { + "epoch": 0.9118589743589743, + "grad_norm": 0.5732316970825195, + "learning_rate": 0.00017557412412068005, + "loss": 1.1248, + "step": 5121 + }, + { + "epoch": 0.9120370370370371, + "grad_norm": 0.6167279481887817, + "learning_rate": 0.0001755649568508416, + "loss": 0.94, + "step": 5122 + }, + { + "epoch": 0.9122150997150997, + "grad_norm": 0.5497499108314514, + "learning_rate": 0.00017555578810046483, + "loss": 1.0112, + "step": 5123 + }, + { + "epoch": 0.9123931623931624, + "grad_norm": 0.540762186050415, + "learning_rate": 0.00017554661786972931, + "loss": 1.1058, + "step": 5124 + }, + { + "epoch": 0.9125712250712251, + "grad_norm": 0.5943556427955627, + "learning_rate": 0.0001755374461588148, + "loss": 0.9086, + "step": 5125 + }, + { + "epoch": 0.9127492877492878, + "grad_norm": 0.5300756692886353, + "learning_rate": 0.0001755282729679009, + "loss": 1.1566, + "step": 5126 + }, + { + "epoch": 0.9129273504273504, + "grad_norm": 0.5390434861183167, + "learning_rate": 0.00017551909829716743, + "loss": 1.1395, + "step": 5127 + }, + { + "epoch": 0.9131054131054132, + "grad_norm": 0.627434492111206, + "learning_rate": 0.00017550992214679405, + "loss": 1.1537, + "step": 5128 + }, + { + "epoch": 0.9132834757834758, + "grad_norm": 0.4806903302669525, + "learning_rate": 0.00017550074451696063, + "loss": 0.7905, + "step": 5129 + }, + { + "epoch": 0.9134615384615384, + "grad_norm": 0.5714817047119141, + "learning_rate": 0.00017549156540784696, + "loss": 1.1042, + "step": 5130 + }, + { + "epoch": 0.9136396011396012, + "grad_norm": 0.5839236378669739, + "learning_rate": 0.0001754823848196329, + "loss": 1.0383, + "step": 5131 + }, + { + "epoch": 0.9138176638176638, + "grad_norm": 0.6089872717857361, + "learning_rate": 0.0001754732027524983, + "loss": 0.9399, + "step": 5132 + }, + { + "epoch": 0.9139957264957265, + "grad_norm": 0.4937956631183624, + "learning_rate": 0.00017546401920662307, + "loss": 0.7382, + "step": 5133 + }, + { + "epoch": 0.9141737891737892, + "grad_norm": 0.5918676257133484, + "learning_rate": 0.00017545483418218716, + "loss": 1.2207, + "step": 5134 + }, + { + "epoch": 0.9143518518518519, + "grad_norm": 0.5825346112251282, + "learning_rate": 0.0001754456476793705, + "loss": 0.9669, + "step": 5135 + }, + { + "epoch": 0.9145299145299145, + "grad_norm": 0.49829617142677307, + "learning_rate": 0.0001754364596983531, + "loss": 1.2247, + "step": 5136 + }, + { + "epoch": 0.9147079772079773, + "grad_norm": 0.5128271579742432, + "learning_rate": 0.00017542727023931497, + "loss": 0.9563, + "step": 5137 + }, + { + "epoch": 0.9148860398860399, + "grad_norm": 0.5789414644241333, + "learning_rate": 0.00017541807930243622, + "loss": 1.22, + "step": 5138 + }, + { + "epoch": 0.9150641025641025, + "grad_norm": 0.44155433773994446, + "learning_rate": 0.00017540888688789683, + "loss": 0.9897, + "step": 5139 + }, + { + "epoch": 0.9152421652421653, + "grad_norm": 0.550464391708374, + "learning_rate": 0.00017539969299587696, + "loss": 1.0624, + "step": 5140 + }, + { + "epoch": 0.9154202279202279, + "grad_norm": 0.5019831657409668, + "learning_rate": 0.0001753904976265567, + "loss": 0.9045, + "step": 5141 + }, + { + "epoch": 0.9155982905982906, + "grad_norm": 0.589658796787262, + "learning_rate": 0.0001753813007801163, + "loss": 1.0454, + "step": 5142 + }, + { + "epoch": 0.9157763532763533, + "grad_norm": 0.5945459008216858, + "learning_rate": 0.00017537210245673586, + "loss": 1.0042, + "step": 5143 + }, + { + "epoch": 0.915954415954416, + "grad_norm": 0.5409809947013855, + "learning_rate": 0.00017536290265659566, + "loss": 1.0609, + "step": 5144 + }, + { + "epoch": 0.9161324786324786, + "grad_norm": 0.5302975177764893, + "learning_rate": 0.00017535370137987597, + "loss": 1.1394, + "step": 5145 + }, + { + "epoch": 0.9163105413105413, + "grad_norm": 0.5253351330757141, + "learning_rate": 0.00017534449862675698, + "loss": 1.2249, + "step": 5146 + }, + { + "epoch": 0.916488603988604, + "grad_norm": 0.6363829970359802, + "learning_rate": 0.00017533529439741908, + "loss": 1.1333, + "step": 5147 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.4703354835510254, + "learning_rate": 0.0001753260886920426, + "loss": 0.9971, + "step": 5148 + }, + { + "epoch": 0.9168447293447294, + "grad_norm": 0.6394907236099243, + "learning_rate": 0.00017531688151080786, + "loss": 1.5942, + "step": 5149 + }, + { + "epoch": 0.917022792022792, + "grad_norm": 0.5573459267616272, + "learning_rate": 0.00017530767285389527, + "loss": 0.9669, + "step": 5150 + }, + { + "epoch": 0.9172008547008547, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00017529846272148532, + "loss": 1.2151, + "step": 5151 + }, + { + "epoch": 0.9173789173789174, + "grad_norm": 0.5550395846366882, + "learning_rate": 0.0001752892511137584, + "loss": 1.1765, + "step": 5152 + }, + { + "epoch": 0.91755698005698, + "grad_norm": 0.5461394786834717, + "learning_rate": 0.00017528003803089496, + "loss": 1.1136, + "step": 5153 + }, + { + "epoch": 0.9177350427350427, + "grad_norm": 0.5512672662734985, + "learning_rate": 0.00017527082347307558, + "loss": 1.1727, + "step": 5154 + }, + { + "epoch": 0.9179131054131054, + "grad_norm": 0.5210778713226318, + "learning_rate": 0.0001752616074404808, + "loss": 1.09, + "step": 5155 + }, + { + "epoch": 0.9180911680911681, + "grad_norm": 0.5214943289756775, + "learning_rate": 0.00017525238993329115, + "loss": 0.9654, + "step": 5156 + }, + { + "epoch": 0.9182692307692307, + "grad_norm": 0.5822862386703491, + "learning_rate": 0.00017524317095168724, + "loss": 1.0951, + "step": 5157 + }, + { + "epoch": 0.9184472934472935, + "grad_norm": 0.43948012590408325, + "learning_rate": 0.0001752339504958497, + "loss": 0.6984, + "step": 5158 + }, + { + "epoch": 0.9186253561253561, + "grad_norm": 0.5024449229240417, + "learning_rate": 0.00017522472856595916, + "loss": 0.983, + "step": 5159 + }, + { + "epoch": 0.9188034188034188, + "grad_norm": 0.5815144181251526, + "learning_rate": 0.00017521550516219636, + "loss": 0.9784, + "step": 5160 + }, + { + "epoch": 0.9189814814814815, + "grad_norm": 0.5519825220108032, + "learning_rate": 0.00017520628028474197, + "loss": 1.064, + "step": 5161 + }, + { + "epoch": 0.9191595441595442, + "grad_norm": 0.5615749955177307, + "learning_rate": 0.00017519705393377675, + "loss": 1.1284, + "step": 5162 + }, + { + "epoch": 0.9193376068376068, + "grad_norm": 0.5929917693138123, + "learning_rate": 0.00017518782610948148, + "loss": 1.1221, + "step": 5163 + }, + { + "epoch": 0.9195156695156695, + "grad_norm": 0.7116361856460571, + "learning_rate": 0.00017517859681203692, + "loss": 1.0188, + "step": 5164 + }, + { + "epoch": 0.9196937321937322, + "grad_norm": 0.5095893740653992, + "learning_rate": 0.00017516936604162396, + "loss": 1.0724, + "step": 5165 + }, + { + "epoch": 0.9198717948717948, + "grad_norm": 0.5701385736465454, + "learning_rate": 0.00017516013379842337, + "loss": 1.0572, + "step": 5166 + }, + { + "epoch": 0.9200498575498576, + "grad_norm": 0.518412709236145, + "learning_rate": 0.00017515090008261613, + "loss": 1.0514, + "step": 5167 + }, + { + "epoch": 0.9202279202279202, + "grad_norm": 0.5324261784553528, + "learning_rate": 0.00017514166489438312, + "loss": 1.1708, + "step": 5168 + }, + { + "epoch": 0.9204059829059829, + "grad_norm": 0.5640990138053894, + "learning_rate": 0.00017513242823390525, + "loss": 1.2846, + "step": 5169 + }, + { + "epoch": 0.9205840455840456, + "grad_norm": 0.510352373123169, + "learning_rate": 0.00017512319010136356, + "loss": 1.0763, + "step": 5170 + }, + { + "epoch": 0.9207621082621082, + "grad_norm": 0.4994175136089325, + "learning_rate": 0.00017511395049693898, + "loss": 0.9665, + "step": 5171 + }, + { + "epoch": 0.9209401709401709, + "grad_norm": 0.43196994066238403, + "learning_rate": 0.00017510470942081258, + "loss": 0.761, + "step": 5172 + }, + { + "epoch": 0.9211182336182336, + "grad_norm": 0.558977484703064, + "learning_rate": 0.00017509546687316543, + "loss": 1.0758, + "step": 5173 + }, + { + "epoch": 0.9212962962962963, + "grad_norm": 0.573302149772644, + "learning_rate": 0.0001750862228541786, + "loss": 0.9635, + "step": 5174 + }, + { + "epoch": 0.9214743589743589, + "grad_norm": 0.5083786845207214, + "learning_rate": 0.00017507697736403321, + "loss": 1.0311, + "step": 5175 + }, + { + "epoch": 0.9216524216524217, + "grad_norm": 0.5478954911231995, + "learning_rate": 0.00017506773040291043, + "loss": 1.074, + "step": 5176 + }, + { + "epoch": 0.9218304843304843, + "grad_norm": 0.522376537322998, + "learning_rate": 0.00017505848197099137, + "loss": 1.1162, + "step": 5177 + }, + { + "epoch": 0.9220085470085471, + "grad_norm": 0.5946292281150818, + "learning_rate": 0.0001750492320684573, + "loss": 0.9494, + "step": 5178 + }, + { + "epoch": 0.9221866096866097, + "grad_norm": 0.5423247814178467, + "learning_rate": 0.00017503998069548943, + "loss": 1.0558, + "step": 5179 + }, + { + "epoch": 0.9223646723646723, + "grad_norm": 0.49960651993751526, + "learning_rate": 0.000175030727852269, + "loss": 1.0748, + "step": 5180 + }, + { + "epoch": 0.9225427350427351, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00017502147353897732, + "loss": 1.2066, + "step": 5181 + }, + { + "epoch": 0.9227207977207977, + "grad_norm": 0.57244473695755, + "learning_rate": 0.00017501221775579576, + "loss": 1.048, + "step": 5182 + }, + { + "epoch": 0.9228988603988604, + "grad_norm": 0.512464165687561, + "learning_rate": 0.00017500296050290557, + "loss": 1.1405, + "step": 5183 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.5380734801292419, + "learning_rate": 0.00017499370178048818, + "loss": 1.0641, + "step": 5184 + }, + { + "epoch": 0.9232549857549858, + "grad_norm": 0.47102874517440796, + "learning_rate": 0.000174984441588725, + "loss": 0.7948, + "step": 5185 + }, + { + "epoch": 0.9234330484330484, + "grad_norm": 0.6702211499214172, + "learning_rate": 0.00017497517992779747, + "loss": 1.3009, + "step": 5186 + }, + { + "epoch": 0.9236111111111112, + "grad_norm": 0.4685834050178528, + "learning_rate": 0.000174965916797887, + "loss": 0.8136, + "step": 5187 + }, + { + "epoch": 0.9237891737891738, + "grad_norm": 0.5414277911186218, + "learning_rate": 0.00017495665219917513, + "loss": 0.9708, + "step": 5188 + }, + { + "epoch": 0.9239672364672364, + "grad_norm": 0.5253050923347473, + "learning_rate": 0.0001749473861318434, + "loss": 1.0691, + "step": 5189 + }, + { + "epoch": 0.9241452991452992, + "grad_norm": 0.6009906530380249, + "learning_rate": 0.00017493811859607328, + "loss": 1.2023, + "step": 5190 + }, + { + "epoch": 0.9243233618233618, + "grad_norm": 0.5519336462020874, + "learning_rate": 0.00017492884959204643, + "loss": 1.189, + "step": 5191 + }, + { + "epoch": 0.9245014245014245, + "grad_norm": 0.5024857521057129, + "learning_rate": 0.0001749195791199444, + "loss": 0.8685, + "step": 5192 + }, + { + "epoch": 0.9246794871794872, + "grad_norm": 0.5735679864883423, + "learning_rate": 0.00017491030717994887, + "loss": 1.1903, + "step": 5193 + }, + { + "epoch": 0.9248575498575499, + "grad_norm": 0.5338658094406128, + "learning_rate": 0.00017490103377224147, + "loss": 1.0442, + "step": 5194 + }, + { + "epoch": 0.9250356125356125, + "grad_norm": 0.46669119596481323, + "learning_rate": 0.0001748917588970039, + "loss": 0.6343, + "step": 5195 + }, + { + "epoch": 0.9252136752136753, + "grad_norm": 0.510910153388977, + "learning_rate": 0.00017488248255441793, + "loss": 0.9334, + "step": 5196 + }, + { + "epoch": 0.9253917378917379, + "grad_norm": 0.5732216238975525, + "learning_rate": 0.00017487320474466524, + "loss": 1.0483, + "step": 5197 + }, + { + "epoch": 0.9255698005698005, + "grad_norm": 0.5864318609237671, + "learning_rate": 0.00017486392546792762, + "loss": 1.0669, + "step": 5198 + }, + { + "epoch": 0.9257478632478633, + "grad_norm": 0.5074281096458435, + "learning_rate": 0.00017485464472438692, + "loss": 1.0636, + "step": 5199 + }, + { + "epoch": 0.9259259259259259, + "grad_norm": 0.5833215117454529, + "learning_rate": 0.00017484536251422496, + "loss": 1.2005, + "step": 5200 + }, + { + "epoch": 0.9261039886039886, + "grad_norm": 0.5624990463256836, + "learning_rate": 0.0001748360788376236, + "loss": 1.1623, + "step": 5201 + }, + { + "epoch": 0.9262820512820513, + "grad_norm": 0.5618230104446411, + "learning_rate": 0.00017482679369476472, + "loss": 1.0495, + "step": 5202 + }, + { + "epoch": 0.926460113960114, + "grad_norm": 0.6254985332489014, + "learning_rate": 0.00017481750708583024, + "loss": 0.9521, + "step": 5203 + }, + { + "epoch": 0.9266381766381766, + "grad_norm": 0.5488203763961792, + "learning_rate": 0.00017480821901100216, + "loss": 1.0689, + "step": 5204 + }, + { + "epoch": 0.9268162393162394, + "grad_norm": 0.6157993674278259, + "learning_rate": 0.00017479892947046245, + "loss": 1.2852, + "step": 5205 + }, + { + "epoch": 0.926994301994302, + "grad_norm": 0.49653390049934387, + "learning_rate": 0.00017478963846439305, + "loss": 0.8616, + "step": 5206 + }, + { + "epoch": 0.9271723646723646, + "grad_norm": 0.5079081058502197, + "learning_rate": 0.00017478034599297603, + "loss": 1.0192, + "step": 5207 + }, + { + "epoch": 0.9273504273504274, + "grad_norm": 0.5392495393753052, + "learning_rate": 0.00017477105205639354, + "loss": 1.115, + "step": 5208 + }, + { + "epoch": 0.92752849002849, + "grad_norm": 0.5336191654205322, + "learning_rate": 0.00017476175665482756, + "loss": 1.1892, + "step": 5209 + }, + { + "epoch": 0.9277065527065527, + "grad_norm": 0.631712019443512, + "learning_rate": 0.00017475245978846026, + "loss": 0.9619, + "step": 5210 + }, + { + "epoch": 0.9278846153846154, + "grad_norm": 0.5123951435089111, + "learning_rate": 0.0001747431614574738, + "loss": 1.1477, + "step": 5211 + }, + { + "epoch": 0.9280626780626781, + "grad_norm": 0.5045743584632874, + "learning_rate": 0.00017473386166205038, + "loss": 0.9749, + "step": 5212 + }, + { + "epoch": 0.9282407407407407, + "grad_norm": 0.5296525359153748, + "learning_rate": 0.00017472456040237217, + "loss": 1.0736, + "step": 5213 + }, + { + "epoch": 0.9284188034188035, + "grad_norm": 0.6304933428764343, + "learning_rate": 0.00017471525767862145, + "loss": 1.2444, + "step": 5214 + }, + { + "epoch": 0.9285968660968661, + "grad_norm": 0.4851958155632019, + "learning_rate": 0.00017470595349098044, + "loss": 0.9049, + "step": 5215 + }, + { + "epoch": 0.9287749287749287, + "grad_norm": 0.5730679631233215, + "learning_rate": 0.00017469664783963148, + "loss": 1.0773, + "step": 5216 + }, + { + "epoch": 0.9289529914529915, + "grad_norm": 0.6020415425300598, + "learning_rate": 0.00017468734072475684, + "loss": 1.3247, + "step": 5217 + }, + { + "epoch": 0.9291310541310541, + "grad_norm": 0.47981077432632446, + "learning_rate": 0.00017467803214653893, + "loss": 1.0009, + "step": 5218 + }, + { + "epoch": 0.9293091168091168, + "grad_norm": 0.5787527561187744, + "learning_rate": 0.0001746687221051601, + "loss": 1.2523, + "step": 5219 + }, + { + "epoch": 0.9294871794871795, + "grad_norm": 0.4495891332626343, + "learning_rate": 0.00017465941060080278, + "loss": 0.7364, + "step": 5220 + }, + { + "epoch": 0.9296652421652422, + "grad_norm": 0.5721768140792847, + "learning_rate": 0.0001746500976336494, + "loss": 1.015, + "step": 5221 + }, + { + "epoch": 0.9298433048433048, + "grad_norm": 0.5500208735466003, + "learning_rate": 0.0001746407832038824, + "loss": 1.053, + "step": 5222 + }, + { + "epoch": 0.9300213675213675, + "grad_norm": 0.5784386992454529, + "learning_rate": 0.00017463146731168437, + "loss": 0.9784, + "step": 5223 + }, + { + "epoch": 0.9301994301994302, + "grad_norm": 0.4960322082042694, + "learning_rate": 0.00017462214995723772, + "loss": 0.8674, + "step": 5224 + }, + { + "epoch": 0.9303774928774928, + "grad_norm": 0.5005537271499634, + "learning_rate": 0.00017461283114072508, + "loss": 1.0486, + "step": 5225 + }, + { + "epoch": 0.9305555555555556, + "grad_norm": 0.5064167380332947, + "learning_rate": 0.000174603510862329, + "loss": 0.9722, + "step": 5226 + }, + { + "epoch": 0.9307336182336182, + "grad_norm": 0.583558976650238, + "learning_rate": 0.0001745941891222321, + "loss": 0.9957, + "step": 5227 + }, + { + "epoch": 0.9309116809116809, + "grad_norm": 0.4982515871524811, + "learning_rate": 0.00017458486592061704, + "loss": 0.958, + "step": 5228 + }, + { + "epoch": 0.9310897435897436, + "grad_norm": 0.526549756526947, + "learning_rate": 0.0001745755412576664, + "loss": 1.1172, + "step": 5229 + }, + { + "epoch": 0.9312678062678063, + "grad_norm": 0.6129719018936157, + "learning_rate": 0.000174566215133563, + "loss": 1.2524, + "step": 5230 + }, + { + "epoch": 0.9314458689458689, + "grad_norm": 0.5385653972625732, + "learning_rate": 0.00017455688754848948, + "loss": 1.1655, + "step": 5231 + }, + { + "epoch": 0.9316239316239316, + "grad_norm": 0.5646410584449768, + "learning_rate": 0.0001745475585026287, + "loss": 0.9026, + "step": 5232 + }, + { + "epoch": 0.9318019943019943, + "grad_norm": 0.549223780632019, + "learning_rate": 0.0001745382279961633, + "loss": 0.804, + "step": 5233 + }, + { + "epoch": 0.9319800569800569, + "grad_norm": 0.48547953367233276, + "learning_rate": 0.0001745288960292762, + "loss": 1.0224, + "step": 5234 + }, + { + "epoch": 0.9321581196581197, + "grad_norm": 0.5260967016220093, + "learning_rate": 0.00017451956260215016, + "loss": 0.9688, + "step": 5235 + }, + { + "epoch": 0.9323361823361823, + "grad_norm": 0.6261999011039734, + "learning_rate": 0.00017451022771496812, + "loss": 1.2539, + "step": 5236 + }, + { + "epoch": 0.9325142450142451, + "grad_norm": 0.5801421999931335, + "learning_rate": 0.00017450089136791298, + "loss": 1.11, + "step": 5237 + }, + { + "epoch": 0.9326923076923077, + "grad_norm": 0.5833573937416077, + "learning_rate": 0.0001744915535611676, + "loss": 0.9328, + "step": 5238 + }, + { + "epoch": 0.9328703703703703, + "grad_norm": 0.5422634482383728, + "learning_rate": 0.00017448221429491496, + "loss": 1.034, + "step": 5239 + }, + { + "epoch": 0.9330484330484331, + "grad_norm": 0.5105658769607544, + "learning_rate": 0.00017447287356933808, + "loss": 0.8924, + "step": 5240 + }, + { + "epoch": 0.9332264957264957, + "grad_norm": 0.5114831924438477, + "learning_rate": 0.00017446353138461995, + "loss": 0.9328, + "step": 5241 + }, + { + "epoch": 0.9334045584045584, + "grad_norm": 0.5105039477348328, + "learning_rate": 0.00017445418774094358, + "loss": 1.0468, + "step": 5242 + }, + { + "epoch": 0.9335826210826211, + "grad_norm": 0.593250036239624, + "learning_rate": 0.00017444484263849208, + "loss": 1.0603, + "step": 5243 + }, + { + "epoch": 0.9337606837606838, + "grad_norm": 0.600788414478302, + "learning_rate": 0.00017443549607744853, + "loss": 1.1506, + "step": 5244 + }, + { + "epoch": 0.9339387464387464, + "grad_norm": 0.5394418239593506, + "learning_rate": 0.00017442614805799605, + "loss": 1.038, + "step": 5245 + }, + { + "epoch": 0.9341168091168092, + "grad_norm": 0.5446375608444214, + "learning_rate": 0.00017441679858031786, + "loss": 1.079, + "step": 5246 + }, + { + "epoch": 0.9342948717948718, + "grad_norm": 0.5859794616699219, + "learning_rate": 0.00017440744764459702, + "loss": 1.1453, + "step": 5247 + }, + { + "epoch": 0.9344729344729344, + "grad_norm": 0.4899081289768219, + "learning_rate": 0.00017439809525101688, + "loss": 1.163, + "step": 5248 + }, + { + "epoch": 0.9346509971509972, + "grad_norm": 0.652846097946167, + "learning_rate": 0.00017438874139976055, + "loss": 1.1819, + "step": 5249 + }, + { + "epoch": 0.9348290598290598, + "grad_norm": 0.5402514934539795, + "learning_rate": 0.00017437938609101138, + "loss": 1.0159, + "step": 5250 + }, + { + "epoch": 0.9350071225071225, + "grad_norm": 0.565864086151123, + "learning_rate": 0.00017437002932495265, + "loss": 1.1121, + "step": 5251 + }, + { + "epoch": 0.9351851851851852, + "grad_norm": 0.611786425113678, + "learning_rate": 0.0001743606711017677, + "loss": 1.2511, + "step": 5252 + }, + { + "epoch": 0.9353632478632479, + "grad_norm": 0.5706882476806641, + "learning_rate": 0.00017435131142163988, + "loss": 1.128, + "step": 5253 + }, + { + "epoch": 0.9355413105413105, + "grad_norm": 0.5369367003440857, + "learning_rate": 0.00017434195028475253, + "loss": 1.0562, + "step": 5254 + }, + { + "epoch": 0.9357193732193733, + "grad_norm": 0.49957552552223206, + "learning_rate": 0.0001743325876912891, + "loss": 1.0568, + "step": 5255 + }, + { + "epoch": 0.9358974358974359, + "grad_norm": 0.5398106575012207, + "learning_rate": 0.00017432322364143305, + "loss": 1.1502, + "step": 5256 + }, + { + "epoch": 0.9360754985754985, + "grad_norm": 0.6522027254104614, + "learning_rate": 0.00017431385813536783, + "loss": 1.0591, + "step": 5257 + }, + { + "epoch": 0.9362535612535613, + "grad_norm": 0.5872012972831726, + "learning_rate": 0.00017430449117327693, + "loss": 1.3737, + "step": 5258 + }, + { + "epoch": 0.9364316239316239, + "grad_norm": 0.5124474167823792, + "learning_rate": 0.00017429512275534382, + "loss": 1.0727, + "step": 5259 + }, + { + "epoch": 0.9366096866096866, + "grad_norm": 0.5103365778923035, + "learning_rate": 0.00017428575288175218, + "loss": 1.0339, + "step": 5260 + }, + { + "epoch": 0.9367877492877493, + "grad_norm": 0.585483729839325, + "learning_rate": 0.0001742763815526855, + "loss": 1.1844, + "step": 5261 + }, + { + "epoch": 0.936965811965812, + "grad_norm": 0.5855562090873718, + "learning_rate": 0.00017426700876832746, + "loss": 1.3234, + "step": 5262 + }, + { + "epoch": 0.9371438746438746, + "grad_norm": 0.5774588584899902, + "learning_rate": 0.00017425763452886162, + "loss": 1.0937, + "step": 5263 + }, + { + "epoch": 0.9373219373219374, + "grad_norm": 0.5718343257904053, + "learning_rate": 0.00017424825883447168, + "loss": 1.0783, + "step": 5264 + }, + { + "epoch": 0.9375, + "grad_norm": 0.5414558053016663, + "learning_rate": 0.00017423888168534136, + "loss": 1.1244, + "step": 5265 + }, + { + "epoch": 0.9376780626780626, + "grad_norm": 0.5818275809288025, + "learning_rate": 0.00017422950308165438, + "loss": 1.247, + "step": 5266 + }, + { + "epoch": 0.9378561253561254, + "grad_norm": 0.586398184299469, + "learning_rate": 0.00017422012302359448, + "loss": 1.0515, + "step": 5267 + }, + { + "epoch": 0.938034188034188, + "grad_norm": 0.5236606001853943, + "learning_rate": 0.00017421074151134544, + "loss": 1.1907, + "step": 5268 + }, + { + "epoch": 0.9382122507122507, + "grad_norm": 0.5108010172843933, + "learning_rate": 0.0001742013585450911, + "loss": 1.1125, + "step": 5269 + }, + { + "epoch": 0.9383903133903134, + "grad_norm": 0.4956454038619995, + "learning_rate": 0.00017419197412501527, + "loss": 1.0305, + "step": 5270 + }, + { + "epoch": 0.9385683760683761, + "grad_norm": 0.5432302951812744, + "learning_rate": 0.0001741825882513018, + "loss": 1.1946, + "step": 5271 + }, + { + "epoch": 0.9387464387464387, + "grad_norm": 0.5119295716285706, + "learning_rate": 0.00017417320092413463, + "loss": 0.875, + "step": 5272 + }, + { + "epoch": 0.9389245014245015, + "grad_norm": 0.49740248918533325, + "learning_rate": 0.0001741638121436977, + "loss": 1.1093, + "step": 5273 + }, + { + "epoch": 0.9391025641025641, + "grad_norm": 0.5069027543067932, + "learning_rate": 0.00017415442191017491, + "loss": 1.2498, + "step": 5274 + }, + { + "epoch": 0.9392806267806267, + "grad_norm": 0.570264995098114, + "learning_rate": 0.00017414503022375027, + "loss": 1.0192, + "step": 5275 + }, + { + "epoch": 0.9394586894586895, + "grad_norm": 0.48129352927207947, + "learning_rate": 0.00017413563708460776, + "loss": 0.8467, + "step": 5276 + }, + { + "epoch": 0.9396367521367521, + "grad_norm": 0.5214534401893616, + "learning_rate": 0.00017412624249293148, + "loss": 0.9723, + "step": 5277 + }, + { + "epoch": 0.9398148148148148, + "grad_norm": 0.5150161385536194, + "learning_rate": 0.00017411684644890544, + "loss": 1.0906, + "step": 5278 + }, + { + "epoch": 0.9399928774928775, + "grad_norm": 0.5695852637290955, + "learning_rate": 0.00017410744895271377, + "loss": 1.2891, + "step": 5279 + }, + { + "epoch": 0.9401709401709402, + "grad_norm": 0.5613594651222229, + "learning_rate": 0.00017409805000454055, + "loss": 1.1373, + "step": 5280 + }, + { + "epoch": 0.9403490028490028, + "grad_norm": 0.5134239196777344, + "learning_rate": 0.00017408864960457004, + "loss": 1.1081, + "step": 5281 + }, + { + "epoch": 0.9405270655270656, + "grad_norm": 0.5256397724151611, + "learning_rate": 0.00017407924775298628, + "loss": 1.058, + "step": 5282 + }, + { + "epoch": 0.9407051282051282, + "grad_norm": 0.5145402550697327, + "learning_rate": 0.00017406984444997357, + "loss": 1.0667, + "step": 5283 + }, + { + "epoch": 0.9408831908831908, + "grad_norm": 0.5435704588890076, + "learning_rate": 0.0001740604396957161, + "loss": 1.2275, + "step": 5284 + }, + { + "epoch": 0.9410612535612536, + "grad_norm": 0.5798762440681458, + "learning_rate": 0.0001740510334903982, + "loss": 1.2061, + "step": 5285 + }, + { + "epoch": 0.9412393162393162, + "grad_norm": 0.5461057424545288, + "learning_rate": 0.00017404162583420414, + "loss": 1.1585, + "step": 5286 + }, + { + "epoch": 0.9414173789173789, + "grad_norm": 0.5090487003326416, + "learning_rate": 0.00017403221672731818, + "loss": 1.2496, + "step": 5287 + }, + { + "epoch": 0.9415954415954416, + "grad_norm": 0.5171035528182983, + "learning_rate": 0.00017402280616992476, + "loss": 1.1947, + "step": 5288 + }, + { + "epoch": 0.9417735042735043, + "grad_norm": 0.5292364358901978, + "learning_rate": 0.00017401339416220818, + "loss": 1.0182, + "step": 5289 + }, + { + "epoch": 0.9419515669515669, + "grad_norm": 0.5011499524116516, + "learning_rate": 0.00017400398070435293, + "loss": 1.3363, + "step": 5290 + }, + { + "epoch": 0.9421296296296297, + "grad_norm": 0.4821554720401764, + "learning_rate": 0.0001739945657965434, + "loss": 0.9077, + "step": 5291 + }, + { + "epoch": 0.9423076923076923, + "grad_norm": 0.5849515199661255, + "learning_rate": 0.00017398514943896403, + "loss": 1.1582, + "step": 5292 + }, + { + "epoch": 0.9424857549857549, + "grad_norm": 0.49826139211654663, + "learning_rate": 0.00017397573163179937, + "loss": 1.1025, + "step": 5293 + }, + { + "epoch": 0.9426638176638177, + "grad_norm": 0.6031842827796936, + "learning_rate": 0.00017396631237523392, + "loss": 1.1932, + "step": 5294 + }, + { + "epoch": 0.9428418803418803, + "grad_norm": 0.6013330221176147, + "learning_rate": 0.00017395689166945224, + "loss": 1.2078, + "step": 5295 + }, + { + "epoch": 0.9430199430199431, + "grad_norm": 0.5147021412849426, + "learning_rate": 0.00017394746951463893, + "loss": 0.9988, + "step": 5296 + }, + { + "epoch": 0.9431980056980057, + "grad_norm": 0.5721762776374817, + "learning_rate": 0.0001739380459109785, + "loss": 1.1442, + "step": 5297 + }, + { + "epoch": 0.9433760683760684, + "grad_norm": 0.49272531270980835, + "learning_rate": 0.0001739286208586557, + "loss": 1.0481, + "step": 5298 + }, + { + "epoch": 0.9435541310541311, + "grad_norm": 0.6545688509941101, + "learning_rate": 0.00017391919435785514, + "loss": 1.1393, + "step": 5299 + }, + { + "epoch": 0.9437321937321937, + "grad_norm": 0.617756724357605, + "learning_rate": 0.00017390976640876152, + "loss": 1.1108, + "step": 5300 + }, + { + "epoch": 0.9439102564102564, + "grad_norm": 0.4870470464229584, + "learning_rate": 0.00017390033701155955, + "loss": 0.9028, + "step": 5301 + }, + { + "epoch": 0.9440883190883191, + "grad_norm": 0.5250138640403748, + "learning_rate": 0.000173890906166434, + "loss": 1.0326, + "step": 5302 + }, + { + "epoch": 0.9442663817663818, + "grad_norm": 0.5879467129707336, + "learning_rate": 0.00017388147387356964, + "loss": 1.1569, + "step": 5303 + }, + { + "epoch": 0.9444444444444444, + "grad_norm": 0.4790486991405487, + "learning_rate": 0.00017387204013315127, + "loss": 0.967, + "step": 5304 + }, + { + "epoch": 0.9446225071225072, + "grad_norm": 0.5884372591972351, + "learning_rate": 0.0001738626049453637, + "loss": 1.1342, + "step": 5305 + }, + { + "epoch": 0.9448005698005698, + "grad_norm": 0.4633975028991699, + "learning_rate": 0.00017385316831039187, + "loss": 0.8942, + "step": 5306 + }, + { + "epoch": 0.9449786324786325, + "grad_norm": 0.5301823019981384, + "learning_rate": 0.0001738437302284206, + "loss": 1.1683, + "step": 5307 + }, + { + "epoch": 0.9451566951566952, + "grad_norm": 0.5476770997047424, + "learning_rate": 0.00017383429069963484, + "loss": 1.1574, + "step": 5308 + }, + { + "epoch": 0.9453347578347578, + "grad_norm": 0.47689101099967957, + "learning_rate": 0.00017382484972421953, + "loss": 1.0792, + "step": 5309 + }, + { + "epoch": 0.9455128205128205, + "grad_norm": 0.526063084602356, + "learning_rate": 0.00017381540730235963, + "loss": 0.9012, + "step": 5310 + }, + { + "epoch": 0.9456908831908832, + "grad_norm": 0.5667058229446411, + "learning_rate": 0.0001738059634342402, + "loss": 1.0908, + "step": 5311 + }, + { + "epoch": 0.9458689458689459, + "grad_norm": 0.5402196645736694, + "learning_rate": 0.00017379651812004623, + "loss": 0.943, + "step": 5312 + }, + { + "epoch": 0.9460470085470085, + "grad_norm": 0.5288932919502258, + "learning_rate": 0.00017378707135996276, + "loss": 1.0055, + "step": 5313 + }, + { + "epoch": 0.9462250712250713, + "grad_norm": 0.5607456564903259, + "learning_rate": 0.00017377762315417492, + "loss": 1.2073, + "step": 5314 + }, + { + "epoch": 0.9464031339031339, + "grad_norm": 0.5737698674201965, + "learning_rate": 0.00017376817350286781, + "loss": 1.0001, + "step": 5315 + }, + { + "epoch": 0.9465811965811965, + "grad_norm": 0.6562079787254333, + "learning_rate": 0.00017375872240622657, + "loss": 1.1503, + "step": 5316 + }, + { + "epoch": 0.9467592592592593, + "grad_norm": 0.5407183170318604, + "learning_rate": 0.0001737492698644364, + "loss": 1.1169, + "step": 5317 + }, + { + "epoch": 0.9469373219373219, + "grad_norm": 0.5504152178764343, + "learning_rate": 0.00017373981587768248, + "loss": 1.0468, + "step": 5318 + }, + { + "epoch": 0.9471153846153846, + "grad_norm": 0.4813530743122101, + "learning_rate": 0.00017373036044615006, + "loss": 0.9707, + "step": 5319 + }, + { + "epoch": 0.9472934472934473, + "grad_norm": 0.5810509920120239, + "learning_rate": 0.00017372090357002437, + "loss": 1.4949, + "step": 5320 + }, + { + "epoch": 0.94747150997151, + "grad_norm": 0.5250222086906433, + "learning_rate": 0.00017371144524949074, + "loss": 1.0818, + "step": 5321 + }, + { + "epoch": 0.9476495726495726, + "grad_norm": 0.4852280914783478, + "learning_rate": 0.00017370198548473444, + "loss": 1.1793, + "step": 5322 + }, + { + "epoch": 0.9478276353276354, + "grad_norm": 0.5392420291900635, + "learning_rate": 0.00017369252427594086, + "loss": 1.153, + "step": 5323 + }, + { + "epoch": 0.948005698005698, + "grad_norm": 0.521294116973877, + "learning_rate": 0.00017368306162329533, + "loss": 0.8572, + "step": 5324 + }, + { + "epoch": 0.9481837606837606, + "grad_norm": 0.5579673647880554, + "learning_rate": 0.0001736735975269833, + "loss": 1.0452, + "step": 5325 + }, + { + "epoch": 0.9483618233618234, + "grad_norm": 0.6027318835258484, + "learning_rate": 0.0001736641319871901, + "loss": 1.3475, + "step": 5326 + }, + { + "epoch": 0.948539886039886, + "grad_norm": 0.5600738525390625, + "learning_rate": 0.00017365466500410132, + "loss": 1.0338, + "step": 5327 + }, + { + "epoch": 0.9487179487179487, + "grad_norm": 0.5691532492637634, + "learning_rate": 0.00017364519657790236, + "loss": 1.129, + "step": 5328 + }, + { + "epoch": 0.9488960113960114, + "grad_norm": 0.5161463022232056, + "learning_rate": 0.0001736357267087788, + "loss": 1.0438, + "step": 5329 + }, + { + "epoch": 0.9490740740740741, + "grad_norm": 0.5049656629562378, + "learning_rate": 0.0001736262553969161, + "loss": 0.9484, + "step": 5330 + }, + { + "epoch": 0.9492521367521367, + "grad_norm": 0.5477150678634644, + "learning_rate": 0.00017361678264249988, + "loss": 0.8995, + "step": 5331 + }, + { + "epoch": 0.9494301994301995, + "grad_norm": 0.5679608583450317, + "learning_rate": 0.0001736073084457157, + "loss": 1.241, + "step": 5332 + }, + { + "epoch": 0.9496082621082621, + "grad_norm": 0.5748196840286255, + "learning_rate": 0.00017359783280674926, + "loss": 1.0046, + "step": 5333 + }, + { + "epoch": 0.9497863247863247, + "grad_norm": 0.5677094459533691, + "learning_rate": 0.00017358835572578617, + "loss": 1.2913, + "step": 5334 + }, + { + "epoch": 0.9499643874643875, + "grad_norm": 0.49663659930229187, + "learning_rate": 0.0001735788772030121, + "loss": 1.0388, + "step": 5335 + }, + { + "epoch": 0.9501424501424501, + "grad_norm": 0.5687218904495239, + "learning_rate": 0.0001735693972386128, + "loss": 1.1631, + "step": 5336 + }, + { + "epoch": 0.9503205128205128, + "grad_norm": 0.520708441734314, + "learning_rate": 0.00017355991583277395, + "loss": 1.0744, + "step": 5337 + }, + { + "epoch": 0.9504985754985755, + "grad_norm": 0.5738952159881592, + "learning_rate": 0.00017355043298568137, + "loss": 1.318, + "step": 5338 + }, + { + "epoch": 0.9506766381766382, + "grad_norm": 0.5378455519676208, + "learning_rate": 0.00017354094869752085, + "loss": 0.9827, + "step": 5339 + }, + { + "epoch": 0.9508547008547008, + "grad_norm": 0.5047366619110107, + "learning_rate": 0.0001735314629684782, + "loss": 1.0966, + "step": 5340 + }, + { + "epoch": 0.9510327635327636, + "grad_norm": 0.5526043772697449, + "learning_rate": 0.0001735219757987393, + "loss": 1.059, + "step": 5341 + }, + { + "epoch": 0.9512108262108262, + "grad_norm": 0.5741400718688965, + "learning_rate": 0.00017351248718849003, + "loss": 1.1232, + "step": 5342 + }, + { + "epoch": 0.9513888888888888, + "grad_norm": 0.5421118140220642, + "learning_rate": 0.00017350299713791626, + "loss": 1.0427, + "step": 5343 + }, + { + "epoch": 0.9515669515669516, + "grad_norm": 0.4857081472873688, + "learning_rate": 0.00017349350564720392, + "loss": 0.8663, + "step": 5344 + }, + { + "epoch": 0.9517450142450142, + "grad_norm": 0.5411618947982788, + "learning_rate": 0.00017348401271653904, + "loss": 1.0317, + "step": 5345 + }, + { + "epoch": 0.9519230769230769, + "grad_norm": 0.5246246457099915, + "learning_rate": 0.00017347451834610756, + "loss": 1.0076, + "step": 5346 + }, + { + "epoch": 0.9521011396011396, + "grad_norm": 0.5278927683830261, + "learning_rate": 0.00017346502253609556, + "loss": 0.931, + "step": 5347 + }, + { + "epoch": 0.9522792022792023, + "grad_norm": 0.5934548377990723, + "learning_rate": 0.00017345552528668902, + "loss": 1.3205, + "step": 5348 + }, + { + "epoch": 0.9524572649572649, + "grad_norm": 0.5466100573539734, + "learning_rate": 0.00017344602659807406, + "loss": 0.8725, + "step": 5349 + }, + { + "epoch": 0.9526353276353277, + "grad_norm": 0.5220118761062622, + "learning_rate": 0.00017343652647043678, + "loss": 1.1642, + "step": 5350 + }, + { + "epoch": 0.9528133903133903, + "grad_norm": 0.6166301965713501, + "learning_rate": 0.0001734270249039633, + "loss": 0.8152, + "step": 5351 + }, + { + "epoch": 0.9529914529914529, + "grad_norm": 0.5173428058624268, + "learning_rate": 0.00017341752189883983, + "loss": 0.9296, + "step": 5352 + }, + { + "epoch": 0.9531695156695157, + "grad_norm": 0.5363461375236511, + "learning_rate": 0.0001734080174552525, + "loss": 1.3546, + "step": 5353 + }, + { + "epoch": 0.9533475783475783, + "grad_norm": 0.5333831906318665, + "learning_rate": 0.0001733985115733876, + "loss": 1.0401, + "step": 5354 + }, + { + "epoch": 0.9535256410256411, + "grad_norm": 0.5179334878921509, + "learning_rate": 0.00017338900425343132, + "loss": 1.1254, + "step": 5355 + }, + { + "epoch": 0.9537037037037037, + "grad_norm": 0.5171303153038025, + "learning_rate": 0.00017337949549556993, + "loss": 1.0518, + "step": 5356 + }, + { + "epoch": 0.9538817663817664, + "grad_norm": 0.5164596438407898, + "learning_rate": 0.00017336998529998978, + "loss": 0.8732, + "step": 5357 + }, + { + "epoch": 0.9540598290598291, + "grad_norm": 0.5555717349052429, + "learning_rate": 0.00017336047366687719, + "loss": 1.2312, + "step": 5358 + }, + { + "epoch": 0.9542378917378918, + "grad_norm": 0.45685622096061707, + "learning_rate": 0.00017335096059641847, + "loss": 0.8882, + "step": 5359 + }, + { + "epoch": 0.9544159544159544, + "grad_norm": 0.5260133743286133, + "learning_rate": 0.0001733414460888001, + "loss": 1.0952, + "step": 5360 + }, + { + "epoch": 0.9545940170940171, + "grad_norm": 0.4597703814506531, + "learning_rate": 0.0001733319301442084, + "loss": 1.0835, + "step": 5361 + }, + { + "epoch": 0.9547720797720798, + "grad_norm": 0.5279495120048523, + "learning_rate": 0.0001733224127628299, + "loss": 1.0295, + "step": 5362 + }, + { + "epoch": 0.9549501424501424, + "grad_norm": 0.48919400572776794, + "learning_rate": 0.00017331289394485104, + "loss": 0.9693, + "step": 5363 + }, + { + "epoch": 0.9551282051282052, + "grad_norm": 0.5639515519142151, + "learning_rate": 0.0001733033736904583, + "loss": 1.0893, + "step": 5364 + }, + { + "epoch": 0.9553062678062678, + "grad_norm": 0.49761319160461426, + "learning_rate": 0.00017329385199983823, + "loss": 1.038, + "step": 5365 + }, + { + "epoch": 0.9554843304843305, + "grad_norm": 0.5503305792808533, + "learning_rate": 0.0001732843288731774, + "loss": 0.9976, + "step": 5366 + }, + { + "epoch": 0.9556623931623932, + "grad_norm": 0.5633028745651245, + "learning_rate": 0.00017327480431066235, + "loss": 1.0602, + "step": 5367 + }, + { + "epoch": 0.9558404558404558, + "grad_norm": 0.48074454069137573, + "learning_rate": 0.00017326527831247973, + "loss": 1.0286, + "step": 5368 + }, + { + "epoch": 0.9560185185185185, + "grad_norm": 0.506597638130188, + "learning_rate": 0.0001732557508788162, + "loss": 0.9061, + "step": 5369 + }, + { + "epoch": 0.9561965811965812, + "grad_norm": 0.6570749282836914, + "learning_rate": 0.0001732462220098584, + "loss": 1.0852, + "step": 5370 + }, + { + "epoch": 0.9563746438746439, + "grad_norm": 0.5607653856277466, + "learning_rate": 0.00017323669170579302, + "loss": 1.0486, + "step": 5371 + }, + { + "epoch": 0.9565527065527065, + "grad_norm": 0.6047050356864929, + "learning_rate": 0.0001732271599668068, + "loss": 1.2175, + "step": 5372 + }, + { + "epoch": 0.9567307692307693, + "grad_norm": 0.5506869554519653, + "learning_rate": 0.00017321762679308651, + "loss": 1.0114, + "step": 5373 + }, + { + "epoch": 0.9569088319088319, + "grad_norm": 0.5868638157844543, + "learning_rate": 0.00017320809218481891, + "loss": 1.2983, + "step": 5374 + }, + { + "epoch": 0.9570868945868946, + "grad_norm": 0.539619505405426, + "learning_rate": 0.00017319855614219084, + "loss": 1.2361, + "step": 5375 + }, + { + "epoch": 0.9572649572649573, + "grad_norm": 0.5525495409965515, + "learning_rate": 0.0001731890186653891, + "loss": 1.1316, + "step": 5376 + }, + { + "epoch": 0.95744301994302, + "grad_norm": 0.5549767017364502, + "learning_rate": 0.0001731794797546006, + "loss": 1.0547, + "step": 5377 + }, + { + "epoch": 0.9576210826210826, + "grad_norm": 0.5356076955795288, + "learning_rate": 0.00017316993941001222, + "loss": 0.9942, + "step": 5378 + }, + { + "epoch": 0.9577991452991453, + "grad_norm": 0.5365784168243408, + "learning_rate": 0.00017316039763181084, + "loss": 1.226, + "step": 5379 + }, + { + "epoch": 0.957977207977208, + "grad_norm": 0.5190927386283875, + "learning_rate": 0.00017315085442018343, + "loss": 1.1704, + "step": 5380 + }, + { + "epoch": 0.9581552706552706, + "grad_norm": 0.526658833026886, + "learning_rate": 0.00017314130977531705, + "loss": 1.109, + "step": 5381 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 0.5373684763908386, + "learning_rate": 0.0001731317636973986, + "loss": 1.0018, + "step": 5382 + }, + { + "epoch": 0.958511396011396, + "grad_norm": 0.5714904069900513, + "learning_rate": 0.00017312221618661516, + "loss": 1.1855, + "step": 5383 + }, + { + "epoch": 0.9586894586894587, + "grad_norm": 0.5707863569259644, + "learning_rate": 0.00017311266724315377, + "loss": 0.9482, + "step": 5384 + }, + { + "epoch": 0.9588675213675214, + "grad_norm": 0.5856872797012329, + "learning_rate": 0.00017310311686720157, + "loss": 0.9543, + "step": 5385 + }, + { + "epoch": 0.959045584045584, + "grad_norm": 0.5041963458061218, + "learning_rate": 0.00017309356505894568, + "loss": 1.1427, + "step": 5386 + }, + { + "epoch": 0.9592236467236467, + "grad_norm": 0.5409179925918579, + "learning_rate": 0.00017308401181857316, + "loss": 0.8432, + "step": 5387 + }, + { + "epoch": 0.9594017094017094, + "grad_norm": 0.5248702764511108, + "learning_rate": 0.00017307445714627128, + "loss": 1.1403, + "step": 5388 + }, + { + "epoch": 0.9595797720797721, + "grad_norm": 0.50718092918396, + "learning_rate": 0.00017306490104222722, + "loss": 0.9066, + "step": 5389 + }, + { + "epoch": 0.9597578347578347, + "grad_norm": 0.5563821196556091, + "learning_rate": 0.0001730553435066282, + "loss": 1.0204, + "step": 5390 + }, + { + "epoch": 0.9599358974358975, + "grad_norm": 0.5696987509727478, + "learning_rate": 0.00017304578453966146, + "loss": 1.1405, + "step": 5391 + }, + { + "epoch": 0.9601139601139601, + "grad_norm": 0.5927395224571228, + "learning_rate": 0.00017303622414151435, + "loss": 1.0398, + "step": 5392 + }, + { + "epoch": 0.9602920227920227, + "grad_norm": 0.5375707745552063, + "learning_rate": 0.0001730266623123741, + "loss": 0.9519, + "step": 5393 + }, + { + "epoch": 0.9604700854700855, + "grad_norm": 0.457998126745224, + "learning_rate": 0.00017301709905242815, + "loss": 0.8743, + "step": 5394 + }, + { + "epoch": 0.9606481481481481, + "grad_norm": 0.5427796244621277, + "learning_rate": 0.00017300753436186382, + "loss": 1.078, + "step": 5395 + }, + { + "epoch": 0.9608262108262108, + "grad_norm": 0.5458595752716064, + "learning_rate": 0.0001729979682408685, + "loss": 1.1081, + "step": 5396 + }, + { + "epoch": 0.9610042735042735, + "grad_norm": 0.5495280027389526, + "learning_rate": 0.00017298840068962962, + "loss": 1.0141, + "step": 5397 + }, + { + "epoch": 0.9611823361823362, + "grad_norm": 0.5878560543060303, + "learning_rate": 0.00017297883170833465, + "loss": 1.302, + "step": 5398 + }, + { + "epoch": 0.9613603988603988, + "grad_norm": 0.5452881455421448, + "learning_rate": 0.00017296926129717108, + "loss": 0.9929, + "step": 5399 + }, + { + "epoch": 0.9615384615384616, + "grad_norm": 0.6021811366081238, + "learning_rate": 0.0001729596894563264, + "loss": 1.2629, + "step": 5400 + }, + { + "epoch": 0.9617165242165242, + "grad_norm": 0.5820204615592957, + "learning_rate": 0.0001729501161859882, + "loss": 1.0662, + "step": 5401 + }, + { + "epoch": 0.9618945868945868, + "grad_norm": 0.4953218102455139, + "learning_rate": 0.000172940541486344, + "loss": 1.047, + "step": 5402 + }, + { + "epoch": 0.9620726495726496, + "grad_norm": 0.5409793853759766, + "learning_rate": 0.00017293096535758143, + "loss": 1.1993, + "step": 5403 + }, + { + "epoch": 0.9622507122507122, + "grad_norm": 0.49702873826026917, + "learning_rate": 0.00017292138779988805, + "loss": 1.2471, + "step": 5404 + }, + { + "epoch": 0.9624287749287749, + "grad_norm": 0.5743489861488342, + "learning_rate": 0.00017291180881345158, + "loss": 1.0816, + "step": 5405 + }, + { + "epoch": 0.9626068376068376, + "grad_norm": 0.5747945308685303, + "learning_rate": 0.00017290222839845968, + "loss": 1.3548, + "step": 5406 + }, + { + "epoch": 0.9627849002849003, + "grad_norm": 0.5341345071792603, + "learning_rate": 0.00017289264655510005, + "loss": 1.0435, + "step": 5407 + }, + { + "epoch": 0.9629629629629629, + "grad_norm": 0.5719689130783081, + "learning_rate": 0.00017288306328356044, + "loss": 1.2319, + "step": 5408 + }, + { + "epoch": 0.9631410256410257, + "grad_norm": 0.4783279597759247, + "learning_rate": 0.0001728734785840286, + "loss": 0.9397, + "step": 5409 + }, + { + "epoch": 0.9633190883190883, + "grad_norm": 0.4730507731437683, + "learning_rate": 0.00017286389245669233, + "loss": 0.9384, + "step": 5410 + }, + { + "epoch": 0.9634971509971509, + "grad_norm": 0.5309939384460449, + "learning_rate": 0.00017285430490173944, + "loss": 1.098, + "step": 5411 + }, + { + "epoch": 0.9636752136752137, + "grad_norm": 0.5177853107452393, + "learning_rate": 0.0001728447159193578, + "loss": 1.2777, + "step": 5412 + }, + { + "epoch": 0.9638532763532763, + "grad_norm": 0.6437913775444031, + "learning_rate": 0.00017283512550973526, + "loss": 1.2661, + "step": 5413 + }, + { + "epoch": 0.9640313390313391, + "grad_norm": 0.6096072196960449, + "learning_rate": 0.00017282553367305975, + "loss": 0.9569, + "step": 5414 + }, + { + "epoch": 0.9642094017094017, + "grad_norm": 0.5104934573173523, + "learning_rate": 0.00017281594040951918, + "loss": 0.9666, + "step": 5415 + }, + { + "epoch": 0.9643874643874644, + "grad_norm": 0.6178240776062012, + "learning_rate": 0.00017280634571930153, + "loss": 1.1277, + "step": 5416 + }, + { + "epoch": 0.9645655270655271, + "grad_norm": 0.5749034881591797, + "learning_rate": 0.0001727967496025948, + "loss": 1.245, + "step": 5417 + }, + { + "epoch": 0.9647435897435898, + "grad_norm": 0.5036978721618652, + "learning_rate": 0.00017278715205958694, + "loss": 1.3049, + "step": 5418 + }, + { + "epoch": 0.9649216524216524, + "grad_norm": 0.5593041777610779, + "learning_rate": 0.00017277755309046605, + "loss": 1.2304, + "step": 5419 + }, + { + "epoch": 0.9650997150997151, + "grad_norm": 0.5446555614471436, + "learning_rate": 0.0001727679526954202, + "loss": 0.732, + "step": 5420 + }, + { + "epoch": 0.9652777777777778, + "grad_norm": 0.6063070297241211, + "learning_rate": 0.00017275835087463747, + "loss": 1.3723, + "step": 5421 + }, + { + "epoch": 0.9654558404558404, + "grad_norm": 0.4994211792945862, + "learning_rate": 0.00017274874762830602, + "loss": 1.0505, + "step": 5422 + }, + { + "epoch": 0.9656339031339032, + "grad_norm": 0.49396973848342896, + "learning_rate": 0.00017273914295661395, + "loss": 0.8691, + "step": 5423 + }, + { + "epoch": 0.9658119658119658, + "grad_norm": 0.5067027807235718, + "learning_rate": 0.0001727295368597495, + "loss": 0.9744, + "step": 5424 + }, + { + "epoch": 0.9659900284900285, + "grad_norm": 0.6720643043518066, + "learning_rate": 0.00017271992933790085, + "loss": 1.1513, + "step": 5425 + }, + { + "epoch": 0.9661680911680912, + "grad_norm": 0.5494341254234314, + "learning_rate": 0.00017271032039125624, + "loss": 0.8295, + "step": 5426 + }, + { + "epoch": 0.9663461538461539, + "grad_norm": 0.644332230091095, + "learning_rate": 0.00017270071002000394, + "loss": 1.0043, + "step": 5427 + }, + { + "epoch": 0.9665242165242165, + "grad_norm": 0.5658500790596008, + "learning_rate": 0.00017269109822433225, + "loss": 1.2575, + "step": 5428 + }, + { + "epoch": 0.9667022792022792, + "grad_norm": 0.5163155794143677, + "learning_rate": 0.00017268148500442952, + "loss": 1.1391, + "step": 5429 + }, + { + "epoch": 0.9668803418803419, + "grad_norm": 0.5113703608512878, + "learning_rate": 0.00017267187036048404, + "loss": 1.0819, + "step": 5430 + }, + { + "epoch": 0.9670584045584045, + "grad_norm": 0.6339422464370728, + "learning_rate": 0.00017266225429268426, + "loss": 1.0733, + "step": 5431 + }, + { + "epoch": 0.9672364672364673, + "grad_norm": 0.5158288478851318, + "learning_rate": 0.0001726526368012185, + "loss": 0.9518, + "step": 5432 + }, + { + "epoch": 0.9674145299145299, + "grad_norm": 0.593717634677887, + "learning_rate": 0.00017264301788627527, + "loss": 0.9416, + "step": 5433 + }, + { + "epoch": 0.9675925925925926, + "grad_norm": 0.49593186378479004, + "learning_rate": 0.00017263339754804301, + "loss": 1.0307, + "step": 5434 + }, + { + "epoch": 0.9677706552706553, + "grad_norm": 0.44032949209213257, + "learning_rate": 0.00017262377578671024, + "loss": 0.7884, + "step": 5435 + }, + { + "epoch": 0.967948717948718, + "grad_norm": 0.513073742389679, + "learning_rate": 0.00017261415260246538, + "loss": 0.9797, + "step": 5436 + }, + { + "epoch": 0.9681267806267806, + "grad_norm": 0.5737422108650208, + "learning_rate": 0.0001726045279954971, + "loss": 1.0487, + "step": 5437 + }, + { + "epoch": 0.9683048433048433, + "grad_norm": 0.5385867953300476, + "learning_rate": 0.0001725949019659939, + "loss": 1.4166, + "step": 5438 + }, + { + "epoch": 0.968482905982906, + "grad_norm": 0.5224326848983765, + "learning_rate": 0.00017258527451414438, + "loss": 1.195, + "step": 5439 + }, + { + "epoch": 0.9686609686609686, + "grad_norm": 0.5305148363113403, + "learning_rate": 0.0001725756456401372, + "loss": 1.0301, + "step": 5440 + }, + { + "epoch": 0.9688390313390314, + "grad_norm": 0.532588005065918, + "learning_rate": 0.000172566015344161, + "loss": 1.1269, + "step": 5441 + }, + { + "epoch": 0.969017094017094, + "grad_norm": 0.5812515020370483, + "learning_rate": 0.0001725563836264045, + "loss": 1.1787, + "step": 5442 + }, + { + "epoch": 0.9691951566951567, + "grad_norm": 0.4962109327316284, + "learning_rate": 0.00017254675048705638, + "loss": 1.0639, + "step": 5443 + }, + { + "epoch": 0.9693732193732194, + "grad_norm": 0.5094883441925049, + "learning_rate": 0.00017253711592630534, + "loss": 1.0922, + "step": 5444 + }, + { + "epoch": 0.969551282051282, + "grad_norm": 0.5728049874305725, + "learning_rate": 0.00017252747994434025, + "loss": 1.1237, + "step": 5445 + }, + { + "epoch": 0.9697293447293447, + "grad_norm": 0.5406180620193481, + "learning_rate": 0.00017251784254134983, + "loss": 1.1161, + "step": 5446 + }, + { + "epoch": 0.9699074074074074, + "grad_norm": 0.5724552869796753, + "learning_rate": 0.00017250820371752292, + "loss": 1.2205, + "step": 5447 + }, + { + "epoch": 0.9700854700854701, + "grad_norm": 0.5698846578598022, + "learning_rate": 0.0001724985634730484, + "loss": 1.1472, + "step": 5448 + }, + { + "epoch": 0.9702635327635327, + "grad_norm": 0.5315805673599243, + "learning_rate": 0.0001724889218081151, + "loss": 1.0253, + "step": 5449 + }, + { + "epoch": 0.9704415954415955, + "grad_norm": 0.5970377326011658, + "learning_rate": 0.000172479278722912, + "loss": 1.3033, + "step": 5450 + }, + { + "epoch": 0.9706196581196581, + "grad_norm": 0.6149488687515259, + "learning_rate": 0.00017246963421762798, + "loss": 1.0689, + "step": 5451 + }, + { + "epoch": 0.9707977207977208, + "grad_norm": 0.4848574995994568, + "learning_rate": 0.00017245998829245202, + "loss": 0.8829, + "step": 5452 + }, + { + "epoch": 0.9709757834757835, + "grad_norm": 0.6073294281959534, + "learning_rate": 0.00017245034094757312, + "loss": 1.2378, + "step": 5453 + }, + { + "epoch": 0.9711538461538461, + "grad_norm": 0.6362034678459167, + "learning_rate": 0.00017244069218318026, + "loss": 1.3606, + "step": 5454 + }, + { + "epoch": 0.9713319088319088, + "grad_norm": 0.5353880524635315, + "learning_rate": 0.00017243104199946257, + "loss": 1.1288, + "step": 5455 + }, + { + "epoch": 0.9715099715099715, + "grad_norm": 0.5096352100372314, + "learning_rate": 0.00017242139039660902, + "loss": 1.0056, + "step": 5456 + }, + { + "epoch": 0.9716880341880342, + "grad_norm": 0.5086682438850403, + "learning_rate": 0.00017241173737480884, + "loss": 1.091, + "step": 5457 + }, + { + "epoch": 0.9718660968660968, + "grad_norm": 0.5034295320510864, + "learning_rate": 0.000172402082934251, + "loss": 0.9749, + "step": 5458 + }, + { + "epoch": 0.9720441595441596, + "grad_norm": 0.5205379724502563, + "learning_rate": 0.0001723924270751248, + "loss": 1.1068, + "step": 5459 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 0.5904826521873474, + "learning_rate": 0.00017238276979761937, + "loss": 1.0613, + "step": 5460 + }, + { + "epoch": 0.9724002849002849, + "grad_norm": 0.6415045261383057, + "learning_rate": 0.0001723731111019239, + "loss": 1.2126, + "step": 5461 + }, + { + "epoch": 0.9725783475783476, + "grad_norm": 0.5769147872924805, + "learning_rate": 0.0001723634509882277, + "loss": 1.337, + "step": 5462 + }, + { + "epoch": 0.9727564102564102, + "grad_norm": 0.5585111975669861, + "learning_rate": 0.00017235378945671998, + "loss": 1.3922, + "step": 5463 + }, + { + "epoch": 0.9729344729344729, + "grad_norm": 0.5788411498069763, + "learning_rate": 0.00017234412650759008, + "loss": 0.8532, + "step": 5464 + }, + { + "epoch": 0.9731125356125356, + "grad_norm": 0.5617673397064209, + "learning_rate": 0.00017233446214102728, + "loss": 1.2575, + "step": 5465 + }, + { + "epoch": 0.9732905982905983, + "grad_norm": 0.4227815568447113, + "learning_rate": 0.00017232479635722093, + "loss": 1.0618, + "step": 5466 + }, + { + "epoch": 0.9734686609686609, + "grad_norm": 0.49751797318458557, + "learning_rate": 0.00017231512915636047, + "loss": 0.7714, + "step": 5467 + }, + { + "epoch": 0.9736467236467237, + "grad_norm": 0.5983800292015076, + "learning_rate": 0.0001723054605386353, + "loss": 1.2297, + "step": 5468 + }, + { + "epoch": 0.9738247863247863, + "grad_norm": 0.543394923210144, + "learning_rate": 0.0001722957905042348, + "loss": 1.0078, + "step": 5469 + }, + { + "epoch": 0.9740028490028491, + "grad_norm": 0.5633566975593567, + "learning_rate": 0.00017228611905334846, + "loss": 1.0938, + "step": 5470 + }, + { + "epoch": 0.9741809116809117, + "grad_norm": 0.49377235770225525, + "learning_rate": 0.00017227644618616578, + "loss": 1.096, + "step": 5471 + }, + { + "epoch": 0.9743589743589743, + "grad_norm": 0.4963362216949463, + "learning_rate": 0.00017226677190287627, + "loss": 1.0003, + "step": 5472 + }, + { + "epoch": 0.9745370370370371, + "grad_norm": 0.4483006000518799, + "learning_rate": 0.00017225709620366953, + "loss": 0.8623, + "step": 5473 + }, + { + "epoch": 0.9747150997150997, + "grad_norm": 0.5429352521896362, + "learning_rate": 0.00017224741908873506, + "loss": 1.1383, + "step": 5474 + }, + { + "epoch": 0.9748931623931624, + "grad_norm": 0.5871657729148865, + "learning_rate": 0.0001722377405582625, + "loss": 1.2005, + "step": 5475 + }, + { + "epoch": 0.9750712250712251, + "grad_norm": 0.6002383828163147, + "learning_rate": 0.0001722280606124415, + "loss": 1.0696, + "step": 5476 + }, + { + "epoch": 0.9752492877492878, + "grad_norm": 0.5351617336273193, + "learning_rate": 0.00017221837925146164, + "loss": 1.243, + "step": 5477 + }, + { + "epoch": 0.9754273504273504, + "grad_norm": 0.46613118052482605, + "learning_rate": 0.00017220869647551268, + "loss": 1.0344, + "step": 5478 + }, + { + "epoch": 0.9756054131054132, + "grad_norm": 0.6015593409538269, + "learning_rate": 0.00017219901228478432, + "loss": 1.082, + "step": 5479 + }, + { + "epoch": 0.9757834757834758, + "grad_norm": 0.5829521417617798, + "learning_rate": 0.0001721893266794663, + "loss": 0.8683, + "step": 5480 + }, + { + "epoch": 0.9759615384615384, + "grad_norm": 0.6344960927963257, + "learning_rate": 0.00017217963965974838, + "loss": 1.1048, + "step": 5481 + }, + { + "epoch": 0.9761396011396012, + "grad_norm": 0.5586308240890503, + "learning_rate": 0.00017216995122582034, + "loss": 0.9657, + "step": 5482 + }, + { + "epoch": 0.9763176638176638, + "grad_norm": 0.48625239729881287, + "learning_rate": 0.00017216026137787204, + "loss": 1.1026, + "step": 5483 + }, + { + "epoch": 0.9764957264957265, + "grad_norm": 0.5625223517417908, + "learning_rate": 0.00017215057011609332, + "loss": 1.1579, + "step": 5484 + }, + { + "epoch": 0.9766737891737892, + "grad_norm": 0.6016653776168823, + "learning_rate": 0.0001721408774406741, + "loss": 1.1777, + "step": 5485 + }, + { + "epoch": 0.9768518518518519, + "grad_norm": 0.5444921851158142, + "learning_rate": 0.00017213118335180418, + "loss": 1.119, + "step": 5486 + }, + { + "epoch": 0.9770299145299145, + "grad_norm": 0.5574755668640137, + "learning_rate": 0.0001721214878496736, + "loss": 1.1128, + "step": 5487 + }, + { + "epoch": 0.9772079772079773, + "grad_norm": 0.5486113429069519, + "learning_rate": 0.00017211179093447226, + "loss": 1.1673, + "step": 5488 + }, + { + "epoch": 0.9773860398860399, + "grad_norm": 0.5545483231544495, + "learning_rate": 0.00017210209260639018, + "loss": 1.1748, + "step": 5489 + }, + { + "epoch": 0.9775641025641025, + "grad_norm": 0.5756667256355286, + "learning_rate": 0.0001720923928656174, + "loss": 1.2377, + "step": 5490 + }, + { + "epoch": 0.9777421652421653, + "grad_norm": 0.5744972229003906, + "learning_rate": 0.00017208269171234392, + "loss": 1.1242, + "step": 5491 + }, + { + "epoch": 0.9779202279202279, + "grad_norm": 0.6109468340873718, + "learning_rate": 0.00017207298914675984, + "loss": 1.1948, + "step": 5492 + }, + { + "epoch": 0.9780982905982906, + "grad_norm": 0.5195167660713196, + "learning_rate": 0.00017206328516905525, + "loss": 1.0941, + "step": 5493 + }, + { + "epoch": 0.9782763532763533, + "grad_norm": 0.5549042224884033, + "learning_rate": 0.0001720535797794203, + "loss": 1.1503, + "step": 5494 + }, + { + "epoch": 0.978454415954416, + "grad_norm": 0.6317743062973022, + "learning_rate": 0.0001720438729780451, + "loss": 1.3468, + "step": 5495 + }, + { + "epoch": 0.9786324786324786, + "grad_norm": 0.5932528972625732, + "learning_rate": 0.0001720341647651199, + "loss": 1.105, + "step": 5496 + }, + { + "epoch": 0.9788105413105413, + "grad_norm": 0.607880175113678, + "learning_rate": 0.00017202445514083488, + "loss": 1.1465, + "step": 5497 + }, + { + "epoch": 0.978988603988604, + "grad_norm": 0.49227309226989746, + "learning_rate": 0.00017201474410538027, + "loss": 0.9075, + "step": 5498 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 0.5059443116188049, + "learning_rate": 0.00017200503165894636, + "loss": 1.0483, + "step": 5499 + }, + { + "epoch": 0.9793447293447294, + "grad_norm": 0.5792799592018127, + "learning_rate": 0.0001719953178017234, + "loss": 1.0987, + "step": 5500 + }, + { + "epoch": 0.979522792022792, + "grad_norm": 0.5010457038879395, + "learning_rate": 0.00017198560253390177, + "loss": 1.1051, + "step": 5501 + }, + { + "epoch": 0.9797008547008547, + "grad_norm": 0.5866543054580688, + "learning_rate": 0.0001719758858556718, + "loss": 1.2824, + "step": 5502 + }, + { + "epoch": 0.9798789173789174, + "grad_norm": 0.5392137169837952, + "learning_rate": 0.00017196616776722382, + "loss": 0.886, + "step": 5503 + }, + { + "epoch": 0.98005698005698, + "grad_norm": 0.5200899839401245, + "learning_rate": 0.00017195644826874834, + "loss": 1.1504, + "step": 5504 + }, + { + "epoch": 0.9802350427350427, + "grad_norm": 0.533159077167511, + "learning_rate": 0.00017194672736043569, + "loss": 1.1216, + "step": 5505 + }, + { + "epoch": 0.9804131054131054, + "grad_norm": 0.5543524622917175, + "learning_rate": 0.0001719370050424764, + "loss": 1.0161, + "step": 5506 + }, + { + "epoch": 0.9805911680911681, + "grad_norm": 0.5315365195274353, + "learning_rate": 0.00017192728131506092, + "loss": 1.0509, + "step": 5507 + }, + { + "epoch": 0.9807692307692307, + "grad_norm": 0.5406147837638855, + "learning_rate": 0.00017191755617837977, + "loss": 1.0695, + "step": 5508 + }, + { + "epoch": 0.9809472934472935, + "grad_norm": 0.4563386142253876, + "learning_rate": 0.00017190782963262354, + "loss": 0.995, + "step": 5509 + }, + { + "epoch": 0.9811253561253561, + "grad_norm": 0.5456405282020569, + "learning_rate": 0.00017189810167798274, + "loss": 1.0546, + "step": 5510 + }, + { + "epoch": 0.9813034188034188, + "grad_norm": 0.6275575160980225, + "learning_rate": 0.00017188837231464795, + "loss": 1.0432, + "step": 5511 + }, + { + "epoch": 0.9814814814814815, + "grad_norm": 0.49735602736473083, + "learning_rate": 0.0001718786415428099, + "loss": 1.035, + "step": 5512 + }, + { + "epoch": 0.9816595441595442, + "grad_norm": 0.5234259963035583, + "learning_rate": 0.00017186890936265916, + "loss": 1.0918, + "step": 5513 + }, + { + "epoch": 0.9818376068376068, + "grad_norm": 0.5091170072555542, + "learning_rate": 0.00017185917577438643, + "loss": 1.0239, + "step": 5514 + }, + { + "epoch": 0.9820156695156695, + "grad_norm": 0.6155703067779541, + "learning_rate": 0.00017184944077818244, + "loss": 1.2366, + "step": 5515 + }, + { + "epoch": 0.9821937321937322, + "grad_norm": 0.5074070692062378, + "learning_rate": 0.0001718397043742379, + "loss": 1.0318, + "step": 5516 + }, + { + "epoch": 0.9823717948717948, + "grad_norm": 0.5234423279762268, + "learning_rate": 0.0001718299665627436, + "loss": 1.0322, + "step": 5517 + }, + { + "epoch": 0.9825498575498576, + "grad_norm": 0.5783474445343018, + "learning_rate": 0.0001718202273438903, + "loss": 0.9486, + "step": 5518 + }, + { + "epoch": 0.9827279202279202, + "grad_norm": 0.5708683133125305, + "learning_rate": 0.00017181048671786886, + "loss": 1.0785, + "step": 5519 + }, + { + "epoch": 0.9829059829059829, + "grad_norm": 0.5985961556434631, + "learning_rate": 0.00017180074468487009, + "loss": 1.198, + "step": 5520 + }, + { + "epoch": 0.9830840455840456, + "grad_norm": 0.5711352229118347, + "learning_rate": 0.0001717910012450849, + "loss": 1.0386, + "step": 5521 + }, + { + "epoch": 0.9832621082621082, + "grad_norm": 0.5338063836097717, + "learning_rate": 0.00017178125639870416, + "loss": 1.1594, + "step": 5522 + }, + { + "epoch": 0.9834401709401709, + "grad_norm": 0.6144943237304688, + "learning_rate": 0.00017177151014591881, + "loss": 1.1083, + "step": 5523 + }, + { + "epoch": 0.9836182336182336, + "grad_norm": 0.547285795211792, + "learning_rate": 0.00017176176248691983, + "loss": 1.1507, + "step": 5524 + }, + { + "epoch": 0.9837962962962963, + "grad_norm": 0.5807644724845886, + "learning_rate": 0.00017175201342189817, + "loss": 1.3044, + "step": 5525 + }, + { + "epoch": 0.9839743589743589, + "grad_norm": 0.5229477882385254, + "learning_rate": 0.00017174226295104485, + "loss": 1.2622, + "step": 5526 + }, + { + "epoch": 0.9841524216524217, + "grad_norm": 0.6100695133209229, + "learning_rate": 0.00017173251107455094, + "loss": 1.2026, + "step": 5527 + }, + { + "epoch": 0.9843304843304843, + "grad_norm": 0.5410884618759155, + "learning_rate": 0.00017172275779260744, + "loss": 1.2964, + "step": 5528 + }, + { + "epoch": 0.9845085470085471, + "grad_norm": 0.5937406420707703, + "learning_rate": 0.00017171300310540554, + "loss": 1.1435, + "step": 5529 + }, + { + "epoch": 0.9846866096866097, + "grad_norm": 0.56817227602005, + "learning_rate": 0.00017170324701313634, + "loss": 1.0099, + "step": 5530 + }, + { + "epoch": 0.9848646723646723, + "grad_norm": 0.5776323080062866, + "learning_rate": 0.00017169348951599092, + "loss": 1.3539, + "step": 5531 + }, + { + "epoch": 0.9850427350427351, + "grad_norm": 0.5208535194396973, + "learning_rate": 0.0001716837306141605, + "loss": 1.2306, + "step": 5532 + }, + { + "epoch": 0.9852207977207977, + "grad_norm": 0.552173376083374, + "learning_rate": 0.0001716739703078363, + "loss": 1.0551, + "step": 5533 + }, + { + "epoch": 0.9853988603988604, + "grad_norm": 0.5327515602111816, + "learning_rate": 0.00017166420859720955, + "loss": 1.2443, + "step": 5534 + }, + { + "epoch": 0.9855769230769231, + "grad_norm": 0.5255244374275208, + "learning_rate": 0.0001716544454824715, + "loss": 1.005, + "step": 5535 + }, + { + "epoch": 0.9857549857549858, + "grad_norm": 0.4753847122192383, + "learning_rate": 0.00017164468096381343, + "loss": 1.0081, + "step": 5536 + }, + { + "epoch": 0.9859330484330484, + "grad_norm": 0.5261829495429993, + "learning_rate": 0.00017163491504142665, + "loss": 1.2249, + "step": 5537 + }, + { + "epoch": 0.9861111111111112, + "grad_norm": 0.46499499678611755, + "learning_rate": 0.00017162514771550255, + "loss": 0.8759, + "step": 5538 + }, + { + "epoch": 0.9862891737891738, + "grad_norm": 0.5233004689216614, + "learning_rate": 0.00017161537898623247, + "loss": 1.0474, + "step": 5539 + }, + { + "epoch": 0.9864672364672364, + "grad_norm": 0.46905553340911865, + "learning_rate": 0.00017160560885380778, + "loss": 0.9033, + "step": 5540 + }, + { + "epoch": 0.9866452991452992, + "grad_norm": 0.5816231369972229, + "learning_rate": 0.00017159583731841998, + "loss": 1.0628, + "step": 5541 + }, + { + "epoch": 0.9868233618233618, + "grad_norm": 0.4575413167476654, + "learning_rate": 0.00017158606438026045, + "loss": 1.0446, + "step": 5542 + }, + { + "epoch": 0.9870014245014245, + "grad_norm": 0.5968109965324402, + "learning_rate": 0.00017157629003952067, + "loss": 1.032, + "step": 5543 + }, + { + "epoch": 0.9871794871794872, + "grad_norm": 0.5316148400306702, + "learning_rate": 0.00017156651429639218, + "loss": 0.9167, + "step": 5544 + }, + { + "epoch": 0.9873575498575499, + "grad_norm": 0.5185125470161438, + "learning_rate": 0.00017155673715106651, + "loss": 1.1527, + "step": 5545 + }, + { + "epoch": 0.9875356125356125, + "grad_norm": 0.5167772769927979, + "learning_rate": 0.00017154695860373525, + "loss": 0.9954, + "step": 5546 + }, + { + "epoch": 0.9877136752136753, + "grad_norm": 0.6406680345535278, + "learning_rate": 0.00017153717865458994, + "loss": 1.2758, + "step": 5547 + }, + { + "epoch": 0.9878917378917379, + "grad_norm": 0.5223956108093262, + "learning_rate": 0.00017152739730382223, + "loss": 1.1526, + "step": 5548 + }, + { + "epoch": 0.9880698005698005, + "grad_norm": 0.6131790280342102, + "learning_rate": 0.00017151761455162375, + "loss": 1.1024, + "step": 5549 + }, + { + "epoch": 0.9882478632478633, + "grad_norm": 0.5574753880500793, + "learning_rate": 0.00017150783039818616, + "loss": 0.9733, + "step": 5550 + }, + { + "epoch": 0.9884259259259259, + "grad_norm": 0.5417882800102234, + "learning_rate": 0.0001714980448437012, + "loss": 1.2244, + "step": 5551 + }, + { + "epoch": 0.9886039886039886, + "grad_norm": 0.6217474341392517, + "learning_rate": 0.0001714882578883606, + "loss": 0.9224, + "step": 5552 + }, + { + "epoch": 0.9887820512820513, + "grad_norm": 0.5846285223960876, + "learning_rate": 0.00017147846953235606, + "loss": 1.2429, + "step": 5553 + }, + { + "epoch": 0.988960113960114, + "grad_norm": 0.5924782752990723, + "learning_rate": 0.00017146867977587936, + "loss": 0.9907, + "step": 5554 + }, + { + "epoch": 0.9891381766381766, + "grad_norm": 0.5756853818893433, + "learning_rate": 0.00017145888861912242, + "loss": 1.1266, + "step": 5555 + }, + { + "epoch": 0.9893162393162394, + "grad_norm": 0.5277376770973206, + "learning_rate": 0.00017144909606227693, + "loss": 1.1676, + "step": 5556 + }, + { + "epoch": 0.989494301994302, + "grad_norm": 0.5138902068138123, + "learning_rate": 0.00017143930210553485, + "loss": 0.9864, + "step": 5557 + }, + { + "epoch": 0.9896723646723646, + "grad_norm": 0.8072507977485657, + "learning_rate": 0.00017142950674908805, + "loss": 1.111, + "step": 5558 + }, + { + "epoch": 0.9898504273504274, + "grad_norm": 0.5641721487045288, + "learning_rate": 0.00017141970999312844, + "loss": 0.9106, + "step": 5559 + }, + { + "epoch": 0.99002849002849, + "grad_norm": 0.5260798931121826, + "learning_rate": 0.000171409911837848, + "loss": 1.1609, + "step": 5560 + }, + { + "epoch": 0.9902065527065527, + "grad_norm": 0.5398530960083008, + "learning_rate": 0.00017140011228343864, + "loss": 1.0368, + "step": 5561 + }, + { + "epoch": 0.9903846153846154, + "grad_norm": 0.6011313199996948, + "learning_rate": 0.00017139031133009245, + "loss": 1.1314, + "step": 5562 + }, + { + "epoch": 0.9905626780626781, + "grad_norm": 0.6194971203804016, + "learning_rate": 0.00017138050897800135, + "loss": 1.3493, + "step": 5563 + }, + { + "epoch": 0.9907407407407407, + "grad_norm": 0.5779356956481934, + "learning_rate": 0.0001713707052273575, + "loss": 0.943, + "step": 5564 + }, + { + "epoch": 0.9909188034188035, + "grad_norm": 0.5321127772331238, + "learning_rate": 0.00017136090007835293, + "loss": 0.7914, + "step": 5565 + }, + { + "epoch": 0.9910968660968661, + "grad_norm": 0.5470426678657532, + "learning_rate": 0.00017135109353117977, + "loss": 1.2113, + "step": 5566 + }, + { + "epoch": 0.9912749287749287, + "grad_norm": 0.5551436543464661, + "learning_rate": 0.00017134128558603012, + "loss": 0.8932, + "step": 5567 + }, + { + "epoch": 0.9914529914529915, + "grad_norm": 0.45770928263664246, + "learning_rate": 0.0001713314762430962, + "loss": 1.0061, + "step": 5568 + }, + { + "epoch": 0.9916310541310541, + "grad_norm": 0.5578967332839966, + "learning_rate": 0.00017132166550257017, + "loss": 1.148, + "step": 5569 + }, + { + "epoch": 0.9918091168091168, + "grad_norm": 0.5086452960968018, + "learning_rate": 0.0001713118533646443, + "loss": 0.9803, + "step": 5570 + }, + { + "epoch": 0.9919871794871795, + "grad_norm": 0.4714745879173279, + "learning_rate": 0.00017130203982951078, + "loss": 1.0176, + "step": 5571 + }, + { + "epoch": 0.9921652421652422, + "grad_norm": 0.6254406571388245, + "learning_rate": 0.0001712922248973619, + "loss": 1.0932, + "step": 5572 + }, + { + "epoch": 0.9923433048433048, + "grad_norm": 0.5005003809928894, + "learning_rate": 0.00017128240856838998, + "loss": 1.0783, + "step": 5573 + }, + { + "epoch": 0.9925213675213675, + "grad_norm": 0.5668206214904785, + "learning_rate": 0.00017127259084278733, + "loss": 1.0404, + "step": 5574 + }, + { + "epoch": 0.9926994301994302, + "grad_norm": 0.4976036250591278, + "learning_rate": 0.00017126277172074632, + "loss": 1.1437, + "step": 5575 + }, + { + "epoch": 0.9928774928774928, + "grad_norm": 0.567546546459198, + "learning_rate": 0.00017125295120245935, + "loss": 1.2188, + "step": 5576 + }, + { + "epoch": 0.9930555555555556, + "grad_norm": 0.5614372491836548, + "learning_rate": 0.0001712431292881188, + "loss": 0.9187, + "step": 5577 + }, + { + "epoch": 0.9932336182336182, + "grad_norm": 0.6117973327636719, + "learning_rate": 0.00017123330597791712, + "loss": 1.1285, + "step": 5578 + }, + { + "epoch": 0.9934116809116809, + "grad_norm": 0.6000342965126038, + "learning_rate": 0.00017122348127204676, + "loss": 0.9837, + "step": 5579 + }, + { + "epoch": 0.9935897435897436, + "grad_norm": 0.5453050136566162, + "learning_rate": 0.0001712136551707003, + "loss": 0.8771, + "step": 5580 + }, + { + "epoch": 0.9937678062678063, + "grad_norm": 0.49603891372680664, + "learning_rate": 0.00017120382767407018, + "loss": 1.0754, + "step": 5581 + }, + { + "epoch": 0.9939458689458689, + "grad_norm": 0.48031488060951233, + "learning_rate": 0.00017119399878234894, + "loss": 0.6933, + "step": 5582 + }, + { + "epoch": 0.9941239316239316, + "grad_norm": 0.6048742532730103, + "learning_rate": 0.0001711841684957292, + "loss": 0.9696, + "step": 5583 + }, + { + "epoch": 0.9943019943019943, + "grad_norm": 0.5183123350143433, + "learning_rate": 0.00017117433681440355, + "loss": 1.1313, + "step": 5584 + }, + { + "epoch": 0.9944800569800569, + "grad_norm": 0.504916250705719, + "learning_rate": 0.00017116450373856466, + "loss": 1.0273, + "step": 5585 + }, + { + "epoch": 0.9946581196581197, + "grad_norm": 0.5804886817932129, + "learning_rate": 0.0001711546692684051, + "loss": 1.1162, + "step": 5586 + }, + { + "epoch": 0.9948361823361823, + "grad_norm": 0.5531938672065735, + "learning_rate": 0.0001711448334041176, + "loss": 1.2893, + "step": 5587 + }, + { + "epoch": 0.9950142450142451, + "grad_norm": 0.5079928636550903, + "learning_rate": 0.00017113499614589492, + "loss": 1.0393, + "step": 5588 + }, + { + "epoch": 0.9951923076923077, + "grad_norm": 0.5421964526176453, + "learning_rate": 0.00017112515749392973, + "loss": 0.8844, + "step": 5589 + }, + { + "epoch": 0.9953703703703703, + "grad_norm": 0.4834558367729187, + "learning_rate": 0.00017111531744841486, + "loss": 1.0187, + "step": 5590 + }, + { + "epoch": 0.9955484330484331, + "grad_norm": 0.6704340577125549, + "learning_rate": 0.00017110547600954307, + "loss": 0.8524, + "step": 5591 + }, + { + "epoch": 0.9957264957264957, + "grad_norm": 0.4578927159309387, + "learning_rate": 0.00017109563317750718, + "loss": 1.059, + "step": 5592 + }, + { + "epoch": 0.9959045584045584, + "grad_norm": 0.5563494563102722, + "learning_rate": 0.00017108578895250006, + "loss": 1.1211, + "step": 5593 + }, + { + "epoch": 0.9960826210826211, + "grad_norm": 0.5272170901298523, + "learning_rate": 0.00017107594333471454, + "loss": 0.9224, + "step": 5594 + }, + { + "epoch": 0.9962606837606838, + "grad_norm": 0.5697501301765442, + "learning_rate": 0.00017106609632434357, + "loss": 1.2223, + "step": 5595 + }, + { + "epoch": 0.9964387464387464, + "grad_norm": 0.5385653376579285, + "learning_rate": 0.00017105624792158007, + "loss": 1.0809, + "step": 5596 + }, + { + "epoch": 0.9966168091168092, + "grad_norm": 0.5608006119728088, + "learning_rate": 0.000171046398126617, + "loss": 1.3936, + "step": 5597 + }, + { + "epoch": 0.9967948717948718, + "grad_norm": 0.5063132643699646, + "learning_rate": 0.00017103654693964736, + "loss": 1.2086, + "step": 5598 + }, + { + "epoch": 0.9969729344729344, + "grad_norm": 0.6014235019683838, + "learning_rate": 0.00017102669436086415, + "loss": 1.1231, + "step": 5599 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 0.49549567699432373, + "learning_rate": 0.00017101684039046036, + "loss": 1.0013, + "step": 5600 + }, + { + "epoch": 0.9973290598290598, + "grad_norm": 0.517464816570282, + "learning_rate": 0.00017100698502862916, + "loss": 1.1143, + "step": 5601 + }, + { + "epoch": 0.9975071225071225, + "grad_norm": 0.514281153678894, + "learning_rate": 0.00017099712827556358, + "loss": 1.0336, + "step": 5602 + }, + { + "epoch": 0.9976851851851852, + "grad_norm": 0.5378567576408386, + "learning_rate": 0.00017098727013145672, + "loss": 0.8278, + "step": 5603 + }, + { + "epoch": 0.9978632478632479, + "grad_norm": 0.5098404884338379, + "learning_rate": 0.0001709774105965018, + "loss": 0.9902, + "step": 5604 + }, + { + "epoch": 0.9980413105413105, + "grad_norm": 0.6231759190559387, + "learning_rate": 0.00017096754967089198, + "loss": 1.0564, + "step": 5605 + }, + { + "epoch": 0.9982193732193733, + "grad_norm": 0.47434380650520325, + "learning_rate": 0.00017095768735482042, + "loss": 0.7457, + "step": 5606 + }, + { + "epoch": 0.9983974358974359, + "grad_norm": 0.5771013498306274, + "learning_rate": 0.00017094782364848035, + "loss": 1.1191, + "step": 5607 + }, + { + "epoch": 0.9985754985754985, + "grad_norm": 0.5617234706878662, + "learning_rate": 0.00017093795855206508, + "loss": 1.0779, + "step": 5608 + }, + { + "epoch": 0.9987535612535613, + "grad_norm": 0.6573554873466492, + "learning_rate": 0.00017092809206576792, + "loss": 1.0191, + "step": 5609 + }, + { + "epoch": 0.9989316239316239, + "grad_norm": 0.482834130525589, + "learning_rate": 0.00017091822418978207, + "loss": 1.0119, + "step": 5610 + }, + { + "epoch": 0.9991096866096866, + "grad_norm": 0.47496405243873596, + "learning_rate": 0.000170908354924301, + "loss": 0.8297, + "step": 5611 + }, + { + "epoch": 0.9992877492877493, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.00017089848426951796, + "loss": 1.1511, + "step": 5612 + }, + { + "epoch": 0.999465811965812, + "grad_norm": 0.5402522683143616, + "learning_rate": 0.00017088861222562643, + "loss": 1.1401, + "step": 5613 + }, + { + "epoch": 0.9996438746438746, + "grad_norm": 0.546302318572998, + "learning_rate": 0.00017087873879281977, + "loss": 0.8611, + "step": 5614 + }, + { + "epoch": 0.9998219373219374, + "grad_norm": 0.44279807806015015, + "learning_rate": 0.0001708688639712915, + "loss": 0.79, + "step": 5615 + }, + { + "epoch": 1.0, + "grad_norm": 0.5514659285545349, + "learning_rate": 0.00017085898776123502, + "loss": 1.0709, + "step": 5616 + }, + { + "epoch": 1.0, + "eval_loss": 1.093075156211853, + "eval_runtime": 24.6155, + "eval_samples_per_second": 42.29, + "eval_steps_per_second": 21.166, + "step": 5616 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.6290156841278076, + "learning_rate": 0.0001708491101628439, + "loss": 1.1786, + "step": 5617 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.4703841209411621, + "learning_rate": 0.00017083923117631162, + "loss": 0.9548, + "step": 5618 + }, + { + "epoch": 1.0003561253561253, + "grad_norm": 0.4518105089664459, + "learning_rate": 0.0001708293508018318, + "loss": 1.0089, + "step": 5619 + }, + { + "epoch": 1.000534188034188, + "grad_norm": 0.5658619403839111, + "learning_rate": 0.00017081946903959794, + "loss": 0.9466, + "step": 5620 + }, + { + "epoch": 1.0007122507122508, + "grad_norm": 0.6153838634490967, + "learning_rate": 0.00017080958588980372, + "loss": 1.2898, + "step": 5621 + }, + { + "epoch": 1.0008903133903133, + "grad_norm": 0.5245628952980042, + "learning_rate": 0.00017079970135264275, + "loss": 1.1702, + "step": 5622 + }, + { + "epoch": 1.001068376068376, + "grad_norm": 0.5291880965232849, + "learning_rate": 0.00017078981542830875, + "loss": 1.0779, + "step": 5623 + }, + { + "epoch": 1.0012464387464388, + "grad_norm": 0.500579297542572, + "learning_rate": 0.0001707799281169953, + "loss": 0.9587, + "step": 5624 + }, + { + "epoch": 1.0014245014245013, + "grad_norm": 0.45739707350730896, + "learning_rate": 0.00017077003941889625, + "loss": 0.9373, + "step": 5625 + }, + { + "epoch": 1.001602564102564, + "grad_norm": 0.5513401031494141, + "learning_rate": 0.00017076014933420526, + "loss": 1.0368, + "step": 5626 + }, + { + "epoch": 1.0017806267806268, + "grad_norm": 0.46513232588768005, + "learning_rate": 0.00017075025786311612, + "loss": 0.9422, + "step": 5627 + }, + { + "epoch": 1.0019586894586894, + "grad_norm": 0.4530394673347473, + "learning_rate": 0.00017074036500582267, + "loss": 0.8211, + "step": 5628 + }, + { + "epoch": 1.0021367521367521, + "grad_norm": 0.5612013339996338, + "learning_rate": 0.00017073047076251872, + "loss": 0.9466, + "step": 5629 + }, + { + "epoch": 1.0023148148148149, + "grad_norm": 0.4976879954338074, + "learning_rate": 0.00017072057513339812, + "loss": 0.8059, + "step": 5630 + }, + { + "epoch": 1.0024928774928774, + "grad_norm": 0.4842833876609802, + "learning_rate": 0.00017071067811865476, + "loss": 0.6554, + "step": 5631 + }, + { + "epoch": 1.0026709401709402, + "grad_norm": 0.5446373224258423, + "learning_rate": 0.00017070077971848257, + "loss": 1.1001, + "step": 5632 + }, + { + "epoch": 1.002849002849003, + "grad_norm": 0.5996584892272949, + "learning_rate": 0.00017069087993307544, + "loss": 1.0317, + "step": 5633 + }, + { + "epoch": 1.0030270655270654, + "grad_norm": 0.5369443297386169, + "learning_rate": 0.00017068097876262738, + "loss": 0.8019, + "step": 5634 + }, + { + "epoch": 1.0032051282051282, + "grad_norm": 0.4985966682434082, + "learning_rate": 0.00017067107620733236, + "loss": 1.0121, + "step": 5635 + }, + { + "epoch": 1.003383190883191, + "grad_norm": 0.5262824892997742, + "learning_rate": 0.0001706611722673844, + "loss": 1.0157, + "step": 5636 + }, + { + "epoch": 1.0035612535612535, + "grad_norm": 0.5912795066833496, + "learning_rate": 0.00017065126694297756, + "loss": 1.0327, + "step": 5637 + }, + { + "epoch": 1.0037393162393162, + "grad_norm": 0.5866343379020691, + "learning_rate": 0.00017064136023430595, + "loss": 1.1194, + "step": 5638 + }, + { + "epoch": 1.003917378917379, + "grad_norm": 0.5009918808937073, + "learning_rate": 0.0001706314521415636, + "loss": 1.0467, + "step": 5639 + }, + { + "epoch": 1.0040954415954415, + "grad_norm": 0.5455304384231567, + "learning_rate": 0.00017062154266494464, + "loss": 0.8749, + "step": 5640 + }, + { + "epoch": 1.0042735042735043, + "grad_norm": 0.5648258328437805, + "learning_rate": 0.00017061163180464328, + "loss": 0.9408, + "step": 5641 + }, + { + "epoch": 1.004451566951567, + "grad_norm": 0.5276365876197815, + "learning_rate": 0.00017060171956085368, + "loss": 0.9681, + "step": 5642 + }, + { + "epoch": 1.0046296296296295, + "grad_norm": 0.5212745070457458, + "learning_rate": 0.00017059180593377007, + "loss": 0.9188, + "step": 5643 + }, + { + "epoch": 1.0048076923076923, + "grad_norm": 0.540626585483551, + "learning_rate": 0.00017058189092358664, + "loss": 1.0809, + "step": 5644 + }, + { + "epoch": 1.004985754985755, + "grad_norm": 0.5592377781867981, + "learning_rate": 0.00017057197453049767, + "loss": 0.8589, + "step": 5645 + }, + { + "epoch": 1.0051638176638176, + "grad_norm": 0.5115051865577698, + "learning_rate": 0.00017056205675469746, + "loss": 0.8006, + "step": 5646 + }, + { + "epoch": 1.0053418803418803, + "grad_norm": 0.5031117796897888, + "learning_rate": 0.00017055213759638034, + "loss": 0.9242, + "step": 5647 + }, + { + "epoch": 1.005519943019943, + "grad_norm": 0.5342774987220764, + "learning_rate": 0.00017054221705574066, + "loss": 0.8268, + "step": 5648 + }, + { + "epoch": 1.0056980056980056, + "grad_norm": 0.44480493664741516, + "learning_rate": 0.00017053229513297276, + "loss": 0.6892, + "step": 5649 + }, + { + "epoch": 1.0058760683760684, + "grad_norm": 0.5032621622085571, + "learning_rate": 0.00017052237182827105, + "loss": 0.971, + "step": 5650 + }, + { + "epoch": 1.006054131054131, + "grad_norm": 0.5611015558242798, + "learning_rate": 0.00017051244714182996, + "loss": 0.9403, + "step": 5651 + }, + { + "epoch": 1.0062321937321936, + "grad_norm": 0.5064613223075867, + "learning_rate": 0.00017050252107384393, + "loss": 0.9718, + "step": 5652 + }, + { + "epoch": 1.0064102564102564, + "grad_norm": 0.6458395719528198, + "learning_rate": 0.0001704925936245075, + "loss": 1.1161, + "step": 5653 + }, + { + "epoch": 1.0065883190883191, + "grad_norm": 0.527418315410614, + "learning_rate": 0.00017048266479401512, + "loss": 0.9315, + "step": 5654 + }, + { + "epoch": 1.0067663817663817, + "grad_norm": 0.5127941370010376, + "learning_rate": 0.00017047273458256133, + "loss": 0.8206, + "step": 5655 + }, + { + "epoch": 1.0069444444444444, + "grad_norm": 0.6257100105285645, + "learning_rate": 0.00017046280299034067, + "loss": 0.9854, + "step": 5656 + }, + { + "epoch": 1.0071225071225072, + "grad_norm": 0.5081700682640076, + "learning_rate": 0.0001704528700175478, + "loss": 0.9478, + "step": 5657 + }, + { + "epoch": 1.0073005698005697, + "grad_norm": 0.598127543926239, + "learning_rate": 0.00017044293566437725, + "loss": 1.0721, + "step": 5658 + }, + { + "epoch": 1.0074786324786325, + "grad_norm": 0.5429877638816833, + "learning_rate": 0.00017043299993102376, + "loss": 0.9732, + "step": 5659 + }, + { + "epoch": 1.0076566951566952, + "grad_norm": 0.6006619334220886, + "learning_rate": 0.00017042306281768194, + "loss": 1.1262, + "step": 5660 + }, + { + "epoch": 1.0078347578347577, + "grad_norm": 0.48933324217796326, + "learning_rate": 0.00017041312432454646, + "loss": 0.8596, + "step": 5661 + }, + { + "epoch": 1.0080128205128205, + "grad_norm": 0.5902166366577148, + "learning_rate": 0.0001704031844518121, + "loss": 1.1035, + "step": 5662 + }, + { + "epoch": 1.0081908831908832, + "grad_norm": 0.523597776889801, + "learning_rate": 0.0001703932431996736, + "loss": 0.7117, + "step": 5663 + }, + { + "epoch": 1.0083689458689458, + "grad_norm": 0.6313928365707397, + "learning_rate": 0.00017038330056832573, + "loss": 1.0204, + "step": 5664 + }, + { + "epoch": 1.0085470085470085, + "grad_norm": 0.5627471804618835, + "learning_rate": 0.00017037335655796328, + "loss": 0.7648, + "step": 5665 + }, + { + "epoch": 1.0087250712250713, + "grad_norm": 0.5817851424217224, + "learning_rate": 0.0001703634111687811, + "loss": 1.0452, + "step": 5666 + }, + { + "epoch": 1.0089031339031338, + "grad_norm": 0.5143535137176514, + "learning_rate": 0.00017035346440097407, + "loss": 0.9788, + "step": 5667 + }, + { + "epoch": 1.0090811965811965, + "grad_norm": 0.5331187844276428, + "learning_rate": 0.000170343516254737, + "loss": 0.7584, + "step": 5668 + }, + { + "epoch": 1.0092592592592593, + "grad_norm": 0.5723634362220764, + "learning_rate": 0.00017033356673026487, + "loss": 0.9435, + "step": 5669 + }, + { + "epoch": 1.0094373219373218, + "grad_norm": 0.6012297868728638, + "learning_rate": 0.00017032361582775265, + "loss": 1.142, + "step": 5670 + }, + { + "epoch": 1.0096153846153846, + "grad_norm": 0.6161282658576965, + "learning_rate": 0.00017031366354739523, + "loss": 1.2823, + "step": 5671 + }, + { + "epoch": 1.0097934472934473, + "grad_norm": 0.5088054537773132, + "learning_rate": 0.00017030370988938763, + "loss": 0.9743, + "step": 5672 + }, + { + "epoch": 1.0099715099715099, + "grad_norm": 0.512003481388092, + "learning_rate": 0.0001702937548539249, + "loss": 0.9112, + "step": 5673 + }, + { + "epoch": 1.0101495726495726, + "grad_norm": 0.5565149784088135, + "learning_rate": 0.00017028379844120207, + "loss": 1.0074, + "step": 5674 + }, + { + "epoch": 1.0103276353276354, + "grad_norm": 0.6463099718093872, + "learning_rate": 0.00017027384065141418, + "loss": 1.175, + "step": 5675 + }, + { + "epoch": 1.010505698005698, + "grad_norm": 0.46999064087867737, + "learning_rate": 0.00017026388148475637, + "loss": 0.8429, + "step": 5676 + }, + { + "epoch": 1.0106837606837606, + "grad_norm": 0.5617384910583496, + "learning_rate": 0.00017025392094142377, + "loss": 1.045, + "step": 5677 + }, + { + "epoch": 1.0108618233618234, + "grad_norm": 0.5156623721122742, + "learning_rate": 0.00017024395902161154, + "loss": 1.016, + "step": 5678 + }, + { + "epoch": 1.0110398860398861, + "grad_norm": 0.5693390369415283, + "learning_rate": 0.00017023399572551484, + "loss": 0.8616, + "step": 5679 + }, + { + "epoch": 1.0112179487179487, + "grad_norm": 0.5234879851341248, + "learning_rate": 0.00017022403105332892, + "loss": 0.9244, + "step": 5680 + }, + { + "epoch": 1.0113960113960114, + "grad_norm": 0.6513097286224365, + "learning_rate": 0.00017021406500524893, + "loss": 0.9565, + "step": 5681 + }, + { + "epoch": 1.0115740740740742, + "grad_norm": 0.5788878202438354, + "learning_rate": 0.00017020409758147022, + "loss": 0.8994, + "step": 5682 + }, + { + "epoch": 1.0117521367521367, + "grad_norm": 0.5495247840881348, + "learning_rate": 0.00017019412878218807, + "loss": 0.9371, + "step": 5683 + }, + { + "epoch": 1.0119301994301995, + "grad_norm": 0.639045238494873, + "learning_rate": 0.00017018415860759777, + "loss": 1.0297, + "step": 5684 + }, + { + "epoch": 1.0121082621082622, + "grad_norm": 0.5167784690856934, + "learning_rate": 0.0001701741870578947, + "loss": 0.8974, + "step": 5685 + }, + { + "epoch": 1.0122863247863247, + "grad_norm": 0.6131011247634888, + "learning_rate": 0.00017016421413327417, + "loss": 1.13, + "step": 5686 + }, + { + "epoch": 1.0124643874643875, + "grad_norm": 0.4804688096046448, + "learning_rate": 0.00017015423983393166, + "loss": 1.0098, + "step": 5687 + }, + { + "epoch": 1.0126424501424502, + "grad_norm": 0.6605221629142761, + "learning_rate": 0.00017014426416006253, + "loss": 1.1123, + "step": 5688 + }, + { + "epoch": 1.0128205128205128, + "grad_norm": 0.5523666739463806, + "learning_rate": 0.00017013428711186226, + "loss": 0.8226, + "step": 5689 + }, + { + "epoch": 1.0129985754985755, + "grad_norm": 0.6012941598892212, + "learning_rate": 0.00017012430868952632, + "loss": 0.8915, + "step": 5690 + }, + { + "epoch": 1.0131766381766383, + "grad_norm": 0.5830875039100647, + "learning_rate": 0.00017011432889325022, + "loss": 1.021, + "step": 5691 + }, + { + "epoch": 1.0133547008547008, + "grad_norm": 0.5546056032180786, + "learning_rate": 0.0001701043477232295, + "loss": 0.7656, + "step": 5692 + }, + { + "epoch": 1.0135327635327636, + "grad_norm": 0.5592601299285889, + "learning_rate": 0.0001700943651796597, + "loss": 1.0172, + "step": 5693 + }, + { + "epoch": 1.0137108262108263, + "grad_norm": 0.5708866715431213, + "learning_rate": 0.00017008438126273645, + "loss": 1.0012, + "step": 5694 + }, + { + "epoch": 1.0138888888888888, + "grad_norm": 0.6856338381767273, + "learning_rate": 0.0001700743959726553, + "loss": 1.1278, + "step": 5695 + }, + { + "epoch": 1.0140669515669516, + "grad_norm": 0.6523802876472473, + "learning_rate": 0.000170064409309612, + "loss": 1.0406, + "step": 5696 + }, + { + "epoch": 1.0142450142450143, + "grad_norm": 0.6653079986572266, + "learning_rate": 0.00017005442127380208, + "loss": 1.1086, + "step": 5697 + }, + { + "epoch": 1.0144230769230769, + "grad_norm": 0.5841104388237, + "learning_rate": 0.00017004443186542133, + "loss": 0.9335, + "step": 5698 + }, + { + "epoch": 1.0146011396011396, + "grad_norm": 0.5696784257888794, + "learning_rate": 0.0001700344410846654, + "loss": 1.0247, + "step": 5699 + }, + { + "epoch": 1.0147792022792024, + "grad_norm": 0.7135653495788574, + "learning_rate": 0.00017002444893173013, + "loss": 1.0259, + "step": 5700 + }, + { + "epoch": 1.014957264957265, + "grad_norm": 0.5806999802589417, + "learning_rate": 0.00017001445540681124, + "loss": 1.0053, + "step": 5701 + }, + { + "epoch": 1.0151353276353277, + "grad_norm": 0.5298715829849243, + "learning_rate": 0.0001700044605101045, + "loss": 0.9415, + "step": 5702 + }, + { + "epoch": 1.0153133903133904, + "grad_norm": 0.5817379951477051, + "learning_rate": 0.0001699944642418058, + "loss": 1.0906, + "step": 5703 + }, + { + "epoch": 1.015491452991453, + "grad_norm": 0.6564923524856567, + "learning_rate": 0.00016998446660211098, + "loss": 0.9933, + "step": 5704 + }, + { + "epoch": 1.0156695156695157, + "grad_norm": 0.6547308564186096, + "learning_rate": 0.00016997446759121592, + "loss": 1.0045, + "step": 5705 + }, + { + "epoch": 1.0158475783475784, + "grad_norm": 0.5763013958930969, + "learning_rate": 0.00016996446720931652, + "loss": 1.0898, + "step": 5706 + }, + { + "epoch": 1.016025641025641, + "grad_norm": 0.6118074059486389, + "learning_rate": 0.00016995446545660871, + "loss": 0.9398, + "step": 5707 + }, + { + "epoch": 1.0162037037037037, + "grad_norm": 0.6810526251792908, + "learning_rate": 0.0001699444623332885, + "loss": 1.0968, + "step": 5708 + }, + { + "epoch": 1.0163817663817665, + "grad_norm": 0.5292752981185913, + "learning_rate": 0.00016993445783955184, + "loss": 0.7549, + "step": 5709 + }, + { + "epoch": 1.016559829059829, + "grad_norm": 0.6014277935028076, + "learning_rate": 0.00016992445197559474, + "loss": 1.1711, + "step": 5710 + }, + { + "epoch": 1.0167378917378918, + "grad_norm": 0.5089772343635559, + "learning_rate": 0.00016991444474161326, + "loss": 0.9188, + "step": 5711 + }, + { + "epoch": 1.0169159544159545, + "grad_norm": 0.567193865776062, + "learning_rate": 0.0001699044361378035, + "loss": 0.7462, + "step": 5712 + }, + { + "epoch": 1.017094017094017, + "grad_norm": 0.5638598799705505, + "learning_rate": 0.00016989442616436147, + "loss": 0.9643, + "step": 5713 + }, + { + "epoch": 1.0172720797720798, + "grad_norm": 0.5634039640426636, + "learning_rate": 0.0001698844148214834, + "loss": 1.0141, + "step": 5714 + }, + { + "epoch": 1.0174501424501425, + "grad_norm": 0.5326652526855469, + "learning_rate": 0.00016987440210936537, + "loss": 0.865, + "step": 5715 + }, + { + "epoch": 1.017628205128205, + "grad_norm": 0.5858046412467957, + "learning_rate": 0.0001698643880282036, + "loss": 0.9561, + "step": 5716 + }, + { + "epoch": 1.0178062678062678, + "grad_norm": 0.6424698829650879, + "learning_rate": 0.00016985437257819428, + "loss": 1.0169, + "step": 5717 + }, + { + "epoch": 1.0179843304843306, + "grad_norm": 0.6294280290603638, + "learning_rate": 0.00016984435575953364, + "loss": 1.0438, + "step": 5718 + }, + { + "epoch": 1.018162393162393, + "grad_norm": 0.5533088445663452, + "learning_rate": 0.00016983433757241788, + "loss": 0.8901, + "step": 5719 + }, + { + "epoch": 1.0183404558404558, + "grad_norm": 0.5148718953132629, + "learning_rate": 0.00016982431801704342, + "loss": 0.9201, + "step": 5720 + }, + { + "epoch": 1.0185185185185186, + "grad_norm": 0.5609371662139893, + "learning_rate": 0.00016981429709360645, + "loss": 0.9347, + "step": 5721 + }, + { + "epoch": 1.0186965811965811, + "grad_norm": 0.5502731204032898, + "learning_rate": 0.00016980427480230338, + "loss": 1.0508, + "step": 5722 + }, + { + "epoch": 1.0188746438746439, + "grad_norm": 0.5880394577980042, + "learning_rate": 0.00016979425114333055, + "loss": 1.1258, + "step": 5723 + }, + { + "epoch": 1.0190527065527066, + "grad_norm": 0.5569866895675659, + "learning_rate": 0.0001697842261168843, + "loss": 0.9186, + "step": 5724 + }, + { + "epoch": 1.0192307692307692, + "grad_norm": 0.7468093037605286, + "learning_rate": 0.00016977419972316116, + "loss": 1.2066, + "step": 5725 + }, + { + "epoch": 1.019408831908832, + "grad_norm": 0.6041515469551086, + "learning_rate": 0.00016976417196235753, + "loss": 0.939, + "step": 5726 + }, + { + "epoch": 1.0195868945868947, + "grad_norm": 0.6102641224861145, + "learning_rate": 0.00016975414283466983, + "loss": 0.8334, + "step": 5727 + }, + { + "epoch": 1.0197649572649572, + "grad_norm": 0.5418640375137329, + "learning_rate": 0.00016974411234029467, + "loss": 0.8072, + "step": 5728 + }, + { + "epoch": 1.01994301994302, + "grad_norm": 0.6569705605506897, + "learning_rate": 0.00016973408047942843, + "loss": 1.103, + "step": 5729 + }, + { + "epoch": 1.0201210826210827, + "grad_norm": 0.5778102278709412, + "learning_rate": 0.00016972404725226778, + "loss": 0.9353, + "step": 5730 + }, + { + "epoch": 1.0202991452991452, + "grad_norm": 0.5474382638931274, + "learning_rate": 0.0001697140126590093, + "loss": 1.0009, + "step": 5731 + }, + { + "epoch": 1.020477207977208, + "grad_norm": 0.5869506597518921, + "learning_rate": 0.00016970397669984947, + "loss": 1.0027, + "step": 5732 + }, + { + "epoch": 1.0206552706552707, + "grad_norm": 0.5078117251396179, + "learning_rate": 0.00016969393937498508, + "loss": 0.8316, + "step": 5733 + }, + { + "epoch": 1.0208333333333333, + "grad_norm": 0.5488452911376953, + "learning_rate": 0.0001696839006846127, + "loss": 0.8438, + "step": 5734 + }, + { + "epoch": 1.021011396011396, + "grad_norm": 0.5921052098274231, + "learning_rate": 0.00016967386062892908, + "loss": 0.9147, + "step": 5735 + }, + { + "epoch": 1.0211894586894588, + "grad_norm": 0.5486881136894226, + "learning_rate": 0.00016966381920813085, + "loss": 0.7619, + "step": 5736 + }, + { + "epoch": 1.0213675213675213, + "grad_norm": 0.5250689387321472, + "learning_rate": 0.00016965377642241483, + "loss": 0.9192, + "step": 5737 + }, + { + "epoch": 1.021545584045584, + "grad_norm": 0.5355087518692017, + "learning_rate": 0.00016964373227197773, + "loss": 0.954, + "step": 5738 + }, + { + "epoch": 1.0217236467236468, + "grad_norm": 0.6758780479431152, + "learning_rate": 0.0001696336867570164, + "loss": 1.1257, + "step": 5739 + }, + { + "epoch": 1.0219017094017093, + "grad_norm": 0.6361044049263, + "learning_rate": 0.00016962363987772756, + "loss": 1.0889, + "step": 5740 + }, + { + "epoch": 1.022079772079772, + "grad_norm": 0.5802326798439026, + "learning_rate": 0.00016961359163430819, + "loss": 0.8966, + "step": 5741 + }, + { + "epoch": 1.0222578347578348, + "grad_norm": 0.5535712242126465, + "learning_rate": 0.00016960354202695508, + "loss": 1.0007, + "step": 5742 + }, + { + "epoch": 1.0224358974358974, + "grad_norm": 0.5469220280647278, + "learning_rate": 0.00016959349105586516, + "loss": 0.8202, + "step": 5743 + }, + { + "epoch": 1.02261396011396, + "grad_norm": 0.5533008575439453, + "learning_rate": 0.00016958343872123534, + "loss": 0.9576, + "step": 5744 + }, + { + "epoch": 1.0227920227920229, + "grad_norm": 0.615132749080658, + "learning_rate": 0.00016957338502326258, + "loss": 0.8719, + "step": 5745 + }, + { + "epoch": 1.0229700854700854, + "grad_norm": 0.519075334072113, + "learning_rate": 0.0001695633299621439, + "loss": 0.8309, + "step": 5746 + }, + { + "epoch": 1.0231481481481481, + "grad_norm": 0.6249759197235107, + "learning_rate": 0.00016955327353807624, + "loss": 1.151, + "step": 5747 + }, + { + "epoch": 1.023326210826211, + "grad_norm": 0.560299277305603, + "learning_rate": 0.00016954321575125668, + "loss": 0.7889, + "step": 5748 + }, + { + "epoch": 1.0235042735042734, + "grad_norm": 0.5735262036323547, + "learning_rate": 0.0001695331566018823, + "loss": 0.8794, + "step": 5749 + }, + { + "epoch": 1.0236823361823362, + "grad_norm": 0.5893994569778442, + "learning_rate": 0.00016952309609015012, + "loss": 0.9696, + "step": 5750 + }, + { + "epoch": 1.023860398860399, + "grad_norm": 0.6064512133598328, + "learning_rate": 0.0001695130342162573, + "loss": 0.9771, + "step": 5751 + }, + { + "epoch": 1.0240384615384615, + "grad_norm": 0.5833427309989929, + "learning_rate": 0.00016950297098040099, + "loss": 1.1768, + "step": 5752 + }, + { + "epoch": 1.0242165242165242, + "grad_norm": 0.5940282344818115, + "learning_rate": 0.00016949290638277833, + "loss": 1.0758, + "step": 5753 + }, + { + "epoch": 1.024394586894587, + "grad_norm": 0.5267124772071838, + "learning_rate": 0.00016948284042358656, + "loss": 0.772, + "step": 5754 + }, + { + "epoch": 1.0245726495726495, + "grad_norm": 0.6217982172966003, + "learning_rate": 0.00016947277310302284, + "loss": 0.8583, + "step": 5755 + }, + { + "epoch": 1.0247507122507122, + "grad_norm": 0.6192215085029602, + "learning_rate": 0.00016946270442128443, + "loss": 0.9148, + "step": 5756 + }, + { + "epoch": 1.024928774928775, + "grad_norm": 0.5337123870849609, + "learning_rate": 0.00016945263437856867, + "loss": 1.0054, + "step": 5757 + }, + { + "epoch": 1.0251068376068375, + "grad_norm": 0.5462040901184082, + "learning_rate": 0.00016944256297507276, + "loss": 1.1097, + "step": 5758 + }, + { + "epoch": 1.0252849002849003, + "grad_norm": 0.5606170892715454, + "learning_rate": 0.00016943249021099415, + "loss": 1.0192, + "step": 5759 + }, + { + "epoch": 1.025462962962963, + "grad_norm": 0.636974573135376, + "learning_rate": 0.00016942241608653008, + "loss": 1.0241, + "step": 5760 + }, + { + "epoch": 1.0256410256410255, + "grad_norm": 0.4895164966583252, + "learning_rate": 0.00016941234060187797, + "loss": 0.9057, + "step": 5761 + }, + { + "epoch": 1.0258190883190883, + "grad_norm": 0.5810303092002869, + "learning_rate": 0.00016940226375723527, + "loss": 1.0809, + "step": 5762 + }, + { + "epoch": 1.025997150997151, + "grad_norm": 0.6043853163719177, + "learning_rate": 0.00016939218555279937, + "loss": 1.0685, + "step": 5763 + }, + { + "epoch": 1.0261752136752136, + "grad_norm": 0.5827188491821289, + "learning_rate": 0.00016938210598876774, + "loss": 1.0236, + "step": 5764 + }, + { + "epoch": 1.0263532763532763, + "grad_norm": 0.6677887439727783, + "learning_rate": 0.0001693720250653379, + "loss": 1.0586, + "step": 5765 + }, + { + "epoch": 1.026531339031339, + "grad_norm": 0.558051347732544, + "learning_rate": 0.0001693619427827073, + "loss": 0.745, + "step": 5766 + }, + { + "epoch": 1.0267094017094016, + "grad_norm": 0.6336706280708313, + "learning_rate": 0.0001693518591410735, + "loss": 1.0658, + "step": 5767 + }, + { + "epoch": 1.0268874643874644, + "grad_norm": 0.7077126502990723, + "learning_rate": 0.00016934177414063416, + "loss": 1.18, + "step": 5768 + }, + { + "epoch": 1.0270655270655271, + "grad_norm": 0.5342326760292053, + "learning_rate": 0.00016933168778158675, + "loss": 0.8347, + "step": 5769 + }, + { + "epoch": 1.0272435897435896, + "grad_norm": 0.6116416454315186, + "learning_rate": 0.00016932160006412895, + "loss": 1.0648, + "step": 5770 + }, + { + "epoch": 1.0274216524216524, + "grad_norm": 0.5411320924758911, + "learning_rate": 0.0001693115109884584, + "loss": 1.0756, + "step": 5771 + }, + { + "epoch": 1.0275997150997151, + "grad_norm": 0.5549847483634949, + "learning_rate": 0.00016930142055477277, + "loss": 0.7259, + "step": 5772 + }, + { + "epoch": 1.0277777777777777, + "grad_norm": 0.549010694026947, + "learning_rate": 0.00016929132876326977, + "loss": 0.9488, + "step": 5773 + }, + { + "epoch": 1.0279558404558404, + "grad_norm": 0.6302017569541931, + "learning_rate": 0.00016928123561414714, + "loss": 0.8851, + "step": 5774 + }, + { + "epoch": 1.0281339031339032, + "grad_norm": 0.5831273198127747, + "learning_rate": 0.00016927114110760257, + "loss": 0.7841, + "step": 5775 + }, + { + "epoch": 1.0283119658119657, + "grad_norm": 0.5528474450111389, + "learning_rate": 0.00016926104524383394, + "loss": 1.0108, + "step": 5776 + }, + { + "epoch": 1.0284900284900285, + "grad_norm": 0.6279126405715942, + "learning_rate": 0.00016925094802303897, + "loss": 0.8632, + "step": 5777 + }, + { + "epoch": 1.0286680911680912, + "grad_norm": 0.6783218383789062, + "learning_rate": 0.00016924084944541554, + "loss": 1.0746, + "step": 5778 + }, + { + "epoch": 1.0288461538461537, + "grad_norm": 0.5823925137519836, + "learning_rate": 0.00016923074951116153, + "loss": 1.0486, + "step": 5779 + }, + { + "epoch": 1.0290242165242165, + "grad_norm": 0.6095981597900391, + "learning_rate": 0.00016922064822047473, + "loss": 0.8113, + "step": 5780 + }, + { + "epoch": 1.0292022792022792, + "grad_norm": 0.7887664437294006, + "learning_rate": 0.00016921054557355317, + "loss": 1.2411, + "step": 5781 + }, + { + "epoch": 1.0293803418803418, + "grad_norm": 0.6511263251304626, + "learning_rate": 0.00016920044157059475, + "loss": 0.924, + "step": 5782 + }, + { + "epoch": 1.0295584045584045, + "grad_norm": 0.6045661568641663, + "learning_rate": 0.00016919033621179744, + "loss": 0.8373, + "step": 5783 + }, + { + "epoch": 1.0297364672364673, + "grad_norm": 0.6914188861846924, + "learning_rate": 0.0001691802294973592, + "loss": 0.9589, + "step": 5784 + }, + { + "epoch": 1.0299145299145298, + "grad_norm": 0.6483730673789978, + "learning_rate": 0.00016917012142747805, + "loss": 0.9871, + "step": 5785 + }, + { + "epoch": 1.0300925925925926, + "grad_norm": 0.5775033235549927, + "learning_rate": 0.0001691600120023521, + "loss": 1.0591, + "step": 5786 + }, + { + "epoch": 1.0302706552706553, + "grad_norm": 0.6206814646720886, + "learning_rate": 0.00016914990122217932, + "loss": 0.9126, + "step": 5787 + }, + { + "epoch": 1.0304487179487178, + "grad_norm": 0.5422028303146362, + "learning_rate": 0.00016913978908715796, + "loss": 0.8227, + "step": 5788 + }, + { + "epoch": 1.0306267806267806, + "grad_norm": 0.5824416875839233, + "learning_rate": 0.000169129675597486, + "loss": 1.111, + "step": 5789 + }, + { + "epoch": 1.0308048433048433, + "grad_norm": 0.5419015884399414, + "learning_rate": 0.00016911956075336165, + "loss": 0.8941, + "step": 5790 + }, + { + "epoch": 1.0309829059829059, + "grad_norm": 0.6171557903289795, + "learning_rate": 0.0001691094445549831, + "loss": 0.8679, + "step": 5791 + }, + { + "epoch": 1.0311609686609686, + "grad_norm": 0.6136980056762695, + "learning_rate": 0.00016909932700254855, + "loss": 0.9266, + "step": 5792 + }, + { + "epoch": 1.0313390313390314, + "grad_norm": 0.6275020241737366, + "learning_rate": 0.00016908920809625624, + "loss": 1.0828, + "step": 5793 + }, + { + "epoch": 1.0315170940170941, + "grad_norm": 0.6538251638412476, + "learning_rate": 0.0001690790878363044, + "loss": 0.8413, + "step": 5794 + }, + { + "epoch": 1.0316951566951567, + "grad_norm": 0.5981295108795166, + "learning_rate": 0.00016906896622289136, + "loss": 0.9845, + "step": 5795 + }, + { + "epoch": 1.0318732193732194, + "grad_norm": 0.5390967130661011, + "learning_rate": 0.00016905884325621538, + "loss": 0.8755, + "step": 5796 + }, + { + "epoch": 1.032051282051282, + "grad_norm": 0.5534448623657227, + "learning_rate": 0.00016904871893647482, + "loss": 1.1868, + "step": 5797 + }, + { + "epoch": 1.0322293447293447, + "grad_norm": 0.664556086063385, + "learning_rate": 0.00016903859326386806, + "loss": 1.1418, + "step": 5798 + }, + { + "epoch": 1.0324074074074074, + "grad_norm": 0.5737143158912659, + "learning_rate": 0.00016902846623859346, + "loss": 1.124, + "step": 5799 + }, + { + "epoch": 1.0325854700854702, + "grad_norm": 0.6499935388565063, + "learning_rate": 0.0001690183378608495, + "loss": 1.0331, + "step": 5800 + }, + { + "epoch": 1.0327635327635327, + "grad_norm": 0.5721518993377686, + "learning_rate": 0.00016900820813083454, + "loss": 0.8664, + "step": 5801 + }, + { + "epoch": 1.0329415954415955, + "grad_norm": 0.5651140809059143, + "learning_rate": 0.0001689980770487471, + "loss": 1.1661, + "step": 5802 + }, + { + "epoch": 1.0331196581196582, + "grad_norm": 0.5935871005058289, + "learning_rate": 0.0001689879446147857, + "loss": 0.8722, + "step": 5803 + }, + { + "epoch": 1.0332977207977208, + "grad_norm": 0.5627842545509338, + "learning_rate": 0.00016897781082914884, + "loss": 1.0036, + "step": 5804 + }, + { + "epoch": 1.0334757834757835, + "grad_norm": 0.5866895914077759, + "learning_rate": 0.00016896767569203502, + "loss": 0.9739, + "step": 5805 + }, + { + "epoch": 1.0336538461538463, + "grad_norm": 0.5568059682846069, + "learning_rate": 0.0001689575392036429, + "loss": 0.7081, + "step": 5806 + }, + { + "epoch": 1.0338319088319088, + "grad_norm": 0.6054235100746155, + "learning_rate": 0.00016894740136417103, + "loss": 1.1168, + "step": 5807 + }, + { + "epoch": 1.0340099715099715, + "grad_norm": 0.5215454697608948, + "learning_rate": 0.00016893726217381805, + "loss": 0.9172, + "step": 5808 + }, + { + "epoch": 1.0341880341880343, + "grad_norm": 0.5415732860565186, + "learning_rate": 0.00016892712163278263, + "loss": 0.7812, + "step": 5809 + }, + { + "epoch": 1.0343660968660968, + "grad_norm": 0.6341692805290222, + "learning_rate": 0.00016891697974126345, + "loss": 1.0658, + "step": 5810 + }, + { + "epoch": 1.0345441595441596, + "grad_norm": 0.6326245665550232, + "learning_rate": 0.00016890683649945922, + "loss": 1.0134, + "step": 5811 + }, + { + "epoch": 1.0347222222222223, + "grad_norm": 0.5729571580886841, + "learning_rate": 0.00016889669190756868, + "loss": 0.9139, + "step": 5812 + }, + { + "epoch": 1.0349002849002849, + "grad_norm": 0.5912853479385376, + "learning_rate": 0.00016888654596579054, + "loss": 1.122, + "step": 5813 + }, + { + "epoch": 1.0350783475783476, + "grad_norm": 0.8410450220108032, + "learning_rate": 0.00016887639867432368, + "loss": 1.3009, + "step": 5814 + }, + { + "epoch": 1.0352564102564104, + "grad_norm": 0.5416620969772339, + "learning_rate": 0.00016886625003336683, + "loss": 0.8751, + "step": 5815 + }, + { + "epoch": 1.0354344729344729, + "grad_norm": 0.6367851495742798, + "learning_rate": 0.0001688561000431189, + "loss": 0.956, + "step": 5816 + }, + { + "epoch": 1.0356125356125356, + "grad_norm": 0.4618827700614929, + "learning_rate": 0.0001688459487037787, + "loss": 0.5313, + "step": 5817 + }, + { + "epoch": 1.0357905982905984, + "grad_norm": 0.7139244079589844, + "learning_rate": 0.00016883579601554516, + "loss": 1.0787, + "step": 5818 + }, + { + "epoch": 1.035968660968661, + "grad_norm": 0.6896135210990906, + "learning_rate": 0.00016882564197861715, + "loss": 0.932, + "step": 5819 + }, + { + "epoch": 1.0361467236467237, + "grad_norm": 0.5889739394187927, + "learning_rate": 0.00016881548659319372, + "loss": 0.8852, + "step": 5820 + }, + { + "epoch": 1.0363247863247864, + "grad_norm": 0.5954701900482178, + "learning_rate": 0.00016880532985947375, + "loss": 0.8192, + "step": 5821 + }, + { + "epoch": 1.036502849002849, + "grad_norm": 0.6665091514587402, + "learning_rate": 0.00016879517177765627, + "loss": 0.9578, + "step": 5822 + }, + { + "epoch": 1.0366809116809117, + "grad_norm": 0.5990539789199829, + "learning_rate": 0.00016878501234794034, + "loss": 0.9797, + "step": 5823 + }, + { + "epoch": 1.0368589743589745, + "grad_norm": 0.596755862236023, + "learning_rate": 0.00016877485157052496, + "loss": 1.173, + "step": 5824 + }, + { + "epoch": 1.037037037037037, + "grad_norm": 0.544658362865448, + "learning_rate": 0.00016876468944560923, + "loss": 1.0742, + "step": 5825 + }, + { + "epoch": 1.0372150997150997, + "grad_norm": 0.5841910243034363, + "learning_rate": 0.00016875452597339225, + "loss": 1.029, + "step": 5826 + }, + { + "epoch": 1.0373931623931625, + "grad_norm": 0.6508592963218689, + "learning_rate": 0.00016874436115407317, + "loss": 0.9883, + "step": 5827 + }, + { + "epoch": 1.037571225071225, + "grad_norm": 0.590050458908081, + "learning_rate": 0.00016873419498785114, + "loss": 1.0713, + "step": 5828 + }, + { + "epoch": 1.0377492877492878, + "grad_norm": 0.5386307239532471, + "learning_rate": 0.00016872402747492534, + "loss": 1.0159, + "step": 5829 + }, + { + "epoch": 1.0379273504273505, + "grad_norm": 0.6173896193504333, + "learning_rate": 0.00016871385861549497, + "loss": 1.0056, + "step": 5830 + }, + { + "epoch": 1.038105413105413, + "grad_norm": 0.5377787351608276, + "learning_rate": 0.0001687036884097593, + "loss": 0.8708, + "step": 5831 + }, + { + "epoch": 1.0382834757834758, + "grad_norm": 0.5753569006919861, + "learning_rate": 0.00016869351685791756, + "loss": 1.0529, + "step": 5832 + }, + { + "epoch": 1.0384615384615385, + "grad_norm": 0.6085895299911499, + "learning_rate": 0.00016868334396016906, + "loss": 1.1017, + "step": 5833 + }, + { + "epoch": 1.038639601139601, + "grad_norm": 0.6320509910583496, + "learning_rate": 0.0001686731697167131, + "loss": 1.0543, + "step": 5834 + }, + { + "epoch": 1.0388176638176638, + "grad_norm": 0.5691760778427124, + "learning_rate": 0.00016866299412774907, + "loss": 0.9975, + "step": 5835 + }, + { + "epoch": 1.0389957264957266, + "grad_norm": 0.5990765690803528, + "learning_rate": 0.0001686528171934763, + "loss": 0.8776, + "step": 5836 + }, + { + "epoch": 1.039173789173789, + "grad_norm": 0.6650477647781372, + "learning_rate": 0.00016864263891409415, + "loss": 1.0652, + "step": 5837 + }, + { + "epoch": 1.0393518518518519, + "grad_norm": 0.6050353646278381, + "learning_rate": 0.00016863245928980212, + "loss": 0.9313, + "step": 5838 + }, + { + "epoch": 1.0395299145299146, + "grad_norm": 0.587505578994751, + "learning_rate": 0.0001686222783207996, + "loss": 0.9892, + "step": 5839 + }, + { + "epoch": 1.0397079772079771, + "grad_norm": 0.6310170292854309, + "learning_rate": 0.00016861209600728608, + "loss": 1.1045, + "step": 5840 + }, + { + "epoch": 1.03988603988604, + "grad_norm": 0.5683430433273315, + "learning_rate": 0.0001686019123494611, + "loss": 1.0507, + "step": 5841 + }, + { + "epoch": 1.0400641025641026, + "grad_norm": 0.6621488332748413, + "learning_rate": 0.00016859172734752414, + "loss": 0.9255, + "step": 5842 + }, + { + "epoch": 1.0402421652421652, + "grad_norm": 0.6197706460952759, + "learning_rate": 0.00016858154100167475, + "loss": 1.0031, + "step": 5843 + }, + { + "epoch": 1.040420227920228, + "grad_norm": 0.6805898547172546, + "learning_rate": 0.00016857135331211257, + "loss": 0.9901, + "step": 5844 + }, + { + "epoch": 1.0405982905982907, + "grad_norm": 0.5512405633926392, + "learning_rate": 0.00016856116427903714, + "loss": 1.0033, + "step": 5845 + }, + { + "epoch": 1.0407763532763532, + "grad_norm": 0.5643384456634521, + "learning_rate": 0.00016855097390264815, + "loss": 0.9136, + "step": 5846 + }, + { + "epoch": 1.040954415954416, + "grad_norm": 0.48351922631263733, + "learning_rate": 0.0001685407821831452, + "loss": 0.6163, + "step": 5847 + }, + { + "epoch": 1.0411324786324787, + "grad_norm": 0.6256039142608643, + "learning_rate": 0.00016853058912072802, + "loss": 0.9409, + "step": 5848 + }, + { + "epoch": 1.0413105413105412, + "grad_norm": 0.6539996862411499, + "learning_rate": 0.00016852039471559627, + "loss": 0.9367, + "step": 5849 + }, + { + "epoch": 1.041488603988604, + "grad_norm": 0.6192609667778015, + "learning_rate": 0.00016851019896794975, + "loss": 0.9631, + "step": 5850 + }, + { + "epoch": 1.0416666666666667, + "grad_norm": 0.613563060760498, + "learning_rate": 0.0001685000018779882, + "loss": 0.9132, + "step": 5851 + }, + { + "epoch": 1.0418447293447293, + "grad_norm": 0.6004200577735901, + "learning_rate": 0.0001684898034459114, + "loss": 1.1313, + "step": 5852 + }, + { + "epoch": 1.042022792022792, + "grad_norm": 0.6158567070960999, + "learning_rate": 0.0001684796036719192, + "loss": 1.0253, + "step": 5853 + }, + { + "epoch": 1.0422008547008548, + "grad_norm": 0.6362335085868835, + "learning_rate": 0.00016846940255621143, + "loss": 0.93, + "step": 5854 + }, + { + "epoch": 1.0423789173789173, + "grad_norm": 0.6148427128791809, + "learning_rate": 0.00016845920009898787, + "loss": 0.9122, + "step": 5855 + }, + { + "epoch": 1.04255698005698, + "grad_norm": 0.5119984149932861, + "learning_rate": 0.00016844899630044858, + "loss": 0.7954, + "step": 5856 + }, + { + "epoch": 1.0427350427350428, + "grad_norm": 0.571849524974823, + "learning_rate": 0.00016843879116079338, + "loss": 0.8588, + "step": 5857 + }, + { + "epoch": 1.0429131054131053, + "grad_norm": 0.6173384785652161, + "learning_rate": 0.00016842858468022221, + "loss": 1.0475, + "step": 5858 + }, + { + "epoch": 1.043091168091168, + "grad_norm": 0.566114068031311, + "learning_rate": 0.0001684183768589351, + "loss": 0.8485, + "step": 5859 + }, + { + "epoch": 1.0432692307692308, + "grad_norm": 0.653134286403656, + "learning_rate": 0.000168408167697132, + "loss": 0.9976, + "step": 5860 + }, + { + "epoch": 1.0434472934472934, + "grad_norm": 0.63815838098526, + "learning_rate": 0.00016839795719501296, + "loss": 0.7091, + "step": 5861 + }, + { + "epoch": 1.0436253561253561, + "grad_norm": 0.5109001994132996, + "learning_rate": 0.00016838774535277805, + "loss": 0.7668, + "step": 5862 + }, + { + "epoch": 1.0438034188034189, + "grad_norm": 0.6741907596588135, + "learning_rate": 0.0001683775321706273, + "loss": 1.0493, + "step": 5863 + }, + { + "epoch": 1.0439814814814814, + "grad_norm": 0.6006115674972534, + "learning_rate": 0.0001683673176487609, + "loss": 0.9784, + "step": 5864 + }, + { + "epoch": 1.0441595441595442, + "grad_norm": 0.5504778027534485, + "learning_rate": 0.0001683571017873789, + "loss": 0.9718, + "step": 5865 + }, + { + "epoch": 1.044337606837607, + "grad_norm": 0.5713102221488953, + "learning_rate": 0.00016834688458668148, + "loss": 1.12, + "step": 5866 + }, + { + "epoch": 1.0445156695156694, + "grad_norm": 0.7878454923629761, + "learning_rate": 0.00016833666604686886, + "loss": 1.1803, + "step": 5867 + }, + { + "epoch": 1.0446937321937322, + "grad_norm": 0.582697331905365, + "learning_rate": 0.00016832644616814122, + "loss": 0.943, + "step": 5868 + }, + { + "epoch": 1.044871794871795, + "grad_norm": 0.5300645232200623, + "learning_rate": 0.00016831622495069878, + "loss": 0.9087, + "step": 5869 + }, + { + "epoch": 1.0450498575498575, + "grad_norm": 0.5627666115760803, + "learning_rate": 0.00016830600239474186, + "loss": 1.081, + "step": 5870 + }, + { + "epoch": 1.0452279202279202, + "grad_norm": 0.6760496497154236, + "learning_rate": 0.0001682957785004707, + "loss": 1.1098, + "step": 5871 + }, + { + "epoch": 1.045405982905983, + "grad_norm": 0.6424084901809692, + "learning_rate": 0.00016828555326808565, + "loss": 0.9657, + "step": 5872 + }, + { + "epoch": 1.0455840455840455, + "grad_norm": 0.5523313283920288, + "learning_rate": 0.000168275326697787, + "loss": 1.0163, + "step": 5873 + }, + { + "epoch": 1.0457621082621082, + "grad_norm": 0.5582337975502014, + "learning_rate": 0.00016826509878977518, + "loss": 0.8825, + "step": 5874 + }, + { + "epoch": 1.045940170940171, + "grad_norm": 0.5603214502334595, + "learning_rate": 0.00016825486954425055, + "loss": 0.9032, + "step": 5875 + }, + { + "epoch": 1.0461182336182335, + "grad_norm": 0.5944222807884216, + "learning_rate": 0.00016824463896141355, + "loss": 0.9384, + "step": 5876 + }, + { + "epoch": 1.0462962962962963, + "grad_norm": 0.6220229268074036, + "learning_rate": 0.00016823440704146457, + "loss": 0.8962, + "step": 5877 + }, + { + "epoch": 1.046474358974359, + "grad_norm": 0.5607972145080566, + "learning_rate": 0.0001682241737846042, + "loss": 0.9385, + "step": 5878 + }, + { + "epoch": 1.0466524216524216, + "grad_norm": 0.6206870079040527, + "learning_rate": 0.00016821393919103282, + "loss": 1.0597, + "step": 5879 + }, + { + "epoch": 1.0468304843304843, + "grad_norm": 0.5126399993896484, + "learning_rate": 0.000168203703260951, + "loss": 0.9403, + "step": 5880 + }, + { + "epoch": 1.047008547008547, + "grad_norm": 0.6569282412528992, + "learning_rate": 0.00016819346599455929, + "loss": 0.8124, + "step": 5881 + }, + { + "epoch": 1.0471866096866096, + "grad_norm": 0.6670137047767639, + "learning_rate": 0.0001681832273920583, + "loss": 1.1927, + "step": 5882 + }, + { + "epoch": 1.0473646723646723, + "grad_norm": 0.5403243899345398, + "learning_rate": 0.00016817298745364862, + "loss": 0.8539, + "step": 5883 + }, + { + "epoch": 1.047542735042735, + "grad_norm": 0.5500505566596985, + "learning_rate": 0.00016816274617953086, + "loss": 1.1064, + "step": 5884 + }, + { + "epoch": 1.0477207977207976, + "grad_norm": 0.5482703447341919, + "learning_rate": 0.00016815250356990566, + "loss": 0.7276, + "step": 5885 + }, + { + "epoch": 1.0478988603988604, + "grad_norm": 0.6290771961212158, + "learning_rate": 0.00016814225962497373, + "loss": 0.9018, + "step": 5886 + }, + { + "epoch": 1.0480769230769231, + "grad_norm": 0.6404094696044922, + "learning_rate": 0.00016813201434493578, + "loss": 1.0638, + "step": 5887 + }, + { + "epoch": 1.0482549857549857, + "grad_norm": 0.5484994053840637, + "learning_rate": 0.0001681217677299926, + "loss": 1.0033, + "step": 5888 + }, + { + "epoch": 1.0484330484330484, + "grad_norm": 0.6474852561950684, + "learning_rate": 0.0001681115197803448, + "loss": 1.1017, + "step": 5889 + }, + { + "epoch": 1.0486111111111112, + "grad_norm": 0.6186243295669556, + "learning_rate": 0.0001681012704961933, + "loss": 0.9978, + "step": 5890 + }, + { + "epoch": 1.0487891737891737, + "grad_norm": 0.6244034767150879, + "learning_rate": 0.00016809101987773887, + "loss": 0.9906, + "step": 5891 + }, + { + "epoch": 1.0489672364672364, + "grad_norm": 0.5893426537513733, + "learning_rate": 0.00016808076792518235, + "loss": 0.9345, + "step": 5892 + }, + { + "epoch": 1.0491452991452992, + "grad_norm": 0.6283876299858093, + "learning_rate": 0.0001680705146387246, + "loss": 1.0041, + "step": 5893 + }, + { + "epoch": 1.0493233618233617, + "grad_norm": 0.6075255870819092, + "learning_rate": 0.00016806026001856656, + "loss": 1.0661, + "step": 5894 + }, + { + "epoch": 1.0495014245014245, + "grad_norm": 0.5350496768951416, + "learning_rate": 0.00016805000406490907, + "loss": 0.6789, + "step": 5895 + }, + { + "epoch": 1.0496794871794872, + "grad_norm": 0.5380373597145081, + "learning_rate": 0.00016803974677795312, + "loss": 0.8889, + "step": 5896 + }, + { + "epoch": 1.0498575498575498, + "grad_norm": 0.6145668029785156, + "learning_rate": 0.0001680294881578997, + "loss": 0.8952, + "step": 5897 + }, + { + "epoch": 1.0500356125356125, + "grad_norm": 0.5666532516479492, + "learning_rate": 0.00016801922820494972, + "loss": 0.9697, + "step": 5898 + }, + { + "epoch": 1.0502136752136753, + "grad_norm": 0.5352747440338135, + "learning_rate": 0.0001680089669193043, + "loss": 0.9619, + "step": 5899 + }, + { + "epoch": 1.0503917378917378, + "grad_norm": 0.5405527949333191, + "learning_rate": 0.00016799870430116444, + "loss": 0.8733, + "step": 5900 + }, + { + "epoch": 1.0505698005698005, + "grad_norm": 0.5936748385429382, + "learning_rate": 0.00016798844035073124, + "loss": 0.8746, + "step": 5901 + }, + { + "epoch": 1.0507478632478633, + "grad_norm": 0.539652943611145, + "learning_rate": 0.00016797817506820578, + "loss": 0.8743, + "step": 5902 + }, + { + "epoch": 1.0509259259259258, + "grad_norm": 0.644528865814209, + "learning_rate": 0.00016796790845378915, + "loss": 0.9251, + "step": 5903 + }, + { + "epoch": 1.0511039886039886, + "grad_norm": 0.5429201126098633, + "learning_rate": 0.00016795764050768258, + "loss": 0.747, + "step": 5904 + }, + { + "epoch": 1.0512820512820513, + "grad_norm": 0.6432006359100342, + "learning_rate": 0.00016794737123008725, + "loss": 0.9166, + "step": 5905 + }, + { + "epoch": 1.0514601139601139, + "grad_norm": 0.6084117293357849, + "learning_rate": 0.00016793710062120427, + "loss": 1.0778, + "step": 5906 + }, + { + "epoch": 1.0516381766381766, + "grad_norm": 0.5351580381393433, + "learning_rate": 0.00016792682868123495, + "loss": 0.9124, + "step": 5907 + }, + { + "epoch": 1.0518162393162394, + "grad_norm": 0.7078854441642761, + "learning_rate": 0.00016791655541038053, + "loss": 1.1209, + "step": 5908 + }, + { + "epoch": 1.051994301994302, + "grad_norm": 0.5943832993507385, + "learning_rate": 0.0001679062808088423, + "loss": 0.9077, + "step": 5909 + }, + { + "epoch": 1.0521723646723646, + "grad_norm": 0.5216894745826721, + "learning_rate": 0.00016789600487682156, + "loss": 0.9866, + "step": 5910 + }, + { + "epoch": 1.0523504273504274, + "grad_norm": 0.738451361656189, + "learning_rate": 0.00016788572761451963, + "loss": 1.1611, + "step": 5911 + }, + { + "epoch": 1.05252849002849, + "grad_norm": 0.6411251425743103, + "learning_rate": 0.00016787544902213791, + "loss": 1.1481, + "step": 5912 + }, + { + "epoch": 1.0527065527065527, + "grad_norm": 0.6768319010734558, + "learning_rate": 0.00016786516909987774, + "loss": 0.8614, + "step": 5913 + }, + { + "epoch": 1.0528846153846154, + "grad_norm": 0.5838070511817932, + "learning_rate": 0.0001678548878479406, + "loss": 0.9719, + "step": 5914 + }, + { + "epoch": 1.0530626780626782, + "grad_norm": 0.541522741317749, + "learning_rate": 0.00016784460526652784, + "loss": 0.767, + "step": 5915 + }, + { + "epoch": 1.0532407407407407, + "grad_norm": 0.6064762473106384, + "learning_rate": 0.000167834321355841, + "loss": 1.0792, + "step": 5916 + }, + { + "epoch": 1.0534188034188035, + "grad_norm": 0.5515492558479309, + "learning_rate": 0.00016782403611608152, + "loss": 0.7897, + "step": 5917 + }, + { + "epoch": 1.0535968660968662, + "grad_norm": 0.6326262950897217, + "learning_rate": 0.000167813749547451, + "loss": 0.9279, + "step": 5918 + }, + { + "epoch": 1.0537749287749287, + "grad_norm": 0.6262009739875793, + "learning_rate": 0.0001678034616501509, + "loss": 0.9752, + "step": 5919 + }, + { + "epoch": 1.0539529914529915, + "grad_norm": 0.6049023270606995, + "learning_rate": 0.00016779317242438278, + "loss": 0.9167, + "step": 5920 + }, + { + "epoch": 1.0541310541310542, + "grad_norm": 0.6286031007766724, + "learning_rate": 0.0001677828818703483, + "loss": 1.1277, + "step": 5921 + }, + { + "epoch": 1.0543091168091168, + "grad_norm": 0.662086009979248, + "learning_rate": 0.00016777258998824907, + "loss": 1.0824, + "step": 5922 + }, + { + "epoch": 1.0544871794871795, + "grad_norm": 0.5358783006668091, + "learning_rate": 0.00016776229677828672, + "loss": 0.825, + "step": 5923 + }, + { + "epoch": 1.0546652421652423, + "grad_norm": 0.490326464176178, + "learning_rate": 0.00016775200224066294, + "loss": 0.7916, + "step": 5924 + }, + { + "epoch": 1.0548433048433048, + "grad_norm": 0.5940443277359009, + "learning_rate": 0.0001677417063755794, + "loss": 1.0121, + "step": 5925 + }, + { + "epoch": 1.0550213675213675, + "grad_norm": 0.5974507927894592, + "learning_rate": 0.00016773140918323787, + "loss": 0.7629, + "step": 5926 + }, + { + "epoch": 1.0551994301994303, + "grad_norm": 0.5747174024581909, + "learning_rate": 0.00016772111066384003, + "loss": 0.9373, + "step": 5927 + }, + { + "epoch": 1.0553774928774928, + "grad_norm": 0.5998024940490723, + "learning_rate": 0.00016771081081758772, + "loss": 0.8543, + "step": 5928 + }, + { + "epoch": 1.0555555555555556, + "grad_norm": 0.5771155953407288, + "learning_rate": 0.00016770050964468275, + "loss": 0.9108, + "step": 5929 + }, + { + "epoch": 1.0557336182336183, + "grad_norm": 0.5695661306381226, + "learning_rate": 0.00016769020714532692, + "loss": 0.8055, + "step": 5930 + }, + { + "epoch": 1.0559116809116809, + "grad_norm": 0.6164212226867676, + "learning_rate": 0.0001676799033197221, + "loss": 1.0917, + "step": 5931 + }, + { + "epoch": 1.0560897435897436, + "grad_norm": 0.6092487573623657, + "learning_rate": 0.00016766959816807018, + "loss": 0.9276, + "step": 5932 + }, + { + "epoch": 1.0562678062678064, + "grad_norm": 0.5595401525497437, + "learning_rate": 0.00016765929169057305, + "loss": 0.9435, + "step": 5933 + }, + { + "epoch": 1.056445868945869, + "grad_norm": 0.5875109434127808, + "learning_rate": 0.00016764898388743263, + "loss": 0.959, + "step": 5934 + }, + { + "epoch": 1.0566239316239316, + "grad_norm": 0.6045668721199036, + "learning_rate": 0.00016763867475885088, + "loss": 0.8636, + "step": 5935 + }, + { + "epoch": 1.0568019943019944, + "grad_norm": 0.6088171005249023, + "learning_rate": 0.00016762836430502987, + "loss": 0.6807, + "step": 5936 + }, + { + "epoch": 1.056980056980057, + "grad_norm": 0.6293274760246277, + "learning_rate": 0.00016761805252617148, + "loss": 1.042, + "step": 5937 + }, + { + "epoch": 1.0571581196581197, + "grad_norm": 0.588472843170166, + "learning_rate": 0.00016760773942247785, + "loss": 0.8896, + "step": 5938 + }, + { + "epoch": 1.0573361823361824, + "grad_norm": 0.4412326216697693, + "learning_rate": 0.000167597424994151, + "loss": 0.6727, + "step": 5939 + }, + { + "epoch": 1.057514245014245, + "grad_norm": 0.6086825132369995, + "learning_rate": 0.00016758710924139302, + "loss": 0.9908, + "step": 5940 + }, + { + "epoch": 1.0576923076923077, + "grad_norm": 0.6424705386161804, + "learning_rate": 0.00016757679216440608, + "loss": 1.0182, + "step": 5941 + }, + { + "epoch": 1.0578703703703705, + "grad_norm": 0.6610676050186157, + "learning_rate": 0.00016756647376339222, + "loss": 0.9645, + "step": 5942 + }, + { + "epoch": 1.058048433048433, + "grad_norm": 0.598292887210846, + "learning_rate": 0.0001675561540385537, + "loss": 0.9694, + "step": 5943 + }, + { + "epoch": 1.0582264957264957, + "grad_norm": 0.6941167116165161, + "learning_rate": 0.00016754583299009266, + "loss": 1.0786, + "step": 5944 + }, + { + "epoch": 1.0584045584045585, + "grad_norm": 0.6543232798576355, + "learning_rate": 0.00016753551061821133, + "loss": 1.0488, + "step": 5945 + }, + { + "epoch": 1.058582621082621, + "grad_norm": 0.606159508228302, + "learning_rate": 0.000167525186923112, + "loss": 0.9448, + "step": 5946 + }, + { + "epoch": 1.0587606837606838, + "grad_norm": 0.5051791071891785, + "learning_rate": 0.00016751486190499685, + "loss": 0.7485, + "step": 5947 + }, + { + "epoch": 1.0589387464387465, + "grad_norm": 0.6459367275238037, + "learning_rate": 0.00016750453556406826, + "loss": 1.0055, + "step": 5948 + }, + { + "epoch": 1.059116809116809, + "grad_norm": 0.551591157913208, + "learning_rate": 0.00016749420790052852, + "loss": 0.9717, + "step": 5949 + }, + { + "epoch": 1.0592948717948718, + "grad_norm": 0.5899214148521423, + "learning_rate": 0.00016748387891458, + "loss": 0.7774, + "step": 5950 + }, + { + "epoch": 1.0594729344729346, + "grad_norm": 0.582379162311554, + "learning_rate": 0.00016747354860642503, + "loss": 0.953, + "step": 5951 + }, + { + "epoch": 1.059650997150997, + "grad_norm": 0.6035816073417664, + "learning_rate": 0.00016746321697626605, + "loss": 1.1175, + "step": 5952 + }, + { + "epoch": 1.0598290598290598, + "grad_norm": 0.6476401686668396, + "learning_rate": 0.00016745288402430548, + "loss": 0.9448, + "step": 5953 + }, + { + "epoch": 1.0600071225071226, + "grad_norm": 0.6126405596733093, + "learning_rate": 0.00016744254975074578, + "loss": 0.882, + "step": 5954 + }, + { + "epoch": 1.0601851851851851, + "grad_norm": 0.5333579182624817, + "learning_rate": 0.0001674322141557894, + "loss": 0.9539, + "step": 5955 + }, + { + "epoch": 1.0603632478632479, + "grad_norm": 0.6085022687911987, + "learning_rate": 0.0001674218772396389, + "loss": 1.0028, + "step": 5956 + }, + { + "epoch": 1.0605413105413106, + "grad_norm": 0.5809528827667236, + "learning_rate": 0.0001674115390024967, + "loss": 0.84, + "step": 5957 + }, + { + "epoch": 1.0607193732193732, + "grad_norm": 0.5820229649543762, + "learning_rate": 0.00016740119944456548, + "loss": 0.9563, + "step": 5958 + }, + { + "epoch": 1.060897435897436, + "grad_norm": 0.6349015831947327, + "learning_rate": 0.00016739085856604775, + "loss": 0.9739, + "step": 5959 + }, + { + "epoch": 1.0610754985754987, + "grad_norm": 0.6346020102500916, + "learning_rate": 0.00016738051636714616, + "loss": 0.907, + "step": 5960 + }, + { + "epoch": 1.0612535612535612, + "grad_norm": 0.5850573778152466, + "learning_rate": 0.0001673701728480633, + "loss": 1.0688, + "step": 5961 + }, + { + "epoch": 1.061431623931624, + "grad_norm": 0.6258122324943542, + "learning_rate": 0.00016735982800900184, + "loss": 0.9997, + "step": 5962 + }, + { + "epoch": 1.0616096866096867, + "grad_norm": 0.6744239330291748, + "learning_rate": 0.00016734948185016452, + "loss": 0.9431, + "step": 5963 + }, + { + "epoch": 1.0617877492877492, + "grad_norm": 0.5769457817077637, + "learning_rate": 0.000167339134371754, + "loss": 0.9658, + "step": 5964 + }, + { + "epoch": 1.061965811965812, + "grad_norm": 0.6385112404823303, + "learning_rate": 0.000167328785573973, + "loss": 1.0199, + "step": 5965 + }, + { + "epoch": 1.0621438746438747, + "grad_norm": 0.536522388458252, + "learning_rate": 0.00016731843545702435, + "loss": 0.8496, + "step": 5966 + }, + { + "epoch": 1.0623219373219372, + "grad_norm": 0.5978497862815857, + "learning_rate": 0.00016730808402111075, + "loss": 0.8536, + "step": 5967 + }, + { + "epoch": 1.0625, + "grad_norm": 0.6091681122779846, + "learning_rate": 0.0001672977312664351, + "loss": 1.0241, + "step": 5968 + }, + { + "epoch": 1.0626780626780628, + "grad_norm": 0.5807273387908936, + "learning_rate": 0.0001672873771932002, + "loss": 1.0522, + "step": 5969 + }, + { + "epoch": 1.0628561253561253, + "grad_norm": 0.6511965990066528, + "learning_rate": 0.0001672770218016089, + "loss": 0.8908, + "step": 5970 + }, + { + "epoch": 1.063034188034188, + "grad_norm": 0.6241721510887146, + "learning_rate": 0.00016726666509186416, + "loss": 0.9854, + "step": 5971 + }, + { + "epoch": 1.0632122507122508, + "grad_norm": 0.6112468242645264, + "learning_rate": 0.0001672563070641688, + "loss": 1.0091, + "step": 5972 + }, + { + "epoch": 1.0633903133903133, + "grad_norm": 0.6135509014129639, + "learning_rate": 0.00016724594771872587, + "loss": 0.8891, + "step": 5973 + }, + { + "epoch": 1.063568376068376, + "grad_norm": 0.608384370803833, + "learning_rate": 0.00016723558705573823, + "loss": 1.017, + "step": 5974 + }, + { + "epoch": 1.0637464387464388, + "grad_norm": 0.6578485369682312, + "learning_rate": 0.00016722522507540895, + "loss": 0.9165, + "step": 5975 + }, + { + "epoch": 1.0639245014245013, + "grad_norm": 0.562588095664978, + "learning_rate": 0.00016721486177794106, + "loss": 0.7989, + "step": 5976 + }, + { + "epoch": 1.064102564102564, + "grad_norm": 0.5541409254074097, + "learning_rate": 0.00016720449716353753, + "loss": 0.8917, + "step": 5977 + }, + { + "epoch": 1.0642806267806268, + "grad_norm": 0.551167905330658, + "learning_rate": 0.0001671941312324015, + "loss": 0.824, + "step": 5978 + }, + { + "epoch": 1.0644586894586894, + "grad_norm": 0.6280582547187805, + "learning_rate": 0.0001671837639847361, + "loss": 0.9708, + "step": 5979 + }, + { + "epoch": 1.0646367521367521, + "grad_norm": 0.6389226913452148, + "learning_rate": 0.00016717339542074436, + "loss": 1.0081, + "step": 5980 + }, + { + "epoch": 1.0648148148148149, + "grad_norm": 0.6677889823913574, + "learning_rate": 0.0001671630255406295, + "loss": 1.2709, + "step": 5981 + }, + { + "epoch": 1.0649928774928774, + "grad_norm": 0.5748161673545837, + "learning_rate": 0.00016715265434459465, + "loss": 0.9157, + "step": 5982 + }, + { + "epoch": 1.0651709401709402, + "grad_norm": 0.6677651405334473, + "learning_rate": 0.00016714228183284304, + "loss": 1.1097, + "step": 5983 + }, + { + "epoch": 1.065349002849003, + "grad_norm": 0.6253604292869568, + "learning_rate": 0.0001671319080055779, + "loss": 0.9819, + "step": 5984 + }, + { + "epoch": 1.0655270655270654, + "grad_norm": 0.5548844337463379, + "learning_rate": 0.0001671215328630025, + "loss": 0.9324, + "step": 5985 + }, + { + "epoch": 1.0657051282051282, + "grad_norm": 0.622062623500824, + "learning_rate": 0.00016711115640532004, + "loss": 0.8749, + "step": 5986 + }, + { + "epoch": 1.065883190883191, + "grad_norm": 0.6496043801307678, + "learning_rate": 0.00016710077863273394, + "loss": 1.0642, + "step": 5987 + }, + { + "epoch": 1.0660612535612535, + "grad_norm": 0.6140534281730652, + "learning_rate": 0.00016709039954544746, + "loss": 0.8928, + "step": 5988 + }, + { + "epoch": 1.0662393162393162, + "grad_norm": 0.6387218236923218, + "learning_rate": 0.00016708001914366393, + "loss": 0.9525, + "step": 5989 + }, + { + "epoch": 1.066417378917379, + "grad_norm": 0.6119858026504517, + "learning_rate": 0.0001670696374275868, + "loss": 0.8663, + "step": 5990 + }, + { + "epoch": 1.0665954415954415, + "grad_norm": 0.6722040772438049, + "learning_rate": 0.00016705925439741947, + "loss": 1.1173, + "step": 5991 + }, + { + "epoch": 1.0667735042735043, + "grad_norm": 0.8226081132888794, + "learning_rate": 0.00016704887005336534, + "loss": 1.0572, + "step": 5992 + }, + { + "epoch": 1.066951566951567, + "grad_norm": 0.7248596549034119, + "learning_rate": 0.00016703848439562785, + "loss": 1.0493, + "step": 5993 + }, + { + "epoch": 1.0671296296296295, + "grad_norm": 0.7185787558555603, + "learning_rate": 0.00016702809742441058, + "loss": 1.1366, + "step": 5994 + }, + { + "epoch": 1.0673076923076923, + "grad_norm": 0.6118780970573425, + "learning_rate": 0.00016701770913991694, + "loss": 0.9557, + "step": 5995 + }, + { + "epoch": 1.067485754985755, + "grad_norm": 0.6472596526145935, + "learning_rate": 0.0001670073195423505, + "loss": 0.9977, + "step": 5996 + }, + { + "epoch": 1.0676638176638176, + "grad_norm": 0.7110133767127991, + "learning_rate": 0.00016699692863191484, + "loss": 1.1932, + "step": 5997 + }, + { + "epoch": 1.0678418803418803, + "grad_norm": 0.5827305912971497, + "learning_rate": 0.00016698653640881354, + "loss": 0.7641, + "step": 5998 + }, + { + "epoch": 1.068019943019943, + "grad_norm": 0.527208149433136, + "learning_rate": 0.00016697614287325017, + "loss": 0.7683, + "step": 5999 + }, + { + "epoch": 1.0681980056980056, + "grad_norm": 0.6680626273155212, + "learning_rate": 0.00016696574802542848, + "loss": 1.1748, + "step": 6000 + }, + { + "epoch": 1.0683760683760684, + "grad_norm": 0.5947227478027344, + "learning_rate": 0.00016695535186555204, + "loss": 1.0894, + "step": 6001 + }, + { + "epoch": 1.068554131054131, + "grad_norm": 0.5828250646591187, + "learning_rate": 0.00016694495439382456, + "loss": 0.9895, + "step": 6002 + }, + { + "epoch": 1.0687321937321936, + "grad_norm": 0.5897728204727173, + "learning_rate": 0.00016693455561044978, + "loss": 0.9686, + "step": 6003 + }, + { + "epoch": 1.0689102564102564, + "grad_norm": 0.5441751480102539, + "learning_rate": 0.0001669241555156314, + "loss": 0.8948, + "step": 6004 + }, + { + "epoch": 1.0690883190883191, + "grad_norm": 0.694199800491333, + "learning_rate": 0.00016691375410957324, + "loss": 1.0824, + "step": 6005 + }, + { + "epoch": 1.0692663817663817, + "grad_norm": 0.6077630519866943, + "learning_rate": 0.00016690335139247906, + "loss": 1.0931, + "step": 6006 + }, + { + "epoch": 1.0694444444444444, + "grad_norm": 0.6558539867401123, + "learning_rate": 0.0001668929473645527, + "loss": 1.0099, + "step": 6007 + }, + { + "epoch": 1.0696225071225072, + "grad_norm": 0.5722812414169312, + "learning_rate": 0.00016688254202599798, + "loss": 0.7999, + "step": 6008 + }, + { + "epoch": 1.0698005698005697, + "grad_norm": 0.5915400981903076, + "learning_rate": 0.0001668721353770188, + "loss": 0.7866, + "step": 6009 + }, + { + "epoch": 1.0699786324786325, + "grad_norm": 0.5290952324867249, + "learning_rate": 0.00016686172741781901, + "loss": 0.793, + "step": 6010 + }, + { + "epoch": 1.0701566951566952, + "grad_norm": 0.5501774549484253, + "learning_rate": 0.00016685131814860263, + "loss": 0.8775, + "step": 6011 + }, + { + "epoch": 1.0703347578347577, + "grad_norm": 0.6192594766616821, + "learning_rate": 0.00016684090756957347, + "loss": 1.1686, + "step": 6012 + }, + { + "epoch": 1.0705128205128205, + "grad_norm": 0.6640267968177795, + "learning_rate": 0.00016683049568093561, + "loss": 1.1789, + "step": 6013 + }, + { + "epoch": 1.0706908831908832, + "grad_norm": 0.552893877029419, + "learning_rate": 0.00016682008248289303, + "loss": 0.7957, + "step": 6014 + }, + { + "epoch": 1.0708689458689458, + "grad_norm": 0.6406302452087402, + "learning_rate": 0.00016680966797564972, + "loss": 1.1174, + "step": 6015 + }, + { + "epoch": 1.0710470085470085, + "grad_norm": Infinity, + "learning_rate": 0.00016680966797564972, + "loss": 0.9168, + "step": 6016 + }, + { + "epoch": 1.0712250712250713, + "grad_norm": 0.6384762525558472, + "learning_rate": 0.00016679925215940975, + "loss": 0.9831, + "step": 6017 + }, + { + "epoch": 1.071403133903134, + "grad_norm": 0.5906224846839905, + "learning_rate": 0.0001667888350343772, + "loss": 0.9167, + "step": 6018 + }, + { + "epoch": 1.0715811965811965, + "grad_norm": 0.658044695854187, + "learning_rate": 0.00016677841660075617, + "loss": 1.0075, + "step": 6019 + }, + { + "epoch": 1.0717592592592593, + "grad_norm": 0.6313242316246033, + "learning_rate": 0.00016676799685875078, + "loss": 0.8551, + "step": 6020 + }, + { + "epoch": 1.0719373219373218, + "grad_norm": 0.5891841053962708, + "learning_rate": 0.00016675757580856518, + "loss": 0.8475, + "step": 6021 + }, + { + "epoch": 1.0721153846153846, + "grad_norm": 0.581317126750946, + "learning_rate": 0.00016674715345040358, + "loss": 0.9308, + "step": 6022 + }, + { + "epoch": 1.0722934472934473, + "grad_norm": 0.5952537655830383, + "learning_rate": 0.00016673672978447017, + "loss": 0.9104, + "step": 6023 + }, + { + "epoch": 1.07247150997151, + "grad_norm": 0.5934227705001831, + "learning_rate": 0.00016672630481096915, + "loss": 0.9882, + "step": 6024 + }, + { + "epoch": 1.0726495726495726, + "grad_norm": 0.5867539048194885, + "learning_rate": 0.00016671587853010482, + "loss": 1.0186, + "step": 6025 + }, + { + "epoch": 1.0728276353276354, + "grad_norm": 0.6002280116081238, + "learning_rate": 0.00016670545094208143, + "loss": 0.92, + "step": 6026 + }, + { + "epoch": 1.073005698005698, + "grad_norm": 0.6261683702468872, + "learning_rate": 0.0001666950220471033, + "loss": 0.9293, + "step": 6027 + }, + { + "epoch": 1.0731837606837606, + "grad_norm": 0.6128147840499878, + "learning_rate": 0.00016668459184537477, + "loss": 1.0787, + "step": 6028 + }, + { + "epoch": 1.0733618233618234, + "grad_norm": 0.62148118019104, + "learning_rate": 0.00016667416033710016, + "loss": 0.8843, + "step": 6029 + }, + { + "epoch": 1.0735398860398861, + "grad_norm": 0.7166166305541992, + "learning_rate": 0.0001666637275224839, + "loss": 0.8877, + "step": 6030 + }, + { + "epoch": 1.0737179487179487, + "grad_norm": 0.5275574922561646, + "learning_rate": 0.0001666532934017304, + "loss": 0.9604, + "step": 6031 + }, + { + "epoch": 1.0738960113960114, + "grad_norm": 0.8132784962654114, + "learning_rate": 0.00016664285797504406, + "loss": 1.0203, + "step": 6032 + }, + { + "epoch": 1.074074074074074, + "grad_norm": 0.5887695550918579, + "learning_rate": 0.00016663242124262935, + "loss": 0.8819, + "step": 6033 + }, + { + "epoch": 1.0742521367521367, + "grad_norm": 0.5552900433540344, + "learning_rate": 0.00016662198320469078, + "loss": 0.7542, + "step": 6034 + }, + { + "epoch": 1.0744301994301995, + "grad_norm": 0.6228970885276794, + "learning_rate": 0.0001666115438614328, + "loss": 1.0362, + "step": 6035 + }, + { + "epoch": 1.0746082621082622, + "grad_norm": 0.7193471789360046, + "learning_rate": 0.00016660110321306003, + "loss": 1.3073, + "step": 6036 + }, + { + "epoch": 1.0747863247863247, + "grad_norm": 0.6167412996292114, + "learning_rate": 0.000166590661259777, + "loss": 0.941, + "step": 6037 + }, + { + "epoch": 1.0749643874643875, + "grad_norm": 0.5716922879219055, + "learning_rate": 0.00016658021800178827, + "loss": 0.83, + "step": 6038 + }, + { + "epoch": 1.0751424501424502, + "grad_norm": 0.6404047012329102, + "learning_rate": 0.00016656977343929848, + "loss": 1.0617, + "step": 6039 + }, + { + "epoch": 1.0753205128205128, + "grad_norm": 0.531395435333252, + "learning_rate": 0.00016655932757251226, + "loss": 0.7785, + "step": 6040 + }, + { + "epoch": 1.0754985754985755, + "grad_norm": 0.6468462347984314, + "learning_rate": 0.0001665488804016343, + "loss": 0.7893, + "step": 6041 + }, + { + "epoch": 1.0756766381766383, + "grad_norm": 0.6539653539657593, + "learning_rate": 0.00016653843192686925, + "loss": 1.1011, + "step": 6042 + }, + { + "epoch": 1.0758547008547008, + "grad_norm": 0.630107045173645, + "learning_rate": 0.0001665279821484219, + "loss": 0.9262, + "step": 6043 + }, + { + "epoch": 1.0760327635327636, + "grad_norm": 0.5875992774963379, + "learning_rate": 0.00016651753106649688, + "loss": 1.0501, + "step": 6044 + }, + { + "epoch": 1.0762108262108263, + "grad_norm": 0.573428750038147, + "learning_rate": 0.00016650707868129904, + "loss": 1.0672, + "step": 6045 + }, + { + "epoch": 1.0763888888888888, + "grad_norm": 0.6215469241142273, + "learning_rate": 0.00016649662499303316, + "loss": 0.868, + "step": 6046 + }, + { + "epoch": 1.0765669515669516, + "grad_norm": 0.6666893362998962, + "learning_rate": 0.00016648617000190402, + "loss": 1.0965, + "step": 6047 + }, + { + "epoch": 1.0767450142450143, + "grad_norm": 0.8343498706817627, + "learning_rate": 0.00016647571370811653, + "loss": 1.2302, + "step": 6048 + }, + { + "epoch": 1.0769230769230769, + "grad_norm": 0.591147780418396, + "learning_rate": 0.0001664652561118755, + "loss": 0.9698, + "step": 6049 + }, + { + "epoch": 1.0771011396011396, + "grad_norm": 0.573375940322876, + "learning_rate": 0.00016645479721338584, + "loss": 0.8798, + "step": 6050 + }, + { + "epoch": 1.0772792022792024, + "grad_norm": 0.4956737160682678, + "learning_rate": 0.00016644433701285246, + "loss": 0.6523, + "step": 6051 + }, + { + "epoch": 1.077457264957265, + "grad_norm": 0.6896619200706482, + "learning_rate": 0.00016643387551048034, + "loss": 0.8911, + "step": 6052 + }, + { + "epoch": 1.0776353276353277, + "grad_norm": 0.5820416808128357, + "learning_rate": 0.00016642341270647445, + "loss": 1.1486, + "step": 6053 + }, + { + "epoch": 1.0778133903133904, + "grad_norm": 0.611132025718689, + "learning_rate": 0.00016641294860103976, + "loss": 1.0705, + "step": 6054 + }, + { + "epoch": 1.077991452991453, + "grad_norm": 0.6705698370933533, + "learning_rate": 0.00016640248319438133, + "loss": 0.9826, + "step": 6055 + }, + { + "epoch": 1.0781695156695157, + "grad_norm": 0.5987013578414917, + "learning_rate": 0.00016639201648670416, + "loss": 1.0409, + "step": 6056 + }, + { + "epoch": 1.0783475783475784, + "grad_norm": 0.6707149744033813, + "learning_rate": 0.00016638154847821332, + "loss": 1.1332, + "step": 6057 + }, + { + "epoch": 1.078525641025641, + "grad_norm": 0.6400678157806396, + "learning_rate": 0.00016637107916911393, + "loss": 1.2559, + "step": 6058 + }, + { + "epoch": 1.0787037037037037, + "grad_norm": 0.6370311379432678, + "learning_rate": 0.00016636060855961115, + "loss": 0.9752, + "step": 6059 + }, + { + "epoch": 1.0788817663817665, + "grad_norm": 0.6116052269935608, + "learning_rate": 0.00016635013664991012, + "loss": 0.8364, + "step": 6060 + }, + { + "epoch": 1.079059829059829, + "grad_norm": 0.7932127714157104, + "learning_rate": 0.00016633966344021593, + "loss": 0.939, + "step": 6061 + }, + { + "epoch": 1.0792378917378918, + "grad_norm": 0.576249897480011, + "learning_rate": 0.00016632918893073385, + "loss": 0.8911, + "step": 6062 + }, + { + "epoch": 1.0794159544159545, + "grad_norm": 0.5456888675689697, + "learning_rate": 0.00016631871312166915, + "loss": 0.8646, + "step": 6063 + }, + { + "epoch": 1.079594017094017, + "grad_norm": 0.717522919178009, + "learning_rate": 0.000166308236013227, + "loss": 1.0814, + "step": 6064 + }, + { + "epoch": 1.0797720797720798, + "grad_norm": 0.6637256145477295, + "learning_rate": 0.0001662977576056127, + "loss": 1.22, + "step": 6065 + }, + { + "epoch": 1.0799501424501425, + "grad_norm": 0.5846666693687439, + "learning_rate": 0.0001662872778990316, + "loss": 1.1745, + "step": 6066 + }, + { + "epoch": 1.080128205128205, + "grad_norm": 0.6611326336860657, + "learning_rate": 0.00016627679689368895, + "loss": 1.1262, + "step": 6067 + }, + { + "epoch": 1.0803062678062678, + "grad_norm": 0.6022892594337463, + "learning_rate": 0.00016626631458979015, + "loss": 0.9741, + "step": 6068 + }, + { + "epoch": 1.0804843304843306, + "grad_norm": 0.5862685441970825, + "learning_rate": 0.00016625583098754058, + "loss": 0.914, + "step": 6069 + }, + { + "epoch": 1.080662393162393, + "grad_norm": 0.7089241147041321, + "learning_rate": 0.00016624534608714563, + "loss": 1.0614, + "step": 6070 + }, + { + "epoch": 1.0808404558404558, + "grad_norm": 0.5286028981208801, + "learning_rate": 0.00016623485988881076, + "loss": 0.8756, + "step": 6071 + }, + { + "epoch": 1.0810185185185186, + "grad_norm": 0.6437101364135742, + "learning_rate": 0.00016622437239274137, + "loss": 0.7222, + "step": 6072 + }, + { + "epoch": 1.0811965811965811, + "grad_norm": 0.6197740435600281, + "learning_rate": 0.000166213883599143, + "loss": 0.7876, + "step": 6073 + }, + { + "epoch": 1.0813746438746439, + "grad_norm": 0.5889328122138977, + "learning_rate": 0.0001662033935082211, + "loss": 0.9587, + "step": 6074 + }, + { + "epoch": 1.0815527065527066, + "grad_norm": 0.5353847742080688, + "learning_rate": 0.00016619290212018125, + "loss": 0.8664, + "step": 6075 + }, + { + "epoch": 1.0817307692307692, + "grad_norm": 0.7202061414718628, + "learning_rate": 0.00016618240943522898, + "loss": 1.0429, + "step": 6076 + }, + { + "epoch": 1.081908831908832, + "grad_norm": 0.5831515192985535, + "learning_rate": 0.0001661719154535699, + "loss": 1.0323, + "step": 6077 + }, + { + "epoch": 1.0820868945868947, + "grad_norm": 0.6270500421524048, + "learning_rate": 0.00016616142017540953, + "loss": 0.9272, + "step": 6078 + }, + { + "epoch": 1.0822649572649572, + "grad_norm": 0.6064695119857788, + "learning_rate": 0.00016615092360095364, + "loss": 1.0629, + "step": 6079 + }, + { + "epoch": 1.08244301994302, + "grad_norm": 0.5578122138977051, + "learning_rate": 0.00016614042573040777, + "loss": 0.8601, + "step": 6080 + }, + { + "epoch": 1.0826210826210827, + "grad_norm": 0.5920688509941101, + "learning_rate": 0.0001661299265639777, + "loss": 1.0082, + "step": 6081 + }, + { + "epoch": 1.0827991452991452, + "grad_norm": 0.6191682815551758, + "learning_rate": 0.0001661194261018691, + "loss": 0.9645, + "step": 6082 + }, + { + "epoch": 1.082977207977208, + "grad_norm": 0.6403279304504395, + "learning_rate": 0.00016610892434428765, + "loss": 0.9263, + "step": 6083 + }, + { + "epoch": 1.0831552706552707, + "grad_norm": 0.579502284526825, + "learning_rate": 0.00016609842129143915, + "loss": 0.8997, + "step": 6084 + }, + { + "epoch": 1.0833333333333333, + "grad_norm": 0.5831437706947327, + "learning_rate": 0.00016608791694352944, + "loss": 1.0703, + "step": 6085 + }, + { + "epoch": 1.083511396011396, + "grad_norm": 0.6188452243804932, + "learning_rate": 0.00016607741130076424, + "loss": 0.8856, + "step": 6086 + }, + { + "epoch": 1.0836894586894588, + "grad_norm": 0.7413692474365234, + "learning_rate": 0.00016606690436334946, + "loss": 1.1995, + "step": 6087 + }, + { + "epoch": 1.0838675213675213, + "grad_norm": 0.5552099347114563, + "learning_rate": 0.00016605639613149093, + "loss": 0.8514, + "step": 6088 + }, + { + "epoch": 1.084045584045584, + "grad_norm": 0.5906503200531006, + "learning_rate": 0.00016604588660539452, + "loss": 0.9431, + "step": 6089 + }, + { + "epoch": 1.0842236467236468, + "grad_norm": 0.5326111316680908, + "learning_rate": 0.0001660353757852662, + "loss": 0.8306, + "step": 6090 + }, + { + "epoch": 1.0844017094017093, + "grad_norm": 0.7273091673851013, + "learning_rate": 0.0001660248636713118, + "loss": 1.1109, + "step": 6091 + }, + { + "epoch": 1.084579772079772, + "grad_norm": 0.66513592004776, + "learning_rate": 0.00016601435026373737, + "loss": 1.0621, + "step": 6092 + }, + { + "epoch": 1.0847578347578348, + "grad_norm": 0.6470831632614136, + "learning_rate": 0.00016600383556274892, + "loss": 1.1075, + "step": 6093 + }, + { + "epoch": 1.0849358974358974, + "grad_norm": 0.6308658719062805, + "learning_rate": 0.0001659933195685524, + "loss": 0.9832, + "step": 6094 + }, + { + "epoch": 1.08511396011396, + "grad_norm": 0.6569336652755737, + "learning_rate": 0.00016598280228135388, + "loss": 0.9754, + "step": 6095 + }, + { + "epoch": 1.0852920227920229, + "grad_norm": 0.5672318339347839, + "learning_rate": 0.0001659722837013594, + "loss": 0.9075, + "step": 6096 + }, + { + "epoch": 1.0854700854700854, + "grad_norm": 0.6397247314453125, + "learning_rate": 0.00016596176382877506, + "loss": 1.0358, + "step": 6097 + }, + { + "epoch": 1.0856481481481481, + "grad_norm": 0.6046154499053955, + "learning_rate": 0.000165951242663807, + "loss": 0.9036, + "step": 6098 + }, + { + "epoch": 1.085826210826211, + "grad_norm": 0.7190790176391602, + "learning_rate": 0.00016594072020666134, + "loss": 1.05, + "step": 6099 + }, + { + "epoch": 1.0860042735042734, + "grad_norm": 0.636986255645752, + "learning_rate": 0.00016593019645754425, + "loss": 1.0648, + "step": 6100 + }, + { + "epoch": 1.0861823361823362, + "grad_norm": 0.7239426374435425, + "learning_rate": 0.00016591967141666193, + "loss": 1.3332, + "step": 6101 + }, + { + "epoch": 1.086360398860399, + "grad_norm": 0.5623281002044678, + "learning_rate": 0.00016590914508422054, + "loss": 0.997, + "step": 6102 + }, + { + "epoch": 1.0865384615384615, + "grad_norm": 0.5559574365615845, + "learning_rate": 0.00016589861746042642, + "loss": 0.9309, + "step": 6103 + }, + { + "epoch": 1.0867165242165242, + "grad_norm": 0.6056998372077942, + "learning_rate": 0.00016588808854548574, + "loss": 1.05, + "step": 6104 + }, + { + "epoch": 1.086894586894587, + "grad_norm": 0.6419603228569031, + "learning_rate": 0.00016587755833960487, + "loss": 0.8933, + "step": 6105 + }, + { + "epoch": 1.0870726495726495, + "grad_norm": 0.5236496329307556, + "learning_rate": 0.00016586702684299006, + "loss": 1.0061, + "step": 6106 + }, + { + "epoch": 1.0872507122507122, + "grad_norm": 0.5764613747596741, + "learning_rate": 0.0001658564940558477, + "loss": 1.0218, + "step": 6107 + }, + { + "epoch": 1.087428774928775, + "grad_norm": 0.6049391627311707, + "learning_rate": 0.00016584595997838416, + "loss": 0.8157, + "step": 6108 + }, + { + "epoch": 1.0876068376068375, + "grad_norm": 0.585422933101654, + "learning_rate": 0.0001658354246108058, + "loss": 1.2761, + "step": 6109 + }, + { + "epoch": 1.0877849002849003, + "grad_norm": 0.6420125365257263, + "learning_rate": 0.00016582488795331907, + "loss": 1.1978, + "step": 6110 + }, + { + "epoch": 1.087962962962963, + "grad_norm": 0.646091878414154, + "learning_rate": 0.00016581435000613038, + "loss": 0.8946, + "step": 6111 + }, + { + "epoch": 1.0881410256410255, + "grad_norm": 0.6563934087753296, + "learning_rate": 0.00016580381076944625, + "loss": 1.0625, + "step": 6112 + }, + { + "epoch": 1.0883190883190883, + "grad_norm": 0.6796613931655884, + "learning_rate": 0.0001657932702434731, + "loss": 0.9401, + "step": 6113 + }, + { + "epoch": 1.088497150997151, + "grad_norm": 0.6248648762702942, + "learning_rate": 0.00016578272842841753, + "loss": 0.8558, + "step": 6114 + }, + { + "epoch": 1.0886752136752136, + "grad_norm": 0.5136269330978394, + "learning_rate": 0.00016577218532448605, + "loss": 0.6424, + "step": 6115 + }, + { + "epoch": 1.0888532763532763, + "grad_norm": 0.5581641793251038, + "learning_rate": 0.00016576164093188523, + "loss": 0.7923, + "step": 6116 + }, + { + "epoch": 1.089031339031339, + "grad_norm": 0.630352258682251, + "learning_rate": 0.0001657510952508216, + "loss": 0.9115, + "step": 6117 + }, + { + "epoch": 1.0892094017094016, + "grad_norm": 0.6167593002319336, + "learning_rate": 0.0001657405482815019, + "loss": 1.1112, + "step": 6118 + }, + { + "epoch": 1.0893874643874644, + "grad_norm": 0.5908578634262085, + "learning_rate": 0.00016573000002413271, + "loss": 1.0359, + "step": 6119 + }, + { + "epoch": 1.0895655270655271, + "grad_norm": 0.6326140761375427, + "learning_rate": 0.00016571945047892073, + "loss": 1.0459, + "step": 6120 + }, + { + "epoch": 1.0897435897435896, + "grad_norm": 0.7273572683334351, + "learning_rate": 0.00016570889964607262, + "loss": 1.0901, + "step": 6121 + }, + { + "epoch": 1.0899216524216524, + "grad_norm": 0.6168062090873718, + "learning_rate": 0.00016569834752579513, + "loss": 0.8739, + "step": 6122 + }, + { + "epoch": 1.0900997150997151, + "grad_norm": 0.5620378255844116, + "learning_rate": 0.00016568779411829497, + "loss": 0.9614, + "step": 6123 + }, + { + "epoch": 1.0902777777777777, + "grad_norm": 0.6319156885147095, + "learning_rate": 0.00016567723942377899, + "loss": 1.1031, + "step": 6124 + }, + { + "epoch": 1.0904558404558404, + "grad_norm": 0.6590072512626648, + "learning_rate": 0.00016566668344245388, + "loss": 1.0086, + "step": 6125 + }, + { + "epoch": 1.0906339031339032, + "grad_norm": 0.5823387503623962, + "learning_rate": 0.00016565612617452656, + "loss": 0.8886, + "step": 6126 + }, + { + "epoch": 1.0908119658119657, + "grad_norm": 0.5795989632606506, + "learning_rate": 0.00016564556762020381, + "loss": 0.7683, + "step": 6127 + }, + { + "epoch": 1.0909900284900285, + "grad_norm": 0.5940101742744446, + "learning_rate": 0.00016563500777969255, + "loss": 0.8873, + "step": 6128 + }, + { + "epoch": 1.0911680911680912, + "grad_norm": 0.5708247423171997, + "learning_rate": 0.00016562444665319963, + "loss": 0.7382, + "step": 6129 + }, + { + "epoch": 1.0913461538461537, + "grad_norm": 0.6339239478111267, + "learning_rate": 0.00016561388424093202, + "loss": 0.9323, + "step": 6130 + }, + { + "epoch": 1.0915242165242165, + "grad_norm": 0.720000147819519, + "learning_rate": 0.00016560332054309663, + "loss": 1.0437, + "step": 6131 + }, + { + "epoch": 1.0917022792022792, + "grad_norm": 0.686580240726471, + "learning_rate": 0.00016559275555990048, + "loss": 0.9841, + "step": 6132 + }, + { + "epoch": 1.091880341880342, + "grad_norm": 0.6067900061607361, + "learning_rate": 0.00016558218929155053, + "loss": 1.0862, + "step": 6133 + }, + { + "epoch": 1.0920584045584045, + "grad_norm": 0.6678896546363831, + "learning_rate": 0.00016557162173825384, + "loss": 0.8509, + "step": 6134 + }, + { + "epoch": 1.0922364672364673, + "grad_norm": 0.53044193983078, + "learning_rate": 0.0001655610529002174, + "loss": 0.9227, + "step": 6135 + }, + { + "epoch": 1.0924145299145298, + "grad_norm": 0.6499412655830383, + "learning_rate": 0.00016555048277764836, + "loss": 1.0867, + "step": 6136 + }, + { + "epoch": 1.0925925925925926, + "grad_norm": 0.6543099284172058, + "learning_rate": 0.00016553991137075374, + "loss": 0.849, + "step": 6137 + }, + { + "epoch": 1.0927706552706553, + "grad_norm": 0.5772737860679626, + "learning_rate": 0.0001655293386797407, + "loss": 0.8475, + "step": 6138 + }, + { + "epoch": 1.092948717948718, + "grad_norm": 0.616348385810852, + "learning_rate": 0.00016551876470481642, + "loss": 0.9205, + "step": 6139 + }, + { + "epoch": 1.0931267806267806, + "grad_norm": 0.7151142954826355, + "learning_rate": 0.00016550818944618801, + "loss": 1.1389, + "step": 6140 + }, + { + "epoch": 1.0933048433048433, + "grad_norm": 0.6566469669342041, + "learning_rate": 0.00016549761290406275, + "loss": 0.8216, + "step": 6141 + }, + { + "epoch": 1.0934829059829059, + "grad_norm": 0.7075428366661072, + "learning_rate": 0.00016548703507864783, + "loss": 1.065, + "step": 6142 + }, + { + "epoch": 1.0936609686609686, + "grad_norm": 0.6589360237121582, + "learning_rate": 0.00016547645597015046, + "loss": 0.9899, + "step": 6143 + }, + { + "epoch": 1.0938390313390314, + "grad_norm": 0.6445585489273071, + "learning_rate": 0.00016546587557877797, + "loss": 1.1629, + "step": 6144 + }, + { + "epoch": 1.0940170940170941, + "grad_norm": 0.6216462850570679, + "learning_rate": 0.00016545529390473763, + "loss": 0.9685, + "step": 6145 + }, + { + "epoch": 1.0941951566951567, + "grad_norm": 0.6195303797721863, + "learning_rate": 0.0001654447109482368, + "loss": 1.144, + "step": 6146 + }, + { + "epoch": 1.0943732193732194, + "grad_norm": 0.6625444293022156, + "learning_rate": 0.0001654341267094828, + "loss": 0.9886, + "step": 6147 + }, + { + "epoch": 1.094551282051282, + "grad_norm": 0.6449851393699646, + "learning_rate": 0.000165423541188683, + "loss": 0.9568, + "step": 6148 + }, + { + "epoch": 1.0947293447293447, + "grad_norm": 0.6490375995635986, + "learning_rate": 0.00016541295438604484, + "loss": 1.1304, + "step": 6149 + }, + { + "epoch": 1.0949074074074074, + "grad_norm": 0.6771987676620483, + "learning_rate": 0.00016540236630177574, + "loss": 1.0426, + "step": 6150 + }, + { + "epoch": 1.0950854700854702, + "grad_norm": 0.5214568376541138, + "learning_rate": 0.00016539177693608307, + "loss": 0.6742, + "step": 6151 + }, + { + "epoch": 1.0952635327635327, + "grad_norm": 0.6005097031593323, + "learning_rate": 0.00016538118628917442, + "loss": 0.9901, + "step": 6152 + }, + { + "epoch": 1.0954415954415955, + "grad_norm": 0.6449539065361023, + "learning_rate": 0.0001653705943612572, + "loss": 0.9654, + "step": 6153 + }, + { + "epoch": 1.095619658119658, + "grad_norm": 0.6443646550178528, + "learning_rate": 0.00016536000115253903, + "loss": 0.9084, + "step": 6154 + }, + { + "epoch": 1.0957977207977208, + "grad_norm": 0.6072495579719543, + "learning_rate": 0.0001653494066632274, + "loss": 0.6308, + "step": 6155 + }, + { + "epoch": 1.0959757834757835, + "grad_norm": 0.5751157999038696, + "learning_rate": 0.00016533881089352988, + "loss": 0.96, + "step": 6156 + }, + { + "epoch": 1.0961538461538463, + "grad_norm": 0.6310713291168213, + "learning_rate": 0.0001653282138436541, + "loss": 1.0997, + "step": 6157 + }, + { + "epoch": 1.0963319088319088, + "grad_norm": 0.5573651790618896, + "learning_rate": 0.00016531761551380765, + "loss": 0.9738, + "step": 6158 + }, + { + "epoch": 1.0965099715099715, + "grad_norm": 0.5615308880805969, + "learning_rate": 0.00016530701590419824, + "loss": 0.9658, + "step": 6159 + }, + { + "epoch": 1.0966880341880343, + "grad_norm": 0.6471942663192749, + "learning_rate": 0.0001652964150150335, + "loss": 1.0763, + "step": 6160 + }, + { + "epoch": 1.0968660968660968, + "grad_norm": 0.6305427551269531, + "learning_rate": 0.00016528581284652117, + "loss": 1.112, + "step": 6161 + }, + { + "epoch": 1.0970441595441596, + "grad_norm": 0.6881145238876343, + "learning_rate": 0.00016527520939886892, + "loss": 0.8476, + "step": 6162 + }, + { + "epoch": 1.0972222222222223, + "grad_norm": 0.6507891416549683, + "learning_rate": 0.00016526460467228458, + "loss": 1.1097, + "step": 6163 + }, + { + "epoch": 1.0974002849002849, + "grad_norm": 0.5960137844085693, + "learning_rate": 0.00016525399866697586, + "loss": 0.9934, + "step": 6164 + }, + { + "epoch": 1.0975783475783476, + "grad_norm": 0.6001808643341064, + "learning_rate": 0.0001652433913831506, + "loss": 1.0782, + "step": 6165 + }, + { + "epoch": 1.0977564102564104, + "grad_norm": 0.5639005303382874, + "learning_rate": 0.00016523278282101663, + "loss": 1.0929, + "step": 6166 + }, + { + "epoch": 1.0979344729344729, + "grad_norm": 0.5962058305740356, + "learning_rate": 0.00016522217298078177, + "loss": 1.0315, + "step": 6167 + }, + { + "epoch": 1.0981125356125356, + "grad_norm": 0.6920329928398132, + "learning_rate": 0.0001652115618626539, + "loss": 0.9176, + "step": 6168 + }, + { + "epoch": 1.0982905982905984, + "grad_norm": 0.6963527202606201, + "learning_rate": 0.00016520094946684098, + "loss": 1.2136, + "step": 6169 + }, + { + "epoch": 1.098468660968661, + "grad_norm": 0.5855711102485657, + "learning_rate": 0.00016519033579355093, + "loss": 0.8453, + "step": 6170 + }, + { + "epoch": 1.0986467236467237, + "grad_norm": 0.6454927325248718, + "learning_rate": 0.0001651797208429916, + "loss": 1.0747, + "step": 6171 + }, + { + "epoch": 1.0988247863247864, + "grad_norm": 0.644585907459259, + "learning_rate": 0.00016516910461537108, + "loss": 0.8165, + "step": 6172 + }, + { + "epoch": 1.099002849002849, + "grad_norm": 0.6488069891929626, + "learning_rate": 0.00016515848711089732, + "loss": 1.1048, + "step": 6173 + }, + { + "epoch": 1.0991809116809117, + "grad_norm": 0.5867953896522522, + "learning_rate": 0.00016514786832977834, + "loss": 0.63, + "step": 6174 + }, + { + "epoch": 1.0993589743589745, + "grad_norm": 0.560591459274292, + "learning_rate": 0.00016513724827222227, + "loss": 0.9255, + "step": 6175 + }, + { + "epoch": 1.099537037037037, + "grad_norm": 0.675262451171875, + "learning_rate": 0.00016512662693843707, + "loss": 0.7637, + "step": 6176 + }, + { + "epoch": 1.0997150997150997, + "grad_norm": 0.6515669822692871, + "learning_rate": 0.00016511600432863091, + "loss": 0.7579, + "step": 6177 + }, + { + "epoch": 1.0998931623931625, + "grad_norm": 0.683409571647644, + "learning_rate": 0.00016510538044301192, + "loss": 0.9183, + "step": 6178 + }, + { + "epoch": 1.100071225071225, + "grad_norm": 0.6194507479667664, + "learning_rate": 0.00016509475528178827, + "loss": 1.16, + "step": 6179 + }, + { + "epoch": 1.1002492877492878, + "grad_norm": 0.6192209720611572, + "learning_rate": 0.0001650841288451681, + "loss": 1.1392, + "step": 6180 + }, + { + "epoch": 1.1004273504273505, + "grad_norm": 0.6029189825057983, + "learning_rate": 0.0001650735011333596, + "loss": 1.1453, + "step": 6181 + }, + { + "epoch": 1.100605413105413, + "grad_norm": 0.7040731310844421, + "learning_rate": 0.00016506287214657105, + "loss": 0.9367, + "step": 6182 + }, + { + "epoch": 1.1007834757834758, + "grad_norm": 0.5909842252731323, + "learning_rate": 0.00016505224188501067, + "loss": 0.6463, + "step": 6183 + }, + { + "epoch": 1.1009615384615385, + "grad_norm": 0.6129698157310486, + "learning_rate": 0.00016504161034888674, + "loss": 0.9432, + "step": 6184 + }, + { + "epoch": 1.101139601139601, + "grad_norm": 0.6181607842445374, + "learning_rate": 0.00016503097753840757, + "loss": 0.9934, + "step": 6185 + }, + { + "epoch": 1.1013176638176638, + "grad_norm": 0.6463226675987244, + "learning_rate": 0.0001650203434537815, + "loss": 0.8471, + "step": 6186 + }, + { + "epoch": 1.1014957264957266, + "grad_norm": 0.5999348163604736, + "learning_rate": 0.00016500970809521688, + "loss": 0.9418, + "step": 6187 + }, + { + "epoch": 1.101673789173789, + "grad_norm": 0.629504919052124, + "learning_rate": 0.00016499907146292204, + "loss": 0.9699, + "step": 6188 + }, + { + "epoch": 1.1018518518518519, + "grad_norm": 0.694767951965332, + "learning_rate": 0.00016498843355710542, + "loss": 0.8793, + "step": 6189 + }, + { + "epoch": 1.1020299145299146, + "grad_norm": 0.6205509901046753, + "learning_rate": 0.00016497779437797547, + "loss": 0.8384, + "step": 6190 + }, + { + "epoch": 1.1022079772079771, + "grad_norm": 0.6256579756736755, + "learning_rate": 0.0001649671539257406, + "loss": 0.9275, + "step": 6191 + }, + { + "epoch": 1.10238603988604, + "grad_norm": 0.6593793034553528, + "learning_rate": 0.00016495651220060933, + "loss": 1.0495, + "step": 6192 + }, + { + "epoch": 1.1025641025641026, + "grad_norm": 0.7809221148490906, + "learning_rate": 0.00016494586920279012, + "loss": 1.0485, + "step": 6193 + }, + { + "epoch": 1.1027421652421652, + "grad_norm": 0.6147717833518982, + "learning_rate": 0.0001649352249324915, + "loss": 0.8739, + "step": 6194 + }, + { + "epoch": 1.102920227920228, + "grad_norm": 0.565411388874054, + "learning_rate": 0.00016492457938992208, + "loss": 0.9759, + "step": 6195 + }, + { + "epoch": 1.1030982905982907, + "grad_norm": 0.596370279788971, + "learning_rate": 0.00016491393257529036, + "loss": 0.9658, + "step": 6196 + }, + { + "epoch": 1.1032763532763532, + "grad_norm": 0.6334326863288879, + "learning_rate": 0.00016490328448880498, + "loss": 0.8785, + "step": 6197 + }, + { + "epoch": 1.103454415954416, + "grad_norm": 0.5538334846496582, + "learning_rate": 0.0001648926351306746, + "loss": 0.7174, + "step": 6198 + }, + { + "epoch": 1.1036324786324787, + "grad_norm": 0.6249658465385437, + "learning_rate": 0.00016488198450110778, + "loss": 0.8579, + "step": 6199 + }, + { + "epoch": 1.1038105413105412, + "grad_norm": 0.6128895878791809, + "learning_rate": 0.00016487133260031329, + "loss": 0.8538, + "step": 6200 + }, + { + "epoch": 1.103988603988604, + "grad_norm": 0.5808702707290649, + "learning_rate": 0.0001648606794284998, + "loss": 0.8143, + "step": 6201 + }, + { + "epoch": 1.1041666666666667, + "grad_norm": 0.671419084072113, + "learning_rate": 0.00016485002498587602, + "loss": 1.1268, + "step": 6202 + }, + { + "epoch": 1.1043447293447293, + "grad_norm": 0.5706788897514343, + "learning_rate": 0.00016483936927265075, + "loss": 0.9558, + "step": 6203 + }, + { + "epoch": 1.104522792022792, + "grad_norm": 0.5700307488441467, + "learning_rate": 0.00016482871228903266, + "loss": 0.9616, + "step": 6204 + }, + { + "epoch": 1.1047008547008548, + "grad_norm": 0.5764816403388977, + "learning_rate": 0.0001648180540352307, + "loss": 0.8692, + "step": 6205 + }, + { + "epoch": 1.1048789173789173, + "grad_norm": 0.5786563754081726, + "learning_rate": 0.00016480739451145358, + "loss": 0.9406, + "step": 6206 + }, + { + "epoch": 1.10505698005698, + "grad_norm": 0.6112591624259949, + "learning_rate": 0.0001647967337179102, + "loss": 0.8999, + "step": 6207 + }, + { + "epoch": 1.1052350427350428, + "grad_norm": 0.5708907246589661, + "learning_rate": 0.00016478607165480944, + "loss": 0.9236, + "step": 6208 + }, + { + "epoch": 1.1054131054131053, + "grad_norm": 0.6742013692855835, + "learning_rate": 0.00016477540832236014, + "loss": 1.0911, + "step": 6209 + }, + { + "epoch": 1.105591168091168, + "grad_norm": 0.6382617354393005, + "learning_rate": 0.0001647647437207713, + "loss": 0.7901, + "step": 6210 + }, + { + "epoch": 1.1057692307692308, + "grad_norm": 0.6241547465324402, + "learning_rate": 0.00016475407785025188, + "loss": 1.0048, + "step": 6211 + }, + { + "epoch": 1.1059472934472934, + "grad_norm": 0.6452877521514893, + "learning_rate": 0.00016474341071101077, + "loss": 0.8902, + "step": 6212 + }, + { + "epoch": 1.1061253561253561, + "grad_norm": 0.6212326288223267, + "learning_rate": 0.00016473274230325704, + "loss": 1.078, + "step": 6213 + }, + { + "epoch": 1.1063034188034189, + "grad_norm": 0.6870912909507751, + "learning_rate": 0.00016472207262719968, + "loss": 0.9127, + "step": 6214 + }, + { + "epoch": 1.1064814814814814, + "grad_norm": 0.6286750435829163, + "learning_rate": 0.00016471140168304777, + "loss": 1.0271, + "step": 6215 + }, + { + "epoch": 1.1066595441595442, + "grad_norm": 0.645806074142456, + "learning_rate": 0.00016470072947101036, + "loss": 1.1514, + "step": 6216 + }, + { + "epoch": 1.106837606837607, + "grad_norm": 0.6800320148468018, + "learning_rate": 0.00016469005599129653, + "loss": 0.9322, + "step": 6217 + }, + { + "epoch": 1.1070156695156694, + "grad_norm": 0.5898309946060181, + "learning_rate": 0.0001646793812441155, + "loss": 1.065, + "step": 6218 + }, + { + "epoch": 1.1071937321937322, + "grad_norm": 0.6000019907951355, + "learning_rate": 0.00016466870522967634, + "loss": 0.911, + "step": 6219 + }, + { + "epoch": 1.107371794871795, + "grad_norm": 0.6164331436157227, + "learning_rate": 0.0001646580279481882, + "loss": 0.8421, + "step": 6220 + }, + { + "epoch": 1.1075498575498575, + "grad_norm": 0.6410242319107056, + "learning_rate": 0.00016464734939986036, + "loss": 0.9688, + "step": 6221 + }, + { + "epoch": 1.1077279202279202, + "grad_norm": 0.7153300046920776, + "learning_rate": 0.00016463666958490197, + "loss": 1.0722, + "step": 6222 + }, + { + "epoch": 1.107905982905983, + "grad_norm": 0.6977026462554932, + "learning_rate": 0.00016462598850352234, + "loss": 1.0192, + "step": 6223 + }, + { + "epoch": 1.1080840455840455, + "grad_norm": 0.6379461884498596, + "learning_rate": 0.0001646153061559307, + "loss": 1.0474, + "step": 6224 + }, + { + "epoch": 1.1082621082621082, + "grad_norm": 0.6135090589523315, + "learning_rate": 0.00016460462254233634, + "loss": 1.0082, + "step": 6225 + }, + { + "epoch": 1.108440170940171, + "grad_norm": 0.6326230764389038, + "learning_rate": 0.00016459393766294866, + "loss": 1.1097, + "step": 6226 + }, + { + "epoch": 1.1086182336182335, + "grad_norm": 0.6636839509010315, + "learning_rate": 0.0001645832515179769, + "loss": 0.9689, + "step": 6227 + }, + { + "epoch": 1.1087962962962963, + "grad_norm": 0.5713129043579102, + "learning_rate": 0.00016457256410763052, + "loss": 0.8642, + "step": 6228 + }, + { + "epoch": 1.108974358974359, + "grad_norm": 0.584204912185669, + "learning_rate": 0.00016456187543211888, + "loss": 0.9957, + "step": 6229 + }, + { + "epoch": 1.1091524216524216, + "grad_norm": 0.5920230746269226, + "learning_rate": 0.0001645511854916514, + "loss": 0.7297, + "step": 6230 + }, + { + "epoch": 1.1093304843304843, + "grad_norm": 0.6207385063171387, + "learning_rate": 0.0001645404942864375, + "loss": 0.868, + "step": 6231 + }, + { + "epoch": 1.109508547008547, + "grad_norm": 0.7267234921455383, + "learning_rate": 0.00016452980181668673, + "loss": 1.0248, + "step": 6232 + }, + { + "epoch": 1.1096866096866096, + "grad_norm": 0.5925650596618652, + "learning_rate": 0.00016451910808260852, + "loss": 1.1075, + "step": 6233 + }, + { + "epoch": 1.1098646723646723, + "grad_norm": 0.5632196664810181, + "learning_rate": 0.00016450841308441244, + "loss": 0.9865, + "step": 6234 + }, + { + "epoch": 1.110042735042735, + "grad_norm": 0.6115161180496216, + "learning_rate": 0.000164497716822308, + "loss": 1.1343, + "step": 6235 + }, + { + "epoch": 1.1102207977207976, + "grad_norm": 0.634398341178894, + "learning_rate": 0.00016448701929650477, + "loss": 1.1039, + "step": 6236 + }, + { + "epoch": 1.1103988603988604, + "grad_norm": 0.5843468308448792, + "learning_rate": 0.00016447632050721237, + "loss": 0.8462, + "step": 6237 + }, + { + "epoch": 1.1105769230769231, + "grad_norm": 0.799375593662262, + "learning_rate": 0.0001644656204546404, + "loss": 0.9861, + "step": 6238 + }, + { + "epoch": 1.1107549857549857, + "grad_norm": 0.600289523601532, + "learning_rate": 0.0001644549191389985, + "loss": 1.0323, + "step": 6239 + }, + { + "epoch": 1.1109330484330484, + "grad_norm": 0.6154919266700745, + "learning_rate": 0.00016444421656049637, + "loss": 0.9158, + "step": 6240 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 0.6685689687728882, + "learning_rate": 0.00016443351271934367, + "loss": 1.0429, + "step": 6241 + }, + { + "epoch": 1.1112891737891737, + "grad_norm": 0.699978232383728, + "learning_rate": 0.00016442280761575016, + "loss": 1.072, + "step": 6242 + }, + { + "epoch": 1.1114672364672364, + "grad_norm": 0.6461396217346191, + "learning_rate": 0.00016441210124992556, + "loss": 0.9758, + "step": 6243 + }, + { + "epoch": 1.1116452991452992, + "grad_norm": 0.6463284492492676, + "learning_rate": 0.00016440139362207962, + "loss": 0.9205, + "step": 6244 + }, + { + "epoch": 1.1118233618233617, + "grad_norm": 0.6587556004524231, + "learning_rate": 0.00016439068473242217, + "loss": 1.0027, + "step": 6245 + }, + { + "epoch": 1.1120014245014245, + "grad_norm": 0.6896520256996155, + "learning_rate": 0.000164379974581163, + "loss": 0.9788, + "step": 6246 + }, + { + "epoch": 1.1121794871794872, + "grad_norm": 0.6766142845153809, + "learning_rate": 0.000164369263168512, + "loss": 0.9647, + "step": 6247 + }, + { + "epoch": 1.1123575498575498, + "grad_norm": 0.7024297118186951, + "learning_rate": 0.00016435855049467898, + "loss": 1.1163, + "step": 6248 + }, + { + "epoch": 1.1125356125356125, + "grad_norm": 0.6654963493347168, + "learning_rate": 0.00016434783655987385, + "loss": 0.9302, + "step": 6249 + }, + { + "epoch": 1.1127136752136753, + "grad_norm": 0.6973692774772644, + "learning_rate": 0.0001643371213643065, + "loss": 0.9585, + "step": 6250 + }, + { + "epoch": 1.1128917378917378, + "grad_norm": 0.7153545022010803, + "learning_rate": 0.000164326404908187, + "loss": 1.0485, + "step": 6251 + }, + { + "epoch": 1.1130698005698005, + "grad_norm": 0.6114685535430908, + "learning_rate": 0.00016431568719172516, + "loss": 0.8881, + "step": 6252 + }, + { + "epoch": 1.1132478632478633, + "grad_norm": 0.6500731706619263, + "learning_rate": 0.00016430496821513103, + "loss": 1.0658, + "step": 6253 + }, + { + "epoch": 1.113425925925926, + "grad_norm": 0.5800092220306396, + "learning_rate": 0.00016429424797861466, + "loss": 0.9158, + "step": 6254 + }, + { + "epoch": 1.1136039886039886, + "grad_norm": 0.6653759479522705, + "learning_rate": 0.00016428352648238602, + "loss": 0.9762, + "step": 6255 + }, + { + "epoch": 1.1137820512820513, + "grad_norm": 0.649208128452301, + "learning_rate": 0.00016427280372665525, + "loss": 1.1184, + "step": 6256 + }, + { + "epoch": 1.1139601139601139, + "grad_norm": 0.6665199398994446, + "learning_rate": 0.00016426207971163238, + "loss": 0.9417, + "step": 6257 + }, + { + "epoch": 1.1141381766381766, + "grad_norm": 0.6110978126525879, + "learning_rate": 0.00016425135443752758, + "loss": 1.1531, + "step": 6258 + }, + { + "epoch": 1.1143162393162394, + "grad_norm": 0.6517077088356018, + "learning_rate": 0.00016424062790455093, + "loss": 0.9055, + "step": 6259 + }, + { + "epoch": 1.114494301994302, + "grad_norm": 0.6278966665267944, + "learning_rate": 0.00016422990011291265, + "loss": 1.0087, + "step": 6260 + }, + { + "epoch": 1.1146723646723646, + "grad_norm": 0.5818809270858765, + "learning_rate": 0.00016421917106282288, + "loss": 1.0202, + "step": 6261 + }, + { + "epoch": 1.1148504273504274, + "grad_norm": 0.5670005679130554, + "learning_rate": 0.00016420844075449187, + "loss": 0.841, + "step": 6262 + }, + { + "epoch": 1.11502849002849, + "grad_norm": 0.6584762334823608, + "learning_rate": 0.00016419770918812984, + "loss": 1.0322, + "step": 6263 + }, + { + "epoch": 1.1152065527065527, + "grad_norm": 0.6023790836334229, + "learning_rate": 0.00016418697636394705, + "loss": 0.9152, + "step": 6264 + }, + { + "epoch": 1.1153846153846154, + "grad_norm": 0.6234691739082336, + "learning_rate": 0.00016417624228215382, + "loss": 0.9555, + "step": 6265 + }, + { + "epoch": 1.1155626780626782, + "grad_norm": 0.6690816879272461, + "learning_rate": 0.00016416550694296045, + "loss": 0.9341, + "step": 6266 + }, + { + "epoch": 1.1157407407407407, + "grad_norm": 0.6030237078666687, + "learning_rate": 0.00016415477034657723, + "loss": 1.0442, + "step": 6267 + }, + { + "epoch": 1.1159188034188035, + "grad_norm": 0.5954633951187134, + "learning_rate": 0.00016414403249321455, + "loss": 0.9132, + "step": 6268 + }, + { + "epoch": 1.116096866096866, + "grad_norm": 0.7876830101013184, + "learning_rate": 0.0001641332933830828, + "loss": 0.9456, + "step": 6269 + }, + { + "epoch": 1.1162749287749287, + "grad_norm": 0.6776009798049927, + "learning_rate": 0.00016412255301639244, + "loss": 0.9022, + "step": 6270 + }, + { + "epoch": 1.1164529914529915, + "grad_norm": 0.6094426512718201, + "learning_rate": 0.0001641118113933538, + "loss": 0.9629, + "step": 6271 + }, + { + "epoch": 1.1166310541310542, + "grad_norm": 0.5818213820457458, + "learning_rate": 0.00016410106851417742, + "loss": 0.9049, + "step": 6272 + }, + { + "epoch": 1.1168091168091168, + "grad_norm": 0.5668078064918518, + "learning_rate": 0.00016409032437907377, + "loss": 1.0011, + "step": 6273 + }, + { + "epoch": 1.1169871794871795, + "grad_norm": 0.6984922289848328, + "learning_rate": 0.00016407957898825334, + "loss": 0.9454, + "step": 6274 + }, + { + "epoch": 1.1171652421652423, + "grad_norm": 0.5509830117225647, + "learning_rate": 0.00016406883234192668, + "loss": 0.9132, + "step": 6275 + }, + { + "epoch": 1.1173433048433048, + "grad_norm": 0.5117461681365967, + "learning_rate": 0.00016405808444030435, + "loss": 0.7675, + "step": 6276 + }, + { + "epoch": 1.1175213675213675, + "grad_norm": 0.6358339786529541, + "learning_rate": 0.00016404733528359688, + "loss": 0.9777, + "step": 6277 + }, + { + "epoch": 1.1176994301994303, + "grad_norm": 0.5870591402053833, + "learning_rate": 0.00016403658487201494, + "loss": 0.8576, + "step": 6278 + }, + { + "epoch": 1.1178774928774928, + "grad_norm": 0.6532407999038696, + "learning_rate": 0.00016402583320576915, + "loss": 1.1787, + "step": 6279 + }, + { + "epoch": 1.1180555555555556, + "grad_norm": 0.6374639272689819, + "learning_rate": 0.00016401508028507017, + "loss": 0.9298, + "step": 6280 + }, + { + "epoch": 1.1182336182336183, + "grad_norm": 0.7280316352844238, + "learning_rate": 0.00016400432611012869, + "loss": 1.1081, + "step": 6281 + }, + { + "epoch": 1.1184116809116809, + "grad_norm": 0.6070699095726013, + "learning_rate": 0.00016399357068115538, + "loss": 0.9107, + "step": 6282 + }, + { + "epoch": 1.1185897435897436, + "grad_norm": 0.6701489686965942, + "learning_rate": 0.00016398281399836097, + "loss": 1.0879, + "step": 6283 + }, + { + "epoch": 1.1187678062678064, + "grad_norm": 0.6343162655830383, + "learning_rate": 0.00016397205606195626, + "loss": 0.8552, + "step": 6284 + }, + { + "epoch": 1.118945868945869, + "grad_norm": 0.6450608968734741, + "learning_rate": 0.00016396129687215198, + "loss": 1.1119, + "step": 6285 + }, + { + "epoch": 1.1191239316239316, + "grad_norm": 0.7219904661178589, + "learning_rate": 0.00016395053642915896, + "loss": 0.9081, + "step": 6286 + }, + { + "epoch": 1.1193019943019944, + "grad_norm": 0.6189733147621155, + "learning_rate": 0.00016393977473318802, + "loss": 0.9818, + "step": 6287 + }, + { + "epoch": 1.119480056980057, + "grad_norm": 0.6310907602310181, + "learning_rate": 0.00016392901178445004, + "loss": 1.0334, + "step": 6288 + }, + { + "epoch": 1.1196581196581197, + "grad_norm": 0.6556720733642578, + "learning_rate": 0.00016391824758315587, + "loss": 1.0452, + "step": 6289 + }, + { + "epoch": 1.1198361823361824, + "grad_norm": 0.6697782278060913, + "learning_rate": 0.00016390748212951638, + "loss": 0.9627, + "step": 6290 + }, + { + "epoch": 1.120014245014245, + "grad_norm": 0.6341549754142761, + "learning_rate": 0.00016389671542374256, + "loss": 1.112, + "step": 6291 + }, + { + "epoch": 1.1201923076923077, + "grad_norm": 0.6913946270942688, + "learning_rate": 0.00016388594746604535, + "loss": 0.9622, + "step": 6292 + }, + { + "epoch": 1.1203703703703705, + "grad_norm": 0.695488691329956, + "learning_rate": 0.0001638751782566357, + "loss": 1.0951, + "step": 6293 + }, + { + "epoch": 1.120548433048433, + "grad_norm": 0.6965359449386597, + "learning_rate": 0.00016386440779572463, + "loss": 1.1742, + "step": 6294 + }, + { + "epoch": 1.1207264957264957, + "grad_norm": 0.624679684638977, + "learning_rate": 0.00016385363608352314, + "loss": 0.9756, + "step": 6295 + }, + { + "epoch": 1.1209045584045585, + "grad_norm": 0.7511318922042847, + "learning_rate": 0.0001638428631202423, + "loss": 0.907, + "step": 6296 + }, + { + "epoch": 1.121082621082621, + "grad_norm": 0.5334641337394714, + "learning_rate": 0.00016383208890609317, + "loss": 0.7932, + "step": 6297 + }, + { + "epoch": 1.1212606837606838, + "grad_norm": 0.7518552541732788, + "learning_rate": 0.00016382131344128687, + "loss": 1.1556, + "step": 6298 + }, + { + "epoch": 1.1214387464387465, + "grad_norm": 0.618618369102478, + "learning_rate": 0.00016381053672603449, + "loss": 1.1027, + "step": 6299 + }, + { + "epoch": 1.121616809116809, + "grad_norm": 0.638956606388092, + "learning_rate": 0.00016379975876054724, + "loss": 1.0377, + "step": 6300 + }, + { + "epoch": 1.1217948717948718, + "grad_norm": 0.8031370639801025, + "learning_rate": 0.0001637889795450362, + "loss": 1.0821, + "step": 6301 + }, + { + "epoch": 1.1219729344729346, + "grad_norm": 0.6710168123245239, + "learning_rate": 0.00016377819907971265, + "loss": 1.2896, + "step": 6302 + }, + { + "epoch": 1.122150997150997, + "grad_norm": 0.5850739479064941, + "learning_rate": 0.00016376741736478777, + "loss": 1.0836, + "step": 6303 + }, + { + "epoch": 1.1223290598290598, + "grad_norm": 0.6410611271858215, + "learning_rate": 0.0001637566344004728, + "loss": 1.0395, + "step": 6304 + }, + { + "epoch": 1.1225071225071226, + "grad_norm": 0.6884660720825195, + "learning_rate": 0.00016374585018697903, + "loss": 0.871, + "step": 6305 + }, + { + "epoch": 1.1226851851851851, + "grad_norm": 0.622207522392273, + "learning_rate": 0.00016373506472451777, + "loss": 0.9897, + "step": 6306 + }, + { + "epoch": 1.1228632478632479, + "grad_norm": 0.6018275618553162, + "learning_rate": 0.00016372427801330028, + "loss": 0.8398, + "step": 6307 + }, + { + "epoch": 1.1230413105413106, + "grad_norm": 0.6451539993286133, + "learning_rate": 0.00016371349005353796, + "loss": 0.9878, + "step": 6308 + }, + { + "epoch": 1.1232193732193732, + "grad_norm": 0.5549424886703491, + "learning_rate": 0.00016370270084544215, + "loss": 0.844, + "step": 6309 + }, + { + "epoch": 1.123397435897436, + "grad_norm": 0.6082940697669983, + "learning_rate": 0.00016369191038922423, + "loss": 1.0704, + "step": 6310 + }, + { + "epoch": 1.1235754985754987, + "grad_norm": 0.6423100829124451, + "learning_rate": 0.00016368111868509563, + "loss": 1.0639, + "step": 6311 + }, + { + "epoch": 1.1237535612535612, + "grad_norm": 0.6274200081825256, + "learning_rate": 0.00016367032573326784, + "loss": 0.9996, + "step": 6312 + }, + { + "epoch": 1.123931623931624, + "grad_norm": 0.6618558168411255, + "learning_rate": 0.00016365953153395227, + "loss": 0.8074, + "step": 6313 + }, + { + "epoch": 1.1241096866096867, + "grad_norm": 0.7624069452285767, + "learning_rate": 0.00016364873608736038, + "loss": 0.9741, + "step": 6314 + }, + { + "epoch": 1.1242877492877492, + "grad_norm": 0.5391361117362976, + "learning_rate": 0.00016363793939370375, + "loss": 0.6992, + "step": 6315 + }, + { + "epoch": 1.124465811965812, + "grad_norm": 0.7564396858215332, + "learning_rate": 0.0001636271414531939, + "loss": 1.1971, + "step": 6316 + }, + { + "epoch": 1.1246438746438747, + "grad_norm": 0.6584066152572632, + "learning_rate": 0.00016361634226604239, + "loss": 1.0842, + "step": 6317 + }, + { + "epoch": 1.1248219373219372, + "grad_norm": 0.6851227283477783, + "learning_rate": 0.00016360554183246078, + "loss": 1.0879, + "step": 6318 + }, + { + "epoch": 1.125, + "grad_norm": 0.5699417591094971, + "learning_rate": 0.00016359474015266074, + "loss": 0.782, + "step": 6319 + }, + { + "epoch": 1.1251780626780628, + "grad_norm": 0.5495570302009583, + "learning_rate": 0.00016358393722685385, + "loss": 1.076, + "step": 6320 + }, + { + "epoch": 1.1253561253561253, + "grad_norm": 0.5872206091880798, + "learning_rate": 0.0001635731330552518, + "loss": 0.8601, + "step": 6321 + }, + { + "epoch": 1.125534188034188, + "grad_norm": 0.7012827396392822, + "learning_rate": 0.00016356232763806627, + "loss": 1.0443, + "step": 6322 + }, + { + "epoch": 1.1257122507122508, + "grad_norm": 0.6645881533622742, + "learning_rate": 0.00016355152097550897, + "loss": 1.0027, + "step": 6323 + }, + { + "epoch": 1.1258903133903133, + "grad_norm": 0.7376120090484619, + "learning_rate": 0.00016354071306779163, + "loss": 1.1941, + "step": 6324 + }, + { + "epoch": 1.126068376068376, + "grad_norm": 0.648932695388794, + "learning_rate": 0.000163529903915126, + "loss": 1.096, + "step": 6325 + }, + { + "epoch": 1.1262464387464388, + "grad_norm": 0.6186314821243286, + "learning_rate": 0.0001635190935177239, + "loss": 1.011, + "step": 6326 + }, + { + "epoch": 1.1264245014245013, + "grad_norm": 0.5964710116386414, + "learning_rate": 0.0001635082818757971, + "loss": 0.8893, + "step": 6327 + }, + { + "epoch": 1.126602564102564, + "grad_norm": 0.5264934301376343, + "learning_rate": 0.00016349746898955747, + "loss": 0.7325, + "step": 6328 + }, + { + "epoch": 1.1267806267806268, + "grad_norm": 0.6523048877716064, + "learning_rate": 0.00016348665485921678, + "loss": 1.0488, + "step": 6329 + }, + { + "epoch": 1.1269586894586894, + "grad_norm": 0.6878600120544434, + "learning_rate": 0.00016347583948498703, + "loss": 1.0926, + "step": 6330 + }, + { + "epoch": 1.1271367521367521, + "grad_norm": 0.592656672000885, + "learning_rate": 0.00016346502286708004, + "loss": 0.978, + "step": 6331 + }, + { + "epoch": 1.1273148148148149, + "grad_norm": 0.6338315606117249, + "learning_rate": 0.00016345420500570777, + "loss": 1.1048, + "step": 6332 + }, + { + "epoch": 1.1274928774928774, + "grad_norm": 0.5955204367637634, + "learning_rate": 0.00016344338590108218, + "loss": 0.88, + "step": 6333 + }, + { + "epoch": 1.1276709401709402, + "grad_norm": 0.690448522567749, + "learning_rate": 0.0001634325655534152, + "loss": 1.0564, + "step": 6334 + }, + { + "epoch": 1.127849002849003, + "grad_norm": 0.6125795841217041, + "learning_rate": 0.00016342174396291888, + "loss": 1.0608, + "step": 6335 + }, + { + "epoch": 1.1280270655270654, + "grad_norm": 0.6387807726860046, + "learning_rate": 0.00016341092112980523, + "loss": 0.9581, + "step": 6336 + }, + { + "epoch": 1.1282051282051282, + "grad_norm": 0.6247823238372803, + "learning_rate": 0.0001634000970542863, + "loss": 0.932, + "step": 6337 + }, + { + "epoch": 1.128383190883191, + "grad_norm": 0.5928077697753906, + "learning_rate": 0.0001633892717365742, + "loss": 0.8963, + "step": 6338 + }, + { + "epoch": 1.1285612535612535, + "grad_norm": 0.5922074913978577, + "learning_rate": 0.000163378445176881, + "loss": 0.9772, + "step": 6339 + }, + { + "epoch": 1.1287393162393162, + "grad_norm": 0.6573056578636169, + "learning_rate": 0.00016336761737541878, + "loss": 0.8233, + "step": 6340 + }, + { + "epoch": 1.128917378917379, + "grad_norm": 0.627772867679596, + "learning_rate": 0.0001633567883323998, + "loss": 0.9618, + "step": 6341 + }, + { + "epoch": 1.1290954415954415, + "grad_norm": 0.6066579818725586, + "learning_rate": 0.0001633459580480361, + "loss": 0.9066, + "step": 6342 + }, + { + "epoch": 1.1292735042735043, + "grad_norm": 0.670295000076294, + "learning_rate": 0.00016333512652253997, + "loss": 0.8003, + "step": 6343 + }, + { + "epoch": 1.129451566951567, + "grad_norm": 0.6402488946914673, + "learning_rate": 0.0001633242937561236, + "loss": 0.998, + "step": 6344 + }, + { + "epoch": 1.1296296296296295, + "grad_norm": 0.7224995493888855, + "learning_rate": 0.00016331345974899923, + "loss": 1.0308, + "step": 6345 + }, + { + "epoch": 1.1298076923076923, + "grad_norm": 0.5019716620445251, + "learning_rate": 0.00016330262450137917, + "loss": 0.6874, + "step": 6346 + }, + { + "epoch": 1.129985754985755, + "grad_norm": 0.5774167776107788, + "learning_rate": 0.00016329178801347566, + "loss": 0.8287, + "step": 6347 + }, + { + "epoch": 1.1301638176638176, + "grad_norm": 0.7797795534133911, + "learning_rate": 0.00016328095028550103, + "loss": 1.2145, + "step": 6348 + }, + { + "epoch": 1.1303418803418803, + "grad_norm": 0.5384017825126648, + "learning_rate": 0.00016327011131766765, + "loss": 0.8022, + "step": 6349 + }, + { + "epoch": 1.130519943019943, + "grad_norm": 0.6350888609886169, + "learning_rate": 0.00016325927111018786, + "loss": 1.1178, + "step": 6350 + }, + { + "epoch": 1.1306980056980056, + "grad_norm": 0.6386831998825073, + "learning_rate": 0.0001632484296632741, + "loss": 0.967, + "step": 6351 + }, + { + "epoch": 1.1308760683760684, + "grad_norm": 0.6214167475700378, + "learning_rate": 0.0001632375869771387, + "loss": 0.9416, + "step": 6352 + }, + { + "epoch": 1.131054131054131, + "grad_norm": 0.6145567297935486, + "learning_rate": 0.00016322674305199416, + "loss": 0.9175, + "step": 6353 + }, + { + "epoch": 1.1312321937321936, + "grad_norm": 0.7027857303619385, + "learning_rate": 0.00016321589788805297, + "loss": 1.0063, + "step": 6354 + }, + { + "epoch": 1.1314102564102564, + "grad_norm": 0.6942669153213501, + "learning_rate": 0.00016320505148552755, + "loss": 0.9191, + "step": 6355 + }, + { + "epoch": 1.1315883190883191, + "grad_norm": 0.6388658285140991, + "learning_rate": 0.0001631942038446304, + "loss": 0.993, + "step": 6356 + }, + { + "epoch": 1.131766381766382, + "grad_norm": 0.6627292633056641, + "learning_rate": 0.00016318335496557415, + "loss": 1.0055, + "step": 6357 + }, + { + "epoch": 1.1319444444444444, + "grad_norm": 0.7997342944145203, + "learning_rate": 0.0001631725048485713, + "loss": 0.9019, + "step": 6358 + }, + { + "epoch": 1.1321225071225072, + "grad_norm": 0.8817830681800842, + "learning_rate": 0.00016316165349383445, + "loss": 0.9793, + "step": 6359 + }, + { + "epoch": 1.1323005698005697, + "grad_norm": 0.5629408955574036, + "learning_rate": 0.00016315080090157621, + "loss": 0.6139, + "step": 6360 + }, + { + "epoch": 1.1324786324786325, + "grad_norm": 0.647220253944397, + "learning_rate": 0.0001631399470720092, + "loss": 0.9776, + "step": 6361 + }, + { + "epoch": 1.1326566951566952, + "grad_norm": 0.6762630939483643, + "learning_rate": 0.0001631290920053461, + "loss": 1.1027, + "step": 6362 + }, + { + "epoch": 1.132834757834758, + "grad_norm": 0.5862727761268616, + "learning_rate": 0.00016311823570179957, + "loss": 1.1359, + "step": 6363 + }, + { + "epoch": 1.1330128205128205, + "grad_norm": 0.7042981386184692, + "learning_rate": 0.00016310737816158235, + "loss": 1.142, + "step": 6364 + }, + { + "epoch": 1.1331908831908832, + "grad_norm": 0.5990639328956604, + "learning_rate": 0.00016309651938490712, + "loss": 0.9306, + "step": 6365 + }, + { + "epoch": 1.1333689458689458, + "grad_norm": 0.5894871950149536, + "learning_rate": 0.00016308565937198669, + "loss": 0.8343, + "step": 6366 + }, + { + "epoch": 1.1335470085470085, + "grad_norm": 0.6863628029823303, + "learning_rate": 0.0001630747981230338, + "loss": 0.9552, + "step": 6367 + }, + { + "epoch": 1.1337250712250713, + "grad_norm": 0.7438958287239075, + "learning_rate": 0.00016306393563826128, + "loss": 1.0422, + "step": 6368 + }, + { + "epoch": 1.133903133903134, + "grad_norm": 0.5695775747299194, + "learning_rate": 0.00016305307191788194, + "loss": 0.8633, + "step": 6369 + }, + { + "epoch": 1.1340811965811965, + "grad_norm": 0.6257741451263428, + "learning_rate": 0.00016304220696210863, + "loss": 1.0333, + "step": 6370 + }, + { + "epoch": 1.1342592592592593, + "grad_norm": 0.6366072297096252, + "learning_rate": 0.00016303134077115425, + "loss": 1.1452, + "step": 6371 + }, + { + "epoch": 1.1344373219373218, + "grad_norm": 0.624569296836853, + "learning_rate": 0.00016302047334523168, + "loss": 1.0569, + "step": 6372 + }, + { + "epoch": 1.1346153846153846, + "grad_norm": 0.5585938096046448, + "learning_rate": 0.00016300960468455382, + "loss": 0.9612, + "step": 6373 + }, + { + "epoch": 1.1347934472934473, + "grad_norm": 0.5738831162452698, + "learning_rate": 0.00016299873478933368, + "loss": 0.9206, + "step": 6374 + }, + { + "epoch": 1.13497150997151, + "grad_norm": 0.6797143220901489, + "learning_rate": 0.00016298786365978417, + "loss": 1.0748, + "step": 6375 + }, + { + "epoch": 1.1351495726495726, + "grad_norm": 0.6341326832771301, + "learning_rate": 0.00016297699129611833, + "loss": 0.9901, + "step": 6376 + }, + { + "epoch": 1.1353276353276354, + "grad_norm": 0.6568490862846375, + "learning_rate": 0.00016296611769854916, + "loss": 1.0598, + "step": 6377 + }, + { + "epoch": 1.135505698005698, + "grad_norm": 0.6151928901672363, + "learning_rate": 0.00016295524286728973, + "loss": 0.8352, + "step": 6378 + }, + { + "epoch": 1.1356837606837606, + "grad_norm": 0.7209593057632446, + "learning_rate": 0.0001629443668025531, + "loss": 0.9945, + "step": 6379 + }, + { + "epoch": 1.1358618233618234, + "grad_norm": 0.6600689888000488, + "learning_rate": 0.00016293348950455235, + "loss": 1.0572, + "step": 6380 + }, + { + "epoch": 1.1360398860398861, + "grad_norm": 0.5587523579597473, + "learning_rate": 0.0001629226109735006, + "loss": 0.8526, + "step": 6381 + }, + { + "epoch": 1.1362179487179487, + "grad_norm": 0.6184542775154114, + "learning_rate": 0.00016291173120961102, + "loss": 0.8246, + "step": 6382 + }, + { + "epoch": 1.1363960113960114, + "grad_norm": 0.6604713797569275, + "learning_rate": 0.00016290085021309673, + "loss": 1.0349, + "step": 6383 + }, + { + "epoch": 1.136574074074074, + "grad_norm": 0.5880835056304932, + "learning_rate": 0.00016288996798417097, + "loss": 0.8726, + "step": 6384 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 0.5770880579948425, + "learning_rate": 0.00016287908452304692, + "loss": 0.7639, + "step": 6385 + }, + { + "epoch": 1.1369301994301995, + "grad_norm": 0.5719713568687439, + "learning_rate": 0.00016286819982993782, + "loss": 0.9717, + "step": 6386 + }, + { + "epoch": 1.1371082621082622, + "grad_norm": 0.7028461694717407, + "learning_rate": 0.00016285731390505695, + "loss": 1.0147, + "step": 6387 + }, + { + "epoch": 1.1372863247863247, + "grad_norm": 0.5396828651428223, + "learning_rate": 0.00016284642674861756, + "loss": 0.8119, + "step": 6388 + }, + { + "epoch": 1.1374643874643875, + "grad_norm": 0.592580258846283, + "learning_rate": 0.00016283553836083303, + "loss": 1.0914, + "step": 6389 + }, + { + "epoch": 1.13764245014245, + "grad_norm": 0.634596586227417, + "learning_rate": 0.00016282464874191663, + "loss": 1.1037, + "step": 6390 + }, + { + "epoch": 1.1378205128205128, + "grad_norm": 0.6462705731391907, + "learning_rate": 0.00016281375789208176, + "loss": 1.1523, + "step": 6391 + }, + { + "epoch": 1.1379985754985755, + "grad_norm": 0.6527917385101318, + "learning_rate": 0.0001628028658115418, + "loss": 1.0415, + "step": 6392 + }, + { + "epoch": 1.1381766381766383, + "grad_norm": 0.6309964060783386, + "learning_rate": 0.00016279197250051013, + "loss": 0.9747, + "step": 6393 + }, + { + "epoch": 1.1383547008547008, + "grad_norm": 0.6342993974685669, + "learning_rate": 0.00016278107795920018, + "loss": 0.9897, + "step": 6394 + }, + { + "epoch": 1.1385327635327636, + "grad_norm": 0.7149887084960938, + "learning_rate": 0.00016277018218782544, + "loss": 0.9659, + "step": 6395 + }, + { + "epoch": 1.138710826210826, + "grad_norm": 0.7219462394714355, + "learning_rate": 0.00016275928518659938, + "loss": 0.9301, + "step": 6396 + }, + { + "epoch": 1.1388888888888888, + "grad_norm": 0.6649485230445862, + "learning_rate": 0.0001627483869557355, + "loss": 0.9012, + "step": 6397 + }, + { + "epoch": 1.1390669515669516, + "grad_norm": 0.6910027861595154, + "learning_rate": 0.00016273748749544731, + "loss": 0.956, + "step": 6398 + }, + { + "epoch": 1.1392450142450143, + "grad_norm": 0.6369016766548157, + "learning_rate": 0.00016272658680594837, + "loss": 0.8027, + "step": 6399 + }, + { + "epoch": 1.1394230769230769, + "grad_norm": 0.6540524959564209, + "learning_rate": 0.00016271568488745227, + "loss": 1.2397, + "step": 6400 + }, + { + "epoch": 1.1396011396011396, + "grad_norm": 0.5912376046180725, + "learning_rate": 0.00016270478174017263, + "loss": 0.8453, + "step": 6401 + }, + { + "epoch": 1.1397792022792024, + "grad_norm": 0.6847240924835205, + "learning_rate": 0.00016269387736432303, + "loss": 0.9776, + "step": 6402 + }, + { + "epoch": 1.139957264957265, + "grad_norm": 0.6465024352073669, + "learning_rate": 0.00016268297176011716, + "loss": 0.8971, + "step": 6403 + }, + { + "epoch": 1.1401353276353277, + "grad_norm": 0.6639063954353333, + "learning_rate": 0.00016267206492776866, + "loss": 0.9756, + "step": 6404 + }, + { + "epoch": 1.1403133903133904, + "grad_norm": 0.6343763470649719, + "learning_rate": 0.00016266115686749123, + "loss": 0.9368, + "step": 6405 + }, + { + "epoch": 1.140491452991453, + "grad_norm": 0.7144993543624878, + "learning_rate": 0.0001626502475794986, + "loss": 0.9285, + "step": 6406 + }, + { + "epoch": 1.1406695156695157, + "grad_norm": 0.6217414736747742, + "learning_rate": 0.00016263933706400451, + "loss": 0.8867, + "step": 6407 + }, + { + "epoch": 1.1408475783475784, + "grad_norm": 0.6843730807304382, + "learning_rate": 0.00016262842532122274, + "loss": 0.9863, + "step": 6408 + }, + { + "epoch": 1.141025641025641, + "grad_norm": 0.6866166591644287, + "learning_rate": 0.00016261751235136705, + "loss": 1.0517, + "step": 6409 + }, + { + "epoch": 1.1412037037037037, + "grad_norm": 0.6650584936141968, + "learning_rate": 0.0001626065981546513, + "loss": 1.0629, + "step": 6410 + }, + { + "epoch": 1.1413817663817665, + "grad_norm": 0.5805012583732605, + "learning_rate": 0.00016259568273128933, + "loss": 0.8175, + "step": 6411 + }, + { + "epoch": 1.141559829059829, + "grad_norm": 0.7005903124809265, + "learning_rate": 0.00016258476608149497, + "loss": 1.0267, + "step": 6412 + }, + { + "epoch": 1.1417378917378918, + "grad_norm": 0.6293461322784424, + "learning_rate": 0.00016257384820548217, + "loss": 1.1034, + "step": 6413 + }, + { + "epoch": 1.1419159544159545, + "grad_norm": 0.6281774640083313, + "learning_rate": 0.00016256292910346476, + "loss": 1.0775, + "step": 6414 + }, + { + "epoch": 1.142094017094017, + "grad_norm": 0.5912862420082092, + "learning_rate": 0.0001625520087756567, + "loss": 0.9589, + "step": 6415 + }, + { + "epoch": 1.1422720797720798, + "grad_norm": 0.5813978314399719, + "learning_rate": 0.00016254108722227198, + "loss": 0.9195, + "step": 6416 + }, + { + "epoch": 1.1424501424501425, + "grad_norm": 0.650805652141571, + "learning_rate": 0.00016253016444352458, + "loss": 1.0207, + "step": 6417 + }, + { + "epoch": 1.142628205128205, + "grad_norm": 0.6909520030021667, + "learning_rate": 0.00016251924043962851, + "loss": 0.9854, + "step": 6418 + }, + { + "epoch": 1.1428062678062678, + "grad_norm": 0.6054595112800598, + "learning_rate": 0.0001625083152107978, + "loss": 0.852, + "step": 6419 + }, + { + "epoch": 1.1429843304843306, + "grad_norm": 0.601078987121582, + "learning_rate": 0.00016249738875724647, + "loss": 0.9609, + "step": 6420 + }, + { + "epoch": 1.143162393162393, + "grad_norm": 0.5340180397033691, + "learning_rate": 0.00016248646107918868, + "loss": 0.8364, + "step": 6421 + }, + { + "epoch": 1.1433404558404558, + "grad_norm": 0.6687821745872498, + "learning_rate": 0.00016247553217683846, + "loss": 1.005, + "step": 6422 + }, + { + "epoch": 1.1435185185185186, + "grad_norm": 0.6347902417182922, + "learning_rate": 0.00016246460205040998, + "loss": 1.026, + "step": 6423 + }, + { + "epoch": 1.1436965811965811, + "grad_norm": 0.6136734485626221, + "learning_rate": 0.00016245367070011736, + "loss": 0.7811, + "step": 6424 + }, + { + "epoch": 1.1438746438746439, + "grad_norm": 0.6591334342956543, + "learning_rate": 0.00016244273812617482, + "loss": 0.991, + "step": 6425 + }, + { + "epoch": 1.1440527065527066, + "grad_norm": 0.6062475442886353, + "learning_rate": 0.00016243180432879656, + "loss": 0.9879, + "step": 6426 + }, + { + "epoch": 1.1442307692307692, + "grad_norm": 0.5941380858421326, + "learning_rate": 0.00016242086930819678, + "loss": 0.9771, + "step": 6427 + }, + { + "epoch": 1.144408831908832, + "grad_norm": 0.7320533990859985, + "learning_rate": 0.00016240993306458973, + "loss": 1.0919, + "step": 6428 + }, + { + "epoch": 1.1445868945868947, + "grad_norm": 0.6998075246810913, + "learning_rate": 0.00016239899559818962, + "loss": 1.0721, + "step": 6429 + }, + { + "epoch": 1.1447649572649572, + "grad_norm": 0.847931444644928, + "learning_rate": 0.0001623880569092109, + "loss": 0.8759, + "step": 6430 + }, + { + "epoch": 1.14494301994302, + "grad_norm": 0.6670104265213013, + "learning_rate": 0.00016237711699786775, + "loss": 1.0515, + "step": 6431 + }, + { + "epoch": 1.1451210826210827, + "grad_norm": 0.601759672164917, + "learning_rate": 0.00016236617586437463, + "loss": 0.7298, + "step": 6432 + }, + { + "epoch": 1.1452991452991452, + "grad_norm": 0.6411594152450562, + "learning_rate": 0.00016235523350894578, + "loss": 0.9336, + "step": 6433 + }, + { + "epoch": 1.145477207977208, + "grad_norm": 0.6485120058059692, + "learning_rate": 0.0001623442899317957, + "loss": 1.1215, + "step": 6434 + }, + { + "epoch": 1.1456552706552707, + "grad_norm": 0.6041508316993713, + "learning_rate": 0.00016233334513313875, + "loss": 0.8917, + "step": 6435 + }, + { + "epoch": 1.1458333333333333, + "grad_norm": 0.6292745471000671, + "learning_rate": 0.0001623223991131894, + "loss": 0.9976, + "step": 6436 + }, + { + "epoch": 1.146011396011396, + "grad_norm": 0.5442200303077698, + "learning_rate": 0.0001623114518721621, + "loss": 0.8072, + "step": 6437 + }, + { + "epoch": 1.1461894586894588, + "grad_norm": 0.6668170094490051, + "learning_rate": 0.00016230050341027136, + "loss": 0.9641, + "step": 6438 + }, + { + "epoch": 1.1463675213675213, + "grad_norm": 0.644186794757843, + "learning_rate": 0.00016228955372773164, + "loss": 0.9248, + "step": 6439 + }, + { + "epoch": 1.146545584045584, + "grad_norm": 0.6661991477012634, + "learning_rate": 0.00016227860282475753, + "loss": 0.8719, + "step": 6440 + }, + { + "epoch": 1.1467236467236468, + "grad_norm": 0.5232062935829163, + "learning_rate": 0.00016226765070156355, + "loss": 0.5418, + "step": 6441 + }, + { + "epoch": 1.1469017094017093, + "grad_norm": 0.573176383972168, + "learning_rate": 0.00016225669735836436, + "loss": 1.0858, + "step": 6442 + }, + { + "epoch": 1.147079772079772, + "grad_norm": 0.6137439608573914, + "learning_rate": 0.00016224574279537446, + "loss": 1.1205, + "step": 6443 + }, + { + "epoch": 1.1472578347578348, + "grad_norm": 0.6328136920928955, + "learning_rate": 0.00016223478701280855, + "loss": 0.8957, + "step": 6444 + }, + { + "epoch": 1.1474358974358974, + "grad_norm": 0.6687374114990234, + "learning_rate": 0.00016222383001088126, + "loss": 1.0318, + "step": 6445 + }, + { + "epoch": 1.14761396011396, + "grad_norm": 0.6057115793228149, + "learning_rate": 0.0001622128717898073, + "loss": 0.9575, + "step": 6446 + }, + { + "epoch": 1.1477920227920229, + "grad_norm": 0.6758735775947571, + "learning_rate": 0.0001622019123498013, + "loss": 1.2273, + "step": 6447 + }, + { + "epoch": 1.1479700854700854, + "grad_norm": 0.6233550310134888, + "learning_rate": 0.0001621909516910781, + "loss": 0.7875, + "step": 6448 + }, + { + "epoch": 1.1481481481481481, + "grad_norm": 0.6371827721595764, + "learning_rate": 0.0001621799898138524, + "loss": 1.0488, + "step": 6449 + }, + { + "epoch": 1.148326210826211, + "grad_norm": 0.6179831624031067, + "learning_rate": 0.00016216902671833892, + "loss": 0.9792, + "step": 6450 + }, + { + "epoch": 1.1485042735042734, + "grad_norm": 0.6234193444252014, + "learning_rate": 0.00016215806240475256, + "loss": 0.927, + "step": 6451 + }, + { + "epoch": 1.1486823361823362, + "grad_norm": 0.6940563917160034, + "learning_rate": 0.00016214709687330803, + "loss": 1.047, + "step": 6452 + }, + { + "epoch": 1.148860398860399, + "grad_norm": 0.6567606925964355, + "learning_rate": 0.00016213613012422027, + "loss": 0.9695, + "step": 6453 + }, + { + "epoch": 1.1490384615384615, + "grad_norm": 0.7374183535575867, + "learning_rate": 0.0001621251621577041, + "loss": 1.0443, + "step": 6454 + }, + { + "epoch": 1.1492165242165242, + "grad_norm": 0.6789869666099548, + "learning_rate": 0.00016211419297397443, + "loss": 1.0319, + "step": 6455 + }, + { + "epoch": 1.149394586894587, + "grad_norm": 0.6225521564483643, + "learning_rate": 0.00016210322257324619, + "loss": 1.0529, + "step": 6456 + }, + { + "epoch": 1.1495726495726495, + "grad_norm": 0.619701623916626, + "learning_rate": 0.00016209225095573432, + "loss": 0.962, + "step": 6457 + }, + { + "epoch": 1.1497507122507122, + "grad_norm": 0.6132834553718567, + "learning_rate": 0.00016208127812165375, + "loss": 0.9588, + "step": 6458 + }, + { + "epoch": 1.149928774928775, + "grad_norm": 0.6005367040634155, + "learning_rate": 0.00016207030407121954, + "loss": 0.9497, + "step": 6459 + }, + { + "epoch": 1.1501068376068375, + "grad_norm": 0.575309157371521, + "learning_rate": 0.00016205932880464664, + "loss": 1.0035, + "step": 6460 + }, + { + "epoch": 1.1502849002849003, + "grad_norm": 0.5958710312843323, + "learning_rate": 0.0001620483523221501, + "loss": 1.0004, + "step": 6461 + }, + { + "epoch": 1.150462962962963, + "grad_norm": 0.5934719443321228, + "learning_rate": 0.000162037374623945, + "loss": 0.8694, + "step": 6462 + }, + { + "epoch": 1.1506410256410255, + "grad_norm": 0.6042510271072388, + "learning_rate": 0.00016202639571024643, + "loss": 0.8598, + "step": 6463 + }, + { + "epoch": 1.1508190883190883, + "grad_norm": 0.6206158399581909, + "learning_rate": 0.00016201541558126946, + "loss": 0.961, + "step": 6464 + }, + { + "epoch": 1.150997150997151, + "grad_norm": 0.5997715592384338, + "learning_rate": 0.00016200443423722925, + "loss": 0.8686, + "step": 6465 + }, + { + "epoch": 1.1511752136752136, + "grad_norm": 0.742457926273346, + "learning_rate": 0.00016199345167834098, + "loss": 1.1113, + "step": 6466 + }, + { + "epoch": 1.1513532763532763, + "grad_norm": 0.6772766709327698, + "learning_rate": 0.00016198246790481976, + "loss": 1.0717, + "step": 6467 + }, + { + "epoch": 1.151531339031339, + "grad_norm": 0.6127712726593018, + "learning_rate": 0.0001619714829168809, + "loss": 0.8887, + "step": 6468 + }, + { + "epoch": 1.1517094017094016, + "grad_norm": 0.5585067272186279, + "learning_rate": 0.00016196049671473954, + "loss": 1.0144, + "step": 6469 + }, + { + "epoch": 1.1518874643874644, + "grad_norm": 0.6269431710243225, + "learning_rate": 0.00016194950929861092, + "loss": 1.0206, + "step": 6470 + }, + { + "epoch": 1.1520655270655271, + "grad_norm": 0.6270785331726074, + "learning_rate": 0.0001619385206687104, + "loss": 1.0517, + "step": 6471 + }, + { + "epoch": 1.1522435897435896, + "grad_norm": 0.744712233543396, + "learning_rate": 0.00016192753082525322, + "loss": 1.0699, + "step": 6472 + }, + { + "epoch": 1.1524216524216524, + "grad_norm": 0.7025929689407349, + "learning_rate": 0.00016191653976845474, + "loss": 0.951, + "step": 6473 + }, + { + "epoch": 1.1525997150997151, + "grad_norm": 0.6175379753112793, + "learning_rate": 0.00016190554749853024, + "loss": 1.2153, + "step": 6474 + }, + { + "epoch": 1.1527777777777777, + "grad_norm": 0.6212149858474731, + "learning_rate": 0.00016189455401569513, + "loss": 1.0428, + "step": 6475 + }, + { + "epoch": 1.1529558404558404, + "grad_norm": 0.6716817617416382, + "learning_rate": 0.00016188355932016484, + "loss": 1.179, + "step": 6476 + }, + { + "epoch": 1.1531339031339032, + "grad_norm": 0.6247739791870117, + "learning_rate": 0.00016187256341215476, + "loss": 0.9451, + "step": 6477 + }, + { + "epoch": 1.153311965811966, + "grad_norm": 0.6223008036613464, + "learning_rate": 0.00016186156629188032, + "loss": 0.9915, + "step": 6478 + }, + { + "epoch": 1.1534900284900285, + "grad_norm": 0.5610866546630859, + "learning_rate": 0.000161850567959557, + "loss": 0.7741, + "step": 6479 + }, + { + "epoch": 1.1536680911680912, + "grad_norm": 0.6241226196289062, + "learning_rate": 0.0001618395684154003, + "loss": 1.2193, + "step": 6480 + }, + { + "epoch": 1.1538461538461537, + "grad_norm": 0.703789472579956, + "learning_rate": 0.00016182856765962567, + "loss": 1.0725, + "step": 6481 + }, + { + "epoch": 1.1540242165242165, + "grad_norm": 0.6802006959915161, + "learning_rate": 0.00016181756569244872, + "loss": 1.0908, + "step": 6482 + }, + { + "epoch": 1.1542022792022792, + "grad_norm": 0.6504136919975281, + "learning_rate": 0.000161806562514085, + "loss": 0.9706, + "step": 6483 + }, + { + "epoch": 1.154380341880342, + "grad_norm": 0.7217034101486206, + "learning_rate": 0.00016179555812475003, + "loss": 0.9084, + "step": 6484 + }, + { + "epoch": 1.1545584045584045, + "grad_norm": 0.5919039249420166, + "learning_rate": 0.0001617845525246595, + "loss": 0.949, + "step": 6485 + }, + { + "epoch": 1.1547364672364673, + "grad_norm": 0.6160184741020203, + "learning_rate": 0.00016177354571402902, + "loss": 0.8144, + "step": 6486 + }, + { + "epoch": 1.1549145299145298, + "grad_norm": 0.7323806285858154, + "learning_rate": 0.00016176253769307426, + "loss": 1.0528, + "step": 6487 + }, + { + "epoch": 1.1550925925925926, + "grad_norm": 0.6051317453384399, + "learning_rate": 0.0001617515284620108, + "loss": 0.9558, + "step": 6488 + }, + { + "epoch": 1.1552706552706553, + "grad_norm": 0.6418905258178711, + "learning_rate": 0.00016174051802105447, + "loss": 1.062, + "step": 6489 + }, + { + "epoch": 1.155448717948718, + "grad_norm": 0.6914883852005005, + "learning_rate": 0.00016172950637042096, + "loss": 0.9999, + "step": 6490 + }, + { + "epoch": 1.1556267806267806, + "grad_norm": 0.5558316707611084, + "learning_rate": 0.000161718493510326, + "loss": 0.9561, + "step": 6491 + }, + { + "epoch": 1.1558048433048433, + "grad_norm": 0.6632496118545532, + "learning_rate": 0.00016170747944098531, + "loss": 1.0133, + "step": 6492 + }, + { + "epoch": 1.1559829059829059, + "grad_norm": 0.6407149434089661, + "learning_rate": 0.00016169646416261478, + "loss": 1.0563, + "step": 6493 + }, + { + "epoch": 1.1561609686609686, + "grad_norm": 0.8128494024276733, + "learning_rate": 0.0001616854476754302, + "loss": 1.1559, + "step": 6494 + }, + { + "epoch": 1.1563390313390314, + "grad_norm": 0.6403429508209229, + "learning_rate": 0.00016167442997964742, + "loss": 1.0983, + "step": 6495 + }, + { + "epoch": 1.1565170940170941, + "grad_norm": 0.76612788438797, + "learning_rate": 0.0001616634110754823, + "loss": 0.973, + "step": 6496 + }, + { + "epoch": 1.1566951566951567, + "grad_norm": 0.6914355754852295, + "learning_rate": 0.0001616523909631507, + "loss": 0.9307, + "step": 6497 + }, + { + "epoch": 1.1568732193732194, + "grad_norm": 0.546602725982666, + "learning_rate": 0.00016164136964286863, + "loss": 1.0328, + "step": 6498 + }, + { + "epoch": 1.157051282051282, + "grad_norm": 0.5695818662643433, + "learning_rate": 0.00016163034711485193, + "loss": 0.9607, + "step": 6499 + }, + { + "epoch": 1.1572293447293447, + "grad_norm": 0.5649738311767578, + "learning_rate": 0.00016161932337931662, + "loss": 1.1521, + "step": 6500 + }, + { + "epoch": 1.1574074074074074, + "grad_norm": 0.6437582969665527, + "learning_rate": 0.00016160829843647867, + "loss": 0.9613, + "step": 6501 + }, + { + "epoch": 1.1575854700854702, + "grad_norm": 0.5841929316520691, + "learning_rate": 0.0001615972722865541, + "loss": 0.8187, + "step": 6502 + }, + { + "epoch": 1.1577635327635327, + "grad_norm": 0.6481246948242188, + "learning_rate": 0.00016158624492975892, + "loss": 1.0447, + "step": 6503 + }, + { + "epoch": 1.1579415954415955, + "grad_norm": 0.629804790019989, + "learning_rate": 0.0001615752163663092, + "loss": 0.9034, + "step": 6504 + }, + { + "epoch": 1.158119658119658, + "grad_norm": 0.5797054171562195, + "learning_rate": 0.00016156418659642104, + "loss": 0.8168, + "step": 6505 + }, + { + "epoch": 1.1582977207977208, + "grad_norm": 0.588424563407898, + "learning_rate": 0.00016155315562031052, + "loss": 0.828, + "step": 6506 + }, + { + "epoch": 1.1584757834757835, + "grad_norm": 0.7120068669319153, + "learning_rate": 0.0001615421234381938, + "loss": 1.0637, + "step": 6507 + }, + { + "epoch": 1.1586538461538463, + "grad_norm": 0.6635081768035889, + "learning_rate": 0.00016153109005028702, + "loss": 0.9838, + "step": 6508 + }, + { + "epoch": 1.1588319088319088, + "grad_norm": 0.6080414056777954, + "learning_rate": 0.00016152005545680634, + "loss": 0.983, + "step": 6509 + }, + { + "epoch": 1.1590099715099715, + "grad_norm": 0.7131237983703613, + "learning_rate": 0.00016150901965796796, + "loss": 1.1053, + "step": 6510 + }, + { + "epoch": 1.159188034188034, + "grad_norm": 0.6051005125045776, + "learning_rate": 0.00016149798265398813, + "loss": 0.9903, + "step": 6511 + }, + { + "epoch": 1.1593660968660968, + "grad_norm": 0.6193733811378479, + "learning_rate": 0.00016148694444508306, + "loss": 1.0478, + "step": 6512 + }, + { + "epoch": 1.1595441595441596, + "grad_norm": 0.567888081073761, + "learning_rate": 0.00016147590503146905, + "loss": 0.7995, + "step": 6513 + }, + { + "epoch": 1.1597222222222223, + "grad_norm": 0.6889783143997192, + "learning_rate": 0.00016146486441336242, + "loss": 0.9684, + "step": 6514 + }, + { + "epoch": 1.1599002849002849, + "grad_norm": 0.6470308303833008, + "learning_rate": 0.0001614538225909794, + "loss": 0.9824, + "step": 6515 + }, + { + "epoch": 1.1600783475783476, + "grad_norm": 0.6833886504173279, + "learning_rate": 0.00016144277956453638, + "loss": 0.9845, + "step": 6516 + }, + { + "epoch": 1.1602564102564104, + "grad_norm": 0.5827815532684326, + "learning_rate": 0.00016143173533424978, + "loss": 0.9476, + "step": 6517 + }, + { + "epoch": 1.1604344729344729, + "grad_norm": 0.6701242327690125, + "learning_rate": 0.00016142068990033593, + "loss": 1.0839, + "step": 6518 + }, + { + "epoch": 1.1606125356125356, + "grad_norm": 0.5844996571540833, + "learning_rate": 0.00016140964326301122, + "loss": 0.8861, + "step": 6519 + }, + { + "epoch": 1.1607905982905984, + "grad_norm": 0.5831994414329529, + "learning_rate": 0.00016139859542249214, + "loss": 0.9817, + "step": 6520 + }, + { + "epoch": 1.160968660968661, + "grad_norm": 0.6830124855041504, + "learning_rate": 0.0001613875463789951, + "loss": 0.8749, + "step": 6521 + }, + { + "epoch": 1.1611467236467237, + "grad_norm": 0.6003018021583557, + "learning_rate": 0.00016137649613273667, + "loss": 0.9593, + "step": 6522 + }, + { + "epoch": 1.1613247863247864, + "grad_norm": 0.5973994731903076, + "learning_rate": 0.00016136544468393327, + "loss": 1.0384, + "step": 6523 + }, + { + "epoch": 1.161502849002849, + "grad_norm": 0.6702523827552795, + "learning_rate": 0.00016135439203280143, + "loss": 1.0431, + "step": 6524 + }, + { + "epoch": 1.1616809116809117, + "grad_norm": 0.6160697937011719, + "learning_rate": 0.00016134333817955775, + "loss": 1.0339, + "step": 6525 + }, + { + "epoch": 1.1618589743589745, + "grad_norm": 0.7078264355659485, + "learning_rate": 0.0001613322831244188, + "loss": 1.0285, + "step": 6526 + }, + { + "epoch": 1.162037037037037, + "grad_norm": 0.5744216442108154, + "learning_rate": 0.00016132122686760117, + "loss": 0.6589, + "step": 6527 + }, + { + "epoch": 1.1622150997150997, + "grad_norm": 0.6802098155021667, + "learning_rate": 0.00016131016940932146, + "loss": 0.9532, + "step": 6528 + }, + { + "epoch": 1.1623931623931625, + "grad_norm": 0.6523237228393555, + "learning_rate": 0.00016129911074979635, + "loss": 0.9409, + "step": 6529 + }, + { + "epoch": 1.162571225071225, + "grad_norm": 0.710307776927948, + "learning_rate": 0.00016128805088924252, + "loss": 1.2536, + "step": 6530 + }, + { + "epoch": 1.1627492877492878, + "grad_norm": 0.6349819898605347, + "learning_rate": 0.0001612769898278766, + "loss": 1.0857, + "step": 6531 + }, + { + "epoch": 1.1629273504273505, + "grad_norm": 0.5348139405250549, + "learning_rate": 0.00016126592756591542, + "loss": 0.5969, + "step": 6532 + }, + { + "epoch": 1.163105413105413, + "grad_norm": 0.635619580745697, + "learning_rate": 0.00016125486410357564, + "loss": 0.9885, + "step": 6533 + }, + { + "epoch": 1.1632834757834758, + "grad_norm": 0.6434559226036072, + "learning_rate": 0.000161243799441074, + "loss": 0.8377, + "step": 6534 + }, + { + "epoch": 1.1634615384615385, + "grad_norm": 0.6509647369384766, + "learning_rate": 0.00016123273357862737, + "loss": 0.8393, + "step": 6535 + }, + { + "epoch": 1.163639601139601, + "grad_norm": 0.6179081797599792, + "learning_rate": 0.0001612216665164525, + "loss": 0.9143, + "step": 6536 + }, + { + "epoch": 1.1638176638176638, + "grad_norm": 0.5923223495483398, + "learning_rate": 0.0001612105982547663, + "loss": 1.0185, + "step": 6537 + }, + { + "epoch": 1.1639957264957266, + "grad_norm": 0.702150285243988, + "learning_rate": 0.00016119952879378556, + "loss": 0.863, + "step": 6538 + }, + { + "epoch": 1.164173789173789, + "grad_norm": 0.6596643328666687, + "learning_rate": 0.00016118845813372715, + "loss": 1.0089, + "step": 6539 + }, + { + "epoch": 1.1643518518518519, + "grad_norm": 0.7675769329071045, + "learning_rate": 0.00016117738627480804, + "loss": 1.0179, + "step": 6540 + }, + { + "epoch": 1.1645299145299146, + "grad_norm": 0.6742541193962097, + "learning_rate": 0.00016116631321724513, + "loss": 1.0663, + "step": 6541 + }, + { + "epoch": 1.1647079772079771, + "grad_norm": 0.7379785776138306, + "learning_rate": 0.0001611552389612554, + "loss": 1.0162, + "step": 6542 + }, + { + "epoch": 1.16488603988604, + "grad_norm": 0.5729365944862366, + "learning_rate": 0.00016114416350705577, + "loss": 0.8146, + "step": 6543 + }, + { + "epoch": 1.1650641025641026, + "grad_norm": 0.6481349468231201, + "learning_rate": 0.00016113308685486327, + "loss": 1.0748, + "step": 6544 + }, + { + "epoch": 1.1652421652421652, + "grad_norm": 0.5588181018829346, + "learning_rate": 0.00016112200900489493, + "loss": 0.7511, + "step": 6545 + }, + { + "epoch": 1.165420227920228, + "grad_norm": 0.674363911151886, + "learning_rate": 0.0001611109299573678, + "loss": 0.9852, + "step": 6546 + }, + { + "epoch": 1.1655982905982907, + "grad_norm": 0.6712620854377747, + "learning_rate": 0.00016109984971249893, + "loss": 0.9558, + "step": 6547 + }, + { + "epoch": 1.1657763532763532, + "grad_norm": 0.5260626077651978, + "learning_rate": 0.00016108876827050544, + "loss": 0.7008, + "step": 6548 + }, + { + "epoch": 1.165954415954416, + "grad_norm": 0.6056292057037354, + "learning_rate": 0.00016107768563160445, + "loss": 0.7756, + "step": 6549 + }, + { + "epoch": 1.1661324786324787, + "grad_norm": 0.5725821256637573, + "learning_rate": 0.00016106660179601308, + "loss": 0.8228, + "step": 6550 + }, + { + "epoch": 1.1663105413105412, + "grad_norm": 0.6708397269248962, + "learning_rate": 0.00016105551676394848, + "loss": 1.0711, + "step": 6551 + }, + { + "epoch": 1.166488603988604, + "grad_norm": 0.645453155040741, + "learning_rate": 0.00016104443053562787, + "loss": 0.9299, + "step": 6552 + }, + { + "epoch": 1.1666666666666667, + "grad_norm": 0.6743524074554443, + "learning_rate": 0.00016103334311126847, + "loss": 0.8977, + "step": 6553 + }, + { + "epoch": 1.1668447293447293, + "grad_norm": 0.7248545289039612, + "learning_rate": 0.0001610222544910875, + "loss": 1.2135, + "step": 6554 + }, + { + "epoch": 1.167022792022792, + "grad_norm": 0.5798853635787964, + "learning_rate": 0.00016101116467530217, + "loss": 0.857, + "step": 6555 + }, + { + "epoch": 1.1672008547008548, + "grad_norm": 0.6828082799911499, + "learning_rate": 0.00016100007366412985, + "loss": 0.9405, + "step": 6556 + }, + { + "epoch": 1.1673789173789173, + "grad_norm": 0.6820163130760193, + "learning_rate": 0.0001609889814577878, + "loss": 0.9144, + "step": 6557 + }, + { + "epoch": 1.16755698005698, + "grad_norm": 0.6482275128364563, + "learning_rate": 0.00016097788805649333, + "loss": 0.8586, + "step": 6558 + }, + { + "epoch": 1.1677350427350428, + "grad_norm": 0.6404715180397034, + "learning_rate": 0.00016096679346046385, + "loss": 0.7018, + "step": 6559 + }, + { + "epoch": 1.1679131054131053, + "grad_norm": 0.6315203309059143, + "learning_rate": 0.0001609556976699167, + "loss": 0.9602, + "step": 6560 + }, + { + "epoch": 1.168091168091168, + "grad_norm": 0.5521387457847595, + "learning_rate": 0.00016094460068506925, + "loss": 0.9294, + "step": 6561 + }, + { + "epoch": 1.1682692307692308, + "grad_norm": 0.583372175693512, + "learning_rate": 0.00016093350250613895, + "loss": 1.077, + "step": 6562 + }, + { + "epoch": 1.1684472934472934, + "grad_norm": 0.5990512371063232, + "learning_rate": 0.00016092240313334325, + "loss": 1.0102, + "step": 6563 + }, + { + "epoch": 1.1686253561253561, + "grad_norm": 0.675128161907196, + "learning_rate": 0.00016091130256689964, + "loss": 1.0407, + "step": 6564 + }, + { + "epoch": 1.1688034188034189, + "grad_norm": 0.48797324299812317, + "learning_rate": 0.00016090020080702556, + "loss": 0.7821, + "step": 6565 + }, + { + "epoch": 1.1689814814814814, + "grad_norm": 0.7487484216690063, + "learning_rate": 0.00016088909785393857, + "loss": 1.0444, + "step": 6566 + }, + { + "epoch": 1.1691595441595442, + "grad_norm": 0.6288858652114868, + "learning_rate": 0.00016087799370785618, + "loss": 1.1854, + "step": 6567 + }, + { + "epoch": 1.169337606837607, + "grad_norm": 0.6639021635055542, + "learning_rate": 0.000160866888368996, + "loss": 0.9632, + "step": 6568 + }, + { + "epoch": 1.1695156695156694, + "grad_norm": 0.6553738713264465, + "learning_rate": 0.00016085578183757556, + "loss": 1.2765, + "step": 6569 + }, + { + "epoch": 1.1696937321937322, + "grad_norm": 0.7489066123962402, + "learning_rate": 0.00016084467411381248, + "loss": 1.0705, + "step": 6570 + }, + { + "epoch": 1.169871794871795, + "grad_norm": 0.7079828381538391, + "learning_rate": 0.00016083356519792444, + "loss": 0.8256, + "step": 6571 + }, + { + "epoch": 1.1700498575498575, + "grad_norm": 0.7065926790237427, + "learning_rate": 0.00016082245509012902, + "loss": 1.0439, + "step": 6572 + }, + { + "epoch": 1.1702279202279202, + "grad_norm": 0.6113346815109253, + "learning_rate": 0.00016081134379064395, + "loss": 0.9153, + "step": 6573 + }, + { + "epoch": 1.170405982905983, + "grad_norm": 0.6094171404838562, + "learning_rate": 0.0001608002312996869, + "loss": 0.9723, + "step": 6574 + }, + { + "epoch": 1.1705840455840455, + "grad_norm": 0.6208072900772095, + "learning_rate": 0.00016078911761747565, + "loss": 0.948, + "step": 6575 + }, + { + "epoch": 1.1707621082621082, + "grad_norm": 0.5736680626869202, + "learning_rate": 0.00016077800274422792, + "loss": 0.9155, + "step": 6576 + }, + { + "epoch": 1.170940170940171, + "grad_norm": 0.6793957948684692, + "learning_rate": 0.0001607668866801615, + "loss": 0.9574, + "step": 6577 + }, + { + "epoch": 1.1711182336182335, + "grad_norm": 0.6251805424690247, + "learning_rate": 0.00016075576942549413, + "loss": 1.0319, + "step": 6578 + }, + { + "epoch": 1.1712962962962963, + "grad_norm": 0.628882110118866, + "learning_rate": 0.0001607446509804437, + "loss": 0.9336, + "step": 6579 + }, + { + "epoch": 1.171474358974359, + "grad_norm": 0.6712356805801392, + "learning_rate": 0.000160733531345228, + "loss": 1.0958, + "step": 6580 + }, + { + "epoch": 1.1716524216524216, + "grad_norm": 0.599365770816803, + "learning_rate": 0.0001607224105200649, + "loss": 0.9814, + "step": 6581 + }, + { + "epoch": 1.1718304843304843, + "grad_norm": 0.5798245668411255, + "learning_rate": 0.00016071128850517235, + "loss": 1.0355, + "step": 6582 + }, + { + "epoch": 1.172008547008547, + "grad_norm": 0.7646229863166809, + "learning_rate": 0.00016070016530076817, + "loss": 0.9976, + "step": 6583 + }, + { + "epoch": 1.1721866096866096, + "grad_norm": 0.6371127367019653, + "learning_rate": 0.0001606890409070704, + "loss": 0.9588, + "step": 6584 + }, + { + "epoch": 1.1723646723646723, + "grad_norm": 0.6497066617012024, + "learning_rate": 0.0001606779153242969, + "loss": 0.8817, + "step": 6585 + }, + { + "epoch": 1.172542735042735, + "grad_norm": 0.7255781888961792, + "learning_rate": 0.0001606667885526657, + "loss": 1.1319, + "step": 6586 + }, + { + "epoch": 1.1727207977207976, + "grad_norm": 0.67711341381073, + "learning_rate": 0.00016065566059239483, + "loss": 1.0755, + "step": 6587 + }, + { + "epoch": 1.1728988603988604, + "grad_norm": 0.6159650087356567, + "learning_rate": 0.00016064453144370227, + "loss": 0.9892, + "step": 6588 + }, + { + "epoch": 1.1730769230769231, + "grad_norm": 0.658938467502594, + "learning_rate": 0.00016063340110680609, + "loss": 0.9131, + "step": 6589 + }, + { + "epoch": 1.1732549857549857, + "grad_norm": 0.6754795908927917, + "learning_rate": 0.00016062226958192438, + "loss": 1.0119, + "step": 6590 + }, + { + "epoch": 1.1734330484330484, + "grad_norm": 0.6453405022621155, + "learning_rate": 0.00016061113686927523, + "loss": 0.997, + "step": 6591 + }, + { + "epoch": 1.1736111111111112, + "grad_norm": 0.6580284237861633, + "learning_rate": 0.00016060000296907675, + "loss": 0.8432, + "step": 6592 + }, + { + "epoch": 1.173789173789174, + "grad_norm": 0.6588153839111328, + "learning_rate": 0.00016058886788154712, + "loss": 1.0725, + "step": 6593 + }, + { + "epoch": 1.1739672364672364, + "grad_norm": 0.6247910857200623, + "learning_rate": 0.00016057773160690447, + "loss": 0.8736, + "step": 6594 + }, + { + "epoch": 1.1741452991452992, + "grad_norm": 0.579594075679779, + "learning_rate": 0.000160566594145367, + "loss": 0.8809, + "step": 6595 + }, + { + "epoch": 1.1743233618233617, + "grad_norm": 0.6738116145133972, + "learning_rate": 0.00016055545549715293, + "loss": 0.825, + "step": 6596 + }, + { + "epoch": 1.1745014245014245, + "grad_norm": 0.6658982634544373, + "learning_rate": 0.00016054431566248054, + "loss": 1.0809, + "step": 6597 + }, + { + "epoch": 1.1746794871794872, + "grad_norm": 0.5367915630340576, + "learning_rate": 0.00016053317464156803, + "loss": 0.9005, + "step": 6598 + }, + { + "epoch": 1.17485754985755, + "grad_norm": 0.7243228554725647, + "learning_rate": 0.00016052203243463372, + "loss": 1.0573, + "step": 6599 + }, + { + "epoch": 1.1750356125356125, + "grad_norm": 0.6359432935714722, + "learning_rate": 0.0001605108890418959, + "loss": 0.8569, + "step": 6600 + }, + { + "epoch": 1.1752136752136753, + "grad_norm": 0.6565225720405579, + "learning_rate": 0.0001604997444635729, + "loss": 0.9748, + "step": 6601 + }, + { + "epoch": 1.1753917378917378, + "grad_norm": 0.7124663591384888, + "learning_rate": 0.0001604885986998831, + "loss": 1.0271, + "step": 6602 + }, + { + "epoch": 1.1755698005698005, + "grad_norm": 0.659766435623169, + "learning_rate": 0.00016047745175104487, + "loss": 1.0635, + "step": 6603 + }, + { + "epoch": 1.1757478632478633, + "grad_norm": 0.5874318480491638, + "learning_rate": 0.00016046630361727656, + "loss": 0.9257, + "step": 6604 + }, + { + "epoch": 1.175925925925926, + "grad_norm": 0.587345540523529, + "learning_rate": 0.0001604551542987967, + "loss": 1.0759, + "step": 6605 + }, + { + "epoch": 1.1761039886039886, + "grad_norm": 0.733567476272583, + "learning_rate": 0.00016044400379582364, + "loss": 0.9877, + "step": 6606 + }, + { + "epoch": 1.1762820512820513, + "grad_norm": 0.6538317203521729, + "learning_rate": 0.0001604328521085759, + "loss": 1.0094, + "step": 6607 + }, + { + "epoch": 1.1764601139601139, + "grad_norm": 0.6279696822166443, + "learning_rate": 0.00016042169923727195, + "loss": 1.1049, + "step": 6608 + }, + { + "epoch": 1.1766381766381766, + "grad_norm": 0.6949752569198608, + "learning_rate": 0.00016041054518213033, + "loss": 1.1418, + "step": 6609 + }, + { + "epoch": 1.1768162393162394, + "grad_norm": 0.6144010424613953, + "learning_rate": 0.00016039938994336957, + "loss": 1.0306, + "step": 6610 + }, + { + "epoch": 1.176994301994302, + "grad_norm": 0.5868683457374573, + "learning_rate": 0.00016038823352120823, + "loss": 0.9894, + "step": 6611 + }, + { + "epoch": 1.1771723646723646, + "grad_norm": 0.7181115746498108, + "learning_rate": 0.0001603770759158649, + "loss": 1.1674, + "step": 6612 + }, + { + "epoch": 1.1773504273504274, + "grad_norm": 0.6271308064460754, + "learning_rate": 0.00016036591712755818, + "loss": 0.9726, + "step": 6613 + }, + { + "epoch": 1.17752849002849, + "grad_norm": 0.6922675371170044, + "learning_rate": 0.00016035475715650668, + "loss": 0.9142, + "step": 6614 + }, + { + "epoch": 1.1777065527065527, + "grad_norm": 0.6838833689689636, + "learning_rate": 0.00016034359600292913, + "loss": 1.1627, + "step": 6615 + }, + { + "epoch": 1.1778846153846154, + "grad_norm": 0.6628252267837524, + "learning_rate": 0.00016033243366704418, + "loss": 0.739, + "step": 6616 + }, + { + "epoch": 1.1780626780626782, + "grad_norm": 0.6367576122283936, + "learning_rate": 0.0001603212701490705, + "loss": 0.9015, + "step": 6617 + }, + { + "epoch": 1.1782407407407407, + "grad_norm": 0.6498967409133911, + "learning_rate": 0.00016031010544922687, + "loss": 0.9645, + "step": 6618 + }, + { + "epoch": 1.1784188034188035, + "grad_norm": 0.468795508146286, + "learning_rate": 0.00016029893956773198, + "loss": 0.7305, + "step": 6619 + }, + { + "epoch": 1.178596866096866, + "grad_norm": 0.6355500817298889, + "learning_rate": 0.00016028777250480465, + "loss": 0.9183, + "step": 6620 + }, + { + "epoch": 1.1787749287749287, + "grad_norm": 0.7582615613937378, + "learning_rate": 0.0001602766042606636, + "loss": 1.1641, + "step": 6621 + }, + { + "epoch": 1.1789529914529915, + "grad_norm": 0.580035924911499, + "learning_rate": 0.00016026543483552776, + "loss": 0.9164, + "step": 6622 + }, + { + "epoch": 1.1791310541310542, + "grad_norm": 0.6198559999465942, + "learning_rate": 0.00016025426422961592, + "loss": 0.9803, + "step": 6623 + }, + { + "epoch": 1.1793091168091168, + "grad_norm": 0.59112149477005, + "learning_rate": 0.0001602430924431469, + "loss": 0.8645, + "step": 6624 + }, + { + "epoch": 1.1794871794871795, + "grad_norm": 0.6200533509254456, + "learning_rate": 0.00016023191947633965, + "loss": 1.068, + "step": 6625 + }, + { + "epoch": 1.179665242165242, + "grad_norm": 0.6077516078948975, + "learning_rate": 0.00016022074532941305, + "loss": 1.0017, + "step": 6626 + }, + { + "epoch": 1.1798433048433048, + "grad_norm": 0.6770145893096924, + "learning_rate": 0.00016020957000258606, + "loss": 0.9022, + "step": 6627 + }, + { + "epoch": 1.1800213675213675, + "grad_norm": 0.6478054523468018, + "learning_rate": 0.0001601983934960776, + "loss": 0.8615, + "step": 6628 + }, + { + "epoch": 1.1801994301994303, + "grad_norm": 0.6528988480567932, + "learning_rate": 0.00016018721581010666, + "loss": 1.0015, + "step": 6629 + }, + { + "epoch": 1.1803774928774928, + "grad_norm": 0.6160712242126465, + "learning_rate": 0.0001601760369448923, + "loss": 0.9382, + "step": 6630 + }, + { + "epoch": 1.1805555555555556, + "grad_norm": 0.5755789875984192, + "learning_rate": 0.00016016485690065345, + "loss": 1.0551, + "step": 6631 + }, + { + "epoch": 1.180733618233618, + "grad_norm": 0.8495022654533386, + "learning_rate": 0.00016015367567760925, + "loss": 0.9295, + "step": 6632 + }, + { + "epoch": 1.1809116809116809, + "grad_norm": 0.6010929346084595, + "learning_rate": 0.0001601424932759787, + "loss": 1.0413, + "step": 6633 + }, + { + "epoch": 1.1810897435897436, + "grad_norm": 0.6953579187393188, + "learning_rate": 0.00016013130969598093, + "loss": 1.0149, + "step": 6634 + }, + { + "epoch": 1.1812678062678064, + "grad_norm": 0.6949529647827148, + "learning_rate": 0.0001601201249378351, + "loss": 0.9992, + "step": 6635 + }, + { + "epoch": 1.181445868945869, + "grad_norm": 0.6471893787384033, + "learning_rate": 0.00016010893900176028, + "loss": 0.7985, + "step": 6636 + }, + { + "epoch": 1.1816239316239316, + "grad_norm": 0.6524858474731445, + "learning_rate": 0.00016009775188797568, + "loss": 0.9517, + "step": 6637 + }, + { + "epoch": 1.1818019943019944, + "grad_norm": 0.639214038848877, + "learning_rate": 0.00016008656359670046, + "loss": 1.0357, + "step": 6638 + }, + { + "epoch": 1.181980056980057, + "grad_norm": 0.6039628386497498, + "learning_rate": 0.00016007537412815386, + "loss": 1.0536, + "step": 6639 + }, + { + "epoch": 1.1821581196581197, + "grad_norm": 0.653540313243866, + "learning_rate": 0.00016006418348255507, + "loss": 0.9414, + "step": 6640 + }, + { + "epoch": 1.1823361823361824, + "grad_norm": 0.6331741809844971, + "learning_rate": 0.0001600529916601234, + "loss": 1.0352, + "step": 6641 + }, + { + "epoch": 1.182514245014245, + "grad_norm": 0.7552719712257385, + "learning_rate": 0.00016004179866107812, + "loss": 1.1103, + "step": 6642 + }, + { + "epoch": 1.1826923076923077, + "grad_norm": 0.6795875430107117, + "learning_rate": 0.00016003060448563852, + "loss": 1.1246, + "step": 6643 + }, + { + "epoch": 1.1828703703703705, + "grad_norm": 0.6308842301368713, + "learning_rate": 0.0001600194091340239, + "loss": 0.9532, + "step": 6644 + }, + { + "epoch": 1.183048433048433, + "grad_norm": 0.5640553832054138, + "learning_rate": 0.00016000821260645366, + "loss": 0.7491, + "step": 6645 + }, + { + "epoch": 1.1832264957264957, + "grad_norm": 0.5611832141876221, + "learning_rate": 0.00015999701490314712, + "loss": 0.9239, + "step": 6646 + }, + { + "epoch": 1.1834045584045585, + "grad_norm": 0.5881187915802002, + "learning_rate": 0.00015998581602432374, + "loss": 0.9246, + "step": 6647 + }, + { + "epoch": 1.183582621082621, + "grad_norm": 0.7291010022163391, + "learning_rate": 0.00015997461597020291, + "loss": 1.0314, + "step": 6648 + }, + { + "epoch": 1.1837606837606838, + "grad_norm": 0.6784794926643372, + "learning_rate": 0.00015996341474100402, + "loss": 1.0011, + "step": 6649 + }, + { + "epoch": 1.1839387464387465, + "grad_norm": 0.7083746194839478, + "learning_rate": 0.00015995221233694663, + "loss": 1.0336, + "step": 6650 + }, + { + "epoch": 1.184116809116809, + "grad_norm": 0.7081790566444397, + "learning_rate": 0.00015994100875825015, + "loss": 1.2386, + "step": 6651 + }, + { + "epoch": 1.1842948717948718, + "grad_norm": 0.5938812494277954, + "learning_rate": 0.00015992980400513415, + "loss": 0.7549, + "step": 6652 + }, + { + "epoch": 1.1844729344729346, + "grad_norm": 0.7084267139434814, + "learning_rate": 0.00015991859807781811, + "loss": 1.1194, + "step": 6653 + }, + { + "epoch": 1.184650997150997, + "grad_norm": 0.6391362547874451, + "learning_rate": 0.0001599073909765216, + "loss": 1.0857, + "step": 6654 + }, + { + "epoch": 1.1848290598290598, + "grad_norm": 0.8074106574058533, + "learning_rate": 0.00015989618270146423, + "loss": 1.1715, + "step": 6655 + }, + { + "epoch": 1.1850071225071226, + "grad_norm": 0.5778565406799316, + "learning_rate": 0.0001598849732528656, + "loss": 0.8843, + "step": 6656 + }, + { + "epoch": 1.1851851851851851, + "grad_norm": 0.6955079436302185, + "learning_rate": 0.00015987376263094526, + "loss": 1.0281, + "step": 6657 + }, + { + "epoch": 1.1853632478632479, + "grad_norm": 0.6789296269416809, + "learning_rate": 0.00015986255083592297, + "loss": 0.9739, + "step": 6658 + }, + { + "epoch": 1.1855413105413106, + "grad_norm": 0.6294292211532593, + "learning_rate": 0.00015985133786801834, + "loss": 1.0692, + "step": 6659 + }, + { + "epoch": 1.1857193732193732, + "grad_norm": 0.5604581832885742, + "learning_rate": 0.00015984012372745107, + "loss": 0.9059, + "step": 6660 + }, + { + "epoch": 1.185897435897436, + "grad_norm": 0.6727550625801086, + "learning_rate": 0.00015982890841444088, + "loss": 1.049, + "step": 6661 + }, + { + "epoch": 1.1860754985754987, + "grad_norm": 0.620914101600647, + "learning_rate": 0.0001598176919292075, + "loss": 1.1021, + "step": 6662 + }, + { + "epoch": 1.1862535612535612, + "grad_norm": 0.6696683168411255, + "learning_rate": 0.00015980647427197076, + "loss": 0.9053, + "step": 6663 + }, + { + "epoch": 1.186431623931624, + "grad_norm": 0.6713385581970215, + "learning_rate": 0.00015979525544295036, + "loss": 0.9596, + "step": 6664 + }, + { + "epoch": 1.1866096866096867, + "grad_norm": 0.7643477320671082, + "learning_rate": 0.00015978403544236614, + "loss": 0.882, + "step": 6665 + }, + { + "epoch": 1.1867877492877492, + "grad_norm": 0.5890966057777405, + "learning_rate": 0.00015977281427043794, + "loss": 1.0215, + "step": 6666 + }, + { + "epoch": 1.186965811965812, + "grad_norm": 0.7287502288818359, + "learning_rate": 0.0001597615919273856, + "loss": 1.0111, + "step": 6667 + }, + { + "epoch": 1.1871438746438747, + "grad_norm": 0.5713803172111511, + "learning_rate": 0.00015975036841342903, + "loss": 1.0068, + "step": 6668 + }, + { + "epoch": 1.1873219373219372, + "grad_norm": 0.5113094449043274, + "learning_rate": 0.0001597391437287881, + "loss": 0.9018, + "step": 6669 + }, + { + "epoch": 1.1875, + "grad_norm": 0.585640013217926, + "learning_rate": 0.00015972791787368276, + "loss": 1.0375, + "step": 6670 + }, + { + "epoch": 1.1876780626780628, + "grad_norm": 0.5778326392173767, + "learning_rate": 0.00015971669084833293, + "loss": 0.9975, + "step": 6671 + }, + { + "epoch": 1.1878561253561253, + "grad_norm": 0.6707763075828552, + "learning_rate": 0.0001597054626529586, + "loss": 1.0048, + "step": 6672 + }, + { + "epoch": 1.188034188034188, + "grad_norm": 0.6113292574882507, + "learning_rate": 0.00015969423328777974, + "loss": 1.1447, + "step": 6673 + }, + { + "epoch": 1.1882122507122508, + "grad_norm": 0.6075651049613953, + "learning_rate": 0.00015968300275301638, + "loss": 0.9212, + "step": 6674 + }, + { + "epoch": 1.1883903133903133, + "grad_norm": 0.6990494132041931, + "learning_rate": 0.00015967177104888857, + "loss": 0.9952, + "step": 6675 + }, + { + "epoch": 1.188568376068376, + "grad_norm": 0.6228706240653992, + "learning_rate": 0.00015966053817561638, + "loss": 1.0187, + "step": 6676 + }, + { + "epoch": 1.1887464387464388, + "grad_norm": 0.6387844085693359, + "learning_rate": 0.00015964930413341985, + "loss": 1.1614, + "step": 6677 + }, + { + "epoch": 1.1889245014245013, + "grad_norm": 0.6501925587654114, + "learning_rate": 0.00015963806892251915, + "loss": 1.0366, + "step": 6678 + }, + { + "epoch": 1.189102564102564, + "grad_norm": 0.6923739910125732, + "learning_rate": 0.00015962683254313435, + "loss": 1.1992, + "step": 6679 + }, + { + "epoch": 1.1892806267806268, + "grad_norm": 0.6640275120735168, + "learning_rate": 0.00015961559499548563, + "loss": 0.8883, + "step": 6680 + }, + { + "epoch": 1.1894586894586894, + "grad_norm": 0.6493857502937317, + "learning_rate": 0.00015960435627979317, + "loss": 1.1368, + "step": 6681 + }, + { + "epoch": 1.1896367521367521, + "grad_norm": 0.6357189416885376, + "learning_rate": 0.0001595931163962772, + "loss": 1.0502, + "step": 6682 + }, + { + "epoch": 1.1898148148148149, + "grad_norm": 0.5756343007087708, + "learning_rate": 0.0001595818753451579, + "loss": 0.9871, + "step": 6683 + }, + { + "epoch": 1.1899928774928774, + "grad_norm": 0.7369210124015808, + "learning_rate": 0.0001595706331266555, + "loss": 1.3229, + "step": 6684 + }, + { + "epoch": 1.1901709401709402, + "grad_norm": 0.7140820622444153, + "learning_rate": 0.0001595593897409903, + "loss": 1.1154, + "step": 6685 + }, + { + "epoch": 1.190349002849003, + "grad_norm": 0.696973443031311, + "learning_rate": 0.00015954814518838255, + "loss": 0.9806, + "step": 6686 + }, + { + "epoch": 1.1905270655270654, + "grad_norm": 0.5299260020256042, + "learning_rate": 0.00015953689946905262, + "loss": 0.771, + "step": 6687 + }, + { + "epoch": 1.1907051282051282, + "grad_norm": 0.6814879775047302, + "learning_rate": 0.00015952565258322085, + "loss": 0.8444, + "step": 6688 + }, + { + "epoch": 1.190883190883191, + "grad_norm": 0.6215870976448059, + "learning_rate": 0.00015951440453110754, + "loss": 1.0743, + "step": 6689 + }, + { + "epoch": 1.1910612535612535, + "grad_norm": 0.7017203569412231, + "learning_rate": 0.00015950315531293308, + "loss": 1.185, + "step": 6690 + }, + { + "epoch": 1.1912393162393162, + "grad_norm": 0.7147250175476074, + "learning_rate": 0.00015949190492891795, + "loss": 1.0646, + "step": 6691 + }, + { + "epoch": 1.191417378917379, + "grad_norm": 0.5867117047309875, + "learning_rate": 0.00015948065337928252, + "loss": 1.0554, + "step": 6692 + }, + { + "epoch": 1.1915954415954415, + "grad_norm": 0.6813527345657349, + "learning_rate": 0.0001594694006642472, + "loss": 1.1451, + "step": 6693 + }, + { + "epoch": 1.1917735042735043, + "grad_norm": 0.5192593932151794, + "learning_rate": 0.00015945814678403256, + "loss": 0.7886, + "step": 6694 + }, + { + "epoch": 1.191951566951567, + "grad_norm": 0.6537744402885437, + "learning_rate": 0.00015944689173885904, + "loss": 0.9905, + "step": 6695 + }, + { + "epoch": 1.1921296296296295, + "grad_norm": 0.7350276112556458, + "learning_rate": 0.00015943563552894716, + "loss": 0.9009, + "step": 6696 + }, + { + "epoch": 1.1923076923076923, + "grad_norm": 0.7086381316184998, + "learning_rate": 0.00015942437815451746, + "loss": 0.9117, + "step": 6697 + }, + { + "epoch": 1.192485754985755, + "grad_norm": 0.6774969696998596, + "learning_rate": 0.00015941311961579054, + "loss": 1.1172, + "step": 6698 + }, + { + "epoch": 1.1926638176638176, + "grad_norm": 0.7034362554550171, + "learning_rate": 0.00015940185991298694, + "loss": 0.8054, + "step": 6699 + }, + { + "epoch": 1.1928418803418803, + "grad_norm": 0.66145920753479, + "learning_rate": 0.00015939059904632728, + "loss": 0.7417, + "step": 6700 + }, + { + "epoch": 1.193019943019943, + "grad_norm": 0.6590890884399414, + "learning_rate": 0.00015937933701603223, + "loss": 0.9169, + "step": 6701 + }, + { + "epoch": 1.1931980056980056, + "grad_norm": 0.7492850422859192, + "learning_rate": 0.0001593680738223224, + "loss": 1.0529, + "step": 6702 + }, + { + "epoch": 1.1933760683760684, + "grad_norm": 0.7103236317634583, + "learning_rate": 0.00015935680946541848, + "loss": 1.1377, + "step": 6703 + }, + { + "epoch": 1.193554131054131, + "grad_norm": 0.6164175868034363, + "learning_rate": 0.00015934554394554122, + "loss": 0.8636, + "step": 6704 + }, + { + "epoch": 1.1937321937321936, + "grad_norm": 0.6667410135269165, + "learning_rate": 0.0001593342772629113, + "loss": 1.0073, + "step": 6705 + }, + { + "epoch": 1.1939102564102564, + "grad_norm": 0.6785695552825928, + "learning_rate": 0.00015932300941774944, + "loss": 1.0752, + "step": 6706 + }, + { + "epoch": 1.1940883190883191, + "grad_norm": 0.6446872353553772, + "learning_rate": 0.0001593117404102765, + "loss": 0.9509, + "step": 6707 + }, + { + "epoch": 1.194266381766382, + "grad_norm": 0.6607686877250671, + "learning_rate": 0.00015930047024071317, + "loss": 1.0902, + "step": 6708 + }, + { + "epoch": 1.1944444444444444, + "grad_norm": 0.664804995059967, + "learning_rate": 0.0001592891989092803, + "loss": 0.9783, + "step": 6709 + }, + { + "epoch": 1.1946225071225072, + "grad_norm": 0.7147907018661499, + "learning_rate": 0.00015927792641619876, + "loss": 1.0558, + "step": 6710 + }, + { + "epoch": 1.1948005698005697, + "grad_norm": 0.6858944296836853, + "learning_rate": 0.0001592666527616894, + "loss": 1.0514, + "step": 6711 + }, + { + "epoch": 1.1949786324786325, + "grad_norm": 0.598463773727417, + "learning_rate": 0.0001592553779459731, + "loss": 0.8927, + "step": 6712 + }, + { + "epoch": 1.1951566951566952, + "grad_norm": 0.6872668862342834, + "learning_rate": 0.00015924410196927076, + "loss": 1.016, + "step": 6713 + }, + { + "epoch": 1.195334757834758, + "grad_norm": 0.6547996401786804, + "learning_rate": 0.00015923282483180326, + "loss": 1.1573, + "step": 6714 + }, + { + "epoch": 1.1955128205128205, + "grad_norm": 0.6254705786705017, + "learning_rate": 0.00015922154653379167, + "loss": 1.0179, + "step": 6715 + }, + { + "epoch": 1.1956908831908832, + "grad_norm": 0.6049207448959351, + "learning_rate": 0.00015921026707545684, + "loss": 1.0713, + "step": 6716 + }, + { + "epoch": 1.1958689458689458, + "grad_norm": 0.6042858958244324, + "learning_rate": 0.0001591989864570199, + "loss": 0.919, + "step": 6717 + }, + { + "epoch": 1.1960470085470085, + "grad_norm": 0.6521187424659729, + "learning_rate": 0.0001591877046787017, + "loss": 1.0112, + "step": 6718 + }, + { + "epoch": 1.1962250712250713, + "grad_norm": 0.766260027885437, + "learning_rate": 0.00015917642174072348, + "loss": 0.9774, + "step": 6719 + }, + { + "epoch": 1.196403133903134, + "grad_norm": 0.7066532373428345, + "learning_rate": 0.00015916513764330613, + "loss": 1.1112, + "step": 6720 + }, + { + "epoch": 1.1965811965811965, + "grad_norm": 0.7351508140563965, + "learning_rate": 0.00015915385238667083, + "loss": 0.9841, + "step": 6721 + }, + { + "epoch": 1.1967592592592593, + "grad_norm": 0.6133812069892883, + "learning_rate": 0.0001591425659710387, + "loss": 0.8629, + "step": 6722 + }, + { + "epoch": 1.1969373219373218, + "grad_norm": 0.7244157791137695, + "learning_rate": 0.00015913127839663083, + "loss": 1.1584, + "step": 6723 + }, + { + "epoch": 1.1971153846153846, + "grad_norm": 0.5986210107803345, + "learning_rate": 0.00015911998966366842, + "loss": 0.8507, + "step": 6724 + }, + { + "epoch": 1.1972934472934473, + "grad_norm": 0.6087439060211182, + "learning_rate": 0.00015910869977237257, + "loss": 0.884, + "step": 6725 + }, + { + "epoch": 1.19747150997151, + "grad_norm": 0.7546007633209229, + "learning_rate": 0.00015909740872296457, + "loss": 1.1449, + "step": 6726 + }, + { + "epoch": 1.1976495726495726, + "grad_norm": 0.6437731385231018, + "learning_rate": 0.0001590861165156656, + "loss": 0.7845, + "step": 6727 + }, + { + "epoch": 1.1978276353276354, + "grad_norm": 0.6281737089157104, + "learning_rate": 0.00015907482315069693, + "loss": 0.8969, + "step": 6728 + }, + { + "epoch": 1.198005698005698, + "grad_norm": 0.6196113228797913, + "learning_rate": 0.00015906352862827983, + "loss": 1.0264, + "step": 6729 + }, + { + "epoch": 1.1981837606837606, + "grad_norm": 0.5990965962409973, + "learning_rate": 0.00015905223294863553, + "loss": 1.0017, + "step": 6730 + }, + { + "epoch": 1.1983618233618234, + "grad_norm": 0.6509191393852234, + "learning_rate": 0.00015904093611198542, + "loss": 1.1066, + "step": 6731 + }, + { + "epoch": 1.1985398860398861, + "grad_norm": 0.6648043990135193, + "learning_rate": 0.00015902963811855085, + "loss": 1.077, + "step": 6732 + }, + { + "epoch": 1.1987179487179487, + "grad_norm": 0.7071963548660278, + "learning_rate": 0.00015901833896855307, + "loss": 1.1346, + "step": 6733 + }, + { + "epoch": 1.1988960113960114, + "grad_norm": 0.5889959335327148, + "learning_rate": 0.0001590070386622136, + "loss": 0.9525, + "step": 6734 + }, + { + "epoch": 1.199074074074074, + "grad_norm": 0.6233037710189819, + "learning_rate": 0.00015899573719975376, + "loss": 1.0513, + "step": 6735 + }, + { + "epoch": 1.1992521367521367, + "grad_norm": 0.7912302613258362, + "learning_rate": 0.000158984434581395, + "loss": 0.8749, + "step": 6736 + }, + { + "epoch": 1.1994301994301995, + "grad_norm": 0.5783160924911499, + "learning_rate": 0.0001589731308073588, + "loss": 0.7173, + "step": 6737 + }, + { + "epoch": 1.1996082621082622, + "grad_norm": 0.718950092792511, + "learning_rate": 0.00015896182587786658, + "loss": 1.0815, + "step": 6738 + }, + { + "epoch": 1.1997863247863247, + "grad_norm": 0.6700926423072815, + "learning_rate": 0.0001589505197931399, + "loss": 1.0817, + "step": 6739 + }, + { + "epoch": 1.1999643874643875, + "grad_norm": 0.7614455223083496, + "learning_rate": 0.0001589392125534002, + "loss": 0.9707, + "step": 6740 + }, + { + "epoch": 1.20014245014245, + "grad_norm": 0.6998619437217712, + "learning_rate": 0.00015892790415886906, + "loss": 1.0541, + "step": 6741 + }, + { + "epoch": 1.2003205128205128, + "grad_norm": 0.6127668619155884, + "learning_rate": 0.0001589165946097681, + "loss": 0.9147, + "step": 6742 + }, + { + "epoch": 1.2004985754985755, + "grad_norm": 0.7112005352973938, + "learning_rate": 0.00015890528390631885, + "loss": 0.868, + "step": 6743 + }, + { + "epoch": 1.2006766381766383, + "grad_norm": 0.6631024479866028, + "learning_rate": 0.0001588939720487429, + "loss": 0.9277, + "step": 6744 + }, + { + "epoch": 1.2008547008547008, + "grad_norm": 0.6106321215629578, + "learning_rate": 0.00015888265903726188, + "loss": 1.0223, + "step": 6745 + }, + { + "epoch": 1.2010327635327636, + "grad_norm": 0.6400851607322693, + "learning_rate": 0.00015887134487209753, + "loss": 1.1279, + "step": 6746 + }, + { + "epoch": 1.201210826210826, + "grad_norm": 0.6298650503158569, + "learning_rate": 0.00015886002955347147, + "loss": 0.9481, + "step": 6747 + }, + { + "epoch": 1.2013888888888888, + "grad_norm": 0.647974967956543, + "learning_rate": 0.00015884871308160538, + "loss": 1.1513, + "step": 6748 + }, + { + "epoch": 1.2015669515669516, + "grad_norm": 0.6770651936531067, + "learning_rate": 0.000158837395456721, + "loss": 0.9914, + "step": 6749 + }, + { + "epoch": 1.2017450142450143, + "grad_norm": 0.6708947420120239, + "learning_rate": 0.0001588260766790401, + "loss": 1.1848, + "step": 6750 + }, + { + "epoch": 1.2019230769230769, + "grad_norm": 0.5624440908432007, + "learning_rate": 0.00015881475674878442, + "loss": 0.9848, + "step": 6751 + }, + { + "epoch": 1.2021011396011396, + "grad_norm": 0.5512633919715881, + "learning_rate": 0.00015880343566617575, + "loss": 1.0308, + "step": 6752 + }, + { + "epoch": 1.2022792022792024, + "grad_norm": 0.5621042251586914, + "learning_rate": 0.0001587921134314359, + "loss": 0.8724, + "step": 6753 + }, + { + "epoch": 1.202457264957265, + "grad_norm": 0.6881251931190491, + "learning_rate": 0.00015878079004478675, + "loss": 0.9771, + "step": 6754 + }, + { + "epoch": 1.2026353276353277, + "grad_norm": 0.729998767375946, + "learning_rate": 0.0001587694655064501, + "loss": 1.002, + "step": 6755 + }, + { + "epoch": 1.2028133903133904, + "grad_norm": 0.5972567200660706, + "learning_rate": 0.00015875813981664787, + "loss": 1.0571, + "step": 6756 + }, + { + "epoch": 1.202991452991453, + "grad_norm": 0.6319229006767273, + "learning_rate": 0.00015874681297560196, + "loss": 0.9294, + "step": 6757 + }, + { + "epoch": 1.2031695156695157, + "grad_norm": 0.6751521825790405, + "learning_rate": 0.00015873548498353428, + "loss": 0.783, + "step": 6758 + }, + { + "epoch": 1.2033475783475784, + "grad_norm": 0.6476554870605469, + "learning_rate": 0.00015872415584066677, + "loss": 0.8939, + "step": 6759 + }, + { + "epoch": 1.203525641025641, + "grad_norm": 0.6530960202217102, + "learning_rate": 0.0001587128255472214, + "loss": 0.9828, + "step": 6760 + }, + { + "epoch": 1.2037037037037037, + "grad_norm": 0.6708502173423767, + "learning_rate": 0.00015870149410342023, + "loss": 0.9285, + "step": 6761 + }, + { + "epoch": 1.2038817663817665, + "grad_norm": 0.7749543190002441, + "learning_rate": 0.0001586901615094852, + "loss": 1.1295, + "step": 6762 + }, + { + "epoch": 1.204059829059829, + "grad_norm": 0.6750495433807373, + "learning_rate": 0.00015867882776563836, + "loss": 1.0562, + "step": 6763 + }, + { + "epoch": 1.2042378917378918, + "grad_norm": 0.6892416477203369, + "learning_rate": 0.00015866749287210178, + "loss": 0.7207, + "step": 6764 + }, + { + "epoch": 1.2044159544159545, + "grad_norm": 0.7066485285758972, + "learning_rate": 0.00015865615682909758, + "loss": 1.0489, + "step": 6765 + }, + { + "epoch": 1.204594017094017, + "grad_norm": 0.5669938325881958, + "learning_rate": 0.00015864481963684783, + "loss": 0.8149, + "step": 6766 + }, + { + "epoch": 1.2047720797720798, + "grad_norm": 0.6467341780662537, + "learning_rate": 0.0001586334812955746, + "loss": 0.9595, + "step": 6767 + }, + { + "epoch": 1.2049501424501425, + "grad_norm": 0.6026045680046082, + "learning_rate": 0.0001586221418055002, + "loss": 0.9832, + "step": 6768 + }, + { + "epoch": 1.205128205128205, + "grad_norm": 0.7655174732208252, + "learning_rate": 0.00015861080116684665, + "loss": 0.9796, + "step": 6769 + }, + { + "epoch": 1.2053062678062678, + "grad_norm": 0.6386621594429016, + "learning_rate": 0.00015859945937983624, + "loss": 0.9368, + "step": 6770 + }, + { + "epoch": 1.2054843304843306, + "grad_norm": 0.7088032364845276, + "learning_rate": 0.0001585881164446911, + "loss": 1.0167, + "step": 6771 + }, + { + "epoch": 1.205662393162393, + "grad_norm": 0.6015275716781616, + "learning_rate": 0.0001585767723616336, + "loss": 0.8551, + "step": 6772 + }, + { + "epoch": 1.2058404558404558, + "grad_norm": 0.7013260722160339, + "learning_rate": 0.00015856542713088583, + "loss": 0.8009, + "step": 6773 + }, + { + "epoch": 1.2060185185185186, + "grad_norm": 0.6931240558624268, + "learning_rate": 0.00015855408075267024, + "loss": 0.9964, + "step": 6774 + }, + { + "epoch": 1.2061965811965811, + "grad_norm": 0.7274388670921326, + "learning_rate": 0.00015854273322720908, + "loss": 1.0991, + "step": 6775 + }, + { + "epoch": 1.2063746438746439, + "grad_norm": 0.6353716254234314, + "learning_rate": 0.00015853138455472466, + "loss": 1.0893, + "step": 6776 + }, + { + "epoch": 1.2065527065527066, + "grad_norm": 0.6958979368209839, + "learning_rate": 0.00015852003473543932, + "loss": 1.0238, + "step": 6777 + }, + { + "epoch": 1.2067307692307692, + "grad_norm": 0.626838743686676, + "learning_rate": 0.00015850868376957551, + "loss": 0.9384, + "step": 6778 + }, + { + "epoch": 1.206908831908832, + "grad_norm": 0.5455024242401123, + "learning_rate": 0.00015849733165735556, + "loss": 0.8068, + "step": 6779 + }, + { + "epoch": 1.2070868945868947, + "grad_norm": 0.6337353587150574, + "learning_rate": 0.0001584859783990019, + "loss": 1.1341, + "step": 6780 + }, + { + "epoch": 1.2072649572649572, + "grad_norm": 0.6318019032478333, + "learning_rate": 0.000158474623994737, + "loss": 1.1095, + "step": 6781 + }, + { + "epoch": 1.20744301994302, + "grad_norm": 0.8183810710906982, + "learning_rate": 0.00015846326844478332, + "loss": 1.1471, + "step": 6782 + }, + { + "epoch": 1.2076210826210827, + "grad_norm": 0.6140483021736145, + "learning_rate": 0.00015845191174936334, + "loss": 0.8538, + "step": 6783 + }, + { + "epoch": 1.2077991452991452, + "grad_norm": 0.7570197582244873, + "learning_rate": 0.0001584405539086996, + "loss": 1.427, + "step": 6784 + }, + { + "epoch": 1.207977207977208, + "grad_norm": 0.7616991996765137, + "learning_rate": 0.00015842919492301455, + "loss": 1.2214, + "step": 6785 + }, + { + "epoch": 1.2081552706552707, + "grad_norm": 0.561996579170227, + "learning_rate": 0.00015841783479253084, + "loss": 0.8916, + "step": 6786 + }, + { + "epoch": 1.2083333333333333, + "grad_norm": 0.6124222874641418, + "learning_rate": 0.000158406473517471, + "loss": 0.9637, + "step": 6787 + }, + { + "epoch": 1.208511396011396, + "grad_norm": 0.6053098440170288, + "learning_rate": 0.00015839511109805762, + "loss": 1.0365, + "step": 6788 + }, + { + "epoch": 1.2086894586894588, + "grad_norm": 0.6451675295829773, + "learning_rate": 0.00015838374753451338, + "loss": 1.0497, + "step": 6789 + }, + { + "epoch": 1.2088675213675213, + "grad_norm": 0.6789399981498718, + "learning_rate": 0.00015837238282706087, + "loss": 0.9286, + "step": 6790 + }, + { + "epoch": 1.209045584045584, + "grad_norm": 0.5742998123168945, + "learning_rate": 0.0001583610169759228, + "loss": 1.082, + "step": 6791 + }, + { + "epoch": 1.2092236467236468, + "grad_norm": 0.6813693642616272, + "learning_rate": 0.0001583496499813218, + "loss": 0.9785, + "step": 6792 + }, + { + "epoch": 1.2094017094017093, + "grad_norm": 0.6150603890419006, + "learning_rate": 0.0001583382818434806, + "loss": 0.9533, + "step": 6793 + }, + { + "epoch": 1.209579772079772, + "grad_norm": 0.6905919909477234, + "learning_rate": 0.000158326912562622, + "loss": 1.0132, + "step": 6794 + }, + { + "epoch": 1.2097578347578348, + "grad_norm": 0.5861411094665527, + "learning_rate": 0.0001583155421389687, + "loss": 0.7071, + "step": 6795 + }, + { + "epoch": 1.2099358974358974, + "grad_norm": 0.6822740435600281, + "learning_rate": 0.0001583041705727435, + "loss": 1.1366, + "step": 6796 + }, + { + "epoch": 1.21011396011396, + "grad_norm": 0.6013675928115845, + "learning_rate": 0.00015829279786416916, + "loss": 0.9232, + "step": 6797 + }, + { + "epoch": 1.2102920227920229, + "grad_norm": 0.650675356388092, + "learning_rate": 0.00015828142401346857, + "loss": 0.887, + "step": 6798 + }, + { + "epoch": 1.2104700854700854, + "grad_norm": 0.6764078736305237, + "learning_rate": 0.00015827004902086456, + "loss": 0.8423, + "step": 6799 + }, + { + "epoch": 1.2106481481481481, + "grad_norm": 0.6460821628570557, + "learning_rate": 0.00015825867288657994, + "loss": 1.0074, + "step": 6800 + }, + { + "epoch": 1.210826210826211, + "grad_norm": 0.692562997341156, + "learning_rate": 0.00015824729561083768, + "loss": 0.7978, + "step": 6801 + }, + { + "epoch": 1.2110042735042734, + "grad_norm": 0.7255034446716309, + "learning_rate": 0.00015823591719386066, + "loss": 1.071, + "step": 6802 + }, + { + "epoch": 1.2111823361823362, + "grad_norm": 0.6598904728889465, + "learning_rate": 0.0001582245376358718, + "loss": 0.9736, + "step": 6803 + }, + { + "epoch": 1.211360398860399, + "grad_norm": 0.6372483968734741, + "learning_rate": 0.0001582131569370941, + "loss": 0.9029, + "step": 6804 + }, + { + "epoch": 1.2115384615384615, + "grad_norm": 0.5907173156738281, + "learning_rate": 0.00015820177509775048, + "loss": 0.918, + "step": 6805 + }, + { + "epoch": 1.2117165242165242, + "grad_norm": 0.6252630949020386, + "learning_rate": 0.00015819039211806404, + "loss": 0.7801, + "step": 6806 + }, + { + "epoch": 1.211894586894587, + "grad_norm": 0.5793096423149109, + "learning_rate": 0.0001581790079982577, + "loss": 0.5769, + "step": 6807 + }, + { + "epoch": 1.2120726495726495, + "grad_norm": 0.7267270684242249, + "learning_rate": 0.00015816762273855454, + "loss": 1.1428, + "step": 6808 + }, + { + "epoch": 1.2122507122507122, + "grad_norm": 0.7481234073638916, + "learning_rate": 0.00015815623633917767, + "loss": 1.0209, + "step": 6809 + }, + { + "epoch": 1.212428774928775, + "grad_norm": 0.6114386916160583, + "learning_rate": 0.00015814484880035017, + "loss": 0.9073, + "step": 6810 + }, + { + "epoch": 1.2126068376068375, + "grad_norm": 0.6871182322502136, + "learning_rate": 0.00015813346012229516, + "loss": 1.151, + "step": 6811 + }, + { + "epoch": 1.2127849002849003, + "grad_norm": 0.6380293965339661, + "learning_rate": 0.0001581220703052357, + "loss": 1.0981, + "step": 6812 + }, + { + "epoch": 1.212962962962963, + "grad_norm": 0.6013718247413635, + "learning_rate": 0.00015811067934939503, + "loss": 0.8832, + "step": 6813 + }, + { + "epoch": 1.2131410256410255, + "grad_norm": 0.5816897749900818, + "learning_rate": 0.00015809928725499632, + "loss": 1.063, + "step": 6814 + }, + { + "epoch": 1.2133190883190883, + "grad_norm": 0.5970914363861084, + "learning_rate": 0.00015808789402226278, + "loss": 1.1177, + "step": 6815 + }, + { + "epoch": 1.213497150997151, + "grad_norm": 0.7624936103820801, + "learning_rate": 0.00015807649965141762, + "loss": 1.048, + "step": 6816 + }, + { + "epoch": 1.2136752136752136, + "grad_norm": 0.636263906955719, + "learning_rate": 0.0001580651041426841, + "loss": 0.9743, + "step": 6817 + }, + { + "epoch": 1.2138532763532763, + "grad_norm": 0.641090452671051, + "learning_rate": 0.00015805370749628547, + "loss": 1.0227, + "step": 6818 + }, + { + "epoch": 1.214031339031339, + "grad_norm": 0.6484021544456482, + "learning_rate": 0.00015804230971244504, + "loss": 0.9615, + "step": 6819 + }, + { + "epoch": 1.2142094017094016, + "grad_norm": 0.6473353505134583, + "learning_rate": 0.00015803091079138613, + "loss": 1.0507, + "step": 6820 + }, + { + "epoch": 1.2143874643874644, + "grad_norm": 0.5477129220962524, + "learning_rate": 0.00015801951073333206, + "loss": 0.7928, + "step": 6821 + }, + { + "epoch": 1.2145655270655271, + "grad_norm": 0.7256210446357727, + "learning_rate": 0.0001580081095385062, + "loss": 1.0172, + "step": 6822 + }, + { + "epoch": 1.2147435897435896, + "grad_norm": 0.5785418748855591, + "learning_rate": 0.00015799670720713195, + "loss": 0.8478, + "step": 6823 + }, + { + "epoch": 1.2149216524216524, + "grad_norm": 0.6782996654510498, + "learning_rate": 0.00015798530373943267, + "loss": 1.1819, + "step": 6824 + }, + { + "epoch": 1.2150997150997151, + "grad_norm": 0.6513699293136597, + "learning_rate": 0.00015797389913563186, + "loss": 0.9626, + "step": 6825 + }, + { + "epoch": 1.2152777777777777, + "grad_norm": 0.6503037214279175, + "learning_rate": 0.0001579624933959529, + "loss": 1.0282, + "step": 6826 + }, + { + "epoch": 1.2154558404558404, + "grad_norm": 0.581501841545105, + "learning_rate": 0.0001579510865206193, + "loss": 0.8976, + "step": 6827 + }, + { + "epoch": 1.2156339031339032, + "grad_norm": 0.6696721911430359, + "learning_rate": 0.00015793967850985454, + "loss": 0.6418, + "step": 6828 + }, + { + "epoch": 1.215811965811966, + "grad_norm": 0.6577274203300476, + "learning_rate": 0.00015792826936388213, + "loss": 1.0615, + "step": 6829 + }, + { + "epoch": 1.2159900284900285, + "grad_norm": 0.66291743516922, + "learning_rate": 0.00015791685908292564, + "loss": 0.8582, + "step": 6830 + }, + { + "epoch": 1.2161680911680912, + "grad_norm": 0.6548362374305725, + "learning_rate": 0.0001579054476672086, + "loss": 1.0343, + "step": 6831 + }, + { + "epoch": 1.2163461538461537, + "grad_norm": 0.6381218433380127, + "learning_rate": 0.00015789403511695457, + "loss": 0.8133, + "step": 6832 + }, + { + "epoch": 1.2165242165242165, + "grad_norm": 0.7217492461204529, + "learning_rate": 0.00015788262143238722, + "loss": 0.9183, + "step": 6833 + }, + { + "epoch": 1.2167022792022792, + "grad_norm": 0.610454797744751, + "learning_rate": 0.00015787120661373013, + "loss": 0.8488, + "step": 6834 + }, + { + "epoch": 1.216880341880342, + "grad_norm": 0.592771053314209, + "learning_rate": 0.00015785979066120696, + "loss": 0.8673, + "step": 6835 + }, + { + "epoch": 1.2170584045584045, + "grad_norm": 0.5787834525108337, + "learning_rate": 0.00015784837357504138, + "loss": 0.7945, + "step": 6836 + }, + { + "epoch": 1.2172364672364673, + "grad_norm": 0.6814196109771729, + "learning_rate": 0.0001578369553554571, + "loss": 0.8906, + "step": 6837 + }, + { + "epoch": 1.2174145299145298, + "grad_norm": 0.6383981108665466, + "learning_rate": 0.00015782553600267787, + "loss": 0.8962, + "step": 6838 + }, + { + "epoch": 1.2175925925925926, + "grad_norm": 0.6733864545822144, + "learning_rate": 0.0001578141155169273, + "loss": 1.2077, + "step": 6839 + }, + { + "epoch": 1.2177706552706553, + "grad_norm": 0.5891284346580505, + "learning_rate": 0.0001578026938984293, + "loss": 0.9477, + "step": 6840 + }, + { + "epoch": 1.217948717948718, + "grad_norm": 0.7220266461372375, + "learning_rate": 0.00015779127114740757, + "loss": 1.0343, + "step": 6841 + }, + { + "epoch": 1.2181267806267806, + "grad_norm": 0.6566546559333801, + "learning_rate": 0.0001577798472640859, + "loss": 0.9576, + "step": 6842 + }, + { + "epoch": 1.2183048433048433, + "grad_norm": 0.6428449153900146, + "learning_rate": 0.0001577684222486882, + "loss": 0.8957, + "step": 6843 + }, + { + "epoch": 1.2184829059829059, + "grad_norm": 0.6542909741401672, + "learning_rate": 0.00015775699610143823, + "loss": 0.9942, + "step": 6844 + }, + { + "epoch": 1.2186609686609686, + "grad_norm": 0.7101675868034363, + "learning_rate": 0.00015774556882255992, + "loss": 1.015, + "step": 6845 + }, + { + "epoch": 1.2188390313390314, + "grad_norm": 0.6606267094612122, + "learning_rate": 0.00015773414041227713, + "loss": 1.1406, + "step": 6846 + }, + { + "epoch": 1.2190170940170941, + "grad_norm": 0.67124342918396, + "learning_rate": 0.00015772271087081383, + "loss": 1.2392, + "step": 6847 + }, + { + "epoch": 1.2191951566951567, + "grad_norm": 0.6615056991577148, + "learning_rate": 0.0001577112801983939, + "loss": 1.1583, + "step": 6848 + }, + { + "epoch": 1.2193732193732194, + "grad_norm": 0.6941317319869995, + "learning_rate": 0.0001576998483952413, + "loss": 1.0255, + "step": 6849 + }, + { + "epoch": 1.219551282051282, + "grad_norm": 0.5740683674812317, + "learning_rate": 0.00015768841546158005, + "loss": 1.0393, + "step": 6850 + }, + { + "epoch": 1.2197293447293447, + "grad_norm": 0.7143667340278625, + "learning_rate": 0.00015767698139763415, + "loss": 0.7564, + "step": 6851 + }, + { + "epoch": 1.2199074074074074, + "grad_norm": 0.6730484366416931, + "learning_rate": 0.00015766554620362758, + "loss": 1.2221, + "step": 6852 + }, + { + "epoch": 1.2200854700854702, + "grad_norm": 0.6883087754249573, + "learning_rate": 0.00015765410987978444, + "loss": 1.0156, + "step": 6853 + }, + { + "epoch": 1.2202635327635327, + "grad_norm": 0.6585961580276489, + "learning_rate": 0.00015764267242632875, + "loss": 1.0888, + "step": 6854 + }, + { + "epoch": 1.2204415954415955, + "grad_norm": 0.6325246691703796, + "learning_rate": 0.00015763123384348465, + "loss": 0.973, + "step": 6855 + }, + { + "epoch": 1.220619658119658, + "grad_norm": 0.5930588245391846, + "learning_rate": 0.00015761979413147627, + "loss": 0.8551, + "step": 6856 + }, + { + "epoch": 1.2207977207977208, + "grad_norm": 0.6440611481666565, + "learning_rate": 0.0001576083532905277, + "loss": 0.8396, + "step": 6857 + }, + { + "epoch": 1.2209757834757835, + "grad_norm": 0.6796659231185913, + "learning_rate": 0.00015759691132086315, + "loss": 1.0662, + "step": 6858 + }, + { + "epoch": 1.2211538461538463, + "grad_norm": 0.6813400983810425, + "learning_rate": 0.00015758546822270674, + "loss": 1.0457, + "step": 6859 + }, + { + "epoch": 1.2213319088319088, + "grad_norm": 0.6871716976165771, + "learning_rate": 0.00015757402399628272, + "loss": 1.1675, + "step": 6860 + }, + { + "epoch": 1.2215099715099715, + "grad_norm": 0.6431481838226318, + "learning_rate": 0.00015756257864181524, + "loss": 0.9366, + "step": 6861 + }, + { + "epoch": 1.221688034188034, + "grad_norm": 0.6061800718307495, + "learning_rate": 0.00015755113215952868, + "loss": 0.9267, + "step": 6862 + }, + { + "epoch": 1.2218660968660968, + "grad_norm": 0.5755770206451416, + "learning_rate": 0.00015753968454964722, + "loss": 0.7342, + "step": 6863 + }, + { + "epoch": 1.2220441595441596, + "grad_norm": 0.571345329284668, + "learning_rate": 0.00015752823581239515, + "loss": 0.8943, + "step": 6864 + }, + { + "epoch": 1.2222222222222223, + "grad_norm": 0.6925615668296814, + "learning_rate": 0.0001575167859479968, + "loss": 0.8801, + "step": 6865 + }, + { + "epoch": 1.2224002849002849, + "grad_norm": 0.6812975406646729, + "learning_rate": 0.00015750533495667655, + "loss": 0.9567, + "step": 6866 + }, + { + "epoch": 1.2225783475783476, + "grad_norm": 0.8216777443885803, + "learning_rate": 0.00015749388283865868, + "loss": 1.0908, + "step": 6867 + }, + { + "epoch": 1.2227564102564104, + "grad_norm": 0.6051010489463806, + "learning_rate": 0.00015748242959416763, + "loss": 0.8851, + "step": 6868 + }, + { + "epoch": 1.2229344729344729, + "grad_norm": 0.7750816345214844, + "learning_rate": 0.00015747097522342775, + "loss": 1.1526, + "step": 6869 + }, + { + "epoch": 1.2231125356125356, + "grad_norm": 0.6240930557250977, + "learning_rate": 0.00015745951972666355, + "loss": 1.0603, + "step": 6870 + }, + { + "epoch": 1.2232905982905984, + "grad_norm": 0.7228875160217285, + "learning_rate": 0.00015744806310409937, + "loss": 1.1028, + "step": 6871 + }, + { + "epoch": 1.223468660968661, + "grad_norm": 0.724075436592102, + "learning_rate": 0.00015743660535595978, + "loss": 0.8983, + "step": 6872 + }, + { + "epoch": 1.2236467236467237, + "grad_norm": 0.6398203372955322, + "learning_rate": 0.00015742514648246916, + "loss": 1.0548, + "step": 6873 + }, + { + "epoch": 1.2238247863247864, + "grad_norm": 0.7024285793304443, + "learning_rate": 0.00015741368648385212, + "loss": 1.0172, + "step": 6874 + }, + { + "epoch": 1.224002849002849, + "grad_norm": 0.6717609763145447, + "learning_rate": 0.00015740222536033316, + "loss": 0.9002, + "step": 6875 + }, + { + "epoch": 1.2241809116809117, + "grad_norm": 0.5886133313179016, + "learning_rate": 0.00015739076311213686, + "loss": 0.8614, + "step": 6876 + }, + { + "epoch": 1.2243589743589745, + "grad_norm": 0.6856684684753418, + "learning_rate": 0.00015737929973948776, + "loss": 1.1633, + "step": 6877 + }, + { + "epoch": 1.224537037037037, + "grad_norm": 0.6771421432495117, + "learning_rate": 0.00015736783524261045, + "loss": 1.0921, + "step": 6878 + }, + { + "epoch": 1.2247150997150997, + "grad_norm": 0.5016412138938904, + "learning_rate": 0.0001573563696217296, + "loss": 0.6732, + "step": 6879 + }, + { + "epoch": 1.2248931623931625, + "grad_norm": 0.7595276236534119, + "learning_rate": 0.00015734490287706984, + "loss": 1.0427, + "step": 6880 + }, + { + "epoch": 1.225071225071225, + "grad_norm": 0.6664281487464905, + "learning_rate": 0.00015733343500885582, + "loss": 1.2836, + "step": 6881 + }, + { + "epoch": 1.2252492877492878, + "grad_norm": 0.6662577390670776, + "learning_rate": 0.00015732196601731224, + "loss": 1.1288, + "step": 6882 + }, + { + "epoch": 1.2254273504273505, + "grad_norm": 0.6238988041877747, + "learning_rate": 0.00015731049590266385, + "loss": 1.0809, + "step": 6883 + }, + { + "epoch": 1.225605413105413, + "grad_norm": 0.6483062505722046, + "learning_rate": 0.00015729902466513532, + "loss": 0.9992, + "step": 6884 + }, + { + "epoch": 1.2257834757834758, + "grad_norm": 0.6890861988067627, + "learning_rate": 0.0001572875523049514, + "loss": 1.1844, + "step": 6885 + }, + { + "epoch": 1.2259615384615385, + "grad_norm": 0.7087607383728027, + "learning_rate": 0.00015727607882233695, + "loss": 1.013, + "step": 6886 + }, + { + "epoch": 1.226139601139601, + "grad_norm": 0.709048867225647, + "learning_rate": 0.00015726460421751668, + "loss": 0.9748, + "step": 6887 + }, + { + "epoch": 1.2263176638176638, + "grad_norm": 0.5918150544166565, + "learning_rate": 0.00015725312849071546, + "loss": 0.9978, + "step": 6888 + }, + { + "epoch": 1.2264957264957266, + "grad_norm": 0.4343377947807312, + "learning_rate": 0.0001572416516421581, + "loss": 0.6233, + "step": 6889 + }, + { + "epoch": 1.226673789173789, + "grad_norm": 0.6360403895378113, + "learning_rate": 0.00015723017367206952, + "loss": 0.9698, + "step": 6890 + }, + { + "epoch": 1.2268518518518519, + "grad_norm": 0.7261984944343567, + "learning_rate": 0.00015721869458067454, + "loss": 1.0426, + "step": 6891 + }, + { + "epoch": 1.2270299145299146, + "grad_norm": 0.6806774139404297, + "learning_rate": 0.0001572072143681981, + "loss": 0.9692, + "step": 6892 + }, + { + "epoch": 1.2272079772079771, + "grad_norm": 0.7140612006187439, + "learning_rate": 0.00015719573303486515, + "loss": 1.0828, + "step": 6893 + }, + { + "epoch": 1.22738603988604, + "grad_norm": 0.5383326411247253, + "learning_rate": 0.0001571842505809006, + "loss": 1.012, + "step": 6894 + }, + { + "epoch": 1.2275641025641026, + "grad_norm": 0.5992259383201599, + "learning_rate": 0.0001571727670065295, + "loss": 0.876, + "step": 6895 + }, + { + "epoch": 1.2277421652421652, + "grad_norm": 0.636696457862854, + "learning_rate": 0.00015716128231197676, + "loss": 1.1001, + "step": 6896 + }, + { + "epoch": 1.227920227920228, + "grad_norm": 0.5980371236801147, + "learning_rate": 0.00015714979649746744, + "loss": 0.937, + "step": 6897 + }, + { + "epoch": 1.2280982905982907, + "grad_norm": 0.7678794860839844, + "learning_rate": 0.00015713830956322656, + "loss": 1.1965, + "step": 6898 + }, + { + "epoch": 1.2282763532763532, + "grad_norm": 0.6918835639953613, + "learning_rate": 0.00015712682150947923, + "loss": 0.8578, + "step": 6899 + }, + { + "epoch": 1.228454415954416, + "grad_norm": 0.6463451385498047, + "learning_rate": 0.00015711533233645048, + "loss": 1.009, + "step": 6900 + }, + { + "epoch": 1.2286324786324787, + "grad_norm": 0.6720646023750305, + "learning_rate": 0.00015710384204436549, + "loss": 1.0031, + "step": 6901 + }, + { + "epoch": 1.2288105413105412, + "grad_norm": 0.6618736982345581, + "learning_rate": 0.00015709235063344926, + "loss": 0.9017, + "step": 6902 + }, + { + "epoch": 1.228988603988604, + "grad_norm": 0.6789427399635315, + "learning_rate": 0.0001570808581039271, + "loss": 1.1289, + "step": 6903 + }, + { + "epoch": 1.2291666666666667, + "grad_norm": 0.6395950317382812, + "learning_rate": 0.00015706936445602403, + "loss": 1.1051, + "step": 6904 + }, + { + "epoch": 1.2293447293447293, + "grad_norm": 0.7023917436599731, + "learning_rate": 0.00015705786968996533, + "loss": 1.2876, + "step": 6905 + }, + { + "epoch": 1.229522792022792, + "grad_norm": 0.7473352551460266, + "learning_rate": 0.00015704637380597623, + "loss": 1.237, + "step": 6906 + }, + { + "epoch": 1.2297008547008548, + "grad_norm": 0.6952672004699707, + "learning_rate": 0.00015703487680428192, + "loss": 1.0674, + "step": 6907 + }, + { + "epoch": 1.2298789173789173, + "grad_norm": 0.5968644022941589, + "learning_rate": 0.0001570233786851077, + "loss": 0.9169, + "step": 6908 + }, + { + "epoch": 1.23005698005698, + "grad_norm": 0.7219798564910889, + "learning_rate": 0.0001570118794486788, + "loss": 1.0556, + "step": 6909 + }, + { + "epoch": 1.2302350427350428, + "grad_norm": 0.6603400707244873, + "learning_rate": 0.0001570003790952206, + "loss": 0.9596, + "step": 6910 + }, + { + "epoch": 1.2304131054131053, + "grad_norm": 0.5972838401794434, + "learning_rate": 0.0001569888776249583, + "loss": 0.9168, + "step": 6911 + }, + { + "epoch": 1.230591168091168, + "grad_norm": 0.792585551738739, + "learning_rate": 0.00015697737503811738, + "loss": 1.1074, + "step": 6912 + }, + { + "epoch": 1.2307692307692308, + "grad_norm": 0.5845609903335571, + "learning_rate": 0.00015696587133492314, + "loss": 0.8413, + "step": 6913 + }, + { + "epoch": 1.2309472934472934, + "grad_norm": 0.6603896021842957, + "learning_rate": 0.000156954366515601, + "loss": 0.9109, + "step": 6914 + }, + { + "epoch": 1.2311253561253561, + "grad_norm": 0.6367142796516418, + "learning_rate": 0.00015694286058037636, + "loss": 1.0119, + "step": 6915 + }, + { + "epoch": 1.2313034188034189, + "grad_norm": 0.693854570388794, + "learning_rate": 0.00015693135352947465, + "loss": 1.0925, + "step": 6916 + }, + { + "epoch": 1.2314814814814814, + "grad_norm": 0.6570404171943665, + "learning_rate": 0.00015691984536312135, + "loss": 0.9731, + "step": 6917 + }, + { + "epoch": 1.2316595441595442, + "grad_norm": 0.6778639554977417, + "learning_rate": 0.0001569083360815419, + "loss": 1.1415, + "step": 6918 + }, + { + "epoch": 1.231837606837607, + "grad_norm": 0.6656233668327332, + "learning_rate": 0.00015689682568496182, + "loss": 0.8603, + "step": 6919 + }, + { + "epoch": 1.2320156695156694, + "grad_norm": 0.6569861173629761, + "learning_rate": 0.00015688531417360665, + "loss": 0.8374, + "step": 6920 + }, + { + "epoch": 1.2321937321937322, + "grad_norm": 0.6746888160705566, + "learning_rate": 0.0001568738015477019, + "loss": 1.1395, + "step": 6921 + }, + { + "epoch": 1.232371794871795, + "grad_norm": 0.6180813908576965, + "learning_rate": 0.00015686228780747316, + "loss": 1.0049, + "step": 6922 + }, + { + "epoch": 1.2325498575498575, + "grad_norm": 0.7326146960258484, + "learning_rate": 0.000156850772953146, + "loss": 1.2389, + "step": 6923 + }, + { + "epoch": 1.2327279202279202, + "grad_norm": 0.5912215709686279, + "learning_rate": 0.00015683925698494608, + "loss": 1.0174, + "step": 6924 + }, + { + "epoch": 1.232905982905983, + "grad_norm": 0.5214745402336121, + "learning_rate": 0.00015682773990309895, + "loss": 0.5778, + "step": 6925 + }, + { + "epoch": 1.2330840455840455, + "grad_norm": 0.6862079501152039, + "learning_rate": 0.00015681622170783034, + "loss": 0.896, + "step": 6926 + }, + { + "epoch": 1.2332621082621082, + "grad_norm": 0.7858926057815552, + "learning_rate": 0.00015680470239936586, + "loss": 1.0714, + "step": 6927 + }, + { + "epoch": 1.233440170940171, + "grad_norm": 0.6706146597862244, + "learning_rate": 0.00015679318197793127, + "loss": 1.0157, + "step": 6928 + }, + { + "epoch": 1.2336182336182335, + "grad_norm": 0.6657105088233948, + "learning_rate": 0.00015678166044375225, + "loss": 0.9674, + "step": 6929 + }, + { + "epoch": 1.2337962962962963, + "grad_norm": 0.6790838837623596, + "learning_rate": 0.0001567701377970545, + "loss": 0.9744, + "step": 6930 + }, + { + "epoch": 1.233974358974359, + "grad_norm": 0.6469771862030029, + "learning_rate": 0.00015675861403806386, + "loss": 1.0205, + "step": 6931 + }, + { + "epoch": 1.2341524216524216, + "grad_norm": 0.4926300346851349, + "learning_rate": 0.0001567470891670061, + "loss": 0.6336, + "step": 6932 + }, + { + "epoch": 1.2343304843304843, + "grad_norm": 0.6762157082557678, + "learning_rate": 0.000156735563184107, + "loss": 1.059, + "step": 6933 + }, + { + "epoch": 1.234508547008547, + "grad_norm": 0.6998521685600281, + "learning_rate": 0.0001567240360895924, + "loss": 1.0586, + "step": 6934 + }, + { + "epoch": 1.2346866096866096, + "grad_norm": 0.5947706699371338, + "learning_rate": 0.00015671250788368814, + "loss": 0.8815, + "step": 6935 + }, + { + "epoch": 1.2348646723646723, + "grad_norm": 0.6966122984886169, + "learning_rate": 0.0001567009785666201, + "loss": 1.0105, + "step": 6936 + }, + { + "epoch": 1.235042735042735, + "grad_norm": 0.6747866272926331, + "learning_rate": 0.0001566894481386142, + "loss": 0.8783, + "step": 6937 + }, + { + "epoch": 1.2352207977207976, + "grad_norm": 0.6348921060562134, + "learning_rate": 0.0001566779165998963, + "loss": 0.7813, + "step": 6938 + }, + { + "epoch": 1.2353988603988604, + "grad_norm": 0.596466600894928, + "learning_rate": 0.00015666638395069236, + "loss": 0.8689, + "step": 6939 + }, + { + "epoch": 1.2355769230769231, + "grad_norm": 0.6926795244216919, + "learning_rate": 0.00015665485019122834, + "loss": 1.0266, + "step": 6940 + }, + { + "epoch": 1.2357549857549857, + "grad_norm": 0.6590100526809692, + "learning_rate": 0.00015664331532173022, + "loss": 1.128, + "step": 6941 + }, + { + "epoch": 1.2359330484330484, + "grad_norm": 0.7422109246253967, + "learning_rate": 0.00015663177934242402, + "loss": 0.8495, + "step": 6942 + }, + { + "epoch": 1.2361111111111112, + "grad_norm": 0.6463228464126587, + "learning_rate": 0.0001566202422535357, + "loss": 1.0941, + "step": 6943 + }, + { + "epoch": 1.236289173789174, + "grad_norm": 0.7278686761856079, + "learning_rate": 0.0001566087040552914, + "loss": 1.2039, + "step": 6944 + }, + { + "epoch": 1.2364672364672364, + "grad_norm": 0.6917086839675903, + "learning_rate": 0.00015659716474791712, + "loss": 1.042, + "step": 6945 + }, + { + "epoch": 1.2366452991452992, + "grad_norm": 0.637205183506012, + "learning_rate": 0.00015658562433163898, + "loss": 1.0379, + "step": 6946 + }, + { + "epoch": 1.2368233618233617, + "grad_norm": 0.6706623435020447, + "learning_rate": 0.00015657408280668307, + "loss": 1.0347, + "step": 6947 + }, + { + "epoch": 1.2370014245014245, + "grad_norm": 0.6435480713844299, + "learning_rate": 0.00015656254017327553, + "loss": 0.7708, + "step": 6948 + }, + { + "epoch": 1.2371794871794872, + "grad_norm": 0.5703113675117493, + "learning_rate": 0.0001565509964316425, + "loss": 0.8786, + "step": 6949 + }, + { + "epoch": 1.23735754985755, + "grad_norm": 0.6438127160072327, + "learning_rate": 0.00015653945158201018, + "loss": 0.9435, + "step": 6950 + }, + { + "epoch": 1.2375356125356125, + "grad_norm": 0.68101966381073, + "learning_rate": 0.00015652790562460474, + "loss": 1.1062, + "step": 6951 + }, + { + "epoch": 1.2377136752136753, + "grad_norm": 0.661230206489563, + "learning_rate": 0.00015651635855965242, + "loss": 1.0113, + "step": 6952 + }, + { + "epoch": 1.2378917378917378, + "grad_norm": 0.6399117708206177, + "learning_rate": 0.0001565048103873795, + "loss": 1.1423, + "step": 6953 + }, + { + "epoch": 1.2380698005698005, + "grad_norm": 0.7614672780036926, + "learning_rate": 0.00015649326110801215, + "loss": 1.0359, + "step": 6954 + }, + { + "epoch": 1.2382478632478633, + "grad_norm": 0.6461986303329468, + "learning_rate": 0.00015648171072177674, + "loss": 1.0145, + "step": 6955 + }, + { + "epoch": 1.238425925925926, + "grad_norm": 0.5902668833732605, + "learning_rate": 0.0001564701592288995, + "loss": 0.9451, + "step": 6956 + }, + { + "epoch": 1.2386039886039886, + "grad_norm": 0.5686020255088806, + "learning_rate": 0.00015645860662960682, + "loss": 0.7512, + "step": 6957 + }, + { + "epoch": 1.2387820512820513, + "grad_norm": 0.6640077829360962, + "learning_rate": 0.00015644705292412503, + "loss": 0.7133, + "step": 6958 + }, + { + "epoch": 1.2389601139601139, + "grad_norm": 0.7402132749557495, + "learning_rate": 0.00015643549811268049, + "loss": 1.0903, + "step": 6959 + }, + { + "epoch": 1.2391381766381766, + "grad_norm": 0.62332683801651, + "learning_rate": 0.00015642394219549962, + "loss": 0.9378, + "step": 6960 + }, + { + "epoch": 1.2393162393162394, + "grad_norm": 0.6374901533126831, + "learning_rate": 0.00015641238517280877, + "loss": 1.0746, + "step": 6961 + }, + { + "epoch": 1.239494301994302, + "grad_norm": 0.5939112901687622, + "learning_rate": 0.00015640082704483443, + "loss": 0.7185, + "step": 6962 + }, + { + "epoch": 1.2396723646723646, + "grad_norm": 0.8378096222877502, + "learning_rate": 0.00015638926781180306, + "loss": 1.1932, + "step": 6963 + }, + { + "epoch": 1.2398504273504274, + "grad_norm": 0.5707982778549194, + "learning_rate": 0.0001563777074739411, + "loss": 0.9834, + "step": 6964 + }, + { + "epoch": 1.24002849002849, + "grad_norm": 0.6339748501777649, + "learning_rate": 0.00015636614603147512, + "loss": 1.0307, + "step": 6965 + }, + { + "epoch": 1.2402065527065527, + "grad_norm": 0.7353155016899109, + "learning_rate": 0.00015635458348463156, + "loss": 1.0311, + "step": 6966 + }, + { + "epoch": 1.2403846153846154, + "grad_norm": 0.8307726979255676, + "learning_rate": 0.00015634301983363704, + "loss": 1.0673, + "step": 6967 + }, + { + "epoch": 1.2405626780626782, + "grad_norm": 0.5299199819564819, + "learning_rate": 0.00015633145507871807, + "loss": 0.6649, + "step": 6968 + }, + { + "epoch": 1.2407407407407407, + "grad_norm": 0.6162533760070801, + "learning_rate": 0.00015631988922010126, + "loss": 0.8096, + "step": 6969 + }, + { + "epoch": 1.2409188034188035, + "grad_norm": 0.6212689876556396, + "learning_rate": 0.0001563083222580132, + "loss": 1.0371, + "step": 6970 + }, + { + "epoch": 1.241096866096866, + "grad_norm": 0.6148123145103455, + "learning_rate": 0.00015629675419268055, + "loss": 1.0439, + "step": 6971 + }, + { + "epoch": 1.2412749287749287, + "grad_norm": 0.6163684129714966, + "learning_rate": 0.00015628518502432994, + "loss": 0.9075, + "step": 6972 + }, + { + "epoch": 1.2414529914529915, + "grad_norm": 0.5127472877502441, + "learning_rate": 0.00015627361475318807, + "loss": 0.6138, + "step": 6973 + }, + { + "epoch": 1.2416310541310542, + "grad_norm": 0.6508103013038635, + "learning_rate": 0.0001562620433794816, + "loss": 0.9608, + "step": 6974 + }, + { + "epoch": 1.2418091168091168, + "grad_norm": 0.6711046695709229, + "learning_rate": 0.0001562504709034373, + "loss": 1.1494, + "step": 6975 + }, + { + "epoch": 1.2419871794871795, + "grad_norm": 0.6831514835357666, + "learning_rate": 0.00015623889732528182, + "loss": 0.9664, + "step": 6976 + }, + { + "epoch": 1.242165242165242, + "grad_norm": 0.693732738494873, + "learning_rate": 0.00015622732264524198, + "loss": 0.9055, + "step": 6977 + }, + { + "epoch": 1.2423433048433048, + "grad_norm": 0.8475173711776733, + "learning_rate": 0.00015621574686354456, + "loss": 1.2014, + "step": 6978 + }, + { + "epoch": 1.2425213675213675, + "grad_norm": 0.6342347264289856, + "learning_rate": 0.0001562041699804164, + "loss": 1.0691, + "step": 6979 + }, + { + "epoch": 1.2426994301994303, + "grad_norm": 0.620517373085022, + "learning_rate": 0.00015619259199608422, + "loss": 0.7318, + "step": 6980 + }, + { + "epoch": 1.2428774928774928, + "grad_norm": 0.589567244052887, + "learning_rate": 0.000156181012910775, + "loss": 1.0656, + "step": 6981 + }, + { + "epoch": 1.2430555555555556, + "grad_norm": 0.7570258975028992, + "learning_rate": 0.00015616943272471546, + "loss": 1.0517, + "step": 6982 + }, + { + "epoch": 1.243233618233618, + "grad_norm": 0.6232032775878906, + "learning_rate": 0.00015615785143813262, + "loss": 0.8867, + "step": 6983 + }, + { + "epoch": 1.2434116809116809, + "grad_norm": 0.630095899105072, + "learning_rate": 0.0001561462690512533, + "loss": 0.9287, + "step": 6984 + }, + { + "epoch": 1.2435897435897436, + "grad_norm": 0.7410848140716553, + "learning_rate": 0.00015613468556430454, + "loss": 1.162, + "step": 6985 + }, + { + "epoch": 1.2437678062678064, + "grad_norm": 0.7574684023857117, + "learning_rate": 0.00015612310097751317, + "loss": 1.2118, + "step": 6986 + }, + { + "epoch": 1.243945868945869, + "grad_norm": 0.580760657787323, + "learning_rate": 0.0001561115152911062, + "loss": 1.0612, + "step": 6987 + }, + { + "epoch": 1.2441239316239316, + "grad_norm": 0.6105104088783264, + "learning_rate": 0.00015609992850531073, + "loss": 0.9262, + "step": 6988 + }, + { + "epoch": 1.2443019943019944, + "grad_norm": 0.669435441493988, + "learning_rate": 0.00015608834062035362, + "loss": 0.9595, + "step": 6989 + }, + { + "epoch": 1.244480056980057, + "grad_norm": 0.6530314683914185, + "learning_rate": 0.00015607675163646206, + "loss": 0.7987, + "step": 6990 + }, + { + "epoch": 1.2446581196581197, + "grad_norm": 0.5801477432250977, + "learning_rate": 0.00015606516155386297, + "loss": 0.7667, + "step": 6991 + }, + { + "epoch": 1.2448361823361824, + "grad_norm": 0.5773885250091553, + "learning_rate": 0.00015605357037278355, + "loss": 0.847, + "step": 6992 + }, + { + "epoch": 1.245014245014245, + "grad_norm": 0.5399810075759888, + "learning_rate": 0.00015604197809345082, + "loss": 0.9284, + "step": 6993 + }, + { + "epoch": 1.2451923076923077, + "grad_norm": 0.5910452604293823, + "learning_rate": 0.000156030384716092, + "loss": 1.0004, + "step": 6994 + }, + { + "epoch": 1.2453703703703705, + "grad_norm": 0.5979224443435669, + "learning_rate": 0.00015601879024093414, + "loss": 0.9027, + "step": 6995 + }, + { + "epoch": 1.245548433048433, + "grad_norm": 0.6092126369476318, + "learning_rate": 0.0001560071946682045, + "loss": 0.9755, + "step": 6996 + }, + { + "epoch": 1.2457264957264957, + "grad_norm": 0.6536708474159241, + "learning_rate": 0.0001559955979981302, + "loss": 1.1828, + "step": 6997 + }, + { + "epoch": 1.2459045584045585, + "grad_norm": 0.6602030992507935, + "learning_rate": 0.00015598400023093847, + "loss": 1.0395, + "step": 6998 + }, + { + "epoch": 1.246082621082621, + "grad_norm": 0.6864825487136841, + "learning_rate": 0.00015597240136685657, + "loss": 1.083, + "step": 6999 + }, + { + "epoch": 1.2462606837606838, + "grad_norm": 0.6194674968719482, + "learning_rate": 0.0001559608014061117, + "loss": 1.0461, + "step": 7000 + }, + { + "epoch": 1.2464387464387465, + "grad_norm": 0.5879074335098267, + "learning_rate": 0.00015594920034893122, + "loss": 1.076, + "step": 7001 + }, + { + "epoch": 1.246616809116809, + "grad_norm": 0.6514387726783752, + "learning_rate": 0.00015593759819554234, + "loss": 1.0396, + "step": 7002 + }, + { + "epoch": 1.2467948717948718, + "grad_norm": 0.5988301634788513, + "learning_rate": 0.00015592599494617247, + "loss": 0.9501, + "step": 7003 + }, + { + "epoch": 1.2469729344729346, + "grad_norm": 0.6282773613929749, + "learning_rate": 0.00015591439060104887, + "loss": 1.1002, + "step": 7004 + }, + { + "epoch": 1.247150997150997, + "grad_norm": 0.6910465955734253, + "learning_rate": 0.00015590278516039896, + "loss": 1.1771, + "step": 7005 + }, + { + "epoch": 1.2473290598290598, + "grad_norm": 0.6097282767295837, + "learning_rate": 0.00015589117862445007, + "loss": 1.0707, + "step": 7006 + }, + { + "epoch": 1.2475071225071226, + "grad_norm": 0.7076875567436218, + "learning_rate": 0.00015587957099342967, + "loss": 1.0078, + "step": 7007 + }, + { + "epoch": 1.2476851851851851, + "grad_norm": 0.6776556372642517, + "learning_rate": 0.00015586796226756518, + "loss": 0.8971, + "step": 7008 + }, + { + "epoch": 1.2478632478632479, + "grad_norm": 0.6506341695785522, + "learning_rate": 0.00015585635244708398, + "loss": 0.9727, + "step": 7009 + }, + { + "epoch": 1.2480413105413106, + "grad_norm": 0.624724805355072, + "learning_rate": 0.00015584474153221357, + "loss": 0.9858, + "step": 7010 + }, + { + "epoch": 1.2482193732193732, + "grad_norm": 0.6070096492767334, + "learning_rate": 0.0001558331295231815, + "loss": 0.9385, + "step": 7011 + }, + { + "epoch": 1.248397435897436, + "grad_norm": 0.6948656439781189, + "learning_rate": 0.00015582151642021524, + "loss": 0.9425, + "step": 7012 + }, + { + "epoch": 1.2485754985754987, + "grad_norm": 0.6559088230133057, + "learning_rate": 0.0001558099022235423, + "loss": 1.0002, + "step": 7013 + }, + { + "epoch": 1.2487535612535612, + "grad_norm": 0.6097117066383362, + "learning_rate": 0.00015579828693339026, + "loss": 1.0234, + "step": 7014 + }, + { + "epoch": 1.248931623931624, + "grad_norm": 0.6612260341644287, + "learning_rate": 0.00015578667054998673, + "loss": 1.1376, + "step": 7015 + }, + { + "epoch": 1.2491096866096867, + "grad_norm": 0.6305607557296753, + "learning_rate": 0.00015577505307355925, + "loss": 0.9127, + "step": 7016 + }, + { + "epoch": 1.2492877492877492, + "grad_norm": 0.6648319959640503, + "learning_rate": 0.00015576343450433549, + "loss": 0.8697, + "step": 7017 + }, + { + "epoch": 1.249465811965812, + "grad_norm": 0.7642946839332581, + "learning_rate": 0.00015575181484254303, + "loss": 1.0998, + "step": 7018 + }, + { + "epoch": 1.2496438746438747, + "grad_norm": 0.6775243282318115, + "learning_rate": 0.00015574019408840962, + "loss": 1.0186, + "step": 7019 + }, + { + "epoch": 1.2498219373219372, + "grad_norm": 0.6075591444969177, + "learning_rate": 0.00015572857224216286, + "loss": 0.9592, + "step": 7020 + }, + { + "epoch": 1.2498219373219372, + "eval_loss": 1.105136752128601, + "eval_runtime": 24.4793, + "eval_samples_per_second": 42.526, + "eval_steps_per_second": 21.283, + "step": 7020 + }, + { + "epoch": 1.25, + "grad_norm": 0.5856962203979492, + "learning_rate": 0.0001557169493040305, + "loss": 0.8336, + "step": 7021 + }, + { + "epoch": 1.2501780626780628, + "grad_norm": 0.6451364159584045, + "learning_rate": 0.00015570532527424028, + "loss": 0.8805, + "step": 7022 + }, + { + "epoch": 1.2503561253561253, + "grad_norm": 0.6266474723815918, + "learning_rate": 0.00015569370015301991, + "loss": 1.0023, + "step": 7023 + }, + { + "epoch": 1.250534188034188, + "grad_norm": 0.5547378063201904, + "learning_rate": 0.00015568207394059722, + "loss": 0.7385, + "step": 7024 + }, + { + "epoch": 1.2507122507122508, + "grad_norm": 0.604169487953186, + "learning_rate": 0.0001556704466371999, + "loss": 0.9194, + "step": 7025 + }, + { + "epoch": 1.2508903133903133, + "grad_norm": 0.7054405212402344, + "learning_rate": 0.00015565881824305586, + "loss": 1.1864, + "step": 7026 + }, + { + "epoch": 1.251068376068376, + "grad_norm": 0.6429929733276367, + "learning_rate": 0.0001556471887583929, + "loss": 1.0129, + "step": 7027 + }, + { + "epoch": 1.2512464387464388, + "grad_norm": 0.695957362651825, + "learning_rate": 0.00015563555818343887, + "loss": 1.2994, + "step": 7028 + }, + { + "epoch": 1.2514245014245013, + "grad_norm": 0.5889938473701477, + "learning_rate": 0.0001556239265184216, + "loss": 1.0109, + "step": 7029 + }, + { + "epoch": 1.251602564102564, + "grad_norm": 0.6424569487571716, + "learning_rate": 0.0001556122937635691, + "loss": 0.8585, + "step": 7030 + }, + { + "epoch": 1.2517806267806268, + "grad_norm": 0.5561244487762451, + "learning_rate": 0.0001556006599191092, + "loss": 0.9994, + "step": 7031 + }, + { + "epoch": 1.2519586894586894, + "grad_norm": 0.6355302333831787, + "learning_rate": 0.00015558902498526988, + "loss": 0.9495, + "step": 7032 + }, + { + "epoch": 1.2521367521367521, + "grad_norm": 0.6272686719894409, + "learning_rate": 0.00015557738896227908, + "loss": 0.7611, + "step": 7033 + }, + { + "epoch": 1.2523148148148149, + "grad_norm": 0.7069199085235596, + "learning_rate": 0.00015556575185036482, + "loss": 1.0612, + "step": 7034 + }, + { + "epoch": 1.2524928774928774, + "grad_norm": 0.6635094285011292, + "learning_rate": 0.00015555411364975505, + "loss": 1.1182, + "step": 7035 + }, + { + "epoch": 1.2526709401709402, + "grad_norm": 0.6112014651298523, + "learning_rate": 0.00015554247436067785, + "loss": 0.8677, + "step": 7036 + }, + { + "epoch": 1.252849002849003, + "grad_norm": 0.678963303565979, + "learning_rate": 0.00015553083398336126, + "loss": 1.1421, + "step": 7037 + }, + { + "epoch": 1.2530270655270654, + "grad_norm": 0.6291939616203308, + "learning_rate": 0.0001555191925180333, + "loss": 0.9157, + "step": 7038 + }, + { + "epoch": 1.2532051282051282, + "grad_norm": 0.6519795656204224, + "learning_rate": 0.0001555075499649221, + "loss": 1.0074, + "step": 7039 + }, + { + "epoch": 1.253383190883191, + "grad_norm": 0.6063529849052429, + "learning_rate": 0.00015549590632425576, + "loss": 1.0205, + "step": 7040 + }, + { + "epoch": 1.2535612535612537, + "grad_norm": 0.7055633664131165, + "learning_rate": 0.00015548426159626242, + "loss": 1.0254, + "step": 7041 + }, + { + "epoch": 1.2537393162393162, + "grad_norm": 0.6783022880554199, + "learning_rate": 0.00015547261578117025, + "loss": 1.1017, + "step": 7042 + }, + { + "epoch": 1.253917378917379, + "grad_norm": 0.7055003643035889, + "learning_rate": 0.0001554609688792074, + "loss": 1.0269, + "step": 7043 + }, + { + "epoch": 1.2540954415954415, + "grad_norm": 0.6465007662773132, + "learning_rate": 0.0001554493208906021, + "loss": 1.0492, + "step": 7044 + }, + { + "epoch": 1.2542735042735043, + "grad_norm": 0.6443775296211243, + "learning_rate": 0.0001554376718155825, + "loss": 0.9778, + "step": 7045 + }, + { + "epoch": 1.254451566951567, + "grad_norm": 0.695214569568634, + "learning_rate": 0.0001554260216543769, + "loss": 0.8792, + "step": 7046 + }, + { + "epoch": 1.2546296296296298, + "grad_norm": 0.6777814626693726, + "learning_rate": 0.00015541437040721354, + "loss": 0.8944, + "step": 7047 + }, + { + "epoch": 1.2548076923076923, + "grad_norm": 0.6269369721412659, + "learning_rate": 0.0001554027180743207, + "loss": 0.8825, + "step": 7048 + }, + { + "epoch": 1.254985754985755, + "grad_norm": 0.6197061538696289, + "learning_rate": 0.0001553910646559267, + "loss": 0.9823, + "step": 7049 + }, + { + "epoch": 1.2551638176638176, + "grad_norm": 0.681347131729126, + "learning_rate": 0.00015537941015225984, + "loss": 0.995, + "step": 7050 + }, + { + "epoch": 1.2553418803418803, + "grad_norm": 0.6224286556243896, + "learning_rate": 0.00015536775456354848, + "loss": 0.7714, + "step": 7051 + }, + { + "epoch": 1.255519943019943, + "grad_norm": 0.6113278269767761, + "learning_rate": 0.00015535609789002098, + "loss": 0.9859, + "step": 7052 + }, + { + "epoch": 1.2556980056980058, + "grad_norm": 0.6985422372817993, + "learning_rate": 0.00015534444013190577, + "loss": 0.8785, + "step": 7053 + }, + { + "epoch": 1.2558760683760684, + "grad_norm": 0.5602933168411255, + "learning_rate": 0.00015533278128943118, + "loss": 0.8341, + "step": 7054 + }, + { + "epoch": 1.256054131054131, + "grad_norm": 0.587684690952301, + "learning_rate": 0.0001553211213628257, + "loss": 0.7933, + "step": 7055 + }, + { + "epoch": 1.2562321937321936, + "grad_norm": 0.692997932434082, + "learning_rate": 0.0001553094603523178, + "loss": 1.0957, + "step": 7056 + }, + { + "epoch": 1.2564102564102564, + "grad_norm": 0.6925587058067322, + "learning_rate": 0.00015529779825813588, + "loss": 0.8602, + "step": 7057 + }, + { + "epoch": 1.2565883190883191, + "grad_norm": 0.6383063197135925, + "learning_rate": 0.0001552861350805085, + "loss": 0.9933, + "step": 7058 + }, + { + "epoch": 1.256766381766382, + "grad_norm": 0.6520544290542603, + "learning_rate": 0.00015527447081966413, + "loss": 0.9498, + "step": 7059 + }, + { + "epoch": 1.2569444444444444, + "grad_norm": 0.7353914380073547, + "learning_rate": 0.00015526280547583133, + "loss": 1.1071, + "step": 7060 + }, + { + "epoch": 1.2571225071225072, + "grad_norm": 0.7141618132591248, + "learning_rate": 0.00015525113904923864, + "loss": 0.8333, + "step": 7061 + }, + { + "epoch": 1.2573005698005697, + "grad_norm": 0.6194499731063843, + "learning_rate": 0.00015523947154011468, + "loss": 0.9421, + "step": 7062 + }, + { + "epoch": 1.2574786324786325, + "grad_norm": 0.7514514327049255, + "learning_rate": 0.00015522780294868803, + "loss": 1.226, + "step": 7063 + }, + { + "epoch": 1.2576566951566952, + "grad_norm": 0.762923538684845, + "learning_rate": 0.0001552161332751873, + "loss": 1.1893, + "step": 7064 + }, + { + "epoch": 1.257834757834758, + "grad_norm": 0.6265730261802673, + "learning_rate": 0.00015520446251984113, + "loss": 0.6604, + "step": 7065 + }, + { + "epoch": 1.2580128205128205, + "grad_norm": 0.6447750329971313, + "learning_rate": 0.0001551927906828782, + "loss": 0.9814, + "step": 7066 + }, + { + "epoch": 1.2581908831908832, + "grad_norm": 0.5791042447090149, + "learning_rate": 0.00015518111776452722, + "loss": 0.8283, + "step": 7067 + }, + { + "epoch": 1.2583689458689458, + "grad_norm": 0.5267777442932129, + "learning_rate": 0.00015516944376501682, + "loss": 0.5748, + "step": 7068 + }, + { + "epoch": 1.2585470085470085, + "grad_norm": 0.7343912720680237, + "learning_rate": 0.0001551577686845758, + "loss": 1.1777, + "step": 7069 + }, + { + "epoch": 1.2587250712250713, + "grad_norm": 0.645746111869812, + "learning_rate": 0.00015514609252343284, + "loss": 0.9356, + "step": 7070 + }, + { + "epoch": 1.258903133903134, + "grad_norm": 0.6993104219436646, + "learning_rate": 0.0001551344152818168, + "loss": 1.06, + "step": 7071 + }, + { + "epoch": 1.2590811965811965, + "grad_norm": 0.6661365628242493, + "learning_rate": 0.0001551227369599564, + "loss": 1.061, + "step": 7072 + }, + { + "epoch": 1.2592592592592593, + "grad_norm": 0.7833736538887024, + "learning_rate": 0.0001551110575580805, + "loss": 0.9674, + "step": 7073 + }, + { + "epoch": 1.2594373219373218, + "grad_norm": 0.5878575444221497, + "learning_rate": 0.00015509937707641787, + "loss": 0.9002, + "step": 7074 + }, + { + "epoch": 1.2596153846153846, + "grad_norm": 0.6402907371520996, + "learning_rate": 0.00015508769551519745, + "loss": 1.0157, + "step": 7075 + }, + { + "epoch": 1.2597934472934473, + "grad_norm": 0.6794611215591431, + "learning_rate": 0.00015507601287464805, + "loss": 1.052, + "step": 7076 + }, + { + "epoch": 1.25997150997151, + "grad_norm": 0.706922173500061, + "learning_rate": 0.0001550643291549986, + "loss": 1.0814, + "step": 7077 + }, + { + "epoch": 1.2601495726495726, + "grad_norm": 0.6722953915596008, + "learning_rate": 0.000155052644356478, + "loss": 1.1402, + "step": 7078 + }, + { + "epoch": 1.2603276353276354, + "grad_norm": 0.6619611978530884, + "learning_rate": 0.00015504095847931518, + "loss": 0.9583, + "step": 7079 + }, + { + "epoch": 1.260505698005698, + "grad_norm": 0.5645583271980286, + "learning_rate": 0.00015502927152373914, + "loss": 0.6746, + "step": 7080 + }, + { + "epoch": 1.2606837606837606, + "grad_norm": 0.6634977459907532, + "learning_rate": 0.00015501758348997882, + "loss": 1.0451, + "step": 7081 + }, + { + "epoch": 1.2608618233618234, + "grad_norm": 0.7167651057243347, + "learning_rate": 0.00015500589437826326, + "loss": 0.931, + "step": 7082 + }, + { + "epoch": 1.2610398860398861, + "grad_norm": 0.6179340481758118, + "learning_rate": 0.00015499420418882146, + "loss": 1.0953, + "step": 7083 + }, + { + "epoch": 1.2612179487179487, + "grad_norm": 0.6948468685150146, + "learning_rate": 0.00015498251292188247, + "loss": 1.0277, + "step": 7084 + }, + { + "epoch": 1.2613960113960114, + "grad_norm": 0.6256045699119568, + "learning_rate": 0.00015497082057767532, + "loss": 1.0154, + "step": 7085 + }, + { + "epoch": 1.261574074074074, + "grad_norm": 0.6457428336143494, + "learning_rate": 0.0001549591271564292, + "loss": 0.9693, + "step": 7086 + }, + { + "epoch": 1.2617521367521367, + "grad_norm": 0.722259521484375, + "learning_rate": 0.0001549474326583731, + "loss": 0.9176, + "step": 7087 + }, + { + "epoch": 1.2619301994301995, + "grad_norm": 0.742477297782898, + "learning_rate": 0.0001549357370837362, + "loss": 0.9813, + "step": 7088 + }, + { + "epoch": 1.2621082621082622, + "grad_norm": 0.5981723666191101, + "learning_rate": 0.0001549240404327477, + "loss": 0.8943, + "step": 7089 + }, + { + "epoch": 1.2622863247863247, + "grad_norm": 0.6266574859619141, + "learning_rate": 0.00015491234270563665, + "loss": 0.8439, + "step": 7090 + }, + { + "epoch": 1.2624643874643875, + "grad_norm": 0.6723998188972473, + "learning_rate": 0.00015490064390263238, + "loss": 1.2278, + "step": 7091 + }, + { + "epoch": 1.26264245014245, + "grad_norm": 0.6628100275993347, + "learning_rate": 0.00015488894402396398, + "loss": 0.9526, + "step": 7092 + }, + { + "epoch": 1.2628205128205128, + "grad_norm": 0.6661350727081299, + "learning_rate": 0.0001548772430698608, + "loss": 0.974, + "step": 7093 + }, + { + "epoch": 1.2629985754985755, + "grad_norm": 0.8210669755935669, + "learning_rate": 0.000154865541040552, + "loss": 1.1142, + "step": 7094 + }, + { + "epoch": 1.2631766381766383, + "grad_norm": 0.6329003572463989, + "learning_rate": 0.0001548538379362669, + "loss": 0.8485, + "step": 7095 + }, + { + "epoch": 1.2633547008547008, + "grad_norm": 0.6288384795188904, + "learning_rate": 0.0001548421337572348, + "loss": 0.816, + "step": 7096 + }, + { + "epoch": 1.2635327635327636, + "grad_norm": 0.631060004234314, + "learning_rate": 0.00015483042850368504, + "loss": 0.8237, + "step": 7097 + }, + { + "epoch": 1.263710826210826, + "grad_norm": 0.7343839406967163, + "learning_rate": 0.0001548187221758469, + "loss": 1.1507, + "step": 7098 + }, + { + "epoch": 1.2638888888888888, + "grad_norm": 0.6313042640686035, + "learning_rate": 0.0001548070147739498, + "loss": 0.7762, + "step": 7099 + }, + { + "epoch": 1.2640669515669516, + "grad_norm": 0.6449850797653198, + "learning_rate": 0.00015479530629822308, + "loss": 0.9225, + "step": 7100 + }, + { + "epoch": 1.2642450142450143, + "grad_norm": 0.6371589303016663, + "learning_rate": 0.00015478359674889617, + "loss": 1.0088, + "step": 7101 + }, + { + "epoch": 1.2644230769230769, + "grad_norm": 0.6483678221702576, + "learning_rate": 0.00015477188612619849, + "loss": 0.6234, + "step": 7102 + }, + { + "epoch": 1.2646011396011396, + "grad_norm": 0.6945441365242004, + "learning_rate": 0.00015476017443035947, + "loss": 1.123, + "step": 7103 + }, + { + "epoch": 1.2647792022792022, + "grad_norm": 0.6356340050697327, + "learning_rate": 0.00015474846166160856, + "loss": 0.9923, + "step": 7104 + }, + { + "epoch": 1.264957264957265, + "grad_norm": 0.6774702668190002, + "learning_rate": 0.00015473674782017532, + "loss": 0.9694, + "step": 7105 + }, + { + "epoch": 1.2651353276353277, + "grad_norm": 0.6332793831825256, + "learning_rate": 0.0001547250329062892, + "loss": 1.0633, + "step": 7106 + }, + { + "epoch": 1.2653133903133904, + "grad_norm": 0.6563684344291687, + "learning_rate": 0.00015471331692017972, + "loss": 1.0893, + "step": 7107 + }, + { + "epoch": 1.265491452991453, + "grad_norm": 0.7318371534347534, + "learning_rate": 0.0001547015998620765, + "loss": 1.1777, + "step": 7108 + }, + { + "epoch": 1.2656695156695157, + "grad_norm": 0.7099173069000244, + "learning_rate": 0.000154689881732209, + "loss": 1.1717, + "step": 7109 + }, + { + "epoch": 1.2658475783475782, + "grad_norm": 0.661078691482544, + "learning_rate": 0.00015467816253080693, + "loss": 1.0448, + "step": 7110 + }, + { + "epoch": 1.266025641025641, + "grad_norm": 0.6206802129745483, + "learning_rate": 0.0001546664422580998, + "loss": 0.9334, + "step": 7111 + }, + { + "epoch": 1.2662037037037037, + "grad_norm": 0.6514355540275574, + "learning_rate": 0.00015465472091431728, + "loss": 0.9533, + "step": 7112 + }, + { + "epoch": 1.2663817663817665, + "grad_norm": 0.6090209484100342, + "learning_rate": 0.0001546429984996891, + "loss": 0.9206, + "step": 7113 + }, + { + "epoch": 1.266559829059829, + "grad_norm": 0.6345987915992737, + "learning_rate": 0.00015463127501444488, + "loss": 1.0537, + "step": 7114 + }, + { + "epoch": 1.2667378917378918, + "grad_norm": 0.6095160245895386, + "learning_rate": 0.0001546195504588143, + "loss": 0.8652, + "step": 7115 + }, + { + "epoch": 1.2669159544159543, + "grad_norm": 0.6751621961593628, + "learning_rate": 0.00015460782483302707, + "loss": 0.9001, + "step": 7116 + }, + { + "epoch": 1.267094017094017, + "grad_norm": 0.6261575222015381, + "learning_rate": 0.00015459609813731295, + "loss": 0.929, + "step": 7117 + }, + { + "epoch": 1.2672720797720798, + "grad_norm": 0.589495837688446, + "learning_rate": 0.0001545843703719017, + "loss": 0.9023, + "step": 7118 + }, + { + "epoch": 1.2674501424501425, + "grad_norm": 0.6364617943763733, + "learning_rate": 0.00015457264153702311, + "loss": 0.8261, + "step": 7119 + }, + { + "epoch": 1.267628205128205, + "grad_norm": 0.6685599684715271, + "learning_rate": 0.00015456091163290698, + "loss": 1.1267, + "step": 7120 + }, + { + "epoch": 1.2678062678062678, + "grad_norm": 0.6440932750701904, + "learning_rate": 0.0001545491806597831, + "loss": 0.9643, + "step": 7121 + }, + { + "epoch": 1.2679843304843303, + "grad_norm": 0.7641597390174866, + "learning_rate": 0.00015453744861788137, + "loss": 1.1577, + "step": 7122 + }, + { + "epoch": 1.268162393162393, + "grad_norm": 0.6965937614440918, + "learning_rate": 0.00015452571550743163, + "loss": 0.7835, + "step": 7123 + }, + { + "epoch": 1.2683404558404558, + "grad_norm": 0.6332844495773315, + "learning_rate": 0.00015451398132866376, + "loss": 0.9794, + "step": 7124 + }, + { + "epoch": 1.2685185185185186, + "grad_norm": 0.6719903349876404, + "learning_rate": 0.00015450224608180765, + "loss": 0.9795, + "step": 7125 + }, + { + "epoch": 1.2686965811965811, + "grad_norm": 0.567414402961731, + "learning_rate": 0.00015449050976709328, + "loss": 0.9737, + "step": 7126 + }, + { + "epoch": 1.2688746438746439, + "grad_norm": 0.6810645461082458, + "learning_rate": 0.0001544787723847505, + "loss": 1.2358, + "step": 7127 + }, + { + "epoch": 1.2690527065527066, + "grad_norm": 0.6693191528320312, + "learning_rate": 0.00015446703393500938, + "loss": 0.9475, + "step": 7128 + }, + { + "epoch": 1.2692307692307692, + "grad_norm": 0.7077522277832031, + "learning_rate": 0.00015445529441809988, + "loss": 1.013, + "step": 7129 + }, + { + "epoch": 1.269408831908832, + "grad_norm": 0.6596258878707886, + "learning_rate": 0.000154443553834252, + "loss": 1.1506, + "step": 7130 + }, + { + "epoch": 1.2695868945868947, + "grad_norm": 0.6721500754356384, + "learning_rate": 0.0001544318121836958, + "loss": 0.8848, + "step": 7131 + }, + { + "epoch": 1.2697649572649572, + "grad_norm": 0.6943998336791992, + "learning_rate": 0.00015442006946666132, + "loss": 1.1118, + "step": 7132 + }, + { + "epoch": 1.26994301994302, + "grad_norm": 0.6132234930992126, + "learning_rate": 0.0001544083256833786, + "loss": 0.9932, + "step": 7133 + }, + { + "epoch": 1.2701210826210827, + "grad_norm": 0.7337939739227295, + "learning_rate": 0.00015439658083407775, + "loss": 1.0973, + "step": 7134 + }, + { + "epoch": 1.2702991452991452, + "grad_norm": 0.6551772356033325, + "learning_rate": 0.00015438483491898893, + "loss": 1.0006, + "step": 7135 + }, + { + "epoch": 1.270477207977208, + "grad_norm": 0.660068929195404, + "learning_rate": 0.00015437308793834223, + "loss": 0.9291, + "step": 7136 + }, + { + "epoch": 1.2706552706552707, + "grad_norm": 0.7622788548469543, + "learning_rate": 0.00015436133989236783, + "loss": 1.0782, + "step": 7137 + }, + { + "epoch": 1.2708333333333333, + "grad_norm": 0.848494291305542, + "learning_rate": 0.00015434959078129587, + "loss": 1.2001, + "step": 7138 + }, + { + "epoch": 1.271011396011396, + "grad_norm": 0.6222602725028992, + "learning_rate": 0.0001543378406053566, + "loss": 1.011, + "step": 7139 + }, + { + "epoch": 1.2711894586894588, + "grad_norm": 0.6164663434028625, + "learning_rate": 0.00015432608936478026, + "loss": 1.0282, + "step": 7140 + }, + { + "epoch": 1.2713675213675213, + "grad_norm": 0.7236546277999878, + "learning_rate": 0.000154314337059797, + "loss": 1.0112, + "step": 7141 + }, + { + "epoch": 1.271545584045584, + "grad_norm": 0.6891111135482788, + "learning_rate": 0.00015430258369063715, + "loss": 1.1191, + "step": 7142 + }, + { + "epoch": 1.2717236467236468, + "grad_norm": 0.6600295901298523, + "learning_rate": 0.00015429082925753099, + "loss": 0.9561, + "step": 7143 + }, + { + "epoch": 1.2719017094017093, + "grad_norm": 0.6819902062416077, + "learning_rate": 0.0001542790737607088, + "loss": 1.0631, + "step": 7144 + }, + { + "epoch": 1.272079772079772, + "grad_norm": 0.6518470644950867, + "learning_rate": 0.0001542673172004009, + "loss": 1.0806, + "step": 7145 + }, + { + "epoch": 1.2722578347578348, + "grad_norm": 0.737501859664917, + "learning_rate": 0.00015425555957683767, + "loss": 1.0144, + "step": 7146 + }, + { + "epoch": 1.2724358974358974, + "grad_norm": 0.6245740652084351, + "learning_rate": 0.00015424380089024944, + "loss": 1.0612, + "step": 7147 + }, + { + "epoch": 1.27261396011396, + "grad_norm": 0.7118125557899475, + "learning_rate": 0.0001542320411408666, + "loss": 1.1458, + "step": 7148 + }, + { + "epoch": 1.2727920227920229, + "grad_norm": 0.6965761780738831, + "learning_rate": 0.00015422028032891958, + "loss": 0.8052, + "step": 7149 + }, + { + "epoch": 1.2729700854700854, + "grad_norm": 0.7661466598510742, + "learning_rate": 0.0001542085184546388, + "loss": 1.1245, + "step": 7150 + }, + { + "epoch": 1.2731481481481481, + "grad_norm": 0.7238876223564148, + "learning_rate": 0.00015419675551825475, + "loss": 0.9346, + "step": 7151 + }, + { + "epoch": 1.273326210826211, + "grad_norm": 0.669562041759491, + "learning_rate": 0.0001541849915199978, + "loss": 0.7816, + "step": 7152 + }, + { + "epoch": 1.2735042735042734, + "grad_norm": 0.6799174547195435, + "learning_rate": 0.00015417322646009855, + "loss": 1.047, + "step": 7153 + }, + { + "epoch": 1.2736823361823362, + "grad_norm": 0.6012796759605408, + "learning_rate": 0.00015416146033878745, + "loss": 1.0101, + "step": 7154 + }, + { + "epoch": 1.273860398860399, + "grad_norm": 0.7008427977561951, + "learning_rate": 0.00015414969315629505, + "loss": 1.1321, + "step": 7155 + }, + { + "epoch": 1.2740384615384617, + "grad_norm": 0.6555556058883667, + "learning_rate": 0.0001541379249128519, + "loss": 0.9926, + "step": 7156 + }, + { + "epoch": 1.2742165242165242, + "grad_norm": 0.6324251294136047, + "learning_rate": 0.00015412615560868854, + "loss": 0.9051, + "step": 7157 + }, + { + "epoch": 1.274394586894587, + "grad_norm": 0.6035568714141846, + "learning_rate": 0.0001541143852440356, + "loss": 0.8248, + "step": 7158 + }, + { + "epoch": 1.2745726495726495, + "grad_norm": 0.6733569502830505, + "learning_rate": 0.0001541026138191237, + "loss": 0.9149, + "step": 7159 + }, + { + "epoch": 1.2747507122507122, + "grad_norm": 0.8306798338890076, + "learning_rate": 0.0001540908413341835, + "loss": 1.0694, + "step": 7160 + }, + { + "epoch": 1.274928774928775, + "grad_norm": 0.6649713516235352, + "learning_rate": 0.00015407906778944563, + "loss": 1.1358, + "step": 7161 + }, + { + "epoch": 1.2751068376068377, + "grad_norm": 0.6889697909355164, + "learning_rate": 0.00015406729318514074, + "loss": 1.0096, + "step": 7162 + }, + { + "epoch": 1.2752849002849003, + "grad_norm": 0.6948645114898682, + "learning_rate": 0.0001540555175214996, + "loss": 1.0649, + "step": 7163 + }, + { + "epoch": 1.275462962962963, + "grad_norm": 0.6844844818115234, + "learning_rate": 0.0001540437407987528, + "loss": 0.884, + "step": 7164 + }, + { + "epoch": 1.2756410256410255, + "grad_norm": 0.7124526500701904, + "learning_rate": 0.00015403196301713124, + "loss": 1.1307, + "step": 7165 + }, + { + "epoch": 1.2758190883190883, + "grad_norm": 0.7328375577926636, + "learning_rate": 0.00015402018417686556, + "loss": 1.0348, + "step": 7166 + }, + { + "epoch": 1.275997150997151, + "grad_norm": 0.5872696042060852, + "learning_rate": 0.00015400840427818663, + "loss": 0.9827, + "step": 7167 + }, + { + "epoch": 1.2761752136752138, + "grad_norm": 0.6370702385902405, + "learning_rate": 0.00015399662332132519, + "loss": 0.9171, + "step": 7168 + }, + { + "epoch": 1.2763532763532763, + "grad_norm": 0.6481866240501404, + "learning_rate": 0.00015398484130651205, + "loss": 0.8704, + "step": 7169 + }, + { + "epoch": 1.276531339031339, + "grad_norm": 0.598739743232727, + "learning_rate": 0.00015397305823397812, + "loss": 0.8097, + "step": 7170 + }, + { + "epoch": 1.2767094017094016, + "grad_norm": 0.5941228270530701, + "learning_rate": 0.00015396127410395423, + "loss": 0.8853, + "step": 7171 + }, + { + "epoch": 1.2768874643874644, + "grad_norm": 0.6485885381698608, + "learning_rate": 0.00015394948891667127, + "loss": 0.702, + "step": 7172 + }, + { + "epoch": 1.2770655270655271, + "grad_norm": 0.5314942598342896, + "learning_rate": 0.00015393770267236017, + "loss": 0.7899, + "step": 7173 + }, + { + "epoch": 1.2772435897435899, + "grad_norm": 0.6113781929016113, + "learning_rate": 0.00015392591537125182, + "loss": 0.9871, + "step": 7174 + }, + { + "epoch": 1.2774216524216524, + "grad_norm": 0.5625866651535034, + "learning_rate": 0.00015391412701357715, + "loss": 0.8246, + "step": 7175 + }, + { + "epoch": 1.2775997150997151, + "grad_norm": 0.6006998419761658, + "learning_rate": 0.00015390233759956718, + "loss": 0.899, + "step": 7176 + }, + { + "epoch": 1.2777777777777777, + "grad_norm": 0.6916918158531189, + "learning_rate": 0.0001538905471294529, + "loss": 1.0443, + "step": 7177 + }, + { + "epoch": 1.2779558404558404, + "grad_norm": 0.6263536810874939, + "learning_rate": 0.00015387875560346525, + "loss": 0.9159, + "step": 7178 + }, + { + "epoch": 1.2781339031339032, + "grad_norm": 0.6563085913658142, + "learning_rate": 0.00015386696302183535, + "loss": 0.994, + "step": 7179 + }, + { + "epoch": 1.278311965811966, + "grad_norm": 0.6312007904052734, + "learning_rate": 0.00015385516938479416, + "loss": 0.9148, + "step": 7180 + }, + { + "epoch": 1.2784900284900285, + "grad_norm": 0.6408209204673767, + "learning_rate": 0.00015384337469257284, + "loss": 1.0508, + "step": 7181 + }, + { + "epoch": 1.2786680911680912, + "grad_norm": 0.656234085559845, + "learning_rate": 0.00015383157894540244, + "loss": 0.9952, + "step": 7182 + }, + { + "epoch": 1.2788461538461537, + "grad_norm": 0.7401639819145203, + "learning_rate": 0.00015381978214351407, + "loss": 1.1615, + "step": 7183 + }, + { + "epoch": 1.2790242165242165, + "grad_norm": 0.5746055841445923, + "learning_rate": 0.00015380798428713885, + "loss": 0.9142, + "step": 7184 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 0.8061720728874207, + "learning_rate": 0.00015379618537650797, + "loss": 1.13, + "step": 7185 + }, + { + "epoch": 1.279380341880342, + "grad_norm": 0.6336073875427246, + "learning_rate": 0.0001537843854118526, + "loss": 1.0581, + "step": 7186 + }, + { + "epoch": 1.2795584045584045, + "grad_norm": 0.6549856066703796, + "learning_rate": 0.0001537725843934039, + "loss": 1.09, + "step": 7187 + }, + { + "epoch": 1.2797364672364673, + "grad_norm": 0.5759010910987854, + "learning_rate": 0.00015376078232139315, + "loss": 0.8441, + "step": 7188 + }, + { + "epoch": 1.2799145299145298, + "grad_norm": 0.5733884572982788, + "learning_rate": 0.00015374897919605152, + "loss": 0.9086, + "step": 7189 + }, + { + "epoch": 1.2800925925925926, + "grad_norm": 0.6505870819091797, + "learning_rate": 0.0001537371750176103, + "loss": 1.1683, + "step": 7190 + }, + { + "epoch": 1.2802706552706553, + "grad_norm": 0.6744688153266907, + "learning_rate": 0.00015372536978630077, + "loss": 0.9483, + "step": 7191 + }, + { + "epoch": 1.280448717948718, + "grad_norm": 0.598098874092102, + "learning_rate": 0.0001537135635023542, + "loss": 0.7747, + "step": 7192 + }, + { + "epoch": 1.2806267806267806, + "grad_norm": 0.6711761951446533, + "learning_rate": 0.00015370175616600195, + "loss": 1.1897, + "step": 7193 + }, + { + "epoch": 1.2808048433048433, + "grad_norm": 0.6207453608512878, + "learning_rate": 0.00015368994777747536, + "loss": 1.0063, + "step": 7194 + }, + { + "epoch": 1.2809829059829059, + "grad_norm": 0.6701686382293701, + "learning_rate": 0.00015367813833700575, + "loss": 1.0864, + "step": 7195 + }, + { + "epoch": 1.2811609686609686, + "grad_norm": 0.5916469693183899, + "learning_rate": 0.00015366632784482456, + "loss": 0.8786, + "step": 7196 + }, + { + "epoch": 1.2813390313390314, + "grad_norm": 0.6567547917366028, + "learning_rate": 0.00015365451630116312, + "loss": 0.9977, + "step": 7197 + }, + { + "epoch": 1.2815170940170941, + "grad_norm": 0.7287433743476868, + "learning_rate": 0.00015364270370625294, + "loss": 1.1248, + "step": 7198 + }, + { + "epoch": 1.2816951566951567, + "grad_norm": 0.7736039161682129, + "learning_rate": 0.0001536308900603254, + "loss": 0.9832, + "step": 7199 + }, + { + "epoch": 1.2818732193732194, + "grad_norm": 0.6799852252006531, + "learning_rate": 0.00015361907536361194, + "loss": 1.0275, + "step": 7200 + }, + { + "epoch": 1.282051282051282, + "grad_norm": 0.5975812673568726, + "learning_rate": 0.00015360725961634407, + "loss": 1.0516, + "step": 7201 + }, + { + "epoch": 1.2822293447293447, + "grad_norm": 0.616307258605957, + "learning_rate": 0.00015359544281875337, + "loss": 0.8095, + "step": 7202 + }, + { + "epoch": 1.2824074074074074, + "grad_norm": 0.6357580423355103, + "learning_rate": 0.00015358362497107126, + "loss": 0.9186, + "step": 7203 + }, + { + "epoch": 1.2825854700854702, + "grad_norm": 0.679333508014679, + "learning_rate": 0.00015357180607352935, + "loss": 0.9433, + "step": 7204 + }, + { + "epoch": 1.2827635327635327, + "grad_norm": 0.6345439553260803, + "learning_rate": 0.00015355998612635914, + "loss": 0.9186, + "step": 7205 + }, + { + "epoch": 1.2829415954415955, + "grad_norm": 0.6256508827209473, + "learning_rate": 0.00015354816512979231, + "loss": 0.9984, + "step": 7206 + }, + { + "epoch": 1.283119658119658, + "grad_norm": 0.7973852753639221, + "learning_rate": 0.00015353634308406044, + "loss": 1.1145, + "step": 7207 + }, + { + "epoch": 1.2832977207977208, + "grad_norm": 0.711125910282135, + "learning_rate": 0.0001535245199893951, + "loss": 1.1947, + "step": 7208 + }, + { + "epoch": 1.2834757834757835, + "grad_norm": 0.6096055507659912, + "learning_rate": 0.00015351269584602798, + "loss": 1.0078, + "step": 7209 + }, + { + "epoch": 1.2836538461538463, + "grad_norm": 0.7089232802391052, + "learning_rate": 0.00015350087065419077, + "loss": 1.112, + "step": 7210 + }, + { + "epoch": 1.2838319088319088, + "grad_norm": 0.716199517250061, + "learning_rate": 0.00015348904441411508, + "loss": 1.1015, + "step": 7211 + }, + { + "epoch": 1.2840099715099715, + "grad_norm": 0.6374632716178894, + "learning_rate": 0.00015347721712603276, + "loss": 1.0519, + "step": 7212 + }, + { + "epoch": 1.284188034188034, + "grad_norm": 0.6500036716461182, + "learning_rate": 0.0001534653887901754, + "loss": 1.1719, + "step": 7213 + }, + { + "epoch": 1.2843660968660968, + "grad_norm": 0.7249937653541565, + "learning_rate": 0.00015345355940677485, + "loss": 1.0188, + "step": 7214 + }, + { + "epoch": 1.2845441595441596, + "grad_norm": 0.6645919680595398, + "learning_rate": 0.00015344172897606285, + "loss": 0.9788, + "step": 7215 + }, + { + "epoch": 1.2847222222222223, + "grad_norm": 0.7032710313796997, + "learning_rate": 0.00015342989749827113, + "loss": 1.1093, + "step": 7216 + }, + { + "epoch": 1.2849002849002849, + "grad_norm": 0.622767984867096, + "learning_rate": 0.0001534180649736316, + "loss": 0.8978, + "step": 7217 + }, + { + "epoch": 1.2850783475783476, + "grad_norm": 0.7499693036079407, + "learning_rate": 0.00015340623140237605, + "loss": 1.2232, + "step": 7218 + }, + { + "epoch": 1.2852564102564101, + "grad_norm": 0.6308625936508179, + "learning_rate": 0.00015339439678473636, + "loss": 0.8621, + "step": 7219 + }, + { + "epoch": 1.2854344729344729, + "grad_norm": 0.6513667106628418, + "learning_rate": 0.00015338256112094434, + "loss": 1.0541, + "step": 7220 + }, + { + "epoch": 1.2856125356125356, + "grad_norm": 0.6080937385559082, + "learning_rate": 0.00015337072441123193, + "loss": 0.8474, + "step": 7221 + }, + { + "epoch": 1.2857905982905984, + "grad_norm": 0.6742652058601379, + "learning_rate": 0.00015335888665583104, + "loss": 1.0172, + "step": 7222 + }, + { + "epoch": 1.285968660968661, + "grad_norm": 0.620810866355896, + "learning_rate": 0.00015334704785497364, + "loss": 1.049, + "step": 7223 + }, + { + "epoch": 1.2861467236467237, + "grad_norm": 0.5733018517494202, + "learning_rate": 0.00015333520800889165, + "loss": 0.7371, + "step": 7224 + }, + { + "epoch": 1.2863247863247862, + "grad_norm": 0.6447640061378479, + "learning_rate": 0.00015332336711781702, + "loss": 0.9925, + "step": 7225 + }, + { + "epoch": 1.286502849002849, + "grad_norm": 0.6764999628067017, + "learning_rate": 0.00015331152518198183, + "loss": 0.9052, + "step": 7226 + }, + { + "epoch": 1.2866809116809117, + "grad_norm": 0.6492836475372314, + "learning_rate": 0.00015329968220161803, + "loss": 0.9493, + "step": 7227 + }, + { + "epoch": 1.2868589743589745, + "grad_norm": 0.666157603263855, + "learning_rate": 0.00015328783817695766, + "loss": 1.0626, + "step": 7228 + }, + { + "epoch": 1.287037037037037, + "grad_norm": 0.7098026871681213, + "learning_rate": 0.00015327599310823283, + "loss": 1.0461, + "step": 7229 + }, + { + "epoch": 1.2872150997150997, + "grad_norm": 0.637778103351593, + "learning_rate": 0.00015326414699567555, + "loss": 0.9383, + "step": 7230 + }, + { + "epoch": 1.2873931623931623, + "grad_norm": 0.6816399693489075, + "learning_rate": 0.00015325229983951798, + "loss": 1.0647, + "step": 7231 + }, + { + "epoch": 1.287571225071225, + "grad_norm": 0.668689489364624, + "learning_rate": 0.0001532404516399922, + "loss": 1.0479, + "step": 7232 + }, + { + "epoch": 1.2877492877492878, + "grad_norm": 0.6459103226661682, + "learning_rate": 0.0001532286023973304, + "loss": 1.1751, + "step": 7233 + }, + { + "epoch": 1.2879273504273505, + "grad_norm": 0.679999589920044, + "learning_rate": 0.00015321675211176468, + "loss": 0.7541, + "step": 7234 + }, + { + "epoch": 1.288105413105413, + "grad_norm": 0.5415067672729492, + "learning_rate": 0.00015320490078352724, + "loss": 0.822, + "step": 7235 + }, + { + "epoch": 1.2882834757834758, + "grad_norm": 0.6817963719367981, + "learning_rate": 0.00015319304841285032, + "loss": 0.9424, + "step": 7236 + }, + { + "epoch": 1.2884615384615383, + "grad_norm": 0.6187505125999451, + "learning_rate": 0.0001531811949999661, + "loss": 0.8596, + "step": 7237 + }, + { + "epoch": 1.288639601139601, + "grad_norm": 0.6737838387489319, + "learning_rate": 0.00015316934054510685, + "loss": 1.0046, + "step": 7238 + }, + { + "epoch": 1.2888176638176638, + "grad_norm": 0.6445996761322021, + "learning_rate": 0.00015315748504850482, + "loss": 1.01, + "step": 7239 + }, + { + "epoch": 1.2889957264957266, + "grad_norm": 0.7279136180877686, + "learning_rate": 0.0001531456285103923, + "loss": 0.9066, + "step": 7240 + }, + { + "epoch": 1.289173789173789, + "grad_norm": 0.6619178652763367, + "learning_rate": 0.00015313377093100153, + "loss": 0.8977, + "step": 7241 + }, + { + "epoch": 1.2893518518518519, + "grad_norm": 0.7644323110580444, + "learning_rate": 0.000153121912310565, + "loss": 1.3085, + "step": 7242 + }, + { + "epoch": 1.2895299145299146, + "grad_norm": 0.645882248878479, + "learning_rate": 0.00015311005264931487, + "loss": 1.0337, + "step": 7243 + }, + { + "epoch": 1.2897079772079771, + "grad_norm": 0.6868017911911011, + "learning_rate": 0.0001530981919474836, + "loss": 0.9616, + "step": 7244 + }, + { + "epoch": 1.28988603988604, + "grad_norm": 0.7176693677902222, + "learning_rate": 0.00015308633020530362, + "loss": 1.1975, + "step": 7245 + }, + { + "epoch": 1.2900641025641026, + "grad_norm": 0.7358015775680542, + "learning_rate": 0.00015307446742300718, + "loss": 0.9308, + "step": 7246 + }, + { + "epoch": 1.2902421652421652, + "grad_norm": 0.7330248355865479, + "learning_rate": 0.00015306260360082688, + "loss": 0.9518, + "step": 7247 + }, + { + "epoch": 1.290420227920228, + "grad_norm": 0.6571981310844421, + "learning_rate": 0.00015305073873899503, + "loss": 0.9531, + "step": 7248 + }, + { + "epoch": 1.2905982905982907, + "grad_norm": 0.5968486666679382, + "learning_rate": 0.00015303887283774417, + "loss": 0.9245, + "step": 7249 + }, + { + "epoch": 1.2907763532763532, + "grad_norm": 0.6398176550865173, + "learning_rate": 0.0001530270058973068, + "loss": 1.0452, + "step": 7250 + }, + { + "epoch": 1.290954415954416, + "grad_norm": 0.5462267994880676, + "learning_rate": 0.00015301513791791542, + "loss": 0.8451, + "step": 7251 + }, + { + "epoch": 1.2911324786324787, + "grad_norm": 0.7536166906356812, + "learning_rate": 0.00015300326889980252, + "loss": 1.0086, + "step": 7252 + }, + { + "epoch": 1.2913105413105412, + "grad_norm": 0.6208569407463074, + "learning_rate": 0.00015299139884320065, + "loss": 0.7437, + "step": 7253 + }, + { + "epoch": 1.291488603988604, + "grad_norm": 0.7025452852249146, + "learning_rate": 0.00015297952774834242, + "loss": 0.8874, + "step": 7254 + }, + { + "epoch": 1.2916666666666667, + "grad_norm": 0.6758308410644531, + "learning_rate": 0.00015296765561546041, + "loss": 1.0378, + "step": 7255 + }, + { + "epoch": 1.2918447293447293, + "grad_norm": 0.7170431613922119, + "learning_rate": 0.00015295578244478724, + "loss": 1.0111, + "step": 7256 + }, + { + "epoch": 1.292022792022792, + "grad_norm": 0.6263511180877686, + "learning_rate": 0.00015294390823655544, + "loss": 0.7836, + "step": 7257 + }, + { + "epoch": 1.2922008547008548, + "grad_norm": 0.5887803435325623, + "learning_rate": 0.0001529320329909978, + "loss": 1.068, + "step": 7258 + }, + { + "epoch": 1.2923789173789173, + "grad_norm": 0.5955889821052551, + "learning_rate": 0.00015292015670834692, + "loss": 0.8903, + "step": 7259 + }, + { + "epoch": 1.29255698005698, + "grad_norm": 0.630449652671814, + "learning_rate": 0.00015290827938883552, + "loss": 1.1096, + "step": 7260 + }, + { + "epoch": 1.2927350427350428, + "grad_norm": 0.7405480146408081, + "learning_rate": 0.00015289640103269625, + "loss": 1.0648, + "step": 7261 + }, + { + "epoch": 1.2929131054131053, + "grad_norm": 0.6082221865653992, + "learning_rate": 0.00015288452164016191, + "loss": 0.9266, + "step": 7262 + }, + { + "epoch": 1.293091168091168, + "grad_norm": 0.6211720108985901, + "learning_rate": 0.00015287264121146524, + "loss": 0.849, + "step": 7263 + }, + { + "epoch": 1.2932692307692308, + "grad_norm": 0.6481043100357056, + "learning_rate": 0.00015286075974683898, + "loss": 0.7761, + "step": 7264 + }, + { + "epoch": 1.2934472934472934, + "grad_norm": 0.5957167744636536, + "learning_rate": 0.00015284887724651593, + "loss": 0.8942, + "step": 7265 + }, + { + "epoch": 1.2936253561253561, + "grad_norm": 0.7272268533706665, + "learning_rate": 0.00015283699371072894, + "loss": 1.0913, + "step": 7266 + }, + { + "epoch": 1.2938034188034189, + "grad_norm": 0.5902758836746216, + "learning_rate": 0.0001528251091397108, + "loss": 1.1045, + "step": 7267 + }, + { + "epoch": 1.2939814814814814, + "grad_norm": 0.6382482051849365, + "learning_rate": 0.00015281322353369436, + "loss": 0.9265, + "step": 7268 + }, + { + "epoch": 1.2941595441595442, + "grad_norm": 0.6556048393249512, + "learning_rate": 0.00015280133689291256, + "loss": 1.0536, + "step": 7269 + }, + { + "epoch": 1.294337606837607, + "grad_norm": 0.680895209312439, + "learning_rate": 0.00015278944921759822, + "loss": 0.9996, + "step": 7270 + }, + { + "epoch": 1.2945156695156697, + "grad_norm": 0.670317530632019, + "learning_rate": 0.00015277756050798428, + "loss": 1.1402, + "step": 7271 + }, + { + "epoch": 1.2946937321937322, + "grad_norm": 0.6312688589096069, + "learning_rate": 0.0001527656707643037, + "loss": 1.0669, + "step": 7272 + }, + { + "epoch": 1.294871794871795, + "grad_norm": 0.6267009973526001, + "learning_rate": 0.0001527537799867894, + "loss": 0.8985, + "step": 7273 + }, + { + "epoch": 1.2950498575498575, + "grad_norm": 0.7069001197814941, + "learning_rate": 0.00015274188817567436, + "loss": 0.9478, + "step": 7274 + }, + { + "epoch": 1.2952279202279202, + "grad_norm": 0.7229067087173462, + "learning_rate": 0.00015272999533119162, + "loss": 0.9005, + "step": 7275 + }, + { + "epoch": 1.295405982905983, + "grad_norm": 0.6254632472991943, + "learning_rate": 0.00015271810145357412, + "loss": 0.9746, + "step": 7276 + }, + { + "epoch": 1.2955840455840457, + "grad_norm": 0.6772669553756714, + "learning_rate": 0.00015270620654305494, + "loss": 1.1714, + "step": 7277 + }, + { + "epoch": 1.2957621082621082, + "grad_norm": 0.605576753616333, + "learning_rate": 0.00015269431059986713, + "loss": 0.7735, + "step": 7278 + }, + { + "epoch": 1.295940170940171, + "grad_norm": 0.7144771814346313, + "learning_rate": 0.00015268241362424378, + "loss": 0.9757, + "step": 7279 + }, + { + "epoch": 1.2961182336182335, + "grad_norm": 0.5275486707687378, + "learning_rate": 0.00015267051561641798, + "loss": 0.5669, + "step": 7280 + }, + { + "epoch": 1.2962962962962963, + "grad_norm": 0.6619452238082886, + "learning_rate": 0.00015265861657662284, + "loss": 0.9511, + "step": 7281 + }, + { + "epoch": 1.296474358974359, + "grad_norm": 0.6788223385810852, + "learning_rate": 0.00015264671650509147, + "loss": 1.2649, + "step": 7282 + }, + { + "epoch": 1.2966524216524218, + "grad_norm": 0.6198732852935791, + "learning_rate": 0.00015263481540205706, + "loss": 1.0659, + "step": 7283 + }, + { + "epoch": 1.2968304843304843, + "grad_norm": 0.6038815975189209, + "learning_rate": 0.0001526229132677528, + "loss": 1.0655, + "step": 7284 + }, + { + "epoch": 1.297008547008547, + "grad_norm": 0.7616196870803833, + "learning_rate": 0.00015261101010241186, + "loss": 1.131, + "step": 7285 + }, + { + "epoch": 1.2971866096866096, + "grad_norm": 0.7002527713775635, + "learning_rate": 0.00015259910590626746, + "loss": 1.1375, + "step": 7286 + }, + { + "epoch": 1.2973646723646723, + "grad_norm": 0.6067437529563904, + "learning_rate": 0.00015258720067955284, + "loss": 0.9306, + "step": 7287 + }, + { + "epoch": 1.297542735042735, + "grad_norm": 0.653232216835022, + "learning_rate": 0.00015257529442250128, + "loss": 1.107, + "step": 7288 + }, + { + "epoch": 1.2977207977207978, + "grad_norm": 0.6969175934791565, + "learning_rate": 0.00015256338713534603, + "loss": 0.8365, + "step": 7289 + }, + { + "epoch": 1.2978988603988604, + "grad_norm": 0.6176731586456299, + "learning_rate": 0.00015255147881832043, + "loss": 0.9707, + "step": 7290 + }, + { + "epoch": 1.2980769230769231, + "grad_norm": 0.6543741822242737, + "learning_rate": 0.00015253956947165772, + "loss": 0.7714, + "step": 7291 + }, + { + "epoch": 1.2982549857549857, + "grad_norm": 0.5224920511245728, + "learning_rate": 0.00015252765909559135, + "loss": 0.7469, + "step": 7292 + }, + { + "epoch": 1.2984330484330484, + "grad_norm": 0.638708770275116, + "learning_rate": 0.00015251574769035455, + "loss": 1.0965, + "step": 7293 + }, + { + "epoch": 1.2986111111111112, + "grad_norm": 0.6742943525314331, + "learning_rate": 0.0001525038352561808, + "loss": 1.1286, + "step": 7294 + }, + { + "epoch": 1.298789173789174, + "grad_norm": 0.6027839183807373, + "learning_rate": 0.00015249192179330346, + "loss": 0.8824, + "step": 7295 + }, + { + "epoch": 1.2989672364672364, + "grad_norm": 0.7462167143821716, + "learning_rate": 0.00015248000730195597, + "loss": 0.94, + "step": 7296 + }, + { + "epoch": 1.2991452991452992, + "grad_norm": 0.6972534656524658, + "learning_rate": 0.00015246809178237172, + "loss": 1.0664, + "step": 7297 + }, + { + "epoch": 1.2993233618233617, + "grad_norm": 0.569949209690094, + "learning_rate": 0.0001524561752347842, + "loss": 0.691, + "step": 7298 + }, + { + "epoch": 1.2995014245014245, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00015244425765942695, + "loss": 1.083, + "step": 7299 + }, + { + "epoch": 1.2996794871794872, + "grad_norm": 0.6927483677864075, + "learning_rate": 0.00015243233905653337, + "loss": 1.0068, + "step": 7300 + }, + { + "epoch": 1.29985754985755, + "grad_norm": 0.752824604511261, + "learning_rate": 0.00015242041942633704, + "loss": 0.9946, + "step": 7301 + }, + { + "epoch": 1.3000356125356125, + "grad_norm": 0.6532080173492432, + "learning_rate": 0.0001524084987690715, + "loss": 1.2326, + "step": 7302 + }, + { + "epoch": 1.3002136752136753, + "grad_norm": 0.7954180836677551, + "learning_rate": 0.0001523965770849703, + "loss": 1.1105, + "step": 7303 + }, + { + "epoch": 1.3003917378917378, + "grad_norm": 0.5971781015396118, + "learning_rate": 0.000152384654374267, + "loss": 1.0984, + "step": 7304 + }, + { + "epoch": 1.3005698005698005, + "grad_norm": 0.7778682112693787, + "learning_rate": 0.0001523727306371952, + "loss": 1.0795, + "step": 7305 + }, + { + "epoch": 1.3007478632478633, + "grad_norm": 0.6712004542350769, + "learning_rate": 0.00015236080587398856, + "loss": 1.0814, + "step": 7306 + }, + { + "epoch": 1.300925925925926, + "grad_norm": 0.581048846244812, + "learning_rate": 0.00015234888008488066, + "loss": 0.9868, + "step": 7307 + }, + { + "epoch": 1.3011039886039886, + "grad_norm": 0.697695791721344, + "learning_rate": 0.00015233695327010523, + "loss": 1.1045, + "step": 7308 + }, + { + "epoch": 1.3012820512820513, + "grad_norm": 0.6858421564102173, + "learning_rate": 0.00015232502542989593, + "loss": 1.0769, + "step": 7309 + }, + { + "epoch": 1.3014601139601139, + "grad_norm": 0.6312826871871948, + "learning_rate": 0.00015231309656448642, + "loss": 0.9523, + "step": 7310 + }, + { + "epoch": 1.3016381766381766, + "grad_norm": 0.9243300557136536, + "learning_rate": 0.0001523011666741105, + "loss": 0.947, + "step": 7311 + }, + { + "epoch": 1.3018162393162394, + "grad_norm": 0.6808217763900757, + "learning_rate": 0.00015228923575900184, + "loss": 0.8631, + "step": 7312 + }, + { + "epoch": 1.301994301994302, + "grad_norm": 0.6713891625404358, + "learning_rate": 0.00015227730381939424, + "loss": 0.9157, + "step": 7313 + }, + { + "epoch": 1.3021723646723646, + "grad_norm": 0.6802582740783691, + "learning_rate": 0.00015226537085552146, + "loss": 1.041, + "step": 7314 + }, + { + "epoch": 1.3023504273504274, + "grad_norm": 0.6543951034545898, + "learning_rate": 0.0001522534368676173, + "loss": 0.8709, + "step": 7315 + }, + { + "epoch": 1.30252849002849, + "grad_norm": 0.6290678381919861, + "learning_rate": 0.0001522415018559156, + "loss": 1.0568, + "step": 7316 + }, + { + "epoch": 1.3027065527065527, + "grad_norm": 0.6590015292167664, + "learning_rate": 0.0001522295658206502, + "loss": 0.9919, + "step": 7317 + }, + { + "epoch": 1.3028846153846154, + "grad_norm": 0.6374103426933289, + "learning_rate": 0.00015221762876205494, + "loss": 0.878, + "step": 7318 + }, + { + "epoch": 1.3030626780626782, + "grad_norm": 0.7247048616409302, + "learning_rate": 0.00015220569068036372, + "loss": 1.061, + "step": 7319 + }, + { + "epoch": 1.3032407407407407, + "grad_norm": 0.6450991630554199, + "learning_rate": 0.00015219375157581047, + "loss": 0.9389, + "step": 7320 + }, + { + "epoch": 1.3034188034188035, + "grad_norm": 0.8039840459823608, + "learning_rate": 0.00015218181144862903, + "loss": 1.0692, + "step": 7321 + }, + { + "epoch": 1.303596866096866, + "grad_norm": 0.6539456248283386, + "learning_rate": 0.00015216987029905346, + "loss": 1.0478, + "step": 7322 + }, + { + "epoch": 1.3037749287749287, + "grad_norm": 0.60880047082901, + "learning_rate": 0.00015215792812731758, + "loss": 0.8412, + "step": 7323 + }, + { + "epoch": 1.3039529914529915, + "grad_norm": 0.6757258176803589, + "learning_rate": 0.0001521459849336555, + "loss": 0.896, + "step": 7324 + }, + { + "epoch": 1.3041310541310542, + "grad_norm": 0.6735622882843018, + "learning_rate": 0.00015213404071830116, + "loss": 1.1078, + "step": 7325 + }, + { + "epoch": 1.3043091168091168, + "grad_norm": 0.7321233749389648, + "learning_rate": 0.00015212209548148858, + "loss": 1.1021, + "step": 7326 + }, + { + "epoch": 1.3044871794871795, + "grad_norm": 0.6678910851478577, + "learning_rate": 0.00015211014922345182, + "loss": 1.0043, + "step": 7327 + }, + { + "epoch": 1.304665242165242, + "grad_norm": 0.6876940727233887, + "learning_rate": 0.0001520982019444249, + "loss": 1.0376, + "step": 7328 + }, + { + "epoch": 1.3048433048433048, + "grad_norm": 0.6171853542327881, + "learning_rate": 0.00015208625364464195, + "loss": 0.839, + "step": 7329 + }, + { + "epoch": 1.3050213675213675, + "grad_norm": 0.6449569463729858, + "learning_rate": 0.0001520743043243371, + "loss": 1.0908, + "step": 7330 + }, + { + "epoch": 1.3051994301994303, + "grad_norm": 0.6894628405570984, + "learning_rate": 0.00015206235398374443, + "loss": 1.0263, + "step": 7331 + }, + { + "epoch": 1.3053774928774928, + "grad_norm": 0.5853552222251892, + "learning_rate": 0.00015205040262309804, + "loss": 0.8342, + "step": 7332 + }, + { + "epoch": 1.3055555555555556, + "grad_norm": 0.5934799313545227, + "learning_rate": 0.00015203845024263214, + "loss": 0.9464, + "step": 7333 + }, + { + "epoch": 1.305733618233618, + "grad_norm": 0.668927788734436, + "learning_rate": 0.00015202649684258095, + "loss": 0.9018, + "step": 7334 + }, + { + "epoch": 1.3059116809116809, + "grad_norm": 0.676810085773468, + "learning_rate": 0.0001520145424231786, + "loss": 0.9284, + "step": 7335 + }, + { + "epoch": 1.3060897435897436, + "grad_norm": 0.6223878264427185, + "learning_rate": 0.00015200258698465935, + "loss": 1.0779, + "step": 7336 + }, + { + "epoch": 1.3062678062678064, + "grad_norm": 0.6092363595962524, + "learning_rate": 0.00015199063052725745, + "loss": 0.8602, + "step": 7337 + }, + { + "epoch": 1.306445868945869, + "grad_norm": 0.7668731212615967, + "learning_rate": 0.00015197867305120712, + "loss": 1.0756, + "step": 7338 + }, + { + "epoch": 1.3066239316239316, + "grad_norm": 0.6485331654548645, + "learning_rate": 0.00015196671455674268, + "loss": 1.0193, + "step": 7339 + }, + { + "epoch": 1.3068019943019942, + "grad_norm": 0.5661036372184753, + "learning_rate": 0.0001519547550440984, + "loss": 0.8321, + "step": 7340 + }, + { + "epoch": 1.306980056980057, + "grad_norm": 0.6270507574081421, + "learning_rate": 0.00015194279451350866, + "loss": 0.6403, + "step": 7341 + }, + { + "epoch": 1.3071581196581197, + "grad_norm": 0.7283764481544495, + "learning_rate": 0.00015193083296520773, + "loss": 1.0401, + "step": 7342 + }, + { + "epoch": 1.3073361823361824, + "grad_norm": 0.658835232257843, + "learning_rate": 0.00015191887039943, + "loss": 1.0172, + "step": 7343 + }, + { + "epoch": 1.307514245014245, + "grad_norm": 0.6288984417915344, + "learning_rate": 0.00015190690681640988, + "loss": 0.8649, + "step": 7344 + }, + { + "epoch": 1.3076923076923077, + "grad_norm": 0.666442334651947, + "learning_rate": 0.00015189494221638176, + "loss": 1.0757, + "step": 7345 + }, + { + "epoch": 1.3078703703703702, + "grad_norm": 0.6116433143615723, + "learning_rate": 0.00015188297659958003, + "loss": 0.9244, + "step": 7346 + }, + { + "epoch": 1.308048433048433, + "grad_norm": 0.6378964185714722, + "learning_rate": 0.0001518710099662392, + "loss": 0.9629, + "step": 7347 + }, + { + "epoch": 1.3082264957264957, + "grad_norm": 0.6258945465087891, + "learning_rate": 0.00015185904231659357, + "loss": 0.8524, + "step": 7348 + }, + { + "epoch": 1.3084045584045585, + "grad_norm": 0.6498504877090454, + "learning_rate": 0.0001518470736508778, + "loss": 0.9685, + "step": 7349 + }, + { + "epoch": 1.308582621082621, + "grad_norm": 0.6928247809410095, + "learning_rate": 0.00015183510396932635, + "loss": 0.9054, + "step": 7350 + }, + { + "epoch": 1.3087606837606838, + "grad_norm": 0.6350936889648438, + "learning_rate": 0.0001518231332721737, + "loss": 1.0039, + "step": 7351 + }, + { + "epoch": 1.3089387464387463, + "grad_norm": 0.6652286648750305, + "learning_rate": 0.00015181116155965437, + "loss": 0.8946, + "step": 7352 + }, + { + "epoch": 1.309116809116809, + "grad_norm": 0.6554864048957825, + "learning_rate": 0.000151799188832003, + "loss": 0.9518, + "step": 7353 + }, + { + "epoch": 1.3092948717948718, + "grad_norm": 0.7523114085197449, + "learning_rate": 0.0001517872150894541, + "loss": 0.9462, + "step": 7354 + }, + { + "epoch": 1.3094729344729346, + "grad_norm": 0.7113336324691772, + "learning_rate": 0.0001517752403322423, + "loss": 1.2347, + "step": 7355 + }, + { + "epoch": 1.309650997150997, + "grad_norm": 0.6461622714996338, + "learning_rate": 0.00015176326456060223, + "loss": 0.8891, + "step": 7356 + }, + { + "epoch": 1.3098290598290598, + "grad_norm": 0.7429143190383911, + "learning_rate": 0.00015175128777476852, + "loss": 1.1944, + "step": 7357 + }, + { + "epoch": 1.3100071225071226, + "grad_norm": 0.6816306114196777, + "learning_rate": 0.00015173930997497585, + "loss": 1.1445, + "step": 7358 + }, + { + "epoch": 1.3101851851851851, + "grad_norm": 0.6644450426101685, + "learning_rate": 0.00015172733116145884, + "loss": 0.9808, + "step": 7359 + }, + { + "epoch": 1.3103632478632479, + "grad_norm": 0.6921063661575317, + "learning_rate": 0.00015171535133445225, + "loss": 1.0162, + "step": 7360 + }, + { + "epoch": 1.3105413105413106, + "grad_norm": 0.6386187672615051, + "learning_rate": 0.00015170337049419082, + "loss": 0.9951, + "step": 7361 + }, + { + "epoch": 1.3107193732193732, + "grad_norm": 0.6505418419837952, + "learning_rate": 0.0001516913886409092, + "loss": 0.8872, + "step": 7362 + }, + { + "epoch": 1.310897435897436, + "grad_norm": 0.6415576934814453, + "learning_rate": 0.00015167940577484222, + "loss": 1.056, + "step": 7363 + }, + { + "epoch": 1.3110754985754987, + "grad_norm": 0.6691195964813232, + "learning_rate": 0.00015166742189622458, + "loss": 1.0561, + "step": 7364 + }, + { + "epoch": 1.3112535612535612, + "grad_norm": 0.6376257538795471, + "learning_rate": 0.00015165543700529122, + "loss": 0.8499, + "step": 7365 + }, + { + "epoch": 1.311431623931624, + "grad_norm": 0.6270790696144104, + "learning_rate": 0.00015164345110227684, + "loss": 1.0244, + "step": 7366 + }, + { + "epoch": 1.3116096866096867, + "grad_norm": 0.7120122313499451, + "learning_rate": 0.0001516314641874163, + "loss": 1.0476, + "step": 7367 + }, + { + "epoch": 1.3117877492877492, + "grad_norm": 0.6152660250663757, + "learning_rate": 0.0001516194762609445, + "loss": 0.897, + "step": 7368 + }, + { + "epoch": 1.311965811965812, + "grad_norm": 0.7578088045120239, + "learning_rate": 0.00015160748732309626, + "loss": 1.1609, + "step": 7369 + }, + { + "epoch": 1.3121438746438747, + "grad_norm": 0.6594924330711365, + "learning_rate": 0.00015159549737410656, + "loss": 1.1706, + "step": 7370 + }, + { + "epoch": 1.3123219373219372, + "grad_norm": 0.6559173464775085, + "learning_rate": 0.00015158350641421024, + "loss": 0.9452, + "step": 7371 + }, + { + "epoch": 1.3125, + "grad_norm": 0.6667516231536865, + "learning_rate": 0.00015157151444364226, + "loss": 0.8153, + "step": 7372 + }, + { + "epoch": 1.3126780626780628, + "grad_norm": 0.7054803371429443, + "learning_rate": 0.00015155952146263761, + "loss": 0.9887, + "step": 7373 + }, + { + "epoch": 1.3128561253561253, + "grad_norm": 0.7035902142524719, + "learning_rate": 0.00015154752747143123, + "loss": 1.1832, + "step": 7374 + }, + { + "epoch": 1.313034188034188, + "grad_norm": 0.6297488212585449, + "learning_rate": 0.00015153553247025813, + "loss": 0.9602, + "step": 7375 + }, + { + "epoch": 1.3132122507122508, + "grad_norm": 0.6851378083229065, + "learning_rate": 0.00015152353645935335, + "loss": 1.0743, + "step": 7376 + }, + { + "epoch": 1.3133903133903133, + "grad_norm": 0.6215537786483765, + "learning_rate": 0.00015151153943895187, + "loss": 0.9484, + "step": 7377 + }, + { + "epoch": 1.313568376068376, + "grad_norm": 0.6848666071891785, + "learning_rate": 0.0001514995414092888, + "loss": 1.0978, + "step": 7378 + }, + { + "epoch": 1.3137464387464388, + "grad_norm": 0.7527492642402649, + "learning_rate": 0.00015148754237059918, + "loss": 1.083, + "step": 7379 + }, + { + "epoch": 1.3139245014245013, + "grad_norm": 0.6264588236808777, + "learning_rate": 0.00015147554232311814, + "loss": 0.9995, + "step": 7380 + }, + { + "epoch": 1.314102564102564, + "grad_norm": 0.6666619181632996, + "learning_rate": 0.00015146354126708075, + "loss": 1.0156, + "step": 7381 + }, + { + "epoch": 1.3142806267806268, + "grad_norm": 0.6626597046852112, + "learning_rate": 0.00015145153920272222, + "loss": 1.0047, + "step": 7382 + }, + { + "epoch": 1.3144586894586894, + "grad_norm": 0.5975428223609924, + "learning_rate": 0.0001514395361302776, + "loss": 0.806, + "step": 7383 + }, + { + "epoch": 1.3146367521367521, + "grad_norm": 0.6509957909584045, + "learning_rate": 0.00015142753204998218, + "loss": 0.8871, + "step": 7384 + }, + { + "epoch": 1.3148148148148149, + "grad_norm": 0.6672926545143127, + "learning_rate": 0.00015141552696207108, + "loss": 0.9616, + "step": 7385 + }, + { + "epoch": 1.3149928774928774, + "grad_norm": 0.6965435147285461, + "learning_rate": 0.00015140352086677954, + "loss": 1.124, + "step": 7386 + }, + { + "epoch": 1.3151709401709402, + "grad_norm": 0.6559258103370667, + "learning_rate": 0.00015139151376434277, + "loss": 1.0271, + "step": 7387 + }, + { + "epoch": 1.315349002849003, + "grad_norm": 0.7613587379455566, + "learning_rate": 0.00015137950565499608, + "loss": 1.0349, + "step": 7388 + }, + { + "epoch": 1.3155270655270654, + "grad_norm": 0.7001944780349731, + "learning_rate": 0.0001513674965389747, + "loss": 0.8551, + "step": 7389 + }, + { + "epoch": 1.3157051282051282, + "grad_norm": 0.6087043285369873, + "learning_rate": 0.0001513554864165139, + "loss": 0.7118, + "step": 7390 + }, + { + "epoch": 1.315883190883191, + "grad_norm": 0.71526700258255, + "learning_rate": 0.00015134347528784908, + "loss": 1.0478, + "step": 7391 + }, + { + "epoch": 1.3160612535612537, + "grad_norm": 0.6182073950767517, + "learning_rate": 0.00015133146315321548, + "loss": 0.9474, + "step": 7392 + }, + { + "epoch": 1.3162393162393162, + "grad_norm": 0.7771387696266174, + "learning_rate": 0.0001513194500128485, + "loss": 1.0544, + "step": 7393 + }, + { + "epoch": 1.316417378917379, + "grad_norm": 0.7108260989189148, + "learning_rate": 0.00015130743586698353, + "loss": 0.8813, + "step": 7394 + }, + { + "epoch": 1.3165954415954415, + "grad_norm": 0.7057309150695801, + "learning_rate": 0.0001512954207158559, + "loss": 0.899, + "step": 7395 + }, + { + "epoch": 1.3167735042735043, + "grad_norm": 0.6139237880706787, + "learning_rate": 0.00015128340455970106, + "loss": 0.8885, + "step": 7396 + }, + { + "epoch": 1.316951566951567, + "grad_norm": 0.7166598439216614, + "learning_rate": 0.00015127138739875443, + "loss": 0.9792, + "step": 7397 + }, + { + "epoch": 1.3171296296296298, + "grad_norm": 0.6916186809539795, + "learning_rate": 0.00015125936923325153, + "loss": 0.8871, + "step": 7398 + }, + { + "epoch": 1.3173076923076923, + "grad_norm": 0.7189087271690369, + "learning_rate": 0.0001512473500634277, + "loss": 0.8302, + "step": 7399 + }, + { + "epoch": 1.317485754985755, + "grad_norm": 0.5739200115203857, + "learning_rate": 0.00015123532988951853, + "loss": 0.9137, + "step": 7400 + }, + { + "epoch": 1.3176638176638176, + "grad_norm": 0.7661057114601135, + "learning_rate": 0.00015122330871175952, + "loss": 1.1255, + "step": 7401 + }, + { + "epoch": 1.3178418803418803, + "grad_norm": 0.6487592458724976, + "learning_rate": 0.00015121128653038617, + "loss": 1.0519, + "step": 7402 + }, + { + "epoch": 1.318019943019943, + "grad_norm": 0.693134605884552, + "learning_rate": 0.00015119926334563406, + "loss": 0.9585, + "step": 7403 + }, + { + "epoch": 1.3181980056980058, + "grad_norm": 0.5895997285842896, + "learning_rate": 0.0001511872391577387, + "loss": 0.8033, + "step": 7404 + }, + { + "epoch": 1.3183760683760684, + "grad_norm": 0.654876172542572, + "learning_rate": 0.00015117521396693575, + "loss": 1.0082, + "step": 7405 + }, + { + "epoch": 1.318554131054131, + "grad_norm": 0.5877239108085632, + "learning_rate": 0.0001511631877734608, + "loss": 1.0147, + "step": 7406 + }, + { + "epoch": 1.3187321937321936, + "grad_norm": 0.6109837889671326, + "learning_rate": 0.00015115116057754944, + "loss": 0.7498, + "step": 7407 + }, + { + "epoch": 1.3189102564102564, + "grad_norm": 0.643856942653656, + "learning_rate": 0.00015113913237943736, + "loss": 1.0417, + "step": 7408 + }, + { + "epoch": 1.3190883190883191, + "grad_norm": 0.654077410697937, + "learning_rate": 0.00015112710317936022, + "loss": 1.1809, + "step": 7409 + }, + { + "epoch": 1.319266381766382, + "grad_norm": 0.6785375475883484, + "learning_rate": 0.00015111507297755367, + "loss": 0.9447, + "step": 7410 + }, + { + "epoch": 1.3194444444444444, + "grad_norm": 0.6513382196426392, + "learning_rate": 0.00015110304177425347, + "loss": 0.8286, + "step": 7411 + }, + { + "epoch": 1.3196225071225072, + "grad_norm": 0.6536405682563782, + "learning_rate": 0.00015109100956969533, + "loss": 1.1959, + "step": 7412 + }, + { + "epoch": 1.3198005698005697, + "grad_norm": 0.6633172035217285, + "learning_rate": 0.00015107897636411498, + "loss": 0.8839, + "step": 7413 + }, + { + "epoch": 1.3199786324786325, + "grad_norm": 0.5773791670799255, + "learning_rate": 0.00015106694215774821, + "loss": 0.9785, + "step": 7414 + }, + { + "epoch": 1.3201566951566952, + "grad_norm": 0.7005468010902405, + "learning_rate": 0.00015105490695083078, + "loss": 1.0752, + "step": 7415 + }, + { + "epoch": 1.320334757834758, + "grad_norm": 0.6509538888931274, + "learning_rate": 0.0001510428707435985, + "loss": 0.9886, + "step": 7416 + }, + { + "epoch": 1.3205128205128205, + "grad_norm": 0.6607788801193237, + "learning_rate": 0.0001510308335362872, + "loss": 0.9756, + "step": 7417 + }, + { + "epoch": 1.3206908831908832, + "grad_norm": 0.5977858304977417, + "learning_rate": 0.00015101879532913274, + "loss": 1.0574, + "step": 7418 + }, + { + "epoch": 1.3208689458689458, + "grad_norm": 0.6478607058525085, + "learning_rate": 0.00015100675612237096, + "loss": 1.0076, + "step": 7419 + }, + { + "epoch": 1.3210470085470085, + "grad_norm": 0.6386681199073792, + "learning_rate": 0.00015099471591623775, + "loss": 0.9639, + "step": 7420 + }, + { + "epoch": 1.3212250712250713, + "grad_norm": 0.6348143815994263, + "learning_rate": 0.000150982674710969, + "loss": 1.0226, + "step": 7421 + }, + { + "epoch": 1.321403133903134, + "grad_norm": 0.6737388372421265, + "learning_rate": 0.00015097063250680068, + "loss": 0.9985, + "step": 7422 + }, + { + "epoch": 1.3215811965811965, + "grad_norm": 0.7302656769752502, + "learning_rate": 0.00015095858930396866, + "loss": 0.9969, + "step": 7423 + }, + { + "epoch": 1.3217592592592593, + "grad_norm": 0.7062691450119019, + "learning_rate": 0.00015094654510270898, + "loss": 0.9137, + "step": 7424 + }, + { + "epoch": 1.3219373219373218, + "grad_norm": 0.6289888620376587, + "learning_rate": 0.00015093449990325754, + "loss": 0.9231, + "step": 7425 + }, + { + "epoch": 1.3221153846153846, + "grad_norm": 0.643284261226654, + "learning_rate": 0.0001509224537058504, + "loss": 0.8981, + "step": 7426 + }, + { + "epoch": 1.3222934472934473, + "grad_norm": 0.7019244432449341, + "learning_rate": 0.00015091040651072355, + "loss": 0.9994, + "step": 7427 + }, + { + "epoch": 1.32247150997151, + "grad_norm": 0.5982088446617126, + "learning_rate": 0.0001508983583181131, + "loss": 0.9365, + "step": 7428 + }, + { + "epoch": 1.3226495726495726, + "grad_norm": 0.6086063385009766, + "learning_rate": 0.00015088630912825498, + "loss": 0.8621, + "step": 7429 + }, + { + "epoch": 1.3228276353276354, + "grad_norm": 0.6829213500022888, + "learning_rate": 0.00015087425894138535, + "loss": 1.1959, + "step": 7430 + }, + { + "epoch": 1.323005698005698, + "grad_norm": 0.6538017392158508, + "learning_rate": 0.00015086220775774033, + "loss": 0.9412, + "step": 7431 + }, + { + "epoch": 1.3231837606837606, + "grad_norm": 0.6334070563316345, + "learning_rate": 0.00015085015557755597, + "loss": 0.9044, + "step": 7432 + }, + { + "epoch": 1.3233618233618234, + "grad_norm": 0.6514624357223511, + "learning_rate": 0.00015083810240106845, + "loss": 0.8859, + "step": 7433 + }, + { + "epoch": 1.3235398860398861, + "grad_norm": 0.7130434513092041, + "learning_rate": 0.00015082604822851397, + "loss": 1.2845, + "step": 7434 + }, + { + "epoch": 1.3237179487179487, + "grad_norm": 0.609419584274292, + "learning_rate": 0.00015081399306012862, + "loss": 1.0725, + "step": 7435 + }, + { + "epoch": 1.3238960113960114, + "grad_norm": 0.586807370185852, + "learning_rate": 0.0001508019368961486, + "loss": 0.9032, + "step": 7436 + }, + { + "epoch": 1.324074074074074, + "grad_norm": 0.6937291026115417, + "learning_rate": 0.0001507898797368102, + "loss": 0.7975, + "step": 7437 + }, + { + "epoch": 1.3242521367521367, + "grad_norm": 0.6804966330528259, + "learning_rate": 0.00015077782158234962, + "loss": 1.1018, + "step": 7438 + }, + { + "epoch": 1.3244301994301995, + "grad_norm": 0.6110677123069763, + "learning_rate": 0.0001507657624330031, + "loss": 0.7988, + "step": 7439 + }, + { + "epoch": 1.3246082621082622, + "grad_norm": 0.6340961456298828, + "learning_rate": 0.0001507537022890069, + "loss": 0.844, + "step": 7440 + }, + { + "epoch": 1.3247863247863247, + "grad_norm": 0.7291021943092346, + "learning_rate": 0.00015074164115059735, + "loss": 0.9867, + "step": 7441 + }, + { + "epoch": 1.3249643874643875, + "grad_norm": 0.6818505525588989, + "learning_rate": 0.00015072957901801076, + "loss": 1.1541, + "step": 7442 + }, + { + "epoch": 1.32514245014245, + "grad_norm": 0.6174707412719727, + "learning_rate": 0.00015071751589148345, + "loss": 1.1679, + "step": 7443 + }, + { + "epoch": 1.3253205128205128, + "grad_norm": 0.6481367945671082, + "learning_rate": 0.00015070545177125176, + "loss": 1.0955, + "step": 7444 + }, + { + "epoch": 1.3254985754985755, + "grad_norm": 0.6752339005470276, + "learning_rate": 0.00015069338665755203, + "loss": 0.8651, + "step": 7445 + }, + { + "epoch": 1.3256766381766383, + "grad_norm": 0.6608055830001831, + "learning_rate": 0.00015068132055062077, + "loss": 0.9553, + "step": 7446 + }, + { + "epoch": 1.3258547008547008, + "grad_norm": 0.5933246612548828, + "learning_rate": 0.00015066925345069425, + "loss": 0.8584, + "step": 7447 + }, + { + "epoch": 1.3260327635327636, + "grad_norm": 0.6301844716072083, + "learning_rate": 0.000150657185358009, + "loss": 0.8583, + "step": 7448 + }, + { + "epoch": 1.326210826210826, + "grad_norm": 0.7359434962272644, + "learning_rate": 0.00015064511627280145, + "loss": 1.0905, + "step": 7449 + }, + { + "epoch": 1.3263888888888888, + "grad_norm": 0.6334579586982727, + "learning_rate": 0.00015063304619530806, + "loss": 0.9814, + "step": 7450 + }, + { + "epoch": 1.3265669515669516, + "grad_norm": 0.6974197626113892, + "learning_rate": 0.00015062097512576528, + "loss": 0.9302, + "step": 7451 + }, + { + "epoch": 1.3267450142450143, + "grad_norm": 0.6895849704742432, + "learning_rate": 0.00015060890306440965, + "loss": 1.0175, + "step": 7452 + }, + { + "epoch": 1.3269230769230769, + "grad_norm": 0.5938003659248352, + "learning_rate": 0.00015059683001147767, + "loss": 0.8084, + "step": 7453 + }, + { + "epoch": 1.3271011396011396, + "grad_norm": 0.6821470856666565, + "learning_rate": 0.00015058475596720596, + "loss": 0.9897, + "step": 7454 + }, + { + "epoch": 1.3272792022792022, + "grad_norm": 0.5507164001464844, + "learning_rate": 0.00015057268093183104, + "loss": 0.7012, + "step": 7455 + }, + { + "epoch": 1.327457264957265, + "grad_norm": 0.6216199398040771, + "learning_rate": 0.00015056060490558945, + "loss": 1.0281, + "step": 7456 + }, + { + "epoch": 1.3276353276353277, + "grad_norm": 0.6674157977104187, + "learning_rate": 0.00015054852788871787, + "loss": 0.8776, + "step": 7457 + }, + { + "epoch": 1.3278133903133904, + "grad_norm": 0.666963517665863, + "learning_rate": 0.0001505364498814529, + "loss": 1.0742, + "step": 7458 + }, + { + "epoch": 1.327991452991453, + "grad_norm": 0.6205331683158875, + "learning_rate": 0.00015052437088403114, + "loss": 1.1109, + "step": 7459 + }, + { + "epoch": 1.3281695156695157, + "grad_norm": 0.6402750611305237, + "learning_rate": 0.00015051229089668933, + "loss": 1.0648, + "step": 7460 + }, + { + "epoch": 1.3283475783475782, + "grad_norm": 0.7445703744888306, + "learning_rate": 0.00015050020991966406, + "loss": 0.8989, + "step": 7461 + }, + { + "epoch": 1.328525641025641, + "grad_norm": 0.8131299614906311, + "learning_rate": 0.00015048812795319212, + "loss": 0.9552, + "step": 7462 + }, + { + "epoch": 1.3287037037037037, + "grad_norm": 0.7007313966751099, + "learning_rate": 0.00015047604499751017, + "loss": 0.9899, + "step": 7463 + }, + { + "epoch": 1.3288817663817665, + "grad_norm": 0.60536789894104, + "learning_rate": 0.000150463961052855, + "loss": 0.7694, + "step": 7464 + }, + { + "epoch": 1.329059829059829, + "grad_norm": 0.6910434365272522, + "learning_rate": 0.00015045187611946331, + "loss": 0.9575, + "step": 7465 + }, + { + "epoch": 1.3292378917378918, + "grad_norm": 0.7693352103233337, + "learning_rate": 0.00015043979019757194, + "loss": 1.1987, + "step": 7466 + }, + { + "epoch": 1.3294159544159543, + "grad_norm": 0.6675218939781189, + "learning_rate": 0.00015042770328741763, + "loss": 1.0099, + "step": 7467 + }, + { + "epoch": 1.329594017094017, + "grad_norm": 0.8040883541107178, + "learning_rate": 0.00015041561538923722, + "loss": 0.9493, + "step": 7468 + }, + { + "epoch": 1.3297720797720798, + "grad_norm": 0.6765826344490051, + "learning_rate": 0.00015040352650326762, + "loss": 1.1035, + "step": 7469 + }, + { + "epoch": 1.3299501424501425, + "grad_norm": 0.7099924087524414, + "learning_rate": 0.0001503914366297456, + "loss": 0.9198, + "step": 7470 + }, + { + "epoch": 1.330128205128205, + "grad_norm": 0.6673682928085327, + "learning_rate": 0.00015037934576890804, + "loss": 1.0234, + "step": 7471 + }, + { + "epoch": 1.3303062678062678, + "grad_norm": 0.7022300958633423, + "learning_rate": 0.00015036725392099184, + "loss": 1.3875, + "step": 7472 + }, + { + "epoch": 1.3304843304843303, + "grad_norm": 0.6997060179710388, + "learning_rate": 0.00015035516108623394, + "loss": 0.8114, + "step": 7473 + }, + { + "epoch": 1.330662393162393, + "grad_norm": 0.6262350678443909, + "learning_rate": 0.00015034306726487127, + "loss": 1.128, + "step": 7474 + }, + { + "epoch": 1.3308404558404558, + "grad_norm": 0.6330382227897644, + "learning_rate": 0.00015033097245714078, + "loss": 0.9032, + "step": 7475 + }, + { + "epoch": 1.3310185185185186, + "grad_norm": 0.6527551412582397, + "learning_rate": 0.00015031887666327944, + "loss": 0.9311, + "step": 7476 + }, + { + "epoch": 1.3311965811965811, + "grad_norm": 0.6754798889160156, + "learning_rate": 0.00015030677988352422, + "loss": 1.0626, + "step": 7477 + }, + { + "epoch": 1.3313746438746439, + "grad_norm": 0.6397945284843445, + "learning_rate": 0.00015029468211811216, + "loss": 0.9222, + "step": 7478 + }, + { + "epoch": 1.3315527065527066, + "grad_norm": 0.8163481950759888, + "learning_rate": 0.0001502825833672803, + "loss": 1.1827, + "step": 7479 + }, + { + "epoch": 1.3317307692307692, + "grad_norm": 0.6645621657371521, + "learning_rate": 0.00015027048363126566, + "loss": 0.9744, + "step": 7480 + }, + { + "epoch": 1.331908831908832, + "grad_norm": 0.6943182349205017, + "learning_rate": 0.0001502583829103053, + "loss": 1.1597, + "step": 7481 + }, + { + "epoch": 1.3320868945868947, + "grad_norm": 0.6283710598945618, + "learning_rate": 0.00015024628120463636, + "loss": 0.9514, + "step": 7482 + }, + { + "epoch": 1.3322649572649572, + "grad_norm": 0.6159678101539612, + "learning_rate": 0.0001502341785144959, + "loss": 0.9752, + "step": 7483 + }, + { + "epoch": 1.33244301994302, + "grad_norm": 0.6259802579879761, + "learning_rate": 0.00015022207484012107, + "loss": 0.9356, + "step": 7484 + }, + { + "epoch": 1.3326210826210827, + "grad_norm": 0.7322365641593933, + "learning_rate": 0.00015020997018174904, + "loss": 1.2072, + "step": 7485 + }, + { + "epoch": 1.3327991452991452, + "grad_norm": 0.6323443651199341, + "learning_rate": 0.0001501978645396169, + "loss": 1.1661, + "step": 7486 + }, + { + "epoch": 1.332977207977208, + "grad_norm": 0.7811527848243713, + "learning_rate": 0.00015018575791396187, + "loss": 1.0304, + "step": 7487 + }, + { + "epoch": 1.3331552706552707, + "grad_norm": 0.7221232056617737, + "learning_rate": 0.0001501736503050212, + "loss": 0.8838, + "step": 7488 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.6980099081993103, + "learning_rate": 0.00015016154171303207, + "loss": 1.1841, + "step": 7489 + }, + { + "epoch": 1.333511396011396, + "grad_norm": 0.6802879571914673, + "learning_rate": 0.00015014943213823175, + "loss": 0.959, + "step": 7490 + }, + { + "epoch": 1.3336894586894588, + "grad_norm": 0.637698233127594, + "learning_rate": 0.00015013732158085746, + "loss": 1.0517, + "step": 7491 + }, + { + "epoch": 1.3338675213675213, + "grad_norm": 0.6386787295341492, + "learning_rate": 0.0001501252100411465, + "loss": 0.7125, + "step": 7492 + }, + { + "epoch": 1.334045584045584, + "grad_norm": 0.6287358403205872, + "learning_rate": 0.0001501130975193362, + "loss": 0.8913, + "step": 7493 + }, + { + "epoch": 1.3342236467236468, + "grad_norm": 0.6142337322235107, + "learning_rate": 0.00015010098401566386, + "loss": 0.8149, + "step": 7494 + }, + { + "epoch": 1.3344017094017093, + "grad_norm": 0.6369916200637817, + "learning_rate": 0.0001500888695303668, + "loss": 1.0186, + "step": 7495 + }, + { + "epoch": 1.334579772079772, + "grad_norm": 0.7526934146881104, + "learning_rate": 0.0001500767540636824, + "loss": 1.2421, + "step": 7496 + }, + { + "epoch": 1.3347578347578348, + "grad_norm": 0.7278095483779907, + "learning_rate": 0.00015006463761584802, + "loss": 0.9856, + "step": 7497 + }, + { + "epoch": 1.3349358974358974, + "grad_norm": 0.6165127158164978, + "learning_rate": 0.00015005252018710104, + "loss": 1.0041, + "step": 7498 + }, + { + "epoch": 1.33511396011396, + "grad_norm": 0.637856662273407, + "learning_rate": 0.00015004040177767896, + "loss": 0.9134, + "step": 7499 + }, + { + "epoch": 1.3352920227920229, + "grad_norm": 0.661227285861969, + "learning_rate": 0.00015002828238781912, + "loss": 1.0393, + "step": 7500 + }, + { + "epoch": 1.3354700854700854, + "grad_norm": 0.6061869859695435, + "learning_rate": 0.000150016162017759, + "loss": 0.8453, + "step": 7501 + }, + { + "epoch": 1.3356481481481481, + "grad_norm": 0.6938419938087463, + "learning_rate": 0.0001500040406677361, + "loss": 1.0338, + "step": 7502 + }, + { + "epoch": 1.335826210826211, + "grad_norm": 0.6672863960266113, + "learning_rate": 0.0001499919183379879, + "loss": 0.8765, + "step": 7503 + }, + { + "epoch": 1.3360042735042734, + "grad_norm": 0.6200515031814575, + "learning_rate": 0.00014997979502875193, + "loss": 0.8286, + "step": 7504 + }, + { + "epoch": 1.3361823361823362, + "grad_norm": 0.6287549138069153, + "learning_rate": 0.00014996767074026567, + "loss": 0.9761, + "step": 7505 + }, + { + "epoch": 1.336360398860399, + "grad_norm": 0.6036837100982666, + "learning_rate": 0.0001499555454727667, + "loss": 1.0506, + "step": 7506 + }, + { + "epoch": 1.3365384615384617, + "grad_norm": 0.6875260472297668, + "learning_rate": 0.0001499434192264926, + "loss": 1.001, + "step": 7507 + }, + { + "epoch": 1.3367165242165242, + "grad_norm": 0.6558469533920288, + "learning_rate": 0.00014993129200168096, + "loss": 0.6874, + "step": 7508 + }, + { + "epoch": 1.336894586894587, + "grad_norm": 0.604167103767395, + "learning_rate": 0.00014991916379856934, + "loss": 1.0173, + "step": 7509 + }, + { + "epoch": 1.3370726495726495, + "grad_norm": 0.5941442251205444, + "learning_rate": 0.00014990703461739544, + "loss": 0.8569, + "step": 7510 + }, + { + "epoch": 1.3372507122507122, + "grad_norm": 0.7645071148872375, + "learning_rate": 0.00014989490445839687, + "loss": 1.0172, + "step": 7511 + }, + { + "epoch": 1.337428774928775, + "grad_norm": 0.5491678714752197, + "learning_rate": 0.00014988277332181126, + "loss": 0.8018, + "step": 7512 + }, + { + "epoch": 1.3376068376068377, + "grad_norm": 0.583322286605835, + "learning_rate": 0.00014987064120787635, + "loss": 0.8704, + "step": 7513 + }, + { + "epoch": 1.3377849002849003, + "grad_norm": 0.7385724186897278, + "learning_rate": 0.00014985850811682984, + "loss": 1.1121, + "step": 7514 + }, + { + "epoch": 1.337962962962963, + "grad_norm": 0.6842585206031799, + "learning_rate": 0.00014984637404890941, + "loss": 0.914, + "step": 7515 + }, + { + "epoch": 1.3381410256410255, + "grad_norm": 0.6771186590194702, + "learning_rate": 0.00014983423900435285, + "loss": 1.0838, + "step": 7516 + }, + { + "epoch": 1.3383190883190883, + "grad_norm": 0.7562049031257629, + "learning_rate": 0.00014982210298339788, + "loss": 1.123, + "step": 7517 + }, + { + "epoch": 1.338497150997151, + "grad_norm": 0.7617804408073425, + "learning_rate": 0.0001498099659862823, + "loss": 0.9438, + "step": 7518 + }, + { + "epoch": 1.3386752136752138, + "grad_norm": 0.561958909034729, + "learning_rate": 0.00014979782801324392, + "loss": 0.8739, + "step": 7519 + }, + { + "epoch": 1.3388532763532763, + "grad_norm": 0.7726154923439026, + "learning_rate": 0.00014978568906452052, + "loss": 1.1306, + "step": 7520 + }, + { + "epoch": 1.339031339031339, + "grad_norm": 0.6658660173416138, + "learning_rate": 0.00014977354914035002, + "loss": 1.0214, + "step": 7521 + }, + { + "epoch": 1.3392094017094016, + "grad_norm": 0.6385402679443359, + "learning_rate": 0.00014976140824097015, + "loss": 0.8851, + "step": 7522 + }, + { + "epoch": 1.3393874643874644, + "grad_norm": 0.6315767168998718, + "learning_rate": 0.0001497492663666189, + "loss": 0.986, + "step": 7523 + }, + { + "epoch": 1.3395655270655271, + "grad_norm": 0.6379088759422302, + "learning_rate": 0.0001497371235175341, + "loss": 0.9322, + "step": 7524 + }, + { + "epoch": 1.3397435897435899, + "grad_norm": 0.6605859994888306, + "learning_rate": 0.0001497249796939537, + "loss": 1.1112, + "step": 7525 + }, + { + "epoch": 1.3399216524216524, + "grad_norm": 0.7342822551727295, + "learning_rate": 0.0001497128348961156, + "loss": 0.9798, + "step": 7526 + }, + { + "epoch": 1.3400997150997151, + "grad_norm": 0.5667192935943604, + "learning_rate": 0.0001497006891242578, + "loss": 0.7493, + "step": 7527 + }, + { + "epoch": 1.3402777777777777, + "grad_norm": 0.6106827855110168, + "learning_rate": 0.0001496885423786182, + "loss": 1.0924, + "step": 7528 + }, + { + "epoch": 1.3404558404558404, + "grad_norm": 0.6207202076911926, + "learning_rate": 0.00014967639465943486, + "loss": 1.1123, + "step": 7529 + }, + { + "epoch": 1.3406339031339032, + "grad_norm": 0.6272760033607483, + "learning_rate": 0.00014966424596694574, + "loss": 0.9275, + "step": 7530 + }, + { + "epoch": 1.340811965811966, + "grad_norm": 0.6485986113548279, + "learning_rate": 0.0001496520963013889, + "loss": 1.1491, + "step": 7531 + }, + { + "epoch": 1.3409900284900285, + "grad_norm": 0.5743561387062073, + "learning_rate": 0.00014963994566300238, + "loss": 1.1101, + "step": 7532 + }, + { + "epoch": 1.3411680911680912, + "grad_norm": 0.6508657336235046, + "learning_rate": 0.00014962779405202424, + "loss": 1.0368, + "step": 7533 + }, + { + "epoch": 1.3413461538461537, + "grad_norm": 0.6598748564720154, + "learning_rate": 0.00014961564146869259, + "loss": 1.1064, + "step": 7534 + }, + { + "epoch": 1.3415242165242165, + "grad_norm": 0.6722840070724487, + "learning_rate": 0.00014960348791324547, + "loss": 0.9758, + "step": 7535 + }, + { + "epoch": 1.3417022792022792, + "grad_norm": 0.5807220935821533, + "learning_rate": 0.00014959133338592108, + "loss": 0.9936, + "step": 7536 + }, + { + "epoch": 1.341880341880342, + "grad_norm": 0.6318647265434265, + "learning_rate": 0.00014957917788695752, + "loss": 0.907, + "step": 7537 + }, + { + "epoch": 1.3420584045584045, + "grad_norm": 0.6725485324859619, + "learning_rate": 0.00014956702141659295, + "loss": 0.988, + "step": 7538 + }, + { + "epoch": 1.3422364672364673, + "grad_norm": 0.6675217747688293, + "learning_rate": 0.0001495548639750656, + "loss": 1.0194, + "step": 7539 + }, + { + "epoch": 1.3424145299145298, + "grad_norm": 0.6976884603500366, + "learning_rate": 0.0001495427055626136, + "loss": 1.2515, + "step": 7540 + }, + { + "epoch": 1.3425925925925926, + "grad_norm": 0.654941737651825, + "learning_rate": 0.0001495305461794752, + "loss": 1.2072, + "step": 7541 + }, + { + "epoch": 1.3427706552706553, + "grad_norm": 0.7085291743278503, + "learning_rate": 0.00014951838582588864, + "loss": 0.9772, + "step": 7542 + }, + { + "epoch": 1.342948717948718, + "grad_norm": 0.6319566965103149, + "learning_rate": 0.00014950622450209217, + "loss": 1.0162, + "step": 7543 + }, + { + "epoch": 1.3431267806267806, + "grad_norm": 0.6272495985031128, + "learning_rate": 0.00014949406220832407, + "loss": 0.7985, + "step": 7544 + }, + { + "epoch": 1.3433048433048433, + "grad_norm": 0.6352069973945618, + "learning_rate": 0.00014948189894482266, + "loss": 1.0041, + "step": 7545 + }, + { + "epoch": 1.3434829059829059, + "grad_norm": 0.6071867346763611, + "learning_rate": 0.0001494697347118262, + "loss": 0.9486, + "step": 7546 + }, + { + "epoch": 1.3436609686609686, + "grad_norm": 0.6458829641342163, + "learning_rate": 0.00014945756950957308, + "loss": 0.9417, + "step": 7547 + }, + { + "epoch": 1.3438390313390314, + "grad_norm": 0.6472262740135193, + "learning_rate": 0.0001494454033383016, + "loss": 1.056, + "step": 7548 + }, + { + "epoch": 1.3440170940170941, + "grad_norm": 0.6985635161399841, + "learning_rate": 0.00014943323619825017, + "loss": 1.0483, + "step": 7549 + }, + { + "epoch": 1.3441951566951567, + "grad_norm": 0.6379460096359253, + "learning_rate": 0.00014942106808965718, + "loss": 0.9552, + "step": 7550 + }, + { + "epoch": 1.3443732193732194, + "grad_norm": 0.7036557793617249, + "learning_rate": 0.00014940889901276098, + "loss": 0.9647, + "step": 7551 + }, + { + "epoch": 1.344551282051282, + "grad_norm": 0.6697289943695068, + "learning_rate": 0.0001493967289678001, + "loss": 0.9029, + "step": 7552 + }, + { + "epoch": 1.3447293447293447, + "grad_norm": 0.6336250901222229, + "learning_rate": 0.00014938455795501286, + "loss": 0.9458, + "step": 7553 + }, + { + "epoch": 1.3449074074074074, + "grad_norm": 0.7279673218727112, + "learning_rate": 0.00014937238597463785, + "loss": 1.0228, + "step": 7554 + }, + { + "epoch": 1.3450854700854702, + "grad_norm": 0.6514406204223633, + "learning_rate": 0.00014936021302691349, + "loss": 0.8265, + "step": 7555 + }, + { + "epoch": 1.3452635327635327, + "grad_norm": 0.6405338644981384, + "learning_rate": 0.0001493480391120783, + "loss": 0.9516, + "step": 7556 + }, + { + "epoch": 1.3454415954415955, + "grad_norm": 0.6442672610282898, + "learning_rate": 0.00014933586423037076, + "loss": 0.9279, + "step": 7557 + }, + { + "epoch": 1.345619658119658, + "grad_norm": 0.7588633894920349, + "learning_rate": 0.00014932368838202945, + "loss": 1.0976, + "step": 7558 + }, + { + "epoch": 1.3457977207977208, + "grad_norm": 0.5536739230155945, + "learning_rate": 0.00014931151156729296, + "loss": 0.713, + "step": 7559 + }, + { + "epoch": 1.3459757834757835, + "grad_norm": 0.6897570490837097, + "learning_rate": 0.00014929933378639981, + "loss": 0.9521, + "step": 7560 + }, + { + "epoch": 1.3461538461538463, + "grad_norm": 0.6654927134513855, + "learning_rate": 0.00014928715503958863, + "loss": 0.8506, + "step": 7561 + }, + { + "epoch": 1.3463319088319088, + "grad_norm": 0.655806839466095, + "learning_rate": 0.00014927497532709808, + "loss": 0.8636, + "step": 7562 + }, + { + "epoch": 1.3465099715099715, + "grad_norm": 0.6547064185142517, + "learning_rate": 0.00014926279464916667, + "loss": 0.9155, + "step": 7563 + }, + { + "epoch": 1.346688034188034, + "grad_norm": 0.7555415034294128, + "learning_rate": 0.00014925061300603316, + "loss": 0.8791, + "step": 7564 + }, + { + "epoch": 1.3468660968660968, + "grad_norm": 0.7439392805099487, + "learning_rate": 0.0001492384303979362, + "loss": 1.1669, + "step": 7565 + }, + { + "epoch": 1.3470441595441596, + "grad_norm": 0.6016925573348999, + "learning_rate": 0.0001492262468251145, + "loss": 0.9811, + "step": 7566 + }, + { + "epoch": 1.3472222222222223, + "grad_norm": 0.644652783870697, + "learning_rate": 0.00014921406228780675, + "loss": 0.7096, + "step": 7567 + }, + { + "epoch": 1.3474002849002849, + "grad_norm": 0.721814751625061, + "learning_rate": 0.00014920187678625166, + "loss": 0.9933, + "step": 7568 + }, + { + "epoch": 1.3475783475783476, + "grad_norm": 0.6212092638015747, + "learning_rate": 0.000149189690320688, + "loss": 0.8499, + "step": 7569 + }, + { + "epoch": 1.3477564102564101, + "grad_norm": 0.6235958337783813, + "learning_rate": 0.00014917750289135455, + "loss": 0.9189, + "step": 7570 + }, + { + "epoch": 1.3479344729344729, + "grad_norm": 0.6309674978256226, + "learning_rate": 0.0001491653144984901, + "loss": 0.9744, + "step": 7571 + }, + { + "epoch": 1.3481125356125356, + "grad_norm": 0.7606496214866638, + "learning_rate": 0.00014915312514233344, + "loss": 1.0181, + "step": 7572 + }, + { + "epoch": 1.3482905982905984, + "grad_norm": 0.6892654895782471, + "learning_rate": 0.00014914093482312342, + "loss": 0.9517, + "step": 7573 + }, + { + "epoch": 1.348468660968661, + "grad_norm": 0.6746503114700317, + "learning_rate": 0.0001491287435410988, + "loss": 1.056, + "step": 7574 + }, + { + "epoch": 1.3486467236467237, + "grad_norm": 0.5892919301986694, + "learning_rate": 0.00014911655129649858, + "loss": 1.0515, + "step": 7575 + }, + { + "epoch": 1.3488247863247862, + "grad_norm": 0.6278096437454224, + "learning_rate": 0.0001491043580895615, + "loss": 0.864, + "step": 7576 + }, + { + "epoch": 1.349002849002849, + "grad_norm": 0.7017706632614136, + "learning_rate": 0.0001490921639205266, + "loss": 1.0618, + "step": 7577 + }, + { + "epoch": 1.3491809116809117, + "grad_norm": 0.7318746447563171, + "learning_rate": 0.00014907996878963268, + "loss": 0.9905, + "step": 7578 + }, + { + "epoch": 1.3493589743589745, + "grad_norm": 0.6485885977745056, + "learning_rate": 0.00014906777269711873, + "loss": 1.0498, + "step": 7579 + }, + { + "epoch": 1.349537037037037, + "grad_norm": 0.644902229309082, + "learning_rate": 0.00014905557564322372, + "loss": 0.885, + "step": 7580 + }, + { + "epoch": 1.3497150997150997, + "grad_norm": 0.6567610502243042, + "learning_rate": 0.0001490433776281866, + "loss": 0.8938, + "step": 7581 + }, + { + "epoch": 1.3498931623931623, + "grad_norm": 0.6233102679252625, + "learning_rate": 0.0001490311786522464, + "loss": 0.9007, + "step": 7582 + }, + { + "epoch": 1.350071225071225, + "grad_norm": 0.6962146759033203, + "learning_rate": 0.00014901897871564206, + "loss": 0.9257, + "step": 7583 + }, + { + "epoch": 1.3502492877492878, + "grad_norm": 0.6986933350563049, + "learning_rate": 0.00014900677781861266, + "loss": 1.0089, + "step": 7584 + }, + { + "epoch": 1.3504273504273505, + "grad_norm": 0.7527925968170166, + "learning_rate": 0.00014899457596139729, + "loss": 1.0762, + "step": 7585 + }, + { + "epoch": 1.350605413105413, + "grad_norm": 0.69191974401474, + "learning_rate": 0.00014898237314423494, + "loss": 0.9829, + "step": 7586 + }, + { + "epoch": 1.3507834757834758, + "grad_norm": 0.7866443395614624, + "learning_rate": 0.00014897016936736478, + "loss": 1.0911, + "step": 7587 + }, + { + "epoch": 1.3509615384615383, + "grad_norm": 0.7087522745132446, + "learning_rate": 0.00014895796463102587, + "loss": 1.0693, + "step": 7588 + }, + { + "epoch": 1.351139601139601, + "grad_norm": 0.704276442527771, + "learning_rate": 0.00014894575893545736, + "loss": 0.9082, + "step": 7589 + }, + { + "epoch": 1.3513176638176638, + "grad_norm": 0.7074487805366516, + "learning_rate": 0.00014893355228089833, + "loss": 0.8731, + "step": 7590 + }, + { + "epoch": 1.3514957264957266, + "grad_norm": 0.6542425155639648, + "learning_rate": 0.00014892134466758803, + "loss": 0.9325, + "step": 7591 + }, + { + "epoch": 1.351673789173789, + "grad_norm": 0.6577230095863342, + "learning_rate": 0.0001489091360957656, + "loss": 0.8468, + "step": 7592 + }, + { + "epoch": 1.3518518518518519, + "grad_norm": 0.638534426689148, + "learning_rate": 0.00014889692656567025, + "loss": 0.8598, + "step": 7593 + }, + { + "epoch": 1.3520299145299146, + "grad_norm": 0.751133918762207, + "learning_rate": 0.0001488847160775412, + "loss": 1.0006, + "step": 7594 + }, + { + "epoch": 1.3522079772079771, + "grad_norm": 0.6272708773612976, + "learning_rate": 0.00014887250463161767, + "loss": 0.8782, + "step": 7595 + }, + { + "epoch": 1.35238603988604, + "grad_norm": 0.7242439985275269, + "learning_rate": 0.00014886029222813897, + "loss": 1.2443, + "step": 7596 + }, + { + "epoch": 1.3525641025641026, + "grad_norm": 0.6199275851249695, + "learning_rate": 0.0001488480788673443, + "loss": 0.9211, + "step": 7597 + }, + { + "epoch": 1.3527421652421652, + "grad_norm": 0.6401306986808777, + "learning_rate": 0.00014883586454947305, + "loss": 0.8808, + "step": 7598 + }, + { + "epoch": 1.352920227920228, + "grad_norm": 0.6340938806533813, + "learning_rate": 0.00014882364927476443, + "loss": 0.9406, + "step": 7599 + }, + { + "epoch": 1.3530982905982907, + "grad_norm": 0.6388604044914246, + "learning_rate": 0.00014881143304345783, + "loss": 1.0674, + "step": 7600 + }, + { + "epoch": 1.3532763532763532, + "grad_norm": 0.7562061548233032, + "learning_rate": 0.00014879921585579263, + "loss": 1.0959, + "step": 7601 + }, + { + "epoch": 1.353454415954416, + "grad_norm": 0.6303606033325195, + "learning_rate": 0.00014878699771200815, + "loss": 0.9641, + "step": 7602 + }, + { + "epoch": 1.3536324786324787, + "grad_norm": 0.8623232841491699, + "learning_rate": 0.00014877477861234382, + "loss": 1.1529, + "step": 7603 + }, + { + "epoch": 1.3538105413105412, + "grad_norm": 0.6607624888420105, + "learning_rate": 0.00014876255855703896, + "loss": 0.6291, + "step": 7604 + }, + { + "epoch": 1.353988603988604, + "grad_norm": 0.6226931214332581, + "learning_rate": 0.0001487503375463331, + "loss": 0.7485, + "step": 7605 + }, + { + "epoch": 1.3541666666666667, + "grad_norm": 0.7626705169677734, + "learning_rate": 0.00014873811558046565, + "loss": 0.9694, + "step": 7606 + }, + { + "epoch": 1.3543447293447293, + "grad_norm": 0.5436057448387146, + "learning_rate": 0.00014872589265967605, + "loss": 0.6173, + "step": 7607 + }, + { + "epoch": 1.354522792022792, + "grad_norm": 0.7822177410125732, + "learning_rate": 0.00014871366878420382, + "loss": 1.0048, + "step": 7608 + }, + { + "epoch": 1.3547008547008548, + "grad_norm": 0.6955201625823975, + "learning_rate": 0.00014870144395428848, + "loss": 0.9487, + "step": 7609 + }, + { + "epoch": 1.3548789173789173, + "grad_norm": 0.6625505685806274, + "learning_rate": 0.00014868921817016943, + "loss": 0.9389, + "step": 7610 + }, + { + "epoch": 1.35505698005698, + "grad_norm": 0.6625354886054993, + "learning_rate": 0.00014867699143208634, + "loss": 0.9538, + "step": 7611 + }, + { + "epoch": 1.3552350427350428, + "grad_norm": 0.7426592707633972, + "learning_rate": 0.00014866476374027874, + "loss": 1.2566, + "step": 7612 + }, + { + "epoch": 1.3554131054131053, + "grad_norm": 0.6856544017791748, + "learning_rate": 0.00014865253509498616, + "loss": 0.9663, + "step": 7613 + }, + { + "epoch": 1.355591168091168, + "grad_norm": 0.6343915462493896, + "learning_rate": 0.00014864030549644825, + "loss": 0.9416, + "step": 7614 + }, + { + "epoch": 1.3557692307692308, + "grad_norm": 0.6319553256034851, + "learning_rate": 0.00014862807494490454, + "loss": 0.9335, + "step": 7615 + }, + { + "epoch": 1.3559472934472934, + "grad_norm": 0.6919772624969482, + "learning_rate": 0.00014861584344059476, + "loss": 0.8516, + "step": 7616 + }, + { + "epoch": 1.3561253561253561, + "grad_norm": 0.6405790448188782, + "learning_rate": 0.00014860361098375851, + "loss": 1.1278, + "step": 7617 + }, + { + "epoch": 1.3563034188034189, + "grad_norm": 0.7591732144355774, + "learning_rate": 0.00014859137757463548, + "loss": 1.0961, + "step": 7618 + }, + { + "epoch": 1.3564814814814814, + "grad_norm": 0.6166727542877197, + "learning_rate": 0.0001485791432134653, + "loss": 0.9358, + "step": 7619 + }, + { + "epoch": 1.3566595441595442, + "grad_norm": 0.7068707346916199, + "learning_rate": 0.00014856690790048777, + "loss": 0.8325, + "step": 7620 + }, + { + "epoch": 1.356837606837607, + "grad_norm": 0.8465402722358704, + "learning_rate": 0.00014855467163594257, + "loss": 1.0047, + "step": 7621 + }, + { + "epoch": 1.3570156695156697, + "grad_norm": 0.7403460741043091, + "learning_rate": 0.00014854243442006943, + "loss": 1.0907, + "step": 7622 + }, + { + "epoch": 1.3571937321937322, + "grad_norm": 0.6939566135406494, + "learning_rate": 0.00014853019625310813, + "loss": 0.9156, + "step": 7623 + }, + { + "epoch": 1.357371794871795, + "grad_norm": 0.6425924897193909, + "learning_rate": 0.0001485179571352984, + "loss": 0.8156, + "step": 7624 + }, + { + "epoch": 1.3575498575498575, + "grad_norm": 0.7091902494430542, + "learning_rate": 0.00014850571706688013, + "loss": 1.0483, + "step": 7625 + }, + { + "epoch": 1.3577279202279202, + "grad_norm": 0.663342297077179, + "learning_rate": 0.00014849347604809312, + "loss": 1.0405, + "step": 7626 + }, + { + "epoch": 1.357905982905983, + "grad_norm": 0.6727671027183533, + "learning_rate": 0.00014848123407917716, + "loss": 1.0389, + "step": 7627 + }, + { + "epoch": 1.3580840455840457, + "grad_norm": 0.6572692394256592, + "learning_rate": 0.0001484689911603721, + "loss": 1.0489, + "step": 7628 + }, + { + "epoch": 1.3582621082621082, + "grad_norm": 0.7629066109657288, + "learning_rate": 0.0001484567472919179, + "loss": 1.0372, + "step": 7629 + }, + { + "epoch": 1.358440170940171, + "grad_norm": 0.7848913669586182, + "learning_rate": 0.00014844450247405435, + "loss": 0.9437, + "step": 7630 + }, + { + "epoch": 1.3586182336182335, + "grad_norm": 0.715949535369873, + "learning_rate": 0.00014843225670702143, + "loss": 1.1949, + "step": 7631 + }, + { + "epoch": 1.3587962962962963, + "grad_norm": 0.6498245596885681, + "learning_rate": 0.00014842000999105905, + "loss": 0.8845, + "step": 7632 + }, + { + "epoch": 1.358974358974359, + "grad_norm": 0.7251074910163879, + "learning_rate": 0.00014840776232640716, + "loss": 1.093, + "step": 7633 + }, + { + "epoch": 1.3591524216524218, + "grad_norm": 0.6223580837249756, + "learning_rate": 0.0001483955137133057, + "loss": 1.0344, + "step": 7634 + }, + { + "epoch": 1.3593304843304843, + "grad_norm": 0.6504943370819092, + "learning_rate": 0.00014838326415199472, + "loss": 1.109, + "step": 7635 + }, + { + "epoch": 1.359508547008547, + "grad_norm": 0.5912374258041382, + "learning_rate": 0.00014837101364271416, + "loss": 1.0756, + "step": 7636 + }, + { + "epoch": 1.3596866096866096, + "grad_norm": 0.6116467714309692, + "learning_rate": 0.00014835876218570408, + "loss": 0.7871, + "step": 7637 + }, + { + "epoch": 1.3598646723646723, + "grad_norm": 0.7013412117958069, + "learning_rate": 0.0001483465097812045, + "loss": 1.0003, + "step": 7638 + }, + { + "epoch": 1.360042735042735, + "grad_norm": 0.5930750370025635, + "learning_rate": 0.00014833425642945552, + "loss": 0.9926, + "step": 7639 + }, + { + "epoch": 1.3602207977207978, + "grad_norm": 0.732955276966095, + "learning_rate": 0.00014832200213069717, + "loss": 1.2801, + "step": 7640 + }, + { + "epoch": 1.3603988603988604, + "grad_norm": 0.6836149096488953, + "learning_rate": 0.00014830974688516958, + "loss": 0.9292, + "step": 7641 + }, + { + "epoch": 1.3605769230769231, + "grad_norm": 0.6531919836997986, + "learning_rate": 0.00014829749069311283, + "loss": 0.9551, + "step": 7642 + }, + { + "epoch": 1.3607549857549857, + "grad_norm": 0.719093382358551, + "learning_rate": 0.0001482852335547671, + "loss": 0.8588, + "step": 7643 + }, + { + "epoch": 1.3609330484330484, + "grad_norm": 0.6144105792045593, + "learning_rate": 0.00014827297547037252, + "loss": 0.9033, + "step": 7644 + }, + { + "epoch": 1.3611111111111112, + "grad_norm": 0.789241373538971, + "learning_rate": 0.00014826071644016926, + "loss": 1.1916, + "step": 7645 + }, + { + "epoch": 1.361289173789174, + "grad_norm": 0.6137418746948242, + "learning_rate": 0.0001482484564643975, + "loss": 0.9648, + "step": 7646 + }, + { + "epoch": 1.3614672364672364, + "grad_norm": 0.6789261698722839, + "learning_rate": 0.00014823619554329745, + "loss": 0.829, + "step": 7647 + }, + { + "epoch": 1.3616452991452992, + "grad_norm": 0.6508790254592896, + "learning_rate": 0.0001482239336771094, + "loss": 0.942, + "step": 7648 + }, + { + "epoch": 1.3618233618233617, + "grad_norm": 0.6725571751594543, + "learning_rate": 0.00014821167086607353, + "loss": 0.8884, + "step": 7649 + }, + { + "epoch": 1.3620014245014245, + "grad_norm": 0.6252003908157349, + "learning_rate": 0.00014819940711043012, + "loss": 0.9778, + "step": 7650 + }, + { + "epoch": 1.3621794871794872, + "grad_norm": 0.6950626969337463, + "learning_rate": 0.00014818714241041943, + "loss": 1.2104, + "step": 7651 + }, + { + "epoch": 1.36235754985755, + "grad_norm": 0.6527379155158997, + "learning_rate": 0.0001481748767662818, + "loss": 0.7845, + "step": 7652 + }, + { + "epoch": 1.3625356125356125, + "grad_norm": 0.7438235282897949, + "learning_rate": 0.00014816261017825755, + "loss": 0.9513, + "step": 7653 + }, + { + "epoch": 1.3627136752136753, + "grad_norm": 0.6412696838378906, + "learning_rate": 0.000148150342646587, + "loss": 0.8478, + "step": 7654 + }, + { + "epoch": 1.3628917378917378, + "grad_norm": 0.658481240272522, + "learning_rate": 0.00014813807417151046, + "loss": 0.6816, + "step": 7655 + }, + { + "epoch": 1.3630698005698005, + "grad_norm": 0.6170126795768738, + "learning_rate": 0.0001481258047532684, + "loss": 0.8862, + "step": 7656 + }, + { + "epoch": 1.3632478632478633, + "grad_norm": 0.7049173712730408, + "learning_rate": 0.0001481135343921012, + "loss": 1.0027, + "step": 7657 + }, + { + "epoch": 1.363425925925926, + "grad_norm": 0.7780741453170776, + "learning_rate": 0.0001481012630882492, + "loss": 1.0183, + "step": 7658 + }, + { + "epoch": 1.3636039886039886, + "grad_norm": 0.6658362746238708, + "learning_rate": 0.00014808899084195286, + "loss": 0.878, + "step": 7659 + }, + { + "epoch": 1.3637820512820513, + "grad_norm": 0.7192076444625854, + "learning_rate": 0.00014807671765345267, + "loss": 1.2269, + "step": 7660 + }, + { + "epoch": 1.3639601139601139, + "grad_norm": 0.7038660049438477, + "learning_rate": 0.00014806444352298903, + "loss": 0.889, + "step": 7661 + }, + { + "epoch": 1.3641381766381766, + "grad_norm": 0.622803270816803, + "learning_rate": 0.00014805216845080249, + "loss": 0.9623, + "step": 7662 + }, + { + "epoch": 1.3643162393162394, + "grad_norm": 0.9157076478004456, + "learning_rate": 0.00014803989243713353, + "loss": 1.106, + "step": 7663 + }, + { + "epoch": 1.364494301994302, + "grad_norm": 0.6369999647140503, + "learning_rate": 0.00014802761548222268, + "loss": 0.9755, + "step": 7664 + }, + { + "epoch": 1.3646723646723646, + "grad_norm": 0.8318394422531128, + "learning_rate": 0.00014801533758631045, + "loss": 1.1786, + "step": 7665 + }, + { + "epoch": 1.3648504273504274, + "grad_norm": 0.7065796852111816, + "learning_rate": 0.00014800305874963744, + "loss": 1.2066, + "step": 7666 + }, + { + "epoch": 1.36502849002849, + "grad_norm": 0.6570265293121338, + "learning_rate": 0.0001479907789724442, + "loss": 1.0084, + "step": 7667 + }, + { + "epoch": 1.3652065527065527, + "grad_norm": 0.637321949005127, + "learning_rate": 0.00014797849825497135, + "loss": 0.9075, + "step": 7668 + }, + { + "epoch": 1.3653846153846154, + "grad_norm": 0.7656470537185669, + "learning_rate": 0.00014796621659745948, + "loss": 1.1497, + "step": 7669 + }, + { + "epoch": 1.3655626780626782, + "grad_norm": 0.6798120737075806, + "learning_rate": 0.0001479539340001493, + "loss": 0.8154, + "step": 7670 + }, + { + "epoch": 1.3657407407407407, + "grad_norm": 0.7004328966140747, + "learning_rate": 0.0001479416504632813, + "loss": 1.0513, + "step": 7671 + }, + { + "epoch": 1.3659188034188035, + "grad_norm": 0.6551713943481445, + "learning_rate": 0.0001479293659870963, + "loss": 0.8735, + "step": 7672 + }, + { + "epoch": 1.366096866096866, + "grad_norm": 0.7685719132423401, + "learning_rate": 0.00014791708057183494, + "loss": 1.111, + "step": 7673 + }, + { + "epoch": 1.3662749287749287, + "grad_norm": 0.673624575138092, + "learning_rate": 0.0001479047942177379, + "loss": 0.9418, + "step": 7674 + }, + { + "epoch": 1.3664529914529915, + "grad_norm": 0.6281047463417053, + "learning_rate": 0.00014789250692504597, + "loss": 1.0938, + "step": 7675 + }, + { + "epoch": 1.3666310541310542, + "grad_norm": 0.5846312642097473, + "learning_rate": 0.0001478802186939998, + "loss": 0.6352, + "step": 7676 + }, + { + "epoch": 1.3668091168091168, + "grad_norm": 0.7037251591682434, + "learning_rate": 0.00014786792952484025, + "loss": 1.1775, + "step": 7677 + }, + { + "epoch": 1.3669871794871795, + "grad_norm": 0.69822758436203, + "learning_rate": 0.00014785563941780808, + "loss": 1.0877, + "step": 7678 + }, + { + "epoch": 1.367165242165242, + "grad_norm": 0.7229313254356384, + "learning_rate": 0.000147843348373144, + "loss": 1.0305, + "step": 7679 + }, + { + "epoch": 1.3673433048433048, + "grad_norm": 0.665771484375, + "learning_rate": 0.00014783105639108897, + "loss": 0.9056, + "step": 7680 + }, + { + "epoch": 1.3675213675213675, + "grad_norm": 0.6418357491493225, + "learning_rate": 0.00014781876347188367, + "loss": 0.9374, + "step": 7681 + }, + { + "epoch": 1.3676994301994303, + "grad_norm": 0.7255483269691467, + "learning_rate": 0.0001478064696157691, + "loss": 0.8533, + "step": 7682 + }, + { + "epoch": 1.3678774928774928, + "grad_norm": 0.668064534664154, + "learning_rate": 0.00014779417482298603, + "loss": 0.9002, + "step": 7683 + }, + { + "epoch": 1.3680555555555556, + "grad_norm": 0.6797603368759155, + "learning_rate": 0.0001477818790937754, + "loss": 0.9733, + "step": 7684 + }, + { + "epoch": 1.368233618233618, + "grad_norm": 0.6905350685119629, + "learning_rate": 0.0001477695824283781, + "loss": 0.7985, + "step": 7685 + }, + { + "epoch": 1.3684116809116809, + "grad_norm": 0.6846137046813965, + "learning_rate": 0.00014775728482703507, + "loss": 0.9154, + "step": 7686 + }, + { + "epoch": 1.3685897435897436, + "grad_norm": 0.6686832904815674, + "learning_rate": 0.00014774498628998726, + "loss": 0.926, + "step": 7687 + }, + { + "epoch": 1.3687678062678064, + "grad_norm": 0.7050234079360962, + "learning_rate": 0.00014773268681747561, + "loss": 0.9386, + "step": 7688 + }, + { + "epoch": 1.368945868945869, + "grad_norm": 0.7048354744911194, + "learning_rate": 0.00014772038640974112, + "loss": 1.1483, + "step": 7689 + }, + { + "epoch": 1.3691239316239316, + "grad_norm": 0.698192298412323, + "learning_rate": 0.0001477080850670248, + "loss": 1.1452, + "step": 7690 + }, + { + "epoch": 1.3693019943019942, + "grad_norm": 0.6838962435722351, + "learning_rate": 0.00014769578278956766, + "loss": 0.9789, + "step": 7691 + }, + { + "epoch": 1.369480056980057, + "grad_norm": 0.6636955142021179, + "learning_rate": 0.00014768347957761074, + "loss": 0.931, + "step": 7692 + }, + { + "epoch": 1.3696581196581197, + "grad_norm": 0.706030547618866, + "learning_rate": 0.0001476711754313951, + "loss": 1.1096, + "step": 7693 + }, + { + "epoch": 1.3698361823361824, + "grad_norm": 0.6771288514137268, + "learning_rate": 0.00014765887035116178, + "loss": 0.9641, + "step": 7694 + }, + { + "epoch": 1.370014245014245, + "grad_norm": 0.6805008053779602, + "learning_rate": 0.00014764656433715188, + "loss": 0.8724, + "step": 7695 + }, + { + "epoch": 1.3701923076923077, + "grad_norm": 0.6599233746528625, + "learning_rate": 0.00014763425738960657, + "loss": 0.8477, + "step": 7696 + }, + { + "epoch": 1.3703703703703702, + "grad_norm": 0.7036116123199463, + "learning_rate": 0.0001476219495087669, + "loss": 1.0991, + "step": 7697 + }, + { + "epoch": 1.370548433048433, + "grad_norm": 0.6677989363670349, + "learning_rate": 0.0001476096406948741, + "loss": 1.2397, + "step": 7698 + }, + { + "epoch": 1.3707264957264957, + "grad_norm": 0.5652269721031189, + "learning_rate": 0.00014759733094816928, + "loss": 0.9302, + "step": 7699 + }, + { + "epoch": 1.3709045584045585, + "grad_norm": 0.6670156121253967, + "learning_rate": 0.00014758502026889362, + "loss": 0.8362, + "step": 7700 + }, + { + "epoch": 1.371082621082621, + "grad_norm": 0.6705406904220581, + "learning_rate": 0.00014757270865728832, + "loss": 0.876, + "step": 7701 + }, + { + "epoch": 1.3712606837606838, + "grad_norm": 0.6020053625106812, + "learning_rate": 0.00014756039611359465, + "loss": 0.9182, + "step": 7702 + }, + { + "epoch": 1.3714387464387463, + "grad_norm": 0.6370134949684143, + "learning_rate": 0.0001475480826380538, + "loss": 1.1063, + "step": 7703 + }, + { + "epoch": 1.371616809116809, + "grad_norm": 0.6906460523605347, + "learning_rate": 0.00014753576823090705, + "loss": 0.988, + "step": 7704 + }, + { + "epoch": 1.3717948717948718, + "grad_norm": 0.6047569513320923, + "learning_rate": 0.00014752345289239567, + "loss": 1.15, + "step": 7705 + }, + { + "epoch": 1.3719729344729346, + "grad_norm": 0.7019868493080139, + "learning_rate": 0.00014751113662276095, + "loss": 1.1185, + "step": 7706 + }, + { + "epoch": 1.372150997150997, + "grad_norm": 0.6534035801887512, + "learning_rate": 0.00014749881942224417, + "loss": 0.9006, + "step": 7707 + }, + { + "epoch": 1.3723290598290598, + "grad_norm": 0.6111651659011841, + "learning_rate": 0.00014748650129108674, + "loss": 0.935, + "step": 7708 + }, + { + "epoch": 1.3725071225071226, + "grad_norm": 0.6678512096405029, + "learning_rate": 0.00014747418222952995, + "loss": 0.8771, + "step": 7709 + }, + { + "epoch": 1.3726851851851851, + "grad_norm": 0.607829749584198, + "learning_rate": 0.00014746186223781518, + "loss": 1.0509, + "step": 7710 + }, + { + "epoch": 1.3728632478632479, + "grad_norm": 0.7274412512779236, + "learning_rate": 0.00014744954131618382, + "loss": 0.9545, + "step": 7711 + }, + { + "epoch": 1.3730413105413106, + "grad_norm": 0.640333354473114, + "learning_rate": 0.00014743721946487723, + "loss": 1.018, + "step": 7712 + }, + { + "epoch": 1.3732193732193732, + "grad_norm": 0.6772079467773438, + "learning_rate": 0.0001474248966841369, + "loss": 1.0983, + "step": 7713 + }, + { + "epoch": 1.373397435897436, + "grad_norm": 0.49630534648895264, + "learning_rate": 0.00014741257297420422, + "loss": 0.5238, + "step": 7714 + }, + { + "epoch": 1.3735754985754987, + "grad_norm": 0.6316596269607544, + "learning_rate": 0.00014740024833532068, + "loss": 1.1342, + "step": 7715 + }, + { + "epoch": 1.3737535612535612, + "grad_norm": 0.5928404331207275, + "learning_rate": 0.00014738792276772775, + "loss": 0.7987, + "step": 7716 + }, + { + "epoch": 1.373931623931624, + "grad_norm": 0.6773418188095093, + "learning_rate": 0.00014737559627166688, + "loss": 0.934, + "step": 7717 + }, + { + "epoch": 1.3741096866096867, + "grad_norm": 0.7895028591156006, + "learning_rate": 0.00014736326884737963, + "loss": 0.984, + "step": 7718 + }, + { + "epoch": 1.3742877492877492, + "grad_norm": 0.7074753046035767, + "learning_rate": 0.00014735094049510752, + "loss": 1.0093, + "step": 7719 + }, + { + "epoch": 1.374465811965812, + "grad_norm": 0.5389847159385681, + "learning_rate": 0.00014733861121509208, + "loss": 0.8138, + "step": 7720 + }, + { + "epoch": 1.3746438746438747, + "grad_norm": 0.6138495206832886, + "learning_rate": 0.00014732628100757493, + "loss": 0.9282, + "step": 7721 + }, + { + "epoch": 1.3748219373219372, + "grad_norm": 0.7609560489654541, + "learning_rate": 0.00014731394987279757, + "loss": 0.9859, + "step": 7722 + }, + { + "epoch": 1.375, + "grad_norm": 0.6806198954582214, + "learning_rate": 0.00014730161781100165, + "loss": 0.8932, + "step": 7723 + }, + { + "epoch": 1.3751780626780628, + "grad_norm": 0.7229103446006775, + "learning_rate": 0.0001472892848224288, + "loss": 0.956, + "step": 7724 + }, + { + "epoch": 1.3753561253561253, + "grad_norm": 0.6157994866371155, + "learning_rate": 0.00014727695090732066, + "loss": 1.0285, + "step": 7725 + }, + { + "epoch": 1.375534188034188, + "grad_norm": 0.5885980129241943, + "learning_rate": 0.00014726461606591885, + "loss": 0.9174, + "step": 7726 + }, + { + "epoch": 1.3757122507122508, + "grad_norm": 0.6655769944190979, + "learning_rate": 0.0001472522802984651, + "loss": 0.9059, + "step": 7727 + }, + { + "epoch": 1.3758903133903133, + "grad_norm": 0.7075541019439697, + "learning_rate": 0.00014723994360520105, + "loss": 1.0055, + "step": 7728 + }, + { + "epoch": 1.376068376068376, + "grad_norm": 0.6947159171104431, + "learning_rate": 0.00014722760598636847, + "loss": 0.9782, + "step": 7729 + }, + { + "epoch": 1.3762464387464388, + "grad_norm": 0.6629964709281921, + "learning_rate": 0.00014721526744220905, + "loss": 0.9427, + "step": 7730 + }, + { + "epoch": 1.3764245014245013, + "grad_norm": 0.7385284304618835, + "learning_rate": 0.00014720292797296453, + "loss": 0.9953, + "step": 7731 + }, + { + "epoch": 1.376602564102564, + "grad_norm": 0.6123563051223755, + "learning_rate": 0.0001471905875788767, + "loss": 1.0103, + "step": 7732 + }, + { + "epoch": 1.3767806267806268, + "grad_norm": 0.6457047462463379, + "learning_rate": 0.00014717824626018732, + "loss": 0.9779, + "step": 7733 + }, + { + "epoch": 1.3769586894586894, + "grad_norm": 0.6196442246437073, + "learning_rate": 0.00014716590401713824, + "loss": 0.8747, + "step": 7734 + }, + { + "epoch": 1.3771367521367521, + "grad_norm": 0.7932298183441162, + "learning_rate": 0.00014715356084997122, + "loss": 1.1617, + "step": 7735 + }, + { + "epoch": 1.3773148148148149, + "grad_norm": 0.787304699420929, + "learning_rate": 0.00014714121675892815, + "loss": 1.1383, + "step": 7736 + }, + { + "epoch": 1.3774928774928774, + "grad_norm": 0.672795295715332, + "learning_rate": 0.00014712887174425085, + "loss": 1.2563, + "step": 7737 + }, + { + "epoch": 1.3776709401709402, + "grad_norm": 0.6505744457244873, + "learning_rate": 0.00014711652580618123, + "loss": 0.9194, + "step": 7738 + }, + { + "epoch": 1.377849002849003, + "grad_norm": 0.8141193985939026, + "learning_rate": 0.00014710417894496115, + "loss": 1.1428, + "step": 7739 + }, + { + "epoch": 1.3780270655270654, + "grad_norm": 0.6269707679748535, + "learning_rate": 0.00014709183116083253, + "loss": 0.7164, + "step": 7740 + }, + { + "epoch": 1.3782051282051282, + "grad_norm": 0.6737076640129089, + "learning_rate": 0.0001470794824540373, + "loss": 0.9965, + "step": 7741 + }, + { + "epoch": 1.378383190883191, + "grad_norm": 0.6451728343963623, + "learning_rate": 0.0001470671328248174, + "loss": 1.0539, + "step": 7742 + }, + { + "epoch": 1.3785612535612537, + "grad_norm": 0.6480295062065125, + "learning_rate": 0.00014705478227341486, + "loss": 0.9118, + "step": 7743 + }, + { + "epoch": 1.3787393162393162, + "grad_norm": 0.7429090738296509, + "learning_rate": 0.00014704243080007154, + "loss": 1.0031, + "step": 7744 + }, + { + "epoch": 1.378917378917379, + "grad_norm": 0.5601376891136169, + "learning_rate": 0.00014703007840502955, + "loss": 0.849, + "step": 7745 + }, + { + "epoch": 1.3790954415954415, + "grad_norm": 0.7067657113075256, + "learning_rate": 0.00014701772508853088, + "loss": 1.3067, + "step": 7746 + }, + { + "epoch": 1.3792735042735043, + "grad_norm": 0.7016390562057495, + "learning_rate": 0.00014700537085081755, + "loss": 1.0236, + "step": 7747 + }, + { + "epoch": 1.379451566951567, + "grad_norm": 0.6505000591278076, + "learning_rate": 0.0001469930156921316, + "loss": 1.0121, + "step": 7748 + }, + { + "epoch": 1.3796296296296298, + "grad_norm": 0.8515380620956421, + "learning_rate": 0.00014698065961271512, + "loss": 1.0413, + "step": 7749 + }, + { + "epoch": 1.3798076923076923, + "grad_norm": 0.6322008371353149, + "learning_rate": 0.00014696830261281025, + "loss": 0.8306, + "step": 7750 + }, + { + "epoch": 1.379985754985755, + "grad_norm": 0.7090431451797485, + "learning_rate": 0.00014695594469265902, + "loss": 1.1829, + "step": 7751 + }, + { + "epoch": 1.3801638176638176, + "grad_norm": 0.5913167595863342, + "learning_rate": 0.00014694358585250363, + "loss": 0.9769, + "step": 7752 + }, + { + "epoch": 1.3803418803418803, + "grad_norm": 0.7345432639122009, + "learning_rate": 0.00014693122609258616, + "loss": 0.9928, + "step": 7753 + }, + { + "epoch": 1.380519943019943, + "grad_norm": 0.6158214211463928, + "learning_rate": 0.00014691886541314884, + "loss": 1.1166, + "step": 7754 + }, + { + "epoch": 1.3806980056980058, + "grad_norm": 0.6874041557312012, + "learning_rate": 0.0001469065038144338, + "loss": 1.0808, + "step": 7755 + }, + { + "epoch": 1.3808760683760684, + "grad_norm": 0.8135195970535278, + "learning_rate": 0.00014689414129668326, + "loss": 0.9482, + "step": 7756 + }, + { + "epoch": 1.381054131054131, + "grad_norm": 0.6389174461364746, + "learning_rate": 0.00014688177786013944, + "loss": 1.039, + "step": 7757 + }, + { + "epoch": 1.3812321937321936, + "grad_norm": 0.6953016519546509, + "learning_rate": 0.00014686941350504454, + "loss": 0.9426, + "step": 7758 + }, + { + "epoch": 1.3814102564102564, + "grad_norm": 0.8171859383583069, + "learning_rate": 0.00014685704823164087, + "loss": 1.0393, + "step": 7759 + }, + { + "epoch": 1.3815883190883191, + "grad_norm": 0.6968414783477783, + "learning_rate": 0.0001468446820401707, + "loss": 1.1167, + "step": 7760 + }, + { + "epoch": 1.381766381766382, + "grad_norm": 0.6916623711585999, + "learning_rate": 0.00014683231493087628, + "loss": 1.1886, + "step": 7761 + }, + { + "epoch": 1.3819444444444444, + "grad_norm": 0.7351683378219604, + "learning_rate": 0.00014681994690399992, + "loss": 0.9893, + "step": 7762 + }, + { + "epoch": 1.3821225071225072, + "grad_norm": 0.6617491245269775, + "learning_rate": 0.00014680757795978395, + "loss": 1.0505, + "step": 7763 + }, + { + "epoch": 1.3823005698005697, + "grad_norm": 0.6627485156059265, + "learning_rate": 0.00014679520809847074, + "loss": 0.9878, + "step": 7764 + }, + { + "epoch": 1.3824786324786325, + "grad_norm": 0.704636812210083, + "learning_rate": 0.00014678283732030264, + "loss": 0.8332, + "step": 7765 + }, + { + "epoch": 1.3826566951566952, + "grad_norm": 0.698853075504303, + "learning_rate": 0.00014677046562552203, + "loss": 1.0926, + "step": 7766 + }, + { + "epoch": 1.382834757834758, + "grad_norm": 0.6695869565010071, + "learning_rate": 0.0001467580930143713, + "loss": 1.0626, + "step": 7767 + }, + { + "epoch": 1.3830128205128205, + "grad_norm": 0.672173023223877, + "learning_rate": 0.00014674571948709286, + "loss": 0.8842, + "step": 7768 + }, + { + "epoch": 1.3831908831908832, + "grad_norm": 0.6735473871231079, + "learning_rate": 0.00014673334504392916, + "loss": 0.9382, + "step": 7769 + }, + { + "epoch": 1.3833689458689458, + "grad_norm": 0.6864013075828552, + "learning_rate": 0.00014672096968512265, + "loss": 1.1369, + "step": 7770 + }, + { + "epoch": 1.3835470085470085, + "grad_norm": 0.7154954075813293, + "learning_rate": 0.0001467085934109158, + "loss": 1.1447, + "step": 7771 + }, + { + "epoch": 1.3837250712250713, + "grad_norm": 0.5934487581253052, + "learning_rate": 0.0001466962162215511, + "loss": 0.8923, + "step": 7772 + }, + { + "epoch": 1.383903133903134, + "grad_norm": 0.8116832971572876, + "learning_rate": 0.00014668383811727097, + "loss": 1.0997, + "step": 7773 + }, + { + "epoch": 1.3840811965811965, + "grad_norm": 0.8661674857139587, + "learning_rate": 0.00014667145909831808, + "loss": 1.0112, + "step": 7774 + }, + { + "epoch": 1.3842592592592593, + "grad_norm": 0.5173856616020203, + "learning_rate": 0.00014665907916493488, + "loss": 0.6571, + "step": 7775 + }, + { + "epoch": 1.3844373219373218, + "grad_norm": 0.6165067553520203, + "learning_rate": 0.00014664669831736395, + "loss": 1.0992, + "step": 7776 + }, + { + "epoch": 1.3846153846153846, + "grad_norm": 0.6564429998397827, + "learning_rate": 0.00014663431655584787, + "loss": 0.9103, + "step": 7777 + }, + { + "epoch": 1.3847934472934473, + "grad_norm": 0.7162124514579773, + "learning_rate": 0.00014662193388062923, + "loss": 1.0645, + "step": 7778 + }, + { + "epoch": 1.38497150997151, + "grad_norm": 0.6391215920448303, + "learning_rate": 0.00014660955029195064, + "loss": 0.902, + "step": 7779 + }, + { + "epoch": 1.3851495726495726, + "grad_norm": 0.6876635551452637, + "learning_rate": 0.00014659716579005475, + "loss": 1.0924, + "step": 7780 + }, + { + "epoch": 1.3853276353276354, + "grad_norm": 0.7254653573036194, + "learning_rate": 0.00014658478037518418, + "loss": 1.0135, + "step": 7781 + }, + { + "epoch": 1.385505698005698, + "grad_norm": 0.6900535225868225, + "learning_rate": 0.00014657239404758162, + "loss": 0.983, + "step": 7782 + }, + { + "epoch": 1.3856837606837606, + "grad_norm": 0.7477042078971863, + "learning_rate": 0.00014656000680748975, + "loss": 1.0707, + "step": 7783 + }, + { + "epoch": 1.3858618233618234, + "grad_norm": 0.5756927132606506, + "learning_rate": 0.00014654761865515124, + "loss": 0.8881, + "step": 7784 + }, + { + "epoch": 1.3860398860398861, + "grad_norm": 0.6736083626747131, + "learning_rate": 0.00014653522959080884, + "loss": 1.0193, + "step": 7785 + }, + { + "epoch": 1.3862179487179487, + "grad_norm": 0.616179883480072, + "learning_rate": 0.0001465228396147053, + "loss": 0.8676, + "step": 7786 + }, + { + "epoch": 1.3863960113960114, + "grad_norm": 0.7956456542015076, + "learning_rate": 0.00014651044872708338, + "loss": 0.9787, + "step": 7787 + }, + { + "epoch": 1.386574074074074, + "grad_norm": 0.6613463163375854, + "learning_rate": 0.00014649805692818578, + "loss": 1.0032, + "step": 7788 + }, + { + "epoch": 1.3867521367521367, + "grad_norm": 0.6215800642967224, + "learning_rate": 0.0001464856642182554, + "loss": 1.0123, + "step": 7789 + }, + { + "epoch": 1.3869301994301995, + "grad_norm": 0.6701171398162842, + "learning_rate": 0.00014647327059753496, + "loss": 0.9108, + "step": 7790 + }, + { + "epoch": 1.3871082621082622, + "grad_norm": 0.6213465929031372, + "learning_rate": 0.00014646087606626736, + "loss": 0.9313, + "step": 7791 + }, + { + "epoch": 1.3872863247863247, + "grad_norm": 0.7535304427146912, + "learning_rate": 0.00014644848062469535, + "loss": 1.0813, + "step": 7792 + }, + { + "epoch": 1.3874643874643875, + "grad_norm": 0.6778230667114258, + "learning_rate": 0.0001464360842730619, + "loss": 1.0405, + "step": 7793 + }, + { + "epoch": 1.38764245014245, + "grad_norm": 0.7816025614738464, + "learning_rate": 0.0001464236870116098, + "loss": 0.9228, + "step": 7794 + }, + { + "epoch": 1.3878205128205128, + "grad_norm": 0.6815229058265686, + "learning_rate": 0.00014641128884058203, + "loss": 0.9607, + "step": 7795 + }, + { + "epoch": 1.3879985754985755, + "grad_norm": 0.7027714848518372, + "learning_rate": 0.00014639888976022145, + "loss": 0.9379, + "step": 7796 + }, + { + "epoch": 1.3881766381766383, + "grad_norm": 0.7636353373527527, + "learning_rate": 0.00014638648977077104, + "loss": 1.1186, + "step": 7797 + }, + { + "epoch": 1.3883547008547008, + "grad_norm": 0.6732974052429199, + "learning_rate": 0.00014637408887247365, + "loss": 1.1378, + "step": 7798 + }, + { + "epoch": 1.3885327635327636, + "grad_norm": 0.7539397478103638, + "learning_rate": 0.0001463616870655724, + "loss": 0.999, + "step": 7799 + }, + { + "epoch": 1.388710826210826, + "grad_norm": 0.6872972846031189, + "learning_rate": 0.00014634928435031013, + "loss": 0.9564, + "step": 7800 + }, + { + "epoch": 1.3888888888888888, + "grad_norm": 0.6823115348815918, + "learning_rate": 0.00014633688072693, + "loss": 0.9745, + "step": 7801 + }, + { + "epoch": 1.3890669515669516, + "grad_norm": 0.6462571620941162, + "learning_rate": 0.00014632447619567488, + "loss": 0.8314, + "step": 7802 + }, + { + "epoch": 1.3892450142450143, + "grad_norm": 0.7245402932167053, + "learning_rate": 0.0001463120707567879, + "loss": 0.8291, + "step": 7803 + }, + { + "epoch": 1.3894230769230769, + "grad_norm": 0.697179913520813, + "learning_rate": 0.00014629966441051208, + "loss": 1.017, + "step": 7804 + }, + { + "epoch": 1.3896011396011396, + "grad_norm": 0.6304250359535217, + "learning_rate": 0.00014628725715709053, + "loss": 0.9262, + "step": 7805 + }, + { + "epoch": 1.3897792022792022, + "grad_norm": 0.5780240297317505, + "learning_rate": 0.00014627484899676634, + "loss": 0.6596, + "step": 7806 + }, + { + "epoch": 1.389957264957265, + "grad_norm": 0.8030684590339661, + "learning_rate": 0.0001462624399297826, + "loss": 0.9977, + "step": 7807 + }, + { + "epoch": 1.3901353276353277, + "grad_norm": 0.7999774813652039, + "learning_rate": 0.00014625002995638246, + "loss": 1.1036, + "step": 7808 + }, + { + "epoch": 1.3903133903133904, + "grad_norm": 0.7054862976074219, + "learning_rate": 0.00014623761907680904, + "loss": 1.1435, + "step": 7809 + }, + { + "epoch": 1.390491452991453, + "grad_norm": 0.6660647392272949, + "learning_rate": 0.00014622520729130556, + "loss": 0.703, + "step": 7810 + }, + { + "epoch": 1.3906695156695157, + "grad_norm": 0.6339690089225769, + "learning_rate": 0.00014621279460011515, + "loss": 1.0451, + "step": 7811 + }, + { + "epoch": 1.3908475783475782, + "grad_norm": 0.8568736910820007, + "learning_rate": 0.00014620038100348102, + "loss": 1.009, + "step": 7812 + }, + { + "epoch": 1.391025641025641, + "grad_norm": 0.7126797437667847, + "learning_rate": 0.00014618796650164642, + "loss": 0.9592, + "step": 7813 + }, + { + "epoch": 1.3912037037037037, + "grad_norm": 0.6768994331359863, + "learning_rate": 0.00014617555109485453, + "loss": 1.09, + "step": 7814 + }, + { + "epoch": 1.3913817663817665, + "grad_norm": 0.7609471678733826, + "learning_rate": 0.00014616313478334864, + "loss": 0.9781, + "step": 7815 + }, + { + "epoch": 1.391559829059829, + "grad_norm": 0.7107006907463074, + "learning_rate": 0.00014615071756737203, + "loss": 0.9769, + "step": 7816 + }, + { + "epoch": 1.3917378917378918, + "grad_norm": 0.6324763894081116, + "learning_rate": 0.00014613829944716802, + "loss": 1.089, + "step": 7817 + }, + { + "epoch": 1.3919159544159543, + "grad_norm": 0.6617186069488525, + "learning_rate": 0.00014612588042297984, + "loss": 1.0466, + "step": 7818 + }, + { + "epoch": 1.392094017094017, + "grad_norm": 0.7881436944007874, + "learning_rate": 0.00014611346049505083, + "loss": 1.003, + "step": 7819 + }, + { + "epoch": 1.3922720797720798, + "grad_norm": 0.7391049861907959, + "learning_rate": 0.00014610103966362437, + "loss": 1.0531, + "step": 7820 + }, + { + "epoch": 1.3924501424501425, + "grad_norm": 0.6299472451210022, + "learning_rate": 0.00014608861792894383, + "loss": 0.8433, + "step": 7821 + }, + { + "epoch": 1.392628205128205, + "grad_norm": 0.6053452491760254, + "learning_rate": 0.00014607619529125255, + "loss": 0.7945, + "step": 7822 + }, + { + "epoch": 1.3928062678062678, + "grad_norm": 0.7160114645957947, + "learning_rate": 0.0001460637717507939, + "loss": 1.1604, + "step": 7823 + }, + { + "epoch": 1.3929843304843303, + "grad_norm": 0.6308854222297668, + "learning_rate": 0.00014605134730781135, + "loss": 1.0918, + "step": 7824 + }, + { + "epoch": 1.393162393162393, + "grad_norm": 0.7187000513076782, + "learning_rate": 0.00014603892196254833, + "loss": 1.0594, + "step": 7825 + }, + { + "epoch": 1.3933404558404558, + "grad_norm": 0.7516581416130066, + "learning_rate": 0.00014602649571524826, + "loss": 0.9222, + "step": 7826 + }, + { + "epoch": 1.3935185185185186, + "grad_norm": 0.6340481638908386, + "learning_rate": 0.00014601406856615463, + "loss": 0.8131, + "step": 7827 + }, + { + "epoch": 1.3936965811965811, + "grad_norm": 0.8161744475364685, + "learning_rate": 0.0001460016405155109, + "loss": 0.8695, + "step": 7828 + }, + { + "epoch": 1.3938746438746439, + "grad_norm": 0.6926971077919006, + "learning_rate": 0.0001459892115635606, + "loss": 0.9548, + "step": 7829 + }, + { + "epoch": 1.3940527065527066, + "grad_norm": 0.6669796109199524, + "learning_rate": 0.0001459767817105472, + "loss": 0.9255, + "step": 7830 + }, + { + "epoch": 1.3942307692307692, + "grad_norm": 0.6626184582710266, + "learning_rate": 0.00014596435095671432, + "loss": 1.1141, + "step": 7831 + }, + { + "epoch": 1.394408831908832, + "grad_norm": 0.6755738854408264, + "learning_rate": 0.00014595191930230546, + "loss": 0.9596, + "step": 7832 + }, + { + "epoch": 1.3945868945868947, + "grad_norm": 0.6034863591194153, + "learning_rate": 0.00014593948674756417, + "loss": 0.8088, + "step": 7833 + }, + { + "epoch": 1.3947649572649572, + "grad_norm": 0.5638226866722107, + "learning_rate": 0.00014592705329273406, + "loss": 0.5828, + "step": 7834 + }, + { + "epoch": 1.39494301994302, + "grad_norm": 0.6902222633361816, + "learning_rate": 0.0001459146189380588, + "loss": 0.7954, + "step": 7835 + }, + { + "epoch": 1.3951210826210827, + "grad_norm": 0.7579947710037231, + "learning_rate": 0.0001459021836837819, + "loss": 1.1301, + "step": 7836 + }, + { + "epoch": 1.3952991452991452, + "grad_norm": 0.6894911527633667, + "learning_rate": 0.00014588974753014712, + "loss": 1.082, + "step": 7837 + }, + { + "epoch": 1.395477207977208, + "grad_norm": 0.6330230832099915, + "learning_rate": 0.000145877310477398, + "loss": 0.7614, + "step": 7838 + }, + { + "epoch": 1.3956552706552707, + "grad_norm": 0.6164960265159607, + "learning_rate": 0.00014586487252577832, + "loss": 0.8981, + "step": 7839 + }, + { + "epoch": 1.3958333333333333, + "grad_norm": 0.6575061678886414, + "learning_rate": 0.0001458524336755317, + "loss": 0.9735, + "step": 7840 + }, + { + "epoch": 1.396011396011396, + "grad_norm": 0.687921941280365, + "learning_rate": 0.00014583999392690195, + "loss": 0.9207, + "step": 7841 + }, + { + "epoch": 1.3961894586894588, + "grad_norm": 0.6175212860107422, + "learning_rate": 0.00014582755328013274, + "loss": 1.0444, + "step": 7842 + }, + { + "epoch": 1.3963675213675213, + "grad_norm": 0.6351733207702637, + "learning_rate": 0.00014581511173546781, + "loss": 1.0143, + "step": 7843 + }, + { + "epoch": 1.396545584045584, + "grad_norm": 0.7235051989555359, + "learning_rate": 0.00014580266929315093, + "loss": 0.9108, + "step": 7844 + }, + { + "epoch": 1.3967236467236468, + "grad_norm": 0.6432043313980103, + "learning_rate": 0.00014579022595342586, + "loss": 0.8674, + "step": 7845 + }, + { + "epoch": 1.3969017094017093, + "grad_norm": 0.7775412797927856, + "learning_rate": 0.00014577778171653648, + "loss": 1.0637, + "step": 7846 + }, + { + "epoch": 1.397079772079772, + "grad_norm": 0.6748763918876648, + "learning_rate": 0.00014576533658272655, + "loss": 1.0356, + "step": 7847 + }, + { + "epoch": 1.3972578347578348, + "grad_norm": 0.6940401196479797, + "learning_rate": 0.00014575289055223994, + "loss": 0.9937, + "step": 7848 + }, + { + "epoch": 1.3974358974358974, + "grad_norm": 0.6971304416656494, + "learning_rate": 0.00014574044362532045, + "loss": 0.9753, + "step": 7849 + }, + { + "epoch": 1.39761396011396, + "grad_norm": 0.6576017141342163, + "learning_rate": 0.00014572799580221197, + "loss": 1.1233, + "step": 7850 + }, + { + "epoch": 1.3977920227920229, + "grad_norm": 0.6270702481269836, + "learning_rate": 0.00014571554708315843, + "loss": 0.9771, + "step": 7851 + }, + { + "epoch": 1.3979700854700854, + "grad_norm": 0.6898425817489624, + "learning_rate": 0.00014570309746840372, + "loss": 0.9235, + "step": 7852 + }, + { + "epoch": 1.3981481481481481, + "grad_norm": 0.7017102241516113, + "learning_rate": 0.00014569064695819174, + "loss": 1.1056, + "step": 7853 + }, + { + "epoch": 1.398326210826211, + "grad_norm": 0.6298288702964783, + "learning_rate": 0.00014567819555276647, + "loss": 0.8635, + "step": 7854 + }, + { + "epoch": 1.3985042735042734, + "grad_norm": 0.7173134684562683, + "learning_rate": 0.00014566574325237182, + "loss": 1.0893, + "step": 7855 + }, + { + "epoch": 1.3986823361823362, + "grad_norm": 0.7541036605834961, + "learning_rate": 0.0001456532900572518, + "loss": 1.0996, + "step": 7856 + }, + { + "epoch": 1.398860398860399, + "grad_norm": 0.6204771399497986, + "learning_rate": 0.0001456408359676504, + "loss": 0.7601, + "step": 7857 + }, + { + "epoch": 1.3990384615384617, + "grad_norm": 0.629557192325592, + "learning_rate": 0.00014562838098381163, + "loss": 0.9239, + "step": 7858 + }, + { + "epoch": 1.3992165242165242, + "grad_norm": 0.6878390908241272, + "learning_rate": 0.00014561592510597954, + "loss": 0.9641, + "step": 7859 + }, + { + "epoch": 1.399394586894587, + "grad_norm": 0.7490049004554749, + "learning_rate": 0.00014560346833439813, + "loss": 1.0198, + "step": 7860 + }, + { + "epoch": 1.3995726495726495, + "grad_norm": 0.6337960958480835, + "learning_rate": 0.0001455910106693115, + "loss": 0.8709, + "step": 7861 + }, + { + "epoch": 1.3997507122507122, + "grad_norm": 0.6210524439811707, + "learning_rate": 0.0001455785521109637, + "loss": 1.1049, + "step": 7862 + }, + { + "epoch": 1.399928774928775, + "grad_norm": 0.7894936203956604, + "learning_rate": 0.00014556609265959887, + "loss": 0.8933, + "step": 7863 + }, + { + "epoch": 1.4001068376068377, + "grad_norm": 0.6888098120689392, + "learning_rate": 0.00014555363231546112, + "loss": 0.9738, + "step": 7864 + }, + { + "epoch": 1.4002849002849003, + "grad_norm": 0.608799934387207, + "learning_rate": 0.00014554117107879456, + "loss": 0.9103, + "step": 7865 + }, + { + "epoch": 1.400462962962963, + "grad_norm": 0.7390474081039429, + "learning_rate": 0.00014552870894984335, + "loss": 1.2484, + "step": 7866 + }, + { + "epoch": 1.4006410256410255, + "grad_norm": 0.6513381600379944, + "learning_rate": 0.00014551624592885169, + "loss": 0.8523, + "step": 7867 + }, + { + "epoch": 1.4008190883190883, + "grad_norm": 0.6357464790344238, + "learning_rate": 0.00014550378201606373, + "loss": 0.9594, + "step": 7868 + }, + { + "epoch": 1.400997150997151, + "grad_norm": 0.6893286108970642, + "learning_rate": 0.0001454913172117237, + "loss": 0.9798, + "step": 7869 + }, + { + "epoch": 1.4011752136752138, + "grad_norm": 0.6566550731658936, + "learning_rate": 0.0001454788515160758, + "loss": 1.0532, + "step": 7870 + }, + { + "epoch": 1.4013532763532763, + "grad_norm": 0.6442158222198486, + "learning_rate": 0.00014546638492936425, + "loss": 1.0789, + "step": 7871 + }, + { + "epoch": 1.401531339031339, + "grad_norm": 0.7570971846580505, + "learning_rate": 0.0001454539174518334, + "loss": 0.9806, + "step": 7872 + }, + { + "epoch": 1.4017094017094016, + "grad_norm": 0.6180047392845154, + "learning_rate": 0.0001454414490837274, + "loss": 0.857, + "step": 7873 + }, + { + "epoch": 1.4018874643874644, + "grad_norm": 0.7143170237541199, + "learning_rate": 0.0001454289798252906, + "loss": 0.8815, + "step": 7874 + }, + { + "epoch": 1.4020655270655271, + "grad_norm": 0.6388922929763794, + "learning_rate": 0.00014541650967676736, + "loss": 0.95, + "step": 7875 + }, + { + "epoch": 1.4022435897435899, + "grad_norm": 0.7137351632118225, + "learning_rate": 0.00014540403863840193, + "loss": 0.8973, + "step": 7876 + }, + { + "epoch": 1.4024216524216524, + "grad_norm": 0.656315267086029, + "learning_rate": 0.0001453915667104387, + "loss": 1.149, + "step": 7877 + }, + { + "epoch": 1.4025997150997151, + "grad_norm": 0.7234711647033691, + "learning_rate": 0.000145379093893122, + "loss": 0.9798, + "step": 7878 + }, + { + "epoch": 1.4027777777777777, + "grad_norm": 0.6595289707183838, + "learning_rate": 0.00014536662018669623, + "loss": 1.2704, + "step": 7879 + }, + { + "epoch": 1.4029558404558404, + "grad_norm": 0.6760551333427429, + "learning_rate": 0.00014535414559140576, + "loss": 0.8672, + "step": 7880 + }, + { + "epoch": 1.4031339031339032, + "grad_norm": 0.5916706919670105, + "learning_rate": 0.000145341670107495, + "loss": 0.888, + "step": 7881 + }, + { + "epoch": 1.403311965811966, + "grad_norm": 0.7272133231163025, + "learning_rate": 0.00014532919373520846, + "loss": 1.0466, + "step": 7882 + }, + { + "epoch": 1.4034900284900285, + "grad_norm": 0.8512467741966248, + "learning_rate": 0.00014531671647479048, + "loss": 1.2482, + "step": 7883 + }, + { + "epoch": 1.4036680911680912, + "grad_norm": 0.5536492466926575, + "learning_rate": 0.0001453042383264856, + "loss": 0.7823, + "step": 7884 + }, + { + "epoch": 1.4038461538461537, + "grad_norm": 0.7262215614318848, + "learning_rate": 0.0001452917592905383, + "loss": 0.9713, + "step": 7885 + }, + { + "epoch": 1.4040242165242165, + "grad_norm": 0.7146059274673462, + "learning_rate": 0.00014527927936719304, + "loss": 1.1064, + "step": 7886 + }, + { + "epoch": 1.4042022792022792, + "grad_norm": 0.5915318131446838, + "learning_rate": 0.00014526679855669436, + "loss": 0.8567, + "step": 7887 + }, + { + "epoch": 1.404380341880342, + "grad_norm": 0.6548298001289368, + "learning_rate": 0.00014525431685928682, + "loss": 1.1359, + "step": 7888 + }, + { + "epoch": 1.4045584045584045, + "grad_norm": 0.7482563853263855, + "learning_rate": 0.0001452418342752149, + "loss": 0.9095, + "step": 7889 + }, + { + "epoch": 1.4047364672364673, + "grad_norm": 0.6660130023956299, + "learning_rate": 0.0001452293508047233, + "loss": 1.2343, + "step": 7890 + }, + { + "epoch": 1.4049145299145298, + "grad_norm": 0.7457148432731628, + "learning_rate": 0.00014521686644805644, + "loss": 1.2086, + "step": 7891 + }, + { + "epoch": 1.4050925925925926, + "grad_norm": 0.5957929491996765, + "learning_rate": 0.00014520438120545906, + "loss": 0.9724, + "step": 7892 + }, + { + "epoch": 1.4052706552706553, + "grad_norm": 0.6832270622253418, + "learning_rate": 0.00014519189507717573, + "loss": 0.9903, + "step": 7893 + }, + { + "epoch": 1.405448717948718, + "grad_norm": 0.6202489733695984, + "learning_rate": 0.00014517940806345109, + "loss": 0.962, + "step": 7894 + }, + { + "epoch": 1.4056267806267806, + "grad_norm": 0.6419472694396973, + "learning_rate": 0.0001451669201645298, + "loss": 0.8147, + "step": 7895 + }, + { + "epoch": 1.4058048433048433, + "grad_norm": 0.61143958568573, + "learning_rate": 0.00014515443138065652, + "loss": 0.8674, + "step": 7896 + }, + { + "epoch": 1.4059829059829059, + "grad_norm": 0.7527356743812561, + "learning_rate": 0.00014514194171207597, + "loss": 1.0581, + "step": 7897 + }, + { + "epoch": 1.4061609686609686, + "grad_norm": 0.7195194363594055, + "learning_rate": 0.00014512945115903285, + "loss": 1.0268, + "step": 7898 + }, + { + "epoch": 1.4063390313390314, + "grad_norm": 0.7919661998748779, + "learning_rate": 0.00014511695972177187, + "loss": 1.0259, + "step": 7899 + }, + { + "epoch": 1.4065170940170941, + "grad_norm": 0.6774758696556091, + "learning_rate": 0.00014510446740053783, + "loss": 1.1214, + "step": 7900 + }, + { + "epoch": 1.4066951566951567, + "grad_norm": 0.6102406978607178, + "learning_rate": 0.0001450919741955754, + "loss": 1.1846, + "step": 7901 + }, + { + "epoch": 1.4068732193732194, + "grad_norm": 0.7189443707466125, + "learning_rate": 0.00014507948010712942, + "loss": 0.7758, + "step": 7902 + }, + { + "epoch": 1.407051282051282, + "grad_norm": 0.654153048992157, + "learning_rate": 0.00014506698513544467, + "loss": 0.899, + "step": 7903 + }, + { + "epoch": 1.4072293447293447, + "grad_norm": 0.637934684753418, + "learning_rate": 0.00014505448928076598, + "loss": 0.8301, + "step": 7904 + }, + { + "epoch": 1.4074074074074074, + "grad_norm": 0.7504615783691406, + "learning_rate": 0.00014504199254333812, + "loss": 0.9883, + "step": 7905 + }, + { + "epoch": 1.4075854700854702, + "grad_norm": 0.7902522683143616, + "learning_rate": 0.00014502949492340602, + "loss": 0.9615, + "step": 7906 + }, + { + "epoch": 1.4077635327635327, + "grad_norm": 0.5832732319831848, + "learning_rate": 0.0001450169964212145, + "loss": 0.7136, + "step": 7907 + }, + { + "epoch": 1.4079415954415955, + "grad_norm": 0.6025400757789612, + "learning_rate": 0.00014500449703700846, + "loss": 0.8812, + "step": 7908 + }, + { + "epoch": 1.408119658119658, + "grad_norm": 0.6412411332130432, + "learning_rate": 0.0001449919967710328, + "loss": 0.9346, + "step": 7909 + }, + { + "epoch": 1.4082977207977208, + "grad_norm": 0.7546970844268799, + "learning_rate": 0.00014497949562353242, + "loss": 1.0794, + "step": 7910 + }, + { + "epoch": 1.4084757834757835, + "grad_norm": 0.6175593733787537, + "learning_rate": 0.00014496699359475222, + "loss": 0.8939, + "step": 7911 + }, + { + "epoch": 1.4086538461538463, + "grad_norm": 0.6571716666221619, + "learning_rate": 0.00014495449068493722, + "loss": 1.1003, + "step": 7912 + }, + { + "epoch": 1.4088319088319088, + "grad_norm": 0.7038990259170532, + "learning_rate": 0.00014494198689433236, + "loss": 0.8844, + "step": 7913 + }, + { + "epoch": 1.4090099715099715, + "grad_norm": 0.7007337212562561, + "learning_rate": 0.00014492948222318263, + "loss": 1.2038, + "step": 7914 + }, + { + "epoch": 1.409188034188034, + "grad_norm": 0.7318591475486755, + "learning_rate": 0.00014491697667173302, + "loss": 1.0388, + "step": 7915 + }, + { + "epoch": 1.4093660968660968, + "grad_norm": 0.7010329961776733, + "learning_rate": 0.00014490447024022855, + "loss": 1.1485, + "step": 7916 + }, + { + "epoch": 1.4095441595441596, + "grad_norm": 0.7844831347465515, + "learning_rate": 0.0001448919629289143, + "loss": 1.1417, + "step": 7917 + }, + { + "epoch": 1.4097222222222223, + "grad_norm": 0.6953392624855042, + "learning_rate": 0.00014487945473803525, + "loss": 0.9546, + "step": 7918 + }, + { + "epoch": 1.4099002849002849, + "grad_norm": 0.6307587623596191, + "learning_rate": 0.00014486694566783655, + "loss": 0.9912, + "step": 7919 + }, + { + "epoch": 1.4100783475783476, + "grad_norm": 0.6200215816497803, + "learning_rate": 0.00014485443571856326, + "loss": 1.0998, + "step": 7920 + }, + { + "epoch": 1.4102564102564101, + "grad_norm": 0.7096502184867859, + "learning_rate": 0.00014484192489046043, + "loss": 0.9587, + "step": 7921 + }, + { + "epoch": 1.4104344729344729, + "grad_norm": 0.6965526342391968, + "learning_rate": 0.00014482941318377327, + "loss": 0.8791, + "step": 7922 + }, + { + "epoch": 1.4106125356125356, + "grad_norm": 0.7303466200828552, + "learning_rate": 0.00014481690059874687, + "loss": 1.084, + "step": 7923 + }, + { + "epoch": 1.4107905982905984, + "grad_norm": 0.6144066452980042, + "learning_rate": 0.00014480438713562638, + "loss": 0.9646, + "step": 7924 + }, + { + "epoch": 1.410968660968661, + "grad_norm": 0.645222008228302, + "learning_rate": 0.00014479187279465704, + "loss": 0.728, + "step": 7925 + }, + { + "epoch": 1.4111467236467237, + "grad_norm": 0.6069912314414978, + "learning_rate": 0.000144779357576084, + "loss": 0.842, + "step": 7926 + }, + { + "epoch": 1.4113247863247862, + "grad_norm": 0.6212135553359985, + "learning_rate": 0.00014476684148015243, + "loss": 0.9817, + "step": 7927 + }, + { + "epoch": 1.411502849002849, + "grad_norm": 0.6893343329429626, + "learning_rate": 0.00014475432450710763, + "loss": 1.0265, + "step": 7928 + }, + { + "epoch": 1.4116809116809117, + "grad_norm": 0.6842793822288513, + "learning_rate": 0.00014474180665719478, + "loss": 1.0593, + "step": 7929 + }, + { + "epoch": 1.4118589743589745, + "grad_norm": 0.74690842628479, + "learning_rate": 0.0001447292879306592, + "loss": 0.9096, + "step": 7930 + }, + { + "epoch": 1.412037037037037, + "grad_norm": 0.6624761819839478, + "learning_rate": 0.00014471676832774613, + "loss": 1.2244, + "step": 7931 + }, + { + "epoch": 1.4122150997150997, + "grad_norm": 0.6205778121948242, + "learning_rate": 0.00014470424784870088, + "loss": 1.1, + "step": 7932 + }, + { + "epoch": 1.4123931623931623, + "grad_norm": 0.7592337131500244, + "learning_rate": 0.00014469172649376875, + "loss": 0.963, + "step": 7933 + }, + { + "epoch": 1.412571225071225, + "grad_norm": 0.673328697681427, + "learning_rate": 0.00014467920426319508, + "loss": 0.8923, + "step": 7934 + }, + { + "epoch": 1.4127492877492878, + "grad_norm": 0.6064394116401672, + "learning_rate": 0.00014466668115722522, + "loss": 0.9679, + "step": 7935 + }, + { + "epoch": 1.4129273504273505, + "grad_norm": 0.7738677859306335, + "learning_rate": 0.00014465415717610454, + "loss": 1.0678, + "step": 7936 + }, + { + "epoch": 1.413105413105413, + "grad_norm": 0.7013397812843323, + "learning_rate": 0.00014464163232007836, + "loss": 0.9017, + "step": 7937 + }, + { + "epoch": 1.4132834757834758, + "grad_norm": 0.713291347026825, + "learning_rate": 0.0001446291065893922, + "loss": 1.1953, + "step": 7938 + }, + { + "epoch": 1.4134615384615383, + "grad_norm": 0.7538655996322632, + "learning_rate": 0.00014461657998429136, + "loss": 1.0571, + "step": 7939 + }, + { + "epoch": 1.413639601139601, + "grad_norm": 0.6358973383903503, + "learning_rate": 0.00014460405250502133, + "loss": 0.8552, + "step": 7940 + }, + { + "epoch": 1.4138176638176638, + "grad_norm": 0.67508864402771, + "learning_rate": 0.00014459152415182756, + "loss": 1.0293, + "step": 7941 + }, + { + "epoch": 1.4139957264957266, + "grad_norm": 0.7074598670005798, + "learning_rate": 0.00014457899492495546, + "loss": 1.2102, + "step": 7942 + }, + { + "epoch": 1.414173789173789, + "grad_norm": 0.7157037854194641, + "learning_rate": 0.00014456646482465058, + "loss": 1.0566, + "step": 7943 + }, + { + "epoch": 1.4143518518518519, + "grad_norm": 0.7918477058410645, + "learning_rate": 0.00014455393385115844, + "loss": 1.3727, + "step": 7944 + }, + { + "epoch": 1.4145299145299146, + "grad_norm": 0.569144606590271, + "learning_rate": 0.0001445414020047245, + "loss": 0.7251, + "step": 7945 + }, + { + "epoch": 1.4147079772079771, + "grad_norm": 0.7589054107666016, + "learning_rate": 0.0001445288692855943, + "loss": 1.0155, + "step": 7946 + }, + { + "epoch": 1.41488603988604, + "grad_norm": 0.7531685829162598, + "learning_rate": 0.0001445163356940134, + "loss": 0.8404, + "step": 7947 + }, + { + "epoch": 1.4150641025641026, + "grad_norm": 0.5730917453765869, + "learning_rate": 0.0001445038012302274, + "loss": 0.8215, + "step": 7948 + }, + { + "epoch": 1.4152421652421652, + "grad_norm": 0.6960710883140564, + "learning_rate": 0.00014449126589448187, + "loss": 0.7902, + "step": 7949 + }, + { + "epoch": 1.415420227920228, + "grad_norm": 0.8207054138183594, + "learning_rate": 0.0001444787296870224, + "loss": 1.493, + "step": 7950 + }, + { + "epoch": 1.4155982905982907, + "grad_norm": 0.5854668617248535, + "learning_rate": 0.00014446619260809462, + "loss": 0.9262, + "step": 7951 + }, + { + "epoch": 1.4157763532763532, + "grad_norm": 0.5458414554595947, + "learning_rate": 0.00014445365465794413, + "loss": 0.8431, + "step": 7952 + }, + { + "epoch": 1.415954415954416, + "grad_norm": 0.6880569458007812, + "learning_rate": 0.00014444111583681666, + "loss": 1.0184, + "step": 7953 + }, + { + "epoch": 1.4161324786324787, + "grad_norm": 0.6391083598136902, + "learning_rate": 0.00014442857614495783, + "loss": 0.88, + "step": 7954 + }, + { + "epoch": 1.4163105413105412, + "grad_norm": 0.6246135234832764, + "learning_rate": 0.00014441603558261335, + "loss": 0.776, + "step": 7955 + }, + { + "epoch": 1.416488603988604, + "grad_norm": 0.6263493895530701, + "learning_rate": 0.00014440349415002893, + "loss": 0.9069, + "step": 7956 + }, + { + "epoch": 1.4166666666666667, + "grad_norm": 0.7123475670814514, + "learning_rate": 0.00014439095184745024, + "loss": 0.8339, + "step": 7957 + }, + { + "epoch": 1.4168447293447293, + "grad_norm": 0.7171050906181335, + "learning_rate": 0.00014437840867512309, + "loss": 1.0633, + "step": 7958 + }, + { + "epoch": 1.417022792022792, + "grad_norm": 0.7097769975662231, + "learning_rate": 0.00014436586463329322, + "loss": 1.0852, + "step": 7959 + }, + { + "epoch": 1.4172008547008548, + "grad_norm": 0.6889223456382751, + "learning_rate": 0.00014435331972220637, + "loss": 0.916, + "step": 7960 + }, + { + "epoch": 1.4173789173789173, + "grad_norm": 0.6674435138702393, + "learning_rate": 0.0001443407739421084, + "loss": 0.9307, + "step": 7961 + }, + { + "epoch": 1.41755698005698, + "grad_norm": 0.6578894853591919, + "learning_rate": 0.00014432822729324503, + "loss": 0.8767, + "step": 7962 + }, + { + "epoch": 1.4177350427350428, + "grad_norm": 0.7145379781723022, + "learning_rate": 0.00014431567977586212, + "loss": 0.9962, + "step": 7963 + }, + { + "epoch": 1.4179131054131053, + "grad_norm": 0.6916680335998535, + "learning_rate": 0.00014430313139020555, + "loss": 1.0464, + "step": 7964 + }, + { + "epoch": 1.418091168091168, + "grad_norm": 0.6296181678771973, + "learning_rate": 0.00014429058213652116, + "loss": 1.0699, + "step": 7965 + }, + { + "epoch": 1.4182692307692308, + "grad_norm": 0.5640227198600769, + "learning_rate": 0.00014427803201505482, + "loss": 0.7006, + "step": 7966 + }, + { + "epoch": 1.4184472934472934, + "grad_norm": 0.7181212306022644, + "learning_rate": 0.0001442654810260524, + "loss": 1.1648, + "step": 7967 + }, + { + "epoch": 1.4186253561253561, + "grad_norm": 0.6830772757530212, + "learning_rate": 0.00014425292916975984, + "loss": 1.0641, + "step": 7968 + }, + { + "epoch": 1.4188034188034189, + "grad_norm": 0.665716290473938, + "learning_rate": 0.00014424037644642307, + "loss": 0.8769, + "step": 7969 + }, + { + "epoch": 1.4189814814814814, + "grad_norm": 0.8088666796684265, + "learning_rate": 0.00014422782285628802, + "loss": 1.1496, + "step": 7970 + }, + { + "epoch": 1.4191595441595442, + "grad_norm": 0.7186072468757629, + "learning_rate": 0.00014421526839960064, + "loss": 0.7421, + "step": 7971 + }, + { + "epoch": 1.419337606837607, + "grad_norm": 0.6405926942825317, + "learning_rate": 0.00014420271307660694, + "loss": 1.0139, + "step": 7972 + }, + { + "epoch": 1.4195156695156697, + "grad_norm": 0.7097104787826538, + "learning_rate": 0.0001441901568875529, + "loss": 1.1582, + "step": 7973 + }, + { + "epoch": 1.4196937321937322, + "grad_norm": 0.7347947359085083, + "learning_rate": 0.00014417759983268452, + "loss": 0.9751, + "step": 7974 + }, + { + "epoch": 1.419871794871795, + "grad_norm": 0.6999621987342834, + "learning_rate": 0.00014416504191224787, + "loss": 0.9419, + "step": 7975 + }, + { + "epoch": 1.4200498575498575, + "grad_norm": 0.6500616073608398, + "learning_rate": 0.00014415248312648897, + "loss": 0.9407, + "step": 7976 + }, + { + "epoch": 1.4202279202279202, + "grad_norm": 0.6368781328201294, + "learning_rate": 0.00014413992347565383, + "loss": 1.1224, + "step": 7977 + }, + { + "epoch": 1.420405982905983, + "grad_norm": 0.6422648429870605, + "learning_rate": 0.00014412736295998864, + "loss": 0.9573, + "step": 7978 + }, + { + "epoch": 1.4205840455840457, + "grad_norm": 0.744057297706604, + "learning_rate": 0.00014411480157973942, + "loss": 1.1384, + "step": 7979 + }, + { + "epoch": 1.4207621082621082, + "grad_norm": 0.5905839204788208, + "learning_rate": 0.00014410223933515232, + "loss": 0.8212, + "step": 7980 + }, + { + "epoch": 1.420940170940171, + "grad_norm": 0.5905438661575317, + "learning_rate": 0.0001440896762264734, + "loss": 0.8281, + "step": 7981 + }, + { + "epoch": 1.4211182336182335, + "grad_norm": 0.7087140679359436, + "learning_rate": 0.00014407711225394892, + "loss": 1.0165, + "step": 7982 + }, + { + "epoch": 1.4212962962962963, + "grad_norm": 0.6173902153968811, + "learning_rate": 0.00014406454741782495, + "loss": 0.8823, + "step": 7983 + }, + { + "epoch": 1.421474358974359, + "grad_norm": 0.6649761199951172, + "learning_rate": 0.00014405198171834772, + "loss": 0.9489, + "step": 7984 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 0.619286835193634, + "learning_rate": 0.00014403941515576344, + "loss": 0.8149, + "step": 7985 + }, + { + "epoch": 1.4218304843304843, + "grad_norm": 0.6358469724655151, + "learning_rate": 0.0001440268477303183, + "loss": 1.0558, + "step": 7986 + }, + { + "epoch": 1.422008547008547, + "grad_norm": 0.7239769697189331, + "learning_rate": 0.0001440142794422585, + "loss": 1.0528, + "step": 7987 + }, + { + "epoch": 1.4221866096866096, + "grad_norm": 0.681168794631958, + "learning_rate": 0.00014400171029183036, + "loss": 1.0867, + "step": 7988 + }, + { + "epoch": 1.4223646723646723, + "grad_norm": 0.6741157174110413, + "learning_rate": 0.0001439891402792801, + "loss": 0.9153, + "step": 7989 + }, + { + "epoch": 1.422542735042735, + "grad_norm": 0.5881659984588623, + "learning_rate": 0.00014397656940485403, + "loss": 0.92, + "step": 7990 + }, + { + "epoch": 1.4227207977207978, + "grad_norm": 0.637093722820282, + "learning_rate": 0.00014396399766879842, + "loss": 0.921, + "step": 7991 + }, + { + "epoch": 1.4228988603988604, + "grad_norm": 0.7760605216026306, + "learning_rate": 0.0001439514250713596, + "loss": 1.1451, + "step": 7992 + }, + { + "epoch": 1.4230769230769231, + "grad_norm": 0.6619600653648376, + "learning_rate": 0.00014393885161278393, + "loss": 1.0365, + "step": 7993 + }, + { + "epoch": 1.4232549857549857, + "grad_norm": 0.5354374051094055, + "learning_rate": 0.0001439262772933177, + "loss": 0.8718, + "step": 7994 + }, + { + "epoch": 1.4234330484330484, + "grad_norm": 0.7063560485839844, + "learning_rate": 0.00014391370211320735, + "loss": 0.8258, + "step": 7995 + }, + { + "epoch": 1.4236111111111112, + "grad_norm": 0.6876368522644043, + "learning_rate": 0.00014390112607269923, + "loss": 0.9579, + "step": 7996 + }, + { + "epoch": 1.423789173789174, + "grad_norm": 0.6976612210273743, + "learning_rate": 0.00014388854917203974, + "loss": 1.0376, + "step": 7997 + }, + { + "epoch": 1.4239672364672364, + "grad_norm": 0.6157355308532715, + "learning_rate": 0.00014387597141147525, + "loss": 0.8743, + "step": 7998 + }, + { + "epoch": 1.4241452991452992, + "grad_norm": 0.7273156046867371, + "learning_rate": 0.0001438633927912523, + "loss": 1.101, + "step": 7999 + }, + { + "epoch": 1.4243233618233617, + "grad_norm": 0.918380618095398, + "learning_rate": 0.0001438508133116173, + "loss": 0.9625, + "step": 8000 + }, + { + "epoch": 1.4245014245014245, + "grad_norm": 0.626040518283844, + "learning_rate": 0.00014383823297281666, + "loss": 0.9552, + "step": 8001 + }, + { + "epoch": 1.4246794871794872, + "grad_norm": 0.7320386171340942, + "learning_rate": 0.00014382565177509693, + "loss": 1.0719, + "step": 8002 + }, + { + "epoch": 1.42485754985755, + "grad_norm": 0.7283148169517517, + "learning_rate": 0.0001438130697187046, + "loss": 1.0455, + "step": 8003 + }, + { + "epoch": 1.4250356125356125, + "grad_norm": 0.6614177823066711, + "learning_rate": 0.00014380048680388613, + "loss": 0.9876, + "step": 8004 + }, + { + "epoch": 1.4252136752136753, + "grad_norm": 0.6726453900337219, + "learning_rate": 0.00014378790303088817, + "loss": 0.9861, + "step": 8005 + }, + { + "epoch": 1.4253917378917378, + "grad_norm": 0.7968725562095642, + "learning_rate": 0.00014377531839995718, + "loss": 1.1662, + "step": 8006 + }, + { + "epoch": 1.4255698005698005, + "grad_norm": 0.6510586738586426, + "learning_rate": 0.0001437627329113398, + "loss": 0.9452, + "step": 8007 + }, + { + "epoch": 1.4257478632478633, + "grad_norm": 0.6933155655860901, + "learning_rate": 0.00014375014656528253, + "loss": 1.0149, + "step": 8008 + }, + { + "epoch": 1.425925925925926, + "grad_norm": 0.7141832113265991, + "learning_rate": 0.00014373755936203204, + "loss": 1.0667, + "step": 8009 + }, + { + "epoch": 1.4261039886039886, + "grad_norm": 0.6352181434631348, + "learning_rate": 0.00014372497130183494, + "loss": 0.8652, + "step": 8010 + }, + { + "epoch": 1.4262820512820513, + "grad_norm": 0.7494860291481018, + "learning_rate": 0.00014371238238493786, + "loss": 0.9592, + "step": 8011 + }, + { + "epoch": 1.4264601139601139, + "grad_norm": 0.610556423664093, + "learning_rate": 0.00014369979261158746, + "loss": 0.7015, + "step": 8012 + }, + { + "epoch": 1.4266381766381766, + "grad_norm": 0.7305756211280823, + "learning_rate": 0.00014368720198203037, + "loss": 0.9681, + "step": 8013 + }, + { + "epoch": 1.4268162393162394, + "grad_norm": 0.6964020133018494, + "learning_rate": 0.0001436746104965133, + "loss": 1.1166, + "step": 8014 + }, + { + "epoch": 1.426994301994302, + "grad_norm": 0.7449237108230591, + "learning_rate": 0.00014366201815528302, + "loss": 1.1331, + "step": 8015 + }, + { + "epoch": 1.4271723646723646, + "grad_norm": 0.625834047794342, + "learning_rate": 0.00014364942495858615, + "loss": 0.8796, + "step": 8016 + }, + { + "epoch": 1.4273504273504274, + "grad_norm": 0.664559006690979, + "learning_rate": 0.0001436368309066695, + "loss": 1.0263, + "step": 8017 + }, + { + "epoch": 1.42752849002849, + "grad_norm": Infinity, + "learning_rate": 0.0001436368309066695, + "loss": 1.0731, + "step": 8018 + }, + { + "epoch": 1.4277065527065527, + "grad_norm": 0.6714464426040649, + "learning_rate": 0.00014362423599977977, + "loss": 0.9345, + "step": 8019 + }, + { + "epoch": 1.4278846153846154, + "grad_norm": 0.7595751285552979, + "learning_rate": 0.00014361164023816376, + "loss": 0.9646, + "step": 8020 + }, + { + "epoch": 1.4280626780626782, + "grad_norm": 0.6413954496383667, + "learning_rate": 0.00014359904362206828, + "loss": 1.0471, + "step": 8021 + }, + { + "epoch": 1.4282407407407407, + "grad_norm": 0.7298843264579773, + "learning_rate": 0.00014358644615174008, + "loss": 0.8932, + "step": 8022 + }, + { + "epoch": 1.4284188034188035, + "grad_norm": 0.8022156953811646, + "learning_rate": 0.00014357384782742602, + "loss": 1.0437, + "step": 8023 + }, + { + "epoch": 1.428596866096866, + "grad_norm": 0.7264443635940552, + "learning_rate": 0.00014356124864937296, + "loss": 0.9368, + "step": 8024 + }, + { + "epoch": 1.4287749287749287, + "grad_norm": 0.6819384098052979, + "learning_rate": 0.00014354864861782768, + "loss": 1.0, + "step": 8025 + }, + { + "epoch": 1.4289529914529915, + "grad_norm": 0.5945104956626892, + "learning_rate": 0.0001435360477330371, + "loss": 0.8108, + "step": 8026 + }, + { + "epoch": 1.4291310541310542, + "grad_norm": 0.6497398018836975, + "learning_rate": 0.0001435234459952481, + "loss": 0.8712, + "step": 8027 + }, + { + "epoch": 1.4293091168091168, + "grad_norm": 0.6424077749252319, + "learning_rate": 0.0001435108434047076, + "loss": 0.9172, + "step": 8028 + }, + { + "epoch": 1.4294871794871795, + "grad_norm": 0.6806963086128235, + "learning_rate": 0.00014349823996166253, + "loss": 1.1648, + "step": 8029 + }, + { + "epoch": 1.429665242165242, + "grad_norm": 0.6601083874702454, + "learning_rate": 0.00014348563566635977, + "loss": 0.9453, + "step": 8030 + }, + { + "epoch": 1.4298433048433048, + "grad_norm": 0.7024385929107666, + "learning_rate": 0.00014347303051904636, + "loss": 1.074, + "step": 8031 + }, + { + "epoch": 1.4300213675213675, + "grad_norm": 0.7094005942344666, + "learning_rate": 0.00014346042451996918, + "loss": 0.9976, + "step": 8032 + }, + { + "epoch": 1.4301994301994303, + "grad_norm": 0.6775936484336853, + "learning_rate": 0.0001434478176693753, + "loss": 0.9039, + "step": 8033 + }, + { + "epoch": 1.4303774928774928, + "grad_norm": 0.6920986771583557, + "learning_rate": 0.00014343520996751166, + "loss": 0.9122, + "step": 8034 + }, + { + "epoch": 1.4305555555555556, + "grad_norm": 0.720690906047821, + "learning_rate": 0.00014342260141462528, + "loss": 1.1028, + "step": 8035 + }, + { + "epoch": 1.430733618233618, + "grad_norm": 0.624546229839325, + "learning_rate": 0.00014340999201096328, + "loss": 0.9083, + "step": 8036 + }, + { + "epoch": 1.4309116809116809, + "grad_norm": 0.6560490727424622, + "learning_rate": 0.00014339738175677265, + "loss": 0.8029, + "step": 8037 + }, + { + "epoch": 1.4310897435897436, + "grad_norm": 0.8266100883483887, + "learning_rate": 0.00014338477065230047, + "loss": 0.9655, + "step": 8038 + }, + { + "epoch": 1.4312678062678064, + "grad_norm": 0.6593570113182068, + "learning_rate": 0.00014337215869779385, + "loss": 1.0299, + "step": 8039 + }, + { + "epoch": 1.431445868945869, + "grad_norm": 0.6321794390678406, + "learning_rate": 0.00014335954589349986, + "loss": 0.8755, + "step": 8040 + }, + { + "epoch": 1.4316239316239316, + "grad_norm": 0.7030870318412781, + "learning_rate": 0.00014334693223966562, + "loss": 1.1226, + "step": 8041 + }, + { + "epoch": 1.4318019943019942, + "grad_norm": 0.7794312238693237, + "learning_rate": 0.0001433343177365383, + "loss": 1.1252, + "step": 8042 + }, + { + "epoch": 1.431980056980057, + "grad_norm": 0.6115018129348755, + "learning_rate": 0.00014332170238436507, + "loss": 0.8753, + "step": 8043 + }, + { + "epoch": 1.4321581196581197, + "grad_norm": 0.8525674939155579, + "learning_rate": 0.00014330908618339304, + "loss": 0.9135, + "step": 8044 + }, + { + "epoch": 1.4323361823361824, + "grad_norm": 0.6869912147521973, + "learning_rate": 0.00014329646913386948, + "loss": 0.868, + "step": 8045 + }, + { + "epoch": 1.432514245014245, + "grad_norm": 0.5877542495727539, + "learning_rate": 0.0001432838512360415, + "loss": 0.9051, + "step": 8046 + }, + { + "epoch": 1.4326923076923077, + "grad_norm": 0.6609327793121338, + "learning_rate": 0.0001432712324901564, + "loss": 0.9084, + "step": 8047 + }, + { + "epoch": 1.4328703703703702, + "grad_norm": 0.6318345069885254, + "learning_rate": 0.0001432586128964614, + "loss": 0.8291, + "step": 8048 + }, + { + "epoch": 1.433048433048433, + "grad_norm": 0.6973567008972168, + "learning_rate": 0.0001432459924552037, + "loss": 0.97, + "step": 8049 + }, + { + "epoch": 1.4332264957264957, + "grad_norm": 0.6838201284408569, + "learning_rate": 0.00014323337116663062, + "loss": 1.0957, + "step": 8050 + }, + { + "epoch": 1.4334045584045585, + "grad_norm": 0.7472857236862183, + "learning_rate": 0.00014322074903098944, + "loss": 1.0981, + "step": 8051 + }, + { + "epoch": 1.433582621082621, + "grad_norm": 0.7723061442375183, + "learning_rate": 0.0001432081260485275, + "loss": 1.2231, + "step": 8052 + }, + { + "epoch": 1.4337606837606838, + "grad_norm": 0.681834876537323, + "learning_rate": 0.00014319550221949208, + "loss": 1.073, + "step": 8053 + }, + { + "epoch": 1.4339387464387463, + "grad_norm": 0.6566045880317688, + "learning_rate": 0.00014318287754413051, + "loss": 1.1298, + "step": 8054 + }, + { + "epoch": 1.434116809116809, + "grad_norm": 0.6792440414428711, + "learning_rate": 0.00014317025202269015, + "loss": 1.2224, + "step": 8055 + }, + { + "epoch": 1.4342948717948718, + "grad_norm": 0.7946709394454956, + "learning_rate": 0.00014315762565541838, + "loss": 1.0728, + "step": 8056 + }, + { + "epoch": 1.4344729344729346, + "grad_norm": 0.633466899394989, + "learning_rate": 0.00014314499844256262, + "loss": 0.944, + "step": 8057 + }, + { + "epoch": 1.434650997150997, + "grad_norm": 0.7308502197265625, + "learning_rate": 0.00014313237038437023, + "loss": 1.0684, + "step": 8058 + }, + { + "epoch": 1.4348290598290598, + "grad_norm": 0.6483737230300903, + "learning_rate": 0.00014311974148108862, + "loss": 1.0843, + "step": 8059 + }, + { + "epoch": 1.4350071225071226, + "grad_norm": 0.6301209926605225, + "learning_rate": 0.00014310711173296526, + "loss": 1.0083, + "step": 8060 + }, + { + "epoch": 1.4351851851851851, + "grad_norm": 0.6674302816390991, + "learning_rate": 0.00014309448114024757, + "loss": 0.9877, + "step": 8061 + }, + { + "epoch": 1.4353632478632479, + "grad_norm": 0.6888732314109802, + "learning_rate": 0.00014308184970318307, + "loss": 0.9937, + "step": 8062 + }, + { + "epoch": 1.4355413105413106, + "grad_norm": 0.6922950148582458, + "learning_rate": 0.00014306921742201923, + "loss": 1.0149, + "step": 8063 + }, + { + "epoch": 1.4357193732193732, + "grad_norm": 0.6050686240196228, + "learning_rate": 0.00014305658429700352, + "loss": 0.7882, + "step": 8064 + }, + { + "epoch": 1.435897435897436, + "grad_norm": 0.5080767869949341, + "learning_rate": 0.00014304395032838348, + "loss": 0.7796, + "step": 8065 + }, + { + "epoch": 1.4360754985754987, + "grad_norm": 0.6382707953453064, + "learning_rate": 0.00014303131551640668, + "loss": 0.965, + "step": 8066 + }, + { + "epoch": 1.4362535612535612, + "grad_norm": 0.7153477668762207, + "learning_rate": 0.00014301867986132063, + "loss": 1.1277, + "step": 8067 + }, + { + "epoch": 1.436431623931624, + "grad_norm": 0.6208404898643494, + "learning_rate": 0.00014300604336337292, + "loss": 0.8246, + "step": 8068 + }, + { + "epoch": 1.4366096866096867, + "grad_norm": 0.719695508480072, + "learning_rate": 0.0001429934060228111, + "loss": 0.7681, + "step": 8069 + }, + { + "epoch": 1.4367877492877492, + "grad_norm": 0.6219030618667603, + "learning_rate": 0.0001429807678398828, + "loss": 1.0425, + "step": 8070 + }, + { + "epoch": 1.436965811965812, + "grad_norm": 0.6080238819122314, + "learning_rate": 0.00014296812881483566, + "loss": 0.8762, + "step": 8071 + }, + { + "epoch": 1.4371438746438747, + "grad_norm": 0.6264194846153259, + "learning_rate": 0.00014295548894791729, + "loss": 1.087, + "step": 8072 + }, + { + "epoch": 1.4373219373219372, + "grad_norm": 0.6503600478172302, + "learning_rate": 0.00014294284823937535, + "loss": 1.0583, + "step": 8073 + }, + { + "epoch": 1.4375, + "grad_norm": 0.7623817324638367, + "learning_rate": 0.0001429302066894575, + "loss": 1.2372, + "step": 8074 + }, + { + "epoch": 1.4376780626780628, + "grad_norm": 0.7020344138145447, + "learning_rate": 0.00014291756429841144, + "loss": 1.2163, + "step": 8075 + }, + { + "epoch": 1.4378561253561253, + "grad_norm": 0.7070338129997253, + "learning_rate": 0.00014290492106648484, + "loss": 0.986, + "step": 8076 + }, + { + "epoch": 1.438034188034188, + "grad_norm": 0.6407621502876282, + "learning_rate": 0.00014289227699392545, + "loss": 0.9329, + "step": 8077 + }, + { + "epoch": 1.4382122507122508, + "grad_norm": 0.6836710572242737, + "learning_rate": 0.00014287963208098098, + "loss": 0.9252, + "step": 8078 + }, + { + "epoch": 1.4383903133903133, + "grad_norm": 0.648642897605896, + "learning_rate": 0.00014286698632789922, + "loss": 1.0457, + "step": 8079 + }, + { + "epoch": 1.438568376068376, + "grad_norm": 0.7015881538391113, + "learning_rate": 0.0001428543397349279, + "loss": 1.0516, + "step": 8080 + }, + { + "epoch": 1.4387464387464388, + "grad_norm": 0.6031532883644104, + "learning_rate": 0.0001428416923023148, + "loss": 0.9423, + "step": 8081 + }, + { + "epoch": 1.4389245014245013, + "grad_norm": 0.8235578536987305, + "learning_rate": 0.00014282904403030772, + "loss": 1.3433, + "step": 8082 + }, + { + "epoch": 1.439102564102564, + "grad_norm": 0.7355761528015137, + "learning_rate": 0.00014281639491915452, + "loss": 1.0128, + "step": 8083 + }, + { + "epoch": 1.4392806267806268, + "grad_norm": 0.7429629564285278, + "learning_rate": 0.00014280374496910303, + "loss": 0.8546, + "step": 8084 + }, + { + "epoch": 1.4394586894586894, + "grad_norm": 0.5831776857376099, + "learning_rate": 0.00014279109418040105, + "loss": 0.9021, + "step": 8085 + }, + { + "epoch": 1.4396367521367521, + "grad_norm": 0.6585184931755066, + "learning_rate": 0.00014277844255329645, + "loss": 0.9256, + "step": 8086 + }, + { + "epoch": 1.4398148148148149, + "grad_norm": 0.6412501931190491, + "learning_rate": 0.00014276579008803717, + "loss": 0.9305, + "step": 8087 + }, + { + "epoch": 1.4399928774928774, + "grad_norm": 0.6305423378944397, + "learning_rate": 0.00014275313678487102, + "loss": 0.9471, + "step": 8088 + }, + { + "epoch": 1.4401709401709402, + "grad_norm": 0.7160914540290833, + "learning_rate": 0.00014274048264404602, + "loss": 0.8798, + "step": 8089 + }, + { + "epoch": 1.440349002849003, + "grad_norm": 0.6740858554840088, + "learning_rate": 0.00014272782766581004, + "loss": 0.9022, + "step": 8090 + }, + { + "epoch": 1.4405270655270654, + "grad_norm": 0.7554821968078613, + "learning_rate": 0.000142715171850411, + "loss": 1.0924, + "step": 8091 + }, + { + "epoch": 1.4407051282051282, + "grad_norm": 0.7361162304878235, + "learning_rate": 0.00014270251519809694, + "loss": 0.9907, + "step": 8092 + }, + { + "epoch": 1.440883190883191, + "grad_norm": 0.731813371181488, + "learning_rate": 0.0001426898577091158, + "loss": 1.1765, + "step": 8093 + }, + { + "epoch": 1.4410612535612537, + "grad_norm": 0.6877756714820862, + "learning_rate": 0.00014267719938371558, + "loss": 1.0536, + "step": 8094 + }, + { + "epoch": 1.4412393162393162, + "grad_norm": 0.6724407076835632, + "learning_rate": 0.00014266454022214426, + "loss": 1.1895, + "step": 8095 + }, + { + "epoch": 1.441417378917379, + "grad_norm": 0.6946671605110168, + "learning_rate": 0.0001426518802246499, + "loss": 1.0437, + "step": 8096 + }, + { + "epoch": 1.4415954415954415, + "grad_norm": 0.7032839059829712, + "learning_rate": 0.00014263921939148058, + "loss": 1.1363, + "step": 8097 + }, + { + "epoch": 1.4417735042735043, + "grad_norm": 0.6942192316055298, + "learning_rate": 0.00014262655772288434, + "loss": 1.315, + "step": 8098 + }, + { + "epoch": 1.441951566951567, + "grad_norm": 0.7002301812171936, + "learning_rate": 0.00014261389521910922, + "loss": 1.0546, + "step": 8099 + }, + { + "epoch": 1.4421296296296298, + "grad_norm": 0.7260788083076477, + "learning_rate": 0.00014260123188040335, + "loss": 0.9374, + "step": 8100 + }, + { + "epoch": 1.4423076923076923, + "grad_norm": 0.6629201173782349, + "learning_rate": 0.00014258856770701486, + "loss": 0.8632, + "step": 8101 + }, + { + "epoch": 1.442485754985755, + "grad_norm": 0.6570318937301636, + "learning_rate": 0.0001425759026991918, + "loss": 1.0102, + "step": 8102 + }, + { + "epoch": 1.4426638176638176, + "grad_norm": 0.7696560621261597, + "learning_rate": 0.00014256323685718242, + "loss": 0.9703, + "step": 8103 + }, + { + "epoch": 1.4428418803418803, + "grad_norm": 0.7206611633300781, + "learning_rate": 0.00014255057018123482, + "loss": 1.1728, + "step": 8104 + }, + { + "epoch": 1.443019943019943, + "grad_norm": 0.6871611475944519, + "learning_rate": 0.0001425379026715972, + "loss": 0.9377, + "step": 8105 + }, + { + "epoch": 1.4431980056980058, + "grad_norm": 0.6027442812919617, + "learning_rate": 0.00014252523432851775, + "loss": 0.9212, + "step": 8106 + }, + { + "epoch": 1.4433760683760684, + "grad_norm": 0.7149752378463745, + "learning_rate": 0.00014251256515224463, + "loss": 0.9654, + "step": 8107 + }, + { + "epoch": 1.443554131054131, + "grad_norm": 0.5949522256851196, + "learning_rate": 0.00014249989514302614, + "loss": 1.0646, + "step": 8108 + }, + { + "epoch": 1.4437321937321936, + "grad_norm": 0.7345452904701233, + "learning_rate": 0.0001424872243011105, + "loss": 0.9801, + "step": 8109 + }, + { + "epoch": 1.4439102564102564, + "grad_norm": 0.8045009970664978, + "learning_rate": 0.00014247455262674592, + "loss": 1.3529, + "step": 8110 + }, + { + "epoch": 1.4440883190883191, + "grad_norm": 0.6712123155593872, + "learning_rate": 0.00014246188012018073, + "loss": 1.0416, + "step": 8111 + }, + { + "epoch": 1.444266381766382, + "grad_norm": 0.7811154127120972, + "learning_rate": 0.00014244920678166322, + "loss": 1.2019, + "step": 8112 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 0.6834486126899719, + "learning_rate": 0.00014243653261144167, + "loss": 0.986, + "step": 8113 + }, + { + "epoch": 1.4446225071225072, + "grad_norm": 0.6901041269302368, + "learning_rate": 0.00014242385760976443, + "loss": 1.0988, + "step": 8114 + }, + { + "epoch": 1.4448005698005697, + "grad_norm": 0.6233634948730469, + "learning_rate": 0.00014241118177687982, + "loss": 0.7748, + "step": 8115 + }, + { + "epoch": 1.4449786324786325, + "grad_norm": 0.6899837851524353, + "learning_rate": 0.00014239850511303624, + "loss": 0.9734, + "step": 8116 + }, + { + "epoch": 1.4451566951566952, + "grad_norm": 0.6316244006156921, + "learning_rate": 0.00014238582761848197, + "loss": 0.7888, + "step": 8117 + }, + { + "epoch": 1.445334757834758, + "grad_norm": 0.6074259877204895, + "learning_rate": 0.00014237314929346545, + "loss": 0.8843, + "step": 8118 + }, + { + "epoch": 1.4455128205128205, + "grad_norm": 0.6112192273139954, + "learning_rate": 0.00014236047013823516, + "loss": 0.8529, + "step": 8119 + }, + { + "epoch": 1.4456908831908832, + "grad_norm": 0.6883894801139832, + "learning_rate": 0.0001423477901530394, + "loss": 0.9506, + "step": 8120 + }, + { + "epoch": 1.4458689458689458, + "grad_norm": 0.7248309254646301, + "learning_rate": 0.00014233510933812666, + "loss": 0.9573, + "step": 8121 + }, + { + "epoch": 1.4460470085470085, + "grad_norm": 0.6853367686271667, + "learning_rate": 0.00014232242769374542, + "loss": 0.9903, + "step": 8122 + }, + { + "epoch": 1.4462250712250713, + "grad_norm": 0.7179274559020996, + "learning_rate": 0.0001423097452201441, + "loss": 0.9157, + "step": 8123 + }, + { + "epoch": 1.446403133903134, + "grad_norm": 0.6704817414283752, + "learning_rate": 0.00014229706191757127, + "loss": 1.1361, + "step": 8124 + }, + { + "epoch": 1.4465811965811965, + "grad_norm": 0.6380739212036133, + "learning_rate": 0.00014228437778627533, + "loss": 0.9336, + "step": 8125 + }, + { + "epoch": 1.4467592592592593, + "grad_norm": 0.6275362372398376, + "learning_rate": 0.00014227169282650487, + "loss": 0.9617, + "step": 8126 + }, + { + "epoch": 1.4469373219373218, + "grad_norm": 0.5644828677177429, + "learning_rate": 0.00014225900703850836, + "loss": 0.7384, + "step": 8127 + }, + { + "epoch": 1.4471153846153846, + "grad_norm": 0.6522284150123596, + "learning_rate": 0.00014224632042253443, + "loss": 1.1098, + "step": 8128 + }, + { + "epoch": 1.4472934472934473, + "grad_norm": 0.6228049993515015, + "learning_rate": 0.0001422336329788316, + "loss": 1.1061, + "step": 8129 + }, + { + "epoch": 1.44747150997151, + "grad_norm": 0.6092000603675842, + "learning_rate": 0.00014222094470764848, + "loss": 0.808, + "step": 8130 + }, + { + "epoch": 1.4476495726495726, + "grad_norm": 0.667435348033905, + "learning_rate": 0.00014220825560923363, + "loss": 1.1223, + "step": 8131 + }, + { + "epoch": 1.4478276353276354, + "grad_norm": 0.6080766320228577, + "learning_rate": 0.0001421955656838357, + "loss": 1.0099, + "step": 8132 + }, + { + "epoch": 1.448005698005698, + "grad_norm": 0.7597638368606567, + "learning_rate": 0.00014218287493170332, + "loss": 0.9718, + "step": 8133 + }, + { + "epoch": 1.4481837606837606, + "grad_norm": 0.574130654335022, + "learning_rate": 0.0001421701833530851, + "loss": 0.7745, + "step": 8134 + }, + { + "epoch": 1.4483618233618234, + "grad_norm": 0.6372822523117065, + "learning_rate": 0.0001421574909482298, + "loss": 1.0088, + "step": 8135 + }, + { + "epoch": 1.4485398860398861, + "grad_norm": 0.6759644746780396, + "learning_rate": 0.000142144797717386, + "loss": 0.9684, + "step": 8136 + }, + { + "epoch": 1.4487179487179487, + "grad_norm": 0.706351637840271, + "learning_rate": 0.00014213210366080244, + "loss": 1.021, + "step": 8137 + }, + { + "epoch": 1.4488960113960114, + "grad_norm": 0.6976894736289978, + "learning_rate": 0.0001421194087787278, + "loss": 1.1038, + "step": 8138 + }, + { + "epoch": 1.449074074074074, + "grad_norm": 0.7322551012039185, + "learning_rate": 0.00014210671307141092, + "loss": 1.0213, + "step": 8139 + }, + { + "epoch": 1.4492521367521367, + "grad_norm": 0.5885626077651978, + "learning_rate": 0.0001420940165391004, + "loss": 0.821, + "step": 8140 + }, + { + "epoch": 1.4494301994301995, + "grad_norm": 0.7009791135787964, + "learning_rate": 0.0001420813191820451, + "loss": 0.8647, + "step": 8141 + }, + { + "epoch": 1.4496082621082622, + "grad_norm": 0.5715423822402954, + "learning_rate": 0.00014206862100049375, + "loss": 0.873, + "step": 8142 + }, + { + "epoch": 1.4497863247863247, + "grad_norm": 1.1452178955078125, + "learning_rate": 0.00014205592199469514, + "loss": 1.2523, + "step": 8143 + }, + { + "epoch": 1.4499643874643875, + "grad_norm": 0.8076814413070679, + "learning_rate": 0.00014204322216489814, + "loss": 1.1071, + "step": 8144 + }, + { + "epoch": 1.45014245014245, + "grad_norm": 0.7325751185417175, + "learning_rate": 0.00014203052151135154, + "loss": 0.9846, + "step": 8145 + }, + { + "epoch": 1.4503205128205128, + "grad_norm": 0.7009061574935913, + "learning_rate": 0.00014201782003430417, + "loss": 0.8153, + "step": 8146 + }, + { + "epoch": 1.4504985754985755, + "grad_norm": 0.6502353549003601, + "learning_rate": 0.0001420051177340049, + "loss": 0.8959, + "step": 8147 + }, + { + "epoch": 1.4506766381766383, + "grad_norm": 0.6134430170059204, + "learning_rate": 0.00014199241461070261, + "loss": 0.9683, + "step": 8148 + }, + { + "epoch": 1.4508547008547008, + "grad_norm": 0.720160722732544, + "learning_rate": 0.0001419797106646462, + "loss": 0.9579, + "step": 8149 + }, + { + "epoch": 1.4510327635327636, + "grad_norm": 0.6141422986984253, + "learning_rate": 0.00014196700589608454, + "loss": 0.9427, + "step": 8150 + }, + { + "epoch": 1.451210826210826, + "grad_norm": 0.6835139393806458, + "learning_rate": 0.00014195430030526656, + "loss": 1.0374, + "step": 8151 + }, + { + "epoch": 1.4513888888888888, + "grad_norm": 0.6829691529273987, + "learning_rate": 0.00014194159389244128, + "loss": 0.9418, + "step": 8152 + }, + { + "epoch": 1.4515669515669516, + "grad_norm": 0.7142195701599121, + "learning_rate": 0.00014192888665785755, + "loss": 1.1876, + "step": 8153 + }, + { + "epoch": 1.4517450142450143, + "grad_norm": 0.6719943284988403, + "learning_rate": 0.0001419161786017644, + "loss": 1.1417, + "step": 8154 + }, + { + "epoch": 1.4519230769230769, + "grad_norm": 0.6478939652442932, + "learning_rate": 0.0001419034697244108, + "loss": 0.943, + "step": 8155 + }, + { + "epoch": 1.4521011396011396, + "grad_norm": 0.6308888792991638, + "learning_rate": 0.00014189076002604575, + "loss": 0.9842, + "step": 8156 + }, + { + "epoch": 1.4522792022792022, + "grad_norm": 0.673559844493866, + "learning_rate": 0.00014187804950691827, + "loss": 0.8108, + "step": 8157 + }, + { + "epoch": 1.452457264957265, + "grad_norm": 0.5895359516143799, + "learning_rate": 0.00014186533816727744, + "loss": 0.8187, + "step": 8158 + }, + { + "epoch": 1.4526353276353277, + "grad_norm": 0.6703287363052368, + "learning_rate": 0.00014185262600737225, + "loss": 0.9012, + "step": 8159 + }, + { + "epoch": 1.4528133903133904, + "grad_norm": 0.697728157043457, + "learning_rate": 0.00014183991302745182, + "loss": 1.2572, + "step": 8160 + }, + { + "epoch": 1.452991452991453, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00014182719922776514, + "loss": 1.078, + "step": 8161 + }, + { + "epoch": 1.4531695156695157, + "grad_norm": 0.6774863600730896, + "learning_rate": 0.00014181448460856143, + "loss": 1.0607, + "step": 8162 + }, + { + "epoch": 1.4533475783475782, + "grad_norm": 0.6872009038925171, + "learning_rate": 0.00014180176917008976, + "loss": 1.0713, + "step": 8163 + }, + { + "epoch": 1.453525641025641, + "grad_norm": 0.7949981093406677, + "learning_rate": 0.00014178905291259926, + "loss": 1.0471, + "step": 8164 + }, + { + "epoch": 1.4537037037037037, + "grad_norm": 0.6592127084732056, + "learning_rate": 0.00014177633583633908, + "loss": 0.8409, + "step": 8165 + }, + { + "epoch": 1.4538817663817665, + "grad_norm": 0.6745635867118835, + "learning_rate": 0.00014176361794155837, + "loss": 1.0859, + "step": 8166 + }, + { + "epoch": 1.454059829059829, + "grad_norm": 0.6661605834960938, + "learning_rate": 0.00014175089922850633, + "loss": 1.0587, + "step": 8167 + }, + { + "epoch": 1.4542378917378918, + "grad_norm": 0.6697571873664856, + "learning_rate": 0.00014173817969743212, + "loss": 0.8876, + "step": 8168 + }, + { + "epoch": 1.4544159544159543, + "grad_norm": 0.6162588000297546, + "learning_rate": 0.000141725459348585, + "loss": 0.9575, + "step": 8169 + }, + { + "epoch": 1.454594017094017, + "grad_norm": 0.6235088109970093, + "learning_rate": 0.00014171273818221422, + "loss": 0.9209, + "step": 8170 + }, + { + "epoch": 1.4547720797720798, + "grad_norm": 0.6744212508201599, + "learning_rate": 0.00014170001619856896, + "loss": 0.9704, + "step": 8171 + }, + { + "epoch": 1.4549501424501425, + "grad_norm": 0.6781345009803772, + "learning_rate": 0.0001416872933978985, + "loss": 1.1507, + "step": 8172 + }, + { + "epoch": 1.455128205128205, + "grad_norm": 0.7160060405731201, + "learning_rate": 0.0001416745697804521, + "loss": 1.2529, + "step": 8173 + }, + { + "epoch": 1.4553062678062678, + "grad_norm": 0.6742389798164368, + "learning_rate": 0.00014166184534647913, + "loss": 1.0168, + "step": 8174 + }, + { + "epoch": 1.4554843304843303, + "grad_norm": 0.6685828566551208, + "learning_rate": 0.0001416491200962288, + "loss": 1.0807, + "step": 8175 + }, + { + "epoch": 1.455662393162393, + "grad_norm": 0.6998327374458313, + "learning_rate": 0.0001416363940299505, + "loss": 1.1711, + "step": 8176 + }, + { + "epoch": 1.4558404558404558, + "grad_norm": 0.7132518291473389, + "learning_rate": 0.00014162366714789358, + "loss": 1.1392, + "step": 8177 + }, + { + "epoch": 1.4560185185185186, + "grad_norm": 0.6995887160301208, + "learning_rate": 0.0001416109394503073, + "loss": 1.3335, + "step": 8178 + }, + { + "epoch": 1.4561965811965811, + "grad_norm": 0.7161234021186829, + "learning_rate": 0.00014159821093744115, + "loss": 0.9725, + "step": 8179 + }, + { + "epoch": 1.4563746438746439, + "grad_norm": 0.7678874135017395, + "learning_rate": 0.00014158548160954446, + "loss": 1.1578, + "step": 8180 + }, + { + "epoch": 1.4565527065527066, + "grad_norm": 0.67372065782547, + "learning_rate": 0.00014157275146686662, + "loss": 1.0867, + "step": 8181 + }, + { + "epoch": 1.4567307692307692, + "grad_norm": 0.7757831811904907, + "learning_rate": 0.00014156002050965712, + "loss": 0.9768, + "step": 8182 + }, + { + "epoch": 1.456908831908832, + "grad_norm": 0.7174801230430603, + "learning_rate": 0.00014154728873816533, + "loss": 1.1712, + "step": 8183 + }, + { + "epoch": 1.4570868945868947, + "grad_norm": 0.5972673892974854, + "learning_rate": 0.0001415345561526407, + "loss": 0.9571, + "step": 8184 + }, + { + "epoch": 1.4572649572649572, + "grad_norm": 0.7999650835990906, + "learning_rate": 0.00014152182275333275, + "loss": 1.0583, + "step": 8185 + }, + { + "epoch": 1.45744301994302, + "grad_norm": 0.6737848520278931, + "learning_rate": 0.00014150908854049091, + "loss": 1.0562, + "step": 8186 + }, + { + "epoch": 1.4576210826210827, + "grad_norm": 0.7756418585777283, + "learning_rate": 0.00014149635351436474, + "loss": 1.2301, + "step": 8187 + }, + { + "epoch": 1.4577991452991452, + "grad_norm": 0.5633914470672607, + "learning_rate": 0.00014148361767520374, + "loss": 0.8847, + "step": 8188 + }, + { + "epoch": 1.457977207977208, + "grad_norm": 0.8462759256362915, + "learning_rate": 0.00014147088102325737, + "loss": 0.8046, + "step": 8189 + }, + { + "epoch": 1.4581552706552707, + "grad_norm": 0.7081632614135742, + "learning_rate": 0.00014145814355877526, + "loss": 1.0764, + "step": 8190 + }, + { + "epoch": 1.4583333333333333, + "grad_norm": 0.7357106804847717, + "learning_rate": 0.00014144540528200698, + "loss": 1.0202, + "step": 8191 + }, + { + "epoch": 1.458511396011396, + "grad_norm": 0.603566586971283, + "learning_rate": 0.00014143266619320204, + "loss": 0.8214, + "step": 8192 + }, + { + "epoch": 1.4586894586894588, + "grad_norm": 0.6829110383987427, + "learning_rate": 0.00014141992629261007, + "loss": 0.9479, + "step": 8193 + }, + { + "epoch": 1.4588675213675213, + "grad_norm": 0.6822739839553833, + "learning_rate": 0.00014140718558048072, + "loss": 0.9117, + "step": 8194 + }, + { + "epoch": 1.459045584045584, + "grad_norm": 0.7383607029914856, + "learning_rate": 0.00014139444405706356, + "loss": 0.9819, + "step": 8195 + }, + { + "epoch": 1.4592236467236468, + "grad_norm": 0.6319897770881653, + "learning_rate": 0.00014138170172260826, + "loss": 1.0508, + "step": 8196 + }, + { + "epoch": 1.4594017094017093, + "grad_norm": 0.6804461479187012, + "learning_rate": 0.0001413689585773645, + "loss": 0.992, + "step": 8197 + }, + { + "epoch": 1.459579772079772, + "grad_norm": 0.6198720335960388, + "learning_rate": 0.0001413562146215819, + "loss": 1.0113, + "step": 8198 + }, + { + "epoch": 1.4597578347578348, + "grad_norm": 0.5968540906906128, + "learning_rate": 0.0001413434698555102, + "loss": 0.7562, + "step": 8199 + }, + { + "epoch": 1.4599358974358974, + "grad_norm": 0.5370334982872009, + "learning_rate": 0.00014133072427939913, + "loss": 0.9238, + "step": 8200 + }, + { + "epoch": 1.46011396011396, + "grad_norm": 0.6652548909187317, + "learning_rate": 0.00014131797789349832, + "loss": 0.9464, + "step": 8201 + }, + { + "epoch": 1.4602920227920229, + "grad_norm": 0.637852668762207, + "learning_rate": 0.00014130523069805757, + "loss": 1.0395, + "step": 8202 + }, + { + "epoch": 1.4604700854700854, + "grad_norm": 0.8186550140380859, + "learning_rate": 0.00014129248269332664, + "loss": 1.2116, + "step": 8203 + }, + { + "epoch": 1.4606481481481481, + "grad_norm": 0.5290196537971497, + "learning_rate": 0.00014127973387955528, + "loss": 0.7331, + "step": 8204 + }, + { + "epoch": 1.460826210826211, + "grad_norm": 0.6516342163085938, + "learning_rate": 0.00014126698425699332, + "loss": 0.9275, + "step": 8205 + }, + { + "epoch": 1.4610042735042734, + "grad_norm": 0.767254114151001, + "learning_rate": 0.00014125423382589048, + "loss": 0.9355, + "step": 8206 + }, + { + "epoch": 1.4611823361823362, + "grad_norm": 0.6476777195930481, + "learning_rate": 0.00014124148258649668, + "loss": 0.9263, + "step": 8207 + }, + { + "epoch": 1.461360398860399, + "grad_norm": 0.6737871766090393, + "learning_rate": 0.00014122873053906167, + "loss": 0.9815, + "step": 8208 + }, + { + "epoch": 1.4615384615384617, + "grad_norm": 0.6311159729957581, + "learning_rate": 0.00014121597768383532, + "loss": 0.9607, + "step": 8209 + }, + { + "epoch": 1.4617165242165242, + "grad_norm": 0.6061250567436218, + "learning_rate": 0.00014120322402106752, + "loss": 0.7428, + "step": 8210 + }, + { + "epoch": 1.461894586894587, + "grad_norm": 0.6916252970695496, + "learning_rate": 0.00014119046955100815, + "loss": 0.9664, + "step": 8211 + }, + { + "epoch": 1.4620726495726495, + "grad_norm": 0.6583660840988159, + "learning_rate": 0.00014117771427390706, + "loss": 1.0645, + "step": 8212 + }, + { + "epoch": 1.4622507122507122, + "grad_norm": 0.7034604549407959, + "learning_rate": 0.00014116495819001425, + "loss": 0.9223, + "step": 8213 + }, + { + "epoch": 1.462428774928775, + "grad_norm": 0.6378605961799622, + "learning_rate": 0.00014115220129957954, + "loss": 0.7963, + "step": 8214 + }, + { + "epoch": 1.4626068376068377, + "grad_norm": 0.6251596212387085, + "learning_rate": 0.00014113944360285297, + "loss": 0.9852, + "step": 8215 + }, + { + "epoch": 1.4627849002849003, + "grad_norm": 0.7055560946464539, + "learning_rate": 0.00014112668510008446, + "loss": 0.9342, + "step": 8216 + }, + { + "epoch": 1.462962962962963, + "grad_norm": 0.6250377893447876, + "learning_rate": 0.00014111392579152396, + "loss": 0.9886, + "step": 8217 + }, + { + "epoch": 1.4631410256410255, + "grad_norm": 0.6011185050010681, + "learning_rate": 0.00014110116567742152, + "loss": 0.8465, + "step": 8218 + }, + { + "epoch": 1.4633190883190883, + "grad_norm": 0.6632489562034607, + "learning_rate": 0.0001410884047580271, + "loss": 0.8619, + "step": 8219 + }, + { + "epoch": 1.463497150997151, + "grad_norm": 0.7194828987121582, + "learning_rate": 0.00014107564303359076, + "loss": 1.1231, + "step": 8220 + }, + { + "epoch": 1.4636752136752138, + "grad_norm": 0.7640393376350403, + "learning_rate": 0.0001410628805043625, + "loss": 1.1955, + "step": 8221 + }, + { + "epoch": 1.4638532763532763, + "grad_norm": 0.9118906259536743, + "learning_rate": 0.0001410501171705924, + "loss": 1.0555, + "step": 8222 + }, + { + "epoch": 1.464031339031339, + "grad_norm": 0.7545066475868225, + "learning_rate": 0.00014103735303253053, + "loss": 0.9425, + "step": 8223 + }, + { + "epoch": 1.4642094017094016, + "grad_norm": 0.6848801970481873, + "learning_rate": 0.000141024588090427, + "loss": 1.0418, + "step": 8224 + }, + { + "epoch": 1.4643874643874644, + "grad_norm": 0.6825160384178162, + "learning_rate": 0.00014101182234453185, + "loss": 0.9615, + "step": 8225 + }, + { + "epoch": 1.4645655270655271, + "grad_norm": 0.8258556723594666, + "learning_rate": 0.00014099905579509527, + "loss": 1.1237, + "step": 8226 + }, + { + "epoch": 1.4647435897435899, + "grad_norm": 0.6427522897720337, + "learning_rate": 0.00014098628844236733, + "loss": 1.0853, + "step": 8227 + }, + { + "epoch": 1.4649216524216524, + "grad_norm": 0.6476351022720337, + "learning_rate": 0.00014097352028659825, + "loss": 1.1286, + "step": 8228 + }, + { + "epoch": 1.4650997150997151, + "grad_norm": 0.7621034383773804, + "learning_rate": 0.00014096075132803812, + "loss": 1.1402, + "step": 8229 + }, + { + "epoch": 1.4652777777777777, + "grad_norm": 0.6629892587661743, + "learning_rate": 0.00014094798156693718, + "loss": 0.7108, + "step": 8230 + }, + { + "epoch": 1.4654558404558404, + "grad_norm": 0.6902043223381042, + "learning_rate": 0.00014093521100354557, + "loss": 1.1761, + "step": 8231 + }, + { + "epoch": 1.4656339031339032, + "grad_norm": 0.7422910928726196, + "learning_rate": 0.00014092243963811357, + "loss": 0.867, + "step": 8232 + }, + { + "epoch": 1.465811965811966, + "grad_norm": 0.7424963712692261, + "learning_rate": 0.00014090966747089137, + "loss": 1.015, + "step": 8233 + }, + { + "epoch": 1.4659900284900285, + "grad_norm": 0.6855891942977905, + "learning_rate": 0.0001408968945021292, + "loss": 0.9624, + "step": 8234 + }, + { + "epoch": 1.4661680911680912, + "grad_norm": 0.5968918204307556, + "learning_rate": 0.00014088412073207736, + "loss": 0.9243, + "step": 8235 + }, + { + "epoch": 1.4663461538461537, + "grad_norm": 0.6153344511985779, + "learning_rate": 0.0001408713461609861, + "loss": 1.0305, + "step": 8236 + }, + { + "epoch": 1.4665242165242165, + "grad_norm": 0.6627458333969116, + "learning_rate": 0.0001408585707891057, + "loss": 1.1102, + "step": 8237 + }, + { + "epoch": 1.4667022792022792, + "grad_norm": 0.6475233435630798, + "learning_rate": 0.0001408457946166865, + "loss": 1.0045, + "step": 8238 + }, + { + "epoch": 1.466880341880342, + "grad_norm": 0.6792858839035034, + "learning_rate": 0.00014083301764397876, + "loss": 1.0092, + "step": 8239 + }, + { + "epoch": 1.4670584045584045, + "grad_norm": 0.6916255354881287, + "learning_rate": 0.00014082023987123293, + "loss": 1.0761, + "step": 8240 + }, + { + "epoch": 1.4672364672364673, + "grad_norm": 0.7901251912117004, + "learning_rate": 0.00014080746129869923, + "loss": 0.8002, + "step": 8241 + }, + { + "epoch": 1.4674145299145298, + "grad_norm": 0.8078263401985168, + "learning_rate": 0.00014079468192662812, + "loss": 0.9738, + "step": 8242 + }, + { + "epoch": 1.4675925925925926, + "grad_norm": 0.6370784640312195, + "learning_rate": 0.00014078190175526996, + "loss": 1.0256, + "step": 8243 + }, + { + "epoch": 1.4677706552706553, + "grad_norm": 0.6087532639503479, + "learning_rate": 0.0001407691207848752, + "loss": 0.9747, + "step": 8244 + }, + { + "epoch": 1.467948717948718, + "grad_norm": 0.6333357691764832, + "learning_rate": 0.00014075633901569414, + "loss": 1.0135, + "step": 8245 + }, + { + "epoch": 1.4681267806267806, + "grad_norm": 0.6914255619049072, + "learning_rate": 0.00014074355644797733, + "loss": 1.0261, + "step": 8246 + }, + { + "epoch": 1.4683048433048433, + "grad_norm": 0.6374734044075012, + "learning_rate": 0.00014073077308197513, + "loss": 0.9197, + "step": 8247 + }, + { + "epoch": 1.4684829059829059, + "grad_norm": 0.8023789525032043, + "learning_rate": 0.00014071798891793807, + "loss": 1.1085, + "step": 8248 + }, + { + "epoch": 1.4686609686609686, + "grad_norm": 0.7722933888435364, + "learning_rate": 0.0001407052039561166, + "loss": 1.2018, + "step": 8249 + }, + { + "epoch": 1.4688390313390314, + "grad_norm": 0.6823393106460571, + "learning_rate": 0.0001406924181967612, + "loss": 1.088, + "step": 8250 + }, + { + "epoch": 1.4690170940170941, + "grad_norm": 0.7037357687950134, + "learning_rate": 0.00014067963164012242, + "loss": 1.0324, + "step": 8251 + }, + { + "epoch": 1.4691951566951567, + "grad_norm": 0.6549737453460693, + "learning_rate": 0.00014066684428645074, + "loss": 1.152, + "step": 8252 + }, + { + "epoch": 1.4693732193732194, + "grad_norm": 0.5349790453910828, + "learning_rate": 0.00014065405613599674, + "loss": 0.6996, + "step": 8253 + }, + { + "epoch": 1.469551282051282, + "grad_norm": 0.6760679483413696, + "learning_rate": 0.00014064126718901096, + "loss": 0.9856, + "step": 8254 + }, + { + "epoch": 1.4697293447293447, + "grad_norm": 0.5912436842918396, + "learning_rate": 0.00014062847744574395, + "loss": 1.0076, + "step": 8255 + }, + { + "epoch": 1.4699074074074074, + "grad_norm": 0.75101637840271, + "learning_rate": 0.00014061568690644632, + "loss": 1.0033, + "step": 8256 + }, + { + "epoch": 1.4700854700854702, + "grad_norm": 0.6233504414558411, + "learning_rate": 0.00014060289557136873, + "loss": 0.8525, + "step": 8257 + }, + { + "epoch": 1.4702635327635327, + "grad_norm": 0.659570038318634, + "learning_rate": 0.00014059010344076171, + "loss": 0.855, + "step": 8258 + }, + { + "epoch": 1.4704415954415955, + "grad_norm": 0.8096539974212646, + "learning_rate": 0.00014057731051487593, + "loss": 0.9905, + "step": 8259 + }, + { + "epoch": 1.470619658119658, + "grad_norm": 0.5829728245735168, + "learning_rate": 0.00014056451679396204, + "loss": 0.7974, + "step": 8260 + }, + { + "epoch": 1.4707977207977208, + "grad_norm": 0.6176979541778564, + "learning_rate": 0.0001405517222782707, + "loss": 0.9556, + "step": 8261 + }, + { + "epoch": 1.4709757834757835, + "grad_norm": 0.6322479248046875, + "learning_rate": 0.00014053892696805264, + "loss": 0.8837, + "step": 8262 + }, + { + "epoch": 1.4711538461538463, + "grad_norm": 0.6886917948722839, + "learning_rate": 0.0001405261308635585, + "loss": 0.9242, + "step": 8263 + }, + { + "epoch": 1.4713319088319088, + "grad_norm": 0.7474521994590759, + "learning_rate": 0.00014051333396503901, + "loss": 0.9906, + "step": 8264 + }, + { + "epoch": 1.4715099715099715, + "grad_norm": 0.7120978832244873, + "learning_rate": 0.00014050053627274488, + "loss": 1.1074, + "step": 8265 + }, + { + "epoch": 1.471688034188034, + "grad_norm": 0.6778998374938965, + "learning_rate": 0.0001404877377869269, + "loss": 1.0027, + "step": 8266 + }, + { + "epoch": 1.4718660968660968, + "grad_norm": 0.6832901239395142, + "learning_rate": 0.0001404749385078358, + "loss": 0.9399, + "step": 8267 + }, + { + "epoch": 1.4720441595441596, + "grad_norm": 0.7428423762321472, + "learning_rate": 0.00014046213843572236, + "loss": 1.0591, + "step": 8268 + }, + { + "epoch": 1.4722222222222223, + "grad_norm": 0.7522720098495483, + "learning_rate": 0.00014044933757083737, + "loss": 1.1184, + "step": 8269 + }, + { + "epoch": 1.4724002849002849, + "grad_norm": 0.7714734673500061, + "learning_rate": 0.00014043653591343163, + "loss": 1.0783, + "step": 8270 + }, + { + "epoch": 1.4725783475783476, + "grad_norm": 0.5860890746116638, + "learning_rate": 0.00014042373346375597, + "loss": 0.8394, + "step": 8271 + }, + { + "epoch": 1.4727564102564101, + "grad_norm": 0.6400395035743713, + "learning_rate": 0.0001404109302220612, + "loss": 0.9153, + "step": 8272 + }, + { + "epoch": 1.4729344729344729, + "grad_norm": 0.7441139817237854, + "learning_rate": 0.00014039812618859827, + "loss": 0.9224, + "step": 8273 + }, + { + "epoch": 1.4731125356125356, + "grad_norm": 0.6030932664871216, + "learning_rate": 0.00014038532136361793, + "loss": 1.0783, + "step": 8274 + }, + { + "epoch": 1.4732905982905984, + "grad_norm": 0.7243345975875854, + "learning_rate": 0.0001403725157473711, + "loss": 0.9894, + "step": 8275 + }, + { + "epoch": 1.473468660968661, + "grad_norm": 0.6880641579627991, + "learning_rate": 0.0001403597093401087, + "loss": 0.9459, + "step": 8276 + }, + { + "epoch": 1.4736467236467237, + "grad_norm": 0.6263882517814636, + "learning_rate": 0.00014034690214208165, + "loss": 0.8781, + "step": 8277 + }, + { + "epoch": 1.4738247863247862, + "grad_norm": 0.7159495949745178, + "learning_rate": 0.00014033409415354085, + "loss": 1.0511, + "step": 8278 + }, + { + "epoch": 1.474002849002849, + "grad_norm": 0.7182226181030273, + "learning_rate": 0.00014032128537473727, + "loss": 1.1196, + "step": 8279 + }, + { + "epoch": 1.4741809116809117, + "grad_norm": 0.744478166103363, + "learning_rate": 0.00014030847580592186, + "loss": 1.0747, + "step": 8280 + }, + { + "epoch": 1.4743589743589745, + "grad_norm": 0.6806797385215759, + "learning_rate": 0.00014029566544734558, + "loss": 1.1519, + "step": 8281 + }, + { + "epoch": 1.474537037037037, + "grad_norm": 0.6813502311706543, + "learning_rate": 0.00014028285429925946, + "loss": 0.968, + "step": 8282 + }, + { + "epoch": 1.4747150997150997, + "grad_norm": 0.639784574508667, + "learning_rate": 0.00014027004236191452, + "loss": 1.0685, + "step": 8283 + }, + { + "epoch": 1.4748931623931623, + "grad_norm": 0.6325878500938416, + "learning_rate": 0.00014025722963556173, + "loss": 1.0358, + "step": 8284 + }, + { + "epoch": 1.475071225071225, + "grad_norm": 0.7012955546379089, + "learning_rate": 0.00014024441612045215, + "loss": 1.1059, + "step": 8285 + }, + { + "epoch": 1.4752492877492878, + "grad_norm": 0.690380334854126, + "learning_rate": 0.00014023160181683684, + "loss": 0.9628, + "step": 8286 + }, + { + "epoch": 1.4754273504273505, + "grad_norm": 0.7178516983985901, + "learning_rate": 0.00014021878672496686, + "loss": 0.963, + "step": 8287 + }, + { + "epoch": 1.475605413105413, + "grad_norm": 0.7049064636230469, + "learning_rate": 0.0001402059708450933, + "loss": 0.8996, + "step": 8288 + }, + { + "epoch": 1.4757834757834758, + "grad_norm": 0.6777819395065308, + "learning_rate": 0.00014019315417746728, + "loss": 1.0696, + "step": 8289 + }, + { + "epoch": 1.4759615384615383, + "grad_norm": 0.5948763489723206, + "learning_rate": 0.00014018033672233987, + "loss": 0.928, + "step": 8290 + }, + { + "epoch": 1.476139601139601, + "grad_norm": 0.7183942198753357, + "learning_rate": 0.00014016751847996224, + "loss": 1.1053, + "step": 8291 + }, + { + "epoch": 1.4763176638176638, + "grad_norm": 0.7426177263259888, + "learning_rate": 0.00014015469945058556, + "loss": 0.9504, + "step": 8292 + }, + { + "epoch": 1.4764957264957266, + "grad_norm": 0.6508159041404724, + "learning_rate": 0.0001401418796344609, + "loss": 1.1176, + "step": 8293 + }, + { + "epoch": 1.476673789173789, + "grad_norm": 0.6954567432403564, + "learning_rate": 0.00014012905903183954, + "loss": 0.9238, + "step": 8294 + }, + { + "epoch": 1.4768518518518519, + "grad_norm": 0.7023960947990417, + "learning_rate": 0.0001401162376429726, + "loss": 1.2032, + "step": 8295 + }, + { + "epoch": 1.4770299145299146, + "grad_norm": 0.7174739837646484, + "learning_rate": 0.00014010341546811134, + "loss": 0.9385, + "step": 8296 + }, + { + "epoch": 1.4772079772079771, + "grad_norm": 0.611980140209198, + "learning_rate": 0.00014009059250750695, + "loss": 0.9469, + "step": 8297 + }, + { + "epoch": 1.47738603988604, + "grad_norm": 0.6362917423248291, + "learning_rate": 0.0001400777687614107, + "loss": 1.1406, + "step": 8298 + }, + { + "epoch": 1.4775641025641026, + "grad_norm": 0.6884697675704956, + "learning_rate": 0.00014006494423007381, + "loss": 0.7915, + "step": 8299 + }, + { + "epoch": 1.4777421652421652, + "grad_norm": 0.6266025304794312, + "learning_rate": 0.00014005211891374755, + "loss": 0.94, + "step": 8300 + }, + { + "epoch": 1.477920227920228, + "grad_norm": 0.6130280494689941, + "learning_rate": 0.00014003929281268323, + "loss": 0.9369, + "step": 8301 + }, + { + "epoch": 1.4780982905982907, + "grad_norm": 0.7244207859039307, + "learning_rate": 0.00014002646592713215, + "loss": 1.1449, + "step": 8302 + }, + { + "epoch": 1.4782763532763532, + "grad_norm": 0.6527345776557922, + "learning_rate": 0.0001400136382573456, + "loss": 0.7792, + "step": 8303 + }, + { + "epoch": 1.478454415954416, + "grad_norm": 0.7102689743041992, + "learning_rate": 0.00014000080980357496, + "loss": 0.9577, + "step": 8304 + }, + { + "epoch": 1.4786324786324787, + "grad_norm": 0.6179325580596924, + "learning_rate": 0.00013998798056607154, + "loss": 0.827, + "step": 8305 + }, + { + "epoch": 1.4788105413105412, + "grad_norm": 0.761234700679779, + "learning_rate": 0.00013997515054508668, + "loss": 1.0576, + "step": 8306 + }, + { + "epoch": 1.478988603988604, + "grad_norm": 0.6200914978981018, + "learning_rate": 0.0001399623197408718, + "loss": 1.0514, + "step": 8307 + }, + { + "epoch": 1.4791666666666667, + "grad_norm": 0.5961193442344666, + "learning_rate": 0.0001399494881536783, + "loss": 0.7846, + "step": 8308 + }, + { + "epoch": 1.4793447293447293, + "grad_norm": 0.645984411239624, + "learning_rate": 0.00013993665578375758, + "loss": 0.9927, + "step": 8309 + }, + { + "epoch": 1.479522792022792, + "grad_norm": 0.7258989810943604, + "learning_rate": 0.000139923822631361, + "loss": 0.7567, + "step": 8310 + }, + { + "epoch": 1.4797008547008548, + "grad_norm": 0.708882212638855, + "learning_rate": 0.00013991098869674007, + "loss": 1.1147, + "step": 8311 + }, + { + "epoch": 1.4798789173789173, + "grad_norm": 0.669262707233429, + "learning_rate": 0.00013989815398014624, + "loss": 0.7142, + "step": 8312 + }, + { + "epoch": 1.48005698005698, + "grad_norm": 0.7398767471313477, + "learning_rate": 0.00013988531848183096, + "loss": 1.043, + "step": 8313 + }, + { + "epoch": 1.4802350427350428, + "grad_norm": 0.753197193145752, + "learning_rate": 0.0001398724822020457, + "loss": 1.058, + "step": 8314 + }, + { + "epoch": 1.4804131054131053, + "grad_norm": 0.663526177406311, + "learning_rate": 0.000139859645141042, + "loss": 1.1272, + "step": 8315 + }, + { + "epoch": 1.480591168091168, + "grad_norm": 0.6537514925003052, + "learning_rate": 0.00013984680729907135, + "loss": 1.011, + "step": 8316 + }, + { + "epoch": 1.4807692307692308, + "grad_norm": 0.707554817199707, + "learning_rate": 0.00013983396867638527, + "loss": 1.0593, + "step": 8317 + }, + { + "epoch": 1.4809472934472934, + "grad_norm": 0.6261475086212158, + "learning_rate": 0.00013982112927323533, + "loss": 1.0731, + "step": 8318 + }, + { + "epoch": 1.4811253561253561, + "grad_norm": 0.6694258451461792, + "learning_rate": 0.00013980828908987308, + "loss": 1.0703, + "step": 8319 + }, + { + "epoch": 1.4813034188034189, + "grad_norm": 0.7793164253234863, + "learning_rate": 0.00013979544812655012, + "loss": 1.0447, + "step": 8320 + }, + { + "epoch": 1.4814814814814814, + "grad_norm": 0.6496448516845703, + "learning_rate": 0.00013978260638351802, + "loss": 1.0208, + "step": 8321 + }, + { + "epoch": 1.4816595441595442, + "grad_norm": 0.5992059111595154, + "learning_rate": 0.00013976976386102834, + "loss": 0.9717, + "step": 8322 + }, + { + "epoch": 1.481837606837607, + "grad_norm": 0.7473567128181458, + "learning_rate": 0.0001397569205593328, + "loss": 0.9612, + "step": 8323 + }, + { + "epoch": 1.4820156695156697, + "grad_norm": 0.657558798789978, + "learning_rate": 0.00013974407647868297, + "loss": 1.2137, + "step": 8324 + }, + { + "epoch": 1.4821937321937322, + "grad_norm": 0.7040614485740662, + "learning_rate": 0.00013973123161933055, + "loss": 1.007, + "step": 8325 + }, + { + "epoch": 1.482371794871795, + "grad_norm": 0.6098681092262268, + "learning_rate": 0.00013971838598152717, + "loss": 1.0595, + "step": 8326 + }, + { + "epoch": 1.4825498575498575, + "grad_norm": 0.7194869518280029, + "learning_rate": 0.0001397055395655245, + "loss": 0.9632, + "step": 8327 + }, + { + "epoch": 1.4827279202279202, + "grad_norm": 0.645972728729248, + "learning_rate": 0.00013969269237157426, + "loss": 1.0712, + "step": 8328 + }, + { + "epoch": 1.482905982905983, + "grad_norm": 0.6580560207366943, + "learning_rate": 0.0001396798443999282, + "loss": 1.2117, + "step": 8329 + }, + { + "epoch": 1.4830840455840457, + "grad_norm": 0.6624418497085571, + "learning_rate": 0.00013966699565083802, + "loss": 0.8529, + "step": 8330 + }, + { + "epoch": 1.4832621082621082, + "grad_norm": 0.659896731376648, + "learning_rate": 0.00013965414612455545, + "loss": 0.9359, + "step": 8331 + }, + { + "epoch": 1.483440170940171, + "grad_norm": 0.6690883636474609, + "learning_rate": 0.00013964129582133222, + "loss": 0.971, + "step": 8332 + }, + { + "epoch": 1.4836182336182335, + "grad_norm": 0.6767334938049316, + "learning_rate": 0.00013962844474142022, + "loss": 1.0137, + "step": 8333 + }, + { + "epoch": 1.4837962962962963, + "grad_norm": 0.6412752270698547, + "learning_rate": 0.0001396155928850711, + "loss": 1.2812, + "step": 8334 + }, + { + "epoch": 1.483974358974359, + "grad_norm": 0.6731469035148621, + "learning_rate": 0.0001396027402525368, + "loss": 0.8723, + "step": 8335 + }, + { + "epoch": 1.4841524216524218, + "grad_norm": 0.7327923774719238, + "learning_rate": 0.000139589886844069, + "loss": 0.9606, + "step": 8336 + }, + { + "epoch": 1.4843304843304843, + "grad_norm": 0.6194515824317932, + "learning_rate": 0.00013957703265991963, + "loss": 0.8514, + "step": 8337 + }, + { + "epoch": 1.484508547008547, + "grad_norm": 0.7250012755393982, + "learning_rate": 0.00013956417770034053, + "loss": 0.9755, + "step": 8338 + }, + { + "epoch": 1.4846866096866096, + "grad_norm": 0.7484263181686401, + "learning_rate": 0.00013955132196558358, + "loss": 1.0376, + "step": 8339 + }, + { + "epoch": 1.4848646723646723, + "grad_norm": 0.7593362331390381, + "learning_rate": 0.00013953846545590058, + "loss": 1.3011, + "step": 8340 + }, + { + "epoch": 1.485042735042735, + "grad_norm": 0.6670466065406799, + "learning_rate": 0.00013952560817154352, + "loss": 0.9726, + "step": 8341 + }, + { + "epoch": 1.4852207977207978, + "grad_norm": 0.8001134395599365, + "learning_rate": 0.00013951275011276425, + "loss": 1.1447, + "step": 8342 + }, + { + "epoch": 1.4853988603988604, + "grad_norm": 0.741450309753418, + "learning_rate": 0.00013949989127981475, + "loss": 1.1101, + "step": 8343 + }, + { + "epoch": 1.4855769230769231, + "grad_norm": 0.6594467163085938, + "learning_rate": 0.00013948703167294694, + "loss": 1.0205, + "step": 8344 + }, + { + "epoch": 1.4857549857549857, + "grad_norm": 0.6303030252456665, + "learning_rate": 0.00013947417129241276, + "loss": 0.9179, + "step": 8345 + }, + { + "epoch": 1.4859330484330484, + "grad_norm": 0.6352720856666565, + "learning_rate": 0.00013946131013846418, + "loss": 1.158, + "step": 8346 + }, + { + "epoch": 1.4861111111111112, + "grad_norm": 0.6720923781394958, + "learning_rate": 0.0001394484482113532, + "loss": 0.8805, + "step": 8347 + }, + { + "epoch": 1.486289173789174, + "grad_norm": 0.7186421751976013, + "learning_rate": 0.00013943558551133186, + "loss": 0.8951, + "step": 8348 + }, + { + "epoch": 1.4864672364672364, + "grad_norm": 0.6038698554039001, + "learning_rate": 0.00013942272203865214, + "loss": 1.0079, + "step": 8349 + }, + { + "epoch": 1.4866452991452992, + "grad_norm": 0.665790319442749, + "learning_rate": 0.00013940985779356606, + "loss": 0.8853, + "step": 8350 + }, + { + "epoch": 1.4868233618233617, + "grad_norm": 0.6941595673561096, + "learning_rate": 0.00013939699277632568, + "loss": 1.1404, + "step": 8351 + }, + { + "epoch": 1.4870014245014245, + "grad_norm": 0.7943871021270752, + "learning_rate": 0.00013938412698718305, + "loss": 0.9961, + "step": 8352 + }, + { + "epoch": 1.4871794871794872, + "grad_norm": 0.6363818645477295, + "learning_rate": 0.00013937126042639028, + "loss": 0.8621, + "step": 8353 + }, + { + "epoch": 1.48735754985755, + "grad_norm": 0.7986421585083008, + "learning_rate": 0.00013935839309419943, + "loss": 1.0547, + "step": 8354 + }, + { + "epoch": 1.4875356125356125, + "grad_norm": 0.5890130400657654, + "learning_rate": 0.00013934552499086266, + "loss": 0.9863, + "step": 8355 + }, + { + "epoch": 1.4877136752136753, + "grad_norm": 0.7915370464324951, + "learning_rate": 0.00013933265611663207, + "loss": 1.0385, + "step": 8356 + }, + { + "epoch": 1.4878917378917378, + "grad_norm": 0.7062503695487976, + "learning_rate": 0.00013931978647175973, + "loss": 1.0984, + "step": 8357 + }, + { + "epoch": 1.4880698005698005, + "grad_norm": 0.6496769785881042, + "learning_rate": 0.00013930691605649792, + "loss": 1.0884, + "step": 8358 + }, + { + "epoch": 1.4882478632478633, + "grad_norm": 0.6527266502380371, + "learning_rate": 0.0001392940448710987, + "loss": 1.0366, + "step": 8359 + }, + { + "epoch": 1.488425925925926, + "grad_norm": 0.6269870400428772, + "learning_rate": 0.00013928117291581431, + "loss": 0.9097, + "step": 8360 + }, + { + "epoch": 1.4886039886039886, + "grad_norm": 0.6581160426139832, + "learning_rate": 0.00013926830019089694, + "loss": 0.8694, + "step": 8361 + }, + { + "epoch": 1.4887820512820513, + "grad_norm": 0.6196219325065613, + "learning_rate": 0.0001392554266965988, + "loss": 0.8054, + "step": 8362 + }, + { + "epoch": 1.4889601139601139, + "grad_norm": 0.6246176362037659, + "learning_rate": 0.0001392425524331721, + "loss": 0.9309, + "step": 8363 + }, + { + "epoch": 1.4891381766381766, + "grad_norm": 0.7293874025344849, + "learning_rate": 0.00013922967740086914, + "loss": 1.051, + "step": 8364 + }, + { + "epoch": 1.4893162393162394, + "grad_norm": 0.6581604480743408, + "learning_rate": 0.00013921680159994213, + "loss": 0.8475, + "step": 8365 + }, + { + "epoch": 1.489494301994302, + "grad_norm": 0.6294612288475037, + "learning_rate": 0.00013920392503064335, + "loss": 0.6946, + "step": 8366 + }, + { + "epoch": 1.4896723646723646, + "grad_norm": 0.5725370645523071, + "learning_rate": 0.00013919104769322512, + "loss": 0.7838, + "step": 8367 + }, + { + "epoch": 1.4898504273504274, + "grad_norm": 0.681520402431488, + "learning_rate": 0.00013917816958793967, + "loss": 0.99, + "step": 8368 + }, + { + "epoch": 1.49002849002849, + "grad_norm": 0.6660219430923462, + "learning_rate": 0.00013916529071503943, + "loss": 0.9113, + "step": 8369 + }, + { + "epoch": 1.4902065527065527, + "grad_norm": 0.7567862272262573, + "learning_rate": 0.00013915241107477665, + "loss": 1.2498, + "step": 8370 + }, + { + "epoch": 1.4903846153846154, + "grad_norm": 0.7366036176681519, + "learning_rate": 0.00013913953066740372, + "loss": 1.115, + "step": 8371 + }, + { + "epoch": 1.4905626780626782, + "grad_norm": 0.6201434135437012, + "learning_rate": 0.00013912664949317297, + "loss": 0.8447, + "step": 8372 + }, + { + "epoch": 1.4907407407407407, + "grad_norm": 0.7618655562400818, + "learning_rate": 0.00013911376755233683, + "loss": 0.9696, + "step": 8373 + }, + { + "epoch": 1.4909188034188035, + "grad_norm": 0.6716726422309875, + "learning_rate": 0.00013910088484514764, + "loss": 0.9753, + "step": 8374 + }, + { + "epoch": 1.491096866096866, + "grad_norm": 0.6745659112930298, + "learning_rate": 0.0001390880013718579, + "loss": 1.134, + "step": 8375 + }, + { + "epoch": 1.4912749287749287, + "grad_norm": 0.7524410486221313, + "learning_rate": 0.0001390751171327199, + "loss": 1.0235, + "step": 8376 + }, + { + "epoch": 1.4914529914529915, + "grad_norm": 0.7409411072731018, + "learning_rate": 0.00013906223212798615, + "loss": 0.752, + "step": 8377 + }, + { + "epoch": 1.4916310541310542, + "grad_norm": 0.7016384601593018, + "learning_rate": 0.00013904934635790913, + "loss": 1.1712, + "step": 8378 + }, + { + "epoch": 1.4918091168091168, + "grad_norm": 0.6537824869155884, + "learning_rate": 0.00013903645982274129, + "loss": 1.1162, + "step": 8379 + }, + { + "epoch": 1.4919871794871795, + "grad_norm": 0.6460806727409363, + "learning_rate": 0.0001390235725227351, + "loss": 0.9389, + "step": 8380 + }, + { + "epoch": 1.492165242165242, + "grad_norm": 0.6405501365661621, + "learning_rate": 0.0001390106844581431, + "loss": 1.0508, + "step": 8381 + }, + { + "epoch": 1.4923433048433048, + "grad_norm": 0.6672594547271729, + "learning_rate": 0.00013899779562921775, + "loss": 1.0018, + "step": 8382 + }, + { + "epoch": 1.4925213675213675, + "grad_norm": 0.6303185820579529, + "learning_rate": 0.0001389849060362116, + "loss": 0.9964, + "step": 8383 + }, + { + "epoch": 1.4926994301994303, + "grad_norm": 0.6981508731842041, + "learning_rate": 0.00013897201567937719, + "loss": 1.174, + "step": 8384 + }, + { + "epoch": 1.4928774928774928, + "grad_norm": 0.6195989847183228, + "learning_rate": 0.0001389591245589671, + "loss": 0.9254, + "step": 8385 + }, + { + "epoch": 1.4930555555555556, + "grad_norm": 0.6232163310050964, + "learning_rate": 0.00013894623267523393, + "loss": 0.7151, + "step": 8386 + }, + { + "epoch": 1.493233618233618, + "grad_norm": 0.673067033290863, + "learning_rate": 0.0001389333400284302, + "loss": 1.0156, + "step": 8387 + }, + { + "epoch": 1.4934116809116809, + "grad_norm": 0.706266462802887, + "learning_rate": 0.00013892044661880856, + "loss": 0.9387, + "step": 8388 + }, + { + "epoch": 1.4935897435897436, + "grad_norm": 0.742640495300293, + "learning_rate": 0.00013890755244662161, + "loss": 1.1597, + "step": 8389 + }, + { + "epoch": 1.4937678062678064, + "grad_norm": 0.6856846809387207, + "learning_rate": 0.000138894657512122, + "loss": 0.9998, + "step": 8390 + }, + { + "epoch": 1.493945868945869, + "grad_norm": 0.7214110493659973, + "learning_rate": 0.0001388817618155624, + "loss": 1.1867, + "step": 8391 + }, + { + "epoch": 1.4941239316239316, + "grad_norm": 0.7346787452697754, + "learning_rate": 0.0001388688653571954, + "loss": 0.9071, + "step": 8392 + }, + { + "epoch": 1.4943019943019942, + "grad_norm": 0.7019181847572327, + "learning_rate": 0.00013885596813727373, + "loss": 1.0472, + "step": 8393 + }, + { + "epoch": 1.494480056980057, + "grad_norm": 0.6780814528465271, + "learning_rate": 0.00013884307015605012, + "loss": 1.0031, + "step": 8394 + }, + { + "epoch": 1.4946581196581197, + "grad_norm": 0.6722873449325562, + "learning_rate": 0.0001388301714137772, + "loss": 0.8889, + "step": 8395 + }, + { + "epoch": 1.4948361823361824, + "grad_norm": 0.6736134886741638, + "learning_rate": 0.00013881727191070777, + "loss": 0.8695, + "step": 8396 + }, + { + "epoch": 1.495014245014245, + "grad_norm": 0.632648766040802, + "learning_rate": 0.00013880437164709452, + "loss": 0.9391, + "step": 8397 + }, + { + "epoch": 1.4951923076923077, + "grad_norm": 0.7004299163818359, + "learning_rate": 0.0001387914706231902, + "loss": 1.1423, + "step": 8398 + }, + { + "epoch": 1.4953703703703702, + "grad_norm": 0.5787134766578674, + "learning_rate": 0.0001387785688392476, + "loss": 0.9953, + "step": 8399 + }, + { + "epoch": 1.495548433048433, + "grad_norm": 0.6671785712242126, + "learning_rate": 0.0001387656662955195, + "loss": 0.9356, + "step": 8400 + }, + { + "epoch": 1.4957264957264957, + "grad_norm": 0.7216096520423889, + "learning_rate": 0.0001387527629922587, + "loss": 0.9065, + "step": 8401 + }, + { + "epoch": 1.4959045584045585, + "grad_norm": 0.6469849348068237, + "learning_rate": 0.00013873985892971801, + "loss": 1.0664, + "step": 8402 + }, + { + "epoch": 1.496082621082621, + "grad_norm": 0.5598217248916626, + "learning_rate": 0.00013872695410815027, + "loss": 0.8834, + "step": 8403 + }, + { + "epoch": 1.4962606837606838, + "grad_norm": 0.6860302686691284, + "learning_rate": 0.00013871404852780828, + "loss": 0.9061, + "step": 8404 + }, + { + "epoch": 1.4964387464387463, + "grad_norm": 0.7101688385009766, + "learning_rate": 0.00013870114218894497, + "loss": 1.0236, + "step": 8405 + }, + { + "epoch": 1.496616809116809, + "grad_norm": 0.6494225859642029, + "learning_rate": 0.00013868823509181313, + "loss": 0.9631, + "step": 8406 + }, + { + "epoch": 1.4967948717948718, + "grad_norm": 0.6804189085960388, + "learning_rate": 0.00013867532723666574, + "loss": 0.9341, + "step": 8407 + }, + { + "epoch": 1.4969729344729346, + "grad_norm": 0.8493942022323608, + "learning_rate": 0.00013866241862375562, + "loss": 1.1451, + "step": 8408 + }, + { + "epoch": 1.497150997150997, + "grad_norm": 0.6248497366905212, + "learning_rate": 0.00013864950925333576, + "loss": 0.8584, + "step": 8409 + }, + { + "epoch": 1.4973290598290598, + "grad_norm": 0.6238769292831421, + "learning_rate": 0.00013863659912565903, + "loss": 1.1612, + "step": 8410 + }, + { + "epoch": 1.4975071225071226, + "grad_norm": 0.8538609147071838, + "learning_rate": 0.0001386236882409784, + "loss": 1.0817, + "step": 8411 + }, + { + "epoch": 1.4976851851851851, + "grad_norm": 0.7301406264305115, + "learning_rate": 0.00013861077659954683, + "loss": 0.943, + "step": 8412 + }, + { + "epoch": 1.4978632478632479, + "grad_norm": 0.6573456525802612, + "learning_rate": 0.0001385978642016173, + "loss": 1.0154, + "step": 8413 + }, + { + "epoch": 1.4980413105413106, + "grad_norm": 0.7634185552597046, + "learning_rate": 0.0001385849510474428, + "loss": 1.0432, + "step": 8414 + }, + { + "epoch": 1.4982193732193732, + "grad_norm": 0.6156686544418335, + "learning_rate": 0.00013857203713727633, + "loss": 1.0442, + "step": 8415 + }, + { + "epoch": 1.498397435897436, + "grad_norm": 0.5386871695518494, + "learning_rate": 0.00013855912247137092, + "loss": 0.9055, + "step": 8416 + }, + { + "epoch": 1.4985754985754987, + "grad_norm": 0.7108574509620667, + "learning_rate": 0.00013854620704997962, + "loss": 0.9705, + "step": 8417 + }, + { + "epoch": 1.4987535612535612, + "grad_norm": 0.7313347458839417, + "learning_rate": 0.00013853329087335547, + "loss": 0.7541, + "step": 8418 + }, + { + "epoch": 1.498931623931624, + "grad_norm": 0.8369119167327881, + "learning_rate": 0.0001385203739417515, + "loss": 1.1317, + "step": 8419 + }, + { + "epoch": 1.4991096866096867, + "grad_norm": 0.6763789057731628, + "learning_rate": 0.00013850745625542085, + "loss": 0.7909, + "step": 8420 + }, + { + "epoch": 1.4992877492877492, + "grad_norm": 0.7369635105133057, + "learning_rate": 0.00013849453781461656, + "loss": 1.1454, + "step": 8421 + }, + { + "epoch": 1.499465811965812, + "grad_norm": 0.7165971398353577, + "learning_rate": 0.0001384816186195918, + "loss": 1.1927, + "step": 8422 + }, + { + "epoch": 1.4996438746438747, + "grad_norm": 0.7502337694168091, + "learning_rate": 0.00013846869867059966, + "loss": 1.0592, + "step": 8423 + }, + { + "epoch": 1.4998219373219372, + "grad_norm": 0.7207813858985901, + "learning_rate": 0.00013845577796789326, + "loss": 1.1133, + "step": 8424 + }, + { + "epoch": 1.4998219373219372, + "eval_loss": 1.1057652235031128, + "eval_runtime": 24.7975, + "eval_samples_per_second": 41.98, + "eval_steps_per_second": 21.01, + "step": 8424 + }, + { + "epoch": 1.5, + "grad_norm": 0.6962727308273315, + "learning_rate": 0.00013844285651172576, + "loss": 1.0711, + "step": 8425 + }, + { + "epoch": 1.5001780626780628, + "grad_norm": 0.6585133075714111, + "learning_rate": 0.00013842993430235038, + "loss": 0.9793, + "step": 8426 + }, + { + "epoch": 1.5003561253561255, + "grad_norm": 0.7045056819915771, + "learning_rate": 0.00013841701134002029, + "loss": 1.0046, + "step": 8427 + }, + { + "epoch": 1.500534188034188, + "grad_norm": 0.6788702011108398, + "learning_rate": 0.00013840408762498863, + "loss": 0.9539, + "step": 8428 + }, + { + "epoch": 1.5007122507122506, + "grad_norm": 0.7253114581108093, + "learning_rate": 0.00013839116315750863, + "loss": 0.9446, + "step": 8429 + }, + { + "epoch": 1.5008903133903133, + "grad_norm": 0.6103765368461609, + "learning_rate": 0.0001383782379378336, + "loss": 0.7862, + "step": 8430 + }, + { + "epoch": 1.501068376068376, + "grad_norm": 0.6662353873252869, + "learning_rate": 0.00013836531196621666, + "loss": 1.2178, + "step": 8431 + }, + { + "epoch": 1.5012464387464388, + "grad_norm": 0.6871803998947144, + "learning_rate": 0.00013835238524291117, + "loss": 0.9263, + "step": 8432 + }, + { + "epoch": 1.5014245014245016, + "grad_norm": 0.62713223695755, + "learning_rate": 0.00013833945776817034, + "loss": 0.8879, + "step": 8433 + }, + { + "epoch": 1.501602564102564, + "grad_norm": 0.6698164343833923, + "learning_rate": 0.00013832652954224748, + "loss": 0.9847, + "step": 8434 + }, + { + "epoch": 1.5017806267806266, + "grad_norm": 0.6855883002281189, + "learning_rate": 0.0001383136005653959, + "loss": 0.8614, + "step": 8435 + }, + { + "epoch": 1.5019586894586894, + "grad_norm": 0.7028802037239075, + "learning_rate": 0.0001383006708378689, + "loss": 1.0153, + "step": 8436 + }, + { + "epoch": 1.5021367521367521, + "grad_norm": 0.6710380911827087, + "learning_rate": 0.00013828774035991981, + "loss": 1.0163, + "step": 8437 + }, + { + "epoch": 1.5023148148148149, + "grad_norm": 0.618984580039978, + "learning_rate": 0.000138274809131802, + "loss": 1.0015, + "step": 8438 + }, + { + "epoch": 1.5024928774928776, + "grad_norm": 0.6881645321846008, + "learning_rate": 0.00013826187715376882, + "loss": 0.9776, + "step": 8439 + }, + { + "epoch": 1.5026709401709402, + "grad_norm": 0.6715859770774841, + "learning_rate": 0.00013824894442607358, + "loss": 0.9129, + "step": 8440 + }, + { + "epoch": 1.5028490028490027, + "grad_norm": 0.5940943360328674, + "learning_rate": 0.0001382360109489698, + "loss": 1.0724, + "step": 8441 + }, + { + "epoch": 1.5030270655270654, + "grad_norm": 0.6536458134651184, + "learning_rate": 0.0001382230767227108, + "loss": 1.0162, + "step": 8442 + }, + { + "epoch": 1.5032051282051282, + "grad_norm": 0.6163156628608704, + "learning_rate": 0.00013821014174755, + "loss": 1.0521, + "step": 8443 + }, + { + "epoch": 1.503383190883191, + "grad_norm": 0.7592282891273499, + "learning_rate": 0.00013819720602374082, + "loss": 0.9525, + "step": 8444 + }, + { + "epoch": 1.5035612535612537, + "grad_norm": 0.6672595143318176, + "learning_rate": 0.0001381842695515368, + "loss": 0.9359, + "step": 8445 + }, + { + "epoch": 1.5037393162393162, + "grad_norm": 0.6395034193992615, + "learning_rate": 0.0001381713323311913, + "loss": 1.166, + "step": 8446 + }, + { + "epoch": 1.5039173789173788, + "grad_norm": 0.5958148837089539, + "learning_rate": 0.00013815839436295783, + "loss": 0.9885, + "step": 8447 + }, + { + "epoch": 1.5040954415954415, + "grad_norm": 0.676555871963501, + "learning_rate": 0.0001381454556470899, + "loss": 1.0637, + "step": 8448 + }, + { + "epoch": 1.5042735042735043, + "grad_norm": 0.642428994178772, + "learning_rate": 0.00013813251618384102, + "loss": 0.9288, + "step": 8449 + }, + { + "epoch": 1.504451566951567, + "grad_norm": 0.6730920076370239, + "learning_rate": 0.00013811957597346467, + "loss": 1.1345, + "step": 8450 + }, + { + "epoch": 1.5046296296296298, + "grad_norm": 0.7824259996414185, + "learning_rate": 0.00013810663501621443, + "loss": 0.7532, + "step": 8451 + }, + { + "epoch": 1.5048076923076923, + "grad_norm": 0.8184825778007507, + "learning_rate": 0.00013809369331234386, + "loss": 1.2674, + "step": 8452 + }, + { + "epoch": 1.5049857549857548, + "grad_norm": 0.7369286417961121, + "learning_rate": 0.00013808075086210647, + "loss": 1.0978, + "step": 8453 + }, + { + "epoch": 1.5051638176638176, + "grad_norm": 0.6336679458618164, + "learning_rate": 0.00013806780766575588, + "loss": 1.0922, + "step": 8454 + }, + { + "epoch": 1.5053418803418803, + "grad_norm": 0.700219452381134, + "learning_rate": 0.0001380548637235457, + "loss": 1.0908, + "step": 8455 + }, + { + "epoch": 1.505519943019943, + "grad_norm": 0.6346127986907959, + "learning_rate": 0.0001380419190357295, + "loss": 1.1265, + "step": 8456 + }, + { + "epoch": 1.5056980056980058, + "grad_norm": 0.8653196096420288, + "learning_rate": 0.00013802897360256093, + "loss": 1.0466, + "step": 8457 + }, + { + "epoch": 1.5058760683760684, + "grad_norm": 0.6589069962501526, + "learning_rate": 0.0001380160274242936, + "loss": 1.245, + "step": 8458 + }, + { + "epoch": 1.506054131054131, + "grad_norm": 0.6527602076530457, + "learning_rate": 0.00013800308050118117, + "loss": 1.1539, + "step": 8459 + }, + { + "epoch": 1.5062321937321936, + "grad_norm": 0.6005436182022095, + "learning_rate": 0.00013799013283347734, + "loss": 0.899, + "step": 8460 + }, + { + "epoch": 1.5064102564102564, + "grad_norm": 0.6954274773597717, + "learning_rate": 0.0001379771844214358, + "loss": 1.1245, + "step": 8461 + }, + { + "epoch": 1.5065883190883191, + "grad_norm": 0.658764660358429, + "learning_rate": 0.00013796423526531019, + "loss": 0.9884, + "step": 8462 + }, + { + "epoch": 1.506766381766382, + "grad_norm": 0.652214527130127, + "learning_rate": 0.0001379512853653543, + "loss": 0.9711, + "step": 8463 + }, + { + "epoch": 1.5069444444444444, + "grad_norm": 0.5680044889450073, + "learning_rate": 0.00013793833472182176, + "loss": 0.9055, + "step": 8464 + }, + { + "epoch": 1.5071225071225072, + "grad_norm": 0.7524166703224182, + "learning_rate": 0.0001379253833349664, + "loss": 1.1163, + "step": 8465 + }, + { + "epoch": 1.5073005698005697, + "grad_norm": 0.692936897277832, + "learning_rate": 0.0001379124312050419, + "loss": 0.899, + "step": 8466 + }, + { + "epoch": 1.5074786324786325, + "grad_norm": 0.6871617436408997, + "learning_rate": 0.00013789947833230207, + "loss": 0.9416, + "step": 8467 + }, + { + "epoch": 1.5076566951566952, + "grad_norm": 0.5983462333679199, + "learning_rate": 0.0001378865247170007, + "loss": 0.9776, + "step": 8468 + }, + { + "epoch": 1.507834757834758, + "grad_norm": 0.6486790180206299, + "learning_rate": 0.0001378735703593916, + "loss": 0.9346, + "step": 8469 + }, + { + "epoch": 1.5080128205128205, + "grad_norm": 0.6843809485435486, + "learning_rate": 0.00013786061525972857, + "loss": 1.1276, + "step": 8470 + }, + { + "epoch": 1.5081908831908832, + "grad_norm": 0.5734516382217407, + "learning_rate": 0.00013784765941826538, + "loss": 0.6939, + "step": 8471 + }, + { + "epoch": 1.5083689458689458, + "grad_norm": 0.6126381754875183, + "learning_rate": 0.00013783470283525596, + "loss": 0.8609, + "step": 8472 + }, + { + "epoch": 1.5085470085470085, + "grad_norm": 0.7570928335189819, + "learning_rate": 0.00013782174551095415, + "loss": 0.8809, + "step": 8473 + }, + { + "epoch": 1.5087250712250713, + "grad_norm": 0.6911360025405884, + "learning_rate": 0.00013780878744561377, + "loss": 0.9916, + "step": 8474 + }, + { + "epoch": 1.508903133903134, + "grad_norm": 0.6651954650878906, + "learning_rate": 0.00013779582863948878, + "loss": 1.0012, + "step": 8475 + }, + { + "epoch": 1.5090811965811965, + "grad_norm": 0.845396876335144, + "learning_rate": 0.000137782869092833, + "loss": 0.8455, + "step": 8476 + }, + { + "epoch": 1.5092592592592593, + "grad_norm": 0.6958050727844238, + "learning_rate": 0.00013776990880590042, + "loss": 1.0264, + "step": 8477 + }, + { + "epoch": 1.5094373219373218, + "grad_norm": 0.6950124502182007, + "learning_rate": 0.00013775694777894493, + "loss": 1.0547, + "step": 8478 + }, + { + "epoch": 1.5096153846153846, + "grad_norm": 0.7243088483810425, + "learning_rate": 0.00013774398601222045, + "loss": 1.0999, + "step": 8479 + }, + { + "epoch": 1.5097934472934473, + "grad_norm": 0.6820448040962219, + "learning_rate": 0.00013773102350598097, + "loss": 0.823, + "step": 8480 + }, + { + "epoch": 1.50997150997151, + "grad_norm": 0.689996063709259, + "learning_rate": 0.0001377180602604805, + "loss": 1.049, + "step": 8481 + }, + { + "epoch": 1.5101495726495726, + "grad_norm": 0.6763314604759216, + "learning_rate": 0.000137705096275973, + "loss": 0.9633, + "step": 8482 + }, + { + "epoch": 1.5103276353276354, + "grad_norm": 0.6760517358779907, + "learning_rate": 0.00013769213155271243, + "loss": 1.0326, + "step": 8483 + }, + { + "epoch": 1.510505698005698, + "grad_norm": 0.7181188464164734, + "learning_rate": 0.00013767916609095285, + "loss": 0.9629, + "step": 8484 + }, + { + "epoch": 1.5106837606837606, + "grad_norm": 0.7102212905883789, + "learning_rate": 0.0001376661998909483, + "loss": 1.2714, + "step": 8485 + }, + { + "epoch": 1.5108618233618234, + "grad_norm": 0.6719805598258972, + "learning_rate": 0.00013765323295295278, + "loss": 0.7848, + "step": 8486 + }, + { + "epoch": 1.5110398860398861, + "grad_norm": 0.6592095494270325, + "learning_rate": 0.0001376402652772204, + "loss": 0.882, + "step": 8487 + }, + { + "epoch": 1.5112179487179487, + "grad_norm": 0.6858693361282349, + "learning_rate": 0.00013762729686400522, + "loss": 0.9418, + "step": 8488 + }, + { + "epoch": 1.5113960113960114, + "grad_norm": 0.7183199524879456, + "learning_rate": 0.0001376143277135613, + "loss": 1.0611, + "step": 8489 + }, + { + "epoch": 1.511574074074074, + "grad_norm": 0.6294263005256653, + "learning_rate": 0.00013760135782614277, + "loss": 0.864, + "step": 8490 + }, + { + "epoch": 1.5117521367521367, + "grad_norm": 0.6762619614601135, + "learning_rate": 0.00013758838720200376, + "loss": 1.0295, + "step": 8491 + }, + { + "epoch": 1.5119301994301995, + "grad_norm": 0.6919726133346558, + "learning_rate": 0.00013757541584139834, + "loss": 1.0803, + "step": 8492 + }, + { + "epoch": 1.5121082621082622, + "grad_norm": 0.6801241040229797, + "learning_rate": 0.00013756244374458075, + "loss": 1.1394, + "step": 8493 + }, + { + "epoch": 1.5122863247863247, + "grad_norm": 0.6758754253387451, + "learning_rate": 0.0001375494709118051, + "loss": 1.0053, + "step": 8494 + }, + { + "epoch": 1.5124643874643875, + "grad_norm": 0.6727001070976257, + "learning_rate": 0.00013753649734332555, + "loss": 1.1407, + "step": 8495 + }, + { + "epoch": 1.51264245014245, + "grad_norm": 0.693913459777832, + "learning_rate": 0.00013752352303939632, + "loss": 1.1804, + "step": 8496 + }, + { + "epoch": 1.5128205128205128, + "grad_norm": 0.6122510433197021, + "learning_rate": 0.0001375105480002716, + "loss": 0.917, + "step": 8497 + }, + { + "epoch": 1.5129985754985755, + "grad_norm": 0.6305009722709656, + "learning_rate": 0.00013749757222620562, + "loss": 1.1075, + "step": 8498 + }, + { + "epoch": 1.5131766381766383, + "grad_norm": 0.7249642610549927, + "learning_rate": 0.0001374845957174526, + "loss": 0.9107, + "step": 8499 + }, + { + "epoch": 1.5133547008547008, + "grad_norm": 0.6922136545181274, + "learning_rate": 0.0001374716184742668, + "loss": 0.9974, + "step": 8500 + }, + { + "epoch": 1.5135327635327636, + "grad_norm": 0.6989904046058655, + "learning_rate": 0.00013745864049690245, + "loss": 0.9866, + "step": 8501 + }, + { + "epoch": 1.513710826210826, + "grad_norm": 0.6284058094024658, + "learning_rate": 0.0001374456617856139, + "loss": 0.8658, + "step": 8502 + }, + { + "epoch": 1.5138888888888888, + "grad_norm": 0.615388810634613, + "learning_rate": 0.00013743268234065535, + "loss": 0.7876, + "step": 8503 + }, + { + "epoch": 1.5140669515669516, + "grad_norm": 0.6212600469589233, + "learning_rate": 0.0001374197021622812, + "loss": 0.855, + "step": 8504 + }, + { + "epoch": 1.5142450142450143, + "grad_norm": 0.6312419772148132, + "learning_rate": 0.00013740672125074567, + "loss": 0.9252, + "step": 8505 + }, + { + "epoch": 1.5144230769230769, + "grad_norm": 0.7094576954841614, + "learning_rate": 0.00013739373960630315, + "loss": 0.7655, + "step": 8506 + }, + { + "epoch": 1.5146011396011396, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.000137380757229208, + "loss": 0.7855, + "step": 8507 + }, + { + "epoch": 1.5147792022792022, + "grad_norm": 0.6798399686813354, + "learning_rate": 0.00013736777411971457, + "loss": 0.9935, + "step": 8508 + }, + { + "epoch": 1.514957264957265, + "grad_norm": 0.7835991978645325, + "learning_rate": 0.00013735479027807723, + "loss": 1.1603, + "step": 8509 + }, + { + "epoch": 1.5151353276353277, + "grad_norm": 0.6230790615081787, + "learning_rate": 0.00013734180570455033, + "loss": 1.1463, + "step": 8510 + }, + { + "epoch": 1.5153133903133904, + "grad_norm": 0.646603524684906, + "learning_rate": 0.00013732882039938835, + "loss": 0.9564, + "step": 8511 + }, + { + "epoch": 1.515491452991453, + "grad_norm": 0.6619647145271301, + "learning_rate": 0.0001373158343628457, + "loss": 0.8492, + "step": 8512 + }, + { + "epoch": 1.5156695156695157, + "grad_norm": 0.6458454132080078, + "learning_rate": 0.00013730284759517675, + "loss": 1.0049, + "step": 8513 + }, + { + "epoch": 1.5158475783475782, + "grad_norm": 0.7415743470191956, + "learning_rate": 0.00013728986009663602, + "loss": 0.872, + "step": 8514 + }, + { + "epoch": 1.516025641025641, + "grad_norm": 0.6198840141296387, + "learning_rate": 0.00013727687186747793, + "loss": 0.8645, + "step": 8515 + }, + { + "epoch": 1.5162037037037037, + "grad_norm": 0.7160853147506714, + "learning_rate": 0.00013726388290795697, + "loss": 1.0144, + "step": 8516 + }, + { + "epoch": 1.5163817663817665, + "grad_norm": 0.6604135632514954, + "learning_rate": 0.00013725089321832765, + "loss": 0.9827, + "step": 8517 + }, + { + "epoch": 1.5165598290598292, + "grad_norm": 0.6480790972709656, + "learning_rate": 0.00013723790279884443, + "loss": 1.0357, + "step": 8518 + }, + { + "epoch": 1.5167378917378918, + "grad_norm": 0.6207128167152405, + "learning_rate": 0.00013722491164976187, + "loss": 0.9467, + "step": 8519 + }, + { + "epoch": 1.5169159544159543, + "grad_norm": 0.6024298667907715, + "learning_rate": 0.00013721191977133452, + "loss": 0.8821, + "step": 8520 + }, + { + "epoch": 1.517094017094017, + "grad_norm": 0.684898316860199, + "learning_rate": 0.00013719892716381688, + "loss": 0.9823, + "step": 8521 + }, + { + "epoch": 1.5172720797720798, + "grad_norm": 0.7460635304450989, + "learning_rate": 0.00013718593382746355, + "loss": 1.2573, + "step": 8522 + }, + { + "epoch": 1.5174501424501425, + "grad_norm": 0.7193243503570557, + "learning_rate": 0.00013717293976252907, + "loss": 1.0162, + "step": 8523 + }, + { + "epoch": 1.5176282051282053, + "grad_norm": 0.6328752040863037, + "learning_rate": 0.0001371599449692681, + "loss": 0.8183, + "step": 8524 + }, + { + "epoch": 1.5178062678062678, + "grad_norm": 0.658784806728363, + "learning_rate": 0.00013714694944793517, + "loss": 0.9315, + "step": 8525 + }, + { + "epoch": 1.5179843304843303, + "grad_norm": 0.7875827550888062, + "learning_rate": 0.00013713395319878493, + "loss": 1.0889, + "step": 8526 + }, + { + "epoch": 1.518162393162393, + "grad_norm": 0.6580079793930054, + "learning_rate": 0.00013712095622207203, + "loss": 1.0276, + "step": 8527 + }, + { + "epoch": 1.5183404558404558, + "grad_norm": 0.6214027404785156, + "learning_rate": 0.00013710795851805106, + "loss": 0.9692, + "step": 8528 + }, + { + "epoch": 1.5185185185185186, + "grad_norm": 0.7839403748512268, + "learning_rate": 0.0001370949600869768, + "loss": 0.7378, + "step": 8529 + }, + { + "epoch": 1.5186965811965814, + "grad_norm": 0.6632764339447021, + "learning_rate": 0.0001370819609291038, + "loss": 0.9431, + "step": 8530 + }, + { + "epoch": 1.5188746438746439, + "grad_norm": 0.7071712017059326, + "learning_rate": 0.00013706896104468682, + "loss": 0.7684, + "step": 8531 + }, + { + "epoch": 1.5190527065527064, + "grad_norm": 0.7494829297065735, + "learning_rate": 0.00013705596043398058, + "loss": 0.9709, + "step": 8532 + }, + { + "epoch": 1.5192307692307692, + "grad_norm": 0.6408106088638306, + "learning_rate": 0.00013704295909723973, + "loss": 0.8494, + "step": 8533 + }, + { + "epoch": 1.519408831908832, + "grad_norm": 0.6043150424957275, + "learning_rate": 0.0001370299570347191, + "loss": 0.7485, + "step": 8534 + }, + { + "epoch": 1.5195868945868947, + "grad_norm": 0.6944992542266846, + "learning_rate": 0.00013701695424667336, + "loss": 0.8403, + "step": 8535 + }, + { + "epoch": 1.5197649572649574, + "grad_norm": 0.7730217576026917, + "learning_rate": 0.00013700395073335726, + "loss": 0.9122, + "step": 8536 + }, + { + "epoch": 1.51994301994302, + "grad_norm": 0.6300255060195923, + "learning_rate": 0.00013699094649502564, + "loss": 0.9185, + "step": 8537 + }, + { + "epoch": 1.5201210826210825, + "grad_norm": 0.648676335811615, + "learning_rate": 0.00013697794153193327, + "loss": 0.9897, + "step": 8538 + }, + { + "epoch": 1.5202991452991452, + "grad_norm": 0.7365788817405701, + "learning_rate": 0.00013696493584433494, + "loss": 0.958, + "step": 8539 + }, + { + "epoch": 1.520477207977208, + "grad_norm": 0.6634557247161865, + "learning_rate": 0.00013695192943248552, + "loss": 0.9389, + "step": 8540 + }, + { + "epoch": 1.5206552706552707, + "grad_norm": 0.6110827922821045, + "learning_rate": 0.00013693892229663977, + "loss": 0.9341, + "step": 8541 + }, + { + "epoch": 1.5208333333333335, + "grad_norm": 0.7207275032997131, + "learning_rate": 0.00013692591443705256, + "loss": 0.9526, + "step": 8542 + }, + { + "epoch": 1.521011396011396, + "grad_norm": 0.7071022391319275, + "learning_rate": 0.0001369129058539788, + "loss": 0.9572, + "step": 8543 + }, + { + "epoch": 1.5211894586894585, + "grad_norm": 0.5898227691650391, + "learning_rate": 0.0001368998965476733, + "loss": 0.921, + "step": 8544 + }, + { + "epoch": 1.5213675213675213, + "grad_norm": 0.7542559504508972, + "learning_rate": 0.000136886886518391, + "loss": 0.7799, + "step": 8545 + }, + { + "epoch": 1.521545584045584, + "grad_norm": 0.6904959678649902, + "learning_rate": 0.00013687387576638674, + "loss": 0.9601, + "step": 8546 + }, + { + "epoch": 1.5217236467236468, + "grad_norm": 0.763414204120636, + "learning_rate": 0.00013686086429191553, + "loss": 1.0046, + "step": 8547 + }, + { + "epoch": 1.5219017094017095, + "grad_norm": 0.6879960298538208, + "learning_rate": 0.00013684785209523224, + "loss": 0.9615, + "step": 8548 + }, + { + "epoch": 1.522079772079772, + "grad_norm": 0.7166057229042053, + "learning_rate": 0.00013683483917659186, + "loss": 0.9481, + "step": 8549 + }, + { + "epoch": 1.5222578347578346, + "grad_norm": 0.6384348273277283, + "learning_rate": 0.0001368218255362493, + "loss": 1.1037, + "step": 8550 + }, + { + "epoch": 1.5224358974358974, + "grad_norm": 0.6564528346061707, + "learning_rate": 0.00013680881117445953, + "loss": 0.951, + "step": 8551 + }, + { + "epoch": 1.52261396011396, + "grad_norm": 0.749301016330719, + "learning_rate": 0.00013679579609147762, + "loss": 0.9324, + "step": 8552 + }, + { + "epoch": 1.5227920227920229, + "grad_norm": 0.8130472898483276, + "learning_rate": 0.00013678278028755848, + "loss": 1.0178, + "step": 8553 + }, + { + "epoch": 1.5229700854700856, + "grad_norm": 0.6763297319412231, + "learning_rate": 0.0001367697637629572, + "loss": 0.9224, + "step": 8554 + }, + { + "epoch": 1.5231481481481481, + "grad_norm": 0.6630885601043701, + "learning_rate": 0.00013675674651792878, + "loss": 1.0254, + "step": 8555 + }, + { + "epoch": 1.5233262108262107, + "grad_norm": 0.7377206087112427, + "learning_rate": 0.00013674372855272825, + "loss": 1.0413, + "step": 8556 + }, + { + "epoch": 1.5235042735042734, + "grad_norm": 0.5270320177078247, + "learning_rate": 0.00013673070986761068, + "loss": 0.7124, + "step": 8557 + }, + { + "epoch": 1.5236823361823362, + "grad_norm": 0.5941976308822632, + "learning_rate": 0.00013671769046283116, + "loss": 1.0281, + "step": 8558 + }, + { + "epoch": 1.523860398860399, + "grad_norm": 0.6131376028060913, + "learning_rate": 0.0001367046703386448, + "loss": 0.7593, + "step": 8559 + }, + { + "epoch": 1.5240384615384617, + "grad_norm": 0.7381763458251953, + "learning_rate": 0.00013669164949530664, + "loss": 1.148, + "step": 8560 + }, + { + "epoch": 1.5242165242165242, + "grad_norm": 0.683274507522583, + "learning_rate": 0.00013667862793307185, + "loss": 0.8354, + "step": 8561 + }, + { + "epoch": 1.5243945868945867, + "grad_norm": 0.6912649273872375, + "learning_rate": 0.0001366656056521955, + "loss": 0.9043, + "step": 8562 + }, + { + "epoch": 1.5245726495726495, + "grad_norm": 0.5999594330787659, + "learning_rate": 0.0001366525826529328, + "loss": 0.6138, + "step": 8563 + }, + { + "epoch": 1.5247507122507122, + "grad_norm": 0.7185927629470825, + "learning_rate": 0.00013663955893553892, + "loss": 0.895, + "step": 8564 + }, + { + "epoch": 1.524928774928775, + "grad_norm": 0.5967002511024475, + "learning_rate": 0.00013662653450026893, + "loss": 0.9636, + "step": 8565 + }, + { + "epoch": 1.5251068376068377, + "grad_norm": 0.7122953534126282, + "learning_rate": 0.00013661350934737813, + "loss": 0.9465, + "step": 8566 + }, + { + "epoch": 1.5252849002849003, + "grad_norm": 0.705326497554779, + "learning_rate": 0.00013660048347712163, + "loss": 1.121, + "step": 8567 + }, + { + "epoch": 1.5254629629629628, + "grad_norm": 0.6023733019828796, + "learning_rate": 0.0001365874568897547, + "loss": 0.9881, + "step": 8568 + }, + { + "epoch": 1.5256410256410255, + "grad_norm": 0.6883122324943542, + "learning_rate": 0.0001365744295855326, + "loss": 1.2372, + "step": 8569 + }, + { + "epoch": 1.5258190883190883, + "grad_norm": 0.718126654624939, + "learning_rate": 0.0001365614015647105, + "loss": 1.0888, + "step": 8570 + }, + { + "epoch": 1.525997150997151, + "grad_norm": 0.6649243831634521, + "learning_rate": 0.00013654837282754367, + "loss": 1.0458, + "step": 8571 + }, + { + "epoch": 1.5261752136752138, + "grad_norm": 0.6959797143936157, + "learning_rate": 0.00013653534337428738, + "loss": 0.9282, + "step": 8572 + }, + { + "epoch": 1.5263532763532763, + "grad_norm": 0.6069976687431335, + "learning_rate": 0.00013652231320519697, + "loss": 0.9706, + "step": 8573 + }, + { + "epoch": 1.526531339031339, + "grad_norm": 0.7085374593734741, + "learning_rate": 0.0001365092823205277, + "loss": 1.1241, + "step": 8574 + }, + { + "epoch": 1.5267094017094016, + "grad_norm": 0.575106143951416, + "learning_rate": 0.00013649625072053488, + "loss": 0.9814, + "step": 8575 + }, + { + "epoch": 1.5268874643874644, + "grad_norm": 0.6541273593902588, + "learning_rate": 0.00013648321840547384, + "loss": 1.0155, + "step": 8576 + }, + { + "epoch": 1.5270655270655271, + "grad_norm": 0.6754382848739624, + "learning_rate": 0.0001364701853755999, + "loss": 1.0284, + "step": 8577 + }, + { + "epoch": 1.5272435897435899, + "grad_norm": 0.6219634413719177, + "learning_rate": 0.00013645715163116846, + "loss": 1.1539, + "step": 8578 + }, + { + "epoch": 1.5274216524216524, + "grad_norm": 0.7625157833099365, + "learning_rate": 0.00013644411717243486, + "loss": 1.1157, + "step": 8579 + }, + { + "epoch": 1.5275997150997151, + "grad_norm": 0.6944296956062317, + "learning_rate": 0.0001364310819996545, + "loss": 0.8309, + "step": 8580 + }, + { + "epoch": 1.5277777777777777, + "grad_norm": 0.7198494672775269, + "learning_rate": 0.00013641804611308277, + "loss": 1.0883, + "step": 8581 + }, + { + "epoch": 1.5279558404558404, + "grad_norm": 0.6398822069168091, + "learning_rate": 0.00013640500951297508, + "loss": 1.0173, + "step": 8582 + }, + { + "epoch": 1.5281339031339032, + "grad_norm": 0.7306683659553528, + "learning_rate": 0.00013639197219958682, + "loss": 0.9979, + "step": 8583 + }, + { + "epoch": 1.528311965811966, + "grad_norm": 0.6873512268066406, + "learning_rate": 0.00013637893417317348, + "loss": 0.7883, + "step": 8584 + }, + { + "epoch": 1.5284900284900285, + "grad_norm": 0.6482085585594177, + "learning_rate": 0.00013636589543399052, + "loss": 0.9367, + "step": 8585 + }, + { + "epoch": 1.5286680911680912, + "grad_norm": 0.8161232471466064, + "learning_rate": 0.00013635285598229336, + "loss": 1.0582, + "step": 8586 + }, + { + "epoch": 1.5288461538461537, + "grad_norm": 0.6722155809402466, + "learning_rate": 0.0001363398158183375, + "loss": 0.9805, + "step": 8587 + }, + { + "epoch": 1.5290242165242165, + "grad_norm": 0.7175397872924805, + "learning_rate": 0.00013632677494237845, + "loss": 1.0747, + "step": 8588 + }, + { + "epoch": 1.5292022792022792, + "grad_norm": 0.6665592789649963, + "learning_rate": 0.00013631373335467172, + "loss": 1.006, + "step": 8589 + }, + { + "epoch": 1.529380341880342, + "grad_norm": 0.7002299427986145, + "learning_rate": 0.0001363006910554728, + "loss": 1.0702, + "step": 8590 + }, + { + "epoch": 1.5295584045584045, + "grad_norm": 0.7712168097496033, + "learning_rate": 0.00013628764804503725, + "loss": 1.0628, + "step": 8591 + }, + { + "epoch": 1.5297364672364673, + "grad_norm": 0.6620795130729675, + "learning_rate": 0.0001362746043236206, + "loss": 1.01, + "step": 8592 + }, + { + "epoch": 1.5299145299145298, + "grad_norm": 0.6374393701553345, + "learning_rate": 0.00013626155989147846, + "loss": 0.9106, + "step": 8593 + }, + { + "epoch": 1.5300925925925926, + "grad_norm": 0.6531631946563721, + "learning_rate": 0.00013624851474886636, + "loss": 1.0488, + "step": 8594 + }, + { + "epoch": 1.5302706552706553, + "grad_norm": 0.6843775510787964, + "learning_rate": 0.00013623546889603993, + "loss": 0.8599, + "step": 8595 + }, + { + "epoch": 1.530448717948718, + "grad_norm": 0.7232706546783447, + "learning_rate": 0.00013622242233325476, + "loss": 1.0875, + "step": 8596 + }, + { + "epoch": 1.5306267806267806, + "grad_norm": 0.695691704750061, + "learning_rate": 0.00013620937506076644, + "loss": 0.9835, + "step": 8597 + }, + { + "epoch": 1.5308048433048433, + "grad_norm": 0.6321248412132263, + "learning_rate": 0.00013619632707883065, + "loss": 0.9778, + "step": 8598 + }, + { + "epoch": 1.5309829059829059, + "grad_norm": 0.6469168663024902, + "learning_rate": 0.00013618327838770303, + "loss": 0.9968, + "step": 8599 + }, + { + "epoch": 1.5311609686609686, + "grad_norm": 0.6798683404922485, + "learning_rate": 0.00013617022898763925, + "loss": 0.78, + "step": 8600 + }, + { + "epoch": 1.5313390313390314, + "grad_norm": 0.6932336091995239, + "learning_rate": 0.00013615717887889496, + "loss": 0.9473, + "step": 8601 + }, + { + "epoch": 1.5315170940170941, + "grad_norm": 0.7304185628890991, + "learning_rate": 0.00013614412806172585, + "loss": 1.0478, + "step": 8602 + }, + { + "epoch": 1.5316951566951567, + "grad_norm": 0.6585272550582886, + "learning_rate": 0.00013613107653638763, + "loss": 0.8563, + "step": 8603 + }, + { + "epoch": 1.5318732193732194, + "grad_norm": 0.6804470419883728, + "learning_rate": 0.00013611802430313604, + "loss": 0.9839, + "step": 8604 + }, + { + "epoch": 1.532051282051282, + "grad_norm": 0.7271378040313721, + "learning_rate": 0.0001361049713622268, + "loss": 1.0906, + "step": 8605 + }, + { + "epoch": 1.5322293447293447, + "grad_norm": 0.7731603980064392, + "learning_rate": 0.00013609191771391562, + "loss": 1.1318, + "step": 8606 + }, + { + "epoch": 1.5324074074074074, + "grad_norm": 0.6143709421157837, + "learning_rate": 0.0001360788633584583, + "loss": 0.8726, + "step": 8607 + }, + { + "epoch": 1.5325854700854702, + "grad_norm": 0.6847203373908997, + "learning_rate": 0.00013606580829611056, + "loss": 0.9963, + "step": 8608 + }, + { + "epoch": 1.5327635327635327, + "grad_norm": 0.7561219334602356, + "learning_rate": 0.0001360527525271283, + "loss": 0.8873, + "step": 8609 + }, + { + "epoch": 1.5329415954415955, + "grad_norm": 0.7997925281524658, + "learning_rate": 0.0001360396960517672, + "loss": 0.7675, + "step": 8610 + }, + { + "epoch": 1.533119658119658, + "grad_norm": 0.7206357717514038, + "learning_rate": 0.00013602663887028315, + "loss": 1.0084, + "step": 8611 + }, + { + "epoch": 1.5332977207977208, + "grad_norm": 0.6454238891601562, + "learning_rate": 0.00013601358098293194, + "loss": 0.8194, + "step": 8612 + }, + { + "epoch": 1.5334757834757835, + "grad_norm": 0.5531884431838989, + "learning_rate": 0.0001360005223899694, + "loss": 0.8596, + "step": 8613 + }, + { + "epoch": 1.5336538461538463, + "grad_norm": 0.659161388874054, + "learning_rate": 0.00013598746309165144, + "loss": 1.0363, + "step": 8614 + }, + { + "epoch": 1.5338319088319088, + "grad_norm": 0.6958948373794556, + "learning_rate": 0.00013597440308823385, + "loss": 0.9852, + "step": 8615 + }, + { + "epoch": 1.5340099715099715, + "grad_norm": 0.7147171497344971, + "learning_rate": 0.0001359613423799726, + "loss": 1.0506, + "step": 8616 + }, + { + "epoch": 1.534188034188034, + "grad_norm": 0.604450523853302, + "learning_rate": 0.00013594828096712353, + "loss": 0.9344, + "step": 8617 + }, + { + "epoch": 1.5343660968660968, + "grad_norm": 0.714547336101532, + "learning_rate": 0.00013593521884994257, + "loss": 1.1583, + "step": 8618 + }, + { + "epoch": 1.5345441595441596, + "grad_norm": 0.6864442825317383, + "learning_rate": 0.00013592215602868565, + "loss": 0.991, + "step": 8619 + }, + { + "epoch": 1.5347222222222223, + "grad_norm": 0.6384446620941162, + "learning_rate": 0.00013590909250360873, + "loss": 0.8799, + "step": 8620 + }, + { + "epoch": 1.5349002849002849, + "grad_norm": 0.7307949662208557, + "learning_rate": 0.00013589602827496772, + "loss": 1.0276, + "step": 8621 + }, + { + "epoch": 1.5350783475783476, + "grad_norm": 0.6620129942893982, + "learning_rate": 0.00013588296334301862, + "loss": 0.9378, + "step": 8622 + }, + { + "epoch": 1.5352564102564101, + "grad_norm": 0.7216851711273193, + "learning_rate": 0.00013586989770801735, + "loss": 0.8984, + "step": 8623 + }, + { + "epoch": 1.5354344729344729, + "grad_norm": 0.7319885492324829, + "learning_rate": 0.00013585683137022, + "loss": 1.0357, + "step": 8624 + }, + { + "epoch": 1.5356125356125356, + "grad_norm": 0.7455703616142273, + "learning_rate": 0.00013584376432988247, + "loss": 0.9727, + "step": 8625 + }, + { + "epoch": 1.5357905982905984, + "grad_norm": 0.7285277247428894, + "learning_rate": 0.0001358306965872609, + "loss": 1.1132, + "step": 8626 + }, + { + "epoch": 1.535968660968661, + "grad_norm": 0.6250096559524536, + "learning_rate": 0.00013581762814261124, + "loss": 0.8538, + "step": 8627 + }, + { + "epoch": 1.5361467236467237, + "grad_norm": 0.6252279281616211, + "learning_rate": 0.0001358045589961895, + "loss": 0.822, + "step": 8628 + }, + { + "epoch": 1.5363247863247862, + "grad_norm": 0.7723368406295776, + "learning_rate": 0.0001357914891482519, + "loss": 0.9841, + "step": 8629 + }, + { + "epoch": 1.536502849002849, + "grad_norm": 0.6855236887931824, + "learning_rate": 0.00013577841859905435, + "loss": 0.9512, + "step": 8630 + }, + { + "epoch": 1.5366809116809117, + "grad_norm": 0.8320944309234619, + "learning_rate": 0.00013576534734885303, + "loss": 1.0324, + "step": 8631 + }, + { + "epoch": 1.5368589743589745, + "grad_norm": 0.6970052123069763, + "learning_rate": 0.00013575227539790405, + "loss": 0.9874, + "step": 8632 + }, + { + "epoch": 1.5370370370370372, + "grad_norm": 0.7774853110313416, + "learning_rate": 0.00013573920274646345, + "loss": 0.962, + "step": 8633 + }, + { + "epoch": 1.5372150997150997, + "grad_norm": 0.6479182839393616, + "learning_rate": 0.0001357261293947875, + "loss": 0.9438, + "step": 8634 + }, + { + "epoch": 1.5373931623931623, + "grad_norm": 0.6855679750442505, + "learning_rate": 0.00013571305534313218, + "loss": 1.0898, + "step": 8635 + }, + { + "epoch": 1.537571225071225, + "grad_norm": 0.6527835726737976, + "learning_rate": 0.00013569998059175377, + "loss": 0.954, + "step": 8636 + }, + { + "epoch": 1.5377492877492878, + "grad_norm": 0.6601176857948303, + "learning_rate": 0.00013568690514090837, + "loss": 1.0183, + "step": 8637 + }, + { + "epoch": 1.5379273504273505, + "grad_norm": 0.6628120541572571, + "learning_rate": 0.0001356738289908522, + "loss": 1.0651, + "step": 8638 + }, + { + "epoch": 1.5381054131054133, + "grad_norm": 0.7492203712463379, + "learning_rate": 0.00013566075214184147, + "loss": 1.2438, + "step": 8639 + }, + { + "epoch": 1.5382834757834758, + "grad_norm": 0.6781343817710876, + "learning_rate": 0.00013564767459413237, + "loss": 0.9413, + "step": 8640 + }, + { + "epoch": 1.5384615384615383, + "grad_norm": 0.6890891790390015, + "learning_rate": 0.00013563459634798115, + "loss": 0.9912, + "step": 8641 + }, + { + "epoch": 1.538639601139601, + "grad_norm": 0.722820520401001, + "learning_rate": 0.00013562151740364404, + "loss": 1.1799, + "step": 8642 + }, + { + "epoch": 1.5388176638176638, + "grad_norm": 0.738369882106781, + "learning_rate": 0.0001356084377613773, + "loss": 1.1313, + "step": 8643 + }, + { + "epoch": 1.5389957264957266, + "grad_norm": 0.6232718229293823, + "learning_rate": 0.00013559535742143717, + "loss": 0.9035, + "step": 8644 + }, + { + "epoch": 1.5391737891737893, + "grad_norm": 0.7371624708175659, + "learning_rate": 0.00013558227638407996, + "loss": 1.3377, + "step": 8645 + }, + { + "epoch": 1.5393518518518519, + "grad_norm": 0.658353865146637, + "learning_rate": 0.00013556919464956197, + "loss": 0.9591, + "step": 8646 + }, + { + "epoch": 1.5395299145299144, + "grad_norm": 0.6205827593803406, + "learning_rate": 0.0001355561122181395, + "loss": 0.9217, + "step": 8647 + }, + { + "epoch": 1.5397079772079771, + "grad_norm": 0.5892502069473267, + "learning_rate": 0.00013554302909006888, + "loss": 0.8893, + "step": 8648 + }, + { + "epoch": 1.53988603988604, + "grad_norm": 1.224568486213684, + "learning_rate": 0.0001355299452656064, + "loss": 0.8237, + "step": 8649 + }, + { + "epoch": 1.5400641025641026, + "grad_norm": 0.7732635736465454, + "learning_rate": 0.0001355168607450085, + "loss": 1.1043, + "step": 8650 + }, + { + "epoch": 1.5402421652421654, + "grad_norm": 0.6365402340888977, + "learning_rate": 0.00013550377552853146, + "loss": 1.0345, + "step": 8651 + }, + { + "epoch": 1.540420227920228, + "grad_norm": 0.7046400904655457, + "learning_rate": 0.00013549068961643171, + "loss": 1.0361, + "step": 8652 + }, + { + "epoch": 1.5405982905982905, + "grad_norm": 0.6760256886482239, + "learning_rate": 0.0001354776030089656, + "loss": 0.9437, + "step": 8653 + }, + { + "epoch": 1.5407763532763532, + "grad_norm": 0.6180984973907471, + "learning_rate": 0.00013546451570638958, + "loss": 0.9737, + "step": 8654 + }, + { + "epoch": 1.540954415954416, + "grad_norm": 0.6221960186958313, + "learning_rate": 0.00013545142770896005, + "loss": 0.9313, + "step": 8655 + }, + { + "epoch": 1.5411324786324787, + "grad_norm": 0.6887816786766052, + "learning_rate": 0.0001354383390169334, + "loss": 1.1736, + "step": 8656 + }, + { + "epoch": 1.5413105413105415, + "grad_norm": 0.5840606093406677, + "learning_rate": 0.00013542524963056614, + "loss": 0.9269, + "step": 8657 + }, + { + "epoch": 1.541488603988604, + "grad_norm": 0.7396654486656189, + "learning_rate": 0.00013541215955011472, + "loss": 1.1189, + "step": 8658 + }, + { + "epoch": 1.5416666666666665, + "grad_norm": 0.780616819858551, + "learning_rate": 0.00013539906877583555, + "loss": 1.1251, + "step": 8659 + }, + { + "epoch": 1.5418447293447293, + "grad_norm": 0.6975206732749939, + "learning_rate": 0.0001353859773079852, + "loss": 1.2134, + "step": 8660 + }, + { + "epoch": 1.542022792022792, + "grad_norm": 0.7572869658470154, + "learning_rate": 0.00013537288514682013, + "loss": 0.9396, + "step": 8661 + }, + { + "epoch": 1.5422008547008548, + "grad_norm": 0.6252159476280212, + "learning_rate": 0.00013535979229259686, + "loss": 0.8449, + "step": 8662 + }, + { + "epoch": 1.5423789173789175, + "grad_norm": 0.7321650981903076, + "learning_rate": 0.0001353466987455719, + "loss": 1.3263, + "step": 8663 + }, + { + "epoch": 1.54255698005698, + "grad_norm": 0.7168700695037842, + "learning_rate": 0.00013533360450600177, + "loss": 0.8923, + "step": 8664 + }, + { + "epoch": 1.5427350427350426, + "grad_norm": 0.5931934714317322, + "learning_rate": 0.00013532050957414313, + "loss": 0.8448, + "step": 8665 + }, + { + "epoch": 1.5429131054131053, + "grad_norm": 0.6621279120445251, + "learning_rate": 0.00013530741395025245, + "loss": 1.1023, + "step": 8666 + }, + { + "epoch": 1.543091168091168, + "grad_norm": 0.7133732438087463, + "learning_rate": 0.00013529431763458633, + "loss": 0.9986, + "step": 8667 + }, + { + "epoch": 1.5432692307692308, + "grad_norm": 0.7589015960693359, + "learning_rate": 0.0001352812206274014, + "loss": 1.0111, + "step": 8668 + }, + { + "epoch": 1.5434472934472936, + "grad_norm": 0.6958192586898804, + "learning_rate": 0.0001352681229289542, + "loss": 0.9466, + "step": 8669 + }, + { + "epoch": 1.5436253561253561, + "grad_norm": 0.7539750337600708, + "learning_rate": 0.0001352550245395014, + "loss": 1.0974, + "step": 8670 + }, + { + "epoch": 1.5438034188034186, + "grad_norm": 0.7003816366195679, + "learning_rate": 0.00013524192545929964, + "loss": 1.0354, + "step": 8671 + }, + { + "epoch": 1.5439814814814814, + "grad_norm": 0.6503025889396667, + "learning_rate": 0.00013522882568860558, + "loss": 1.0476, + "step": 8672 + }, + { + "epoch": 1.5441595441595442, + "grad_norm": 0.6757345199584961, + "learning_rate": 0.00013521572522767584, + "loss": 0.864, + "step": 8673 + }, + { + "epoch": 1.544337606837607, + "grad_norm": 0.6857611536979675, + "learning_rate": 0.0001352026240767671, + "loss": 1.1627, + "step": 8674 + }, + { + "epoch": 1.5445156695156697, + "grad_norm": 0.5775430798530579, + "learning_rate": 0.0001351895222361361, + "loss": 0.7444, + "step": 8675 + }, + { + "epoch": 1.5446937321937322, + "grad_norm": 0.7511499524116516, + "learning_rate": 0.00013517641970603952, + "loss": 1.1547, + "step": 8676 + }, + { + "epoch": 1.5448717948717947, + "grad_norm": 0.6727504730224609, + "learning_rate": 0.00013516331648673403, + "loss": 1.0829, + "step": 8677 + }, + { + "epoch": 1.5450498575498575, + "grad_norm": 0.6128812432289124, + "learning_rate": 0.00013515021257847642, + "loss": 0.9318, + "step": 8678 + }, + { + "epoch": 1.5452279202279202, + "grad_norm": 0.7309781312942505, + "learning_rate": 0.00013513710798152343, + "loss": 1.0844, + "step": 8679 + }, + { + "epoch": 1.545405982905983, + "grad_norm": 0.695655882358551, + "learning_rate": 0.00013512400269613176, + "loss": 1.113, + "step": 8680 + }, + { + "epoch": 1.5455840455840457, + "grad_norm": 0.696441650390625, + "learning_rate": 0.00013511089672255824, + "loss": 1.0499, + "step": 8681 + }, + { + "epoch": 1.5457621082621082, + "grad_norm": 0.6309961080551147, + "learning_rate": 0.00013509779006105964, + "loss": 0.8759, + "step": 8682 + }, + { + "epoch": 1.5459401709401708, + "grad_norm": 0.6155984401702881, + "learning_rate": 0.00013508468271189277, + "loss": 0.8967, + "step": 8683 + }, + { + "epoch": 1.5461182336182335, + "grad_norm": 0.6786884665489197, + "learning_rate": 0.00013507157467531442, + "loss": 1.0806, + "step": 8684 + }, + { + "epoch": 1.5462962962962963, + "grad_norm": 0.6494075059890747, + "learning_rate": 0.00013505846595158138, + "loss": 1.0196, + "step": 8685 + }, + { + "epoch": 1.546474358974359, + "grad_norm": 0.7599824070930481, + "learning_rate": 0.00013504535654095055, + "loss": 0.8662, + "step": 8686 + }, + { + "epoch": 1.5466524216524218, + "grad_norm": 0.6017210483551025, + "learning_rate": 0.00013503224644367877, + "loss": 0.872, + "step": 8687 + }, + { + "epoch": 1.5468304843304843, + "grad_norm": 0.7972410321235657, + "learning_rate": 0.00013501913566002288, + "loss": 1.0958, + "step": 8688 + }, + { + "epoch": 1.547008547008547, + "grad_norm": 0.7572960257530212, + "learning_rate": 0.00013500602419023978, + "loss": 1.0219, + "step": 8689 + }, + { + "epoch": 1.5471866096866096, + "grad_norm": 0.6329224109649658, + "learning_rate": 0.00013499291203458635, + "loss": 0.8636, + "step": 8690 + }, + { + "epoch": 1.5473646723646723, + "grad_norm": 0.6777113080024719, + "learning_rate": 0.0001349797991933195, + "loss": 1.0297, + "step": 8691 + }, + { + "epoch": 1.547542735042735, + "grad_norm": 0.6449527144432068, + "learning_rate": 0.00013496668566669617, + "loss": 1.0296, + "step": 8692 + }, + { + "epoch": 1.5477207977207978, + "grad_norm": 0.8236973881721497, + "learning_rate": 0.00013495357145497326, + "loss": 0.8569, + "step": 8693 + }, + { + "epoch": 1.5478988603988604, + "grad_norm": 0.6753743290901184, + "learning_rate": 0.0001349404565584077, + "loss": 1.0733, + "step": 8694 + }, + { + "epoch": 1.5480769230769231, + "grad_norm": 0.6642967462539673, + "learning_rate": 0.0001349273409772565, + "loss": 0.9437, + "step": 8695 + }, + { + "epoch": 1.5482549857549857, + "grad_norm": 0.6470823884010315, + "learning_rate": 0.00013491422471177661, + "loss": 0.999, + "step": 8696 + }, + { + "epoch": 1.5484330484330484, + "grad_norm": 0.7287036776542664, + "learning_rate": 0.000134901107762225, + "loss": 0.9396, + "step": 8697 + }, + { + "epoch": 1.5486111111111112, + "grad_norm": 0.6258324980735779, + "learning_rate": 0.00013488799012885872, + "loss": 1.045, + "step": 8698 + }, + { + "epoch": 1.548789173789174, + "grad_norm": 0.6540539860725403, + "learning_rate": 0.00013487487181193473, + "loss": 0.9939, + "step": 8699 + }, + { + "epoch": 1.5489672364672364, + "grad_norm": 0.7129563093185425, + "learning_rate": 0.00013486175281171003, + "loss": 1.2079, + "step": 8700 + }, + { + "epoch": 1.5491452991452992, + "grad_norm": 0.6383145451545715, + "learning_rate": 0.00013484863312844173, + "loss": 0.9999, + "step": 8701 + }, + { + "epoch": 1.5493233618233617, + "grad_norm": 0.6310200691223145, + "learning_rate": 0.0001348355127623869, + "loss": 1.1193, + "step": 8702 + }, + { + "epoch": 1.5495014245014245, + "grad_norm": 0.6370054483413696, + "learning_rate": 0.0001348223917138025, + "loss": 1.0213, + "step": 8703 + }, + { + "epoch": 1.5496794871794872, + "grad_norm": 0.7052688598632812, + "learning_rate": 0.00013480926998294573, + "loss": 0.8773, + "step": 8704 + }, + { + "epoch": 1.54985754985755, + "grad_norm": 0.6369579434394836, + "learning_rate": 0.00013479614757007355, + "loss": 1.0072, + "step": 8705 + }, + { + "epoch": 1.5500356125356125, + "grad_norm": 0.7152075171470642, + "learning_rate": 0.0001347830244754432, + "loss": 1.0409, + "step": 8706 + }, + { + "epoch": 1.5502136752136753, + "grad_norm": 0.654183566570282, + "learning_rate": 0.00013476990069931173, + "loss": 0.9363, + "step": 8707 + }, + { + "epoch": 1.5503917378917378, + "grad_norm": 0.6700537204742432, + "learning_rate": 0.00013475677624193627, + "loss": 0.985, + "step": 8708 + }, + { + "epoch": 1.5505698005698005, + "grad_norm": 0.7195445895195007, + "learning_rate": 0.00013474365110357402, + "loss": 0.988, + "step": 8709 + }, + { + "epoch": 1.5507478632478633, + "grad_norm": 0.6019890904426575, + "learning_rate": 0.00013473052528448201, + "loss": 0.9915, + "step": 8710 + }, + { + "epoch": 1.550925925925926, + "grad_norm": 0.7787565588951111, + "learning_rate": 0.0001347173987849176, + "loss": 0.9676, + "step": 8711 + }, + { + "epoch": 1.5511039886039886, + "grad_norm": 0.6997103691101074, + "learning_rate": 0.00013470427160513782, + "loss": 1.1158, + "step": 8712 + }, + { + "epoch": 1.5512820512820513, + "grad_norm": 0.6259464025497437, + "learning_rate": 0.00013469114374539998, + "loss": 0.8784, + "step": 8713 + }, + { + "epoch": 1.5514601139601139, + "grad_norm": 0.6159056425094604, + "learning_rate": 0.00013467801520596122, + "loss": 0.9184, + "step": 8714 + }, + { + "epoch": 1.5516381766381766, + "grad_norm": 0.6823606491088867, + "learning_rate": 0.00013466488598707876, + "loss": 0.9542, + "step": 8715 + }, + { + "epoch": 1.5518162393162394, + "grad_norm": 0.6781585812568665, + "learning_rate": 0.0001346517560890099, + "loss": 1.1761, + "step": 8716 + }, + { + "epoch": 1.551994301994302, + "grad_norm": 0.6313831806182861, + "learning_rate": 0.00013463862551201184, + "loss": 0.8935, + "step": 8717 + }, + { + "epoch": 1.5521723646723646, + "grad_norm": 0.7466186881065369, + "learning_rate": 0.0001346254942563419, + "loss": 1.0583, + "step": 8718 + }, + { + "epoch": 1.5523504273504274, + "grad_norm": 0.7073680758476257, + "learning_rate": 0.0001346123623222573, + "loss": 0.9863, + "step": 8719 + }, + { + "epoch": 1.55252849002849, + "grad_norm": 0.6286870241165161, + "learning_rate": 0.00013459922971001536, + "loss": 0.9921, + "step": 8720 + }, + { + "epoch": 1.5527065527065527, + "grad_norm": 0.6047035455703735, + "learning_rate": 0.0001345860964198734, + "loss": 0.9155, + "step": 8721 + }, + { + "epoch": 1.5528846153846154, + "grad_norm": 0.5909964442253113, + "learning_rate": 0.00013457296245208874, + "loss": 0.9593, + "step": 8722 + }, + { + "epoch": 1.5530626780626782, + "grad_norm": 0.7838597893714905, + "learning_rate": 0.00013455982780691869, + "loss": 0.8872, + "step": 8723 + }, + { + "epoch": 1.5532407407407407, + "grad_norm": 0.6914706230163574, + "learning_rate": 0.00013454669248462063, + "loss": 0.9104, + "step": 8724 + }, + { + "epoch": 1.5534188034188035, + "grad_norm": 0.6777952909469604, + "learning_rate": 0.00013453355648545182, + "loss": 0.9839, + "step": 8725 + }, + { + "epoch": 1.553596866096866, + "grad_norm": 0.7482799291610718, + "learning_rate": 0.00013452041980966978, + "loss": 1.1164, + "step": 8726 + }, + { + "epoch": 1.5537749287749287, + "grad_norm": 0.6616327166557312, + "learning_rate": 0.0001345072824575318, + "loss": 0.9574, + "step": 8727 + }, + { + "epoch": 1.5539529914529915, + "grad_norm": 0.7193203568458557, + "learning_rate": 0.00013449414442929532, + "loss": 1.0609, + "step": 8728 + }, + { + "epoch": 1.5541310541310542, + "grad_norm": 0.6599446535110474, + "learning_rate": 0.0001344810057252177, + "loss": 0.9574, + "step": 8729 + }, + { + "epoch": 1.5543091168091168, + "grad_norm": 0.7221707105636597, + "learning_rate": 0.00013446786634555642, + "loss": 0.9819, + "step": 8730 + }, + { + "epoch": 1.5544871794871795, + "grad_norm": 0.6531312465667725, + "learning_rate": 0.0001344547262905689, + "loss": 0.9986, + "step": 8731 + }, + { + "epoch": 1.554665242165242, + "grad_norm": 0.6879804730415344, + "learning_rate": 0.0001344415855605126, + "loss": 1.1078, + "step": 8732 + }, + { + "epoch": 1.5548433048433048, + "grad_norm": 0.708907425403595, + "learning_rate": 0.00013442844415564498, + "loss": 1.0221, + "step": 8733 + }, + { + "epoch": 1.5550213675213675, + "grad_norm": 0.7957375645637512, + "learning_rate": 0.0001344153020762235, + "loss": 1.3101, + "step": 8734 + }, + { + "epoch": 1.5551994301994303, + "grad_norm": 0.7068197727203369, + "learning_rate": 0.00013440215932250567, + "loss": 0.8995, + "step": 8735 + }, + { + "epoch": 1.5553774928774928, + "grad_norm": 0.6455841064453125, + "learning_rate": 0.00013438901589474898, + "loss": 0.7244, + "step": 8736 + }, + { + "epoch": 1.5555555555555556, + "grad_norm": 0.7500516772270203, + "learning_rate": 0.00013437587179321097, + "loss": 1.0161, + "step": 8737 + }, + { + "epoch": 1.555733618233618, + "grad_norm": 0.5983143448829651, + "learning_rate": 0.00013436272701814917, + "loss": 0.9922, + "step": 8738 + }, + { + "epoch": 1.5559116809116809, + "grad_norm": 0.8761729598045349, + "learning_rate": 0.0001343495815698211, + "loss": 1.022, + "step": 8739 + }, + { + "epoch": 1.5560897435897436, + "grad_norm": 0.6901857852935791, + "learning_rate": 0.00013433643544848438, + "loss": 1.0668, + "step": 8740 + }, + { + "epoch": 1.5562678062678064, + "grad_norm": 0.6770836114883423, + "learning_rate": 0.00013432328865439647, + "loss": 0.9516, + "step": 8741 + }, + { + "epoch": 1.556445868945869, + "grad_norm": 0.6138805150985718, + "learning_rate": 0.00013431014118781505, + "loss": 0.8682, + "step": 8742 + }, + { + "epoch": 1.5566239316239316, + "grad_norm": 0.6796693801879883, + "learning_rate": 0.00013429699304899772, + "loss": 1.1132, + "step": 8743 + }, + { + "epoch": 1.5568019943019942, + "grad_norm": 0.6626394987106323, + "learning_rate": 0.000134283844238202, + "loss": 0.9273, + "step": 8744 + }, + { + "epoch": 1.556980056980057, + "grad_norm": 0.7088519334793091, + "learning_rate": 0.00013427069475568563, + "loss": 0.8915, + "step": 8745 + }, + { + "epoch": 1.5571581196581197, + "grad_norm": 0.6244857311248779, + "learning_rate": 0.0001342575446017061, + "loss": 0.9466, + "step": 8746 + }, + { + "epoch": 1.5573361823361824, + "grad_norm": 0.6969038248062134, + "learning_rate": 0.00013424439377652123, + "loss": 1.2307, + "step": 8747 + }, + { + "epoch": 1.5575142450142452, + "grad_norm": 0.6636740565299988, + "learning_rate": 0.0001342312422803886, + "loss": 0.9456, + "step": 8748 + }, + { + "epoch": 1.5576923076923077, + "grad_norm": 0.7863389253616333, + "learning_rate": 0.00013421809011356586, + "loss": 1.1888, + "step": 8749 + }, + { + "epoch": 1.5578703703703702, + "grad_norm": 0.7504058480262756, + "learning_rate": 0.00013420493727631073, + "loss": 1.2602, + "step": 8750 + }, + { + "epoch": 1.558048433048433, + "grad_norm": 0.7173139452934265, + "learning_rate": 0.00013419178376888085, + "loss": 1.0726, + "step": 8751 + }, + { + "epoch": 1.5582264957264957, + "grad_norm": 0.6517474055290222, + "learning_rate": 0.00013417862959153406, + "loss": 1.1299, + "step": 8752 + }, + { + "epoch": 1.5584045584045585, + "grad_norm": 0.8911739587783813, + "learning_rate": 0.00013416547474452803, + "loss": 1.105, + "step": 8753 + }, + { + "epoch": 1.5585826210826212, + "grad_norm": 0.7116649150848389, + "learning_rate": 0.00013415231922812049, + "loss": 0.8037, + "step": 8754 + }, + { + "epoch": 1.5587606837606838, + "grad_norm": 0.6935904026031494, + "learning_rate": 0.00013413916304256916, + "loss": 1.2778, + "step": 8755 + }, + { + "epoch": 1.5589387464387463, + "grad_norm": 0.652763843536377, + "learning_rate": 0.00013412600618813186, + "loss": 0.9188, + "step": 8756 + }, + { + "epoch": 1.559116809116809, + "grad_norm": 0.6545276641845703, + "learning_rate": 0.00013411284866506637, + "loss": 1.0116, + "step": 8757 + }, + { + "epoch": 1.5592948717948718, + "grad_norm": 0.632165253162384, + "learning_rate": 0.0001340996904736305, + "loss": 0.8538, + "step": 8758 + }, + { + "epoch": 1.5594729344729346, + "grad_norm": 0.6719664931297302, + "learning_rate": 0.000134086531614082, + "loss": 1.1877, + "step": 8759 + }, + { + "epoch": 1.5596509971509973, + "grad_norm": 0.6691158413887024, + "learning_rate": 0.00013407337208667873, + "loss": 1.0411, + "step": 8760 + }, + { + "epoch": 1.5598290598290598, + "grad_norm": 0.7711479067802429, + "learning_rate": 0.0001340602118916785, + "loss": 0.9995, + "step": 8761 + }, + { + "epoch": 1.5600071225071224, + "grad_norm": 0.7229881286621094, + "learning_rate": 0.0001340470510293392, + "loss": 1.1751, + "step": 8762 + }, + { + "epoch": 1.5601851851851851, + "grad_norm": 0.7183271646499634, + "learning_rate": 0.00013403388949991864, + "loss": 0.9371, + "step": 8763 + }, + { + "epoch": 1.5603632478632479, + "grad_norm": 0.8142383098602295, + "learning_rate": 0.00013402072730367475, + "loss": 1.0199, + "step": 8764 + }, + { + "epoch": 1.5605413105413106, + "grad_norm": 0.6349362134933472, + "learning_rate": 0.00013400756444086534, + "loss": 0.8453, + "step": 8765 + }, + { + "epoch": 1.5607193732193734, + "grad_norm": 0.651900589466095, + "learning_rate": 0.00013399440091174834, + "loss": 0.8952, + "step": 8766 + }, + { + "epoch": 1.560897435897436, + "grad_norm": 0.6873346567153931, + "learning_rate": 0.00013398123671658172, + "loss": 0.9438, + "step": 8767 + }, + { + "epoch": 1.5610754985754984, + "grad_norm": 0.7404754757881165, + "learning_rate": 0.00013396807185562333, + "loss": 1.123, + "step": 8768 + }, + { + "epoch": 1.5612535612535612, + "grad_norm": 0.7449641227722168, + "learning_rate": 0.00013395490632913111, + "loss": 0.9407, + "step": 8769 + }, + { + "epoch": 1.561431623931624, + "grad_norm": 0.7393384575843811, + "learning_rate": 0.0001339417401373631, + "loss": 1.0209, + "step": 8770 + }, + { + "epoch": 1.5616096866096867, + "grad_norm": 0.6787426471710205, + "learning_rate": 0.00013392857328057713, + "loss": 0.9768, + "step": 8771 + }, + { + "epoch": 1.5617877492877494, + "grad_norm": 0.6295693516731262, + "learning_rate": 0.00013391540575903127, + "loss": 0.9011, + "step": 8772 + }, + { + "epoch": 1.561965811965812, + "grad_norm": 0.7114503979682922, + "learning_rate": 0.00013390223757298354, + "loss": 1.0696, + "step": 8773 + }, + { + "epoch": 1.5621438746438745, + "grad_norm": 0.7540110349655151, + "learning_rate": 0.00013388906872269184, + "loss": 1.0071, + "step": 8774 + }, + { + "epoch": 1.5623219373219372, + "grad_norm": 0.6472305059432983, + "learning_rate": 0.00013387589920841423, + "loss": 1.105, + "step": 8775 + }, + { + "epoch": 1.5625, + "grad_norm": 0.6936793327331543, + "learning_rate": 0.00013386272903040874, + "loss": 0.885, + "step": 8776 + }, + { + "epoch": 1.5626780626780628, + "grad_norm": 0.7487989068031311, + "learning_rate": 0.00013384955818893343, + "loss": 0.7842, + "step": 8777 + }, + { + "epoch": 1.5628561253561255, + "grad_norm": 0.6109505891799927, + "learning_rate": 0.00013383638668424633, + "loss": 0.9461, + "step": 8778 + }, + { + "epoch": 1.563034188034188, + "grad_norm": 0.6650055646896362, + "learning_rate": 0.00013382321451660558, + "loss": 1.0463, + "step": 8779 + }, + { + "epoch": 1.5632122507122506, + "grad_norm": 0.7147329449653625, + "learning_rate": 0.00013381004168626915, + "loss": 0.946, + "step": 8780 + }, + { + "epoch": 1.5633903133903133, + "grad_norm": 0.6919382810592651, + "learning_rate": 0.00013379686819349522, + "loss": 0.8946, + "step": 8781 + }, + { + "epoch": 1.563568376068376, + "grad_norm": 0.7339401245117188, + "learning_rate": 0.00013378369403854184, + "loss": 0.9625, + "step": 8782 + }, + { + "epoch": 1.5637464387464388, + "grad_norm": 0.6337129473686218, + "learning_rate": 0.00013377051922166717, + "loss": 1.0854, + "step": 8783 + }, + { + "epoch": 1.5639245014245016, + "grad_norm": 0.7301266193389893, + "learning_rate": 0.0001337573437431293, + "loss": 1.017, + "step": 8784 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 0.689540684223175, + "learning_rate": 0.00013374416760318644, + "loss": 0.8734, + "step": 8785 + }, + { + "epoch": 1.5642806267806266, + "grad_norm": 0.7121307849884033, + "learning_rate": 0.0001337309908020967, + "loss": 1.0827, + "step": 8786 + }, + { + "epoch": 1.5644586894586894, + "grad_norm": 0.6715386509895325, + "learning_rate": 0.00013371781334011826, + "loss": 0.946, + "step": 8787 + }, + { + "epoch": 1.5646367521367521, + "grad_norm": 0.6895501613616943, + "learning_rate": 0.00013370463521750932, + "loss": 1.1113, + "step": 8788 + }, + { + "epoch": 1.5648148148148149, + "grad_norm": 0.6592531204223633, + "learning_rate": 0.00013369145643452805, + "loss": 0.9952, + "step": 8789 + }, + { + "epoch": 1.5649928774928776, + "grad_norm": 0.7495190501213074, + "learning_rate": 0.0001336782769914327, + "loss": 1.0936, + "step": 8790 + }, + { + "epoch": 1.5651709401709402, + "grad_norm": 0.7273977398872375, + "learning_rate": 0.00013366509688848147, + "loss": 1.1749, + "step": 8791 + }, + { + "epoch": 1.5653490028490027, + "grad_norm": 0.6447354555130005, + "learning_rate": 0.0001336519161259326, + "loss": 0.8638, + "step": 8792 + }, + { + "epoch": 1.5655270655270654, + "grad_norm": 0.6572020053863525, + "learning_rate": 0.00013363873470404432, + "loss": 0.8005, + "step": 8793 + }, + { + "epoch": 1.5657051282051282, + "grad_norm": 0.676418662071228, + "learning_rate": 0.00013362555262307491, + "loss": 0.7651, + "step": 8794 + }, + { + "epoch": 1.565883190883191, + "grad_norm": 0.6886745095252991, + "learning_rate": 0.0001336123698832827, + "loss": 1.0765, + "step": 8795 + }, + { + "epoch": 1.5660612535612537, + "grad_norm": 0.8134182095527649, + "learning_rate": 0.00013359918648492584, + "loss": 1.2228, + "step": 8796 + }, + { + "epoch": 1.5662393162393162, + "grad_norm": 0.7210384011268616, + "learning_rate": 0.00013358600242826277, + "loss": 0.8247, + "step": 8797 + }, + { + "epoch": 1.5664173789173788, + "grad_norm": 0.7086136341094971, + "learning_rate": 0.00013357281771355175, + "loss": 1.0323, + "step": 8798 + }, + { + "epoch": 1.5665954415954415, + "grad_norm": 0.7419785857200623, + "learning_rate": 0.0001335596323410511, + "loss": 1.213, + "step": 8799 + }, + { + "epoch": 1.5667735042735043, + "grad_norm": 0.6390291452407837, + "learning_rate": 0.0001335464463110192, + "loss": 1.0403, + "step": 8800 + }, + { + "epoch": 1.566951566951567, + "grad_norm": 0.6111941337585449, + "learning_rate": 0.00013353325962371434, + "loss": 0.9747, + "step": 8801 + }, + { + "epoch": 1.5671296296296298, + "grad_norm": 0.6792671084403992, + "learning_rate": 0.00013352007227939488, + "loss": 1.1179, + "step": 8802 + }, + { + "epoch": 1.5673076923076923, + "grad_norm": 0.6656535863876343, + "learning_rate": 0.0001335068842783193, + "loss": 0.9214, + "step": 8803 + }, + { + "epoch": 1.5674857549857548, + "grad_norm": 0.6910907626152039, + "learning_rate": 0.0001334936956207459, + "loss": 1.0609, + "step": 8804 + }, + { + "epoch": 1.5676638176638176, + "grad_norm": 0.65049147605896, + "learning_rate": 0.00013348050630693315, + "loss": 0.7189, + "step": 8805 + }, + { + "epoch": 1.5678418803418803, + "grad_norm": 0.6258065104484558, + "learning_rate": 0.0001334673163371394, + "loss": 1.0683, + "step": 8806 + }, + { + "epoch": 1.568019943019943, + "grad_norm": 0.7518934607505798, + "learning_rate": 0.00013345412571162305, + "loss": 1.2415, + "step": 8807 + }, + { + "epoch": 1.5681980056980058, + "grad_norm": 0.7395275235176086, + "learning_rate": 0.00013344093443064267, + "loss": 0.9153, + "step": 8808 + }, + { + "epoch": 1.5683760683760684, + "grad_norm": 0.6789839267730713, + "learning_rate": 0.00013342774249445663, + "loss": 0.8051, + "step": 8809 + }, + { + "epoch": 1.568554131054131, + "grad_norm": 0.786247193813324, + "learning_rate": 0.00013341454990332342, + "loss": 1.203, + "step": 8810 + }, + { + "epoch": 1.5687321937321936, + "grad_norm": 0.6858161687850952, + "learning_rate": 0.00013340135665750153, + "loss": 0.9494, + "step": 8811 + }, + { + "epoch": 1.5689102564102564, + "grad_norm": 0.7245797514915466, + "learning_rate": 0.0001333881627572494, + "loss": 1.0544, + "step": 8812 + }, + { + "epoch": 1.5690883190883191, + "grad_norm": 0.6176164150238037, + "learning_rate": 0.00013337496820282563, + "loss": 0.9084, + "step": 8813 + }, + { + "epoch": 1.569266381766382, + "grad_norm": 0.7342953681945801, + "learning_rate": 0.00013336177299448868, + "loss": 1.0006, + "step": 8814 + }, + { + "epoch": 1.5694444444444444, + "grad_norm": 0.5183523297309875, + "learning_rate": 0.00013334857713249708, + "loss": 0.6295, + "step": 8815 + }, + { + "epoch": 1.5696225071225072, + "grad_norm": 0.6664513349533081, + "learning_rate": 0.00013333538061710936, + "loss": 0.7569, + "step": 8816 + }, + { + "epoch": 1.5698005698005697, + "grad_norm": 0.7051160931587219, + "learning_rate": 0.0001333221834485841, + "loss": 0.9917, + "step": 8817 + }, + { + "epoch": 1.5699786324786325, + "grad_norm": 0.7888057231903076, + "learning_rate": 0.0001333089856271799, + "loss": 1.0337, + "step": 8818 + }, + { + "epoch": 1.5701566951566952, + "grad_norm": 0.6796144247055054, + "learning_rate": 0.00013329578715315534, + "loss": 1.0915, + "step": 8819 + }, + { + "epoch": 1.570334757834758, + "grad_norm": 0.7442883849143982, + "learning_rate": 0.000133282588026769, + "loss": 1.1695, + "step": 8820 + }, + { + "epoch": 1.5705128205128205, + "grad_norm": 0.6164735555648804, + "learning_rate": 0.00013326938824827946, + "loss": 1.0143, + "step": 8821 + }, + { + "epoch": 1.5706908831908832, + "grad_norm": 0.6526502966880798, + "learning_rate": 0.00013325618781794539, + "loss": 0.8402, + "step": 8822 + }, + { + "epoch": 1.5708689458689458, + "grad_norm": 0.6376087069511414, + "learning_rate": 0.00013324298673602535, + "loss": 0.7582, + "step": 8823 + }, + { + "epoch": 1.5710470085470085, + "grad_norm": 0.6888708472251892, + "learning_rate": 0.00013322978500277807, + "loss": 0.997, + "step": 8824 + }, + { + "epoch": 1.5712250712250713, + "grad_norm": 0.553656280040741, + "learning_rate": 0.0001332165826184622, + "loss": 0.6917, + "step": 8825 + }, + { + "epoch": 1.571403133903134, + "grad_norm": 0.643285870552063, + "learning_rate": 0.0001332033795833364, + "loss": 0.8689, + "step": 8826 + }, + { + "epoch": 1.5715811965811965, + "grad_norm": 0.6210280060768127, + "learning_rate": 0.00013319017589765933, + "loss": 0.9047, + "step": 8827 + }, + { + "epoch": 1.5717592592592593, + "grad_norm": 0.7612366676330566, + "learning_rate": 0.0001331769715616897, + "loss": 0.9818, + "step": 8828 + }, + { + "epoch": 1.5719373219373218, + "grad_norm": 0.5970702171325684, + "learning_rate": 0.00013316376657568628, + "loss": 0.82, + "step": 8829 + }, + { + "epoch": 1.5721153846153846, + "grad_norm": 0.7182583808898926, + "learning_rate": 0.0001331505609399077, + "loss": 1.0633, + "step": 8830 + }, + { + "epoch": 1.5722934472934473, + "grad_norm": 0.7230739593505859, + "learning_rate": 0.00013313735465461278, + "loss": 0.977, + "step": 8831 + }, + { + "epoch": 1.57247150997151, + "grad_norm": 0.6752985119819641, + "learning_rate": 0.00013312414772006018, + "loss": 0.9666, + "step": 8832 + }, + { + "epoch": 1.5726495726495726, + "grad_norm": 0.7724275588989258, + "learning_rate": 0.00013311094013650877, + "loss": 1.148, + "step": 8833 + }, + { + "epoch": 1.5728276353276354, + "grad_norm": 0.7216386198997498, + "learning_rate": 0.00013309773190421724, + "loss": 0.9935, + "step": 8834 + }, + { + "epoch": 1.573005698005698, + "grad_norm": 0.6422320008277893, + "learning_rate": 0.0001330845230234444, + "loss": 0.9383, + "step": 8835 + }, + { + "epoch": 1.5731837606837606, + "grad_norm": 0.669538140296936, + "learning_rate": 0.00013307131349444906, + "loss": 1.0866, + "step": 8836 + }, + { + "epoch": 1.5733618233618234, + "grad_norm": 0.6994584798812866, + "learning_rate": 0.00013305810331749003, + "loss": 0.7882, + "step": 8837 + }, + { + "epoch": 1.5735398860398861, + "grad_norm": 0.8094269633293152, + "learning_rate": 0.00013304489249282617, + "loss": 1.2316, + "step": 8838 + }, + { + "epoch": 1.5737179487179487, + "grad_norm": 0.7180120348930359, + "learning_rate": 0.00013303168102071625, + "loss": 0.9795, + "step": 8839 + }, + { + "epoch": 1.5738960113960114, + "grad_norm": 0.6191438436508179, + "learning_rate": 0.00013301846890141918, + "loss": 0.8957, + "step": 8840 + }, + { + "epoch": 1.574074074074074, + "grad_norm": 0.671094536781311, + "learning_rate": 0.00013300525613519382, + "loss": 1.059, + "step": 8841 + }, + { + "epoch": 1.5742521367521367, + "grad_norm": 0.8062624931335449, + "learning_rate": 0.000132992042722299, + "loss": 0.9782, + "step": 8842 + }, + { + "epoch": 1.5744301994301995, + "grad_norm": 0.6674807667732239, + "learning_rate": 0.00013297882866299362, + "loss": 0.7765, + "step": 8843 + }, + { + "epoch": 1.5746082621082622, + "grad_norm": 0.6369131803512573, + "learning_rate": 0.00013296561395753664, + "loss": 0.97, + "step": 8844 + }, + { + "epoch": 1.5747863247863247, + "grad_norm": 0.7913636565208435, + "learning_rate": 0.00013295239860618691, + "loss": 1.0458, + "step": 8845 + }, + { + "epoch": 1.5749643874643875, + "grad_norm": 0.6722261905670166, + "learning_rate": 0.0001329391826092034, + "loss": 1.1118, + "step": 8846 + }, + { + "epoch": 1.57514245014245, + "grad_norm": 0.6936299800872803, + "learning_rate": 0.00013292596596684502, + "loss": 1.009, + "step": 8847 + }, + { + "epoch": 1.5753205128205128, + "grad_norm": 0.7009961009025574, + "learning_rate": 0.00013291274867937073, + "loss": 0.9904, + "step": 8848 + }, + { + "epoch": 1.5754985754985755, + "grad_norm": 0.6900732517242432, + "learning_rate": 0.0001328995307470395, + "loss": 1.0488, + "step": 8849 + }, + { + "epoch": 1.5756766381766383, + "grad_norm": 0.6389018297195435, + "learning_rate": 0.00013288631217011032, + "loss": 0.9444, + "step": 8850 + }, + { + "epoch": 1.5758547008547008, + "grad_norm": 0.6370900869369507, + "learning_rate": 0.00013287309294884216, + "loss": 0.7465, + "step": 8851 + }, + { + "epoch": 1.5760327635327636, + "grad_norm": 0.6463848948478699, + "learning_rate": 0.00013285987308349405, + "loss": 0.896, + "step": 8852 + }, + { + "epoch": 1.576210826210826, + "grad_norm": 0.6022449731826782, + "learning_rate": 0.00013284665257432495, + "loss": 0.8822, + "step": 8853 + }, + { + "epoch": 1.5763888888888888, + "grad_norm": 0.768189013004303, + "learning_rate": 0.00013283343142159396, + "loss": 0.9862, + "step": 8854 + }, + { + "epoch": 1.5765669515669516, + "grad_norm": 0.6642358303070068, + "learning_rate": 0.00013282020962556007, + "loss": 1.0713, + "step": 8855 + }, + { + "epoch": 1.5767450142450143, + "grad_norm": 0.6883034706115723, + "learning_rate": 0.00013280698718648234, + "loss": 1.0351, + "step": 8856 + }, + { + "epoch": 1.5769230769230769, + "grad_norm": 0.602808952331543, + "learning_rate": 0.00013279376410461988, + "loss": 0.7615, + "step": 8857 + }, + { + "epoch": 1.5771011396011396, + "grad_norm": 0.5968614220619202, + "learning_rate": 0.0001327805403802317, + "loss": 0.9443, + "step": 8858 + }, + { + "epoch": 1.5772792022792022, + "grad_norm": 0.7314837574958801, + "learning_rate": 0.00013276731601357696, + "loss": 0.8784, + "step": 8859 + }, + { + "epoch": 1.577457264957265, + "grad_norm": 0.619754433631897, + "learning_rate": 0.0001327540910049147, + "loss": 0.954, + "step": 8860 + }, + { + "epoch": 1.5776353276353277, + "grad_norm": 0.7195139527320862, + "learning_rate": 0.0001327408653545041, + "loss": 1.0227, + "step": 8861 + }, + { + "epoch": 1.5778133903133904, + "grad_norm": 0.6796214580535889, + "learning_rate": 0.0001327276390626042, + "loss": 1.0593, + "step": 8862 + }, + { + "epoch": 1.577991452991453, + "grad_norm": 0.6576255559921265, + "learning_rate": 0.00013271441212947427, + "loss": 0.7921, + "step": 8863 + }, + { + "epoch": 1.5781695156695157, + "grad_norm": 0.7222092151641846, + "learning_rate": 0.00013270118455537336, + "loss": 1.0545, + "step": 8864 + }, + { + "epoch": 1.5783475783475782, + "grad_norm": 0.7159737348556519, + "learning_rate": 0.00013268795634056066, + "loss": 0.9664, + "step": 8865 + }, + { + "epoch": 1.578525641025641, + "grad_norm": 0.7120481133460999, + "learning_rate": 0.00013267472748529536, + "loss": 1.0148, + "step": 8866 + }, + { + "epoch": 1.5787037037037037, + "grad_norm": 0.7353253364562988, + "learning_rate": 0.00013266149798983666, + "loss": 0.9288, + "step": 8867 + }, + { + "epoch": 1.5788817663817665, + "grad_norm": 0.6652441620826721, + "learning_rate": 0.00013264826785444375, + "loss": 0.8246, + "step": 8868 + }, + { + "epoch": 1.5790598290598292, + "grad_norm": 0.7254189252853394, + "learning_rate": 0.00013263503707937584, + "loss": 0.9892, + "step": 8869 + }, + { + "epoch": 1.5792378917378918, + "grad_norm": 0.6305747032165527, + "learning_rate": 0.00013262180566489223, + "loss": 0.8931, + "step": 8870 + }, + { + "epoch": 1.5794159544159543, + "grad_norm": 0.6560617089271545, + "learning_rate": 0.00013260857361125205, + "loss": 0.9245, + "step": 8871 + }, + { + "epoch": 1.579594017094017, + "grad_norm": 0.7304151654243469, + "learning_rate": 0.00013259534091871462, + "loss": 1.009, + "step": 8872 + }, + { + "epoch": 1.5797720797720798, + "grad_norm": 0.782636821269989, + "learning_rate": 0.00013258210758753918, + "loss": 1.1123, + "step": 8873 + }, + { + "epoch": 1.5799501424501425, + "grad_norm": 0.6992011070251465, + "learning_rate": 0.00013256887361798504, + "loss": 1.099, + "step": 8874 + }, + { + "epoch": 1.5801282051282053, + "grad_norm": 0.7159731984138489, + "learning_rate": 0.00013255563901031148, + "loss": 1.0257, + "step": 8875 + }, + { + "epoch": 1.5803062678062678, + "grad_norm": 0.6055454611778259, + "learning_rate": 0.0001325424037647778, + "loss": 0.9199, + "step": 8876 + }, + { + "epoch": 1.5804843304843303, + "grad_norm": 0.6838310360908508, + "learning_rate": 0.00013252916788164334, + "loss": 0.8644, + "step": 8877 + }, + { + "epoch": 1.580662393162393, + "grad_norm": 0.7067445516586304, + "learning_rate": 0.00013251593136116738, + "loss": 1.0285, + "step": 8878 + }, + { + "epoch": 1.5808404558404558, + "grad_norm": 0.7021774649620056, + "learning_rate": 0.00013250269420360928, + "loss": 1.1263, + "step": 8879 + }, + { + "epoch": 1.5810185185185186, + "grad_norm": 0.6586757302284241, + "learning_rate": 0.00013248945640922843, + "loss": 0.906, + "step": 8880 + }, + { + "epoch": 1.5811965811965814, + "grad_norm": 0.6673910021781921, + "learning_rate": 0.00013247621797828418, + "loss": 1.0652, + "step": 8881 + }, + { + "epoch": 1.5813746438746439, + "grad_norm": 0.6763964295387268, + "learning_rate": 0.00013246297891103588, + "loss": 1.0227, + "step": 8882 + }, + { + "epoch": 1.5815527065527064, + "grad_norm": 0.6536892056465149, + "learning_rate": 0.00013244973920774298, + "loss": 0.9026, + "step": 8883 + }, + { + "epoch": 1.5817307692307692, + "grad_norm": 0.8010411858558655, + "learning_rate": 0.0001324364988686648, + "loss": 1.1167, + "step": 8884 + }, + { + "epoch": 1.581908831908832, + "grad_norm": 0.8159251809120178, + "learning_rate": 0.00013242325789406082, + "loss": 1.233, + "step": 8885 + }, + { + "epoch": 1.5820868945868947, + "grad_norm": 0.6487745046615601, + "learning_rate": 0.00013241001628419048, + "loss": 0.9888, + "step": 8886 + }, + { + "epoch": 1.5822649572649574, + "grad_norm": 0.6750285029411316, + "learning_rate": 0.00013239677403931318, + "loss": 0.8874, + "step": 8887 + }, + { + "epoch": 1.58244301994302, + "grad_norm": 0.7164602875709534, + "learning_rate": 0.0001323835311596884, + "loss": 1.2029, + "step": 8888 + }, + { + "epoch": 1.5826210826210825, + "grad_norm": 0.6081351041793823, + "learning_rate": 0.00013237028764557558, + "loss": 0.9593, + "step": 8889 + }, + { + "epoch": 1.5827991452991452, + "grad_norm": 0.7235409021377563, + "learning_rate": 0.00013235704349723424, + "loss": 1.5324, + "step": 8890 + }, + { + "epoch": 1.582977207977208, + "grad_norm": 0.6658480763435364, + "learning_rate": 0.0001323437987149238, + "loss": 0.9756, + "step": 8891 + }, + { + "epoch": 1.5831552706552707, + "grad_norm": 0.7924265265464783, + "learning_rate": 0.00013233055329890387, + "loss": 0.9329, + "step": 8892 + }, + { + "epoch": 1.5833333333333335, + "grad_norm": 0.6262093186378479, + "learning_rate": 0.0001323173072494339, + "loss": 0.8288, + "step": 8893 + }, + { + "epoch": 1.583511396011396, + "grad_norm": 0.6851989030838013, + "learning_rate": 0.0001323040605667734, + "loss": 0.9822, + "step": 8894 + }, + { + "epoch": 1.5836894586894585, + "grad_norm": 0.6963728666305542, + "learning_rate": 0.00013229081325118194, + "loss": 1.0416, + "step": 8895 + }, + { + "epoch": 1.5838675213675213, + "grad_norm": 0.6017457842826843, + "learning_rate": 0.0001322775653029191, + "loss": 0.8123, + "step": 8896 + }, + { + "epoch": 1.584045584045584, + "grad_norm": 0.7396472096443176, + "learning_rate": 0.0001322643167222444, + "loss": 1.0339, + "step": 8897 + }, + { + "epoch": 1.5842236467236468, + "grad_norm": 0.6360299587249756, + "learning_rate": 0.00013225106750941744, + "loss": 0.9463, + "step": 8898 + }, + { + "epoch": 1.5844017094017095, + "grad_norm": 0.6297624111175537, + "learning_rate": 0.00013223781766469783, + "loss": 0.9921, + "step": 8899 + }, + { + "epoch": 1.584579772079772, + "grad_norm": 0.7722037434577942, + "learning_rate": 0.0001322245671883451, + "loss": 0.8394, + "step": 8900 + }, + { + "epoch": 1.5847578347578346, + "grad_norm": 0.677364706993103, + "learning_rate": 0.00013221131608061895, + "loss": 1.0954, + "step": 8901 + }, + { + "epoch": 1.5849358974358974, + "grad_norm": 0.6954908967018127, + "learning_rate": 0.00013219806434177899, + "loss": 1.0637, + "step": 8902 + }, + { + "epoch": 1.58511396011396, + "grad_norm": 0.7079192996025085, + "learning_rate": 0.00013218481197208484, + "loss": 1.039, + "step": 8903 + }, + { + "epoch": 1.5852920227920229, + "grad_norm": 0.7070451378822327, + "learning_rate": 0.00013217155897179611, + "loss": 1.0025, + "step": 8904 + }, + { + "epoch": 1.5854700854700856, + "grad_norm": 0.6940776705741882, + "learning_rate": 0.00013215830534117257, + "loss": 0.8039, + "step": 8905 + }, + { + "epoch": 1.5856481481481481, + "grad_norm": 0.6545892953872681, + "learning_rate": 0.00013214505108047382, + "loss": 0.9347, + "step": 8906 + }, + { + "epoch": 1.5858262108262107, + "grad_norm": 0.6769635081291199, + "learning_rate": 0.00013213179618995957, + "loss": 1.0321, + "step": 8907 + }, + { + "epoch": 1.5860042735042734, + "grad_norm": 0.6505448222160339, + "learning_rate": 0.00013211854066988953, + "loss": 1.0558, + "step": 8908 + }, + { + "epoch": 1.5861823361823362, + "grad_norm": 0.6764090061187744, + "learning_rate": 0.00013210528452052336, + "loss": 0.8407, + "step": 8909 + }, + { + "epoch": 1.586360398860399, + "grad_norm": 0.6454851627349854, + "learning_rate": 0.00013209202774212088, + "loss": 0.7439, + "step": 8910 + }, + { + "epoch": 1.5865384615384617, + "grad_norm": 0.6911695599555969, + "learning_rate": 0.00013207877033494177, + "loss": 0.9625, + "step": 8911 + }, + { + "epoch": 1.5867165242165242, + "grad_norm": 0.7405226826667786, + "learning_rate": 0.0001320655122992458, + "loss": 1.054, + "step": 8912 + }, + { + "epoch": 1.5868945868945867, + "grad_norm": 0.7362869381904602, + "learning_rate": 0.00013205225363529274, + "loss": 1.0516, + "step": 8913 + }, + { + "epoch": 1.5870726495726495, + "grad_norm": 0.6923766136169434, + "learning_rate": 0.0001320389943433423, + "loss": 1.2323, + "step": 8914 + }, + { + "epoch": 1.5872507122507122, + "grad_norm": 0.7980395555496216, + "learning_rate": 0.00013202573442365435, + "loss": 1.0229, + "step": 8915 + }, + { + "epoch": 1.587428774928775, + "grad_norm": 0.7211610078811646, + "learning_rate": 0.00013201247387648868, + "loss": 1.0666, + "step": 8916 + }, + { + "epoch": 1.5876068376068377, + "grad_norm": 0.6728795766830444, + "learning_rate": 0.00013199921270210506, + "loss": 1.0322, + "step": 8917 + }, + { + "epoch": 1.5877849002849003, + "grad_norm": 0.6226436495780945, + "learning_rate": 0.00013198595090076337, + "loss": 1.0517, + "step": 8918 + }, + { + "epoch": 1.5879629629629628, + "grad_norm": 0.6396511197090149, + "learning_rate": 0.0001319726884727234, + "loss": 0.8662, + "step": 8919 + }, + { + "epoch": 1.5881410256410255, + "grad_norm": 0.5664374828338623, + "learning_rate": 0.00013195942541824497, + "loss": 0.6601, + "step": 8920 + }, + { + "epoch": 1.5883190883190883, + "grad_norm": 0.6556946039199829, + "learning_rate": 0.00013194616173758806, + "loss": 0.9662, + "step": 8921 + }, + { + "epoch": 1.588497150997151, + "grad_norm": 0.7332060933113098, + "learning_rate": 0.00013193289743101245, + "loss": 0.7687, + "step": 8922 + }, + { + "epoch": 1.5886752136752138, + "grad_norm": 0.6103306412696838, + "learning_rate": 0.00013191963249877805, + "loss": 0.8329, + "step": 8923 + }, + { + "epoch": 1.5888532763532763, + "grad_norm": 0.63165283203125, + "learning_rate": 0.00013190636694114475, + "loss": 0.8336, + "step": 8924 + }, + { + "epoch": 1.589031339031339, + "grad_norm": 0.6955820322036743, + "learning_rate": 0.00013189310075837246, + "loss": 1.0457, + "step": 8925 + }, + { + "epoch": 1.5892094017094016, + "grad_norm": 0.6911605596542358, + "learning_rate": 0.00013187983395072114, + "loss": 0.9389, + "step": 8926 + }, + { + "epoch": 1.5893874643874644, + "grad_norm": 0.6493414640426636, + "learning_rate": 0.00013186656651845068, + "loss": 0.9821, + "step": 8927 + }, + { + "epoch": 1.5895655270655271, + "grad_norm": 0.6168226599693298, + "learning_rate": 0.00013185329846182107, + "loss": 1.0259, + "step": 8928 + }, + { + "epoch": 1.5897435897435899, + "grad_norm": 0.6460188627243042, + "learning_rate": 0.0001318400297810922, + "loss": 0.9836, + "step": 8929 + }, + { + "epoch": 1.5899216524216524, + "grad_norm": 0.6630695462226868, + "learning_rate": 0.0001318267604765241, + "loss": 0.8936, + "step": 8930 + }, + { + "epoch": 1.5900997150997151, + "grad_norm": 0.6308651566505432, + "learning_rate": 0.00013181349054837676, + "loss": 0.9583, + "step": 8931 + }, + { + "epoch": 1.5902777777777777, + "grad_norm": 0.6508499979972839, + "learning_rate": 0.00013180021999691018, + "loss": 0.7647, + "step": 8932 + }, + { + "epoch": 1.5904558404558404, + "grad_norm": 0.6625795960426331, + "learning_rate": 0.00013178694882238432, + "loss": 1.0329, + "step": 8933 + }, + { + "epoch": 1.5906339031339032, + "grad_norm": 0.6721987128257751, + "learning_rate": 0.00013177367702505924, + "loss": 0.9377, + "step": 8934 + }, + { + "epoch": 1.590811965811966, + "grad_norm": 0.7295519709587097, + "learning_rate": 0.00013176040460519497, + "loss": 0.9396, + "step": 8935 + }, + { + "epoch": 1.5909900284900285, + "grad_norm": 0.6673944592475891, + "learning_rate": 0.0001317471315630515, + "loss": 1.0284, + "step": 8936 + }, + { + "epoch": 1.5911680911680912, + "grad_norm": 0.6858960390090942, + "learning_rate": 0.00013173385789888898, + "loss": 1.2022, + "step": 8937 + }, + { + "epoch": 1.5913461538461537, + "grad_norm": 0.5836796164512634, + "learning_rate": 0.00013172058361296743, + "loss": 1.0078, + "step": 8938 + }, + { + "epoch": 1.5915242165242165, + "grad_norm": 0.7732513546943665, + "learning_rate": 0.00013170730870554694, + "loss": 1.0912, + "step": 8939 + }, + { + "epoch": 1.5917022792022792, + "grad_norm": 0.7095892429351807, + "learning_rate": 0.0001316940331768876, + "loss": 1.0506, + "step": 8940 + }, + { + "epoch": 1.591880341880342, + "grad_norm": 0.757534384727478, + "learning_rate": 0.00013168075702724952, + "loss": 1.036, + "step": 8941 + }, + { + "epoch": 1.5920584045584045, + "grad_norm": 0.6719361543655396, + "learning_rate": 0.00013166748025689282, + "loss": 0.9406, + "step": 8942 + }, + { + "epoch": 1.5922364672364673, + "grad_norm": 0.6955735087394714, + "learning_rate": 0.00013165420286607763, + "loss": 0.9325, + "step": 8943 + }, + { + "epoch": 1.5924145299145298, + "grad_norm": 0.6810322999954224, + "learning_rate": 0.00013164092485506407, + "loss": 1.0402, + "step": 8944 + }, + { + "epoch": 1.5925925925925926, + "grad_norm": 0.6346224546432495, + "learning_rate": 0.00013162764622411233, + "loss": 0.9725, + "step": 8945 + }, + { + "epoch": 1.5927706552706553, + "grad_norm": 0.728705883026123, + "learning_rate": 0.00013161436697348258, + "loss": 0.9665, + "step": 8946 + }, + { + "epoch": 1.592948717948718, + "grad_norm": 0.6838595271110535, + "learning_rate": 0.00013160108710343494, + "loss": 0.9771, + "step": 8947 + }, + { + "epoch": 1.5931267806267806, + "grad_norm": 0.7052602767944336, + "learning_rate": 0.00013158780661422966, + "loss": 0.8819, + "step": 8948 + }, + { + "epoch": 1.5933048433048433, + "grad_norm": 0.7237630486488342, + "learning_rate": 0.00013157452550612697, + "loss": 1.0609, + "step": 8949 + }, + { + "epoch": 1.5934829059829059, + "grad_norm": 0.6554936766624451, + "learning_rate": 0.00013156124377938699, + "loss": 0.8592, + "step": 8950 + }, + { + "epoch": 1.5936609686609686, + "grad_norm": 0.6125665307044983, + "learning_rate": 0.00013154796143427, + "loss": 0.8399, + "step": 8951 + }, + { + "epoch": 1.5938390313390314, + "grad_norm": 0.6930897235870361, + "learning_rate": 0.0001315346784710363, + "loss": 0.9965, + "step": 8952 + }, + { + "epoch": 1.5940170940170941, + "grad_norm": 0.7808064818382263, + "learning_rate": 0.00013152139488994605, + "loss": 1.0527, + "step": 8953 + }, + { + "epoch": 1.5941951566951567, + "grad_norm": 0.6125522255897522, + "learning_rate": 0.0001315081106912595, + "loss": 1.1159, + "step": 8954 + }, + { + "epoch": 1.5943732193732194, + "grad_norm": 0.5863428711891174, + "learning_rate": 0.00013149482587523703, + "loss": 0.84, + "step": 8955 + }, + { + "epoch": 1.594551282051282, + "grad_norm": 0.7170202732086182, + "learning_rate": 0.00013148154044213882, + "loss": 1.0821, + "step": 8956 + }, + { + "epoch": 1.5947293447293447, + "grad_norm": 0.6409463882446289, + "learning_rate": 0.00013146825439222528, + "loss": 1.0097, + "step": 8957 + }, + { + "epoch": 1.5949074074074074, + "grad_norm": 0.7037690281867981, + "learning_rate": 0.00013145496772575666, + "loss": 1.1511, + "step": 8958 + }, + { + "epoch": 1.5950854700854702, + "grad_norm": 0.6400953531265259, + "learning_rate": 0.00013144168044299326, + "loss": 1.0809, + "step": 8959 + }, + { + "epoch": 1.5952635327635327, + "grad_norm": 0.6129940152168274, + "learning_rate": 0.00013142839254419545, + "loss": 0.8481, + "step": 8960 + }, + { + "epoch": 1.5954415954415955, + "grad_norm": 0.7452271580696106, + "learning_rate": 0.00013141510402962358, + "loss": 1.0649, + "step": 8961 + }, + { + "epoch": 1.595619658119658, + "grad_norm": 0.7407623529434204, + "learning_rate": 0.000131401814899538, + "loss": 0.9084, + "step": 8962 + }, + { + "epoch": 1.5957977207977208, + "grad_norm": 0.7103050947189331, + "learning_rate": 0.0001313885251541991, + "loss": 0.946, + "step": 8963 + }, + { + "epoch": 1.5959757834757835, + "grad_norm": 0.5566636323928833, + "learning_rate": 0.00013137523479386727, + "loss": 0.6781, + "step": 8964 + }, + { + "epoch": 1.5961538461538463, + "grad_norm": 0.8137457966804504, + "learning_rate": 0.00013136194381880288, + "loss": 0.9273, + "step": 8965 + }, + { + "epoch": 1.5963319088319088, + "grad_norm": 0.779330849647522, + "learning_rate": 0.0001313486522292663, + "loss": 1.1105, + "step": 8966 + }, + { + "epoch": 1.5965099715099715, + "grad_norm": 0.6807126998901367, + "learning_rate": 0.00013133536002551808, + "loss": 1.0728, + "step": 8967 + }, + { + "epoch": 1.596688034188034, + "grad_norm": 0.7371507287025452, + "learning_rate": 0.00013132206720781853, + "loss": 0.979, + "step": 8968 + }, + { + "epoch": 1.5968660968660968, + "grad_norm": 0.6811465620994568, + "learning_rate": 0.00013130877377642814, + "loss": 0.9821, + "step": 8969 + }, + { + "epoch": 1.5970441595441596, + "grad_norm": 0.6732743978500366, + "learning_rate": 0.00013129547973160738, + "loss": 0.8511, + "step": 8970 + }, + { + "epoch": 1.5972222222222223, + "grad_norm": 0.594901978969574, + "learning_rate": 0.0001312821850736167, + "loss": 0.9674, + "step": 8971 + }, + { + "epoch": 1.5974002849002849, + "grad_norm": 0.6743764281272888, + "learning_rate": 0.00013126888980271657, + "loss": 0.9268, + "step": 8972 + }, + { + "epoch": 1.5975783475783476, + "grad_norm": 0.7532161474227905, + "learning_rate": 0.00013125559391916752, + "loss": 1.0474, + "step": 8973 + }, + { + "epoch": 1.5977564102564101, + "grad_norm": 0.6331499814987183, + "learning_rate": 0.00013124229742323, + "loss": 1.05, + "step": 8974 + }, + { + "epoch": 1.5979344729344729, + "grad_norm": 0.7418690323829651, + "learning_rate": 0.0001312290003151646, + "loss": 0.9475, + "step": 8975 + }, + { + "epoch": 1.5981125356125356, + "grad_norm": 0.6511179804801941, + "learning_rate": 0.0001312157025952318, + "loss": 0.9206, + "step": 8976 + }, + { + "epoch": 1.5982905982905984, + "grad_norm": 0.6380775570869446, + "learning_rate": 0.00013120240426369215, + "loss": 0.9953, + "step": 8977 + }, + { + "epoch": 1.598468660968661, + "grad_norm": 0.8483675122261047, + "learning_rate": 0.00013118910532080623, + "loss": 0.9454, + "step": 8978 + }, + { + "epoch": 1.5986467236467237, + "grad_norm": 0.6700518727302551, + "learning_rate": 0.00013117580576683455, + "loss": 1.0413, + "step": 8979 + }, + { + "epoch": 1.5988247863247862, + "grad_norm": 0.7750083208084106, + "learning_rate": 0.00013116250560203774, + "loss": 1.1868, + "step": 8980 + }, + { + "epoch": 1.599002849002849, + "grad_norm": 0.7474972009658813, + "learning_rate": 0.00013114920482667635, + "loss": 1.0876, + "step": 8981 + }, + { + "epoch": 1.5991809116809117, + "grad_norm": 0.6920070052146912, + "learning_rate": 0.000131135903441011, + "loss": 1.0787, + "step": 8982 + }, + { + "epoch": 1.5993589743589745, + "grad_norm": 0.7572436928749084, + "learning_rate": 0.00013112260144530232, + "loss": 0.9798, + "step": 8983 + }, + { + "epoch": 1.5995370370370372, + "grad_norm": 0.6983019709587097, + "learning_rate": 0.00013110929883981088, + "loss": 1.1115, + "step": 8984 + }, + { + "epoch": 1.5997150997150997, + "grad_norm": 0.6352120041847229, + "learning_rate": 0.0001310959956247974, + "loss": 0.9962, + "step": 8985 + }, + { + "epoch": 1.5998931623931623, + "grad_norm": 0.596858561038971, + "learning_rate": 0.00013108269180052244, + "loss": 0.8686, + "step": 8986 + }, + { + "epoch": 1.600071225071225, + "grad_norm": 0.6237605214118958, + "learning_rate": 0.00013106938736724672, + "loss": 0.9166, + "step": 8987 + }, + { + "epoch": 1.6002492877492878, + "grad_norm": 0.6818585395812988, + "learning_rate": 0.0001310560823252309, + "loss": 0.9993, + "step": 8988 + }, + { + "epoch": 1.6004273504273505, + "grad_norm": 0.6372287273406982, + "learning_rate": 0.00013104277667473564, + "loss": 0.8589, + "step": 8989 + }, + { + "epoch": 1.6006054131054133, + "grad_norm": 0.6057302355766296, + "learning_rate": 0.0001310294704160217, + "loss": 0.9325, + "step": 8990 + }, + { + "epoch": 1.6007834757834758, + "grad_norm": 0.6999384164810181, + "learning_rate": 0.0001310161635493497, + "loss": 0.8691, + "step": 8991 + }, + { + "epoch": 1.6009615384615383, + "grad_norm": 0.6182113289833069, + "learning_rate": 0.00013100285607498045, + "loss": 1.0271, + "step": 8992 + }, + { + "epoch": 1.601139601139601, + "grad_norm": 0.6681149005889893, + "learning_rate": 0.0001309895479931746, + "loss": 0.989, + "step": 8993 + }, + { + "epoch": 1.6013176638176638, + "grad_norm": 0.6187826991081238, + "learning_rate": 0.00013097623930419293, + "loss": 0.8051, + "step": 8994 + }, + { + "epoch": 1.6014957264957266, + "grad_norm": 0.698793888092041, + "learning_rate": 0.00013096293000829621, + "loss": 1.0762, + "step": 8995 + }, + { + "epoch": 1.6016737891737893, + "grad_norm": 0.693149745464325, + "learning_rate": 0.0001309496201057452, + "loss": 1.0894, + "step": 8996 + }, + { + "epoch": 1.6018518518518519, + "grad_norm": 0.6664052605628967, + "learning_rate": 0.00013093630959680068, + "loss": 0.9835, + "step": 8997 + }, + { + "epoch": 1.6020299145299144, + "grad_norm": 0.6919469833374023, + "learning_rate": 0.0001309229984817234, + "loss": 0.9062, + "step": 8998 + }, + { + "epoch": 1.6022079772079771, + "grad_norm": 0.704781174659729, + "learning_rate": 0.00013090968676077427, + "loss": 0.8582, + "step": 8999 + }, + { + "epoch": 1.60238603988604, + "grad_norm": 0.8055264949798584, + "learning_rate": 0.000130896374434214, + "loss": 0.9813, + "step": 9000 + }, + { + "epoch": 1.6025641025641026, + "grad_norm": 0.6301952004432678, + "learning_rate": 0.00013088306150230348, + "loss": 0.7056, + "step": 9001 + }, + { + "epoch": 1.6027421652421654, + "grad_norm": 0.698544442653656, + "learning_rate": 0.00013086974796530347, + "loss": 0.9806, + "step": 9002 + }, + { + "epoch": 1.602920227920228, + "grad_norm": 0.669548511505127, + "learning_rate": 0.00013085643382347491, + "loss": 1.0317, + "step": 9003 + }, + { + "epoch": 1.6030982905982905, + "grad_norm": 0.6404716372489929, + "learning_rate": 0.00013084311907707864, + "loss": 0.8885, + "step": 9004 + }, + { + "epoch": 1.6032763532763532, + "grad_norm": 0.6968616843223572, + "learning_rate": 0.0001308298037263755, + "loss": 1.0665, + "step": 9005 + }, + { + "epoch": 1.603454415954416, + "grad_norm": 0.849311113357544, + "learning_rate": 0.00013081648777162644, + "loss": 1.1404, + "step": 9006 + }, + { + "epoch": 1.6036324786324787, + "grad_norm": 0.6603094935417175, + "learning_rate": 0.00013080317121309223, + "loss": 0.8341, + "step": 9007 + }, + { + "epoch": 1.6038105413105415, + "grad_norm": 0.6777810454368591, + "learning_rate": 0.00013078985405103394, + "loss": 1.044, + "step": 9008 + }, + { + "epoch": 1.603988603988604, + "grad_norm": 0.6783546209335327, + "learning_rate": 0.0001307765362857124, + "loss": 1.042, + "step": 9009 + }, + { + "epoch": 1.6041666666666665, + "grad_norm": 0.7251788377761841, + "learning_rate": 0.00013076321791738858, + "loss": 0.9004, + "step": 9010 + }, + { + "epoch": 1.6043447293447293, + "grad_norm": 0.7885342240333557, + "learning_rate": 0.00013074989894632338, + "loss": 1.1966, + "step": 9011 + }, + { + "epoch": 1.604522792022792, + "grad_norm": 0.7171013355255127, + "learning_rate": 0.0001307365793727778, + "loss": 1.2242, + "step": 9012 + }, + { + "epoch": 1.6047008547008548, + "grad_norm": 0.6027249693870544, + "learning_rate": 0.00013072325919701283, + "loss": 0.917, + "step": 9013 + }, + { + "epoch": 1.6048789173789175, + "grad_norm": 0.5957151055335999, + "learning_rate": 0.00013070993841928936, + "loss": 0.9154, + "step": 9014 + }, + { + "epoch": 1.60505698005698, + "grad_norm": 0.6190659403800964, + "learning_rate": 0.00013069661703986847, + "loss": 0.7071, + "step": 9015 + }, + { + "epoch": 1.6052350427350426, + "grad_norm": 0.6454868316650391, + "learning_rate": 0.00013068329505901117, + "loss": 0.8381, + "step": 9016 + }, + { + "epoch": 1.6054131054131053, + "grad_norm": 0.6255491375923157, + "learning_rate": 0.00013066997247697837, + "loss": 0.7515, + "step": 9017 + }, + { + "epoch": 1.605591168091168, + "grad_norm": 0.6214072108268738, + "learning_rate": 0.0001306566492940312, + "loss": 1.0101, + "step": 9018 + }, + { + "epoch": 1.6057692307692308, + "grad_norm": 0.7244150638580322, + "learning_rate": 0.0001306433255104307, + "loss": 1.2558, + "step": 9019 + }, + { + "epoch": 1.6059472934472936, + "grad_norm": 0.6162270903587341, + "learning_rate": 0.00013063000112643785, + "loss": 1.1009, + "step": 9020 + }, + { + "epoch": 1.6061253561253561, + "grad_norm": 0.7309414744377136, + "learning_rate": 0.0001306166761423138, + "loss": 1.1973, + "step": 9021 + }, + { + "epoch": 1.6063034188034186, + "grad_norm": 0.7150956392288208, + "learning_rate": 0.00013060335055831957, + "loss": 0.9136, + "step": 9022 + }, + { + "epoch": 1.6064814814814814, + "grad_norm": 0.8187742829322815, + "learning_rate": 0.00013059002437471623, + "loss": 1.0524, + "step": 9023 + }, + { + "epoch": 1.6066595441595442, + "grad_norm": 0.7928692698478699, + "learning_rate": 0.00013057669759176493, + "loss": 1.0249, + "step": 9024 + }, + { + "epoch": 1.606837606837607, + "grad_norm": 0.6929279565811157, + "learning_rate": 0.00013056337020972677, + "loss": 1.1804, + "step": 9025 + }, + { + "epoch": 1.6070156695156697, + "grad_norm": 0.6771654486656189, + "learning_rate": 0.00013055004222886285, + "loss": 1.0284, + "step": 9026 + }, + { + "epoch": 1.6071937321937322, + "grad_norm": 0.6689024567604065, + "learning_rate": 0.0001305367136494343, + "loss": 1.0431, + "step": 9027 + }, + { + "epoch": 1.6073717948717947, + "grad_norm": 0.71135413646698, + "learning_rate": 0.0001305233844717023, + "loss": 0.9692, + "step": 9028 + }, + { + "epoch": 1.6075498575498575, + "grad_norm": 0.5459749698638916, + "learning_rate": 0.00013051005469592796, + "loss": 0.5643, + "step": 9029 + }, + { + "epoch": 1.6077279202279202, + "grad_norm": 0.7225865125656128, + "learning_rate": 0.00013049672432237253, + "loss": 1.0954, + "step": 9030 + }, + { + "epoch": 1.607905982905983, + "grad_norm": 0.6878093481063843, + "learning_rate": 0.0001304833933512971, + "loss": 0.894, + "step": 9031 + }, + { + "epoch": 1.6080840455840457, + "grad_norm": 0.6967248320579529, + "learning_rate": 0.00013047006178296288, + "loss": 1.0356, + "step": 9032 + }, + { + "epoch": 1.6082621082621082, + "grad_norm": 0.6404993534088135, + "learning_rate": 0.00013045672961763114, + "loss": 0.8528, + "step": 9033 + }, + { + "epoch": 1.6084401709401708, + "grad_norm": 0.5919156074523926, + "learning_rate": 0.000130443396855563, + "loss": 0.7196, + "step": 9034 + }, + { + "epoch": 1.6086182336182335, + "grad_norm": 0.6792302131652832, + "learning_rate": 0.00013043006349701977, + "loss": 0.9519, + "step": 9035 + }, + { + "epoch": 1.6087962962962963, + "grad_norm": 0.6263542175292969, + "learning_rate": 0.00013041672954226268, + "loss": 1.0483, + "step": 9036 + }, + { + "epoch": 1.608974358974359, + "grad_norm": 0.5865579843521118, + "learning_rate": 0.00013040339499155294, + "loss": 0.8794, + "step": 9037 + }, + { + "epoch": 1.6091524216524218, + "grad_norm": 0.8383142948150635, + "learning_rate": 0.00013039005984515181, + "loss": 0.8929, + "step": 9038 + }, + { + "epoch": 1.6093304843304843, + "grad_norm": 0.6438691020011902, + "learning_rate": 0.00013037672410332063, + "loss": 0.9957, + "step": 9039 + }, + { + "epoch": 1.609508547008547, + "grad_norm": 0.74748694896698, + "learning_rate": 0.0001303633877663206, + "loss": 0.9809, + "step": 9040 + }, + { + "epoch": 1.6096866096866096, + "grad_norm": 0.6697205901145935, + "learning_rate": 0.00013035005083441312, + "loss": 0.9556, + "step": 9041 + }, + { + "epoch": 1.6098646723646723, + "grad_norm": 0.6577828526496887, + "learning_rate": 0.00013033671330785941, + "loss": 0.8956, + "step": 9042 + }, + { + "epoch": 1.610042735042735, + "grad_norm": 0.6423429846763611, + "learning_rate": 0.0001303233751869208, + "loss": 0.8467, + "step": 9043 + }, + { + "epoch": 1.6102207977207978, + "grad_norm": 0.6552175879478455, + "learning_rate": 0.00013031003647185867, + "loss": 0.8656, + "step": 9044 + }, + { + "epoch": 1.6103988603988604, + "grad_norm": 0.6755174398422241, + "learning_rate": 0.00013029669716293433, + "loss": 0.7836, + "step": 9045 + }, + { + "epoch": 1.6105769230769231, + "grad_norm": 0.6832906007766724, + "learning_rate": 0.00013028335726040914, + "loss": 1.1531, + "step": 9046 + }, + { + "epoch": 1.6107549857549857, + "grad_norm": 0.6498637795448303, + "learning_rate": 0.00013027001676454446, + "loss": 0.8637, + "step": 9047 + }, + { + "epoch": 1.6109330484330484, + "grad_norm": 0.6792944073677063, + "learning_rate": 0.0001302566756756017, + "loss": 1.0865, + "step": 9048 + }, + { + "epoch": 1.6111111111111112, + "grad_norm": 0.6801337003707886, + "learning_rate": 0.00013024333399384226, + "loss": 1.0738, + "step": 9049 + }, + { + "epoch": 1.611289173789174, + "grad_norm": 0.675216794013977, + "learning_rate": 0.0001302299917195275, + "loss": 1.1074, + "step": 9050 + }, + { + "epoch": 1.6114672364672364, + "grad_norm": 0.6418983340263367, + "learning_rate": 0.00013021664885291885, + "loss": 1.0025, + "step": 9051 + }, + { + "epoch": 1.6116452991452992, + "grad_norm": 0.7778789401054382, + "learning_rate": 0.0001302033053942777, + "loss": 1.0847, + "step": 9052 + }, + { + "epoch": 1.6118233618233617, + "grad_norm": 0.7672827243804932, + "learning_rate": 0.00013018996134386555, + "loss": 1.0565, + "step": 9053 + }, + { + "epoch": 1.6120014245014245, + "grad_norm": 0.6770617961883545, + "learning_rate": 0.00013017661670194382, + "loss": 0.9069, + "step": 9054 + }, + { + "epoch": 1.6121794871794872, + "grad_norm": 0.7161242961883545, + "learning_rate": 0.00013016327146877393, + "loss": 1.1301, + "step": 9055 + }, + { + "epoch": 1.61235754985755, + "grad_norm": 0.6923251152038574, + "learning_rate": 0.00013014992564461746, + "loss": 0.9546, + "step": 9056 + }, + { + "epoch": 1.6125356125356125, + "grad_norm": 0.622953474521637, + "learning_rate": 0.0001301365792297358, + "loss": 0.8152, + "step": 9057 + }, + { + "epoch": 1.6127136752136753, + "grad_norm": 0.7477008104324341, + "learning_rate": 0.00013012323222439046, + "loss": 0.8428, + "step": 9058 + }, + { + "epoch": 1.6128917378917378, + "grad_norm": 0.6612883806228638, + "learning_rate": 0.000130109884628843, + "loss": 1.0678, + "step": 9059 + }, + { + "epoch": 1.6130698005698005, + "grad_norm": 0.6406781077384949, + "learning_rate": 0.00013009653644335486, + "loss": 0.6792, + "step": 9060 + }, + { + "epoch": 1.6132478632478633, + "grad_norm": 0.6279141902923584, + "learning_rate": 0.00013008318766818763, + "loss": 0.9826, + "step": 9061 + }, + { + "epoch": 1.613425925925926, + "grad_norm": 0.6616412401199341, + "learning_rate": 0.00013006983830360285, + "loss": 1.0691, + "step": 9062 + }, + { + "epoch": 1.6136039886039886, + "grad_norm": 0.6520406603813171, + "learning_rate": 0.000130056488349862, + "loss": 0.9487, + "step": 9063 + }, + { + "epoch": 1.6137820512820513, + "grad_norm": 0.6378647089004517, + "learning_rate": 0.00013004313780722672, + "loss": 0.8557, + "step": 9064 + }, + { + "epoch": 1.6139601139601139, + "grad_norm": 0.6547569036483765, + "learning_rate": 0.00013002978667595857, + "loss": 0.879, + "step": 9065 + }, + { + "epoch": 1.6141381766381766, + "grad_norm": 0.7347842454910278, + "learning_rate": 0.00013001643495631914, + "loss": 1.0757, + "step": 9066 + }, + { + "epoch": 1.6143162393162394, + "grad_norm": 0.5988406538963318, + "learning_rate": 0.00013000308264857002, + "loss": 0.6754, + "step": 9067 + }, + { + "epoch": 1.614494301994302, + "grad_norm": 0.6949366331100464, + "learning_rate": 0.00012998972975297282, + "loss": 1.1236, + "step": 9068 + }, + { + "epoch": 1.6146723646723646, + "grad_norm": 0.7095484137535095, + "learning_rate": 0.00012997637626978913, + "loss": 1.0124, + "step": 9069 + }, + { + "epoch": 1.6148504273504274, + "grad_norm": 0.6634095311164856, + "learning_rate": 0.00012996302219928064, + "loss": 1.2018, + "step": 9070 + }, + { + "epoch": 1.61502849002849, + "grad_norm": 0.6894524693489075, + "learning_rate": 0.000129949667541709, + "loss": 0.9959, + "step": 9071 + }, + { + "epoch": 1.6152065527065527, + "grad_norm": 0.672334611415863, + "learning_rate": 0.00012993631229733582, + "loss": 1.0369, + "step": 9072 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.725759744644165, + "learning_rate": 0.00012992295646642278, + "loss": 1.0079, + "step": 9073 + }, + { + "epoch": 1.6155626780626782, + "grad_norm": 0.7941585779190063, + "learning_rate": 0.00012990960004923154, + "loss": 0.9468, + "step": 9074 + }, + { + "epoch": 1.6157407407407407, + "grad_norm": 0.6556950807571411, + "learning_rate": 0.00012989624304602385, + "loss": 0.9915, + "step": 9075 + }, + { + "epoch": 1.6159188034188035, + "grad_norm": 0.7515892386436462, + "learning_rate": 0.0001298828854570614, + "loss": 1.0924, + "step": 9076 + }, + { + "epoch": 1.616096866096866, + "grad_norm": 0.6944101452827454, + "learning_rate": 0.00012986952728260586, + "loss": 0.9632, + "step": 9077 + }, + { + "epoch": 1.6162749287749287, + "grad_norm": 0.6286170482635498, + "learning_rate": 0.000129856168522919, + "loss": 1.0311, + "step": 9078 + }, + { + "epoch": 1.6164529914529915, + "grad_norm": 0.8362757563591003, + "learning_rate": 0.0001298428091782625, + "loss": 1.1232, + "step": 9079 + }, + { + "epoch": 1.6166310541310542, + "grad_norm": 0.6199851632118225, + "learning_rate": 0.0001298294492488982, + "loss": 0.9454, + "step": 9080 + }, + { + "epoch": 1.6168091168091168, + "grad_norm": 0.7541791796684265, + "learning_rate": 0.0001298160887350878, + "loss": 0.9759, + "step": 9081 + }, + { + "epoch": 1.6169871794871795, + "grad_norm": 0.6940878033638, + "learning_rate": 0.00012980272763709304, + "loss": 0.9258, + "step": 9082 + }, + { + "epoch": 1.617165242165242, + "grad_norm": 0.6934045553207397, + "learning_rate": 0.00012978936595517575, + "loss": 1.0142, + "step": 9083 + }, + { + "epoch": 1.6173433048433048, + "grad_norm": 0.8147503733634949, + "learning_rate": 0.00012977600368959774, + "loss": 0.964, + "step": 9084 + }, + { + "epoch": 1.6175213675213675, + "grad_norm": 0.6583107709884644, + "learning_rate": 0.00012976264084062079, + "loss": 1.0315, + "step": 9085 + }, + { + "epoch": 1.6176994301994303, + "grad_norm": 0.7192013263702393, + "learning_rate": 0.0001297492774085067, + "loss": 0.9528, + "step": 9086 + }, + { + "epoch": 1.6178774928774928, + "grad_norm": 0.665888786315918, + "learning_rate": 0.00012973591339351733, + "loss": 1.0188, + "step": 9087 + }, + { + "epoch": 1.6180555555555556, + "grad_norm": 0.7170987725257874, + "learning_rate": 0.0001297225487959145, + "loss": 0.8969, + "step": 9088 + }, + { + "epoch": 1.618233618233618, + "grad_norm": 0.6768732070922852, + "learning_rate": 0.00012970918361596007, + "loss": 1.1951, + "step": 9089 + }, + { + "epoch": 1.6184116809116809, + "grad_norm": 0.6640290021896362, + "learning_rate": 0.00012969581785391592, + "loss": 0.9649, + "step": 9090 + }, + { + "epoch": 1.6185897435897436, + "grad_norm": 0.6200813055038452, + "learning_rate": 0.00012968245151004392, + "loss": 0.9446, + "step": 9091 + }, + { + "epoch": 1.6187678062678064, + "grad_norm": 0.6815837621688843, + "learning_rate": 0.0001296690845846059, + "loss": 1.0506, + "step": 9092 + }, + { + "epoch": 1.618945868945869, + "grad_norm": 0.7252637147903442, + "learning_rate": 0.0001296557170778638, + "loss": 1.1977, + "step": 9093 + }, + { + "epoch": 1.6191239316239316, + "grad_norm": 0.5609107613563538, + "learning_rate": 0.00012964234899007955, + "loss": 0.8009, + "step": 9094 + }, + { + "epoch": 1.6193019943019942, + "grad_norm": 0.6539437770843506, + "learning_rate": 0.00012962898032151506, + "loss": 0.8482, + "step": 9095 + }, + { + "epoch": 1.619480056980057, + "grad_norm": 0.6993300914764404, + "learning_rate": 0.0001296156110724322, + "loss": 1.0725, + "step": 9096 + }, + { + "epoch": 1.6196581196581197, + "grad_norm": 0.6768273711204529, + "learning_rate": 0.000129602241243093, + "loss": 0.9247, + "step": 9097 + }, + { + "epoch": 1.6198361823361824, + "grad_norm": 0.6896265745162964, + "learning_rate": 0.00012958887083375939, + "loss": 0.9526, + "step": 9098 + }, + { + "epoch": 1.6200142450142452, + "grad_norm": 0.7475146651268005, + "learning_rate": 0.00012957549984469327, + "loss": 0.8302, + "step": 9099 + }, + { + "epoch": 1.6201923076923077, + "grad_norm": 0.6622769236564636, + "learning_rate": 0.00012956212827615674, + "loss": 0.9505, + "step": 9100 + }, + { + "epoch": 1.6203703703703702, + "grad_norm": 0.6938058137893677, + "learning_rate": 0.00012954875612841167, + "loss": 0.9757, + "step": 9101 + }, + { + "epoch": 1.620548433048433, + "grad_norm": 0.7453510761260986, + "learning_rate": 0.0001295353834017201, + "loss": 1.0919, + "step": 9102 + }, + { + "epoch": 1.6207264957264957, + "grad_norm": 0.7868932485580444, + "learning_rate": 0.0001295220100963441, + "loss": 0.9265, + "step": 9103 + }, + { + "epoch": 1.6209045584045585, + "grad_norm": 0.6779825091362, + "learning_rate": 0.00012950863621254558, + "loss": 0.98, + "step": 9104 + }, + { + "epoch": 1.6210826210826212, + "grad_norm": 0.6825897097587585, + "learning_rate": 0.00012949526175058662, + "loss": 0.9218, + "step": 9105 + }, + { + "epoch": 1.6212606837606838, + "grad_norm": 0.6686047911643982, + "learning_rate": 0.00012948188671072934, + "loss": 0.9546, + "step": 9106 + }, + { + "epoch": 1.6214387464387463, + "grad_norm": 0.7456090450286865, + "learning_rate": 0.0001294685110932357, + "loss": 1.0819, + "step": 9107 + }, + { + "epoch": 1.621616809116809, + "grad_norm": 0.7111441493034363, + "learning_rate": 0.0001294551348983678, + "loss": 0.9916, + "step": 9108 + }, + { + "epoch": 1.6217948717948718, + "grad_norm": 0.6534699201583862, + "learning_rate": 0.00012944175812638773, + "loss": 1.0374, + "step": 9109 + }, + { + "epoch": 1.6219729344729346, + "grad_norm": 0.6046397089958191, + "learning_rate": 0.00012942838077755758, + "loss": 0.7922, + "step": 9110 + }, + { + "epoch": 1.6221509971509973, + "grad_norm": 0.7736679911613464, + "learning_rate": 0.00012941500285213942, + "loss": 1.0056, + "step": 9111 + }, + { + "epoch": 1.6223290598290598, + "grad_norm": 0.6850929260253906, + "learning_rate": 0.00012940162435039538, + "loss": 0.9538, + "step": 9112 + }, + { + "epoch": 1.6225071225071224, + "grad_norm": 0.6305751800537109, + "learning_rate": 0.00012938824527258756, + "loss": 0.9341, + "step": 9113 + }, + { + "epoch": 1.6226851851851851, + "grad_norm": 0.6740923523902893, + "learning_rate": 0.0001293748656189782, + "loss": 1.0037, + "step": 9114 + }, + { + "epoch": 1.6228632478632479, + "grad_norm": 0.6579762101173401, + "learning_rate": 0.00012936148538982928, + "loss": 1.0022, + "step": 9115 + }, + { + "epoch": 1.6230413105413106, + "grad_norm": 0.6500434279441833, + "learning_rate": 0.0001293481045854031, + "loss": 0.8589, + "step": 9116 + }, + { + "epoch": 1.6232193732193734, + "grad_norm": 0.7825912237167358, + "learning_rate": 0.00012933472320596177, + "loss": 1.0345, + "step": 9117 + }, + { + "epoch": 1.623397435897436, + "grad_norm": 0.8341414332389832, + "learning_rate": 0.0001293213412517675, + "loss": 1.0314, + "step": 9118 + }, + { + "epoch": 1.6235754985754984, + "grad_norm": 0.63664311170578, + "learning_rate": 0.00012930795872308242, + "loss": 0.819, + "step": 9119 + }, + { + "epoch": 1.6237535612535612, + "grad_norm": 0.6800840497016907, + "learning_rate": 0.00012929457562016878, + "loss": 0.95, + "step": 9120 + }, + { + "epoch": 1.623931623931624, + "grad_norm": 0.754165530204773, + "learning_rate": 0.0001292811919432888, + "loss": 1.1193, + "step": 9121 + }, + { + "epoch": 1.6241096866096867, + "grad_norm": 0.678871750831604, + "learning_rate": 0.00012926780769270465, + "loss": 0.9015, + "step": 9122 + }, + { + "epoch": 1.6242877492877494, + "grad_norm": 0.6642945408821106, + "learning_rate": 0.00012925442286867866, + "loss": 0.9095, + "step": 9123 + }, + { + "epoch": 1.624465811965812, + "grad_norm": 0.6089697480201721, + "learning_rate": 0.000129241037471473, + "loss": 0.8994, + "step": 9124 + }, + { + "epoch": 1.6246438746438745, + "grad_norm": 0.7320881485939026, + "learning_rate": 0.00012922765150134995, + "loss": 1.0518, + "step": 9125 + }, + { + "epoch": 1.6248219373219372, + "grad_norm": 0.7308032512664795, + "learning_rate": 0.0001292142649585718, + "loss": 1.0557, + "step": 9126 + }, + { + "epoch": 1.625, + "grad_norm": 0.6896602511405945, + "learning_rate": 0.0001292008778434008, + "loss": 1.145, + "step": 9127 + }, + { + "epoch": 1.6251780626780628, + "grad_norm": 0.6112532615661621, + "learning_rate": 0.00012918749015609926, + "loss": 0.9611, + "step": 9128 + }, + { + "epoch": 1.6253561253561255, + "grad_norm": 0.6856057643890381, + "learning_rate": 0.00012917410189692947, + "loss": 1.0124, + "step": 9129 + }, + { + "epoch": 1.625534188034188, + "grad_norm": 0.699252188205719, + "learning_rate": 0.00012916071306615378, + "loss": 0.8854, + "step": 9130 + }, + { + "epoch": 1.6257122507122506, + "grad_norm": 0.6306683421134949, + "learning_rate": 0.0001291473236640345, + "loss": 1.0722, + "step": 9131 + }, + { + "epoch": 1.6258903133903133, + "grad_norm": 0.6358118653297424, + "learning_rate": 0.00012913393369083393, + "loss": 0.889, + "step": 9132 + }, + { + "epoch": 1.626068376068376, + "grad_norm": 0.6953601837158203, + "learning_rate": 0.00012912054314681445, + "loss": 1.0168, + "step": 9133 + }, + { + "epoch": 1.6262464387464388, + "grad_norm": 0.6742331385612488, + "learning_rate": 0.00012910715203223844, + "loss": 0.8152, + "step": 9134 + }, + { + "epoch": 1.6264245014245016, + "grad_norm": 0.5872861742973328, + "learning_rate": 0.00012909376034736823, + "loss": 0.8702, + "step": 9135 + }, + { + "epoch": 1.626602564102564, + "grad_norm": 0.7580631971359253, + "learning_rate": 0.00012908036809246623, + "loss": 0.994, + "step": 9136 + }, + { + "epoch": 1.6267806267806266, + "grad_norm": 0.7544930577278137, + "learning_rate": 0.00012906697526779488, + "loss": 0.7475, + "step": 9137 + }, + { + "epoch": 1.6269586894586894, + "grad_norm": 0.6850766539573669, + "learning_rate": 0.00012905358187361647, + "loss": 1.0943, + "step": 9138 + }, + { + "epoch": 1.6271367521367521, + "grad_norm": 0.6821565628051758, + "learning_rate": 0.0001290401879101935, + "loss": 1.2928, + "step": 9139 + }, + { + "epoch": 1.6273148148148149, + "grad_norm": 0.6961034536361694, + "learning_rate": 0.00012902679337778835, + "loss": 0.8694, + "step": 9140 + }, + { + "epoch": 1.6274928774928776, + "grad_norm": 0.7159550786018372, + "learning_rate": 0.00012901339827666353, + "loss": 0.8827, + "step": 9141 + }, + { + "epoch": 1.6276709401709402, + "grad_norm": 0.7491081953048706, + "learning_rate": 0.0001290000026070814, + "loss": 0.8159, + "step": 9142 + }, + { + "epoch": 1.6278490028490027, + "grad_norm": 0.7107849717140198, + "learning_rate": 0.00012898660636930447, + "loss": 1.0625, + "step": 9143 + }, + { + "epoch": 1.6280270655270654, + "grad_norm": 0.7227210998535156, + "learning_rate": 0.0001289732095635952, + "loss": 0.9744, + "step": 9144 + }, + { + "epoch": 1.6282051282051282, + "grad_norm": 0.7141995429992676, + "learning_rate": 0.00012895981219021607, + "loss": 0.9836, + "step": 9145 + }, + { + "epoch": 1.628383190883191, + "grad_norm": 0.6445552706718445, + "learning_rate": 0.00012894641424942958, + "loss": 1.0183, + "step": 9146 + }, + { + "epoch": 1.6285612535612537, + "grad_norm": 0.698783278465271, + "learning_rate": 0.00012893301574149824, + "loss": 0.8392, + "step": 9147 + }, + { + "epoch": 1.6287393162393162, + "grad_norm": 0.6529116034507751, + "learning_rate": 0.00012891961666668458, + "loss": 0.9317, + "step": 9148 + }, + { + "epoch": 1.6289173789173788, + "grad_norm": 0.7780548930168152, + "learning_rate": 0.0001289062170252511, + "loss": 1.2406, + "step": 9149 + }, + { + "epoch": 1.6290954415954415, + "grad_norm": 0.6500990986824036, + "learning_rate": 0.0001288928168174603, + "loss": 1.0381, + "step": 9150 + }, + { + "epoch": 1.6292735042735043, + "grad_norm": 0.7098208665847778, + "learning_rate": 0.00012887941604357482, + "loss": 1.2126, + "step": 9151 + }, + { + "epoch": 1.629451566951567, + "grad_norm": 0.730648398399353, + "learning_rate": 0.0001288660147038572, + "loss": 0.8351, + "step": 9152 + }, + { + "epoch": 1.6296296296296298, + "grad_norm": 0.5520278215408325, + "learning_rate": 0.0001288526127985699, + "loss": 0.5877, + "step": 9153 + }, + { + "epoch": 1.6298076923076923, + "grad_norm": 0.7611770033836365, + "learning_rate": 0.00012883921032797563, + "loss": 1.2227, + "step": 9154 + }, + { + "epoch": 1.6299857549857548, + "grad_norm": 0.636820375919342, + "learning_rate": 0.00012882580729233696, + "loss": 0.8305, + "step": 9155 + }, + { + "epoch": 1.6301638176638176, + "grad_norm": 0.694492518901825, + "learning_rate": 0.00012881240369191644, + "loss": 1.0452, + "step": 9156 + }, + { + "epoch": 1.6303418803418803, + "grad_norm": 0.67826908826828, + "learning_rate": 0.00012879899952697677, + "loss": 0.8345, + "step": 9157 + }, + { + "epoch": 1.630519943019943, + "grad_norm": 0.5891323685646057, + "learning_rate": 0.00012878559479778052, + "loss": 0.8367, + "step": 9158 + }, + { + "epoch": 1.6306980056980058, + "grad_norm": 0.6766192317008972, + "learning_rate": 0.0001287721895045903, + "loss": 0.8319, + "step": 9159 + }, + { + "epoch": 1.6308760683760684, + "grad_norm": 0.5306392908096313, + "learning_rate": 0.0001287587836476688, + "loss": 0.7945, + "step": 9160 + }, + { + "epoch": 1.631054131054131, + "grad_norm": 0.6677970290184021, + "learning_rate": 0.0001287453772272787, + "loss": 1.1228, + "step": 9161 + }, + { + "epoch": 1.6312321937321936, + "grad_norm": 0.810052752494812, + "learning_rate": 0.00012873197024368266, + "loss": 0.8395, + "step": 9162 + }, + { + "epoch": 1.6314102564102564, + "grad_norm": 0.7619220018386841, + "learning_rate": 0.00012871856269714333, + "loss": 1.3713, + "step": 9163 + }, + { + "epoch": 1.6315883190883191, + "grad_norm": 0.6564521193504333, + "learning_rate": 0.00012870515458792342, + "loss": 1.0513, + "step": 9164 + }, + { + "epoch": 1.631766381766382, + "grad_norm": 0.6874445676803589, + "learning_rate": 0.00012869174591628564, + "loss": 1.0255, + "step": 9165 + }, + { + "epoch": 1.6319444444444444, + "grad_norm": 0.6958737373352051, + "learning_rate": 0.0001286783366824927, + "loss": 0.9361, + "step": 9166 + }, + { + "epoch": 1.6321225071225072, + "grad_norm": 0.6909199357032776, + "learning_rate": 0.0001286649268868073, + "loss": 0.9855, + "step": 9167 + }, + { + "epoch": 1.6323005698005697, + "grad_norm": 0.7671375274658203, + "learning_rate": 0.00012865151652949225, + "loss": 1.084, + "step": 9168 + }, + { + "epoch": 1.6324786324786325, + "grad_norm": 0.750200092792511, + "learning_rate": 0.00012863810561081023, + "loss": 0.9341, + "step": 9169 + }, + { + "epoch": 1.6326566951566952, + "grad_norm": 0.6595860123634338, + "learning_rate": 0.00012862469413102402, + "loss": 0.9386, + "step": 9170 + }, + { + "epoch": 1.632834757834758, + "grad_norm": 0.622373640537262, + "learning_rate": 0.0001286112820903964, + "loss": 0.7697, + "step": 9171 + }, + { + "epoch": 1.6330128205128205, + "grad_norm": 0.9628498554229736, + "learning_rate": 0.00012859786948919014, + "loss": 1.2629, + "step": 9172 + }, + { + "epoch": 1.6331908831908832, + "grad_norm": 0.7610561847686768, + "learning_rate": 0.000128584456327668, + "loss": 0.9748, + "step": 9173 + }, + { + "epoch": 1.6333689458689458, + "grad_norm": 0.6585374474525452, + "learning_rate": 0.00012857104260609285, + "loss": 0.9049, + "step": 9174 + }, + { + "epoch": 1.6335470085470085, + "grad_norm": 0.6996221542358398, + "learning_rate": 0.00012855762832472746, + "loss": 0.8893, + "step": 9175 + }, + { + "epoch": 1.6337250712250713, + "grad_norm": 0.6226270198822021, + "learning_rate": 0.00012854421348383466, + "loss": 0.8913, + "step": 9176 + }, + { + "epoch": 1.633903133903134, + "grad_norm": 0.6570866107940674, + "learning_rate": 0.00012853079808367731, + "loss": 0.8632, + "step": 9177 + }, + { + "epoch": 1.6340811965811965, + "grad_norm": 0.6899664402008057, + "learning_rate": 0.00012851738212451826, + "loss": 0.8177, + "step": 9178 + }, + { + "epoch": 1.6342592592592593, + "grad_norm": 0.75257807970047, + "learning_rate": 0.0001285039656066203, + "loss": 0.9096, + "step": 9179 + }, + { + "epoch": 1.6344373219373218, + "grad_norm": 0.6614963412284851, + "learning_rate": 0.00012849054853024638, + "loss": 0.9255, + "step": 9180 + }, + { + "epoch": 1.6346153846153846, + "grad_norm": 0.7245957851409912, + "learning_rate": 0.00012847713089565933, + "loss": 1.0122, + "step": 9181 + }, + { + "epoch": 1.6347934472934473, + "grad_norm": 0.7332839369773865, + "learning_rate": 0.00012846371270312204, + "loss": 0.8484, + "step": 9182 + }, + { + "epoch": 1.63497150997151, + "grad_norm": 0.628089189529419, + "learning_rate": 0.00012845029395289748, + "loss": 1.0171, + "step": 9183 + }, + { + "epoch": 1.6351495726495726, + "grad_norm": 0.7493528723716736, + "learning_rate": 0.00012843687464524848, + "loss": 1.1635, + "step": 9184 + }, + { + "epoch": 1.6353276353276354, + "grad_norm": 0.6328163146972656, + "learning_rate": 0.00012842345478043799, + "loss": 1.1254, + "step": 9185 + }, + { + "epoch": 1.635505698005698, + "grad_norm": 0.6720291376113892, + "learning_rate": 0.00012841003435872894, + "loss": 0.9729, + "step": 9186 + }, + { + "epoch": 1.6356837606837606, + "grad_norm": 0.6657332181930542, + "learning_rate": 0.00012839661338038427, + "loss": 1.1047, + "step": 9187 + }, + { + "epoch": 1.6358618233618234, + "grad_norm": 0.7416180968284607, + "learning_rate": 0.000128383191845667, + "loss": 0.9505, + "step": 9188 + }, + { + "epoch": 1.6360398860398861, + "grad_norm": 0.8737816214561462, + "learning_rate": 0.00012836976975484, + "loss": 1.0518, + "step": 9189 + }, + { + "epoch": 1.6362179487179487, + "grad_norm": 0.7351877093315125, + "learning_rate": 0.0001283563471081663, + "loss": 1.1152, + "step": 9190 + }, + { + "epoch": 1.6363960113960114, + "grad_norm": 0.6442788243293762, + "learning_rate": 0.00012834292390590893, + "loss": 0.9432, + "step": 9191 + }, + { + "epoch": 1.636574074074074, + "grad_norm": 0.6848029494285583, + "learning_rate": 0.0001283295001483308, + "loss": 0.8528, + "step": 9192 + }, + { + "epoch": 1.6367521367521367, + "grad_norm": 0.6627060174942017, + "learning_rate": 0.00012831607583569497, + "loss": 1.0222, + "step": 9193 + }, + { + "epoch": 1.6369301994301995, + "grad_norm": 0.7319555878639221, + "learning_rate": 0.00012830265096826446, + "loss": 0.9392, + "step": 9194 + }, + { + "epoch": 1.6371082621082622, + "grad_norm": 0.6986424326896667, + "learning_rate": 0.0001282892255463023, + "loss": 1.2095, + "step": 9195 + }, + { + "epoch": 1.6372863247863247, + "grad_norm": 0.6649929881095886, + "learning_rate": 0.0001282757995700715, + "loss": 0.9426, + "step": 9196 + }, + { + "epoch": 1.6374643874643875, + "grad_norm": 0.6789031624794006, + "learning_rate": 0.0001282623730398352, + "loss": 0.9705, + "step": 9197 + }, + { + "epoch": 1.63764245014245, + "grad_norm": 0.6388779878616333, + "learning_rate": 0.00012824894595585637, + "loss": 1.0698, + "step": 9198 + }, + { + "epoch": 1.6378205128205128, + "grad_norm": 0.636832594871521, + "learning_rate": 0.00012823551831839814, + "loss": 0.9445, + "step": 9199 + }, + { + "epoch": 1.6379985754985755, + "grad_norm": 0.670190691947937, + "learning_rate": 0.0001282220901277236, + "loss": 0.9847, + "step": 9200 + }, + { + "epoch": 1.6381766381766383, + "grad_norm": 0.6020209193229675, + "learning_rate": 0.0001282086613840958, + "loss": 1.0047, + "step": 9201 + }, + { + "epoch": 1.6383547008547008, + "grad_norm": 0.6648211479187012, + "learning_rate": 0.0001281952320877779, + "loss": 0.8717, + "step": 9202 + }, + { + "epoch": 1.6385327635327636, + "grad_norm": 0.7207710146903992, + "learning_rate": 0.000128181802239033, + "loss": 1.1232, + "step": 9203 + }, + { + "epoch": 1.638710826210826, + "grad_norm": 0.800992488861084, + "learning_rate": 0.0001281683718381242, + "loss": 1.0688, + "step": 9204 + }, + { + "epoch": 1.6388888888888888, + "grad_norm": 0.789398193359375, + "learning_rate": 0.0001281549408853147, + "loss": 1.1772, + "step": 9205 + }, + { + "epoch": 1.6390669515669516, + "grad_norm": 0.6514480710029602, + "learning_rate": 0.0001281415093808676, + "loss": 1.1685, + "step": 9206 + }, + { + "epoch": 1.6392450142450143, + "grad_norm": 0.6914686560630798, + "learning_rate": 0.00012812807732504608, + "loss": 1.1307, + "step": 9207 + }, + { + "epoch": 1.6394230769230769, + "grad_norm": 0.6788144111633301, + "learning_rate": 0.00012811464471811334, + "loss": 1.1735, + "step": 9208 + }, + { + "epoch": 1.6396011396011396, + "grad_norm": 0.7049870491027832, + "learning_rate": 0.00012810121156033252, + "loss": 1.0128, + "step": 9209 + }, + { + "epoch": 1.6397792022792022, + "grad_norm": 0.7156766057014465, + "learning_rate": 0.00012808777785196687, + "loss": 0.9503, + "step": 9210 + }, + { + "epoch": 1.639957264957265, + "grad_norm": 0.651716411113739, + "learning_rate": 0.0001280743435932795, + "loss": 1.1227, + "step": 9211 + }, + { + "epoch": 1.6401353276353277, + "grad_norm": 0.7276262044906616, + "learning_rate": 0.0001280609087845337, + "loss": 1.06, + "step": 9212 + }, + { + "epoch": 1.6403133903133904, + "grad_norm": 0.6591095924377441, + "learning_rate": 0.0001280474734259927, + "loss": 1.0861, + "step": 9213 + }, + { + "epoch": 1.640491452991453, + "grad_norm": 0.6675926446914673, + "learning_rate": 0.00012803403751791975, + "loss": 0.9815, + "step": 9214 + }, + { + "epoch": 1.6406695156695157, + "grad_norm": 0.6391474008560181, + "learning_rate": 0.00012802060106057803, + "loss": 0.8027, + "step": 9215 + }, + { + "epoch": 1.6408475783475782, + "grad_norm": 0.6384556293487549, + "learning_rate": 0.00012800716405423086, + "loss": 0.7877, + "step": 9216 + }, + { + "epoch": 1.641025641025641, + "grad_norm": 0.661191463470459, + "learning_rate": 0.00012799372649914146, + "loss": 0.9725, + "step": 9217 + }, + { + "epoch": 1.6412037037037037, + "grad_norm": 0.7418332695960999, + "learning_rate": 0.0001279802883955732, + "loss": 1.1756, + "step": 9218 + }, + { + "epoch": 1.6413817663817665, + "grad_norm": 0.6588954329490662, + "learning_rate": 0.00012796684974378928, + "loss": 1.0428, + "step": 9219 + }, + { + "epoch": 1.6415598290598292, + "grad_norm": 0.7566093802452087, + "learning_rate": 0.000127953410544053, + "loss": 1.1254, + "step": 9220 + }, + { + "epoch": 1.6417378917378918, + "grad_norm": 0.6801039576530457, + "learning_rate": 0.00012793997079662777, + "loss": 1.0854, + "step": 9221 + }, + { + "epoch": 1.6419159544159543, + "grad_norm": 0.7262716889381409, + "learning_rate": 0.0001279265305017768, + "loss": 0.9343, + "step": 9222 + }, + { + "epoch": 1.642094017094017, + "grad_norm": 0.628625750541687, + "learning_rate": 0.0001279130896597635, + "loss": 0.8942, + "step": 9223 + }, + { + "epoch": 1.6422720797720798, + "grad_norm": 0.6183576583862305, + "learning_rate": 0.0001278996482708512, + "loss": 0.9284, + "step": 9224 + }, + { + "epoch": 1.6424501424501425, + "grad_norm": 0.7912000417709351, + "learning_rate": 0.00012788620633530327, + "loss": 1.3043, + "step": 9225 + }, + { + "epoch": 1.6426282051282053, + "grad_norm": 0.6982026100158691, + "learning_rate": 0.00012787276385338298, + "loss": 1.0224, + "step": 9226 + }, + { + "epoch": 1.6428062678062678, + "grad_norm": 0.6734985709190369, + "learning_rate": 0.00012785932082535386, + "loss": 0.8781, + "step": 9227 + }, + { + "epoch": 1.6429843304843303, + "grad_norm": 0.8799532055854797, + "learning_rate": 0.0001278458772514792, + "loss": 1.1482, + "step": 9228 + }, + { + "epoch": 1.643162393162393, + "grad_norm": 0.590295672416687, + "learning_rate": 0.0001278324331320224, + "loss": 0.9502, + "step": 9229 + }, + { + "epoch": 1.6433404558404558, + "grad_norm": 0.6562125086784363, + "learning_rate": 0.0001278189884672469, + "loss": 0.9834, + "step": 9230 + }, + { + "epoch": 1.6435185185185186, + "grad_norm": 0.6848936676979065, + "learning_rate": 0.00012780554325741612, + "loss": 1.0414, + "step": 9231 + }, + { + "epoch": 1.6436965811965814, + "grad_norm": 0.5985032320022583, + "learning_rate": 0.00012779209750279344, + "loss": 0.9469, + "step": 9232 + }, + { + "epoch": 1.6438746438746439, + "grad_norm": 0.7500917911529541, + "learning_rate": 0.00012777865120364238, + "loss": 0.9626, + "step": 9233 + }, + { + "epoch": 1.6440527065527064, + "grad_norm": 0.6565709114074707, + "learning_rate": 0.00012776520436022634, + "loss": 1.0594, + "step": 9234 + }, + { + "epoch": 1.6442307692307692, + "grad_norm": 0.8005441427230835, + "learning_rate": 0.00012775175697280882, + "loss": 1.2379, + "step": 9235 + }, + { + "epoch": 1.644408831908832, + "grad_norm": 0.6734150648117065, + "learning_rate": 0.00012773830904165326, + "loss": 0.9171, + "step": 9236 + }, + { + "epoch": 1.6445868945868947, + "grad_norm": 0.6950868368148804, + "learning_rate": 0.00012772486056702314, + "loss": 1.1782, + "step": 9237 + }, + { + "epoch": 1.6447649572649574, + "grad_norm": 0.8009599447250366, + "learning_rate": 0.000127711411549182, + "loss": 1.0288, + "step": 9238 + }, + { + "epoch": 1.64494301994302, + "grad_norm": 0.6227970719337463, + "learning_rate": 0.0001276979619883933, + "loss": 0.9327, + "step": 9239 + }, + { + "epoch": 1.6451210826210825, + "grad_norm": 0.6828190088272095, + "learning_rate": 0.00012768451188492058, + "loss": 0.9816, + "step": 9240 + }, + { + "epoch": 1.6452991452991452, + "grad_norm": 0.9689767360687256, + "learning_rate": 0.00012767106123902738, + "loss": 0.9049, + "step": 9241 + }, + { + "epoch": 1.645477207977208, + "grad_norm": 0.677061140537262, + "learning_rate": 0.00012765761005097717, + "loss": 0.9472, + "step": 9242 + }, + { + "epoch": 1.6456552706552707, + "grad_norm": 0.7227110862731934, + "learning_rate": 0.00012764415832103356, + "loss": 1.0384, + "step": 9243 + }, + { + "epoch": 1.6458333333333335, + "grad_norm": 0.6540094614028931, + "learning_rate": 0.0001276307060494601, + "loss": 0.8166, + "step": 9244 + }, + { + "epoch": 1.646011396011396, + "grad_norm": 0.6921904683113098, + "learning_rate": 0.00012761725323652033, + "loss": 0.9746, + "step": 9245 + }, + { + "epoch": 1.6461894586894585, + "grad_norm": 0.6742660999298096, + "learning_rate": 0.0001276037998824779, + "loss": 0.8441, + "step": 9246 + }, + { + "epoch": 1.6463675213675213, + "grad_norm": 0.6611103415489197, + "learning_rate": 0.0001275903459875963, + "loss": 1.087, + "step": 9247 + }, + { + "epoch": 1.646545584045584, + "grad_norm": 0.6805498003959656, + "learning_rate": 0.00012757689155213923, + "loss": 0.923, + "step": 9248 + }, + { + "epoch": 1.6467236467236468, + "grad_norm": 0.6598179340362549, + "learning_rate": 0.00012756343657637024, + "loss": 0.9371, + "step": 9249 + }, + { + "epoch": 1.6469017094017095, + "grad_norm": 0.7147273421287537, + "learning_rate": 0.00012754998106055297, + "loss": 1.053, + "step": 9250 + }, + { + "epoch": 1.647079772079772, + "grad_norm": 0.72414630651474, + "learning_rate": 0.00012753652500495103, + "loss": 1.0547, + "step": 9251 + }, + { + "epoch": 1.6472578347578346, + "grad_norm": 0.7784913182258606, + "learning_rate": 0.00012752306840982811, + "loss": 0.9012, + "step": 9252 + }, + { + "epoch": 1.6474358974358974, + "grad_norm": 0.644026517868042, + "learning_rate": 0.0001275096112754478, + "loss": 1.0911, + "step": 9253 + }, + { + "epoch": 1.64761396011396, + "grad_norm": 0.691124677658081, + "learning_rate": 0.00012749615360207382, + "loss": 0.9918, + "step": 9254 + }, + { + "epoch": 1.6477920227920229, + "grad_norm": 0.6632972359657288, + "learning_rate": 0.00012748269538996986, + "loss": 0.9438, + "step": 9255 + }, + { + "epoch": 1.6479700854700856, + "grad_norm": 0.6548733115196228, + "learning_rate": 0.00012746923663939955, + "loss": 1.1082, + "step": 9256 + }, + { + "epoch": 1.6481481481481481, + "grad_norm": 0.6737542748451233, + "learning_rate": 0.00012745577735062664, + "loss": 0.9255, + "step": 9257 + }, + { + "epoch": 1.6483262108262107, + "grad_norm": 0.686862051486969, + "learning_rate": 0.00012744231752391479, + "loss": 0.9493, + "step": 9258 + }, + { + "epoch": 1.6485042735042734, + "grad_norm": 0.6096474528312683, + "learning_rate": 0.00012742885715952772, + "loss": 0.6849, + "step": 9259 + }, + { + "epoch": 1.6486823361823362, + "grad_norm": 0.702751636505127, + "learning_rate": 0.00012741539625772918, + "loss": 1.0335, + "step": 9260 + }, + { + "epoch": 1.648860398860399, + "grad_norm": 0.7470958232879639, + "learning_rate": 0.0001274019348187829, + "loss": 1.105, + "step": 9261 + }, + { + "epoch": 1.6490384615384617, + "grad_norm": 0.6642739176750183, + "learning_rate": 0.0001273884728429526, + "loss": 1.01, + "step": 9262 + }, + { + "epoch": 1.6492165242165242, + "grad_norm": 0.6470904350280762, + "learning_rate": 0.00012737501033050213, + "loss": 0.9009, + "step": 9263 + }, + { + "epoch": 1.6493945868945867, + "grad_norm": 0.7487246990203857, + "learning_rate": 0.00012736154728169518, + "loss": 0.9832, + "step": 9264 + }, + { + "epoch": 1.6495726495726495, + "grad_norm": 0.7370779514312744, + "learning_rate": 0.00012734808369679553, + "loss": 1.0464, + "step": 9265 + }, + { + "epoch": 1.6497507122507122, + "grad_norm": 0.7942814826965332, + "learning_rate": 0.00012733461957606702, + "loss": 1.102, + "step": 9266 + }, + { + "epoch": 1.649928774928775, + "grad_norm": 0.6535606980323792, + "learning_rate": 0.00012732115491977336, + "loss": 1.0655, + "step": 9267 + }, + { + "epoch": 1.6501068376068377, + "grad_norm": 0.601716935634613, + "learning_rate": 0.00012730768972817847, + "loss": 0.8236, + "step": 9268 + }, + { + "epoch": 1.6502849002849003, + "grad_norm": 0.7375118732452393, + "learning_rate": 0.00012729422400154614, + "loss": 0.9313, + "step": 9269 + }, + { + "epoch": 1.6504629629629628, + "grad_norm": 0.7360411882400513, + "learning_rate": 0.00012728075774014018, + "loss": 0.9254, + "step": 9270 + }, + { + "epoch": 1.6506410256410255, + "grad_norm": 0.8453929424285889, + "learning_rate": 0.00012726729094422444, + "loss": 1.0975, + "step": 9271 + }, + { + "epoch": 1.6508190883190883, + "grad_norm": 0.5615501999855042, + "learning_rate": 0.00012725382361406274, + "loss": 0.8243, + "step": 9272 + }, + { + "epoch": 1.650997150997151, + "grad_norm": 0.6494898796081543, + "learning_rate": 0.000127240355749919, + "loss": 0.9766, + "step": 9273 + }, + { + "epoch": 1.6511752136752138, + "grad_norm": 0.6544778347015381, + "learning_rate": 0.0001272268873520571, + "loss": 0.9969, + "step": 9274 + }, + { + "epoch": 1.6513532763532763, + "grad_norm": 0.6937400698661804, + "learning_rate": 0.00012721341842074092, + "loss": 1.0626, + "step": 9275 + }, + { + "epoch": 1.651531339031339, + "grad_norm": 0.7068421244621277, + "learning_rate": 0.0001271999489562343, + "loss": 1.0068, + "step": 9276 + }, + { + "epoch": 1.6517094017094016, + "grad_norm": 0.6425052285194397, + "learning_rate": 0.0001271864789588012, + "loss": 0.8716, + "step": 9277 + }, + { + "epoch": 1.6518874643874644, + "grad_norm": 0.6895090341567993, + "learning_rate": 0.0001271730084287055, + "loss": 1.081, + "step": 9278 + }, + { + "epoch": 1.6520655270655271, + "grad_norm": 0.6773712038993835, + "learning_rate": 0.00012715953736621116, + "loss": 0.7586, + "step": 9279 + }, + { + "epoch": 1.6522435897435899, + "grad_norm": 0.6085716485977173, + "learning_rate": 0.0001271460657715821, + "loss": 0.8627, + "step": 9280 + }, + { + "epoch": 1.6524216524216524, + "grad_norm": 0.6415461897850037, + "learning_rate": 0.00012713259364508227, + "loss": 0.9751, + "step": 9281 + }, + { + "epoch": 1.6525997150997151, + "grad_norm": 0.6460939645767212, + "learning_rate": 0.00012711912098697565, + "loss": 0.9578, + "step": 9282 + }, + { + "epoch": 1.6527777777777777, + "grad_norm": 0.6076797246932983, + "learning_rate": 0.00012710564779752615, + "loss": 0.9627, + "step": 9283 + }, + { + "epoch": 1.6529558404558404, + "grad_norm": 0.710782527923584, + "learning_rate": 0.00012709217407699783, + "loss": 0.8725, + "step": 9284 + }, + { + "epoch": 1.6531339031339032, + "grad_norm": 0.6793623566627502, + "learning_rate": 0.00012707869982565463, + "loss": 0.908, + "step": 9285 + }, + { + "epoch": 1.653311965811966, + "grad_norm": 0.6841681003570557, + "learning_rate": 0.00012706522504376055, + "loss": 0.8546, + "step": 9286 + }, + { + "epoch": 1.6534900284900285, + "grad_norm": 0.7908675670623779, + "learning_rate": 0.0001270517497315796, + "loss": 0.9409, + "step": 9287 + }, + { + "epoch": 1.6536680911680912, + "grad_norm": 0.6918683648109436, + "learning_rate": 0.0001270382738893758, + "loss": 1.0493, + "step": 9288 + }, + { + "epoch": 1.6538461538461537, + "grad_norm": 0.6891819834709167, + "learning_rate": 0.00012702479751741322, + "loss": 1.0675, + "step": 9289 + }, + { + "epoch": 1.6540242165242165, + "grad_norm": 0.6965166926383972, + "learning_rate": 0.00012701132061595586, + "loss": 0.8563, + "step": 9290 + }, + { + "epoch": 1.6542022792022792, + "grad_norm": 0.7549001574516296, + "learning_rate": 0.00012699784318526779, + "loss": 1.1572, + "step": 9291 + }, + { + "epoch": 1.654380341880342, + "grad_norm": 0.6100513339042664, + "learning_rate": 0.00012698436522561303, + "loss": 0.897, + "step": 9292 + }, + { + "epoch": 1.6545584045584045, + "grad_norm": 0.6477037668228149, + "learning_rate": 0.00012697088673725574, + "loss": 0.7961, + "step": 9293 + }, + { + "epoch": 1.6547364672364673, + "grad_norm": 0.7402619123458862, + "learning_rate": 0.0001269574077204599, + "loss": 1.2001, + "step": 9294 + }, + { + "epoch": 1.6549145299145298, + "grad_norm": 0.7162346243858337, + "learning_rate": 0.0001269439281754897, + "loss": 0.9963, + "step": 9295 + }, + { + "epoch": 1.6550925925925926, + "grad_norm": 0.6757413744926453, + "learning_rate": 0.0001269304481026092, + "loss": 1.0476, + "step": 9296 + }, + { + "epoch": 1.6552706552706553, + "grad_norm": 0.6455655097961426, + "learning_rate": 0.0001269169675020825, + "loss": 0.9716, + "step": 9297 + }, + { + "epoch": 1.655448717948718, + "grad_norm": 0.7705031037330627, + "learning_rate": 0.0001269034863741737, + "loss": 0.9886, + "step": 9298 + }, + { + "epoch": 1.6556267806267806, + "grad_norm": 0.6084272861480713, + "learning_rate": 0.000126890004719147, + "loss": 0.8231, + "step": 9299 + }, + { + "epoch": 1.6558048433048433, + "grad_norm": 0.7051045298576355, + "learning_rate": 0.00012687652253726652, + "loss": 0.8673, + "step": 9300 + }, + { + "epoch": 1.6559829059829059, + "grad_norm": 0.731675386428833, + "learning_rate": 0.0001268630398287964, + "loss": 0.8609, + "step": 9301 + }, + { + "epoch": 1.6561609686609686, + "grad_norm": 0.6796799302101135, + "learning_rate": 0.00012684955659400087, + "loss": 1.0157, + "step": 9302 + }, + { + "epoch": 1.6563390313390314, + "grad_norm": 0.6270264983177185, + "learning_rate": 0.000126836072833144, + "loss": 0.8924, + "step": 9303 + }, + { + "epoch": 1.6565170940170941, + "grad_norm": 0.7235464453697205, + "learning_rate": 0.00012682258854649004, + "loss": 0.8904, + "step": 9304 + }, + { + "epoch": 1.6566951566951567, + "grad_norm": 0.7644724249839783, + "learning_rate": 0.00012680910373430318, + "loss": 0.9119, + "step": 9305 + }, + { + "epoch": 1.6568732193732194, + "grad_norm": 0.661411702632904, + "learning_rate": 0.00012679561839684764, + "loss": 1.0066, + "step": 9306 + }, + { + "epoch": 1.657051282051282, + "grad_norm": 0.6981723308563232, + "learning_rate": 0.0001267821325343876, + "loss": 1.2579, + "step": 9307 + }, + { + "epoch": 1.6572293447293447, + "grad_norm": 0.6469807028770447, + "learning_rate": 0.0001267686461471873, + "loss": 0.8678, + "step": 9308 + }, + { + "epoch": 1.6574074074074074, + "grad_norm": 0.8255495429039001, + "learning_rate": 0.000126755159235511, + "loss": 0.9053, + "step": 9309 + }, + { + "epoch": 1.6575854700854702, + "grad_norm": 0.6882261037826538, + "learning_rate": 0.00012674167179962294, + "loss": 0.8364, + "step": 9310 + }, + { + "epoch": 1.6577635327635327, + "grad_norm": 0.6816701889038086, + "learning_rate": 0.00012672818383978733, + "loss": 0.9627, + "step": 9311 + }, + { + "epoch": 1.6579415954415955, + "grad_norm": 0.6993424892425537, + "learning_rate": 0.00012671469535626852, + "loss": 0.8337, + "step": 9312 + }, + { + "epoch": 1.658119658119658, + "grad_norm": 0.6271458864212036, + "learning_rate": 0.00012670120634933075, + "loss": 0.8322, + "step": 9313 + }, + { + "epoch": 1.6582977207977208, + "grad_norm": 0.7012003660202026, + "learning_rate": 0.00012668771681923827, + "loss": 0.8895, + "step": 9314 + }, + { + "epoch": 1.6584757834757835, + "grad_norm": 0.6704670190811157, + "learning_rate": 0.00012667422676625547, + "loss": 1.0544, + "step": 9315 + }, + { + "epoch": 1.6586538461538463, + "grad_norm": 0.6189491748809814, + "learning_rate": 0.0001266607361906466, + "loss": 0.9623, + "step": 9316 + }, + { + "epoch": 1.6588319088319088, + "grad_norm": 0.7065694332122803, + "learning_rate": 0.000126647245092676, + "loss": 0.8874, + "step": 9317 + }, + { + "epoch": 1.6590099715099715, + "grad_norm": 0.7473452687263489, + "learning_rate": 0.00012663375347260795, + "loss": 1.0576, + "step": 9318 + }, + { + "epoch": 1.659188034188034, + "grad_norm": 0.6839408874511719, + "learning_rate": 0.0001266202613307068, + "loss": 0.9127, + "step": 9319 + }, + { + "epoch": 1.6593660968660968, + "grad_norm": 0.7154020071029663, + "learning_rate": 0.00012660676866723699, + "loss": 1.1174, + "step": 9320 + }, + { + "epoch": 1.6595441595441596, + "grad_norm": 0.7123729586601257, + "learning_rate": 0.0001265932754824628, + "loss": 0.9617, + "step": 9321 + }, + { + "epoch": 1.6597222222222223, + "grad_norm": 0.7537810802459717, + "learning_rate": 0.0001265797817766486, + "loss": 1.0333, + "step": 9322 + }, + { + "epoch": 1.6599002849002849, + "grad_norm": 0.706551730632782, + "learning_rate": 0.00012656628755005884, + "loss": 1.0838, + "step": 9323 + }, + { + "epoch": 1.6600783475783476, + "grad_norm": 0.8104004859924316, + "learning_rate": 0.0001265527928029578, + "loss": 0.9807, + "step": 9324 + }, + { + "epoch": 1.6602564102564101, + "grad_norm": 0.6892881989479065, + "learning_rate": 0.00012653929753560998, + "loss": 0.9941, + "step": 9325 + }, + { + "epoch": 1.6604344729344729, + "grad_norm": 0.5919203758239746, + "learning_rate": 0.00012652580174827974, + "loss": 0.9268, + "step": 9326 + }, + { + "epoch": 1.6606125356125356, + "grad_norm": 0.6715863347053528, + "learning_rate": 0.00012651230544123154, + "loss": 1.0912, + "step": 9327 + }, + { + "epoch": 1.6607905982905984, + "grad_norm": 0.6765137314796448, + "learning_rate": 0.0001264988086147298, + "loss": 1.1576, + "step": 9328 + }, + { + "epoch": 1.660968660968661, + "grad_norm": 0.6781638860702515, + "learning_rate": 0.00012648531126903888, + "loss": 1.1162, + "step": 9329 + }, + { + "epoch": 1.6611467236467237, + "grad_norm": 0.715871274471283, + "learning_rate": 0.00012647181340442337, + "loss": 0.714, + "step": 9330 + }, + { + "epoch": 1.6613247863247862, + "grad_norm": 0.6237258315086365, + "learning_rate": 0.00012645831502114762, + "loss": 0.8512, + "step": 9331 + }, + { + "epoch": 1.661502849002849, + "grad_norm": 0.6668339967727661, + "learning_rate": 0.0001264448161194762, + "loss": 1.0384, + "step": 9332 + }, + { + "epoch": 1.6616809116809117, + "grad_norm": 0.8316730260848999, + "learning_rate": 0.00012643131669967352, + "loss": 0.8931, + "step": 9333 + }, + { + "epoch": 1.6618589743589745, + "grad_norm": 0.7013183832168579, + "learning_rate": 0.00012641781676200406, + "loss": 1.0548, + "step": 9334 + }, + { + "epoch": 1.6620370370370372, + "grad_norm": 0.6980466842651367, + "learning_rate": 0.00012640431630673243, + "loss": 0.8988, + "step": 9335 + }, + { + "epoch": 1.6622150997150997, + "grad_norm": 0.7045995593070984, + "learning_rate": 0.000126390815334123, + "loss": 1.107, + "step": 9336 + }, + { + "epoch": 1.6623931623931623, + "grad_norm": 0.6699773669242859, + "learning_rate": 0.00012637731384444043, + "loss": 1.1757, + "step": 9337 + }, + { + "epoch": 1.662571225071225, + "grad_norm": 0.6489999294281006, + "learning_rate": 0.00012636381183794916, + "loss": 0.9282, + "step": 9338 + }, + { + "epoch": 1.6627492877492878, + "grad_norm": 0.7085952758789062, + "learning_rate": 0.00012635030931491375, + "loss": 1.0221, + "step": 9339 + }, + { + "epoch": 1.6629273504273505, + "grad_norm": 0.6893135905265808, + "learning_rate": 0.00012633680627559878, + "loss": 1.0517, + "step": 9340 + }, + { + "epoch": 1.6631054131054133, + "grad_norm": 0.5659682154655457, + "learning_rate": 0.00012632330272026882, + "loss": 0.6294, + "step": 9341 + }, + { + "epoch": 1.6632834757834758, + "grad_norm": 0.6889018416404724, + "learning_rate": 0.00012630979864918838, + "loss": 1.0735, + "step": 9342 + }, + { + "epoch": 1.6634615384615383, + "grad_norm": 0.7333424687385559, + "learning_rate": 0.00012629629406262212, + "loss": 0.9079, + "step": 9343 + }, + { + "epoch": 1.663639601139601, + "grad_norm": 0.6340580582618713, + "learning_rate": 0.00012628278896083462, + "loss": 0.9738, + "step": 9344 + }, + { + "epoch": 1.6638176638176638, + "grad_norm": 0.7042564749717712, + "learning_rate": 0.00012626928334409044, + "loss": 0.959, + "step": 9345 + }, + { + "epoch": 1.6639957264957266, + "grad_norm": 0.711757242679596, + "learning_rate": 0.00012625577721265424, + "loss": 0.8113, + "step": 9346 + }, + { + "epoch": 1.6641737891737893, + "grad_norm": 0.7723299264907837, + "learning_rate": 0.0001262422705667906, + "loss": 1.1724, + "step": 9347 + }, + { + "epoch": 1.6643518518518519, + "grad_norm": 0.711334228515625, + "learning_rate": 0.00012622876340676422, + "loss": 1.0121, + "step": 9348 + }, + { + "epoch": 1.6645299145299144, + "grad_norm": 0.6954590678215027, + "learning_rate": 0.0001262152557328397, + "loss": 1.2093, + "step": 9349 + }, + { + "epoch": 1.6647079772079771, + "grad_norm": 0.6341620087623596, + "learning_rate": 0.00012620174754528166, + "loss": 1.0535, + "step": 9350 + }, + { + "epoch": 1.66488603988604, + "grad_norm": 0.6434268355369568, + "learning_rate": 0.00012618823884435484, + "loss": 0.8964, + "step": 9351 + }, + { + "epoch": 1.6650641025641026, + "grad_norm": 0.7685084939002991, + "learning_rate": 0.00012617472963032385, + "loss": 1.0639, + "step": 9352 + }, + { + "epoch": 1.6652421652421654, + "grad_norm": 0.6347958445549011, + "learning_rate": 0.00012616121990345345, + "loss": 1.0252, + "step": 9353 + }, + { + "epoch": 1.665420227920228, + "grad_norm": 0.647722601890564, + "learning_rate": 0.0001261477096640083, + "loss": 0.9527, + "step": 9354 + }, + { + "epoch": 1.6655982905982905, + "grad_norm": 0.5942047834396362, + "learning_rate": 0.000126134198912253, + "loss": 1.0062, + "step": 9355 + }, + { + "epoch": 1.6657763532763532, + "grad_norm": 0.683555006980896, + "learning_rate": 0.00012612068764845247, + "loss": 0.8101, + "step": 9356 + }, + { + "epoch": 1.665954415954416, + "grad_norm": 0.6832289099693298, + "learning_rate": 0.00012610717587287128, + "loss": 1.1436, + "step": 9357 + }, + { + "epoch": 1.6661324786324787, + "grad_norm": 0.7035253047943115, + "learning_rate": 0.00012609366358577422, + "loss": 0.9724, + "step": 9358 + }, + { + "epoch": 1.6663105413105415, + "grad_norm": 0.6471409797668457, + "learning_rate": 0.00012608015078742604, + "loss": 0.776, + "step": 9359 + }, + { + "epoch": 1.666488603988604, + "grad_norm": 0.7069687247276306, + "learning_rate": 0.00012606663747809145, + "loss": 0.9667, + "step": 9360 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.6744135618209839, + "learning_rate": 0.00012605312365803525, + "loss": 1.1152, + "step": 9361 + }, + { + "epoch": 1.6668447293447293, + "grad_norm": 0.7212334275245667, + "learning_rate": 0.00012603960932752227, + "loss": 1.1543, + "step": 9362 + }, + { + "epoch": 1.667022792022792, + "grad_norm": 0.6501669883728027, + "learning_rate": 0.0001260260944868172, + "loss": 0.8595, + "step": 9363 + }, + { + "epoch": 1.6672008547008548, + "grad_norm": 0.6970864534378052, + "learning_rate": 0.00012601257913618486, + "loss": 0.9364, + "step": 9364 + }, + { + "epoch": 1.6673789173789175, + "grad_norm": 0.6802223324775696, + "learning_rate": 0.00012599906327589007, + "loss": 0.8429, + "step": 9365 + }, + { + "epoch": 1.66755698005698, + "grad_norm": 0.6842933893203735, + "learning_rate": 0.00012598554690619764, + "loss": 1.1255, + "step": 9366 + }, + { + "epoch": 1.6677350427350426, + "grad_norm": 0.6547088623046875, + "learning_rate": 0.0001259720300273724, + "loss": 0.983, + "step": 9367 + }, + { + "epoch": 1.6679131054131053, + "grad_norm": 0.620424211025238, + "learning_rate": 0.0001259585126396792, + "loss": 0.918, + "step": 9368 + }, + { + "epoch": 1.668091168091168, + "grad_norm": 0.5659816861152649, + "learning_rate": 0.00012594499474338287, + "loss": 0.7788, + "step": 9369 + }, + { + "epoch": 1.6682692307692308, + "grad_norm": 0.5904595255851746, + "learning_rate": 0.00012593147633874826, + "loss": 0.801, + "step": 9370 + }, + { + "epoch": 1.6684472934472936, + "grad_norm": 0.6444024443626404, + "learning_rate": 0.0001259179574260402, + "loss": 1.0997, + "step": 9371 + }, + { + "epoch": 1.6686253561253561, + "grad_norm": 0.6408827304840088, + "learning_rate": 0.00012590443800552365, + "loss": 0.9839, + "step": 9372 + }, + { + "epoch": 1.6688034188034186, + "grad_norm": 0.752391517162323, + "learning_rate": 0.00012589091807746345, + "loss": 1.0249, + "step": 9373 + }, + { + "epoch": 1.6689814814814814, + "grad_norm": 0.8256397247314453, + "learning_rate": 0.00012587739764212448, + "loss": 0.9541, + "step": 9374 + }, + { + "epoch": 1.6691595441595442, + "grad_norm": 0.7878768444061279, + "learning_rate": 0.00012586387669977166, + "loss": 1.0071, + "step": 9375 + }, + { + "epoch": 1.669337606837607, + "grad_norm": 0.6179735660552979, + "learning_rate": 0.0001258503552506699, + "loss": 0.8495, + "step": 9376 + }, + { + "epoch": 1.6695156695156697, + "grad_norm": 0.6699580550193787, + "learning_rate": 0.00012583683329508413, + "loss": 0.8999, + "step": 9377 + }, + { + "epoch": 1.6696937321937322, + "grad_norm": 0.6542006731033325, + "learning_rate": 0.00012582331083327929, + "loss": 1.0357, + "step": 9378 + }, + { + "epoch": 1.6698717948717947, + "grad_norm": 0.7275210618972778, + "learning_rate": 0.0001258097878655203, + "loss": 1.0259, + "step": 9379 + }, + { + "epoch": 1.6700498575498575, + "grad_norm": 0.6836326122283936, + "learning_rate": 0.00012579626439207216, + "loss": 1.0428, + "step": 9380 + }, + { + "epoch": 1.6702279202279202, + "grad_norm": 0.760123610496521, + "learning_rate": 0.00012578274041319978, + "loss": 0.9716, + "step": 9381 + }, + { + "epoch": 1.670405982905983, + "grad_norm": 0.5525194406509399, + "learning_rate": 0.00012576921592916818, + "loss": 0.8253, + "step": 9382 + }, + { + "epoch": 1.6705840455840457, + "grad_norm": 0.6881270408630371, + "learning_rate": 0.00012575569094024232, + "loss": 1.0571, + "step": 9383 + }, + { + "epoch": 1.6707621082621082, + "grad_norm": 0.6776245832443237, + "learning_rate": 0.0001257421654466872, + "loss": 0.9119, + "step": 9384 + }, + { + "epoch": 1.6709401709401708, + "grad_norm": 0.7903014421463013, + "learning_rate": 0.0001257286394487678, + "loss": 1.0626, + "step": 9385 + }, + { + "epoch": 1.6711182336182335, + "grad_norm": 0.61158287525177, + "learning_rate": 0.0001257151129467492, + "loss": 0.9378, + "step": 9386 + }, + { + "epoch": 1.6712962962962963, + "grad_norm": 0.655189573764801, + "learning_rate": 0.00012570158594089637, + "loss": 0.9334, + "step": 9387 + }, + { + "epoch": 1.671474358974359, + "grad_norm": 0.6707320809364319, + "learning_rate": 0.0001256880584314743, + "loss": 1.1802, + "step": 9388 + }, + { + "epoch": 1.6716524216524218, + "grad_norm": 0.847341775894165, + "learning_rate": 0.00012567453041874814, + "loss": 1.1169, + "step": 9389 + }, + { + "epoch": 1.6718304843304843, + "grad_norm": 0.6136410236358643, + "learning_rate": 0.00012566100190298287, + "loss": 0.8959, + "step": 9390 + }, + { + "epoch": 1.672008547008547, + "grad_norm": 0.7203437089920044, + "learning_rate": 0.00012564747288444357, + "loss": 0.9803, + "step": 9391 + }, + { + "epoch": 1.6721866096866096, + "grad_norm": 0.7832576632499695, + "learning_rate": 0.00012563394336339534, + "loss": 0.8696, + "step": 9392 + }, + { + "epoch": 1.6723646723646723, + "grad_norm": 0.6940804719924927, + "learning_rate": 0.00012562041334010323, + "loss": 1.0571, + "step": 9393 + }, + { + "epoch": 1.672542735042735, + "grad_norm": 0.6042298674583435, + "learning_rate": 0.00012560688281483234, + "loss": 0.8835, + "step": 9394 + }, + { + "epoch": 1.6727207977207978, + "grad_norm": 0.7870675921440125, + "learning_rate": 0.00012559335178784776, + "loss": 1.1585, + "step": 9395 + }, + { + "epoch": 1.6728988603988604, + "grad_norm": 0.7448568940162659, + "learning_rate": 0.00012557982025941463, + "loss": 0.9699, + "step": 9396 + }, + { + "epoch": 1.6730769230769231, + "grad_norm": 0.7226544618606567, + "learning_rate": 0.00012556628822979807, + "loss": 0.7817, + "step": 9397 + }, + { + "epoch": 1.6732549857549857, + "grad_norm": 0.5652043223381042, + "learning_rate": 0.0001255527556992632, + "loss": 0.8077, + "step": 9398 + }, + { + "epoch": 1.6734330484330484, + "grad_norm": 0.6459930539131165, + "learning_rate": 0.00012553922266807517, + "loss": 1.22, + "step": 9399 + }, + { + "epoch": 1.6736111111111112, + "grad_norm": 0.7568991780281067, + "learning_rate": 0.00012552568913649912, + "loss": 1.1559, + "step": 9400 + }, + { + "epoch": 1.673789173789174, + "grad_norm": 0.7462680339813232, + "learning_rate": 0.0001255121551048002, + "loss": 1.1438, + "step": 9401 + }, + { + "epoch": 1.6739672364672364, + "grad_norm": 0.6653871536254883, + "learning_rate": 0.0001254986205732436, + "loss": 0.9468, + "step": 9402 + }, + { + "epoch": 1.6741452991452992, + "grad_norm": 0.6261825561523438, + "learning_rate": 0.0001254850855420945, + "loss": 0.8558, + "step": 9403 + }, + { + "epoch": 1.6743233618233617, + "grad_norm": 0.6442354321479797, + "learning_rate": 0.0001254715500116181, + "loss": 0.8605, + "step": 9404 + }, + { + "epoch": 1.6745014245014245, + "grad_norm": 0.7483665943145752, + "learning_rate": 0.00012545801398207958, + "loss": 0.9089, + "step": 9405 + }, + { + "epoch": 1.6746794871794872, + "grad_norm": 0.7319819927215576, + "learning_rate": 0.00012544447745374416, + "loss": 0.9937, + "step": 9406 + }, + { + "epoch": 1.67485754985755, + "grad_norm": 0.703014075756073, + "learning_rate": 0.00012543094042687708, + "loss": 0.9597, + "step": 9407 + }, + { + "epoch": 1.6750356125356125, + "grad_norm": 0.6593887209892273, + "learning_rate": 0.00012541740290174353, + "loss": 0.844, + "step": 9408 + }, + { + "epoch": 1.6752136752136753, + "grad_norm": 0.6567463874816895, + "learning_rate": 0.00012540386487860879, + "loss": 1.0744, + "step": 9409 + }, + { + "epoch": 1.6753917378917378, + "grad_norm": 0.7784611582756042, + "learning_rate": 0.00012539032635773805, + "loss": 0.974, + "step": 9410 + }, + { + "epoch": 1.6755698005698005, + "grad_norm": 0.6760087609291077, + "learning_rate": 0.00012537678733939663, + "loss": 0.8948, + "step": 9411 + }, + { + "epoch": 1.6757478632478633, + "grad_norm": 0.825965940952301, + "learning_rate": 0.0001253632478238498, + "loss": 1.1196, + "step": 9412 + }, + { + "epoch": 1.675925925925926, + "grad_norm": 0.7215564250946045, + "learning_rate": 0.00012534970781136277, + "loss": 1.1774, + "step": 9413 + }, + { + "epoch": 1.6761039886039886, + "grad_norm": 0.6548578143119812, + "learning_rate": 0.00012533616730220094, + "loss": 0.8671, + "step": 9414 + }, + { + "epoch": 1.6762820512820513, + "grad_norm": 0.7257684469223022, + "learning_rate": 0.00012532262629662947, + "loss": 1.105, + "step": 9415 + }, + { + "epoch": 1.6764601139601139, + "grad_norm": 0.6695847511291504, + "learning_rate": 0.00012530908479491378, + "loss": 0.9189, + "step": 9416 + }, + { + "epoch": 1.6766381766381766, + "grad_norm": 0.684695303440094, + "learning_rate": 0.00012529554279731915, + "loss": 1.066, + "step": 9417 + }, + { + "epoch": 1.6768162393162394, + "grad_norm": 0.7107276320457458, + "learning_rate": 0.0001252820003041109, + "loss": 0.9311, + "step": 9418 + }, + { + "epoch": 1.676994301994302, + "grad_norm": 0.6755440831184387, + "learning_rate": 0.0001252684573155544, + "loss": 1.1036, + "step": 9419 + }, + { + "epoch": 1.6771723646723646, + "grad_norm": 0.7571110725402832, + "learning_rate": 0.00012525491383191491, + "loss": 1.0244, + "step": 9420 + }, + { + "epoch": 1.6773504273504274, + "grad_norm": 0.6960614323616028, + "learning_rate": 0.0001252413698534579, + "loss": 0.9077, + "step": 9421 + }, + { + "epoch": 1.67752849002849, + "grad_norm": 0.6675550937652588, + "learning_rate": 0.00012522782538044867, + "loss": 1.0543, + "step": 9422 + }, + { + "epoch": 1.6777065527065527, + "grad_norm": 0.6637391448020935, + "learning_rate": 0.0001252142804131526, + "loss": 0.9471, + "step": 9423 + }, + { + "epoch": 1.6778846153846154, + "grad_norm": 0.6382880210876465, + "learning_rate": 0.00012520073495183508, + "loss": 0.9729, + "step": 9424 + }, + { + "epoch": 1.6780626780626782, + "grad_norm": 0.731922447681427, + "learning_rate": 0.0001251871889967615, + "loss": 1.0385, + "step": 9425 + }, + { + "epoch": 1.6782407407407407, + "grad_norm": 0.5868890285491943, + "learning_rate": 0.00012517364254819728, + "loss": 0.8466, + "step": 9426 + }, + { + "epoch": 1.6784188034188035, + "grad_norm": 0.8535677790641785, + "learning_rate": 0.00012516009560640786, + "loss": 1.1009, + "step": 9427 + }, + { + "epoch": 1.678596866096866, + "grad_norm": 0.7044199705123901, + "learning_rate": 0.0001251465481716586, + "loss": 1.0862, + "step": 9428 + }, + { + "epoch": 1.6787749287749287, + "grad_norm": 0.7207323312759399, + "learning_rate": 0.00012513300024421498, + "loss": 1.064, + "step": 9429 + }, + { + "epoch": 1.6789529914529915, + "grad_norm": 0.7739703059196472, + "learning_rate": 0.0001251194518243424, + "loss": 1.1738, + "step": 9430 + }, + { + "epoch": 1.6791310541310542, + "grad_norm": 0.6829344630241394, + "learning_rate": 0.00012510590291230637, + "loss": 1.0555, + "step": 9431 + }, + { + "epoch": 1.6793091168091168, + "grad_norm": 0.6760238409042358, + "learning_rate": 0.0001250923535083723, + "loss": 1.2177, + "step": 9432 + }, + { + "epoch": 1.6794871794871795, + "grad_norm": 0.6666911840438843, + "learning_rate": 0.0001250788036128057, + "loss": 0.8957, + "step": 9433 + }, + { + "epoch": 1.679665242165242, + "grad_norm": 0.747797429561615, + "learning_rate": 0.00012506525322587207, + "loss": 0.9793, + "step": 9434 + }, + { + "epoch": 1.6798433048433048, + "grad_norm": 0.6261107325553894, + "learning_rate": 0.00012505170234783686, + "loss": 0.7781, + "step": 9435 + }, + { + "epoch": 1.6800213675213675, + "grad_norm": 0.7055163979530334, + "learning_rate": 0.00012503815097896555, + "loss": 1.0617, + "step": 9436 + }, + { + "epoch": 1.6801994301994303, + "grad_norm": 0.5567409992218018, + "learning_rate": 0.00012502459911952371, + "loss": 0.7911, + "step": 9437 + }, + { + "epoch": 1.6803774928774928, + "grad_norm": 0.7410423159599304, + "learning_rate": 0.0001250110467697768, + "loss": 1.1041, + "step": 9438 + }, + { + "epoch": 1.6805555555555556, + "grad_norm": 0.6185283064842224, + "learning_rate": 0.00012499749392999045, + "loss": 0.8101, + "step": 9439 + }, + { + "epoch": 1.680733618233618, + "grad_norm": 0.6988311409950256, + "learning_rate": 0.0001249839406004301, + "loss": 0.8579, + "step": 9440 + }, + { + "epoch": 1.6809116809116809, + "grad_norm": 0.5588746070861816, + "learning_rate": 0.00012497038678136132, + "loss": 0.8035, + "step": 9441 + }, + { + "epoch": 1.6810897435897436, + "grad_norm": 0.6568905711174011, + "learning_rate": 0.0001249568324730497, + "loss": 0.7455, + "step": 9442 + }, + { + "epoch": 1.6812678062678064, + "grad_norm": 0.6924821138381958, + "learning_rate": 0.00012494327767576078, + "loss": 1.134, + "step": 9443 + }, + { + "epoch": 1.681445868945869, + "grad_norm": 0.6940170526504517, + "learning_rate": 0.00012492972238976018, + "loss": 0.9719, + "step": 9444 + }, + { + "epoch": 1.6816239316239316, + "grad_norm": 0.667465090751648, + "learning_rate": 0.00012491616661531343, + "loss": 0.953, + "step": 9445 + }, + { + "epoch": 1.6818019943019942, + "grad_norm": 0.7693275809288025, + "learning_rate": 0.00012490261035268612, + "loss": 1.1342, + "step": 9446 + }, + { + "epoch": 1.681980056980057, + "grad_norm": 0.7243115305900574, + "learning_rate": 0.00012488905360214393, + "loss": 1.1847, + "step": 9447 + }, + { + "epoch": 1.6821581196581197, + "grad_norm": 0.657357931137085, + "learning_rate": 0.00012487549636395245, + "loss": 0.8747, + "step": 9448 + }, + { + "epoch": 1.6823361823361824, + "grad_norm": 0.7471592426300049, + "learning_rate": 0.00012486193863837727, + "loss": 1.0472, + "step": 9449 + }, + { + "epoch": 1.6825142450142452, + "grad_norm": 0.7476530075073242, + "learning_rate": 0.00012484838042568406, + "loss": 1.0708, + "step": 9450 + }, + { + "epoch": 1.6826923076923077, + "grad_norm": 0.6031121611595154, + "learning_rate": 0.00012483482172613846, + "loss": 0.8243, + "step": 9451 + }, + { + "epoch": 1.6828703703703702, + "grad_norm": 0.6733492016792297, + "learning_rate": 0.00012482126254000607, + "loss": 0.7808, + "step": 9452 + }, + { + "epoch": 1.683048433048433, + "grad_norm": 0.5865318179130554, + "learning_rate": 0.00012480770286755265, + "loss": 0.829, + "step": 9453 + }, + { + "epoch": 1.6832264957264957, + "grad_norm": 0.6805713772773743, + "learning_rate": 0.0001247941427090438, + "loss": 0.7206, + "step": 9454 + }, + { + "epoch": 1.6834045584045585, + "grad_norm": 0.6514836549758911, + "learning_rate": 0.0001247805820647453, + "loss": 0.9499, + "step": 9455 + }, + { + "epoch": 1.6835826210826212, + "grad_norm": 0.7432990074157715, + "learning_rate": 0.0001247670209349227, + "loss": 1.1324, + "step": 9456 + }, + { + "epoch": 1.6837606837606838, + "grad_norm": 0.6348414421081543, + "learning_rate": 0.00012475345931984178, + "loss": 0.8246, + "step": 9457 + }, + { + "epoch": 1.6839387464387463, + "grad_norm": 0.7194374203681946, + "learning_rate": 0.00012473989721976825, + "loss": 0.9634, + "step": 9458 + }, + { + "epoch": 1.684116809116809, + "grad_norm": 0.7869647741317749, + "learning_rate": 0.00012472633463496785, + "loss": 1.2115, + "step": 9459 + }, + { + "epoch": 1.6842948717948718, + "grad_norm": 0.6672070026397705, + "learning_rate": 0.00012471277156570623, + "loss": 0.9842, + "step": 9460 + }, + { + "epoch": 1.6844729344729346, + "grad_norm": 0.6611466407775879, + "learning_rate": 0.00012469920801224925, + "loss": 0.9343, + "step": 9461 + }, + { + "epoch": 1.6846509971509973, + "grad_norm": 0.6715068221092224, + "learning_rate": 0.0001246856439748626, + "loss": 0.6852, + "step": 9462 + }, + { + "epoch": 1.6848290598290598, + "grad_norm": 0.641942024230957, + "learning_rate": 0.00012467207945381198, + "loss": 0.8863, + "step": 9463 + }, + { + "epoch": 1.6850071225071224, + "grad_norm": 0.8414762616157532, + "learning_rate": 0.00012465851444936325, + "loss": 1.3404, + "step": 9464 + }, + { + "epoch": 1.6851851851851851, + "grad_norm": 0.715752363204956, + "learning_rate": 0.00012464494896178216, + "loss": 1.123, + "step": 9465 + }, + { + "epoch": 1.6853632478632479, + "grad_norm": 0.6913973093032837, + "learning_rate": 0.00012463138299133447, + "loss": 1.0659, + "step": 9466 + }, + { + "epoch": 1.6855413105413106, + "grad_norm": 0.6998484134674072, + "learning_rate": 0.000124617816538286, + "loss": 1.0555, + "step": 9467 + }, + { + "epoch": 1.6857193732193734, + "grad_norm": 0.7313308119773865, + "learning_rate": 0.00012460424960290256, + "loss": 1.0915, + "step": 9468 + }, + { + "epoch": 1.685897435897436, + "grad_norm": 0.6790569424629211, + "learning_rate": 0.00012459068218544995, + "loss": 1.0214, + "step": 9469 + }, + { + "epoch": 1.6860754985754984, + "grad_norm": 0.6494466662406921, + "learning_rate": 0.00012457711428619402, + "loss": 0.9476, + "step": 9470 + }, + { + "epoch": 1.6862535612535612, + "grad_norm": 0.8048526048660278, + "learning_rate": 0.0001245635459054006, + "loss": 1.1852, + "step": 9471 + }, + { + "epoch": 1.686431623931624, + "grad_norm": 0.6237879395484924, + "learning_rate": 0.0001245499770433355, + "loss": 1.0106, + "step": 9472 + }, + { + "epoch": 1.6866096866096867, + "grad_norm": 0.6282906532287598, + "learning_rate": 0.0001245364077002646, + "loss": 0.9858, + "step": 9473 + }, + { + "epoch": 1.6867877492877494, + "grad_norm": 0.7239370346069336, + "learning_rate": 0.00012452283787645375, + "loss": 0.9586, + "step": 9474 + }, + { + "epoch": 1.686965811965812, + "grad_norm": 0.6438776850700378, + "learning_rate": 0.00012450926757216887, + "loss": 0.9198, + "step": 9475 + }, + { + "epoch": 1.6871438746438745, + "grad_norm": 0.6451360583305359, + "learning_rate": 0.00012449569678767578, + "loss": 1.0183, + "step": 9476 + }, + { + "epoch": 1.6873219373219372, + "grad_norm": 0.6950216293334961, + "learning_rate": 0.0001244821255232404, + "loss": 0.9048, + "step": 9477 + }, + { + "epoch": 1.6875, + "grad_norm": 0.710489809513092, + "learning_rate": 0.00012446855377912865, + "loss": 1.1596, + "step": 9478 + }, + { + "epoch": 1.6876780626780628, + "grad_norm": 0.6819305419921875, + "learning_rate": 0.0001244549815556064, + "loss": 0.8486, + "step": 9479 + }, + { + "epoch": 1.6878561253561255, + "grad_norm": 0.7185879945755005, + "learning_rate": 0.00012444140885293958, + "loss": 0.9539, + "step": 9480 + }, + { + "epoch": 1.688034188034188, + "grad_norm": 0.8181464672088623, + "learning_rate": 0.00012442783567139415, + "loss": 1.0038, + "step": 9481 + }, + { + "epoch": 1.6882122507122506, + "grad_norm": 0.47161349654197693, + "learning_rate": 0.000124414262011236, + "loss": 0.67, + "step": 9482 + }, + { + "epoch": 1.6883903133903133, + "grad_norm": 0.7752482295036316, + "learning_rate": 0.00012440068787273112, + "loss": 0.9944, + "step": 9483 + }, + { + "epoch": 1.688568376068376, + "grad_norm": 0.7119397521018982, + "learning_rate": 0.00012438711325614543, + "loss": 0.9098, + "step": 9484 + }, + { + "epoch": 1.6887464387464388, + "grad_norm": 0.7161153554916382, + "learning_rate": 0.00012437353816174493, + "loss": 1.0003, + "step": 9485 + }, + { + "epoch": 1.6889245014245016, + "grad_norm": 0.5989507436752319, + "learning_rate": 0.0001243599625897956, + "loss": 1.0301, + "step": 9486 + }, + { + "epoch": 1.689102564102564, + "grad_norm": 0.7906841039657593, + "learning_rate": 0.00012434638654056334, + "loss": 1.0388, + "step": 9487 + }, + { + "epoch": 1.6892806267806266, + "grad_norm": 0.6679551601409912, + "learning_rate": 0.00012433281001431428, + "loss": 0.9505, + "step": 9488 + }, + { + "epoch": 1.6894586894586894, + "grad_norm": 0.7090578675270081, + "learning_rate": 0.0001243192330113143, + "loss": 0.8616, + "step": 9489 + }, + { + "epoch": 1.6896367521367521, + "grad_norm": 0.6401308178901672, + "learning_rate": 0.00012430565553182949, + "loss": 0.9099, + "step": 9490 + }, + { + "epoch": 1.6898148148148149, + "grad_norm": 0.7360149621963501, + "learning_rate": 0.00012429207757612586, + "loss": 1.0233, + "step": 9491 + }, + { + "epoch": 1.6899928774928776, + "grad_norm": 0.6736137270927429, + "learning_rate": 0.00012427849914446946, + "loss": 0.9803, + "step": 9492 + }, + { + "epoch": 1.6901709401709402, + "grad_norm": 0.7728668451309204, + "learning_rate": 0.00012426492023712623, + "loss": 1.2316, + "step": 9493 + }, + { + "epoch": 1.6903490028490027, + "grad_norm": 0.789718508720398, + "learning_rate": 0.00012425134085436234, + "loss": 1.1218, + "step": 9494 + }, + { + "epoch": 1.6905270655270654, + "grad_norm": 0.7314121723175049, + "learning_rate": 0.0001242377609964438, + "loss": 1.1294, + "step": 9495 + }, + { + "epoch": 1.6907051282051282, + "grad_norm": 0.7222046256065369, + "learning_rate": 0.0001242241806636367, + "loss": 1.0288, + "step": 9496 + }, + { + "epoch": 1.690883190883191, + "grad_norm": 0.7546363472938538, + "learning_rate": 0.00012421059985620708, + "loss": 0.8781, + "step": 9497 + }, + { + "epoch": 1.6910612535612537, + "grad_norm": 0.7502550482749939, + "learning_rate": 0.00012419701857442104, + "loss": 0.927, + "step": 9498 + }, + { + "epoch": 1.6912393162393162, + "grad_norm": 0.6244059205055237, + "learning_rate": 0.00012418343681854473, + "loss": 0.9689, + "step": 9499 + }, + { + "epoch": 1.6914173789173788, + "grad_norm": 0.7214263677597046, + "learning_rate": 0.00012416985458884417, + "loss": 1.0842, + "step": 9500 + }, + { + "epoch": 1.6915954415954415, + "grad_norm": 0.6960242390632629, + "learning_rate": 0.00012415627188558555, + "loss": 0.9766, + "step": 9501 + }, + { + "epoch": 1.6917735042735043, + "grad_norm": 0.6687830686569214, + "learning_rate": 0.00012414268870903494, + "loss": 1.0222, + "step": 9502 + }, + { + "epoch": 1.691951566951567, + "grad_norm": 0.8611155152320862, + "learning_rate": 0.00012412910505945848, + "loss": 1.1792, + "step": 9503 + }, + { + "epoch": 1.6921296296296298, + "grad_norm": 0.6655587553977966, + "learning_rate": 0.00012411552093712235, + "loss": 0.8763, + "step": 9504 + }, + { + "epoch": 1.6923076923076923, + "grad_norm": 0.7829837799072266, + "learning_rate": 0.00012410193634229268, + "loss": 1.0803, + "step": 9505 + }, + { + "epoch": 1.6924857549857548, + "grad_norm": 0.7951042652130127, + "learning_rate": 0.00012408835127523566, + "loss": 1.0925, + "step": 9506 + }, + { + "epoch": 1.6926638176638176, + "grad_norm": 0.715495228767395, + "learning_rate": 0.0001240747657362174, + "loss": 1.2411, + "step": 9507 + }, + { + "epoch": 1.6928418803418803, + "grad_norm": 0.6779513359069824, + "learning_rate": 0.00012406117972550414, + "loss": 0.8886, + "step": 9508 + }, + { + "epoch": 1.693019943019943, + "grad_norm": 0.647588312625885, + "learning_rate": 0.00012404759324336203, + "loss": 1.107, + "step": 9509 + }, + { + "epoch": 1.6931980056980058, + "grad_norm": 0.7398989796638489, + "learning_rate": 0.00012403400629005726, + "loss": 1.0256, + "step": 9510 + }, + { + "epoch": 1.6933760683760684, + "grad_norm": 0.7572638392448425, + "learning_rate": 0.0001240204188658561, + "loss": 0.9662, + "step": 9511 + }, + { + "epoch": 1.693554131054131, + "grad_norm": 0.7044163346290588, + "learning_rate": 0.00012400683097102473, + "loss": 1.1388, + "step": 9512 + }, + { + "epoch": 1.6937321937321936, + "grad_norm": 0.7889094948768616, + "learning_rate": 0.00012399324260582936, + "loss": 1.0453, + "step": 9513 + }, + { + "epoch": 1.6939102564102564, + "grad_norm": 0.7977854609489441, + "learning_rate": 0.00012397965377053627, + "loss": 1.015, + "step": 9514 + }, + { + "epoch": 1.6940883190883191, + "grad_norm": 0.6223814487457275, + "learning_rate": 0.00012396606446541165, + "loss": 0.7985, + "step": 9515 + }, + { + "epoch": 1.694266381766382, + "grad_norm": 0.8307462334632874, + "learning_rate": 0.0001239524746907218, + "loss": 0.8899, + "step": 9516 + }, + { + "epoch": 1.6944444444444444, + "grad_norm": 0.7780544757843018, + "learning_rate": 0.00012393888444673295, + "loss": 0.9406, + "step": 9517 + }, + { + "epoch": 1.6946225071225072, + "grad_norm": 0.6894499659538269, + "learning_rate": 0.0001239252937337114, + "loss": 0.9412, + "step": 9518 + }, + { + "epoch": 1.6948005698005697, + "grad_norm": 0.7000680565834045, + "learning_rate": 0.00012391170255192342, + "loss": 1.0314, + "step": 9519 + }, + { + "epoch": 1.6949786324786325, + "grad_norm": 0.6772416830062866, + "learning_rate": 0.0001238981109016353, + "loss": 0.9153, + "step": 9520 + }, + { + "epoch": 1.6951566951566952, + "grad_norm": 0.7069609761238098, + "learning_rate": 0.00012388451878311333, + "loss": 1.1777, + "step": 9521 + }, + { + "epoch": 1.695334757834758, + "grad_norm": 0.6138432621955872, + "learning_rate": 0.00012387092619662386, + "loss": 0.8085, + "step": 9522 + }, + { + "epoch": 1.6955128205128205, + "grad_norm": 0.6122859716415405, + "learning_rate": 0.00012385733314243313, + "loss": 0.8534, + "step": 9523 + }, + { + "epoch": 1.6956908831908832, + "grad_norm": 0.7499903440475464, + "learning_rate": 0.00012384373962080755, + "loss": 0.9329, + "step": 9524 + }, + { + "epoch": 1.6958689458689458, + "grad_norm": 0.6413441896438599, + "learning_rate": 0.00012383014563201343, + "loss": 0.9609, + "step": 9525 + }, + { + "epoch": 1.6960470085470085, + "grad_norm": 0.7467969059944153, + "learning_rate": 0.0001238165511763171, + "loss": 0.9142, + "step": 9526 + }, + { + "epoch": 1.6962250712250713, + "grad_norm": 0.6540884375572205, + "learning_rate": 0.00012380295625398494, + "loss": 0.9503, + "step": 9527 + }, + { + "epoch": 1.696403133903134, + "grad_norm": 0.6298567652702332, + "learning_rate": 0.00012378936086528326, + "loss": 0.8853, + "step": 9528 + }, + { + "epoch": 1.6965811965811965, + "grad_norm": 0.8003417253494263, + "learning_rate": 0.00012377576501047845, + "loss": 0.969, + "step": 9529 + }, + { + "epoch": 1.6967592592592593, + "grad_norm": 0.8318493962287903, + "learning_rate": 0.00012376216868983697, + "loss": 1.1413, + "step": 9530 + }, + { + "epoch": 1.6969373219373218, + "grad_norm": 0.8294426202774048, + "learning_rate": 0.00012374857190362515, + "loss": 1.1885, + "step": 9531 + }, + { + "epoch": 1.6971153846153846, + "grad_norm": 0.7502955198287964, + "learning_rate": 0.0001237349746521094, + "loss": 1.233, + "step": 9532 + }, + { + "epoch": 1.6972934472934473, + "grad_norm": 0.6306588649749756, + "learning_rate": 0.00012372137693555612, + "loss": 1.2255, + "step": 9533 + }, + { + "epoch": 1.69747150997151, + "grad_norm": 0.7802746891975403, + "learning_rate": 0.0001237077787542317, + "loss": 1.2054, + "step": 9534 + }, + { + "epoch": 1.6976495726495726, + "grad_norm": 0.685114860534668, + "learning_rate": 0.00012369418010840265, + "loss": 0.9865, + "step": 9535 + }, + { + "epoch": 1.6978276353276354, + "grad_norm": 0.6656857132911682, + "learning_rate": 0.00012368058099833536, + "loss": 1.1579, + "step": 9536 + }, + { + "epoch": 1.698005698005698, + "grad_norm": 0.6596674919128418, + "learning_rate": 0.00012366698142429625, + "loss": 0.9104, + "step": 9537 + }, + { + "epoch": 1.6981837606837606, + "grad_norm": 0.6025584936141968, + "learning_rate": 0.00012365338138655183, + "loss": 1.117, + "step": 9538 + }, + { + "epoch": 1.6983618233618234, + "grad_norm": 0.671585202217102, + "learning_rate": 0.0001236397808853685, + "loss": 1.0271, + "step": 9539 + }, + { + "epoch": 1.6985398860398861, + "grad_norm": 0.7467984557151794, + "learning_rate": 0.0001236261799210128, + "loss": 1.0411, + "step": 9540 + }, + { + "epoch": 1.6987179487179487, + "grad_norm": 0.6251640915870667, + "learning_rate": 0.0001236125784937512, + "loss": 0.7154, + "step": 9541 + }, + { + "epoch": 1.6988960113960114, + "grad_norm": 0.7560956478118896, + "learning_rate": 0.00012359897660385016, + "loss": 1.0048, + "step": 9542 + }, + { + "epoch": 1.699074074074074, + "grad_norm": 0.6144903302192688, + "learning_rate": 0.00012358537425157618, + "loss": 1.1294, + "step": 9543 + }, + { + "epoch": 1.6992521367521367, + "grad_norm": 0.7839425206184387, + "learning_rate": 0.00012357177143719578, + "loss": 1.0725, + "step": 9544 + }, + { + "epoch": 1.6994301994301995, + "grad_norm": 0.6488651037216187, + "learning_rate": 0.00012355816816097553, + "loss": 0.9267, + "step": 9545 + }, + { + "epoch": 1.6996082621082622, + "grad_norm": 0.6848782896995544, + "learning_rate": 0.00012354456442318187, + "loss": 1.0426, + "step": 9546 + }, + { + "epoch": 1.6997863247863247, + "grad_norm": 0.7164611220359802, + "learning_rate": 0.0001235309602240814, + "loss": 0.8208, + "step": 9547 + }, + { + "epoch": 1.6999643874643875, + "grad_norm": 0.6725530624389648, + "learning_rate": 0.0001235173555639406, + "loss": 0.9366, + "step": 9548 + }, + { + "epoch": 1.70014245014245, + "grad_norm": 0.6958004236221313, + "learning_rate": 0.00012350375044302612, + "loss": 1.0185, + "step": 9549 + }, + { + "epoch": 1.7003205128205128, + "grad_norm": 0.8035947680473328, + "learning_rate": 0.00012349014486160445, + "loss": 1.065, + "step": 9550 + }, + { + "epoch": 1.7004985754985755, + "grad_norm": 0.6705633997917175, + "learning_rate": 0.00012347653881994222, + "loss": 0.8381, + "step": 9551 + }, + { + "epoch": 1.7006766381766383, + "grad_norm": 0.6652300357818604, + "learning_rate": 0.00012346293231830596, + "loss": 1.1428, + "step": 9552 + }, + { + "epoch": 1.7008547008547008, + "grad_norm": 0.6719335913658142, + "learning_rate": 0.0001234493253569623, + "loss": 1.0138, + "step": 9553 + }, + { + "epoch": 1.7010327635327636, + "grad_norm": 0.746981680393219, + "learning_rate": 0.0001234357179361778, + "loss": 1.1169, + "step": 9554 + }, + { + "epoch": 1.701210826210826, + "grad_norm": 0.6768170595169067, + "learning_rate": 0.0001234221100562191, + "loss": 0.9065, + "step": 9555 + }, + { + "epoch": 1.7013888888888888, + "grad_norm": 0.7127171754837036, + "learning_rate": 0.00012340850171735278, + "loss": 0.9467, + "step": 9556 + }, + { + "epoch": 1.7015669515669516, + "grad_norm": 0.6802694797515869, + "learning_rate": 0.00012339489291984554, + "loss": 0.8938, + "step": 9557 + }, + { + "epoch": 1.7017450142450143, + "grad_norm": 0.7101455926895142, + "learning_rate": 0.00012338128366396394, + "loss": 1.1939, + "step": 9558 + }, + { + "epoch": 1.7019230769230769, + "grad_norm": 0.621223509311676, + "learning_rate": 0.00012336767394997467, + "loss": 0.7583, + "step": 9559 + }, + { + "epoch": 1.7021011396011396, + "grad_norm": 0.7130763530731201, + "learning_rate": 0.00012335406377814439, + "loss": 0.8684, + "step": 9560 + }, + { + "epoch": 1.7022792022792022, + "grad_norm": 0.6761086583137512, + "learning_rate": 0.00012334045314873972, + "loss": 1.0197, + "step": 9561 + }, + { + "epoch": 1.702457264957265, + "grad_norm": 0.7030459642410278, + "learning_rate": 0.00012332684206202736, + "loss": 0.8627, + "step": 9562 + }, + { + "epoch": 1.7026353276353277, + "grad_norm": 0.6278037428855896, + "learning_rate": 0.000123313230518274, + "loss": 0.8953, + "step": 9563 + }, + { + "epoch": 1.7028133903133904, + "grad_norm": 0.6450623869895935, + "learning_rate": 0.00012329961851774627, + "loss": 0.8826, + "step": 9564 + }, + { + "epoch": 1.702991452991453, + "grad_norm": 0.7324244976043701, + "learning_rate": 0.00012328600606071097, + "loss": 1.0133, + "step": 9565 + }, + { + "epoch": 1.7031695156695157, + "grad_norm": 0.6560033559799194, + "learning_rate": 0.00012327239314743473, + "loss": 0.9601, + "step": 9566 + }, + { + "epoch": 1.7033475783475782, + "grad_norm": 0.6693514585494995, + "learning_rate": 0.0001232587797781843, + "loss": 0.9447, + "step": 9567 + }, + { + "epoch": 1.703525641025641, + "grad_norm": 0.6403199434280396, + "learning_rate": 0.00012324516595322638, + "loss": 0.8554, + "step": 9568 + }, + { + "epoch": 1.7037037037037037, + "grad_norm": 0.8290280103683472, + "learning_rate": 0.00012323155167282774, + "loss": 1.1877, + "step": 9569 + }, + { + "epoch": 1.7038817663817665, + "grad_norm": 0.7207778692245483, + "learning_rate": 0.00012321793693725509, + "loss": 1.0978, + "step": 9570 + }, + { + "epoch": 1.7040598290598292, + "grad_norm": 0.8794265985488892, + "learning_rate": 0.00012320432174677519, + "loss": 0.9387, + "step": 9571 + }, + { + "epoch": 1.7042378917378918, + "grad_norm": 0.6683359146118164, + "learning_rate": 0.00012319070610165484, + "loss": 0.9227, + "step": 9572 + }, + { + "epoch": 1.7044159544159543, + "grad_norm": 0.7342001795768738, + "learning_rate": 0.00012317709000216076, + "loss": 0.9453, + "step": 9573 + }, + { + "epoch": 1.704594017094017, + "grad_norm": 0.6315770149230957, + "learning_rate": 0.00012316347344855973, + "loss": 0.8263, + "step": 9574 + }, + { + "epoch": 1.7047720797720798, + "grad_norm": 0.7697155475616455, + "learning_rate": 0.00012314985644111857, + "loss": 1.0238, + "step": 9575 + }, + { + "epoch": 1.7049501424501425, + "grad_norm": 0.6674068570137024, + "learning_rate": 0.00012313623898010408, + "loss": 1.0823, + "step": 9576 + }, + { + "epoch": 1.7051282051282053, + "grad_norm": 0.6995484232902527, + "learning_rate": 0.00012312262106578304, + "loss": 1.2001, + "step": 9577 + }, + { + "epoch": 1.7053062678062678, + "grad_norm": 0.7639257907867432, + "learning_rate": 0.00012310900269842226, + "loss": 1.3438, + "step": 9578 + }, + { + "epoch": 1.7054843304843303, + "grad_norm": 0.6486390233039856, + "learning_rate": 0.00012309538387828857, + "loss": 0.9924, + "step": 9579 + }, + { + "epoch": 1.705662393162393, + "grad_norm": 0.6737813949584961, + "learning_rate": 0.00012308176460564885, + "loss": 0.8722, + "step": 9580 + }, + { + "epoch": 1.7058404558404558, + "grad_norm": 0.6462090611457825, + "learning_rate": 0.00012306814488076987, + "loss": 1.1013, + "step": 9581 + }, + { + "epoch": 1.7060185185185186, + "grad_norm": 0.7887832522392273, + "learning_rate": 0.00012305452470391852, + "loss": 0.9998, + "step": 9582 + }, + { + "epoch": 1.7061965811965814, + "grad_norm": 0.6345070004463196, + "learning_rate": 0.00012304090407536165, + "loss": 1.0305, + "step": 9583 + }, + { + "epoch": 1.7063746438746439, + "grad_norm": 0.6398460268974304, + "learning_rate": 0.0001230272829953661, + "loss": 1.2243, + "step": 9584 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 0.6501944065093994, + "learning_rate": 0.00012301366146419879, + "loss": 0.9425, + "step": 9585 + }, + { + "epoch": 1.7067307692307692, + "grad_norm": 0.6406761407852173, + "learning_rate": 0.00012300003948212661, + "loss": 0.948, + "step": 9586 + }, + { + "epoch": 1.706908831908832, + "grad_norm": 0.7114266157150269, + "learning_rate": 0.00012298641704941644, + "loss": 1.1291, + "step": 9587 + }, + { + "epoch": 1.7070868945868947, + "grad_norm": 0.6653099656105042, + "learning_rate": 0.00012297279416633515, + "loss": 1.0156, + "step": 9588 + }, + { + "epoch": 1.7072649572649574, + "grad_norm": 0.5970917344093323, + "learning_rate": 0.0001229591708331497, + "loss": 0.9424, + "step": 9589 + }, + { + "epoch": 1.70744301994302, + "grad_norm": 0.6861461400985718, + "learning_rate": 0.00012294554705012694, + "loss": 0.7581, + "step": 9590 + }, + { + "epoch": 1.7076210826210825, + "grad_norm": 0.6930568218231201, + "learning_rate": 0.00012293192281753393, + "loss": 1.0544, + "step": 9591 + }, + { + "epoch": 1.7077991452991452, + "grad_norm": 0.7420656085014343, + "learning_rate": 0.00012291829813563748, + "loss": 0.7092, + "step": 9592 + }, + { + "epoch": 1.707977207977208, + "grad_norm": 0.6607801914215088, + "learning_rate": 0.0001229046730047046, + "loss": 0.5544, + "step": 9593 + }, + { + "epoch": 1.7081552706552707, + "grad_norm": 0.8419139385223389, + "learning_rate": 0.00012289104742500224, + "loss": 1.0443, + "step": 9594 + }, + { + "epoch": 1.7083333333333335, + "grad_norm": 0.6774617433547974, + "learning_rate": 0.00012287742139679734, + "loss": 1.0098, + "step": 9595 + }, + { + "epoch": 1.708511396011396, + "grad_norm": 0.7517698407173157, + "learning_rate": 0.0001228637949203569, + "loss": 1.1145, + "step": 9596 + }, + { + "epoch": 1.7086894586894585, + "grad_norm": 0.6048635840415955, + "learning_rate": 0.00012285016799594791, + "loss": 0.7398, + "step": 9597 + }, + { + "epoch": 1.7088675213675213, + "grad_norm": 0.8054425716400146, + "learning_rate": 0.00012283654062383734, + "loss": 1.0893, + "step": 9598 + }, + { + "epoch": 1.709045584045584, + "grad_norm": 0.8694897294044495, + "learning_rate": 0.0001228229128042922, + "loss": 1.2366, + "step": 9599 + }, + { + "epoch": 1.7092236467236468, + "grad_norm": 0.7460638880729675, + "learning_rate": 0.00012280928453757946, + "loss": 1.1753, + "step": 9600 + }, + { + "epoch": 1.7094017094017095, + "grad_norm": 0.6714958548545837, + "learning_rate": 0.00012279565582396618, + "loss": 1.0473, + "step": 9601 + }, + { + "epoch": 1.709579772079772, + "grad_norm": 0.6893340945243835, + "learning_rate": 0.00012278202666371937, + "loss": 1.2761, + "step": 9602 + }, + { + "epoch": 1.7097578347578346, + "grad_norm": 0.6816153526306152, + "learning_rate": 0.00012276839705710612, + "loss": 0.991, + "step": 9603 + }, + { + "epoch": 1.7099358974358974, + "grad_norm": 0.6961633563041687, + "learning_rate": 0.0001227547670043934, + "loss": 1.0634, + "step": 9604 + }, + { + "epoch": 1.71011396011396, + "grad_norm": 0.643734872341156, + "learning_rate": 0.0001227411365058483, + "loss": 0.8672, + "step": 9605 + }, + { + "epoch": 1.7102920227920229, + "grad_norm": 0.7313315272331238, + "learning_rate": 0.00012272750556173784, + "loss": 1.1152, + "step": 9606 + }, + { + "epoch": 1.7104700854700856, + "grad_norm": 0.6464954614639282, + "learning_rate": 0.00012271387417232916, + "loss": 0.8798, + "step": 9607 + }, + { + "epoch": 1.7106481481481481, + "grad_norm": 0.8365204334259033, + "learning_rate": 0.00012270024233788929, + "loss": 1.213, + "step": 9608 + }, + { + "epoch": 1.7108262108262107, + "grad_norm": 0.6460705995559692, + "learning_rate": 0.0001226866100586853, + "loss": 0.9232, + "step": 9609 + }, + { + "epoch": 1.7110042735042734, + "grad_norm": 0.6446022987365723, + "learning_rate": 0.00012267297733498434, + "loss": 0.8295, + "step": 9610 + }, + { + "epoch": 1.7111823361823362, + "grad_norm": 0.7692012190818787, + "learning_rate": 0.00012265934416705345, + "loss": 1.0715, + "step": 9611 + }, + { + "epoch": 1.711360398860399, + "grad_norm": 0.671154260635376, + "learning_rate": 0.0001226457105551598, + "loss": 0.9752, + "step": 9612 + }, + { + "epoch": 1.7115384615384617, + "grad_norm": 0.6525935530662537, + "learning_rate": 0.00012263207649957053, + "loss": 1.09, + "step": 9613 + }, + { + "epoch": 1.7117165242165242, + "grad_norm": 0.6984749436378479, + "learning_rate": 0.0001226184420005527, + "loss": 0.9956, + "step": 9614 + }, + { + "epoch": 1.7118945868945867, + "grad_norm": 0.6769809126853943, + "learning_rate": 0.0001226048070583735, + "loss": 1.0151, + "step": 9615 + }, + { + "epoch": 1.7120726495726495, + "grad_norm": 0.6085978746414185, + "learning_rate": 0.00012259117167330005, + "loss": 0.8706, + "step": 9616 + }, + { + "epoch": 1.7122507122507122, + "grad_norm": 0.7335749268531799, + "learning_rate": 0.00012257753584559952, + "loss": 1.0575, + "step": 9617 + }, + { + "epoch": 1.712428774928775, + "grad_norm": 0.7392038106918335, + "learning_rate": 0.0001225638995755391, + "loss": 0.8763, + "step": 9618 + }, + { + "epoch": 1.7126068376068377, + "grad_norm": 0.6708608865737915, + "learning_rate": 0.00012255026286338592, + "loss": 1.131, + "step": 9619 + }, + { + "epoch": 1.7127849002849003, + "grad_norm": 0.726657509803772, + "learning_rate": 0.0001225366257094072, + "loss": 1.0569, + "step": 9620 + }, + { + "epoch": 1.7129629629629628, + "grad_norm": 0.749098002910614, + "learning_rate": 0.0001225229881138701, + "loss": 0.9196, + "step": 9621 + }, + { + "epoch": 1.7131410256410255, + "grad_norm": 0.6550580263137817, + "learning_rate": 0.00012250935007704182, + "loss": 1.0244, + "step": 9622 + }, + { + "epoch": 1.7133190883190883, + "grad_norm": 0.7714282274246216, + "learning_rate": 0.00012249571159918962, + "loss": 1.1025, + "step": 9623 + }, + { + "epoch": 1.713497150997151, + "grad_norm": 0.7869850397109985, + "learning_rate": 0.00012248207268058064, + "loss": 0.9238, + "step": 9624 + }, + { + "epoch": 1.7136752136752138, + "grad_norm": 0.7187856435775757, + "learning_rate": 0.00012246843332148216, + "loss": 1.081, + "step": 9625 + }, + { + "epoch": 1.7138532763532763, + "grad_norm": 0.6634210348129272, + "learning_rate": 0.00012245479352216142, + "loss": 1.1944, + "step": 9626 + }, + { + "epoch": 1.714031339031339, + "grad_norm": 0.6609212160110474, + "learning_rate": 0.00012244115328288567, + "loss": 0.9613, + "step": 9627 + }, + { + "epoch": 1.7142094017094016, + "grad_norm": 0.7906867861747742, + "learning_rate": 0.0001224275126039221, + "loss": 1.2692, + "step": 9628 + }, + { + "epoch": 1.7143874643874644, + "grad_norm": 0.8037096858024597, + "learning_rate": 0.000122413871485538, + "loss": 0.9823, + "step": 9629 + }, + { + "epoch": 1.7145655270655271, + "grad_norm": 0.7740145921707153, + "learning_rate": 0.00012240022992800068, + "loss": 1.1937, + "step": 9630 + }, + { + "epoch": 1.7147435897435899, + "grad_norm": 0.595372200012207, + "learning_rate": 0.00012238658793157738, + "loss": 0.9153, + "step": 9631 + }, + { + "epoch": 1.7149216524216524, + "grad_norm": 0.6671900749206543, + "learning_rate": 0.0001223729454965354, + "loss": 1.0895, + "step": 9632 + }, + { + "epoch": 1.7150997150997151, + "grad_norm": 0.5805774927139282, + "learning_rate": 0.000122359302623142, + "loss": 1.0001, + "step": 9633 + }, + { + "epoch": 1.7152777777777777, + "grad_norm": 0.8851602673530579, + "learning_rate": 0.00012234565931166456, + "loss": 1.2828, + "step": 9634 + }, + { + "epoch": 1.7154558404558404, + "grad_norm": 0.6960011720657349, + "learning_rate": 0.0001223320155623703, + "loss": 1.0622, + "step": 9635 + }, + { + "epoch": 1.7156339031339032, + "grad_norm": 0.5587009191513062, + "learning_rate": 0.0001223183713755266, + "loss": 0.83, + "step": 9636 + }, + { + "epoch": 1.715811965811966, + "grad_norm": 0.6892730593681335, + "learning_rate": 0.00012230472675140076, + "loss": 0.9214, + "step": 9637 + }, + { + "epoch": 1.7159900284900285, + "grad_norm": 0.6545090079307556, + "learning_rate": 0.00012229108169026017, + "loss": 0.829, + "step": 9638 + }, + { + "epoch": 1.7161680911680912, + "grad_norm": 0.6539101600646973, + "learning_rate": 0.00012227743619237213, + "loss": 1.0686, + "step": 9639 + }, + { + "epoch": 1.7163461538461537, + "grad_norm": 0.5887274146080017, + "learning_rate": 0.000122263790258004, + "loss": 0.9285, + "step": 9640 + }, + { + "epoch": 1.7165242165242165, + "grad_norm": 0.6328918933868408, + "learning_rate": 0.00012225014388742313, + "loss": 0.9684, + "step": 9641 + }, + { + "epoch": 1.7167022792022792, + "grad_norm": 0.6377436518669128, + "learning_rate": 0.00012223649708089694, + "loss": 0.9425, + "step": 9642 + }, + { + "epoch": 1.716880341880342, + "grad_norm": 0.6967392563819885, + "learning_rate": 0.00012222284983869275, + "loss": 0.9342, + "step": 9643 + }, + { + "epoch": 1.7170584045584045, + "grad_norm": 0.7051317691802979, + "learning_rate": 0.00012220920216107802, + "loss": 1.1843, + "step": 9644 + }, + { + "epoch": 1.7172364672364673, + "grad_norm": 0.6864503622055054, + "learning_rate": 0.00012219555404832007, + "loss": 1.0371, + "step": 9645 + }, + { + "epoch": 1.7174145299145298, + "grad_norm": 0.583454430103302, + "learning_rate": 0.00012218190550068638, + "loss": 0.6774, + "step": 9646 + }, + { + "epoch": 1.7175925925925926, + "grad_norm": 0.6755677461624146, + "learning_rate": 0.0001221682565184443, + "loss": 0.9517, + "step": 9647 + }, + { + "epoch": 1.7177706552706553, + "grad_norm": 0.7230031490325928, + "learning_rate": 0.0001221546071018613, + "loss": 1.0385, + "step": 9648 + }, + { + "epoch": 1.717948717948718, + "grad_norm": 0.7381200194358826, + "learning_rate": 0.0001221409572512048, + "loss": 0.9893, + "step": 9649 + }, + { + "epoch": 1.7181267806267806, + "grad_norm": 0.7079094648361206, + "learning_rate": 0.0001221273069667422, + "loss": 0.7793, + "step": 9650 + }, + { + "epoch": 1.7183048433048433, + "grad_norm": 0.6666881442070007, + "learning_rate": 0.00012211365624874106, + "loss": 0.9752, + "step": 9651 + }, + { + "epoch": 1.7184829059829059, + "grad_norm": 0.6196922659873962, + "learning_rate": 0.00012210000509746868, + "loss": 0.922, + "step": 9652 + }, + { + "epoch": 1.7186609686609686, + "grad_norm": 0.657879650592804, + "learning_rate": 0.00012208635351319266, + "loss": 1.2583, + "step": 9653 + }, + { + "epoch": 1.7188390313390314, + "grad_norm": 0.7240459322929382, + "learning_rate": 0.00012207270149618043, + "loss": 0.8479, + "step": 9654 + }, + { + "epoch": 1.7190170940170941, + "grad_norm": 0.8293825387954712, + "learning_rate": 0.00012205904904669945, + "loss": 0.9092, + "step": 9655 + }, + { + "epoch": 1.7191951566951567, + "grad_norm": 0.6907553672790527, + "learning_rate": 0.0001220453961650172, + "loss": 1.0543, + "step": 9656 + }, + { + "epoch": 1.7193732193732194, + "grad_norm": 0.7178300023078918, + "learning_rate": 0.00012203174285140124, + "loss": 0.9147, + "step": 9657 + }, + { + "epoch": 1.719551282051282, + "grad_norm": 0.7037166357040405, + "learning_rate": 0.00012201808910611905, + "loss": 0.8685, + "step": 9658 + }, + { + "epoch": 1.7197293447293447, + "grad_norm": 0.5850751996040344, + "learning_rate": 0.00012200443492943813, + "loss": 0.72, + "step": 9659 + }, + { + "epoch": 1.7199074074074074, + "grad_norm": 0.744239330291748, + "learning_rate": 0.00012199078032162603, + "loss": 0.9717, + "step": 9660 + }, + { + "epoch": 1.7200854700854702, + "grad_norm": 0.6509126424789429, + "learning_rate": 0.00012197712528295025, + "loss": 0.9768, + "step": 9661 + }, + { + "epoch": 1.7202635327635327, + "grad_norm": 0.623220682144165, + "learning_rate": 0.00012196346981367837, + "loss": 0.9824, + "step": 9662 + }, + { + "epoch": 1.7204415954415955, + "grad_norm": 0.6376451849937439, + "learning_rate": 0.00012194981391407792, + "loss": 0.8228, + "step": 9663 + }, + { + "epoch": 1.720619658119658, + "grad_norm": 0.794830322265625, + "learning_rate": 0.00012193615758441648, + "loss": 0.9168, + "step": 9664 + }, + { + "epoch": 1.7207977207977208, + "grad_norm": 0.7812975645065308, + "learning_rate": 0.0001219225008249616, + "loss": 0.8625, + "step": 9665 + }, + { + "epoch": 1.7209757834757835, + "grad_norm": 0.6843218207359314, + "learning_rate": 0.0001219088436359808, + "loss": 1.0176, + "step": 9666 + }, + { + "epoch": 1.7211538461538463, + "grad_norm": 0.6924905180931091, + "learning_rate": 0.00012189518601774178, + "loss": 0.855, + "step": 9667 + }, + { + "epoch": 1.7213319088319088, + "grad_norm": 0.6348826289176941, + "learning_rate": 0.00012188152797051202, + "loss": 1.1596, + "step": 9668 + }, + { + "epoch": 1.7215099715099715, + "grad_norm": 0.7170482873916626, + "learning_rate": 0.00012186786949455922, + "loss": 0.9811, + "step": 9669 + }, + { + "epoch": 1.721688034188034, + "grad_norm": 0.7471763491630554, + "learning_rate": 0.00012185421059015094, + "loss": 1.0925, + "step": 9670 + }, + { + "epoch": 1.7218660968660968, + "grad_norm": 0.6771119236946106, + "learning_rate": 0.00012184055125755481, + "loss": 0.9403, + "step": 9671 + }, + { + "epoch": 1.7220441595441596, + "grad_norm": 0.4335343539714813, + "learning_rate": 0.0001218268914970384, + "loss": 0.4925, + "step": 9672 + }, + { + "epoch": 1.7222222222222223, + "grad_norm": 0.6652585864067078, + "learning_rate": 0.00012181323130886943, + "loss": 0.7684, + "step": 9673 + }, + { + "epoch": 1.7224002849002849, + "grad_norm": 0.6465467810630798, + "learning_rate": 0.00012179957069331548, + "loss": 0.9011, + "step": 9674 + }, + { + "epoch": 1.7225783475783476, + "grad_norm": 0.6725688576698303, + "learning_rate": 0.00012178590965064427, + "loss": 0.9563, + "step": 9675 + }, + { + "epoch": 1.7227564102564101, + "grad_norm": 0.6223418712615967, + "learning_rate": 0.00012177224818112341, + "loss": 0.9099, + "step": 9676 + }, + { + "epoch": 1.7229344729344729, + "grad_norm": 0.79325270652771, + "learning_rate": 0.00012175858628502053, + "loss": 1.0318, + "step": 9677 + }, + { + "epoch": 1.7231125356125356, + "grad_norm": 0.6735602617263794, + "learning_rate": 0.0001217449239626034, + "loss": 1.0797, + "step": 9678 + }, + { + "epoch": 1.7232905982905984, + "grad_norm": 0.7082492113113403, + "learning_rate": 0.00012173126121413962, + "loss": 1.1341, + "step": 9679 + }, + { + "epoch": 1.723468660968661, + "grad_norm": 0.6563859581947327, + "learning_rate": 0.00012171759803989696, + "loss": 0.8778, + "step": 9680 + }, + { + "epoch": 1.7236467236467237, + "grad_norm": 0.6867792010307312, + "learning_rate": 0.00012170393444014306, + "loss": 0.8301, + "step": 9681 + }, + { + "epoch": 1.7238247863247862, + "grad_norm": 0.7870511412620544, + "learning_rate": 0.00012169027041514562, + "loss": 0.9165, + "step": 9682 + }, + { + "epoch": 1.724002849002849, + "grad_norm": 0.8006493449211121, + "learning_rate": 0.00012167660596517241, + "loss": 1.0395, + "step": 9683 + }, + { + "epoch": 1.7241809116809117, + "grad_norm": 0.6936125159263611, + "learning_rate": 0.00012166294109049114, + "loss": 1.1037, + "step": 9684 + }, + { + "epoch": 1.7243589743589745, + "grad_norm": 0.8176514506340027, + "learning_rate": 0.00012164927579136956, + "loss": 0.8791, + "step": 9685 + }, + { + "epoch": 1.7245370370370372, + "grad_norm": 0.6948300004005432, + "learning_rate": 0.00012163561006807537, + "loss": 0.9292, + "step": 9686 + }, + { + "epoch": 1.7247150997150997, + "grad_norm": 0.6237453818321228, + "learning_rate": 0.00012162194392087634, + "loss": 0.8553, + "step": 9687 + }, + { + "epoch": 1.7248931623931623, + "grad_norm": 0.6198007464408875, + "learning_rate": 0.00012160827735004021, + "loss": 0.9599, + "step": 9688 + }, + { + "epoch": 1.725071225071225, + "grad_norm": 0.639838695526123, + "learning_rate": 0.00012159461035583482, + "loss": 0.9328, + "step": 9689 + }, + { + "epoch": 1.7252492877492878, + "grad_norm": 0.7264436483383179, + "learning_rate": 0.00012158094293852789, + "loss": 1.0247, + "step": 9690 + }, + { + "epoch": 1.7254273504273505, + "grad_norm": 0.6320534348487854, + "learning_rate": 0.00012156727509838721, + "loss": 1.1222, + "step": 9691 + }, + { + "epoch": 1.7256054131054133, + "grad_norm": 0.6204122304916382, + "learning_rate": 0.00012155360683568056, + "loss": 0.9765, + "step": 9692 + }, + { + "epoch": 1.7257834757834758, + "grad_norm": 0.7026457190513611, + "learning_rate": 0.00012153993815067579, + "loss": 1.0178, + "step": 9693 + }, + { + "epoch": 1.7259615384615383, + "grad_norm": 0.6471006870269775, + "learning_rate": 0.00012152626904364067, + "loss": 1.0035, + "step": 9694 + }, + { + "epoch": 1.726139601139601, + "grad_norm": 0.6875706911087036, + "learning_rate": 0.00012151259951484301, + "loss": 0.7921, + "step": 9695 + }, + { + "epoch": 1.7263176638176638, + "grad_norm": 0.6963251233100891, + "learning_rate": 0.00012149892956455067, + "loss": 0.9677, + "step": 9696 + }, + { + "epoch": 1.7264957264957266, + "grad_norm": 0.9077282547950745, + "learning_rate": 0.00012148525919303142, + "loss": 0.9362, + "step": 9697 + }, + { + "epoch": 1.7266737891737893, + "grad_norm": 0.7347434163093567, + "learning_rate": 0.00012147158840055319, + "loss": 0.8712, + "step": 9698 + }, + { + "epoch": 1.7268518518518519, + "grad_norm": 0.7206630110740662, + "learning_rate": 0.00012145791718738377, + "loss": 1.032, + "step": 9699 + }, + { + "epoch": 1.7270299145299144, + "grad_norm": 0.7174662947654724, + "learning_rate": 0.00012144424555379106, + "loss": 0.954, + "step": 9700 + }, + { + "epoch": 1.7272079772079771, + "grad_norm": 0.7442345023155212, + "learning_rate": 0.0001214305735000429, + "loss": 1.0709, + "step": 9701 + }, + { + "epoch": 1.72738603988604, + "grad_norm": 0.6154376268386841, + "learning_rate": 0.00012141690102640715, + "loss": 0.9365, + "step": 9702 + }, + { + "epoch": 1.7275641025641026, + "grad_norm": 0.6213796734809875, + "learning_rate": 0.00012140322813315172, + "loss": 0.8337, + "step": 9703 + }, + { + "epoch": 1.7277421652421654, + "grad_norm": 0.7682011127471924, + "learning_rate": 0.0001213895548205445, + "loss": 1.1579, + "step": 9704 + }, + { + "epoch": 1.727920227920228, + "grad_norm": 0.6796970963478088, + "learning_rate": 0.0001213758810888534, + "loss": 0.8875, + "step": 9705 + }, + { + "epoch": 1.7280982905982905, + "grad_norm": 0.7203732132911682, + "learning_rate": 0.0001213622069383463, + "loss": 0.7827, + "step": 9706 + }, + { + "epoch": 1.7282763532763532, + "grad_norm": 0.6151877045631409, + "learning_rate": 0.00012134853236929111, + "loss": 1.0282, + "step": 9707 + }, + { + "epoch": 1.728454415954416, + "grad_norm": 0.6665124297142029, + "learning_rate": 0.0001213348573819558, + "loss": 1.0636, + "step": 9708 + }, + { + "epoch": 1.7286324786324787, + "grad_norm": 0.7334614396095276, + "learning_rate": 0.00012132118197660829, + "loss": 1.0889, + "step": 9709 + }, + { + "epoch": 1.7288105413105415, + "grad_norm": 0.7267759442329407, + "learning_rate": 0.00012130750615351649, + "loss": 1.096, + "step": 9710 + }, + { + "epoch": 1.728988603988604, + "grad_norm": 0.6542944312095642, + "learning_rate": 0.00012129382991294837, + "loss": 1.0855, + "step": 9711 + }, + { + "epoch": 1.7291666666666665, + "grad_norm": 0.694523274898529, + "learning_rate": 0.00012128015325517193, + "loss": 0.8482, + "step": 9712 + }, + { + "epoch": 1.7293447293447293, + "grad_norm": 0.7879082560539246, + "learning_rate": 0.00012126647618045504, + "loss": 1.2356, + "step": 9713 + }, + { + "epoch": 1.729522792022792, + "grad_norm": 0.7108420729637146, + "learning_rate": 0.00012125279868906574, + "loss": 1.0185, + "step": 9714 + }, + { + "epoch": 1.7297008547008548, + "grad_norm": 0.6928725838661194, + "learning_rate": 0.000121239120781272, + "loss": 1.1507, + "step": 9715 + }, + { + "epoch": 1.7298789173789175, + "grad_norm": 0.6195241212844849, + "learning_rate": 0.00012122544245734182, + "loss": 0.8656, + "step": 9716 + }, + { + "epoch": 1.73005698005698, + "grad_norm": 0.5962017774581909, + "learning_rate": 0.00012121176371754317, + "loss": 0.918, + "step": 9717 + }, + { + "epoch": 1.7302350427350426, + "grad_norm": 0.7409394979476929, + "learning_rate": 0.00012119808456214407, + "loss": 1.0283, + "step": 9718 + }, + { + "epoch": 1.7304131054131053, + "grad_norm": 0.6571973562240601, + "learning_rate": 0.00012118440499141257, + "loss": 1.1015, + "step": 9719 + }, + { + "epoch": 1.730591168091168, + "grad_norm": 0.681394100189209, + "learning_rate": 0.00012117072500561664, + "loss": 0.8247, + "step": 9720 + }, + { + "epoch": 1.7307692307692308, + "grad_norm": 0.7278251647949219, + "learning_rate": 0.00012115704460502432, + "loss": 1.0693, + "step": 9721 + }, + { + "epoch": 1.7309472934472936, + "grad_norm": 0.6569405794143677, + "learning_rate": 0.0001211433637899037, + "loss": 0.8992, + "step": 9722 + }, + { + "epoch": 1.7311253561253561, + "grad_norm": 0.6305136680603027, + "learning_rate": 0.00012112968256052272, + "loss": 0.8543, + "step": 9723 + }, + { + "epoch": 1.7313034188034186, + "grad_norm": 0.6111339330673218, + "learning_rate": 0.00012111600091714956, + "loss": 0.991, + "step": 9724 + }, + { + "epoch": 1.7314814814814814, + "grad_norm": 0.646973192691803, + "learning_rate": 0.00012110231886005223, + "loss": 0.8855, + "step": 9725 + }, + { + "epoch": 1.7316595441595442, + "grad_norm": 0.7054407000541687, + "learning_rate": 0.00012108863638949879, + "loss": 1.0816, + "step": 9726 + }, + { + "epoch": 1.731837606837607, + "grad_norm": 0.6592162847518921, + "learning_rate": 0.00012107495350575729, + "loss": 1.0961, + "step": 9727 + }, + { + "epoch": 1.7320156695156697, + "grad_norm": 0.6615595817565918, + "learning_rate": 0.00012106127020909587, + "loss": 0.9669, + "step": 9728 + }, + { + "epoch": 1.7321937321937322, + "grad_norm": 0.9030881524085999, + "learning_rate": 0.00012104758649978263, + "loss": 0.9438, + "step": 9729 + }, + { + "epoch": 1.7323717948717947, + "grad_norm": 0.6776516437530518, + "learning_rate": 0.00012103390237808566, + "loss": 0.8967, + "step": 9730 + }, + { + "epoch": 1.7325498575498575, + "grad_norm": 0.6010605096817017, + "learning_rate": 0.00012102021784427306, + "loss": 0.8893, + "step": 9731 + }, + { + "epoch": 1.7327279202279202, + "grad_norm": 0.6540384292602539, + "learning_rate": 0.00012100653289861295, + "loss": 0.9328, + "step": 9732 + }, + { + "epoch": 1.732905982905983, + "grad_norm": 0.6836950182914734, + "learning_rate": 0.00012099284754137345, + "loss": 0.9019, + "step": 9733 + }, + { + "epoch": 1.7330840455840457, + "grad_norm": 0.7597874402999878, + "learning_rate": 0.00012097916177282274, + "loss": 1.0093, + "step": 9734 + }, + { + "epoch": 1.7332621082621082, + "grad_norm": 0.7686513066291809, + "learning_rate": 0.00012096547559322892, + "loss": 0.8685, + "step": 9735 + }, + { + "epoch": 1.7334401709401708, + "grad_norm": 0.613777220249176, + "learning_rate": 0.0001209517890028602, + "loss": 0.8317, + "step": 9736 + }, + { + "epoch": 1.7336182336182335, + "grad_norm": 0.6788455843925476, + "learning_rate": 0.00012093810200198466, + "loss": 0.866, + "step": 9737 + }, + { + "epoch": 1.7337962962962963, + "grad_norm": 0.616801381111145, + "learning_rate": 0.00012092441459087047, + "loss": 0.8299, + "step": 9738 + }, + { + "epoch": 1.733974358974359, + "grad_norm": 0.731987476348877, + "learning_rate": 0.00012091072676978589, + "loss": 1.089, + "step": 9739 + }, + { + "epoch": 1.7341524216524218, + "grad_norm": 0.7042871117591858, + "learning_rate": 0.00012089703853899905, + "loss": 0.8667, + "step": 9740 + }, + { + "epoch": 1.7343304843304843, + "grad_norm": 0.62722247838974, + "learning_rate": 0.00012088334989877817, + "loss": 0.9185, + "step": 9741 + }, + { + "epoch": 1.734508547008547, + "grad_norm": 0.6354684829711914, + "learning_rate": 0.0001208696608493914, + "loss": 0.9951, + "step": 9742 + }, + { + "epoch": 1.7346866096866096, + "grad_norm": 0.658647894859314, + "learning_rate": 0.00012085597139110698, + "loss": 0.9324, + "step": 9743 + }, + { + "epoch": 1.7348646723646723, + "grad_norm": 0.84359210729599, + "learning_rate": 0.00012084228152419312, + "loss": 1.0861, + "step": 9744 + }, + { + "epoch": 1.735042735042735, + "grad_norm": 0.6293938755989075, + "learning_rate": 0.00012082859124891807, + "loss": 0.9676, + "step": 9745 + }, + { + "epoch": 1.7352207977207978, + "grad_norm": 0.6398760676383972, + "learning_rate": 0.00012081490056555004, + "loss": 0.8502, + "step": 9746 + }, + { + "epoch": 1.7353988603988604, + "grad_norm": 0.6918041706085205, + "learning_rate": 0.00012080120947435726, + "loss": 1.0081, + "step": 9747 + }, + { + "epoch": 1.7355769230769231, + "grad_norm": 0.7374079823493958, + "learning_rate": 0.00012078751797560798, + "loss": 0.9485, + "step": 9748 + }, + { + "epoch": 1.7357549857549857, + "grad_norm": 0.7392128705978394, + "learning_rate": 0.00012077382606957049, + "loss": 0.9283, + "step": 9749 + }, + { + "epoch": 1.7359330484330484, + "grad_norm": 0.701320230960846, + "learning_rate": 0.00012076013375651303, + "loss": 1.0339, + "step": 9750 + }, + { + "epoch": 1.7361111111111112, + "grad_norm": 0.6316696405410767, + "learning_rate": 0.00012074644103670387, + "loss": 0.9097, + "step": 9751 + }, + { + "epoch": 1.736289173789174, + "grad_norm": 0.6892024278640747, + "learning_rate": 0.00012073274791041132, + "loss": 1.0863, + "step": 9752 + }, + { + "epoch": 1.7364672364672364, + "grad_norm": 0.6032847762107849, + "learning_rate": 0.00012071905437790361, + "loss": 0.9305, + "step": 9753 + }, + { + "epoch": 1.7366452991452992, + "grad_norm": 0.6659184098243713, + "learning_rate": 0.00012070536043944907, + "loss": 0.9793, + "step": 9754 + }, + { + "epoch": 1.7368233618233617, + "grad_norm": 0.7413665056228638, + "learning_rate": 0.00012069166609531602, + "loss": 1.0523, + "step": 9755 + }, + { + "epoch": 1.7370014245014245, + "grad_norm": 0.7814368009567261, + "learning_rate": 0.00012067797134577275, + "loss": 0.9988, + "step": 9756 + }, + { + "epoch": 1.7371794871794872, + "grad_norm": 0.6174948811531067, + "learning_rate": 0.00012066427619108757, + "loss": 0.9002, + "step": 9757 + }, + { + "epoch": 1.73735754985755, + "grad_norm": 0.6521819233894348, + "learning_rate": 0.00012065058063152885, + "loss": 1.1307, + "step": 9758 + }, + { + "epoch": 1.7375356125356125, + "grad_norm": 0.6797493696212769, + "learning_rate": 0.00012063688466736489, + "loss": 0.84, + "step": 9759 + }, + { + "epoch": 1.7377136752136753, + "grad_norm": 0.6496474146842957, + "learning_rate": 0.00012062318829886404, + "loss": 0.86, + "step": 9760 + }, + { + "epoch": 1.7378917378917378, + "grad_norm": 0.6701306104660034, + "learning_rate": 0.00012060949152629467, + "loss": 0.9422, + "step": 9761 + }, + { + "epoch": 1.7380698005698005, + "grad_norm": 0.7331172823905945, + "learning_rate": 0.00012059579434992512, + "loss": 1.1648, + "step": 9762 + }, + { + "epoch": 1.7382478632478633, + "grad_norm": 0.63930743932724, + "learning_rate": 0.00012058209677002375, + "loss": 1.0617, + "step": 9763 + }, + { + "epoch": 1.738425925925926, + "grad_norm": 0.668851912021637, + "learning_rate": 0.00012056839878685895, + "loss": 0.8219, + "step": 9764 + }, + { + "epoch": 1.7386039886039886, + "grad_norm": 0.7305747270584106, + "learning_rate": 0.00012055470040069912, + "loss": 1.0416, + "step": 9765 + }, + { + "epoch": 1.7387820512820513, + "grad_norm": 0.6931866407394409, + "learning_rate": 0.00012054100161181264, + "loss": 1.0588, + "step": 9766 + }, + { + "epoch": 1.7389601139601139, + "grad_norm": 0.6565485000610352, + "learning_rate": 0.00012052730242046785, + "loss": 0.7885, + "step": 9767 + }, + { + "epoch": 1.7391381766381766, + "grad_norm": 0.739985466003418, + "learning_rate": 0.00012051360282693327, + "loss": 1.0973, + "step": 9768 + }, + { + "epoch": 1.7393162393162394, + "grad_norm": 0.6477079391479492, + "learning_rate": 0.00012049990283147723, + "loss": 0.9841, + "step": 9769 + }, + { + "epoch": 1.739494301994302, + "grad_norm": 0.7018330097198486, + "learning_rate": 0.00012048620243436819, + "loss": 1.0869, + "step": 9770 + }, + { + "epoch": 1.7396723646723646, + "grad_norm": 0.7087421417236328, + "learning_rate": 0.00012047250163587456, + "loss": 0.916, + "step": 9771 + }, + { + "epoch": 1.7398504273504274, + "grad_norm": 0.8747151494026184, + "learning_rate": 0.00012045880043626481, + "loss": 0.8245, + "step": 9772 + }, + { + "epoch": 1.74002849002849, + "grad_norm": 0.777498722076416, + "learning_rate": 0.00012044509883580735, + "loss": 1.071, + "step": 9773 + }, + { + "epoch": 1.7402065527065527, + "grad_norm": 0.6668971180915833, + "learning_rate": 0.00012043139683477062, + "loss": 1.0447, + "step": 9774 + }, + { + "epoch": 1.7403846153846154, + "grad_norm": 0.6702026724815369, + "learning_rate": 0.00012041769443342317, + "loss": 0.8688, + "step": 9775 + }, + { + "epoch": 1.7405626780626782, + "grad_norm": 0.7866267561912537, + "learning_rate": 0.00012040399163203337, + "loss": 1.0842, + "step": 9776 + }, + { + "epoch": 1.7407407407407407, + "grad_norm": 0.7655110955238342, + "learning_rate": 0.00012039028843086977, + "loss": 1.2417, + "step": 9777 + }, + { + "epoch": 1.7409188034188035, + "grad_norm": 0.7084119915962219, + "learning_rate": 0.0001203765848302008, + "loss": 0.9844, + "step": 9778 + }, + { + "epoch": 1.741096866096866, + "grad_norm": 0.7135398983955383, + "learning_rate": 0.00012036288083029497, + "loss": 1.1102, + "step": 9779 + }, + { + "epoch": 1.7412749287749287, + "grad_norm": 0.6784615516662598, + "learning_rate": 0.0001203491764314208, + "loss": 1.0349, + "step": 9780 + }, + { + "epoch": 1.7414529914529915, + "grad_norm": 0.7170301079750061, + "learning_rate": 0.00012033547163384682, + "loss": 1.0899, + "step": 9781 + }, + { + "epoch": 1.7416310541310542, + "grad_norm": 0.6692060828208923, + "learning_rate": 0.0001203217664378415, + "loss": 1.0486, + "step": 9782 + }, + { + "epoch": 1.7418091168091168, + "grad_norm": 0.6730037927627563, + "learning_rate": 0.00012030806084367336, + "loss": 0.9684, + "step": 9783 + }, + { + "epoch": 1.7419871794871795, + "grad_norm": 0.5983504056930542, + "learning_rate": 0.00012029435485161096, + "loss": 0.7106, + "step": 9784 + }, + { + "epoch": 1.742165242165242, + "grad_norm": 0.6834231615066528, + "learning_rate": 0.00012028064846192284, + "loss": 0.803, + "step": 9785 + }, + { + "epoch": 1.7423433048433048, + "grad_norm": 0.621046245098114, + "learning_rate": 0.00012026694167487755, + "loss": 0.9129, + "step": 9786 + }, + { + "epoch": 1.7425213675213675, + "grad_norm": 0.6348989605903625, + "learning_rate": 0.00012025323449074361, + "loss": 1.0076, + "step": 9787 + }, + { + "epoch": 1.7426994301994303, + "grad_norm": 0.6139974594116211, + "learning_rate": 0.00012023952690978966, + "loss": 1.0756, + "step": 9788 + }, + { + "epoch": 1.7428774928774928, + "grad_norm": 0.6473259925842285, + "learning_rate": 0.00012022581893228419, + "loss": 1.0568, + "step": 9789 + }, + { + "epoch": 1.7430555555555556, + "grad_norm": 0.6133778095245361, + "learning_rate": 0.00012021211055849581, + "loss": 0.8722, + "step": 9790 + }, + { + "epoch": 1.743233618233618, + "grad_norm": 0.6934139728546143, + "learning_rate": 0.00012019840178869315, + "loss": 1.0329, + "step": 9791 + }, + { + "epoch": 1.7434116809116809, + "grad_norm": 0.6730150580406189, + "learning_rate": 0.00012018469262314474, + "loss": 0.9326, + "step": 9792 + }, + { + "epoch": 1.7435897435897436, + "grad_norm": 0.6805521249771118, + "learning_rate": 0.0001201709830621192, + "loss": 1.0527, + "step": 9793 + }, + { + "epoch": 1.7437678062678064, + "grad_norm": 0.6972569823265076, + "learning_rate": 0.00012015727310588516, + "loss": 1.0024, + "step": 9794 + }, + { + "epoch": 1.743945868945869, + "grad_norm": 0.7329187989234924, + "learning_rate": 0.00012014356275471122, + "loss": 1.1864, + "step": 9795 + }, + { + "epoch": 1.7441239316239316, + "grad_norm": 0.7220240831375122, + "learning_rate": 0.00012012985200886602, + "loss": 0.8831, + "step": 9796 + }, + { + "epoch": 1.7443019943019942, + "grad_norm": 0.7829749584197998, + "learning_rate": 0.00012011614086861818, + "loss": 1.0365, + "step": 9797 + }, + { + "epoch": 1.744480056980057, + "grad_norm": 0.7148944735527039, + "learning_rate": 0.00012010242933423637, + "loss": 1.0413, + "step": 9798 + }, + { + "epoch": 1.7446581196581197, + "grad_norm": 0.5607262253761292, + "learning_rate": 0.00012008871740598917, + "loss": 0.8154, + "step": 9799 + }, + { + "epoch": 1.7448361823361824, + "grad_norm": 0.754626452922821, + "learning_rate": 0.00012007500508414531, + "loss": 1.0569, + "step": 9800 + }, + { + "epoch": 1.7450142450142452, + "grad_norm": 0.7216293215751648, + "learning_rate": 0.00012006129236897343, + "loss": 1.1641, + "step": 9801 + }, + { + "epoch": 1.7451923076923077, + "grad_norm": 0.6575515270233154, + "learning_rate": 0.0001200475792607422, + "loss": 0.9063, + "step": 9802 + }, + { + "epoch": 1.7453703703703702, + "grad_norm": 0.7411505579948425, + "learning_rate": 0.00012003386575972031, + "loss": 0.9791, + "step": 9803 + }, + { + "epoch": 1.745548433048433, + "grad_norm": 0.6945903301239014, + "learning_rate": 0.0001200201518661764, + "loss": 0.8111, + "step": 9804 + }, + { + "epoch": 1.7457264957264957, + "grad_norm": 0.5760970115661621, + "learning_rate": 0.00012000643758037924, + "loss": 1.1054, + "step": 9805 + }, + { + "epoch": 1.7459045584045585, + "grad_norm": 0.6732224225997925, + "learning_rate": 0.00011999272290259748, + "loss": 0.8992, + "step": 9806 + }, + { + "epoch": 1.7460826210826212, + "grad_norm": 0.673270046710968, + "learning_rate": 0.00011997900783309983, + "loss": 1.0554, + "step": 9807 + }, + { + "epoch": 1.7462606837606838, + "grad_norm": 0.7233314514160156, + "learning_rate": 0.00011996529237215503, + "loss": 1.066, + "step": 9808 + }, + { + "epoch": 1.7464387464387463, + "grad_norm": 0.7016494274139404, + "learning_rate": 0.00011995157652003183, + "loss": 0.891, + "step": 9809 + }, + { + "epoch": 1.746616809116809, + "grad_norm": 0.9377092719078064, + "learning_rate": 0.00011993786027699889, + "loss": 0.8626, + "step": 9810 + }, + { + "epoch": 1.7467948717948718, + "grad_norm": 0.6825845241546631, + "learning_rate": 0.00011992414364332503, + "loss": 0.8996, + "step": 9811 + }, + { + "epoch": 1.7469729344729346, + "grad_norm": 0.6836053729057312, + "learning_rate": 0.00011991042661927896, + "loss": 0.9338, + "step": 9812 + }, + { + "epoch": 1.7471509971509973, + "grad_norm": 0.6462908387184143, + "learning_rate": 0.00011989670920512943, + "loss": 1.1185, + "step": 9813 + }, + { + "epoch": 1.7473290598290598, + "grad_norm": 0.7191921472549438, + "learning_rate": 0.00011988299140114522, + "loss": 0.9084, + "step": 9814 + }, + { + "epoch": 1.7475071225071224, + "grad_norm": 0.6951598525047302, + "learning_rate": 0.00011986927320759508, + "loss": 1.0653, + "step": 9815 + }, + { + "epoch": 1.7476851851851851, + "grad_norm": 0.7512598037719727, + "learning_rate": 0.00011985555462474784, + "loss": 1.0259, + "step": 9816 + }, + { + "epoch": 1.7478632478632479, + "grad_norm": 0.6885492205619812, + "learning_rate": 0.00011984183565287226, + "loss": 0.7148, + "step": 9817 + }, + { + "epoch": 1.7480413105413106, + "grad_norm": 0.6880139708518982, + "learning_rate": 0.00011982811629223709, + "loss": 1.1567, + "step": 9818 + }, + { + "epoch": 1.7482193732193734, + "grad_norm": 0.7381170392036438, + "learning_rate": 0.0001198143965431112, + "loss": 0.8483, + "step": 9819 + }, + { + "epoch": 1.748397435897436, + "grad_norm": 0.6761063933372498, + "learning_rate": 0.00011980067640576333, + "loss": 0.9498, + "step": 9820 + }, + { + "epoch": 1.7485754985754984, + "grad_norm": 0.6454669237136841, + "learning_rate": 0.00011978695588046238, + "loss": 0.7336, + "step": 9821 + }, + { + "epoch": 1.7487535612535612, + "grad_norm": 0.6026871800422668, + "learning_rate": 0.00011977323496747712, + "loss": 0.8618, + "step": 9822 + }, + { + "epoch": 1.748931623931624, + "grad_norm": 0.6877408027648926, + "learning_rate": 0.0001197595136670764, + "loss": 0.9146, + "step": 9823 + }, + { + "epoch": 1.7491096866096867, + "grad_norm": 0.6874892115592957, + "learning_rate": 0.00011974579197952906, + "loss": 1.1628, + "step": 9824 + }, + { + "epoch": 1.7492877492877494, + "grad_norm": 0.7464384436607361, + "learning_rate": 0.00011973206990510393, + "loss": 1.007, + "step": 9825 + }, + { + "epoch": 1.749465811965812, + "grad_norm": 0.7281473278999329, + "learning_rate": 0.00011971834744406986, + "loss": 1.0776, + "step": 9826 + }, + { + "epoch": 1.7496438746438745, + "grad_norm": 0.6112284660339355, + "learning_rate": 0.00011970462459669575, + "loss": 0.7616, + "step": 9827 + }, + { + "epoch": 1.7498219373219372, + "grad_norm": 0.6498035192489624, + "learning_rate": 0.00011969090136325048, + "loss": 0.884, + "step": 9828 + }, + { + "epoch": 1.7498219373219372, + "eval_loss": 1.1018389463424683, + "eval_runtime": 24.5594, + "eval_samples_per_second": 42.387, + "eval_steps_per_second": 21.214, + "step": 9828 + }, + { + "epoch": 1.75, + "grad_norm": 0.6746426224708557, + "learning_rate": 0.00011967717774400289, + "loss": 0.9023, + "step": 9829 + }, + { + "epoch": 1.7501780626780628, + "grad_norm": 0.6513423323631287, + "learning_rate": 0.00011966345373922188, + "loss": 0.9786, + "step": 9830 + }, + { + "epoch": 1.7503561253561255, + "grad_norm": 0.7053804397583008, + "learning_rate": 0.00011964972934917632, + "loss": 1.0667, + "step": 9831 + }, + { + "epoch": 1.750534188034188, + "grad_norm": 0.6769008040428162, + "learning_rate": 0.00011963600457413513, + "loss": 0.8596, + "step": 9832 + }, + { + "epoch": 1.7507122507122506, + "grad_norm": 0.7162246108055115, + "learning_rate": 0.00011962227941436725, + "loss": 1.0746, + "step": 9833 + }, + { + "epoch": 1.7508903133903133, + "grad_norm": 0.7665811777114868, + "learning_rate": 0.00011960855387014156, + "loss": 1.0056, + "step": 9834 + }, + { + "epoch": 1.751068376068376, + "grad_norm": 0.6186950206756592, + "learning_rate": 0.00011959482794172696, + "loss": 0.9016, + "step": 9835 + }, + { + "epoch": 1.7512464387464388, + "grad_norm": 0.8018904328346252, + "learning_rate": 0.00011958110162939245, + "loss": 0.9534, + "step": 9836 + }, + { + "epoch": 1.7514245014245016, + "grad_norm": 0.8239033818244934, + "learning_rate": 0.0001195673749334069, + "loss": 1.214, + "step": 9837 + }, + { + "epoch": 1.751602564102564, + "grad_norm": 0.7886297106742859, + "learning_rate": 0.00011955364785403931, + "loss": 0.9672, + "step": 9838 + }, + { + "epoch": 1.7517806267806266, + "grad_norm": 0.6463177800178528, + "learning_rate": 0.00011953992039155862, + "loss": 0.9184, + "step": 9839 + }, + { + "epoch": 1.7519586894586894, + "grad_norm": 0.7374706864356995, + "learning_rate": 0.00011952619254623374, + "loss": 0.9988, + "step": 9840 + }, + { + "epoch": 1.7521367521367521, + "grad_norm": 0.7456657886505127, + "learning_rate": 0.00011951246431833369, + "loss": 1.2197, + "step": 9841 + }, + { + "epoch": 1.7523148148148149, + "grad_norm": 0.6644248962402344, + "learning_rate": 0.00011949873570812746, + "loss": 0.9449, + "step": 9842 + }, + { + "epoch": 1.7524928774928776, + "grad_norm": 0.707919180393219, + "learning_rate": 0.000119485006715884, + "loss": 0.774, + "step": 9843 + }, + { + "epoch": 1.7526709401709402, + "grad_norm": 0.6273906826972961, + "learning_rate": 0.00011947127734187231, + "loss": 0.8682, + "step": 9844 + }, + { + "epoch": 1.7528490028490027, + "grad_norm": 0.8335350155830383, + "learning_rate": 0.00011945754758636136, + "loss": 1.2282, + "step": 9845 + }, + { + "epoch": 1.7530270655270654, + "grad_norm": 0.6849051117897034, + "learning_rate": 0.00011944381744962022, + "loss": 1.1091, + "step": 9846 + }, + { + "epoch": 1.7532051282051282, + "grad_norm": 0.8571760058403015, + "learning_rate": 0.00011943008693191781, + "loss": 0.9806, + "step": 9847 + }, + { + "epoch": 1.753383190883191, + "grad_norm": 0.7045019268989563, + "learning_rate": 0.00011941635603352328, + "loss": 0.9217, + "step": 9848 + }, + { + "epoch": 1.7535612535612537, + "grad_norm": 0.6820187568664551, + "learning_rate": 0.00011940262475470556, + "loss": 0.9983, + "step": 9849 + }, + { + "epoch": 1.7537393162393162, + "grad_norm": 0.7400697469711304, + "learning_rate": 0.00011938889309573374, + "loss": 0.9521, + "step": 9850 + }, + { + "epoch": 1.7539173789173788, + "grad_norm": 0.7027658820152283, + "learning_rate": 0.00011937516105687678, + "loss": 1.0749, + "step": 9851 + }, + { + "epoch": 1.7540954415954415, + "grad_norm": 0.6778307557106018, + "learning_rate": 0.00011936142863840382, + "loss": 1.0249, + "step": 9852 + }, + { + "epoch": 1.7542735042735043, + "grad_norm": 0.6787961721420288, + "learning_rate": 0.00011934769584058389, + "loss": 1.0014, + "step": 9853 + }, + { + "epoch": 1.754451566951567, + "grad_norm": 0.7515636086463928, + "learning_rate": 0.00011933396266368606, + "loss": 1.0351, + "step": 9854 + }, + { + "epoch": 1.7546296296296298, + "grad_norm": 0.6620134115219116, + "learning_rate": 0.00011932022910797938, + "loss": 1.0294, + "step": 9855 + }, + { + "epoch": 1.7548076923076923, + "grad_norm": 0.8260951638221741, + "learning_rate": 0.00011930649517373294, + "loss": 0.9078, + "step": 9856 + }, + { + "epoch": 1.7549857549857548, + "grad_norm": 0.7680675983428955, + "learning_rate": 0.00011929276086121584, + "loss": 0.92, + "step": 9857 + }, + { + "epoch": 1.7551638176638176, + "grad_norm": 0.7104191184043884, + "learning_rate": 0.00011927902617069717, + "loss": 0.9937, + "step": 9858 + }, + { + "epoch": 1.7553418803418803, + "grad_norm": 0.7185840606689453, + "learning_rate": 0.00011926529110244603, + "loss": 0.9775, + "step": 9859 + }, + { + "epoch": 1.755519943019943, + "grad_norm": 0.7114652991294861, + "learning_rate": 0.00011925155565673151, + "loss": 0.883, + "step": 9860 + }, + { + "epoch": 1.7556980056980058, + "grad_norm": 0.6906639337539673, + "learning_rate": 0.00011923781983382276, + "loss": 0.9789, + "step": 9861 + }, + { + "epoch": 1.7558760683760684, + "grad_norm": 0.706908106803894, + "learning_rate": 0.00011922408363398892, + "loss": 1.1186, + "step": 9862 + }, + { + "epoch": 1.756054131054131, + "grad_norm": 0.7532939910888672, + "learning_rate": 0.00011921034705749908, + "loss": 0.977, + "step": 9863 + }, + { + "epoch": 1.7562321937321936, + "grad_norm": 0.7397763729095459, + "learning_rate": 0.0001191966101046224, + "loss": 1.1121, + "step": 9864 + }, + { + "epoch": 1.7564102564102564, + "grad_norm": 0.6955398321151733, + "learning_rate": 0.00011918287277562801, + "loss": 1.0439, + "step": 9865 + }, + { + "epoch": 1.7565883190883191, + "grad_norm": 0.7485929727554321, + "learning_rate": 0.00011916913507078507, + "loss": 1.1644, + "step": 9866 + }, + { + "epoch": 1.756766381766382, + "grad_norm": 0.6337487101554871, + "learning_rate": 0.00011915539699036274, + "loss": 0.8216, + "step": 9867 + }, + { + "epoch": 1.7569444444444444, + "grad_norm": 0.6628872752189636, + "learning_rate": 0.00011914165853463022, + "loss": 0.9584, + "step": 9868 + }, + { + "epoch": 1.7571225071225072, + "grad_norm": 0.6577547788619995, + "learning_rate": 0.00011912791970385666, + "loss": 0.9484, + "step": 9869 + }, + { + "epoch": 1.7573005698005697, + "grad_norm": 0.6409304738044739, + "learning_rate": 0.00011911418049831127, + "loss": 1.1256, + "step": 9870 + }, + { + "epoch": 1.7574786324786325, + "grad_norm": 0.7499844431877136, + "learning_rate": 0.00011910044091826319, + "loss": 0.7991, + "step": 9871 + }, + { + "epoch": 1.7576566951566952, + "grad_norm": 0.6786715388298035, + "learning_rate": 0.00011908670096398165, + "loss": 1.0368, + "step": 9872 + }, + { + "epoch": 1.757834757834758, + "grad_norm": 0.6432101130485535, + "learning_rate": 0.00011907296063573585, + "loss": 0.9059, + "step": 9873 + }, + { + "epoch": 1.7580128205128205, + "grad_norm": 0.6542613506317139, + "learning_rate": 0.00011905921993379503, + "loss": 0.9866, + "step": 9874 + }, + { + "epoch": 1.7581908831908832, + "grad_norm": 0.6048218011856079, + "learning_rate": 0.00011904547885842838, + "loss": 0.9488, + "step": 9875 + }, + { + "epoch": 1.7583689458689458, + "grad_norm": 0.7694938778877258, + "learning_rate": 0.00011903173740990512, + "loss": 1.1026, + "step": 9876 + }, + { + "epoch": 1.7585470085470085, + "grad_norm": 0.6621627807617188, + "learning_rate": 0.00011901799558849451, + "loss": 1.135, + "step": 9877 + }, + { + "epoch": 1.7587250712250713, + "grad_norm": 0.6561587452888489, + "learning_rate": 0.0001190042533944658, + "loss": 0.9322, + "step": 9878 + }, + { + "epoch": 1.758903133903134, + "grad_norm": 0.7846759557723999, + "learning_rate": 0.00011899051082808821, + "loss": 0.9324, + "step": 9879 + }, + { + "epoch": 1.7590811965811965, + "grad_norm": 0.6004071831703186, + "learning_rate": 0.00011897676788963101, + "loss": 0.9641, + "step": 9880 + }, + { + "epoch": 1.7592592592592593, + "grad_norm": 0.6731070280075073, + "learning_rate": 0.00011896302457936344, + "loss": 1.1437, + "step": 9881 + }, + { + "epoch": 1.7594373219373218, + "grad_norm": 0.6768675446510315, + "learning_rate": 0.00011894928089755481, + "loss": 1.0707, + "step": 9882 + }, + { + "epoch": 1.7596153846153846, + "grad_norm": 0.8368878960609436, + "learning_rate": 0.0001189355368444744, + "loss": 1.0435, + "step": 9883 + }, + { + "epoch": 1.7597934472934473, + "grad_norm": 0.6132324934005737, + "learning_rate": 0.00011892179242039149, + "loss": 0.8889, + "step": 9884 + }, + { + "epoch": 1.75997150997151, + "grad_norm": 0.7598093152046204, + "learning_rate": 0.00011890804762557535, + "loss": 1.151, + "step": 9885 + }, + { + "epoch": 1.7601495726495726, + "grad_norm": 0.7317715883255005, + "learning_rate": 0.00011889430246029527, + "loss": 0.9992, + "step": 9886 + }, + { + "epoch": 1.7603276353276354, + "grad_norm": 0.7664858102798462, + "learning_rate": 0.00011888055692482059, + "loss": 0.8398, + "step": 9887 + }, + { + "epoch": 1.760505698005698, + "grad_norm": 0.6916853189468384, + "learning_rate": 0.00011886681101942063, + "loss": 0.9507, + "step": 9888 + }, + { + "epoch": 1.7606837606837606, + "grad_norm": 0.7103399634361267, + "learning_rate": 0.0001188530647443647, + "loss": 0.915, + "step": 9889 + }, + { + "epoch": 1.7608618233618234, + "grad_norm": 0.6177804470062256, + "learning_rate": 0.00011883931809992215, + "loss": 0.721, + "step": 9890 + }, + { + "epoch": 1.7610398860398861, + "grad_norm": 0.7523959279060364, + "learning_rate": 0.00011882557108636227, + "loss": 0.99, + "step": 9891 + }, + { + "epoch": 1.7612179487179487, + "grad_norm": 0.6211134791374207, + "learning_rate": 0.00011881182370395442, + "loss": 0.8089, + "step": 9892 + }, + { + "epoch": 1.7613960113960114, + "grad_norm": 0.6660307049751282, + "learning_rate": 0.00011879807595296802, + "loss": 1.1062, + "step": 9893 + }, + { + "epoch": 1.761574074074074, + "grad_norm": 0.7039240598678589, + "learning_rate": 0.00011878432783367232, + "loss": 0.9739, + "step": 9894 + }, + { + "epoch": 1.7617521367521367, + "grad_norm": 0.658064603805542, + "learning_rate": 0.00011877057934633675, + "loss": 0.9438, + "step": 9895 + }, + { + "epoch": 1.7619301994301995, + "grad_norm": 0.8227152228355408, + "learning_rate": 0.00011875683049123068, + "loss": 0.8385, + "step": 9896 + }, + { + "epoch": 1.7621082621082622, + "grad_norm": 0.6622483730316162, + "learning_rate": 0.00011874308126862346, + "loss": 0.9432, + "step": 9897 + }, + { + "epoch": 1.7622863247863247, + "grad_norm": 0.7211357951164246, + "learning_rate": 0.00011872933167878453, + "loss": 1.2471, + "step": 9898 + }, + { + "epoch": 1.7624643874643875, + "grad_norm": 0.6177424192428589, + "learning_rate": 0.00011871558172198322, + "loss": 0.8892, + "step": 9899 + }, + { + "epoch": 1.76264245014245, + "grad_norm": 0.6924285888671875, + "learning_rate": 0.00011870183139848898, + "loss": 1.021, + "step": 9900 + }, + { + "epoch": 1.7628205128205128, + "grad_norm": 0.6168648600578308, + "learning_rate": 0.0001186880807085712, + "loss": 0.9013, + "step": 9901 + }, + { + "epoch": 1.7629985754985755, + "grad_norm": 0.6410452723503113, + "learning_rate": 0.00011867432965249929, + "loss": 0.6686, + "step": 9902 + }, + { + "epoch": 1.7631766381766383, + "grad_norm": 0.6959559917449951, + "learning_rate": 0.0001186605782305427, + "loss": 0.9814, + "step": 9903 + }, + { + "epoch": 1.7633547008547008, + "grad_norm": 0.7456178069114685, + "learning_rate": 0.00011864682644297085, + "loss": 1.0151, + "step": 9904 + }, + { + "epoch": 1.7635327635327636, + "grad_norm": 0.6499991416931152, + "learning_rate": 0.00011863307429005317, + "loss": 0.83, + "step": 9905 + }, + { + "epoch": 1.763710826210826, + "grad_norm": 0.643344521522522, + "learning_rate": 0.00011861932177205908, + "loss": 0.8853, + "step": 9906 + }, + { + "epoch": 1.7638888888888888, + "grad_norm": 0.6570441722869873, + "learning_rate": 0.00011860556888925804, + "loss": 0.9179, + "step": 9907 + }, + { + "epoch": 1.7640669515669516, + "grad_norm": 0.6892307996749878, + "learning_rate": 0.00011859181564191957, + "loss": 0.9657, + "step": 9908 + }, + { + "epoch": 1.7642450142450143, + "grad_norm": 0.648158073425293, + "learning_rate": 0.0001185780620303131, + "loss": 0.9179, + "step": 9909 + }, + { + "epoch": 1.7644230769230769, + "grad_norm": 0.5833603143692017, + "learning_rate": 0.00011856430805470808, + "loss": 0.8505, + "step": 9910 + }, + { + "epoch": 1.7646011396011396, + "grad_norm": 0.8302416205406189, + "learning_rate": 0.000118550553715374, + "loss": 0.8948, + "step": 9911 + }, + { + "epoch": 1.7647792022792022, + "grad_norm": 0.7075300216674805, + "learning_rate": 0.00011853679901258035, + "loss": 1.2467, + "step": 9912 + }, + { + "epoch": 1.764957264957265, + "grad_norm": 0.81916344165802, + "learning_rate": 0.00011852304394659666, + "loss": 0.9963, + "step": 9913 + }, + { + "epoch": 1.7651353276353277, + "grad_norm": 0.6492435932159424, + "learning_rate": 0.00011850928851769239, + "loss": 1.0704, + "step": 9914 + }, + { + "epoch": 1.7653133903133904, + "grad_norm": 0.7301090359687805, + "learning_rate": 0.00011849553272613704, + "loss": 1.0477, + "step": 9915 + }, + { + "epoch": 1.765491452991453, + "grad_norm": 0.7280275821685791, + "learning_rate": 0.00011848177657220019, + "loss": 0.9124, + "step": 9916 + }, + { + "epoch": 1.7656695156695157, + "grad_norm": 0.6948845386505127, + "learning_rate": 0.00011846802005615127, + "loss": 1.2275, + "step": 9917 + }, + { + "epoch": 1.7658475783475782, + "grad_norm": 0.6553834676742554, + "learning_rate": 0.0001184542631782599, + "loss": 1.2311, + "step": 9918 + }, + { + "epoch": 1.766025641025641, + "grad_norm": 0.6899739503860474, + "learning_rate": 0.00011844050593879556, + "loss": 0.8936, + "step": 9919 + }, + { + "epoch": 1.7662037037037037, + "grad_norm": 0.6076815128326416, + "learning_rate": 0.00011842674833802782, + "loss": 0.8432, + "step": 9920 + }, + { + "epoch": 1.7663817663817665, + "grad_norm": 0.7650902271270752, + "learning_rate": 0.00011841299037622624, + "loss": 1.0447, + "step": 9921 + }, + { + "epoch": 1.7665598290598292, + "grad_norm": 0.6864938735961914, + "learning_rate": 0.00011839923205366032, + "loss": 0.936, + "step": 9922 + }, + { + "epoch": 1.7667378917378918, + "grad_norm": 0.7176852226257324, + "learning_rate": 0.0001183854733705997, + "loss": 0.9764, + "step": 9923 + }, + { + "epoch": 1.7669159544159543, + "grad_norm": 0.6513439416885376, + "learning_rate": 0.00011837171432731393, + "loss": 1.0095, + "step": 9924 + }, + { + "epoch": 1.767094017094017, + "grad_norm": 0.8031024932861328, + "learning_rate": 0.00011835795492407256, + "loss": 1.1348, + "step": 9925 + }, + { + "epoch": 1.7672720797720798, + "grad_norm": 0.7659830451011658, + "learning_rate": 0.00011834419516114518, + "loss": 0.9058, + "step": 9926 + }, + { + "epoch": 1.7674501424501425, + "grad_norm": 0.8864039778709412, + "learning_rate": 0.00011833043503880145, + "loss": 1.0342, + "step": 9927 + }, + { + "epoch": 1.7676282051282053, + "grad_norm": 0.6870512962341309, + "learning_rate": 0.00011831667455731088, + "loss": 0.9361, + "step": 9928 + }, + { + "epoch": 1.7678062678062678, + "grad_norm": 0.6458830833435059, + "learning_rate": 0.00011830291371694315, + "loss": 0.8215, + "step": 9929 + }, + { + "epoch": 1.7679843304843303, + "grad_norm": 0.7456086874008179, + "learning_rate": 0.00011828915251796787, + "loss": 1.1243, + "step": 9930 + }, + { + "epoch": 1.768162393162393, + "grad_norm": 0.6834850311279297, + "learning_rate": 0.00011827539096065459, + "loss": 0.9536, + "step": 9931 + }, + { + "epoch": 1.7683404558404558, + "grad_norm": 0.643864631652832, + "learning_rate": 0.00011826162904527302, + "loss": 1.1707, + "step": 9932 + }, + { + "epoch": 1.7685185185185186, + "grad_norm": 0.6312864422798157, + "learning_rate": 0.00011824786677209275, + "loss": 0.7937, + "step": 9933 + }, + { + "epoch": 1.7686965811965814, + "grad_norm": 0.6092729568481445, + "learning_rate": 0.00011823410414138343, + "loss": 0.8787, + "step": 9934 + }, + { + "epoch": 1.7688746438746439, + "grad_norm": 0.6859988570213318, + "learning_rate": 0.00011822034115341474, + "loss": 0.9691, + "step": 9935 + }, + { + "epoch": 1.7690527065527064, + "grad_norm": 0.7219935059547424, + "learning_rate": 0.0001182065778084563, + "loss": 1.0606, + "step": 9936 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.6596202850341797, + "learning_rate": 0.00011819281410677778, + "loss": 1.0543, + "step": 9937 + }, + { + "epoch": 1.769408831908832, + "grad_norm": 0.6616338491439819, + "learning_rate": 0.00011817905004864887, + "loss": 0.9757, + "step": 9938 + }, + { + "epoch": 1.7695868945868947, + "grad_norm": 0.6637360453605652, + "learning_rate": 0.00011816528563433924, + "loss": 0.925, + "step": 9939 + }, + { + "epoch": 1.7697649572649574, + "grad_norm": 0.8422333002090454, + "learning_rate": 0.00011815152086411859, + "loss": 1.1343, + "step": 9940 + }, + { + "epoch": 1.76994301994302, + "grad_norm": 0.6638204455375671, + "learning_rate": 0.00011813775573825656, + "loss": 1.2136, + "step": 9941 + }, + { + "epoch": 1.7701210826210825, + "grad_norm": 0.7258831858634949, + "learning_rate": 0.0001181239902570229, + "loss": 0.7308, + "step": 9942 + }, + { + "epoch": 1.7702991452991452, + "grad_norm": 0.730582594871521, + "learning_rate": 0.0001181102244206873, + "loss": 1.1097, + "step": 9943 + }, + { + "epoch": 1.770477207977208, + "grad_norm": 0.7324019074440002, + "learning_rate": 0.00011809645822951946, + "loss": 0.9802, + "step": 9944 + }, + { + "epoch": 1.7706552706552707, + "grad_norm": 0.5565997958183289, + "learning_rate": 0.00011808269168378914, + "loss": 0.7079, + "step": 9945 + }, + { + "epoch": 1.7708333333333335, + "grad_norm": 0.6395503282546997, + "learning_rate": 0.00011806892478376601, + "loss": 1.0048, + "step": 9946 + }, + { + "epoch": 1.771011396011396, + "grad_norm": 0.7670905590057373, + "learning_rate": 0.00011805515752971985, + "loss": 1.2509, + "step": 9947 + }, + { + "epoch": 1.7711894586894585, + "grad_norm": 0.5945813655853271, + "learning_rate": 0.00011804138992192037, + "loss": 0.8856, + "step": 9948 + }, + { + "epoch": 1.7713675213675213, + "grad_norm": 0.7355493307113647, + "learning_rate": 0.00011802762196063737, + "loss": 0.9629, + "step": 9949 + }, + { + "epoch": 1.771545584045584, + "grad_norm": 0.7024806141853333, + "learning_rate": 0.00011801385364614055, + "loss": 1.1351, + "step": 9950 + }, + { + "epoch": 1.7717236467236468, + "grad_norm": 0.6553003191947937, + "learning_rate": 0.00011800008497869968, + "loss": 0.911, + "step": 9951 + }, + { + "epoch": 1.7719017094017095, + "grad_norm": 0.6883971691131592, + "learning_rate": 0.00011798631595858454, + "loss": 1.0099, + "step": 9952 + }, + { + "epoch": 1.772079772079772, + "grad_norm": 0.7106832265853882, + "learning_rate": 0.00011797254658606489, + "loss": 1.0298, + "step": 9953 + }, + { + "epoch": 1.7722578347578346, + "grad_norm": 0.7902877926826477, + "learning_rate": 0.00011795877686141055, + "loss": 1.0572, + "step": 9954 + }, + { + "epoch": 1.7724358974358974, + "grad_norm": 0.7105007171630859, + "learning_rate": 0.00011794500678489126, + "loss": 1.1725, + "step": 9955 + }, + { + "epoch": 1.77261396011396, + "grad_norm": 0.7314959764480591, + "learning_rate": 0.00011793123635677685, + "loss": 1.1074, + "step": 9956 + }, + { + "epoch": 1.7727920227920229, + "grad_norm": 0.6358618140220642, + "learning_rate": 0.00011791746557733712, + "loss": 0.8786, + "step": 9957 + }, + { + "epoch": 1.7729700854700856, + "grad_norm": 0.6441367864608765, + "learning_rate": 0.00011790369444684187, + "loss": 1.1332, + "step": 9958 + }, + { + "epoch": 1.7731481481481481, + "grad_norm": 0.686787486076355, + "learning_rate": 0.0001178899229655609, + "loss": 0.9566, + "step": 9959 + }, + { + "epoch": 1.7733262108262107, + "grad_norm": 0.653840184211731, + "learning_rate": 0.00011787615113376407, + "loss": 0.8763, + "step": 9960 + }, + { + "epoch": 1.7735042735042734, + "grad_norm": 0.7106643915176392, + "learning_rate": 0.00011786237895172119, + "loss": 0.9929, + "step": 9961 + }, + { + "epoch": 1.7736823361823362, + "grad_norm": 0.6634044051170349, + "learning_rate": 0.0001178486064197021, + "loss": 0.7467, + "step": 9962 + }, + { + "epoch": 1.773860398860399, + "grad_norm": 0.7087352871894836, + "learning_rate": 0.00011783483353797663, + "loss": 1.0104, + "step": 9963 + }, + { + "epoch": 1.7740384615384617, + "grad_norm": 0.8088061213493347, + "learning_rate": 0.00011782106030681466, + "loss": 1.0376, + "step": 9964 + }, + { + "epoch": 1.7742165242165242, + "grad_norm": 0.7204688787460327, + "learning_rate": 0.00011780728672648604, + "loss": 0.8556, + "step": 9965 + }, + { + "epoch": 1.7743945868945867, + "grad_norm": 0.7893314957618713, + "learning_rate": 0.0001177935127972606, + "loss": 0.9764, + "step": 9966 + }, + { + "epoch": 1.7745726495726495, + "grad_norm": 0.6098896265029907, + "learning_rate": 0.00011777973851940826, + "loss": 0.9407, + "step": 9967 + }, + { + "epoch": 1.7747507122507122, + "grad_norm": 0.6420868039131165, + "learning_rate": 0.0001177659638931989, + "loss": 1.1328, + "step": 9968 + }, + { + "epoch": 1.774928774928775, + "grad_norm": 0.7732378244400024, + "learning_rate": 0.00011775218891890234, + "loss": 1.1236, + "step": 9969 + }, + { + "epoch": 1.7751068376068377, + "grad_norm": 0.6591582894325256, + "learning_rate": 0.00011773841359678855, + "loss": 1.1523, + "step": 9970 + }, + { + "epoch": 1.7752849002849003, + "grad_norm": 0.6337170004844666, + "learning_rate": 0.00011772463792712738, + "loss": 1.1998, + "step": 9971 + }, + { + "epoch": 1.7754629629629628, + "grad_norm": 0.6400532126426697, + "learning_rate": 0.00011771086191018874, + "loss": 0.9543, + "step": 9972 + }, + { + "epoch": 1.7756410256410255, + "grad_norm": 0.6431527733802795, + "learning_rate": 0.00011769708554624257, + "loss": 0.8164, + "step": 9973 + }, + { + "epoch": 1.7758190883190883, + "grad_norm": 0.7303599119186401, + "learning_rate": 0.00011768330883555876, + "loss": 0.9553, + "step": 9974 + }, + { + "epoch": 1.775997150997151, + "grad_norm": 0.7838605642318726, + "learning_rate": 0.00011766953177840725, + "loss": 0.9759, + "step": 9975 + }, + { + "epoch": 1.7761752136752138, + "grad_norm": 0.6505265831947327, + "learning_rate": 0.00011765575437505796, + "loss": 0.8527, + "step": 9976 + }, + { + "epoch": 1.7763532763532763, + "grad_norm": 0.7336180806159973, + "learning_rate": 0.00011764197662578086, + "loss": 1.1098, + "step": 9977 + }, + { + "epoch": 1.776531339031339, + "grad_norm": 0.7040138244628906, + "learning_rate": 0.00011762819853084586, + "loss": 1.1289, + "step": 9978 + }, + { + "epoch": 1.7767094017094016, + "grad_norm": 0.6414867043495178, + "learning_rate": 0.00011761442009052293, + "loss": 1.0826, + "step": 9979 + }, + { + "epoch": 1.7768874643874644, + "grad_norm": 0.6760666370391846, + "learning_rate": 0.00011760064130508204, + "loss": 1.0188, + "step": 9980 + }, + { + "epoch": 1.7770655270655271, + "grad_norm": 0.7864978909492493, + "learning_rate": 0.00011758686217479316, + "loss": 1.1938, + "step": 9981 + }, + { + "epoch": 1.7772435897435899, + "grad_norm": 0.7964870929718018, + "learning_rate": 0.00011757308269992622, + "loss": 0.9876, + "step": 9982 + }, + { + "epoch": 1.7774216524216524, + "grad_norm": 0.5158692002296448, + "learning_rate": 0.00011755930288075123, + "loss": 0.6508, + "step": 9983 + }, + { + "epoch": 1.7775997150997151, + "grad_norm": 0.7208606600761414, + "learning_rate": 0.00011754552271753819, + "loss": 1.0738, + "step": 9984 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 0.6811334490776062, + "learning_rate": 0.00011753174221055705, + "loss": 1.1216, + "step": 9985 + }, + { + "epoch": 1.7779558404558404, + "grad_norm": 0.6389986276626587, + "learning_rate": 0.00011751796136007787, + "loss": 0.9664, + "step": 9986 + }, + { + "epoch": 1.7781339031339032, + "grad_norm": 0.7081875205039978, + "learning_rate": 0.00011750418016637064, + "loss": 0.9365, + "step": 9987 + }, + { + "epoch": 1.778311965811966, + "grad_norm": 0.7291778326034546, + "learning_rate": 0.00011749039862970535, + "loss": 1.3222, + "step": 9988 + }, + { + "epoch": 1.7784900284900285, + "grad_norm": 0.6790453791618347, + "learning_rate": 0.000117476616750352, + "loss": 0.9537, + "step": 9989 + }, + { + "epoch": 1.7786680911680912, + "grad_norm": 0.6271076202392578, + "learning_rate": 0.00011746283452858069, + "loss": 0.9842, + "step": 9990 + }, + { + "epoch": 1.7788461538461537, + "grad_norm": 0.675628662109375, + "learning_rate": 0.00011744905196466138, + "loss": 0.8675, + "step": 9991 + }, + { + "epoch": 1.7790242165242165, + "grad_norm": 0.7328314185142517, + "learning_rate": 0.00011743526905886417, + "loss": 0.9793, + "step": 9992 + }, + { + "epoch": 1.7792022792022792, + "grad_norm": 0.698764979839325, + "learning_rate": 0.00011742148581145908, + "loss": 0.9527, + "step": 9993 + }, + { + "epoch": 1.779380341880342, + "grad_norm": 0.6911364793777466, + "learning_rate": 0.00011740770222271616, + "loss": 1.1069, + "step": 9994 + }, + { + "epoch": 1.7795584045584045, + "grad_norm": 0.6990836262702942, + "learning_rate": 0.00011739391829290547, + "loss": 0.9132, + "step": 9995 + }, + { + "epoch": 1.7797364672364673, + "grad_norm": 0.7056801319122314, + "learning_rate": 0.0001173801340222971, + "loss": 1.053, + "step": 9996 + }, + { + "epoch": 1.7799145299145298, + "grad_norm": 0.7453791499137878, + "learning_rate": 0.0001173663494111611, + "loss": 0.8806, + "step": 9997 + }, + { + "epoch": 1.7800925925925926, + "grad_norm": 0.7211771011352539, + "learning_rate": 0.00011735256445976757, + "loss": 0.9968, + "step": 9998 + }, + { + "epoch": 1.7802706552706553, + "grad_norm": 0.7259734272956848, + "learning_rate": 0.00011733877916838656, + "loss": 1.167, + "step": 9999 + }, + { + "epoch": 1.780448717948718, + "grad_norm": 0.6931926012039185, + "learning_rate": 0.00011732499353728821, + "loss": 1.0634, + "step": 10000 + }, + { + "epoch": 1.7806267806267806, + "grad_norm": 0.6900074481964111, + "learning_rate": 0.00011731120756674259, + "loss": 0.9718, + "step": 10001 + }, + { + "epoch": 1.7808048433048433, + "grad_norm": 0.6817582845687866, + "learning_rate": 0.00011729742125701984, + "loss": 1.0896, + "step": 10002 + }, + { + "epoch": 1.7809829059829059, + "grad_norm": 0.6901891231536865, + "learning_rate": 0.00011728363460839003, + "loss": 1.0163, + "step": 10003 + }, + { + "epoch": 1.7811609686609686, + "grad_norm": 0.9138323664665222, + "learning_rate": 0.00011726984762112328, + "loss": 1.1713, + "step": 10004 + }, + { + "epoch": 1.7813390313390314, + "grad_norm": 0.6105810403823853, + "learning_rate": 0.00011725606029548977, + "loss": 0.9331, + "step": 10005 + }, + { + "epoch": 1.7815170940170941, + "grad_norm": 0.5605259537696838, + "learning_rate": 0.0001172422726317596, + "loss": 0.7154, + "step": 10006 + }, + { + "epoch": 1.7816951566951567, + "grad_norm": 0.6950963735580444, + "learning_rate": 0.00011722848463020292, + "loss": 1.0093, + "step": 10007 + }, + { + "epoch": 1.7818732193732194, + "grad_norm": 0.6806309819221497, + "learning_rate": 0.00011721469629108988, + "loss": 0.8662, + "step": 10008 + }, + { + "epoch": 1.782051282051282, + "grad_norm": 0.7528520226478577, + "learning_rate": 0.00011720090761469063, + "loss": 0.8567, + "step": 10009 + }, + { + "epoch": 1.7822293447293447, + "grad_norm": 0.6617229580879211, + "learning_rate": 0.00011718711860127529, + "loss": 1.0378, + "step": 10010 + }, + { + "epoch": 1.7824074074074074, + "grad_norm": 0.6468376517295837, + "learning_rate": 0.00011717332925111411, + "loss": 1.0658, + "step": 10011 + }, + { + "epoch": 1.7825854700854702, + "grad_norm": 0.7141897082328796, + "learning_rate": 0.00011715953956447721, + "loss": 1.023, + "step": 10012 + }, + { + "epoch": 1.7827635327635327, + "grad_norm": 0.5777570605278015, + "learning_rate": 0.00011714574954163475, + "loss": 0.9154, + "step": 10013 + }, + { + "epoch": 1.7829415954415955, + "grad_norm": 0.7536137700080872, + "learning_rate": 0.00011713195918285695, + "loss": 0.9651, + "step": 10014 + }, + { + "epoch": 1.783119658119658, + "grad_norm": 0.6977683305740356, + "learning_rate": 0.00011711816848841402, + "loss": 0.7977, + "step": 10015 + }, + { + "epoch": 1.7832977207977208, + "grad_norm": 0.6522472500801086, + "learning_rate": 0.00011710437745857614, + "loss": 0.8834, + "step": 10016 + }, + { + "epoch": 1.7834757834757835, + "grad_norm": 0.6263057589530945, + "learning_rate": 0.0001170905860936135, + "loss": 1.0576, + "step": 10017 + }, + { + "epoch": 1.7836538461538463, + "grad_norm": 0.6470699310302734, + "learning_rate": 0.00011707679439379635, + "loss": 0.9412, + "step": 10018 + }, + { + "epoch": 1.7838319088319088, + "grad_norm": Infinity, + "learning_rate": 0.00011707679439379635, + "loss": 1.1746, + "step": 10019 + }, + { + "epoch": 1.7840099715099715, + "grad_norm": 0.6022017002105713, + "learning_rate": 0.00011706300235939485, + "loss": 0.8945, + "step": 10020 + }, + { + "epoch": 1.784188034188034, + "grad_norm": 0.637208104133606, + "learning_rate": 0.00011704920999067927, + "loss": 1.0215, + "step": 10021 + }, + { + "epoch": 1.7843660968660968, + "grad_norm": 0.7467851042747498, + "learning_rate": 0.00011703541728791987, + "loss": 1.0341, + "step": 10022 + }, + { + "epoch": 1.7845441595441596, + "grad_norm": 0.7562711238861084, + "learning_rate": 0.00011702162425138683, + "loss": 0.9748, + "step": 10023 + }, + { + "epoch": 1.7847222222222223, + "grad_norm": 0.6480089426040649, + "learning_rate": 0.00011700783088135043, + "loss": 1.05, + "step": 10024 + }, + { + "epoch": 1.7849002849002849, + "grad_norm": 0.6293981671333313, + "learning_rate": 0.00011699403717808091, + "loss": 1.0376, + "step": 10025 + }, + { + "epoch": 1.7850783475783476, + "grad_norm": 0.6821253895759583, + "learning_rate": 0.00011698024314184853, + "loss": 1.0542, + "step": 10026 + }, + { + "epoch": 1.7852564102564101, + "grad_norm": 0.6681216359138489, + "learning_rate": 0.00011696644877292356, + "loss": 1.0018, + "step": 10027 + }, + { + "epoch": 1.7854344729344729, + "grad_norm": 0.6788804531097412, + "learning_rate": 0.00011695265407157628, + "loss": 1.1823, + "step": 10028 + }, + { + "epoch": 1.7856125356125356, + "grad_norm": 0.6147881150245667, + "learning_rate": 0.00011693885903807697, + "loss": 0.9246, + "step": 10029 + }, + { + "epoch": 1.7857905982905984, + "grad_norm": 0.7952296137809753, + "learning_rate": 0.00011692506367269588, + "loss": 1.0528, + "step": 10030 + }, + { + "epoch": 1.785968660968661, + "grad_norm": 0.6985954642295837, + "learning_rate": 0.00011691126797570333, + "loss": 0.9173, + "step": 10031 + }, + { + "epoch": 1.7861467236467237, + "grad_norm": 0.6211223602294922, + "learning_rate": 0.00011689747194736961, + "loss": 0.7527, + "step": 10032 + }, + { + "epoch": 1.7863247863247862, + "grad_norm": 0.7531208992004395, + "learning_rate": 0.00011688367558796507, + "loss": 1.1087, + "step": 10033 + }, + { + "epoch": 1.786502849002849, + "grad_norm": 0.7742924690246582, + "learning_rate": 0.00011686987889775996, + "loss": 1.1512, + "step": 10034 + }, + { + "epoch": 1.7866809116809117, + "grad_norm": 0.7046231627464294, + "learning_rate": 0.00011685608187702459, + "loss": 1.0516, + "step": 10035 + }, + { + "epoch": 1.7868589743589745, + "grad_norm": 0.6264076232910156, + "learning_rate": 0.00011684228452602933, + "loss": 0.8938, + "step": 10036 + }, + { + "epoch": 1.7870370370370372, + "grad_norm": 0.6342145800590515, + "learning_rate": 0.00011682848684504448, + "loss": 0.8177, + "step": 10037 + }, + { + "epoch": 1.7872150997150997, + "grad_norm": 0.6609861254692078, + "learning_rate": 0.00011681468883434041, + "loss": 0.9692, + "step": 10038 + }, + { + "epoch": 1.7873931623931623, + "grad_norm": 0.7918622493743896, + "learning_rate": 0.00011680089049418743, + "loss": 0.8246, + "step": 10039 + }, + { + "epoch": 1.787571225071225, + "grad_norm": 0.697712779045105, + "learning_rate": 0.00011678709182485592, + "loss": 0.8981, + "step": 10040 + }, + { + "epoch": 1.7877492877492878, + "grad_norm": 0.6747658252716064, + "learning_rate": 0.00011677329282661617, + "loss": 1.1243, + "step": 10041 + }, + { + "epoch": 1.7879273504273505, + "grad_norm": 0.6525771617889404, + "learning_rate": 0.00011675949349973863, + "loss": 0.852, + "step": 10042 + }, + { + "epoch": 1.7881054131054133, + "grad_norm": 0.7062464952468872, + "learning_rate": 0.00011674569384449363, + "loss": 1.2582, + "step": 10043 + }, + { + "epoch": 1.7882834757834758, + "grad_norm": 0.6453786492347717, + "learning_rate": 0.00011673189386115154, + "loss": 0.868, + "step": 10044 + }, + { + "epoch": 1.7884615384615383, + "grad_norm": 0.7939708232879639, + "learning_rate": 0.00011671809354998273, + "loss": 0.7553, + "step": 10045 + }, + { + "epoch": 1.788639601139601, + "grad_norm": 0.6466066837310791, + "learning_rate": 0.00011670429291125761, + "loss": 0.942, + "step": 10046 + }, + { + "epoch": 1.7888176638176638, + "grad_norm": 0.7380510568618774, + "learning_rate": 0.00011669049194524657, + "loss": 1.044, + "step": 10047 + }, + { + "epoch": 1.7889957264957266, + "grad_norm": 0.6719707250595093, + "learning_rate": 0.00011667669065222002, + "loss": 1.1624, + "step": 10048 + }, + { + "epoch": 1.7891737891737893, + "grad_norm": 0.6996603012084961, + "learning_rate": 0.00011666288903244837, + "loss": 1.001, + "step": 10049 + }, + { + "epoch": 1.7893518518518519, + "grad_norm": 0.696590006351471, + "learning_rate": 0.00011664908708620202, + "loss": 1.17, + "step": 10050 + }, + { + "epoch": 1.7895299145299144, + "grad_norm": 0.7226764559745789, + "learning_rate": 0.00011663528481375137, + "loss": 1.0762, + "step": 10051 + }, + { + "epoch": 1.7897079772079771, + "grad_norm": 0.6117866635322571, + "learning_rate": 0.00011662148221536689, + "loss": 0.9199, + "step": 10052 + }, + { + "epoch": 1.78988603988604, + "grad_norm": 0.6424985527992249, + "learning_rate": 0.000116607679291319, + "loss": 1.1672, + "step": 10053 + }, + { + "epoch": 1.7900641025641026, + "grad_norm": 0.6390290856361389, + "learning_rate": 0.00011659387604187813, + "loss": 1.1895, + "step": 10054 + }, + { + "epoch": 1.7902421652421654, + "grad_norm": 0.6553205251693726, + "learning_rate": 0.00011658007246731473, + "loss": 1.0967, + "step": 10055 + }, + { + "epoch": 1.790420227920228, + "grad_norm": 0.7737570405006409, + "learning_rate": 0.00011656626856789922, + "loss": 0.9637, + "step": 10056 + }, + { + "epoch": 1.7905982905982905, + "grad_norm": 0.644296407699585, + "learning_rate": 0.00011655246434390212, + "loss": 0.9933, + "step": 10057 + }, + { + "epoch": 1.7907763532763532, + "grad_norm": 0.8154410123825073, + "learning_rate": 0.00011653865979559388, + "loss": 0.9623, + "step": 10058 + }, + { + "epoch": 1.790954415954416, + "grad_norm": 0.7181384563446045, + "learning_rate": 0.00011652485492324495, + "loss": 0.9113, + "step": 10059 + }, + { + "epoch": 1.7911324786324787, + "grad_norm": 0.7835097908973694, + "learning_rate": 0.00011651104972712582, + "loss": 1.0804, + "step": 10060 + }, + { + "epoch": 1.7913105413105415, + "grad_norm": 0.6843693852424622, + "learning_rate": 0.00011649724420750691, + "loss": 1.0242, + "step": 10061 + }, + { + "epoch": 1.791488603988604, + "grad_norm": 0.8364703059196472, + "learning_rate": 0.00011648343836465885, + "loss": 0.8445, + "step": 10062 + }, + { + "epoch": 1.7916666666666665, + "grad_norm": 0.7122092843055725, + "learning_rate": 0.00011646963219885201, + "loss": 1.0453, + "step": 10063 + }, + { + "epoch": 1.7918447293447293, + "grad_norm": 0.7018755078315735, + "learning_rate": 0.00011645582571035696, + "loss": 0.9753, + "step": 10064 + }, + { + "epoch": 1.792022792022792, + "grad_norm": 0.6522594094276428, + "learning_rate": 0.00011644201889944419, + "loss": 1.0328, + "step": 10065 + }, + { + "epoch": 1.7922008547008548, + "grad_norm": 0.70301353931427, + "learning_rate": 0.00011642821176638419, + "loss": 0.9143, + "step": 10066 + }, + { + "epoch": 1.7923789173789175, + "grad_norm": 0.6255469918251038, + "learning_rate": 0.0001164144043114475, + "loss": 0.9527, + "step": 10067 + }, + { + "epoch": 1.79255698005698, + "grad_norm": 0.6780602931976318, + "learning_rate": 0.0001164005965349047, + "loss": 0.9192, + "step": 10068 + }, + { + "epoch": 1.7927350427350426, + "grad_norm": 0.6025984287261963, + "learning_rate": 0.00011638678843702626, + "loss": 0.9055, + "step": 10069 + }, + { + "epoch": 1.7929131054131053, + "grad_norm": 0.6430829763412476, + "learning_rate": 0.00011637298001808275, + "loss": 0.9359, + "step": 10070 + }, + { + "epoch": 1.793091168091168, + "grad_norm": 0.6388106942176819, + "learning_rate": 0.0001163591712783447, + "loss": 0.8847, + "step": 10071 + }, + { + "epoch": 1.7932692307692308, + "grad_norm": 0.706347644329071, + "learning_rate": 0.00011634536221808265, + "loss": 0.9055, + "step": 10072 + }, + { + "epoch": 1.7934472934472936, + "grad_norm": 0.661226749420166, + "learning_rate": 0.00011633155283756721, + "loss": 1.118, + "step": 10073 + }, + { + "epoch": 1.7936253561253561, + "grad_norm": 0.543207049369812, + "learning_rate": 0.00011631774313706891, + "loss": 0.8856, + "step": 10074 + }, + { + "epoch": 1.7938034188034186, + "grad_norm": 0.6514154672622681, + "learning_rate": 0.00011630393311685835, + "loss": 0.8967, + "step": 10075 + }, + { + "epoch": 1.7939814814814814, + "grad_norm": 0.8669198155403137, + "learning_rate": 0.00011629012277720607, + "loss": 1.0362, + "step": 10076 + }, + { + "epoch": 1.7941595441595442, + "grad_norm": 0.7256068587303162, + "learning_rate": 0.00011627631211838266, + "loss": 1.1948, + "step": 10077 + }, + { + "epoch": 1.794337606837607, + "grad_norm": 0.6504935622215271, + "learning_rate": 0.00011626250114065875, + "loss": 0.8309, + "step": 10078 + }, + { + "epoch": 1.7945156695156697, + "grad_norm": 0.6964160799980164, + "learning_rate": 0.0001162486898443049, + "loss": 0.9593, + "step": 10079 + }, + { + "epoch": 1.7946937321937322, + "grad_norm": 0.668727695941925, + "learning_rate": 0.00011623487822959174, + "loss": 0.8897, + "step": 10080 + }, + { + "epoch": 1.7948717948717947, + "grad_norm": 0.6907223463058472, + "learning_rate": 0.00011622106629678986, + "loss": 0.897, + "step": 10081 + }, + { + "epoch": 1.7950498575498575, + "grad_norm": 0.6652865409851074, + "learning_rate": 0.00011620725404616985, + "loss": 0.9321, + "step": 10082 + }, + { + "epoch": 1.7952279202279202, + "grad_norm": 0.6523811221122742, + "learning_rate": 0.00011619344147800239, + "loss": 0.8991, + "step": 10083 + }, + { + "epoch": 1.795405982905983, + "grad_norm": 0.6162952184677124, + "learning_rate": 0.0001161796285925581, + "loss": 0.8061, + "step": 10084 + }, + { + "epoch": 1.7955840455840457, + "grad_norm": 0.670606791973114, + "learning_rate": 0.0001161658153901076, + "loss": 0.9341, + "step": 10085 + }, + { + "epoch": 1.7957621082621082, + "grad_norm": 0.6372489333152771, + "learning_rate": 0.00011615200187092148, + "loss": 1.1049, + "step": 10086 + }, + { + "epoch": 1.7959401709401708, + "grad_norm": 0.7311037182807922, + "learning_rate": 0.00011613818803527045, + "loss": 1.0881, + "step": 10087 + }, + { + "epoch": 1.7961182336182335, + "grad_norm": 0.7440751194953918, + "learning_rate": 0.00011612437388342518, + "loss": 0.9487, + "step": 10088 + }, + { + "epoch": 1.7962962962962963, + "grad_norm": 0.6605934500694275, + "learning_rate": 0.00011611055941565629, + "loss": 0.8757, + "step": 10089 + }, + { + "epoch": 1.796474358974359, + "grad_norm": 0.7546001076698303, + "learning_rate": 0.00011609674463223446, + "loss": 0.9368, + "step": 10090 + }, + { + "epoch": 1.7966524216524218, + "grad_norm": 0.7001389861106873, + "learning_rate": 0.00011608292953343036, + "loss": 0.9098, + "step": 10091 + }, + { + "epoch": 1.7968304843304843, + "grad_norm": 0.6898102760314941, + "learning_rate": 0.00011606911411951462, + "loss": 0.8821, + "step": 10092 + }, + { + "epoch": 1.797008547008547, + "grad_norm": 0.7020773887634277, + "learning_rate": 0.00011605529839075801, + "loss": 1.2775, + "step": 10093 + }, + { + "epoch": 1.7971866096866096, + "grad_norm": 0.6061446070671082, + "learning_rate": 0.0001160414823474312, + "loss": 1.0156, + "step": 10094 + }, + { + "epoch": 1.7973646723646723, + "grad_norm": 0.6746069192886353, + "learning_rate": 0.00011602766598980484, + "loss": 0.8223, + "step": 10095 + }, + { + "epoch": 1.797542735042735, + "grad_norm": 0.655829131603241, + "learning_rate": 0.00011601384931814967, + "loss": 0.9482, + "step": 10096 + }, + { + "epoch": 1.7977207977207978, + "grad_norm": 0.6762703061103821, + "learning_rate": 0.00011600003233273636, + "loss": 1.0191, + "step": 10097 + }, + { + "epoch": 1.7978988603988604, + "grad_norm": 0.7610527276992798, + "learning_rate": 0.00011598621503383566, + "loss": 1.0771, + "step": 10098 + }, + { + "epoch": 1.7980769230769231, + "grad_norm": 0.6857240200042725, + "learning_rate": 0.0001159723974217183, + "loss": 0.8325, + "step": 10099 + }, + { + "epoch": 1.7982549857549857, + "grad_norm": 0.6897954940795898, + "learning_rate": 0.00011595857949665501, + "loss": 1.0064, + "step": 10100 + }, + { + "epoch": 1.7984330484330484, + "grad_norm": 0.7023211717605591, + "learning_rate": 0.00011594476125891649, + "loss": 1.1346, + "step": 10101 + }, + { + "epoch": 1.7986111111111112, + "grad_norm": 0.8131003975868225, + "learning_rate": 0.00011593094270877347, + "loss": 1.0384, + "step": 10102 + }, + { + "epoch": 1.798789173789174, + "grad_norm": 0.6504445672035217, + "learning_rate": 0.00011591712384649676, + "loss": 0.8172, + "step": 10103 + }, + { + "epoch": 1.7989672364672364, + "grad_norm": 0.7379748821258545, + "learning_rate": 0.00011590330467235704, + "loss": 1.0118, + "step": 10104 + }, + { + "epoch": 1.7991452991452992, + "grad_norm": 0.8867329955101013, + "learning_rate": 0.0001158894851866251, + "loss": 1.023, + "step": 10105 + }, + { + "epoch": 1.7993233618233617, + "grad_norm": 0.7057412266731262, + "learning_rate": 0.00011587566538957173, + "loss": 0.8415, + "step": 10106 + }, + { + "epoch": 1.7995014245014245, + "grad_norm": 0.7479654550552368, + "learning_rate": 0.00011586184528146769, + "loss": 0.9663, + "step": 10107 + }, + { + "epoch": 1.7996794871794872, + "grad_norm": 0.6280845403671265, + "learning_rate": 0.00011584802486258368, + "loss": 0.973, + "step": 10108 + }, + { + "epoch": 1.79985754985755, + "grad_norm": 0.6735749840736389, + "learning_rate": 0.00011583420413319059, + "loss": 0.8631, + "step": 10109 + }, + { + "epoch": 1.8000356125356125, + "grad_norm": 0.5940406918525696, + "learning_rate": 0.00011582038309355918, + "loss": 0.8533, + "step": 10110 + }, + { + "epoch": 1.8002136752136753, + "grad_norm": 0.6923874020576477, + "learning_rate": 0.00011580656174396021, + "loss": 1.1105, + "step": 10111 + }, + { + "epoch": 1.8003917378917378, + "grad_norm": 0.6996715664863586, + "learning_rate": 0.00011579274008466447, + "loss": 0.9952, + "step": 10112 + }, + { + "epoch": 1.8005698005698005, + "grad_norm": 0.656561553478241, + "learning_rate": 0.00011577891811594281, + "loss": 0.9621, + "step": 10113 + }, + { + "epoch": 1.8007478632478633, + "grad_norm": 0.7121242880821228, + "learning_rate": 0.00011576509583806605, + "loss": 0.8658, + "step": 10114 + }, + { + "epoch": 1.800925925925926, + "grad_norm": 0.7864459753036499, + "learning_rate": 0.00011575127325130498, + "loss": 0.9867, + "step": 10115 + }, + { + "epoch": 1.8011039886039886, + "grad_norm": 0.6086452007293701, + "learning_rate": 0.00011573745035593042, + "loss": 0.8625, + "step": 10116 + }, + { + "epoch": 1.8012820512820513, + "grad_norm": 0.6553642749786377, + "learning_rate": 0.00011572362715221321, + "loss": 0.8475, + "step": 10117 + }, + { + "epoch": 1.8014601139601139, + "grad_norm": 0.6677348017692566, + "learning_rate": 0.00011570980364042419, + "loss": 0.9672, + "step": 10118 + }, + { + "epoch": 1.8016381766381766, + "grad_norm": 0.6275015473365784, + "learning_rate": 0.0001156959798208342, + "loss": 0.8663, + "step": 10119 + }, + { + "epoch": 1.8018162393162394, + "grad_norm": 0.787568211555481, + "learning_rate": 0.0001156821556937141, + "loss": 1.0188, + "step": 10120 + }, + { + "epoch": 1.801994301994302, + "grad_norm": 0.6983163356781006, + "learning_rate": 0.00011566833125933473, + "loss": 1.0767, + "step": 10121 + }, + { + "epoch": 1.8021723646723646, + "grad_norm": 0.7008936405181885, + "learning_rate": 0.00011565450651796695, + "loss": 1.0116, + "step": 10122 + }, + { + "epoch": 1.8023504273504274, + "grad_norm": 0.7694976925849915, + "learning_rate": 0.00011564068146988163, + "loss": 1.0227, + "step": 10123 + }, + { + "epoch": 1.80252849002849, + "grad_norm": 0.9530014991760254, + "learning_rate": 0.00011562685611534967, + "loss": 0.907, + "step": 10124 + }, + { + "epoch": 1.8027065527065527, + "grad_norm": 0.6714984178543091, + "learning_rate": 0.00011561303045464189, + "loss": 0.9501, + "step": 10125 + }, + { + "epoch": 1.8028846153846154, + "grad_norm": 0.7233797311782837, + "learning_rate": 0.00011559920448802925, + "loss": 1.021, + "step": 10126 + }, + { + "epoch": 1.8030626780626782, + "grad_norm": 0.7600540518760681, + "learning_rate": 0.0001155853782157826, + "loss": 1.1056, + "step": 10127 + }, + { + "epoch": 1.8032407407407407, + "grad_norm": 0.7836297750473022, + "learning_rate": 0.00011557155163817281, + "loss": 0.9906, + "step": 10128 + }, + { + "epoch": 1.8034188034188035, + "grad_norm": 0.7161104083061218, + "learning_rate": 0.00011555772475547084, + "loss": 0.9541, + "step": 10129 + }, + { + "epoch": 1.803596866096866, + "grad_norm": 0.6613732576370239, + "learning_rate": 0.00011554389756794757, + "loss": 0.9188, + "step": 10130 + }, + { + "epoch": 1.8037749287749287, + "grad_norm": 0.6415915489196777, + "learning_rate": 0.00011553007007587391, + "loss": 0.9928, + "step": 10131 + }, + { + "epoch": 1.8039529914529915, + "grad_norm": 0.7730516195297241, + "learning_rate": 0.0001155162422795208, + "loss": 1.0654, + "step": 10132 + }, + { + "epoch": 1.8041310541310542, + "grad_norm": 0.6769654750823975, + "learning_rate": 0.00011550241417915913, + "loss": 1.0678, + "step": 10133 + }, + { + "epoch": 1.8043091168091168, + "grad_norm": 0.6542425751686096, + "learning_rate": 0.00011548858577505988, + "loss": 0.9796, + "step": 10134 + }, + { + "epoch": 1.8044871794871795, + "grad_norm": 0.7282404899597168, + "learning_rate": 0.00011547475706749395, + "loss": 1.0314, + "step": 10135 + }, + { + "epoch": 1.804665242165242, + "grad_norm": 0.6450245976448059, + "learning_rate": 0.00011546092805673232, + "loss": 0.9564, + "step": 10136 + }, + { + "epoch": 1.8048433048433048, + "grad_norm": 0.65577632188797, + "learning_rate": 0.0001154470987430459, + "loss": 1.0219, + "step": 10137 + }, + { + "epoch": 1.8050213675213675, + "grad_norm": 0.7151737809181213, + "learning_rate": 0.00011543326912670567, + "loss": 0.9245, + "step": 10138 + }, + { + "epoch": 1.8051994301994303, + "grad_norm": 0.6695905327796936, + "learning_rate": 0.00011541943920798259, + "loss": 0.9535, + "step": 10139 + }, + { + "epoch": 1.8053774928774928, + "grad_norm": 0.7443813681602478, + "learning_rate": 0.00011540560898714767, + "loss": 1.1697, + "step": 10140 + }, + { + "epoch": 1.8055555555555556, + "grad_norm": 0.5701992511749268, + "learning_rate": 0.0001153917784644718, + "loss": 0.7868, + "step": 10141 + }, + { + "epoch": 1.805733618233618, + "grad_norm": 0.6992354989051819, + "learning_rate": 0.00011537794764022605, + "loss": 0.9856, + "step": 10142 + }, + { + "epoch": 1.8059116809116809, + "grad_norm": 0.6354477405548096, + "learning_rate": 0.00011536411651468131, + "loss": 0.8752, + "step": 10143 + }, + { + "epoch": 1.8060897435897436, + "grad_norm": 0.6952932476997375, + "learning_rate": 0.00011535028508810864, + "loss": 0.9446, + "step": 10144 + }, + { + "epoch": 1.8062678062678064, + "grad_norm": 0.5527541637420654, + "learning_rate": 0.00011533645336077901, + "loss": 0.5486, + "step": 10145 + }, + { + "epoch": 1.806445868945869, + "grad_norm": 0.685046374797821, + "learning_rate": 0.00011532262133296345, + "loss": 0.9529, + "step": 10146 + }, + { + "epoch": 1.8066239316239316, + "grad_norm": 0.6927558779716492, + "learning_rate": 0.00011530878900493296, + "loss": 1.1758, + "step": 10147 + }, + { + "epoch": 1.8068019943019942, + "grad_norm": 0.6758309602737427, + "learning_rate": 0.00011529495637695855, + "loss": 1.0076, + "step": 10148 + }, + { + "epoch": 1.806980056980057, + "grad_norm": 0.6739441156387329, + "learning_rate": 0.00011528112344931121, + "loss": 1.1914, + "step": 10149 + }, + { + "epoch": 1.8071581196581197, + "grad_norm": 0.7031944394111633, + "learning_rate": 0.00011526729022226204, + "loss": 0.783, + "step": 10150 + }, + { + "epoch": 1.8073361823361824, + "grad_norm": 0.6476930975914001, + "learning_rate": 0.00011525345669608202, + "loss": 0.9595, + "step": 10151 + }, + { + "epoch": 1.8075142450142452, + "grad_norm": 0.710498571395874, + "learning_rate": 0.00011523962287104222, + "loss": 0.8821, + "step": 10152 + }, + { + "epoch": 1.8076923076923077, + "grad_norm": 0.6664412617683411, + "learning_rate": 0.00011522578874741365, + "loss": 1.0182, + "step": 10153 + }, + { + "epoch": 1.8078703703703702, + "grad_norm": 0.8374263048171997, + "learning_rate": 0.00011521195432546737, + "loss": 0.9394, + "step": 10154 + }, + { + "epoch": 1.808048433048433, + "grad_norm": 0.6770764589309692, + "learning_rate": 0.00011519811960547447, + "loss": 1.0568, + "step": 10155 + }, + { + "epoch": 1.8082264957264957, + "grad_norm": 0.7014045715332031, + "learning_rate": 0.00011518428458770595, + "loss": 1.1705, + "step": 10156 + }, + { + "epoch": 1.8084045584045585, + "grad_norm": 0.6590061187744141, + "learning_rate": 0.00011517044927243295, + "loss": 1.1233, + "step": 10157 + }, + { + "epoch": 1.8085826210826212, + "grad_norm": 0.6093801856040955, + "learning_rate": 0.00011515661365992647, + "loss": 0.953, + "step": 10158 + }, + { + "epoch": 1.8087606837606838, + "grad_norm": 0.6197089552879333, + "learning_rate": 0.00011514277775045768, + "loss": 0.9414, + "step": 10159 + }, + { + "epoch": 1.8089387464387463, + "grad_norm": 0.7530463337898254, + "learning_rate": 0.00011512894154429759, + "loss": 0.9168, + "step": 10160 + }, + { + "epoch": 1.809116809116809, + "grad_norm": 0.6051347851753235, + "learning_rate": 0.00011511510504171735, + "loss": 0.9132, + "step": 10161 + }, + { + "epoch": 1.8092948717948718, + "grad_norm": 0.6388311982154846, + "learning_rate": 0.000115101268242988, + "loss": 0.6551, + "step": 10162 + }, + { + "epoch": 1.8094729344729346, + "grad_norm": 0.7040972709655762, + "learning_rate": 0.00011508743114838063, + "loss": 0.9409, + "step": 10163 + }, + { + "epoch": 1.8096509971509973, + "grad_norm": 0.7669548392295837, + "learning_rate": 0.00011507359375816644, + "loss": 1.0376, + "step": 10164 + }, + { + "epoch": 1.8098290598290598, + "grad_norm": 0.7309662699699402, + "learning_rate": 0.00011505975607261646, + "loss": 0.9071, + "step": 10165 + }, + { + "epoch": 1.8100071225071224, + "grad_norm": 0.6624547839164734, + "learning_rate": 0.00011504591809200187, + "loss": 1.0765, + "step": 10166 + }, + { + "epoch": 1.8101851851851851, + "grad_norm": 0.7719045281410217, + "learning_rate": 0.00011503207981659376, + "loss": 0.9244, + "step": 10167 + }, + { + "epoch": 1.8103632478632479, + "grad_norm": 0.6701484322547913, + "learning_rate": 0.0001150182412466633, + "loss": 0.9475, + "step": 10168 + }, + { + "epoch": 1.8105413105413106, + "grad_norm": 0.5604981184005737, + "learning_rate": 0.00011500440238248154, + "loss": 0.6268, + "step": 10169 + }, + { + "epoch": 1.8107193732193734, + "grad_norm": 0.6736510992050171, + "learning_rate": 0.00011499056322431973, + "loss": 0.9088, + "step": 10170 + }, + { + "epoch": 1.810897435897436, + "grad_norm": 0.7428455948829651, + "learning_rate": 0.00011497672377244897, + "loss": 0.9298, + "step": 10171 + }, + { + "epoch": 1.8110754985754984, + "grad_norm": 0.6543142795562744, + "learning_rate": 0.00011496288402714042, + "loss": 0.8863, + "step": 10172 + }, + { + "epoch": 1.8112535612535612, + "grad_norm": 0.6809250712394714, + "learning_rate": 0.00011494904398866524, + "loss": 0.977, + "step": 10173 + }, + { + "epoch": 1.811431623931624, + "grad_norm": 0.8105120062828064, + "learning_rate": 0.00011493520365729456, + "loss": 1.2115, + "step": 10174 + }, + { + "epoch": 1.8116096866096867, + "grad_norm": 0.6985095143318176, + "learning_rate": 0.00011492136303329964, + "loss": 0.8233, + "step": 10175 + }, + { + "epoch": 1.8117877492877494, + "grad_norm": 0.7198361754417419, + "learning_rate": 0.00011490752211695158, + "loss": 1.0552, + "step": 10176 + }, + { + "epoch": 1.811965811965812, + "grad_norm": 0.7077036499977112, + "learning_rate": 0.0001148936809085216, + "loss": 0.9171, + "step": 10177 + }, + { + "epoch": 1.8121438746438745, + "grad_norm": 0.9362925887107849, + "learning_rate": 0.00011487983940828089, + "loss": 0.9042, + "step": 10178 + }, + { + "epoch": 1.8123219373219372, + "grad_norm": 0.6732819676399231, + "learning_rate": 0.0001148659976165006, + "loss": 1.1033, + "step": 10179 + }, + { + "epoch": 1.8125, + "grad_norm": 0.747702419757843, + "learning_rate": 0.00011485215553345201, + "loss": 1.0692, + "step": 10180 + }, + { + "epoch": 1.8126780626780628, + "grad_norm": 0.7011259198188782, + "learning_rate": 0.00011483831315940627, + "loss": 0.9278, + "step": 10181 + }, + { + "epoch": 1.8128561253561255, + "grad_norm": 0.8542702198028564, + "learning_rate": 0.00011482447049463462, + "loss": 0.9476, + "step": 10182 + }, + { + "epoch": 1.813034188034188, + "grad_norm": 0.6975166201591492, + "learning_rate": 0.00011481062753940825, + "loss": 0.9486, + "step": 10183 + }, + { + "epoch": 1.8132122507122506, + "grad_norm": 0.8239036798477173, + "learning_rate": 0.0001147967842939984, + "loss": 1.0518, + "step": 10184 + }, + { + "epoch": 1.8133903133903133, + "grad_norm": 0.7559717297554016, + "learning_rate": 0.00011478294075867628, + "loss": 1.1877, + "step": 10185 + }, + { + "epoch": 1.813568376068376, + "grad_norm": 0.6755532026290894, + "learning_rate": 0.00011476909693371318, + "loss": 0.9287, + "step": 10186 + }, + { + "epoch": 1.8137464387464388, + "grad_norm": 0.6561332941055298, + "learning_rate": 0.0001147552528193803, + "loss": 0.83, + "step": 10187 + }, + { + "epoch": 1.8139245014245016, + "grad_norm": 0.7223508954048157, + "learning_rate": 0.00011474140841594887, + "loss": 1.1259, + "step": 10188 + }, + { + "epoch": 1.814102564102564, + "grad_norm": 0.7920593023300171, + "learning_rate": 0.0001147275637236902, + "loss": 1.0925, + "step": 10189 + }, + { + "epoch": 1.8142806267806266, + "grad_norm": 0.6896616816520691, + "learning_rate": 0.00011471371874287546, + "loss": 1.0204, + "step": 10190 + }, + { + "epoch": 1.8144586894586894, + "grad_norm": 0.6149865388870239, + "learning_rate": 0.00011469987347377602, + "loss": 1.1249, + "step": 10191 + }, + { + "epoch": 1.8146367521367521, + "grad_norm": 0.6650002598762512, + "learning_rate": 0.00011468602791666307, + "loss": 0.9723, + "step": 10192 + }, + { + "epoch": 1.8148148148148149, + "grad_norm": 0.7298738956451416, + "learning_rate": 0.00011467218207180792, + "loss": 1.0225, + "step": 10193 + }, + { + "epoch": 1.8149928774928776, + "grad_norm": 0.8075628876686096, + "learning_rate": 0.00011465833593948183, + "loss": 1.0429, + "step": 10194 + }, + { + "epoch": 1.8151709401709402, + "grad_norm": 0.8196593523025513, + "learning_rate": 0.0001146444895199561, + "loss": 0.9148, + "step": 10195 + }, + { + "epoch": 1.8153490028490027, + "grad_norm": 0.6394698023796082, + "learning_rate": 0.00011463064281350204, + "loss": 0.9781, + "step": 10196 + }, + { + "epoch": 1.8155270655270654, + "grad_norm": 0.7302836775779724, + "learning_rate": 0.00011461679582039091, + "loss": 1.0394, + "step": 10197 + }, + { + "epoch": 1.8157051282051282, + "grad_norm": 0.7066670060157776, + "learning_rate": 0.00011460294854089404, + "loss": 1.1153, + "step": 10198 + }, + { + "epoch": 1.815883190883191, + "grad_norm": 0.6471068263053894, + "learning_rate": 0.0001145891009752827, + "loss": 1.1533, + "step": 10199 + }, + { + "epoch": 1.8160612535612537, + "grad_norm": 0.6842355132102966, + "learning_rate": 0.00011457525312382826, + "loss": 0.953, + "step": 10200 + }, + { + "epoch": 1.8162393162393162, + "grad_norm": 0.6720319986343384, + "learning_rate": 0.00011456140498680202, + "loss": 1.003, + "step": 10201 + }, + { + "epoch": 1.8164173789173788, + "grad_norm": 0.632017970085144, + "learning_rate": 0.00011454755656447527, + "loss": 0.8148, + "step": 10202 + }, + { + "epoch": 1.8165954415954415, + "grad_norm": 0.7193828225135803, + "learning_rate": 0.00011453370785711939, + "loss": 1.0098, + "step": 10203 + }, + { + "epoch": 1.8167735042735043, + "grad_norm": 0.7098045349121094, + "learning_rate": 0.00011451985886500566, + "loss": 1.1276, + "step": 10204 + }, + { + "epoch": 1.816951566951567, + "grad_norm": 0.7076733708381653, + "learning_rate": 0.00011450600958840547, + "loss": 1.1216, + "step": 10205 + }, + { + "epoch": 1.8171296296296298, + "grad_norm": 0.6864610314369202, + "learning_rate": 0.00011449216002759018, + "loss": 0.9896, + "step": 10206 + }, + { + "epoch": 1.8173076923076923, + "grad_norm": 0.737727701663971, + "learning_rate": 0.0001144783101828311, + "loss": 0.9447, + "step": 10207 + }, + { + "epoch": 1.8174857549857548, + "grad_norm": 0.6562525033950806, + "learning_rate": 0.00011446446005439964, + "loss": 1.1208, + "step": 10208 + }, + { + "epoch": 1.8176638176638176, + "grad_norm": 0.7203826308250427, + "learning_rate": 0.0001144506096425671, + "loss": 1.1339, + "step": 10209 + }, + { + "epoch": 1.8178418803418803, + "grad_norm": 0.6657233834266663, + "learning_rate": 0.00011443675894760489, + "loss": 0.8307, + "step": 10210 + }, + { + "epoch": 1.818019943019943, + "grad_norm": 0.7032586932182312, + "learning_rate": 0.00011442290796978437, + "loss": 0.8546, + "step": 10211 + }, + { + "epoch": 1.8181980056980058, + "grad_norm": 0.6989460587501526, + "learning_rate": 0.00011440905670937696, + "loss": 1.0749, + "step": 10212 + }, + { + "epoch": 1.8183760683760684, + "grad_norm": 0.6461085677146912, + "learning_rate": 0.00011439520516665399, + "loss": 0.984, + "step": 10213 + }, + { + "epoch": 1.818554131054131, + "grad_norm": 0.7077372670173645, + "learning_rate": 0.00011438135334188689, + "loss": 1.0813, + "step": 10214 + }, + { + "epoch": 1.8187321937321936, + "grad_norm": 0.6724075675010681, + "learning_rate": 0.00011436750123534704, + "loss": 0.9975, + "step": 10215 + }, + { + "epoch": 1.8189102564102564, + "grad_norm": 0.6205753684043884, + "learning_rate": 0.00011435364884730583, + "loss": 0.7414, + "step": 10216 + }, + { + "epoch": 1.8190883190883191, + "grad_norm": 0.6416093707084656, + "learning_rate": 0.00011433979617803472, + "loss": 1.0024, + "step": 10217 + }, + { + "epoch": 1.819266381766382, + "grad_norm": 0.7817183136940002, + "learning_rate": 0.00011432594322780508, + "loss": 1.0577, + "step": 10218 + }, + { + "epoch": 1.8194444444444444, + "grad_norm": 0.688220202922821, + "learning_rate": 0.00011431208999688835, + "loss": 1.0301, + "step": 10219 + }, + { + "epoch": 1.8196225071225072, + "grad_norm": 0.6464754343032837, + "learning_rate": 0.0001142982364855559, + "loss": 1.0608, + "step": 10220 + }, + { + "epoch": 1.8198005698005697, + "grad_norm": 0.6607306599617004, + "learning_rate": 0.00011428438269407926, + "loss": 1.1203, + "step": 10221 + }, + { + "epoch": 1.8199786324786325, + "grad_norm": 0.5779942870140076, + "learning_rate": 0.00011427052862272982, + "loss": 0.7895, + "step": 10222 + }, + { + "epoch": 1.8201566951566952, + "grad_norm": 0.7599068880081177, + "learning_rate": 0.000114256674271779, + "loss": 0.883, + "step": 10223 + }, + { + "epoch": 1.820334757834758, + "grad_norm": 0.6578865051269531, + "learning_rate": 0.00011424281964149824, + "loss": 1.101, + "step": 10224 + }, + { + "epoch": 1.8205128205128205, + "grad_norm": 0.7090746760368347, + "learning_rate": 0.00011422896473215905, + "loss": 0.9514, + "step": 10225 + }, + { + "epoch": 1.8206908831908832, + "grad_norm": 0.7537758946418762, + "learning_rate": 0.00011421510954403281, + "loss": 1.2193, + "step": 10226 + }, + { + "epoch": 1.8208689458689458, + "grad_norm": 0.670183002948761, + "learning_rate": 0.00011420125407739106, + "loss": 1.1408, + "step": 10227 + }, + { + "epoch": 1.8210470085470085, + "grad_norm": 0.742520809173584, + "learning_rate": 0.00011418739833250524, + "loss": 0.8826, + "step": 10228 + }, + { + "epoch": 1.8212250712250713, + "grad_norm": 0.6542800664901733, + "learning_rate": 0.00011417354230964683, + "loss": 1.0039, + "step": 10229 + }, + { + "epoch": 1.821403133903134, + "grad_norm": 0.6713709235191345, + "learning_rate": 0.00011415968600908727, + "loss": 0.9351, + "step": 10230 + }, + { + "epoch": 1.8215811965811965, + "grad_norm": 0.6794951558113098, + "learning_rate": 0.0001141458294310981, + "loss": 0.9491, + "step": 10231 + }, + { + "epoch": 1.8217592592592593, + "grad_norm": 0.6921972632408142, + "learning_rate": 0.00011413197257595079, + "loss": 1.1342, + "step": 10232 + }, + { + "epoch": 1.8219373219373218, + "grad_norm": 0.702586829662323, + "learning_rate": 0.00011411811544391682, + "loss": 0.9992, + "step": 10233 + }, + { + "epoch": 1.8221153846153846, + "grad_norm": 0.8147975206375122, + "learning_rate": 0.00011410425803526772, + "loss": 1.0507, + "step": 10234 + }, + { + "epoch": 1.8222934472934473, + "grad_norm": 0.66419517993927, + "learning_rate": 0.00011409040035027496, + "loss": 1.0426, + "step": 10235 + }, + { + "epoch": 1.82247150997151, + "grad_norm": 0.6132485866546631, + "learning_rate": 0.00011407654238921011, + "loss": 0.9859, + "step": 10236 + }, + { + "epoch": 1.8226495726495726, + "grad_norm": 0.7522366046905518, + "learning_rate": 0.00011406268415234462, + "loss": 0.9379, + "step": 10237 + }, + { + "epoch": 1.8228276353276354, + "grad_norm": 0.6335554122924805, + "learning_rate": 0.00011404882563995007, + "loss": 0.9322, + "step": 10238 + }, + { + "epoch": 1.823005698005698, + "grad_norm": 0.7577497363090515, + "learning_rate": 0.00011403496685229797, + "loss": 1.1383, + "step": 10239 + }, + { + "epoch": 1.8231837606837606, + "grad_norm": 0.6796886920928955, + "learning_rate": 0.00011402110778965982, + "loss": 1.0092, + "step": 10240 + }, + { + "epoch": 1.8233618233618234, + "grad_norm": 0.7676617503166199, + "learning_rate": 0.0001140072484523072, + "loss": 1.0137, + "step": 10241 + }, + { + "epoch": 1.8235398860398861, + "grad_norm": 0.7807821035385132, + "learning_rate": 0.00011399338884051165, + "loss": 0.8987, + "step": 10242 + }, + { + "epoch": 1.8237179487179487, + "grad_norm": 0.7169568538665771, + "learning_rate": 0.00011397952895454473, + "loss": 0.8984, + "step": 10243 + }, + { + "epoch": 1.8238960113960114, + "grad_norm": 0.6564654111862183, + "learning_rate": 0.00011396566879467793, + "loss": 1.0255, + "step": 10244 + }, + { + "epoch": 1.824074074074074, + "grad_norm": 0.7290034294128418, + "learning_rate": 0.00011395180836118292, + "loss": 0.9962, + "step": 10245 + }, + { + "epoch": 1.8242521367521367, + "grad_norm": 0.6610758900642395, + "learning_rate": 0.00011393794765433115, + "loss": 1.102, + "step": 10246 + }, + { + "epoch": 1.8244301994301995, + "grad_norm": 0.6875932216644287, + "learning_rate": 0.0001139240866743943, + "loss": 0.9963, + "step": 10247 + }, + { + "epoch": 1.8246082621082622, + "grad_norm": 0.7595645189285278, + "learning_rate": 0.00011391022542164387, + "loss": 1.1285, + "step": 10248 + }, + { + "epoch": 1.8247863247863247, + "grad_norm": 0.6752721667289734, + "learning_rate": 0.0001138963638963515, + "loss": 0.9447, + "step": 10249 + }, + { + "epoch": 1.8249643874643875, + "grad_norm": 0.6697955131530762, + "learning_rate": 0.00011388250209878873, + "loss": 1.0804, + "step": 10250 + }, + { + "epoch": 1.82514245014245, + "grad_norm": 0.6546956896781921, + "learning_rate": 0.00011386864002922713, + "loss": 0.9626, + "step": 10251 + }, + { + "epoch": 1.8253205128205128, + "grad_norm": 0.8002896904945374, + "learning_rate": 0.00011385477768793838, + "loss": 1.1933, + "step": 10252 + }, + { + "epoch": 1.8254985754985755, + "grad_norm": 0.6566781401634216, + "learning_rate": 0.00011384091507519403, + "loss": 0.9802, + "step": 10253 + }, + { + "epoch": 1.8256766381766383, + "grad_norm": 0.617420494556427, + "learning_rate": 0.00011382705219126572, + "loss": 1.1098, + "step": 10254 + }, + { + "epoch": 1.8258547008547008, + "grad_norm": 0.6558036208152771, + "learning_rate": 0.00011381318903642504, + "loss": 1.0291, + "step": 10255 + }, + { + "epoch": 1.8260327635327636, + "grad_norm": 0.6295637488365173, + "learning_rate": 0.00011379932561094358, + "loss": 1.0792, + "step": 10256 + }, + { + "epoch": 1.826210826210826, + "grad_norm": 0.7475154399871826, + "learning_rate": 0.00011378546191509303, + "loss": 1.1362, + "step": 10257 + }, + { + "epoch": 1.8263888888888888, + "grad_norm": 0.6814939379692078, + "learning_rate": 0.00011377159794914498, + "loss": 0.9131, + "step": 10258 + }, + { + "epoch": 1.8265669515669516, + "grad_norm": 0.6726876497268677, + "learning_rate": 0.00011375773371337111, + "loss": 0.9147, + "step": 10259 + }, + { + "epoch": 1.8267450142450143, + "grad_norm": 0.785943865776062, + "learning_rate": 0.00011374386920804298, + "loss": 1.0137, + "step": 10260 + }, + { + "epoch": 1.8269230769230769, + "grad_norm": 0.7614478468894958, + "learning_rate": 0.0001137300044334323, + "loss": 1.2118, + "step": 10261 + }, + { + "epoch": 1.8271011396011396, + "grad_norm": 0.7317564487457275, + "learning_rate": 0.00011371613938981072, + "loss": 1.0602, + "step": 10262 + }, + { + "epoch": 1.8272792022792022, + "grad_norm": 0.6716432571411133, + "learning_rate": 0.00011370227407744987, + "loss": 0.952, + "step": 10263 + }, + { + "epoch": 1.827457264957265, + "grad_norm": 0.6946425437927246, + "learning_rate": 0.00011368840849662139, + "loss": 1.0554, + "step": 10264 + }, + { + "epoch": 1.8276353276353277, + "grad_norm": 0.6692264080047607, + "learning_rate": 0.00011367454264759703, + "loss": 0.8944, + "step": 10265 + }, + { + "epoch": 1.8278133903133904, + "grad_norm": 0.6931505799293518, + "learning_rate": 0.00011366067653064838, + "loss": 0.9045, + "step": 10266 + }, + { + "epoch": 1.827991452991453, + "grad_norm": 0.7233194708824158, + "learning_rate": 0.00011364681014604716, + "loss": 0.9441, + "step": 10267 + }, + { + "epoch": 1.8281695156695157, + "grad_norm": 0.6451242566108704, + "learning_rate": 0.00011363294349406506, + "loss": 0.9948, + "step": 10268 + }, + { + "epoch": 1.8283475783475782, + "grad_norm": 0.6993351578712463, + "learning_rate": 0.00011361907657497375, + "loss": 1.1057, + "step": 10269 + }, + { + "epoch": 1.828525641025641, + "grad_norm": 0.7241137623786926, + "learning_rate": 0.00011360520938904493, + "loss": 0.974, + "step": 10270 + }, + { + "epoch": 1.8287037037037037, + "grad_norm": 0.6349480152130127, + "learning_rate": 0.00011359134193655027, + "loss": 0.9026, + "step": 10271 + }, + { + "epoch": 1.8288817663817665, + "grad_norm": 0.6916826963424683, + "learning_rate": 0.00011357747421776151, + "loss": 0.9153, + "step": 10272 + }, + { + "epoch": 1.8290598290598292, + "grad_norm": 0.879770040512085, + "learning_rate": 0.00011356360623295037, + "loss": 1.0818, + "step": 10273 + }, + { + "epoch": 1.8292378917378918, + "grad_norm": 0.6293807029724121, + "learning_rate": 0.00011354973798238853, + "loss": 1.1164, + "step": 10274 + }, + { + "epoch": 1.8294159544159543, + "grad_norm": 0.7070622444152832, + "learning_rate": 0.0001135358694663477, + "loss": 0.8795, + "step": 10275 + }, + { + "epoch": 1.829594017094017, + "grad_norm": 0.6847673654556274, + "learning_rate": 0.00011352200068509962, + "loss": 0.9173, + "step": 10276 + }, + { + "epoch": 1.8297720797720798, + "grad_norm": 0.6552146077156067, + "learning_rate": 0.00011350813163891605, + "loss": 1.0425, + "step": 10277 + }, + { + "epoch": 1.8299501424501425, + "grad_norm": 0.6432808041572571, + "learning_rate": 0.0001134942623280687, + "loss": 0.9418, + "step": 10278 + }, + { + "epoch": 1.8301282051282053, + "grad_norm": 0.7412393093109131, + "learning_rate": 0.00011348039275282931, + "loss": 1.1212, + "step": 10279 + }, + { + "epoch": 1.8303062678062678, + "grad_norm": 0.6543423533439636, + "learning_rate": 0.00011346652291346965, + "loss": 1.0553, + "step": 10280 + }, + { + "epoch": 1.8304843304843303, + "grad_norm": 0.7159286141395569, + "learning_rate": 0.00011345265281026138, + "loss": 1.0582, + "step": 10281 + }, + { + "epoch": 1.830662393162393, + "grad_norm": 0.6443323493003845, + "learning_rate": 0.00011343878244347639, + "loss": 0.9462, + "step": 10282 + }, + { + "epoch": 1.8308404558404558, + "grad_norm": 0.7592014074325562, + "learning_rate": 0.00011342491181338634, + "loss": 1.2718, + "step": 10283 + }, + { + "epoch": 1.8310185185185186, + "grad_norm": 0.627109944820404, + "learning_rate": 0.00011341104092026302, + "loss": 1.0177, + "step": 10284 + }, + { + "epoch": 1.8311965811965814, + "grad_norm": 0.8061598539352417, + "learning_rate": 0.00011339716976437827, + "loss": 0.9416, + "step": 10285 + }, + { + "epoch": 1.8313746438746439, + "grad_norm": 0.6584261059761047, + "learning_rate": 0.00011338329834600377, + "loss": 0.8297, + "step": 10286 + }, + { + "epoch": 1.8315527065527064, + "grad_norm": 0.6329470276832581, + "learning_rate": 0.00011336942666541133, + "loss": 0.8386, + "step": 10287 + }, + { + "epoch": 1.8317307692307692, + "grad_norm": 0.6833979487419128, + "learning_rate": 0.00011335555472287275, + "loss": 0.9407, + "step": 10288 + }, + { + "epoch": 1.831908831908832, + "grad_norm": 0.7663840651512146, + "learning_rate": 0.00011334168251865985, + "loss": 1.0018, + "step": 10289 + }, + { + "epoch": 1.8320868945868947, + "grad_norm": 0.7751262784004211, + "learning_rate": 0.00011332781005304436, + "loss": 1.0576, + "step": 10290 + }, + { + "epoch": 1.8322649572649574, + "grad_norm": 0.6857370138168335, + "learning_rate": 0.00011331393732629814, + "loss": 0.9888, + "step": 10291 + }, + { + "epoch": 1.83244301994302, + "grad_norm": 0.7534535527229309, + "learning_rate": 0.00011330006433869296, + "loss": 1.0834, + "step": 10292 + }, + { + "epoch": 1.8326210826210825, + "grad_norm": 0.6785250306129456, + "learning_rate": 0.00011328619109050065, + "loss": 1.0471, + "step": 10293 + }, + { + "epoch": 1.8327991452991452, + "grad_norm": 0.7023689150810242, + "learning_rate": 0.00011327231758199303, + "loss": 1.0652, + "step": 10294 + }, + { + "epoch": 1.832977207977208, + "grad_norm": 0.6776610612869263, + "learning_rate": 0.00011325844381344192, + "loss": 0.9504, + "step": 10295 + }, + { + "epoch": 1.8331552706552707, + "grad_norm": 0.7704112529754639, + "learning_rate": 0.00011324456978511917, + "loss": 0.9712, + "step": 10296 + }, + { + "epoch": 1.8333333333333335, + "grad_norm": 0.601502537727356, + "learning_rate": 0.00011323069549729654, + "loss": 1.075, + "step": 10297 + }, + { + "epoch": 1.833511396011396, + "grad_norm": 0.6282439231872559, + "learning_rate": 0.00011321682095024596, + "loss": 0.9238, + "step": 10298 + }, + { + "epoch": 1.8336894586894585, + "grad_norm": 0.6873499155044556, + "learning_rate": 0.00011320294614423921, + "loss": 1.0464, + "step": 10299 + }, + { + "epoch": 1.8338675213675213, + "grad_norm": 0.6063792705535889, + "learning_rate": 0.00011318907107954815, + "loss": 0.9732, + "step": 10300 + }, + { + "epoch": 1.834045584045584, + "grad_norm": 0.5830921530723572, + "learning_rate": 0.00011317519575644464, + "loss": 0.7568, + "step": 10301 + }, + { + "epoch": 1.8342236467236468, + "grad_norm": 0.6394222378730774, + "learning_rate": 0.00011316132017520053, + "loss": 0.9958, + "step": 10302 + }, + { + "epoch": 1.8344017094017095, + "grad_norm": 0.7052412033081055, + "learning_rate": 0.00011314744433608773, + "loss": 0.9129, + "step": 10303 + }, + { + "epoch": 1.834579772079772, + "grad_norm": 0.7287624478340149, + "learning_rate": 0.00011313356823937801, + "loss": 0.8608, + "step": 10304 + }, + { + "epoch": 1.8347578347578346, + "grad_norm": 0.702937662601471, + "learning_rate": 0.00011311969188534334, + "loss": 1.3074, + "step": 10305 + }, + { + "epoch": 1.8349358974358974, + "grad_norm": 0.6693850159645081, + "learning_rate": 0.00011310581527425557, + "loss": 0.928, + "step": 10306 + }, + { + "epoch": 1.83511396011396, + "grad_norm": 0.8153932094573975, + "learning_rate": 0.00011309193840638654, + "loss": 1.1771, + "step": 10307 + }, + { + "epoch": 1.8352920227920229, + "grad_norm": 0.6517418622970581, + "learning_rate": 0.00011307806128200821, + "loss": 0.9634, + "step": 10308 + }, + { + "epoch": 1.8354700854700856, + "grad_norm": 0.6626226305961609, + "learning_rate": 0.00011306418390139245, + "loss": 0.9371, + "step": 10309 + }, + { + "epoch": 1.8356481481481481, + "grad_norm": 0.7397477030754089, + "learning_rate": 0.0001130503062648111, + "loss": 0.9398, + "step": 10310 + }, + { + "epoch": 1.8358262108262107, + "grad_norm": 0.6790265440940857, + "learning_rate": 0.00011303642837253614, + "loss": 0.9728, + "step": 10311 + }, + { + "epoch": 1.8360042735042734, + "grad_norm": 0.6266449093818665, + "learning_rate": 0.00011302255022483941, + "loss": 0.847, + "step": 10312 + }, + { + "epoch": 1.8361823361823362, + "grad_norm": 0.791657030582428, + "learning_rate": 0.00011300867182199288, + "loss": 0.8342, + "step": 10313 + }, + { + "epoch": 1.836360398860399, + "grad_norm": 0.7128583788871765, + "learning_rate": 0.00011299479316426846, + "loss": 0.9591, + "step": 10314 + }, + { + "epoch": 1.8365384615384617, + "grad_norm": 0.659928023815155, + "learning_rate": 0.00011298091425193806, + "loss": 1.0282, + "step": 10315 + }, + { + "epoch": 1.8367165242165242, + "grad_norm": 0.6641396284103394, + "learning_rate": 0.00011296703508527363, + "loss": 1.0161, + "step": 10316 + }, + { + "epoch": 1.8368945868945867, + "grad_norm": 0.7921316027641296, + "learning_rate": 0.00011295315566454702, + "loss": 0.8897, + "step": 10317 + }, + { + "epoch": 1.8370726495726495, + "grad_norm": 0.6900694966316223, + "learning_rate": 0.00011293927599003029, + "loss": 1.0094, + "step": 10318 + }, + { + "epoch": 1.8372507122507122, + "grad_norm": 0.8054366707801819, + "learning_rate": 0.0001129253960619953, + "loss": 0.9489, + "step": 10319 + }, + { + "epoch": 1.837428774928775, + "grad_norm": 0.6623767018318176, + "learning_rate": 0.00011291151588071405, + "loss": 0.92, + "step": 10320 + }, + { + "epoch": 1.8376068376068377, + "grad_norm": 0.6143901348114014, + "learning_rate": 0.00011289763544645846, + "loss": 0.8093, + "step": 10321 + }, + { + "epoch": 1.8377849002849003, + "grad_norm": 0.8207027316093445, + "learning_rate": 0.00011288375475950046, + "loss": 1.2402, + "step": 10322 + }, + { + "epoch": 1.8379629629629628, + "grad_norm": 0.6759985685348511, + "learning_rate": 0.00011286987382011209, + "loss": 0.9179, + "step": 10323 + }, + { + "epoch": 1.8381410256410255, + "grad_norm": 0.745439887046814, + "learning_rate": 0.00011285599262856523, + "loss": 0.8157, + "step": 10324 + }, + { + "epoch": 1.8383190883190883, + "grad_norm": 0.6873317360877991, + "learning_rate": 0.00011284211118513194, + "loss": 0.8681, + "step": 10325 + }, + { + "epoch": 1.838497150997151, + "grad_norm": 0.7060160040855408, + "learning_rate": 0.00011282822949008416, + "loss": 1.0833, + "step": 10326 + }, + { + "epoch": 1.8386752136752138, + "grad_norm": 0.8079642653465271, + "learning_rate": 0.00011281434754369389, + "loss": 0.8639, + "step": 10327 + }, + { + "epoch": 1.8388532763532763, + "grad_norm": 0.6434001922607422, + "learning_rate": 0.00011280046534623303, + "loss": 0.9269, + "step": 10328 + }, + { + "epoch": 1.839031339031339, + "grad_norm": 0.7005292773246765, + "learning_rate": 0.0001127865828979737, + "loss": 1.1475, + "step": 10329 + }, + { + "epoch": 1.8392094017094016, + "grad_norm": 0.7004852890968323, + "learning_rate": 0.00011277270019918784, + "loss": 0.9467, + "step": 10330 + }, + { + "epoch": 1.8393874643874644, + "grad_norm": 0.7542549967765808, + "learning_rate": 0.00011275881725014743, + "loss": 1.0371, + "step": 10331 + }, + { + "epoch": 1.8395655270655271, + "grad_norm": 0.674051821231842, + "learning_rate": 0.00011274493405112452, + "loss": 1.1097, + "step": 10332 + }, + { + "epoch": 1.8397435897435899, + "grad_norm": 0.8136405348777771, + "learning_rate": 0.00011273105060239107, + "loss": 0.9718, + "step": 10333 + }, + { + "epoch": 1.8399216524216524, + "grad_norm": 0.6524073481559753, + "learning_rate": 0.00011271716690421916, + "loss": 0.9953, + "step": 10334 + }, + { + "epoch": 1.8400997150997151, + "grad_norm": 0.7436625957489014, + "learning_rate": 0.00011270328295688077, + "loss": 1.0722, + "step": 10335 + }, + { + "epoch": 1.8402777777777777, + "grad_norm": 0.6815723180770874, + "learning_rate": 0.00011268939876064795, + "loss": 1.0924, + "step": 10336 + }, + { + "epoch": 1.8404558404558404, + "grad_norm": 0.6923388242721558, + "learning_rate": 0.0001126755143157927, + "loss": 0.921, + "step": 10337 + }, + { + "epoch": 1.8406339031339032, + "grad_norm": 0.7464849948883057, + "learning_rate": 0.00011266162962258708, + "loss": 1.0549, + "step": 10338 + }, + { + "epoch": 1.840811965811966, + "grad_norm": 0.6621805429458618, + "learning_rate": 0.00011264774468130315, + "loss": 1.0764, + "step": 10339 + }, + { + "epoch": 1.8409900284900285, + "grad_norm": 0.7370132803916931, + "learning_rate": 0.00011263385949221295, + "loss": 0.7818, + "step": 10340 + }, + { + "epoch": 1.8411680911680912, + "grad_norm": 0.673100471496582, + "learning_rate": 0.00011261997405558848, + "loss": 1.04, + "step": 10341 + }, + { + "epoch": 1.8413461538461537, + "grad_norm": 0.5978201031684875, + "learning_rate": 0.00011260608837170183, + "loss": 0.9644, + "step": 10342 + }, + { + "epoch": 1.8415242165242165, + "grad_norm": 0.6868628263473511, + "learning_rate": 0.00011259220244082507, + "loss": 0.9533, + "step": 10343 + }, + { + "epoch": 1.8417022792022792, + "grad_norm": 0.6580314636230469, + "learning_rate": 0.0001125783162632303, + "loss": 0.9506, + "step": 10344 + }, + { + "epoch": 1.841880341880342, + "grad_norm": 0.7238291501998901, + "learning_rate": 0.00011256442983918951, + "loss": 0.8663, + "step": 10345 + }, + { + "epoch": 1.8420584045584045, + "grad_norm": 0.5838520526885986, + "learning_rate": 0.00011255054316897484, + "loss": 0.9606, + "step": 10346 + }, + { + "epoch": 1.8422364672364673, + "grad_norm": 0.7102842926979065, + "learning_rate": 0.00011253665625285836, + "loss": 0.801, + "step": 10347 + }, + { + "epoch": 1.8424145299145298, + "grad_norm": 0.6449147462844849, + "learning_rate": 0.0001125227690911121, + "loss": 1.0827, + "step": 10348 + }, + { + "epoch": 1.8425925925925926, + "grad_norm": 0.6355304718017578, + "learning_rate": 0.00011250888168400823, + "loss": 1.0369, + "step": 10349 + }, + { + "epoch": 1.8427706552706553, + "grad_norm": 0.678977906703949, + "learning_rate": 0.0001124949940318188, + "loss": 0.9491, + "step": 10350 + }, + { + "epoch": 1.842948717948718, + "grad_norm": 0.6366633772850037, + "learning_rate": 0.00011248110613481592, + "loss": 0.7272, + "step": 10351 + }, + { + "epoch": 1.8431267806267806, + "grad_norm": 0.6639098525047302, + "learning_rate": 0.00011246721799327171, + "loss": 1.0313, + "step": 10352 + }, + { + "epoch": 1.8433048433048433, + "grad_norm": 0.6034720540046692, + "learning_rate": 0.00011245332960745822, + "loss": 0.7141, + "step": 10353 + }, + { + "epoch": 1.8434829059829059, + "grad_norm": 0.8118346333503723, + "learning_rate": 0.00011243944097764763, + "loss": 1.171, + "step": 10354 + }, + { + "epoch": 1.8436609686609686, + "grad_norm": 0.6706618070602417, + "learning_rate": 0.00011242555210411203, + "loss": 0.9578, + "step": 10355 + }, + { + "epoch": 1.8438390313390314, + "grad_norm": 0.619562029838562, + "learning_rate": 0.00011241166298712355, + "loss": 0.9883, + "step": 10356 + }, + { + "epoch": 1.8440170940170941, + "grad_norm": 0.6471936106681824, + "learning_rate": 0.00011239777362695434, + "loss": 0.8897, + "step": 10357 + }, + { + "epoch": 1.8441951566951567, + "grad_norm": 0.7179005742073059, + "learning_rate": 0.00011238388402387645, + "loss": 0.9646, + "step": 10358 + }, + { + "epoch": 1.8443732193732194, + "grad_norm": 0.7726966738700867, + "learning_rate": 0.00011236999417816214, + "loss": 0.8855, + "step": 10359 + }, + { + "epoch": 1.844551282051282, + "grad_norm": 0.6733565330505371, + "learning_rate": 0.00011235610409008346, + "loss": 1.0379, + "step": 10360 + }, + { + "epoch": 1.8447293447293447, + "grad_norm": 0.7317814826965332, + "learning_rate": 0.0001123422137599126, + "loss": 0.8528, + "step": 10361 + }, + { + "epoch": 1.8449074074074074, + "grad_norm": 0.6727005839347839, + "learning_rate": 0.0001123283231879217, + "loss": 0.9612, + "step": 10362 + }, + { + "epoch": 1.8450854700854702, + "grad_norm": 0.6350542306900024, + "learning_rate": 0.00011231443237438289, + "loss": 0.9939, + "step": 10363 + }, + { + "epoch": 1.8452635327635327, + "grad_norm": 0.693148672580719, + "learning_rate": 0.00011230054131956836, + "loss": 1.0149, + "step": 10364 + }, + { + "epoch": 1.8454415954415955, + "grad_norm": 0.7263579368591309, + "learning_rate": 0.0001122866500237503, + "loss": 1.1044, + "step": 10365 + }, + { + "epoch": 1.845619658119658, + "grad_norm": 0.7044230699539185, + "learning_rate": 0.00011227275848720085, + "loss": 1.0677, + "step": 10366 + }, + { + "epoch": 1.8457977207977208, + "grad_norm": 0.6895326972007751, + "learning_rate": 0.00011225886671019219, + "loss": 1.1025, + "step": 10367 + }, + { + "epoch": 1.8459757834757835, + "grad_norm": 0.6045145988464355, + "learning_rate": 0.00011224497469299651, + "loss": 0.8079, + "step": 10368 + }, + { + "epoch": 1.8461538461538463, + "grad_norm": 0.6613210439682007, + "learning_rate": 0.00011223108243588599, + "loss": 1.0345, + "step": 10369 + }, + { + "epoch": 1.8463319088319088, + "grad_norm": 0.6288960576057434, + "learning_rate": 0.0001122171899391328, + "loss": 1.0166, + "step": 10370 + }, + { + "epoch": 1.8465099715099715, + "grad_norm": 0.6158748865127563, + "learning_rate": 0.00011220329720300917, + "loss": 0.895, + "step": 10371 + }, + { + "epoch": 1.846688034188034, + "grad_norm": 0.6583057641983032, + "learning_rate": 0.00011218940422778728, + "loss": 0.8059, + "step": 10372 + }, + { + "epoch": 1.8468660968660968, + "grad_norm": 0.6761550903320312, + "learning_rate": 0.00011217551101373932, + "loss": 0.9253, + "step": 10373 + }, + { + "epoch": 1.8470441595441596, + "grad_norm": 0.5969263315200806, + "learning_rate": 0.0001121616175611375, + "loss": 0.8549, + "step": 10374 + }, + { + "epoch": 1.8472222222222223, + "grad_norm": 0.7994722723960876, + "learning_rate": 0.00011214772387025407, + "loss": 0.9918, + "step": 10375 + }, + { + "epoch": 1.8474002849002849, + "grad_norm": 0.6949167847633362, + "learning_rate": 0.00011213382994136123, + "loss": 1.1853, + "step": 10376 + }, + { + "epoch": 1.8475783475783476, + "grad_norm": 0.7356176376342773, + "learning_rate": 0.00011211993577473121, + "loss": 0.8809, + "step": 10377 + }, + { + "epoch": 1.8477564102564101, + "grad_norm": 0.7110268473625183, + "learning_rate": 0.0001121060413706362, + "loss": 0.9805, + "step": 10378 + }, + { + "epoch": 1.8479344729344729, + "grad_norm": 0.6509962677955627, + "learning_rate": 0.00011209214672934846, + "loss": 0.8899, + "step": 10379 + }, + { + "epoch": 1.8481125356125356, + "grad_norm": 0.6103082299232483, + "learning_rate": 0.00011207825185114025, + "loss": 0.8576, + "step": 10380 + }, + { + "epoch": 1.8482905982905984, + "grad_norm": 0.6261070966720581, + "learning_rate": 0.00011206435673628377, + "loss": 0.8884, + "step": 10381 + }, + { + "epoch": 1.848468660968661, + "grad_norm": 0.7629222273826599, + "learning_rate": 0.00011205046138505126, + "loss": 1.1714, + "step": 10382 + }, + { + "epoch": 1.8486467236467237, + "grad_norm": 0.617957353591919, + "learning_rate": 0.000112036565797715, + "loss": 0.9546, + "step": 10383 + }, + { + "epoch": 1.8488247863247862, + "grad_norm": 0.6926987171173096, + "learning_rate": 0.00011202266997454724, + "loss": 0.8842, + "step": 10384 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 0.602758526802063, + "learning_rate": 0.00011200877391582025, + "loss": 0.9782, + "step": 10385 + }, + { + "epoch": 1.8491809116809117, + "grad_norm": 0.706731915473938, + "learning_rate": 0.00011199487762180627, + "loss": 0.8176, + "step": 10386 + }, + { + "epoch": 1.8493589743589745, + "grad_norm": 0.7135118842124939, + "learning_rate": 0.0001119809810927776, + "loss": 0.9277, + "step": 10387 + }, + { + "epoch": 1.8495370370370372, + "grad_norm": 0.7484592199325562, + "learning_rate": 0.00011196708432900647, + "loss": 1.0733, + "step": 10388 + }, + { + "epoch": 1.8497150997150997, + "grad_norm": 0.7087157964706421, + "learning_rate": 0.00011195318733076519, + "loss": 0.9443, + "step": 10389 + }, + { + "epoch": 1.8498931623931623, + "grad_norm": 0.6511468291282654, + "learning_rate": 0.00011193929009832602, + "loss": 0.955, + "step": 10390 + }, + { + "epoch": 1.850071225071225, + "grad_norm": 0.6386628746986389, + "learning_rate": 0.0001119253926319613, + "loss": 1.0357, + "step": 10391 + }, + { + "epoch": 1.8502492877492878, + "grad_norm": 0.6400021314620972, + "learning_rate": 0.00011191149493194327, + "loss": 0.8094, + "step": 10392 + }, + { + "epoch": 1.8504273504273505, + "grad_norm": 0.7942537069320679, + "learning_rate": 0.00011189759699854423, + "loss": 0.9717, + "step": 10393 + }, + { + "epoch": 1.8506054131054133, + "grad_norm": 0.7230474948883057, + "learning_rate": 0.00011188369883203647, + "loss": 0.9043, + "step": 10394 + }, + { + "epoch": 1.8507834757834758, + "grad_norm": 0.8837162852287292, + "learning_rate": 0.00011186980043269235, + "loss": 1.2821, + "step": 10395 + }, + { + "epoch": 1.8509615384615383, + "grad_norm": 0.7260291576385498, + "learning_rate": 0.00011185590180078413, + "loss": 1.1672, + "step": 10396 + }, + { + "epoch": 1.851139601139601, + "grad_norm": 0.6290066242218018, + "learning_rate": 0.00011184200293658415, + "loss": 0.8942, + "step": 10397 + }, + { + "epoch": 1.8513176638176638, + "grad_norm": 0.6571013331413269, + "learning_rate": 0.00011182810384036475, + "loss": 1.0753, + "step": 10398 + }, + { + "epoch": 1.8514957264957266, + "grad_norm": 0.6494737267494202, + "learning_rate": 0.00011181420451239817, + "loss": 0.8833, + "step": 10399 + }, + { + "epoch": 1.8516737891737893, + "grad_norm": 0.7383694648742676, + "learning_rate": 0.00011180030495295684, + "loss": 1.0094, + "step": 10400 + }, + { + "epoch": 1.8518518518518519, + "grad_norm": 0.6713876724243164, + "learning_rate": 0.00011178640516231302, + "loss": 0.975, + "step": 10401 + }, + { + "epoch": 1.8520299145299144, + "grad_norm": 0.8041042685508728, + "learning_rate": 0.00011177250514073912, + "loss": 1.1419, + "step": 10402 + }, + { + "epoch": 1.8522079772079771, + "grad_norm": 0.7035061120986938, + "learning_rate": 0.00011175860488850738, + "loss": 1.0921, + "step": 10403 + }, + { + "epoch": 1.85238603988604, + "grad_norm": 0.6135673522949219, + "learning_rate": 0.00011174470440589022, + "loss": 0.9611, + "step": 10404 + }, + { + "epoch": 1.8525641025641026, + "grad_norm": 0.7868386507034302, + "learning_rate": 0.00011173080369315999, + "loss": 0.8561, + "step": 10405 + }, + { + "epoch": 1.8527421652421654, + "grad_norm": 0.6575735211372375, + "learning_rate": 0.00011171690275058902, + "loss": 1.0256, + "step": 10406 + }, + { + "epoch": 1.852920227920228, + "grad_norm": 0.7514392137527466, + "learning_rate": 0.00011170300157844969, + "loss": 1.0868, + "step": 10407 + }, + { + "epoch": 1.8530982905982905, + "grad_norm": 0.6915257573127747, + "learning_rate": 0.00011168910017701436, + "loss": 1.1223, + "step": 10408 + }, + { + "epoch": 1.8532763532763532, + "grad_norm": 0.7406772971153259, + "learning_rate": 0.00011167519854655535, + "loss": 1.0922, + "step": 10409 + }, + { + "epoch": 1.853454415954416, + "grad_norm": 0.6632742881774902, + "learning_rate": 0.0001116612966873451, + "loss": 0.9082, + "step": 10410 + }, + { + "epoch": 1.8536324786324787, + "grad_norm": 0.8154461979866028, + "learning_rate": 0.00011164739459965598, + "loss": 1.1126, + "step": 10411 + }, + { + "epoch": 1.8538105413105415, + "grad_norm": 0.895764172077179, + "learning_rate": 0.00011163349228376037, + "loss": 1.0589, + "step": 10412 + }, + { + "epoch": 1.853988603988604, + "grad_norm": 0.6746504902839661, + "learning_rate": 0.00011161958973993063, + "loss": 1.0184, + "step": 10413 + }, + { + "epoch": 1.8541666666666665, + "grad_norm": 0.7271263003349304, + "learning_rate": 0.00011160568696843916, + "loss": 0.9989, + "step": 10414 + }, + { + "epoch": 1.8543447293447293, + "grad_norm": 0.7503132820129395, + "learning_rate": 0.00011159178396955836, + "loss": 1.0783, + "step": 10415 + }, + { + "epoch": 1.854522792022792, + "grad_norm": 0.6768177151679993, + "learning_rate": 0.00011157788074356066, + "loss": 0.9916, + "step": 10416 + }, + { + "epoch": 1.8547008547008548, + "grad_norm": 0.6804978251457214, + "learning_rate": 0.00011156397729071842, + "loss": 0.9534, + "step": 10417 + }, + { + "epoch": 1.8548789173789175, + "grad_norm": 0.7144617438316345, + "learning_rate": 0.00011155007361130408, + "loss": 0.991, + "step": 10418 + }, + { + "epoch": 1.85505698005698, + "grad_norm": 0.6816750168800354, + "learning_rate": 0.00011153616970559, + "loss": 0.9551, + "step": 10419 + }, + { + "epoch": 1.8552350427350426, + "grad_norm": 0.6620030999183655, + "learning_rate": 0.00011152226557384866, + "loss": 0.8854, + "step": 10420 + }, + { + "epoch": 1.8554131054131053, + "grad_norm": 0.8400058746337891, + "learning_rate": 0.00011150836121635249, + "loss": 1.1593, + "step": 10421 + }, + { + "epoch": 1.855591168091168, + "grad_norm": 0.6666815280914307, + "learning_rate": 0.00011149445663337385, + "loss": 1.2112, + "step": 10422 + }, + { + "epoch": 1.8557692307692308, + "grad_norm": 0.7298431396484375, + "learning_rate": 0.00011148055182518522, + "loss": 0.9721, + "step": 10423 + }, + { + "epoch": 1.8559472934472936, + "grad_norm": 0.66816645860672, + "learning_rate": 0.00011146664679205903, + "loss": 1.0945, + "step": 10424 + }, + { + "epoch": 1.8561253561253561, + "grad_norm": 0.5979483127593994, + "learning_rate": 0.00011145274153426771, + "loss": 1.0176, + "step": 10425 + }, + { + "epoch": 1.8563034188034186, + "grad_norm": 0.6579445600509644, + "learning_rate": 0.00011143883605208372, + "loss": 0.9143, + "step": 10426 + }, + { + "epoch": 1.8564814814814814, + "grad_norm": 0.6871697902679443, + "learning_rate": 0.0001114249303457795, + "loss": 1.071, + "step": 10427 + }, + { + "epoch": 1.8566595441595442, + "grad_norm": 0.6683333516120911, + "learning_rate": 0.0001114110244156275, + "loss": 0.7809, + "step": 10428 + }, + { + "epoch": 1.856837606837607, + "grad_norm": 0.6122907996177673, + "learning_rate": 0.0001113971182619002, + "loss": 0.8329, + "step": 10429 + }, + { + "epoch": 1.8570156695156697, + "grad_norm": 0.6510575413703918, + "learning_rate": 0.00011138321188487, + "loss": 1.0068, + "step": 10430 + }, + { + "epoch": 1.8571937321937322, + "grad_norm": 0.6417793035507202, + "learning_rate": 0.00011136930528480945, + "loss": 1.0093, + "step": 10431 + }, + { + "epoch": 1.8573717948717947, + "grad_norm": 0.595824658870697, + "learning_rate": 0.00011135539846199096, + "loss": 0.9856, + "step": 10432 + }, + { + "epoch": 1.8575498575498575, + "grad_norm": 0.7594470381736755, + "learning_rate": 0.00011134149141668704, + "loss": 0.8173, + "step": 10433 + }, + { + "epoch": 1.8577279202279202, + "grad_norm": 0.7078324556350708, + "learning_rate": 0.00011132758414917016, + "loss": 1.0236, + "step": 10434 + }, + { + "epoch": 1.857905982905983, + "grad_norm": 0.6830437779426575, + "learning_rate": 0.00011131367665971275, + "loss": 0.8483, + "step": 10435 + }, + { + "epoch": 1.8580840455840457, + "grad_norm": 0.6856399774551392, + "learning_rate": 0.0001112997689485874, + "loss": 0.8729, + "step": 10436 + }, + { + "epoch": 1.8582621082621082, + "grad_norm": 0.6530426144599915, + "learning_rate": 0.00011128586101606653, + "loss": 0.8616, + "step": 10437 + }, + { + "epoch": 1.8584401709401708, + "grad_norm": 0.6341808438301086, + "learning_rate": 0.00011127195286242267, + "loss": 0.896, + "step": 10438 + }, + { + "epoch": 1.8586182336182335, + "grad_norm": 0.6278257966041565, + "learning_rate": 0.00011125804448792831, + "loss": 0.8309, + "step": 10439 + }, + { + "epoch": 1.8587962962962963, + "grad_norm": 0.708705723285675, + "learning_rate": 0.00011124413589285594, + "loss": 1.1065, + "step": 10440 + }, + { + "epoch": 1.858974358974359, + "grad_norm": 0.6845232248306274, + "learning_rate": 0.00011123022707747808, + "loss": 0.9292, + "step": 10441 + }, + { + "epoch": 1.8591524216524218, + "grad_norm": 0.749204695224762, + "learning_rate": 0.00011121631804206726, + "loss": 1.0487, + "step": 10442 + }, + { + "epoch": 1.8593304843304843, + "grad_norm": 0.7123128771781921, + "learning_rate": 0.00011120240878689599, + "loss": 0.9138, + "step": 10443 + }, + { + "epoch": 1.859508547008547, + "grad_norm": 0.6862115263938904, + "learning_rate": 0.00011118849931223679, + "loss": 1.0675, + "step": 10444 + }, + { + "epoch": 1.8596866096866096, + "grad_norm": 0.7245760560035706, + "learning_rate": 0.00011117458961836215, + "loss": 0.9643, + "step": 10445 + }, + { + "epoch": 1.8598646723646723, + "grad_norm": 0.701574444770813, + "learning_rate": 0.0001111606797055447, + "loss": 1.0022, + "step": 10446 + }, + { + "epoch": 1.860042735042735, + "grad_norm": 0.7292088270187378, + "learning_rate": 0.0001111467695740569, + "loss": 0.9465, + "step": 10447 + }, + { + "epoch": 1.8602207977207978, + "grad_norm": 0.7045044302940369, + "learning_rate": 0.0001111328592241713, + "loss": 1.0942, + "step": 10448 + }, + { + "epoch": 1.8603988603988604, + "grad_norm": 0.7181426286697388, + "learning_rate": 0.00011111894865616046, + "loss": 1.2108, + "step": 10449 + }, + { + "epoch": 1.8605769230769231, + "grad_norm": 0.6083306074142456, + "learning_rate": 0.00011110503787029689, + "loss": 0.929, + "step": 10450 + }, + { + "epoch": 1.8607549857549857, + "grad_norm": 0.6847347617149353, + "learning_rate": 0.00011109112686685319, + "loss": 1.0911, + "step": 10451 + }, + { + "epoch": 1.8609330484330484, + "grad_norm": 0.7131744027137756, + "learning_rate": 0.0001110772156461019, + "loss": 0.9649, + "step": 10452 + }, + { + "epoch": 1.8611111111111112, + "grad_norm": 0.7920312881469727, + "learning_rate": 0.00011106330420831559, + "loss": 0.9965, + "step": 10453 + }, + { + "epoch": 1.861289173789174, + "grad_norm": 0.6640987992286682, + "learning_rate": 0.00011104939255376681, + "loss": 1.2346, + "step": 10454 + }, + { + "epoch": 1.8614672364672364, + "grad_norm": 0.5878208875656128, + "learning_rate": 0.00011103548068272811, + "loss": 0.8565, + "step": 10455 + }, + { + "epoch": 1.8616452991452992, + "grad_norm": 0.6636882424354553, + "learning_rate": 0.0001110215685954721, + "loss": 0.8556, + "step": 10456 + }, + { + "epoch": 1.8618233618233617, + "grad_norm": 0.5985570549964905, + "learning_rate": 0.00011100765629227137, + "loss": 1.0291, + "step": 10457 + }, + { + "epoch": 1.8620014245014245, + "grad_norm": 0.7546643614768982, + "learning_rate": 0.00011099374377339846, + "loss": 1.0199, + "step": 10458 + }, + { + "epoch": 1.8621794871794872, + "grad_norm": 0.6529727578163147, + "learning_rate": 0.00011097983103912602, + "loss": 1.0826, + "step": 10459 + }, + { + "epoch": 1.86235754985755, + "grad_norm": 0.6394338607788086, + "learning_rate": 0.00011096591808972654, + "loss": 0.9896, + "step": 10460 + }, + { + "epoch": 1.8625356125356125, + "grad_norm": 0.6508805751800537, + "learning_rate": 0.00011095200492547271, + "loss": 0.9659, + "step": 10461 + }, + { + "epoch": 1.8627136752136753, + "grad_norm": 0.7085812091827393, + "learning_rate": 0.00011093809154663705, + "loss": 0.9998, + "step": 10462 + }, + { + "epoch": 1.8628917378917378, + "grad_norm": 0.6488457322120667, + "learning_rate": 0.00011092417795349226, + "loss": 0.9757, + "step": 10463 + }, + { + "epoch": 1.8630698005698005, + "grad_norm": 0.6405763626098633, + "learning_rate": 0.0001109102641463109, + "loss": 0.8188, + "step": 10464 + }, + { + "epoch": 1.8632478632478633, + "grad_norm": 0.713361918926239, + "learning_rate": 0.00011089635012536554, + "loss": 0.886, + "step": 10465 + }, + { + "epoch": 1.863425925925926, + "grad_norm": 0.5752255916595459, + "learning_rate": 0.00011088243589092886, + "loss": 1.0223, + "step": 10466 + }, + { + "epoch": 1.8636039886039886, + "grad_norm": 0.6722734570503235, + "learning_rate": 0.00011086852144327344, + "loss": 0.9499, + "step": 10467 + }, + { + "epoch": 1.8637820512820513, + "grad_norm": 0.5516420006752014, + "learning_rate": 0.00011085460678267194, + "loss": 0.7767, + "step": 10468 + }, + { + "epoch": 1.8639601139601139, + "grad_norm": 0.731257438659668, + "learning_rate": 0.00011084069190939697, + "loss": 1.2299, + "step": 10469 + }, + { + "epoch": 1.8641381766381766, + "grad_norm": 0.7977055907249451, + "learning_rate": 0.00011082677682372114, + "loss": 0.9109, + "step": 10470 + }, + { + "epoch": 1.8643162393162394, + "grad_norm": 0.679900586605072, + "learning_rate": 0.0001108128615259171, + "loss": 0.9319, + "step": 10471 + }, + { + "epoch": 1.864494301994302, + "grad_norm": 0.7428545951843262, + "learning_rate": 0.00011079894601625754, + "loss": 0.8585, + "step": 10472 + }, + { + "epoch": 1.8646723646723646, + "grad_norm": 0.6560967564582825, + "learning_rate": 0.00011078503029501504, + "loss": 1.0069, + "step": 10473 + }, + { + "epoch": 1.8648504273504274, + "grad_norm": 0.636202871799469, + "learning_rate": 0.00011077111436246228, + "loss": 1.0329, + "step": 10474 + }, + { + "epoch": 1.86502849002849, + "grad_norm": 0.6666205525398254, + "learning_rate": 0.00011075719821887191, + "loss": 1.0123, + "step": 10475 + }, + { + "epoch": 1.8652065527065527, + "grad_norm": 0.7089471220970154, + "learning_rate": 0.00011074328186451657, + "loss": 0.7851, + "step": 10476 + }, + { + "epoch": 1.8653846153846154, + "grad_norm": 0.6054788827896118, + "learning_rate": 0.00011072936529966895, + "loss": 0.8224, + "step": 10477 + }, + { + "epoch": 1.8655626780626782, + "grad_norm": 0.6009029150009155, + "learning_rate": 0.00011071544852460172, + "loss": 0.865, + "step": 10478 + }, + { + "epoch": 1.8657407407407407, + "grad_norm": 0.6238716244697571, + "learning_rate": 0.00011070153153958753, + "loss": 0.8685, + "step": 10479 + }, + { + "epoch": 1.8659188034188035, + "grad_norm": 0.719985842704773, + "learning_rate": 0.00011068761434489903, + "loss": 1.2204, + "step": 10480 + }, + { + "epoch": 1.866096866096866, + "grad_norm": 0.72972172498703, + "learning_rate": 0.00011067369694080895, + "loss": 1.0454, + "step": 10481 + }, + { + "epoch": 1.8662749287749287, + "grad_norm": 0.6741998791694641, + "learning_rate": 0.00011065977932758995, + "loss": 0.9992, + "step": 10482 + }, + { + "epoch": 1.8664529914529915, + "grad_norm": 0.6150268912315369, + "learning_rate": 0.00011064586150551472, + "loss": 0.8866, + "step": 10483 + }, + { + "epoch": 1.8666310541310542, + "grad_norm": 0.8253782391548157, + "learning_rate": 0.00011063194347485597, + "loss": 1.1173, + "step": 10484 + }, + { + "epoch": 1.8668091168091168, + "grad_norm": 0.7176247835159302, + "learning_rate": 0.00011061802523588636, + "loss": 1.0414, + "step": 10485 + }, + { + "epoch": 1.8669871794871795, + "grad_norm": 0.6372736096382141, + "learning_rate": 0.00011060410678887858, + "loss": 1.0548, + "step": 10486 + }, + { + "epoch": 1.867165242165242, + "grad_norm": 0.7107454538345337, + "learning_rate": 0.00011059018813410538, + "loss": 1.2298, + "step": 10487 + }, + { + "epoch": 1.8673433048433048, + "grad_norm": 0.7113911509513855, + "learning_rate": 0.00011057626927183944, + "loss": 0.9598, + "step": 10488 + }, + { + "epoch": 1.8675213675213675, + "grad_norm": 0.6734410524368286, + "learning_rate": 0.00011056235020235346, + "loss": 0.9475, + "step": 10489 + }, + { + "epoch": 1.8676994301994303, + "grad_norm": 0.6875202655792236, + "learning_rate": 0.0001105484309259202, + "loss": 1.0735, + "step": 10490 + }, + { + "epoch": 1.8678774928774928, + "grad_norm": 0.6908353567123413, + "learning_rate": 0.0001105345114428123, + "loss": 1.0558, + "step": 10491 + }, + { + "epoch": 1.8680555555555556, + "grad_norm": 0.6283324360847473, + "learning_rate": 0.00011052059175330256, + "loss": 0.8872, + "step": 10492 + }, + { + "epoch": 1.868233618233618, + "grad_norm": 0.6422587633132935, + "learning_rate": 0.00011050667185766368, + "loss": 1.1022, + "step": 10493 + }, + { + "epoch": 1.8684116809116809, + "grad_norm": 0.7075859904289246, + "learning_rate": 0.0001104927517561684, + "loss": 1.1389, + "step": 10494 + }, + { + "epoch": 1.8685897435897436, + "grad_norm": 0.5896905064582825, + "learning_rate": 0.00011047883144908944, + "loss": 0.7732, + "step": 10495 + }, + { + "epoch": 1.8687678062678064, + "grad_norm": 0.7647629976272583, + "learning_rate": 0.00011046491093669953, + "loss": 0.9983, + "step": 10496 + }, + { + "epoch": 1.868945868945869, + "grad_norm": 0.5864735841751099, + "learning_rate": 0.00011045099021927144, + "loss": 0.8427, + "step": 10497 + }, + { + "epoch": 1.8691239316239316, + "grad_norm": 0.6766837239265442, + "learning_rate": 0.00011043706929707791, + "loss": 0.9595, + "step": 10498 + }, + { + "epoch": 1.8693019943019942, + "grad_norm": 0.5480074286460876, + "learning_rate": 0.00011042314817039168, + "loss": 0.691, + "step": 10499 + }, + { + "epoch": 1.869480056980057, + "grad_norm": 0.6259615421295166, + "learning_rate": 0.00011040922683948553, + "loss": 0.9991, + "step": 10500 + }, + { + "epoch": 1.8696581196581197, + "grad_norm": 0.5950598120689392, + "learning_rate": 0.00011039530530463218, + "loss": 0.7413, + "step": 10501 + }, + { + "epoch": 1.8698361823361824, + "grad_norm": 0.8099377751350403, + "learning_rate": 0.00011038138356610441, + "loss": 1.1351, + "step": 10502 + }, + { + "epoch": 1.8700142450142452, + "grad_norm": 0.6716185212135315, + "learning_rate": 0.00011036746162417501, + "loss": 1.1057, + "step": 10503 + }, + { + "epoch": 1.8701923076923077, + "grad_norm": 0.7993219494819641, + "learning_rate": 0.00011035353947911675, + "loss": 1.2095, + "step": 10504 + }, + { + "epoch": 1.8703703703703702, + "grad_norm": 0.6381276249885559, + "learning_rate": 0.00011033961713120237, + "loss": 1.0261, + "step": 10505 + }, + { + "epoch": 1.870548433048433, + "grad_norm": 0.6326032280921936, + "learning_rate": 0.00011032569458070469, + "loss": 0.8664, + "step": 10506 + }, + { + "epoch": 1.8707264957264957, + "grad_norm": 0.6864820718765259, + "learning_rate": 0.00011031177182789644, + "loss": 0.9959, + "step": 10507 + }, + { + "epoch": 1.8709045584045585, + "grad_norm": 0.6341838240623474, + "learning_rate": 0.00011029784887305048, + "loss": 0.8029, + "step": 10508 + }, + { + "epoch": 1.8710826210826212, + "grad_norm": 0.6559172868728638, + "learning_rate": 0.00011028392571643957, + "loss": 0.9282, + "step": 10509 + }, + { + "epoch": 1.8712606837606838, + "grad_norm": 0.6976849436759949, + "learning_rate": 0.0001102700023583365, + "loss": 1.0198, + "step": 10510 + }, + { + "epoch": 1.8714387464387463, + "grad_norm": 0.7159395217895508, + "learning_rate": 0.00011025607879901402, + "loss": 1.1585, + "step": 10511 + }, + { + "epoch": 1.871616809116809, + "grad_norm": 0.7168624997138977, + "learning_rate": 0.000110242155038745, + "loss": 1.0558, + "step": 10512 + }, + { + "epoch": 1.8717948717948718, + "grad_norm": 0.5784319043159485, + "learning_rate": 0.00011022823107780224, + "loss": 0.9481, + "step": 10513 + }, + { + "epoch": 1.8719729344729346, + "grad_norm": 0.6602259874343872, + "learning_rate": 0.00011021430691645856, + "loss": 1.0538, + "step": 10514 + }, + { + "epoch": 1.8721509971509973, + "grad_norm": 0.6874588131904602, + "learning_rate": 0.00011020038255498672, + "loss": 1.1396, + "step": 10515 + }, + { + "epoch": 1.8723290598290598, + "grad_norm": 0.7311663031578064, + "learning_rate": 0.00011018645799365956, + "loss": 1.084, + "step": 10516 + }, + { + "epoch": 1.8725071225071224, + "grad_norm": 0.7097118496894836, + "learning_rate": 0.00011017253323274996, + "loss": 0.9872, + "step": 10517 + }, + { + "epoch": 1.8726851851851851, + "grad_norm": 0.6667875051498413, + "learning_rate": 0.00011015860827253068, + "loss": 1.105, + "step": 10518 + }, + { + "epoch": 1.8728632478632479, + "grad_norm": 0.6807677745819092, + "learning_rate": 0.0001101446831132746, + "loss": 0.9093, + "step": 10519 + }, + { + "epoch": 1.8730413105413106, + "grad_norm": 0.6885797381401062, + "learning_rate": 0.0001101307577552545, + "loss": 0.8479, + "step": 10520 + }, + { + "epoch": 1.8732193732193734, + "grad_norm": 0.6269213557243347, + "learning_rate": 0.00011011683219874323, + "loss": 0.9457, + "step": 10521 + }, + { + "epoch": 1.873397435897436, + "grad_norm": 0.7096766829490662, + "learning_rate": 0.00011010290644401364, + "loss": 1.0971, + "step": 10522 + }, + { + "epoch": 1.8735754985754984, + "grad_norm": 0.6909209489822388, + "learning_rate": 0.00011008898049133863, + "loss": 0.9928, + "step": 10523 + }, + { + "epoch": 1.8737535612535612, + "grad_norm": 0.6586211323738098, + "learning_rate": 0.000110075054340991, + "loss": 0.818, + "step": 10524 + }, + { + "epoch": 1.873931623931624, + "grad_norm": 0.5934817790985107, + "learning_rate": 0.0001100611279932436, + "loss": 0.7698, + "step": 10525 + }, + { + "epoch": 1.8741096866096867, + "grad_norm": 0.6361709237098694, + "learning_rate": 0.00011004720144836931, + "loss": 0.9465, + "step": 10526 + }, + { + "epoch": 1.8742877492877494, + "grad_norm": 0.6742212176322937, + "learning_rate": 0.00011003327470664095, + "loss": 1.0998, + "step": 10527 + }, + { + "epoch": 1.874465811965812, + "grad_norm": 0.6634946465492249, + "learning_rate": 0.00011001934776833143, + "loss": 0.8328, + "step": 10528 + }, + { + "epoch": 1.8746438746438745, + "grad_norm": 0.6754063963890076, + "learning_rate": 0.0001100054206337136, + "loss": 1.147, + "step": 10529 + }, + { + "epoch": 1.8748219373219372, + "grad_norm": 0.5951135158538818, + "learning_rate": 0.00010999149330306036, + "loss": 0.8956, + "step": 10530 + }, + { + "epoch": 1.875, + "grad_norm": 0.6140317320823669, + "learning_rate": 0.00010997756577664455, + "loss": 0.9368, + "step": 10531 + }, + { + "epoch": 1.8751780626780628, + "grad_norm": 0.6419258713722229, + "learning_rate": 0.00010996363805473904, + "loss": 0.9817, + "step": 10532 + }, + { + "epoch": 1.8753561253561255, + "grad_norm": 0.7173396348953247, + "learning_rate": 0.00010994971013761677, + "loss": 0.9638, + "step": 10533 + }, + { + "epoch": 1.875534188034188, + "grad_norm": 0.8125925660133362, + "learning_rate": 0.0001099357820255506, + "loss": 1.0996, + "step": 10534 + }, + { + "epoch": 1.8757122507122506, + "grad_norm": 0.6191564798355103, + "learning_rate": 0.00010992185371881341, + "loss": 0.8266, + "step": 10535 + }, + { + "epoch": 1.8758903133903133, + "grad_norm": 0.6632885336875916, + "learning_rate": 0.0001099079252176781, + "loss": 1.1884, + "step": 10536 + }, + { + "epoch": 1.876068376068376, + "grad_norm": 0.7323372960090637, + "learning_rate": 0.00010989399652241759, + "loss": 1.0842, + "step": 10537 + }, + { + "epoch": 1.8762464387464388, + "grad_norm": 0.7553854584693909, + "learning_rate": 0.00010988006763330476, + "loss": 0.9948, + "step": 10538 + }, + { + "epoch": 1.8764245014245016, + "grad_norm": 0.5887658596038818, + "learning_rate": 0.00010986613855061255, + "loss": 0.7653, + "step": 10539 + }, + { + "epoch": 1.876602564102564, + "grad_norm": 0.6849574446678162, + "learning_rate": 0.00010985220927461384, + "loss": 1.152, + "step": 10540 + }, + { + "epoch": 1.8767806267806266, + "grad_norm": 0.6985000371932983, + "learning_rate": 0.00010983827980558155, + "loss": 0.9869, + "step": 10541 + }, + { + "epoch": 1.8769586894586894, + "grad_norm": 0.6885373592376709, + "learning_rate": 0.00010982435014378858, + "loss": 1.1803, + "step": 10542 + }, + { + "epoch": 1.8771367521367521, + "grad_norm": 0.7610142827033997, + "learning_rate": 0.00010981042028950788, + "loss": 0.9219, + "step": 10543 + }, + { + "epoch": 1.8773148148148149, + "grad_norm": 0.6545612215995789, + "learning_rate": 0.00010979649024301242, + "loss": 1.0337, + "step": 10544 + }, + { + "epoch": 1.8774928774928776, + "grad_norm": 0.7307698130607605, + "learning_rate": 0.00010978256000457505, + "loss": 0.9726, + "step": 10545 + }, + { + "epoch": 1.8776709401709402, + "grad_norm": 0.68310546875, + "learning_rate": 0.00010976862957446877, + "loss": 1.161, + "step": 10546 + }, + { + "epoch": 1.8778490028490027, + "grad_norm": 0.6114758253097534, + "learning_rate": 0.00010975469895296646, + "loss": 0.8863, + "step": 10547 + }, + { + "epoch": 1.8780270655270654, + "grad_norm": 0.732390820980072, + "learning_rate": 0.00010974076814034106, + "loss": 1.0339, + "step": 10548 + }, + { + "epoch": 1.8782051282051282, + "grad_norm": 0.6741712689399719, + "learning_rate": 0.0001097268371368656, + "loss": 1.0024, + "step": 10549 + }, + { + "epoch": 1.878383190883191, + "grad_norm": 0.6374897360801697, + "learning_rate": 0.00010971290594281294, + "loss": 0.91, + "step": 10550 + }, + { + "epoch": 1.8785612535612537, + "grad_norm": 0.6434261202812195, + "learning_rate": 0.00010969897455845608, + "loss": 1.0048, + "step": 10551 + }, + { + "epoch": 1.8787393162393162, + "grad_norm": 0.6573047041893005, + "learning_rate": 0.00010968504298406794, + "loss": 1.118, + "step": 10552 + }, + { + "epoch": 1.8789173789173788, + "grad_norm": 0.6686552166938782, + "learning_rate": 0.00010967111121992152, + "loss": 1.089, + "step": 10553 + }, + { + "epoch": 1.8790954415954415, + "grad_norm": 0.7899606823921204, + "learning_rate": 0.00010965717926628976, + "loss": 1.059, + "step": 10554 + }, + { + "epoch": 1.8792735042735043, + "grad_norm": 0.5808879733085632, + "learning_rate": 0.00010964324712344564, + "loss": 0.9369, + "step": 10555 + }, + { + "epoch": 1.879451566951567, + "grad_norm": 0.6322834491729736, + "learning_rate": 0.00010962931479166211, + "loss": 0.8783, + "step": 10556 + }, + { + "epoch": 1.8796296296296298, + "grad_norm": 0.647002637386322, + "learning_rate": 0.00010961538227121218, + "loss": 0.9468, + "step": 10557 + }, + { + "epoch": 1.8798076923076923, + "grad_norm": 0.6581854820251465, + "learning_rate": 0.0001096014495623688, + "loss": 1.0077, + "step": 10558 + }, + { + "epoch": 1.8799857549857548, + "grad_norm": 0.6879259943962097, + "learning_rate": 0.00010958751666540496, + "loss": 0.976, + "step": 10559 + }, + { + "epoch": 1.8801638176638176, + "grad_norm": 0.7055090665817261, + "learning_rate": 0.00010957358358059364, + "loss": 0.8903, + "step": 10560 + }, + { + "epoch": 1.8803418803418803, + "grad_norm": 0.6865016222000122, + "learning_rate": 0.00010955965030820782, + "loss": 0.9872, + "step": 10561 + }, + { + "epoch": 1.880519943019943, + "grad_norm": 0.663436770439148, + "learning_rate": 0.00010954571684852055, + "loss": 1.0485, + "step": 10562 + }, + { + "epoch": 1.8806980056980058, + "grad_norm": 0.6861656904220581, + "learning_rate": 0.00010953178320180475, + "loss": 1.0691, + "step": 10563 + }, + { + "epoch": 1.8808760683760684, + "grad_norm": 0.8045449256896973, + "learning_rate": 0.0001095178493683335, + "loss": 1.1534, + "step": 10564 + }, + { + "epoch": 1.881054131054131, + "grad_norm": 0.6493151187896729, + "learning_rate": 0.00010950391534837973, + "loss": 0.8756, + "step": 10565 + }, + { + "epoch": 1.8812321937321936, + "grad_norm": 0.7057121992111206, + "learning_rate": 0.00010948998114221651, + "loss": 1.1709, + "step": 10566 + }, + { + "epoch": 1.8814102564102564, + "grad_norm": 0.7708197236061096, + "learning_rate": 0.0001094760467501168, + "loss": 1.0037, + "step": 10567 + }, + { + "epoch": 1.8815883190883191, + "grad_norm": 0.7234642505645752, + "learning_rate": 0.00010946211217235364, + "loss": 1.0757, + "step": 10568 + }, + { + "epoch": 1.881766381766382, + "grad_norm": 0.6964395642280579, + "learning_rate": 0.00010944817740920006, + "loss": 1.0769, + "step": 10569 + }, + { + "epoch": 1.8819444444444444, + "grad_norm": 0.7465848922729492, + "learning_rate": 0.00010943424246092906, + "loss": 0.9772, + "step": 10570 + }, + { + "epoch": 1.8821225071225072, + "grad_norm": 0.7145788073539734, + "learning_rate": 0.0001094203073278137, + "loss": 0.9638, + "step": 10571 + }, + { + "epoch": 1.8823005698005697, + "grad_norm": 0.7421764135360718, + "learning_rate": 0.00010940637201012698, + "loss": 1.0324, + "step": 10572 + }, + { + "epoch": 1.8824786324786325, + "grad_norm": 0.7373253107070923, + "learning_rate": 0.0001093924365081419, + "loss": 1.1554, + "step": 10573 + }, + { + "epoch": 1.8826566951566952, + "grad_norm": 0.6861984729766846, + "learning_rate": 0.00010937850082213156, + "loss": 0.9899, + "step": 10574 + }, + { + "epoch": 1.882834757834758, + "grad_norm": 0.6173393130302429, + "learning_rate": 0.000109364564952369, + "loss": 0.8495, + "step": 10575 + }, + { + "epoch": 1.8830128205128205, + "grad_norm": 0.6871610879898071, + "learning_rate": 0.00010935062889912723, + "loss": 1.2164, + "step": 10576 + }, + { + "epoch": 1.8831908831908832, + "grad_norm": 0.7062903642654419, + "learning_rate": 0.00010933669266267931, + "loss": 1.1077, + "step": 10577 + }, + { + "epoch": 1.8833689458689458, + "grad_norm": 0.6574689745903015, + "learning_rate": 0.00010932275624329828, + "loss": 0.9326, + "step": 10578 + }, + { + "epoch": 1.8835470085470085, + "grad_norm": 0.636385440826416, + "learning_rate": 0.00010930881964125723, + "loss": 1.0581, + "step": 10579 + }, + { + "epoch": 1.8837250712250713, + "grad_norm": 0.6178432106971741, + "learning_rate": 0.0001092948828568292, + "loss": 1.1288, + "step": 10580 + }, + { + "epoch": 1.883903133903134, + "grad_norm": 0.6509431600570679, + "learning_rate": 0.00010928094589028721, + "loss": 1.0113, + "step": 10581 + }, + { + "epoch": 1.8840811965811965, + "grad_norm": 0.6543706059455872, + "learning_rate": 0.00010926700874190441, + "loss": 1.0041, + "step": 10582 + }, + { + "epoch": 1.8842592592592593, + "grad_norm": 0.6815463304519653, + "learning_rate": 0.0001092530714119538, + "loss": 1.0892, + "step": 10583 + }, + { + "epoch": 1.8844373219373218, + "grad_norm": 0.6787421107292175, + "learning_rate": 0.00010923913390070846, + "loss": 1.2693, + "step": 10584 + }, + { + "epoch": 1.8846153846153846, + "grad_norm": 0.6953850984573364, + "learning_rate": 0.00010922519620844151, + "loss": 0.9848, + "step": 10585 + }, + { + "epoch": 1.8847934472934473, + "grad_norm": 0.7061360478401184, + "learning_rate": 0.000109211258335426, + "loss": 0.949, + "step": 10586 + }, + { + "epoch": 1.88497150997151, + "grad_norm": 0.6845372915267944, + "learning_rate": 0.00010919732028193504, + "loss": 0.9554, + "step": 10587 + }, + { + "epoch": 1.8851495726495726, + "grad_norm": 0.6524720788002014, + "learning_rate": 0.00010918338204824165, + "loss": 1.1037, + "step": 10588 + }, + { + "epoch": 1.8853276353276354, + "grad_norm": 0.6410523653030396, + "learning_rate": 0.00010916944363461899, + "loss": 0.9085, + "step": 10589 + }, + { + "epoch": 1.885505698005698, + "grad_norm": 0.7109059691429138, + "learning_rate": 0.00010915550504134014, + "loss": 1.0526, + "step": 10590 + }, + { + "epoch": 1.8856837606837606, + "grad_norm": 0.7781991362571716, + "learning_rate": 0.00010914156626867818, + "loss": 0.9737, + "step": 10591 + }, + { + "epoch": 1.8858618233618234, + "grad_norm": 0.7173767685890198, + "learning_rate": 0.00010912762731690623, + "loss": 0.8862, + "step": 10592 + }, + { + "epoch": 1.8860398860398861, + "grad_norm": 0.7650504112243652, + "learning_rate": 0.00010911368818629732, + "loss": 1.2175, + "step": 10593 + }, + { + "epoch": 1.8862179487179487, + "grad_norm": 0.6316116452217102, + "learning_rate": 0.00010909974887712468, + "loss": 0.8332, + "step": 10594 + }, + { + "epoch": 1.8863960113960114, + "grad_norm": 0.6504800319671631, + "learning_rate": 0.00010908580938966138, + "loss": 0.8864, + "step": 10595 + }, + { + "epoch": 1.886574074074074, + "grad_norm": 0.675507128238678, + "learning_rate": 0.00010907186972418049, + "loss": 0.8523, + "step": 10596 + }, + { + "epoch": 1.8867521367521367, + "grad_norm": 0.6535763144493103, + "learning_rate": 0.00010905792988095515, + "loss": 1.0786, + "step": 10597 + }, + { + "epoch": 1.8869301994301995, + "grad_norm": 0.7071853280067444, + "learning_rate": 0.0001090439898602585, + "loss": 0.9319, + "step": 10598 + }, + { + "epoch": 1.8871082621082622, + "grad_norm": 0.699466347694397, + "learning_rate": 0.00010903004966236365, + "loss": 0.9573, + "step": 10599 + }, + { + "epoch": 1.8872863247863247, + "grad_norm": 0.7099201083183289, + "learning_rate": 0.00010901610928754375, + "loss": 0.9447, + "step": 10600 + }, + { + "epoch": 1.8874643874643875, + "grad_norm": 0.6140450835227966, + "learning_rate": 0.00010900216873607189, + "loss": 1.0227, + "step": 10601 + }, + { + "epoch": 1.88764245014245, + "grad_norm": 0.6613629460334778, + "learning_rate": 0.00010898822800822127, + "loss": 1.0152, + "step": 10602 + }, + { + "epoch": 1.8878205128205128, + "grad_norm": 0.7334819436073303, + "learning_rate": 0.00010897428710426498, + "loss": 1.1452, + "step": 10603 + }, + { + "epoch": 1.8879985754985755, + "grad_norm": 0.6819368004798889, + "learning_rate": 0.00010896034602447616, + "loss": 1.0504, + "step": 10604 + }, + { + "epoch": 1.8881766381766383, + "grad_norm": 0.6781361103057861, + "learning_rate": 0.00010894640476912799, + "loss": 0.8719, + "step": 10605 + }, + { + "epoch": 1.8883547008547008, + "grad_norm": 0.621960461139679, + "learning_rate": 0.00010893246333849361, + "loss": 0.9264, + "step": 10606 + }, + { + "epoch": 1.8885327635327636, + "grad_norm": 0.6350592374801636, + "learning_rate": 0.00010891852173284615, + "loss": 1.0042, + "step": 10607 + }, + { + "epoch": 1.888710826210826, + "grad_norm": 0.6650694012641907, + "learning_rate": 0.00010890457995245879, + "loss": 1.1387, + "step": 10608 + }, + { + "epoch": 1.8888888888888888, + "grad_norm": 0.6515723466873169, + "learning_rate": 0.00010889063799760468, + "loss": 0.9508, + "step": 10609 + }, + { + "epoch": 1.8890669515669516, + "grad_norm": 0.6368890404701233, + "learning_rate": 0.000108876695868557, + "loss": 0.8051, + "step": 10610 + }, + { + "epoch": 1.8892450142450143, + "grad_norm": 0.7971013188362122, + "learning_rate": 0.00010886275356558888, + "loss": 0.8629, + "step": 10611 + }, + { + "epoch": 1.8894230769230769, + "grad_norm": 0.6739095449447632, + "learning_rate": 0.00010884881108897353, + "loss": 0.9606, + "step": 10612 + }, + { + "epoch": 1.8896011396011396, + "grad_norm": 0.7754076719284058, + "learning_rate": 0.00010883486843898412, + "loss": 1.0751, + "step": 10613 + }, + { + "epoch": 1.8897792022792022, + "grad_norm": 0.6538285613059998, + "learning_rate": 0.00010882092561589379, + "loss": 0.9288, + "step": 10614 + }, + { + "epoch": 1.889957264957265, + "grad_norm": 0.7373257875442505, + "learning_rate": 0.00010880698261997577, + "loss": 0.9884, + "step": 10615 + }, + { + "epoch": 1.8901353276353277, + "grad_norm": 0.6575660109519958, + "learning_rate": 0.00010879303945150321, + "loss": 1.0307, + "step": 10616 + }, + { + "epoch": 1.8903133903133904, + "grad_norm": 0.7500179409980774, + "learning_rate": 0.00010877909611074932, + "loss": 1.0812, + "step": 10617 + }, + { + "epoch": 1.890491452991453, + "grad_norm": 0.7607308030128479, + "learning_rate": 0.00010876515259798727, + "loss": 0.9746, + "step": 10618 + }, + { + "epoch": 1.8906695156695157, + "grad_norm": 0.7930253744125366, + "learning_rate": 0.00010875120891349024, + "loss": 0.7911, + "step": 10619 + }, + { + "epoch": 1.8908475783475782, + "grad_norm": 0.635254979133606, + "learning_rate": 0.00010873726505753148, + "loss": 1.0468, + "step": 10620 + }, + { + "epoch": 1.891025641025641, + "grad_norm": 0.7579759359359741, + "learning_rate": 0.00010872332103038414, + "loss": 0.9558, + "step": 10621 + }, + { + "epoch": 1.8912037037037037, + "grad_norm": 0.5841903686523438, + "learning_rate": 0.00010870937683232146, + "loss": 0.913, + "step": 10622 + }, + { + "epoch": 1.8913817663817665, + "grad_norm": 0.7088860273361206, + "learning_rate": 0.00010869543246361664, + "loss": 1.0814, + "step": 10623 + }, + { + "epoch": 1.8915598290598292, + "grad_norm": 0.6713772416114807, + "learning_rate": 0.00010868148792454285, + "loss": 0.9972, + "step": 10624 + }, + { + "epoch": 1.8917378917378918, + "grad_norm": 0.6733243465423584, + "learning_rate": 0.00010866754321537338, + "loss": 0.9596, + "step": 10625 + }, + { + "epoch": 1.8919159544159543, + "grad_norm": 0.7747747898101807, + "learning_rate": 0.00010865359833638138, + "loss": 1.0871, + "step": 10626 + }, + { + "epoch": 1.892094017094017, + "grad_norm": 0.677175760269165, + "learning_rate": 0.00010863965328784011, + "loss": 0.9939, + "step": 10627 + }, + { + "epoch": 1.8922720797720798, + "grad_norm": 0.7883930206298828, + "learning_rate": 0.00010862570807002279, + "loss": 1.0708, + "step": 10628 + }, + { + "epoch": 1.8924501424501425, + "grad_norm": 0.7003030180931091, + "learning_rate": 0.00010861176268320261, + "loss": 0.9791, + "step": 10629 + }, + { + "epoch": 1.8926282051282053, + "grad_norm": 0.7450358271598816, + "learning_rate": 0.00010859781712765284, + "loss": 0.9672, + "step": 10630 + }, + { + "epoch": 1.8928062678062678, + "grad_norm": 0.7776696085929871, + "learning_rate": 0.00010858387140364672, + "loss": 1.1037, + "step": 10631 + }, + { + "epoch": 1.8929843304843303, + "grad_norm": 0.6896173357963562, + "learning_rate": 0.00010856992551145745, + "loss": 1.0048, + "step": 10632 + }, + { + "epoch": 1.893162393162393, + "grad_norm": 0.5997697710990906, + "learning_rate": 0.00010855597945135834, + "loss": 0.8025, + "step": 10633 + }, + { + "epoch": 1.8933404558404558, + "grad_norm": 0.8781484365463257, + "learning_rate": 0.00010854203322362251, + "loss": 1.0014, + "step": 10634 + }, + { + "epoch": 1.8935185185185186, + "grad_norm": 0.6348843574523926, + "learning_rate": 0.00010852808682852334, + "loss": 0.9857, + "step": 10635 + }, + { + "epoch": 1.8936965811965814, + "grad_norm": 0.9704267978668213, + "learning_rate": 0.000108514140266334, + "loss": 1.0522, + "step": 10636 + }, + { + "epoch": 1.8938746438746439, + "grad_norm": 0.70372074842453, + "learning_rate": 0.00010850019353732779, + "loss": 1.1044, + "step": 10637 + }, + { + "epoch": 1.8940527065527064, + "grad_norm": 0.6528043150901794, + "learning_rate": 0.00010848624664177793, + "loss": 0.9328, + "step": 10638 + }, + { + "epoch": 1.8942307692307692, + "grad_norm": 0.6299768090248108, + "learning_rate": 0.00010847229957995768, + "loss": 1.0099, + "step": 10639 + }, + { + "epoch": 1.894408831908832, + "grad_norm": 0.6347038149833679, + "learning_rate": 0.00010845835235214034, + "loss": 1.1354, + "step": 10640 + }, + { + "epoch": 1.8945868945868947, + "grad_norm": 0.7087811827659607, + "learning_rate": 0.00010844440495859913, + "loss": 1.0543, + "step": 10641 + }, + { + "epoch": 1.8947649572649574, + "grad_norm": 0.7386305332183838, + "learning_rate": 0.00010843045739960738, + "loss": 0.9192, + "step": 10642 + }, + { + "epoch": 1.89494301994302, + "grad_norm": 0.6047097444534302, + "learning_rate": 0.00010841650967543833, + "loss": 0.8668, + "step": 10643 + }, + { + "epoch": 1.8951210826210825, + "grad_norm": 0.6779503226280212, + "learning_rate": 0.00010840256178636523, + "loss": 0.9263, + "step": 10644 + }, + { + "epoch": 1.8952991452991452, + "grad_norm": 0.7398194670677185, + "learning_rate": 0.00010838861373266138, + "loss": 0.9534, + "step": 10645 + }, + { + "epoch": 1.895477207977208, + "grad_norm": 0.8138558864593506, + "learning_rate": 0.00010837466551460011, + "loss": 0.9835, + "step": 10646 + }, + { + "epoch": 1.8956552706552707, + "grad_norm": 0.8847818374633789, + "learning_rate": 0.00010836071713245466, + "loss": 0.9769, + "step": 10647 + }, + { + "epoch": 1.8958333333333335, + "grad_norm": 0.6824164390563965, + "learning_rate": 0.0001083467685864983, + "loss": 0.9901, + "step": 10648 + }, + { + "epoch": 1.896011396011396, + "grad_norm": 0.6318182945251465, + "learning_rate": 0.00010833281987700436, + "loss": 0.7677, + "step": 10649 + }, + { + "epoch": 1.8961894586894585, + "grad_norm": 0.7372074127197266, + "learning_rate": 0.00010831887100424612, + "loss": 0.9858, + "step": 10650 + }, + { + "epoch": 1.8963675213675213, + "grad_norm": 0.7246516346931458, + "learning_rate": 0.00010830492196849688, + "loss": 0.9644, + "step": 10651 + }, + { + "epoch": 1.896545584045584, + "grad_norm": 0.6517095565795898, + "learning_rate": 0.00010829097277002997, + "loss": 1.1733, + "step": 10652 + }, + { + "epoch": 1.8967236467236468, + "grad_norm": 0.6931695342063904, + "learning_rate": 0.00010827702340911867, + "loss": 0.9923, + "step": 10653 + }, + { + "epoch": 1.8969017094017095, + "grad_norm": 0.6210272312164307, + "learning_rate": 0.00010826307388603628, + "loss": 0.8757, + "step": 10654 + }, + { + "epoch": 1.897079772079772, + "grad_norm": 0.7011165618896484, + "learning_rate": 0.00010824912420105611, + "loss": 1.0011, + "step": 10655 + }, + { + "epoch": 1.8972578347578346, + "grad_norm": 0.7431246638298035, + "learning_rate": 0.0001082351743544515, + "loss": 1.1498, + "step": 10656 + }, + { + "epoch": 1.8974358974358974, + "grad_norm": 0.7099978923797607, + "learning_rate": 0.00010822122434649576, + "loss": 1.0673, + "step": 10657 + }, + { + "epoch": 1.89761396011396, + "grad_norm": 0.7375551462173462, + "learning_rate": 0.00010820727417746219, + "loss": 1.0157, + "step": 10658 + }, + { + "epoch": 1.8977920227920229, + "grad_norm": 0.8155642151832581, + "learning_rate": 0.00010819332384762413, + "loss": 1.229, + "step": 10659 + }, + { + "epoch": 1.8979700854700856, + "grad_norm": 0.6917914748191833, + "learning_rate": 0.00010817937335725493, + "loss": 0.9701, + "step": 10660 + }, + { + "epoch": 1.8981481481481481, + "grad_norm": 0.8498218059539795, + "learning_rate": 0.00010816542270662786, + "loss": 1.0123, + "step": 10661 + }, + { + "epoch": 1.8983262108262107, + "grad_norm": 0.7234359979629517, + "learning_rate": 0.00010815147189601634, + "loss": 1.0755, + "step": 10662 + }, + { + "epoch": 1.8985042735042734, + "grad_norm": 0.6997553110122681, + "learning_rate": 0.00010813752092569365, + "loss": 1.1594, + "step": 10663 + }, + { + "epoch": 1.8986823361823362, + "grad_norm": 0.6519457101821899, + "learning_rate": 0.00010812356979593314, + "loss": 0.9609, + "step": 10664 + }, + { + "epoch": 1.898860398860399, + "grad_norm": 0.7215374708175659, + "learning_rate": 0.00010810961850700813, + "loss": 1.1392, + "step": 10665 + }, + { + "epoch": 1.8990384615384617, + "grad_norm": 0.7766093611717224, + "learning_rate": 0.00010809566705919202, + "loss": 1.0256, + "step": 10666 + }, + { + "epoch": 1.8992165242165242, + "grad_norm": 0.6520358920097351, + "learning_rate": 0.00010808171545275814, + "loss": 1.0434, + "step": 10667 + }, + { + "epoch": 1.8993945868945867, + "grad_norm": 0.7454953193664551, + "learning_rate": 0.00010806776368797982, + "loss": 1.2323, + "step": 10668 + }, + { + "epoch": 1.8995726495726495, + "grad_norm": 0.6891530752182007, + "learning_rate": 0.00010805381176513043, + "loss": 1.1104, + "step": 10669 + }, + { + "epoch": 1.8997507122507122, + "grad_norm": 0.6609626412391663, + "learning_rate": 0.00010803985968448331, + "loss": 0.8565, + "step": 10670 + }, + { + "epoch": 1.899928774928775, + "grad_norm": 0.6650999188423157, + "learning_rate": 0.00010802590744631187, + "loss": 1.1003, + "step": 10671 + }, + { + "epoch": 1.9001068376068377, + "grad_norm": 0.5794292092323303, + "learning_rate": 0.00010801195505088945, + "loss": 0.528, + "step": 10672 + }, + { + "epoch": 1.9002849002849003, + "grad_norm": 1.0802743434906006, + "learning_rate": 0.00010799800249848939, + "loss": 0.8861, + "step": 10673 + }, + { + "epoch": 1.9004629629629628, + "grad_norm": 0.650833249092102, + "learning_rate": 0.00010798404978938513, + "loss": 0.9962, + "step": 10674 + }, + { + "epoch": 1.9006410256410255, + "grad_norm": 0.7290451526641846, + "learning_rate": 0.00010797009692384994, + "loss": 1.0764, + "step": 10675 + }, + { + "epoch": 1.9008190883190883, + "grad_norm": 0.6273928880691528, + "learning_rate": 0.00010795614390215727, + "loss": 0.9478, + "step": 10676 + }, + { + "epoch": 1.900997150997151, + "grad_norm": 0.6939455270767212, + "learning_rate": 0.00010794219072458052, + "loss": 0.8991, + "step": 10677 + }, + { + "epoch": 1.9011752136752138, + "grad_norm": 0.7455828189849854, + "learning_rate": 0.00010792823739139302, + "loss": 0.8902, + "step": 10678 + }, + { + "epoch": 1.9013532763532763, + "grad_norm": 0.6894607543945312, + "learning_rate": 0.00010791428390286817, + "loss": 0.9355, + "step": 10679 + }, + { + "epoch": 1.901531339031339, + "grad_norm": 0.6844658851623535, + "learning_rate": 0.00010790033025927936, + "loss": 0.9835, + "step": 10680 + }, + { + "epoch": 1.9017094017094016, + "grad_norm": 0.6646730899810791, + "learning_rate": 0.00010788637646090001, + "loss": 0.9376, + "step": 10681 + }, + { + "epoch": 1.9018874643874644, + "grad_norm": 0.6494864225387573, + "learning_rate": 0.00010787242250800349, + "loss": 0.8533, + "step": 10682 + }, + { + "epoch": 1.9020655270655271, + "grad_norm": 0.686198353767395, + "learning_rate": 0.0001078584684008632, + "loss": 0.8075, + "step": 10683 + }, + { + "epoch": 1.9022435897435899, + "grad_norm": 0.7014855742454529, + "learning_rate": 0.00010784451413975256, + "loss": 1.0805, + "step": 10684 + }, + { + "epoch": 1.9024216524216524, + "grad_norm": 0.7191864252090454, + "learning_rate": 0.00010783055972494496, + "loss": 0.9375, + "step": 10685 + }, + { + "epoch": 1.9025997150997151, + "grad_norm": 0.8114212155342102, + "learning_rate": 0.00010781660515671379, + "loss": 0.9716, + "step": 10686 + }, + { + "epoch": 1.9027777777777777, + "grad_norm": 0.7423529028892517, + "learning_rate": 0.0001078026504353325, + "loss": 0.9066, + "step": 10687 + }, + { + "epoch": 1.9029558404558404, + "grad_norm": 0.6517882347106934, + "learning_rate": 0.00010778869556107447, + "loss": 0.9908, + "step": 10688 + }, + { + "epoch": 1.9031339031339032, + "grad_norm": 0.6983367800712585, + "learning_rate": 0.00010777474053421315, + "loss": 1.1048, + "step": 10689 + }, + { + "epoch": 1.903311965811966, + "grad_norm": 0.597766101360321, + "learning_rate": 0.00010776078535502193, + "loss": 0.84, + "step": 10690 + }, + { + "epoch": 1.9034900284900285, + "grad_norm": 0.7335455417633057, + "learning_rate": 0.00010774683002377422, + "loss": 1.0387, + "step": 10691 + }, + { + "epoch": 1.9036680911680912, + "grad_norm": 0.6742176413536072, + "learning_rate": 0.0001077328745407435, + "loss": 0.9743, + "step": 10692 + }, + { + "epoch": 1.9038461538461537, + "grad_norm": 0.7954961657524109, + "learning_rate": 0.00010771891890620316, + "loss": 1.1025, + "step": 10693 + }, + { + "epoch": 1.9040242165242165, + "grad_norm": 0.733351469039917, + "learning_rate": 0.00010770496312042664, + "loss": 1.028, + "step": 10694 + }, + { + "epoch": 1.9042022792022792, + "grad_norm": 0.7059772610664368, + "learning_rate": 0.00010769100718368734, + "loss": 1.0103, + "step": 10695 + }, + { + "epoch": 1.904380341880342, + "grad_norm": 0.6234813332557678, + "learning_rate": 0.00010767705109625877, + "loss": 0.6893, + "step": 10696 + }, + { + "epoch": 1.9045584045584045, + "grad_norm": 0.6670311689376831, + "learning_rate": 0.0001076630948584143, + "loss": 1.1386, + "step": 10697 + }, + { + "epoch": 1.9047364672364673, + "grad_norm": 0.7444894909858704, + "learning_rate": 0.00010764913847042744, + "loss": 0.8524, + "step": 10698 + }, + { + "epoch": 1.9049145299145298, + "grad_norm": 0.6252964735031128, + "learning_rate": 0.00010763518193257158, + "loss": 0.9407, + "step": 10699 + }, + { + "epoch": 1.9050925925925926, + "grad_norm": 0.7794382572174072, + "learning_rate": 0.0001076212252451202, + "loss": 1.05, + "step": 10700 + }, + { + "epoch": 1.9052706552706553, + "grad_norm": 0.6313693523406982, + "learning_rate": 0.00010760726840834671, + "loss": 0.8667, + "step": 10701 + }, + { + "epoch": 1.905448717948718, + "grad_norm": 0.6766461730003357, + "learning_rate": 0.00010759331142252462, + "loss": 0.9675, + "step": 10702 + }, + { + "epoch": 1.9056267806267806, + "grad_norm": 0.7457365393638611, + "learning_rate": 0.00010757935428792739, + "loss": 0.9177, + "step": 10703 + }, + { + "epoch": 1.9058048433048433, + "grad_norm": 0.6649872064590454, + "learning_rate": 0.00010756539700482844, + "loss": 0.8703, + "step": 10704 + }, + { + "epoch": 1.9059829059829059, + "grad_norm": 0.8418740034103394, + "learning_rate": 0.00010755143957350127, + "loss": 0.8993, + "step": 10705 + }, + { + "epoch": 1.9061609686609686, + "grad_norm": 0.6767167448997498, + "learning_rate": 0.00010753748199421929, + "loss": 1.0063, + "step": 10706 + }, + { + "epoch": 1.9063390313390314, + "grad_norm": 0.6959242820739746, + "learning_rate": 0.00010752352426725603, + "loss": 1.0516, + "step": 10707 + }, + { + "epoch": 1.9065170940170941, + "grad_norm": 0.7106529474258423, + "learning_rate": 0.00010750956639288493, + "loss": 0.9596, + "step": 10708 + }, + { + "epoch": 1.9066951566951567, + "grad_norm": 0.7611243724822998, + "learning_rate": 0.00010749560837137949, + "loss": 1.0739, + "step": 10709 + }, + { + "epoch": 1.9068732193732194, + "grad_norm": 0.6684338450431824, + "learning_rate": 0.00010748165020301317, + "loss": 1.1437, + "step": 10710 + }, + { + "epoch": 1.907051282051282, + "grad_norm": 0.5957385897636414, + "learning_rate": 0.00010746769188805945, + "loss": 0.8802, + "step": 10711 + }, + { + "epoch": 1.9072293447293447, + "grad_norm": 0.69919353723526, + "learning_rate": 0.00010745373342679184, + "loss": 1.1891, + "step": 10712 + }, + { + "epoch": 1.9074074074074074, + "grad_norm": 0.7562127709388733, + "learning_rate": 0.0001074397748194838, + "loss": 0.8717, + "step": 10713 + }, + { + "epoch": 1.9075854700854702, + "grad_norm": 0.6420038938522339, + "learning_rate": 0.00010742581606640882, + "loss": 1.1196, + "step": 10714 + }, + { + "epoch": 1.9077635327635327, + "grad_norm": 0.7545611262321472, + "learning_rate": 0.00010741185716784039, + "loss": 1.161, + "step": 10715 + }, + { + "epoch": 1.9079415954415955, + "grad_norm": 0.6467727422714233, + "learning_rate": 0.000107397898124052, + "loss": 0.8029, + "step": 10716 + }, + { + "epoch": 1.908119658119658, + "grad_norm": 0.6129235625267029, + "learning_rate": 0.00010738393893531722, + "loss": 0.8802, + "step": 10717 + }, + { + "epoch": 1.9082977207977208, + "grad_norm": 0.6416113376617432, + "learning_rate": 0.00010736997960190946, + "loss": 0.8465, + "step": 10718 + }, + { + "epoch": 1.9084757834757835, + "grad_norm": 0.6609050631523132, + "learning_rate": 0.00010735602012410229, + "loss": 0.9484, + "step": 10719 + }, + { + "epoch": 1.9086538461538463, + "grad_norm": 0.6302639842033386, + "learning_rate": 0.00010734206050216913, + "loss": 0.898, + "step": 10720 + }, + { + "epoch": 1.9088319088319088, + "grad_norm": 0.7291215658187866, + "learning_rate": 0.00010732810073638358, + "loss": 0.9544, + "step": 10721 + }, + { + "epoch": 1.9090099715099715, + "grad_norm": 0.6436966061592102, + "learning_rate": 0.0001073141408270191, + "loss": 0.956, + "step": 10722 + }, + { + "epoch": 1.909188034188034, + "grad_norm": 0.6247875094413757, + "learning_rate": 0.00010730018077434924, + "loss": 0.8704, + "step": 10723 + }, + { + "epoch": 1.9093660968660968, + "grad_norm": 0.7599029541015625, + "learning_rate": 0.00010728622057864753, + "loss": 1.2024, + "step": 10724 + }, + { + "epoch": 1.9095441595441596, + "grad_norm": 0.6894544959068298, + "learning_rate": 0.00010727226024018744, + "loss": 1.1226, + "step": 10725 + }, + { + "epoch": 1.9097222222222223, + "grad_norm": 0.6920733451843262, + "learning_rate": 0.0001072582997592425, + "loss": 0.7682, + "step": 10726 + }, + { + "epoch": 1.9099002849002849, + "grad_norm": 0.6013005375862122, + "learning_rate": 0.00010724433913608627, + "loss": 0.9462, + "step": 10727 + }, + { + "epoch": 1.9100783475783476, + "grad_norm": 0.7466302514076233, + "learning_rate": 0.00010723037837099225, + "loss": 0.9507, + "step": 10728 + }, + { + "epoch": 1.9102564102564101, + "grad_norm": 0.7070091962814331, + "learning_rate": 0.00010721641746423401, + "loss": 1.0704, + "step": 10729 + }, + { + "epoch": 1.9104344729344729, + "grad_norm": 0.6747950315475464, + "learning_rate": 0.00010720245641608506, + "loss": 0.7899, + "step": 10730 + }, + { + "epoch": 1.9106125356125356, + "grad_norm": 0.7338371276855469, + "learning_rate": 0.00010718849522681891, + "loss": 0.9574, + "step": 10731 + }, + { + "epoch": 1.9107905982905984, + "grad_norm": 0.6923216581344604, + "learning_rate": 0.00010717453389670915, + "loss": 1.0725, + "step": 10732 + }, + { + "epoch": 1.910968660968661, + "grad_norm": 0.6050783395767212, + "learning_rate": 0.0001071605724260293, + "loss": 0.9224, + "step": 10733 + }, + { + "epoch": 1.9111467236467237, + "grad_norm": 0.6854597330093384, + "learning_rate": 0.00010714661081505291, + "loss": 0.9749, + "step": 10734 + }, + { + "epoch": 1.9113247863247862, + "grad_norm": 0.7661508321762085, + "learning_rate": 0.00010713264906405351, + "loss": 1.1564, + "step": 10735 + }, + { + "epoch": 1.911502849002849, + "grad_norm": 0.6389622688293457, + "learning_rate": 0.00010711868717330467, + "loss": 0.8148, + "step": 10736 + }, + { + "epoch": 1.9116809116809117, + "grad_norm": 0.6318161487579346, + "learning_rate": 0.00010710472514307996, + "loss": 0.7833, + "step": 10737 + }, + { + "epoch": 1.9118589743589745, + "grad_norm": 0.8646727800369263, + "learning_rate": 0.00010709076297365292, + "loss": 1.2682, + "step": 10738 + }, + { + "epoch": 1.9120370370370372, + "grad_norm": 0.6085501909255981, + "learning_rate": 0.0001070768006652971, + "loss": 0.8706, + "step": 10739 + }, + { + "epoch": 1.9122150997150997, + "grad_norm": 0.8259731531143188, + "learning_rate": 0.00010706283821828607, + "loss": 0.9014, + "step": 10740 + }, + { + "epoch": 1.9123931623931623, + "grad_norm": 0.6509148478507996, + "learning_rate": 0.0001070488756328934, + "loss": 0.8814, + "step": 10741 + }, + { + "epoch": 1.912571225071225, + "grad_norm": 0.7241966128349304, + "learning_rate": 0.00010703491290939264, + "loss": 0.9925, + "step": 10742 + }, + { + "epoch": 1.9127492877492878, + "grad_norm": 0.7736822366714478, + "learning_rate": 0.00010702095004805738, + "loss": 1.0881, + "step": 10743 + }, + { + "epoch": 1.9129273504273505, + "grad_norm": 0.6912824511528015, + "learning_rate": 0.00010700698704916123, + "loss": 1.2334, + "step": 10744 + }, + { + "epoch": 1.9131054131054133, + "grad_norm": 0.825065553188324, + "learning_rate": 0.0001069930239129777, + "loss": 0.9783, + "step": 10745 + }, + { + "epoch": 1.9132834757834758, + "grad_norm": 0.7650560140609741, + "learning_rate": 0.00010697906063978038, + "loss": 0.9788, + "step": 10746 + }, + { + "epoch": 1.9134615384615383, + "grad_norm": 0.7368232607841492, + "learning_rate": 0.00010696509722984287, + "loss": 0.8704, + "step": 10747 + }, + { + "epoch": 1.913639601139601, + "grad_norm": 0.6630628108978271, + "learning_rate": 0.00010695113368343875, + "loss": 1.1993, + "step": 10748 + }, + { + "epoch": 1.9138176638176638, + "grad_norm": 0.6842190027236938, + "learning_rate": 0.0001069371700008416, + "loss": 0.9128, + "step": 10749 + }, + { + "epoch": 1.9139957264957266, + "grad_norm": 0.591655969619751, + "learning_rate": 0.00010692320618232503, + "loss": 1.0607, + "step": 10750 + }, + { + "epoch": 1.9141737891737893, + "grad_norm": 0.74644535779953, + "learning_rate": 0.0001069092422281626, + "loss": 1.0937, + "step": 10751 + }, + { + "epoch": 1.9143518518518519, + "grad_norm": 0.7123813629150391, + "learning_rate": 0.00010689527813862792, + "loss": 0.9043, + "step": 10752 + }, + { + "epoch": 1.9145299145299144, + "grad_norm": 0.6850089430809021, + "learning_rate": 0.0001068813139139946, + "loss": 1.0908, + "step": 10753 + }, + { + "epoch": 1.9147079772079771, + "grad_norm": 0.5882078409194946, + "learning_rate": 0.00010686734955453623, + "loss": 0.829, + "step": 10754 + }, + { + "epoch": 1.91488603988604, + "grad_norm": 0.6741717457771301, + "learning_rate": 0.00010685338506052642, + "loss": 0.9197, + "step": 10755 + }, + { + "epoch": 1.9150641025641026, + "grad_norm": 0.6597354412078857, + "learning_rate": 0.00010683942043223876, + "loss": 0.8778, + "step": 10756 + }, + { + "epoch": 1.9152421652421654, + "grad_norm": 0.6682151556015015, + "learning_rate": 0.00010682545566994684, + "loss": 0.9305, + "step": 10757 + }, + { + "epoch": 1.915420227920228, + "grad_norm": 0.8283176422119141, + "learning_rate": 0.00010681149077392431, + "loss": 1.0164, + "step": 10758 + }, + { + "epoch": 1.9155982905982905, + "grad_norm": 0.648845374584198, + "learning_rate": 0.00010679752574444477, + "loss": 1.0114, + "step": 10759 + }, + { + "epoch": 1.9157763532763532, + "grad_norm": 0.755913496017456, + "learning_rate": 0.00010678356058178182, + "loss": 1.1142, + "step": 10760 + }, + { + "epoch": 1.915954415954416, + "grad_norm": 0.7334780097007751, + "learning_rate": 0.00010676959528620911, + "loss": 0.8758, + "step": 10761 + }, + { + "epoch": 1.9161324786324787, + "grad_norm": 0.9132041335105896, + "learning_rate": 0.00010675562985800025, + "loss": 0.995, + "step": 10762 + }, + { + "epoch": 1.9163105413105415, + "grad_norm": 0.7070860266685486, + "learning_rate": 0.00010674166429742882, + "loss": 0.9856, + "step": 10763 + }, + { + "epoch": 1.916488603988604, + "grad_norm": 0.7143638134002686, + "learning_rate": 0.00010672769860476853, + "loss": 1.0612, + "step": 10764 + }, + { + "epoch": 1.9166666666666665, + "grad_norm": 0.815717339515686, + "learning_rate": 0.00010671373278029293, + "loss": 1.1539, + "step": 10765 + }, + { + "epoch": 1.9168447293447293, + "grad_norm": 0.6379499435424805, + "learning_rate": 0.0001066997668242757, + "loss": 0.8295, + "step": 10766 + }, + { + "epoch": 1.917022792022792, + "grad_norm": 0.6482511758804321, + "learning_rate": 0.00010668580073699044, + "loss": 1.0079, + "step": 10767 + }, + { + "epoch": 1.9172008547008548, + "grad_norm": 0.7382873296737671, + "learning_rate": 0.00010667183451871082, + "loss": 0.8973, + "step": 10768 + }, + { + "epoch": 1.9173789173789175, + "grad_norm": 0.7818579077720642, + "learning_rate": 0.00010665786816971044, + "loss": 1.2131, + "step": 10769 + }, + { + "epoch": 1.91755698005698, + "grad_norm": 0.6960901021957397, + "learning_rate": 0.000106643901690263, + "loss": 1.1466, + "step": 10770 + }, + { + "epoch": 1.9177350427350426, + "grad_norm": 0.696966826915741, + "learning_rate": 0.00010662993508064208, + "loss": 0.854, + "step": 10771 + }, + { + "epoch": 1.9179131054131053, + "grad_norm": 0.6745442152023315, + "learning_rate": 0.00010661596834112133, + "loss": 0.9559, + "step": 10772 + }, + { + "epoch": 1.918091168091168, + "grad_norm": 0.7436230778694153, + "learning_rate": 0.00010660200147197447, + "loss": 1.1367, + "step": 10773 + }, + { + "epoch": 1.9182692307692308, + "grad_norm": 0.6051676869392395, + "learning_rate": 0.00010658803447347509, + "loss": 1.05, + "step": 10774 + }, + { + "epoch": 1.9184472934472936, + "grad_norm": 0.5662530660629272, + "learning_rate": 0.00010657406734589686, + "loss": 0.8697, + "step": 10775 + }, + { + "epoch": 1.9186253561253561, + "grad_norm": 0.6640757322311401, + "learning_rate": 0.00010656010008951344, + "loss": 1.0636, + "step": 10776 + }, + { + "epoch": 1.9188034188034186, + "grad_norm": 0.6994011998176575, + "learning_rate": 0.00010654613270459848, + "loss": 0.9326, + "step": 10777 + }, + { + "epoch": 1.9189814814814814, + "grad_norm": 0.6827420592308044, + "learning_rate": 0.00010653216519142563, + "loss": 0.8667, + "step": 10778 + }, + { + "epoch": 1.9191595441595442, + "grad_norm": 0.6814691424369812, + "learning_rate": 0.00010651819755026862, + "loss": 0.828, + "step": 10779 + }, + { + "epoch": 1.919337606837607, + "grad_norm": 0.7033611536026001, + "learning_rate": 0.00010650422978140103, + "loss": 1.0427, + "step": 10780 + }, + { + "epoch": 1.9195156695156697, + "grad_norm": 0.7098833322525024, + "learning_rate": 0.00010649026188509657, + "loss": 1.1723, + "step": 10781 + }, + { + "epoch": 1.9196937321937322, + "grad_norm": 0.7184767723083496, + "learning_rate": 0.00010647629386162893, + "loss": 0.852, + "step": 10782 + }, + { + "epoch": 1.9198717948717947, + "grad_norm": 0.6682565808296204, + "learning_rate": 0.00010646232571127175, + "loss": 0.8827, + "step": 10783 + }, + { + "epoch": 1.9200498575498575, + "grad_norm": 0.6699280142784119, + "learning_rate": 0.00010644835743429873, + "loss": 0.8346, + "step": 10784 + }, + { + "epoch": 1.9202279202279202, + "grad_norm": 0.8041857481002808, + "learning_rate": 0.00010643438903098355, + "loss": 0.9622, + "step": 10785 + }, + { + "epoch": 1.920405982905983, + "grad_norm": 0.7315110564231873, + "learning_rate": 0.00010642042050159986, + "loss": 1.0443, + "step": 10786 + }, + { + "epoch": 1.9205840455840457, + "grad_norm": 0.5850204229354858, + "learning_rate": 0.0001064064518464214, + "loss": 1.0155, + "step": 10787 + }, + { + "epoch": 1.9207621082621082, + "grad_norm": 0.7320640683174133, + "learning_rate": 0.00010639248306572178, + "loss": 1.1556, + "step": 10788 + }, + { + "epoch": 1.9209401709401708, + "grad_norm": 0.689804196357727, + "learning_rate": 0.00010637851415977478, + "loss": 1.1058, + "step": 10789 + }, + { + "epoch": 1.9211182336182335, + "grad_norm": 0.6433262228965759, + "learning_rate": 0.000106364545128854, + "loss": 1.0916, + "step": 10790 + }, + { + "epoch": 1.9212962962962963, + "grad_norm": 0.6802626252174377, + "learning_rate": 0.00010635057597323323, + "loss": 1.126, + "step": 10791 + }, + { + "epoch": 1.921474358974359, + "grad_norm": 0.7503384351730347, + "learning_rate": 0.00010633660669318608, + "loss": 0.9354, + "step": 10792 + }, + { + "epoch": 1.9216524216524218, + "grad_norm": 0.6370253562927246, + "learning_rate": 0.00010632263728898629, + "loss": 0.9976, + "step": 10793 + }, + { + "epoch": 1.9218304843304843, + "grad_norm": 0.7566042542457581, + "learning_rate": 0.00010630866776090755, + "loss": 1.0311, + "step": 10794 + }, + { + "epoch": 1.922008547008547, + "grad_norm": 0.7011943459510803, + "learning_rate": 0.0001062946981092236, + "loss": 0.8777, + "step": 10795 + }, + { + "epoch": 1.9221866096866096, + "grad_norm": 0.6621114015579224, + "learning_rate": 0.00010628072833420811, + "loss": 0.9615, + "step": 10796 + }, + { + "epoch": 1.9223646723646723, + "grad_norm": 0.6863150000572205, + "learning_rate": 0.00010626675843613478, + "loss": 1.071, + "step": 10797 + }, + { + "epoch": 1.922542735042735, + "grad_norm": 0.597970724105835, + "learning_rate": 0.00010625278841527733, + "loss": 0.8661, + "step": 10798 + }, + { + "epoch": 1.9227207977207978, + "grad_norm": 0.5958755612373352, + "learning_rate": 0.00010623881827190947, + "loss": 0.9075, + "step": 10799 + }, + { + "epoch": 1.9228988603988604, + "grad_norm": 0.7764523029327393, + "learning_rate": 0.00010622484800630494, + "loss": 1.0576, + "step": 10800 + }, + { + "epoch": 1.9230769230769231, + "grad_norm": 0.774156391620636, + "learning_rate": 0.00010621087761873748, + "loss": 0.9273, + "step": 10801 + }, + { + "epoch": 1.9232549857549857, + "grad_norm": 0.6321687698364258, + "learning_rate": 0.00010619690710948074, + "loss": 0.8805, + "step": 10802 + }, + { + "epoch": 1.9234330484330484, + "grad_norm": 0.659538984298706, + "learning_rate": 0.00010618293647880846, + "loss": 0.9845, + "step": 10803 + }, + { + "epoch": 1.9236111111111112, + "grad_norm": 0.6931299567222595, + "learning_rate": 0.00010616896572699442, + "loss": 1.2005, + "step": 10804 + }, + { + "epoch": 1.923789173789174, + "grad_norm": 0.6054762005805969, + "learning_rate": 0.00010615499485431228, + "loss": 0.825, + "step": 10805 + }, + { + "epoch": 1.9239672364672364, + "grad_norm": 0.6631526947021484, + "learning_rate": 0.00010614102386103584, + "loss": 0.9149, + "step": 10806 + }, + { + "epoch": 1.9241452991452992, + "grad_norm": 0.6667893528938293, + "learning_rate": 0.00010612705274743878, + "loss": 1.014, + "step": 10807 + }, + { + "epoch": 1.9243233618233617, + "grad_norm": 0.861302375793457, + "learning_rate": 0.00010611308151379482, + "loss": 1.0809, + "step": 10808 + }, + { + "epoch": 1.9245014245014245, + "grad_norm": 0.6997994780540466, + "learning_rate": 0.00010609911016037777, + "loss": 0.8897, + "step": 10809 + }, + { + "epoch": 1.9246794871794872, + "grad_norm": 0.5689206123352051, + "learning_rate": 0.00010608513868746131, + "loss": 0.7517, + "step": 10810 + }, + { + "epoch": 1.92485754985755, + "grad_norm": 0.5972287654876709, + "learning_rate": 0.00010607116709531918, + "loss": 0.9015, + "step": 10811 + }, + { + "epoch": 1.9250356125356125, + "grad_norm": 0.7115643620491028, + "learning_rate": 0.00010605719538422519, + "loss": 0.6974, + "step": 10812 + }, + { + "epoch": 1.9252136752136753, + "grad_norm": 0.6548098921775818, + "learning_rate": 0.00010604322355445297, + "loss": 0.7075, + "step": 10813 + }, + { + "epoch": 1.9253917378917378, + "grad_norm": 0.6666337847709656, + "learning_rate": 0.00010602925160627639, + "loss": 1.0389, + "step": 10814 + }, + { + "epoch": 1.9255698005698005, + "grad_norm": 0.7754444479942322, + "learning_rate": 0.00010601527953996913, + "loss": 1.0674, + "step": 10815 + }, + { + "epoch": 1.9257478632478633, + "grad_norm": 0.6602712869644165, + "learning_rate": 0.00010600130735580498, + "loss": 1.2622, + "step": 10816 + }, + { + "epoch": 1.925925925925926, + "grad_norm": 0.6974020004272461, + "learning_rate": 0.00010598733505405767, + "loss": 0.9748, + "step": 10817 + }, + { + "epoch": 1.9261039886039886, + "grad_norm": 0.6236271858215332, + "learning_rate": 0.00010597336263500095, + "loss": 0.9463, + "step": 10818 + }, + { + "epoch": 1.9262820512820513, + "grad_norm": 0.6856079697608948, + "learning_rate": 0.00010595939009890859, + "loss": 0.9484, + "step": 10819 + }, + { + "epoch": 1.9264601139601139, + "grad_norm": 0.7300925850868225, + "learning_rate": 0.00010594541744605437, + "loss": 0.9702, + "step": 10820 + }, + { + "epoch": 1.9266381766381766, + "grad_norm": 0.6546478867530823, + "learning_rate": 0.00010593144467671208, + "loss": 0.8235, + "step": 10821 + }, + { + "epoch": 1.9268162393162394, + "grad_norm": 0.7215169072151184, + "learning_rate": 0.00010591747179115543, + "loss": 0.9986, + "step": 10822 + }, + { + "epoch": 1.926994301994302, + "grad_norm": 0.7304712533950806, + "learning_rate": 0.00010590349878965822, + "loss": 1.099, + "step": 10823 + }, + { + "epoch": 1.9271723646723646, + "grad_norm": 0.5883305668830872, + "learning_rate": 0.0001058895256724942, + "loss": 1.0647, + "step": 10824 + }, + { + "epoch": 1.9273504273504274, + "grad_norm": 0.8067272305488586, + "learning_rate": 0.00010587555243993716, + "loss": 1.0295, + "step": 10825 + }, + { + "epoch": 1.92752849002849, + "grad_norm": 0.6607550978660583, + "learning_rate": 0.00010586157909226089, + "loss": 0.8669, + "step": 10826 + }, + { + "epoch": 1.9277065527065527, + "grad_norm": 0.7256106734275818, + "learning_rate": 0.00010584760562973914, + "loss": 1.1674, + "step": 10827 + }, + { + "epoch": 1.9278846153846154, + "grad_norm": 0.6584621071815491, + "learning_rate": 0.00010583363205264574, + "loss": 0.8901, + "step": 10828 + }, + { + "epoch": 1.9280626780626782, + "grad_norm": 0.7200617790222168, + "learning_rate": 0.00010581965836125439, + "loss": 1.0463, + "step": 10829 + }, + { + "epoch": 1.9282407407407407, + "grad_norm": 0.7244223952293396, + "learning_rate": 0.00010580568455583894, + "loss": 1.0973, + "step": 10830 + }, + { + "epoch": 1.9284188034188035, + "grad_norm": 0.7678009867668152, + "learning_rate": 0.00010579171063667317, + "loss": 1.1753, + "step": 10831 + }, + { + "epoch": 1.928596866096866, + "grad_norm": 0.6455881595611572, + "learning_rate": 0.00010577773660403085, + "loss": 0.8988, + "step": 10832 + }, + { + "epoch": 1.9287749287749287, + "grad_norm": 0.6804864406585693, + "learning_rate": 0.0001057637624581858, + "loss": 0.8156, + "step": 10833 + }, + { + "epoch": 1.9289529914529915, + "grad_norm": 0.7874828577041626, + "learning_rate": 0.00010574978819941176, + "loss": 1.1876, + "step": 10834 + }, + { + "epoch": 1.9291310541310542, + "grad_norm": 0.7396490573883057, + "learning_rate": 0.00010573581382798261, + "loss": 0.8709, + "step": 10835 + }, + { + "epoch": 1.9293091168091168, + "grad_norm": 0.6800381541252136, + "learning_rate": 0.00010572183934417209, + "loss": 0.9906, + "step": 10836 + }, + { + "epoch": 1.9294871794871795, + "grad_norm": 0.7077754139900208, + "learning_rate": 0.000105707864748254, + "loss": 0.9785, + "step": 10837 + }, + { + "epoch": 1.929665242165242, + "grad_norm": 0.693249523639679, + "learning_rate": 0.00010569389004050216, + "loss": 0.9515, + "step": 10838 + }, + { + "epoch": 1.9298433048433048, + "grad_norm": 0.706924319267273, + "learning_rate": 0.00010567991522119037, + "loss": 1.074, + "step": 10839 + }, + { + "epoch": 1.9300213675213675, + "grad_norm": 0.6504101157188416, + "learning_rate": 0.00010566594029059244, + "loss": 1.0635, + "step": 10840 + }, + { + "epoch": 1.9301994301994303, + "grad_norm": 0.7620238661766052, + "learning_rate": 0.00010565196524898219, + "loss": 0.944, + "step": 10841 + }, + { + "epoch": 1.9303774928774928, + "grad_norm": 0.6713484525680542, + "learning_rate": 0.00010563799009663344, + "loss": 0.749, + "step": 10842 + }, + { + "epoch": 1.9305555555555556, + "grad_norm": 0.9279242157936096, + "learning_rate": 0.00010562401483381997, + "loss": 0.961, + "step": 10843 + }, + { + "epoch": 1.930733618233618, + "grad_norm": 0.6710723638534546, + "learning_rate": 0.00010561003946081558, + "loss": 1.1288, + "step": 10844 + }, + { + "epoch": 1.9309116809116809, + "grad_norm": 0.7751701474189758, + "learning_rate": 0.00010559606397789416, + "loss": 0.9435, + "step": 10845 + }, + { + "epoch": 1.9310897435897436, + "grad_norm": 0.6741766929626465, + "learning_rate": 0.00010558208838532948, + "loss": 1.0299, + "step": 10846 + }, + { + "epoch": 1.9312678062678064, + "grad_norm": 0.6988041400909424, + "learning_rate": 0.00010556811268339539, + "loss": 1.0236, + "step": 10847 + }, + { + "epoch": 1.931445868945869, + "grad_norm": 0.6353505253791809, + "learning_rate": 0.00010555413687236568, + "loss": 1.0361, + "step": 10848 + }, + { + "epoch": 1.9316239316239316, + "grad_norm": 0.7162703275680542, + "learning_rate": 0.0001055401609525142, + "loss": 1.0931, + "step": 10849 + }, + { + "epoch": 1.9318019943019942, + "grad_norm": 0.61545330286026, + "learning_rate": 0.00010552618492411476, + "loss": 0.8829, + "step": 10850 + }, + { + "epoch": 1.931980056980057, + "grad_norm": 0.6304612159729004, + "learning_rate": 0.00010551220878744124, + "loss": 0.8574, + "step": 10851 + }, + { + "epoch": 1.9321581196581197, + "grad_norm": 0.6372067928314209, + "learning_rate": 0.00010549823254276743, + "loss": 1.0949, + "step": 10852 + }, + { + "epoch": 1.9323361823361824, + "grad_norm": 0.6952856779098511, + "learning_rate": 0.00010548425619036715, + "loss": 0.9232, + "step": 10853 + }, + { + "epoch": 1.9325142450142452, + "grad_norm": 0.6510106325149536, + "learning_rate": 0.00010547027973051427, + "loss": 1.0753, + "step": 10854 + }, + { + "epoch": 1.9326923076923077, + "grad_norm": 0.6377716064453125, + "learning_rate": 0.00010545630316348263, + "loss": 0.8466, + "step": 10855 + }, + { + "epoch": 1.9328703703703702, + "grad_norm": 0.7366968393325806, + "learning_rate": 0.00010544232648954606, + "loss": 0.9351, + "step": 10856 + }, + { + "epoch": 1.933048433048433, + "grad_norm": 0.703652024269104, + "learning_rate": 0.00010542834970897843, + "loss": 1.0032, + "step": 10857 + }, + { + "epoch": 1.9332264957264957, + "grad_norm": 0.6685494780540466, + "learning_rate": 0.00010541437282205355, + "loss": 0.8818, + "step": 10858 + }, + { + "epoch": 1.9334045584045585, + "grad_norm": 0.6594362854957581, + "learning_rate": 0.00010540039582904527, + "loss": 0.9535, + "step": 10859 + }, + { + "epoch": 1.9335826210826212, + "grad_norm": 0.8003259301185608, + "learning_rate": 0.00010538641873022744, + "loss": 0.8852, + "step": 10860 + }, + { + "epoch": 1.9337606837606838, + "grad_norm": 0.6567012071609497, + "learning_rate": 0.00010537244152587393, + "loss": 1.0832, + "step": 10861 + }, + { + "epoch": 1.9339387464387463, + "grad_norm": 0.6714941263198853, + "learning_rate": 0.00010535846421625862, + "loss": 1.1047, + "step": 10862 + }, + { + "epoch": 1.934116809116809, + "grad_norm": 0.6998924612998962, + "learning_rate": 0.00010534448680165531, + "loss": 0.8827, + "step": 10863 + }, + { + "epoch": 1.9342948717948718, + "grad_norm": 0.6065765619277954, + "learning_rate": 0.0001053305092823379, + "loss": 0.5773, + "step": 10864 + }, + { + "epoch": 1.9344729344729346, + "grad_norm": 0.7678273916244507, + "learning_rate": 0.0001053165316585802, + "loss": 0.9199, + "step": 10865 + }, + { + "epoch": 1.9346509971509973, + "grad_norm": 0.7071540951728821, + "learning_rate": 0.00010530255393065613, + "loss": 1.0292, + "step": 10866 + }, + { + "epoch": 1.9348290598290598, + "grad_norm": 0.6329835057258606, + "learning_rate": 0.00010528857609883956, + "loss": 0.9915, + "step": 10867 + }, + { + "epoch": 1.9350071225071224, + "grad_norm": 0.6274038553237915, + "learning_rate": 0.00010527459816340427, + "loss": 0.8499, + "step": 10868 + }, + { + "epoch": 1.9351851851851851, + "grad_norm": 0.6564371585845947, + "learning_rate": 0.00010526062012462424, + "loss": 1.1707, + "step": 10869 + }, + { + "epoch": 1.9353632478632479, + "grad_norm": 0.8561269044876099, + "learning_rate": 0.00010524664198277326, + "loss": 1.148, + "step": 10870 + }, + { + "epoch": 1.9355413105413106, + "grad_norm": 0.6322671175003052, + "learning_rate": 0.00010523266373812521, + "loss": 0.9165, + "step": 10871 + }, + { + "epoch": 1.9357193732193734, + "grad_norm": 0.7602947354316711, + "learning_rate": 0.00010521868539095403, + "loss": 0.9647, + "step": 10872 + }, + { + "epoch": 1.935897435897436, + "grad_norm": 0.5962168574333191, + "learning_rate": 0.00010520470694153353, + "loss": 0.8585, + "step": 10873 + }, + { + "epoch": 1.9360754985754984, + "grad_norm": 0.7498637437820435, + "learning_rate": 0.00010519072839013757, + "loss": 0.9828, + "step": 10874 + }, + { + "epoch": 1.9362535612535612, + "grad_norm": 0.6841256022453308, + "learning_rate": 0.00010517674973704012, + "loss": 0.9991, + "step": 10875 + }, + { + "epoch": 1.936431623931624, + "grad_norm": 0.8281826972961426, + "learning_rate": 0.00010516277098251499, + "loss": 1.028, + "step": 10876 + }, + { + "epoch": 1.9366096866096867, + "grad_norm": 0.6673563718795776, + "learning_rate": 0.0001051487921268361, + "loss": 1.1594, + "step": 10877 + }, + { + "epoch": 1.9367877492877494, + "grad_norm": 0.7833667993545532, + "learning_rate": 0.00010513481317027733, + "loss": 0.7675, + "step": 10878 + }, + { + "epoch": 1.936965811965812, + "grad_norm": 0.6087225675582886, + "learning_rate": 0.00010512083411311253, + "loss": 0.7803, + "step": 10879 + }, + { + "epoch": 1.9371438746438745, + "grad_norm": 0.6758120656013489, + "learning_rate": 0.00010510685495561563, + "loss": 1.0621, + "step": 10880 + }, + { + "epoch": 1.9373219373219372, + "grad_norm": 0.6720096468925476, + "learning_rate": 0.00010509287569806055, + "loss": 0.8502, + "step": 10881 + }, + { + "epoch": 1.9375, + "grad_norm": 0.6233887672424316, + "learning_rate": 0.00010507889634072113, + "loss": 1.0127, + "step": 10882 + }, + { + "epoch": 1.9376780626780628, + "grad_norm": 0.667742908000946, + "learning_rate": 0.00010506491688387127, + "loss": 0.9086, + "step": 10883 + }, + { + "epoch": 1.9378561253561255, + "grad_norm": 0.6533677577972412, + "learning_rate": 0.00010505093732778492, + "loss": 0.9724, + "step": 10884 + }, + { + "epoch": 1.938034188034188, + "grad_norm": 0.7171359062194824, + "learning_rate": 0.00010503695767273591, + "loss": 0.9915, + "step": 10885 + }, + { + "epoch": 1.9382122507122506, + "grad_norm": 0.723655641078949, + "learning_rate": 0.0001050229779189982, + "loss": 0.8981, + "step": 10886 + }, + { + "epoch": 1.9383903133903133, + "grad_norm": 0.6863494515419006, + "learning_rate": 0.00010500899806684568, + "loss": 1.2577, + "step": 10887 + }, + { + "epoch": 1.938568376068376, + "grad_norm": 0.8174706697463989, + "learning_rate": 0.00010499501811655224, + "loss": 0.9848, + "step": 10888 + }, + { + "epoch": 1.9387464387464388, + "grad_norm": 0.6378024220466614, + "learning_rate": 0.00010498103806839179, + "loss": 0.8499, + "step": 10889 + }, + { + "epoch": 1.9389245014245016, + "grad_norm": 0.6734544634819031, + "learning_rate": 0.00010496705792263823, + "loss": 0.8446, + "step": 10890 + }, + { + "epoch": 1.939102564102564, + "grad_norm": 0.6802361607551575, + "learning_rate": 0.00010495307767956551, + "loss": 0.9285, + "step": 10891 + }, + { + "epoch": 1.9392806267806266, + "grad_norm": 0.7821299433708191, + "learning_rate": 0.00010493909733944752, + "loss": 1.08, + "step": 10892 + }, + { + "epoch": 1.9394586894586894, + "grad_norm": 0.6204990148544312, + "learning_rate": 0.00010492511690255818, + "loss": 0.7861, + "step": 10893 + }, + { + "epoch": 1.9396367521367521, + "grad_norm": 0.6386391520500183, + "learning_rate": 0.0001049111363691714, + "loss": 0.9162, + "step": 10894 + }, + { + "epoch": 1.9398148148148149, + "grad_norm": 0.6885092854499817, + "learning_rate": 0.0001048971557395611, + "loss": 1.0026, + "step": 10895 + }, + { + "epoch": 1.9399928774928776, + "grad_norm": 0.6962558627128601, + "learning_rate": 0.00010488317501400122, + "loss": 1.146, + "step": 10896 + }, + { + "epoch": 1.9401709401709402, + "grad_norm": 0.6283716559410095, + "learning_rate": 0.00010486919419276566, + "loss": 1.0268, + "step": 10897 + }, + { + "epoch": 1.9403490028490027, + "grad_norm": 0.7183622717857361, + "learning_rate": 0.00010485521327612835, + "loss": 1.0123, + "step": 10898 + }, + { + "epoch": 1.9405270655270654, + "grad_norm": 0.6354197263717651, + "learning_rate": 0.00010484123226436321, + "loss": 0.871, + "step": 10899 + }, + { + "epoch": 1.9407051282051282, + "grad_norm": 0.804358184337616, + "learning_rate": 0.00010482725115774421, + "loss": 1.1001, + "step": 10900 + }, + { + "epoch": 1.940883190883191, + "grad_norm": 0.6896754503250122, + "learning_rate": 0.00010481326995654524, + "loss": 1.0976, + "step": 10901 + }, + { + "epoch": 1.9410612535612537, + "grad_norm": 0.9108015894889832, + "learning_rate": 0.00010479928866104023, + "loss": 0.8785, + "step": 10902 + }, + { + "epoch": 1.9412393162393162, + "grad_norm": 0.6963121294975281, + "learning_rate": 0.00010478530727150316, + "loss": 1.0458, + "step": 10903 + }, + { + "epoch": 1.9414173789173788, + "grad_norm": 0.6657114624977112, + "learning_rate": 0.00010477132578820792, + "loss": 0.8188, + "step": 10904 + }, + { + "epoch": 1.9415954415954415, + "grad_norm": 0.671716034412384, + "learning_rate": 0.00010475734421142847, + "loss": 1.0915, + "step": 10905 + }, + { + "epoch": 1.9417735042735043, + "grad_norm": 0.6790717244148254, + "learning_rate": 0.0001047433625414387, + "loss": 0.9688, + "step": 10906 + }, + { + "epoch": 1.941951566951567, + "grad_norm": 0.6411764621734619, + "learning_rate": 0.00010472938077851264, + "loss": 1.0387, + "step": 10907 + }, + { + "epoch": 1.9421296296296298, + "grad_norm": 0.8579615950584412, + "learning_rate": 0.00010471539892292417, + "loss": 1.1635, + "step": 10908 + }, + { + "epoch": 1.9423076923076923, + "grad_norm": 0.7031029462814331, + "learning_rate": 0.00010470141697494726, + "loss": 0.9813, + "step": 10909 + }, + { + "epoch": 1.9424857549857548, + "grad_norm": 0.6657388806343079, + "learning_rate": 0.00010468743493485584, + "loss": 0.7947, + "step": 10910 + }, + { + "epoch": 1.9426638176638176, + "grad_norm": 0.6364194750785828, + "learning_rate": 0.00010467345280292389, + "loss": 0.8554, + "step": 10911 + }, + { + "epoch": 1.9428418803418803, + "grad_norm": 0.7394127249717712, + "learning_rate": 0.00010465947057942534, + "loss": 0.822, + "step": 10912 + }, + { + "epoch": 1.943019943019943, + "grad_norm": 0.6557473540306091, + "learning_rate": 0.00010464548826463411, + "loss": 1.0025, + "step": 10913 + }, + { + "epoch": 1.9431980056980058, + "grad_norm": 0.6530601382255554, + "learning_rate": 0.00010463150585882422, + "loss": 1.0828, + "step": 10914 + }, + { + "epoch": 1.9433760683760684, + "grad_norm": 0.7376404404640198, + "learning_rate": 0.00010461752336226957, + "loss": 0.9413, + "step": 10915 + }, + { + "epoch": 1.943554131054131, + "grad_norm": 0.7110656499862671, + "learning_rate": 0.00010460354077524417, + "loss": 0.9162, + "step": 10916 + }, + { + "epoch": 1.9437321937321936, + "grad_norm": 0.6515666246414185, + "learning_rate": 0.00010458955809802194, + "loss": 0.9211, + "step": 10917 + }, + { + "epoch": 1.9439102564102564, + "grad_norm": 0.6888720989227295, + "learning_rate": 0.00010457557533087683, + "loss": 1.0632, + "step": 10918 + }, + { + "epoch": 1.9440883190883191, + "grad_norm": 0.7246627807617188, + "learning_rate": 0.00010456159247408286, + "loss": 0.9807, + "step": 10919 + }, + { + "epoch": 1.944266381766382, + "grad_norm": 0.727834165096283, + "learning_rate": 0.00010454760952791394, + "loss": 1.0793, + "step": 10920 + }, + { + "epoch": 1.9444444444444444, + "grad_norm": 0.6365306377410889, + "learning_rate": 0.00010453362649264407, + "loss": 1.0415, + "step": 10921 + }, + { + "epoch": 1.9446225071225072, + "grad_norm": 0.7187839150428772, + "learning_rate": 0.0001045196433685472, + "loss": 1.007, + "step": 10922 + }, + { + "epoch": 1.9448005698005697, + "grad_norm": 0.5905138254165649, + "learning_rate": 0.00010450566015589732, + "loss": 0.9818, + "step": 10923 + }, + { + "epoch": 1.9449786324786325, + "grad_norm": 0.7008894085884094, + "learning_rate": 0.00010449167685496837, + "loss": 0.8444, + "step": 10924 + }, + { + "epoch": 1.9451566951566952, + "grad_norm": 0.6126312017440796, + "learning_rate": 0.00010447769346603435, + "loss": 0.7207, + "step": 10925 + }, + { + "epoch": 1.945334757834758, + "grad_norm": 0.7513176202774048, + "learning_rate": 0.00010446370998936922, + "loss": 0.8693, + "step": 10926 + }, + { + "epoch": 1.9455128205128205, + "grad_norm": 0.6382531523704529, + "learning_rate": 0.00010444972642524697, + "loss": 0.8379, + "step": 10927 + }, + { + "epoch": 1.9456908831908832, + "grad_norm": 0.7062170505523682, + "learning_rate": 0.0001044357427739416, + "loss": 1.0525, + "step": 10928 + }, + { + "epoch": 1.9458689458689458, + "grad_norm": 0.6954067349433899, + "learning_rate": 0.00010442175903572703, + "loss": 1.0238, + "step": 10929 + }, + { + "epoch": 1.9460470085470085, + "grad_norm": 0.7257117033004761, + "learning_rate": 0.00010440777521087731, + "loss": 1.1413, + "step": 10930 + }, + { + "epoch": 1.9462250712250713, + "grad_norm": 0.6617701053619385, + "learning_rate": 0.00010439379129966635, + "loss": 1.0089, + "step": 10931 + }, + { + "epoch": 1.946403133903134, + "grad_norm": 0.6860800385475159, + "learning_rate": 0.00010437980730236821, + "loss": 1.1778, + "step": 10932 + }, + { + "epoch": 1.9465811965811965, + "grad_norm": 0.846235454082489, + "learning_rate": 0.00010436582321925684, + "loss": 0.9851, + "step": 10933 + }, + { + "epoch": 1.9467592592592593, + "grad_norm": 0.6385617852210999, + "learning_rate": 0.00010435183905060623, + "loss": 0.9542, + "step": 10934 + }, + { + "epoch": 1.9469373219373218, + "grad_norm": 0.7137401700019836, + "learning_rate": 0.00010433785479669038, + "loss": 1.0499, + "step": 10935 + }, + { + "epoch": 1.9471153846153846, + "grad_norm": 0.6269308924674988, + "learning_rate": 0.00010432387045778324, + "loss": 0.8929, + "step": 10936 + }, + { + "epoch": 1.9472934472934473, + "grad_norm": 0.7903163433074951, + "learning_rate": 0.00010430988603415888, + "loss": 0.9812, + "step": 10937 + }, + { + "epoch": 1.94747150997151, + "grad_norm": 0.6006736159324646, + "learning_rate": 0.00010429590152609121, + "loss": 0.7959, + "step": 10938 + }, + { + "epoch": 1.9476495726495726, + "grad_norm": 0.6061521768569946, + "learning_rate": 0.00010428191693385431, + "loss": 0.8748, + "step": 10939 + }, + { + "epoch": 1.9478276353276354, + "grad_norm": 0.6637623906135559, + "learning_rate": 0.00010426793225772216, + "loss": 0.7047, + "step": 10940 + }, + { + "epoch": 1.948005698005698, + "grad_norm": 0.7650586366653442, + "learning_rate": 0.00010425394749796874, + "loss": 1.0018, + "step": 10941 + }, + { + "epoch": 1.9481837606837606, + "grad_norm": 0.6575125455856323, + "learning_rate": 0.000104239962654868, + "loss": 0.8915, + "step": 10942 + }, + { + "epoch": 1.9483618233618234, + "grad_norm": 0.6315393447875977, + "learning_rate": 0.00010422597772869404, + "loss": 1.1884, + "step": 10943 + }, + { + "epoch": 1.9485398860398861, + "grad_norm": 0.7607148885726929, + "learning_rate": 0.00010421199271972083, + "loss": 0.9341, + "step": 10944 + }, + { + "epoch": 1.9487179487179487, + "grad_norm": 0.6491827964782715, + "learning_rate": 0.00010419800762822239, + "loss": 0.9991, + "step": 10945 + }, + { + "epoch": 1.9488960113960114, + "grad_norm": 0.6294243335723877, + "learning_rate": 0.00010418402245447265, + "loss": 0.9253, + "step": 10946 + }, + { + "epoch": 1.949074074074074, + "grad_norm": 0.6472215056419373, + "learning_rate": 0.00010417003719874571, + "loss": 1.0402, + "step": 10947 + }, + { + "epoch": 1.9492521367521367, + "grad_norm": 0.7377899885177612, + "learning_rate": 0.00010415605186131559, + "loss": 1.046, + "step": 10948 + }, + { + "epoch": 1.9494301994301995, + "grad_norm": 0.6391907334327698, + "learning_rate": 0.00010414206644245623, + "loss": 0.8529, + "step": 10949 + }, + { + "epoch": 1.9496082621082622, + "grad_norm": 0.7101355195045471, + "learning_rate": 0.0001041280809424417, + "loss": 0.925, + "step": 10950 + }, + { + "epoch": 1.9497863247863247, + "grad_norm": 0.7891978025436401, + "learning_rate": 0.00010411409536154597, + "loss": 1.0691, + "step": 10951 + }, + { + "epoch": 1.9499643874643875, + "grad_norm": 0.7225242853164673, + "learning_rate": 0.00010410010970004311, + "loss": 1.158, + "step": 10952 + }, + { + "epoch": 1.95014245014245, + "grad_norm": 0.6073256731033325, + "learning_rate": 0.00010408612395820714, + "loss": 0.9977, + "step": 10953 + }, + { + "epoch": 1.9503205128205128, + "grad_norm": 0.6373769044876099, + "learning_rate": 0.00010407213813631203, + "loss": 1.019, + "step": 10954 + }, + { + "epoch": 1.9504985754985755, + "grad_norm": 0.7451884746551514, + "learning_rate": 0.00010405815223463184, + "loss": 0.9497, + "step": 10955 + }, + { + "epoch": 1.9506766381766383, + "grad_norm": 0.7760418057441711, + "learning_rate": 0.00010404416625344058, + "loss": 1.0378, + "step": 10956 + }, + { + "epoch": 1.9508547008547008, + "grad_norm": 0.7057808041572571, + "learning_rate": 0.00010403018019301228, + "loss": 0.8953, + "step": 10957 + }, + { + "epoch": 1.9510327635327636, + "grad_norm": 0.6599584817886353, + "learning_rate": 0.00010401619405362095, + "loss": 0.8859, + "step": 10958 + }, + { + "epoch": 1.951210826210826, + "grad_norm": 0.6977253556251526, + "learning_rate": 0.00010400220783554069, + "loss": 0.9038, + "step": 10959 + }, + { + "epoch": 1.9513888888888888, + "grad_norm": 0.6930267810821533, + "learning_rate": 0.00010398822153904546, + "loss": 1.1547, + "step": 10960 + }, + { + "epoch": 1.9515669515669516, + "grad_norm": 0.6301694512367249, + "learning_rate": 0.00010397423516440931, + "loss": 0.8875, + "step": 10961 + }, + { + "epoch": 1.9517450142450143, + "grad_norm": 0.7447484135627747, + "learning_rate": 0.00010396024871190628, + "loss": 1.0454, + "step": 10962 + }, + { + "epoch": 1.9519230769230769, + "grad_norm": 0.8666765093803406, + "learning_rate": 0.00010394626218181041, + "loss": 1.2211, + "step": 10963 + }, + { + "epoch": 1.9521011396011396, + "grad_norm": 0.599354088306427, + "learning_rate": 0.00010393227557439573, + "loss": 1.0419, + "step": 10964 + }, + { + "epoch": 1.9522792022792022, + "grad_norm": 0.6991702914237976, + "learning_rate": 0.00010391828888993627, + "loss": 0.8217, + "step": 10965 + }, + { + "epoch": 1.952457264957265, + "grad_norm": 0.7467028498649597, + "learning_rate": 0.0001039043021287061, + "loss": 0.8708, + "step": 10966 + }, + { + "epoch": 1.9526353276353277, + "grad_norm": 0.6806215047836304, + "learning_rate": 0.0001038903152909792, + "loss": 1.218, + "step": 10967 + }, + { + "epoch": 1.9528133903133904, + "grad_norm": 0.6704212427139282, + "learning_rate": 0.00010387632837702968, + "loss": 0.8428, + "step": 10968 + }, + { + "epoch": 1.952991452991453, + "grad_norm": 0.6843154430389404, + "learning_rate": 0.00010386234138713155, + "loss": 0.9729, + "step": 10969 + }, + { + "epoch": 1.9531695156695157, + "grad_norm": 0.6619821190834045, + "learning_rate": 0.00010384835432155888, + "loss": 1.021, + "step": 10970 + }, + { + "epoch": 1.9533475783475782, + "grad_norm": 0.6249803900718689, + "learning_rate": 0.0001038343671805857, + "loss": 0.9321, + "step": 10971 + }, + { + "epoch": 1.953525641025641, + "grad_norm": 0.7361689805984497, + "learning_rate": 0.00010382037996448604, + "loss": 0.9451, + "step": 10972 + }, + { + "epoch": 1.9537037037037037, + "grad_norm": 0.6464847922325134, + "learning_rate": 0.00010380639267353398, + "loss": 1.0188, + "step": 10973 + }, + { + "epoch": 1.9538817663817665, + "grad_norm": 0.5975635647773743, + "learning_rate": 0.00010379240530800356, + "loss": 0.9025, + "step": 10974 + }, + { + "epoch": 1.9540598290598292, + "grad_norm": 0.6734475493431091, + "learning_rate": 0.00010377841786816884, + "loss": 1.0742, + "step": 10975 + }, + { + "epoch": 1.9542378917378918, + "grad_norm": 0.7318592667579651, + "learning_rate": 0.00010376443035430386, + "loss": 1.1082, + "step": 10976 + }, + { + "epoch": 1.9544159544159543, + "grad_norm": 0.7696142792701721, + "learning_rate": 0.00010375044276668271, + "loss": 0.8421, + "step": 10977 + }, + { + "epoch": 1.954594017094017, + "grad_norm": 0.68442302942276, + "learning_rate": 0.00010373645510557939, + "loss": 1.0794, + "step": 10978 + }, + { + "epoch": 1.9547720797720798, + "grad_norm": 0.7582547068595886, + "learning_rate": 0.00010372246737126801, + "loss": 1.0332, + "step": 10979 + }, + { + "epoch": 1.9549501424501425, + "grad_norm": 0.6529998183250427, + "learning_rate": 0.00010370847956402262, + "loss": 1.1833, + "step": 10980 + }, + { + "epoch": 1.9551282051282053, + "grad_norm": 0.7565605044364929, + "learning_rate": 0.00010369449168411729, + "loss": 1.0494, + "step": 10981 + }, + { + "epoch": 1.9553062678062678, + "grad_norm": 0.6346915364265442, + "learning_rate": 0.00010368050373182605, + "loss": 1.0052, + "step": 10982 + }, + { + "epoch": 1.9554843304843303, + "grad_norm": 0.7021830081939697, + "learning_rate": 0.00010366651570742298, + "loss": 0.9716, + "step": 10983 + }, + { + "epoch": 1.955662393162393, + "grad_norm": 0.6464530825614929, + "learning_rate": 0.00010365252761118218, + "loss": 0.9802, + "step": 10984 + }, + { + "epoch": 1.9558404558404558, + "grad_norm": 0.6845090985298157, + "learning_rate": 0.00010363853944337768, + "loss": 0.9529, + "step": 10985 + }, + { + "epoch": 1.9560185185185186, + "grad_norm": 0.7178115248680115, + "learning_rate": 0.00010362455120428356, + "loss": 0.9968, + "step": 10986 + }, + { + "epoch": 1.9561965811965814, + "grad_norm": 0.6131038069725037, + "learning_rate": 0.00010361056289417385, + "loss": 1.0559, + "step": 10987 + }, + { + "epoch": 1.9563746438746439, + "grad_norm": 0.6946909427642822, + "learning_rate": 0.0001035965745133227, + "loss": 1.0457, + "step": 10988 + }, + { + "epoch": 1.9565527065527064, + "grad_norm": 0.7376706600189209, + "learning_rate": 0.00010358258606200413, + "loss": 0.7775, + "step": 10989 + }, + { + "epoch": 1.9567307692307692, + "grad_norm": 0.6864920854568481, + "learning_rate": 0.00010356859754049225, + "loss": 0.8798, + "step": 10990 + }, + { + "epoch": 1.956908831908832, + "grad_norm": 0.6301153302192688, + "learning_rate": 0.0001035546089490611, + "loss": 0.8757, + "step": 10991 + }, + { + "epoch": 1.9570868945868947, + "grad_norm": 0.7184807062149048, + "learning_rate": 0.00010354062028798474, + "loss": 1.0783, + "step": 10992 + }, + { + "epoch": 1.9572649572649574, + "grad_norm": 0.7138563394546509, + "learning_rate": 0.00010352663155753732, + "loss": 1.0328, + "step": 10993 + }, + { + "epoch": 1.95744301994302, + "grad_norm": 0.6565547585487366, + "learning_rate": 0.00010351264275799286, + "loss": 1.1312, + "step": 10994 + }, + { + "epoch": 1.9576210826210825, + "grad_norm": 0.7055862545967102, + "learning_rate": 0.00010349865388962547, + "loss": 1.0787, + "step": 10995 + }, + { + "epoch": 1.9577991452991452, + "grad_norm": 0.6184022426605225, + "learning_rate": 0.00010348466495270926, + "loss": 0.9635, + "step": 10996 + }, + { + "epoch": 1.957977207977208, + "grad_norm": 0.6563652753829956, + "learning_rate": 0.0001034706759475182, + "loss": 0.772, + "step": 10997 + }, + { + "epoch": 1.9581552706552707, + "grad_norm": 0.6103591322898865, + "learning_rate": 0.00010345668687432651, + "loss": 0.8113, + "step": 10998 + }, + { + "epoch": 1.9583333333333335, + "grad_norm": 0.6715512275695801, + "learning_rate": 0.0001034426977334082, + "loss": 1.1841, + "step": 10999 + }, + { + "epoch": 1.958511396011396, + "grad_norm": 0.680092453956604, + "learning_rate": 0.00010342870852503739, + "loss": 0.9992, + "step": 11000 + }, + { + "epoch": 1.9586894586894585, + "grad_norm": 0.828472375869751, + "learning_rate": 0.00010341471924948816, + "loss": 1.0975, + "step": 11001 + }, + { + "epoch": 1.9588675213675213, + "grad_norm": 0.758441686630249, + "learning_rate": 0.00010340072990703463, + "loss": 1.0632, + "step": 11002 + }, + { + "epoch": 1.959045584045584, + "grad_norm": 0.6847560405731201, + "learning_rate": 0.00010338674049795079, + "loss": 1.0054, + "step": 11003 + }, + { + "epoch": 1.9592236467236468, + "grad_norm": 0.707626223564148, + "learning_rate": 0.00010337275102251085, + "loss": 0.9427, + "step": 11004 + }, + { + "epoch": 1.9594017094017095, + "grad_norm": 0.769036591053009, + "learning_rate": 0.00010335876148098887, + "loss": 1.0424, + "step": 11005 + }, + { + "epoch": 1.959579772079772, + "grad_norm": 0.822695791721344, + "learning_rate": 0.00010334477187365892, + "loss": 1.1573, + "step": 11006 + }, + { + "epoch": 1.9597578347578346, + "grad_norm": 0.6290286183357239, + "learning_rate": 0.00010333078220079513, + "loss": 0.936, + "step": 11007 + }, + { + "epoch": 1.9599358974358974, + "grad_norm": 0.6802252531051636, + "learning_rate": 0.00010331679246267155, + "loss": 0.8049, + "step": 11008 + }, + { + "epoch": 1.96011396011396, + "grad_norm": 0.6652607321739197, + "learning_rate": 0.00010330280265956232, + "loss": 0.926, + "step": 11009 + }, + { + "epoch": 1.9602920227920229, + "grad_norm": 0.7057216763496399, + "learning_rate": 0.00010328881279174154, + "loss": 0.9464, + "step": 11010 + }, + { + "epoch": 1.9604700854700856, + "grad_norm": 0.6951601505279541, + "learning_rate": 0.00010327482285948331, + "loss": 0.9882, + "step": 11011 + }, + { + "epoch": 1.9606481481481481, + "grad_norm": 0.6537632942199707, + "learning_rate": 0.00010326083286306174, + "loss": 0.8663, + "step": 11012 + }, + { + "epoch": 1.9608262108262107, + "grad_norm": 0.7252047657966614, + "learning_rate": 0.0001032468428027509, + "loss": 1.1377, + "step": 11013 + }, + { + "epoch": 1.9610042735042734, + "grad_norm": 0.6494104266166687, + "learning_rate": 0.00010323285267882492, + "loss": 0.8072, + "step": 11014 + }, + { + "epoch": 1.9611823361823362, + "grad_norm": 0.8463460206985474, + "learning_rate": 0.00010321886249155792, + "loss": 1.22, + "step": 11015 + }, + { + "epoch": 1.961360398860399, + "grad_norm": 0.6071396470069885, + "learning_rate": 0.00010320487224122401, + "loss": 0.7975, + "step": 11016 + }, + { + "epoch": 1.9615384615384617, + "grad_norm": 0.6546960473060608, + "learning_rate": 0.00010319088192809725, + "loss": 1.1729, + "step": 11017 + }, + { + "epoch": 1.9617165242165242, + "grad_norm": 0.7399442791938782, + "learning_rate": 0.00010317689155245178, + "loss": 1.092, + "step": 11018 + }, + { + "epoch": 1.9618945868945867, + "grad_norm": 0.7103837728500366, + "learning_rate": 0.00010316290111456175, + "loss": 0.8436, + "step": 11019 + }, + { + "epoch": 1.9620726495726495, + "grad_norm": 0.6990065574645996, + "learning_rate": 0.00010314891061470125, + "loss": 0.9003, + "step": 11020 + }, + { + "epoch": 1.9622507122507122, + "grad_norm": 0.7945666313171387, + "learning_rate": 0.00010313492005314438, + "loss": 0.8812, + "step": 11021 + }, + { + "epoch": 1.962428774928775, + "grad_norm": 0.6177538633346558, + "learning_rate": 0.00010312092943016527, + "loss": 1.0091, + "step": 11022 + }, + { + "epoch": 1.9626068376068377, + "grad_norm": 0.7260771989822388, + "learning_rate": 0.000103106938746038, + "loss": 0.9376, + "step": 11023 + }, + { + "epoch": 1.9627849002849003, + "grad_norm": 0.6726518273353577, + "learning_rate": 0.00010309294800103674, + "loss": 0.8048, + "step": 11024 + }, + { + "epoch": 1.9629629629629628, + "grad_norm": 0.8759992122650146, + "learning_rate": 0.00010307895719543562, + "loss": 1.0248, + "step": 11025 + }, + { + "epoch": 1.9631410256410255, + "grad_norm": 0.683437168598175, + "learning_rate": 0.00010306496632950868, + "loss": 1.0314, + "step": 11026 + }, + { + "epoch": 1.9633190883190883, + "grad_norm": 0.7255756258964539, + "learning_rate": 0.00010305097540353012, + "loss": 0.9828, + "step": 11027 + }, + { + "epoch": 1.963497150997151, + "grad_norm": 0.6904804706573486, + "learning_rate": 0.000103036984417774, + "loss": 0.9054, + "step": 11028 + }, + { + "epoch": 1.9636752136752138, + "grad_norm": 0.6906846761703491, + "learning_rate": 0.00010302299337251451, + "loss": 1.0287, + "step": 11029 + }, + { + "epoch": 1.9638532763532763, + "grad_norm": 0.6677078008651733, + "learning_rate": 0.00010300900226802575, + "loss": 0.8742, + "step": 11030 + }, + { + "epoch": 1.964031339031339, + "grad_norm": 0.6144888997077942, + "learning_rate": 0.00010299501110458183, + "loss": 0.6942, + "step": 11031 + }, + { + "epoch": 1.9642094017094016, + "grad_norm": 0.753010094165802, + "learning_rate": 0.0001029810198824569, + "loss": 0.9018, + "step": 11032 + }, + { + "epoch": 1.9643874643874644, + "grad_norm": 0.6872276663780212, + "learning_rate": 0.00010296702860192505, + "loss": 1.1647, + "step": 11033 + }, + { + "epoch": 1.9645655270655271, + "grad_norm": 0.709000289440155, + "learning_rate": 0.00010295303726326047, + "loss": 0.9143, + "step": 11034 + }, + { + "epoch": 1.9647435897435899, + "grad_norm": 0.6507021188735962, + "learning_rate": 0.00010293904586673723, + "loss": 1.006, + "step": 11035 + }, + { + "epoch": 1.9649216524216524, + "grad_norm": 0.6789946556091309, + "learning_rate": 0.00010292505441262952, + "loss": 0.9049, + "step": 11036 + }, + { + "epoch": 1.9650997150997151, + "grad_norm": 0.7156081795692444, + "learning_rate": 0.00010291106290121143, + "loss": 0.9195, + "step": 11037 + }, + { + "epoch": 1.9652777777777777, + "grad_norm": 0.6770932078361511, + "learning_rate": 0.0001028970713327571, + "loss": 0.9524, + "step": 11038 + }, + { + "epoch": 1.9654558404558404, + "grad_norm": 0.7304288148880005, + "learning_rate": 0.00010288307970754067, + "loss": 0.9276, + "step": 11039 + }, + { + "epoch": 1.9656339031339032, + "grad_norm": 0.7603645324707031, + "learning_rate": 0.0001028690880258363, + "loss": 1.2157, + "step": 11040 + }, + { + "epoch": 1.965811965811966, + "grad_norm": 0.6875246167182922, + "learning_rate": 0.00010285509628791811, + "loss": 1.0269, + "step": 11041 + }, + { + "epoch": 1.9659900284900285, + "grad_norm": 0.7234818935394287, + "learning_rate": 0.00010284110449406026, + "loss": 0.9695, + "step": 11042 + }, + { + "epoch": 1.9661680911680912, + "grad_norm": 0.7322804927825928, + "learning_rate": 0.00010282711264453684, + "loss": 0.9752, + "step": 11043 + }, + { + "epoch": 1.9663461538461537, + "grad_norm": 0.7524822950363159, + "learning_rate": 0.00010281312073962202, + "loss": 1.2144, + "step": 11044 + }, + { + "epoch": 1.9665242165242165, + "grad_norm": 0.6623101234436035, + "learning_rate": 0.00010279912877958995, + "loss": 1.1334, + "step": 11045 + }, + { + "epoch": 1.9667022792022792, + "grad_norm": 0.7814893126487732, + "learning_rate": 0.00010278513676471477, + "loss": 1.266, + "step": 11046 + }, + { + "epoch": 1.966880341880342, + "grad_norm": 0.7129884362220764, + "learning_rate": 0.00010277114469527063, + "loss": 1.0918, + "step": 11047 + }, + { + "epoch": 1.9670584045584045, + "grad_norm": 0.6996828317642212, + "learning_rate": 0.00010275715257153164, + "loss": 0.9269, + "step": 11048 + }, + { + "epoch": 1.9672364672364673, + "grad_norm": 0.6439059972763062, + "learning_rate": 0.00010274316039377198, + "loss": 1.1998, + "step": 11049 + }, + { + "epoch": 1.9674145299145298, + "grad_norm": 0.6837672591209412, + "learning_rate": 0.00010272916816226581, + "loss": 0.8899, + "step": 11050 + }, + { + "epoch": 1.9675925925925926, + "grad_norm": 0.702583909034729, + "learning_rate": 0.00010271517587728726, + "loss": 1.1862, + "step": 11051 + }, + { + "epoch": 1.9677706552706553, + "grad_norm": 0.6627798676490784, + "learning_rate": 0.00010270118353911047, + "loss": 0.898, + "step": 11052 + }, + { + "epoch": 1.967948717948718, + "grad_norm": 0.7628579139709473, + "learning_rate": 0.00010268719114800957, + "loss": 1.006, + "step": 11053 + }, + { + "epoch": 1.9681267806267806, + "grad_norm": 0.6425395607948303, + "learning_rate": 0.00010267319870425877, + "loss": 0.962, + "step": 11054 + }, + { + "epoch": 1.9683048433048433, + "grad_norm": 0.7462666630744934, + "learning_rate": 0.00010265920620813219, + "loss": 1.0703, + "step": 11055 + }, + { + "epoch": 1.9684829059829059, + "grad_norm": 0.67641681432724, + "learning_rate": 0.00010264521365990401, + "loss": 1.1077, + "step": 11056 + }, + { + "epoch": 1.9686609686609686, + "grad_norm": 0.6716381311416626, + "learning_rate": 0.0001026312210598483, + "loss": 1.1048, + "step": 11057 + }, + { + "epoch": 1.9688390313390314, + "grad_norm": 0.7207448482513428, + "learning_rate": 0.00010261722840823935, + "loss": 0.9236, + "step": 11058 + }, + { + "epoch": 1.9690170940170941, + "grad_norm": 0.7208544015884399, + "learning_rate": 0.0001026032357053512, + "loss": 1.0814, + "step": 11059 + }, + { + "epoch": 1.9691951566951567, + "grad_norm": 0.6076363325119019, + "learning_rate": 0.00010258924295145807, + "loss": 0.9388, + "step": 11060 + }, + { + "epoch": 1.9693732193732194, + "grad_norm": 0.6460439562797546, + "learning_rate": 0.00010257525014683411, + "loss": 0.9506, + "step": 11061 + }, + { + "epoch": 1.969551282051282, + "grad_norm": 0.7449939250946045, + "learning_rate": 0.00010256125729175348, + "loss": 1.0209, + "step": 11062 + }, + { + "epoch": 1.9697293447293447, + "grad_norm": 0.640885055065155, + "learning_rate": 0.00010254726438649031, + "loss": 1.0235, + "step": 11063 + }, + { + "epoch": 1.9699074074074074, + "grad_norm": 0.6872261166572571, + "learning_rate": 0.00010253327143131879, + "loss": 0.9217, + "step": 11064 + }, + { + "epoch": 1.9700854700854702, + "grad_norm": 0.6213285326957703, + "learning_rate": 0.0001025192784265131, + "loss": 0.8204, + "step": 11065 + }, + { + "epoch": 1.9702635327635327, + "grad_norm": 0.6594449281692505, + "learning_rate": 0.00010250528537234736, + "loss": 0.9789, + "step": 11066 + }, + { + "epoch": 1.9704415954415955, + "grad_norm": 0.7098729610443115, + "learning_rate": 0.00010249129226909577, + "loss": 1.2551, + "step": 11067 + }, + { + "epoch": 1.970619658119658, + "grad_norm": 0.7455953359603882, + "learning_rate": 0.0001024772991170325, + "loss": 1.0281, + "step": 11068 + }, + { + "epoch": 1.9707977207977208, + "grad_norm": 0.6657416224479675, + "learning_rate": 0.00010246330591643166, + "loss": 0.9421, + "step": 11069 + }, + { + "epoch": 1.9709757834757835, + "grad_norm": 0.6480659246444702, + "learning_rate": 0.00010244931266756748, + "loss": 0.9424, + "step": 11070 + }, + { + "epoch": 1.9711538461538463, + "grad_norm": 0.6440510749816895, + "learning_rate": 0.00010243531937071411, + "loss": 0.9651, + "step": 11071 + }, + { + "epoch": 1.9713319088319088, + "grad_norm": 0.6329794526100159, + "learning_rate": 0.00010242132602614571, + "loss": 0.9233, + "step": 11072 + }, + { + "epoch": 1.9715099715099715, + "grad_norm": 0.6694819927215576, + "learning_rate": 0.00010240733263413646, + "loss": 0.884, + "step": 11073 + }, + { + "epoch": 1.971688034188034, + "grad_norm": 0.7702556848526001, + "learning_rate": 0.0001023933391949605, + "loss": 1.216, + "step": 11074 + }, + { + "epoch": 1.9718660968660968, + "grad_norm": 0.6587536931037903, + "learning_rate": 0.00010237934570889207, + "loss": 0.9324, + "step": 11075 + }, + { + "epoch": 1.9720441595441596, + "grad_norm": 0.7919837832450867, + "learning_rate": 0.00010236535217620529, + "loss": 1.0011, + "step": 11076 + }, + { + "epoch": 1.9722222222222223, + "grad_norm": 0.6604606509208679, + "learning_rate": 0.00010235135859717433, + "loss": 0.929, + "step": 11077 + }, + { + "epoch": 1.9724002849002849, + "grad_norm": 0.7158446907997131, + "learning_rate": 0.0001023373649720734, + "loss": 0.8912, + "step": 11078 + }, + { + "epoch": 1.9725783475783476, + "grad_norm": 0.7450904846191406, + "learning_rate": 0.00010232337130117666, + "loss": 1.0782, + "step": 11079 + }, + { + "epoch": 1.9727564102564101, + "grad_norm": 0.6687077283859253, + "learning_rate": 0.00010230937758475827, + "loss": 1.0662, + "step": 11080 + }, + { + "epoch": 1.9729344729344729, + "grad_norm": 0.7188364267349243, + "learning_rate": 0.00010229538382309245, + "loss": 1.024, + "step": 11081 + }, + { + "epoch": 1.9731125356125356, + "grad_norm": 0.6787814497947693, + "learning_rate": 0.00010228139001645334, + "loss": 0.9559, + "step": 11082 + }, + { + "epoch": 1.9732905982905984, + "grad_norm": 0.6834072470664978, + "learning_rate": 0.00010226739616511513, + "loss": 0.8143, + "step": 11083 + }, + { + "epoch": 1.973468660968661, + "grad_norm": 0.6651090979576111, + "learning_rate": 0.00010225340226935201, + "loss": 1.05, + "step": 11084 + }, + { + "epoch": 1.9736467236467237, + "grad_norm": 0.7125018835067749, + "learning_rate": 0.00010223940832943813, + "loss": 1.0275, + "step": 11085 + }, + { + "epoch": 1.9738247863247862, + "grad_norm": 0.6886870861053467, + "learning_rate": 0.00010222541434564772, + "loss": 1.0972, + "step": 11086 + }, + { + "epoch": 1.974002849002849, + "grad_norm": 0.7068913578987122, + "learning_rate": 0.00010221142031825492, + "loss": 0.9248, + "step": 11087 + }, + { + "epoch": 1.9741809116809117, + "grad_norm": 0.7752319574356079, + "learning_rate": 0.00010219742624753397, + "loss": 0.9754, + "step": 11088 + }, + { + "epoch": 1.9743589743589745, + "grad_norm": 0.7915459871292114, + "learning_rate": 0.00010218343213375896, + "loss": 1.2589, + "step": 11089 + }, + { + "epoch": 1.9745370370370372, + "grad_norm": 0.6597068309783936, + "learning_rate": 0.00010216943797720418, + "loss": 1.0004, + "step": 11090 + }, + { + "epoch": 1.9747150997150997, + "grad_norm": 0.7060620188713074, + "learning_rate": 0.00010215544377814375, + "loss": 0.9968, + "step": 11091 + }, + { + "epoch": 1.9748931623931623, + "grad_norm": 0.6815677881240845, + "learning_rate": 0.0001021414495368519, + "loss": 0.8889, + "step": 11092 + }, + { + "epoch": 1.975071225071225, + "grad_norm": 0.6872935891151428, + "learning_rate": 0.00010212745525360277, + "loss": 1.1582, + "step": 11093 + }, + { + "epoch": 1.9752492877492878, + "grad_norm": 0.6781140565872192, + "learning_rate": 0.00010211346092867056, + "loss": 0.9988, + "step": 11094 + }, + { + "epoch": 1.9754273504273505, + "grad_norm": 0.6959224343299866, + "learning_rate": 0.00010209946656232949, + "loss": 1.1097, + "step": 11095 + }, + { + "epoch": 1.9756054131054133, + "grad_norm": 0.7205058336257935, + "learning_rate": 0.00010208547215485376, + "loss": 0.9951, + "step": 11096 + }, + { + "epoch": 1.9757834757834758, + "grad_norm": 0.6968751549720764, + "learning_rate": 0.00010207147770651748, + "loss": 0.9313, + "step": 11097 + }, + { + "epoch": 1.9759615384615383, + "grad_norm": 0.6688823103904724, + "learning_rate": 0.00010205748321759494, + "loss": 0.9439, + "step": 11098 + }, + { + "epoch": 1.976139601139601, + "grad_norm": 0.6169568300247192, + "learning_rate": 0.00010204348868836028, + "loss": 1.123, + "step": 11099 + }, + { + "epoch": 1.9763176638176638, + "grad_norm": 0.6995537281036377, + "learning_rate": 0.00010202949411908768, + "loss": 1.1928, + "step": 11100 + }, + { + "epoch": 1.9764957264957266, + "grad_norm": 0.7102637887001038, + "learning_rate": 0.00010201549951005138, + "loss": 1.0265, + "step": 11101 + }, + { + "epoch": 1.9766737891737893, + "grad_norm": 0.6820045113563538, + "learning_rate": 0.00010200150486152558, + "loss": 0.9309, + "step": 11102 + }, + { + "epoch": 1.9768518518518519, + "grad_norm": 0.7050938010215759, + "learning_rate": 0.00010198751017378443, + "loss": 1.0047, + "step": 11103 + }, + { + "epoch": 1.9770299145299144, + "grad_norm": 0.6418201923370361, + "learning_rate": 0.00010197351544710214, + "loss": 1.1172, + "step": 11104 + }, + { + "epoch": 1.9772079772079771, + "grad_norm": 0.6681215763092041, + "learning_rate": 0.0001019595206817529, + "loss": 1.0621, + "step": 11105 + }, + { + "epoch": 1.97738603988604, + "grad_norm": 0.7725709676742554, + "learning_rate": 0.00010194552587801094, + "loss": 1.0044, + "step": 11106 + }, + { + "epoch": 1.9775641025641026, + "grad_norm": 0.6870455741882324, + "learning_rate": 0.00010193153103615045, + "loss": 1.2652, + "step": 11107 + }, + { + "epoch": 1.9777421652421654, + "grad_norm": 0.6352108120918274, + "learning_rate": 0.00010191753615644561, + "loss": 1.1081, + "step": 11108 + }, + { + "epoch": 1.977920227920228, + "grad_norm": 0.7322626113891602, + "learning_rate": 0.00010190354123917066, + "loss": 1.0003, + "step": 11109 + }, + { + "epoch": 1.9780982905982905, + "grad_norm": 0.6240935921669006, + "learning_rate": 0.00010188954628459972, + "loss": 0.8925, + "step": 11110 + }, + { + "epoch": 1.9782763532763532, + "grad_norm": 0.6648945212364197, + "learning_rate": 0.00010187555129300708, + "loss": 1.0882, + "step": 11111 + }, + { + "epoch": 1.978454415954416, + "grad_norm": 0.6704208850860596, + "learning_rate": 0.00010186155626466692, + "loss": 0.8873, + "step": 11112 + }, + { + "epoch": 1.9786324786324787, + "grad_norm": 0.6716459393501282, + "learning_rate": 0.00010184756119985341, + "loss": 1.0045, + "step": 11113 + }, + { + "epoch": 1.9788105413105415, + "grad_norm": 0.81277996301651, + "learning_rate": 0.0001018335660988408, + "loss": 0.8867, + "step": 11114 + }, + { + "epoch": 1.978988603988604, + "grad_norm": 0.7008311748504639, + "learning_rate": 0.00010181957096190323, + "loss": 0.9391, + "step": 11115 + }, + { + "epoch": 1.9791666666666665, + "grad_norm": 0.727676272392273, + "learning_rate": 0.00010180557578931498, + "loss": 1.0157, + "step": 11116 + }, + { + "epoch": 1.9793447293447293, + "grad_norm": 0.7058015465736389, + "learning_rate": 0.00010179158058135018, + "loss": 1.0, + "step": 11117 + }, + { + "epoch": 1.979522792022792, + "grad_norm": 0.7770412564277649, + "learning_rate": 0.00010177758533828312, + "loss": 1.0428, + "step": 11118 + }, + { + "epoch": 1.9797008547008548, + "grad_norm": 0.6557414531707764, + "learning_rate": 0.00010176359006038798, + "loss": 0.8557, + "step": 11119 + }, + { + "epoch": 1.9798789173789175, + "grad_norm": 0.7681090235710144, + "learning_rate": 0.00010174959474793894, + "loss": 0.867, + "step": 11120 + }, + { + "epoch": 1.98005698005698, + "grad_norm": 0.7915860414505005, + "learning_rate": 0.0001017355994012102, + "loss": 0.9961, + "step": 11121 + }, + { + "epoch": 1.9802350427350426, + "grad_norm": 0.8039166927337646, + "learning_rate": 0.00010172160402047604, + "loss": 1.1378, + "step": 11122 + }, + { + "epoch": 1.9804131054131053, + "grad_norm": 0.6641189455986023, + "learning_rate": 0.0001017076086060106, + "loss": 0.8914, + "step": 11123 + }, + { + "epoch": 1.980591168091168, + "grad_norm": 0.7673811316490173, + "learning_rate": 0.00010169361315808812, + "loss": 1.018, + "step": 11124 + }, + { + "epoch": 1.9807692307692308, + "grad_norm": 0.7320558428764343, + "learning_rate": 0.00010167961767698279, + "loss": 1.0515, + "step": 11125 + }, + { + "epoch": 1.9809472934472936, + "grad_norm": 0.5717357993125916, + "learning_rate": 0.00010166562216296886, + "loss": 0.7619, + "step": 11126 + }, + { + "epoch": 1.9811253561253561, + "grad_norm": 0.6638465523719788, + "learning_rate": 0.00010165162661632052, + "loss": 1.0161, + "step": 11127 + }, + { + "epoch": 1.9813034188034186, + "grad_norm": 0.7293243408203125, + "learning_rate": 0.00010163763103731201, + "loss": 1.063, + "step": 11128 + }, + { + "epoch": 1.9814814814814814, + "grad_norm": 0.634694516658783, + "learning_rate": 0.00010162363542621752, + "loss": 0.8945, + "step": 11129 + }, + { + "epoch": 1.9816595441595442, + "grad_norm": 0.7086902856826782, + "learning_rate": 0.00010160963978331122, + "loss": 1.0542, + "step": 11130 + }, + { + "epoch": 1.981837606837607, + "grad_norm": 0.5939825773239136, + "learning_rate": 0.00010159564410886742, + "loss": 0.7822, + "step": 11131 + }, + { + "epoch": 1.9820156695156697, + "grad_norm": 0.722183346748352, + "learning_rate": 0.00010158164840316027, + "loss": 1.0252, + "step": 11132 + }, + { + "epoch": 1.9821937321937322, + "grad_norm": 0.7300103306770325, + "learning_rate": 0.000101567652666464, + "loss": 0.9099, + "step": 11133 + }, + { + "epoch": 1.9823717948717947, + "grad_norm": 0.7148736119270325, + "learning_rate": 0.00010155365689905285, + "loss": 1.0149, + "step": 11134 + }, + { + "epoch": 1.9825498575498575, + "grad_norm": 0.8214462995529175, + "learning_rate": 0.000101539661101201, + "loss": 1.0127, + "step": 11135 + }, + { + "epoch": 1.9827279202279202, + "grad_norm": 0.7111126780509949, + "learning_rate": 0.00010152566527318265, + "loss": 1.045, + "step": 11136 + }, + { + "epoch": 1.982905982905983, + "grad_norm": 0.6640021800994873, + "learning_rate": 0.00010151166941527213, + "loss": 0.9618, + "step": 11137 + }, + { + "epoch": 1.9830840455840457, + "grad_norm": 0.7177722454071045, + "learning_rate": 0.00010149767352774358, + "loss": 1.0373, + "step": 11138 + }, + { + "epoch": 1.9832621082621082, + "grad_norm": 0.6728883981704712, + "learning_rate": 0.00010148367761087121, + "loss": 0.9886, + "step": 11139 + }, + { + "epoch": 1.9834401709401708, + "grad_norm": 0.7060428857803345, + "learning_rate": 0.00010146968166492926, + "loss": 1.042, + "step": 11140 + }, + { + "epoch": 1.9836182336182335, + "grad_norm": 0.706253707408905, + "learning_rate": 0.00010145568569019192, + "loss": 1.2249, + "step": 11141 + }, + { + "epoch": 1.9837962962962963, + "grad_norm": 0.618221640586853, + "learning_rate": 0.00010144168968693348, + "loss": 0.9223, + "step": 11142 + }, + { + "epoch": 1.983974358974359, + "grad_norm": 0.7005748748779297, + "learning_rate": 0.00010142769365542814, + "loss": 1.2735, + "step": 11143 + }, + { + "epoch": 1.9841524216524218, + "grad_norm": 0.6059799194335938, + "learning_rate": 0.0001014136975959501, + "loss": 0.7216, + "step": 11144 + }, + { + "epoch": 1.9843304843304843, + "grad_norm": 0.7169116735458374, + "learning_rate": 0.00010139970150877358, + "loss": 0.9541, + "step": 11145 + }, + { + "epoch": 1.984508547008547, + "grad_norm": 0.7402058839797974, + "learning_rate": 0.00010138570539417281, + "loss": 1.1268, + "step": 11146 + }, + { + "epoch": 1.9846866096866096, + "grad_norm": 0.7204117178916931, + "learning_rate": 0.00010137170925242201, + "loss": 1.1557, + "step": 11147 + }, + { + "epoch": 1.9848646723646723, + "grad_norm": 0.589163064956665, + "learning_rate": 0.00010135771308379545, + "loss": 0.9863, + "step": 11148 + }, + { + "epoch": 1.985042735042735, + "grad_norm": 0.6342785358428955, + "learning_rate": 0.00010134371688856732, + "loss": 0.9294, + "step": 11149 + }, + { + "epoch": 1.9852207977207978, + "grad_norm": 0.7144256234169006, + "learning_rate": 0.00010132972066701183, + "loss": 0.9428, + "step": 11150 + }, + { + "epoch": 1.9853988603988604, + "grad_norm": 0.658032238483429, + "learning_rate": 0.00010131572441940322, + "loss": 0.9749, + "step": 11151 + }, + { + "epoch": 1.9855769230769231, + "grad_norm": 0.7609163522720337, + "learning_rate": 0.00010130172814601576, + "loss": 1.1771, + "step": 11152 + }, + { + "epoch": 1.9857549857549857, + "grad_norm": 0.6531760692596436, + "learning_rate": 0.00010128773184712361, + "loss": 0.8529, + "step": 11153 + }, + { + "epoch": 1.9859330484330484, + "grad_norm": 0.6983599066734314, + "learning_rate": 0.00010127373552300103, + "loss": 1.0307, + "step": 11154 + }, + { + "epoch": 1.9861111111111112, + "grad_norm": 0.7121559381484985, + "learning_rate": 0.00010125973917392224, + "loss": 0.9426, + "step": 11155 + }, + { + "epoch": 1.986289173789174, + "grad_norm": 0.6282170414924622, + "learning_rate": 0.0001012457428001615, + "loss": 0.8983, + "step": 11156 + }, + { + "epoch": 1.9864672364672364, + "grad_norm": 0.6960387825965881, + "learning_rate": 0.000101231746401993, + "loss": 0.9001, + "step": 11157 + }, + { + "epoch": 1.9866452991452992, + "grad_norm": 0.7523152232170105, + "learning_rate": 0.000101217749979691, + "loss": 1.3462, + "step": 11158 + }, + { + "epoch": 1.9868233618233617, + "grad_norm": 0.71713787317276, + "learning_rate": 0.00010120375353352971, + "loss": 1.0147, + "step": 11159 + }, + { + "epoch": 1.9870014245014245, + "grad_norm": 0.7304390072822571, + "learning_rate": 0.00010118975706378339, + "loss": 0.8436, + "step": 11160 + }, + { + "epoch": 1.9871794871794872, + "grad_norm": 0.789968729019165, + "learning_rate": 0.00010117576057072622, + "loss": 1.1162, + "step": 11161 + }, + { + "epoch": 1.98735754985755, + "grad_norm": 0.6752170920372009, + "learning_rate": 0.00010116176405463249, + "loss": 1.0619, + "step": 11162 + }, + { + "epoch": 1.9875356125356125, + "grad_norm": 0.681398868560791, + "learning_rate": 0.0001011477675157764, + "loss": 0.8981, + "step": 11163 + }, + { + "epoch": 1.9877136752136753, + "grad_norm": 0.61469566822052, + "learning_rate": 0.0001011337709544322, + "loss": 1.0139, + "step": 11164 + }, + { + "epoch": 1.9878917378917378, + "grad_norm": 0.7524265050888062, + "learning_rate": 0.0001011197743708741, + "loss": 1.1571, + "step": 11165 + }, + { + "epoch": 1.9880698005698005, + "grad_norm": 0.6289594173431396, + "learning_rate": 0.00010110577776537633, + "loss": 0.93, + "step": 11166 + }, + { + "epoch": 1.9882478632478633, + "grad_norm": 0.6991903781890869, + "learning_rate": 0.00010109178113821318, + "loss": 1.1176, + "step": 11167 + }, + { + "epoch": 1.988425925925926, + "grad_norm": 0.7604053020477295, + "learning_rate": 0.00010107778448965883, + "loss": 1.0497, + "step": 11168 + }, + { + "epoch": 1.9886039886039886, + "grad_norm": 0.7166453003883362, + "learning_rate": 0.00010106378781998753, + "loss": 1.1237, + "step": 11169 + }, + { + "epoch": 1.9887820512820513, + "grad_norm": 0.6071686744689941, + "learning_rate": 0.00010104979112947352, + "loss": 0.8934, + "step": 11170 + }, + { + "epoch": 1.9889601139601139, + "grad_norm": 0.6618169546127319, + "learning_rate": 0.00010103579441839101, + "loss": 1.0596, + "step": 11171 + }, + { + "epoch": 1.9891381766381766, + "grad_norm": 0.6838458776473999, + "learning_rate": 0.0001010217976870143, + "loss": 1.0167, + "step": 11172 + }, + { + "epoch": 1.9893162393162394, + "grad_norm": 0.6369979381561279, + "learning_rate": 0.00010100780093561757, + "loss": 0.9001, + "step": 11173 + }, + { + "epoch": 1.989494301994302, + "grad_norm": 0.661313533782959, + "learning_rate": 0.00010099380416447508, + "loss": 0.8952, + "step": 11174 + }, + { + "epoch": 1.9896723646723646, + "grad_norm": 0.6991600394248962, + "learning_rate": 0.00010097980737386106, + "loss": 1.0083, + "step": 11175 + }, + { + "epoch": 1.9898504273504274, + "grad_norm": 0.618748664855957, + "learning_rate": 0.00010096581056404972, + "loss": 0.8797, + "step": 11176 + }, + { + "epoch": 1.99002849002849, + "grad_norm": 0.7039223909378052, + "learning_rate": 0.00010095181373531535, + "loss": 1.0385, + "step": 11177 + }, + { + "epoch": 1.9902065527065527, + "grad_norm": 0.7598999738693237, + "learning_rate": 0.00010093781688793216, + "loss": 0.9205, + "step": 11178 + }, + { + "epoch": 1.9903846153846154, + "grad_norm": 0.6355955600738525, + "learning_rate": 0.00010092382002217441, + "loss": 0.8646, + "step": 11179 + }, + { + "epoch": 1.9905626780626782, + "grad_norm": 0.8024569153785706, + "learning_rate": 0.00010090982313831634, + "loss": 1.1678, + "step": 11180 + }, + { + "epoch": 1.9907407407407407, + "grad_norm": 0.5960529446601868, + "learning_rate": 0.00010089582623663216, + "loss": 0.8277, + "step": 11181 + }, + { + "epoch": 1.9909188034188035, + "grad_norm": 0.6323728561401367, + "learning_rate": 0.00010088182931739609, + "loss": 0.948, + "step": 11182 + }, + { + "epoch": 1.991096866096866, + "grad_norm": 0.7532381415367126, + "learning_rate": 0.00010086783238088244, + "loss": 1.2948, + "step": 11183 + }, + { + "epoch": 1.9912749287749287, + "grad_norm": 0.5740166306495667, + "learning_rate": 0.00010085383542736543, + "loss": 0.7019, + "step": 11184 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 0.616985559463501, + "learning_rate": 0.00010083983845711929, + "loss": 1.0802, + "step": 11185 + }, + { + "epoch": 1.9916310541310542, + "grad_norm": 0.7505929470062256, + "learning_rate": 0.00010082584147041824, + "loss": 1.0523, + "step": 11186 + }, + { + "epoch": 1.9918091168091168, + "grad_norm": 0.7147656679153442, + "learning_rate": 0.00010081184446753653, + "loss": 1.0019, + "step": 11187 + }, + { + "epoch": 1.9919871794871795, + "grad_norm": 0.7301992774009705, + "learning_rate": 0.00010079784744874845, + "loss": 1.0329, + "step": 11188 + }, + { + "epoch": 1.992165242165242, + "grad_norm": 0.6847206354141235, + "learning_rate": 0.00010078385041432819, + "loss": 1.0367, + "step": 11189 + }, + { + "epoch": 1.9923433048433048, + "grad_norm": 0.7310990691184998, + "learning_rate": 0.00010076985336455, + "loss": 1.1675, + "step": 11190 + }, + { + "epoch": 1.9925213675213675, + "grad_norm": 0.6916858553886414, + "learning_rate": 0.00010075585629968813, + "loss": 0.8615, + "step": 11191 + }, + { + "epoch": 1.9926994301994303, + "grad_norm": 0.6519390344619751, + "learning_rate": 0.00010074185922001685, + "loss": 0.8105, + "step": 11192 + }, + { + "epoch": 1.9928774928774928, + "grad_norm": 0.7437400817871094, + "learning_rate": 0.00010072786212581036, + "loss": 0.9993, + "step": 11193 + }, + { + "epoch": 1.9930555555555556, + "grad_norm": 0.5048928260803223, + "learning_rate": 0.00010071386501734292, + "loss": 0.7912, + "step": 11194 + }, + { + "epoch": 1.993233618233618, + "grad_norm": 0.8042343258857727, + "learning_rate": 0.00010069986789488882, + "loss": 0.9156, + "step": 11195 + }, + { + "epoch": 1.9934116809116809, + "grad_norm": 0.7188669443130493, + "learning_rate": 0.0001006858707587222, + "loss": 1.0474, + "step": 11196 + }, + { + "epoch": 1.9935897435897436, + "grad_norm": 0.7377660870552063, + "learning_rate": 0.00010067187360911738, + "loss": 0.7013, + "step": 11197 + }, + { + "epoch": 1.9937678062678064, + "grad_norm": 0.6684696078300476, + "learning_rate": 0.00010065787644634861, + "loss": 0.9199, + "step": 11198 + }, + { + "epoch": 1.993945868945869, + "grad_norm": 0.7341524958610535, + "learning_rate": 0.00010064387927069012, + "loss": 1.0925, + "step": 11199 + }, + { + "epoch": 1.9941239316239316, + "grad_norm": 0.685745120048523, + "learning_rate": 0.00010062988208241614, + "loss": 1.083, + "step": 11200 + }, + { + "epoch": 1.9943019943019942, + "grad_norm": 0.6923556327819824, + "learning_rate": 0.00010061588488180096, + "loss": 1.2728, + "step": 11201 + }, + { + "epoch": 1.994480056980057, + "grad_norm": 0.6663293242454529, + "learning_rate": 0.00010060188766911876, + "loss": 1.0937, + "step": 11202 + }, + { + "epoch": 1.9946581196581197, + "grad_norm": 0.7963639497756958, + "learning_rate": 0.00010058789044464383, + "loss": 1.0592, + "step": 11203 + }, + { + "epoch": 1.9948361823361824, + "grad_norm": 0.6362990140914917, + "learning_rate": 0.00010057389320865042, + "loss": 0.8872, + "step": 11204 + }, + { + "epoch": 1.9950142450142452, + "grad_norm": 0.7752974033355713, + "learning_rate": 0.00010055989596141278, + "loss": 1.043, + "step": 11205 + }, + { + "epoch": 1.9951923076923077, + "grad_norm": 0.7125133275985718, + "learning_rate": 0.00010054589870320512, + "loss": 1.0015, + "step": 11206 + }, + { + "epoch": 1.9953703703703702, + "grad_norm": 0.7102736830711365, + "learning_rate": 0.00010053190143430169, + "loss": 1.0052, + "step": 11207 + }, + { + "epoch": 1.995548433048433, + "grad_norm": 0.8628628849983215, + "learning_rate": 0.00010051790415497677, + "loss": 1.2351, + "step": 11208 + }, + { + "epoch": 1.9957264957264957, + "grad_norm": 0.7233129739761353, + "learning_rate": 0.00010050390686550462, + "loss": 1.0848, + "step": 11209 + }, + { + "epoch": 1.9959045584045585, + "grad_norm": 0.5936228036880493, + "learning_rate": 0.00010048990956615944, + "loss": 0.7998, + "step": 11210 + }, + { + "epoch": 1.9960826210826212, + "grad_norm": 0.7345388531684875, + "learning_rate": 0.0001004759122572155, + "loss": 1.0329, + "step": 11211 + }, + { + "epoch": 1.9962606837606838, + "grad_norm": 0.7344130873680115, + "learning_rate": 0.00010046191493894703, + "loss": 1.1563, + "step": 11212 + }, + { + "epoch": 1.9964387464387463, + "grad_norm": 0.6979942321777344, + "learning_rate": 0.00010044791761162833, + "loss": 0.9269, + "step": 11213 + }, + { + "epoch": 1.996616809116809, + "grad_norm": 0.67514967918396, + "learning_rate": 0.0001004339202755336, + "loss": 0.9028, + "step": 11214 + }, + { + "epoch": 1.9967948717948718, + "grad_norm": 0.6379111409187317, + "learning_rate": 0.00010041992293093712, + "loss": 0.7816, + "step": 11215 + }, + { + "epoch": 1.9969729344729346, + "grad_norm": 0.693976104259491, + "learning_rate": 0.00010040592557811308, + "loss": 0.8411, + "step": 11216 + }, + { + "epoch": 1.9971509971509973, + "grad_norm": 0.5952646732330322, + "learning_rate": 0.0001003919282173358, + "loss": 0.8681, + "step": 11217 + }, + { + "epoch": 1.9973290598290598, + "grad_norm": 0.7452160716056824, + "learning_rate": 0.00010037793084887948, + "loss": 1.0198, + "step": 11218 + }, + { + "epoch": 1.9975071225071224, + "grad_norm": 0.6683938503265381, + "learning_rate": 0.00010036393347301841, + "loss": 0.8162, + "step": 11219 + }, + { + "epoch": 1.9976851851851851, + "grad_norm": 0.6849120855331421, + "learning_rate": 0.00010034993609002683, + "loss": 1.0668, + "step": 11220 + }, + { + "epoch": 1.9978632478632479, + "grad_norm": 0.8782517910003662, + "learning_rate": 0.00010033593870017897, + "loss": 1.222, + "step": 11221 + }, + { + "epoch": 1.9980413105413106, + "grad_norm": 0.6482772827148438, + "learning_rate": 0.00010032194130374908, + "loss": 0.7722, + "step": 11222 + }, + { + "epoch": 1.9982193732193734, + "grad_norm": 0.8595399260520935, + "learning_rate": 0.00010030794390101142, + "loss": 1.3004, + "step": 11223 + }, + { + "epoch": 1.998397435897436, + "grad_norm": 0.7258931994438171, + "learning_rate": 0.00010029394649224024, + "loss": 0.8825, + "step": 11224 + }, + { + "epoch": 1.9985754985754984, + "grad_norm": 0.6291348934173584, + "learning_rate": 0.00010027994907770981, + "loss": 0.8681, + "step": 11225 + }, + { + "epoch": 1.9987535612535612, + "grad_norm": 0.7528844475746155, + "learning_rate": 0.00010026595165769434, + "loss": 1.1443, + "step": 11226 + }, + { + "epoch": 1.998931623931624, + "grad_norm": 0.654017984867096, + "learning_rate": 0.0001002519542324681, + "loss": 0.8585, + "step": 11227 + }, + { + "epoch": 1.9991096866096867, + "grad_norm": 0.6812533736228943, + "learning_rate": 0.00010023795680230532, + "loss": 0.8757, + "step": 11228 + }, + { + "epoch": 1.9992877492877494, + "grad_norm": 0.7120179533958435, + "learning_rate": 0.0001002239593674803, + "loss": 1.0159, + "step": 11229 + }, + { + "epoch": 1.999465811965812, + "grad_norm": 0.6943802237510681, + "learning_rate": 0.00010020996192826725, + "loss": 1.0193, + "step": 11230 + }, + { + "epoch": 1.9996438746438745, + "grad_norm": 0.7227906584739685, + "learning_rate": 0.00010019596448494047, + "loss": 1.1536, + "step": 11231 + }, + { + "epoch": 1.9998219373219372, + "grad_norm": 0.6233312487602234, + "learning_rate": 0.00010018196703777411, + "loss": 0.9117, + "step": 11232 + }, + { + "epoch": 1.9998219373219372, + "eval_loss": 1.0963108539581299, + "eval_runtime": 24.4478, + "eval_samples_per_second": 42.58, + "eval_steps_per_second": 21.311, + "step": 11232 + } + ], + "logging_steps": 1, + "max_steps": 22464, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 5616, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.3382624499400704e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-11232/training_args.bin b/checkpoint-11232/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..1245f6a2afbe9a6eefbb6d141231d555e0b0bf84 --- /dev/null +++ b/checkpoint-11232/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86de370014ed2be86ea27c820b434ceec5e097da2b5f9b08d0eac9aa564d8961 +size 6200 diff --git a/checkpoint-16848/README.md b/checkpoint-16848/README.md new file mode 100644 index 0000000000000000000000000000000000000000..719b4726992f7d0707a4253e9123dec35e4de390 --- /dev/null +++ b/checkpoint-16848/README.md @@ -0,0 +1,202 @@ +--- +base_model: openlm-research/open_llama_3b_v2 +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.13.2 \ No newline at end of file diff --git a/checkpoint-16848/adapter_config.json b/checkpoint-16848/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6f20a570fc808390da3f2e001093ac1e56c1da --- /dev/null +++ b/checkpoint-16848/adapter_config.json @@ -0,0 +1,34 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "openlm-research/open_llama_3b_v2", + "bias": "none", + "fan_in_fan_out": null, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.0, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "v_proj", + "down_proj", + "o_proj", + "q_proj", + "up_proj", + "k_proj" + ], + "task_type": "CAUSAL_LM", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/checkpoint-16848/adapter_model.safetensors b/checkpoint-16848/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3e168b3124d125c747860a13472082cb4b3c7ff5 --- /dev/null +++ b/checkpoint-16848/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03085ad0ad1bd99c21ebcb38449bb38ae5c59f4fd968a0b6bb26d8cd80e9e417 +size 50899792 diff --git a/checkpoint-16848/optimizer.pt b/checkpoint-16848/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..3a2c6b78d5691087b2f44d23513f47d44f07c4a6 --- /dev/null +++ b/checkpoint-16848/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71fb7e3d382d3193410d495df1ec8962896c45111c77639022c41a01d17ffee4 +size 26231684 diff --git a/checkpoint-16848/rng_state.pth b/checkpoint-16848/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..9402f707925fbc8129d9caeef7089abcb232912c --- /dev/null +++ b/checkpoint-16848/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52cca5856c568bc52c683b690919168fa27bfbdfefc6e0a62355afa6011157c3 +size 14244 diff --git a/checkpoint-16848/scheduler.pt b/checkpoint-16848/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..911075f926d55fdd3ae06b342e6974c78d70efb0 --- /dev/null +++ b/checkpoint-16848/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3acc4594645e05c91ea7d2972486dd125352a9c7ef49354924ff3431645d5b82 +size 1064 diff --git a/checkpoint-16848/special_tokens_map.json b/checkpoint-16848/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/checkpoint-16848/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoint-16848/tokenizer.model b/checkpoint-16848/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..98866ff8ae3631f331c57923c921a0c9ad22b97d --- /dev/null +++ b/checkpoint-16848/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b289e85fa20fd375d8b33dc12f77616f18abc6359804471d1fafcb425fecb8 +size 511574 diff --git a/checkpoint-16848/tokenizer_config.json b/checkpoint-16848/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/checkpoint-16848/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/checkpoint-16848/trainer_state.json b/checkpoint-16848/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a81badf41324718b70079c4e113b372df15bd5ea --- /dev/null +++ b/checkpoint-16848/trainer_state.json @@ -0,0 +1,118073 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.9996438746438745, + "eval_steps": 1404, + "global_step": 16848, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00017806267806267807, + "grad_norm": 0.2854898273944855, + "learning_rate": 1e-05, + "loss": 1.1997, + "step": 1 + }, + { + "epoch": 0.00017806267806267807, + "eval_loss": 1.3698358535766602, + "eval_runtime": 24.1591, + "eval_samples_per_second": 43.089, + "eval_steps_per_second": 21.565, + "step": 1 + }, + { + "epoch": 0.00035612535612535614, + "grad_norm": 0.3508087396621704, + "learning_rate": 2e-05, + "loss": 1.4134, + "step": 2 + }, + { + "epoch": 0.0005341880341880342, + "grad_norm": 0.27050870656967163, + "learning_rate": 3e-05, + "loss": 1.3447, + "step": 3 + }, + { + "epoch": 0.0007122507122507123, + "grad_norm": 0.27706292271614075, + "learning_rate": 4e-05, + "loss": 1.0354, + "step": 4 + }, + { + "epoch": 0.0008903133903133903, + "grad_norm": 0.30398961901664734, + "learning_rate": 5e-05, + "loss": 1.1441, + "step": 5 + }, + { + "epoch": 0.0010683760683760685, + "grad_norm": 0.3103881776332855, + "learning_rate": 6e-05, + "loss": 1.341, + "step": 6 + }, + { + "epoch": 0.0012464387464387464, + "grad_norm": 0.5191189646720886, + "learning_rate": 7e-05, + "loss": 1.3457, + "step": 7 + }, + { + "epoch": 0.0014245014245014246, + "grad_norm": 0.4449467360973358, + "learning_rate": 8e-05, + "loss": 1.5051, + "step": 8 + }, + { + "epoch": 0.0016025641025641025, + "grad_norm": 0.3914581537246704, + "learning_rate": 9e-05, + "loss": 1.5525, + "step": 9 + }, + { + "epoch": 0.0017806267806267807, + "grad_norm": 0.37746086716651917, + "learning_rate": 0.0001, + "loss": 1.3266, + "step": 10 + }, + { + "epoch": 0.001958689458689459, + "grad_norm": 0.35226109623908997, + "learning_rate": 0.00011000000000000002, + "loss": 1.5416, + "step": 11 + }, + { + "epoch": 0.002136752136752137, + "grad_norm": 0.3343672454357147, + "learning_rate": 0.00012, + "loss": 1.3221, + "step": 12 + }, + { + "epoch": 0.0023148148148148147, + "grad_norm": 0.47298333048820496, + "learning_rate": 0.00013000000000000002, + "loss": 1.2999, + "step": 13 + }, + { + "epoch": 0.002492877492877493, + "grad_norm": 0.377814918756485, + "learning_rate": 0.00014, + "loss": 1.1688, + "step": 14 + }, + { + "epoch": 0.002670940170940171, + "grad_norm": 0.46344801783561707, + "learning_rate": 0.00015000000000000001, + "loss": 1.3565, + "step": 15 + }, + { + "epoch": 0.002849002849002849, + "grad_norm": 0.49615249037742615, + "learning_rate": 0.00016, + "loss": 1.5692, + "step": 16 + }, + { + "epoch": 0.003027065527065527, + "grad_norm": 0.5109946131706238, + "learning_rate": 0.00017, + "loss": 1.2991, + "step": 17 + }, + { + "epoch": 0.003205128205128205, + "grad_norm": 0.5125070214271545, + "learning_rate": 0.00018, + "loss": 1.3309, + "step": 18 + }, + { + "epoch": 0.003383190883190883, + "grad_norm": 0.4517767131328583, + "learning_rate": 0.00019, + "loss": 1.357, + "step": 19 + }, + { + "epoch": 0.0035612535612535613, + "grad_norm": 0.47267794609069824, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 20 + }, + { + "epoch": 0.0037393162393162395, + "grad_norm": 0.46823424100875854, + "learning_rate": 0.00019999999902035388, + "loss": 1.1195, + "step": 21 + }, + { + "epoch": 0.003917378917378918, + "grad_norm": 0.440036803483963, + "learning_rate": 0.00019999999608141548, + "loss": 1.2822, + "step": 22 + }, + { + "epoch": 0.004095441595441595, + "grad_norm": 0.371101975440979, + "learning_rate": 0.00019999999118318492, + "loss": 1.132, + "step": 23 + }, + { + "epoch": 0.004273504273504274, + "grad_norm": 0.44691094756126404, + "learning_rate": 0.00019999998432566226, + "loss": 1.2968, + "step": 24 + }, + { + "epoch": 0.004451566951566952, + "grad_norm": 0.5462725162506104, + "learning_rate": 0.0001999999755088476, + "loss": 1.1714, + "step": 25 + }, + { + "epoch": 0.004629629629629629, + "grad_norm": 0.39860013127326965, + "learning_rate": 0.0001999999647327412, + "loss": 1.0407, + "step": 26 + }, + { + "epoch": 0.004807692307692308, + "grad_norm": 0.5031934380531311, + "learning_rate": 0.0001999999519973432, + "loss": 1.2773, + "step": 27 + }, + { + "epoch": 0.004985754985754986, + "grad_norm": 0.42162764072418213, + "learning_rate": 0.0001999999373026539, + "loss": 1.2824, + "step": 28 + }, + { + "epoch": 0.005163817663817663, + "grad_norm": 0.40964868664741516, + "learning_rate": 0.00019999992064867353, + "loss": 1.226, + "step": 29 + }, + { + "epoch": 0.005341880341880342, + "grad_norm": 0.41650915145874023, + "learning_rate": 0.00019999990203540245, + "loss": 1.2677, + "step": 30 + }, + { + "epoch": 0.00551994301994302, + "grad_norm": 0.40052226185798645, + "learning_rate": 0.00019999988146284103, + "loss": 0.9443, + "step": 31 + }, + { + "epoch": 0.005698005698005698, + "grad_norm": 0.5198387503623962, + "learning_rate": 0.00019999985893098964, + "loss": 1.3043, + "step": 32 + }, + { + "epoch": 0.005876068376068376, + "grad_norm": 0.50941002368927, + "learning_rate": 0.00019999983443984878, + "loss": 1.2002, + "step": 33 + }, + { + "epoch": 0.006054131054131054, + "grad_norm": 0.30082932114601135, + "learning_rate": 0.00019999980798941888, + "loss": 0.9904, + "step": 34 + }, + { + "epoch": 0.006232193732193732, + "grad_norm": 0.4228935241699219, + "learning_rate": 0.00019999977957970048, + "loss": 1.1137, + "step": 35 + }, + { + "epoch": 0.00641025641025641, + "grad_norm": 0.41294750571250916, + "learning_rate": 0.0001999997492106941, + "loss": 1.3385, + "step": 36 + }, + { + "epoch": 0.006588319088319089, + "grad_norm": 0.4415493905544281, + "learning_rate": 0.00019999971688240041, + "loss": 1.1695, + "step": 37 + }, + { + "epoch": 0.006766381766381766, + "grad_norm": 0.3726460933685303, + "learning_rate": 0.00019999968259482, + "loss": 1.1734, + "step": 38 + }, + { + "epoch": 0.006944444444444444, + "grad_norm": 0.3969627320766449, + "learning_rate": 0.0001999996463479535, + "loss": 1.1209, + "step": 39 + }, + { + "epoch": 0.007122507122507123, + "grad_norm": 0.3779667913913727, + "learning_rate": 0.0001999996081418017, + "loss": 1.1635, + "step": 40 + }, + { + "epoch": 0.0073005698005698, + "grad_norm": 0.3933636546134949, + "learning_rate": 0.0001999995679763653, + "loss": 1.1514, + "step": 41 + }, + { + "epoch": 0.007478632478632479, + "grad_norm": 0.3567957282066345, + "learning_rate": 0.00019999952585164507, + "loss": 1.2488, + "step": 42 + }, + { + "epoch": 0.007656695156695157, + "grad_norm": 0.32506081461906433, + "learning_rate": 0.00019999948176764186, + "loss": 1.149, + "step": 43 + }, + { + "epoch": 0.007834757834757835, + "grad_norm": 0.46588361263275146, + "learning_rate": 0.0001999994357243566, + "loss": 1.4263, + "step": 44 + }, + { + "epoch": 0.008012820512820512, + "grad_norm": 0.5070307850837708, + "learning_rate": 0.00019999938772179005, + "loss": 1.0698, + "step": 45 + }, + { + "epoch": 0.00819088319088319, + "grad_norm": 0.38199326395988464, + "learning_rate": 0.00019999933775994327, + "loss": 0.9907, + "step": 46 + }, + { + "epoch": 0.00836894586894587, + "grad_norm": 0.43684661388397217, + "learning_rate": 0.0001999992858388172, + "loss": 1.2905, + "step": 47 + }, + { + "epoch": 0.008547008547008548, + "grad_norm": 0.44482162594795227, + "learning_rate": 0.00019999923195841284, + "loss": 1.2153, + "step": 48 + }, + { + "epoch": 0.008725071225071225, + "grad_norm": 0.4259667694568634, + "learning_rate": 0.0001999991761187313, + "loss": 1.1582, + "step": 49 + }, + { + "epoch": 0.008903133903133903, + "grad_norm": 0.41649091243743896, + "learning_rate": 0.00019999911831977357, + "loss": 1.0185, + "step": 50 + }, + { + "epoch": 0.009081196581196582, + "grad_norm": 0.4179716110229492, + "learning_rate": 0.0001999990585615409, + "loss": 1.3579, + "step": 51 + }, + { + "epoch": 0.009259259259259259, + "grad_norm": 0.3372558355331421, + "learning_rate": 0.00019999899684403438, + "loss": 1.0638, + "step": 52 + }, + { + "epoch": 0.009437321937321937, + "grad_norm": 0.41294020414352417, + "learning_rate": 0.00019999893316725525, + "loss": 1.1932, + "step": 53 + }, + { + "epoch": 0.009615384615384616, + "grad_norm": 0.4407919645309448, + "learning_rate": 0.00019999886753120473, + "loss": 1.4129, + "step": 54 + }, + { + "epoch": 0.009793447293447293, + "grad_norm": 0.47948843240737915, + "learning_rate": 0.00019999879993588414, + "loss": 1.2424, + "step": 55 + }, + { + "epoch": 0.009971509971509971, + "grad_norm": 0.3535355031490326, + "learning_rate": 0.00019999873038129484, + "loss": 1.0145, + "step": 56 + }, + { + "epoch": 0.01014957264957265, + "grad_norm": 0.5067078471183777, + "learning_rate": 0.00019999865886743813, + "loss": 1.4708, + "step": 57 + }, + { + "epoch": 0.010327635327635327, + "grad_norm": 0.42862898111343384, + "learning_rate": 0.0001999985853943154, + "loss": 1.0399, + "step": 58 + }, + { + "epoch": 0.010505698005698005, + "grad_norm": 0.4769059419631958, + "learning_rate": 0.00019999850996192816, + "loss": 1.1258, + "step": 59 + }, + { + "epoch": 0.010683760683760684, + "grad_norm": 0.4065442383289337, + "learning_rate": 0.0001999984325702778, + "loss": 1.2077, + "step": 60 + }, + { + "epoch": 0.010861823361823363, + "grad_norm": 0.5318329930305481, + "learning_rate": 0.0001999983532193659, + "loss": 1.2298, + "step": 61 + }, + { + "epoch": 0.01103988603988604, + "grad_norm": 0.4777173101902008, + "learning_rate": 0.000199998271909194, + "loss": 1.3195, + "step": 62 + }, + { + "epoch": 0.011217948717948718, + "grad_norm": 0.37553808093070984, + "learning_rate": 0.0001999981886397637, + "loss": 1.1188, + "step": 63 + }, + { + "epoch": 0.011396011396011397, + "grad_norm": 0.3920556902885437, + "learning_rate": 0.0001999981034110766, + "loss": 1.1448, + "step": 64 + }, + { + "epoch": 0.011574074074074073, + "grad_norm": 0.454272598028183, + "learning_rate": 0.0001999980162231344, + "loss": 1.0812, + "step": 65 + }, + { + "epoch": 0.011752136752136752, + "grad_norm": 0.4354456663131714, + "learning_rate": 0.00019999792707593882, + "loss": 1.1174, + "step": 66 + }, + { + "epoch": 0.01193019943019943, + "grad_norm": 0.5030252933502197, + "learning_rate": 0.00019999783596949156, + "loss": 1.2925, + "step": 67 + }, + { + "epoch": 0.012108262108262107, + "grad_norm": 0.5141571164131165, + "learning_rate": 0.00019999774290379446, + "loss": 1.6193, + "step": 68 + }, + { + "epoch": 0.012286324786324786, + "grad_norm": 0.417298287153244, + "learning_rate": 0.0001999976478788493, + "loss": 1.1875, + "step": 69 + }, + { + "epoch": 0.012464387464387465, + "grad_norm": 0.4642415940761566, + "learning_rate": 0.00019999755089465795, + "loss": 1.4138, + "step": 70 + }, + { + "epoch": 0.012642450142450143, + "grad_norm": 0.43184754252433777, + "learning_rate": 0.0001999974519512223, + "loss": 1.0697, + "step": 71 + }, + { + "epoch": 0.01282051282051282, + "grad_norm": 0.46698349714279175, + "learning_rate": 0.00019999735104854436, + "loss": 0.709, + "step": 72 + }, + { + "epoch": 0.012998575498575499, + "grad_norm": 0.37253814935684204, + "learning_rate": 0.000199997248186626, + "loss": 1.2084, + "step": 73 + }, + { + "epoch": 0.013176638176638177, + "grad_norm": 0.3851388692855835, + "learning_rate": 0.0001999971433654693, + "loss": 1.0548, + "step": 74 + }, + { + "epoch": 0.013354700854700854, + "grad_norm": 0.4434688985347748, + "learning_rate": 0.00019999703658507635, + "loss": 1.4084, + "step": 75 + }, + { + "epoch": 0.013532763532763533, + "grad_norm": 0.43164482712745667, + "learning_rate": 0.00019999692784544913, + "loss": 1.4872, + "step": 76 + }, + { + "epoch": 0.013710826210826211, + "grad_norm": 0.4224303364753723, + "learning_rate": 0.00019999681714658984, + "loss": 1.2221, + "step": 77 + }, + { + "epoch": 0.013888888888888888, + "grad_norm": 0.35588955879211426, + "learning_rate": 0.00019999670448850069, + "loss": 0.84, + "step": 78 + }, + { + "epoch": 0.014066951566951567, + "grad_norm": 0.3970590829849243, + "learning_rate": 0.0001999965898711838, + "loss": 1.1886, + "step": 79 + }, + { + "epoch": 0.014245014245014245, + "grad_norm": 0.4331924319267273, + "learning_rate": 0.00019999647329464146, + "loss": 1.179, + "step": 80 + }, + { + "epoch": 0.014423076923076924, + "grad_norm": 0.4226946234703064, + "learning_rate": 0.00019999635475887598, + "loss": 1.1496, + "step": 81 + }, + { + "epoch": 0.0146011396011396, + "grad_norm": 0.381592720746994, + "learning_rate": 0.00019999623426388962, + "loss": 1.1774, + "step": 82 + }, + { + "epoch": 0.01477920227920228, + "grad_norm": 0.4190855622291565, + "learning_rate": 0.00019999611180968478, + "loss": 1.1491, + "step": 83 + }, + { + "epoch": 0.014957264957264958, + "grad_norm": 0.3904292583465576, + "learning_rate": 0.00019999598739626389, + "loss": 1.1275, + "step": 84 + }, + { + "epoch": 0.015135327635327635, + "grad_norm": 0.4515478014945984, + "learning_rate": 0.0001999958610236293, + "loss": 1.2404, + "step": 85 + }, + { + "epoch": 0.015313390313390313, + "grad_norm": 0.48341724276542664, + "learning_rate": 0.00019999573269178359, + "loss": 1.3572, + "step": 86 + }, + { + "epoch": 0.015491452991452992, + "grad_norm": 0.42150333523750305, + "learning_rate": 0.00019999560240072914, + "loss": 1.0203, + "step": 87 + }, + { + "epoch": 0.01566951566951567, + "grad_norm": 0.45445525646209717, + "learning_rate": 0.00019999547015046867, + "loss": 1.0677, + "step": 88 + }, + { + "epoch": 0.01584757834757835, + "grad_norm": 0.3581015467643738, + "learning_rate": 0.00019999533594100463, + "loss": 1.0693, + "step": 89 + }, + { + "epoch": 0.016025641025641024, + "grad_norm": 0.4430878758430481, + "learning_rate": 0.00019999519977233971, + "loss": 1.1591, + "step": 90 + }, + { + "epoch": 0.016203703703703703, + "grad_norm": 0.3940352201461792, + "learning_rate": 0.0001999950616444766, + "loss": 1.1325, + "step": 91 + }, + { + "epoch": 0.01638176638176638, + "grad_norm": 0.4521673321723938, + "learning_rate": 0.00019999492155741794, + "loss": 1.3288, + "step": 92 + }, + { + "epoch": 0.01655982905982906, + "grad_norm": 0.3988296687602997, + "learning_rate": 0.00019999477951116658, + "loss": 1.0023, + "step": 93 + }, + { + "epoch": 0.01673789173789174, + "grad_norm": 0.38709723949432373, + "learning_rate": 0.00019999463550572516, + "loss": 1.2623, + "step": 94 + }, + { + "epoch": 0.016915954415954417, + "grad_norm": 0.35376182198524475, + "learning_rate": 0.00019999448954109662, + "loss": 1.0643, + "step": 95 + }, + { + "epoch": 0.017094017094017096, + "grad_norm": 0.49547120928764343, + "learning_rate": 0.00019999434161728377, + "loss": 1.2121, + "step": 96 + }, + { + "epoch": 0.01727207977207977, + "grad_norm": 0.49593672156333923, + "learning_rate": 0.00019999419173428952, + "loss": 1.1635, + "step": 97 + }, + { + "epoch": 0.01745014245014245, + "grad_norm": 0.4146541953086853, + "learning_rate": 0.0001999940398921168, + "loss": 1.1452, + "step": 98 + }, + { + "epoch": 0.017628205128205128, + "grad_norm": 0.5177254676818848, + "learning_rate": 0.00019999388609076858, + "loss": 1.2178, + "step": 99 + }, + { + "epoch": 0.017806267806267807, + "grad_norm": 0.4012768864631653, + "learning_rate": 0.0001999937303302479, + "loss": 0.9222, + "step": 100 + }, + { + "epoch": 0.017984330484330485, + "grad_norm": 0.4597131907939911, + "learning_rate": 0.00019999357261055777, + "loss": 0.979, + "step": 101 + }, + { + "epoch": 0.018162393162393164, + "grad_norm": 0.6190966963768005, + "learning_rate": 0.00019999341293170132, + "loss": 1.3909, + "step": 102 + }, + { + "epoch": 0.01834045584045584, + "grad_norm": 0.4576462209224701, + "learning_rate": 0.00019999325129368164, + "loss": 1.073, + "step": 103 + }, + { + "epoch": 0.018518518518518517, + "grad_norm": 0.4036749005317688, + "learning_rate": 0.00019999308769650192, + "loss": 1.1354, + "step": 104 + }, + { + "epoch": 0.018696581196581196, + "grad_norm": 0.4722452759742737, + "learning_rate": 0.00019999292214016538, + "loss": 1.2039, + "step": 105 + }, + { + "epoch": 0.018874643874643875, + "grad_norm": 0.5338274240493774, + "learning_rate": 0.00019999275462467527, + "loss": 1.225, + "step": 106 + }, + { + "epoch": 0.019052706552706553, + "grad_norm": 0.4301491677761078, + "learning_rate": 0.00019999258515003484, + "loss": 1.0601, + "step": 107 + }, + { + "epoch": 0.019230769230769232, + "grad_norm": 0.33271175622940063, + "learning_rate": 0.0001999924137162474, + "loss": 0.8441, + "step": 108 + }, + { + "epoch": 0.01940883190883191, + "grad_norm": 0.4648784399032593, + "learning_rate": 0.0001999922403233163, + "loss": 1.2038, + "step": 109 + }, + { + "epoch": 0.019586894586894586, + "grad_norm": 0.37915176153182983, + "learning_rate": 0.00019999206497124504, + "loss": 1.0923, + "step": 110 + }, + { + "epoch": 0.019764957264957264, + "grad_norm": 0.3865506052970886, + "learning_rate": 0.00019999188766003695, + "loss": 0.9535, + "step": 111 + }, + { + "epoch": 0.019943019943019943, + "grad_norm": 0.35739636421203613, + "learning_rate": 0.0001999917083896955, + "loss": 1.2688, + "step": 112 + }, + { + "epoch": 0.02012108262108262, + "grad_norm": 0.3943796157836914, + "learning_rate": 0.0001999915271602243, + "loss": 1.1097, + "step": 113 + }, + { + "epoch": 0.0202991452991453, + "grad_norm": 0.44758161902427673, + "learning_rate": 0.0001999913439716268, + "loss": 1.2698, + "step": 114 + }, + { + "epoch": 0.02047720797720798, + "grad_norm": 0.3749747574329376, + "learning_rate": 0.00019999115882390664, + "loss": 1.1091, + "step": 115 + }, + { + "epoch": 0.020655270655270654, + "grad_norm": 0.3479487895965576, + "learning_rate": 0.00019999097171706745, + "loss": 1.0049, + "step": 116 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 0.4491243064403534, + "learning_rate": 0.00019999078265111285, + "loss": 1.1857, + "step": 117 + }, + { + "epoch": 0.02101139601139601, + "grad_norm": 0.345289021730423, + "learning_rate": 0.00019999059162604662, + "loss": 1.1397, + "step": 118 + }, + { + "epoch": 0.02118945868945869, + "grad_norm": 0.5467649698257446, + "learning_rate": 0.00019999039864187243, + "loss": 1.2196, + "step": 119 + }, + { + "epoch": 0.021367521367521368, + "grad_norm": 0.36446481943130493, + "learning_rate": 0.00019999020369859409, + "loss": 0.796, + "step": 120 + }, + { + "epoch": 0.021545584045584047, + "grad_norm": 0.4225841760635376, + "learning_rate": 0.00019999000679621543, + "loss": 0.9684, + "step": 121 + }, + { + "epoch": 0.021723646723646725, + "grad_norm": 0.4205594062805176, + "learning_rate": 0.0001999898079347403, + "loss": 1.2762, + "step": 122 + }, + { + "epoch": 0.0219017094017094, + "grad_norm": 0.43773892521858215, + "learning_rate": 0.00019998960711417257, + "loss": 1.117, + "step": 123 + }, + { + "epoch": 0.02207977207977208, + "grad_norm": 0.41279685497283936, + "learning_rate": 0.00019998940433451623, + "loss": 1.1502, + "step": 124 + }, + { + "epoch": 0.022257834757834757, + "grad_norm": 0.4090803563594818, + "learning_rate": 0.0001999891995957752, + "loss": 1.2591, + "step": 125 + }, + { + "epoch": 0.022435897435897436, + "grad_norm": 0.6000410914421082, + "learning_rate": 0.0001999889928979535, + "loss": 1.4321, + "step": 126 + }, + { + "epoch": 0.022613960113960115, + "grad_norm": 0.524264395236969, + "learning_rate": 0.00019998878424105524, + "loss": 1.1849, + "step": 127 + }, + { + "epoch": 0.022792022792022793, + "grad_norm": 0.4581047296524048, + "learning_rate": 0.00019998857362508443, + "loss": 1.0598, + "step": 128 + }, + { + "epoch": 0.022970085470085472, + "grad_norm": 0.42663446068763733, + "learning_rate": 0.00019998836105004526, + "loss": 1.1909, + "step": 129 + }, + { + "epoch": 0.023148148148148147, + "grad_norm": 0.45709118247032166, + "learning_rate": 0.00019998814651594183, + "loss": 1.2104, + "step": 130 + }, + { + "epoch": 0.023326210826210825, + "grad_norm": 0.39528369903564453, + "learning_rate": 0.0001999879300227784, + "loss": 1.3073, + "step": 131 + }, + { + "epoch": 0.023504273504273504, + "grad_norm": 0.46896448731422424, + "learning_rate": 0.00019998771157055914, + "loss": 1.3202, + "step": 132 + }, + { + "epoch": 0.023682336182336183, + "grad_norm": 0.4386129677295685, + "learning_rate": 0.00019998749115928842, + "loss": 1.2196, + "step": 133 + }, + { + "epoch": 0.02386039886039886, + "grad_norm": 0.45920488238334656, + "learning_rate": 0.00019998726878897051, + "loss": 1.3668, + "step": 134 + }, + { + "epoch": 0.02403846153846154, + "grad_norm": 0.4115797281265259, + "learning_rate": 0.0001999870444596098, + "loss": 1.1052, + "step": 135 + }, + { + "epoch": 0.024216524216524215, + "grad_norm": 0.3860839903354645, + "learning_rate": 0.0001999868181712106, + "loss": 1.0344, + "step": 136 + }, + { + "epoch": 0.024394586894586893, + "grad_norm": 0.42514732480049133, + "learning_rate": 0.00019998658992377742, + "loss": 1.1979, + "step": 137 + }, + { + "epoch": 0.024572649572649572, + "grad_norm": 0.36001840233802795, + "learning_rate": 0.00019998635971731475, + "loss": 1.4536, + "step": 138 + }, + { + "epoch": 0.02475071225071225, + "grad_norm": 0.3739112317562103, + "learning_rate": 0.00019998612755182707, + "loss": 1.0097, + "step": 139 + }, + { + "epoch": 0.02492877492877493, + "grad_norm": 0.37545472383499146, + "learning_rate": 0.00019998589342731888, + "loss": 0.829, + "step": 140 + }, + { + "epoch": 0.025106837606837608, + "grad_norm": 0.38660728931427, + "learning_rate": 0.0001999856573437948, + "loss": 1.1324, + "step": 141 + }, + { + "epoch": 0.025284900284900286, + "grad_norm": 0.3741356432437897, + "learning_rate": 0.00019998541930125953, + "loss": 1.0934, + "step": 142 + }, + { + "epoch": 0.02546296296296296, + "grad_norm": 0.41900336742401123, + "learning_rate": 0.00019998517929971764, + "loss": 1.0336, + "step": 143 + }, + { + "epoch": 0.02564102564102564, + "grad_norm": 0.4167572259902954, + "learning_rate": 0.00019998493733917384, + "loss": 1.2571, + "step": 144 + }, + { + "epoch": 0.02581908831908832, + "grad_norm": 0.39437636733055115, + "learning_rate": 0.0001999846934196329, + "loss": 1.2283, + "step": 145 + }, + { + "epoch": 0.025997150997150997, + "grad_norm": 0.39129480719566345, + "learning_rate": 0.00019998444754109964, + "loss": 0.9893, + "step": 146 + }, + { + "epoch": 0.026175213675213676, + "grad_norm": 0.45533549785614014, + "learning_rate": 0.0001999841997035788, + "loss": 1.0793, + "step": 147 + }, + { + "epoch": 0.026353276353276354, + "grad_norm": 0.3741768002510071, + "learning_rate": 0.00019998394990707524, + "loss": 1.2179, + "step": 148 + }, + { + "epoch": 0.026531339031339033, + "grad_norm": 0.4066533148288727, + "learning_rate": 0.0001999836981515939, + "loss": 1.1443, + "step": 149 + }, + { + "epoch": 0.026709401709401708, + "grad_norm": 0.4851688742637634, + "learning_rate": 0.0001999834444371397, + "loss": 1.1668, + "step": 150 + }, + { + "epoch": 0.026887464387464387, + "grad_norm": 0.428091436624527, + "learning_rate": 0.0001999831887637176, + "loss": 1.2676, + "step": 151 + }, + { + "epoch": 0.027065527065527065, + "grad_norm": 0.4024655222892761, + "learning_rate": 0.0001999829311313326, + "loss": 1.3115, + "step": 152 + }, + { + "epoch": 0.027243589743589744, + "grad_norm": 0.43983033299446106, + "learning_rate": 0.00019998267153998976, + "loss": 1.1019, + "step": 153 + }, + { + "epoch": 0.027421652421652423, + "grad_norm": 0.4317505359649658, + "learning_rate": 0.0001999824099896942, + "loss": 1.3129, + "step": 154 + }, + { + "epoch": 0.0275997150997151, + "grad_norm": 0.43107882142066956, + "learning_rate": 0.000199982146480451, + "loss": 1.2134, + "step": 155 + }, + { + "epoch": 0.027777777777777776, + "grad_norm": 0.3939448297023773, + "learning_rate": 0.00019998188101226532, + "loss": 1.0321, + "step": 156 + }, + { + "epoch": 0.027955840455840455, + "grad_norm": 0.4641847610473633, + "learning_rate": 0.00019998161358514237, + "loss": 1.2369, + "step": 157 + }, + { + "epoch": 0.028133903133903133, + "grad_norm": 0.3538529872894287, + "learning_rate": 0.0001999813441990874, + "loss": 1.2061, + "step": 158 + }, + { + "epoch": 0.028311965811965812, + "grad_norm": 0.3277950584888458, + "learning_rate": 0.0001999810728541057, + "loss": 0.9419, + "step": 159 + }, + { + "epoch": 0.02849002849002849, + "grad_norm": 0.424710750579834, + "learning_rate": 0.00019998079955020254, + "loss": 1.3302, + "step": 160 + }, + { + "epoch": 0.02866809116809117, + "grad_norm": 0.4120834469795227, + "learning_rate": 0.00019998052428738333, + "loss": 1.079, + "step": 161 + }, + { + "epoch": 0.028846153846153848, + "grad_norm": 0.45811930298805237, + "learning_rate": 0.00019998024706565346, + "loss": 1.1259, + "step": 162 + }, + { + "epoch": 0.029024216524216523, + "grad_norm": 0.3873266875743866, + "learning_rate": 0.0001999799678850183, + "loss": 1.2124, + "step": 163 + }, + { + "epoch": 0.0292022792022792, + "grad_norm": 0.5806412696838379, + "learning_rate": 0.00019997968674548337, + "loss": 1.3467, + "step": 164 + }, + { + "epoch": 0.02938034188034188, + "grad_norm": 0.3906802833080292, + "learning_rate": 0.00019997940364705418, + "loss": 1.1438, + "step": 165 + }, + { + "epoch": 0.02955840455840456, + "grad_norm": 0.45201995968818665, + "learning_rate": 0.00019997911858973626, + "loss": 1.1469, + "step": 166 + }, + { + "epoch": 0.029736467236467237, + "grad_norm": 0.4965892732143402, + "learning_rate": 0.0001999788315735352, + "loss": 1.0829, + "step": 167 + }, + { + "epoch": 0.029914529914529916, + "grad_norm": 0.32578057050704956, + "learning_rate": 0.0001999785425984566, + "loss": 1.0432, + "step": 168 + }, + { + "epoch": 0.03009259259259259, + "grad_norm": 0.4146028161048889, + "learning_rate": 0.00019997825166450617, + "loss": 1.1657, + "step": 169 + }, + { + "epoch": 0.03027065527065527, + "grad_norm": 0.4342964291572571, + "learning_rate": 0.0001999779587716896, + "loss": 1.2038, + "step": 170 + }, + { + "epoch": 0.030448717948717948, + "grad_norm": 0.40128546953201294, + "learning_rate": 0.00019997766392001258, + "loss": 1.3044, + "step": 171 + }, + { + "epoch": 0.030626780626780627, + "grad_norm": 0.4357539117336273, + "learning_rate": 0.00019997736710948094, + "loss": 1.2143, + "step": 172 + }, + { + "epoch": 0.030804843304843305, + "grad_norm": 0.4821035861968994, + "learning_rate": 0.00019997706834010045, + "loss": 1.0469, + "step": 173 + }, + { + "epoch": 0.030982905982905984, + "grad_norm": 0.3966675102710724, + "learning_rate": 0.000199976767611877, + "loss": 1.2122, + "step": 174 + }, + { + "epoch": 0.031160968660968662, + "grad_norm": 0.4265064299106598, + "learning_rate": 0.00019997646492481648, + "loss": 1.0871, + "step": 175 + }, + { + "epoch": 0.03133903133903134, + "grad_norm": 0.3445652723312378, + "learning_rate": 0.00019997616027892485, + "loss": 1.0412, + "step": 176 + }, + { + "epoch": 0.031517094017094016, + "grad_norm": 0.47187718749046326, + "learning_rate": 0.000199975853674208, + "loss": 1.0822, + "step": 177 + }, + { + "epoch": 0.0316951566951567, + "grad_norm": 0.37751707434654236, + "learning_rate": 0.000199975545110672, + "loss": 1.1439, + "step": 178 + }, + { + "epoch": 0.03187321937321937, + "grad_norm": 0.38792455196380615, + "learning_rate": 0.00019997523458832286, + "loss": 0.8604, + "step": 179 + }, + { + "epoch": 0.03205128205128205, + "grad_norm": 0.35199594497680664, + "learning_rate": 0.00019997492210716667, + "loss": 1.0819, + "step": 180 + }, + { + "epoch": 0.03222934472934473, + "grad_norm": 0.4828922748565674, + "learning_rate": 0.00019997460766720958, + "loss": 1.1879, + "step": 181 + }, + { + "epoch": 0.032407407407407406, + "grad_norm": 0.46153363585472107, + "learning_rate": 0.00019997429126845774, + "loss": 1.1592, + "step": 182 + }, + { + "epoch": 0.03258547008547009, + "grad_norm": 0.4844890832901001, + "learning_rate": 0.0001999739729109173, + "loss": 1.1334, + "step": 183 + }, + { + "epoch": 0.03276353276353276, + "grad_norm": 0.414617121219635, + "learning_rate": 0.00019997365259459457, + "loss": 1.0547, + "step": 184 + }, + { + "epoch": 0.032941595441595445, + "grad_norm": 0.46544626355171204, + "learning_rate": 0.00019997333031949581, + "loss": 1.4067, + "step": 185 + }, + { + "epoch": 0.03311965811965812, + "grad_norm": 0.48489415645599365, + "learning_rate": 0.0001999730060856273, + "loss": 1.4027, + "step": 186 + }, + { + "epoch": 0.033297720797720795, + "grad_norm": 0.3963346481323242, + "learning_rate": 0.0001999726798929954, + "loss": 1.1327, + "step": 187 + }, + { + "epoch": 0.03347578347578348, + "grad_norm": 0.3809385895729065, + "learning_rate": 0.00019997235174160652, + "loss": 1.3475, + "step": 188 + }, + { + "epoch": 0.03365384615384615, + "grad_norm": 0.3866960406303406, + "learning_rate": 0.0001999720216314671, + "loss": 1.1576, + "step": 189 + }, + { + "epoch": 0.033831908831908834, + "grad_norm": 0.34976935386657715, + "learning_rate": 0.00019997168956258356, + "loss": 0.9361, + "step": 190 + }, + { + "epoch": 0.03400997150997151, + "grad_norm": 0.38681939244270325, + "learning_rate": 0.00019997135553496243, + "loss": 1.1796, + "step": 191 + }, + { + "epoch": 0.03418803418803419, + "grad_norm": 0.41905197501182556, + "learning_rate": 0.0001999710195486103, + "loss": 1.1714, + "step": 192 + }, + { + "epoch": 0.03436609686609687, + "grad_norm": 0.42356589436531067, + "learning_rate": 0.0001999706816035337, + "loss": 1.0022, + "step": 193 + }, + { + "epoch": 0.03454415954415954, + "grad_norm": 0.3929740786552429, + "learning_rate": 0.00019997034169973925, + "loss": 1.3769, + "step": 194 + }, + { + "epoch": 0.034722222222222224, + "grad_norm": 0.4325186312198639, + "learning_rate": 0.00019996999983723366, + "loss": 1.3057, + "step": 195 + }, + { + "epoch": 0.0349002849002849, + "grad_norm": 0.3954029381275177, + "learning_rate": 0.00019996965601602355, + "loss": 1.1958, + "step": 196 + }, + { + "epoch": 0.03507834757834758, + "grad_norm": 0.34454262256622314, + "learning_rate": 0.00019996931023611572, + "loss": 1.0972, + "step": 197 + }, + { + "epoch": 0.035256410256410256, + "grad_norm": 0.48900291323661804, + "learning_rate": 0.0001999689624975169, + "loss": 1.213, + "step": 198 + }, + { + "epoch": 0.03543447293447293, + "grad_norm": 0.35214388370513916, + "learning_rate": 0.00019996861280023397, + "loss": 1.0285, + "step": 199 + }, + { + "epoch": 0.03561253561253561, + "grad_norm": 0.49393126368522644, + "learning_rate": 0.00019996826114427373, + "loss": 1.2313, + "step": 200 + }, + { + "epoch": 0.03579059829059829, + "grad_norm": 0.3994458019733429, + "learning_rate": 0.00019996790752964305, + "loss": 1.0474, + "step": 201 + }, + { + "epoch": 0.03596866096866097, + "grad_norm": 0.5387318730354309, + "learning_rate": 0.0001999675519563489, + "loss": 1.3067, + "step": 202 + }, + { + "epoch": 0.036146723646723646, + "grad_norm": 0.4976751208305359, + "learning_rate": 0.00019996719442439824, + "loss": 1.2593, + "step": 203 + }, + { + "epoch": 0.03632478632478633, + "grad_norm": 0.47052907943725586, + "learning_rate": 0.0001999668349337981, + "loss": 1.1036, + "step": 204 + }, + { + "epoch": 0.036502849002849, + "grad_norm": 0.39616644382476807, + "learning_rate": 0.00019996647348455543, + "loss": 1.0481, + "step": 205 + }, + { + "epoch": 0.03668091168091168, + "grad_norm": 0.42987677454948425, + "learning_rate": 0.00019996611007667742, + "loss": 1.0923, + "step": 206 + }, + { + "epoch": 0.03685897435897436, + "grad_norm": 0.47065848112106323, + "learning_rate": 0.00019996574471017113, + "loss": 1.1403, + "step": 207 + }, + { + "epoch": 0.037037037037037035, + "grad_norm": 0.4363015592098236, + "learning_rate": 0.00019996537738504373, + "loss": 1.253, + "step": 208 + }, + { + "epoch": 0.03721509971509972, + "grad_norm": 0.4038296937942505, + "learning_rate": 0.00019996500810130243, + "loss": 1.1679, + "step": 209 + }, + { + "epoch": 0.03739316239316239, + "grad_norm": 0.5038532018661499, + "learning_rate": 0.00019996463685895445, + "loss": 1.1182, + "step": 210 + }, + { + "epoch": 0.037571225071225074, + "grad_norm": 0.37740692496299744, + "learning_rate": 0.00019996426365800706, + "loss": 1.0465, + "step": 211 + }, + { + "epoch": 0.03774928774928775, + "grad_norm": 0.47794604301452637, + "learning_rate": 0.00019996388849846759, + "loss": 1.2836, + "step": 212 + }, + { + "epoch": 0.037927350427350424, + "grad_norm": 0.38460609316825867, + "learning_rate": 0.0001999635113803434, + "loss": 1.2099, + "step": 213 + }, + { + "epoch": 0.038105413105413107, + "grad_norm": 0.42016157507896423, + "learning_rate": 0.0001999631323036418, + "loss": 1.152, + "step": 214 + }, + { + "epoch": 0.03828347578347578, + "grad_norm": 0.4024946391582489, + "learning_rate": 0.00019996275126837033, + "loss": 1.1534, + "step": 215 + }, + { + "epoch": 0.038461538461538464, + "grad_norm": 0.4573793411254883, + "learning_rate": 0.00019996236827453642, + "loss": 1.2019, + "step": 216 + }, + { + "epoch": 0.03863960113960114, + "grad_norm": 0.3642503321170807, + "learning_rate": 0.0001999619833221475, + "loss": 1.0541, + "step": 217 + }, + { + "epoch": 0.03881766381766382, + "grad_norm": 0.38492897152900696, + "learning_rate": 0.0001999615964112112, + "loss": 1.1269, + "step": 218 + }, + { + "epoch": 0.038995726495726496, + "grad_norm": 0.427219420671463, + "learning_rate": 0.0001999612075417351, + "loss": 1.1126, + "step": 219 + }, + { + "epoch": 0.03917378917378917, + "grad_norm": 0.40781742334365845, + "learning_rate": 0.00019996081671372676, + "loss": 1.2207, + "step": 220 + }, + { + "epoch": 0.03935185185185185, + "grad_norm": 0.39229512214660645, + "learning_rate": 0.00019996042392719386, + "loss": 1.0403, + "step": 221 + }, + { + "epoch": 0.03952991452991453, + "grad_norm": 0.42038577795028687, + "learning_rate": 0.0001999600291821441, + "loss": 1.2157, + "step": 222 + }, + { + "epoch": 0.03970797720797721, + "grad_norm": 0.3963491916656494, + "learning_rate": 0.00019995963247858525, + "loss": 1.0532, + "step": 223 + }, + { + "epoch": 0.039886039886039885, + "grad_norm": 0.4389874041080475, + "learning_rate": 0.00019995923381652502, + "loss": 1.4279, + "step": 224 + }, + { + "epoch": 0.04006410256410257, + "grad_norm": 0.357312947511673, + "learning_rate": 0.00019995883319597123, + "loss": 0.9871, + "step": 225 + }, + { + "epoch": 0.04024216524216524, + "grad_norm": 0.3644427955150604, + "learning_rate": 0.00019995843061693181, + "loss": 1.0879, + "step": 226 + }, + { + "epoch": 0.04042022792022792, + "grad_norm": 0.4074651002883911, + "learning_rate": 0.00019995802607941453, + "loss": 1.2138, + "step": 227 + }, + { + "epoch": 0.0405982905982906, + "grad_norm": 0.40709465742111206, + "learning_rate": 0.0001999576195834274, + "loss": 1.1905, + "step": 228 + }, + { + "epoch": 0.040776353276353275, + "grad_norm": 0.4280182719230652, + "learning_rate": 0.00019995721112897838, + "loss": 1.2331, + "step": 229 + }, + { + "epoch": 0.04095441595441596, + "grad_norm": 0.37846076488494873, + "learning_rate": 0.00019995680071607544, + "loss": 1.078, + "step": 230 + }, + { + "epoch": 0.04113247863247863, + "grad_norm": 0.3877260088920593, + "learning_rate": 0.0001999563883447266, + "loss": 1.0309, + "step": 231 + }, + { + "epoch": 0.04131054131054131, + "grad_norm": 0.42886826395988464, + "learning_rate": 0.00019995597401494, + "loss": 1.0403, + "step": 232 + }, + { + "epoch": 0.04148860398860399, + "grad_norm": 0.4316534101963043, + "learning_rate": 0.00019995555772672372, + "loss": 1.2418, + "step": 233 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 0.45768865942955017, + "learning_rate": 0.00019995513948008593, + "loss": 1.233, + "step": 234 + }, + { + "epoch": 0.041844729344729346, + "grad_norm": 0.5647913813591003, + "learning_rate": 0.00019995471927503481, + "loss": 1.1346, + "step": 235 + }, + { + "epoch": 0.04202279202279202, + "grad_norm": 0.3797492980957031, + "learning_rate": 0.00019995429711157863, + "loss": 1.1574, + "step": 236 + }, + { + "epoch": 0.042200854700854704, + "grad_norm": 0.4392767548561096, + "learning_rate": 0.00019995387298972562, + "loss": 0.8988, + "step": 237 + }, + { + "epoch": 0.04237891737891738, + "grad_norm": 0.37331557273864746, + "learning_rate": 0.0001999534469094841, + "loss": 1.0439, + "step": 238 + }, + { + "epoch": 0.042556980056980054, + "grad_norm": 0.3785935938358307, + "learning_rate": 0.00019995301887086245, + "loss": 0.9839, + "step": 239 + }, + { + "epoch": 0.042735042735042736, + "grad_norm": 0.4351862668991089, + "learning_rate": 0.00019995258887386898, + "loss": 1.2653, + "step": 240 + }, + { + "epoch": 0.04291310541310541, + "grad_norm": 0.399475634098053, + "learning_rate": 0.0001999521569185122, + "loss": 0.9877, + "step": 241 + }, + { + "epoch": 0.04309116809116809, + "grad_norm": 0.42332810163497925, + "learning_rate": 0.00019995172300480053, + "loss": 1.2403, + "step": 242 + }, + { + "epoch": 0.04326923076923077, + "grad_norm": 0.4397708475589752, + "learning_rate": 0.00019995128713274247, + "loss": 0.9316, + "step": 243 + }, + { + "epoch": 0.04344729344729345, + "grad_norm": 0.3614110052585602, + "learning_rate": 0.00019995084930234658, + "loss": 1.1088, + "step": 244 + }, + { + "epoch": 0.043625356125356125, + "grad_norm": 0.39433717727661133, + "learning_rate": 0.0001999504095136214, + "loss": 1.2002, + "step": 245 + }, + { + "epoch": 0.0438034188034188, + "grad_norm": 0.33088216185569763, + "learning_rate": 0.0001999499677665756, + "loss": 0.8796, + "step": 246 + }, + { + "epoch": 0.04398148148148148, + "grad_norm": 0.5239143967628479, + "learning_rate": 0.00019994952406121784, + "loss": 1.2808, + "step": 247 + }, + { + "epoch": 0.04415954415954416, + "grad_norm": 0.42156723141670227, + "learning_rate": 0.00019994907839755675, + "loss": 1.1775, + "step": 248 + }, + { + "epoch": 0.04433760683760684, + "grad_norm": 0.42569902539253235, + "learning_rate": 0.0001999486307756011, + "loss": 1.001, + "step": 249 + }, + { + "epoch": 0.044515669515669515, + "grad_norm": 0.38241544365882874, + "learning_rate": 0.00019994818119535964, + "loss": 1.1064, + "step": 250 + }, + { + "epoch": 0.0446937321937322, + "grad_norm": 0.4185071885585785, + "learning_rate": 0.0001999477296568412, + "loss": 1.2109, + "step": 251 + }, + { + "epoch": 0.04487179487179487, + "grad_norm": 0.4189644157886505, + "learning_rate": 0.00019994727616005464, + "loss": 1.2902, + "step": 252 + }, + { + "epoch": 0.04504985754985755, + "grad_norm": 0.34671884775161743, + "learning_rate": 0.0001999468207050088, + "loss": 0.9429, + "step": 253 + }, + { + "epoch": 0.04522792022792023, + "grad_norm": 0.42391687631607056, + "learning_rate": 0.00019994636329171266, + "loss": 0.7179, + "step": 254 + }, + { + "epoch": 0.045405982905982904, + "grad_norm": 0.3803195655345917, + "learning_rate": 0.00019994590392017513, + "loss": 1.0318, + "step": 255 + }, + { + "epoch": 0.045584045584045586, + "grad_norm": 0.3389956057071686, + "learning_rate": 0.00019994544259040525, + "loss": 1.0485, + "step": 256 + }, + { + "epoch": 0.04576210826210826, + "grad_norm": 0.4927038550376892, + "learning_rate": 0.000199944979302412, + "loss": 1.3426, + "step": 257 + }, + { + "epoch": 0.045940170940170943, + "grad_norm": 0.33200421929359436, + "learning_rate": 0.00019994451405620453, + "loss": 1.0071, + "step": 258 + }, + { + "epoch": 0.04611823361823362, + "grad_norm": 0.38028615713119507, + "learning_rate": 0.00019994404685179195, + "loss": 1.0985, + "step": 259 + }, + { + "epoch": 0.046296296296296294, + "grad_norm": 0.3752151429653168, + "learning_rate": 0.00019994357768918333, + "loss": 0.9209, + "step": 260 + }, + { + "epoch": 0.046474358974358976, + "grad_norm": 0.43030866980552673, + "learning_rate": 0.00019994310656838796, + "loss": 0.9921, + "step": 261 + }, + { + "epoch": 0.04665242165242165, + "grad_norm": 0.4402460753917694, + "learning_rate": 0.00019994263348941502, + "loss": 1.1051, + "step": 262 + }, + { + "epoch": 0.04683048433048433, + "grad_norm": 0.43012720346450806, + "learning_rate": 0.0001999421584522738, + "loss": 1.1839, + "step": 263 + }, + { + "epoch": 0.04700854700854701, + "grad_norm": 0.4195305407047272, + "learning_rate": 0.0001999416814569736, + "loss": 1.1749, + "step": 264 + }, + { + "epoch": 0.04718660968660968, + "grad_norm": 0.45623287558555603, + "learning_rate": 0.00019994120250352372, + "loss": 1.2433, + "step": 265 + }, + { + "epoch": 0.047364672364672365, + "grad_norm": 0.4736156761646271, + "learning_rate": 0.00019994072159193363, + "loss": 1.2882, + "step": 266 + }, + { + "epoch": 0.04754273504273504, + "grad_norm": 0.36698561906814575, + "learning_rate": 0.0001999402387222127, + "loss": 1.1486, + "step": 267 + }, + { + "epoch": 0.04772079772079772, + "grad_norm": 0.3854144215583801, + "learning_rate": 0.00019993975389437038, + "loss": 0.8115, + "step": 268 + }, + { + "epoch": 0.0478988603988604, + "grad_norm": 0.41512808203697205, + "learning_rate": 0.0001999392671084162, + "loss": 1.0959, + "step": 269 + }, + { + "epoch": 0.04807692307692308, + "grad_norm": 0.3869563341140747, + "learning_rate": 0.0001999387783643597, + "loss": 1.087, + "step": 270 + }, + { + "epoch": 0.048254985754985755, + "grad_norm": 0.4649744927883148, + "learning_rate": 0.00019993828766221044, + "loss": 1.0011, + "step": 271 + }, + { + "epoch": 0.04843304843304843, + "grad_norm": 0.40331923961639404, + "learning_rate": 0.00019993779500197803, + "loss": 1.1463, + "step": 272 + }, + { + "epoch": 0.04861111111111111, + "grad_norm": 0.3826279938220978, + "learning_rate": 0.0001999373003836721, + "loss": 1.1491, + "step": 273 + }, + { + "epoch": 0.04878917378917379, + "grad_norm": 0.3967166543006897, + "learning_rate": 0.00019993680380730243, + "loss": 1.1462, + "step": 274 + }, + { + "epoch": 0.04896723646723647, + "grad_norm": 0.4298507869243622, + "learning_rate": 0.00019993630527287865, + "loss": 1.2471, + "step": 275 + }, + { + "epoch": 0.049145299145299144, + "grad_norm": 0.41486215591430664, + "learning_rate": 0.0001999358047804106, + "loss": 1.287, + "step": 276 + }, + { + "epoch": 0.049323361823361826, + "grad_norm": 0.3914124369621277, + "learning_rate": 0.00019993530232990803, + "loss": 1.0935, + "step": 277 + }, + { + "epoch": 0.0495014245014245, + "grad_norm": 0.39888378977775574, + "learning_rate": 0.00019993479792138082, + "loss": 1.2347, + "step": 278 + }, + { + "epoch": 0.049679487179487176, + "grad_norm": 0.3911665678024292, + "learning_rate": 0.00019993429155483884, + "loss": 1.0917, + "step": 279 + }, + { + "epoch": 0.04985754985754986, + "grad_norm": 0.42871445417404175, + "learning_rate": 0.00019993378323029197, + "loss": 1.0277, + "step": 280 + }, + { + "epoch": 0.050035612535612534, + "grad_norm": 0.35397860407829285, + "learning_rate": 0.00019993327294775027, + "loss": 0.9549, + "step": 281 + }, + { + "epoch": 0.050213675213675216, + "grad_norm": 0.4528059959411621, + "learning_rate": 0.00019993276070722364, + "loss": 1.2338, + "step": 282 + }, + { + "epoch": 0.05039173789173789, + "grad_norm": 0.354735791683197, + "learning_rate": 0.00019993224650872218, + "loss": 1.1892, + "step": 283 + }, + { + "epoch": 0.05056980056980057, + "grad_norm": 0.44407567381858826, + "learning_rate": 0.00019993173035225592, + "loss": 1.1621, + "step": 284 + }, + { + "epoch": 0.05074786324786325, + "grad_norm": 0.4177244305610657, + "learning_rate": 0.000199931212237835, + "loss": 1.1184, + "step": 285 + }, + { + "epoch": 0.05092592592592592, + "grad_norm": 0.5627759695053101, + "learning_rate": 0.0001999306921654696, + "loss": 1.0755, + "step": 286 + }, + { + "epoch": 0.051103988603988605, + "grad_norm": 0.46767523884773254, + "learning_rate": 0.00019993017013516986, + "loss": 1.2654, + "step": 287 + }, + { + "epoch": 0.05128205128205128, + "grad_norm": 0.4163128733634949, + "learning_rate": 0.000199929646146946, + "loss": 1.1307, + "step": 288 + }, + { + "epoch": 0.05146011396011396, + "grad_norm": 0.36954161524772644, + "learning_rate": 0.00019992912020080832, + "loss": 0.8274, + "step": 289 + }, + { + "epoch": 0.05163817663817664, + "grad_norm": 0.4770594835281372, + "learning_rate": 0.00019992859229676712, + "loss": 1.2235, + "step": 290 + }, + { + "epoch": 0.05181623931623932, + "grad_norm": 0.4174608290195465, + "learning_rate": 0.00019992806243483274, + "loss": 1.2893, + "step": 291 + }, + { + "epoch": 0.051994301994301995, + "grad_norm": 0.3794898986816406, + "learning_rate": 0.00019992753061501555, + "loss": 1.104, + "step": 292 + }, + { + "epoch": 0.05217236467236467, + "grad_norm": 0.3912592828273773, + "learning_rate": 0.000199926996837326, + "loss": 1.0043, + "step": 293 + }, + { + "epoch": 0.05235042735042735, + "grad_norm": 0.39641159772872925, + "learning_rate": 0.00019992646110177448, + "loss": 1.083, + "step": 294 + }, + { + "epoch": 0.05252849002849003, + "grad_norm": 0.3518857955932617, + "learning_rate": 0.00019992592340837157, + "loss": 0.9275, + "step": 295 + }, + { + "epoch": 0.05270655270655271, + "grad_norm": 0.3955721855163574, + "learning_rate": 0.00019992538375712777, + "loss": 1.0153, + "step": 296 + }, + { + "epoch": 0.052884615384615384, + "grad_norm": 0.3837333023548126, + "learning_rate": 0.00019992484214805364, + "loss": 1.1664, + "step": 297 + }, + { + "epoch": 0.053062678062678066, + "grad_norm": 0.39400920271873474, + "learning_rate": 0.0001999242985811598, + "loss": 1.0532, + "step": 298 + }, + { + "epoch": 0.05324074074074074, + "grad_norm": 0.39258649945259094, + "learning_rate": 0.00019992375305645692, + "loss": 1.0081, + "step": 299 + }, + { + "epoch": 0.053418803418803416, + "grad_norm": 0.49768248200416565, + "learning_rate": 0.00019992320557395566, + "loss": 1.2553, + "step": 300 + }, + { + "epoch": 0.0535968660968661, + "grad_norm": 0.364776074886322, + "learning_rate": 0.00019992265613366677, + "loss": 1.0582, + "step": 301 + }, + { + "epoch": 0.053774928774928774, + "grad_norm": 0.47317907214164734, + "learning_rate": 0.00019992210473560097, + "loss": 1.3114, + "step": 302 + }, + { + "epoch": 0.053952991452991456, + "grad_norm": 0.3706119656562805, + "learning_rate": 0.00019992155137976917, + "loss": 0.9554, + "step": 303 + }, + { + "epoch": 0.05413105413105413, + "grad_norm": 0.42809563875198364, + "learning_rate": 0.0001999209960661821, + "loss": 1.306, + "step": 304 + }, + { + "epoch": 0.054309116809116806, + "grad_norm": 0.4514487385749817, + "learning_rate": 0.00019992043879485066, + "loss": 1.0147, + "step": 305 + }, + { + "epoch": 0.05448717948717949, + "grad_norm": 0.36672836542129517, + "learning_rate": 0.0001999198795657858, + "loss": 1.1392, + "step": 306 + }, + { + "epoch": 0.05466524216524216, + "grad_norm": 0.4206554889678955, + "learning_rate": 0.00019991931837899847, + "loss": 1.2405, + "step": 307 + }, + { + "epoch": 0.054843304843304845, + "grad_norm": 0.46168261766433716, + "learning_rate": 0.00019991875523449966, + "loss": 1.2707, + "step": 308 + }, + { + "epoch": 0.05502136752136752, + "grad_norm": 0.39503365755081177, + "learning_rate": 0.00019991819013230039, + "loss": 1.0776, + "step": 309 + }, + { + "epoch": 0.0551994301994302, + "grad_norm": 0.35244834423065186, + "learning_rate": 0.00019991762307241178, + "loss": 1.0864, + "step": 310 + }, + { + "epoch": 0.05537749287749288, + "grad_norm": 0.3865319490432739, + "learning_rate": 0.0001999170540548449, + "loss": 1.3659, + "step": 311 + }, + { + "epoch": 0.05555555555555555, + "grad_norm": 0.3666876554489136, + "learning_rate": 0.0001999164830796109, + "loss": 0.9884, + "step": 312 + }, + { + "epoch": 0.055733618233618235, + "grad_norm": 0.4278281629085541, + "learning_rate": 0.00019991591014672096, + "loss": 1.1522, + "step": 313 + }, + { + "epoch": 0.05591168091168091, + "grad_norm": 0.4172627031803131, + "learning_rate": 0.0001999153352561863, + "loss": 1.2527, + "step": 314 + }, + { + "epoch": 0.05608974358974359, + "grad_norm": 0.38872212171554565, + "learning_rate": 0.00019991475840801823, + "loss": 1.2985, + "step": 315 + }, + { + "epoch": 0.05626780626780627, + "grad_norm": 0.4160458445549011, + "learning_rate": 0.00019991417960222804, + "loss": 1.1347, + "step": 316 + }, + { + "epoch": 0.05644586894586895, + "grad_norm": 0.5169723033905029, + "learning_rate": 0.00019991359883882705, + "loss": 1.0819, + "step": 317 + }, + { + "epoch": 0.056623931623931624, + "grad_norm": 0.42306259274482727, + "learning_rate": 0.0001999130161178266, + "loss": 1.3139, + "step": 318 + }, + { + "epoch": 0.0568019943019943, + "grad_norm": 0.41975873708724976, + "learning_rate": 0.00019991243143923816, + "loss": 1.2277, + "step": 319 + }, + { + "epoch": 0.05698005698005698, + "grad_norm": 0.3873472511768341, + "learning_rate": 0.00019991184480307324, + "loss": 1.156, + "step": 320 + }, + { + "epoch": 0.057158119658119656, + "grad_norm": 0.43656104803085327, + "learning_rate": 0.0001999112562093432, + "loss": 1.2344, + "step": 321 + }, + { + "epoch": 0.05733618233618234, + "grad_norm": 0.3738791048526764, + "learning_rate": 0.00019991066565805968, + "loss": 0.9573, + "step": 322 + }, + { + "epoch": 0.05751424501424501, + "grad_norm": 0.3838156461715698, + "learning_rate": 0.00019991007314923418, + "loss": 0.9274, + "step": 323 + }, + { + "epoch": 0.057692307692307696, + "grad_norm": 0.4564770758152008, + "learning_rate": 0.00019990947868287837, + "loss": 1.0756, + "step": 324 + }, + { + "epoch": 0.05787037037037037, + "grad_norm": 0.4560079872608185, + "learning_rate": 0.00019990888225900386, + "loss": 1.1508, + "step": 325 + }, + { + "epoch": 0.058048433048433046, + "grad_norm": 0.44356057047843933, + "learning_rate": 0.00019990828387762236, + "loss": 1.2323, + "step": 326 + }, + { + "epoch": 0.05822649572649573, + "grad_norm": 0.46390119194984436, + "learning_rate": 0.00019990768353874553, + "loss": 1.0031, + "step": 327 + }, + { + "epoch": 0.0584045584045584, + "grad_norm": 0.4502357244491577, + "learning_rate": 0.00019990708124238525, + "loss": 1.3454, + "step": 328 + }, + { + "epoch": 0.058582621082621085, + "grad_norm": 0.3979945182800293, + "learning_rate": 0.0001999064769885532, + "loss": 1.2833, + "step": 329 + }, + { + "epoch": 0.05876068376068376, + "grad_norm": 0.3899286687374115, + "learning_rate": 0.00019990587077726128, + "loss": 1.0175, + "step": 330 + }, + { + "epoch": 0.05893874643874644, + "grad_norm": 0.41422948241233826, + "learning_rate": 0.00019990526260852139, + "loss": 1.1151, + "step": 331 + }, + { + "epoch": 0.05911680911680912, + "grad_norm": 0.4266608953475952, + "learning_rate": 0.0001999046524823454, + "loss": 1.1119, + "step": 332 + }, + { + "epoch": 0.05929487179487179, + "grad_norm": 0.46563324332237244, + "learning_rate": 0.00019990404039874524, + "loss": 1.2358, + "step": 333 + }, + { + "epoch": 0.059472934472934474, + "grad_norm": 0.4404347240924835, + "learning_rate": 0.00019990342635773297, + "loss": 1.1748, + "step": 334 + }, + { + "epoch": 0.05965099715099715, + "grad_norm": 0.5133237838745117, + "learning_rate": 0.00019990281035932062, + "loss": 1.1649, + "step": 335 + }, + { + "epoch": 0.05982905982905983, + "grad_norm": 0.3593895435333252, + "learning_rate": 0.00019990219240352018, + "loss": 1.0318, + "step": 336 + }, + { + "epoch": 0.06000712250712251, + "grad_norm": 0.40554583072662354, + "learning_rate": 0.00019990157249034384, + "loss": 1.1202, + "step": 337 + }, + { + "epoch": 0.06018518518518518, + "grad_norm": 0.3770706057548523, + "learning_rate": 0.00019990095061980372, + "loss": 0.9908, + "step": 338 + }, + { + "epoch": 0.060363247863247864, + "grad_norm": 0.39676955342292786, + "learning_rate": 0.000199900326791912, + "loss": 0.8176, + "step": 339 + }, + { + "epoch": 0.06054131054131054, + "grad_norm": 0.41448578238487244, + "learning_rate": 0.00019989970100668086, + "loss": 1.2877, + "step": 340 + }, + { + "epoch": 0.06071937321937322, + "grad_norm": 0.4200015068054199, + "learning_rate": 0.00019989907326412265, + "loss": 1.2293, + "step": 341 + }, + { + "epoch": 0.060897435897435896, + "grad_norm": 0.47350621223449707, + "learning_rate": 0.0001998984435642496, + "loss": 1.2331, + "step": 342 + }, + { + "epoch": 0.06107549857549858, + "grad_norm": 0.47050634026527405, + "learning_rate": 0.00019989781190707406, + "loss": 0.8888, + "step": 343 + }, + { + "epoch": 0.06125356125356125, + "grad_norm": 0.4994896948337555, + "learning_rate": 0.00019989717829260842, + "loss": 1.0921, + "step": 344 + }, + { + "epoch": 0.06143162393162393, + "grad_norm": 0.36340200901031494, + "learning_rate": 0.0001998965427208651, + "loss": 0.9777, + "step": 345 + }, + { + "epoch": 0.06160968660968661, + "grad_norm": 0.3538152873516083, + "learning_rate": 0.00019989590519185654, + "loss": 1.0055, + "step": 346 + }, + { + "epoch": 0.061787749287749286, + "grad_norm": 0.5388944149017334, + "learning_rate": 0.00019989526570559526, + "loss": 1.1001, + "step": 347 + }, + { + "epoch": 0.06196581196581197, + "grad_norm": 0.4411574602127075, + "learning_rate": 0.00019989462426209373, + "loss": 1.0038, + "step": 348 + }, + { + "epoch": 0.06214387464387464, + "grad_norm": 0.3930876851081848, + "learning_rate": 0.00019989398086136455, + "loss": 1.1534, + "step": 349 + }, + { + "epoch": 0.062321937321937325, + "grad_norm": 0.47357070446014404, + "learning_rate": 0.00019989333550342033, + "loss": 1.2687, + "step": 350 + }, + { + "epoch": 0.0625, + "grad_norm": 0.40302303433418274, + "learning_rate": 0.00019989268818827372, + "loss": 1.1894, + "step": 351 + }, + { + "epoch": 0.06267806267806268, + "grad_norm": 0.4470510184764862, + "learning_rate": 0.00019989203891593738, + "loss": 1.2207, + "step": 352 + }, + { + "epoch": 0.06285612535612535, + "grad_norm": 0.42235100269317627, + "learning_rate": 0.00019989138768642406, + "loss": 1.2086, + "step": 353 + }, + { + "epoch": 0.06303418803418803, + "grad_norm": 0.38305309414863586, + "learning_rate": 0.0001998907344997465, + "loss": 1.0473, + "step": 354 + }, + { + "epoch": 0.06321225071225071, + "grad_norm": 0.3893027901649475, + "learning_rate": 0.0001998900793559175, + "loss": 1.1746, + "step": 355 + }, + { + "epoch": 0.0633903133903134, + "grad_norm": 0.41206735372543335, + "learning_rate": 0.0001998894222549499, + "loss": 1.188, + "step": 356 + }, + { + "epoch": 0.06356837606837606, + "grad_norm": 0.3700513243675232, + "learning_rate": 0.00019988876319685658, + "loss": 0.9862, + "step": 357 + }, + { + "epoch": 0.06374643874643875, + "grad_norm": 0.3708794116973877, + "learning_rate": 0.0001998881021816504, + "loss": 1.2003, + "step": 358 + }, + { + "epoch": 0.06392450142450143, + "grad_norm": 0.4058014154434204, + "learning_rate": 0.00019988743920934442, + "loss": 1.2311, + "step": 359 + }, + { + "epoch": 0.0641025641025641, + "grad_norm": 0.39134132862091064, + "learning_rate": 0.00019988677427995155, + "loss": 1.001, + "step": 360 + }, + { + "epoch": 0.06428062678062678, + "grad_norm": 0.3853437602519989, + "learning_rate": 0.00019988610739348484, + "loss": 1.0725, + "step": 361 + }, + { + "epoch": 0.06445868945868946, + "grad_norm": 0.47114330530166626, + "learning_rate": 0.00019988543854995735, + "loss": 1.2196, + "step": 362 + }, + { + "epoch": 0.06463675213675214, + "grad_norm": 0.40465688705444336, + "learning_rate": 0.00019988476774938216, + "loss": 1.1869, + "step": 363 + }, + { + "epoch": 0.06481481481481481, + "grad_norm": 0.40301886200904846, + "learning_rate": 0.00019988409499177245, + "loss": 1.1765, + "step": 364 + }, + { + "epoch": 0.0649928774928775, + "grad_norm": 0.43443185091018677, + "learning_rate": 0.0001998834202771414, + "loss": 1.2022, + "step": 365 + }, + { + "epoch": 0.06517094017094018, + "grad_norm": 0.4712986350059509, + "learning_rate": 0.00019988274360550217, + "loss": 1.156, + "step": 366 + }, + { + "epoch": 0.06534900284900284, + "grad_norm": 0.4524450898170471, + "learning_rate": 0.00019988206497686815, + "loss": 1.2917, + "step": 367 + }, + { + "epoch": 0.06552706552706553, + "grad_norm": 0.40302205085754395, + "learning_rate": 0.0001998813843912525, + "loss": 0.9993, + "step": 368 + }, + { + "epoch": 0.06570512820512821, + "grad_norm": 0.39435216784477234, + "learning_rate": 0.00019988070184866864, + "loss": 1.0914, + "step": 369 + }, + { + "epoch": 0.06588319088319089, + "grad_norm": 0.39267390966415405, + "learning_rate": 0.00019988001734912988, + "loss": 1.3138, + "step": 370 + }, + { + "epoch": 0.06606125356125356, + "grad_norm": 0.38351675868034363, + "learning_rate": 0.00019987933089264968, + "loss": 1.0997, + "step": 371 + }, + { + "epoch": 0.06623931623931624, + "grad_norm": 0.3294839859008789, + "learning_rate": 0.00019987864247924145, + "loss": 0.9656, + "step": 372 + }, + { + "epoch": 0.06641737891737892, + "grad_norm": 0.45333364605903625, + "learning_rate": 0.00019987795210891872, + "loss": 1.095, + "step": 373 + }, + { + "epoch": 0.06659544159544159, + "grad_norm": 0.4362282454967499, + "learning_rate": 0.00019987725978169501, + "loss": 1.2103, + "step": 374 + }, + { + "epoch": 0.06677350427350427, + "grad_norm": 0.41314780712127686, + "learning_rate": 0.00019987656549758385, + "loss": 1.2115, + "step": 375 + }, + { + "epoch": 0.06695156695156695, + "grad_norm": 0.4230864644050598, + "learning_rate": 0.00019987586925659888, + "loss": 1.17, + "step": 376 + }, + { + "epoch": 0.06712962962962964, + "grad_norm": 0.4703855812549591, + "learning_rate": 0.00019987517105875372, + "loss": 1.367, + "step": 377 + }, + { + "epoch": 0.0673076923076923, + "grad_norm": 0.4671297073364258, + "learning_rate": 0.00019987447090406206, + "loss": 1.2543, + "step": 378 + }, + { + "epoch": 0.06748575498575499, + "grad_norm": 0.43746981024742126, + "learning_rate": 0.0001998737687925376, + "loss": 1.214, + "step": 379 + }, + { + "epoch": 0.06766381766381767, + "grad_norm": 0.40889596939086914, + "learning_rate": 0.00019987306472419412, + "loss": 1.0496, + "step": 380 + }, + { + "epoch": 0.06784188034188034, + "grad_norm": 0.3677358627319336, + "learning_rate": 0.0001998723586990454, + "loss": 1.1242, + "step": 381 + }, + { + "epoch": 0.06801994301994302, + "grad_norm": 0.3892628848552704, + "learning_rate": 0.00019987165071710527, + "loss": 1.0246, + "step": 382 + }, + { + "epoch": 0.0681980056980057, + "grad_norm": 0.4281293749809265, + "learning_rate": 0.00019987094077838764, + "loss": 1.2817, + "step": 383 + }, + { + "epoch": 0.06837606837606838, + "grad_norm": 0.45030340552330017, + "learning_rate": 0.00019987022888290636, + "loss": 1.159, + "step": 384 + }, + { + "epoch": 0.06855413105413105, + "grad_norm": 0.6327905058860779, + "learning_rate": 0.00019986951503067545, + "loss": 0.9577, + "step": 385 + }, + { + "epoch": 0.06873219373219373, + "grad_norm": 0.40339627861976624, + "learning_rate": 0.0001998687992217088, + "loss": 1.138, + "step": 386 + }, + { + "epoch": 0.06891025641025642, + "grad_norm": 0.4018291234970093, + "learning_rate": 0.00019986808145602052, + "loss": 0.9109, + "step": 387 + }, + { + "epoch": 0.06908831908831908, + "grad_norm": 0.41566264629364014, + "learning_rate": 0.00019986736173362464, + "loss": 1.1516, + "step": 388 + }, + { + "epoch": 0.06926638176638177, + "grad_norm": 0.3569067418575287, + "learning_rate": 0.00019986664005453527, + "loss": 1.2329, + "step": 389 + }, + { + "epoch": 0.06944444444444445, + "grad_norm": 0.3959648907184601, + "learning_rate": 0.0001998659164187665, + "loss": 1.1041, + "step": 390 + }, + { + "epoch": 0.06962250712250712, + "grad_norm": 0.42853206396102905, + "learning_rate": 0.00019986519082633257, + "loss": 1.0859, + "step": 391 + }, + { + "epoch": 0.0698005698005698, + "grad_norm": 0.42005518078804016, + "learning_rate": 0.0001998644632772477, + "loss": 1.2017, + "step": 392 + }, + { + "epoch": 0.06997863247863248, + "grad_norm": 0.4296947419643402, + "learning_rate": 0.00019986373377152612, + "loss": 1.1464, + "step": 393 + }, + { + "epoch": 0.07015669515669516, + "grad_norm": 0.394747793674469, + "learning_rate": 0.0001998630023091821, + "loss": 1.0316, + "step": 394 + }, + { + "epoch": 0.07033475783475783, + "grad_norm": 0.3779357969760895, + "learning_rate": 0.00019986226889023002, + "loss": 1.1081, + "step": 395 + }, + { + "epoch": 0.07051282051282051, + "grad_norm": 0.4271804690361023, + "learning_rate": 0.00019986153351468424, + "loss": 0.985, + "step": 396 + }, + { + "epoch": 0.0706908831908832, + "grad_norm": 0.49412235617637634, + "learning_rate": 0.00019986079618255912, + "loss": 1.2606, + "step": 397 + }, + { + "epoch": 0.07086894586894586, + "grad_norm": 0.43657439947128296, + "learning_rate": 0.00019986005689386915, + "loss": 1.2266, + "step": 398 + }, + { + "epoch": 0.07104700854700854, + "grad_norm": 0.4060729444026947, + "learning_rate": 0.0001998593156486288, + "loss": 1.1787, + "step": 399 + }, + { + "epoch": 0.07122507122507123, + "grad_norm": 0.387046217918396, + "learning_rate": 0.00019985857244685264, + "loss": 0.9411, + "step": 400 + }, + { + "epoch": 0.07140313390313391, + "grad_norm": 0.4243999123573303, + "learning_rate": 0.00019985782728855516, + "loss": 1.2024, + "step": 401 + }, + { + "epoch": 0.07158119658119658, + "grad_norm": 0.43113812804222107, + "learning_rate": 0.000199857080173751, + "loss": 1.1246, + "step": 402 + }, + { + "epoch": 0.07175925925925926, + "grad_norm": 0.4653271436691284, + "learning_rate": 0.0001998563311024548, + "loss": 1.2343, + "step": 403 + }, + { + "epoch": 0.07193732193732194, + "grad_norm": 0.43260812759399414, + "learning_rate": 0.0001998555800746812, + "loss": 0.9543, + "step": 404 + }, + { + "epoch": 0.07211538461538461, + "grad_norm": 0.4635484516620636, + "learning_rate": 0.00019985482709044495, + "loss": 1.1091, + "step": 405 + }, + { + "epoch": 0.07229344729344729, + "grad_norm": 0.38362643122673035, + "learning_rate": 0.00019985407214976076, + "loss": 1.2584, + "step": 406 + }, + { + "epoch": 0.07247150997150997, + "grad_norm": 0.4068310558795929, + "learning_rate": 0.00019985331525264351, + "loss": 1.1944, + "step": 407 + }, + { + "epoch": 0.07264957264957266, + "grad_norm": 0.43909943103790283, + "learning_rate": 0.00019985255639910795, + "loss": 1.3748, + "step": 408 + }, + { + "epoch": 0.07282763532763532, + "grad_norm": 0.48674601316452026, + "learning_rate": 0.000199851795589169, + "loss": 1.2684, + "step": 409 + }, + { + "epoch": 0.073005698005698, + "grad_norm": 0.4218580722808838, + "learning_rate": 0.0001998510328228415, + "loss": 1.168, + "step": 410 + }, + { + "epoch": 0.07318376068376069, + "grad_norm": 0.4688236117362976, + "learning_rate": 0.00019985026810014046, + "loss": 1.3088, + "step": 411 + }, + { + "epoch": 0.07336182336182336, + "grad_norm": 0.3863612711429596, + "learning_rate": 0.00019984950142108083, + "loss": 1.0261, + "step": 412 + }, + { + "epoch": 0.07353988603988604, + "grad_norm": 0.4177640378475189, + "learning_rate": 0.00019984873278567765, + "loss": 1.1985, + "step": 413 + }, + { + "epoch": 0.07371794871794872, + "grad_norm": 0.4645586311817169, + "learning_rate": 0.00019984796219394592, + "loss": 1.2463, + "step": 414 + }, + { + "epoch": 0.0738960113960114, + "grad_norm": 0.5051766633987427, + "learning_rate": 0.00019984718964590083, + "loss": 1.3031, + "step": 415 + }, + { + "epoch": 0.07407407407407407, + "grad_norm": 0.4200040400028229, + "learning_rate": 0.0001998464151415575, + "loss": 1.0842, + "step": 416 + }, + { + "epoch": 0.07425213675213675, + "grad_norm": 0.34211036562919617, + "learning_rate": 0.000199845638680931, + "loss": 0.9659, + "step": 417 + }, + { + "epoch": 0.07443019943019943, + "grad_norm": 0.3553323447704315, + "learning_rate": 0.00019984486026403668, + "loss": 1.0102, + "step": 418 + }, + { + "epoch": 0.0746082621082621, + "grad_norm": 0.4967300295829773, + "learning_rate": 0.00019984407989088974, + "loss": 1.3125, + "step": 419 + }, + { + "epoch": 0.07478632478632478, + "grad_norm": 0.41649797558784485, + "learning_rate": 0.00019984329756150544, + "loss": 1.3092, + "step": 420 + }, + { + "epoch": 0.07496438746438747, + "grad_norm": 0.43825802206993103, + "learning_rate": 0.00019984251327589912, + "loss": 1.3678, + "step": 421 + }, + { + "epoch": 0.07514245014245015, + "grad_norm": 0.363394170999527, + "learning_rate": 0.00019984172703408617, + "loss": 1.305, + "step": 422 + }, + { + "epoch": 0.07532051282051282, + "grad_norm": 0.411563903093338, + "learning_rate": 0.000199840938836082, + "loss": 1.4248, + "step": 423 + }, + { + "epoch": 0.0754985754985755, + "grad_norm": 0.40548190474510193, + "learning_rate": 0.000199840148681902, + "loss": 1.1081, + "step": 424 + }, + { + "epoch": 0.07567663817663818, + "grad_norm": 0.3781099021434784, + "learning_rate": 0.00019983935657156171, + "loss": 1.185, + "step": 425 + }, + { + "epoch": 0.07585470085470085, + "grad_norm": 0.46597573161125183, + "learning_rate": 0.00019983856250507662, + "loss": 1.119, + "step": 426 + }, + { + "epoch": 0.07603276353276353, + "grad_norm": 0.3988197147846222, + "learning_rate": 0.00019983776648246232, + "loss": 1.206, + "step": 427 + }, + { + "epoch": 0.07621082621082621, + "grad_norm": 0.41210901737213135, + "learning_rate": 0.00019983696850373433, + "loss": 1.1843, + "step": 428 + }, + { + "epoch": 0.0763888888888889, + "grad_norm": 0.41870948672294617, + "learning_rate": 0.00019983616856890837, + "loss": 1.2248, + "step": 429 + }, + { + "epoch": 0.07656695156695156, + "grad_norm": 0.4320056140422821, + "learning_rate": 0.00019983536667800007, + "loss": 0.9743, + "step": 430 + }, + { + "epoch": 0.07674501424501425, + "grad_norm": 0.48455503582954407, + "learning_rate": 0.00019983456283102517, + "loss": 1.0438, + "step": 431 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 0.38712427020072937, + "learning_rate": 0.00019983375702799935, + "loss": 1.2041, + "step": 432 + }, + { + "epoch": 0.0771011396011396, + "grad_norm": 0.3578857481479645, + "learning_rate": 0.0001998329492689385, + "loss": 1.1623, + "step": 433 + }, + { + "epoch": 0.07727920227920228, + "grad_norm": 0.43065932393074036, + "learning_rate": 0.00019983213955385834, + "loss": 1.3033, + "step": 434 + }, + { + "epoch": 0.07745726495726496, + "grad_norm": 0.4882095754146576, + "learning_rate": 0.00019983132788277484, + "loss": 1.1635, + "step": 435 + }, + { + "epoch": 0.07763532763532764, + "grad_norm": 0.3429015874862671, + "learning_rate": 0.00019983051425570382, + "loss": 0.7289, + "step": 436 + }, + { + "epoch": 0.07781339031339031, + "grad_norm": 0.4320310056209564, + "learning_rate": 0.00019982969867266128, + "loss": 1.3685, + "step": 437 + }, + { + "epoch": 0.07799145299145299, + "grad_norm": 0.39891982078552246, + "learning_rate": 0.00019982888113366314, + "loss": 1.0444, + "step": 438 + }, + { + "epoch": 0.07816951566951567, + "grad_norm": 0.3675695061683655, + "learning_rate": 0.00019982806163872547, + "loss": 1.0527, + "step": 439 + }, + { + "epoch": 0.07834757834757834, + "grad_norm": 0.42824694514274597, + "learning_rate": 0.0001998272401878643, + "loss": 1.166, + "step": 440 + }, + { + "epoch": 0.07852564102564102, + "grad_norm": 0.3721694350242615, + "learning_rate": 0.00019982641678109575, + "loss": 1.1328, + "step": 441 + }, + { + "epoch": 0.0787037037037037, + "grad_norm": 0.33899208903312683, + "learning_rate": 0.00019982559141843592, + "loss": 1.016, + "step": 442 + }, + { + "epoch": 0.07888176638176639, + "grad_norm": 0.4029340147972107, + "learning_rate": 0.000199824764099901, + "loss": 1.0076, + "step": 443 + }, + { + "epoch": 0.07905982905982906, + "grad_norm": 0.4169132113456726, + "learning_rate": 0.0001998239348255072, + "loss": 1.208, + "step": 444 + }, + { + "epoch": 0.07923789173789174, + "grad_norm": 0.3865824043750763, + "learning_rate": 0.00019982310359527075, + "loss": 1.067, + "step": 445 + }, + { + "epoch": 0.07941595441595442, + "grad_norm": 0.4218919277191162, + "learning_rate": 0.00019982227040920796, + "loss": 1.195, + "step": 446 + }, + { + "epoch": 0.07959401709401709, + "grad_norm": 0.40504586696624756, + "learning_rate": 0.00019982143526733512, + "loss": 1.0188, + "step": 447 + }, + { + "epoch": 0.07977207977207977, + "grad_norm": 0.38330578804016113, + "learning_rate": 0.00019982059816966863, + "loss": 1.0484, + "step": 448 + }, + { + "epoch": 0.07995014245014245, + "grad_norm": 0.43731689453125, + "learning_rate": 0.00019981975911622488, + "loss": 1.074, + "step": 449 + }, + { + "epoch": 0.08012820512820513, + "grad_norm": 0.40858447551727295, + "learning_rate": 0.00019981891810702033, + "loss": 1.0008, + "step": 450 + }, + { + "epoch": 0.0803062678062678, + "grad_norm": 0.4031754732131958, + "learning_rate": 0.00019981807514207143, + "loss": 1.2179, + "step": 451 + }, + { + "epoch": 0.08048433048433049, + "grad_norm": 0.41920867562294006, + "learning_rate": 0.00019981723022139466, + "loss": 1.1406, + "step": 452 + }, + { + "epoch": 0.08066239316239317, + "grad_norm": 0.40305474400520325, + "learning_rate": 0.00019981638334500668, + "loss": 1.098, + "step": 453 + }, + { + "epoch": 0.08084045584045584, + "grad_norm": 0.4564182460308075, + "learning_rate": 0.00019981553451292396, + "loss": 1.419, + "step": 454 + }, + { + "epoch": 0.08101851851851852, + "grad_norm": 0.3832945227622986, + "learning_rate": 0.00019981468372516322, + "loss": 1.0919, + "step": 455 + }, + { + "epoch": 0.0811965811965812, + "grad_norm": 0.43062624335289, + "learning_rate": 0.0001998138309817411, + "loss": 1.0458, + "step": 456 + }, + { + "epoch": 0.08137464387464387, + "grad_norm": 0.3871173560619354, + "learning_rate": 0.0001998129762826743, + "loss": 1.1391, + "step": 457 + }, + { + "epoch": 0.08155270655270655, + "grad_norm": 0.43423157930374146, + "learning_rate": 0.0001998121196279796, + "loss": 1.1132, + "step": 458 + }, + { + "epoch": 0.08173076923076923, + "grad_norm": 0.4341012239456177, + "learning_rate": 0.00019981126101767372, + "loss": 1.113, + "step": 459 + }, + { + "epoch": 0.08190883190883191, + "grad_norm": 0.36748576164245605, + "learning_rate": 0.00019981040045177352, + "loss": 0.8108, + "step": 460 + }, + { + "epoch": 0.08208689458689458, + "grad_norm": 0.43133220076560974, + "learning_rate": 0.00019980953793029586, + "loss": 1.1861, + "step": 461 + }, + { + "epoch": 0.08226495726495726, + "grad_norm": 0.37204909324645996, + "learning_rate": 0.00019980867345325767, + "loss": 0.9222, + "step": 462 + }, + { + "epoch": 0.08244301994301995, + "grad_norm": 0.43370047211647034, + "learning_rate": 0.00019980780702067582, + "loss": 1.2984, + "step": 463 + }, + { + "epoch": 0.08262108262108261, + "grad_norm": 0.4991510808467865, + "learning_rate": 0.00019980693863256736, + "loss": 1.2222, + "step": 464 + }, + { + "epoch": 0.0827991452991453, + "grad_norm": 0.44318175315856934, + "learning_rate": 0.00019980606828894927, + "loss": 1.2262, + "step": 465 + }, + { + "epoch": 0.08297720797720798, + "grad_norm": 0.380231648683548, + "learning_rate": 0.0001998051959898386, + "loss": 1.0274, + "step": 466 + }, + { + "epoch": 0.08315527065527066, + "grad_norm": 0.39519667625427246, + "learning_rate": 0.0001998043217352524, + "loss": 1.2499, + "step": 467 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.457499235868454, + "learning_rate": 0.0001998034455252079, + "loss": 1.0751, + "step": 468 + }, + { + "epoch": 0.08351139601139601, + "grad_norm": 0.368522584438324, + "learning_rate": 0.00019980256735972215, + "loss": 1.0776, + "step": 469 + }, + { + "epoch": 0.08368945868945869, + "grad_norm": 0.3768427073955536, + "learning_rate": 0.00019980168723881243, + "loss": 1.2198, + "step": 470 + }, + { + "epoch": 0.08386752136752136, + "grad_norm": 0.37045565247535706, + "learning_rate": 0.000199800805162496, + "loss": 1.1816, + "step": 471 + }, + { + "epoch": 0.08404558404558404, + "grad_norm": 0.4219281077384949, + "learning_rate": 0.0001997999211307901, + "loss": 1.0515, + "step": 472 + }, + { + "epoch": 0.08422364672364673, + "grad_norm": 0.3815271258354187, + "learning_rate": 0.00019979903514371207, + "loss": 1.1709, + "step": 473 + }, + { + "epoch": 0.08440170940170941, + "grad_norm": 0.4566493630409241, + "learning_rate": 0.00019979814720127924, + "loss": 1.3063, + "step": 474 + }, + { + "epoch": 0.08457977207977208, + "grad_norm": 0.4043879806995392, + "learning_rate": 0.000199797257303509, + "loss": 1.0549, + "step": 475 + }, + { + "epoch": 0.08475783475783476, + "grad_norm": 0.3897830545902252, + "learning_rate": 0.00019979636545041886, + "loss": 1.1483, + "step": 476 + }, + { + "epoch": 0.08493589743589744, + "grad_norm": 0.36097025871276855, + "learning_rate": 0.00019979547164202622, + "loss": 1.1196, + "step": 477 + }, + { + "epoch": 0.08511396011396011, + "grad_norm": 0.3766986131668091, + "learning_rate": 0.00019979457587834863, + "loss": 1.0131, + "step": 478 + }, + { + "epoch": 0.08529202279202279, + "grad_norm": 0.39460286498069763, + "learning_rate": 0.00019979367815940364, + "loss": 1.1729, + "step": 479 + }, + { + "epoch": 0.08547008547008547, + "grad_norm": 0.4137469232082367, + "learning_rate": 0.00019979277848520885, + "loss": 1.2569, + "step": 480 + }, + { + "epoch": 0.08564814814814815, + "grad_norm": 0.464688777923584, + "learning_rate": 0.00019979187685578183, + "loss": 1.2064, + "step": 481 + }, + { + "epoch": 0.08582621082621082, + "grad_norm": 0.4245518147945404, + "learning_rate": 0.0001997909732711403, + "loss": 0.9812, + "step": 482 + }, + { + "epoch": 0.0860042735042735, + "grad_norm": 0.43368837237358093, + "learning_rate": 0.00019979006773130197, + "loss": 1.2822, + "step": 483 + }, + { + "epoch": 0.08618233618233619, + "grad_norm": 0.4232824444770813, + "learning_rate": 0.00019978916023628452, + "loss": 1.1446, + "step": 484 + }, + { + "epoch": 0.08636039886039885, + "grad_norm": 0.4183506369590759, + "learning_rate": 0.00019978825078610578, + "loss": 1.2605, + "step": 485 + }, + { + "epoch": 0.08653846153846154, + "grad_norm": 0.4391268491744995, + "learning_rate": 0.00019978733938078356, + "loss": 1.2165, + "step": 486 + }, + { + "epoch": 0.08671652421652422, + "grad_norm": 0.4139612317085266, + "learning_rate": 0.0001997864260203357, + "loss": 0.9389, + "step": 487 + }, + { + "epoch": 0.0868945868945869, + "grad_norm": 0.4058656096458435, + "learning_rate": 0.00019978551070478013, + "loss": 1.0652, + "step": 488 + }, + { + "epoch": 0.08707264957264957, + "grad_norm": 0.42333099246025085, + "learning_rate": 0.00019978459343413473, + "loss": 1.119, + "step": 489 + }, + { + "epoch": 0.08725071225071225, + "grad_norm": 0.4573031961917877, + "learning_rate": 0.00019978367420841754, + "loss": 1.1546, + "step": 490 + }, + { + "epoch": 0.08742877492877493, + "grad_norm": 0.4161617159843445, + "learning_rate": 0.00019978275302764655, + "loss": 1.0836, + "step": 491 + }, + { + "epoch": 0.0876068376068376, + "grad_norm": 0.422145277261734, + "learning_rate": 0.00019978182989183977, + "loss": 1.1908, + "step": 492 + }, + { + "epoch": 0.08778490028490028, + "grad_norm": 0.4588126838207245, + "learning_rate": 0.00019978090480101532, + "loss": 1.1758, + "step": 493 + }, + { + "epoch": 0.08796296296296297, + "grad_norm": 0.4425722062587738, + "learning_rate": 0.00019977997775519132, + "loss": 1.088, + "step": 494 + }, + { + "epoch": 0.08814102564102565, + "grad_norm": 0.37860307097435, + "learning_rate": 0.00019977904875438594, + "loss": 1.1532, + "step": 495 + }, + { + "epoch": 0.08831908831908832, + "grad_norm": 0.40435823798179626, + "learning_rate": 0.00019977811779861733, + "loss": 1.1271, + "step": 496 + }, + { + "epoch": 0.088497150997151, + "grad_norm": 0.42578884959220886, + "learning_rate": 0.0001997771848879038, + "loss": 0.9889, + "step": 497 + }, + { + "epoch": 0.08867521367521368, + "grad_norm": 0.3439478874206543, + "learning_rate": 0.00019977625002226361, + "loss": 1.1273, + "step": 498 + }, + { + "epoch": 0.08885327635327635, + "grad_norm": 0.362341970205307, + "learning_rate": 0.00019977531320171504, + "loss": 1.0214, + "step": 499 + }, + { + "epoch": 0.08903133903133903, + "grad_norm": 0.4305768609046936, + "learning_rate": 0.0001997743744262765, + "loss": 1.2648, + "step": 500 + }, + { + "epoch": 0.08920940170940171, + "grad_norm": 0.35900023579597473, + "learning_rate": 0.00019977343369596636, + "loss": 1.0274, + "step": 501 + }, + { + "epoch": 0.0893874643874644, + "grad_norm": 0.4950818717479706, + "learning_rate": 0.00019977249101080306, + "loss": 1.1483, + "step": 502 + }, + { + "epoch": 0.08956552706552706, + "grad_norm": 0.3800346553325653, + "learning_rate": 0.00019977154637080503, + "loss": 1.0636, + "step": 503 + }, + { + "epoch": 0.08974358974358974, + "grad_norm": 0.46202352643013, + "learning_rate": 0.0001997705997759908, + "loss": 1.1544, + "step": 504 + }, + { + "epoch": 0.08992165242165243, + "grad_norm": 0.36818403005599976, + "learning_rate": 0.00019976965122637895, + "loss": 0.9824, + "step": 505 + }, + { + "epoch": 0.0900997150997151, + "grad_norm": 0.40248095989227295, + "learning_rate": 0.00019976870072198805, + "loss": 1.1002, + "step": 506 + }, + { + "epoch": 0.09027777777777778, + "grad_norm": 0.3841850459575653, + "learning_rate": 0.00019976774826283667, + "loss": 1.2433, + "step": 507 + }, + { + "epoch": 0.09045584045584046, + "grad_norm": 0.46892330050468445, + "learning_rate": 0.0001997667938489435, + "loss": 1.3194, + "step": 508 + }, + { + "epoch": 0.09063390313390314, + "grad_norm": 0.39059561491012573, + "learning_rate": 0.0001997658374803273, + "loss": 1.1778, + "step": 509 + }, + { + "epoch": 0.09081196581196581, + "grad_norm": 0.3793235421180725, + "learning_rate": 0.00019976487915700672, + "loss": 1.0659, + "step": 510 + }, + { + "epoch": 0.09099002849002849, + "grad_norm": 0.39067742228507996, + "learning_rate": 0.00019976391887900058, + "loss": 1.107, + "step": 511 + }, + { + "epoch": 0.09116809116809117, + "grad_norm": 0.40121713280677795, + "learning_rate": 0.00019976295664632772, + "loss": 1.102, + "step": 512 + }, + { + "epoch": 0.09134615384615384, + "grad_norm": 0.49830010533332825, + "learning_rate": 0.00019976199245900697, + "loss": 1.1701, + "step": 513 + }, + { + "epoch": 0.09152421652421652, + "grad_norm": 0.4536968171596527, + "learning_rate": 0.0001997610263170572, + "loss": 1.1067, + "step": 514 + }, + { + "epoch": 0.0917022792022792, + "grad_norm": 0.3832971453666687, + "learning_rate": 0.00019976005822049735, + "loss": 1.0991, + "step": 515 + }, + { + "epoch": 0.09188034188034189, + "grad_norm": 0.4093509614467621, + "learning_rate": 0.0001997590881693464, + "loss": 1.0565, + "step": 516 + }, + { + "epoch": 0.09205840455840456, + "grad_norm": 0.46073687076568604, + "learning_rate": 0.0001997581161636233, + "loss": 1.0057, + "step": 517 + }, + { + "epoch": 0.09223646723646724, + "grad_norm": 0.5001922845840454, + "learning_rate": 0.0001997571422033472, + "loss": 1.2639, + "step": 518 + }, + { + "epoch": 0.09241452991452992, + "grad_norm": 0.4620618224143982, + "learning_rate": 0.00019975616628853713, + "loss": 1.0966, + "step": 519 + }, + { + "epoch": 0.09259259259259259, + "grad_norm": 0.3788183927536011, + "learning_rate": 0.0001997551884192122, + "loss": 0.9783, + "step": 520 + }, + { + "epoch": 0.09277065527065527, + "grad_norm": 0.45589539408683777, + "learning_rate": 0.00019975420859539154, + "loss": 1.2194, + "step": 521 + }, + { + "epoch": 0.09294871794871795, + "grad_norm": 0.40747523307800293, + "learning_rate": 0.00019975322681709443, + "loss": 1.0349, + "step": 522 + }, + { + "epoch": 0.09312678062678063, + "grad_norm": 0.5045142769813538, + "learning_rate": 0.00019975224308434002, + "loss": 1.1373, + "step": 523 + }, + { + "epoch": 0.0933048433048433, + "grad_norm": 0.40352702140808105, + "learning_rate": 0.00019975125739714767, + "loss": 1.1236, + "step": 524 + }, + { + "epoch": 0.09348290598290598, + "grad_norm": 0.4301735758781433, + "learning_rate": 0.0001997502697555366, + "loss": 1.2932, + "step": 525 + }, + { + "epoch": 0.09366096866096867, + "grad_norm": 0.36800238490104675, + "learning_rate": 0.00019974928015952624, + "loss": 1.0734, + "step": 526 + }, + { + "epoch": 0.09383903133903133, + "grad_norm": 0.4027230143547058, + "learning_rate": 0.00019974828860913594, + "loss": 1.2776, + "step": 527 + }, + { + "epoch": 0.09401709401709402, + "grad_norm": 0.42497140169143677, + "learning_rate": 0.0001997472951043851, + "loss": 1.248, + "step": 528 + }, + { + "epoch": 0.0941951566951567, + "grad_norm": 0.3888593018054962, + "learning_rate": 0.00019974629964529325, + "loss": 1.0231, + "step": 529 + }, + { + "epoch": 0.09437321937321937, + "grad_norm": 0.3761361241340637, + "learning_rate": 0.00019974530223187986, + "loss": 1.0216, + "step": 530 + }, + { + "epoch": 0.09455128205128205, + "grad_norm": 0.42192980647087097, + "learning_rate": 0.00019974430286416448, + "loss": 1.0731, + "step": 531 + }, + { + "epoch": 0.09472934472934473, + "grad_norm": 0.44244512915611267, + "learning_rate": 0.00019974330154216667, + "loss": 1.2793, + "step": 532 + }, + { + "epoch": 0.09490740740740741, + "grad_norm": 0.378252774477005, + "learning_rate": 0.0001997422982659061, + "loss": 1.0462, + "step": 533 + }, + { + "epoch": 0.09508547008547008, + "grad_norm": 0.45589110255241394, + "learning_rate": 0.00019974129303540236, + "loss": 1.1884, + "step": 534 + }, + { + "epoch": 0.09526353276353276, + "grad_norm": 0.33930808305740356, + "learning_rate": 0.0001997402858506752, + "loss": 0.8381, + "step": 535 + }, + { + "epoch": 0.09544159544159544, + "grad_norm": 0.45408427715301514, + "learning_rate": 0.0001997392767117443, + "loss": 1.2379, + "step": 536 + }, + { + "epoch": 0.09561965811965811, + "grad_norm": 0.44125741720199585, + "learning_rate": 0.0001997382656186295, + "loss": 1.1941, + "step": 537 + }, + { + "epoch": 0.0957977207977208, + "grad_norm": 0.4075697660446167, + "learning_rate": 0.00019973725257135054, + "loss": 1.0142, + "step": 538 + }, + { + "epoch": 0.09597578347578348, + "grad_norm": 0.4258415102958679, + "learning_rate": 0.00019973623756992733, + "loss": 1.0447, + "step": 539 + }, + { + "epoch": 0.09615384615384616, + "grad_norm": 0.2738485038280487, + "learning_rate": 0.0001997352206143797, + "loss": 0.5521, + "step": 540 + }, + { + "epoch": 0.09633190883190883, + "grad_norm": 0.38815587759017944, + "learning_rate": 0.00019973420170472762, + "loss": 1.1052, + "step": 541 + }, + { + "epoch": 0.09650997150997151, + "grad_norm": 0.3909834027290344, + "learning_rate": 0.00019973318084099106, + "loss": 1.0494, + "step": 542 + }, + { + "epoch": 0.09668803418803419, + "grad_norm": 0.4517597258090973, + "learning_rate": 0.00019973215802318996, + "loss": 1.0611, + "step": 543 + }, + { + "epoch": 0.09686609686609686, + "grad_norm": 0.48659002780914307, + "learning_rate": 0.00019973113325134442, + "loss": 0.9967, + "step": 544 + }, + { + "epoch": 0.09704415954415954, + "grad_norm": 0.4039791524410248, + "learning_rate": 0.0001997301065254745, + "loss": 1.251, + "step": 545 + }, + { + "epoch": 0.09722222222222222, + "grad_norm": 0.3985383212566376, + "learning_rate": 0.0001997290778456003, + "loss": 1.2263, + "step": 546 + }, + { + "epoch": 0.0974002849002849, + "grad_norm": 0.4540637731552124, + "learning_rate": 0.00019972804721174199, + "loss": 1.2084, + "step": 547 + }, + { + "epoch": 0.09757834757834757, + "grad_norm": 0.36867982149124146, + "learning_rate": 0.00019972701462391977, + "loss": 0.9704, + "step": 548 + }, + { + "epoch": 0.09775641025641026, + "grad_norm": 0.40199780464172363, + "learning_rate": 0.00019972598008215385, + "loss": 1.1121, + "step": 549 + }, + { + "epoch": 0.09793447293447294, + "grad_norm": 0.42728984355926514, + "learning_rate": 0.00019972494358646455, + "loss": 1.1606, + "step": 550 + }, + { + "epoch": 0.0981125356125356, + "grad_norm": 0.4212374687194824, + "learning_rate": 0.0001997239051368721, + "loss": 1.3093, + "step": 551 + }, + { + "epoch": 0.09829059829059829, + "grad_norm": 0.3972226083278656, + "learning_rate": 0.0001997228647333969, + "loss": 1.1218, + "step": 552 + }, + { + "epoch": 0.09846866096866097, + "grad_norm": 0.43649932742118835, + "learning_rate": 0.00019972182237605935, + "loss": 1.2532, + "step": 553 + }, + { + "epoch": 0.09864672364672365, + "grad_norm": 0.3812280595302582, + "learning_rate": 0.0001997207780648798, + "loss": 1.0409, + "step": 554 + }, + { + "epoch": 0.09882478632478632, + "grad_norm": 0.41684821248054504, + "learning_rate": 0.00019971973179987878, + "loss": 0.9569, + "step": 555 + }, + { + "epoch": 0.099002849002849, + "grad_norm": 0.38081470131874084, + "learning_rate": 0.00019971868358107674, + "loss": 1.1615, + "step": 556 + }, + { + "epoch": 0.09918091168091168, + "grad_norm": 0.3702073097229004, + "learning_rate": 0.0001997176334084943, + "loss": 1.3907, + "step": 557 + }, + { + "epoch": 0.09935897435897435, + "grad_norm": 0.3625728189945221, + "learning_rate": 0.00019971658128215193, + "loss": 1.1897, + "step": 558 + }, + { + "epoch": 0.09953703703703703, + "grad_norm": 0.3815405070781708, + "learning_rate": 0.0001997155272020703, + "loss": 1.1473, + "step": 559 + }, + { + "epoch": 0.09971509971509972, + "grad_norm": 0.48664286732673645, + "learning_rate": 0.00019971447116827004, + "loss": 1.2462, + "step": 560 + }, + { + "epoch": 0.0998931623931624, + "grad_norm": 0.3708696663379669, + "learning_rate": 0.0001997134131807719, + "loss": 1.0979, + "step": 561 + }, + { + "epoch": 0.10007122507122507, + "grad_norm": 0.44511324167251587, + "learning_rate": 0.00019971235323959654, + "loss": 1.2313, + "step": 562 + }, + { + "epoch": 0.10024928774928775, + "grad_norm": 0.3687448799610138, + "learning_rate": 0.00019971129134476473, + "loss": 1.1526, + "step": 563 + }, + { + "epoch": 0.10042735042735043, + "grad_norm": 0.4506866931915283, + "learning_rate": 0.00019971022749629735, + "loss": 1.0003, + "step": 564 + }, + { + "epoch": 0.1006054131054131, + "grad_norm": 0.41910406947135925, + "learning_rate": 0.00019970916169421515, + "loss": 1.013, + "step": 565 + }, + { + "epoch": 0.10078347578347578, + "grad_norm": 0.39728936553001404, + "learning_rate": 0.0001997080939385391, + "loss": 1.0501, + "step": 566 + }, + { + "epoch": 0.10096153846153846, + "grad_norm": 0.41415902972221375, + "learning_rate": 0.00019970702422929005, + "loss": 1.0791, + "step": 567 + }, + { + "epoch": 0.10113960113960115, + "grad_norm": 0.45630788803100586, + "learning_rate": 0.00019970595256648896, + "loss": 1.2884, + "step": 568 + }, + { + "epoch": 0.10131766381766381, + "grad_norm": 0.4371698796749115, + "learning_rate": 0.00019970487895015686, + "loss": 1.0684, + "step": 569 + }, + { + "epoch": 0.1014957264957265, + "grad_norm": 0.4350591003894806, + "learning_rate": 0.00019970380338031477, + "loss": 1.2415, + "step": 570 + }, + { + "epoch": 0.10167378917378918, + "grad_norm": 0.4232708215713501, + "learning_rate": 0.00019970272585698382, + "loss": 1.2656, + "step": 571 + }, + { + "epoch": 0.10185185185185185, + "grad_norm": 0.3917689919471741, + "learning_rate": 0.00019970164638018502, + "loss": 1.0178, + "step": 572 + }, + { + "epoch": 0.10202991452991453, + "grad_norm": 0.4262804388999939, + "learning_rate": 0.0001997005649499396, + "loss": 1.1805, + "step": 573 + }, + { + "epoch": 0.10220797720797721, + "grad_norm": 0.5217884182929993, + "learning_rate": 0.0001996994815662687, + "loss": 1.2392, + "step": 574 + }, + { + "epoch": 0.10238603988603989, + "grad_norm": 0.4273875057697296, + "learning_rate": 0.00019969839622919358, + "loss": 1.0844, + "step": 575 + }, + { + "epoch": 0.10256410256410256, + "grad_norm": 0.41588085889816284, + "learning_rate": 0.00019969730893873547, + "loss": 1.2437, + "step": 576 + }, + { + "epoch": 0.10274216524216524, + "grad_norm": 0.41617709398269653, + "learning_rate": 0.0001996962196949157, + "loss": 0.9519, + "step": 577 + }, + { + "epoch": 0.10292022792022792, + "grad_norm": 0.4832979142665863, + "learning_rate": 0.00019969512849775565, + "loss": 1.1889, + "step": 578 + }, + { + "epoch": 0.10309829059829059, + "grad_norm": 0.3936060965061188, + "learning_rate": 0.0001996940353472766, + "loss": 0.9888, + "step": 579 + }, + { + "epoch": 0.10327635327635327, + "grad_norm": 0.4147680997848511, + "learning_rate": 0.00019969294024350004, + "loss": 1.0733, + "step": 580 + }, + { + "epoch": 0.10345441595441596, + "grad_norm": 0.37791356444358826, + "learning_rate": 0.00019969184318644742, + "loss": 1.212, + "step": 581 + }, + { + "epoch": 0.10363247863247864, + "grad_norm": 0.44297221302986145, + "learning_rate": 0.00019969074417614023, + "loss": 1.0535, + "step": 582 + }, + { + "epoch": 0.10381054131054131, + "grad_norm": 0.4032835066318512, + "learning_rate": 0.0001996896432126, + "loss": 1.1869, + "step": 583 + }, + { + "epoch": 0.10398860398860399, + "grad_norm": 0.49271953105926514, + "learning_rate": 0.00019968854029584827, + "loss": 1.1661, + "step": 584 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 0.362699031829834, + "learning_rate": 0.0001996874354259067, + "loss": 0.868, + "step": 585 + }, + { + "epoch": 0.10434472934472934, + "grad_norm": 0.401795357465744, + "learning_rate": 0.0001996863286027969, + "loss": 1.1045, + "step": 586 + }, + { + "epoch": 0.10452279202279202, + "grad_norm": 0.45380479097366333, + "learning_rate": 0.00019968521982654058, + "loss": 0.8503, + "step": 587 + }, + { + "epoch": 0.1047008547008547, + "grad_norm": 0.49759066104888916, + "learning_rate": 0.00019968410909715947, + "loss": 1.4073, + "step": 588 + }, + { + "epoch": 0.10487891737891739, + "grad_norm": 0.4421198070049286, + "learning_rate": 0.0001996829964146753, + "loss": 1.1512, + "step": 589 + }, + { + "epoch": 0.10505698005698005, + "grad_norm": 0.46675658226013184, + "learning_rate": 0.00019968188177910988, + "loss": 1.0132, + "step": 590 + }, + { + "epoch": 0.10523504273504274, + "grad_norm": 0.5710657238960266, + "learning_rate": 0.00019968076519048507, + "loss": 1.267, + "step": 591 + }, + { + "epoch": 0.10541310541310542, + "grad_norm": 0.4655563235282898, + "learning_rate": 0.00019967964664882276, + "loss": 1.1204, + "step": 592 + }, + { + "epoch": 0.10559116809116809, + "grad_norm": 0.3895256519317627, + "learning_rate": 0.00019967852615414478, + "loss": 1.0814, + "step": 593 + }, + { + "epoch": 0.10576923076923077, + "grad_norm": 0.424216091632843, + "learning_rate": 0.00019967740370647322, + "loss": 1.1663, + "step": 594 + }, + { + "epoch": 0.10594729344729345, + "grad_norm": 0.3978985846042633, + "learning_rate": 0.00019967627930582996, + "loss": 0.909, + "step": 595 + }, + { + "epoch": 0.10612535612535613, + "grad_norm": 0.47064995765686035, + "learning_rate": 0.00019967515295223705, + "loss": 1.2351, + "step": 596 + }, + { + "epoch": 0.1063034188034188, + "grad_norm": 0.42449644207954407, + "learning_rate": 0.0001996740246457166, + "loss": 0.9739, + "step": 597 + }, + { + "epoch": 0.10648148148148148, + "grad_norm": 0.39033401012420654, + "learning_rate": 0.00019967289438629066, + "loss": 1.0933, + "step": 598 + }, + { + "epoch": 0.10665954415954416, + "grad_norm": 0.4398612678050995, + "learning_rate": 0.00019967176217398143, + "loss": 1.2479, + "step": 599 + }, + { + "epoch": 0.10683760683760683, + "grad_norm": 0.3946632742881775, + "learning_rate": 0.00019967062800881107, + "loss": 1.0417, + "step": 600 + }, + { + "epoch": 0.10701566951566951, + "grad_norm": 0.5083445906639099, + "learning_rate": 0.0001996694918908018, + "loss": 1.1109, + "step": 601 + }, + { + "epoch": 0.1071937321937322, + "grad_norm": 0.477724552154541, + "learning_rate": 0.00019966835381997585, + "loss": 1.2891, + "step": 602 + }, + { + "epoch": 0.10737179487179487, + "grad_norm": 0.4110167920589447, + "learning_rate": 0.0001996672137963556, + "loss": 1.0555, + "step": 603 + }, + { + "epoch": 0.10754985754985755, + "grad_norm": 0.44078320264816284, + "learning_rate": 0.00019966607181996334, + "loss": 0.9188, + "step": 604 + }, + { + "epoch": 0.10772792022792023, + "grad_norm": 0.41251105070114136, + "learning_rate": 0.00019966492789082142, + "loss": 1.2592, + "step": 605 + }, + { + "epoch": 0.10790598290598291, + "grad_norm": 0.37701505422592163, + "learning_rate": 0.00019966378200895227, + "loss": 1.0233, + "step": 606 + }, + { + "epoch": 0.10808404558404558, + "grad_norm": 0.44624966382980347, + "learning_rate": 0.00019966263417437835, + "loss": 1.2273, + "step": 607 + }, + { + "epoch": 0.10826210826210826, + "grad_norm": 0.3618549108505249, + "learning_rate": 0.00019966148438712214, + "loss": 0.9101, + "step": 608 + }, + { + "epoch": 0.10844017094017094, + "grad_norm": 0.384574294090271, + "learning_rate": 0.00019966033264720616, + "loss": 1.1769, + "step": 609 + }, + { + "epoch": 0.10861823361823361, + "grad_norm": 0.50872403383255, + "learning_rate": 0.000199659178954653, + "loss": 1.1213, + "step": 610 + }, + { + "epoch": 0.1087962962962963, + "grad_norm": 0.39736685156822205, + "learning_rate": 0.00019965802330948527, + "loss": 1.275, + "step": 611 + }, + { + "epoch": 0.10897435897435898, + "grad_norm": 0.484660267829895, + "learning_rate": 0.00019965686571172557, + "loss": 1.1671, + "step": 612 + }, + { + "epoch": 0.10915242165242166, + "grad_norm": 0.41420218348503113, + "learning_rate": 0.0001996557061613966, + "loss": 0.9541, + "step": 613 + }, + { + "epoch": 0.10933048433048433, + "grad_norm": 0.4057196080684662, + "learning_rate": 0.00019965454465852112, + "loss": 1.0145, + "step": 614 + }, + { + "epoch": 0.10950854700854701, + "grad_norm": 0.4559510052204132, + "learning_rate": 0.00019965338120312182, + "loss": 1.0889, + "step": 615 + }, + { + "epoch": 0.10968660968660969, + "grad_norm": 0.40960055589675903, + "learning_rate": 0.00019965221579522154, + "loss": 1.1447, + "step": 616 + }, + { + "epoch": 0.10986467236467236, + "grad_norm": 0.4701732099056244, + "learning_rate": 0.0001996510484348431, + "loss": 1.2871, + "step": 617 + }, + { + "epoch": 0.11004273504273504, + "grad_norm": 0.38420796394348145, + "learning_rate": 0.0001996498791220094, + "loss": 1.058, + "step": 618 + }, + { + "epoch": 0.11022079772079772, + "grad_norm": 0.4014730453491211, + "learning_rate": 0.00019964870785674327, + "loss": 1.023, + "step": 619 + }, + { + "epoch": 0.1103988603988604, + "grad_norm": 0.38846179842948914, + "learning_rate": 0.00019964753463906773, + "loss": 0.9834, + "step": 620 + }, + { + "epoch": 0.11057692307692307, + "grad_norm": 0.5120236277580261, + "learning_rate": 0.00019964635946900577, + "loss": 1.2347, + "step": 621 + }, + { + "epoch": 0.11075498575498575, + "grad_norm": 0.40483301877975464, + "learning_rate": 0.00019964518234658038, + "loss": 1.131, + "step": 622 + }, + { + "epoch": 0.11093304843304844, + "grad_norm": 0.445782870054245, + "learning_rate": 0.00019964400327181464, + "loss": 0.9349, + "step": 623 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 0.490460604429245, + "learning_rate": 0.00019964282224473165, + "loss": 1.0257, + "step": 624 + }, + { + "epoch": 0.11128917378917379, + "grad_norm": 0.37585243582725525, + "learning_rate": 0.00019964163926535454, + "loss": 0.9724, + "step": 625 + }, + { + "epoch": 0.11146723646723647, + "grad_norm": 0.4160473346710205, + "learning_rate": 0.00019964045433370651, + "loss": 0.874, + "step": 626 + }, + { + "epoch": 0.11164529914529915, + "grad_norm": 0.442425012588501, + "learning_rate": 0.00019963926744981074, + "loss": 1.064, + "step": 627 + }, + { + "epoch": 0.11182336182336182, + "grad_norm": 0.4451471269130707, + "learning_rate": 0.00019963807861369054, + "loss": 1.2343, + "step": 628 + }, + { + "epoch": 0.1120014245014245, + "grad_norm": 0.5018183588981628, + "learning_rate": 0.00019963688782536913, + "loss": 1.1226, + "step": 629 + }, + { + "epoch": 0.11217948717948718, + "grad_norm": 0.43723925948143005, + "learning_rate": 0.0001996356950848699, + "loss": 1.0178, + "step": 630 + }, + { + "epoch": 0.11235754985754985, + "grad_norm": 0.4794611930847168, + "learning_rate": 0.0001996345003922162, + "loss": 0.9695, + "step": 631 + }, + { + "epoch": 0.11253561253561253, + "grad_norm": 0.5021790266036987, + "learning_rate": 0.00019963330374743143, + "loss": 1.1748, + "step": 632 + }, + { + "epoch": 0.11271367521367522, + "grad_norm": 0.47228625416755676, + "learning_rate": 0.00019963210515053906, + "loss": 1.2138, + "step": 633 + }, + { + "epoch": 0.1128917378917379, + "grad_norm": 0.4261155128479004, + "learning_rate": 0.00019963090460156256, + "loss": 0.9428, + "step": 634 + }, + { + "epoch": 0.11306980056980057, + "grad_norm": 0.3279525339603424, + "learning_rate": 0.00019962970210052542, + "loss": 0.7803, + "step": 635 + }, + { + "epoch": 0.11324786324786325, + "grad_norm": 0.5106086730957031, + "learning_rate": 0.00019962849764745125, + "loss": 1.113, + "step": 636 + }, + { + "epoch": 0.11342592592592593, + "grad_norm": 0.38272222876548767, + "learning_rate": 0.00019962729124236363, + "loss": 0.896, + "step": 637 + }, + { + "epoch": 0.1136039886039886, + "grad_norm": 0.39532098174095154, + "learning_rate": 0.0001996260828852862, + "loss": 0.9308, + "step": 638 + }, + { + "epoch": 0.11378205128205128, + "grad_norm": 0.44947221875190735, + "learning_rate": 0.00019962487257624262, + "loss": 1.207, + "step": 639 + }, + { + "epoch": 0.11396011396011396, + "grad_norm": 0.40684598684310913, + "learning_rate": 0.00019962366031525664, + "loss": 1.11, + "step": 640 + }, + { + "epoch": 0.11413817663817664, + "grad_norm": 0.4296625852584839, + "learning_rate": 0.00019962244610235194, + "loss": 1.2784, + "step": 641 + }, + { + "epoch": 0.11431623931623931, + "grad_norm": 0.4560794532299042, + "learning_rate": 0.0001996212299375524, + "loss": 1.1191, + "step": 642 + }, + { + "epoch": 0.114494301994302, + "grad_norm": 0.40246087312698364, + "learning_rate": 0.00019962001182088177, + "loss": 1.1401, + "step": 643 + }, + { + "epoch": 0.11467236467236468, + "grad_norm": 0.3938910663127899, + "learning_rate": 0.000199618791752364, + "loss": 1.0959, + "step": 644 + }, + { + "epoch": 0.11485042735042734, + "grad_norm": 0.4123380184173584, + "learning_rate": 0.00019961756973202287, + "loss": 1.2824, + "step": 645 + }, + { + "epoch": 0.11502849002849003, + "grad_norm": 0.41085442900657654, + "learning_rate": 0.00019961634575988243, + "loss": 1.1137, + "step": 646 + }, + { + "epoch": 0.11520655270655271, + "grad_norm": 0.38276201486587524, + "learning_rate": 0.0001996151198359667, + "loss": 1.0747, + "step": 647 + }, + { + "epoch": 0.11538461538461539, + "grad_norm": 0.49269407987594604, + "learning_rate": 0.00019961389196029953, + "loss": 1.1731, + "step": 648 + }, + { + "epoch": 0.11556267806267806, + "grad_norm": 0.5152469277381897, + "learning_rate": 0.00019961266213290512, + "loss": 1.3574, + "step": 649 + }, + { + "epoch": 0.11574074074074074, + "grad_norm": 0.4835714101791382, + "learning_rate": 0.0001996114303538075, + "loss": 1.2859, + "step": 650 + }, + { + "epoch": 0.11591880341880342, + "grad_norm": 0.4284524917602539, + "learning_rate": 0.00019961019662303087, + "loss": 1.1103, + "step": 651 + }, + { + "epoch": 0.11609686609686609, + "grad_norm": 0.3933276832103729, + "learning_rate": 0.00019960896094059933, + "loss": 1.2647, + "step": 652 + }, + { + "epoch": 0.11627492877492877, + "grad_norm": 0.33749741315841675, + "learning_rate": 0.00019960772330653712, + "loss": 0.819, + "step": 653 + }, + { + "epoch": 0.11645299145299146, + "grad_norm": 0.48122069239616394, + "learning_rate": 0.00019960648372086852, + "loss": 1.2781, + "step": 654 + }, + { + "epoch": 0.11663105413105414, + "grad_norm": 0.4681607186794281, + "learning_rate": 0.00019960524218361775, + "loss": 0.9723, + "step": 655 + }, + { + "epoch": 0.1168091168091168, + "grad_norm": 0.3974960148334503, + "learning_rate": 0.0001996039986948092, + "loss": 1.0302, + "step": 656 + }, + { + "epoch": 0.11698717948717949, + "grad_norm": 0.43180662393569946, + "learning_rate": 0.0001996027532544672, + "loss": 1.3265, + "step": 657 + }, + { + "epoch": 0.11716524216524217, + "grad_norm": 0.4481917917728424, + "learning_rate": 0.00019960150586261613, + "loss": 1.136, + "step": 658 + }, + { + "epoch": 0.11734330484330484, + "grad_norm": 0.43428945541381836, + "learning_rate": 0.00019960025651928045, + "loss": 1.2412, + "step": 659 + }, + { + "epoch": 0.11752136752136752, + "grad_norm": 0.36211395263671875, + "learning_rate": 0.00019959900522448467, + "loss": 0.9563, + "step": 660 + }, + { + "epoch": 0.1176994301994302, + "grad_norm": 0.43585848808288574, + "learning_rate": 0.0001995977519782533, + "loss": 1.1677, + "step": 661 + }, + { + "epoch": 0.11787749287749288, + "grad_norm": 0.4232597351074219, + "learning_rate": 0.00019959649678061086, + "loss": 1.1187, + "step": 662 + }, + { + "epoch": 0.11805555555555555, + "grad_norm": 0.3304753303527832, + "learning_rate": 0.00019959523963158194, + "loss": 0.8473, + "step": 663 + }, + { + "epoch": 0.11823361823361823, + "grad_norm": 0.37600061297416687, + "learning_rate": 0.0001995939805311912, + "loss": 1.1227, + "step": 664 + }, + { + "epoch": 0.11841168091168092, + "grad_norm": 0.33417847752571106, + "learning_rate": 0.0001995927194794633, + "loss": 1.0315, + "step": 665 + }, + { + "epoch": 0.11858974358974358, + "grad_norm": 0.46799129247665405, + "learning_rate": 0.00019959145647642298, + "loss": 1.135, + "step": 666 + }, + { + "epoch": 0.11876780626780627, + "grad_norm": 0.4141576886177063, + "learning_rate": 0.0001995901915220949, + "loss": 1.0956, + "step": 667 + }, + { + "epoch": 0.11894586894586895, + "grad_norm": 0.3824596405029297, + "learning_rate": 0.0001995889246165039, + "loss": 1.1782, + "step": 668 + }, + { + "epoch": 0.11912393162393162, + "grad_norm": 0.4087786376476288, + "learning_rate": 0.00019958765575967484, + "loss": 0.9704, + "step": 669 + }, + { + "epoch": 0.1193019943019943, + "grad_norm": 0.5161317586898804, + "learning_rate": 0.00019958638495163252, + "loss": 1.2207, + "step": 670 + }, + { + "epoch": 0.11948005698005698, + "grad_norm": 0.4782274067401886, + "learning_rate": 0.0001995851121924019, + "loss": 1.1257, + "step": 671 + }, + { + "epoch": 0.11965811965811966, + "grad_norm": 0.40617331862449646, + "learning_rate": 0.00019958383748200782, + "loss": 1.1153, + "step": 672 + }, + { + "epoch": 0.11983618233618233, + "grad_norm": 0.40149980783462524, + "learning_rate": 0.00019958256082047533, + "loss": 0.9785, + "step": 673 + }, + { + "epoch": 0.12001424501424501, + "grad_norm": 0.4378886818885803, + "learning_rate": 0.00019958128220782942, + "loss": 1.1355, + "step": 674 + }, + { + "epoch": 0.1201923076923077, + "grad_norm": 0.4449596703052521, + "learning_rate": 0.0001995800016440952, + "loss": 1.0325, + "step": 675 + }, + { + "epoch": 0.12037037037037036, + "grad_norm": 0.4268079698085785, + "learning_rate": 0.00019957871912929765, + "loss": 1.1901, + "step": 676 + }, + { + "epoch": 0.12054843304843305, + "grad_norm": 0.4250091016292572, + "learning_rate": 0.00019957743466346198, + "loss": 1.0084, + "step": 677 + }, + { + "epoch": 0.12072649572649573, + "grad_norm": 0.40724286437034607, + "learning_rate": 0.0001995761482466133, + "loss": 1.0866, + "step": 678 + }, + { + "epoch": 0.12090455840455841, + "grad_norm": 0.42478349804878235, + "learning_rate": 0.00019957485987877688, + "loss": 1.1909, + "step": 679 + }, + { + "epoch": 0.12108262108262108, + "grad_norm": 0.371362566947937, + "learning_rate": 0.0001995735695599779, + "loss": 1.083, + "step": 680 + }, + { + "epoch": 0.12126068376068376, + "grad_norm": 0.4715283513069153, + "learning_rate": 0.0001995722772902417, + "loss": 1.2942, + "step": 681 + }, + { + "epoch": 0.12143874643874644, + "grad_norm": 0.3611983060836792, + "learning_rate": 0.00019957098306959355, + "loss": 0.9878, + "step": 682 + }, + { + "epoch": 0.12161680911680911, + "grad_norm": 0.4764883816242218, + "learning_rate": 0.00019956968689805883, + "loss": 1.0082, + "step": 683 + }, + { + "epoch": 0.12179487179487179, + "grad_norm": 0.33170604705810547, + "learning_rate": 0.00019956838877566293, + "loss": 0.8529, + "step": 684 + }, + { + "epoch": 0.12197293447293447, + "grad_norm": 0.46896886825561523, + "learning_rate": 0.00019956708870243133, + "loss": 1.0745, + "step": 685 + }, + { + "epoch": 0.12215099715099716, + "grad_norm": 0.4120674431324005, + "learning_rate": 0.00019956578667838941, + "loss": 1.1828, + "step": 686 + }, + { + "epoch": 0.12232905982905982, + "grad_norm": 0.45671191811561584, + "learning_rate": 0.00019956448270356275, + "loss": 1.3484, + "step": 687 + }, + { + "epoch": 0.1225071225071225, + "grad_norm": 0.4023838937282562, + "learning_rate": 0.00019956317677797687, + "loss": 0.9623, + "step": 688 + }, + { + "epoch": 0.12268518518518519, + "grad_norm": 0.5205856561660767, + "learning_rate": 0.00019956186890165737, + "loss": 1.2221, + "step": 689 + }, + { + "epoch": 0.12286324786324786, + "grad_norm": 0.43956050276756287, + "learning_rate": 0.00019956055907462987, + "loss": 1.1051, + "step": 690 + }, + { + "epoch": 0.12304131054131054, + "grad_norm": 0.4341758191585541, + "learning_rate": 0.00019955924729692003, + "loss": 0.8972, + "step": 691 + }, + { + "epoch": 0.12321937321937322, + "grad_norm": 0.42025020718574524, + "learning_rate": 0.00019955793356855357, + "loss": 1.1137, + "step": 692 + }, + { + "epoch": 0.1233974358974359, + "grad_norm": 0.44375079870224, + "learning_rate": 0.0001995566178895562, + "loss": 1.2783, + "step": 693 + }, + { + "epoch": 0.12357549857549857, + "grad_norm": 0.4703320264816284, + "learning_rate": 0.00019955530025995372, + "loss": 1.1991, + "step": 694 + }, + { + "epoch": 0.12375356125356125, + "grad_norm": 0.43781620264053345, + "learning_rate": 0.00019955398067977195, + "loss": 1.2316, + "step": 695 + }, + { + "epoch": 0.12393162393162394, + "grad_norm": 0.4362877607345581, + "learning_rate": 0.0001995526591490367, + "loss": 1.1374, + "step": 696 + }, + { + "epoch": 0.1241096866096866, + "grad_norm": 0.4434499442577362, + "learning_rate": 0.00019955133566777392, + "loss": 1.1034, + "step": 697 + }, + { + "epoch": 0.12428774928774929, + "grad_norm": 0.46613508462905884, + "learning_rate": 0.00019955001023600955, + "loss": 1.2252, + "step": 698 + }, + { + "epoch": 0.12446581196581197, + "grad_norm": 0.46226736903190613, + "learning_rate": 0.00019954868285376945, + "loss": 1.0296, + "step": 699 + }, + { + "epoch": 0.12464387464387465, + "grad_norm": 0.4460904002189636, + "learning_rate": 0.00019954735352107977, + "loss": 1.0553, + "step": 700 + }, + { + "epoch": 0.12482193732193732, + "grad_norm": 0.36708924174308777, + "learning_rate": 0.00019954602223796648, + "loss": 0.9384, + "step": 701 + }, + { + "epoch": 0.125, + "grad_norm": 0.3780093491077423, + "learning_rate": 0.00019954468900445566, + "loss": 0.9062, + "step": 702 + }, + { + "epoch": 0.12517806267806267, + "grad_norm": 0.41797417402267456, + "learning_rate": 0.00019954335382057345, + "loss": 1.0344, + "step": 703 + }, + { + "epoch": 0.12535612535612536, + "grad_norm": 0.43710798025131226, + "learning_rate": 0.00019954201668634597, + "loss": 1.1324, + "step": 704 + }, + { + "epoch": 0.12553418803418803, + "grad_norm": 0.4732789695262909, + "learning_rate": 0.00019954067760179952, + "loss": 1.1419, + "step": 705 + }, + { + "epoch": 0.1257122507122507, + "grad_norm": 0.43248575925827026, + "learning_rate": 0.00019953933656696022, + "loss": 1.5112, + "step": 706 + }, + { + "epoch": 0.1258903133903134, + "grad_norm": 0.4074753522872925, + "learning_rate": 0.00019953799358185442, + "loss": 0.9751, + "step": 707 + }, + { + "epoch": 0.12606837606837606, + "grad_norm": 0.4586823880672455, + "learning_rate": 0.0001995366486465084, + "loss": 1.267, + "step": 708 + }, + { + "epoch": 0.12624643874643873, + "grad_norm": 0.4716857075691223, + "learning_rate": 0.0001995353017609485, + "loss": 1.1636, + "step": 709 + }, + { + "epoch": 0.12642450142450143, + "grad_norm": 0.5214398503303528, + "learning_rate": 0.00019953395292520115, + "loss": 1.2317, + "step": 710 + }, + { + "epoch": 0.1266025641025641, + "grad_norm": 0.42961129546165466, + "learning_rate": 0.00019953260213929276, + "loss": 1.0271, + "step": 711 + }, + { + "epoch": 0.1267806267806268, + "grad_norm": 0.4764653444290161, + "learning_rate": 0.00019953124940324979, + "loss": 1.1747, + "step": 712 + }, + { + "epoch": 0.12695868945868946, + "grad_norm": 0.4420304000377655, + "learning_rate": 0.00019952989471709874, + "loss": 0.9783, + "step": 713 + }, + { + "epoch": 0.12713675213675213, + "grad_norm": 0.44114625453948975, + "learning_rate": 0.00019952853808086616, + "loss": 1.1953, + "step": 714 + }, + { + "epoch": 0.12731481481481483, + "grad_norm": 0.501923143863678, + "learning_rate": 0.0001995271794945786, + "loss": 0.9886, + "step": 715 + }, + { + "epoch": 0.1274928774928775, + "grad_norm": 0.42266538739204407, + "learning_rate": 0.00019952581895826276, + "loss": 1.2033, + "step": 716 + }, + { + "epoch": 0.12767094017094016, + "grad_norm": 0.37770554423332214, + "learning_rate": 0.00019952445647194523, + "loss": 1.0164, + "step": 717 + }, + { + "epoch": 0.12784900284900286, + "grad_norm": 0.369266152381897, + "learning_rate": 0.00019952309203565268, + "loss": 0.9186, + "step": 718 + }, + { + "epoch": 0.12802706552706553, + "grad_norm": 0.40446221828460693, + "learning_rate": 0.00019952172564941193, + "loss": 1.1576, + "step": 719 + }, + { + "epoch": 0.1282051282051282, + "grad_norm": 0.504172146320343, + "learning_rate": 0.00019952035731324967, + "loss": 1.2695, + "step": 720 + }, + { + "epoch": 0.1283831908831909, + "grad_norm": 0.37284108996391296, + "learning_rate": 0.0001995189870271928, + "loss": 1.0288, + "step": 721 + }, + { + "epoch": 0.12856125356125356, + "grad_norm": 0.41811618208885193, + "learning_rate": 0.00019951761479126805, + "loss": 1.2241, + "step": 722 + }, + { + "epoch": 0.12873931623931623, + "grad_norm": 0.44706249237060547, + "learning_rate": 0.0001995162406055024, + "loss": 1.0831, + "step": 723 + }, + { + "epoch": 0.12891737891737892, + "grad_norm": 0.426572322845459, + "learning_rate": 0.00019951486446992273, + "loss": 1.0047, + "step": 724 + }, + { + "epoch": 0.1290954415954416, + "grad_norm": 0.4446277618408203, + "learning_rate": 0.00019951348638455602, + "loss": 1.0827, + "step": 725 + }, + { + "epoch": 0.12927350427350429, + "grad_norm": 0.3934919834136963, + "learning_rate": 0.00019951210634942926, + "loss": 0.9808, + "step": 726 + }, + { + "epoch": 0.12945156695156695, + "grad_norm": 0.4316558241844177, + "learning_rate": 0.0001995107243645695, + "loss": 1.3341, + "step": 727 + }, + { + "epoch": 0.12962962962962962, + "grad_norm": 0.43074217438697815, + "learning_rate": 0.00019950934043000382, + "loss": 1.007, + "step": 728 + }, + { + "epoch": 0.12980769230769232, + "grad_norm": 0.5212171673774719, + "learning_rate": 0.0001995079545457593, + "loss": 1.1822, + "step": 729 + }, + { + "epoch": 0.129985754985755, + "grad_norm": 0.3749600946903229, + "learning_rate": 0.00019950656671186313, + "loss": 0.9657, + "step": 730 + }, + { + "epoch": 0.13016381766381765, + "grad_norm": 0.36626043915748596, + "learning_rate": 0.00019950517692834252, + "loss": 1.1274, + "step": 731 + }, + { + "epoch": 0.13034188034188035, + "grad_norm": 0.4635467529296875, + "learning_rate": 0.00019950378519522467, + "loss": 1.2305, + "step": 732 + }, + { + "epoch": 0.13051994301994302, + "grad_norm": 0.4077455699443817, + "learning_rate": 0.00019950239151253683, + "loss": 0.9485, + "step": 733 + }, + { + "epoch": 0.1306980056980057, + "grad_norm": 0.4222758114337921, + "learning_rate": 0.0001995009958803063, + "loss": 1.0376, + "step": 734 + }, + { + "epoch": 0.13087606837606838, + "grad_norm": 0.4330402612686157, + "learning_rate": 0.0001994995982985605, + "loss": 1.1774, + "step": 735 + }, + { + "epoch": 0.13105413105413105, + "grad_norm": 0.42275673151016235, + "learning_rate": 0.00019949819876732673, + "loss": 1.1238, + "step": 736 + }, + { + "epoch": 0.13123219373219372, + "grad_norm": 0.45576968789100647, + "learning_rate": 0.00019949679728663246, + "loss": 1.0428, + "step": 737 + }, + { + "epoch": 0.13141025641025642, + "grad_norm": 0.5508752465248108, + "learning_rate": 0.00019949539385650514, + "loss": 1.3221, + "step": 738 + }, + { + "epoch": 0.13158831908831908, + "grad_norm": 0.4115872383117676, + "learning_rate": 0.00019949398847697225, + "loss": 1.0301, + "step": 739 + }, + { + "epoch": 0.13176638176638178, + "grad_norm": 0.4662442207336426, + "learning_rate": 0.00019949258114806132, + "loss": 1.3263, + "step": 740 + }, + { + "epoch": 0.13194444444444445, + "grad_norm": 0.6077266931533813, + "learning_rate": 0.00019949117186979999, + "loss": 1.0269, + "step": 741 + }, + { + "epoch": 0.13212250712250712, + "grad_norm": 0.47039318084716797, + "learning_rate": 0.00019948976064221579, + "loss": 1.3782, + "step": 742 + }, + { + "epoch": 0.1323005698005698, + "grad_norm": 0.4773450493812561, + "learning_rate": 0.0001994883474653364, + "loss": 1.289, + "step": 743 + }, + { + "epoch": 0.13247863247863248, + "grad_norm": 0.40180155634880066, + "learning_rate": 0.00019948693233918952, + "loss": 0.8691, + "step": 744 + }, + { + "epoch": 0.13265669515669515, + "grad_norm": 0.45216289162635803, + "learning_rate": 0.00019948551526380288, + "loss": 1.071, + "step": 745 + }, + { + "epoch": 0.13283475783475784, + "grad_norm": 0.4289272427558899, + "learning_rate": 0.0001994840962392042, + "loss": 1.0422, + "step": 746 + }, + { + "epoch": 0.1330128205128205, + "grad_norm": 0.4617730379104614, + "learning_rate": 0.00019948267526542134, + "loss": 1.0835, + "step": 747 + }, + { + "epoch": 0.13319088319088318, + "grad_norm": 0.42710617184638977, + "learning_rate": 0.00019948125234248208, + "loss": 1.0535, + "step": 748 + }, + { + "epoch": 0.13336894586894588, + "grad_norm": 0.43433234095573425, + "learning_rate": 0.0001994798274704144, + "loss": 0.9313, + "step": 749 + }, + { + "epoch": 0.13354700854700854, + "grad_norm": 0.46270284056663513, + "learning_rate": 0.0001994784006492461, + "loss": 1.0903, + "step": 750 + }, + { + "epoch": 0.1337250712250712, + "grad_norm": 0.5319814682006836, + "learning_rate": 0.00019947697187900517, + "loss": 1.2329, + "step": 751 + }, + { + "epoch": 0.1339031339031339, + "grad_norm": 0.3511372208595276, + "learning_rate": 0.00019947554115971967, + "loss": 0.7116, + "step": 752 + }, + { + "epoch": 0.13408119658119658, + "grad_norm": 0.4103890359401703, + "learning_rate": 0.00019947410849141756, + "loss": 1.1527, + "step": 753 + }, + { + "epoch": 0.13425925925925927, + "grad_norm": 0.5390757322311401, + "learning_rate": 0.00019947267387412695, + "loss": 1.1682, + "step": 754 + }, + { + "epoch": 0.13443732193732194, + "grad_norm": 0.29939723014831543, + "learning_rate": 0.0001994712373078759, + "loss": 0.5848, + "step": 755 + }, + { + "epoch": 0.1346153846153846, + "grad_norm": 0.4605920612812042, + "learning_rate": 0.0001994697987926926, + "loss": 0.9448, + "step": 756 + }, + { + "epoch": 0.1347934472934473, + "grad_norm": 0.426213800907135, + "learning_rate": 0.00019946835832860527, + "loss": 1.0487, + "step": 757 + }, + { + "epoch": 0.13497150997150997, + "grad_norm": 0.4209515154361725, + "learning_rate": 0.00019946691591564203, + "loss": 1.0951, + "step": 758 + }, + { + "epoch": 0.13514957264957264, + "grad_norm": 0.39555591344833374, + "learning_rate": 0.0001994654715538312, + "loss": 0.8754, + "step": 759 + }, + { + "epoch": 0.13532763532763534, + "grad_norm": 0.4065483510494232, + "learning_rate": 0.0001994640252432011, + "loss": 0.9451, + "step": 760 + }, + { + "epoch": 0.135505698005698, + "grad_norm": 0.4489104151725769, + "learning_rate": 0.00019946257698378003, + "loss": 1.2031, + "step": 761 + }, + { + "epoch": 0.13568376068376067, + "grad_norm": 0.39928409457206726, + "learning_rate": 0.0001994611267755964, + "loss": 1.1124, + "step": 762 + }, + { + "epoch": 0.13586182336182337, + "grad_norm": 0.4145409166812897, + "learning_rate": 0.00019945967461867858, + "loss": 1.083, + "step": 763 + }, + { + "epoch": 0.13603988603988604, + "grad_norm": 0.43508613109588623, + "learning_rate": 0.00019945822051305507, + "loss": 1.1119, + "step": 764 + }, + { + "epoch": 0.1362179487179487, + "grad_norm": 0.5186598300933838, + "learning_rate": 0.0001994567644587543, + "loss": 1.3256, + "step": 765 + }, + { + "epoch": 0.1363960113960114, + "grad_norm": 0.4615778625011444, + "learning_rate": 0.00019945530645580487, + "loss": 1.3906, + "step": 766 + }, + { + "epoch": 0.13657407407407407, + "grad_norm": 0.4838152527809143, + "learning_rate": 0.00019945384650423532, + "loss": 0.8169, + "step": 767 + }, + { + "epoch": 0.13675213675213677, + "grad_norm": 0.49253368377685547, + "learning_rate": 0.0001994523846040742, + "loss": 1.1613, + "step": 768 + }, + { + "epoch": 0.13693019943019943, + "grad_norm": 0.4697009325027466, + "learning_rate": 0.00019945092075535024, + "loss": 1.1722, + "step": 769 + }, + { + "epoch": 0.1371082621082621, + "grad_norm": 0.47162383794784546, + "learning_rate": 0.00019944945495809204, + "loss": 1.054, + "step": 770 + }, + { + "epoch": 0.1372863247863248, + "grad_norm": 0.4653547704219818, + "learning_rate": 0.00019944798721232835, + "loss": 1.1791, + "step": 771 + }, + { + "epoch": 0.13746438746438747, + "grad_norm": 0.4244011640548706, + "learning_rate": 0.000199446517518088, + "loss": 1.1557, + "step": 772 + }, + { + "epoch": 0.13764245014245013, + "grad_norm": 0.43812859058380127, + "learning_rate": 0.00019944504587539967, + "loss": 1.1567, + "step": 773 + }, + { + "epoch": 0.13782051282051283, + "grad_norm": 0.3984275162220001, + "learning_rate": 0.00019944357228429227, + "loss": 1.0715, + "step": 774 + }, + { + "epoch": 0.1379985754985755, + "grad_norm": 0.3794248104095459, + "learning_rate": 0.0001994420967447946, + "loss": 0.9377, + "step": 775 + }, + { + "epoch": 0.13817663817663817, + "grad_norm": 0.4214578866958618, + "learning_rate": 0.00019944061925693566, + "loss": 1.0112, + "step": 776 + }, + { + "epoch": 0.13835470085470086, + "grad_norm": 0.4738999605178833, + "learning_rate": 0.00019943913982074435, + "loss": 0.8718, + "step": 777 + }, + { + "epoch": 0.13853276353276353, + "grad_norm": 0.43455326557159424, + "learning_rate": 0.00019943765843624965, + "loss": 1.1343, + "step": 778 + }, + { + "epoch": 0.1387108262108262, + "grad_norm": 0.44973456859588623, + "learning_rate": 0.00019943617510348062, + "loss": 1.0487, + "step": 779 + }, + { + "epoch": 0.1388888888888889, + "grad_norm": 0.4216597080230713, + "learning_rate": 0.00019943468982246628, + "loss": 1.0765, + "step": 780 + }, + { + "epoch": 0.13906695156695156, + "grad_norm": 0.5089883208274841, + "learning_rate": 0.00019943320259323578, + "loss": 1.3137, + "step": 781 + }, + { + "epoch": 0.13924501424501423, + "grad_norm": 0.4358222782611847, + "learning_rate": 0.00019943171341581822, + "loss": 1.1891, + "step": 782 + }, + { + "epoch": 0.13942307692307693, + "grad_norm": 0.40918609499931335, + "learning_rate": 0.00019943022229024275, + "loss": 1.279, + "step": 783 + }, + { + "epoch": 0.1396011396011396, + "grad_norm": 0.4614863395690918, + "learning_rate": 0.00019942872921653866, + "loss": 1.2477, + "step": 784 + }, + { + "epoch": 0.1397792022792023, + "grad_norm": 0.4141528904438019, + "learning_rate": 0.00019942723419473515, + "loss": 0.9622, + "step": 785 + }, + { + "epoch": 0.13995726495726496, + "grad_norm": 0.536139726638794, + "learning_rate": 0.00019942573722486154, + "loss": 1.2127, + "step": 786 + }, + { + "epoch": 0.14013532763532763, + "grad_norm": 0.4968845546245575, + "learning_rate": 0.0001994242383069471, + "loss": 1.2965, + "step": 787 + }, + { + "epoch": 0.14031339031339032, + "grad_norm": 0.3897174894809723, + "learning_rate": 0.00019942273744102132, + "loss": 0.9907, + "step": 788 + }, + { + "epoch": 0.140491452991453, + "grad_norm": 0.466307669878006, + "learning_rate": 0.0001994212346271135, + "loss": 1.2021, + "step": 789 + }, + { + "epoch": 0.14066951566951566, + "grad_norm": 0.49283576011657715, + "learning_rate": 0.0001994197298652531, + "loss": 1.0969, + "step": 790 + }, + { + "epoch": 0.14084757834757836, + "grad_norm": 0.4686102271080017, + "learning_rate": 0.00019941822315546964, + "loss": 1.0125, + "step": 791 + }, + { + "epoch": 0.14102564102564102, + "grad_norm": 0.4389997124671936, + "learning_rate": 0.0001994167144977926, + "loss": 1.1294, + "step": 792 + }, + { + "epoch": 0.1412037037037037, + "grad_norm": 0.38539355993270874, + "learning_rate": 0.00019941520389225162, + "loss": 1.1231, + "step": 793 + }, + { + "epoch": 0.1413817663817664, + "grad_norm": 0.4860847592353821, + "learning_rate": 0.00019941369133887618, + "loss": 1.2268, + "step": 794 + }, + { + "epoch": 0.14155982905982906, + "grad_norm": 0.4567467272281647, + "learning_rate": 0.00019941217683769598, + "loss": 1.1482, + "step": 795 + }, + { + "epoch": 0.14173789173789172, + "grad_norm": 0.5549420714378357, + "learning_rate": 0.00019941066038874067, + "loss": 1.1899, + "step": 796 + }, + { + "epoch": 0.14191595441595442, + "grad_norm": 0.3950003385543823, + "learning_rate": 0.00019940914199204, + "loss": 0.96, + "step": 797 + }, + { + "epoch": 0.1420940170940171, + "grad_norm": 0.43845999240875244, + "learning_rate": 0.00019940762164762373, + "loss": 1.0338, + "step": 798 + }, + { + "epoch": 0.14227207977207978, + "grad_norm": 0.468537300825119, + "learning_rate": 0.00019940609935552157, + "loss": 1.2416, + "step": 799 + }, + { + "epoch": 0.14245014245014245, + "grad_norm": 0.4292038679122925, + "learning_rate": 0.0001994045751157634, + "loss": 1.1397, + "step": 800 + }, + { + "epoch": 0.14262820512820512, + "grad_norm": 0.3800995647907257, + "learning_rate": 0.00019940304892837908, + "loss": 0.939, + "step": 801 + }, + { + "epoch": 0.14280626780626782, + "grad_norm": 0.38004353642463684, + "learning_rate": 0.00019940152079339852, + "loss": 1.0485, + "step": 802 + }, + { + "epoch": 0.14298433048433049, + "grad_norm": 0.4658142924308777, + "learning_rate": 0.00019939999071085163, + "loss": 1.1561, + "step": 803 + }, + { + "epoch": 0.14316239316239315, + "grad_norm": 0.4235048294067383, + "learning_rate": 0.0001993984586807684, + "loss": 1.0516, + "step": 804 + }, + { + "epoch": 0.14334045584045585, + "grad_norm": 0.42925819754600525, + "learning_rate": 0.00019939692470317887, + "loss": 1.2238, + "step": 805 + }, + { + "epoch": 0.14351851851851852, + "grad_norm": 0.43701639771461487, + "learning_rate": 0.00019939538877811308, + "loss": 1.0129, + "step": 806 + }, + { + "epoch": 0.14369658119658119, + "grad_norm": 0.42786353826522827, + "learning_rate": 0.00019939385090560113, + "loss": 1.1355, + "step": 807 + }, + { + "epoch": 0.14387464387464388, + "grad_norm": 0.371218740940094, + "learning_rate": 0.00019939231108567312, + "loss": 0.9712, + "step": 808 + }, + { + "epoch": 0.14405270655270655, + "grad_norm": 0.4834294617176056, + "learning_rate": 0.00019939076931835926, + "loss": 1.1375, + "step": 809 + }, + { + "epoch": 0.14423076923076922, + "grad_norm": 0.4700150191783905, + "learning_rate": 0.00019938922560368974, + "loss": 1.1943, + "step": 810 + }, + { + "epoch": 0.14440883190883191, + "grad_norm": 0.4430996775627136, + "learning_rate": 0.0001993876799416948, + "loss": 1.1976, + "step": 811 + }, + { + "epoch": 0.14458689458689458, + "grad_norm": 0.4161672592163086, + "learning_rate": 0.00019938613233240476, + "loss": 1.0291, + "step": 812 + }, + { + "epoch": 0.14476495726495728, + "grad_norm": 0.39838850498199463, + "learning_rate": 0.0001993845827758499, + "loss": 1.2103, + "step": 813 + }, + { + "epoch": 0.14494301994301995, + "grad_norm": 0.429198294878006, + "learning_rate": 0.00019938303127206057, + "loss": 0.9971, + "step": 814 + }, + { + "epoch": 0.14512108262108261, + "grad_norm": 0.4589254856109619, + "learning_rate": 0.00019938147782106719, + "loss": 1.2392, + "step": 815 + }, + { + "epoch": 0.1452991452991453, + "grad_norm": 0.42506635189056396, + "learning_rate": 0.00019937992242290023, + "loss": 1.0827, + "step": 816 + }, + { + "epoch": 0.14547720797720798, + "grad_norm": 0.3778113126754761, + "learning_rate": 0.00019937836507759012, + "loss": 1.021, + "step": 817 + }, + { + "epoch": 0.14565527065527065, + "grad_norm": 0.43071216344833374, + "learning_rate": 0.0001993768057851674, + "loss": 1.273, + "step": 818 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 0.4944681227207184, + "learning_rate": 0.00019937524454566262, + "loss": 1.3037, + "step": 819 + }, + { + "epoch": 0.146011396011396, + "grad_norm": 0.4438824951648712, + "learning_rate": 0.00019937368135910632, + "loss": 1.1383, + "step": 820 + }, + { + "epoch": 0.14618945868945868, + "grad_norm": 0.400215744972229, + "learning_rate": 0.0001993721162255292, + "loss": 1.0669, + "step": 821 + }, + { + "epoch": 0.14636752136752137, + "grad_norm": 0.4341452121734619, + "learning_rate": 0.00019937054914496185, + "loss": 1.1431, + "step": 822 + }, + { + "epoch": 0.14654558404558404, + "grad_norm": 0.3941744267940521, + "learning_rate": 0.00019936898011743503, + "loss": 1.1593, + "step": 823 + }, + { + "epoch": 0.1467236467236467, + "grad_norm": 0.4318541884422302, + "learning_rate": 0.00019936740914297947, + "loss": 1.2814, + "step": 824 + }, + { + "epoch": 0.1469017094017094, + "grad_norm": 0.44488632678985596, + "learning_rate": 0.00019936583622162595, + "loss": 1.1054, + "step": 825 + }, + { + "epoch": 0.14707977207977208, + "grad_norm": 0.38701096177101135, + "learning_rate": 0.00019936426135340528, + "loss": 1.1086, + "step": 826 + }, + { + "epoch": 0.14725783475783477, + "grad_norm": 0.45794424414634705, + "learning_rate": 0.0001993626845383483, + "loss": 1.2395, + "step": 827 + }, + { + "epoch": 0.14743589743589744, + "grad_norm": 0.49237680435180664, + "learning_rate": 0.00019936110577648596, + "loss": 1.3483, + "step": 828 + }, + { + "epoch": 0.1476139601139601, + "grad_norm": 0.481666624546051, + "learning_rate": 0.00019935952506784914, + "loss": 1.1848, + "step": 829 + }, + { + "epoch": 0.1477920227920228, + "grad_norm": 0.4015209376811981, + "learning_rate": 0.00019935794241246883, + "loss": 1.0624, + "step": 830 + }, + { + "epoch": 0.14797008547008547, + "grad_norm": 0.47975999116897583, + "learning_rate": 0.00019935635781037606, + "loss": 1.1595, + "step": 831 + }, + { + "epoch": 0.14814814814814814, + "grad_norm": 0.4440356492996216, + "learning_rate": 0.00019935477126160181, + "loss": 1.1325, + "step": 832 + }, + { + "epoch": 0.14832621082621084, + "grad_norm": 0.4167410731315613, + "learning_rate": 0.00019935318276617723, + "loss": 1.0662, + "step": 833 + }, + { + "epoch": 0.1485042735042735, + "grad_norm": 0.4107447862625122, + "learning_rate": 0.0001993515923241334, + "loss": 0.8816, + "step": 834 + }, + { + "epoch": 0.14868233618233617, + "grad_norm": 0.4020158648490906, + "learning_rate": 0.00019934999993550154, + "loss": 0.9797, + "step": 835 + }, + { + "epoch": 0.14886039886039887, + "grad_norm": 0.4186473786830902, + "learning_rate": 0.0001993484056003128, + "loss": 1.1243, + "step": 836 + }, + { + "epoch": 0.14903846153846154, + "grad_norm": 0.5534794926643372, + "learning_rate": 0.00019934680931859842, + "loss": 1.1189, + "step": 837 + }, + { + "epoch": 0.1492165242165242, + "grad_norm": 0.37901270389556885, + "learning_rate": 0.0001993452110903897, + "loss": 0.9241, + "step": 838 + }, + { + "epoch": 0.1493945868945869, + "grad_norm": 0.41773587465286255, + "learning_rate": 0.00019934361091571793, + "loss": 0.9467, + "step": 839 + }, + { + "epoch": 0.14957264957264957, + "grad_norm": 0.4962073564529419, + "learning_rate": 0.00019934200879461448, + "loss": 1.2423, + "step": 840 + }, + { + "epoch": 0.14975071225071226, + "grad_norm": 0.38565897941589355, + "learning_rate": 0.00019934040472711074, + "loss": 1.1545, + "step": 841 + }, + { + "epoch": 0.14992877492877493, + "grad_norm": 0.4295346736907959, + "learning_rate": 0.0001993387987132381, + "loss": 1.2482, + "step": 842 + }, + { + "epoch": 0.1501068376068376, + "grad_norm": 0.4279189705848694, + "learning_rate": 0.0001993371907530281, + "loss": 1.1135, + "step": 843 + }, + { + "epoch": 0.1502849002849003, + "grad_norm": 0.44649168848991394, + "learning_rate": 0.0001993355808465122, + "loss": 1.0734, + "step": 844 + }, + { + "epoch": 0.15046296296296297, + "grad_norm": 0.453707218170166, + "learning_rate": 0.0001993339689937219, + "loss": 1.0992, + "step": 845 + }, + { + "epoch": 0.15064102564102563, + "grad_norm": 0.5113263726234436, + "learning_rate": 0.00019933235519468886, + "loss": 1.1792, + "step": 846 + }, + { + "epoch": 0.15081908831908833, + "grad_norm": 0.5822970271110535, + "learning_rate": 0.00019933073944944466, + "loss": 1.367, + "step": 847 + }, + { + "epoch": 0.150997150997151, + "grad_norm": 0.3946528732776642, + "learning_rate": 0.00019932912175802097, + "loss": 0.9781, + "step": 848 + }, + { + "epoch": 0.15117521367521367, + "grad_norm": 0.5429860949516296, + "learning_rate": 0.00019932750212044945, + "loss": 0.9783, + "step": 849 + }, + { + "epoch": 0.15135327635327636, + "grad_norm": 0.45847952365875244, + "learning_rate": 0.0001993258805367619, + "loss": 1.1352, + "step": 850 + }, + { + "epoch": 0.15153133903133903, + "grad_norm": 0.42770692706108093, + "learning_rate": 0.00019932425700699004, + "loss": 1.2365, + "step": 851 + }, + { + "epoch": 0.1517094017094017, + "grad_norm": 0.41845405101776123, + "learning_rate": 0.00019932263153116565, + "loss": 1.2642, + "step": 852 + }, + { + "epoch": 0.1518874643874644, + "grad_norm": 0.4641731083393097, + "learning_rate": 0.00019932100410932066, + "loss": 1.2009, + "step": 853 + }, + { + "epoch": 0.15206552706552706, + "grad_norm": 0.4128672778606415, + "learning_rate": 0.00019931937474148689, + "loss": 1.1981, + "step": 854 + }, + { + "epoch": 0.15224358974358973, + "grad_norm": 0.4730764925479889, + "learning_rate": 0.00019931774342769632, + "loss": 1.2145, + "step": 855 + }, + { + "epoch": 0.15242165242165243, + "grad_norm": 0.36611825227737427, + "learning_rate": 0.00019931611016798089, + "loss": 0.8504, + "step": 856 + }, + { + "epoch": 0.1525997150997151, + "grad_norm": 0.40944692492485046, + "learning_rate": 0.00019931447496237254, + "loss": 1.2853, + "step": 857 + }, + { + "epoch": 0.1527777777777778, + "grad_norm": 0.4521993398666382, + "learning_rate": 0.0001993128378109034, + "loss": 1.0198, + "step": 858 + }, + { + "epoch": 0.15295584045584046, + "grad_norm": 0.42113015055656433, + "learning_rate": 0.0001993111987136055, + "loss": 1.1284, + "step": 859 + }, + { + "epoch": 0.15313390313390313, + "grad_norm": 0.4117624759674072, + "learning_rate": 0.00019930955767051098, + "loss": 1.0445, + "step": 860 + }, + { + "epoch": 0.15331196581196582, + "grad_norm": 0.4807964265346527, + "learning_rate": 0.00019930791468165197, + "loss": 1.1378, + "step": 861 + }, + { + "epoch": 0.1534900284900285, + "grad_norm": 0.4186483323574066, + "learning_rate": 0.00019930626974706063, + "loss": 1.1636, + "step": 862 + }, + { + "epoch": 0.15366809116809116, + "grad_norm": 0.3764737844467163, + "learning_rate": 0.00019930462286676926, + "loss": 0.9523, + "step": 863 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.4283556044101715, + "learning_rate": 0.00019930297404081008, + "loss": 1.1008, + "step": 864 + }, + { + "epoch": 0.15402421652421652, + "grad_norm": 0.4485796093940735, + "learning_rate": 0.00019930132326921541, + "loss": 1.0834, + "step": 865 + }, + { + "epoch": 0.1542022792022792, + "grad_norm": 0.3882720172405243, + "learning_rate": 0.0001992996705520176, + "loss": 1.1086, + "step": 866 + }, + { + "epoch": 0.1543803418803419, + "grad_norm": 0.44698455929756165, + "learning_rate": 0.00019929801588924902, + "loss": 1.1437, + "step": 867 + }, + { + "epoch": 0.15455840455840456, + "grad_norm": 0.46978411078453064, + "learning_rate": 0.00019929635928094208, + "loss": 1.091, + "step": 868 + }, + { + "epoch": 0.15473646723646722, + "grad_norm": 0.4717854857444763, + "learning_rate": 0.00019929470072712927, + "loss": 1.1959, + "step": 869 + }, + { + "epoch": 0.15491452991452992, + "grad_norm": 0.4324854016304016, + "learning_rate": 0.00019929304022784305, + "loss": 1.2062, + "step": 870 + }, + { + "epoch": 0.1550925925925926, + "grad_norm": 0.3948180675506592, + "learning_rate": 0.00019929137778311597, + "loss": 1.1101, + "step": 871 + }, + { + "epoch": 0.15527065527065528, + "grad_norm": 0.40345287322998047, + "learning_rate": 0.0001992897133929806, + "loss": 0.8894, + "step": 872 + }, + { + "epoch": 0.15544871794871795, + "grad_norm": 0.44931963086128235, + "learning_rate": 0.00019928804705746957, + "loss": 0.9389, + "step": 873 + }, + { + "epoch": 0.15562678062678062, + "grad_norm": 0.529196560382843, + "learning_rate": 0.0001992863787766155, + "loss": 1.3362, + "step": 874 + }, + { + "epoch": 0.15580484330484332, + "grad_norm": 0.41218671202659607, + "learning_rate": 0.0001992847085504511, + "loss": 1.0727, + "step": 875 + }, + { + "epoch": 0.15598290598290598, + "grad_norm": 0.44074541330337524, + "learning_rate": 0.00019928303637900907, + "loss": 1.1091, + "step": 876 + }, + { + "epoch": 0.15616096866096865, + "grad_norm": 0.5264310240745544, + "learning_rate": 0.00019928136226232218, + "loss": 1.201, + "step": 877 + }, + { + "epoch": 0.15633903133903135, + "grad_norm": 0.4255099594593048, + "learning_rate": 0.00019927968620042324, + "loss": 1.2514, + "step": 878 + }, + { + "epoch": 0.15651709401709402, + "grad_norm": 0.4030280113220215, + "learning_rate": 0.0001992780081933451, + "loss": 1.0422, + "step": 879 + }, + { + "epoch": 0.15669515669515668, + "grad_norm": 0.5270203948020935, + "learning_rate": 0.00019927632824112058, + "loss": 1.2476, + "step": 880 + }, + { + "epoch": 0.15687321937321938, + "grad_norm": 0.37767237424850464, + "learning_rate": 0.00019927464634378268, + "loss": 1.0768, + "step": 881 + }, + { + "epoch": 0.15705128205128205, + "grad_norm": 0.4535936415195465, + "learning_rate": 0.0001992729625013643, + "loss": 1.2097, + "step": 882 + }, + { + "epoch": 0.15722934472934472, + "grad_norm": 0.4282119870185852, + "learning_rate": 0.00019927127671389843, + "loss": 1.0904, + "step": 883 + }, + { + "epoch": 0.1574074074074074, + "grad_norm": 0.3924157917499542, + "learning_rate": 0.0001992695889814181, + "loss": 0.9692, + "step": 884 + }, + { + "epoch": 0.15758547008547008, + "grad_norm": 0.525075376033783, + "learning_rate": 0.0001992678993039564, + "loss": 1.0292, + "step": 885 + }, + { + "epoch": 0.15776353276353278, + "grad_norm": 0.4388505518436432, + "learning_rate": 0.00019926620768154644, + "loss": 1.1944, + "step": 886 + }, + { + "epoch": 0.15794159544159544, + "grad_norm": 0.4362235963344574, + "learning_rate": 0.00019926451411422132, + "loss": 0.97, + "step": 887 + }, + { + "epoch": 0.1581196581196581, + "grad_norm": 0.4265296459197998, + "learning_rate": 0.0001992628186020143, + "loss": 0.9196, + "step": 888 + }, + { + "epoch": 0.1582977207977208, + "grad_norm": 0.4019876718521118, + "learning_rate": 0.0001992611211449585, + "loss": 1.1368, + "step": 889 + }, + { + "epoch": 0.15847578347578348, + "grad_norm": 0.5003397464752197, + "learning_rate": 0.00019925942174308726, + "loss": 1.2582, + "step": 890 + }, + { + "epoch": 0.15865384615384615, + "grad_norm": 0.4774404466152191, + "learning_rate": 0.00019925772039643382, + "loss": 1.2277, + "step": 891 + }, + { + "epoch": 0.15883190883190884, + "grad_norm": 0.4590449333190918, + "learning_rate": 0.00019925601710503153, + "loss": 1.1679, + "step": 892 + }, + { + "epoch": 0.1590099715099715, + "grad_norm": 0.4221442639827728, + "learning_rate": 0.0001992543118689138, + "loss": 1.1626, + "step": 893 + }, + { + "epoch": 0.15918803418803418, + "grad_norm": 0.47613003849983215, + "learning_rate": 0.00019925260468811403, + "loss": 1.1509, + "step": 894 + }, + { + "epoch": 0.15936609686609687, + "grad_norm": 0.41706812381744385, + "learning_rate": 0.0001992508955626656, + "loss": 1.0366, + "step": 895 + }, + { + "epoch": 0.15954415954415954, + "grad_norm": 0.5064654350280762, + "learning_rate": 0.00019924918449260205, + "loss": 1.0729, + "step": 896 + }, + { + "epoch": 0.1597222222222222, + "grad_norm": 0.5019610524177551, + "learning_rate": 0.00019924747147795696, + "loss": 1.0642, + "step": 897 + }, + { + "epoch": 0.1599002849002849, + "grad_norm": 0.4345671534538269, + "learning_rate": 0.00019924575651876378, + "loss": 1.1747, + "step": 898 + }, + { + "epoch": 0.16007834757834757, + "grad_norm": 0.4397568702697754, + "learning_rate": 0.0001992440396150562, + "loss": 1.282, + "step": 899 + }, + { + "epoch": 0.16025641025641027, + "grad_norm": 0.520187497138977, + "learning_rate": 0.0001992423207668678, + "loss": 0.976, + "step": 900 + }, + { + "epoch": 0.16043447293447294, + "grad_norm": 0.39329993724823, + "learning_rate": 0.0001992405999742323, + "loss": 0.9829, + "step": 901 + }, + { + "epoch": 0.1606125356125356, + "grad_norm": 0.42361345887184143, + "learning_rate": 0.00019923887723718339, + "loss": 1.139, + "step": 902 + }, + { + "epoch": 0.1607905982905983, + "grad_norm": 0.3846314251422882, + "learning_rate": 0.00019923715255575482, + "loss": 0.8262, + "step": 903 + }, + { + "epoch": 0.16096866096866097, + "grad_norm": 0.39258381724357605, + "learning_rate": 0.0001992354259299804, + "loss": 0.9638, + "step": 904 + }, + { + "epoch": 0.16114672364672364, + "grad_norm": 0.4000850319862366, + "learning_rate": 0.00019923369735989397, + "loss": 0.91, + "step": 905 + }, + { + "epoch": 0.16132478632478633, + "grad_norm": 0.46303513646125793, + "learning_rate": 0.00019923196684552936, + "loss": 1.1447, + "step": 906 + }, + { + "epoch": 0.161502849002849, + "grad_norm": 0.38437438011169434, + "learning_rate": 0.0001992302343869205, + "loss": 1.0212, + "step": 907 + }, + { + "epoch": 0.16168091168091167, + "grad_norm": 0.44585472345352173, + "learning_rate": 0.00019922849998410135, + "loss": 1.1964, + "step": 908 + }, + { + "epoch": 0.16185897435897437, + "grad_norm": 0.41959813237190247, + "learning_rate": 0.00019922676363710583, + "loss": 0.9925, + "step": 909 + }, + { + "epoch": 0.16203703703703703, + "grad_norm": 0.47442761063575745, + "learning_rate": 0.00019922502534596803, + "loss": 0.9237, + "step": 910 + }, + { + "epoch": 0.1622150997150997, + "grad_norm": 0.5065128207206726, + "learning_rate": 0.00019922328511072198, + "loss": 1.2573, + "step": 911 + }, + { + "epoch": 0.1623931623931624, + "grad_norm": 0.4739879369735718, + "learning_rate": 0.0001992215429314018, + "loss": 1.4416, + "step": 912 + }, + { + "epoch": 0.16257122507122507, + "grad_norm": 0.48763832449913025, + "learning_rate": 0.00019921979880804157, + "loss": 1.0408, + "step": 913 + }, + { + "epoch": 0.16274928774928774, + "grad_norm": 0.4841614067554474, + "learning_rate": 0.0001992180527406755, + "loss": 1.1826, + "step": 914 + }, + { + "epoch": 0.16292735042735043, + "grad_norm": 0.49433308839797974, + "learning_rate": 0.0001992163047293378, + "loss": 1.3552, + "step": 915 + }, + { + "epoch": 0.1631054131054131, + "grad_norm": 0.4985002875328064, + "learning_rate": 0.0001992145547740627, + "loss": 1.2639, + "step": 916 + }, + { + "epoch": 0.1632834757834758, + "grad_norm": 0.40348032116889954, + "learning_rate": 0.00019921280287488448, + "loss": 1.1731, + "step": 917 + }, + { + "epoch": 0.16346153846153846, + "grad_norm": 0.5166002511978149, + "learning_rate": 0.0001992110490318375, + "loss": 1.0692, + "step": 918 + }, + { + "epoch": 0.16363960113960113, + "grad_norm": 0.44233468174934387, + "learning_rate": 0.00019920929324495615, + "loss": 1.0488, + "step": 919 + }, + { + "epoch": 0.16381766381766383, + "grad_norm": 0.43709903955459595, + "learning_rate": 0.00019920753551427476, + "loss": 0.8884, + "step": 920 + }, + { + "epoch": 0.1639957264957265, + "grad_norm": 0.4054167568683624, + "learning_rate": 0.00019920577583982778, + "loss": 0.9872, + "step": 921 + }, + { + "epoch": 0.16417378917378916, + "grad_norm": 0.4657362997531891, + "learning_rate": 0.0001992040142216497, + "loss": 1.4402, + "step": 922 + }, + { + "epoch": 0.16435185185185186, + "grad_norm": 0.42550426721572876, + "learning_rate": 0.0001992022506597751, + "loss": 1.0456, + "step": 923 + }, + { + "epoch": 0.16452991452991453, + "grad_norm": 0.49346762895584106, + "learning_rate": 0.00019920048515423842, + "loss": 1.527, + "step": 924 + }, + { + "epoch": 0.1647079772079772, + "grad_norm": 0.3970337510108948, + "learning_rate": 0.0001991987177050743, + "loss": 1.0363, + "step": 925 + }, + { + "epoch": 0.1648860398860399, + "grad_norm": 0.4027378559112549, + "learning_rate": 0.0001991969483123174, + "loss": 0.8416, + "step": 926 + }, + { + "epoch": 0.16506410256410256, + "grad_norm": 0.4181644916534424, + "learning_rate": 0.00019919517697600237, + "loss": 1.2253, + "step": 927 + }, + { + "epoch": 0.16524216524216523, + "grad_norm": 0.43686383962631226, + "learning_rate": 0.0001991934036961639, + "loss": 1.0808, + "step": 928 + }, + { + "epoch": 0.16542022792022792, + "grad_norm": 0.4242876172065735, + "learning_rate": 0.0001991916284728367, + "loss": 0.9483, + "step": 929 + }, + { + "epoch": 0.1655982905982906, + "grad_norm": 0.3690609037876129, + "learning_rate": 0.00019918985130605563, + "loss": 0.9495, + "step": 930 + }, + { + "epoch": 0.1657763532763533, + "grad_norm": 0.42184555530548096, + "learning_rate": 0.00019918807219585546, + "loss": 1.0966, + "step": 931 + }, + { + "epoch": 0.16595441595441596, + "grad_norm": 0.4342746138572693, + "learning_rate": 0.00019918629114227106, + "loss": 1.0875, + "step": 932 + }, + { + "epoch": 0.16613247863247863, + "grad_norm": 0.4191494286060333, + "learning_rate": 0.00019918450814533737, + "loss": 1.0777, + "step": 933 + }, + { + "epoch": 0.16631054131054132, + "grad_norm": 0.37124550342559814, + "learning_rate": 0.00019918272320508922, + "loss": 1.0131, + "step": 934 + }, + { + "epoch": 0.166488603988604, + "grad_norm": 0.4475722014904022, + "learning_rate": 0.00019918093632156168, + "loss": 1.1185, + "step": 935 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.4629058241844177, + "learning_rate": 0.0001991791474947897, + "loss": 1.0353, + "step": 936 + }, + { + "epoch": 0.16684472934472935, + "grad_norm": 0.48192909359931946, + "learning_rate": 0.00019917735672480834, + "loss": 1.1628, + "step": 937 + }, + { + "epoch": 0.16702279202279202, + "grad_norm": 0.5542252063751221, + "learning_rate": 0.00019917556401165273, + "loss": 1.3133, + "step": 938 + }, + { + "epoch": 0.1672008547008547, + "grad_norm": 0.4172651171684265, + "learning_rate": 0.00019917376935535796, + "loss": 1.1733, + "step": 939 + }, + { + "epoch": 0.16737891737891739, + "grad_norm": 0.4424920380115509, + "learning_rate": 0.0001991719727559592, + "loss": 1.0262, + "step": 940 + }, + { + "epoch": 0.16755698005698005, + "grad_norm": 0.4551742970943451, + "learning_rate": 0.00019917017421349162, + "loss": 1.0883, + "step": 941 + }, + { + "epoch": 0.16773504273504272, + "grad_norm": 0.45929640531539917, + "learning_rate": 0.00019916837372799048, + "loss": 1.1836, + "step": 942 + }, + { + "epoch": 0.16791310541310542, + "grad_norm": 0.4609353542327881, + "learning_rate": 0.0001991665712994911, + "loss": 1.0682, + "step": 943 + }, + { + "epoch": 0.16809116809116809, + "grad_norm": 0.42617303133010864, + "learning_rate": 0.00019916476692802873, + "loss": 1.074, + "step": 944 + }, + { + "epoch": 0.16826923076923078, + "grad_norm": 0.41919493675231934, + "learning_rate": 0.00019916296061363875, + "loss": 1.0969, + "step": 945 + }, + { + "epoch": 0.16844729344729345, + "grad_norm": 0.450979083776474, + "learning_rate": 0.00019916115235635656, + "loss": 1.1686, + "step": 946 + }, + { + "epoch": 0.16862535612535612, + "grad_norm": 0.42166751623153687, + "learning_rate": 0.00019915934215621758, + "loss": 0.9273, + "step": 947 + }, + { + "epoch": 0.16880341880341881, + "grad_norm": 0.4404160976409912, + "learning_rate": 0.00019915753001325729, + "loss": 1.1663, + "step": 948 + }, + { + "epoch": 0.16898148148148148, + "grad_norm": 0.42025226354599, + "learning_rate": 0.0001991557159275111, + "loss": 0.9433, + "step": 949 + }, + { + "epoch": 0.16915954415954415, + "grad_norm": 0.4277796745300293, + "learning_rate": 0.00019915389989901474, + "loss": 0.8475, + "step": 950 + }, + { + "epoch": 0.16933760683760685, + "grad_norm": 0.5162755250930786, + "learning_rate": 0.00019915208192780365, + "loss": 1.1155, + "step": 951 + }, + { + "epoch": 0.16951566951566951, + "grad_norm": 0.4214856028556824, + "learning_rate": 0.00019915026201391346, + "loss": 1.173, + "step": 952 + }, + { + "epoch": 0.16969373219373218, + "grad_norm": 0.4713292419910431, + "learning_rate": 0.00019914844015737985, + "loss": 1.1615, + "step": 953 + }, + { + "epoch": 0.16987179487179488, + "grad_norm": 0.461179256439209, + "learning_rate": 0.00019914661635823854, + "loss": 1.1169, + "step": 954 + }, + { + "epoch": 0.17004985754985755, + "grad_norm": 0.46200552582740784, + "learning_rate": 0.00019914479061652527, + "loss": 1.0274, + "step": 955 + }, + { + "epoch": 0.17022792022792022, + "grad_norm": 0.40968334674835205, + "learning_rate": 0.00019914296293227572, + "loss": 1.066, + "step": 956 + }, + { + "epoch": 0.1704059829059829, + "grad_norm": 0.40877434611320496, + "learning_rate": 0.0001991411333055258, + "loss": 1.1595, + "step": 957 + }, + { + "epoch": 0.17058404558404558, + "grad_norm": 0.42940187454223633, + "learning_rate": 0.00019913930173631132, + "loss": 1.0364, + "step": 958 + }, + { + "epoch": 0.17076210826210828, + "grad_norm": 0.49648910760879517, + "learning_rate": 0.00019913746822466819, + "loss": 1.0763, + "step": 959 + }, + { + "epoch": 0.17094017094017094, + "grad_norm": 0.4353426396846771, + "learning_rate": 0.00019913563277063228, + "loss": 0.9698, + "step": 960 + }, + { + "epoch": 0.1711182336182336, + "grad_norm": 0.45079681277275085, + "learning_rate": 0.00019913379537423958, + "loss": 1.2244, + "step": 961 + }, + { + "epoch": 0.1712962962962963, + "grad_norm": 0.4276828467845917, + "learning_rate": 0.00019913195603552607, + "loss": 0.9976, + "step": 962 + }, + { + "epoch": 0.17147435897435898, + "grad_norm": 0.41122403740882874, + "learning_rate": 0.00019913011475452785, + "loss": 1.0077, + "step": 963 + }, + { + "epoch": 0.17165242165242164, + "grad_norm": 0.43170276284217834, + "learning_rate": 0.00019912827153128096, + "loss": 1.1402, + "step": 964 + }, + { + "epoch": 0.17183048433048434, + "grad_norm": 0.37950268387794495, + "learning_rate": 0.0001991264263658215, + "loss": 0.9818, + "step": 965 + }, + { + "epoch": 0.172008547008547, + "grad_norm": 0.477333128452301, + "learning_rate": 0.00019912457925818562, + "loss": 1.1756, + "step": 966 + }, + { + "epoch": 0.17218660968660968, + "grad_norm": 0.4326401352882385, + "learning_rate": 0.00019912273020840954, + "loss": 1.3718, + "step": 967 + }, + { + "epoch": 0.17236467236467237, + "grad_norm": 0.37711042165756226, + "learning_rate": 0.00019912087921652945, + "loss": 0.9011, + "step": 968 + }, + { + "epoch": 0.17254273504273504, + "grad_norm": 0.50013667345047, + "learning_rate": 0.00019911902628258162, + "loss": 1.1163, + "step": 969 + }, + { + "epoch": 0.1727207977207977, + "grad_norm": 0.41913339495658875, + "learning_rate": 0.0001991171714066024, + "loss": 1.2614, + "step": 970 + }, + { + "epoch": 0.1728988603988604, + "grad_norm": 0.4075855612754822, + "learning_rate": 0.00019911531458862813, + "loss": 0.8984, + "step": 971 + }, + { + "epoch": 0.17307692307692307, + "grad_norm": 0.40277954936027527, + "learning_rate": 0.00019911345582869513, + "loss": 1.0851, + "step": 972 + }, + { + "epoch": 0.17325498575498577, + "grad_norm": 0.4312847852706909, + "learning_rate": 0.00019911159512683987, + "loss": 1.1273, + "step": 973 + }, + { + "epoch": 0.17343304843304844, + "grad_norm": 0.40303611755371094, + "learning_rate": 0.0001991097324830988, + "loss": 0.9645, + "step": 974 + }, + { + "epoch": 0.1736111111111111, + "grad_norm": 0.45560577511787415, + "learning_rate": 0.00019910786789750838, + "loss": 1.0864, + "step": 975 + }, + { + "epoch": 0.1737891737891738, + "grad_norm": 0.43775680661201477, + "learning_rate": 0.00019910600137010517, + "loss": 1.028, + "step": 976 + }, + { + "epoch": 0.17396723646723647, + "grad_norm": 0.3917224407196045, + "learning_rate": 0.00019910413290092572, + "loss": 1.0491, + "step": 977 + }, + { + "epoch": 0.17414529914529914, + "grad_norm": 0.4068751037120819, + "learning_rate": 0.0001991022624900067, + "loss": 1.0476, + "step": 978 + }, + { + "epoch": 0.17432336182336183, + "grad_norm": 0.4463370144367218, + "learning_rate": 0.0001991003901373847, + "loss": 1.0612, + "step": 979 + }, + { + "epoch": 0.1745014245014245, + "grad_norm": 0.46949052810668945, + "learning_rate": 0.0001990985158430964, + "loss": 1.3099, + "step": 980 + }, + { + "epoch": 0.17467948717948717, + "grad_norm": 0.4250012934207916, + "learning_rate": 0.00019909663960717856, + "loss": 0.9903, + "step": 981 + }, + { + "epoch": 0.17485754985754987, + "grad_norm": 0.5293903946876526, + "learning_rate": 0.0001990947614296679, + "loss": 0.9908, + "step": 982 + }, + { + "epoch": 0.17503561253561253, + "grad_norm": 0.3838284909725189, + "learning_rate": 0.0001990928813106013, + "loss": 0.716, + "step": 983 + }, + { + "epoch": 0.1752136752136752, + "grad_norm": 0.4597751200199127, + "learning_rate": 0.0001990909992500155, + "loss": 1.0126, + "step": 984 + }, + { + "epoch": 0.1753917378917379, + "grad_norm": 0.4844081699848175, + "learning_rate": 0.0001990891152479474, + "loss": 1.1043, + "step": 985 + }, + { + "epoch": 0.17556980056980057, + "grad_norm": 0.4763399660587311, + "learning_rate": 0.00019908722930443392, + "loss": 1.019, + "step": 986 + }, + { + "epoch": 0.17574786324786323, + "grad_norm": 0.4670077860355377, + "learning_rate": 0.00019908534141951204, + "loss": 1.1382, + "step": 987 + }, + { + "epoch": 0.17592592592592593, + "grad_norm": 0.39372730255126953, + "learning_rate": 0.00019908345159321873, + "loss": 1.1219, + "step": 988 + }, + { + "epoch": 0.1761039886039886, + "grad_norm": 0.41869843006134033, + "learning_rate": 0.00019908155982559098, + "loss": 0.9461, + "step": 989 + }, + { + "epoch": 0.1762820512820513, + "grad_norm": 0.4398406147956848, + "learning_rate": 0.00019907966611666593, + "loss": 1.1328, + "step": 990 + }, + { + "epoch": 0.17646011396011396, + "grad_norm": 0.4315733015537262, + "learning_rate": 0.0001990777704664806, + "loss": 1.0974, + "step": 991 + }, + { + "epoch": 0.17663817663817663, + "grad_norm": 0.42859575152397156, + "learning_rate": 0.00019907587287507222, + "loss": 1.2637, + "step": 992 + }, + { + "epoch": 0.17681623931623933, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.0001990739733424779, + "loss": 1.0699, + "step": 993 + }, + { + "epoch": 0.176994301994302, + "grad_norm": 0.4443826973438263, + "learning_rate": 0.00019907207186873488, + "loss": 1.0547, + "step": 994 + }, + { + "epoch": 0.17717236467236466, + "grad_norm": 0.4108099937438965, + "learning_rate": 0.00019907016845388043, + "loss": 1.1401, + "step": 995 + }, + { + "epoch": 0.17735042735042736, + "grad_norm": 0.4474675953388214, + "learning_rate": 0.00019906826309795182, + "loss": 1.0712, + "step": 996 + }, + { + "epoch": 0.17752849002849003, + "grad_norm": 0.4149756133556366, + "learning_rate": 0.00019906635580098638, + "loss": 0.9585, + "step": 997 + }, + { + "epoch": 0.1777065527065527, + "grad_norm": 0.4875968098640442, + "learning_rate": 0.00019906444656302152, + "loss": 1.0659, + "step": 998 + }, + { + "epoch": 0.1778846153846154, + "grad_norm": 0.5494784116744995, + "learning_rate": 0.0001990625353840946, + "loss": 1.2858, + "step": 999 + }, + { + "epoch": 0.17806267806267806, + "grad_norm": 0.425062358379364, + "learning_rate": 0.0001990606222642431, + "loss": 1.1826, + "step": 1000 + }, + { + "epoch": 0.17824074074074073, + "grad_norm": 0.3890725374221802, + "learning_rate": 0.00019905870720350445, + "loss": 0.9568, + "step": 1001 + }, + { + "epoch": 0.17841880341880342, + "grad_norm": 0.3884070813655853, + "learning_rate": 0.00019905679020191624, + "loss": 0.9674, + "step": 1002 + }, + { + "epoch": 0.1785968660968661, + "grad_norm": 0.49496129155158997, + "learning_rate": 0.00019905487125951597, + "loss": 0.9143, + "step": 1003 + }, + { + "epoch": 0.1787749287749288, + "grad_norm": 0.43448135256767273, + "learning_rate": 0.00019905295037634128, + "loss": 1.2677, + "step": 1004 + }, + { + "epoch": 0.17895299145299146, + "grad_norm": 0.47327905893325806, + "learning_rate": 0.00019905102755242982, + "loss": 0.9089, + "step": 1005 + }, + { + "epoch": 0.17913105413105412, + "grad_norm": 0.4962378442287445, + "learning_rate": 0.00019904910278781922, + "loss": 1.1748, + "step": 1006 + }, + { + "epoch": 0.17930911680911682, + "grad_norm": 0.4343934655189514, + "learning_rate": 0.0001990471760825472, + "loss": 1.2176, + "step": 1007 + }, + { + "epoch": 0.1794871794871795, + "grad_norm": 0.4695793092250824, + "learning_rate": 0.0001990452474366515, + "loss": 1.1822, + "step": 1008 + }, + { + "epoch": 0.17966524216524216, + "grad_norm": 0.4156060516834259, + "learning_rate": 0.00019904331685016995, + "loss": 0.8231, + "step": 1009 + }, + { + "epoch": 0.17984330484330485, + "grad_norm": 0.5068191885948181, + "learning_rate": 0.00019904138432314035, + "loss": 1.1363, + "step": 1010 + }, + { + "epoch": 0.18002136752136752, + "grad_norm": 0.5189786553382874, + "learning_rate": 0.00019903944985560058, + "loss": 1.3131, + "step": 1011 + }, + { + "epoch": 0.1801994301994302, + "grad_norm": 0.5126828551292419, + "learning_rate": 0.00019903751344758848, + "loss": 1.0305, + "step": 1012 + }, + { + "epoch": 0.18037749287749288, + "grad_norm": 0.41045933961868286, + "learning_rate": 0.00019903557509914205, + "loss": 1.2726, + "step": 1013 + }, + { + "epoch": 0.18055555555555555, + "grad_norm": 0.4141713082790375, + "learning_rate": 0.0001990336348102993, + "loss": 0.9606, + "step": 1014 + }, + { + "epoch": 0.18073361823361822, + "grad_norm": 0.42652079463005066, + "learning_rate": 0.00019903169258109812, + "loss": 1.0235, + "step": 1015 + }, + { + "epoch": 0.18091168091168092, + "grad_norm": 0.42098379135131836, + "learning_rate": 0.0001990297484115767, + "loss": 1.0602, + "step": 1016 + }, + { + "epoch": 0.18108974358974358, + "grad_norm": 0.49920013546943665, + "learning_rate": 0.0001990278023017731, + "loss": 1.3322, + "step": 1017 + }, + { + "epoch": 0.18126780626780628, + "grad_norm": 0.412304550409317, + "learning_rate": 0.00019902585425172537, + "loss": 1.1011, + "step": 1018 + }, + { + "epoch": 0.18144586894586895, + "grad_norm": 0.44226935505867004, + "learning_rate": 0.00019902390426147177, + "loss": 0.9777, + "step": 1019 + }, + { + "epoch": 0.18162393162393162, + "grad_norm": 0.4685269594192505, + "learning_rate": 0.00019902195233105046, + "loss": 1.3587, + "step": 1020 + }, + { + "epoch": 0.1818019943019943, + "grad_norm": 0.4500584304332733, + "learning_rate": 0.00019901999846049968, + "loss": 0.9888, + "step": 1021 + }, + { + "epoch": 0.18198005698005698, + "grad_norm": 0.48566994071006775, + "learning_rate": 0.00019901804264985774, + "loss": 1.2364, + "step": 1022 + }, + { + "epoch": 0.18215811965811965, + "grad_norm": 0.4063156247138977, + "learning_rate": 0.00019901608489916294, + "loss": 1.2224, + "step": 1023 + }, + { + "epoch": 0.18233618233618235, + "grad_norm": 0.471276193857193, + "learning_rate": 0.00019901412520845367, + "loss": 0.9926, + "step": 1024 + }, + { + "epoch": 0.182514245014245, + "grad_norm": 0.5165421366691589, + "learning_rate": 0.00019901216357776829, + "loss": 0.9595, + "step": 1025 + }, + { + "epoch": 0.18269230769230768, + "grad_norm": 0.4746754467487335, + "learning_rate": 0.0001990102000071452, + "loss": 1.2057, + "step": 1026 + }, + { + "epoch": 0.18287037037037038, + "grad_norm": 0.44803035259246826, + "learning_rate": 0.00019900823449662297, + "loss": 1.2114, + "step": 1027 + }, + { + "epoch": 0.18304843304843305, + "grad_norm": 0.47256240248680115, + "learning_rate": 0.00019900626704624005, + "loss": 1.112, + "step": 1028 + }, + { + "epoch": 0.18322649572649571, + "grad_norm": 0.4253387153148651, + "learning_rate": 0.000199004297656035, + "loss": 0.9899, + "step": 1029 + }, + { + "epoch": 0.1834045584045584, + "grad_norm": 0.44958099722862244, + "learning_rate": 0.00019900232632604636, + "loss": 1.1445, + "step": 1030 + }, + { + "epoch": 0.18358262108262108, + "grad_norm": 0.5296537280082703, + "learning_rate": 0.00019900035305631285, + "loss": 1.2502, + "step": 1031 + }, + { + "epoch": 0.18376068376068377, + "grad_norm": 0.5057148933410645, + "learning_rate": 0.00019899837784687302, + "loss": 1.1426, + "step": 1032 + }, + { + "epoch": 0.18393874643874644, + "grad_norm": 0.41463762521743774, + "learning_rate": 0.00019899640069776566, + "loss": 1.1854, + "step": 1033 + }, + { + "epoch": 0.1841168091168091, + "grad_norm": 0.45800045132637024, + "learning_rate": 0.00019899442160902945, + "loss": 1.2438, + "step": 1034 + }, + { + "epoch": 0.1842948717948718, + "grad_norm": 0.43450453877449036, + "learning_rate": 0.00019899244058070324, + "loss": 1.0598, + "step": 1035 + }, + { + "epoch": 0.18447293447293447, + "grad_norm": 0.4141148626804352, + "learning_rate": 0.00019899045761282577, + "loss": 1.0465, + "step": 1036 + }, + { + "epoch": 0.18465099715099714, + "grad_norm": 0.3938458263874054, + "learning_rate": 0.0001989884727054359, + "loss": 1.0142, + "step": 1037 + }, + { + "epoch": 0.18482905982905984, + "grad_norm": 0.43898263573646545, + "learning_rate": 0.00019898648585857257, + "loss": 0.9212, + "step": 1038 + }, + { + "epoch": 0.1850071225071225, + "grad_norm": 0.4425487816333771, + "learning_rate": 0.00019898449707227465, + "loss": 1.2987, + "step": 1039 + }, + { + "epoch": 0.18518518518518517, + "grad_norm": 0.4537975490093231, + "learning_rate": 0.00019898250634658115, + "loss": 1.2023, + "step": 1040 + }, + { + "epoch": 0.18536324786324787, + "grad_norm": 0.4107198119163513, + "learning_rate": 0.00019898051368153104, + "loss": 0.8443, + "step": 1041 + }, + { + "epoch": 0.18554131054131054, + "grad_norm": 0.4389404058456421, + "learning_rate": 0.0001989785190771634, + "loss": 1.0502, + "step": 1042 + }, + { + "epoch": 0.1857193732193732, + "grad_norm": 0.4288824796676636, + "learning_rate": 0.00019897652253351726, + "loss": 1.01, + "step": 1043 + }, + { + "epoch": 0.1858974358974359, + "grad_norm": 0.50815349817276, + "learning_rate": 0.00019897452405063178, + "loss": 1.0308, + "step": 1044 + }, + { + "epoch": 0.18607549857549857, + "grad_norm": 0.45252710580825806, + "learning_rate": 0.0001989725236285461, + "loss": 1.0967, + "step": 1045 + }, + { + "epoch": 0.18625356125356127, + "grad_norm": 0.45049402117729187, + "learning_rate": 0.00019897052126729943, + "loss": 1.0141, + "step": 1046 + }, + { + "epoch": 0.18643162393162394, + "grad_norm": 0.49637508392333984, + "learning_rate": 0.00019896851696693098, + "loss": 1.0997, + "step": 1047 + }, + { + "epoch": 0.1866096866096866, + "grad_norm": 0.4465886056423187, + "learning_rate": 0.00019896651072748005, + "loss": 1.1415, + "step": 1048 + }, + { + "epoch": 0.1867877492877493, + "grad_norm": 0.5309500694274902, + "learning_rate": 0.00019896450254898592, + "loss": 1.1028, + "step": 1049 + }, + { + "epoch": 0.18696581196581197, + "grad_norm": 0.3516653776168823, + "learning_rate": 0.00019896249243148793, + "loss": 0.9841, + "step": 1050 + }, + { + "epoch": 0.18714387464387464, + "grad_norm": 0.4529176950454712, + "learning_rate": 0.0001989604803750255, + "loss": 1.1335, + "step": 1051 + }, + { + "epoch": 0.18732193732193733, + "grad_norm": 0.47694942355155945, + "learning_rate": 0.000198958466379638, + "loss": 1.2383, + "step": 1052 + }, + { + "epoch": 0.1875, + "grad_norm": 0.5524206757545471, + "learning_rate": 0.0001989564504453649, + "loss": 1.3668, + "step": 1053 + }, + { + "epoch": 0.18767806267806267, + "grad_norm": 0.39203691482543945, + "learning_rate": 0.00019895443257224576, + "loss": 1.2203, + "step": 1054 + }, + { + "epoch": 0.18785612535612536, + "grad_norm": 0.4164120852947235, + "learning_rate": 0.00019895241276032005, + "loss": 0.8954, + "step": 1055 + }, + { + "epoch": 0.18803418803418803, + "grad_norm": 0.41217970848083496, + "learning_rate": 0.0001989503910096274, + "loss": 1.0238, + "step": 1056 + }, + { + "epoch": 0.1882122507122507, + "grad_norm": 0.44038307666778564, + "learning_rate": 0.00019894836732020735, + "loss": 0.8159, + "step": 1057 + }, + { + "epoch": 0.1883903133903134, + "grad_norm": 0.45780670642852783, + "learning_rate": 0.0001989463416920996, + "loss": 1.2864, + "step": 1058 + }, + { + "epoch": 0.18856837606837606, + "grad_norm": 0.5197559595108032, + "learning_rate": 0.00019894431412534384, + "loss": 1.0756, + "step": 1059 + }, + { + "epoch": 0.18874643874643873, + "grad_norm": 0.43283385038375854, + "learning_rate": 0.00019894228461997979, + "loss": 1.0642, + "step": 1060 + }, + { + "epoch": 0.18892450142450143, + "grad_norm": 0.4657376706600189, + "learning_rate": 0.00019894025317604717, + "loss": 1.1159, + "step": 1061 + }, + { + "epoch": 0.1891025641025641, + "grad_norm": 0.4474908113479614, + "learning_rate": 0.00019893821979358588, + "loss": 1.2006, + "step": 1062 + }, + { + "epoch": 0.1892806267806268, + "grad_norm": 0.43878164887428284, + "learning_rate": 0.00019893618447263566, + "loss": 1.1599, + "step": 1063 + }, + { + "epoch": 0.18945868945868946, + "grad_norm": 0.4598735272884369, + "learning_rate": 0.00019893414721323645, + "loss": 1.3346, + "step": 1064 + }, + { + "epoch": 0.18963675213675213, + "grad_norm": 0.3947420120239258, + "learning_rate": 0.00019893210801542812, + "loss": 1.1201, + "step": 1065 + }, + { + "epoch": 0.18981481481481483, + "grad_norm": 0.3401558995246887, + "learning_rate": 0.00019893006687925064, + "loss": 0.7568, + "step": 1066 + }, + { + "epoch": 0.1899928774928775, + "grad_norm": 0.4400341808795929, + "learning_rate": 0.00019892802380474405, + "loss": 1.1706, + "step": 1067 + }, + { + "epoch": 0.19017094017094016, + "grad_norm": 0.42394164204597473, + "learning_rate": 0.00019892597879194829, + "loss": 1.0163, + "step": 1068 + }, + { + "epoch": 0.19034900284900286, + "grad_norm": 0.42904096841812134, + "learning_rate": 0.00019892393184090353, + "loss": 0.9193, + "step": 1069 + }, + { + "epoch": 0.19052706552706553, + "grad_norm": 0.497601181268692, + "learning_rate": 0.00019892188295164977, + "loss": 1.0377, + "step": 1070 + }, + { + "epoch": 0.1907051282051282, + "grad_norm": 0.4536020755767822, + "learning_rate": 0.00019891983212422723, + "loss": 1.0946, + "step": 1071 + }, + { + "epoch": 0.1908831908831909, + "grad_norm": 0.44916942715644836, + "learning_rate": 0.00019891777935867607, + "loss": 1.0563, + "step": 1072 + }, + { + "epoch": 0.19106125356125356, + "grad_norm": 0.4256889820098877, + "learning_rate": 0.0001989157246550365, + "loss": 1.0988, + "step": 1073 + }, + { + "epoch": 0.19123931623931623, + "grad_norm": 0.5559163689613342, + "learning_rate": 0.0001989136680133488, + "loss": 0.9155, + "step": 1074 + }, + { + "epoch": 0.19141737891737892, + "grad_norm": 0.391804963350296, + "learning_rate": 0.00019891160943365322, + "loss": 0.9314, + "step": 1075 + }, + { + "epoch": 0.1915954415954416, + "grad_norm": 0.4535716474056244, + "learning_rate": 0.00019890954891599015, + "loss": 1.0768, + "step": 1076 + }, + { + "epoch": 0.19177350427350429, + "grad_norm": 0.46770521998405457, + "learning_rate": 0.00019890748646039991, + "loss": 0.8406, + "step": 1077 + }, + { + "epoch": 0.19195156695156695, + "grad_norm": 0.4875394403934479, + "learning_rate": 0.00019890542206692295, + "loss": 1.1055, + "step": 1078 + }, + { + "epoch": 0.19212962962962962, + "grad_norm": 0.5072727203369141, + "learning_rate": 0.0001989033557355997, + "loss": 1.3093, + "step": 1079 + }, + { + "epoch": 0.19230769230769232, + "grad_norm": 0.4419287443161011, + "learning_rate": 0.00019890128746647068, + "loss": 1.1916, + "step": 1080 + }, + { + "epoch": 0.192485754985755, + "grad_norm": 0.45803651213645935, + "learning_rate": 0.00019889921725957637, + "loss": 1.2579, + "step": 1081 + }, + { + "epoch": 0.19266381766381765, + "grad_norm": 0.4832262098789215, + "learning_rate": 0.0001988971451149573, + "loss": 1.3217, + "step": 1082 + }, + { + "epoch": 0.19284188034188035, + "grad_norm": 0.4819786250591278, + "learning_rate": 0.00019889507103265416, + "loss": 1.0979, + "step": 1083 + }, + { + "epoch": 0.19301994301994302, + "grad_norm": 0.49360713362693787, + "learning_rate": 0.0001988929950127075, + "loss": 1.0987, + "step": 1084 + }, + { + "epoch": 0.1931980056980057, + "grad_norm": 0.44209200143814087, + "learning_rate": 0.00019889091705515806, + "loss": 1.2616, + "step": 1085 + }, + { + "epoch": 0.19337606837606838, + "grad_norm": 0.41626206040382385, + "learning_rate": 0.00019888883716004654, + "loss": 1.0922, + "step": 1086 + }, + { + "epoch": 0.19355413105413105, + "grad_norm": 0.4916635751724243, + "learning_rate": 0.00019888675532741366, + "loss": 0.9331, + "step": 1087 + }, + { + "epoch": 0.19373219373219372, + "grad_norm": 0.4493125379085541, + "learning_rate": 0.00019888467155730025, + "loss": 1.1261, + "step": 1088 + }, + { + "epoch": 0.19391025641025642, + "grad_norm": 0.3755671977996826, + "learning_rate": 0.00019888258584974708, + "loss": 0.9821, + "step": 1089 + }, + { + "epoch": 0.19408831908831908, + "grad_norm": 0.41917556524276733, + "learning_rate": 0.00019888049820479507, + "loss": 1.251, + "step": 1090 + }, + { + "epoch": 0.19426638176638178, + "grad_norm": 0.46184420585632324, + "learning_rate": 0.0001988784086224851, + "loss": 1.1731, + "step": 1091 + }, + { + "epoch": 0.19444444444444445, + "grad_norm": 0.4783691465854645, + "learning_rate": 0.00019887631710285812, + "loss": 1.1635, + "step": 1092 + }, + { + "epoch": 0.19462250712250712, + "grad_norm": 0.4710482060909271, + "learning_rate": 0.00019887422364595512, + "loss": 1.0229, + "step": 1093 + }, + { + "epoch": 0.1948005698005698, + "grad_norm": 0.4738706648349762, + "learning_rate": 0.00019887212825181707, + "loss": 1.128, + "step": 1094 + }, + { + "epoch": 0.19497863247863248, + "grad_norm": 0.45665010809898376, + "learning_rate": 0.00019887003092048508, + "loss": 1.0425, + "step": 1095 + }, + { + "epoch": 0.19515669515669515, + "grad_norm": 0.42740485072135925, + "learning_rate": 0.0001988679316520002, + "loss": 1.0738, + "step": 1096 + }, + { + "epoch": 0.19533475783475784, + "grad_norm": 0.5977092385292053, + "learning_rate": 0.0001988658304464036, + "loss": 1.2687, + "step": 1097 + }, + { + "epoch": 0.1955128205128205, + "grad_norm": 0.4411074221134186, + "learning_rate": 0.0001988637273037364, + "loss": 1.287, + "step": 1098 + }, + { + "epoch": 0.19569088319088318, + "grad_norm": 0.4409518539905548, + "learning_rate": 0.00019886162222403986, + "loss": 1.0515, + "step": 1099 + }, + { + "epoch": 0.19586894586894588, + "grad_norm": 0.4926736652851105, + "learning_rate": 0.0001988595152073552, + "loss": 1.1388, + "step": 1100 + }, + { + "epoch": 0.19604700854700854, + "grad_norm": 0.4607115387916565, + "learning_rate": 0.00019885740625372368, + "loss": 0.9803, + "step": 1101 + }, + { + "epoch": 0.1962250712250712, + "grad_norm": 0.4725342094898224, + "learning_rate": 0.0001988552953631867, + "loss": 1.199, + "step": 1102 + }, + { + "epoch": 0.1964031339031339, + "grad_norm": 0.48014503717422485, + "learning_rate": 0.00019885318253578548, + "loss": 1.1868, + "step": 1103 + }, + { + "epoch": 0.19658119658119658, + "grad_norm": 0.3872644603252411, + "learning_rate": 0.00019885106777156155, + "loss": 0.9182, + "step": 1104 + }, + { + "epoch": 0.19675925925925927, + "grad_norm": 0.4737720787525177, + "learning_rate": 0.00019884895107055627, + "loss": 1.1513, + "step": 1105 + }, + { + "epoch": 0.19693732193732194, + "grad_norm": 0.4144562780857086, + "learning_rate": 0.00019884683243281116, + "loss": 1.1711, + "step": 1106 + }, + { + "epoch": 0.1971153846153846, + "grad_norm": 0.4672079384326935, + "learning_rate": 0.00019884471185836769, + "loss": 1.0386, + "step": 1107 + }, + { + "epoch": 0.1972934472934473, + "grad_norm": 0.4558824598789215, + "learning_rate": 0.0001988425893472674, + "loss": 1.0535, + "step": 1108 + }, + { + "epoch": 0.19747150997150997, + "grad_norm": 0.5149834752082825, + "learning_rate": 0.00019884046489955192, + "loss": 1.0296, + "step": 1109 + }, + { + "epoch": 0.19764957264957264, + "grad_norm": 0.43444496393203735, + "learning_rate": 0.00019883833851526287, + "loss": 1.1475, + "step": 1110 + }, + { + "epoch": 0.19782763532763534, + "grad_norm": 0.46062374114990234, + "learning_rate": 0.00019883621019444188, + "loss": 1.183, + "step": 1111 + }, + { + "epoch": 0.198005698005698, + "grad_norm": 0.4893282949924469, + "learning_rate": 0.00019883407993713065, + "loss": 1.3733, + "step": 1112 + }, + { + "epoch": 0.19818376068376067, + "grad_norm": 0.5434843897819519, + "learning_rate": 0.00019883194774337096, + "loss": 1.2505, + "step": 1113 + }, + { + "epoch": 0.19836182336182337, + "grad_norm": 0.4698035418987274, + "learning_rate": 0.00019882981361320456, + "loss": 1.0152, + "step": 1114 + }, + { + "epoch": 0.19853988603988604, + "grad_norm": 0.4582163989543915, + "learning_rate": 0.00019882767754667325, + "loss": 1.1718, + "step": 1115 + }, + { + "epoch": 0.1987179487179487, + "grad_norm": 0.48744696378707886, + "learning_rate": 0.0001988255395438189, + "loss": 1.2923, + "step": 1116 + }, + { + "epoch": 0.1988960113960114, + "grad_norm": 0.4172030985355377, + "learning_rate": 0.0001988233996046834, + "loss": 0.8098, + "step": 1117 + }, + { + "epoch": 0.19907407407407407, + "grad_norm": 0.4556557834148407, + "learning_rate": 0.00019882125772930867, + "loss": 0.9654, + "step": 1118 + }, + { + "epoch": 0.19925213675213677, + "grad_norm": 0.4363219141960144, + "learning_rate": 0.00019881911391773666, + "loss": 1.0333, + "step": 1119 + }, + { + "epoch": 0.19943019943019943, + "grad_norm": 0.4336536228656769, + "learning_rate": 0.0001988169681700094, + "loss": 1.091, + "step": 1120 + }, + { + "epoch": 0.1996082621082621, + "grad_norm": 0.42073166370391846, + "learning_rate": 0.00019881482048616893, + "loss": 0.9687, + "step": 1121 + }, + { + "epoch": 0.1997863247863248, + "grad_norm": 0.4330587685108185, + "learning_rate": 0.00019881267086625733, + "loss": 1.0512, + "step": 1122 + }, + { + "epoch": 0.19996438746438747, + "grad_norm": 0.4602276682853699, + "learning_rate": 0.0001988105193103167, + "loss": 1.1806, + "step": 1123 + }, + { + "epoch": 0.20014245014245013, + "grad_norm": 0.4271257817745209, + "learning_rate": 0.0001988083658183892, + "loss": 1.1079, + "step": 1124 + }, + { + "epoch": 0.20032051282051283, + "grad_norm": 0.35446426272392273, + "learning_rate": 0.00019880621039051707, + "loss": 0.6769, + "step": 1125 + }, + { + "epoch": 0.2004985754985755, + "grad_norm": 0.413753479719162, + "learning_rate": 0.00019880405302674244, + "loss": 1.1088, + "step": 1126 + }, + { + "epoch": 0.20067663817663817, + "grad_norm": 0.4423675835132599, + "learning_rate": 0.00019880189372710767, + "loss": 1.1371, + "step": 1127 + }, + { + "epoch": 0.20085470085470086, + "grad_norm": 0.41865605115890503, + "learning_rate": 0.00019879973249165502, + "loss": 1.0027, + "step": 1128 + }, + { + "epoch": 0.20103276353276353, + "grad_norm": 0.4109594225883484, + "learning_rate": 0.00019879756932042686, + "loss": 0.8734, + "step": 1129 + }, + { + "epoch": 0.2012108262108262, + "grad_norm": 0.42326363921165466, + "learning_rate": 0.00019879540421346555, + "loss": 0.9722, + "step": 1130 + }, + { + "epoch": 0.2013888888888889, + "grad_norm": 0.4601542055606842, + "learning_rate": 0.00019879323717081354, + "loss": 1.1251, + "step": 1131 + }, + { + "epoch": 0.20156695156695156, + "grad_norm": 0.4704367518424988, + "learning_rate": 0.00019879106819251327, + "loss": 0.9457, + "step": 1132 + }, + { + "epoch": 0.20174501424501423, + "grad_norm": 0.465023934841156, + "learning_rate": 0.00019878889727860724, + "loss": 0.9633, + "step": 1133 + }, + { + "epoch": 0.20192307692307693, + "grad_norm": 0.4572450518608093, + "learning_rate": 0.00019878672442913796, + "loss": 1.1965, + "step": 1134 + }, + { + "epoch": 0.2021011396011396, + "grad_norm": 0.4323410391807556, + "learning_rate": 0.00019878454964414807, + "loss": 1.1296, + "step": 1135 + }, + { + "epoch": 0.2022792022792023, + "grad_norm": 0.4513751268386841, + "learning_rate": 0.00019878237292368013, + "loss": 1.0571, + "step": 1136 + }, + { + "epoch": 0.20245726495726496, + "grad_norm": 0.45504096150398254, + "learning_rate": 0.00019878019426777677, + "loss": 1.0316, + "step": 1137 + }, + { + "epoch": 0.20263532763532763, + "grad_norm": 0.45715275406837463, + "learning_rate": 0.0001987780136764807, + "loss": 1.0528, + "step": 1138 + }, + { + "epoch": 0.20281339031339032, + "grad_norm": 0.4934465289115906, + "learning_rate": 0.00019877583114983466, + "loss": 1.3238, + "step": 1139 + }, + { + "epoch": 0.202991452991453, + "grad_norm": 0.4304082989692688, + "learning_rate": 0.0001987736466878814, + "loss": 1.1774, + "step": 1140 + }, + { + "epoch": 0.20316951566951566, + "grad_norm": 0.49721968173980713, + "learning_rate": 0.00019877146029066372, + "loss": 1.1767, + "step": 1141 + }, + { + "epoch": 0.20334757834757836, + "grad_norm": 0.3629468083381653, + "learning_rate": 0.00019876927195822445, + "loss": 0.8588, + "step": 1142 + }, + { + "epoch": 0.20352564102564102, + "grad_norm": 0.49310383200645447, + "learning_rate": 0.00019876708169060648, + "loss": 1.0588, + "step": 1143 + }, + { + "epoch": 0.2037037037037037, + "grad_norm": 0.4270328879356384, + "learning_rate": 0.00019876488948785271, + "loss": 1.1523, + "step": 1144 + }, + { + "epoch": 0.2038817663817664, + "grad_norm": 0.4559730887413025, + "learning_rate": 0.0001987626953500061, + "loss": 1.1736, + "step": 1145 + }, + { + "epoch": 0.20405982905982906, + "grad_norm": 0.5335259437561035, + "learning_rate": 0.00019876049927710962, + "loss": 0.991, + "step": 1146 + }, + { + "epoch": 0.20423789173789172, + "grad_norm": 0.43500083684921265, + "learning_rate": 0.0001987583012692063, + "loss": 1.0631, + "step": 1147 + }, + { + "epoch": 0.20441595441595442, + "grad_norm": 0.4135417938232422, + "learning_rate": 0.00019875610132633927, + "loss": 1.0896, + "step": 1148 + }, + { + "epoch": 0.2045940170940171, + "grad_norm": 0.4078896641731262, + "learning_rate": 0.00019875389944855153, + "loss": 1.0395, + "step": 1149 + }, + { + "epoch": 0.20477207977207978, + "grad_norm": 0.46612194180488586, + "learning_rate": 0.00019875169563588632, + "loss": 1.0541, + "step": 1150 + }, + { + "epoch": 0.20495014245014245, + "grad_norm": 0.5093224048614502, + "learning_rate": 0.00019874948988838674, + "loss": 1.1486, + "step": 1151 + }, + { + "epoch": 0.20512820512820512, + "grad_norm": 0.5079755187034607, + "learning_rate": 0.00019874728220609607, + "loss": 1.2614, + "step": 1152 + }, + { + "epoch": 0.20530626780626782, + "grad_norm": 0.43663498759269714, + "learning_rate": 0.0001987450725890575, + "loss": 1.0683, + "step": 1153 + }, + { + "epoch": 0.20548433048433049, + "grad_norm": 0.5029327273368835, + "learning_rate": 0.00019874286103731435, + "loss": 1.1934, + "step": 1154 + }, + { + "epoch": 0.20566239316239315, + "grad_norm": 0.48770397901535034, + "learning_rate": 0.00019874064755090999, + "loss": 1.1634, + "step": 1155 + }, + { + "epoch": 0.20584045584045585, + "grad_norm": 0.46826690435409546, + "learning_rate": 0.00019873843212988776, + "loss": 1.0621, + "step": 1156 + }, + { + "epoch": 0.20601851851851852, + "grad_norm": 0.4810047149658203, + "learning_rate": 0.00019873621477429105, + "loss": 1.0879, + "step": 1157 + }, + { + "epoch": 0.20619658119658119, + "grad_norm": 0.4769522249698639, + "learning_rate": 0.00019873399548416335, + "loss": 1.1365, + "step": 1158 + }, + { + "epoch": 0.20637464387464388, + "grad_norm": 0.4221782982349396, + "learning_rate": 0.00019873177425954806, + "loss": 1.1168, + "step": 1159 + }, + { + "epoch": 0.20655270655270655, + "grad_norm": 0.4084923565387726, + "learning_rate": 0.00019872955110048876, + "loss": 1.2364, + "step": 1160 + }, + { + "epoch": 0.20673076923076922, + "grad_norm": 0.4781704545021057, + "learning_rate": 0.00019872732600702904, + "loss": 1.19, + "step": 1161 + }, + { + "epoch": 0.20690883190883191, + "grad_norm": 0.3984242081642151, + "learning_rate": 0.0001987250989792124, + "loss": 1.0568, + "step": 1162 + }, + { + "epoch": 0.20708689458689458, + "grad_norm": 0.4601972997188568, + "learning_rate": 0.00019872287001708257, + "loss": 1.1625, + "step": 1163 + }, + { + "epoch": 0.20726495726495728, + "grad_norm": 0.4853581190109253, + "learning_rate": 0.00019872063912068316, + "loss": 1.2304, + "step": 1164 + }, + { + "epoch": 0.20744301994301995, + "grad_norm": 0.41779839992523193, + "learning_rate": 0.0001987184062900579, + "loss": 0.9807, + "step": 1165 + }, + { + "epoch": 0.20762108262108261, + "grad_norm": 0.4945356249809265, + "learning_rate": 0.00019871617152525056, + "loss": 1.1861, + "step": 1166 + }, + { + "epoch": 0.2077991452991453, + "grad_norm": 0.47432294487953186, + "learning_rate": 0.00019871393482630487, + "loss": 1.1448, + "step": 1167 + }, + { + "epoch": 0.20797720797720798, + "grad_norm": 0.44647398591041565, + "learning_rate": 0.00019871169619326473, + "loss": 1.096, + "step": 1168 + }, + { + "epoch": 0.20815527065527065, + "grad_norm": 0.4643072783946991, + "learning_rate": 0.00019870945562617393, + "loss": 1.1561, + "step": 1169 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 0.4544340968132019, + "learning_rate": 0.0001987072131250764, + "loss": 1.0764, + "step": 1170 + }, + { + "epoch": 0.208511396011396, + "grad_norm": 0.6036561727523804, + "learning_rate": 0.00019870496869001607, + "loss": 1.3961, + "step": 1171 + }, + { + "epoch": 0.20868945868945868, + "grad_norm": 0.41348758339881897, + "learning_rate": 0.00019870272232103695, + "loss": 1.2219, + "step": 1172 + }, + { + "epoch": 0.20886752136752137, + "grad_norm": 0.4184056222438812, + "learning_rate": 0.000198700474018183, + "loss": 1.1115, + "step": 1173 + }, + { + "epoch": 0.20904558404558404, + "grad_norm": 0.41920599341392517, + "learning_rate": 0.0001986982237814983, + "loss": 0.9207, + "step": 1174 + }, + { + "epoch": 0.2092236467236467, + "grad_norm": 0.4710249602794647, + "learning_rate": 0.00019869597161102694, + "loss": 1.1342, + "step": 1175 + }, + { + "epoch": 0.2094017094017094, + "grad_norm": 0.46897777915000916, + "learning_rate": 0.000198693717506813, + "loss": 0.983, + "step": 1176 + }, + { + "epoch": 0.20957977207977208, + "grad_norm": 0.4817039370536804, + "learning_rate": 0.00019869146146890074, + "loss": 1.0923, + "step": 1177 + }, + { + "epoch": 0.20975783475783477, + "grad_norm": 0.4806751012802124, + "learning_rate": 0.00019868920349733427, + "loss": 1.2296, + "step": 1178 + }, + { + "epoch": 0.20993589743589744, + "grad_norm": 0.44182994961738586, + "learning_rate": 0.0001986869435921579, + "loss": 1.1856, + "step": 1179 + }, + { + "epoch": 0.2101139601139601, + "grad_norm": 0.4282805621623993, + "learning_rate": 0.00019868468175341584, + "loss": 1.0046, + "step": 1180 + }, + { + "epoch": 0.2102920227920228, + "grad_norm": 0.5011838674545288, + "learning_rate": 0.00019868241798115242, + "loss": 1.2401, + "step": 1181 + }, + { + "epoch": 0.21047008547008547, + "grad_norm": 0.4282447397708893, + "learning_rate": 0.00019868015227541208, + "loss": 0.9338, + "step": 1182 + }, + { + "epoch": 0.21064814814814814, + "grad_norm": 0.4348810911178589, + "learning_rate": 0.00019867788463623912, + "loss": 0.926, + "step": 1183 + }, + { + "epoch": 0.21082621082621084, + "grad_norm": 0.41518425941467285, + "learning_rate": 0.00019867561506367799, + "loss": 1.2723, + "step": 1184 + }, + { + "epoch": 0.2110042735042735, + "grad_norm": 0.47346001863479614, + "learning_rate": 0.00019867334355777315, + "loss": 1.1931, + "step": 1185 + }, + { + "epoch": 0.21118233618233617, + "grad_norm": 0.4071715474128723, + "learning_rate": 0.00019867107011856914, + "loss": 0.9619, + "step": 1186 + }, + { + "epoch": 0.21136039886039887, + "grad_norm": 0.4803447425365448, + "learning_rate": 0.00019866879474611046, + "loss": 1.2, + "step": 1187 + }, + { + "epoch": 0.21153846153846154, + "grad_norm": 0.4827699661254883, + "learning_rate": 0.00019866651744044172, + "loss": 1.0938, + "step": 1188 + }, + { + "epoch": 0.2117165242165242, + "grad_norm": 0.4528424143791199, + "learning_rate": 0.00019866423820160756, + "loss": 0.9721, + "step": 1189 + }, + { + "epoch": 0.2118945868945869, + "grad_norm": 0.43566834926605225, + "learning_rate": 0.0001986619570296526, + "loss": 1.0352, + "step": 1190 + }, + { + "epoch": 0.21207264957264957, + "grad_norm": 0.4516540467739105, + "learning_rate": 0.0001986596739246215, + "loss": 1.1333, + "step": 1191 + }, + { + "epoch": 0.21225071225071226, + "grad_norm": 0.4456641376018524, + "learning_rate": 0.00019865738888655908, + "loss": 1.2813, + "step": 1192 + }, + { + "epoch": 0.21242877492877493, + "grad_norm": 0.47048309445381165, + "learning_rate": 0.00019865510191551008, + "loss": 1.1067, + "step": 1193 + }, + { + "epoch": 0.2126068376068376, + "grad_norm": 0.4604061543941498, + "learning_rate": 0.00019865281301151928, + "loss": 0.925, + "step": 1194 + }, + { + "epoch": 0.2127849002849003, + "grad_norm": 0.49341437220573425, + "learning_rate": 0.00019865052217463153, + "loss": 1.2319, + "step": 1195 + }, + { + "epoch": 0.21296296296296297, + "grad_norm": 0.5099014639854431, + "learning_rate": 0.00019864822940489173, + "loss": 1.139, + "step": 1196 + }, + { + "epoch": 0.21314102564102563, + "grad_norm": 0.41396936774253845, + "learning_rate": 0.0001986459347023448, + "loss": 1.0594, + "step": 1197 + }, + { + "epoch": 0.21331908831908833, + "grad_norm": 0.46071869134902954, + "learning_rate": 0.0001986436380670357, + "loss": 1.0815, + "step": 1198 + }, + { + "epoch": 0.213497150997151, + "grad_norm": 0.507882297039032, + "learning_rate": 0.00019864133949900942, + "loss": 1.3841, + "step": 1199 + }, + { + "epoch": 0.21367521367521367, + "grad_norm": 0.45680439472198486, + "learning_rate": 0.00019863903899831103, + "loss": 1.0945, + "step": 1200 + }, + { + "epoch": 0.21385327635327636, + "grad_norm": 0.44277429580688477, + "learning_rate": 0.00019863673656498555, + "loss": 1.1655, + "step": 1201 + }, + { + "epoch": 0.21403133903133903, + "grad_norm": 0.43890756368637085, + "learning_rate": 0.00019863443219907812, + "loss": 1.1186, + "step": 1202 + }, + { + "epoch": 0.2142094017094017, + "grad_norm": 0.3910178542137146, + "learning_rate": 0.0001986321259006339, + "loss": 1.0817, + "step": 1203 + }, + { + "epoch": 0.2143874643874644, + "grad_norm": 0.3803878128528595, + "learning_rate": 0.00019862981766969803, + "loss": 0.8022, + "step": 1204 + }, + { + "epoch": 0.21456552706552706, + "grad_norm": 0.4495108425617218, + "learning_rate": 0.0001986275075063158, + "loss": 1.2212, + "step": 1205 + }, + { + "epoch": 0.21474358974358973, + "grad_norm": 0.5211976766586304, + "learning_rate": 0.00019862519541053244, + "loss": 1.2771, + "step": 1206 + }, + { + "epoch": 0.21492165242165243, + "grad_norm": 0.4313061535358429, + "learning_rate": 0.00019862288138239325, + "loss": 1.1205, + "step": 1207 + }, + { + "epoch": 0.2150997150997151, + "grad_norm": 0.47110888361930847, + "learning_rate": 0.00019862056542194355, + "loss": 1.1835, + "step": 1208 + }, + { + "epoch": 0.2152777777777778, + "grad_norm": 0.5129403471946716, + "learning_rate": 0.00019861824752922876, + "loss": 1.1655, + "step": 1209 + }, + { + "epoch": 0.21545584045584046, + "grad_norm": 0.4353938102722168, + "learning_rate": 0.00019861592770429427, + "loss": 1.2794, + "step": 1210 + }, + { + "epoch": 0.21563390313390313, + "grad_norm": 0.48590636253356934, + "learning_rate": 0.0001986136059471855, + "loss": 1.2003, + "step": 1211 + }, + { + "epoch": 0.21581196581196582, + "grad_norm": 0.4738406836986542, + "learning_rate": 0.00019861128225794804, + "loss": 1.2271, + "step": 1212 + }, + { + "epoch": 0.2159900284900285, + "grad_norm": 0.45983126759529114, + "learning_rate": 0.0001986089566366273, + "loss": 1.1896, + "step": 1213 + }, + { + "epoch": 0.21616809116809116, + "grad_norm": 0.37296006083488464, + "learning_rate": 0.00019860662908326892, + "loss": 1.079, + "step": 1214 + }, + { + "epoch": 0.21634615384615385, + "grad_norm": 0.4442676305770874, + "learning_rate": 0.00019860429959791845, + "loss": 1.1754, + "step": 1215 + }, + { + "epoch": 0.21652421652421652, + "grad_norm": 0.4950128495693207, + "learning_rate": 0.0001986019681806216, + "loss": 1.1571, + "step": 1216 + }, + { + "epoch": 0.2167022792022792, + "grad_norm": 0.4374556541442871, + "learning_rate": 0.000198599634831424, + "loss": 1.1003, + "step": 1217 + }, + { + "epoch": 0.2168803418803419, + "grad_norm": 0.47301414608955383, + "learning_rate": 0.00019859729955037136, + "loss": 1.1426, + "step": 1218 + }, + { + "epoch": 0.21705840455840456, + "grad_norm": 0.41213178634643555, + "learning_rate": 0.00019859496233750947, + "loss": 1.0659, + "step": 1219 + }, + { + "epoch": 0.21723646723646722, + "grad_norm": 0.41601964831352234, + "learning_rate": 0.0001985926231928841, + "loss": 1.0248, + "step": 1220 + }, + { + "epoch": 0.21741452991452992, + "grad_norm": 0.46328839659690857, + "learning_rate": 0.0001985902821165411, + "loss": 1.0405, + "step": 1221 + }, + { + "epoch": 0.2175925925925926, + "grad_norm": 0.43287959694862366, + "learning_rate": 0.0001985879391085263, + "loss": 0.9202, + "step": 1222 + }, + { + "epoch": 0.21777065527065528, + "grad_norm": 0.4770444631576538, + "learning_rate": 0.00019858559416888568, + "loss": 1.0911, + "step": 1223 + }, + { + "epoch": 0.21794871794871795, + "grad_norm": 0.4756585955619812, + "learning_rate": 0.00019858324729766507, + "loss": 1.1566, + "step": 1224 + }, + { + "epoch": 0.21812678062678062, + "grad_norm": 0.4337233006954193, + "learning_rate": 0.00019858089849491054, + "loss": 0.9084, + "step": 1225 + }, + { + "epoch": 0.21830484330484332, + "grad_norm": 0.5165579319000244, + "learning_rate": 0.00019857854776066813, + "loss": 1.4154, + "step": 1226 + }, + { + "epoch": 0.21848290598290598, + "grad_norm": 0.4280378520488739, + "learning_rate": 0.00019857619509498382, + "loss": 1.1291, + "step": 1227 + }, + { + "epoch": 0.21866096866096865, + "grad_norm": 0.5375089049339294, + "learning_rate": 0.00019857384049790376, + "loss": 1.2985, + "step": 1228 + }, + { + "epoch": 0.21883903133903135, + "grad_norm": 0.4708811640739441, + "learning_rate": 0.00019857148396947401, + "loss": 1.0589, + "step": 1229 + }, + { + "epoch": 0.21901709401709402, + "grad_norm": 0.4744570255279541, + "learning_rate": 0.00019856912550974084, + "loss": 1.1269, + "step": 1230 + }, + { + "epoch": 0.21919515669515668, + "grad_norm": 0.5355265736579895, + "learning_rate": 0.00019856676511875043, + "loss": 1.1441, + "step": 1231 + }, + { + "epoch": 0.21937321937321938, + "grad_norm": 0.42718183994293213, + "learning_rate": 0.00019856440279654897, + "loss": 1.0244, + "step": 1232 + }, + { + "epoch": 0.21955128205128205, + "grad_norm": 0.5162127614021301, + "learning_rate": 0.00019856203854318283, + "loss": 1.2674, + "step": 1233 + }, + { + "epoch": 0.21972934472934472, + "grad_norm": 0.5180695652961731, + "learning_rate": 0.00019855967235869827, + "loss": 1.2472, + "step": 1234 + }, + { + "epoch": 0.2199074074074074, + "grad_norm": 0.4290023744106293, + "learning_rate": 0.00019855730424314167, + "loss": 1.0502, + "step": 1235 + }, + { + "epoch": 0.22008547008547008, + "grad_norm": 0.4418254792690277, + "learning_rate": 0.00019855493419655945, + "loss": 1.0589, + "step": 1236 + }, + { + "epoch": 0.22026353276353278, + "grad_norm": 0.4074663817882538, + "learning_rate": 0.000198552562218998, + "loss": 0.9197, + "step": 1237 + }, + { + "epoch": 0.22044159544159544, + "grad_norm": 0.4526660740375519, + "learning_rate": 0.00019855018831050383, + "loss": 1.2578, + "step": 1238 + }, + { + "epoch": 0.2206196581196581, + "grad_norm": 0.4747827649116516, + "learning_rate": 0.00019854781247112343, + "loss": 1.0841, + "step": 1239 + }, + { + "epoch": 0.2207977207977208, + "grad_norm": 0.41567128896713257, + "learning_rate": 0.00019854543470090334, + "loss": 1.0737, + "step": 1240 + }, + { + "epoch": 0.22097578347578348, + "grad_norm": 0.4793100953102112, + "learning_rate": 0.00019854305499989022, + "loss": 1.1972, + "step": 1241 + }, + { + "epoch": 0.22115384615384615, + "grad_norm": 0.41755473613739014, + "learning_rate": 0.00019854067336813058, + "loss": 1.2529, + "step": 1242 + }, + { + "epoch": 0.22133190883190884, + "grad_norm": 0.40421152114868164, + "learning_rate": 0.0001985382898056712, + "loss": 1.0549, + "step": 1243 + }, + { + "epoch": 0.2215099715099715, + "grad_norm": 0.45779645442962646, + "learning_rate": 0.0001985359043125587, + "loss": 1.1586, + "step": 1244 + }, + { + "epoch": 0.22168803418803418, + "grad_norm": 0.4380546808242798, + "learning_rate": 0.00019853351688883987, + "loss": 1.1024, + "step": 1245 + }, + { + "epoch": 0.22186609686609687, + "grad_norm": 0.39917269349098206, + "learning_rate": 0.00019853112753456142, + "loss": 0.9823, + "step": 1246 + }, + { + "epoch": 0.22204415954415954, + "grad_norm": 0.4228038489818573, + "learning_rate": 0.00019852873624977022, + "loss": 1.1684, + "step": 1247 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 0.4462146759033203, + "learning_rate": 0.00019852634303451315, + "loss": 0.9027, + "step": 1248 + }, + { + "epoch": 0.2224002849002849, + "grad_norm": 0.5682163834571838, + "learning_rate": 0.000198523947888837, + "loss": 1.141, + "step": 1249 + }, + { + "epoch": 0.22257834757834757, + "grad_norm": 0.44866830110549927, + "learning_rate": 0.0001985215508127888, + "loss": 1.0759, + "step": 1250 + }, + { + "epoch": 0.22275641025641027, + "grad_norm": 0.4034106135368347, + "learning_rate": 0.00019851915180641548, + "loss": 1.0675, + "step": 1251 + }, + { + "epoch": 0.22293447293447294, + "grad_norm": 0.4780726432800293, + "learning_rate": 0.00019851675086976397, + "loss": 1.0283, + "step": 1252 + }, + { + "epoch": 0.2231125356125356, + "grad_norm": 0.48892372846603394, + "learning_rate": 0.00019851434800288145, + "loss": 1.1159, + "step": 1253 + }, + { + "epoch": 0.2232905982905983, + "grad_norm": 0.42629215121269226, + "learning_rate": 0.0001985119432058149, + "loss": 1.0292, + "step": 1254 + }, + { + "epoch": 0.22346866096866097, + "grad_norm": 0.4496444761753082, + "learning_rate": 0.00019850953647861146, + "loss": 1.0252, + "step": 1255 + }, + { + "epoch": 0.22364672364672364, + "grad_norm": 0.4371408224105835, + "learning_rate": 0.00019850712782131828, + "loss": 1.1104, + "step": 1256 + }, + { + "epoch": 0.22382478632478633, + "grad_norm": 0.4910794496536255, + "learning_rate": 0.00019850471723398258, + "loss": 1.1928, + "step": 1257 + }, + { + "epoch": 0.224002849002849, + "grad_norm": 0.41235068440437317, + "learning_rate": 0.00019850230471665157, + "loss": 1.1261, + "step": 1258 + }, + { + "epoch": 0.22418091168091167, + "grad_norm": 0.4507700502872467, + "learning_rate": 0.0001984998902693725, + "loss": 1.0602, + "step": 1259 + }, + { + "epoch": 0.22435897435897437, + "grad_norm": 0.4654198884963989, + "learning_rate": 0.00019849747389219272, + "loss": 1.1258, + "step": 1260 + }, + { + "epoch": 0.22453703703703703, + "grad_norm": 0.439807653427124, + "learning_rate": 0.00019849505558515952, + "loss": 1.2312, + "step": 1261 + }, + { + "epoch": 0.2247150997150997, + "grad_norm": 0.4309258759021759, + "learning_rate": 0.00019849263534832035, + "loss": 1.0083, + "step": 1262 + }, + { + "epoch": 0.2248931623931624, + "grad_norm": 0.4920141100883484, + "learning_rate": 0.00019849021318172255, + "loss": 1.0254, + "step": 1263 + }, + { + "epoch": 0.22507122507122507, + "grad_norm": 0.5333457589149475, + "learning_rate": 0.00019848778908541367, + "loss": 1.3017, + "step": 1264 + }, + { + "epoch": 0.22524928774928774, + "grad_norm": 0.4096757769584656, + "learning_rate": 0.0001984853630594411, + "loss": 0.9531, + "step": 1265 + }, + { + "epoch": 0.22542735042735043, + "grad_norm": 0.5744075775146484, + "learning_rate": 0.00019848293510385244, + "loss": 1.1414, + "step": 1266 + }, + { + "epoch": 0.2256054131054131, + "grad_norm": 0.44707193970680237, + "learning_rate": 0.00019848050521869529, + "loss": 1.1926, + "step": 1267 + }, + { + "epoch": 0.2257834757834758, + "grad_norm": 0.4162999391555786, + "learning_rate": 0.00019847807340401716, + "loss": 1.1354, + "step": 1268 + }, + { + "epoch": 0.22596153846153846, + "grad_norm": 0.4273204207420349, + "learning_rate": 0.0001984756396598658, + "loss": 0.9956, + "step": 1269 + }, + { + "epoch": 0.22613960113960113, + "grad_norm": 0.5670466423034668, + "learning_rate": 0.00019847320398628878, + "loss": 1.2384, + "step": 1270 + }, + { + "epoch": 0.22631766381766383, + "grad_norm": 0.424544095993042, + "learning_rate": 0.00019847076638333395, + "loss": 0.9963, + "step": 1271 + }, + { + "epoch": 0.2264957264957265, + "grad_norm": 0.3716120719909668, + "learning_rate": 0.000198468326851049, + "loss": 0.865, + "step": 1272 + }, + { + "epoch": 0.22667378917378916, + "grad_norm": 0.4472847282886505, + "learning_rate": 0.00019846588538948172, + "loss": 1.174, + "step": 1273 + }, + { + "epoch": 0.22685185185185186, + "grad_norm": 0.4599195718765259, + "learning_rate": 0.00019846344199867994, + "loss": 1.289, + "step": 1274 + }, + { + "epoch": 0.22702991452991453, + "grad_norm": 0.4303213357925415, + "learning_rate": 0.0001984609966786916, + "loss": 1.1606, + "step": 1275 + }, + { + "epoch": 0.2272079772079772, + "grad_norm": 0.44893527030944824, + "learning_rate": 0.00019845854942956455, + "loss": 1.1043, + "step": 1276 + }, + { + "epoch": 0.2273860398860399, + "grad_norm": 0.40033379197120667, + "learning_rate": 0.00019845610025134676, + "loss": 1.1434, + "step": 1277 + }, + { + "epoch": 0.22756410256410256, + "grad_norm": 0.4385402202606201, + "learning_rate": 0.00019845364914408616, + "loss": 0.9943, + "step": 1278 + }, + { + "epoch": 0.22774216524216523, + "grad_norm": 0.42123618721961975, + "learning_rate": 0.0001984511961078309, + "loss": 1.0911, + "step": 1279 + }, + { + "epoch": 0.22792022792022792, + "grad_norm": 0.5558577179908752, + "learning_rate": 0.00019844874114262893, + "loss": 1.3893, + "step": 1280 + }, + { + "epoch": 0.2280982905982906, + "grad_norm": 0.3996453583240509, + "learning_rate": 0.00019844628424852835, + "loss": 0.8951, + "step": 1281 + }, + { + "epoch": 0.2282763532763533, + "grad_norm": 0.3943425714969635, + "learning_rate": 0.0001984438254255774, + "loss": 1.0595, + "step": 1282 + }, + { + "epoch": 0.22845441595441596, + "grad_norm": 0.4429021179676056, + "learning_rate": 0.00019844136467382414, + "loss": 1.0853, + "step": 1283 + }, + { + "epoch": 0.22863247863247863, + "grad_norm": 0.4515686631202698, + "learning_rate": 0.00019843890199331687, + "loss": 1.0829, + "step": 1284 + }, + { + "epoch": 0.22881054131054132, + "grad_norm": 0.5157768726348877, + "learning_rate": 0.00019843643738410378, + "loss": 1.334, + "step": 1285 + }, + { + "epoch": 0.228988603988604, + "grad_norm": 0.45833173394203186, + "learning_rate": 0.0001984339708462332, + "loss": 1.1353, + "step": 1286 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 0.46610337495803833, + "learning_rate": 0.00019843150237975344, + "loss": 1.1338, + "step": 1287 + }, + { + "epoch": 0.22934472934472935, + "grad_norm": 0.5076978802680969, + "learning_rate": 0.00019842903198471286, + "loss": 1.1811, + "step": 1288 + }, + { + "epoch": 0.22952279202279202, + "grad_norm": 0.4297824800014496, + "learning_rate": 0.00019842655966115986, + "loss": 1.1799, + "step": 1289 + }, + { + "epoch": 0.2297008547008547, + "grad_norm": 0.5304586291313171, + "learning_rate": 0.0001984240854091429, + "loss": 1.1315, + "step": 1290 + }, + { + "epoch": 0.22987891737891739, + "grad_norm": 0.45359212160110474, + "learning_rate": 0.00019842160922871042, + "loss": 1.1037, + "step": 1291 + }, + { + "epoch": 0.23005698005698005, + "grad_norm": 0.4416881203651428, + "learning_rate": 0.00019841913111991096, + "loss": 1.122, + "step": 1292 + }, + { + "epoch": 0.23023504273504272, + "grad_norm": 0.46682995557785034, + "learning_rate": 0.0001984166510827931, + "loss": 0.9808, + "step": 1293 + }, + { + "epoch": 0.23041310541310542, + "grad_norm": 0.44172337651252747, + "learning_rate": 0.00019841416911740538, + "loss": 0.9167, + "step": 1294 + }, + { + "epoch": 0.23059116809116809, + "grad_norm": 0.40562742948532104, + "learning_rate": 0.0001984116852237965, + "loss": 0.9547, + "step": 1295 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.4040384888648987, + "learning_rate": 0.00019840919940201503, + "loss": 1.1039, + "step": 1296 + }, + { + "epoch": 0.23094729344729345, + "grad_norm": 0.5094077587127686, + "learning_rate": 0.00019840671165210973, + "loss": 1.2283, + "step": 1297 + }, + { + "epoch": 0.23112535612535612, + "grad_norm": 0.48553213477134705, + "learning_rate": 0.00019840422197412938, + "loss": 1.0927, + "step": 1298 + }, + { + "epoch": 0.23130341880341881, + "grad_norm": 0.5197509527206421, + "learning_rate": 0.00019840173036812266, + "loss": 1.2154, + "step": 1299 + }, + { + "epoch": 0.23148148148148148, + "grad_norm": 0.42069005966186523, + "learning_rate": 0.0001983992368341385, + "loss": 1.0076, + "step": 1300 + }, + { + "epoch": 0.23165954415954415, + "grad_norm": 0.475204735994339, + "learning_rate": 0.00019839674137222567, + "loss": 1.1682, + "step": 1301 + }, + { + "epoch": 0.23183760683760685, + "grad_norm": 0.55730140209198, + "learning_rate": 0.0001983942439824331, + "loss": 1.2948, + "step": 1302 + }, + { + "epoch": 0.23201566951566951, + "grad_norm": 0.4533313512802124, + "learning_rate": 0.00019839174466480973, + "loss": 1.2691, + "step": 1303 + }, + { + "epoch": 0.23219373219373218, + "grad_norm": 0.4733520746231079, + "learning_rate": 0.0001983892434194045, + "loss": 1.2232, + "step": 1304 + }, + { + "epoch": 0.23237179487179488, + "grad_norm": 0.5085756182670593, + "learning_rate": 0.00019838674024626643, + "loss": 1.1347, + "step": 1305 + }, + { + "epoch": 0.23254985754985755, + "grad_norm": 0.4679976999759674, + "learning_rate": 0.00019838423514544456, + "loss": 1.0018, + "step": 1306 + }, + { + "epoch": 0.23272792022792022, + "grad_norm": 0.4234481751918793, + "learning_rate": 0.00019838172811698795, + "loss": 1.0472, + "step": 1307 + }, + { + "epoch": 0.2329059829059829, + "grad_norm": 0.5749204158782959, + "learning_rate": 0.00019837921916094579, + "loss": 1.2239, + "step": 1308 + }, + { + "epoch": 0.23308404558404558, + "grad_norm": 0.46715882420539856, + "learning_rate": 0.0001983767082773672, + "loss": 1.1924, + "step": 1309 + }, + { + "epoch": 0.23326210826210828, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.00019837419546630137, + "loss": 1.1086, + "step": 1310 + }, + { + "epoch": 0.23344017094017094, + "grad_norm": 0.4419243037700653, + "learning_rate": 0.0001983716807277975, + "loss": 1.1911, + "step": 1311 + }, + { + "epoch": 0.2336182336182336, + "grad_norm": 0.5107570290565491, + "learning_rate": 0.00019836916406190493, + "loss": 1.1071, + "step": 1312 + }, + { + "epoch": 0.2337962962962963, + "grad_norm": 0.5295659303665161, + "learning_rate": 0.00019836664546867293, + "loss": 1.2905, + "step": 1313 + }, + { + "epoch": 0.23397435897435898, + "grad_norm": 0.4844837784767151, + "learning_rate": 0.00019836412494815084, + "loss": 1.3507, + "step": 1314 + }, + { + "epoch": 0.23415242165242164, + "grad_norm": 0.6166049242019653, + "learning_rate": 0.00019836160250038808, + "loss": 1.2822, + "step": 1315 + }, + { + "epoch": 0.23433048433048434, + "grad_norm": 0.3229198753833771, + "learning_rate": 0.00019835907812543402, + "loss": 0.4959, + "step": 1316 + }, + { + "epoch": 0.234508547008547, + "grad_norm": 0.5788772702217102, + "learning_rate": 0.00019835655182333815, + "loss": 1.0832, + "step": 1317 + }, + { + "epoch": 0.23468660968660968, + "grad_norm": 0.525705099105835, + "learning_rate": 0.00019835402359414997, + "loss": 1.0968, + "step": 1318 + }, + { + "epoch": 0.23486467236467237, + "grad_norm": 0.5007779002189636, + "learning_rate": 0.000198351493437919, + "loss": 1.2788, + "step": 1319 + }, + { + "epoch": 0.23504273504273504, + "grad_norm": 0.4276871383190155, + "learning_rate": 0.00019834896135469484, + "loss": 1.0419, + "step": 1320 + }, + { + "epoch": 0.2352207977207977, + "grad_norm": 0.5359070301055908, + "learning_rate": 0.00019834642734452708, + "loss": 1.1308, + "step": 1321 + }, + { + "epoch": 0.2353988603988604, + "grad_norm": 0.4854908883571625, + "learning_rate": 0.0001983438914074654, + "loss": 1.1211, + "step": 1322 + }, + { + "epoch": 0.23557692307692307, + "grad_norm": 0.4913707375526428, + "learning_rate": 0.0001983413535435594, + "loss": 1.2392, + "step": 1323 + }, + { + "epoch": 0.23575498575498577, + "grad_norm": 0.46755748987197876, + "learning_rate": 0.0001983388137528589, + "loss": 0.9348, + "step": 1324 + }, + { + "epoch": 0.23593304843304844, + "grad_norm": 0.4592570960521698, + "learning_rate": 0.0001983362720354136, + "loss": 1.1339, + "step": 1325 + }, + { + "epoch": 0.2361111111111111, + "grad_norm": 0.5121711492538452, + "learning_rate": 0.00019833372839127335, + "loss": 1.2973, + "step": 1326 + }, + { + "epoch": 0.2362891737891738, + "grad_norm": 0.4809017479419708, + "learning_rate": 0.000198331182820488, + "loss": 0.9849, + "step": 1327 + }, + { + "epoch": 0.23646723646723647, + "grad_norm": 0.42340895533561707, + "learning_rate": 0.00019832863532310733, + "loss": 1.0731, + "step": 1328 + }, + { + "epoch": 0.23664529914529914, + "grad_norm": 0.5388045310974121, + "learning_rate": 0.00019832608589918135, + "loss": 1.0729, + "step": 1329 + }, + { + "epoch": 0.23682336182336183, + "grad_norm": 0.43075770139694214, + "learning_rate": 0.00019832353454875992, + "loss": 1.1684, + "step": 1330 + }, + { + "epoch": 0.2370014245014245, + "grad_norm": 0.554927408695221, + "learning_rate": 0.00019832098127189313, + "loss": 1.0842, + "step": 1331 + }, + { + "epoch": 0.23717948717948717, + "grad_norm": 0.5359260439872742, + "learning_rate": 0.0001983184260686309, + "loss": 1.2399, + "step": 1332 + }, + { + "epoch": 0.23735754985754987, + "grad_norm": 0.5141251087188721, + "learning_rate": 0.0001983158689390234, + "loss": 1.3752, + "step": 1333 + }, + { + "epoch": 0.23753561253561253, + "grad_norm": 0.4578750431537628, + "learning_rate": 0.00019831330988312067, + "loss": 1.0965, + "step": 1334 + }, + { + "epoch": 0.2377136752136752, + "grad_norm": 0.47974497079849243, + "learning_rate": 0.00019831074890097286, + "loss": 1.3379, + "step": 1335 + }, + { + "epoch": 0.2378917378917379, + "grad_norm": 0.4618176817893982, + "learning_rate": 0.00019830818599263014, + "loss": 1.274, + "step": 1336 + }, + { + "epoch": 0.23806980056980057, + "grad_norm": 0.4279816448688507, + "learning_rate": 0.00019830562115814276, + "loss": 0.996, + "step": 1337 + }, + { + "epoch": 0.23824786324786323, + "grad_norm": 0.4255026876926422, + "learning_rate": 0.0001983030543975609, + "loss": 0.969, + "step": 1338 + }, + { + "epoch": 0.23842592592592593, + "grad_norm": 0.4551412761211395, + "learning_rate": 0.00019830048571093493, + "loss": 1.0204, + "step": 1339 + }, + { + "epoch": 0.2386039886039886, + "grad_norm": 0.4747903048992157, + "learning_rate": 0.00019829791509831513, + "loss": 1.1816, + "step": 1340 + }, + { + "epoch": 0.2387820512820513, + "grad_norm": 0.47187140583992004, + "learning_rate": 0.00019829534255975188, + "loss": 1.1205, + "step": 1341 + }, + { + "epoch": 0.23896011396011396, + "grad_norm": 0.49332180619239807, + "learning_rate": 0.0001982927680952956, + "loss": 1.2657, + "step": 1342 + }, + { + "epoch": 0.23913817663817663, + "grad_norm": 0.5162837505340576, + "learning_rate": 0.0001982901917049967, + "loss": 1.2247, + "step": 1343 + }, + { + "epoch": 0.23931623931623933, + "grad_norm": 0.43407055735588074, + "learning_rate": 0.0001982876133889057, + "loss": 1.0038, + "step": 1344 + }, + { + "epoch": 0.239494301994302, + "grad_norm": 0.5132251977920532, + "learning_rate": 0.00019828503314707306, + "loss": 1.0678, + "step": 1345 + }, + { + "epoch": 0.23967236467236466, + "grad_norm": 0.46295464038848877, + "learning_rate": 0.00019828245097954937, + "loss": 1.1802, + "step": 1346 + }, + { + "epoch": 0.23985042735042736, + "grad_norm": 0.4682658314704895, + "learning_rate": 0.00019827986688638523, + "loss": 1.0249, + "step": 1347 + }, + { + "epoch": 0.24002849002849003, + "grad_norm": 0.49990561604499817, + "learning_rate": 0.00019827728086763125, + "loss": 1.0691, + "step": 1348 + }, + { + "epoch": 0.2402065527065527, + "grad_norm": 0.39090847969055176, + "learning_rate": 0.00019827469292333806, + "loss": 0.8367, + "step": 1349 + }, + { + "epoch": 0.2403846153846154, + "grad_norm": 0.5023905634880066, + "learning_rate": 0.00019827210305355645, + "loss": 1.0675, + "step": 1350 + }, + { + "epoch": 0.24056267806267806, + "grad_norm": 0.4744076430797577, + "learning_rate": 0.00019826951125833715, + "loss": 1.3166, + "step": 1351 + }, + { + "epoch": 0.24074074074074073, + "grad_norm": 0.44914689660072327, + "learning_rate": 0.00019826691753773088, + "loss": 0.9818, + "step": 1352 + }, + { + "epoch": 0.24091880341880342, + "grad_norm": 0.44391971826553345, + "learning_rate": 0.00019826432189178853, + "loss": 1.0448, + "step": 1353 + }, + { + "epoch": 0.2410968660968661, + "grad_norm": 0.46102839708328247, + "learning_rate": 0.00019826172432056086, + "loss": 0.9952, + "step": 1354 + }, + { + "epoch": 0.2412749287749288, + "grad_norm": 0.4796878695487976, + "learning_rate": 0.00019825912482409884, + "loss": 1.0977, + "step": 1355 + }, + { + "epoch": 0.24145299145299146, + "grad_norm": 0.5003768801689148, + "learning_rate": 0.0001982565234024534, + "loss": 1.3149, + "step": 1356 + }, + { + "epoch": 0.24163105413105412, + "grad_norm": 0.43475663661956787, + "learning_rate": 0.00019825392005567551, + "loss": 1.0527, + "step": 1357 + }, + { + "epoch": 0.24180911680911682, + "grad_norm": 0.46120527386665344, + "learning_rate": 0.00019825131478381613, + "loss": 1.2333, + "step": 1358 + }, + { + "epoch": 0.2419871794871795, + "grad_norm": 0.43748101592063904, + "learning_rate": 0.00019824870758692638, + "loss": 0.9788, + "step": 1359 + }, + { + "epoch": 0.24216524216524216, + "grad_norm": 0.5275192856788635, + "learning_rate": 0.00019824609846505727, + "loss": 1.1473, + "step": 1360 + }, + { + "epoch": 0.24234330484330485, + "grad_norm": 0.346463143825531, + "learning_rate": 0.00019824348741825993, + "loss": 0.6824, + "step": 1361 + }, + { + "epoch": 0.24252136752136752, + "grad_norm": 0.5004115700721741, + "learning_rate": 0.00019824087444658556, + "loss": 1.1853, + "step": 1362 + }, + { + "epoch": 0.2426994301994302, + "grad_norm": 0.42746666073799133, + "learning_rate": 0.00019823825955008533, + "loss": 0.9355, + "step": 1363 + }, + { + "epoch": 0.24287749287749288, + "grad_norm": 0.4099743068218231, + "learning_rate": 0.00019823564272881047, + "loss": 1.0753, + "step": 1364 + }, + { + "epoch": 0.24305555555555555, + "grad_norm": 0.5262967944145203, + "learning_rate": 0.00019823302398281226, + "loss": 1.2324, + "step": 1365 + }, + { + "epoch": 0.24323361823361822, + "grad_norm": 0.436069518327713, + "learning_rate": 0.000198230403312142, + "loss": 1.1887, + "step": 1366 + }, + { + "epoch": 0.24341168091168092, + "grad_norm": 0.38252368569374084, + "learning_rate": 0.00019822778071685107, + "loss": 1.0211, + "step": 1367 + }, + { + "epoch": 0.24358974358974358, + "grad_norm": 0.48024141788482666, + "learning_rate": 0.00019822515619699081, + "loss": 1.065, + "step": 1368 + }, + { + "epoch": 0.24376780626780628, + "grad_norm": 0.47421589493751526, + "learning_rate": 0.00019822252975261267, + "loss": 1.0433, + "step": 1369 + }, + { + "epoch": 0.24394586894586895, + "grad_norm": 0.46094807982444763, + "learning_rate": 0.00019821990138376808, + "loss": 1.1427, + "step": 1370 + }, + { + "epoch": 0.24412393162393162, + "grad_norm": 0.5093680620193481, + "learning_rate": 0.00019821727109050856, + "loss": 1.1086, + "step": 1371 + }, + { + "epoch": 0.2443019943019943, + "grad_norm": 0.41084879636764526, + "learning_rate": 0.00019821463887288566, + "loss": 1.0068, + "step": 1372 + }, + { + "epoch": 0.24448005698005698, + "grad_norm": 0.4991084635257721, + "learning_rate": 0.0001982120047309509, + "loss": 1.1884, + "step": 1373 + }, + { + "epoch": 0.24465811965811965, + "grad_norm": 0.39198383688926697, + "learning_rate": 0.00019820936866475595, + "loss": 0.9776, + "step": 1374 + }, + { + "epoch": 0.24483618233618235, + "grad_norm": 0.4517424702644348, + "learning_rate": 0.00019820673067435244, + "loss": 1.1491, + "step": 1375 + }, + { + "epoch": 0.245014245014245, + "grad_norm": 0.45881983637809753, + "learning_rate": 0.00019820409075979202, + "loss": 1.1198, + "step": 1376 + }, + { + "epoch": 0.24519230769230768, + "grad_norm": 0.4498792290687561, + "learning_rate": 0.00019820144892112646, + "loss": 1.0897, + "step": 1377 + }, + { + "epoch": 0.24537037037037038, + "grad_norm": 0.4128037393093109, + "learning_rate": 0.00019819880515840752, + "loss": 0.9415, + "step": 1378 + }, + { + "epoch": 0.24554843304843305, + "grad_norm": 0.4340885281562805, + "learning_rate": 0.00019819615947168698, + "loss": 1.201, + "step": 1379 + }, + { + "epoch": 0.24572649572649571, + "grad_norm": 0.43814027309417725, + "learning_rate": 0.00019819351186101667, + "loss": 1.1039, + "step": 1380 + }, + { + "epoch": 0.2459045584045584, + "grad_norm": 0.40115082263946533, + "learning_rate": 0.00019819086232644845, + "loss": 1.2599, + "step": 1381 + }, + { + "epoch": 0.24608262108262108, + "grad_norm": 0.4947351813316345, + "learning_rate": 0.00019818821086803426, + "loss": 1.252, + "step": 1382 + }, + { + "epoch": 0.24626068376068377, + "grad_norm": 0.45179441571235657, + "learning_rate": 0.0001981855574858261, + "loss": 1.1323, + "step": 1383 + }, + { + "epoch": 0.24643874643874644, + "grad_norm": 0.47159844636917114, + "learning_rate": 0.00019818290217987587, + "loss": 1.2053, + "step": 1384 + }, + { + "epoch": 0.2466168091168091, + "grad_norm": 0.4358448386192322, + "learning_rate": 0.0001981802449502356, + "loss": 1.1174, + "step": 1385 + }, + { + "epoch": 0.2467948717948718, + "grad_norm": 0.4588233530521393, + "learning_rate": 0.00019817758579695745, + "loss": 1.1098, + "step": 1386 + }, + { + "epoch": 0.24697293447293447, + "grad_norm": 0.4955112636089325, + "learning_rate": 0.00019817492472009338, + "loss": 1.258, + "step": 1387 + }, + { + "epoch": 0.24715099715099714, + "grad_norm": 0.4226941764354706, + "learning_rate": 0.00019817226171969565, + "loss": 1.0976, + "step": 1388 + }, + { + "epoch": 0.24732905982905984, + "grad_norm": 0.4076840579509735, + "learning_rate": 0.00019816959679581637, + "loss": 1.0121, + "step": 1389 + }, + { + "epoch": 0.2475071225071225, + "grad_norm": 0.4395063519477844, + "learning_rate": 0.0001981669299485078, + "loss": 1.3153, + "step": 1390 + }, + { + "epoch": 0.24768518518518517, + "grad_norm": 0.41010400652885437, + "learning_rate": 0.0001981642611778221, + "loss": 1.0717, + "step": 1391 + }, + { + "epoch": 0.24786324786324787, + "grad_norm": 0.43459352850914, + "learning_rate": 0.00019816159048381167, + "loss": 1.1077, + "step": 1392 + }, + { + "epoch": 0.24804131054131054, + "grad_norm": 0.46291449666023254, + "learning_rate": 0.00019815891786652875, + "loss": 1.0257, + "step": 1393 + }, + { + "epoch": 0.2482193732193732, + "grad_norm": 0.46408146619796753, + "learning_rate": 0.00019815624332602578, + "loss": 0.7899, + "step": 1394 + }, + { + "epoch": 0.2483974358974359, + "grad_norm": 0.4763357937335968, + "learning_rate": 0.00019815356686235508, + "loss": 0.9857, + "step": 1395 + }, + { + "epoch": 0.24857549857549857, + "grad_norm": 0.4766457676887512, + "learning_rate": 0.00019815088847556918, + "loss": 1.0589, + "step": 1396 + }, + { + "epoch": 0.24875356125356127, + "grad_norm": 0.4486583173274994, + "learning_rate": 0.0001981482081657205, + "loss": 1.2572, + "step": 1397 + }, + { + "epoch": 0.24893162393162394, + "grad_norm": 0.468878835439682, + "learning_rate": 0.00019814552593286155, + "loss": 1.101, + "step": 1398 + }, + { + "epoch": 0.2491096866096866, + "grad_norm": 0.4230278730392456, + "learning_rate": 0.0001981428417770449, + "loss": 0.9457, + "step": 1399 + }, + { + "epoch": 0.2492877492877493, + "grad_norm": 0.45630761981010437, + "learning_rate": 0.00019814015569832315, + "loss": 1.0665, + "step": 1400 + }, + { + "epoch": 0.24946581196581197, + "grad_norm": 0.5780113935470581, + "learning_rate": 0.00019813746769674893, + "loss": 1.1064, + "step": 1401 + }, + { + "epoch": 0.24964387464387464, + "grad_norm": 0.4343436658382416, + "learning_rate": 0.0001981347777723749, + "loss": 1.1132, + "step": 1402 + }, + { + "epoch": 0.24982193732193733, + "grad_norm": 0.4879056513309479, + "learning_rate": 0.0001981320859252537, + "loss": 1.1301, + "step": 1403 + }, + { + "epoch": 0.25, + "grad_norm": 0.5248328447341919, + "learning_rate": 0.00019812939215543818, + "loss": 1.1468, + "step": 1404 + }, + { + "epoch": 0.25, + "eval_loss": 1.115895390510559, + "eval_runtime": 25.0474, + "eval_samples_per_second": 41.561, + "eval_steps_per_second": 20.801, + "step": 1404 + }, + { + "epoch": 0.2501780626780627, + "grad_norm": 0.5076769590377808, + "learning_rate": 0.00019812669646298106, + "loss": 1.1428, + "step": 1405 + }, + { + "epoch": 0.25035612535612534, + "grad_norm": 0.5510252714157104, + "learning_rate": 0.00019812399884793514, + "loss": 1.3383, + "step": 1406 + }, + { + "epoch": 0.25053418803418803, + "grad_norm": 0.48918986320495605, + "learning_rate": 0.0001981212993103533, + "loss": 1.1507, + "step": 1407 + }, + { + "epoch": 0.25071225071225073, + "grad_norm": 0.4678935110569, + "learning_rate": 0.00019811859785028846, + "loss": 1.13, + "step": 1408 + }, + { + "epoch": 0.25089031339031337, + "grad_norm": 0.5155254602432251, + "learning_rate": 0.0001981158944677935, + "loss": 1.1194, + "step": 1409 + }, + { + "epoch": 0.25106837606837606, + "grad_norm": 0.4533839523792267, + "learning_rate": 0.00019811318916292142, + "loss": 0.9464, + "step": 1410 + }, + { + "epoch": 0.25124643874643876, + "grad_norm": 0.5142433047294617, + "learning_rate": 0.00019811048193572517, + "loss": 1.0837, + "step": 1411 + }, + { + "epoch": 0.2514245014245014, + "grad_norm": 0.4330446124076843, + "learning_rate": 0.00019810777278625788, + "loss": 0.9117, + "step": 1412 + }, + { + "epoch": 0.2516025641025641, + "grad_norm": 0.44806256890296936, + "learning_rate": 0.00019810506171457254, + "loss": 1.1643, + "step": 1413 + }, + { + "epoch": 0.2517806267806268, + "grad_norm": 0.43526285886764526, + "learning_rate": 0.00019810234872072235, + "loss": 0.9776, + "step": 1414 + }, + { + "epoch": 0.25195868945868943, + "grad_norm": 0.47394511103630066, + "learning_rate": 0.00019809963380476039, + "loss": 1.0935, + "step": 1415 + }, + { + "epoch": 0.25213675213675213, + "grad_norm": 0.48961278796195984, + "learning_rate": 0.00019809691696673993, + "loss": 1.179, + "step": 1416 + }, + { + "epoch": 0.2523148148148148, + "grad_norm": 0.43153589963912964, + "learning_rate": 0.00019809419820671412, + "loss": 0.906, + "step": 1417 + }, + { + "epoch": 0.25249287749287747, + "grad_norm": 0.41187527775764465, + "learning_rate": 0.00019809147752473632, + "loss": 0.899, + "step": 1418 + }, + { + "epoch": 0.25267094017094016, + "grad_norm": 0.5003183484077454, + "learning_rate": 0.00019808875492085973, + "loss": 1.0606, + "step": 1419 + }, + { + "epoch": 0.25284900284900286, + "grad_norm": 0.4430316984653473, + "learning_rate": 0.00019808603039513778, + "loss": 0.9167, + "step": 1420 + }, + { + "epoch": 0.25302706552706555, + "grad_norm": 0.4577699601650238, + "learning_rate": 0.00019808330394762382, + "loss": 1.1184, + "step": 1421 + }, + { + "epoch": 0.2532051282051282, + "grad_norm": 0.42656826972961426, + "learning_rate": 0.0001980805755783713, + "loss": 0.9335, + "step": 1422 + }, + { + "epoch": 0.2533831908831909, + "grad_norm": 0.40980881452560425, + "learning_rate": 0.0001980778452874336, + "loss": 0.9756, + "step": 1423 + }, + { + "epoch": 0.2535612535612536, + "grad_norm": 0.5752090811729431, + "learning_rate": 0.00019807511307486423, + "loss": 1.1694, + "step": 1424 + }, + { + "epoch": 0.2537393162393162, + "grad_norm": 0.5000349283218384, + "learning_rate": 0.00019807237894071681, + "loss": 0.9515, + "step": 1425 + }, + { + "epoch": 0.2539173789173789, + "grad_norm": 0.5159069299697876, + "learning_rate": 0.00019806964288504483, + "loss": 1.4014, + "step": 1426 + }, + { + "epoch": 0.2540954415954416, + "grad_norm": 0.5377941131591797, + "learning_rate": 0.00019806690490790194, + "loss": 1.2832, + "step": 1427 + }, + { + "epoch": 0.25427350427350426, + "grad_norm": 0.4565938711166382, + "learning_rate": 0.00019806416500934174, + "loss": 1.0629, + "step": 1428 + }, + { + "epoch": 0.25445156695156695, + "grad_norm": 0.49867144227027893, + "learning_rate": 0.00019806142318941797, + "loss": 1.2011, + "step": 1429 + }, + { + "epoch": 0.25462962962962965, + "grad_norm": 0.5111994743347168, + "learning_rate": 0.00019805867944818427, + "loss": 0.8925, + "step": 1430 + }, + { + "epoch": 0.2548076923076923, + "grad_norm": 0.5204268097877502, + "learning_rate": 0.00019805593378569448, + "loss": 1.2956, + "step": 1431 + }, + { + "epoch": 0.254985754985755, + "grad_norm": 0.3889026939868927, + "learning_rate": 0.00019805318620200234, + "loss": 1.0355, + "step": 1432 + }, + { + "epoch": 0.2551638176638177, + "grad_norm": 0.46825656294822693, + "learning_rate": 0.00019805043669716174, + "loss": 1.0444, + "step": 1433 + }, + { + "epoch": 0.2553418803418803, + "grad_norm": 0.4509420394897461, + "learning_rate": 0.00019804768527122648, + "loss": 1.0423, + "step": 1434 + }, + { + "epoch": 0.255519943019943, + "grad_norm": 0.4514774978160858, + "learning_rate": 0.0001980449319242505, + "loss": 1.1588, + "step": 1435 + }, + { + "epoch": 0.2556980056980057, + "grad_norm": 0.43019044399261475, + "learning_rate": 0.0001980421766562878, + "loss": 0.9939, + "step": 1436 + }, + { + "epoch": 0.25587606837606836, + "grad_norm": 0.5056091547012329, + "learning_rate": 0.00019803941946739228, + "loss": 1.1238, + "step": 1437 + }, + { + "epoch": 0.25605413105413105, + "grad_norm": 0.48664605617523193, + "learning_rate": 0.000198036660357618, + "loss": 1.0702, + "step": 1438 + }, + { + "epoch": 0.25623219373219375, + "grad_norm": 0.4500972032546997, + "learning_rate": 0.000198033899327019, + "loss": 0.9365, + "step": 1439 + }, + { + "epoch": 0.2564102564102564, + "grad_norm": 0.4800589382648468, + "learning_rate": 0.0001980311363756494, + "loss": 1.1159, + "step": 1440 + }, + { + "epoch": 0.2565883190883191, + "grad_norm": 0.3486495316028595, + "learning_rate": 0.0001980283715035633, + "loss": 0.6029, + "step": 1441 + }, + { + "epoch": 0.2567663817663818, + "grad_norm": 0.46258702874183655, + "learning_rate": 0.00019802560471081493, + "loss": 1.025, + "step": 1442 + }, + { + "epoch": 0.2569444444444444, + "grad_norm": 0.4846673607826233, + "learning_rate": 0.00019802283599745844, + "loss": 1.1105, + "step": 1443 + }, + { + "epoch": 0.2571225071225071, + "grad_norm": 0.4586990475654602, + "learning_rate": 0.00019802006536354813, + "loss": 0.9897, + "step": 1444 + }, + { + "epoch": 0.2573005698005698, + "grad_norm": 0.5177786350250244, + "learning_rate": 0.00019801729280913825, + "loss": 1.2558, + "step": 1445 + }, + { + "epoch": 0.25747863247863245, + "grad_norm": 0.43213751912117004, + "learning_rate": 0.00019801451833428312, + "loss": 1.0961, + "step": 1446 + }, + { + "epoch": 0.25765669515669515, + "grad_norm": 0.42974478006362915, + "learning_rate": 0.00019801174193903714, + "loss": 1.0659, + "step": 1447 + }, + { + "epoch": 0.25783475783475784, + "grad_norm": 0.4424504339694977, + "learning_rate": 0.00019800896362345464, + "loss": 0.9805, + "step": 1448 + }, + { + "epoch": 0.25801282051282054, + "grad_norm": 0.4734833836555481, + "learning_rate": 0.0001980061833875901, + "loss": 1.255, + "step": 1449 + }, + { + "epoch": 0.2581908831908832, + "grad_norm": 0.41024845838546753, + "learning_rate": 0.000198003401231498, + "loss": 1.0908, + "step": 1450 + }, + { + "epoch": 0.2583689458689459, + "grad_norm": 0.43603816628456116, + "learning_rate": 0.00019800061715523283, + "loss": 1.0611, + "step": 1451 + }, + { + "epoch": 0.25854700854700857, + "grad_norm": 0.4871339499950409, + "learning_rate": 0.00019799783115884915, + "loss": 1.1851, + "step": 1452 + }, + { + "epoch": 0.2587250712250712, + "grad_norm": 0.49758270382881165, + "learning_rate": 0.00019799504324240157, + "loss": 1.1936, + "step": 1453 + }, + { + "epoch": 0.2589031339031339, + "grad_norm": 0.4201010763645172, + "learning_rate": 0.00019799225340594466, + "loss": 1.1567, + "step": 1454 + }, + { + "epoch": 0.2590811965811966, + "grad_norm": 0.4200313091278076, + "learning_rate": 0.00019798946164953309, + "loss": 0.9666, + "step": 1455 + }, + { + "epoch": 0.25925925925925924, + "grad_norm": 0.43001702427864075, + "learning_rate": 0.0001979866679732216, + "loss": 1.0104, + "step": 1456 + }, + { + "epoch": 0.25943732193732194, + "grad_norm": 0.46733465790748596, + "learning_rate": 0.0001979838723770649, + "loss": 1.0927, + "step": 1457 + }, + { + "epoch": 0.25961538461538464, + "grad_norm": 0.4513280391693115, + "learning_rate": 0.00019798107486111773, + "loss": 1.0282, + "step": 1458 + }, + { + "epoch": 0.2597934472934473, + "grad_norm": 0.40411749482154846, + "learning_rate": 0.00019797827542543495, + "loss": 1.0789, + "step": 1459 + }, + { + "epoch": 0.25997150997151, + "grad_norm": 0.4359099268913269, + "learning_rate": 0.0001979754740700714, + "loss": 1.0616, + "step": 1460 + }, + { + "epoch": 0.26014957264957267, + "grad_norm": 0.4979047477245331, + "learning_rate": 0.00019797267079508198, + "loss": 1.2948, + "step": 1461 + }, + { + "epoch": 0.2603276353276353, + "grad_norm": 0.44698619842529297, + "learning_rate": 0.0001979698656005216, + "loss": 0.9198, + "step": 1462 + }, + { + "epoch": 0.260505698005698, + "grad_norm": 0.48437631130218506, + "learning_rate": 0.00019796705848644516, + "loss": 1.3207, + "step": 1463 + }, + { + "epoch": 0.2606837606837607, + "grad_norm": 0.4382587671279907, + "learning_rate": 0.00019796424945290778, + "loss": 1.1315, + "step": 1464 + }, + { + "epoch": 0.26086182336182334, + "grad_norm": 0.4565944969654083, + "learning_rate": 0.0001979614384999644, + "loss": 1.1893, + "step": 1465 + }, + { + "epoch": 0.26103988603988604, + "grad_norm": 0.4705163836479187, + "learning_rate": 0.00019795862562767017, + "loss": 1.1132, + "step": 1466 + }, + { + "epoch": 0.26121794871794873, + "grad_norm": 0.525184690952301, + "learning_rate": 0.00019795581083608012, + "loss": 1.2111, + "step": 1467 + }, + { + "epoch": 0.2613960113960114, + "grad_norm": 0.45215457677841187, + "learning_rate": 0.00019795299412524945, + "loss": 1.1851, + "step": 1468 + }, + { + "epoch": 0.26157407407407407, + "grad_norm": 0.4336663484573364, + "learning_rate": 0.00019795017549523335, + "loss": 1.0147, + "step": 1469 + }, + { + "epoch": 0.26175213675213677, + "grad_norm": 0.5327649712562561, + "learning_rate": 0.00019794735494608703, + "loss": 1.1743, + "step": 1470 + }, + { + "epoch": 0.2619301994301994, + "grad_norm": 0.49972307682037354, + "learning_rate": 0.00019794453247786578, + "loss": 1.1624, + "step": 1471 + }, + { + "epoch": 0.2621082621082621, + "grad_norm": 0.43475785851478577, + "learning_rate": 0.00019794170809062485, + "loss": 0.9888, + "step": 1472 + }, + { + "epoch": 0.2622863247863248, + "grad_norm": 0.428838849067688, + "learning_rate": 0.0001979388817844196, + "loss": 0.9154, + "step": 1473 + }, + { + "epoch": 0.26246438746438744, + "grad_norm": 0.508568286895752, + "learning_rate": 0.00019793605355930544, + "loss": 1.1679, + "step": 1474 + }, + { + "epoch": 0.26264245014245013, + "grad_norm": 0.47791770100593567, + "learning_rate": 0.00019793322341533776, + "loss": 1.1375, + "step": 1475 + }, + { + "epoch": 0.26282051282051283, + "grad_norm": 0.41909220814704895, + "learning_rate": 0.00019793039135257196, + "loss": 1.0235, + "step": 1476 + }, + { + "epoch": 0.26299857549857547, + "grad_norm": 0.5564408302307129, + "learning_rate": 0.00019792755737106361, + "loss": 1.0756, + "step": 1477 + }, + { + "epoch": 0.26317663817663817, + "grad_norm": 0.42813625931739807, + "learning_rate": 0.0001979247214708682, + "loss": 0.8213, + "step": 1478 + }, + { + "epoch": 0.26335470085470086, + "grad_norm": 0.44495970010757446, + "learning_rate": 0.00019792188365204126, + "loss": 0.9654, + "step": 1479 + }, + { + "epoch": 0.26353276353276356, + "grad_norm": 0.47473424673080444, + "learning_rate": 0.00019791904391463846, + "loss": 1.1643, + "step": 1480 + }, + { + "epoch": 0.2637108262108262, + "grad_norm": 0.40189051628112793, + "learning_rate": 0.0001979162022587154, + "loss": 0.8687, + "step": 1481 + }, + { + "epoch": 0.2638888888888889, + "grad_norm": 0.44629937410354614, + "learning_rate": 0.00019791335868432776, + "loss": 1.0284, + "step": 1482 + }, + { + "epoch": 0.2640669515669516, + "grad_norm": 0.511275053024292, + "learning_rate": 0.00019791051319153124, + "loss": 1.2217, + "step": 1483 + }, + { + "epoch": 0.26424501424501423, + "grad_norm": 0.5136445164680481, + "learning_rate": 0.00019790766578038163, + "loss": 1.1129, + "step": 1484 + }, + { + "epoch": 0.2644230769230769, + "grad_norm": 0.4450451135635376, + "learning_rate": 0.00019790481645093469, + "loss": 0.9912, + "step": 1485 + }, + { + "epoch": 0.2646011396011396, + "grad_norm": 0.39455199241638184, + "learning_rate": 0.00019790196520324621, + "loss": 1.0887, + "step": 1486 + }, + { + "epoch": 0.26477920227920226, + "grad_norm": 0.4444045126438141, + "learning_rate": 0.00019789911203737216, + "loss": 1.1559, + "step": 1487 + }, + { + "epoch": 0.26495726495726496, + "grad_norm": 0.4769677221775055, + "learning_rate": 0.0001978962569533683, + "loss": 1.147, + "step": 1488 + }, + { + "epoch": 0.26513532763532766, + "grad_norm": 0.40226617455482483, + "learning_rate": 0.0001978933999512907, + "loss": 1.0966, + "step": 1489 + }, + { + "epoch": 0.2653133903133903, + "grad_norm": 0.4640974700450897, + "learning_rate": 0.00019789054103119526, + "loss": 1.1002, + "step": 1490 + }, + { + "epoch": 0.265491452991453, + "grad_norm": 0.48251107335090637, + "learning_rate": 0.00019788768019313806, + "loss": 1.07, + "step": 1491 + }, + { + "epoch": 0.2656695156695157, + "grad_norm": 0.4836949408054352, + "learning_rate": 0.00019788481743717506, + "loss": 1.2992, + "step": 1492 + }, + { + "epoch": 0.26584757834757833, + "grad_norm": 0.4253857135772705, + "learning_rate": 0.00019788195276336244, + "loss": 1.1326, + "step": 1493 + }, + { + "epoch": 0.266025641025641, + "grad_norm": 0.5161862373352051, + "learning_rate": 0.0001978790861717563, + "loss": 1.2131, + "step": 1494 + }, + { + "epoch": 0.2662037037037037, + "grad_norm": 0.5223346948623657, + "learning_rate": 0.00019787621766241274, + "loss": 1.0933, + "step": 1495 + }, + { + "epoch": 0.26638176638176636, + "grad_norm": 0.37622541189193726, + "learning_rate": 0.000197873347235388, + "loss": 0.8919, + "step": 1496 + }, + { + "epoch": 0.26655982905982906, + "grad_norm": 0.4425419569015503, + "learning_rate": 0.0001978704748907384, + "loss": 1.0411, + "step": 1497 + }, + { + "epoch": 0.26673789173789175, + "grad_norm": 0.4536985456943512, + "learning_rate": 0.00019786760062852015, + "loss": 1.2747, + "step": 1498 + }, + { + "epoch": 0.2669159544159544, + "grad_norm": 0.4998049736022949, + "learning_rate": 0.00019786472444878955, + "loss": 1.3214, + "step": 1499 + }, + { + "epoch": 0.2670940170940171, + "grad_norm": 0.42104312777519226, + "learning_rate": 0.00019786184635160295, + "loss": 0.7878, + "step": 1500 + }, + { + "epoch": 0.2672720797720798, + "grad_norm": 0.5354288220405579, + "learning_rate": 0.00019785896633701678, + "loss": 1.0642, + "step": 1501 + }, + { + "epoch": 0.2674501424501424, + "grad_norm": 0.4681485891342163, + "learning_rate": 0.00019785608440508744, + "loss": 1.1737, + "step": 1502 + }, + { + "epoch": 0.2676282051282051, + "grad_norm": 0.49107062816619873, + "learning_rate": 0.0001978532005558714, + "loss": 1.1507, + "step": 1503 + }, + { + "epoch": 0.2678062678062678, + "grad_norm": 0.4173283576965332, + "learning_rate": 0.0001978503147894252, + "loss": 1.0538, + "step": 1504 + }, + { + "epoch": 0.26798433048433046, + "grad_norm": 0.49354055523872375, + "learning_rate": 0.0001978474271058053, + "loss": 1.1043, + "step": 1505 + }, + { + "epoch": 0.26816239316239315, + "grad_norm": 0.5787215232849121, + "learning_rate": 0.00019784453750506834, + "loss": 0.9245, + "step": 1506 + }, + { + "epoch": 0.26834045584045585, + "grad_norm": 0.48982590436935425, + "learning_rate": 0.00019784164598727095, + "loss": 1.2007, + "step": 1507 + }, + { + "epoch": 0.26851851851851855, + "grad_norm": 0.4971007704734802, + "learning_rate": 0.00019783875255246973, + "loss": 1.1174, + "step": 1508 + }, + { + "epoch": 0.2686965811965812, + "grad_norm": 0.5200340151786804, + "learning_rate": 0.00019783585720072142, + "loss": 1.1967, + "step": 1509 + }, + { + "epoch": 0.2688746438746439, + "grad_norm": 0.47911885380744934, + "learning_rate": 0.00019783295993208271, + "loss": 1.162, + "step": 1510 + }, + { + "epoch": 0.2690527065527066, + "grad_norm": 0.4764275848865509, + "learning_rate": 0.00019783006074661037, + "loss": 1.1358, + "step": 1511 + }, + { + "epoch": 0.2692307692307692, + "grad_norm": 0.478545606136322, + "learning_rate": 0.00019782715964436124, + "loss": 1.0096, + "step": 1512 + }, + { + "epoch": 0.2694088319088319, + "grad_norm": 0.5512787699699402, + "learning_rate": 0.00019782425662539212, + "loss": 1.1799, + "step": 1513 + }, + { + "epoch": 0.2695868945868946, + "grad_norm": 0.5495108962059021, + "learning_rate": 0.00019782135168975988, + "loss": 1.0959, + "step": 1514 + }, + { + "epoch": 0.26976495726495725, + "grad_norm": 0.42052868008613586, + "learning_rate": 0.0001978184448375215, + "loss": 1.1872, + "step": 1515 + }, + { + "epoch": 0.26994301994301995, + "grad_norm": 0.4994426965713501, + "learning_rate": 0.0001978155360687339, + "loss": 1.0568, + "step": 1516 + }, + { + "epoch": 0.27012108262108264, + "grad_norm": 0.459577351808548, + "learning_rate": 0.00019781262538345402, + "loss": 1.0315, + "step": 1517 + }, + { + "epoch": 0.2702991452991453, + "grad_norm": 0.4792841374874115, + "learning_rate": 0.00019780971278173895, + "loss": 1.2055, + "step": 1518 + }, + { + "epoch": 0.270477207977208, + "grad_norm": 0.5017708539962769, + "learning_rate": 0.00019780679826364575, + "loss": 1.157, + "step": 1519 + }, + { + "epoch": 0.2706552706552707, + "grad_norm": 0.5197349786758423, + "learning_rate": 0.00019780388182923152, + "loss": 0.9101, + "step": 1520 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 0.4226742684841156, + "learning_rate": 0.00019780096347855338, + "loss": 1.0525, + "step": 1521 + }, + { + "epoch": 0.271011396011396, + "grad_norm": 0.5058164596557617, + "learning_rate": 0.00019779804321166852, + "loss": 0.931, + "step": 1522 + }, + { + "epoch": 0.2711894586894587, + "grad_norm": 0.44492244720458984, + "learning_rate": 0.00019779512102863418, + "loss": 1.0641, + "step": 1523 + }, + { + "epoch": 0.27136752136752135, + "grad_norm": 0.5348989963531494, + "learning_rate": 0.00019779219692950758, + "loss": 1.1692, + "step": 1524 + }, + { + "epoch": 0.27154558404558404, + "grad_norm": 0.4631774425506592, + "learning_rate": 0.00019778927091434602, + "loss": 1.0876, + "step": 1525 + }, + { + "epoch": 0.27172364672364674, + "grad_norm": 0.45957499742507935, + "learning_rate": 0.00019778634298320684, + "loss": 0.9527, + "step": 1526 + }, + { + "epoch": 0.2719017094017094, + "grad_norm": 0.4506755769252777, + "learning_rate": 0.00019778341313614743, + "loss": 1.086, + "step": 1527 + }, + { + "epoch": 0.2720797720797721, + "grad_norm": 0.4900587797164917, + "learning_rate": 0.00019778048137322513, + "loss": 0.9911, + "step": 1528 + }, + { + "epoch": 0.27225783475783477, + "grad_norm": 0.478127658367157, + "learning_rate": 0.00019777754769449745, + "loss": 1.2083, + "step": 1529 + }, + { + "epoch": 0.2724358974358974, + "grad_norm": 0.47220897674560547, + "learning_rate": 0.00019777461210002183, + "loss": 1.0313, + "step": 1530 + }, + { + "epoch": 0.2726139601139601, + "grad_norm": 0.4526277184486389, + "learning_rate": 0.0001977716745898558, + "loss": 1.2648, + "step": 1531 + }, + { + "epoch": 0.2727920227920228, + "grad_norm": 0.42907601594924927, + "learning_rate": 0.00019776873516405688, + "loss": 0.8645, + "step": 1532 + }, + { + "epoch": 0.27297008547008544, + "grad_norm": 0.43440163135528564, + "learning_rate": 0.00019776579382268272, + "loss": 0.9702, + "step": 1533 + }, + { + "epoch": 0.27314814814814814, + "grad_norm": 0.48213550448417664, + "learning_rate": 0.0001977628505657909, + "loss": 0.998, + "step": 1534 + }, + { + "epoch": 0.27332621082621084, + "grad_norm": 0.43385565280914307, + "learning_rate": 0.00019775990539343914, + "loss": 1.0575, + "step": 1535 + }, + { + "epoch": 0.27350427350427353, + "grad_norm": 0.45706847310066223, + "learning_rate": 0.00019775695830568507, + "loss": 1.3024, + "step": 1536 + }, + { + "epoch": 0.27368233618233617, + "grad_norm": 0.45769137144088745, + "learning_rate": 0.00019775400930258652, + "loss": 1.0987, + "step": 1537 + }, + { + "epoch": 0.27386039886039887, + "grad_norm": 0.44682395458221436, + "learning_rate": 0.00019775105838420117, + "loss": 1.1327, + "step": 1538 + }, + { + "epoch": 0.27403846153846156, + "grad_norm": 0.5923072099685669, + "learning_rate": 0.00019774810555058694, + "loss": 1.4766, + "step": 1539 + }, + { + "epoch": 0.2742165242165242, + "grad_norm": 0.4327206015586853, + "learning_rate": 0.0001977451508018016, + "loss": 1.1175, + "step": 1540 + }, + { + "epoch": 0.2743945868945869, + "grad_norm": 0.48036691546440125, + "learning_rate": 0.00019774219413790315, + "loss": 1.1189, + "step": 1541 + }, + { + "epoch": 0.2745726495726496, + "grad_norm": 0.41371914744377136, + "learning_rate": 0.00019773923555894935, + "loss": 1.1366, + "step": 1542 + }, + { + "epoch": 0.27475071225071224, + "grad_norm": 0.4452378749847412, + "learning_rate": 0.00019773627506499832, + "loss": 0.9517, + "step": 1543 + }, + { + "epoch": 0.27492877492877493, + "grad_norm": 0.469098299741745, + "learning_rate": 0.00019773331265610802, + "loss": 1.0848, + "step": 1544 + }, + { + "epoch": 0.27510683760683763, + "grad_norm": 0.5390294790267944, + "learning_rate": 0.00019773034833233646, + "loss": 0.8589, + "step": 1545 + }, + { + "epoch": 0.27528490028490027, + "grad_norm": 0.5368238091468811, + "learning_rate": 0.00019772738209374174, + "loss": 1.2954, + "step": 1546 + }, + { + "epoch": 0.27546296296296297, + "grad_norm": 0.4705318510532379, + "learning_rate": 0.00019772441394038198, + "loss": 1.2252, + "step": 1547 + }, + { + "epoch": 0.27564102564102566, + "grad_norm": 0.4682813286781311, + "learning_rate": 0.00019772144387231533, + "loss": 1.0855, + "step": 1548 + }, + { + "epoch": 0.2758190883190883, + "grad_norm": 0.46876460313796997, + "learning_rate": 0.0001977184718896, + "loss": 1.1959, + "step": 1549 + }, + { + "epoch": 0.275997150997151, + "grad_norm": 0.4172806441783905, + "learning_rate": 0.00019771549799229416, + "loss": 1.2166, + "step": 1550 + }, + { + "epoch": 0.2761752136752137, + "grad_norm": 0.5088075399398804, + "learning_rate": 0.0001977125221804562, + "loss": 1.1285, + "step": 1551 + }, + { + "epoch": 0.27635327635327633, + "grad_norm": 0.4728628396987915, + "learning_rate": 0.0001977095444541443, + "loss": 1.2985, + "step": 1552 + }, + { + "epoch": 0.27653133903133903, + "grad_norm": 0.4431236684322357, + "learning_rate": 0.00019770656481341684, + "loss": 1.1298, + "step": 1553 + }, + { + "epoch": 0.2767094017094017, + "grad_norm": 0.474065363407135, + "learning_rate": 0.00019770358325833223, + "loss": 1.1915, + "step": 1554 + }, + { + "epoch": 0.27688746438746437, + "grad_norm": 0.45718875527381897, + "learning_rate": 0.00019770059978894885, + "loss": 1.0626, + "step": 1555 + }, + { + "epoch": 0.27706552706552706, + "grad_norm": 0.49300211668014526, + "learning_rate": 0.00019769761440532522, + "loss": 1.0134, + "step": 1556 + }, + { + "epoch": 0.27724358974358976, + "grad_norm": 0.4389498829841614, + "learning_rate": 0.00019769462710751974, + "loss": 1.0292, + "step": 1557 + }, + { + "epoch": 0.2774216524216524, + "grad_norm": 0.47330448031425476, + "learning_rate": 0.000197691637895591, + "loss": 1.1273, + "step": 1558 + }, + { + "epoch": 0.2775997150997151, + "grad_norm": 0.5322058200836182, + "learning_rate": 0.00019768864676959755, + "loss": 1.059, + "step": 1559 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 0.4714536964893341, + "learning_rate": 0.000197685653729598, + "loss": 1.1987, + "step": 1560 + }, + { + "epoch": 0.27795584045584043, + "grad_norm": 0.48687809705734253, + "learning_rate": 0.00019768265877565097, + "loss": 1.3206, + "step": 1561 + }, + { + "epoch": 0.2781339031339031, + "grad_norm": 0.46066713333129883, + "learning_rate": 0.00019767966190781518, + "loss": 1.0845, + "step": 1562 + }, + { + "epoch": 0.2783119658119658, + "grad_norm": 0.44372090697288513, + "learning_rate": 0.00019767666312614935, + "loss": 1.0942, + "step": 1563 + }, + { + "epoch": 0.27849002849002846, + "grad_norm": 0.4615907073020935, + "learning_rate": 0.00019767366243071216, + "loss": 1.071, + "step": 1564 + }, + { + "epoch": 0.27866809116809116, + "grad_norm": 0.502097487449646, + "learning_rate": 0.0001976706598215625, + "loss": 1.1164, + "step": 1565 + }, + { + "epoch": 0.27884615384615385, + "grad_norm": 0.4371815621852875, + "learning_rate": 0.00019766765529875913, + "loss": 1.0252, + "step": 1566 + }, + { + "epoch": 0.27902421652421655, + "grad_norm": 0.43035808205604553, + "learning_rate": 0.00019766464886236093, + "loss": 1.073, + "step": 1567 + }, + { + "epoch": 0.2792022792022792, + "grad_norm": 0.49721601605415344, + "learning_rate": 0.00019766164051242683, + "loss": 1.0316, + "step": 1568 + }, + { + "epoch": 0.2793803418803419, + "grad_norm": 0.44866231083869934, + "learning_rate": 0.00019765863024901576, + "loss": 1.0951, + "step": 1569 + }, + { + "epoch": 0.2795584045584046, + "grad_norm": 0.46318337321281433, + "learning_rate": 0.0001976556180721867, + "loss": 0.9836, + "step": 1570 + }, + { + "epoch": 0.2797364672364672, + "grad_norm": 0.4227696657180786, + "learning_rate": 0.00019765260398199868, + "loss": 1.0414, + "step": 1571 + }, + { + "epoch": 0.2799145299145299, + "grad_norm": 0.6062980890274048, + "learning_rate": 0.00019764958797851073, + "loss": 1.137, + "step": 1572 + }, + { + "epoch": 0.2800925925925926, + "grad_norm": 0.4856833219528198, + "learning_rate": 0.00019764657006178196, + "loss": 1.1361, + "step": 1573 + }, + { + "epoch": 0.28027065527065526, + "grad_norm": 0.45612895488739014, + "learning_rate": 0.00019764355023187146, + "loss": 1.0005, + "step": 1574 + }, + { + "epoch": 0.28044871794871795, + "grad_norm": 0.4143696129322052, + "learning_rate": 0.00019764052848883845, + "loss": 1.051, + "step": 1575 + }, + { + "epoch": 0.28062678062678065, + "grad_norm": 0.4532071352005005, + "learning_rate": 0.00019763750483274212, + "loss": 1.0595, + "step": 1576 + }, + { + "epoch": 0.2808048433048433, + "grad_norm": 0.4940357208251953, + "learning_rate": 0.0001976344792636417, + "loss": 1.0983, + "step": 1577 + }, + { + "epoch": 0.280982905982906, + "grad_norm": 0.44405099749565125, + "learning_rate": 0.0001976314517815965, + "loss": 1.0846, + "step": 1578 + }, + { + "epoch": 0.2811609686609687, + "grad_norm": 0.5508625507354736, + "learning_rate": 0.00019762842238666578, + "loss": 1.1722, + "step": 1579 + }, + { + "epoch": 0.2813390313390313, + "grad_norm": 0.5241084694862366, + "learning_rate": 0.00019762539107890894, + "loss": 1.351, + "step": 1580 + }, + { + "epoch": 0.281517094017094, + "grad_norm": 0.5307353734970093, + "learning_rate": 0.00019762235785838537, + "loss": 1.1868, + "step": 1581 + }, + { + "epoch": 0.2816951566951567, + "grad_norm": 0.45697924494743347, + "learning_rate": 0.00019761932272515447, + "loss": 1.1982, + "step": 1582 + }, + { + "epoch": 0.28187321937321935, + "grad_norm": 0.412483811378479, + "learning_rate": 0.00019761628567927574, + "loss": 1.0433, + "step": 1583 + }, + { + "epoch": 0.28205128205128205, + "grad_norm": 0.4614165425300598, + "learning_rate": 0.00019761324672080868, + "loss": 1.104, + "step": 1584 + }, + { + "epoch": 0.28222934472934474, + "grad_norm": 0.47644901275634766, + "learning_rate": 0.00019761020584981284, + "loss": 1.1037, + "step": 1585 + }, + { + "epoch": 0.2824074074074074, + "grad_norm": 0.4985184669494629, + "learning_rate": 0.00019760716306634773, + "loss": 1.2213, + "step": 1586 + }, + { + "epoch": 0.2825854700854701, + "grad_norm": 0.508301317691803, + "learning_rate": 0.00019760411837047305, + "loss": 1.1315, + "step": 1587 + }, + { + "epoch": 0.2827635327635328, + "grad_norm": 0.5346587300300598, + "learning_rate": 0.00019760107176224845, + "loss": 1.2281, + "step": 1588 + }, + { + "epoch": 0.2829415954415954, + "grad_norm": 0.5106825232505798, + "learning_rate": 0.00019759802324173357, + "loss": 1.2904, + "step": 1589 + }, + { + "epoch": 0.2831196581196581, + "grad_norm": 0.46458688378334045, + "learning_rate": 0.00019759497280898817, + "loss": 1.0861, + "step": 1590 + }, + { + "epoch": 0.2832977207977208, + "grad_norm": 0.49115365743637085, + "learning_rate": 0.00019759192046407201, + "loss": 1.0529, + "step": 1591 + }, + { + "epoch": 0.28347578347578345, + "grad_norm": 0.5114167332649231, + "learning_rate": 0.0001975888662070449, + "loss": 1.2555, + "step": 1592 + }, + { + "epoch": 0.28365384615384615, + "grad_norm": 0.45844775438308716, + "learning_rate": 0.0001975858100379667, + "loss": 1.0662, + "step": 1593 + }, + { + "epoch": 0.28383190883190884, + "grad_norm": 0.4684161841869354, + "learning_rate": 0.00019758275195689727, + "loss": 1.0537, + "step": 1594 + }, + { + "epoch": 0.28400997150997154, + "grad_norm": 0.4816220998764038, + "learning_rate": 0.0001975796919638965, + "loss": 1.126, + "step": 1595 + }, + { + "epoch": 0.2841880341880342, + "grad_norm": 0.46578118205070496, + "learning_rate": 0.0001975766300590244, + "loss": 0.9651, + "step": 1596 + }, + { + "epoch": 0.2843660968660969, + "grad_norm": 0.4181675612926483, + "learning_rate": 0.0001975735662423409, + "loss": 1.0888, + "step": 1597 + }, + { + "epoch": 0.28454415954415957, + "grad_norm": 0.49417954683303833, + "learning_rate": 0.00019757050051390609, + "loss": 1.1878, + "step": 1598 + }, + { + "epoch": 0.2847222222222222, + "grad_norm": 0.47264960408210754, + "learning_rate": 0.00019756743287377998, + "loss": 1.027, + "step": 1599 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 0.47686338424682617, + "learning_rate": 0.0001975643633220227, + "loss": 1.1307, + "step": 1600 + }, + { + "epoch": 0.2850783475783476, + "grad_norm": 0.5571266412734985, + "learning_rate": 0.00019756129185869443, + "loss": 0.984, + "step": 1601 + }, + { + "epoch": 0.28525641025641024, + "grad_norm": 0.46942809224128723, + "learning_rate": 0.00019755821848385527, + "loss": 1.0397, + "step": 1602 + }, + { + "epoch": 0.28543447293447294, + "grad_norm": 0.6325890421867371, + "learning_rate": 0.00019755514319756551, + "loss": 1.0918, + "step": 1603 + }, + { + "epoch": 0.28561253561253563, + "grad_norm": 0.5297608375549316, + "learning_rate": 0.00019755206599988533, + "loss": 0.9911, + "step": 1604 + }, + { + "epoch": 0.2857905982905983, + "grad_norm": 0.4736945331096649, + "learning_rate": 0.00019754898689087512, + "loss": 1.0786, + "step": 1605 + }, + { + "epoch": 0.28596866096866097, + "grad_norm": 0.5048685669898987, + "learning_rate": 0.00019754590587059512, + "loss": 0.9834, + "step": 1606 + }, + { + "epoch": 0.28614672364672367, + "grad_norm": 0.3823149502277374, + "learning_rate": 0.00019754282293910574, + "loss": 0.8341, + "step": 1607 + }, + { + "epoch": 0.2863247863247863, + "grad_norm": 0.44071945548057556, + "learning_rate": 0.00019753973809646738, + "loss": 1.131, + "step": 1608 + }, + { + "epoch": 0.286502849002849, + "grad_norm": 0.44182759523391724, + "learning_rate": 0.00019753665134274043, + "loss": 1.0321, + "step": 1609 + }, + { + "epoch": 0.2866809116809117, + "grad_norm": 0.4486250877380371, + "learning_rate": 0.00019753356267798546, + "loss": 0.9941, + "step": 1610 + }, + { + "epoch": 0.28685897435897434, + "grad_norm": 0.42796584963798523, + "learning_rate": 0.00019753047210226292, + "loss": 1.0235, + "step": 1611 + }, + { + "epoch": 0.28703703703703703, + "grad_norm": 0.47294023633003235, + "learning_rate": 0.00019752737961563336, + "loss": 1.11, + "step": 1612 + }, + { + "epoch": 0.28721509971509973, + "grad_norm": 0.44550734758377075, + "learning_rate": 0.00019752428521815742, + "loss": 1.0849, + "step": 1613 + }, + { + "epoch": 0.28739316239316237, + "grad_norm": 0.44189929962158203, + "learning_rate": 0.0001975211889098957, + "loss": 0.8904, + "step": 1614 + }, + { + "epoch": 0.28757122507122507, + "grad_norm": 0.5302733182907104, + "learning_rate": 0.00019751809069090885, + "loss": 1.2348, + "step": 1615 + }, + { + "epoch": 0.28774928774928776, + "grad_norm": 0.5951390862464905, + "learning_rate": 0.00019751499056125762, + "loss": 1.3035, + "step": 1616 + }, + { + "epoch": 0.2879273504273504, + "grad_norm": 0.5431534647941589, + "learning_rate": 0.0001975118885210027, + "loss": 1.0016, + "step": 1617 + }, + { + "epoch": 0.2881054131054131, + "grad_norm": 0.47301986813545227, + "learning_rate": 0.00019750878457020489, + "loss": 1.2245, + "step": 1618 + }, + { + "epoch": 0.2882834757834758, + "grad_norm": 0.44785359501838684, + "learning_rate": 0.00019750567870892497, + "loss": 1.122, + "step": 1619 + }, + { + "epoch": 0.28846153846153844, + "grad_norm": 0.49494361877441406, + "learning_rate": 0.00019750257093722383, + "loss": 0.9421, + "step": 1620 + }, + { + "epoch": 0.28863960113960113, + "grad_norm": 0.4484521150588989, + "learning_rate": 0.00019749946125516242, + "loss": 1.2146, + "step": 1621 + }, + { + "epoch": 0.28881766381766383, + "grad_norm": 0.4635269343852997, + "learning_rate": 0.00019749634966280156, + "loss": 0.976, + "step": 1622 + }, + { + "epoch": 0.28899572649572647, + "grad_norm": 0.5532249808311462, + "learning_rate": 0.00019749323616020226, + "loss": 1.1818, + "step": 1623 + }, + { + "epoch": 0.28917378917378916, + "grad_norm": 0.4730629622936249, + "learning_rate": 0.00019749012074742552, + "loss": 1.0321, + "step": 1624 + }, + { + "epoch": 0.28935185185185186, + "grad_norm": 0.47437289357185364, + "learning_rate": 0.0001974870034245324, + "loss": 1.1572, + "step": 1625 + }, + { + "epoch": 0.28952991452991456, + "grad_norm": 0.4796304404735565, + "learning_rate": 0.00019748388419158394, + "loss": 1.1667, + "step": 1626 + }, + { + "epoch": 0.2897079772079772, + "grad_norm": 0.42686304450035095, + "learning_rate": 0.0001974807630486413, + "loss": 0.9824, + "step": 1627 + }, + { + "epoch": 0.2898860398860399, + "grad_norm": 0.4444865584373474, + "learning_rate": 0.00019747763999576558, + "loss": 1.2789, + "step": 1628 + }, + { + "epoch": 0.2900641025641026, + "grad_norm": 0.5039985179901123, + "learning_rate": 0.000197474515033018, + "loss": 1.1488, + "step": 1629 + }, + { + "epoch": 0.29024216524216523, + "grad_norm": 0.581479549407959, + "learning_rate": 0.00019747138816045978, + "loss": 1.1232, + "step": 1630 + }, + { + "epoch": 0.2904202279202279, + "grad_norm": 0.5415821075439453, + "learning_rate": 0.00019746825937815222, + "loss": 1.2326, + "step": 1631 + }, + { + "epoch": 0.2905982905982906, + "grad_norm": 0.45528364181518555, + "learning_rate": 0.00019746512868615656, + "loss": 1.0246, + "step": 1632 + }, + { + "epoch": 0.29077635327635326, + "grad_norm": 0.5255574584007263, + "learning_rate": 0.00019746199608453418, + "loss": 1.0592, + "step": 1633 + }, + { + "epoch": 0.29095441595441596, + "grad_norm": 0.5064096450805664, + "learning_rate": 0.00019745886157334646, + "loss": 1.3439, + "step": 1634 + }, + { + "epoch": 0.29113247863247865, + "grad_norm": 0.500848650932312, + "learning_rate": 0.00019745572515265475, + "loss": 1.1212, + "step": 1635 + }, + { + "epoch": 0.2913105413105413, + "grad_norm": 0.5229088068008423, + "learning_rate": 0.00019745258682252062, + "loss": 1.1019, + "step": 1636 + }, + { + "epoch": 0.291488603988604, + "grad_norm": 0.4494398832321167, + "learning_rate": 0.00019744944658300545, + "loss": 1.1298, + "step": 1637 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 0.48383277654647827, + "learning_rate": 0.00019744630443417082, + "loss": 1.206, + "step": 1638 + }, + { + "epoch": 0.2918447293447293, + "grad_norm": 0.4870131313800812, + "learning_rate": 0.00019744316037607828, + "loss": 1.2096, + "step": 1639 + }, + { + "epoch": 0.292022792022792, + "grad_norm": 0.4153090715408325, + "learning_rate": 0.00019744001440878944, + "loss": 1.0478, + "step": 1640 + }, + { + "epoch": 0.2922008547008547, + "grad_norm": 0.4262249171733856, + "learning_rate": 0.0001974368665323659, + "loss": 1.0393, + "step": 1641 + }, + { + "epoch": 0.29237891737891736, + "grad_norm": 0.46131134033203125, + "learning_rate": 0.00019743371674686938, + "loss": 1.0908, + "step": 1642 + }, + { + "epoch": 0.29255698005698005, + "grad_norm": 0.44877463579177856, + "learning_rate": 0.0001974305650523616, + "loss": 1.1906, + "step": 1643 + }, + { + "epoch": 0.29273504273504275, + "grad_norm": 0.5199326276779175, + "learning_rate": 0.00019742741144890432, + "loss": 1.1147, + "step": 1644 + }, + { + "epoch": 0.2929131054131054, + "grad_norm": 0.48142504692077637, + "learning_rate": 0.00019742425593655924, + "loss": 1.1951, + "step": 1645 + }, + { + "epoch": 0.2930911680911681, + "grad_norm": 0.5672988891601562, + "learning_rate": 0.0001974210985153883, + "loss": 1.1817, + "step": 1646 + }, + { + "epoch": 0.2932692307692308, + "grad_norm": 0.38135233521461487, + "learning_rate": 0.00019741793918545326, + "loss": 0.8567, + "step": 1647 + }, + { + "epoch": 0.2934472934472934, + "grad_norm": 0.6153588891029358, + "learning_rate": 0.0001974147779468161, + "loss": 1.0593, + "step": 1648 + }, + { + "epoch": 0.2936253561253561, + "grad_norm": 0.38935527205467224, + "learning_rate": 0.0001974116147995387, + "loss": 0.9907, + "step": 1649 + }, + { + "epoch": 0.2938034188034188, + "grad_norm": 0.467351496219635, + "learning_rate": 0.0001974084497436831, + "loss": 1.091, + "step": 1650 + }, + { + "epoch": 0.29398148148148145, + "grad_norm": 0.45613420009613037, + "learning_rate": 0.00019740528277931128, + "loss": 0.6789, + "step": 1651 + }, + { + "epoch": 0.29415954415954415, + "grad_norm": 0.4045158326625824, + "learning_rate": 0.00019740211390648524, + "loss": 1.0727, + "step": 1652 + }, + { + "epoch": 0.29433760683760685, + "grad_norm": 0.5122803449630737, + "learning_rate": 0.00019739894312526714, + "loss": 1.2297, + "step": 1653 + }, + { + "epoch": 0.29451566951566954, + "grad_norm": 0.44304123520851135, + "learning_rate": 0.00019739577043571908, + "loss": 0.9562, + "step": 1654 + }, + { + "epoch": 0.2946937321937322, + "grad_norm": 0.6070618629455566, + "learning_rate": 0.00019739259583790322, + "loss": 1.2745, + "step": 1655 + }, + { + "epoch": 0.2948717948717949, + "grad_norm": 0.48815637826919556, + "learning_rate": 0.00019738941933188176, + "loss": 1.0574, + "step": 1656 + }, + { + "epoch": 0.2950498575498576, + "grad_norm": 0.5067802667617798, + "learning_rate": 0.00019738624091771693, + "loss": 1.1874, + "step": 1657 + }, + { + "epoch": 0.2952279202279202, + "grad_norm": 0.4956928491592407, + "learning_rate": 0.000197383060595471, + "loss": 1.1085, + "step": 1658 + }, + { + "epoch": 0.2954059829059829, + "grad_norm": 0.46313008666038513, + "learning_rate": 0.00019737987836520633, + "loss": 1.0548, + "step": 1659 + }, + { + "epoch": 0.2955840455840456, + "grad_norm": 0.49944064021110535, + "learning_rate": 0.0001973766942269852, + "loss": 1.1485, + "step": 1660 + }, + { + "epoch": 0.29576210826210825, + "grad_norm": 0.4743517339229584, + "learning_rate": 0.00019737350818087003, + "loss": 0.9279, + "step": 1661 + }, + { + "epoch": 0.29594017094017094, + "grad_norm": 0.45935431122779846, + "learning_rate": 0.00019737032022692326, + "loss": 0.9574, + "step": 1662 + }, + { + "epoch": 0.29611823361823364, + "grad_norm": 0.4550873637199402, + "learning_rate": 0.00019736713036520734, + "loss": 1.1642, + "step": 1663 + }, + { + "epoch": 0.2962962962962963, + "grad_norm": 0.45252951979637146, + "learning_rate": 0.00019736393859578474, + "loss": 1.0113, + "step": 1664 + }, + { + "epoch": 0.296474358974359, + "grad_norm": 0.5147238969802856, + "learning_rate": 0.00019736074491871804, + "loss": 1.1604, + "step": 1665 + }, + { + "epoch": 0.29665242165242167, + "grad_norm": 0.5122934579849243, + "learning_rate": 0.00019735754933406977, + "loss": 0.9525, + "step": 1666 + }, + { + "epoch": 0.2968304843304843, + "grad_norm": 0.438620001077652, + "learning_rate": 0.00019735435184190257, + "loss": 1.0728, + "step": 1667 + }, + { + "epoch": 0.297008547008547, + "grad_norm": 0.41970670223236084, + "learning_rate": 0.00019735115244227908, + "loss": 0.9782, + "step": 1668 + }, + { + "epoch": 0.2971866096866097, + "grad_norm": 0.5447152256965637, + "learning_rate": 0.000197347951135262, + "loss": 1.0633, + "step": 1669 + }, + { + "epoch": 0.29736467236467234, + "grad_norm": 0.4846996068954468, + "learning_rate": 0.00019734474792091407, + "loss": 0.9019, + "step": 1670 + }, + { + "epoch": 0.29754273504273504, + "grad_norm": 0.4721437990665436, + "learning_rate": 0.00019734154279929796, + "loss": 1.1793, + "step": 1671 + }, + { + "epoch": 0.29772079772079774, + "grad_norm": 0.4659852385520935, + "learning_rate": 0.00019733833577047655, + "loss": 1.1503, + "step": 1672 + }, + { + "epoch": 0.2978988603988604, + "grad_norm": 0.3733183443546295, + "learning_rate": 0.00019733512683451268, + "loss": 0.7763, + "step": 1673 + }, + { + "epoch": 0.2980769230769231, + "grad_norm": 0.4898292124271393, + "learning_rate": 0.0001973319159914692, + "loss": 1.3146, + "step": 1674 + }, + { + "epoch": 0.29825498575498577, + "grad_norm": 0.41774725914001465, + "learning_rate": 0.00019732870324140899, + "loss": 1.2069, + "step": 1675 + }, + { + "epoch": 0.2984330484330484, + "grad_norm": 0.4607912003993988, + "learning_rate": 0.000197325488584395, + "loss": 1.2255, + "step": 1676 + }, + { + "epoch": 0.2986111111111111, + "grad_norm": 0.4692424237728119, + "learning_rate": 0.00019732227202049025, + "loss": 1.0793, + "step": 1677 + }, + { + "epoch": 0.2987891737891738, + "grad_norm": 0.5925022959709167, + "learning_rate": 0.00019731905354975778, + "loss": 1.0297, + "step": 1678 + }, + { + "epoch": 0.29896723646723644, + "grad_norm": 0.44047990441322327, + "learning_rate": 0.00019731583317226056, + "loss": 1.0982, + "step": 1679 + }, + { + "epoch": 0.29914529914529914, + "grad_norm": 0.5863066911697388, + "learning_rate": 0.0001973126108880618, + "loss": 1.0284, + "step": 1680 + }, + { + "epoch": 0.29932336182336183, + "grad_norm": 0.48962152004241943, + "learning_rate": 0.00019730938669722457, + "loss": 1.1861, + "step": 1681 + }, + { + "epoch": 0.29950142450142453, + "grad_norm": 0.5445577502250671, + "learning_rate": 0.00019730616059981205, + "loss": 1.2574, + "step": 1682 + }, + { + "epoch": 0.29967948717948717, + "grad_norm": 0.49327564239501953, + "learning_rate": 0.00019730293259588743, + "loss": 0.9578, + "step": 1683 + }, + { + "epoch": 0.29985754985754987, + "grad_norm": 0.4252840578556061, + "learning_rate": 0.00019729970268551398, + "loss": 1.0083, + "step": 1684 + }, + { + "epoch": 0.30003561253561256, + "grad_norm": 0.5140926241874695, + "learning_rate": 0.000197296470868755, + "loss": 1.3263, + "step": 1685 + }, + { + "epoch": 0.3002136752136752, + "grad_norm": 0.5143948197364807, + "learning_rate": 0.00019729323714567375, + "loss": 1.0424, + "step": 1686 + }, + { + "epoch": 0.3003917378917379, + "grad_norm": 0.3811354339122772, + "learning_rate": 0.00019729000151633367, + "loss": 0.6319, + "step": 1687 + }, + { + "epoch": 0.3005698005698006, + "grad_norm": 0.5249716639518738, + "learning_rate": 0.0001972867639807981, + "loss": 1.0173, + "step": 1688 + }, + { + "epoch": 0.30074786324786323, + "grad_norm": 0.41832098364830017, + "learning_rate": 0.00019728352453913048, + "loss": 1.0503, + "step": 1689 + }, + { + "epoch": 0.30092592592592593, + "grad_norm": 0.5961149334907532, + "learning_rate": 0.00019728028319139428, + "loss": 1.1843, + "step": 1690 + }, + { + "epoch": 0.3011039886039886, + "grad_norm": 0.44083690643310547, + "learning_rate": 0.00019727703993765303, + "loss": 1.1311, + "step": 1691 + }, + { + "epoch": 0.30128205128205127, + "grad_norm": 0.4368111491203308, + "learning_rate": 0.00019727379477797022, + "loss": 0.9463, + "step": 1692 + }, + { + "epoch": 0.30146011396011396, + "grad_norm": 0.5289376974105835, + "learning_rate": 0.00019727054771240954, + "loss": 0.9836, + "step": 1693 + }, + { + "epoch": 0.30163817663817666, + "grad_norm": 0.4132843613624573, + "learning_rate": 0.00019726729874103448, + "loss": 1.1052, + "step": 1694 + }, + { + "epoch": 0.3018162393162393, + "grad_norm": 0.4919086992740631, + "learning_rate": 0.00019726404786390877, + "loss": 1.2219, + "step": 1695 + }, + { + "epoch": 0.301994301994302, + "grad_norm": 0.42561691999435425, + "learning_rate": 0.0001972607950810961, + "loss": 1.0756, + "step": 1696 + }, + { + "epoch": 0.3021723646723647, + "grad_norm": 0.5030396580696106, + "learning_rate": 0.0001972575403926602, + "loss": 1.2207, + "step": 1697 + }, + { + "epoch": 0.30235042735042733, + "grad_norm": 0.4779801666736603, + "learning_rate": 0.0001972542837986648, + "loss": 1.194, + "step": 1698 + }, + { + "epoch": 0.30252849002849, + "grad_norm": 0.45395568013191223, + "learning_rate": 0.00019725102529917377, + "loss": 1.0775, + "step": 1699 + }, + { + "epoch": 0.3027065527065527, + "grad_norm": 0.6540699005126953, + "learning_rate": 0.0001972477648942509, + "loss": 1.181, + "step": 1700 + }, + { + "epoch": 0.30288461538461536, + "grad_norm": 0.46281275153160095, + "learning_rate": 0.00019724450258396008, + "loss": 0.629, + "step": 1701 + }, + { + "epoch": 0.30306267806267806, + "grad_norm": 0.3452845811843872, + "learning_rate": 0.00019724123836836527, + "loss": 0.51, + "step": 1702 + }, + { + "epoch": 0.30324074074074076, + "grad_norm": 0.4507991671562195, + "learning_rate": 0.00019723797224753038, + "loss": 1.0258, + "step": 1703 + }, + { + "epoch": 0.3034188034188034, + "grad_norm": 0.5385412573814392, + "learning_rate": 0.0001972347042215194, + "loss": 1.0232, + "step": 1704 + }, + { + "epoch": 0.3035968660968661, + "grad_norm": 0.4460466504096985, + "learning_rate": 0.00019723143429039642, + "loss": 1.1307, + "step": 1705 + }, + { + "epoch": 0.3037749287749288, + "grad_norm": 0.5229718685150146, + "learning_rate": 0.00019722816245422545, + "loss": 1.0964, + "step": 1706 + }, + { + "epoch": 0.30395299145299143, + "grad_norm": 0.4776979088783264, + "learning_rate": 0.00019722488871307058, + "loss": 1.2678, + "step": 1707 + }, + { + "epoch": 0.3041310541310541, + "grad_norm": 0.5371831655502319, + "learning_rate": 0.00019722161306699601, + "loss": 1.2808, + "step": 1708 + }, + { + "epoch": 0.3043091168091168, + "grad_norm": 0.45322108268737793, + "learning_rate": 0.0001972183355160659, + "loss": 1.0775, + "step": 1709 + }, + { + "epoch": 0.30448717948717946, + "grad_norm": 0.5036569833755493, + "learning_rate": 0.00019721505606034448, + "loss": 1.1859, + "step": 1710 + }, + { + "epoch": 0.30466524216524216, + "grad_norm": 0.5425969958305359, + "learning_rate": 0.00019721177469989593, + "loss": 1.0173, + "step": 1711 + }, + { + "epoch": 0.30484330484330485, + "grad_norm": 0.5638980269432068, + "learning_rate": 0.00019720849143478462, + "loss": 1.182, + "step": 1712 + }, + { + "epoch": 0.30502136752136755, + "grad_norm": 0.5160546898841858, + "learning_rate": 0.00019720520626507486, + "loss": 0.9853, + "step": 1713 + }, + { + "epoch": 0.3051994301994302, + "grad_norm": 0.5079004168510437, + "learning_rate": 0.000197201919190831, + "loss": 1.3154, + "step": 1714 + }, + { + "epoch": 0.3053774928774929, + "grad_norm": 0.4590355455875397, + "learning_rate": 0.00019719863021211745, + "loss": 1.007, + "step": 1715 + }, + { + "epoch": 0.3055555555555556, + "grad_norm": 0.49656423926353455, + "learning_rate": 0.00019719533932899865, + "loss": 1.2187, + "step": 1716 + }, + { + "epoch": 0.3057336182336182, + "grad_norm": 0.46426209807395935, + "learning_rate": 0.0001971920465415391, + "loss": 1.3007, + "step": 1717 + }, + { + "epoch": 0.3059116809116809, + "grad_norm": 0.5211917757987976, + "learning_rate": 0.00019718875184980328, + "loss": 1.2256, + "step": 1718 + }, + { + "epoch": 0.3060897435897436, + "grad_norm": 0.42953309416770935, + "learning_rate": 0.00019718545525385578, + "loss": 1.2838, + "step": 1719 + }, + { + "epoch": 0.30626780626780625, + "grad_norm": 0.4893105924129486, + "learning_rate": 0.00019718215675376116, + "loss": 1.052, + "step": 1720 + }, + { + "epoch": 0.30644586894586895, + "grad_norm": 0.4833602011203766, + "learning_rate": 0.00019717885634958405, + "loss": 1.069, + "step": 1721 + }, + { + "epoch": 0.30662393162393164, + "grad_norm": 0.502176821231842, + "learning_rate": 0.0001971755540413891, + "loss": 1.1659, + "step": 1722 + }, + { + "epoch": 0.3068019943019943, + "grad_norm": 0.4648856818675995, + "learning_rate": 0.00019717224982924108, + "loss": 1.1873, + "step": 1723 + }, + { + "epoch": 0.306980056980057, + "grad_norm": 0.405429869890213, + "learning_rate": 0.00019716894371320465, + "loss": 0.99, + "step": 1724 + }, + { + "epoch": 0.3071581196581197, + "grad_norm": 0.4306945204734802, + "learning_rate": 0.00019716563569334463, + "loss": 0.8751, + "step": 1725 + }, + { + "epoch": 0.3073361823361823, + "grad_norm": 0.49424824118614197, + "learning_rate": 0.00019716232576972583, + "loss": 0.9205, + "step": 1726 + }, + { + "epoch": 0.307514245014245, + "grad_norm": 0.5044034123420715, + "learning_rate": 0.00019715901394241306, + "loss": 1.2042, + "step": 1727 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 0.512180507183075, + "learning_rate": 0.00019715570021147126, + "loss": 1.1644, + "step": 1728 + }, + { + "epoch": 0.30787037037037035, + "grad_norm": 0.4377981126308441, + "learning_rate": 0.00019715238457696538, + "loss": 1.1625, + "step": 1729 + }, + { + "epoch": 0.30804843304843305, + "grad_norm": 0.49107855558395386, + "learning_rate": 0.00019714906703896027, + "loss": 1.1037, + "step": 1730 + }, + { + "epoch": 0.30822649572649574, + "grad_norm": 0.47342559695243835, + "learning_rate": 0.00019714574759752105, + "loss": 1.3186, + "step": 1731 + }, + { + "epoch": 0.3084045584045584, + "grad_norm": 0.487177312374115, + "learning_rate": 0.0001971424262527127, + "loss": 1.1196, + "step": 1732 + }, + { + "epoch": 0.3085826210826211, + "grad_norm": 0.5290025472640991, + "learning_rate": 0.0001971391030046003, + "loss": 1.2103, + "step": 1733 + }, + { + "epoch": 0.3087606837606838, + "grad_norm": 0.4587760269641876, + "learning_rate": 0.00019713577785324896, + "loss": 1.1017, + "step": 1734 + }, + { + "epoch": 0.3089387464387464, + "grad_norm": 0.45323294401168823, + "learning_rate": 0.00019713245079872388, + "loss": 1.0, + "step": 1735 + }, + { + "epoch": 0.3091168091168091, + "grad_norm": 0.43414804339408875, + "learning_rate": 0.00019712912184109013, + "loss": 1.0341, + "step": 1736 + }, + { + "epoch": 0.3092948717948718, + "grad_norm": 0.49604663252830505, + "learning_rate": 0.00019712579098041304, + "loss": 0.9437, + "step": 1737 + }, + { + "epoch": 0.30947293447293445, + "grad_norm": 0.48580703139305115, + "learning_rate": 0.00019712245821675785, + "loss": 1.2622, + "step": 1738 + }, + { + "epoch": 0.30965099715099714, + "grad_norm": 0.45333603024482727, + "learning_rate": 0.00019711912355018982, + "loss": 1.2063, + "step": 1739 + }, + { + "epoch": 0.30982905982905984, + "grad_norm": 0.5990764498710632, + "learning_rate": 0.00019711578698077432, + "loss": 1.5097, + "step": 1740 + }, + { + "epoch": 0.31000712250712253, + "grad_norm": 0.4386102259159088, + "learning_rate": 0.0001971124485085767, + "loss": 1.1283, + "step": 1741 + }, + { + "epoch": 0.3101851851851852, + "grad_norm": 0.4476035237312317, + "learning_rate": 0.00019710910813366242, + "loss": 0.8922, + "step": 1742 + }, + { + "epoch": 0.31036324786324787, + "grad_norm": 0.5276228785514832, + "learning_rate": 0.00019710576585609685, + "loss": 1.2373, + "step": 1743 + }, + { + "epoch": 0.31054131054131057, + "grad_norm": 0.4885637164115906, + "learning_rate": 0.00019710242167594557, + "loss": 1.0881, + "step": 1744 + }, + { + "epoch": 0.3107193732193732, + "grad_norm": 0.421132355928421, + "learning_rate": 0.000197099075593274, + "loss": 1.0544, + "step": 1745 + }, + { + "epoch": 0.3108974358974359, + "grad_norm": 0.5257927179336548, + "learning_rate": 0.00019709572760814777, + "loss": 1.265, + "step": 1746 + }, + { + "epoch": 0.3110754985754986, + "grad_norm": 0.5164850950241089, + "learning_rate": 0.00019709237772063247, + "loss": 0.9593, + "step": 1747 + }, + { + "epoch": 0.31125356125356124, + "grad_norm": 0.5176383256912231, + "learning_rate": 0.00019708902593079374, + "loss": 1.0194, + "step": 1748 + }, + { + "epoch": 0.31143162393162394, + "grad_norm": 0.4620790481567383, + "learning_rate": 0.00019708567223869716, + "loss": 0.9241, + "step": 1749 + }, + { + "epoch": 0.31160968660968663, + "grad_norm": 0.48307979106903076, + "learning_rate": 0.00019708231664440854, + "loss": 1.2314, + "step": 1750 + }, + { + "epoch": 0.31178774928774927, + "grad_norm": 0.4931468069553375, + "learning_rate": 0.00019707895914799364, + "loss": 1.2065, + "step": 1751 + }, + { + "epoch": 0.31196581196581197, + "grad_norm": 0.5035979747772217, + "learning_rate": 0.00019707559974951818, + "loss": 1.1867, + "step": 1752 + }, + { + "epoch": 0.31214387464387466, + "grad_norm": 0.47543632984161377, + "learning_rate": 0.00019707223844904795, + "loss": 1.0603, + "step": 1753 + }, + { + "epoch": 0.3123219373219373, + "grad_norm": 0.49929797649383545, + "learning_rate": 0.00019706887524664892, + "loss": 1.0597, + "step": 1754 + }, + { + "epoch": 0.3125, + "grad_norm": 0.5075222253799438, + "learning_rate": 0.00019706551014238687, + "loss": 1.1398, + "step": 1755 + }, + { + "epoch": 0.3126780626780627, + "grad_norm": 0.5096884369850159, + "learning_rate": 0.00019706214313632784, + "loss": 1.1382, + "step": 1756 + }, + { + "epoch": 0.31285612535612534, + "grad_norm": 0.4629988372325897, + "learning_rate": 0.0001970587742285377, + "loss": 1.0009, + "step": 1757 + }, + { + "epoch": 0.31303418803418803, + "grad_norm": 0.5244084596633911, + "learning_rate": 0.00019705540341908253, + "loss": 1.047, + "step": 1758 + }, + { + "epoch": 0.31321225071225073, + "grad_norm": 0.5136716961860657, + "learning_rate": 0.00019705203070802832, + "loss": 1.29, + "step": 1759 + }, + { + "epoch": 0.31339031339031337, + "grad_norm": 0.43991541862487793, + "learning_rate": 0.0001970486560954412, + "loss": 0.9605, + "step": 1760 + }, + { + "epoch": 0.31356837606837606, + "grad_norm": 0.4633477032184601, + "learning_rate": 0.00019704527958138725, + "loss": 1.1507, + "step": 1761 + }, + { + "epoch": 0.31374643874643876, + "grad_norm": 0.4419999420642853, + "learning_rate": 0.00019704190116593266, + "loss": 0.9262, + "step": 1762 + }, + { + "epoch": 0.3139245014245014, + "grad_norm": 0.49359434843063354, + "learning_rate": 0.00019703852084914357, + "loss": 0.9348, + "step": 1763 + }, + { + "epoch": 0.3141025641025641, + "grad_norm": 0.5072139501571655, + "learning_rate": 0.00019703513863108627, + "loss": 1.1592, + "step": 1764 + }, + { + "epoch": 0.3142806267806268, + "grad_norm": 0.45969831943511963, + "learning_rate": 0.00019703175451182698, + "loss": 1.1519, + "step": 1765 + }, + { + "epoch": 0.31445868945868943, + "grad_norm": 0.5148758292198181, + "learning_rate": 0.00019702836849143208, + "loss": 1.1673, + "step": 1766 + }, + { + "epoch": 0.31463675213675213, + "grad_norm": 0.43033209443092346, + "learning_rate": 0.0001970249805699678, + "loss": 0.9256, + "step": 1767 + }, + { + "epoch": 0.3148148148148148, + "grad_norm": 0.48143425583839417, + "learning_rate": 0.00019702159074750058, + "loss": 1.08, + "step": 1768 + }, + { + "epoch": 0.31499287749287747, + "grad_norm": 0.4780619740486145, + "learning_rate": 0.00019701819902409685, + "loss": 1.1198, + "step": 1769 + }, + { + "epoch": 0.31517094017094016, + "grad_norm": 0.4662075936794281, + "learning_rate": 0.00019701480539982305, + "loss": 0.8424, + "step": 1770 + }, + { + "epoch": 0.31534900284900286, + "grad_norm": 0.503901481628418, + "learning_rate": 0.00019701140987474566, + "loss": 1.1026, + "step": 1771 + }, + { + "epoch": 0.31552706552706555, + "grad_norm": 0.5197132229804993, + "learning_rate": 0.00019700801244893124, + "loss": 1.2148, + "step": 1772 + }, + { + "epoch": 0.3157051282051282, + "grad_norm": 0.4746309220790863, + "learning_rate": 0.00019700461312244634, + "loss": 1.0906, + "step": 1773 + }, + { + "epoch": 0.3158831908831909, + "grad_norm": 0.5277339816093445, + "learning_rate": 0.00019700121189535752, + "loss": 1.0588, + "step": 1774 + }, + { + "epoch": 0.3160612535612536, + "grad_norm": 0.436002254486084, + "learning_rate": 0.00019699780876773147, + "loss": 1.0341, + "step": 1775 + }, + { + "epoch": 0.3162393162393162, + "grad_norm": 0.5171145796775818, + "learning_rate": 0.00019699440373963486, + "loss": 1.282, + "step": 1776 + }, + { + "epoch": 0.3164173789173789, + "grad_norm": 0.38382846117019653, + "learning_rate": 0.00019699099681113436, + "loss": 0.8908, + "step": 1777 + }, + { + "epoch": 0.3165954415954416, + "grad_norm": 0.4621630609035492, + "learning_rate": 0.0001969875879822968, + "loss": 1.1074, + "step": 1778 + }, + { + "epoch": 0.31677350427350426, + "grad_norm": 0.5543130040168762, + "learning_rate": 0.00019698417725318892, + "loss": 0.9682, + "step": 1779 + }, + { + "epoch": 0.31695156695156695, + "grad_norm": 0.49534836411476135, + "learning_rate": 0.00019698076462387753, + "loss": 1.107, + "step": 1780 + }, + { + "epoch": 0.31712962962962965, + "grad_norm": 0.48844948410987854, + "learning_rate": 0.00019697735009442956, + "loss": 1.1295, + "step": 1781 + }, + { + "epoch": 0.3173076923076923, + "grad_norm": 0.5070686936378479, + "learning_rate": 0.00019697393366491185, + "loss": 1.083, + "step": 1782 + }, + { + "epoch": 0.317485754985755, + "grad_norm": 0.47817620635032654, + "learning_rate": 0.00019697051533539134, + "loss": 1.3014, + "step": 1783 + }, + { + "epoch": 0.3176638176638177, + "grad_norm": 0.538488507270813, + "learning_rate": 0.00019696709510593502, + "loss": 1.0354, + "step": 1784 + }, + { + "epoch": 0.3178418803418803, + "grad_norm": 0.5141439437866211, + "learning_rate": 0.0001969636729766099, + "loss": 1.2912, + "step": 1785 + }, + { + "epoch": 0.318019943019943, + "grad_norm": 0.5009665489196777, + "learning_rate": 0.00019696024894748306, + "loss": 0.9014, + "step": 1786 + }, + { + "epoch": 0.3181980056980057, + "grad_norm": 0.46199744939804077, + "learning_rate": 0.00019695682301862155, + "loss": 1.0532, + "step": 1787 + }, + { + "epoch": 0.31837606837606836, + "grad_norm": 0.4649423062801361, + "learning_rate": 0.0001969533951900925, + "loss": 0.8608, + "step": 1788 + }, + { + "epoch": 0.31855413105413105, + "grad_norm": 0.516909658908844, + "learning_rate": 0.0001969499654619631, + "loss": 1.1385, + "step": 1789 + }, + { + "epoch": 0.31873219373219375, + "grad_norm": 0.46016669273376465, + "learning_rate": 0.00019694653383430048, + "loss": 0.9168, + "step": 1790 + }, + { + "epoch": 0.3189102564102564, + "grad_norm": 0.4794938564300537, + "learning_rate": 0.00019694310030717193, + "loss": 1.0244, + "step": 1791 + }, + { + "epoch": 0.3190883190883191, + "grad_norm": 0.46577662229537964, + "learning_rate": 0.00019693966488064471, + "loss": 1.0954, + "step": 1792 + }, + { + "epoch": 0.3192663817663818, + "grad_norm": 0.4866746962070465, + "learning_rate": 0.00019693622755478614, + "loss": 1.2925, + "step": 1793 + }, + { + "epoch": 0.3194444444444444, + "grad_norm": 0.4841702878475189, + "learning_rate": 0.00019693278832966357, + "loss": 1.119, + "step": 1794 + }, + { + "epoch": 0.3196225071225071, + "grad_norm": 0.4835243821144104, + "learning_rate": 0.00019692934720534435, + "loss": 1.1702, + "step": 1795 + }, + { + "epoch": 0.3198005698005698, + "grad_norm": 0.5200608968734741, + "learning_rate": 0.00019692590418189594, + "loss": 1.1989, + "step": 1796 + }, + { + "epoch": 0.31997863247863245, + "grad_norm": 0.5147821307182312, + "learning_rate": 0.00019692245925938577, + "loss": 1.1417, + "step": 1797 + }, + { + "epoch": 0.32015669515669515, + "grad_norm": 0.5145614743232727, + "learning_rate": 0.00019691901243788136, + "loss": 1.0571, + "step": 1798 + }, + { + "epoch": 0.32033475783475784, + "grad_norm": 0.5416026711463928, + "learning_rate": 0.00019691556371745022, + "loss": 1.188, + "step": 1799 + }, + { + "epoch": 0.32051282051282054, + "grad_norm": 0.5140644311904907, + "learning_rate": 0.00019691211309815995, + "loss": 1.1795, + "step": 1800 + }, + { + "epoch": 0.3206908831908832, + "grad_norm": 0.44219106435775757, + "learning_rate": 0.00019690866058007817, + "loss": 0.9215, + "step": 1801 + }, + { + "epoch": 0.3208689458689459, + "grad_norm": 0.49523603916168213, + "learning_rate": 0.00019690520616327245, + "loss": 1.1117, + "step": 1802 + }, + { + "epoch": 0.32104700854700857, + "grad_norm": 0.5818293690681458, + "learning_rate": 0.0001969017498478105, + "loss": 1.16, + "step": 1803 + }, + { + "epoch": 0.3212250712250712, + "grad_norm": 0.5175749659538269, + "learning_rate": 0.0001968982916337601, + "loss": 1.1999, + "step": 1804 + }, + { + "epoch": 0.3214031339031339, + "grad_norm": 0.49916017055511475, + "learning_rate": 0.00019689483152118898, + "loss": 0.9505, + "step": 1805 + }, + { + "epoch": 0.3215811965811966, + "grad_norm": 0.46849536895751953, + "learning_rate": 0.00019689136951016488, + "loss": 0.9627, + "step": 1806 + }, + { + "epoch": 0.32175925925925924, + "grad_norm": 0.4226818382740021, + "learning_rate": 0.00019688790560075568, + "loss": 1.037, + "step": 1807 + }, + { + "epoch": 0.32193732193732194, + "grad_norm": 0.4697103798389435, + "learning_rate": 0.00019688443979302923, + "loss": 1.1431, + "step": 1808 + }, + { + "epoch": 0.32211538461538464, + "grad_norm": 0.4999365508556366, + "learning_rate": 0.00019688097208705343, + "loss": 1.171, + "step": 1809 + }, + { + "epoch": 0.3222934472934473, + "grad_norm": 0.5229731798171997, + "learning_rate": 0.00019687750248289625, + "loss": 1.3395, + "step": 1810 + }, + { + "epoch": 0.32247150997151, + "grad_norm": 0.512525737285614, + "learning_rate": 0.00019687403098062566, + "loss": 1.1438, + "step": 1811 + }, + { + "epoch": 0.32264957264957267, + "grad_norm": 0.4558548927307129, + "learning_rate": 0.00019687055758030967, + "loss": 1.0012, + "step": 1812 + }, + { + "epoch": 0.3228276353276353, + "grad_norm": 0.45195743441581726, + "learning_rate": 0.00019686708228201636, + "loss": 1.0222, + "step": 1813 + }, + { + "epoch": 0.323005698005698, + "grad_norm": 0.5023126602172852, + "learning_rate": 0.00019686360508581373, + "loss": 1.2128, + "step": 1814 + }, + { + "epoch": 0.3231837606837607, + "grad_norm": 0.46516045928001404, + "learning_rate": 0.00019686012599177003, + "loss": 0.989, + "step": 1815 + }, + { + "epoch": 0.32336182336182334, + "grad_norm": 0.4142672121524811, + "learning_rate": 0.00019685664499995338, + "loss": 1.0144, + "step": 1816 + }, + { + "epoch": 0.32353988603988604, + "grad_norm": 0.4511009752750397, + "learning_rate": 0.0001968531621104319, + "loss": 0.885, + "step": 1817 + }, + { + "epoch": 0.32371794871794873, + "grad_norm": 0.49583545327186584, + "learning_rate": 0.00019684967732327396, + "loss": 1.0986, + "step": 1818 + }, + { + "epoch": 0.3238960113960114, + "grad_norm": 0.5872161388397217, + "learning_rate": 0.0001968461906385478, + "loss": 1.1482, + "step": 1819 + }, + { + "epoch": 0.32407407407407407, + "grad_norm": 0.4509563148021698, + "learning_rate": 0.00019684270205632168, + "loss": 1.0578, + "step": 1820 + }, + { + "epoch": 0.32425213675213677, + "grad_norm": 0.501345157623291, + "learning_rate": 0.00019683921157666402, + "loss": 1.1792, + "step": 1821 + }, + { + "epoch": 0.3244301994301994, + "grad_norm": 0.48257577419281006, + "learning_rate": 0.00019683571919964314, + "loss": 1.0448, + "step": 1822 + }, + { + "epoch": 0.3246082621082621, + "grad_norm": 0.5399422645568848, + "learning_rate": 0.00019683222492532752, + "loss": 1.0579, + "step": 1823 + }, + { + "epoch": 0.3247863247863248, + "grad_norm": 0.4382506012916565, + "learning_rate": 0.0001968287287537856, + "loss": 1.0246, + "step": 1824 + }, + { + "epoch": 0.32496438746438744, + "grad_norm": 0.49247491359710693, + "learning_rate": 0.00019682523068508586, + "loss": 1.318, + "step": 1825 + }, + { + "epoch": 0.32514245014245013, + "grad_norm": 0.49067625403404236, + "learning_rate": 0.0001968217307192969, + "loss": 1.1028, + "step": 1826 + }, + { + "epoch": 0.32532051282051283, + "grad_norm": 0.4832286238670349, + "learning_rate": 0.00019681822885648723, + "loss": 1.0996, + "step": 1827 + }, + { + "epoch": 0.32549857549857547, + "grad_norm": 0.47144386172294617, + "learning_rate": 0.0001968147250967255, + "loss": 1.0707, + "step": 1828 + }, + { + "epoch": 0.32567663817663817, + "grad_norm": 0.46299225091934204, + "learning_rate": 0.0001968112194400803, + "loss": 1.0461, + "step": 1829 + }, + { + "epoch": 0.32585470085470086, + "grad_norm": 0.4880816340446472, + "learning_rate": 0.00019680771188662044, + "loss": 1.1198, + "step": 1830 + }, + { + "epoch": 0.32603276353276356, + "grad_norm": 0.43837276101112366, + "learning_rate": 0.00019680420243641452, + "loss": 1.0599, + "step": 1831 + }, + { + "epoch": 0.3262108262108262, + "grad_norm": 0.453168660402298, + "learning_rate": 0.0001968006910895314, + "loss": 1.0327, + "step": 1832 + }, + { + "epoch": 0.3263888888888889, + "grad_norm": 0.45183828473091125, + "learning_rate": 0.00019679717784603975, + "loss": 1.1381, + "step": 1833 + }, + { + "epoch": 0.3265669515669516, + "grad_norm": 0.5326765775680542, + "learning_rate": 0.00019679366270600852, + "loss": 1.3169, + "step": 1834 + }, + { + "epoch": 0.32674501424501423, + "grad_norm": 0.47468429803848267, + "learning_rate": 0.00019679014566950653, + "loss": 1.1816, + "step": 1835 + }, + { + "epoch": 0.3269230769230769, + "grad_norm": 0.5096879005432129, + "learning_rate": 0.0001967866267366027, + "loss": 1.1162, + "step": 1836 + }, + { + "epoch": 0.3271011396011396, + "grad_norm": 0.491514652967453, + "learning_rate": 0.00019678310590736598, + "loss": 1.2793, + "step": 1837 + }, + { + "epoch": 0.32727920227920226, + "grad_norm": 0.601439356803894, + "learning_rate": 0.00019677958318186533, + "loss": 0.9851, + "step": 1838 + }, + { + "epoch": 0.32745726495726496, + "grad_norm": 0.45270970463752747, + "learning_rate": 0.0001967760585601698, + "loss": 1.0042, + "step": 1839 + }, + { + "epoch": 0.32763532763532766, + "grad_norm": 0.48864325881004333, + "learning_rate": 0.00019677253204234847, + "loss": 1.0835, + "step": 1840 + }, + { + "epoch": 0.3278133903133903, + "grad_norm": 0.5855685472488403, + "learning_rate": 0.00019676900362847037, + "loss": 1.193, + "step": 1841 + }, + { + "epoch": 0.327991452991453, + "grad_norm": 0.7181013822555542, + "learning_rate": 0.00019676547331860466, + "loss": 1.2028, + "step": 1842 + }, + { + "epoch": 0.3281695156695157, + "grad_norm": 0.4517378807067871, + "learning_rate": 0.00019676194111282054, + "loss": 1.013, + "step": 1843 + }, + { + "epoch": 0.32834757834757833, + "grad_norm": 0.5477756857872009, + "learning_rate": 0.00019675840701118718, + "loss": 1.2311, + "step": 1844 + }, + { + "epoch": 0.328525641025641, + "grad_norm": 0.5194997191429138, + "learning_rate": 0.00019675487101377382, + "loss": 1.0953, + "step": 1845 + }, + { + "epoch": 0.3287037037037037, + "grad_norm": 0.44454067945480347, + "learning_rate": 0.00019675133312064977, + "loss": 0.8505, + "step": 1846 + }, + { + "epoch": 0.32888176638176636, + "grad_norm": 0.3938713073730469, + "learning_rate": 0.00019674779333188428, + "loss": 0.8525, + "step": 1847 + }, + { + "epoch": 0.32905982905982906, + "grad_norm": 0.4927884340286255, + "learning_rate": 0.00019674425164754682, + "loss": 1.2477, + "step": 1848 + }, + { + "epoch": 0.32923789173789175, + "grad_norm": 0.4516635239124298, + "learning_rate": 0.0001967407080677067, + "loss": 0.8333, + "step": 1849 + }, + { + "epoch": 0.3294159544159544, + "grad_norm": 0.47105780243873596, + "learning_rate": 0.00019673716259243336, + "loss": 1.0989, + "step": 1850 + }, + { + "epoch": 0.3295940170940171, + "grad_norm": 0.5192127823829651, + "learning_rate": 0.00019673361522179627, + "loss": 1.1164, + "step": 1851 + }, + { + "epoch": 0.3297720797720798, + "grad_norm": 0.5222696661949158, + "learning_rate": 0.00019673006595586495, + "loss": 1.3191, + "step": 1852 + }, + { + "epoch": 0.3299501424501424, + "grad_norm": 0.6046679019927979, + "learning_rate": 0.0001967265147947089, + "loss": 0.9782, + "step": 1853 + }, + { + "epoch": 0.3301282051282051, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.00019672296173839775, + "loss": 1.2247, + "step": 1854 + }, + { + "epoch": 0.3303062678062678, + "grad_norm": 0.5435982346534729, + "learning_rate": 0.00019671940678700107, + "loss": 1.1647, + "step": 1855 + }, + { + "epoch": 0.33048433048433046, + "grad_norm": 0.46878984570503235, + "learning_rate": 0.00019671584994058856, + "loss": 1.132, + "step": 1856 + }, + { + "epoch": 0.33066239316239315, + "grad_norm": 0.5336877107620239, + "learning_rate": 0.00019671229119922986, + "loss": 1.0583, + "step": 1857 + }, + { + "epoch": 0.33084045584045585, + "grad_norm": 0.4811093807220459, + "learning_rate": 0.0001967087305629947, + "loss": 1.0089, + "step": 1858 + }, + { + "epoch": 0.33101851851851855, + "grad_norm": 0.5140184760093689, + "learning_rate": 0.0001967051680319529, + "loss": 1.2335, + "step": 1859 + }, + { + "epoch": 0.3311965811965812, + "grad_norm": 0.5855883955955505, + "learning_rate": 0.00019670160360617418, + "loss": 1.1107, + "step": 1860 + }, + { + "epoch": 0.3313746438746439, + "grad_norm": 0.5081531405448914, + "learning_rate": 0.00019669803728572844, + "loss": 1.0669, + "step": 1861 + }, + { + "epoch": 0.3315527065527066, + "grad_norm": 0.48749417066574097, + "learning_rate": 0.0001966944690706855, + "loss": 1.1465, + "step": 1862 + }, + { + "epoch": 0.3317307692307692, + "grad_norm": 0.5175687670707703, + "learning_rate": 0.00019669089896111536, + "loss": 1.254, + "step": 1863 + }, + { + "epoch": 0.3319088319088319, + "grad_norm": 0.4198860824108124, + "learning_rate": 0.0001966873269570879, + "loss": 0.9811, + "step": 1864 + }, + { + "epoch": 0.3320868945868946, + "grad_norm": 0.5220273733139038, + "learning_rate": 0.0001966837530586731, + "loss": 1.277, + "step": 1865 + }, + { + "epoch": 0.33226495726495725, + "grad_norm": 0.551954448223114, + "learning_rate": 0.00019668017726594101, + "loss": 1.0627, + "step": 1866 + }, + { + "epoch": 0.33244301994301995, + "grad_norm": 0.5289301872253418, + "learning_rate": 0.00019667659957896166, + "loss": 1.4525, + "step": 1867 + }, + { + "epoch": 0.33262108262108264, + "grad_norm": 0.5190161466598511, + "learning_rate": 0.00019667301999780522, + "loss": 1.1064, + "step": 1868 + }, + { + "epoch": 0.3327991452991453, + "grad_norm": 0.437637060880661, + "learning_rate": 0.00019666943852254172, + "loss": 1.1304, + "step": 1869 + }, + { + "epoch": 0.332977207977208, + "grad_norm": 0.4801286458969116, + "learning_rate": 0.00019666585515324138, + "loss": 1.032, + "step": 1870 + }, + { + "epoch": 0.3331552706552707, + "grad_norm": 0.5041908621788025, + "learning_rate": 0.00019666226988997445, + "loss": 1.2611, + "step": 1871 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.4529375731945038, + "learning_rate": 0.00019665868273281115, + "loss": 1.1346, + "step": 1872 + }, + { + "epoch": 0.333511396011396, + "grad_norm": 0.4797019064426422, + "learning_rate": 0.00019665509368182172, + "loss": 1.1716, + "step": 1873 + }, + { + "epoch": 0.3336894586894587, + "grad_norm": 0.5505055785179138, + "learning_rate": 0.00019665150273707652, + "loss": 0.9729, + "step": 1874 + }, + { + "epoch": 0.33386752136752135, + "grad_norm": 0.4228051006793976, + "learning_rate": 0.00019664790989864592, + "loss": 0.9023, + "step": 1875 + }, + { + "epoch": 0.33404558404558404, + "grad_norm": 0.4926959276199341, + "learning_rate": 0.00019664431516660028, + "loss": 1.0999, + "step": 1876 + }, + { + "epoch": 0.33422364672364674, + "grad_norm": 0.4273219704627991, + "learning_rate": 0.00019664071854101005, + "loss": 1.1039, + "step": 1877 + }, + { + "epoch": 0.3344017094017094, + "grad_norm": 0.48438936471939087, + "learning_rate": 0.00019663712002194566, + "loss": 1.1308, + "step": 1878 + }, + { + "epoch": 0.3345797720797721, + "grad_norm": 0.5102053284645081, + "learning_rate": 0.0001966335196094777, + "loss": 1.0618, + "step": 1879 + }, + { + "epoch": 0.33475783475783477, + "grad_norm": 0.4357300400733948, + "learning_rate": 0.00019662991730367663, + "loss": 1.0521, + "step": 1880 + }, + { + "epoch": 0.3349358974358974, + "grad_norm": 0.5052695870399475, + "learning_rate": 0.00019662631310461308, + "loss": 0.9579, + "step": 1881 + }, + { + "epoch": 0.3351139601139601, + "grad_norm": 0.4889117181301117, + "learning_rate": 0.00019662270701235762, + "loss": 1.0304, + "step": 1882 + }, + { + "epoch": 0.3352920227920228, + "grad_norm": 0.4671195149421692, + "learning_rate": 0.000196619099026981, + "loss": 1.2228, + "step": 1883 + }, + { + "epoch": 0.33547008547008544, + "grad_norm": 0.4700174331665039, + "learning_rate": 0.0001966154891485538, + "loss": 0.9634, + "step": 1884 + }, + { + "epoch": 0.33564814814814814, + "grad_norm": 0.488817423582077, + "learning_rate": 0.00019661187737714676, + "loss": 1.2499, + "step": 1885 + }, + { + "epoch": 0.33582621082621084, + "grad_norm": 0.5336169600486755, + "learning_rate": 0.00019660826371283073, + "loss": 1.251, + "step": 1886 + }, + { + "epoch": 0.33600427350427353, + "grad_norm": 0.5054540038108826, + "learning_rate": 0.00019660464815567642, + "loss": 1.221, + "step": 1887 + }, + { + "epoch": 0.33618233618233617, + "grad_norm": 0.5078747868537903, + "learning_rate": 0.00019660103070575472, + "loss": 0.9792, + "step": 1888 + }, + { + "epoch": 0.33636039886039887, + "grad_norm": 0.498571515083313, + "learning_rate": 0.0001965974113631365, + "loss": 1.1682, + "step": 1889 + }, + { + "epoch": 0.33653846153846156, + "grad_norm": 0.49969518184661865, + "learning_rate": 0.00019659379012789264, + "loss": 1.0012, + "step": 1890 + }, + { + "epoch": 0.3367165242165242, + "grad_norm": 0.4238094687461853, + "learning_rate": 0.00019659016700009416, + "loss": 1.0455, + "step": 1891 + }, + { + "epoch": 0.3368945868945869, + "grad_norm": 0.5139104723930359, + "learning_rate": 0.000196586541979812, + "loss": 0.9979, + "step": 1892 + }, + { + "epoch": 0.3370726495726496, + "grad_norm": 0.5446547269821167, + "learning_rate": 0.00019658291506711715, + "loss": 0.9271, + "step": 1893 + }, + { + "epoch": 0.33725071225071224, + "grad_norm": 0.5284572839736938, + "learning_rate": 0.00019657928626208077, + "loss": 1.0356, + "step": 1894 + }, + { + "epoch": 0.33742877492877493, + "grad_norm": 0.49936217069625854, + "learning_rate": 0.00019657565556477387, + "loss": 0.9785, + "step": 1895 + }, + { + "epoch": 0.33760683760683763, + "grad_norm": 0.4678729772567749, + "learning_rate": 0.00019657202297526763, + "loss": 1.2135, + "step": 1896 + }, + { + "epoch": 0.33778490028490027, + "grad_norm": 0.46844249963760376, + "learning_rate": 0.0001965683884936332, + "loss": 0.9369, + "step": 1897 + }, + { + "epoch": 0.33796296296296297, + "grad_norm": 0.4307389557361603, + "learning_rate": 0.0001965647521199418, + "loss": 0.9301, + "step": 1898 + }, + { + "epoch": 0.33814102564102566, + "grad_norm": 0.48227834701538086, + "learning_rate": 0.00019656111385426468, + "loss": 1.3169, + "step": 1899 + }, + { + "epoch": 0.3383190883190883, + "grad_norm": 0.45860713720321655, + "learning_rate": 0.00019655747369667315, + "loss": 0.9835, + "step": 1900 + }, + { + "epoch": 0.338497150997151, + "grad_norm": 0.5522414445877075, + "learning_rate": 0.00019655383164723846, + "loss": 1.363, + "step": 1901 + }, + { + "epoch": 0.3386752136752137, + "grad_norm": 0.5283710360527039, + "learning_rate": 0.000196550187706032, + "loss": 1.1499, + "step": 1902 + }, + { + "epoch": 0.33885327635327633, + "grad_norm": 0.4419134259223938, + "learning_rate": 0.00019654654187312525, + "loss": 1.2039, + "step": 1903 + }, + { + "epoch": 0.33903133903133903, + "grad_norm": 0.49066096544265747, + "learning_rate": 0.00019654289414858952, + "loss": 0.9707, + "step": 1904 + }, + { + "epoch": 0.3392094017094017, + "grad_norm": 0.4619338810443878, + "learning_rate": 0.00019653924453249633, + "loss": 1.0849, + "step": 1905 + }, + { + "epoch": 0.33938746438746437, + "grad_norm": 0.5191119313240051, + "learning_rate": 0.0001965355930249172, + "loss": 1.1387, + "step": 1906 + }, + { + "epoch": 0.33956552706552706, + "grad_norm": 0.5245711207389832, + "learning_rate": 0.00019653193962592368, + "loss": 1.3435, + "step": 1907 + }, + { + "epoch": 0.33974358974358976, + "grad_norm": 0.49562904238700867, + "learning_rate": 0.0001965282843355873, + "loss": 1.2781, + "step": 1908 + }, + { + "epoch": 0.3399216524216524, + "grad_norm": 0.4661353826522827, + "learning_rate": 0.0001965246271539797, + "loss": 0.9317, + "step": 1909 + }, + { + "epoch": 0.3400997150997151, + "grad_norm": 0.4723222851753235, + "learning_rate": 0.00019652096808117254, + "loss": 1.0733, + "step": 1910 + }, + { + "epoch": 0.3402777777777778, + "grad_norm": 0.4358505308628082, + "learning_rate": 0.00019651730711723754, + "loss": 1.1461, + "step": 1911 + }, + { + "epoch": 0.34045584045584043, + "grad_norm": 0.462422251701355, + "learning_rate": 0.00019651364426224638, + "loss": 1.0914, + "step": 1912 + }, + { + "epoch": 0.3406339031339031, + "grad_norm": 0.47952914237976074, + "learning_rate": 0.0001965099795162709, + "loss": 1.0392, + "step": 1913 + }, + { + "epoch": 0.3408119658119658, + "grad_norm": 0.5036373734474182, + "learning_rate": 0.00019650631287938282, + "loss": 1.4002, + "step": 1914 + }, + { + "epoch": 0.34099002849002846, + "grad_norm": 0.5130090713500977, + "learning_rate": 0.000196502644351654, + "loss": 1.3499, + "step": 1915 + }, + { + "epoch": 0.34116809116809116, + "grad_norm": 0.4426332414150238, + "learning_rate": 0.00019649897393315635, + "loss": 1.0726, + "step": 1916 + }, + { + "epoch": 0.34134615384615385, + "grad_norm": 0.5580727458000183, + "learning_rate": 0.00019649530162396176, + "loss": 1.1164, + "step": 1917 + }, + { + "epoch": 0.34152421652421655, + "grad_norm": 0.545001745223999, + "learning_rate": 0.00019649162742414218, + "loss": 0.962, + "step": 1918 + }, + { + "epoch": 0.3417022792022792, + "grad_norm": 0.5225808024406433, + "learning_rate": 0.00019648795133376962, + "loss": 1.1415, + "step": 1919 + }, + { + "epoch": 0.3418803418803419, + "grad_norm": 0.48210129141807556, + "learning_rate": 0.0001964842733529161, + "loss": 1.1188, + "step": 1920 + }, + { + "epoch": 0.3420584045584046, + "grad_norm": 0.4515395164489746, + "learning_rate": 0.00019648059348165365, + "loss": 1.0828, + "step": 1921 + }, + { + "epoch": 0.3422364672364672, + "grad_norm": 0.5802633166313171, + "learning_rate": 0.0001964769117200544, + "loss": 1.3137, + "step": 1922 + }, + { + "epoch": 0.3424145299145299, + "grad_norm": 0.4432032108306885, + "learning_rate": 0.00019647322806819046, + "loss": 1.0523, + "step": 1923 + }, + { + "epoch": 0.3425925925925926, + "grad_norm": 0.4697614908218384, + "learning_rate": 0.00019646954252613402, + "loss": 0.8426, + "step": 1924 + }, + { + "epoch": 0.34277065527065526, + "grad_norm": 0.4610968232154846, + "learning_rate": 0.0001964658550939573, + "loss": 0.9826, + "step": 1925 + }, + { + "epoch": 0.34294871794871795, + "grad_norm": 0.5278257727622986, + "learning_rate": 0.00019646216577173258, + "loss": 1.1064, + "step": 1926 + }, + { + "epoch": 0.34312678062678065, + "grad_norm": 0.5686144232749939, + "learning_rate": 0.00019645847455953205, + "loss": 0.9138, + "step": 1927 + }, + { + "epoch": 0.3433048433048433, + "grad_norm": 0.42894792556762695, + "learning_rate": 0.0001964547814574281, + "loss": 1.0461, + "step": 1928 + }, + { + "epoch": 0.343482905982906, + "grad_norm": 0.5567317605018616, + "learning_rate": 0.0001964510864654931, + "loss": 0.8787, + "step": 1929 + }, + { + "epoch": 0.3436609686609687, + "grad_norm": 0.5015586614608765, + "learning_rate": 0.0001964473895837994, + "loss": 1.1406, + "step": 1930 + }, + { + "epoch": 0.3438390313390313, + "grad_norm": 0.47391530871391296, + "learning_rate": 0.00019644369081241948, + "loss": 1.0685, + "step": 1931 + }, + { + "epoch": 0.344017094017094, + "grad_norm": 0.546037495136261, + "learning_rate": 0.00019643999015142574, + "loss": 1.2349, + "step": 1932 + }, + { + "epoch": 0.3441951566951567, + "grad_norm": 0.4724953770637512, + "learning_rate": 0.00019643628760089078, + "loss": 1.0621, + "step": 1933 + }, + { + "epoch": 0.34437321937321935, + "grad_norm": 0.5644593834877014, + "learning_rate": 0.00019643258316088703, + "loss": 1.2559, + "step": 1934 + }, + { + "epoch": 0.34455128205128205, + "grad_norm": 0.500815749168396, + "learning_rate": 0.00019642887683148718, + "loss": 1.0439, + "step": 1935 + }, + { + "epoch": 0.34472934472934474, + "grad_norm": 0.4932316541671753, + "learning_rate": 0.0001964251686127638, + "loss": 1.0404, + "step": 1936 + }, + { + "epoch": 0.3449074074074074, + "grad_norm": 0.48494651913642883, + "learning_rate": 0.00019642145850478954, + "loss": 0.9951, + "step": 1937 + }, + { + "epoch": 0.3450854700854701, + "grad_norm": 0.5191963315010071, + "learning_rate": 0.00019641774650763706, + "loss": 1.1258, + "step": 1938 + }, + { + "epoch": 0.3452635327635328, + "grad_norm": 0.4439312815666199, + "learning_rate": 0.00019641403262137918, + "loss": 1.1158, + "step": 1939 + }, + { + "epoch": 0.3454415954415954, + "grad_norm": 0.4829137921333313, + "learning_rate": 0.0001964103168460886, + "loss": 1.0531, + "step": 1940 + }, + { + "epoch": 0.3456196581196581, + "grad_norm": 0.49433329701423645, + "learning_rate": 0.00019640659918183811, + "loss": 1.1295, + "step": 1941 + }, + { + "epoch": 0.3457977207977208, + "grad_norm": 0.5351347923278809, + "learning_rate": 0.00019640287962870062, + "loss": 1.2379, + "step": 1942 + }, + { + "epoch": 0.34597578347578345, + "grad_norm": 0.4845680892467499, + "learning_rate": 0.00019639915818674895, + "loss": 1.0197, + "step": 1943 + }, + { + "epoch": 0.34615384615384615, + "grad_norm": 0.5312514901161194, + "learning_rate": 0.00019639543485605604, + "loss": 0.9734, + "step": 1944 + }, + { + "epoch": 0.34633190883190884, + "grad_norm": 0.4571874737739563, + "learning_rate": 0.00019639170963669478, + "loss": 1.1012, + "step": 1945 + }, + { + "epoch": 0.34650997150997154, + "grad_norm": 0.4449031949043274, + "learning_rate": 0.00019638798252873824, + "loss": 1.1393, + "step": 1946 + }, + { + "epoch": 0.3466880341880342, + "grad_norm": 0.47470834851264954, + "learning_rate": 0.0001963842535322594, + "loss": 0.981, + "step": 1947 + }, + { + "epoch": 0.3468660968660969, + "grad_norm": 0.5386981964111328, + "learning_rate": 0.00019638052264733132, + "loss": 1.1247, + "step": 1948 + }, + { + "epoch": 0.34704415954415957, + "grad_norm": 0.535589873790741, + "learning_rate": 0.00019637678987402714, + "loss": 1.3157, + "step": 1949 + }, + { + "epoch": 0.3472222222222222, + "grad_norm": 0.49338245391845703, + "learning_rate": 0.00019637305521242, + "loss": 1.1066, + "step": 1950 + }, + { + "epoch": 0.3474002849002849, + "grad_norm": 0.4247688353061676, + "learning_rate": 0.00019636931866258298, + "loss": 1.0039, + "step": 1951 + }, + { + "epoch": 0.3475783475783476, + "grad_norm": 0.5351517200469971, + "learning_rate": 0.00019636558022458934, + "loss": 1.0344, + "step": 1952 + }, + { + "epoch": 0.34775641025641024, + "grad_norm": 0.4633362889289856, + "learning_rate": 0.00019636183989851238, + "loss": 1.1383, + "step": 1953 + }, + { + "epoch": 0.34793447293447294, + "grad_norm": 0.553709089756012, + "learning_rate": 0.00019635809768442535, + "loss": 1.0389, + "step": 1954 + }, + { + "epoch": 0.34811253561253563, + "grad_norm": 0.479374498128891, + "learning_rate": 0.00019635435358240154, + "loss": 1.1774, + "step": 1955 + }, + { + "epoch": 0.3482905982905983, + "grad_norm": 0.5274081230163574, + "learning_rate": 0.0001963506075925143, + "loss": 1.1809, + "step": 1956 + }, + { + "epoch": 0.34846866096866097, + "grad_norm": 0.45398542284965515, + "learning_rate": 0.0001963468597148371, + "loss": 1.0502, + "step": 1957 + }, + { + "epoch": 0.34864672364672367, + "grad_norm": 0.48201611638069153, + "learning_rate": 0.00019634310994944332, + "loss": 1.0557, + "step": 1958 + }, + { + "epoch": 0.3488247863247863, + "grad_norm": 0.6407544016838074, + "learning_rate": 0.00019633935829640642, + "loss": 1.2138, + "step": 1959 + }, + { + "epoch": 0.349002849002849, + "grad_norm": 0.5385687351226807, + "learning_rate": 0.00019633560475579995, + "loss": 1.3496, + "step": 1960 + }, + { + "epoch": 0.3491809116809117, + "grad_norm": 0.5260964035987854, + "learning_rate": 0.0001963318493276974, + "loss": 1.0253, + "step": 1961 + }, + { + "epoch": 0.34935897435897434, + "grad_norm": 0.48478585481643677, + "learning_rate": 0.00019632809201217238, + "loss": 1.137, + "step": 1962 + }, + { + "epoch": 0.34953703703703703, + "grad_norm": 0.620033860206604, + "learning_rate": 0.0001963243328092985, + "loss": 1.3445, + "step": 1963 + }, + { + "epoch": 0.34971509971509973, + "grad_norm": 0.5149700045585632, + "learning_rate": 0.00019632057171914942, + "loss": 1.1042, + "step": 1964 + }, + { + "epoch": 0.34989316239316237, + "grad_norm": 0.42695048451423645, + "learning_rate": 0.0001963168087417988, + "loss": 0.8789, + "step": 1965 + }, + { + "epoch": 0.35007122507122507, + "grad_norm": 0.5281283855438232, + "learning_rate": 0.00019631304387732044, + "loss": 1.1155, + "step": 1966 + }, + { + "epoch": 0.35024928774928776, + "grad_norm": 0.4994089901447296, + "learning_rate": 0.00019630927712578804, + "loss": 1.1226, + "step": 1967 + }, + { + "epoch": 0.3504273504273504, + "grad_norm": 0.4433288276195526, + "learning_rate": 0.0001963055084872754, + "loss": 1.0262, + "step": 1968 + }, + { + "epoch": 0.3506054131054131, + "grad_norm": 0.46541857719421387, + "learning_rate": 0.0001963017379618564, + "loss": 1.1438, + "step": 1969 + }, + { + "epoch": 0.3507834757834758, + "grad_norm": 0.5097604393959045, + "learning_rate": 0.00019629796554960488, + "loss": 0.9641, + "step": 1970 + }, + { + "epoch": 0.35096153846153844, + "grad_norm": 0.49461981654167175, + "learning_rate": 0.00019629419125059478, + "loss": 1.1765, + "step": 1971 + }, + { + "epoch": 0.35113960113960113, + "grad_norm": 0.4763339161872864, + "learning_rate": 0.00019629041506490005, + "loss": 1.0527, + "step": 1972 + }, + { + "epoch": 0.35131766381766383, + "grad_norm": 0.4528443217277527, + "learning_rate": 0.00019628663699259463, + "loss": 1.1409, + "step": 1973 + }, + { + "epoch": 0.35149572649572647, + "grad_norm": 0.4436309039592743, + "learning_rate": 0.00019628285703375258, + "loss": 1.0459, + "step": 1974 + }, + { + "epoch": 0.35167378917378916, + "grad_norm": 0.5146129727363586, + "learning_rate": 0.00019627907518844797, + "loss": 1.2527, + "step": 1975 + }, + { + "epoch": 0.35185185185185186, + "grad_norm": 0.5202171802520752, + "learning_rate": 0.0001962752914567549, + "loss": 1.226, + "step": 1976 + }, + { + "epoch": 0.35202991452991456, + "grad_norm": 0.5267411470413208, + "learning_rate": 0.00019627150583874747, + "loss": 1.0898, + "step": 1977 + }, + { + "epoch": 0.3522079772079772, + "grad_norm": 0.546840250492096, + "learning_rate": 0.00019626771833449987, + "loss": 1.1716, + "step": 1978 + }, + { + "epoch": 0.3523860398860399, + "grad_norm": 0.5525290966033936, + "learning_rate": 0.0001962639289440863, + "loss": 1.1762, + "step": 1979 + }, + { + "epoch": 0.3525641025641026, + "grad_norm": 0.48967215418815613, + "learning_rate": 0.000196260137667581, + "loss": 1.1884, + "step": 1980 + }, + { + "epoch": 0.35274216524216523, + "grad_norm": 0.5908235907554626, + "learning_rate": 0.0001962563445050583, + "loss": 1.1887, + "step": 1981 + }, + { + "epoch": 0.3529202279202279, + "grad_norm": 0.46708086133003235, + "learning_rate": 0.00019625254945659245, + "loss": 0.8842, + "step": 1982 + }, + { + "epoch": 0.3530982905982906, + "grad_norm": 0.41652458906173706, + "learning_rate": 0.00019624875252225788, + "loss": 1.0268, + "step": 1983 + }, + { + "epoch": 0.35327635327635326, + "grad_norm": 0.5084529519081116, + "learning_rate": 0.00019624495370212892, + "loss": 1.0547, + "step": 1984 + }, + { + "epoch": 0.35345441595441596, + "grad_norm": 0.5667507648468018, + "learning_rate": 0.00019624115299628003, + "loss": 1.0656, + "step": 1985 + }, + { + "epoch": 0.35363247863247865, + "grad_norm": 0.5022873282432556, + "learning_rate": 0.00019623735040478568, + "loss": 1.0627, + "step": 1986 + }, + { + "epoch": 0.3538105413105413, + "grad_norm": 0.48342058062553406, + "learning_rate": 0.00019623354592772035, + "loss": 1.0976, + "step": 1987 + }, + { + "epoch": 0.353988603988604, + "grad_norm": 0.48117366433143616, + "learning_rate": 0.0001962297395651586, + "loss": 1.0515, + "step": 1988 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 0.492564857006073, + "learning_rate": 0.000196225931317175, + "loss": 1.1957, + "step": 1989 + }, + { + "epoch": 0.3543447293447293, + "grad_norm": 0.4756208658218384, + "learning_rate": 0.00019622212118384417, + "loss": 1.007, + "step": 1990 + }, + { + "epoch": 0.354522792022792, + "grad_norm": 0.581930935382843, + "learning_rate": 0.00019621830916524076, + "loss": 1.232, + "step": 1991 + }, + { + "epoch": 0.3547008547008547, + "grad_norm": 0.480064332485199, + "learning_rate": 0.00019621449526143947, + "loss": 1.2693, + "step": 1992 + }, + { + "epoch": 0.35487891737891736, + "grad_norm": 0.5679123401641846, + "learning_rate": 0.000196210679472515, + "loss": 1.2985, + "step": 1993 + }, + { + "epoch": 0.35505698005698005, + "grad_norm": 0.43757280707359314, + "learning_rate": 0.00019620686179854213, + "loss": 1.1387, + "step": 1994 + }, + { + "epoch": 0.35523504273504275, + "grad_norm": 0.4950634837150574, + "learning_rate": 0.00019620304223959566, + "loss": 1.1809, + "step": 1995 + }, + { + "epoch": 0.3554131054131054, + "grad_norm": 0.5574113726615906, + "learning_rate": 0.00019619922079575043, + "loss": 1.2434, + "step": 1996 + }, + { + "epoch": 0.3555911680911681, + "grad_norm": 0.5154930949211121, + "learning_rate": 0.00019619539746708128, + "loss": 1.1747, + "step": 1997 + }, + { + "epoch": 0.3557692307692308, + "grad_norm": 0.4377825856208801, + "learning_rate": 0.00019619157225366315, + "loss": 0.9547, + "step": 1998 + }, + { + "epoch": 0.3559472934472934, + "grad_norm": 0.530714213848114, + "learning_rate": 0.00019618774515557097, + "loss": 1.2057, + "step": 1999 + }, + { + "epoch": 0.3561253561253561, + "grad_norm": 0.5703464150428772, + "learning_rate": 0.00019618391617287978, + "loss": 1.3068, + "step": 2000 + }, + { + "epoch": 0.3563034188034188, + "grad_norm": 0.4862228333950043, + "learning_rate": 0.0001961800853056645, + "loss": 1.0077, + "step": 2001 + }, + { + "epoch": 0.35648148148148145, + "grad_norm": 0.5575395822525024, + "learning_rate": 0.00019617625255400028, + "loss": 1.03, + "step": 2002 + }, + { + "epoch": 0.35665954415954415, + "grad_norm": 0.4826279580593109, + "learning_rate": 0.0001961724179179622, + "loss": 1.268, + "step": 2003 + }, + { + "epoch": 0.35683760683760685, + "grad_norm": 0.49423274397850037, + "learning_rate": 0.00019616858139762534, + "loss": 1.1305, + "step": 2004 + }, + { + "epoch": 0.35701566951566954, + "grad_norm": 0.5208541750907898, + "learning_rate": 0.00019616474299306491, + "loss": 1.1651, + "step": 2005 + }, + { + "epoch": 0.3571937321937322, + "grad_norm": 0.5324164032936096, + "learning_rate": 0.0001961609027043561, + "loss": 1.1406, + "step": 2006 + }, + { + "epoch": 0.3573717948717949, + "grad_norm": 0.45385462045669556, + "learning_rate": 0.00019615706053157416, + "loss": 1.0716, + "step": 2007 + }, + { + "epoch": 0.3575498575498576, + "grad_norm": 0.5016173720359802, + "learning_rate": 0.00019615321647479438, + "loss": 1.0878, + "step": 2008 + }, + { + "epoch": 0.3577279202279202, + "grad_norm": 0.5073097348213196, + "learning_rate": 0.00019614937053409205, + "loss": 1.237, + "step": 2009 + }, + { + "epoch": 0.3579059829059829, + "grad_norm": 0.48880141973495483, + "learning_rate": 0.00019614552270954256, + "loss": 0.8794, + "step": 2010 + }, + { + "epoch": 0.3580840455840456, + "grad_norm": 0.43902209401130676, + "learning_rate": 0.00019614167300122126, + "loss": 0.912, + "step": 2011 + }, + { + "epoch": 0.35826210826210825, + "grad_norm": 0.42809322476387024, + "learning_rate": 0.0001961378214092036, + "loss": 0.7804, + "step": 2012 + }, + { + "epoch": 0.35844017094017094, + "grad_norm": 0.4464281499385834, + "learning_rate": 0.00019613396793356503, + "loss": 1.0004, + "step": 2013 + }, + { + "epoch": 0.35861823361823364, + "grad_norm": 0.49085676670074463, + "learning_rate": 0.00019613011257438109, + "loss": 1.1087, + "step": 2014 + }, + { + "epoch": 0.3587962962962963, + "grad_norm": 0.4997732937335968, + "learning_rate": 0.00019612625533172725, + "loss": 0.9591, + "step": 2015 + }, + { + "epoch": 0.358974358974359, + "grad_norm": 0.48442545533180237, + "learning_rate": 0.00019612239620567912, + "loss": 0.9744, + "step": 2016 + }, + { + "epoch": 0.35915242165242167, + "grad_norm": 0.4989205002784729, + "learning_rate": 0.00019611853519631233, + "loss": 0.9844, + "step": 2017 + }, + { + "epoch": 0.3593304843304843, + "grad_norm": 0.6107521653175354, + "learning_rate": 0.00019611467230370248, + "loss": 1.147, + "step": 2018 + }, + { + "epoch": 0.359508547008547, + "grad_norm": 0.5594844818115234, + "learning_rate": 0.00019611080752792535, + "loss": 1.3195, + "step": 2019 + }, + { + "epoch": 0.3596866096866097, + "grad_norm": 0.4786946475505829, + "learning_rate": 0.00019610694086905656, + "loss": 1.2108, + "step": 2020 + }, + { + "epoch": 0.35986467236467234, + "grad_norm": 0.5186030268669128, + "learning_rate": 0.0001961030723271719, + "loss": 1.0008, + "step": 2021 + }, + { + "epoch": 0.36004273504273504, + "grad_norm": 0.4520573318004608, + "learning_rate": 0.0001960992019023472, + "loss": 1.1307, + "step": 2022 + }, + { + "epoch": 0.36022079772079774, + "grad_norm": 0.4983210563659668, + "learning_rate": 0.00019609532959465823, + "loss": 1.1486, + "step": 2023 + }, + { + "epoch": 0.3603988603988604, + "grad_norm": 0.6209200024604797, + "learning_rate": 0.00019609145540418094, + "loss": 1.2566, + "step": 2024 + }, + { + "epoch": 0.3605769230769231, + "grad_norm": 0.47047603130340576, + "learning_rate": 0.00019608757933099117, + "loss": 1.1588, + "step": 2025 + }, + { + "epoch": 0.36075498575498577, + "grad_norm": 0.5147389769554138, + "learning_rate": 0.0001960837013751649, + "loss": 1.2113, + "step": 2026 + }, + { + "epoch": 0.3609330484330484, + "grad_norm": 0.45826098322868347, + "learning_rate": 0.00019607982153677808, + "loss": 1.13, + "step": 2027 + }, + { + "epoch": 0.3611111111111111, + "grad_norm": 0.5699561834335327, + "learning_rate": 0.00019607593981590675, + "loss": 1.2476, + "step": 2028 + }, + { + "epoch": 0.3612891737891738, + "grad_norm": 0.5349239110946655, + "learning_rate": 0.000196072056212627, + "loss": 1.2295, + "step": 2029 + }, + { + "epoch": 0.36146723646723644, + "grad_norm": 0.6212165355682373, + "learning_rate": 0.00019606817072701484, + "loss": 1.1965, + "step": 2030 + }, + { + "epoch": 0.36164529914529914, + "grad_norm": 0.4870990216732025, + "learning_rate": 0.00019606428335914645, + "loss": 1.4464, + "step": 2031 + }, + { + "epoch": 0.36182336182336183, + "grad_norm": 0.42427828907966614, + "learning_rate": 0.00019606039410909797, + "loss": 1.1546, + "step": 2032 + }, + { + "epoch": 0.36200142450142453, + "grad_norm": 0.5081788301467896, + "learning_rate": 0.0001960565029769456, + "loss": 1.1867, + "step": 2033 + }, + { + "epoch": 0.36217948717948717, + "grad_norm": 0.4813104271888733, + "learning_rate": 0.00019605260996276565, + "loss": 1.3726, + "step": 2034 + }, + { + "epoch": 0.36235754985754987, + "grad_norm": 0.4648851156234741, + "learning_rate": 0.0001960487150666343, + "loss": 1.2434, + "step": 2035 + }, + { + "epoch": 0.36253561253561256, + "grad_norm": 0.484161913394928, + "learning_rate": 0.00019604481828862792, + "loss": 1.1309, + "step": 2036 + }, + { + "epoch": 0.3627136752136752, + "grad_norm": 0.4929439127445221, + "learning_rate": 0.00019604091962882283, + "loss": 1.1007, + "step": 2037 + }, + { + "epoch": 0.3628917378917379, + "grad_norm": 0.45599642395973206, + "learning_rate": 0.00019603701908729544, + "loss": 1.2628, + "step": 2038 + }, + { + "epoch": 0.3630698005698006, + "grad_norm": 0.45295149087905884, + "learning_rate": 0.00019603311666412213, + "loss": 0.9808, + "step": 2039 + }, + { + "epoch": 0.36324786324786323, + "grad_norm": 0.48681163787841797, + "learning_rate": 0.00019602921235937942, + "loss": 1.0574, + "step": 2040 + }, + { + "epoch": 0.36342592592592593, + "grad_norm": 0.41232365369796753, + "learning_rate": 0.00019602530617314378, + "loss": 1.0454, + "step": 2041 + }, + { + "epoch": 0.3636039886039886, + "grad_norm": 0.46214723587036133, + "learning_rate": 0.00019602139810549174, + "loss": 0.9985, + "step": 2042 + }, + { + "epoch": 0.36378205128205127, + "grad_norm": 0.44307878613471985, + "learning_rate": 0.00019601748815649989, + "loss": 0.9683, + "step": 2043 + }, + { + "epoch": 0.36396011396011396, + "grad_norm": 0.4809451401233673, + "learning_rate": 0.00019601357632624477, + "loss": 1.028, + "step": 2044 + }, + { + "epoch": 0.36413817663817666, + "grad_norm": 0.4638497531414032, + "learning_rate": 0.0001960096626148031, + "loss": 0.9851, + "step": 2045 + }, + { + "epoch": 0.3643162393162393, + "grad_norm": 0.5942164063453674, + "learning_rate": 0.00019600574702225153, + "loss": 1.1606, + "step": 2046 + }, + { + "epoch": 0.364494301994302, + "grad_norm": 0.5171293616294861, + "learning_rate": 0.00019600182954866675, + "loss": 1.2335, + "step": 2047 + }, + { + "epoch": 0.3646723646723647, + "grad_norm": 0.5294404625892639, + "learning_rate": 0.00019599791019412558, + "loss": 1.0966, + "step": 2048 + }, + { + "epoch": 0.36485042735042733, + "grad_norm": 0.46117448806762695, + "learning_rate": 0.00019599398895870477, + "loss": 1.0565, + "step": 2049 + }, + { + "epoch": 0.36502849002849, + "grad_norm": 0.5385118126869202, + "learning_rate": 0.00019599006584248118, + "loss": 1.0076, + "step": 2050 + }, + { + "epoch": 0.3652065527065527, + "grad_norm": 0.4915166199207306, + "learning_rate": 0.00019598614084553165, + "loss": 0.9686, + "step": 2051 + }, + { + "epoch": 0.36538461538461536, + "grad_norm": 0.46769094467163086, + "learning_rate": 0.00019598221396793303, + "loss": 1.1217, + "step": 2052 + }, + { + "epoch": 0.36556267806267806, + "grad_norm": 0.5440493822097778, + "learning_rate": 0.00019597828520976236, + "loss": 1.2344, + "step": 2053 + }, + { + "epoch": 0.36574074074074076, + "grad_norm": 0.616727352142334, + "learning_rate": 0.00019597435457109657, + "loss": 1.2953, + "step": 2054 + }, + { + "epoch": 0.3659188034188034, + "grad_norm": 0.4859183430671692, + "learning_rate": 0.00019597042205201265, + "loss": 1.16, + "step": 2055 + }, + { + "epoch": 0.3660968660968661, + "grad_norm": 0.47056329250335693, + "learning_rate": 0.0001959664876525877, + "loss": 0.9982, + "step": 2056 + }, + { + "epoch": 0.3662749287749288, + "grad_norm": 0.48347967863082886, + "learning_rate": 0.00019596255137289875, + "loss": 1.0966, + "step": 2057 + }, + { + "epoch": 0.36645299145299143, + "grad_norm": 0.5068454742431641, + "learning_rate": 0.00019595861321302296, + "loss": 1.2891, + "step": 2058 + }, + { + "epoch": 0.3666310541310541, + "grad_norm": 0.5702359080314636, + "learning_rate": 0.00019595467317303747, + "loss": 1.1394, + "step": 2059 + }, + { + "epoch": 0.3668091168091168, + "grad_norm": 0.5028812885284424, + "learning_rate": 0.0001959507312530195, + "loss": 1.2324, + "step": 2060 + }, + { + "epoch": 0.36698717948717946, + "grad_norm": 0.4672880172729492, + "learning_rate": 0.00019594678745304628, + "loss": 1.0581, + "step": 2061 + }, + { + "epoch": 0.36716524216524216, + "grad_norm": 0.5233900547027588, + "learning_rate": 0.00019594284177319504, + "loss": 1.138, + "step": 2062 + }, + { + "epoch": 0.36734330484330485, + "grad_norm": 0.46871712803840637, + "learning_rate": 0.00019593889421354316, + "loss": 1.2159, + "step": 2063 + }, + { + "epoch": 0.36752136752136755, + "grad_norm": 0.5180533528327942, + "learning_rate": 0.00019593494477416793, + "loss": 1.1116, + "step": 2064 + }, + { + "epoch": 0.3676994301994302, + "grad_norm": 0.5398494005203247, + "learning_rate": 0.0001959309934551467, + "loss": 1.2038, + "step": 2065 + }, + { + "epoch": 0.3678774928774929, + "grad_norm": 0.4850373864173889, + "learning_rate": 0.000195927040256557, + "loss": 1.4315, + "step": 2066 + }, + { + "epoch": 0.3680555555555556, + "grad_norm": 0.49190905690193176, + "learning_rate": 0.0001959230851784762, + "loss": 0.9993, + "step": 2067 + }, + { + "epoch": 0.3682336182336182, + "grad_norm": 0.4546903073787689, + "learning_rate": 0.00019591912822098178, + "loss": 1.0979, + "step": 2068 + }, + { + "epoch": 0.3684116809116809, + "grad_norm": 0.4726468622684479, + "learning_rate": 0.00019591516938415133, + "loss": 1.1629, + "step": 2069 + }, + { + "epoch": 0.3685897435897436, + "grad_norm": 0.47856009006500244, + "learning_rate": 0.00019591120866806235, + "loss": 1.2048, + "step": 2070 + }, + { + "epoch": 0.36876780626780625, + "grad_norm": 0.46847718954086304, + "learning_rate": 0.0001959072460727925, + "loss": 1.0958, + "step": 2071 + }, + { + "epoch": 0.36894586894586895, + "grad_norm": 0.47164350748062134, + "learning_rate": 0.0001959032815984194, + "loss": 1.1912, + "step": 2072 + }, + { + "epoch": 0.36912393162393164, + "grad_norm": 0.4838213324546814, + "learning_rate": 0.0001958993152450207, + "loss": 1.1466, + "step": 2073 + }, + { + "epoch": 0.3693019943019943, + "grad_norm": 0.47234636545181274, + "learning_rate": 0.00019589534701267412, + "loss": 0.9475, + "step": 2074 + }, + { + "epoch": 0.369480056980057, + "grad_norm": 0.4913126826286316, + "learning_rate": 0.00019589137690145746, + "loss": 1.1571, + "step": 2075 + }, + { + "epoch": 0.3696581196581197, + "grad_norm": 0.4696233570575714, + "learning_rate": 0.00019588740491144842, + "loss": 0.9797, + "step": 2076 + }, + { + "epoch": 0.3698361823361823, + "grad_norm": 0.46146106719970703, + "learning_rate": 0.00019588343104272492, + "loss": 1.027, + "step": 2077 + }, + { + "epoch": 0.370014245014245, + "grad_norm": 0.4920627176761627, + "learning_rate": 0.00019587945529536474, + "loss": 1.1008, + "step": 2078 + }, + { + "epoch": 0.3701923076923077, + "grad_norm": 0.4854249954223633, + "learning_rate": 0.0001958754776694458, + "loss": 1.0759, + "step": 2079 + }, + { + "epoch": 0.37037037037037035, + "grad_norm": 0.4884897768497467, + "learning_rate": 0.00019587149816504608, + "loss": 1.1403, + "step": 2080 + }, + { + "epoch": 0.37054843304843305, + "grad_norm": 0.5062584280967712, + "learning_rate": 0.00019586751678224345, + "loss": 1.0185, + "step": 2081 + }, + { + "epoch": 0.37072649572649574, + "grad_norm": 0.44697675108909607, + "learning_rate": 0.000195863533521116, + "loss": 1.0462, + "step": 2082 + }, + { + "epoch": 0.3709045584045584, + "grad_norm": 0.5122885704040527, + "learning_rate": 0.00019585954838174176, + "loss": 1.108, + "step": 2083 + }, + { + "epoch": 0.3710826210826211, + "grad_norm": 0.486650288105011, + "learning_rate": 0.0001958555613641988, + "loss": 1.126, + "step": 2084 + }, + { + "epoch": 0.3712606837606838, + "grad_norm": 0.5296297669410706, + "learning_rate": 0.00019585157246856523, + "loss": 1.1757, + "step": 2085 + }, + { + "epoch": 0.3714387464387464, + "grad_norm": 0.4935721457004547, + "learning_rate": 0.0001958475816949192, + "loss": 1.1654, + "step": 2086 + }, + { + "epoch": 0.3716168091168091, + "grad_norm": 0.6226509213447571, + "learning_rate": 0.00019584358904333891, + "loss": 1.1981, + "step": 2087 + }, + { + "epoch": 0.3717948717948718, + "grad_norm": 0.44094228744506836, + "learning_rate": 0.0001958395945139026, + "loss": 0.8468, + "step": 2088 + }, + { + "epoch": 0.37197293447293445, + "grad_norm": 0.5335884690284729, + "learning_rate": 0.00019583559810668858, + "loss": 1.1597, + "step": 2089 + }, + { + "epoch": 0.37215099715099714, + "grad_norm": 0.4585414528846741, + "learning_rate": 0.000195831599821775, + "loss": 0.9343, + "step": 2090 + }, + { + "epoch": 0.37232905982905984, + "grad_norm": 0.533087432384491, + "learning_rate": 0.00019582759965924035, + "loss": 1.1209, + "step": 2091 + }, + { + "epoch": 0.37250712250712253, + "grad_norm": 0.5302683711051941, + "learning_rate": 0.00019582359761916295, + "loss": 1.236, + "step": 2092 + }, + { + "epoch": 0.3726851851851852, + "grad_norm": 0.4522508382797241, + "learning_rate": 0.00019581959370162122, + "loss": 1.0196, + "step": 2093 + }, + { + "epoch": 0.37286324786324787, + "grad_norm": 0.52391517162323, + "learning_rate": 0.00019581558790669358, + "loss": 1.0077, + "step": 2094 + }, + { + "epoch": 0.37304131054131057, + "grad_norm": 0.47144797444343567, + "learning_rate": 0.00019581158023445854, + "loss": 1.0956, + "step": 2095 + }, + { + "epoch": 0.3732193732193732, + "grad_norm": 0.4486723244190216, + "learning_rate": 0.00019580757068499459, + "loss": 0.8697, + "step": 2096 + }, + { + "epoch": 0.3733974358974359, + "grad_norm": 0.4626580476760864, + "learning_rate": 0.00019580355925838034, + "loss": 0.8489, + "step": 2097 + }, + { + "epoch": 0.3735754985754986, + "grad_norm": 0.5647920370101929, + "learning_rate": 0.00019579954595469438, + "loss": 1.1458, + "step": 2098 + }, + { + "epoch": 0.37375356125356124, + "grad_norm": 0.4734349846839905, + "learning_rate": 0.00019579553077401528, + "loss": 1.1036, + "step": 2099 + }, + { + "epoch": 0.37393162393162394, + "grad_norm": 0.5624295473098755, + "learning_rate": 0.00019579151371642176, + "loss": 0.9793, + "step": 2100 + }, + { + "epoch": 0.37410968660968663, + "grad_norm": 0.47507283091545105, + "learning_rate": 0.00019578749478199256, + "loss": 1.0371, + "step": 2101 + }, + { + "epoch": 0.37428774928774927, + "grad_norm": 0.550865113735199, + "learning_rate": 0.00019578347397080633, + "loss": 1.046, + "step": 2102 + }, + { + "epoch": 0.37446581196581197, + "grad_norm": 0.5249403715133667, + "learning_rate": 0.00019577945128294193, + "loss": 1.3185, + "step": 2103 + }, + { + "epoch": 0.37464387464387466, + "grad_norm": 0.4921024739742279, + "learning_rate": 0.00019577542671847815, + "loss": 1.0758, + "step": 2104 + }, + { + "epoch": 0.3748219373219373, + "grad_norm": 0.5351784825325012, + "learning_rate": 0.00019577140027749384, + "loss": 1.067, + "step": 2105 + }, + { + "epoch": 0.375, + "grad_norm": 0.44420507550239563, + "learning_rate": 0.00019576737196006787, + "loss": 1.1065, + "step": 2106 + }, + { + "epoch": 0.3751780626780627, + "grad_norm": 0.531384289264679, + "learning_rate": 0.0001957633417662792, + "loss": 1.1634, + "step": 2107 + }, + { + "epoch": 0.37535612535612534, + "grad_norm": 0.5167618989944458, + "learning_rate": 0.00019575930969620677, + "loss": 1.1646, + "step": 2108 + }, + { + "epoch": 0.37553418803418803, + "grad_norm": 0.41487228870391846, + "learning_rate": 0.0001957552757499296, + "loss": 0.793, + "step": 2109 + }, + { + "epoch": 0.37571225071225073, + "grad_norm": 0.5110787153244019, + "learning_rate": 0.00019575123992752672, + "loss": 1.1752, + "step": 2110 + }, + { + "epoch": 0.37589031339031337, + "grad_norm": 0.4422051012516022, + "learning_rate": 0.00019574720222907717, + "loss": 1.0102, + "step": 2111 + }, + { + "epoch": 0.37606837606837606, + "grad_norm": 0.4757538139820099, + "learning_rate": 0.0001957431626546601, + "loss": 1.0467, + "step": 2112 + }, + { + "epoch": 0.37624643874643876, + "grad_norm": 0.4736764430999756, + "learning_rate": 0.00019573912120435466, + "loss": 1.3048, + "step": 2113 + }, + { + "epoch": 0.3764245014245014, + "grad_norm": 0.49894335865974426, + "learning_rate": 0.00019573507787824004, + "loss": 1.0502, + "step": 2114 + }, + { + "epoch": 0.3766025641025641, + "grad_norm": 0.48120981454849243, + "learning_rate": 0.00019573103267639543, + "loss": 1.2405, + "step": 2115 + }, + { + "epoch": 0.3767806267806268, + "grad_norm": 0.4826737642288208, + "learning_rate": 0.0001957269855989001, + "loss": 1.1189, + "step": 2116 + }, + { + "epoch": 0.37695868945868943, + "grad_norm": 0.4736921489238739, + "learning_rate": 0.0001957229366458333, + "loss": 1.2862, + "step": 2117 + }, + { + "epoch": 0.37713675213675213, + "grad_norm": 0.3895208537578583, + "learning_rate": 0.00019571888581727446, + "loss": 1.0573, + "step": 2118 + }, + { + "epoch": 0.3773148148148148, + "grad_norm": 0.5107510089874268, + "learning_rate": 0.00019571483311330284, + "loss": 1.2913, + "step": 2119 + }, + { + "epoch": 0.37749287749287747, + "grad_norm": 0.4543241262435913, + "learning_rate": 0.00019571077853399794, + "loss": 0.949, + "step": 2120 + }, + { + "epoch": 0.37767094017094016, + "grad_norm": 0.46897491812705994, + "learning_rate": 0.00019570672207943913, + "loss": 1.2235, + "step": 2121 + }, + { + "epoch": 0.37784900284900286, + "grad_norm": 0.4812130630016327, + "learning_rate": 0.0001957026637497059, + "loss": 0.8857, + "step": 2122 + }, + { + "epoch": 0.37802706552706555, + "grad_norm": 0.47452476620674133, + "learning_rate": 0.00019569860354487782, + "loss": 1.0549, + "step": 2123 + }, + { + "epoch": 0.3782051282051282, + "grad_norm": 0.49879950284957886, + "learning_rate": 0.00019569454146503438, + "loss": 1.0475, + "step": 2124 + }, + { + "epoch": 0.3783831908831909, + "grad_norm": 0.4246445894241333, + "learning_rate": 0.00019569047751025518, + "loss": 0.8788, + "step": 2125 + }, + { + "epoch": 0.3785612535612536, + "grad_norm": 0.4868565499782562, + "learning_rate": 0.00019568641168061986, + "loss": 1.1801, + "step": 2126 + }, + { + "epoch": 0.3787393162393162, + "grad_norm": 0.46723654866218567, + "learning_rate": 0.0001956823439762081, + "loss": 1.1661, + "step": 2127 + }, + { + "epoch": 0.3789173789173789, + "grad_norm": 0.4989059269428253, + "learning_rate": 0.00019567827439709954, + "loss": 1.3037, + "step": 2128 + }, + { + "epoch": 0.3790954415954416, + "grad_norm": 0.441307932138443, + "learning_rate": 0.00019567420294337395, + "loss": 1.0197, + "step": 2129 + }, + { + "epoch": 0.37927350427350426, + "grad_norm": 0.5200160145759583, + "learning_rate": 0.0001956701296151111, + "loss": 1.3366, + "step": 2130 + }, + { + "epoch": 0.37945156695156695, + "grad_norm": 0.43610256910324097, + "learning_rate": 0.00019566605441239082, + "loss": 1.0148, + "step": 2131 + }, + { + "epoch": 0.37962962962962965, + "grad_norm": 0.4160982370376587, + "learning_rate": 0.00019566197733529293, + "loss": 1.0758, + "step": 2132 + }, + { + "epoch": 0.3798076923076923, + "grad_norm": 0.5007950663566589, + "learning_rate": 0.00019565789838389726, + "loss": 1.1937, + "step": 2133 + }, + { + "epoch": 0.379985754985755, + "grad_norm": 0.4991525113582611, + "learning_rate": 0.00019565381755828385, + "loss": 1.1788, + "step": 2134 + }, + { + "epoch": 0.3801638176638177, + "grad_norm": 0.6313113570213318, + "learning_rate": 0.00019564973485853258, + "loss": 1.1241, + "step": 2135 + }, + { + "epoch": 0.3803418803418803, + "grad_norm": 0.49736538529396057, + "learning_rate": 0.0001956456502847234, + "loss": 1.0299, + "step": 2136 + }, + { + "epoch": 0.380519943019943, + "grad_norm": 0.4384380578994751, + "learning_rate": 0.00019564156383693643, + "loss": 1.132, + "step": 2137 + }, + { + "epoch": 0.3806980056980057, + "grad_norm": 0.4696183502674103, + "learning_rate": 0.00019563747551525168, + "loss": 1.1145, + "step": 2138 + }, + { + "epoch": 0.38087606837606836, + "grad_norm": 0.42039749026298523, + "learning_rate": 0.0001956333853197493, + "loss": 0.9549, + "step": 2139 + }, + { + "epoch": 0.38105413105413105, + "grad_norm": 0.5547221899032593, + "learning_rate": 0.00019562929325050936, + "loss": 1.0476, + "step": 2140 + }, + { + "epoch": 0.38123219373219375, + "grad_norm": 0.4803301692008972, + "learning_rate": 0.0001956251993076121, + "loss": 1.1285, + "step": 2141 + }, + { + "epoch": 0.3814102564102564, + "grad_norm": 0.609501838684082, + "learning_rate": 0.00019562110349113766, + "loss": 1.2375, + "step": 2142 + }, + { + "epoch": 0.3815883190883191, + "grad_norm": 0.5134759545326233, + "learning_rate": 0.00019561700580116639, + "loss": 1.0895, + "step": 2143 + }, + { + "epoch": 0.3817663817663818, + "grad_norm": 0.5086711049079895, + "learning_rate": 0.00019561290623777846, + "loss": 1.1139, + "step": 2144 + }, + { + "epoch": 0.3819444444444444, + "grad_norm": 0.5371596813201904, + "learning_rate": 0.00019560880480105428, + "loss": 0.9302, + "step": 2145 + }, + { + "epoch": 0.3821225071225071, + "grad_norm": 0.4966319799423218, + "learning_rate": 0.00019560470149107418, + "loss": 1.2485, + "step": 2146 + }, + { + "epoch": 0.3823005698005698, + "grad_norm": 0.5296950340270996, + "learning_rate": 0.00019560059630791855, + "loss": 1.4449, + "step": 2147 + }, + { + "epoch": 0.38247863247863245, + "grad_norm": 0.5564194321632385, + "learning_rate": 0.00019559648925166783, + "loss": 1.0817, + "step": 2148 + }, + { + "epoch": 0.38265669515669515, + "grad_norm": 0.5763841867446899, + "learning_rate": 0.0001955923803224025, + "loss": 1.1915, + "step": 2149 + }, + { + "epoch": 0.38283475783475784, + "grad_norm": 0.4782295823097229, + "learning_rate": 0.00019558826952020304, + "loss": 1.1317, + "step": 2150 + }, + { + "epoch": 0.38301282051282054, + "grad_norm": 0.4876856207847595, + "learning_rate": 0.00019558415684515002, + "loss": 1.2113, + "step": 2151 + }, + { + "epoch": 0.3831908831908832, + "grad_norm": 0.4894421398639679, + "learning_rate": 0.00019558004229732398, + "loss": 1.0761, + "step": 2152 + }, + { + "epoch": 0.3833689458689459, + "grad_norm": 0.47914227843284607, + "learning_rate": 0.0001955759258768056, + "loss": 1.0869, + "step": 2153 + }, + { + "epoch": 0.38354700854700857, + "grad_norm": 0.43933629989624023, + "learning_rate": 0.00019557180758367543, + "loss": 1.0581, + "step": 2154 + }, + { + "epoch": 0.3837250712250712, + "grad_norm": 0.4078103005886078, + "learning_rate": 0.00019556768741801428, + "loss": 1.065, + "step": 2155 + }, + { + "epoch": 0.3839031339031339, + "grad_norm": 0.5112793445587158, + "learning_rate": 0.00019556356537990278, + "loss": 1.2023, + "step": 2156 + }, + { + "epoch": 0.3840811965811966, + "grad_norm": 0.4699678122997284, + "learning_rate": 0.00019555944146942177, + "loss": 1.2459, + "step": 2157 + }, + { + "epoch": 0.38425925925925924, + "grad_norm": 0.4723528027534485, + "learning_rate": 0.00019555531568665198, + "loss": 1.2204, + "step": 2158 + }, + { + "epoch": 0.38443732193732194, + "grad_norm": 0.4648225009441376, + "learning_rate": 0.00019555118803167432, + "loss": 1.1355, + "step": 2159 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 0.49861815571784973, + "learning_rate": 0.00019554705850456961, + "loss": 1.1301, + "step": 2160 + }, + { + "epoch": 0.3847934472934473, + "grad_norm": 0.4076344966888428, + "learning_rate": 0.00019554292710541874, + "loss": 0.8997, + "step": 2161 + }, + { + "epoch": 0.38497150997151, + "grad_norm": 0.5510796308517456, + "learning_rate": 0.00019553879383430272, + "loss": 1.0594, + "step": 2162 + }, + { + "epoch": 0.38514957264957267, + "grad_norm": 0.55793696641922, + "learning_rate": 0.00019553465869130249, + "loss": 1.1284, + "step": 2163 + }, + { + "epoch": 0.3853276353276353, + "grad_norm": 0.5096491575241089, + "learning_rate": 0.00019553052167649906, + "loss": 1.0419, + "step": 2164 + }, + { + "epoch": 0.385505698005698, + "grad_norm": 0.49077361822128296, + "learning_rate": 0.0001955263827899735, + "loss": 1.1632, + "step": 2165 + }, + { + "epoch": 0.3856837606837607, + "grad_norm": 0.5546894073486328, + "learning_rate": 0.00019552224203180693, + "loss": 1.1487, + "step": 2166 + }, + { + "epoch": 0.38586182336182334, + "grad_norm": 0.4930037260055542, + "learning_rate": 0.00019551809940208047, + "loss": 1.2668, + "step": 2167 + }, + { + "epoch": 0.38603988603988604, + "grad_norm": 0.5600671172142029, + "learning_rate": 0.00019551395490087525, + "loss": 1.3988, + "step": 2168 + }, + { + "epoch": 0.38621794871794873, + "grad_norm": 0.45897629857063293, + "learning_rate": 0.0001955098085282725, + "loss": 0.7792, + "step": 2169 + }, + { + "epoch": 0.3863960113960114, + "grad_norm": 0.46138936281204224, + "learning_rate": 0.00019550566028435346, + "loss": 1.1749, + "step": 2170 + }, + { + "epoch": 0.38657407407407407, + "grad_norm": 0.5136167407035828, + "learning_rate": 0.0001955015101691994, + "loss": 1.0153, + "step": 2171 + }, + { + "epoch": 0.38675213675213677, + "grad_norm": 0.4886440336704254, + "learning_rate": 0.00019549735818289165, + "loss": 1.0006, + "step": 2172 + }, + { + "epoch": 0.3869301994301994, + "grad_norm": 0.4339776635169983, + "learning_rate": 0.00019549320432551154, + "loss": 1.0109, + "step": 2173 + }, + { + "epoch": 0.3871082621082621, + "grad_norm": 0.48729443550109863, + "learning_rate": 0.00019548904859714044, + "loss": 1.2016, + "step": 2174 + }, + { + "epoch": 0.3872863247863248, + "grad_norm": 0.5128757357597351, + "learning_rate": 0.0001954848909978598, + "loss": 1.085, + "step": 2175 + }, + { + "epoch": 0.38746438746438744, + "grad_norm": 0.49636292457580566, + "learning_rate": 0.0001954807315277511, + "loss": 1.0671, + "step": 2176 + }, + { + "epoch": 0.38764245014245013, + "grad_norm": 0.4946988821029663, + "learning_rate": 0.00019547657018689578, + "loss": 1.2091, + "step": 2177 + }, + { + "epoch": 0.38782051282051283, + "grad_norm": 0.49004554748535156, + "learning_rate": 0.00019547240697537544, + "loss": 1.0241, + "step": 2178 + }, + { + "epoch": 0.38799857549857547, + "grad_norm": 0.48750075697898865, + "learning_rate": 0.00019546824189327157, + "loss": 1.1082, + "step": 2179 + }, + { + "epoch": 0.38817663817663817, + "grad_norm": 0.47726166248321533, + "learning_rate": 0.00019546407494066585, + "loss": 1.1275, + "step": 2180 + }, + { + "epoch": 0.38835470085470086, + "grad_norm": 0.5253444910049438, + "learning_rate": 0.00019545990611763986, + "loss": 1.0164, + "step": 2181 + }, + { + "epoch": 0.38853276353276356, + "grad_norm": 0.4470371603965759, + "learning_rate": 0.00019545573542427533, + "loss": 1.0138, + "step": 2182 + }, + { + "epoch": 0.3887108262108262, + "grad_norm": 0.6645087599754333, + "learning_rate": 0.00019545156286065397, + "loss": 1.0884, + "step": 2183 + }, + { + "epoch": 0.3888888888888889, + "grad_norm": 0.498775839805603, + "learning_rate": 0.0001954473884268575, + "loss": 1.1035, + "step": 2184 + }, + { + "epoch": 0.3890669515669516, + "grad_norm": 0.5830566883087158, + "learning_rate": 0.00019544321212296772, + "loss": 1.1665, + "step": 2185 + }, + { + "epoch": 0.38924501424501423, + "grad_norm": 0.48162809014320374, + "learning_rate": 0.00019543903394906646, + "loss": 1.1035, + "step": 2186 + }, + { + "epoch": 0.3894230769230769, + "grad_norm": 0.46334075927734375, + "learning_rate": 0.0001954348539052356, + "loss": 0.9764, + "step": 2187 + }, + { + "epoch": 0.3896011396011396, + "grad_norm": 0.6343515515327454, + "learning_rate": 0.00019543067199155704, + "loss": 0.9474, + "step": 2188 + }, + { + "epoch": 0.38977920227920226, + "grad_norm": 0.4867806136608124, + "learning_rate": 0.0001954264882081127, + "loss": 1.1161, + "step": 2189 + }, + { + "epoch": 0.38995726495726496, + "grad_norm": 0.49305734038352966, + "learning_rate": 0.00019542230255498454, + "loss": 1.1825, + "step": 2190 + }, + { + "epoch": 0.39013532763532766, + "grad_norm": 0.518465518951416, + "learning_rate": 0.00019541811503225457, + "loss": 1.0695, + "step": 2191 + }, + { + "epoch": 0.3903133903133903, + "grad_norm": 0.4892457127571106, + "learning_rate": 0.00019541392564000488, + "loss": 1.3113, + "step": 2192 + }, + { + "epoch": 0.390491452991453, + "grad_norm": 0.5150920152664185, + "learning_rate": 0.00019540973437831753, + "loss": 1.0735, + "step": 2193 + }, + { + "epoch": 0.3906695156695157, + "grad_norm": 0.5414708256721497, + "learning_rate": 0.00019540554124727462, + "loss": 1.0773, + "step": 2194 + }, + { + "epoch": 0.39084757834757833, + "grad_norm": 0.49826398491859436, + "learning_rate": 0.0001954013462469583, + "loss": 1.0542, + "step": 2195 + }, + { + "epoch": 0.391025641025641, + "grad_norm": 0.5203596949577332, + "learning_rate": 0.0001953971493774508, + "loss": 1.178, + "step": 2196 + }, + { + "epoch": 0.3912037037037037, + "grad_norm": 0.45095738768577576, + "learning_rate": 0.00019539295063883432, + "loss": 1.1254, + "step": 2197 + }, + { + "epoch": 0.39138176638176636, + "grad_norm": 0.4938857853412628, + "learning_rate": 0.00019538875003119113, + "loss": 1.1061, + "step": 2198 + }, + { + "epoch": 0.39155982905982906, + "grad_norm": 0.5260919332504272, + "learning_rate": 0.00019538454755460354, + "loss": 1.3292, + "step": 2199 + }, + { + "epoch": 0.39173789173789175, + "grad_norm": 0.46527108550071716, + "learning_rate": 0.00019538034320915388, + "loss": 1.2074, + "step": 2200 + }, + { + "epoch": 0.3919159544159544, + "grad_norm": 0.5608304738998413, + "learning_rate": 0.00019537613699492453, + "loss": 1.0385, + "step": 2201 + }, + { + "epoch": 0.3920940170940171, + "grad_norm": 0.5056684613227844, + "learning_rate": 0.00019537192891199792, + "loss": 1.1513, + "step": 2202 + }, + { + "epoch": 0.3922720797720798, + "grad_norm": 0.3764426112174988, + "learning_rate": 0.00019536771896045644, + "loss": 0.8966, + "step": 2203 + }, + { + "epoch": 0.3924501424501424, + "grad_norm": 0.4983638823032379, + "learning_rate": 0.0001953635071403827, + "loss": 1.097, + "step": 2204 + }, + { + "epoch": 0.3926282051282051, + "grad_norm": 0.5733919739723206, + "learning_rate": 0.00019535929345185904, + "loss": 1.4992, + "step": 2205 + }, + { + "epoch": 0.3928062678062678, + "grad_norm": 0.632064163684845, + "learning_rate": 0.00019535507789496817, + "loss": 1.0611, + "step": 2206 + }, + { + "epoch": 0.39298433048433046, + "grad_norm": 0.409978449344635, + "learning_rate": 0.00019535086046979262, + "loss": 0.7172, + "step": 2207 + }, + { + "epoch": 0.39316239316239315, + "grad_norm": 0.40910813212394714, + "learning_rate": 0.00019534664117641502, + "loss": 0.8803, + "step": 2208 + }, + { + "epoch": 0.39334045584045585, + "grad_norm": 0.4696179926395416, + "learning_rate": 0.00019534242001491807, + "loss": 1.1551, + "step": 2209 + }, + { + "epoch": 0.39351851851851855, + "grad_norm": 0.538425862789154, + "learning_rate": 0.00019533819698538444, + "loss": 1.1296, + "step": 2210 + }, + { + "epoch": 0.3936965811965812, + "grad_norm": 0.5913630723953247, + "learning_rate": 0.00019533397208789692, + "loss": 0.9757, + "step": 2211 + }, + { + "epoch": 0.3938746438746439, + "grad_norm": 0.5649870038032532, + "learning_rate": 0.00019532974532253822, + "loss": 0.9976, + "step": 2212 + }, + { + "epoch": 0.3940527065527066, + "grad_norm": 0.5012063980102539, + "learning_rate": 0.00019532551668939121, + "loss": 0.9969, + "step": 2213 + }, + { + "epoch": 0.3942307692307692, + "grad_norm": 0.5098594427108765, + "learning_rate": 0.00019532128618853872, + "loss": 1.1229, + "step": 2214 + }, + { + "epoch": 0.3944088319088319, + "grad_norm": 0.4753342568874359, + "learning_rate": 0.0001953170538200636, + "loss": 1.0808, + "step": 2215 + }, + { + "epoch": 0.3945868945868946, + "grad_norm": 0.4770098626613617, + "learning_rate": 0.00019531281958404888, + "loss": 1.0656, + "step": 2216 + }, + { + "epoch": 0.39476495726495725, + "grad_norm": 0.6007979512214661, + "learning_rate": 0.00019530858348057746, + "loss": 1.0093, + "step": 2217 + }, + { + "epoch": 0.39494301994301995, + "grad_norm": 0.4501650929450989, + "learning_rate": 0.00019530434550973227, + "loss": 0.8557, + "step": 2218 + }, + { + "epoch": 0.39512108262108264, + "grad_norm": 0.5123980641365051, + "learning_rate": 0.00019530010567159645, + "loss": 0.9833, + "step": 2219 + }, + { + "epoch": 0.3952991452991453, + "grad_norm": 0.4623969495296478, + "learning_rate": 0.000195295863966253, + "loss": 0.913, + "step": 2220 + }, + { + "epoch": 0.395477207977208, + "grad_norm": 0.4341880679130554, + "learning_rate": 0.0001952916203937851, + "loss": 1.0234, + "step": 2221 + }, + { + "epoch": 0.3956552706552707, + "grad_norm": 0.5935006141662598, + "learning_rate": 0.00019528737495427581, + "loss": 1.061, + "step": 2222 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 0.44835174083709717, + "learning_rate": 0.00019528312764780837, + "loss": 1.1567, + "step": 2223 + }, + { + "epoch": 0.396011396011396, + "grad_norm": 0.5476976633071899, + "learning_rate": 0.00019527887847446595, + "loss": 1.2304, + "step": 2224 + }, + { + "epoch": 0.3961894586894587, + "grad_norm": 0.4487939774990082, + "learning_rate": 0.00019527462743433187, + "loss": 1.1813, + "step": 2225 + }, + { + "epoch": 0.39636752136752135, + "grad_norm": 0.4053241014480591, + "learning_rate": 0.00019527037452748936, + "loss": 0.7899, + "step": 2226 + }, + { + "epoch": 0.39654558404558404, + "grad_norm": 0.534570574760437, + "learning_rate": 0.00019526611975402176, + "loss": 1.0681, + "step": 2227 + }, + { + "epoch": 0.39672364672364674, + "grad_norm": 0.46096158027648926, + "learning_rate": 0.00019526186311401246, + "loss": 0.9234, + "step": 2228 + }, + { + "epoch": 0.3969017094017094, + "grad_norm": 0.47363516688346863, + "learning_rate": 0.00019525760460754483, + "loss": 1.0197, + "step": 2229 + }, + { + "epoch": 0.3970797720797721, + "grad_norm": 0.46317258477211, + "learning_rate": 0.00019525334423470234, + "loss": 1.2103, + "step": 2230 + }, + { + "epoch": 0.39725783475783477, + "grad_norm": 0.4924237132072449, + "learning_rate": 0.0001952490819955684, + "loss": 1.3299, + "step": 2231 + }, + { + "epoch": 0.3974358974358974, + "grad_norm": 0.5419978499412537, + "learning_rate": 0.0001952448178902266, + "loss": 1.2526, + "step": 2232 + }, + { + "epoch": 0.3976139601139601, + "grad_norm": 0.5003267526626587, + "learning_rate": 0.00019524055191876043, + "loss": 1.1073, + "step": 2233 + }, + { + "epoch": 0.3977920227920228, + "grad_norm": 0.621789276599884, + "learning_rate": 0.00019523628408125347, + "loss": 1.3409, + "step": 2234 + }, + { + "epoch": 0.39797008547008544, + "grad_norm": 0.44235602021217346, + "learning_rate": 0.0001952320143777894, + "loss": 0.9799, + "step": 2235 + }, + { + "epoch": 0.39814814814814814, + "grad_norm": 0.49954718351364136, + "learning_rate": 0.0001952277428084518, + "loss": 1.2227, + "step": 2236 + }, + { + "epoch": 0.39832621082621084, + "grad_norm": 0.5113739967346191, + "learning_rate": 0.00019522346937332443, + "loss": 1.1644, + "step": 2237 + }, + { + "epoch": 0.39850427350427353, + "grad_norm": 0.5026139616966248, + "learning_rate": 0.00019521919407249096, + "loss": 1.0823, + "step": 2238 + }, + { + "epoch": 0.39868233618233617, + "grad_norm": 0.4943205714225769, + "learning_rate": 0.0001952149169060352, + "loss": 1.0961, + "step": 2239 + }, + { + "epoch": 0.39886039886039887, + "grad_norm": 0.4680631458759308, + "learning_rate": 0.00019521063787404094, + "loss": 0.9787, + "step": 2240 + }, + { + "epoch": 0.39903846153846156, + "grad_norm": 0.5511566400527954, + "learning_rate": 0.00019520635697659202, + "loss": 1.2543, + "step": 2241 + }, + { + "epoch": 0.3992165242165242, + "grad_norm": 0.5494263172149658, + "learning_rate": 0.00019520207421377229, + "loss": 1.1978, + "step": 2242 + }, + { + "epoch": 0.3993945868945869, + "grad_norm": 0.4850340485572815, + "learning_rate": 0.00019519778958566568, + "loss": 0.8531, + "step": 2243 + }, + { + "epoch": 0.3995726495726496, + "grad_norm": 0.47168150544166565, + "learning_rate": 0.00019519350309235613, + "loss": 1.0746, + "step": 2244 + }, + { + "epoch": 0.39975071225071224, + "grad_norm": 0.571133553981781, + "learning_rate": 0.00019518921473392765, + "loss": 1.2984, + "step": 2245 + }, + { + "epoch": 0.39992877492877493, + "grad_norm": 0.4636089503765106, + "learning_rate": 0.00019518492451046427, + "loss": 1.019, + "step": 2246 + }, + { + "epoch": 0.40010683760683763, + "grad_norm": 0.4573518931865692, + "learning_rate": 0.00019518063242205, + "loss": 1.1042, + "step": 2247 + }, + { + "epoch": 0.40028490028490027, + "grad_norm": 0.49098989367485046, + "learning_rate": 0.00019517633846876894, + "loss": 1.1224, + "step": 2248 + }, + { + "epoch": 0.40046296296296297, + "grad_norm": 0.5475491881370544, + "learning_rate": 0.00019517204265070523, + "loss": 1.0984, + "step": 2249 + }, + { + "epoch": 0.40064102564102566, + "grad_norm": 0.45498281717300415, + "learning_rate": 0.00019516774496794307, + "loss": 0.8883, + "step": 2250 + }, + { + "epoch": 0.4008190883190883, + "grad_norm": 0.4908423125743866, + "learning_rate": 0.00019516344542056666, + "loss": 1.328, + "step": 2251 + }, + { + "epoch": 0.400997150997151, + "grad_norm": 0.5474920272827148, + "learning_rate": 0.0001951591440086602, + "loss": 1.3825, + "step": 2252 + }, + { + "epoch": 0.4011752136752137, + "grad_norm": 0.5165615081787109, + "learning_rate": 0.000195154840732308, + "loss": 1.33, + "step": 2253 + }, + { + "epoch": 0.40135327635327633, + "grad_norm": 0.5185585021972656, + "learning_rate": 0.00019515053559159435, + "loss": 1.1689, + "step": 2254 + }, + { + "epoch": 0.40153133903133903, + "grad_norm": 0.5468854904174805, + "learning_rate": 0.00019514622858660363, + "loss": 1.2708, + "step": 2255 + }, + { + "epoch": 0.4017094017094017, + "grad_norm": 0.47556906938552856, + "learning_rate": 0.0001951419197174202, + "loss": 1.0488, + "step": 2256 + }, + { + "epoch": 0.40188746438746437, + "grad_norm": 0.5521323084831238, + "learning_rate": 0.0001951376089841285, + "loss": 1.0868, + "step": 2257 + }, + { + "epoch": 0.40206552706552706, + "grad_norm": 0.6029638051986694, + "learning_rate": 0.00019513329638681296, + "loss": 1.1735, + "step": 2258 + }, + { + "epoch": 0.40224358974358976, + "grad_norm": 0.4897766411304474, + "learning_rate": 0.00019512898192555812, + "loss": 1.1687, + "step": 2259 + }, + { + "epoch": 0.4024216524216524, + "grad_norm": 0.45527184009552, + "learning_rate": 0.00019512466560044848, + "loss": 1.0352, + "step": 2260 + }, + { + "epoch": 0.4025997150997151, + "grad_norm": 0.5025625824928284, + "learning_rate": 0.00019512034741156863, + "loss": 1.2503, + "step": 2261 + }, + { + "epoch": 0.4027777777777778, + "grad_norm": 0.46415451169013977, + "learning_rate": 0.00019511602735900317, + "loss": 1.032, + "step": 2262 + }, + { + "epoch": 0.40295584045584043, + "grad_norm": 0.4812934398651123, + "learning_rate": 0.00019511170544283678, + "loss": 1.0523, + "step": 2263 + }, + { + "epoch": 0.4031339031339031, + "grad_norm": 0.49937039613723755, + "learning_rate": 0.00019510738166315404, + "loss": 1.2238, + "step": 2264 + }, + { + "epoch": 0.4033119658119658, + "grad_norm": 0.5428698062896729, + "learning_rate": 0.00019510305602003975, + "loss": 1.0361, + "step": 2265 + }, + { + "epoch": 0.40349002849002846, + "grad_norm": 0.44836854934692383, + "learning_rate": 0.0001950987285135786, + "loss": 1.169, + "step": 2266 + }, + { + "epoch": 0.40366809116809116, + "grad_norm": 0.5071489214897156, + "learning_rate": 0.00019509439914385549, + "loss": 1.1567, + "step": 2267 + }, + { + "epoch": 0.40384615384615385, + "grad_norm": 0.5204613208770752, + "learning_rate": 0.00019509006791095513, + "loss": 0.9949, + "step": 2268 + }, + { + "epoch": 0.40402421652421655, + "grad_norm": 0.4583234488964081, + "learning_rate": 0.00019508573481496238, + "loss": 0.9051, + "step": 2269 + }, + { + "epoch": 0.4042022792022792, + "grad_norm": 0.5436791181564331, + "learning_rate": 0.00019508139985596222, + "loss": 1.3239, + "step": 2270 + }, + { + "epoch": 0.4043803418803419, + "grad_norm": 0.48774269223213196, + "learning_rate": 0.00019507706303403954, + "loss": 1.2102, + "step": 2271 + }, + { + "epoch": 0.4045584045584046, + "grad_norm": 0.4742540717124939, + "learning_rate": 0.00019507272434927933, + "loss": 1.1137, + "step": 2272 + }, + { + "epoch": 0.4047364672364672, + "grad_norm": 0.531148374080658, + "learning_rate": 0.00019506838380176658, + "loss": 1.3162, + "step": 2273 + }, + { + "epoch": 0.4049145299145299, + "grad_norm": 0.5002314448356628, + "learning_rate": 0.0001950640413915863, + "loss": 1.0743, + "step": 2274 + }, + { + "epoch": 0.4050925925925926, + "grad_norm": 0.39826446771621704, + "learning_rate": 0.00019505969711882366, + "loss": 0.7698, + "step": 2275 + }, + { + "epoch": 0.40527065527065526, + "grad_norm": 0.5177471041679382, + "learning_rate": 0.00019505535098356371, + "loss": 1.1821, + "step": 2276 + }, + { + "epoch": 0.40544871794871795, + "grad_norm": 0.467241108417511, + "learning_rate": 0.00019505100298589158, + "loss": 0.8036, + "step": 2277 + }, + { + "epoch": 0.40562678062678065, + "grad_norm": 0.43711844086647034, + "learning_rate": 0.00019504665312589255, + "loss": 0.8667, + "step": 2278 + }, + { + "epoch": 0.4058048433048433, + "grad_norm": 0.4929116368293762, + "learning_rate": 0.00019504230140365177, + "loss": 1.1279, + "step": 2279 + }, + { + "epoch": 0.405982905982906, + "grad_norm": 0.5279183983802795, + "learning_rate": 0.00019503794781925452, + "loss": 1.1318, + "step": 2280 + }, + { + "epoch": 0.4061609686609687, + "grad_norm": 0.549217939376831, + "learning_rate": 0.00019503359237278608, + "loss": 1.2007, + "step": 2281 + }, + { + "epoch": 0.4063390313390313, + "grad_norm": 0.5485880374908447, + "learning_rate": 0.00019502923506433187, + "loss": 1.1079, + "step": 2282 + }, + { + "epoch": 0.406517094017094, + "grad_norm": 0.48379644751548767, + "learning_rate": 0.0001950248758939772, + "loss": 0.9978, + "step": 2283 + }, + { + "epoch": 0.4066951566951567, + "grad_norm": 0.5943657755851746, + "learning_rate": 0.00019502051486180744, + "loss": 1.0466, + "step": 2284 + }, + { + "epoch": 0.40687321937321935, + "grad_norm": 0.5721273422241211, + "learning_rate": 0.00019501615196790812, + "loss": 1.2674, + "step": 2285 + }, + { + "epoch": 0.40705128205128205, + "grad_norm": 0.47624221444129944, + "learning_rate": 0.00019501178721236464, + "loss": 1.089, + "step": 2286 + }, + { + "epoch": 0.40722934472934474, + "grad_norm": 0.5091297030448914, + "learning_rate": 0.0001950074205952626, + "loss": 1.2035, + "step": 2287 + }, + { + "epoch": 0.4074074074074074, + "grad_norm": 0.45206236839294434, + "learning_rate": 0.0001950030521166875, + "loss": 0.9188, + "step": 2288 + }, + { + "epoch": 0.4075854700854701, + "grad_norm": 0.5563844442367554, + "learning_rate": 0.00019499868177672497, + "loss": 1.3444, + "step": 2289 + }, + { + "epoch": 0.4077635327635328, + "grad_norm": 0.4971138536930084, + "learning_rate": 0.00019499430957546055, + "loss": 1.1615, + "step": 2290 + }, + { + "epoch": 0.4079415954415954, + "grad_norm": 0.49355944991111755, + "learning_rate": 0.00019498993551298, + "loss": 1.1528, + "step": 2291 + }, + { + "epoch": 0.4081196581196581, + "grad_norm": 0.534705638885498, + "learning_rate": 0.000194985559589369, + "loss": 1.197, + "step": 2292 + }, + { + "epoch": 0.4082977207977208, + "grad_norm": 0.5113020539283752, + "learning_rate": 0.0001949811818047133, + "loss": 1.109, + "step": 2293 + }, + { + "epoch": 0.40847578347578345, + "grad_norm": 0.4823366701602936, + "learning_rate": 0.00019497680215909858, + "loss": 1.168, + "step": 2294 + }, + { + "epoch": 0.40865384615384615, + "grad_norm": 0.500792920589447, + "learning_rate": 0.00019497242065261077, + "loss": 1.1567, + "step": 2295 + }, + { + "epoch": 0.40883190883190884, + "grad_norm": 0.5047918558120728, + "learning_rate": 0.00019496803728533566, + "loss": 1.0515, + "step": 2296 + }, + { + "epoch": 0.40900997150997154, + "grad_norm": 0.474624365568161, + "learning_rate": 0.00019496365205735913, + "loss": 1.1747, + "step": 2297 + }, + { + "epoch": 0.4091880341880342, + "grad_norm": 0.5522183179855347, + "learning_rate": 0.0001949592649687671, + "loss": 1.1506, + "step": 2298 + }, + { + "epoch": 0.4093660968660969, + "grad_norm": 0.4526083767414093, + "learning_rate": 0.00019495487601964553, + "loss": 0.9968, + "step": 2299 + }, + { + "epoch": 0.40954415954415957, + "grad_norm": 0.545845091342926, + "learning_rate": 0.00019495048521008044, + "loss": 1.146, + "step": 2300 + }, + { + "epoch": 0.4097222222222222, + "grad_norm": 0.5475544333457947, + "learning_rate": 0.00019494609254015784, + "loss": 1.0101, + "step": 2301 + }, + { + "epoch": 0.4099002849002849, + "grad_norm": 0.43419042229652405, + "learning_rate": 0.00019494169800996373, + "loss": 1.065, + "step": 2302 + }, + { + "epoch": 0.4100783475783476, + "grad_norm": 0.44998374581336975, + "learning_rate": 0.00019493730161958435, + "loss": 0.9948, + "step": 2303 + }, + { + "epoch": 0.41025641025641024, + "grad_norm": 0.5401661992073059, + "learning_rate": 0.0001949329033691057, + "loss": 1.0473, + "step": 2304 + }, + { + "epoch": 0.41043447293447294, + "grad_norm": 0.48064103722572327, + "learning_rate": 0.00019492850325861404, + "loss": 1.0486, + "step": 2305 + }, + { + "epoch": 0.41061253561253563, + "grad_norm": 0.5398300290107727, + "learning_rate": 0.00019492410128819557, + "loss": 1.0314, + "step": 2306 + }, + { + "epoch": 0.4107905982905983, + "grad_norm": 0.4771125912666321, + "learning_rate": 0.0001949196974579365, + "loss": 0.9855, + "step": 2307 + }, + { + "epoch": 0.41096866096866097, + "grad_norm": 0.5375809669494629, + "learning_rate": 0.00019491529176792315, + "loss": 1.0777, + "step": 2308 + }, + { + "epoch": 0.41114672364672367, + "grad_norm": 0.48424094915390015, + "learning_rate": 0.00019491088421824183, + "loss": 1.0751, + "step": 2309 + }, + { + "epoch": 0.4113247863247863, + "grad_norm": 0.5054880380630493, + "learning_rate": 0.00019490647480897887, + "loss": 1.2457, + "step": 2310 + }, + { + "epoch": 0.411502849002849, + "grad_norm": 0.47118356823921204, + "learning_rate": 0.0001949020635402207, + "loss": 1.0445, + "step": 2311 + }, + { + "epoch": 0.4116809116809117, + "grad_norm": 0.47171851992607117, + "learning_rate": 0.00019489765041205375, + "loss": 1.0062, + "step": 2312 + }, + { + "epoch": 0.41185897435897434, + "grad_norm": 0.5703238844871521, + "learning_rate": 0.00019489323542456447, + "loss": 1.5639, + "step": 2313 + }, + { + "epoch": 0.41203703703703703, + "grad_norm": 0.5045075416564941, + "learning_rate": 0.00019488881857783935, + "loss": 1.1665, + "step": 2314 + }, + { + "epoch": 0.41221509971509973, + "grad_norm": 0.46835362911224365, + "learning_rate": 0.00019488439987196495, + "loss": 1.2078, + "step": 2315 + }, + { + "epoch": 0.41239316239316237, + "grad_norm": 0.5187196731567383, + "learning_rate": 0.00019487997930702785, + "loss": 1.1049, + "step": 2316 + }, + { + "epoch": 0.41257122507122507, + "grad_norm": 0.5190554857254028, + "learning_rate": 0.00019487555688311463, + "loss": 1.331, + "step": 2317 + }, + { + "epoch": 0.41274928774928776, + "grad_norm": 0.7394969463348389, + "learning_rate": 0.00019487113260031197, + "loss": 0.9646, + "step": 2318 + }, + { + "epoch": 0.4129273504273504, + "grad_norm": 0.532982349395752, + "learning_rate": 0.00019486670645870656, + "loss": 1.166, + "step": 2319 + }, + { + "epoch": 0.4131054131054131, + "grad_norm": 0.48659515380859375, + "learning_rate": 0.00019486227845838509, + "loss": 1.0016, + "step": 2320 + }, + { + "epoch": 0.4132834757834758, + "grad_norm": 0.5364453196525574, + "learning_rate": 0.00019485784859943434, + "loss": 1.3877, + "step": 2321 + }, + { + "epoch": 0.41346153846153844, + "grad_norm": 0.49788740277290344, + "learning_rate": 0.0001948534168819411, + "loss": 1.2949, + "step": 2322 + }, + { + "epoch": 0.41363960113960113, + "grad_norm": 0.5125377774238586, + "learning_rate": 0.00019484898330599217, + "loss": 0.9769, + "step": 2323 + }, + { + "epoch": 0.41381766381766383, + "grad_norm": 0.5434861779212952, + "learning_rate": 0.00019484454787167447, + "loss": 1.254, + "step": 2324 + }, + { + "epoch": 0.41399572649572647, + "grad_norm": 0.5324583053588867, + "learning_rate": 0.00019484011057907487, + "loss": 0.9788, + "step": 2325 + }, + { + "epoch": 0.41417378917378916, + "grad_norm": 0.4806961715221405, + "learning_rate": 0.00019483567142828033, + "loss": 1.0089, + "step": 2326 + }, + { + "epoch": 0.41435185185185186, + "grad_norm": 0.5152947306632996, + "learning_rate": 0.0001948312304193778, + "loss": 1.15, + "step": 2327 + }, + { + "epoch": 0.41452991452991456, + "grad_norm": 0.6030138731002808, + "learning_rate": 0.0001948267875524543, + "loss": 1.196, + "step": 2328 + }, + { + "epoch": 0.4147079772079772, + "grad_norm": 0.4504946768283844, + "learning_rate": 0.0001948223428275969, + "loss": 0.8742, + "step": 2329 + }, + { + "epoch": 0.4148860398860399, + "grad_norm": 0.5195745825767517, + "learning_rate": 0.00019481789624489263, + "loss": 1.0104, + "step": 2330 + }, + { + "epoch": 0.4150641025641026, + "grad_norm": 0.5269250869750977, + "learning_rate": 0.0001948134478044287, + "loss": 1.2284, + "step": 2331 + }, + { + "epoch": 0.41524216524216523, + "grad_norm": 0.5302315354347229, + "learning_rate": 0.00019480899750629218, + "loss": 1.1374, + "step": 2332 + }, + { + "epoch": 0.4154202279202279, + "grad_norm": 0.5501471161842346, + "learning_rate": 0.0001948045453505703, + "loss": 1.214, + "step": 2333 + }, + { + "epoch": 0.4155982905982906, + "grad_norm": 0.4674588739871979, + "learning_rate": 0.0001948000913373503, + "loss": 1.0568, + "step": 2334 + }, + { + "epoch": 0.41577635327635326, + "grad_norm": 0.5262266993522644, + "learning_rate": 0.0001947956354667195, + "loss": 1.111, + "step": 2335 + }, + { + "epoch": 0.41595441595441596, + "grad_norm": 0.4549071788787842, + "learning_rate": 0.00019479117773876507, + "loss": 1.2655, + "step": 2336 + }, + { + "epoch": 0.41613247863247865, + "grad_norm": 0.48897311091423035, + "learning_rate": 0.00019478671815357447, + "loss": 1.0543, + "step": 2337 + }, + { + "epoch": 0.4163105413105413, + "grad_norm": 0.5544867515563965, + "learning_rate": 0.000194782256711235, + "loss": 1.2276, + "step": 2338 + }, + { + "epoch": 0.416488603988604, + "grad_norm": 0.5050773024559021, + "learning_rate": 0.0001947777934118341, + "loss": 0.9781, + "step": 2339 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 0.4831899106502533, + "learning_rate": 0.00019477332825545925, + "loss": 1.0213, + "step": 2340 + }, + { + "epoch": 0.4168447293447293, + "grad_norm": 0.5392552614212036, + "learning_rate": 0.0001947688612421979, + "loss": 1.3251, + "step": 2341 + }, + { + "epoch": 0.417022792022792, + "grad_norm": 0.5003608465194702, + "learning_rate": 0.00019476439237213754, + "loss": 1.0714, + "step": 2342 + }, + { + "epoch": 0.4172008547008547, + "grad_norm": 0.5016986727714539, + "learning_rate": 0.00019475992164536582, + "loss": 1.0656, + "step": 2343 + }, + { + "epoch": 0.41737891737891736, + "grad_norm": 0.5139234066009521, + "learning_rate": 0.00019475544906197024, + "loss": 1.1317, + "step": 2344 + }, + { + "epoch": 0.41755698005698005, + "grad_norm": 0.582478940486908, + "learning_rate": 0.00019475097462203847, + "loss": 1.4209, + "step": 2345 + }, + { + "epoch": 0.41773504273504275, + "grad_norm": 0.5248767137527466, + "learning_rate": 0.00019474649832565823, + "loss": 1.2965, + "step": 2346 + }, + { + "epoch": 0.4179131054131054, + "grad_norm": 0.4977390170097351, + "learning_rate": 0.00019474202017291713, + "loss": 1.3319, + "step": 2347 + }, + { + "epoch": 0.4180911680911681, + "grad_norm": 0.4868984818458557, + "learning_rate": 0.00019473754016390298, + "loss": 1.0595, + "step": 2348 + }, + { + "epoch": 0.4182692307692308, + "grad_norm": 0.5965346693992615, + "learning_rate": 0.00019473305829870353, + "loss": 1.2289, + "step": 2349 + }, + { + "epoch": 0.4184472934472934, + "grad_norm": 0.46590209007263184, + "learning_rate": 0.0001947285745774066, + "loss": 1.0468, + "step": 2350 + }, + { + "epoch": 0.4186253561253561, + "grad_norm": 0.497811883687973, + "learning_rate": 0.0001947240890001, + "loss": 1.1247, + "step": 2351 + }, + { + "epoch": 0.4188034188034188, + "grad_norm": 0.5348289012908936, + "learning_rate": 0.0001947196015668717, + "loss": 0.9496, + "step": 2352 + }, + { + "epoch": 0.41898148148148145, + "grad_norm": 0.5086174607276917, + "learning_rate": 0.0001947151122778095, + "loss": 0.8869, + "step": 2353 + }, + { + "epoch": 0.41915954415954415, + "grad_norm": 0.4844677150249481, + "learning_rate": 0.00019471062113300146, + "loss": 0.847, + "step": 2354 + }, + { + "epoch": 0.41933760683760685, + "grad_norm": 0.5395866632461548, + "learning_rate": 0.00019470612813253556, + "loss": 0.9684, + "step": 2355 + }, + { + "epoch": 0.41951566951566954, + "grad_norm": 0.479403018951416, + "learning_rate": 0.0001947016332764998, + "loss": 1.0532, + "step": 2356 + }, + { + "epoch": 0.4196937321937322, + "grad_norm": 0.5499961376190186, + "learning_rate": 0.00019469713656498227, + "loss": 1.2565, + "step": 2357 + }, + { + "epoch": 0.4198717948717949, + "grad_norm": 0.5865352153778076, + "learning_rate": 0.00019469263799807104, + "loss": 1.1349, + "step": 2358 + }, + { + "epoch": 0.4200498575498576, + "grad_norm": 0.4454309046268463, + "learning_rate": 0.00019468813757585432, + "loss": 0.9631, + "step": 2359 + }, + { + "epoch": 0.4202279202279202, + "grad_norm": 0.48426875472068787, + "learning_rate": 0.00019468363529842023, + "loss": 0.9795, + "step": 2360 + }, + { + "epoch": 0.4204059829059829, + "grad_norm": 0.47428226470947266, + "learning_rate": 0.00019467913116585697, + "loss": 0.9316, + "step": 2361 + }, + { + "epoch": 0.4205840455840456, + "grad_norm": 0.5193758010864258, + "learning_rate": 0.00019467462517825282, + "loss": 1.235, + "step": 2362 + }, + { + "epoch": 0.42076210826210825, + "grad_norm": 0.49845513701438904, + "learning_rate": 0.00019467011733569607, + "loss": 1.2413, + "step": 2363 + }, + { + "epoch": 0.42094017094017094, + "grad_norm": 0.45483845472335815, + "learning_rate": 0.00019466560763827502, + "loss": 1.2817, + "step": 2364 + }, + { + "epoch": 0.42111823361823364, + "grad_norm": 0.43345287442207336, + "learning_rate": 0.00019466109608607806, + "loss": 0.8568, + "step": 2365 + }, + { + "epoch": 0.4212962962962963, + "grad_norm": 0.4467088282108307, + "learning_rate": 0.00019465658267919352, + "loss": 1.1408, + "step": 2366 + }, + { + "epoch": 0.421474358974359, + "grad_norm": 0.6705610156059265, + "learning_rate": 0.00019465206741770992, + "loss": 1.445, + "step": 2367 + }, + { + "epoch": 0.42165242165242167, + "grad_norm": 0.5037859678268433, + "learning_rate": 0.00019464755030171565, + "loss": 0.8682, + "step": 2368 + }, + { + "epoch": 0.4218304843304843, + "grad_norm": 0.49576324224472046, + "learning_rate": 0.00019464303133129928, + "loss": 0.8387, + "step": 2369 + }, + { + "epoch": 0.422008547008547, + "grad_norm": 0.5222806334495544, + "learning_rate": 0.00019463851050654927, + "loss": 1.1443, + "step": 2370 + }, + { + "epoch": 0.4221866096866097, + "grad_norm": 0.4966863989830017, + "learning_rate": 0.00019463398782755426, + "loss": 1.1555, + "step": 2371 + }, + { + "epoch": 0.42236467236467234, + "grad_norm": 0.6140168309211731, + "learning_rate": 0.00019462946329440285, + "loss": 1.2264, + "step": 2372 + }, + { + "epoch": 0.42254273504273504, + "grad_norm": 0.4906651973724365, + "learning_rate": 0.0001946249369071837, + "loss": 1.2459, + "step": 2373 + }, + { + "epoch": 0.42272079772079774, + "grad_norm": 0.5956700444221497, + "learning_rate": 0.00019462040866598544, + "loss": 1.1521, + "step": 2374 + }, + { + "epoch": 0.4228988603988604, + "grad_norm": 0.46044886112213135, + "learning_rate": 0.00019461587857089687, + "loss": 1.2084, + "step": 2375 + }, + { + "epoch": 0.4230769230769231, + "grad_norm": 0.5109430551528931, + "learning_rate": 0.00019461134662200668, + "loss": 1.2684, + "step": 2376 + }, + { + "epoch": 0.42325498575498577, + "grad_norm": 0.4373733103275299, + "learning_rate": 0.0001946068128194037, + "loss": 1.0451, + "step": 2377 + }, + { + "epoch": 0.4234330484330484, + "grad_norm": 0.553817868232727, + "learning_rate": 0.00019460227716317673, + "loss": 1.1052, + "step": 2378 + }, + { + "epoch": 0.4236111111111111, + "grad_norm": 0.5742647647857666, + "learning_rate": 0.00019459773965341468, + "loss": 1.1647, + "step": 2379 + }, + { + "epoch": 0.4237891737891738, + "grad_norm": 0.5461940169334412, + "learning_rate": 0.00019459320029020642, + "loss": 1.0953, + "step": 2380 + }, + { + "epoch": 0.42396723646723644, + "grad_norm": 0.5837802290916443, + "learning_rate": 0.0001945886590736409, + "loss": 1.1303, + "step": 2381 + }, + { + "epoch": 0.42414529914529914, + "grad_norm": 0.5316985249519348, + "learning_rate": 0.0001945841160038071, + "loss": 1.1204, + "step": 2382 + }, + { + "epoch": 0.42432336182336183, + "grad_norm": 0.5846191048622131, + "learning_rate": 0.00019457957108079404, + "loss": 1.2622, + "step": 2383 + }, + { + "epoch": 0.42450142450142453, + "grad_norm": 0.43266957998275757, + "learning_rate": 0.00019457502430469075, + "loss": 0.9834, + "step": 2384 + }, + { + "epoch": 0.42467948717948717, + "grad_norm": 0.514081597328186, + "learning_rate": 0.00019457047567558632, + "loss": 0.8413, + "step": 2385 + }, + { + "epoch": 0.42485754985754987, + "grad_norm": 0.4831700325012207, + "learning_rate": 0.00019456592519356987, + "loss": 0.9244, + "step": 2386 + }, + { + "epoch": 0.42503561253561256, + "grad_norm": 0.5612850785255432, + "learning_rate": 0.00019456137285873057, + "loss": 0.9438, + "step": 2387 + }, + { + "epoch": 0.4252136752136752, + "grad_norm": 0.5197352766990662, + "learning_rate": 0.00019455681867115758, + "loss": 1.1095, + "step": 2388 + }, + { + "epoch": 0.4253917378917379, + "grad_norm": 0.5045261979103088, + "learning_rate": 0.00019455226263094018, + "loss": 1.0007, + "step": 2389 + }, + { + "epoch": 0.4255698005698006, + "grad_norm": 0.5167570114135742, + "learning_rate": 0.00019454770473816758, + "loss": 1.1335, + "step": 2390 + }, + { + "epoch": 0.42574786324786323, + "grad_norm": 0.49262070655822754, + "learning_rate": 0.00019454314499292913, + "loss": 1.0436, + "step": 2391 + }, + { + "epoch": 0.42592592592592593, + "grad_norm": 0.4489207863807678, + "learning_rate": 0.00019453858339531417, + "loss": 1.0138, + "step": 2392 + }, + { + "epoch": 0.4261039886039886, + "grad_norm": 0.6024920344352722, + "learning_rate": 0.00019453401994541203, + "loss": 1.1921, + "step": 2393 + }, + { + "epoch": 0.42628205128205127, + "grad_norm": 0.46807861328125, + "learning_rate": 0.00019452945464331215, + "loss": 1.0947, + "step": 2394 + }, + { + "epoch": 0.42646011396011396, + "grad_norm": 0.48776543140411377, + "learning_rate": 0.00019452488748910397, + "loss": 1.0029, + "step": 2395 + }, + { + "epoch": 0.42663817663817666, + "grad_norm": 0.4798663556575775, + "learning_rate": 0.000194520318482877, + "loss": 0.7863, + "step": 2396 + }, + { + "epoch": 0.4268162393162393, + "grad_norm": 0.5067816972732544, + "learning_rate": 0.0001945157476247207, + "loss": 1.0049, + "step": 2397 + }, + { + "epoch": 0.426994301994302, + "grad_norm": 0.5179638266563416, + "learning_rate": 0.00019451117491472468, + "loss": 1.1851, + "step": 2398 + }, + { + "epoch": 0.4271723646723647, + "grad_norm": 0.4782430827617645, + "learning_rate": 0.00019450660035297854, + "loss": 1.125, + "step": 2399 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 0.560077965259552, + "learning_rate": 0.00019450202393957186, + "loss": 1.1843, + "step": 2400 + }, + { + "epoch": 0.42752849002849, + "grad_norm": 0.5247970223426819, + "learning_rate": 0.00019449744567459436, + "loss": 1.1576, + "step": 2401 + }, + { + "epoch": 0.4277065527065527, + "grad_norm": 0.6414062976837158, + "learning_rate": 0.00019449286555813568, + "loss": 1.1833, + "step": 2402 + }, + { + "epoch": 0.42788461538461536, + "grad_norm": 0.5006586909294128, + "learning_rate": 0.00019448828359028563, + "loss": 1.1778, + "step": 2403 + }, + { + "epoch": 0.42806267806267806, + "grad_norm": 0.4946450889110565, + "learning_rate": 0.0001944836997711339, + "loss": 1.1611, + "step": 2404 + }, + { + "epoch": 0.42824074074074076, + "grad_norm": 0.4601200222969055, + "learning_rate": 0.00019447911410077037, + "loss": 1.2456, + "step": 2405 + }, + { + "epoch": 0.4284188034188034, + "grad_norm": 0.4653947651386261, + "learning_rate": 0.00019447452657928485, + "loss": 1.0941, + "step": 2406 + }, + { + "epoch": 0.4285968660968661, + "grad_norm": 0.5015713572502136, + "learning_rate": 0.00019446993720676726, + "loss": 1.3113, + "step": 2407 + }, + { + "epoch": 0.4287749287749288, + "grad_norm": 0.5803143978118896, + "learning_rate": 0.0001944653459833075, + "loss": 1.0568, + "step": 2408 + }, + { + "epoch": 0.42895299145299143, + "grad_norm": 0.5259647965431213, + "learning_rate": 0.0001944607529089955, + "loss": 1.1243, + "step": 2409 + }, + { + "epoch": 0.4291310541310541, + "grad_norm": 0.5150414109230042, + "learning_rate": 0.00019445615798392124, + "loss": 1.0676, + "step": 2410 + }, + { + "epoch": 0.4293091168091168, + "grad_norm": 0.5848649740219116, + "learning_rate": 0.0001944515612081748, + "loss": 1.0671, + "step": 2411 + }, + { + "epoch": 0.42948717948717946, + "grad_norm": 0.5696990489959717, + "learning_rate": 0.00019444696258184626, + "loss": 1.3323, + "step": 2412 + }, + { + "epoch": 0.42966524216524216, + "grad_norm": 0.49822330474853516, + "learning_rate": 0.00019444236210502567, + "loss": 1.1004, + "step": 2413 + }, + { + "epoch": 0.42984330484330485, + "grad_norm": 0.4683490991592407, + "learning_rate": 0.00019443775977780317, + "loss": 0.9768, + "step": 2414 + }, + { + "epoch": 0.43002136752136755, + "grad_norm": 0.5703811049461365, + "learning_rate": 0.00019443315560026893, + "loss": 1.154, + "step": 2415 + }, + { + "epoch": 0.4301994301994302, + "grad_norm": 0.5121861100196838, + "learning_rate": 0.0001944285495725132, + "loss": 1.1388, + "step": 2416 + }, + { + "epoch": 0.4303774928774929, + "grad_norm": 0.4864094853401184, + "learning_rate": 0.00019442394169462619, + "loss": 0.9214, + "step": 2417 + }, + { + "epoch": 0.4305555555555556, + "grad_norm": 0.5234864354133606, + "learning_rate": 0.0001944193319666982, + "loss": 1.2787, + "step": 2418 + }, + { + "epoch": 0.4307336182336182, + "grad_norm": 0.5137650370597839, + "learning_rate": 0.00019441472038881955, + "loss": 1.1406, + "step": 2419 + }, + { + "epoch": 0.4309116809116809, + "grad_norm": 0.49687784910202026, + "learning_rate": 0.00019441010696108054, + "loss": 0.93, + "step": 2420 + }, + { + "epoch": 0.4310897435897436, + "grad_norm": 0.5078722834587097, + "learning_rate": 0.00019440549168357163, + "loss": 1.1417, + "step": 2421 + }, + { + "epoch": 0.43126780626780625, + "grad_norm": 0.4483391046524048, + "learning_rate": 0.00019440087455638324, + "loss": 0.9016, + "step": 2422 + }, + { + "epoch": 0.43144586894586895, + "grad_norm": 0.5963045954704285, + "learning_rate": 0.00019439625557960576, + "loss": 1.1567, + "step": 2423 + }, + { + "epoch": 0.43162393162393164, + "grad_norm": 0.5534471273422241, + "learning_rate": 0.0001943916347533298, + "loss": 1.1409, + "step": 2424 + }, + { + "epoch": 0.4318019943019943, + "grad_norm": 0.6400241851806641, + "learning_rate": 0.0001943870120776458, + "loss": 1.2041, + "step": 2425 + }, + { + "epoch": 0.431980056980057, + "grad_norm": 0.4599420726299286, + "learning_rate": 0.0001943823875526444, + "loss": 1.023, + "step": 2426 + }, + { + "epoch": 0.4321581196581197, + "grad_norm": 0.4799708425998688, + "learning_rate": 0.00019437776117841614, + "loss": 1.0872, + "step": 2427 + }, + { + "epoch": 0.4323361823361823, + "grad_norm": 0.5138532519340515, + "learning_rate": 0.00019437313295505172, + "loss": 1.1175, + "step": 2428 + }, + { + "epoch": 0.432514245014245, + "grad_norm": 0.538223147392273, + "learning_rate": 0.00019436850288264183, + "loss": 1.1203, + "step": 2429 + }, + { + "epoch": 0.4326923076923077, + "grad_norm": 0.458044171333313, + "learning_rate": 0.00019436387096127713, + "loss": 1.0383, + "step": 2430 + }, + { + "epoch": 0.43287037037037035, + "grad_norm": 0.5928303599357605, + "learning_rate": 0.00019435923719104842, + "loss": 1.1191, + "step": 2431 + }, + { + "epoch": 0.43304843304843305, + "grad_norm": 0.5818437933921814, + "learning_rate": 0.00019435460157204645, + "loss": 1.0352, + "step": 2432 + }, + { + "epoch": 0.43322649572649574, + "grad_norm": 0.487341046333313, + "learning_rate": 0.0001943499641043621, + "loss": 1.2608, + "step": 2433 + }, + { + "epoch": 0.4334045584045584, + "grad_norm": 0.4737292230129242, + "learning_rate": 0.0001943453247880862, + "loss": 1.0084, + "step": 2434 + }, + { + "epoch": 0.4335826210826211, + "grad_norm": 0.4251207709312439, + "learning_rate": 0.0001943406836233096, + "loss": 0.9163, + "step": 2435 + }, + { + "epoch": 0.4337606837606838, + "grad_norm": 0.49468478560447693, + "learning_rate": 0.00019433604061012331, + "loss": 1.0293, + "step": 2436 + }, + { + "epoch": 0.4339387464387464, + "grad_norm": 0.47120022773742676, + "learning_rate": 0.00019433139574861826, + "loss": 1.0097, + "step": 2437 + }, + { + "epoch": 0.4341168091168091, + "grad_norm": 0.5060358047485352, + "learning_rate": 0.00019432674903888548, + "loss": 1.0683, + "step": 2438 + }, + { + "epoch": 0.4342948717948718, + "grad_norm": 0.5455917119979858, + "learning_rate": 0.00019432210048101598, + "loss": 0.8886, + "step": 2439 + }, + { + "epoch": 0.43447293447293445, + "grad_norm": 0.7960546612739563, + "learning_rate": 0.00019431745007510086, + "loss": 0.8648, + "step": 2440 + }, + { + "epoch": 0.43465099715099714, + "grad_norm": 0.5069689154624939, + "learning_rate": 0.00019431279782123126, + "loss": 1.1315, + "step": 2441 + }, + { + "epoch": 0.43482905982905984, + "grad_norm": 0.5597776174545288, + "learning_rate": 0.0001943081437194983, + "loss": 1.2281, + "step": 2442 + }, + { + "epoch": 0.43500712250712253, + "grad_norm": 0.4527420997619629, + "learning_rate": 0.00019430348776999315, + "loss": 0.7576, + "step": 2443 + }, + { + "epoch": 0.4351851851851852, + "grad_norm": 0.5625936388969421, + "learning_rate": 0.00019429882997280706, + "loss": 1.0302, + "step": 2444 + }, + { + "epoch": 0.43536324786324787, + "grad_norm": 0.5173513293266296, + "learning_rate": 0.0001942941703280313, + "loss": 1.2255, + "step": 2445 + }, + { + "epoch": 0.43554131054131057, + "grad_norm": 0.45889151096343994, + "learning_rate": 0.00019428950883575714, + "loss": 0.9322, + "step": 2446 + }, + { + "epoch": 0.4357193732193732, + "grad_norm": 0.5288477540016174, + "learning_rate": 0.00019428484549607593, + "loss": 1.0572, + "step": 2447 + }, + { + "epoch": 0.4358974358974359, + "grad_norm": 0.48328033089637756, + "learning_rate": 0.00019428018030907902, + "loss": 1.1213, + "step": 2448 + }, + { + "epoch": 0.4360754985754986, + "grad_norm": 0.5146737098693848, + "learning_rate": 0.00019427551327485786, + "loss": 0.9633, + "step": 2449 + }, + { + "epoch": 0.43625356125356124, + "grad_norm": 0.5138360261917114, + "learning_rate": 0.00019427084439350382, + "loss": 1.0561, + "step": 2450 + }, + { + "epoch": 0.43643162393162394, + "grad_norm": 0.5192533135414124, + "learning_rate": 0.00019426617366510843, + "loss": 1.1704, + "step": 2451 + }, + { + "epoch": 0.43660968660968663, + "grad_norm": 0.4819495379924774, + "learning_rate": 0.00019426150108976318, + "loss": 1.0958, + "step": 2452 + }, + { + "epoch": 0.43678774928774927, + "grad_norm": 0.4626680910587311, + "learning_rate": 0.00019425682666755965, + "loss": 1.1872, + "step": 2453 + }, + { + "epoch": 0.43696581196581197, + "grad_norm": 0.5773931741714478, + "learning_rate": 0.00019425215039858937, + "loss": 1.0722, + "step": 2454 + }, + { + "epoch": 0.43714387464387466, + "grad_norm": 0.5003872513771057, + "learning_rate": 0.00019424747228294402, + "loss": 1.0561, + "step": 2455 + }, + { + "epoch": 0.4373219373219373, + "grad_norm": 0.47370314598083496, + "learning_rate": 0.0001942427923207152, + "loss": 1.1619, + "step": 2456 + }, + { + "epoch": 0.4375, + "grad_norm": 0.466421514749527, + "learning_rate": 0.00019423811051199466, + "loss": 1.1311, + "step": 2457 + }, + { + "epoch": 0.4376780626780627, + "grad_norm": 0.44564682245254517, + "learning_rate": 0.00019423342685687413, + "loss": 1.1889, + "step": 2458 + }, + { + "epoch": 0.43785612535612534, + "grad_norm": 0.40986698865890503, + "learning_rate": 0.00019422874135544533, + "loss": 0.7312, + "step": 2459 + }, + { + "epoch": 0.43803418803418803, + "grad_norm": 0.4714358448982239, + "learning_rate": 0.0001942240540078001, + "loss": 0.9273, + "step": 2460 + }, + { + "epoch": 0.43821225071225073, + "grad_norm": 0.5298398733139038, + "learning_rate": 0.00019421936481403025, + "loss": 1.3377, + "step": 2461 + }, + { + "epoch": 0.43839031339031337, + "grad_norm": 0.6326695680618286, + "learning_rate": 0.0001942146737742277, + "loss": 1.0258, + "step": 2462 + }, + { + "epoch": 0.43856837606837606, + "grad_norm": 0.5087653994560242, + "learning_rate": 0.00019420998088848427, + "loss": 1.0007, + "step": 2463 + }, + { + "epoch": 0.43874643874643876, + "grad_norm": 0.4895429313182831, + "learning_rate": 0.00019420528615689202, + "loss": 1.0032, + "step": 2464 + }, + { + "epoch": 0.4389245014245014, + "grad_norm": 0.5029937028884888, + "learning_rate": 0.00019420058957954285, + "loss": 1.2877, + "step": 2465 + }, + { + "epoch": 0.4391025641025641, + "grad_norm": 0.4953192174434662, + "learning_rate": 0.00019419589115652884, + "loss": 1.0759, + "step": 2466 + }, + { + "epoch": 0.4392806267806268, + "grad_norm": 0.5081778168678284, + "learning_rate": 0.000194191190887942, + "loss": 0.8816, + "step": 2467 + }, + { + "epoch": 0.43945868945868943, + "grad_norm": 0.5065913200378418, + "learning_rate": 0.00019418648877387446, + "loss": 1.0362, + "step": 2468 + }, + { + "epoch": 0.43963675213675213, + "grad_norm": 0.540600061416626, + "learning_rate": 0.00019418178481441832, + "loss": 1.0911, + "step": 2469 + }, + { + "epoch": 0.4398148148148148, + "grad_norm": 0.5122954845428467, + "learning_rate": 0.00019417707900966572, + "loss": 0.9866, + "step": 2470 + }, + { + "epoch": 0.43999287749287747, + "grad_norm": 0.5380190014839172, + "learning_rate": 0.00019417237135970893, + "loss": 1.2775, + "step": 2471 + }, + { + "epoch": 0.44017094017094016, + "grad_norm": 1.2977570295333862, + "learning_rate": 0.00019416766186464016, + "loss": 1.3993, + "step": 2472 + }, + { + "epoch": 0.44034900284900286, + "grad_norm": 0.48105308413505554, + "learning_rate": 0.00019416295052455165, + "loss": 0.9369, + "step": 2473 + }, + { + "epoch": 0.44052706552706555, + "grad_norm": 0.4742157459259033, + "learning_rate": 0.00019415823733953574, + "loss": 1.101, + "step": 2474 + }, + { + "epoch": 0.4407051282051282, + "grad_norm": 0.4958631694316864, + "learning_rate": 0.00019415352230968473, + "loss": 0.9906, + "step": 2475 + }, + { + "epoch": 0.4408831908831909, + "grad_norm": 0.5808146595954895, + "learning_rate": 0.00019414880543509107, + "loss": 1.2315, + "step": 2476 + }, + { + "epoch": 0.4410612535612536, + "grad_norm": 0.4294755160808563, + "learning_rate": 0.00019414408671584714, + "loss": 0.8275, + "step": 2477 + }, + { + "epoch": 0.4412393162393162, + "grad_norm": 0.5346055626869202, + "learning_rate": 0.0001941393661520454, + "loss": 1.2432, + "step": 2478 + }, + { + "epoch": 0.4414173789173789, + "grad_norm": 0.5827590227127075, + "learning_rate": 0.00019413464374377833, + "loss": 1.3204, + "step": 2479 + }, + { + "epoch": 0.4415954415954416, + "grad_norm": 0.45688143372535706, + "learning_rate": 0.00019412991949113847, + "loss": 0.9307, + "step": 2480 + }, + { + "epoch": 0.44177350427350426, + "grad_norm": 0.512999415397644, + "learning_rate": 0.0001941251933942184, + "loss": 1.2808, + "step": 2481 + }, + { + "epoch": 0.44195156695156695, + "grad_norm": 0.4546334445476532, + "learning_rate": 0.00019412046545311064, + "loss": 1.0156, + "step": 2482 + }, + { + "epoch": 0.44212962962962965, + "grad_norm": 0.48552581667900085, + "learning_rate": 0.00019411573566790793, + "loss": 1.3798, + "step": 2483 + }, + { + "epoch": 0.4423076923076923, + "grad_norm": 0.511970579624176, + "learning_rate": 0.00019411100403870287, + "loss": 1.065, + "step": 2484 + }, + { + "epoch": 0.442485754985755, + "grad_norm": 0.6367824077606201, + "learning_rate": 0.00019410627056558815, + "loss": 1.3242, + "step": 2485 + }, + { + "epoch": 0.4426638176638177, + "grad_norm": 0.48913368582725525, + "learning_rate": 0.00019410153524865659, + "loss": 0.9761, + "step": 2486 + }, + { + "epoch": 0.4428418803418803, + "grad_norm": 0.5077710151672363, + "learning_rate": 0.0001940967980880009, + "loss": 1.1023, + "step": 2487 + }, + { + "epoch": 0.443019943019943, + "grad_norm": 0.4956335723400116, + "learning_rate": 0.00019409205908371395, + "loss": 1.1788, + "step": 2488 + }, + { + "epoch": 0.4431980056980057, + "grad_norm": 0.4726616442203522, + "learning_rate": 0.00019408731823588853, + "loss": 1.1445, + "step": 2489 + }, + { + "epoch": 0.44337606837606836, + "grad_norm": 0.5676438212394714, + "learning_rate": 0.00019408257554461757, + "loss": 1.0344, + "step": 2490 + }, + { + "epoch": 0.44355413105413105, + "grad_norm": 0.537656843662262, + "learning_rate": 0.000194077831009994, + "loss": 0.9876, + "step": 2491 + }, + { + "epoch": 0.44373219373219375, + "grad_norm": 0.517905592918396, + "learning_rate": 0.00019407308463211074, + "loss": 1.1389, + "step": 2492 + }, + { + "epoch": 0.4439102564102564, + "grad_norm": 0.49227026104927063, + "learning_rate": 0.0001940683364110608, + "loss": 1.0351, + "step": 2493 + }, + { + "epoch": 0.4440883190883191, + "grad_norm": 0.5131173729896545, + "learning_rate": 0.00019406358634693725, + "loss": 1.0351, + "step": 2494 + }, + { + "epoch": 0.4442663817663818, + "grad_norm": 0.5064495205879211, + "learning_rate": 0.0001940588344398331, + "loss": 1.0248, + "step": 2495 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 0.44107526540756226, + "learning_rate": 0.00019405408068984148, + "loss": 0.8068, + "step": 2496 + }, + { + "epoch": 0.4446225071225071, + "grad_norm": 0.6711848378181458, + "learning_rate": 0.00019404932509705554, + "loss": 1.059, + "step": 2497 + }, + { + "epoch": 0.4448005698005698, + "grad_norm": 0.5862596035003662, + "learning_rate": 0.00019404456766156845, + "loss": 1.2012, + "step": 2498 + }, + { + "epoch": 0.44497863247863245, + "grad_norm": 0.5528512001037598, + "learning_rate": 0.0001940398083834734, + "loss": 1.1121, + "step": 2499 + }, + { + "epoch": 0.44515669515669515, + "grad_norm": 0.5326655507087708, + "learning_rate": 0.0001940350472628637, + "loss": 1.166, + "step": 2500 + }, + { + "epoch": 0.44533475783475784, + "grad_norm": 0.5384873747825623, + "learning_rate": 0.00019403028429983252, + "loss": 1.4111, + "step": 2501 + }, + { + "epoch": 0.44551282051282054, + "grad_norm": 0.5142310857772827, + "learning_rate": 0.0001940255194944733, + "loss": 1.3353, + "step": 2502 + }, + { + "epoch": 0.4456908831908832, + "grad_norm": 0.49124231934547424, + "learning_rate": 0.0001940207528468793, + "loss": 1.1443, + "step": 2503 + }, + { + "epoch": 0.4458689458689459, + "grad_norm": 0.509713888168335, + "learning_rate": 0.000194015984357144, + "loss": 1.1857, + "step": 2504 + }, + { + "epoch": 0.44604700854700857, + "grad_norm": 0.5211275219917297, + "learning_rate": 0.00019401121402536078, + "loss": 0.9911, + "step": 2505 + }, + { + "epoch": 0.4462250712250712, + "grad_norm": 0.480340838432312, + "learning_rate": 0.00019400644185162312, + "loss": 1.1018, + "step": 2506 + }, + { + "epoch": 0.4464031339031339, + "grad_norm": 0.4212559163570404, + "learning_rate": 0.00019400166783602448, + "loss": 0.7501, + "step": 2507 + }, + { + "epoch": 0.4465811965811966, + "grad_norm": 0.5110511183738708, + "learning_rate": 0.00019399689197865846, + "loss": 1.1244, + "step": 2508 + }, + { + "epoch": 0.44675925925925924, + "grad_norm": 0.5604230165481567, + "learning_rate": 0.0001939921142796186, + "loss": 1.1066, + "step": 2509 + }, + { + "epoch": 0.44693732193732194, + "grad_norm": 0.5578675270080566, + "learning_rate": 0.0001939873347389985, + "loss": 1.0514, + "step": 2510 + }, + { + "epoch": 0.44711538461538464, + "grad_norm": 0.520908772945404, + "learning_rate": 0.00019398255335689184, + "loss": 1.1217, + "step": 2511 + }, + { + "epoch": 0.4472934472934473, + "grad_norm": 0.4405131936073303, + "learning_rate": 0.00019397777013339224, + "loss": 1.043, + "step": 2512 + }, + { + "epoch": 0.44747150997151, + "grad_norm": 0.5217751860618591, + "learning_rate": 0.0001939729850685935, + "loss": 1.1301, + "step": 2513 + }, + { + "epoch": 0.44764957264957267, + "grad_norm": 0.6151493191719055, + "learning_rate": 0.00019396819816258932, + "loss": 1.3498, + "step": 2514 + }, + { + "epoch": 0.4478276353276353, + "grad_norm": 0.5622836947441101, + "learning_rate": 0.0001939634094154735, + "loss": 1.146, + "step": 2515 + }, + { + "epoch": 0.448005698005698, + "grad_norm": 0.4671688973903656, + "learning_rate": 0.00019395861882733984, + "loss": 0.9456, + "step": 2516 + }, + { + "epoch": 0.4481837606837607, + "grad_norm": 0.453951358795166, + "learning_rate": 0.00019395382639828223, + "loss": 1.0042, + "step": 2517 + }, + { + "epoch": 0.44836182336182334, + "grad_norm": 0.5150699615478516, + "learning_rate": 0.0001939490321283946, + "loss": 1.166, + "step": 2518 + }, + { + "epoch": 0.44853988603988604, + "grad_norm": 0.5718298554420471, + "learning_rate": 0.0001939442360177708, + "loss": 1.2033, + "step": 2519 + }, + { + "epoch": 0.44871794871794873, + "grad_norm": 0.5306782126426697, + "learning_rate": 0.00019393943806650488, + "loss": 1.0765, + "step": 2520 + }, + { + "epoch": 0.4488960113960114, + "grad_norm": 0.47633033990859985, + "learning_rate": 0.0001939346382746908, + "loss": 0.9957, + "step": 2521 + }, + { + "epoch": 0.44907407407407407, + "grad_norm": 0.496441513299942, + "learning_rate": 0.00019392983664242262, + "loss": 1.2016, + "step": 2522 + }, + { + "epoch": 0.44925213675213677, + "grad_norm": 0.45956477522850037, + "learning_rate": 0.00019392503316979442, + "loss": 1.026, + "step": 2523 + }, + { + "epoch": 0.4494301994301994, + "grad_norm": 0.5400575995445251, + "learning_rate": 0.0001939202278569003, + "loss": 1.0785, + "step": 2524 + }, + { + "epoch": 0.4496082621082621, + "grad_norm": 0.4847868084907532, + "learning_rate": 0.00019391542070383442, + "loss": 1.013, + "step": 2525 + }, + { + "epoch": 0.4497863247863248, + "grad_norm": 0.4694063663482666, + "learning_rate": 0.00019391061171069094, + "loss": 0.8793, + "step": 2526 + }, + { + "epoch": 0.44996438746438744, + "grad_norm": 0.5158169269561768, + "learning_rate": 0.00019390580087756413, + "loss": 0.9602, + "step": 2527 + }, + { + "epoch": 0.45014245014245013, + "grad_norm": 0.5404585003852844, + "learning_rate": 0.00019390098820454822, + "loss": 1.2247, + "step": 2528 + }, + { + "epoch": 0.45032051282051283, + "grad_norm": 0.5302738547325134, + "learning_rate": 0.00019389617369173752, + "loss": 0.918, + "step": 2529 + }, + { + "epoch": 0.45049857549857547, + "grad_norm": 0.5065485835075378, + "learning_rate": 0.00019389135733922634, + "loss": 1.0934, + "step": 2530 + }, + { + "epoch": 0.45067663817663817, + "grad_norm": 0.5491471886634827, + "learning_rate": 0.00019388653914710903, + "loss": 1.0736, + "step": 2531 + }, + { + "epoch": 0.45085470085470086, + "grad_norm": 0.4850206971168518, + "learning_rate": 0.00019388171911548005, + "loss": 1.2401, + "step": 2532 + }, + { + "epoch": 0.45103276353276356, + "grad_norm": 0.5419789552688599, + "learning_rate": 0.0001938768972444338, + "loss": 1.269, + "step": 2533 + }, + { + "epoch": 0.4512108262108262, + "grad_norm": 0.4209023714065552, + "learning_rate": 0.00019387207353406476, + "loss": 1.0544, + "step": 2534 + }, + { + "epoch": 0.4513888888888889, + "grad_norm": 0.578588604927063, + "learning_rate": 0.00019386724798446743, + "loss": 1.0564, + "step": 2535 + }, + { + "epoch": 0.4515669515669516, + "grad_norm": 0.5277524590492249, + "learning_rate": 0.00019386242059573638, + "loss": 1.1497, + "step": 2536 + }, + { + "epoch": 0.45174501424501423, + "grad_norm": 0.5536073446273804, + "learning_rate": 0.0001938575913679662, + "loss": 1.2213, + "step": 2537 + }, + { + "epoch": 0.4519230769230769, + "grad_norm": 0.5572254657745361, + "learning_rate": 0.00019385276030125143, + "loss": 1.0231, + "step": 2538 + }, + { + "epoch": 0.4521011396011396, + "grad_norm": 0.493847131729126, + "learning_rate": 0.00019384792739568686, + "loss": 0.9385, + "step": 2539 + }, + { + "epoch": 0.45227920227920226, + "grad_norm": 0.4641396403312683, + "learning_rate": 0.00019384309265136707, + "loss": 0.9332, + "step": 2540 + }, + { + "epoch": 0.45245726495726496, + "grad_norm": 0.5439442992210388, + "learning_rate": 0.00019383825606838681, + "loss": 1.317, + "step": 2541 + }, + { + "epoch": 0.45263532763532766, + "grad_norm": 0.7050970792770386, + "learning_rate": 0.00019383341764684086, + "loss": 0.9508, + "step": 2542 + }, + { + "epoch": 0.4528133903133903, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.000193828577386824, + "loss": 1.2704, + "step": 2543 + }, + { + "epoch": 0.452991452991453, + "grad_norm": 0.47641924023628235, + "learning_rate": 0.0001938237352884311, + "loss": 1.0101, + "step": 2544 + }, + { + "epoch": 0.4531695156695157, + "grad_norm": 0.5223637819290161, + "learning_rate": 0.000193818891351757, + "loss": 1.0548, + "step": 2545 + }, + { + "epoch": 0.45334757834757833, + "grad_norm": 0.49065667390823364, + "learning_rate": 0.0001938140455768966, + "loss": 1.0927, + "step": 2546 + }, + { + "epoch": 0.453525641025641, + "grad_norm": 0.4808312654495239, + "learning_rate": 0.0001938091979639449, + "loss": 1.0599, + "step": 2547 + }, + { + "epoch": 0.4537037037037037, + "grad_norm": 0.5157489776611328, + "learning_rate": 0.0001938043485129968, + "loss": 1.2596, + "step": 2548 + }, + { + "epoch": 0.45388176638176636, + "grad_norm": 0.5983387231826782, + "learning_rate": 0.0001937994972241474, + "loss": 1.2276, + "step": 2549 + }, + { + "epoch": 0.45405982905982906, + "grad_norm": 0.49776506423950195, + "learning_rate": 0.00019379464409749163, + "loss": 1.3666, + "step": 2550 + }, + { + "epoch": 0.45423789173789175, + "grad_norm": 0.4693490266799927, + "learning_rate": 0.00019378978913312471, + "loss": 1.087, + "step": 2551 + }, + { + "epoch": 0.4544159544159544, + "grad_norm": 0.4754335880279541, + "learning_rate": 0.00019378493233114167, + "loss": 1.1282, + "step": 2552 + }, + { + "epoch": 0.4545940170940171, + "grad_norm": 0.5852862000465393, + "learning_rate": 0.00019378007369163776, + "loss": 1.1113, + "step": 2553 + }, + { + "epoch": 0.4547720797720798, + "grad_norm": 0.47442635893821716, + "learning_rate": 0.00019377521321470805, + "loss": 0.983, + "step": 2554 + }, + { + "epoch": 0.4549501424501424, + "grad_norm": 0.47432273626327515, + "learning_rate": 0.00019377035090044787, + "loss": 1.0169, + "step": 2555 + }, + { + "epoch": 0.4551282051282051, + "grad_norm": 0.4929196834564209, + "learning_rate": 0.00019376548674895246, + "loss": 1.0182, + "step": 2556 + }, + { + "epoch": 0.4553062678062678, + "grad_norm": 0.5433184504508972, + "learning_rate": 0.00019376062076031708, + "loss": 1.1339, + "step": 2557 + }, + { + "epoch": 0.45548433048433046, + "grad_norm": 0.47430408000946045, + "learning_rate": 0.00019375575293463715, + "loss": 1.1589, + "step": 2558 + }, + { + "epoch": 0.45566239316239315, + "grad_norm": 0.46641045808792114, + "learning_rate": 0.000193750883272008, + "loss": 1.029, + "step": 2559 + }, + { + "epoch": 0.45584045584045585, + "grad_norm": 0.44476228952407837, + "learning_rate": 0.00019374601177252502, + "loss": 0.8494, + "step": 2560 + }, + { + "epoch": 0.45601851851851855, + "grad_norm": 0.4886183440685272, + "learning_rate": 0.00019374113843628366, + "loss": 1.1374, + "step": 2561 + }, + { + "epoch": 0.4561965811965812, + "grad_norm": 0.4786703288555145, + "learning_rate": 0.00019373626326337946, + "loss": 1.2861, + "step": 2562 + }, + { + "epoch": 0.4563746438746439, + "grad_norm": 0.5752716660499573, + "learning_rate": 0.0001937313862539079, + "loss": 1.2365, + "step": 2563 + }, + { + "epoch": 0.4565527065527066, + "grad_norm": 0.519176185131073, + "learning_rate": 0.00019372650740796452, + "loss": 1.2264, + "step": 2564 + }, + { + "epoch": 0.4567307692307692, + "grad_norm": 0.5927292704582214, + "learning_rate": 0.00019372162672564493, + "loss": 0.8979, + "step": 2565 + }, + { + "epoch": 0.4569088319088319, + "grad_norm": 0.5467435121536255, + "learning_rate": 0.00019371674420704478, + "loss": 1.1016, + "step": 2566 + }, + { + "epoch": 0.4570868945868946, + "grad_norm": 0.49593284726142883, + "learning_rate": 0.00019371185985225968, + "loss": 0.982, + "step": 2567 + }, + { + "epoch": 0.45726495726495725, + "grad_norm": 0.5696587562561035, + "learning_rate": 0.00019370697366138538, + "loss": 0.979, + "step": 2568 + }, + { + "epoch": 0.45744301994301995, + "grad_norm": 0.4455752968788147, + "learning_rate": 0.00019370208563451757, + "loss": 0.8832, + "step": 2569 + }, + { + "epoch": 0.45762108262108264, + "grad_norm": 0.5072923302650452, + "learning_rate": 0.00019369719577175203, + "loss": 1.1046, + "step": 2570 + }, + { + "epoch": 0.4577991452991453, + "grad_norm": 0.45119982957839966, + "learning_rate": 0.0001936923040731846, + "loss": 1.0083, + "step": 2571 + }, + { + "epoch": 0.457977207977208, + "grad_norm": 0.5062251091003418, + "learning_rate": 0.00019368741053891108, + "loss": 1.2771, + "step": 2572 + }, + { + "epoch": 0.4581552706552707, + "grad_norm": 0.5511104464530945, + "learning_rate": 0.0001936825151690274, + "loss": 1.0039, + "step": 2573 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 0.4721006453037262, + "learning_rate": 0.0001936776179636294, + "loss": 1.3246, + "step": 2574 + }, + { + "epoch": 0.458511396011396, + "grad_norm": 0.5021488666534424, + "learning_rate": 0.0001936727189228131, + "loss": 1.1733, + "step": 2575 + }, + { + "epoch": 0.4586894586894587, + "grad_norm": 0.5755292177200317, + "learning_rate": 0.0001936678180466745, + "loss": 1.2241, + "step": 2576 + }, + { + "epoch": 0.45886752136752135, + "grad_norm": 0.4501610994338989, + "learning_rate": 0.00019366291533530952, + "loss": 1.0503, + "step": 2577 + }, + { + "epoch": 0.45904558404558404, + "grad_norm": 0.4067458212375641, + "learning_rate": 0.00019365801078881432, + "loss": 0.8259, + "step": 2578 + }, + { + "epoch": 0.45922364672364674, + "grad_norm": 0.539730429649353, + "learning_rate": 0.0001936531044072849, + "loss": 1.1964, + "step": 2579 + }, + { + "epoch": 0.4594017094017094, + "grad_norm": 0.5624797344207764, + "learning_rate": 0.0001936481961908175, + "loss": 1.2059, + "step": 2580 + }, + { + "epoch": 0.4595797720797721, + "grad_norm": 0.43679240345954895, + "learning_rate": 0.00019364328613950824, + "loss": 1.1371, + "step": 2581 + }, + { + "epoch": 0.45975783475783477, + "grad_norm": 0.5214769244194031, + "learning_rate": 0.00019363837425345328, + "loss": 1.109, + "step": 2582 + }, + { + "epoch": 0.4599358974358974, + "grad_norm": 0.4522894024848938, + "learning_rate": 0.00019363346053274892, + "loss": 1.0532, + "step": 2583 + }, + { + "epoch": 0.4601139601139601, + "grad_norm": 0.44980281591415405, + "learning_rate": 0.0001936285449774914, + "loss": 0.9352, + "step": 2584 + }, + { + "epoch": 0.4602920227920228, + "grad_norm": 0.5697414875030518, + "learning_rate": 0.00019362362758777705, + "loss": 1.2171, + "step": 2585 + }, + { + "epoch": 0.46047008547008544, + "grad_norm": 0.4636315107345581, + "learning_rate": 0.00019361870836370217, + "loss": 1.0662, + "step": 2586 + }, + { + "epoch": 0.46064814814814814, + "grad_norm": 0.5144017338752747, + "learning_rate": 0.00019361378730536321, + "loss": 1.0681, + "step": 2587 + }, + { + "epoch": 0.46082621082621084, + "grad_norm": 0.5007636547088623, + "learning_rate": 0.00019360886441285654, + "loss": 1.2058, + "step": 2588 + }, + { + "epoch": 0.46100427350427353, + "grad_norm": 0.5024117231369019, + "learning_rate": 0.00019360393968627864, + "loss": 1.065, + "step": 2589 + }, + { + "epoch": 0.46118233618233617, + "grad_norm": 0.48105588555336, + "learning_rate": 0.00019359901312572596, + "loss": 1.0887, + "step": 2590 + }, + { + "epoch": 0.46136039886039887, + "grad_norm": 0.5381982326507568, + "learning_rate": 0.00019359408473129506, + "loss": 1.2754, + "step": 2591 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 0.5051333904266357, + "learning_rate": 0.0001935891545030825, + "loss": 0.9334, + "step": 2592 + }, + { + "epoch": 0.4617165242165242, + "grad_norm": 0.43818601965904236, + "learning_rate": 0.0001935842224411849, + "loss": 1.0967, + "step": 2593 + }, + { + "epoch": 0.4618945868945869, + "grad_norm": 0.4727257490158081, + "learning_rate": 0.0001935792885456988, + "loss": 0.8136, + "step": 2594 + }, + { + "epoch": 0.4620726495726496, + "grad_norm": 0.5505291223526001, + "learning_rate": 0.00019357435281672098, + "loss": 1.3113, + "step": 2595 + }, + { + "epoch": 0.46225071225071224, + "grad_norm": 0.4705682396888733, + "learning_rate": 0.0001935694152543481, + "loss": 0.9863, + "step": 2596 + }, + { + "epoch": 0.46242877492877493, + "grad_norm": 0.49653419852256775, + "learning_rate": 0.0001935644758586769, + "loss": 1.035, + "step": 2597 + }, + { + "epoch": 0.46260683760683763, + "grad_norm": 0.4788367748260498, + "learning_rate": 0.00019355953462980415, + "loss": 1.1253, + "step": 2598 + }, + { + "epoch": 0.46278490028490027, + "grad_norm": 0.5295125842094421, + "learning_rate": 0.00019355459156782668, + "loss": 1.0853, + "step": 2599 + }, + { + "epoch": 0.46296296296296297, + "grad_norm": 0.4878056049346924, + "learning_rate": 0.00019354964667284133, + "loss": 1.1381, + "step": 2600 + }, + { + "epoch": 0.46314102564102566, + "grad_norm": 0.5442031025886536, + "learning_rate": 0.00019354469994494497, + "loss": 1.1349, + "step": 2601 + }, + { + "epoch": 0.4633190883190883, + "grad_norm": 0.4845225214958191, + "learning_rate": 0.00019353975138423457, + "loss": 1.0538, + "step": 2602 + }, + { + "epoch": 0.463497150997151, + "grad_norm": 0.4957871437072754, + "learning_rate": 0.00019353480099080703, + "loss": 1.2765, + "step": 2603 + }, + { + "epoch": 0.4636752136752137, + "grad_norm": 0.5414339303970337, + "learning_rate": 0.00019352984876475936, + "loss": 1.1015, + "step": 2604 + }, + { + "epoch": 0.46385327635327633, + "grad_norm": 0.5171043872833252, + "learning_rate": 0.0001935248947061886, + "loss": 0.9995, + "step": 2605 + }, + { + "epoch": 0.46403133903133903, + "grad_norm": 0.46040529012680054, + "learning_rate": 0.0001935199388151918, + "loss": 1.1126, + "step": 2606 + }, + { + "epoch": 0.4642094017094017, + "grad_norm": 0.5327033400535583, + "learning_rate": 0.00019351498109186613, + "loss": 1.1983, + "step": 2607 + }, + { + "epoch": 0.46438746438746437, + "grad_norm": 0.4451361298561096, + "learning_rate": 0.0001935100215363086, + "loss": 0.9689, + "step": 2608 + }, + { + "epoch": 0.46456552706552706, + "grad_norm": 0.5462809801101685, + "learning_rate": 0.00019350506014861646, + "loss": 1.036, + "step": 2609 + }, + { + "epoch": 0.46474358974358976, + "grad_norm": 0.4907000958919525, + "learning_rate": 0.00019350009692888694, + "loss": 1.0724, + "step": 2610 + }, + { + "epoch": 0.4649216524216524, + "grad_norm": 0.47523510456085205, + "learning_rate": 0.00019349513187721723, + "loss": 0.9214, + "step": 2611 + }, + { + "epoch": 0.4650997150997151, + "grad_norm": 0.539732813835144, + "learning_rate": 0.0001934901649937046, + "loss": 1.1166, + "step": 2612 + }, + { + "epoch": 0.4652777777777778, + "grad_norm": 0.4827860891819, + "learning_rate": 0.00019348519627844643, + "loss": 1.1613, + "step": 2613 + }, + { + "epoch": 0.46545584045584043, + "grad_norm": 0.5385223031044006, + "learning_rate": 0.00019348022573154, + "loss": 1.0105, + "step": 2614 + }, + { + "epoch": 0.4656339031339031, + "grad_norm": 0.4629383087158203, + "learning_rate": 0.0001934752533530828, + "loss": 1.0298, + "step": 2615 + }, + { + "epoch": 0.4658119658119658, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00019347027914317212, + "loss": 1.3158, + "step": 2616 + }, + { + "epoch": 0.46599002849002846, + "grad_norm": 0.5954698324203491, + "learning_rate": 0.00019346530310190553, + "loss": 1.1882, + "step": 2617 + }, + { + "epoch": 0.46616809116809116, + "grad_norm": 0.49185171723365784, + "learning_rate": 0.00019346032522938046, + "loss": 1.0977, + "step": 2618 + }, + { + "epoch": 0.46634615384615385, + "grad_norm": 0.5145422220230103, + "learning_rate": 0.0001934553455256945, + "loss": 0.9948, + "step": 2619 + }, + { + "epoch": 0.46652421652421655, + "grad_norm": 0.6809412837028503, + "learning_rate": 0.00019345036399094517, + "loss": 1.5798, + "step": 2620 + }, + { + "epoch": 0.4667022792022792, + "grad_norm": 0.4606841206550598, + "learning_rate": 0.00019344538062523005, + "loss": 0.7357, + "step": 2621 + }, + { + "epoch": 0.4668803418803419, + "grad_norm": 0.49036628007888794, + "learning_rate": 0.00019344039542864685, + "loss": 1.1518, + "step": 2622 + }, + { + "epoch": 0.4670584045584046, + "grad_norm": 0.47904539108276367, + "learning_rate": 0.0001934354084012932, + "loss": 0.9929, + "step": 2623 + }, + { + "epoch": 0.4672364672364672, + "grad_norm": 0.5224666595458984, + "learning_rate": 0.0001934304195432668, + "loss": 1.2544, + "step": 2624 + }, + { + "epoch": 0.4674145299145299, + "grad_norm": 0.4902483820915222, + "learning_rate": 0.00019342542885466543, + "loss": 1.0301, + "step": 2625 + }, + { + "epoch": 0.4675925925925926, + "grad_norm": 0.46824702620506287, + "learning_rate": 0.00019342043633558683, + "loss": 0.9364, + "step": 2626 + }, + { + "epoch": 0.46777065527065526, + "grad_norm": 0.46272051334381104, + "learning_rate": 0.00019341544198612888, + "loss": 1.056, + "step": 2627 + }, + { + "epoch": 0.46794871794871795, + "grad_norm": 0.6216606497764587, + "learning_rate": 0.0001934104458063894, + "loss": 1.0825, + "step": 2628 + }, + { + "epoch": 0.46812678062678065, + "grad_norm": 0.5024014115333557, + "learning_rate": 0.00019340544779646623, + "loss": 1.1832, + "step": 2629 + }, + { + "epoch": 0.4683048433048433, + "grad_norm": 0.5547130107879639, + "learning_rate": 0.00019340044795645737, + "loss": 1.1335, + "step": 2630 + }, + { + "epoch": 0.468482905982906, + "grad_norm": 0.5439161658287048, + "learning_rate": 0.0001933954462864608, + "loss": 1.0229, + "step": 2631 + }, + { + "epoch": 0.4686609686609687, + "grad_norm": 0.4782990515232086, + "learning_rate": 0.0001933904427865744, + "loss": 1.2318, + "step": 2632 + }, + { + "epoch": 0.4688390313390313, + "grad_norm": 0.5872140526771545, + "learning_rate": 0.00019338543745689633, + "loss": 1.0132, + "step": 2633 + }, + { + "epoch": 0.469017094017094, + "grad_norm": 0.44163307547569275, + "learning_rate": 0.00019338043029752458, + "loss": 1.0091, + "step": 2634 + }, + { + "epoch": 0.4691951566951567, + "grad_norm": 0.541081428527832, + "learning_rate": 0.0001933754213085573, + "loss": 1.2155, + "step": 2635 + }, + { + "epoch": 0.46937321937321935, + "grad_norm": 0.4761527478694916, + "learning_rate": 0.00019337041049009255, + "loss": 1.1138, + "step": 2636 + }, + { + "epoch": 0.46955128205128205, + "grad_norm": 0.46414369344711304, + "learning_rate": 0.0001933653978422286, + "loss": 0.9903, + "step": 2637 + }, + { + "epoch": 0.46972934472934474, + "grad_norm": 0.5337086915969849, + "learning_rate": 0.00019336038336506363, + "loss": 1.2873, + "step": 2638 + }, + { + "epoch": 0.4699074074074074, + "grad_norm": 0.5065379738807678, + "learning_rate": 0.00019335536705869592, + "loss": 1.1436, + "step": 2639 + }, + { + "epoch": 0.4700854700854701, + "grad_norm": 0.5539217591285706, + "learning_rate": 0.0001933503489232237, + "loss": 1.2881, + "step": 2640 + }, + { + "epoch": 0.4702635327635328, + "grad_norm": 0.48303213715553284, + "learning_rate": 0.0001933453289587453, + "loss": 1.0209, + "step": 2641 + }, + { + "epoch": 0.4704415954415954, + "grad_norm": 0.6986871957778931, + "learning_rate": 0.00019334030716535908, + "loss": 1.1979, + "step": 2642 + }, + { + "epoch": 0.4706196581196581, + "grad_norm": 0.46137234568595886, + "learning_rate": 0.00019333528354316347, + "loss": 1.0682, + "step": 2643 + }, + { + "epoch": 0.4707977207977208, + "grad_norm": 0.4726654291152954, + "learning_rate": 0.00019333025809225684, + "loss": 1.1712, + "step": 2644 + }, + { + "epoch": 0.47097578347578345, + "grad_norm": 0.46188637614250183, + "learning_rate": 0.0001933252308127377, + "loss": 1.0183, + "step": 2645 + }, + { + "epoch": 0.47115384615384615, + "grad_norm": 0.5323259830474854, + "learning_rate": 0.0001933202017047045, + "loss": 0.935, + "step": 2646 + }, + { + "epoch": 0.47133190883190884, + "grad_norm": 0.5004189014434814, + "learning_rate": 0.00019331517076825582, + "loss": 1.1331, + "step": 2647 + }, + { + "epoch": 0.47150997150997154, + "grad_norm": 0.5443634986877441, + "learning_rate": 0.0001933101380034902, + "loss": 1.0514, + "step": 2648 + }, + { + "epoch": 0.4716880341880342, + "grad_norm": 0.504180371761322, + "learning_rate": 0.0001933051034105063, + "loss": 1.3099, + "step": 2649 + }, + { + "epoch": 0.4718660968660969, + "grad_norm": 0.5092344284057617, + "learning_rate": 0.0001933000669894027, + "loss": 1.0716, + "step": 2650 + }, + { + "epoch": 0.47204415954415957, + "grad_norm": 0.5236422419548035, + "learning_rate": 0.0001932950287402781, + "loss": 1.0981, + "step": 2651 + }, + { + "epoch": 0.4722222222222222, + "grad_norm": 0.6228063702583313, + "learning_rate": 0.0001932899886632312, + "loss": 1.3398, + "step": 2652 + }, + { + "epoch": 0.4724002849002849, + "grad_norm": 0.5112748146057129, + "learning_rate": 0.00019328494675836078, + "loss": 1.0151, + "step": 2653 + }, + { + "epoch": 0.4725783475783476, + "grad_norm": 0.5554201602935791, + "learning_rate": 0.00019327990302576563, + "loss": 1.404, + "step": 2654 + }, + { + "epoch": 0.47275641025641024, + "grad_norm": 0.5050725340843201, + "learning_rate": 0.0001932748574655445, + "loss": 0.951, + "step": 2655 + }, + { + "epoch": 0.47293447293447294, + "grad_norm": 0.5161749720573425, + "learning_rate": 0.00019326981007779636, + "loss": 1.2425, + "step": 2656 + }, + { + "epoch": 0.47311253561253563, + "grad_norm": 0.4865442216396332, + "learning_rate": 0.00019326476086262002, + "loss": 1.1175, + "step": 2657 + }, + { + "epoch": 0.4732905982905983, + "grad_norm": 0.5276186466217041, + "learning_rate": 0.0001932597098201144, + "loss": 1.3687, + "step": 2658 + }, + { + "epoch": 0.47346866096866097, + "grad_norm": 0.509139358997345, + "learning_rate": 0.00019325465695037855, + "loss": 1.0546, + "step": 2659 + }, + { + "epoch": 0.47364672364672367, + "grad_norm": 0.49815434217453003, + "learning_rate": 0.00019324960225351138, + "loss": 1.0807, + "step": 2660 + }, + { + "epoch": 0.4738247863247863, + "grad_norm": 0.5059618353843689, + "learning_rate": 0.00019324454572961197, + "loss": 1.0827, + "step": 2661 + }, + { + "epoch": 0.474002849002849, + "grad_norm": 0.5698565244674683, + "learning_rate": 0.00019323948737877942, + "loss": 1.2019, + "step": 2662 + }, + { + "epoch": 0.4741809116809117, + "grad_norm": 0.49661511182785034, + "learning_rate": 0.00019323442720111276, + "loss": 1.1447, + "step": 2663 + }, + { + "epoch": 0.47435897435897434, + "grad_norm": 0.46442747116088867, + "learning_rate": 0.0001932293651967112, + "loss": 0.8796, + "step": 2664 + }, + { + "epoch": 0.47453703703703703, + "grad_norm": 0.48306044936180115, + "learning_rate": 0.00019322430136567388, + "loss": 1.1358, + "step": 2665 + }, + { + "epoch": 0.47471509971509973, + "grad_norm": 0.5677350759506226, + "learning_rate": 0.00019321923570810005, + "loss": 1.1026, + "step": 2666 + }, + { + "epoch": 0.47489316239316237, + "grad_norm": 0.3700144588947296, + "learning_rate": 0.0001932141682240889, + "loss": 0.7514, + "step": 2667 + }, + { + "epoch": 0.47507122507122507, + "grad_norm": 0.6003054976463318, + "learning_rate": 0.0001932090989137398, + "loss": 1.1591, + "step": 2668 + }, + { + "epoch": 0.47524928774928776, + "grad_norm": 0.520298421382904, + "learning_rate": 0.00019320402777715204, + "loss": 1.339, + "step": 2669 + }, + { + "epoch": 0.4754273504273504, + "grad_norm": 0.46453598141670227, + "learning_rate": 0.00019319895481442493, + "loss": 0.9879, + "step": 2670 + }, + { + "epoch": 0.4756054131054131, + "grad_norm": 0.5247363448143005, + "learning_rate": 0.00019319388002565793, + "loss": 0.9862, + "step": 2671 + }, + { + "epoch": 0.4757834757834758, + "grad_norm": 0.5498613715171814, + "learning_rate": 0.00019318880341095046, + "loss": 1.2224, + "step": 2672 + }, + { + "epoch": 0.47596153846153844, + "grad_norm": 0.565838098526001, + "learning_rate": 0.00019318372497040192, + "loss": 1.0712, + "step": 2673 + }, + { + "epoch": 0.47613960113960113, + "grad_norm": 0.5797489881515503, + "learning_rate": 0.00019317864470411191, + "loss": 1.0176, + "step": 2674 + }, + { + "epoch": 0.47631766381766383, + "grad_norm": 0.5114326477050781, + "learning_rate": 0.0001931735626121799, + "loss": 1.1027, + "step": 2675 + }, + { + "epoch": 0.47649572649572647, + "grad_norm": 0.5396515727043152, + "learning_rate": 0.00019316847869470547, + "loss": 1.1782, + "step": 2676 + }, + { + "epoch": 0.47667378917378916, + "grad_norm": 0.4812076985836029, + "learning_rate": 0.00019316339295178824, + "loss": 1.1196, + "step": 2677 + }, + { + "epoch": 0.47685185185185186, + "grad_norm": 0.4875647723674774, + "learning_rate": 0.00019315830538352787, + "loss": 1.1407, + "step": 2678 + }, + { + "epoch": 0.47702991452991456, + "grad_norm": 0.5036377906799316, + "learning_rate": 0.00019315321599002404, + "loss": 0.9842, + "step": 2679 + }, + { + "epoch": 0.4772079772079772, + "grad_norm": 0.5054177641868591, + "learning_rate": 0.00019314812477137645, + "loss": 0.8196, + "step": 2680 + }, + { + "epoch": 0.4773860398860399, + "grad_norm": 0.5050665736198425, + "learning_rate": 0.00019314303172768483, + "loss": 0.8463, + "step": 2681 + }, + { + "epoch": 0.4775641025641026, + "grad_norm": 0.5179004669189453, + "learning_rate": 0.000193137936859049, + "loss": 1.2485, + "step": 2682 + }, + { + "epoch": 0.47774216524216523, + "grad_norm": 0.44986143708229065, + "learning_rate": 0.00019313284016556876, + "loss": 0.9855, + "step": 2683 + }, + { + "epoch": 0.4779202279202279, + "grad_norm": 0.5594347715377808, + "learning_rate": 0.00019312774164734398, + "loss": 1.0987, + "step": 2684 + }, + { + "epoch": 0.4780982905982906, + "grad_norm": 0.4837244749069214, + "learning_rate": 0.0001931226413044746, + "loss": 1.1119, + "step": 2685 + }, + { + "epoch": 0.47827635327635326, + "grad_norm": 0.489145427942276, + "learning_rate": 0.0001931175391370605, + "loss": 1.1962, + "step": 2686 + }, + { + "epoch": 0.47845441595441596, + "grad_norm": 0.503568708896637, + "learning_rate": 0.00019311243514520164, + "loss": 0.9668, + "step": 2687 + }, + { + "epoch": 0.47863247863247865, + "grad_norm": 0.5401005744934082, + "learning_rate": 0.00019310732932899805, + "loss": 1.3072, + "step": 2688 + }, + { + "epoch": 0.4788105413105413, + "grad_norm": 0.526523768901825, + "learning_rate": 0.00019310222168854971, + "loss": 1.1387, + "step": 2689 + }, + { + "epoch": 0.478988603988604, + "grad_norm": 0.5223183631896973, + "learning_rate": 0.00019309711222395678, + "loss": 1.1391, + "step": 2690 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 0.5840879082679749, + "learning_rate": 0.00019309200093531933, + "loss": 1.1543, + "step": 2691 + }, + { + "epoch": 0.4793447293447293, + "grad_norm": 0.5173699259757996, + "learning_rate": 0.00019308688782273753, + "loss": 1.1889, + "step": 2692 + }, + { + "epoch": 0.479522792022792, + "grad_norm": 0.5417894124984741, + "learning_rate": 0.00019308177288631146, + "loss": 1.299, + "step": 2693 + }, + { + "epoch": 0.4797008547008547, + "grad_norm": 0.4890797734260559, + "learning_rate": 0.0001930766561261415, + "loss": 1.1516, + "step": 2694 + }, + { + "epoch": 0.47987891737891736, + "grad_norm": 0.5422119498252869, + "learning_rate": 0.00019307153754232772, + "loss": 1.0301, + "step": 2695 + }, + { + "epoch": 0.48005698005698005, + "grad_norm": 0.5838702917098999, + "learning_rate": 0.00019306641713497057, + "loss": 1.265, + "step": 2696 + }, + { + "epoch": 0.48023504273504275, + "grad_norm": 0.5020943284034729, + "learning_rate": 0.00019306129490417027, + "loss": 1.1119, + "step": 2697 + }, + { + "epoch": 0.4804131054131054, + "grad_norm": 0.412993460893631, + "learning_rate": 0.00019305617085002723, + "loss": 0.8083, + "step": 2698 + }, + { + "epoch": 0.4805911680911681, + "grad_norm": 0.6270101070404053, + "learning_rate": 0.00019305104497264184, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 0.4807692307692308, + "grad_norm": 0.45256730914115906, + "learning_rate": 0.0001930459172721145, + "loss": 1.0368, + "step": 2700 + }, + { + "epoch": 0.4809472934472934, + "grad_norm": 0.5351749658584595, + "learning_rate": 0.0001930407877485457, + "loss": 1.135, + "step": 2701 + }, + { + "epoch": 0.4811253561253561, + "grad_norm": 0.49324163794517517, + "learning_rate": 0.00019303565640203593, + "loss": 0.9383, + "step": 2702 + }, + { + "epoch": 0.4813034188034188, + "grad_norm": 0.5434361100196838, + "learning_rate": 0.00019303052323268576, + "loss": 1.2605, + "step": 2703 + }, + { + "epoch": 0.48148148148148145, + "grad_norm": 0.5858064889907837, + "learning_rate": 0.00019302538824059572, + "loss": 1.0846, + "step": 2704 + }, + { + "epoch": 0.48165954415954415, + "grad_norm": 0.5753700733184814, + "learning_rate": 0.00019302025142586647, + "loss": 1.0371, + "step": 2705 + }, + { + "epoch": 0.48183760683760685, + "grad_norm": 0.43102699518203735, + "learning_rate": 0.00019301511278859858, + "loss": 0.9189, + "step": 2706 + }, + { + "epoch": 0.48201566951566954, + "grad_norm": 0.4731025993824005, + "learning_rate": 0.0001930099723288928, + "loss": 1.1291, + "step": 2707 + }, + { + "epoch": 0.4821937321937322, + "grad_norm": 0.5685615539550781, + "learning_rate": 0.00019300483004684987, + "loss": 1.1006, + "step": 2708 + }, + { + "epoch": 0.4823717948717949, + "grad_norm": 0.4368155896663666, + "learning_rate": 0.00019299968594257044, + "loss": 0.9959, + "step": 2709 + }, + { + "epoch": 0.4825498575498576, + "grad_norm": 0.5594738125801086, + "learning_rate": 0.00019299454001615537, + "loss": 1.0826, + "step": 2710 + }, + { + "epoch": 0.4827279202279202, + "grad_norm": 0.48876598477363586, + "learning_rate": 0.00019298939226770548, + "loss": 1.1556, + "step": 2711 + }, + { + "epoch": 0.4829059829059829, + "grad_norm": 0.548039436340332, + "learning_rate": 0.00019298424269732157, + "loss": 1.158, + "step": 2712 + }, + { + "epoch": 0.4830840455840456, + "grad_norm": 0.4957645535469055, + "learning_rate": 0.00019297909130510464, + "loss": 0.9824, + "step": 2713 + }, + { + "epoch": 0.48326210826210825, + "grad_norm": 0.5197011232376099, + "learning_rate": 0.00019297393809115555, + "loss": 1.1074, + "step": 2714 + }, + { + "epoch": 0.48344017094017094, + "grad_norm": 0.5742064118385315, + "learning_rate": 0.00019296878305557526, + "loss": 1.0431, + "step": 2715 + }, + { + "epoch": 0.48361823361823364, + "grad_norm": 0.5698413252830505, + "learning_rate": 0.0001929636261984648, + "loss": 1.0713, + "step": 2716 + }, + { + "epoch": 0.4837962962962963, + "grad_norm": 0.48126333951950073, + "learning_rate": 0.0001929584675199252, + "loss": 0.9274, + "step": 2717 + }, + { + "epoch": 0.483974358974359, + "grad_norm": 0.49299830198287964, + "learning_rate": 0.00019295330702005754, + "loss": 0.9392, + "step": 2718 + }, + { + "epoch": 0.48415242165242167, + "grad_norm": 0.4780774414539337, + "learning_rate": 0.0001929481446989629, + "loss": 1.1459, + "step": 2719 + }, + { + "epoch": 0.4843304843304843, + "grad_norm": 0.5462654829025269, + "learning_rate": 0.00019294298055674248, + "loss": 1.0635, + "step": 2720 + }, + { + "epoch": 0.484508547008547, + "grad_norm": 0.5371061563491821, + "learning_rate": 0.00019293781459349743, + "loss": 1.3578, + "step": 2721 + }, + { + "epoch": 0.4846866096866097, + "grad_norm": 0.46308520436286926, + "learning_rate": 0.00019293264680932893, + "loss": 0.9001, + "step": 2722 + }, + { + "epoch": 0.48486467236467234, + "grad_norm": 0.5149807929992676, + "learning_rate": 0.0001929274772043383, + "loss": 0.6908, + "step": 2723 + }, + { + "epoch": 0.48504273504273504, + "grad_norm": 0.5435031056404114, + "learning_rate": 0.00019292230577862678, + "loss": 1.2143, + "step": 2724 + }, + { + "epoch": 0.48522079772079774, + "grad_norm": 0.44217726588249207, + "learning_rate": 0.00019291713253229568, + "loss": 0.9303, + "step": 2725 + }, + { + "epoch": 0.4853988603988604, + "grad_norm": 0.6120226383209229, + "learning_rate": 0.00019291195746544643, + "loss": 1.3801, + "step": 2726 + }, + { + "epoch": 0.4855769230769231, + "grad_norm": 0.5014316439628601, + "learning_rate": 0.00019290678057818037, + "loss": 1.0631, + "step": 2727 + }, + { + "epoch": 0.48575498575498577, + "grad_norm": 0.5667829513549805, + "learning_rate": 0.00019290160187059895, + "loss": 1.3166, + "step": 2728 + }, + { + "epoch": 0.4859330484330484, + "grad_norm": 0.5011509656906128, + "learning_rate": 0.0001928964213428036, + "loss": 1.1887, + "step": 2729 + }, + { + "epoch": 0.4861111111111111, + "grad_norm": 0.48317405581474304, + "learning_rate": 0.00019289123899489586, + "loss": 1.1125, + "step": 2730 + }, + { + "epoch": 0.4862891737891738, + "grad_norm": 0.4669005870819092, + "learning_rate": 0.00019288605482697726, + "loss": 1.0091, + "step": 2731 + }, + { + "epoch": 0.48646723646723644, + "grad_norm": 0.4330739974975586, + "learning_rate": 0.00019288086883914937, + "loss": 0.9789, + "step": 2732 + }, + { + "epoch": 0.48664529914529914, + "grad_norm": 0.48482781648635864, + "learning_rate": 0.0001928756810315138, + "loss": 1.1922, + "step": 2733 + }, + { + "epoch": 0.48682336182336183, + "grad_norm": 0.5781838297843933, + "learning_rate": 0.0001928704914041722, + "loss": 1.1793, + "step": 2734 + }, + { + "epoch": 0.48700142450142453, + "grad_norm": 0.5955413579940796, + "learning_rate": 0.00019286529995722623, + "loss": 1.1001, + "step": 2735 + }, + { + "epoch": 0.48717948717948717, + "grad_norm": 0.49204322695732117, + "learning_rate": 0.00019286010669077763, + "loss": 0.9219, + "step": 2736 + }, + { + "epoch": 0.48735754985754987, + "grad_norm": 0.5853500962257385, + "learning_rate": 0.00019285491160492813, + "loss": 1.1133, + "step": 2737 + }, + { + "epoch": 0.48753561253561256, + "grad_norm": 0.5555846095085144, + "learning_rate": 0.0001928497146997795, + "loss": 1.0915, + "step": 2738 + }, + { + "epoch": 0.4877136752136752, + "grad_norm": 0.5166759490966797, + "learning_rate": 0.00019284451597543364, + "loss": 0.9349, + "step": 2739 + }, + { + "epoch": 0.4878917378917379, + "grad_norm": 0.47816506028175354, + "learning_rate": 0.00019283931543199234, + "loss": 0.8978, + "step": 2740 + }, + { + "epoch": 0.4880698005698006, + "grad_norm": 0.5632442831993103, + "learning_rate": 0.0001928341130695575, + "loss": 1.0491, + "step": 2741 + }, + { + "epoch": 0.48824786324786323, + "grad_norm": 0.6532769799232483, + "learning_rate": 0.00019282890888823107, + "loss": 1.2779, + "step": 2742 + }, + { + "epoch": 0.48842592592592593, + "grad_norm": 0.5733640789985657, + "learning_rate": 0.000192823702888115, + "loss": 1.4127, + "step": 2743 + }, + { + "epoch": 0.4886039886039886, + "grad_norm": 0.5701746344566345, + "learning_rate": 0.00019281849506931132, + "loss": 1.138, + "step": 2744 + }, + { + "epoch": 0.48878205128205127, + "grad_norm": 0.5227449536323547, + "learning_rate": 0.000192813285431922, + "loss": 1.1831, + "step": 2745 + }, + { + "epoch": 0.48896011396011396, + "grad_norm": 0.48457080125808716, + "learning_rate": 0.00019280807397604915, + "loss": 1.2468, + "step": 2746 + }, + { + "epoch": 0.48913817663817666, + "grad_norm": 0.4596176743507385, + "learning_rate": 0.0001928028607017949, + "loss": 1.1098, + "step": 2747 + }, + { + "epoch": 0.4893162393162393, + "grad_norm": 0.5204966068267822, + "learning_rate": 0.00019279764560926142, + "loss": 1.1501, + "step": 2748 + }, + { + "epoch": 0.489494301994302, + "grad_norm": 0.5179490447044373, + "learning_rate": 0.0001927924286985508, + "loss": 1.2601, + "step": 2749 + }, + { + "epoch": 0.4896723646723647, + "grad_norm": 0.4563423693180084, + "learning_rate": 0.00019278720996976533, + "loss": 1.081, + "step": 2750 + }, + { + "epoch": 0.48985042735042733, + "grad_norm": 0.4906339943408966, + "learning_rate": 0.00019278198942300717, + "loss": 1.157, + "step": 2751 + }, + { + "epoch": 0.49002849002849, + "grad_norm": 0.42241403460502625, + "learning_rate": 0.00019277676705837873, + "loss": 1.0333, + "step": 2752 + }, + { + "epoch": 0.4902065527065527, + "grad_norm": 0.6310175657272339, + "learning_rate": 0.00019277154287598226, + "loss": 1.1225, + "step": 2753 + }, + { + "epoch": 0.49038461538461536, + "grad_norm": 0.5109034776687622, + "learning_rate": 0.0001927663168759201, + "loss": 1.1619, + "step": 2754 + }, + { + "epoch": 0.49056267806267806, + "grad_norm": 0.4809598922729492, + "learning_rate": 0.00019276108905829465, + "loss": 1.0423, + "step": 2755 + }, + { + "epoch": 0.49074074074074076, + "grad_norm": 0.557502806186676, + "learning_rate": 0.00019275585942320837, + "loss": 0.8783, + "step": 2756 + }, + { + "epoch": 0.4909188034188034, + "grad_norm": 0.5434393882751465, + "learning_rate": 0.0001927506279707637, + "loss": 1.1701, + "step": 2757 + }, + { + "epoch": 0.4910968660968661, + "grad_norm": 0.49278944730758667, + "learning_rate": 0.00019274539470106317, + "loss": 1.0447, + "step": 2758 + }, + { + "epoch": 0.4912749287749288, + "grad_norm": 0.5634264349937439, + "learning_rate": 0.00019274015961420927, + "loss": 1.0639, + "step": 2759 + }, + { + "epoch": 0.49145299145299143, + "grad_norm": 0.5632645487785339, + "learning_rate": 0.00019273492271030464, + "loss": 0.9223, + "step": 2760 + }, + { + "epoch": 0.4916310541310541, + "grad_norm": 0.5949172377586365, + "learning_rate": 0.00019272968398945177, + "loss": 0.894, + "step": 2761 + }, + { + "epoch": 0.4918091168091168, + "grad_norm": 0.5375374555587769, + "learning_rate": 0.00019272444345175342, + "loss": 1.0311, + "step": 2762 + }, + { + "epoch": 0.49198717948717946, + "grad_norm": 0.5211305022239685, + "learning_rate": 0.00019271920109731222, + "loss": 1.1531, + "step": 2763 + }, + { + "epoch": 0.49216524216524216, + "grad_norm": 0.44022253155708313, + "learning_rate": 0.00019271395692623084, + "loss": 0.9147, + "step": 2764 + }, + { + "epoch": 0.49234330484330485, + "grad_norm": 0.4682174623012543, + "learning_rate": 0.0001927087109386121, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 0.49252136752136755, + "grad_norm": 0.4971517324447632, + "learning_rate": 0.0001927034631345588, + "loss": 1.1017, + "step": 2766 + }, + { + "epoch": 0.4926994301994302, + "grad_norm": 0.5015294551849365, + "learning_rate": 0.00019269821351417364, + "loss": 1.1093, + "step": 2767 + }, + { + "epoch": 0.4928774928774929, + "grad_norm": 0.5512694716453552, + "learning_rate": 0.00019269296207755958, + "loss": 0.9657, + "step": 2768 + }, + { + "epoch": 0.4930555555555556, + "grad_norm": 0.4914868474006653, + "learning_rate": 0.00019268770882481948, + "loss": 1.0379, + "step": 2769 + }, + { + "epoch": 0.4932336182336182, + "grad_norm": 0.567337691783905, + "learning_rate": 0.00019268245375605626, + "loss": 1.004, + "step": 2770 + }, + { + "epoch": 0.4934116809116809, + "grad_norm": 0.518489420413971, + "learning_rate": 0.0001926771968713729, + "loss": 1.0734, + "step": 2771 + }, + { + "epoch": 0.4935897435897436, + "grad_norm": 0.567742109298706, + "learning_rate": 0.00019267193817087237, + "loss": 1.1276, + "step": 2772 + }, + { + "epoch": 0.49376780626780625, + "grad_norm": 0.5287964344024658, + "learning_rate": 0.00019266667765465773, + "loss": 1.1429, + "step": 2773 + }, + { + "epoch": 0.49394586894586895, + "grad_norm": 0.5302085876464844, + "learning_rate": 0.00019266141532283207, + "loss": 1.0934, + "step": 2774 + }, + { + "epoch": 0.49412393162393164, + "grad_norm": 0.5569987297058105, + "learning_rate": 0.00019265615117549842, + "loss": 1.1453, + "step": 2775 + }, + { + "epoch": 0.4943019943019943, + "grad_norm": 0.519695520401001, + "learning_rate": 0.00019265088521275997, + "loss": 1.1255, + "step": 2776 + }, + { + "epoch": 0.494480056980057, + "grad_norm": 0.5073211193084717, + "learning_rate": 0.0001926456174347199, + "loss": 1.0609, + "step": 2777 + }, + { + "epoch": 0.4946581196581197, + "grad_norm": 0.45028239488601685, + "learning_rate": 0.00019264034784148142, + "loss": 0.9098, + "step": 2778 + }, + { + "epoch": 0.4948361823361823, + "grad_norm": 0.6641215682029724, + "learning_rate": 0.00019263507643314776, + "loss": 0.8903, + "step": 2779 + }, + { + "epoch": 0.495014245014245, + "grad_norm": 0.5281413793563843, + "learning_rate": 0.00019262980320982224, + "loss": 1.2906, + "step": 2780 + }, + { + "epoch": 0.4951923076923077, + "grad_norm": 0.6256437301635742, + "learning_rate": 0.0001926245281716081, + "loss": 1.4142, + "step": 2781 + }, + { + "epoch": 0.49537037037037035, + "grad_norm": 0.5422517657279968, + "learning_rate": 0.00019261925131860877, + "loss": 1.1606, + "step": 2782 + }, + { + "epoch": 0.49554843304843305, + "grad_norm": 0.46938949823379517, + "learning_rate": 0.0001926139726509276, + "loss": 1.0333, + "step": 2783 + }, + { + "epoch": 0.49572649572649574, + "grad_norm": 0.5799683928489685, + "learning_rate": 0.000192608692168668, + "loss": 1.0333, + "step": 2784 + }, + { + "epoch": 0.4959045584045584, + "grad_norm": 0.5231602787971497, + "learning_rate": 0.0001926034098719335, + "loss": 1.1847, + "step": 2785 + }, + { + "epoch": 0.4960826210826211, + "grad_norm": 0.477845698595047, + "learning_rate": 0.00019259812576082752, + "loss": 1.0746, + "step": 2786 + }, + { + "epoch": 0.4962606837606838, + "grad_norm": 0.5490350723266602, + "learning_rate": 0.00019259283983545365, + "loss": 1.2462, + "step": 2787 + }, + { + "epoch": 0.4964387464387464, + "grad_norm": 0.5788847208023071, + "learning_rate": 0.0001925875520959154, + "loss": 1.3485, + "step": 2788 + }, + { + "epoch": 0.4966168091168091, + "grad_norm": 0.46184736490249634, + "learning_rate": 0.00019258226254231643, + "loss": 0.8673, + "step": 2789 + }, + { + "epoch": 0.4967948717948718, + "grad_norm": 0.4890633225440979, + "learning_rate": 0.0001925769711747603, + "loss": 0.9474, + "step": 2790 + }, + { + "epoch": 0.49697293447293445, + "grad_norm": 0.5719282627105713, + "learning_rate": 0.00019257167799335078, + "loss": 1.2532, + "step": 2791 + }, + { + "epoch": 0.49715099715099714, + "grad_norm": 0.5385584235191345, + "learning_rate": 0.0001925663829981915, + "loss": 1.1326, + "step": 2792 + }, + { + "epoch": 0.49732905982905984, + "grad_norm": 0.5339545011520386, + "learning_rate": 0.00019256108618938625, + "loss": 1.1362, + "step": 2793 + }, + { + "epoch": 0.49750712250712253, + "grad_norm": 0.5017803907394409, + "learning_rate": 0.00019255578756703878, + "loss": 1.0449, + "step": 2794 + }, + { + "epoch": 0.4976851851851852, + "grad_norm": 0.6004226803779602, + "learning_rate": 0.00019255048713125294, + "loss": 0.9346, + "step": 2795 + }, + { + "epoch": 0.49786324786324787, + "grad_norm": 0.44581490755081177, + "learning_rate": 0.00019254518488213255, + "loss": 1.038, + "step": 2796 + }, + { + "epoch": 0.49804131054131057, + "grad_norm": 0.5180951356887817, + "learning_rate": 0.00019253988081978151, + "loss": 1.0479, + "step": 2797 + }, + { + "epoch": 0.4982193732193732, + "grad_norm": 0.53944993019104, + "learning_rate": 0.00019253457494430376, + "loss": 1.2598, + "step": 2798 + }, + { + "epoch": 0.4983974358974359, + "grad_norm": 0.5633010268211365, + "learning_rate": 0.00019252926725580322, + "loss": 1.205, + "step": 2799 + }, + { + "epoch": 0.4985754985754986, + "grad_norm": 0.6653175950050354, + "learning_rate": 0.0001925239577543839, + "loss": 1.2383, + "step": 2800 + }, + { + "epoch": 0.49875356125356124, + "grad_norm": 0.5083333849906921, + "learning_rate": 0.00019251864644014984, + "loss": 1.0649, + "step": 2801 + }, + { + "epoch": 0.49893162393162394, + "grad_norm": 0.4842020571231842, + "learning_rate": 0.00019251333331320506, + "loss": 1.1991, + "step": 2802 + }, + { + "epoch": 0.49910968660968663, + "grad_norm": 0.47987112402915955, + "learning_rate": 0.00019250801837365373, + "loss": 1.1686, + "step": 2803 + }, + { + "epoch": 0.49928774928774927, + "grad_norm": 0.5316333770751953, + "learning_rate": 0.00019250270162159992, + "loss": 1.1759, + "step": 2804 + }, + { + "epoch": 0.49946581196581197, + "grad_norm": 0.5015079379081726, + "learning_rate": 0.00019249738305714787, + "loss": 0.9424, + "step": 2805 + }, + { + "epoch": 0.49964387464387466, + "grad_norm": 0.6488274931907654, + "learning_rate": 0.00019249206268040172, + "loss": 1.066, + "step": 2806 + }, + { + "epoch": 0.4998219373219373, + "grad_norm": 0.40364864468574524, + "learning_rate": 0.00019248674049146574, + "loss": 0.6998, + "step": 2807 + }, + { + "epoch": 0.5, + "grad_norm": 0.5535672903060913, + "learning_rate": 0.00019248141649044423, + "loss": 1.2207, + "step": 2808 + }, + { + "epoch": 0.5, + "eval_loss": 1.1072274446487427, + "eval_runtime": 28.6913, + "eval_samples_per_second": 36.283, + "eval_steps_per_second": 18.159, + "step": 2808 + }, + { + "epoch": 0.5001780626780626, + "grad_norm": 0.4834389090538025, + "learning_rate": 0.00019247609067744143, + "loss": 1.1686, + "step": 2809 + }, + { + "epoch": 0.5003561253561254, + "grad_norm": 0.5007249712944031, + "learning_rate": 0.00019247076305256176, + "loss": 1.1343, + "step": 2810 + }, + { + "epoch": 0.500534188034188, + "grad_norm": 0.4773348271846771, + "learning_rate": 0.00019246543361590957, + "loss": 0.9324, + "step": 2811 + }, + { + "epoch": 0.5007122507122507, + "grad_norm": 0.47324609756469727, + "learning_rate": 0.0001924601023675893, + "loss": 1.0223, + "step": 2812 + }, + { + "epoch": 0.5008903133903134, + "grad_norm": 0.5583845973014832, + "learning_rate": 0.00019245476930770537, + "loss": 1.1328, + "step": 2813 + }, + { + "epoch": 0.5010683760683761, + "grad_norm": 0.4814579486846924, + "learning_rate": 0.00019244943443636232, + "loss": 1.0528, + "step": 2814 + }, + { + "epoch": 0.5012464387464387, + "grad_norm": 0.4996104836463928, + "learning_rate": 0.00019244409775366465, + "loss": 1.2482, + "step": 2815 + }, + { + "epoch": 0.5014245014245015, + "grad_norm": 0.47870904207229614, + "learning_rate": 0.0001924387592597169, + "loss": 0.9452, + "step": 2816 + }, + { + "epoch": 0.5016025641025641, + "grad_norm": 0.5617441534996033, + "learning_rate": 0.0001924334189546237, + "loss": 1.378, + "step": 2817 + }, + { + "epoch": 0.5017806267806267, + "grad_norm": 0.4872083365917206, + "learning_rate": 0.00019242807683848967, + "loss": 1.1571, + "step": 2818 + }, + { + "epoch": 0.5019586894586895, + "grad_norm": 0.5147804021835327, + "learning_rate": 0.00019242273291141947, + "loss": 1.1086, + "step": 2819 + }, + { + "epoch": 0.5021367521367521, + "grad_norm": 0.4698995351791382, + "learning_rate": 0.00019241738717351784, + "loss": 1.1579, + "step": 2820 + }, + { + "epoch": 0.5023148148148148, + "grad_norm": 0.5158926844596863, + "learning_rate": 0.00019241203962488946, + "loss": 1.2763, + "step": 2821 + }, + { + "epoch": 0.5024928774928775, + "grad_norm": 0.5218976736068726, + "learning_rate": 0.00019240669026563914, + "loss": 1.0633, + "step": 2822 + }, + { + "epoch": 0.5026709401709402, + "grad_norm": 0.5511452555656433, + "learning_rate": 0.0001924013390958717, + "loss": 0.9939, + "step": 2823 + }, + { + "epoch": 0.5028490028490028, + "grad_norm": 0.5227555632591248, + "learning_rate": 0.00019239598611569191, + "loss": 1.2478, + "step": 2824 + }, + { + "epoch": 0.5030270655270656, + "grad_norm": 0.5444719791412354, + "learning_rate": 0.00019239063132520475, + "loss": 1.1574, + "step": 2825 + }, + { + "epoch": 0.5032051282051282, + "grad_norm": 0.4752781093120575, + "learning_rate": 0.0001923852747245151, + "loss": 0.9034, + "step": 2826 + }, + { + "epoch": 0.5033831908831908, + "grad_norm": 0.5286496877670288, + "learning_rate": 0.00019237991631372792, + "loss": 1.1391, + "step": 2827 + }, + { + "epoch": 0.5035612535612536, + "grad_norm": 0.5009933710098267, + "learning_rate": 0.00019237455609294815, + "loss": 1.2178, + "step": 2828 + }, + { + "epoch": 0.5037393162393162, + "grad_norm": 0.5012276768684387, + "learning_rate": 0.00019236919406228085, + "loss": 0.9877, + "step": 2829 + }, + { + "epoch": 0.5039173789173789, + "grad_norm": 0.576508104801178, + "learning_rate": 0.00019236383022183106, + "loss": 1.1299, + "step": 2830 + }, + { + "epoch": 0.5040954415954416, + "grad_norm": 0.4716590642929077, + "learning_rate": 0.0001923584645717039, + "loss": 1.0451, + "step": 2831 + }, + { + "epoch": 0.5042735042735043, + "grad_norm": 0.5817418098449707, + "learning_rate": 0.00019235309711200448, + "loss": 1.0911, + "step": 2832 + }, + { + "epoch": 0.5044515669515669, + "grad_norm": 0.5695745944976807, + "learning_rate": 0.000192347727842838, + "loss": 1.0229, + "step": 2833 + }, + { + "epoch": 0.5046296296296297, + "grad_norm": 0.49127066135406494, + "learning_rate": 0.00019234235676430958, + "loss": 1.1377, + "step": 2834 + }, + { + "epoch": 0.5048076923076923, + "grad_norm": 0.5426172614097595, + "learning_rate": 0.00019233698387652453, + "loss": 1.2427, + "step": 2835 + }, + { + "epoch": 0.5049857549857549, + "grad_norm": 0.5342385172843933, + "learning_rate": 0.0001923316091795881, + "loss": 1.1427, + "step": 2836 + }, + { + "epoch": 0.5051638176638177, + "grad_norm": 0.5480486750602722, + "learning_rate": 0.00019232623267360558, + "loss": 1.0647, + "step": 2837 + }, + { + "epoch": 0.5053418803418803, + "grad_norm": 0.4584530293941498, + "learning_rate": 0.00019232085435868235, + "loss": 1.0461, + "step": 2838 + }, + { + "epoch": 0.5055199430199431, + "grad_norm": 0.5992119908332825, + "learning_rate": 0.00019231547423492371, + "loss": 1.1456, + "step": 2839 + }, + { + "epoch": 0.5056980056980057, + "grad_norm": 0.514018177986145, + "learning_rate": 0.00019231009230243515, + "loss": 1.2559, + "step": 2840 + }, + { + "epoch": 0.5058760683760684, + "grad_norm": 0.5392283797264099, + "learning_rate": 0.0001923047085613221, + "loss": 1.044, + "step": 2841 + }, + { + "epoch": 0.5060541310541311, + "grad_norm": 0.4486566483974457, + "learning_rate": 0.00019229932301169, + "loss": 1.0679, + "step": 2842 + }, + { + "epoch": 0.5062321937321937, + "grad_norm": 0.4523460566997528, + "learning_rate": 0.00019229393565364442, + "loss": 1.1651, + "step": 2843 + }, + { + "epoch": 0.5064102564102564, + "grad_norm": 0.6032688021659851, + "learning_rate": 0.0001922885464872909, + "loss": 1.15, + "step": 2844 + }, + { + "epoch": 0.5065883190883191, + "grad_norm": 0.5883688926696777, + "learning_rate": 0.000192283155512735, + "loss": 1.2179, + "step": 2845 + }, + { + "epoch": 0.5067663817663818, + "grad_norm": 0.5534378886222839, + "learning_rate": 0.00019227776273008238, + "loss": 1.0387, + "step": 2846 + }, + { + "epoch": 0.5069444444444444, + "grad_norm": 0.5899033546447754, + "learning_rate": 0.00019227236813943872, + "loss": 1.0812, + "step": 2847 + }, + { + "epoch": 0.5071225071225072, + "grad_norm": 0.5718855261802673, + "learning_rate": 0.00019226697174090965, + "loss": 1.1375, + "step": 2848 + }, + { + "epoch": 0.5073005698005698, + "grad_norm": 0.5080967545509338, + "learning_rate": 0.00019226157353460094, + "loss": 1.1421, + "step": 2849 + }, + { + "epoch": 0.5074786324786325, + "grad_norm": 0.5253677368164062, + "learning_rate": 0.0001922561735206184, + "loss": 1.0166, + "step": 2850 + }, + { + "epoch": 0.5076566951566952, + "grad_norm": 0.47797444462776184, + "learning_rate": 0.00019225077169906772, + "loss": 1.0504, + "step": 2851 + }, + { + "epoch": 0.5078347578347578, + "grad_norm": 0.4911690652370453, + "learning_rate": 0.0001922453680700548, + "loss": 1.0629, + "step": 2852 + }, + { + "epoch": 0.5080128205128205, + "grad_norm": 0.49678200483322144, + "learning_rate": 0.00019223996263368557, + "loss": 1.1672, + "step": 2853 + }, + { + "epoch": 0.5081908831908832, + "grad_norm": 0.5451810359954834, + "learning_rate": 0.00019223455539006586, + "loss": 1.3031, + "step": 2854 + }, + { + "epoch": 0.5083689458689459, + "grad_norm": 0.5708984136581421, + "learning_rate": 0.00019222914633930166, + "loss": 1.0986, + "step": 2855 + }, + { + "epoch": 0.5085470085470085, + "grad_norm": 0.47232356667518616, + "learning_rate": 0.00019222373548149888, + "loss": 1.0449, + "step": 2856 + }, + { + "epoch": 0.5087250712250713, + "grad_norm": 0.6027610898017883, + "learning_rate": 0.0001922183228167636, + "loss": 0.862, + "step": 2857 + }, + { + "epoch": 0.5089031339031339, + "grad_norm": 0.5211802124977112, + "learning_rate": 0.00019221290834520188, + "loss": 1.1048, + "step": 2858 + }, + { + "epoch": 0.5090811965811965, + "grad_norm": 0.45101237297058105, + "learning_rate": 0.00019220749206691972, + "loss": 1.0046, + "step": 2859 + }, + { + "epoch": 0.5092592592592593, + "grad_norm": 0.5526158213615417, + "learning_rate": 0.00019220207398202335, + "loss": 1.2275, + "step": 2860 + }, + { + "epoch": 0.5094373219373219, + "grad_norm": 0.48322010040283203, + "learning_rate": 0.00019219665409061885, + "loss": 0.9974, + "step": 2861 + }, + { + "epoch": 0.5096153846153846, + "grad_norm": 0.4775219261646271, + "learning_rate": 0.00019219123239281244, + "loss": 1.1852, + "step": 2862 + }, + { + "epoch": 0.5097934472934473, + "grad_norm": 0.46184200048446655, + "learning_rate": 0.00019218580888871034, + "loss": 0.9393, + "step": 2863 + }, + { + "epoch": 0.50997150997151, + "grad_norm": 0.47495174407958984, + "learning_rate": 0.00019218038357841883, + "loss": 0.9631, + "step": 2864 + }, + { + "epoch": 0.5101495726495726, + "grad_norm": 0.48600029945373535, + "learning_rate": 0.00019217495646204418, + "loss": 1.0498, + "step": 2865 + }, + { + "epoch": 0.5103276353276354, + "grad_norm": 0.5801547169685364, + "learning_rate": 0.00019216952753969274, + "loss": 1.2181, + "step": 2866 + }, + { + "epoch": 0.510505698005698, + "grad_norm": 0.5082106590270996, + "learning_rate": 0.00019216409681147085, + "loss": 1.2009, + "step": 2867 + }, + { + "epoch": 0.5106837606837606, + "grad_norm": 0.4184330701828003, + "learning_rate": 0.00019215866427748493, + "loss": 0.8462, + "step": 2868 + }, + { + "epoch": 0.5108618233618234, + "grad_norm": 0.518099844455719, + "learning_rate": 0.00019215322993784147, + "loss": 1.2091, + "step": 2869 + }, + { + "epoch": 0.511039886039886, + "grad_norm": 0.569464921951294, + "learning_rate": 0.0001921477937926469, + "loss": 1.0264, + "step": 2870 + }, + { + "epoch": 0.5112179487179487, + "grad_norm": 0.526767909526825, + "learning_rate": 0.00019214235584200768, + "loss": 1.1192, + "step": 2871 + }, + { + "epoch": 0.5113960113960114, + "grad_norm": 0.6511057019233704, + "learning_rate": 0.00019213691608603047, + "loss": 1.3193, + "step": 2872 + }, + { + "epoch": 0.5115740740740741, + "grad_norm": 0.48536401987075806, + "learning_rate": 0.00019213147452482173, + "loss": 1.1671, + "step": 2873 + }, + { + "epoch": 0.5117521367521367, + "grad_norm": 0.7972469329833984, + "learning_rate": 0.00019212603115848818, + "loss": 1.1393, + "step": 2874 + }, + { + "epoch": 0.5119301994301995, + "grad_norm": 0.5543264746665955, + "learning_rate": 0.00019212058598713642, + "loss": 1.1436, + "step": 2875 + }, + { + "epoch": 0.5121082621082621, + "grad_norm": 0.49688720703125, + "learning_rate": 0.0001921151390108731, + "loss": 1.0897, + "step": 2876 + }, + { + "epoch": 0.5122863247863247, + "grad_norm": 0.4928736090660095, + "learning_rate": 0.000192109690229805, + "loss": 1.2426, + "step": 2877 + }, + { + "epoch": 0.5124643874643875, + "grad_norm": 0.4917896091938019, + "learning_rate": 0.0001921042396440389, + "loss": 1.0047, + "step": 2878 + }, + { + "epoch": 0.5126424501424501, + "grad_norm": 0.5485204458236694, + "learning_rate": 0.00019209878725368152, + "loss": 1.2615, + "step": 2879 + }, + { + "epoch": 0.5128205128205128, + "grad_norm": 0.5229470133781433, + "learning_rate": 0.0001920933330588397, + "loss": 1.3249, + "step": 2880 + }, + { + "epoch": 0.5129985754985755, + "grad_norm": 0.4783077538013458, + "learning_rate": 0.00019208787705962037, + "loss": 1.2004, + "step": 2881 + }, + { + "epoch": 0.5131766381766382, + "grad_norm": 0.5106910467147827, + "learning_rate": 0.00019208241925613035, + "loss": 1.1745, + "step": 2882 + }, + { + "epoch": 0.5133547008547008, + "grad_norm": 0.5308730006217957, + "learning_rate": 0.00019207695964847666, + "loss": 0.9706, + "step": 2883 + }, + { + "epoch": 0.5135327635327636, + "grad_norm": 0.5489775538444519, + "learning_rate": 0.00019207149823676617, + "loss": 1.0073, + "step": 2884 + }, + { + "epoch": 0.5137108262108262, + "grad_norm": 0.4992835521697998, + "learning_rate": 0.00019206603502110596, + "loss": 1.1053, + "step": 2885 + }, + { + "epoch": 0.5138888888888888, + "grad_norm": 0.5304922461509705, + "learning_rate": 0.00019206057000160302, + "loss": 1.0565, + "step": 2886 + }, + { + "epoch": 0.5140669515669516, + "grad_norm": 0.46411609649658203, + "learning_rate": 0.00019205510317836448, + "loss": 0.9202, + "step": 2887 + }, + { + "epoch": 0.5142450142450142, + "grad_norm": 0.5236835479736328, + "learning_rate": 0.0001920496345514974, + "loss": 0.9075, + "step": 2888 + }, + { + "epoch": 0.5144230769230769, + "grad_norm": 0.4416964054107666, + "learning_rate": 0.00019204416412110895, + "loss": 0.9225, + "step": 2889 + }, + { + "epoch": 0.5146011396011396, + "grad_norm": 0.5470940470695496, + "learning_rate": 0.00019203869188730633, + "loss": 1.2195, + "step": 2890 + }, + { + "epoch": 0.5147792022792023, + "grad_norm": 0.5380414128303528, + "learning_rate": 0.0001920332178501967, + "loss": 1.0731, + "step": 2891 + }, + { + "epoch": 0.5149572649572649, + "grad_norm": 0.4405716359615326, + "learning_rate": 0.00019202774200988737, + "loss": 0.8739, + "step": 2892 + }, + { + "epoch": 0.5151353276353277, + "grad_norm": 0.5222984552383423, + "learning_rate": 0.0001920222643664856, + "loss": 1.1806, + "step": 2893 + }, + { + "epoch": 0.5153133903133903, + "grad_norm": 0.48545539379119873, + "learning_rate": 0.0001920167849200987, + "loss": 0.9939, + "step": 2894 + }, + { + "epoch": 0.5154914529914529, + "grad_norm": 0.45078009366989136, + "learning_rate": 0.0001920113036708341, + "loss": 1.0085, + "step": 2895 + }, + { + "epoch": 0.5156695156695157, + "grad_norm": 0.5029830932617188, + "learning_rate": 0.00019200582061879913, + "loss": 1.1095, + "step": 2896 + }, + { + "epoch": 0.5158475783475783, + "grad_norm": 0.5316143035888672, + "learning_rate": 0.00019200033576410118, + "loss": 0.9883, + "step": 2897 + }, + { + "epoch": 0.5160256410256411, + "grad_norm": 0.5282100439071655, + "learning_rate": 0.0001919948491068478, + "loss": 1.1441, + "step": 2898 + }, + { + "epoch": 0.5162037037037037, + "grad_norm": 0.5145367980003357, + "learning_rate": 0.00019198936064714647, + "loss": 1.1999, + "step": 2899 + }, + { + "epoch": 0.5163817663817664, + "grad_norm": 0.5385651588439941, + "learning_rate": 0.00019198387038510468, + "loss": 1.1831, + "step": 2900 + }, + { + "epoch": 0.5165598290598291, + "grad_norm": 0.4971916377544403, + "learning_rate": 0.00019197837832083002, + "loss": 1.2518, + "step": 2901 + }, + { + "epoch": 0.5167378917378918, + "grad_norm": 0.5253807306289673, + "learning_rate": 0.00019197288445443016, + "loss": 1.0788, + "step": 2902 + }, + { + "epoch": 0.5169159544159544, + "grad_norm": 0.49724945425987244, + "learning_rate": 0.00019196738878601263, + "loss": 1.0985, + "step": 2903 + }, + { + "epoch": 0.5170940170940171, + "grad_norm": 0.5327325463294983, + "learning_rate": 0.0001919618913156852, + "loss": 1.2862, + "step": 2904 + }, + { + "epoch": 0.5172720797720798, + "grad_norm": 0.639999270439148, + "learning_rate": 0.00019195639204355554, + "loss": 1.2052, + "step": 2905 + }, + { + "epoch": 0.5174501424501424, + "grad_norm": 0.4630785584449768, + "learning_rate": 0.0001919508909697314, + "loss": 1.1157, + "step": 2906 + }, + { + "epoch": 0.5176282051282052, + "grad_norm": 0.513949990272522, + "learning_rate": 0.00019194538809432055, + "loss": 1.0047, + "step": 2907 + }, + { + "epoch": 0.5178062678062678, + "grad_norm": 0.488034725189209, + "learning_rate": 0.0001919398834174308, + "loss": 0.9008, + "step": 2908 + }, + { + "epoch": 0.5179843304843305, + "grad_norm": 0.4892788529396057, + "learning_rate": 0.00019193437693917006, + "loss": 1.1024, + "step": 2909 + }, + { + "epoch": 0.5181623931623932, + "grad_norm": 0.5503842830657959, + "learning_rate": 0.00019192886865964618, + "loss": 1.2283, + "step": 2910 + }, + { + "epoch": 0.5183404558404558, + "grad_norm": 0.48885393142700195, + "learning_rate": 0.00019192335857896707, + "loss": 0.9522, + "step": 2911 + }, + { + "epoch": 0.5185185185185185, + "grad_norm": 0.5479527115821838, + "learning_rate": 0.00019191784669724072, + "loss": 1.1616, + "step": 2912 + }, + { + "epoch": 0.5186965811965812, + "grad_norm": 0.42701148986816406, + "learning_rate": 0.00019191233301457506, + "loss": 0.8434, + "step": 2913 + }, + { + "epoch": 0.5188746438746439, + "grad_norm": 0.4273422658443451, + "learning_rate": 0.00019190681753107822, + "loss": 0.8316, + "step": 2914 + }, + { + "epoch": 0.5190527065527065, + "grad_norm": 0.5047736763954163, + "learning_rate": 0.00019190130024685818, + "loss": 1.171, + "step": 2915 + }, + { + "epoch": 0.5192307692307693, + "grad_norm": 0.5221177935600281, + "learning_rate": 0.00019189578116202307, + "loss": 1.0256, + "step": 2916 + }, + { + "epoch": 0.5194088319088319, + "grad_norm": 0.4782322943210602, + "learning_rate": 0.00019189026027668105, + "loss": 0.8598, + "step": 2917 + }, + { + "epoch": 0.5195868945868946, + "grad_norm": 0.5627185702323914, + "learning_rate": 0.00019188473759094022, + "loss": 1.1825, + "step": 2918 + }, + { + "epoch": 0.5197649572649573, + "grad_norm": 0.5036423206329346, + "learning_rate": 0.00019187921310490888, + "loss": 1.0881, + "step": 2919 + }, + { + "epoch": 0.51994301994302, + "grad_norm": 0.4271143972873688, + "learning_rate": 0.0001918736868186952, + "loss": 0.9265, + "step": 2920 + }, + { + "epoch": 0.5201210826210826, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00019186815873240747, + "loss": 1.196, + "step": 2921 + }, + { + "epoch": 0.5202991452991453, + "grad_norm": 0.5494198203086853, + "learning_rate": 0.00019186262884615402, + "loss": 1.1207, + "step": 2922 + }, + { + "epoch": 0.520477207977208, + "grad_norm": 0.5305119752883911, + "learning_rate": 0.0001918570971600432, + "loss": 1.0393, + "step": 2923 + }, + { + "epoch": 0.5206552706552706, + "grad_norm": 0.46713170409202576, + "learning_rate": 0.00019185156367418333, + "loss": 0.9583, + "step": 2924 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 0.597776472568512, + "learning_rate": 0.00019184602838868292, + "loss": 1.2978, + "step": 2925 + }, + { + "epoch": 0.521011396011396, + "grad_norm": 0.520976722240448, + "learning_rate": 0.00019184049130365036, + "loss": 1.0515, + "step": 2926 + }, + { + "epoch": 0.5211894586894587, + "grad_norm": 0.5266290307044983, + "learning_rate": 0.00019183495241919415, + "loss": 1.0437, + "step": 2927 + }, + { + "epoch": 0.5213675213675214, + "grad_norm": 0.50911545753479, + "learning_rate": 0.00019182941173542285, + "loss": 0.9977, + "step": 2928 + }, + { + "epoch": 0.521545584045584, + "grad_norm": 0.4924670457839966, + "learning_rate": 0.00019182386925244496, + "loss": 0.9309, + "step": 2929 + }, + { + "epoch": 0.5217236467236467, + "grad_norm": 0.4979301393032074, + "learning_rate": 0.00019181832497036912, + "loss": 0.87, + "step": 2930 + }, + { + "epoch": 0.5219017094017094, + "grad_norm": 0.6307916045188904, + "learning_rate": 0.0001918127788893039, + "loss": 1.2159, + "step": 2931 + }, + { + "epoch": 0.5220797720797721, + "grad_norm": 0.4915660619735718, + "learning_rate": 0.00019180723100935802, + "loss": 1.0828, + "step": 2932 + }, + { + "epoch": 0.5222578347578347, + "grad_norm": 0.4312742352485657, + "learning_rate": 0.00019180168133064017, + "loss": 1.0496, + "step": 2933 + }, + { + "epoch": 0.5224358974358975, + "grad_norm": 0.6006124019622803, + "learning_rate": 0.00019179612985325908, + "loss": 1.0751, + "step": 2934 + }, + { + "epoch": 0.5226139601139601, + "grad_norm": 0.5332220196723938, + "learning_rate": 0.0001917905765773235, + "loss": 1.2601, + "step": 2935 + }, + { + "epoch": 0.5227920227920227, + "grad_norm": 0.4877954423427582, + "learning_rate": 0.00019178502150294223, + "loss": 1.2279, + "step": 2936 + }, + { + "epoch": 0.5229700854700855, + "grad_norm": 0.5975968837738037, + "learning_rate": 0.00019177946463022418, + "loss": 1.3371, + "step": 2937 + }, + { + "epoch": 0.5231481481481481, + "grad_norm": 0.5363923907279968, + "learning_rate": 0.00019177390595927815, + "loss": 1.0705, + "step": 2938 + }, + { + "epoch": 0.5233262108262108, + "grad_norm": 0.4314909875392914, + "learning_rate": 0.0001917683454902131, + "loss": 0.9172, + "step": 2939 + }, + { + "epoch": 0.5235042735042735, + "grad_norm": 0.46187883615493774, + "learning_rate": 0.0001917627832231379, + "loss": 1.1201, + "step": 2940 + }, + { + "epoch": 0.5236823361823362, + "grad_norm": 0.4648260772228241, + "learning_rate": 0.00019175721915816162, + "loss": 1.1307, + "step": 2941 + }, + { + "epoch": 0.5238603988603988, + "grad_norm": 0.4427165687084198, + "learning_rate": 0.00019175165329539325, + "loss": 0.9459, + "step": 2942 + }, + { + "epoch": 0.5240384615384616, + "grad_norm": 0.4645056128501892, + "learning_rate": 0.0001917460856349418, + "loss": 0.9176, + "step": 2943 + }, + { + "epoch": 0.5242165242165242, + "grad_norm": 0.4939568042755127, + "learning_rate": 0.0001917405161769164, + "loss": 1.1056, + "step": 2944 + }, + { + "epoch": 0.5243945868945868, + "grad_norm": 0.6057310104370117, + "learning_rate": 0.00019173494492142617, + "loss": 1.2714, + "step": 2945 + }, + { + "epoch": 0.5245726495726496, + "grad_norm": 0.5038546323776245, + "learning_rate": 0.00019172937186858025, + "loss": 0.911, + "step": 2946 + }, + { + "epoch": 0.5247507122507122, + "grad_norm": 0.5521321296691895, + "learning_rate": 0.00019172379701848784, + "loss": 1.0781, + "step": 2947 + }, + { + "epoch": 0.5249287749287749, + "grad_norm": 0.516979455947876, + "learning_rate": 0.00019171822037125817, + "loss": 1.1051, + "step": 2948 + }, + { + "epoch": 0.5251068376068376, + "grad_norm": 0.5443150997161865, + "learning_rate": 0.0001917126419270005, + "loss": 1.0802, + "step": 2949 + }, + { + "epoch": 0.5252849002849003, + "grad_norm": 0.5373311042785645, + "learning_rate": 0.00019170706168582412, + "loss": 0.9313, + "step": 2950 + }, + { + "epoch": 0.5254629629629629, + "grad_norm": 0.7511917948722839, + "learning_rate": 0.0001917014796478384, + "loss": 1.1958, + "step": 2951 + }, + { + "epoch": 0.5256410256410257, + "grad_norm": 0.49893468618392944, + "learning_rate": 0.00019169589581315263, + "loss": 0.9387, + "step": 2952 + }, + { + "epoch": 0.5258190883190883, + "grad_norm": 0.48010289669036865, + "learning_rate": 0.00019169031018187628, + "loss": 1.2459, + "step": 2953 + }, + { + "epoch": 0.5259971509971509, + "grad_norm": 0.48768678307533264, + "learning_rate": 0.0001916847227541188, + "loss": 1.0127, + "step": 2954 + }, + { + "epoch": 0.5261752136752137, + "grad_norm": 0.5973068475723267, + "learning_rate": 0.00019167913352998963, + "loss": 1.1685, + "step": 2955 + }, + { + "epoch": 0.5263532763532763, + "grad_norm": 0.5567806959152222, + "learning_rate": 0.00019167354250959826, + "loss": 1.142, + "step": 2956 + }, + { + "epoch": 0.5265313390313391, + "grad_norm": 0.47819700837135315, + "learning_rate": 0.00019166794969305428, + "loss": 0.712, + "step": 2957 + }, + { + "epoch": 0.5267094017094017, + "grad_norm": 0.5191744565963745, + "learning_rate": 0.00019166235508046725, + "loss": 1.2208, + "step": 2958 + }, + { + "epoch": 0.5268874643874644, + "grad_norm": 0.4987856149673462, + "learning_rate": 0.00019165675867194675, + "loss": 1.0466, + "step": 2959 + }, + { + "epoch": 0.5270655270655271, + "grad_norm": 0.5017665028572083, + "learning_rate": 0.0001916511604676025, + "loss": 1.1236, + "step": 2960 + }, + { + "epoch": 0.5272435897435898, + "grad_norm": 0.5115348696708679, + "learning_rate": 0.00019164556046754415, + "loss": 1.1497, + "step": 2961 + }, + { + "epoch": 0.5274216524216524, + "grad_norm": 0.4934345781803131, + "learning_rate": 0.0001916399586718814, + "loss": 1.0183, + "step": 2962 + }, + { + "epoch": 0.5275997150997151, + "grad_norm": 0.5033719539642334, + "learning_rate": 0.00019163435508072404, + "loss": 1.0256, + "step": 2963 + }, + { + "epoch": 0.5277777777777778, + "grad_norm": 0.5325372219085693, + "learning_rate": 0.00019162874969418184, + "loss": 1.1384, + "step": 2964 + }, + { + "epoch": 0.5279558404558404, + "grad_norm": 0.4901772141456604, + "learning_rate": 0.00019162314251236465, + "loss": 1.0831, + "step": 2965 + }, + { + "epoch": 0.5281339031339032, + "grad_norm": 0.4743805229663849, + "learning_rate": 0.0001916175335353823, + "loss": 1.1894, + "step": 2966 + }, + { + "epoch": 0.5283119658119658, + "grad_norm": 0.5439450740814209, + "learning_rate": 0.00019161192276334466, + "loss": 1.2066, + "step": 2967 + }, + { + "epoch": 0.5284900284900285, + "grad_norm": 0.5123090744018555, + "learning_rate": 0.00019160631019636174, + "loss": 1.1829, + "step": 2968 + }, + { + "epoch": 0.5286680911680912, + "grad_norm": 0.5995343923568726, + "learning_rate": 0.00019160069583454346, + "loss": 1.4872, + "step": 2969 + }, + { + "epoch": 0.5288461538461539, + "grad_norm": 0.4596657156944275, + "learning_rate": 0.00019159507967799985, + "loss": 0.8948, + "step": 2970 + }, + { + "epoch": 0.5290242165242165, + "grad_norm": 0.5533682107925415, + "learning_rate": 0.0001915894617268409, + "loss": 1.1779, + "step": 2971 + }, + { + "epoch": 0.5292022792022792, + "grad_norm": 0.3860718309879303, + "learning_rate": 0.00019158384198117673, + "loss": 0.6424, + "step": 2972 + }, + { + "epoch": 0.5293803418803419, + "grad_norm": 0.47424063086509705, + "learning_rate": 0.0001915782204411174, + "loss": 1.1592, + "step": 2973 + }, + { + "epoch": 0.5295584045584045, + "grad_norm": 0.5050228834152222, + "learning_rate": 0.00019157259710677309, + "loss": 1.1971, + "step": 2974 + }, + { + "epoch": 0.5297364672364673, + "grad_norm": 0.6080113649368286, + "learning_rate": 0.00019156697197825396, + "loss": 1.1511, + "step": 2975 + }, + { + "epoch": 0.5299145299145299, + "grad_norm": 0.4805932641029358, + "learning_rate": 0.00019156134505567024, + "loss": 1.1033, + "step": 2976 + }, + { + "epoch": 0.5300925925925926, + "grad_norm": 0.4835345447063446, + "learning_rate": 0.00019155571633913215, + "loss": 1.1832, + "step": 2977 + }, + { + "epoch": 0.5302706552706553, + "grad_norm": 0.5183725953102112, + "learning_rate": 0.00019155008582875, + "loss": 0.9221, + "step": 2978 + }, + { + "epoch": 0.530448717948718, + "grad_norm": 0.48015761375427246, + "learning_rate": 0.00019154445352463412, + "loss": 1.045, + "step": 2979 + }, + { + "epoch": 0.5306267806267806, + "grad_norm": 0.4670043885707855, + "learning_rate": 0.0001915388194268948, + "loss": 0.9025, + "step": 2980 + }, + { + "epoch": 0.5308048433048433, + "grad_norm": 0.5048824548721313, + "learning_rate": 0.0001915331835356425, + "loss": 1.0681, + "step": 2981 + }, + { + "epoch": 0.530982905982906, + "grad_norm": 0.4785633981227875, + "learning_rate": 0.00019152754585098758, + "loss": 1.0097, + "step": 2982 + }, + { + "epoch": 0.5311609686609686, + "grad_norm": 0.4829573333263397, + "learning_rate": 0.00019152190637304056, + "loss": 1.0856, + "step": 2983 + }, + { + "epoch": 0.5313390313390314, + "grad_norm": 0.5425563454627991, + "learning_rate": 0.00019151626510191189, + "loss": 1.2313, + "step": 2984 + }, + { + "epoch": 0.531517094017094, + "grad_norm": 0.5532251596450806, + "learning_rate": 0.0001915106220377121, + "loss": 1.0328, + "step": 2985 + }, + { + "epoch": 0.5316951566951567, + "grad_norm": 0.47016972303390503, + "learning_rate": 0.0001915049771805518, + "loss": 1.2003, + "step": 2986 + }, + { + "epoch": 0.5318732193732194, + "grad_norm": 0.5241743326187134, + "learning_rate": 0.00019149933053054153, + "loss": 1.046, + "step": 2987 + }, + { + "epoch": 0.532051282051282, + "grad_norm": 0.5043526887893677, + "learning_rate": 0.00019149368208779197, + "loss": 1.0022, + "step": 2988 + }, + { + "epoch": 0.5322293447293447, + "grad_norm": 0.5563312768936157, + "learning_rate": 0.00019148803185241374, + "loss": 1.1017, + "step": 2989 + }, + { + "epoch": 0.5324074074074074, + "grad_norm": 0.5414231419563293, + "learning_rate": 0.00019148237982451763, + "loss": 0.9649, + "step": 2990 + }, + { + "epoch": 0.5325854700854701, + "grad_norm": 0.5452231764793396, + "learning_rate": 0.0001914767260042143, + "loss": 1.2281, + "step": 2991 + }, + { + "epoch": 0.5327635327635327, + "grad_norm": 0.5500698685646057, + "learning_rate": 0.00019147107039161454, + "loss": 1.2865, + "step": 2992 + }, + { + "epoch": 0.5329415954415955, + "grad_norm": 0.49747416377067566, + "learning_rate": 0.00019146541298682918, + "loss": 1.1296, + "step": 2993 + }, + { + "epoch": 0.5331196581196581, + "grad_norm": 0.5684167742729187, + "learning_rate": 0.00019145975378996903, + "loss": 1.0685, + "step": 2994 + }, + { + "epoch": 0.5332977207977208, + "grad_norm": 0.5411235690116882, + "learning_rate": 0.00019145409280114502, + "loss": 1.1372, + "step": 2995 + }, + { + "epoch": 0.5334757834757835, + "grad_norm": 0.5006675720214844, + "learning_rate": 0.00019144843002046806, + "loss": 1.0688, + "step": 2996 + }, + { + "epoch": 0.5336538461538461, + "grad_norm": 0.4591315686702728, + "learning_rate": 0.00019144276544804908, + "loss": 1.1071, + "step": 2997 + }, + { + "epoch": 0.5338319088319088, + "grad_norm": 0.5615306496620178, + "learning_rate": 0.000191437099083999, + "loss": 1.1033, + "step": 2998 + }, + { + "epoch": 0.5340099715099715, + "grad_norm": 0.4986817240715027, + "learning_rate": 0.00019143143092842897, + "loss": 1.176, + "step": 2999 + }, + { + "epoch": 0.5341880341880342, + "grad_norm": 0.5017120242118835, + "learning_rate": 0.00019142576098144995, + "loss": 1.0174, + "step": 3000 + }, + { + "epoch": 0.5343660968660968, + "grad_norm": 0.508298397064209, + "learning_rate": 0.0001914200892431731, + "loss": 1.164, + "step": 3001 + }, + { + "epoch": 0.5345441595441596, + "grad_norm": 0.48068809509277344, + "learning_rate": 0.0001914144157137095, + "loss": 0.7959, + "step": 3002 + }, + { + "epoch": 0.5347222222222222, + "grad_norm": 0.6347028017044067, + "learning_rate": 0.0001914087403931703, + "loss": 1.1727, + "step": 3003 + }, + { + "epoch": 0.5349002849002849, + "grad_norm": 0.5558401942253113, + "learning_rate": 0.00019140306328166676, + "loss": 1.2282, + "step": 3004 + }, + { + "epoch": 0.5350783475783476, + "grad_norm": 0.5093596577644348, + "learning_rate": 0.00019139738437931004, + "loss": 1.3258, + "step": 3005 + }, + { + "epoch": 0.5352564102564102, + "grad_norm": 0.4653106927871704, + "learning_rate": 0.0001913917036862114, + "loss": 1.1062, + "step": 3006 + }, + { + "epoch": 0.5354344729344729, + "grad_norm": 0.48085781931877136, + "learning_rate": 0.00019138602120248222, + "loss": 0.9019, + "step": 3007 + }, + { + "epoch": 0.5356125356125356, + "grad_norm": 0.5174745321273804, + "learning_rate": 0.0001913803369282338, + "loss": 1.044, + "step": 3008 + }, + { + "epoch": 0.5357905982905983, + "grad_norm": 0.5359669327735901, + "learning_rate": 0.00019137465086357746, + "loss": 1.0723, + "step": 3009 + }, + { + "epoch": 0.5359686609686609, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.00019136896300862467, + "loss": 1.2192, + "step": 3010 + }, + { + "epoch": 0.5361467236467237, + "grad_norm": 0.4905693829059601, + "learning_rate": 0.00019136327336348688, + "loss": 1.2372, + "step": 3011 + }, + { + "epoch": 0.5363247863247863, + "grad_norm": 0.5741264820098877, + "learning_rate": 0.0001913575819282755, + "loss": 1.1703, + "step": 3012 + }, + { + "epoch": 0.5365028490028491, + "grad_norm": 0.577033281326294, + "learning_rate": 0.0001913518887031021, + "loss": 1.1555, + "step": 3013 + }, + { + "epoch": 0.5366809116809117, + "grad_norm": 0.46795153617858887, + "learning_rate": 0.00019134619368807822, + "loss": 0.8583, + "step": 3014 + }, + { + "epoch": 0.5368589743589743, + "grad_norm": 0.5973345637321472, + "learning_rate": 0.0001913404968833154, + "loss": 1.1509, + "step": 3015 + }, + { + "epoch": 0.5370370370370371, + "grad_norm": 0.62020343542099, + "learning_rate": 0.00019133479828892531, + "loss": 1.0781, + "step": 3016 + }, + { + "epoch": 0.5372150997150997, + "grad_norm": 0.5342286229133606, + "learning_rate": 0.00019132909790501958, + "loss": 1.1556, + "step": 3017 + }, + { + "epoch": 0.5373931623931624, + "grad_norm": 0.49612846970558167, + "learning_rate": 0.0001913233957317099, + "loss": 0.9027, + "step": 3018 + }, + { + "epoch": 0.5375712250712251, + "grad_norm": 0.5403908491134644, + "learning_rate": 0.00019131769176910796, + "loss": 1.1125, + "step": 3019 + }, + { + "epoch": 0.5377492877492878, + "grad_norm": 0.4952050447463989, + "learning_rate": 0.0001913119860173256, + "loss": 1.2329, + "step": 3020 + }, + { + "epoch": 0.5379273504273504, + "grad_norm": 0.5877819657325745, + "learning_rate": 0.0001913062784764745, + "loss": 1.2855, + "step": 3021 + }, + { + "epoch": 0.5381054131054132, + "grad_norm": 0.49312907457351685, + "learning_rate": 0.00019130056914666655, + "loss": 1.0212, + "step": 3022 + }, + { + "epoch": 0.5382834757834758, + "grad_norm": 0.45544490218162537, + "learning_rate": 0.00019129485802801366, + "loss": 0.9748, + "step": 3023 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.5535242557525635, + "learning_rate": 0.00019128914512062762, + "loss": 1.2134, + "step": 3024 + }, + { + "epoch": 0.5386396011396012, + "grad_norm": 0.45369696617126465, + "learning_rate": 0.00019128343042462044, + "loss": 0.9964, + "step": 3025 + }, + { + "epoch": 0.5388176638176638, + "grad_norm": 0.6240725517272949, + "learning_rate": 0.00019127771394010406, + "loss": 1.425, + "step": 3026 + }, + { + "epoch": 0.5389957264957265, + "grad_norm": 0.4859573245048523, + "learning_rate": 0.0001912719956671905, + "loss": 1.087, + "step": 3027 + }, + { + "epoch": 0.5391737891737892, + "grad_norm": 0.47529762983322144, + "learning_rate": 0.0001912662756059918, + "loss": 0.9517, + "step": 3028 + }, + { + "epoch": 0.5393518518518519, + "grad_norm": 0.5317288637161255, + "learning_rate": 0.00019126055375661997, + "loss": 1.0945, + "step": 3029 + }, + { + "epoch": 0.5395299145299145, + "grad_norm": 0.55974280834198, + "learning_rate": 0.00019125483011918722, + "loss": 1.0794, + "step": 3030 + }, + { + "epoch": 0.5397079772079773, + "grad_norm": 0.48579123616218567, + "learning_rate": 0.0001912491046938056, + "loss": 1.1421, + "step": 3031 + }, + { + "epoch": 0.5398860398860399, + "grad_norm": 0.4917181134223938, + "learning_rate": 0.00019124337748058733, + "loss": 0.9708, + "step": 3032 + }, + { + "epoch": 0.5400641025641025, + "grad_norm": 0.525291383266449, + "learning_rate": 0.00019123764847964466, + "loss": 1.064, + "step": 3033 + }, + { + "epoch": 0.5402421652421653, + "grad_norm": 0.5733301639556885, + "learning_rate": 0.00019123191769108977, + "loss": 1.2142, + "step": 3034 + }, + { + "epoch": 0.5404202279202279, + "grad_norm": 0.5400987863540649, + "learning_rate": 0.00019122618511503494, + "loss": 1.1309, + "step": 3035 + }, + { + "epoch": 0.5405982905982906, + "grad_norm": 0.6261051893234253, + "learning_rate": 0.00019122045075159257, + "loss": 1.2112, + "step": 3036 + }, + { + "epoch": 0.5407763532763533, + "grad_norm": 0.5483576059341431, + "learning_rate": 0.0001912147146008749, + "loss": 1.2705, + "step": 3037 + }, + { + "epoch": 0.540954415954416, + "grad_norm": 0.5442137122154236, + "learning_rate": 0.00019120897666299443, + "loss": 1.2512, + "step": 3038 + }, + { + "epoch": 0.5411324786324786, + "grad_norm": 0.5680811405181885, + "learning_rate": 0.00019120323693806355, + "loss": 1.392, + "step": 3039 + }, + { + "epoch": 0.5413105413105413, + "grad_norm": 0.5237287878990173, + "learning_rate": 0.00019119749542619466, + "loss": 1.1599, + "step": 3040 + }, + { + "epoch": 0.541488603988604, + "grad_norm": 0.48119300603866577, + "learning_rate": 0.00019119175212750032, + "loss": 1.0976, + "step": 3041 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 0.507033109664917, + "learning_rate": 0.00019118600704209302, + "loss": 1.0181, + "step": 3042 + }, + { + "epoch": 0.5418447293447294, + "grad_norm": 0.484672874212265, + "learning_rate": 0.00019118026017008531, + "loss": 1.1636, + "step": 3043 + }, + { + "epoch": 0.542022792022792, + "grad_norm": 0.4923502206802368, + "learning_rate": 0.00019117451151158985, + "loss": 1.0388, + "step": 3044 + }, + { + "epoch": 0.5422008547008547, + "grad_norm": 0.4882057309150696, + "learning_rate": 0.00019116876106671922, + "loss": 1.131, + "step": 3045 + }, + { + "epoch": 0.5423789173789174, + "grad_norm": 0.6068355441093445, + "learning_rate": 0.0001911630088355861, + "loss": 1.3218, + "step": 3046 + }, + { + "epoch": 0.54255698005698, + "grad_norm": 0.5012881755828857, + "learning_rate": 0.0001911572548183032, + "loss": 1.0514, + "step": 3047 + }, + { + "epoch": 0.5427350427350427, + "grad_norm": 0.49849793314933777, + "learning_rate": 0.00019115149901498328, + "loss": 1.0003, + "step": 3048 + }, + { + "epoch": 0.5429131054131054, + "grad_norm": 0.4934251010417938, + "learning_rate": 0.00019114574142573904, + "loss": 1.0319, + "step": 3049 + }, + { + "epoch": 0.5430911680911681, + "grad_norm": 0.4947762191295624, + "learning_rate": 0.00019113998205068334, + "loss": 1.0906, + "step": 3050 + }, + { + "epoch": 0.5432692307692307, + "grad_norm": 0.5449416041374207, + "learning_rate": 0.00019113422088992907, + "loss": 0.9093, + "step": 3051 + }, + { + "epoch": 0.5434472934472935, + "grad_norm": 0.49395284056663513, + "learning_rate": 0.00019112845794358902, + "loss": 1.0071, + "step": 3052 + }, + { + "epoch": 0.5436253561253561, + "grad_norm": 0.5478728413581848, + "learning_rate": 0.00019112269321177613, + "loss": 1.2124, + "step": 3053 + }, + { + "epoch": 0.5438034188034188, + "grad_norm": 0.6205173134803772, + "learning_rate": 0.0001911169266946034, + "loss": 1.021, + "step": 3054 + }, + { + "epoch": 0.5439814814814815, + "grad_norm": 0.4777783751487732, + "learning_rate": 0.00019111115839218372, + "loss": 0.9192, + "step": 3055 + }, + { + "epoch": 0.5441595441595442, + "grad_norm": 0.5541689991950989, + "learning_rate": 0.00019110538830463018, + "loss": 1.1248, + "step": 3056 + }, + { + "epoch": 0.5443376068376068, + "grad_norm": 0.4750942289829254, + "learning_rate": 0.0001910996164320558, + "loss": 1.3147, + "step": 3057 + }, + { + "epoch": 0.5445156695156695, + "grad_norm": 0.6283948421478271, + "learning_rate": 0.0001910938427745737, + "loss": 1.0919, + "step": 3058 + }, + { + "epoch": 0.5446937321937322, + "grad_norm": 0.552725076675415, + "learning_rate": 0.00019108806733229698, + "loss": 1.3807, + "step": 3059 + }, + { + "epoch": 0.5448717948717948, + "grad_norm": 0.4832848310470581, + "learning_rate": 0.0001910822901053388, + "loss": 1.0705, + "step": 3060 + }, + { + "epoch": 0.5450498575498576, + "grad_norm": 0.6468375325202942, + "learning_rate": 0.00019107651109381233, + "loss": 1.0766, + "step": 3061 + }, + { + "epoch": 0.5452279202279202, + "grad_norm": 0.5464920401573181, + "learning_rate": 0.00019107073029783083, + "loss": 1.0453, + "step": 3062 + }, + { + "epoch": 0.5454059829059829, + "grad_norm": 0.5321210026741028, + "learning_rate": 0.0001910649477175076, + "loss": 1.2326, + "step": 3063 + }, + { + "epoch": 0.5455840455840456, + "grad_norm": 0.5572962164878845, + "learning_rate": 0.00019105916335295582, + "loss": 1.0673, + "step": 3064 + }, + { + "epoch": 0.5457621082621082, + "grad_norm": 0.5239177942276001, + "learning_rate": 0.00019105337720428894, + "loss": 1.04, + "step": 3065 + }, + { + "epoch": 0.5459401709401709, + "grad_norm": 0.5633319616317749, + "learning_rate": 0.00019104758927162023, + "loss": 0.9606, + "step": 3066 + }, + { + "epoch": 0.5461182336182336, + "grad_norm": 0.5317914485931396, + "learning_rate": 0.0001910417995550632, + "loss": 1.0651, + "step": 3067 + }, + { + "epoch": 0.5462962962962963, + "grad_norm": 0.5126453638076782, + "learning_rate": 0.00019103600805473118, + "loss": 1.0316, + "step": 3068 + }, + { + "epoch": 0.5464743589743589, + "grad_norm": 0.5262107253074646, + "learning_rate": 0.00019103021477073773, + "loss": 1.0752, + "step": 3069 + }, + { + "epoch": 0.5466524216524217, + "grad_norm": 0.5384877324104309, + "learning_rate": 0.0001910244197031963, + "loss": 1.1731, + "step": 3070 + }, + { + "epoch": 0.5468304843304843, + "grad_norm": 0.5126553773880005, + "learning_rate": 0.00019101862285222048, + "loss": 1.2229, + "step": 3071 + }, + { + "epoch": 0.5470085470085471, + "grad_norm": 0.4841194450855255, + "learning_rate": 0.0001910128242179238, + "loss": 0.9955, + "step": 3072 + }, + { + "epoch": 0.5471866096866097, + "grad_norm": 0.526546061038971, + "learning_rate": 0.00019100702380041987, + "loss": 1.2436, + "step": 3073 + }, + { + "epoch": 0.5473646723646723, + "grad_norm": 0.5085833072662354, + "learning_rate": 0.0001910012215998224, + "loss": 1.011, + "step": 3074 + }, + { + "epoch": 0.5475427350427351, + "grad_norm": 0.5149994492530823, + "learning_rate": 0.000190995417616245, + "loss": 0.8632, + "step": 3075 + }, + { + "epoch": 0.5477207977207977, + "grad_norm": 0.48079630732536316, + "learning_rate": 0.00019098961184980145, + "loss": 1.1115, + "step": 3076 + }, + { + "epoch": 0.5478988603988604, + "grad_norm": 0.5769477486610413, + "learning_rate": 0.00019098380430060546, + "loss": 0.9544, + "step": 3077 + }, + { + "epoch": 0.5480769230769231, + "grad_norm": 0.5260093808174133, + "learning_rate": 0.0001909779949687708, + "loss": 1.2354, + "step": 3078 + }, + { + "epoch": 0.5482549857549858, + "grad_norm": 0.5518734455108643, + "learning_rate": 0.00019097218385441135, + "loss": 1.1944, + "step": 3079 + }, + { + "epoch": 0.5484330484330484, + "grad_norm": 0.5436808466911316, + "learning_rate": 0.00019096637095764095, + "loss": 1.0717, + "step": 3080 + }, + { + "epoch": 0.5486111111111112, + "grad_norm": 0.4749584197998047, + "learning_rate": 0.00019096055627857344, + "loss": 1.0417, + "step": 3081 + }, + { + "epoch": 0.5487891737891738, + "grad_norm": 0.5485591292381287, + "learning_rate": 0.0001909547398173228, + "loss": 1.2515, + "step": 3082 + }, + { + "epoch": 0.5489672364672364, + "grad_norm": 0.5751016736030579, + "learning_rate": 0.00019094892157400296, + "loss": 1.2112, + "step": 3083 + }, + { + "epoch": 0.5491452991452992, + "grad_norm": 0.5404475331306458, + "learning_rate": 0.00019094310154872795, + "loss": 0.4334, + "step": 3084 + }, + { + "epoch": 0.5493233618233618, + "grad_norm": 0.5198020935058594, + "learning_rate": 0.00019093727974161178, + "loss": 0.9759, + "step": 3085 + }, + { + "epoch": 0.5495014245014245, + "grad_norm": 0.4893439710140228, + "learning_rate": 0.0001909314561527685, + "loss": 1.1287, + "step": 3086 + }, + { + "epoch": 0.5496794871794872, + "grad_norm": 0.5675956606864929, + "learning_rate": 0.00019092563078231228, + "loss": 1.234, + "step": 3087 + }, + { + "epoch": 0.5498575498575499, + "grad_norm": 0.5539132356643677, + "learning_rate": 0.00019091980363035714, + "loss": 1.2378, + "step": 3088 + }, + { + "epoch": 0.5500356125356125, + "grad_norm": 0.5194353461265564, + "learning_rate": 0.00019091397469701735, + "loss": 1.1338, + "step": 3089 + }, + { + "epoch": 0.5502136752136753, + "grad_norm": 0.5143756866455078, + "learning_rate": 0.0001909081439824071, + "loss": 0.9118, + "step": 3090 + }, + { + "epoch": 0.5503917378917379, + "grad_norm": 0.5624327659606934, + "learning_rate": 0.0001909023114866406, + "loss": 1.035, + "step": 3091 + }, + { + "epoch": 0.5505698005698005, + "grad_norm": 0.5285067558288574, + "learning_rate": 0.0001908964772098321, + "loss": 1.0451, + "step": 3092 + }, + { + "epoch": 0.5507478632478633, + "grad_norm": 0.5730587244033813, + "learning_rate": 0.000190890641152096, + "loss": 1.0672, + "step": 3093 + }, + { + "epoch": 0.5509259259259259, + "grad_norm": 0.5822951197624207, + "learning_rate": 0.0001908848033135466, + "loss": 1.1791, + "step": 3094 + }, + { + "epoch": 0.5511039886039886, + "grad_norm": 0.596161961555481, + "learning_rate": 0.00019087896369429826, + "loss": 1.0954, + "step": 3095 + }, + { + "epoch": 0.5512820512820513, + "grad_norm": 0.5138190984725952, + "learning_rate": 0.00019087312229446542, + "loss": 0.896, + "step": 3096 + }, + { + "epoch": 0.551460113960114, + "grad_norm": 0.5061872601509094, + "learning_rate": 0.0001908672791141625, + "loss": 1.1017, + "step": 3097 + }, + { + "epoch": 0.5516381766381766, + "grad_norm": 0.5189547538757324, + "learning_rate": 0.00019086143415350404, + "loss": 1.2906, + "step": 3098 + }, + { + "epoch": 0.5518162393162394, + "grad_norm": 0.5640039443969727, + "learning_rate": 0.00019085558741260448, + "loss": 1.1001, + "step": 3099 + }, + { + "epoch": 0.551994301994302, + "grad_norm": 0.453867107629776, + "learning_rate": 0.00019084973889157844, + "loss": 0.9731, + "step": 3100 + }, + { + "epoch": 0.5521723646723646, + "grad_norm": 0.5431303977966309, + "learning_rate": 0.0001908438885905405, + "loss": 1.3511, + "step": 3101 + }, + { + "epoch": 0.5523504273504274, + "grad_norm": 0.47693368792533875, + "learning_rate": 0.00019083803650960527, + "loss": 1.0426, + "step": 3102 + }, + { + "epoch": 0.55252849002849, + "grad_norm": 0.4663422703742981, + "learning_rate": 0.00019083218264888743, + "loss": 1.05, + "step": 3103 + }, + { + "epoch": 0.5527065527065527, + "grad_norm": 0.561354398727417, + "learning_rate": 0.00019082632700850164, + "loss": 0.9608, + "step": 3104 + }, + { + "epoch": 0.5528846153846154, + "grad_norm": 0.4981916844844818, + "learning_rate": 0.00019082046958856266, + "loss": 1.1935, + "step": 3105 + }, + { + "epoch": 0.5530626780626781, + "grad_norm": 0.5301326513290405, + "learning_rate": 0.0001908146103891852, + "loss": 1.0646, + "step": 3106 + }, + { + "epoch": 0.5532407407407407, + "grad_norm": 0.5023610591888428, + "learning_rate": 0.00019080874941048416, + "loss": 1.127, + "step": 3107 + }, + { + "epoch": 0.5534188034188035, + "grad_norm": 0.5172514319419861, + "learning_rate": 0.00019080288665257426, + "loss": 1.0435, + "step": 3108 + }, + { + "epoch": 0.5535968660968661, + "grad_norm": 0.6340598464012146, + "learning_rate": 0.00019079702211557048, + "loss": 1.3528, + "step": 3109 + }, + { + "epoch": 0.5537749287749287, + "grad_norm": 0.46882256865501404, + "learning_rate": 0.0001907911557995876, + "loss": 1.1361, + "step": 3110 + }, + { + "epoch": 0.5539529914529915, + "grad_norm": 0.6401382088661194, + "learning_rate": 0.00019078528770474068, + "loss": 1.2415, + "step": 3111 + }, + { + "epoch": 0.5541310541310541, + "grad_norm": 0.5141328573226929, + "learning_rate": 0.00019077941783114463, + "loss": 1.0505, + "step": 3112 + }, + { + "epoch": 0.5543091168091168, + "grad_norm": 0.522318959236145, + "learning_rate": 0.00019077354617891444, + "loss": 1.0964, + "step": 3113 + }, + { + "epoch": 0.5544871794871795, + "grad_norm": 0.539551854133606, + "learning_rate": 0.00019076767274816517, + "loss": 1.0735, + "step": 3114 + }, + { + "epoch": 0.5546652421652422, + "grad_norm": 0.495320200920105, + "learning_rate": 0.00019076179753901195, + "loss": 0.9754, + "step": 3115 + }, + { + "epoch": 0.5548433048433048, + "grad_norm": 0.5499199628829956, + "learning_rate": 0.00019075592055156984, + "loss": 1.0043, + "step": 3116 + }, + { + "epoch": 0.5550213675213675, + "grad_norm": 0.5352509617805481, + "learning_rate": 0.00019075004178595396, + "loss": 1.1701, + "step": 3117 + }, + { + "epoch": 0.5551994301994302, + "grad_norm": 0.5392300486564636, + "learning_rate": 0.00019074416124227953, + "loss": 1.1612, + "step": 3118 + }, + { + "epoch": 0.5553774928774928, + "grad_norm": 0.5195050835609436, + "learning_rate": 0.0001907382789206618, + "loss": 1.0934, + "step": 3119 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 0.5276884436607361, + "learning_rate": 0.000190732394821216, + "loss": 0.9011, + "step": 3120 + }, + { + "epoch": 0.5557336182336182, + "grad_norm": 0.6115903258323669, + "learning_rate": 0.00019072650894405734, + "loss": 1.3065, + "step": 3121 + }, + { + "epoch": 0.5559116809116809, + "grad_norm": 0.5752483010292053, + "learning_rate": 0.00019072062128930127, + "loss": 1.0063, + "step": 3122 + }, + { + "epoch": 0.5560897435897436, + "grad_norm": 0.5508273243904114, + "learning_rate": 0.00019071473185706302, + "loss": 1.2598, + "step": 3123 + }, + { + "epoch": 0.5562678062678063, + "grad_norm": 0.49712198972702026, + "learning_rate": 0.00019070884064745808, + "loss": 0.924, + "step": 3124 + }, + { + "epoch": 0.5564458689458689, + "grad_norm": 0.572849452495575, + "learning_rate": 0.00019070294766060185, + "loss": 0.9683, + "step": 3125 + }, + { + "epoch": 0.5566239316239316, + "grad_norm": 0.4807920753955841, + "learning_rate": 0.00019069705289660976, + "loss": 1.0998, + "step": 3126 + }, + { + "epoch": 0.5568019943019943, + "grad_norm": 0.5543031096458435, + "learning_rate": 0.0001906911563555973, + "loss": 1.0878, + "step": 3127 + }, + { + "epoch": 0.5569800569800569, + "grad_norm": 0.5710418820381165, + "learning_rate": 0.00019068525803768007, + "loss": 1.0381, + "step": 3128 + }, + { + "epoch": 0.5571581196581197, + "grad_norm": 0.5169163346290588, + "learning_rate": 0.00019067935794297357, + "loss": 1.1149, + "step": 3129 + }, + { + "epoch": 0.5573361823361823, + "grad_norm": 0.6474376916885376, + "learning_rate": 0.00019067345607159345, + "loss": 0.9828, + "step": 3130 + }, + { + "epoch": 0.5575142450142451, + "grad_norm": 0.5029847621917725, + "learning_rate": 0.0001906675524236553, + "loss": 0.797, + "step": 3131 + }, + { + "epoch": 0.5576923076923077, + "grad_norm": 0.5681431293487549, + "learning_rate": 0.00019066164699927478, + "loss": 1.1565, + "step": 3132 + }, + { + "epoch": 0.5578703703703703, + "grad_norm": 0.5654549598693848, + "learning_rate": 0.00019065573979856764, + "loss": 1.2488, + "step": 3133 + }, + { + "epoch": 0.5580484330484331, + "grad_norm": 0.47653043270111084, + "learning_rate": 0.0001906498308216496, + "loss": 1.0428, + "step": 3134 + }, + { + "epoch": 0.5582264957264957, + "grad_norm": 0.5068467259407043, + "learning_rate": 0.00019064392006863643, + "loss": 0.9659, + "step": 3135 + }, + { + "epoch": 0.5584045584045584, + "grad_norm": 0.7076661586761475, + "learning_rate": 0.00019063800753964393, + "loss": 1.1289, + "step": 3136 + }, + { + "epoch": 0.5585826210826211, + "grad_norm": 0.551456868648529, + "learning_rate": 0.000190632093234788, + "loss": 1.1925, + "step": 3137 + }, + { + "epoch": 0.5587606837606838, + "grad_norm": 0.518276035785675, + "learning_rate": 0.00019062617715418442, + "loss": 0.8681, + "step": 3138 + }, + { + "epoch": 0.5589387464387464, + "grad_norm": 0.5272278785705566, + "learning_rate": 0.0001906202592979492, + "loss": 1.0865, + "step": 3139 + }, + { + "epoch": 0.5591168091168092, + "grad_norm": 0.5344942212104797, + "learning_rate": 0.00019061433966619822, + "loss": 1.1647, + "step": 3140 + }, + { + "epoch": 0.5592948717948718, + "grad_norm": 0.5833460092544556, + "learning_rate": 0.00019060841825904753, + "loss": 1.3403, + "step": 3141 + }, + { + "epoch": 0.5594729344729344, + "grad_norm": 0.5707054734230042, + "learning_rate": 0.00019060249507661306, + "loss": 1.1236, + "step": 3142 + }, + { + "epoch": 0.5596509971509972, + "grad_norm": 0.5446065664291382, + "learning_rate": 0.00019059657011901094, + "loss": 1.017, + "step": 3143 + }, + { + "epoch": 0.5598290598290598, + "grad_norm": 0.5285109281539917, + "learning_rate": 0.0001905906433863572, + "loss": 1.3186, + "step": 3144 + }, + { + "epoch": 0.5600071225071225, + "grad_norm": 0.5308659672737122, + "learning_rate": 0.00019058471487876802, + "loss": 0.8464, + "step": 3145 + }, + { + "epoch": 0.5601851851851852, + "grad_norm": 0.5218054056167603, + "learning_rate": 0.00019057878459635948, + "loss": 1.0219, + "step": 3146 + }, + { + "epoch": 0.5603632478632479, + "grad_norm": 0.45067787170410156, + "learning_rate": 0.00019057285253924785, + "loss": 1.0364, + "step": 3147 + }, + { + "epoch": 0.5605413105413105, + "grad_norm": 0.4856041669845581, + "learning_rate": 0.0001905669187075493, + "loss": 1.1928, + "step": 3148 + }, + { + "epoch": 0.5607193732193733, + "grad_norm": 0.506912112236023, + "learning_rate": 0.00019056098310138016, + "loss": 1.119, + "step": 3149 + }, + { + "epoch": 0.5608974358974359, + "grad_norm": 0.49049463868141174, + "learning_rate": 0.00019055504572085662, + "loss": 1.2165, + "step": 3150 + }, + { + "epoch": 0.5610754985754985, + "grad_norm": 0.5250293612480164, + "learning_rate": 0.0001905491065660951, + "loss": 1.1427, + "step": 3151 + }, + { + "epoch": 0.5612535612535613, + "grad_norm": 0.43438446521759033, + "learning_rate": 0.00019054316563721195, + "loss": 0.884, + "step": 3152 + }, + { + "epoch": 0.5614316239316239, + "grad_norm": 0.5386807918548584, + "learning_rate": 0.00019053722293432354, + "loss": 1.1494, + "step": 3153 + }, + { + "epoch": 0.5616096866096866, + "grad_norm": 0.5403809547424316, + "learning_rate": 0.00019053127845754632, + "loss": 1.1743, + "step": 3154 + }, + { + "epoch": 0.5617877492877493, + "grad_norm": 0.4759823977947235, + "learning_rate": 0.00019052533220699678, + "loss": 1.0716, + "step": 3155 + }, + { + "epoch": 0.561965811965812, + "grad_norm": 0.45332327485084534, + "learning_rate": 0.0001905193841827914, + "loss": 0.8405, + "step": 3156 + }, + { + "epoch": 0.5621438746438746, + "grad_norm": 0.5617053508758545, + "learning_rate": 0.00019051343438504671, + "loss": 1.0422, + "step": 3157 + }, + { + "epoch": 0.5623219373219374, + "grad_norm": 0.5088049173355103, + "learning_rate": 0.00019050748281387931, + "loss": 1.0067, + "step": 3158 + }, + { + "epoch": 0.5625, + "grad_norm": 0.5174484848976135, + "learning_rate": 0.00019050152946940578, + "loss": 1.0623, + "step": 3159 + }, + { + "epoch": 0.5626780626780626, + "grad_norm": 0.6093568801879883, + "learning_rate": 0.0001904955743517428, + "loss": 1.24, + "step": 3160 + }, + { + "epoch": 0.5628561253561254, + "grad_norm": 0.49063584208488464, + "learning_rate": 0.00019048961746100703, + "loss": 0.8563, + "step": 3161 + }, + { + "epoch": 0.563034188034188, + "grad_norm": 0.583940863609314, + "learning_rate": 0.00019048365879731517, + "loss": 1.0695, + "step": 3162 + }, + { + "epoch": 0.5632122507122507, + "grad_norm": 0.4943268597126007, + "learning_rate": 0.000190477698360784, + "loss": 0.8606, + "step": 3163 + }, + { + "epoch": 0.5633903133903134, + "grad_norm": 0.5050932168960571, + "learning_rate": 0.00019047173615153028, + "loss": 1.1591, + "step": 3164 + }, + { + "epoch": 0.5635683760683761, + "grad_norm": 0.5445677638053894, + "learning_rate": 0.0001904657721696708, + "loss": 1.262, + "step": 3165 + }, + { + "epoch": 0.5637464387464387, + "grad_norm": 0.5445297360420227, + "learning_rate": 0.00019045980641532246, + "loss": 1.223, + "step": 3166 + }, + { + "epoch": 0.5639245014245015, + "grad_norm": 0.5098413228988647, + "learning_rate": 0.00019045383888860213, + "loss": 1.0829, + "step": 3167 + }, + { + "epoch": 0.5641025641025641, + "grad_norm": 0.484998881816864, + "learning_rate": 0.0001904478695896267, + "loss": 1.0711, + "step": 3168 + }, + { + "epoch": 0.5642806267806267, + "grad_norm": 0.5515334010124207, + "learning_rate": 0.0001904418985185132, + "loss": 1.1583, + "step": 3169 + }, + { + "epoch": 0.5644586894586895, + "grad_norm": 0.545460045337677, + "learning_rate": 0.00019043592567537853, + "loss": 1.2321, + "step": 3170 + }, + { + "epoch": 0.5646367521367521, + "grad_norm": 0.5463964343070984, + "learning_rate": 0.0001904299510603398, + "loss": 1.1019, + "step": 3171 + }, + { + "epoch": 0.5648148148148148, + "grad_norm": 0.5619220733642578, + "learning_rate": 0.000190423974673514, + "loss": 1.1001, + "step": 3172 + }, + { + "epoch": 0.5649928774928775, + "grad_norm": 0.4448916018009186, + "learning_rate": 0.00019041799651501825, + "loss": 1.057, + "step": 3173 + }, + { + "epoch": 0.5651709401709402, + "grad_norm": 0.6073006987571716, + "learning_rate": 0.00019041201658496975, + "loss": 1.0306, + "step": 3174 + }, + { + "epoch": 0.5653490028490028, + "grad_norm": 0.5342072248458862, + "learning_rate": 0.0001904060348834855, + "loss": 0.9231, + "step": 3175 + }, + { + "epoch": 0.5655270655270656, + "grad_norm": 0.4505697786808014, + "learning_rate": 0.0001904000514106829, + "loss": 1.1134, + "step": 3176 + }, + { + "epoch": 0.5657051282051282, + "grad_norm": 0.5627852082252502, + "learning_rate": 0.00019039406616667902, + "loss": 1.2138, + "step": 3177 + }, + { + "epoch": 0.5658831908831908, + "grad_norm": 0.499734103679657, + "learning_rate": 0.0001903880791515912, + "loss": 1.1074, + "step": 3178 + }, + { + "epoch": 0.5660612535612536, + "grad_norm": 0.4768189489841461, + "learning_rate": 0.00019038209036553676, + "loss": 0.9442, + "step": 3179 + }, + { + "epoch": 0.5662393162393162, + "grad_norm": 0.5265373587608337, + "learning_rate": 0.00019037609980863298, + "loss": 1.0907, + "step": 3180 + }, + { + "epoch": 0.5664173789173789, + "grad_norm": 0.5506128072738647, + "learning_rate": 0.00019037010748099728, + "loss": 1.2541, + "step": 3181 + }, + { + "epoch": 0.5665954415954416, + "grad_norm": 0.44860872626304626, + "learning_rate": 0.00019036411338274703, + "loss": 0.893, + "step": 3182 + }, + { + "epoch": 0.5667735042735043, + "grad_norm": 0.4901522994041443, + "learning_rate": 0.00019035811751399973, + "loss": 1.0469, + "step": 3183 + }, + { + "epoch": 0.5669515669515669, + "grad_norm": 0.500868022441864, + "learning_rate": 0.0001903521198748728, + "loss": 1.0527, + "step": 3184 + }, + { + "epoch": 0.5671296296296297, + "grad_norm": 0.5508102774620056, + "learning_rate": 0.00019034612046548376, + "loss": 1.283, + "step": 3185 + }, + { + "epoch": 0.5673076923076923, + "grad_norm": 0.5079495906829834, + "learning_rate": 0.0001903401192859502, + "loss": 1.0808, + "step": 3186 + }, + { + "epoch": 0.5674857549857549, + "grad_norm": 0.5758788585662842, + "learning_rate": 0.00019033411633638964, + "loss": 1.1301, + "step": 3187 + }, + { + "epoch": 0.5676638176638177, + "grad_norm": 0.46557924151420593, + "learning_rate": 0.00019032811161691972, + "loss": 1.0205, + "step": 3188 + }, + { + "epoch": 0.5678418803418803, + "grad_norm": 0.5665056109428406, + "learning_rate": 0.0001903221051276581, + "loss": 1.1926, + "step": 3189 + }, + { + "epoch": 0.5680199430199431, + "grad_norm": 0.5948992967605591, + "learning_rate": 0.00019031609686872246, + "loss": 1.2724, + "step": 3190 + }, + { + "epoch": 0.5681980056980057, + "grad_norm": 0.6189367771148682, + "learning_rate": 0.00019031008684023055, + "loss": 1.2762, + "step": 3191 + }, + { + "epoch": 0.5683760683760684, + "grad_norm": 0.49511992931365967, + "learning_rate": 0.00019030407504230006, + "loss": 1.0117, + "step": 3192 + }, + { + "epoch": 0.5685541310541311, + "grad_norm": 0.5358837842941284, + "learning_rate": 0.00019029806147504878, + "loss": 0.944, + "step": 3193 + }, + { + "epoch": 0.5687321937321937, + "grad_norm": 0.458636999130249, + "learning_rate": 0.00019029204613859463, + "loss": 0.8174, + "step": 3194 + }, + { + "epoch": 0.5689102564102564, + "grad_norm": 0.5168304443359375, + "learning_rate": 0.00019028602903305535, + "loss": 1.1533, + "step": 3195 + }, + { + "epoch": 0.5690883190883191, + "grad_norm": 0.5334134697914124, + "learning_rate": 0.00019028001015854892, + "loss": 1.1868, + "step": 3196 + }, + { + "epoch": 0.5692663817663818, + "grad_norm": 0.5649123191833496, + "learning_rate": 0.0001902739895151932, + "loss": 0.9876, + "step": 3197 + }, + { + "epoch": 0.5694444444444444, + "grad_norm": 0.5647651553153992, + "learning_rate": 0.0001902679671031062, + "loss": 1.0805, + "step": 3198 + }, + { + "epoch": 0.5696225071225072, + "grad_norm": 0.5251876711845398, + "learning_rate": 0.00019026194292240587, + "loss": 1.2335, + "step": 3199 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 0.5268014669418335, + "learning_rate": 0.0001902559169732103, + "loss": 1.19, + "step": 3200 + }, + { + "epoch": 0.5699786324786325, + "grad_norm": 0.5301041007041931, + "learning_rate": 0.00019024988925563752, + "loss": 1.1173, + "step": 3201 + }, + { + "epoch": 0.5701566951566952, + "grad_norm": 0.4531562030315399, + "learning_rate": 0.00019024385976980566, + "loss": 0.7576, + "step": 3202 + }, + { + "epoch": 0.5703347578347578, + "grad_norm": 0.5779716372489929, + "learning_rate": 0.00019023782851583282, + "loss": 1.1719, + "step": 3203 + }, + { + "epoch": 0.5705128205128205, + "grad_norm": 0.4886093735694885, + "learning_rate": 0.00019023179549383716, + "loss": 1.085, + "step": 3204 + }, + { + "epoch": 0.5706908831908832, + "grad_norm": 0.510117769241333, + "learning_rate": 0.0001902257607039369, + "loss": 0.8931, + "step": 3205 + }, + { + "epoch": 0.5708689458689459, + "grad_norm": 0.5195479393005371, + "learning_rate": 0.00019021972414625036, + "loss": 0.9922, + "step": 3206 + }, + { + "epoch": 0.5710470085470085, + "grad_norm": 0.5791407227516174, + "learning_rate": 0.00019021368582089568, + "loss": 1.112, + "step": 3207 + }, + { + "epoch": 0.5712250712250713, + "grad_norm": 0.5056005716323853, + "learning_rate": 0.00019020764572799122, + "loss": 0.8474, + "step": 3208 + }, + { + "epoch": 0.5714031339031339, + "grad_norm": 0.5060068964958191, + "learning_rate": 0.00019020160386765537, + "loss": 1.071, + "step": 3209 + }, + { + "epoch": 0.5715811965811965, + "grad_norm": 0.5396568775177002, + "learning_rate": 0.00019019556024000648, + "loss": 1.0436, + "step": 3210 + }, + { + "epoch": 0.5717592592592593, + "grad_norm": 0.6552190780639648, + "learning_rate": 0.0001901895148451629, + "loss": 0.9869, + "step": 3211 + }, + { + "epoch": 0.5719373219373219, + "grad_norm": 0.5177004337310791, + "learning_rate": 0.00019018346768324314, + "loss": 1.0193, + "step": 3212 + }, + { + "epoch": 0.5721153846153846, + "grad_norm": 0.5192117094993591, + "learning_rate": 0.0001901774187543657, + "loss": 1.1263, + "step": 3213 + }, + { + "epoch": 0.5722934472934473, + "grad_norm": 0.4857729971408844, + "learning_rate": 0.00019017136805864906, + "loss": 0.9808, + "step": 3214 + }, + { + "epoch": 0.57247150997151, + "grad_norm": 0.5800918936729431, + "learning_rate": 0.00019016531559621177, + "loss": 1.2334, + "step": 3215 + }, + { + "epoch": 0.5726495726495726, + "grad_norm": 0.4812086522579193, + "learning_rate": 0.00019015926136717242, + "loss": 1.2409, + "step": 3216 + }, + { + "epoch": 0.5728276353276354, + "grad_norm": 0.5128398537635803, + "learning_rate": 0.00019015320537164963, + "loss": 0.9036, + "step": 3217 + }, + { + "epoch": 0.573005698005698, + "grad_norm": 0.4761141538619995, + "learning_rate": 0.00019014714760976205, + "loss": 1.1058, + "step": 3218 + }, + { + "epoch": 0.5731837606837606, + "grad_norm": 0.5850459933280945, + "learning_rate": 0.0001901410880816284, + "loss": 1.1011, + "step": 3219 + }, + { + "epoch": 0.5733618233618234, + "grad_norm": 0.5648714303970337, + "learning_rate": 0.00019013502678736738, + "loss": 1.0479, + "step": 3220 + }, + { + "epoch": 0.573539886039886, + "grad_norm": 0.5835902094841003, + "learning_rate": 0.00019012896372709774, + "loss": 1.0555, + "step": 3221 + }, + { + "epoch": 0.5737179487179487, + "grad_norm": 0.5155113935470581, + "learning_rate": 0.00019012289890093828, + "loss": 0.9488, + "step": 3222 + }, + { + "epoch": 0.5738960113960114, + "grad_norm": 0.5064889788627625, + "learning_rate": 0.00019011683230900784, + "loss": 0.9144, + "step": 3223 + }, + { + "epoch": 0.5740740740740741, + "grad_norm": 0.53825843334198, + "learning_rate": 0.00019011076395142527, + "loss": 1.0713, + "step": 3224 + }, + { + "epoch": 0.5742521367521367, + "grad_norm": 0.5341386198997498, + "learning_rate": 0.00019010469382830947, + "loss": 1.1438, + "step": 3225 + }, + { + "epoch": 0.5744301994301995, + "grad_norm": 0.5300050973892212, + "learning_rate": 0.00019009862193977936, + "loss": 1.0114, + "step": 3226 + }, + { + "epoch": 0.5746082621082621, + "grad_norm": 0.6033682823181152, + "learning_rate": 0.0001900925482859539, + "loss": 1.0458, + "step": 3227 + }, + { + "epoch": 0.5747863247863247, + "grad_norm": 0.5108983516693115, + "learning_rate": 0.00019008647286695215, + "loss": 1.1211, + "step": 3228 + }, + { + "epoch": 0.5749643874643875, + "grad_norm": 0.5263782739639282, + "learning_rate": 0.00019008039568289308, + "loss": 0.8647, + "step": 3229 + }, + { + "epoch": 0.5751424501424501, + "grad_norm": 0.47119566798210144, + "learning_rate": 0.0001900743167338958, + "loss": 1.019, + "step": 3230 + }, + { + "epoch": 0.5753205128205128, + "grad_norm": 0.56391841173172, + "learning_rate": 0.00019006823602007937, + "loss": 0.9791, + "step": 3231 + }, + { + "epoch": 0.5754985754985755, + "grad_norm": 0.5364985466003418, + "learning_rate": 0.000190062153541563, + "loss": 1.1355, + "step": 3232 + }, + { + "epoch": 0.5756766381766382, + "grad_norm": 0.5098565220832825, + "learning_rate": 0.00019005606929846578, + "loss": 0.987, + "step": 3233 + }, + { + "epoch": 0.5758547008547008, + "grad_norm": 0.6640968918800354, + "learning_rate": 0.00019004998329090692, + "loss": 1.1165, + "step": 3234 + }, + { + "epoch": 0.5760327635327636, + "grad_norm": 0.5044721961021423, + "learning_rate": 0.00019004389551900578, + "loss": 0.8643, + "step": 3235 + }, + { + "epoch": 0.5762108262108262, + "grad_norm": 0.4822785258293152, + "learning_rate": 0.00019003780598288153, + "loss": 1.0735, + "step": 3236 + }, + { + "epoch": 0.5763888888888888, + "grad_norm": 0.505261242389679, + "learning_rate": 0.00019003171468265348, + "loss": 1.0001, + "step": 3237 + }, + { + "epoch": 0.5765669515669516, + "grad_norm": 0.5020412802696228, + "learning_rate": 0.00019002562161844102, + "loss": 0.9601, + "step": 3238 + }, + { + "epoch": 0.5767450142450142, + "grad_norm": 0.4920475482940674, + "learning_rate": 0.00019001952679036354, + "loss": 1.0111, + "step": 3239 + }, + { + "epoch": 0.5769230769230769, + "grad_norm": 0.5638813376426697, + "learning_rate": 0.00019001343019854042, + "loss": 1.1456, + "step": 3240 + }, + { + "epoch": 0.5771011396011396, + "grad_norm": 0.5519235134124756, + "learning_rate": 0.0001900073318430911, + "loss": 0.9258, + "step": 3241 + }, + { + "epoch": 0.5772792022792023, + "grad_norm": 0.5207770466804504, + "learning_rate": 0.0001900012317241351, + "loss": 0.9859, + "step": 3242 + }, + { + "epoch": 0.5774572649572649, + "grad_norm": 0.5493707656860352, + "learning_rate": 0.00018999512984179195, + "loss": 1.1183, + "step": 3243 + }, + { + "epoch": 0.5776353276353277, + "grad_norm": 0.4504764676094055, + "learning_rate": 0.00018998902619618116, + "loss": 0.9363, + "step": 3244 + }, + { + "epoch": 0.5778133903133903, + "grad_norm": 0.5232836604118347, + "learning_rate": 0.00018998292078742233, + "loss": 1.1887, + "step": 3245 + }, + { + "epoch": 0.5779914529914529, + "grad_norm": 0.5715088248252869, + "learning_rate": 0.0001899768136156351, + "loss": 1.4524, + "step": 3246 + }, + { + "epoch": 0.5781695156695157, + "grad_norm": 0.59555584192276, + "learning_rate": 0.0001899707046809391, + "loss": 1.0922, + "step": 3247 + }, + { + "epoch": 0.5783475783475783, + "grad_norm": 0.4500894546508789, + "learning_rate": 0.00018996459398345404, + "loss": 1.0087, + "step": 3248 + }, + { + "epoch": 0.5785256410256411, + "grad_norm": 0.49126625061035156, + "learning_rate": 0.00018995848152329967, + "loss": 1.1512, + "step": 3249 + }, + { + "epoch": 0.5787037037037037, + "grad_norm": 0.4096335172653198, + "learning_rate": 0.00018995236730059574, + "loss": 0.7633, + "step": 3250 + }, + { + "epoch": 0.5788817663817664, + "grad_norm": 0.5364313721656799, + "learning_rate": 0.00018994625131546199, + "loss": 1.295, + "step": 3251 + }, + { + "epoch": 0.5790598290598291, + "grad_norm": 0.4897502660751343, + "learning_rate": 0.00018994013356801834, + "loss": 1.2197, + "step": 3252 + }, + { + "epoch": 0.5792378917378918, + "grad_norm": 0.5101368427276611, + "learning_rate": 0.00018993401405838456, + "loss": 1.1129, + "step": 3253 + }, + { + "epoch": 0.5794159544159544, + "grad_norm": 0.5426377654075623, + "learning_rate": 0.00018992789278668063, + "loss": 1.188, + "step": 3254 + }, + { + "epoch": 0.5795940170940171, + "grad_norm": 0.5066362023353577, + "learning_rate": 0.00018992176975302644, + "loss": 1.2802, + "step": 3255 + }, + { + "epoch": 0.5797720797720798, + "grad_norm": 0.5418947339057922, + "learning_rate": 0.00018991564495754196, + "loss": 1.1675, + "step": 3256 + }, + { + "epoch": 0.5799501424501424, + "grad_norm": 0.5139963626861572, + "learning_rate": 0.0001899095184003472, + "loss": 0.9717, + "step": 3257 + }, + { + "epoch": 0.5801282051282052, + "grad_norm": 0.5167285799980164, + "learning_rate": 0.00018990339008156219, + "loss": 1.1529, + "step": 3258 + }, + { + "epoch": 0.5803062678062678, + "grad_norm": 0.53471440076828, + "learning_rate": 0.00018989726000130704, + "loss": 1.0711, + "step": 3259 + }, + { + "epoch": 0.5804843304843305, + "grad_norm": 0.49875229597091675, + "learning_rate": 0.0001898911281597018, + "loss": 1.1095, + "step": 3260 + }, + { + "epoch": 0.5806623931623932, + "grad_norm": 0.4473155438899994, + "learning_rate": 0.00018988499455686663, + "loss": 0.836, + "step": 3261 + }, + { + "epoch": 0.5808404558404558, + "grad_norm": 0.6181996464729309, + "learning_rate": 0.00018987885919292174, + "loss": 1.2787, + "step": 3262 + }, + { + "epoch": 0.5810185185185185, + "grad_norm": 0.4996899664402008, + "learning_rate": 0.00018987272206798733, + "loss": 1.2132, + "step": 3263 + }, + { + "epoch": 0.5811965811965812, + "grad_norm": 0.49979713559150696, + "learning_rate": 0.00018986658318218358, + "loss": 0.8388, + "step": 3264 + }, + { + "epoch": 0.5813746438746439, + "grad_norm": 0.5288876295089722, + "learning_rate": 0.00018986044253563084, + "loss": 1.1871, + "step": 3265 + }, + { + "epoch": 0.5815527065527065, + "grad_norm": 0.534063458442688, + "learning_rate": 0.00018985430012844937, + "loss": 0.96, + "step": 3266 + }, + { + "epoch": 0.5817307692307693, + "grad_norm": 0.5081285834312439, + "learning_rate": 0.00018984815596075953, + "loss": 1.1577, + "step": 3267 + }, + { + "epoch": 0.5819088319088319, + "grad_norm": 0.5648202896118164, + "learning_rate": 0.00018984201003268176, + "loss": 1.2235, + "step": 3268 + }, + { + "epoch": 0.5820868945868946, + "grad_norm": 0.495061993598938, + "learning_rate": 0.00018983586234433642, + "loss": 1.056, + "step": 3269 + }, + { + "epoch": 0.5822649572649573, + "grad_norm": 0.47149857878685, + "learning_rate": 0.000189829712895844, + "loss": 1.0844, + "step": 3270 + }, + { + "epoch": 0.58244301994302, + "grad_norm": 0.6107062697410583, + "learning_rate": 0.00018982356168732492, + "loss": 0.9868, + "step": 3271 + }, + { + "epoch": 0.5826210826210826, + "grad_norm": 0.7355940341949463, + "learning_rate": 0.00018981740871889974, + "loss": 1.1448, + "step": 3272 + }, + { + "epoch": 0.5827991452991453, + "grad_norm": 0.5950441956520081, + "learning_rate": 0.00018981125399068907, + "loss": 0.9618, + "step": 3273 + }, + { + "epoch": 0.582977207977208, + "grad_norm": 0.47607290744781494, + "learning_rate": 0.0001898050975028134, + "loss": 0.957, + "step": 3274 + }, + { + "epoch": 0.5831552706552706, + "grad_norm": 0.541164755821228, + "learning_rate": 0.00018979893925539338, + "loss": 1.1426, + "step": 3275 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 0.5240640044212341, + "learning_rate": 0.00018979277924854974, + "loss": 1.1421, + "step": 3276 + }, + { + "epoch": 0.583511396011396, + "grad_norm": 0.48155727982521057, + "learning_rate": 0.00018978661748240307, + "loss": 1.0069, + "step": 3277 + }, + { + "epoch": 0.5836894586894587, + "grad_norm": 0.5559938549995422, + "learning_rate": 0.00018978045395707418, + "loss": 1.1227, + "step": 3278 + }, + { + "epoch": 0.5838675213675214, + "grad_norm": 0.5244291424751282, + "learning_rate": 0.0001897742886726838, + "loss": 1.1103, + "step": 3279 + }, + { + "epoch": 0.584045584045584, + "grad_norm": 0.5277758240699768, + "learning_rate": 0.00018976812162935268, + "loss": 1.2125, + "step": 3280 + }, + { + "epoch": 0.5842236467236467, + "grad_norm": 0.5415039658546448, + "learning_rate": 0.00018976195282720173, + "loss": 1.146, + "step": 3281 + }, + { + "epoch": 0.5844017094017094, + "grad_norm": 0.5152051448822021, + "learning_rate": 0.00018975578226635177, + "loss": 1.0092, + "step": 3282 + }, + { + "epoch": 0.5845797720797721, + "grad_norm": 0.5489452481269836, + "learning_rate": 0.00018974960994692371, + "loss": 1.2425, + "step": 3283 + }, + { + "epoch": 0.5847578347578347, + "grad_norm": 0.491274356842041, + "learning_rate": 0.00018974343586903848, + "loss": 0.9559, + "step": 3284 + }, + { + "epoch": 0.5849358974358975, + "grad_norm": 0.5783739686012268, + "learning_rate": 0.00018973726003281707, + "loss": 1.1971, + "step": 3285 + }, + { + "epoch": 0.5851139601139601, + "grad_norm": 0.5056472420692444, + "learning_rate": 0.00018973108243838045, + "loss": 1.0313, + "step": 3286 + }, + { + "epoch": 0.5852920227920227, + "grad_norm": 0.4939729571342468, + "learning_rate": 0.00018972490308584962, + "loss": 1.1061, + "step": 3287 + }, + { + "epoch": 0.5854700854700855, + "grad_norm": 0.4889580011367798, + "learning_rate": 0.00018971872197534576, + "loss": 0.9157, + "step": 3288 + }, + { + "epoch": 0.5856481481481481, + "grad_norm": 0.40889349579811096, + "learning_rate": 0.00018971253910698993, + "loss": 0.8083, + "step": 3289 + }, + { + "epoch": 0.5858262108262108, + "grad_norm": 0.5221503973007202, + "learning_rate": 0.00018970635448090322, + "loss": 0.9995, + "step": 3290 + }, + { + "epoch": 0.5860042735042735, + "grad_norm": 0.47060561180114746, + "learning_rate": 0.00018970016809720687, + "loss": 0.9738, + "step": 3291 + }, + { + "epoch": 0.5861823361823362, + "grad_norm": 0.6083170771598816, + "learning_rate": 0.000189693979956022, + "loss": 1.188, + "step": 3292 + }, + { + "epoch": 0.5863603988603988, + "grad_norm": 0.4696751534938812, + "learning_rate": 0.00018968779005746998, + "loss": 1.089, + "step": 3293 + }, + { + "epoch": 0.5865384615384616, + "grad_norm": 0.5081014633178711, + "learning_rate": 0.00018968159840167202, + "loss": 1.1869, + "step": 3294 + }, + { + "epoch": 0.5867165242165242, + "grad_norm": 0.48042431473731995, + "learning_rate": 0.0001896754049887494, + "loss": 0.964, + "step": 3295 + }, + { + "epoch": 0.5868945868945868, + "grad_norm": 0.5075193643569946, + "learning_rate": 0.00018966920981882353, + "loss": 1.1884, + "step": 3296 + }, + { + "epoch": 0.5870726495726496, + "grad_norm": 0.5734842419624329, + "learning_rate": 0.00018966301289201576, + "loss": 1.1475, + "step": 3297 + }, + { + "epoch": 0.5872507122507122, + "grad_norm": 0.5525311231613159, + "learning_rate": 0.00018965681420844753, + "loss": 1.241, + "step": 3298 + }, + { + "epoch": 0.5874287749287749, + "grad_norm": 0.48142680525779724, + "learning_rate": 0.00018965061376824025, + "loss": 1.0871, + "step": 3299 + }, + { + "epoch": 0.5876068376068376, + "grad_norm": 0.5360350608825684, + "learning_rate": 0.00018964441157151544, + "loss": 1.1895, + "step": 3300 + }, + { + "epoch": 0.5877849002849003, + "grad_norm": 0.5207685232162476, + "learning_rate": 0.00018963820761839457, + "loss": 0.9323, + "step": 3301 + }, + { + "epoch": 0.5879629629629629, + "grad_norm": 0.453620970249176, + "learning_rate": 0.00018963200190899926, + "loss": 0.802, + "step": 3302 + }, + { + "epoch": 0.5881410256410257, + "grad_norm": 0.5198796391487122, + "learning_rate": 0.00018962579444345106, + "loss": 1.0243, + "step": 3303 + }, + { + "epoch": 0.5883190883190883, + "grad_norm": 0.5597525835037231, + "learning_rate": 0.0001896195852218716, + "loss": 0.9351, + "step": 3304 + }, + { + "epoch": 0.5884971509971509, + "grad_norm": 0.5738299489021301, + "learning_rate": 0.00018961337424438254, + "loss": 1.3737, + "step": 3305 + }, + { + "epoch": 0.5886752136752137, + "grad_norm": 0.5569949150085449, + "learning_rate": 0.00018960716151110554, + "loss": 1.0469, + "step": 3306 + }, + { + "epoch": 0.5888532763532763, + "grad_norm": 0.5088010430335999, + "learning_rate": 0.00018960094702216238, + "loss": 1.0982, + "step": 3307 + }, + { + "epoch": 0.5890313390313391, + "grad_norm": 0.5127636790275574, + "learning_rate": 0.0001895947307776748, + "loss": 0.9986, + "step": 3308 + }, + { + "epoch": 0.5892094017094017, + "grad_norm": 0.5160682797431946, + "learning_rate": 0.00018958851277776456, + "loss": 1.0219, + "step": 3309 + }, + { + "epoch": 0.5893874643874644, + "grad_norm": 0.5380711555480957, + "learning_rate": 0.00018958229302255356, + "loss": 1.118, + "step": 3310 + }, + { + "epoch": 0.5895655270655271, + "grad_norm": 0.5571228861808777, + "learning_rate": 0.0001895760715121636, + "loss": 1.0302, + "step": 3311 + }, + { + "epoch": 0.5897435897435898, + "grad_norm": 0.542266309261322, + "learning_rate": 0.00018956984824671657, + "loss": 1.0372, + "step": 3312 + }, + { + "epoch": 0.5899216524216524, + "grad_norm": 0.48350459337234497, + "learning_rate": 0.00018956362322633446, + "loss": 1.2, + "step": 3313 + }, + { + "epoch": 0.5900997150997151, + "grad_norm": 0.5001645088195801, + "learning_rate": 0.0001895573964511392, + "loss": 0.9749, + "step": 3314 + }, + { + "epoch": 0.5902777777777778, + "grad_norm": 0.5227531790733337, + "learning_rate": 0.00018955116792125276, + "loss": 1.025, + "step": 3315 + }, + { + "epoch": 0.5904558404558404, + "grad_norm": 0.522251546382904, + "learning_rate": 0.00018954493763679727, + "loss": 1.0821, + "step": 3316 + }, + { + "epoch": 0.5906339031339032, + "grad_norm": 0.5423251390457153, + "learning_rate": 0.00018953870559789467, + "loss": 1.0961, + "step": 3317 + }, + { + "epoch": 0.5908119658119658, + "grad_norm": 0.5615720748901367, + "learning_rate": 0.0001895324718046672, + "loss": 1.1209, + "step": 3318 + }, + { + "epoch": 0.5909900284900285, + "grad_norm": 0.44746771454811096, + "learning_rate": 0.00018952623625723692, + "loss": 0.9935, + "step": 3319 + }, + { + "epoch": 0.5911680911680912, + "grad_norm": 0.5993229150772095, + "learning_rate": 0.00018951999895572597, + "loss": 1.1409, + "step": 3320 + }, + { + "epoch": 0.5913461538461539, + "grad_norm": 0.4969801902770996, + "learning_rate": 0.00018951375990025666, + "loss": 1.1568, + "step": 3321 + }, + { + "epoch": 0.5915242165242165, + "grad_norm": 0.6001267433166504, + "learning_rate": 0.00018950751909095116, + "loss": 1.1135, + "step": 3322 + }, + { + "epoch": 0.5917022792022792, + "grad_norm": 0.5386021733283997, + "learning_rate": 0.00018950127652793172, + "loss": 0.947, + "step": 3323 + }, + { + "epoch": 0.5918803418803419, + "grad_norm": 0.49043843150138855, + "learning_rate": 0.00018949503221132074, + "loss": 0.9581, + "step": 3324 + }, + { + "epoch": 0.5920584045584045, + "grad_norm": 0.5241141319274902, + "learning_rate": 0.00018948878614124048, + "loss": 1.0797, + "step": 3325 + }, + { + "epoch": 0.5922364672364673, + "grad_norm": 0.5755026340484619, + "learning_rate": 0.00018948253831781338, + "loss": 1.1046, + "step": 3326 + }, + { + "epoch": 0.5924145299145299, + "grad_norm": 0.5004449486732483, + "learning_rate": 0.00018947628874116179, + "loss": 1.1416, + "step": 3327 + }, + { + "epoch": 0.5925925925925926, + "grad_norm": 0.53347247838974, + "learning_rate": 0.00018947003741140821, + "loss": 1.2718, + "step": 3328 + }, + { + "epoch": 0.5927706552706553, + "grad_norm": 0.6473469138145447, + "learning_rate": 0.0001894637843286751, + "loss": 1.2255, + "step": 3329 + }, + { + "epoch": 0.592948717948718, + "grad_norm": 0.4750518798828125, + "learning_rate": 0.00018945752949308498, + "loss": 1.0537, + "step": 3330 + }, + { + "epoch": 0.5931267806267806, + "grad_norm": 0.5636306405067444, + "learning_rate": 0.00018945127290476043, + "loss": 0.9906, + "step": 3331 + }, + { + "epoch": 0.5933048433048433, + "grad_norm": 0.4871736466884613, + "learning_rate": 0.00018944501456382397, + "loss": 1.0549, + "step": 3332 + }, + { + "epoch": 0.593482905982906, + "grad_norm": 0.5554637312889099, + "learning_rate": 0.0001894387544703983, + "loss": 1.1587, + "step": 3333 + }, + { + "epoch": 0.5936609686609686, + "grad_norm": 0.5385799407958984, + "learning_rate": 0.000189432492624606, + "loss": 0.9565, + "step": 3334 + }, + { + "epoch": 0.5938390313390314, + "grad_norm": 0.4996553063392639, + "learning_rate": 0.00018942622902656976, + "loss": 1.0456, + "step": 3335 + }, + { + "epoch": 0.594017094017094, + "grad_norm": 0.46810707449913025, + "learning_rate": 0.00018941996367641237, + "loss": 1.119, + "step": 3336 + }, + { + "epoch": 0.5941951566951567, + "grad_norm": 0.5672653913497925, + "learning_rate": 0.0001894136965742565, + "loss": 1.1317, + "step": 3337 + }, + { + "epoch": 0.5943732193732194, + "grad_norm": 0.4790053367614746, + "learning_rate": 0.00018940742772022504, + "loss": 1.0967, + "step": 3338 + }, + { + "epoch": 0.594551282051282, + "grad_norm": 0.5935906171798706, + "learning_rate": 0.00018940115711444072, + "loss": 1.3044, + "step": 3339 + }, + { + "epoch": 0.5947293447293447, + "grad_norm": 0.4790516793727875, + "learning_rate": 0.00018939488475702647, + "loss": 1.074, + "step": 3340 + }, + { + "epoch": 0.5949074074074074, + "grad_norm": 0.474588006734848, + "learning_rate": 0.00018938861064810516, + "loss": 1.1476, + "step": 3341 + }, + { + "epoch": 0.5950854700854701, + "grad_norm": 0.4908665120601654, + "learning_rate": 0.0001893823347877997, + "loss": 1.216, + "step": 3342 + }, + { + "epoch": 0.5952635327635327, + "grad_norm": 0.531650960445404, + "learning_rate": 0.00018937605717623307, + "loss": 1.1057, + "step": 3343 + }, + { + "epoch": 0.5954415954415955, + "grad_norm": 0.5581082105636597, + "learning_rate": 0.00018936977781352823, + "loss": 0.7972, + "step": 3344 + }, + { + "epoch": 0.5956196581196581, + "grad_norm": 0.42370662093162537, + "learning_rate": 0.00018936349669980827, + "loss": 0.8888, + "step": 3345 + }, + { + "epoch": 0.5957977207977208, + "grad_norm": 0.5817318558692932, + "learning_rate": 0.00018935721383519624, + "loss": 1.2801, + "step": 3346 + }, + { + "epoch": 0.5959757834757835, + "grad_norm": 0.4766376316547394, + "learning_rate": 0.00018935092921981524, + "loss": 1.0918, + "step": 3347 + }, + { + "epoch": 0.5961538461538461, + "grad_norm": 0.5567346811294556, + "learning_rate": 0.00018934464285378836, + "loss": 1.0269, + "step": 3348 + }, + { + "epoch": 0.5963319088319088, + "grad_norm": 0.5285565257072449, + "learning_rate": 0.0001893383547372388, + "loss": 1.1887, + "step": 3349 + }, + { + "epoch": 0.5965099715099715, + "grad_norm": 0.49052694439888, + "learning_rate": 0.00018933206487028979, + "loss": 1.0773, + "step": 3350 + }, + { + "epoch": 0.5966880341880342, + "grad_norm": 0.6175199151039124, + "learning_rate": 0.0001893257732530645, + "loss": 1.0192, + "step": 3351 + }, + { + "epoch": 0.5968660968660968, + "grad_norm": 0.56049644947052, + "learning_rate": 0.00018931947988568628, + "loss": 0.9516, + "step": 3352 + }, + { + "epoch": 0.5970441595441596, + "grad_norm": 0.47873660922050476, + "learning_rate": 0.00018931318476827838, + "loss": 0.8174, + "step": 3353 + }, + { + "epoch": 0.5972222222222222, + "grad_norm": 0.4748854339122772, + "learning_rate": 0.00018930688790096416, + "loss": 1.0238, + "step": 3354 + }, + { + "epoch": 0.5974002849002849, + "grad_norm": 0.5382232666015625, + "learning_rate": 0.00018930058928386698, + "loss": 1.0815, + "step": 3355 + }, + { + "epoch": 0.5975783475783476, + "grad_norm": 0.5038299560546875, + "learning_rate": 0.00018929428891711027, + "loss": 1.0472, + "step": 3356 + }, + { + "epoch": 0.5977564102564102, + "grad_norm": 0.5185908079147339, + "learning_rate": 0.00018928798680081744, + "loss": 1.0435, + "step": 3357 + }, + { + "epoch": 0.5979344729344729, + "grad_norm": 0.5169877409934998, + "learning_rate": 0.00018928168293511202, + "loss": 1.0437, + "step": 3358 + }, + { + "epoch": 0.5981125356125356, + "grad_norm": 0.5218369960784912, + "learning_rate": 0.00018927537732011749, + "loss": 1.082, + "step": 3359 + }, + { + "epoch": 0.5982905982905983, + "grad_norm": 0.5358219742774963, + "learning_rate": 0.0001892690699559574, + "loss": 1.2523, + "step": 3360 + }, + { + "epoch": 0.5984686609686609, + "grad_norm": 0.47716647386550903, + "learning_rate": 0.0001892627608427553, + "loss": 1.2069, + "step": 3361 + }, + { + "epoch": 0.5986467236467237, + "grad_norm": 0.5484169125556946, + "learning_rate": 0.00018925644998063482, + "loss": 1.2016, + "step": 3362 + }, + { + "epoch": 0.5988247863247863, + "grad_norm": 0.46814846992492676, + "learning_rate": 0.00018925013736971965, + "loss": 0.7989, + "step": 3363 + }, + { + "epoch": 0.5990028490028491, + "grad_norm": 0.5391258001327515, + "learning_rate": 0.0001892438230101334, + "loss": 1.224, + "step": 3364 + }, + { + "epoch": 0.5991809116809117, + "grad_norm": 0.5248384475708008, + "learning_rate": 0.00018923750690199987, + "loss": 1.1532, + "step": 3365 + }, + { + "epoch": 0.5993589743589743, + "grad_norm": 0.5074637532234192, + "learning_rate": 0.00018923118904544273, + "loss": 1.0968, + "step": 3366 + }, + { + "epoch": 0.5995370370370371, + "grad_norm": 0.5260029435157776, + "learning_rate": 0.00018922486944058581, + "loss": 1.1311, + "step": 3367 + }, + { + "epoch": 0.5997150997150997, + "grad_norm": 0.48497965931892395, + "learning_rate": 0.00018921854808755294, + "loss": 1.1208, + "step": 3368 + }, + { + "epoch": 0.5998931623931624, + "grad_norm": 0.5108651518821716, + "learning_rate": 0.00018921222498646792, + "loss": 1.147, + "step": 3369 + }, + { + "epoch": 0.6000712250712251, + "grad_norm": 0.5243437886238098, + "learning_rate": 0.00018920590013745471, + "loss": 0.9614, + "step": 3370 + }, + { + "epoch": 0.6002492877492878, + "grad_norm": 0.47022634744644165, + "learning_rate": 0.00018919957354063719, + "loss": 1.0579, + "step": 3371 + }, + { + "epoch": 0.6004273504273504, + "grad_norm": 0.6461413502693176, + "learning_rate": 0.00018919324519613931, + "loss": 1.2126, + "step": 3372 + }, + { + "epoch": 0.6006054131054132, + "grad_norm": 0.4654616713523865, + "learning_rate": 0.00018918691510408508, + "loss": 1.1476, + "step": 3373 + }, + { + "epoch": 0.6007834757834758, + "grad_norm": 0.48571303486824036, + "learning_rate": 0.00018918058326459854, + "loss": 1.2093, + "step": 3374 + }, + { + "epoch": 0.6009615384615384, + "grad_norm": 0.5255016684532166, + "learning_rate": 0.00018917424967780368, + "loss": 1.1538, + "step": 3375 + }, + { + "epoch": 0.6011396011396012, + "grad_norm": 0.5059894323348999, + "learning_rate": 0.00018916791434382468, + "loss": 1.0556, + "step": 3376 + }, + { + "epoch": 0.6013176638176638, + "grad_norm": 0.4581229090690613, + "learning_rate": 0.00018916157726278561, + "loss": 1.1468, + "step": 3377 + }, + { + "epoch": 0.6014957264957265, + "grad_norm": 0.5701818466186523, + "learning_rate": 0.00018915523843481067, + "loss": 1.3641, + "step": 3378 + }, + { + "epoch": 0.6016737891737892, + "grad_norm": 0.5007243752479553, + "learning_rate": 0.00018914889786002403, + "loss": 1.2705, + "step": 3379 + }, + { + "epoch": 0.6018518518518519, + "grad_norm": 0.5192995071411133, + "learning_rate": 0.0001891425555385499, + "loss": 0.9922, + "step": 3380 + }, + { + "epoch": 0.6020299145299145, + "grad_norm": 0.5880612134933472, + "learning_rate": 0.00018913621147051258, + "loss": 0.8783, + "step": 3381 + }, + { + "epoch": 0.6022079772079773, + "grad_norm": 0.5161563158035278, + "learning_rate": 0.0001891298656560364, + "loss": 0.9634, + "step": 3382 + }, + { + "epoch": 0.6023860398860399, + "grad_norm": 0.48450782895088196, + "learning_rate": 0.00018912351809524563, + "loss": 0.809, + "step": 3383 + }, + { + "epoch": 0.6025641025641025, + "grad_norm": 0.621537983417511, + "learning_rate": 0.00018911716878826465, + "loss": 1.2031, + "step": 3384 + }, + { + "epoch": 0.6027421652421653, + "grad_norm": 0.6014544367790222, + "learning_rate": 0.00018911081773521787, + "loss": 1.1552, + "step": 3385 + }, + { + "epoch": 0.6029202279202279, + "grad_norm": 0.49995481967926025, + "learning_rate": 0.00018910446493622976, + "loss": 0.8569, + "step": 3386 + }, + { + "epoch": 0.6030982905982906, + "grad_norm": 0.5157307386398315, + "learning_rate": 0.00018909811039142472, + "loss": 0.9515, + "step": 3387 + }, + { + "epoch": 0.6032763532763533, + "grad_norm": 0.5164140462875366, + "learning_rate": 0.0001890917541009273, + "loss": 0.9803, + "step": 3388 + }, + { + "epoch": 0.603454415954416, + "grad_norm": 0.5555596947669983, + "learning_rate": 0.00018908539606486206, + "loss": 1.2994, + "step": 3389 + }, + { + "epoch": 0.6036324786324786, + "grad_norm": 0.605697512626648, + "learning_rate": 0.00018907903628335353, + "loss": 1.2865, + "step": 3390 + }, + { + "epoch": 0.6038105413105413, + "grad_norm": 0.5700713992118835, + "learning_rate": 0.0001890726747565263, + "loss": 1.2493, + "step": 3391 + }, + { + "epoch": 0.603988603988604, + "grad_norm": 0.5516746044158936, + "learning_rate": 0.0001890663114845051, + "loss": 1.2743, + "step": 3392 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 0.5233162641525269, + "learning_rate": 0.0001890599464674145, + "loss": 0.9237, + "step": 3393 + }, + { + "epoch": 0.6043447293447294, + "grad_norm": 0.5709942579269409, + "learning_rate": 0.00018905357970537925, + "loss": 0.9922, + "step": 3394 + }, + { + "epoch": 0.604522792022792, + "grad_norm": 0.48403796553611755, + "learning_rate": 0.0001890472111985241, + "loss": 1.1255, + "step": 3395 + }, + { + "epoch": 0.6047008547008547, + "grad_norm": 0.628718376159668, + "learning_rate": 0.00018904084094697386, + "loss": 1.1458, + "step": 3396 + }, + { + "epoch": 0.6048789173789174, + "grad_norm": 0.46822869777679443, + "learning_rate": 0.00018903446895085328, + "loss": 0.8727, + "step": 3397 + }, + { + "epoch": 0.60505698005698, + "grad_norm": 0.505584180355072, + "learning_rate": 0.00018902809521028724, + "loss": 1.1595, + "step": 3398 + }, + { + "epoch": 0.6052350427350427, + "grad_norm": 0.4494974911212921, + "learning_rate": 0.00018902171972540058, + "loss": 0.6685, + "step": 3399 + }, + { + "epoch": 0.6054131054131054, + "grad_norm": 0.5101519227027893, + "learning_rate": 0.0001890153424963183, + "loss": 0.9313, + "step": 3400 + }, + { + "epoch": 0.6055911680911681, + "grad_norm": 0.5081079602241516, + "learning_rate": 0.00018900896352316528, + "loss": 1.2588, + "step": 3401 + }, + { + "epoch": 0.6057692307692307, + "grad_norm": 0.5784309506416321, + "learning_rate": 0.00018900258280606653, + "loss": 1.2077, + "step": 3402 + }, + { + "epoch": 0.6059472934472935, + "grad_norm": 0.4506312608718872, + "learning_rate": 0.00018899620034514705, + "loss": 1.05, + "step": 3403 + }, + { + "epoch": 0.6061253561253561, + "grad_norm": 0.5243048071861267, + "learning_rate": 0.0001889898161405319, + "loss": 1.2295, + "step": 3404 + }, + { + "epoch": 0.6063034188034188, + "grad_norm": 0.5447196364402771, + "learning_rate": 0.00018898343019234615, + "loss": 1.1476, + "step": 3405 + }, + { + "epoch": 0.6064814814814815, + "grad_norm": 0.46813663840293884, + "learning_rate": 0.00018897704250071492, + "loss": 1.2113, + "step": 3406 + }, + { + "epoch": 0.6066595441595442, + "grad_norm": 0.5340631604194641, + "learning_rate": 0.00018897065306576342, + "loss": 1.1656, + "step": 3407 + }, + { + "epoch": 0.6068376068376068, + "grad_norm": 0.513708233833313, + "learning_rate": 0.00018896426188761675, + "loss": 1.1616, + "step": 3408 + }, + { + "epoch": 0.6070156695156695, + "grad_norm": 0.594601035118103, + "learning_rate": 0.00018895786896640023, + "loss": 1.2564, + "step": 3409 + }, + { + "epoch": 0.6071937321937322, + "grad_norm": 0.45067599415779114, + "learning_rate": 0.000188951474302239, + "loss": 1.0107, + "step": 3410 + }, + { + "epoch": 0.6073717948717948, + "grad_norm": 0.5394250750541687, + "learning_rate": 0.00018894507789525843, + "loss": 1.4081, + "step": 3411 + }, + { + "epoch": 0.6075498575498576, + "grad_norm": 0.5612049102783203, + "learning_rate": 0.00018893867974558383, + "loss": 1.1015, + "step": 3412 + }, + { + "epoch": 0.6077279202279202, + "grad_norm": 0.4794061779975891, + "learning_rate": 0.00018893227985334056, + "loss": 1.2103, + "step": 3413 + }, + { + "epoch": 0.6079059829059829, + "grad_norm": 0.6060562133789062, + "learning_rate": 0.00018892587821865402, + "loss": 1.3693, + "step": 3414 + }, + { + "epoch": 0.6080840455840456, + "grad_norm": 0.44624534249305725, + "learning_rate": 0.00018891947484164963, + "loss": 0.8209, + "step": 3415 + }, + { + "epoch": 0.6082621082621082, + "grad_norm": 0.49297213554382324, + "learning_rate": 0.0001889130697224528, + "loss": 1.2027, + "step": 3416 + }, + { + "epoch": 0.6084401709401709, + "grad_norm": 0.4431746304035187, + "learning_rate": 0.0001889066628611891, + "loss": 1.0347, + "step": 3417 + }, + { + "epoch": 0.6086182336182336, + "grad_norm": 0.5425933599472046, + "learning_rate": 0.00018890025425798404, + "loss": 1.0556, + "step": 3418 + }, + { + "epoch": 0.6087962962962963, + "grad_norm": 0.5502763390541077, + "learning_rate": 0.00018889384391296315, + "loss": 1.2362, + "step": 3419 + }, + { + "epoch": 0.6089743589743589, + "grad_norm": 0.5442292094230652, + "learning_rate": 0.00018888743182625203, + "loss": 1.1306, + "step": 3420 + }, + { + "epoch": 0.6091524216524217, + "grad_norm": 0.4651123583316803, + "learning_rate": 0.00018888101799797636, + "loss": 0.9305, + "step": 3421 + }, + { + "epoch": 0.6093304843304843, + "grad_norm": 0.4713892340660095, + "learning_rate": 0.00018887460242826177, + "loss": 1.0789, + "step": 3422 + }, + { + "epoch": 0.6095085470085471, + "grad_norm": 0.5283244848251343, + "learning_rate": 0.00018886818511723398, + "loss": 1.345, + "step": 3423 + }, + { + "epoch": 0.6096866096866097, + "grad_norm": 0.5527324080467224, + "learning_rate": 0.0001888617660650187, + "loss": 1.1297, + "step": 3424 + }, + { + "epoch": 0.6098646723646723, + "grad_norm": 0.5412901043891907, + "learning_rate": 0.00018885534527174168, + "loss": 1.1213, + "step": 3425 + }, + { + "epoch": 0.6100427350427351, + "grad_norm": 0.5295354127883911, + "learning_rate": 0.00018884892273752878, + "loss": 1.1217, + "step": 3426 + }, + { + "epoch": 0.6102207977207977, + "grad_norm": 0.461900532245636, + "learning_rate": 0.0001888424984625058, + "loss": 0.827, + "step": 3427 + }, + { + "epoch": 0.6103988603988604, + "grad_norm": 0.4922671616077423, + "learning_rate": 0.00018883607244679865, + "loss": 1.2216, + "step": 3428 + }, + { + "epoch": 0.6105769230769231, + "grad_norm": 0.5080927014350891, + "learning_rate": 0.00018882964469053317, + "loss": 1.2446, + "step": 3429 + }, + { + "epoch": 0.6107549857549858, + "grad_norm": 0.5523943901062012, + "learning_rate": 0.00018882321519383534, + "loss": 1.3346, + "step": 3430 + }, + { + "epoch": 0.6109330484330484, + "grad_norm": 0.5105271935462952, + "learning_rate": 0.0001888167839568311, + "loss": 1.1311, + "step": 3431 + }, + { + "epoch": 0.6111111111111112, + "grad_norm": 0.5635872483253479, + "learning_rate": 0.0001888103509796465, + "loss": 1.1875, + "step": 3432 + }, + { + "epoch": 0.6112891737891738, + "grad_norm": 0.4619547426700592, + "learning_rate": 0.00018880391626240755, + "loss": 0.9176, + "step": 3433 + }, + { + "epoch": 0.6114672364672364, + "grad_norm": 0.5896356105804443, + "learning_rate": 0.00018879747980524034, + "loss": 1.0251, + "step": 3434 + }, + { + "epoch": 0.6116452991452992, + "grad_norm": 0.49062737822532654, + "learning_rate": 0.000188791041608271, + "loss": 1.1598, + "step": 3435 + }, + { + "epoch": 0.6118233618233618, + "grad_norm": 0.45717164874076843, + "learning_rate": 0.00018878460167162558, + "loss": 0.8647, + "step": 3436 + }, + { + "epoch": 0.6120014245014245, + "grad_norm": 0.5903525352478027, + "learning_rate": 0.00018877815999543038, + "loss": 0.9671, + "step": 3437 + }, + { + "epoch": 0.6121794871794872, + "grad_norm": 0.5315384268760681, + "learning_rate": 0.00018877171657981153, + "loss": 1.1759, + "step": 3438 + }, + { + "epoch": 0.6123575498575499, + "grad_norm": 0.5650150775909424, + "learning_rate": 0.0001887652714248953, + "loss": 1.0128, + "step": 3439 + }, + { + "epoch": 0.6125356125356125, + "grad_norm": 0.49841752648353577, + "learning_rate": 0.000188758824530808, + "loss": 1.1259, + "step": 3440 + }, + { + "epoch": 0.6127136752136753, + "grad_norm": 0.4985620975494385, + "learning_rate": 0.00018875237589767593, + "loss": 1.0158, + "step": 3441 + }, + { + "epoch": 0.6128917378917379, + "grad_norm": 0.45266565680503845, + "learning_rate": 0.00018874592552562536, + "loss": 0.93, + "step": 3442 + }, + { + "epoch": 0.6130698005698005, + "grad_norm": 0.5696130990982056, + "learning_rate": 0.00018873947341478274, + "loss": 1.1432, + "step": 3443 + }, + { + "epoch": 0.6132478632478633, + "grad_norm": 0.5211645364761353, + "learning_rate": 0.00018873301956527451, + "loss": 1.1317, + "step": 3444 + }, + { + "epoch": 0.6134259259259259, + "grad_norm": 0.4991866946220398, + "learning_rate": 0.00018872656397722707, + "loss": 1.0362, + "step": 3445 + }, + { + "epoch": 0.6136039886039886, + "grad_norm": 0.5109508037567139, + "learning_rate": 0.00018872010665076694, + "loss": 1.2728, + "step": 3446 + }, + { + "epoch": 0.6137820512820513, + "grad_norm": 0.5838373899459839, + "learning_rate": 0.00018871364758602058, + "loss": 1.1131, + "step": 3447 + }, + { + "epoch": 0.613960113960114, + "grad_norm": 0.5139824151992798, + "learning_rate": 0.00018870718678311462, + "loss": 1.238, + "step": 3448 + }, + { + "epoch": 0.6141381766381766, + "grad_norm": 0.4852082431316376, + "learning_rate": 0.00018870072424217562, + "loss": 1.0677, + "step": 3449 + }, + { + "epoch": 0.6143162393162394, + "grad_norm": 0.5312315225601196, + "learning_rate": 0.00018869425996333018, + "loss": 1.178, + "step": 3450 + }, + { + "epoch": 0.614494301994302, + "grad_norm": 0.6343565583229065, + "learning_rate": 0.00018868779394670492, + "loss": 0.8839, + "step": 3451 + }, + { + "epoch": 0.6146723646723646, + "grad_norm": 0.6029773950576782, + "learning_rate": 0.00018868132619242662, + "loss": 1.1188, + "step": 3452 + }, + { + "epoch": 0.6148504273504274, + "grad_norm": 0.5246016383171082, + "learning_rate": 0.00018867485670062193, + "loss": 1.0797, + "step": 3453 + }, + { + "epoch": 0.61502849002849, + "grad_norm": 0.49307698011398315, + "learning_rate": 0.00018866838547141763, + "loss": 0.9749, + "step": 3454 + }, + { + "epoch": 0.6152065527065527, + "grad_norm": 0.5232903361320496, + "learning_rate": 0.00018866191250494052, + "loss": 1.0785, + "step": 3455 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 0.5545645356178284, + "learning_rate": 0.0001886554378013174, + "loss": 1.0496, + "step": 3456 + }, + { + "epoch": 0.6155626780626781, + "grad_norm": 0.493945837020874, + "learning_rate": 0.00018864896136067515, + "loss": 0.9248, + "step": 3457 + }, + { + "epoch": 0.6157407407407407, + "grad_norm": 0.5223548412322998, + "learning_rate": 0.00018864248318314065, + "loss": 1.0617, + "step": 3458 + }, + { + "epoch": 0.6159188034188035, + "grad_norm": 0.5666514039039612, + "learning_rate": 0.00018863600326884082, + "loss": 0.9981, + "step": 3459 + }, + { + "epoch": 0.6160968660968661, + "grad_norm": 0.4648127257823944, + "learning_rate": 0.00018862952161790265, + "loss": 0.917, + "step": 3460 + }, + { + "epoch": 0.6162749287749287, + "grad_norm": 0.590326189994812, + "learning_rate": 0.0001886230382304531, + "loss": 1.044, + "step": 3461 + }, + { + "epoch": 0.6164529914529915, + "grad_norm": 0.5511625409126282, + "learning_rate": 0.00018861655310661925, + "loss": 1.0988, + "step": 3462 + }, + { + "epoch": 0.6166310541310541, + "grad_norm": 0.567182183265686, + "learning_rate": 0.0001886100662465281, + "loss": 1.3017, + "step": 3463 + }, + { + "epoch": 0.6168091168091168, + "grad_norm": 0.5708897709846497, + "learning_rate": 0.0001886035776503068, + "loss": 0.9123, + "step": 3464 + }, + { + "epoch": 0.6169871794871795, + "grad_norm": 0.4945180416107178, + "learning_rate": 0.0001885970873180824, + "loss": 1.1645, + "step": 3465 + }, + { + "epoch": 0.6171652421652422, + "grad_norm": 0.4713336229324341, + "learning_rate": 0.00018859059524998215, + "loss": 1.0546, + "step": 3466 + }, + { + "epoch": 0.6173433048433048, + "grad_norm": 0.532859206199646, + "learning_rate": 0.0001885841014461332, + "loss": 1.0795, + "step": 3467 + }, + { + "epoch": 0.6175213675213675, + "grad_norm": 0.5165733695030212, + "learning_rate": 0.00018857760590666284, + "loss": 1.1284, + "step": 3468 + }, + { + "epoch": 0.6176994301994302, + "grad_norm": 0.48623126745224, + "learning_rate": 0.00018857110863169826, + "loss": 0.8618, + "step": 3469 + }, + { + "epoch": 0.6178774928774928, + "grad_norm": 0.628559947013855, + "learning_rate": 0.0001885646096213668, + "loss": 1.1089, + "step": 3470 + }, + { + "epoch": 0.6180555555555556, + "grad_norm": 0.503545880317688, + "learning_rate": 0.0001885581088757958, + "loss": 1.2311, + "step": 3471 + }, + { + "epoch": 0.6182336182336182, + "grad_norm": 0.6172101497650146, + "learning_rate": 0.00018855160639511264, + "loss": 1.2651, + "step": 3472 + }, + { + "epoch": 0.6184116809116809, + "grad_norm": 0.49572527408599854, + "learning_rate": 0.00018854510217944465, + "loss": 1.1026, + "step": 3473 + }, + { + "epoch": 0.6185897435897436, + "grad_norm": 0.5373549461364746, + "learning_rate": 0.00018853859622891938, + "loss": 1.2562, + "step": 3474 + }, + { + "epoch": 0.6187678062678063, + "grad_norm": 0.5272396206855774, + "learning_rate": 0.0001885320885436642, + "loss": 1.1763, + "step": 3475 + }, + { + "epoch": 0.6189458689458689, + "grad_norm": 0.46584269404411316, + "learning_rate": 0.00018852557912380665, + "loss": 1.1762, + "step": 3476 + }, + { + "epoch": 0.6191239316239316, + "grad_norm": 0.4798245131969452, + "learning_rate": 0.0001885190679694743, + "loss": 0.9229, + "step": 3477 + }, + { + "epoch": 0.6193019943019943, + "grad_norm": 0.5221366286277771, + "learning_rate": 0.0001885125550807947, + "loss": 1.1078, + "step": 3478 + }, + { + "epoch": 0.6194800569800569, + "grad_norm": 0.5051897168159485, + "learning_rate": 0.0001885060404578954, + "loss": 1.0055, + "step": 3479 + }, + { + "epoch": 0.6196581196581197, + "grad_norm": 0.492662250995636, + "learning_rate": 0.00018849952410090413, + "loss": 1.1172, + "step": 3480 + }, + { + "epoch": 0.6198361823361823, + "grad_norm": 0.4906775951385498, + "learning_rate": 0.00018849300600994853, + "loss": 1.1223, + "step": 3481 + }, + { + "epoch": 0.6200142450142451, + "grad_norm": 0.5032641291618347, + "learning_rate": 0.0001884864861851563, + "loss": 0.9541, + "step": 3482 + }, + { + "epoch": 0.6201923076923077, + "grad_norm": 0.5262296795845032, + "learning_rate": 0.00018847996462665521, + "loss": 1.021, + "step": 3483 + }, + { + "epoch": 0.6203703703703703, + "grad_norm": 0.5253522992134094, + "learning_rate": 0.00018847344133457295, + "loss": 0.9075, + "step": 3484 + }, + { + "epoch": 0.6205484330484331, + "grad_norm": 0.4204299747943878, + "learning_rate": 0.00018846691630903744, + "loss": 0.895, + "step": 3485 + }, + { + "epoch": 0.6207264957264957, + "grad_norm": 0.557604193687439, + "learning_rate": 0.0001884603895501765, + "loss": 1.1758, + "step": 3486 + }, + { + "epoch": 0.6209045584045584, + "grad_norm": 0.5981321930885315, + "learning_rate": 0.00018845386105811795, + "loss": 1.1087, + "step": 3487 + }, + { + "epoch": 0.6210826210826211, + "grad_norm": 0.5285581946372986, + "learning_rate": 0.00018844733083298975, + "loss": 1.0692, + "step": 3488 + }, + { + "epoch": 0.6212606837606838, + "grad_norm": 0.5403170585632324, + "learning_rate": 0.00018844079887491986, + "loss": 1.1998, + "step": 3489 + }, + { + "epoch": 0.6214387464387464, + "grad_norm": 0.5471615791320801, + "learning_rate": 0.0001884342651840362, + "loss": 0.9556, + "step": 3490 + }, + { + "epoch": 0.6216168091168092, + "grad_norm": 0.6126871705055237, + "learning_rate": 0.00018842772976046686, + "loss": 1.2629, + "step": 3491 + }, + { + "epoch": 0.6217948717948718, + "grad_norm": 0.45669353008270264, + "learning_rate": 0.00018842119260433982, + "loss": 1.0203, + "step": 3492 + }, + { + "epoch": 0.6219729344729344, + "grad_norm": 0.4998520612716675, + "learning_rate": 0.0001884146537157832, + "loss": 1.0271, + "step": 3493 + }, + { + "epoch": 0.6221509971509972, + "grad_norm": 0.5820242166519165, + "learning_rate": 0.00018840811309492507, + "loss": 1.0321, + "step": 3494 + }, + { + "epoch": 0.6223290598290598, + "grad_norm": 0.581676185131073, + "learning_rate": 0.00018840157074189367, + "loss": 0.9219, + "step": 3495 + }, + { + "epoch": 0.6225071225071225, + "grad_norm": 0.6044120788574219, + "learning_rate": 0.0001883950266568171, + "loss": 1.1621, + "step": 3496 + }, + { + "epoch": 0.6226851851851852, + "grad_norm": 0.5448858737945557, + "learning_rate": 0.0001883884808398236, + "loss": 1.0686, + "step": 3497 + }, + { + "epoch": 0.6228632478632479, + "grad_norm": 0.4921551048755646, + "learning_rate": 0.00018838193329104143, + "loss": 1.2259, + "step": 3498 + }, + { + "epoch": 0.6230413105413105, + "grad_norm": 0.5374335646629333, + "learning_rate": 0.00018837538401059888, + "loss": 1.2608, + "step": 3499 + }, + { + "epoch": 0.6232193732193733, + "grad_norm": 0.5123008489608765, + "learning_rate": 0.0001883688329986243, + "loss": 0.8682, + "step": 3500 + }, + { + "epoch": 0.6233974358974359, + "grad_norm": 0.566145122051239, + "learning_rate": 0.00018836228025524595, + "loss": 1.1807, + "step": 3501 + }, + { + "epoch": 0.6235754985754985, + "grad_norm": 0.6658587455749512, + "learning_rate": 0.00018835572578059233, + "loss": 1.1641, + "step": 3502 + }, + { + "epoch": 0.6237535612535613, + "grad_norm": 0.4992465078830719, + "learning_rate": 0.00018834916957479177, + "loss": 0.9125, + "step": 3503 + }, + { + "epoch": 0.6239316239316239, + "grad_norm": 0.5081812739372253, + "learning_rate": 0.00018834261163797278, + "loss": 1.0939, + "step": 3504 + }, + { + "epoch": 0.6241096866096866, + "grad_norm": 0.5168607234954834, + "learning_rate": 0.0001883360519702638, + "loss": 1.2382, + "step": 3505 + }, + { + "epoch": 0.6242877492877493, + "grad_norm": 0.5517697334289551, + "learning_rate": 0.00018832949057179344, + "loss": 1.206, + "step": 3506 + }, + { + "epoch": 0.624465811965812, + "grad_norm": 0.4505497217178345, + "learning_rate": 0.00018832292744269013, + "loss": 0.8485, + "step": 3507 + }, + { + "epoch": 0.6246438746438746, + "grad_norm": 0.5230690240859985, + "learning_rate": 0.0001883163625830826, + "loss": 1.1701, + "step": 3508 + }, + { + "epoch": 0.6248219373219374, + "grad_norm": 0.5062205195426941, + "learning_rate": 0.00018830979599309937, + "loss": 1.0602, + "step": 3509 + }, + { + "epoch": 0.625, + "grad_norm": 0.49922460317611694, + "learning_rate": 0.00018830322767286913, + "loss": 1.1937, + "step": 3510 + }, + { + "epoch": 0.6251780626780626, + "grad_norm": 0.4637366831302643, + "learning_rate": 0.0001882966576225206, + "loss": 1.038, + "step": 3511 + }, + { + "epoch": 0.6253561253561254, + "grad_norm": 0.5330080389976501, + "learning_rate": 0.00018829008584218246, + "loss": 0.9308, + "step": 3512 + }, + { + "epoch": 0.625534188034188, + "grad_norm": 0.5443428754806519, + "learning_rate": 0.0001882835123319835, + "loss": 1.0006, + "step": 3513 + }, + { + "epoch": 0.6257122507122507, + "grad_norm": 0.5534018874168396, + "learning_rate": 0.00018827693709205253, + "loss": 1.2383, + "step": 3514 + }, + { + "epoch": 0.6258903133903134, + "grad_norm": 0.49207547307014465, + "learning_rate": 0.00018827036012251832, + "loss": 0.9804, + "step": 3515 + }, + { + "epoch": 0.6260683760683761, + "grad_norm": 0.4900086224079132, + "learning_rate": 0.0001882637814235098, + "loss": 1.012, + "step": 3516 + }, + { + "epoch": 0.6262464387464387, + "grad_norm": 0.5267475247383118, + "learning_rate": 0.00018825720099515585, + "loss": 1.1104, + "step": 3517 + }, + { + "epoch": 0.6264245014245015, + "grad_norm": 0.5711902379989624, + "learning_rate": 0.00018825061883758534, + "loss": 1.0616, + "step": 3518 + }, + { + "epoch": 0.6266025641025641, + "grad_norm": 0.5007771849632263, + "learning_rate": 0.0001882440349509273, + "loss": 0.9578, + "step": 3519 + }, + { + "epoch": 0.6267806267806267, + "grad_norm": 0.5657192468643188, + "learning_rate": 0.00018823744933531075, + "loss": 1.2768, + "step": 3520 + }, + { + "epoch": 0.6269586894586895, + "grad_norm": 0.6077173352241516, + "learning_rate": 0.00018823086199086462, + "loss": 1.147, + "step": 3521 + }, + { + "epoch": 0.6271367521367521, + "grad_norm": 0.5114718079566956, + "learning_rate": 0.000188224272917718, + "loss": 1.1176, + "step": 3522 + }, + { + "epoch": 0.6273148148148148, + "grad_norm": 0.4831676185131073, + "learning_rate": 0.0001882176821160001, + "loss": 0.8021, + "step": 3523 + }, + { + "epoch": 0.6274928774928775, + "grad_norm": 0.6327390670776367, + "learning_rate": 0.00018821108958583994, + "loss": 0.9449, + "step": 3524 + }, + { + "epoch": 0.6276709401709402, + "grad_norm": 0.5541796684265137, + "learning_rate": 0.00018820449532736672, + "loss": 1.2018, + "step": 3525 + }, + { + "epoch": 0.6278490028490028, + "grad_norm": 0.5224639773368835, + "learning_rate": 0.00018819789934070968, + "loss": 1.0138, + "step": 3526 + }, + { + "epoch": 0.6280270655270656, + "grad_norm": 0.49359360337257385, + "learning_rate": 0.00018819130162599798, + "loss": 1.0768, + "step": 3527 + }, + { + "epoch": 0.6282051282051282, + "grad_norm": 0.5525050759315491, + "learning_rate": 0.00018818470218336092, + "loss": 1.0883, + "step": 3528 + }, + { + "epoch": 0.6283831908831908, + "grad_norm": 0.5563427209854126, + "learning_rate": 0.00018817810101292787, + "loss": 1.1491, + "step": 3529 + }, + { + "epoch": 0.6285612535612536, + "grad_norm": 0.49363306164741516, + "learning_rate": 0.00018817149811482803, + "loss": 1.1409, + "step": 3530 + }, + { + "epoch": 0.6287393162393162, + "grad_norm": 0.5102340579032898, + "learning_rate": 0.00018816489348919086, + "loss": 1.1914, + "step": 3531 + }, + { + "epoch": 0.6289173789173789, + "grad_norm": 0.5173332691192627, + "learning_rate": 0.00018815828713614576, + "loss": 0.9308, + "step": 3532 + }, + { + "epoch": 0.6290954415954416, + "grad_norm": 0.5093010067939758, + "learning_rate": 0.00018815167905582216, + "loss": 0.9429, + "step": 3533 + }, + { + "epoch": 0.6292735042735043, + "grad_norm": 0.5453153848648071, + "learning_rate": 0.00018814506924834954, + "loss": 1.0147, + "step": 3534 + }, + { + "epoch": 0.6294515669515669, + "grad_norm": 0.5850773453712463, + "learning_rate": 0.00018813845771385737, + "loss": 1.3372, + "step": 3535 + }, + { + "epoch": 0.6296296296296297, + "grad_norm": 0.5095621943473816, + "learning_rate": 0.00018813184445247525, + "loss": 1.0515, + "step": 3536 + }, + { + "epoch": 0.6298076923076923, + "grad_norm": 0.6216054558753967, + "learning_rate": 0.00018812522946433266, + "loss": 0.8703, + "step": 3537 + }, + { + "epoch": 0.6299857549857549, + "grad_norm": 0.4945531189441681, + "learning_rate": 0.00018811861274955932, + "loss": 1.1485, + "step": 3538 + }, + { + "epoch": 0.6301638176638177, + "grad_norm": 0.47882601618766785, + "learning_rate": 0.00018811199430828477, + "loss": 1.1107, + "step": 3539 + }, + { + "epoch": 0.6303418803418803, + "grad_norm": 0.5005326867103577, + "learning_rate": 0.00018810537414063876, + "loss": 1.0237, + "step": 3540 + }, + { + "epoch": 0.6305199430199431, + "grad_norm": 0.5382370352745056, + "learning_rate": 0.00018809875224675093, + "loss": 0.9965, + "step": 3541 + }, + { + "epoch": 0.6306980056980057, + "grad_norm": 0.47002625465393066, + "learning_rate": 0.0001880921286267511, + "loss": 1.065, + "step": 3542 + }, + { + "epoch": 0.6308760683760684, + "grad_norm": 0.4519105851650238, + "learning_rate": 0.00018808550328076897, + "loss": 0.9312, + "step": 3543 + }, + { + "epoch": 0.6310541310541311, + "grad_norm": 0.45360881090164185, + "learning_rate": 0.0001880788762089344, + "loss": 1.0739, + "step": 3544 + }, + { + "epoch": 0.6312321937321937, + "grad_norm": 0.5578218698501587, + "learning_rate": 0.00018807224741137723, + "loss": 1.2478, + "step": 3545 + }, + { + "epoch": 0.6314102564102564, + "grad_norm": 0.4838615655899048, + "learning_rate": 0.0001880656168882273, + "loss": 1.0221, + "step": 3546 + }, + { + "epoch": 0.6315883190883191, + "grad_norm": 0.5733556747436523, + "learning_rate": 0.0001880589846396146, + "loss": 1.1249, + "step": 3547 + }, + { + "epoch": 0.6317663817663818, + "grad_norm": 0.4939686954021454, + "learning_rate": 0.00018805235066566894, + "loss": 0.8559, + "step": 3548 + }, + { + "epoch": 0.6319444444444444, + "grad_norm": 0.5072234869003296, + "learning_rate": 0.00018804571496652044, + "loss": 1.0842, + "step": 3549 + }, + { + "epoch": 0.6321225071225072, + "grad_norm": 0.4640493392944336, + "learning_rate": 0.00018803907754229903, + "loss": 1.0728, + "step": 3550 + }, + { + "epoch": 0.6323005698005698, + "grad_norm": 0.5314788818359375, + "learning_rate": 0.00018803243839313481, + "loss": 1.0752, + "step": 3551 + }, + { + "epoch": 0.6324786324786325, + "grad_norm": 0.5511462092399597, + "learning_rate": 0.0001880257975191578, + "loss": 1.0238, + "step": 3552 + }, + { + "epoch": 0.6326566951566952, + "grad_norm": 0.4980711042881012, + "learning_rate": 0.00018801915492049816, + "loss": 1.0981, + "step": 3553 + }, + { + "epoch": 0.6328347578347578, + "grad_norm": 0.7746123671531677, + "learning_rate": 0.00018801251059728604, + "loss": 1.0968, + "step": 3554 + }, + { + "epoch": 0.6330128205128205, + "grad_norm": 0.5006106495857239, + "learning_rate": 0.00018800586454965155, + "loss": 1.1802, + "step": 3555 + }, + { + "epoch": 0.6331908831908832, + "grad_norm": 0.49427780508995056, + "learning_rate": 0.000187999216777725, + "loss": 1.1257, + "step": 3556 + }, + { + "epoch": 0.6333689458689459, + "grad_norm": 0.5484146475791931, + "learning_rate": 0.00018799256728163662, + "loss": 1.1344, + "step": 3557 + }, + { + "epoch": 0.6335470085470085, + "grad_norm": 0.5007877349853516, + "learning_rate": 0.00018798591606151662, + "loss": 1.1328, + "step": 3558 + }, + { + "epoch": 0.6337250712250713, + "grad_norm": 0.5068148970603943, + "learning_rate": 0.00018797926311749544, + "loss": 0.976, + "step": 3559 + }, + { + "epoch": 0.6339031339031339, + "grad_norm": 0.44936859607696533, + "learning_rate": 0.00018797260844970334, + "loss": 0.9735, + "step": 3560 + }, + { + "epoch": 0.6340811965811965, + "grad_norm": 0.4592931866645813, + "learning_rate": 0.0001879659520582707, + "loss": 1.1306, + "step": 3561 + }, + { + "epoch": 0.6342592592592593, + "grad_norm": 0.4664020836353302, + "learning_rate": 0.00018795929394332795, + "loss": 1.0577, + "step": 3562 + }, + { + "epoch": 0.6344373219373219, + "grad_norm": 0.5638116002082825, + "learning_rate": 0.00018795263410500556, + "loss": 1.1747, + "step": 3563 + }, + { + "epoch": 0.6346153846153846, + "grad_norm": 0.524736225605011, + "learning_rate": 0.00018794597254343401, + "loss": 0.8964, + "step": 3564 + }, + { + "epoch": 0.6347934472934473, + "grad_norm": 0.4645404517650604, + "learning_rate": 0.00018793930925874386, + "loss": 0.8673, + "step": 3565 + }, + { + "epoch": 0.63497150997151, + "grad_norm": 0.4800064265727997, + "learning_rate": 0.00018793264425106558, + "loss": 1.0334, + "step": 3566 + }, + { + "epoch": 0.6351495726495726, + "grad_norm": 0.6202501058578491, + "learning_rate": 0.0001879259775205298, + "loss": 1.1061, + "step": 3567 + }, + { + "epoch": 0.6353276353276354, + "grad_norm": 0.503383457660675, + "learning_rate": 0.00018791930906726718, + "loss": 0.8545, + "step": 3568 + }, + { + "epoch": 0.635505698005698, + "grad_norm": 0.5256780982017517, + "learning_rate": 0.00018791263889140832, + "loss": 1.0785, + "step": 3569 + }, + { + "epoch": 0.6356837606837606, + "grad_norm": 0.47562023997306824, + "learning_rate": 0.00018790596699308392, + "loss": 1.0041, + "step": 3570 + }, + { + "epoch": 0.6358618233618234, + "grad_norm": 0.5103238224983215, + "learning_rate": 0.00018789929337242469, + "loss": 1.1488, + "step": 3571 + }, + { + "epoch": 0.636039886039886, + "grad_norm": 0.5023695826530457, + "learning_rate": 0.0001878926180295614, + "loss": 1.0696, + "step": 3572 + }, + { + "epoch": 0.6362179487179487, + "grad_norm": 0.5302290916442871, + "learning_rate": 0.00018788594096462487, + "loss": 1.0554, + "step": 3573 + }, + { + "epoch": 0.6363960113960114, + "grad_norm": 0.4798361361026764, + "learning_rate": 0.00018787926217774588, + "loss": 0.8872, + "step": 3574 + }, + { + "epoch": 0.6365740740740741, + "grad_norm": 0.5529209971427917, + "learning_rate": 0.00018787258166905527, + "loss": 1.0976, + "step": 3575 + }, + { + "epoch": 0.6367521367521367, + "grad_norm": 0.49757125973701477, + "learning_rate": 0.00018786589943868402, + "loss": 1.0049, + "step": 3576 + }, + { + "epoch": 0.6369301994301995, + "grad_norm": 0.5497848391532898, + "learning_rate": 0.00018785921548676295, + "loss": 1.2272, + "step": 3577 + }, + { + "epoch": 0.6371082621082621, + "grad_norm": 0.5061752200126648, + "learning_rate": 0.0001878525298134231, + "loss": 1.0307, + "step": 3578 + }, + { + "epoch": 0.6372863247863247, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00018784584241879538, + "loss": 1.1064, + "step": 3579 + }, + { + "epoch": 0.6374643874643875, + "grad_norm": 0.48312774300575256, + "learning_rate": 0.0001878391533030109, + "loss": 1.078, + "step": 3580 + }, + { + "epoch": 0.6376424501424501, + "grad_norm": 0.5059898495674133, + "learning_rate": 0.00018783246246620067, + "loss": 1.0922, + "step": 3581 + }, + { + "epoch": 0.6378205128205128, + "grad_norm": 0.5144124031066895, + "learning_rate": 0.00018782576990849581, + "loss": 1.0909, + "step": 3582 + }, + { + "epoch": 0.6379985754985755, + "grad_norm": 0.5535032153129578, + "learning_rate": 0.0001878190756300274, + "loss": 1.2579, + "step": 3583 + }, + { + "epoch": 0.6381766381766382, + "grad_norm": 0.49145692586898804, + "learning_rate": 0.00018781237963092667, + "loss": 1.0823, + "step": 3584 + }, + { + "epoch": 0.6383547008547008, + "grad_norm": 0.5245576500892639, + "learning_rate": 0.00018780568191132472, + "loss": 0.9595, + "step": 3585 + }, + { + "epoch": 0.6385327635327636, + "grad_norm": 0.5026637315750122, + "learning_rate": 0.00018779898247135287, + "loss": 1.153, + "step": 3586 + }, + { + "epoch": 0.6387108262108262, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018779228131114234, + "loss": 1.0661, + "step": 3587 + }, + { + "epoch": 0.6388888888888888, + "grad_norm": 0.517387330532074, + "learning_rate": 0.00018778557843082444, + "loss": 1.0113, + "step": 3588 + }, + { + "epoch": 0.6390669515669516, + "grad_norm": 0.5149948000907898, + "learning_rate": 0.00018777887383053047, + "loss": 0.9483, + "step": 3589 + }, + { + "epoch": 0.6392450142450142, + "grad_norm": 0.4854544997215271, + "learning_rate": 0.00018777216751039185, + "loss": 1.22, + "step": 3590 + }, + { + "epoch": 0.6394230769230769, + "grad_norm": 0.5317271947860718, + "learning_rate": 0.0001877654594705399, + "loss": 1.2483, + "step": 3591 + }, + { + "epoch": 0.6396011396011396, + "grad_norm": 0.4554755687713623, + "learning_rate": 0.0001877587497111061, + "loss": 0.9864, + "step": 3592 + }, + { + "epoch": 0.6397792022792023, + "grad_norm": 0.4833736717700958, + "learning_rate": 0.0001877520382322219, + "loss": 0.8895, + "step": 3593 + }, + { + "epoch": 0.6399572649572649, + "grad_norm": 0.5018072724342346, + "learning_rate": 0.00018774532503401878, + "loss": 1.2523, + "step": 3594 + }, + { + "epoch": 0.6401353276353277, + "grad_norm": 0.4478762447834015, + "learning_rate": 0.00018773861011662832, + "loss": 0.8833, + "step": 3595 + }, + { + "epoch": 0.6403133903133903, + "grad_norm": 0.5686985850334167, + "learning_rate": 0.00018773189348018205, + "loss": 0.9934, + "step": 3596 + }, + { + "epoch": 0.6404914529914529, + "grad_norm": 0.5144175291061401, + "learning_rate": 0.00018772517512481157, + "loss": 0.8149, + "step": 3597 + }, + { + "epoch": 0.6406695156695157, + "grad_norm": 0.5359936356544495, + "learning_rate": 0.00018771845505064852, + "loss": 1.1822, + "step": 3598 + }, + { + "epoch": 0.6408475783475783, + "grad_norm": 0.532573938369751, + "learning_rate": 0.00018771173325782457, + "loss": 1.0361, + "step": 3599 + }, + { + "epoch": 0.6410256410256411, + "grad_norm": 0.46121537685394287, + "learning_rate": 0.00018770500974647138, + "loss": 1.0792, + "step": 3600 + }, + { + "epoch": 0.6412037037037037, + "grad_norm": 0.4804821312427521, + "learning_rate": 0.00018769828451672076, + "loss": 1.1119, + "step": 3601 + }, + { + "epoch": 0.6413817663817664, + "grad_norm": 0.4955114722251892, + "learning_rate": 0.00018769155756870443, + "loss": 0.9312, + "step": 3602 + }, + { + "epoch": 0.6415598290598291, + "grad_norm": 0.4987298250198364, + "learning_rate": 0.00018768482890255415, + "loss": 1.2326, + "step": 3603 + }, + { + "epoch": 0.6417378917378918, + "grad_norm": 0.47216179966926575, + "learning_rate": 0.0001876780985184018, + "loss": 1.0114, + "step": 3604 + }, + { + "epoch": 0.6419159544159544, + "grad_norm": 0.5891931653022766, + "learning_rate": 0.0001876713664163793, + "loss": 1.2963, + "step": 3605 + }, + { + "epoch": 0.6420940170940171, + "grad_norm": 0.4645081162452698, + "learning_rate": 0.00018766463259661846, + "loss": 1.0874, + "step": 3606 + }, + { + "epoch": 0.6422720797720798, + "grad_norm": 0.5275476574897766, + "learning_rate": 0.00018765789705925125, + "loss": 0.9453, + "step": 3607 + }, + { + "epoch": 0.6424501424501424, + "grad_norm": 0.5884957313537598, + "learning_rate": 0.00018765115980440964, + "loss": 1.0796, + "step": 3608 + }, + { + "epoch": 0.6426282051282052, + "grad_norm": 0.4843178987503052, + "learning_rate": 0.00018764442083222567, + "loss": 1.1657, + "step": 3609 + }, + { + "epoch": 0.6428062678062678, + "grad_norm": 0.5188381671905518, + "learning_rate": 0.00018763768014283126, + "loss": 1.1109, + "step": 3610 + }, + { + "epoch": 0.6429843304843305, + "grad_norm": 0.4101468324661255, + "learning_rate": 0.00018763093773635863, + "loss": 0.895, + "step": 3611 + }, + { + "epoch": 0.6431623931623932, + "grad_norm": 0.4552084505558014, + "learning_rate": 0.00018762419361293979, + "loss": 0.9418, + "step": 3612 + }, + { + "epoch": 0.6433404558404558, + "grad_norm": 0.5924661159515381, + "learning_rate": 0.0001876174477727069, + "loss": 1.2562, + "step": 3613 + }, + { + "epoch": 0.6435185185185185, + "grad_norm": 0.5072348713874817, + "learning_rate": 0.00018761070021579212, + "loss": 1.1501, + "step": 3614 + }, + { + "epoch": 0.6436965811965812, + "grad_norm": 0.5312697887420654, + "learning_rate": 0.0001876039509423277, + "loss": 1.0751, + "step": 3615 + }, + { + "epoch": 0.6438746438746439, + "grad_norm": 0.6046462059020996, + "learning_rate": 0.0001875971999524458, + "loss": 1.0927, + "step": 3616 + }, + { + "epoch": 0.6440527065527065, + "grad_norm": 0.4992375373840332, + "learning_rate": 0.00018759044724627876, + "loss": 0.96, + "step": 3617 + }, + { + "epoch": 0.6442307692307693, + "grad_norm": 0.4983134865760803, + "learning_rate": 0.00018758369282395886, + "loss": 1.0599, + "step": 3618 + }, + { + "epoch": 0.6444088319088319, + "grad_norm": 0.5655683279037476, + "learning_rate": 0.00018757693668561843, + "loss": 1.2372, + "step": 3619 + }, + { + "epoch": 0.6445868945868946, + "grad_norm": 0.4968827962875366, + "learning_rate": 0.00018757017883138985, + "loss": 1.1639, + "step": 3620 + }, + { + "epoch": 0.6447649572649573, + "grad_norm": 0.5831420421600342, + "learning_rate": 0.00018756341926140553, + "loss": 0.9002, + "step": 3621 + }, + { + "epoch": 0.64494301994302, + "grad_norm": 0.4828467071056366, + "learning_rate": 0.0001875566579757979, + "loss": 0.9201, + "step": 3622 + }, + { + "epoch": 0.6451210826210826, + "grad_norm": 0.5067087411880493, + "learning_rate": 0.00018754989497469943, + "loss": 0.9874, + "step": 3623 + }, + { + "epoch": 0.6452991452991453, + "grad_norm": 0.5182318091392517, + "learning_rate": 0.00018754313025824267, + "loss": 1.1291, + "step": 3624 + }, + { + "epoch": 0.645477207977208, + "grad_norm": 0.472200483083725, + "learning_rate": 0.0001875363638265601, + "loss": 1.0286, + "step": 3625 + }, + { + "epoch": 0.6456552706552706, + "grad_norm": 0.4597308039665222, + "learning_rate": 0.0001875295956797843, + "loss": 0.7517, + "step": 3626 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 0.5358221530914307, + "learning_rate": 0.00018752282581804798, + "loss": 1.2264, + "step": 3627 + }, + { + "epoch": 0.646011396011396, + "grad_norm": 0.5268992781639099, + "learning_rate": 0.00018751605424148363, + "loss": 1.0801, + "step": 3628 + }, + { + "epoch": 0.6461894586894587, + "grad_norm": 0.5917379260063171, + "learning_rate": 0.00018750928095022403, + "loss": 0.9538, + "step": 3629 + }, + { + "epoch": 0.6463675213675214, + "grad_norm": 0.44506707787513733, + "learning_rate": 0.00018750250594440183, + "loss": 0.9818, + "step": 3630 + }, + { + "epoch": 0.646545584045584, + "grad_norm": 0.5578880906105042, + "learning_rate": 0.00018749572922414982, + "loss": 0.9958, + "step": 3631 + }, + { + "epoch": 0.6467236467236467, + "grad_norm": 0.5155318975448608, + "learning_rate": 0.00018748895078960076, + "loss": 1.2888, + "step": 3632 + }, + { + "epoch": 0.6469017094017094, + "grad_norm": 0.5117297768592834, + "learning_rate": 0.0001874821706408874, + "loss": 1.0452, + "step": 3633 + }, + { + "epoch": 0.6470797720797721, + "grad_norm": 0.5169841647148132, + "learning_rate": 0.00018747538877814267, + "loss": 1.1649, + "step": 3634 + }, + { + "epoch": 0.6472578347578347, + "grad_norm": 0.5001181960105896, + "learning_rate": 0.00018746860520149942, + "loss": 1.1472, + "step": 3635 + }, + { + "epoch": 0.6474358974358975, + "grad_norm": 0.6289856433868408, + "learning_rate": 0.00018746181991109056, + "loss": 1.0351, + "step": 3636 + }, + { + "epoch": 0.6476139601139601, + "grad_norm": 0.5490612983703613, + "learning_rate": 0.00018745503290704897, + "loss": 0.8938, + "step": 3637 + }, + { + "epoch": 0.6477920227920227, + "grad_norm": 0.47378283739089966, + "learning_rate": 0.00018744824418950775, + "loss": 0.937, + "step": 3638 + }, + { + "epoch": 0.6479700854700855, + "grad_norm": 0.6079059839248657, + "learning_rate": 0.0001874414537585998, + "loss": 1.0486, + "step": 3639 + }, + { + "epoch": 0.6481481481481481, + "grad_norm": 0.5351769924163818, + "learning_rate": 0.00018743466161445823, + "loss": 1.0316, + "step": 3640 + }, + { + "epoch": 0.6483262108262108, + "grad_norm": 0.5516425967216492, + "learning_rate": 0.0001874278677572161, + "loss": 1.1552, + "step": 3641 + }, + { + "epoch": 0.6485042735042735, + "grad_norm": 0.5027523636817932, + "learning_rate": 0.0001874210721870065, + "loss": 1.0491, + "step": 3642 + }, + { + "epoch": 0.6486823361823362, + "grad_norm": 0.5596168041229248, + "learning_rate": 0.00018741427490396258, + "loss": 1.0256, + "step": 3643 + }, + { + "epoch": 0.6488603988603988, + "grad_norm": 0.5601046681404114, + "learning_rate": 0.00018740747590821751, + "loss": 1.1604, + "step": 3644 + }, + { + "epoch": 0.6490384615384616, + "grad_norm": 0.49749523401260376, + "learning_rate": 0.0001874006751999046, + "loss": 1.0532, + "step": 3645 + }, + { + "epoch": 0.6492165242165242, + "grad_norm": 0.6226113438606262, + "learning_rate": 0.00018739387277915697, + "loss": 1.1402, + "step": 3646 + }, + { + "epoch": 0.6493945868945868, + "grad_norm": 0.6142009496688843, + "learning_rate": 0.00018738706864610794, + "loss": 1.2437, + "step": 3647 + }, + { + "epoch": 0.6495726495726496, + "grad_norm": 0.48814916610717773, + "learning_rate": 0.00018738026280089084, + "loss": 0.8429, + "step": 3648 + }, + { + "epoch": 0.6497507122507122, + "grad_norm": 0.5717982053756714, + "learning_rate": 0.00018737345524363902, + "loss": 1.1095, + "step": 3649 + }, + { + "epoch": 0.6499287749287749, + "grad_norm": 0.5150009989738464, + "learning_rate": 0.00018736664597448582, + "loss": 1.199, + "step": 3650 + }, + { + "epoch": 0.6501068376068376, + "grad_norm": 0.58461594581604, + "learning_rate": 0.00018735983499356472, + "loss": 1.0704, + "step": 3651 + }, + { + "epoch": 0.6502849002849003, + "grad_norm": 0.5108643770217896, + "learning_rate": 0.0001873530223010091, + "loss": 1.2039, + "step": 3652 + }, + { + "epoch": 0.6504629629629629, + "grad_norm": 0.513306736946106, + "learning_rate": 0.00018734620789695247, + "loss": 1.1448, + "step": 3653 + }, + { + "epoch": 0.6506410256410257, + "grad_norm": 0.5139986872673035, + "learning_rate": 0.00018733939178152835, + "loss": 1.0023, + "step": 3654 + }, + { + "epoch": 0.6508190883190883, + "grad_norm": 0.5187703967094421, + "learning_rate": 0.00018733257395487027, + "loss": 1.1304, + "step": 3655 + }, + { + "epoch": 0.6509971509971509, + "grad_norm": 0.5470501184463501, + "learning_rate": 0.00018732575441711183, + "loss": 1.0272, + "step": 3656 + }, + { + "epoch": 0.6511752136752137, + "grad_norm": 0.537309467792511, + "learning_rate": 0.00018731893316838665, + "loss": 1.0806, + "step": 3657 + }, + { + "epoch": 0.6513532763532763, + "grad_norm": 0.5187864899635315, + "learning_rate": 0.00018731211020882836, + "loss": 1.0154, + "step": 3658 + }, + { + "epoch": 0.6515313390313391, + "grad_norm": 0.48373252153396606, + "learning_rate": 0.00018730528553857062, + "loss": 1.0135, + "step": 3659 + }, + { + "epoch": 0.6517094017094017, + "grad_norm": 0.5645000338554382, + "learning_rate": 0.00018729845915774716, + "loss": 0.8924, + "step": 3660 + }, + { + "epoch": 0.6518874643874644, + "grad_norm": 0.5722129940986633, + "learning_rate": 0.00018729163106649178, + "loss": 1.2416, + "step": 3661 + }, + { + "epoch": 0.6520655270655271, + "grad_norm": 0.5904877185821533, + "learning_rate": 0.00018728480126493823, + "loss": 0.9792, + "step": 3662 + }, + { + "epoch": 0.6522435897435898, + "grad_norm": 0.5224713087081909, + "learning_rate": 0.00018727796975322026, + "loss": 1.079, + "step": 3663 + }, + { + "epoch": 0.6524216524216524, + "grad_norm": 0.5667217969894409, + "learning_rate": 0.00018727113653147184, + "loss": 1.1397, + "step": 3664 + }, + { + "epoch": 0.6525997150997151, + "grad_norm": 0.5274622440338135, + "learning_rate": 0.00018726430159982677, + "loss": 1.0569, + "step": 3665 + }, + { + "epoch": 0.6527777777777778, + "grad_norm": 0.5745310187339783, + "learning_rate": 0.00018725746495841896, + "loss": 1.2129, + "step": 3666 + }, + { + "epoch": 0.6529558404558404, + "grad_norm": 0.6123398542404175, + "learning_rate": 0.0001872506266073824, + "loss": 1.186, + "step": 3667 + }, + { + "epoch": 0.6531339031339032, + "grad_norm": 0.4983387291431427, + "learning_rate": 0.00018724378654685106, + "loss": 1.1957, + "step": 3668 + }, + { + "epoch": 0.6533119658119658, + "grad_norm": 0.5584192276000977, + "learning_rate": 0.00018723694477695897, + "loss": 1.0939, + "step": 3669 + }, + { + "epoch": 0.6534900284900285, + "grad_norm": 0.5318745374679565, + "learning_rate": 0.00018723010129784016, + "loss": 1.1869, + "step": 3670 + }, + { + "epoch": 0.6536680911680912, + "grad_norm": 0.4607617259025574, + "learning_rate": 0.0001872232561096287, + "loss": 0.8447, + "step": 3671 + }, + { + "epoch": 0.6538461538461539, + "grad_norm": 0.5312213897705078, + "learning_rate": 0.00018721640921245874, + "loss": 1.0623, + "step": 3672 + }, + { + "epoch": 0.6540242165242165, + "grad_norm": 0.5099136233329773, + "learning_rate": 0.0001872095606064644, + "loss": 0.7174, + "step": 3673 + }, + { + "epoch": 0.6542022792022792, + "grad_norm": 0.6894404888153076, + "learning_rate": 0.0001872027102917799, + "loss": 1.0251, + "step": 3674 + }, + { + "epoch": 0.6543803418803419, + "grad_norm": 0.5758535861968994, + "learning_rate": 0.00018719585826853944, + "loss": 1.1655, + "step": 3675 + }, + { + "epoch": 0.6545584045584045, + "grad_norm": 0.521824061870575, + "learning_rate": 0.0001871890045368773, + "loss": 1.1653, + "step": 3676 + }, + { + "epoch": 0.6547364672364673, + "grad_norm": 0.5370712280273438, + "learning_rate": 0.00018718214909692771, + "loss": 1.3152, + "step": 3677 + }, + { + "epoch": 0.6549145299145299, + "grad_norm": 0.4459827244281769, + "learning_rate": 0.000187175291948825, + "loss": 1.0953, + "step": 3678 + }, + { + "epoch": 0.6550925925925926, + "grad_norm": 0.44131460785865784, + "learning_rate": 0.00018716843309270353, + "loss": 0.8568, + "step": 3679 + }, + { + "epoch": 0.6552706552706553, + "grad_norm": 0.5529624819755554, + "learning_rate": 0.00018716157252869772, + "loss": 1.2085, + "step": 3680 + }, + { + "epoch": 0.655448717948718, + "grad_norm": 0.44604751467704773, + "learning_rate": 0.00018715471025694194, + "loss": 0.9605, + "step": 3681 + }, + { + "epoch": 0.6556267806267806, + "grad_norm": 0.4662449359893799, + "learning_rate": 0.0001871478462775707, + "loss": 1.2092, + "step": 3682 + }, + { + "epoch": 0.6558048433048433, + "grad_norm": 0.42632922530174255, + "learning_rate": 0.0001871409805907184, + "loss": 0.9141, + "step": 3683 + }, + { + "epoch": 0.655982905982906, + "grad_norm": 0.534009575843811, + "learning_rate": 0.00018713411319651958, + "loss": 1.0147, + "step": 3684 + }, + { + "epoch": 0.6561609686609686, + "grad_norm": 0.5433241724967957, + "learning_rate": 0.00018712724409510888, + "loss": 1.1998, + "step": 3685 + }, + { + "epoch": 0.6563390313390314, + "grad_norm": 0.4771319627761841, + "learning_rate": 0.0001871203732866208, + "loss": 1.0384, + "step": 3686 + }, + { + "epoch": 0.656517094017094, + "grad_norm": 0.507641077041626, + "learning_rate": 0.00018711350077119, + "loss": 0.9608, + "step": 3687 + }, + { + "epoch": 0.6566951566951567, + "grad_norm": 0.5069413185119629, + "learning_rate": 0.00018710662654895108, + "loss": 1.055, + "step": 3688 + }, + { + "epoch": 0.6568732193732194, + "grad_norm": 0.512340247631073, + "learning_rate": 0.00018709975062003876, + "loss": 0.9506, + "step": 3689 + }, + { + "epoch": 0.657051282051282, + "grad_norm": 0.5156390070915222, + "learning_rate": 0.00018709287298458778, + "loss": 1.0089, + "step": 3690 + }, + { + "epoch": 0.6572293447293447, + "grad_norm": 0.5101696252822876, + "learning_rate": 0.0001870859936427329, + "loss": 1.0441, + "step": 3691 + }, + { + "epoch": 0.6574074074074074, + "grad_norm": 0.4394689202308655, + "learning_rate": 0.00018707911259460884, + "loss": 0.9124, + "step": 3692 + }, + { + "epoch": 0.6575854700854701, + "grad_norm": 0.4842554032802582, + "learning_rate": 0.00018707222984035043, + "loss": 1.0051, + "step": 3693 + }, + { + "epoch": 0.6577635327635327, + "grad_norm": 0.6418108344078064, + "learning_rate": 0.00018706534538009262, + "loss": 1.1165, + "step": 3694 + }, + { + "epoch": 0.6579415954415955, + "grad_norm": 0.5596832036972046, + "learning_rate": 0.00018705845921397022, + "loss": 1.1127, + "step": 3695 + }, + { + "epoch": 0.6581196581196581, + "grad_norm": 0.6692909002304077, + "learning_rate": 0.00018705157134211813, + "loss": 1.2403, + "step": 3696 + }, + { + "epoch": 0.6582977207977208, + "grad_norm": 0.5046468377113342, + "learning_rate": 0.00018704468176467134, + "loss": 1.1016, + "step": 3697 + }, + { + "epoch": 0.6584757834757835, + "grad_norm": 0.6723586320877075, + "learning_rate": 0.00018703779048176485, + "loss": 1.1777, + "step": 3698 + }, + { + "epoch": 0.6586538461538461, + "grad_norm": 0.5269754528999329, + "learning_rate": 0.00018703089749353365, + "loss": 1.1441, + "step": 3699 + }, + { + "epoch": 0.6588319088319088, + "grad_norm": 0.5303323268890381, + "learning_rate": 0.0001870240028001128, + "loss": 1.07, + "step": 3700 + }, + { + "epoch": 0.6590099715099715, + "grad_norm": 0.4795511066913605, + "learning_rate": 0.00018701710640163738, + "loss": 1.0189, + "step": 3701 + }, + { + "epoch": 0.6591880341880342, + "grad_norm": 0.514659583568573, + "learning_rate": 0.00018701020829824255, + "loss": 1.0792, + "step": 3702 + }, + { + "epoch": 0.6593660968660968, + "grad_norm": 0.5407463312149048, + "learning_rate": 0.0001870033084900634, + "loss": 0.9346, + "step": 3703 + }, + { + "epoch": 0.6595441595441596, + "grad_norm": 0.5358424186706543, + "learning_rate": 0.0001869964069772352, + "loss": 1.1242, + "step": 3704 + }, + { + "epoch": 0.6597222222222222, + "grad_norm": 0.470825731754303, + "learning_rate": 0.00018698950375989307, + "loss": 0.9952, + "step": 3705 + }, + { + "epoch": 0.6599002849002849, + "grad_norm": 0.5711592435836792, + "learning_rate": 0.00018698259883817236, + "loss": 1.1678, + "step": 3706 + }, + { + "epoch": 0.6600783475783476, + "grad_norm": 0.5298995971679688, + "learning_rate": 0.00018697569221220832, + "loss": 0.869, + "step": 3707 + }, + { + "epoch": 0.6602564102564102, + "grad_norm": 0.5453875064849854, + "learning_rate": 0.00018696878388213626, + "loss": 0.9706, + "step": 3708 + }, + { + "epoch": 0.6604344729344729, + "grad_norm": 0.6219926476478577, + "learning_rate": 0.00018696187384809154, + "loss": 1.1902, + "step": 3709 + }, + { + "epoch": 0.6606125356125356, + "grad_norm": 0.5972491502761841, + "learning_rate": 0.00018695496211020953, + "loss": 1.2054, + "step": 3710 + }, + { + "epoch": 0.6607905982905983, + "grad_norm": 0.5048904418945312, + "learning_rate": 0.0001869480486686257, + "loss": 1.0405, + "step": 3711 + }, + { + "epoch": 0.6609686609686609, + "grad_norm": 0.5474200248718262, + "learning_rate": 0.00018694113352347546, + "loss": 1.09, + "step": 3712 + }, + { + "epoch": 0.6611467236467237, + "grad_norm": 0.5073318481445312, + "learning_rate": 0.00018693421667489432, + "loss": 1.0698, + "step": 3713 + }, + { + "epoch": 0.6613247863247863, + "grad_norm": 0.5693208575248718, + "learning_rate": 0.0001869272981230178, + "loss": 0.9664, + "step": 3714 + }, + { + "epoch": 0.6615028490028491, + "grad_norm": 0.5678503513336182, + "learning_rate": 0.00018692037786798143, + "loss": 1.0895, + "step": 3715 + }, + { + "epoch": 0.6616809116809117, + "grad_norm": 0.4950976073741913, + "learning_rate": 0.00018691345590992082, + "loss": 0.9584, + "step": 3716 + }, + { + "epoch": 0.6618589743589743, + "grad_norm": 0.4944666624069214, + "learning_rate": 0.0001869065322489716, + "loss": 0.8607, + "step": 3717 + }, + { + "epoch": 0.6620370370370371, + "grad_norm": 0.5197804570198059, + "learning_rate": 0.0001868996068852694, + "loss": 1.2335, + "step": 3718 + }, + { + "epoch": 0.6622150997150997, + "grad_norm": 0.6550365686416626, + "learning_rate": 0.00018689267981894994, + "loss": 1.0441, + "step": 3719 + }, + { + "epoch": 0.6623931623931624, + "grad_norm": 0.5331503748893738, + "learning_rate": 0.00018688575105014888, + "loss": 1.1696, + "step": 3720 + }, + { + "epoch": 0.6625712250712251, + "grad_norm": 0.47304239869117737, + "learning_rate": 0.00018687882057900207, + "loss": 0.9695, + "step": 3721 + }, + { + "epoch": 0.6627492877492878, + "grad_norm": 0.5653772354125977, + "learning_rate": 0.00018687188840564524, + "loss": 1.2082, + "step": 3722 + }, + { + "epoch": 0.6629273504273504, + "grad_norm": 0.5323491096496582, + "learning_rate": 0.00018686495453021417, + "loss": 0.9106, + "step": 3723 + }, + { + "epoch": 0.6631054131054132, + "grad_norm": 0.5612817406654358, + "learning_rate": 0.00018685801895284483, + "loss": 1.1302, + "step": 3724 + }, + { + "epoch": 0.6632834757834758, + "grad_norm": 0.4562164545059204, + "learning_rate": 0.000186851081673673, + "loss": 0.8886, + "step": 3725 + }, + { + "epoch": 0.6634615384615384, + "grad_norm": 0.5006430745124817, + "learning_rate": 0.00018684414269283463, + "loss": 0.9128, + "step": 3726 + }, + { + "epoch": 0.6636396011396012, + "grad_norm": 0.5305442810058594, + "learning_rate": 0.0001868372020104657, + "loss": 1.1766, + "step": 3727 + }, + { + "epoch": 0.6638176638176638, + "grad_norm": 0.6129274368286133, + "learning_rate": 0.0001868302596267022, + "loss": 1.04, + "step": 3728 + }, + { + "epoch": 0.6639957264957265, + "grad_norm": 0.5530399084091187, + "learning_rate": 0.00018682331554168013, + "loss": 1.4114, + "step": 3729 + }, + { + "epoch": 0.6641737891737892, + "grad_norm": 0.5397193431854248, + "learning_rate": 0.00018681636975553557, + "loss": 1.1945, + "step": 3730 + }, + { + "epoch": 0.6643518518518519, + "grad_norm": 0.5510205030441284, + "learning_rate": 0.00018680942226840456, + "loss": 1.0489, + "step": 3731 + }, + { + "epoch": 0.6645299145299145, + "grad_norm": 0.5519221425056458, + "learning_rate": 0.00018680247308042324, + "loss": 1.1633, + "step": 3732 + }, + { + "epoch": 0.6647079772079773, + "grad_norm": 0.4848768711090088, + "learning_rate": 0.00018679552219172784, + "loss": 0.8716, + "step": 3733 + }, + { + "epoch": 0.6648860398860399, + "grad_norm": 0.5490246415138245, + "learning_rate": 0.0001867885696024544, + "loss": 1.1347, + "step": 3734 + }, + { + "epoch": 0.6650641025641025, + "grad_norm": 0.5281458497047424, + "learning_rate": 0.00018678161531273928, + "loss": 1.0987, + "step": 3735 + }, + { + "epoch": 0.6652421652421653, + "grad_norm": 0.5313079953193665, + "learning_rate": 0.00018677465932271867, + "loss": 0.9705, + "step": 3736 + }, + { + "epoch": 0.6654202279202279, + "grad_norm": 0.5425750017166138, + "learning_rate": 0.0001867677016325289, + "loss": 1.1847, + "step": 3737 + }, + { + "epoch": 0.6655982905982906, + "grad_norm": 0.5796298980712891, + "learning_rate": 0.0001867607422423062, + "loss": 1.2639, + "step": 3738 + }, + { + "epoch": 0.6657763532763533, + "grad_norm": 0.49738675355911255, + "learning_rate": 0.00018675378115218702, + "loss": 1.0536, + "step": 3739 + }, + { + "epoch": 0.665954415954416, + "grad_norm": 0.665250301361084, + "learning_rate": 0.0001867468183623077, + "loss": 1.2836, + "step": 3740 + }, + { + "epoch": 0.6661324786324786, + "grad_norm": 0.5184717178344727, + "learning_rate": 0.00018673985387280469, + "loss": 1.0497, + "step": 3741 + }, + { + "epoch": 0.6663105413105413, + "grad_norm": 0.5129656791687012, + "learning_rate": 0.00018673288768381442, + "loss": 1.2041, + "step": 3742 + }, + { + "epoch": 0.666488603988604, + "grad_norm": 0.5308768153190613, + "learning_rate": 0.00018672591979547337, + "loss": 1.2092, + "step": 3743 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.5059141516685486, + "learning_rate": 0.00018671895020791812, + "loss": 1.1929, + "step": 3744 + }, + { + "epoch": 0.6668447293447294, + "grad_norm": 0.5237857103347778, + "learning_rate": 0.00018671197892128517, + "loss": 1.2538, + "step": 3745 + }, + { + "epoch": 0.667022792022792, + "grad_norm": 0.450000137090683, + "learning_rate": 0.0001867050059357111, + "loss": 0.7138, + "step": 3746 + }, + { + "epoch": 0.6672008547008547, + "grad_norm": 0.5413795709609985, + "learning_rate": 0.00018669803125133258, + "loss": 1.1383, + "step": 3747 + }, + { + "epoch": 0.6673789173789174, + "grad_norm": 0.4657825529575348, + "learning_rate": 0.00018669105486828622, + "loss": 1.0518, + "step": 3748 + }, + { + "epoch": 0.66755698005698, + "grad_norm": 0.6198551654815674, + "learning_rate": 0.00018668407678670875, + "loss": 1.2697, + "step": 3749 + }, + { + "epoch": 0.6677350427350427, + "grad_norm": 0.5112186074256897, + "learning_rate": 0.00018667709700673685, + "loss": 0.9907, + "step": 3750 + }, + { + "epoch": 0.6679131054131054, + "grad_norm": 0.5446593761444092, + "learning_rate": 0.00018667011552850728, + "loss": 1.0708, + "step": 3751 + }, + { + "epoch": 0.6680911680911681, + "grad_norm": 0.5673866271972656, + "learning_rate": 0.00018666313235215682, + "loss": 1.05, + "step": 3752 + }, + { + "epoch": 0.6682692307692307, + "grad_norm": 0.4821988046169281, + "learning_rate": 0.00018665614747782235, + "loss": 1.0543, + "step": 3753 + }, + { + "epoch": 0.6684472934472935, + "grad_norm": 0.5158842206001282, + "learning_rate": 0.00018664916090564067, + "loss": 1.0331, + "step": 3754 + }, + { + "epoch": 0.6686253561253561, + "grad_norm": 0.45486921072006226, + "learning_rate": 0.00018664217263574865, + "loss": 0.9262, + "step": 3755 + }, + { + "epoch": 0.6688034188034188, + "grad_norm": 0.46193036437034607, + "learning_rate": 0.00018663518266828327, + "loss": 0.9858, + "step": 3756 + }, + { + "epoch": 0.6689814814814815, + "grad_norm": 0.5144094824790955, + "learning_rate": 0.00018662819100338148, + "loss": 1.0302, + "step": 3757 + }, + { + "epoch": 0.6691595441595442, + "grad_norm": 0.5246134400367737, + "learning_rate": 0.0001866211976411802, + "loss": 1.064, + "step": 3758 + }, + { + "epoch": 0.6693376068376068, + "grad_norm": 0.4853166937828064, + "learning_rate": 0.0001866142025818165, + "loss": 0.9481, + "step": 3759 + }, + { + "epoch": 0.6695156695156695, + "grad_norm": 0.5029586553573608, + "learning_rate": 0.00018660720582542743, + "loss": 0.9443, + "step": 3760 + }, + { + "epoch": 0.6696937321937322, + "grad_norm": 0.5373172163963318, + "learning_rate": 0.0001866002073721501, + "loss": 1.1401, + "step": 3761 + }, + { + "epoch": 0.6698717948717948, + "grad_norm": 0.6236287951469421, + "learning_rate": 0.00018659320722212158, + "loss": 1.1255, + "step": 3762 + }, + { + "epoch": 0.6700498575498576, + "grad_norm": 0.5470684766769409, + "learning_rate": 0.00018658620537547903, + "loss": 1.0622, + "step": 3763 + }, + { + "epoch": 0.6702279202279202, + "grad_norm": 0.63177090883255, + "learning_rate": 0.00018657920183235964, + "loss": 0.9736, + "step": 3764 + }, + { + "epoch": 0.6704059829059829, + "grad_norm": 0.5456309914588928, + "learning_rate": 0.00018657219659290068, + "loss": 1.027, + "step": 3765 + }, + { + "epoch": 0.6705840455840456, + "grad_norm": 0.4816138744354248, + "learning_rate": 0.00018656518965723935, + "loss": 0.7801, + "step": 3766 + }, + { + "epoch": 0.6707621082621082, + "grad_norm": 0.4811640679836273, + "learning_rate": 0.00018655818102551294, + "loss": 1.0535, + "step": 3767 + }, + { + "epoch": 0.6709401709401709, + "grad_norm": 0.4677673280239105, + "learning_rate": 0.00018655117069785884, + "loss": 1.1043, + "step": 3768 + }, + { + "epoch": 0.6711182336182336, + "grad_norm": 0.5628635883331299, + "learning_rate": 0.0001865441586744143, + "loss": 1.0392, + "step": 3769 + }, + { + "epoch": 0.6712962962962963, + "grad_norm": 0.5484504103660583, + "learning_rate": 0.00018653714495531673, + "loss": 1.1533, + "step": 3770 + }, + { + "epoch": 0.6714743589743589, + "grad_norm": 0.5830571055412292, + "learning_rate": 0.0001865301295407036, + "loss": 1.2479, + "step": 3771 + }, + { + "epoch": 0.6716524216524217, + "grad_norm": 0.5516841411590576, + "learning_rate": 0.00018652311243071235, + "loss": 1.2152, + "step": 3772 + }, + { + "epoch": 0.6718304843304843, + "grad_norm": 0.6360766291618347, + "learning_rate": 0.0001865160936254804, + "loss": 1.0752, + "step": 3773 + }, + { + "epoch": 0.6720085470085471, + "grad_norm": 0.6038610935211182, + "learning_rate": 0.00018650907312514533, + "loss": 1.2425, + "step": 3774 + }, + { + "epoch": 0.6721866096866097, + "grad_norm": 0.49572908878326416, + "learning_rate": 0.0001865020509298447, + "loss": 1.0057, + "step": 3775 + }, + { + "epoch": 0.6723646723646723, + "grad_norm": 0.4551616311073303, + "learning_rate": 0.00018649502703971607, + "loss": 1.0763, + "step": 3776 + }, + { + "epoch": 0.6725427350427351, + "grad_norm": 0.6621482372283936, + "learning_rate": 0.00018648800145489706, + "loss": 1.0306, + "step": 3777 + }, + { + "epoch": 0.6727207977207977, + "grad_norm": 0.5523806810379028, + "learning_rate": 0.0001864809741755253, + "loss": 0.9906, + "step": 3778 + }, + { + "epoch": 0.6728988603988604, + "grad_norm": 0.5527048110961914, + "learning_rate": 0.00018647394520173856, + "loss": 1.0734, + "step": 3779 + }, + { + "epoch": 0.6730769230769231, + "grad_norm": 0.573573887348175, + "learning_rate": 0.00018646691453367444, + "loss": 1.1409, + "step": 3780 + }, + { + "epoch": 0.6732549857549858, + "grad_norm": 0.6273239254951477, + "learning_rate": 0.00018645988217147079, + "loss": 0.9682, + "step": 3781 + }, + { + "epoch": 0.6734330484330484, + "grad_norm": 0.4917762279510498, + "learning_rate": 0.00018645284811526534, + "loss": 0.9681, + "step": 3782 + }, + { + "epoch": 0.6736111111111112, + "grad_norm": 0.4901154339313507, + "learning_rate": 0.0001864458123651959, + "loss": 1.1828, + "step": 3783 + }, + { + "epoch": 0.6737891737891738, + "grad_norm": 0.6292546391487122, + "learning_rate": 0.00018643877492140036, + "loss": 1.1987, + "step": 3784 + }, + { + "epoch": 0.6739672364672364, + "grad_norm": 0.5334137678146362, + "learning_rate": 0.0001864317357840166, + "loss": 1.0347, + "step": 3785 + }, + { + "epoch": 0.6741452991452992, + "grad_norm": 0.6064338684082031, + "learning_rate": 0.0001864246949531825, + "loss": 1.4154, + "step": 3786 + }, + { + "epoch": 0.6743233618233618, + "grad_norm": 0.5442034602165222, + "learning_rate": 0.000186417652429036, + "loss": 1.2604, + "step": 3787 + }, + { + "epoch": 0.6745014245014245, + "grad_norm": 0.490858793258667, + "learning_rate": 0.00018641060821171518, + "loss": 1.1511, + "step": 3788 + }, + { + "epoch": 0.6746794871794872, + "grad_norm": 0.571116030216217, + "learning_rate": 0.00018640356230135798, + "loss": 1.1479, + "step": 3789 + }, + { + "epoch": 0.6748575498575499, + "grad_norm": 0.4857785105705261, + "learning_rate": 0.00018639651469810247, + "loss": 0.9, + "step": 3790 + }, + { + "epoch": 0.6750356125356125, + "grad_norm": 0.5320703983306885, + "learning_rate": 0.0001863894654020867, + "loss": 1.2284, + "step": 3791 + }, + { + "epoch": 0.6752136752136753, + "grad_norm": 0.5586925745010376, + "learning_rate": 0.0001863824144134488, + "loss": 1.1183, + "step": 3792 + }, + { + "epoch": 0.6753917378917379, + "grad_norm": 0.47740885615348816, + "learning_rate": 0.000186375361732327, + "loss": 1.1512, + "step": 3793 + }, + { + "epoch": 0.6755698005698005, + "grad_norm": 0.5867732167243958, + "learning_rate": 0.00018636830735885935, + "loss": 1.1903, + "step": 3794 + }, + { + "epoch": 0.6757478632478633, + "grad_norm": 0.5013887882232666, + "learning_rate": 0.0001863612512931842, + "loss": 0.8581, + "step": 3795 + }, + { + "epoch": 0.6759259259259259, + "grad_norm": 0.6026871204376221, + "learning_rate": 0.0001863541935354397, + "loss": 0.9581, + "step": 3796 + }, + { + "epoch": 0.6761039886039886, + "grad_norm": 0.5238468647003174, + "learning_rate": 0.00018634713408576415, + "loss": 1.0949, + "step": 3797 + }, + { + "epoch": 0.6762820512820513, + "grad_norm": 0.5128598213195801, + "learning_rate": 0.00018634007294429585, + "loss": 0.8992, + "step": 3798 + }, + { + "epoch": 0.676460113960114, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018633301011117324, + "loss": 1.0793, + "step": 3799 + }, + { + "epoch": 0.6766381766381766, + "grad_norm": 0.592566728591919, + "learning_rate": 0.00018632594558653457, + "loss": 1.3242, + "step": 3800 + }, + { + "epoch": 0.6768162393162394, + "grad_norm": 0.4953067898750305, + "learning_rate": 0.0001863188793705184, + "loss": 0.9925, + "step": 3801 + }, + { + "epoch": 0.676994301994302, + "grad_norm": 0.4989747107028961, + "learning_rate": 0.00018631181146326305, + "loss": 1.0677, + "step": 3802 + }, + { + "epoch": 0.6771723646723646, + "grad_norm": 0.5375261902809143, + "learning_rate": 0.00018630474186490705, + "loss": 1.0556, + "step": 3803 + }, + { + "epoch": 0.6773504273504274, + "grad_norm": 0.6512624025344849, + "learning_rate": 0.00018629767057558894, + "loss": 1.2041, + "step": 3804 + }, + { + "epoch": 0.67752849002849, + "grad_norm": 0.5428260564804077, + "learning_rate": 0.00018629059759544723, + "loss": 0.9645, + "step": 3805 + }, + { + "epoch": 0.6777065527065527, + "grad_norm": 0.5598662495613098, + "learning_rate": 0.00018628352292462052, + "loss": 1.1683, + "step": 3806 + }, + { + "epoch": 0.6778846153846154, + "grad_norm": 0.49351340532302856, + "learning_rate": 0.0001862764465632474, + "loss": 1.1622, + "step": 3807 + }, + { + "epoch": 0.6780626780626781, + "grad_norm": 0.4796701669692993, + "learning_rate": 0.00018626936851146657, + "loss": 1.0017, + "step": 3808 + }, + { + "epoch": 0.6782407407407407, + "grad_norm": 0.444533109664917, + "learning_rate": 0.00018626228876941664, + "loss": 0.9145, + "step": 3809 + }, + { + "epoch": 0.6784188034188035, + "grad_norm": 0.5197392702102661, + "learning_rate": 0.00018625520733723635, + "loss": 1.283, + "step": 3810 + }, + { + "epoch": 0.6785968660968661, + "grad_norm": 0.48785829544067383, + "learning_rate": 0.00018624812421506447, + "loss": 1.1084, + "step": 3811 + }, + { + "epoch": 0.6787749287749287, + "grad_norm": 0.5083680152893066, + "learning_rate": 0.00018624103940303974, + "loss": 0.9071, + "step": 3812 + }, + { + "epoch": 0.6789529914529915, + "grad_norm": 0.553819477558136, + "learning_rate": 0.00018623395290130103, + "loss": 0.9986, + "step": 3813 + }, + { + "epoch": 0.6791310541310541, + "grad_norm": 0.5347508788108826, + "learning_rate": 0.00018622686470998713, + "loss": 1.0148, + "step": 3814 + }, + { + "epoch": 0.6793091168091168, + "grad_norm": 0.5080769062042236, + "learning_rate": 0.00018621977482923693, + "loss": 1.0169, + "step": 3815 + }, + { + "epoch": 0.6794871794871795, + "grad_norm": 0.5444077849388123, + "learning_rate": 0.00018621268325918938, + "loss": 1.172, + "step": 3816 + }, + { + "epoch": 0.6796652421652422, + "grad_norm": 0.521946132183075, + "learning_rate": 0.00018620558999998335, + "loss": 1.0247, + "step": 3817 + }, + { + "epoch": 0.6798433048433048, + "grad_norm": 0.5257413983345032, + "learning_rate": 0.00018619849505175786, + "loss": 1.1574, + "step": 3818 + }, + { + "epoch": 0.6800213675213675, + "grad_norm": 0.5473007559776306, + "learning_rate": 0.00018619139841465193, + "loss": 1.1254, + "step": 3819 + }, + { + "epoch": 0.6801994301994302, + "grad_norm": 0.5479872226715088, + "learning_rate": 0.00018618430008880463, + "loss": 1.0196, + "step": 3820 + }, + { + "epoch": 0.6803774928774928, + "grad_norm": 0.5918973088264465, + "learning_rate": 0.00018617720007435497, + "loss": 1.082, + "step": 3821 + }, + { + "epoch": 0.6805555555555556, + "grad_norm": 0.5411791801452637, + "learning_rate": 0.0001861700983714421, + "loss": 0.7723, + "step": 3822 + }, + { + "epoch": 0.6807336182336182, + "grad_norm": 0.5466326475143433, + "learning_rate": 0.00018616299498020516, + "loss": 1.0979, + "step": 3823 + }, + { + "epoch": 0.6809116809116809, + "grad_norm": 0.5405182838439941, + "learning_rate": 0.00018615588990078332, + "loss": 0.8891, + "step": 3824 + }, + { + "epoch": 0.6810897435897436, + "grad_norm": 0.5415780544281006, + "learning_rate": 0.00018614878313331579, + "loss": 1.0927, + "step": 3825 + }, + { + "epoch": 0.6812678062678063, + "grad_norm": 0.5284909605979919, + "learning_rate": 0.00018614167467794182, + "loss": 1.0684, + "step": 3826 + }, + { + "epoch": 0.6814458689458689, + "grad_norm": 0.4873995780944824, + "learning_rate": 0.00018613456453480062, + "loss": 1.1653, + "step": 3827 + }, + { + "epoch": 0.6816239316239316, + "grad_norm": 0.5506551265716553, + "learning_rate": 0.0001861274527040316, + "loss": 0.9876, + "step": 3828 + }, + { + "epoch": 0.6818019943019943, + "grad_norm": 0.5031297206878662, + "learning_rate": 0.0001861203391857741, + "loss": 1.067, + "step": 3829 + }, + { + "epoch": 0.6819800569800569, + "grad_norm": 0.622346043586731, + "learning_rate": 0.0001861132239801674, + "loss": 1.1514, + "step": 3830 + }, + { + "epoch": 0.6821581196581197, + "grad_norm": 0.47706183791160583, + "learning_rate": 0.000186106107087351, + "loss": 0.9857, + "step": 3831 + }, + { + "epoch": 0.6823361823361823, + "grad_norm": 0.5082845091819763, + "learning_rate": 0.00018609898850746424, + "loss": 1.123, + "step": 3832 + }, + { + "epoch": 0.6825142450142451, + "grad_norm": 0.5119805932044983, + "learning_rate": 0.00018609186824064671, + "loss": 1.1386, + "step": 3833 + }, + { + "epoch": 0.6826923076923077, + "grad_norm": 0.5247541069984436, + "learning_rate": 0.00018608474628703788, + "loss": 0.9433, + "step": 3834 + }, + { + "epoch": 0.6828703703703703, + "grad_norm": 0.4618282616138458, + "learning_rate": 0.00018607762264677722, + "loss": 0.8727, + "step": 3835 + }, + { + "epoch": 0.6830484330484331, + "grad_norm": 0.6014040112495422, + "learning_rate": 0.00018607049732000436, + "loss": 1.1823, + "step": 3836 + }, + { + "epoch": 0.6832264957264957, + "grad_norm": 0.6489043831825256, + "learning_rate": 0.00018606337030685892, + "loss": 1.1466, + "step": 3837 + }, + { + "epoch": 0.6834045584045584, + "grad_norm": 0.5527763366699219, + "learning_rate": 0.00018605624160748053, + "loss": 1.3015, + "step": 3838 + }, + { + "epoch": 0.6835826210826211, + "grad_norm": 0.5628284215927124, + "learning_rate": 0.0001860491112220088, + "loss": 1.1504, + "step": 3839 + }, + { + "epoch": 0.6837606837606838, + "grad_norm": 0.5414566993713379, + "learning_rate": 0.00018604197915058355, + "loss": 1.0155, + "step": 3840 + }, + { + "epoch": 0.6839387464387464, + "grad_norm": 0.5378929376602173, + "learning_rate": 0.00018603484539334443, + "loss": 0.8917, + "step": 3841 + }, + { + "epoch": 0.6841168091168092, + "grad_norm": 0.5953748822212219, + "learning_rate": 0.00018602770995043125, + "loss": 1.1971, + "step": 3842 + }, + { + "epoch": 0.6842948717948718, + "grad_norm": 0.511813759803772, + "learning_rate": 0.00018602057282198376, + "loss": 1.1345, + "step": 3843 + }, + { + "epoch": 0.6844729344729344, + "grad_norm": 0.5145484209060669, + "learning_rate": 0.00018601343400814185, + "loss": 1.0786, + "step": 3844 + }, + { + "epoch": 0.6846509971509972, + "grad_norm": 0.5199604034423828, + "learning_rate": 0.00018600629350904542, + "loss": 1.2063, + "step": 3845 + }, + { + "epoch": 0.6848290598290598, + "grad_norm": 0.5653825998306274, + "learning_rate": 0.0001859991513248343, + "loss": 1.0314, + "step": 3846 + }, + { + "epoch": 0.6850071225071225, + "grad_norm": 0.5660843849182129, + "learning_rate": 0.00018599200745564843, + "loss": 1.2754, + "step": 3847 + }, + { + "epoch": 0.6851851851851852, + "grad_norm": 0.5225719809532166, + "learning_rate": 0.00018598486190162788, + "loss": 1.0837, + "step": 3848 + }, + { + "epoch": 0.6853632478632479, + "grad_norm": 0.5011669397354126, + "learning_rate": 0.00018597771466291252, + "loss": 1.1, + "step": 3849 + }, + { + "epoch": 0.6855413105413105, + "grad_norm": 0.5923115015029907, + "learning_rate": 0.00018597056573964245, + "loss": 1.1875, + "step": 3850 + }, + { + "epoch": 0.6857193732193733, + "grad_norm": 0.5666482448577881, + "learning_rate": 0.00018596341513195776, + "loss": 1.1663, + "step": 3851 + }, + { + "epoch": 0.6858974358974359, + "grad_norm": 0.5396790504455566, + "learning_rate": 0.0001859562628399985, + "loss": 1.1179, + "step": 3852 + }, + { + "epoch": 0.6860754985754985, + "grad_norm": 0.5709532499313354, + "learning_rate": 0.00018594910886390485, + "loss": 1.0369, + "step": 3853 + }, + { + "epoch": 0.6862535612535613, + "grad_norm": 0.45524322986602783, + "learning_rate": 0.00018594195320381692, + "loss": 1.0171, + "step": 3854 + }, + { + "epoch": 0.6864316239316239, + "grad_norm": 0.6130724549293518, + "learning_rate": 0.00018593479585987498, + "loss": 1.1944, + "step": 3855 + }, + { + "epoch": 0.6866096866096866, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.0001859276368322192, + "loss": 1.2567, + "step": 3856 + }, + { + "epoch": 0.6867877492877493, + "grad_norm": 0.49919846653938293, + "learning_rate": 0.00018592047612098992, + "loss": 0.9459, + "step": 3857 + }, + { + "epoch": 0.686965811965812, + "grad_norm": 0.5776857733726501, + "learning_rate": 0.00018591331372632734, + "loss": 1.2456, + "step": 3858 + }, + { + "epoch": 0.6871438746438746, + "grad_norm": 0.4740692377090454, + "learning_rate": 0.00018590614964837188, + "loss": 1.0401, + "step": 3859 + }, + { + "epoch": 0.6873219373219374, + "grad_norm": 0.5015742182731628, + "learning_rate": 0.00018589898388726389, + "loss": 1.2052, + "step": 3860 + }, + { + "epoch": 0.6875, + "grad_norm": 0.4819730818271637, + "learning_rate": 0.0001858918164431437, + "loss": 1.007, + "step": 3861 + }, + { + "epoch": 0.6876780626780626, + "grad_norm": 0.5510426163673401, + "learning_rate": 0.00018588464731615184, + "loss": 1.0123, + "step": 3862 + }, + { + "epoch": 0.6878561253561254, + "grad_norm": 0.4950829744338989, + "learning_rate": 0.00018587747650642867, + "loss": 1.033, + "step": 3863 + }, + { + "epoch": 0.688034188034188, + "grad_norm": 0.5278680920600891, + "learning_rate": 0.0001858703040141148, + "loss": 1.0912, + "step": 3864 + }, + { + "epoch": 0.6882122507122507, + "grad_norm": 0.6359158158302307, + "learning_rate": 0.00018586312983935068, + "loss": 1.2868, + "step": 3865 + }, + { + "epoch": 0.6883903133903134, + "grad_norm": 0.5098239183425903, + "learning_rate": 0.0001858559539822769, + "loss": 0.8364, + "step": 3866 + }, + { + "epoch": 0.6885683760683761, + "grad_norm": 0.5651038289070129, + "learning_rate": 0.000185848776443034, + "loss": 1.1983, + "step": 3867 + }, + { + "epoch": 0.6887464387464387, + "grad_norm": 0.5305678248405457, + "learning_rate": 0.00018584159722176272, + "loss": 1.32, + "step": 3868 + }, + { + "epoch": 0.6889245014245015, + "grad_norm": 0.5481845140457153, + "learning_rate": 0.00018583441631860368, + "loss": 1.013, + "step": 3869 + }, + { + "epoch": 0.6891025641025641, + "grad_norm": 0.5214795470237732, + "learning_rate": 0.00018582723373369753, + "loss": 1.172, + "step": 3870 + }, + { + "epoch": 0.6892806267806267, + "grad_norm": 0.6282780766487122, + "learning_rate": 0.00018582004946718502, + "loss": 1.7304, + "step": 3871 + }, + { + "epoch": 0.6894586894586895, + "grad_norm": 0.5266988277435303, + "learning_rate": 0.0001858128635192069, + "loss": 1.1418, + "step": 3872 + }, + { + "epoch": 0.6896367521367521, + "grad_norm": 0.4761001467704773, + "learning_rate": 0.000185805675889904, + "loss": 0.8585, + "step": 3873 + }, + { + "epoch": 0.6898148148148148, + "grad_norm": 0.528779923915863, + "learning_rate": 0.00018579848657941715, + "loss": 1.0036, + "step": 3874 + }, + { + "epoch": 0.6899928774928775, + "grad_norm": 0.5427684783935547, + "learning_rate": 0.00018579129558788716, + "loss": 0.9769, + "step": 3875 + }, + { + "epoch": 0.6901709401709402, + "grad_norm": 0.6229544281959534, + "learning_rate": 0.00018578410291545495, + "loss": 1.2848, + "step": 3876 + }, + { + "epoch": 0.6903490028490028, + "grad_norm": 0.6602693200111389, + "learning_rate": 0.00018577690856226147, + "loss": 1.2713, + "step": 3877 + }, + { + "epoch": 0.6905270655270656, + "grad_norm": 0.45884042978286743, + "learning_rate": 0.0001857697125284476, + "loss": 0.9143, + "step": 3878 + }, + { + "epoch": 0.6907051282051282, + "grad_norm": 0.4956444203853607, + "learning_rate": 0.00018576251481415443, + "loss": 0.9646, + "step": 3879 + }, + { + "epoch": 0.6908831908831908, + "grad_norm": 0.473561555147171, + "learning_rate": 0.00018575531541952292, + "loss": 0.843, + "step": 3880 + }, + { + "epoch": 0.6910612535612536, + "grad_norm": 0.4676312506198883, + "learning_rate": 0.00018574811434469415, + "loss": 0.9464, + "step": 3881 + }, + { + "epoch": 0.6912393162393162, + "grad_norm": 0.5452045202255249, + "learning_rate": 0.00018574091158980922, + "loss": 0.985, + "step": 3882 + }, + { + "epoch": 0.6914173789173789, + "grad_norm": 0.6274946331977844, + "learning_rate": 0.0001857337071550092, + "loss": 1.0357, + "step": 3883 + }, + { + "epoch": 0.6915954415954416, + "grad_norm": 0.5533788800239563, + "learning_rate": 0.00018572650104043531, + "loss": 1.2636, + "step": 3884 + }, + { + "epoch": 0.6917735042735043, + "grad_norm": 0.48312318325042725, + "learning_rate": 0.00018571929324622872, + "loss": 1.2402, + "step": 3885 + }, + { + "epoch": 0.6919515669515669, + "grad_norm": 0.6087453961372375, + "learning_rate": 0.00018571208377253062, + "loss": 1.2961, + "step": 3886 + }, + { + "epoch": 0.6921296296296297, + "grad_norm": 0.49156486988067627, + "learning_rate": 0.00018570487261948234, + "loss": 0.9585, + "step": 3887 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 0.5200015902519226, + "learning_rate": 0.0001856976597872251, + "loss": 0.9274, + "step": 3888 + }, + { + "epoch": 0.6924857549857549, + "grad_norm": 0.5185118913650513, + "learning_rate": 0.0001856904452759002, + "loss": 1.0015, + "step": 3889 + }, + { + "epoch": 0.6926638176638177, + "grad_norm": 0.5859049558639526, + "learning_rate": 0.00018568322908564904, + "loss": 1.0959, + "step": 3890 + }, + { + "epoch": 0.6928418803418803, + "grad_norm": 0.5882301926612854, + "learning_rate": 0.00018567601121661302, + "loss": 1.3214, + "step": 3891 + }, + { + "epoch": 0.6930199430199431, + "grad_norm": 0.6475503444671631, + "learning_rate": 0.0001856687916689335, + "loss": 1.3265, + "step": 3892 + }, + { + "epoch": 0.6931980056980057, + "grad_norm": 0.46175432205200195, + "learning_rate": 0.000185661570442752, + "loss": 0.8547, + "step": 3893 + }, + { + "epoch": 0.6933760683760684, + "grad_norm": 0.5362716913223267, + "learning_rate": 0.00018565434753820998, + "loss": 0.974, + "step": 3894 + }, + { + "epoch": 0.6935541310541311, + "grad_norm": 0.4317963719367981, + "learning_rate": 0.00018564712295544896, + "loss": 0.7653, + "step": 3895 + }, + { + "epoch": 0.6937321937321937, + "grad_norm": 0.5679717659950256, + "learning_rate": 0.00018563989669461047, + "loss": 1.0691, + "step": 3896 + }, + { + "epoch": 0.6939102564102564, + "grad_norm": 0.5058363676071167, + "learning_rate": 0.00018563266875583608, + "loss": 1.0665, + "step": 3897 + }, + { + "epoch": 0.6940883190883191, + "grad_norm": 0.5365496277809143, + "learning_rate": 0.00018562543913926746, + "loss": 0.9963, + "step": 3898 + }, + { + "epoch": 0.6942663817663818, + "grad_norm": 0.49945300817489624, + "learning_rate": 0.0001856182078450462, + "loss": 0.8668, + "step": 3899 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 0.5869430899620056, + "learning_rate": 0.00018561097487331405, + "loss": 1.1942, + "step": 3900 + }, + { + "epoch": 0.6946225071225072, + "grad_norm": 0.5188950300216675, + "learning_rate": 0.0001856037402242127, + "loss": 0.9493, + "step": 3901 + }, + { + "epoch": 0.6948005698005698, + "grad_norm": 0.510788083076477, + "learning_rate": 0.00018559650389788384, + "loss": 0.9989, + "step": 3902 + }, + { + "epoch": 0.6949786324786325, + "grad_norm": 0.5360601544380188, + "learning_rate": 0.0001855892658944693, + "loss": 1.2766, + "step": 3903 + }, + { + "epoch": 0.6951566951566952, + "grad_norm": 0.522502601146698, + "learning_rate": 0.00018558202621411093, + "loss": 0.8774, + "step": 3904 + }, + { + "epoch": 0.6953347578347578, + "grad_norm": 0.5330635905265808, + "learning_rate": 0.00018557478485695052, + "loss": 0.972, + "step": 3905 + }, + { + "epoch": 0.6955128205128205, + "grad_norm": 0.5387479066848755, + "learning_rate": 0.00018556754182312996, + "loss": 1.0574, + "step": 3906 + }, + { + "epoch": 0.6956908831908832, + "grad_norm": 0.5357984900474548, + "learning_rate": 0.00018556029711279116, + "loss": 1.396, + "step": 3907 + }, + { + "epoch": 0.6958689458689459, + "grad_norm": 0.5647178292274475, + "learning_rate": 0.00018555305072607612, + "loss": 1.3304, + "step": 3908 + }, + { + "epoch": 0.6960470085470085, + "grad_norm": 0.46460914611816406, + "learning_rate": 0.00018554580266312673, + "loss": 0.9574, + "step": 3909 + }, + { + "epoch": 0.6962250712250713, + "grad_norm": 0.6206206679344177, + "learning_rate": 0.00018553855292408503, + "loss": 1.1637, + "step": 3910 + }, + { + "epoch": 0.6964031339031339, + "grad_norm": 0.5899842977523804, + "learning_rate": 0.00018553130150909312, + "loss": 1.1067, + "step": 3911 + }, + { + "epoch": 0.6965811965811965, + "grad_norm": 0.47294262051582336, + "learning_rate": 0.000185524048418293, + "loss": 1.1516, + "step": 3912 + }, + { + "epoch": 0.6967592592592593, + "grad_norm": 0.5791197419166565, + "learning_rate": 0.00018551679365182684, + "loss": 1.0007, + "step": 3913 + }, + { + "epoch": 0.6969373219373219, + "grad_norm": 0.5678651332855225, + "learning_rate": 0.00018550953720983672, + "loss": 1.2698, + "step": 3914 + }, + { + "epoch": 0.6971153846153846, + "grad_norm": 0.6509683728218079, + "learning_rate": 0.0001855022790924649, + "loss": 1.0354, + "step": 3915 + }, + { + "epoch": 0.6972934472934473, + "grad_norm": 0.5176648497581482, + "learning_rate": 0.0001854950192998535, + "loss": 1.1243, + "step": 3916 + }, + { + "epoch": 0.69747150997151, + "grad_norm": 0.520631730556488, + "learning_rate": 0.00018548775783214477, + "loss": 1.1371, + "step": 3917 + }, + { + "epoch": 0.6976495726495726, + "grad_norm": 0.5408333539962769, + "learning_rate": 0.00018548049468948108, + "loss": 1.1185, + "step": 3918 + }, + { + "epoch": 0.6978276353276354, + "grad_norm": 0.5423790216445923, + "learning_rate": 0.00018547322987200461, + "loss": 1.1539, + "step": 3919 + }, + { + "epoch": 0.698005698005698, + "grad_norm": 0.5422113537788391, + "learning_rate": 0.0001854659633798578, + "loss": 1.171, + "step": 3920 + }, + { + "epoch": 0.6981837606837606, + "grad_norm": 0.5113416314125061, + "learning_rate": 0.00018545869521318292, + "loss": 1.0597, + "step": 3921 + }, + { + "epoch": 0.6983618233618234, + "grad_norm": 0.49901214241981506, + "learning_rate": 0.00018545142537212248, + "loss": 1.1043, + "step": 3922 + }, + { + "epoch": 0.698539886039886, + "grad_norm": 0.6606622338294983, + "learning_rate": 0.00018544415385681885, + "loss": 1.1797, + "step": 3923 + }, + { + "epoch": 0.6987179487179487, + "grad_norm": 0.4786234498023987, + "learning_rate": 0.00018543688066741454, + "loss": 0.9532, + "step": 3924 + }, + { + "epoch": 0.6988960113960114, + "grad_norm": 0.5900700688362122, + "learning_rate": 0.00018542960580405203, + "loss": 1.1171, + "step": 3925 + }, + { + "epoch": 0.6990740740740741, + "grad_norm": 0.53485506772995, + "learning_rate": 0.00018542232926687383, + "loss": 1.1535, + "step": 3926 + }, + { + "epoch": 0.6992521367521367, + "grad_norm": 0.5269177556037903, + "learning_rate": 0.00018541505105602255, + "loss": 1.0287, + "step": 3927 + }, + { + "epoch": 0.6994301994301995, + "grad_norm": 0.5185505151748657, + "learning_rate": 0.0001854077711716408, + "loss": 1.2526, + "step": 3928 + }, + { + "epoch": 0.6996082621082621, + "grad_norm": 0.5615512132644653, + "learning_rate": 0.00018540048961387115, + "loss": 1.0189, + "step": 3929 + }, + { + "epoch": 0.6997863247863247, + "grad_norm": 0.4492493271827698, + "learning_rate": 0.00018539320638285637, + "loss": 0.8917, + "step": 3930 + }, + { + "epoch": 0.6999643874643875, + "grad_norm": 0.5062302947044373, + "learning_rate": 0.00018538592147873906, + "loss": 1.053, + "step": 3931 + }, + { + "epoch": 0.7001424501424501, + "grad_norm": 0.5508798956871033, + "learning_rate": 0.000185378634901662, + "loss": 0.9638, + "step": 3932 + }, + { + "epoch": 0.7003205128205128, + "grad_norm": 0.463980108499527, + "learning_rate": 0.00018537134665176793, + "loss": 1.0945, + "step": 3933 + }, + { + "epoch": 0.7004985754985755, + "grad_norm": 0.5027088522911072, + "learning_rate": 0.0001853640567291997, + "loss": 1.1745, + "step": 3934 + }, + { + "epoch": 0.7006766381766382, + "grad_norm": 0.5006551146507263, + "learning_rate": 0.00018535676513410009, + "loss": 0.8521, + "step": 3935 + }, + { + "epoch": 0.7008547008547008, + "grad_norm": 0.5870724320411682, + "learning_rate": 0.000185349471866612, + "loss": 0.9197, + "step": 3936 + }, + { + "epoch": 0.7010327635327636, + "grad_norm": 0.5030696392059326, + "learning_rate": 0.00018534217692687825, + "loss": 1.1049, + "step": 3937 + }, + { + "epoch": 0.7012108262108262, + "grad_norm": 0.5212681889533997, + "learning_rate": 0.00018533488031504186, + "loss": 1.3397, + "step": 3938 + }, + { + "epoch": 0.7013888888888888, + "grad_norm": 0.5649709105491638, + "learning_rate": 0.0001853275820312458, + "loss": 1.1994, + "step": 3939 + }, + { + "epoch": 0.7015669515669516, + "grad_norm": 0.4892779290676117, + "learning_rate": 0.00018532028207563297, + "loss": 1.1511, + "step": 3940 + }, + { + "epoch": 0.7017450142450142, + "grad_norm": 0.4929407835006714, + "learning_rate": 0.00018531298044834643, + "loss": 1.0792, + "step": 3941 + }, + { + "epoch": 0.7019230769230769, + "grad_norm": 0.5645940899848938, + "learning_rate": 0.00018530567714952932, + "loss": 1.0937, + "step": 3942 + }, + { + "epoch": 0.7021011396011396, + "grad_norm": 0.5471178293228149, + "learning_rate": 0.00018529837217932466, + "loss": 1.193, + "step": 3943 + }, + { + "epoch": 0.7022792022792023, + "grad_norm": 0.576627790927887, + "learning_rate": 0.00018529106553787558, + "loss": 1.1032, + "step": 3944 + }, + { + "epoch": 0.7024572649572649, + "grad_norm": 0.5015735626220703, + "learning_rate": 0.00018528375722532526, + "loss": 1.066, + "step": 3945 + }, + { + "epoch": 0.7026353276353277, + "grad_norm": 0.5315404534339905, + "learning_rate": 0.00018527644724181683, + "loss": 1.2059, + "step": 3946 + }, + { + "epoch": 0.7028133903133903, + "grad_norm": 0.5516065955162048, + "learning_rate": 0.0001852691355874936, + "loss": 1.161, + "step": 3947 + }, + { + "epoch": 0.7029914529914529, + "grad_norm": 0.5026212930679321, + "learning_rate": 0.0001852618222624988, + "loss": 1.2616, + "step": 3948 + }, + { + "epoch": 0.7031695156695157, + "grad_norm": 0.49874603748321533, + "learning_rate": 0.0001852545072669757, + "loss": 0.805, + "step": 3949 + }, + { + "epoch": 0.7033475783475783, + "grad_norm": 0.47698748111724854, + "learning_rate": 0.00018524719060106763, + "loss": 1.2321, + "step": 3950 + }, + { + "epoch": 0.7035256410256411, + "grad_norm": 0.5201322436332703, + "learning_rate": 0.00018523987226491792, + "loss": 1.1577, + "step": 3951 + }, + { + "epoch": 0.7037037037037037, + "grad_norm": 0.5506543517112732, + "learning_rate": 0.00018523255225867002, + "loss": 1.2289, + "step": 3952 + }, + { + "epoch": 0.7038817663817664, + "grad_norm": 0.5691256523132324, + "learning_rate": 0.0001852252305824673, + "loss": 1.1945, + "step": 3953 + }, + { + "epoch": 0.7040598290598291, + "grad_norm": 0.5324838757514954, + "learning_rate": 0.00018521790723645322, + "loss": 1.1037, + "step": 3954 + }, + { + "epoch": 0.7042378917378918, + "grad_norm": 0.5238786339759827, + "learning_rate": 0.00018521058222077127, + "loss": 1.2075, + "step": 3955 + }, + { + "epoch": 0.7044159544159544, + "grad_norm": 0.4936453402042389, + "learning_rate": 0.00018520325553556498, + "loss": 1.0537, + "step": 3956 + }, + { + "epoch": 0.7045940170940171, + "grad_norm": 0.6198282837867737, + "learning_rate": 0.00018519592718097791, + "loss": 1.0728, + "step": 3957 + }, + { + "epoch": 0.7047720797720798, + "grad_norm": 0.44729140400886536, + "learning_rate": 0.0001851885971571536, + "loss": 0.8432, + "step": 3958 + }, + { + "epoch": 0.7049501424501424, + "grad_norm": 0.5884211659431458, + "learning_rate": 0.00018518126546423572, + "loss": 0.9515, + "step": 3959 + }, + { + "epoch": 0.7051282051282052, + "grad_norm": 0.5293807983398438, + "learning_rate": 0.00018517393210236788, + "loss": 1.1178, + "step": 3960 + }, + { + "epoch": 0.7053062678062678, + "grad_norm": 0.6036825180053711, + "learning_rate": 0.00018516659707169374, + "loss": 1.0408, + "step": 3961 + }, + { + "epoch": 0.7054843304843305, + "grad_norm": 0.5157122015953064, + "learning_rate": 0.0001851592603723571, + "loss": 1.2136, + "step": 3962 + }, + { + "epoch": 0.7056623931623932, + "grad_norm": 0.5354781150817871, + "learning_rate": 0.00018515192200450163, + "loss": 0.7165, + "step": 3963 + }, + { + "epoch": 0.7058404558404558, + "grad_norm": 0.6073734760284424, + "learning_rate": 0.00018514458196827111, + "loss": 1.3079, + "step": 3964 + }, + { + "epoch": 0.7060185185185185, + "grad_norm": 0.4324839413166046, + "learning_rate": 0.0001851372402638094, + "loss": 0.7903, + "step": 3965 + }, + { + "epoch": 0.7061965811965812, + "grad_norm": 0.6530333161354065, + "learning_rate": 0.00018512989689126034, + "loss": 1.3179, + "step": 3966 + }, + { + "epoch": 0.7063746438746439, + "grad_norm": 0.5500404238700867, + "learning_rate": 0.00018512255185076782, + "loss": 1.0624, + "step": 3967 + }, + { + "epoch": 0.7065527065527065, + "grad_norm": 0.6277863383293152, + "learning_rate": 0.00018511520514247567, + "loss": 1.1056, + "step": 3968 + }, + { + "epoch": 0.7067307692307693, + "grad_norm": 0.580544650554657, + "learning_rate": 0.0001851078567665279, + "loss": 0.9849, + "step": 3969 + }, + { + "epoch": 0.7069088319088319, + "grad_norm": 0.4880999028682709, + "learning_rate": 0.00018510050672306848, + "loss": 1.0185, + "step": 3970 + }, + { + "epoch": 0.7070868945868946, + "grad_norm": 0.4919959306716919, + "learning_rate": 0.0001850931550122414, + "loss": 1.0334, + "step": 3971 + }, + { + "epoch": 0.7072649572649573, + "grad_norm": 0.6001213192939758, + "learning_rate": 0.0001850858016341907, + "loss": 1.0729, + "step": 3972 + }, + { + "epoch": 0.70744301994302, + "grad_norm": 0.538690447807312, + "learning_rate": 0.00018507844658906052, + "loss": 1.0733, + "step": 3973 + }, + { + "epoch": 0.7076210826210826, + "grad_norm": 0.5427643656730652, + "learning_rate": 0.00018507108987699487, + "loss": 1.1207, + "step": 3974 + }, + { + "epoch": 0.7077991452991453, + "grad_norm": 0.43014347553253174, + "learning_rate": 0.00018506373149813795, + "loss": 0.7958, + "step": 3975 + }, + { + "epoch": 0.707977207977208, + "grad_norm": 0.56591796875, + "learning_rate": 0.00018505637145263394, + "loss": 1.2199, + "step": 3976 + }, + { + "epoch": 0.7081552706552706, + "grad_norm": 0.59147047996521, + "learning_rate": 0.000185049009740627, + "loss": 1.2354, + "step": 3977 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 0.5078346133232117, + "learning_rate": 0.00018504164636226137, + "loss": 0.976, + "step": 3978 + }, + { + "epoch": 0.708511396011396, + "grad_norm": 0.533302366733551, + "learning_rate": 0.00018503428131768135, + "loss": 0.9653, + "step": 3979 + }, + { + "epoch": 0.7086894586894587, + "grad_norm": 0.4985341727733612, + "learning_rate": 0.00018502691460703122, + "loss": 1.1485, + "step": 3980 + }, + { + "epoch": 0.7088675213675214, + "grad_norm": 0.5143141150474548, + "learning_rate": 0.00018501954623045532, + "loss": 1.148, + "step": 3981 + }, + { + "epoch": 0.709045584045584, + "grad_norm": 0.507189154624939, + "learning_rate": 0.00018501217618809804, + "loss": 0.9306, + "step": 3982 + }, + { + "epoch": 0.7092236467236467, + "grad_norm": 0.5246604084968567, + "learning_rate": 0.00018500480448010377, + "loss": 0.9116, + "step": 3983 + }, + { + "epoch": 0.7094017094017094, + "grad_norm": 0.5321049094200134, + "learning_rate": 0.00018499743110661693, + "loss": 0.9607, + "step": 3984 + }, + { + "epoch": 0.7095797720797721, + "grad_norm": 0.62645423412323, + "learning_rate": 0.000184990056067782, + "loss": 1.5834, + "step": 3985 + }, + { + "epoch": 0.7097578347578347, + "grad_norm": 0.486557275056839, + "learning_rate": 0.0001849826793637435, + "loss": 1.0598, + "step": 3986 + }, + { + "epoch": 0.7099358974358975, + "grad_norm": 0.5122783184051514, + "learning_rate": 0.0001849753009946459, + "loss": 1.2213, + "step": 3987 + }, + { + "epoch": 0.7101139601139601, + "grad_norm": 0.4864068627357483, + "learning_rate": 0.0001849679209606338, + "loss": 1.2708, + "step": 3988 + }, + { + "epoch": 0.7102920227920227, + "grad_norm": 0.5860990881919861, + "learning_rate": 0.00018496053926185183, + "loss": 1.2421, + "step": 3989 + }, + { + "epoch": 0.7104700854700855, + "grad_norm": 0.471194326877594, + "learning_rate": 0.00018495315589844453, + "loss": 0.879, + "step": 3990 + }, + { + "epoch": 0.7106481481481481, + "grad_norm": 0.5626323819160461, + "learning_rate": 0.00018494577087055662, + "loss": 1.1297, + "step": 3991 + }, + { + "epoch": 0.7108262108262108, + "grad_norm": 0.4706762135028839, + "learning_rate": 0.0001849383841783328, + "loss": 1.0444, + "step": 3992 + }, + { + "epoch": 0.7110042735042735, + "grad_norm": 0.5776444673538208, + "learning_rate": 0.00018493099582191783, + "loss": 1.1773, + "step": 3993 + }, + { + "epoch": 0.7111823361823362, + "grad_norm": 0.5493253469467163, + "learning_rate": 0.00018492360580145637, + "loss": 1.0354, + "step": 3994 + }, + { + "epoch": 0.7113603988603988, + "grad_norm": 0.5328514575958252, + "learning_rate": 0.0001849162141170933, + "loss": 0.9251, + "step": 3995 + }, + { + "epoch": 0.7115384615384616, + "grad_norm": 0.5814893841743469, + "learning_rate": 0.0001849088207689734, + "loss": 1.1066, + "step": 3996 + }, + { + "epoch": 0.7117165242165242, + "grad_norm": 0.5476071834564209, + "learning_rate": 0.00018490142575724154, + "loss": 1.1613, + "step": 3997 + }, + { + "epoch": 0.7118945868945868, + "grad_norm": 0.5216463208198547, + "learning_rate": 0.00018489402908204258, + "loss": 1.2574, + "step": 3998 + }, + { + "epoch": 0.7120726495726496, + "grad_norm": 0.5110020637512207, + "learning_rate": 0.00018488663074352153, + "loss": 1.0663, + "step": 3999 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 0.448090523481369, + "learning_rate": 0.00018487923074182326, + "loss": 0.6687, + "step": 4000 + }, + { + "epoch": 0.7124287749287749, + "grad_norm": 0.4980565011501312, + "learning_rate": 0.00018487182907709279, + "loss": 1.2365, + "step": 4001 + }, + { + "epoch": 0.7126068376068376, + "grad_norm": 0.485831081867218, + "learning_rate": 0.00018486442574947511, + "loss": 1.0941, + "step": 4002 + }, + { + "epoch": 0.7127849002849003, + "grad_norm": 0.4955040216445923, + "learning_rate": 0.00018485702075911534, + "loss": 1.248, + "step": 4003 + }, + { + "epoch": 0.7129629629629629, + "grad_norm": 0.5168375968933105, + "learning_rate": 0.00018484961410615845, + "loss": 1.1118, + "step": 4004 + }, + { + "epoch": 0.7131410256410257, + "grad_norm": 0.5255687832832336, + "learning_rate": 0.00018484220579074968, + "loss": 1.0558, + "step": 4005 + }, + { + "epoch": 0.7133190883190883, + "grad_norm": 0.5502219796180725, + "learning_rate": 0.00018483479581303416, + "loss": 1.1604, + "step": 4006 + }, + { + "epoch": 0.7134971509971509, + "grad_norm": 0.5155881643295288, + "learning_rate": 0.000184827384173157, + "loss": 0.8246, + "step": 4007 + }, + { + "epoch": 0.7136752136752137, + "grad_norm": 0.5321542024612427, + "learning_rate": 0.0001848199708712635, + "loss": 1.2058, + "step": 4008 + }, + { + "epoch": 0.7138532763532763, + "grad_norm": 0.4929848313331604, + "learning_rate": 0.00018481255590749884, + "loss": 1.4023, + "step": 4009 + }, + { + "epoch": 0.7140313390313391, + "grad_norm": 0.5070937871932983, + "learning_rate": 0.00018480513928200836, + "loss": 1.0561, + "step": 4010 + }, + { + "epoch": 0.7142094017094017, + "grad_norm": 0.5750083327293396, + "learning_rate": 0.00018479772099493728, + "loss": 1.0276, + "step": 4011 + }, + { + "epoch": 0.7143874643874644, + "grad_norm": 0.5265933275222778, + "learning_rate": 0.00018479030104643108, + "loss": 1.0295, + "step": 4012 + }, + { + "epoch": 0.7145655270655271, + "grad_norm": 0.526830792427063, + "learning_rate": 0.00018478287943663504, + "loss": 1.0157, + "step": 4013 + }, + { + "epoch": 0.7147435897435898, + "grad_norm": 0.5344091653823853, + "learning_rate": 0.00018477545616569458, + "loss": 1.1997, + "step": 4014 + }, + { + "epoch": 0.7149216524216524, + "grad_norm": 0.4935445189476013, + "learning_rate": 0.0001847680312337552, + "loss": 1.1858, + "step": 4015 + }, + { + "epoch": 0.7150997150997151, + "grad_norm": 0.5291212797164917, + "learning_rate": 0.0001847606046409623, + "loss": 0.926, + "step": 4016 + }, + { + "epoch": 0.7152777777777778, + "grad_norm": 0.559050977230072, + "learning_rate": 0.00018475317638746142, + "loss": 1.0947, + "step": 4017 + }, + { + "epoch": 0.7154558404558404, + "grad_norm": 0.4566570222377777, + "learning_rate": 0.00018474574647339814, + "loss": 1.0334, + "step": 4018 + }, + { + "epoch": 0.7156339031339032, + "grad_norm": 0.5156155824661255, + "learning_rate": 0.000184738314898918, + "loss": 1.0076, + "step": 4019 + }, + { + "epoch": 0.7158119658119658, + "grad_norm": 0.5008716583251953, + "learning_rate": 0.00018473088166416662, + "loss": 1.0378, + "step": 4020 + }, + { + "epoch": 0.7159900284900285, + "grad_norm": 0.49556368589401245, + "learning_rate": 0.0001847234467692896, + "loss": 1.15, + "step": 4021 + }, + { + "epoch": 0.7161680911680912, + "grad_norm": 0.5464680790901184, + "learning_rate": 0.00018471601021443265, + "loss": 1.2975, + "step": 4022 + }, + { + "epoch": 0.7163461538461539, + "grad_norm": 0.6291980147361755, + "learning_rate": 0.00018470857199974144, + "loss": 1.05, + "step": 4023 + }, + { + "epoch": 0.7165242165242165, + "grad_norm": 0.5566631555557251, + "learning_rate": 0.00018470113212536176, + "loss": 1.1296, + "step": 4024 + }, + { + "epoch": 0.7167022792022792, + "grad_norm": 0.5569562911987305, + "learning_rate": 0.00018469369059143933, + "loss": 1.2484, + "step": 4025 + }, + { + "epoch": 0.7168803418803419, + "grad_norm": 0.5804716944694519, + "learning_rate": 0.00018468624739812, + "loss": 1.0547, + "step": 4026 + }, + { + "epoch": 0.7170584045584045, + "grad_norm": 0.6316802501678467, + "learning_rate": 0.00018467880254554952, + "loss": 1.1188, + "step": 4027 + }, + { + "epoch": 0.7172364672364673, + "grad_norm": 0.6131419539451599, + "learning_rate": 0.00018467135603387385, + "loss": 1.1662, + "step": 4028 + }, + { + "epoch": 0.7174145299145299, + "grad_norm": 0.4703124761581421, + "learning_rate": 0.00018466390786323883, + "loss": 1.038, + "step": 4029 + }, + { + "epoch": 0.7175925925925926, + "grad_norm": 0.5718469023704529, + "learning_rate": 0.0001846564580337904, + "loss": 1.0786, + "step": 4030 + }, + { + "epoch": 0.7177706552706553, + "grad_norm": 0.5227612853050232, + "learning_rate": 0.00018464900654567457, + "loss": 1.0561, + "step": 4031 + }, + { + "epoch": 0.717948717948718, + "grad_norm": 0.5800358057022095, + "learning_rate": 0.00018464155339903727, + "loss": 1.0944, + "step": 4032 + }, + { + "epoch": 0.7181267806267806, + "grad_norm": 0.5562314987182617, + "learning_rate": 0.00018463409859402455, + "loss": 0.8573, + "step": 4033 + }, + { + "epoch": 0.7183048433048433, + "grad_norm": 0.6420153379440308, + "learning_rate": 0.0001846266421307825, + "loss": 1.088, + "step": 4034 + }, + { + "epoch": 0.718482905982906, + "grad_norm": 0.4745902717113495, + "learning_rate": 0.00018461918400945718, + "loss": 1.1679, + "step": 4035 + }, + { + "epoch": 0.7186609686609686, + "grad_norm": 0.5070300102233887, + "learning_rate": 0.00018461172423019475, + "loss": 1.1984, + "step": 4036 + }, + { + "epoch": 0.7188390313390314, + "grad_norm": 0.5339375138282776, + "learning_rate": 0.00018460426279314133, + "loss": 1.3038, + "step": 4037 + }, + { + "epoch": 0.719017094017094, + "grad_norm": 0.5947147607803345, + "learning_rate": 0.00018459679969844313, + "loss": 1.0103, + "step": 4038 + }, + { + "epoch": 0.7191951566951567, + "grad_norm": 0.5493791699409485, + "learning_rate": 0.00018458933494624642, + "loss": 1.1001, + "step": 4039 + }, + { + "epoch": 0.7193732193732194, + "grad_norm": 0.5700310468673706, + "learning_rate": 0.00018458186853669736, + "loss": 0.9006, + "step": 4040 + }, + { + "epoch": 0.719551282051282, + "grad_norm": 0.60371994972229, + "learning_rate": 0.0001845744004699423, + "loss": 1.3001, + "step": 4041 + }, + { + "epoch": 0.7197293447293447, + "grad_norm": 0.5469261407852173, + "learning_rate": 0.00018456693074612757, + "loss": 1.1745, + "step": 4042 + }, + { + "epoch": 0.7199074074074074, + "grad_norm": 0.5179165601730347, + "learning_rate": 0.00018455945936539947, + "loss": 0.9883, + "step": 4043 + }, + { + "epoch": 0.7200854700854701, + "grad_norm": 0.5396696329116821, + "learning_rate": 0.00018455198632790447, + "loss": 1.1277, + "step": 4044 + }, + { + "epoch": 0.7202635327635327, + "grad_norm": 0.4559909403324127, + "learning_rate": 0.00018454451163378888, + "loss": 0.9644, + "step": 4045 + }, + { + "epoch": 0.7204415954415955, + "grad_norm": 0.49863892793655396, + "learning_rate": 0.00018453703528319927, + "loss": 1.1276, + "step": 4046 + }, + { + "epoch": 0.7206196581196581, + "grad_norm": 0.4790710508823395, + "learning_rate": 0.000184529557276282, + "loss": 0.9443, + "step": 4047 + }, + { + "epoch": 0.7207977207977208, + "grad_norm": 0.541999876499176, + "learning_rate": 0.0001845220776131837, + "loss": 1.0681, + "step": 4048 + }, + { + "epoch": 0.7209757834757835, + "grad_norm": 0.5119109153747559, + "learning_rate": 0.00018451459629405088, + "loss": 1.2078, + "step": 4049 + }, + { + "epoch": 0.7211538461538461, + "grad_norm": 0.6141307353973389, + "learning_rate": 0.00018450711331903006, + "loss": 1.1071, + "step": 4050 + }, + { + "epoch": 0.7213319088319088, + "grad_norm": 0.48679864406585693, + "learning_rate": 0.00018449962868826795, + "loss": 0.9713, + "step": 4051 + }, + { + "epoch": 0.7215099715099715, + "grad_norm": 0.5548661947250366, + "learning_rate": 0.0001844921424019111, + "loss": 1.2099, + "step": 4052 + }, + { + "epoch": 0.7216880341880342, + "grad_norm": 0.5000107884407043, + "learning_rate": 0.00018448465446010626, + "loss": 1.0184, + "step": 4053 + }, + { + "epoch": 0.7218660968660968, + "grad_norm": 0.6131454110145569, + "learning_rate": 0.00018447716486300013, + "loss": 1.2581, + "step": 4054 + }, + { + "epoch": 0.7220441595441596, + "grad_norm": 0.5145987868309021, + "learning_rate": 0.0001844696736107394, + "loss": 1.1646, + "step": 4055 + }, + { + "epoch": 0.7222222222222222, + "grad_norm": 0.4361337125301361, + "learning_rate": 0.00018446218070347094, + "loss": 0.8239, + "step": 4056 + }, + { + "epoch": 0.7224002849002849, + "grad_norm": 0.5549173355102539, + "learning_rate": 0.00018445468614134146, + "loss": 1.1935, + "step": 4057 + }, + { + "epoch": 0.7225783475783476, + "grad_norm": 0.5569297671318054, + "learning_rate": 0.00018444718992449789, + "loss": 1.0137, + "step": 4058 + }, + { + "epoch": 0.7227564102564102, + "grad_norm": 0.44866305589675903, + "learning_rate": 0.00018443969205308704, + "loss": 0.987, + "step": 4059 + }, + { + "epoch": 0.7229344729344729, + "grad_norm": 0.5142943263053894, + "learning_rate": 0.0001844321925272558, + "loss": 1.0837, + "step": 4060 + }, + { + "epoch": 0.7231125356125356, + "grad_norm": 0.4922119379043579, + "learning_rate": 0.0001844246913471512, + "loss": 0.8477, + "step": 4061 + }, + { + "epoch": 0.7232905982905983, + "grad_norm": 0.5245375633239746, + "learning_rate": 0.0001844171885129201, + "loss": 0.9985, + "step": 4062 + }, + { + "epoch": 0.7234686609686609, + "grad_norm": 0.45562678575515747, + "learning_rate": 0.00018440968402470956, + "loss": 0.8678, + "step": 4063 + }, + { + "epoch": 0.7236467236467237, + "grad_norm": 0.5388376712799072, + "learning_rate": 0.0001844021778826666, + "loss": 1.0586, + "step": 4064 + }, + { + "epoch": 0.7238247863247863, + "grad_norm": 0.48945263028144836, + "learning_rate": 0.00018439467008693833, + "loss": 1.0547, + "step": 4065 + }, + { + "epoch": 0.7240028490028491, + "grad_norm": 0.5202330350875854, + "learning_rate": 0.00018438716063767178, + "loss": 1.3142, + "step": 4066 + }, + { + "epoch": 0.7241809116809117, + "grad_norm": 0.5432567000389099, + "learning_rate": 0.00018437964953501413, + "loss": 1.0192, + "step": 4067 + }, + { + "epoch": 0.7243589743589743, + "grad_norm": 0.5220325589179993, + "learning_rate": 0.00018437213677911253, + "loss": 1.0904, + "step": 4068 + }, + { + "epoch": 0.7245370370370371, + "grad_norm": 0.45711690187454224, + "learning_rate": 0.00018436462237011417, + "loss": 1.0417, + "step": 4069 + }, + { + "epoch": 0.7247150997150997, + "grad_norm": 0.560778021812439, + "learning_rate": 0.0001843571063081663, + "loss": 1.2316, + "step": 4070 + }, + { + "epoch": 0.7248931623931624, + "grad_norm": 0.591533362865448, + "learning_rate": 0.0001843495885934162, + "loss": 1.0294, + "step": 4071 + }, + { + "epoch": 0.7250712250712251, + "grad_norm": 0.5550443530082703, + "learning_rate": 0.00018434206922601106, + "loss": 1.0162, + "step": 4072 + }, + { + "epoch": 0.7252492877492878, + "grad_norm": 0.5744053721427917, + "learning_rate": 0.00018433454820609833, + "loss": 1.2774, + "step": 4073 + }, + { + "epoch": 0.7254273504273504, + "grad_norm": 0.6210703253746033, + "learning_rate": 0.0001843270255338253, + "loss": 1.2526, + "step": 4074 + }, + { + "epoch": 0.7256054131054132, + "grad_norm": 0.49684277176856995, + "learning_rate": 0.0001843195012093394, + "loss": 1.0786, + "step": 4075 + }, + { + "epoch": 0.7257834757834758, + "grad_norm": 0.5851606130599976, + "learning_rate": 0.00018431197523278802, + "loss": 1.14, + "step": 4076 + }, + { + "epoch": 0.7259615384615384, + "grad_norm": 0.5494425296783447, + "learning_rate": 0.00018430444760431862, + "loss": 1.211, + "step": 4077 + }, + { + "epoch": 0.7261396011396012, + "grad_norm": 0.5247658491134644, + "learning_rate": 0.00018429691832407867, + "loss": 0.8031, + "step": 4078 + }, + { + "epoch": 0.7263176638176638, + "grad_norm": 0.5012249946594238, + "learning_rate": 0.00018428938739221574, + "loss": 1.1258, + "step": 4079 + }, + { + "epoch": 0.7264957264957265, + "grad_norm": 0.5226427912712097, + "learning_rate": 0.0001842818548088774, + "loss": 1.0029, + "step": 4080 + }, + { + "epoch": 0.7266737891737892, + "grad_norm": 0.45008543133735657, + "learning_rate": 0.00018427432057421114, + "loss": 1.0681, + "step": 4081 + }, + { + "epoch": 0.7268518518518519, + "grad_norm": 0.5127285122871399, + "learning_rate": 0.00018426678468836467, + "loss": 1.1069, + "step": 4082 + }, + { + "epoch": 0.7270299145299145, + "grad_norm": 0.5406150221824646, + "learning_rate": 0.0001842592471514856, + "loss": 1.052, + "step": 4083 + }, + { + "epoch": 0.7272079772079773, + "grad_norm": 0.5001157522201538, + "learning_rate": 0.0001842517079637216, + "loss": 0.9157, + "step": 4084 + }, + { + "epoch": 0.7273860398860399, + "grad_norm": 0.6169779300689697, + "learning_rate": 0.00018424416712522042, + "loss": 1.3133, + "step": 4085 + }, + { + "epoch": 0.7275641025641025, + "grad_norm": 0.4891316890716553, + "learning_rate": 0.00018423662463612974, + "loss": 0.9505, + "step": 4086 + }, + { + "epoch": 0.7277421652421653, + "grad_norm": 0.5883708596229553, + "learning_rate": 0.00018422908049659743, + "loss": 1.2797, + "step": 4087 + }, + { + "epoch": 0.7279202279202279, + "grad_norm": 0.6679072976112366, + "learning_rate": 0.00018422153470677125, + "loss": 1.1096, + "step": 4088 + }, + { + "epoch": 0.7280982905982906, + "grad_norm": 0.5178479552268982, + "learning_rate": 0.00018421398726679904, + "loss": 1.0299, + "step": 4089 + }, + { + "epoch": 0.7282763532763533, + "grad_norm": 0.6343900561332703, + "learning_rate": 0.0001842064381768287, + "loss": 1.2983, + "step": 4090 + }, + { + "epoch": 0.728454415954416, + "grad_norm": 0.43816515803337097, + "learning_rate": 0.0001841988874370081, + "loss": 0.9452, + "step": 4091 + }, + { + "epoch": 0.7286324786324786, + "grad_norm": 0.579790472984314, + "learning_rate": 0.00018419133504748528, + "loss": 1.1037, + "step": 4092 + }, + { + "epoch": 0.7288105413105413, + "grad_norm": 0.571374773979187, + "learning_rate": 0.00018418378100840807, + "loss": 1.1655, + "step": 4093 + }, + { + "epoch": 0.728988603988604, + "grad_norm": 0.5163514018058777, + "learning_rate": 0.0001841762253199246, + "loss": 1.1579, + "step": 4094 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 0.6553022265434265, + "learning_rate": 0.0001841686679821828, + "loss": 0.9664, + "step": 4095 + }, + { + "epoch": 0.7293447293447294, + "grad_norm": 0.5072969198226929, + "learning_rate": 0.00018416110899533084, + "loss": 0.9416, + "step": 4096 + }, + { + "epoch": 0.729522792022792, + "grad_norm": 0.5103251338005066, + "learning_rate": 0.00018415354835951675, + "loss": 1.0715, + "step": 4097 + }, + { + "epoch": 0.7297008547008547, + "grad_norm": 0.49752289056777954, + "learning_rate": 0.00018414598607488874, + "loss": 1.1848, + "step": 4098 + }, + { + "epoch": 0.7298789173789174, + "grad_norm": 0.5361882448196411, + "learning_rate": 0.00018413842214159488, + "loss": 1.1035, + "step": 4099 + }, + { + "epoch": 0.73005698005698, + "grad_norm": 0.5167670249938965, + "learning_rate": 0.00018413085655978343, + "loss": 1.0015, + "step": 4100 + }, + { + "epoch": 0.7302350427350427, + "grad_norm": 0.5930629372596741, + "learning_rate": 0.00018412328932960263, + "loss": 0.9766, + "step": 4101 + }, + { + "epoch": 0.7304131054131054, + "grad_norm": 0.5234778523445129, + "learning_rate": 0.00018411572045120073, + "loss": 1.0317, + "step": 4102 + }, + { + "epoch": 0.7305911680911681, + "grad_norm": 0.5361374020576477, + "learning_rate": 0.000184108149924726, + "loss": 1.1228, + "step": 4103 + }, + { + "epoch": 0.7307692307692307, + "grad_norm": 0.5845770239830017, + "learning_rate": 0.0001841005777503268, + "loss": 0.9541, + "step": 4104 + }, + { + "epoch": 0.7309472934472935, + "grad_norm": 0.49320483207702637, + "learning_rate": 0.0001840930039281515, + "loss": 0.9445, + "step": 4105 + }, + { + "epoch": 0.7311253561253561, + "grad_norm": 0.5391250252723694, + "learning_rate": 0.00018408542845834845, + "loss": 1.1983, + "step": 4106 + }, + { + "epoch": 0.7313034188034188, + "grad_norm": 0.4890393316745758, + "learning_rate": 0.00018407785134106613, + "loss": 0.8353, + "step": 4107 + }, + { + "epoch": 0.7314814814814815, + "grad_norm": 0.5839747190475464, + "learning_rate": 0.00018407027257645296, + "loss": 1.4074, + "step": 4108 + }, + { + "epoch": 0.7316595441595442, + "grad_norm": 0.5957708358764648, + "learning_rate": 0.0001840626921646574, + "loss": 1.1032, + "step": 4109 + }, + { + "epoch": 0.7318376068376068, + "grad_norm": 0.5029017925262451, + "learning_rate": 0.00018405511010582805, + "loss": 1.095, + "step": 4110 + }, + { + "epoch": 0.7320156695156695, + "grad_norm": 0.6054347157478333, + "learning_rate": 0.00018404752640011345, + "loss": 1.0366, + "step": 4111 + }, + { + "epoch": 0.7321937321937322, + "grad_norm": 0.5476830005645752, + "learning_rate": 0.00018403994104766212, + "loss": 1.0976, + "step": 4112 + }, + { + "epoch": 0.7323717948717948, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00018403235404862277, + "loss": 1.0809, + "step": 4113 + }, + { + "epoch": 0.7325498575498576, + "grad_norm": 0.5119251012802124, + "learning_rate": 0.00018402476540314394, + "loss": 1.0176, + "step": 4114 + }, + { + "epoch": 0.7327279202279202, + "grad_norm": 0.5825830698013306, + "learning_rate": 0.00018401717511137445, + "loss": 1.2357, + "step": 4115 + }, + { + "epoch": 0.7329059829059829, + "grad_norm": 0.5702941417694092, + "learning_rate": 0.0001840095831734629, + "loss": 1.1549, + "step": 4116 + }, + { + "epoch": 0.7330840455840456, + "grad_norm": 0.5660699605941772, + "learning_rate": 0.00018400198958955807, + "loss": 1.1778, + "step": 4117 + }, + { + "epoch": 0.7332621082621082, + "grad_norm": 0.5241161584854126, + "learning_rate": 0.0001839943943598088, + "loss": 0.8587, + "step": 4118 + }, + { + "epoch": 0.7334401709401709, + "grad_norm": 0.581194281578064, + "learning_rate": 0.0001839867974843638, + "loss": 1.2169, + "step": 4119 + }, + { + "epoch": 0.7336182336182336, + "grad_norm": 0.4342379570007324, + "learning_rate": 0.00018397919896337198, + "loss": 0.9182, + "step": 4120 + }, + { + "epoch": 0.7337962962962963, + "grad_norm": 0.5708567500114441, + "learning_rate": 0.00018397159879698224, + "loss": 1.1781, + "step": 4121 + }, + { + "epoch": 0.7339743589743589, + "grad_norm": 0.5827265977859497, + "learning_rate": 0.00018396399698534344, + "loss": 1.2905, + "step": 4122 + }, + { + "epoch": 0.7341524216524217, + "grad_norm": 0.5274056792259216, + "learning_rate": 0.00018395639352860457, + "loss": 1.1786, + "step": 4123 + }, + { + "epoch": 0.7343304843304843, + "grad_norm": 0.5094266533851624, + "learning_rate": 0.00018394878842691452, + "loss": 1.2016, + "step": 4124 + }, + { + "epoch": 0.7345085470085471, + "grad_norm": 0.48779475688934326, + "learning_rate": 0.0001839411816804224, + "loss": 1.0562, + "step": 4125 + }, + { + "epoch": 0.7346866096866097, + "grad_norm": 0.5805709958076477, + "learning_rate": 0.00018393357328927716, + "loss": 1.1705, + "step": 4126 + }, + { + "epoch": 0.7348646723646723, + "grad_norm": 0.4910700023174286, + "learning_rate": 0.00018392596325362791, + "loss": 1.0682, + "step": 4127 + }, + { + "epoch": 0.7350427350427351, + "grad_norm": 0.5297428369522095, + "learning_rate": 0.0001839183515736238, + "loss": 0.9505, + "step": 4128 + }, + { + "epoch": 0.7352207977207977, + "grad_norm": 0.45442086458206177, + "learning_rate": 0.00018391073824941385, + "loss": 0.9548, + "step": 4129 + }, + { + "epoch": 0.7353988603988604, + "grad_norm": 0.49299946427345276, + "learning_rate": 0.00018390312328114733, + "loss": 1.0868, + "step": 4130 + }, + { + "epoch": 0.7355769230769231, + "grad_norm": 0.4839940369129181, + "learning_rate": 0.0001838955066689734, + "loss": 0.9565, + "step": 4131 + }, + { + "epoch": 0.7357549857549858, + "grad_norm": 0.48600608110427856, + "learning_rate": 0.00018388788841304128, + "loss": 1.2353, + "step": 4132 + }, + { + "epoch": 0.7359330484330484, + "grad_norm": 0.4893583357334137, + "learning_rate": 0.0001838802685135003, + "loss": 0.9595, + "step": 4133 + }, + { + "epoch": 0.7361111111111112, + "grad_norm": 0.4587398171424866, + "learning_rate": 0.00018387264697049963, + "loss": 1.1222, + "step": 4134 + }, + { + "epoch": 0.7362891737891738, + "grad_norm": 0.5361055731773376, + "learning_rate": 0.00018386502378418872, + "loss": 1.3304, + "step": 4135 + }, + { + "epoch": 0.7364672364672364, + "grad_norm": 0.5556629300117493, + "learning_rate": 0.00018385739895471686, + "loss": 1.0358, + "step": 4136 + }, + { + "epoch": 0.7366452991452992, + "grad_norm": 0.45555856823921204, + "learning_rate": 0.00018384977248223346, + "loss": 1.0081, + "step": 4137 + }, + { + "epoch": 0.7368233618233618, + "grad_norm": 0.5606052875518799, + "learning_rate": 0.00018384214436688797, + "loss": 0.9367, + "step": 4138 + }, + { + "epoch": 0.7370014245014245, + "grad_norm": 0.5428356528282166, + "learning_rate": 0.00018383451460882982, + "loss": 1.1391, + "step": 4139 + }, + { + "epoch": 0.7371794871794872, + "grad_norm": 0.4891330897808075, + "learning_rate": 0.00018382688320820853, + "loss": 0.9805, + "step": 4140 + }, + { + "epoch": 0.7373575498575499, + "grad_norm": 0.5407996773719788, + "learning_rate": 0.0001838192501651736, + "loss": 1.0532, + "step": 4141 + }, + { + "epoch": 0.7375356125356125, + "grad_norm": 0.5241971611976624, + "learning_rate": 0.00018381161547987454, + "loss": 0.9509, + "step": 4142 + }, + { + "epoch": 0.7377136752136753, + "grad_norm": 0.5370210409164429, + "learning_rate": 0.000183803979152461, + "loss": 1.2342, + "step": 4143 + }, + { + "epoch": 0.7378917378917379, + "grad_norm": 0.5470060706138611, + "learning_rate": 0.00018379634118308259, + "loss": 0.9621, + "step": 4144 + }, + { + "epoch": 0.7380698005698005, + "grad_norm": 0.546313464641571, + "learning_rate": 0.00018378870157188893, + "loss": 1.1253, + "step": 4145 + }, + { + "epoch": 0.7382478632478633, + "grad_norm": 0.502027153968811, + "learning_rate": 0.00018378106031902974, + "loss": 1.1919, + "step": 4146 + }, + { + "epoch": 0.7384259259259259, + "grad_norm": 0.5282283425331116, + "learning_rate": 0.0001837734174246547, + "loss": 1.0088, + "step": 4147 + }, + { + "epoch": 0.7386039886039886, + "grad_norm": 0.5152897238731384, + "learning_rate": 0.00018376577288891355, + "loss": 1.0813, + "step": 4148 + }, + { + "epoch": 0.7387820512820513, + "grad_norm": 0.5002804398536682, + "learning_rate": 0.0001837581267119561, + "loss": 0.9797, + "step": 4149 + }, + { + "epoch": 0.738960113960114, + "grad_norm": 0.5698176026344299, + "learning_rate": 0.00018375047889393215, + "loss": 1.1099, + "step": 4150 + }, + { + "epoch": 0.7391381766381766, + "grad_norm": 0.5384604930877686, + "learning_rate": 0.00018374282943499156, + "loss": 1.1944, + "step": 4151 + }, + { + "epoch": 0.7393162393162394, + "grad_norm": 0.5483044385910034, + "learning_rate": 0.00018373517833528418, + "loss": 1.1734, + "step": 4152 + }, + { + "epoch": 0.739494301994302, + "grad_norm": 0.4824066162109375, + "learning_rate": 0.0001837275255949599, + "loss": 0.9515, + "step": 4153 + }, + { + "epoch": 0.7396723646723646, + "grad_norm": 0.45413634181022644, + "learning_rate": 0.00018371987121416873, + "loss": 0.7534, + "step": 4154 + }, + { + "epoch": 0.7398504273504274, + "grad_norm": 0.5874246954917908, + "learning_rate": 0.00018371221519306055, + "loss": 0.9464, + "step": 4155 + }, + { + "epoch": 0.74002849002849, + "grad_norm": 0.5219913125038147, + "learning_rate": 0.00018370455753178544, + "loss": 1.0494, + "step": 4156 + }, + { + "epoch": 0.7402065527065527, + "grad_norm": 0.5937709212303162, + "learning_rate": 0.00018369689823049341, + "loss": 1.0529, + "step": 4157 + }, + { + "epoch": 0.7403846153846154, + "grad_norm": 0.5204295516014099, + "learning_rate": 0.00018368923728933449, + "loss": 1.0602, + "step": 4158 + }, + { + "epoch": 0.7405626780626781, + "grad_norm": 0.5422890186309814, + "learning_rate": 0.00018368157470845885, + "loss": 0.9261, + "step": 4159 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.6163852214813232, + "learning_rate": 0.00018367391048801655, + "loss": 1.2771, + "step": 4160 + }, + { + "epoch": 0.7409188034188035, + "grad_norm": 0.5070751309394836, + "learning_rate": 0.00018366624462815785, + "loss": 1.0401, + "step": 4161 + }, + { + "epoch": 0.7410968660968661, + "grad_norm": 0.4477100968360901, + "learning_rate": 0.00018365857712903283, + "loss": 1.1463, + "step": 4162 + }, + { + "epoch": 0.7412749287749287, + "grad_norm": 0.5421462655067444, + "learning_rate": 0.0001836509079907918, + "loss": 0.9373, + "step": 4163 + }, + { + "epoch": 0.7414529914529915, + "grad_norm": 0.6162141561508179, + "learning_rate": 0.000183643237213585, + "loss": 1.1827, + "step": 4164 + }, + { + "epoch": 0.7416310541310541, + "grad_norm": 0.5653836131095886, + "learning_rate": 0.00018363556479756272, + "loss": 1.0689, + "step": 4165 + }, + { + "epoch": 0.7418091168091168, + "grad_norm": 0.57053542137146, + "learning_rate": 0.00018362789074287527, + "loss": 1.0289, + "step": 4166 + }, + { + "epoch": 0.7419871794871795, + "grad_norm": 0.5603055953979492, + "learning_rate": 0.00018362021504967304, + "loss": 1.1926, + "step": 4167 + }, + { + "epoch": 0.7421652421652422, + "grad_norm": 0.5460166335105896, + "learning_rate": 0.0001836125377181064, + "loss": 1.1488, + "step": 4168 + }, + { + "epoch": 0.7423433048433048, + "grad_norm": 0.5097107887268066, + "learning_rate": 0.00018360485874832579, + "loss": 1.0781, + "step": 4169 + }, + { + "epoch": 0.7425213675213675, + "grad_norm": 0.6280624270439148, + "learning_rate": 0.00018359717814048164, + "loss": 1.3625, + "step": 4170 + }, + { + "epoch": 0.7426994301994302, + "grad_norm": 0.4528210759162903, + "learning_rate": 0.0001835894958947244, + "loss": 0.8417, + "step": 4171 + }, + { + "epoch": 0.7428774928774928, + "grad_norm": 0.48735132813453674, + "learning_rate": 0.00018358181201120468, + "loss": 0.9544, + "step": 4172 + }, + { + "epoch": 0.7430555555555556, + "grad_norm": 0.48388174176216125, + "learning_rate": 0.00018357412649007296, + "loss": 1.0663, + "step": 4173 + }, + { + "epoch": 0.7432336182336182, + "grad_norm": 0.5435357689857483, + "learning_rate": 0.00018356643933147986, + "loss": 1.2074, + "step": 4174 + }, + { + "epoch": 0.7434116809116809, + "grad_norm": 0.49890074133872986, + "learning_rate": 0.00018355875053557594, + "loss": 1.1322, + "step": 4175 + }, + { + "epoch": 0.7435897435897436, + "grad_norm": 0.5680708885192871, + "learning_rate": 0.0001835510601025119, + "loss": 1.1964, + "step": 4176 + }, + { + "epoch": 0.7437678062678063, + "grad_norm": 0.5002360939979553, + "learning_rate": 0.00018354336803243842, + "loss": 1.1396, + "step": 4177 + }, + { + "epoch": 0.7439458689458689, + "grad_norm": 0.5202965140342712, + "learning_rate": 0.00018353567432550616, + "loss": 1.1498, + "step": 4178 + }, + { + "epoch": 0.7441239316239316, + "grad_norm": 0.514492928981781, + "learning_rate": 0.00018352797898186588, + "loss": 1.0959, + "step": 4179 + }, + { + "epoch": 0.7443019943019943, + "grad_norm": 0.6395383477210999, + "learning_rate": 0.0001835202820016684, + "loss": 1.2867, + "step": 4180 + }, + { + "epoch": 0.7444800569800569, + "grad_norm": 0.5489062070846558, + "learning_rate": 0.00018351258338506447, + "loss": 1.1638, + "step": 4181 + }, + { + "epoch": 0.7446581196581197, + "grad_norm": 0.5705671906471252, + "learning_rate": 0.00018350488313220498, + "loss": 0.9493, + "step": 4182 + }, + { + "epoch": 0.7448361823361823, + "grad_norm": 0.5404297709465027, + "learning_rate": 0.00018349718124324076, + "loss": 0.9876, + "step": 4183 + }, + { + "epoch": 0.7450142450142451, + "grad_norm": 0.5841003060340881, + "learning_rate": 0.0001834894777183227, + "loss": 1.1225, + "step": 4184 + }, + { + "epoch": 0.7451923076923077, + "grad_norm": 0.49774688482284546, + "learning_rate": 0.00018348177255760178, + "loss": 1.1442, + "step": 4185 + }, + { + "epoch": 0.7453703703703703, + "grad_norm": 0.5212422609329224, + "learning_rate": 0.00018347406576122894, + "loss": 1.101, + "step": 4186 + }, + { + "epoch": 0.7455484330484331, + "grad_norm": 0.615024983882904, + "learning_rate": 0.00018346635732935517, + "loss": 1.4188, + "step": 4187 + }, + { + "epoch": 0.7457264957264957, + "grad_norm": 0.46818843483924866, + "learning_rate": 0.00018345864726213154, + "loss": 1.0071, + "step": 4188 + }, + { + "epoch": 0.7459045584045584, + "grad_norm": 0.4921121895313263, + "learning_rate": 0.00018345093555970906, + "loss": 1.015, + "step": 4189 + }, + { + "epoch": 0.7460826210826211, + "grad_norm": 0.5042136311531067, + "learning_rate": 0.00018344322222223889, + "loss": 0.9974, + "step": 4190 + }, + { + "epoch": 0.7462606837606838, + "grad_norm": 0.5872490406036377, + "learning_rate": 0.0001834355072498721, + "loss": 1.3166, + "step": 4191 + }, + { + "epoch": 0.7464387464387464, + "grad_norm": 0.559117317199707, + "learning_rate": 0.00018342779064275984, + "loss": 1.2227, + "step": 4192 + }, + { + "epoch": 0.7466168091168092, + "grad_norm": 0.5269635319709778, + "learning_rate": 0.00018342007240105336, + "loss": 1.0281, + "step": 4193 + }, + { + "epoch": 0.7467948717948718, + "grad_norm": 0.4608335793018341, + "learning_rate": 0.00018341235252490387, + "loss": 0.98, + "step": 4194 + }, + { + "epoch": 0.7469729344729344, + "grad_norm": 0.5818259119987488, + "learning_rate": 0.00018340463101446255, + "loss": 1.1544, + "step": 4195 + }, + { + "epoch": 0.7471509971509972, + "grad_norm": 0.5577529668807983, + "learning_rate": 0.00018339690786988079, + "loss": 1.3059, + "step": 4196 + }, + { + "epoch": 0.7473290598290598, + "grad_norm": 0.5430468320846558, + "learning_rate": 0.00018338918309130983, + "loss": 1.2766, + "step": 4197 + }, + { + "epoch": 0.7475071225071225, + "grad_norm": 0.4941701591014862, + "learning_rate": 0.0001833814566789011, + "loss": 1.193, + "step": 4198 + }, + { + "epoch": 0.7476851851851852, + "grad_norm": 0.5471884608268738, + "learning_rate": 0.00018337372863280589, + "loss": 1.2261, + "step": 4199 + }, + { + "epoch": 0.7478632478632479, + "grad_norm": 0.4641438126564026, + "learning_rate": 0.0001833659989531757, + "loss": 0.7953, + "step": 4200 + }, + { + "epoch": 0.7480413105413105, + "grad_norm": 0.5244714617729187, + "learning_rate": 0.0001833582676401619, + "loss": 0.9344, + "step": 4201 + }, + { + "epoch": 0.7482193732193733, + "grad_norm": 0.5964360237121582, + "learning_rate": 0.00018335053469391603, + "loss": 1.2072, + "step": 4202 + }, + { + "epoch": 0.7483974358974359, + "grad_norm": 0.4929158091545105, + "learning_rate": 0.00018334280011458954, + "loss": 1.2183, + "step": 4203 + }, + { + "epoch": 0.7485754985754985, + "grad_norm": 0.46221864223480225, + "learning_rate": 0.00018333506390233405, + "loss": 1.1957, + "step": 4204 + }, + { + "epoch": 0.7487535612535613, + "grad_norm": 0.6301732659339905, + "learning_rate": 0.0001833273260573011, + "loss": 1.0582, + "step": 4205 + }, + { + "epoch": 0.7489316239316239, + "grad_norm": 0.5606021881103516, + "learning_rate": 0.0001833195865796423, + "loss": 1.4034, + "step": 4206 + }, + { + "epoch": 0.7491096866096866, + "grad_norm": 0.44856077432632446, + "learning_rate": 0.00018331184546950926, + "loss": 0.8421, + "step": 4207 + }, + { + "epoch": 0.7492877492877493, + "grad_norm": 0.5487226247787476, + "learning_rate": 0.00018330410272705366, + "loss": 1.238, + "step": 4208 + }, + { + "epoch": 0.749465811965812, + "grad_norm": 0.6043636798858643, + "learning_rate": 0.00018329635835242724, + "loss": 1.1215, + "step": 4209 + }, + { + "epoch": 0.7496438746438746, + "grad_norm": 0.5145319104194641, + "learning_rate": 0.00018328861234578173, + "loss": 1.1002, + "step": 4210 + }, + { + "epoch": 0.7498219373219374, + "grad_norm": 0.5667078495025635, + "learning_rate": 0.00018328086470726884, + "loss": 1.2994, + "step": 4211 + }, + { + "epoch": 0.75, + "grad_norm": 0.5117634534835815, + "learning_rate": 0.00018327311543704043, + "loss": 0.9448, + "step": 4212 + }, + { + "epoch": 0.75, + "eval_loss": 1.0982474088668823, + "eval_runtime": 24.6617, + "eval_samples_per_second": 42.211, + "eval_steps_per_second": 21.126, + "step": 4212 + }, + { + "epoch": 0.7501780626780626, + "grad_norm": 0.5451585054397583, + "learning_rate": 0.00018326536453524826, + "loss": 0.9023, + "step": 4213 + }, + { + "epoch": 0.7503561253561254, + "grad_norm": 0.6585208773612976, + "learning_rate": 0.0001832576120020443, + "loss": 1.2798, + "step": 4214 + }, + { + "epoch": 0.750534188034188, + "grad_norm": 0.6444812417030334, + "learning_rate": 0.00018324985783758037, + "loss": 1.3999, + "step": 4215 + }, + { + "epoch": 0.7507122507122507, + "grad_norm": 0.6178330779075623, + "learning_rate": 0.0001832421020420084, + "loss": 1.1846, + "step": 4216 + }, + { + "epoch": 0.7508903133903134, + "grad_norm": 0.509969174861908, + "learning_rate": 0.00018323434461548036, + "loss": 1.1831, + "step": 4217 + }, + { + "epoch": 0.7510683760683761, + "grad_norm": 0.5558911561965942, + "learning_rate": 0.00018322658555814826, + "loss": 1.1599, + "step": 4218 + }, + { + "epoch": 0.7512464387464387, + "grad_norm": 0.5714917778968811, + "learning_rate": 0.0001832188248701641, + "loss": 0.9702, + "step": 4219 + }, + { + "epoch": 0.7514245014245015, + "grad_norm": 0.6136442422866821, + "learning_rate": 0.00018321106255167995, + "loss": 0.9376, + "step": 4220 + }, + { + "epoch": 0.7516025641025641, + "grad_norm": 0.5832077264785767, + "learning_rate": 0.00018320329860284785, + "loss": 1.2564, + "step": 4221 + }, + { + "epoch": 0.7517806267806267, + "grad_norm": 0.45330923795700073, + "learning_rate": 0.00018319553302381997, + "loss": 0.9321, + "step": 4222 + }, + { + "epoch": 0.7519586894586895, + "grad_norm": 0.5278468132019043, + "learning_rate": 0.00018318776581474847, + "loss": 1.1334, + "step": 4223 + }, + { + "epoch": 0.7521367521367521, + "grad_norm": 0.49267473816871643, + "learning_rate": 0.00018317999697578549, + "loss": 1.1577, + "step": 4224 + }, + { + "epoch": 0.7523148148148148, + "grad_norm": 0.5372124314308167, + "learning_rate": 0.00018317222650708325, + "loss": 1.037, + "step": 4225 + }, + { + "epoch": 0.7524928774928775, + "grad_norm": 0.5879829525947571, + "learning_rate": 0.000183164454408794, + "loss": 1.1312, + "step": 4226 + }, + { + "epoch": 0.7526709401709402, + "grad_norm": 0.5363932251930237, + "learning_rate": 0.00018315668068107004, + "loss": 1.174, + "step": 4227 + }, + { + "epoch": 0.7528490028490028, + "grad_norm": 0.5585991740226746, + "learning_rate": 0.00018314890532406366, + "loss": 1.2106, + "step": 4228 + }, + { + "epoch": 0.7530270655270656, + "grad_norm": 0.49395787715911865, + "learning_rate": 0.0001831411283379272, + "loss": 1.1163, + "step": 4229 + }, + { + "epoch": 0.7532051282051282, + "grad_norm": 0.5081066489219666, + "learning_rate": 0.00018313334972281306, + "loss": 1.184, + "step": 4230 + }, + { + "epoch": 0.7533831908831908, + "grad_norm": 0.40304034948349, + "learning_rate": 0.0001831255694788736, + "loss": 0.7548, + "step": 4231 + }, + { + "epoch": 0.7535612535612536, + "grad_norm": 0.4999815821647644, + "learning_rate": 0.0001831177876062613, + "loss": 1.0092, + "step": 4232 + }, + { + "epoch": 0.7537393162393162, + "grad_norm": 0.48917025327682495, + "learning_rate": 0.00018311000410512862, + "loss": 1.0354, + "step": 4233 + }, + { + "epoch": 0.7539173789173789, + "grad_norm": 0.475606769323349, + "learning_rate": 0.00018310221897562806, + "loss": 0.8728, + "step": 4234 + }, + { + "epoch": 0.7540954415954416, + "grad_norm": 0.630439817905426, + "learning_rate": 0.00018309443221791214, + "loss": 1.1436, + "step": 4235 + }, + { + "epoch": 0.7542735042735043, + "grad_norm": 0.524740993976593, + "learning_rate": 0.00018308664383213344, + "loss": 1.0487, + "step": 4236 + }, + { + "epoch": 0.7544515669515669, + "grad_norm": 0.4734523892402649, + "learning_rate": 0.0001830788538184445, + "loss": 1.0681, + "step": 4237 + }, + { + "epoch": 0.7546296296296297, + "grad_norm": 0.5767266750335693, + "learning_rate": 0.00018307106217699807, + "loss": 1.0599, + "step": 4238 + }, + { + "epoch": 0.7548076923076923, + "grad_norm": 0.6276642084121704, + "learning_rate": 0.0001830632689079467, + "loss": 1.2837, + "step": 4239 + }, + { + "epoch": 0.7549857549857549, + "grad_norm": 0.5539988279342651, + "learning_rate": 0.00018305547401144316, + "loss": 0.9072, + "step": 4240 + }, + { + "epoch": 0.7551638176638177, + "grad_norm": 0.4551292061805725, + "learning_rate": 0.00018304767748764014, + "loss": 1.0204, + "step": 4241 + }, + { + "epoch": 0.7553418803418803, + "grad_norm": 0.47344550490379333, + "learning_rate": 0.00018303987933669034, + "loss": 1.0473, + "step": 4242 + }, + { + "epoch": 0.7555199430199431, + "grad_norm": 0.6050213575363159, + "learning_rate": 0.00018303207955874665, + "loss": 1.1552, + "step": 4243 + }, + { + "epoch": 0.7556980056980057, + "grad_norm": 0.48943889141082764, + "learning_rate": 0.00018302427815396186, + "loss": 1.0002, + "step": 4244 + }, + { + "epoch": 0.7558760683760684, + "grad_norm": 0.5664682984352112, + "learning_rate": 0.00018301647512248878, + "loss": 1.1865, + "step": 4245 + }, + { + "epoch": 0.7560541310541311, + "grad_norm": 0.5702242255210876, + "learning_rate": 0.00018300867046448034, + "loss": 1.3029, + "step": 4246 + }, + { + "epoch": 0.7562321937321937, + "grad_norm": 0.593207836151123, + "learning_rate": 0.00018300086418008942, + "loss": 1.109, + "step": 4247 + }, + { + "epoch": 0.7564102564102564, + "grad_norm": 0.5887887477874756, + "learning_rate": 0.000182993056269469, + "loss": 1.3022, + "step": 4248 + }, + { + "epoch": 0.7565883190883191, + "grad_norm": 0.5277966260910034, + "learning_rate": 0.00018298524673277203, + "loss": 1.1738, + "step": 4249 + }, + { + "epoch": 0.7567663817663818, + "grad_norm": 0.589347779750824, + "learning_rate": 0.00018297743557015155, + "loss": 1.0185, + "step": 4250 + }, + { + "epoch": 0.7569444444444444, + "grad_norm": 0.49920859932899475, + "learning_rate": 0.0001829696227817606, + "loss": 1.118, + "step": 4251 + }, + { + "epoch": 0.7571225071225072, + "grad_norm": 0.502565324306488, + "learning_rate": 0.0001829618083677522, + "loss": 1.1856, + "step": 4252 + }, + { + "epoch": 0.7573005698005698, + "grad_norm": 0.49814435839653015, + "learning_rate": 0.00018295399232827955, + "loss": 1.0432, + "step": 4253 + }, + { + "epoch": 0.7574786324786325, + "grad_norm": 0.5087502598762512, + "learning_rate": 0.00018294617466349574, + "loss": 1.2325, + "step": 4254 + }, + { + "epoch": 0.7576566951566952, + "grad_norm": 0.5107288956642151, + "learning_rate": 0.00018293835537355394, + "loss": 1.0487, + "step": 4255 + }, + { + "epoch": 0.7578347578347578, + "grad_norm": 0.524725615978241, + "learning_rate": 0.00018293053445860732, + "loss": 1.1821, + "step": 4256 + }, + { + "epoch": 0.7580128205128205, + "grad_norm": 0.5234082937240601, + "learning_rate": 0.0001829227119188092, + "loss": 0.8896, + "step": 4257 + }, + { + "epoch": 0.7581908831908832, + "grad_norm": 0.5102918744087219, + "learning_rate": 0.00018291488775431275, + "loss": 1.0246, + "step": 4258 + }, + { + "epoch": 0.7583689458689459, + "grad_norm": 0.5552714467048645, + "learning_rate": 0.00018290706196527135, + "loss": 1.0193, + "step": 4259 + }, + { + "epoch": 0.7585470085470085, + "grad_norm": 0.5395022630691528, + "learning_rate": 0.00018289923455183825, + "loss": 1.3203, + "step": 4260 + }, + { + "epoch": 0.7587250712250713, + "grad_norm": 0.7474865913391113, + "learning_rate": 0.00018289140551416692, + "loss": 1.182, + "step": 4261 + }, + { + "epoch": 0.7589031339031339, + "grad_norm": 0.4892016649246216, + "learning_rate": 0.00018288357485241066, + "loss": 0.968, + "step": 4262 + }, + { + "epoch": 0.7590811965811965, + "grad_norm": 0.4627816081047058, + "learning_rate": 0.00018287574256672291, + "loss": 0.6895, + "step": 4263 + }, + { + "epoch": 0.7592592592592593, + "grad_norm": 0.6221280097961426, + "learning_rate": 0.00018286790865725715, + "loss": 0.9691, + "step": 4264 + }, + { + "epoch": 0.7594373219373219, + "grad_norm": 0.5542295575141907, + "learning_rate": 0.0001828600731241669, + "loss": 0.9996, + "step": 4265 + }, + { + "epoch": 0.7596153846153846, + "grad_norm": 0.5570770502090454, + "learning_rate": 0.00018285223596760562, + "loss": 1.1996, + "step": 4266 + }, + { + "epoch": 0.7597934472934473, + "grad_norm": 0.5495262742042542, + "learning_rate": 0.00018284439718772687, + "loss": 1.1572, + "step": 4267 + }, + { + "epoch": 0.75997150997151, + "grad_norm": 0.5006741881370544, + "learning_rate": 0.00018283655678468427, + "loss": 1.1215, + "step": 4268 + }, + { + "epoch": 0.7601495726495726, + "grad_norm": 0.4682157635688782, + "learning_rate": 0.00018282871475863144, + "loss": 1.0547, + "step": 4269 + }, + { + "epoch": 0.7603276353276354, + "grad_norm": 0.6275840997695923, + "learning_rate": 0.00018282087110972197, + "loss": 1.3855, + "step": 4270 + }, + { + "epoch": 0.760505698005698, + "grad_norm": 0.5341474413871765, + "learning_rate": 0.0001828130258381096, + "loss": 1.2024, + "step": 4271 + }, + { + "epoch": 0.7606837606837606, + "grad_norm": 0.4330833852291107, + "learning_rate": 0.000182805178943948, + "loss": 1.0508, + "step": 4272 + }, + { + "epoch": 0.7608618233618234, + "grad_norm": 0.6276537179946899, + "learning_rate": 0.00018279733042739094, + "loss": 1.1635, + "step": 4273 + }, + { + "epoch": 0.761039886039886, + "grad_norm": 0.5370199084281921, + "learning_rate": 0.00018278948028859217, + "loss": 1.0579, + "step": 4274 + }, + { + "epoch": 0.7612179487179487, + "grad_norm": 0.524959921836853, + "learning_rate": 0.00018278162852770552, + "loss": 1.0972, + "step": 4275 + }, + { + "epoch": 0.7613960113960114, + "grad_norm": 0.5029389262199402, + "learning_rate": 0.00018277377514488486, + "loss": 0.959, + "step": 4276 + }, + { + "epoch": 0.7615740740740741, + "grad_norm": 0.49772894382476807, + "learning_rate": 0.00018276592014028397, + "loss": 1.2773, + "step": 4277 + }, + { + "epoch": 0.7617521367521367, + "grad_norm": 0.5195719003677368, + "learning_rate": 0.00018275806351405685, + "loss": 1.0676, + "step": 4278 + }, + { + "epoch": 0.7619301994301995, + "grad_norm": 0.5167942643165588, + "learning_rate": 0.00018275020526635735, + "loss": 1.0615, + "step": 4279 + }, + { + "epoch": 0.7621082621082621, + "grad_norm": 0.4958035945892334, + "learning_rate": 0.0001827423453973395, + "loss": 0.9605, + "step": 4280 + }, + { + "epoch": 0.7622863247863247, + "grad_norm": 0.6256808042526245, + "learning_rate": 0.00018273448390715728, + "loss": 1.2526, + "step": 4281 + }, + { + "epoch": 0.7624643874643875, + "grad_norm": 0.5062580108642578, + "learning_rate": 0.0001827266207959647, + "loss": 1.0604, + "step": 4282 + }, + { + "epoch": 0.7626424501424501, + "grad_norm": 0.5080778002738953, + "learning_rate": 0.00018271875606391583, + "loss": 1.1246, + "step": 4283 + }, + { + "epoch": 0.7628205128205128, + "grad_norm": 0.5069389939308167, + "learning_rate": 0.00018271088971116479, + "loss": 1.3158, + "step": 4284 + }, + { + "epoch": 0.7629985754985755, + "grad_norm": 0.7280121445655823, + "learning_rate": 0.00018270302173786567, + "loss": 1.2066, + "step": 4285 + }, + { + "epoch": 0.7631766381766382, + "grad_norm": 0.6523470282554626, + "learning_rate": 0.00018269515214417267, + "loss": 1.3236, + "step": 4286 + }, + { + "epoch": 0.7633547008547008, + "grad_norm": 0.5799322724342346, + "learning_rate": 0.00018268728093023988, + "loss": 0.9786, + "step": 4287 + }, + { + "epoch": 0.7635327635327636, + "grad_norm": 0.46675166487693787, + "learning_rate": 0.00018267940809622163, + "loss": 0.8131, + "step": 4288 + }, + { + "epoch": 0.7637108262108262, + "grad_norm": 0.5566182732582092, + "learning_rate": 0.00018267153364227214, + "loss": 1.0565, + "step": 4289 + }, + { + "epoch": 0.7638888888888888, + "grad_norm": 0.532028079032898, + "learning_rate": 0.00018266365756854566, + "loss": 0.952, + "step": 4290 + }, + { + "epoch": 0.7640669515669516, + "grad_norm": 0.5082666873931885, + "learning_rate": 0.00018265577987519653, + "loss": 1.0704, + "step": 4291 + }, + { + "epoch": 0.7642450142450142, + "grad_norm": 0.5223562717437744, + "learning_rate": 0.00018264790056237912, + "loss": 1.1161, + "step": 4292 + }, + { + "epoch": 0.7644230769230769, + "grad_norm": 0.48472318053245544, + "learning_rate": 0.00018264001963024778, + "loss": 0.8784, + "step": 4293 + }, + { + "epoch": 0.7646011396011396, + "grad_norm": 0.5901281833648682, + "learning_rate": 0.0001826321370789569, + "loss": 1.1031, + "step": 4294 + }, + { + "epoch": 0.7647792022792023, + "grad_norm": 0.570350706577301, + "learning_rate": 0.000182624252908661, + "loss": 0.9047, + "step": 4295 + }, + { + "epoch": 0.7649572649572649, + "grad_norm": 0.568373441696167, + "learning_rate": 0.00018261636711951445, + "loss": 1.0106, + "step": 4296 + }, + { + "epoch": 0.7651353276353277, + "grad_norm": 0.6175880432128906, + "learning_rate": 0.00018260847971167182, + "loss": 1.3531, + "step": 4297 + }, + { + "epoch": 0.7653133903133903, + "grad_norm": 0.5682594776153564, + "learning_rate": 0.00018260059068528762, + "loss": 1.1261, + "step": 4298 + }, + { + "epoch": 0.7654914529914529, + "grad_norm": 0.5050225257873535, + "learning_rate": 0.00018259270004051644, + "loss": 1.0921, + "step": 4299 + }, + { + "epoch": 0.7656695156695157, + "grad_norm": 0.5416565537452698, + "learning_rate": 0.0001825848077775129, + "loss": 1.0881, + "step": 4300 + }, + { + "epoch": 0.7658475783475783, + "grad_norm": 0.5418867468833923, + "learning_rate": 0.0001825769138964316, + "loss": 1.2069, + "step": 4301 + }, + { + "epoch": 0.7660256410256411, + "grad_norm": 0.5447866320610046, + "learning_rate": 0.00018256901839742718, + "loss": 1.1827, + "step": 4302 + }, + { + "epoch": 0.7662037037037037, + "grad_norm": 0.5482802987098694, + "learning_rate": 0.00018256112128065439, + "loss": 1.0492, + "step": 4303 + }, + { + "epoch": 0.7663817663817664, + "grad_norm": 0.5059601664543152, + "learning_rate": 0.0001825532225462679, + "loss": 1.0996, + "step": 4304 + }, + { + "epoch": 0.7665598290598291, + "grad_norm": 0.5153701901435852, + "learning_rate": 0.00018254532219442258, + "loss": 1.3237, + "step": 4305 + }, + { + "epoch": 0.7667378917378918, + "grad_norm": 0.5370768904685974, + "learning_rate": 0.0001825374202252731, + "loss": 0.9925, + "step": 4306 + }, + { + "epoch": 0.7669159544159544, + "grad_norm": 0.4516580402851105, + "learning_rate": 0.00018252951663897432, + "loss": 1.0749, + "step": 4307 + }, + { + "epoch": 0.7670940170940171, + "grad_norm": 0.5565171837806702, + "learning_rate": 0.0001825216114356811, + "loss": 1.1617, + "step": 4308 + }, + { + "epoch": 0.7672720797720798, + "grad_norm": 0.5212662220001221, + "learning_rate": 0.00018251370461554834, + "loss": 1.1108, + "step": 4309 + }, + { + "epoch": 0.7674501424501424, + "grad_norm": 0.49061715602874756, + "learning_rate": 0.00018250579617873095, + "loss": 1.0881, + "step": 4310 + }, + { + "epoch": 0.7676282051282052, + "grad_norm": 0.5535751581192017, + "learning_rate": 0.00018249788612538387, + "loss": 0.9341, + "step": 4311 + }, + { + "epoch": 0.7678062678062678, + "grad_norm": 0.5425209403038025, + "learning_rate": 0.00018248997445566208, + "loss": 1.1858, + "step": 4312 + }, + { + "epoch": 0.7679843304843305, + "grad_norm": 0.6224395036697388, + "learning_rate": 0.0001824820611697206, + "loss": 1.0836, + "step": 4313 + }, + { + "epoch": 0.7681623931623932, + "grad_norm": 0.4895690977573395, + "learning_rate": 0.00018247414626771445, + "loss": 0.8598, + "step": 4314 + }, + { + "epoch": 0.7683404558404558, + "grad_norm": 0.5279615521430969, + "learning_rate": 0.00018246622974979877, + "loss": 1.1742, + "step": 4315 + }, + { + "epoch": 0.7685185185185185, + "grad_norm": 0.45300471782684326, + "learning_rate": 0.0001824583116161286, + "loss": 0.8872, + "step": 4316 + }, + { + "epoch": 0.7686965811965812, + "grad_norm": 0.6499692797660828, + "learning_rate": 0.00018245039186685916, + "loss": 1.2495, + "step": 4317 + }, + { + "epoch": 0.7688746438746439, + "grad_norm": 0.48151278495788574, + "learning_rate": 0.00018244247050214552, + "loss": 1.2382, + "step": 4318 + }, + { + "epoch": 0.7690527065527065, + "grad_norm": 0.6597028374671936, + "learning_rate": 0.0001824345475221429, + "loss": 1.3453, + "step": 4319 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.4536992609500885, + "learning_rate": 0.0001824266229270066, + "loss": 1.1141, + "step": 4320 + }, + { + "epoch": 0.7694088319088319, + "grad_norm": 0.5489405393600464, + "learning_rate": 0.00018241869671689184, + "loss": 1.0333, + "step": 4321 + }, + { + "epoch": 0.7695868945868946, + "grad_norm": 0.5741586089134216, + "learning_rate": 0.00018241076889195394, + "loss": 0.9939, + "step": 4322 + }, + { + "epoch": 0.7697649572649573, + "grad_norm": 0.47170960903167725, + "learning_rate": 0.00018240283945234823, + "loss": 0.9878, + "step": 4323 + }, + { + "epoch": 0.76994301994302, + "grad_norm": 0.4729093313217163, + "learning_rate": 0.00018239490839823004, + "loss": 1.0087, + "step": 4324 + }, + { + "epoch": 0.7701210826210826, + "grad_norm": 0.49869823455810547, + "learning_rate": 0.0001823869757297548, + "loss": 1.169, + "step": 4325 + }, + { + "epoch": 0.7702991452991453, + "grad_norm": 0.5118468403816223, + "learning_rate": 0.0001823790414470779, + "loss": 1.1092, + "step": 4326 + }, + { + "epoch": 0.770477207977208, + "grad_norm": 0.5076048970222473, + "learning_rate": 0.0001823711055503548, + "loss": 1.1028, + "step": 4327 + }, + { + "epoch": 0.7706552706552706, + "grad_norm": 0.5661569237709045, + "learning_rate": 0.00018236316803974098, + "loss": 1.1114, + "step": 4328 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 0.5542354583740234, + "learning_rate": 0.000182355228915392, + "loss": 1.0931, + "step": 4329 + }, + { + "epoch": 0.771011396011396, + "grad_norm": 0.5476680994033813, + "learning_rate": 0.0001823472881774634, + "loss": 1.036, + "step": 4330 + }, + { + "epoch": 0.7711894586894587, + "grad_norm": 0.5449798703193665, + "learning_rate": 0.00018233934582611073, + "loss": 1.0682, + "step": 4331 + }, + { + "epoch": 0.7713675213675214, + "grad_norm": 0.61089026927948, + "learning_rate": 0.00018233140186148963, + "loss": 1.0748, + "step": 4332 + }, + { + "epoch": 0.771545584045584, + "grad_norm": 0.5015206336975098, + "learning_rate": 0.00018232345628375576, + "loss": 1.2032, + "step": 4333 + }, + { + "epoch": 0.7717236467236467, + "grad_norm": 0.579289972782135, + "learning_rate": 0.00018231550909306475, + "loss": 1.0764, + "step": 4334 + }, + { + "epoch": 0.7719017094017094, + "grad_norm": 0.5889299511909485, + "learning_rate": 0.00018230756028957235, + "loss": 1.1768, + "step": 4335 + }, + { + "epoch": 0.7720797720797721, + "grad_norm": 0.5328249335289001, + "learning_rate": 0.00018229960987343428, + "loss": 1.0055, + "step": 4336 + }, + { + "epoch": 0.7722578347578347, + "grad_norm": 0.5766382217407227, + "learning_rate": 0.0001822916578448063, + "loss": 0.9923, + "step": 4337 + }, + { + "epoch": 0.7724358974358975, + "grad_norm": 0.6448187828063965, + "learning_rate": 0.00018228370420384423, + "loss": 1.1135, + "step": 4338 + }, + { + "epoch": 0.7726139601139601, + "grad_norm": 0.5505210757255554, + "learning_rate": 0.00018227574895070394, + "loss": 1.2048, + "step": 4339 + }, + { + "epoch": 0.7727920227920227, + "grad_norm": 0.6278925538063049, + "learning_rate": 0.00018226779208554126, + "loss": 1.1045, + "step": 4340 + }, + { + "epoch": 0.7729700854700855, + "grad_norm": 0.5345009565353394, + "learning_rate": 0.00018225983360851207, + "loss": 1.0102, + "step": 4341 + }, + { + "epoch": 0.7731481481481481, + "grad_norm": 0.566633403301239, + "learning_rate": 0.00018225187351977233, + "loss": 1.0038, + "step": 4342 + }, + { + "epoch": 0.7733262108262108, + "grad_norm": 0.5066078901290894, + "learning_rate": 0.000182243911819478, + "loss": 1.0339, + "step": 4343 + }, + { + "epoch": 0.7735042735042735, + "grad_norm": 0.5614920258522034, + "learning_rate": 0.00018223594850778503, + "loss": 1.1021, + "step": 4344 + }, + { + "epoch": 0.7736823361823362, + "grad_norm": 0.7747337818145752, + "learning_rate": 0.0001822279835848495, + "loss": 1.1129, + "step": 4345 + }, + { + "epoch": 0.7738603988603988, + "grad_norm": 0.7066529989242554, + "learning_rate": 0.00018222001705082744, + "loss": 1.3234, + "step": 4346 + }, + { + "epoch": 0.7740384615384616, + "grad_norm": 0.6340884566307068, + "learning_rate": 0.00018221204890587497, + "loss": 1.0726, + "step": 4347 + }, + { + "epoch": 0.7742165242165242, + "grad_norm": 0.5401145815849304, + "learning_rate": 0.00018220407915014818, + "loss": 0.9904, + "step": 4348 + }, + { + "epoch": 0.7743945868945868, + "grad_norm": 0.5069159269332886, + "learning_rate": 0.00018219610778380315, + "loss": 1.0654, + "step": 4349 + }, + { + "epoch": 0.7745726495726496, + "grad_norm": 0.5422839522361755, + "learning_rate": 0.00018218813480699623, + "loss": 1.1741, + "step": 4350 + }, + { + "epoch": 0.7747507122507122, + "grad_norm": 0.5550300478935242, + "learning_rate": 0.0001821801602198835, + "loss": 1.0033, + "step": 4351 + }, + { + "epoch": 0.7749287749287749, + "grad_norm": 0.5987736582756042, + "learning_rate": 0.00018217218402262123, + "loss": 0.935, + "step": 4352 + }, + { + "epoch": 0.7751068376068376, + "grad_norm": 0.6137008666992188, + "learning_rate": 0.00018216420621536573, + "loss": 1.17, + "step": 4353 + }, + { + "epoch": 0.7752849002849003, + "grad_norm": 0.47124359011650085, + "learning_rate": 0.0001821562267982733, + "loss": 0.8316, + "step": 4354 + }, + { + "epoch": 0.7754629629629629, + "grad_norm": 0.5057868361473083, + "learning_rate": 0.00018214824577150024, + "loss": 1.0246, + "step": 4355 + }, + { + "epoch": 0.7756410256410257, + "grad_norm": 0.604055643081665, + "learning_rate": 0.00018214026313520299, + "loss": 1.1272, + "step": 4356 + }, + { + "epoch": 0.7758190883190883, + "grad_norm": 0.6690384149551392, + "learning_rate": 0.0001821322788895379, + "loss": 1.0464, + "step": 4357 + }, + { + "epoch": 0.7759971509971509, + "grad_norm": 0.5458958745002747, + "learning_rate": 0.0001821242930346614, + "loss": 1.1712, + "step": 4358 + }, + { + "epoch": 0.7761752136752137, + "grad_norm": 0.6448663473129272, + "learning_rate": 0.00018211630557073, + "loss": 1.1125, + "step": 4359 + }, + { + "epoch": 0.7763532763532763, + "grad_norm": 0.49889448285102844, + "learning_rate": 0.00018210831649790018, + "loss": 1.097, + "step": 4360 + }, + { + "epoch": 0.7765313390313391, + "grad_norm": 0.5118046998977661, + "learning_rate": 0.00018210032581632843, + "loss": 1.009, + "step": 4361 + }, + { + "epoch": 0.7767094017094017, + "grad_norm": 0.5450068116188049, + "learning_rate": 0.00018209233352617135, + "loss": 1.1138, + "step": 4362 + }, + { + "epoch": 0.7768874643874644, + "grad_norm": 0.6147481203079224, + "learning_rate": 0.00018208433962758558, + "loss": 1.212, + "step": 4363 + }, + { + "epoch": 0.7770655270655271, + "grad_norm": 0.554176926612854, + "learning_rate": 0.00018207634412072764, + "loss": 1.1271, + "step": 4364 + }, + { + "epoch": 0.7772435897435898, + "grad_norm": 0.5872851014137268, + "learning_rate": 0.00018206834700575426, + "loss": 1.2793, + "step": 4365 + }, + { + "epoch": 0.7774216524216524, + "grad_norm": 0.5135685205459595, + "learning_rate": 0.00018206034828282207, + "loss": 0.9642, + "step": 4366 + }, + { + "epoch": 0.7775997150997151, + "grad_norm": 0.5699490308761597, + "learning_rate": 0.00018205234795208786, + "loss": 0.9086, + "step": 4367 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 0.5908057689666748, + "learning_rate": 0.00018204434601370832, + "loss": 1.1973, + "step": 4368 + }, + { + "epoch": 0.7779558404558404, + "grad_norm": 0.5777581334114075, + "learning_rate": 0.00018203634246784025, + "loss": 1.0447, + "step": 4369 + }, + { + "epoch": 0.7781339031339032, + "grad_norm": 0.4822927713394165, + "learning_rate": 0.00018202833731464048, + "loss": 0.814, + "step": 4370 + }, + { + "epoch": 0.7783119658119658, + "grad_norm": 0.5343610644340515, + "learning_rate": 0.0001820203305542658, + "loss": 1.2785, + "step": 4371 + }, + { + "epoch": 0.7784900284900285, + "grad_norm": 0.5462222695350647, + "learning_rate": 0.00018201232218687316, + "loss": 1.1785, + "step": 4372 + }, + { + "epoch": 0.7786680911680912, + "grad_norm": 0.5177609324455261, + "learning_rate": 0.00018200431221261943, + "loss": 1.111, + "step": 4373 + }, + { + "epoch": 0.7788461538461539, + "grad_norm": 0.5324625968933105, + "learning_rate": 0.00018199630063166157, + "loss": 1.0738, + "step": 4374 + }, + { + "epoch": 0.7790242165242165, + "grad_norm": 0.6392876505851746, + "learning_rate": 0.0001819882874441565, + "loss": 1.1758, + "step": 4375 + }, + { + "epoch": 0.7792022792022792, + "grad_norm": 0.49964696168899536, + "learning_rate": 0.00018198027265026127, + "loss": 1.0556, + "step": 4376 + }, + { + "epoch": 0.7793803418803419, + "grad_norm": 0.6090660691261292, + "learning_rate": 0.00018197225625013287, + "loss": 1.0102, + "step": 4377 + }, + { + "epoch": 0.7795584045584045, + "grad_norm": 0.5242345929145813, + "learning_rate": 0.00018196423824392842, + "loss": 0.8335, + "step": 4378 + }, + { + "epoch": 0.7797364672364673, + "grad_norm": 0.5265036225318909, + "learning_rate": 0.00018195621863180498, + "loss": 1.0781, + "step": 4379 + }, + { + "epoch": 0.7799145299145299, + "grad_norm": 0.5115378499031067, + "learning_rate": 0.0001819481974139197, + "loss": 1.1658, + "step": 4380 + }, + { + "epoch": 0.7800925925925926, + "grad_norm": 0.6489549875259399, + "learning_rate": 0.00018194017459042972, + "loss": 1.0572, + "step": 4381 + }, + { + "epoch": 0.7802706552706553, + "grad_norm": 0.5800202488899231, + "learning_rate": 0.0001819321501614922, + "loss": 0.9593, + "step": 4382 + }, + { + "epoch": 0.780448717948718, + "grad_norm": 0.5608528256416321, + "learning_rate": 0.00018192412412726443, + "loss": 1.0324, + "step": 4383 + }, + { + "epoch": 0.7806267806267806, + "grad_norm": 0.5596401691436768, + "learning_rate": 0.00018191609648790362, + "loss": 1.071, + "step": 4384 + }, + { + "epoch": 0.7808048433048433, + "grad_norm": 0.5712903141975403, + "learning_rate": 0.00018190806724356707, + "loss": 0.9011, + "step": 4385 + }, + { + "epoch": 0.780982905982906, + "grad_norm": 0.5079438090324402, + "learning_rate": 0.0001819000363944121, + "loss": 1.1194, + "step": 4386 + }, + { + "epoch": 0.7811609686609686, + "grad_norm": 0.5785079598426819, + "learning_rate": 0.00018189200394059602, + "loss": 1.1703, + "step": 4387 + }, + { + "epoch": 0.7813390313390314, + "grad_norm": 0.6901816129684448, + "learning_rate": 0.00018188396988227625, + "loss": 1.6689, + "step": 4388 + }, + { + "epoch": 0.781517094017094, + "grad_norm": 0.48107922077178955, + "learning_rate": 0.00018187593421961022, + "loss": 1.0116, + "step": 4389 + }, + { + "epoch": 0.7816951566951567, + "grad_norm": 0.5843084454536438, + "learning_rate": 0.0001818678969527553, + "loss": 1.1172, + "step": 4390 + }, + { + "epoch": 0.7818732193732194, + "grad_norm": 0.479034423828125, + "learning_rate": 0.00018185985808186902, + "loss": 0.811, + "step": 4391 + }, + { + "epoch": 0.782051282051282, + "grad_norm": 0.5864158272743225, + "learning_rate": 0.00018185181760710888, + "loss": 0.9522, + "step": 4392 + }, + { + "epoch": 0.7822293447293447, + "grad_norm": 0.4824625551700592, + "learning_rate": 0.00018184377552863242, + "loss": 0.9039, + "step": 4393 + }, + { + "epoch": 0.7824074074074074, + "grad_norm": 0.580102801322937, + "learning_rate": 0.00018183573184659717, + "loss": 1.2382, + "step": 4394 + }, + { + "epoch": 0.7825854700854701, + "grad_norm": 0.5300056338310242, + "learning_rate": 0.00018182768656116073, + "loss": 1.2268, + "step": 4395 + }, + { + "epoch": 0.7827635327635327, + "grad_norm": 0.5548123121261597, + "learning_rate": 0.00018181963967248078, + "loss": 1.0628, + "step": 4396 + }, + { + "epoch": 0.7829415954415955, + "grad_norm": 0.5485070943832397, + "learning_rate": 0.00018181159118071496, + "loss": 0.9628, + "step": 4397 + }, + { + "epoch": 0.7831196581196581, + "grad_norm": 0.47405415773391724, + "learning_rate": 0.00018180354108602095, + "loss": 1.1413, + "step": 4398 + }, + { + "epoch": 0.7832977207977208, + "grad_norm": 0.5545752644538879, + "learning_rate": 0.0001817954893885565, + "loss": 1.3807, + "step": 4399 + }, + { + "epoch": 0.7834757834757835, + "grad_norm": 0.5339497327804565, + "learning_rate": 0.00018178743608847933, + "loss": 0.9978, + "step": 4400 + }, + { + "epoch": 0.7836538461538461, + "grad_norm": 0.5006352663040161, + "learning_rate": 0.00018177938118594725, + "loss": 0.8873, + "step": 4401 + }, + { + "epoch": 0.7838319088319088, + "grad_norm": 0.4845179319381714, + "learning_rate": 0.00018177132468111812, + "loss": 0.8866, + "step": 4402 + }, + { + "epoch": 0.7840099715099715, + "grad_norm": 0.5240967869758606, + "learning_rate": 0.0001817632665741497, + "loss": 1.0347, + "step": 4403 + }, + { + "epoch": 0.7841880341880342, + "grad_norm": 0.5311884880065918, + "learning_rate": 0.00018175520686519993, + "loss": 1.2065, + "step": 4404 + }, + { + "epoch": 0.7843660968660968, + "grad_norm": 0.5562815070152283, + "learning_rate": 0.00018174714555442673, + "loss": 1.1272, + "step": 4405 + }, + { + "epoch": 0.7845441595441596, + "grad_norm": 0.5524366497993469, + "learning_rate": 0.00018173908264198802, + "loss": 1.2337, + "step": 4406 + }, + { + "epoch": 0.7847222222222222, + "grad_norm": 0.5612216591835022, + "learning_rate": 0.0001817310181280418, + "loss": 1.1809, + "step": 4407 + }, + { + "epoch": 0.7849002849002849, + "grad_norm": 0.5315343737602234, + "learning_rate": 0.000181722952012746, + "loss": 1.0491, + "step": 4408 + }, + { + "epoch": 0.7850783475783476, + "grad_norm": 0.5233435034751892, + "learning_rate": 0.00018171488429625878, + "loss": 1.0457, + "step": 4409 + }, + { + "epoch": 0.7852564102564102, + "grad_norm": 0.7809093594551086, + "learning_rate": 0.00018170681497873813, + "loss": 1.1578, + "step": 4410 + }, + { + "epoch": 0.7854344729344729, + "grad_norm": 0.49659839272499084, + "learning_rate": 0.00018169874406034217, + "loss": 1.0815, + "step": 4411 + }, + { + "epoch": 0.7856125356125356, + "grad_norm": 0.5020765066146851, + "learning_rate": 0.00018169067154122904, + "loss": 1.1985, + "step": 4412 + }, + { + "epoch": 0.7857905982905983, + "grad_norm": 0.6408432126045227, + "learning_rate": 0.0001816825974215569, + "loss": 1.2272, + "step": 4413 + }, + { + "epoch": 0.7859686609686609, + "grad_norm": 0.5062605142593384, + "learning_rate": 0.00018167452170148396, + "loss": 0.9663, + "step": 4414 + }, + { + "epoch": 0.7861467236467237, + "grad_norm": 0.5100119113922119, + "learning_rate": 0.0001816664443811684, + "loss": 1.0256, + "step": 4415 + }, + { + "epoch": 0.7863247863247863, + "grad_norm": 0.5277643799781799, + "learning_rate": 0.00018165836546076854, + "loss": 1.2885, + "step": 4416 + }, + { + "epoch": 0.7865028490028491, + "grad_norm": 0.5568150281906128, + "learning_rate": 0.0001816502849404426, + "loss": 1.2673, + "step": 4417 + }, + { + "epoch": 0.7866809116809117, + "grad_norm": 0.5061392188072205, + "learning_rate": 0.00018164220282034896, + "loss": 1.072, + "step": 4418 + }, + { + "epoch": 0.7868589743589743, + "grad_norm": 0.5383077263832092, + "learning_rate": 0.00018163411910064597, + "loss": 1.0621, + "step": 4419 + }, + { + "epoch": 0.7870370370370371, + "grad_norm": 0.5167948007583618, + "learning_rate": 0.00018162603378149198, + "loss": 1.099, + "step": 4420 + }, + { + "epoch": 0.7872150997150997, + "grad_norm": 0.5084534287452698, + "learning_rate": 0.0001816179468630454, + "loss": 1.3984, + "step": 4421 + }, + { + "epoch": 0.7873931623931624, + "grad_norm": 0.608762264251709, + "learning_rate": 0.00018160985834546475, + "loss": 1.3553, + "step": 4422 + }, + { + "epoch": 0.7875712250712251, + "grad_norm": 0.4900866746902466, + "learning_rate": 0.00018160176822890842, + "loss": 1.0009, + "step": 4423 + }, + { + "epoch": 0.7877492877492878, + "grad_norm": 0.5928917527198792, + "learning_rate": 0.00018159367651353496, + "loss": 1.0523, + "step": 4424 + }, + { + "epoch": 0.7879273504273504, + "grad_norm": 0.624422013759613, + "learning_rate": 0.0001815855831995029, + "loss": 1.0519, + "step": 4425 + }, + { + "epoch": 0.7881054131054132, + "grad_norm": 0.5140150785446167, + "learning_rate": 0.00018157748828697082, + "loss": 1.048, + "step": 4426 + }, + { + "epoch": 0.7882834757834758, + "grad_norm": 0.47006943821907043, + "learning_rate": 0.00018156939177609732, + "loss": 1.0067, + "step": 4427 + }, + { + "epoch": 0.7884615384615384, + "grad_norm": 0.5178864002227783, + "learning_rate": 0.00018156129366704105, + "loss": 1.0583, + "step": 4428 + }, + { + "epoch": 0.7886396011396012, + "grad_norm": 0.5279985666275024, + "learning_rate": 0.00018155319395996066, + "loss": 1.3023, + "step": 4429 + }, + { + "epoch": 0.7888176638176638, + "grad_norm": 0.5238787531852722, + "learning_rate": 0.00018154509265501482, + "loss": 1.0851, + "step": 4430 + }, + { + "epoch": 0.7889957264957265, + "grad_norm": 0.5914917588233948, + "learning_rate": 0.00018153698975236228, + "loss": 0.9291, + "step": 4431 + }, + { + "epoch": 0.7891737891737892, + "grad_norm": 0.5046082735061646, + "learning_rate": 0.00018152888525216183, + "loss": 0.9951, + "step": 4432 + }, + { + "epoch": 0.7893518518518519, + "grad_norm": 0.5042256116867065, + "learning_rate": 0.00018152077915457225, + "loss": 1.0243, + "step": 4433 + }, + { + "epoch": 0.7895299145299145, + "grad_norm": 0.5950339436531067, + "learning_rate": 0.0001815126714597523, + "loss": 0.9803, + "step": 4434 + }, + { + "epoch": 0.7897079772079773, + "grad_norm": 0.5163764953613281, + "learning_rate": 0.0001815045621678609, + "loss": 1.0353, + "step": 4435 + }, + { + "epoch": 0.7898860398860399, + "grad_norm": 0.5166211128234863, + "learning_rate": 0.00018149645127905691, + "loss": 0.9649, + "step": 4436 + }, + { + "epoch": 0.7900641025641025, + "grad_norm": 0.5239769220352173, + "learning_rate": 0.00018148833879349927, + "loss": 0.9747, + "step": 4437 + }, + { + "epoch": 0.7902421652421653, + "grad_norm": 0.5803237557411194, + "learning_rate": 0.00018148022471134692, + "loss": 1.315, + "step": 4438 + }, + { + "epoch": 0.7904202279202279, + "grad_norm": 0.5141370296478271, + "learning_rate": 0.00018147210903275877, + "loss": 1.0547, + "step": 4439 + }, + { + "epoch": 0.7905982905982906, + "grad_norm": 0.545788586139679, + "learning_rate": 0.00018146399175789394, + "loss": 1.0797, + "step": 4440 + }, + { + "epoch": 0.7907763532763533, + "grad_norm": 0.5273314714431763, + "learning_rate": 0.0001814558728869114, + "loss": 0.7928, + "step": 4441 + }, + { + "epoch": 0.790954415954416, + "grad_norm": 0.4614652693271637, + "learning_rate": 0.00018144775241997024, + "loss": 0.8826, + "step": 4442 + }, + { + "epoch": 0.7911324786324786, + "grad_norm": 0.6203590631484985, + "learning_rate": 0.00018143963035722958, + "loss": 1.2891, + "step": 4443 + }, + { + "epoch": 0.7913105413105413, + "grad_norm": 0.4870408773422241, + "learning_rate": 0.0001814315066988485, + "loss": 1.0717, + "step": 4444 + }, + { + "epoch": 0.791488603988604, + "grad_norm": 0.6468982696533203, + "learning_rate": 0.00018142338144498625, + "loss": 1.3398, + "step": 4445 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 0.4727918207645416, + "learning_rate": 0.00018141525459580197, + "loss": 1.0195, + "step": 4446 + }, + { + "epoch": 0.7918447293447294, + "grad_norm": 0.5080479979515076, + "learning_rate": 0.0001814071261514549, + "loss": 1.0163, + "step": 4447 + }, + { + "epoch": 0.792022792022792, + "grad_norm": 0.5380908250808716, + "learning_rate": 0.0001813989961121043, + "loss": 1.1673, + "step": 4448 + }, + { + "epoch": 0.7922008547008547, + "grad_norm": 0.5020384192466736, + "learning_rate": 0.00018139086447790945, + "loss": 0.8591, + "step": 4449 + }, + { + "epoch": 0.7923789173789174, + "grad_norm": 0.5279949903488159, + "learning_rate": 0.0001813827312490297, + "loss": 1.1221, + "step": 4450 + }, + { + "epoch": 0.79255698005698, + "grad_norm": 0.6739233732223511, + "learning_rate": 0.00018137459642562437, + "loss": 1.2704, + "step": 4451 + }, + { + "epoch": 0.7927350427350427, + "grad_norm": 0.5112259984016418, + "learning_rate": 0.00018136646000785288, + "loss": 1.1161, + "step": 4452 + }, + { + "epoch": 0.7929131054131054, + "grad_norm": 0.5244031548500061, + "learning_rate": 0.00018135832199587463, + "loss": 0.7866, + "step": 4453 + }, + { + "epoch": 0.7930911680911681, + "grad_norm": 0.5803347229957581, + "learning_rate": 0.0001813501823898491, + "loss": 0.994, + "step": 4454 + }, + { + "epoch": 0.7932692307692307, + "grad_norm": 0.6191152930259705, + "learning_rate": 0.00018134204118993568, + "loss": 1.0725, + "step": 4455 + }, + { + "epoch": 0.7934472934472935, + "grad_norm": 0.549735963344574, + "learning_rate": 0.00018133389839629396, + "loss": 0.9915, + "step": 4456 + }, + { + "epoch": 0.7936253561253561, + "grad_norm": 0.4940381646156311, + "learning_rate": 0.00018132575400908347, + "loss": 1.1815, + "step": 4457 + }, + { + "epoch": 0.7938034188034188, + "grad_norm": 0.5009099245071411, + "learning_rate": 0.00018131760802846377, + "loss": 1.0833, + "step": 4458 + }, + { + "epoch": 0.7939814814814815, + "grad_norm": 0.595853865146637, + "learning_rate": 0.00018130946045459445, + "loss": 1.2774, + "step": 4459 + }, + { + "epoch": 0.7941595441595442, + "grad_norm": 0.534794807434082, + "learning_rate": 0.00018130131128763513, + "loss": 1.0891, + "step": 4460 + }, + { + "epoch": 0.7943376068376068, + "grad_norm": 0.5828582048416138, + "learning_rate": 0.00018129316052774557, + "loss": 1.0786, + "step": 4461 + }, + { + "epoch": 0.7945156695156695, + "grad_norm": 0.4750654697418213, + "learning_rate": 0.00018128500817508533, + "loss": 1.0818, + "step": 4462 + }, + { + "epoch": 0.7946937321937322, + "grad_norm": 0.5626576542854309, + "learning_rate": 0.00018127685422981426, + "loss": 1.0807, + "step": 4463 + }, + { + "epoch": 0.7948717948717948, + "grad_norm": 0.6434760093688965, + "learning_rate": 0.00018126869869209203, + "loss": 1.0908, + "step": 4464 + }, + { + "epoch": 0.7950498575498576, + "grad_norm": 0.5577414631843567, + "learning_rate": 0.00018126054156207853, + "loss": 1.0281, + "step": 4465 + }, + { + "epoch": 0.7952279202279202, + "grad_norm": 0.5001249313354492, + "learning_rate": 0.00018125238283993347, + "loss": 0.9083, + "step": 4466 + }, + { + "epoch": 0.7954059829059829, + "grad_norm": 0.5298314690589905, + "learning_rate": 0.00018124422252581676, + "loss": 0.971, + "step": 4467 + }, + { + "epoch": 0.7955840455840456, + "grad_norm": 0.4872737228870392, + "learning_rate": 0.00018123606061988832, + "loss": 1.0515, + "step": 4468 + }, + { + "epoch": 0.7957621082621082, + "grad_norm": 0.5895398259162903, + "learning_rate": 0.00018122789712230798, + "loss": 1.0771, + "step": 4469 + }, + { + "epoch": 0.7959401709401709, + "grad_norm": 0.5212514996528625, + "learning_rate": 0.00018121973203323577, + "loss": 1.0365, + "step": 4470 + }, + { + "epoch": 0.7961182336182336, + "grad_norm": 0.4679451584815979, + "learning_rate": 0.0001812115653528316, + "loss": 0.9445, + "step": 4471 + }, + { + "epoch": 0.7962962962962963, + "grad_norm": 0.5852653980255127, + "learning_rate": 0.00018120339708125552, + "loss": 1.1781, + "step": 4472 + }, + { + "epoch": 0.7964743589743589, + "grad_norm": 0.6081342697143555, + "learning_rate": 0.00018119522721866756, + "loss": 1.3881, + "step": 4473 + }, + { + "epoch": 0.7966524216524217, + "grad_norm": 0.5254155993461609, + "learning_rate": 0.00018118705576522777, + "loss": 1.2198, + "step": 4474 + }, + { + "epoch": 0.7968304843304843, + "grad_norm": 0.5959419012069702, + "learning_rate": 0.00018117888272109632, + "loss": 1.0922, + "step": 4475 + }, + { + "epoch": 0.7970085470085471, + "grad_norm": 0.6243147253990173, + "learning_rate": 0.0001811707080864333, + "loss": 1.1782, + "step": 4476 + }, + { + "epoch": 0.7971866096866097, + "grad_norm": 0.5336906909942627, + "learning_rate": 0.0001811625318613988, + "loss": 1.167, + "step": 4477 + }, + { + "epoch": 0.7973646723646723, + "grad_norm": 0.5287907719612122, + "learning_rate": 0.00018115435404615315, + "loss": 0.9923, + "step": 4478 + }, + { + "epoch": 0.7975427350427351, + "grad_norm": 0.48941442370414734, + "learning_rate": 0.0001811461746408565, + "loss": 0.863, + "step": 4479 + }, + { + "epoch": 0.7977207977207977, + "grad_norm": 0.48465651273727417, + "learning_rate": 0.0001811379936456691, + "loss": 1.147, + "step": 4480 + }, + { + "epoch": 0.7978988603988604, + "grad_norm": 0.5676067471504211, + "learning_rate": 0.0001811298110607513, + "loss": 1.3121, + "step": 4481 + }, + { + "epoch": 0.7980769230769231, + "grad_norm": 0.4894018769264221, + "learning_rate": 0.00018112162688626337, + "loss": 1.1831, + "step": 4482 + }, + { + "epoch": 0.7982549857549858, + "grad_norm": 0.5626382827758789, + "learning_rate": 0.0001811134411223657, + "loss": 1.1977, + "step": 4483 + }, + { + "epoch": 0.7984330484330484, + "grad_norm": 0.564119815826416, + "learning_rate": 0.00018110525376921862, + "loss": 1.2686, + "step": 4484 + }, + { + "epoch": 0.7986111111111112, + "grad_norm": 0.6385740041732788, + "learning_rate": 0.00018109706482698256, + "loss": 1.2418, + "step": 4485 + }, + { + "epoch": 0.7987891737891738, + "grad_norm": 0.5550164580345154, + "learning_rate": 0.00018108887429581802, + "loss": 1.081, + "step": 4486 + }, + { + "epoch": 0.7989672364672364, + "grad_norm": 0.5583973526954651, + "learning_rate": 0.00018108068217588544, + "loss": 1.1757, + "step": 4487 + }, + { + "epoch": 0.7991452991452992, + "grad_norm": 0.5533342957496643, + "learning_rate": 0.00018107248846734527, + "loss": 1.1947, + "step": 4488 + }, + { + "epoch": 0.7993233618233618, + "grad_norm": 0.5291479229927063, + "learning_rate": 0.00018106429317035815, + "loss": 1.2769, + "step": 4489 + }, + { + "epoch": 0.7995014245014245, + "grad_norm": 0.4680160582065582, + "learning_rate": 0.00018105609628508458, + "loss": 0.7059, + "step": 4490 + }, + { + "epoch": 0.7996794871794872, + "grad_norm": 0.5364881157875061, + "learning_rate": 0.00018104789781168517, + "loss": 1.0566, + "step": 4491 + }, + { + "epoch": 0.7998575498575499, + "grad_norm": 0.5917307734489441, + "learning_rate": 0.0001810396977503206, + "loss": 1.2263, + "step": 4492 + }, + { + "epoch": 0.8000356125356125, + "grad_norm": 0.6013199090957642, + "learning_rate": 0.0001810314961011515, + "loss": 1.2053, + "step": 4493 + }, + { + "epoch": 0.8002136752136753, + "grad_norm": 0.6005663275718689, + "learning_rate": 0.0001810232928643385, + "loss": 1.2241, + "step": 4494 + }, + { + "epoch": 0.8003917378917379, + "grad_norm": 0.49207603931427, + "learning_rate": 0.00018101508804004246, + "loss": 1.0661, + "step": 4495 + }, + { + "epoch": 0.8005698005698005, + "grad_norm": 0.4834063947200775, + "learning_rate": 0.00018100688162842401, + "loss": 1.1745, + "step": 4496 + }, + { + "epoch": 0.8007478632478633, + "grad_norm": 0.5347156524658203, + "learning_rate": 0.000180998673629644, + "loss": 1.0679, + "step": 4497 + }, + { + "epoch": 0.8009259259259259, + "grad_norm": 0.5815600156784058, + "learning_rate": 0.00018099046404386327, + "loss": 1.2652, + "step": 4498 + }, + { + "epoch": 0.8011039886039886, + "grad_norm": 0.5291135311126709, + "learning_rate": 0.00018098225287124263, + "loss": 1.2072, + "step": 4499 + }, + { + "epoch": 0.8012820512820513, + "grad_norm": 0.5779497027397156, + "learning_rate": 0.000180974040111943, + "loss": 1.3277, + "step": 4500 + }, + { + "epoch": 0.801460113960114, + "grad_norm": 0.44566696882247925, + "learning_rate": 0.0001809658257661252, + "loss": 0.7702, + "step": 4501 + }, + { + "epoch": 0.8016381766381766, + "grad_norm": 0.5407577753067017, + "learning_rate": 0.00018095760983395027, + "loss": 1.2894, + "step": 4502 + }, + { + "epoch": 0.8018162393162394, + "grad_norm": 0.4771903455257416, + "learning_rate": 0.00018094939231557916, + "loss": 1.045, + "step": 4503 + }, + { + "epoch": 0.801994301994302, + "grad_norm": 0.5970945358276367, + "learning_rate": 0.00018094117321117286, + "loss": 1.2059, + "step": 4504 + }, + { + "epoch": 0.8021723646723646, + "grad_norm": 0.4959338903427124, + "learning_rate": 0.0001809329525208924, + "loss": 1.155, + "step": 4505 + }, + { + "epoch": 0.8023504273504274, + "grad_norm": 0.5142548084259033, + "learning_rate": 0.00018092473024489887, + "loss": 0.9413, + "step": 4506 + }, + { + "epoch": 0.80252849002849, + "grad_norm": 0.5336433053016663, + "learning_rate": 0.00018091650638335334, + "loss": 1.0699, + "step": 4507 + }, + { + "epoch": 0.8027065527065527, + "grad_norm": 0.47770628333091736, + "learning_rate": 0.00018090828093641698, + "loss": 1.1515, + "step": 4508 + }, + { + "epoch": 0.8028846153846154, + "grad_norm": 0.5443438291549683, + "learning_rate": 0.00018090005390425091, + "loss": 1.189, + "step": 4509 + }, + { + "epoch": 0.8030626780626781, + "grad_norm": 0.523179829120636, + "learning_rate": 0.00018089182528701632, + "loss": 1.1272, + "step": 4510 + }, + { + "epoch": 0.8032407407407407, + "grad_norm": 0.49628451466560364, + "learning_rate": 0.00018088359508487448, + "loss": 0.9754, + "step": 4511 + }, + { + "epoch": 0.8034188034188035, + "grad_norm": 0.5933086276054382, + "learning_rate": 0.00018087536329798663, + "loss": 1.2111, + "step": 4512 + }, + { + "epoch": 0.8035968660968661, + "grad_norm": 0.4565310776233673, + "learning_rate": 0.00018086712992651402, + "loss": 0.7729, + "step": 4513 + }, + { + "epoch": 0.8037749287749287, + "grad_norm": 0.5013461112976074, + "learning_rate": 0.00018085889497061798, + "loss": 1.2178, + "step": 4514 + }, + { + "epoch": 0.8039529914529915, + "grad_norm": 0.5170024633407593, + "learning_rate": 0.00018085065843045987, + "loss": 0.9181, + "step": 4515 + }, + { + "epoch": 0.8041310541310541, + "grad_norm": 0.583363950252533, + "learning_rate": 0.00018084242030620104, + "loss": 1.1542, + "step": 4516 + }, + { + "epoch": 0.8043091168091168, + "grad_norm": 0.46835777163505554, + "learning_rate": 0.00018083418059800297, + "loss": 0.8954, + "step": 4517 + }, + { + "epoch": 0.8044871794871795, + "grad_norm": 0.5145657062530518, + "learning_rate": 0.000180825939306027, + "loss": 1.0417, + "step": 4518 + }, + { + "epoch": 0.8046652421652422, + "grad_norm": 0.47216105461120605, + "learning_rate": 0.00018081769643043467, + "loss": 0.9516, + "step": 4519 + }, + { + "epoch": 0.8048433048433048, + "grad_norm": 0.5059915781021118, + "learning_rate": 0.0001808094519713875, + "loss": 1.1643, + "step": 4520 + }, + { + "epoch": 0.8050213675213675, + "grad_norm": 0.5406439900398254, + "learning_rate": 0.00018080120592904692, + "loss": 1.2038, + "step": 4521 + }, + { + "epoch": 0.8051994301994302, + "grad_norm": 0.6123420000076294, + "learning_rate": 0.0001807929583035746, + "loss": 1.4004, + "step": 4522 + }, + { + "epoch": 0.8053774928774928, + "grad_norm": 0.49699845910072327, + "learning_rate": 0.00018078470909513208, + "loss": 1.0347, + "step": 4523 + }, + { + "epoch": 0.8055555555555556, + "grad_norm": 0.5369421243667603, + "learning_rate": 0.000180776458303881, + "loss": 1.0418, + "step": 4524 + }, + { + "epoch": 0.8057336182336182, + "grad_norm": 0.5407396554946899, + "learning_rate": 0.00018076820592998301, + "loss": 0.9546, + "step": 4525 + }, + { + "epoch": 0.8059116809116809, + "grad_norm": 0.5749752521514893, + "learning_rate": 0.00018075995197359984, + "loss": 1.1438, + "step": 4526 + }, + { + "epoch": 0.8060897435897436, + "grad_norm": 0.5523102283477783, + "learning_rate": 0.00018075169643489317, + "loss": 1.1312, + "step": 4527 + }, + { + "epoch": 0.8062678062678063, + "grad_norm": 0.5767508149147034, + "learning_rate": 0.00018074343931402472, + "loss": 1.1951, + "step": 4528 + }, + { + "epoch": 0.8064458689458689, + "grad_norm": 0.5262924432754517, + "learning_rate": 0.00018073518061115633, + "loss": 1.1985, + "step": 4529 + }, + { + "epoch": 0.8066239316239316, + "grad_norm": 0.4742378294467926, + "learning_rate": 0.0001807269203264498, + "loss": 1.0126, + "step": 4530 + }, + { + "epoch": 0.8068019943019943, + "grad_norm": 0.5190158486366272, + "learning_rate": 0.00018071865846006692, + "loss": 0.9985, + "step": 4531 + }, + { + "epoch": 0.8069800569800569, + "grad_norm": 0.5910618305206299, + "learning_rate": 0.00018071039501216964, + "loss": 1.2776, + "step": 4532 + }, + { + "epoch": 0.8071581196581197, + "grad_norm": 0.5363098382949829, + "learning_rate": 0.00018070212998291983, + "loss": 1.3346, + "step": 4533 + }, + { + "epoch": 0.8073361823361823, + "grad_norm": 0.47711408138275146, + "learning_rate": 0.0001806938633724794, + "loss": 1.04, + "step": 4534 + }, + { + "epoch": 0.8075142450142451, + "grad_norm": 0.5092964172363281, + "learning_rate": 0.0001806855951810104, + "loss": 1.1409, + "step": 4535 + }, + { + "epoch": 0.8076923076923077, + "grad_norm": 0.5828777551651001, + "learning_rate": 0.00018067732540867472, + "loss": 1.3048, + "step": 4536 + }, + { + "epoch": 0.8078703703703703, + "grad_norm": 0.5779826045036316, + "learning_rate": 0.00018066905405563445, + "loss": 1.1599, + "step": 4537 + }, + { + "epoch": 0.8080484330484331, + "grad_norm": 0.49908435344696045, + "learning_rate": 0.00018066078112205167, + "loss": 1.1502, + "step": 4538 + }, + { + "epoch": 0.8082264957264957, + "grad_norm": 0.4772704839706421, + "learning_rate": 0.0001806525066080884, + "loss": 0.7925, + "step": 4539 + }, + { + "epoch": 0.8084045584045584, + "grad_norm": 0.4298383295536041, + "learning_rate": 0.00018064423051390683, + "loss": 0.7322, + "step": 4540 + }, + { + "epoch": 0.8085826210826211, + "grad_norm": 0.49349579215049744, + "learning_rate": 0.0001806359528396691, + "loss": 1.0021, + "step": 4541 + }, + { + "epoch": 0.8087606837606838, + "grad_norm": 0.4698609411716461, + "learning_rate": 0.00018062767358553735, + "loss": 0.9751, + "step": 4542 + }, + { + "epoch": 0.8089387464387464, + "grad_norm": 0.4949014186859131, + "learning_rate": 0.00018061939275167385, + "loss": 0.9553, + "step": 4543 + }, + { + "epoch": 0.8091168091168092, + "grad_norm": 0.5604463815689087, + "learning_rate": 0.0001806111103382408, + "loss": 0.9894, + "step": 4544 + }, + { + "epoch": 0.8092948717948718, + "grad_norm": 0.5761561989784241, + "learning_rate": 0.00018060282634540053, + "loss": 1.258, + "step": 4545 + }, + { + "epoch": 0.8094729344729344, + "grad_norm": 0.5239115357398987, + "learning_rate": 0.00018059454077331527, + "loss": 0.9189, + "step": 4546 + }, + { + "epoch": 0.8096509971509972, + "grad_norm": 0.47902220487594604, + "learning_rate": 0.00018058625362214742, + "loss": 1.0389, + "step": 4547 + }, + { + "epoch": 0.8098290598290598, + "grad_norm": 0.6274173259735107, + "learning_rate": 0.00018057796489205936, + "loss": 1.3368, + "step": 4548 + }, + { + "epoch": 0.8100071225071225, + "grad_norm": 0.5789401531219482, + "learning_rate": 0.00018056967458321345, + "loss": 1.1473, + "step": 4549 + }, + { + "epoch": 0.8101851851851852, + "grad_norm": 0.5850043296813965, + "learning_rate": 0.0001805613826957721, + "loss": 1.2224, + "step": 4550 + }, + { + "epoch": 0.8103632478632479, + "grad_norm": 0.6310738921165466, + "learning_rate": 0.00018055308922989788, + "loss": 1.0707, + "step": 4551 + }, + { + "epoch": 0.8105413105413105, + "grad_norm": 0.5198429822921753, + "learning_rate": 0.00018054479418575317, + "loss": 0.8984, + "step": 4552 + }, + { + "epoch": 0.8107193732193733, + "grad_norm": 0.5757743120193481, + "learning_rate": 0.00018053649756350054, + "loss": 1.2007, + "step": 4553 + }, + { + "epoch": 0.8108974358974359, + "grad_norm": 0.5109567642211914, + "learning_rate": 0.0001805281993633025, + "loss": 1.0696, + "step": 4554 + }, + { + "epoch": 0.8110754985754985, + "grad_norm": 0.5030225515365601, + "learning_rate": 0.00018051989958532173, + "loss": 0.9667, + "step": 4555 + }, + { + "epoch": 0.8112535612535613, + "grad_norm": 0.5291743874549866, + "learning_rate": 0.00018051159822972079, + "loss": 1.0219, + "step": 4556 + }, + { + "epoch": 0.8114316239316239, + "grad_norm": 0.5874896049499512, + "learning_rate": 0.00018050329529666233, + "loss": 0.8589, + "step": 4557 + }, + { + "epoch": 0.8116096866096866, + "grad_norm": 0.673284113407135, + "learning_rate": 0.000180494990786309, + "loss": 1.1902, + "step": 4558 + }, + { + "epoch": 0.8117877492877493, + "grad_norm": 0.4742524027824402, + "learning_rate": 0.00018048668469882354, + "loss": 1.0578, + "step": 4559 + }, + { + "epoch": 0.811965811965812, + "grad_norm": 0.5519167184829712, + "learning_rate": 0.0001804783770343687, + "loss": 1.083, + "step": 4560 + }, + { + "epoch": 0.8121438746438746, + "grad_norm": 0.5669941306114197, + "learning_rate": 0.00018047006779310727, + "loss": 1.0784, + "step": 4561 + }, + { + "epoch": 0.8123219373219374, + "grad_norm": 0.512759804725647, + "learning_rate": 0.000180461756975202, + "loss": 1.0361, + "step": 4562 + }, + { + "epoch": 0.8125, + "grad_norm": 0.5721749067306519, + "learning_rate": 0.00018045344458081575, + "loss": 1.0246, + "step": 4563 + }, + { + "epoch": 0.8126780626780626, + "grad_norm": 0.566430389881134, + "learning_rate": 0.00018044513061011137, + "loss": 1.1452, + "step": 4564 + }, + { + "epoch": 0.8128561253561254, + "grad_norm": 0.49391916394233704, + "learning_rate": 0.00018043681506325177, + "loss": 0.89, + "step": 4565 + }, + { + "epoch": 0.813034188034188, + "grad_norm": 0.5379437804222107, + "learning_rate": 0.00018042849794039988, + "loss": 1.1289, + "step": 4566 + }, + { + "epoch": 0.8132122507122507, + "grad_norm": 0.5667982697486877, + "learning_rate": 0.00018042017924171865, + "loss": 1.1596, + "step": 4567 + }, + { + "epoch": 0.8133903133903134, + "grad_norm": 0.6214209794998169, + "learning_rate": 0.00018041185896737109, + "loss": 1.0622, + "step": 4568 + }, + { + "epoch": 0.8135683760683761, + "grad_norm": 0.5442491173744202, + "learning_rate": 0.00018040353711752015, + "loss": 1.0536, + "step": 4569 + }, + { + "epoch": 0.8137464387464387, + "grad_norm": 0.5266172885894775, + "learning_rate": 0.00018039521369232894, + "loss": 1.0576, + "step": 4570 + }, + { + "epoch": 0.8139245014245015, + "grad_norm": 0.6057912111282349, + "learning_rate": 0.00018038688869196053, + "loss": 1.3067, + "step": 4571 + }, + { + "epoch": 0.8141025641025641, + "grad_norm": 0.489869087934494, + "learning_rate": 0.00018037856211657803, + "loss": 1.0279, + "step": 4572 + }, + { + "epoch": 0.8142806267806267, + "grad_norm": 0.5497978329658508, + "learning_rate": 0.00018037023396634457, + "loss": 1.1568, + "step": 4573 + }, + { + "epoch": 0.8144586894586895, + "grad_norm": 0.5243251919746399, + "learning_rate": 0.0001803619042414233, + "loss": 0.9767, + "step": 4574 + }, + { + "epoch": 0.8146367521367521, + "grad_norm": 0.503032922744751, + "learning_rate": 0.0001803535729419775, + "loss": 1.065, + "step": 4575 + }, + { + "epoch": 0.8148148148148148, + "grad_norm": 0.49955418705940247, + "learning_rate": 0.00018034524006817034, + "loss": 1.2752, + "step": 4576 + }, + { + "epoch": 0.8149928774928775, + "grad_norm": 0.5746406316757202, + "learning_rate": 0.00018033690562016508, + "loss": 1.098, + "step": 4577 + }, + { + "epoch": 0.8151709401709402, + "grad_norm": 0.5224192142486572, + "learning_rate": 0.00018032856959812507, + "loss": 1.1284, + "step": 4578 + }, + { + "epoch": 0.8153490028490028, + "grad_norm": 0.5484535694122314, + "learning_rate": 0.00018032023200221362, + "loss": 0.9182, + "step": 4579 + }, + { + "epoch": 0.8155270655270656, + "grad_norm": 0.5003355741500854, + "learning_rate": 0.00018031189283259405, + "loss": 1.136, + "step": 4580 + }, + { + "epoch": 0.8157051282051282, + "grad_norm": 0.5395768284797668, + "learning_rate": 0.00018030355208942977, + "loss": 1.2349, + "step": 4581 + }, + { + "epoch": 0.8158831908831908, + "grad_norm": 0.561966598033905, + "learning_rate": 0.0001802952097728842, + "loss": 0.999, + "step": 4582 + }, + { + "epoch": 0.8160612535612536, + "grad_norm": 0.4886479675769806, + "learning_rate": 0.00018028686588312083, + "loss": 0.9165, + "step": 4583 + }, + { + "epoch": 0.8162393162393162, + "grad_norm": 0.4769509732723236, + "learning_rate": 0.00018027852042030307, + "loss": 1.1377, + "step": 4584 + }, + { + "epoch": 0.8164173789173789, + "grad_norm": 0.4723633825778961, + "learning_rate": 0.00018027017338459448, + "loss": 1.0274, + "step": 4585 + }, + { + "epoch": 0.8165954415954416, + "grad_norm": 0.5773285627365112, + "learning_rate": 0.00018026182477615859, + "loss": 1.1468, + "step": 4586 + }, + { + "epoch": 0.8167735042735043, + "grad_norm": 0.5529203414916992, + "learning_rate": 0.00018025347459515895, + "loss": 1.0815, + "step": 4587 + }, + { + "epoch": 0.8169515669515669, + "grad_norm": 0.5449469685554504, + "learning_rate": 0.00018024512284175922, + "loss": 1.1637, + "step": 4588 + }, + { + "epoch": 0.8171296296296297, + "grad_norm": 0.5155341625213623, + "learning_rate": 0.00018023676951612298, + "loss": 1.1842, + "step": 4589 + }, + { + "epoch": 0.8173076923076923, + "grad_norm": 0.5569564700126648, + "learning_rate": 0.00018022841461841393, + "loss": 0.9254, + "step": 4590 + }, + { + "epoch": 0.8174857549857549, + "grad_norm": 0.45203131437301636, + "learning_rate": 0.00018022005814879573, + "loss": 0.9561, + "step": 4591 + }, + { + "epoch": 0.8176638176638177, + "grad_norm": 0.5735056400299072, + "learning_rate": 0.00018021170010743218, + "loss": 1.1402, + "step": 4592 + }, + { + "epoch": 0.8178418803418803, + "grad_norm": 0.6075260043144226, + "learning_rate": 0.00018020334049448697, + "loss": 0.8601, + "step": 4593 + }, + { + "epoch": 0.8180199430199431, + "grad_norm": 0.522682785987854, + "learning_rate": 0.0001801949793101239, + "loss": 1.0088, + "step": 4594 + }, + { + "epoch": 0.8181980056980057, + "grad_norm": 0.5648437142372131, + "learning_rate": 0.00018018661655450682, + "loss": 0.8359, + "step": 4595 + }, + { + "epoch": 0.8183760683760684, + "grad_norm": 0.5406472086906433, + "learning_rate": 0.00018017825222779954, + "loss": 1.1553, + "step": 4596 + }, + { + "epoch": 0.8185541310541311, + "grad_norm": 0.4917788803577423, + "learning_rate": 0.000180169886330166, + "loss": 1.2198, + "step": 4597 + }, + { + "epoch": 0.8187321937321937, + "grad_norm": 0.6293069124221802, + "learning_rate": 0.00018016151886177004, + "loss": 1.0245, + "step": 4598 + }, + { + "epoch": 0.8189102564102564, + "grad_norm": 0.47277843952178955, + "learning_rate": 0.00018015314982277564, + "loss": 1.1141, + "step": 4599 + }, + { + "epoch": 0.8190883190883191, + "grad_norm": 0.6132395267486572, + "learning_rate": 0.0001801447792133468, + "loss": 1.1227, + "step": 4600 + }, + { + "epoch": 0.8192663817663818, + "grad_norm": 0.46839597821235657, + "learning_rate": 0.00018013640703364747, + "loss": 0.9239, + "step": 4601 + }, + { + "epoch": 0.8194444444444444, + "grad_norm": 0.5055009722709656, + "learning_rate": 0.00018012803328384171, + "loss": 0.8486, + "step": 4602 + }, + { + "epoch": 0.8196225071225072, + "grad_norm": 0.5094841718673706, + "learning_rate": 0.00018011965796409362, + "loss": 0.9969, + "step": 4603 + }, + { + "epoch": 0.8198005698005698, + "grad_norm": 0.6177363395690918, + "learning_rate": 0.00018011128107456726, + "loss": 1.242, + "step": 4604 + }, + { + "epoch": 0.8199786324786325, + "grad_norm": 0.5280042290687561, + "learning_rate": 0.00018010290261542676, + "loss": 1.1569, + "step": 4605 + }, + { + "epoch": 0.8201566951566952, + "grad_norm": 0.5259367227554321, + "learning_rate": 0.00018009452258683625, + "loss": 0.9993, + "step": 4606 + }, + { + "epoch": 0.8203347578347578, + "grad_norm": 0.464469850063324, + "learning_rate": 0.00018008614098896, + "loss": 1.0288, + "step": 4607 + }, + { + "epoch": 0.8205128205128205, + "grad_norm": 0.6136324405670166, + "learning_rate": 0.00018007775782196214, + "loss": 1.1541, + "step": 4608 + }, + { + "epoch": 0.8206908831908832, + "grad_norm": 0.5376590490341187, + "learning_rate": 0.000180069373086007, + "loss": 1.0624, + "step": 4609 + }, + { + "epoch": 0.8208689458689459, + "grad_norm": 0.662916362285614, + "learning_rate": 0.0001800609867812588, + "loss": 1.1502, + "step": 4610 + }, + { + "epoch": 0.8210470085470085, + "grad_norm": 0.5153383612632751, + "learning_rate": 0.00018005259890788188, + "loss": 0.9789, + "step": 4611 + }, + { + "epoch": 0.8212250712250713, + "grad_norm": 0.5042359232902527, + "learning_rate": 0.00018004420946604057, + "loss": 0.9585, + "step": 4612 + }, + { + "epoch": 0.8214031339031339, + "grad_norm": 0.5395993590354919, + "learning_rate": 0.00018003581845589927, + "loss": 1.159, + "step": 4613 + }, + { + "epoch": 0.8215811965811965, + "grad_norm": 0.5561928749084473, + "learning_rate": 0.00018002742587762237, + "loss": 1.1604, + "step": 4614 + }, + { + "epoch": 0.8217592592592593, + "grad_norm": 0.5602710843086243, + "learning_rate": 0.00018001903173137432, + "loss": 0.9922, + "step": 4615 + }, + { + "epoch": 0.8219373219373219, + "grad_norm": 0.5529088377952576, + "learning_rate": 0.00018001063601731955, + "loss": 1.0943, + "step": 4616 + }, + { + "epoch": 0.8221153846153846, + "grad_norm": 0.5156456828117371, + "learning_rate": 0.00018000223873562254, + "loss": 1.1399, + "step": 4617 + }, + { + "epoch": 0.8222934472934473, + "grad_norm": 0.4868306517601013, + "learning_rate": 0.0001799938398864479, + "loss": 1.0692, + "step": 4618 + }, + { + "epoch": 0.82247150997151, + "grad_norm": 0.5372915267944336, + "learning_rate": 0.0001799854394699601, + "loss": 1.2675, + "step": 4619 + }, + { + "epoch": 0.8226495726495726, + "grad_norm": 0.6101839542388916, + "learning_rate": 0.0001799770374863238, + "loss": 0.9586, + "step": 4620 + }, + { + "epoch": 0.8228276353276354, + "grad_norm": 0.5034586787223816, + "learning_rate": 0.00017996863393570357, + "loss": 1.0885, + "step": 4621 + }, + { + "epoch": 0.823005698005698, + "grad_norm": 0.5608823299407959, + "learning_rate": 0.0001799602288182641, + "loss": 1.0002, + "step": 4622 + }, + { + "epoch": 0.8231837606837606, + "grad_norm": 0.5700048208236694, + "learning_rate": 0.00017995182213417, + "loss": 1.1484, + "step": 4623 + }, + { + "epoch": 0.8233618233618234, + "grad_norm": 0.5283229351043701, + "learning_rate": 0.00017994341388358608, + "loss": 1.0744, + "step": 4624 + }, + { + "epoch": 0.823539886039886, + "grad_norm": 0.5215758681297302, + "learning_rate": 0.00017993500406667703, + "loss": 1.2686, + "step": 4625 + }, + { + "epoch": 0.8237179487179487, + "grad_norm": 0.528883159160614, + "learning_rate": 0.0001799265926836076, + "loss": 1.1393, + "step": 4626 + }, + { + "epoch": 0.8238960113960114, + "grad_norm": 0.5589834451675415, + "learning_rate": 0.00017991817973454265, + "loss": 1.1744, + "step": 4627 + }, + { + "epoch": 0.8240740740740741, + "grad_norm": 0.49817174673080444, + "learning_rate": 0.00017990976521964697, + "loss": 1.0544, + "step": 4628 + }, + { + "epoch": 0.8242521367521367, + "grad_norm": 0.613961398601532, + "learning_rate": 0.00017990134913908542, + "loss": 1.0951, + "step": 4629 + }, + { + "epoch": 0.8244301994301995, + "grad_norm": 0.47278255224227905, + "learning_rate": 0.00017989293149302295, + "loss": 0.9742, + "step": 4630 + }, + { + "epoch": 0.8246082621082621, + "grad_norm": 0.49807092547416687, + "learning_rate": 0.00017988451228162443, + "loss": 1.0985, + "step": 4631 + }, + { + "epoch": 0.8247863247863247, + "grad_norm": 0.5624374747276306, + "learning_rate": 0.00017987609150505485, + "loss": 1.2446, + "step": 4632 + }, + { + "epoch": 0.8249643874643875, + "grad_norm": 0.4863535761833191, + "learning_rate": 0.00017986766916347916, + "loss": 1.0239, + "step": 4633 + }, + { + "epoch": 0.8251424501424501, + "grad_norm": 0.679585874080658, + "learning_rate": 0.00017985924525706245, + "loss": 1.1698, + "step": 4634 + }, + { + "epoch": 0.8253205128205128, + "grad_norm": 0.5545455813407898, + "learning_rate": 0.00017985081978596967, + "loss": 1.0926, + "step": 4635 + }, + { + "epoch": 0.8254985754985755, + "grad_norm": 0.5303109288215637, + "learning_rate": 0.000179842392750366, + "loss": 1.0978, + "step": 4636 + }, + { + "epoch": 0.8256766381766382, + "grad_norm": 0.6053299307823181, + "learning_rate": 0.00017983396415041644, + "loss": 1.0596, + "step": 4637 + }, + { + "epoch": 0.8258547008547008, + "grad_norm": 0.5241885185241699, + "learning_rate": 0.00017982553398628625, + "loss": 0.8541, + "step": 4638 + }, + { + "epoch": 0.8260327635327636, + "grad_norm": 0.5934443473815918, + "learning_rate": 0.00017981710225814052, + "loss": 1.145, + "step": 4639 + }, + { + "epoch": 0.8262108262108262, + "grad_norm": 0.5341619849205017, + "learning_rate": 0.00017980866896614447, + "loss": 1.0745, + "step": 4640 + }, + { + "epoch": 0.8263888888888888, + "grad_norm": 0.6732913851737976, + "learning_rate": 0.00017980023411046336, + "loss": 1.0775, + "step": 4641 + }, + { + "epoch": 0.8265669515669516, + "grad_norm": 0.5134359002113342, + "learning_rate": 0.0001797917976912624, + "loss": 1.0298, + "step": 4642 + }, + { + "epoch": 0.8267450142450142, + "grad_norm": 0.5234783887863159, + "learning_rate": 0.00017978335970870698, + "loss": 1.1069, + "step": 4643 + }, + { + "epoch": 0.8269230769230769, + "grad_norm": 0.4776439964771271, + "learning_rate": 0.00017977492016296232, + "loss": 0.6367, + "step": 4644 + }, + { + "epoch": 0.8271011396011396, + "grad_norm": 0.53763347864151, + "learning_rate": 0.0001797664790541938, + "loss": 1.1356, + "step": 4645 + }, + { + "epoch": 0.8272792022792023, + "grad_norm": 0.5082212686538696, + "learning_rate": 0.00017975803638256682, + "loss": 0.7873, + "step": 4646 + }, + { + "epoch": 0.8274572649572649, + "grad_norm": 0.5156424641609192, + "learning_rate": 0.00017974959214824685, + "loss": 1.084, + "step": 4647 + }, + { + "epoch": 0.8276353276353277, + "grad_norm": 0.5275198817253113, + "learning_rate": 0.00017974114635139926, + "loss": 1.1219, + "step": 4648 + }, + { + "epoch": 0.8278133903133903, + "grad_norm": 0.5548223257064819, + "learning_rate": 0.00017973269899218956, + "loss": 1.0808, + "step": 4649 + }, + { + "epoch": 0.8279914529914529, + "grad_norm": 0.535347580909729, + "learning_rate": 0.00017972425007078323, + "loss": 1.1211, + "step": 4650 + }, + { + "epoch": 0.8281695156695157, + "grad_norm": 0.5299580693244934, + "learning_rate": 0.00017971579958734587, + "loss": 0.9911, + "step": 4651 + }, + { + "epoch": 0.8283475783475783, + "grad_norm": 0.4863550066947937, + "learning_rate": 0.000179707347542043, + "loss": 0.9122, + "step": 4652 + }, + { + "epoch": 0.8285256410256411, + "grad_norm": 0.5284972190856934, + "learning_rate": 0.00017969889393504022, + "loss": 1.0424, + "step": 4653 + }, + { + "epoch": 0.8287037037037037, + "grad_norm": 0.5305661559104919, + "learning_rate": 0.00017969043876650317, + "loss": 1.1122, + "step": 4654 + }, + { + "epoch": 0.8288817663817664, + "grad_norm": 0.5645657777786255, + "learning_rate": 0.00017968198203659755, + "loss": 1.2195, + "step": 4655 + }, + { + "epoch": 0.8290598290598291, + "grad_norm": 0.521649181842804, + "learning_rate": 0.000179673523745489, + "loss": 1.2684, + "step": 4656 + }, + { + "epoch": 0.8292378917378918, + "grad_norm": 0.5984422564506531, + "learning_rate": 0.00017966506389334322, + "loss": 0.9894, + "step": 4657 + }, + { + "epoch": 0.8294159544159544, + "grad_norm": 0.5318729281425476, + "learning_rate": 0.00017965660248032603, + "loss": 1.2929, + "step": 4658 + }, + { + "epoch": 0.8295940170940171, + "grad_norm": 0.4666081368923187, + "learning_rate": 0.0001796481395066032, + "loss": 0.9646, + "step": 4659 + }, + { + "epoch": 0.8297720797720798, + "grad_norm": 0.5780388116836548, + "learning_rate": 0.00017963967497234054, + "loss": 1.1043, + "step": 4660 + }, + { + "epoch": 0.8299501424501424, + "grad_norm": 0.44089245796203613, + "learning_rate": 0.00017963120887770387, + "loss": 0.8932, + "step": 4661 + }, + { + "epoch": 0.8301282051282052, + "grad_norm": 0.5198349356651306, + "learning_rate": 0.0001796227412228591, + "loss": 0.9378, + "step": 4662 + }, + { + "epoch": 0.8303062678062678, + "grad_norm": 0.5298343896865845, + "learning_rate": 0.00017961427200797206, + "loss": 1.0272, + "step": 4663 + }, + { + "epoch": 0.8304843304843305, + "grad_norm": 0.5087099671363831, + "learning_rate": 0.0001796058012332088, + "loss": 0.989, + "step": 4664 + }, + { + "epoch": 0.8306623931623932, + "grad_norm": 0.504228949546814, + "learning_rate": 0.0001795973288987352, + "loss": 1.0134, + "step": 4665 + }, + { + "epoch": 0.8308404558404558, + "grad_norm": 0.6788033843040466, + "learning_rate": 0.00017958885500471728, + "loss": 0.8856, + "step": 4666 + }, + { + "epoch": 0.8310185185185185, + "grad_norm": 0.5166172385215759, + "learning_rate": 0.00017958037955132113, + "loss": 0.8711, + "step": 4667 + }, + { + "epoch": 0.8311965811965812, + "grad_norm": 0.5712400078773499, + "learning_rate": 0.00017957190253871272, + "loss": 1.0418, + "step": 4668 + }, + { + "epoch": 0.8313746438746439, + "grad_norm": 0.5531231164932251, + "learning_rate": 0.0001795634239670582, + "loss": 0.9021, + "step": 4669 + }, + { + "epoch": 0.8315527065527065, + "grad_norm": 0.6165615916252136, + "learning_rate": 0.00017955494383652365, + "loss": 1.0927, + "step": 4670 + }, + { + "epoch": 0.8317307692307693, + "grad_norm": 0.5920368432998657, + "learning_rate": 0.00017954646214727525, + "loss": 1.231, + "step": 4671 + }, + { + "epoch": 0.8319088319088319, + "grad_norm": 0.5037244558334351, + "learning_rate": 0.00017953797889947915, + "loss": 0.85, + "step": 4672 + }, + { + "epoch": 0.8320868945868946, + "grad_norm": 0.5618211627006531, + "learning_rate": 0.0001795294940933016, + "loss": 1.145, + "step": 4673 + }, + { + "epoch": 0.8322649572649573, + "grad_norm": 0.6275593042373657, + "learning_rate": 0.00017952100772890877, + "loss": 0.9061, + "step": 4674 + }, + { + "epoch": 0.83244301994302, + "grad_norm": 0.5376096367835999, + "learning_rate": 0.00017951251980646702, + "loss": 1.1948, + "step": 4675 + }, + { + "epoch": 0.8326210826210826, + "grad_norm": 0.5162268877029419, + "learning_rate": 0.0001795040303261426, + "loss": 1.2158, + "step": 4676 + }, + { + "epoch": 0.8327991452991453, + "grad_norm": 0.5730512142181396, + "learning_rate": 0.0001794955392881019, + "loss": 0.9962, + "step": 4677 + }, + { + "epoch": 0.832977207977208, + "grad_norm": 0.5128712058067322, + "learning_rate": 0.00017948704669251122, + "loss": 1.2797, + "step": 4678 + }, + { + "epoch": 0.8331552706552706, + "grad_norm": 0.5173979997634888, + "learning_rate": 0.00017947855253953697, + "loss": 1.1093, + "step": 4679 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 0.504646897315979, + "learning_rate": 0.0001794700568293456, + "loss": 1.3171, + "step": 4680 + }, + { + "epoch": 0.833511396011396, + "grad_norm": 0.5638105869293213, + "learning_rate": 0.00017946155956210356, + "loss": 0.9224, + "step": 4681 + }, + { + "epoch": 0.8336894586894587, + "grad_norm": 0.5289680361747742, + "learning_rate": 0.00017945306073797733, + "loss": 0.8919, + "step": 4682 + }, + { + "epoch": 0.8338675213675214, + "grad_norm": 0.5224629044532776, + "learning_rate": 0.0001794445603571334, + "loss": 1.0345, + "step": 4683 + }, + { + "epoch": 0.834045584045584, + "grad_norm": 0.5342282056808472, + "learning_rate": 0.00017943605841973836, + "loss": 1.2305, + "step": 4684 + }, + { + "epoch": 0.8342236467236467, + "grad_norm": 0.6118032336235046, + "learning_rate": 0.00017942755492595874, + "loss": 1.0316, + "step": 4685 + }, + { + "epoch": 0.8344017094017094, + "grad_norm": 0.49112311005592346, + "learning_rate": 0.00017941904987596121, + "loss": 0.9809, + "step": 4686 + }, + { + "epoch": 0.8345797720797721, + "grad_norm": 0.5044063925743103, + "learning_rate": 0.0001794105432699124, + "loss": 0.834, + "step": 4687 + }, + { + "epoch": 0.8347578347578347, + "grad_norm": 0.4849987328052521, + "learning_rate": 0.00017940203510797892, + "loss": 0.9971, + "step": 4688 + }, + { + "epoch": 0.8349358974358975, + "grad_norm": 0.5539469122886658, + "learning_rate": 0.00017939352539032748, + "loss": 1.1599, + "step": 4689 + }, + { + "epoch": 0.8351139601139601, + "grad_norm": 0.5474258065223694, + "learning_rate": 0.00017938501411712485, + "loss": 1.25, + "step": 4690 + }, + { + "epoch": 0.8352920227920227, + "grad_norm": 0.4880213737487793, + "learning_rate": 0.0001793765012885378, + "loss": 1.1471, + "step": 4691 + }, + { + "epoch": 0.8354700854700855, + "grad_norm": 0.5602759718894958, + "learning_rate": 0.00017936798690473309, + "loss": 1.0723, + "step": 4692 + }, + { + "epoch": 0.8356481481481481, + "grad_norm": 0.627775251865387, + "learning_rate": 0.00017935947096587755, + "loss": 1.3768, + "step": 4693 + }, + { + "epoch": 0.8358262108262108, + "grad_norm": 0.5324847102165222, + "learning_rate": 0.00017935095347213804, + "loss": 0.9945, + "step": 4694 + }, + { + "epoch": 0.8360042735042735, + "grad_norm": 0.5244048237800598, + "learning_rate": 0.0001793424344236814, + "loss": 1.1725, + "step": 4695 + }, + { + "epoch": 0.8361823361823362, + "grad_norm": 0.5420708656311035, + "learning_rate": 0.00017933391382067462, + "loss": 1.1267, + "step": 4696 + }, + { + "epoch": 0.8363603988603988, + "grad_norm": 0.5285456776618958, + "learning_rate": 0.00017932539166328458, + "loss": 1.0368, + "step": 4697 + }, + { + "epoch": 0.8365384615384616, + "grad_norm": 0.5330373048782349, + "learning_rate": 0.00017931686795167825, + "loss": 1.1082, + "step": 4698 + }, + { + "epoch": 0.8367165242165242, + "grad_norm": 0.5516682267189026, + "learning_rate": 0.0001793083426860227, + "loss": 1.1833, + "step": 4699 + }, + { + "epoch": 0.8368945868945868, + "grad_norm": 0.5229935646057129, + "learning_rate": 0.0001792998158664849, + "loss": 0.8527, + "step": 4700 + }, + { + "epoch": 0.8370726495726496, + "grad_norm": 0.4821490943431854, + "learning_rate": 0.00017929128749323195, + "loss": 1.1201, + "step": 4701 + }, + { + "epoch": 0.8372507122507122, + "grad_norm": 0.6276404857635498, + "learning_rate": 0.0001792827575664309, + "loss": 1.0986, + "step": 4702 + }, + { + "epoch": 0.8374287749287749, + "grad_norm": 0.5681334733963013, + "learning_rate": 0.00017927422608624897, + "loss": 1.3821, + "step": 4703 + }, + { + "epoch": 0.8376068376068376, + "grad_norm": 0.5257087349891663, + "learning_rate": 0.00017926569305285324, + "loss": 1.1033, + "step": 4704 + }, + { + "epoch": 0.8377849002849003, + "grad_norm": 0.5665168166160583, + "learning_rate": 0.0001792571584664109, + "loss": 1.104, + "step": 4705 + }, + { + "epoch": 0.8379629629629629, + "grad_norm": 0.5202076435089111, + "learning_rate": 0.00017924862232708918, + "loss": 1.052, + "step": 4706 + }, + { + "epoch": 0.8381410256410257, + "grad_norm": 0.5103010535240173, + "learning_rate": 0.00017924008463505534, + "loss": 1.1348, + "step": 4707 + }, + { + "epoch": 0.8383190883190883, + "grad_norm": 0.6811865568161011, + "learning_rate": 0.00017923154539047667, + "loss": 1.2804, + "step": 4708 + }, + { + "epoch": 0.8384971509971509, + "grad_norm": 0.46808311343193054, + "learning_rate": 0.00017922300459352042, + "loss": 0.9302, + "step": 4709 + }, + { + "epoch": 0.8386752136752137, + "grad_norm": 0.47713059186935425, + "learning_rate": 0.00017921446224435398, + "loss": 0.78, + "step": 4710 + }, + { + "epoch": 0.8388532763532763, + "grad_norm": 0.7579890489578247, + "learning_rate": 0.0001792059183431447, + "loss": 1.4776, + "step": 4711 + }, + { + "epoch": 0.8390313390313391, + "grad_norm": 0.6009423136711121, + "learning_rate": 0.00017919737289006, + "loss": 1.2679, + "step": 4712 + }, + { + "epoch": 0.8392094017094017, + "grad_norm": 0.56390780210495, + "learning_rate": 0.00017918882588526729, + "loss": 1.0402, + "step": 4713 + }, + { + "epoch": 0.8393874643874644, + "grad_norm": 0.5698862075805664, + "learning_rate": 0.00017918027732893404, + "loss": 1.2336, + "step": 4714 + }, + { + "epoch": 0.8395655270655271, + "grad_norm": 0.5016305446624756, + "learning_rate": 0.0001791717272212277, + "loss": 1.0373, + "step": 4715 + }, + { + "epoch": 0.8397435897435898, + "grad_norm": 0.5886971950531006, + "learning_rate": 0.0001791631755623159, + "loss": 1.1062, + "step": 4716 + }, + { + "epoch": 0.8399216524216524, + "grad_norm": 0.647833526134491, + "learning_rate": 0.00017915462235236607, + "loss": 1.0464, + "step": 4717 + }, + { + "epoch": 0.8400997150997151, + "grad_norm": 0.4961194396018982, + "learning_rate": 0.00017914606759154587, + "loss": 1.0763, + "step": 4718 + }, + { + "epoch": 0.8402777777777778, + "grad_norm": 0.47041359543800354, + "learning_rate": 0.00017913751128002288, + "loss": 1.0685, + "step": 4719 + }, + { + "epoch": 0.8404558404558404, + "grad_norm": 0.5752858519554138, + "learning_rate": 0.00017912895341796475, + "loss": 1.0577, + "step": 4720 + }, + { + "epoch": 0.8406339031339032, + "grad_norm": 0.5233224034309387, + "learning_rate": 0.00017912039400553914, + "loss": 1.1484, + "step": 4721 + }, + { + "epoch": 0.8408119658119658, + "grad_norm": 0.5327485203742981, + "learning_rate": 0.00017911183304291378, + "loss": 1.0028, + "step": 4722 + }, + { + "epoch": 0.8409900284900285, + "grad_norm": 0.5320752263069153, + "learning_rate": 0.00017910327053025638, + "loss": 1.1247, + "step": 4723 + }, + { + "epoch": 0.8411680911680912, + "grad_norm": 0.529617965221405, + "learning_rate": 0.00017909470646773477, + "loss": 1.1698, + "step": 4724 + }, + { + "epoch": 0.8413461538461539, + "grad_norm": 0.5055609345436096, + "learning_rate": 0.00017908614085551664, + "loss": 1.0925, + "step": 4725 + }, + { + "epoch": 0.8415242165242165, + "grad_norm": 0.5356255769729614, + "learning_rate": 0.00017907757369376985, + "loss": 1.0354, + "step": 4726 + }, + { + "epoch": 0.8417022792022792, + "grad_norm": 0.582834780216217, + "learning_rate": 0.00017906900498266233, + "loss": 1.1248, + "step": 4727 + }, + { + "epoch": 0.8418803418803419, + "grad_norm": 0.5750834941864014, + "learning_rate": 0.00017906043472236188, + "loss": 1.0119, + "step": 4728 + }, + { + "epoch": 0.8420584045584045, + "grad_norm": 0.5923320055007935, + "learning_rate": 0.00017905186291303644, + "loss": 1.0662, + "step": 4729 + }, + { + "epoch": 0.8422364672364673, + "grad_norm": 0.4767811894416809, + "learning_rate": 0.00017904328955485396, + "loss": 1.0911, + "step": 4730 + }, + { + "epoch": 0.8424145299145299, + "grad_norm": 0.5294556021690369, + "learning_rate": 0.00017903471464798245, + "loss": 1.2861, + "step": 4731 + }, + { + "epoch": 0.8425925925925926, + "grad_norm": 0.599117636680603, + "learning_rate": 0.00017902613819258985, + "loss": 1.1707, + "step": 4732 + }, + { + "epoch": 0.8427706552706553, + "grad_norm": 0.5912977457046509, + "learning_rate": 0.00017901756018884424, + "loss": 1.1884, + "step": 4733 + }, + { + "epoch": 0.842948717948718, + "grad_norm": 0.587676465511322, + "learning_rate": 0.0001790089806369137, + "loss": 1.1054, + "step": 4734 + }, + { + "epoch": 0.8431267806267806, + "grad_norm": 0.6271800398826599, + "learning_rate": 0.0001790003995369663, + "loss": 1.2094, + "step": 4735 + }, + { + "epoch": 0.8433048433048433, + "grad_norm": 0.47198590636253357, + "learning_rate": 0.00017899181688917017, + "loss": 0.9561, + "step": 4736 + }, + { + "epoch": 0.843482905982906, + "grad_norm": 0.690732479095459, + "learning_rate": 0.00017898323269369351, + "loss": 1.1629, + "step": 4737 + }, + { + "epoch": 0.8436609686609686, + "grad_norm": 0.4926888048648834, + "learning_rate": 0.00017897464695070445, + "loss": 1.1097, + "step": 4738 + }, + { + "epoch": 0.8438390313390314, + "grad_norm": 0.7071278691291809, + "learning_rate": 0.00017896605966037128, + "loss": 1.195, + "step": 4739 + }, + { + "epoch": 0.844017094017094, + "grad_norm": 0.5650486350059509, + "learning_rate": 0.00017895747082286216, + "loss": 1.0107, + "step": 4740 + }, + { + "epoch": 0.8441951566951567, + "grad_norm": 0.5291931629180908, + "learning_rate": 0.00017894888043834545, + "loss": 1.0104, + "step": 4741 + }, + { + "epoch": 0.8443732193732194, + "grad_norm": 0.5751241445541382, + "learning_rate": 0.00017894028850698942, + "loss": 1.2482, + "step": 4742 + }, + { + "epoch": 0.844551282051282, + "grad_norm": 0.5833632349967957, + "learning_rate": 0.0001789316950289624, + "loss": 1.0552, + "step": 4743 + }, + { + "epoch": 0.8447293447293447, + "grad_norm": 0.543729841709137, + "learning_rate": 0.00017892310000443282, + "loss": 1.1453, + "step": 4744 + }, + { + "epoch": 0.8449074074074074, + "grad_norm": 0.5674204230308533, + "learning_rate": 0.00017891450343356902, + "loss": 1.0757, + "step": 4745 + }, + { + "epoch": 0.8450854700854701, + "grad_norm": 0.5161892771720886, + "learning_rate": 0.00017890590531653946, + "loss": 1.1163, + "step": 4746 + }, + { + "epoch": 0.8452635327635327, + "grad_norm": 0.49907612800598145, + "learning_rate": 0.00017889730565351258, + "loss": 1.0356, + "step": 4747 + }, + { + "epoch": 0.8454415954415955, + "grad_norm": 0.4994732439517975, + "learning_rate": 0.00017888870444465692, + "loss": 1.026, + "step": 4748 + }, + { + "epoch": 0.8456196581196581, + "grad_norm": 0.6397520303726196, + "learning_rate": 0.00017888010169014095, + "loss": 0.957, + "step": 4749 + }, + { + "epoch": 0.8457977207977208, + "grad_norm": 0.5379729270935059, + "learning_rate": 0.00017887149739013327, + "loss": 1.1664, + "step": 4750 + }, + { + "epoch": 0.8459757834757835, + "grad_norm": 0.4487382769584656, + "learning_rate": 0.00017886289154480246, + "loss": 0.9377, + "step": 4751 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.5645943880081177, + "learning_rate": 0.00017885428415431707, + "loss": 1.273, + "step": 4752 + }, + { + "epoch": 0.8463319088319088, + "grad_norm": 0.5535289645195007, + "learning_rate": 0.00017884567521884577, + "loss": 1.1779, + "step": 4753 + }, + { + "epoch": 0.8465099715099715, + "grad_norm": 0.5039721131324768, + "learning_rate": 0.0001788370647385573, + "loss": 1.0237, + "step": 4754 + }, + { + "epoch": 0.8466880341880342, + "grad_norm": 0.4543854892253876, + "learning_rate": 0.00017882845271362032, + "loss": 0.8149, + "step": 4755 + }, + { + "epoch": 0.8468660968660968, + "grad_norm": 0.5095639824867249, + "learning_rate": 0.00017881983914420352, + "loss": 1.0141, + "step": 4756 + }, + { + "epoch": 0.8470441595441596, + "grad_norm": 0.5341798663139343, + "learning_rate": 0.00017881122403047575, + "loss": 1.1885, + "step": 4757 + }, + { + "epoch": 0.8472222222222222, + "grad_norm": 0.5595062971115112, + "learning_rate": 0.00017880260737260573, + "loss": 0.8939, + "step": 4758 + }, + { + "epoch": 0.8474002849002849, + "grad_norm": 0.5355880260467529, + "learning_rate": 0.00017879398917076232, + "loss": 1.2434, + "step": 4759 + }, + { + "epoch": 0.8475783475783476, + "grad_norm": 0.49477261304855347, + "learning_rate": 0.0001787853694251144, + "loss": 0.979, + "step": 4760 + }, + { + "epoch": 0.8477564102564102, + "grad_norm": 0.5154359340667725, + "learning_rate": 0.00017877674813583078, + "loss": 1.0957, + "step": 4761 + }, + { + "epoch": 0.8479344729344729, + "grad_norm": 0.5651070475578308, + "learning_rate": 0.00017876812530308046, + "loss": 1.1884, + "step": 4762 + }, + { + "epoch": 0.8481125356125356, + "grad_norm": 0.537277340888977, + "learning_rate": 0.00017875950092703232, + "loss": 1.0272, + "step": 4763 + }, + { + "epoch": 0.8482905982905983, + "grad_norm": 0.5259691476821899, + "learning_rate": 0.00017875087500785538, + "loss": 1.1493, + "step": 4764 + }, + { + "epoch": 0.8484686609686609, + "grad_norm": 0.5491300225257874, + "learning_rate": 0.00017874224754571867, + "loss": 0.8316, + "step": 4765 + }, + { + "epoch": 0.8486467236467237, + "grad_norm": 0.5493744611740112, + "learning_rate": 0.00017873361854079116, + "loss": 1.2328, + "step": 4766 + }, + { + "epoch": 0.8488247863247863, + "grad_norm": 0.571002185344696, + "learning_rate": 0.00017872498799324197, + "loss": 1.1384, + "step": 4767 + }, + { + "epoch": 0.8490028490028491, + "grad_norm": 0.538152813911438, + "learning_rate": 0.00017871635590324013, + "loss": 1.0581, + "step": 4768 + }, + { + "epoch": 0.8491809116809117, + "grad_norm": 0.5214923620223999, + "learning_rate": 0.00017870772227095486, + "loss": 1.0612, + "step": 4769 + }, + { + "epoch": 0.8493589743589743, + "grad_norm": 0.5714883804321289, + "learning_rate": 0.0001786990870965553, + "loss": 0.9076, + "step": 4770 + }, + { + "epoch": 0.8495370370370371, + "grad_norm": 0.4181775450706482, + "learning_rate": 0.00017869045038021054, + "loss": 0.8366, + "step": 4771 + }, + { + "epoch": 0.8497150997150997, + "grad_norm": 0.6266027688980103, + "learning_rate": 0.00017868181212208993, + "loss": 1.2047, + "step": 4772 + }, + { + "epoch": 0.8498931623931624, + "grad_norm": 0.5423732399940491, + "learning_rate": 0.0001786731723223626, + "loss": 1.3878, + "step": 4773 + }, + { + "epoch": 0.8500712250712251, + "grad_norm": 0.5512300133705139, + "learning_rate": 0.00017866453098119793, + "loss": 1.1132, + "step": 4774 + }, + { + "epoch": 0.8502492877492878, + "grad_norm": 0.5767185688018799, + "learning_rate": 0.00017865588809876519, + "loss": 0.97, + "step": 4775 + }, + { + "epoch": 0.8504273504273504, + "grad_norm": 0.5305790305137634, + "learning_rate": 0.00017864724367523368, + "loss": 1.1158, + "step": 4776 + }, + { + "epoch": 0.8506054131054132, + "grad_norm": 0.49702391028404236, + "learning_rate": 0.00017863859771077284, + "loss": 0.9669, + "step": 4777 + }, + { + "epoch": 0.8507834757834758, + "grad_norm": 0.5490063428878784, + "learning_rate": 0.00017862995020555205, + "loss": 1.0646, + "step": 4778 + }, + { + "epoch": 0.8509615384615384, + "grad_norm": 0.5308689475059509, + "learning_rate": 0.00017862130115974068, + "loss": 0.8922, + "step": 4779 + }, + { + "epoch": 0.8511396011396012, + "grad_norm": 0.5412983894348145, + "learning_rate": 0.00017861265057350826, + "loss": 1.1444, + "step": 4780 + }, + { + "epoch": 0.8513176638176638, + "grad_norm": 0.5857377052307129, + "learning_rate": 0.00017860399844702425, + "loss": 1.1643, + "step": 4781 + }, + { + "epoch": 0.8514957264957265, + "grad_norm": 0.599273681640625, + "learning_rate": 0.00017859534478045815, + "loss": 1.169, + "step": 4782 + }, + { + "epoch": 0.8516737891737892, + "grad_norm": 0.5677087903022766, + "learning_rate": 0.00017858668957397957, + "loss": 1.0793, + "step": 4783 + }, + { + "epoch": 0.8518518518518519, + "grad_norm": 0.5648362636566162, + "learning_rate": 0.00017857803282775807, + "loss": 1.1932, + "step": 4784 + }, + { + "epoch": 0.8520299145299145, + "grad_norm": 0.5138826966285706, + "learning_rate": 0.00017856937454196323, + "loss": 1.0011, + "step": 4785 + }, + { + "epoch": 0.8522079772079773, + "grad_norm": 0.5951429009437561, + "learning_rate": 0.0001785607147167647, + "loss": 1.3198, + "step": 4786 + }, + { + "epoch": 0.8523860398860399, + "grad_norm": 0.5341953039169312, + "learning_rate": 0.00017855205335233216, + "loss": 0.9094, + "step": 4787 + }, + { + "epoch": 0.8525641025641025, + "grad_norm": 0.5193579196929932, + "learning_rate": 0.00017854339044883535, + "loss": 0.892, + "step": 4788 + }, + { + "epoch": 0.8527421652421653, + "grad_norm": 0.5053097009658813, + "learning_rate": 0.00017853472600644392, + "loss": 1.0589, + "step": 4789 + }, + { + "epoch": 0.8529202279202279, + "grad_norm": 0.5819617509841919, + "learning_rate": 0.0001785260600253277, + "loss": 1.2646, + "step": 4790 + }, + { + "epoch": 0.8530982905982906, + "grad_norm": 0.5327470302581787, + "learning_rate": 0.00017851739250565645, + "loss": 1.056, + "step": 4791 + }, + { + "epoch": 0.8532763532763533, + "grad_norm": 0.5131269097328186, + "learning_rate": 0.0001785087234476, + "loss": 1.1192, + "step": 4792 + }, + { + "epoch": 0.853454415954416, + "grad_norm": 0.4698086977005005, + "learning_rate": 0.00017850005285132821, + "loss": 0.9849, + "step": 4793 + }, + { + "epoch": 0.8536324786324786, + "grad_norm": 0.5503947734832764, + "learning_rate": 0.00017849138071701092, + "loss": 1.1139, + "step": 4794 + }, + { + "epoch": 0.8538105413105413, + "grad_norm": 0.5120903849601746, + "learning_rate": 0.0001784827070448181, + "loss": 0.9801, + "step": 4795 + }, + { + "epoch": 0.853988603988604, + "grad_norm": 0.47650405764579773, + "learning_rate": 0.00017847403183491968, + "loss": 1.0268, + "step": 4796 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 0.5773387551307678, + "learning_rate": 0.0001784653550874856, + "loss": 1.0336, + "step": 4797 + }, + { + "epoch": 0.8543447293447294, + "grad_norm": 0.545531153678894, + "learning_rate": 0.00017845667680268593, + "loss": 1.0532, + "step": 4798 + }, + { + "epoch": 0.854522792022792, + "grad_norm": 0.533161461353302, + "learning_rate": 0.0001784479969806906, + "loss": 1.1964, + "step": 4799 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 0.5880789160728455, + "learning_rate": 0.00017843931562166977, + "loss": 1.1588, + "step": 4800 + }, + { + "epoch": 0.8548789173789174, + "grad_norm": 0.5381524562835693, + "learning_rate": 0.00017843063272579346, + "loss": 1.1533, + "step": 4801 + }, + { + "epoch": 0.85505698005698, + "grad_norm": 0.6280176639556885, + "learning_rate": 0.00017842194829323187, + "loss": 1.0084, + "step": 4802 + }, + { + "epoch": 0.8552350427350427, + "grad_norm": 0.5098552703857422, + "learning_rate": 0.0001784132623241551, + "loss": 1.0804, + "step": 4803 + }, + { + "epoch": 0.8554131054131054, + "grad_norm": 0.5406526923179626, + "learning_rate": 0.00017840457481873328, + "loss": 1.2571, + "step": 4804 + }, + { + "epoch": 0.8555911680911681, + "grad_norm": 0.5859003663063049, + "learning_rate": 0.00017839588577713678, + "loss": 1.2462, + "step": 4805 + }, + { + "epoch": 0.8557692307692307, + "grad_norm": 0.6209002137184143, + "learning_rate": 0.00017838719519953572, + "loss": 1.307, + "step": 4806 + }, + { + "epoch": 0.8559472934472935, + "grad_norm": 0.525753915309906, + "learning_rate": 0.00017837850308610037, + "loss": 1.2957, + "step": 4807 + }, + { + "epoch": 0.8561253561253561, + "grad_norm": 0.5096195340156555, + "learning_rate": 0.0001783698094370011, + "loss": 1.1433, + "step": 4808 + }, + { + "epoch": 0.8563034188034188, + "grad_norm": 0.5873076915740967, + "learning_rate": 0.0001783611142524082, + "loss": 1.2271, + "step": 4809 + }, + { + "epoch": 0.8564814814814815, + "grad_norm": 0.5093944668769836, + "learning_rate": 0.0001783524175324921, + "loss": 0.8788, + "step": 4810 + }, + { + "epoch": 0.8566595441595442, + "grad_norm": 0.5485084652900696, + "learning_rate": 0.00017834371927742307, + "loss": 1.256, + "step": 4811 + }, + { + "epoch": 0.8568376068376068, + "grad_norm": 0.5808873772621155, + "learning_rate": 0.00017833501948737163, + "loss": 0.9287, + "step": 4812 + }, + { + "epoch": 0.8570156695156695, + "grad_norm": 0.5113978385925293, + "learning_rate": 0.00017832631816250822, + "loss": 1.0372, + "step": 4813 + }, + { + "epoch": 0.8571937321937322, + "grad_norm": 0.5877016186714172, + "learning_rate": 0.0001783176153030033, + "loss": 1.3023, + "step": 4814 + }, + { + "epoch": 0.8573717948717948, + "grad_norm": 0.534328043460846, + "learning_rate": 0.00017830891090902742, + "loss": 1.1023, + "step": 4815 + }, + { + "epoch": 0.8575498575498576, + "grad_norm": 0.5781638026237488, + "learning_rate": 0.0001783002049807511, + "loss": 0.9562, + "step": 4816 + }, + { + "epoch": 0.8577279202279202, + "grad_norm": 0.5760263204574585, + "learning_rate": 0.00017829149751834487, + "loss": 0.8733, + "step": 4817 + }, + { + "epoch": 0.8579059829059829, + "grad_norm": 0.3887255787849426, + "learning_rate": 0.00017828278852197944, + "loss": 0.5949, + "step": 4818 + }, + { + "epoch": 0.8580840455840456, + "grad_norm": 0.47814446687698364, + "learning_rate": 0.00017827407799182537, + "loss": 1.0698, + "step": 4819 + }, + { + "epoch": 0.8582621082621082, + "grad_norm": 0.5520272254943848, + "learning_rate": 0.00017826536592805334, + "loss": 1.1314, + "step": 4820 + }, + { + "epoch": 0.8584401709401709, + "grad_norm": 0.5285319685935974, + "learning_rate": 0.00017825665233083405, + "loss": 1.1618, + "step": 4821 + }, + { + "epoch": 0.8586182336182336, + "grad_norm": 0.6080102324485779, + "learning_rate": 0.0001782479372003382, + "loss": 1.3817, + "step": 4822 + }, + { + "epoch": 0.8587962962962963, + "grad_norm": 0.7474410533905029, + "learning_rate": 0.00017823922053673662, + "loss": 1.1321, + "step": 4823 + }, + { + "epoch": 0.8589743589743589, + "grad_norm": 0.559283435344696, + "learning_rate": 0.0001782305023402, + "loss": 1.1894, + "step": 4824 + }, + { + "epoch": 0.8591524216524217, + "grad_norm": 0.5620571374893188, + "learning_rate": 0.00017822178261089918, + "loss": 1.134, + "step": 4825 + }, + { + "epoch": 0.8593304843304843, + "grad_norm": 0.5553044676780701, + "learning_rate": 0.00017821306134900504, + "loss": 1.3222, + "step": 4826 + }, + { + "epoch": 0.8595085470085471, + "grad_norm": 0.6177778244018555, + "learning_rate": 0.00017820433855468846, + "loss": 1.2545, + "step": 4827 + }, + { + "epoch": 0.8596866096866097, + "grad_norm": 0.656233012676239, + "learning_rate": 0.0001781956142281203, + "loss": 1.1346, + "step": 4828 + }, + { + "epoch": 0.8598646723646723, + "grad_norm": 0.6710973381996155, + "learning_rate": 0.0001781868883694715, + "loss": 1.1361, + "step": 4829 + }, + { + "epoch": 0.8600427350427351, + "grad_norm": 0.5093601942062378, + "learning_rate": 0.0001781781609789131, + "loss": 1.0509, + "step": 4830 + }, + { + "epoch": 0.8602207977207977, + "grad_norm": 0.5707578063011169, + "learning_rate": 0.00017816943205661598, + "loss": 1.0964, + "step": 4831 + }, + { + "epoch": 0.8603988603988604, + "grad_norm": 0.6159597635269165, + "learning_rate": 0.00017816070160275125, + "loss": 1.0322, + "step": 4832 + }, + { + "epoch": 0.8605769230769231, + "grad_norm": 0.5430580377578735, + "learning_rate": 0.0001781519696174899, + "loss": 1.2464, + "step": 4833 + }, + { + "epoch": 0.8607549857549858, + "grad_norm": 0.48104700446128845, + "learning_rate": 0.0001781432361010031, + "loss": 1.1031, + "step": 4834 + }, + { + "epoch": 0.8609330484330484, + "grad_norm": 0.5304946303367615, + "learning_rate": 0.0001781345010534619, + "loss": 1.0281, + "step": 4835 + }, + { + "epoch": 0.8611111111111112, + "grad_norm": 0.5230711698532104, + "learning_rate": 0.00017812576447503742, + "loss": 0.9499, + "step": 4836 + }, + { + "epoch": 0.8612891737891738, + "grad_norm": 0.5363606214523315, + "learning_rate": 0.00017811702636590093, + "loss": 1.1358, + "step": 4837 + }, + { + "epoch": 0.8614672364672364, + "grad_norm": 0.5880044102668762, + "learning_rate": 0.00017810828672622358, + "loss": 1.1765, + "step": 4838 + }, + { + "epoch": 0.8616452991452992, + "grad_norm": 0.5194395184516907, + "learning_rate": 0.0001780995455561766, + "loss": 1.1622, + "step": 4839 + }, + { + "epoch": 0.8618233618233618, + "grad_norm": 0.5114264488220215, + "learning_rate": 0.00017809080285593126, + "loss": 1.0081, + "step": 4840 + }, + { + "epoch": 0.8620014245014245, + "grad_norm": 0.6174240112304688, + "learning_rate": 0.00017808205862565886, + "loss": 1.0745, + "step": 4841 + }, + { + "epoch": 0.8621794871794872, + "grad_norm": 0.5662630200386047, + "learning_rate": 0.0001780733128655307, + "loss": 1.3369, + "step": 4842 + }, + { + "epoch": 0.8623575498575499, + "grad_norm": 0.5917882919311523, + "learning_rate": 0.00017806456557571817, + "loss": 1.1631, + "step": 4843 + }, + { + "epoch": 0.8625356125356125, + "grad_norm": 0.5305736660957336, + "learning_rate": 0.00017805581675639265, + "loss": 0.9875, + "step": 4844 + }, + { + "epoch": 0.8627136752136753, + "grad_norm": 0.5181219577789307, + "learning_rate": 0.00017804706640772556, + "loss": 0.9918, + "step": 4845 + }, + { + "epoch": 0.8628917378917379, + "grad_norm": 0.5467997789382935, + "learning_rate": 0.00017803831452988832, + "loss": 1.1395, + "step": 4846 + }, + { + "epoch": 0.8630698005698005, + "grad_norm": 0.5494031310081482, + "learning_rate": 0.00017802956112305241, + "loss": 1.0312, + "step": 4847 + }, + { + "epoch": 0.8632478632478633, + "grad_norm": 0.5804065465927124, + "learning_rate": 0.00017802080618738931, + "loss": 1.1555, + "step": 4848 + }, + { + "epoch": 0.8634259259259259, + "grad_norm": 0.5424801111221313, + "learning_rate": 0.00017801204972307067, + "loss": 1.0215, + "step": 4849 + }, + { + "epoch": 0.8636039886039886, + "grad_norm": 0.5321891903877258, + "learning_rate": 0.0001780032917302679, + "loss": 1.0187, + "step": 4850 + }, + { + "epoch": 0.8637820512820513, + "grad_norm": 0.5543400049209595, + "learning_rate": 0.0001779945322091527, + "loss": 1.1972, + "step": 4851 + }, + { + "epoch": 0.863960113960114, + "grad_norm": 0.566649317741394, + "learning_rate": 0.00017798577115989668, + "loss": 1.0758, + "step": 4852 + }, + { + "epoch": 0.8641381766381766, + "grad_norm": 0.5538444519042969, + "learning_rate": 0.00017797700858267145, + "loss": 1.1338, + "step": 4853 + }, + { + "epoch": 0.8643162393162394, + "grad_norm": 0.5641313791275024, + "learning_rate": 0.0001779682444776487, + "loss": 1.256, + "step": 4854 + }, + { + "epoch": 0.864494301994302, + "grad_norm": 0.6377350091934204, + "learning_rate": 0.00017795947884500016, + "loss": 1.144, + "step": 4855 + }, + { + "epoch": 0.8646723646723646, + "grad_norm": 0.5581876039505005, + "learning_rate": 0.0001779507116848976, + "loss": 1.3163, + "step": 4856 + }, + { + "epoch": 0.8648504273504274, + "grad_norm": 0.5416772365570068, + "learning_rate": 0.0001779419429975128, + "loss": 1.0219, + "step": 4857 + }, + { + "epoch": 0.86502849002849, + "grad_norm": 0.5450608730316162, + "learning_rate": 0.0001779331727830175, + "loss": 1.0093, + "step": 4858 + }, + { + "epoch": 0.8652065527065527, + "grad_norm": 0.5151242017745972, + "learning_rate": 0.00017792440104158358, + "loss": 1.067, + "step": 4859 + }, + { + "epoch": 0.8653846153846154, + "grad_norm": 0.5225046873092651, + "learning_rate": 0.0001779156277733829, + "loss": 1.0432, + "step": 4860 + }, + { + "epoch": 0.8655626780626781, + "grad_norm": 0.5168602466583252, + "learning_rate": 0.00017790685297858737, + "loss": 0.9665, + "step": 4861 + }, + { + "epoch": 0.8657407407407407, + "grad_norm": 0.5749059319496155, + "learning_rate": 0.00017789807665736889, + "loss": 1.1607, + "step": 4862 + }, + { + "epoch": 0.8659188034188035, + "grad_norm": 0.45656394958496094, + "learning_rate": 0.00017788929880989938, + "loss": 0.8362, + "step": 4863 + }, + { + "epoch": 0.8660968660968661, + "grad_norm": 0.5090615749359131, + "learning_rate": 0.00017788051943635086, + "loss": 0.9553, + "step": 4864 + }, + { + "epoch": 0.8662749287749287, + "grad_norm": 0.5381240248680115, + "learning_rate": 0.0001778717385368954, + "loss": 1.1391, + "step": 4865 + }, + { + "epoch": 0.8664529914529915, + "grad_norm": 0.522720456123352, + "learning_rate": 0.00017786295611170493, + "loss": 1.1869, + "step": 4866 + }, + { + "epoch": 0.8666310541310541, + "grad_norm": 0.530986487865448, + "learning_rate": 0.0001778541721609516, + "loss": 1.1046, + "step": 4867 + }, + { + "epoch": 0.8668091168091168, + "grad_norm": 0.5065864324569702, + "learning_rate": 0.0001778453866848075, + "loss": 1.008, + "step": 4868 + }, + { + "epoch": 0.8669871794871795, + "grad_norm": 0.5541394352912903, + "learning_rate": 0.00017783659968344476, + "loss": 1.0004, + "step": 4869 + }, + { + "epoch": 0.8671652421652422, + "grad_norm": 0.5059576630592346, + "learning_rate": 0.00017782781115703556, + "loss": 1.128, + "step": 4870 + }, + { + "epoch": 0.8673433048433048, + "grad_norm": 0.5052187442779541, + "learning_rate": 0.00017781902110575203, + "loss": 0.8544, + "step": 4871 + }, + { + "epoch": 0.8675213675213675, + "grad_norm": 0.5383397340774536, + "learning_rate": 0.00017781022952976646, + "loss": 1.1411, + "step": 4872 + }, + { + "epoch": 0.8676994301994302, + "grad_norm": 0.4760429859161377, + "learning_rate": 0.00017780143642925106, + "loss": 0.8246, + "step": 4873 + }, + { + "epoch": 0.8678774928774928, + "grad_norm": 0.5480535626411438, + "learning_rate": 0.00017779264180437817, + "loss": 1.013, + "step": 4874 + }, + { + "epoch": 0.8680555555555556, + "grad_norm": 0.5303317904472351, + "learning_rate": 0.00017778384565532004, + "loss": 1.0201, + "step": 4875 + }, + { + "epoch": 0.8682336182336182, + "grad_norm": 0.5365355014801025, + "learning_rate": 0.00017777504798224903, + "loss": 1.1107, + "step": 4876 + }, + { + "epoch": 0.8684116809116809, + "grad_norm": 0.5173360705375671, + "learning_rate": 0.00017776624878533754, + "loss": 1.0808, + "step": 4877 + }, + { + "epoch": 0.8685897435897436, + "grad_norm": 0.5088842511177063, + "learning_rate": 0.00017775744806475792, + "loss": 0.995, + "step": 4878 + }, + { + "epoch": 0.8687678062678063, + "grad_norm": 0.5796698927879333, + "learning_rate": 0.00017774864582068264, + "loss": 1.1485, + "step": 4879 + }, + { + "epoch": 0.8689458689458689, + "grad_norm": 0.5719375610351562, + "learning_rate": 0.00017773984205328417, + "loss": 1.0133, + "step": 4880 + }, + { + "epoch": 0.8691239316239316, + "grad_norm": 0.6396418213844299, + "learning_rate": 0.00017773103676273498, + "loss": 1.0932, + "step": 4881 + }, + { + "epoch": 0.8693019943019943, + "grad_norm": 0.5602468252182007, + "learning_rate": 0.00017772222994920763, + "loss": 0.9702, + "step": 4882 + }, + { + "epoch": 0.8694800569800569, + "grad_norm": 0.5167748332023621, + "learning_rate": 0.00017771342161287457, + "loss": 1.0528, + "step": 4883 + }, + { + "epoch": 0.8696581196581197, + "grad_norm": 0.5572916865348816, + "learning_rate": 0.00017770461175390848, + "loss": 1.1341, + "step": 4884 + }, + { + "epoch": 0.8698361823361823, + "grad_norm": 0.6666276454925537, + "learning_rate": 0.00017769580037248195, + "loss": 1.1948, + "step": 4885 + }, + { + "epoch": 0.8700142450142451, + "grad_norm": 0.5348601937294006, + "learning_rate": 0.0001776869874687676, + "loss": 1.0562, + "step": 4886 + }, + { + "epoch": 0.8701923076923077, + "grad_norm": 0.5449648499488831, + "learning_rate": 0.00017767817304293812, + "loss": 0.988, + "step": 4887 + }, + { + "epoch": 0.8703703703703703, + "grad_norm": 0.5995045304298401, + "learning_rate": 0.0001776693570951662, + "loss": 1.2526, + "step": 4888 + }, + { + "epoch": 0.8705484330484331, + "grad_norm": 0.6575320959091187, + "learning_rate": 0.00017766053962562457, + "loss": 1.1717, + "step": 4889 + }, + { + "epoch": 0.8707264957264957, + "grad_norm": 0.5882139801979065, + "learning_rate": 0.00017765172063448597, + "loss": 1.238, + "step": 4890 + }, + { + "epoch": 0.8709045584045584, + "grad_norm": 0.5908389091491699, + "learning_rate": 0.00017764290012192325, + "loss": 1.0606, + "step": 4891 + }, + { + "epoch": 0.8710826210826211, + "grad_norm": 0.6169339418411255, + "learning_rate": 0.00017763407808810917, + "loss": 1.1456, + "step": 4892 + }, + { + "epoch": 0.8712606837606838, + "grad_norm": 0.5916035771369934, + "learning_rate": 0.0001776252545332166, + "loss": 1.0026, + "step": 4893 + }, + { + "epoch": 0.8714387464387464, + "grad_norm": 0.539995551109314, + "learning_rate": 0.00017761642945741843, + "loss": 1.2397, + "step": 4894 + }, + { + "epoch": 0.8716168091168092, + "grad_norm": 0.5346137881278992, + "learning_rate": 0.00017760760286088755, + "loss": 1.1232, + "step": 4895 + }, + { + "epoch": 0.8717948717948718, + "grad_norm": 0.570202112197876, + "learning_rate": 0.00017759877474379692, + "loss": 1.0708, + "step": 4896 + }, + { + "epoch": 0.8719729344729344, + "grad_norm": 0.5023398399353027, + "learning_rate": 0.00017758994510631948, + "loss": 1.1056, + "step": 4897 + }, + { + "epoch": 0.8721509971509972, + "grad_norm": 0.5447137951850891, + "learning_rate": 0.00017758111394862826, + "loss": 0.8776, + "step": 4898 + }, + { + "epoch": 0.8723290598290598, + "grad_norm": 0.5193906426429749, + "learning_rate": 0.00017757228127089625, + "loss": 0.9959, + "step": 4899 + }, + { + "epoch": 0.8725071225071225, + "grad_norm": 0.5958787798881531, + "learning_rate": 0.00017756344707329656, + "loss": 1.092, + "step": 4900 + }, + { + "epoch": 0.8726851851851852, + "grad_norm": 0.521045982837677, + "learning_rate": 0.00017755461135600221, + "loss": 0.9864, + "step": 4901 + }, + { + "epoch": 0.8728632478632479, + "grad_norm": 0.5257635116577148, + "learning_rate": 0.00017754577411918638, + "loss": 1.216, + "step": 4902 + }, + { + "epoch": 0.8730413105413105, + "grad_norm": 0.5425964593887329, + "learning_rate": 0.0001775369353630222, + "loss": 1.1432, + "step": 4903 + }, + { + "epoch": 0.8732193732193733, + "grad_norm": 0.47995322942733765, + "learning_rate": 0.00017752809508768286, + "loss": 1.0227, + "step": 4904 + }, + { + "epoch": 0.8733974358974359, + "grad_norm": 0.5747429728507996, + "learning_rate": 0.0001775192532933415, + "loss": 0.9984, + "step": 4905 + }, + { + "epoch": 0.8735754985754985, + "grad_norm": 0.5745723247528076, + "learning_rate": 0.00017751040998017142, + "loss": 1.2559, + "step": 4906 + }, + { + "epoch": 0.8737535612535613, + "grad_norm": 0.6114141941070557, + "learning_rate": 0.0001775015651483459, + "loss": 1.3224, + "step": 4907 + }, + { + "epoch": 0.8739316239316239, + "grad_norm": 0.4757187068462372, + "learning_rate": 0.00017749271879803817, + "loss": 1.0352, + "step": 4908 + }, + { + "epoch": 0.8741096866096866, + "grad_norm": 0.48644450306892395, + "learning_rate": 0.0001774838709294216, + "loss": 1.0876, + "step": 4909 + }, + { + "epoch": 0.8742877492877493, + "grad_norm": 0.5652037262916565, + "learning_rate": 0.00017747502154266955, + "loss": 0.9189, + "step": 4910 + }, + { + "epoch": 0.874465811965812, + "grad_norm": 0.5289644002914429, + "learning_rate": 0.00017746617063795538, + "loss": 0.9431, + "step": 4911 + }, + { + "epoch": 0.8746438746438746, + "grad_norm": 0.594656229019165, + "learning_rate": 0.00017745731821545253, + "loss": 1.2408, + "step": 4912 + }, + { + "epoch": 0.8748219373219374, + "grad_norm": 0.5693240165710449, + "learning_rate": 0.0001774484642753344, + "loss": 1.347, + "step": 4913 + }, + { + "epoch": 0.875, + "grad_norm": 0.5291008949279785, + "learning_rate": 0.00017743960881777456, + "loss": 1.161, + "step": 4914 + }, + { + "epoch": 0.8751780626780626, + "grad_norm": 0.5958300232887268, + "learning_rate": 0.00017743075184294642, + "loss": 1.2058, + "step": 4915 + }, + { + "epoch": 0.8753561253561254, + "grad_norm": 0.513884425163269, + "learning_rate": 0.00017742189335102354, + "loss": 1.0952, + "step": 4916 + }, + { + "epoch": 0.875534188034188, + "grad_norm": 0.5860681533813477, + "learning_rate": 0.00017741303334217948, + "loss": 1.1801, + "step": 4917 + }, + { + "epoch": 0.8757122507122507, + "grad_norm": 0.47962820529937744, + "learning_rate": 0.00017740417181658788, + "loss": 1.0785, + "step": 4918 + }, + { + "epoch": 0.8758903133903134, + "grad_norm": 0.5110440254211426, + "learning_rate": 0.00017739530877442227, + "loss": 1.1385, + "step": 4919 + }, + { + "epoch": 0.8760683760683761, + "grad_norm": 0.5106285214424133, + "learning_rate": 0.00017738644421585643, + "loss": 1.1204, + "step": 4920 + }, + { + "epoch": 0.8762464387464387, + "grad_norm": 0.5709205865859985, + "learning_rate": 0.00017737757814106393, + "loss": 1.0108, + "step": 4921 + }, + { + "epoch": 0.8764245014245015, + "grad_norm": 0.5850250124931335, + "learning_rate": 0.0001773687105502185, + "loss": 1.0059, + "step": 4922 + }, + { + "epoch": 0.8766025641025641, + "grad_norm": 0.5194727778434753, + "learning_rate": 0.00017735984144349396, + "loss": 0.9466, + "step": 4923 + }, + { + "epoch": 0.8767806267806267, + "grad_norm": 0.5246787667274475, + "learning_rate": 0.000177350970821064, + "loss": 1.1336, + "step": 4924 + }, + { + "epoch": 0.8769586894586895, + "grad_norm": 0.5798323154449463, + "learning_rate": 0.00017734209868310244, + "loss": 1.1641, + "step": 4925 + }, + { + "epoch": 0.8771367521367521, + "grad_norm": 0.5188565850257874, + "learning_rate": 0.00017733322502978314, + "loss": 0.9959, + "step": 4926 + }, + { + "epoch": 0.8773148148148148, + "grad_norm": 0.5969653725624084, + "learning_rate": 0.00017732434986127995, + "loss": 1.2162, + "step": 4927 + }, + { + "epoch": 0.8774928774928775, + "grad_norm": 0.5520089268684387, + "learning_rate": 0.00017731547317776674, + "loss": 1.0163, + "step": 4928 + }, + { + "epoch": 0.8776709401709402, + "grad_norm": 0.48789507150650024, + "learning_rate": 0.00017730659497941745, + "loss": 0.9757, + "step": 4929 + }, + { + "epoch": 0.8778490028490028, + "grad_norm": 0.6034960746765137, + "learning_rate": 0.000177297715266406, + "loss": 1.1278, + "step": 4930 + }, + { + "epoch": 0.8780270655270656, + "grad_norm": 0.53016597032547, + "learning_rate": 0.00017728883403890638, + "loss": 1.0637, + "step": 4931 + }, + { + "epoch": 0.8782051282051282, + "grad_norm": 0.5073726177215576, + "learning_rate": 0.00017727995129709266, + "loss": 1.1491, + "step": 4932 + }, + { + "epoch": 0.8783831908831908, + "grad_norm": 0.540605366230011, + "learning_rate": 0.00017727106704113878, + "loss": 1.0133, + "step": 4933 + }, + { + "epoch": 0.8785612535612536, + "grad_norm": 0.5346775054931641, + "learning_rate": 0.0001772621812712189, + "loss": 1.1781, + "step": 4934 + }, + { + "epoch": 0.8787393162393162, + "grad_norm": 0.5659036040306091, + "learning_rate": 0.00017725329398750702, + "loss": 1.1023, + "step": 4935 + }, + { + "epoch": 0.8789173789173789, + "grad_norm": 0.591063380241394, + "learning_rate": 0.00017724440519017738, + "loss": 1.0298, + "step": 4936 + }, + { + "epoch": 0.8790954415954416, + "grad_norm": 0.5173781514167786, + "learning_rate": 0.0001772355148794041, + "loss": 1.0483, + "step": 4937 + }, + { + "epoch": 0.8792735042735043, + "grad_norm": 0.5405352711677551, + "learning_rate": 0.0001772266230553613, + "loss": 1.0716, + "step": 4938 + }, + { + "epoch": 0.8794515669515669, + "grad_norm": 0.518442690372467, + "learning_rate": 0.00017721772971822323, + "loss": 1.1373, + "step": 4939 + }, + { + "epoch": 0.8796296296296297, + "grad_norm": 0.533673107624054, + "learning_rate": 0.0001772088348681642, + "loss": 1.0489, + "step": 4940 + }, + { + "epoch": 0.8798076923076923, + "grad_norm": 0.46117857098579407, + "learning_rate": 0.0001771999385053584, + "loss": 1.0297, + "step": 4941 + }, + { + "epoch": 0.8799857549857549, + "grad_norm": 0.4687997102737427, + "learning_rate": 0.0001771910406299802, + "loss": 1.071, + "step": 4942 + }, + { + "epoch": 0.8801638176638177, + "grad_norm": 0.5064153075218201, + "learning_rate": 0.0001771821412422039, + "loss": 0.9518, + "step": 4943 + }, + { + "epoch": 0.8803418803418803, + "grad_norm": 0.6561978459358215, + "learning_rate": 0.00017717324034220385, + "loss": 1.11, + "step": 4944 + }, + { + "epoch": 0.8805199430199431, + "grad_norm": 0.5551498532295227, + "learning_rate": 0.00017716433793015454, + "loss": 0.9719, + "step": 4945 + }, + { + "epoch": 0.8806980056980057, + "grad_norm": 0.47059500217437744, + "learning_rate": 0.00017715543400623025, + "loss": 0.8891, + "step": 4946 + }, + { + "epoch": 0.8808760683760684, + "grad_norm": 0.5035740733146667, + "learning_rate": 0.00017714652857060554, + "loss": 0.9671, + "step": 4947 + }, + { + "epoch": 0.8810541310541311, + "grad_norm": 0.4599960446357727, + "learning_rate": 0.00017713762162345487, + "loss": 0.9588, + "step": 4948 + }, + { + "epoch": 0.8812321937321937, + "grad_norm": 0.5087231397628784, + "learning_rate": 0.0001771287131649527, + "loss": 1.1433, + "step": 4949 + }, + { + "epoch": 0.8814102564102564, + "grad_norm": 0.5609854459762573, + "learning_rate": 0.00017711980319527366, + "loss": 1.2022, + "step": 4950 + }, + { + "epoch": 0.8815883190883191, + "grad_norm": 0.49460700154304504, + "learning_rate": 0.00017711089171459227, + "loss": 1.019, + "step": 4951 + }, + { + "epoch": 0.8817663817663818, + "grad_norm": 0.5047259330749512, + "learning_rate": 0.00017710197872308314, + "loss": 0.8301, + "step": 4952 + }, + { + "epoch": 0.8819444444444444, + "grad_norm": 0.5784406065940857, + "learning_rate": 0.0001770930642209209, + "loss": 0.9336, + "step": 4953 + }, + { + "epoch": 0.8821225071225072, + "grad_norm": 0.5037121772766113, + "learning_rate": 0.00017708414820828022, + "loss": 1.0199, + "step": 4954 + }, + { + "epoch": 0.8823005698005698, + "grad_norm": 0.5683804750442505, + "learning_rate": 0.00017707523068533575, + "loss": 0.9758, + "step": 4955 + }, + { + "epoch": 0.8824786324786325, + "grad_norm": 0.5167922973632812, + "learning_rate": 0.0001770663116522623, + "loss": 1.0389, + "step": 4956 + }, + { + "epoch": 0.8826566951566952, + "grad_norm": 0.5813606381416321, + "learning_rate": 0.0001770573911092345, + "loss": 1.3998, + "step": 4957 + }, + { + "epoch": 0.8828347578347578, + "grad_norm": 0.5280475616455078, + "learning_rate": 0.00017704846905642723, + "loss": 1.0545, + "step": 4958 + }, + { + "epoch": 0.8830128205128205, + "grad_norm": 0.5421732068061829, + "learning_rate": 0.00017703954549401528, + "loss": 0.899, + "step": 4959 + }, + { + "epoch": 0.8831908831908832, + "grad_norm": 0.5177720189094543, + "learning_rate": 0.00017703062042217344, + "loss": 0.975, + "step": 4960 + }, + { + "epoch": 0.8833689458689459, + "grad_norm": 0.639327883720398, + "learning_rate": 0.00017702169384107666, + "loss": 1.1936, + "step": 4961 + }, + { + "epoch": 0.8835470085470085, + "grad_norm": 0.5201572179794312, + "learning_rate": 0.00017701276575089975, + "loss": 0.9891, + "step": 4962 + }, + { + "epoch": 0.8837250712250713, + "grad_norm": 0.5304145216941833, + "learning_rate": 0.00017700383615181767, + "loss": 1.0569, + "step": 4963 + }, + { + "epoch": 0.8839031339031339, + "grad_norm": 0.6068132519721985, + "learning_rate": 0.00017699490504400538, + "loss": 1.2653, + "step": 4964 + }, + { + "epoch": 0.8840811965811965, + "grad_norm": 0.597895085811615, + "learning_rate": 0.00017698597242763787, + "loss": 1.2577, + "step": 4965 + }, + { + "epoch": 0.8842592592592593, + "grad_norm": 0.5356902480125427, + "learning_rate": 0.00017697703830289017, + "loss": 1.1056, + "step": 4966 + }, + { + "epoch": 0.8844373219373219, + "grad_norm": 0.5429540872573853, + "learning_rate": 0.0001769681026699373, + "loss": 1.0951, + "step": 4967 + }, + { + "epoch": 0.8846153846153846, + "grad_norm": 0.5789309144020081, + "learning_rate": 0.00017695916552895436, + "loss": 1.0786, + "step": 4968 + }, + { + "epoch": 0.8847934472934473, + "grad_norm": 0.5621341466903687, + "learning_rate": 0.0001769502268801164, + "loss": 1.0645, + "step": 4969 + }, + { + "epoch": 0.88497150997151, + "grad_norm": 0.5879453420639038, + "learning_rate": 0.00017694128672359865, + "loss": 1.2171, + "step": 4970 + }, + { + "epoch": 0.8851495726495726, + "grad_norm": 0.5005951523780823, + "learning_rate": 0.0001769323450595762, + "loss": 1.0725, + "step": 4971 + }, + { + "epoch": 0.8853276353276354, + "grad_norm": 0.5439660549163818, + "learning_rate": 0.00017692340188822425, + "loss": 1.162, + "step": 4972 + }, + { + "epoch": 0.885505698005698, + "grad_norm": 0.6309837698936462, + "learning_rate": 0.00017691445720971802, + "loss": 1.2861, + "step": 4973 + }, + { + "epoch": 0.8856837606837606, + "grad_norm": 0.4997463822364807, + "learning_rate": 0.00017690551102423282, + "loss": 1.1887, + "step": 4974 + }, + { + "epoch": 0.8858618233618234, + "grad_norm": 0.5430852174758911, + "learning_rate": 0.00017689656333194385, + "loss": 1.1231, + "step": 4975 + }, + { + "epoch": 0.886039886039886, + "grad_norm": 0.5414215922355652, + "learning_rate": 0.00017688761413302644, + "loss": 1.2345, + "step": 4976 + }, + { + "epoch": 0.8862179487179487, + "grad_norm": 0.5594443082809448, + "learning_rate": 0.00017687866342765601, + "loss": 1.0775, + "step": 4977 + }, + { + "epoch": 0.8863960113960114, + "grad_norm": 0.5827134847640991, + "learning_rate": 0.00017686971121600787, + "loss": 1.0609, + "step": 4978 + }, + { + "epoch": 0.8865740740740741, + "grad_norm": 0.5075414776802063, + "learning_rate": 0.00017686075749825738, + "loss": 0.796, + "step": 4979 + }, + { + "epoch": 0.8867521367521367, + "grad_norm": 0.6007544994354248, + "learning_rate": 0.00017685180227458003, + "loss": 1.1716, + "step": 4980 + }, + { + "epoch": 0.8869301994301995, + "grad_norm": 0.6458030343055725, + "learning_rate": 0.00017684284554515128, + "loss": 1.1945, + "step": 4981 + }, + { + "epoch": 0.8871082621082621, + "grad_norm": 0.5519212484359741, + "learning_rate": 0.00017683388731014657, + "loss": 1.2571, + "step": 4982 + }, + { + "epoch": 0.8872863247863247, + "grad_norm": 0.5079960227012634, + "learning_rate": 0.00017682492756974146, + "loss": 1.1186, + "step": 4983 + }, + { + "epoch": 0.8874643874643875, + "grad_norm": 0.63576740026474, + "learning_rate": 0.00017681596632411147, + "loss": 1.389, + "step": 4984 + }, + { + "epoch": 0.8876424501424501, + "grad_norm": 0.43325698375701904, + "learning_rate": 0.0001768070035734322, + "loss": 0.7757, + "step": 4985 + }, + { + "epoch": 0.8878205128205128, + "grad_norm": 0.49492064118385315, + "learning_rate": 0.00017679803931787923, + "loss": 1.0096, + "step": 4986 + }, + { + "epoch": 0.8879985754985755, + "grad_norm": 0.5561224222183228, + "learning_rate": 0.00017678907355762825, + "loss": 0.952, + "step": 4987 + }, + { + "epoch": 0.8881766381766382, + "grad_norm": 0.5392457246780396, + "learning_rate": 0.00017678010629285486, + "loss": 1.0442, + "step": 4988 + }, + { + "epoch": 0.8883547008547008, + "grad_norm": 0.4659234881401062, + "learning_rate": 0.00017677113752373482, + "loss": 0.8668, + "step": 4989 + }, + { + "epoch": 0.8885327635327636, + "grad_norm": 0.5139175057411194, + "learning_rate": 0.0001767621672504438, + "loss": 0.8386, + "step": 4990 + }, + { + "epoch": 0.8887108262108262, + "grad_norm": 0.5395823121070862, + "learning_rate": 0.00017675319547315755, + "loss": 0.9754, + "step": 4991 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 0.4751867949962616, + "learning_rate": 0.0001767442221920519, + "loss": 0.8775, + "step": 4992 + }, + { + "epoch": 0.8890669515669516, + "grad_norm": 0.5728281736373901, + "learning_rate": 0.00017673524740730265, + "loss": 1.2807, + "step": 4993 + }, + { + "epoch": 0.8892450142450142, + "grad_norm": 0.5545622110366821, + "learning_rate": 0.00017672627111908558, + "loss": 1.0039, + "step": 4994 + }, + { + "epoch": 0.8894230769230769, + "grad_norm": 0.5127374529838562, + "learning_rate": 0.00017671729332757665, + "loss": 1.0505, + "step": 4995 + }, + { + "epoch": 0.8896011396011396, + "grad_norm": 0.5238714218139648, + "learning_rate": 0.00017670831403295175, + "loss": 1.1775, + "step": 4996 + }, + { + "epoch": 0.8897792022792023, + "grad_norm": 0.5610160827636719, + "learning_rate": 0.00017669933323538674, + "loss": 1.0555, + "step": 4997 + }, + { + "epoch": 0.8899572649572649, + "grad_norm": 0.5481634736061096, + "learning_rate": 0.00017669035093505762, + "loss": 1.0802, + "step": 4998 + }, + { + "epoch": 0.8901353276353277, + "grad_norm": 0.4725174307823181, + "learning_rate": 0.0001766813671321404, + "loss": 0.9611, + "step": 4999 + }, + { + "epoch": 0.8903133903133903, + "grad_norm": 0.5184635519981384, + "learning_rate": 0.0001766723818268111, + "loss": 1.1659, + "step": 5000 + }, + { + "epoch": 0.8904914529914529, + "grad_norm": 0.5503578186035156, + "learning_rate": 0.00017666339501924575, + "loss": 1.2165, + "step": 5001 + }, + { + "epoch": 0.8906695156695157, + "grad_norm": 0.5299594402313232, + "learning_rate": 0.0001766544067096204, + "loss": 1.0196, + "step": 5002 + }, + { + "epoch": 0.8908475783475783, + "grad_norm": 0.5673944354057312, + "learning_rate": 0.00017664541689811118, + "loss": 1.2058, + "step": 5003 + }, + { + "epoch": 0.8910256410256411, + "grad_norm": 0.6057320833206177, + "learning_rate": 0.00017663642558489426, + "loss": 1.0136, + "step": 5004 + }, + { + "epoch": 0.8912037037037037, + "grad_norm": 0.4767026901245117, + "learning_rate": 0.00017662743277014578, + "loss": 0.8522, + "step": 5005 + }, + { + "epoch": 0.8913817663817664, + "grad_norm": 0.5346270203590393, + "learning_rate": 0.00017661843845404192, + "loss": 1.1568, + "step": 5006 + }, + { + "epoch": 0.8915598290598291, + "grad_norm": 0.5365738868713379, + "learning_rate": 0.00017660944263675891, + "loss": 1.0488, + "step": 5007 + }, + { + "epoch": 0.8917378917378918, + "grad_norm": 0.5536269545555115, + "learning_rate": 0.00017660044531847305, + "loss": 1.1216, + "step": 5008 + }, + { + "epoch": 0.8919159544159544, + "grad_norm": 0.6325978636741638, + "learning_rate": 0.00017659144649936055, + "loss": 1.2843, + "step": 5009 + }, + { + "epoch": 0.8920940170940171, + "grad_norm": 0.5890641212463379, + "learning_rate": 0.00017658244617959777, + "loss": 1.1976, + "step": 5010 + }, + { + "epoch": 0.8922720797720798, + "grad_norm": 0.604870080947876, + "learning_rate": 0.00017657344435936107, + "loss": 1.2881, + "step": 5011 + }, + { + "epoch": 0.8924501424501424, + "grad_norm": 0.49805206060409546, + "learning_rate": 0.00017656444103882676, + "loss": 0.8998, + "step": 5012 + }, + { + "epoch": 0.8926282051282052, + "grad_norm": 0.506926953792572, + "learning_rate": 0.0001765554362181713, + "loss": 1.0731, + "step": 5013 + }, + { + "epoch": 0.8928062678062678, + "grad_norm": 0.5353260636329651, + "learning_rate": 0.0001765464298975711, + "loss": 1.0676, + "step": 5014 + }, + { + "epoch": 0.8929843304843305, + "grad_norm": 0.5641853213310242, + "learning_rate": 0.0001765374220772026, + "loss": 0.9606, + "step": 5015 + }, + { + "epoch": 0.8931623931623932, + "grad_norm": 0.5049327611923218, + "learning_rate": 0.00017652841275724233, + "loss": 1.009, + "step": 5016 + }, + { + "epoch": 0.8933404558404558, + "grad_norm": 0.6255155205726624, + "learning_rate": 0.0001765194019378668, + "loss": 1.138, + "step": 5017 + }, + { + "epoch": 0.8935185185185185, + "grad_norm": 0.5816851854324341, + "learning_rate": 0.00017651038961925247, + "loss": 1.3398, + "step": 5018 + }, + { + "epoch": 0.8936965811965812, + "grad_norm": 0.5188020467758179, + "learning_rate": 0.00017650137580157605, + "loss": 1.0126, + "step": 5019 + }, + { + "epoch": 0.8938746438746439, + "grad_norm": 0.5231554508209229, + "learning_rate": 0.00017649236048501406, + "loss": 1.0328, + "step": 5020 + }, + { + "epoch": 0.8940527065527065, + "grad_norm": 0.7638634443283081, + "learning_rate": 0.0001764833436697432, + "loss": 1.3016, + "step": 5021 + }, + { + "epoch": 0.8942307692307693, + "grad_norm": 0.5354094505310059, + "learning_rate": 0.00017647432535594008, + "loss": 1.0646, + "step": 5022 + }, + { + "epoch": 0.8944088319088319, + "grad_norm": 0.6938086748123169, + "learning_rate": 0.0001764653055437814, + "loss": 1.2051, + "step": 5023 + }, + { + "epoch": 0.8945868945868946, + "grad_norm": 0.5546849370002747, + "learning_rate": 0.00017645628423344393, + "loss": 1.0671, + "step": 5024 + }, + { + "epoch": 0.8947649572649573, + "grad_norm": 0.49294665455818176, + "learning_rate": 0.0001764472614251044, + "loss": 1.0328, + "step": 5025 + }, + { + "epoch": 0.89494301994302, + "grad_norm": 0.5965796113014221, + "learning_rate": 0.00017643823711893956, + "loss": 1.0741, + "step": 5026 + }, + { + "epoch": 0.8951210826210826, + "grad_norm": 0.4846448302268982, + "learning_rate": 0.00017642921131512626, + "loss": 1.0409, + "step": 5027 + }, + { + "epoch": 0.8952991452991453, + "grad_norm": 0.5767390131950378, + "learning_rate": 0.00017642018401384135, + "loss": 1.018, + "step": 5028 + }, + { + "epoch": 0.895477207977208, + "grad_norm": 0.503027617931366, + "learning_rate": 0.00017641115521526167, + "loss": 1.0002, + "step": 5029 + }, + { + "epoch": 0.8956552706552706, + "grad_norm": 0.6668619513511658, + "learning_rate": 0.00017640212491956412, + "loss": 1.2154, + "step": 5030 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 0.5544148683547974, + "learning_rate": 0.00017639309312692566, + "loss": 1.2701, + "step": 5031 + }, + { + "epoch": 0.896011396011396, + "grad_norm": 0.6026872992515564, + "learning_rate": 0.00017638405983752323, + "loss": 0.9335, + "step": 5032 + }, + { + "epoch": 0.8961894586894587, + "grad_norm": 0.6288694143295288, + "learning_rate": 0.00017637502505153384, + "loss": 0.9075, + "step": 5033 + }, + { + "epoch": 0.8963675213675214, + "grad_norm": 0.4890204966068268, + "learning_rate": 0.00017636598876913446, + "loss": 0.8492, + "step": 5034 + }, + { + "epoch": 0.896545584045584, + "grad_norm": 0.5746598243713379, + "learning_rate": 0.00017635695099050218, + "loss": 1.1557, + "step": 5035 + }, + { + "epoch": 0.8967236467236467, + "grad_norm": 0.5165683031082153, + "learning_rate": 0.00017634791171581405, + "loss": 1.0899, + "step": 5036 + }, + { + "epoch": 0.8969017094017094, + "grad_norm": 0.4621037244796753, + "learning_rate": 0.0001763388709452472, + "loss": 1.0457, + "step": 5037 + }, + { + "epoch": 0.8970797720797721, + "grad_norm": 0.532358705997467, + "learning_rate": 0.00017632982867897876, + "loss": 1.139, + "step": 5038 + }, + { + "epoch": 0.8972578347578347, + "grad_norm": 0.5794399976730347, + "learning_rate": 0.00017632078491718587, + "loss": 1.031, + "step": 5039 + }, + { + "epoch": 0.8974358974358975, + "grad_norm": 0.5031905174255371, + "learning_rate": 0.00017631173966004576, + "loss": 0.9508, + "step": 5040 + }, + { + "epoch": 0.8976139601139601, + "grad_norm": 0.6528840065002441, + "learning_rate": 0.00017630269290773564, + "loss": 0.9974, + "step": 5041 + }, + { + "epoch": 0.8977920227920227, + "grad_norm": 0.6007558703422546, + "learning_rate": 0.00017629364466043273, + "loss": 1.0993, + "step": 5042 + }, + { + "epoch": 0.8979700854700855, + "grad_norm": 0.5104095339775085, + "learning_rate": 0.00017628459491831437, + "loss": 0.9175, + "step": 5043 + }, + { + "epoch": 0.8981481481481481, + "grad_norm": 0.5285516977310181, + "learning_rate": 0.00017627554368155782, + "loss": 0.998, + "step": 5044 + }, + { + "epoch": 0.8983262108262108, + "grad_norm": 0.5629046559333801, + "learning_rate": 0.00017626649095034045, + "loss": 1.2021, + "step": 5045 + }, + { + "epoch": 0.8985042735042735, + "grad_norm": 0.57548987865448, + "learning_rate": 0.00017625743672483962, + "loss": 1.2076, + "step": 5046 + }, + { + "epoch": 0.8986823361823362, + "grad_norm": 0.4883024990558624, + "learning_rate": 0.0001762483810052327, + "loss": 0.9761, + "step": 5047 + }, + { + "epoch": 0.8988603988603988, + "grad_norm": 0.6378034949302673, + "learning_rate": 0.0001762393237916972, + "loss": 1.2266, + "step": 5048 + }, + { + "epoch": 0.8990384615384616, + "grad_norm": 0.5201624035835266, + "learning_rate": 0.0001762302650844105, + "loss": 1.247, + "step": 5049 + }, + { + "epoch": 0.8992165242165242, + "grad_norm": 0.5438048243522644, + "learning_rate": 0.0001762212048835501, + "loss": 0.993, + "step": 5050 + }, + { + "epoch": 0.8993945868945868, + "grad_norm": 0.5928253531455994, + "learning_rate": 0.00017621214318929354, + "loss": 1.0469, + "step": 5051 + }, + { + "epoch": 0.8995726495726496, + "grad_norm": 0.6437996625900269, + "learning_rate": 0.00017620308000181831, + "loss": 1.3136, + "step": 5052 + }, + { + "epoch": 0.8997507122507122, + "grad_norm": 0.5961456298828125, + "learning_rate": 0.00017619401532130208, + "loss": 1.1495, + "step": 5053 + }, + { + "epoch": 0.8999287749287749, + "grad_norm": 0.497388631105423, + "learning_rate": 0.0001761849491479224, + "loss": 0.7783, + "step": 5054 + }, + { + "epoch": 0.9001068376068376, + "grad_norm": 0.5984451174736023, + "learning_rate": 0.00017617588148185687, + "loss": 1.3115, + "step": 5055 + }, + { + "epoch": 0.9002849002849003, + "grad_norm": 0.549163818359375, + "learning_rate": 0.0001761668123232832, + "loss": 1.1649, + "step": 5056 + }, + { + "epoch": 0.9004629629629629, + "grad_norm": 0.5831968188285828, + "learning_rate": 0.00017615774167237903, + "loss": 1.1749, + "step": 5057 + }, + { + "epoch": 0.9006410256410257, + "grad_norm": 0.5111076235771179, + "learning_rate": 0.00017614866952932214, + "loss": 0.8936, + "step": 5058 + }, + { + "epoch": 0.9008190883190883, + "grad_norm": 0.5740947723388672, + "learning_rate": 0.00017613959589429028, + "loss": 1.2606, + "step": 5059 + }, + { + "epoch": 0.9009971509971509, + "grad_norm": 0.5881099700927734, + "learning_rate": 0.0001761305207674612, + "loss": 1.3682, + "step": 5060 + }, + { + "epoch": 0.9011752136752137, + "grad_norm": 0.5007091760635376, + "learning_rate": 0.00017612144414901268, + "loss": 0.7788, + "step": 5061 + }, + { + "epoch": 0.9013532763532763, + "grad_norm": 0.5127760171890259, + "learning_rate": 0.00017611236603912262, + "loss": 1.0519, + "step": 5062 + }, + { + "epoch": 0.9015313390313391, + "grad_norm": 0.6185184121131897, + "learning_rate": 0.00017610328643796882, + "loss": 1.1672, + "step": 5063 + }, + { + "epoch": 0.9017094017094017, + "grad_norm": 0.49707287549972534, + "learning_rate": 0.00017609420534572926, + "loss": 1.1865, + "step": 5064 + }, + { + "epoch": 0.9018874643874644, + "grad_norm": 0.5667552351951599, + "learning_rate": 0.0001760851227625818, + "loss": 1.1388, + "step": 5065 + }, + { + "epoch": 0.9020655270655271, + "grad_norm": 0.50298011302948, + "learning_rate": 0.00017607603868870442, + "loss": 0.9552, + "step": 5066 + }, + { + "epoch": 0.9022435897435898, + "grad_norm": 0.5709219574928284, + "learning_rate": 0.0001760669531242751, + "loss": 1.2636, + "step": 5067 + }, + { + "epoch": 0.9024216524216524, + "grad_norm": 0.4943496286869049, + "learning_rate": 0.0001760578660694718, + "loss": 0.8951, + "step": 5068 + }, + { + "epoch": 0.9025997150997151, + "grad_norm": 0.5475931167602539, + "learning_rate": 0.00017604877752447267, + "loss": 1.1442, + "step": 5069 + }, + { + "epoch": 0.9027777777777778, + "grad_norm": 0.5280239582061768, + "learning_rate": 0.0001760396874894557, + "loss": 0.9537, + "step": 5070 + }, + { + "epoch": 0.9029558404558404, + "grad_norm": 0.5480797290802002, + "learning_rate": 0.000176030595964599, + "loss": 1.1557, + "step": 5071 + }, + { + "epoch": 0.9031339031339032, + "grad_norm": 0.5232734680175781, + "learning_rate": 0.00017602150295008073, + "loss": 1.0219, + "step": 5072 + }, + { + "epoch": 0.9033119658119658, + "grad_norm": 0.5448359251022339, + "learning_rate": 0.000176012408446079, + "loss": 1.1964, + "step": 5073 + }, + { + "epoch": 0.9034900284900285, + "grad_norm": 0.4841914474964142, + "learning_rate": 0.00017600331245277206, + "loss": 1.0667, + "step": 5074 + }, + { + "epoch": 0.9036680911680912, + "grad_norm": 0.5407083630561829, + "learning_rate": 0.0001759942149703381, + "loss": 1.1895, + "step": 5075 + }, + { + "epoch": 0.9038461538461539, + "grad_norm": 0.5140416026115417, + "learning_rate": 0.00017598511599895534, + "loss": 0.9402, + "step": 5076 + }, + { + "epoch": 0.9040242165242165, + "grad_norm": 0.6333765983581543, + "learning_rate": 0.00017597601553880207, + "loss": 1.239, + "step": 5077 + }, + { + "epoch": 0.9042022792022792, + "grad_norm": 0.4996028244495392, + "learning_rate": 0.00017596691359005664, + "loss": 1.0259, + "step": 5078 + }, + { + "epoch": 0.9043803418803419, + "grad_norm": 0.591892421245575, + "learning_rate": 0.00017595781015289732, + "loss": 1.2148, + "step": 5079 + }, + { + "epoch": 0.9045584045584045, + "grad_norm": 0.736499011516571, + "learning_rate": 0.0001759487052275025, + "loss": 1.1373, + "step": 5080 + }, + { + "epoch": 0.9047364672364673, + "grad_norm": 0.5951572060585022, + "learning_rate": 0.00017593959881405057, + "loss": 1.1833, + "step": 5081 + }, + { + "epoch": 0.9049145299145299, + "grad_norm": 0.5092006325721741, + "learning_rate": 0.00017593049091271996, + "loss": 0.8841, + "step": 5082 + }, + { + "epoch": 0.9050925925925926, + "grad_norm": 0.5679013729095459, + "learning_rate": 0.0001759213815236891, + "loss": 1.1056, + "step": 5083 + }, + { + "epoch": 0.9052706552706553, + "grad_norm": 0.5708174109458923, + "learning_rate": 0.0001759122706471365, + "loss": 1.1952, + "step": 5084 + }, + { + "epoch": 0.905448717948718, + "grad_norm": 0.5726733803749084, + "learning_rate": 0.00017590315828324067, + "loss": 1.1013, + "step": 5085 + }, + { + "epoch": 0.9056267806267806, + "grad_norm": 0.5821273326873779, + "learning_rate": 0.00017589404443218008, + "loss": 1.2323, + "step": 5086 + }, + { + "epoch": 0.9058048433048433, + "grad_norm": 0.5811445713043213, + "learning_rate": 0.00017588492909413337, + "loss": 1.2241, + "step": 5087 + }, + { + "epoch": 0.905982905982906, + "grad_norm": 0.5377545952796936, + "learning_rate": 0.0001758758122692791, + "loss": 0.9777, + "step": 5088 + }, + { + "epoch": 0.9061609686609686, + "grad_norm": 0.5985640287399292, + "learning_rate": 0.0001758666939577959, + "loss": 0.9737, + "step": 5089 + }, + { + "epoch": 0.9063390313390314, + "grad_norm": 0.6038222908973694, + "learning_rate": 0.00017585757415986247, + "loss": 1.2116, + "step": 5090 + }, + { + "epoch": 0.906517094017094, + "grad_norm": 0.6752246022224426, + "learning_rate": 0.00017584845287565743, + "loss": 1.1975, + "step": 5091 + }, + { + "epoch": 0.9066951566951567, + "grad_norm": 0.5400625467300415, + "learning_rate": 0.0001758393301053595, + "loss": 0.9669, + "step": 5092 + }, + { + "epoch": 0.9068732193732194, + "grad_norm": 0.5637784004211426, + "learning_rate": 0.00017583020584914746, + "loss": 1.2672, + "step": 5093 + }, + { + "epoch": 0.907051282051282, + "grad_norm": 0.4825877249240875, + "learning_rate": 0.00017582108010720006, + "loss": 0.9719, + "step": 5094 + }, + { + "epoch": 0.9072293447293447, + "grad_norm": 0.49902790784835815, + "learning_rate": 0.00017581195287969613, + "loss": 0.7941, + "step": 5095 + }, + { + "epoch": 0.9074074074074074, + "grad_norm": 0.5991541743278503, + "learning_rate": 0.0001758028241668144, + "loss": 1.049, + "step": 5096 + }, + { + "epoch": 0.9075854700854701, + "grad_norm": 0.5788859724998474, + "learning_rate": 0.00017579369396873384, + "loss": 1.0318, + "step": 5097 + }, + { + "epoch": 0.9077635327635327, + "grad_norm": 0.5914160013198853, + "learning_rate": 0.0001757845622856333, + "loss": 1.1007, + "step": 5098 + }, + { + "epoch": 0.9079415954415955, + "grad_norm": 0.5361711382865906, + "learning_rate": 0.00017577542911769166, + "loss": 1.0694, + "step": 5099 + }, + { + "epoch": 0.9081196581196581, + "grad_norm": 0.5752849578857422, + "learning_rate": 0.00017576629446508792, + "loss": 1.1184, + "step": 5100 + }, + { + "epoch": 0.9082977207977208, + "grad_norm": 0.6042249798774719, + "learning_rate": 0.000175757158328001, + "loss": 1.2808, + "step": 5101 + }, + { + "epoch": 0.9084757834757835, + "grad_norm": 0.508352518081665, + "learning_rate": 0.00017574802070661, + "loss": 1.0038, + "step": 5102 + }, + { + "epoch": 0.9086538461538461, + "grad_norm": 0.5667358040809631, + "learning_rate": 0.00017573888160109385, + "loss": 1.0208, + "step": 5103 + }, + { + "epoch": 0.9088319088319088, + "grad_norm": 0.653619647026062, + "learning_rate": 0.00017572974101163165, + "loss": 1.2053, + "step": 5104 + }, + { + "epoch": 0.9090099715099715, + "grad_norm": 0.5069597363471985, + "learning_rate": 0.00017572059893840246, + "loss": 0.8634, + "step": 5105 + }, + { + "epoch": 0.9091880341880342, + "grad_norm": 0.6160602569580078, + "learning_rate": 0.00017571145538158547, + "loss": 1.2626, + "step": 5106 + }, + { + "epoch": 0.9093660968660968, + "grad_norm": 0.6335833668708801, + "learning_rate": 0.00017570231034135978, + "loss": 1.3381, + "step": 5107 + }, + { + "epoch": 0.9095441595441596, + "grad_norm": 0.5140398740768433, + "learning_rate": 0.00017569316381790454, + "loss": 1.1258, + "step": 5108 + }, + { + "epoch": 0.9097222222222222, + "grad_norm": 0.5682975649833679, + "learning_rate": 0.00017568401581139905, + "loss": 1.3367, + "step": 5109 + }, + { + "epoch": 0.9099002849002849, + "grad_norm": 0.49765729904174805, + "learning_rate": 0.00017567486632202246, + "loss": 1.1891, + "step": 5110 + }, + { + "epoch": 0.9100783475783476, + "grad_norm": 0.5139224529266357, + "learning_rate": 0.00017566571534995406, + "loss": 0.9768, + "step": 5111 + }, + { + "epoch": 0.9102564102564102, + "grad_norm": 0.5510922074317932, + "learning_rate": 0.00017565656289537316, + "loss": 1.1552, + "step": 5112 + }, + { + "epoch": 0.9104344729344729, + "grad_norm": 0.6243364810943604, + "learning_rate": 0.00017564740895845908, + "loss": 1.1341, + "step": 5113 + }, + { + "epoch": 0.9106125356125356, + "grad_norm": 0.5334977507591248, + "learning_rate": 0.00017563825353939116, + "loss": 1.0894, + "step": 5114 + }, + { + "epoch": 0.9107905982905983, + "grad_norm": 0.5195826292037964, + "learning_rate": 0.00017562909663834878, + "loss": 1.1011, + "step": 5115 + }, + { + "epoch": 0.9109686609686609, + "grad_norm": 0.5298168063163757, + "learning_rate": 0.00017561993825551138, + "loss": 1.0079, + "step": 5116 + }, + { + "epoch": 0.9111467236467237, + "grad_norm": 0.5858965516090393, + "learning_rate": 0.00017561077839105835, + "loss": 1.2746, + "step": 5117 + }, + { + "epoch": 0.9113247863247863, + "grad_norm": 0.5572476387023926, + "learning_rate": 0.0001756016170451692, + "loss": 0.8169, + "step": 5118 + }, + { + "epoch": 0.9115028490028491, + "grad_norm": 0.5247095823287964, + "learning_rate": 0.0001755924542180234, + "loss": 1.1206, + "step": 5119 + }, + { + "epoch": 0.9116809116809117, + "grad_norm": 0.5605118274688721, + "learning_rate": 0.0001755832899098005, + "loss": 1.371, + "step": 5120 + }, + { + "epoch": 0.9118589743589743, + "grad_norm": 0.5732316970825195, + "learning_rate": 0.00017557412412068005, + "loss": 1.1248, + "step": 5121 + }, + { + "epoch": 0.9120370370370371, + "grad_norm": 0.6167279481887817, + "learning_rate": 0.0001755649568508416, + "loss": 0.94, + "step": 5122 + }, + { + "epoch": 0.9122150997150997, + "grad_norm": 0.5497499108314514, + "learning_rate": 0.00017555578810046483, + "loss": 1.0112, + "step": 5123 + }, + { + "epoch": 0.9123931623931624, + "grad_norm": 0.540762186050415, + "learning_rate": 0.00017554661786972931, + "loss": 1.1058, + "step": 5124 + }, + { + "epoch": 0.9125712250712251, + "grad_norm": 0.5943556427955627, + "learning_rate": 0.0001755374461588148, + "loss": 0.9086, + "step": 5125 + }, + { + "epoch": 0.9127492877492878, + "grad_norm": 0.5300756692886353, + "learning_rate": 0.0001755282729679009, + "loss": 1.1566, + "step": 5126 + }, + { + "epoch": 0.9129273504273504, + "grad_norm": 0.5390434861183167, + "learning_rate": 0.00017551909829716743, + "loss": 1.1395, + "step": 5127 + }, + { + "epoch": 0.9131054131054132, + "grad_norm": 0.627434492111206, + "learning_rate": 0.00017550992214679405, + "loss": 1.1537, + "step": 5128 + }, + { + "epoch": 0.9132834757834758, + "grad_norm": 0.4806903302669525, + "learning_rate": 0.00017550074451696063, + "loss": 0.7905, + "step": 5129 + }, + { + "epoch": 0.9134615384615384, + "grad_norm": 0.5714817047119141, + "learning_rate": 0.00017549156540784696, + "loss": 1.1042, + "step": 5130 + }, + { + "epoch": 0.9136396011396012, + "grad_norm": 0.5839236378669739, + "learning_rate": 0.0001754823848196329, + "loss": 1.0383, + "step": 5131 + }, + { + "epoch": 0.9138176638176638, + "grad_norm": 0.6089872717857361, + "learning_rate": 0.0001754732027524983, + "loss": 0.9399, + "step": 5132 + }, + { + "epoch": 0.9139957264957265, + "grad_norm": 0.4937956631183624, + "learning_rate": 0.00017546401920662307, + "loss": 0.7382, + "step": 5133 + }, + { + "epoch": 0.9141737891737892, + "grad_norm": 0.5918676257133484, + "learning_rate": 0.00017545483418218716, + "loss": 1.2207, + "step": 5134 + }, + { + "epoch": 0.9143518518518519, + "grad_norm": 0.5825346112251282, + "learning_rate": 0.0001754456476793705, + "loss": 0.9669, + "step": 5135 + }, + { + "epoch": 0.9145299145299145, + "grad_norm": 0.49829617142677307, + "learning_rate": 0.0001754364596983531, + "loss": 1.2247, + "step": 5136 + }, + { + "epoch": 0.9147079772079773, + "grad_norm": 0.5128271579742432, + "learning_rate": 0.00017542727023931497, + "loss": 0.9563, + "step": 5137 + }, + { + "epoch": 0.9148860398860399, + "grad_norm": 0.5789414644241333, + "learning_rate": 0.00017541807930243622, + "loss": 1.22, + "step": 5138 + }, + { + "epoch": 0.9150641025641025, + "grad_norm": 0.44155433773994446, + "learning_rate": 0.00017540888688789683, + "loss": 0.9897, + "step": 5139 + }, + { + "epoch": 0.9152421652421653, + "grad_norm": 0.550464391708374, + "learning_rate": 0.00017539969299587696, + "loss": 1.0624, + "step": 5140 + }, + { + "epoch": 0.9154202279202279, + "grad_norm": 0.5019831657409668, + "learning_rate": 0.0001753904976265567, + "loss": 0.9045, + "step": 5141 + }, + { + "epoch": 0.9155982905982906, + "grad_norm": 0.589658796787262, + "learning_rate": 0.0001753813007801163, + "loss": 1.0454, + "step": 5142 + }, + { + "epoch": 0.9157763532763533, + "grad_norm": 0.5945459008216858, + "learning_rate": 0.00017537210245673586, + "loss": 1.0042, + "step": 5143 + }, + { + "epoch": 0.915954415954416, + "grad_norm": 0.5409809947013855, + "learning_rate": 0.00017536290265659566, + "loss": 1.0609, + "step": 5144 + }, + { + "epoch": 0.9161324786324786, + "grad_norm": 0.5302975177764893, + "learning_rate": 0.00017535370137987597, + "loss": 1.1394, + "step": 5145 + }, + { + "epoch": 0.9163105413105413, + "grad_norm": 0.5253351330757141, + "learning_rate": 0.00017534449862675698, + "loss": 1.2249, + "step": 5146 + }, + { + "epoch": 0.916488603988604, + "grad_norm": 0.6363829970359802, + "learning_rate": 0.00017533529439741908, + "loss": 1.1333, + "step": 5147 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.4703354835510254, + "learning_rate": 0.0001753260886920426, + "loss": 0.9971, + "step": 5148 + }, + { + "epoch": 0.9168447293447294, + "grad_norm": 0.6394907236099243, + "learning_rate": 0.00017531688151080786, + "loss": 1.5942, + "step": 5149 + }, + { + "epoch": 0.917022792022792, + "grad_norm": 0.5573459267616272, + "learning_rate": 0.00017530767285389527, + "loss": 0.9669, + "step": 5150 + }, + { + "epoch": 0.9172008547008547, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00017529846272148532, + "loss": 1.2151, + "step": 5151 + }, + { + "epoch": 0.9173789173789174, + "grad_norm": 0.5550395846366882, + "learning_rate": 0.0001752892511137584, + "loss": 1.1765, + "step": 5152 + }, + { + "epoch": 0.91755698005698, + "grad_norm": 0.5461394786834717, + "learning_rate": 0.00017528003803089496, + "loss": 1.1136, + "step": 5153 + }, + { + "epoch": 0.9177350427350427, + "grad_norm": 0.5512672662734985, + "learning_rate": 0.00017527082347307558, + "loss": 1.1727, + "step": 5154 + }, + { + "epoch": 0.9179131054131054, + "grad_norm": 0.5210778713226318, + "learning_rate": 0.0001752616074404808, + "loss": 1.09, + "step": 5155 + }, + { + "epoch": 0.9180911680911681, + "grad_norm": 0.5214943289756775, + "learning_rate": 0.00017525238993329115, + "loss": 0.9654, + "step": 5156 + }, + { + "epoch": 0.9182692307692307, + "grad_norm": 0.5822862386703491, + "learning_rate": 0.00017524317095168724, + "loss": 1.0951, + "step": 5157 + }, + { + "epoch": 0.9184472934472935, + "grad_norm": 0.43948012590408325, + "learning_rate": 0.0001752339504958497, + "loss": 0.6984, + "step": 5158 + }, + { + "epoch": 0.9186253561253561, + "grad_norm": 0.5024449229240417, + "learning_rate": 0.00017522472856595916, + "loss": 0.983, + "step": 5159 + }, + { + "epoch": 0.9188034188034188, + "grad_norm": 0.5815144181251526, + "learning_rate": 0.00017521550516219636, + "loss": 0.9784, + "step": 5160 + }, + { + "epoch": 0.9189814814814815, + "grad_norm": 0.5519825220108032, + "learning_rate": 0.00017520628028474197, + "loss": 1.064, + "step": 5161 + }, + { + "epoch": 0.9191595441595442, + "grad_norm": 0.5615749955177307, + "learning_rate": 0.00017519705393377675, + "loss": 1.1284, + "step": 5162 + }, + { + "epoch": 0.9193376068376068, + "grad_norm": 0.5929917693138123, + "learning_rate": 0.00017518782610948148, + "loss": 1.1221, + "step": 5163 + }, + { + "epoch": 0.9195156695156695, + "grad_norm": 0.7116361856460571, + "learning_rate": 0.00017517859681203692, + "loss": 1.0188, + "step": 5164 + }, + { + "epoch": 0.9196937321937322, + "grad_norm": 0.5095893740653992, + "learning_rate": 0.00017516936604162396, + "loss": 1.0724, + "step": 5165 + }, + { + "epoch": 0.9198717948717948, + "grad_norm": 0.5701385736465454, + "learning_rate": 0.00017516013379842337, + "loss": 1.0572, + "step": 5166 + }, + { + "epoch": 0.9200498575498576, + "grad_norm": 0.518412709236145, + "learning_rate": 0.00017515090008261613, + "loss": 1.0514, + "step": 5167 + }, + { + "epoch": 0.9202279202279202, + "grad_norm": 0.5324261784553528, + "learning_rate": 0.00017514166489438312, + "loss": 1.1708, + "step": 5168 + }, + { + "epoch": 0.9204059829059829, + "grad_norm": 0.5640990138053894, + "learning_rate": 0.00017513242823390525, + "loss": 1.2846, + "step": 5169 + }, + { + "epoch": 0.9205840455840456, + "grad_norm": 0.510352373123169, + "learning_rate": 0.00017512319010136356, + "loss": 1.0763, + "step": 5170 + }, + { + "epoch": 0.9207621082621082, + "grad_norm": 0.4994175136089325, + "learning_rate": 0.00017511395049693898, + "loss": 0.9665, + "step": 5171 + }, + { + "epoch": 0.9209401709401709, + "grad_norm": 0.43196994066238403, + "learning_rate": 0.00017510470942081258, + "loss": 0.761, + "step": 5172 + }, + { + "epoch": 0.9211182336182336, + "grad_norm": 0.558977484703064, + "learning_rate": 0.00017509546687316543, + "loss": 1.0758, + "step": 5173 + }, + { + "epoch": 0.9212962962962963, + "grad_norm": 0.573302149772644, + "learning_rate": 0.0001750862228541786, + "loss": 0.9635, + "step": 5174 + }, + { + "epoch": 0.9214743589743589, + "grad_norm": 0.5083786845207214, + "learning_rate": 0.00017507697736403321, + "loss": 1.0311, + "step": 5175 + }, + { + "epoch": 0.9216524216524217, + "grad_norm": 0.5478954911231995, + "learning_rate": 0.00017506773040291043, + "loss": 1.074, + "step": 5176 + }, + { + "epoch": 0.9218304843304843, + "grad_norm": 0.522376537322998, + "learning_rate": 0.00017505848197099137, + "loss": 1.1162, + "step": 5177 + }, + { + "epoch": 0.9220085470085471, + "grad_norm": 0.5946292281150818, + "learning_rate": 0.0001750492320684573, + "loss": 0.9494, + "step": 5178 + }, + { + "epoch": 0.9221866096866097, + "grad_norm": 0.5423247814178467, + "learning_rate": 0.00017503998069548943, + "loss": 1.0558, + "step": 5179 + }, + { + "epoch": 0.9223646723646723, + "grad_norm": 0.49960651993751526, + "learning_rate": 0.000175030727852269, + "loss": 1.0748, + "step": 5180 + }, + { + "epoch": 0.9225427350427351, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00017502147353897732, + "loss": 1.2066, + "step": 5181 + }, + { + "epoch": 0.9227207977207977, + "grad_norm": 0.57244473695755, + "learning_rate": 0.00017501221775579576, + "loss": 1.048, + "step": 5182 + }, + { + "epoch": 0.9228988603988604, + "grad_norm": 0.512464165687561, + "learning_rate": 0.00017500296050290557, + "loss": 1.1405, + "step": 5183 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.5380734801292419, + "learning_rate": 0.00017499370178048818, + "loss": 1.0641, + "step": 5184 + }, + { + "epoch": 0.9232549857549858, + "grad_norm": 0.47102874517440796, + "learning_rate": 0.000174984441588725, + "loss": 0.7948, + "step": 5185 + }, + { + "epoch": 0.9234330484330484, + "grad_norm": 0.6702211499214172, + "learning_rate": 0.00017497517992779747, + "loss": 1.3009, + "step": 5186 + }, + { + "epoch": 0.9236111111111112, + "grad_norm": 0.4685834050178528, + "learning_rate": 0.000174965916797887, + "loss": 0.8136, + "step": 5187 + }, + { + "epoch": 0.9237891737891738, + "grad_norm": 0.5414277911186218, + "learning_rate": 0.00017495665219917513, + "loss": 0.9708, + "step": 5188 + }, + { + "epoch": 0.9239672364672364, + "grad_norm": 0.5253050923347473, + "learning_rate": 0.0001749473861318434, + "loss": 1.0691, + "step": 5189 + }, + { + "epoch": 0.9241452991452992, + "grad_norm": 0.6009906530380249, + "learning_rate": 0.00017493811859607328, + "loss": 1.2023, + "step": 5190 + }, + { + "epoch": 0.9243233618233618, + "grad_norm": 0.5519336462020874, + "learning_rate": 0.00017492884959204643, + "loss": 1.189, + "step": 5191 + }, + { + "epoch": 0.9245014245014245, + "grad_norm": 0.5024857521057129, + "learning_rate": 0.0001749195791199444, + "loss": 0.8685, + "step": 5192 + }, + { + "epoch": 0.9246794871794872, + "grad_norm": 0.5735679864883423, + "learning_rate": 0.00017491030717994887, + "loss": 1.1903, + "step": 5193 + }, + { + "epoch": 0.9248575498575499, + "grad_norm": 0.5338658094406128, + "learning_rate": 0.00017490103377224147, + "loss": 1.0442, + "step": 5194 + }, + { + "epoch": 0.9250356125356125, + "grad_norm": 0.46669119596481323, + "learning_rate": 0.0001748917588970039, + "loss": 0.6343, + "step": 5195 + }, + { + "epoch": 0.9252136752136753, + "grad_norm": 0.510910153388977, + "learning_rate": 0.00017488248255441793, + "loss": 0.9334, + "step": 5196 + }, + { + "epoch": 0.9253917378917379, + "grad_norm": 0.5732216238975525, + "learning_rate": 0.00017487320474466524, + "loss": 1.0483, + "step": 5197 + }, + { + "epoch": 0.9255698005698005, + "grad_norm": 0.5864318609237671, + "learning_rate": 0.00017486392546792762, + "loss": 1.0669, + "step": 5198 + }, + { + "epoch": 0.9257478632478633, + "grad_norm": 0.5074281096458435, + "learning_rate": 0.00017485464472438692, + "loss": 1.0636, + "step": 5199 + }, + { + "epoch": 0.9259259259259259, + "grad_norm": 0.5833215117454529, + "learning_rate": 0.00017484536251422496, + "loss": 1.2005, + "step": 5200 + }, + { + "epoch": 0.9261039886039886, + "grad_norm": 0.5624990463256836, + "learning_rate": 0.0001748360788376236, + "loss": 1.1623, + "step": 5201 + }, + { + "epoch": 0.9262820512820513, + "grad_norm": 0.5618230104446411, + "learning_rate": 0.00017482679369476472, + "loss": 1.0495, + "step": 5202 + }, + { + "epoch": 0.926460113960114, + "grad_norm": 0.6254985332489014, + "learning_rate": 0.00017481750708583024, + "loss": 0.9521, + "step": 5203 + }, + { + "epoch": 0.9266381766381766, + "grad_norm": 0.5488203763961792, + "learning_rate": 0.00017480821901100216, + "loss": 1.0689, + "step": 5204 + }, + { + "epoch": 0.9268162393162394, + "grad_norm": 0.6157993674278259, + "learning_rate": 0.00017479892947046245, + "loss": 1.2852, + "step": 5205 + }, + { + "epoch": 0.926994301994302, + "grad_norm": 0.49653390049934387, + "learning_rate": 0.00017478963846439305, + "loss": 0.8616, + "step": 5206 + }, + { + "epoch": 0.9271723646723646, + "grad_norm": 0.5079081058502197, + "learning_rate": 0.00017478034599297603, + "loss": 1.0192, + "step": 5207 + }, + { + "epoch": 0.9273504273504274, + "grad_norm": 0.5392495393753052, + "learning_rate": 0.00017477105205639354, + "loss": 1.115, + "step": 5208 + }, + { + "epoch": 0.92752849002849, + "grad_norm": 0.5336191654205322, + "learning_rate": 0.00017476175665482756, + "loss": 1.1892, + "step": 5209 + }, + { + "epoch": 0.9277065527065527, + "grad_norm": 0.631712019443512, + "learning_rate": 0.00017475245978846026, + "loss": 0.9619, + "step": 5210 + }, + { + "epoch": 0.9278846153846154, + "grad_norm": 0.5123951435089111, + "learning_rate": 0.0001747431614574738, + "loss": 1.1477, + "step": 5211 + }, + { + "epoch": 0.9280626780626781, + "grad_norm": 0.5045743584632874, + "learning_rate": 0.00017473386166205038, + "loss": 0.9749, + "step": 5212 + }, + { + "epoch": 0.9282407407407407, + "grad_norm": 0.5296525359153748, + "learning_rate": 0.00017472456040237217, + "loss": 1.0736, + "step": 5213 + }, + { + "epoch": 0.9284188034188035, + "grad_norm": 0.6304933428764343, + "learning_rate": 0.00017471525767862145, + "loss": 1.2444, + "step": 5214 + }, + { + "epoch": 0.9285968660968661, + "grad_norm": 0.4851958155632019, + "learning_rate": 0.00017470595349098044, + "loss": 0.9049, + "step": 5215 + }, + { + "epoch": 0.9287749287749287, + "grad_norm": 0.5730679631233215, + "learning_rate": 0.00017469664783963148, + "loss": 1.0773, + "step": 5216 + }, + { + "epoch": 0.9289529914529915, + "grad_norm": 0.6020415425300598, + "learning_rate": 0.00017468734072475684, + "loss": 1.3247, + "step": 5217 + }, + { + "epoch": 0.9291310541310541, + "grad_norm": 0.47981077432632446, + "learning_rate": 0.00017467803214653893, + "loss": 1.0009, + "step": 5218 + }, + { + "epoch": 0.9293091168091168, + "grad_norm": 0.5787527561187744, + "learning_rate": 0.0001746687221051601, + "loss": 1.2523, + "step": 5219 + }, + { + "epoch": 0.9294871794871795, + "grad_norm": 0.4495891332626343, + "learning_rate": 0.00017465941060080278, + "loss": 0.7364, + "step": 5220 + }, + { + "epoch": 0.9296652421652422, + "grad_norm": 0.5721768140792847, + "learning_rate": 0.0001746500976336494, + "loss": 1.015, + "step": 5221 + }, + { + "epoch": 0.9298433048433048, + "grad_norm": 0.5500208735466003, + "learning_rate": 0.0001746407832038824, + "loss": 1.053, + "step": 5222 + }, + { + "epoch": 0.9300213675213675, + "grad_norm": 0.5784386992454529, + "learning_rate": 0.00017463146731168437, + "loss": 0.9784, + "step": 5223 + }, + { + "epoch": 0.9301994301994302, + "grad_norm": 0.4960322082042694, + "learning_rate": 0.00017462214995723772, + "loss": 0.8674, + "step": 5224 + }, + { + "epoch": 0.9303774928774928, + "grad_norm": 0.5005537271499634, + "learning_rate": 0.00017461283114072508, + "loss": 1.0486, + "step": 5225 + }, + { + "epoch": 0.9305555555555556, + "grad_norm": 0.5064167380332947, + "learning_rate": 0.000174603510862329, + "loss": 0.9722, + "step": 5226 + }, + { + "epoch": 0.9307336182336182, + "grad_norm": 0.583558976650238, + "learning_rate": 0.0001745941891222321, + "loss": 0.9957, + "step": 5227 + }, + { + "epoch": 0.9309116809116809, + "grad_norm": 0.4982515871524811, + "learning_rate": 0.00017458486592061704, + "loss": 0.958, + "step": 5228 + }, + { + "epoch": 0.9310897435897436, + "grad_norm": 0.526549756526947, + "learning_rate": 0.0001745755412576664, + "loss": 1.1172, + "step": 5229 + }, + { + "epoch": 0.9312678062678063, + "grad_norm": 0.6129719018936157, + "learning_rate": 0.000174566215133563, + "loss": 1.2524, + "step": 5230 + }, + { + "epoch": 0.9314458689458689, + "grad_norm": 0.5385653972625732, + "learning_rate": 0.00017455688754848948, + "loss": 1.1655, + "step": 5231 + }, + { + "epoch": 0.9316239316239316, + "grad_norm": 0.5646410584449768, + "learning_rate": 0.0001745475585026287, + "loss": 0.9026, + "step": 5232 + }, + { + "epoch": 0.9318019943019943, + "grad_norm": 0.549223780632019, + "learning_rate": 0.0001745382279961633, + "loss": 0.804, + "step": 5233 + }, + { + "epoch": 0.9319800569800569, + "grad_norm": 0.48547953367233276, + "learning_rate": 0.0001745288960292762, + "loss": 1.0224, + "step": 5234 + }, + { + "epoch": 0.9321581196581197, + "grad_norm": 0.5260967016220093, + "learning_rate": 0.00017451956260215016, + "loss": 0.9688, + "step": 5235 + }, + { + "epoch": 0.9323361823361823, + "grad_norm": 0.6261999011039734, + "learning_rate": 0.00017451022771496812, + "loss": 1.2539, + "step": 5236 + }, + { + "epoch": 0.9325142450142451, + "grad_norm": 0.5801421999931335, + "learning_rate": 0.00017450089136791298, + "loss": 1.11, + "step": 5237 + }, + { + "epoch": 0.9326923076923077, + "grad_norm": 0.5833573937416077, + "learning_rate": 0.0001744915535611676, + "loss": 0.9328, + "step": 5238 + }, + { + "epoch": 0.9328703703703703, + "grad_norm": 0.5422634482383728, + "learning_rate": 0.00017448221429491496, + "loss": 1.034, + "step": 5239 + }, + { + "epoch": 0.9330484330484331, + "grad_norm": 0.5105658769607544, + "learning_rate": 0.00017447287356933808, + "loss": 0.8924, + "step": 5240 + }, + { + "epoch": 0.9332264957264957, + "grad_norm": 0.5114831924438477, + "learning_rate": 0.00017446353138461995, + "loss": 0.9328, + "step": 5241 + }, + { + "epoch": 0.9334045584045584, + "grad_norm": 0.5105039477348328, + "learning_rate": 0.00017445418774094358, + "loss": 1.0468, + "step": 5242 + }, + { + "epoch": 0.9335826210826211, + "grad_norm": 0.593250036239624, + "learning_rate": 0.00017444484263849208, + "loss": 1.0603, + "step": 5243 + }, + { + "epoch": 0.9337606837606838, + "grad_norm": 0.600788414478302, + "learning_rate": 0.00017443549607744853, + "loss": 1.1506, + "step": 5244 + }, + { + "epoch": 0.9339387464387464, + "grad_norm": 0.5394418239593506, + "learning_rate": 0.00017442614805799605, + "loss": 1.038, + "step": 5245 + }, + { + "epoch": 0.9341168091168092, + "grad_norm": 0.5446375608444214, + "learning_rate": 0.00017441679858031786, + "loss": 1.079, + "step": 5246 + }, + { + "epoch": 0.9342948717948718, + "grad_norm": 0.5859794616699219, + "learning_rate": 0.00017440744764459702, + "loss": 1.1453, + "step": 5247 + }, + { + "epoch": 0.9344729344729344, + "grad_norm": 0.4899081289768219, + "learning_rate": 0.00017439809525101688, + "loss": 1.163, + "step": 5248 + }, + { + "epoch": 0.9346509971509972, + "grad_norm": 0.652846097946167, + "learning_rate": 0.00017438874139976055, + "loss": 1.1819, + "step": 5249 + }, + { + "epoch": 0.9348290598290598, + "grad_norm": 0.5402514934539795, + "learning_rate": 0.00017437938609101138, + "loss": 1.0159, + "step": 5250 + }, + { + "epoch": 0.9350071225071225, + "grad_norm": 0.565864086151123, + "learning_rate": 0.00017437002932495265, + "loss": 1.1121, + "step": 5251 + }, + { + "epoch": 0.9351851851851852, + "grad_norm": 0.611786425113678, + "learning_rate": 0.0001743606711017677, + "loss": 1.2511, + "step": 5252 + }, + { + "epoch": 0.9353632478632479, + "grad_norm": 0.5706882476806641, + "learning_rate": 0.00017435131142163988, + "loss": 1.128, + "step": 5253 + }, + { + "epoch": 0.9355413105413105, + "grad_norm": 0.5369367003440857, + "learning_rate": 0.00017434195028475253, + "loss": 1.0562, + "step": 5254 + }, + { + "epoch": 0.9357193732193733, + "grad_norm": 0.49957552552223206, + "learning_rate": 0.0001743325876912891, + "loss": 1.0568, + "step": 5255 + }, + { + "epoch": 0.9358974358974359, + "grad_norm": 0.5398106575012207, + "learning_rate": 0.00017432322364143305, + "loss": 1.1502, + "step": 5256 + }, + { + "epoch": 0.9360754985754985, + "grad_norm": 0.6522027254104614, + "learning_rate": 0.00017431385813536783, + "loss": 1.0591, + "step": 5257 + }, + { + "epoch": 0.9362535612535613, + "grad_norm": 0.5872012972831726, + "learning_rate": 0.00017430449117327693, + "loss": 1.3737, + "step": 5258 + }, + { + "epoch": 0.9364316239316239, + "grad_norm": 0.5124474167823792, + "learning_rate": 0.00017429512275534382, + "loss": 1.0727, + "step": 5259 + }, + { + "epoch": 0.9366096866096866, + "grad_norm": 0.5103365778923035, + "learning_rate": 0.00017428575288175218, + "loss": 1.0339, + "step": 5260 + }, + { + "epoch": 0.9367877492877493, + "grad_norm": 0.585483729839325, + "learning_rate": 0.0001742763815526855, + "loss": 1.1844, + "step": 5261 + }, + { + "epoch": 0.936965811965812, + "grad_norm": 0.5855562090873718, + "learning_rate": 0.00017426700876832746, + "loss": 1.3234, + "step": 5262 + }, + { + "epoch": 0.9371438746438746, + "grad_norm": 0.5774588584899902, + "learning_rate": 0.00017425763452886162, + "loss": 1.0937, + "step": 5263 + }, + { + "epoch": 0.9373219373219374, + "grad_norm": 0.5718343257904053, + "learning_rate": 0.00017424825883447168, + "loss": 1.0783, + "step": 5264 + }, + { + "epoch": 0.9375, + "grad_norm": 0.5414558053016663, + "learning_rate": 0.00017423888168534136, + "loss": 1.1244, + "step": 5265 + }, + { + "epoch": 0.9376780626780626, + "grad_norm": 0.5818275809288025, + "learning_rate": 0.00017422950308165438, + "loss": 1.247, + "step": 5266 + }, + { + "epoch": 0.9378561253561254, + "grad_norm": 0.586398184299469, + "learning_rate": 0.00017422012302359448, + "loss": 1.0515, + "step": 5267 + }, + { + "epoch": 0.938034188034188, + "grad_norm": 0.5236606001853943, + "learning_rate": 0.00017421074151134544, + "loss": 1.1907, + "step": 5268 + }, + { + "epoch": 0.9382122507122507, + "grad_norm": 0.5108010172843933, + "learning_rate": 0.0001742013585450911, + "loss": 1.1125, + "step": 5269 + }, + { + "epoch": 0.9383903133903134, + "grad_norm": 0.4956454038619995, + "learning_rate": 0.00017419197412501527, + "loss": 1.0305, + "step": 5270 + }, + { + "epoch": 0.9385683760683761, + "grad_norm": 0.5432302951812744, + "learning_rate": 0.0001741825882513018, + "loss": 1.1946, + "step": 5271 + }, + { + "epoch": 0.9387464387464387, + "grad_norm": 0.5119295716285706, + "learning_rate": 0.00017417320092413463, + "loss": 0.875, + "step": 5272 + }, + { + "epoch": 0.9389245014245015, + "grad_norm": 0.49740248918533325, + "learning_rate": 0.0001741638121436977, + "loss": 1.1093, + "step": 5273 + }, + { + "epoch": 0.9391025641025641, + "grad_norm": 0.5069027543067932, + "learning_rate": 0.00017415442191017491, + "loss": 1.2498, + "step": 5274 + }, + { + "epoch": 0.9392806267806267, + "grad_norm": 0.570264995098114, + "learning_rate": 0.00017414503022375027, + "loss": 1.0192, + "step": 5275 + }, + { + "epoch": 0.9394586894586895, + "grad_norm": 0.48129352927207947, + "learning_rate": 0.00017413563708460776, + "loss": 0.8467, + "step": 5276 + }, + { + "epoch": 0.9396367521367521, + "grad_norm": 0.5214534401893616, + "learning_rate": 0.00017412624249293148, + "loss": 0.9723, + "step": 5277 + }, + { + "epoch": 0.9398148148148148, + "grad_norm": 0.5150161385536194, + "learning_rate": 0.00017411684644890544, + "loss": 1.0906, + "step": 5278 + }, + { + "epoch": 0.9399928774928775, + "grad_norm": 0.5695852637290955, + "learning_rate": 0.00017410744895271377, + "loss": 1.2891, + "step": 5279 + }, + { + "epoch": 0.9401709401709402, + "grad_norm": 0.5613594651222229, + "learning_rate": 0.00017409805000454055, + "loss": 1.1373, + "step": 5280 + }, + { + "epoch": 0.9403490028490028, + "grad_norm": 0.5134239196777344, + "learning_rate": 0.00017408864960457004, + "loss": 1.1081, + "step": 5281 + }, + { + "epoch": 0.9405270655270656, + "grad_norm": 0.5256397724151611, + "learning_rate": 0.00017407924775298628, + "loss": 1.058, + "step": 5282 + }, + { + "epoch": 0.9407051282051282, + "grad_norm": 0.5145402550697327, + "learning_rate": 0.00017406984444997357, + "loss": 1.0667, + "step": 5283 + }, + { + "epoch": 0.9408831908831908, + "grad_norm": 0.5435704588890076, + "learning_rate": 0.0001740604396957161, + "loss": 1.2275, + "step": 5284 + }, + { + "epoch": 0.9410612535612536, + "grad_norm": 0.5798762440681458, + "learning_rate": 0.0001740510334903982, + "loss": 1.2061, + "step": 5285 + }, + { + "epoch": 0.9412393162393162, + "grad_norm": 0.5461057424545288, + "learning_rate": 0.00017404162583420414, + "loss": 1.1585, + "step": 5286 + }, + { + "epoch": 0.9414173789173789, + "grad_norm": 0.5090487003326416, + "learning_rate": 0.00017403221672731818, + "loss": 1.2496, + "step": 5287 + }, + { + "epoch": 0.9415954415954416, + "grad_norm": 0.5171035528182983, + "learning_rate": 0.00017402280616992476, + "loss": 1.1947, + "step": 5288 + }, + { + "epoch": 0.9417735042735043, + "grad_norm": 0.5292364358901978, + "learning_rate": 0.00017401339416220818, + "loss": 1.0182, + "step": 5289 + }, + { + "epoch": 0.9419515669515669, + "grad_norm": 0.5011499524116516, + "learning_rate": 0.00017400398070435293, + "loss": 1.3363, + "step": 5290 + }, + { + "epoch": 0.9421296296296297, + "grad_norm": 0.4821554720401764, + "learning_rate": 0.0001739945657965434, + "loss": 0.9077, + "step": 5291 + }, + { + "epoch": 0.9423076923076923, + "grad_norm": 0.5849515199661255, + "learning_rate": 0.00017398514943896403, + "loss": 1.1582, + "step": 5292 + }, + { + "epoch": 0.9424857549857549, + "grad_norm": 0.49826139211654663, + "learning_rate": 0.00017397573163179937, + "loss": 1.1025, + "step": 5293 + }, + { + "epoch": 0.9426638176638177, + "grad_norm": 0.6031842827796936, + "learning_rate": 0.00017396631237523392, + "loss": 1.1932, + "step": 5294 + }, + { + "epoch": 0.9428418803418803, + "grad_norm": 0.6013330221176147, + "learning_rate": 0.00017395689166945224, + "loss": 1.2078, + "step": 5295 + }, + { + "epoch": 0.9430199430199431, + "grad_norm": 0.5147021412849426, + "learning_rate": 0.00017394746951463893, + "loss": 0.9988, + "step": 5296 + }, + { + "epoch": 0.9431980056980057, + "grad_norm": 0.5721762776374817, + "learning_rate": 0.0001739380459109785, + "loss": 1.1442, + "step": 5297 + }, + { + "epoch": 0.9433760683760684, + "grad_norm": 0.49272531270980835, + "learning_rate": 0.0001739286208586557, + "loss": 1.0481, + "step": 5298 + }, + { + "epoch": 0.9435541310541311, + "grad_norm": 0.6545688509941101, + "learning_rate": 0.00017391919435785514, + "loss": 1.1393, + "step": 5299 + }, + { + "epoch": 0.9437321937321937, + "grad_norm": 0.617756724357605, + "learning_rate": 0.00017390976640876152, + "loss": 1.1108, + "step": 5300 + }, + { + "epoch": 0.9439102564102564, + "grad_norm": 0.4870470464229584, + "learning_rate": 0.00017390033701155955, + "loss": 0.9028, + "step": 5301 + }, + { + "epoch": 0.9440883190883191, + "grad_norm": 0.5250138640403748, + "learning_rate": 0.000173890906166434, + "loss": 1.0326, + "step": 5302 + }, + { + "epoch": 0.9442663817663818, + "grad_norm": 0.5879467129707336, + "learning_rate": 0.00017388147387356964, + "loss": 1.1569, + "step": 5303 + }, + { + "epoch": 0.9444444444444444, + "grad_norm": 0.4790486991405487, + "learning_rate": 0.00017387204013315127, + "loss": 0.967, + "step": 5304 + }, + { + "epoch": 0.9446225071225072, + "grad_norm": 0.5884372591972351, + "learning_rate": 0.0001738626049453637, + "loss": 1.1342, + "step": 5305 + }, + { + "epoch": 0.9448005698005698, + "grad_norm": 0.4633975028991699, + "learning_rate": 0.00017385316831039187, + "loss": 0.8942, + "step": 5306 + }, + { + "epoch": 0.9449786324786325, + "grad_norm": 0.5301823019981384, + "learning_rate": 0.0001738437302284206, + "loss": 1.1683, + "step": 5307 + }, + { + "epoch": 0.9451566951566952, + "grad_norm": 0.5476770997047424, + "learning_rate": 0.00017383429069963484, + "loss": 1.1574, + "step": 5308 + }, + { + "epoch": 0.9453347578347578, + "grad_norm": 0.47689101099967957, + "learning_rate": 0.00017382484972421953, + "loss": 1.0792, + "step": 5309 + }, + { + "epoch": 0.9455128205128205, + "grad_norm": 0.526063084602356, + "learning_rate": 0.00017381540730235963, + "loss": 0.9012, + "step": 5310 + }, + { + "epoch": 0.9456908831908832, + "grad_norm": 0.5667058229446411, + "learning_rate": 0.0001738059634342402, + "loss": 1.0908, + "step": 5311 + }, + { + "epoch": 0.9458689458689459, + "grad_norm": 0.5402196645736694, + "learning_rate": 0.00017379651812004623, + "loss": 0.943, + "step": 5312 + }, + { + "epoch": 0.9460470085470085, + "grad_norm": 0.5288932919502258, + "learning_rate": 0.00017378707135996276, + "loss": 1.0055, + "step": 5313 + }, + { + "epoch": 0.9462250712250713, + "grad_norm": 0.5607456564903259, + "learning_rate": 0.00017377762315417492, + "loss": 1.2073, + "step": 5314 + }, + { + "epoch": 0.9464031339031339, + "grad_norm": 0.5737698674201965, + "learning_rate": 0.00017376817350286781, + "loss": 1.0001, + "step": 5315 + }, + { + "epoch": 0.9465811965811965, + "grad_norm": 0.6562079787254333, + "learning_rate": 0.00017375872240622657, + "loss": 1.1503, + "step": 5316 + }, + { + "epoch": 0.9467592592592593, + "grad_norm": 0.5407183170318604, + "learning_rate": 0.0001737492698644364, + "loss": 1.1169, + "step": 5317 + }, + { + "epoch": 0.9469373219373219, + "grad_norm": 0.5504152178764343, + "learning_rate": 0.00017373981587768248, + "loss": 1.0468, + "step": 5318 + }, + { + "epoch": 0.9471153846153846, + "grad_norm": 0.4813530743122101, + "learning_rate": 0.00017373036044615006, + "loss": 0.9707, + "step": 5319 + }, + { + "epoch": 0.9472934472934473, + "grad_norm": 0.5810509920120239, + "learning_rate": 0.00017372090357002437, + "loss": 1.4949, + "step": 5320 + }, + { + "epoch": 0.94747150997151, + "grad_norm": 0.5250222086906433, + "learning_rate": 0.00017371144524949074, + "loss": 1.0818, + "step": 5321 + }, + { + "epoch": 0.9476495726495726, + "grad_norm": 0.4852280914783478, + "learning_rate": 0.00017370198548473444, + "loss": 1.1793, + "step": 5322 + }, + { + "epoch": 0.9478276353276354, + "grad_norm": 0.5392420291900635, + "learning_rate": 0.00017369252427594086, + "loss": 1.153, + "step": 5323 + }, + { + "epoch": 0.948005698005698, + "grad_norm": 0.521294116973877, + "learning_rate": 0.00017368306162329533, + "loss": 0.8572, + "step": 5324 + }, + { + "epoch": 0.9481837606837606, + "grad_norm": 0.5579673647880554, + "learning_rate": 0.0001736735975269833, + "loss": 1.0452, + "step": 5325 + }, + { + "epoch": 0.9483618233618234, + "grad_norm": 0.6027318835258484, + "learning_rate": 0.0001736641319871901, + "loss": 1.3475, + "step": 5326 + }, + { + "epoch": 0.948539886039886, + "grad_norm": 0.5600738525390625, + "learning_rate": 0.00017365466500410132, + "loss": 1.0338, + "step": 5327 + }, + { + "epoch": 0.9487179487179487, + "grad_norm": 0.5691532492637634, + "learning_rate": 0.00017364519657790236, + "loss": 1.129, + "step": 5328 + }, + { + "epoch": 0.9488960113960114, + "grad_norm": 0.5161463022232056, + "learning_rate": 0.0001736357267087788, + "loss": 1.0438, + "step": 5329 + }, + { + "epoch": 0.9490740740740741, + "grad_norm": 0.5049656629562378, + "learning_rate": 0.0001736262553969161, + "loss": 0.9484, + "step": 5330 + }, + { + "epoch": 0.9492521367521367, + "grad_norm": 0.5477150678634644, + "learning_rate": 0.00017361678264249988, + "loss": 0.8995, + "step": 5331 + }, + { + "epoch": 0.9494301994301995, + "grad_norm": 0.5679608583450317, + "learning_rate": 0.0001736073084457157, + "loss": 1.241, + "step": 5332 + }, + { + "epoch": 0.9496082621082621, + "grad_norm": 0.5748196840286255, + "learning_rate": 0.00017359783280674926, + "loss": 1.0046, + "step": 5333 + }, + { + "epoch": 0.9497863247863247, + "grad_norm": 0.5677094459533691, + "learning_rate": 0.00017358835572578617, + "loss": 1.2913, + "step": 5334 + }, + { + "epoch": 0.9499643874643875, + "grad_norm": 0.49663659930229187, + "learning_rate": 0.0001735788772030121, + "loss": 1.0388, + "step": 5335 + }, + { + "epoch": 0.9501424501424501, + "grad_norm": 0.5687218904495239, + "learning_rate": 0.0001735693972386128, + "loss": 1.1631, + "step": 5336 + }, + { + "epoch": 0.9503205128205128, + "grad_norm": 0.520708441734314, + "learning_rate": 0.00017355991583277395, + "loss": 1.0744, + "step": 5337 + }, + { + "epoch": 0.9504985754985755, + "grad_norm": 0.5738952159881592, + "learning_rate": 0.00017355043298568137, + "loss": 1.318, + "step": 5338 + }, + { + "epoch": 0.9506766381766382, + "grad_norm": 0.5378455519676208, + "learning_rate": 0.00017354094869752085, + "loss": 0.9827, + "step": 5339 + }, + { + "epoch": 0.9508547008547008, + "grad_norm": 0.5047366619110107, + "learning_rate": 0.0001735314629684782, + "loss": 1.0966, + "step": 5340 + }, + { + "epoch": 0.9510327635327636, + "grad_norm": 0.5526043772697449, + "learning_rate": 0.0001735219757987393, + "loss": 1.059, + "step": 5341 + }, + { + "epoch": 0.9512108262108262, + "grad_norm": 0.5741400718688965, + "learning_rate": 0.00017351248718849003, + "loss": 1.1232, + "step": 5342 + }, + { + "epoch": 0.9513888888888888, + "grad_norm": 0.5421118140220642, + "learning_rate": 0.00017350299713791626, + "loss": 1.0427, + "step": 5343 + }, + { + "epoch": 0.9515669515669516, + "grad_norm": 0.4857081472873688, + "learning_rate": 0.00017349350564720392, + "loss": 0.8663, + "step": 5344 + }, + { + "epoch": 0.9517450142450142, + "grad_norm": 0.5411618947982788, + "learning_rate": 0.00017348401271653904, + "loss": 1.0317, + "step": 5345 + }, + { + "epoch": 0.9519230769230769, + "grad_norm": 0.5246246457099915, + "learning_rate": 0.00017347451834610756, + "loss": 1.0076, + "step": 5346 + }, + { + "epoch": 0.9521011396011396, + "grad_norm": 0.5278927683830261, + "learning_rate": 0.00017346502253609556, + "loss": 0.931, + "step": 5347 + }, + { + "epoch": 0.9522792022792023, + "grad_norm": 0.5934548377990723, + "learning_rate": 0.00017345552528668902, + "loss": 1.3205, + "step": 5348 + }, + { + "epoch": 0.9524572649572649, + "grad_norm": 0.5466100573539734, + "learning_rate": 0.00017344602659807406, + "loss": 0.8725, + "step": 5349 + }, + { + "epoch": 0.9526353276353277, + "grad_norm": 0.5220118761062622, + "learning_rate": 0.00017343652647043678, + "loss": 1.1642, + "step": 5350 + }, + { + "epoch": 0.9528133903133903, + "grad_norm": 0.6166301965713501, + "learning_rate": 0.0001734270249039633, + "loss": 0.8152, + "step": 5351 + }, + { + "epoch": 0.9529914529914529, + "grad_norm": 0.5173428058624268, + "learning_rate": 0.00017341752189883983, + "loss": 0.9296, + "step": 5352 + }, + { + "epoch": 0.9531695156695157, + "grad_norm": 0.5363461375236511, + "learning_rate": 0.0001734080174552525, + "loss": 1.3546, + "step": 5353 + }, + { + "epoch": 0.9533475783475783, + "grad_norm": 0.5333831906318665, + "learning_rate": 0.0001733985115733876, + "loss": 1.0401, + "step": 5354 + }, + { + "epoch": 0.9535256410256411, + "grad_norm": 0.5179334878921509, + "learning_rate": 0.00017338900425343132, + "loss": 1.1254, + "step": 5355 + }, + { + "epoch": 0.9537037037037037, + "grad_norm": 0.5171303153038025, + "learning_rate": 0.00017337949549556993, + "loss": 1.0518, + "step": 5356 + }, + { + "epoch": 0.9538817663817664, + "grad_norm": 0.5164596438407898, + "learning_rate": 0.00017336998529998978, + "loss": 0.8732, + "step": 5357 + }, + { + "epoch": 0.9540598290598291, + "grad_norm": 0.5555717349052429, + "learning_rate": 0.00017336047366687719, + "loss": 1.2312, + "step": 5358 + }, + { + "epoch": 0.9542378917378918, + "grad_norm": 0.45685622096061707, + "learning_rate": 0.00017335096059641847, + "loss": 0.8882, + "step": 5359 + }, + { + "epoch": 0.9544159544159544, + "grad_norm": 0.5260133743286133, + "learning_rate": 0.0001733414460888001, + "loss": 1.0952, + "step": 5360 + }, + { + "epoch": 0.9545940170940171, + "grad_norm": 0.4597703814506531, + "learning_rate": 0.0001733319301442084, + "loss": 1.0835, + "step": 5361 + }, + { + "epoch": 0.9547720797720798, + "grad_norm": 0.5279495120048523, + "learning_rate": 0.0001733224127628299, + "loss": 1.0295, + "step": 5362 + }, + { + "epoch": 0.9549501424501424, + "grad_norm": 0.48919400572776794, + "learning_rate": 0.00017331289394485104, + "loss": 0.9693, + "step": 5363 + }, + { + "epoch": 0.9551282051282052, + "grad_norm": 0.5639515519142151, + "learning_rate": 0.0001733033736904583, + "loss": 1.0893, + "step": 5364 + }, + { + "epoch": 0.9553062678062678, + "grad_norm": 0.49761319160461426, + "learning_rate": 0.00017329385199983823, + "loss": 1.038, + "step": 5365 + }, + { + "epoch": 0.9554843304843305, + "grad_norm": 0.5503305792808533, + "learning_rate": 0.0001732843288731774, + "loss": 0.9976, + "step": 5366 + }, + { + "epoch": 0.9556623931623932, + "grad_norm": 0.5633028745651245, + "learning_rate": 0.00017327480431066235, + "loss": 1.0602, + "step": 5367 + }, + { + "epoch": 0.9558404558404558, + "grad_norm": 0.48074454069137573, + "learning_rate": 0.00017326527831247973, + "loss": 1.0286, + "step": 5368 + }, + { + "epoch": 0.9560185185185185, + "grad_norm": 0.506597638130188, + "learning_rate": 0.0001732557508788162, + "loss": 0.9061, + "step": 5369 + }, + { + "epoch": 0.9561965811965812, + "grad_norm": 0.6570749282836914, + "learning_rate": 0.0001732462220098584, + "loss": 1.0852, + "step": 5370 + }, + { + "epoch": 0.9563746438746439, + "grad_norm": 0.5607653856277466, + "learning_rate": 0.00017323669170579302, + "loss": 1.0486, + "step": 5371 + }, + { + "epoch": 0.9565527065527065, + "grad_norm": 0.6047050356864929, + "learning_rate": 0.0001732271599668068, + "loss": 1.2175, + "step": 5372 + }, + { + "epoch": 0.9567307692307693, + "grad_norm": 0.5506869554519653, + "learning_rate": 0.00017321762679308651, + "loss": 1.0114, + "step": 5373 + }, + { + "epoch": 0.9569088319088319, + "grad_norm": 0.5868638157844543, + "learning_rate": 0.00017320809218481891, + "loss": 1.2983, + "step": 5374 + }, + { + "epoch": 0.9570868945868946, + "grad_norm": 0.539619505405426, + "learning_rate": 0.00017319855614219084, + "loss": 1.2361, + "step": 5375 + }, + { + "epoch": 0.9572649572649573, + "grad_norm": 0.5525495409965515, + "learning_rate": 0.0001731890186653891, + "loss": 1.1316, + "step": 5376 + }, + { + "epoch": 0.95744301994302, + "grad_norm": 0.5549767017364502, + "learning_rate": 0.0001731794797546006, + "loss": 1.0547, + "step": 5377 + }, + { + "epoch": 0.9576210826210826, + "grad_norm": 0.5356076955795288, + "learning_rate": 0.00017316993941001222, + "loss": 0.9942, + "step": 5378 + }, + { + "epoch": 0.9577991452991453, + "grad_norm": 0.5365784168243408, + "learning_rate": 0.00017316039763181084, + "loss": 1.226, + "step": 5379 + }, + { + "epoch": 0.957977207977208, + "grad_norm": 0.5190927386283875, + "learning_rate": 0.00017315085442018343, + "loss": 1.1704, + "step": 5380 + }, + { + "epoch": 0.9581552706552706, + "grad_norm": 0.526658833026886, + "learning_rate": 0.00017314130977531705, + "loss": 1.109, + "step": 5381 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 0.5373684763908386, + "learning_rate": 0.0001731317636973986, + "loss": 1.0018, + "step": 5382 + }, + { + "epoch": 0.958511396011396, + "grad_norm": 0.5714904069900513, + "learning_rate": 0.00017312221618661516, + "loss": 1.1855, + "step": 5383 + }, + { + "epoch": 0.9586894586894587, + "grad_norm": 0.5707863569259644, + "learning_rate": 0.00017311266724315377, + "loss": 0.9482, + "step": 5384 + }, + { + "epoch": 0.9588675213675214, + "grad_norm": 0.5856872797012329, + "learning_rate": 0.00017310311686720157, + "loss": 0.9543, + "step": 5385 + }, + { + "epoch": 0.959045584045584, + "grad_norm": 0.5041963458061218, + "learning_rate": 0.00017309356505894568, + "loss": 1.1427, + "step": 5386 + }, + { + "epoch": 0.9592236467236467, + "grad_norm": 0.5409179925918579, + "learning_rate": 0.00017308401181857316, + "loss": 0.8432, + "step": 5387 + }, + { + "epoch": 0.9594017094017094, + "grad_norm": 0.5248702764511108, + "learning_rate": 0.00017307445714627128, + "loss": 1.1403, + "step": 5388 + }, + { + "epoch": 0.9595797720797721, + "grad_norm": 0.50718092918396, + "learning_rate": 0.00017306490104222722, + "loss": 0.9066, + "step": 5389 + }, + { + "epoch": 0.9597578347578347, + "grad_norm": 0.5563821196556091, + "learning_rate": 0.0001730553435066282, + "loss": 1.0204, + "step": 5390 + }, + { + "epoch": 0.9599358974358975, + "grad_norm": 0.5696987509727478, + "learning_rate": 0.00017304578453966146, + "loss": 1.1405, + "step": 5391 + }, + { + "epoch": 0.9601139601139601, + "grad_norm": 0.5927395224571228, + "learning_rate": 0.00017303622414151435, + "loss": 1.0398, + "step": 5392 + }, + { + "epoch": 0.9602920227920227, + "grad_norm": 0.5375707745552063, + "learning_rate": 0.0001730266623123741, + "loss": 0.9519, + "step": 5393 + }, + { + "epoch": 0.9604700854700855, + "grad_norm": 0.457998126745224, + "learning_rate": 0.00017301709905242815, + "loss": 0.8743, + "step": 5394 + }, + { + "epoch": 0.9606481481481481, + "grad_norm": 0.5427796244621277, + "learning_rate": 0.00017300753436186382, + "loss": 1.078, + "step": 5395 + }, + { + "epoch": 0.9608262108262108, + "grad_norm": 0.5458595752716064, + "learning_rate": 0.0001729979682408685, + "loss": 1.1081, + "step": 5396 + }, + { + "epoch": 0.9610042735042735, + "grad_norm": 0.5495280027389526, + "learning_rate": 0.00017298840068962962, + "loss": 1.0141, + "step": 5397 + }, + { + "epoch": 0.9611823361823362, + "grad_norm": 0.5878560543060303, + "learning_rate": 0.00017297883170833465, + "loss": 1.302, + "step": 5398 + }, + { + "epoch": 0.9613603988603988, + "grad_norm": 0.5452881455421448, + "learning_rate": 0.00017296926129717108, + "loss": 0.9929, + "step": 5399 + }, + { + "epoch": 0.9615384615384616, + "grad_norm": 0.6021811366081238, + "learning_rate": 0.0001729596894563264, + "loss": 1.2629, + "step": 5400 + }, + { + "epoch": 0.9617165242165242, + "grad_norm": 0.5820204615592957, + "learning_rate": 0.0001729501161859882, + "loss": 1.0662, + "step": 5401 + }, + { + "epoch": 0.9618945868945868, + "grad_norm": 0.4953218102455139, + "learning_rate": 0.000172940541486344, + "loss": 1.047, + "step": 5402 + }, + { + "epoch": 0.9620726495726496, + "grad_norm": 0.5409793853759766, + "learning_rate": 0.00017293096535758143, + "loss": 1.1993, + "step": 5403 + }, + { + "epoch": 0.9622507122507122, + "grad_norm": 0.49702873826026917, + "learning_rate": 0.00017292138779988805, + "loss": 1.2471, + "step": 5404 + }, + { + "epoch": 0.9624287749287749, + "grad_norm": 0.5743489861488342, + "learning_rate": 0.00017291180881345158, + "loss": 1.0816, + "step": 5405 + }, + { + "epoch": 0.9626068376068376, + "grad_norm": 0.5747945308685303, + "learning_rate": 0.00017290222839845968, + "loss": 1.3548, + "step": 5406 + }, + { + "epoch": 0.9627849002849003, + "grad_norm": 0.5341345071792603, + "learning_rate": 0.00017289264655510005, + "loss": 1.0435, + "step": 5407 + }, + { + "epoch": 0.9629629629629629, + "grad_norm": 0.5719689130783081, + "learning_rate": 0.00017288306328356044, + "loss": 1.2319, + "step": 5408 + }, + { + "epoch": 0.9631410256410257, + "grad_norm": 0.4783279597759247, + "learning_rate": 0.0001728734785840286, + "loss": 0.9397, + "step": 5409 + }, + { + "epoch": 0.9633190883190883, + "grad_norm": 0.4730507731437683, + "learning_rate": 0.00017286389245669233, + "loss": 0.9384, + "step": 5410 + }, + { + "epoch": 0.9634971509971509, + "grad_norm": 0.5309939384460449, + "learning_rate": 0.00017285430490173944, + "loss": 1.098, + "step": 5411 + }, + { + "epoch": 0.9636752136752137, + "grad_norm": 0.5177853107452393, + "learning_rate": 0.0001728447159193578, + "loss": 1.2777, + "step": 5412 + }, + { + "epoch": 0.9638532763532763, + "grad_norm": 0.6437913775444031, + "learning_rate": 0.00017283512550973526, + "loss": 1.2661, + "step": 5413 + }, + { + "epoch": 0.9640313390313391, + "grad_norm": 0.6096072196960449, + "learning_rate": 0.00017282553367305975, + "loss": 0.9569, + "step": 5414 + }, + { + "epoch": 0.9642094017094017, + "grad_norm": 0.5104934573173523, + "learning_rate": 0.00017281594040951918, + "loss": 0.9666, + "step": 5415 + }, + { + "epoch": 0.9643874643874644, + "grad_norm": 0.6178240776062012, + "learning_rate": 0.00017280634571930153, + "loss": 1.1277, + "step": 5416 + }, + { + "epoch": 0.9645655270655271, + "grad_norm": 0.5749034881591797, + "learning_rate": 0.0001727967496025948, + "loss": 1.245, + "step": 5417 + }, + { + "epoch": 0.9647435897435898, + "grad_norm": 0.5036978721618652, + "learning_rate": 0.00017278715205958694, + "loss": 1.3049, + "step": 5418 + }, + { + "epoch": 0.9649216524216524, + "grad_norm": 0.5593041777610779, + "learning_rate": 0.00017277755309046605, + "loss": 1.2304, + "step": 5419 + }, + { + "epoch": 0.9650997150997151, + "grad_norm": 0.5446555614471436, + "learning_rate": 0.0001727679526954202, + "loss": 0.732, + "step": 5420 + }, + { + "epoch": 0.9652777777777778, + "grad_norm": 0.6063070297241211, + "learning_rate": 0.00017275835087463747, + "loss": 1.3723, + "step": 5421 + }, + { + "epoch": 0.9654558404558404, + "grad_norm": 0.4994211792945862, + "learning_rate": 0.00017274874762830602, + "loss": 1.0505, + "step": 5422 + }, + { + "epoch": 0.9656339031339032, + "grad_norm": 0.49396973848342896, + "learning_rate": 0.00017273914295661395, + "loss": 0.8691, + "step": 5423 + }, + { + "epoch": 0.9658119658119658, + "grad_norm": 0.5067027807235718, + "learning_rate": 0.0001727295368597495, + "loss": 0.9744, + "step": 5424 + }, + { + "epoch": 0.9659900284900285, + "grad_norm": 0.6720643043518066, + "learning_rate": 0.00017271992933790085, + "loss": 1.1513, + "step": 5425 + }, + { + "epoch": 0.9661680911680912, + "grad_norm": 0.5494341254234314, + "learning_rate": 0.00017271032039125624, + "loss": 0.8295, + "step": 5426 + }, + { + "epoch": 0.9663461538461539, + "grad_norm": 0.644332230091095, + "learning_rate": 0.00017270071002000394, + "loss": 1.0043, + "step": 5427 + }, + { + "epoch": 0.9665242165242165, + "grad_norm": 0.5658500790596008, + "learning_rate": 0.00017269109822433225, + "loss": 1.2575, + "step": 5428 + }, + { + "epoch": 0.9667022792022792, + "grad_norm": 0.5163155794143677, + "learning_rate": 0.00017268148500442952, + "loss": 1.1391, + "step": 5429 + }, + { + "epoch": 0.9668803418803419, + "grad_norm": 0.5113703608512878, + "learning_rate": 0.00017267187036048404, + "loss": 1.0819, + "step": 5430 + }, + { + "epoch": 0.9670584045584045, + "grad_norm": 0.6339422464370728, + "learning_rate": 0.00017266225429268426, + "loss": 1.0733, + "step": 5431 + }, + { + "epoch": 0.9672364672364673, + "grad_norm": 0.5158288478851318, + "learning_rate": 0.0001726526368012185, + "loss": 0.9518, + "step": 5432 + }, + { + "epoch": 0.9674145299145299, + "grad_norm": 0.593717634677887, + "learning_rate": 0.00017264301788627527, + "loss": 0.9416, + "step": 5433 + }, + { + "epoch": 0.9675925925925926, + "grad_norm": 0.49593186378479004, + "learning_rate": 0.00017263339754804301, + "loss": 1.0307, + "step": 5434 + }, + { + "epoch": 0.9677706552706553, + "grad_norm": 0.44032949209213257, + "learning_rate": 0.00017262377578671024, + "loss": 0.7884, + "step": 5435 + }, + { + "epoch": 0.967948717948718, + "grad_norm": 0.513073742389679, + "learning_rate": 0.00017261415260246538, + "loss": 0.9797, + "step": 5436 + }, + { + "epoch": 0.9681267806267806, + "grad_norm": 0.5737422108650208, + "learning_rate": 0.0001726045279954971, + "loss": 1.0487, + "step": 5437 + }, + { + "epoch": 0.9683048433048433, + "grad_norm": 0.5385867953300476, + "learning_rate": 0.0001725949019659939, + "loss": 1.4166, + "step": 5438 + }, + { + "epoch": 0.968482905982906, + "grad_norm": 0.5224326848983765, + "learning_rate": 0.00017258527451414438, + "loss": 1.195, + "step": 5439 + }, + { + "epoch": 0.9686609686609686, + "grad_norm": 0.5305148363113403, + "learning_rate": 0.0001725756456401372, + "loss": 1.0301, + "step": 5440 + }, + { + "epoch": 0.9688390313390314, + "grad_norm": 0.532588005065918, + "learning_rate": 0.000172566015344161, + "loss": 1.1269, + "step": 5441 + }, + { + "epoch": 0.969017094017094, + "grad_norm": 0.5812515020370483, + "learning_rate": 0.0001725563836264045, + "loss": 1.1787, + "step": 5442 + }, + { + "epoch": 0.9691951566951567, + "grad_norm": 0.4962109327316284, + "learning_rate": 0.00017254675048705638, + "loss": 1.0639, + "step": 5443 + }, + { + "epoch": 0.9693732193732194, + "grad_norm": 0.5094883441925049, + "learning_rate": 0.00017253711592630534, + "loss": 1.0922, + "step": 5444 + }, + { + "epoch": 0.969551282051282, + "grad_norm": 0.5728049874305725, + "learning_rate": 0.00017252747994434025, + "loss": 1.1237, + "step": 5445 + }, + { + "epoch": 0.9697293447293447, + "grad_norm": 0.5406180620193481, + "learning_rate": 0.00017251784254134983, + "loss": 1.1161, + "step": 5446 + }, + { + "epoch": 0.9699074074074074, + "grad_norm": 0.5724552869796753, + "learning_rate": 0.00017250820371752292, + "loss": 1.2205, + "step": 5447 + }, + { + "epoch": 0.9700854700854701, + "grad_norm": 0.5698846578598022, + "learning_rate": 0.0001724985634730484, + "loss": 1.1472, + "step": 5448 + }, + { + "epoch": 0.9702635327635327, + "grad_norm": 0.5315805673599243, + "learning_rate": 0.0001724889218081151, + "loss": 1.0253, + "step": 5449 + }, + { + "epoch": 0.9704415954415955, + "grad_norm": 0.5970377326011658, + "learning_rate": 0.000172479278722912, + "loss": 1.3033, + "step": 5450 + }, + { + "epoch": 0.9706196581196581, + "grad_norm": 0.6149488687515259, + "learning_rate": 0.00017246963421762798, + "loss": 1.0689, + "step": 5451 + }, + { + "epoch": 0.9707977207977208, + "grad_norm": 0.4848574995994568, + "learning_rate": 0.00017245998829245202, + "loss": 0.8829, + "step": 5452 + }, + { + "epoch": 0.9709757834757835, + "grad_norm": 0.6073294281959534, + "learning_rate": 0.00017245034094757312, + "loss": 1.2378, + "step": 5453 + }, + { + "epoch": 0.9711538461538461, + "grad_norm": 0.6362034678459167, + "learning_rate": 0.00017244069218318026, + "loss": 1.3606, + "step": 5454 + }, + { + "epoch": 0.9713319088319088, + "grad_norm": 0.5353880524635315, + "learning_rate": 0.00017243104199946257, + "loss": 1.1288, + "step": 5455 + }, + { + "epoch": 0.9715099715099715, + "grad_norm": 0.5096352100372314, + "learning_rate": 0.00017242139039660902, + "loss": 1.0056, + "step": 5456 + }, + { + "epoch": 0.9716880341880342, + "grad_norm": 0.5086682438850403, + "learning_rate": 0.00017241173737480884, + "loss": 1.091, + "step": 5457 + }, + { + "epoch": 0.9718660968660968, + "grad_norm": 0.5034295320510864, + "learning_rate": 0.000172402082934251, + "loss": 0.9749, + "step": 5458 + }, + { + "epoch": 0.9720441595441596, + "grad_norm": 0.5205379724502563, + "learning_rate": 0.0001723924270751248, + "loss": 1.1068, + "step": 5459 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 0.5904826521873474, + "learning_rate": 0.00017238276979761937, + "loss": 1.0613, + "step": 5460 + }, + { + "epoch": 0.9724002849002849, + "grad_norm": 0.6415045261383057, + "learning_rate": 0.0001723731111019239, + "loss": 1.2126, + "step": 5461 + }, + { + "epoch": 0.9725783475783476, + "grad_norm": 0.5769147872924805, + "learning_rate": 0.0001723634509882277, + "loss": 1.337, + "step": 5462 + }, + { + "epoch": 0.9727564102564102, + "grad_norm": 0.5585111975669861, + "learning_rate": 0.00017235378945671998, + "loss": 1.3922, + "step": 5463 + }, + { + "epoch": 0.9729344729344729, + "grad_norm": 0.5788411498069763, + "learning_rate": 0.00017234412650759008, + "loss": 0.8532, + "step": 5464 + }, + { + "epoch": 0.9731125356125356, + "grad_norm": 0.5617673397064209, + "learning_rate": 0.00017233446214102728, + "loss": 1.2575, + "step": 5465 + }, + { + "epoch": 0.9732905982905983, + "grad_norm": 0.4227815568447113, + "learning_rate": 0.00017232479635722093, + "loss": 1.0618, + "step": 5466 + }, + { + "epoch": 0.9734686609686609, + "grad_norm": 0.49751797318458557, + "learning_rate": 0.00017231512915636047, + "loss": 0.7714, + "step": 5467 + }, + { + "epoch": 0.9736467236467237, + "grad_norm": 0.5983800292015076, + "learning_rate": 0.0001723054605386353, + "loss": 1.2297, + "step": 5468 + }, + { + "epoch": 0.9738247863247863, + "grad_norm": 0.543394923210144, + "learning_rate": 0.0001722957905042348, + "loss": 1.0078, + "step": 5469 + }, + { + "epoch": 0.9740028490028491, + "grad_norm": 0.5633566975593567, + "learning_rate": 0.00017228611905334846, + "loss": 1.0938, + "step": 5470 + }, + { + "epoch": 0.9741809116809117, + "grad_norm": 0.49377235770225525, + "learning_rate": 0.00017227644618616578, + "loss": 1.096, + "step": 5471 + }, + { + "epoch": 0.9743589743589743, + "grad_norm": 0.4963362216949463, + "learning_rate": 0.00017226677190287627, + "loss": 1.0003, + "step": 5472 + }, + { + "epoch": 0.9745370370370371, + "grad_norm": 0.4483006000518799, + "learning_rate": 0.00017225709620366953, + "loss": 0.8623, + "step": 5473 + }, + { + "epoch": 0.9747150997150997, + "grad_norm": 0.5429352521896362, + "learning_rate": 0.00017224741908873506, + "loss": 1.1383, + "step": 5474 + }, + { + "epoch": 0.9748931623931624, + "grad_norm": 0.5871657729148865, + "learning_rate": 0.0001722377405582625, + "loss": 1.2005, + "step": 5475 + }, + { + "epoch": 0.9750712250712251, + "grad_norm": 0.6002383828163147, + "learning_rate": 0.0001722280606124415, + "loss": 1.0696, + "step": 5476 + }, + { + "epoch": 0.9752492877492878, + "grad_norm": 0.5351617336273193, + "learning_rate": 0.00017221837925146164, + "loss": 1.243, + "step": 5477 + }, + { + "epoch": 0.9754273504273504, + "grad_norm": 0.46613118052482605, + "learning_rate": 0.00017220869647551268, + "loss": 1.0344, + "step": 5478 + }, + { + "epoch": 0.9756054131054132, + "grad_norm": 0.6015593409538269, + "learning_rate": 0.00017219901228478432, + "loss": 1.082, + "step": 5479 + }, + { + "epoch": 0.9757834757834758, + "grad_norm": 0.5829521417617798, + "learning_rate": 0.0001721893266794663, + "loss": 0.8683, + "step": 5480 + }, + { + "epoch": 0.9759615384615384, + "grad_norm": 0.6344960927963257, + "learning_rate": 0.00017217963965974838, + "loss": 1.1048, + "step": 5481 + }, + { + "epoch": 0.9761396011396012, + "grad_norm": 0.5586308240890503, + "learning_rate": 0.00017216995122582034, + "loss": 0.9657, + "step": 5482 + }, + { + "epoch": 0.9763176638176638, + "grad_norm": 0.48625239729881287, + "learning_rate": 0.00017216026137787204, + "loss": 1.1026, + "step": 5483 + }, + { + "epoch": 0.9764957264957265, + "grad_norm": 0.5625223517417908, + "learning_rate": 0.00017215057011609332, + "loss": 1.1579, + "step": 5484 + }, + { + "epoch": 0.9766737891737892, + "grad_norm": 0.6016653776168823, + "learning_rate": 0.0001721408774406741, + "loss": 1.1777, + "step": 5485 + }, + { + "epoch": 0.9768518518518519, + "grad_norm": 0.5444921851158142, + "learning_rate": 0.00017213118335180418, + "loss": 1.119, + "step": 5486 + }, + { + "epoch": 0.9770299145299145, + "grad_norm": 0.5574755668640137, + "learning_rate": 0.0001721214878496736, + "loss": 1.1128, + "step": 5487 + }, + { + "epoch": 0.9772079772079773, + "grad_norm": 0.5486113429069519, + "learning_rate": 0.00017211179093447226, + "loss": 1.1673, + "step": 5488 + }, + { + "epoch": 0.9773860398860399, + "grad_norm": 0.5545483231544495, + "learning_rate": 0.00017210209260639018, + "loss": 1.1748, + "step": 5489 + }, + { + "epoch": 0.9775641025641025, + "grad_norm": 0.5756667256355286, + "learning_rate": 0.0001720923928656174, + "loss": 1.2377, + "step": 5490 + }, + { + "epoch": 0.9777421652421653, + "grad_norm": 0.5744972229003906, + "learning_rate": 0.00017208269171234392, + "loss": 1.1242, + "step": 5491 + }, + { + "epoch": 0.9779202279202279, + "grad_norm": 0.6109468340873718, + "learning_rate": 0.00017207298914675984, + "loss": 1.1948, + "step": 5492 + }, + { + "epoch": 0.9780982905982906, + "grad_norm": 0.5195167660713196, + "learning_rate": 0.00017206328516905525, + "loss": 1.0941, + "step": 5493 + }, + { + "epoch": 0.9782763532763533, + "grad_norm": 0.5549042224884033, + "learning_rate": 0.0001720535797794203, + "loss": 1.1503, + "step": 5494 + }, + { + "epoch": 0.978454415954416, + "grad_norm": 0.6317743062973022, + "learning_rate": 0.0001720438729780451, + "loss": 1.3468, + "step": 5495 + }, + { + "epoch": 0.9786324786324786, + "grad_norm": 0.5932528972625732, + "learning_rate": 0.0001720341647651199, + "loss": 1.105, + "step": 5496 + }, + { + "epoch": 0.9788105413105413, + "grad_norm": 0.607880175113678, + "learning_rate": 0.00017202445514083488, + "loss": 1.1465, + "step": 5497 + }, + { + "epoch": 0.978988603988604, + "grad_norm": 0.49227309226989746, + "learning_rate": 0.00017201474410538027, + "loss": 0.9075, + "step": 5498 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 0.5059443116188049, + "learning_rate": 0.00017200503165894636, + "loss": 1.0483, + "step": 5499 + }, + { + "epoch": 0.9793447293447294, + "grad_norm": 0.5792799592018127, + "learning_rate": 0.0001719953178017234, + "loss": 1.0987, + "step": 5500 + }, + { + "epoch": 0.979522792022792, + "grad_norm": 0.5010457038879395, + "learning_rate": 0.00017198560253390177, + "loss": 1.1051, + "step": 5501 + }, + { + "epoch": 0.9797008547008547, + "grad_norm": 0.5866543054580688, + "learning_rate": 0.0001719758858556718, + "loss": 1.2824, + "step": 5502 + }, + { + "epoch": 0.9798789173789174, + "grad_norm": 0.5392137169837952, + "learning_rate": 0.00017196616776722382, + "loss": 0.886, + "step": 5503 + }, + { + "epoch": 0.98005698005698, + "grad_norm": 0.5200899839401245, + "learning_rate": 0.00017195644826874834, + "loss": 1.1504, + "step": 5504 + }, + { + "epoch": 0.9802350427350427, + "grad_norm": 0.533159077167511, + "learning_rate": 0.00017194672736043569, + "loss": 1.1216, + "step": 5505 + }, + { + "epoch": 0.9804131054131054, + "grad_norm": 0.5543524622917175, + "learning_rate": 0.0001719370050424764, + "loss": 1.0161, + "step": 5506 + }, + { + "epoch": 0.9805911680911681, + "grad_norm": 0.5315365195274353, + "learning_rate": 0.00017192728131506092, + "loss": 1.0509, + "step": 5507 + }, + { + "epoch": 0.9807692307692307, + "grad_norm": 0.5406147837638855, + "learning_rate": 0.00017191755617837977, + "loss": 1.0695, + "step": 5508 + }, + { + "epoch": 0.9809472934472935, + "grad_norm": 0.4563386142253876, + "learning_rate": 0.00017190782963262354, + "loss": 0.995, + "step": 5509 + }, + { + "epoch": 0.9811253561253561, + "grad_norm": 0.5456405282020569, + "learning_rate": 0.00017189810167798274, + "loss": 1.0546, + "step": 5510 + }, + { + "epoch": 0.9813034188034188, + "grad_norm": 0.6275575160980225, + "learning_rate": 0.00017188837231464795, + "loss": 1.0432, + "step": 5511 + }, + { + "epoch": 0.9814814814814815, + "grad_norm": 0.49735602736473083, + "learning_rate": 0.0001718786415428099, + "loss": 1.035, + "step": 5512 + }, + { + "epoch": 0.9816595441595442, + "grad_norm": 0.5234259963035583, + "learning_rate": 0.00017186890936265916, + "loss": 1.0918, + "step": 5513 + }, + { + "epoch": 0.9818376068376068, + "grad_norm": 0.5091170072555542, + "learning_rate": 0.00017185917577438643, + "loss": 1.0239, + "step": 5514 + }, + { + "epoch": 0.9820156695156695, + "grad_norm": 0.6155703067779541, + "learning_rate": 0.00017184944077818244, + "loss": 1.2366, + "step": 5515 + }, + { + "epoch": 0.9821937321937322, + "grad_norm": 0.5074070692062378, + "learning_rate": 0.0001718397043742379, + "loss": 1.0318, + "step": 5516 + }, + { + "epoch": 0.9823717948717948, + "grad_norm": 0.5234423279762268, + "learning_rate": 0.0001718299665627436, + "loss": 1.0322, + "step": 5517 + }, + { + "epoch": 0.9825498575498576, + "grad_norm": 0.5783474445343018, + "learning_rate": 0.0001718202273438903, + "loss": 0.9486, + "step": 5518 + }, + { + "epoch": 0.9827279202279202, + "grad_norm": 0.5708683133125305, + "learning_rate": 0.00017181048671786886, + "loss": 1.0785, + "step": 5519 + }, + { + "epoch": 0.9829059829059829, + "grad_norm": 0.5985961556434631, + "learning_rate": 0.00017180074468487009, + "loss": 1.198, + "step": 5520 + }, + { + "epoch": 0.9830840455840456, + "grad_norm": 0.5711352229118347, + "learning_rate": 0.0001717910012450849, + "loss": 1.0386, + "step": 5521 + }, + { + "epoch": 0.9832621082621082, + "grad_norm": 0.5338063836097717, + "learning_rate": 0.00017178125639870416, + "loss": 1.1594, + "step": 5522 + }, + { + "epoch": 0.9834401709401709, + "grad_norm": 0.6144943237304688, + "learning_rate": 0.00017177151014591881, + "loss": 1.1083, + "step": 5523 + }, + { + "epoch": 0.9836182336182336, + "grad_norm": 0.547285795211792, + "learning_rate": 0.00017176176248691983, + "loss": 1.1507, + "step": 5524 + }, + { + "epoch": 0.9837962962962963, + "grad_norm": 0.5807644724845886, + "learning_rate": 0.00017175201342189817, + "loss": 1.3044, + "step": 5525 + }, + { + "epoch": 0.9839743589743589, + "grad_norm": 0.5229477882385254, + "learning_rate": 0.00017174226295104485, + "loss": 1.2622, + "step": 5526 + }, + { + "epoch": 0.9841524216524217, + "grad_norm": 0.6100695133209229, + "learning_rate": 0.00017173251107455094, + "loss": 1.2026, + "step": 5527 + }, + { + "epoch": 0.9843304843304843, + "grad_norm": 0.5410884618759155, + "learning_rate": 0.00017172275779260744, + "loss": 1.2964, + "step": 5528 + }, + { + "epoch": 0.9845085470085471, + "grad_norm": 0.5937406420707703, + "learning_rate": 0.00017171300310540554, + "loss": 1.1435, + "step": 5529 + }, + { + "epoch": 0.9846866096866097, + "grad_norm": 0.56817227602005, + "learning_rate": 0.00017170324701313634, + "loss": 1.0099, + "step": 5530 + }, + { + "epoch": 0.9848646723646723, + "grad_norm": 0.5776323080062866, + "learning_rate": 0.00017169348951599092, + "loss": 1.3539, + "step": 5531 + }, + { + "epoch": 0.9850427350427351, + "grad_norm": 0.5208535194396973, + "learning_rate": 0.0001716837306141605, + "loss": 1.2306, + "step": 5532 + }, + { + "epoch": 0.9852207977207977, + "grad_norm": 0.552173376083374, + "learning_rate": 0.0001716739703078363, + "loss": 1.0551, + "step": 5533 + }, + { + "epoch": 0.9853988603988604, + "grad_norm": 0.5327515602111816, + "learning_rate": 0.00017166420859720955, + "loss": 1.2443, + "step": 5534 + }, + { + "epoch": 0.9855769230769231, + "grad_norm": 0.5255244374275208, + "learning_rate": 0.0001716544454824715, + "loss": 1.005, + "step": 5535 + }, + { + "epoch": 0.9857549857549858, + "grad_norm": 0.4753847122192383, + "learning_rate": 0.00017164468096381343, + "loss": 1.0081, + "step": 5536 + }, + { + "epoch": 0.9859330484330484, + "grad_norm": 0.5261829495429993, + "learning_rate": 0.00017163491504142665, + "loss": 1.2249, + "step": 5537 + }, + { + "epoch": 0.9861111111111112, + "grad_norm": 0.46499499678611755, + "learning_rate": 0.00017162514771550255, + "loss": 0.8759, + "step": 5538 + }, + { + "epoch": 0.9862891737891738, + "grad_norm": 0.5233004689216614, + "learning_rate": 0.00017161537898623247, + "loss": 1.0474, + "step": 5539 + }, + { + "epoch": 0.9864672364672364, + "grad_norm": 0.46905553340911865, + "learning_rate": 0.00017160560885380778, + "loss": 0.9033, + "step": 5540 + }, + { + "epoch": 0.9866452991452992, + "grad_norm": 0.5816231369972229, + "learning_rate": 0.00017159583731841998, + "loss": 1.0628, + "step": 5541 + }, + { + "epoch": 0.9868233618233618, + "grad_norm": 0.4575413167476654, + "learning_rate": 0.00017158606438026045, + "loss": 1.0446, + "step": 5542 + }, + { + "epoch": 0.9870014245014245, + "grad_norm": 0.5968109965324402, + "learning_rate": 0.00017157629003952067, + "loss": 1.032, + "step": 5543 + }, + { + "epoch": 0.9871794871794872, + "grad_norm": 0.5316148400306702, + "learning_rate": 0.00017156651429639218, + "loss": 0.9167, + "step": 5544 + }, + { + "epoch": 0.9873575498575499, + "grad_norm": 0.5185125470161438, + "learning_rate": 0.00017155673715106651, + "loss": 1.1527, + "step": 5545 + }, + { + "epoch": 0.9875356125356125, + "grad_norm": 0.5167772769927979, + "learning_rate": 0.00017154695860373525, + "loss": 0.9954, + "step": 5546 + }, + { + "epoch": 0.9877136752136753, + "grad_norm": 0.6406680345535278, + "learning_rate": 0.00017153717865458994, + "loss": 1.2758, + "step": 5547 + }, + { + "epoch": 0.9878917378917379, + "grad_norm": 0.5223956108093262, + "learning_rate": 0.00017152739730382223, + "loss": 1.1526, + "step": 5548 + }, + { + "epoch": 0.9880698005698005, + "grad_norm": 0.6131790280342102, + "learning_rate": 0.00017151761455162375, + "loss": 1.1024, + "step": 5549 + }, + { + "epoch": 0.9882478632478633, + "grad_norm": 0.5574753880500793, + "learning_rate": 0.00017150783039818616, + "loss": 0.9733, + "step": 5550 + }, + { + "epoch": 0.9884259259259259, + "grad_norm": 0.5417882800102234, + "learning_rate": 0.0001714980448437012, + "loss": 1.2244, + "step": 5551 + }, + { + "epoch": 0.9886039886039886, + "grad_norm": 0.6217474341392517, + "learning_rate": 0.0001714882578883606, + "loss": 0.9224, + "step": 5552 + }, + { + "epoch": 0.9887820512820513, + "grad_norm": 0.5846285223960876, + "learning_rate": 0.00017147846953235606, + "loss": 1.2429, + "step": 5553 + }, + { + "epoch": 0.988960113960114, + "grad_norm": 0.5924782752990723, + "learning_rate": 0.00017146867977587936, + "loss": 0.9907, + "step": 5554 + }, + { + "epoch": 0.9891381766381766, + "grad_norm": 0.5756853818893433, + "learning_rate": 0.00017145888861912242, + "loss": 1.1266, + "step": 5555 + }, + { + "epoch": 0.9893162393162394, + "grad_norm": 0.5277376770973206, + "learning_rate": 0.00017144909606227693, + "loss": 1.1676, + "step": 5556 + }, + { + "epoch": 0.989494301994302, + "grad_norm": 0.5138902068138123, + "learning_rate": 0.00017143930210553485, + "loss": 0.9864, + "step": 5557 + }, + { + "epoch": 0.9896723646723646, + "grad_norm": 0.8072507977485657, + "learning_rate": 0.00017142950674908805, + "loss": 1.111, + "step": 5558 + }, + { + "epoch": 0.9898504273504274, + "grad_norm": 0.5641721487045288, + "learning_rate": 0.00017141970999312844, + "loss": 0.9106, + "step": 5559 + }, + { + "epoch": 0.99002849002849, + "grad_norm": 0.5260798931121826, + "learning_rate": 0.000171409911837848, + "loss": 1.1609, + "step": 5560 + }, + { + "epoch": 0.9902065527065527, + "grad_norm": 0.5398530960083008, + "learning_rate": 0.00017140011228343864, + "loss": 1.0368, + "step": 5561 + }, + { + "epoch": 0.9903846153846154, + "grad_norm": 0.6011313199996948, + "learning_rate": 0.00017139031133009245, + "loss": 1.1314, + "step": 5562 + }, + { + "epoch": 0.9905626780626781, + "grad_norm": 0.6194971203804016, + "learning_rate": 0.00017138050897800135, + "loss": 1.3493, + "step": 5563 + }, + { + "epoch": 0.9907407407407407, + "grad_norm": 0.5779356956481934, + "learning_rate": 0.0001713707052273575, + "loss": 0.943, + "step": 5564 + }, + { + "epoch": 0.9909188034188035, + "grad_norm": 0.5321127772331238, + "learning_rate": 0.00017136090007835293, + "loss": 0.7914, + "step": 5565 + }, + { + "epoch": 0.9910968660968661, + "grad_norm": 0.5470426678657532, + "learning_rate": 0.00017135109353117977, + "loss": 1.2113, + "step": 5566 + }, + { + "epoch": 0.9912749287749287, + "grad_norm": 0.5551436543464661, + "learning_rate": 0.00017134128558603012, + "loss": 0.8932, + "step": 5567 + }, + { + "epoch": 0.9914529914529915, + "grad_norm": 0.45770928263664246, + "learning_rate": 0.0001713314762430962, + "loss": 1.0061, + "step": 5568 + }, + { + "epoch": 0.9916310541310541, + "grad_norm": 0.5578967332839966, + "learning_rate": 0.00017132166550257017, + "loss": 1.148, + "step": 5569 + }, + { + "epoch": 0.9918091168091168, + "grad_norm": 0.5086452960968018, + "learning_rate": 0.0001713118533646443, + "loss": 0.9803, + "step": 5570 + }, + { + "epoch": 0.9919871794871795, + "grad_norm": 0.4714745879173279, + "learning_rate": 0.00017130203982951078, + "loss": 1.0176, + "step": 5571 + }, + { + "epoch": 0.9921652421652422, + "grad_norm": 0.6254406571388245, + "learning_rate": 0.0001712922248973619, + "loss": 1.0932, + "step": 5572 + }, + { + "epoch": 0.9923433048433048, + "grad_norm": 0.5005003809928894, + "learning_rate": 0.00017128240856838998, + "loss": 1.0783, + "step": 5573 + }, + { + "epoch": 0.9925213675213675, + "grad_norm": 0.5668206214904785, + "learning_rate": 0.00017127259084278733, + "loss": 1.0404, + "step": 5574 + }, + { + "epoch": 0.9926994301994302, + "grad_norm": 0.4976036250591278, + "learning_rate": 0.00017126277172074632, + "loss": 1.1437, + "step": 5575 + }, + { + "epoch": 0.9928774928774928, + "grad_norm": 0.567546546459198, + "learning_rate": 0.00017125295120245935, + "loss": 1.2188, + "step": 5576 + }, + { + "epoch": 0.9930555555555556, + "grad_norm": 0.5614372491836548, + "learning_rate": 0.0001712431292881188, + "loss": 0.9187, + "step": 5577 + }, + { + "epoch": 0.9932336182336182, + "grad_norm": 0.6117973327636719, + "learning_rate": 0.00017123330597791712, + "loss": 1.1285, + "step": 5578 + }, + { + "epoch": 0.9934116809116809, + "grad_norm": 0.6000342965126038, + "learning_rate": 0.00017122348127204676, + "loss": 0.9837, + "step": 5579 + }, + { + "epoch": 0.9935897435897436, + "grad_norm": 0.5453050136566162, + "learning_rate": 0.0001712136551707003, + "loss": 0.8771, + "step": 5580 + }, + { + "epoch": 0.9937678062678063, + "grad_norm": 0.49603891372680664, + "learning_rate": 0.00017120382767407018, + "loss": 1.0754, + "step": 5581 + }, + { + "epoch": 0.9939458689458689, + "grad_norm": 0.48031488060951233, + "learning_rate": 0.00017119399878234894, + "loss": 0.6933, + "step": 5582 + }, + { + "epoch": 0.9941239316239316, + "grad_norm": 0.6048742532730103, + "learning_rate": 0.0001711841684957292, + "loss": 0.9696, + "step": 5583 + }, + { + "epoch": 0.9943019943019943, + "grad_norm": 0.5183123350143433, + "learning_rate": 0.00017117433681440355, + "loss": 1.1313, + "step": 5584 + }, + { + "epoch": 0.9944800569800569, + "grad_norm": 0.504916250705719, + "learning_rate": 0.00017116450373856466, + "loss": 1.0273, + "step": 5585 + }, + { + "epoch": 0.9946581196581197, + "grad_norm": 0.5804886817932129, + "learning_rate": 0.0001711546692684051, + "loss": 1.1162, + "step": 5586 + }, + { + "epoch": 0.9948361823361823, + "grad_norm": 0.5531938672065735, + "learning_rate": 0.0001711448334041176, + "loss": 1.2893, + "step": 5587 + }, + { + "epoch": 0.9950142450142451, + "grad_norm": 0.5079928636550903, + "learning_rate": 0.00017113499614589492, + "loss": 1.0393, + "step": 5588 + }, + { + "epoch": 0.9951923076923077, + "grad_norm": 0.5421964526176453, + "learning_rate": 0.00017112515749392973, + "loss": 0.8844, + "step": 5589 + }, + { + "epoch": 0.9953703703703703, + "grad_norm": 0.4834558367729187, + "learning_rate": 0.00017111531744841486, + "loss": 1.0187, + "step": 5590 + }, + { + "epoch": 0.9955484330484331, + "grad_norm": 0.6704340577125549, + "learning_rate": 0.00017110547600954307, + "loss": 0.8524, + "step": 5591 + }, + { + "epoch": 0.9957264957264957, + "grad_norm": 0.4578927159309387, + "learning_rate": 0.00017109563317750718, + "loss": 1.059, + "step": 5592 + }, + { + "epoch": 0.9959045584045584, + "grad_norm": 0.5563494563102722, + "learning_rate": 0.00017108578895250006, + "loss": 1.1211, + "step": 5593 + }, + { + "epoch": 0.9960826210826211, + "grad_norm": 0.5272170901298523, + "learning_rate": 0.00017107594333471454, + "loss": 0.9224, + "step": 5594 + }, + { + "epoch": 0.9962606837606838, + "grad_norm": 0.5697501301765442, + "learning_rate": 0.00017106609632434357, + "loss": 1.2223, + "step": 5595 + }, + { + "epoch": 0.9964387464387464, + "grad_norm": 0.5385653376579285, + "learning_rate": 0.00017105624792158007, + "loss": 1.0809, + "step": 5596 + }, + { + "epoch": 0.9966168091168092, + "grad_norm": 0.5608006119728088, + "learning_rate": 0.000171046398126617, + "loss": 1.3936, + "step": 5597 + }, + { + "epoch": 0.9967948717948718, + "grad_norm": 0.5063132643699646, + "learning_rate": 0.00017103654693964736, + "loss": 1.2086, + "step": 5598 + }, + { + "epoch": 0.9969729344729344, + "grad_norm": 0.6014235019683838, + "learning_rate": 0.00017102669436086415, + "loss": 1.1231, + "step": 5599 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 0.49549567699432373, + "learning_rate": 0.00017101684039046036, + "loss": 1.0013, + "step": 5600 + }, + { + "epoch": 0.9973290598290598, + "grad_norm": 0.517464816570282, + "learning_rate": 0.00017100698502862916, + "loss": 1.1143, + "step": 5601 + }, + { + "epoch": 0.9975071225071225, + "grad_norm": 0.514281153678894, + "learning_rate": 0.00017099712827556358, + "loss": 1.0336, + "step": 5602 + }, + { + "epoch": 0.9976851851851852, + "grad_norm": 0.5378567576408386, + "learning_rate": 0.00017098727013145672, + "loss": 0.8278, + "step": 5603 + }, + { + "epoch": 0.9978632478632479, + "grad_norm": 0.5098404884338379, + "learning_rate": 0.0001709774105965018, + "loss": 0.9902, + "step": 5604 + }, + { + "epoch": 0.9980413105413105, + "grad_norm": 0.6231759190559387, + "learning_rate": 0.00017096754967089198, + "loss": 1.0564, + "step": 5605 + }, + { + "epoch": 0.9982193732193733, + "grad_norm": 0.47434380650520325, + "learning_rate": 0.00017095768735482042, + "loss": 0.7457, + "step": 5606 + }, + { + "epoch": 0.9983974358974359, + "grad_norm": 0.5771013498306274, + "learning_rate": 0.00017094782364848035, + "loss": 1.1191, + "step": 5607 + }, + { + "epoch": 0.9985754985754985, + "grad_norm": 0.5617234706878662, + "learning_rate": 0.00017093795855206508, + "loss": 1.0779, + "step": 5608 + }, + { + "epoch": 0.9987535612535613, + "grad_norm": 0.6573554873466492, + "learning_rate": 0.00017092809206576792, + "loss": 1.0191, + "step": 5609 + }, + { + "epoch": 0.9989316239316239, + "grad_norm": 0.482834130525589, + "learning_rate": 0.00017091822418978207, + "loss": 1.0119, + "step": 5610 + }, + { + "epoch": 0.9991096866096866, + "grad_norm": 0.47496405243873596, + "learning_rate": 0.000170908354924301, + "loss": 0.8297, + "step": 5611 + }, + { + "epoch": 0.9992877492877493, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.00017089848426951796, + "loss": 1.1511, + "step": 5612 + }, + { + "epoch": 0.999465811965812, + "grad_norm": 0.5402522683143616, + "learning_rate": 0.00017088861222562643, + "loss": 1.1401, + "step": 5613 + }, + { + "epoch": 0.9996438746438746, + "grad_norm": 0.546302318572998, + "learning_rate": 0.00017087873879281977, + "loss": 0.8611, + "step": 5614 + }, + { + "epoch": 0.9998219373219374, + "grad_norm": 0.44279807806015015, + "learning_rate": 0.0001708688639712915, + "loss": 0.79, + "step": 5615 + }, + { + "epoch": 1.0, + "grad_norm": 0.5514659285545349, + "learning_rate": 0.00017085898776123502, + "loss": 1.0709, + "step": 5616 + }, + { + "epoch": 1.0, + "eval_loss": 1.093075156211853, + "eval_runtime": 24.6155, + "eval_samples_per_second": 42.29, + "eval_steps_per_second": 21.166, + "step": 5616 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.6290156841278076, + "learning_rate": 0.0001708491101628439, + "loss": 1.1786, + "step": 5617 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.4703841209411621, + "learning_rate": 0.00017083923117631162, + "loss": 0.9548, + "step": 5618 + }, + { + "epoch": 1.0003561253561253, + "grad_norm": 0.4518105089664459, + "learning_rate": 0.0001708293508018318, + "loss": 1.0089, + "step": 5619 + }, + { + "epoch": 1.000534188034188, + "grad_norm": 0.5658619403839111, + "learning_rate": 0.00017081946903959794, + "loss": 0.9466, + "step": 5620 + }, + { + "epoch": 1.0007122507122508, + "grad_norm": 0.6153838634490967, + "learning_rate": 0.00017080958588980372, + "loss": 1.2898, + "step": 5621 + }, + { + "epoch": 1.0008903133903133, + "grad_norm": 0.5245628952980042, + "learning_rate": 0.00017079970135264275, + "loss": 1.1702, + "step": 5622 + }, + { + "epoch": 1.001068376068376, + "grad_norm": 0.5291880965232849, + "learning_rate": 0.00017078981542830875, + "loss": 1.0779, + "step": 5623 + }, + { + "epoch": 1.0012464387464388, + "grad_norm": 0.500579297542572, + "learning_rate": 0.0001707799281169953, + "loss": 0.9587, + "step": 5624 + }, + { + "epoch": 1.0014245014245013, + "grad_norm": 0.45739707350730896, + "learning_rate": 0.00017077003941889625, + "loss": 0.9373, + "step": 5625 + }, + { + "epoch": 1.001602564102564, + "grad_norm": 0.5513401031494141, + "learning_rate": 0.00017076014933420526, + "loss": 1.0368, + "step": 5626 + }, + { + "epoch": 1.0017806267806268, + "grad_norm": 0.46513232588768005, + "learning_rate": 0.00017075025786311612, + "loss": 0.9422, + "step": 5627 + }, + { + "epoch": 1.0019586894586894, + "grad_norm": 0.4530394673347473, + "learning_rate": 0.00017074036500582267, + "loss": 0.8211, + "step": 5628 + }, + { + "epoch": 1.0021367521367521, + "grad_norm": 0.5612013339996338, + "learning_rate": 0.00017073047076251872, + "loss": 0.9466, + "step": 5629 + }, + { + "epoch": 1.0023148148148149, + "grad_norm": 0.4976879954338074, + "learning_rate": 0.00017072057513339812, + "loss": 0.8059, + "step": 5630 + }, + { + "epoch": 1.0024928774928774, + "grad_norm": 0.4842833876609802, + "learning_rate": 0.00017071067811865476, + "loss": 0.6554, + "step": 5631 + }, + { + "epoch": 1.0026709401709402, + "grad_norm": 0.5446373224258423, + "learning_rate": 0.00017070077971848257, + "loss": 1.1001, + "step": 5632 + }, + { + "epoch": 1.002849002849003, + "grad_norm": 0.5996584892272949, + "learning_rate": 0.00017069087993307544, + "loss": 1.0317, + "step": 5633 + }, + { + "epoch": 1.0030270655270654, + "grad_norm": 0.5369443297386169, + "learning_rate": 0.00017068097876262738, + "loss": 0.8019, + "step": 5634 + }, + { + "epoch": 1.0032051282051282, + "grad_norm": 0.4985966682434082, + "learning_rate": 0.00017067107620733236, + "loss": 1.0121, + "step": 5635 + }, + { + "epoch": 1.003383190883191, + "grad_norm": 0.5262824892997742, + "learning_rate": 0.0001706611722673844, + "loss": 1.0157, + "step": 5636 + }, + { + "epoch": 1.0035612535612535, + "grad_norm": 0.5912795066833496, + "learning_rate": 0.00017065126694297756, + "loss": 1.0327, + "step": 5637 + }, + { + "epoch": 1.0037393162393162, + "grad_norm": 0.5866343379020691, + "learning_rate": 0.00017064136023430595, + "loss": 1.1194, + "step": 5638 + }, + { + "epoch": 1.003917378917379, + "grad_norm": 0.5009918808937073, + "learning_rate": 0.0001706314521415636, + "loss": 1.0467, + "step": 5639 + }, + { + "epoch": 1.0040954415954415, + "grad_norm": 0.5455304384231567, + "learning_rate": 0.00017062154266494464, + "loss": 0.8749, + "step": 5640 + }, + { + "epoch": 1.0042735042735043, + "grad_norm": 0.5648258328437805, + "learning_rate": 0.00017061163180464328, + "loss": 0.9408, + "step": 5641 + }, + { + "epoch": 1.004451566951567, + "grad_norm": 0.5276365876197815, + "learning_rate": 0.00017060171956085368, + "loss": 0.9681, + "step": 5642 + }, + { + "epoch": 1.0046296296296295, + "grad_norm": 0.5212745070457458, + "learning_rate": 0.00017059180593377007, + "loss": 0.9188, + "step": 5643 + }, + { + "epoch": 1.0048076923076923, + "grad_norm": 0.540626585483551, + "learning_rate": 0.00017058189092358664, + "loss": 1.0809, + "step": 5644 + }, + { + "epoch": 1.004985754985755, + "grad_norm": 0.5592377781867981, + "learning_rate": 0.00017057197453049767, + "loss": 0.8589, + "step": 5645 + }, + { + "epoch": 1.0051638176638176, + "grad_norm": 0.5115051865577698, + "learning_rate": 0.00017056205675469746, + "loss": 0.8006, + "step": 5646 + }, + { + "epoch": 1.0053418803418803, + "grad_norm": 0.5031117796897888, + "learning_rate": 0.00017055213759638034, + "loss": 0.9242, + "step": 5647 + }, + { + "epoch": 1.005519943019943, + "grad_norm": 0.5342774987220764, + "learning_rate": 0.00017054221705574066, + "loss": 0.8268, + "step": 5648 + }, + { + "epoch": 1.0056980056980056, + "grad_norm": 0.44480493664741516, + "learning_rate": 0.00017053229513297276, + "loss": 0.6892, + "step": 5649 + }, + { + "epoch": 1.0058760683760684, + "grad_norm": 0.5032621622085571, + "learning_rate": 0.00017052237182827105, + "loss": 0.971, + "step": 5650 + }, + { + "epoch": 1.006054131054131, + "grad_norm": 0.5611015558242798, + "learning_rate": 0.00017051244714182996, + "loss": 0.9403, + "step": 5651 + }, + { + "epoch": 1.0062321937321936, + "grad_norm": 0.5064613223075867, + "learning_rate": 0.00017050252107384393, + "loss": 0.9718, + "step": 5652 + }, + { + "epoch": 1.0064102564102564, + "grad_norm": 0.6458395719528198, + "learning_rate": 0.0001704925936245075, + "loss": 1.1161, + "step": 5653 + }, + { + "epoch": 1.0065883190883191, + "grad_norm": 0.527418315410614, + "learning_rate": 0.00017048266479401512, + "loss": 0.9315, + "step": 5654 + }, + { + "epoch": 1.0067663817663817, + "grad_norm": 0.5127941370010376, + "learning_rate": 0.00017047273458256133, + "loss": 0.8206, + "step": 5655 + }, + { + "epoch": 1.0069444444444444, + "grad_norm": 0.6257100105285645, + "learning_rate": 0.00017046280299034067, + "loss": 0.9854, + "step": 5656 + }, + { + "epoch": 1.0071225071225072, + "grad_norm": 0.5081700682640076, + "learning_rate": 0.0001704528700175478, + "loss": 0.9478, + "step": 5657 + }, + { + "epoch": 1.0073005698005697, + "grad_norm": 0.598127543926239, + "learning_rate": 0.00017044293566437725, + "loss": 1.0721, + "step": 5658 + }, + { + "epoch": 1.0074786324786325, + "grad_norm": 0.5429877638816833, + "learning_rate": 0.00017043299993102376, + "loss": 0.9732, + "step": 5659 + }, + { + "epoch": 1.0076566951566952, + "grad_norm": 0.6006619334220886, + "learning_rate": 0.00017042306281768194, + "loss": 1.1262, + "step": 5660 + }, + { + "epoch": 1.0078347578347577, + "grad_norm": 0.48933324217796326, + "learning_rate": 0.00017041312432454646, + "loss": 0.8596, + "step": 5661 + }, + { + "epoch": 1.0080128205128205, + "grad_norm": 0.5902166366577148, + "learning_rate": 0.0001704031844518121, + "loss": 1.1035, + "step": 5662 + }, + { + "epoch": 1.0081908831908832, + "grad_norm": 0.523597776889801, + "learning_rate": 0.0001703932431996736, + "loss": 0.7117, + "step": 5663 + }, + { + "epoch": 1.0083689458689458, + "grad_norm": 0.6313928365707397, + "learning_rate": 0.00017038330056832573, + "loss": 1.0204, + "step": 5664 + }, + { + "epoch": 1.0085470085470085, + "grad_norm": 0.5627471804618835, + "learning_rate": 0.00017037335655796328, + "loss": 0.7648, + "step": 5665 + }, + { + "epoch": 1.0087250712250713, + "grad_norm": 0.5817851424217224, + "learning_rate": 0.0001703634111687811, + "loss": 1.0452, + "step": 5666 + }, + { + "epoch": 1.0089031339031338, + "grad_norm": 0.5143535137176514, + "learning_rate": 0.00017035346440097407, + "loss": 0.9788, + "step": 5667 + }, + { + "epoch": 1.0090811965811965, + "grad_norm": 0.5331187844276428, + "learning_rate": 0.000170343516254737, + "loss": 0.7584, + "step": 5668 + }, + { + "epoch": 1.0092592592592593, + "grad_norm": 0.5723634362220764, + "learning_rate": 0.00017033356673026487, + "loss": 0.9435, + "step": 5669 + }, + { + "epoch": 1.0094373219373218, + "grad_norm": 0.6012297868728638, + "learning_rate": 0.00017032361582775265, + "loss": 1.142, + "step": 5670 + }, + { + "epoch": 1.0096153846153846, + "grad_norm": 0.6161282658576965, + "learning_rate": 0.00017031366354739523, + "loss": 1.2823, + "step": 5671 + }, + { + "epoch": 1.0097934472934473, + "grad_norm": 0.5088054537773132, + "learning_rate": 0.00017030370988938763, + "loss": 0.9743, + "step": 5672 + }, + { + "epoch": 1.0099715099715099, + "grad_norm": 0.512003481388092, + "learning_rate": 0.0001702937548539249, + "loss": 0.9112, + "step": 5673 + }, + { + "epoch": 1.0101495726495726, + "grad_norm": 0.5565149784088135, + "learning_rate": 0.00017028379844120207, + "loss": 1.0074, + "step": 5674 + }, + { + "epoch": 1.0103276353276354, + "grad_norm": 0.6463099718093872, + "learning_rate": 0.00017027384065141418, + "loss": 1.175, + "step": 5675 + }, + { + "epoch": 1.010505698005698, + "grad_norm": 0.46999064087867737, + "learning_rate": 0.00017026388148475637, + "loss": 0.8429, + "step": 5676 + }, + { + "epoch": 1.0106837606837606, + "grad_norm": 0.5617384910583496, + "learning_rate": 0.00017025392094142377, + "loss": 1.045, + "step": 5677 + }, + { + "epoch": 1.0108618233618234, + "grad_norm": 0.5156623721122742, + "learning_rate": 0.00017024395902161154, + "loss": 1.016, + "step": 5678 + }, + { + "epoch": 1.0110398860398861, + "grad_norm": 0.5693390369415283, + "learning_rate": 0.00017023399572551484, + "loss": 0.8616, + "step": 5679 + }, + { + "epoch": 1.0112179487179487, + "grad_norm": 0.5234879851341248, + "learning_rate": 0.00017022403105332892, + "loss": 0.9244, + "step": 5680 + }, + { + "epoch": 1.0113960113960114, + "grad_norm": 0.6513097286224365, + "learning_rate": 0.00017021406500524893, + "loss": 0.9565, + "step": 5681 + }, + { + "epoch": 1.0115740740740742, + "grad_norm": 0.5788878202438354, + "learning_rate": 0.00017020409758147022, + "loss": 0.8994, + "step": 5682 + }, + { + "epoch": 1.0117521367521367, + "grad_norm": 0.5495247840881348, + "learning_rate": 0.00017019412878218807, + "loss": 0.9371, + "step": 5683 + }, + { + "epoch": 1.0119301994301995, + "grad_norm": 0.639045238494873, + "learning_rate": 0.00017018415860759777, + "loss": 1.0297, + "step": 5684 + }, + { + "epoch": 1.0121082621082622, + "grad_norm": 0.5167784690856934, + "learning_rate": 0.0001701741870578947, + "loss": 0.8974, + "step": 5685 + }, + { + "epoch": 1.0122863247863247, + "grad_norm": 0.6131011247634888, + "learning_rate": 0.00017016421413327417, + "loss": 1.13, + "step": 5686 + }, + { + "epoch": 1.0124643874643875, + "grad_norm": 0.4804688096046448, + "learning_rate": 0.00017015423983393166, + "loss": 1.0098, + "step": 5687 + }, + { + "epoch": 1.0126424501424502, + "grad_norm": 0.6605221629142761, + "learning_rate": 0.00017014426416006253, + "loss": 1.1123, + "step": 5688 + }, + { + "epoch": 1.0128205128205128, + "grad_norm": 0.5523666739463806, + "learning_rate": 0.00017013428711186226, + "loss": 0.8226, + "step": 5689 + }, + { + "epoch": 1.0129985754985755, + "grad_norm": 0.6012941598892212, + "learning_rate": 0.00017012430868952632, + "loss": 0.8915, + "step": 5690 + }, + { + "epoch": 1.0131766381766383, + "grad_norm": 0.5830875039100647, + "learning_rate": 0.00017011432889325022, + "loss": 1.021, + "step": 5691 + }, + { + "epoch": 1.0133547008547008, + "grad_norm": 0.5546056032180786, + "learning_rate": 0.0001701043477232295, + "loss": 0.7656, + "step": 5692 + }, + { + "epoch": 1.0135327635327636, + "grad_norm": 0.5592601299285889, + "learning_rate": 0.0001700943651796597, + "loss": 1.0172, + "step": 5693 + }, + { + "epoch": 1.0137108262108263, + "grad_norm": 0.5708866715431213, + "learning_rate": 0.00017008438126273645, + "loss": 1.0012, + "step": 5694 + }, + { + "epoch": 1.0138888888888888, + "grad_norm": 0.6856338381767273, + "learning_rate": 0.0001700743959726553, + "loss": 1.1278, + "step": 5695 + }, + { + "epoch": 1.0140669515669516, + "grad_norm": 0.6523802876472473, + "learning_rate": 0.000170064409309612, + "loss": 1.0406, + "step": 5696 + }, + { + "epoch": 1.0142450142450143, + "grad_norm": 0.6653079986572266, + "learning_rate": 0.00017005442127380208, + "loss": 1.1086, + "step": 5697 + }, + { + "epoch": 1.0144230769230769, + "grad_norm": 0.5841104388237, + "learning_rate": 0.00017004443186542133, + "loss": 0.9335, + "step": 5698 + }, + { + "epoch": 1.0146011396011396, + "grad_norm": 0.5696784257888794, + "learning_rate": 0.0001700344410846654, + "loss": 1.0247, + "step": 5699 + }, + { + "epoch": 1.0147792022792024, + "grad_norm": 0.7135653495788574, + "learning_rate": 0.00017002444893173013, + "loss": 1.0259, + "step": 5700 + }, + { + "epoch": 1.014957264957265, + "grad_norm": 0.5806999802589417, + "learning_rate": 0.00017001445540681124, + "loss": 1.0053, + "step": 5701 + }, + { + "epoch": 1.0151353276353277, + "grad_norm": 0.5298715829849243, + "learning_rate": 0.0001700044605101045, + "loss": 0.9415, + "step": 5702 + }, + { + "epoch": 1.0153133903133904, + "grad_norm": 0.5817379951477051, + "learning_rate": 0.0001699944642418058, + "loss": 1.0906, + "step": 5703 + }, + { + "epoch": 1.015491452991453, + "grad_norm": 0.6564923524856567, + "learning_rate": 0.00016998446660211098, + "loss": 0.9933, + "step": 5704 + }, + { + "epoch": 1.0156695156695157, + "grad_norm": 0.6547308564186096, + "learning_rate": 0.00016997446759121592, + "loss": 1.0045, + "step": 5705 + }, + { + "epoch": 1.0158475783475784, + "grad_norm": 0.5763013958930969, + "learning_rate": 0.00016996446720931652, + "loss": 1.0898, + "step": 5706 + }, + { + "epoch": 1.016025641025641, + "grad_norm": 0.6118074059486389, + "learning_rate": 0.00016995446545660871, + "loss": 0.9398, + "step": 5707 + }, + { + "epoch": 1.0162037037037037, + "grad_norm": 0.6810526251792908, + "learning_rate": 0.0001699444623332885, + "loss": 1.0968, + "step": 5708 + }, + { + "epoch": 1.0163817663817665, + "grad_norm": 0.5292752981185913, + "learning_rate": 0.00016993445783955184, + "loss": 0.7549, + "step": 5709 + }, + { + "epoch": 1.016559829059829, + "grad_norm": 0.6014277935028076, + "learning_rate": 0.00016992445197559474, + "loss": 1.1711, + "step": 5710 + }, + { + "epoch": 1.0167378917378918, + "grad_norm": 0.5089772343635559, + "learning_rate": 0.00016991444474161326, + "loss": 0.9188, + "step": 5711 + }, + { + "epoch": 1.0169159544159545, + "grad_norm": 0.567193865776062, + "learning_rate": 0.0001699044361378035, + "loss": 0.7462, + "step": 5712 + }, + { + "epoch": 1.017094017094017, + "grad_norm": 0.5638598799705505, + "learning_rate": 0.00016989442616436147, + "loss": 0.9643, + "step": 5713 + }, + { + "epoch": 1.0172720797720798, + "grad_norm": 0.5634039640426636, + "learning_rate": 0.0001698844148214834, + "loss": 1.0141, + "step": 5714 + }, + { + "epoch": 1.0174501424501425, + "grad_norm": 0.5326652526855469, + "learning_rate": 0.00016987440210936537, + "loss": 0.865, + "step": 5715 + }, + { + "epoch": 1.017628205128205, + "grad_norm": 0.5858046412467957, + "learning_rate": 0.0001698643880282036, + "loss": 0.9561, + "step": 5716 + }, + { + "epoch": 1.0178062678062678, + "grad_norm": 0.6424698829650879, + "learning_rate": 0.00016985437257819428, + "loss": 1.0169, + "step": 5717 + }, + { + "epoch": 1.0179843304843306, + "grad_norm": 0.6294280290603638, + "learning_rate": 0.00016984435575953364, + "loss": 1.0438, + "step": 5718 + }, + { + "epoch": 1.018162393162393, + "grad_norm": 0.5533088445663452, + "learning_rate": 0.00016983433757241788, + "loss": 0.8901, + "step": 5719 + }, + { + "epoch": 1.0183404558404558, + "grad_norm": 0.5148718953132629, + "learning_rate": 0.00016982431801704342, + "loss": 0.9201, + "step": 5720 + }, + { + "epoch": 1.0185185185185186, + "grad_norm": 0.5609371662139893, + "learning_rate": 0.00016981429709360645, + "loss": 0.9347, + "step": 5721 + }, + { + "epoch": 1.0186965811965811, + "grad_norm": 0.5502731204032898, + "learning_rate": 0.00016980427480230338, + "loss": 1.0508, + "step": 5722 + }, + { + "epoch": 1.0188746438746439, + "grad_norm": 0.5880394577980042, + "learning_rate": 0.00016979425114333055, + "loss": 1.1258, + "step": 5723 + }, + { + "epoch": 1.0190527065527066, + "grad_norm": 0.5569866895675659, + "learning_rate": 0.0001697842261168843, + "loss": 0.9186, + "step": 5724 + }, + { + "epoch": 1.0192307692307692, + "grad_norm": 0.7468093037605286, + "learning_rate": 0.00016977419972316116, + "loss": 1.2066, + "step": 5725 + }, + { + "epoch": 1.019408831908832, + "grad_norm": 0.6041515469551086, + "learning_rate": 0.00016976417196235753, + "loss": 0.939, + "step": 5726 + }, + { + "epoch": 1.0195868945868947, + "grad_norm": 0.6102641224861145, + "learning_rate": 0.00016975414283466983, + "loss": 0.8334, + "step": 5727 + }, + { + "epoch": 1.0197649572649572, + "grad_norm": 0.5418640375137329, + "learning_rate": 0.00016974411234029467, + "loss": 0.8072, + "step": 5728 + }, + { + "epoch": 1.01994301994302, + "grad_norm": 0.6569705605506897, + "learning_rate": 0.00016973408047942843, + "loss": 1.103, + "step": 5729 + }, + { + "epoch": 1.0201210826210827, + "grad_norm": 0.5778102278709412, + "learning_rate": 0.00016972404725226778, + "loss": 0.9353, + "step": 5730 + }, + { + "epoch": 1.0202991452991452, + "grad_norm": 0.5474382638931274, + "learning_rate": 0.0001697140126590093, + "loss": 1.0009, + "step": 5731 + }, + { + "epoch": 1.020477207977208, + "grad_norm": 0.5869506597518921, + "learning_rate": 0.00016970397669984947, + "loss": 1.0027, + "step": 5732 + }, + { + "epoch": 1.0206552706552707, + "grad_norm": 0.5078117251396179, + "learning_rate": 0.00016969393937498508, + "loss": 0.8316, + "step": 5733 + }, + { + "epoch": 1.0208333333333333, + "grad_norm": 0.5488452911376953, + "learning_rate": 0.0001696839006846127, + "loss": 0.8438, + "step": 5734 + }, + { + "epoch": 1.021011396011396, + "grad_norm": 0.5921052098274231, + "learning_rate": 0.00016967386062892908, + "loss": 0.9147, + "step": 5735 + }, + { + "epoch": 1.0211894586894588, + "grad_norm": 0.5486881136894226, + "learning_rate": 0.00016966381920813085, + "loss": 0.7619, + "step": 5736 + }, + { + "epoch": 1.0213675213675213, + "grad_norm": 0.5250689387321472, + "learning_rate": 0.00016965377642241483, + "loss": 0.9192, + "step": 5737 + }, + { + "epoch": 1.021545584045584, + "grad_norm": 0.5355087518692017, + "learning_rate": 0.00016964373227197773, + "loss": 0.954, + "step": 5738 + }, + { + "epoch": 1.0217236467236468, + "grad_norm": 0.6758780479431152, + "learning_rate": 0.0001696336867570164, + "loss": 1.1257, + "step": 5739 + }, + { + "epoch": 1.0219017094017093, + "grad_norm": 0.6361044049263, + "learning_rate": 0.00016962363987772756, + "loss": 1.0889, + "step": 5740 + }, + { + "epoch": 1.022079772079772, + "grad_norm": 0.5802326798439026, + "learning_rate": 0.00016961359163430819, + "loss": 0.8966, + "step": 5741 + }, + { + "epoch": 1.0222578347578348, + "grad_norm": 0.5535712242126465, + "learning_rate": 0.00016960354202695508, + "loss": 1.0007, + "step": 5742 + }, + { + "epoch": 1.0224358974358974, + "grad_norm": 0.5469220280647278, + "learning_rate": 0.00016959349105586516, + "loss": 0.8202, + "step": 5743 + }, + { + "epoch": 1.02261396011396, + "grad_norm": 0.5533008575439453, + "learning_rate": 0.00016958343872123534, + "loss": 0.9576, + "step": 5744 + }, + { + "epoch": 1.0227920227920229, + "grad_norm": 0.615132749080658, + "learning_rate": 0.00016957338502326258, + "loss": 0.8719, + "step": 5745 + }, + { + "epoch": 1.0229700854700854, + "grad_norm": 0.519075334072113, + "learning_rate": 0.0001695633299621439, + "loss": 0.8309, + "step": 5746 + }, + { + "epoch": 1.0231481481481481, + "grad_norm": 0.6249759197235107, + "learning_rate": 0.00016955327353807624, + "loss": 1.151, + "step": 5747 + }, + { + "epoch": 1.023326210826211, + "grad_norm": 0.560299277305603, + "learning_rate": 0.00016954321575125668, + "loss": 0.7889, + "step": 5748 + }, + { + "epoch": 1.0235042735042734, + "grad_norm": 0.5735262036323547, + "learning_rate": 0.0001695331566018823, + "loss": 0.8794, + "step": 5749 + }, + { + "epoch": 1.0236823361823362, + "grad_norm": 0.5893994569778442, + "learning_rate": 0.00016952309609015012, + "loss": 0.9696, + "step": 5750 + }, + { + "epoch": 1.023860398860399, + "grad_norm": 0.6064512133598328, + "learning_rate": 0.0001695130342162573, + "loss": 0.9771, + "step": 5751 + }, + { + "epoch": 1.0240384615384615, + "grad_norm": 0.5833427309989929, + "learning_rate": 0.00016950297098040099, + "loss": 1.1768, + "step": 5752 + }, + { + "epoch": 1.0242165242165242, + "grad_norm": 0.5940282344818115, + "learning_rate": 0.00016949290638277833, + "loss": 1.0758, + "step": 5753 + }, + { + "epoch": 1.024394586894587, + "grad_norm": 0.5267124772071838, + "learning_rate": 0.00016948284042358656, + "loss": 0.772, + "step": 5754 + }, + { + "epoch": 1.0245726495726495, + "grad_norm": 0.6217982172966003, + "learning_rate": 0.00016947277310302284, + "loss": 0.8583, + "step": 5755 + }, + { + "epoch": 1.0247507122507122, + "grad_norm": 0.6192215085029602, + "learning_rate": 0.00016946270442128443, + "loss": 0.9148, + "step": 5756 + }, + { + "epoch": 1.024928774928775, + "grad_norm": 0.5337123870849609, + "learning_rate": 0.00016945263437856867, + "loss": 1.0054, + "step": 5757 + }, + { + "epoch": 1.0251068376068375, + "grad_norm": 0.5462040901184082, + "learning_rate": 0.00016944256297507276, + "loss": 1.1097, + "step": 5758 + }, + { + "epoch": 1.0252849002849003, + "grad_norm": 0.5606170892715454, + "learning_rate": 0.00016943249021099415, + "loss": 1.0192, + "step": 5759 + }, + { + "epoch": 1.025462962962963, + "grad_norm": 0.636974573135376, + "learning_rate": 0.00016942241608653008, + "loss": 1.0241, + "step": 5760 + }, + { + "epoch": 1.0256410256410255, + "grad_norm": 0.4895164966583252, + "learning_rate": 0.00016941234060187797, + "loss": 0.9057, + "step": 5761 + }, + { + "epoch": 1.0258190883190883, + "grad_norm": 0.5810303092002869, + "learning_rate": 0.00016940226375723527, + "loss": 1.0809, + "step": 5762 + }, + { + "epoch": 1.025997150997151, + "grad_norm": 0.6043853163719177, + "learning_rate": 0.00016939218555279937, + "loss": 1.0685, + "step": 5763 + }, + { + "epoch": 1.0261752136752136, + "grad_norm": 0.5827188491821289, + "learning_rate": 0.00016938210598876774, + "loss": 1.0236, + "step": 5764 + }, + { + "epoch": 1.0263532763532763, + "grad_norm": 0.6677887439727783, + "learning_rate": 0.0001693720250653379, + "loss": 1.0586, + "step": 5765 + }, + { + "epoch": 1.026531339031339, + "grad_norm": 0.558051347732544, + "learning_rate": 0.0001693619427827073, + "loss": 0.745, + "step": 5766 + }, + { + "epoch": 1.0267094017094016, + "grad_norm": 0.6336706280708313, + "learning_rate": 0.0001693518591410735, + "loss": 1.0658, + "step": 5767 + }, + { + "epoch": 1.0268874643874644, + "grad_norm": 0.7077126502990723, + "learning_rate": 0.00016934177414063416, + "loss": 1.18, + "step": 5768 + }, + { + "epoch": 1.0270655270655271, + "grad_norm": 0.5342326760292053, + "learning_rate": 0.00016933168778158675, + "loss": 0.8347, + "step": 5769 + }, + { + "epoch": 1.0272435897435896, + "grad_norm": 0.6116416454315186, + "learning_rate": 0.00016932160006412895, + "loss": 1.0648, + "step": 5770 + }, + { + "epoch": 1.0274216524216524, + "grad_norm": 0.5411320924758911, + "learning_rate": 0.0001693115109884584, + "loss": 1.0756, + "step": 5771 + }, + { + "epoch": 1.0275997150997151, + "grad_norm": 0.5549847483634949, + "learning_rate": 0.00016930142055477277, + "loss": 0.7259, + "step": 5772 + }, + { + "epoch": 1.0277777777777777, + "grad_norm": 0.549010694026947, + "learning_rate": 0.00016929132876326977, + "loss": 0.9488, + "step": 5773 + }, + { + "epoch": 1.0279558404558404, + "grad_norm": 0.6302017569541931, + "learning_rate": 0.00016928123561414714, + "loss": 0.8851, + "step": 5774 + }, + { + "epoch": 1.0281339031339032, + "grad_norm": 0.5831273198127747, + "learning_rate": 0.00016927114110760257, + "loss": 0.7841, + "step": 5775 + }, + { + "epoch": 1.0283119658119657, + "grad_norm": 0.5528474450111389, + "learning_rate": 0.00016926104524383394, + "loss": 1.0108, + "step": 5776 + }, + { + "epoch": 1.0284900284900285, + "grad_norm": 0.6279126405715942, + "learning_rate": 0.00016925094802303897, + "loss": 0.8632, + "step": 5777 + }, + { + "epoch": 1.0286680911680912, + "grad_norm": 0.6783218383789062, + "learning_rate": 0.00016924084944541554, + "loss": 1.0746, + "step": 5778 + }, + { + "epoch": 1.0288461538461537, + "grad_norm": 0.5823925137519836, + "learning_rate": 0.00016923074951116153, + "loss": 1.0486, + "step": 5779 + }, + { + "epoch": 1.0290242165242165, + "grad_norm": 0.6095981597900391, + "learning_rate": 0.00016922064822047473, + "loss": 0.8113, + "step": 5780 + }, + { + "epoch": 1.0292022792022792, + "grad_norm": 0.7887664437294006, + "learning_rate": 0.00016921054557355317, + "loss": 1.2411, + "step": 5781 + }, + { + "epoch": 1.0293803418803418, + "grad_norm": 0.6511263251304626, + "learning_rate": 0.00016920044157059475, + "loss": 0.924, + "step": 5782 + }, + { + "epoch": 1.0295584045584045, + "grad_norm": 0.6045661568641663, + "learning_rate": 0.00016919033621179744, + "loss": 0.8373, + "step": 5783 + }, + { + "epoch": 1.0297364672364673, + "grad_norm": 0.6914188861846924, + "learning_rate": 0.0001691802294973592, + "loss": 0.9589, + "step": 5784 + }, + { + "epoch": 1.0299145299145298, + "grad_norm": 0.6483730673789978, + "learning_rate": 0.00016917012142747805, + "loss": 0.9871, + "step": 5785 + }, + { + "epoch": 1.0300925925925926, + "grad_norm": 0.5775033235549927, + "learning_rate": 0.0001691600120023521, + "loss": 1.0591, + "step": 5786 + }, + { + "epoch": 1.0302706552706553, + "grad_norm": 0.6206814646720886, + "learning_rate": 0.00016914990122217932, + "loss": 0.9126, + "step": 5787 + }, + { + "epoch": 1.0304487179487178, + "grad_norm": 0.5422028303146362, + "learning_rate": 0.00016913978908715796, + "loss": 0.8227, + "step": 5788 + }, + { + "epoch": 1.0306267806267806, + "grad_norm": 0.5824416875839233, + "learning_rate": 0.000169129675597486, + "loss": 1.111, + "step": 5789 + }, + { + "epoch": 1.0308048433048433, + "grad_norm": 0.5419015884399414, + "learning_rate": 0.00016911956075336165, + "loss": 0.8941, + "step": 5790 + }, + { + "epoch": 1.0309829059829059, + "grad_norm": 0.6171557903289795, + "learning_rate": 0.0001691094445549831, + "loss": 0.8679, + "step": 5791 + }, + { + "epoch": 1.0311609686609686, + "grad_norm": 0.6136980056762695, + "learning_rate": 0.00016909932700254855, + "loss": 0.9266, + "step": 5792 + }, + { + "epoch": 1.0313390313390314, + "grad_norm": 0.6275020241737366, + "learning_rate": 0.00016908920809625624, + "loss": 1.0828, + "step": 5793 + }, + { + "epoch": 1.0315170940170941, + "grad_norm": 0.6538251638412476, + "learning_rate": 0.0001690790878363044, + "loss": 0.8413, + "step": 5794 + }, + { + "epoch": 1.0316951566951567, + "grad_norm": 0.5981295108795166, + "learning_rate": 0.00016906896622289136, + "loss": 0.9845, + "step": 5795 + }, + { + "epoch": 1.0318732193732194, + "grad_norm": 0.5390967130661011, + "learning_rate": 0.00016905884325621538, + "loss": 0.8755, + "step": 5796 + }, + { + "epoch": 1.032051282051282, + "grad_norm": 0.5534448623657227, + "learning_rate": 0.00016904871893647482, + "loss": 1.1868, + "step": 5797 + }, + { + "epoch": 1.0322293447293447, + "grad_norm": 0.664556086063385, + "learning_rate": 0.00016903859326386806, + "loss": 1.1418, + "step": 5798 + }, + { + "epoch": 1.0324074074074074, + "grad_norm": 0.5737143158912659, + "learning_rate": 0.00016902846623859346, + "loss": 1.124, + "step": 5799 + }, + { + "epoch": 1.0325854700854702, + "grad_norm": 0.6499935388565063, + "learning_rate": 0.0001690183378608495, + "loss": 1.0331, + "step": 5800 + }, + { + "epoch": 1.0327635327635327, + "grad_norm": 0.5721518993377686, + "learning_rate": 0.00016900820813083454, + "loss": 0.8664, + "step": 5801 + }, + { + "epoch": 1.0329415954415955, + "grad_norm": 0.5651140809059143, + "learning_rate": 0.0001689980770487471, + "loss": 1.1661, + "step": 5802 + }, + { + "epoch": 1.0331196581196582, + "grad_norm": 0.5935871005058289, + "learning_rate": 0.0001689879446147857, + "loss": 0.8722, + "step": 5803 + }, + { + "epoch": 1.0332977207977208, + "grad_norm": 0.5627842545509338, + "learning_rate": 0.00016897781082914884, + "loss": 1.0036, + "step": 5804 + }, + { + "epoch": 1.0334757834757835, + "grad_norm": 0.5866895914077759, + "learning_rate": 0.00016896767569203502, + "loss": 0.9739, + "step": 5805 + }, + { + "epoch": 1.0336538461538463, + "grad_norm": 0.5568059682846069, + "learning_rate": 0.0001689575392036429, + "loss": 0.7081, + "step": 5806 + }, + { + "epoch": 1.0338319088319088, + "grad_norm": 0.6054235100746155, + "learning_rate": 0.00016894740136417103, + "loss": 1.1168, + "step": 5807 + }, + { + "epoch": 1.0340099715099715, + "grad_norm": 0.5215454697608948, + "learning_rate": 0.00016893726217381805, + "loss": 0.9172, + "step": 5808 + }, + { + "epoch": 1.0341880341880343, + "grad_norm": 0.5415732860565186, + "learning_rate": 0.00016892712163278263, + "loss": 0.7812, + "step": 5809 + }, + { + "epoch": 1.0343660968660968, + "grad_norm": 0.6341692805290222, + "learning_rate": 0.00016891697974126345, + "loss": 1.0658, + "step": 5810 + }, + { + "epoch": 1.0345441595441596, + "grad_norm": 0.6326245665550232, + "learning_rate": 0.00016890683649945922, + "loss": 1.0134, + "step": 5811 + }, + { + "epoch": 1.0347222222222223, + "grad_norm": 0.5729571580886841, + "learning_rate": 0.00016889669190756868, + "loss": 0.9139, + "step": 5812 + }, + { + "epoch": 1.0349002849002849, + "grad_norm": 0.5912853479385376, + "learning_rate": 0.00016888654596579054, + "loss": 1.122, + "step": 5813 + }, + { + "epoch": 1.0350783475783476, + "grad_norm": 0.8410450220108032, + "learning_rate": 0.00016887639867432368, + "loss": 1.3009, + "step": 5814 + }, + { + "epoch": 1.0352564102564104, + "grad_norm": 0.5416620969772339, + "learning_rate": 0.00016886625003336683, + "loss": 0.8751, + "step": 5815 + }, + { + "epoch": 1.0354344729344729, + "grad_norm": 0.6367851495742798, + "learning_rate": 0.0001688561000431189, + "loss": 0.956, + "step": 5816 + }, + { + "epoch": 1.0356125356125356, + "grad_norm": 0.4618827700614929, + "learning_rate": 0.0001688459487037787, + "loss": 0.5313, + "step": 5817 + }, + { + "epoch": 1.0357905982905984, + "grad_norm": 0.7139244079589844, + "learning_rate": 0.00016883579601554516, + "loss": 1.0787, + "step": 5818 + }, + { + "epoch": 1.035968660968661, + "grad_norm": 0.6896135210990906, + "learning_rate": 0.00016882564197861715, + "loss": 0.932, + "step": 5819 + }, + { + "epoch": 1.0361467236467237, + "grad_norm": 0.5889739394187927, + "learning_rate": 0.00016881548659319372, + "loss": 0.8852, + "step": 5820 + }, + { + "epoch": 1.0363247863247864, + "grad_norm": 0.5954701900482178, + "learning_rate": 0.00016880532985947375, + "loss": 0.8192, + "step": 5821 + }, + { + "epoch": 1.036502849002849, + "grad_norm": 0.6665091514587402, + "learning_rate": 0.00016879517177765627, + "loss": 0.9578, + "step": 5822 + }, + { + "epoch": 1.0366809116809117, + "grad_norm": 0.5990539789199829, + "learning_rate": 0.00016878501234794034, + "loss": 0.9797, + "step": 5823 + }, + { + "epoch": 1.0368589743589745, + "grad_norm": 0.596755862236023, + "learning_rate": 0.00016877485157052496, + "loss": 1.173, + "step": 5824 + }, + { + "epoch": 1.037037037037037, + "grad_norm": 0.544658362865448, + "learning_rate": 0.00016876468944560923, + "loss": 1.0742, + "step": 5825 + }, + { + "epoch": 1.0372150997150997, + "grad_norm": 0.5841910243034363, + "learning_rate": 0.00016875452597339225, + "loss": 1.029, + "step": 5826 + }, + { + "epoch": 1.0373931623931625, + "grad_norm": 0.6508592963218689, + "learning_rate": 0.00016874436115407317, + "loss": 0.9883, + "step": 5827 + }, + { + "epoch": 1.037571225071225, + "grad_norm": 0.590050458908081, + "learning_rate": 0.00016873419498785114, + "loss": 1.0713, + "step": 5828 + }, + { + "epoch": 1.0377492877492878, + "grad_norm": 0.5386307239532471, + "learning_rate": 0.00016872402747492534, + "loss": 1.0159, + "step": 5829 + }, + { + "epoch": 1.0379273504273505, + "grad_norm": 0.6173896193504333, + "learning_rate": 0.00016871385861549497, + "loss": 1.0056, + "step": 5830 + }, + { + "epoch": 1.038105413105413, + "grad_norm": 0.5377787351608276, + "learning_rate": 0.0001687036884097593, + "loss": 0.8708, + "step": 5831 + }, + { + "epoch": 1.0382834757834758, + "grad_norm": 0.5753569006919861, + "learning_rate": 0.00016869351685791756, + "loss": 1.0529, + "step": 5832 + }, + { + "epoch": 1.0384615384615385, + "grad_norm": 0.6085895299911499, + "learning_rate": 0.00016868334396016906, + "loss": 1.1017, + "step": 5833 + }, + { + "epoch": 1.038639601139601, + "grad_norm": 0.6320509910583496, + "learning_rate": 0.0001686731697167131, + "loss": 1.0543, + "step": 5834 + }, + { + "epoch": 1.0388176638176638, + "grad_norm": 0.5691760778427124, + "learning_rate": 0.00016866299412774907, + "loss": 0.9975, + "step": 5835 + }, + { + "epoch": 1.0389957264957266, + "grad_norm": 0.5990765690803528, + "learning_rate": 0.0001686528171934763, + "loss": 0.8776, + "step": 5836 + }, + { + "epoch": 1.039173789173789, + "grad_norm": 0.6650477647781372, + "learning_rate": 0.00016864263891409415, + "loss": 1.0652, + "step": 5837 + }, + { + "epoch": 1.0393518518518519, + "grad_norm": 0.6050353646278381, + "learning_rate": 0.00016863245928980212, + "loss": 0.9313, + "step": 5838 + }, + { + "epoch": 1.0395299145299146, + "grad_norm": 0.587505578994751, + "learning_rate": 0.0001686222783207996, + "loss": 0.9892, + "step": 5839 + }, + { + "epoch": 1.0397079772079771, + "grad_norm": 0.6310170292854309, + "learning_rate": 0.00016861209600728608, + "loss": 1.1045, + "step": 5840 + }, + { + "epoch": 1.03988603988604, + "grad_norm": 0.5683430433273315, + "learning_rate": 0.0001686019123494611, + "loss": 1.0507, + "step": 5841 + }, + { + "epoch": 1.0400641025641026, + "grad_norm": 0.6621488332748413, + "learning_rate": 0.00016859172734752414, + "loss": 0.9255, + "step": 5842 + }, + { + "epoch": 1.0402421652421652, + "grad_norm": 0.6197706460952759, + "learning_rate": 0.00016858154100167475, + "loss": 1.0031, + "step": 5843 + }, + { + "epoch": 1.040420227920228, + "grad_norm": 0.6805898547172546, + "learning_rate": 0.00016857135331211257, + "loss": 0.9901, + "step": 5844 + }, + { + "epoch": 1.0405982905982907, + "grad_norm": 0.5512405633926392, + "learning_rate": 0.00016856116427903714, + "loss": 1.0033, + "step": 5845 + }, + { + "epoch": 1.0407763532763532, + "grad_norm": 0.5643384456634521, + "learning_rate": 0.00016855097390264815, + "loss": 0.9136, + "step": 5846 + }, + { + "epoch": 1.040954415954416, + "grad_norm": 0.48351922631263733, + "learning_rate": 0.0001685407821831452, + "loss": 0.6163, + "step": 5847 + }, + { + "epoch": 1.0411324786324787, + "grad_norm": 0.6256039142608643, + "learning_rate": 0.00016853058912072802, + "loss": 0.9409, + "step": 5848 + }, + { + "epoch": 1.0413105413105412, + "grad_norm": 0.6539996862411499, + "learning_rate": 0.00016852039471559627, + "loss": 0.9367, + "step": 5849 + }, + { + "epoch": 1.041488603988604, + "grad_norm": 0.6192609667778015, + "learning_rate": 0.00016851019896794975, + "loss": 0.9631, + "step": 5850 + }, + { + "epoch": 1.0416666666666667, + "grad_norm": 0.613563060760498, + "learning_rate": 0.0001685000018779882, + "loss": 0.9132, + "step": 5851 + }, + { + "epoch": 1.0418447293447293, + "grad_norm": 0.6004200577735901, + "learning_rate": 0.0001684898034459114, + "loss": 1.1313, + "step": 5852 + }, + { + "epoch": 1.042022792022792, + "grad_norm": 0.6158567070960999, + "learning_rate": 0.0001684796036719192, + "loss": 1.0253, + "step": 5853 + }, + { + "epoch": 1.0422008547008548, + "grad_norm": 0.6362335085868835, + "learning_rate": 0.00016846940255621143, + "loss": 0.93, + "step": 5854 + }, + { + "epoch": 1.0423789173789173, + "grad_norm": 0.6148427128791809, + "learning_rate": 0.00016845920009898787, + "loss": 0.9122, + "step": 5855 + }, + { + "epoch": 1.04255698005698, + "grad_norm": 0.5119984149932861, + "learning_rate": 0.00016844899630044858, + "loss": 0.7954, + "step": 5856 + }, + { + "epoch": 1.0427350427350428, + "grad_norm": 0.571849524974823, + "learning_rate": 0.00016843879116079338, + "loss": 0.8588, + "step": 5857 + }, + { + "epoch": 1.0429131054131053, + "grad_norm": 0.6173384785652161, + "learning_rate": 0.00016842858468022221, + "loss": 1.0475, + "step": 5858 + }, + { + "epoch": 1.043091168091168, + "grad_norm": 0.566114068031311, + "learning_rate": 0.0001684183768589351, + "loss": 0.8485, + "step": 5859 + }, + { + "epoch": 1.0432692307692308, + "grad_norm": 0.653134286403656, + "learning_rate": 0.000168408167697132, + "loss": 0.9976, + "step": 5860 + }, + { + "epoch": 1.0434472934472934, + "grad_norm": 0.63815838098526, + "learning_rate": 0.00016839795719501296, + "loss": 0.7091, + "step": 5861 + }, + { + "epoch": 1.0436253561253561, + "grad_norm": 0.5109001994132996, + "learning_rate": 0.00016838774535277805, + "loss": 0.7668, + "step": 5862 + }, + { + "epoch": 1.0438034188034189, + "grad_norm": 0.6741907596588135, + "learning_rate": 0.0001683775321706273, + "loss": 1.0493, + "step": 5863 + }, + { + "epoch": 1.0439814814814814, + "grad_norm": 0.6006115674972534, + "learning_rate": 0.0001683673176487609, + "loss": 0.9784, + "step": 5864 + }, + { + "epoch": 1.0441595441595442, + "grad_norm": 0.5504778027534485, + "learning_rate": 0.0001683571017873789, + "loss": 0.9718, + "step": 5865 + }, + { + "epoch": 1.044337606837607, + "grad_norm": 0.5713102221488953, + "learning_rate": 0.00016834688458668148, + "loss": 1.12, + "step": 5866 + }, + { + "epoch": 1.0445156695156694, + "grad_norm": 0.7878454923629761, + "learning_rate": 0.00016833666604686886, + "loss": 1.1803, + "step": 5867 + }, + { + "epoch": 1.0446937321937322, + "grad_norm": 0.582697331905365, + "learning_rate": 0.00016832644616814122, + "loss": 0.943, + "step": 5868 + }, + { + "epoch": 1.044871794871795, + "grad_norm": 0.5300645232200623, + "learning_rate": 0.00016831622495069878, + "loss": 0.9087, + "step": 5869 + }, + { + "epoch": 1.0450498575498575, + "grad_norm": 0.5627666115760803, + "learning_rate": 0.00016830600239474186, + "loss": 1.081, + "step": 5870 + }, + { + "epoch": 1.0452279202279202, + "grad_norm": 0.6760496497154236, + "learning_rate": 0.0001682957785004707, + "loss": 1.1098, + "step": 5871 + }, + { + "epoch": 1.045405982905983, + "grad_norm": 0.6424084901809692, + "learning_rate": 0.00016828555326808565, + "loss": 0.9657, + "step": 5872 + }, + { + "epoch": 1.0455840455840455, + "grad_norm": 0.5523313283920288, + "learning_rate": 0.000168275326697787, + "loss": 1.0163, + "step": 5873 + }, + { + "epoch": 1.0457621082621082, + "grad_norm": 0.5582337975502014, + "learning_rate": 0.00016826509878977518, + "loss": 0.8825, + "step": 5874 + }, + { + "epoch": 1.045940170940171, + "grad_norm": 0.5603214502334595, + "learning_rate": 0.00016825486954425055, + "loss": 0.9032, + "step": 5875 + }, + { + "epoch": 1.0461182336182335, + "grad_norm": 0.5944222807884216, + "learning_rate": 0.00016824463896141355, + "loss": 0.9384, + "step": 5876 + }, + { + "epoch": 1.0462962962962963, + "grad_norm": 0.6220229268074036, + "learning_rate": 0.00016823440704146457, + "loss": 0.8962, + "step": 5877 + }, + { + "epoch": 1.046474358974359, + "grad_norm": 0.5607972145080566, + "learning_rate": 0.0001682241737846042, + "loss": 0.9385, + "step": 5878 + }, + { + "epoch": 1.0466524216524216, + "grad_norm": 0.6206870079040527, + "learning_rate": 0.00016821393919103282, + "loss": 1.0597, + "step": 5879 + }, + { + "epoch": 1.0468304843304843, + "grad_norm": 0.5126399993896484, + "learning_rate": 0.000168203703260951, + "loss": 0.9403, + "step": 5880 + }, + { + "epoch": 1.047008547008547, + "grad_norm": 0.6569282412528992, + "learning_rate": 0.00016819346599455929, + "loss": 0.8124, + "step": 5881 + }, + { + "epoch": 1.0471866096866096, + "grad_norm": 0.6670137047767639, + "learning_rate": 0.0001681832273920583, + "loss": 1.1927, + "step": 5882 + }, + { + "epoch": 1.0473646723646723, + "grad_norm": 0.5403243899345398, + "learning_rate": 0.00016817298745364862, + "loss": 0.8539, + "step": 5883 + }, + { + "epoch": 1.047542735042735, + "grad_norm": 0.5500505566596985, + "learning_rate": 0.00016816274617953086, + "loss": 1.1064, + "step": 5884 + }, + { + "epoch": 1.0477207977207976, + "grad_norm": 0.5482703447341919, + "learning_rate": 0.00016815250356990566, + "loss": 0.7276, + "step": 5885 + }, + { + "epoch": 1.0478988603988604, + "grad_norm": 0.6290771961212158, + "learning_rate": 0.00016814225962497373, + "loss": 0.9018, + "step": 5886 + }, + { + "epoch": 1.0480769230769231, + "grad_norm": 0.6404094696044922, + "learning_rate": 0.00016813201434493578, + "loss": 1.0638, + "step": 5887 + }, + { + "epoch": 1.0482549857549857, + "grad_norm": 0.5484994053840637, + "learning_rate": 0.0001681217677299926, + "loss": 1.0033, + "step": 5888 + }, + { + "epoch": 1.0484330484330484, + "grad_norm": 0.6474852561950684, + "learning_rate": 0.0001681115197803448, + "loss": 1.1017, + "step": 5889 + }, + { + "epoch": 1.0486111111111112, + "grad_norm": 0.6186243295669556, + "learning_rate": 0.0001681012704961933, + "loss": 0.9978, + "step": 5890 + }, + { + "epoch": 1.0487891737891737, + "grad_norm": 0.6244034767150879, + "learning_rate": 0.00016809101987773887, + "loss": 0.9906, + "step": 5891 + }, + { + "epoch": 1.0489672364672364, + "grad_norm": 0.5893426537513733, + "learning_rate": 0.00016808076792518235, + "loss": 0.9345, + "step": 5892 + }, + { + "epoch": 1.0491452991452992, + "grad_norm": 0.6283876299858093, + "learning_rate": 0.0001680705146387246, + "loss": 1.0041, + "step": 5893 + }, + { + "epoch": 1.0493233618233617, + "grad_norm": 0.6075255870819092, + "learning_rate": 0.00016806026001856656, + "loss": 1.0661, + "step": 5894 + }, + { + "epoch": 1.0495014245014245, + "grad_norm": 0.5350496768951416, + "learning_rate": 0.00016805000406490907, + "loss": 0.6789, + "step": 5895 + }, + { + "epoch": 1.0496794871794872, + "grad_norm": 0.5380373597145081, + "learning_rate": 0.00016803974677795312, + "loss": 0.8889, + "step": 5896 + }, + { + "epoch": 1.0498575498575498, + "grad_norm": 0.6145668029785156, + "learning_rate": 0.0001680294881578997, + "loss": 0.8952, + "step": 5897 + }, + { + "epoch": 1.0500356125356125, + "grad_norm": 0.5666532516479492, + "learning_rate": 0.00016801922820494972, + "loss": 0.9697, + "step": 5898 + }, + { + "epoch": 1.0502136752136753, + "grad_norm": 0.5352747440338135, + "learning_rate": 0.0001680089669193043, + "loss": 0.9619, + "step": 5899 + }, + { + "epoch": 1.0503917378917378, + "grad_norm": 0.5405527949333191, + "learning_rate": 0.00016799870430116444, + "loss": 0.8733, + "step": 5900 + }, + { + "epoch": 1.0505698005698005, + "grad_norm": 0.5936748385429382, + "learning_rate": 0.00016798844035073124, + "loss": 0.8746, + "step": 5901 + }, + { + "epoch": 1.0507478632478633, + "grad_norm": 0.539652943611145, + "learning_rate": 0.00016797817506820578, + "loss": 0.8743, + "step": 5902 + }, + { + "epoch": 1.0509259259259258, + "grad_norm": 0.644528865814209, + "learning_rate": 0.00016796790845378915, + "loss": 0.9251, + "step": 5903 + }, + { + "epoch": 1.0511039886039886, + "grad_norm": 0.5429201126098633, + "learning_rate": 0.00016795764050768258, + "loss": 0.747, + "step": 5904 + }, + { + "epoch": 1.0512820512820513, + "grad_norm": 0.6432006359100342, + "learning_rate": 0.00016794737123008725, + "loss": 0.9166, + "step": 5905 + }, + { + "epoch": 1.0514601139601139, + "grad_norm": 0.6084117293357849, + "learning_rate": 0.00016793710062120427, + "loss": 1.0778, + "step": 5906 + }, + { + "epoch": 1.0516381766381766, + "grad_norm": 0.5351580381393433, + "learning_rate": 0.00016792682868123495, + "loss": 0.9124, + "step": 5907 + }, + { + "epoch": 1.0518162393162394, + "grad_norm": 0.7078854441642761, + "learning_rate": 0.00016791655541038053, + "loss": 1.1209, + "step": 5908 + }, + { + "epoch": 1.051994301994302, + "grad_norm": 0.5943832993507385, + "learning_rate": 0.0001679062808088423, + "loss": 0.9077, + "step": 5909 + }, + { + "epoch": 1.0521723646723646, + "grad_norm": 0.5216894745826721, + "learning_rate": 0.00016789600487682156, + "loss": 0.9866, + "step": 5910 + }, + { + "epoch": 1.0523504273504274, + "grad_norm": 0.738451361656189, + "learning_rate": 0.00016788572761451963, + "loss": 1.1611, + "step": 5911 + }, + { + "epoch": 1.05252849002849, + "grad_norm": 0.6411251425743103, + "learning_rate": 0.00016787544902213791, + "loss": 1.1481, + "step": 5912 + }, + { + "epoch": 1.0527065527065527, + "grad_norm": 0.6768319010734558, + "learning_rate": 0.00016786516909987774, + "loss": 0.8614, + "step": 5913 + }, + { + "epoch": 1.0528846153846154, + "grad_norm": 0.5838070511817932, + "learning_rate": 0.0001678548878479406, + "loss": 0.9719, + "step": 5914 + }, + { + "epoch": 1.0530626780626782, + "grad_norm": 0.541522741317749, + "learning_rate": 0.00016784460526652784, + "loss": 0.767, + "step": 5915 + }, + { + "epoch": 1.0532407407407407, + "grad_norm": 0.6064762473106384, + "learning_rate": 0.000167834321355841, + "loss": 1.0792, + "step": 5916 + }, + { + "epoch": 1.0534188034188035, + "grad_norm": 0.5515492558479309, + "learning_rate": 0.00016782403611608152, + "loss": 0.7897, + "step": 5917 + }, + { + "epoch": 1.0535968660968662, + "grad_norm": 0.6326262950897217, + "learning_rate": 0.000167813749547451, + "loss": 0.9279, + "step": 5918 + }, + { + "epoch": 1.0537749287749287, + "grad_norm": 0.6262009739875793, + "learning_rate": 0.0001678034616501509, + "loss": 0.9752, + "step": 5919 + }, + { + "epoch": 1.0539529914529915, + "grad_norm": 0.6049023270606995, + "learning_rate": 0.00016779317242438278, + "loss": 0.9167, + "step": 5920 + }, + { + "epoch": 1.0541310541310542, + "grad_norm": 0.6286031007766724, + "learning_rate": 0.0001677828818703483, + "loss": 1.1277, + "step": 5921 + }, + { + "epoch": 1.0543091168091168, + "grad_norm": 0.662086009979248, + "learning_rate": 0.00016777258998824907, + "loss": 1.0824, + "step": 5922 + }, + { + "epoch": 1.0544871794871795, + "grad_norm": 0.5358783006668091, + "learning_rate": 0.00016776229677828672, + "loss": 0.825, + "step": 5923 + }, + { + "epoch": 1.0546652421652423, + "grad_norm": 0.490326464176178, + "learning_rate": 0.00016775200224066294, + "loss": 0.7916, + "step": 5924 + }, + { + "epoch": 1.0548433048433048, + "grad_norm": 0.5940443277359009, + "learning_rate": 0.0001677417063755794, + "loss": 1.0121, + "step": 5925 + }, + { + "epoch": 1.0550213675213675, + "grad_norm": 0.5974507927894592, + "learning_rate": 0.00016773140918323787, + "loss": 0.7629, + "step": 5926 + }, + { + "epoch": 1.0551994301994303, + "grad_norm": 0.5747174024581909, + "learning_rate": 0.00016772111066384003, + "loss": 0.9373, + "step": 5927 + }, + { + "epoch": 1.0553774928774928, + "grad_norm": 0.5998024940490723, + "learning_rate": 0.00016771081081758772, + "loss": 0.8543, + "step": 5928 + }, + { + "epoch": 1.0555555555555556, + "grad_norm": 0.5771155953407288, + "learning_rate": 0.00016770050964468275, + "loss": 0.9108, + "step": 5929 + }, + { + "epoch": 1.0557336182336183, + "grad_norm": 0.5695661306381226, + "learning_rate": 0.00016769020714532692, + "loss": 0.8055, + "step": 5930 + }, + { + "epoch": 1.0559116809116809, + "grad_norm": 0.6164212226867676, + "learning_rate": 0.0001676799033197221, + "loss": 1.0917, + "step": 5931 + }, + { + "epoch": 1.0560897435897436, + "grad_norm": 0.6092487573623657, + "learning_rate": 0.00016766959816807018, + "loss": 0.9276, + "step": 5932 + }, + { + "epoch": 1.0562678062678064, + "grad_norm": 0.5595401525497437, + "learning_rate": 0.00016765929169057305, + "loss": 0.9435, + "step": 5933 + }, + { + "epoch": 1.056445868945869, + "grad_norm": 0.5875109434127808, + "learning_rate": 0.00016764898388743263, + "loss": 0.959, + "step": 5934 + }, + { + "epoch": 1.0566239316239316, + "grad_norm": 0.6045668721199036, + "learning_rate": 0.00016763867475885088, + "loss": 0.8636, + "step": 5935 + }, + { + "epoch": 1.0568019943019944, + "grad_norm": 0.6088171005249023, + "learning_rate": 0.00016762836430502987, + "loss": 0.6807, + "step": 5936 + }, + { + "epoch": 1.056980056980057, + "grad_norm": 0.6293274760246277, + "learning_rate": 0.00016761805252617148, + "loss": 1.042, + "step": 5937 + }, + { + "epoch": 1.0571581196581197, + "grad_norm": 0.588472843170166, + "learning_rate": 0.00016760773942247785, + "loss": 0.8896, + "step": 5938 + }, + { + "epoch": 1.0573361823361824, + "grad_norm": 0.4412326216697693, + "learning_rate": 0.000167597424994151, + "loss": 0.6727, + "step": 5939 + }, + { + "epoch": 1.057514245014245, + "grad_norm": 0.6086825132369995, + "learning_rate": 0.00016758710924139302, + "loss": 0.9908, + "step": 5940 + }, + { + "epoch": 1.0576923076923077, + "grad_norm": 0.6424705386161804, + "learning_rate": 0.00016757679216440608, + "loss": 1.0182, + "step": 5941 + }, + { + "epoch": 1.0578703703703705, + "grad_norm": 0.6610676050186157, + "learning_rate": 0.00016756647376339222, + "loss": 0.9645, + "step": 5942 + }, + { + "epoch": 1.058048433048433, + "grad_norm": 0.598292887210846, + "learning_rate": 0.0001675561540385537, + "loss": 0.9694, + "step": 5943 + }, + { + "epoch": 1.0582264957264957, + "grad_norm": 0.6941167116165161, + "learning_rate": 0.00016754583299009266, + "loss": 1.0786, + "step": 5944 + }, + { + "epoch": 1.0584045584045585, + "grad_norm": 0.6543232798576355, + "learning_rate": 0.00016753551061821133, + "loss": 1.0488, + "step": 5945 + }, + { + "epoch": 1.058582621082621, + "grad_norm": 0.606159508228302, + "learning_rate": 0.000167525186923112, + "loss": 0.9448, + "step": 5946 + }, + { + "epoch": 1.0587606837606838, + "grad_norm": 0.5051791071891785, + "learning_rate": 0.00016751486190499685, + "loss": 0.7485, + "step": 5947 + }, + { + "epoch": 1.0589387464387465, + "grad_norm": 0.6459367275238037, + "learning_rate": 0.00016750453556406826, + "loss": 1.0055, + "step": 5948 + }, + { + "epoch": 1.059116809116809, + "grad_norm": 0.551591157913208, + "learning_rate": 0.00016749420790052852, + "loss": 0.9717, + "step": 5949 + }, + { + "epoch": 1.0592948717948718, + "grad_norm": 0.5899214148521423, + "learning_rate": 0.00016748387891458, + "loss": 0.7774, + "step": 5950 + }, + { + "epoch": 1.0594729344729346, + "grad_norm": 0.582379162311554, + "learning_rate": 0.00016747354860642503, + "loss": 0.953, + "step": 5951 + }, + { + "epoch": 1.059650997150997, + "grad_norm": 0.6035816073417664, + "learning_rate": 0.00016746321697626605, + "loss": 1.1175, + "step": 5952 + }, + { + "epoch": 1.0598290598290598, + "grad_norm": 0.6476401686668396, + "learning_rate": 0.00016745288402430548, + "loss": 0.9448, + "step": 5953 + }, + { + "epoch": 1.0600071225071226, + "grad_norm": 0.6126405596733093, + "learning_rate": 0.00016744254975074578, + "loss": 0.882, + "step": 5954 + }, + { + "epoch": 1.0601851851851851, + "grad_norm": 0.5333579182624817, + "learning_rate": 0.0001674322141557894, + "loss": 0.9539, + "step": 5955 + }, + { + "epoch": 1.0603632478632479, + "grad_norm": 0.6085022687911987, + "learning_rate": 0.0001674218772396389, + "loss": 1.0028, + "step": 5956 + }, + { + "epoch": 1.0605413105413106, + "grad_norm": 0.5809528827667236, + "learning_rate": 0.0001674115390024967, + "loss": 0.84, + "step": 5957 + }, + { + "epoch": 1.0607193732193732, + "grad_norm": 0.5820229649543762, + "learning_rate": 0.00016740119944456548, + "loss": 0.9563, + "step": 5958 + }, + { + "epoch": 1.060897435897436, + "grad_norm": 0.6349015831947327, + "learning_rate": 0.00016739085856604775, + "loss": 0.9739, + "step": 5959 + }, + { + "epoch": 1.0610754985754987, + "grad_norm": 0.6346020102500916, + "learning_rate": 0.00016738051636714616, + "loss": 0.907, + "step": 5960 + }, + { + "epoch": 1.0612535612535612, + "grad_norm": 0.5850573778152466, + "learning_rate": 0.0001673701728480633, + "loss": 1.0688, + "step": 5961 + }, + { + "epoch": 1.061431623931624, + "grad_norm": 0.6258122324943542, + "learning_rate": 0.00016735982800900184, + "loss": 0.9997, + "step": 5962 + }, + { + "epoch": 1.0616096866096867, + "grad_norm": 0.6744239330291748, + "learning_rate": 0.00016734948185016452, + "loss": 0.9431, + "step": 5963 + }, + { + "epoch": 1.0617877492877492, + "grad_norm": 0.5769457817077637, + "learning_rate": 0.000167339134371754, + "loss": 0.9658, + "step": 5964 + }, + { + "epoch": 1.061965811965812, + "grad_norm": 0.6385112404823303, + "learning_rate": 0.000167328785573973, + "loss": 1.0199, + "step": 5965 + }, + { + "epoch": 1.0621438746438747, + "grad_norm": 0.536522388458252, + "learning_rate": 0.00016731843545702435, + "loss": 0.8496, + "step": 5966 + }, + { + "epoch": 1.0623219373219372, + "grad_norm": 0.5978497862815857, + "learning_rate": 0.00016730808402111075, + "loss": 0.8536, + "step": 5967 + }, + { + "epoch": 1.0625, + "grad_norm": 0.6091681122779846, + "learning_rate": 0.0001672977312664351, + "loss": 1.0241, + "step": 5968 + }, + { + "epoch": 1.0626780626780628, + "grad_norm": 0.5807273387908936, + "learning_rate": 0.0001672873771932002, + "loss": 1.0522, + "step": 5969 + }, + { + "epoch": 1.0628561253561253, + "grad_norm": 0.6511965990066528, + "learning_rate": 0.0001672770218016089, + "loss": 0.8908, + "step": 5970 + }, + { + "epoch": 1.063034188034188, + "grad_norm": 0.6241721510887146, + "learning_rate": 0.00016726666509186416, + "loss": 0.9854, + "step": 5971 + }, + { + "epoch": 1.0632122507122508, + "grad_norm": 0.6112468242645264, + "learning_rate": 0.0001672563070641688, + "loss": 1.0091, + "step": 5972 + }, + { + "epoch": 1.0633903133903133, + "grad_norm": 0.6135509014129639, + "learning_rate": 0.00016724594771872587, + "loss": 0.8891, + "step": 5973 + }, + { + "epoch": 1.063568376068376, + "grad_norm": 0.608384370803833, + "learning_rate": 0.00016723558705573823, + "loss": 1.017, + "step": 5974 + }, + { + "epoch": 1.0637464387464388, + "grad_norm": 0.6578485369682312, + "learning_rate": 0.00016722522507540895, + "loss": 0.9165, + "step": 5975 + }, + { + "epoch": 1.0639245014245013, + "grad_norm": 0.562588095664978, + "learning_rate": 0.00016721486177794106, + "loss": 0.7989, + "step": 5976 + }, + { + "epoch": 1.064102564102564, + "grad_norm": 0.5541409254074097, + "learning_rate": 0.00016720449716353753, + "loss": 0.8917, + "step": 5977 + }, + { + "epoch": 1.0642806267806268, + "grad_norm": 0.551167905330658, + "learning_rate": 0.0001671941312324015, + "loss": 0.824, + "step": 5978 + }, + { + "epoch": 1.0644586894586894, + "grad_norm": 0.6280582547187805, + "learning_rate": 0.0001671837639847361, + "loss": 0.9708, + "step": 5979 + }, + { + "epoch": 1.0646367521367521, + "grad_norm": 0.6389226913452148, + "learning_rate": 0.00016717339542074436, + "loss": 1.0081, + "step": 5980 + }, + { + "epoch": 1.0648148148148149, + "grad_norm": 0.6677889823913574, + "learning_rate": 0.0001671630255406295, + "loss": 1.2709, + "step": 5981 + }, + { + "epoch": 1.0649928774928774, + "grad_norm": 0.5748161673545837, + "learning_rate": 0.00016715265434459465, + "loss": 0.9157, + "step": 5982 + }, + { + "epoch": 1.0651709401709402, + "grad_norm": 0.6677651405334473, + "learning_rate": 0.00016714228183284304, + "loss": 1.1097, + "step": 5983 + }, + { + "epoch": 1.065349002849003, + "grad_norm": 0.6253604292869568, + "learning_rate": 0.0001671319080055779, + "loss": 0.9819, + "step": 5984 + }, + { + "epoch": 1.0655270655270654, + "grad_norm": 0.5548844337463379, + "learning_rate": 0.0001671215328630025, + "loss": 0.9324, + "step": 5985 + }, + { + "epoch": 1.0657051282051282, + "grad_norm": 0.622062623500824, + "learning_rate": 0.00016711115640532004, + "loss": 0.8749, + "step": 5986 + }, + { + "epoch": 1.065883190883191, + "grad_norm": 0.6496043801307678, + "learning_rate": 0.00016710077863273394, + "loss": 1.0642, + "step": 5987 + }, + { + "epoch": 1.0660612535612535, + "grad_norm": 0.6140534281730652, + "learning_rate": 0.00016709039954544746, + "loss": 0.8928, + "step": 5988 + }, + { + "epoch": 1.0662393162393162, + "grad_norm": 0.6387218236923218, + "learning_rate": 0.00016708001914366393, + "loss": 0.9525, + "step": 5989 + }, + { + "epoch": 1.066417378917379, + "grad_norm": 0.6119858026504517, + "learning_rate": 0.0001670696374275868, + "loss": 0.8663, + "step": 5990 + }, + { + "epoch": 1.0665954415954415, + "grad_norm": 0.6722040772438049, + "learning_rate": 0.00016705925439741947, + "loss": 1.1173, + "step": 5991 + }, + { + "epoch": 1.0667735042735043, + "grad_norm": 0.8226081132888794, + "learning_rate": 0.00016704887005336534, + "loss": 1.0572, + "step": 5992 + }, + { + "epoch": 1.066951566951567, + "grad_norm": 0.7248596549034119, + "learning_rate": 0.00016703848439562785, + "loss": 1.0493, + "step": 5993 + }, + { + "epoch": 1.0671296296296295, + "grad_norm": 0.7185787558555603, + "learning_rate": 0.00016702809742441058, + "loss": 1.1366, + "step": 5994 + }, + { + "epoch": 1.0673076923076923, + "grad_norm": 0.6118780970573425, + "learning_rate": 0.00016701770913991694, + "loss": 0.9557, + "step": 5995 + }, + { + "epoch": 1.067485754985755, + "grad_norm": 0.6472596526145935, + "learning_rate": 0.0001670073195423505, + "loss": 0.9977, + "step": 5996 + }, + { + "epoch": 1.0676638176638176, + "grad_norm": 0.7110133767127991, + "learning_rate": 0.00016699692863191484, + "loss": 1.1932, + "step": 5997 + }, + { + "epoch": 1.0678418803418803, + "grad_norm": 0.5827305912971497, + "learning_rate": 0.00016698653640881354, + "loss": 0.7641, + "step": 5998 + }, + { + "epoch": 1.068019943019943, + "grad_norm": 0.527208149433136, + "learning_rate": 0.00016697614287325017, + "loss": 0.7683, + "step": 5999 + }, + { + "epoch": 1.0681980056980056, + "grad_norm": 0.6680626273155212, + "learning_rate": 0.00016696574802542848, + "loss": 1.1748, + "step": 6000 + }, + { + "epoch": 1.0683760683760684, + "grad_norm": 0.5947227478027344, + "learning_rate": 0.00016695535186555204, + "loss": 1.0894, + "step": 6001 + }, + { + "epoch": 1.068554131054131, + "grad_norm": 0.5828250646591187, + "learning_rate": 0.00016694495439382456, + "loss": 0.9895, + "step": 6002 + }, + { + "epoch": 1.0687321937321936, + "grad_norm": 0.5897728204727173, + "learning_rate": 0.00016693455561044978, + "loss": 0.9686, + "step": 6003 + }, + { + "epoch": 1.0689102564102564, + "grad_norm": 0.5441751480102539, + "learning_rate": 0.0001669241555156314, + "loss": 0.8948, + "step": 6004 + }, + { + "epoch": 1.0690883190883191, + "grad_norm": 0.694199800491333, + "learning_rate": 0.00016691375410957324, + "loss": 1.0824, + "step": 6005 + }, + { + "epoch": 1.0692663817663817, + "grad_norm": 0.6077630519866943, + "learning_rate": 0.00016690335139247906, + "loss": 1.0931, + "step": 6006 + }, + { + "epoch": 1.0694444444444444, + "grad_norm": 0.6558539867401123, + "learning_rate": 0.0001668929473645527, + "loss": 1.0099, + "step": 6007 + }, + { + "epoch": 1.0696225071225072, + "grad_norm": 0.5722812414169312, + "learning_rate": 0.00016688254202599798, + "loss": 0.7999, + "step": 6008 + }, + { + "epoch": 1.0698005698005697, + "grad_norm": 0.5915400981903076, + "learning_rate": 0.0001668721353770188, + "loss": 0.7866, + "step": 6009 + }, + { + "epoch": 1.0699786324786325, + "grad_norm": 0.5290952324867249, + "learning_rate": 0.00016686172741781901, + "loss": 0.793, + "step": 6010 + }, + { + "epoch": 1.0701566951566952, + "grad_norm": 0.5501774549484253, + "learning_rate": 0.00016685131814860263, + "loss": 0.8775, + "step": 6011 + }, + { + "epoch": 1.0703347578347577, + "grad_norm": 0.6192594766616821, + "learning_rate": 0.00016684090756957347, + "loss": 1.1686, + "step": 6012 + }, + { + "epoch": 1.0705128205128205, + "grad_norm": 0.6640267968177795, + "learning_rate": 0.00016683049568093561, + "loss": 1.1789, + "step": 6013 + }, + { + "epoch": 1.0706908831908832, + "grad_norm": 0.552893877029419, + "learning_rate": 0.00016682008248289303, + "loss": 0.7957, + "step": 6014 + }, + { + "epoch": 1.0708689458689458, + "grad_norm": 0.6406302452087402, + "learning_rate": 0.00016680966797564972, + "loss": 1.1174, + "step": 6015 + }, + { + "epoch": 1.0710470085470085, + "grad_norm": Infinity, + "learning_rate": 0.00016680966797564972, + "loss": 0.9168, + "step": 6016 + }, + { + "epoch": 1.0712250712250713, + "grad_norm": 0.6384762525558472, + "learning_rate": 0.00016679925215940975, + "loss": 0.9831, + "step": 6017 + }, + { + "epoch": 1.071403133903134, + "grad_norm": 0.5906224846839905, + "learning_rate": 0.0001667888350343772, + "loss": 0.9167, + "step": 6018 + }, + { + "epoch": 1.0715811965811965, + "grad_norm": 0.658044695854187, + "learning_rate": 0.00016677841660075617, + "loss": 1.0075, + "step": 6019 + }, + { + "epoch": 1.0717592592592593, + "grad_norm": 0.6313242316246033, + "learning_rate": 0.00016676799685875078, + "loss": 0.8551, + "step": 6020 + }, + { + "epoch": 1.0719373219373218, + "grad_norm": 0.5891841053962708, + "learning_rate": 0.00016675757580856518, + "loss": 0.8475, + "step": 6021 + }, + { + "epoch": 1.0721153846153846, + "grad_norm": 0.581317126750946, + "learning_rate": 0.00016674715345040358, + "loss": 0.9308, + "step": 6022 + }, + { + "epoch": 1.0722934472934473, + "grad_norm": 0.5952537655830383, + "learning_rate": 0.00016673672978447017, + "loss": 0.9104, + "step": 6023 + }, + { + "epoch": 1.07247150997151, + "grad_norm": 0.5934227705001831, + "learning_rate": 0.00016672630481096915, + "loss": 0.9882, + "step": 6024 + }, + { + "epoch": 1.0726495726495726, + "grad_norm": 0.5867539048194885, + "learning_rate": 0.00016671587853010482, + "loss": 1.0186, + "step": 6025 + }, + { + "epoch": 1.0728276353276354, + "grad_norm": 0.6002280116081238, + "learning_rate": 0.00016670545094208143, + "loss": 0.92, + "step": 6026 + }, + { + "epoch": 1.073005698005698, + "grad_norm": 0.6261683702468872, + "learning_rate": 0.0001666950220471033, + "loss": 0.9293, + "step": 6027 + }, + { + "epoch": 1.0731837606837606, + "grad_norm": 0.6128147840499878, + "learning_rate": 0.00016668459184537477, + "loss": 1.0787, + "step": 6028 + }, + { + "epoch": 1.0733618233618234, + "grad_norm": 0.62148118019104, + "learning_rate": 0.00016667416033710016, + "loss": 0.8843, + "step": 6029 + }, + { + "epoch": 1.0735398860398861, + "grad_norm": 0.7166166305541992, + "learning_rate": 0.0001666637275224839, + "loss": 0.8877, + "step": 6030 + }, + { + "epoch": 1.0737179487179487, + "grad_norm": 0.5275574922561646, + "learning_rate": 0.0001666532934017304, + "loss": 0.9604, + "step": 6031 + }, + { + "epoch": 1.0738960113960114, + "grad_norm": 0.8132784962654114, + "learning_rate": 0.00016664285797504406, + "loss": 1.0203, + "step": 6032 + }, + { + "epoch": 1.074074074074074, + "grad_norm": 0.5887695550918579, + "learning_rate": 0.00016663242124262935, + "loss": 0.8819, + "step": 6033 + }, + { + "epoch": 1.0742521367521367, + "grad_norm": 0.5552900433540344, + "learning_rate": 0.00016662198320469078, + "loss": 0.7542, + "step": 6034 + }, + { + "epoch": 1.0744301994301995, + "grad_norm": 0.6228970885276794, + "learning_rate": 0.0001666115438614328, + "loss": 1.0362, + "step": 6035 + }, + { + "epoch": 1.0746082621082622, + "grad_norm": 0.7193471789360046, + "learning_rate": 0.00016660110321306003, + "loss": 1.3073, + "step": 6036 + }, + { + "epoch": 1.0747863247863247, + "grad_norm": 0.6167412996292114, + "learning_rate": 0.000166590661259777, + "loss": 0.941, + "step": 6037 + }, + { + "epoch": 1.0749643874643875, + "grad_norm": 0.5716922879219055, + "learning_rate": 0.00016658021800178827, + "loss": 0.83, + "step": 6038 + }, + { + "epoch": 1.0751424501424502, + "grad_norm": 0.6404047012329102, + "learning_rate": 0.00016656977343929848, + "loss": 1.0617, + "step": 6039 + }, + { + "epoch": 1.0753205128205128, + "grad_norm": 0.531395435333252, + "learning_rate": 0.00016655932757251226, + "loss": 0.7785, + "step": 6040 + }, + { + "epoch": 1.0754985754985755, + "grad_norm": 0.6468462347984314, + "learning_rate": 0.0001665488804016343, + "loss": 0.7893, + "step": 6041 + }, + { + "epoch": 1.0756766381766383, + "grad_norm": 0.6539653539657593, + "learning_rate": 0.00016653843192686925, + "loss": 1.1011, + "step": 6042 + }, + { + "epoch": 1.0758547008547008, + "grad_norm": 0.630107045173645, + "learning_rate": 0.0001665279821484219, + "loss": 0.9262, + "step": 6043 + }, + { + "epoch": 1.0760327635327636, + "grad_norm": 0.5875992774963379, + "learning_rate": 0.00016651753106649688, + "loss": 1.0501, + "step": 6044 + }, + { + "epoch": 1.0762108262108263, + "grad_norm": 0.573428750038147, + "learning_rate": 0.00016650707868129904, + "loss": 1.0672, + "step": 6045 + }, + { + "epoch": 1.0763888888888888, + "grad_norm": 0.6215469241142273, + "learning_rate": 0.00016649662499303316, + "loss": 0.868, + "step": 6046 + }, + { + "epoch": 1.0765669515669516, + "grad_norm": 0.6666893362998962, + "learning_rate": 0.00016648617000190402, + "loss": 1.0965, + "step": 6047 + }, + { + "epoch": 1.0767450142450143, + "grad_norm": 0.8343498706817627, + "learning_rate": 0.00016647571370811653, + "loss": 1.2302, + "step": 6048 + }, + { + "epoch": 1.0769230769230769, + "grad_norm": 0.591147780418396, + "learning_rate": 0.0001664652561118755, + "loss": 0.9698, + "step": 6049 + }, + { + "epoch": 1.0771011396011396, + "grad_norm": 0.573375940322876, + "learning_rate": 0.00016645479721338584, + "loss": 0.8798, + "step": 6050 + }, + { + "epoch": 1.0772792022792024, + "grad_norm": 0.4956737160682678, + "learning_rate": 0.00016644433701285246, + "loss": 0.6523, + "step": 6051 + }, + { + "epoch": 1.077457264957265, + "grad_norm": 0.6896619200706482, + "learning_rate": 0.00016643387551048034, + "loss": 0.8911, + "step": 6052 + }, + { + "epoch": 1.0776353276353277, + "grad_norm": 0.5820416808128357, + "learning_rate": 0.00016642341270647445, + "loss": 1.1486, + "step": 6053 + }, + { + "epoch": 1.0778133903133904, + "grad_norm": 0.611132025718689, + "learning_rate": 0.00016641294860103976, + "loss": 1.0705, + "step": 6054 + }, + { + "epoch": 1.077991452991453, + "grad_norm": 0.6705698370933533, + "learning_rate": 0.00016640248319438133, + "loss": 0.9826, + "step": 6055 + }, + { + "epoch": 1.0781695156695157, + "grad_norm": 0.5987013578414917, + "learning_rate": 0.00016639201648670416, + "loss": 1.0409, + "step": 6056 + }, + { + "epoch": 1.0783475783475784, + "grad_norm": 0.6707149744033813, + "learning_rate": 0.00016638154847821332, + "loss": 1.1332, + "step": 6057 + }, + { + "epoch": 1.078525641025641, + "grad_norm": 0.6400678157806396, + "learning_rate": 0.00016637107916911393, + "loss": 1.2559, + "step": 6058 + }, + { + "epoch": 1.0787037037037037, + "grad_norm": 0.6370311379432678, + "learning_rate": 0.00016636060855961115, + "loss": 0.9752, + "step": 6059 + }, + { + "epoch": 1.0788817663817665, + "grad_norm": 0.6116052269935608, + "learning_rate": 0.00016635013664991012, + "loss": 0.8364, + "step": 6060 + }, + { + "epoch": 1.079059829059829, + "grad_norm": 0.7932127714157104, + "learning_rate": 0.00016633966344021593, + "loss": 0.939, + "step": 6061 + }, + { + "epoch": 1.0792378917378918, + "grad_norm": 0.576249897480011, + "learning_rate": 0.00016632918893073385, + "loss": 0.8911, + "step": 6062 + }, + { + "epoch": 1.0794159544159545, + "grad_norm": 0.5456888675689697, + "learning_rate": 0.00016631871312166915, + "loss": 0.8646, + "step": 6063 + }, + { + "epoch": 1.079594017094017, + "grad_norm": 0.717522919178009, + "learning_rate": 0.000166308236013227, + "loss": 1.0814, + "step": 6064 + }, + { + "epoch": 1.0797720797720798, + "grad_norm": 0.6637256145477295, + "learning_rate": 0.0001662977576056127, + "loss": 1.22, + "step": 6065 + }, + { + "epoch": 1.0799501424501425, + "grad_norm": 0.5846666693687439, + "learning_rate": 0.0001662872778990316, + "loss": 1.1745, + "step": 6066 + }, + { + "epoch": 1.080128205128205, + "grad_norm": 0.6611326336860657, + "learning_rate": 0.00016627679689368895, + "loss": 1.1262, + "step": 6067 + }, + { + "epoch": 1.0803062678062678, + "grad_norm": 0.6022892594337463, + "learning_rate": 0.00016626631458979015, + "loss": 0.9741, + "step": 6068 + }, + { + "epoch": 1.0804843304843306, + "grad_norm": 0.5862685441970825, + "learning_rate": 0.00016625583098754058, + "loss": 0.914, + "step": 6069 + }, + { + "epoch": 1.080662393162393, + "grad_norm": 0.7089241147041321, + "learning_rate": 0.00016624534608714563, + "loss": 1.0614, + "step": 6070 + }, + { + "epoch": 1.0808404558404558, + "grad_norm": 0.5286028981208801, + "learning_rate": 0.00016623485988881076, + "loss": 0.8756, + "step": 6071 + }, + { + "epoch": 1.0810185185185186, + "grad_norm": 0.6437101364135742, + "learning_rate": 0.00016622437239274137, + "loss": 0.7222, + "step": 6072 + }, + { + "epoch": 1.0811965811965811, + "grad_norm": 0.6197740435600281, + "learning_rate": 0.000166213883599143, + "loss": 0.7876, + "step": 6073 + }, + { + "epoch": 1.0813746438746439, + "grad_norm": 0.5889328122138977, + "learning_rate": 0.0001662033935082211, + "loss": 0.9587, + "step": 6074 + }, + { + "epoch": 1.0815527065527066, + "grad_norm": 0.5353847742080688, + "learning_rate": 0.00016619290212018125, + "loss": 0.8664, + "step": 6075 + }, + { + "epoch": 1.0817307692307692, + "grad_norm": 0.7202061414718628, + "learning_rate": 0.00016618240943522898, + "loss": 1.0429, + "step": 6076 + }, + { + "epoch": 1.081908831908832, + "grad_norm": 0.5831515192985535, + "learning_rate": 0.0001661719154535699, + "loss": 1.0323, + "step": 6077 + }, + { + "epoch": 1.0820868945868947, + "grad_norm": 0.6270500421524048, + "learning_rate": 0.00016616142017540953, + "loss": 0.9272, + "step": 6078 + }, + { + "epoch": 1.0822649572649572, + "grad_norm": 0.6064695119857788, + "learning_rate": 0.00016615092360095364, + "loss": 1.0629, + "step": 6079 + }, + { + "epoch": 1.08244301994302, + "grad_norm": 0.5578122138977051, + "learning_rate": 0.00016614042573040777, + "loss": 0.8601, + "step": 6080 + }, + { + "epoch": 1.0826210826210827, + "grad_norm": 0.5920688509941101, + "learning_rate": 0.0001661299265639777, + "loss": 1.0082, + "step": 6081 + }, + { + "epoch": 1.0827991452991452, + "grad_norm": 0.6191682815551758, + "learning_rate": 0.0001661194261018691, + "loss": 0.9645, + "step": 6082 + }, + { + "epoch": 1.082977207977208, + "grad_norm": 0.6403279304504395, + "learning_rate": 0.00016610892434428765, + "loss": 0.9263, + "step": 6083 + }, + { + "epoch": 1.0831552706552707, + "grad_norm": 0.579502284526825, + "learning_rate": 0.00016609842129143915, + "loss": 0.8997, + "step": 6084 + }, + { + "epoch": 1.0833333333333333, + "grad_norm": 0.5831437706947327, + "learning_rate": 0.00016608791694352944, + "loss": 1.0703, + "step": 6085 + }, + { + "epoch": 1.083511396011396, + "grad_norm": 0.6188452243804932, + "learning_rate": 0.00016607741130076424, + "loss": 0.8856, + "step": 6086 + }, + { + "epoch": 1.0836894586894588, + "grad_norm": 0.7413692474365234, + "learning_rate": 0.00016606690436334946, + "loss": 1.1995, + "step": 6087 + }, + { + "epoch": 1.0838675213675213, + "grad_norm": 0.5552099347114563, + "learning_rate": 0.00016605639613149093, + "loss": 0.8514, + "step": 6088 + }, + { + "epoch": 1.084045584045584, + "grad_norm": 0.5906503200531006, + "learning_rate": 0.00016604588660539452, + "loss": 0.9431, + "step": 6089 + }, + { + "epoch": 1.0842236467236468, + "grad_norm": 0.5326111316680908, + "learning_rate": 0.0001660353757852662, + "loss": 0.8306, + "step": 6090 + }, + { + "epoch": 1.0844017094017093, + "grad_norm": 0.7273091673851013, + "learning_rate": 0.0001660248636713118, + "loss": 1.1109, + "step": 6091 + }, + { + "epoch": 1.084579772079772, + "grad_norm": 0.66513592004776, + "learning_rate": 0.00016601435026373737, + "loss": 1.0621, + "step": 6092 + }, + { + "epoch": 1.0847578347578348, + "grad_norm": 0.6470831632614136, + "learning_rate": 0.00016600383556274892, + "loss": 1.1075, + "step": 6093 + }, + { + "epoch": 1.0849358974358974, + "grad_norm": 0.6308658719062805, + "learning_rate": 0.0001659933195685524, + "loss": 0.9832, + "step": 6094 + }, + { + "epoch": 1.08511396011396, + "grad_norm": 0.6569336652755737, + "learning_rate": 0.00016598280228135388, + "loss": 0.9754, + "step": 6095 + }, + { + "epoch": 1.0852920227920229, + "grad_norm": 0.5672318339347839, + "learning_rate": 0.0001659722837013594, + "loss": 0.9075, + "step": 6096 + }, + { + "epoch": 1.0854700854700854, + "grad_norm": 0.6397247314453125, + "learning_rate": 0.00016596176382877506, + "loss": 1.0358, + "step": 6097 + }, + { + "epoch": 1.0856481481481481, + "grad_norm": 0.6046154499053955, + "learning_rate": 0.000165951242663807, + "loss": 0.9036, + "step": 6098 + }, + { + "epoch": 1.085826210826211, + "grad_norm": 0.7190790176391602, + "learning_rate": 0.00016594072020666134, + "loss": 1.05, + "step": 6099 + }, + { + "epoch": 1.0860042735042734, + "grad_norm": 0.636986255645752, + "learning_rate": 0.00016593019645754425, + "loss": 1.0648, + "step": 6100 + }, + { + "epoch": 1.0861823361823362, + "grad_norm": 0.7239426374435425, + "learning_rate": 0.00016591967141666193, + "loss": 1.3332, + "step": 6101 + }, + { + "epoch": 1.086360398860399, + "grad_norm": 0.5623281002044678, + "learning_rate": 0.00016590914508422054, + "loss": 0.997, + "step": 6102 + }, + { + "epoch": 1.0865384615384615, + "grad_norm": 0.5559574365615845, + "learning_rate": 0.00016589861746042642, + "loss": 0.9309, + "step": 6103 + }, + { + "epoch": 1.0867165242165242, + "grad_norm": 0.6056998372077942, + "learning_rate": 0.00016588808854548574, + "loss": 1.05, + "step": 6104 + }, + { + "epoch": 1.086894586894587, + "grad_norm": 0.6419603228569031, + "learning_rate": 0.00016587755833960487, + "loss": 0.8933, + "step": 6105 + }, + { + "epoch": 1.0870726495726495, + "grad_norm": 0.5236496329307556, + "learning_rate": 0.00016586702684299006, + "loss": 1.0061, + "step": 6106 + }, + { + "epoch": 1.0872507122507122, + "grad_norm": 0.5764613747596741, + "learning_rate": 0.0001658564940558477, + "loss": 1.0218, + "step": 6107 + }, + { + "epoch": 1.087428774928775, + "grad_norm": 0.6049391627311707, + "learning_rate": 0.00016584595997838416, + "loss": 0.8157, + "step": 6108 + }, + { + "epoch": 1.0876068376068375, + "grad_norm": 0.585422933101654, + "learning_rate": 0.0001658354246108058, + "loss": 1.2761, + "step": 6109 + }, + { + "epoch": 1.0877849002849003, + "grad_norm": 0.6420125365257263, + "learning_rate": 0.00016582488795331907, + "loss": 1.1978, + "step": 6110 + }, + { + "epoch": 1.087962962962963, + "grad_norm": 0.646091878414154, + "learning_rate": 0.00016581435000613038, + "loss": 0.8946, + "step": 6111 + }, + { + "epoch": 1.0881410256410255, + "grad_norm": 0.6563934087753296, + "learning_rate": 0.00016580381076944625, + "loss": 1.0625, + "step": 6112 + }, + { + "epoch": 1.0883190883190883, + "grad_norm": 0.6796613931655884, + "learning_rate": 0.0001657932702434731, + "loss": 0.9401, + "step": 6113 + }, + { + "epoch": 1.088497150997151, + "grad_norm": 0.6248648762702942, + "learning_rate": 0.00016578272842841753, + "loss": 0.8558, + "step": 6114 + }, + { + "epoch": 1.0886752136752136, + "grad_norm": 0.5136269330978394, + "learning_rate": 0.00016577218532448605, + "loss": 0.6424, + "step": 6115 + }, + { + "epoch": 1.0888532763532763, + "grad_norm": 0.5581641793251038, + "learning_rate": 0.00016576164093188523, + "loss": 0.7923, + "step": 6116 + }, + { + "epoch": 1.089031339031339, + "grad_norm": 0.630352258682251, + "learning_rate": 0.0001657510952508216, + "loss": 0.9115, + "step": 6117 + }, + { + "epoch": 1.0892094017094016, + "grad_norm": 0.6167593002319336, + "learning_rate": 0.0001657405482815019, + "loss": 1.1112, + "step": 6118 + }, + { + "epoch": 1.0893874643874644, + "grad_norm": 0.5908578634262085, + "learning_rate": 0.00016573000002413271, + "loss": 1.0359, + "step": 6119 + }, + { + "epoch": 1.0895655270655271, + "grad_norm": 0.6326140761375427, + "learning_rate": 0.00016571945047892073, + "loss": 1.0459, + "step": 6120 + }, + { + "epoch": 1.0897435897435896, + "grad_norm": 0.7273572683334351, + "learning_rate": 0.00016570889964607262, + "loss": 1.0901, + "step": 6121 + }, + { + "epoch": 1.0899216524216524, + "grad_norm": 0.6168062090873718, + "learning_rate": 0.00016569834752579513, + "loss": 0.8739, + "step": 6122 + }, + { + "epoch": 1.0900997150997151, + "grad_norm": 0.5620378255844116, + "learning_rate": 0.00016568779411829497, + "loss": 0.9614, + "step": 6123 + }, + { + "epoch": 1.0902777777777777, + "grad_norm": 0.6319156885147095, + "learning_rate": 0.00016567723942377899, + "loss": 1.1031, + "step": 6124 + }, + { + "epoch": 1.0904558404558404, + "grad_norm": 0.6590072512626648, + "learning_rate": 0.00016566668344245388, + "loss": 1.0086, + "step": 6125 + }, + { + "epoch": 1.0906339031339032, + "grad_norm": 0.5823387503623962, + "learning_rate": 0.00016565612617452656, + "loss": 0.8886, + "step": 6126 + }, + { + "epoch": 1.0908119658119657, + "grad_norm": 0.5795989632606506, + "learning_rate": 0.00016564556762020381, + "loss": 0.7683, + "step": 6127 + }, + { + "epoch": 1.0909900284900285, + "grad_norm": 0.5940101742744446, + "learning_rate": 0.00016563500777969255, + "loss": 0.8873, + "step": 6128 + }, + { + "epoch": 1.0911680911680912, + "grad_norm": 0.5708247423171997, + "learning_rate": 0.00016562444665319963, + "loss": 0.7382, + "step": 6129 + }, + { + "epoch": 1.0913461538461537, + "grad_norm": 0.6339239478111267, + "learning_rate": 0.00016561388424093202, + "loss": 0.9323, + "step": 6130 + }, + { + "epoch": 1.0915242165242165, + "grad_norm": 0.720000147819519, + "learning_rate": 0.00016560332054309663, + "loss": 1.0437, + "step": 6131 + }, + { + "epoch": 1.0917022792022792, + "grad_norm": 0.686580240726471, + "learning_rate": 0.00016559275555990048, + "loss": 0.9841, + "step": 6132 + }, + { + "epoch": 1.091880341880342, + "grad_norm": 0.6067900061607361, + "learning_rate": 0.00016558218929155053, + "loss": 1.0862, + "step": 6133 + }, + { + "epoch": 1.0920584045584045, + "grad_norm": 0.6678896546363831, + "learning_rate": 0.00016557162173825384, + "loss": 0.8509, + "step": 6134 + }, + { + "epoch": 1.0922364672364673, + "grad_norm": 0.53044193983078, + "learning_rate": 0.0001655610529002174, + "loss": 0.9227, + "step": 6135 + }, + { + "epoch": 1.0924145299145298, + "grad_norm": 0.6499412655830383, + "learning_rate": 0.00016555048277764836, + "loss": 1.0867, + "step": 6136 + }, + { + "epoch": 1.0925925925925926, + "grad_norm": 0.6543099284172058, + "learning_rate": 0.00016553991137075374, + "loss": 0.849, + "step": 6137 + }, + { + "epoch": 1.0927706552706553, + "grad_norm": 0.5772737860679626, + "learning_rate": 0.0001655293386797407, + "loss": 0.8475, + "step": 6138 + }, + { + "epoch": 1.092948717948718, + "grad_norm": 0.616348385810852, + "learning_rate": 0.00016551876470481642, + "loss": 0.9205, + "step": 6139 + }, + { + "epoch": 1.0931267806267806, + "grad_norm": 0.7151142954826355, + "learning_rate": 0.00016550818944618801, + "loss": 1.1389, + "step": 6140 + }, + { + "epoch": 1.0933048433048433, + "grad_norm": 0.6566469669342041, + "learning_rate": 0.00016549761290406275, + "loss": 0.8216, + "step": 6141 + }, + { + "epoch": 1.0934829059829059, + "grad_norm": 0.7075428366661072, + "learning_rate": 0.00016548703507864783, + "loss": 1.065, + "step": 6142 + }, + { + "epoch": 1.0936609686609686, + "grad_norm": 0.6589360237121582, + "learning_rate": 0.00016547645597015046, + "loss": 0.9899, + "step": 6143 + }, + { + "epoch": 1.0938390313390314, + "grad_norm": 0.6445585489273071, + "learning_rate": 0.00016546587557877797, + "loss": 1.1629, + "step": 6144 + }, + { + "epoch": 1.0940170940170941, + "grad_norm": 0.6216462850570679, + "learning_rate": 0.00016545529390473763, + "loss": 0.9685, + "step": 6145 + }, + { + "epoch": 1.0941951566951567, + "grad_norm": 0.6195303797721863, + "learning_rate": 0.0001654447109482368, + "loss": 1.144, + "step": 6146 + }, + { + "epoch": 1.0943732193732194, + "grad_norm": 0.6625444293022156, + "learning_rate": 0.0001654341267094828, + "loss": 0.9886, + "step": 6147 + }, + { + "epoch": 1.094551282051282, + "grad_norm": 0.6449851393699646, + "learning_rate": 0.000165423541188683, + "loss": 0.9568, + "step": 6148 + }, + { + "epoch": 1.0947293447293447, + "grad_norm": 0.6490375995635986, + "learning_rate": 0.00016541295438604484, + "loss": 1.1304, + "step": 6149 + }, + { + "epoch": 1.0949074074074074, + "grad_norm": 0.6771987676620483, + "learning_rate": 0.00016540236630177574, + "loss": 1.0426, + "step": 6150 + }, + { + "epoch": 1.0950854700854702, + "grad_norm": 0.5214568376541138, + "learning_rate": 0.00016539177693608307, + "loss": 0.6742, + "step": 6151 + }, + { + "epoch": 1.0952635327635327, + "grad_norm": 0.6005097031593323, + "learning_rate": 0.00016538118628917442, + "loss": 0.9901, + "step": 6152 + }, + { + "epoch": 1.0954415954415955, + "grad_norm": 0.6449539065361023, + "learning_rate": 0.0001653705943612572, + "loss": 0.9654, + "step": 6153 + }, + { + "epoch": 1.095619658119658, + "grad_norm": 0.6443646550178528, + "learning_rate": 0.00016536000115253903, + "loss": 0.9084, + "step": 6154 + }, + { + "epoch": 1.0957977207977208, + "grad_norm": 0.6072495579719543, + "learning_rate": 0.0001653494066632274, + "loss": 0.6308, + "step": 6155 + }, + { + "epoch": 1.0959757834757835, + "grad_norm": 0.5751157999038696, + "learning_rate": 0.00016533881089352988, + "loss": 0.96, + "step": 6156 + }, + { + "epoch": 1.0961538461538463, + "grad_norm": 0.6310713291168213, + "learning_rate": 0.0001653282138436541, + "loss": 1.0997, + "step": 6157 + }, + { + "epoch": 1.0963319088319088, + "grad_norm": 0.5573651790618896, + "learning_rate": 0.00016531761551380765, + "loss": 0.9738, + "step": 6158 + }, + { + "epoch": 1.0965099715099715, + "grad_norm": 0.5615308880805969, + "learning_rate": 0.00016530701590419824, + "loss": 0.9658, + "step": 6159 + }, + { + "epoch": 1.0966880341880343, + "grad_norm": 0.6471942663192749, + "learning_rate": 0.0001652964150150335, + "loss": 1.0763, + "step": 6160 + }, + { + "epoch": 1.0968660968660968, + "grad_norm": 0.6305427551269531, + "learning_rate": 0.00016528581284652117, + "loss": 1.112, + "step": 6161 + }, + { + "epoch": 1.0970441595441596, + "grad_norm": 0.6881145238876343, + "learning_rate": 0.00016527520939886892, + "loss": 0.8476, + "step": 6162 + }, + { + "epoch": 1.0972222222222223, + "grad_norm": 0.6507891416549683, + "learning_rate": 0.00016526460467228458, + "loss": 1.1097, + "step": 6163 + }, + { + "epoch": 1.0974002849002849, + "grad_norm": 0.5960137844085693, + "learning_rate": 0.00016525399866697586, + "loss": 0.9934, + "step": 6164 + }, + { + "epoch": 1.0975783475783476, + "grad_norm": 0.6001808643341064, + "learning_rate": 0.0001652433913831506, + "loss": 1.0782, + "step": 6165 + }, + { + "epoch": 1.0977564102564104, + "grad_norm": 0.5639005303382874, + "learning_rate": 0.00016523278282101663, + "loss": 1.0929, + "step": 6166 + }, + { + "epoch": 1.0979344729344729, + "grad_norm": 0.5962058305740356, + "learning_rate": 0.00016522217298078177, + "loss": 1.0315, + "step": 6167 + }, + { + "epoch": 1.0981125356125356, + "grad_norm": 0.6920329928398132, + "learning_rate": 0.0001652115618626539, + "loss": 0.9176, + "step": 6168 + }, + { + "epoch": 1.0982905982905984, + "grad_norm": 0.6963527202606201, + "learning_rate": 0.00016520094946684098, + "loss": 1.2136, + "step": 6169 + }, + { + "epoch": 1.098468660968661, + "grad_norm": 0.5855711102485657, + "learning_rate": 0.00016519033579355093, + "loss": 0.8453, + "step": 6170 + }, + { + "epoch": 1.0986467236467237, + "grad_norm": 0.6454927325248718, + "learning_rate": 0.0001651797208429916, + "loss": 1.0747, + "step": 6171 + }, + { + "epoch": 1.0988247863247864, + "grad_norm": 0.644585907459259, + "learning_rate": 0.00016516910461537108, + "loss": 0.8165, + "step": 6172 + }, + { + "epoch": 1.099002849002849, + "grad_norm": 0.6488069891929626, + "learning_rate": 0.00016515848711089732, + "loss": 1.1048, + "step": 6173 + }, + { + "epoch": 1.0991809116809117, + "grad_norm": 0.5867953896522522, + "learning_rate": 0.00016514786832977834, + "loss": 0.63, + "step": 6174 + }, + { + "epoch": 1.0993589743589745, + "grad_norm": 0.560591459274292, + "learning_rate": 0.00016513724827222227, + "loss": 0.9255, + "step": 6175 + }, + { + "epoch": 1.099537037037037, + "grad_norm": 0.675262451171875, + "learning_rate": 0.00016512662693843707, + "loss": 0.7637, + "step": 6176 + }, + { + "epoch": 1.0997150997150997, + "grad_norm": 0.6515669822692871, + "learning_rate": 0.00016511600432863091, + "loss": 0.7579, + "step": 6177 + }, + { + "epoch": 1.0998931623931625, + "grad_norm": 0.683409571647644, + "learning_rate": 0.00016510538044301192, + "loss": 0.9183, + "step": 6178 + }, + { + "epoch": 1.100071225071225, + "grad_norm": 0.6194507479667664, + "learning_rate": 0.00016509475528178827, + "loss": 1.16, + "step": 6179 + }, + { + "epoch": 1.1002492877492878, + "grad_norm": 0.6192209720611572, + "learning_rate": 0.0001650841288451681, + "loss": 1.1392, + "step": 6180 + }, + { + "epoch": 1.1004273504273505, + "grad_norm": 0.6029189825057983, + "learning_rate": 0.0001650735011333596, + "loss": 1.1453, + "step": 6181 + }, + { + "epoch": 1.100605413105413, + "grad_norm": 0.7040731310844421, + "learning_rate": 0.00016506287214657105, + "loss": 0.9367, + "step": 6182 + }, + { + "epoch": 1.1007834757834758, + "grad_norm": 0.5909842252731323, + "learning_rate": 0.00016505224188501067, + "loss": 0.6463, + "step": 6183 + }, + { + "epoch": 1.1009615384615385, + "grad_norm": 0.6129698157310486, + "learning_rate": 0.00016504161034888674, + "loss": 0.9432, + "step": 6184 + }, + { + "epoch": 1.101139601139601, + "grad_norm": 0.6181607842445374, + "learning_rate": 0.00016503097753840757, + "loss": 0.9934, + "step": 6185 + }, + { + "epoch": 1.1013176638176638, + "grad_norm": 0.6463226675987244, + "learning_rate": 0.0001650203434537815, + "loss": 0.8471, + "step": 6186 + }, + { + "epoch": 1.1014957264957266, + "grad_norm": 0.5999348163604736, + "learning_rate": 0.00016500970809521688, + "loss": 0.9418, + "step": 6187 + }, + { + "epoch": 1.101673789173789, + "grad_norm": 0.629504919052124, + "learning_rate": 0.00016499907146292204, + "loss": 0.9699, + "step": 6188 + }, + { + "epoch": 1.1018518518518519, + "grad_norm": 0.694767951965332, + "learning_rate": 0.00016498843355710542, + "loss": 0.8793, + "step": 6189 + }, + { + "epoch": 1.1020299145299146, + "grad_norm": 0.6205509901046753, + "learning_rate": 0.00016497779437797547, + "loss": 0.8384, + "step": 6190 + }, + { + "epoch": 1.1022079772079771, + "grad_norm": 0.6256579756736755, + "learning_rate": 0.0001649671539257406, + "loss": 0.9275, + "step": 6191 + }, + { + "epoch": 1.10238603988604, + "grad_norm": 0.6593793034553528, + "learning_rate": 0.00016495651220060933, + "loss": 1.0495, + "step": 6192 + }, + { + "epoch": 1.1025641025641026, + "grad_norm": 0.7809221148490906, + "learning_rate": 0.00016494586920279012, + "loss": 1.0485, + "step": 6193 + }, + { + "epoch": 1.1027421652421652, + "grad_norm": 0.6147717833518982, + "learning_rate": 0.0001649352249324915, + "loss": 0.8739, + "step": 6194 + }, + { + "epoch": 1.102920227920228, + "grad_norm": 0.565411388874054, + "learning_rate": 0.00016492457938992208, + "loss": 0.9759, + "step": 6195 + }, + { + "epoch": 1.1030982905982907, + "grad_norm": 0.596370279788971, + "learning_rate": 0.00016491393257529036, + "loss": 0.9658, + "step": 6196 + }, + { + "epoch": 1.1032763532763532, + "grad_norm": 0.6334326863288879, + "learning_rate": 0.00016490328448880498, + "loss": 0.8785, + "step": 6197 + }, + { + "epoch": 1.103454415954416, + "grad_norm": 0.5538334846496582, + "learning_rate": 0.0001648926351306746, + "loss": 0.7174, + "step": 6198 + }, + { + "epoch": 1.1036324786324787, + "grad_norm": 0.6249658465385437, + "learning_rate": 0.00016488198450110778, + "loss": 0.8579, + "step": 6199 + }, + { + "epoch": 1.1038105413105412, + "grad_norm": 0.6128895878791809, + "learning_rate": 0.00016487133260031329, + "loss": 0.8538, + "step": 6200 + }, + { + "epoch": 1.103988603988604, + "grad_norm": 0.5808702707290649, + "learning_rate": 0.0001648606794284998, + "loss": 0.8143, + "step": 6201 + }, + { + "epoch": 1.1041666666666667, + "grad_norm": 0.671419084072113, + "learning_rate": 0.00016485002498587602, + "loss": 1.1268, + "step": 6202 + }, + { + "epoch": 1.1043447293447293, + "grad_norm": 0.5706788897514343, + "learning_rate": 0.00016483936927265075, + "loss": 0.9558, + "step": 6203 + }, + { + "epoch": 1.104522792022792, + "grad_norm": 0.5700307488441467, + "learning_rate": 0.00016482871228903266, + "loss": 0.9616, + "step": 6204 + }, + { + "epoch": 1.1047008547008548, + "grad_norm": 0.5764816403388977, + "learning_rate": 0.0001648180540352307, + "loss": 0.8692, + "step": 6205 + }, + { + "epoch": 1.1048789173789173, + "grad_norm": 0.5786563754081726, + "learning_rate": 0.00016480739451145358, + "loss": 0.9406, + "step": 6206 + }, + { + "epoch": 1.10505698005698, + "grad_norm": 0.6112591624259949, + "learning_rate": 0.0001647967337179102, + "loss": 0.8999, + "step": 6207 + }, + { + "epoch": 1.1052350427350428, + "grad_norm": 0.5708907246589661, + "learning_rate": 0.00016478607165480944, + "loss": 0.9236, + "step": 6208 + }, + { + "epoch": 1.1054131054131053, + "grad_norm": 0.6742013692855835, + "learning_rate": 0.00016477540832236014, + "loss": 1.0911, + "step": 6209 + }, + { + "epoch": 1.105591168091168, + "grad_norm": 0.6382617354393005, + "learning_rate": 0.0001647647437207713, + "loss": 0.7901, + "step": 6210 + }, + { + "epoch": 1.1057692307692308, + "grad_norm": 0.6241547465324402, + "learning_rate": 0.00016475407785025188, + "loss": 1.0048, + "step": 6211 + }, + { + "epoch": 1.1059472934472934, + "grad_norm": 0.6452877521514893, + "learning_rate": 0.00016474341071101077, + "loss": 0.8902, + "step": 6212 + }, + { + "epoch": 1.1061253561253561, + "grad_norm": 0.6212326288223267, + "learning_rate": 0.00016473274230325704, + "loss": 1.078, + "step": 6213 + }, + { + "epoch": 1.1063034188034189, + "grad_norm": 0.6870912909507751, + "learning_rate": 0.00016472207262719968, + "loss": 0.9127, + "step": 6214 + }, + { + "epoch": 1.1064814814814814, + "grad_norm": 0.6286750435829163, + "learning_rate": 0.00016471140168304777, + "loss": 1.0271, + "step": 6215 + }, + { + "epoch": 1.1066595441595442, + "grad_norm": 0.645806074142456, + "learning_rate": 0.00016470072947101036, + "loss": 1.1514, + "step": 6216 + }, + { + "epoch": 1.106837606837607, + "grad_norm": 0.6800320148468018, + "learning_rate": 0.00016469005599129653, + "loss": 0.9322, + "step": 6217 + }, + { + "epoch": 1.1070156695156694, + "grad_norm": 0.5898309946060181, + "learning_rate": 0.0001646793812441155, + "loss": 1.065, + "step": 6218 + }, + { + "epoch": 1.1071937321937322, + "grad_norm": 0.6000019907951355, + "learning_rate": 0.00016466870522967634, + "loss": 0.911, + "step": 6219 + }, + { + "epoch": 1.107371794871795, + "grad_norm": 0.6164331436157227, + "learning_rate": 0.0001646580279481882, + "loss": 0.8421, + "step": 6220 + }, + { + "epoch": 1.1075498575498575, + "grad_norm": 0.6410242319107056, + "learning_rate": 0.00016464734939986036, + "loss": 0.9688, + "step": 6221 + }, + { + "epoch": 1.1077279202279202, + "grad_norm": 0.7153300046920776, + "learning_rate": 0.00016463666958490197, + "loss": 1.0722, + "step": 6222 + }, + { + "epoch": 1.107905982905983, + "grad_norm": 0.6977026462554932, + "learning_rate": 0.00016462598850352234, + "loss": 1.0192, + "step": 6223 + }, + { + "epoch": 1.1080840455840455, + "grad_norm": 0.6379461884498596, + "learning_rate": 0.0001646153061559307, + "loss": 1.0474, + "step": 6224 + }, + { + "epoch": 1.1082621082621082, + "grad_norm": 0.6135090589523315, + "learning_rate": 0.00016460462254233634, + "loss": 1.0082, + "step": 6225 + }, + { + "epoch": 1.108440170940171, + "grad_norm": 0.6326230764389038, + "learning_rate": 0.00016459393766294866, + "loss": 1.1097, + "step": 6226 + }, + { + "epoch": 1.1086182336182335, + "grad_norm": 0.6636839509010315, + "learning_rate": 0.0001645832515179769, + "loss": 0.9689, + "step": 6227 + }, + { + "epoch": 1.1087962962962963, + "grad_norm": 0.5713129043579102, + "learning_rate": 0.00016457256410763052, + "loss": 0.8642, + "step": 6228 + }, + { + "epoch": 1.108974358974359, + "grad_norm": 0.584204912185669, + "learning_rate": 0.00016456187543211888, + "loss": 0.9957, + "step": 6229 + }, + { + "epoch": 1.1091524216524216, + "grad_norm": 0.5920230746269226, + "learning_rate": 0.0001645511854916514, + "loss": 0.7297, + "step": 6230 + }, + { + "epoch": 1.1093304843304843, + "grad_norm": 0.6207385063171387, + "learning_rate": 0.0001645404942864375, + "loss": 0.868, + "step": 6231 + }, + { + "epoch": 1.109508547008547, + "grad_norm": 0.7267234921455383, + "learning_rate": 0.00016452980181668673, + "loss": 1.0248, + "step": 6232 + }, + { + "epoch": 1.1096866096866096, + "grad_norm": 0.5925650596618652, + "learning_rate": 0.00016451910808260852, + "loss": 1.1075, + "step": 6233 + }, + { + "epoch": 1.1098646723646723, + "grad_norm": 0.5632196664810181, + "learning_rate": 0.00016450841308441244, + "loss": 0.9865, + "step": 6234 + }, + { + "epoch": 1.110042735042735, + "grad_norm": 0.6115161180496216, + "learning_rate": 0.000164497716822308, + "loss": 1.1343, + "step": 6235 + }, + { + "epoch": 1.1102207977207976, + "grad_norm": 0.634398341178894, + "learning_rate": 0.00016448701929650477, + "loss": 1.1039, + "step": 6236 + }, + { + "epoch": 1.1103988603988604, + "grad_norm": 0.5843468308448792, + "learning_rate": 0.00016447632050721237, + "loss": 0.8462, + "step": 6237 + }, + { + "epoch": 1.1105769230769231, + "grad_norm": 0.799375593662262, + "learning_rate": 0.0001644656204546404, + "loss": 0.9861, + "step": 6238 + }, + { + "epoch": 1.1107549857549857, + "grad_norm": 0.600289523601532, + "learning_rate": 0.0001644549191389985, + "loss": 1.0323, + "step": 6239 + }, + { + "epoch": 1.1109330484330484, + "grad_norm": 0.6154919266700745, + "learning_rate": 0.00016444421656049637, + "loss": 0.9158, + "step": 6240 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 0.6685689687728882, + "learning_rate": 0.00016443351271934367, + "loss": 1.0429, + "step": 6241 + }, + { + "epoch": 1.1112891737891737, + "grad_norm": 0.699978232383728, + "learning_rate": 0.00016442280761575016, + "loss": 1.072, + "step": 6242 + }, + { + "epoch": 1.1114672364672364, + "grad_norm": 0.6461396217346191, + "learning_rate": 0.00016441210124992556, + "loss": 0.9758, + "step": 6243 + }, + { + "epoch": 1.1116452991452992, + "grad_norm": 0.6463284492492676, + "learning_rate": 0.00016440139362207962, + "loss": 0.9205, + "step": 6244 + }, + { + "epoch": 1.1118233618233617, + "grad_norm": 0.6587556004524231, + "learning_rate": 0.00016439068473242217, + "loss": 1.0027, + "step": 6245 + }, + { + "epoch": 1.1120014245014245, + "grad_norm": 0.6896520256996155, + "learning_rate": 0.000164379974581163, + "loss": 0.9788, + "step": 6246 + }, + { + "epoch": 1.1121794871794872, + "grad_norm": 0.6766142845153809, + "learning_rate": 0.000164369263168512, + "loss": 0.9647, + "step": 6247 + }, + { + "epoch": 1.1123575498575498, + "grad_norm": 0.7024297118186951, + "learning_rate": 0.00016435855049467898, + "loss": 1.1163, + "step": 6248 + }, + { + "epoch": 1.1125356125356125, + "grad_norm": 0.6654963493347168, + "learning_rate": 0.00016434783655987385, + "loss": 0.9302, + "step": 6249 + }, + { + "epoch": 1.1127136752136753, + "grad_norm": 0.6973692774772644, + "learning_rate": 0.0001643371213643065, + "loss": 0.9585, + "step": 6250 + }, + { + "epoch": 1.1128917378917378, + "grad_norm": 0.7153545022010803, + "learning_rate": 0.000164326404908187, + "loss": 1.0485, + "step": 6251 + }, + { + "epoch": 1.1130698005698005, + "grad_norm": 0.6114685535430908, + "learning_rate": 0.00016431568719172516, + "loss": 0.8881, + "step": 6252 + }, + { + "epoch": 1.1132478632478633, + "grad_norm": 0.6500731706619263, + "learning_rate": 0.00016430496821513103, + "loss": 1.0658, + "step": 6253 + }, + { + "epoch": 1.113425925925926, + "grad_norm": 0.5800092220306396, + "learning_rate": 0.00016429424797861466, + "loss": 0.9158, + "step": 6254 + }, + { + "epoch": 1.1136039886039886, + "grad_norm": 0.6653759479522705, + "learning_rate": 0.00016428352648238602, + "loss": 0.9762, + "step": 6255 + }, + { + "epoch": 1.1137820512820513, + "grad_norm": 0.649208128452301, + "learning_rate": 0.00016427280372665525, + "loss": 1.1184, + "step": 6256 + }, + { + "epoch": 1.1139601139601139, + "grad_norm": 0.6665199398994446, + "learning_rate": 0.00016426207971163238, + "loss": 0.9417, + "step": 6257 + }, + { + "epoch": 1.1141381766381766, + "grad_norm": 0.6110978126525879, + "learning_rate": 0.00016425135443752758, + "loss": 1.1531, + "step": 6258 + }, + { + "epoch": 1.1143162393162394, + "grad_norm": 0.6517077088356018, + "learning_rate": 0.00016424062790455093, + "loss": 0.9055, + "step": 6259 + }, + { + "epoch": 1.114494301994302, + "grad_norm": 0.6278966665267944, + "learning_rate": 0.00016422990011291265, + "loss": 1.0087, + "step": 6260 + }, + { + "epoch": 1.1146723646723646, + "grad_norm": 0.5818809270858765, + "learning_rate": 0.00016421917106282288, + "loss": 1.0202, + "step": 6261 + }, + { + "epoch": 1.1148504273504274, + "grad_norm": 0.5670005679130554, + "learning_rate": 0.00016420844075449187, + "loss": 0.841, + "step": 6262 + }, + { + "epoch": 1.11502849002849, + "grad_norm": 0.6584762334823608, + "learning_rate": 0.00016419770918812984, + "loss": 1.0322, + "step": 6263 + }, + { + "epoch": 1.1152065527065527, + "grad_norm": 0.6023790836334229, + "learning_rate": 0.00016418697636394705, + "loss": 0.9152, + "step": 6264 + }, + { + "epoch": 1.1153846153846154, + "grad_norm": 0.6234691739082336, + "learning_rate": 0.00016417624228215382, + "loss": 0.9555, + "step": 6265 + }, + { + "epoch": 1.1155626780626782, + "grad_norm": 0.6690816879272461, + "learning_rate": 0.00016416550694296045, + "loss": 0.9341, + "step": 6266 + }, + { + "epoch": 1.1157407407407407, + "grad_norm": 0.6030237078666687, + "learning_rate": 0.00016415477034657723, + "loss": 1.0442, + "step": 6267 + }, + { + "epoch": 1.1159188034188035, + "grad_norm": 0.5954633951187134, + "learning_rate": 0.00016414403249321455, + "loss": 0.9132, + "step": 6268 + }, + { + "epoch": 1.116096866096866, + "grad_norm": 0.7876830101013184, + "learning_rate": 0.0001641332933830828, + "loss": 0.9456, + "step": 6269 + }, + { + "epoch": 1.1162749287749287, + "grad_norm": 0.6776009798049927, + "learning_rate": 0.00016412255301639244, + "loss": 0.9022, + "step": 6270 + }, + { + "epoch": 1.1164529914529915, + "grad_norm": 0.6094426512718201, + "learning_rate": 0.0001641118113933538, + "loss": 0.9629, + "step": 6271 + }, + { + "epoch": 1.1166310541310542, + "grad_norm": 0.5818213820457458, + "learning_rate": 0.00016410106851417742, + "loss": 0.9049, + "step": 6272 + }, + { + "epoch": 1.1168091168091168, + "grad_norm": 0.5668078064918518, + "learning_rate": 0.00016409032437907377, + "loss": 1.0011, + "step": 6273 + }, + { + "epoch": 1.1169871794871795, + "grad_norm": 0.6984922289848328, + "learning_rate": 0.00016407957898825334, + "loss": 0.9454, + "step": 6274 + }, + { + "epoch": 1.1171652421652423, + "grad_norm": 0.5509830117225647, + "learning_rate": 0.00016406883234192668, + "loss": 0.9132, + "step": 6275 + }, + { + "epoch": 1.1173433048433048, + "grad_norm": 0.5117461681365967, + "learning_rate": 0.00016405808444030435, + "loss": 0.7675, + "step": 6276 + }, + { + "epoch": 1.1175213675213675, + "grad_norm": 0.6358339786529541, + "learning_rate": 0.00016404733528359688, + "loss": 0.9777, + "step": 6277 + }, + { + "epoch": 1.1176994301994303, + "grad_norm": 0.5870591402053833, + "learning_rate": 0.00016403658487201494, + "loss": 0.8576, + "step": 6278 + }, + { + "epoch": 1.1178774928774928, + "grad_norm": 0.6532407999038696, + "learning_rate": 0.00016402583320576915, + "loss": 1.1787, + "step": 6279 + }, + { + "epoch": 1.1180555555555556, + "grad_norm": 0.6374639272689819, + "learning_rate": 0.00016401508028507017, + "loss": 0.9298, + "step": 6280 + }, + { + "epoch": 1.1182336182336183, + "grad_norm": 0.7280316352844238, + "learning_rate": 0.00016400432611012869, + "loss": 1.1081, + "step": 6281 + }, + { + "epoch": 1.1184116809116809, + "grad_norm": 0.6070699095726013, + "learning_rate": 0.00016399357068115538, + "loss": 0.9107, + "step": 6282 + }, + { + "epoch": 1.1185897435897436, + "grad_norm": 0.6701489686965942, + "learning_rate": 0.00016398281399836097, + "loss": 1.0879, + "step": 6283 + }, + { + "epoch": 1.1187678062678064, + "grad_norm": 0.6343162655830383, + "learning_rate": 0.00016397205606195626, + "loss": 0.8552, + "step": 6284 + }, + { + "epoch": 1.118945868945869, + "grad_norm": 0.6450608968734741, + "learning_rate": 0.00016396129687215198, + "loss": 1.1119, + "step": 6285 + }, + { + "epoch": 1.1191239316239316, + "grad_norm": 0.7219904661178589, + "learning_rate": 0.00016395053642915896, + "loss": 0.9081, + "step": 6286 + }, + { + "epoch": 1.1193019943019944, + "grad_norm": 0.6189733147621155, + "learning_rate": 0.00016393977473318802, + "loss": 0.9818, + "step": 6287 + }, + { + "epoch": 1.119480056980057, + "grad_norm": 0.6310907602310181, + "learning_rate": 0.00016392901178445004, + "loss": 1.0334, + "step": 6288 + }, + { + "epoch": 1.1196581196581197, + "grad_norm": 0.6556720733642578, + "learning_rate": 0.00016391824758315587, + "loss": 1.0452, + "step": 6289 + }, + { + "epoch": 1.1198361823361824, + "grad_norm": 0.6697782278060913, + "learning_rate": 0.00016390748212951638, + "loss": 0.9627, + "step": 6290 + }, + { + "epoch": 1.120014245014245, + "grad_norm": 0.6341549754142761, + "learning_rate": 0.00016389671542374256, + "loss": 1.112, + "step": 6291 + }, + { + "epoch": 1.1201923076923077, + "grad_norm": 0.6913946270942688, + "learning_rate": 0.00016388594746604535, + "loss": 0.9622, + "step": 6292 + }, + { + "epoch": 1.1203703703703705, + "grad_norm": 0.695488691329956, + "learning_rate": 0.0001638751782566357, + "loss": 1.0951, + "step": 6293 + }, + { + "epoch": 1.120548433048433, + "grad_norm": 0.6965359449386597, + "learning_rate": 0.00016386440779572463, + "loss": 1.1742, + "step": 6294 + }, + { + "epoch": 1.1207264957264957, + "grad_norm": 0.624679684638977, + "learning_rate": 0.00016385363608352314, + "loss": 0.9756, + "step": 6295 + }, + { + "epoch": 1.1209045584045585, + "grad_norm": 0.7511318922042847, + "learning_rate": 0.0001638428631202423, + "loss": 0.907, + "step": 6296 + }, + { + "epoch": 1.121082621082621, + "grad_norm": 0.5334641337394714, + "learning_rate": 0.00016383208890609317, + "loss": 0.7932, + "step": 6297 + }, + { + "epoch": 1.1212606837606838, + "grad_norm": 0.7518552541732788, + "learning_rate": 0.00016382131344128687, + "loss": 1.1556, + "step": 6298 + }, + { + "epoch": 1.1214387464387465, + "grad_norm": 0.618618369102478, + "learning_rate": 0.00016381053672603449, + "loss": 1.1027, + "step": 6299 + }, + { + "epoch": 1.121616809116809, + "grad_norm": 0.638956606388092, + "learning_rate": 0.00016379975876054724, + "loss": 1.0377, + "step": 6300 + }, + { + "epoch": 1.1217948717948718, + "grad_norm": 0.8031370639801025, + "learning_rate": 0.0001637889795450362, + "loss": 1.0821, + "step": 6301 + }, + { + "epoch": 1.1219729344729346, + "grad_norm": 0.6710168123245239, + "learning_rate": 0.00016377819907971265, + "loss": 1.2896, + "step": 6302 + }, + { + "epoch": 1.122150997150997, + "grad_norm": 0.5850739479064941, + "learning_rate": 0.00016376741736478777, + "loss": 1.0836, + "step": 6303 + }, + { + "epoch": 1.1223290598290598, + "grad_norm": 0.6410611271858215, + "learning_rate": 0.0001637566344004728, + "loss": 1.0395, + "step": 6304 + }, + { + "epoch": 1.1225071225071226, + "grad_norm": 0.6884660720825195, + "learning_rate": 0.00016374585018697903, + "loss": 0.871, + "step": 6305 + }, + { + "epoch": 1.1226851851851851, + "grad_norm": 0.622207522392273, + "learning_rate": 0.00016373506472451777, + "loss": 0.9897, + "step": 6306 + }, + { + "epoch": 1.1228632478632479, + "grad_norm": 0.6018275618553162, + "learning_rate": 0.00016372427801330028, + "loss": 0.8398, + "step": 6307 + }, + { + "epoch": 1.1230413105413106, + "grad_norm": 0.6451539993286133, + "learning_rate": 0.00016371349005353796, + "loss": 0.9878, + "step": 6308 + }, + { + "epoch": 1.1232193732193732, + "grad_norm": 0.5549424886703491, + "learning_rate": 0.00016370270084544215, + "loss": 0.844, + "step": 6309 + }, + { + "epoch": 1.123397435897436, + "grad_norm": 0.6082940697669983, + "learning_rate": 0.00016369191038922423, + "loss": 1.0704, + "step": 6310 + }, + { + "epoch": 1.1235754985754987, + "grad_norm": 0.6423100829124451, + "learning_rate": 0.00016368111868509563, + "loss": 1.0639, + "step": 6311 + }, + { + "epoch": 1.1237535612535612, + "grad_norm": 0.6274200081825256, + "learning_rate": 0.00016367032573326784, + "loss": 0.9996, + "step": 6312 + }, + { + "epoch": 1.123931623931624, + "grad_norm": 0.6618558168411255, + "learning_rate": 0.00016365953153395227, + "loss": 0.8074, + "step": 6313 + }, + { + "epoch": 1.1241096866096867, + "grad_norm": 0.7624069452285767, + "learning_rate": 0.00016364873608736038, + "loss": 0.9741, + "step": 6314 + }, + { + "epoch": 1.1242877492877492, + "grad_norm": 0.5391361117362976, + "learning_rate": 0.00016363793939370375, + "loss": 0.6992, + "step": 6315 + }, + { + "epoch": 1.124465811965812, + "grad_norm": 0.7564396858215332, + "learning_rate": 0.0001636271414531939, + "loss": 1.1971, + "step": 6316 + }, + { + "epoch": 1.1246438746438747, + "grad_norm": 0.6584066152572632, + "learning_rate": 0.00016361634226604239, + "loss": 1.0842, + "step": 6317 + }, + { + "epoch": 1.1248219373219372, + "grad_norm": 0.6851227283477783, + "learning_rate": 0.00016360554183246078, + "loss": 1.0879, + "step": 6318 + }, + { + "epoch": 1.125, + "grad_norm": 0.5699417591094971, + "learning_rate": 0.00016359474015266074, + "loss": 0.782, + "step": 6319 + }, + { + "epoch": 1.1251780626780628, + "grad_norm": 0.5495570302009583, + "learning_rate": 0.00016358393722685385, + "loss": 1.076, + "step": 6320 + }, + { + "epoch": 1.1253561253561253, + "grad_norm": 0.5872206091880798, + "learning_rate": 0.0001635731330552518, + "loss": 0.8601, + "step": 6321 + }, + { + "epoch": 1.125534188034188, + "grad_norm": 0.7012827396392822, + "learning_rate": 0.00016356232763806627, + "loss": 1.0443, + "step": 6322 + }, + { + "epoch": 1.1257122507122508, + "grad_norm": 0.6645881533622742, + "learning_rate": 0.00016355152097550897, + "loss": 1.0027, + "step": 6323 + }, + { + "epoch": 1.1258903133903133, + "grad_norm": 0.7376120090484619, + "learning_rate": 0.00016354071306779163, + "loss": 1.1941, + "step": 6324 + }, + { + "epoch": 1.126068376068376, + "grad_norm": 0.648932695388794, + "learning_rate": 0.000163529903915126, + "loss": 1.096, + "step": 6325 + }, + { + "epoch": 1.1262464387464388, + "grad_norm": 0.6186314821243286, + "learning_rate": 0.0001635190935177239, + "loss": 1.011, + "step": 6326 + }, + { + "epoch": 1.1264245014245013, + "grad_norm": 0.5964710116386414, + "learning_rate": 0.0001635082818757971, + "loss": 0.8893, + "step": 6327 + }, + { + "epoch": 1.126602564102564, + "grad_norm": 0.5264934301376343, + "learning_rate": 0.00016349746898955747, + "loss": 0.7325, + "step": 6328 + }, + { + "epoch": 1.1267806267806268, + "grad_norm": 0.6523048877716064, + "learning_rate": 0.00016348665485921678, + "loss": 1.0488, + "step": 6329 + }, + { + "epoch": 1.1269586894586894, + "grad_norm": 0.6878600120544434, + "learning_rate": 0.00016347583948498703, + "loss": 1.0926, + "step": 6330 + }, + { + "epoch": 1.1271367521367521, + "grad_norm": 0.592656672000885, + "learning_rate": 0.00016346502286708004, + "loss": 0.978, + "step": 6331 + }, + { + "epoch": 1.1273148148148149, + "grad_norm": 0.6338315606117249, + "learning_rate": 0.00016345420500570777, + "loss": 1.1048, + "step": 6332 + }, + { + "epoch": 1.1274928774928774, + "grad_norm": 0.5955204367637634, + "learning_rate": 0.00016344338590108218, + "loss": 0.88, + "step": 6333 + }, + { + "epoch": 1.1276709401709402, + "grad_norm": 0.690448522567749, + "learning_rate": 0.0001634325655534152, + "loss": 1.0564, + "step": 6334 + }, + { + "epoch": 1.127849002849003, + "grad_norm": 0.6125795841217041, + "learning_rate": 0.00016342174396291888, + "loss": 1.0608, + "step": 6335 + }, + { + "epoch": 1.1280270655270654, + "grad_norm": 0.6387807726860046, + "learning_rate": 0.00016341092112980523, + "loss": 0.9581, + "step": 6336 + }, + { + "epoch": 1.1282051282051282, + "grad_norm": 0.6247823238372803, + "learning_rate": 0.0001634000970542863, + "loss": 0.932, + "step": 6337 + }, + { + "epoch": 1.128383190883191, + "grad_norm": 0.5928077697753906, + "learning_rate": 0.0001633892717365742, + "loss": 0.8963, + "step": 6338 + }, + { + "epoch": 1.1285612535612535, + "grad_norm": 0.5922074913978577, + "learning_rate": 0.000163378445176881, + "loss": 0.9772, + "step": 6339 + }, + { + "epoch": 1.1287393162393162, + "grad_norm": 0.6573056578636169, + "learning_rate": 0.00016336761737541878, + "loss": 0.8233, + "step": 6340 + }, + { + "epoch": 1.128917378917379, + "grad_norm": 0.627772867679596, + "learning_rate": 0.0001633567883323998, + "loss": 0.9618, + "step": 6341 + }, + { + "epoch": 1.1290954415954415, + "grad_norm": 0.6066579818725586, + "learning_rate": 0.0001633459580480361, + "loss": 0.9066, + "step": 6342 + }, + { + "epoch": 1.1292735042735043, + "grad_norm": 0.670295000076294, + "learning_rate": 0.00016333512652253997, + "loss": 0.8003, + "step": 6343 + }, + { + "epoch": 1.129451566951567, + "grad_norm": 0.6402488946914673, + "learning_rate": 0.0001633242937561236, + "loss": 0.998, + "step": 6344 + }, + { + "epoch": 1.1296296296296295, + "grad_norm": 0.7224995493888855, + "learning_rate": 0.00016331345974899923, + "loss": 1.0308, + "step": 6345 + }, + { + "epoch": 1.1298076923076923, + "grad_norm": 0.5019716620445251, + "learning_rate": 0.00016330262450137917, + "loss": 0.6874, + "step": 6346 + }, + { + "epoch": 1.129985754985755, + "grad_norm": 0.5774167776107788, + "learning_rate": 0.00016329178801347566, + "loss": 0.8287, + "step": 6347 + }, + { + "epoch": 1.1301638176638176, + "grad_norm": 0.7797795534133911, + "learning_rate": 0.00016328095028550103, + "loss": 1.2145, + "step": 6348 + }, + { + "epoch": 1.1303418803418803, + "grad_norm": 0.5384017825126648, + "learning_rate": 0.00016327011131766765, + "loss": 0.8022, + "step": 6349 + }, + { + "epoch": 1.130519943019943, + "grad_norm": 0.6350888609886169, + "learning_rate": 0.00016325927111018786, + "loss": 1.1178, + "step": 6350 + }, + { + "epoch": 1.1306980056980056, + "grad_norm": 0.6386831998825073, + "learning_rate": 0.0001632484296632741, + "loss": 0.967, + "step": 6351 + }, + { + "epoch": 1.1308760683760684, + "grad_norm": 0.6214167475700378, + "learning_rate": 0.0001632375869771387, + "loss": 0.9416, + "step": 6352 + }, + { + "epoch": 1.131054131054131, + "grad_norm": 0.6145567297935486, + "learning_rate": 0.00016322674305199416, + "loss": 0.9175, + "step": 6353 + }, + { + "epoch": 1.1312321937321936, + "grad_norm": 0.7027857303619385, + "learning_rate": 0.00016321589788805297, + "loss": 1.0063, + "step": 6354 + }, + { + "epoch": 1.1314102564102564, + "grad_norm": 0.6942669153213501, + "learning_rate": 0.00016320505148552755, + "loss": 0.9191, + "step": 6355 + }, + { + "epoch": 1.1315883190883191, + "grad_norm": 0.6388658285140991, + "learning_rate": 0.0001631942038446304, + "loss": 0.993, + "step": 6356 + }, + { + "epoch": 1.131766381766382, + "grad_norm": 0.6627292633056641, + "learning_rate": 0.00016318335496557415, + "loss": 1.0055, + "step": 6357 + }, + { + "epoch": 1.1319444444444444, + "grad_norm": 0.7997342944145203, + "learning_rate": 0.0001631725048485713, + "loss": 0.9019, + "step": 6358 + }, + { + "epoch": 1.1321225071225072, + "grad_norm": 0.8817830681800842, + "learning_rate": 0.00016316165349383445, + "loss": 0.9793, + "step": 6359 + }, + { + "epoch": 1.1323005698005697, + "grad_norm": 0.5629408955574036, + "learning_rate": 0.00016315080090157621, + "loss": 0.6139, + "step": 6360 + }, + { + "epoch": 1.1324786324786325, + "grad_norm": 0.647220253944397, + "learning_rate": 0.0001631399470720092, + "loss": 0.9776, + "step": 6361 + }, + { + "epoch": 1.1326566951566952, + "grad_norm": 0.6762630939483643, + "learning_rate": 0.0001631290920053461, + "loss": 1.1027, + "step": 6362 + }, + { + "epoch": 1.132834757834758, + "grad_norm": 0.5862727761268616, + "learning_rate": 0.00016311823570179957, + "loss": 1.1359, + "step": 6363 + }, + { + "epoch": 1.1330128205128205, + "grad_norm": 0.7042981386184692, + "learning_rate": 0.00016310737816158235, + "loss": 1.142, + "step": 6364 + }, + { + "epoch": 1.1331908831908832, + "grad_norm": 0.5990639328956604, + "learning_rate": 0.00016309651938490712, + "loss": 0.9306, + "step": 6365 + }, + { + "epoch": 1.1333689458689458, + "grad_norm": 0.5894871950149536, + "learning_rate": 0.00016308565937198669, + "loss": 0.8343, + "step": 6366 + }, + { + "epoch": 1.1335470085470085, + "grad_norm": 0.6863628029823303, + "learning_rate": 0.0001630747981230338, + "loss": 0.9552, + "step": 6367 + }, + { + "epoch": 1.1337250712250713, + "grad_norm": 0.7438958287239075, + "learning_rate": 0.00016306393563826128, + "loss": 1.0422, + "step": 6368 + }, + { + "epoch": 1.133903133903134, + "grad_norm": 0.5695775747299194, + "learning_rate": 0.00016305307191788194, + "loss": 0.8633, + "step": 6369 + }, + { + "epoch": 1.1340811965811965, + "grad_norm": 0.6257741451263428, + "learning_rate": 0.00016304220696210863, + "loss": 1.0333, + "step": 6370 + }, + { + "epoch": 1.1342592592592593, + "grad_norm": 0.6366072297096252, + "learning_rate": 0.00016303134077115425, + "loss": 1.1452, + "step": 6371 + }, + { + "epoch": 1.1344373219373218, + "grad_norm": 0.624569296836853, + "learning_rate": 0.00016302047334523168, + "loss": 1.0569, + "step": 6372 + }, + { + "epoch": 1.1346153846153846, + "grad_norm": 0.5585938096046448, + "learning_rate": 0.00016300960468455382, + "loss": 0.9612, + "step": 6373 + }, + { + "epoch": 1.1347934472934473, + "grad_norm": 0.5738831162452698, + "learning_rate": 0.00016299873478933368, + "loss": 0.9206, + "step": 6374 + }, + { + "epoch": 1.13497150997151, + "grad_norm": 0.6797143220901489, + "learning_rate": 0.00016298786365978417, + "loss": 1.0748, + "step": 6375 + }, + { + "epoch": 1.1351495726495726, + "grad_norm": 0.6341326832771301, + "learning_rate": 0.00016297699129611833, + "loss": 0.9901, + "step": 6376 + }, + { + "epoch": 1.1353276353276354, + "grad_norm": 0.6568490862846375, + "learning_rate": 0.00016296611769854916, + "loss": 1.0598, + "step": 6377 + }, + { + "epoch": 1.135505698005698, + "grad_norm": 0.6151928901672363, + "learning_rate": 0.00016295524286728973, + "loss": 0.8352, + "step": 6378 + }, + { + "epoch": 1.1356837606837606, + "grad_norm": 0.7209593057632446, + "learning_rate": 0.0001629443668025531, + "loss": 0.9945, + "step": 6379 + }, + { + "epoch": 1.1358618233618234, + "grad_norm": 0.6600689888000488, + "learning_rate": 0.00016293348950455235, + "loss": 1.0572, + "step": 6380 + }, + { + "epoch": 1.1360398860398861, + "grad_norm": 0.5587523579597473, + "learning_rate": 0.0001629226109735006, + "loss": 0.8526, + "step": 6381 + }, + { + "epoch": 1.1362179487179487, + "grad_norm": 0.6184542775154114, + "learning_rate": 0.00016291173120961102, + "loss": 0.8246, + "step": 6382 + }, + { + "epoch": 1.1363960113960114, + "grad_norm": 0.6604713797569275, + "learning_rate": 0.00016290085021309673, + "loss": 1.0349, + "step": 6383 + }, + { + "epoch": 1.136574074074074, + "grad_norm": 0.5880835056304932, + "learning_rate": 0.00016288996798417097, + "loss": 0.8726, + "step": 6384 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 0.5770880579948425, + "learning_rate": 0.00016287908452304692, + "loss": 0.7639, + "step": 6385 + }, + { + "epoch": 1.1369301994301995, + "grad_norm": 0.5719713568687439, + "learning_rate": 0.00016286819982993782, + "loss": 0.9717, + "step": 6386 + }, + { + "epoch": 1.1371082621082622, + "grad_norm": 0.7028461694717407, + "learning_rate": 0.00016285731390505695, + "loss": 1.0147, + "step": 6387 + }, + { + "epoch": 1.1372863247863247, + "grad_norm": 0.5396828651428223, + "learning_rate": 0.00016284642674861756, + "loss": 0.8119, + "step": 6388 + }, + { + "epoch": 1.1374643874643875, + "grad_norm": 0.592580258846283, + "learning_rate": 0.00016283553836083303, + "loss": 1.0914, + "step": 6389 + }, + { + "epoch": 1.13764245014245, + "grad_norm": 0.634596586227417, + "learning_rate": 0.00016282464874191663, + "loss": 1.1037, + "step": 6390 + }, + { + "epoch": 1.1378205128205128, + "grad_norm": 0.6462705731391907, + "learning_rate": 0.00016281375789208176, + "loss": 1.1523, + "step": 6391 + }, + { + "epoch": 1.1379985754985755, + "grad_norm": 0.6527917385101318, + "learning_rate": 0.0001628028658115418, + "loss": 1.0415, + "step": 6392 + }, + { + "epoch": 1.1381766381766383, + "grad_norm": 0.6309964060783386, + "learning_rate": 0.00016279197250051013, + "loss": 0.9747, + "step": 6393 + }, + { + "epoch": 1.1383547008547008, + "grad_norm": 0.6342993974685669, + "learning_rate": 0.00016278107795920018, + "loss": 0.9897, + "step": 6394 + }, + { + "epoch": 1.1385327635327636, + "grad_norm": 0.7149887084960938, + "learning_rate": 0.00016277018218782544, + "loss": 0.9659, + "step": 6395 + }, + { + "epoch": 1.138710826210826, + "grad_norm": 0.7219462394714355, + "learning_rate": 0.00016275928518659938, + "loss": 0.9301, + "step": 6396 + }, + { + "epoch": 1.1388888888888888, + "grad_norm": 0.6649485230445862, + "learning_rate": 0.0001627483869557355, + "loss": 0.9012, + "step": 6397 + }, + { + "epoch": 1.1390669515669516, + "grad_norm": 0.6910027861595154, + "learning_rate": 0.00016273748749544731, + "loss": 0.956, + "step": 6398 + }, + { + "epoch": 1.1392450142450143, + "grad_norm": 0.6369016766548157, + "learning_rate": 0.00016272658680594837, + "loss": 0.8027, + "step": 6399 + }, + { + "epoch": 1.1394230769230769, + "grad_norm": 0.6540524959564209, + "learning_rate": 0.00016271568488745227, + "loss": 1.2397, + "step": 6400 + }, + { + "epoch": 1.1396011396011396, + "grad_norm": 0.5912376046180725, + "learning_rate": 0.00016270478174017263, + "loss": 0.8453, + "step": 6401 + }, + { + "epoch": 1.1397792022792024, + "grad_norm": 0.6847240924835205, + "learning_rate": 0.00016269387736432303, + "loss": 0.9776, + "step": 6402 + }, + { + "epoch": 1.139957264957265, + "grad_norm": 0.6465024352073669, + "learning_rate": 0.00016268297176011716, + "loss": 0.8971, + "step": 6403 + }, + { + "epoch": 1.1401353276353277, + "grad_norm": 0.6639063954353333, + "learning_rate": 0.00016267206492776866, + "loss": 0.9756, + "step": 6404 + }, + { + "epoch": 1.1403133903133904, + "grad_norm": 0.6343763470649719, + "learning_rate": 0.00016266115686749123, + "loss": 0.9368, + "step": 6405 + }, + { + "epoch": 1.140491452991453, + "grad_norm": 0.7144993543624878, + "learning_rate": 0.0001626502475794986, + "loss": 0.9285, + "step": 6406 + }, + { + "epoch": 1.1406695156695157, + "grad_norm": 0.6217414736747742, + "learning_rate": 0.00016263933706400451, + "loss": 0.8867, + "step": 6407 + }, + { + "epoch": 1.1408475783475784, + "grad_norm": 0.6843730807304382, + "learning_rate": 0.00016262842532122274, + "loss": 0.9863, + "step": 6408 + }, + { + "epoch": 1.141025641025641, + "grad_norm": 0.6866166591644287, + "learning_rate": 0.00016261751235136705, + "loss": 1.0517, + "step": 6409 + }, + { + "epoch": 1.1412037037037037, + "grad_norm": 0.6650584936141968, + "learning_rate": 0.0001626065981546513, + "loss": 1.0629, + "step": 6410 + }, + { + "epoch": 1.1413817663817665, + "grad_norm": 0.5805012583732605, + "learning_rate": 0.00016259568273128933, + "loss": 0.8175, + "step": 6411 + }, + { + "epoch": 1.141559829059829, + "grad_norm": 0.7005903124809265, + "learning_rate": 0.00016258476608149497, + "loss": 1.0267, + "step": 6412 + }, + { + "epoch": 1.1417378917378918, + "grad_norm": 0.6293461322784424, + "learning_rate": 0.00016257384820548217, + "loss": 1.1034, + "step": 6413 + }, + { + "epoch": 1.1419159544159545, + "grad_norm": 0.6281774640083313, + "learning_rate": 0.00016256292910346476, + "loss": 1.0775, + "step": 6414 + }, + { + "epoch": 1.142094017094017, + "grad_norm": 0.5912862420082092, + "learning_rate": 0.0001625520087756567, + "loss": 0.9589, + "step": 6415 + }, + { + "epoch": 1.1422720797720798, + "grad_norm": 0.5813978314399719, + "learning_rate": 0.00016254108722227198, + "loss": 0.9195, + "step": 6416 + }, + { + "epoch": 1.1424501424501425, + "grad_norm": 0.650805652141571, + "learning_rate": 0.00016253016444352458, + "loss": 1.0207, + "step": 6417 + }, + { + "epoch": 1.142628205128205, + "grad_norm": 0.6909520030021667, + "learning_rate": 0.00016251924043962851, + "loss": 0.9854, + "step": 6418 + }, + { + "epoch": 1.1428062678062678, + "grad_norm": 0.6054595112800598, + "learning_rate": 0.0001625083152107978, + "loss": 0.852, + "step": 6419 + }, + { + "epoch": 1.1429843304843306, + "grad_norm": 0.601078987121582, + "learning_rate": 0.00016249738875724647, + "loss": 0.9609, + "step": 6420 + }, + { + "epoch": 1.143162393162393, + "grad_norm": 0.5340180397033691, + "learning_rate": 0.00016248646107918868, + "loss": 0.8364, + "step": 6421 + }, + { + "epoch": 1.1433404558404558, + "grad_norm": 0.6687821745872498, + "learning_rate": 0.00016247553217683846, + "loss": 1.005, + "step": 6422 + }, + { + "epoch": 1.1435185185185186, + "grad_norm": 0.6347902417182922, + "learning_rate": 0.00016246460205040998, + "loss": 1.026, + "step": 6423 + }, + { + "epoch": 1.1436965811965811, + "grad_norm": 0.6136734485626221, + "learning_rate": 0.00016245367070011736, + "loss": 0.7811, + "step": 6424 + }, + { + "epoch": 1.1438746438746439, + "grad_norm": 0.6591334342956543, + "learning_rate": 0.00016244273812617482, + "loss": 0.991, + "step": 6425 + }, + { + "epoch": 1.1440527065527066, + "grad_norm": 0.6062475442886353, + "learning_rate": 0.00016243180432879656, + "loss": 0.9879, + "step": 6426 + }, + { + "epoch": 1.1442307692307692, + "grad_norm": 0.5941380858421326, + "learning_rate": 0.00016242086930819678, + "loss": 0.9771, + "step": 6427 + }, + { + "epoch": 1.144408831908832, + "grad_norm": 0.7320533990859985, + "learning_rate": 0.00016240993306458973, + "loss": 1.0919, + "step": 6428 + }, + { + "epoch": 1.1445868945868947, + "grad_norm": 0.6998075246810913, + "learning_rate": 0.00016239899559818962, + "loss": 1.0721, + "step": 6429 + }, + { + "epoch": 1.1447649572649572, + "grad_norm": 0.847931444644928, + "learning_rate": 0.0001623880569092109, + "loss": 0.8759, + "step": 6430 + }, + { + "epoch": 1.14494301994302, + "grad_norm": 0.6670104265213013, + "learning_rate": 0.00016237711699786775, + "loss": 1.0515, + "step": 6431 + }, + { + "epoch": 1.1451210826210827, + "grad_norm": 0.601759672164917, + "learning_rate": 0.00016236617586437463, + "loss": 0.7298, + "step": 6432 + }, + { + "epoch": 1.1452991452991452, + "grad_norm": 0.6411594152450562, + "learning_rate": 0.00016235523350894578, + "loss": 0.9336, + "step": 6433 + }, + { + "epoch": 1.145477207977208, + "grad_norm": 0.6485120058059692, + "learning_rate": 0.0001623442899317957, + "loss": 1.1215, + "step": 6434 + }, + { + "epoch": 1.1456552706552707, + "grad_norm": 0.6041508316993713, + "learning_rate": 0.00016233334513313875, + "loss": 0.8917, + "step": 6435 + }, + { + "epoch": 1.1458333333333333, + "grad_norm": 0.6292745471000671, + "learning_rate": 0.0001623223991131894, + "loss": 0.9976, + "step": 6436 + }, + { + "epoch": 1.146011396011396, + "grad_norm": 0.5442200303077698, + "learning_rate": 0.0001623114518721621, + "loss": 0.8072, + "step": 6437 + }, + { + "epoch": 1.1461894586894588, + "grad_norm": 0.6668170094490051, + "learning_rate": 0.00016230050341027136, + "loss": 0.9641, + "step": 6438 + }, + { + "epoch": 1.1463675213675213, + "grad_norm": 0.644186794757843, + "learning_rate": 0.00016228955372773164, + "loss": 0.9248, + "step": 6439 + }, + { + "epoch": 1.146545584045584, + "grad_norm": 0.6661991477012634, + "learning_rate": 0.00016227860282475753, + "loss": 0.8719, + "step": 6440 + }, + { + "epoch": 1.1467236467236468, + "grad_norm": 0.5232062935829163, + "learning_rate": 0.00016226765070156355, + "loss": 0.5418, + "step": 6441 + }, + { + "epoch": 1.1469017094017093, + "grad_norm": 0.573176383972168, + "learning_rate": 0.00016225669735836436, + "loss": 1.0858, + "step": 6442 + }, + { + "epoch": 1.147079772079772, + "grad_norm": 0.6137439608573914, + "learning_rate": 0.00016224574279537446, + "loss": 1.1205, + "step": 6443 + }, + { + "epoch": 1.1472578347578348, + "grad_norm": 0.6328136920928955, + "learning_rate": 0.00016223478701280855, + "loss": 0.8957, + "step": 6444 + }, + { + "epoch": 1.1474358974358974, + "grad_norm": 0.6687374114990234, + "learning_rate": 0.00016222383001088126, + "loss": 1.0318, + "step": 6445 + }, + { + "epoch": 1.14761396011396, + "grad_norm": 0.6057115793228149, + "learning_rate": 0.0001622128717898073, + "loss": 0.9575, + "step": 6446 + }, + { + "epoch": 1.1477920227920229, + "grad_norm": 0.6758735775947571, + "learning_rate": 0.0001622019123498013, + "loss": 1.2273, + "step": 6447 + }, + { + "epoch": 1.1479700854700854, + "grad_norm": 0.6233550310134888, + "learning_rate": 0.0001621909516910781, + "loss": 0.7875, + "step": 6448 + }, + { + "epoch": 1.1481481481481481, + "grad_norm": 0.6371827721595764, + "learning_rate": 0.0001621799898138524, + "loss": 1.0488, + "step": 6449 + }, + { + "epoch": 1.148326210826211, + "grad_norm": 0.6179831624031067, + "learning_rate": 0.00016216902671833892, + "loss": 0.9792, + "step": 6450 + }, + { + "epoch": 1.1485042735042734, + "grad_norm": 0.6234193444252014, + "learning_rate": 0.00016215806240475256, + "loss": 0.927, + "step": 6451 + }, + { + "epoch": 1.1486823361823362, + "grad_norm": 0.6940563917160034, + "learning_rate": 0.00016214709687330803, + "loss": 1.047, + "step": 6452 + }, + { + "epoch": 1.148860398860399, + "grad_norm": 0.6567606925964355, + "learning_rate": 0.00016213613012422027, + "loss": 0.9695, + "step": 6453 + }, + { + "epoch": 1.1490384615384615, + "grad_norm": 0.7374183535575867, + "learning_rate": 0.0001621251621577041, + "loss": 1.0443, + "step": 6454 + }, + { + "epoch": 1.1492165242165242, + "grad_norm": 0.6789869666099548, + "learning_rate": 0.00016211419297397443, + "loss": 1.0319, + "step": 6455 + }, + { + "epoch": 1.149394586894587, + "grad_norm": 0.6225521564483643, + "learning_rate": 0.00016210322257324619, + "loss": 1.0529, + "step": 6456 + }, + { + "epoch": 1.1495726495726495, + "grad_norm": 0.619701623916626, + "learning_rate": 0.00016209225095573432, + "loss": 0.962, + "step": 6457 + }, + { + "epoch": 1.1497507122507122, + "grad_norm": 0.6132834553718567, + "learning_rate": 0.00016208127812165375, + "loss": 0.9588, + "step": 6458 + }, + { + "epoch": 1.149928774928775, + "grad_norm": 0.6005367040634155, + "learning_rate": 0.00016207030407121954, + "loss": 0.9497, + "step": 6459 + }, + { + "epoch": 1.1501068376068375, + "grad_norm": 0.575309157371521, + "learning_rate": 0.00016205932880464664, + "loss": 1.0035, + "step": 6460 + }, + { + "epoch": 1.1502849002849003, + "grad_norm": 0.5958710312843323, + "learning_rate": 0.0001620483523221501, + "loss": 1.0004, + "step": 6461 + }, + { + "epoch": 1.150462962962963, + "grad_norm": 0.5934719443321228, + "learning_rate": 0.000162037374623945, + "loss": 0.8694, + "step": 6462 + }, + { + "epoch": 1.1506410256410255, + "grad_norm": 0.6042510271072388, + "learning_rate": 0.00016202639571024643, + "loss": 0.8598, + "step": 6463 + }, + { + "epoch": 1.1508190883190883, + "grad_norm": 0.6206158399581909, + "learning_rate": 0.00016201541558126946, + "loss": 0.961, + "step": 6464 + }, + { + "epoch": 1.150997150997151, + "grad_norm": 0.5997715592384338, + "learning_rate": 0.00016200443423722925, + "loss": 0.8686, + "step": 6465 + }, + { + "epoch": 1.1511752136752136, + "grad_norm": 0.742457926273346, + "learning_rate": 0.00016199345167834098, + "loss": 1.1113, + "step": 6466 + }, + { + "epoch": 1.1513532763532763, + "grad_norm": 0.6772766709327698, + "learning_rate": 0.00016198246790481976, + "loss": 1.0717, + "step": 6467 + }, + { + "epoch": 1.151531339031339, + "grad_norm": 0.6127712726593018, + "learning_rate": 0.0001619714829168809, + "loss": 0.8887, + "step": 6468 + }, + { + "epoch": 1.1517094017094016, + "grad_norm": 0.5585067272186279, + "learning_rate": 0.00016196049671473954, + "loss": 1.0144, + "step": 6469 + }, + { + "epoch": 1.1518874643874644, + "grad_norm": 0.6269431710243225, + "learning_rate": 0.00016194950929861092, + "loss": 1.0206, + "step": 6470 + }, + { + "epoch": 1.1520655270655271, + "grad_norm": 0.6270785331726074, + "learning_rate": 0.0001619385206687104, + "loss": 1.0517, + "step": 6471 + }, + { + "epoch": 1.1522435897435896, + "grad_norm": 0.744712233543396, + "learning_rate": 0.00016192753082525322, + "loss": 1.0699, + "step": 6472 + }, + { + "epoch": 1.1524216524216524, + "grad_norm": 0.7025929689407349, + "learning_rate": 0.00016191653976845474, + "loss": 0.951, + "step": 6473 + }, + { + "epoch": 1.1525997150997151, + "grad_norm": 0.6175379753112793, + "learning_rate": 0.00016190554749853024, + "loss": 1.2153, + "step": 6474 + }, + { + "epoch": 1.1527777777777777, + "grad_norm": 0.6212149858474731, + "learning_rate": 0.00016189455401569513, + "loss": 1.0428, + "step": 6475 + }, + { + "epoch": 1.1529558404558404, + "grad_norm": 0.6716817617416382, + "learning_rate": 0.00016188355932016484, + "loss": 1.179, + "step": 6476 + }, + { + "epoch": 1.1531339031339032, + "grad_norm": 0.6247739791870117, + "learning_rate": 0.00016187256341215476, + "loss": 0.9451, + "step": 6477 + }, + { + "epoch": 1.153311965811966, + "grad_norm": 0.6223008036613464, + "learning_rate": 0.00016186156629188032, + "loss": 0.9915, + "step": 6478 + }, + { + "epoch": 1.1534900284900285, + "grad_norm": 0.5610866546630859, + "learning_rate": 0.000161850567959557, + "loss": 0.7741, + "step": 6479 + }, + { + "epoch": 1.1536680911680912, + "grad_norm": 0.6241226196289062, + "learning_rate": 0.0001618395684154003, + "loss": 1.2193, + "step": 6480 + }, + { + "epoch": 1.1538461538461537, + "grad_norm": 0.703789472579956, + "learning_rate": 0.00016182856765962567, + "loss": 1.0725, + "step": 6481 + }, + { + "epoch": 1.1540242165242165, + "grad_norm": 0.6802006959915161, + "learning_rate": 0.00016181756569244872, + "loss": 1.0908, + "step": 6482 + }, + { + "epoch": 1.1542022792022792, + "grad_norm": 0.6504136919975281, + "learning_rate": 0.000161806562514085, + "loss": 0.9706, + "step": 6483 + }, + { + "epoch": 1.154380341880342, + "grad_norm": 0.7217034101486206, + "learning_rate": 0.00016179555812475003, + "loss": 0.9084, + "step": 6484 + }, + { + "epoch": 1.1545584045584045, + "grad_norm": 0.5919039249420166, + "learning_rate": 0.0001617845525246595, + "loss": 0.949, + "step": 6485 + }, + { + "epoch": 1.1547364672364673, + "grad_norm": 0.6160184741020203, + "learning_rate": 0.00016177354571402902, + "loss": 0.8144, + "step": 6486 + }, + { + "epoch": 1.1549145299145298, + "grad_norm": 0.7323806285858154, + "learning_rate": 0.00016176253769307426, + "loss": 1.0528, + "step": 6487 + }, + { + "epoch": 1.1550925925925926, + "grad_norm": 0.6051317453384399, + "learning_rate": 0.0001617515284620108, + "loss": 0.9558, + "step": 6488 + }, + { + "epoch": 1.1552706552706553, + "grad_norm": 0.6418905258178711, + "learning_rate": 0.00016174051802105447, + "loss": 1.062, + "step": 6489 + }, + { + "epoch": 1.155448717948718, + "grad_norm": 0.6914883852005005, + "learning_rate": 0.00016172950637042096, + "loss": 0.9999, + "step": 6490 + }, + { + "epoch": 1.1556267806267806, + "grad_norm": 0.5558316707611084, + "learning_rate": 0.000161718493510326, + "loss": 0.9561, + "step": 6491 + }, + { + "epoch": 1.1558048433048433, + "grad_norm": 0.6632496118545532, + "learning_rate": 0.00016170747944098531, + "loss": 1.0133, + "step": 6492 + }, + { + "epoch": 1.1559829059829059, + "grad_norm": 0.6407149434089661, + "learning_rate": 0.00016169646416261478, + "loss": 1.0563, + "step": 6493 + }, + { + "epoch": 1.1561609686609686, + "grad_norm": 0.8128494024276733, + "learning_rate": 0.0001616854476754302, + "loss": 1.1559, + "step": 6494 + }, + { + "epoch": 1.1563390313390314, + "grad_norm": 0.6403429508209229, + "learning_rate": 0.00016167442997964742, + "loss": 1.0983, + "step": 6495 + }, + { + "epoch": 1.1565170940170941, + "grad_norm": 0.76612788438797, + "learning_rate": 0.0001616634110754823, + "loss": 0.973, + "step": 6496 + }, + { + "epoch": 1.1566951566951567, + "grad_norm": 0.6914355754852295, + "learning_rate": 0.0001616523909631507, + "loss": 0.9307, + "step": 6497 + }, + { + "epoch": 1.1568732193732194, + "grad_norm": 0.546602725982666, + "learning_rate": 0.00016164136964286863, + "loss": 1.0328, + "step": 6498 + }, + { + "epoch": 1.157051282051282, + "grad_norm": 0.5695818662643433, + "learning_rate": 0.00016163034711485193, + "loss": 0.9607, + "step": 6499 + }, + { + "epoch": 1.1572293447293447, + "grad_norm": 0.5649738311767578, + "learning_rate": 0.00016161932337931662, + "loss": 1.1521, + "step": 6500 + }, + { + "epoch": 1.1574074074074074, + "grad_norm": 0.6437582969665527, + "learning_rate": 0.00016160829843647867, + "loss": 0.9613, + "step": 6501 + }, + { + "epoch": 1.1575854700854702, + "grad_norm": 0.5841929316520691, + "learning_rate": 0.0001615972722865541, + "loss": 0.8187, + "step": 6502 + }, + { + "epoch": 1.1577635327635327, + "grad_norm": 0.6481246948242188, + "learning_rate": 0.00016158624492975892, + "loss": 1.0447, + "step": 6503 + }, + { + "epoch": 1.1579415954415955, + "grad_norm": 0.629804790019989, + "learning_rate": 0.0001615752163663092, + "loss": 0.9034, + "step": 6504 + }, + { + "epoch": 1.158119658119658, + "grad_norm": 0.5797054171562195, + "learning_rate": 0.00016156418659642104, + "loss": 0.8168, + "step": 6505 + }, + { + "epoch": 1.1582977207977208, + "grad_norm": 0.588424563407898, + "learning_rate": 0.00016155315562031052, + "loss": 0.828, + "step": 6506 + }, + { + "epoch": 1.1584757834757835, + "grad_norm": 0.7120068669319153, + "learning_rate": 0.0001615421234381938, + "loss": 1.0637, + "step": 6507 + }, + { + "epoch": 1.1586538461538463, + "grad_norm": 0.6635081768035889, + "learning_rate": 0.00016153109005028702, + "loss": 0.9838, + "step": 6508 + }, + { + "epoch": 1.1588319088319088, + "grad_norm": 0.6080414056777954, + "learning_rate": 0.00016152005545680634, + "loss": 0.983, + "step": 6509 + }, + { + "epoch": 1.1590099715099715, + "grad_norm": 0.7131237983703613, + "learning_rate": 0.00016150901965796796, + "loss": 1.1053, + "step": 6510 + }, + { + "epoch": 1.159188034188034, + "grad_norm": 0.6051005125045776, + "learning_rate": 0.00016149798265398813, + "loss": 0.9903, + "step": 6511 + }, + { + "epoch": 1.1593660968660968, + "grad_norm": 0.6193733811378479, + "learning_rate": 0.00016148694444508306, + "loss": 1.0478, + "step": 6512 + }, + { + "epoch": 1.1595441595441596, + "grad_norm": 0.567888081073761, + "learning_rate": 0.00016147590503146905, + "loss": 0.7995, + "step": 6513 + }, + { + "epoch": 1.1597222222222223, + "grad_norm": 0.6889783143997192, + "learning_rate": 0.00016146486441336242, + "loss": 0.9684, + "step": 6514 + }, + { + "epoch": 1.1599002849002849, + "grad_norm": 0.6470308303833008, + "learning_rate": 0.0001614538225909794, + "loss": 0.9824, + "step": 6515 + }, + { + "epoch": 1.1600783475783476, + "grad_norm": 0.6833886504173279, + "learning_rate": 0.00016144277956453638, + "loss": 0.9845, + "step": 6516 + }, + { + "epoch": 1.1602564102564104, + "grad_norm": 0.5827815532684326, + "learning_rate": 0.00016143173533424978, + "loss": 0.9476, + "step": 6517 + }, + { + "epoch": 1.1604344729344729, + "grad_norm": 0.6701242327690125, + "learning_rate": 0.00016142068990033593, + "loss": 1.0839, + "step": 6518 + }, + { + "epoch": 1.1606125356125356, + "grad_norm": 0.5844996571540833, + "learning_rate": 0.00016140964326301122, + "loss": 0.8861, + "step": 6519 + }, + { + "epoch": 1.1607905982905984, + "grad_norm": 0.5831994414329529, + "learning_rate": 0.00016139859542249214, + "loss": 0.9817, + "step": 6520 + }, + { + "epoch": 1.160968660968661, + "grad_norm": 0.6830124855041504, + "learning_rate": 0.0001613875463789951, + "loss": 0.8749, + "step": 6521 + }, + { + "epoch": 1.1611467236467237, + "grad_norm": 0.6003018021583557, + "learning_rate": 0.00016137649613273667, + "loss": 0.9593, + "step": 6522 + }, + { + "epoch": 1.1613247863247864, + "grad_norm": 0.5973994731903076, + "learning_rate": 0.00016136544468393327, + "loss": 1.0384, + "step": 6523 + }, + { + "epoch": 1.161502849002849, + "grad_norm": 0.6702523827552795, + "learning_rate": 0.00016135439203280143, + "loss": 1.0431, + "step": 6524 + }, + { + "epoch": 1.1616809116809117, + "grad_norm": 0.6160697937011719, + "learning_rate": 0.00016134333817955775, + "loss": 1.0339, + "step": 6525 + }, + { + "epoch": 1.1618589743589745, + "grad_norm": 0.7078264355659485, + "learning_rate": 0.0001613322831244188, + "loss": 1.0285, + "step": 6526 + }, + { + "epoch": 1.162037037037037, + "grad_norm": 0.5744216442108154, + "learning_rate": 0.00016132122686760117, + "loss": 0.6589, + "step": 6527 + }, + { + "epoch": 1.1622150997150997, + "grad_norm": 0.6802098155021667, + "learning_rate": 0.00016131016940932146, + "loss": 0.9532, + "step": 6528 + }, + { + "epoch": 1.1623931623931625, + "grad_norm": 0.6523237228393555, + "learning_rate": 0.00016129911074979635, + "loss": 0.9409, + "step": 6529 + }, + { + "epoch": 1.162571225071225, + "grad_norm": 0.710307776927948, + "learning_rate": 0.00016128805088924252, + "loss": 1.2536, + "step": 6530 + }, + { + "epoch": 1.1627492877492878, + "grad_norm": 0.6349819898605347, + "learning_rate": 0.0001612769898278766, + "loss": 1.0857, + "step": 6531 + }, + { + "epoch": 1.1629273504273505, + "grad_norm": 0.5348139405250549, + "learning_rate": 0.00016126592756591542, + "loss": 0.5969, + "step": 6532 + }, + { + "epoch": 1.163105413105413, + "grad_norm": 0.635619580745697, + "learning_rate": 0.00016125486410357564, + "loss": 0.9885, + "step": 6533 + }, + { + "epoch": 1.1632834757834758, + "grad_norm": 0.6434559226036072, + "learning_rate": 0.000161243799441074, + "loss": 0.8377, + "step": 6534 + }, + { + "epoch": 1.1634615384615385, + "grad_norm": 0.6509647369384766, + "learning_rate": 0.00016123273357862737, + "loss": 0.8393, + "step": 6535 + }, + { + "epoch": 1.163639601139601, + "grad_norm": 0.6179081797599792, + "learning_rate": 0.0001612216665164525, + "loss": 0.9143, + "step": 6536 + }, + { + "epoch": 1.1638176638176638, + "grad_norm": 0.5923223495483398, + "learning_rate": 0.0001612105982547663, + "loss": 1.0185, + "step": 6537 + }, + { + "epoch": 1.1639957264957266, + "grad_norm": 0.702150285243988, + "learning_rate": 0.00016119952879378556, + "loss": 0.863, + "step": 6538 + }, + { + "epoch": 1.164173789173789, + "grad_norm": 0.6596643328666687, + "learning_rate": 0.00016118845813372715, + "loss": 1.0089, + "step": 6539 + }, + { + "epoch": 1.1643518518518519, + "grad_norm": 0.7675769329071045, + "learning_rate": 0.00016117738627480804, + "loss": 1.0179, + "step": 6540 + }, + { + "epoch": 1.1645299145299146, + "grad_norm": 0.6742541193962097, + "learning_rate": 0.00016116631321724513, + "loss": 1.0663, + "step": 6541 + }, + { + "epoch": 1.1647079772079771, + "grad_norm": 0.7379785776138306, + "learning_rate": 0.0001611552389612554, + "loss": 1.0162, + "step": 6542 + }, + { + "epoch": 1.16488603988604, + "grad_norm": 0.5729365944862366, + "learning_rate": 0.00016114416350705577, + "loss": 0.8146, + "step": 6543 + }, + { + "epoch": 1.1650641025641026, + "grad_norm": 0.6481349468231201, + "learning_rate": 0.00016113308685486327, + "loss": 1.0748, + "step": 6544 + }, + { + "epoch": 1.1652421652421652, + "grad_norm": 0.5588181018829346, + "learning_rate": 0.00016112200900489493, + "loss": 0.7511, + "step": 6545 + }, + { + "epoch": 1.165420227920228, + "grad_norm": 0.674363911151886, + "learning_rate": 0.0001611109299573678, + "loss": 0.9852, + "step": 6546 + }, + { + "epoch": 1.1655982905982907, + "grad_norm": 0.6712620854377747, + "learning_rate": 0.00016109984971249893, + "loss": 0.9558, + "step": 6547 + }, + { + "epoch": 1.1657763532763532, + "grad_norm": 0.5260626077651978, + "learning_rate": 0.00016108876827050544, + "loss": 0.7008, + "step": 6548 + }, + { + "epoch": 1.165954415954416, + "grad_norm": 0.6056292057037354, + "learning_rate": 0.00016107768563160445, + "loss": 0.7756, + "step": 6549 + }, + { + "epoch": 1.1661324786324787, + "grad_norm": 0.5725821256637573, + "learning_rate": 0.00016106660179601308, + "loss": 0.8228, + "step": 6550 + }, + { + "epoch": 1.1663105413105412, + "grad_norm": 0.6708397269248962, + "learning_rate": 0.00016105551676394848, + "loss": 1.0711, + "step": 6551 + }, + { + "epoch": 1.166488603988604, + "grad_norm": 0.645453155040741, + "learning_rate": 0.00016104443053562787, + "loss": 0.9299, + "step": 6552 + }, + { + "epoch": 1.1666666666666667, + "grad_norm": 0.6743524074554443, + "learning_rate": 0.00016103334311126847, + "loss": 0.8977, + "step": 6553 + }, + { + "epoch": 1.1668447293447293, + "grad_norm": 0.7248545289039612, + "learning_rate": 0.0001610222544910875, + "loss": 1.2135, + "step": 6554 + }, + { + "epoch": 1.167022792022792, + "grad_norm": 0.5798853635787964, + "learning_rate": 0.00016101116467530217, + "loss": 0.857, + "step": 6555 + }, + { + "epoch": 1.1672008547008548, + "grad_norm": 0.6828082799911499, + "learning_rate": 0.00016100007366412985, + "loss": 0.9405, + "step": 6556 + }, + { + "epoch": 1.1673789173789173, + "grad_norm": 0.6820163130760193, + "learning_rate": 0.0001609889814577878, + "loss": 0.9144, + "step": 6557 + }, + { + "epoch": 1.16755698005698, + "grad_norm": 0.6482275128364563, + "learning_rate": 0.00016097788805649333, + "loss": 0.8586, + "step": 6558 + }, + { + "epoch": 1.1677350427350428, + "grad_norm": 0.6404715180397034, + "learning_rate": 0.00016096679346046385, + "loss": 0.7018, + "step": 6559 + }, + { + "epoch": 1.1679131054131053, + "grad_norm": 0.6315203309059143, + "learning_rate": 0.0001609556976699167, + "loss": 0.9602, + "step": 6560 + }, + { + "epoch": 1.168091168091168, + "grad_norm": 0.5521387457847595, + "learning_rate": 0.00016094460068506925, + "loss": 0.9294, + "step": 6561 + }, + { + "epoch": 1.1682692307692308, + "grad_norm": 0.583372175693512, + "learning_rate": 0.00016093350250613895, + "loss": 1.077, + "step": 6562 + }, + { + "epoch": 1.1684472934472934, + "grad_norm": 0.5990512371063232, + "learning_rate": 0.00016092240313334325, + "loss": 1.0102, + "step": 6563 + }, + { + "epoch": 1.1686253561253561, + "grad_norm": 0.675128161907196, + "learning_rate": 0.00016091130256689964, + "loss": 1.0407, + "step": 6564 + }, + { + "epoch": 1.1688034188034189, + "grad_norm": 0.48797324299812317, + "learning_rate": 0.00016090020080702556, + "loss": 0.7821, + "step": 6565 + }, + { + "epoch": 1.1689814814814814, + "grad_norm": 0.7487484216690063, + "learning_rate": 0.00016088909785393857, + "loss": 1.0444, + "step": 6566 + }, + { + "epoch": 1.1691595441595442, + "grad_norm": 0.6288858652114868, + "learning_rate": 0.00016087799370785618, + "loss": 1.1854, + "step": 6567 + }, + { + "epoch": 1.169337606837607, + "grad_norm": 0.6639021635055542, + "learning_rate": 0.000160866888368996, + "loss": 0.9632, + "step": 6568 + }, + { + "epoch": 1.1695156695156694, + "grad_norm": 0.6553738713264465, + "learning_rate": 0.00016085578183757556, + "loss": 1.2765, + "step": 6569 + }, + { + "epoch": 1.1696937321937322, + "grad_norm": 0.7489066123962402, + "learning_rate": 0.00016084467411381248, + "loss": 1.0705, + "step": 6570 + }, + { + "epoch": 1.169871794871795, + "grad_norm": 0.7079828381538391, + "learning_rate": 0.00016083356519792444, + "loss": 0.8256, + "step": 6571 + }, + { + "epoch": 1.1700498575498575, + "grad_norm": 0.7065926790237427, + "learning_rate": 0.00016082245509012902, + "loss": 1.0439, + "step": 6572 + }, + { + "epoch": 1.1702279202279202, + "grad_norm": 0.6113346815109253, + "learning_rate": 0.00016081134379064395, + "loss": 0.9153, + "step": 6573 + }, + { + "epoch": 1.170405982905983, + "grad_norm": 0.6094171404838562, + "learning_rate": 0.0001608002312996869, + "loss": 0.9723, + "step": 6574 + }, + { + "epoch": 1.1705840455840455, + "grad_norm": 0.6208072900772095, + "learning_rate": 0.00016078911761747565, + "loss": 0.948, + "step": 6575 + }, + { + "epoch": 1.1707621082621082, + "grad_norm": 0.5736680626869202, + "learning_rate": 0.00016077800274422792, + "loss": 0.9155, + "step": 6576 + }, + { + "epoch": 1.170940170940171, + "grad_norm": 0.6793957948684692, + "learning_rate": 0.0001607668866801615, + "loss": 0.9574, + "step": 6577 + }, + { + "epoch": 1.1711182336182335, + "grad_norm": 0.6251805424690247, + "learning_rate": 0.00016075576942549413, + "loss": 1.0319, + "step": 6578 + }, + { + "epoch": 1.1712962962962963, + "grad_norm": 0.628882110118866, + "learning_rate": 0.0001607446509804437, + "loss": 0.9336, + "step": 6579 + }, + { + "epoch": 1.171474358974359, + "grad_norm": 0.6712356805801392, + "learning_rate": 0.000160733531345228, + "loss": 1.0958, + "step": 6580 + }, + { + "epoch": 1.1716524216524216, + "grad_norm": 0.599365770816803, + "learning_rate": 0.0001607224105200649, + "loss": 0.9814, + "step": 6581 + }, + { + "epoch": 1.1718304843304843, + "grad_norm": 0.5798245668411255, + "learning_rate": 0.00016071128850517235, + "loss": 1.0355, + "step": 6582 + }, + { + "epoch": 1.172008547008547, + "grad_norm": 0.7646229863166809, + "learning_rate": 0.00016070016530076817, + "loss": 0.9976, + "step": 6583 + }, + { + "epoch": 1.1721866096866096, + "grad_norm": 0.6371127367019653, + "learning_rate": 0.0001606890409070704, + "loss": 0.9588, + "step": 6584 + }, + { + "epoch": 1.1723646723646723, + "grad_norm": 0.6497066617012024, + "learning_rate": 0.0001606779153242969, + "loss": 0.8817, + "step": 6585 + }, + { + "epoch": 1.172542735042735, + "grad_norm": 0.7255781888961792, + "learning_rate": 0.0001606667885526657, + "loss": 1.1319, + "step": 6586 + }, + { + "epoch": 1.1727207977207976, + "grad_norm": 0.67711341381073, + "learning_rate": 0.00016065566059239483, + "loss": 1.0755, + "step": 6587 + }, + { + "epoch": 1.1728988603988604, + "grad_norm": 0.6159650087356567, + "learning_rate": 0.00016064453144370227, + "loss": 0.9892, + "step": 6588 + }, + { + "epoch": 1.1730769230769231, + "grad_norm": 0.658938467502594, + "learning_rate": 0.00016063340110680609, + "loss": 0.9131, + "step": 6589 + }, + { + "epoch": 1.1732549857549857, + "grad_norm": 0.6754795908927917, + "learning_rate": 0.00016062226958192438, + "loss": 1.0119, + "step": 6590 + }, + { + "epoch": 1.1734330484330484, + "grad_norm": 0.6453405022621155, + "learning_rate": 0.00016061113686927523, + "loss": 0.997, + "step": 6591 + }, + { + "epoch": 1.1736111111111112, + "grad_norm": 0.6580284237861633, + "learning_rate": 0.00016060000296907675, + "loss": 0.8432, + "step": 6592 + }, + { + "epoch": 1.173789173789174, + "grad_norm": 0.6588153839111328, + "learning_rate": 0.00016058886788154712, + "loss": 1.0725, + "step": 6593 + }, + { + "epoch": 1.1739672364672364, + "grad_norm": 0.6247910857200623, + "learning_rate": 0.00016057773160690447, + "loss": 0.8736, + "step": 6594 + }, + { + "epoch": 1.1741452991452992, + "grad_norm": 0.579594075679779, + "learning_rate": 0.000160566594145367, + "loss": 0.8809, + "step": 6595 + }, + { + "epoch": 1.1743233618233617, + "grad_norm": 0.6738116145133972, + "learning_rate": 0.00016055545549715293, + "loss": 0.825, + "step": 6596 + }, + { + "epoch": 1.1745014245014245, + "grad_norm": 0.6658982634544373, + "learning_rate": 0.00016054431566248054, + "loss": 1.0809, + "step": 6597 + }, + { + "epoch": 1.1746794871794872, + "grad_norm": 0.5367915630340576, + "learning_rate": 0.00016053317464156803, + "loss": 0.9005, + "step": 6598 + }, + { + "epoch": 1.17485754985755, + "grad_norm": 0.7243228554725647, + "learning_rate": 0.00016052203243463372, + "loss": 1.0573, + "step": 6599 + }, + { + "epoch": 1.1750356125356125, + "grad_norm": 0.6359432935714722, + "learning_rate": 0.0001605108890418959, + "loss": 0.8569, + "step": 6600 + }, + { + "epoch": 1.1752136752136753, + "grad_norm": 0.6565225720405579, + "learning_rate": 0.0001604997444635729, + "loss": 0.9748, + "step": 6601 + }, + { + "epoch": 1.1753917378917378, + "grad_norm": 0.7124663591384888, + "learning_rate": 0.0001604885986998831, + "loss": 1.0271, + "step": 6602 + }, + { + "epoch": 1.1755698005698005, + "grad_norm": 0.659766435623169, + "learning_rate": 0.00016047745175104487, + "loss": 1.0635, + "step": 6603 + }, + { + "epoch": 1.1757478632478633, + "grad_norm": 0.5874318480491638, + "learning_rate": 0.00016046630361727656, + "loss": 0.9257, + "step": 6604 + }, + { + "epoch": 1.175925925925926, + "grad_norm": 0.587345540523529, + "learning_rate": 0.0001604551542987967, + "loss": 1.0759, + "step": 6605 + }, + { + "epoch": 1.1761039886039886, + "grad_norm": 0.733567476272583, + "learning_rate": 0.00016044400379582364, + "loss": 0.9877, + "step": 6606 + }, + { + "epoch": 1.1762820512820513, + "grad_norm": 0.6538317203521729, + "learning_rate": 0.0001604328521085759, + "loss": 1.0094, + "step": 6607 + }, + { + "epoch": 1.1764601139601139, + "grad_norm": 0.6279696822166443, + "learning_rate": 0.00016042169923727195, + "loss": 1.1049, + "step": 6608 + }, + { + "epoch": 1.1766381766381766, + "grad_norm": 0.6949752569198608, + "learning_rate": 0.00016041054518213033, + "loss": 1.1418, + "step": 6609 + }, + { + "epoch": 1.1768162393162394, + "grad_norm": 0.6144010424613953, + "learning_rate": 0.00016039938994336957, + "loss": 1.0306, + "step": 6610 + }, + { + "epoch": 1.176994301994302, + "grad_norm": 0.5868683457374573, + "learning_rate": 0.00016038823352120823, + "loss": 0.9894, + "step": 6611 + }, + { + "epoch": 1.1771723646723646, + "grad_norm": 0.7181115746498108, + "learning_rate": 0.0001603770759158649, + "loss": 1.1674, + "step": 6612 + }, + { + "epoch": 1.1773504273504274, + "grad_norm": 0.6271308064460754, + "learning_rate": 0.00016036591712755818, + "loss": 0.9726, + "step": 6613 + }, + { + "epoch": 1.17752849002849, + "grad_norm": 0.6922675371170044, + "learning_rate": 0.00016035475715650668, + "loss": 0.9142, + "step": 6614 + }, + { + "epoch": 1.1777065527065527, + "grad_norm": 0.6838833689689636, + "learning_rate": 0.00016034359600292913, + "loss": 1.1627, + "step": 6615 + }, + { + "epoch": 1.1778846153846154, + "grad_norm": 0.6628252267837524, + "learning_rate": 0.00016033243366704418, + "loss": 0.739, + "step": 6616 + }, + { + "epoch": 1.1780626780626782, + "grad_norm": 0.6367576122283936, + "learning_rate": 0.0001603212701490705, + "loss": 0.9015, + "step": 6617 + }, + { + "epoch": 1.1782407407407407, + "grad_norm": 0.6498967409133911, + "learning_rate": 0.00016031010544922687, + "loss": 0.9645, + "step": 6618 + }, + { + "epoch": 1.1784188034188035, + "grad_norm": 0.468795508146286, + "learning_rate": 0.00016029893956773198, + "loss": 0.7305, + "step": 6619 + }, + { + "epoch": 1.178596866096866, + "grad_norm": 0.6355500817298889, + "learning_rate": 0.00016028777250480465, + "loss": 0.9183, + "step": 6620 + }, + { + "epoch": 1.1787749287749287, + "grad_norm": 0.7582615613937378, + "learning_rate": 0.0001602766042606636, + "loss": 1.1641, + "step": 6621 + }, + { + "epoch": 1.1789529914529915, + "grad_norm": 0.580035924911499, + "learning_rate": 0.00016026543483552776, + "loss": 0.9164, + "step": 6622 + }, + { + "epoch": 1.1791310541310542, + "grad_norm": 0.6198559999465942, + "learning_rate": 0.00016025426422961592, + "loss": 0.9803, + "step": 6623 + }, + { + "epoch": 1.1793091168091168, + "grad_norm": 0.59112149477005, + "learning_rate": 0.0001602430924431469, + "loss": 0.8645, + "step": 6624 + }, + { + "epoch": 1.1794871794871795, + "grad_norm": 0.6200533509254456, + "learning_rate": 0.00016023191947633965, + "loss": 1.068, + "step": 6625 + }, + { + "epoch": 1.179665242165242, + "grad_norm": 0.6077516078948975, + "learning_rate": 0.00016022074532941305, + "loss": 1.0017, + "step": 6626 + }, + { + "epoch": 1.1798433048433048, + "grad_norm": 0.6770145893096924, + "learning_rate": 0.00016020957000258606, + "loss": 0.9022, + "step": 6627 + }, + { + "epoch": 1.1800213675213675, + "grad_norm": 0.6478054523468018, + "learning_rate": 0.0001601983934960776, + "loss": 0.8615, + "step": 6628 + }, + { + "epoch": 1.1801994301994303, + "grad_norm": 0.6528988480567932, + "learning_rate": 0.00016018721581010666, + "loss": 1.0015, + "step": 6629 + }, + { + "epoch": 1.1803774928774928, + "grad_norm": 0.6160712242126465, + "learning_rate": 0.0001601760369448923, + "loss": 0.9382, + "step": 6630 + }, + { + "epoch": 1.1805555555555556, + "grad_norm": 0.5755789875984192, + "learning_rate": 0.00016016485690065345, + "loss": 1.0551, + "step": 6631 + }, + { + "epoch": 1.180733618233618, + "grad_norm": 0.8495022654533386, + "learning_rate": 0.00016015367567760925, + "loss": 0.9295, + "step": 6632 + }, + { + "epoch": 1.1809116809116809, + "grad_norm": 0.6010929346084595, + "learning_rate": 0.0001601424932759787, + "loss": 1.0413, + "step": 6633 + }, + { + "epoch": 1.1810897435897436, + "grad_norm": 0.6953579187393188, + "learning_rate": 0.00016013130969598093, + "loss": 1.0149, + "step": 6634 + }, + { + "epoch": 1.1812678062678064, + "grad_norm": 0.6949529647827148, + "learning_rate": 0.0001601201249378351, + "loss": 0.9992, + "step": 6635 + }, + { + "epoch": 1.181445868945869, + "grad_norm": 0.6471893787384033, + "learning_rate": 0.00016010893900176028, + "loss": 0.7985, + "step": 6636 + }, + { + "epoch": 1.1816239316239316, + "grad_norm": 0.6524858474731445, + "learning_rate": 0.00016009775188797568, + "loss": 0.9517, + "step": 6637 + }, + { + "epoch": 1.1818019943019944, + "grad_norm": 0.639214038848877, + "learning_rate": 0.00016008656359670046, + "loss": 1.0357, + "step": 6638 + }, + { + "epoch": 1.181980056980057, + "grad_norm": 0.6039628386497498, + "learning_rate": 0.00016007537412815386, + "loss": 1.0536, + "step": 6639 + }, + { + "epoch": 1.1821581196581197, + "grad_norm": 0.653540313243866, + "learning_rate": 0.00016006418348255507, + "loss": 0.9414, + "step": 6640 + }, + { + "epoch": 1.1823361823361824, + "grad_norm": 0.6331741809844971, + "learning_rate": 0.0001600529916601234, + "loss": 1.0352, + "step": 6641 + }, + { + "epoch": 1.182514245014245, + "grad_norm": 0.7552719712257385, + "learning_rate": 0.00016004179866107812, + "loss": 1.1103, + "step": 6642 + }, + { + "epoch": 1.1826923076923077, + "grad_norm": 0.6795875430107117, + "learning_rate": 0.00016003060448563852, + "loss": 1.1246, + "step": 6643 + }, + { + "epoch": 1.1828703703703705, + "grad_norm": 0.6308842301368713, + "learning_rate": 0.0001600194091340239, + "loss": 0.9532, + "step": 6644 + }, + { + "epoch": 1.183048433048433, + "grad_norm": 0.5640553832054138, + "learning_rate": 0.00016000821260645366, + "loss": 0.7491, + "step": 6645 + }, + { + "epoch": 1.1832264957264957, + "grad_norm": 0.5611832141876221, + "learning_rate": 0.00015999701490314712, + "loss": 0.9239, + "step": 6646 + }, + { + "epoch": 1.1834045584045585, + "grad_norm": 0.5881187915802002, + "learning_rate": 0.00015998581602432374, + "loss": 0.9246, + "step": 6647 + }, + { + "epoch": 1.183582621082621, + "grad_norm": 0.7291010022163391, + "learning_rate": 0.00015997461597020291, + "loss": 1.0314, + "step": 6648 + }, + { + "epoch": 1.1837606837606838, + "grad_norm": 0.6784794926643372, + "learning_rate": 0.00015996341474100402, + "loss": 1.0011, + "step": 6649 + }, + { + "epoch": 1.1839387464387465, + "grad_norm": 0.7083746194839478, + "learning_rate": 0.00015995221233694663, + "loss": 1.0336, + "step": 6650 + }, + { + "epoch": 1.184116809116809, + "grad_norm": 0.7081790566444397, + "learning_rate": 0.00015994100875825015, + "loss": 1.2386, + "step": 6651 + }, + { + "epoch": 1.1842948717948718, + "grad_norm": 0.5938812494277954, + "learning_rate": 0.00015992980400513415, + "loss": 0.7549, + "step": 6652 + }, + { + "epoch": 1.1844729344729346, + "grad_norm": 0.7084267139434814, + "learning_rate": 0.00015991859807781811, + "loss": 1.1194, + "step": 6653 + }, + { + "epoch": 1.184650997150997, + "grad_norm": 0.6391362547874451, + "learning_rate": 0.0001599073909765216, + "loss": 1.0857, + "step": 6654 + }, + { + "epoch": 1.1848290598290598, + "grad_norm": 0.8074106574058533, + "learning_rate": 0.00015989618270146423, + "loss": 1.1715, + "step": 6655 + }, + { + "epoch": 1.1850071225071226, + "grad_norm": 0.5778565406799316, + "learning_rate": 0.0001598849732528656, + "loss": 0.8843, + "step": 6656 + }, + { + "epoch": 1.1851851851851851, + "grad_norm": 0.6955079436302185, + "learning_rate": 0.00015987376263094526, + "loss": 1.0281, + "step": 6657 + }, + { + "epoch": 1.1853632478632479, + "grad_norm": 0.6789296269416809, + "learning_rate": 0.00015986255083592297, + "loss": 0.9739, + "step": 6658 + }, + { + "epoch": 1.1855413105413106, + "grad_norm": 0.6294292211532593, + "learning_rate": 0.00015985133786801834, + "loss": 1.0692, + "step": 6659 + }, + { + "epoch": 1.1857193732193732, + "grad_norm": 0.5604581832885742, + "learning_rate": 0.00015984012372745107, + "loss": 0.9059, + "step": 6660 + }, + { + "epoch": 1.185897435897436, + "grad_norm": 0.6727550625801086, + "learning_rate": 0.00015982890841444088, + "loss": 1.049, + "step": 6661 + }, + { + "epoch": 1.1860754985754987, + "grad_norm": 0.620914101600647, + "learning_rate": 0.0001598176919292075, + "loss": 1.1021, + "step": 6662 + }, + { + "epoch": 1.1862535612535612, + "grad_norm": 0.6696683168411255, + "learning_rate": 0.00015980647427197076, + "loss": 0.9053, + "step": 6663 + }, + { + "epoch": 1.186431623931624, + "grad_norm": 0.6713385581970215, + "learning_rate": 0.00015979525544295036, + "loss": 0.9596, + "step": 6664 + }, + { + "epoch": 1.1866096866096867, + "grad_norm": 0.7643477320671082, + "learning_rate": 0.00015978403544236614, + "loss": 0.882, + "step": 6665 + }, + { + "epoch": 1.1867877492877492, + "grad_norm": 0.5890966057777405, + "learning_rate": 0.00015977281427043794, + "loss": 1.0215, + "step": 6666 + }, + { + "epoch": 1.186965811965812, + "grad_norm": 0.7287502288818359, + "learning_rate": 0.0001597615919273856, + "loss": 1.0111, + "step": 6667 + }, + { + "epoch": 1.1871438746438747, + "grad_norm": 0.5713803172111511, + "learning_rate": 0.00015975036841342903, + "loss": 1.0068, + "step": 6668 + }, + { + "epoch": 1.1873219373219372, + "grad_norm": 0.5113094449043274, + "learning_rate": 0.0001597391437287881, + "loss": 0.9018, + "step": 6669 + }, + { + "epoch": 1.1875, + "grad_norm": 0.585640013217926, + "learning_rate": 0.00015972791787368276, + "loss": 1.0375, + "step": 6670 + }, + { + "epoch": 1.1876780626780628, + "grad_norm": 0.5778326392173767, + "learning_rate": 0.00015971669084833293, + "loss": 0.9975, + "step": 6671 + }, + { + "epoch": 1.1878561253561253, + "grad_norm": 0.6707763075828552, + "learning_rate": 0.0001597054626529586, + "loss": 1.0048, + "step": 6672 + }, + { + "epoch": 1.188034188034188, + "grad_norm": 0.6113292574882507, + "learning_rate": 0.00015969423328777974, + "loss": 1.1447, + "step": 6673 + }, + { + "epoch": 1.1882122507122508, + "grad_norm": 0.6075651049613953, + "learning_rate": 0.00015968300275301638, + "loss": 0.9212, + "step": 6674 + }, + { + "epoch": 1.1883903133903133, + "grad_norm": 0.6990494132041931, + "learning_rate": 0.00015967177104888857, + "loss": 0.9952, + "step": 6675 + }, + { + "epoch": 1.188568376068376, + "grad_norm": 0.6228706240653992, + "learning_rate": 0.00015966053817561638, + "loss": 1.0187, + "step": 6676 + }, + { + "epoch": 1.1887464387464388, + "grad_norm": 0.6387844085693359, + "learning_rate": 0.00015964930413341985, + "loss": 1.1614, + "step": 6677 + }, + { + "epoch": 1.1889245014245013, + "grad_norm": 0.6501925587654114, + "learning_rate": 0.00015963806892251915, + "loss": 1.0366, + "step": 6678 + }, + { + "epoch": 1.189102564102564, + "grad_norm": 0.6923739910125732, + "learning_rate": 0.00015962683254313435, + "loss": 1.1992, + "step": 6679 + }, + { + "epoch": 1.1892806267806268, + "grad_norm": 0.6640275120735168, + "learning_rate": 0.00015961559499548563, + "loss": 0.8883, + "step": 6680 + }, + { + "epoch": 1.1894586894586894, + "grad_norm": 0.6493857502937317, + "learning_rate": 0.00015960435627979317, + "loss": 1.1368, + "step": 6681 + }, + { + "epoch": 1.1896367521367521, + "grad_norm": 0.6357189416885376, + "learning_rate": 0.0001595931163962772, + "loss": 1.0502, + "step": 6682 + }, + { + "epoch": 1.1898148148148149, + "grad_norm": 0.5756343007087708, + "learning_rate": 0.0001595818753451579, + "loss": 0.9871, + "step": 6683 + }, + { + "epoch": 1.1899928774928774, + "grad_norm": 0.7369210124015808, + "learning_rate": 0.0001595706331266555, + "loss": 1.3229, + "step": 6684 + }, + { + "epoch": 1.1901709401709402, + "grad_norm": 0.7140820622444153, + "learning_rate": 0.0001595593897409903, + "loss": 1.1154, + "step": 6685 + }, + { + "epoch": 1.190349002849003, + "grad_norm": 0.696973443031311, + "learning_rate": 0.00015954814518838255, + "loss": 0.9806, + "step": 6686 + }, + { + "epoch": 1.1905270655270654, + "grad_norm": 0.5299260020256042, + "learning_rate": 0.00015953689946905262, + "loss": 0.771, + "step": 6687 + }, + { + "epoch": 1.1907051282051282, + "grad_norm": 0.6814879775047302, + "learning_rate": 0.00015952565258322085, + "loss": 0.8444, + "step": 6688 + }, + { + "epoch": 1.190883190883191, + "grad_norm": 0.6215870976448059, + "learning_rate": 0.00015951440453110754, + "loss": 1.0743, + "step": 6689 + }, + { + "epoch": 1.1910612535612535, + "grad_norm": 0.7017203569412231, + "learning_rate": 0.00015950315531293308, + "loss": 1.185, + "step": 6690 + }, + { + "epoch": 1.1912393162393162, + "grad_norm": 0.7147250175476074, + "learning_rate": 0.00015949190492891795, + "loss": 1.0646, + "step": 6691 + }, + { + "epoch": 1.191417378917379, + "grad_norm": 0.5867117047309875, + "learning_rate": 0.00015948065337928252, + "loss": 1.0554, + "step": 6692 + }, + { + "epoch": 1.1915954415954415, + "grad_norm": 0.6813527345657349, + "learning_rate": 0.0001594694006642472, + "loss": 1.1451, + "step": 6693 + }, + { + "epoch": 1.1917735042735043, + "grad_norm": 0.5192593932151794, + "learning_rate": 0.00015945814678403256, + "loss": 0.7886, + "step": 6694 + }, + { + "epoch": 1.191951566951567, + "grad_norm": 0.6537744402885437, + "learning_rate": 0.00015944689173885904, + "loss": 0.9905, + "step": 6695 + }, + { + "epoch": 1.1921296296296295, + "grad_norm": 0.7350276112556458, + "learning_rate": 0.00015943563552894716, + "loss": 0.9009, + "step": 6696 + }, + { + "epoch": 1.1923076923076923, + "grad_norm": 0.7086381316184998, + "learning_rate": 0.00015942437815451746, + "loss": 0.9117, + "step": 6697 + }, + { + "epoch": 1.192485754985755, + "grad_norm": 0.6774969696998596, + "learning_rate": 0.00015941311961579054, + "loss": 1.1172, + "step": 6698 + }, + { + "epoch": 1.1926638176638176, + "grad_norm": 0.7034362554550171, + "learning_rate": 0.00015940185991298694, + "loss": 0.8054, + "step": 6699 + }, + { + "epoch": 1.1928418803418803, + "grad_norm": 0.66145920753479, + "learning_rate": 0.00015939059904632728, + "loss": 0.7417, + "step": 6700 + }, + { + "epoch": 1.193019943019943, + "grad_norm": 0.6590890884399414, + "learning_rate": 0.00015937933701603223, + "loss": 0.9169, + "step": 6701 + }, + { + "epoch": 1.1931980056980056, + "grad_norm": 0.7492850422859192, + "learning_rate": 0.0001593680738223224, + "loss": 1.0529, + "step": 6702 + }, + { + "epoch": 1.1933760683760684, + "grad_norm": 0.7103236317634583, + "learning_rate": 0.00015935680946541848, + "loss": 1.1377, + "step": 6703 + }, + { + "epoch": 1.193554131054131, + "grad_norm": 0.6164175868034363, + "learning_rate": 0.00015934554394554122, + "loss": 0.8636, + "step": 6704 + }, + { + "epoch": 1.1937321937321936, + "grad_norm": 0.6667410135269165, + "learning_rate": 0.0001593342772629113, + "loss": 1.0073, + "step": 6705 + }, + { + "epoch": 1.1939102564102564, + "grad_norm": 0.6785695552825928, + "learning_rate": 0.00015932300941774944, + "loss": 1.0752, + "step": 6706 + }, + { + "epoch": 1.1940883190883191, + "grad_norm": 0.6446872353553772, + "learning_rate": 0.0001593117404102765, + "loss": 0.9509, + "step": 6707 + }, + { + "epoch": 1.194266381766382, + "grad_norm": 0.6607686877250671, + "learning_rate": 0.00015930047024071317, + "loss": 1.0902, + "step": 6708 + }, + { + "epoch": 1.1944444444444444, + "grad_norm": 0.664804995059967, + "learning_rate": 0.0001592891989092803, + "loss": 0.9783, + "step": 6709 + }, + { + "epoch": 1.1946225071225072, + "grad_norm": 0.7147907018661499, + "learning_rate": 0.00015927792641619876, + "loss": 1.0558, + "step": 6710 + }, + { + "epoch": 1.1948005698005697, + "grad_norm": 0.6858944296836853, + "learning_rate": 0.0001592666527616894, + "loss": 1.0514, + "step": 6711 + }, + { + "epoch": 1.1949786324786325, + "grad_norm": 0.598463773727417, + "learning_rate": 0.0001592553779459731, + "loss": 0.8927, + "step": 6712 + }, + { + "epoch": 1.1951566951566952, + "grad_norm": 0.6872668862342834, + "learning_rate": 0.00015924410196927076, + "loss": 1.016, + "step": 6713 + }, + { + "epoch": 1.195334757834758, + "grad_norm": 0.6547996401786804, + "learning_rate": 0.00015923282483180326, + "loss": 1.1573, + "step": 6714 + }, + { + "epoch": 1.1955128205128205, + "grad_norm": 0.6254705786705017, + "learning_rate": 0.00015922154653379167, + "loss": 1.0179, + "step": 6715 + }, + { + "epoch": 1.1956908831908832, + "grad_norm": 0.6049207448959351, + "learning_rate": 0.00015921026707545684, + "loss": 1.0713, + "step": 6716 + }, + { + "epoch": 1.1958689458689458, + "grad_norm": 0.6042858958244324, + "learning_rate": 0.0001591989864570199, + "loss": 0.919, + "step": 6717 + }, + { + "epoch": 1.1960470085470085, + "grad_norm": 0.6521187424659729, + "learning_rate": 0.0001591877046787017, + "loss": 1.0112, + "step": 6718 + }, + { + "epoch": 1.1962250712250713, + "grad_norm": 0.766260027885437, + "learning_rate": 0.00015917642174072348, + "loss": 0.9774, + "step": 6719 + }, + { + "epoch": 1.196403133903134, + "grad_norm": 0.7066532373428345, + "learning_rate": 0.00015916513764330613, + "loss": 1.1112, + "step": 6720 + }, + { + "epoch": 1.1965811965811965, + "grad_norm": 0.7351508140563965, + "learning_rate": 0.00015915385238667083, + "loss": 0.9841, + "step": 6721 + }, + { + "epoch": 1.1967592592592593, + "grad_norm": 0.6133812069892883, + "learning_rate": 0.0001591425659710387, + "loss": 0.8629, + "step": 6722 + }, + { + "epoch": 1.1969373219373218, + "grad_norm": 0.7244157791137695, + "learning_rate": 0.00015913127839663083, + "loss": 1.1584, + "step": 6723 + }, + { + "epoch": 1.1971153846153846, + "grad_norm": 0.5986210107803345, + "learning_rate": 0.00015911998966366842, + "loss": 0.8507, + "step": 6724 + }, + { + "epoch": 1.1972934472934473, + "grad_norm": 0.6087439060211182, + "learning_rate": 0.00015910869977237257, + "loss": 0.884, + "step": 6725 + }, + { + "epoch": 1.19747150997151, + "grad_norm": 0.7546007633209229, + "learning_rate": 0.00015909740872296457, + "loss": 1.1449, + "step": 6726 + }, + { + "epoch": 1.1976495726495726, + "grad_norm": 0.6437731385231018, + "learning_rate": 0.0001590861165156656, + "loss": 0.7845, + "step": 6727 + }, + { + "epoch": 1.1978276353276354, + "grad_norm": 0.6281737089157104, + "learning_rate": 0.00015907482315069693, + "loss": 0.8969, + "step": 6728 + }, + { + "epoch": 1.198005698005698, + "grad_norm": 0.6196113228797913, + "learning_rate": 0.00015906352862827983, + "loss": 1.0264, + "step": 6729 + }, + { + "epoch": 1.1981837606837606, + "grad_norm": 0.5990965962409973, + "learning_rate": 0.00015905223294863553, + "loss": 1.0017, + "step": 6730 + }, + { + "epoch": 1.1983618233618234, + "grad_norm": 0.6509191393852234, + "learning_rate": 0.00015904093611198542, + "loss": 1.1066, + "step": 6731 + }, + { + "epoch": 1.1985398860398861, + "grad_norm": 0.6648043990135193, + "learning_rate": 0.00015902963811855085, + "loss": 1.077, + "step": 6732 + }, + { + "epoch": 1.1987179487179487, + "grad_norm": 0.7071963548660278, + "learning_rate": 0.00015901833896855307, + "loss": 1.1346, + "step": 6733 + }, + { + "epoch": 1.1988960113960114, + "grad_norm": 0.5889959335327148, + "learning_rate": 0.0001590070386622136, + "loss": 0.9525, + "step": 6734 + }, + { + "epoch": 1.199074074074074, + "grad_norm": 0.6233037710189819, + "learning_rate": 0.00015899573719975376, + "loss": 1.0513, + "step": 6735 + }, + { + "epoch": 1.1992521367521367, + "grad_norm": 0.7912302613258362, + "learning_rate": 0.000158984434581395, + "loss": 0.8749, + "step": 6736 + }, + { + "epoch": 1.1994301994301995, + "grad_norm": 0.5783160924911499, + "learning_rate": 0.0001589731308073588, + "loss": 0.7173, + "step": 6737 + }, + { + "epoch": 1.1996082621082622, + "grad_norm": 0.718950092792511, + "learning_rate": 0.00015896182587786658, + "loss": 1.0815, + "step": 6738 + }, + { + "epoch": 1.1997863247863247, + "grad_norm": 0.6700926423072815, + "learning_rate": 0.0001589505197931399, + "loss": 1.0817, + "step": 6739 + }, + { + "epoch": 1.1999643874643875, + "grad_norm": 0.7614455223083496, + "learning_rate": 0.0001589392125534002, + "loss": 0.9707, + "step": 6740 + }, + { + "epoch": 1.20014245014245, + "grad_norm": 0.6998619437217712, + "learning_rate": 0.00015892790415886906, + "loss": 1.0541, + "step": 6741 + }, + { + "epoch": 1.2003205128205128, + "grad_norm": 0.6127668619155884, + "learning_rate": 0.0001589165946097681, + "loss": 0.9147, + "step": 6742 + }, + { + "epoch": 1.2004985754985755, + "grad_norm": 0.7112005352973938, + "learning_rate": 0.00015890528390631885, + "loss": 0.868, + "step": 6743 + }, + { + "epoch": 1.2006766381766383, + "grad_norm": 0.6631024479866028, + "learning_rate": 0.0001588939720487429, + "loss": 0.9277, + "step": 6744 + }, + { + "epoch": 1.2008547008547008, + "grad_norm": 0.6106321215629578, + "learning_rate": 0.00015888265903726188, + "loss": 1.0223, + "step": 6745 + }, + { + "epoch": 1.2010327635327636, + "grad_norm": 0.6400851607322693, + "learning_rate": 0.00015887134487209753, + "loss": 1.1279, + "step": 6746 + }, + { + "epoch": 1.201210826210826, + "grad_norm": 0.6298650503158569, + "learning_rate": 0.00015886002955347147, + "loss": 0.9481, + "step": 6747 + }, + { + "epoch": 1.2013888888888888, + "grad_norm": 0.647974967956543, + "learning_rate": 0.00015884871308160538, + "loss": 1.1513, + "step": 6748 + }, + { + "epoch": 1.2015669515669516, + "grad_norm": 0.6770651936531067, + "learning_rate": 0.000158837395456721, + "loss": 0.9914, + "step": 6749 + }, + { + "epoch": 1.2017450142450143, + "grad_norm": 0.6708947420120239, + "learning_rate": 0.0001588260766790401, + "loss": 1.1848, + "step": 6750 + }, + { + "epoch": 1.2019230769230769, + "grad_norm": 0.5624440908432007, + "learning_rate": 0.00015881475674878442, + "loss": 0.9848, + "step": 6751 + }, + { + "epoch": 1.2021011396011396, + "grad_norm": 0.5512633919715881, + "learning_rate": 0.00015880343566617575, + "loss": 1.0308, + "step": 6752 + }, + { + "epoch": 1.2022792022792024, + "grad_norm": 0.5621042251586914, + "learning_rate": 0.0001587921134314359, + "loss": 0.8724, + "step": 6753 + }, + { + "epoch": 1.202457264957265, + "grad_norm": 0.6881251931190491, + "learning_rate": 0.00015878079004478675, + "loss": 0.9771, + "step": 6754 + }, + { + "epoch": 1.2026353276353277, + "grad_norm": 0.729998767375946, + "learning_rate": 0.0001587694655064501, + "loss": 1.002, + "step": 6755 + }, + { + "epoch": 1.2028133903133904, + "grad_norm": 0.5972567200660706, + "learning_rate": 0.00015875813981664787, + "loss": 1.0571, + "step": 6756 + }, + { + "epoch": 1.202991452991453, + "grad_norm": 0.6319229006767273, + "learning_rate": 0.00015874681297560196, + "loss": 0.9294, + "step": 6757 + }, + { + "epoch": 1.2031695156695157, + "grad_norm": 0.6751521825790405, + "learning_rate": 0.00015873548498353428, + "loss": 0.783, + "step": 6758 + }, + { + "epoch": 1.2033475783475784, + "grad_norm": 0.6476554870605469, + "learning_rate": 0.00015872415584066677, + "loss": 0.8939, + "step": 6759 + }, + { + "epoch": 1.203525641025641, + "grad_norm": 0.6530960202217102, + "learning_rate": 0.0001587128255472214, + "loss": 0.9828, + "step": 6760 + }, + { + "epoch": 1.2037037037037037, + "grad_norm": 0.6708502173423767, + "learning_rate": 0.00015870149410342023, + "loss": 0.9285, + "step": 6761 + }, + { + "epoch": 1.2038817663817665, + "grad_norm": 0.7749543190002441, + "learning_rate": 0.0001586901615094852, + "loss": 1.1295, + "step": 6762 + }, + { + "epoch": 1.204059829059829, + "grad_norm": 0.6750495433807373, + "learning_rate": 0.00015867882776563836, + "loss": 1.0562, + "step": 6763 + }, + { + "epoch": 1.2042378917378918, + "grad_norm": 0.6892416477203369, + "learning_rate": 0.00015866749287210178, + "loss": 0.7207, + "step": 6764 + }, + { + "epoch": 1.2044159544159545, + "grad_norm": 0.7066485285758972, + "learning_rate": 0.00015865615682909758, + "loss": 1.0489, + "step": 6765 + }, + { + "epoch": 1.204594017094017, + "grad_norm": 0.5669938325881958, + "learning_rate": 0.00015864481963684783, + "loss": 0.8149, + "step": 6766 + }, + { + "epoch": 1.2047720797720798, + "grad_norm": 0.6467341780662537, + "learning_rate": 0.0001586334812955746, + "loss": 0.9595, + "step": 6767 + }, + { + "epoch": 1.2049501424501425, + "grad_norm": 0.6026045680046082, + "learning_rate": 0.0001586221418055002, + "loss": 0.9832, + "step": 6768 + }, + { + "epoch": 1.205128205128205, + "grad_norm": 0.7655174732208252, + "learning_rate": 0.00015861080116684665, + "loss": 0.9796, + "step": 6769 + }, + { + "epoch": 1.2053062678062678, + "grad_norm": 0.6386621594429016, + "learning_rate": 0.00015859945937983624, + "loss": 0.9368, + "step": 6770 + }, + { + "epoch": 1.2054843304843306, + "grad_norm": 0.7088032364845276, + "learning_rate": 0.0001585881164446911, + "loss": 1.0167, + "step": 6771 + }, + { + "epoch": 1.205662393162393, + "grad_norm": 0.6015275716781616, + "learning_rate": 0.0001585767723616336, + "loss": 0.8551, + "step": 6772 + }, + { + "epoch": 1.2058404558404558, + "grad_norm": 0.7013260722160339, + "learning_rate": 0.00015856542713088583, + "loss": 0.8009, + "step": 6773 + }, + { + "epoch": 1.2060185185185186, + "grad_norm": 0.6931240558624268, + "learning_rate": 0.00015855408075267024, + "loss": 0.9964, + "step": 6774 + }, + { + "epoch": 1.2061965811965811, + "grad_norm": 0.7274388670921326, + "learning_rate": 0.00015854273322720908, + "loss": 1.0991, + "step": 6775 + }, + { + "epoch": 1.2063746438746439, + "grad_norm": 0.6353716254234314, + "learning_rate": 0.00015853138455472466, + "loss": 1.0893, + "step": 6776 + }, + { + "epoch": 1.2065527065527066, + "grad_norm": 0.6958979368209839, + "learning_rate": 0.00015852003473543932, + "loss": 1.0238, + "step": 6777 + }, + { + "epoch": 1.2067307692307692, + "grad_norm": 0.626838743686676, + "learning_rate": 0.00015850868376957551, + "loss": 0.9384, + "step": 6778 + }, + { + "epoch": 1.206908831908832, + "grad_norm": 0.5455024242401123, + "learning_rate": 0.00015849733165735556, + "loss": 0.8068, + "step": 6779 + }, + { + "epoch": 1.2070868945868947, + "grad_norm": 0.6337353587150574, + "learning_rate": 0.0001584859783990019, + "loss": 1.1341, + "step": 6780 + }, + { + "epoch": 1.2072649572649572, + "grad_norm": 0.6318019032478333, + "learning_rate": 0.000158474623994737, + "loss": 1.1095, + "step": 6781 + }, + { + "epoch": 1.20744301994302, + "grad_norm": 0.8183810710906982, + "learning_rate": 0.00015846326844478332, + "loss": 1.1471, + "step": 6782 + }, + { + "epoch": 1.2076210826210827, + "grad_norm": 0.6140483021736145, + "learning_rate": 0.00015845191174936334, + "loss": 0.8538, + "step": 6783 + }, + { + "epoch": 1.2077991452991452, + "grad_norm": 0.7570197582244873, + "learning_rate": 0.0001584405539086996, + "loss": 1.427, + "step": 6784 + }, + { + "epoch": 1.207977207977208, + "grad_norm": 0.7616991996765137, + "learning_rate": 0.00015842919492301455, + "loss": 1.2214, + "step": 6785 + }, + { + "epoch": 1.2081552706552707, + "grad_norm": 0.561996579170227, + "learning_rate": 0.00015841783479253084, + "loss": 0.8916, + "step": 6786 + }, + { + "epoch": 1.2083333333333333, + "grad_norm": 0.6124222874641418, + "learning_rate": 0.000158406473517471, + "loss": 0.9637, + "step": 6787 + }, + { + "epoch": 1.208511396011396, + "grad_norm": 0.6053098440170288, + "learning_rate": 0.00015839511109805762, + "loss": 1.0365, + "step": 6788 + }, + { + "epoch": 1.2086894586894588, + "grad_norm": 0.6451675295829773, + "learning_rate": 0.00015838374753451338, + "loss": 1.0497, + "step": 6789 + }, + { + "epoch": 1.2088675213675213, + "grad_norm": 0.6789399981498718, + "learning_rate": 0.00015837238282706087, + "loss": 0.9286, + "step": 6790 + }, + { + "epoch": 1.209045584045584, + "grad_norm": 0.5742998123168945, + "learning_rate": 0.0001583610169759228, + "loss": 1.082, + "step": 6791 + }, + { + "epoch": 1.2092236467236468, + "grad_norm": 0.6813693642616272, + "learning_rate": 0.0001583496499813218, + "loss": 0.9785, + "step": 6792 + }, + { + "epoch": 1.2094017094017093, + "grad_norm": 0.6150603890419006, + "learning_rate": 0.0001583382818434806, + "loss": 0.9533, + "step": 6793 + }, + { + "epoch": 1.209579772079772, + "grad_norm": 0.6905919909477234, + "learning_rate": 0.000158326912562622, + "loss": 1.0132, + "step": 6794 + }, + { + "epoch": 1.2097578347578348, + "grad_norm": 0.5861411094665527, + "learning_rate": 0.0001583155421389687, + "loss": 0.7071, + "step": 6795 + }, + { + "epoch": 1.2099358974358974, + "grad_norm": 0.6822740435600281, + "learning_rate": 0.0001583041705727435, + "loss": 1.1366, + "step": 6796 + }, + { + "epoch": 1.21011396011396, + "grad_norm": 0.6013675928115845, + "learning_rate": 0.00015829279786416916, + "loss": 0.9232, + "step": 6797 + }, + { + "epoch": 1.2102920227920229, + "grad_norm": 0.650675356388092, + "learning_rate": 0.00015828142401346857, + "loss": 0.887, + "step": 6798 + }, + { + "epoch": 1.2104700854700854, + "grad_norm": 0.6764078736305237, + "learning_rate": 0.00015827004902086456, + "loss": 0.8423, + "step": 6799 + }, + { + "epoch": 1.2106481481481481, + "grad_norm": 0.6460821628570557, + "learning_rate": 0.00015825867288657994, + "loss": 1.0074, + "step": 6800 + }, + { + "epoch": 1.210826210826211, + "grad_norm": 0.692562997341156, + "learning_rate": 0.00015824729561083768, + "loss": 0.7978, + "step": 6801 + }, + { + "epoch": 1.2110042735042734, + "grad_norm": 0.7255034446716309, + "learning_rate": 0.00015823591719386066, + "loss": 1.071, + "step": 6802 + }, + { + "epoch": 1.2111823361823362, + "grad_norm": 0.6598904728889465, + "learning_rate": 0.0001582245376358718, + "loss": 0.9736, + "step": 6803 + }, + { + "epoch": 1.211360398860399, + "grad_norm": 0.6372483968734741, + "learning_rate": 0.0001582131569370941, + "loss": 0.9029, + "step": 6804 + }, + { + "epoch": 1.2115384615384615, + "grad_norm": 0.5907173156738281, + "learning_rate": 0.00015820177509775048, + "loss": 0.918, + "step": 6805 + }, + { + "epoch": 1.2117165242165242, + "grad_norm": 0.6252630949020386, + "learning_rate": 0.00015819039211806404, + "loss": 0.7801, + "step": 6806 + }, + { + "epoch": 1.211894586894587, + "grad_norm": 0.5793096423149109, + "learning_rate": 0.0001581790079982577, + "loss": 0.5769, + "step": 6807 + }, + { + "epoch": 1.2120726495726495, + "grad_norm": 0.7267270684242249, + "learning_rate": 0.00015816762273855454, + "loss": 1.1428, + "step": 6808 + }, + { + "epoch": 1.2122507122507122, + "grad_norm": 0.7481234073638916, + "learning_rate": 0.00015815623633917767, + "loss": 1.0209, + "step": 6809 + }, + { + "epoch": 1.212428774928775, + "grad_norm": 0.6114386916160583, + "learning_rate": 0.00015814484880035017, + "loss": 0.9073, + "step": 6810 + }, + { + "epoch": 1.2126068376068375, + "grad_norm": 0.6871182322502136, + "learning_rate": 0.00015813346012229516, + "loss": 1.151, + "step": 6811 + }, + { + "epoch": 1.2127849002849003, + "grad_norm": 0.6380293965339661, + "learning_rate": 0.0001581220703052357, + "loss": 1.0981, + "step": 6812 + }, + { + "epoch": 1.212962962962963, + "grad_norm": 0.6013718247413635, + "learning_rate": 0.00015811067934939503, + "loss": 0.8832, + "step": 6813 + }, + { + "epoch": 1.2131410256410255, + "grad_norm": 0.5816897749900818, + "learning_rate": 0.00015809928725499632, + "loss": 1.063, + "step": 6814 + }, + { + "epoch": 1.2133190883190883, + "grad_norm": 0.5970914363861084, + "learning_rate": 0.00015808789402226278, + "loss": 1.1177, + "step": 6815 + }, + { + "epoch": 1.213497150997151, + "grad_norm": 0.7624936103820801, + "learning_rate": 0.00015807649965141762, + "loss": 1.048, + "step": 6816 + }, + { + "epoch": 1.2136752136752136, + "grad_norm": 0.636263906955719, + "learning_rate": 0.0001580651041426841, + "loss": 0.9743, + "step": 6817 + }, + { + "epoch": 1.2138532763532763, + "grad_norm": 0.641090452671051, + "learning_rate": 0.00015805370749628547, + "loss": 1.0227, + "step": 6818 + }, + { + "epoch": 1.214031339031339, + "grad_norm": 0.6484021544456482, + "learning_rate": 0.00015804230971244504, + "loss": 0.9615, + "step": 6819 + }, + { + "epoch": 1.2142094017094016, + "grad_norm": 0.6473353505134583, + "learning_rate": 0.00015803091079138613, + "loss": 1.0507, + "step": 6820 + }, + { + "epoch": 1.2143874643874644, + "grad_norm": 0.5477129220962524, + "learning_rate": 0.00015801951073333206, + "loss": 0.7928, + "step": 6821 + }, + { + "epoch": 1.2145655270655271, + "grad_norm": 0.7256210446357727, + "learning_rate": 0.0001580081095385062, + "loss": 1.0172, + "step": 6822 + }, + { + "epoch": 1.2147435897435896, + "grad_norm": 0.5785418748855591, + "learning_rate": 0.00015799670720713195, + "loss": 0.8478, + "step": 6823 + }, + { + "epoch": 1.2149216524216524, + "grad_norm": 0.6782996654510498, + "learning_rate": 0.00015798530373943267, + "loss": 1.1819, + "step": 6824 + }, + { + "epoch": 1.2150997150997151, + "grad_norm": 0.6513699293136597, + "learning_rate": 0.00015797389913563186, + "loss": 0.9626, + "step": 6825 + }, + { + "epoch": 1.2152777777777777, + "grad_norm": 0.6503037214279175, + "learning_rate": 0.0001579624933959529, + "loss": 1.0282, + "step": 6826 + }, + { + "epoch": 1.2154558404558404, + "grad_norm": 0.581501841545105, + "learning_rate": 0.0001579510865206193, + "loss": 0.8976, + "step": 6827 + }, + { + "epoch": 1.2156339031339032, + "grad_norm": 0.6696721911430359, + "learning_rate": 0.00015793967850985454, + "loss": 0.6418, + "step": 6828 + }, + { + "epoch": 1.215811965811966, + "grad_norm": 0.6577274203300476, + "learning_rate": 0.00015792826936388213, + "loss": 1.0615, + "step": 6829 + }, + { + "epoch": 1.2159900284900285, + "grad_norm": 0.66291743516922, + "learning_rate": 0.00015791685908292564, + "loss": 0.8582, + "step": 6830 + }, + { + "epoch": 1.2161680911680912, + "grad_norm": 0.6548362374305725, + "learning_rate": 0.0001579054476672086, + "loss": 1.0343, + "step": 6831 + }, + { + "epoch": 1.2163461538461537, + "grad_norm": 0.6381218433380127, + "learning_rate": 0.00015789403511695457, + "loss": 0.8133, + "step": 6832 + }, + { + "epoch": 1.2165242165242165, + "grad_norm": 0.7217492461204529, + "learning_rate": 0.00015788262143238722, + "loss": 0.9183, + "step": 6833 + }, + { + "epoch": 1.2167022792022792, + "grad_norm": 0.610454797744751, + "learning_rate": 0.00015787120661373013, + "loss": 0.8488, + "step": 6834 + }, + { + "epoch": 1.216880341880342, + "grad_norm": 0.592771053314209, + "learning_rate": 0.00015785979066120696, + "loss": 0.8673, + "step": 6835 + }, + { + "epoch": 1.2170584045584045, + "grad_norm": 0.5787834525108337, + "learning_rate": 0.00015784837357504138, + "loss": 0.7945, + "step": 6836 + }, + { + "epoch": 1.2172364672364673, + "grad_norm": 0.6814196109771729, + "learning_rate": 0.0001578369553554571, + "loss": 0.8906, + "step": 6837 + }, + { + "epoch": 1.2174145299145298, + "grad_norm": 0.6383981108665466, + "learning_rate": 0.00015782553600267787, + "loss": 0.8962, + "step": 6838 + }, + { + "epoch": 1.2175925925925926, + "grad_norm": 0.6733864545822144, + "learning_rate": 0.0001578141155169273, + "loss": 1.2077, + "step": 6839 + }, + { + "epoch": 1.2177706552706553, + "grad_norm": 0.5891284346580505, + "learning_rate": 0.0001578026938984293, + "loss": 0.9477, + "step": 6840 + }, + { + "epoch": 1.217948717948718, + "grad_norm": 0.7220266461372375, + "learning_rate": 0.00015779127114740757, + "loss": 1.0343, + "step": 6841 + }, + { + "epoch": 1.2181267806267806, + "grad_norm": 0.6566546559333801, + "learning_rate": 0.0001577798472640859, + "loss": 0.9576, + "step": 6842 + }, + { + "epoch": 1.2183048433048433, + "grad_norm": 0.6428449153900146, + "learning_rate": 0.0001577684222486882, + "loss": 0.8957, + "step": 6843 + }, + { + "epoch": 1.2184829059829059, + "grad_norm": 0.6542909741401672, + "learning_rate": 0.00015775699610143823, + "loss": 0.9942, + "step": 6844 + }, + { + "epoch": 1.2186609686609686, + "grad_norm": 0.7101675868034363, + "learning_rate": 0.00015774556882255992, + "loss": 1.015, + "step": 6845 + }, + { + "epoch": 1.2188390313390314, + "grad_norm": 0.6606267094612122, + "learning_rate": 0.00015773414041227713, + "loss": 1.1406, + "step": 6846 + }, + { + "epoch": 1.2190170940170941, + "grad_norm": 0.67124342918396, + "learning_rate": 0.00015772271087081383, + "loss": 1.2392, + "step": 6847 + }, + { + "epoch": 1.2191951566951567, + "grad_norm": 0.6615056991577148, + "learning_rate": 0.0001577112801983939, + "loss": 1.1583, + "step": 6848 + }, + { + "epoch": 1.2193732193732194, + "grad_norm": 0.6941317319869995, + "learning_rate": 0.0001576998483952413, + "loss": 1.0255, + "step": 6849 + }, + { + "epoch": 1.219551282051282, + "grad_norm": 0.5740683674812317, + "learning_rate": 0.00015768841546158005, + "loss": 1.0393, + "step": 6850 + }, + { + "epoch": 1.2197293447293447, + "grad_norm": 0.7143667340278625, + "learning_rate": 0.00015767698139763415, + "loss": 0.7564, + "step": 6851 + }, + { + "epoch": 1.2199074074074074, + "grad_norm": 0.6730484366416931, + "learning_rate": 0.00015766554620362758, + "loss": 1.2221, + "step": 6852 + }, + { + "epoch": 1.2200854700854702, + "grad_norm": 0.6883087754249573, + "learning_rate": 0.00015765410987978444, + "loss": 1.0156, + "step": 6853 + }, + { + "epoch": 1.2202635327635327, + "grad_norm": 0.6585961580276489, + "learning_rate": 0.00015764267242632875, + "loss": 1.0888, + "step": 6854 + }, + { + "epoch": 1.2204415954415955, + "grad_norm": 0.6325246691703796, + "learning_rate": 0.00015763123384348465, + "loss": 0.973, + "step": 6855 + }, + { + "epoch": 1.220619658119658, + "grad_norm": 0.5930588245391846, + "learning_rate": 0.00015761979413147627, + "loss": 0.8551, + "step": 6856 + }, + { + "epoch": 1.2207977207977208, + "grad_norm": 0.6440611481666565, + "learning_rate": 0.0001576083532905277, + "loss": 0.8396, + "step": 6857 + }, + { + "epoch": 1.2209757834757835, + "grad_norm": 0.6796659231185913, + "learning_rate": 0.00015759691132086315, + "loss": 1.0662, + "step": 6858 + }, + { + "epoch": 1.2211538461538463, + "grad_norm": 0.6813400983810425, + "learning_rate": 0.00015758546822270674, + "loss": 1.0457, + "step": 6859 + }, + { + "epoch": 1.2213319088319088, + "grad_norm": 0.6871716976165771, + "learning_rate": 0.00015757402399628272, + "loss": 1.1675, + "step": 6860 + }, + { + "epoch": 1.2215099715099715, + "grad_norm": 0.6431481838226318, + "learning_rate": 0.00015756257864181524, + "loss": 0.9366, + "step": 6861 + }, + { + "epoch": 1.221688034188034, + "grad_norm": 0.6061800718307495, + "learning_rate": 0.00015755113215952868, + "loss": 0.9267, + "step": 6862 + }, + { + "epoch": 1.2218660968660968, + "grad_norm": 0.5755770206451416, + "learning_rate": 0.00015753968454964722, + "loss": 0.7342, + "step": 6863 + }, + { + "epoch": 1.2220441595441596, + "grad_norm": 0.571345329284668, + "learning_rate": 0.00015752823581239515, + "loss": 0.8943, + "step": 6864 + }, + { + "epoch": 1.2222222222222223, + "grad_norm": 0.6925615668296814, + "learning_rate": 0.0001575167859479968, + "loss": 0.8801, + "step": 6865 + }, + { + "epoch": 1.2224002849002849, + "grad_norm": 0.6812975406646729, + "learning_rate": 0.00015750533495667655, + "loss": 0.9567, + "step": 6866 + }, + { + "epoch": 1.2225783475783476, + "grad_norm": 0.8216777443885803, + "learning_rate": 0.00015749388283865868, + "loss": 1.0908, + "step": 6867 + }, + { + "epoch": 1.2227564102564104, + "grad_norm": 0.6051010489463806, + "learning_rate": 0.00015748242959416763, + "loss": 0.8851, + "step": 6868 + }, + { + "epoch": 1.2229344729344729, + "grad_norm": 0.7750816345214844, + "learning_rate": 0.00015747097522342775, + "loss": 1.1526, + "step": 6869 + }, + { + "epoch": 1.2231125356125356, + "grad_norm": 0.6240930557250977, + "learning_rate": 0.00015745951972666355, + "loss": 1.0603, + "step": 6870 + }, + { + "epoch": 1.2232905982905984, + "grad_norm": 0.7228875160217285, + "learning_rate": 0.00015744806310409937, + "loss": 1.1028, + "step": 6871 + }, + { + "epoch": 1.223468660968661, + "grad_norm": 0.724075436592102, + "learning_rate": 0.00015743660535595978, + "loss": 0.8983, + "step": 6872 + }, + { + "epoch": 1.2236467236467237, + "grad_norm": 0.6398203372955322, + "learning_rate": 0.00015742514648246916, + "loss": 1.0548, + "step": 6873 + }, + { + "epoch": 1.2238247863247864, + "grad_norm": 0.7024285793304443, + "learning_rate": 0.00015741368648385212, + "loss": 1.0172, + "step": 6874 + }, + { + "epoch": 1.224002849002849, + "grad_norm": 0.6717609763145447, + "learning_rate": 0.00015740222536033316, + "loss": 0.9002, + "step": 6875 + }, + { + "epoch": 1.2241809116809117, + "grad_norm": 0.5886133313179016, + "learning_rate": 0.00015739076311213686, + "loss": 0.8614, + "step": 6876 + }, + { + "epoch": 1.2243589743589745, + "grad_norm": 0.6856684684753418, + "learning_rate": 0.00015737929973948776, + "loss": 1.1633, + "step": 6877 + }, + { + "epoch": 1.224537037037037, + "grad_norm": 0.6771421432495117, + "learning_rate": 0.00015736783524261045, + "loss": 1.0921, + "step": 6878 + }, + { + "epoch": 1.2247150997150997, + "grad_norm": 0.5016412138938904, + "learning_rate": 0.0001573563696217296, + "loss": 0.6732, + "step": 6879 + }, + { + "epoch": 1.2248931623931625, + "grad_norm": 0.7595276236534119, + "learning_rate": 0.00015734490287706984, + "loss": 1.0427, + "step": 6880 + }, + { + "epoch": 1.225071225071225, + "grad_norm": 0.6664281487464905, + "learning_rate": 0.00015733343500885582, + "loss": 1.2836, + "step": 6881 + }, + { + "epoch": 1.2252492877492878, + "grad_norm": 0.6662577390670776, + "learning_rate": 0.00015732196601731224, + "loss": 1.1288, + "step": 6882 + }, + { + "epoch": 1.2254273504273505, + "grad_norm": 0.6238988041877747, + "learning_rate": 0.00015731049590266385, + "loss": 1.0809, + "step": 6883 + }, + { + "epoch": 1.225605413105413, + "grad_norm": 0.6483062505722046, + "learning_rate": 0.00015729902466513532, + "loss": 0.9992, + "step": 6884 + }, + { + "epoch": 1.2257834757834758, + "grad_norm": 0.6890861988067627, + "learning_rate": 0.0001572875523049514, + "loss": 1.1844, + "step": 6885 + }, + { + "epoch": 1.2259615384615385, + "grad_norm": 0.7087607383728027, + "learning_rate": 0.00015727607882233695, + "loss": 1.013, + "step": 6886 + }, + { + "epoch": 1.226139601139601, + "grad_norm": 0.709048867225647, + "learning_rate": 0.00015726460421751668, + "loss": 0.9748, + "step": 6887 + }, + { + "epoch": 1.2263176638176638, + "grad_norm": 0.5918150544166565, + "learning_rate": 0.00015725312849071546, + "loss": 0.9978, + "step": 6888 + }, + { + "epoch": 1.2264957264957266, + "grad_norm": 0.4343377947807312, + "learning_rate": 0.0001572416516421581, + "loss": 0.6233, + "step": 6889 + }, + { + "epoch": 1.226673789173789, + "grad_norm": 0.6360403895378113, + "learning_rate": 0.00015723017367206952, + "loss": 0.9698, + "step": 6890 + }, + { + "epoch": 1.2268518518518519, + "grad_norm": 0.7261984944343567, + "learning_rate": 0.00015721869458067454, + "loss": 1.0426, + "step": 6891 + }, + { + "epoch": 1.2270299145299146, + "grad_norm": 0.6806774139404297, + "learning_rate": 0.0001572072143681981, + "loss": 0.9692, + "step": 6892 + }, + { + "epoch": 1.2272079772079771, + "grad_norm": 0.7140612006187439, + "learning_rate": 0.00015719573303486515, + "loss": 1.0828, + "step": 6893 + }, + { + "epoch": 1.22738603988604, + "grad_norm": 0.5383326411247253, + "learning_rate": 0.0001571842505809006, + "loss": 1.012, + "step": 6894 + }, + { + "epoch": 1.2275641025641026, + "grad_norm": 0.5992259383201599, + "learning_rate": 0.0001571727670065295, + "loss": 0.876, + "step": 6895 + }, + { + "epoch": 1.2277421652421652, + "grad_norm": 0.636696457862854, + "learning_rate": 0.00015716128231197676, + "loss": 1.1001, + "step": 6896 + }, + { + "epoch": 1.227920227920228, + "grad_norm": 0.5980371236801147, + "learning_rate": 0.00015714979649746744, + "loss": 0.937, + "step": 6897 + }, + { + "epoch": 1.2280982905982907, + "grad_norm": 0.7678794860839844, + "learning_rate": 0.00015713830956322656, + "loss": 1.1965, + "step": 6898 + }, + { + "epoch": 1.2282763532763532, + "grad_norm": 0.6918835639953613, + "learning_rate": 0.00015712682150947923, + "loss": 0.8578, + "step": 6899 + }, + { + "epoch": 1.228454415954416, + "grad_norm": 0.6463451385498047, + "learning_rate": 0.00015711533233645048, + "loss": 1.009, + "step": 6900 + }, + { + "epoch": 1.2286324786324787, + "grad_norm": 0.6720646023750305, + "learning_rate": 0.00015710384204436549, + "loss": 1.0031, + "step": 6901 + }, + { + "epoch": 1.2288105413105412, + "grad_norm": 0.6618736982345581, + "learning_rate": 0.00015709235063344926, + "loss": 0.9017, + "step": 6902 + }, + { + "epoch": 1.228988603988604, + "grad_norm": 0.6789427399635315, + "learning_rate": 0.0001570808581039271, + "loss": 1.1289, + "step": 6903 + }, + { + "epoch": 1.2291666666666667, + "grad_norm": 0.6395950317382812, + "learning_rate": 0.00015706936445602403, + "loss": 1.1051, + "step": 6904 + }, + { + "epoch": 1.2293447293447293, + "grad_norm": 0.7023917436599731, + "learning_rate": 0.00015705786968996533, + "loss": 1.2876, + "step": 6905 + }, + { + "epoch": 1.229522792022792, + "grad_norm": 0.7473352551460266, + "learning_rate": 0.00015704637380597623, + "loss": 1.237, + "step": 6906 + }, + { + "epoch": 1.2297008547008548, + "grad_norm": 0.6952672004699707, + "learning_rate": 0.00015703487680428192, + "loss": 1.0674, + "step": 6907 + }, + { + "epoch": 1.2298789173789173, + "grad_norm": 0.5968644022941589, + "learning_rate": 0.0001570233786851077, + "loss": 0.9169, + "step": 6908 + }, + { + "epoch": 1.23005698005698, + "grad_norm": 0.7219798564910889, + "learning_rate": 0.0001570118794486788, + "loss": 1.0556, + "step": 6909 + }, + { + "epoch": 1.2302350427350428, + "grad_norm": 0.6603400707244873, + "learning_rate": 0.0001570003790952206, + "loss": 0.9596, + "step": 6910 + }, + { + "epoch": 1.2304131054131053, + "grad_norm": 0.5972838401794434, + "learning_rate": 0.0001569888776249583, + "loss": 0.9168, + "step": 6911 + }, + { + "epoch": 1.230591168091168, + "grad_norm": 0.792585551738739, + "learning_rate": 0.00015697737503811738, + "loss": 1.1074, + "step": 6912 + }, + { + "epoch": 1.2307692307692308, + "grad_norm": 0.5845609903335571, + "learning_rate": 0.00015696587133492314, + "loss": 0.8413, + "step": 6913 + }, + { + "epoch": 1.2309472934472934, + "grad_norm": 0.6603896021842957, + "learning_rate": 0.000156954366515601, + "loss": 0.9109, + "step": 6914 + }, + { + "epoch": 1.2311253561253561, + "grad_norm": 0.6367142796516418, + "learning_rate": 0.00015694286058037636, + "loss": 1.0119, + "step": 6915 + }, + { + "epoch": 1.2313034188034189, + "grad_norm": 0.693854570388794, + "learning_rate": 0.00015693135352947465, + "loss": 1.0925, + "step": 6916 + }, + { + "epoch": 1.2314814814814814, + "grad_norm": 0.6570404171943665, + "learning_rate": 0.00015691984536312135, + "loss": 0.9731, + "step": 6917 + }, + { + "epoch": 1.2316595441595442, + "grad_norm": 0.6778639554977417, + "learning_rate": 0.0001569083360815419, + "loss": 1.1415, + "step": 6918 + }, + { + "epoch": 1.231837606837607, + "grad_norm": 0.6656233668327332, + "learning_rate": 0.00015689682568496182, + "loss": 0.8603, + "step": 6919 + }, + { + "epoch": 1.2320156695156694, + "grad_norm": 0.6569861173629761, + "learning_rate": 0.00015688531417360665, + "loss": 0.8374, + "step": 6920 + }, + { + "epoch": 1.2321937321937322, + "grad_norm": 0.6746888160705566, + "learning_rate": 0.0001568738015477019, + "loss": 1.1395, + "step": 6921 + }, + { + "epoch": 1.232371794871795, + "grad_norm": 0.6180813908576965, + "learning_rate": 0.00015686228780747316, + "loss": 1.0049, + "step": 6922 + }, + { + "epoch": 1.2325498575498575, + "grad_norm": 0.7326146960258484, + "learning_rate": 0.000156850772953146, + "loss": 1.2389, + "step": 6923 + }, + { + "epoch": 1.2327279202279202, + "grad_norm": 0.5912215709686279, + "learning_rate": 0.00015683925698494608, + "loss": 1.0174, + "step": 6924 + }, + { + "epoch": 1.232905982905983, + "grad_norm": 0.5214745402336121, + "learning_rate": 0.00015682773990309895, + "loss": 0.5778, + "step": 6925 + }, + { + "epoch": 1.2330840455840455, + "grad_norm": 0.6862079501152039, + "learning_rate": 0.00015681622170783034, + "loss": 0.896, + "step": 6926 + }, + { + "epoch": 1.2332621082621082, + "grad_norm": 0.7858926057815552, + "learning_rate": 0.00015680470239936586, + "loss": 1.0714, + "step": 6927 + }, + { + "epoch": 1.233440170940171, + "grad_norm": 0.6706146597862244, + "learning_rate": 0.00015679318197793127, + "loss": 1.0157, + "step": 6928 + }, + { + "epoch": 1.2336182336182335, + "grad_norm": 0.6657105088233948, + "learning_rate": 0.00015678166044375225, + "loss": 0.9674, + "step": 6929 + }, + { + "epoch": 1.2337962962962963, + "grad_norm": 0.6790838837623596, + "learning_rate": 0.0001567701377970545, + "loss": 0.9744, + "step": 6930 + }, + { + "epoch": 1.233974358974359, + "grad_norm": 0.6469771862030029, + "learning_rate": 0.00015675861403806386, + "loss": 1.0205, + "step": 6931 + }, + { + "epoch": 1.2341524216524216, + "grad_norm": 0.4926300346851349, + "learning_rate": 0.0001567470891670061, + "loss": 0.6336, + "step": 6932 + }, + { + "epoch": 1.2343304843304843, + "grad_norm": 0.6762157082557678, + "learning_rate": 0.000156735563184107, + "loss": 1.059, + "step": 6933 + }, + { + "epoch": 1.234508547008547, + "grad_norm": 0.6998521685600281, + "learning_rate": 0.0001567240360895924, + "loss": 1.0586, + "step": 6934 + }, + { + "epoch": 1.2346866096866096, + "grad_norm": 0.5947706699371338, + "learning_rate": 0.00015671250788368814, + "loss": 0.8815, + "step": 6935 + }, + { + "epoch": 1.2348646723646723, + "grad_norm": 0.6966122984886169, + "learning_rate": 0.0001567009785666201, + "loss": 1.0105, + "step": 6936 + }, + { + "epoch": 1.235042735042735, + "grad_norm": 0.6747866272926331, + "learning_rate": 0.0001566894481386142, + "loss": 0.8783, + "step": 6937 + }, + { + "epoch": 1.2352207977207976, + "grad_norm": 0.6348921060562134, + "learning_rate": 0.0001566779165998963, + "loss": 0.7813, + "step": 6938 + }, + { + "epoch": 1.2353988603988604, + "grad_norm": 0.596466600894928, + "learning_rate": 0.00015666638395069236, + "loss": 0.8689, + "step": 6939 + }, + { + "epoch": 1.2355769230769231, + "grad_norm": 0.6926795244216919, + "learning_rate": 0.00015665485019122834, + "loss": 1.0266, + "step": 6940 + }, + { + "epoch": 1.2357549857549857, + "grad_norm": 0.6590100526809692, + "learning_rate": 0.00015664331532173022, + "loss": 1.128, + "step": 6941 + }, + { + "epoch": 1.2359330484330484, + "grad_norm": 0.7422109246253967, + "learning_rate": 0.00015663177934242402, + "loss": 0.8495, + "step": 6942 + }, + { + "epoch": 1.2361111111111112, + "grad_norm": 0.6463228464126587, + "learning_rate": 0.0001566202422535357, + "loss": 1.0941, + "step": 6943 + }, + { + "epoch": 1.236289173789174, + "grad_norm": 0.7278686761856079, + "learning_rate": 0.0001566087040552914, + "loss": 1.2039, + "step": 6944 + }, + { + "epoch": 1.2364672364672364, + "grad_norm": 0.6917086839675903, + "learning_rate": 0.00015659716474791712, + "loss": 1.042, + "step": 6945 + }, + { + "epoch": 1.2366452991452992, + "grad_norm": 0.637205183506012, + "learning_rate": 0.00015658562433163898, + "loss": 1.0379, + "step": 6946 + }, + { + "epoch": 1.2368233618233617, + "grad_norm": 0.6706623435020447, + "learning_rate": 0.00015657408280668307, + "loss": 1.0347, + "step": 6947 + }, + { + "epoch": 1.2370014245014245, + "grad_norm": 0.6435480713844299, + "learning_rate": 0.00015656254017327553, + "loss": 0.7708, + "step": 6948 + }, + { + "epoch": 1.2371794871794872, + "grad_norm": 0.5703113675117493, + "learning_rate": 0.0001565509964316425, + "loss": 0.8786, + "step": 6949 + }, + { + "epoch": 1.23735754985755, + "grad_norm": 0.6438127160072327, + "learning_rate": 0.00015653945158201018, + "loss": 0.9435, + "step": 6950 + }, + { + "epoch": 1.2375356125356125, + "grad_norm": 0.68101966381073, + "learning_rate": 0.00015652790562460474, + "loss": 1.1062, + "step": 6951 + }, + { + "epoch": 1.2377136752136753, + "grad_norm": 0.661230206489563, + "learning_rate": 0.00015651635855965242, + "loss": 1.0113, + "step": 6952 + }, + { + "epoch": 1.2378917378917378, + "grad_norm": 0.6399117708206177, + "learning_rate": 0.0001565048103873795, + "loss": 1.1423, + "step": 6953 + }, + { + "epoch": 1.2380698005698005, + "grad_norm": 0.7614672780036926, + "learning_rate": 0.00015649326110801215, + "loss": 1.0359, + "step": 6954 + }, + { + "epoch": 1.2382478632478633, + "grad_norm": 0.6461986303329468, + "learning_rate": 0.00015648171072177674, + "loss": 1.0145, + "step": 6955 + }, + { + "epoch": 1.238425925925926, + "grad_norm": 0.5902668833732605, + "learning_rate": 0.0001564701592288995, + "loss": 0.9451, + "step": 6956 + }, + { + "epoch": 1.2386039886039886, + "grad_norm": 0.5686020255088806, + "learning_rate": 0.00015645860662960682, + "loss": 0.7512, + "step": 6957 + }, + { + "epoch": 1.2387820512820513, + "grad_norm": 0.6640077829360962, + "learning_rate": 0.00015644705292412503, + "loss": 0.7133, + "step": 6958 + }, + { + "epoch": 1.2389601139601139, + "grad_norm": 0.7402132749557495, + "learning_rate": 0.00015643549811268049, + "loss": 1.0903, + "step": 6959 + }, + { + "epoch": 1.2391381766381766, + "grad_norm": 0.62332683801651, + "learning_rate": 0.00015642394219549962, + "loss": 0.9378, + "step": 6960 + }, + { + "epoch": 1.2393162393162394, + "grad_norm": 0.6374901533126831, + "learning_rate": 0.00015641238517280877, + "loss": 1.0746, + "step": 6961 + }, + { + "epoch": 1.239494301994302, + "grad_norm": 0.5939112901687622, + "learning_rate": 0.00015640082704483443, + "loss": 0.7185, + "step": 6962 + }, + { + "epoch": 1.2396723646723646, + "grad_norm": 0.8378096222877502, + "learning_rate": 0.00015638926781180306, + "loss": 1.1932, + "step": 6963 + }, + { + "epoch": 1.2398504273504274, + "grad_norm": 0.5707982778549194, + "learning_rate": 0.0001563777074739411, + "loss": 0.9834, + "step": 6964 + }, + { + "epoch": 1.24002849002849, + "grad_norm": 0.6339748501777649, + "learning_rate": 0.00015636614603147512, + "loss": 1.0307, + "step": 6965 + }, + { + "epoch": 1.2402065527065527, + "grad_norm": 0.7353155016899109, + "learning_rate": 0.00015635458348463156, + "loss": 1.0311, + "step": 6966 + }, + { + "epoch": 1.2403846153846154, + "grad_norm": 0.8307726979255676, + "learning_rate": 0.00015634301983363704, + "loss": 1.0673, + "step": 6967 + }, + { + "epoch": 1.2405626780626782, + "grad_norm": 0.5299199819564819, + "learning_rate": 0.00015633145507871807, + "loss": 0.6649, + "step": 6968 + }, + { + "epoch": 1.2407407407407407, + "grad_norm": 0.6162533760070801, + "learning_rate": 0.00015631988922010126, + "loss": 0.8096, + "step": 6969 + }, + { + "epoch": 1.2409188034188035, + "grad_norm": 0.6212689876556396, + "learning_rate": 0.0001563083222580132, + "loss": 1.0371, + "step": 6970 + }, + { + "epoch": 1.241096866096866, + "grad_norm": 0.6148123145103455, + "learning_rate": 0.00015629675419268055, + "loss": 1.0439, + "step": 6971 + }, + { + "epoch": 1.2412749287749287, + "grad_norm": 0.6163684129714966, + "learning_rate": 0.00015628518502432994, + "loss": 0.9075, + "step": 6972 + }, + { + "epoch": 1.2414529914529915, + "grad_norm": 0.5127472877502441, + "learning_rate": 0.00015627361475318807, + "loss": 0.6138, + "step": 6973 + }, + { + "epoch": 1.2416310541310542, + "grad_norm": 0.6508103013038635, + "learning_rate": 0.0001562620433794816, + "loss": 0.9608, + "step": 6974 + }, + { + "epoch": 1.2418091168091168, + "grad_norm": 0.6711046695709229, + "learning_rate": 0.0001562504709034373, + "loss": 1.1494, + "step": 6975 + }, + { + "epoch": 1.2419871794871795, + "grad_norm": 0.6831514835357666, + "learning_rate": 0.00015623889732528182, + "loss": 0.9664, + "step": 6976 + }, + { + "epoch": 1.242165242165242, + "grad_norm": 0.693732738494873, + "learning_rate": 0.00015622732264524198, + "loss": 0.9055, + "step": 6977 + }, + { + "epoch": 1.2423433048433048, + "grad_norm": 0.8475173711776733, + "learning_rate": 0.00015621574686354456, + "loss": 1.2014, + "step": 6978 + }, + { + "epoch": 1.2425213675213675, + "grad_norm": 0.6342347264289856, + "learning_rate": 0.0001562041699804164, + "loss": 1.0691, + "step": 6979 + }, + { + "epoch": 1.2426994301994303, + "grad_norm": 0.620517373085022, + "learning_rate": 0.00015619259199608422, + "loss": 0.7318, + "step": 6980 + }, + { + "epoch": 1.2428774928774928, + "grad_norm": 0.589567244052887, + "learning_rate": 0.000156181012910775, + "loss": 1.0656, + "step": 6981 + }, + { + "epoch": 1.2430555555555556, + "grad_norm": 0.7570258975028992, + "learning_rate": 0.00015616943272471546, + "loss": 1.0517, + "step": 6982 + }, + { + "epoch": 1.243233618233618, + "grad_norm": 0.6232032775878906, + "learning_rate": 0.00015615785143813262, + "loss": 0.8867, + "step": 6983 + }, + { + "epoch": 1.2434116809116809, + "grad_norm": 0.630095899105072, + "learning_rate": 0.0001561462690512533, + "loss": 0.9287, + "step": 6984 + }, + { + "epoch": 1.2435897435897436, + "grad_norm": 0.7410848140716553, + "learning_rate": 0.00015613468556430454, + "loss": 1.162, + "step": 6985 + }, + { + "epoch": 1.2437678062678064, + "grad_norm": 0.7574684023857117, + "learning_rate": 0.00015612310097751317, + "loss": 1.2118, + "step": 6986 + }, + { + "epoch": 1.243945868945869, + "grad_norm": 0.580760657787323, + "learning_rate": 0.0001561115152911062, + "loss": 1.0612, + "step": 6987 + }, + { + "epoch": 1.2441239316239316, + "grad_norm": 0.6105104088783264, + "learning_rate": 0.00015609992850531073, + "loss": 0.9262, + "step": 6988 + }, + { + "epoch": 1.2443019943019944, + "grad_norm": 0.669435441493988, + "learning_rate": 0.00015608834062035362, + "loss": 0.9595, + "step": 6989 + }, + { + "epoch": 1.244480056980057, + "grad_norm": 0.6530314683914185, + "learning_rate": 0.00015607675163646206, + "loss": 0.7987, + "step": 6990 + }, + { + "epoch": 1.2446581196581197, + "grad_norm": 0.5801477432250977, + "learning_rate": 0.00015606516155386297, + "loss": 0.7667, + "step": 6991 + }, + { + "epoch": 1.2448361823361824, + "grad_norm": 0.5773885250091553, + "learning_rate": 0.00015605357037278355, + "loss": 0.847, + "step": 6992 + }, + { + "epoch": 1.245014245014245, + "grad_norm": 0.5399810075759888, + "learning_rate": 0.00015604197809345082, + "loss": 0.9284, + "step": 6993 + }, + { + "epoch": 1.2451923076923077, + "grad_norm": 0.5910452604293823, + "learning_rate": 0.000156030384716092, + "loss": 1.0004, + "step": 6994 + }, + { + "epoch": 1.2453703703703705, + "grad_norm": 0.5979224443435669, + "learning_rate": 0.00015601879024093414, + "loss": 0.9027, + "step": 6995 + }, + { + "epoch": 1.245548433048433, + "grad_norm": 0.6092126369476318, + "learning_rate": 0.0001560071946682045, + "loss": 0.9755, + "step": 6996 + }, + { + "epoch": 1.2457264957264957, + "grad_norm": 0.6536708474159241, + "learning_rate": 0.0001559955979981302, + "loss": 1.1828, + "step": 6997 + }, + { + "epoch": 1.2459045584045585, + "grad_norm": 0.6602030992507935, + "learning_rate": 0.00015598400023093847, + "loss": 1.0395, + "step": 6998 + }, + { + "epoch": 1.246082621082621, + "grad_norm": 0.6864825487136841, + "learning_rate": 0.00015597240136685657, + "loss": 1.083, + "step": 6999 + }, + { + "epoch": 1.2462606837606838, + "grad_norm": 0.6194674968719482, + "learning_rate": 0.0001559608014061117, + "loss": 1.0461, + "step": 7000 + }, + { + "epoch": 1.2464387464387465, + "grad_norm": 0.5879074335098267, + "learning_rate": 0.00015594920034893122, + "loss": 1.076, + "step": 7001 + }, + { + "epoch": 1.246616809116809, + "grad_norm": 0.6514387726783752, + "learning_rate": 0.00015593759819554234, + "loss": 1.0396, + "step": 7002 + }, + { + "epoch": 1.2467948717948718, + "grad_norm": 0.5988301634788513, + "learning_rate": 0.00015592599494617247, + "loss": 0.9501, + "step": 7003 + }, + { + "epoch": 1.2469729344729346, + "grad_norm": 0.6282773613929749, + "learning_rate": 0.00015591439060104887, + "loss": 1.1002, + "step": 7004 + }, + { + "epoch": 1.247150997150997, + "grad_norm": 0.6910465955734253, + "learning_rate": 0.00015590278516039896, + "loss": 1.1771, + "step": 7005 + }, + { + "epoch": 1.2473290598290598, + "grad_norm": 0.6097282767295837, + "learning_rate": 0.00015589117862445007, + "loss": 1.0707, + "step": 7006 + }, + { + "epoch": 1.2475071225071226, + "grad_norm": 0.7076875567436218, + "learning_rate": 0.00015587957099342967, + "loss": 1.0078, + "step": 7007 + }, + { + "epoch": 1.2476851851851851, + "grad_norm": 0.6776556372642517, + "learning_rate": 0.00015586796226756518, + "loss": 0.8971, + "step": 7008 + }, + { + "epoch": 1.2478632478632479, + "grad_norm": 0.6506341695785522, + "learning_rate": 0.00015585635244708398, + "loss": 0.9727, + "step": 7009 + }, + { + "epoch": 1.2480413105413106, + "grad_norm": 0.624724805355072, + "learning_rate": 0.00015584474153221357, + "loss": 0.9858, + "step": 7010 + }, + { + "epoch": 1.2482193732193732, + "grad_norm": 0.6070096492767334, + "learning_rate": 0.0001558331295231815, + "loss": 0.9385, + "step": 7011 + }, + { + "epoch": 1.248397435897436, + "grad_norm": 0.6948656439781189, + "learning_rate": 0.00015582151642021524, + "loss": 0.9425, + "step": 7012 + }, + { + "epoch": 1.2485754985754987, + "grad_norm": 0.6559088230133057, + "learning_rate": 0.0001558099022235423, + "loss": 1.0002, + "step": 7013 + }, + { + "epoch": 1.2487535612535612, + "grad_norm": 0.6097117066383362, + "learning_rate": 0.00015579828693339026, + "loss": 1.0234, + "step": 7014 + }, + { + "epoch": 1.248931623931624, + "grad_norm": 0.6612260341644287, + "learning_rate": 0.00015578667054998673, + "loss": 1.1376, + "step": 7015 + }, + { + "epoch": 1.2491096866096867, + "grad_norm": 0.6305607557296753, + "learning_rate": 0.00015577505307355925, + "loss": 0.9127, + "step": 7016 + }, + { + "epoch": 1.2492877492877492, + "grad_norm": 0.6648319959640503, + "learning_rate": 0.00015576343450433549, + "loss": 0.8697, + "step": 7017 + }, + { + "epoch": 1.249465811965812, + "grad_norm": 0.7642946839332581, + "learning_rate": 0.00015575181484254303, + "loss": 1.0998, + "step": 7018 + }, + { + "epoch": 1.2496438746438747, + "grad_norm": 0.6775243282318115, + "learning_rate": 0.00015574019408840962, + "loss": 1.0186, + "step": 7019 + }, + { + "epoch": 1.2498219373219372, + "grad_norm": 0.6075591444969177, + "learning_rate": 0.00015572857224216286, + "loss": 0.9592, + "step": 7020 + }, + { + "epoch": 1.2498219373219372, + "eval_loss": 1.105136752128601, + "eval_runtime": 24.4793, + "eval_samples_per_second": 42.526, + "eval_steps_per_second": 21.283, + "step": 7020 + }, + { + "epoch": 1.25, + "grad_norm": 0.5856962203979492, + "learning_rate": 0.0001557169493040305, + "loss": 0.8336, + "step": 7021 + }, + { + "epoch": 1.2501780626780628, + "grad_norm": 0.6451364159584045, + "learning_rate": 0.00015570532527424028, + "loss": 0.8805, + "step": 7022 + }, + { + "epoch": 1.2503561253561253, + "grad_norm": 0.6266474723815918, + "learning_rate": 0.00015569370015301991, + "loss": 1.0023, + "step": 7023 + }, + { + "epoch": 1.250534188034188, + "grad_norm": 0.5547378063201904, + "learning_rate": 0.00015568207394059722, + "loss": 0.7385, + "step": 7024 + }, + { + "epoch": 1.2507122507122508, + "grad_norm": 0.604169487953186, + "learning_rate": 0.0001556704466371999, + "loss": 0.9194, + "step": 7025 + }, + { + "epoch": 1.2508903133903133, + "grad_norm": 0.7054405212402344, + "learning_rate": 0.00015565881824305586, + "loss": 1.1864, + "step": 7026 + }, + { + "epoch": 1.251068376068376, + "grad_norm": 0.6429929733276367, + "learning_rate": 0.0001556471887583929, + "loss": 1.0129, + "step": 7027 + }, + { + "epoch": 1.2512464387464388, + "grad_norm": 0.695957362651825, + "learning_rate": 0.00015563555818343887, + "loss": 1.2994, + "step": 7028 + }, + { + "epoch": 1.2514245014245013, + "grad_norm": 0.5889938473701477, + "learning_rate": 0.0001556239265184216, + "loss": 1.0109, + "step": 7029 + }, + { + "epoch": 1.251602564102564, + "grad_norm": 0.6424569487571716, + "learning_rate": 0.0001556122937635691, + "loss": 0.8585, + "step": 7030 + }, + { + "epoch": 1.2517806267806268, + "grad_norm": 0.5561244487762451, + "learning_rate": 0.0001556006599191092, + "loss": 0.9994, + "step": 7031 + }, + { + "epoch": 1.2519586894586894, + "grad_norm": 0.6355302333831787, + "learning_rate": 0.00015558902498526988, + "loss": 0.9495, + "step": 7032 + }, + { + "epoch": 1.2521367521367521, + "grad_norm": 0.6272686719894409, + "learning_rate": 0.00015557738896227908, + "loss": 0.7611, + "step": 7033 + }, + { + "epoch": 1.2523148148148149, + "grad_norm": 0.7069199085235596, + "learning_rate": 0.00015556575185036482, + "loss": 1.0612, + "step": 7034 + }, + { + "epoch": 1.2524928774928774, + "grad_norm": 0.6635094285011292, + "learning_rate": 0.00015555411364975505, + "loss": 1.1182, + "step": 7035 + }, + { + "epoch": 1.2526709401709402, + "grad_norm": 0.6112014651298523, + "learning_rate": 0.00015554247436067785, + "loss": 0.8677, + "step": 7036 + }, + { + "epoch": 1.252849002849003, + "grad_norm": 0.678963303565979, + "learning_rate": 0.00015553083398336126, + "loss": 1.1421, + "step": 7037 + }, + { + "epoch": 1.2530270655270654, + "grad_norm": 0.6291939616203308, + "learning_rate": 0.0001555191925180333, + "loss": 0.9157, + "step": 7038 + }, + { + "epoch": 1.2532051282051282, + "grad_norm": 0.6519795656204224, + "learning_rate": 0.0001555075499649221, + "loss": 1.0074, + "step": 7039 + }, + { + "epoch": 1.253383190883191, + "grad_norm": 0.6063529849052429, + "learning_rate": 0.00015549590632425576, + "loss": 1.0205, + "step": 7040 + }, + { + "epoch": 1.2535612535612537, + "grad_norm": 0.7055633664131165, + "learning_rate": 0.00015548426159626242, + "loss": 1.0254, + "step": 7041 + }, + { + "epoch": 1.2537393162393162, + "grad_norm": 0.6783022880554199, + "learning_rate": 0.00015547261578117025, + "loss": 1.1017, + "step": 7042 + }, + { + "epoch": 1.253917378917379, + "grad_norm": 0.7055003643035889, + "learning_rate": 0.0001554609688792074, + "loss": 1.0269, + "step": 7043 + }, + { + "epoch": 1.2540954415954415, + "grad_norm": 0.6465007662773132, + "learning_rate": 0.0001554493208906021, + "loss": 1.0492, + "step": 7044 + }, + { + "epoch": 1.2542735042735043, + "grad_norm": 0.6443775296211243, + "learning_rate": 0.0001554376718155825, + "loss": 0.9778, + "step": 7045 + }, + { + "epoch": 1.254451566951567, + "grad_norm": 0.695214569568634, + "learning_rate": 0.0001554260216543769, + "loss": 0.8792, + "step": 7046 + }, + { + "epoch": 1.2546296296296298, + "grad_norm": 0.6777814626693726, + "learning_rate": 0.00015541437040721354, + "loss": 0.8944, + "step": 7047 + }, + { + "epoch": 1.2548076923076923, + "grad_norm": 0.6269369721412659, + "learning_rate": 0.0001554027180743207, + "loss": 0.8825, + "step": 7048 + }, + { + "epoch": 1.254985754985755, + "grad_norm": 0.6197061538696289, + "learning_rate": 0.0001553910646559267, + "loss": 0.9823, + "step": 7049 + }, + { + "epoch": 1.2551638176638176, + "grad_norm": 0.681347131729126, + "learning_rate": 0.00015537941015225984, + "loss": 0.995, + "step": 7050 + }, + { + "epoch": 1.2553418803418803, + "grad_norm": 0.6224286556243896, + "learning_rate": 0.00015536775456354848, + "loss": 0.7714, + "step": 7051 + }, + { + "epoch": 1.255519943019943, + "grad_norm": 0.6113278269767761, + "learning_rate": 0.00015535609789002098, + "loss": 0.9859, + "step": 7052 + }, + { + "epoch": 1.2556980056980058, + "grad_norm": 0.6985422372817993, + "learning_rate": 0.00015534444013190577, + "loss": 0.8785, + "step": 7053 + }, + { + "epoch": 1.2558760683760684, + "grad_norm": 0.5602933168411255, + "learning_rate": 0.00015533278128943118, + "loss": 0.8341, + "step": 7054 + }, + { + "epoch": 1.256054131054131, + "grad_norm": 0.587684690952301, + "learning_rate": 0.0001553211213628257, + "loss": 0.7933, + "step": 7055 + }, + { + "epoch": 1.2562321937321936, + "grad_norm": 0.692997932434082, + "learning_rate": 0.0001553094603523178, + "loss": 1.0957, + "step": 7056 + }, + { + "epoch": 1.2564102564102564, + "grad_norm": 0.6925587058067322, + "learning_rate": 0.00015529779825813588, + "loss": 0.8602, + "step": 7057 + }, + { + "epoch": 1.2565883190883191, + "grad_norm": 0.6383063197135925, + "learning_rate": 0.0001552861350805085, + "loss": 0.9933, + "step": 7058 + }, + { + "epoch": 1.256766381766382, + "grad_norm": 0.6520544290542603, + "learning_rate": 0.00015527447081966413, + "loss": 0.9498, + "step": 7059 + }, + { + "epoch": 1.2569444444444444, + "grad_norm": 0.7353914380073547, + "learning_rate": 0.00015526280547583133, + "loss": 1.1071, + "step": 7060 + }, + { + "epoch": 1.2571225071225072, + "grad_norm": 0.7141618132591248, + "learning_rate": 0.00015525113904923864, + "loss": 0.8333, + "step": 7061 + }, + { + "epoch": 1.2573005698005697, + "grad_norm": 0.6194499731063843, + "learning_rate": 0.00015523947154011468, + "loss": 0.9421, + "step": 7062 + }, + { + "epoch": 1.2574786324786325, + "grad_norm": 0.7514514327049255, + "learning_rate": 0.00015522780294868803, + "loss": 1.226, + "step": 7063 + }, + { + "epoch": 1.2576566951566952, + "grad_norm": 0.762923538684845, + "learning_rate": 0.0001552161332751873, + "loss": 1.1893, + "step": 7064 + }, + { + "epoch": 1.257834757834758, + "grad_norm": 0.6265730261802673, + "learning_rate": 0.00015520446251984113, + "loss": 0.6604, + "step": 7065 + }, + { + "epoch": 1.2580128205128205, + "grad_norm": 0.6447750329971313, + "learning_rate": 0.0001551927906828782, + "loss": 0.9814, + "step": 7066 + }, + { + "epoch": 1.2581908831908832, + "grad_norm": 0.5791042447090149, + "learning_rate": 0.00015518111776452722, + "loss": 0.8283, + "step": 7067 + }, + { + "epoch": 1.2583689458689458, + "grad_norm": 0.5267777442932129, + "learning_rate": 0.00015516944376501682, + "loss": 0.5748, + "step": 7068 + }, + { + "epoch": 1.2585470085470085, + "grad_norm": 0.7343912720680237, + "learning_rate": 0.0001551577686845758, + "loss": 1.1777, + "step": 7069 + }, + { + "epoch": 1.2587250712250713, + "grad_norm": 0.645746111869812, + "learning_rate": 0.00015514609252343284, + "loss": 0.9356, + "step": 7070 + }, + { + "epoch": 1.258903133903134, + "grad_norm": 0.6993104219436646, + "learning_rate": 0.0001551344152818168, + "loss": 1.06, + "step": 7071 + }, + { + "epoch": 1.2590811965811965, + "grad_norm": 0.6661365628242493, + "learning_rate": 0.0001551227369599564, + "loss": 1.061, + "step": 7072 + }, + { + "epoch": 1.2592592592592593, + "grad_norm": 0.7833736538887024, + "learning_rate": 0.0001551110575580805, + "loss": 0.9674, + "step": 7073 + }, + { + "epoch": 1.2594373219373218, + "grad_norm": 0.5878575444221497, + "learning_rate": 0.00015509937707641787, + "loss": 0.9002, + "step": 7074 + }, + { + "epoch": 1.2596153846153846, + "grad_norm": 0.6402907371520996, + "learning_rate": 0.00015508769551519745, + "loss": 1.0157, + "step": 7075 + }, + { + "epoch": 1.2597934472934473, + "grad_norm": 0.6794611215591431, + "learning_rate": 0.00015507601287464805, + "loss": 1.052, + "step": 7076 + }, + { + "epoch": 1.25997150997151, + "grad_norm": 0.706922173500061, + "learning_rate": 0.0001550643291549986, + "loss": 1.0814, + "step": 7077 + }, + { + "epoch": 1.2601495726495726, + "grad_norm": 0.6722953915596008, + "learning_rate": 0.000155052644356478, + "loss": 1.1402, + "step": 7078 + }, + { + "epoch": 1.2603276353276354, + "grad_norm": 0.6619611978530884, + "learning_rate": 0.00015504095847931518, + "loss": 0.9583, + "step": 7079 + }, + { + "epoch": 1.260505698005698, + "grad_norm": 0.5645583271980286, + "learning_rate": 0.00015502927152373914, + "loss": 0.6746, + "step": 7080 + }, + { + "epoch": 1.2606837606837606, + "grad_norm": 0.6634977459907532, + "learning_rate": 0.00015501758348997882, + "loss": 1.0451, + "step": 7081 + }, + { + "epoch": 1.2608618233618234, + "grad_norm": 0.7167651057243347, + "learning_rate": 0.00015500589437826326, + "loss": 0.931, + "step": 7082 + }, + { + "epoch": 1.2610398860398861, + "grad_norm": 0.6179340481758118, + "learning_rate": 0.00015499420418882146, + "loss": 1.0953, + "step": 7083 + }, + { + "epoch": 1.2612179487179487, + "grad_norm": 0.6948468685150146, + "learning_rate": 0.00015498251292188247, + "loss": 1.0277, + "step": 7084 + }, + { + "epoch": 1.2613960113960114, + "grad_norm": 0.6256045699119568, + "learning_rate": 0.00015497082057767532, + "loss": 1.0154, + "step": 7085 + }, + { + "epoch": 1.261574074074074, + "grad_norm": 0.6457428336143494, + "learning_rate": 0.0001549591271564292, + "loss": 0.9693, + "step": 7086 + }, + { + "epoch": 1.2617521367521367, + "grad_norm": 0.722259521484375, + "learning_rate": 0.0001549474326583731, + "loss": 0.9176, + "step": 7087 + }, + { + "epoch": 1.2619301994301995, + "grad_norm": 0.742477297782898, + "learning_rate": 0.0001549357370837362, + "loss": 0.9813, + "step": 7088 + }, + { + "epoch": 1.2621082621082622, + "grad_norm": 0.5981723666191101, + "learning_rate": 0.0001549240404327477, + "loss": 0.8943, + "step": 7089 + }, + { + "epoch": 1.2622863247863247, + "grad_norm": 0.6266574859619141, + "learning_rate": 0.00015491234270563665, + "loss": 0.8439, + "step": 7090 + }, + { + "epoch": 1.2624643874643875, + "grad_norm": 0.6723998188972473, + "learning_rate": 0.00015490064390263238, + "loss": 1.2278, + "step": 7091 + }, + { + "epoch": 1.26264245014245, + "grad_norm": 0.6628100275993347, + "learning_rate": 0.00015488894402396398, + "loss": 0.9526, + "step": 7092 + }, + { + "epoch": 1.2628205128205128, + "grad_norm": 0.6661350727081299, + "learning_rate": 0.0001548772430698608, + "loss": 0.974, + "step": 7093 + }, + { + "epoch": 1.2629985754985755, + "grad_norm": 0.8210669755935669, + "learning_rate": 0.000154865541040552, + "loss": 1.1142, + "step": 7094 + }, + { + "epoch": 1.2631766381766383, + "grad_norm": 0.6329003572463989, + "learning_rate": 0.0001548538379362669, + "loss": 0.8485, + "step": 7095 + }, + { + "epoch": 1.2633547008547008, + "grad_norm": 0.6288384795188904, + "learning_rate": 0.0001548421337572348, + "loss": 0.816, + "step": 7096 + }, + { + "epoch": 1.2635327635327636, + "grad_norm": 0.631060004234314, + "learning_rate": 0.00015483042850368504, + "loss": 0.8237, + "step": 7097 + }, + { + "epoch": 1.263710826210826, + "grad_norm": 0.7343839406967163, + "learning_rate": 0.0001548187221758469, + "loss": 1.1507, + "step": 7098 + }, + { + "epoch": 1.2638888888888888, + "grad_norm": 0.6313042640686035, + "learning_rate": 0.0001548070147739498, + "loss": 0.7762, + "step": 7099 + }, + { + "epoch": 1.2640669515669516, + "grad_norm": 0.6449850797653198, + "learning_rate": 0.00015479530629822308, + "loss": 0.9225, + "step": 7100 + }, + { + "epoch": 1.2642450142450143, + "grad_norm": 0.6371589303016663, + "learning_rate": 0.00015478359674889617, + "loss": 1.0088, + "step": 7101 + }, + { + "epoch": 1.2644230769230769, + "grad_norm": 0.6483678221702576, + "learning_rate": 0.00015477188612619849, + "loss": 0.6234, + "step": 7102 + }, + { + "epoch": 1.2646011396011396, + "grad_norm": 0.6945441365242004, + "learning_rate": 0.00015476017443035947, + "loss": 1.123, + "step": 7103 + }, + { + "epoch": 1.2647792022792022, + "grad_norm": 0.6356340050697327, + "learning_rate": 0.00015474846166160856, + "loss": 0.9923, + "step": 7104 + }, + { + "epoch": 1.264957264957265, + "grad_norm": 0.6774702668190002, + "learning_rate": 0.00015473674782017532, + "loss": 0.9694, + "step": 7105 + }, + { + "epoch": 1.2651353276353277, + "grad_norm": 0.6332793831825256, + "learning_rate": 0.0001547250329062892, + "loss": 1.0633, + "step": 7106 + }, + { + "epoch": 1.2653133903133904, + "grad_norm": 0.6563684344291687, + "learning_rate": 0.00015471331692017972, + "loss": 1.0893, + "step": 7107 + }, + { + "epoch": 1.265491452991453, + "grad_norm": 0.7318371534347534, + "learning_rate": 0.0001547015998620765, + "loss": 1.1777, + "step": 7108 + }, + { + "epoch": 1.2656695156695157, + "grad_norm": 0.7099173069000244, + "learning_rate": 0.000154689881732209, + "loss": 1.1717, + "step": 7109 + }, + { + "epoch": 1.2658475783475782, + "grad_norm": 0.661078691482544, + "learning_rate": 0.00015467816253080693, + "loss": 1.0448, + "step": 7110 + }, + { + "epoch": 1.266025641025641, + "grad_norm": 0.6206802129745483, + "learning_rate": 0.0001546664422580998, + "loss": 0.9334, + "step": 7111 + }, + { + "epoch": 1.2662037037037037, + "grad_norm": 0.6514355540275574, + "learning_rate": 0.00015465472091431728, + "loss": 0.9533, + "step": 7112 + }, + { + "epoch": 1.2663817663817665, + "grad_norm": 0.6090209484100342, + "learning_rate": 0.0001546429984996891, + "loss": 0.9206, + "step": 7113 + }, + { + "epoch": 1.266559829059829, + "grad_norm": 0.6345987915992737, + "learning_rate": 0.00015463127501444488, + "loss": 1.0537, + "step": 7114 + }, + { + "epoch": 1.2667378917378918, + "grad_norm": 0.6095160245895386, + "learning_rate": 0.0001546195504588143, + "loss": 0.8652, + "step": 7115 + }, + { + "epoch": 1.2669159544159543, + "grad_norm": 0.6751621961593628, + "learning_rate": 0.00015460782483302707, + "loss": 0.9001, + "step": 7116 + }, + { + "epoch": 1.267094017094017, + "grad_norm": 0.6261575222015381, + "learning_rate": 0.00015459609813731295, + "loss": 0.929, + "step": 7117 + }, + { + "epoch": 1.2672720797720798, + "grad_norm": 0.589495837688446, + "learning_rate": 0.0001545843703719017, + "loss": 0.9023, + "step": 7118 + }, + { + "epoch": 1.2674501424501425, + "grad_norm": 0.6364617943763733, + "learning_rate": 0.00015457264153702311, + "loss": 0.8261, + "step": 7119 + }, + { + "epoch": 1.267628205128205, + "grad_norm": 0.6685599684715271, + "learning_rate": 0.00015456091163290698, + "loss": 1.1267, + "step": 7120 + }, + { + "epoch": 1.2678062678062678, + "grad_norm": 0.6440932750701904, + "learning_rate": 0.0001545491806597831, + "loss": 0.9643, + "step": 7121 + }, + { + "epoch": 1.2679843304843303, + "grad_norm": 0.7641597390174866, + "learning_rate": 0.00015453744861788137, + "loss": 1.1577, + "step": 7122 + }, + { + "epoch": 1.268162393162393, + "grad_norm": 0.6965937614440918, + "learning_rate": 0.00015452571550743163, + "loss": 0.7835, + "step": 7123 + }, + { + "epoch": 1.2683404558404558, + "grad_norm": 0.6332844495773315, + "learning_rate": 0.00015451398132866376, + "loss": 0.9794, + "step": 7124 + }, + { + "epoch": 1.2685185185185186, + "grad_norm": 0.6719903349876404, + "learning_rate": 0.00015450224608180765, + "loss": 0.9795, + "step": 7125 + }, + { + "epoch": 1.2686965811965811, + "grad_norm": 0.567414402961731, + "learning_rate": 0.00015449050976709328, + "loss": 0.9737, + "step": 7126 + }, + { + "epoch": 1.2688746438746439, + "grad_norm": 0.6810645461082458, + "learning_rate": 0.0001544787723847505, + "loss": 1.2358, + "step": 7127 + }, + { + "epoch": 1.2690527065527066, + "grad_norm": 0.6693191528320312, + "learning_rate": 0.00015446703393500938, + "loss": 0.9475, + "step": 7128 + }, + { + "epoch": 1.2692307692307692, + "grad_norm": 0.7077522277832031, + "learning_rate": 0.00015445529441809988, + "loss": 1.013, + "step": 7129 + }, + { + "epoch": 1.269408831908832, + "grad_norm": 0.6596258878707886, + "learning_rate": 0.000154443553834252, + "loss": 1.1506, + "step": 7130 + }, + { + "epoch": 1.2695868945868947, + "grad_norm": 0.6721500754356384, + "learning_rate": 0.0001544318121836958, + "loss": 0.8848, + "step": 7131 + }, + { + "epoch": 1.2697649572649572, + "grad_norm": 0.6943998336791992, + "learning_rate": 0.00015442006946666132, + "loss": 1.1118, + "step": 7132 + }, + { + "epoch": 1.26994301994302, + "grad_norm": 0.6132234930992126, + "learning_rate": 0.0001544083256833786, + "loss": 0.9932, + "step": 7133 + }, + { + "epoch": 1.2701210826210827, + "grad_norm": 0.7337939739227295, + "learning_rate": 0.00015439658083407775, + "loss": 1.0973, + "step": 7134 + }, + { + "epoch": 1.2702991452991452, + "grad_norm": 0.6551772356033325, + "learning_rate": 0.00015438483491898893, + "loss": 1.0006, + "step": 7135 + }, + { + "epoch": 1.270477207977208, + "grad_norm": 0.660068929195404, + "learning_rate": 0.00015437308793834223, + "loss": 0.9291, + "step": 7136 + }, + { + "epoch": 1.2706552706552707, + "grad_norm": 0.7622788548469543, + "learning_rate": 0.00015436133989236783, + "loss": 1.0782, + "step": 7137 + }, + { + "epoch": 1.2708333333333333, + "grad_norm": 0.848494291305542, + "learning_rate": 0.00015434959078129587, + "loss": 1.2001, + "step": 7138 + }, + { + "epoch": 1.271011396011396, + "grad_norm": 0.6222602725028992, + "learning_rate": 0.0001543378406053566, + "loss": 1.011, + "step": 7139 + }, + { + "epoch": 1.2711894586894588, + "grad_norm": 0.6164663434028625, + "learning_rate": 0.00015432608936478026, + "loss": 1.0282, + "step": 7140 + }, + { + "epoch": 1.2713675213675213, + "grad_norm": 0.7236546277999878, + "learning_rate": 0.000154314337059797, + "loss": 1.0112, + "step": 7141 + }, + { + "epoch": 1.271545584045584, + "grad_norm": 0.6891111135482788, + "learning_rate": 0.00015430258369063715, + "loss": 1.1191, + "step": 7142 + }, + { + "epoch": 1.2717236467236468, + "grad_norm": 0.6600295901298523, + "learning_rate": 0.00015429082925753099, + "loss": 0.9561, + "step": 7143 + }, + { + "epoch": 1.2719017094017093, + "grad_norm": 0.6819902062416077, + "learning_rate": 0.0001542790737607088, + "loss": 1.0631, + "step": 7144 + }, + { + "epoch": 1.272079772079772, + "grad_norm": 0.6518470644950867, + "learning_rate": 0.0001542673172004009, + "loss": 1.0806, + "step": 7145 + }, + { + "epoch": 1.2722578347578348, + "grad_norm": 0.737501859664917, + "learning_rate": 0.00015425555957683767, + "loss": 1.0144, + "step": 7146 + }, + { + "epoch": 1.2724358974358974, + "grad_norm": 0.6245740652084351, + "learning_rate": 0.00015424380089024944, + "loss": 1.0612, + "step": 7147 + }, + { + "epoch": 1.27261396011396, + "grad_norm": 0.7118125557899475, + "learning_rate": 0.0001542320411408666, + "loss": 1.1458, + "step": 7148 + }, + { + "epoch": 1.2727920227920229, + "grad_norm": 0.6965761780738831, + "learning_rate": 0.00015422028032891958, + "loss": 0.8052, + "step": 7149 + }, + { + "epoch": 1.2729700854700854, + "grad_norm": 0.7661466598510742, + "learning_rate": 0.0001542085184546388, + "loss": 1.1245, + "step": 7150 + }, + { + "epoch": 1.2731481481481481, + "grad_norm": 0.7238876223564148, + "learning_rate": 0.00015419675551825475, + "loss": 0.9346, + "step": 7151 + }, + { + "epoch": 1.273326210826211, + "grad_norm": 0.669562041759491, + "learning_rate": 0.0001541849915199978, + "loss": 0.7816, + "step": 7152 + }, + { + "epoch": 1.2735042735042734, + "grad_norm": 0.6799174547195435, + "learning_rate": 0.00015417322646009855, + "loss": 1.047, + "step": 7153 + }, + { + "epoch": 1.2736823361823362, + "grad_norm": 0.6012796759605408, + "learning_rate": 0.00015416146033878745, + "loss": 1.0101, + "step": 7154 + }, + { + "epoch": 1.273860398860399, + "grad_norm": 0.7008427977561951, + "learning_rate": 0.00015414969315629505, + "loss": 1.1321, + "step": 7155 + }, + { + "epoch": 1.2740384615384617, + "grad_norm": 0.6555556058883667, + "learning_rate": 0.0001541379249128519, + "loss": 0.9926, + "step": 7156 + }, + { + "epoch": 1.2742165242165242, + "grad_norm": 0.6324251294136047, + "learning_rate": 0.00015412615560868854, + "loss": 0.9051, + "step": 7157 + }, + { + "epoch": 1.274394586894587, + "grad_norm": 0.6035568714141846, + "learning_rate": 0.0001541143852440356, + "loss": 0.8248, + "step": 7158 + }, + { + "epoch": 1.2745726495726495, + "grad_norm": 0.6733569502830505, + "learning_rate": 0.0001541026138191237, + "loss": 0.9149, + "step": 7159 + }, + { + "epoch": 1.2747507122507122, + "grad_norm": 0.8306798338890076, + "learning_rate": 0.0001540908413341835, + "loss": 1.0694, + "step": 7160 + }, + { + "epoch": 1.274928774928775, + "grad_norm": 0.6649713516235352, + "learning_rate": 0.00015407906778944563, + "loss": 1.1358, + "step": 7161 + }, + { + "epoch": 1.2751068376068377, + "grad_norm": 0.6889697909355164, + "learning_rate": 0.00015406729318514074, + "loss": 1.0096, + "step": 7162 + }, + { + "epoch": 1.2752849002849003, + "grad_norm": 0.6948645114898682, + "learning_rate": 0.0001540555175214996, + "loss": 1.0649, + "step": 7163 + }, + { + "epoch": 1.275462962962963, + "grad_norm": 0.6844844818115234, + "learning_rate": 0.0001540437407987528, + "loss": 0.884, + "step": 7164 + }, + { + "epoch": 1.2756410256410255, + "grad_norm": 0.7124526500701904, + "learning_rate": 0.00015403196301713124, + "loss": 1.1307, + "step": 7165 + }, + { + "epoch": 1.2758190883190883, + "grad_norm": 0.7328375577926636, + "learning_rate": 0.00015402018417686556, + "loss": 1.0348, + "step": 7166 + }, + { + "epoch": 1.275997150997151, + "grad_norm": 0.5872696042060852, + "learning_rate": 0.00015400840427818663, + "loss": 0.9827, + "step": 7167 + }, + { + "epoch": 1.2761752136752138, + "grad_norm": 0.6370702385902405, + "learning_rate": 0.00015399662332132519, + "loss": 0.9171, + "step": 7168 + }, + { + "epoch": 1.2763532763532763, + "grad_norm": 0.6481866240501404, + "learning_rate": 0.00015398484130651205, + "loss": 0.8704, + "step": 7169 + }, + { + "epoch": 1.276531339031339, + "grad_norm": 0.598739743232727, + "learning_rate": 0.00015397305823397812, + "loss": 0.8097, + "step": 7170 + }, + { + "epoch": 1.2767094017094016, + "grad_norm": 0.5941228270530701, + "learning_rate": 0.00015396127410395423, + "loss": 0.8853, + "step": 7171 + }, + { + "epoch": 1.2768874643874644, + "grad_norm": 0.6485885381698608, + "learning_rate": 0.00015394948891667127, + "loss": 0.702, + "step": 7172 + }, + { + "epoch": 1.2770655270655271, + "grad_norm": 0.5314942598342896, + "learning_rate": 0.00015393770267236017, + "loss": 0.7899, + "step": 7173 + }, + { + "epoch": 1.2772435897435899, + "grad_norm": 0.6113781929016113, + "learning_rate": 0.00015392591537125182, + "loss": 0.9871, + "step": 7174 + }, + { + "epoch": 1.2774216524216524, + "grad_norm": 0.5625866651535034, + "learning_rate": 0.00015391412701357715, + "loss": 0.8246, + "step": 7175 + }, + { + "epoch": 1.2775997150997151, + "grad_norm": 0.6006998419761658, + "learning_rate": 0.00015390233759956718, + "loss": 0.899, + "step": 7176 + }, + { + "epoch": 1.2777777777777777, + "grad_norm": 0.6916918158531189, + "learning_rate": 0.0001538905471294529, + "loss": 1.0443, + "step": 7177 + }, + { + "epoch": 1.2779558404558404, + "grad_norm": 0.6263536810874939, + "learning_rate": 0.00015387875560346525, + "loss": 0.9159, + "step": 7178 + }, + { + "epoch": 1.2781339031339032, + "grad_norm": 0.6563085913658142, + "learning_rate": 0.00015386696302183535, + "loss": 0.994, + "step": 7179 + }, + { + "epoch": 1.278311965811966, + "grad_norm": 0.6312007904052734, + "learning_rate": 0.00015385516938479416, + "loss": 0.9148, + "step": 7180 + }, + { + "epoch": 1.2784900284900285, + "grad_norm": 0.6408209204673767, + "learning_rate": 0.00015384337469257284, + "loss": 1.0508, + "step": 7181 + }, + { + "epoch": 1.2786680911680912, + "grad_norm": 0.656234085559845, + "learning_rate": 0.00015383157894540244, + "loss": 0.9952, + "step": 7182 + }, + { + "epoch": 1.2788461538461537, + "grad_norm": 0.7401639819145203, + "learning_rate": 0.00015381978214351407, + "loss": 1.1615, + "step": 7183 + }, + { + "epoch": 1.2790242165242165, + "grad_norm": 0.5746055841445923, + "learning_rate": 0.00015380798428713885, + "loss": 0.9142, + "step": 7184 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 0.8061720728874207, + "learning_rate": 0.00015379618537650797, + "loss": 1.13, + "step": 7185 + }, + { + "epoch": 1.279380341880342, + "grad_norm": 0.6336073875427246, + "learning_rate": 0.0001537843854118526, + "loss": 1.0581, + "step": 7186 + }, + { + "epoch": 1.2795584045584045, + "grad_norm": 0.6549856066703796, + "learning_rate": 0.0001537725843934039, + "loss": 1.09, + "step": 7187 + }, + { + "epoch": 1.2797364672364673, + "grad_norm": 0.5759010910987854, + "learning_rate": 0.00015376078232139315, + "loss": 0.8441, + "step": 7188 + }, + { + "epoch": 1.2799145299145298, + "grad_norm": 0.5733884572982788, + "learning_rate": 0.00015374897919605152, + "loss": 0.9086, + "step": 7189 + }, + { + "epoch": 1.2800925925925926, + "grad_norm": 0.6505870819091797, + "learning_rate": 0.0001537371750176103, + "loss": 1.1683, + "step": 7190 + }, + { + "epoch": 1.2802706552706553, + "grad_norm": 0.6744688153266907, + "learning_rate": 0.00015372536978630077, + "loss": 0.9483, + "step": 7191 + }, + { + "epoch": 1.280448717948718, + "grad_norm": 0.598098874092102, + "learning_rate": 0.0001537135635023542, + "loss": 0.7747, + "step": 7192 + }, + { + "epoch": 1.2806267806267806, + "grad_norm": 0.6711761951446533, + "learning_rate": 0.00015370175616600195, + "loss": 1.1897, + "step": 7193 + }, + { + "epoch": 1.2808048433048433, + "grad_norm": 0.6207453608512878, + "learning_rate": 0.00015368994777747536, + "loss": 1.0063, + "step": 7194 + }, + { + "epoch": 1.2809829059829059, + "grad_norm": 0.6701686382293701, + "learning_rate": 0.00015367813833700575, + "loss": 1.0864, + "step": 7195 + }, + { + "epoch": 1.2811609686609686, + "grad_norm": 0.5916469693183899, + "learning_rate": 0.00015366632784482456, + "loss": 0.8786, + "step": 7196 + }, + { + "epoch": 1.2813390313390314, + "grad_norm": 0.6567547917366028, + "learning_rate": 0.00015365451630116312, + "loss": 0.9977, + "step": 7197 + }, + { + "epoch": 1.2815170940170941, + "grad_norm": 0.7287433743476868, + "learning_rate": 0.00015364270370625294, + "loss": 1.1248, + "step": 7198 + }, + { + "epoch": 1.2816951566951567, + "grad_norm": 0.7736039161682129, + "learning_rate": 0.0001536308900603254, + "loss": 0.9832, + "step": 7199 + }, + { + "epoch": 1.2818732193732194, + "grad_norm": 0.6799852252006531, + "learning_rate": 0.00015361907536361194, + "loss": 1.0275, + "step": 7200 + }, + { + "epoch": 1.282051282051282, + "grad_norm": 0.5975812673568726, + "learning_rate": 0.00015360725961634407, + "loss": 1.0516, + "step": 7201 + }, + { + "epoch": 1.2822293447293447, + "grad_norm": 0.616307258605957, + "learning_rate": 0.00015359544281875337, + "loss": 0.8095, + "step": 7202 + }, + { + "epoch": 1.2824074074074074, + "grad_norm": 0.6357580423355103, + "learning_rate": 0.00015358362497107126, + "loss": 0.9186, + "step": 7203 + }, + { + "epoch": 1.2825854700854702, + "grad_norm": 0.679333508014679, + "learning_rate": 0.00015357180607352935, + "loss": 0.9433, + "step": 7204 + }, + { + "epoch": 1.2827635327635327, + "grad_norm": 0.6345439553260803, + "learning_rate": 0.00015355998612635914, + "loss": 0.9186, + "step": 7205 + }, + { + "epoch": 1.2829415954415955, + "grad_norm": 0.6256508827209473, + "learning_rate": 0.00015354816512979231, + "loss": 0.9984, + "step": 7206 + }, + { + "epoch": 1.283119658119658, + "grad_norm": 0.7973852753639221, + "learning_rate": 0.00015353634308406044, + "loss": 1.1145, + "step": 7207 + }, + { + "epoch": 1.2832977207977208, + "grad_norm": 0.711125910282135, + "learning_rate": 0.0001535245199893951, + "loss": 1.1947, + "step": 7208 + }, + { + "epoch": 1.2834757834757835, + "grad_norm": 0.6096055507659912, + "learning_rate": 0.00015351269584602798, + "loss": 1.0078, + "step": 7209 + }, + { + "epoch": 1.2836538461538463, + "grad_norm": 0.7089232802391052, + "learning_rate": 0.00015350087065419077, + "loss": 1.112, + "step": 7210 + }, + { + "epoch": 1.2838319088319088, + "grad_norm": 0.716199517250061, + "learning_rate": 0.00015348904441411508, + "loss": 1.1015, + "step": 7211 + }, + { + "epoch": 1.2840099715099715, + "grad_norm": 0.6374632716178894, + "learning_rate": 0.00015347721712603276, + "loss": 1.0519, + "step": 7212 + }, + { + "epoch": 1.284188034188034, + "grad_norm": 0.6500036716461182, + "learning_rate": 0.0001534653887901754, + "loss": 1.1719, + "step": 7213 + }, + { + "epoch": 1.2843660968660968, + "grad_norm": 0.7249937653541565, + "learning_rate": 0.00015345355940677485, + "loss": 1.0188, + "step": 7214 + }, + { + "epoch": 1.2845441595441596, + "grad_norm": 0.6645919680595398, + "learning_rate": 0.00015344172897606285, + "loss": 0.9788, + "step": 7215 + }, + { + "epoch": 1.2847222222222223, + "grad_norm": 0.7032710313796997, + "learning_rate": 0.00015342989749827113, + "loss": 1.1093, + "step": 7216 + }, + { + "epoch": 1.2849002849002849, + "grad_norm": 0.622767984867096, + "learning_rate": 0.0001534180649736316, + "loss": 0.8978, + "step": 7217 + }, + { + "epoch": 1.2850783475783476, + "grad_norm": 0.7499693036079407, + "learning_rate": 0.00015340623140237605, + "loss": 1.2232, + "step": 7218 + }, + { + "epoch": 1.2852564102564101, + "grad_norm": 0.6308625936508179, + "learning_rate": 0.00015339439678473636, + "loss": 0.8621, + "step": 7219 + }, + { + "epoch": 1.2854344729344729, + "grad_norm": 0.6513667106628418, + "learning_rate": 0.00015338256112094434, + "loss": 1.0541, + "step": 7220 + }, + { + "epoch": 1.2856125356125356, + "grad_norm": 0.6080937385559082, + "learning_rate": 0.00015337072441123193, + "loss": 0.8474, + "step": 7221 + }, + { + "epoch": 1.2857905982905984, + "grad_norm": 0.6742652058601379, + "learning_rate": 0.00015335888665583104, + "loss": 1.0172, + "step": 7222 + }, + { + "epoch": 1.285968660968661, + "grad_norm": 0.620810866355896, + "learning_rate": 0.00015334704785497364, + "loss": 1.049, + "step": 7223 + }, + { + "epoch": 1.2861467236467237, + "grad_norm": 0.5733018517494202, + "learning_rate": 0.00015333520800889165, + "loss": 0.7371, + "step": 7224 + }, + { + "epoch": 1.2863247863247862, + "grad_norm": 0.6447640061378479, + "learning_rate": 0.00015332336711781702, + "loss": 0.9925, + "step": 7225 + }, + { + "epoch": 1.286502849002849, + "grad_norm": 0.6764999628067017, + "learning_rate": 0.00015331152518198183, + "loss": 0.9052, + "step": 7226 + }, + { + "epoch": 1.2866809116809117, + "grad_norm": 0.6492836475372314, + "learning_rate": 0.00015329968220161803, + "loss": 0.9493, + "step": 7227 + }, + { + "epoch": 1.2868589743589745, + "grad_norm": 0.666157603263855, + "learning_rate": 0.00015328783817695766, + "loss": 1.0626, + "step": 7228 + }, + { + "epoch": 1.287037037037037, + "grad_norm": 0.7098026871681213, + "learning_rate": 0.00015327599310823283, + "loss": 1.0461, + "step": 7229 + }, + { + "epoch": 1.2872150997150997, + "grad_norm": 0.637778103351593, + "learning_rate": 0.00015326414699567555, + "loss": 0.9383, + "step": 7230 + }, + { + "epoch": 1.2873931623931623, + "grad_norm": 0.6816399693489075, + "learning_rate": 0.00015325229983951798, + "loss": 1.0647, + "step": 7231 + }, + { + "epoch": 1.287571225071225, + "grad_norm": 0.668689489364624, + "learning_rate": 0.0001532404516399922, + "loss": 1.0479, + "step": 7232 + }, + { + "epoch": 1.2877492877492878, + "grad_norm": 0.6459103226661682, + "learning_rate": 0.0001532286023973304, + "loss": 1.1751, + "step": 7233 + }, + { + "epoch": 1.2879273504273505, + "grad_norm": 0.679999589920044, + "learning_rate": 0.00015321675211176468, + "loss": 0.7541, + "step": 7234 + }, + { + "epoch": 1.288105413105413, + "grad_norm": 0.5415067672729492, + "learning_rate": 0.00015320490078352724, + "loss": 0.822, + "step": 7235 + }, + { + "epoch": 1.2882834757834758, + "grad_norm": 0.6817963719367981, + "learning_rate": 0.00015319304841285032, + "loss": 0.9424, + "step": 7236 + }, + { + "epoch": 1.2884615384615383, + "grad_norm": 0.6187505125999451, + "learning_rate": 0.0001531811949999661, + "loss": 0.8596, + "step": 7237 + }, + { + "epoch": 1.288639601139601, + "grad_norm": 0.6737838387489319, + "learning_rate": 0.00015316934054510685, + "loss": 1.0046, + "step": 7238 + }, + { + "epoch": 1.2888176638176638, + "grad_norm": 0.6445996761322021, + "learning_rate": 0.00015315748504850482, + "loss": 1.01, + "step": 7239 + }, + { + "epoch": 1.2889957264957266, + "grad_norm": 0.7279136180877686, + "learning_rate": 0.0001531456285103923, + "loss": 0.9066, + "step": 7240 + }, + { + "epoch": 1.289173789173789, + "grad_norm": 0.6619178652763367, + "learning_rate": 0.00015313377093100153, + "loss": 0.8977, + "step": 7241 + }, + { + "epoch": 1.2893518518518519, + "grad_norm": 0.7644323110580444, + "learning_rate": 0.000153121912310565, + "loss": 1.3085, + "step": 7242 + }, + { + "epoch": 1.2895299145299146, + "grad_norm": 0.645882248878479, + "learning_rate": 0.00015311005264931487, + "loss": 1.0337, + "step": 7243 + }, + { + "epoch": 1.2897079772079771, + "grad_norm": 0.6868017911911011, + "learning_rate": 0.0001530981919474836, + "loss": 0.9616, + "step": 7244 + }, + { + "epoch": 1.28988603988604, + "grad_norm": 0.7176693677902222, + "learning_rate": 0.00015308633020530362, + "loss": 1.1975, + "step": 7245 + }, + { + "epoch": 1.2900641025641026, + "grad_norm": 0.7358015775680542, + "learning_rate": 0.00015307446742300718, + "loss": 0.9308, + "step": 7246 + }, + { + "epoch": 1.2902421652421652, + "grad_norm": 0.7330248355865479, + "learning_rate": 0.00015306260360082688, + "loss": 0.9518, + "step": 7247 + }, + { + "epoch": 1.290420227920228, + "grad_norm": 0.6571981310844421, + "learning_rate": 0.00015305073873899503, + "loss": 0.9531, + "step": 7248 + }, + { + "epoch": 1.2905982905982907, + "grad_norm": 0.5968486666679382, + "learning_rate": 0.00015303887283774417, + "loss": 0.9245, + "step": 7249 + }, + { + "epoch": 1.2907763532763532, + "grad_norm": 0.6398176550865173, + "learning_rate": 0.0001530270058973068, + "loss": 1.0452, + "step": 7250 + }, + { + "epoch": 1.290954415954416, + "grad_norm": 0.5462267994880676, + "learning_rate": 0.00015301513791791542, + "loss": 0.8451, + "step": 7251 + }, + { + "epoch": 1.2911324786324787, + "grad_norm": 0.7536166906356812, + "learning_rate": 0.00015300326889980252, + "loss": 1.0086, + "step": 7252 + }, + { + "epoch": 1.2913105413105412, + "grad_norm": 0.6208569407463074, + "learning_rate": 0.00015299139884320065, + "loss": 0.7437, + "step": 7253 + }, + { + "epoch": 1.291488603988604, + "grad_norm": 0.7025452852249146, + "learning_rate": 0.00015297952774834242, + "loss": 0.8874, + "step": 7254 + }, + { + "epoch": 1.2916666666666667, + "grad_norm": 0.6758308410644531, + "learning_rate": 0.00015296765561546041, + "loss": 1.0378, + "step": 7255 + }, + { + "epoch": 1.2918447293447293, + "grad_norm": 0.7170431613922119, + "learning_rate": 0.00015295578244478724, + "loss": 1.0111, + "step": 7256 + }, + { + "epoch": 1.292022792022792, + "grad_norm": 0.6263511180877686, + "learning_rate": 0.00015294390823655544, + "loss": 0.7836, + "step": 7257 + }, + { + "epoch": 1.2922008547008548, + "grad_norm": 0.5887803435325623, + "learning_rate": 0.0001529320329909978, + "loss": 1.068, + "step": 7258 + }, + { + "epoch": 1.2923789173789173, + "grad_norm": 0.5955889821052551, + "learning_rate": 0.00015292015670834692, + "loss": 0.8903, + "step": 7259 + }, + { + "epoch": 1.29255698005698, + "grad_norm": 0.630449652671814, + "learning_rate": 0.00015290827938883552, + "loss": 1.1096, + "step": 7260 + }, + { + "epoch": 1.2927350427350428, + "grad_norm": 0.7405480146408081, + "learning_rate": 0.00015289640103269625, + "loss": 1.0648, + "step": 7261 + }, + { + "epoch": 1.2929131054131053, + "grad_norm": 0.6082221865653992, + "learning_rate": 0.00015288452164016191, + "loss": 0.9266, + "step": 7262 + }, + { + "epoch": 1.293091168091168, + "grad_norm": 0.6211720108985901, + "learning_rate": 0.00015287264121146524, + "loss": 0.849, + "step": 7263 + }, + { + "epoch": 1.2932692307692308, + "grad_norm": 0.6481043100357056, + "learning_rate": 0.00015286075974683898, + "loss": 0.7761, + "step": 7264 + }, + { + "epoch": 1.2934472934472934, + "grad_norm": 0.5957167744636536, + "learning_rate": 0.00015284887724651593, + "loss": 0.8942, + "step": 7265 + }, + { + "epoch": 1.2936253561253561, + "grad_norm": 0.7272268533706665, + "learning_rate": 0.00015283699371072894, + "loss": 1.0913, + "step": 7266 + }, + { + "epoch": 1.2938034188034189, + "grad_norm": 0.5902758836746216, + "learning_rate": 0.0001528251091397108, + "loss": 1.1045, + "step": 7267 + }, + { + "epoch": 1.2939814814814814, + "grad_norm": 0.6382482051849365, + "learning_rate": 0.00015281322353369436, + "loss": 0.9265, + "step": 7268 + }, + { + "epoch": 1.2941595441595442, + "grad_norm": 0.6556048393249512, + "learning_rate": 0.00015280133689291256, + "loss": 1.0536, + "step": 7269 + }, + { + "epoch": 1.294337606837607, + "grad_norm": 0.680895209312439, + "learning_rate": 0.00015278944921759822, + "loss": 0.9996, + "step": 7270 + }, + { + "epoch": 1.2945156695156697, + "grad_norm": 0.670317530632019, + "learning_rate": 0.00015277756050798428, + "loss": 1.1402, + "step": 7271 + }, + { + "epoch": 1.2946937321937322, + "grad_norm": 0.6312688589096069, + "learning_rate": 0.0001527656707643037, + "loss": 1.0669, + "step": 7272 + }, + { + "epoch": 1.294871794871795, + "grad_norm": 0.6267009973526001, + "learning_rate": 0.0001527537799867894, + "loss": 0.8985, + "step": 7273 + }, + { + "epoch": 1.2950498575498575, + "grad_norm": 0.7069001197814941, + "learning_rate": 0.00015274188817567436, + "loss": 0.9478, + "step": 7274 + }, + { + "epoch": 1.2952279202279202, + "grad_norm": 0.7229067087173462, + "learning_rate": 0.00015272999533119162, + "loss": 0.9005, + "step": 7275 + }, + { + "epoch": 1.295405982905983, + "grad_norm": 0.6254632472991943, + "learning_rate": 0.00015271810145357412, + "loss": 0.9746, + "step": 7276 + }, + { + "epoch": 1.2955840455840457, + "grad_norm": 0.6772669553756714, + "learning_rate": 0.00015270620654305494, + "loss": 1.1714, + "step": 7277 + }, + { + "epoch": 1.2957621082621082, + "grad_norm": 0.605576753616333, + "learning_rate": 0.00015269431059986713, + "loss": 0.7735, + "step": 7278 + }, + { + "epoch": 1.295940170940171, + "grad_norm": 0.7144771814346313, + "learning_rate": 0.00015268241362424378, + "loss": 0.9757, + "step": 7279 + }, + { + "epoch": 1.2961182336182335, + "grad_norm": 0.5275486707687378, + "learning_rate": 0.00015267051561641798, + "loss": 0.5669, + "step": 7280 + }, + { + "epoch": 1.2962962962962963, + "grad_norm": 0.6619452238082886, + "learning_rate": 0.00015265861657662284, + "loss": 0.9511, + "step": 7281 + }, + { + "epoch": 1.296474358974359, + "grad_norm": 0.6788223385810852, + "learning_rate": 0.00015264671650509147, + "loss": 1.2649, + "step": 7282 + }, + { + "epoch": 1.2966524216524218, + "grad_norm": 0.6198732852935791, + "learning_rate": 0.00015263481540205706, + "loss": 1.0659, + "step": 7283 + }, + { + "epoch": 1.2968304843304843, + "grad_norm": 0.6038815975189209, + "learning_rate": 0.0001526229132677528, + "loss": 1.0655, + "step": 7284 + }, + { + "epoch": 1.297008547008547, + "grad_norm": 0.7616196870803833, + "learning_rate": 0.00015261101010241186, + "loss": 1.131, + "step": 7285 + }, + { + "epoch": 1.2971866096866096, + "grad_norm": 0.7002527713775635, + "learning_rate": 0.00015259910590626746, + "loss": 1.1375, + "step": 7286 + }, + { + "epoch": 1.2973646723646723, + "grad_norm": 0.6067437529563904, + "learning_rate": 0.00015258720067955284, + "loss": 0.9306, + "step": 7287 + }, + { + "epoch": 1.297542735042735, + "grad_norm": 0.653232216835022, + "learning_rate": 0.00015257529442250128, + "loss": 1.107, + "step": 7288 + }, + { + "epoch": 1.2977207977207978, + "grad_norm": 0.6969175934791565, + "learning_rate": 0.00015256338713534603, + "loss": 0.8365, + "step": 7289 + }, + { + "epoch": 1.2978988603988604, + "grad_norm": 0.6176731586456299, + "learning_rate": 0.00015255147881832043, + "loss": 0.9707, + "step": 7290 + }, + { + "epoch": 1.2980769230769231, + "grad_norm": 0.6543741822242737, + "learning_rate": 0.00015253956947165772, + "loss": 0.7714, + "step": 7291 + }, + { + "epoch": 1.2982549857549857, + "grad_norm": 0.5224920511245728, + "learning_rate": 0.00015252765909559135, + "loss": 0.7469, + "step": 7292 + }, + { + "epoch": 1.2984330484330484, + "grad_norm": 0.638708770275116, + "learning_rate": 0.00015251574769035455, + "loss": 1.0965, + "step": 7293 + }, + { + "epoch": 1.2986111111111112, + "grad_norm": 0.6742943525314331, + "learning_rate": 0.0001525038352561808, + "loss": 1.1286, + "step": 7294 + }, + { + "epoch": 1.298789173789174, + "grad_norm": 0.6027839183807373, + "learning_rate": 0.00015249192179330346, + "loss": 0.8824, + "step": 7295 + }, + { + "epoch": 1.2989672364672364, + "grad_norm": 0.7462167143821716, + "learning_rate": 0.00015248000730195597, + "loss": 0.94, + "step": 7296 + }, + { + "epoch": 1.2991452991452992, + "grad_norm": 0.6972534656524658, + "learning_rate": 0.00015246809178237172, + "loss": 1.0664, + "step": 7297 + }, + { + "epoch": 1.2993233618233617, + "grad_norm": 0.569949209690094, + "learning_rate": 0.0001524561752347842, + "loss": 0.691, + "step": 7298 + }, + { + "epoch": 1.2995014245014245, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00015244425765942695, + "loss": 1.083, + "step": 7299 + }, + { + "epoch": 1.2996794871794872, + "grad_norm": 0.6927483677864075, + "learning_rate": 0.00015243233905653337, + "loss": 1.0068, + "step": 7300 + }, + { + "epoch": 1.29985754985755, + "grad_norm": 0.752824604511261, + "learning_rate": 0.00015242041942633704, + "loss": 0.9946, + "step": 7301 + }, + { + "epoch": 1.3000356125356125, + "grad_norm": 0.6532080173492432, + "learning_rate": 0.0001524084987690715, + "loss": 1.2326, + "step": 7302 + }, + { + "epoch": 1.3002136752136753, + "grad_norm": 0.7954180836677551, + "learning_rate": 0.0001523965770849703, + "loss": 1.1105, + "step": 7303 + }, + { + "epoch": 1.3003917378917378, + "grad_norm": 0.5971781015396118, + "learning_rate": 0.000152384654374267, + "loss": 1.0984, + "step": 7304 + }, + { + "epoch": 1.3005698005698005, + "grad_norm": 0.7778682112693787, + "learning_rate": 0.0001523727306371952, + "loss": 1.0795, + "step": 7305 + }, + { + "epoch": 1.3007478632478633, + "grad_norm": 0.6712004542350769, + "learning_rate": 0.00015236080587398856, + "loss": 1.0814, + "step": 7306 + }, + { + "epoch": 1.300925925925926, + "grad_norm": 0.581048846244812, + "learning_rate": 0.00015234888008488066, + "loss": 0.9868, + "step": 7307 + }, + { + "epoch": 1.3011039886039886, + "grad_norm": 0.697695791721344, + "learning_rate": 0.00015233695327010523, + "loss": 1.1045, + "step": 7308 + }, + { + "epoch": 1.3012820512820513, + "grad_norm": 0.6858421564102173, + "learning_rate": 0.00015232502542989593, + "loss": 1.0769, + "step": 7309 + }, + { + "epoch": 1.3014601139601139, + "grad_norm": 0.6312826871871948, + "learning_rate": 0.00015231309656448642, + "loss": 0.9523, + "step": 7310 + }, + { + "epoch": 1.3016381766381766, + "grad_norm": 0.9243300557136536, + "learning_rate": 0.0001523011666741105, + "loss": 0.947, + "step": 7311 + }, + { + "epoch": 1.3018162393162394, + "grad_norm": 0.6808217763900757, + "learning_rate": 0.00015228923575900184, + "loss": 0.8631, + "step": 7312 + }, + { + "epoch": 1.301994301994302, + "grad_norm": 0.6713891625404358, + "learning_rate": 0.00015227730381939424, + "loss": 0.9157, + "step": 7313 + }, + { + "epoch": 1.3021723646723646, + "grad_norm": 0.6802582740783691, + "learning_rate": 0.00015226537085552146, + "loss": 1.041, + "step": 7314 + }, + { + "epoch": 1.3023504273504274, + "grad_norm": 0.6543951034545898, + "learning_rate": 0.0001522534368676173, + "loss": 0.8709, + "step": 7315 + }, + { + "epoch": 1.30252849002849, + "grad_norm": 0.6290678381919861, + "learning_rate": 0.0001522415018559156, + "loss": 1.0568, + "step": 7316 + }, + { + "epoch": 1.3027065527065527, + "grad_norm": 0.6590015292167664, + "learning_rate": 0.0001522295658206502, + "loss": 0.9919, + "step": 7317 + }, + { + "epoch": 1.3028846153846154, + "grad_norm": 0.6374103426933289, + "learning_rate": 0.00015221762876205494, + "loss": 0.878, + "step": 7318 + }, + { + "epoch": 1.3030626780626782, + "grad_norm": 0.7247048616409302, + "learning_rate": 0.00015220569068036372, + "loss": 1.061, + "step": 7319 + }, + { + "epoch": 1.3032407407407407, + "grad_norm": 0.6450991630554199, + "learning_rate": 0.00015219375157581047, + "loss": 0.9389, + "step": 7320 + }, + { + "epoch": 1.3034188034188035, + "grad_norm": 0.8039840459823608, + "learning_rate": 0.00015218181144862903, + "loss": 1.0692, + "step": 7321 + }, + { + "epoch": 1.303596866096866, + "grad_norm": 0.6539456248283386, + "learning_rate": 0.00015216987029905346, + "loss": 1.0478, + "step": 7322 + }, + { + "epoch": 1.3037749287749287, + "grad_norm": 0.60880047082901, + "learning_rate": 0.00015215792812731758, + "loss": 0.8412, + "step": 7323 + }, + { + "epoch": 1.3039529914529915, + "grad_norm": 0.6757258176803589, + "learning_rate": 0.0001521459849336555, + "loss": 0.896, + "step": 7324 + }, + { + "epoch": 1.3041310541310542, + "grad_norm": 0.6735622882843018, + "learning_rate": 0.00015213404071830116, + "loss": 1.1078, + "step": 7325 + }, + { + "epoch": 1.3043091168091168, + "grad_norm": 0.7321233749389648, + "learning_rate": 0.00015212209548148858, + "loss": 1.1021, + "step": 7326 + }, + { + "epoch": 1.3044871794871795, + "grad_norm": 0.6678910851478577, + "learning_rate": 0.00015211014922345182, + "loss": 1.0043, + "step": 7327 + }, + { + "epoch": 1.304665242165242, + "grad_norm": 0.6876940727233887, + "learning_rate": 0.0001520982019444249, + "loss": 1.0376, + "step": 7328 + }, + { + "epoch": 1.3048433048433048, + "grad_norm": 0.6171853542327881, + "learning_rate": 0.00015208625364464195, + "loss": 0.839, + "step": 7329 + }, + { + "epoch": 1.3050213675213675, + "grad_norm": 0.6449569463729858, + "learning_rate": 0.0001520743043243371, + "loss": 1.0908, + "step": 7330 + }, + { + "epoch": 1.3051994301994303, + "grad_norm": 0.6894628405570984, + "learning_rate": 0.00015206235398374443, + "loss": 1.0263, + "step": 7331 + }, + { + "epoch": 1.3053774928774928, + "grad_norm": 0.5853552222251892, + "learning_rate": 0.00015205040262309804, + "loss": 0.8342, + "step": 7332 + }, + { + "epoch": 1.3055555555555556, + "grad_norm": 0.5934799313545227, + "learning_rate": 0.00015203845024263214, + "loss": 0.9464, + "step": 7333 + }, + { + "epoch": 1.305733618233618, + "grad_norm": 0.668927788734436, + "learning_rate": 0.00015202649684258095, + "loss": 0.9018, + "step": 7334 + }, + { + "epoch": 1.3059116809116809, + "grad_norm": 0.676810085773468, + "learning_rate": 0.0001520145424231786, + "loss": 0.9284, + "step": 7335 + }, + { + "epoch": 1.3060897435897436, + "grad_norm": 0.6223878264427185, + "learning_rate": 0.00015200258698465935, + "loss": 1.0779, + "step": 7336 + }, + { + "epoch": 1.3062678062678064, + "grad_norm": 0.6092363595962524, + "learning_rate": 0.00015199063052725745, + "loss": 0.8602, + "step": 7337 + }, + { + "epoch": 1.306445868945869, + "grad_norm": 0.7668731212615967, + "learning_rate": 0.00015197867305120712, + "loss": 1.0756, + "step": 7338 + }, + { + "epoch": 1.3066239316239316, + "grad_norm": 0.6485331654548645, + "learning_rate": 0.00015196671455674268, + "loss": 1.0193, + "step": 7339 + }, + { + "epoch": 1.3068019943019942, + "grad_norm": 0.5661036372184753, + "learning_rate": 0.0001519547550440984, + "loss": 0.8321, + "step": 7340 + }, + { + "epoch": 1.306980056980057, + "grad_norm": 0.6270507574081421, + "learning_rate": 0.00015194279451350866, + "loss": 0.6403, + "step": 7341 + }, + { + "epoch": 1.3071581196581197, + "grad_norm": 0.7283764481544495, + "learning_rate": 0.00015193083296520773, + "loss": 1.0401, + "step": 7342 + }, + { + "epoch": 1.3073361823361824, + "grad_norm": 0.658835232257843, + "learning_rate": 0.00015191887039943, + "loss": 1.0172, + "step": 7343 + }, + { + "epoch": 1.307514245014245, + "grad_norm": 0.6288984417915344, + "learning_rate": 0.00015190690681640988, + "loss": 0.8649, + "step": 7344 + }, + { + "epoch": 1.3076923076923077, + "grad_norm": 0.666442334651947, + "learning_rate": 0.00015189494221638176, + "loss": 1.0757, + "step": 7345 + }, + { + "epoch": 1.3078703703703702, + "grad_norm": 0.6116433143615723, + "learning_rate": 0.00015188297659958003, + "loss": 0.9244, + "step": 7346 + }, + { + "epoch": 1.308048433048433, + "grad_norm": 0.6378964185714722, + "learning_rate": 0.0001518710099662392, + "loss": 0.9629, + "step": 7347 + }, + { + "epoch": 1.3082264957264957, + "grad_norm": 0.6258945465087891, + "learning_rate": 0.00015185904231659357, + "loss": 0.8524, + "step": 7348 + }, + { + "epoch": 1.3084045584045585, + "grad_norm": 0.6498504877090454, + "learning_rate": 0.0001518470736508778, + "loss": 0.9685, + "step": 7349 + }, + { + "epoch": 1.308582621082621, + "grad_norm": 0.6928247809410095, + "learning_rate": 0.00015183510396932635, + "loss": 0.9054, + "step": 7350 + }, + { + "epoch": 1.3087606837606838, + "grad_norm": 0.6350936889648438, + "learning_rate": 0.0001518231332721737, + "loss": 1.0039, + "step": 7351 + }, + { + "epoch": 1.3089387464387463, + "grad_norm": 0.6652286648750305, + "learning_rate": 0.00015181116155965437, + "loss": 0.8946, + "step": 7352 + }, + { + "epoch": 1.309116809116809, + "grad_norm": 0.6554864048957825, + "learning_rate": 0.000151799188832003, + "loss": 0.9518, + "step": 7353 + }, + { + "epoch": 1.3092948717948718, + "grad_norm": 0.7523114085197449, + "learning_rate": 0.0001517872150894541, + "loss": 0.9462, + "step": 7354 + }, + { + "epoch": 1.3094729344729346, + "grad_norm": 0.7113336324691772, + "learning_rate": 0.0001517752403322423, + "loss": 1.2347, + "step": 7355 + }, + { + "epoch": 1.309650997150997, + "grad_norm": 0.6461622714996338, + "learning_rate": 0.00015176326456060223, + "loss": 0.8891, + "step": 7356 + }, + { + "epoch": 1.3098290598290598, + "grad_norm": 0.7429143190383911, + "learning_rate": 0.00015175128777476852, + "loss": 1.1944, + "step": 7357 + }, + { + "epoch": 1.3100071225071226, + "grad_norm": 0.6816306114196777, + "learning_rate": 0.00015173930997497585, + "loss": 1.1445, + "step": 7358 + }, + { + "epoch": 1.3101851851851851, + "grad_norm": 0.6644450426101685, + "learning_rate": 0.00015172733116145884, + "loss": 0.9808, + "step": 7359 + }, + { + "epoch": 1.3103632478632479, + "grad_norm": 0.6921063661575317, + "learning_rate": 0.00015171535133445225, + "loss": 1.0162, + "step": 7360 + }, + { + "epoch": 1.3105413105413106, + "grad_norm": 0.6386187672615051, + "learning_rate": 0.00015170337049419082, + "loss": 0.9951, + "step": 7361 + }, + { + "epoch": 1.3107193732193732, + "grad_norm": 0.6505418419837952, + "learning_rate": 0.0001516913886409092, + "loss": 0.8872, + "step": 7362 + }, + { + "epoch": 1.310897435897436, + "grad_norm": 0.6415576934814453, + "learning_rate": 0.00015167940577484222, + "loss": 1.056, + "step": 7363 + }, + { + "epoch": 1.3110754985754987, + "grad_norm": 0.6691195964813232, + "learning_rate": 0.00015166742189622458, + "loss": 1.0561, + "step": 7364 + }, + { + "epoch": 1.3112535612535612, + "grad_norm": 0.6376257538795471, + "learning_rate": 0.00015165543700529122, + "loss": 0.8499, + "step": 7365 + }, + { + "epoch": 1.311431623931624, + "grad_norm": 0.6270790696144104, + "learning_rate": 0.00015164345110227684, + "loss": 1.0244, + "step": 7366 + }, + { + "epoch": 1.3116096866096867, + "grad_norm": 0.7120122313499451, + "learning_rate": 0.0001516314641874163, + "loss": 1.0476, + "step": 7367 + }, + { + "epoch": 1.3117877492877492, + "grad_norm": 0.6152660250663757, + "learning_rate": 0.0001516194762609445, + "loss": 0.897, + "step": 7368 + }, + { + "epoch": 1.311965811965812, + "grad_norm": 0.7578088045120239, + "learning_rate": 0.00015160748732309626, + "loss": 1.1609, + "step": 7369 + }, + { + "epoch": 1.3121438746438747, + "grad_norm": 0.6594924330711365, + "learning_rate": 0.00015159549737410656, + "loss": 1.1706, + "step": 7370 + }, + { + "epoch": 1.3123219373219372, + "grad_norm": 0.6559173464775085, + "learning_rate": 0.00015158350641421024, + "loss": 0.9452, + "step": 7371 + }, + { + "epoch": 1.3125, + "grad_norm": 0.6667516231536865, + "learning_rate": 0.00015157151444364226, + "loss": 0.8153, + "step": 7372 + }, + { + "epoch": 1.3126780626780628, + "grad_norm": 0.7054803371429443, + "learning_rate": 0.00015155952146263761, + "loss": 0.9887, + "step": 7373 + }, + { + "epoch": 1.3128561253561253, + "grad_norm": 0.7035902142524719, + "learning_rate": 0.00015154752747143123, + "loss": 1.1832, + "step": 7374 + }, + { + "epoch": 1.313034188034188, + "grad_norm": 0.6297488212585449, + "learning_rate": 0.00015153553247025813, + "loss": 0.9602, + "step": 7375 + }, + { + "epoch": 1.3132122507122508, + "grad_norm": 0.6851378083229065, + "learning_rate": 0.00015152353645935335, + "loss": 1.0743, + "step": 7376 + }, + { + "epoch": 1.3133903133903133, + "grad_norm": 0.6215537786483765, + "learning_rate": 0.00015151153943895187, + "loss": 0.9484, + "step": 7377 + }, + { + "epoch": 1.313568376068376, + "grad_norm": 0.6848666071891785, + "learning_rate": 0.0001514995414092888, + "loss": 1.0978, + "step": 7378 + }, + { + "epoch": 1.3137464387464388, + "grad_norm": 0.7527492642402649, + "learning_rate": 0.00015148754237059918, + "loss": 1.083, + "step": 7379 + }, + { + "epoch": 1.3139245014245013, + "grad_norm": 0.6264588236808777, + "learning_rate": 0.00015147554232311814, + "loss": 0.9995, + "step": 7380 + }, + { + "epoch": 1.314102564102564, + "grad_norm": 0.6666619181632996, + "learning_rate": 0.00015146354126708075, + "loss": 1.0156, + "step": 7381 + }, + { + "epoch": 1.3142806267806268, + "grad_norm": 0.6626597046852112, + "learning_rate": 0.00015145153920272222, + "loss": 1.0047, + "step": 7382 + }, + { + "epoch": 1.3144586894586894, + "grad_norm": 0.5975428223609924, + "learning_rate": 0.0001514395361302776, + "loss": 0.806, + "step": 7383 + }, + { + "epoch": 1.3146367521367521, + "grad_norm": 0.6509957909584045, + "learning_rate": 0.00015142753204998218, + "loss": 0.8871, + "step": 7384 + }, + { + "epoch": 1.3148148148148149, + "grad_norm": 0.6672926545143127, + "learning_rate": 0.00015141552696207108, + "loss": 0.9616, + "step": 7385 + }, + { + "epoch": 1.3149928774928774, + "grad_norm": 0.6965435147285461, + "learning_rate": 0.00015140352086677954, + "loss": 1.124, + "step": 7386 + }, + { + "epoch": 1.3151709401709402, + "grad_norm": 0.6559258103370667, + "learning_rate": 0.00015139151376434277, + "loss": 1.0271, + "step": 7387 + }, + { + "epoch": 1.315349002849003, + "grad_norm": 0.7613587379455566, + "learning_rate": 0.00015137950565499608, + "loss": 1.0349, + "step": 7388 + }, + { + "epoch": 1.3155270655270654, + "grad_norm": 0.7001944780349731, + "learning_rate": 0.0001513674965389747, + "loss": 0.8551, + "step": 7389 + }, + { + "epoch": 1.3157051282051282, + "grad_norm": 0.6087043285369873, + "learning_rate": 0.0001513554864165139, + "loss": 0.7118, + "step": 7390 + }, + { + "epoch": 1.315883190883191, + "grad_norm": 0.71526700258255, + "learning_rate": 0.00015134347528784908, + "loss": 1.0478, + "step": 7391 + }, + { + "epoch": 1.3160612535612537, + "grad_norm": 0.6182073950767517, + "learning_rate": 0.00015133146315321548, + "loss": 0.9474, + "step": 7392 + }, + { + "epoch": 1.3162393162393162, + "grad_norm": 0.7771387696266174, + "learning_rate": 0.0001513194500128485, + "loss": 1.0544, + "step": 7393 + }, + { + "epoch": 1.316417378917379, + "grad_norm": 0.7108260989189148, + "learning_rate": 0.00015130743586698353, + "loss": 0.8813, + "step": 7394 + }, + { + "epoch": 1.3165954415954415, + "grad_norm": 0.7057309150695801, + "learning_rate": 0.0001512954207158559, + "loss": 0.899, + "step": 7395 + }, + { + "epoch": 1.3167735042735043, + "grad_norm": 0.6139237880706787, + "learning_rate": 0.00015128340455970106, + "loss": 0.8885, + "step": 7396 + }, + { + "epoch": 1.316951566951567, + "grad_norm": 0.7166598439216614, + "learning_rate": 0.00015127138739875443, + "loss": 0.9792, + "step": 7397 + }, + { + "epoch": 1.3171296296296298, + "grad_norm": 0.6916186809539795, + "learning_rate": 0.00015125936923325153, + "loss": 0.8871, + "step": 7398 + }, + { + "epoch": 1.3173076923076923, + "grad_norm": 0.7189087271690369, + "learning_rate": 0.0001512473500634277, + "loss": 0.8302, + "step": 7399 + }, + { + "epoch": 1.317485754985755, + "grad_norm": 0.5739200115203857, + "learning_rate": 0.00015123532988951853, + "loss": 0.9137, + "step": 7400 + }, + { + "epoch": 1.3176638176638176, + "grad_norm": 0.7661057114601135, + "learning_rate": 0.00015122330871175952, + "loss": 1.1255, + "step": 7401 + }, + { + "epoch": 1.3178418803418803, + "grad_norm": 0.6487592458724976, + "learning_rate": 0.00015121128653038617, + "loss": 1.0519, + "step": 7402 + }, + { + "epoch": 1.318019943019943, + "grad_norm": 0.693134605884552, + "learning_rate": 0.00015119926334563406, + "loss": 0.9585, + "step": 7403 + }, + { + "epoch": 1.3181980056980058, + "grad_norm": 0.5895997285842896, + "learning_rate": 0.0001511872391577387, + "loss": 0.8033, + "step": 7404 + }, + { + "epoch": 1.3183760683760684, + "grad_norm": 0.654876172542572, + "learning_rate": 0.00015117521396693575, + "loss": 1.0082, + "step": 7405 + }, + { + "epoch": 1.318554131054131, + "grad_norm": 0.5877239108085632, + "learning_rate": 0.0001511631877734608, + "loss": 1.0147, + "step": 7406 + }, + { + "epoch": 1.3187321937321936, + "grad_norm": 0.6109837889671326, + "learning_rate": 0.00015115116057754944, + "loss": 0.7498, + "step": 7407 + }, + { + "epoch": 1.3189102564102564, + "grad_norm": 0.643856942653656, + "learning_rate": 0.00015113913237943736, + "loss": 1.0417, + "step": 7408 + }, + { + "epoch": 1.3190883190883191, + "grad_norm": 0.654077410697937, + "learning_rate": 0.00015112710317936022, + "loss": 1.1809, + "step": 7409 + }, + { + "epoch": 1.319266381766382, + "grad_norm": 0.6785375475883484, + "learning_rate": 0.00015111507297755367, + "loss": 0.9447, + "step": 7410 + }, + { + "epoch": 1.3194444444444444, + "grad_norm": 0.6513382196426392, + "learning_rate": 0.00015110304177425347, + "loss": 0.8286, + "step": 7411 + }, + { + "epoch": 1.3196225071225072, + "grad_norm": 0.6536405682563782, + "learning_rate": 0.00015109100956969533, + "loss": 1.1959, + "step": 7412 + }, + { + "epoch": 1.3198005698005697, + "grad_norm": 0.6633172035217285, + "learning_rate": 0.00015107897636411498, + "loss": 0.8839, + "step": 7413 + }, + { + "epoch": 1.3199786324786325, + "grad_norm": 0.5773791670799255, + "learning_rate": 0.00015106694215774821, + "loss": 0.9785, + "step": 7414 + }, + { + "epoch": 1.3201566951566952, + "grad_norm": 0.7005468010902405, + "learning_rate": 0.00015105490695083078, + "loss": 1.0752, + "step": 7415 + }, + { + "epoch": 1.320334757834758, + "grad_norm": 0.6509538888931274, + "learning_rate": 0.0001510428707435985, + "loss": 0.9886, + "step": 7416 + }, + { + "epoch": 1.3205128205128205, + "grad_norm": 0.6607788801193237, + "learning_rate": 0.0001510308335362872, + "loss": 0.9756, + "step": 7417 + }, + { + "epoch": 1.3206908831908832, + "grad_norm": 0.5977858304977417, + "learning_rate": 0.00015101879532913274, + "loss": 1.0574, + "step": 7418 + }, + { + "epoch": 1.3208689458689458, + "grad_norm": 0.6478607058525085, + "learning_rate": 0.00015100675612237096, + "loss": 1.0076, + "step": 7419 + }, + { + "epoch": 1.3210470085470085, + "grad_norm": 0.6386681199073792, + "learning_rate": 0.00015099471591623775, + "loss": 0.9639, + "step": 7420 + }, + { + "epoch": 1.3212250712250713, + "grad_norm": 0.6348143815994263, + "learning_rate": 0.000150982674710969, + "loss": 1.0226, + "step": 7421 + }, + { + "epoch": 1.321403133903134, + "grad_norm": 0.6737388372421265, + "learning_rate": 0.00015097063250680068, + "loss": 0.9985, + "step": 7422 + }, + { + "epoch": 1.3215811965811965, + "grad_norm": 0.7302656769752502, + "learning_rate": 0.00015095858930396866, + "loss": 0.9969, + "step": 7423 + }, + { + "epoch": 1.3217592592592593, + "grad_norm": 0.7062691450119019, + "learning_rate": 0.00015094654510270898, + "loss": 0.9137, + "step": 7424 + }, + { + "epoch": 1.3219373219373218, + "grad_norm": 0.6289888620376587, + "learning_rate": 0.00015093449990325754, + "loss": 0.9231, + "step": 7425 + }, + { + "epoch": 1.3221153846153846, + "grad_norm": 0.643284261226654, + "learning_rate": 0.0001509224537058504, + "loss": 0.8981, + "step": 7426 + }, + { + "epoch": 1.3222934472934473, + "grad_norm": 0.7019244432449341, + "learning_rate": 0.00015091040651072355, + "loss": 0.9994, + "step": 7427 + }, + { + "epoch": 1.32247150997151, + "grad_norm": 0.5982088446617126, + "learning_rate": 0.0001508983583181131, + "loss": 0.9365, + "step": 7428 + }, + { + "epoch": 1.3226495726495726, + "grad_norm": 0.6086063385009766, + "learning_rate": 0.00015088630912825498, + "loss": 0.8621, + "step": 7429 + }, + { + "epoch": 1.3228276353276354, + "grad_norm": 0.6829213500022888, + "learning_rate": 0.00015087425894138535, + "loss": 1.1959, + "step": 7430 + }, + { + "epoch": 1.323005698005698, + "grad_norm": 0.6538017392158508, + "learning_rate": 0.00015086220775774033, + "loss": 0.9412, + "step": 7431 + }, + { + "epoch": 1.3231837606837606, + "grad_norm": 0.6334070563316345, + "learning_rate": 0.00015085015557755597, + "loss": 0.9044, + "step": 7432 + }, + { + "epoch": 1.3233618233618234, + "grad_norm": 0.6514624357223511, + "learning_rate": 0.00015083810240106845, + "loss": 0.8859, + "step": 7433 + }, + { + "epoch": 1.3235398860398861, + "grad_norm": 0.7130434513092041, + "learning_rate": 0.00015082604822851397, + "loss": 1.2845, + "step": 7434 + }, + { + "epoch": 1.3237179487179487, + "grad_norm": 0.609419584274292, + "learning_rate": 0.00015081399306012862, + "loss": 1.0725, + "step": 7435 + }, + { + "epoch": 1.3238960113960114, + "grad_norm": 0.586807370185852, + "learning_rate": 0.0001508019368961486, + "loss": 0.9032, + "step": 7436 + }, + { + "epoch": 1.324074074074074, + "grad_norm": 0.6937291026115417, + "learning_rate": 0.0001507898797368102, + "loss": 0.7975, + "step": 7437 + }, + { + "epoch": 1.3242521367521367, + "grad_norm": 0.6804966330528259, + "learning_rate": 0.00015077782158234962, + "loss": 1.1018, + "step": 7438 + }, + { + "epoch": 1.3244301994301995, + "grad_norm": 0.6110677123069763, + "learning_rate": 0.0001507657624330031, + "loss": 0.7988, + "step": 7439 + }, + { + "epoch": 1.3246082621082622, + "grad_norm": 0.6340961456298828, + "learning_rate": 0.0001507537022890069, + "loss": 0.844, + "step": 7440 + }, + { + "epoch": 1.3247863247863247, + "grad_norm": 0.7291021943092346, + "learning_rate": 0.00015074164115059735, + "loss": 0.9867, + "step": 7441 + }, + { + "epoch": 1.3249643874643875, + "grad_norm": 0.6818505525588989, + "learning_rate": 0.00015072957901801076, + "loss": 1.1541, + "step": 7442 + }, + { + "epoch": 1.32514245014245, + "grad_norm": 0.6174707412719727, + "learning_rate": 0.00015071751589148345, + "loss": 1.1679, + "step": 7443 + }, + { + "epoch": 1.3253205128205128, + "grad_norm": 0.6481367945671082, + "learning_rate": 0.00015070545177125176, + "loss": 1.0955, + "step": 7444 + }, + { + "epoch": 1.3254985754985755, + "grad_norm": 0.6752339005470276, + "learning_rate": 0.00015069338665755203, + "loss": 0.8651, + "step": 7445 + }, + { + "epoch": 1.3256766381766383, + "grad_norm": 0.6608055830001831, + "learning_rate": 0.00015068132055062077, + "loss": 0.9553, + "step": 7446 + }, + { + "epoch": 1.3258547008547008, + "grad_norm": 0.5933246612548828, + "learning_rate": 0.00015066925345069425, + "loss": 0.8584, + "step": 7447 + }, + { + "epoch": 1.3260327635327636, + "grad_norm": 0.6301844716072083, + "learning_rate": 0.000150657185358009, + "loss": 0.8583, + "step": 7448 + }, + { + "epoch": 1.326210826210826, + "grad_norm": 0.7359434962272644, + "learning_rate": 0.00015064511627280145, + "loss": 1.0905, + "step": 7449 + }, + { + "epoch": 1.3263888888888888, + "grad_norm": 0.6334579586982727, + "learning_rate": 0.00015063304619530806, + "loss": 0.9814, + "step": 7450 + }, + { + "epoch": 1.3265669515669516, + "grad_norm": 0.6974197626113892, + "learning_rate": 0.00015062097512576528, + "loss": 0.9302, + "step": 7451 + }, + { + "epoch": 1.3267450142450143, + "grad_norm": 0.6895849704742432, + "learning_rate": 0.00015060890306440965, + "loss": 1.0175, + "step": 7452 + }, + { + "epoch": 1.3269230769230769, + "grad_norm": 0.5938003659248352, + "learning_rate": 0.00015059683001147767, + "loss": 0.8084, + "step": 7453 + }, + { + "epoch": 1.3271011396011396, + "grad_norm": 0.6821470856666565, + "learning_rate": 0.00015058475596720596, + "loss": 0.9897, + "step": 7454 + }, + { + "epoch": 1.3272792022792022, + "grad_norm": 0.5507164001464844, + "learning_rate": 0.00015057268093183104, + "loss": 0.7012, + "step": 7455 + }, + { + "epoch": 1.327457264957265, + "grad_norm": 0.6216199398040771, + "learning_rate": 0.00015056060490558945, + "loss": 1.0281, + "step": 7456 + }, + { + "epoch": 1.3276353276353277, + "grad_norm": 0.6674157977104187, + "learning_rate": 0.00015054852788871787, + "loss": 0.8776, + "step": 7457 + }, + { + "epoch": 1.3278133903133904, + "grad_norm": 0.666963517665863, + "learning_rate": 0.0001505364498814529, + "loss": 1.0742, + "step": 7458 + }, + { + "epoch": 1.327991452991453, + "grad_norm": 0.6205331683158875, + "learning_rate": 0.00015052437088403114, + "loss": 1.1109, + "step": 7459 + }, + { + "epoch": 1.3281695156695157, + "grad_norm": 0.6402750611305237, + "learning_rate": 0.00015051229089668933, + "loss": 1.0648, + "step": 7460 + }, + { + "epoch": 1.3283475783475782, + "grad_norm": 0.7445703744888306, + "learning_rate": 0.00015050020991966406, + "loss": 0.8989, + "step": 7461 + }, + { + "epoch": 1.328525641025641, + "grad_norm": 0.8131299614906311, + "learning_rate": 0.00015048812795319212, + "loss": 0.9552, + "step": 7462 + }, + { + "epoch": 1.3287037037037037, + "grad_norm": 0.7007313966751099, + "learning_rate": 0.00015047604499751017, + "loss": 0.9899, + "step": 7463 + }, + { + "epoch": 1.3288817663817665, + "grad_norm": 0.60536789894104, + "learning_rate": 0.000150463961052855, + "loss": 0.7694, + "step": 7464 + }, + { + "epoch": 1.329059829059829, + "grad_norm": 0.6910434365272522, + "learning_rate": 0.00015045187611946331, + "loss": 0.9575, + "step": 7465 + }, + { + "epoch": 1.3292378917378918, + "grad_norm": 0.7693352103233337, + "learning_rate": 0.00015043979019757194, + "loss": 1.1987, + "step": 7466 + }, + { + "epoch": 1.3294159544159543, + "grad_norm": 0.6675218939781189, + "learning_rate": 0.00015042770328741763, + "loss": 1.0099, + "step": 7467 + }, + { + "epoch": 1.329594017094017, + "grad_norm": 0.8040883541107178, + "learning_rate": 0.00015041561538923722, + "loss": 0.9493, + "step": 7468 + }, + { + "epoch": 1.3297720797720798, + "grad_norm": 0.6765826344490051, + "learning_rate": 0.00015040352650326762, + "loss": 1.1035, + "step": 7469 + }, + { + "epoch": 1.3299501424501425, + "grad_norm": 0.7099924087524414, + "learning_rate": 0.0001503914366297456, + "loss": 0.9198, + "step": 7470 + }, + { + "epoch": 1.330128205128205, + "grad_norm": 0.6673682928085327, + "learning_rate": 0.00015037934576890804, + "loss": 1.0234, + "step": 7471 + }, + { + "epoch": 1.3303062678062678, + "grad_norm": 0.7022300958633423, + "learning_rate": 0.00015036725392099184, + "loss": 1.3875, + "step": 7472 + }, + { + "epoch": 1.3304843304843303, + "grad_norm": 0.6997060179710388, + "learning_rate": 0.00015035516108623394, + "loss": 0.8114, + "step": 7473 + }, + { + "epoch": 1.330662393162393, + "grad_norm": 0.6262350678443909, + "learning_rate": 0.00015034306726487127, + "loss": 1.128, + "step": 7474 + }, + { + "epoch": 1.3308404558404558, + "grad_norm": 0.6330382227897644, + "learning_rate": 0.00015033097245714078, + "loss": 0.9032, + "step": 7475 + }, + { + "epoch": 1.3310185185185186, + "grad_norm": 0.6527551412582397, + "learning_rate": 0.00015031887666327944, + "loss": 0.9311, + "step": 7476 + }, + { + "epoch": 1.3311965811965811, + "grad_norm": 0.6754798889160156, + "learning_rate": 0.00015030677988352422, + "loss": 1.0626, + "step": 7477 + }, + { + "epoch": 1.3313746438746439, + "grad_norm": 0.6397945284843445, + "learning_rate": 0.00015029468211811216, + "loss": 0.9222, + "step": 7478 + }, + { + "epoch": 1.3315527065527066, + "grad_norm": 0.8163481950759888, + "learning_rate": 0.0001502825833672803, + "loss": 1.1827, + "step": 7479 + }, + { + "epoch": 1.3317307692307692, + "grad_norm": 0.6645621657371521, + "learning_rate": 0.00015027048363126566, + "loss": 0.9744, + "step": 7480 + }, + { + "epoch": 1.331908831908832, + "grad_norm": 0.6943182349205017, + "learning_rate": 0.0001502583829103053, + "loss": 1.1597, + "step": 7481 + }, + { + "epoch": 1.3320868945868947, + "grad_norm": 0.6283710598945618, + "learning_rate": 0.00015024628120463636, + "loss": 0.9514, + "step": 7482 + }, + { + "epoch": 1.3322649572649572, + "grad_norm": 0.6159678101539612, + "learning_rate": 0.0001502341785144959, + "loss": 0.9752, + "step": 7483 + }, + { + "epoch": 1.33244301994302, + "grad_norm": 0.6259802579879761, + "learning_rate": 0.00015022207484012107, + "loss": 0.9356, + "step": 7484 + }, + { + "epoch": 1.3326210826210827, + "grad_norm": 0.7322365641593933, + "learning_rate": 0.00015020997018174904, + "loss": 1.2072, + "step": 7485 + }, + { + "epoch": 1.3327991452991452, + "grad_norm": 0.6323443651199341, + "learning_rate": 0.0001501978645396169, + "loss": 1.1661, + "step": 7486 + }, + { + "epoch": 1.332977207977208, + "grad_norm": 0.7811527848243713, + "learning_rate": 0.00015018575791396187, + "loss": 1.0304, + "step": 7487 + }, + { + "epoch": 1.3331552706552707, + "grad_norm": 0.7221232056617737, + "learning_rate": 0.0001501736503050212, + "loss": 0.8838, + "step": 7488 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.6980099081993103, + "learning_rate": 0.00015016154171303207, + "loss": 1.1841, + "step": 7489 + }, + { + "epoch": 1.333511396011396, + "grad_norm": 0.6802879571914673, + "learning_rate": 0.00015014943213823175, + "loss": 0.959, + "step": 7490 + }, + { + "epoch": 1.3336894586894588, + "grad_norm": 0.637698233127594, + "learning_rate": 0.00015013732158085746, + "loss": 1.0517, + "step": 7491 + }, + { + "epoch": 1.3338675213675213, + "grad_norm": 0.6386787295341492, + "learning_rate": 0.0001501252100411465, + "loss": 0.7125, + "step": 7492 + }, + { + "epoch": 1.334045584045584, + "grad_norm": 0.6287358403205872, + "learning_rate": 0.0001501130975193362, + "loss": 0.8913, + "step": 7493 + }, + { + "epoch": 1.3342236467236468, + "grad_norm": 0.6142337322235107, + "learning_rate": 0.00015010098401566386, + "loss": 0.8149, + "step": 7494 + }, + { + "epoch": 1.3344017094017093, + "grad_norm": 0.6369916200637817, + "learning_rate": 0.0001500888695303668, + "loss": 1.0186, + "step": 7495 + }, + { + "epoch": 1.334579772079772, + "grad_norm": 0.7526934146881104, + "learning_rate": 0.0001500767540636824, + "loss": 1.2421, + "step": 7496 + }, + { + "epoch": 1.3347578347578348, + "grad_norm": 0.7278095483779907, + "learning_rate": 0.00015006463761584802, + "loss": 0.9856, + "step": 7497 + }, + { + "epoch": 1.3349358974358974, + "grad_norm": 0.6165127158164978, + "learning_rate": 0.00015005252018710104, + "loss": 1.0041, + "step": 7498 + }, + { + "epoch": 1.33511396011396, + "grad_norm": 0.637856662273407, + "learning_rate": 0.00015004040177767896, + "loss": 0.9134, + "step": 7499 + }, + { + "epoch": 1.3352920227920229, + "grad_norm": 0.661227285861969, + "learning_rate": 0.00015002828238781912, + "loss": 1.0393, + "step": 7500 + }, + { + "epoch": 1.3354700854700854, + "grad_norm": 0.6061869859695435, + "learning_rate": 0.000150016162017759, + "loss": 0.8453, + "step": 7501 + }, + { + "epoch": 1.3356481481481481, + "grad_norm": 0.6938419938087463, + "learning_rate": 0.0001500040406677361, + "loss": 1.0338, + "step": 7502 + }, + { + "epoch": 1.335826210826211, + "grad_norm": 0.6672863960266113, + "learning_rate": 0.0001499919183379879, + "loss": 0.8765, + "step": 7503 + }, + { + "epoch": 1.3360042735042734, + "grad_norm": 0.6200515031814575, + "learning_rate": 0.00014997979502875193, + "loss": 0.8286, + "step": 7504 + }, + { + "epoch": 1.3361823361823362, + "grad_norm": 0.6287549138069153, + "learning_rate": 0.00014996767074026567, + "loss": 0.9761, + "step": 7505 + }, + { + "epoch": 1.336360398860399, + "grad_norm": 0.6036837100982666, + "learning_rate": 0.0001499555454727667, + "loss": 1.0506, + "step": 7506 + }, + { + "epoch": 1.3365384615384617, + "grad_norm": 0.6875260472297668, + "learning_rate": 0.0001499434192264926, + "loss": 1.001, + "step": 7507 + }, + { + "epoch": 1.3367165242165242, + "grad_norm": 0.6558469533920288, + "learning_rate": 0.00014993129200168096, + "loss": 0.6874, + "step": 7508 + }, + { + "epoch": 1.336894586894587, + "grad_norm": 0.604167103767395, + "learning_rate": 0.00014991916379856934, + "loss": 1.0173, + "step": 7509 + }, + { + "epoch": 1.3370726495726495, + "grad_norm": 0.5941442251205444, + "learning_rate": 0.00014990703461739544, + "loss": 0.8569, + "step": 7510 + }, + { + "epoch": 1.3372507122507122, + "grad_norm": 0.7645071148872375, + "learning_rate": 0.00014989490445839687, + "loss": 1.0172, + "step": 7511 + }, + { + "epoch": 1.337428774928775, + "grad_norm": 0.5491678714752197, + "learning_rate": 0.00014988277332181126, + "loss": 0.8018, + "step": 7512 + }, + { + "epoch": 1.3376068376068377, + "grad_norm": 0.583322286605835, + "learning_rate": 0.00014987064120787635, + "loss": 0.8704, + "step": 7513 + }, + { + "epoch": 1.3377849002849003, + "grad_norm": 0.7385724186897278, + "learning_rate": 0.00014985850811682984, + "loss": 1.1121, + "step": 7514 + }, + { + "epoch": 1.337962962962963, + "grad_norm": 0.6842585206031799, + "learning_rate": 0.00014984637404890941, + "loss": 0.914, + "step": 7515 + }, + { + "epoch": 1.3381410256410255, + "grad_norm": 0.6771186590194702, + "learning_rate": 0.00014983423900435285, + "loss": 1.0838, + "step": 7516 + }, + { + "epoch": 1.3383190883190883, + "grad_norm": 0.7562049031257629, + "learning_rate": 0.00014982210298339788, + "loss": 1.123, + "step": 7517 + }, + { + "epoch": 1.338497150997151, + "grad_norm": 0.7617804408073425, + "learning_rate": 0.0001498099659862823, + "loss": 0.9438, + "step": 7518 + }, + { + "epoch": 1.3386752136752138, + "grad_norm": 0.561958909034729, + "learning_rate": 0.00014979782801324392, + "loss": 0.8739, + "step": 7519 + }, + { + "epoch": 1.3388532763532763, + "grad_norm": 0.7726154923439026, + "learning_rate": 0.00014978568906452052, + "loss": 1.1306, + "step": 7520 + }, + { + "epoch": 1.339031339031339, + "grad_norm": 0.6658660173416138, + "learning_rate": 0.00014977354914035002, + "loss": 1.0214, + "step": 7521 + }, + { + "epoch": 1.3392094017094016, + "grad_norm": 0.6385402679443359, + "learning_rate": 0.00014976140824097015, + "loss": 0.8851, + "step": 7522 + }, + { + "epoch": 1.3393874643874644, + "grad_norm": 0.6315767168998718, + "learning_rate": 0.0001497492663666189, + "loss": 0.986, + "step": 7523 + }, + { + "epoch": 1.3395655270655271, + "grad_norm": 0.6379088759422302, + "learning_rate": 0.0001497371235175341, + "loss": 0.9322, + "step": 7524 + }, + { + "epoch": 1.3397435897435899, + "grad_norm": 0.6605859994888306, + "learning_rate": 0.0001497249796939537, + "loss": 1.1112, + "step": 7525 + }, + { + "epoch": 1.3399216524216524, + "grad_norm": 0.7342822551727295, + "learning_rate": 0.0001497128348961156, + "loss": 0.9798, + "step": 7526 + }, + { + "epoch": 1.3400997150997151, + "grad_norm": 0.5667192935943604, + "learning_rate": 0.0001497006891242578, + "loss": 0.7493, + "step": 7527 + }, + { + "epoch": 1.3402777777777777, + "grad_norm": 0.6106827855110168, + "learning_rate": 0.0001496885423786182, + "loss": 1.0924, + "step": 7528 + }, + { + "epoch": 1.3404558404558404, + "grad_norm": 0.6207202076911926, + "learning_rate": 0.00014967639465943486, + "loss": 1.1123, + "step": 7529 + }, + { + "epoch": 1.3406339031339032, + "grad_norm": 0.6272760033607483, + "learning_rate": 0.00014966424596694574, + "loss": 0.9275, + "step": 7530 + }, + { + "epoch": 1.340811965811966, + "grad_norm": 0.6485986113548279, + "learning_rate": 0.0001496520963013889, + "loss": 1.1491, + "step": 7531 + }, + { + "epoch": 1.3409900284900285, + "grad_norm": 0.5743561387062073, + "learning_rate": 0.00014963994566300238, + "loss": 1.1101, + "step": 7532 + }, + { + "epoch": 1.3411680911680912, + "grad_norm": 0.6508657336235046, + "learning_rate": 0.00014962779405202424, + "loss": 1.0368, + "step": 7533 + }, + { + "epoch": 1.3413461538461537, + "grad_norm": 0.6598748564720154, + "learning_rate": 0.00014961564146869259, + "loss": 1.1064, + "step": 7534 + }, + { + "epoch": 1.3415242165242165, + "grad_norm": 0.6722840070724487, + "learning_rate": 0.00014960348791324547, + "loss": 0.9758, + "step": 7535 + }, + { + "epoch": 1.3417022792022792, + "grad_norm": 0.5807220935821533, + "learning_rate": 0.00014959133338592108, + "loss": 0.9936, + "step": 7536 + }, + { + "epoch": 1.341880341880342, + "grad_norm": 0.6318647265434265, + "learning_rate": 0.00014957917788695752, + "loss": 0.907, + "step": 7537 + }, + { + "epoch": 1.3420584045584045, + "grad_norm": 0.6725485324859619, + "learning_rate": 0.00014956702141659295, + "loss": 0.988, + "step": 7538 + }, + { + "epoch": 1.3422364672364673, + "grad_norm": 0.6675217747688293, + "learning_rate": 0.0001495548639750656, + "loss": 1.0194, + "step": 7539 + }, + { + "epoch": 1.3424145299145298, + "grad_norm": 0.6976884603500366, + "learning_rate": 0.0001495427055626136, + "loss": 1.2515, + "step": 7540 + }, + { + "epoch": 1.3425925925925926, + "grad_norm": 0.654941737651825, + "learning_rate": 0.0001495305461794752, + "loss": 1.2072, + "step": 7541 + }, + { + "epoch": 1.3427706552706553, + "grad_norm": 0.7085291743278503, + "learning_rate": 0.00014951838582588864, + "loss": 0.9772, + "step": 7542 + }, + { + "epoch": 1.342948717948718, + "grad_norm": 0.6319566965103149, + "learning_rate": 0.00014950622450209217, + "loss": 1.0162, + "step": 7543 + }, + { + "epoch": 1.3431267806267806, + "grad_norm": 0.6272495985031128, + "learning_rate": 0.00014949406220832407, + "loss": 0.7985, + "step": 7544 + }, + { + "epoch": 1.3433048433048433, + "grad_norm": 0.6352069973945618, + "learning_rate": 0.00014948189894482266, + "loss": 1.0041, + "step": 7545 + }, + { + "epoch": 1.3434829059829059, + "grad_norm": 0.6071867346763611, + "learning_rate": 0.0001494697347118262, + "loss": 0.9486, + "step": 7546 + }, + { + "epoch": 1.3436609686609686, + "grad_norm": 0.6458829641342163, + "learning_rate": 0.00014945756950957308, + "loss": 0.9417, + "step": 7547 + }, + { + "epoch": 1.3438390313390314, + "grad_norm": 0.6472262740135193, + "learning_rate": 0.0001494454033383016, + "loss": 1.056, + "step": 7548 + }, + { + "epoch": 1.3440170940170941, + "grad_norm": 0.6985635161399841, + "learning_rate": 0.00014943323619825017, + "loss": 1.0483, + "step": 7549 + }, + { + "epoch": 1.3441951566951567, + "grad_norm": 0.6379460096359253, + "learning_rate": 0.00014942106808965718, + "loss": 0.9552, + "step": 7550 + }, + { + "epoch": 1.3443732193732194, + "grad_norm": 0.7036557793617249, + "learning_rate": 0.00014940889901276098, + "loss": 0.9647, + "step": 7551 + }, + { + "epoch": 1.344551282051282, + "grad_norm": 0.6697289943695068, + "learning_rate": 0.0001493967289678001, + "loss": 0.9029, + "step": 7552 + }, + { + "epoch": 1.3447293447293447, + "grad_norm": 0.6336250901222229, + "learning_rate": 0.00014938455795501286, + "loss": 0.9458, + "step": 7553 + }, + { + "epoch": 1.3449074074074074, + "grad_norm": 0.7279673218727112, + "learning_rate": 0.00014937238597463785, + "loss": 1.0228, + "step": 7554 + }, + { + "epoch": 1.3450854700854702, + "grad_norm": 0.6514406204223633, + "learning_rate": 0.00014936021302691349, + "loss": 0.8265, + "step": 7555 + }, + { + "epoch": 1.3452635327635327, + "grad_norm": 0.6405338644981384, + "learning_rate": 0.0001493480391120783, + "loss": 0.9516, + "step": 7556 + }, + { + "epoch": 1.3454415954415955, + "grad_norm": 0.6442672610282898, + "learning_rate": 0.00014933586423037076, + "loss": 0.9279, + "step": 7557 + }, + { + "epoch": 1.345619658119658, + "grad_norm": 0.7588633894920349, + "learning_rate": 0.00014932368838202945, + "loss": 1.0976, + "step": 7558 + }, + { + "epoch": 1.3457977207977208, + "grad_norm": 0.5536739230155945, + "learning_rate": 0.00014931151156729296, + "loss": 0.713, + "step": 7559 + }, + { + "epoch": 1.3459757834757835, + "grad_norm": 0.6897570490837097, + "learning_rate": 0.00014929933378639981, + "loss": 0.9521, + "step": 7560 + }, + { + "epoch": 1.3461538461538463, + "grad_norm": 0.6654927134513855, + "learning_rate": 0.00014928715503958863, + "loss": 0.8506, + "step": 7561 + }, + { + "epoch": 1.3463319088319088, + "grad_norm": 0.655806839466095, + "learning_rate": 0.00014927497532709808, + "loss": 0.8636, + "step": 7562 + }, + { + "epoch": 1.3465099715099715, + "grad_norm": 0.6547064185142517, + "learning_rate": 0.00014926279464916667, + "loss": 0.9155, + "step": 7563 + }, + { + "epoch": 1.346688034188034, + "grad_norm": 0.7555415034294128, + "learning_rate": 0.00014925061300603316, + "loss": 0.8791, + "step": 7564 + }, + { + "epoch": 1.3468660968660968, + "grad_norm": 0.7439392805099487, + "learning_rate": 0.0001492384303979362, + "loss": 1.1669, + "step": 7565 + }, + { + "epoch": 1.3470441595441596, + "grad_norm": 0.6016925573348999, + "learning_rate": 0.0001492262468251145, + "loss": 0.9811, + "step": 7566 + }, + { + "epoch": 1.3472222222222223, + "grad_norm": 0.644652783870697, + "learning_rate": 0.00014921406228780675, + "loss": 0.7096, + "step": 7567 + }, + { + "epoch": 1.3474002849002849, + "grad_norm": 0.721814751625061, + "learning_rate": 0.00014920187678625166, + "loss": 0.9933, + "step": 7568 + }, + { + "epoch": 1.3475783475783476, + "grad_norm": 0.6212092638015747, + "learning_rate": 0.000149189690320688, + "loss": 0.8499, + "step": 7569 + }, + { + "epoch": 1.3477564102564101, + "grad_norm": 0.6235958337783813, + "learning_rate": 0.00014917750289135455, + "loss": 0.9189, + "step": 7570 + }, + { + "epoch": 1.3479344729344729, + "grad_norm": 0.6309674978256226, + "learning_rate": 0.0001491653144984901, + "loss": 0.9744, + "step": 7571 + }, + { + "epoch": 1.3481125356125356, + "grad_norm": 0.7606496214866638, + "learning_rate": 0.00014915312514233344, + "loss": 1.0181, + "step": 7572 + }, + { + "epoch": 1.3482905982905984, + "grad_norm": 0.6892654895782471, + "learning_rate": 0.00014914093482312342, + "loss": 0.9517, + "step": 7573 + }, + { + "epoch": 1.348468660968661, + "grad_norm": 0.6746503114700317, + "learning_rate": 0.0001491287435410988, + "loss": 1.056, + "step": 7574 + }, + { + "epoch": 1.3486467236467237, + "grad_norm": 0.5892919301986694, + "learning_rate": 0.00014911655129649858, + "loss": 1.0515, + "step": 7575 + }, + { + "epoch": 1.3488247863247862, + "grad_norm": 0.6278096437454224, + "learning_rate": 0.0001491043580895615, + "loss": 0.864, + "step": 7576 + }, + { + "epoch": 1.349002849002849, + "grad_norm": 0.7017706632614136, + "learning_rate": 0.0001490921639205266, + "loss": 1.0618, + "step": 7577 + }, + { + "epoch": 1.3491809116809117, + "grad_norm": 0.7318746447563171, + "learning_rate": 0.00014907996878963268, + "loss": 0.9905, + "step": 7578 + }, + { + "epoch": 1.3493589743589745, + "grad_norm": 0.6485885977745056, + "learning_rate": 0.00014906777269711873, + "loss": 1.0498, + "step": 7579 + }, + { + "epoch": 1.349537037037037, + "grad_norm": 0.644902229309082, + "learning_rate": 0.00014905557564322372, + "loss": 0.885, + "step": 7580 + }, + { + "epoch": 1.3497150997150997, + "grad_norm": 0.6567610502243042, + "learning_rate": 0.0001490433776281866, + "loss": 0.8938, + "step": 7581 + }, + { + "epoch": 1.3498931623931623, + "grad_norm": 0.6233102679252625, + "learning_rate": 0.0001490311786522464, + "loss": 0.9007, + "step": 7582 + }, + { + "epoch": 1.350071225071225, + "grad_norm": 0.6962146759033203, + "learning_rate": 0.00014901897871564206, + "loss": 0.9257, + "step": 7583 + }, + { + "epoch": 1.3502492877492878, + "grad_norm": 0.6986933350563049, + "learning_rate": 0.00014900677781861266, + "loss": 1.0089, + "step": 7584 + }, + { + "epoch": 1.3504273504273505, + "grad_norm": 0.7527925968170166, + "learning_rate": 0.00014899457596139729, + "loss": 1.0762, + "step": 7585 + }, + { + "epoch": 1.350605413105413, + "grad_norm": 0.69191974401474, + "learning_rate": 0.00014898237314423494, + "loss": 0.9829, + "step": 7586 + }, + { + "epoch": 1.3507834757834758, + "grad_norm": 0.7866443395614624, + "learning_rate": 0.00014897016936736478, + "loss": 1.0911, + "step": 7587 + }, + { + "epoch": 1.3509615384615383, + "grad_norm": 0.7087522745132446, + "learning_rate": 0.00014895796463102587, + "loss": 1.0693, + "step": 7588 + }, + { + "epoch": 1.351139601139601, + "grad_norm": 0.704276442527771, + "learning_rate": 0.00014894575893545736, + "loss": 0.9082, + "step": 7589 + }, + { + "epoch": 1.3513176638176638, + "grad_norm": 0.7074487805366516, + "learning_rate": 0.00014893355228089833, + "loss": 0.8731, + "step": 7590 + }, + { + "epoch": 1.3514957264957266, + "grad_norm": 0.6542425155639648, + "learning_rate": 0.00014892134466758803, + "loss": 0.9325, + "step": 7591 + }, + { + "epoch": 1.351673789173789, + "grad_norm": 0.6577230095863342, + "learning_rate": 0.0001489091360957656, + "loss": 0.8468, + "step": 7592 + }, + { + "epoch": 1.3518518518518519, + "grad_norm": 0.638534426689148, + "learning_rate": 0.00014889692656567025, + "loss": 0.8598, + "step": 7593 + }, + { + "epoch": 1.3520299145299146, + "grad_norm": 0.751133918762207, + "learning_rate": 0.0001488847160775412, + "loss": 1.0006, + "step": 7594 + }, + { + "epoch": 1.3522079772079771, + "grad_norm": 0.6272708773612976, + "learning_rate": 0.00014887250463161767, + "loss": 0.8782, + "step": 7595 + }, + { + "epoch": 1.35238603988604, + "grad_norm": 0.7242439985275269, + "learning_rate": 0.00014886029222813897, + "loss": 1.2443, + "step": 7596 + }, + { + "epoch": 1.3525641025641026, + "grad_norm": 0.6199275851249695, + "learning_rate": 0.0001488480788673443, + "loss": 0.9211, + "step": 7597 + }, + { + "epoch": 1.3527421652421652, + "grad_norm": 0.6401306986808777, + "learning_rate": 0.00014883586454947305, + "loss": 0.8808, + "step": 7598 + }, + { + "epoch": 1.352920227920228, + "grad_norm": 0.6340938806533813, + "learning_rate": 0.00014882364927476443, + "loss": 0.9406, + "step": 7599 + }, + { + "epoch": 1.3530982905982907, + "grad_norm": 0.6388604044914246, + "learning_rate": 0.00014881143304345783, + "loss": 1.0674, + "step": 7600 + }, + { + "epoch": 1.3532763532763532, + "grad_norm": 0.7562061548233032, + "learning_rate": 0.00014879921585579263, + "loss": 1.0959, + "step": 7601 + }, + { + "epoch": 1.353454415954416, + "grad_norm": 0.6303606033325195, + "learning_rate": 0.00014878699771200815, + "loss": 0.9641, + "step": 7602 + }, + { + "epoch": 1.3536324786324787, + "grad_norm": 0.8623232841491699, + "learning_rate": 0.00014877477861234382, + "loss": 1.1529, + "step": 7603 + }, + { + "epoch": 1.3538105413105412, + "grad_norm": 0.6607624888420105, + "learning_rate": 0.00014876255855703896, + "loss": 0.6291, + "step": 7604 + }, + { + "epoch": 1.353988603988604, + "grad_norm": 0.6226931214332581, + "learning_rate": 0.0001487503375463331, + "loss": 0.7485, + "step": 7605 + }, + { + "epoch": 1.3541666666666667, + "grad_norm": 0.7626705169677734, + "learning_rate": 0.00014873811558046565, + "loss": 0.9694, + "step": 7606 + }, + { + "epoch": 1.3543447293447293, + "grad_norm": 0.5436057448387146, + "learning_rate": 0.00014872589265967605, + "loss": 0.6173, + "step": 7607 + }, + { + "epoch": 1.354522792022792, + "grad_norm": 0.7822177410125732, + "learning_rate": 0.00014871366878420382, + "loss": 1.0048, + "step": 7608 + }, + { + "epoch": 1.3547008547008548, + "grad_norm": 0.6955201625823975, + "learning_rate": 0.00014870144395428848, + "loss": 0.9487, + "step": 7609 + }, + { + "epoch": 1.3548789173789173, + "grad_norm": 0.6625505685806274, + "learning_rate": 0.00014868921817016943, + "loss": 0.9389, + "step": 7610 + }, + { + "epoch": 1.35505698005698, + "grad_norm": 0.6625354886054993, + "learning_rate": 0.00014867699143208634, + "loss": 0.9538, + "step": 7611 + }, + { + "epoch": 1.3552350427350428, + "grad_norm": 0.7426592707633972, + "learning_rate": 0.00014866476374027874, + "loss": 1.2566, + "step": 7612 + }, + { + "epoch": 1.3554131054131053, + "grad_norm": 0.6856544017791748, + "learning_rate": 0.00014865253509498616, + "loss": 0.9663, + "step": 7613 + }, + { + "epoch": 1.355591168091168, + "grad_norm": 0.6343915462493896, + "learning_rate": 0.00014864030549644825, + "loss": 0.9416, + "step": 7614 + }, + { + "epoch": 1.3557692307692308, + "grad_norm": 0.6319553256034851, + "learning_rate": 0.00014862807494490454, + "loss": 0.9335, + "step": 7615 + }, + { + "epoch": 1.3559472934472934, + "grad_norm": 0.6919772624969482, + "learning_rate": 0.00014861584344059476, + "loss": 0.8516, + "step": 7616 + }, + { + "epoch": 1.3561253561253561, + "grad_norm": 0.6405790448188782, + "learning_rate": 0.00014860361098375851, + "loss": 1.1278, + "step": 7617 + }, + { + "epoch": 1.3563034188034189, + "grad_norm": 0.7591732144355774, + "learning_rate": 0.00014859137757463548, + "loss": 1.0961, + "step": 7618 + }, + { + "epoch": 1.3564814814814814, + "grad_norm": 0.6166727542877197, + "learning_rate": 0.0001485791432134653, + "loss": 0.9358, + "step": 7619 + }, + { + "epoch": 1.3566595441595442, + "grad_norm": 0.7068707346916199, + "learning_rate": 0.00014856690790048777, + "loss": 0.8325, + "step": 7620 + }, + { + "epoch": 1.356837606837607, + "grad_norm": 0.8465402722358704, + "learning_rate": 0.00014855467163594257, + "loss": 1.0047, + "step": 7621 + }, + { + "epoch": 1.3570156695156697, + "grad_norm": 0.7403460741043091, + "learning_rate": 0.00014854243442006943, + "loss": 1.0907, + "step": 7622 + }, + { + "epoch": 1.3571937321937322, + "grad_norm": 0.6939566135406494, + "learning_rate": 0.00014853019625310813, + "loss": 0.9156, + "step": 7623 + }, + { + "epoch": 1.357371794871795, + "grad_norm": 0.6425924897193909, + "learning_rate": 0.0001485179571352984, + "loss": 0.8156, + "step": 7624 + }, + { + "epoch": 1.3575498575498575, + "grad_norm": 0.7091902494430542, + "learning_rate": 0.00014850571706688013, + "loss": 1.0483, + "step": 7625 + }, + { + "epoch": 1.3577279202279202, + "grad_norm": 0.663342297077179, + "learning_rate": 0.00014849347604809312, + "loss": 1.0405, + "step": 7626 + }, + { + "epoch": 1.357905982905983, + "grad_norm": 0.6727671027183533, + "learning_rate": 0.00014848123407917716, + "loss": 1.0389, + "step": 7627 + }, + { + "epoch": 1.3580840455840457, + "grad_norm": 0.6572692394256592, + "learning_rate": 0.0001484689911603721, + "loss": 1.0489, + "step": 7628 + }, + { + "epoch": 1.3582621082621082, + "grad_norm": 0.7629066109657288, + "learning_rate": 0.0001484567472919179, + "loss": 1.0372, + "step": 7629 + }, + { + "epoch": 1.358440170940171, + "grad_norm": 0.7848913669586182, + "learning_rate": 0.00014844450247405435, + "loss": 0.9437, + "step": 7630 + }, + { + "epoch": 1.3586182336182335, + "grad_norm": 0.715949535369873, + "learning_rate": 0.00014843225670702143, + "loss": 1.1949, + "step": 7631 + }, + { + "epoch": 1.3587962962962963, + "grad_norm": 0.6498245596885681, + "learning_rate": 0.00014842000999105905, + "loss": 0.8845, + "step": 7632 + }, + { + "epoch": 1.358974358974359, + "grad_norm": 0.7251074910163879, + "learning_rate": 0.00014840776232640716, + "loss": 1.093, + "step": 7633 + }, + { + "epoch": 1.3591524216524218, + "grad_norm": 0.6223580837249756, + "learning_rate": 0.0001483955137133057, + "loss": 1.0344, + "step": 7634 + }, + { + "epoch": 1.3593304843304843, + "grad_norm": 0.6504943370819092, + "learning_rate": 0.00014838326415199472, + "loss": 1.109, + "step": 7635 + }, + { + "epoch": 1.359508547008547, + "grad_norm": 0.5912374258041382, + "learning_rate": 0.00014837101364271416, + "loss": 1.0756, + "step": 7636 + }, + { + "epoch": 1.3596866096866096, + "grad_norm": 0.6116467714309692, + "learning_rate": 0.00014835876218570408, + "loss": 0.7871, + "step": 7637 + }, + { + "epoch": 1.3598646723646723, + "grad_norm": 0.7013412117958069, + "learning_rate": 0.0001483465097812045, + "loss": 1.0003, + "step": 7638 + }, + { + "epoch": 1.360042735042735, + "grad_norm": 0.5930750370025635, + "learning_rate": 0.00014833425642945552, + "loss": 0.9926, + "step": 7639 + }, + { + "epoch": 1.3602207977207978, + "grad_norm": 0.732955276966095, + "learning_rate": 0.00014832200213069717, + "loss": 1.2801, + "step": 7640 + }, + { + "epoch": 1.3603988603988604, + "grad_norm": 0.6836149096488953, + "learning_rate": 0.00014830974688516958, + "loss": 0.9292, + "step": 7641 + }, + { + "epoch": 1.3605769230769231, + "grad_norm": 0.6531919836997986, + "learning_rate": 0.00014829749069311283, + "loss": 0.9551, + "step": 7642 + }, + { + "epoch": 1.3607549857549857, + "grad_norm": 0.719093382358551, + "learning_rate": 0.0001482852335547671, + "loss": 0.8588, + "step": 7643 + }, + { + "epoch": 1.3609330484330484, + "grad_norm": 0.6144105792045593, + "learning_rate": 0.00014827297547037252, + "loss": 0.9033, + "step": 7644 + }, + { + "epoch": 1.3611111111111112, + "grad_norm": 0.789241373538971, + "learning_rate": 0.00014826071644016926, + "loss": 1.1916, + "step": 7645 + }, + { + "epoch": 1.361289173789174, + "grad_norm": 0.6137418746948242, + "learning_rate": 0.0001482484564643975, + "loss": 0.9648, + "step": 7646 + }, + { + "epoch": 1.3614672364672364, + "grad_norm": 0.6789261698722839, + "learning_rate": 0.00014823619554329745, + "loss": 0.829, + "step": 7647 + }, + { + "epoch": 1.3616452991452992, + "grad_norm": 0.6508790254592896, + "learning_rate": 0.0001482239336771094, + "loss": 0.942, + "step": 7648 + }, + { + "epoch": 1.3618233618233617, + "grad_norm": 0.6725571751594543, + "learning_rate": 0.00014821167086607353, + "loss": 0.8884, + "step": 7649 + }, + { + "epoch": 1.3620014245014245, + "grad_norm": 0.6252003908157349, + "learning_rate": 0.00014819940711043012, + "loss": 0.9778, + "step": 7650 + }, + { + "epoch": 1.3621794871794872, + "grad_norm": 0.6950626969337463, + "learning_rate": 0.00014818714241041943, + "loss": 1.2104, + "step": 7651 + }, + { + "epoch": 1.36235754985755, + "grad_norm": 0.6527379155158997, + "learning_rate": 0.0001481748767662818, + "loss": 0.7845, + "step": 7652 + }, + { + "epoch": 1.3625356125356125, + "grad_norm": 0.7438235282897949, + "learning_rate": 0.00014816261017825755, + "loss": 0.9513, + "step": 7653 + }, + { + "epoch": 1.3627136752136753, + "grad_norm": 0.6412696838378906, + "learning_rate": 0.000148150342646587, + "loss": 0.8478, + "step": 7654 + }, + { + "epoch": 1.3628917378917378, + "grad_norm": 0.658481240272522, + "learning_rate": 0.00014813807417151046, + "loss": 0.6816, + "step": 7655 + }, + { + "epoch": 1.3630698005698005, + "grad_norm": 0.6170126795768738, + "learning_rate": 0.0001481258047532684, + "loss": 0.8862, + "step": 7656 + }, + { + "epoch": 1.3632478632478633, + "grad_norm": 0.7049173712730408, + "learning_rate": 0.0001481135343921012, + "loss": 1.0027, + "step": 7657 + }, + { + "epoch": 1.363425925925926, + "grad_norm": 0.7780741453170776, + "learning_rate": 0.0001481012630882492, + "loss": 1.0183, + "step": 7658 + }, + { + "epoch": 1.3636039886039886, + "grad_norm": 0.6658362746238708, + "learning_rate": 0.00014808899084195286, + "loss": 0.878, + "step": 7659 + }, + { + "epoch": 1.3637820512820513, + "grad_norm": 0.7192076444625854, + "learning_rate": 0.00014807671765345267, + "loss": 1.2269, + "step": 7660 + }, + { + "epoch": 1.3639601139601139, + "grad_norm": 0.7038660049438477, + "learning_rate": 0.00014806444352298903, + "loss": 0.889, + "step": 7661 + }, + { + "epoch": 1.3641381766381766, + "grad_norm": 0.622803270816803, + "learning_rate": 0.00014805216845080249, + "loss": 0.9623, + "step": 7662 + }, + { + "epoch": 1.3643162393162394, + "grad_norm": 0.9157076478004456, + "learning_rate": 0.00014803989243713353, + "loss": 1.106, + "step": 7663 + }, + { + "epoch": 1.364494301994302, + "grad_norm": 0.6369999647140503, + "learning_rate": 0.00014802761548222268, + "loss": 0.9755, + "step": 7664 + }, + { + "epoch": 1.3646723646723646, + "grad_norm": 0.8318394422531128, + "learning_rate": 0.00014801533758631045, + "loss": 1.1786, + "step": 7665 + }, + { + "epoch": 1.3648504273504274, + "grad_norm": 0.7065796852111816, + "learning_rate": 0.00014800305874963744, + "loss": 1.2066, + "step": 7666 + }, + { + "epoch": 1.36502849002849, + "grad_norm": 0.6570265293121338, + "learning_rate": 0.0001479907789724442, + "loss": 1.0084, + "step": 7667 + }, + { + "epoch": 1.3652065527065527, + "grad_norm": 0.637321949005127, + "learning_rate": 0.00014797849825497135, + "loss": 0.9075, + "step": 7668 + }, + { + "epoch": 1.3653846153846154, + "grad_norm": 0.7656470537185669, + "learning_rate": 0.00014796621659745948, + "loss": 1.1497, + "step": 7669 + }, + { + "epoch": 1.3655626780626782, + "grad_norm": 0.6798120737075806, + "learning_rate": 0.0001479539340001493, + "loss": 0.8154, + "step": 7670 + }, + { + "epoch": 1.3657407407407407, + "grad_norm": 0.7004328966140747, + "learning_rate": 0.0001479416504632813, + "loss": 1.0513, + "step": 7671 + }, + { + "epoch": 1.3659188034188035, + "grad_norm": 0.6551713943481445, + "learning_rate": 0.0001479293659870963, + "loss": 0.8735, + "step": 7672 + }, + { + "epoch": 1.366096866096866, + "grad_norm": 0.7685719132423401, + "learning_rate": 0.00014791708057183494, + "loss": 1.111, + "step": 7673 + }, + { + "epoch": 1.3662749287749287, + "grad_norm": 0.673624575138092, + "learning_rate": 0.0001479047942177379, + "loss": 0.9418, + "step": 7674 + }, + { + "epoch": 1.3664529914529915, + "grad_norm": 0.6281047463417053, + "learning_rate": 0.00014789250692504597, + "loss": 1.0938, + "step": 7675 + }, + { + "epoch": 1.3666310541310542, + "grad_norm": 0.5846312642097473, + "learning_rate": 0.0001478802186939998, + "loss": 0.6352, + "step": 7676 + }, + { + "epoch": 1.3668091168091168, + "grad_norm": 0.7037251591682434, + "learning_rate": 0.00014786792952484025, + "loss": 1.1775, + "step": 7677 + }, + { + "epoch": 1.3669871794871795, + "grad_norm": 0.69822758436203, + "learning_rate": 0.00014785563941780808, + "loss": 1.0877, + "step": 7678 + }, + { + "epoch": 1.367165242165242, + "grad_norm": 0.7229313254356384, + "learning_rate": 0.000147843348373144, + "loss": 1.0305, + "step": 7679 + }, + { + "epoch": 1.3673433048433048, + "grad_norm": 0.665771484375, + "learning_rate": 0.00014783105639108897, + "loss": 0.9056, + "step": 7680 + }, + { + "epoch": 1.3675213675213675, + "grad_norm": 0.6418357491493225, + "learning_rate": 0.00014781876347188367, + "loss": 0.9374, + "step": 7681 + }, + { + "epoch": 1.3676994301994303, + "grad_norm": 0.7255483269691467, + "learning_rate": 0.0001478064696157691, + "loss": 0.8533, + "step": 7682 + }, + { + "epoch": 1.3678774928774928, + "grad_norm": 0.668064534664154, + "learning_rate": 0.00014779417482298603, + "loss": 0.9002, + "step": 7683 + }, + { + "epoch": 1.3680555555555556, + "grad_norm": 0.6797603368759155, + "learning_rate": 0.0001477818790937754, + "loss": 0.9733, + "step": 7684 + }, + { + "epoch": 1.368233618233618, + "grad_norm": 0.6905350685119629, + "learning_rate": 0.0001477695824283781, + "loss": 0.7985, + "step": 7685 + }, + { + "epoch": 1.3684116809116809, + "grad_norm": 0.6846137046813965, + "learning_rate": 0.00014775728482703507, + "loss": 0.9154, + "step": 7686 + }, + { + "epoch": 1.3685897435897436, + "grad_norm": 0.6686832904815674, + "learning_rate": 0.00014774498628998726, + "loss": 0.926, + "step": 7687 + }, + { + "epoch": 1.3687678062678064, + "grad_norm": 0.7050234079360962, + "learning_rate": 0.00014773268681747561, + "loss": 0.9386, + "step": 7688 + }, + { + "epoch": 1.368945868945869, + "grad_norm": 0.7048354744911194, + "learning_rate": 0.00014772038640974112, + "loss": 1.1483, + "step": 7689 + }, + { + "epoch": 1.3691239316239316, + "grad_norm": 0.698192298412323, + "learning_rate": 0.0001477080850670248, + "loss": 1.1452, + "step": 7690 + }, + { + "epoch": 1.3693019943019942, + "grad_norm": 0.6838962435722351, + "learning_rate": 0.00014769578278956766, + "loss": 0.9789, + "step": 7691 + }, + { + "epoch": 1.369480056980057, + "grad_norm": 0.6636955142021179, + "learning_rate": 0.00014768347957761074, + "loss": 0.931, + "step": 7692 + }, + { + "epoch": 1.3696581196581197, + "grad_norm": 0.706030547618866, + "learning_rate": 0.0001476711754313951, + "loss": 1.1096, + "step": 7693 + }, + { + "epoch": 1.3698361823361824, + "grad_norm": 0.6771288514137268, + "learning_rate": 0.00014765887035116178, + "loss": 0.9641, + "step": 7694 + }, + { + "epoch": 1.370014245014245, + "grad_norm": 0.6805008053779602, + "learning_rate": 0.00014764656433715188, + "loss": 0.8724, + "step": 7695 + }, + { + "epoch": 1.3701923076923077, + "grad_norm": 0.6599233746528625, + "learning_rate": 0.00014763425738960657, + "loss": 0.8477, + "step": 7696 + }, + { + "epoch": 1.3703703703703702, + "grad_norm": 0.7036116123199463, + "learning_rate": 0.0001476219495087669, + "loss": 1.0991, + "step": 7697 + }, + { + "epoch": 1.370548433048433, + "grad_norm": 0.6677989363670349, + "learning_rate": 0.0001476096406948741, + "loss": 1.2397, + "step": 7698 + }, + { + "epoch": 1.3707264957264957, + "grad_norm": 0.5652269721031189, + "learning_rate": 0.00014759733094816928, + "loss": 0.9302, + "step": 7699 + }, + { + "epoch": 1.3709045584045585, + "grad_norm": 0.6670156121253967, + "learning_rate": 0.00014758502026889362, + "loss": 0.8362, + "step": 7700 + }, + { + "epoch": 1.371082621082621, + "grad_norm": 0.6705406904220581, + "learning_rate": 0.00014757270865728832, + "loss": 0.876, + "step": 7701 + }, + { + "epoch": 1.3712606837606838, + "grad_norm": 0.6020053625106812, + "learning_rate": 0.00014756039611359465, + "loss": 0.9182, + "step": 7702 + }, + { + "epoch": 1.3714387464387463, + "grad_norm": 0.6370134949684143, + "learning_rate": 0.0001475480826380538, + "loss": 1.1063, + "step": 7703 + }, + { + "epoch": 1.371616809116809, + "grad_norm": 0.6906460523605347, + "learning_rate": 0.00014753576823090705, + "loss": 0.988, + "step": 7704 + }, + { + "epoch": 1.3717948717948718, + "grad_norm": 0.6047569513320923, + "learning_rate": 0.00014752345289239567, + "loss": 1.15, + "step": 7705 + }, + { + "epoch": 1.3719729344729346, + "grad_norm": 0.7019868493080139, + "learning_rate": 0.00014751113662276095, + "loss": 1.1185, + "step": 7706 + }, + { + "epoch": 1.372150997150997, + "grad_norm": 0.6534035801887512, + "learning_rate": 0.00014749881942224417, + "loss": 0.9006, + "step": 7707 + }, + { + "epoch": 1.3723290598290598, + "grad_norm": 0.6111651659011841, + "learning_rate": 0.00014748650129108674, + "loss": 0.935, + "step": 7708 + }, + { + "epoch": 1.3725071225071226, + "grad_norm": 0.6678512096405029, + "learning_rate": 0.00014747418222952995, + "loss": 0.8771, + "step": 7709 + }, + { + "epoch": 1.3726851851851851, + "grad_norm": 0.607829749584198, + "learning_rate": 0.00014746186223781518, + "loss": 1.0509, + "step": 7710 + }, + { + "epoch": 1.3728632478632479, + "grad_norm": 0.7274412512779236, + "learning_rate": 0.00014744954131618382, + "loss": 0.9545, + "step": 7711 + }, + { + "epoch": 1.3730413105413106, + "grad_norm": 0.640333354473114, + "learning_rate": 0.00014743721946487723, + "loss": 1.018, + "step": 7712 + }, + { + "epoch": 1.3732193732193732, + "grad_norm": 0.6772079467773438, + "learning_rate": 0.0001474248966841369, + "loss": 1.0983, + "step": 7713 + }, + { + "epoch": 1.373397435897436, + "grad_norm": 0.49630534648895264, + "learning_rate": 0.00014741257297420422, + "loss": 0.5238, + "step": 7714 + }, + { + "epoch": 1.3735754985754987, + "grad_norm": 0.6316596269607544, + "learning_rate": 0.00014740024833532068, + "loss": 1.1342, + "step": 7715 + }, + { + "epoch": 1.3737535612535612, + "grad_norm": 0.5928404331207275, + "learning_rate": 0.00014738792276772775, + "loss": 0.7987, + "step": 7716 + }, + { + "epoch": 1.373931623931624, + "grad_norm": 0.6773418188095093, + "learning_rate": 0.00014737559627166688, + "loss": 0.934, + "step": 7717 + }, + { + "epoch": 1.3741096866096867, + "grad_norm": 0.7895028591156006, + "learning_rate": 0.00014736326884737963, + "loss": 0.984, + "step": 7718 + }, + { + "epoch": 1.3742877492877492, + "grad_norm": 0.7074753046035767, + "learning_rate": 0.00014735094049510752, + "loss": 1.0093, + "step": 7719 + }, + { + "epoch": 1.374465811965812, + "grad_norm": 0.5389847159385681, + "learning_rate": 0.00014733861121509208, + "loss": 0.8138, + "step": 7720 + }, + { + "epoch": 1.3746438746438747, + "grad_norm": 0.6138495206832886, + "learning_rate": 0.00014732628100757493, + "loss": 0.9282, + "step": 7721 + }, + { + "epoch": 1.3748219373219372, + "grad_norm": 0.7609560489654541, + "learning_rate": 0.00014731394987279757, + "loss": 0.9859, + "step": 7722 + }, + { + "epoch": 1.375, + "grad_norm": 0.6806198954582214, + "learning_rate": 0.00014730161781100165, + "loss": 0.8932, + "step": 7723 + }, + { + "epoch": 1.3751780626780628, + "grad_norm": 0.7229103446006775, + "learning_rate": 0.0001472892848224288, + "loss": 0.956, + "step": 7724 + }, + { + "epoch": 1.3753561253561253, + "grad_norm": 0.6157994866371155, + "learning_rate": 0.00014727695090732066, + "loss": 1.0285, + "step": 7725 + }, + { + "epoch": 1.375534188034188, + "grad_norm": 0.5885980129241943, + "learning_rate": 0.00014726461606591885, + "loss": 0.9174, + "step": 7726 + }, + { + "epoch": 1.3757122507122508, + "grad_norm": 0.6655769944190979, + "learning_rate": 0.0001472522802984651, + "loss": 0.9059, + "step": 7727 + }, + { + "epoch": 1.3758903133903133, + "grad_norm": 0.7075541019439697, + "learning_rate": 0.00014723994360520105, + "loss": 1.0055, + "step": 7728 + }, + { + "epoch": 1.376068376068376, + "grad_norm": 0.6947159171104431, + "learning_rate": 0.00014722760598636847, + "loss": 0.9782, + "step": 7729 + }, + { + "epoch": 1.3762464387464388, + "grad_norm": 0.6629964709281921, + "learning_rate": 0.00014721526744220905, + "loss": 0.9427, + "step": 7730 + }, + { + "epoch": 1.3764245014245013, + "grad_norm": 0.7385284304618835, + "learning_rate": 0.00014720292797296453, + "loss": 0.9953, + "step": 7731 + }, + { + "epoch": 1.376602564102564, + "grad_norm": 0.6123563051223755, + "learning_rate": 0.0001471905875788767, + "loss": 1.0103, + "step": 7732 + }, + { + "epoch": 1.3767806267806268, + "grad_norm": 0.6457047462463379, + "learning_rate": 0.00014717824626018732, + "loss": 0.9779, + "step": 7733 + }, + { + "epoch": 1.3769586894586894, + "grad_norm": 0.6196442246437073, + "learning_rate": 0.00014716590401713824, + "loss": 0.8747, + "step": 7734 + }, + { + "epoch": 1.3771367521367521, + "grad_norm": 0.7932298183441162, + "learning_rate": 0.00014715356084997122, + "loss": 1.1617, + "step": 7735 + }, + { + "epoch": 1.3773148148148149, + "grad_norm": 0.787304699420929, + "learning_rate": 0.00014714121675892815, + "loss": 1.1383, + "step": 7736 + }, + { + "epoch": 1.3774928774928774, + "grad_norm": 0.672795295715332, + "learning_rate": 0.00014712887174425085, + "loss": 1.2563, + "step": 7737 + }, + { + "epoch": 1.3776709401709402, + "grad_norm": 0.6505744457244873, + "learning_rate": 0.00014711652580618123, + "loss": 0.9194, + "step": 7738 + }, + { + "epoch": 1.377849002849003, + "grad_norm": 0.8141193985939026, + "learning_rate": 0.00014710417894496115, + "loss": 1.1428, + "step": 7739 + }, + { + "epoch": 1.3780270655270654, + "grad_norm": 0.6269707679748535, + "learning_rate": 0.00014709183116083253, + "loss": 0.7164, + "step": 7740 + }, + { + "epoch": 1.3782051282051282, + "grad_norm": 0.6737076640129089, + "learning_rate": 0.0001470794824540373, + "loss": 0.9965, + "step": 7741 + }, + { + "epoch": 1.378383190883191, + "grad_norm": 0.6451728343963623, + "learning_rate": 0.0001470671328248174, + "loss": 1.0539, + "step": 7742 + }, + { + "epoch": 1.3785612535612537, + "grad_norm": 0.6480295062065125, + "learning_rate": 0.00014705478227341486, + "loss": 0.9118, + "step": 7743 + }, + { + "epoch": 1.3787393162393162, + "grad_norm": 0.7429090738296509, + "learning_rate": 0.00014704243080007154, + "loss": 1.0031, + "step": 7744 + }, + { + "epoch": 1.378917378917379, + "grad_norm": 0.5601376891136169, + "learning_rate": 0.00014703007840502955, + "loss": 0.849, + "step": 7745 + }, + { + "epoch": 1.3790954415954415, + "grad_norm": 0.7067657113075256, + "learning_rate": 0.00014701772508853088, + "loss": 1.3067, + "step": 7746 + }, + { + "epoch": 1.3792735042735043, + "grad_norm": 0.7016390562057495, + "learning_rate": 0.00014700537085081755, + "loss": 1.0236, + "step": 7747 + }, + { + "epoch": 1.379451566951567, + "grad_norm": 0.6505000591278076, + "learning_rate": 0.0001469930156921316, + "loss": 1.0121, + "step": 7748 + }, + { + "epoch": 1.3796296296296298, + "grad_norm": 0.8515380620956421, + "learning_rate": 0.00014698065961271512, + "loss": 1.0413, + "step": 7749 + }, + { + "epoch": 1.3798076923076923, + "grad_norm": 0.6322008371353149, + "learning_rate": 0.00014696830261281025, + "loss": 0.8306, + "step": 7750 + }, + { + "epoch": 1.379985754985755, + "grad_norm": 0.7090431451797485, + "learning_rate": 0.00014695594469265902, + "loss": 1.1829, + "step": 7751 + }, + { + "epoch": 1.3801638176638176, + "grad_norm": 0.5913167595863342, + "learning_rate": 0.00014694358585250363, + "loss": 0.9769, + "step": 7752 + }, + { + "epoch": 1.3803418803418803, + "grad_norm": 0.7345432639122009, + "learning_rate": 0.00014693122609258616, + "loss": 0.9928, + "step": 7753 + }, + { + "epoch": 1.380519943019943, + "grad_norm": 0.6158214211463928, + "learning_rate": 0.00014691886541314884, + "loss": 1.1166, + "step": 7754 + }, + { + "epoch": 1.3806980056980058, + "grad_norm": 0.6874041557312012, + "learning_rate": 0.0001469065038144338, + "loss": 1.0808, + "step": 7755 + }, + { + "epoch": 1.3808760683760684, + "grad_norm": 0.8135195970535278, + "learning_rate": 0.00014689414129668326, + "loss": 0.9482, + "step": 7756 + }, + { + "epoch": 1.381054131054131, + "grad_norm": 0.6389174461364746, + "learning_rate": 0.00014688177786013944, + "loss": 1.039, + "step": 7757 + }, + { + "epoch": 1.3812321937321936, + "grad_norm": 0.6953016519546509, + "learning_rate": 0.00014686941350504454, + "loss": 0.9426, + "step": 7758 + }, + { + "epoch": 1.3814102564102564, + "grad_norm": 0.8171859383583069, + "learning_rate": 0.00014685704823164087, + "loss": 1.0393, + "step": 7759 + }, + { + "epoch": 1.3815883190883191, + "grad_norm": 0.6968414783477783, + "learning_rate": 0.0001468446820401707, + "loss": 1.1167, + "step": 7760 + }, + { + "epoch": 1.381766381766382, + "grad_norm": 0.6916623711585999, + "learning_rate": 0.00014683231493087628, + "loss": 1.1886, + "step": 7761 + }, + { + "epoch": 1.3819444444444444, + "grad_norm": 0.7351683378219604, + "learning_rate": 0.00014681994690399992, + "loss": 0.9893, + "step": 7762 + }, + { + "epoch": 1.3821225071225072, + "grad_norm": 0.6617491245269775, + "learning_rate": 0.00014680757795978395, + "loss": 1.0505, + "step": 7763 + }, + { + "epoch": 1.3823005698005697, + "grad_norm": 0.6627485156059265, + "learning_rate": 0.00014679520809847074, + "loss": 0.9878, + "step": 7764 + }, + { + "epoch": 1.3824786324786325, + "grad_norm": 0.704636812210083, + "learning_rate": 0.00014678283732030264, + "loss": 0.8332, + "step": 7765 + }, + { + "epoch": 1.3826566951566952, + "grad_norm": 0.698853075504303, + "learning_rate": 0.00014677046562552203, + "loss": 1.0926, + "step": 7766 + }, + { + "epoch": 1.382834757834758, + "grad_norm": 0.6695869565010071, + "learning_rate": 0.0001467580930143713, + "loss": 1.0626, + "step": 7767 + }, + { + "epoch": 1.3830128205128205, + "grad_norm": 0.672173023223877, + "learning_rate": 0.00014674571948709286, + "loss": 0.8842, + "step": 7768 + }, + { + "epoch": 1.3831908831908832, + "grad_norm": 0.6735473871231079, + "learning_rate": 0.00014673334504392916, + "loss": 0.9382, + "step": 7769 + }, + { + "epoch": 1.3833689458689458, + "grad_norm": 0.6864013075828552, + "learning_rate": 0.00014672096968512265, + "loss": 1.1369, + "step": 7770 + }, + { + "epoch": 1.3835470085470085, + "grad_norm": 0.7154954075813293, + "learning_rate": 0.0001467085934109158, + "loss": 1.1447, + "step": 7771 + }, + { + "epoch": 1.3837250712250713, + "grad_norm": 0.5934487581253052, + "learning_rate": 0.0001466962162215511, + "loss": 0.8923, + "step": 7772 + }, + { + "epoch": 1.383903133903134, + "grad_norm": 0.8116832971572876, + "learning_rate": 0.00014668383811727097, + "loss": 1.0997, + "step": 7773 + }, + { + "epoch": 1.3840811965811965, + "grad_norm": 0.8661674857139587, + "learning_rate": 0.00014667145909831808, + "loss": 1.0112, + "step": 7774 + }, + { + "epoch": 1.3842592592592593, + "grad_norm": 0.5173856616020203, + "learning_rate": 0.00014665907916493488, + "loss": 0.6571, + "step": 7775 + }, + { + "epoch": 1.3844373219373218, + "grad_norm": 0.6165067553520203, + "learning_rate": 0.00014664669831736395, + "loss": 1.0992, + "step": 7776 + }, + { + "epoch": 1.3846153846153846, + "grad_norm": 0.6564429998397827, + "learning_rate": 0.00014663431655584787, + "loss": 0.9103, + "step": 7777 + }, + { + "epoch": 1.3847934472934473, + "grad_norm": 0.7162124514579773, + "learning_rate": 0.00014662193388062923, + "loss": 1.0645, + "step": 7778 + }, + { + "epoch": 1.38497150997151, + "grad_norm": 0.6391215920448303, + "learning_rate": 0.00014660955029195064, + "loss": 0.902, + "step": 7779 + }, + { + "epoch": 1.3851495726495726, + "grad_norm": 0.6876635551452637, + "learning_rate": 0.00014659716579005475, + "loss": 1.0924, + "step": 7780 + }, + { + "epoch": 1.3853276353276354, + "grad_norm": 0.7254653573036194, + "learning_rate": 0.00014658478037518418, + "loss": 1.0135, + "step": 7781 + }, + { + "epoch": 1.385505698005698, + "grad_norm": 0.6900535225868225, + "learning_rate": 0.00014657239404758162, + "loss": 0.983, + "step": 7782 + }, + { + "epoch": 1.3856837606837606, + "grad_norm": 0.7477042078971863, + "learning_rate": 0.00014656000680748975, + "loss": 1.0707, + "step": 7783 + }, + { + "epoch": 1.3858618233618234, + "grad_norm": 0.5756927132606506, + "learning_rate": 0.00014654761865515124, + "loss": 0.8881, + "step": 7784 + }, + { + "epoch": 1.3860398860398861, + "grad_norm": 0.6736083626747131, + "learning_rate": 0.00014653522959080884, + "loss": 1.0193, + "step": 7785 + }, + { + "epoch": 1.3862179487179487, + "grad_norm": 0.616179883480072, + "learning_rate": 0.0001465228396147053, + "loss": 0.8676, + "step": 7786 + }, + { + "epoch": 1.3863960113960114, + "grad_norm": 0.7956456542015076, + "learning_rate": 0.00014651044872708338, + "loss": 0.9787, + "step": 7787 + }, + { + "epoch": 1.386574074074074, + "grad_norm": 0.6613463163375854, + "learning_rate": 0.00014649805692818578, + "loss": 1.0032, + "step": 7788 + }, + { + "epoch": 1.3867521367521367, + "grad_norm": 0.6215800642967224, + "learning_rate": 0.0001464856642182554, + "loss": 1.0123, + "step": 7789 + }, + { + "epoch": 1.3869301994301995, + "grad_norm": 0.6701171398162842, + "learning_rate": 0.00014647327059753496, + "loss": 0.9108, + "step": 7790 + }, + { + "epoch": 1.3871082621082622, + "grad_norm": 0.6213465929031372, + "learning_rate": 0.00014646087606626736, + "loss": 0.9313, + "step": 7791 + }, + { + "epoch": 1.3872863247863247, + "grad_norm": 0.7535304427146912, + "learning_rate": 0.00014644848062469535, + "loss": 1.0813, + "step": 7792 + }, + { + "epoch": 1.3874643874643875, + "grad_norm": 0.6778230667114258, + "learning_rate": 0.0001464360842730619, + "loss": 1.0405, + "step": 7793 + }, + { + "epoch": 1.38764245014245, + "grad_norm": 0.7816025614738464, + "learning_rate": 0.0001464236870116098, + "loss": 0.9228, + "step": 7794 + }, + { + "epoch": 1.3878205128205128, + "grad_norm": 0.6815229058265686, + "learning_rate": 0.00014641128884058203, + "loss": 0.9607, + "step": 7795 + }, + { + "epoch": 1.3879985754985755, + "grad_norm": 0.7027714848518372, + "learning_rate": 0.00014639888976022145, + "loss": 0.9379, + "step": 7796 + }, + { + "epoch": 1.3881766381766383, + "grad_norm": 0.7636353373527527, + "learning_rate": 0.00014638648977077104, + "loss": 1.1186, + "step": 7797 + }, + { + "epoch": 1.3883547008547008, + "grad_norm": 0.6732974052429199, + "learning_rate": 0.00014637408887247365, + "loss": 1.1378, + "step": 7798 + }, + { + "epoch": 1.3885327635327636, + "grad_norm": 0.7539397478103638, + "learning_rate": 0.0001463616870655724, + "loss": 0.999, + "step": 7799 + }, + { + "epoch": 1.388710826210826, + "grad_norm": 0.6872972846031189, + "learning_rate": 0.00014634928435031013, + "loss": 0.9564, + "step": 7800 + }, + { + "epoch": 1.3888888888888888, + "grad_norm": 0.6823115348815918, + "learning_rate": 0.00014633688072693, + "loss": 0.9745, + "step": 7801 + }, + { + "epoch": 1.3890669515669516, + "grad_norm": 0.6462571620941162, + "learning_rate": 0.00014632447619567488, + "loss": 0.8314, + "step": 7802 + }, + { + "epoch": 1.3892450142450143, + "grad_norm": 0.7245402932167053, + "learning_rate": 0.0001463120707567879, + "loss": 0.8291, + "step": 7803 + }, + { + "epoch": 1.3894230769230769, + "grad_norm": 0.697179913520813, + "learning_rate": 0.00014629966441051208, + "loss": 1.017, + "step": 7804 + }, + { + "epoch": 1.3896011396011396, + "grad_norm": 0.6304250359535217, + "learning_rate": 0.00014628725715709053, + "loss": 0.9262, + "step": 7805 + }, + { + "epoch": 1.3897792022792022, + "grad_norm": 0.5780240297317505, + "learning_rate": 0.00014627484899676634, + "loss": 0.6596, + "step": 7806 + }, + { + "epoch": 1.389957264957265, + "grad_norm": 0.8030684590339661, + "learning_rate": 0.0001462624399297826, + "loss": 0.9977, + "step": 7807 + }, + { + "epoch": 1.3901353276353277, + "grad_norm": 0.7999774813652039, + "learning_rate": 0.00014625002995638246, + "loss": 1.1036, + "step": 7808 + }, + { + "epoch": 1.3903133903133904, + "grad_norm": 0.7054862976074219, + "learning_rate": 0.00014623761907680904, + "loss": 1.1435, + "step": 7809 + }, + { + "epoch": 1.390491452991453, + "grad_norm": 0.6660647392272949, + "learning_rate": 0.00014622520729130556, + "loss": 0.703, + "step": 7810 + }, + { + "epoch": 1.3906695156695157, + "grad_norm": 0.6339690089225769, + "learning_rate": 0.00014621279460011515, + "loss": 1.0451, + "step": 7811 + }, + { + "epoch": 1.3908475783475782, + "grad_norm": 0.8568736910820007, + "learning_rate": 0.00014620038100348102, + "loss": 1.009, + "step": 7812 + }, + { + "epoch": 1.391025641025641, + "grad_norm": 0.7126797437667847, + "learning_rate": 0.00014618796650164642, + "loss": 0.9592, + "step": 7813 + }, + { + "epoch": 1.3912037037037037, + "grad_norm": 0.6768994331359863, + "learning_rate": 0.00014617555109485453, + "loss": 1.09, + "step": 7814 + }, + { + "epoch": 1.3913817663817665, + "grad_norm": 0.7609471678733826, + "learning_rate": 0.00014616313478334864, + "loss": 0.9781, + "step": 7815 + }, + { + "epoch": 1.391559829059829, + "grad_norm": 0.7107006907463074, + "learning_rate": 0.00014615071756737203, + "loss": 0.9769, + "step": 7816 + }, + { + "epoch": 1.3917378917378918, + "grad_norm": 0.6324763894081116, + "learning_rate": 0.00014613829944716802, + "loss": 1.089, + "step": 7817 + }, + { + "epoch": 1.3919159544159543, + "grad_norm": 0.6617186069488525, + "learning_rate": 0.00014612588042297984, + "loss": 1.0466, + "step": 7818 + }, + { + "epoch": 1.392094017094017, + "grad_norm": 0.7881436944007874, + "learning_rate": 0.00014611346049505083, + "loss": 1.003, + "step": 7819 + }, + { + "epoch": 1.3922720797720798, + "grad_norm": 0.7391049861907959, + "learning_rate": 0.00014610103966362437, + "loss": 1.0531, + "step": 7820 + }, + { + "epoch": 1.3924501424501425, + "grad_norm": 0.6299472451210022, + "learning_rate": 0.00014608861792894383, + "loss": 0.8433, + "step": 7821 + }, + { + "epoch": 1.392628205128205, + "grad_norm": 0.6053452491760254, + "learning_rate": 0.00014607619529125255, + "loss": 0.7945, + "step": 7822 + }, + { + "epoch": 1.3928062678062678, + "grad_norm": 0.7160114645957947, + "learning_rate": 0.0001460637717507939, + "loss": 1.1604, + "step": 7823 + }, + { + "epoch": 1.3929843304843303, + "grad_norm": 0.6308854222297668, + "learning_rate": 0.00014605134730781135, + "loss": 1.0918, + "step": 7824 + }, + { + "epoch": 1.393162393162393, + "grad_norm": 0.7187000513076782, + "learning_rate": 0.00014603892196254833, + "loss": 1.0594, + "step": 7825 + }, + { + "epoch": 1.3933404558404558, + "grad_norm": 0.7516581416130066, + "learning_rate": 0.00014602649571524826, + "loss": 0.9222, + "step": 7826 + }, + { + "epoch": 1.3935185185185186, + "grad_norm": 0.6340481638908386, + "learning_rate": 0.00014601406856615463, + "loss": 0.8131, + "step": 7827 + }, + { + "epoch": 1.3936965811965811, + "grad_norm": 0.8161744475364685, + "learning_rate": 0.0001460016405155109, + "loss": 0.8695, + "step": 7828 + }, + { + "epoch": 1.3938746438746439, + "grad_norm": 0.6926971077919006, + "learning_rate": 0.0001459892115635606, + "loss": 0.9548, + "step": 7829 + }, + { + "epoch": 1.3940527065527066, + "grad_norm": 0.6669796109199524, + "learning_rate": 0.0001459767817105472, + "loss": 0.9255, + "step": 7830 + }, + { + "epoch": 1.3942307692307692, + "grad_norm": 0.6626184582710266, + "learning_rate": 0.00014596435095671432, + "loss": 1.1141, + "step": 7831 + }, + { + "epoch": 1.394408831908832, + "grad_norm": 0.6755738854408264, + "learning_rate": 0.00014595191930230546, + "loss": 0.9596, + "step": 7832 + }, + { + "epoch": 1.3945868945868947, + "grad_norm": 0.6034863591194153, + "learning_rate": 0.00014593948674756417, + "loss": 0.8088, + "step": 7833 + }, + { + "epoch": 1.3947649572649572, + "grad_norm": 0.5638226866722107, + "learning_rate": 0.00014592705329273406, + "loss": 0.5828, + "step": 7834 + }, + { + "epoch": 1.39494301994302, + "grad_norm": 0.6902222633361816, + "learning_rate": 0.0001459146189380588, + "loss": 0.7954, + "step": 7835 + }, + { + "epoch": 1.3951210826210827, + "grad_norm": 0.7579947710037231, + "learning_rate": 0.0001459021836837819, + "loss": 1.1301, + "step": 7836 + }, + { + "epoch": 1.3952991452991452, + "grad_norm": 0.6894911527633667, + "learning_rate": 0.00014588974753014712, + "loss": 1.082, + "step": 7837 + }, + { + "epoch": 1.395477207977208, + "grad_norm": 0.6330230832099915, + "learning_rate": 0.000145877310477398, + "loss": 0.7614, + "step": 7838 + }, + { + "epoch": 1.3956552706552707, + "grad_norm": 0.6164960265159607, + "learning_rate": 0.00014586487252577832, + "loss": 0.8981, + "step": 7839 + }, + { + "epoch": 1.3958333333333333, + "grad_norm": 0.6575061678886414, + "learning_rate": 0.0001458524336755317, + "loss": 0.9735, + "step": 7840 + }, + { + "epoch": 1.396011396011396, + "grad_norm": 0.687921941280365, + "learning_rate": 0.00014583999392690195, + "loss": 0.9207, + "step": 7841 + }, + { + "epoch": 1.3961894586894588, + "grad_norm": 0.6175212860107422, + "learning_rate": 0.00014582755328013274, + "loss": 1.0444, + "step": 7842 + }, + { + "epoch": 1.3963675213675213, + "grad_norm": 0.6351733207702637, + "learning_rate": 0.00014581511173546781, + "loss": 1.0143, + "step": 7843 + }, + { + "epoch": 1.396545584045584, + "grad_norm": 0.7235051989555359, + "learning_rate": 0.00014580266929315093, + "loss": 0.9108, + "step": 7844 + }, + { + "epoch": 1.3967236467236468, + "grad_norm": 0.6432043313980103, + "learning_rate": 0.00014579022595342586, + "loss": 0.8674, + "step": 7845 + }, + { + "epoch": 1.3969017094017093, + "grad_norm": 0.7775412797927856, + "learning_rate": 0.00014577778171653648, + "loss": 1.0637, + "step": 7846 + }, + { + "epoch": 1.397079772079772, + "grad_norm": 0.6748763918876648, + "learning_rate": 0.00014576533658272655, + "loss": 1.0356, + "step": 7847 + }, + { + "epoch": 1.3972578347578348, + "grad_norm": 0.6940401196479797, + "learning_rate": 0.00014575289055223994, + "loss": 0.9937, + "step": 7848 + }, + { + "epoch": 1.3974358974358974, + "grad_norm": 0.6971304416656494, + "learning_rate": 0.00014574044362532045, + "loss": 0.9753, + "step": 7849 + }, + { + "epoch": 1.39761396011396, + "grad_norm": 0.6576017141342163, + "learning_rate": 0.00014572799580221197, + "loss": 1.1233, + "step": 7850 + }, + { + "epoch": 1.3977920227920229, + "grad_norm": 0.6270702481269836, + "learning_rate": 0.00014571554708315843, + "loss": 0.9771, + "step": 7851 + }, + { + "epoch": 1.3979700854700854, + "grad_norm": 0.6898425817489624, + "learning_rate": 0.00014570309746840372, + "loss": 0.9235, + "step": 7852 + }, + { + "epoch": 1.3981481481481481, + "grad_norm": 0.7017102241516113, + "learning_rate": 0.00014569064695819174, + "loss": 1.1056, + "step": 7853 + }, + { + "epoch": 1.398326210826211, + "grad_norm": 0.6298288702964783, + "learning_rate": 0.00014567819555276647, + "loss": 0.8635, + "step": 7854 + }, + { + "epoch": 1.3985042735042734, + "grad_norm": 0.7173134684562683, + "learning_rate": 0.00014566574325237182, + "loss": 1.0893, + "step": 7855 + }, + { + "epoch": 1.3986823361823362, + "grad_norm": 0.7541036605834961, + "learning_rate": 0.0001456532900572518, + "loss": 1.0996, + "step": 7856 + }, + { + "epoch": 1.398860398860399, + "grad_norm": 0.6204771399497986, + "learning_rate": 0.0001456408359676504, + "loss": 0.7601, + "step": 7857 + }, + { + "epoch": 1.3990384615384617, + "grad_norm": 0.629557192325592, + "learning_rate": 0.00014562838098381163, + "loss": 0.9239, + "step": 7858 + }, + { + "epoch": 1.3992165242165242, + "grad_norm": 0.6878390908241272, + "learning_rate": 0.00014561592510597954, + "loss": 0.9641, + "step": 7859 + }, + { + "epoch": 1.399394586894587, + "grad_norm": 0.7490049004554749, + "learning_rate": 0.00014560346833439813, + "loss": 1.0198, + "step": 7860 + }, + { + "epoch": 1.3995726495726495, + "grad_norm": 0.6337960958480835, + "learning_rate": 0.0001455910106693115, + "loss": 0.8709, + "step": 7861 + }, + { + "epoch": 1.3997507122507122, + "grad_norm": 0.6210524439811707, + "learning_rate": 0.0001455785521109637, + "loss": 1.1049, + "step": 7862 + }, + { + "epoch": 1.399928774928775, + "grad_norm": 0.7894936203956604, + "learning_rate": 0.00014556609265959887, + "loss": 0.8933, + "step": 7863 + }, + { + "epoch": 1.4001068376068377, + "grad_norm": 0.6888098120689392, + "learning_rate": 0.00014555363231546112, + "loss": 0.9738, + "step": 7864 + }, + { + "epoch": 1.4002849002849003, + "grad_norm": 0.608799934387207, + "learning_rate": 0.00014554117107879456, + "loss": 0.9103, + "step": 7865 + }, + { + "epoch": 1.400462962962963, + "grad_norm": 0.7390474081039429, + "learning_rate": 0.00014552870894984335, + "loss": 1.2484, + "step": 7866 + }, + { + "epoch": 1.4006410256410255, + "grad_norm": 0.6513381600379944, + "learning_rate": 0.00014551624592885169, + "loss": 0.8523, + "step": 7867 + }, + { + "epoch": 1.4008190883190883, + "grad_norm": 0.6357464790344238, + "learning_rate": 0.00014550378201606373, + "loss": 0.9594, + "step": 7868 + }, + { + "epoch": 1.400997150997151, + "grad_norm": 0.6893286108970642, + "learning_rate": 0.0001454913172117237, + "loss": 0.9798, + "step": 7869 + }, + { + "epoch": 1.4011752136752138, + "grad_norm": 0.6566550731658936, + "learning_rate": 0.0001454788515160758, + "loss": 1.0532, + "step": 7870 + }, + { + "epoch": 1.4013532763532763, + "grad_norm": 0.6442158222198486, + "learning_rate": 0.00014546638492936425, + "loss": 1.0789, + "step": 7871 + }, + { + "epoch": 1.401531339031339, + "grad_norm": 0.7570971846580505, + "learning_rate": 0.0001454539174518334, + "loss": 0.9806, + "step": 7872 + }, + { + "epoch": 1.4017094017094016, + "grad_norm": 0.6180047392845154, + "learning_rate": 0.0001454414490837274, + "loss": 0.857, + "step": 7873 + }, + { + "epoch": 1.4018874643874644, + "grad_norm": 0.7143170237541199, + "learning_rate": 0.0001454289798252906, + "loss": 0.8815, + "step": 7874 + }, + { + "epoch": 1.4020655270655271, + "grad_norm": 0.6388922929763794, + "learning_rate": 0.00014541650967676736, + "loss": 0.95, + "step": 7875 + }, + { + "epoch": 1.4022435897435899, + "grad_norm": 0.7137351632118225, + "learning_rate": 0.00014540403863840193, + "loss": 0.8973, + "step": 7876 + }, + { + "epoch": 1.4024216524216524, + "grad_norm": 0.656315267086029, + "learning_rate": 0.0001453915667104387, + "loss": 1.149, + "step": 7877 + }, + { + "epoch": 1.4025997150997151, + "grad_norm": 0.7234711647033691, + "learning_rate": 0.000145379093893122, + "loss": 0.9798, + "step": 7878 + }, + { + "epoch": 1.4027777777777777, + "grad_norm": 0.6595289707183838, + "learning_rate": 0.00014536662018669623, + "loss": 1.2704, + "step": 7879 + }, + { + "epoch": 1.4029558404558404, + "grad_norm": 0.6760551333427429, + "learning_rate": 0.00014535414559140576, + "loss": 0.8672, + "step": 7880 + }, + { + "epoch": 1.4031339031339032, + "grad_norm": 0.5916706919670105, + "learning_rate": 0.000145341670107495, + "loss": 0.888, + "step": 7881 + }, + { + "epoch": 1.403311965811966, + "grad_norm": 0.7272133231163025, + "learning_rate": 0.00014532919373520846, + "loss": 1.0466, + "step": 7882 + }, + { + "epoch": 1.4034900284900285, + "grad_norm": 0.8512467741966248, + "learning_rate": 0.00014531671647479048, + "loss": 1.2482, + "step": 7883 + }, + { + "epoch": 1.4036680911680912, + "grad_norm": 0.5536492466926575, + "learning_rate": 0.0001453042383264856, + "loss": 0.7823, + "step": 7884 + }, + { + "epoch": 1.4038461538461537, + "grad_norm": 0.7262215614318848, + "learning_rate": 0.0001452917592905383, + "loss": 0.9713, + "step": 7885 + }, + { + "epoch": 1.4040242165242165, + "grad_norm": 0.7146059274673462, + "learning_rate": 0.00014527927936719304, + "loss": 1.1064, + "step": 7886 + }, + { + "epoch": 1.4042022792022792, + "grad_norm": 0.5915318131446838, + "learning_rate": 0.00014526679855669436, + "loss": 0.8567, + "step": 7887 + }, + { + "epoch": 1.404380341880342, + "grad_norm": 0.6548298001289368, + "learning_rate": 0.00014525431685928682, + "loss": 1.1359, + "step": 7888 + }, + { + "epoch": 1.4045584045584045, + "grad_norm": 0.7482563853263855, + "learning_rate": 0.0001452418342752149, + "loss": 0.9095, + "step": 7889 + }, + { + "epoch": 1.4047364672364673, + "grad_norm": 0.6660130023956299, + "learning_rate": 0.0001452293508047233, + "loss": 1.2343, + "step": 7890 + }, + { + "epoch": 1.4049145299145298, + "grad_norm": 0.7457148432731628, + "learning_rate": 0.00014521686644805644, + "loss": 1.2086, + "step": 7891 + }, + { + "epoch": 1.4050925925925926, + "grad_norm": 0.5957929491996765, + "learning_rate": 0.00014520438120545906, + "loss": 0.9724, + "step": 7892 + }, + { + "epoch": 1.4052706552706553, + "grad_norm": 0.6832270622253418, + "learning_rate": 0.00014519189507717573, + "loss": 0.9903, + "step": 7893 + }, + { + "epoch": 1.405448717948718, + "grad_norm": 0.6202489733695984, + "learning_rate": 0.00014517940806345109, + "loss": 0.962, + "step": 7894 + }, + { + "epoch": 1.4056267806267806, + "grad_norm": 0.6419472694396973, + "learning_rate": 0.0001451669201645298, + "loss": 0.8147, + "step": 7895 + }, + { + "epoch": 1.4058048433048433, + "grad_norm": 0.61143958568573, + "learning_rate": 0.00014515443138065652, + "loss": 0.8674, + "step": 7896 + }, + { + "epoch": 1.4059829059829059, + "grad_norm": 0.7527356743812561, + "learning_rate": 0.00014514194171207597, + "loss": 1.0581, + "step": 7897 + }, + { + "epoch": 1.4061609686609686, + "grad_norm": 0.7195194363594055, + "learning_rate": 0.00014512945115903285, + "loss": 1.0268, + "step": 7898 + }, + { + "epoch": 1.4063390313390314, + "grad_norm": 0.7919661998748779, + "learning_rate": 0.00014511695972177187, + "loss": 1.0259, + "step": 7899 + }, + { + "epoch": 1.4065170940170941, + "grad_norm": 0.6774758696556091, + "learning_rate": 0.00014510446740053783, + "loss": 1.1214, + "step": 7900 + }, + { + "epoch": 1.4066951566951567, + "grad_norm": 0.6102406978607178, + "learning_rate": 0.0001450919741955754, + "loss": 1.1846, + "step": 7901 + }, + { + "epoch": 1.4068732193732194, + "grad_norm": 0.7189443707466125, + "learning_rate": 0.00014507948010712942, + "loss": 0.7758, + "step": 7902 + }, + { + "epoch": 1.407051282051282, + "grad_norm": 0.654153048992157, + "learning_rate": 0.00014506698513544467, + "loss": 0.899, + "step": 7903 + }, + { + "epoch": 1.4072293447293447, + "grad_norm": 0.637934684753418, + "learning_rate": 0.00014505448928076598, + "loss": 0.8301, + "step": 7904 + }, + { + "epoch": 1.4074074074074074, + "grad_norm": 0.7504615783691406, + "learning_rate": 0.00014504199254333812, + "loss": 0.9883, + "step": 7905 + }, + { + "epoch": 1.4075854700854702, + "grad_norm": 0.7902522683143616, + "learning_rate": 0.00014502949492340602, + "loss": 0.9615, + "step": 7906 + }, + { + "epoch": 1.4077635327635327, + "grad_norm": 0.5832732319831848, + "learning_rate": 0.0001450169964212145, + "loss": 0.7136, + "step": 7907 + }, + { + "epoch": 1.4079415954415955, + "grad_norm": 0.6025400757789612, + "learning_rate": 0.00014500449703700846, + "loss": 0.8812, + "step": 7908 + }, + { + "epoch": 1.408119658119658, + "grad_norm": 0.6412411332130432, + "learning_rate": 0.0001449919967710328, + "loss": 0.9346, + "step": 7909 + }, + { + "epoch": 1.4082977207977208, + "grad_norm": 0.7546970844268799, + "learning_rate": 0.00014497949562353242, + "loss": 1.0794, + "step": 7910 + }, + { + "epoch": 1.4084757834757835, + "grad_norm": 0.6175593733787537, + "learning_rate": 0.00014496699359475222, + "loss": 0.8939, + "step": 7911 + }, + { + "epoch": 1.4086538461538463, + "grad_norm": 0.6571716666221619, + "learning_rate": 0.00014495449068493722, + "loss": 1.1003, + "step": 7912 + }, + { + "epoch": 1.4088319088319088, + "grad_norm": 0.7038990259170532, + "learning_rate": 0.00014494198689433236, + "loss": 0.8844, + "step": 7913 + }, + { + "epoch": 1.4090099715099715, + "grad_norm": 0.7007337212562561, + "learning_rate": 0.00014492948222318263, + "loss": 1.2038, + "step": 7914 + }, + { + "epoch": 1.409188034188034, + "grad_norm": 0.7318591475486755, + "learning_rate": 0.00014491697667173302, + "loss": 1.0388, + "step": 7915 + }, + { + "epoch": 1.4093660968660968, + "grad_norm": 0.7010329961776733, + "learning_rate": 0.00014490447024022855, + "loss": 1.1485, + "step": 7916 + }, + { + "epoch": 1.4095441595441596, + "grad_norm": 0.7844831347465515, + "learning_rate": 0.0001448919629289143, + "loss": 1.1417, + "step": 7917 + }, + { + "epoch": 1.4097222222222223, + "grad_norm": 0.6953392624855042, + "learning_rate": 0.00014487945473803525, + "loss": 0.9546, + "step": 7918 + }, + { + "epoch": 1.4099002849002849, + "grad_norm": 0.6307587623596191, + "learning_rate": 0.00014486694566783655, + "loss": 0.9912, + "step": 7919 + }, + { + "epoch": 1.4100783475783476, + "grad_norm": 0.6200215816497803, + "learning_rate": 0.00014485443571856326, + "loss": 1.0998, + "step": 7920 + }, + { + "epoch": 1.4102564102564101, + "grad_norm": 0.7096502184867859, + "learning_rate": 0.00014484192489046043, + "loss": 0.9587, + "step": 7921 + }, + { + "epoch": 1.4104344729344729, + "grad_norm": 0.6965526342391968, + "learning_rate": 0.00014482941318377327, + "loss": 0.8791, + "step": 7922 + }, + { + "epoch": 1.4106125356125356, + "grad_norm": 0.7303466200828552, + "learning_rate": 0.00014481690059874687, + "loss": 1.084, + "step": 7923 + }, + { + "epoch": 1.4107905982905984, + "grad_norm": 0.6144066452980042, + "learning_rate": 0.00014480438713562638, + "loss": 0.9646, + "step": 7924 + }, + { + "epoch": 1.410968660968661, + "grad_norm": 0.645222008228302, + "learning_rate": 0.00014479187279465704, + "loss": 0.728, + "step": 7925 + }, + { + "epoch": 1.4111467236467237, + "grad_norm": 0.6069912314414978, + "learning_rate": 0.000144779357576084, + "loss": 0.842, + "step": 7926 + }, + { + "epoch": 1.4113247863247862, + "grad_norm": 0.6212135553359985, + "learning_rate": 0.00014476684148015243, + "loss": 0.9817, + "step": 7927 + }, + { + "epoch": 1.411502849002849, + "grad_norm": 0.6893343329429626, + "learning_rate": 0.00014475432450710763, + "loss": 1.0265, + "step": 7928 + }, + { + "epoch": 1.4116809116809117, + "grad_norm": 0.6842793822288513, + "learning_rate": 0.00014474180665719478, + "loss": 1.0593, + "step": 7929 + }, + { + "epoch": 1.4118589743589745, + "grad_norm": 0.74690842628479, + "learning_rate": 0.0001447292879306592, + "loss": 0.9096, + "step": 7930 + }, + { + "epoch": 1.412037037037037, + "grad_norm": 0.6624761819839478, + "learning_rate": 0.00014471676832774613, + "loss": 1.2244, + "step": 7931 + }, + { + "epoch": 1.4122150997150997, + "grad_norm": 0.6205778121948242, + "learning_rate": 0.00014470424784870088, + "loss": 1.1, + "step": 7932 + }, + { + "epoch": 1.4123931623931623, + "grad_norm": 0.7592337131500244, + "learning_rate": 0.00014469172649376875, + "loss": 0.963, + "step": 7933 + }, + { + "epoch": 1.412571225071225, + "grad_norm": 0.673328697681427, + "learning_rate": 0.00014467920426319508, + "loss": 0.8923, + "step": 7934 + }, + { + "epoch": 1.4127492877492878, + "grad_norm": 0.6064394116401672, + "learning_rate": 0.00014466668115722522, + "loss": 0.9679, + "step": 7935 + }, + { + "epoch": 1.4129273504273505, + "grad_norm": 0.7738677859306335, + "learning_rate": 0.00014465415717610454, + "loss": 1.0678, + "step": 7936 + }, + { + "epoch": 1.413105413105413, + "grad_norm": 0.7013397812843323, + "learning_rate": 0.00014464163232007836, + "loss": 0.9017, + "step": 7937 + }, + { + "epoch": 1.4132834757834758, + "grad_norm": 0.713291347026825, + "learning_rate": 0.0001446291065893922, + "loss": 1.1953, + "step": 7938 + }, + { + "epoch": 1.4134615384615383, + "grad_norm": 0.7538655996322632, + "learning_rate": 0.00014461657998429136, + "loss": 1.0571, + "step": 7939 + }, + { + "epoch": 1.413639601139601, + "grad_norm": 0.6358973383903503, + "learning_rate": 0.00014460405250502133, + "loss": 0.8552, + "step": 7940 + }, + { + "epoch": 1.4138176638176638, + "grad_norm": 0.67508864402771, + "learning_rate": 0.00014459152415182756, + "loss": 1.0293, + "step": 7941 + }, + { + "epoch": 1.4139957264957266, + "grad_norm": 0.7074598670005798, + "learning_rate": 0.00014457899492495546, + "loss": 1.2102, + "step": 7942 + }, + { + "epoch": 1.414173789173789, + "grad_norm": 0.7157037854194641, + "learning_rate": 0.00014456646482465058, + "loss": 1.0566, + "step": 7943 + }, + { + "epoch": 1.4143518518518519, + "grad_norm": 0.7918477058410645, + "learning_rate": 0.00014455393385115844, + "loss": 1.3727, + "step": 7944 + }, + { + "epoch": 1.4145299145299146, + "grad_norm": 0.569144606590271, + "learning_rate": 0.0001445414020047245, + "loss": 0.7251, + "step": 7945 + }, + { + "epoch": 1.4147079772079771, + "grad_norm": 0.7589054107666016, + "learning_rate": 0.0001445288692855943, + "loss": 1.0155, + "step": 7946 + }, + { + "epoch": 1.41488603988604, + "grad_norm": 0.7531685829162598, + "learning_rate": 0.0001445163356940134, + "loss": 0.8404, + "step": 7947 + }, + { + "epoch": 1.4150641025641026, + "grad_norm": 0.5730917453765869, + "learning_rate": 0.0001445038012302274, + "loss": 0.8215, + "step": 7948 + }, + { + "epoch": 1.4152421652421652, + "grad_norm": 0.6960710883140564, + "learning_rate": 0.00014449126589448187, + "loss": 0.7902, + "step": 7949 + }, + { + "epoch": 1.415420227920228, + "grad_norm": 0.8207054138183594, + "learning_rate": 0.0001444787296870224, + "loss": 1.493, + "step": 7950 + }, + { + "epoch": 1.4155982905982907, + "grad_norm": 0.5854668617248535, + "learning_rate": 0.00014446619260809462, + "loss": 0.9262, + "step": 7951 + }, + { + "epoch": 1.4157763532763532, + "grad_norm": 0.5458414554595947, + "learning_rate": 0.00014445365465794413, + "loss": 0.8431, + "step": 7952 + }, + { + "epoch": 1.415954415954416, + "grad_norm": 0.6880569458007812, + "learning_rate": 0.00014444111583681666, + "loss": 1.0184, + "step": 7953 + }, + { + "epoch": 1.4161324786324787, + "grad_norm": 0.6391083598136902, + "learning_rate": 0.00014442857614495783, + "loss": 0.88, + "step": 7954 + }, + { + "epoch": 1.4163105413105412, + "grad_norm": 0.6246135234832764, + "learning_rate": 0.00014441603558261335, + "loss": 0.776, + "step": 7955 + }, + { + "epoch": 1.416488603988604, + "grad_norm": 0.6263493895530701, + "learning_rate": 0.00014440349415002893, + "loss": 0.9069, + "step": 7956 + }, + { + "epoch": 1.4166666666666667, + "grad_norm": 0.7123475670814514, + "learning_rate": 0.00014439095184745024, + "loss": 0.8339, + "step": 7957 + }, + { + "epoch": 1.4168447293447293, + "grad_norm": 0.7171050906181335, + "learning_rate": 0.00014437840867512309, + "loss": 1.0633, + "step": 7958 + }, + { + "epoch": 1.417022792022792, + "grad_norm": 0.7097769975662231, + "learning_rate": 0.00014436586463329322, + "loss": 1.0852, + "step": 7959 + }, + { + "epoch": 1.4172008547008548, + "grad_norm": 0.6889223456382751, + "learning_rate": 0.00014435331972220637, + "loss": 0.916, + "step": 7960 + }, + { + "epoch": 1.4173789173789173, + "grad_norm": 0.6674435138702393, + "learning_rate": 0.0001443407739421084, + "loss": 0.9307, + "step": 7961 + }, + { + "epoch": 1.41755698005698, + "grad_norm": 0.6578894853591919, + "learning_rate": 0.00014432822729324503, + "loss": 0.8767, + "step": 7962 + }, + { + "epoch": 1.4177350427350428, + "grad_norm": 0.7145379781723022, + "learning_rate": 0.00014431567977586212, + "loss": 0.9962, + "step": 7963 + }, + { + "epoch": 1.4179131054131053, + "grad_norm": 0.6916680335998535, + "learning_rate": 0.00014430313139020555, + "loss": 1.0464, + "step": 7964 + }, + { + "epoch": 1.418091168091168, + "grad_norm": 0.6296181678771973, + "learning_rate": 0.00014429058213652116, + "loss": 1.0699, + "step": 7965 + }, + { + "epoch": 1.4182692307692308, + "grad_norm": 0.5640227198600769, + "learning_rate": 0.00014427803201505482, + "loss": 0.7006, + "step": 7966 + }, + { + "epoch": 1.4184472934472934, + "grad_norm": 0.7181212306022644, + "learning_rate": 0.0001442654810260524, + "loss": 1.1648, + "step": 7967 + }, + { + "epoch": 1.4186253561253561, + "grad_norm": 0.6830772757530212, + "learning_rate": 0.00014425292916975984, + "loss": 1.0641, + "step": 7968 + }, + { + "epoch": 1.4188034188034189, + "grad_norm": 0.665716290473938, + "learning_rate": 0.00014424037644642307, + "loss": 0.8769, + "step": 7969 + }, + { + "epoch": 1.4189814814814814, + "grad_norm": 0.8088666796684265, + "learning_rate": 0.00014422782285628802, + "loss": 1.1496, + "step": 7970 + }, + { + "epoch": 1.4191595441595442, + "grad_norm": 0.7186072468757629, + "learning_rate": 0.00014421526839960064, + "loss": 0.7421, + "step": 7971 + }, + { + "epoch": 1.419337606837607, + "grad_norm": 0.6405926942825317, + "learning_rate": 0.00014420271307660694, + "loss": 1.0139, + "step": 7972 + }, + { + "epoch": 1.4195156695156697, + "grad_norm": 0.7097104787826538, + "learning_rate": 0.0001441901568875529, + "loss": 1.1582, + "step": 7973 + }, + { + "epoch": 1.4196937321937322, + "grad_norm": 0.7347947359085083, + "learning_rate": 0.00014417759983268452, + "loss": 0.9751, + "step": 7974 + }, + { + "epoch": 1.419871794871795, + "grad_norm": 0.6999621987342834, + "learning_rate": 0.00014416504191224787, + "loss": 0.9419, + "step": 7975 + }, + { + "epoch": 1.4200498575498575, + "grad_norm": 0.6500616073608398, + "learning_rate": 0.00014415248312648897, + "loss": 0.9407, + "step": 7976 + }, + { + "epoch": 1.4202279202279202, + "grad_norm": 0.6368781328201294, + "learning_rate": 0.00014413992347565383, + "loss": 1.1224, + "step": 7977 + }, + { + "epoch": 1.420405982905983, + "grad_norm": 0.6422648429870605, + "learning_rate": 0.00014412736295998864, + "loss": 0.9573, + "step": 7978 + }, + { + "epoch": 1.4205840455840457, + "grad_norm": 0.744057297706604, + "learning_rate": 0.00014411480157973942, + "loss": 1.1384, + "step": 7979 + }, + { + "epoch": 1.4207621082621082, + "grad_norm": 0.5905839204788208, + "learning_rate": 0.00014410223933515232, + "loss": 0.8212, + "step": 7980 + }, + { + "epoch": 1.420940170940171, + "grad_norm": 0.5905438661575317, + "learning_rate": 0.0001440896762264734, + "loss": 0.8281, + "step": 7981 + }, + { + "epoch": 1.4211182336182335, + "grad_norm": 0.7087140679359436, + "learning_rate": 0.00014407711225394892, + "loss": 1.0165, + "step": 7982 + }, + { + "epoch": 1.4212962962962963, + "grad_norm": 0.6173902153968811, + "learning_rate": 0.00014406454741782495, + "loss": 0.8823, + "step": 7983 + }, + { + "epoch": 1.421474358974359, + "grad_norm": 0.6649761199951172, + "learning_rate": 0.00014405198171834772, + "loss": 0.9489, + "step": 7984 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 0.619286835193634, + "learning_rate": 0.00014403941515576344, + "loss": 0.8149, + "step": 7985 + }, + { + "epoch": 1.4218304843304843, + "grad_norm": 0.6358469724655151, + "learning_rate": 0.0001440268477303183, + "loss": 1.0558, + "step": 7986 + }, + { + "epoch": 1.422008547008547, + "grad_norm": 0.7239769697189331, + "learning_rate": 0.0001440142794422585, + "loss": 1.0528, + "step": 7987 + }, + { + "epoch": 1.4221866096866096, + "grad_norm": 0.681168794631958, + "learning_rate": 0.00014400171029183036, + "loss": 1.0867, + "step": 7988 + }, + { + "epoch": 1.4223646723646723, + "grad_norm": 0.6741157174110413, + "learning_rate": 0.0001439891402792801, + "loss": 0.9153, + "step": 7989 + }, + { + "epoch": 1.422542735042735, + "grad_norm": 0.5881659984588623, + "learning_rate": 0.00014397656940485403, + "loss": 0.92, + "step": 7990 + }, + { + "epoch": 1.4227207977207978, + "grad_norm": 0.637093722820282, + "learning_rate": 0.00014396399766879842, + "loss": 0.921, + "step": 7991 + }, + { + "epoch": 1.4228988603988604, + "grad_norm": 0.7760605216026306, + "learning_rate": 0.0001439514250713596, + "loss": 1.1451, + "step": 7992 + }, + { + "epoch": 1.4230769230769231, + "grad_norm": 0.6619600653648376, + "learning_rate": 0.00014393885161278393, + "loss": 1.0365, + "step": 7993 + }, + { + "epoch": 1.4232549857549857, + "grad_norm": 0.5354374051094055, + "learning_rate": 0.0001439262772933177, + "loss": 0.8718, + "step": 7994 + }, + { + "epoch": 1.4234330484330484, + "grad_norm": 0.7063560485839844, + "learning_rate": 0.00014391370211320735, + "loss": 0.8258, + "step": 7995 + }, + { + "epoch": 1.4236111111111112, + "grad_norm": 0.6876368522644043, + "learning_rate": 0.00014390112607269923, + "loss": 0.9579, + "step": 7996 + }, + { + "epoch": 1.423789173789174, + "grad_norm": 0.6976612210273743, + "learning_rate": 0.00014388854917203974, + "loss": 1.0376, + "step": 7997 + }, + { + "epoch": 1.4239672364672364, + "grad_norm": 0.6157355308532715, + "learning_rate": 0.00014387597141147525, + "loss": 0.8743, + "step": 7998 + }, + { + "epoch": 1.4241452991452992, + "grad_norm": 0.7273156046867371, + "learning_rate": 0.0001438633927912523, + "loss": 1.101, + "step": 7999 + }, + { + "epoch": 1.4243233618233617, + "grad_norm": 0.918380618095398, + "learning_rate": 0.0001438508133116173, + "loss": 0.9625, + "step": 8000 + }, + { + "epoch": 1.4245014245014245, + "grad_norm": 0.626040518283844, + "learning_rate": 0.00014383823297281666, + "loss": 0.9552, + "step": 8001 + }, + { + "epoch": 1.4246794871794872, + "grad_norm": 0.7320386171340942, + "learning_rate": 0.00014382565177509693, + "loss": 1.0719, + "step": 8002 + }, + { + "epoch": 1.42485754985755, + "grad_norm": 0.7283148169517517, + "learning_rate": 0.0001438130697187046, + "loss": 1.0455, + "step": 8003 + }, + { + "epoch": 1.4250356125356125, + "grad_norm": 0.6614177823066711, + "learning_rate": 0.00014380048680388613, + "loss": 0.9876, + "step": 8004 + }, + { + "epoch": 1.4252136752136753, + "grad_norm": 0.6726453900337219, + "learning_rate": 0.00014378790303088817, + "loss": 0.9861, + "step": 8005 + }, + { + "epoch": 1.4253917378917378, + "grad_norm": 0.7968725562095642, + "learning_rate": 0.00014377531839995718, + "loss": 1.1662, + "step": 8006 + }, + { + "epoch": 1.4255698005698005, + "grad_norm": 0.6510586738586426, + "learning_rate": 0.0001437627329113398, + "loss": 0.9452, + "step": 8007 + }, + { + "epoch": 1.4257478632478633, + "grad_norm": 0.6933155655860901, + "learning_rate": 0.00014375014656528253, + "loss": 1.0149, + "step": 8008 + }, + { + "epoch": 1.425925925925926, + "grad_norm": 0.7141832113265991, + "learning_rate": 0.00014373755936203204, + "loss": 1.0667, + "step": 8009 + }, + { + "epoch": 1.4261039886039886, + "grad_norm": 0.6352181434631348, + "learning_rate": 0.00014372497130183494, + "loss": 0.8652, + "step": 8010 + }, + { + "epoch": 1.4262820512820513, + "grad_norm": 0.7494860291481018, + "learning_rate": 0.00014371238238493786, + "loss": 0.9592, + "step": 8011 + }, + { + "epoch": 1.4264601139601139, + "grad_norm": 0.610556423664093, + "learning_rate": 0.00014369979261158746, + "loss": 0.7015, + "step": 8012 + }, + { + "epoch": 1.4266381766381766, + "grad_norm": 0.7305756211280823, + "learning_rate": 0.00014368720198203037, + "loss": 0.9681, + "step": 8013 + }, + { + "epoch": 1.4268162393162394, + "grad_norm": 0.6964020133018494, + "learning_rate": 0.0001436746104965133, + "loss": 1.1166, + "step": 8014 + }, + { + "epoch": 1.426994301994302, + "grad_norm": 0.7449237108230591, + "learning_rate": 0.00014366201815528302, + "loss": 1.1331, + "step": 8015 + }, + { + "epoch": 1.4271723646723646, + "grad_norm": 0.625834047794342, + "learning_rate": 0.00014364942495858615, + "loss": 0.8796, + "step": 8016 + }, + { + "epoch": 1.4273504273504274, + "grad_norm": 0.664559006690979, + "learning_rate": 0.0001436368309066695, + "loss": 1.0263, + "step": 8017 + }, + { + "epoch": 1.42752849002849, + "grad_norm": Infinity, + "learning_rate": 0.0001436368309066695, + "loss": 1.0731, + "step": 8018 + }, + { + "epoch": 1.4277065527065527, + "grad_norm": 0.6714464426040649, + "learning_rate": 0.00014362423599977977, + "loss": 0.9345, + "step": 8019 + }, + { + "epoch": 1.4278846153846154, + "grad_norm": 0.7595751285552979, + "learning_rate": 0.00014361164023816376, + "loss": 0.9646, + "step": 8020 + }, + { + "epoch": 1.4280626780626782, + "grad_norm": 0.6413954496383667, + "learning_rate": 0.00014359904362206828, + "loss": 1.0471, + "step": 8021 + }, + { + "epoch": 1.4282407407407407, + "grad_norm": 0.7298843264579773, + "learning_rate": 0.00014358644615174008, + "loss": 0.8932, + "step": 8022 + }, + { + "epoch": 1.4284188034188035, + "grad_norm": 0.8022156953811646, + "learning_rate": 0.00014357384782742602, + "loss": 1.0437, + "step": 8023 + }, + { + "epoch": 1.428596866096866, + "grad_norm": 0.7264443635940552, + "learning_rate": 0.00014356124864937296, + "loss": 0.9368, + "step": 8024 + }, + { + "epoch": 1.4287749287749287, + "grad_norm": 0.6819384098052979, + "learning_rate": 0.00014354864861782768, + "loss": 1.0, + "step": 8025 + }, + { + "epoch": 1.4289529914529915, + "grad_norm": 0.5945104956626892, + "learning_rate": 0.0001435360477330371, + "loss": 0.8108, + "step": 8026 + }, + { + "epoch": 1.4291310541310542, + "grad_norm": 0.6497398018836975, + "learning_rate": 0.0001435234459952481, + "loss": 0.8712, + "step": 8027 + }, + { + "epoch": 1.4293091168091168, + "grad_norm": 0.6424077749252319, + "learning_rate": 0.0001435108434047076, + "loss": 0.9172, + "step": 8028 + }, + { + "epoch": 1.4294871794871795, + "grad_norm": 0.6806963086128235, + "learning_rate": 0.00014349823996166253, + "loss": 1.1648, + "step": 8029 + }, + { + "epoch": 1.429665242165242, + "grad_norm": 0.6601083874702454, + "learning_rate": 0.00014348563566635977, + "loss": 0.9453, + "step": 8030 + }, + { + "epoch": 1.4298433048433048, + "grad_norm": 0.7024385929107666, + "learning_rate": 0.00014347303051904636, + "loss": 1.074, + "step": 8031 + }, + { + "epoch": 1.4300213675213675, + "grad_norm": 0.7094005942344666, + "learning_rate": 0.00014346042451996918, + "loss": 0.9976, + "step": 8032 + }, + { + "epoch": 1.4301994301994303, + "grad_norm": 0.6775936484336853, + "learning_rate": 0.0001434478176693753, + "loss": 0.9039, + "step": 8033 + }, + { + "epoch": 1.4303774928774928, + "grad_norm": 0.6920986771583557, + "learning_rate": 0.00014343520996751166, + "loss": 0.9122, + "step": 8034 + }, + { + "epoch": 1.4305555555555556, + "grad_norm": 0.720690906047821, + "learning_rate": 0.00014342260141462528, + "loss": 1.1028, + "step": 8035 + }, + { + "epoch": 1.430733618233618, + "grad_norm": 0.624546229839325, + "learning_rate": 0.00014340999201096328, + "loss": 0.9083, + "step": 8036 + }, + { + "epoch": 1.4309116809116809, + "grad_norm": 0.6560490727424622, + "learning_rate": 0.00014339738175677265, + "loss": 0.8029, + "step": 8037 + }, + { + "epoch": 1.4310897435897436, + "grad_norm": 0.8266100883483887, + "learning_rate": 0.00014338477065230047, + "loss": 0.9655, + "step": 8038 + }, + { + "epoch": 1.4312678062678064, + "grad_norm": 0.6593570113182068, + "learning_rate": 0.00014337215869779385, + "loss": 1.0299, + "step": 8039 + }, + { + "epoch": 1.431445868945869, + "grad_norm": 0.6321794390678406, + "learning_rate": 0.00014335954589349986, + "loss": 0.8755, + "step": 8040 + }, + { + "epoch": 1.4316239316239316, + "grad_norm": 0.7030870318412781, + "learning_rate": 0.00014334693223966562, + "loss": 1.1226, + "step": 8041 + }, + { + "epoch": 1.4318019943019942, + "grad_norm": 0.7794312238693237, + "learning_rate": 0.0001433343177365383, + "loss": 1.1252, + "step": 8042 + }, + { + "epoch": 1.431980056980057, + "grad_norm": 0.6115018129348755, + "learning_rate": 0.00014332170238436507, + "loss": 0.8753, + "step": 8043 + }, + { + "epoch": 1.4321581196581197, + "grad_norm": 0.8525674939155579, + "learning_rate": 0.00014330908618339304, + "loss": 0.9135, + "step": 8044 + }, + { + "epoch": 1.4323361823361824, + "grad_norm": 0.6869912147521973, + "learning_rate": 0.00014329646913386948, + "loss": 0.868, + "step": 8045 + }, + { + "epoch": 1.432514245014245, + "grad_norm": 0.5877542495727539, + "learning_rate": 0.0001432838512360415, + "loss": 0.9051, + "step": 8046 + }, + { + "epoch": 1.4326923076923077, + "grad_norm": 0.6609327793121338, + "learning_rate": 0.0001432712324901564, + "loss": 0.9084, + "step": 8047 + }, + { + "epoch": 1.4328703703703702, + "grad_norm": 0.6318345069885254, + "learning_rate": 0.0001432586128964614, + "loss": 0.8291, + "step": 8048 + }, + { + "epoch": 1.433048433048433, + "grad_norm": 0.6973567008972168, + "learning_rate": 0.0001432459924552037, + "loss": 0.97, + "step": 8049 + }, + { + "epoch": 1.4332264957264957, + "grad_norm": 0.6838201284408569, + "learning_rate": 0.00014323337116663062, + "loss": 1.0957, + "step": 8050 + }, + { + "epoch": 1.4334045584045585, + "grad_norm": 0.7472857236862183, + "learning_rate": 0.00014322074903098944, + "loss": 1.0981, + "step": 8051 + }, + { + "epoch": 1.433582621082621, + "grad_norm": 0.7723061442375183, + "learning_rate": 0.0001432081260485275, + "loss": 1.2231, + "step": 8052 + }, + { + "epoch": 1.4337606837606838, + "grad_norm": 0.681834876537323, + "learning_rate": 0.00014319550221949208, + "loss": 1.073, + "step": 8053 + }, + { + "epoch": 1.4339387464387463, + "grad_norm": 0.6566045880317688, + "learning_rate": 0.00014318287754413051, + "loss": 1.1298, + "step": 8054 + }, + { + "epoch": 1.434116809116809, + "grad_norm": 0.6792440414428711, + "learning_rate": 0.00014317025202269015, + "loss": 1.2224, + "step": 8055 + }, + { + "epoch": 1.4342948717948718, + "grad_norm": 0.7946709394454956, + "learning_rate": 0.00014315762565541838, + "loss": 1.0728, + "step": 8056 + }, + { + "epoch": 1.4344729344729346, + "grad_norm": 0.633466899394989, + "learning_rate": 0.00014314499844256262, + "loss": 0.944, + "step": 8057 + }, + { + "epoch": 1.434650997150997, + "grad_norm": 0.7308502197265625, + "learning_rate": 0.00014313237038437023, + "loss": 1.0684, + "step": 8058 + }, + { + "epoch": 1.4348290598290598, + "grad_norm": 0.6483737230300903, + "learning_rate": 0.00014311974148108862, + "loss": 1.0843, + "step": 8059 + }, + { + "epoch": 1.4350071225071226, + "grad_norm": 0.6301209926605225, + "learning_rate": 0.00014310711173296526, + "loss": 1.0083, + "step": 8060 + }, + { + "epoch": 1.4351851851851851, + "grad_norm": 0.6674302816390991, + "learning_rate": 0.00014309448114024757, + "loss": 0.9877, + "step": 8061 + }, + { + "epoch": 1.4353632478632479, + "grad_norm": 0.6888732314109802, + "learning_rate": 0.00014308184970318307, + "loss": 0.9937, + "step": 8062 + }, + { + "epoch": 1.4355413105413106, + "grad_norm": 0.6922950148582458, + "learning_rate": 0.00014306921742201923, + "loss": 1.0149, + "step": 8063 + }, + { + "epoch": 1.4357193732193732, + "grad_norm": 0.6050686240196228, + "learning_rate": 0.00014305658429700352, + "loss": 0.7882, + "step": 8064 + }, + { + "epoch": 1.435897435897436, + "grad_norm": 0.5080767869949341, + "learning_rate": 0.00014304395032838348, + "loss": 0.7796, + "step": 8065 + }, + { + "epoch": 1.4360754985754987, + "grad_norm": 0.6382707953453064, + "learning_rate": 0.00014303131551640668, + "loss": 0.965, + "step": 8066 + }, + { + "epoch": 1.4362535612535612, + "grad_norm": 0.7153477668762207, + "learning_rate": 0.00014301867986132063, + "loss": 1.1277, + "step": 8067 + }, + { + "epoch": 1.436431623931624, + "grad_norm": 0.6208404898643494, + "learning_rate": 0.00014300604336337292, + "loss": 0.8246, + "step": 8068 + }, + { + "epoch": 1.4366096866096867, + "grad_norm": 0.719695508480072, + "learning_rate": 0.0001429934060228111, + "loss": 0.7681, + "step": 8069 + }, + { + "epoch": 1.4367877492877492, + "grad_norm": 0.6219030618667603, + "learning_rate": 0.0001429807678398828, + "loss": 1.0425, + "step": 8070 + }, + { + "epoch": 1.436965811965812, + "grad_norm": 0.6080238819122314, + "learning_rate": 0.00014296812881483566, + "loss": 0.8762, + "step": 8071 + }, + { + "epoch": 1.4371438746438747, + "grad_norm": 0.6264194846153259, + "learning_rate": 0.00014295548894791729, + "loss": 1.087, + "step": 8072 + }, + { + "epoch": 1.4373219373219372, + "grad_norm": 0.6503600478172302, + "learning_rate": 0.00014294284823937535, + "loss": 1.0583, + "step": 8073 + }, + { + "epoch": 1.4375, + "grad_norm": 0.7623817324638367, + "learning_rate": 0.0001429302066894575, + "loss": 1.2372, + "step": 8074 + }, + { + "epoch": 1.4376780626780628, + "grad_norm": 0.7020344138145447, + "learning_rate": 0.00014291756429841144, + "loss": 1.2163, + "step": 8075 + }, + { + "epoch": 1.4378561253561253, + "grad_norm": 0.7070338129997253, + "learning_rate": 0.00014290492106648484, + "loss": 0.986, + "step": 8076 + }, + { + "epoch": 1.438034188034188, + "grad_norm": 0.6407621502876282, + "learning_rate": 0.00014289227699392545, + "loss": 0.9329, + "step": 8077 + }, + { + "epoch": 1.4382122507122508, + "grad_norm": 0.6836710572242737, + "learning_rate": 0.00014287963208098098, + "loss": 0.9252, + "step": 8078 + }, + { + "epoch": 1.4383903133903133, + "grad_norm": 0.648642897605896, + "learning_rate": 0.00014286698632789922, + "loss": 1.0457, + "step": 8079 + }, + { + "epoch": 1.438568376068376, + "grad_norm": 0.7015881538391113, + "learning_rate": 0.0001428543397349279, + "loss": 1.0516, + "step": 8080 + }, + { + "epoch": 1.4387464387464388, + "grad_norm": 0.6031532883644104, + "learning_rate": 0.0001428416923023148, + "loss": 0.9423, + "step": 8081 + }, + { + "epoch": 1.4389245014245013, + "grad_norm": 0.8235578536987305, + "learning_rate": 0.00014282904403030772, + "loss": 1.3433, + "step": 8082 + }, + { + "epoch": 1.439102564102564, + "grad_norm": 0.7355761528015137, + "learning_rate": 0.00014281639491915452, + "loss": 1.0128, + "step": 8083 + }, + { + "epoch": 1.4392806267806268, + "grad_norm": 0.7429629564285278, + "learning_rate": 0.00014280374496910303, + "loss": 0.8546, + "step": 8084 + }, + { + "epoch": 1.4394586894586894, + "grad_norm": 0.5831776857376099, + "learning_rate": 0.00014279109418040105, + "loss": 0.9021, + "step": 8085 + }, + { + "epoch": 1.4396367521367521, + "grad_norm": 0.6585184931755066, + "learning_rate": 0.00014277844255329645, + "loss": 0.9256, + "step": 8086 + }, + { + "epoch": 1.4398148148148149, + "grad_norm": 0.6412501931190491, + "learning_rate": 0.00014276579008803717, + "loss": 0.9305, + "step": 8087 + }, + { + "epoch": 1.4399928774928774, + "grad_norm": 0.6305423378944397, + "learning_rate": 0.00014275313678487102, + "loss": 0.9471, + "step": 8088 + }, + { + "epoch": 1.4401709401709402, + "grad_norm": 0.7160914540290833, + "learning_rate": 0.00014274048264404602, + "loss": 0.8798, + "step": 8089 + }, + { + "epoch": 1.440349002849003, + "grad_norm": 0.6740858554840088, + "learning_rate": 0.00014272782766581004, + "loss": 0.9022, + "step": 8090 + }, + { + "epoch": 1.4405270655270654, + "grad_norm": 0.7554821968078613, + "learning_rate": 0.000142715171850411, + "loss": 1.0924, + "step": 8091 + }, + { + "epoch": 1.4407051282051282, + "grad_norm": 0.7361162304878235, + "learning_rate": 0.00014270251519809694, + "loss": 0.9907, + "step": 8092 + }, + { + "epoch": 1.440883190883191, + "grad_norm": 0.731813371181488, + "learning_rate": 0.0001426898577091158, + "loss": 1.1765, + "step": 8093 + }, + { + "epoch": 1.4410612535612537, + "grad_norm": 0.6877756714820862, + "learning_rate": 0.00014267719938371558, + "loss": 1.0536, + "step": 8094 + }, + { + "epoch": 1.4412393162393162, + "grad_norm": 0.6724407076835632, + "learning_rate": 0.00014266454022214426, + "loss": 1.1895, + "step": 8095 + }, + { + "epoch": 1.441417378917379, + "grad_norm": 0.6946671605110168, + "learning_rate": 0.0001426518802246499, + "loss": 1.0437, + "step": 8096 + }, + { + "epoch": 1.4415954415954415, + "grad_norm": 0.7032839059829712, + "learning_rate": 0.00014263921939148058, + "loss": 1.1363, + "step": 8097 + }, + { + "epoch": 1.4417735042735043, + "grad_norm": 0.6942192316055298, + "learning_rate": 0.00014262655772288434, + "loss": 1.315, + "step": 8098 + }, + { + "epoch": 1.441951566951567, + "grad_norm": 0.7002301812171936, + "learning_rate": 0.00014261389521910922, + "loss": 1.0546, + "step": 8099 + }, + { + "epoch": 1.4421296296296298, + "grad_norm": 0.7260788083076477, + "learning_rate": 0.00014260123188040335, + "loss": 0.9374, + "step": 8100 + }, + { + "epoch": 1.4423076923076923, + "grad_norm": 0.6629201173782349, + "learning_rate": 0.00014258856770701486, + "loss": 0.8632, + "step": 8101 + }, + { + "epoch": 1.442485754985755, + "grad_norm": 0.6570318937301636, + "learning_rate": 0.0001425759026991918, + "loss": 1.0102, + "step": 8102 + }, + { + "epoch": 1.4426638176638176, + "grad_norm": 0.7696560621261597, + "learning_rate": 0.00014256323685718242, + "loss": 0.9703, + "step": 8103 + }, + { + "epoch": 1.4428418803418803, + "grad_norm": 0.7206611633300781, + "learning_rate": 0.00014255057018123482, + "loss": 1.1728, + "step": 8104 + }, + { + "epoch": 1.443019943019943, + "grad_norm": 0.6871611475944519, + "learning_rate": 0.0001425379026715972, + "loss": 0.9377, + "step": 8105 + }, + { + "epoch": 1.4431980056980058, + "grad_norm": 0.6027442812919617, + "learning_rate": 0.00014252523432851775, + "loss": 0.9212, + "step": 8106 + }, + { + "epoch": 1.4433760683760684, + "grad_norm": 0.7149752378463745, + "learning_rate": 0.00014251256515224463, + "loss": 0.9654, + "step": 8107 + }, + { + "epoch": 1.443554131054131, + "grad_norm": 0.5949522256851196, + "learning_rate": 0.00014249989514302614, + "loss": 1.0646, + "step": 8108 + }, + { + "epoch": 1.4437321937321936, + "grad_norm": 0.7345452904701233, + "learning_rate": 0.0001424872243011105, + "loss": 0.9801, + "step": 8109 + }, + { + "epoch": 1.4439102564102564, + "grad_norm": 0.8045009970664978, + "learning_rate": 0.00014247455262674592, + "loss": 1.3529, + "step": 8110 + }, + { + "epoch": 1.4440883190883191, + "grad_norm": 0.6712123155593872, + "learning_rate": 0.00014246188012018073, + "loss": 1.0416, + "step": 8111 + }, + { + "epoch": 1.444266381766382, + "grad_norm": 0.7811154127120972, + "learning_rate": 0.00014244920678166322, + "loss": 1.2019, + "step": 8112 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 0.6834486126899719, + "learning_rate": 0.00014243653261144167, + "loss": 0.986, + "step": 8113 + }, + { + "epoch": 1.4446225071225072, + "grad_norm": 0.6901041269302368, + "learning_rate": 0.00014242385760976443, + "loss": 1.0988, + "step": 8114 + }, + { + "epoch": 1.4448005698005697, + "grad_norm": 0.6233634948730469, + "learning_rate": 0.00014241118177687982, + "loss": 0.7748, + "step": 8115 + }, + { + "epoch": 1.4449786324786325, + "grad_norm": 0.6899837851524353, + "learning_rate": 0.00014239850511303624, + "loss": 0.9734, + "step": 8116 + }, + { + "epoch": 1.4451566951566952, + "grad_norm": 0.6316244006156921, + "learning_rate": 0.00014238582761848197, + "loss": 0.7888, + "step": 8117 + }, + { + "epoch": 1.445334757834758, + "grad_norm": 0.6074259877204895, + "learning_rate": 0.00014237314929346545, + "loss": 0.8843, + "step": 8118 + }, + { + "epoch": 1.4455128205128205, + "grad_norm": 0.6112192273139954, + "learning_rate": 0.00014236047013823516, + "loss": 0.8529, + "step": 8119 + }, + { + "epoch": 1.4456908831908832, + "grad_norm": 0.6883894801139832, + "learning_rate": 0.0001423477901530394, + "loss": 0.9506, + "step": 8120 + }, + { + "epoch": 1.4458689458689458, + "grad_norm": 0.7248309254646301, + "learning_rate": 0.00014233510933812666, + "loss": 0.9573, + "step": 8121 + }, + { + "epoch": 1.4460470085470085, + "grad_norm": 0.6853367686271667, + "learning_rate": 0.00014232242769374542, + "loss": 0.9903, + "step": 8122 + }, + { + "epoch": 1.4462250712250713, + "grad_norm": 0.7179274559020996, + "learning_rate": 0.0001423097452201441, + "loss": 0.9157, + "step": 8123 + }, + { + "epoch": 1.446403133903134, + "grad_norm": 0.6704817414283752, + "learning_rate": 0.00014229706191757127, + "loss": 1.1361, + "step": 8124 + }, + { + "epoch": 1.4465811965811965, + "grad_norm": 0.6380739212036133, + "learning_rate": 0.00014228437778627533, + "loss": 0.9336, + "step": 8125 + }, + { + "epoch": 1.4467592592592593, + "grad_norm": 0.6275362372398376, + "learning_rate": 0.00014227169282650487, + "loss": 0.9617, + "step": 8126 + }, + { + "epoch": 1.4469373219373218, + "grad_norm": 0.5644828677177429, + "learning_rate": 0.00014225900703850836, + "loss": 0.7384, + "step": 8127 + }, + { + "epoch": 1.4471153846153846, + "grad_norm": 0.6522284150123596, + "learning_rate": 0.00014224632042253443, + "loss": 1.1098, + "step": 8128 + }, + { + "epoch": 1.4472934472934473, + "grad_norm": 0.6228049993515015, + "learning_rate": 0.0001422336329788316, + "loss": 1.1061, + "step": 8129 + }, + { + "epoch": 1.44747150997151, + "grad_norm": 0.6092000603675842, + "learning_rate": 0.00014222094470764848, + "loss": 0.808, + "step": 8130 + }, + { + "epoch": 1.4476495726495726, + "grad_norm": 0.667435348033905, + "learning_rate": 0.00014220825560923363, + "loss": 1.1223, + "step": 8131 + }, + { + "epoch": 1.4478276353276354, + "grad_norm": 0.6080766320228577, + "learning_rate": 0.0001421955656838357, + "loss": 1.0099, + "step": 8132 + }, + { + "epoch": 1.448005698005698, + "grad_norm": 0.7597638368606567, + "learning_rate": 0.00014218287493170332, + "loss": 0.9718, + "step": 8133 + }, + { + "epoch": 1.4481837606837606, + "grad_norm": 0.574130654335022, + "learning_rate": 0.0001421701833530851, + "loss": 0.7745, + "step": 8134 + }, + { + "epoch": 1.4483618233618234, + "grad_norm": 0.6372822523117065, + "learning_rate": 0.0001421574909482298, + "loss": 1.0088, + "step": 8135 + }, + { + "epoch": 1.4485398860398861, + "grad_norm": 0.6759644746780396, + "learning_rate": 0.000142144797717386, + "loss": 0.9684, + "step": 8136 + }, + { + "epoch": 1.4487179487179487, + "grad_norm": 0.706351637840271, + "learning_rate": 0.00014213210366080244, + "loss": 1.021, + "step": 8137 + }, + { + "epoch": 1.4488960113960114, + "grad_norm": 0.6976894736289978, + "learning_rate": 0.0001421194087787278, + "loss": 1.1038, + "step": 8138 + }, + { + "epoch": 1.449074074074074, + "grad_norm": 0.7322551012039185, + "learning_rate": 0.00014210671307141092, + "loss": 1.0213, + "step": 8139 + }, + { + "epoch": 1.4492521367521367, + "grad_norm": 0.5885626077651978, + "learning_rate": 0.0001420940165391004, + "loss": 0.821, + "step": 8140 + }, + { + "epoch": 1.4494301994301995, + "grad_norm": 0.7009791135787964, + "learning_rate": 0.0001420813191820451, + "loss": 0.8647, + "step": 8141 + }, + { + "epoch": 1.4496082621082622, + "grad_norm": 0.5715423822402954, + "learning_rate": 0.00014206862100049375, + "loss": 0.873, + "step": 8142 + }, + { + "epoch": 1.4497863247863247, + "grad_norm": 1.1452178955078125, + "learning_rate": 0.00014205592199469514, + "loss": 1.2523, + "step": 8143 + }, + { + "epoch": 1.4499643874643875, + "grad_norm": 0.8076814413070679, + "learning_rate": 0.00014204322216489814, + "loss": 1.1071, + "step": 8144 + }, + { + "epoch": 1.45014245014245, + "grad_norm": 0.7325751185417175, + "learning_rate": 0.00014203052151135154, + "loss": 0.9846, + "step": 8145 + }, + { + "epoch": 1.4503205128205128, + "grad_norm": 0.7009061574935913, + "learning_rate": 0.00014201782003430417, + "loss": 0.8153, + "step": 8146 + }, + { + "epoch": 1.4504985754985755, + "grad_norm": 0.6502353549003601, + "learning_rate": 0.0001420051177340049, + "loss": 0.8959, + "step": 8147 + }, + { + "epoch": 1.4506766381766383, + "grad_norm": 0.6134430170059204, + "learning_rate": 0.00014199241461070261, + "loss": 0.9683, + "step": 8148 + }, + { + "epoch": 1.4508547008547008, + "grad_norm": 0.720160722732544, + "learning_rate": 0.0001419797106646462, + "loss": 0.9579, + "step": 8149 + }, + { + "epoch": 1.4510327635327636, + "grad_norm": 0.6141422986984253, + "learning_rate": 0.00014196700589608454, + "loss": 0.9427, + "step": 8150 + }, + { + "epoch": 1.451210826210826, + "grad_norm": 0.6835139393806458, + "learning_rate": 0.00014195430030526656, + "loss": 1.0374, + "step": 8151 + }, + { + "epoch": 1.4513888888888888, + "grad_norm": 0.6829691529273987, + "learning_rate": 0.00014194159389244128, + "loss": 0.9418, + "step": 8152 + }, + { + "epoch": 1.4515669515669516, + "grad_norm": 0.7142195701599121, + "learning_rate": 0.00014192888665785755, + "loss": 1.1876, + "step": 8153 + }, + { + "epoch": 1.4517450142450143, + "grad_norm": 0.6719943284988403, + "learning_rate": 0.0001419161786017644, + "loss": 1.1417, + "step": 8154 + }, + { + "epoch": 1.4519230769230769, + "grad_norm": 0.6478939652442932, + "learning_rate": 0.0001419034697244108, + "loss": 0.943, + "step": 8155 + }, + { + "epoch": 1.4521011396011396, + "grad_norm": 0.6308888792991638, + "learning_rate": 0.00014189076002604575, + "loss": 0.9842, + "step": 8156 + }, + { + "epoch": 1.4522792022792022, + "grad_norm": 0.673559844493866, + "learning_rate": 0.00014187804950691827, + "loss": 0.8108, + "step": 8157 + }, + { + "epoch": 1.452457264957265, + "grad_norm": 0.5895359516143799, + "learning_rate": 0.00014186533816727744, + "loss": 0.8187, + "step": 8158 + }, + { + "epoch": 1.4526353276353277, + "grad_norm": 0.6703287363052368, + "learning_rate": 0.00014185262600737225, + "loss": 0.9012, + "step": 8159 + }, + { + "epoch": 1.4528133903133904, + "grad_norm": 0.697728157043457, + "learning_rate": 0.00014183991302745182, + "loss": 1.2572, + "step": 8160 + }, + { + "epoch": 1.452991452991453, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00014182719922776514, + "loss": 1.078, + "step": 8161 + }, + { + "epoch": 1.4531695156695157, + "grad_norm": 0.6774863600730896, + "learning_rate": 0.00014181448460856143, + "loss": 1.0607, + "step": 8162 + }, + { + "epoch": 1.4533475783475782, + "grad_norm": 0.6872009038925171, + "learning_rate": 0.00014180176917008976, + "loss": 1.0713, + "step": 8163 + }, + { + "epoch": 1.453525641025641, + "grad_norm": 0.7949981093406677, + "learning_rate": 0.00014178905291259926, + "loss": 1.0471, + "step": 8164 + }, + { + "epoch": 1.4537037037037037, + "grad_norm": 0.6592127084732056, + "learning_rate": 0.00014177633583633908, + "loss": 0.8409, + "step": 8165 + }, + { + "epoch": 1.4538817663817665, + "grad_norm": 0.6745635867118835, + "learning_rate": 0.00014176361794155837, + "loss": 1.0859, + "step": 8166 + }, + { + "epoch": 1.454059829059829, + "grad_norm": 0.6661605834960938, + "learning_rate": 0.00014175089922850633, + "loss": 1.0587, + "step": 8167 + }, + { + "epoch": 1.4542378917378918, + "grad_norm": 0.6697571873664856, + "learning_rate": 0.00014173817969743212, + "loss": 0.8876, + "step": 8168 + }, + { + "epoch": 1.4544159544159543, + "grad_norm": 0.6162588000297546, + "learning_rate": 0.000141725459348585, + "loss": 0.9575, + "step": 8169 + }, + { + "epoch": 1.454594017094017, + "grad_norm": 0.6235088109970093, + "learning_rate": 0.00014171273818221422, + "loss": 0.9209, + "step": 8170 + }, + { + "epoch": 1.4547720797720798, + "grad_norm": 0.6744212508201599, + "learning_rate": 0.00014170001619856896, + "loss": 0.9704, + "step": 8171 + }, + { + "epoch": 1.4549501424501425, + "grad_norm": 0.6781345009803772, + "learning_rate": 0.0001416872933978985, + "loss": 1.1507, + "step": 8172 + }, + { + "epoch": 1.455128205128205, + "grad_norm": 0.7160060405731201, + "learning_rate": 0.0001416745697804521, + "loss": 1.2529, + "step": 8173 + }, + { + "epoch": 1.4553062678062678, + "grad_norm": 0.6742389798164368, + "learning_rate": 0.00014166184534647913, + "loss": 1.0168, + "step": 8174 + }, + { + "epoch": 1.4554843304843303, + "grad_norm": 0.6685828566551208, + "learning_rate": 0.0001416491200962288, + "loss": 1.0807, + "step": 8175 + }, + { + "epoch": 1.455662393162393, + "grad_norm": 0.6998327374458313, + "learning_rate": 0.0001416363940299505, + "loss": 1.1711, + "step": 8176 + }, + { + "epoch": 1.4558404558404558, + "grad_norm": 0.7132518291473389, + "learning_rate": 0.00014162366714789358, + "loss": 1.1392, + "step": 8177 + }, + { + "epoch": 1.4560185185185186, + "grad_norm": 0.6995887160301208, + "learning_rate": 0.0001416109394503073, + "loss": 1.3335, + "step": 8178 + }, + { + "epoch": 1.4561965811965811, + "grad_norm": 0.7161234021186829, + "learning_rate": 0.00014159821093744115, + "loss": 0.9725, + "step": 8179 + }, + { + "epoch": 1.4563746438746439, + "grad_norm": 0.7678874135017395, + "learning_rate": 0.00014158548160954446, + "loss": 1.1578, + "step": 8180 + }, + { + "epoch": 1.4565527065527066, + "grad_norm": 0.67372065782547, + "learning_rate": 0.00014157275146686662, + "loss": 1.0867, + "step": 8181 + }, + { + "epoch": 1.4567307692307692, + "grad_norm": 0.7757831811904907, + "learning_rate": 0.00014156002050965712, + "loss": 0.9768, + "step": 8182 + }, + { + "epoch": 1.456908831908832, + "grad_norm": 0.7174801230430603, + "learning_rate": 0.00014154728873816533, + "loss": 1.1712, + "step": 8183 + }, + { + "epoch": 1.4570868945868947, + "grad_norm": 0.5972673892974854, + "learning_rate": 0.0001415345561526407, + "loss": 0.9571, + "step": 8184 + }, + { + "epoch": 1.4572649572649572, + "grad_norm": 0.7999650835990906, + "learning_rate": 0.00014152182275333275, + "loss": 1.0583, + "step": 8185 + }, + { + "epoch": 1.45744301994302, + "grad_norm": 0.6737848520278931, + "learning_rate": 0.00014150908854049091, + "loss": 1.0562, + "step": 8186 + }, + { + "epoch": 1.4576210826210827, + "grad_norm": 0.7756418585777283, + "learning_rate": 0.00014149635351436474, + "loss": 1.2301, + "step": 8187 + }, + { + "epoch": 1.4577991452991452, + "grad_norm": 0.5633914470672607, + "learning_rate": 0.00014148361767520374, + "loss": 0.8847, + "step": 8188 + }, + { + "epoch": 1.457977207977208, + "grad_norm": 0.8462759256362915, + "learning_rate": 0.00014147088102325737, + "loss": 0.8046, + "step": 8189 + }, + { + "epoch": 1.4581552706552707, + "grad_norm": 0.7081632614135742, + "learning_rate": 0.00014145814355877526, + "loss": 1.0764, + "step": 8190 + }, + { + "epoch": 1.4583333333333333, + "grad_norm": 0.7357106804847717, + "learning_rate": 0.00014144540528200698, + "loss": 1.0202, + "step": 8191 + }, + { + "epoch": 1.458511396011396, + "grad_norm": 0.603566586971283, + "learning_rate": 0.00014143266619320204, + "loss": 0.8214, + "step": 8192 + }, + { + "epoch": 1.4586894586894588, + "grad_norm": 0.6829110383987427, + "learning_rate": 0.00014141992629261007, + "loss": 0.9479, + "step": 8193 + }, + { + "epoch": 1.4588675213675213, + "grad_norm": 0.6822739839553833, + "learning_rate": 0.00014140718558048072, + "loss": 0.9117, + "step": 8194 + }, + { + "epoch": 1.459045584045584, + "grad_norm": 0.7383607029914856, + "learning_rate": 0.00014139444405706356, + "loss": 0.9819, + "step": 8195 + }, + { + "epoch": 1.4592236467236468, + "grad_norm": 0.6319897770881653, + "learning_rate": 0.00014138170172260826, + "loss": 1.0508, + "step": 8196 + }, + { + "epoch": 1.4594017094017093, + "grad_norm": 0.6804461479187012, + "learning_rate": 0.0001413689585773645, + "loss": 0.992, + "step": 8197 + }, + { + "epoch": 1.459579772079772, + "grad_norm": 0.6198720335960388, + "learning_rate": 0.0001413562146215819, + "loss": 1.0113, + "step": 8198 + }, + { + "epoch": 1.4597578347578348, + "grad_norm": 0.5968540906906128, + "learning_rate": 0.0001413434698555102, + "loss": 0.7562, + "step": 8199 + }, + { + "epoch": 1.4599358974358974, + "grad_norm": 0.5370334982872009, + "learning_rate": 0.00014133072427939913, + "loss": 0.9238, + "step": 8200 + }, + { + "epoch": 1.46011396011396, + "grad_norm": 0.6652548909187317, + "learning_rate": 0.00014131797789349832, + "loss": 0.9464, + "step": 8201 + }, + { + "epoch": 1.4602920227920229, + "grad_norm": 0.637852668762207, + "learning_rate": 0.00014130523069805757, + "loss": 1.0395, + "step": 8202 + }, + { + "epoch": 1.4604700854700854, + "grad_norm": 0.8186550140380859, + "learning_rate": 0.00014129248269332664, + "loss": 1.2116, + "step": 8203 + }, + { + "epoch": 1.4606481481481481, + "grad_norm": 0.5290196537971497, + "learning_rate": 0.00014127973387955528, + "loss": 0.7331, + "step": 8204 + }, + { + "epoch": 1.460826210826211, + "grad_norm": 0.6516342163085938, + "learning_rate": 0.00014126698425699332, + "loss": 0.9275, + "step": 8205 + }, + { + "epoch": 1.4610042735042734, + "grad_norm": 0.767254114151001, + "learning_rate": 0.00014125423382589048, + "loss": 0.9355, + "step": 8206 + }, + { + "epoch": 1.4611823361823362, + "grad_norm": 0.6476777195930481, + "learning_rate": 0.00014124148258649668, + "loss": 0.9263, + "step": 8207 + }, + { + "epoch": 1.461360398860399, + "grad_norm": 0.6737871766090393, + "learning_rate": 0.00014122873053906167, + "loss": 0.9815, + "step": 8208 + }, + { + "epoch": 1.4615384615384617, + "grad_norm": 0.6311159729957581, + "learning_rate": 0.00014121597768383532, + "loss": 0.9607, + "step": 8209 + }, + { + "epoch": 1.4617165242165242, + "grad_norm": 0.6061250567436218, + "learning_rate": 0.00014120322402106752, + "loss": 0.7428, + "step": 8210 + }, + { + "epoch": 1.461894586894587, + "grad_norm": 0.6916252970695496, + "learning_rate": 0.00014119046955100815, + "loss": 0.9664, + "step": 8211 + }, + { + "epoch": 1.4620726495726495, + "grad_norm": 0.6583660840988159, + "learning_rate": 0.00014117771427390706, + "loss": 1.0645, + "step": 8212 + }, + { + "epoch": 1.4622507122507122, + "grad_norm": 0.7034604549407959, + "learning_rate": 0.00014116495819001425, + "loss": 0.9223, + "step": 8213 + }, + { + "epoch": 1.462428774928775, + "grad_norm": 0.6378605961799622, + "learning_rate": 0.00014115220129957954, + "loss": 0.7963, + "step": 8214 + }, + { + "epoch": 1.4626068376068377, + "grad_norm": 0.6251596212387085, + "learning_rate": 0.00014113944360285297, + "loss": 0.9852, + "step": 8215 + }, + { + "epoch": 1.4627849002849003, + "grad_norm": 0.7055560946464539, + "learning_rate": 0.00014112668510008446, + "loss": 0.9342, + "step": 8216 + }, + { + "epoch": 1.462962962962963, + "grad_norm": 0.6250377893447876, + "learning_rate": 0.00014111392579152396, + "loss": 0.9886, + "step": 8217 + }, + { + "epoch": 1.4631410256410255, + "grad_norm": 0.6011185050010681, + "learning_rate": 0.00014110116567742152, + "loss": 0.8465, + "step": 8218 + }, + { + "epoch": 1.4633190883190883, + "grad_norm": 0.6632489562034607, + "learning_rate": 0.0001410884047580271, + "loss": 0.8619, + "step": 8219 + }, + { + "epoch": 1.463497150997151, + "grad_norm": 0.7194828987121582, + "learning_rate": 0.00014107564303359076, + "loss": 1.1231, + "step": 8220 + }, + { + "epoch": 1.4636752136752138, + "grad_norm": 0.7640393376350403, + "learning_rate": 0.0001410628805043625, + "loss": 1.1955, + "step": 8221 + }, + { + "epoch": 1.4638532763532763, + "grad_norm": 0.9118906259536743, + "learning_rate": 0.0001410501171705924, + "loss": 1.0555, + "step": 8222 + }, + { + "epoch": 1.464031339031339, + "grad_norm": 0.7545066475868225, + "learning_rate": 0.00014103735303253053, + "loss": 0.9425, + "step": 8223 + }, + { + "epoch": 1.4642094017094016, + "grad_norm": 0.6848801970481873, + "learning_rate": 0.000141024588090427, + "loss": 1.0418, + "step": 8224 + }, + { + "epoch": 1.4643874643874644, + "grad_norm": 0.6825160384178162, + "learning_rate": 0.00014101182234453185, + "loss": 0.9615, + "step": 8225 + }, + { + "epoch": 1.4645655270655271, + "grad_norm": 0.8258556723594666, + "learning_rate": 0.00014099905579509527, + "loss": 1.1237, + "step": 8226 + }, + { + "epoch": 1.4647435897435899, + "grad_norm": 0.6427522897720337, + "learning_rate": 0.00014098628844236733, + "loss": 1.0853, + "step": 8227 + }, + { + "epoch": 1.4649216524216524, + "grad_norm": 0.6476351022720337, + "learning_rate": 0.00014097352028659825, + "loss": 1.1286, + "step": 8228 + }, + { + "epoch": 1.4650997150997151, + "grad_norm": 0.7621034383773804, + "learning_rate": 0.00014096075132803812, + "loss": 1.1402, + "step": 8229 + }, + { + "epoch": 1.4652777777777777, + "grad_norm": 0.6629892587661743, + "learning_rate": 0.00014094798156693718, + "loss": 0.7108, + "step": 8230 + }, + { + "epoch": 1.4654558404558404, + "grad_norm": 0.6902043223381042, + "learning_rate": 0.00014093521100354557, + "loss": 1.1761, + "step": 8231 + }, + { + "epoch": 1.4656339031339032, + "grad_norm": 0.7422910928726196, + "learning_rate": 0.00014092243963811357, + "loss": 0.867, + "step": 8232 + }, + { + "epoch": 1.465811965811966, + "grad_norm": 0.7424963712692261, + "learning_rate": 0.00014090966747089137, + "loss": 1.015, + "step": 8233 + }, + { + "epoch": 1.4659900284900285, + "grad_norm": 0.6855891942977905, + "learning_rate": 0.0001408968945021292, + "loss": 0.9624, + "step": 8234 + }, + { + "epoch": 1.4661680911680912, + "grad_norm": 0.5968918204307556, + "learning_rate": 0.00014088412073207736, + "loss": 0.9243, + "step": 8235 + }, + { + "epoch": 1.4663461538461537, + "grad_norm": 0.6153344511985779, + "learning_rate": 0.0001408713461609861, + "loss": 1.0305, + "step": 8236 + }, + { + "epoch": 1.4665242165242165, + "grad_norm": 0.6627458333969116, + "learning_rate": 0.0001408585707891057, + "loss": 1.1102, + "step": 8237 + }, + { + "epoch": 1.4667022792022792, + "grad_norm": 0.6475233435630798, + "learning_rate": 0.0001408457946166865, + "loss": 1.0045, + "step": 8238 + }, + { + "epoch": 1.466880341880342, + "grad_norm": 0.6792858839035034, + "learning_rate": 0.00014083301764397876, + "loss": 1.0092, + "step": 8239 + }, + { + "epoch": 1.4670584045584045, + "grad_norm": 0.6916255354881287, + "learning_rate": 0.00014082023987123293, + "loss": 1.0761, + "step": 8240 + }, + { + "epoch": 1.4672364672364673, + "grad_norm": 0.7901251912117004, + "learning_rate": 0.00014080746129869923, + "loss": 0.8002, + "step": 8241 + }, + { + "epoch": 1.4674145299145298, + "grad_norm": 0.8078263401985168, + "learning_rate": 0.00014079468192662812, + "loss": 0.9738, + "step": 8242 + }, + { + "epoch": 1.4675925925925926, + "grad_norm": 0.6370784640312195, + "learning_rate": 0.00014078190175526996, + "loss": 1.0256, + "step": 8243 + }, + { + "epoch": 1.4677706552706553, + "grad_norm": 0.6087532639503479, + "learning_rate": 0.0001407691207848752, + "loss": 0.9747, + "step": 8244 + }, + { + "epoch": 1.467948717948718, + "grad_norm": 0.6333357691764832, + "learning_rate": 0.00014075633901569414, + "loss": 1.0135, + "step": 8245 + }, + { + "epoch": 1.4681267806267806, + "grad_norm": 0.6914255619049072, + "learning_rate": 0.00014074355644797733, + "loss": 1.0261, + "step": 8246 + }, + { + "epoch": 1.4683048433048433, + "grad_norm": 0.6374734044075012, + "learning_rate": 0.00014073077308197513, + "loss": 0.9197, + "step": 8247 + }, + { + "epoch": 1.4684829059829059, + "grad_norm": 0.8023789525032043, + "learning_rate": 0.00014071798891793807, + "loss": 1.1085, + "step": 8248 + }, + { + "epoch": 1.4686609686609686, + "grad_norm": 0.7722933888435364, + "learning_rate": 0.0001407052039561166, + "loss": 1.2018, + "step": 8249 + }, + { + "epoch": 1.4688390313390314, + "grad_norm": 0.6823393106460571, + "learning_rate": 0.0001406924181967612, + "loss": 1.088, + "step": 8250 + }, + { + "epoch": 1.4690170940170941, + "grad_norm": 0.7037357687950134, + "learning_rate": 0.00014067963164012242, + "loss": 1.0324, + "step": 8251 + }, + { + "epoch": 1.4691951566951567, + "grad_norm": 0.6549737453460693, + "learning_rate": 0.00014066684428645074, + "loss": 1.152, + "step": 8252 + }, + { + "epoch": 1.4693732193732194, + "grad_norm": 0.5349790453910828, + "learning_rate": 0.00014065405613599674, + "loss": 0.6996, + "step": 8253 + }, + { + "epoch": 1.469551282051282, + "grad_norm": 0.6760679483413696, + "learning_rate": 0.00014064126718901096, + "loss": 0.9856, + "step": 8254 + }, + { + "epoch": 1.4697293447293447, + "grad_norm": 0.5912436842918396, + "learning_rate": 0.00014062847744574395, + "loss": 1.0076, + "step": 8255 + }, + { + "epoch": 1.4699074074074074, + "grad_norm": 0.75101637840271, + "learning_rate": 0.00014061568690644632, + "loss": 1.0033, + "step": 8256 + }, + { + "epoch": 1.4700854700854702, + "grad_norm": 0.6233504414558411, + "learning_rate": 0.00014060289557136873, + "loss": 0.8525, + "step": 8257 + }, + { + "epoch": 1.4702635327635327, + "grad_norm": 0.659570038318634, + "learning_rate": 0.00014059010344076171, + "loss": 0.855, + "step": 8258 + }, + { + "epoch": 1.4704415954415955, + "grad_norm": 0.8096539974212646, + "learning_rate": 0.00014057731051487593, + "loss": 0.9905, + "step": 8259 + }, + { + "epoch": 1.470619658119658, + "grad_norm": 0.5829728245735168, + "learning_rate": 0.00014056451679396204, + "loss": 0.7974, + "step": 8260 + }, + { + "epoch": 1.4707977207977208, + "grad_norm": 0.6176979541778564, + "learning_rate": 0.0001405517222782707, + "loss": 0.9556, + "step": 8261 + }, + { + "epoch": 1.4709757834757835, + "grad_norm": 0.6322479248046875, + "learning_rate": 0.00014053892696805264, + "loss": 0.8837, + "step": 8262 + }, + { + "epoch": 1.4711538461538463, + "grad_norm": 0.6886917948722839, + "learning_rate": 0.0001405261308635585, + "loss": 0.9242, + "step": 8263 + }, + { + "epoch": 1.4713319088319088, + "grad_norm": 0.7474521994590759, + "learning_rate": 0.00014051333396503901, + "loss": 0.9906, + "step": 8264 + }, + { + "epoch": 1.4715099715099715, + "grad_norm": 0.7120978832244873, + "learning_rate": 0.00014050053627274488, + "loss": 1.1074, + "step": 8265 + }, + { + "epoch": 1.471688034188034, + "grad_norm": 0.6778998374938965, + "learning_rate": 0.0001404877377869269, + "loss": 1.0027, + "step": 8266 + }, + { + "epoch": 1.4718660968660968, + "grad_norm": 0.6832901239395142, + "learning_rate": 0.0001404749385078358, + "loss": 0.9399, + "step": 8267 + }, + { + "epoch": 1.4720441595441596, + "grad_norm": 0.7428423762321472, + "learning_rate": 0.00014046213843572236, + "loss": 1.0591, + "step": 8268 + }, + { + "epoch": 1.4722222222222223, + "grad_norm": 0.7522720098495483, + "learning_rate": 0.00014044933757083737, + "loss": 1.1184, + "step": 8269 + }, + { + "epoch": 1.4724002849002849, + "grad_norm": 0.7714734673500061, + "learning_rate": 0.00014043653591343163, + "loss": 1.0783, + "step": 8270 + }, + { + "epoch": 1.4725783475783476, + "grad_norm": 0.5860890746116638, + "learning_rate": 0.00014042373346375597, + "loss": 0.8394, + "step": 8271 + }, + { + "epoch": 1.4727564102564101, + "grad_norm": 0.6400395035743713, + "learning_rate": 0.0001404109302220612, + "loss": 0.9153, + "step": 8272 + }, + { + "epoch": 1.4729344729344729, + "grad_norm": 0.7441139817237854, + "learning_rate": 0.00014039812618859827, + "loss": 0.9224, + "step": 8273 + }, + { + "epoch": 1.4731125356125356, + "grad_norm": 0.6030932664871216, + "learning_rate": 0.00014038532136361793, + "loss": 1.0783, + "step": 8274 + }, + { + "epoch": 1.4732905982905984, + "grad_norm": 0.7243345975875854, + "learning_rate": 0.0001403725157473711, + "loss": 0.9894, + "step": 8275 + }, + { + "epoch": 1.473468660968661, + "grad_norm": 0.6880641579627991, + "learning_rate": 0.0001403597093401087, + "loss": 0.9459, + "step": 8276 + }, + { + "epoch": 1.4736467236467237, + "grad_norm": 0.6263882517814636, + "learning_rate": 0.00014034690214208165, + "loss": 0.8781, + "step": 8277 + }, + { + "epoch": 1.4738247863247862, + "grad_norm": 0.7159495949745178, + "learning_rate": 0.00014033409415354085, + "loss": 1.0511, + "step": 8278 + }, + { + "epoch": 1.474002849002849, + "grad_norm": 0.7182226181030273, + "learning_rate": 0.00014032128537473727, + "loss": 1.1196, + "step": 8279 + }, + { + "epoch": 1.4741809116809117, + "grad_norm": 0.744478166103363, + "learning_rate": 0.00014030847580592186, + "loss": 1.0747, + "step": 8280 + }, + { + "epoch": 1.4743589743589745, + "grad_norm": 0.6806797385215759, + "learning_rate": 0.00014029566544734558, + "loss": 1.1519, + "step": 8281 + }, + { + "epoch": 1.474537037037037, + "grad_norm": 0.6813502311706543, + "learning_rate": 0.00014028285429925946, + "loss": 0.968, + "step": 8282 + }, + { + "epoch": 1.4747150997150997, + "grad_norm": 0.639784574508667, + "learning_rate": 0.00014027004236191452, + "loss": 1.0685, + "step": 8283 + }, + { + "epoch": 1.4748931623931623, + "grad_norm": 0.6325878500938416, + "learning_rate": 0.00014025722963556173, + "loss": 1.0358, + "step": 8284 + }, + { + "epoch": 1.475071225071225, + "grad_norm": 0.7012955546379089, + "learning_rate": 0.00014024441612045215, + "loss": 1.1059, + "step": 8285 + }, + { + "epoch": 1.4752492877492878, + "grad_norm": 0.690380334854126, + "learning_rate": 0.00014023160181683684, + "loss": 0.9628, + "step": 8286 + }, + { + "epoch": 1.4754273504273505, + "grad_norm": 0.7178516983985901, + "learning_rate": 0.00014021878672496686, + "loss": 0.963, + "step": 8287 + }, + { + "epoch": 1.475605413105413, + "grad_norm": 0.7049064636230469, + "learning_rate": 0.0001402059708450933, + "loss": 0.8996, + "step": 8288 + }, + { + "epoch": 1.4757834757834758, + "grad_norm": 0.6777819395065308, + "learning_rate": 0.00014019315417746728, + "loss": 1.0696, + "step": 8289 + }, + { + "epoch": 1.4759615384615383, + "grad_norm": 0.5948763489723206, + "learning_rate": 0.00014018033672233987, + "loss": 0.928, + "step": 8290 + }, + { + "epoch": 1.476139601139601, + "grad_norm": 0.7183942198753357, + "learning_rate": 0.00014016751847996224, + "loss": 1.1053, + "step": 8291 + }, + { + "epoch": 1.4763176638176638, + "grad_norm": 0.7426177263259888, + "learning_rate": 0.00014015469945058556, + "loss": 0.9504, + "step": 8292 + }, + { + "epoch": 1.4764957264957266, + "grad_norm": 0.6508159041404724, + "learning_rate": 0.0001401418796344609, + "loss": 1.1176, + "step": 8293 + }, + { + "epoch": 1.476673789173789, + "grad_norm": 0.6954567432403564, + "learning_rate": 0.00014012905903183954, + "loss": 0.9238, + "step": 8294 + }, + { + "epoch": 1.4768518518518519, + "grad_norm": 0.7023960947990417, + "learning_rate": 0.0001401162376429726, + "loss": 1.2032, + "step": 8295 + }, + { + "epoch": 1.4770299145299146, + "grad_norm": 0.7174739837646484, + "learning_rate": 0.00014010341546811134, + "loss": 0.9385, + "step": 8296 + }, + { + "epoch": 1.4772079772079771, + "grad_norm": 0.611980140209198, + "learning_rate": 0.00014009059250750695, + "loss": 0.9469, + "step": 8297 + }, + { + "epoch": 1.47738603988604, + "grad_norm": 0.6362917423248291, + "learning_rate": 0.0001400777687614107, + "loss": 1.1406, + "step": 8298 + }, + { + "epoch": 1.4775641025641026, + "grad_norm": 0.6884697675704956, + "learning_rate": 0.00014006494423007381, + "loss": 0.7915, + "step": 8299 + }, + { + "epoch": 1.4777421652421652, + "grad_norm": 0.6266025304794312, + "learning_rate": 0.00014005211891374755, + "loss": 0.94, + "step": 8300 + }, + { + "epoch": 1.477920227920228, + "grad_norm": 0.6130280494689941, + "learning_rate": 0.00014003929281268323, + "loss": 0.9369, + "step": 8301 + }, + { + "epoch": 1.4780982905982907, + "grad_norm": 0.7244207859039307, + "learning_rate": 0.00014002646592713215, + "loss": 1.1449, + "step": 8302 + }, + { + "epoch": 1.4782763532763532, + "grad_norm": 0.6527345776557922, + "learning_rate": 0.0001400136382573456, + "loss": 0.7792, + "step": 8303 + }, + { + "epoch": 1.478454415954416, + "grad_norm": 0.7102689743041992, + "learning_rate": 0.00014000080980357496, + "loss": 0.9577, + "step": 8304 + }, + { + "epoch": 1.4786324786324787, + "grad_norm": 0.6179325580596924, + "learning_rate": 0.00013998798056607154, + "loss": 0.827, + "step": 8305 + }, + { + "epoch": 1.4788105413105412, + "grad_norm": 0.761234700679779, + "learning_rate": 0.00013997515054508668, + "loss": 1.0576, + "step": 8306 + }, + { + "epoch": 1.478988603988604, + "grad_norm": 0.6200914978981018, + "learning_rate": 0.0001399623197408718, + "loss": 1.0514, + "step": 8307 + }, + { + "epoch": 1.4791666666666667, + "grad_norm": 0.5961193442344666, + "learning_rate": 0.0001399494881536783, + "loss": 0.7846, + "step": 8308 + }, + { + "epoch": 1.4793447293447293, + "grad_norm": 0.645984411239624, + "learning_rate": 0.00013993665578375758, + "loss": 0.9927, + "step": 8309 + }, + { + "epoch": 1.479522792022792, + "grad_norm": 0.7258989810943604, + "learning_rate": 0.000139923822631361, + "loss": 0.7567, + "step": 8310 + }, + { + "epoch": 1.4797008547008548, + "grad_norm": 0.708882212638855, + "learning_rate": 0.00013991098869674007, + "loss": 1.1147, + "step": 8311 + }, + { + "epoch": 1.4798789173789173, + "grad_norm": 0.669262707233429, + "learning_rate": 0.00013989815398014624, + "loss": 0.7142, + "step": 8312 + }, + { + "epoch": 1.48005698005698, + "grad_norm": 0.7398767471313477, + "learning_rate": 0.00013988531848183096, + "loss": 1.043, + "step": 8313 + }, + { + "epoch": 1.4802350427350428, + "grad_norm": 0.753197193145752, + "learning_rate": 0.0001398724822020457, + "loss": 1.058, + "step": 8314 + }, + { + "epoch": 1.4804131054131053, + "grad_norm": 0.663526177406311, + "learning_rate": 0.000139859645141042, + "loss": 1.1272, + "step": 8315 + }, + { + "epoch": 1.480591168091168, + "grad_norm": 0.6537514925003052, + "learning_rate": 0.00013984680729907135, + "loss": 1.011, + "step": 8316 + }, + { + "epoch": 1.4807692307692308, + "grad_norm": 0.707554817199707, + "learning_rate": 0.00013983396867638527, + "loss": 1.0593, + "step": 8317 + }, + { + "epoch": 1.4809472934472934, + "grad_norm": 0.6261475086212158, + "learning_rate": 0.00013982112927323533, + "loss": 1.0731, + "step": 8318 + }, + { + "epoch": 1.4811253561253561, + "grad_norm": 0.6694258451461792, + "learning_rate": 0.00013980828908987308, + "loss": 1.0703, + "step": 8319 + }, + { + "epoch": 1.4813034188034189, + "grad_norm": 0.7793164253234863, + "learning_rate": 0.00013979544812655012, + "loss": 1.0447, + "step": 8320 + }, + { + "epoch": 1.4814814814814814, + "grad_norm": 0.6496448516845703, + "learning_rate": 0.00013978260638351802, + "loss": 1.0208, + "step": 8321 + }, + { + "epoch": 1.4816595441595442, + "grad_norm": 0.5992059111595154, + "learning_rate": 0.00013976976386102834, + "loss": 0.9717, + "step": 8322 + }, + { + "epoch": 1.481837606837607, + "grad_norm": 0.7473567128181458, + "learning_rate": 0.0001397569205593328, + "loss": 0.9612, + "step": 8323 + }, + { + "epoch": 1.4820156695156697, + "grad_norm": 0.657558798789978, + "learning_rate": 0.00013974407647868297, + "loss": 1.2137, + "step": 8324 + }, + { + "epoch": 1.4821937321937322, + "grad_norm": 0.7040614485740662, + "learning_rate": 0.00013973123161933055, + "loss": 1.007, + "step": 8325 + }, + { + "epoch": 1.482371794871795, + "grad_norm": 0.6098681092262268, + "learning_rate": 0.00013971838598152717, + "loss": 1.0595, + "step": 8326 + }, + { + "epoch": 1.4825498575498575, + "grad_norm": 0.7194869518280029, + "learning_rate": 0.0001397055395655245, + "loss": 0.9632, + "step": 8327 + }, + { + "epoch": 1.4827279202279202, + "grad_norm": 0.645972728729248, + "learning_rate": 0.00013969269237157426, + "loss": 1.0712, + "step": 8328 + }, + { + "epoch": 1.482905982905983, + "grad_norm": 0.6580560207366943, + "learning_rate": 0.0001396798443999282, + "loss": 1.2117, + "step": 8329 + }, + { + "epoch": 1.4830840455840457, + "grad_norm": 0.6624418497085571, + "learning_rate": 0.00013966699565083802, + "loss": 0.8529, + "step": 8330 + }, + { + "epoch": 1.4832621082621082, + "grad_norm": 0.659896731376648, + "learning_rate": 0.00013965414612455545, + "loss": 0.9359, + "step": 8331 + }, + { + "epoch": 1.483440170940171, + "grad_norm": 0.6690883636474609, + "learning_rate": 0.00013964129582133222, + "loss": 0.971, + "step": 8332 + }, + { + "epoch": 1.4836182336182335, + "grad_norm": 0.6767334938049316, + "learning_rate": 0.00013962844474142022, + "loss": 1.0137, + "step": 8333 + }, + { + "epoch": 1.4837962962962963, + "grad_norm": 0.6412752270698547, + "learning_rate": 0.0001396155928850711, + "loss": 1.2812, + "step": 8334 + }, + { + "epoch": 1.483974358974359, + "grad_norm": 0.6731469035148621, + "learning_rate": 0.0001396027402525368, + "loss": 0.8723, + "step": 8335 + }, + { + "epoch": 1.4841524216524218, + "grad_norm": 0.7327923774719238, + "learning_rate": 0.000139589886844069, + "loss": 0.9606, + "step": 8336 + }, + { + "epoch": 1.4843304843304843, + "grad_norm": 0.6194515824317932, + "learning_rate": 0.00013957703265991963, + "loss": 0.8514, + "step": 8337 + }, + { + "epoch": 1.484508547008547, + "grad_norm": 0.7250012755393982, + "learning_rate": 0.00013956417770034053, + "loss": 0.9755, + "step": 8338 + }, + { + "epoch": 1.4846866096866096, + "grad_norm": 0.7484263181686401, + "learning_rate": 0.00013955132196558358, + "loss": 1.0376, + "step": 8339 + }, + { + "epoch": 1.4848646723646723, + "grad_norm": 0.7593362331390381, + "learning_rate": 0.00013953846545590058, + "loss": 1.3011, + "step": 8340 + }, + { + "epoch": 1.485042735042735, + "grad_norm": 0.6670466065406799, + "learning_rate": 0.00013952560817154352, + "loss": 0.9726, + "step": 8341 + }, + { + "epoch": 1.4852207977207978, + "grad_norm": 0.8001134395599365, + "learning_rate": 0.00013951275011276425, + "loss": 1.1447, + "step": 8342 + }, + { + "epoch": 1.4853988603988604, + "grad_norm": 0.741450309753418, + "learning_rate": 0.00013949989127981475, + "loss": 1.1101, + "step": 8343 + }, + { + "epoch": 1.4855769230769231, + "grad_norm": 0.6594467163085938, + "learning_rate": 0.00013948703167294694, + "loss": 1.0205, + "step": 8344 + }, + { + "epoch": 1.4857549857549857, + "grad_norm": 0.6303030252456665, + "learning_rate": 0.00013947417129241276, + "loss": 0.9179, + "step": 8345 + }, + { + "epoch": 1.4859330484330484, + "grad_norm": 0.6352720856666565, + "learning_rate": 0.00013946131013846418, + "loss": 1.158, + "step": 8346 + }, + { + "epoch": 1.4861111111111112, + "grad_norm": 0.6720923781394958, + "learning_rate": 0.0001394484482113532, + "loss": 0.8805, + "step": 8347 + }, + { + "epoch": 1.486289173789174, + "grad_norm": 0.7186421751976013, + "learning_rate": 0.00013943558551133186, + "loss": 0.8951, + "step": 8348 + }, + { + "epoch": 1.4864672364672364, + "grad_norm": 0.6038698554039001, + "learning_rate": 0.00013942272203865214, + "loss": 1.0079, + "step": 8349 + }, + { + "epoch": 1.4866452991452992, + "grad_norm": 0.665790319442749, + "learning_rate": 0.00013940985779356606, + "loss": 0.8853, + "step": 8350 + }, + { + "epoch": 1.4868233618233617, + "grad_norm": 0.6941595673561096, + "learning_rate": 0.00013939699277632568, + "loss": 1.1404, + "step": 8351 + }, + { + "epoch": 1.4870014245014245, + "grad_norm": 0.7943871021270752, + "learning_rate": 0.00013938412698718305, + "loss": 0.9961, + "step": 8352 + }, + { + "epoch": 1.4871794871794872, + "grad_norm": 0.6363818645477295, + "learning_rate": 0.00013937126042639028, + "loss": 0.8621, + "step": 8353 + }, + { + "epoch": 1.48735754985755, + "grad_norm": 0.7986421585083008, + "learning_rate": 0.00013935839309419943, + "loss": 1.0547, + "step": 8354 + }, + { + "epoch": 1.4875356125356125, + "grad_norm": 0.5890130400657654, + "learning_rate": 0.00013934552499086266, + "loss": 0.9863, + "step": 8355 + }, + { + "epoch": 1.4877136752136753, + "grad_norm": 0.7915370464324951, + "learning_rate": 0.00013933265611663207, + "loss": 1.0385, + "step": 8356 + }, + { + "epoch": 1.4878917378917378, + "grad_norm": 0.7062503695487976, + "learning_rate": 0.00013931978647175973, + "loss": 1.0984, + "step": 8357 + }, + { + "epoch": 1.4880698005698005, + "grad_norm": 0.6496769785881042, + "learning_rate": 0.00013930691605649792, + "loss": 1.0884, + "step": 8358 + }, + { + "epoch": 1.4882478632478633, + "grad_norm": 0.6527266502380371, + "learning_rate": 0.0001392940448710987, + "loss": 1.0366, + "step": 8359 + }, + { + "epoch": 1.488425925925926, + "grad_norm": 0.6269870400428772, + "learning_rate": 0.00013928117291581431, + "loss": 0.9097, + "step": 8360 + }, + { + "epoch": 1.4886039886039886, + "grad_norm": 0.6581160426139832, + "learning_rate": 0.00013926830019089694, + "loss": 0.8694, + "step": 8361 + }, + { + "epoch": 1.4887820512820513, + "grad_norm": 0.6196219325065613, + "learning_rate": 0.0001392554266965988, + "loss": 0.8054, + "step": 8362 + }, + { + "epoch": 1.4889601139601139, + "grad_norm": 0.6246176362037659, + "learning_rate": 0.0001392425524331721, + "loss": 0.9309, + "step": 8363 + }, + { + "epoch": 1.4891381766381766, + "grad_norm": 0.7293874025344849, + "learning_rate": 0.00013922967740086914, + "loss": 1.051, + "step": 8364 + }, + { + "epoch": 1.4893162393162394, + "grad_norm": 0.6581604480743408, + "learning_rate": 0.00013921680159994213, + "loss": 0.8475, + "step": 8365 + }, + { + "epoch": 1.489494301994302, + "grad_norm": 0.6294612288475037, + "learning_rate": 0.00013920392503064335, + "loss": 0.6946, + "step": 8366 + }, + { + "epoch": 1.4896723646723646, + "grad_norm": 0.5725370645523071, + "learning_rate": 0.00013919104769322512, + "loss": 0.7838, + "step": 8367 + }, + { + "epoch": 1.4898504273504274, + "grad_norm": 0.681520402431488, + "learning_rate": 0.00013917816958793967, + "loss": 0.99, + "step": 8368 + }, + { + "epoch": 1.49002849002849, + "grad_norm": 0.6660219430923462, + "learning_rate": 0.00013916529071503943, + "loss": 0.9113, + "step": 8369 + }, + { + "epoch": 1.4902065527065527, + "grad_norm": 0.7567862272262573, + "learning_rate": 0.00013915241107477665, + "loss": 1.2498, + "step": 8370 + }, + { + "epoch": 1.4903846153846154, + "grad_norm": 0.7366036176681519, + "learning_rate": 0.00013913953066740372, + "loss": 1.115, + "step": 8371 + }, + { + "epoch": 1.4905626780626782, + "grad_norm": 0.6201434135437012, + "learning_rate": 0.00013912664949317297, + "loss": 0.8447, + "step": 8372 + }, + { + "epoch": 1.4907407407407407, + "grad_norm": 0.7618655562400818, + "learning_rate": 0.00013911376755233683, + "loss": 0.9696, + "step": 8373 + }, + { + "epoch": 1.4909188034188035, + "grad_norm": 0.6716726422309875, + "learning_rate": 0.00013910088484514764, + "loss": 0.9753, + "step": 8374 + }, + { + "epoch": 1.491096866096866, + "grad_norm": 0.6745659112930298, + "learning_rate": 0.0001390880013718579, + "loss": 1.134, + "step": 8375 + }, + { + "epoch": 1.4912749287749287, + "grad_norm": 0.7524410486221313, + "learning_rate": 0.0001390751171327199, + "loss": 1.0235, + "step": 8376 + }, + { + "epoch": 1.4914529914529915, + "grad_norm": 0.7409411072731018, + "learning_rate": 0.00013906223212798615, + "loss": 0.752, + "step": 8377 + }, + { + "epoch": 1.4916310541310542, + "grad_norm": 0.7016384601593018, + "learning_rate": 0.00013904934635790913, + "loss": 1.1712, + "step": 8378 + }, + { + "epoch": 1.4918091168091168, + "grad_norm": 0.6537824869155884, + "learning_rate": 0.00013903645982274129, + "loss": 1.1162, + "step": 8379 + }, + { + "epoch": 1.4919871794871795, + "grad_norm": 0.6460806727409363, + "learning_rate": 0.0001390235725227351, + "loss": 0.9389, + "step": 8380 + }, + { + "epoch": 1.492165242165242, + "grad_norm": 0.6405501365661621, + "learning_rate": 0.0001390106844581431, + "loss": 1.0508, + "step": 8381 + }, + { + "epoch": 1.4923433048433048, + "grad_norm": 0.6672594547271729, + "learning_rate": 0.00013899779562921775, + "loss": 1.0018, + "step": 8382 + }, + { + "epoch": 1.4925213675213675, + "grad_norm": 0.6303185820579529, + "learning_rate": 0.0001389849060362116, + "loss": 0.9964, + "step": 8383 + }, + { + "epoch": 1.4926994301994303, + "grad_norm": 0.6981508731842041, + "learning_rate": 0.00013897201567937719, + "loss": 1.174, + "step": 8384 + }, + { + "epoch": 1.4928774928774928, + "grad_norm": 0.6195989847183228, + "learning_rate": 0.0001389591245589671, + "loss": 0.9254, + "step": 8385 + }, + { + "epoch": 1.4930555555555556, + "grad_norm": 0.6232163310050964, + "learning_rate": 0.00013894623267523393, + "loss": 0.7151, + "step": 8386 + }, + { + "epoch": 1.493233618233618, + "grad_norm": 0.673067033290863, + "learning_rate": 0.0001389333400284302, + "loss": 1.0156, + "step": 8387 + }, + { + "epoch": 1.4934116809116809, + "grad_norm": 0.706266462802887, + "learning_rate": 0.00013892044661880856, + "loss": 0.9387, + "step": 8388 + }, + { + "epoch": 1.4935897435897436, + "grad_norm": 0.742640495300293, + "learning_rate": 0.00013890755244662161, + "loss": 1.1597, + "step": 8389 + }, + { + "epoch": 1.4937678062678064, + "grad_norm": 0.6856846809387207, + "learning_rate": 0.000138894657512122, + "loss": 0.9998, + "step": 8390 + }, + { + "epoch": 1.493945868945869, + "grad_norm": 0.7214110493659973, + "learning_rate": 0.0001388817618155624, + "loss": 1.1867, + "step": 8391 + }, + { + "epoch": 1.4941239316239316, + "grad_norm": 0.7346787452697754, + "learning_rate": 0.0001388688653571954, + "loss": 0.9071, + "step": 8392 + }, + { + "epoch": 1.4943019943019942, + "grad_norm": 0.7019181847572327, + "learning_rate": 0.00013885596813727373, + "loss": 1.0472, + "step": 8393 + }, + { + "epoch": 1.494480056980057, + "grad_norm": 0.6780814528465271, + "learning_rate": 0.00013884307015605012, + "loss": 1.0031, + "step": 8394 + }, + { + "epoch": 1.4946581196581197, + "grad_norm": 0.6722873449325562, + "learning_rate": 0.0001388301714137772, + "loss": 0.8889, + "step": 8395 + }, + { + "epoch": 1.4948361823361824, + "grad_norm": 0.6736134886741638, + "learning_rate": 0.00013881727191070777, + "loss": 0.8695, + "step": 8396 + }, + { + "epoch": 1.495014245014245, + "grad_norm": 0.632648766040802, + "learning_rate": 0.00013880437164709452, + "loss": 0.9391, + "step": 8397 + }, + { + "epoch": 1.4951923076923077, + "grad_norm": 0.7004299163818359, + "learning_rate": 0.0001387914706231902, + "loss": 1.1423, + "step": 8398 + }, + { + "epoch": 1.4953703703703702, + "grad_norm": 0.5787134766578674, + "learning_rate": 0.0001387785688392476, + "loss": 0.9953, + "step": 8399 + }, + { + "epoch": 1.495548433048433, + "grad_norm": 0.6671785712242126, + "learning_rate": 0.0001387656662955195, + "loss": 0.9356, + "step": 8400 + }, + { + "epoch": 1.4957264957264957, + "grad_norm": 0.7216096520423889, + "learning_rate": 0.0001387527629922587, + "loss": 0.9065, + "step": 8401 + }, + { + "epoch": 1.4959045584045585, + "grad_norm": 0.6469849348068237, + "learning_rate": 0.00013873985892971801, + "loss": 1.0664, + "step": 8402 + }, + { + "epoch": 1.496082621082621, + "grad_norm": 0.5598217248916626, + "learning_rate": 0.00013872695410815027, + "loss": 0.8834, + "step": 8403 + }, + { + "epoch": 1.4962606837606838, + "grad_norm": 0.6860302686691284, + "learning_rate": 0.00013871404852780828, + "loss": 0.9061, + "step": 8404 + }, + { + "epoch": 1.4964387464387463, + "grad_norm": 0.7101688385009766, + "learning_rate": 0.00013870114218894497, + "loss": 1.0236, + "step": 8405 + }, + { + "epoch": 1.496616809116809, + "grad_norm": 0.6494225859642029, + "learning_rate": 0.00013868823509181313, + "loss": 0.9631, + "step": 8406 + }, + { + "epoch": 1.4967948717948718, + "grad_norm": 0.6804189085960388, + "learning_rate": 0.00013867532723666574, + "loss": 0.9341, + "step": 8407 + }, + { + "epoch": 1.4969729344729346, + "grad_norm": 0.8493942022323608, + "learning_rate": 0.00013866241862375562, + "loss": 1.1451, + "step": 8408 + }, + { + "epoch": 1.497150997150997, + "grad_norm": 0.6248497366905212, + "learning_rate": 0.00013864950925333576, + "loss": 0.8584, + "step": 8409 + }, + { + "epoch": 1.4973290598290598, + "grad_norm": 0.6238769292831421, + "learning_rate": 0.00013863659912565903, + "loss": 1.1612, + "step": 8410 + }, + { + "epoch": 1.4975071225071226, + "grad_norm": 0.8538609147071838, + "learning_rate": 0.0001386236882409784, + "loss": 1.0817, + "step": 8411 + }, + { + "epoch": 1.4976851851851851, + "grad_norm": 0.7301406264305115, + "learning_rate": 0.00013861077659954683, + "loss": 0.943, + "step": 8412 + }, + { + "epoch": 1.4978632478632479, + "grad_norm": 0.6573456525802612, + "learning_rate": 0.0001385978642016173, + "loss": 1.0154, + "step": 8413 + }, + { + "epoch": 1.4980413105413106, + "grad_norm": 0.7634185552597046, + "learning_rate": 0.0001385849510474428, + "loss": 1.0432, + "step": 8414 + }, + { + "epoch": 1.4982193732193732, + "grad_norm": 0.6156686544418335, + "learning_rate": 0.00013857203713727633, + "loss": 1.0442, + "step": 8415 + }, + { + "epoch": 1.498397435897436, + "grad_norm": 0.5386871695518494, + "learning_rate": 0.00013855912247137092, + "loss": 0.9055, + "step": 8416 + }, + { + "epoch": 1.4985754985754987, + "grad_norm": 0.7108574509620667, + "learning_rate": 0.00013854620704997962, + "loss": 0.9705, + "step": 8417 + }, + { + "epoch": 1.4987535612535612, + "grad_norm": 0.7313347458839417, + "learning_rate": 0.00013853329087335547, + "loss": 0.7541, + "step": 8418 + }, + { + "epoch": 1.498931623931624, + "grad_norm": 0.8369119167327881, + "learning_rate": 0.0001385203739417515, + "loss": 1.1317, + "step": 8419 + }, + { + "epoch": 1.4991096866096867, + "grad_norm": 0.6763789057731628, + "learning_rate": 0.00013850745625542085, + "loss": 0.7909, + "step": 8420 + }, + { + "epoch": 1.4992877492877492, + "grad_norm": 0.7369635105133057, + "learning_rate": 0.00013849453781461656, + "loss": 1.1454, + "step": 8421 + }, + { + "epoch": 1.499465811965812, + "grad_norm": 0.7165971398353577, + "learning_rate": 0.0001384816186195918, + "loss": 1.1927, + "step": 8422 + }, + { + "epoch": 1.4996438746438747, + "grad_norm": 0.7502337694168091, + "learning_rate": 0.00013846869867059966, + "loss": 1.0592, + "step": 8423 + }, + { + "epoch": 1.4998219373219372, + "grad_norm": 0.7207813858985901, + "learning_rate": 0.00013845577796789326, + "loss": 1.1133, + "step": 8424 + }, + { + "epoch": 1.4998219373219372, + "eval_loss": 1.1057652235031128, + "eval_runtime": 24.7975, + "eval_samples_per_second": 41.98, + "eval_steps_per_second": 21.01, + "step": 8424 + }, + { + "epoch": 1.5, + "grad_norm": 0.6962727308273315, + "learning_rate": 0.00013844285651172576, + "loss": 1.0711, + "step": 8425 + }, + { + "epoch": 1.5001780626780628, + "grad_norm": 0.6585133075714111, + "learning_rate": 0.00013842993430235038, + "loss": 0.9793, + "step": 8426 + }, + { + "epoch": 1.5003561253561255, + "grad_norm": 0.7045056819915771, + "learning_rate": 0.00013841701134002029, + "loss": 1.0046, + "step": 8427 + }, + { + "epoch": 1.500534188034188, + "grad_norm": 0.6788702011108398, + "learning_rate": 0.00013840408762498863, + "loss": 0.9539, + "step": 8428 + }, + { + "epoch": 1.5007122507122506, + "grad_norm": 0.7253114581108093, + "learning_rate": 0.00013839116315750863, + "loss": 0.9446, + "step": 8429 + }, + { + "epoch": 1.5008903133903133, + "grad_norm": 0.6103765368461609, + "learning_rate": 0.0001383782379378336, + "loss": 0.7862, + "step": 8430 + }, + { + "epoch": 1.501068376068376, + "grad_norm": 0.6662353873252869, + "learning_rate": 0.00013836531196621666, + "loss": 1.2178, + "step": 8431 + }, + { + "epoch": 1.5012464387464388, + "grad_norm": 0.6871803998947144, + "learning_rate": 0.00013835238524291117, + "loss": 0.9263, + "step": 8432 + }, + { + "epoch": 1.5014245014245016, + "grad_norm": 0.62713223695755, + "learning_rate": 0.00013833945776817034, + "loss": 0.8879, + "step": 8433 + }, + { + "epoch": 1.501602564102564, + "grad_norm": 0.6698164343833923, + "learning_rate": 0.00013832652954224748, + "loss": 0.9847, + "step": 8434 + }, + { + "epoch": 1.5017806267806266, + "grad_norm": 0.6855883002281189, + "learning_rate": 0.0001383136005653959, + "loss": 0.8614, + "step": 8435 + }, + { + "epoch": 1.5019586894586894, + "grad_norm": 0.7028802037239075, + "learning_rate": 0.0001383006708378689, + "loss": 1.0153, + "step": 8436 + }, + { + "epoch": 1.5021367521367521, + "grad_norm": 0.6710380911827087, + "learning_rate": 0.00013828774035991981, + "loss": 1.0163, + "step": 8437 + }, + { + "epoch": 1.5023148148148149, + "grad_norm": 0.618984580039978, + "learning_rate": 0.000138274809131802, + "loss": 1.0015, + "step": 8438 + }, + { + "epoch": 1.5024928774928776, + "grad_norm": 0.6881645321846008, + "learning_rate": 0.00013826187715376882, + "loss": 0.9776, + "step": 8439 + }, + { + "epoch": 1.5026709401709402, + "grad_norm": 0.6715859770774841, + "learning_rate": 0.00013824894442607358, + "loss": 0.9129, + "step": 8440 + }, + { + "epoch": 1.5028490028490027, + "grad_norm": 0.5940943360328674, + "learning_rate": 0.0001382360109489698, + "loss": 1.0724, + "step": 8441 + }, + { + "epoch": 1.5030270655270654, + "grad_norm": 0.6536458134651184, + "learning_rate": 0.0001382230767227108, + "loss": 1.0162, + "step": 8442 + }, + { + "epoch": 1.5032051282051282, + "grad_norm": 0.6163156628608704, + "learning_rate": 0.00013821014174755, + "loss": 1.0521, + "step": 8443 + }, + { + "epoch": 1.503383190883191, + "grad_norm": 0.7592282891273499, + "learning_rate": 0.00013819720602374082, + "loss": 0.9525, + "step": 8444 + }, + { + "epoch": 1.5035612535612537, + "grad_norm": 0.6672595143318176, + "learning_rate": 0.0001381842695515368, + "loss": 0.9359, + "step": 8445 + }, + { + "epoch": 1.5037393162393162, + "grad_norm": 0.6395034193992615, + "learning_rate": 0.0001381713323311913, + "loss": 1.166, + "step": 8446 + }, + { + "epoch": 1.5039173789173788, + "grad_norm": 0.5958148837089539, + "learning_rate": 0.00013815839436295783, + "loss": 0.9885, + "step": 8447 + }, + { + "epoch": 1.5040954415954415, + "grad_norm": 0.676555871963501, + "learning_rate": 0.0001381454556470899, + "loss": 1.0637, + "step": 8448 + }, + { + "epoch": 1.5042735042735043, + "grad_norm": 0.642428994178772, + "learning_rate": 0.00013813251618384102, + "loss": 0.9288, + "step": 8449 + }, + { + "epoch": 1.504451566951567, + "grad_norm": 0.6730920076370239, + "learning_rate": 0.00013811957597346467, + "loss": 1.1345, + "step": 8450 + }, + { + "epoch": 1.5046296296296298, + "grad_norm": 0.7824259996414185, + "learning_rate": 0.00013810663501621443, + "loss": 0.7532, + "step": 8451 + }, + { + "epoch": 1.5048076923076923, + "grad_norm": 0.8184825778007507, + "learning_rate": 0.00013809369331234386, + "loss": 1.2674, + "step": 8452 + }, + { + "epoch": 1.5049857549857548, + "grad_norm": 0.7369286417961121, + "learning_rate": 0.00013808075086210647, + "loss": 1.0978, + "step": 8453 + }, + { + "epoch": 1.5051638176638176, + "grad_norm": 0.6336679458618164, + "learning_rate": 0.00013806780766575588, + "loss": 1.0922, + "step": 8454 + }, + { + "epoch": 1.5053418803418803, + "grad_norm": 0.700219452381134, + "learning_rate": 0.0001380548637235457, + "loss": 1.0908, + "step": 8455 + }, + { + "epoch": 1.505519943019943, + "grad_norm": 0.6346127986907959, + "learning_rate": 0.0001380419190357295, + "loss": 1.1265, + "step": 8456 + }, + { + "epoch": 1.5056980056980058, + "grad_norm": 0.8653196096420288, + "learning_rate": 0.00013802897360256093, + "loss": 1.0466, + "step": 8457 + }, + { + "epoch": 1.5058760683760684, + "grad_norm": 0.6589069962501526, + "learning_rate": 0.0001380160274242936, + "loss": 1.245, + "step": 8458 + }, + { + "epoch": 1.506054131054131, + "grad_norm": 0.6527602076530457, + "learning_rate": 0.00013800308050118117, + "loss": 1.1539, + "step": 8459 + }, + { + "epoch": 1.5062321937321936, + "grad_norm": 0.6005436182022095, + "learning_rate": 0.00013799013283347734, + "loss": 0.899, + "step": 8460 + }, + { + "epoch": 1.5064102564102564, + "grad_norm": 0.6954274773597717, + "learning_rate": 0.0001379771844214358, + "loss": 1.1245, + "step": 8461 + }, + { + "epoch": 1.5065883190883191, + "grad_norm": 0.658764660358429, + "learning_rate": 0.00013796423526531019, + "loss": 0.9884, + "step": 8462 + }, + { + "epoch": 1.506766381766382, + "grad_norm": 0.652214527130127, + "learning_rate": 0.0001379512853653543, + "loss": 0.9711, + "step": 8463 + }, + { + "epoch": 1.5069444444444444, + "grad_norm": 0.5680044889450073, + "learning_rate": 0.00013793833472182176, + "loss": 0.9055, + "step": 8464 + }, + { + "epoch": 1.5071225071225072, + "grad_norm": 0.7524166703224182, + "learning_rate": 0.0001379253833349664, + "loss": 1.1163, + "step": 8465 + }, + { + "epoch": 1.5073005698005697, + "grad_norm": 0.692936897277832, + "learning_rate": 0.0001379124312050419, + "loss": 0.899, + "step": 8466 + }, + { + "epoch": 1.5074786324786325, + "grad_norm": 0.6871617436408997, + "learning_rate": 0.00013789947833230207, + "loss": 0.9416, + "step": 8467 + }, + { + "epoch": 1.5076566951566952, + "grad_norm": 0.5983462333679199, + "learning_rate": 0.0001378865247170007, + "loss": 0.9776, + "step": 8468 + }, + { + "epoch": 1.507834757834758, + "grad_norm": 0.6486790180206299, + "learning_rate": 0.0001378735703593916, + "loss": 0.9346, + "step": 8469 + }, + { + "epoch": 1.5080128205128205, + "grad_norm": 0.6843809485435486, + "learning_rate": 0.00013786061525972857, + "loss": 1.1276, + "step": 8470 + }, + { + "epoch": 1.5081908831908832, + "grad_norm": 0.5734516382217407, + "learning_rate": 0.00013784765941826538, + "loss": 0.6939, + "step": 8471 + }, + { + "epoch": 1.5083689458689458, + "grad_norm": 0.6126381754875183, + "learning_rate": 0.00013783470283525596, + "loss": 0.8609, + "step": 8472 + }, + { + "epoch": 1.5085470085470085, + "grad_norm": 0.7570928335189819, + "learning_rate": 0.00013782174551095415, + "loss": 0.8809, + "step": 8473 + }, + { + "epoch": 1.5087250712250713, + "grad_norm": 0.6911360025405884, + "learning_rate": 0.00013780878744561377, + "loss": 0.9916, + "step": 8474 + }, + { + "epoch": 1.508903133903134, + "grad_norm": 0.6651954650878906, + "learning_rate": 0.00013779582863948878, + "loss": 1.0012, + "step": 8475 + }, + { + "epoch": 1.5090811965811965, + "grad_norm": 0.845396876335144, + "learning_rate": 0.000137782869092833, + "loss": 0.8455, + "step": 8476 + }, + { + "epoch": 1.5092592592592593, + "grad_norm": 0.6958050727844238, + "learning_rate": 0.00013776990880590042, + "loss": 1.0264, + "step": 8477 + }, + { + "epoch": 1.5094373219373218, + "grad_norm": 0.6950124502182007, + "learning_rate": 0.00013775694777894493, + "loss": 1.0547, + "step": 8478 + }, + { + "epoch": 1.5096153846153846, + "grad_norm": 0.7243088483810425, + "learning_rate": 0.00013774398601222045, + "loss": 1.0999, + "step": 8479 + }, + { + "epoch": 1.5097934472934473, + "grad_norm": 0.6820448040962219, + "learning_rate": 0.00013773102350598097, + "loss": 0.823, + "step": 8480 + }, + { + "epoch": 1.50997150997151, + "grad_norm": 0.689996063709259, + "learning_rate": 0.0001377180602604805, + "loss": 1.049, + "step": 8481 + }, + { + "epoch": 1.5101495726495726, + "grad_norm": 0.6763314604759216, + "learning_rate": 0.000137705096275973, + "loss": 0.9633, + "step": 8482 + }, + { + "epoch": 1.5103276353276354, + "grad_norm": 0.6760517358779907, + "learning_rate": 0.00013769213155271243, + "loss": 1.0326, + "step": 8483 + }, + { + "epoch": 1.510505698005698, + "grad_norm": 0.7181188464164734, + "learning_rate": 0.00013767916609095285, + "loss": 0.9629, + "step": 8484 + }, + { + "epoch": 1.5106837606837606, + "grad_norm": 0.7102212905883789, + "learning_rate": 0.0001376661998909483, + "loss": 1.2714, + "step": 8485 + }, + { + "epoch": 1.5108618233618234, + "grad_norm": 0.6719805598258972, + "learning_rate": 0.00013765323295295278, + "loss": 0.7848, + "step": 8486 + }, + { + "epoch": 1.5110398860398861, + "grad_norm": 0.6592095494270325, + "learning_rate": 0.0001376402652772204, + "loss": 0.882, + "step": 8487 + }, + { + "epoch": 1.5112179487179487, + "grad_norm": 0.6858693361282349, + "learning_rate": 0.00013762729686400522, + "loss": 0.9418, + "step": 8488 + }, + { + "epoch": 1.5113960113960114, + "grad_norm": 0.7183199524879456, + "learning_rate": 0.0001376143277135613, + "loss": 1.0611, + "step": 8489 + }, + { + "epoch": 1.511574074074074, + "grad_norm": 0.6294263005256653, + "learning_rate": 0.00013760135782614277, + "loss": 0.864, + "step": 8490 + }, + { + "epoch": 1.5117521367521367, + "grad_norm": 0.6762619614601135, + "learning_rate": 0.00013758838720200376, + "loss": 1.0295, + "step": 8491 + }, + { + "epoch": 1.5119301994301995, + "grad_norm": 0.6919726133346558, + "learning_rate": 0.00013757541584139834, + "loss": 1.0803, + "step": 8492 + }, + { + "epoch": 1.5121082621082622, + "grad_norm": 0.6801241040229797, + "learning_rate": 0.00013756244374458075, + "loss": 1.1394, + "step": 8493 + }, + { + "epoch": 1.5122863247863247, + "grad_norm": 0.6758754253387451, + "learning_rate": 0.0001375494709118051, + "loss": 1.0053, + "step": 8494 + }, + { + "epoch": 1.5124643874643875, + "grad_norm": 0.6727001070976257, + "learning_rate": 0.00013753649734332555, + "loss": 1.1407, + "step": 8495 + }, + { + "epoch": 1.51264245014245, + "grad_norm": 0.693913459777832, + "learning_rate": 0.00013752352303939632, + "loss": 1.1804, + "step": 8496 + }, + { + "epoch": 1.5128205128205128, + "grad_norm": 0.6122510433197021, + "learning_rate": 0.0001375105480002716, + "loss": 0.917, + "step": 8497 + }, + { + "epoch": 1.5129985754985755, + "grad_norm": 0.6305009722709656, + "learning_rate": 0.00013749757222620562, + "loss": 1.1075, + "step": 8498 + }, + { + "epoch": 1.5131766381766383, + "grad_norm": 0.7249642610549927, + "learning_rate": 0.0001374845957174526, + "loss": 0.9107, + "step": 8499 + }, + { + "epoch": 1.5133547008547008, + "grad_norm": 0.6922136545181274, + "learning_rate": 0.0001374716184742668, + "loss": 0.9974, + "step": 8500 + }, + { + "epoch": 1.5135327635327636, + "grad_norm": 0.6989904046058655, + "learning_rate": 0.00013745864049690245, + "loss": 0.9866, + "step": 8501 + }, + { + "epoch": 1.513710826210826, + "grad_norm": 0.6284058094024658, + "learning_rate": 0.0001374456617856139, + "loss": 0.8658, + "step": 8502 + }, + { + "epoch": 1.5138888888888888, + "grad_norm": 0.615388810634613, + "learning_rate": 0.00013743268234065535, + "loss": 0.7876, + "step": 8503 + }, + { + "epoch": 1.5140669515669516, + "grad_norm": 0.6212600469589233, + "learning_rate": 0.0001374197021622812, + "loss": 0.855, + "step": 8504 + }, + { + "epoch": 1.5142450142450143, + "grad_norm": 0.6312419772148132, + "learning_rate": 0.00013740672125074567, + "loss": 0.9252, + "step": 8505 + }, + { + "epoch": 1.5144230769230769, + "grad_norm": 0.7094576954841614, + "learning_rate": 0.00013739373960630315, + "loss": 0.7655, + "step": 8506 + }, + { + "epoch": 1.5146011396011396, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.000137380757229208, + "loss": 0.7855, + "step": 8507 + }, + { + "epoch": 1.5147792022792022, + "grad_norm": 0.6798399686813354, + "learning_rate": 0.00013736777411971457, + "loss": 0.9935, + "step": 8508 + }, + { + "epoch": 1.514957264957265, + "grad_norm": 0.7835991978645325, + "learning_rate": 0.00013735479027807723, + "loss": 1.1603, + "step": 8509 + }, + { + "epoch": 1.5151353276353277, + "grad_norm": 0.6230790615081787, + "learning_rate": 0.00013734180570455033, + "loss": 1.1463, + "step": 8510 + }, + { + "epoch": 1.5153133903133904, + "grad_norm": 0.646603524684906, + "learning_rate": 0.00013732882039938835, + "loss": 0.9564, + "step": 8511 + }, + { + "epoch": 1.515491452991453, + "grad_norm": 0.6619647145271301, + "learning_rate": 0.0001373158343628457, + "loss": 0.8492, + "step": 8512 + }, + { + "epoch": 1.5156695156695157, + "grad_norm": 0.6458454132080078, + "learning_rate": 0.00013730284759517675, + "loss": 1.0049, + "step": 8513 + }, + { + "epoch": 1.5158475783475782, + "grad_norm": 0.7415743470191956, + "learning_rate": 0.00013728986009663602, + "loss": 0.872, + "step": 8514 + }, + { + "epoch": 1.516025641025641, + "grad_norm": 0.6198840141296387, + "learning_rate": 0.00013727687186747793, + "loss": 0.8645, + "step": 8515 + }, + { + "epoch": 1.5162037037037037, + "grad_norm": 0.7160853147506714, + "learning_rate": 0.00013726388290795697, + "loss": 1.0144, + "step": 8516 + }, + { + "epoch": 1.5163817663817665, + "grad_norm": 0.6604135632514954, + "learning_rate": 0.00013725089321832765, + "loss": 0.9827, + "step": 8517 + }, + { + "epoch": 1.5165598290598292, + "grad_norm": 0.6480790972709656, + "learning_rate": 0.00013723790279884443, + "loss": 1.0357, + "step": 8518 + }, + { + "epoch": 1.5167378917378918, + "grad_norm": 0.6207128167152405, + "learning_rate": 0.00013722491164976187, + "loss": 0.9467, + "step": 8519 + }, + { + "epoch": 1.5169159544159543, + "grad_norm": 0.6024298667907715, + "learning_rate": 0.00013721191977133452, + "loss": 0.8821, + "step": 8520 + }, + { + "epoch": 1.517094017094017, + "grad_norm": 0.684898316860199, + "learning_rate": 0.00013719892716381688, + "loss": 0.9823, + "step": 8521 + }, + { + "epoch": 1.5172720797720798, + "grad_norm": 0.7460635304450989, + "learning_rate": 0.00013718593382746355, + "loss": 1.2573, + "step": 8522 + }, + { + "epoch": 1.5174501424501425, + "grad_norm": 0.7193243503570557, + "learning_rate": 0.00013717293976252907, + "loss": 1.0162, + "step": 8523 + }, + { + "epoch": 1.5176282051282053, + "grad_norm": 0.6328752040863037, + "learning_rate": 0.0001371599449692681, + "loss": 0.8183, + "step": 8524 + }, + { + "epoch": 1.5178062678062678, + "grad_norm": 0.658784806728363, + "learning_rate": 0.00013714694944793517, + "loss": 0.9315, + "step": 8525 + }, + { + "epoch": 1.5179843304843303, + "grad_norm": 0.7875827550888062, + "learning_rate": 0.00013713395319878493, + "loss": 1.0889, + "step": 8526 + }, + { + "epoch": 1.518162393162393, + "grad_norm": 0.6580079793930054, + "learning_rate": 0.00013712095622207203, + "loss": 1.0276, + "step": 8527 + }, + { + "epoch": 1.5183404558404558, + "grad_norm": 0.6214027404785156, + "learning_rate": 0.00013710795851805106, + "loss": 0.9692, + "step": 8528 + }, + { + "epoch": 1.5185185185185186, + "grad_norm": 0.7839403748512268, + "learning_rate": 0.0001370949600869768, + "loss": 0.7378, + "step": 8529 + }, + { + "epoch": 1.5186965811965814, + "grad_norm": 0.6632764339447021, + "learning_rate": 0.0001370819609291038, + "loss": 0.9431, + "step": 8530 + }, + { + "epoch": 1.5188746438746439, + "grad_norm": 0.7071712017059326, + "learning_rate": 0.00013706896104468682, + "loss": 0.7684, + "step": 8531 + }, + { + "epoch": 1.5190527065527064, + "grad_norm": 0.7494829297065735, + "learning_rate": 0.00013705596043398058, + "loss": 0.9709, + "step": 8532 + }, + { + "epoch": 1.5192307692307692, + "grad_norm": 0.6408106088638306, + "learning_rate": 0.00013704295909723973, + "loss": 0.8494, + "step": 8533 + }, + { + "epoch": 1.519408831908832, + "grad_norm": 0.6043150424957275, + "learning_rate": 0.0001370299570347191, + "loss": 0.7485, + "step": 8534 + }, + { + "epoch": 1.5195868945868947, + "grad_norm": 0.6944992542266846, + "learning_rate": 0.00013701695424667336, + "loss": 0.8403, + "step": 8535 + }, + { + "epoch": 1.5197649572649574, + "grad_norm": 0.7730217576026917, + "learning_rate": 0.00013700395073335726, + "loss": 0.9122, + "step": 8536 + }, + { + "epoch": 1.51994301994302, + "grad_norm": 0.6300255060195923, + "learning_rate": 0.00013699094649502564, + "loss": 0.9185, + "step": 8537 + }, + { + "epoch": 1.5201210826210825, + "grad_norm": 0.648676335811615, + "learning_rate": 0.00013697794153193327, + "loss": 0.9897, + "step": 8538 + }, + { + "epoch": 1.5202991452991452, + "grad_norm": 0.7365788817405701, + "learning_rate": 0.00013696493584433494, + "loss": 0.958, + "step": 8539 + }, + { + "epoch": 1.520477207977208, + "grad_norm": 0.6634557247161865, + "learning_rate": 0.00013695192943248552, + "loss": 0.9389, + "step": 8540 + }, + { + "epoch": 1.5206552706552707, + "grad_norm": 0.6110827922821045, + "learning_rate": 0.00013693892229663977, + "loss": 0.9341, + "step": 8541 + }, + { + "epoch": 1.5208333333333335, + "grad_norm": 0.7207275032997131, + "learning_rate": 0.00013692591443705256, + "loss": 0.9526, + "step": 8542 + }, + { + "epoch": 1.521011396011396, + "grad_norm": 0.7071022391319275, + "learning_rate": 0.0001369129058539788, + "loss": 0.9572, + "step": 8543 + }, + { + "epoch": 1.5211894586894585, + "grad_norm": 0.5898227691650391, + "learning_rate": 0.0001368998965476733, + "loss": 0.921, + "step": 8544 + }, + { + "epoch": 1.5213675213675213, + "grad_norm": 0.7542559504508972, + "learning_rate": 0.000136886886518391, + "loss": 0.7799, + "step": 8545 + }, + { + "epoch": 1.521545584045584, + "grad_norm": 0.6904959678649902, + "learning_rate": 0.00013687387576638674, + "loss": 0.9601, + "step": 8546 + }, + { + "epoch": 1.5217236467236468, + "grad_norm": 0.763414204120636, + "learning_rate": 0.00013686086429191553, + "loss": 1.0046, + "step": 8547 + }, + { + "epoch": 1.5219017094017095, + "grad_norm": 0.6879960298538208, + "learning_rate": 0.00013684785209523224, + "loss": 0.9615, + "step": 8548 + }, + { + "epoch": 1.522079772079772, + "grad_norm": 0.7166057229042053, + "learning_rate": 0.00013683483917659186, + "loss": 0.9481, + "step": 8549 + }, + { + "epoch": 1.5222578347578346, + "grad_norm": 0.6384348273277283, + "learning_rate": 0.0001368218255362493, + "loss": 1.1037, + "step": 8550 + }, + { + "epoch": 1.5224358974358974, + "grad_norm": 0.6564528346061707, + "learning_rate": 0.00013680881117445953, + "loss": 0.951, + "step": 8551 + }, + { + "epoch": 1.52261396011396, + "grad_norm": 0.749301016330719, + "learning_rate": 0.00013679579609147762, + "loss": 0.9324, + "step": 8552 + }, + { + "epoch": 1.5227920227920229, + "grad_norm": 0.8130472898483276, + "learning_rate": 0.00013678278028755848, + "loss": 1.0178, + "step": 8553 + }, + { + "epoch": 1.5229700854700856, + "grad_norm": 0.6763297319412231, + "learning_rate": 0.0001367697637629572, + "loss": 0.9224, + "step": 8554 + }, + { + "epoch": 1.5231481481481481, + "grad_norm": 0.6630885601043701, + "learning_rate": 0.00013675674651792878, + "loss": 1.0254, + "step": 8555 + }, + { + "epoch": 1.5233262108262107, + "grad_norm": 0.7377206087112427, + "learning_rate": 0.00013674372855272825, + "loss": 1.0413, + "step": 8556 + }, + { + "epoch": 1.5235042735042734, + "grad_norm": 0.5270320177078247, + "learning_rate": 0.00013673070986761068, + "loss": 0.7124, + "step": 8557 + }, + { + "epoch": 1.5236823361823362, + "grad_norm": 0.5941976308822632, + "learning_rate": 0.00013671769046283116, + "loss": 1.0281, + "step": 8558 + }, + { + "epoch": 1.523860398860399, + "grad_norm": 0.6131376028060913, + "learning_rate": 0.0001367046703386448, + "loss": 0.7593, + "step": 8559 + }, + { + "epoch": 1.5240384615384617, + "grad_norm": 0.7381763458251953, + "learning_rate": 0.00013669164949530664, + "loss": 1.148, + "step": 8560 + }, + { + "epoch": 1.5242165242165242, + "grad_norm": 0.683274507522583, + "learning_rate": 0.00013667862793307185, + "loss": 0.8354, + "step": 8561 + }, + { + "epoch": 1.5243945868945867, + "grad_norm": 0.6912649273872375, + "learning_rate": 0.0001366656056521955, + "loss": 0.9043, + "step": 8562 + }, + { + "epoch": 1.5245726495726495, + "grad_norm": 0.5999594330787659, + "learning_rate": 0.0001366525826529328, + "loss": 0.6138, + "step": 8563 + }, + { + "epoch": 1.5247507122507122, + "grad_norm": 0.7185927629470825, + "learning_rate": 0.00013663955893553892, + "loss": 0.895, + "step": 8564 + }, + { + "epoch": 1.524928774928775, + "grad_norm": 0.5967002511024475, + "learning_rate": 0.00013662653450026893, + "loss": 0.9636, + "step": 8565 + }, + { + "epoch": 1.5251068376068377, + "grad_norm": 0.7122953534126282, + "learning_rate": 0.00013661350934737813, + "loss": 0.9465, + "step": 8566 + }, + { + "epoch": 1.5252849002849003, + "grad_norm": 0.705326497554779, + "learning_rate": 0.00013660048347712163, + "loss": 1.121, + "step": 8567 + }, + { + "epoch": 1.5254629629629628, + "grad_norm": 0.6023733019828796, + "learning_rate": 0.0001365874568897547, + "loss": 0.9881, + "step": 8568 + }, + { + "epoch": 1.5256410256410255, + "grad_norm": 0.6883122324943542, + "learning_rate": 0.0001365744295855326, + "loss": 1.2372, + "step": 8569 + }, + { + "epoch": 1.5258190883190883, + "grad_norm": 0.718126654624939, + "learning_rate": 0.0001365614015647105, + "loss": 1.0888, + "step": 8570 + }, + { + "epoch": 1.525997150997151, + "grad_norm": 0.6649243831634521, + "learning_rate": 0.00013654837282754367, + "loss": 1.0458, + "step": 8571 + }, + { + "epoch": 1.5261752136752138, + "grad_norm": 0.6959797143936157, + "learning_rate": 0.00013653534337428738, + "loss": 0.9282, + "step": 8572 + }, + { + "epoch": 1.5263532763532763, + "grad_norm": 0.6069976687431335, + "learning_rate": 0.00013652231320519697, + "loss": 0.9706, + "step": 8573 + }, + { + "epoch": 1.526531339031339, + "grad_norm": 0.7085374593734741, + "learning_rate": 0.0001365092823205277, + "loss": 1.1241, + "step": 8574 + }, + { + "epoch": 1.5267094017094016, + "grad_norm": 0.575106143951416, + "learning_rate": 0.00013649625072053488, + "loss": 0.9814, + "step": 8575 + }, + { + "epoch": 1.5268874643874644, + "grad_norm": 0.6541273593902588, + "learning_rate": 0.00013648321840547384, + "loss": 1.0155, + "step": 8576 + }, + { + "epoch": 1.5270655270655271, + "grad_norm": 0.6754382848739624, + "learning_rate": 0.0001364701853755999, + "loss": 1.0284, + "step": 8577 + }, + { + "epoch": 1.5272435897435899, + "grad_norm": 0.6219634413719177, + "learning_rate": 0.00013645715163116846, + "loss": 1.1539, + "step": 8578 + }, + { + "epoch": 1.5274216524216524, + "grad_norm": 0.7625157833099365, + "learning_rate": 0.00013644411717243486, + "loss": 1.1157, + "step": 8579 + }, + { + "epoch": 1.5275997150997151, + "grad_norm": 0.6944296956062317, + "learning_rate": 0.0001364310819996545, + "loss": 0.8309, + "step": 8580 + }, + { + "epoch": 1.5277777777777777, + "grad_norm": 0.7198494672775269, + "learning_rate": 0.00013641804611308277, + "loss": 1.0883, + "step": 8581 + }, + { + "epoch": 1.5279558404558404, + "grad_norm": 0.6398822069168091, + "learning_rate": 0.00013640500951297508, + "loss": 1.0173, + "step": 8582 + }, + { + "epoch": 1.5281339031339032, + "grad_norm": 0.7306683659553528, + "learning_rate": 0.00013639197219958682, + "loss": 0.9979, + "step": 8583 + }, + { + "epoch": 1.528311965811966, + "grad_norm": 0.6873512268066406, + "learning_rate": 0.00013637893417317348, + "loss": 0.7883, + "step": 8584 + }, + { + "epoch": 1.5284900284900285, + "grad_norm": 0.6482085585594177, + "learning_rate": 0.00013636589543399052, + "loss": 0.9367, + "step": 8585 + }, + { + "epoch": 1.5286680911680912, + "grad_norm": 0.8161232471466064, + "learning_rate": 0.00013635285598229336, + "loss": 1.0582, + "step": 8586 + }, + { + "epoch": 1.5288461538461537, + "grad_norm": 0.6722155809402466, + "learning_rate": 0.0001363398158183375, + "loss": 0.9805, + "step": 8587 + }, + { + "epoch": 1.5290242165242165, + "grad_norm": 0.7175397872924805, + "learning_rate": 0.00013632677494237845, + "loss": 1.0747, + "step": 8588 + }, + { + "epoch": 1.5292022792022792, + "grad_norm": 0.6665592789649963, + "learning_rate": 0.00013631373335467172, + "loss": 1.006, + "step": 8589 + }, + { + "epoch": 1.529380341880342, + "grad_norm": 0.7002299427986145, + "learning_rate": 0.0001363006910554728, + "loss": 1.0702, + "step": 8590 + }, + { + "epoch": 1.5295584045584045, + "grad_norm": 0.7712168097496033, + "learning_rate": 0.00013628764804503725, + "loss": 1.0628, + "step": 8591 + }, + { + "epoch": 1.5297364672364673, + "grad_norm": 0.6620795130729675, + "learning_rate": 0.0001362746043236206, + "loss": 1.01, + "step": 8592 + }, + { + "epoch": 1.5299145299145298, + "grad_norm": 0.6374393701553345, + "learning_rate": 0.00013626155989147846, + "loss": 0.9106, + "step": 8593 + }, + { + "epoch": 1.5300925925925926, + "grad_norm": 0.6531631946563721, + "learning_rate": 0.00013624851474886636, + "loss": 1.0488, + "step": 8594 + }, + { + "epoch": 1.5302706552706553, + "grad_norm": 0.6843775510787964, + "learning_rate": 0.00013623546889603993, + "loss": 0.8599, + "step": 8595 + }, + { + "epoch": 1.530448717948718, + "grad_norm": 0.7232706546783447, + "learning_rate": 0.00013622242233325476, + "loss": 1.0875, + "step": 8596 + }, + { + "epoch": 1.5306267806267806, + "grad_norm": 0.695691704750061, + "learning_rate": 0.00013620937506076644, + "loss": 0.9835, + "step": 8597 + }, + { + "epoch": 1.5308048433048433, + "grad_norm": 0.6321248412132263, + "learning_rate": 0.00013619632707883065, + "loss": 0.9778, + "step": 8598 + }, + { + "epoch": 1.5309829059829059, + "grad_norm": 0.6469168663024902, + "learning_rate": 0.00013618327838770303, + "loss": 0.9968, + "step": 8599 + }, + { + "epoch": 1.5311609686609686, + "grad_norm": 0.6798683404922485, + "learning_rate": 0.00013617022898763925, + "loss": 0.78, + "step": 8600 + }, + { + "epoch": 1.5313390313390314, + "grad_norm": 0.6932336091995239, + "learning_rate": 0.00013615717887889496, + "loss": 0.9473, + "step": 8601 + }, + { + "epoch": 1.5315170940170941, + "grad_norm": 0.7304185628890991, + "learning_rate": 0.00013614412806172585, + "loss": 1.0478, + "step": 8602 + }, + { + "epoch": 1.5316951566951567, + "grad_norm": 0.6585272550582886, + "learning_rate": 0.00013613107653638763, + "loss": 0.8563, + "step": 8603 + }, + { + "epoch": 1.5318732193732194, + "grad_norm": 0.6804470419883728, + "learning_rate": 0.00013611802430313604, + "loss": 0.9839, + "step": 8604 + }, + { + "epoch": 1.532051282051282, + "grad_norm": 0.7271378040313721, + "learning_rate": 0.0001361049713622268, + "loss": 1.0906, + "step": 8605 + }, + { + "epoch": 1.5322293447293447, + "grad_norm": 0.7731603980064392, + "learning_rate": 0.00013609191771391562, + "loss": 1.1318, + "step": 8606 + }, + { + "epoch": 1.5324074074074074, + "grad_norm": 0.6143709421157837, + "learning_rate": 0.0001360788633584583, + "loss": 0.8726, + "step": 8607 + }, + { + "epoch": 1.5325854700854702, + "grad_norm": 0.6847203373908997, + "learning_rate": 0.00013606580829611056, + "loss": 0.9963, + "step": 8608 + }, + { + "epoch": 1.5327635327635327, + "grad_norm": 0.7561219334602356, + "learning_rate": 0.0001360527525271283, + "loss": 0.8873, + "step": 8609 + }, + { + "epoch": 1.5329415954415955, + "grad_norm": 0.7997925281524658, + "learning_rate": 0.0001360396960517672, + "loss": 0.7675, + "step": 8610 + }, + { + "epoch": 1.533119658119658, + "grad_norm": 0.7206357717514038, + "learning_rate": 0.00013602663887028315, + "loss": 1.0084, + "step": 8611 + }, + { + "epoch": 1.5332977207977208, + "grad_norm": 0.6454238891601562, + "learning_rate": 0.00013601358098293194, + "loss": 0.8194, + "step": 8612 + }, + { + "epoch": 1.5334757834757835, + "grad_norm": 0.5531884431838989, + "learning_rate": 0.0001360005223899694, + "loss": 0.8596, + "step": 8613 + }, + { + "epoch": 1.5336538461538463, + "grad_norm": 0.659161388874054, + "learning_rate": 0.00013598746309165144, + "loss": 1.0363, + "step": 8614 + }, + { + "epoch": 1.5338319088319088, + "grad_norm": 0.6958948373794556, + "learning_rate": 0.00013597440308823385, + "loss": 0.9852, + "step": 8615 + }, + { + "epoch": 1.5340099715099715, + "grad_norm": 0.7147171497344971, + "learning_rate": 0.0001359613423799726, + "loss": 1.0506, + "step": 8616 + }, + { + "epoch": 1.534188034188034, + "grad_norm": 0.604450523853302, + "learning_rate": 0.00013594828096712353, + "loss": 0.9344, + "step": 8617 + }, + { + "epoch": 1.5343660968660968, + "grad_norm": 0.714547336101532, + "learning_rate": 0.00013593521884994257, + "loss": 1.1583, + "step": 8618 + }, + { + "epoch": 1.5345441595441596, + "grad_norm": 0.6864442825317383, + "learning_rate": 0.00013592215602868565, + "loss": 0.991, + "step": 8619 + }, + { + "epoch": 1.5347222222222223, + "grad_norm": 0.6384446620941162, + "learning_rate": 0.00013590909250360873, + "loss": 0.8799, + "step": 8620 + }, + { + "epoch": 1.5349002849002849, + "grad_norm": 0.7307949662208557, + "learning_rate": 0.00013589602827496772, + "loss": 1.0276, + "step": 8621 + }, + { + "epoch": 1.5350783475783476, + "grad_norm": 0.6620129942893982, + "learning_rate": 0.00013588296334301862, + "loss": 0.9378, + "step": 8622 + }, + { + "epoch": 1.5352564102564101, + "grad_norm": 0.7216851711273193, + "learning_rate": 0.00013586989770801735, + "loss": 0.8984, + "step": 8623 + }, + { + "epoch": 1.5354344729344729, + "grad_norm": 0.7319885492324829, + "learning_rate": 0.00013585683137022, + "loss": 1.0357, + "step": 8624 + }, + { + "epoch": 1.5356125356125356, + "grad_norm": 0.7455703616142273, + "learning_rate": 0.00013584376432988247, + "loss": 0.9727, + "step": 8625 + }, + { + "epoch": 1.5357905982905984, + "grad_norm": 0.7285277247428894, + "learning_rate": 0.0001358306965872609, + "loss": 1.1132, + "step": 8626 + }, + { + "epoch": 1.535968660968661, + "grad_norm": 0.6250096559524536, + "learning_rate": 0.00013581762814261124, + "loss": 0.8538, + "step": 8627 + }, + { + "epoch": 1.5361467236467237, + "grad_norm": 0.6252279281616211, + "learning_rate": 0.0001358045589961895, + "loss": 0.822, + "step": 8628 + }, + { + "epoch": 1.5363247863247862, + "grad_norm": 0.7723368406295776, + "learning_rate": 0.0001357914891482519, + "loss": 0.9841, + "step": 8629 + }, + { + "epoch": 1.536502849002849, + "grad_norm": 0.6855236887931824, + "learning_rate": 0.00013577841859905435, + "loss": 0.9512, + "step": 8630 + }, + { + "epoch": 1.5366809116809117, + "grad_norm": 0.8320944309234619, + "learning_rate": 0.00013576534734885303, + "loss": 1.0324, + "step": 8631 + }, + { + "epoch": 1.5368589743589745, + "grad_norm": 0.6970052123069763, + "learning_rate": 0.00013575227539790405, + "loss": 0.9874, + "step": 8632 + }, + { + "epoch": 1.5370370370370372, + "grad_norm": 0.7774853110313416, + "learning_rate": 0.00013573920274646345, + "loss": 0.962, + "step": 8633 + }, + { + "epoch": 1.5372150997150997, + "grad_norm": 0.6479182839393616, + "learning_rate": 0.0001357261293947875, + "loss": 0.9438, + "step": 8634 + }, + { + "epoch": 1.5373931623931623, + "grad_norm": 0.6855679750442505, + "learning_rate": 0.00013571305534313218, + "loss": 1.0898, + "step": 8635 + }, + { + "epoch": 1.537571225071225, + "grad_norm": 0.6527835726737976, + "learning_rate": 0.00013569998059175377, + "loss": 0.954, + "step": 8636 + }, + { + "epoch": 1.5377492877492878, + "grad_norm": 0.6601176857948303, + "learning_rate": 0.00013568690514090837, + "loss": 1.0183, + "step": 8637 + }, + { + "epoch": 1.5379273504273505, + "grad_norm": 0.6628120541572571, + "learning_rate": 0.0001356738289908522, + "loss": 1.0651, + "step": 8638 + }, + { + "epoch": 1.5381054131054133, + "grad_norm": 0.7492203712463379, + "learning_rate": 0.00013566075214184147, + "loss": 1.2438, + "step": 8639 + }, + { + "epoch": 1.5382834757834758, + "grad_norm": 0.6781343817710876, + "learning_rate": 0.00013564767459413237, + "loss": 0.9413, + "step": 8640 + }, + { + "epoch": 1.5384615384615383, + "grad_norm": 0.6890891790390015, + "learning_rate": 0.00013563459634798115, + "loss": 0.9912, + "step": 8641 + }, + { + "epoch": 1.538639601139601, + "grad_norm": 0.722820520401001, + "learning_rate": 0.00013562151740364404, + "loss": 1.1799, + "step": 8642 + }, + { + "epoch": 1.5388176638176638, + "grad_norm": 0.738369882106781, + "learning_rate": 0.0001356084377613773, + "loss": 1.1313, + "step": 8643 + }, + { + "epoch": 1.5389957264957266, + "grad_norm": 0.6232718229293823, + "learning_rate": 0.00013559535742143717, + "loss": 0.9035, + "step": 8644 + }, + { + "epoch": 1.5391737891737893, + "grad_norm": 0.7371624708175659, + "learning_rate": 0.00013558227638407996, + "loss": 1.3377, + "step": 8645 + }, + { + "epoch": 1.5393518518518519, + "grad_norm": 0.658353865146637, + "learning_rate": 0.00013556919464956197, + "loss": 0.9591, + "step": 8646 + }, + { + "epoch": 1.5395299145299144, + "grad_norm": 0.6205827593803406, + "learning_rate": 0.0001355561122181395, + "loss": 0.9217, + "step": 8647 + }, + { + "epoch": 1.5397079772079771, + "grad_norm": 0.5892502069473267, + "learning_rate": 0.00013554302909006888, + "loss": 0.8893, + "step": 8648 + }, + { + "epoch": 1.53988603988604, + "grad_norm": 1.224568486213684, + "learning_rate": 0.0001355299452656064, + "loss": 0.8237, + "step": 8649 + }, + { + "epoch": 1.5400641025641026, + "grad_norm": 0.7732635736465454, + "learning_rate": 0.0001355168607450085, + "loss": 1.1043, + "step": 8650 + }, + { + "epoch": 1.5402421652421654, + "grad_norm": 0.6365402340888977, + "learning_rate": 0.00013550377552853146, + "loss": 1.0345, + "step": 8651 + }, + { + "epoch": 1.540420227920228, + "grad_norm": 0.7046400904655457, + "learning_rate": 0.00013549068961643171, + "loss": 1.0361, + "step": 8652 + }, + { + "epoch": 1.5405982905982905, + "grad_norm": 0.6760256886482239, + "learning_rate": 0.0001354776030089656, + "loss": 0.9437, + "step": 8653 + }, + { + "epoch": 1.5407763532763532, + "grad_norm": 0.6180984973907471, + "learning_rate": 0.00013546451570638958, + "loss": 0.9737, + "step": 8654 + }, + { + "epoch": 1.540954415954416, + "grad_norm": 0.6221960186958313, + "learning_rate": 0.00013545142770896005, + "loss": 0.9313, + "step": 8655 + }, + { + "epoch": 1.5411324786324787, + "grad_norm": 0.6887816786766052, + "learning_rate": 0.0001354383390169334, + "loss": 1.1736, + "step": 8656 + }, + { + "epoch": 1.5413105413105415, + "grad_norm": 0.5840606093406677, + "learning_rate": 0.00013542524963056614, + "loss": 0.9269, + "step": 8657 + }, + { + "epoch": 1.541488603988604, + "grad_norm": 0.7396654486656189, + "learning_rate": 0.00013541215955011472, + "loss": 1.1189, + "step": 8658 + }, + { + "epoch": 1.5416666666666665, + "grad_norm": 0.780616819858551, + "learning_rate": 0.00013539906877583555, + "loss": 1.1251, + "step": 8659 + }, + { + "epoch": 1.5418447293447293, + "grad_norm": 0.6975206732749939, + "learning_rate": 0.0001353859773079852, + "loss": 1.2134, + "step": 8660 + }, + { + "epoch": 1.542022792022792, + "grad_norm": 0.7572869658470154, + "learning_rate": 0.00013537288514682013, + "loss": 0.9396, + "step": 8661 + }, + { + "epoch": 1.5422008547008548, + "grad_norm": 0.6252159476280212, + "learning_rate": 0.00013535979229259686, + "loss": 0.8449, + "step": 8662 + }, + { + "epoch": 1.5423789173789175, + "grad_norm": 0.7321650981903076, + "learning_rate": 0.0001353466987455719, + "loss": 1.3263, + "step": 8663 + }, + { + "epoch": 1.54255698005698, + "grad_norm": 0.7168700695037842, + "learning_rate": 0.00013533360450600177, + "loss": 0.8923, + "step": 8664 + }, + { + "epoch": 1.5427350427350426, + "grad_norm": 0.5931934714317322, + "learning_rate": 0.00013532050957414313, + "loss": 0.8448, + "step": 8665 + }, + { + "epoch": 1.5429131054131053, + "grad_norm": 0.6621279120445251, + "learning_rate": 0.00013530741395025245, + "loss": 1.1023, + "step": 8666 + }, + { + "epoch": 1.543091168091168, + "grad_norm": 0.7133732438087463, + "learning_rate": 0.00013529431763458633, + "loss": 0.9986, + "step": 8667 + }, + { + "epoch": 1.5432692307692308, + "grad_norm": 0.7589015960693359, + "learning_rate": 0.0001352812206274014, + "loss": 1.0111, + "step": 8668 + }, + { + "epoch": 1.5434472934472936, + "grad_norm": 0.6958192586898804, + "learning_rate": 0.0001352681229289542, + "loss": 0.9466, + "step": 8669 + }, + { + "epoch": 1.5436253561253561, + "grad_norm": 0.7539750337600708, + "learning_rate": 0.0001352550245395014, + "loss": 1.0974, + "step": 8670 + }, + { + "epoch": 1.5438034188034186, + "grad_norm": 0.7003816366195679, + "learning_rate": 0.00013524192545929964, + "loss": 1.0354, + "step": 8671 + }, + { + "epoch": 1.5439814814814814, + "grad_norm": 0.6503025889396667, + "learning_rate": 0.00013522882568860558, + "loss": 1.0476, + "step": 8672 + }, + { + "epoch": 1.5441595441595442, + "grad_norm": 0.6757345199584961, + "learning_rate": 0.00013521572522767584, + "loss": 0.864, + "step": 8673 + }, + { + "epoch": 1.544337606837607, + "grad_norm": 0.6857611536979675, + "learning_rate": 0.0001352026240767671, + "loss": 1.1627, + "step": 8674 + }, + { + "epoch": 1.5445156695156697, + "grad_norm": 0.5775430798530579, + "learning_rate": 0.0001351895222361361, + "loss": 0.7444, + "step": 8675 + }, + { + "epoch": 1.5446937321937322, + "grad_norm": 0.7511499524116516, + "learning_rate": 0.00013517641970603952, + "loss": 1.1547, + "step": 8676 + }, + { + "epoch": 1.5448717948717947, + "grad_norm": 0.6727504730224609, + "learning_rate": 0.00013516331648673403, + "loss": 1.0829, + "step": 8677 + }, + { + "epoch": 1.5450498575498575, + "grad_norm": 0.6128812432289124, + "learning_rate": 0.00013515021257847642, + "loss": 0.9318, + "step": 8678 + }, + { + "epoch": 1.5452279202279202, + "grad_norm": 0.7309781312942505, + "learning_rate": 0.00013513710798152343, + "loss": 1.0844, + "step": 8679 + }, + { + "epoch": 1.545405982905983, + "grad_norm": 0.695655882358551, + "learning_rate": 0.00013512400269613176, + "loss": 1.113, + "step": 8680 + }, + { + "epoch": 1.5455840455840457, + "grad_norm": 0.696441650390625, + "learning_rate": 0.00013511089672255824, + "loss": 1.0499, + "step": 8681 + }, + { + "epoch": 1.5457621082621082, + "grad_norm": 0.6309961080551147, + "learning_rate": 0.00013509779006105964, + "loss": 0.8759, + "step": 8682 + }, + { + "epoch": 1.5459401709401708, + "grad_norm": 0.6155984401702881, + "learning_rate": 0.00013508468271189277, + "loss": 0.8967, + "step": 8683 + }, + { + "epoch": 1.5461182336182335, + "grad_norm": 0.6786884665489197, + "learning_rate": 0.00013507157467531442, + "loss": 1.0806, + "step": 8684 + }, + { + "epoch": 1.5462962962962963, + "grad_norm": 0.6494075059890747, + "learning_rate": 0.00013505846595158138, + "loss": 1.0196, + "step": 8685 + }, + { + "epoch": 1.546474358974359, + "grad_norm": 0.7599824070930481, + "learning_rate": 0.00013504535654095055, + "loss": 0.8662, + "step": 8686 + }, + { + "epoch": 1.5466524216524218, + "grad_norm": 0.6017210483551025, + "learning_rate": 0.00013503224644367877, + "loss": 0.872, + "step": 8687 + }, + { + "epoch": 1.5468304843304843, + "grad_norm": 0.7972410321235657, + "learning_rate": 0.00013501913566002288, + "loss": 1.0958, + "step": 8688 + }, + { + "epoch": 1.547008547008547, + "grad_norm": 0.7572960257530212, + "learning_rate": 0.00013500602419023978, + "loss": 1.0219, + "step": 8689 + }, + { + "epoch": 1.5471866096866096, + "grad_norm": 0.6329224109649658, + "learning_rate": 0.00013499291203458635, + "loss": 0.8636, + "step": 8690 + }, + { + "epoch": 1.5473646723646723, + "grad_norm": 0.6777113080024719, + "learning_rate": 0.0001349797991933195, + "loss": 1.0297, + "step": 8691 + }, + { + "epoch": 1.547542735042735, + "grad_norm": 0.6449527144432068, + "learning_rate": 0.00013496668566669617, + "loss": 1.0296, + "step": 8692 + }, + { + "epoch": 1.5477207977207978, + "grad_norm": 0.8236973881721497, + "learning_rate": 0.00013495357145497326, + "loss": 0.8569, + "step": 8693 + }, + { + "epoch": 1.5478988603988604, + "grad_norm": 0.6753743290901184, + "learning_rate": 0.0001349404565584077, + "loss": 1.0733, + "step": 8694 + }, + { + "epoch": 1.5480769230769231, + "grad_norm": 0.6642967462539673, + "learning_rate": 0.0001349273409772565, + "loss": 0.9437, + "step": 8695 + }, + { + "epoch": 1.5482549857549857, + "grad_norm": 0.6470823884010315, + "learning_rate": 0.00013491422471177661, + "loss": 0.999, + "step": 8696 + }, + { + "epoch": 1.5484330484330484, + "grad_norm": 0.7287036776542664, + "learning_rate": 0.000134901107762225, + "loss": 0.9396, + "step": 8697 + }, + { + "epoch": 1.5486111111111112, + "grad_norm": 0.6258324980735779, + "learning_rate": 0.00013488799012885872, + "loss": 1.045, + "step": 8698 + }, + { + "epoch": 1.548789173789174, + "grad_norm": 0.6540539860725403, + "learning_rate": 0.00013487487181193473, + "loss": 0.9939, + "step": 8699 + }, + { + "epoch": 1.5489672364672364, + "grad_norm": 0.7129563093185425, + "learning_rate": 0.00013486175281171003, + "loss": 1.2079, + "step": 8700 + }, + { + "epoch": 1.5491452991452992, + "grad_norm": 0.6383145451545715, + "learning_rate": 0.00013484863312844173, + "loss": 0.9999, + "step": 8701 + }, + { + "epoch": 1.5493233618233617, + "grad_norm": 0.6310200691223145, + "learning_rate": 0.0001348355127623869, + "loss": 1.1193, + "step": 8702 + }, + { + "epoch": 1.5495014245014245, + "grad_norm": 0.6370054483413696, + "learning_rate": 0.0001348223917138025, + "loss": 1.0213, + "step": 8703 + }, + { + "epoch": 1.5496794871794872, + "grad_norm": 0.7052688598632812, + "learning_rate": 0.00013480926998294573, + "loss": 0.8773, + "step": 8704 + }, + { + "epoch": 1.54985754985755, + "grad_norm": 0.6369579434394836, + "learning_rate": 0.00013479614757007355, + "loss": 1.0072, + "step": 8705 + }, + { + "epoch": 1.5500356125356125, + "grad_norm": 0.7152075171470642, + "learning_rate": 0.0001347830244754432, + "loss": 1.0409, + "step": 8706 + }, + { + "epoch": 1.5502136752136753, + "grad_norm": 0.654183566570282, + "learning_rate": 0.00013476990069931173, + "loss": 0.9363, + "step": 8707 + }, + { + "epoch": 1.5503917378917378, + "grad_norm": 0.6700537204742432, + "learning_rate": 0.00013475677624193627, + "loss": 0.985, + "step": 8708 + }, + { + "epoch": 1.5505698005698005, + "grad_norm": 0.7195445895195007, + "learning_rate": 0.00013474365110357402, + "loss": 0.988, + "step": 8709 + }, + { + "epoch": 1.5507478632478633, + "grad_norm": 0.6019890904426575, + "learning_rate": 0.00013473052528448201, + "loss": 0.9915, + "step": 8710 + }, + { + "epoch": 1.550925925925926, + "grad_norm": 0.7787565588951111, + "learning_rate": 0.0001347173987849176, + "loss": 0.9676, + "step": 8711 + }, + { + "epoch": 1.5511039886039886, + "grad_norm": 0.6997103691101074, + "learning_rate": 0.00013470427160513782, + "loss": 1.1158, + "step": 8712 + }, + { + "epoch": 1.5512820512820513, + "grad_norm": 0.6259464025497437, + "learning_rate": 0.00013469114374539998, + "loss": 0.8784, + "step": 8713 + }, + { + "epoch": 1.5514601139601139, + "grad_norm": 0.6159056425094604, + "learning_rate": 0.00013467801520596122, + "loss": 0.9184, + "step": 8714 + }, + { + "epoch": 1.5516381766381766, + "grad_norm": 0.6823606491088867, + "learning_rate": 0.00013466488598707876, + "loss": 0.9542, + "step": 8715 + }, + { + "epoch": 1.5518162393162394, + "grad_norm": 0.6781585812568665, + "learning_rate": 0.0001346517560890099, + "loss": 1.1761, + "step": 8716 + }, + { + "epoch": 1.551994301994302, + "grad_norm": 0.6313831806182861, + "learning_rate": 0.00013463862551201184, + "loss": 0.8935, + "step": 8717 + }, + { + "epoch": 1.5521723646723646, + "grad_norm": 0.7466186881065369, + "learning_rate": 0.0001346254942563419, + "loss": 1.0583, + "step": 8718 + }, + { + "epoch": 1.5523504273504274, + "grad_norm": 0.7073680758476257, + "learning_rate": 0.0001346123623222573, + "loss": 0.9863, + "step": 8719 + }, + { + "epoch": 1.55252849002849, + "grad_norm": 0.6286870241165161, + "learning_rate": 0.00013459922971001536, + "loss": 0.9921, + "step": 8720 + }, + { + "epoch": 1.5527065527065527, + "grad_norm": 0.6047035455703735, + "learning_rate": 0.0001345860964198734, + "loss": 0.9155, + "step": 8721 + }, + { + "epoch": 1.5528846153846154, + "grad_norm": 0.5909964442253113, + "learning_rate": 0.00013457296245208874, + "loss": 0.9593, + "step": 8722 + }, + { + "epoch": 1.5530626780626782, + "grad_norm": 0.7838597893714905, + "learning_rate": 0.00013455982780691869, + "loss": 0.8872, + "step": 8723 + }, + { + "epoch": 1.5532407407407407, + "grad_norm": 0.6914706230163574, + "learning_rate": 0.00013454669248462063, + "loss": 0.9104, + "step": 8724 + }, + { + "epoch": 1.5534188034188035, + "grad_norm": 0.6777952909469604, + "learning_rate": 0.00013453355648545182, + "loss": 0.9839, + "step": 8725 + }, + { + "epoch": 1.553596866096866, + "grad_norm": 0.7482799291610718, + "learning_rate": 0.00013452041980966978, + "loss": 1.1164, + "step": 8726 + }, + { + "epoch": 1.5537749287749287, + "grad_norm": 0.6616327166557312, + "learning_rate": 0.0001345072824575318, + "loss": 0.9574, + "step": 8727 + }, + { + "epoch": 1.5539529914529915, + "grad_norm": 0.7193203568458557, + "learning_rate": 0.00013449414442929532, + "loss": 1.0609, + "step": 8728 + }, + { + "epoch": 1.5541310541310542, + "grad_norm": 0.6599446535110474, + "learning_rate": 0.0001344810057252177, + "loss": 0.9574, + "step": 8729 + }, + { + "epoch": 1.5543091168091168, + "grad_norm": 0.7221707105636597, + "learning_rate": 0.00013446786634555642, + "loss": 0.9819, + "step": 8730 + }, + { + "epoch": 1.5544871794871795, + "grad_norm": 0.6531312465667725, + "learning_rate": 0.0001344547262905689, + "loss": 0.9986, + "step": 8731 + }, + { + "epoch": 1.554665242165242, + "grad_norm": 0.6879804730415344, + "learning_rate": 0.0001344415855605126, + "loss": 1.1078, + "step": 8732 + }, + { + "epoch": 1.5548433048433048, + "grad_norm": 0.708907425403595, + "learning_rate": 0.00013442844415564498, + "loss": 1.0221, + "step": 8733 + }, + { + "epoch": 1.5550213675213675, + "grad_norm": 0.7957375645637512, + "learning_rate": 0.0001344153020762235, + "loss": 1.3101, + "step": 8734 + }, + { + "epoch": 1.5551994301994303, + "grad_norm": 0.7068197727203369, + "learning_rate": 0.00013440215932250567, + "loss": 0.8995, + "step": 8735 + }, + { + "epoch": 1.5553774928774928, + "grad_norm": 0.6455841064453125, + "learning_rate": 0.00013438901589474898, + "loss": 0.7244, + "step": 8736 + }, + { + "epoch": 1.5555555555555556, + "grad_norm": 0.7500516772270203, + "learning_rate": 0.00013437587179321097, + "loss": 1.0161, + "step": 8737 + }, + { + "epoch": 1.555733618233618, + "grad_norm": 0.5983143448829651, + "learning_rate": 0.00013436272701814917, + "loss": 0.9922, + "step": 8738 + }, + { + "epoch": 1.5559116809116809, + "grad_norm": 0.8761729598045349, + "learning_rate": 0.0001343495815698211, + "loss": 1.022, + "step": 8739 + }, + { + "epoch": 1.5560897435897436, + "grad_norm": 0.6901857852935791, + "learning_rate": 0.00013433643544848438, + "loss": 1.0668, + "step": 8740 + }, + { + "epoch": 1.5562678062678064, + "grad_norm": 0.6770836114883423, + "learning_rate": 0.00013432328865439647, + "loss": 0.9516, + "step": 8741 + }, + { + "epoch": 1.556445868945869, + "grad_norm": 0.6138805150985718, + "learning_rate": 0.00013431014118781505, + "loss": 0.8682, + "step": 8742 + }, + { + "epoch": 1.5566239316239316, + "grad_norm": 0.6796693801879883, + "learning_rate": 0.00013429699304899772, + "loss": 1.1132, + "step": 8743 + }, + { + "epoch": 1.5568019943019942, + "grad_norm": 0.6626394987106323, + "learning_rate": 0.000134283844238202, + "loss": 0.9273, + "step": 8744 + }, + { + "epoch": 1.556980056980057, + "grad_norm": 0.7088519334793091, + "learning_rate": 0.00013427069475568563, + "loss": 0.8915, + "step": 8745 + }, + { + "epoch": 1.5571581196581197, + "grad_norm": 0.6244857311248779, + "learning_rate": 0.0001342575446017061, + "loss": 0.9466, + "step": 8746 + }, + { + "epoch": 1.5573361823361824, + "grad_norm": 0.6969038248062134, + "learning_rate": 0.00013424439377652123, + "loss": 1.2307, + "step": 8747 + }, + { + "epoch": 1.5575142450142452, + "grad_norm": 0.6636740565299988, + "learning_rate": 0.0001342312422803886, + "loss": 0.9456, + "step": 8748 + }, + { + "epoch": 1.5576923076923077, + "grad_norm": 0.7863389253616333, + "learning_rate": 0.00013421809011356586, + "loss": 1.1888, + "step": 8749 + }, + { + "epoch": 1.5578703703703702, + "grad_norm": 0.7504058480262756, + "learning_rate": 0.00013420493727631073, + "loss": 1.2602, + "step": 8750 + }, + { + "epoch": 1.558048433048433, + "grad_norm": 0.7173139452934265, + "learning_rate": 0.00013419178376888085, + "loss": 1.0726, + "step": 8751 + }, + { + "epoch": 1.5582264957264957, + "grad_norm": 0.6517474055290222, + "learning_rate": 0.00013417862959153406, + "loss": 1.1299, + "step": 8752 + }, + { + "epoch": 1.5584045584045585, + "grad_norm": 0.8911739587783813, + "learning_rate": 0.00013416547474452803, + "loss": 1.105, + "step": 8753 + }, + { + "epoch": 1.5585826210826212, + "grad_norm": 0.7116649150848389, + "learning_rate": 0.00013415231922812049, + "loss": 0.8037, + "step": 8754 + }, + { + "epoch": 1.5587606837606838, + "grad_norm": 0.6935904026031494, + "learning_rate": 0.00013413916304256916, + "loss": 1.2778, + "step": 8755 + }, + { + "epoch": 1.5589387464387463, + "grad_norm": 0.652763843536377, + "learning_rate": 0.00013412600618813186, + "loss": 0.9188, + "step": 8756 + }, + { + "epoch": 1.559116809116809, + "grad_norm": 0.6545276641845703, + "learning_rate": 0.00013411284866506637, + "loss": 1.0116, + "step": 8757 + }, + { + "epoch": 1.5592948717948718, + "grad_norm": 0.632165253162384, + "learning_rate": 0.0001340996904736305, + "loss": 0.8538, + "step": 8758 + }, + { + "epoch": 1.5594729344729346, + "grad_norm": 0.6719664931297302, + "learning_rate": 0.000134086531614082, + "loss": 1.1877, + "step": 8759 + }, + { + "epoch": 1.5596509971509973, + "grad_norm": 0.6691158413887024, + "learning_rate": 0.00013407337208667873, + "loss": 1.0411, + "step": 8760 + }, + { + "epoch": 1.5598290598290598, + "grad_norm": 0.7711479067802429, + "learning_rate": 0.0001340602118916785, + "loss": 0.9995, + "step": 8761 + }, + { + "epoch": 1.5600071225071224, + "grad_norm": 0.7229881286621094, + "learning_rate": 0.0001340470510293392, + "loss": 1.1751, + "step": 8762 + }, + { + "epoch": 1.5601851851851851, + "grad_norm": 0.7183271646499634, + "learning_rate": 0.00013403388949991864, + "loss": 0.9371, + "step": 8763 + }, + { + "epoch": 1.5603632478632479, + "grad_norm": 0.8142383098602295, + "learning_rate": 0.00013402072730367475, + "loss": 1.0199, + "step": 8764 + }, + { + "epoch": 1.5605413105413106, + "grad_norm": 0.6349362134933472, + "learning_rate": 0.00013400756444086534, + "loss": 0.8453, + "step": 8765 + }, + { + "epoch": 1.5607193732193734, + "grad_norm": 0.651900589466095, + "learning_rate": 0.00013399440091174834, + "loss": 0.8952, + "step": 8766 + }, + { + "epoch": 1.560897435897436, + "grad_norm": 0.6873346567153931, + "learning_rate": 0.00013398123671658172, + "loss": 0.9438, + "step": 8767 + }, + { + "epoch": 1.5610754985754984, + "grad_norm": 0.7404754757881165, + "learning_rate": 0.00013396807185562333, + "loss": 1.123, + "step": 8768 + }, + { + "epoch": 1.5612535612535612, + "grad_norm": 0.7449641227722168, + "learning_rate": 0.00013395490632913111, + "loss": 0.9407, + "step": 8769 + }, + { + "epoch": 1.561431623931624, + "grad_norm": 0.7393384575843811, + "learning_rate": 0.0001339417401373631, + "loss": 1.0209, + "step": 8770 + }, + { + "epoch": 1.5616096866096867, + "grad_norm": 0.6787426471710205, + "learning_rate": 0.00013392857328057713, + "loss": 0.9768, + "step": 8771 + }, + { + "epoch": 1.5617877492877494, + "grad_norm": 0.6295693516731262, + "learning_rate": 0.00013391540575903127, + "loss": 0.9011, + "step": 8772 + }, + { + "epoch": 1.561965811965812, + "grad_norm": 0.7114503979682922, + "learning_rate": 0.00013390223757298354, + "loss": 1.0696, + "step": 8773 + }, + { + "epoch": 1.5621438746438745, + "grad_norm": 0.7540110349655151, + "learning_rate": 0.00013388906872269184, + "loss": 1.0071, + "step": 8774 + }, + { + "epoch": 1.5623219373219372, + "grad_norm": 0.6472305059432983, + "learning_rate": 0.00013387589920841423, + "loss": 1.105, + "step": 8775 + }, + { + "epoch": 1.5625, + "grad_norm": 0.6936793327331543, + "learning_rate": 0.00013386272903040874, + "loss": 0.885, + "step": 8776 + }, + { + "epoch": 1.5626780626780628, + "grad_norm": 0.7487989068031311, + "learning_rate": 0.00013384955818893343, + "loss": 0.7842, + "step": 8777 + }, + { + "epoch": 1.5628561253561255, + "grad_norm": 0.6109505891799927, + "learning_rate": 0.00013383638668424633, + "loss": 0.9461, + "step": 8778 + }, + { + "epoch": 1.563034188034188, + "grad_norm": 0.6650055646896362, + "learning_rate": 0.00013382321451660558, + "loss": 1.0463, + "step": 8779 + }, + { + "epoch": 1.5632122507122506, + "grad_norm": 0.7147329449653625, + "learning_rate": 0.00013381004168626915, + "loss": 0.946, + "step": 8780 + }, + { + "epoch": 1.5633903133903133, + "grad_norm": 0.6919382810592651, + "learning_rate": 0.00013379686819349522, + "loss": 0.8946, + "step": 8781 + }, + { + "epoch": 1.563568376068376, + "grad_norm": 0.7339401245117188, + "learning_rate": 0.00013378369403854184, + "loss": 0.9625, + "step": 8782 + }, + { + "epoch": 1.5637464387464388, + "grad_norm": 0.6337129473686218, + "learning_rate": 0.00013377051922166717, + "loss": 1.0854, + "step": 8783 + }, + { + "epoch": 1.5639245014245016, + "grad_norm": 0.7301266193389893, + "learning_rate": 0.0001337573437431293, + "loss": 1.017, + "step": 8784 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 0.689540684223175, + "learning_rate": 0.00013374416760318644, + "loss": 0.8734, + "step": 8785 + }, + { + "epoch": 1.5642806267806266, + "grad_norm": 0.7121307849884033, + "learning_rate": 0.0001337309908020967, + "loss": 1.0827, + "step": 8786 + }, + { + "epoch": 1.5644586894586894, + "grad_norm": 0.6715386509895325, + "learning_rate": 0.00013371781334011826, + "loss": 0.946, + "step": 8787 + }, + { + "epoch": 1.5646367521367521, + "grad_norm": 0.6895501613616943, + "learning_rate": 0.00013370463521750932, + "loss": 1.1113, + "step": 8788 + }, + { + "epoch": 1.5648148148148149, + "grad_norm": 0.6592531204223633, + "learning_rate": 0.00013369145643452805, + "loss": 0.9952, + "step": 8789 + }, + { + "epoch": 1.5649928774928776, + "grad_norm": 0.7495190501213074, + "learning_rate": 0.0001336782769914327, + "loss": 1.0936, + "step": 8790 + }, + { + "epoch": 1.5651709401709402, + "grad_norm": 0.7273977398872375, + "learning_rate": 0.00013366509688848147, + "loss": 1.1749, + "step": 8791 + }, + { + "epoch": 1.5653490028490027, + "grad_norm": 0.6447354555130005, + "learning_rate": 0.0001336519161259326, + "loss": 0.8638, + "step": 8792 + }, + { + "epoch": 1.5655270655270654, + "grad_norm": 0.6572020053863525, + "learning_rate": 0.00013363873470404432, + "loss": 0.8005, + "step": 8793 + }, + { + "epoch": 1.5657051282051282, + "grad_norm": 0.676418662071228, + "learning_rate": 0.00013362555262307491, + "loss": 0.7651, + "step": 8794 + }, + { + "epoch": 1.565883190883191, + "grad_norm": 0.6886745095252991, + "learning_rate": 0.0001336123698832827, + "loss": 1.0765, + "step": 8795 + }, + { + "epoch": 1.5660612535612537, + "grad_norm": 0.8134182095527649, + "learning_rate": 0.00013359918648492584, + "loss": 1.2228, + "step": 8796 + }, + { + "epoch": 1.5662393162393162, + "grad_norm": 0.7210384011268616, + "learning_rate": 0.00013358600242826277, + "loss": 0.8247, + "step": 8797 + }, + { + "epoch": 1.5664173789173788, + "grad_norm": 0.7086136341094971, + "learning_rate": 0.00013357281771355175, + "loss": 1.0323, + "step": 8798 + }, + { + "epoch": 1.5665954415954415, + "grad_norm": 0.7419785857200623, + "learning_rate": 0.0001335596323410511, + "loss": 1.213, + "step": 8799 + }, + { + "epoch": 1.5667735042735043, + "grad_norm": 0.6390291452407837, + "learning_rate": 0.0001335464463110192, + "loss": 1.0403, + "step": 8800 + }, + { + "epoch": 1.566951566951567, + "grad_norm": 0.6111941337585449, + "learning_rate": 0.00013353325962371434, + "loss": 0.9747, + "step": 8801 + }, + { + "epoch": 1.5671296296296298, + "grad_norm": 0.6792671084403992, + "learning_rate": 0.00013352007227939488, + "loss": 1.1179, + "step": 8802 + }, + { + "epoch": 1.5673076923076923, + "grad_norm": 0.6656535863876343, + "learning_rate": 0.0001335068842783193, + "loss": 0.9214, + "step": 8803 + }, + { + "epoch": 1.5674857549857548, + "grad_norm": 0.6910907626152039, + "learning_rate": 0.0001334936956207459, + "loss": 1.0609, + "step": 8804 + }, + { + "epoch": 1.5676638176638176, + "grad_norm": 0.65049147605896, + "learning_rate": 0.00013348050630693315, + "loss": 0.7189, + "step": 8805 + }, + { + "epoch": 1.5678418803418803, + "grad_norm": 0.6258065104484558, + "learning_rate": 0.0001334673163371394, + "loss": 1.0683, + "step": 8806 + }, + { + "epoch": 1.568019943019943, + "grad_norm": 0.7518934607505798, + "learning_rate": 0.00013345412571162305, + "loss": 1.2415, + "step": 8807 + }, + { + "epoch": 1.5681980056980058, + "grad_norm": 0.7395275235176086, + "learning_rate": 0.00013344093443064267, + "loss": 0.9153, + "step": 8808 + }, + { + "epoch": 1.5683760683760684, + "grad_norm": 0.6789839267730713, + "learning_rate": 0.00013342774249445663, + "loss": 0.8051, + "step": 8809 + }, + { + "epoch": 1.568554131054131, + "grad_norm": 0.786247193813324, + "learning_rate": 0.00013341454990332342, + "loss": 1.203, + "step": 8810 + }, + { + "epoch": 1.5687321937321936, + "grad_norm": 0.6858161687850952, + "learning_rate": 0.00013340135665750153, + "loss": 0.9494, + "step": 8811 + }, + { + "epoch": 1.5689102564102564, + "grad_norm": 0.7245797514915466, + "learning_rate": 0.0001333881627572494, + "loss": 1.0544, + "step": 8812 + }, + { + "epoch": 1.5690883190883191, + "grad_norm": 0.6176164150238037, + "learning_rate": 0.00013337496820282563, + "loss": 0.9084, + "step": 8813 + }, + { + "epoch": 1.569266381766382, + "grad_norm": 0.7342953681945801, + "learning_rate": 0.00013336177299448868, + "loss": 1.0006, + "step": 8814 + }, + { + "epoch": 1.5694444444444444, + "grad_norm": 0.5183523297309875, + "learning_rate": 0.00013334857713249708, + "loss": 0.6295, + "step": 8815 + }, + { + "epoch": 1.5696225071225072, + "grad_norm": 0.6664513349533081, + "learning_rate": 0.00013333538061710936, + "loss": 0.7569, + "step": 8816 + }, + { + "epoch": 1.5698005698005697, + "grad_norm": 0.7051160931587219, + "learning_rate": 0.0001333221834485841, + "loss": 0.9917, + "step": 8817 + }, + { + "epoch": 1.5699786324786325, + "grad_norm": 0.7888057231903076, + "learning_rate": 0.0001333089856271799, + "loss": 1.0337, + "step": 8818 + }, + { + "epoch": 1.5701566951566952, + "grad_norm": 0.6796144247055054, + "learning_rate": 0.00013329578715315534, + "loss": 1.0915, + "step": 8819 + }, + { + "epoch": 1.570334757834758, + "grad_norm": 0.7442883849143982, + "learning_rate": 0.000133282588026769, + "loss": 1.1695, + "step": 8820 + }, + { + "epoch": 1.5705128205128205, + "grad_norm": 0.6164735555648804, + "learning_rate": 0.00013326938824827946, + "loss": 1.0143, + "step": 8821 + }, + { + "epoch": 1.5706908831908832, + "grad_norm": 0.6526502966880798, + "learning_rate": 0.00013325618781794539, + "loss": 0.8402, + "step": 8822 + }, + { + "epoch": 1.5708689458689458, + "grad_norm": 0.6376087069511414, + "learning_rate": 0.00013324298673602535, + "loss": 0.7582, + "step": 8823 + }, + { + "epoch": 1.5710470085470085, + "grad_norm": 0.6888708472251892, + "learning_rate": 0.00013322978500277807, + "loss": 0.997, + "step": 8824 + }, + { + "epoch": 1.5712250712250713, + "grad_norm": 0.553656280040741, + "learning_rate": 0.0001332165826184622, + "loss": 0.6917, + "step": 8825 + }, + { + "epoch": 1.571403133903134, + "grad_norm": 0.643285870552063, + "learning_rate": 0.0001332033795833364, + "loss": 0.8689, + "step": 8826 + }, + { + "epoch": 1.5715811965811965, + "grad_norm": 0.6210280060768127, + "learning_rate": 0.00013319017589765933, + "loss": 0.9047, + "step": 8827 + }, + { + "epoch": 1.5717592592592593, + "grad_norm": 0.7612366676330566, + "learning_rate": 0.0001331769715616897, + "loss": 0.9818, + "step": 8828 + }, + { + "epoch": 1.5719373219373218, + "grad_norm": 0.5970702171325684, + "learning_rate": 0.00013316376657568628, + "loss": 0.82, + "step": 8829 + }, + { + "epoch": 1.5721153846153846, + "grad_norm": 0.7182583808898926, + "learning_rate": 0.0001331505609399077, + "loss": 1.0633, + "step": 8830 + }, + { + "epoch": 1.5722934472934473, + "grad_norm": 0.7230739593505859, + "learning_rate": 0.00013313735465461278, + "loss": 0.977, + "step": 8831 + }, + { + "epoch": 1.57247150997151, + "grad_norm": 0.6752985119819641, + "learning_rate": 0.00013312414772006018, + "loss": 0.9666, + "step": 8832 + }, + { + "epoch": 1.5726495726495726, + "grad_norm": 0.7724275588989258, + "learning_rate": 0.00013311094013650877, + "loss": 1.148, + "step": 8833 + }, + { + "epoch": 1.5728276353276354, + "grad_norm": 0.7216386198997498, + "learning_rate": 0.00013309773190421724, + "loss": 0.9935, + "step": 8834 + }, + { + "epoch": 1.573005698005698, + "grad_norm": 0.6422320008277893, + "learning_rate": 0.0001330845230234444, + "loss": 0.9383, + "step": 8835 + }, + { + "epoch": 1.5731837606837606, + "grad_norm": 0.669538140296936, + "learning_rate": 0.00013307131349444906, + "loss": 1.0866, + "step": 8836 + }, + { + "epoch": 1.5733618233618234, + "grad_norm": 0.6994584798812866, + "learning_rate": 0.00013305810331749003, + "loss": 0.7882, + "step": 8837 + }, + { + "epoch": 1.5735398860398861, + "grad_norm": 0.8094269633293152, + "learning_rate": 0.00013304489249282617, + "loss": 1.2316, + "step": 8838 + }, + { + "epoch": 1.5737179487179487, + "grad_norm": 0.7180120348930359, + "learning_rate": 0.00013303168102071625, + "loss": 0.9795, + "step": 8839 + }, + { + "epoch": 1.5738960113960114, + "grad_norm": 0.6191438436508179, + "learning_rate": 0.00013301846890141918, + "loss": 0.8957, + "step": 8840 + }, + { + "epoch": 1.574074074074074, + "grad_norm": 0.671094536781311, + "learning_rate": 0.00013300525613519382, + "loss": 1.059, + "step": 8841 + }, + { + "epoch": 1.5742521367521367, + "grad_norm": 0.8062624931335449, + "learning_rate": 0.000132992042722299, + "loss": 0.9782, + "step": 8842 + }, + { + "epoch": 1.5744301994301995, + "grad_norm": 0.6674807667732239, + "learning_rate": 0.00013297882866299362, + "loss": 0.7765, + "step": 8843 + }, + { + "epoch": 1.5746082621082622, + "grad_norm": 0.6369131803512573, + "learning_rate": 0.00013296561395753664, + "loss": 0.97, + "step": 8844 + }, + { + "epoch": 1.5747863247863247, + "grad_norm": 0.7913636565208435, + "learning_rate": 0.00013295239860618691, + "loss": 1.0458, + "step": 8845 + }, + { + "epoch": 1.5749643874643875, + "grad_norm": 0.6722261905670166, + "learning_rate": 0.0001329391826092034, + "loss": 1.1118, + "step": 8846 + }, + { + "epoch": 1.57514245014245, + "grad_norm": 0.6936299800872803, + "learning_rate": 0.00013292596596684502, + "loss": 1.009, + "step": 8847 + }, + { + "epoch": 1.5753205128205128, + "grad_norm": 0.7009961009025574, + "learning_rate": 0.00013291274867937073, + "loss": 0.9904, + "step": 8848 + }, + { + "epoch": 1.5754985754985755, + "grad_norm": 0.6900732517242432, + "learning_rate": 0.0001328995307470395, + "loss": 1.0488, + "step": 8849 + }, + { + "epoch": 1.5756766381766383, + "grad_norm": 0.6389018297195435, + "learning_rate": 0.00013288631217011032, + "loss": 0.9444, + "step": 8850 + }, + { + "epoch": 1.5758547008547008, + "grad_norm": 0.6370900869369507, + "learning_rate": 0.00013287309294884216, + "loss": 0.7465, + "step": 8851 + }, + { + "epoch": 1.5760327635327636, + "grad_norm": 0.6463848948478699, + "learning_rate": 0.00013285987308349405, + "loss": 0.896, + "step": 8852 + }, + { + "epoch": 1.576210826210826, + "grad_norm": 0.6022449731826782, + "learning_rate": 0.00013284665257432495, + "loss": 0.8822, + "step": 8853 + }, + { + "epoch": 1.5763888888888888, + "grad_norm": 0.768189013004303, + "learning_rate": 0.00013283343142159396, + "loss": 0.9862, + "step": 8854 + }, + { + "epoch": 1.5765669515669516, + "grad_norm": 0.6642358303070068, + "learning_rate": 0.00013282020962556007, + "loss": 1.0713, + "step": 8855 + }, + { + "epoch": 1.5767450142450143, + "grad_norm": 0.6883034706115723, + "learning_rate": 0.00013280698718648234, + "loss": 1.0351, + "step": 8856 + }, + { + "epoch": 1.5769230769230769, + "grad_norm": 0.602808952331543, + "learning_rate": 0.00013279376410461988, + "loss": 0.7615, + "step": 8857 + }, + { + "epoch": 1.5771011396011396, + "grad_norm": 0.5968614220619202, + "learning_rate": 0.0001327805403802317, + "loss": 0.9443, + "step": 8858 + }, + { + "epoch": 1.5772792022792022, + "grad_norm": 0.7314837574958801, + "learning_rate": 0.00013276731601357696, + "loss": 0.8784, + "step": 8859 + }, + { + "epoch": 1.577457264957265, + "grad_norm": 0.619754433631897, + "learning_rate": 0.0001327540910049147, + "loss": 0.954, + "step": 8860 + }, + { + "epoch": 1.5776353276353277, + "grad_norm": 0.7195139527320862, + "learning_rate": 0.0001327408653545041, + "loss": 1.0227, + "step": 8861 + }, + { + "epoch": 1.5778133903133904, + "grad_norm": 0.6796214580535889, + "learning_rate": 0.0001327276390626042, + "loss": 1.0593, + "step": 8862 + }, + { + "epoch": 1.577991452991453, + "grad_norm": 0.6576255559921265, + "learning_rate": 0.00013271441212947427, + "loss": 0.7921, + "step": 8863 + }, + { + "epoch": 1.5781695156695157, + "grad_norm": 0.7222092151641846, + "learning_rate": 0.00013270118455537336, + "loss": 1.0545, + "step": 8864 + }, + { + "epoch": 1.5783475783475782, + "grad_norm": 0.7159737348556519, + "learning_rate": 0.00013268795634056066, + "loss": 0.9664, + "step": 8865 + }, + { + "epoch": 1.578525641025641, + "grad_norm": 0.7120481133460999, + "learning_rate": 0.00013267472748529536, + "loss": 1.0148, + "step": 8866 + }, + { + "epoch": 1.5787037037037037, + "grad_norm": 0.7353253364562988, + "learning_rate": 0.00013266149798983666, + "loss": 0.9288, + "step": 8867 + }, + { + "epoch": 1.5788817663817665, + "grad_norm": 0.6652441620826721, + "learning_rate": 0.00013264826785444375, + "loss": 0.8246, + "step": 8868 + }, + { + "epoch": 1.5790598290598292, + "grad_norm": 0.7254189252853394, + "learning_rate": 0.00013263503707937584, + "loss": 0.9892, + "step": 8869 + }, + { + "epoch": 1.5792378917378918, + "grad_norm": 0.6305747032165527, + "learning_rate": 0.00013262180566489223, + "loss": 0.8931, + "step": 8870 + }, + { + "epoch": 1.5794159544159543, + "grad_norm": 0.6560617089271545, + "learning_rate": 0.00013260857361125205, + "loss": 0.9245, + "step": 8871 + }, + { + "epoch": 1.579594017094017, + "grad_norm": 0.7304151654243469, + "learning_rate": 0.00013259534091871462, + "loss": 1.009, + "step": 8872 + }, + { + "epoch": 1.5797720797720798, + "grad_norm": 0.782636821269989, + "learning_rate": 0.00013258210758753918, + "loss": 1.1123, + "step": 8873 + }, + { + "epoch": 1.5799501424501425, + "grad_norm": 0.6992011070251465, + "learning_rate": 0.00013256887361798504, + "loss": 1.099, + "step": 8874 + }, + { + "epoch": 1.5801282051282053, + "grad_norm": 0.7159731984138489, + "learning_rate": 0.00013255563901031148, + "loss": 1.0257, + "step": 8875 + }, + { + "epoch": 1.5803062678062678, + "grad_norm": 0.6055454611778259, + "learning_rate": 0.0001325424037647778, + "loss": 0.9199, + "step": 8876 + }, + { + "epoch": 1.5804843304843303, + "grad_norm": 0.6838310360908508, + "learning_rate": 0.00013252916788164334, + "loss": 0.8644, + "step": 8877 + }, + { + "epoch": 1.580662393162393, + "grad_norm": 0.7067445516586304, + "learning_rate": 0.00013251593136116738, + "loss": 1.0285, + "step": 8878 + }, + { + "epoch": 1.5808404558404558, + "grad_norm": 0.7021774649620056, + "learning_rate": 0.00013250269420360928, + "loss": 1.1263, + "step": 8879 + }, + { + "epoch": 1.5810185185185186, + "grad_norm": 0.6586757302284241, + "learning_rate": 0.00013248945640922843, + "loss": 0.906, + "step": 8880 + }, + { + "epoch": 1.5811965811965814, + "grad_norm": 0.6673910021781921, + "learning_rate": 0.00013247621797828418, + "loss": 1.0652, + "step": 8881 + }, + { + "epoch": 1.5813746438746439, + "grad_norm": 0.6763964295387268, + "learning_rate": 0.00013246297891103588, + "loss": 1.0227, + "step": 8882 + }, + { + "epoch": 1.5815527065527064, + "grad_norm": 0.6536892056465149, + "learning_rate": 0.00013244973920774298, + "loss": 0.9026, + "step": 8883 + }, + { + "epoch": 1.5817307692307692, + "grad_norm": 0.8010411858558655, + "learning_rate": 0.0001324364988686648, + "loss": 1.1167, + "step": 8884 + }, + { + "epoch": 1.581908831908832, + "grad_norm": 0.8159251809120178, + "learning_rate": 0.00013242325789406082, + "loss": 1.233, + "step": 8885 + }, + { + "epoch": 1.5820868945868947, + "grad_norm": 0.6487745046615601, + "learning_rate": 0.00013241001628419048, + "loss": 0.9888, + "step": 8886 + }, + { + "epoch": 1.5822649572649574, + "grad_norm": 0.6750285029411316, + "learning_rate": 0.00013239677403931318, + "loss": 0.8874, + "step": 8887 + }, + { + "epoch": 1.58244301994302, + "grad_norm": 0.7164602875709534, + "learning_rate": 0.0001323835311596884, + "loss": 1.2029, + "step": 8888 + }, + { + "epoch": 1.5826210826210825, + "grad_norm": 0.6081351041793823, + "learning_rate": 0.00013237028764557558, + "loss": 0.9593, + "step": 8889 + }, + { + "epoch": 1.5827991452991452, + "grad_norm": 0.7235409021377563, + "learning_rate": 0.00013235704349723424, + "loss": 1.5324, + "step": 8890 + }, + { + "epoch": 1.582977207977208, + "grad_norm": 0.6658480763435364, + "learning_rate": 0.0001323437987149238, + "loss": 0.9756, + "step": 8891 + }, + { + "epoch": 1.5831552706552707, + "grad_norm": 0.7924265265464783, + "learning_rate": 0.00013233055329890387, + "loss": 0.9329, + "step": 8892 + }, + { + "epoch": 1.5833333333333335, + "grad_norm": 0.6262093186378479, + "learning_rate": 0.0001323173072494339, + "loss": 0.8288, + "step": 8893 + }, + { + "epoch": 1.583511396011396, + "grad_norm": 0.6851989030838013, + "learning_rate": 0.0001323040605667734, + "loss": 0.9822, + "step": 8894 + }, + { + "epoch": 1.5836894586894585, + "grad_norm": 0.6963728666305542, + "learning_rate": 0.00013229081325118194, + "loss": 1.0416, + "step": 8895 + }, + { + "epoch": 1.5838675213675213, + "grad_norm": 0.6017457842826843, + "learning_rate": 0.0001322775653029191, + "loss": 0.8123, + "step": 8896 + }, + { + "epoch": 1.584045584045584, + "grad_norm": 0.7396472096443176, + "learning_rate": 0.0001322643167222444, + "loss": 1.0339, + "step": 8897 + }, + { + "epoch": 1.5842236467236468, + "grad_norm": 0.6360299587249756, + "learning_rate": 0.00013225106750941744, + "loss": 0.9463, + "step": 8898 + }, + { + "epoch": 1.5844017094017095, + "grad_norm": 0.6297624111175537, + "learning_rate": 0.00013223781766469783, + "loss": 0.9921, + "step": 8899 + }, + { + "epoch": 1.584579772079772, + "grad_norm": 0.7722037434577942, + "learning_rate": 0.0001322245671883451, + "loss": 0.8394, + "step": 8900 + }, + { + "epoch": 1.5847578347578346, + "grad_norm": 0.677364706993103, + "learning_rate": 0.00013221131608061895, + "loss": 1.0954, + "step": 8901 + }, + { + "epoch": 1.5849358974358974, + "grad_norm": 0.6954908967018127, + "learning_rate": 0.00013219806434177899, + "loss": 1.0637, + "step": 8902 + }, + { + "epoch": 1.58511396011396, + "grad_norm": 0.7079192996025085, + "learning_rate": 0.00013218481197208484, + "loss": 1.039, + "step": 8903 + }, + { + "epoch": 1.5852920227920229, + "grad_norm": 0.7070451378822327, + "learning_rate": 0.00013217155897179611, + "loss": 1.0025, + "step": 8904 + }, + { + "epoch": 1.5854700854700856, + "grad_norm": 0.6940776705741882, + "learning_rate": 0.00013215830534117257, + "loss": 0.8039, + "step": 8905 + }, + { + "epoch": 1.5856481481481481, + "grad_norm": 0.6545892953872681, + "learning_rate": 0.00013214505108047382, + "loss": 0.9347, + "step": 8906 + }, + { + "epoch": 1.5858262108262107, + "grad_norm": 0.6769635081291199, + "learning_rate": 0.00013213179618995957, + "loss": 1.0321, + "step": 8907 + }, + { + "epoch": 1.5860042735042734, + "grad_norm": 0.6505448222160339, + "learning_rate": 0.00013211854066988953, + "loss": 1.0558, + "step": 8908 + }, + { + "epoch": 1.5861823361823362, + "grad_norm": 0.6764090061187744, + "learning_rate": 0.00013210528452052336, + "loss": 0.8407, + "step": 8909 + }, + { + "epoch": 1.586360398860399, + "grad_norm": 0.6454851627349854, + "learning_rate": 0.00013209202774212088, + "loss": 0.7439, + "step": 8910 + }, + { + "epoch": 1.5865384615384617, + "grad_norm": 0.6911695599555969, + "learning_rate": 0.00013207877033494177, + "loss": 0.9625, + "step": 8911 + }, + { + "epoch": 1.5867165242165242, + "grad_norm": 0.7405226826667786, + "learning_rate": 0.0001320655122992458, + "loss": 1.054, + "step": 8912 + }, + { + "epoch": 1.5868945868945867, + "grad_norm": 0.7362869381904602, + "learning_rate": 0.00013205225363529274, + "loss": 1.0516, + "step": 8913 + }, + { + "epoch": 1.5870726495726495, + "grad_norm": 0.6923766136169434, + "learning_rate": 0.0001320389943433423, + "loss": 1.2323, + "step": 8914 + }, + { + "epoch": 1.5872507122507122, + "grad_norm": 0.7980395555496216, + "learning_rate": 0.00013202573442365435, + "loss": 1.0229, + "step": 8915 + }, + { + "epoch": 1.587428774928775, + "grad_norm": 0.7211610078811646, + "learning_rate": 0.00013201247387648868, + "loss": 1.0666, + "step": 8916 + }, + { + "epoch": 1.5876068376068377, + "grad_norm": 0.6728795766830444, + "learning_rate": 0.00013199921270210506, + "loss": 1.0322, + "step": 8917 + }, + { + "epoch": 1.5877849002849003, + "grad_norm": 0.6226436495780945, + "learning_rate": 0.00013198595090076337, + "loss": 1.0517, + "step": 8918 + }, + { + "epoch": 1.5879629629629628, + "grad_norm": 0.6396511197090149, + "learning_rate": 0.0001319726884727234, + "loss": 0.8662, + "step": 8919 + }, + { + "epoch": 1.5881410256410255, + "grad_norm": 0.5664374828338623, + "learning_rate": 0.00013195942541824497, + "loss": 0.6601, + "step": 8920 + }, + { + "epoch": 1.5883190883190883, + "grad_norm": 0.6556946039199829, + "learning_rate": 0.00013194616173758806, + "loss": 0.9662, + "step": 8921 + }, + { + "epoch": 1.588497150997151, + "grad_norm": 0.7332060933113098, + "learning_rate": 0.00013193289743101245, + "loss": 0.7687, + "step": 8922 + }, + { + "epoch": 1.5886752136752138, + "grad_norm": 0.6103306412696838, + "learning_rate": 0.00013191963249877805, + "loss": 0.8329, + "step": 8923 + }, + { + "epoch": 1.5888532763532763, + "grad_norm": 0.63165283203125, + "learning_rate": 0.00013190636694114475, + "loss": 0.8336, + "step": 8924 + }, + { + "epoch": 1.589031339031339, + "grad_norm": 0.6955820322036743, + "learning_rate": 0.00013189310075837246, + "loss": 1.0457, + "step": 8925 + }, + { + "epoch": 1.5892094017094016, + "grad_norm": 0.6911605596542358, + "learning_rate": 0.00013187983395072114, + "loss": 0.9389, + "step": 8926 + }, + { + "epoch": 1.5893874643874644, + "grad_norm": 0.6493414640426636, + "learning_rate": 0.00013186656651845068, + "loss": 0.9821, + "step": 8927 + }, + { + "epoch": 1.5895655270655271, + "grad_norm": 0.6168226599693298, + "learning_rate": 0.00013185329846182107, + "loss": 1.0259, + "step": 8928 + }, + { + "epoch": 1.5897435897435899, + "grad_norm": 0.6460188627243042, + "learning_rate": 0.0001318400297810922, + "loss": 0.9836, + "step": 8929 + }, + { + "epoch": 1.5899216524216524, + "grad_norm": 0.6630695462226868, + "learning_rate": 0.0001318267604765241, + "loss": 0.8936, + "step": 8930 + }, + { + "epoch": 1.5900997150997151, + "grad_norm": 0.6308651566505432, + "learning_rate": 0.00013181349054837676, + "loss": 0.9583, + "step": 8931 + }, + { + "epoch": 1.5902777777777777, + "grad_norm": 0.6508499979972839, + "learning_rate": 0.00013180021999691018, + "loss": 0.7647, + "step": 8932 + }, + { + "epoch": 1.5904558404558404, + "grad_norm": 0.6625795960426331, + "learning_rate": 0.00013178694882238432, + "loss": 1.0329, + "step": 8933 + }, + { + "epoch": 1.5906339031339032, + "grad_norm": 0.6721987128257751, + "learning_rate": 0.00013177367702505924, + "loss": 0.9377, + "step": 8934 + }, + { + "epoch": 1.590811965811966, + "grad_norm": 0.7295519709587097, + "learning_rate": 0.00013176040460519497, + "loss": 0.9396, + "step": 8935 + }, + { + "epoch": 1.5909900284900285, + "grad_norm": 0.6673944592475891, + "learning_rate": 0.0001317471315630515, + "loss": 1.0284, + "step": 8936 + }, + { + "epoch": 1.5911680911680912, + "grad_norm": 0.6858960390090942, + "learning_rate": 0.00013173385789888898, + "loss": 1.2022, + "step": 8937 + }, + { + "epoch": 1.5913461538461537, + "grad_norm": 0.5836796164512634, + "learning_rate": 0.00013172058361296743, + "loss": 1.0078, + "step": 8938 + }, + { + "epoch": 1.5915242165242165, + "grad_norm": 0.7732513546943665, + "learning_rate": 0.00013170730870554694, + "loss": 1.0912, + "step": 8939 + }, + { + "epoch": 1.5917022792022792, + "grad_norm": 0.7095892429351807, + "learning_rate": 0.0001316940331768876, + "loss": 1.0506, + "step": 8940 + }, + { + "epoch": 1.591880341880342, + "grad_norm": 0.757534384727478, + "learning_rate": 0.00013168075702724952, + "loss": 1.036, + "step": 8941 + }, + { + "epoch": 1.5920584045584045, + "grad_norm": 0.6719361543655396, + "learning_rate": 0.00013166748025689282, + "loss": 0.9406, + "step": 8942 + }, + { + "epoch": 1.5922364672364673, + "grad_norm": 0.6955735087394714, + "learning_rate": 0.00013165420286607763, + "loss": 0.9325, + "step": 8943 + }, + { + "epoch": 1.5924145299145298, + "grad_norm": 0.6810322999954224, + "learning_rate": 0.00013164092485506407, + "loss": 1.0402, + "step": 8944 + }, + { + "epoch": 1.5925925925925926, + "grad_norm": 0.6346224546432495, + "learning_rate": 0.00013162764622411233, + "loss": 0.9725, + "step": 8945 + }, + { + "epoch": 1.5927706552706553, + "grad_norm": 0.728705883026123, + "learning_rate": 0.00013161436697348258, + "loss": 0.9665, + "step": 8946 + }, + { + "epoch": 1.592948717948718, + "grad_norm": 0.6838595271110535, + "learning_rate": 0.00013160108710343494, + "loss": 0.9771, + "step": 8947 + }, + { + "epoch": 1.5931267806267806, + "grad_norm": 0.7052602767944336, + "learning_rate": 0.00013158780661422966, + "loss": 0.8819, + "step": 8948 + }, + { + "epoch": 1.5933048433048433, + "grad_norm": 0.7237630486488342, + "learning_rate": 0.00013157452550612697, + "loss": 1.0609, + "step": 8949 + }, + { + "epoch": 1.5934829059829059, + "grad_norm": 0.6554936766624451, + "learning_rate": 0.00013156124377938699, + "loss": 0.8592, + "step": 8950 + }, + { + "epoch": 1.5936609686609686, + "grad_norm": 0.6125665307044983, + "learning_rate": 0.00013154796143427, + "loss": 0.8399, + "step": 8951 + }, + { + "epoch": 1.5938390313390314, + "grad_norm": 0.6930897235870361, + "learning_rate": 0.0001315346784710363, + "loss": 0.9965, + "step": 8952 + }, + { + "epoch": 1.5940170940170941, + "grad_norm": 0.7808064818382263, + "learning_rate": 0.00013152139488994605, + "loss": 1.0527, + "step": 8953 + }, + { + "epoch": 1.5941951566951567, + "grad_norm": 0.6125522255897522, + "learning_rate": 0.0001315081106912595, + "loss": 1.1159, + "step": 8954 + }, + { + "epoch": 1.5943732193732194, + "grad_norm": 0.5863428711891174, + "learning_rate": 0.00013149482587523703, + "loss": 0.84, + "step": 8955 + }, + { + "epoch": 1.594551282051282, + "grad_norm": 0.7170202732086182, + "learning_rate": 0.00013148154044213882, + "loss": 1.0821, + "step": 8956 + }, + { + "epoch": 1.5947293447293447, + "grad_norm": 0.6409463882446289, + "learning_rate": 0.00013146825439222528, + "loss": 1.0097, + "step": 8957 + }, + { + "epoch": 1.5949074074074074, + "grad_norm": 0.7037690281867981, + "learning_rate": 0.00013145496772575666, + "loss": 1.1511, + "step": 8958 + }, + { + "epoch": 1.5950854700854702, + "grad_norm": 0.6400953531265259, + "learning_rate": 0.00013144168044299326, + "loss": 1.0809, + "step": 8959 + }, + { + "epoch": 1.5952635327635327, + "grad_norm": 0.6129940152168274, + "learning_rate": 0.00013142839254419545, + "loss": 0.8481, + "step": 8960 + }, + { + "epoch": 1.5954415954415955, + "grad_norm": 0.7452271580696106, + "learning_rate": 0.00013141510402962358, + "loss": 1.0649, + "step": 8961 + }, + { + "epoch": 1.595619658119658, + "grad_norm": 0.7407623529434204, + "learning_rate": 0.000131401814899538, + "loss": 0.9084, + "step": 8962 + }, + { + "epoch": 1.5957977207977208, + "grad_norm": 0.7103050947189331, + "learning_rate": 0.0001313885251541991, + "loss": 0.946, + "step": 8963 + }, + { + "epoch": 1.5959757834757835, + "grad_norm": 0.5566636323928833, + "learning_rate": 0.00013137523479386727, + "loss": 0.6781, + "step": 8964 + }, + { + "epoch": 1.5961538461538463, + "grad_norm": 0.8137457966804504, + "learning_rate": 0.00013136194381880288, + "loss": 0.9273, + "step": 8965 + }, + { + "epoch": 1.5963319088319088, + "grad_norm": 0.779330849647522, + "learning_rate": 0.0001313486522292663, + "loss": 1.1105, + "step": 8966 + }, + { + "epoch": 1.5965099715099715, + "grad_norm": 0.6807126998901367, + "learning_rate": 0.00013133536002551808, + "loss": 1.0728, + "step": 8967 + }, + { + "epoch": 1.596688034188034, + "grad_norm": 0.7371507287025452, + "learning_rate": 0.00013132206720781853, + "loss": 0.979, + "step": 8968 + }, + { + "epoch": 1.5968660968660968, + "grad_norm": 0.6811465620994568, + "learning_rate": 0.00013130877377642814, + "loss": 0.9821, + "step": 8969 + }, + { + "epoch": 1.5970441595441596, + "grad_norm": 0.6732743978500366, + "learning_rate": 0.00013129547973160738, + "loss": 0.8511, + "step": 8970 + }, + { + "epoch": 1.5972222222222223, + "grad_norm": 0.594901978969574, + "learning_rate": 0.0001312821850736167, + "loss": 0.9674, + "step": 8971 + }, + { + "epoch": 1.5974002849002849, + "grad_norm": 0.6743764281272888, + "learning_rate": 0.00013126888980271657, + "loss": 0.9268, + "step": 8972 + }, + { + "epoch": 1.5975783475783476, + "grad_norm": 0.7532161474227905, + "learning_rate": 0.00013125559391916752, + "loss": 1.0474, + "step": 8973 + }, + { + "epoch": 1.5977564102564101, + "grad_norm": 0.6331499814987183, + "learning_rate": 0.00013124229742323, + "loss": 1.05, + "step": 8974 + }, + { + "epoch": 1.5979344729344729, + "grad_norm": 0.7418690323829651, + "learning_rate": 0.0001312290003151646, + "loss": 0.9475, + "step": 8975 + }, + { + "epoch": 1.5981125356125356, + "grad_norm": 0.6511179804801941, + "learning_rate": 0.0001312157025952318, + "loss": 0.9206, + "step": 8976 + }, + { + "epoch": 1.5982905982905984, + "grad_norm": 0.6380775570869446, + "learning_rate": 0.00013120240426369215, + "loss": 0.9953, + "step": 8977 + }, + { + "epoch": 1.598468660968661, + "grad_norm": 0.8483675122261047, + "learning_rate": 0.00013118910532080623, + "loss": 0.9454, + "step": 8978 + }, + { + "epoch": 1.5986467236467237, + "grad_norm": 0.6700518727302551, + "learning_rate": 0.00013117580576683455, + "loss": 1.0413, + "step": 8979 + }, + { + "epoch": 1.5988247863247862, + "grad_norm": 0.7750083208084106, + "learning_rate": 0.00013116250560203774, + "loss": 1.1868, + "step": 8980 + }, + { + "epoch": 1.599002849002849, + "grad_norm": 0.7474972009658813, + "learning_rate": 0.00013114920482667635, + "loss": 1.0876, + "step": 8981 + }, + { + "epoch": 1.5991809116809117, + "grad_norm": 0.6920070052146912, + "learning_rate": 0.000131135903441011, + "loss": 1.0787, + "step": 8982 + }, + { + "epoch": 1.5993589743589745, + "grad_norm": 0.7572436928749084, + "learning_rate": 0.00013112260144530232, + "loss": 0.9798, + "step": 8983 + }, + { + "epoch": 1.5995370370370372, + "grad_norm": 0.6983019709587097, + "learning_rate": 0.00013110929883981088, + "loss": 1.1115, + "step": 8984 + }, + { + "epoch": 1.5997150997150997, + "grad_norm": 0.6352120041847229, + "learning_rate": 0.0001310959956247974, + "loss": 0.9962, + "step": 8985 + }, + { + "epoch": 1.5998931623931623, + "grad_norm": 0.596858561038971, + "learning_rate": 0.00013108269180052244, + "loss": 0.8686, + "step": 8986 + }, + { + "epoch": 1.600071225071225, + "grad_norm": 0.6237605214118958, + "learning_rate": 0.00013106938736724672, + "loss": 0.9166, + "step": 8987 + }, + { + "epoch": 1.6002492877492878, + "grad_norm": 0.6818585395812988, + "learning_rate": 0.0001310560823252309, + "loss": 0.9993, + "step": 8988 + }, + { + "epoch": 1.6004273504273505, + "grad_norm": 0.6372287273406982, + "learning_rate": 0.00013104277667473564, + "loss": 0.8589, + "step": 8989 + }, + { + "epoch": 1.6006054131054133, + "grad_norm": 0.6057302355766296, + "learning_rate": 0.0001310294704160217, + "loss": 0.9325, + "step": 8990 + }, + { + "epoch": 1.6007834757834758, + "grad_norm": 0.6999384164810181, + "learning_rate": 0.0001310161635493497, + "loss": 0.8691, + "step": 8991 + }, + { + "epoch": 1.6009615384615383, + "grad_norm": 0.6182113289833069, + "learning_rate": 0.00013100285607498045, + "loss": 1.0271, + "step": 8992 + }, + { + "epoch": 1.601139601139601, + "grad_norm": 0.6681149005889893, + "learning_rate": 0.0001309895479931746, + "loss": 0.989, + "step": 8993 + }, + { + "epoch": 1.6013176638176638, + "grad_norm": 0.6187826991081238, + "learning_rate": 0.00013097623930419293, + "loss": 0.8051, + "step": 8994 + }, + { + "epoch": 1.6014957264957266, + "grad_norm": 0.698793888092041, + "learning_rate": 0.00013096293000829621, + "loss": 1.0762, + "step": 8995 + }, + { + "epoch": 1.6016737891737893, + "grad_norm": 0.693149745464325, + "learning_rate": 0.0001309496201057452, + "loss": 1.0894, + "step": 8996 + }, + { + "epoch": 1.6018518518518519, + "grad_norm": 0.6664052605628967, + "learning_rate": 0.00013093630959680068, + "loss": 0.9835, + "step": 8997 + }, + { + "epoch": 1.6020299145299144, + "grad_norm": 0.6919469833374023, + "learning_rate": 0.0001309229984817234, + "loss": 0.9062, + "step": 8998 + }, + { + "epoch": 1.6022079772079771, + "grad_norm": 0.704781174659729, + "learning_rate": 0.00013090968676077427, + "loss": 0.8582, + "step": 8999 + }, + { + "epoch": 1.60238603988604, + "grad_norm": 0.8055264949798584, + "learning_rate": 0.000130896374434214, + "loss": 0.9813, + "step": 9000 + }, + { + "epoch": 1.6025641025641026, + "grad_norm": 0.6301952004432678, + "learning_rate": 0.00013088306150230348, + "loss": 0.7056, + "step": 9001 + }, + { + "epoch": 1.6027421652421654, + "grad_norm": 0.698544442653656, + "learning_rate": 0.00013086974796530347, + "loss": 0.9806, + "step": 9002 + }, + { + "epoch": 1.602920227920228, + "grad_norm": 0.669548511505127, + "learning_rate": 0.00013085643382347491, + "loss": 1.0317, + "step": 9003 + }, + { + "epoch": 1.6030982905982905, + "grad_norm": 0.6404716372489929, + "learning_rate": 0.00013084311907707864, + "loss": 0.8885, + "step": 9004 + }, + { + "epoch": 1.6032763532763532, + "grad_norm": 0.6968616843223572, + "learning_rate": 0.0001308298037263755, + "loss": 1.0665, + "step": 9005 + }, + { + "epoch": 1.603454415954416, + "grad_norm": 0.849311113357544, + "learning_rate": 0.00013081648777162644, + "loss": 1.1404, + "step": 9006 + }, + { + "epoch": 1.6036324786324787, + "grad_norm": 0.6603094935417175, + "learning_rate": 0.00013080317121309223, + "loss": 0.8341, + "step": 9007 + }, + { + "epoch": 1.6038105413105415, + "grad_norm": 0.6777810454368591, + "learning_rate": 0.00013078985405103394, + "loss": 1.044, + "step": 9008 + }, + { + "epoch": 1.603988603988604, + "grad_norm": 0.6783546209335327, + "learning_rate": 0.0001307765362857124, + "loss": 1.042, + "step": 9009 + }, + { + "epoch": 1.6041666666666665, + "grad_norm": 0.7251788377761841, + "learning_rate": 0.00013076321791738858, + "loss": 0.9004, + "step": 9010 + }, + { + "epoch": 1.6043447293447293, + "grad_norm": 0.7885342240333557, + "learning_rate": 0.00013074989894632338, + "loss": 1.1966, + "step": 9011 + }, + { + "epoch": 1.604522792022792, + "grad_norm": 0.7171013355255127, + "learning_rate": 0.0001307365793727778, + "loss": 1.2242, + "step": 9012 + }, + { + "epoch": 1.6047008547008548, + "grad_norm": 0.6027249693870544, + "learning_rate": 0.00013072325919701283, + "loss": 0.917, + "step": 9013 + }, + { + "epoch": 1.6048789173789175, + "grad_norm": 0.5957151055335999, + "learning_rate": 0.00013070993841928936, + "loss": 0.9154, + "step": 9014 + }, + { + "epoch": 1.60505698005698, + "grad_norm": 0.6190659403800964, + "learning_rate": 0.00013069661703986847, + "loss": 0.7071, + "step": 9015 + }, + { + "epoch": 1.6052350427350426, + "grad_norm": 0.6454868316650391, + "learning_rate": 0.00013068329505901117, + "loss": 0.8381, + "step": 9016 + }, + { + "epoch": 1.6054131054131053, + "grad_norm": 0.6255491375923157, + "learning_rate": 0.00013066997247697837, + "loss": 0.7515, + "step": 9017 + }, + { + "epoch": 1.605591168091168, + "grad_norm": 0.6214072108268738, + "learning_rate": 0.0001306566492940312, + "loss": 1.0101, + "step": 9018 + }, + { + "epoch": 1.6057692307692308, + "grad_norm": 0.7244150638580322, + "learning_rate": 0.0001306433255104307, + "loss": 1.2558, + "step": 9019 + }, + { + "epoch": 1.6059472934472936, + "grad_norm": 0.6162270903587341, + "learning_rate": 0.00013063000112643785, + "loss": 1.1009, + "step": 9020 + }, + { + "epoch": 1.6061253561253561, + "grad_norm": 0.7309414744377136, + "learning_rate": 0.0001306166761423138, + "loss": 1.1973, + "step": 9021 + }, + { + "epoch": 1.6063034188034186, + "grad_norm": 0.7150956392288208, + "learning_rate": 0.00013060335055831957, + "loss": 0.9136, + "step": 9022 + }, + { + "epoch": 1.6064814814814814, + "grad_norm": 0.8187742829322815, + "learning_rate": 0.00013059002437471623, + "loss": 1.0524, + "step": 9023 + }, + { + "epoch": 1.6066595441595442, + "grad_norm": 0.7928692698478699, + "learning_rate": 0.00013057669759176493, + "loss": 1.0249, + "step": 9024 + }, + { + "epoch": 1.606837606837607, + "grad_norm": 0.6929279565811157, + "learning_rate": 0.00013056337020972677, + "loss": 1.1804, + "step": 9025 + }, + { + "epoch": 1.6070156695156697, + "grad_norm": 0.6771654486656189, + "learning_rate": 0.00013055004222886285, + "loss": 1.0284, + "step": 9026 + }, + { + "epoch": 1.6071937321937322, + "grad_norm": 0.6689024567604065, + "learning_rate": 0.0001305367136494343, + "loss": 1.0431, + "step": 9027 + }, + { + "epoch": 1.6073717948717947, + "grad_norm": 0.71135413646698, + "learning_rate": 0.0001305233844717023, + "loss": 0.9692, + "step": 9028 + }, + { + "epoch": 1.6075498575498575, + "grad_norm": 0.5459749698638916, + "learning_rate": 0.00013051005469592796, + "loss": 0.5643, + "step": 9029 + }, + { + "epoch": 1.6077279202279202, + "grad_norm": 0.7225865125656128, + "learning_rate": 0.00013049672432237253, + "loss": 1.0954, + "step": 9030 + }, + { + "epoch": 1.607905982905983, + "grad_norm": 0.6878093481063843, + "learning_rate": 0.0001304833933512971, + "loss": 0.894, + "step": 9031 + }, + { + "epoch": 1.6080840455840457, + "grad_norm": 0.6967248320579529, + "learning_rate": 0.00013047006178296288, + "loss": 1.0356, + "step": 9032 + }, + { + "epoch": 1.6082621082621082, + "grad_norm": 0.6404993534088135, + "learning_rate": 0.00013045672961763114, + "loss": 0.8528, + "step": 9033 + }, + { + "epoch": 1.6084401709401708, + "grad_norm": 0.5919156074523926, + "learning_rate": 0.000130443396855563, + "loss": 0.7196, + "step": 9034 + }, + { + "epoch": 1.6086182336182335, + "grad_norm": 0.6792302131652832, + "learning_rate": 0.00013043006349701977, + "loss": 0.9519, + "step": 9035 + }, + { + "epoch": 1.6087962962962963, + "grad_norm": 0.6263542175292969, + "learning_rate": 0.00013041672954226268, + "loss": 1.0483, + "step": 9036 + }, + { + "epoch": 1.608974358974359, + "grad_norm": 0.5865579843521118, + "learning_rate": 0.00013040339499155294, + "loss": 0.8794, + "step": 9037 + }, + { + "epoch": 1.6091524216524218, + "grad_norm": 0.8383142948150635, + "learning_rate": 0.00013039005984515181, + "loss": 0.8929, + "step": 9038 + }, + { + "epoch": 1.6093304843304843, + "grad_norm": 0.6438691020011902, + "learning_rate": 0.00013037672410332063, + "loss": 0.9957, + "step": 9039 + }, + { + "epoch": 1.609508547008547, + "grad_norm": 0.74748694896698, + "learning_rate": 0.0001303633877663206, + "loss": 0.9809, + "step": 9040 + }, + { + "epoch": 1.6096866096866096, + "grad_norm": 0.6697205901145935, + "learning_rate": 0.00013035005083441312, + "loss": 0.9556, + "step": 9041 + }, + { + "epoch": 1.6098646723646723, + "grad_norm": 0.6577828526496887, + "learning_rate": 0.00013033671330785941, + "loss": 0.8956, + "step": 9042 + }, + { + "epoch": 1.610042735042735, + "grad_norm": 0.6423429846763611, + "learning_rate": 0.0001303233751869208, + "loss": 0.8467, + "step": 9043 + }, + { + "epoch": 1.6102207977207978, + "grad_norm": 0.6552175879478455, + "learning_rate": 0.00013031003647185867, + "loss": 0.8656, + "step": 9044 + }, + { + "epoch": 1.6103988603988604, + "grad_norm": 0.6755174398422241, + "learning_rate": 0.00013029669716293433, + "loss": 0.7836, + "step": 9045 + }, + { + "epoch": 1.6105769230769231, + "grad_norm": 0.6832906007766724, + "learning_rate": 0.00013028335726040914, + "loss": 1.1531, + "step": 9046 + }, + { + "epoch": 1.6107549857549857, + "grad_norm": 0.6498637795448303, + "learning_rate": 0.00013027001676454446, + "loss": 0.8637, + "step": 9047 + }, + { + "epoch": 1.6109330484330484, + "grad_norm": 0.6792944073677063, + "learning_rate": 0.0001302566756756017, + "loss": 1.0865, + "step": 9048 + }, + { + "epoch": 1.6111111111111112, + "grad_norm": 0.6801337003707886, + "learning_rate": 0.00013024333399384226, + "loss": 1.0738, + "step": 9049 + }, + { + "epoch": 1.611289173789174, + "grad_norm": 0.675216794013977, + "learning_rate": 0.0001302299917195275, + "loss": 1.1074, + "step": 9050 + }, + { + "epoch": 1.6114672364672364, + "grad_norm": 0.6418983340263367, + "learning_rate": 0.00013021664885291885, + "loss": 1.0025, + "step": 9051 + }, + { + "epoch": 1.6116452991452992, + "grad_norm": 0.7778789401054382, + "learning_rate": 0.0001302033053942777, + "loss": 1.0847, + "step": 9052 + }, + { + "epoch": 1.6118233618233617, + "grad_norm": 0.7672827243804932, + "learning_rate": 0.00013018996134386555, + "loss": 1.0565, + "step": 9053 + }, + { + "epoch": 1.6120014245014245, + "grad_norm": 0.6770617961883545, + "learning_rate": 0.00013017661670194382, + "loss": 0.9069, + "step": 9054 + }, + { + "epoch": 1.6121794871794872, + "grad_norm": 0.7161242961883545, + "learning_rate": 0.00013016327146877393, + "loss": 1.1301, + "step": 9055 + }, + { + "epoch": 1.61235754985755, + "grad_norm": 0.6923251152038574, + "learning_rate": 0.00013014992564461746, + "loss": 0.9546, + "step": 9056 + }, + { + "epoch": 1.6125356125356125, + "grad_norm": 0.622953474521637, + "learning_rate": 0.0001301365792297358, + "loss": 0.8152, + "step": 9057 + }, + { + "epoch": 1.6127136752136753, + "grad_norm": 0.7477008104324341, + "learning_rate": 0.00013012323222439046, + "loss": 0.8428, + "step": 9058 + }, + { + "epoch": 1.6128917378917378, + "grad_norm": 0.6612883806228638, + "learning_rate": 0.000130109884628843, + "loss": 1.0678, + "step": 9059 + }, + { + "epoch": 1.6130698005698005, + "grad_norm": 0.6406781077384949, + "learning_rate": 0.00013009653644335486, + "loss": 0.6792, + "step": 9060 + }, + { + "epoch": 1.6132478632478633, + "grad_norm": 0.6279141902923584, + "learning_rate": 0.00013008318766818763, + "loss": 0.9826, + "step": 9061 + }, + { + "epoch": 1.613425925925926, + "grad_norm": 0.6616412401199341, + "learning_rate": 0.00013006983830360285, + "loss": 1.0691, + "step": 9062 + }, + { + "epoch": 1.6136039886039886, + "grad_norm": 0.6520406603813171, + "learning_rate": 0.000130056488349862, + "loss": 0.9487, + "step": 9063 + }, + { + "epoch": 1.6137820512820513, + "grad_norm": 0.6378647089004517, + "learning_rate": 0.00013004313780722672, + "loss": 0.8557, + "step": 9064 + }, + { + "epoch": 1.6139601139601139, + "grad_norm": 0.6547569036483765, + "learning_rate": 0.00013002978667595857, + "loss": 0.879, + "step": 9065 + }, + { + "epoch": 1.6141381766381766, + "grad_norm": 0.7347842454910278, + "learning_rate": 0.00013001643495631914, + "loss": 1.0757, + "step": 9066 + }, + { + "epoch": 1.6143162393162394, + "grad_norm": 0.5988406538963318, + "learning_rate": 0.00013000308264857002, + "loss": 0.6754, + "step": 9067 + }, + { + "epoch": 1.614494301994302, + "grad_norm": 0.6949366331100464, + "learning_rate": 0.00012998972975297282, + "loss": 1.1236, + "step": 9068 + }, + { + "epoch": 1.6146723646723646, + "grad_norm": 0.7095484137535095, + "learning_rate": 0.00012997637626978913, + "loss": 1.0124, + "step": 9069 + }, + { + "epoch": 1.6148504273504274, + "grad_norm": 0.6634095311164856, + "learning_rate": 0.00012996302219928064, + "loss": 1.2018, + "step": 9070 + }, + { + "epoch": 1.61502849002849, + "grad_norm": 0.6894524693489075, + "learning_rate": 0.000129949667541709, + "loss": 0.9959, + "step": 9071 + }, + { + "epoch": 1.6152065527065527, + "grad_norm": 0.672334611415863, + "learning_rate": 0.00012993631229733582, + "loss": 1.0369, + "step": 9072 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.725759744644165, + "learning_rate": 0.00012992295646642278, + "loss": 1.0079, + "step": 9073 + }, + { + "epoch": 1.6155626780626782, + "grad_norm": 0.7941585779190063, + "learning_rate": 0.00012990960004923154, + "loss": 0.9468, + "step": 9074 + }, + { + "epoch": 1.6157407407407407, + "grad_norm": 0.6556950807571411, + "learning_rate": 0.00012989624304602385, + "loss": 0.9915, + "step": 9075 + }, + { + "epoch": 1.6159188034188035, + "grad_norm": 0.7515892386436462, + "learning_rate": 0.0001298828854570614, + "loss": 1.0924, + "step": 9076 + }, + { + "epoch": 1.616096866096866, + "grad_norm": 0.6944101452827454, + "learning_rate": 0.00012986952728260586, + "loss": 0.9632, + "step": 9077 + }, + { + "epoch": 1.6162749287749287, + "grad_norm": 0.6286170482635498, + "learning_rate": 0.000129856168522919, + "loss": 1.0311, + "step": 9078 + }, + { + "epoch": 1.6164529914529915, + "grad_norm": 0.8362757563591003, + "learning_rate": 0.0001298428091782625, + "loss": 1.1232, + "step": 9079 + }, + { + "epoch": 1.6166310541310542, + "grad_norm": 0.6199851632118225, + "learning_rate": 0.0001298294492488982, + "loss": 0.9454, + "step": 9080 + }, + { + "epoch": 1.6168091168091168, + "grad_norm": 0.7541791796684265, + "learning_rate": 0.0001298160887350878, + "loss": 0.9759, + "step": 9081 + }, + { + "epoch": 1.6169871794871795, + "grad_norm": 0.6940878033638, + "learning_rate": 0.00012980272763709304, + "loss": 0.9258, + "step": 9082 + }, + { + "epoch": 1.617165242165242, + "grad_norm": 0.6934045553207397, + "learning_rate": 0.00012978936595517575, + "loss": 1.0142, + "step": 9083 + }, + { + "epoch": 1.6173433048433048, + "grad_norm": 0.8147503733634949, + "learning_rate": 0.00012977600368959774, + "loss": 0.964, + "step": 9084 + }, + { + "epoch": 1.6175213675213675, + "grad_norm": 0.6583107709884644, + "learning_rate": 0.00012976264084062079, + "loss": 1.0315, + "step": 9085 + }, + { + "epoch": 1.6176994301994303, + "grad_norm": 0.7192013263702393, + "learning_rate": 0.0001297492774085067, + "loss": 0.9528, + "step": 9086 + }, + { + "epoch": 1.6178774928774928, + "grad_norm": 0.665888786315918, + "learning_rate": 0.00012973591339351733, + "loss": 1.0188, + "step": 9087 + }, + { + "epoch": 1.6180555555555556, + "grad_norm": 0.7170987725257874, + "learning_rate": 0.0001297225487959145, + "loss": 0.8969, + "step": 9088 + }, + { + "epoch": 1.618233618233618, + "grad_norm": 0.6768732070922852, + "learning_rate": 0.00012970918361596007, + "loss": 1.1951, + "step": 9089 + }, + { + "epoch": 1.6184116809116809, + "grad_norm": 0.6640290021896362, + "learning_rate": 0.00012969581785391592, + "loss": 0.9649, + "step": 9090 + }, + { + "epoch": 1.6185897435897436, + "grad_norm": 0.6200813055038452, + "learning_rate": 0.00012968245151004392, + "loss": 0.9446, + "step": 9091 + }, + { + "epoch": 1.6187678062678064, + "grad_norm": 0.6815837621688843, + "learning_rate": 0.0001296690845846059, + "loss": 1.0506, + "step": 9092 + }, + { + "epoch": 1.618945868945869, + "grad_norm": 0.7252637147903442, + "learning_rate": 0.0001296557170778638, + "loss": 1.1977, + "step": 9093 + }, + { + "epoch": 1.6191239316239316, + "grad_norm": 0.5609107613563538, + "learning_rate": 0.00012964234899007955, + "loss": 0.8009, + "step": 9094 + }, + { + "epoch": 1.6193019943019942, + "grad_norm": 0.6539437770843506, + "learning_rate": 0.00012962898032151506, + "loss": 0.8482, + "step": 9095 + }, + { + "epoch": 1.619480056980057, + "grad_norm": 0.6993300914764404, + "learning_rate": 0.0001296156110724322, + "loss": 1.0725, + "step": 9096 + }, + { + "epoch": 1.6196581196581197, + "grad_norm": 0.6768273711204529, + "learning_rate": 0.000129602241243093, + "loss": 0.9247, + "step": 9097 + }, + { + "epoch": 1.6198361823361824, + "grad_norm": 0.6896265745162964, + "learning_rate": 0.00012958887083375939, + "loss": 0.9526, + "step": 9098 + }, + { + "epoch": 1.6200142450142452, + "grad_norm": 0.7475146651268005, + "learning_rate": 0.00012957549984469327, + "loss": 0.8302, + "step": 9099 + }, + { + "epoch": 1.6201923076923077, + "grad_norm": 0.6622769236564636, + "learning_rate": 0.00012956212827615674, + "loss": 0.9505, + "step": 9100 + }, + { + "epoch": 1.6203703703703702, + "grad_norm": 0.6938058137893677, + "learning_rate": 0.00012954875612841167, + "loss": 0.9757, + "step": 9101 + }, + { + "epoch": 1.620548433048433, + "grad_norm": 0.7453510761260986, + "learning_rate": 0.0001295353834017201, + "loss": 1.0919, + "step": 9102 + }, + { + "epoch": 1.6207264957264957, + "grad_norm": 0.7868932485580444, + "learning_rate": 0.0001295220100963441, + "loss": 0.9265, + "step": 9103 + }, + { + "epoch": 1.6209045584045585, + "grad_norm": 0.6779825091362, + "learning_rate": 0.00012950863621254558, + "loss": 0.98, + "step": 9104 + }, + { + "epoch": 1.6210826210826212, + "grad_norm": 0.6825897097587585, + "learning_rate": 0.00012949526175058662, + "loss": 0.9218, + "step": 9105 + }, + { + "epoch": 1.6212606837606838, + "grad_norm": 0.6686047911643982, + "learning_rate": 0.00012948188671072934, + "loss": 0.9546, + "step": 9106 + }, + { + "epoch": 1.6214387464387463, + "grad_norm": 0.7456090450286865, + "learning_rate": 0.0001294685110932357, + "loss": 1.0819, + "step": 9107 + }, + { + "epoch": 1.621616809116809, + "grad_norm": 0.7111441493034363, + "learning_rate": 0.0001294551348983678, + "loss": 0.9916, + "step": 9108 + }, + { + "epoch": 1.6217948717948718, + "grad_norm": 0.6534699201583862, + "learning_rate": 0.00012944175812638773, + "loss": 1.0374, + "step": 9109 + }, + { + "epoch": 1.6219729344729346, + "grad_norm": 0.6046397089958191, + "learning_rate": 0.00012942838077755758, + "loss": 0.7922, + "step": 9110 + }, + { + "epoch": 1.6221509971509973, + "grad_norm": 0.7736679911613464, + "learning_rate": 0.00012941500285213942, + "loss": 1.0056, + "step": 9111 + }, + { + "epoch": 1.6223290598290598, + "grad_norm": 0.6850929260253906, + "learning_rate": 0.00012940162435039538, + "loss": 0.9538, + "step": 9112 + }, + { + "epoch": 1.6225071225071224, + "grad_norm": 0.6305751800537109, + "learning_rate": 0.00012938824527258756, + "loss": 0.9341, + "step": 9113 + }, + { + "epoch": 1.6226851851851851, + "grad_norm": 0.6740923523902893, + "learning_rate": 0.0001293748656189782, + "loss": 1.0037, + "step": 9114 + }, + { + "epoch": 1.6228632478632479, + "grad_norm": 0.6579762101173401, + "learning_rate": 0.00012936148538982928, + "loss": 1.0022, + "step": 9115 + }, + { + "epoch": 1.6230413105413106, + "grad_norm": 0.6500434279441833, + "learning_rate": 0.0001293481045854031, + "loss": 0.8589, + "step": 9116 + }, + { + "epoch": 1.6232193732193734, + "grad_norm": 0.7825912237167358, + "learning_rate": 0.00012933472320596177, + "loss": 1.0345, + "step": 9117 + }, + { + "epoch": 1.623397435897436, + "grad_norm": 0.8341414332389832, + "learning_rate": 0.0001293213412517675, + "loss": 1.0314, + "step": 9118 + }, + { + "epoch": 1.6235754985754984, + "grad_norm": 0.63664311170578, + "learning_rate": 0.00012930795872308242, + "loss": 0.819, + "step": 9119 + }, + { + "epoch": 1.6237535612535612, + "grad_norm": 0.6800840497016907, + "learning_rate": 0.00012929457562016878, + "loss": 0.95, + "step": 9120 + }, + { + "epoch": 1.623931623931624, + "grad_norm": 0.754165530204773, + "learning_rate": 0.0001292811919432888, + "loss": 1.1193, + "step": 9121 + }, + { + "epoch": 1.6241096866096867, + "grad_norm": 0.678871750831604, + "learning_rate": 0.00012926780769270465, + "loss": 0.9015, + "step": 9122 + }, + { + "epoch": 1.6242877492877494, + "grad_norm": 0.6642945408821106, + "learning_rate": 0.00012925442286867866, + "loss": 0.9095, + "step": 9123 + }, + { + "epoch": 1.624465811965812, + "grad_norm": 0.6089697480201721, + "learning_rate": 0.000129241037471473, + "loss": 0.8994, + "step": 9124 + }, + { + "epoch": 1.6246438746438745, + "grad_norm": 0.7320881485939026, + "learning_rate": 0.00012922765150134995, + "loss": 1.0518, + "step": 9125 + }, + { + "epoch": 1.6248219373219372, + "grad_norm": 0.7308032512664795, + "learning_rate": 0.0001292142649585718, + "loss": 1.0557, + "step": 9126 + }, + { + "epoch": 1.625, + "grad_norm": 0.6896602511405945, + "learning_rate": 0.0001292008778434008, + "loss": 1.145, + "step": 9127 + }, + { + "epoch": 1.6251780626780628, + "grad_norm": 0.6112532615661621, + "learning_rate": 0.00012918749015609926, + "loss": 0.9611, + "step": 9128 + }, + { + "epoch": 1.6253561253561255, + "grad_norm": 0.6856057643890381, + "learning_rate": 0.00012917410189692947, + "loss": 1.0124, + "step": 9129 + }, + { + "epoch": 1.625534188034188, + "grad_norm": 0.699252188205719, + "learning_rate": 0.00012916071306615378, + "loss": 0.8854, + "step": 9130 + }, + { + "epoch": 1.6257122507122506, + "grad_norm": 0.6306683421134949, + "learning_rate": 0.0001291473236640345, + "loss": 1.0722, + "step": 9131 + }, + { + "epoch": 1.6258903133903133, + "grad_norm": 0.6358118653297424, + "learning_rate": 0.00012913393369083393, + "loss": 0.889, + "step": 9132 + }, + { + "epoch": 1.626068376068376, + "grad_norm": 0.6953601837158203, + "learning_rate": 0.00012912054314681445, + "loss": 1.0168, + "step": 9133 + }, + { + "epoch": 1.6262464387464388, + "grad_norm": 0.6742331385612488, + "learning_rate": 0.00012910715203223844, + "loss": 0.8152, + "step": 9134 + }, + { + "epoch": 1.6264245014245016, + "grad_norm": 0.5872861742973328, + "learning_rate": 0.00012909376034736823, + "loss": 0.8702, + "step": 9135 + }, + { + "epoch": 1.626602564102564, + "grad_norm": 0.7580631971359253, + "learning_rate": 0.00012908036809246623, + "loss": 0.994, + "step": 9136 + }, + { + "epoch": 1.6267806267806266, + "grad_norm": 0.7544930577278137, + "learning_rate": 0.00012906697526779488, + "loss": 0.7475, + "step": 9137 + }, + { + "epoch": 1.6269586894586894, + "grad_norm": 0.6850766539573669, + "learning_rate": 0.00012905358187361647, + "loss": 1.0943, + "step": 9138 + }, + { + "epoch": 1.6271367521367521, + "grad_norm": 0.6821565628051758, + "learning_rate": 0.0001290401879101935, + "loss": 1.2928, + "step": 9139 + }, + { + "epoch": 1.6273148148148149, + "grad_norm": 0.6961034536361694, + "learning_rate": 0.00012902679337778835, + "loss": 0.8694, + "step": 9140 + }, + { + "epoch": 1.6274928774928776, + "grad_norm": 0.7159550786018372, + "learning_rate": 0.00012901339827666353, + "loss": 0.8827, + "step": 9141 + }, + { + "epoch": 1.6276709401709402, + "grad_norm": 0.7491081953048706, + "learning_rate": 0.0001290000026070814, + "loss": 0.8159, + "step": 9142 + }, + { + "epoch": 1.6278490028490027, + "grad_norm": 0.7107849717140198, + "learning_rate": 0.00012898660636930447, + "loss": 1.0625, + "step": 9143 + }, + { + "epoch": 1.6280270655270654, + "grad_norm": 0.7227210998535156, + "learning_rate": 0.0001289732095635952, + "loss": 0.9744, + "step": 9144 + }, + { + "epoch": 1.6282051282051282, + "grad_norm": 0.7141995429992676, + "learning_rate": 0.00012895981219021607, + "loss": 0.9836, + "step": 9145 + }, + { + "epoch": 1.628383190883191, + "grad_norm": 0.6445552706718445, + "learning_rate": 0.00012894641424942958, + "loss": 1.0183, + "step": 9146 + }, + { + "epoch": 1.6285612535612537, + "grad_norm": 0.698783278465271, + "learning_rate": 0.00012893301574149824, + "loss": 0.8392, + "step": 9147 + }, + { + "epoch": 1.6287393162393162, + "grad_norm": 0.6529116034507751, + "learning_rate": 0.00012891961666668458, + "loss": 0.9317, + "step": 9148 + }, + { + "epoch": 1.6289173789173788, + "grad_norm": 0.7780548930168152, + "learning_rate": 0.0001289062170252511, + "loss": 1.2406, + "step": 9149 + }, + { + "epoch": 1.6290954415954415, + "grad_norm": 0.6500990986824036, + "learning_rate": 0.0001288928168174603, + "loss": 1.0381, + "step": 9150 + }, + { + "epoch": 1.6292735042735043, + "grad_norm": 0.7098208665847778, + "learning_rate": 0.00012887941604357482, + "loss": 1.2126, + "step": 9151 + }, + { + "epoch": 1.629451566951567, + "grad_norm": 0.730648398399353, + "learning_rate": 0.0001288660147038572, + "loss": 0.8351, + "step": 9152 + }, + { + "epoch": 1.6296296296296298, + "grad_norm": 0.5520278215408325, + "learning_rate": 0.0001288526127985699, + "loss": 0.5877, + "step": 9153 + }, + { + "epoch": 1.6298076923076923, + "grad_norm": 0.7611770033836365, + "learning_rate": 0.00012883921032797563, + "loss": 1.2227, + "step": 9154 + }, + { + "epoch": 1.6299857549857548, + "grad_norm": 0.636820375919342, + "learning_rate": 0.00012882580729233696, + "loss": 0.8305, + "step": 9155 + }, + { + "epoch": 1.6301638176638176, + "grad_norm": 0.694492518901825, + "learning_rate": 0.00012881240369191644, + "loss": 1.0452, + "step": 9156 + }, + { + "epoch": 1.6303418803418803, + "grad_norm": 0.67826908826828, + "learning_rate": 0.00012879899952697677, + "loss": 0.8345, + "step": 9157 + }, + { + "epoch": 1.630519943019943, + "grad_norm": 0.5891323685646057, + "learning_rate": 0.00012878559479778052, + "loss": 0.8367, + "step": 9158 + }, + { + "epoch": 1.6306980056980058, + "grad_norm": 0.6766192317008972, + "learning_rate": 0.0001287721895045903, + "loss": 0.8319, + "step": 9159 + }, + { + "epoch": 1.6308760683760684, + "grad_norm": 0.5306392908096313, + "learning_rate": 0.0001287587836476688, + "loss": 0.7945, + "step": 9160 + }, + { + "epoch": 1.631054131054131, + "grad_norm": 0.6677970290184021, + "learning_rate": 0.0001287453772272787, + "loss": 1.1228, + "step": 9161 + }, + { + "epoch": 1.6312321937321936, + "grad_norm": 0.810052752494812, + "learning_rate": 0.00012873197024368266, + "loss": 0.8395, + "step": 9162 + }, + { + "epoch": 1.6314102564102564, + "grad_norm": 0.7619220018386841, + "learning_rate": 0.00012871856269714333, + "loss": 1.3713, + "step": 9163 + }, + { + "epoch": 1.6315883190883191, + "grad_norm": 0.6564521193504333, + "learning_rate": 0.00012870515458792342, + "loss": 1.0513, + "step": 9164 + }, + { + "epoch": 1.631766381766382, + "grad_norm": 0.6874445676803589, + "learning_rate": 0.00012869174591628564, + "loss": 1.0255, + "step": 9165 + }, + { + "epoch": 1.6319444444444444, + "grad_norm": 0.6958737373352051, + "learning_rate": 0.0001286783366824927, + "loss": 0.9361, + "step": 9166 + }, + { + "epoch": 1.6321225071225072, + "grad_norm": 0.6909199357032776, + "learning_rate": 0.0001286649268868073, + "loss": 0.9855, + "step": 9167 + }, + { + "epoch": 1.6323005698005697, + "grad_norm": 0.7671375274658203, + "learning_rate": 0.00012865151652949225, + "loss": 1.084, + "step": 9168 + }, + { + "epoch": 1.6324786324786325, + "grad_norm": 0.750200092792511, + "learning_rate": 0.00012863810561081023, + "loss": 0.9341, + "step": 9169 + }, + { + "epoch": 1.6326566951566952, + "grad_norm": 0.6595860123634338, + "learning_rate": 0.00012862469413102402, + "loss": 0.9386, + "step": 9170 + }, + { + "epoch": 1.632834757834758, + "grad_norm": 0.622373640537262, + "learning_rate": 0.0001286112820903964, + "loss": 0.7697, + "step": 9171 + }, + { + "epoch": 1.6330128205128205, + "grad_norm": 0.9628498554229736, + "learning_rate": 0.00012859786948919014, + "loss": 1.2629, + "step": 9172 + }, + { + "epoch": 1.6331908831908832, + "grad_norm": 0.7610561847686768, + "learning_rate": 0.000128584456327668, + "loss": 0.9748, + "step": 9173 + }, + { + "epoch": 1.6333689458689458, + "grad_norm": 0.6585374474525452, + "learning_rate": 0.00012857104260609285, + "loss": 0.9049, + "step": 9174 + }, + { + "epoch": 1.6335470085470085, + "grad_norm": 0.6996221542358398, + "learning_rate": 0.00012855762832472746, + "loss": 0.8893, + "step": 9175 + }, + { + "epoch": 1.6337250712250713, + "grad_norm": 0.6226270198822021, + "learning_rate": 0.00012854421348383466, + "loss": 0.8913, + "step": 9176 + }, + { + "epoch": 1.633903133903134, + "grad_norm": 0.6570866107940674, + "learning_rate": 0.00012853079808367731, + "loss": 0.8632, + "step": 9177 + }, + { + "epoch": 1.6340811965811965, + "grad_norm": 0.6899664402008057, + "learning_rate": 0.00012851738212451826, + "loss": 0.8177, + "step": 9178 + }, + { + "epoch": 1.6342592592592593, + "grad_norm": 0.75257807970047, + "learning_rate": 0.0001285039656066203, + "loss": 0.9096, + "step": 9179 + }, + { + "epoch": 1.6344373219373218, + "grad_norm": 0.6614963412284851, + "learning_rate": 0.00012849054853024638, + "loss": 0.9255, + "step": 9180 + }, + { + "epoch": 1.6346153846153846, + "grad_norm": 0.7245957851409912, + "learning_rate": 0.00012847713089565933, + "loss": 1.0122, + "step": 9181 + }, + { + "epoch": 1.6347934472934473, + "grad_norm": 0.7332839369773865, + "learning_rate": 0.00012846371270312204, + "loss": 0.8484, + "step": 9182 + }, + { + "epoch": 1.63497150997151, + "grad_norm": 0.628089189529419, + "learning_rate": 0.00012845029395289748, + "loss": 1.0171, + "step": 9183 + }, + { + "epoch": 1.6351495726495726, + "grad_norm": 0.7493528723716736, + "learning_rate": 0.00012843687464524848, + "loss": 1.1635, + "step": 9184 + }, + { + "epoch": 1.6353276353276354, + "grad_norm": 0.6328163146972656, + "learning_rate": 0.00012842345478043799, + "loss": 1.1254, + "step": 9185 + }, + { + "epoch": 1.635505698005698, + "grad_norm": 0.6720291376113892, + "learning_rate": 0.00012841003435872894, + "loss": 0.9729, + "step": 9186 + }, + { + "epoch": 1.6356837606837606, + "grad_norm": 0.6657332181930542, + "learning_rate": 0.00012839661338038427, + "loss": 1.1047, + "step": 9187 + }, + { + "epoch": 1.6358618233618234, + "grad_norm": 0.7416180968284607, + "learning_rate": 0.000128383191845667, + "loss": 0.9505, + "step": 9188 + }, + { + "epoch": 1.6360398860398861, + "grad_norm": 0.8737816214561462, + "learning_rate": 0.00012836976975484, + "loss": 1.0518, + "step": 9189 + }, + { + "epoch": 1.6362179487179487, + "grad_norm": 0.7351877093315125, + "learning_rate": 0.0001283563471081663, + "loss": 1.1152, + "step": 9190 + }, + { + "epoch": 1.6363960113960114, + "grad_norm": 0.6442788243293762, + "learning_rate": 0.00012834292390590893, + "loss": 0.9432, + "step": 9191 + }, + { + "epoch": 1.636574074074074, + "grad_norm": 0.6848029494285583, + "learning_rate": 0.0001283295001483308, + "loss": 0.8528, + "step": 9192 + }, + { + "epoch": 1.6367521367521367, + "grad_norm": 0.6627060174942017, + "learning_rate": 0.00012831607583569497, + "loss": 1.0222, + "step": 9193 + }, + { + "epoch": 1.6369301994301995, + "grad_norm": 0.7319555878639221, + "learning_rate": 0.00012830265096826446, + "loss": 0.9392, + "step": 9194 + }, + { + "epoch": 1.6371082621082622, + "grad_norm": 0.6986424326896667, + "learning_rate": 0.0001282892255463023, + "loss": 1.2095, + "step": 9195 + }, + { + "epoch": 1.6372863247863247, + "grad_norm": 0.6649929881095886, + "learning_rate": 0.0001282757995700715, + "loss": 0.9426, + "step": 9196 + }, + { + "epoch": 1.6374643874643875, + "grad_norm": 0.6789031624794006, + "learning_rate": 0.0001282623730398352, + "loss": 0.9705, + "step": 9197 + }, + { + "epoch": 1.63764245014245, + "grad_norm": 0.6388779878616333, + "learning_rate": 0.00012824894595585637, + "loss": 1.0698, + "step": 9198 + }, + { + "epoch": 1.6378205128205128, + "grad_norm": 0.636832594871521, + "learning_rate": 0.00012823551831839814, + "loss": 0.9445, + "step": 9199 + }, + { + "epoch": 1.6379985754985755, + "grad_norm": 0.670190691947937, + "learning_rate": 0.0001282220901277236, + "loss": 0.9847, + "step": 9200 + }, + { + "epoch": 1.6381766381766383, + "grad_norm": 0.6020209193229675, + "learning_rate": 0.0001282086613840958, + "loss": 1.0047, + "step": 9201 + }, + { + "epoch": 1.6383547008547008, + "grad_norm": 0.6648211479187012, + "learning_rate": 0.0001281952320877779, + "loss": 0.8717, + "step": 9202 + }, + { + "epoch": 1.6385327635327636, + "grad_norm": 0.7207710146903992, + "learning_rate": 0.000128181802239033, + "loss": 1.1232, + "step": 9203 + }, + { + "epoch": 1.638710826210826, + "grad_norm": 0.800992488861084, + "learning_rate": 0.0001281683718381242, + "loss": 1.0688, + "step": 9204 + }, + { + "epoch": 1.6388888888888888, + "grad_norm": 0.789398193359375, + "learning_rate": 0.0001281549408853147, + "loss": 1.1772, + "step": 9205 + }, + { + "epoch": 1.6390669515669516, + "grad_norm": 0.6514480710029602, + "learning_rate": 0.0001281415093808676, + "loss": 1.1685, + "step": 9206 + }, + { + "epoch": 1.6392450142450143, + "grad_norm": 0.6914686560630798, + "learning_rate": 0.00012812807732504608, + "loss": 1.1307, + "step": 9207 + }, + { + "epoch": 1.6394230769230769, + "grad_norm": 0.6788144111633301, + "learning_rate": 0.00012811464471811334, + "loss": 1.1735, + "step": 9208 + }, + { + "epoch": 1.6396011396011396, + "grad_norm": 0.7049870491027832, + "learning_rate": 0.00012810121156033252, + "loss": 1.0128, + "step": 9209 + }, + { + "epoch": 1.6397792022792022, + "grad_norm": 0.7156766057014465, + "learning_rate": 0.00012808777785196687, + "loss": 0.9503, + "step": 9210 + }, + { + "epoch": 1.639957264957265, + "grad_norm": 0.651716411113739, + "learning_rate": 0.0001280743435932795, + "loss": 1.1227, + "step": 9211 + }, + { + "epoch": 1.6401353276353277, + "grad_norm": 0.7276262044906616, + "learning_rate": 0.0001280609087845337, + "loss": 1.06, + "step": 9212 + }, + { + "epoch": 1.6403133903133904, + "grad_norm": 0.6591095924377441, + "learning_rate": 0.0001280474734259927, + "loss": 1.0861, + "step": 9213 + }, + { + "epoch": 1.640491452991453, + "grad_norm": 0.6675926446914673, + "learning_rate": 0.00012803403751791975, + "loss": 0.9815, + "step": 9214 + }, + { + "epoch": 1.6406695156695157, + "grad_norm": 0.6391474008560181, + "learning_rate": 0.00012802060106057803, + "loss": 0.8027, + "step": 9215 + }, + { + "epoch": 1.6408475783475782, + "grad_norm": 0.6384556293487549, + "learning_rate": 0.00012800716405423086, + "loss": 0.7877, + "step": 9216 + }, + { + "epoch": 1.641025641025641, + "grad_norm": 0.661191463470459, + "learning_rate": 0.00012799372649914146, + "loss": 0.9725, + "step": 9217 + }, + { + "epoch": 1.6412037037037037, + "grad_norm": 0.7418332695960999, + "learning_rate": 0.0001279802883955732, + "loss": 1.1756, + "step": 9218 + }, + { + "epoch": 1.6413817663817665, + "grad_norm": 0.6588954329490662, + "learning_rate": 0.00012796684974378928, + "loss": 1.0428, + "step": 9219 + }, + { + "epoch": 1.6415598290598292, + "grad_norm": 0.7566093802452087, + "learning_rate": 0.000127953410544053, + "loss": 1.1254, + "step": 9220 + }, + { + "epoch": 1.6417378917378918, + "grad_norm": 0.6801039576530457, + "learning_rate": 0.00012793997079662777, + "loss": 1.0854, + "step": 9221 + }, + { + "epoch": 1.6419159544159543, + "grad_norm": 0.7262716889381409, + "learning_rate": 0.0001279265305017768, + "loss": 0.9343, + "step": 9222 + }, + { + "epoch": 1.642094017094017, + "grad_norm": 0.628625750541687, + "learning_rate": 0.0001279130896597635, + "loss": 0.8942, + "step": 9223 + }, + { + "epoch": 1.6422720797720798, + "grad_norm": 0.6183576583862305, + "learning_rate": 0.0001278996482708512, + "loss": 0.9284, + "step": 9224 + }, + { + "epoch": 1.6424501424501425, + "grad_norm": 0.7912000417709351, + "learning_rate": 0.00012788620633530327, + "loss": 1.3043, + "step": 9225 + }, + { + "epoch": 1.6426282051282053, + "grad_norm": 0.6982026100158691, + "learning_rate": 0.00012787276385338298, + "loss": 1.0224, + "step": 9226 + }, + { + "epoch": 1.6428062678062678, + "grad_norm": 0.6734985709190369, + "learning_rate": 0.00012785932082535386, + "loss": 0.8781, + "step": 9227 + }, + { + "epoch": 1.6429843304843303, + "grad_norm": 0.8799532055854797, + "learning_rate": 0.0001278458772514792, + "loss": 1.1482, + "step": 9228 + }, + { + "epoch": 1.643162393162393, + "grad_norm": 0.590295672416687, + "learning_rate": 0.0001278324331320224, + "loss": 0.9502, + "step": 9229 + }, + { + "epoch": 1.6433404558404558, + "grad_norm": 0.6562125086784363, + "learning_rate": 0.0001278189884672469, + "loss": 0.9834, + "step": 9230 + }, + { + "epoch": 1.6435185185185186, + "grad_norm": 0.6848936676979065, + "learning_rate": 0.00012780554325741612, + "loss": 1.0414, + "step": 9231 + }, + { + "epoch": 1.6436965811965814, + "grad_norm": 0.5985032320022583, + "learning_rate": 0.00012779209750279344, + "loss": 0.9469, + "step": 9232 + }, + { + "epoch": 1.6438746438746439, + "grad_norm": 0.7500917911529541, + "learning_rate": 0.00012777865120364238, + "loss": 0.9626, + "step": 9233 + }, + { + "epoch": 1.6440527065527064, + "grad_norm": 0.6565709114074707, + "learning_rate": 0.00012776520436022634, + "loss": 1.0594, + "step": 9234 + }, + { + "epoch": 1.6442307692307692, + "grad_norm": 0.8005441427230835, + "learning_rate": 0.00012775175697280882, + "loss": 1.2379, + "step": 9235 + }, + { + "epoch": 1.644408831908832, + "grad_norm": 0.6734150648117065, + "learning_rate": 0.00012773830904165326, + "loss": 0.9171, + "step": 9236 + }, + { + "epoch": 1.6445868945868947, + "grad_norm": 0.6950868368148804, + "learning_rate": 0.00012772486056702314, + "loss": 1.1782, + "step": 9237 + }, + { + "epoch": 1.6447649572649574, + "grad_norm": 0.8009599447250366, + "learning_rate": 0.000127711411549182, + "loss": 1.0288, + "step": 9238 + }, + { + "epoch": 1.64494301994302, + "grad_norm": 0.6227970719337463, + "learning_rate": 0.0001276979619883933, + "loss": 0.9327, + "step": 9239 + }, + { + "epoch": 1.6451210826210825, + "grad_norm": 0.6828190088272095, + "learning_rate": 0.00012768451188492058, + "loss": 0.9816, + "step": 9240 + }, + { + "epoch": 1.6452991452991452, + "grad_norm": 0.9689767360687256, + "learning_rate": 0.00012767106123902738, + "loss": 0.9049, + "step": 9241 + }, + { + "epoch": 1.645477207977208, + "grad_norm": 0.677061140537262, + "learning_rate": 0.00012765761005097717, + "loss": 0.9472, + "step": 9242 + }, + { + "epoch": 1.6456552706552707, + "grad_norm": 0.7227110862731934, + "learning_rate": 0.00012764415832103356, + "loss": 1.0384, + "step": 9243 + }, + { + "epoch": 1.6458333333333335, + "grad_norm": 0.6540094614028931, + "learning_rate": 0.0001276307060494601, + "loss": 0.8166, + "step": 9244 + }, + { + "epoch": 1.646011396011396, + "grad_norm": 0.6921904683113098, + "learning_rate": 0.00012761725323652033, + "loss": 0.9746, + "step": 9245 + }, + { + "epoch": 1.6461894586894585, + "grad_norm": 0.6742660999298096, + "learning_rate": 0.0001276037998824779, + "loss": 0.8441, + "step": 9246 + }, + { + "epoch": 1.6463675213675213, + "grad_norm": 0.6611103415489197, + "learning_rate": 0.0001275903459875963, + "loss": 1.087, + "step": 9247 + }, + { + "epoch": 1.646545584045584, + "grad_norm": 0.6805498003959656, + "learning_rate": 0.00012757689155213923, + "loss": 0.923, + "step": 9248 + }, + { + "epoch": 1.6467236467236468, + "grad_norm": 0.6598179340362549, + "learning_rate": 0.00012756343657637024, + "loss": 0.9371, + "step": 9249 + }, + { + "epoch": 1.6469017094017095, + "grad_norm": 0.7147273421287537, + "learning_rate": 0.00012754998106055297, + "loss": 1.053, + "step": 9250 + }, + { + "epoch": 1.647079772079772, + "grad_norm": 0.72414630651474, + "learning_rate": 0.00012753652500495103, + "loss": 1.0547, + "step": 9251 + }, + { + "epoch": 1.6472578347578346, + "grad_norm": 0.7784913182258606, + "learning_rate": 0.00012752306840982811, + "loss": 0.9012, + "step": 9252 + }, + { + "epoch": 1.6474358974358974, + "grad_norm": 0.644026517868042, + "learning_rate": 0.0001275096112754478, + "loss": 1.0911, + "step": 9253 + }, + { + "epoch": 1.64761396011396, + "grad_norm": 0.691124677658081, + "learning_rate": 0.00012749615360207382, + "loss": 0.9918, + "step": 9254 + }, + { + "epoch": 1.6477920227920229, + "grad_norm": 0.6632972359657288, + "learning_rate": 0.00012748269538996986, + "loss": 0.9438, + "step": 9255 + }, + { + "epoch": 1.6479700854700856, + "grad_norm": 0.6548733115196228, + "learning_rate": 0.00012746923663939955, + "loss": 1.1082, + "step": 9256 + }, + { + "epoch": 1.6481481481481481, + "grad_norm": 0.6737542748451233, + "learning_rate": 0.00012745577735062664, + "loss": 0.9255, + "step": 9257 + }, + { + "epoch": 1.6483262108262107, + "grad_norm": 0.686862051486969, + "learning_rate": 0.00012744231752391479, + "loss": 0.9493, + "step": 9258 + }, + { + "epoch": 1.6485042735042734, + "grad_norm": 0.6096474528312683, + "learning_rate": 0.00012742885715952772, + "loss": 0.6849, + "step": 9259 + }, + { + "epoch": 1.6486823361823362, + "grad_norm": 0.702751636505127, + "learning_rate": 0.00012741539625772918, + "loss": 1.0335, + "step": 9260 + }, + { + "epoch": 1.648860398860399, + "grad_norm": 0.7470958232879639, + "learning_rate": 0.0001274019348187829, + "loss": 1.105, + "step": 9261 + }, + { + "epoch": 1.6490384615384617, + "grad_norm": 0.6642739176750183, + "learning_rate": 0.0001273884728429526, + "loss": 1.01, + "step": 9262 + }, + { + "epoch": 1.6492165242165242, + "grad_norm": 0.6470904350280762, + "learning_rate": 0.00012737501033050213, + "loss": 0.9009, + "step": 9263 + }, + { + "epoch": 1.6493945868945867, + "grad_norm": 0.7487246990203857, + "learning_rate": 0.00012736154728169518, + "loss": 0.9832, + "step": 9264 + }, + { + "epoch": 1.6495726495726495, + "grad_norm": 0.7370779514312744, + "learning_rate": 0.00012734808369679553, + "loss": 1.0464, + "step": 9265 + }, + { + "epoch": 1.6497507122507122, + "grad_norm": 0.7942814826965332, + "learning_rate": 0.00012733461957606702, + "loss": 1.102, + "step": 9266 + }, + { + "epoch": 1.649928774928775, + "grad_norm": 0.6535606980323792, + "learning_rate": 0.00012732115491977336, + "loss": 1.0655, + "step": 9267 + }, + { + "epoch": 1.6501068376068377, + "grad_norm": 0.601716935634613, + "learning_rate": 0.00012730768972817847, + "loss": 0.8236, + "step": 9268 + }, + { + "epoch": 1.6502849002849003, + "grad_norm": 0.7375118732452393, + "learning_rate": 0.00012729422400154614, + "loss": 0.9313, + "step": 9269 + }, + { + "epoch": 1.6504629629629628, + "grad_norm": 0.7360411882400513, + "learning_rate": 0.00012728075774014018, + "loss": 0.9254, + "step": 9270 + }, + { + "epoch": 1.6506410256410255, + "grad_norm": 0.8453929424285889, + "learning_rate": 0.00012726729094422444, + "loss": 1.0975, + "step": 9271 + }, + { + "epoch": 1.6508190883190883, + "grad_norm": 0.5615501999855042, + "learning_rate": 0.00012725382361406274, + "loss": 0.8243, + "step": 9272 + }, + { + "epoch": 1.650997150997151, + "grad_norm": 0.6494898796081543, + "learning_rate": 0.000127240355749919, + "loss": 0.9766, + "step": 9273 + }, + { + "epoch": 1.6511752136752138, + "grad_norm": 0.6544778347015381, + "learning_rate": 0.0001272268873520571, + "loss": 0.9969, + "step": 9274 + }, + { + "epoch": 1.6513532763532763, + "grad_norm": 0.6937400698661804, + "learning_rate": 0.00012721341842074092, + "loss": 1.0626, + "step": 9275 + }, + { + "epoch": 1.651531339031339, + "grad_norm": 0.7068421244621277, + "learning_rate": 0.0001271999489562343, + "loss": 1.0068, + "step": 9276 + }, + { + "epoch": 1.6517094017094016, + "grad_norm": 0.6425052285194397, + "learning_rate": 0.0001271864789588012, + "loss": 0.8716, + "step": 9277 + }, + { + "epoch": 1.6518874643874644, + "grad_norm": 0.6895090341567993, + "learning_rate": 0.0001271730084287055, + "loss": 1.081, + "step": 9278 + }, + { + "epoch": 1.6520655270655271, + "grad_norm": 0.6773712038993835, + "learning_rate": 0.00012715953736621116, + "loss": 0.7586, + "step": 9279 + }, + { + "epoch": 1.6522435897435899, + "grad_norm": 0.6085716485977173, + "learning_rate": 0.0001271460657715821, + "loss": 0.8627, + "step": 9280 + }, + { + "epoch": 1.6524216524216524, + "grad_norm": 0.6415461897850037, + "learning_rate": 0.00012713259364508227, + "loss": 0.9751, + "step": 9281 + }, + { + "epoch": 1.6525997150997151, + "grad_norm": 0.6460939645767212, + "learning_rate": 0.00012711912098697565, + "loss": 0.9578, + "step": 9282 + }, + { + "epoch": 1.6527777777777777, + "grad_norm": 0.6076797246932983, + "learning_rate": 0.00012710564779752615, + "loss": 0.9627, + "step": 9283 + }, + { + "epoch": 1.6529558404558404, + "grad_norm": 0.710782527923584, + "learning_rate": 0.00012709217407699783, + "loss": 0.8725, + "step": 9284 + }, + { + "epoch": 1.6531339031339032, + "grad_norm": 0.6793623566627502, + "learning_rate": 0.00012707869982565463, + "loss": 0.908, + "step": 9285 + }, + { + "epoch": 1.653311965811966, + "grad_norm": 0.6841681003570557, + "learning_rate": 0.00012706522504376055, + "loss": 0.8546, + "step": 9286 + }, + { + "epoch": 1.6534900284900285, + "grad_norm": 0.7908675670623779, + "learning_rate": 0.0001270517497315796, + "loss": 0.9409, + "step": 9287 + }, + { + "epoch": 1.6536680911680912, + "grad_norm": 0.6918683648109436, + "learning_rate": 0.0001270382738893758, + "loss": 1.0493, + "step": 9288 + }, + { + "epoch": 1.6538461538461537, + "grad_norm": 0.6891819834709167, + "learning_rate": 0.00012702479751741322, + "loss": 1.0675, + "step": 9289 + }, + { + "epoch": 1.6540242165242165, + "grad_norm": 0.6965166926383972, + "learning_rate": 0.00012701132061595586, + "loss": 0.8563, + "step": 9290 + }, + { + "epoch": 1.6542022792022792, + "grad_norm": 0.7549001574516296, + "learning_rate": 0.00012699784318526779, + "loss": 1.1572, + "step": 9291 + }, + { + "epoch": 1.654380341880342, + "grad_norm": 0.6100513339042664, + "learning_rate": 0.00012698436522561303, + "loss": 0.897, + "step": 9292 + }, + { + "epoch": 1.6545584045584045, + "grad_norm": 0.6477037668228149, + "learning_rate": 0.00012697088673725574, + "loss": 0.7961, + "step": 9293 + }, + { + "epoch": 1.6547364672364673, + "grad_norm": 0.7402619123458862, + "learning_rate": 0.0001269574077204599, + "loss": 1.2001, + "step": 9294 + }, + { + "epoch": 1.6549145299145298, + "grad_norm": 0.7162346243858337, + "learning_rate": 0.0001269439281754897, + "loss": 0.9963, + "step": 9295 + }, + { + "epoch": 1.6550925925925926, + "grad_norm": 0.6757413744926453, + "learning_rate": 0.0001269304481026092, + "loss": 1.0476, + "step": 9296 + }, + { + "epoch": 1.6552706552706553, + "grad_norm": 0.6455655097961426, + "learning_rate": 0.0001269169675020825, + "loss": 0.9716, + "step": 9297 + }, + { + "epoch": 1.655448717948718, + "grad_norm": 0.7705031037330627, + "learning_rate": 0.0001269034863741737, + "loss": 0.9886, + "step": 9298 + }, + { + "epoch": 1.6556267806267806, + "grad_norm": 0.6084272861480713, + "learning_rate": 0.000126890004719147, + "loss": 0.8231, + "step": 9299 + }, + { + "epoch": 1.6558048433048433, + "grad_norm": 0.7051045298576355, + "learning_rate": 0.00012687652253726652, + "loss": 0.8673, + "step": 9300 + }, + { + "epoch": 1.6559829059829059, + "grad_norm": 0.731675386428833, + "learning_rate": 0.0001268630398287964, + "loss": 0.8609, + "step": 9301 + }, + { + "epoch": 1.6561609686609686, + "grad_norm": 0.6796799302101135, + "learning_rate": 0.00012684955659400087, + "loss": 1.0157, + "step": 9302 + }, + { + "epoch": 1.6563390313390314, + "grad_norm": 0.6270264983177185, + "learning_rate": 0.000126836072833144, + "loss": 0.8924, + "step": 9303 + }, + { + "epoch": 1.6565170940170941, + "grad_norm": 0.7235464453697205, + "learning_rate": 0.00012682258854649004, + "loss": 0.8904, + "step": 9304 + }, + { + "epoch": 1.6566951566951567, + "grad_norm": 0.7644724249839783, + "learning_rate": 0.00012680910373430318, + "loss": 0.9119, + "step": 9305 + }, + { + "epoch": 1.6568732193732194, + "grad_norm": 0.661411702632904, + "learning_rate": 0.00012679561839684764, + "loss": 1.0066, + "step": 9306 + }, + { + "epoch": 1.657051282051282, + "grad_norm": 0.6981723308563232, + "learning_rate": 0.0001267821325343876, + "loss": 1.2579, + "step": 9307 + }, + { + "epoch": 1.6572293447293447, + "grad_norm": 0.6469807028770447, + "learning_rate": 0.0001267686461471873, + "loss": 0.8678, + "step": 9308 + }, + { + "epoch": 1.6574074074074074, + "grad_norm": 0.8255495429039001, + "learning_rate": 0.000126755159235511, + "loss": 0.9053, + "step": 9309 + }, + { + "epoch": 1.6575854700854702, + "grad_norm": 0.6882261037826538, + "learning_rate": 0.00012674167179962294, + "loss": 0.8364, + "step": 9310 + }, + { + "epoch": 1.6577635327635327, + "grad_norm": 0.6816701889038086, + "learning_rate": 0.00012672818383978733, + "loss": 0.9627, + "step": 9311 + }, + { + "epoch": 1.6579415954415955, + "grad_norm": 0.6993424892425537, + "learning_rate": 0.00012671469535626852, + "loss": 0.8337, + "step": 9312 + }, + { + "epoch": 1.658119658119658, + "grad_norm": 0.6271458864212036, + "learning_rate": 0.00012670120634933075, + "loss": 0.8322, + "step": 9313 + }, + { + "epoch": 1.6582977207977208, + "grad_norm": 0.7012003660202026, + "learning_rate": 0.00012668771681923827, + "loss": 0.8895, + "step": 9314 + }, + { + "epoch": 1.6584757834757835, + "grad_norm": 0.6704670190811157, + "learning_rate": 0.00012667422676625547, + "loss": 1.0544, + "step": 9315 + }, + { + "epoch": 1.6586538461538463, + "grad_norm": 0.6189491748809814, + "learning_rate": 0.0001266607361906466, + "loss": 0.9623, + "step": 9316 + }, + { + "epoch": 1.6588319088319088, + "grad_norm": 0.7065694332122803, + "learning_rate": 0.000126647245092676, + "loss": 0.8874, + "step": 9317 + }, + { + "epoch": 1.6590099715099715, + "grad_norm": 0.7473452687263489, + "learning_rate": 0.00012663375347260795, + "loss": 1.0576, + "step": 9318 + }, + { + "epoch": 1.659188034188034, + "grad_norm": 0.6839408874511719, + "learning_rate": 0.0001266202613307068, + "loss": 0.9127, + "step": 9319 + }, + { + "epoch": 1.6593660968660968, + "grad_norm": 0.7154020071029663, + "learning_rate": 0.00012660676866723699, + "loss": 1.1174, + "step": 9320 + }, + { + "epoch": 1.6595441595441596, + "grad_norm": 0.7123729586601257, + "learning_rate": 0.0001265932754824628, + "loss": 0.9617, + "step": 9321 + }, + { + "epoch": 1.6597222222222223, + "grad_norm": 0.7537810802459717, + "learning_rate": 0.0001265797817766486, + "loss": 1.0333, + "step": 9322 + }, + { + "epoch": 1.6599002849002849, + "grad_norm": 0.706551730632782, + "learning_rate": 0.00012656628755005884, + "loss": 1.0838, + "step": 9323 + }, + { + "epoch": 1.6600783475783476, + "grad_norm": 0.8104004859924316, + "learning_rate": 0.0001265527928029578, + "loss": 0.9807, + "step": 9324 + }, + { + "epoch": 1.6602564102564101, + "grad_norm": 0.6892881989479065, + "learning_rate": 0.00012653929753560998, + "loss": 0.9941, + "step": 9325 + }, + { + "epoch": 1.6604344729344729, + "grad_norm": 0.5919203758239746, + "learning_rate": 0.00012652580174827974, + "loss": 0.9268, + "step": 9326 + }, + { + "epoch": 1.6606125356125356, + "grad_norm": 0.6715863347053528, + "learning_rate": 0.00012651230544123154, + "loss": 1.0912, + "step": 9327 + }, + { + "epoch": 1.6607905982905984, + "grad_norm": 0.6765137314796448, + "learning_rate": 0.0001264988086147298, + "loss": 1.1576, + "step": 9328 + }, + { + "epoch": 1.660968660968661, + "grad_norm": 0.6781638860702515, + "learning_rate": 0.00012648531126903888, + "loss": 1.1162, + "step": 9329 + }, + { + "epoch": 1.6611467236467237, + "grad_norm": 0.715871274471283, + "learning_rate": 0.00012647181340442337, + "loss": 0.714, + "step": 9330 + }, + { + "epoch": 1.6613247863247862, + "grad_norm": 0.6237258315086365, + "learning_rate": 0.00012645831502114762, + "loss": 0.8512, + "step": 9331 + }, + { + "epoch": 1.661502849002849, + "grad_norm": 0.6668339967727661, + "learning_rate": 0.0001264448161194762, + "loss": 1.0384, + "step": 9332 + }, + { + "epoch": 1.6616809116809117, + "grad_norm": 0.8316730260848999, + "learning_rate": 0.00012643131669967352, + "loss": 0.8931, + "step": 9333 + }, + { + "epoch": 1.6618589743589745, + "grad_norm": 0.7013183832168579, + "learning_rate": 0.00012641781676200406, + "loss": 1.0548, + "step": 9334 + }, + { + "epoch": 1.6620370370370372, + "grad_norm": 0.6980466842651367, + "learning_rate": 0.00012640431630673243, + "loss": 0.8988, + "step": 9335 + }, + { + "epoch": 1.6622150997150997, + "grad_norm": 0.7045995593070984, + "learning_rate": 0.000126390815334123, + "loss": 1.107, + "step": 9336 + }, + { + "epoch": 1.6623931623931623, + "grad_norm": 0.6699773669242859, + "learning_rate": 0.00012637731384444043, + "loss": 1.1757, + "step": 9337 + }, + { + "epoch": 1.662571225071225, + "grad_norm": 0.6489999294281006, + "learning_rate": 0.00012636381183794916, + "loss": 0.9282, + "step": 9338 + }, + { + "epoch": 1.6627492877492878, + "grad_norm": 0.7085952758789062, + "learning_rate": 0.00012635030931491375, + "loss": 1.0221, + "step": 9339 + }, + { + "epoch": 1.6629273504273505, + "grad_norm": 0.6893135905265808, + "learning_rate": 0.00012633680627559878, + "loss": 1.0517, + "step": 9340 + }, + { + "epoch": 1.6631054131054133, + "grad_norm": 0.5659682154655457, + "learning_rate": 0.00012632330272026882, + "loss": 0.6294, + "step": 9341 + }, + { + "epoch": 1.6632834757834758, + "grad_norm": 0.6889018416404724, + "learning_rate": 0.00012630979864918838, + "loss": 1.0735, + "step": 9342 + }, + { + "epoch": 1.6634615384615383, + "grad_norm": 0.7333424687385559, + "learning_rate": 0.00012629629406262212, + "loss": 0.9079, + "step": 9343 + }, + { + "epoch": 1.663639601139601, + "grad_norm": 0.6340580582618713, + "learning_rate": 0.00012628278896083462, + "loss": 0.9738, + "step": 9344 + }, + { + "epoch": 1.6638176638176638, + "grad_norm": 0.7042564749717712, + "learning_rate": 0.00012626928334409044, + "loss": 0.959, + "step": 9345 + }, + { + "epoch": 1.6639957264957266, + "grad_norm": 0.711757242679596, + "learning_rate": 0.00012625577721265424, + "loss": 0.8113, + "step": 9346 + }, + { + "epoch": 1.6641737891737893, + "grad_norm": 0.7723299264907837, + "learning_rate": 0.0001262422705667906, + "loss": 1.1724, + "step": 9347 + }, + { + "epoch": 1.6643518518518519, + "grad_norm": 0.711334228515625, + "learning_rate": 0.00012622876340676422, + "loss": 1.0121, + "step": 9348 + }, + { + "epoch": 1.6645299145299144, + "grad_norm": 0.6954590678215027, + "learning_rate": 0.0001262152557328397, + "loss": 1.2093, + "step": 9349 + }, + { + "epoch": 1.6647079772079771, + "grad_norm": 0.6341620087623596, + "learning_rate": 0.00012620174754528166, + "loss": 1.0535, + "step": 9350 + }, + { + "epoch": 1.66488603988604, + "grad_norm": 0.6434268355369568, + "learning_rate": 0.00012618823884435484, + "loss": 0.8964, + "step": 9351 + }, + { + "epoch": 1.6650641025641026, + "grad_norm": 0.7685084939002991, + "learning_rate": 0.00012617472963032385, + "loss": 1.0639, + "step": 9352 + }, + { + "epoch": 1.6652421652421654, + "grad_norm": 0.6347958445549011, + "learning_rate": 0.00012616121990345345, + "loss": 1.0252, + "step": 9353 + }, + { + "epoch": 1.665420227920228, + "grad_norm": 0.647722601890564, + "learning_rate": 0.0001261477096640083, + "loss": 0.9527, + "step": 9354 + }, + { + "epoch": 1.6655982905982905, + "grad_norm": 0.5942047834396362, + "learning_rate": 0.000126134198912253, + "loss": 1.0062, + "step": 9355 + }, + { + "epoch": 1.6657763532763532, + "grad_norm": 0.683555006980896, + "learning_rate": 0.00012612068764845247, + "loss": 0.8101, + "step": 9356 + }, + { + "epoch": 1.665954415954416, + "grad_norm": 0.6832289099693298, + "learning_rate": 0.00012610717587287128, + "loss": 1.1436, + "step": 9357 + }, + { + "epoch": 1.6661324786324787, + "grad_norm": 0.7035253047943115, + "learning_rate": 0.00012609366358577422, + "loss": 0.9724, + "step": 9358 + }, + { + "epoch": 1.6663105413105415, + "grad_norm": 0.6471409797668457, + "learning_rate": 0.00012608015078742604, + "loss": 0.776, + "step": 9359 + }, + { + "epoch": 1.666488603988604, + "grad_norm": 0.7069687247276306, + "learning_rate": 0.00012606663747809145, + "loss": 0.9667, + "step": 9360 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.6744135618209839, + "learning_rate": 0.00012605312365803525, + "loss": 1.1152, + "step": 9361 + }, + { + "epoch": 1.6668447293447293, + "grad_norm": 0.7212334275245667, + "learning_rate": 0.00012603960932752227, + "loss": 1.1543, + "step": 9362 + }, + { + "epoch": 1.667022792022792, + "grad_norm": 0.6501669883728027, + "learning_rate": 0.0001260260944868172, + "loss": 0.8595, + "step": 9363 + }, + { + "epoch": 1.6672008547008548, + "grad_norm": 0.6970864534378052, + "learning_rate": 0.00012601257913618486, + "loss": 0.9364, + "step": 9364 + }, + { + "epoch": 1.6673789173789175, + "grad_norm": 0.6802223324775696, + "learning_rate": 0.00012599906327589007, + "loss": 0.8429, + "step": 9365 + }, + { + "epoch": 1.66755698005698, + "grad_norm": 0.6842933893203735, + "learning_rate": 0.00012598554690619764, + "loss": 1.1255, + "step": 9366 + }, + { + "epoch": 1.6677350427350426, + "grad_norm": 0.6547088623046875, + "learning_rate": 0.0001259720300273724, + "loss": 0.983, + "step": 9367 + }, + { + "epoch": 1.6679131054131053, + "grad_norm": 0.620424211025238, + "learning_rate": 0.0001259585126396792, + "loss": 0.918, + "step": 9368 + }, + { + "epoch": 1.668091168091168, + "grad_norm": 0.5659816861152649, + "learning_rate": 0.00012594499474338287, + "loss": 0.7788, + "step": 9369 + }, + { + "epoch": 1.6682692307692308, + "grad_norm": 0.5904595255851746, + "learning_rate": 0.00012593147633874826, + "loss": 0.801, + "step": 9370 + }, + { + "epoch": 1.6684472934472936, + "grad_norm": 0.6444024443626404, + "learning_rate": 0.0001259179574260402, + "loss": 1.0997, + "step": 9371 + }, + { + "epoch": 1.6686253561253561, + "grad_norm": 0.6408827304840088, + "learning_rate": 0.00012590443800552365, + "loss": 0.9839, + "step": 9372 + }, + { + "epoch": 1.6688034188034186, + "grad_norm": 0.752391517162323, + "learning_rate": 0.00012589091807746345, + "loss": 1.0249, + "step": 9373 + }, + { + "epoch": 1.6689814814814814, + "grad_norm": 0.8256397247314453, + "learning_rate": 0.00012587739764212448, + "loss": 0.9541, + "step": 9374 + }, + { + "epoch": 1.6691595441595442, + "grad_norm": 0.7878768444061279, + "learning_rate": 0.00012586387669977166, + "loss": 1.0071, + "step": 9375 + }, + { + "epoch": 1.669337606837607, + "grad_norm": 0.6179735660552979, + "learning_rate": 0.0001258503552506699, + "loss": 0.8495, + "step": 9376 + }, + { + "epoch": 1.6695156695156697, + "grad_norm": 0.6699580550193787, + "learning_rate": 0.00012583683329508413, + "loss": 0.8999, + "step": 9377 + }, + { + "epoch": 1.6696937321937322, + "grad_norm": 0.6542006731033325, + "learning_rate": 0.00012582331083327929, + "loss": 1.0357, + "step": 9378 + }, + { + "epoch": 1.6698717948717947, + "grad_norm": 0.7275210618972778, + "learning_rate": 0.0001258097878655203, + "loss": 1.0259, + "step": 9379 + }, + { + "epoch": 1.6700498575498575, + "grad_norm": 0.6836326122283936, + "learning_rate": 0.00012579626439207216, + "loss": 1.0428, + "step": 9380 + }, + { + "epoch": 1.6702279202279202, + "grad_norm": 0.760123610496521, + "learning_rate": 0.00012578274041319978, + "loss": 0.9716, + "step": 9381 + }, + { + "epoch": 1.670405982905983, + "grad_norm": 0.5525194406509399, + "learning_rate": 0.00012576921592916818, + "loss": 0.8253, + "step": 9382 + }, + { + "epoch": 1.6705840455840457, + "grad_norm": 0.6881270408630371, + "learning_rate": 0.00012575569094024232, + "loss": 1.0571, + "step": 9383 + }, + { + "epoch": 1.6707621082621082, + "grad_norm": 0.6776245832443237, + "learning_rate": 0.0001257421654466872, + "loss": 0.9119, + "step": 9384 + }, + { + "epoch": 1.6709401709401708, + "grad_norm": 0.7903014421463013, + "learning_rate": 0.0001257286394487678, + "loss": 1.0626, + "step": 9385 + }, + { + "epoch": 1.6711182336182335, + "grad_norm": 0.61158287525177, + "learning_rate": 0.0001257151129467492, + "loss": 0.9378, + "step": 9386 + }, + { + "epoch": 1.6712962962962963, + "grad_norm": 0.655189573764801, + "learning_rate": 0.00012570158594089637, + "loss": 0.9334, + "step": 9387 + }, + { + "epoch": 1.671474358974359, + "grad_norm": 0.6707320809364319, + "learning_rate": 0.0001256880584314743, + "loss": 1.1802, + "step": 9388 + }, + { + "epoch": 1.6716524216524218, + "grad_norm": 0.847341775894165, + "learning_rate": 0.00012567453041874814, + "loss": 1.1169, + "step": 9389 + }, + { + "epoch": 1.6718304843304843, + "grad_norm": 0.6136410236358643, + "learning_rate": 0.00012566100190298287, + "loss": 0.8959, + "step": 9390 + }, + { + "epoch": 1.672008547008547, + "grad_norm": 0.7203437089920044, + "learning_rate": 0.00012564747288444357, + "loss": 0.9803, + "step": 9391 + }, + { + "epoch": 1.6721866096866096, + "grad_norm": 0.7832576632499695, + "learning_rate": 0.00012563394336339534, + "loss": 0.8696, + "step": 9392 + }, + { + "epoch": 1.6723646723646723, + "grad_norm": 0.6940804719924927, + "learning_rate": 0.00012562041334010323, + "loss": 1.0571, + "step": 9393 + }, + { + "epoch": 1.672542735042735, + "grad_norm": 0.6042298674583435, + "learning_rate": 0.00012560688281483234, + "loss": 0.8835, + "step": 9394 + }, + { + "epoch": 1.6727207977207978, + "grad_norm": 0.7870675921440125, + "learning_rate": 0.00012559335178784776, + "loss": 1.1585, + "step": 9395 + }, + { + "epoch": 1.6728988603988604, + "grad_norm": 0.7448568940162659, + "learning_rate": 0.00012557982025941463, + "loss": 0.9699, + "step": 9396 + }, + { + "epoch": 1.6730769230769231, + "grad_norm": 0.7226544618606567, + "learning_rate": 0.00012556628822979807, + "loss": 0.7817, + "step": 9397 + }, + { + "epoch": 1.6732549857549857, + "grad_norm": 0.5652043223381042, + "learning_rate": 0.0001255527556992632, + "loss": 0.8077, + "step": 9398 + }, + { + "epoch": 1.6734330484330484, + "grad_norm": 0.6459930539131165, + "learning_rate": 0.00012553922266807517, + "loss": 1.22, + "step": 9399 + }, + { + "epoch": 1.6736111111111112, + "grad_norm": 0.7568991780281067, + "learning_rate": 0.00012552568913649912, + "loss": 1.1559, + "step": 9400 + }, + { + "epoch": 1.673789173789174, + "grad_norm": 0.7462680339813232, + "learning_rate": 0.0001255121551048002, + "loss": 1.1438, + "step": 9401 + }, + { + "epoch": 1.6739672364672364, + "grad_norm": 0.6653871536254883, + "learning_rate": 0.0001254986205732436, + "loss": 0.9468, + "step": 9402 + }, + { + "epoch": 1.6741452991452992, + "grad_norm": 0.6261825561523438, + "learning_rate": 0.0001254850855420945, + "loss": 0.8558, + "step": 9403 + }, + { + "epoch": 1.6743233618233617, + "grad_norm": 0.6442354321479797, + "learning_rate": 0.0001254715500116181, + "loss": 0.8605, + "step": 9404 + }, + { + "epoch": 1.6745014245014245, + "grad_norm": 0.7483665943145752, + "learning_rate": 0.00012545801398207958, + "loss": 0.9089, + "step": 9405 + }, + { + "epoch": 1.6746794871794872, + "grad_norm": 0.7319819927215576, + "learning_rate": 0.00012544447745374416, + "loss": 0.9937, + "step": 9406 + }, + { + "epoch": 1.67485754985755, + "grad_norm": 0.703014075756073, + "learning_rate": 0.00012543094042687708, + "loss": 0.9597, + "step": 9407 + }, + { + "epoch": 1.6750356125356125, + "grad_norm": 0.6593887209892273, + "learning_rate": 0.00012541740290174353, + "loss": 0.844, + "step": 9408 + }, + { + "epoch": 1.6752136752136753, + "grad_norm": 0.6567463874816895, + "learning_rate": 0.00012540386487860879, + "loss": 1.0744, + "step": 9409 + }, + { + "epoch": 1.6753917378917378, + "grad_norm": 0.7784611582756042, + "learning_rate": 0.00012539032635773805, + "loss": 0.974, + "step": 9410 + }, + { + "epoch": 1.6755698005698005, + "grad_norm": 0.6760087609291077, + "learning_rate": 0.00012537678733939663, + "loss": 0.8948, + "step": 9411 + }, + { + "epoch": 1.6757478632478633, + "grad_norm": 0.825965940952301, + "learning_rate": 0.0001253632478238498, + "loss": 1.1196, + "step": 9412 + }, + { + "epoch": 1.675925925925926, + "grad_norm": 0.7215564250946045, + "learning_rate": 0.00012534970781136277, + "loss": 1.1774, + "step": 9413 + }, + { + "epoch": 1.6761039886039886, + "grad_norm": 0.6548578143119812, + "learning_rate": 0.00012533616730220094, + "loss": 0.8671, + "step": 9414 + }, + { + "epoch": 1.6762820512820513, + "grad_norm": 0.7257684469223022, + "learning_rate": 0.00012532262629662947, + "loss": 1.105, + "step": 9415 + }, + { + "epoch": 1.6764601139601139, + "grad_norm": 0.6695847511291504, + "learning_rate": 0.00012530908479491378, + "loss": 0.9189, + "step": 9416 + }, + { + "epoch": 1.6766381766381766, + "grad_norm": 0.684695303440094, + "learning_rate": 0.00012529554279731915, + "loss": 1.066, + "step": 9417 + }, + { + "epoch": 1.6768162393162394, + "grad_norm": 0.7107276320457458, + "learning_rate": 0.0001252820003041109, + "loss": 0.9311, + "step": 9418 + }, + { + "epoch": 1.676994301994302, + "grad_norm": 0.6755440831184387, + "learning_rate": 0.0001252684573155544, + "loss": 1.1036, + "step": 9419 + }, + { + "epoch": 1.6771723646723646, + "grad_norm": 0.7571110725402832, + "learning_rate": 0.00012525491383191491, + "loss": 1.0244, + "step": 9420 + }, + { + "epoch": 1.6773504273504274, + "grad_norm": 0.6960614323616028, + "learning_rate": 0.0001252413698534579, + "loss": 0.9077, + "step": 9421 + }, + { + "epoch": 1.67752849002849, + "grad_norm": 0.6675550937652588, + "learning_rate": 0.00012522782538044867, + "loss": 1.0543, + "step": 9422 + }, + { + "epoch": 1.6777065527065527, + "grad_norm": 0.6637391448020935, + "learning_rate": 0.0001252142804131526, + "loss": 0.9471, + "step": 9423 + }, + { + "epoch": 1.6778846153846154, + "grad_norm": 0.6382880210876465, + "learning_rate": 0.00012520073495183508, + "loss": 0.9729, + "step": 9424 + }, + { + "epoch": 1.6780626780626782, + "grad_norm": 0.731922447681427, + "learning_rate": 0.0001251871889967615, + "loss": 1.0385, + "step": 9425 + }, + { + "epoch": 1.6782407407407407, + "grad_norm": 0.5868890285491943, + "learning_rate": 0.00012517364254819728, + "loss": 0.8466, + "step": 9426 + }, + { + "epoch": 1.6784188034188035, + "grad_norm": 0.8535677790641785, + "learning_rate": 0.00012516009560640786, + "loss": 1.1009, + "step": 9427 + }, + { + "epoch": 1.678596866096866, + "grad_norm": 0.7044199705123901, + "learning_rate": 0.0001251465481716586, + "loss": 1.0862, + "step": 9428 + }, + { + "epoch": 1.6787749287749287, + "grad_norm": 0.7207323312759399, + "learning_rate": 0.00012513300024421498, + "loss": 1.064, + "step": 9429 + }, + { + "epoch": 1.6789529914529915, + "grad_norm": 0.7739703059196472, + "learning_rate": 0.0001251194518243424, + "loss": 1.1738, + "step": 9430 + }, + { + "epoch": 1.6791310541310542, + "grad_norm": 0.6829344630241394, + "learning_rate": 0.00012510590291230637, + "loss": 1.0555, + "step": 9431 + }, + { + "epoch": 1.6793091168091168, + "grad_norm": 0.6760238409042358, + "learning_rate": 0.0001250923535083723, + "loss": 1.2177, + "step": 9432 + }, + { + "epoch": 1.6794871794871795, + "grad_norm": 0.6666911840438843, + "learning_rate": 0.0001250788036128057, + "loss": 0.8957, + "step": 9433 + }, + { + "epoch": 1.679665242165242, + "grad_norm": 0.747797429561615, + "learning_rate": 0.00012506525322587207, + "loss": 0.9793, + "step": 9434 + }, + { + "epoch": 1.6798433048433048, + "grad_norm": 0.6261107325553894, + "learning_rate": 0.00012505170234783686, + "loss": 0.7781, + "step": 9435 + }, + { + "epoch": 1.6800213675213675, + "grad_norm": 0.7055163979530334, + "learning_rate": 0.00012503815097896555, + "loss": 1.0617, + "step": 9436 + }, + { + "epoch": 1.6801994301994303, + "grad_norm": 0.5567409992218018, + "learning_rate": 0.00012502459911952371, + "loss": 0.7911, + "step": 9437 + }, + { + "epoch": 1.6803774928774928, + "grad_norm": 0.7410423159599304, + "learning_rate": 0.0001250110467697768, + "loss": 1.1041, + "step": 9438 + }, + { + "epoch": 1.6805555555555556, + "grad_norm": 0.6185283064842224, + "learning_rate": 0.00012499749392999045, + "loss": 0.8101, + "step": 9439 + }, + { + "epoch": 1.680733618233618, + "grad_norm": 0.6988311409950256, + "learning_rate": 0.0001249839406004301, + "loss": 0.8579, + "step": 9440 + }, + { + "epoch": 1.6809116809116809, + "grad_norm": 0.5588746070861816, + "learning_rate": 0.00012497038678136132, + "loss": 0.8035, + "step": 9441 + }, + { + "epoch": 1.6810897435897436, + "grad_norm": 0.6568905711174011, + "learning_rate": 0.0001249568324730497, + "loss": 0.7455, + "step": 9442 + }, + { + "epoch": 1.6812678062678064, + "grad_norm": 0.6924821138381958, + "learning_rate": 0.00012494327767576078, + "loss": 1.134, + "step": 9443 + }, + { + "epoch": 1.681445868945869, + "grad_norm": 0.6940170526504517, + "learning_rate": 0.00012492972238976018, + "loss": 0.9719, + "step": 9444 + }, + { + "epoch": 1.6816239316239316, + "grad_norm": 0.667465090751648, + "learning_rate": 0.00012491616661531343, + "loss": 0.953, + "step": 9445 + }, + { + "epoch": 1.6818019943019942, + "grad_norm": 0.7693275809288025, + "learning_rate": 0.00012490261035268612, + "loss": 1.1342, + "step": 9446 + }, + { + "epoch": 1.681980056980057, + "grad_norm": 0.7243115305900574, + "learning_rate": 0.00012488905360214393, + "loss": 1.1847, + "step": 9447 + }, + { + "epoch": 1.6821581196581197, + "grad_norm": 0.657357931137085, + "learning_rate": 0.00012487549636395245, + "loss": 0.8747, + "step": 9448 + }, + { + "epoch": 1.6823361823361824, + "grad_norm": 0.7471592426300049, + "learning_rate": 0.00012486193863837727, + "loss": 1.0472, + "step": 9449 + }, + { + "epoch": 1.6825142450142452, + "grad_norm": 0.7476530075073242, + "learning_rate": 0.00012484838042568406, + "loss": 1.0708, + "step": 9450 + }, + { + "epoch": 1.6826923076923077, + "grad_norm": 0.6031121611595154, + "learning_rate": 0.00012483482172613846, + "loss": 0.8243, + "step": 9451 + }, + { + "epoch": 1.6828703703703702, + "grad_norm": 0.6733492016792297, + "learning_rate": 0.00012482126254000607, + "loss": 0.7808, + "step": 9452 + }, + { + "epoch": 1.683048433048433, + "grad_norm": 0.5865318179130554, + "learning_rate": 0.00012480770286755265, + "loss": 0.829, + "step": 9453 + }, + { + "epoch": 1.6832264957264957, + "grad_norm": 0.6805713772773743, + "learning_rate": 0.0001247941427090438, + "loss": 0.7206, + "step": 9454 + }, + { + "epoch": 1.6834045584045585, + "grad_norm": 0.6514836549758911, + "learning_rate": 0.0001247805820647453, + "loss": 0.9499, + "step": 9455 + }, + { + "epoch": 1.6835826210826212, + "grad_norm": 0.7432990074157715, + "learning_rate": 0.0001247670209349227, + "loss": 1.1324, + "step": 9456 + }, + { + "epoch": 1.6837606837606838, + "grad_norm": 0.6348414421081543, + "learning_rate": 0.00012475345931984178, + "loss": 0.8246, + "step": 9457 + }, + { + "epoch": 1.6839387464387463, + "grad_norm": 0.7194374203681946, + "learning_rate": 0.00012473989721976825, + "loss": 0.9634, + "step": 9458 + }, + { + "epoch": 1.684116809116809, + "grad_norm": 0.7869647741317749, + "learning_rate": 0.00012472633463496785, + "loss": 1.2115, + "step": 9459 + }, + { + "epoch": 1.6842948717948718, + "grad_norm": 0.6672070026397705, + "learning_rate": 0.00012471277156570623, + "loss": 0.9842, + "step": 9460 + }, + { + "epoch": 1.6844729344729346, + "grad_norm": 0.6611466407775879, + "learning_rate": 0.00012469920801224925, + "loss": 0.9343, + "step": 9461 + }, + { + "epoch": 1.6846509971509973, + "grad_norm": 0.6715068221092224, + "learning_rate": 0.0001246856439748626, + "loss": 0.6852, + "step": 9462 + }, + { + "epoch": 1.6848290598290598, + "grad_norm": 0.641942024230957, + "learning_rate": 0.00012467207945381198, + "loss": 0.8863, + "step": 9463 + }, + { + "epoch": 1.6850071225071224, + "grad_norm": 0.8414762616157532, + "learning_rate": 0.00012465851444936325, + "loss": 1.3404, + "step": 9464 + }, + { + "epoch": 1.6851851851851851, + "grad_norm": 0.715752363204956, + "learning_rate": 0.00012464494896178216, + "loss": 1.123, + "step": 9465 + }, + { + "epoch": 1.6853632478632479, + "grad_norm": 0.6913973093032837, + "learning_rate": 0.00012463138299133447, + "loss": 1.0659, + "step": 9466 + }, + { + "epoch": 1.6855413105413106, + "grad_norm": 0.6998484134674072, + "learning_rate": 0.000124617816538286, + "loss": 1.0555, + "step": 9467 + }, + { + "epoch": 1.6857193732193734, + "grad_norm": 0.7313308119773865, + "learning_rate": 0.00012460424960290256, + "loss": 1.0915, + "step": 9468 + }, + { + "epoch": 1.685897435897436, + "grad_norm": 0.6790569424629211, + "learning_rate": 0.00012459068218544995, + "loss": 1.0214, + "step": 9469 + }, + { + "epoch": 1.6860754985754984, + "grad_norm": 0.6494466662406921, + "learning_rate": 0.00012457711428619402, + "loss": 0.9476, + "step": 9470 + }, + { + "epoch": 1.6862535612535612, + "grad_norm": 0.8048526048660278, + "learning_rate": 0.0001245635459054006, + "loss": 1.1852, + "step": 9471 + }, + { + "epoch": 1.686431623931624, + "grad_norm": 0.6237879395484924, + "learning_rate": 0.0001245499770433355, + "loss": 1.0106, + "step": 9472 + }, + { + "epoch": 1.6866096866096867, + "grad_norm": 0.6282906532287598, + "learning_rate": 0.0001245364077002646, + "loss": 0.9858, + "step": 9473 + }, + { + "epoch": 1.6867877492877494, + "grad_norm": 0.7239370346069336, + "learning_rate": 0.00012452283787645375, + "loss": 0.9586, + "step": 9474 + }, + { + "epoch": 1.686965811965812, + "grad_norm": 0.6438776850700378, + "learning_rate": 0.00012450926757216887, + "loss": 0.9198, + "step": 9475 + }, + { + "epoch": 1.6871438746438745, + "grad_norm": 0.6451360583305359, + "learning_rate": 0.00012449569678767578, + "loss": 1.0183, + "step": 9476 + }, + { + "epoch": 1.6873219373219372, + "grad_norm": 0.6950216293334961, + "learning_rate": 0.0001244821255232404, + "loss": 0.9048, + "step": 9477 + }, + { + "epoch": 1.6875, + "grad_norm": 0.710489809513092, + "learning_rate": 0.00012446855377912865, + "loss": 1.1596, + "step": 9478 + }, + { + "epoch": 1.6876780626780628, + "grad_norm": 0.6819305419921875, + "learning_rate": 0.0001244549815556064, + "loss": 0.8486, + "step": 9479 + }, + { + "epoch": 1.6878561253561255, + "grad_norm": 0.7185879945755005, + "learning_rate": 0.00012444140885293958, + "loss": 0.9539, + "step": 9480 + }, + { + "epoch": 1.688034188034188, + "grad_norm": 0.8181464672088623, + "learning_rate": 0.00012442783567139415, + "loss": 1.0038, + "step": 9481 + }, + { + "epoch": 1.6882122507122506, + "grad_norm": 0.47161349654197693, + "learning_rate": 0.000124414262011236, + "loss": 0.67, + "step": 9482 + }, + { + "epoch": 1.6883903133903133, + "grad_norm": 0.7752482295036316, + "learning_rate": 0.00012440068787273112, + "loss": 0.9944, + "step": 9483 + }, + { + "epoch": 1.688568376068376, + "grad_norm": 0.7119397521018982, + "learning_rate": 0.00012438711325614543, + "loss": 0.9098, + "step": 9484 + }, + { + "epoch": 1.6887464387464388, + "grad_norm": 0.7161153554916382, + "learning_rate": 0.00012437353816174493, + "loss": 1.0003, + "step": 9485 + }, + { + "epoch": 1.6889245014245016, + "grad_norm": 0.5989507436752319, + "learning_rate": 0.0001243599625897956, + "loss": 1.0301, + "step": 9486 + }, + { + "epoch": 1.689102564102564, + "grad_norm": 0.7906841039657593, + "learning_rate": 0.00012434638654056334, + "loss": 1.0388, + "step": 9487 + }, + { + "epoch": 1.6892806267806266, + "grad_norm": 0.6679551601409912, + "learning_rate": 0.00012433281001431428, + "loss": 0.9505, + "step": 9488 + }, + { + "epoch": 1.6894586894586894, + "grad_norm": 0.7090578675270081, + "learning_rate": 0.0001243192330113143, + "loss": 0.8616, + "step": 9489 + }, + { + "epoch": 1.6896367521367521, + "grad_norm": 0.6401308178901672, + "learning_rate": 0.00012430565553182949, + "loss": 0.9099, + "step": 9490 + }, + { + "epoch": 1.6898148148148149, + "grad_norm": 0.7360149621963501, + "learning_rate": 0.00012429207757612586, + "loss": 1.0233, + "step": 9491 + }, + { + "epoch": 1.6899928774928776, + "grad_norm": 0.6736137270927429, + "learning_rate": 0.00012427849914446946, + "loss": 0.9803, + "step": 9492 + }, + { + "epoch": 1.6901709401709402, + "grad_norm": 0.7728668451309204, + "learning_rate": 0.00012426492023712623, + "loss": 1.2316, + "step": 9493 + }, + { + "epoch": 1.6903490028490027, + "grad_norm": 0.789718508720398, + "learning_rate": 0.00012425134085436234, + "loss": 1.1218, + "step": 9494 + }, + { + "epoch": 1.6905270655270654, + "grad_norm": 0.7314121723175049, + "learning_rate": 0.0001242377609964438, + "loss": 1.1294, + "step": 9495 + }, + { + "epoch": 1.6907051282051282, + "grad_norm": 0.7222046256065369, + "learning_rate": 0.0001242241806636367, + "loss": 1.0288, + "step": 9496 + }, + { + "epoch": 1.690883190883191, + "grad_norm": 0.7546363472938538, + "learning_rate": 0.00012421059985620708, + "loss": 0.8781, + "step": 9497 + }, + { + "epoch": 1.6910612535612537, + "grad_norm": 0.7502550482749939, + "learning_rate": 0.00012419701857442104, + "loss": 0.927, + "step": 9498 + }, + { + "epoch": 1.6912393162393162, + "grad_norm": 0.6244059205055237, + "learning_rate": 0.00012418343681854473, + "loss": 0.9689, + "step": 9499 + }, + { + "epoch": 1.6914173789173788, + "grad_norm": 0.7214263677597046, + "learning_rate": 0.00012416985458884417, + "loss": 1.0842, + "step": 9500 + }, + { + "epoch": 1.6915954415954415, + "grad_norm": 0.6960242390632629, + "learning_rate": 0.00012415627188558555, + "loss": 0.9766, + "step": 9501 + }, + { + "epoch": 1.6917735042735043, + "grad_norm": 0.6687830686569214, + "learning_rate": 0.00012414268870903494, + "loss": 1.0222, + "step": 9502 + }, + { + "epoch": 1.691951566951567, + "grad_norm": 0.8611155152320862, + "learning_rate": 0.00012412910505945848, + "loss": 1.1792, + "step": 9503 + }, + { + "epoch": 1.6921296296296298, + "grad_norm": 0.6655587553977966, + "learning_rate": 0.00012411552093712235, + "loss": 0.8763, + "step": 9504 + }, + { + "epoch": 1.6923076923076923, + "grad_norm": 0.7829837799072266, + "learning_rate": 0.00012410193634229268, + "loss": 1.0803, + "step": 9505 + }, + { + "epoch": 1.6924857549857548, + "grad_norm": 0.7951042652130127, + "learning_rate": 0.00012408835127523566, + "loss": 1.0925, + "step": 9506 + }, + { + "epoch": 1.6926638176638176, + "grad_norm": 0.715495228767395, + "learning_rate": 0.0001240747657362174, + "loss": 1.2411, + "step": 9507 + }, + { + "epoch": 1.6928418803418803, + "grad_norm": 0.6779513359069824, + "learning_rate": 0.00012406117972550414, + "loss": 0.8886, + "step": 9508 + }, + { + "epoch": 1.693019943019943, + "grad_norm": 0.647588312625885, + "learning_rate": 0.00012404759324336203, + "loss": 1.107, + "step": 9509 + }, + { + "epoch": 1.6931980056980058, + "grad_norm": 0.7398989796638489, + "learning_rate": 0.00012403400629005726, + "loss": 1.0256, + "step": 9510 + }, + { + "epoch": 1.6933760683760684, + "grad_norm": 0.7572638392448425, + "learning_rate": 0.0001240204188658561, + "loss": 0.9662, + "step": 9511 + }, + { + "epoch": 1.693554131054131, + "grad_norm": 0.7044163346290588, + "learning_rate": 0.00012400683097102473, + "loss": 1.1388, + "step": 9512 + }, + { + "epoch": 1.6937321937321936, + "grad_norm": 0.7889094948768616, + "learning_rate": 0.00012399324260582936, + "loss": 1.0453, + "step": 9513 + }, + { + "epoch": 1.6939102564102564, + "grad_norm": 0.7977854609489441, + "learning_rate": 0.00012397965377053627, + "loss": 1.015, + "step": 9514 + }, + { + "epoch": 1.6940883190883191, + "grad_norm": 0.6223814487457275, + "learning_rate": 0.00012396606446541165, + "loss": 0.7985, + "step": 9515 + }, + { + "epoch": 1.694266381766382, + "grad_norm": 0.8307462334632874, + "learning_rate": 0.0001239524746907218, + "loss": 0.8899, + "step": 9516 + }, + { + "epoch": 1.6944444444444444, + "grad_norm": 0.7780544757843018, + "learning_rate": 0.00012393888444673295, + "loss": 0.9406, + "step": 9517 + }, + { + "epoch": 1.6946225071225072, + "grad_norm": 0.6894499659538269, + "learning_rate": 0.0001239252937337114, + "loss": 0.9412, + "step": 9518 + }, + { + "epoch": 1.6948005698005697, + "grad_norm": 0.7000680565834045, + "learning_rate": 0.00012391170255192342, + "loss": 1.0314, + "step": 9519 + }, + { + "epoch": 1.6949786324786325, + "grad_norm": 0.6772416830062866, + "learning_rate": 0.0001238981109016353, + "loss": 0.9153, + "step": 9520 + }, + { + "epoch": 1.6951566951566952, + "grad_norm": 0.7069609761238098, + "learning_rate": 0.00012388451878311333, + "loss": 1.1777, + "step": 9521 + }, + { + "epoch": 1.695334757834758, + "grad_norm": 0.6138432621955872, + "learning_rate": 0.00012387092619662386, + "loss": 0.8085, + "step": 9522 + }, + { + "epoch": 1.6955128205128205, + "grad_norm": 0.6122859716415405, + "learning_rate": 0.00012385733314243313, + "loss": 0.8534, + "step": 9523 + }, + { + "epoch": 1.6956908831908832, + "grad_norm": 0.7499903440475464, + "learning_rate": 0.00012384373962080755, + "loss": 0.9329, + "step": 9524 + }, + { + "epoch": 1.6958689458689458, + "grad_norm": 0.6413441896438599, + "learning_rate": 0.00012383014563201343, + "loss": 0.9609, + "step": 9525 + }, + { + "epoch": 1.6960470085470085, + "grad_norm": 0.7467969059944153, + "learning_rate": 0.0001238165511763171, + "loss": 0.9142, + "step": 9526 + }, + { + "epoch": 1.6962250712250713, + "grad_norm": 0.6540884375572205, + "learning_rate": 0.00012380295625398494, + "loss": 0.9503, + "step": 9527 + }, + { + "epoch": 1.696403133903134, + "grad_norm": 0.6298567652702332, + "learning_rate": 0.00012378936086528326, + "loss": 0.8853, + "step": 9528 + }, + { + "epoch": 1.6965811965811965, + "grad_norm": 0.8003417253494263, + "learning_rate": 0.00012377576501047845, + "loss": 0.969, + "step": 9529 + }, + { + "epoch": 1.6967592592592593, + "grad_norm": 0.8318493962287903, + "learning_rate": 0.00012376216868983697, + "loss": 1.1413, + "step": 9530 + }, + { + "epoch": 1.6969373219373218, + "grad_norm": 0.8294426202774048, + "learning_rate": 0.00012374857190362515, + "loss": 1.1885, + "step": 9531 + }, + { + "epoch": 1.6971153846153846, + "grad_norm": 0.7502955198287964, + "learning_rate": 0.0001237349746521094, + "loss": 1.233, + "step": 9532 + }, + { + "epoch": 1.6972934472934473, + "grad_norm": 0.6306588649749756, + "learning_rate": 0.00012372137693555612, + "loss": 1.2255, + "step": 9533 + }, + { + "epoch": 1.69747150997151, + "grad_norm": 0.7802746891975403, + "learning_rate": 0.0001237077787542317, + "loss": 1.2054, + "step": 9534 + }, + { + "epoch": 1.6976495726495726, + "grad_norm": 0.685114860534668, + "learning_rate": 0.00012369418010840265, + "loss": 0.9865, + "step": 9535 + }, + { + "epoch": 1.6978276353276354, + "grad_norm": 0.6656857132911682, + "learning_rate": 0.00012368058099833536, + "loss": 1.1579, + "step": 9536 + }, + { + "epoch": 1.698005698005698, + "grad_norm": 0.6596674919128418, + "learning_rate": 0.00012366698142429625, + "loss": 0.9104, + "step": 9537 + }, + { + "epoch": 1.6981837606837606, + "grad_norm": 0.6025584936141968, + "learning_rate": 0.00012365338138655183, + "loss": 1.117, + "step": 9538 + }, + { + "epoch": 1.6983618233618234, + "grad_norm": 0.671585202217102, + "learning_rate": 0.0001236397808853685, + "loss": 1.0271, + "step": 9539 + }, + { + "epoch": 1.6985398860398861, + "grad_norm": 0.7467984557151794, + "learning_rate": 0.0001236261799210128, + "loss": 1.0411, + "step": 9540 + }, + { + "epoch": 1.6987179487179487, + "grad_norm": 0.6251640915870667, + "learning_rate": 0.0001236125784937512, + "loss": 0.7154, + "step": 9541 + }, + { + "epoch": 1.6988960113960114, + "grad_norm": 0.7560956478118896, + "learning_rate": 0.00012359897660385016, + "loss": 1.0048, + "step": 9542 + }, + { + "epoch": 1.699074074074074, + "grad_norm": 0.6144903302192688, + "learning_rate": 0.00012358537425157618, + "loss": 1.1294, + "step": 9543 + }, + { + "epoch": 1.6992521367521367, + "grad_norm": 0.7839425206184387, + "learning_rate": 0.00012357177143719578, + "loss": 1.0725, + "step": 9544 + }, + { + "epoch": 1.6994301994301995, + "grad_norm": 0.6488651037216187, + "learning_rate": 0.00012355816816097553, + "loss": 0.9267, + "step": 9545 + }, + { + "epoch": 1.6996082621082622, + "grad_norm": 0.6848782896995544, + "learning_rate": 0.00012354456442318187, + "loss": 1.0426, + "step": 9546 + }, + { + "epoch": 1.6997863247863247, + "grad_norm": 0.7164611220359802, + "learning_rate": 0.0001235309602240814, + "loss": 0.8208, + "step": 9547 + }, + { + "epoch": 1.6999643874643875, + "grad_norm": 0.6725530624389648, + "learning_rate": 0.0001235173555639406, + "loss": 0.9366, + "step": 9548 + }, + { + "epoch": 1.70014245014245, + "grad_norm": 0.6958004236221313, + "learning_rate": 0.00012350375044302612, + "loss": 1.0185, + "step": 9549 + }, + { + "epoch": 1.7003205128205128, + "grad_norm": 0.8035947680473328, + "learning_rate": 0.00012349014486160445, + "loss": 1.065, + "step": 9550 + }, + { + "epoch": 1.7004985754985755, + "grad_norm": 0.6705633997917175, + "learning_rate": 0.00012347653881994222, + "loss": 0.8381, + "step": 9551 + }, + { + "epoch": 1.7006766381766383, + "grad_norm": 0.6652300357818604, + "learning_rate": 0.00012346293231830596, + "loss": 1.1428, + "step": 9552 + }, + { + "epoch": 1.7008547008547008, + "grad_norm": 0.6719335913658142, + "learning_rate": 0.0001234493253569623, + "loss": 1.0138, + "step": 9553 + }, + { + "epoch": 1.7010327635327636, + "grad_norm": 0.746981680393219, + "learning_rate": 0.0001234357179361778, + "loss": 1.1169, + "step": 9554 + }, + { + "epoch": 1.701210826210826, + "grad_norm": 0.6768170595169067, + "learning_rate": 0.0001234221100562191, + "loss": 0.9065, + "step": 9555 + }, + { + "epoch": 1.7013888888888888, + "grad_norm": 0.7127171754837036, + "learning_rate": 0.00012340850171735278, + "loss": 0.9467, + "step": 9556 + }, + { + "epoch": 1.7015669515669516, + "grad_norm": 0.6802694797515869, + "learning_rate": 0.00012339489291984554, + "loss": 0.8938, + "step": 9557 + }, + { + "epoch": 1.7017450142450143, + "grad_norm": 0.7101455926895142, + "learning_rate": 0.00012338128366396394, + "loss": 1.1939, + "step": 9558 + }, + { + "epoch": 1.7019230769230769, + "grad_norm": 0.621223509311676, + "learning_rate": 0.00012336767394997467, + "loss": 0.7583, + "step": 9559 + }, + { + "epoch": 1.7021011396011396, + "grad_norm": 0.7130763530731201, + "learning_rate": 0.00012335406377814439, + "loss": 0.8684, + "step": 9560 + }, + { + "epoch": 1.7022792022792022, + "grad_norm": 0.6761086583137512, + "learning_rate": 0.00012334045314873972, + "loss": 1.0197, + "step": 9561 + }, + { + "epoch": 1.702457264957265, + "grad_norm": 0.7030459642410278, + "learning_rate": 0.00012332684206202736, + "loss": 0.8627, + "step": 9562 + }, + { + "epoch": 1.7026353276353277, + "grad_norm": 0.6278037428855896, + "learning_rate": 0.000123313230518274, + "loss": 0.8953, + "step": 9563 + }, + { + "epoch": 1.7028133903133904, + "grad_norm": 0.6450623869895935, + "learning_rate": 0.00012329961851774627, + "loss": 0.8826, + "step": 9564 + }, + { + "epoch": 1.702991452991453, + "grad_norm": 0.7324244976043701, + "learning_rate": 0.00012328600606071097, + "loss": 1.0133, + "step": 9565 + }, + { + "epoch": 1.7031695156695157, + "grad_norm": 0.6560033559799194, + "learning_rate": 0.00012327239314743473, + "loss": 0.9601, + "step": 9566 + }, + { + "epoch": 1.7033475783475782, + "grad_norm": 0.6693514585494995, + "learning_rate": 0.0001232587797781843, + "loss": 0.9447, + "step": 9567 + }, + { + "epoch": 1.703525641025641, + "grad_norm": 0.6403199434280396, + "learning_rate": 0.00012324516595322638, + "loss": 0.8554, + "step": 9568 + }, + { + "epoch": 1.7037037037037037, + "grad_norm": 0.8290280103683472, + "learning_rate": 0.00012323155167282774, + "loss": 1.1877, + "step": 9569 + }, + { + "epoch": 1.7038817663817665, + "grad_norm": 0.7207778692245483, + "learning_rate": 0.00012321793693725509, + "loss": 1.0978, + "step": 9570 + }, + { + "epoch": 1.7040598290598292, + "grad_norm": 0.8794265985488892, + "learning_rate": 0.00012320432174677519, + "loss": 0.9387, + "step": 9571 + }, + { + "epoch": 1.7042378917378918, + "grad_norm": 0.6683359146118164, + "learning_rate": 0.00012319070610165484, + "loss": 0.9227, + "step": 9572 + }, + { + "epoch": 1.7044159544159543, + "grad_norm": 0.7342001795768738, + "learning_rate": 0.00012317709000216076, + "loss": 0.9453, + "step": 9573 + }, + { + "epoch": 1.704594017094017, + "grad_norm": 0.6315770149230957, + "learning_rate": 0.00012316347344855973, + "loss": 0.8263, + "step": 9574 + }, + { + "epoch": 1.7047720797720798, + "grad_norm": 0.7697155475616455, + "learning_rate": 0.00012314985644111857, + "loss": 1.0238, + "step": 9575 + }, + { + "epoch": 1.7049501424501425, + "grad_norm": 0.6674068570137024, + "learning_rate": 0.00012313623898010408, + "loss": 1.0823, + "step": 9576 + }, + { + "epoch": 1.7051282051282053, + "grad_norm": 0.6995484232902527, + "learning_rate": 0.00012312262106578304, + "loss": 1.2001, + "step": 9577 + }, + { + "epoch": 1.7053062678062678, + "grad_norm": 0.7639257907867432, + "learning_rate": 0.00012310900269842226, + "loss": 1.3438, + "step": 9578 + }, + { + "epoch": 1.7054843304843303, + "grad_norm": 0.6486390233039856, + "learning_rate": 0.00012309538387828857, + "loss": 0.9924, + "step": 9579 + }, + { + "epoch": 1.705662393162393, + "grad_norm": 0.6737813949584961, + "learning_rate": 0.00012308176460564885, + "loss": 0.8722, + "step": 9580 + }, + { + "epoch": 1.7058404558404558, + "grad_norm": 0.6462090611457825, + "learning_rate": 0.00012306814488076987, + "loss": 1.1013, + "step": 9581 + }, + { + "epoch": 1.7060185185185186, + "grad_norm": 0.7887832522392273, + "learning_rate": 0.00012305452470391852, + "loss": 0.9998, + "step": 9582 + }, + { + "epoch": 1.7061965811965814, + "grad_norm": 0.6345070004463196, + "learning_rate": 0.00012304090407536165, + "loss": 1.0305, + "step": 9583 + }, + { + "epoch": 1.7063746438746439, + "grad_norm": 0.6398460268974304, + "learning_rate": 0.0001230272829953661, + "loss": 1.2243, + "step": 9584 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 0.6501944065093994, + "learning_rate": 0.00012301366146419879, + "loss": 0.9425, + "step": 9585 + }, + { + "epoch": 1.7067307692307692, + "grad_norm": 0.6406761407852173, + "learning_rate": 0.00012300003948212661, + "loss": 0.948, + "step": 9586 + }, + { + "epoch": 1.706908831908832, + "grad_norm": 0.7114266157150269, + "learning_rate": 0.00012298641704941644, + "loss": 1.1291, + "step": 9587 + }, + { + "epoch": 1.7070868945868947, + "grad_norm": 0.6653099656105042, + "learning_rate": 0.00012297279416633515, + "loss": 1.0156, + "step": 9588 + }, + { + "epoch": 1.7072649572649574, + "grad_norm": 0.5970917344093323, + "learning_rate": 0.0001229591708331497, + "loss": 0.9424, + "step": 9589 + }, + { + "epoch": 1.70744301994302, + "grad_norm": 0.6861461400985718, + "learning_rate": 0.00012294554705012694, + "loss": 0.7581, + "step": 9590 + }, + { + "epoch": 1.7076210826210825, + "grad_norm": 0.6930568218231201, + "learning_rate": 0.00012293192281753393, + "loss": 1.0544, + "step": 9591 + }, + { + "epoch": 1.7077991452991452, + "grad_norm": 0.7420656085014343, + "learning_rate": 0.00012291829813563748, + "loss": 0.7092, + "step": 9592 + }, + { + "epoch": 1.707977207977208, + "grad_norm": 0.6607801914215088, + "learning_rate": 0.0001229046730047046, + "loss": 0.5544, + "step": 9593 + }, + { + "epoch": 1.7081552706552707, + "grad_norm": 0.8419139385223389, + "learning_rate": 0.00012289104742500224, + "loss": 1.0443, + "step": 9594 + }, + { + "epoch": 1.7083333333333335, + "grad_norm": 0.6774617433547974, + "learning_rate": 0.00012287742139679734, + "loss": 1.0098, + "step": 9595 + }, + { + "epoch": 1.708511396011396, + "grad_norm": 0.7517698407173157, + "learning_rate": 0.0001228637949203569, + "loss": 1.1145, + "step": 9596 + }, + { + "epoch": 1.7086894586894585, + "grad_norm": 0.6048635840415955, + "learning_rate": 0.00012285016799594791, + "loss": 0.7398, + "step": 9597 + }, + { + "epoch": 1.7088675213675213, + "grad_norm": 0.8054425716400146, + "learning_rate": 0.00012283654062383734, + "loss": 1.0893, + "step": 9598 + }, + { + "epoch": 1.709045584045584, + "grad_norm": 0.8694897294044495, + "learning_rate": 0.0001228229128042922, + "loss": 1.2366, + "step": 9599 + }, + { + "epoch": 1.7092236467236468, + "grad_norm": 0.7460638880729675, + "learning_rate": 0.00012280928453757946, + "loss": 1.1753, + "step": 9600 + }, + { + "epoch": 1.7094017094017095, + "grad_norm": 0.6714958548545837, + "learning_rate": 0.00012279565582396618, + "loss": 1.0473, + "step": 9601 + }, + { + "epoch": 1.709579772079772, + "grad_norm": 0.6893340945243835, + "learning_rate": 0.00012278202666371937, + "loss": 1.2761, + "step": 9602 + }, + { + "epoch": 1.7097578347578346, + "grad_norm": 0.6816153526306152, + "learning_rate": 0.00012276839705710612, + "loss": 0.991, + "step": 9603 + }, + { + "epoch": 1.7099358974358974, + "grad_norm": 0.6961633563041687, + "learning_rate": 0.0001227547670043934, + "loss": 1.0634, + "step": 9604 + }, + { + "epoch": 1.71011396011396, + "grad_norm": 0.643734872341156, + "learning_rate": 0.0001227411365058483, + "loss": 0.8672, + "step": 9605 + }, + { + "epoch": 1.7102920227920229, + "grad_norm": 0.7313315272331238, + "learning_rate": 0.00012272750556173784, + "loss": 1.1152, + "step": 9606 + }, + { + "epoch": 1.7104700854700856, + "grad_norm": 0.6464954614639282, + "learning_rate": 0.00012271387417232916, + "loss": 0.8798, + "step": 9607 + }, + { + "epoch": 1.7106481481481481, + "grad_norm": 0.8365204334259033, + "learning_rate": 0.00012270024233788929, + "loss": 1.213, + "step": 9608 + }, + { + "epoch": 1.7108262108262107, + "grad_norm": 0.6460705995559692, + "learning_rate": 0.0001226866100586853, + "loss": 0.9232, + "step": 9609 + }, + { + "epoch": 1.7110042735042734, + "grad_norm": 0.6446022987365723, + "learning_rate": 0.00012267297733498434, + "loss": 0.8295, + "step": 9610 + }, + { + "epoch": 1.7111823361823362, + "grad_norm": 0.7692012190818787, + "learning_rate": 0.00012265934416705345, + "loss": 1.0715, + "step": 9611 + }, + { + "epoch": 1.711360398860399, + "grad_norm": 0.671154260635376, + "learning_rate": 0.0001226457105551598, + "loss": 0.9752, + "step": 9612 + }, + { + "epoch": 1.7115384615384617, + "grad_norm": 0.6525935530662537, + "learning_rate": 0.00012263207649957053, + "loss": 1.09, + "step": 9613 + }, + { + "epoch": 1.7117165242165242, + "grad_norm": 0.6984749436378479, + "learning_rate": 0.0001226184420005527, + "loss": 0.9956, + "step": 9614 + }, + { + "epoch": 1.7118945868945867, + "grad_norm": 0.6769809126853943, + "learning_rate": 0.0001226048070583735, + "loss": 1.0151, + "step": 9615 + }, + { + "epoch": 1.7120726495726495, + "grad_norm": 0.6085978746414185, + "learning_rate": 0.00012259117167330005, + "loss": 0.8706, + "step": 9616 + }, + { + "epoch": 1.7122507122507122, + "grad_norm": 0.7335749268531799, + "learning_rate": 0.00012257753584559952, + "loss": 1.0575, + "step": 9617 + }, + { + "epoch": 1.712428774928775, + "grad_norm": 0.7392038106918335, + "learning_rate": 0.0001225638995755391, + "loss": 0.8763, + "step": 9618 + }, + { + "epoch": 1.7126068376068377, + "grad_norm": 0.6708608865737915, + "learning_rate": 0.00012255026286338592, + "loss": 1.131, + "step": 9619 + }, + { + "epoch": 1.7127849002849003, + "grad_norm": 0.726657509803772, + "learning_rate": 0.0001225366257094072, + "loss": 1.0569, + "step": 9620 + }, + { + "epoch": 1.7129629629629628, + "grad_norm": 0.749098002910614, + "learning_rate": 0.0001225229881138701, + "loss": 0.9196, + "step": 9621 + }, + { + "epoch": 1.7131410256410255, + "grad_norm": 0.6550580263137817, + "learning_rate": 0.00012250935007704182, + "loss": 1.0244, + "step": 9622 + }, + { + "epoch": 1.7133190883190883, + "grad_norm": 0.7714282274246216, + "learning_rate": 0.00012249571159918962, + "loss": 1.1025, + "step": 9623 + }, + { + "epoch": 1.713497150997151, + "grad_norm": 0.7869850397109985, + "learning_rate": 0.00012248207268058064, + "loss": 0.9238, + "step": 9624 + }, + { + "epoch": 1.7136752136752138, + "grad_norm": 0.7187856435775757, + "learning_rate": 0.00012246843332148216, + "loss": 1.081, + "step": 9625 + }, + { + "epoch": 1.7138532763532763, + "grad_norm": 0.6634210348129272, + "learning_rate": 0.00012245479352216142, + "loss": 1.1944, + "step": 9626 + }, + { + "epoch": 1.714031339031339, + "grad_norm": 0.6609212160110474, + "learning_rate": 0.00012244115328288567, + "loss": 0.9613, + "step": 9627 + }, + { + "epoch": 1.7142094017094016, + "grad_norm": 0.7906867861747742, + "learning_rate": 0.0001224275126039221, + "loss": 1.2692, + "step": 9628 + }, + { + "epoch": 1.7143874643874644, + "grad_norm": 0.8037096858024597, + "learning_rate": 0.000122413871485538, + "loss": 0.9823, + "step": 9629 + }, + { + "epoch": 1.7145655270655271, + "grad_norm": 0.7740145921707153, + "learning_rate": 0.00012240022992800068, + "loss": 1.1937, + "step": 9630 + }, + { + "epoch": 1.7147435897435899, + "grad_norm": 0.595372200012207, + "learning_rate": 0.00012238658793157738, + "loss": 0.9153, + "step": 9631 + }, + { + "epoch": 1.7149216524216524, + "grad_norm": 0.6671900749206543, + "learning_rate": 0.0001223729454965354, + "loss": 1.0895, + "step": 9632 + }, + { + "epoch": 1.7150997150997151, + "grad_norm": 0.5805774927139282, + "learning_rate": 0.000122359302623142, + "loss": 1.0001, + "step": 9633 + }, + { + "epoch": 1.7152777777777777, + "grad_norm": 0.8851602673530579, + "learning_rate": 0.00012234565931166456, + "loss": 1.2828, + "step": 9634 + }, + { + "epoch": 1.7154558404558404, + "grad_norm": 0.6960011720657349, + "learning_rate": 0.0001223320155623703, + "loss": 1.0622, + "step": 9635 + }, + { + "epoch": 1.7156339031339032, + "grad_norm": 0.5587009191513062, + "learning_rate": 0.0001223183713755266, + "loss": 0.83, + "step": 9636 + }, + { + "epoch": 1.715811965811966, + "grad_norm": 0.6892730593681335, + "learning_rate": 0.00012230472675140076, + "loss": 0.9214, + "step": 9637 + }, + { + "epoch": 1.7159900284900285, + "grad_norm": 0.6545090079307556, + "learning_rate": 0.00012229108169026017, + "loss": 0.829, + "step": 9638 + }, + { + "epoch": 1.7161680911680912, + "grad_norm": 0.6539101600646973, + "learning_rate": 0.00012227743619237213, + "loss": 1.0686, + "step": 9639 + }, + { + "epoch": 1.7163461538461537, + "grad_norm": 0.5887274146080017, + "learning_rate": 0.000122263790258004, + "loss": 0.9285, + "step": 9640 + }, + { + "epoch": 1.7165242165242165, + "grad_norm": 0.6328918933868408, + "learning_rate": 0.00012225014388742313, + "loss": 0.9684, + "step": 9641 + }, + { + "epoch": 1.7167022792022792, + "grad_norm": 0.6377436518669128, + "learning_rate": 0.00012223649708089694, + "loss": 0.9425, + "step": 9642 + }, + { + "epoch": 1.716880341880342, + "grad_norm": 0.6967392563819885, + "learning_rate": 0.00012222284983869275, + "loss": 0.9342, + "step": 9643 + }, + { + "epoch": 1.7170584045584045, + "grad_norm": 0.7051317691802979, + "learning_rate": 0.00012220920216107802, + "loss": 1.1843, + "step": 9644 + }, + { + "epoch": 1.7172364672364673, + "grad_norm": 0.6864503622055054, + "learning_rate": 0.00012219555404832007, + "loss": 1.0371, + "step": 9645 + }, + { + "epoch": 1.7174145299145298, + "grad_norm": 0.583454430103302, + "learning_rate": 0.00012218190550068638, + "loss": 0.6774, + "step": 9646 + }, + { + "epoch": 1.7175925925925926, + "grad_norm": 0.6755677461624146, + "learning_rate": 0.0001221682565184443, + "loss": 0.9517, + "step": 9647 + }, + { + "epoch": 1.7177706552706553, + "grad_norm": 0.7230031490325928, + "learning_rate": 0.0001221546071018613, + "loss": 1.0385, + "step": 9648 + }, + { + "epoch": 1.717948717948718, + "grad_norm": 0.7381200194358826, + "learning_rate": 0.0001221409572512048, + "loss": 0.9893, + "step": 9649 + }, + { + "epoch": 1.7181267806267806, + "grad_norm": 0.7079094648361206, + "learning_rate": 0.0001221273069667422, + "loss": 0.7793, + "step": 9650 + }, + { + "epoch": 1.7183048433048433, + "grad_norm": 0.6666881442070007, + "learning_rate": 0.00012211365624874106, + "loss": 0.9752, + "step": 9651 + }, + { + "epoch": 1.7184829059829059, + "grad_norm": 0.6196922659873962, + "learning_rate": 0.00012210000509746868, + "loss": 0.922, + "step": 9652 + }, + { + "epoch": 1.7186609686609686, + "grad_norm": 0.657879650592804, + "learning_rate": 0.00012208635351319266, + "loss": 1.2583, + "step": 9653 + }, + { + "epoch": 1.7188390313390314, + "grad_norm": 0.7240459322929382, + "learning_rate": 0.00012207270149618043, + "loss": 0.8479, + "step": 9654 + }, + { + "epoch": 1.7190170940170941, + "grad_norm": 0.8293825387954712, + "learning_rate": 0.00012205904904669945, + "loss": 0.9092, + "step": 9655 + }, + { + "epoch": 1.7191951566951567, + "grad_norm": 0.6907553672790527, + "learning_rate": 0.0001220453961650172, + "loss": 1.0543, + "step": 9656 + }, + { + "epoch": 1.7193732193732194, + "grad_norm": 0.7178300023078918, + "learning_rate": 0.00012203174285140124, + "loss": 0.9147, + "step": 9657 + }, + { + "epoch": 1.719551282051282, + "grad_norm": 0.7037166357040405, + "learning_rate": 0.00012201808910611905, + "loss": 0.8685, + "step": 9658 + }, + { + "epoch": 1.7197293447293447, + "grad_norm": 0.5850751996040344, + "learning_rate": 0.00012200443492943813, + "loss": 0.72, + "step": 9659 + }, + { + "epoch": 1.7199074074074074, + "grad_norm": 0.744239330291748, + "learning_rate": 0.00012199078032162603, + "loss": 0.9717, + "step": 9660 + }, + { + "epoch": 1.7200854700854702, + "grad_norm": 0.6509126424789429, + "learning_rate": 0.00012197712528295025, + "loss": 0.9768, + "step": 9661 + }, + { + "epoch": 1.7202635327635327, + "grad_norm": 0.623220682144165, + "learning_rate": 0.00012196346981367837, + "loss": 0.9824, + "step": 9662 + }, + { + "epoch": 1.7204415954415955, + "grad_norm": 0.6376451849937439, + "learning_rate": 0.00012194981391407792, + "loss": 0.8228, + "step": 9663 + }, + { + "epoch": 1.720619658119658, + "grad_norm": 0.794830322265625, + "learning_rate": 0.00012193615758441648, + "loss": 0.9168, + "step": 9664 + }, + { + "epoch": 1.7207977207977208, + "grad_norm": 0.7812975645065308, + "learning_rate": 0.0001219225008249616, + "loss": 0.8625, + "step": 9665 + }, + { + "epoch": 1.7209757834757835, + "grad_norm": 0.6843218207359314, + "learning_rate": 0.0001219088436359808, + "loss": 1.0176, + "step": 9666 + }, + { + "epoch": 1.7211538461538463, + "grad_norm": 0.6924905180931091, + "learning_rate": 0.00012189518601774178, + "loss": 0.855, + "step": 9667 + }, + { + "epoch": 1.7213319088319088, + "grad_norm": 0.6348826289176941, + "learning_rate": 0.00012188152797051202, + "loss": 1.1596, + "step": 9668 + }, + { + "epoch": 1.7215099715099715, + "grad_norm": 0.7170482873916626, + "learning_rate": 0.00012186786949455922, + "loss": 0.9811, + "step": 9669 + }, + { + "epoch": 1.721688034188034, + "grad_norm": 0.7471763491630554, + "learning_rate": 0.00012185421059015094, + "loss": 1.0925, + "step": 9670 + }, + { + "epoch": 1.7218660968660968, + "grad_norm": 0.6771119236946106, + "learning_rate": 0.00012184055125755481, + "loss": 0.9403, + "step": 9671 + }, + { + "epoch": 1.7220441595441596, + "grad_norm": 0.4335343539714813, + "learning_rate": 0.0001218268914970384, + "loss": 0.4925, + "step": 9672 + }, + { + "epoch": 1.7222222222222223, + "grad_norm": 0.6652585864067078, + "learning_rate": 0.00012181323130886943, + "loss": 0.7684, + "step": 9673 + }, + { + "epoch": 1.7224002849002849, + "grad_norm": 0.6465467810630798, + "learning_rate": 0.00012179957069331548, + "loss": 0.9011, + "step": 9674 + }, + { + "epoch": 1.7225783475783476, + "grad_norm": 0.6725688576698303, + "learning_rate": 0.00012178590965064427, + "loss": 0.9563, + "step": 9675 + }, + { + "epoch": 1.7227564102564101, + "grad_norm": 0.6223418712615967, + "learning_rate": 0.00012177224818112341, + "loss": 0.9099, + "step": 9676 + }, + { + "epoch": 1.7229344729344729, + "grad_norm": 0.79325270652771, + "learning_rate": 0.00012175858628502053, + "loss": 1.0318, + "step": 9677 + }, + { + "epoch": 1.7231125356125356, + "grad_norm": 0.6735602617263794, + "learning_rate": 0.0001217449239626034, + "loss": 1.0797, + "step": 9678 + }, + { + "epoch": 1.7232905982905984, + "grad_norm": 0.7082492113113403, + "learning_rate": 0.00012173126121413962, + "loss": 1.1341, + "step": 9679 + }, + { + "epoch": 1.723468660968661, + "grad_norm": 0.6563859581947327, + "learning_rate": 0.00012171759803989696, + "loss": 0.8778, + "step": 9680 + }, + { + "epoch": 1.7236467236467237, + "grad_norm": 0.6867792010307312, + "learning_rate": 0.00012170393444014306, + "loss": 0.8301, + "step": 9681 + }, + { + "epoch": 1.7238247863247862, + "grad_norm": 0.7870511412620544, + "learning_rate": 0.00012169027041514562, + "loss": 0.9165, + "step": 9682 + }, + { + "epoch": 1.724002849002849, + "grad_norm": 0.8006493449211121, + "learning_rate": 0.00012167660596517241, + "loss": 1.0395, + "step": 9683 + }, + { + "epoch": 1.7241809116809117, + "grad_norm": 0.6936125159263611, + "learning_rate": 0.00012166294109049114, + "loss": 1.1037, + "step": 9684 + }, + { + "epoch": 1.7243589743589745, + "grad_norm": 0.8176514506340027, + "learning_rate": 0.00012164927579136956, + "loss": 0.8791, + "step": 9685 + }, + { + "epoch": 1.7245370370370372, + "grad_norm": 0.6948300004005432, + "learning_rate": 0.00012163561006807537, + "loss": 0.9292, + "step": 9686 + }, + { + "epoch": 1.7247150997150997, + "grad_norm": 0.6237453818321228, + "learning_rate": 0.00012162194392087634, + "loss": 0.8553, + "step": 9687 + }, + { + "epoch": 1.7248931623931623, + "grad_norm": 0.6198007464408875, + "learning_rate": 0.00012160827735004021, + "loss": 0.9599, + "step": 9688 + }, + { + "epoch": 1.725071225071225, + "grad_norm": 0.639838695526123, + "learning_rate": 0.00012159461035583482, + "loss": 0.9328, + "step": 9689 + }, + { + "epoch": 1.7252492877492878, + "grad_norm": 0.7264436483383179, + "learning_rate": 0.00012158094293852789, + "loss": 1.0247, + "step": 9690 + }, + { + "epoch": 1.7254273504273505, + "grad_norm": 0.6320534348487854, + "learning_rate": 0.00012156727509838721, + "loss": 1.1222, + "step": 9691 + }, + { + "epoch": 1.7256054131054133, + "grad_norm": 0.6204122304916382, + "learning_rate": 0.00012155360683568056, + "loss": 0.9765, + "step": 9692 + }, + { + "epoch": 1.7257834757834758, + "grad_norm": 0.7026457190513611, + "learning_rate": 0.00012153993815067579, + "loss": 1.0178, + "step": 9693 + }, + { + "epoch": 1.7259615384615383, + "grad_norm": 0.6471006870269775, + "learning_rate": 0.00012152626904364067, + "loss": 1.0035, + "step": 9694 + }, + { + "epoch": 1.726139601139601, + "grad_norm": 0.6875706911087036, + "learning_rate": 0.00012151259951484301, + "loss": 0.7921, + "step": 9695 + }, + { + "epoch": 1.7263176638176638, + "grad_norm": 0.6963251233100891, + "learning_rate": 0.00012149892956455067, + "loss": 0.9677, + "step": 9696 + }, + { + "epoch": 1.7264957264957266, + "grad_norm": 0.9077282547950745, + "learning_rate": 0.00012148525919303142, + "loss": 0.9362, + "step": 9697 + }, + { + "epoch": 1.7266737891737893, + "grad_norm": 0.7347434163093567, + "learning_rate": 0.00012147158840055319, + "loss": 0.8712, + "step": 9698 + }, + { + "epoch": 1.7268518518518519, + "grad_norm": 0.7206630110740662, + "learning_rate": 0.00012145791718738377, + "loss": 1.032, + "step": 9699 + }, + { + "epoch": 1.7270299145299144, + "grad_norm": 0.7174662947654724, + "learning_rate": 0.00012144424555379106, + "loss": 0.954, + "step": 9700 + }, + { + "epoch": 1.7272079772079771, + "grad_norm": 0.7442345023155212, + "learning_rate": 0.0001214305735000429, + "loss": 1.0709, + "step": 9701 + }, + { + "epoch": 1.72738603988604, + "grad_norm": 0.6154376268386841, + "learning_rate": 0.00012141690102640715, + "loss": 0.9365, + "step": 9702 + }, + { + "epoch": 1.7275641025641026, + "grad_norm": 0.6213796734809875, + "learning_rate": 0.00012140322813315172, + "loss": 0.8337, + "step": 9703 + }, + { + "epoch": 1.7277421652421654, + "grad_norm": 0.7682011127471924, + "learning_rate": 0.0001213895548205445, + "loss": 1.1579, + "step": 9704 + }, + { + "epoch": 1.727920227920228, + "grad_norm": 0.6796970963478088, + "learning_rate": 0.0001213758810888534, + "loss": 0.8875, + "step": 9705 + }, + { + "epoch": 1.7280982905982905, + "grad_norm": 0.7203732132911682, + "learning_rate": 0.0001213622069383463, + "loss": 0.7827, + "step": 9706 + }, + { + "epoch": 1.7282763532763532, + "grad_norm": 0.6151877045631409, + "learning_rate": 0.00012134853236929111, + "loss": 1.0282, + "step": 9707 + }, + { + "epoch": 1.728454415954416, + "grad_norm": 0.6665124297142029, + "learning_rate": 0.0001213348573819558, + "loss": 1.0636, + "step": 9708 + }, + { + "epoch": 1.7286324786324787, + "grad_norm": 0.7334614396095276, + "learning_rate": 0.00012132118197660829, + "loss": 1.0889, + "step": 9709 + }, + { + "epoch": 1.7288105413105415, + "grad_norm": 0.7267759442329407, + "learning_rate": 0.00012130750615351649, + "loss": 1.096, + "step": 9710 + }, + { + "epoch": 1.728988603988604, + "grad_norm": 0.6542944312095642, + "learning_rate": 0.00012129382991294837, + "loss": 1.0855, + "step": 9711 + }, + { + "epoch": 1.7291666666666665, + "grad_norm": 0.694523274898529, + "learning_rate": 0.00012128015325517193, + "loss": 0.8482, + "step": 9712 + }, + { + "epoch": 1.7293447293447293, + "grad_norm": 0.7879082560539246, + "learning_rate": 0.00012126647618045504, + "loss": 1.2356, + "step": 9713 + }, + { + "epoch": 1.729522792022792, + "grad_norm": 0.7108420729637146, + "learning_rate": 0.00012125279868906574, + "loss": 1.0185, + "step": 9714 + }, + { + "epoch": 1.7297008547008548, + "grad_norm": 0.6928725838661194, + "learning_rate": 0.000121239120781272, + "loss": 1.1507, + "step": 9715 + }, + { + "epoch": 1.7298789173789175, + "grad_norm": 0.6195241212844849, + "learning_rate": 0.00012122544245734182, + "loss": 0.8656, + "step": 9716 + }, + { + "epoch": 1.73005698005698, + "grad_norm": 0.5962017774581909, + "learning_rate": 0.00012121176371754317, + "loss": 0.918, + "step": 9717 + }, + { + "epoch": 1.7302350427350426, + "grad_norm": 0.7409394979476929, + "learning_rate": 0.00012119808456214407, + "loss": 1.0283, + "step": 9718 + }, + { + "epoch": 1.7304131054131053, + "grad_norm": 0.6571973562240601, + "learning_rate": 0.00012118440499141257, + "loss": 1.1015, + "step": 9719 + }, + { + "epoch": 1.730591168091168, + "grad_norm": 0.681394100189209, + "learning_rate": 0.00012117072500561664, + "loss": 0.8247, + "step": 9720 + }, + { + "epoch": 1.7307692307692308, + "grad_norm": 0.7278251647949219, + "learning_rate": 0.00012115704460502432, + "loss": 1.0693, + "step": 9721 + }, + { + "epoch": 1.7309472934472936, + "grad_norm": 0.6569405794143677, + "learning_rate": 0.0001211433637899037, + "loss": 0.8992, + "step": 9722 + }, + { + "epoch": 1.7311253561253561, + "grad_norm": 0.6305136680603027, + "learning_rate": 0.00012112968256052272, + "loss": 0.8543, + "step": 9723 + }, + { + "epoch": 1.7313034188034186, + "grad_norm": 0.6111339330673218, + "learning_rate": 0.00012111600091714956, + "loss": 0.991, + "step": 9724 + }, + { + "epoch": 1.7314814814814814, + "grad_norm": 0.646973192691803, + "learning_rate": 0.00012110231886005223, + "loss": 0.8855, + "step": 9725 + }, + { + "epoch": 1.7316595441595442, + "grad_norm": 0.7054407000541687, + "learning_rate": 0.00012108863638949879, + "loss": 1.0816, + "step": 9726 + }, + { + "epoch": 1.731837606837607, + "grad_norm": 0.6592162847518921, + "learning_rate": 0.00012107495350575729, + "loss": 1.0961, + "step": 9727 + }, + { + "epoch": 1.7320156695156697, + "grad_norm": 0.6615595817565918, + "learning_rate": 0.00012106127020909587, + "loss": 0.9669, + "step": 9728 + }, + { + "epoch": 1.7321937321937322, + "grad_norm": 0.9030881524085999, + "learning_rate": 0.00012104758649978263, + "loss": 0.9438, + "step": 9729 + }, + { + "epoch": 1.7323717948717947, + "grad_norm": 0.6776516437530518, + "learning_rate": 0.00012103390237808566, + "loss": 0.8967, + "step": 9730 + }, + { + "epoch": 1.7325498575498575, + "grad_norm": 0.6010605096817017, + "learning_rate": 0.00012102021784427306, + "loss": 0.8893, + "step": 9731 + }, + { + "epoch": 1.7327279202279202, + "grad_norm": 0.6540384292602539, + "learning_rate": 0.00012100653289861295, + "loss": 0.9328, + "step": 9732 + }, + { + "epoch": 1.732905982905983, + "grad_norm": 0.6836950182914734, + "learning_rate": 0.00012099284754137345, + "loss": 0.9019, + "step": 9733 + }, + { + "epoch": 1.7330840455840457, + "grad_norm": 0.7597874402999878, + "learning_rate": 0.00012097916177282274, + "loss": 1.0093, + "step": 9734 + }, + { + "epoch": 1.7332621082621082, + "grad_norm": 0.7686513066291809, + "learning_rate": 0.00012096547559322892, + "loss": 0.8685, + "step": 9735 + }, + { + "epoch": 1.7334401709401708, + "grad_norm": 0.613777220249176, + "learning_rate": 0.0001209517890028602, + "loss": 0.8317, + "step": 9736 + }, + { + "epoch": 1.7336182336182335, + "grad_norm": 0.6788455843925476, + "learning_rate": 0.00012093810200198466, + "loss": 0.866, + "step": 9737 + }, + { + "epoch": 1.7337962962962963, + "grad_norm": 0.616801381111145, + "learning_rate": 0.00012092441459087047, + "loss": 0.8299, + "step": 9738 + }, + { + "epoch": 1.733974358974359, + "grad_norm": 0.731987476348877, + "learning_rate": 0.00012091072676978589, + "loss": 1.089, + "step": 9739 + }, + { + "epoch": 1.7341524216524218, + "grad_norm": 0.7042871117591858, + "learning_rate": 0.00012089703853899905, + "loss": 0.8667, + "step": 9740 + }, + { + "epoch": 1.7343304843304843, + "grad_norm": 0.62722247838974, + "learning_rate": 0.00012088334989877817, + "loss": 0.9185, + "step": 9741 + }, + { + "epoch": 1.734508547008547, + "grad_norm": 0.6354684829711914, + "learning_rate": 0.0001208696608493914, + "loss": 0.9951, + "step": 9742 + }, + { + "epoch": 1.7346866096866096, + "grad_norm": 0.658647894859314, + "learning_rate": 0.00012085597139110698, + "loss": 0.9324, + "step": 9743 + }, + { + "epoch": 1.7348646723646723, + "grad_norm": 0.84359210729599, + "learning_rate": 0.00012084228152419312, + "loss": 1.0861, + "step": 9744 + }, + { + "epoch": 1.735042735042735, + "grad_norm": 0.6293938755989075, + "learning_rate": 0.00012082859124891807, + "loss": 0.9676, + "step": 9745 + }, + { + "epoch": 1.7352207977207978, + "grad_norm": 0.6398760676383972, + "learning_rate": 0.00012081490056555004, + "loss": 0.8502, + "step": 9746 + }, + { + "epoch": 1.7353988603988604, + "grad_norm": 0.6918041706085205, + "learning_rate": 0.00012080120947435726, + "loss": 1.0081, + "step": 9747 + }, + { + "epoch": 1.7355769230769231, + "grad_norm": 0.7374079823493958, + "learning_rate": 0.00012078751797560798, + "loss": 0.9485, + "step": 9748 + }, + { + "epoch": 1.7357549857549857, + "grad_norm": 0.7392128705978394, + "learning_rate": 0.00012077382606957049, + "loss": 0.9283, + "step": 9749 + }, + { + "epoch": 1.7359330484330484, + "grad_norm": 0.701320230960846, + "learning_rate": 0.00012076013375651303, + "loss": 1.0339, + "step": 9750 + }, + { + "epoch": 1.7361111111111112, + "grad_norm": 0.6316696405410767, + "learning_rate": 0.00012074644103670387, + "loss": 0.9097, + "step": 9751 + }, + { + "epoch": 1.736289173789174, + "grad_norm": 0.6892024278640747, + "learning_rate": 0.00012073274791041132, + "loss": 1.0863, + "step": 9752 + }, + { + "epoch": 1.7364672364672364, + "grad_norm": 0.6032847762107849, + "learning_rate": 0.00012071905437790361, + "loss": 0.9305, + "step": 9753 + }, + { + "epoch": 1.7366452991452992, + "grad_norm": 0.6659184098243713, + "learning_rate": 0.00012070536043944907, + "loss": 0.9793, + "step": 9754 + }, + { + "epoch": 1.7368233618233617, + "grad_norm": 0.7413665056228638, + "learning_rate": 0.00012069166609531602, + "loss": 1.0523, + "step": 9755 + }, + { + "epoch": 1.7370014245014245, + "grad_norm": 0.7814368009567261, + "learning_rate": 0.00012067797134577275, + "loss": 0.9988, + "step": 9756 + }, + { + "epoch": 1.7371794871794872, + "grad_norm": 0.6174948811531067, + "learning_rate": 0.00012066427619108757, + "loss": 0.9002, + "step": 9757 + }, + { + "epoch": 1.73735754985755, + "grad_norm": 0.6521819233894348, + "learning_rate": 0.00012065058063152885, + "loss": 1.1307, + "step": 9758 + }, + { + "epoch": 1.7375356125356125, + "grad_norm": 0.6797493696212769, + "learning_rate": 0.00012063688466736489, + "loss": 0.84, + "step": 9759 + }, + { + "epoch": 1.7377136752136753, + "grad_norm": 0.6496474146842957, + "learning_rate": 0.00012062318829886404, + "loss": 0.86, + "step": 9760 + }, + { + "epoch": 1.7378917378917378, + "grad_norm": 0.6701306104660034, + "learning_rate": 0.00012060949152629467, + "loss": 0.9422, + "step": 9761 + }, + { + "epoch": 1.7380698005698005, + "grad_norm": 0.7331172823905945, + "learning_rate": 0.00012059579434992512, + "loss": 1.1648, + "step": 9762 + }, + { + "epoch": 1.7382478632478633, + "grad_norm": 0.63930743932724, + "learning_rate": 0.00012058209677002375, + "loss": 1.0617, + "step": 9763 + }, + { + "epoch": 1.738425925925926, + "grad_norm": 0.668851912021637, + "learning_rate": 0.00012056839878685895, + "loss": 0.8219, + "step": 9764 + }, + { + "epoch": 1.7386039886039886, + "grad_norm": 0.7305747270584106, + "learning_rate": 0.00012055470040069912, + "loss": 1.0416, + "step": 9765 + }, + { + "epoch": 1.7387820512820513, + "grad_norm": 0.6931866407394409, + "learning_rate": 0.00012054100161181264, + "loss": 1.0588, + "step": 9766 + }, + { + "epoch": 1.7389601139601139, + "grad_norm": 0.6565485000610352, + "learning_rate": 0.00012052730242046785, + "loss": 0.7885, + "step": 9767 + }, + { + "epoch": 1.7391381766381766, + "grad_norm": 0.739985466003418, + "learning_rate": 0.00012051360282693327, + "loss": 1.0973, + "step": 9768 + }, + { + "epoch": 1.7393162393162394, + "grad_norm": 0.6477079391479492, + "learning_rate": 0.00012049990283147723, + "loss": 0.9841, + "step": 9769 + }, + { + "epoch": 1.739494301994302, + "grad_norm": 0.7018330097198486, + "learning_rate": 0.00012048620243436819, + "loss": 1.0869, + "step": 9770 + }, + { + "epoch": 1.7396723646723646, + "grad_norm": 0.7087421417236328, + "learning_rate": 0.00012047250163587456, + "loss": 0.916, + "step": 9771 + }, + { + "epoch": 1.7398504273504274, + "grad_norm": 0.8747151494026184, + "learning_rate": 0.00012045880043626481, + "loss": 0.8245, + "step": 9772 + }, + { + "epoch": 1.74002849002849, + "grad_norm": 0.777498722076416, + "learning_rate": 0.00012044509883580735, + "loss": 1.071, + "step": 9773 + }, + { + "epoch": 1.7402065527065527, + "grad_norm": 0.6668971180915833, + "learning_rate": 0.00012043139683477062, + "loss": 1.0447, + "step": 9774 + }, + { + "epoch": 1.7403846153846154, + "grad_norm": 0.6702026724815369, + "learning_rate": 0.00012041769443342317, + "loss": 0.8688, + "step": 9775 + }, + { + "epoch": 1.7405626780626782, + "grad_norm": 0.7866267561912537, + "learning_rate": 0.00012040399163203337, + "loss": 1.0842, + "step": 9776 + }, + { + "epoch": 1.7407407407407407, + "grad_norm": 0.7655110955238342, + "learning_rate": 0.00012039028843086977, + "loss": 1.2417, + "step": 9777 + }, + { + "epoch": 1.7409188034188035, + "grad_norm": 0.7084119915962219, + "learning_rate": 0.0001203765848302008, + "loss": 0.9844, + "step": 9778 + }, + { + "epoch": 1.741096866096866, + "grad_norm": 0.7135398983955383, + "learning_rate": 0.00012036288083029497, + "loss": 1.1102, + "step": 9779 + }, + { + "epoch": 1.7412749287749287, + "grad_norm": 0.6784615516662598, + "learning_rate": 0.0001203491764314208, + "loss": 1.0349, + "step": 9780 + }, + { + "epoch": 1.7414529914529915, + "grad_norm": 0.7170301079750061, + "learning_rate": 0.00012033547163384682, + "loss": 1.0899, + "step": 9781 + }, + { + "epoch": 1.7416310541310542, + "grad_norm": 0.6692060828208923, + "learning_rate": 0.0001203217664378415, + "loss": 1.0486, + "step": 9782 + }, + { + "epoch": 1.7418091168091168, + "grad_norm": 0.6730037927627563, + "learning_rate": 0.00012030806084367336, + "loss": 0.9684, + "step": 9783 + }, + { + "epoch": 1.7419871794871795, + "grad_norm": 0.5983504056930542, + "learning_rate": 0.00012029435485161096, + "loss": 0.7106, + "step": 9784 + }, + { + "epoch": 1.742165242165242, + "grad_norm": 0.6834231615066528, + "learning_rate": 0.00012028064846192284, + "loss": 0.803, + "step": 9785 + }, + { + "epoch": 1.7423433048433048, + "grad_norm": 0.621046245098114, + "learning_rate": 0.00012026694167487755, + "loss": 0.9129, + "step": 9786 + }, + { + "epoch": 1.7425213675213675, + "grad_norm": 0.6348989605903625, + "learning_rate": 0.00012025323449074361, + "loss": 1.0076, + "step": 9787 + }, + { + "epoch": 1.7426994301994303, + "grad_norm": 0.6139974594116211, + "learning_rate": 0.00012023952690978966, + "loss": 1.0756, + "step": 9788 + }, + { + "epoch": 1.7428774928774928, + "grad_norm": 0.6473259925842285, + "learning_rate": 0.00012022581893228419, + "loss": 1.0568, + "step": 9789 + }, + { + "epoch": 1.7430555555555556, + "grad_norm": 0.6133778095245361, + "learning_rate": 0.00012021211055849581, + "loss": 0.8722, + "step": 9790 + }, + { + "epoch": 1.743233618233618, + "grad_norm": 0.6934139728546143, + "learning_rate": 0.00012019840178869315, + "loss": 1.0329, + "step": 9791 + }, + { + "epoch": 1.7434116809116809, + "grad_norm": 0.6730150580406189, + "learning_rate": 0.00012018469262314474, + "loss": 0.9326, + "step": 9792 + }, + { + "epoch": 1.7435897435897436, + "grad_norm": 0.6805521249771118, + "learning_rate": 0.0001201709830621192, + "loss": 1.0527, + "step": 9793 + }, + { + "epoch": 1.7437678062678064, + "grad_norm": 0.6972569823265076, + "learning_rate": 0.00012015727310588516, + "loss": 1.0024, + "step": 9794 + }, + { + "epoch": 1.743945868945869, + "grad_norm": 0.7329187989234924, + "learning_rate": 0.00012014356275471122, + "loss": 1.1864, + "step": 9795 + }, + { + "epoch": 1.7441239316239316, + "grad_norm": 0.7220240831375122, + "learning_rate": 0.00012012985200886602, + "loss": 0.8831, + "step": 9796 + }, + { + "epoch": 1.7443019943019942, + "grad_norm": 0.7829749584197998, + "learning_rate": 0.00012011614086861818, + "loss": 1.0365, + "step": 9797 + }, + { + "epoch": 1.744480056980057, + "grad_norm": 0.7148944735527039, + "learning_rate": 0.00012010242933423637, + "loss": 1.0413, + "step": 9798 + }, + { + "epoch": 1.7446581196581197, + "grad_norm": 0.5607262253761292, + "learning_rate": 0.00012008871740598917, + "loss": 0.8154, + "step": 9799 + }, + { + "epoch": 1.7448361823361824, + "grad_norm": 0.754626452922821, + "learning_rate": 0.00012007500508414531, + "loss": 1.0569, + "step": 9800 + }, + { + "epoch": 1.7450142450142452, + "grad_norm": 0.7216293215751648, + "learning_rate": 0.00012006129236897343, + "loss": 1.1641, + "step": 9801 + }, + { + "epoch": 1.7451923076923077, + "grad_norm": 0.6575515270233154, + "learning_rate": 0.0001200475792607422, + "loss": 0.9063, + "step": 9802 + }, + { + "epoch": 1.7453703703703702, + "grad_norm": 0.7411505579948425, + "learning_rate": 0.00012003386575972031, + "loss": 0.9791, + "step": 9803 + }, + { + "epoch": 1.745548433048433, + "grad_norm": 0.6945903301239014, + "learning_rate": 0.0001200201518661764, + "loss": 0.8111, + "step": 9804 + }, + { + "epoch": 1.7457264957264957, + "grad_norm": 0.5760970115661621, + "learning_rate": 0.00012000643758037924, + "loss": 1.1054, + "step": 9805 + }, + { + "epoch": 1.7459045584045585, + "grad_norm": 0.6732224225997925, + "learning_rate": 0.00011999272290259748, + "loss": 0.8992, + "step": 9806 + }, + { + "epoch": 1.7460826210826212, + "grad_norm": 0.673270046710968, + "learning_rate": 0.00011997900783309983, + "loss": 1.0554, + "step": 9807 + }, + { + "epoch": 1.7462606837606838, + "grad_norm": 0.7233314514160156, + "learning_rate": 0.00011996529237215503, + "loss": 1.066, + "step": 9808 + }, + { + "epoch": 1.7464387464387463, + "grad_norm": 0.7016494274139404, + "learning_rate": 0.00011995157652003183, + "loss": 0.891, + "step": 9809 + }, + { + "epoch": 1.746616809116809, + "grad_norm": 0.9377092719078064, + "learning_rate": 0.00011993786027699889, + "loss": 0.8626, + "step": 9810 + }, + { + "epoch": 1.7467948717948718, + "grad_norm": 0.6825845241546631, + "learning_rate": 0.00011992414364332503, + "loss": 0.8996, + "step": 9811 + }, + { + "epoch": 1.7469729344729346, + "grad_norm": 0.6836053729057312, + "learning_rate": 0.00011991042661927896, + "loss": 0.9338, + "step": 9812 + }, + { + "epoch": 1.7471509971509973, + "grad_norm": 0.6462908387184143, + "learning_rate": 0.00011989670920512943, + "loss": 1.1185, + "step": 9813 + }, + { + "epoch": 1.7473290598290598, + "grad_norm": 0.7191921472549438, + "learning_rate": 0.00011988299140114522, + "loss": 0.9084, + "step": 9814 + }, + { + "epoch": 1.7475071225071224, + "grad_norm": 0.6951598525047302, + "learning_rate": 0.00011986927320759508, + "loss": 1.0653, + "step": 9815 + }, + { + "epoch": 1.7476851851851851, + "grad_norm": 0.7512598037719727, + "learning_rate": 0.00011985555462474784, + "loss": 1.0259, + "step": 9816 + }, + { + "epoch": 1.7478632478632479, + "grad_norm": 0.6885492205619812, + "learning_rate": 0.00011984183565287226, + "loss": 0.7148, + "step": 9817 + }, + { + "epoch": 1.7480413105413106, + "grad_norm": 0.6880139708518982, + "learning_rate": 0.00011982811629223709, + "loss": 1.1567, + "step": 9818 + }, + { + "epoch": 1.7482193732193734, + "grad_norm": 0.7381170392036438, + "learning_rate": 0.0001198143965431112, + "loss": 0.8483, + "step": 9819 + }, + { + "epoch": 1.748397435897436, + "grad_norm": 0.6761063933372498, + "learning_rate": 0.00011980067640576333, + "loss": 0.9498, + "step": 9820 + }, + { + "epoch": 1.7485754985754984, + "grad_norm": 0.6454669237136841, + "learning_rate": 0.00011978695588046238, + "loss": 0.7336, + "step": 9821 + }, + { + "epoch": 1.7487535612535612, + "grad_norm": 0.6026871800422668, + "learning_rate": 0.00011977323496747712, + "loss": 0.8618, + "step": 9822 + }, + { + "epoch": 1.748931623931624, + "grad_norm": 0.6877408027648926, + "learning_rate": 0.0001197595136670764, + "loss": 0.9146, + "step": 9823 + }, + { + "epoch": 1.7491096866096867, + "grad_norm": 0.6874892115592957, + "learning_rate": 0.00011974579197952906, + "loss": 1.1628, + "step": 9824 + }, + { + "epoch": 1.7492877492877494, + "grad_norm": 0.7464384436607361, + "learning_rate": 0.00011973206990510393, + "loss": 1.007, + "step": 9825 + }, + { + "epoch": 1.749465811965812, + "grad_norm": 0.7281473278999329, + "learning_rate": 0.00011971834744406986, + "loss": 1.0776, + "step": 9826 + }, + { + "epoch": 1.7496438746438745, + "grad_norm": 0.6112284660339355, + "learning_rate": 0.00011970462459669575, + "loss": 0.7616, + "step": 9827 + }, + { + "epoch": 1.7498219373219372, + "grad_norm": 0.6498035192489624, + "learning_rate": 0.00011969090136325048, + "loss": 0.884, + "step": 9828 + }, + { + "epoch": 1.7498219373219372, + "eval_loss": 1.1018389463424683, + "eval_runtime": 24.5594, + "eval_samples_per_second": 42.387, + "eval_steps_per_second": 21.214, + "step": 9828 + }, + { + "epoch": 1.75, + "grad_norm": 0.6746426224708557, + "learning_rate": 0.00011967717774400289, + "loss": 0.9023, + "step": 9829 + }, + { + "epoch": 1.7501780626780628, + "grad_norm": 0.6513423323631287, + "learning_rate": 0.00011966345373922188, + "loss": 0.9786, + "step": 9830 + }, + { + "epoch": 1.7503561253561255, + "grad_norm": 0.7053804397583008, + "learning_rate": 0.00011964972934917632, + "loss": 1.0667, + "step": 9831 + }, + { + "epoch": 1.750534188034188, + "grad_norm": 0.6769008040428162, + "learning_rate": 0.00011963600457413513, + "loss": 0.8596, + "step": 9832 + }, + { + "epoch": 1.7507122507122506, + "grad_norm": 0.7162246108055115, + "learning_rate": 0.00011962227941436725, + "loss": 1.0746, + "step": 9833 + }, + { + "epoch": 1.7508903133903133, + "grad_norm": 0.7665811777114868, + "learning_rate": 0.00011960855387014156, + "loss": 1.0056, + "step": 9834 + }, + { + "epoch": 1.751068376068376, + "grad_norm": 0.6186950206756592, + "learning_rate": 0.00011959482794172696, + "loss": 0.9016, + "step": 9835 + }, + { + "epoch": 1.7512464387464388, + "grad_norm": 0.8018904328346252, + "learning_rate": 0.00011958110162939245, + "loss": 0.9534, + "step": 9836 + }, + { + "epoch": 1.7514245014245016, + "grad_norm": 0.8239033818244934, + "learning_rate": 0.0001195673749334069, + "loss": 1.214, + "step": 9837 + }, + { + "epoch": 1.751602564102564, + "grad_norm": 0.7886297106742859, + "learning_rate": 0.00011955364785403931, + "loss": 0.9672, + "step": 9838 + }, + { + "epoch": 1.7517806267806266, + "grad_norm": 0.6463177800178528, + "learning_rate": 0.00011953992039155862, + "loss": 0.9184, + "step": 9839 + }, + { + "epoch": 1.7519586894586894, + "grad_norm": 0.7374706864356995, + "learning_rate": 0.00011952619254623374, + "loss": 0.9988, + "step": 9840 + }, + { + "epoch": 1.7521367521367521, + "grad_norm": 0.7456657886505127, + "learning_rate": 0.00011951246431833369, + "loss": 1.2197, + "step": 9841 + }, + { + "epoch": 1.7523148148148149, + "grad_norm": 0.6644248962402344, + "learning_rate": 0.00011949873570812746, + "loss": 0.9449, + "step": 9842 + }, + { + "epoch": 1.7524928774928776, + "grad_norm": 0.707919180393219, + "learning_rate": 0.000119485006715884, + "loss": 0.774, + "step": 9843 + }, + { + "epoch": 1.7526709401709402, + "grad_norm": 0.6273906826972961, + "learning_rate": 0.00011947127734187231, + "loss": 0.8682, + "step": 9844 + }, + { + "epoch": 1.7528490028490027, + "grad_norm": 0.8335350155830383, + "learning_rate": 0.00011945754758636136, + "loss": 1.2282, + "step": 9845 + }, + { + "epoch": 1.7530270655270654, + "grad_norm": 0.6849051117897034, + "learning_rate": 0.00011944381744962022, + "loss": 1.1091, + "step": 9846 + }, + { + "epoch": 1.7532051282051282, + "grad_norm": 0.8571760058403015, + "learning_rate": 0.00011943008693191781, + "loss": 0.9806, + "step": 9847 + }, + { + "epoch": 1.753383190883191, + "grad_norm": 0.7045019268989563, + "learning_rate": 0.00011941635603352328, + "loss": 0.9217, + "step": 9848 + }, + { + "epoch": 1.7535612535612537, + "grad_norm": 0.6820187568664551, + "learning_rate": 0.00011940262475470556, + "loss": 0.9983, + "step": 9849 + }, + { + "epoch": 1.7537393162393162, + "grad_norm": 0.7400697469711304, + "learning_rate": 0.00011938889309573374, + "loss": 0.9521, + "step": 9850 + }, + { + "epoch": 1.7539173789173788, + "grad_norm": 0.7027658820152283, + "learning_rate": 0.00011937516105687678, + "loss": 1.0749, + "step": 9851 + }, + { + "epoch": 1.7540954415954415, + "grad_norm": 0.6778307557106018, + "learning_rate": 0.00011936142863840382, + "loss": 1.0249, + "step": 9852 + }, + { + "epoch": 1.7542735042735043, + "grad_norm": 0.6787961721420288, + "learning_rate": 0.00011934769584058389, + "loss": 1.0014, + "step": 9853 + }, + { + "epoch": 1.754451566951567, + "grad_norm": 0.7515636086463928, + "learning_rate": 0.00011933396266368606, + "loss": 1.0351, + "step": 9854 + }, + { + "epoch": 1.7546296296296298, + "grad_norm": 0.6620134115219116, + "learning_rate": 0.00011932022910797938, + "loss": 1.0294, + "step": 9855 + }, + { + "epoch": 1.7548076923076923, + "grad_norm": 0.8260951638221741, + "learning_rate": 0.00011930649517373294, + "loss": 0.9078, + "step": 9856 + }, + { + "epoch": 1.7549857549857548, + "grad_norm": 0.7680675983428955, + "learning_rate": 0.00011929276086121584, + "loss": 0.92, + "step": 9857 + }, + { + "epoch": 1.7551638176638176, + "grad_norm": 0.7104191184043884, + "learning_rate": 0.00011927902617069717, + "loss": 0.9937, + "step": 9858 + }, + { + "epoch": 1.7553418803418803, + "grad_norm": 0.7185840606689453, + "learning_rate": 0.00011926529110244603, + "loss": 0.9775, + "step": 9859 + }, + { + "epoch": 1.755519943019943, + "grad_norm": 0.7114652991294861, + "learning_rate": 0.00011925155565673151, + "loss": 0.883, + "step": 9860 + }, + { + "epoch": 1.7556980056980058, + "grad_norm": 0.6906639337539673, + "learning_rate": 0.00011923781983382276, + "loss": 0.9789, + "step": 9861 + }, + { + "epoch": 1.7558760683760684, + "grad_norm": 0.706908106803894, + "learning_rate": 0.00011922408363398892, + "loss": 1.1186, + "step": 9862 + }, + { + "epoch": 1.756054131054131, + "grad_norm": 0.7532939910888672, + "learning_rate": 0.00011921034705749908, + "loss": 0.977, + "step": 9863 + }, + { + "epoch": 1.7562321937321936, + "grad_norm": 0.7397763729095459, + "learning_rate": 0.0001191966101046224, + "loss": 1.1121, + "step": 9864 + }, + { + "epoch": 1.7564102564102564, + "grad_norm": 0.6955398321151733, + "learning_rate": 0.00011918287277562801, + "loss": 1.0439, + "step": 9865 + }, + { + "epoch": 1.7565883190883191, + "grad_norm": 0.7485929727554321, + "learning_rate": 0.00011916913507078507, + "loss": 1.1644, + "step": 9866 + }, + { + "epoch": 1.756766381766382, + "grad_norm": 0.6337487101554871, + "learning_rate": 0.00011915539699036274, + "loss": 0.8216, + "step": 9867 + }, + { + "epoch": 1.7569444444444444, + "grad_norm": 0.6628872752189636, + "learning_rate": 0.00011914165853463022, + "loss": 0.9584, + "step": 9868 + }, + { + "epoch": 1.7571225071225072, + "grad_norm": 0.6577547788619995, + "learning_rate": 0.00011912791970385666, + "loss": 0.9484, + "step": 9869 + }, + { + "epoch": 1.7573005698005697, + "grad_norm": 0.6409304738044739, + "learning_rate": 0.00011911418049831127, + "loss": 1.1256, + "step": 9870 + }, + { + "epoch": 1.7574786324786325, + "grad_norm": 0.7499844431877136, + "learning_rate": 0.00011910044091826319, + "loss": 0.7991, + "step": 9871 + }, + { + "epoch": 1.7576566951566952, + "grad_norm": 0.6786715388298035, + "learning_rate": 0.00011908670096398165, + "loss": 1.0368, + "step": 9872 + }, + { + "epoch": 1.757834757834758, + "grad_norm": 0.6432101130485535, + "learning_rate": 0.00011907296063573585, + "loss": 0.9059, + "step": 9873 + }, + { + "epoch": 1.7580128205128205, + "grad_norm": 0.6542613506317139, + "learning_rate": 0.00011905921993379503, + "loss": 0.9866, + "step": 9874 + }, + { + "epoch": 1.7581908831908832, + "grad_norm": 0.6048218011856079, + "learning_rate": 0.00011904547885842838, + "loss": 0.9488, + "step": 9875 + }, + { + "epoch": 1.7583689458689458, + "grad_norm": 0.7694938778877258, + "learning_rate": 0.00011903173740990512, + "loss": 1.1026, + "step": 9876 + }, + { + "epoch": 1.7585470085470085, + "grad_norm": 0.6621627807617188, + "learning_rate": 0.00011901799558849451, + "loss": 1.135, + "step": 9877 + }, + { + "epoch": 1.7587250712250713, + "grad_norm": 0.6561587452888489, + "learning_rate": 0.0001190042533944658, + "loss": 0.9322, + "step": 9878 + }, + { + "epoch": 1.758903133903134, + "grad_norm": 0.7846759557723999, + "learning_rate": 0.00011899051082808821, + "loss": 0.9324, + "step": 9879 + }, + { + "epoch": 1.7590811965811965, + "grad_norm": 0.6004071831703186, + "learning_rate": 0.00011897676788963101, + "loss": 0.9641, + "step": 9880 + }, + { + "epoch": 1.7592592592592593, + "grad_norm": 0.6731070280075073, + "learning_rate": 0.00011896302457936344, + "loss": 1.1437, + "step": 9881 + }, + { + "epoch": 1.7594373219373218, + "grad_norm": 0.6768675446510315, + "learning_rate": 0.00011894928089755481, + "loss": 1.0707, + "step": 9882 + }, + { + "epoch": 1.7596153846153846, + "grad_norm": 0.8368878960609436, + "learning_rate": 0.0001189355368444744, + "loss": 1.0435, + "step": 9883 + }, + { + "epoch": 1.7597934472934473, + "grad_norm": 0.6132324934005737, + "learning_rate": 0.00011892179242039149, + "loss": 0.8889, + "step": 9884 + }, + { + "epoch": 1.75997150997151, + "grad_norm": 0.7598093152046204, + "learning_rate": 0.00011890804762557535, + "loss": 1.151, + "step": 9885 + }, + { + "epoch": 1.7601495726495726, + "grad_norm": 0.7317715883255005, + "learning_rate": 0.00011889430246029527, + "loss": 0.9992, + "step": 9886 + }, + { + "epoch": 1.7603276353276354, + "grad_norm": 0.7664858102798462, + "learning_rate": 0.00011888055692482059, + "loss": 0.8398, + "step": 9887 + }, + { + "epoch": 1.760505698005698, + "grad_norm": 0.6916853189468384, + "learning_rate": 0.00011886681101942063, + "loss": 0.9507, + "step": 9888 + }, + { + "epoch": 1.7606837606837606, + "grad_norm": 0.7103399634361267, + "learning_rate": 0.0001188530647443647, + "loss": 0.915, + "step": 9889 + }, + { + "epoch": 1.7608618233618234, + "grad_norm": 0.6177804470062256, + "learning_rate": 0.00011883931809992215, + "loss": 0.721, + "step": 9890 + }, + { + "epoch": 1.7610398860398861, + "grad_norm": 0.7523959279060364, + "learning_rate": 0.00011882557108636227, + "loss": 0.99, + "step": 9891 + }, + { + "epoch": 1.7612179487179487, + "grad_norm": 0.6211134791374207, + "learning_rate": 0.00011881182370395442, + "loss": 0.8089, + "step": 9892 + }, + { + "epoch": 1.7613960113960114, + "grad_norm": 0.6660307049751282, + "learning_rate": 0.00011879807595296802, + "loss": 1.1062, + "step": 9893 + }, + { + "epoch": 1.761574074074074, + "grad_norm": 0.7039240598678589, + "learning_rate": 0.00011878432783367232, + "loss": 0.9739, + "step": 9894 + }, + { + "epoch": 1.7617521367521367, + "grad_norm": 0.658064603805542, + "learning_rate": 0.00011877057934633675, + "loss": 0.9438, + "step": 9895 + }, + { + "epoch": 1.7619301994301995, + "grad_norm": 0.8227152228355408, + "learning_rate": 0.00011875683049123068, + "loss": 0.8385, + "step": 9896 + }, + { + "epoch": 1.7621082621082622, + "grad_norm": 0.6622483730316162, + "learning_rate": 0.00011874308126862346, + "loss": 0.9432, + "step": 9897 + }, + { + "epoch": 1.7622863247863247, + "grad_norm": 0.7211357951164246, + "learning_rate": 0.00011872933167878453, + "loss": 1.2471, + "step": 9898 + }, + { + "epoch": 1.7624643874643875, + "grad_norm": 0.6177424192428589, + "learning_rate": 0.00011871558172198322, + "loss": 0.8892, + "step": 9899 + }, + { + "epoch": 1.76264245014245, + "grad_norm": 0.6924285888671875, + "learning_rate": 0.00011870183139848898, + "loss": 1.021, + "step": 9900 + }, + { + "epoch": 1.7628205128205128, + "grad_norm": 0.6168648600578308, + "learning_rate": 0.0001186880807085712, + "loss": 0.9013, + "step": 9901 + }, + { + "epoch": 1.7629985754985755, + "grad_norm": 0.6410452723503113, + "learning_rate": 0.00011867432965249929, + "loss": 0.6686, + "step": 9902 + }, + { + "epoch": 1.7631766381766383, + "grad_norm": 0.6959559917449951, + "learning_rate": 0.0001186605782305427, + "loss": 0.9814, + "step": 9903 + }, + { + "epoch": 1.7633547008547008, + "grad_norm": 0.7456178069114685, + "learning_rate": 0.00011864682644297085, + "loss": 1.0151, + "step": 9904 + }, + { + "epoch": 1.7635327635327636, + "grad_norm": 0.6499991416931152, + "learning_rate": 0.00011863307429005317, + "loss": 0.83, + "step": 9905 + }, + { + "epoch": 1.763710826210826, + "grad_norm": 0.643344521522522, + "learning_rate": 0.00011861932177205908, + "loss": 0.8853, + "step": 9906 + }, + { + "epoch": 1.7638888888888888, + "grad_norm": 0.6570441722869873, + "learning_rate": 0.00011860556888925804, + "loss": 0.9179, + "step": 9907 + }, + { + "epoch": 1.7640669515669516, + "grad_norm": 0.6892307996749878, + "learning_rate": 0.00011859181564191957, + "loss": 0.9657, + "step": 9908 + }, + { + "epoch": 1.7642450142450143, + "grad_norm": 0.648158073425293, + "learning_rate": 0.0001185780620303131, + "loss": 0.9179, + "step": 9909 + }, + { + "epoch": 1.7644230769230769, + "grad_norm": 0.5833603143692017, + "learning_rate": 0.00011856430805470808, + "loss": 0.8505, + "step": 9910 + }, + { + "epoch": 1.7646011396011396, + "grad_norm": 0.8302416205406189, + "learning_rate": 0.000118550553715374, + "loss": 0.8948, + "step": 9911 + }, + { + "epoch": 1.7647792022792022, + "grad_norm": 0.7075300216674805, + "learning_rate": 0.00011853679901258035, + "loss": 1.2467, + "step": 9912 + }, + { + "epoch": 1.764957264957265, + "grad_norm": 0.81916344165802, + "learning_rate": 0.00011852304394659666, + "loss": 0.9963, + "step": 9913 + }, + { + "epoch": 1.7651353276353277, + "grad_norm": 0.6492435932159424, + "learning_rate": 0.00011850928851769239, + "loss": 1.0704, + "step": 9914 + }, + { + "epoch": 1.7653133903133904, + "grad_norm": 0.7301090359687805, + "learning_rate": 0.00011849553272613704, + "loss": 1.0477, + "step": 9915 + }, + { + "epoch": 1.765491452991453, + "grad_norm": 0.7280275821685791, + "learning_rate": 0.00011848177657220019, + "loss": 0.9124, + "step": 9916 + }, + { + "epoch": 1.7656695156695157, + "grad_norm": 0.6948845386505127, + "learning_rate": 0.00011846802005615127, + "loss": 1.2275, + "step": 9917 + }, + { + "epoch": 1.7658475783475782, + "grad_norm": 0.6553834676742554, + "learning_rate": 0.0001184542631782599, + "loss": 1.2311, + "step": 9918 + }, + { + "epoch": 1.766025641025641, + "grad_norm": 0.6899739503860474, + "learning_rate": 0.00011844050593879556, + "loss": 0.8936, + "step": 9919 + }, + { + "epoch": 1.7662037037037037, + "grad_norm": 0.6076815128326416, + "learning_rate": 0.00011842674833802782, + "loss": 0.8432, + "step": 9920 + }, + { + "epoch": 1.7663817663817665, + "grad_norm": 0.7650902271270752, + "learning_rate": 0.00011841299037622624, + "loss": 1.0447, + "step": 9921 + }, + { + "epoch": 1.7665598290598292, + "grad_norm": 0.6864938735961914, + "learning_rate": 0.00011839923205366032, + "loss": 0.936, + "step": 9922 + }, + { + "epoch": 1.7667378917378918, + "grad_norm": 0.7176852226257324, + "learning_rate": 0.0001183854733705997, + "loss": 0.9764, + "step": 9923 + }, + { + "epoch": 1.7669159544159543, + "grad_norm": 0.6513439416885376, + "learning_rate": 0.00011837171432731393, + "loss": 1.0095, + "step": 9924 + }, + { + "epoch": 1.767094017094017, + "grad_norm": 0.8031024932861328, + "learning_rate": 0.00011835795492407256, + "loss": 1.1348, + "step": 9925 + }, + { + "epoch": 1.7672720797720798, + "grad_norm": 0.7659830451011658, + "learning_rate": 0.00011834419516114518, + "loss": 0.9058, + "step": 9926 + }, + { + "epoch": 1.7674501424501425, + "grad_norm": 0.8864039778709412, + "learning_rate": 0.00011833043503880145, + "loss": 1.0342, + "step": 9927 + }, + { + "epoch": 1.7676282051282053, + "grad_norm": 0.6870512962341309, + "learning_rate": 0.00011831667455731088, + "loss": 0.9361, + "step": 9928 + }, + { + "epoch": 1.7678062678062678, + "grad_norm": 0.6458830833435059, + "learning_rate": 0.00011830291371694315, + "loss": 0.8215, + "step": 9929 + }, + { + "epoch": 1.7679843304843303, + "grad_norm": 0.7456086874008179, + "learning_rate": 0.00011828915251796787, + "loss": 1.1243, + "step": 9930 + }, + { + "epoch": 1.768162393162393, + "grad_norm": 0.6834850311279297, + "learning_rate": 0.00011827539096065459, + "loss": 0.9536, + "step": 9931 + }, + { + "epoch": 1.7683404558404558, + "grad_norm": 0.643864631652832, + "learning_rate": 0.00011826162904527302, + "loss": 1.1707, + "step": 9932 + }, + { + "epoch": 1.7685185185185186, + "grad_norm": 0.6312864422798157, + "learning_rate": 0.00011824786677209275, + "loss": 0.7937, + "step": 9933 + }, + { + "epoch": 1.7686965811965814, + "grad_norm": 0.6092729568481445, + "learning_rate": 0.00011823410414138343, + "loss": 0.8787, + "step": 9934 + }, + { + "epoch": 1.7688746438746439, + "grad_norm": 0.6859988570213318, + "learning_rate": 0.00011822034115341474, + "loss": 0.9691, + "step": 9935 + }, + { + "epoch": 1.7690527065527064, + "grad_norm": 0.7219935059547424, + "learning_rate": 0.0001182065778084563, + "loss": 1.0606, + "step": 9936 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.6596202850341797, + "learning_rate": 0.00011819281410677778, + "loss": 1.0543, + "step": 9937 + }, + { + "epoch": 1.769408831908832, + "grad_norm": 0.6616338491439819, + "learning_rate": 0.00011817905004864887, + "loss": 0.9757, + "step": 9938 + }, + { + "epoch": 1.7695868945868947, + "grad_norm": 0.6637360453605652, + "learning_rate": 0.00011816528563433924, + "loss": 0.925, + "step": 9939 + }, + { + "epoch": 1.7697649572649574, + "grad_norm": 0.8422333002090454, + "learning_rate": 0.00011815152086411859, + "loss": 1.1343, + "step": 9940 + }, + { + "epoch": 1.76994301994302, + "grad_norm": 0.6638204455375671, + "learning_rate": 0.00011813775573825656, + "loss": 1.2136, + "step": 9941 + }, + { + "epoch": 1.7701210826210825, + "grad_norm": 0.7258831858634949, + "learning_rate": 0.0001181239902570229, + "loss": 0.7308, + "step": 9942 + }, + { + "epoch": 1.7702991452991452, + "grad_norm": 0.730582594871521, + "learning_rate": 0.0001181102244206873, + "loss": 1.1097, + "step": 9943 + }, + { + "epoch": 1.770477207977208, + "grad_norm": 0.7324019074440002, + "learning_rate": 0.00011809645822951946, + "loss": 0.9802, + "step": 9944 + }, + { + "epoch": 1.7706552706552707, + "grad_norm": 0.5565997958183289, + "learning_rate": 0.00011808269168378914, + "loss": 0.7079, + "step": 9945 + }, + { + "epoch": 1.7708333333333335, + "grad_norm": 0.6395503282546997, + "learning_rate": 0.00011806892478376601, + "loss": 1.0048, + "step": 9946 + }, + { + "epoch": 1.771011396011396, + "grad_norm": 0.7670905590057373, + "learning_rate": 0.00011805515752971985, + "loss": 1.2509, + "step": 9947 + }, + { + "epoch": 1.7711894586894585, + "grad_norm": 0.5945813655853271, + "learning_rate": 0.00011804138992192037, + "loss": 0.8856, + "step": 9948 + }, + { + "epoch": 1.7713675213675213, + "grad_norm": 0.7355493307113647, + "learning_rate": 0.00011802762196063737, + "loss": 0.9629, + "step": 9949 + }, + { + "epoch": 1.771545584045584, + "grad_norm": 0.7024806141853333, + "learning_rate": 0.00011801385364614055, + "loss": 1.1351, + "step": 9950 + }, + { + "epoch": 1.7717236467236468, + "grad_norm": 0.6553003191947937, + "learning_rate": 0.00011800008497869968, + "loss": 0.911, + "step": 9951 + }, + { + "epoch": 1.7719017094017095, + "grad_norm": 0.6883971691131592, + "learning_rate": 0.00011798631595858454, + "loss": 1.0099, + "step": 9952 + }, + { + "epoch": 1.772079772079772, + "grad_norm": 0.7106832265853882, + "learning_rate": 0.00011797254658606489, + "loss": 1.0298, + "step": 9953 + }, + { + "epoch": 1.7722578347578346, + "grad_norm": 0.7902877926826477, + "learning_rate": 0.00011795877686141055, + "loss": 1.0572, + "step": 9954 + }, + { + "epoch": 1.7724358974358974, + "grad_norm": 0.7105007171630859, + "learning_rate": 0.00011794500678489126, + "loss": 1.1725, + "step": 9955 + }, + { + "epoch": 1.77261396011396, + "grad_norm": 0.7314959764480591, + "learning_rate": 0.00011793123635677685, + "loss": 1.1074, + "step": 9956 + }, + { + "epoch": 1.7727920227920229, + "grad_norm": 0.6358618140220642, + "learning_rate": 0.00011791746557733712, + "loss": 0.8786, + "step": 9957 + }, + { + "epoch": 1.7729700854700856, + "grad_norm": 0.6441367864608765, + "learning_rate": 0.00011790369444684187, + "loss": 1.1332, + "step": 9958 + }, + { + "epoch": 1.7731481481481481, + "grad_norm": 0.686787486076355, + "learning_rate": 0.0001178899229655609, + "loss": 0.9566, + "step": 9959 + }, + { + "epoch": 1.7733262108262107, + "grad_norm": 0.653840184211731, + "learning_rate": 0.00011787615113376407, + "loss": 0.8763, + "step": 9960 + }, + { + "epoch": 1.7735042735042734, + "grad_norm": 0.7106643915176392, + "learning_rate": 0.00011786237895172119, + "loss": 0.9929, + "step": 9961 + }, + { + "epoch": 1.7736823361823362, + "grad_norm": 0.6634044051170349, + "learning_rate": 0.0001178486064197021, + "loss": 0.7467, + "step": 9962 + }, + { + "epoch": 1.773860398860399, + "grad_norm": 0.7087352871894836, + "learning_rate": 0.00011783483353797663, + "loss": 1.0104, + "step": 9963 + }, + { + "epoch": 1.7740384615384617, + "grad_norm": 0.8088061213493347, + "learning_rate": 0.00011782106030681466, + "loss": 1.0376, + "step": 9964 + }, + { + "epoch": 1.7742165242165242, + "grad_norm": 0.7204688787460327, + "learning_rate": 0.00011780728672648604, + "loss": 0.8556, + "step": 9965 + }, + { + "epoch": 1.7743945868945867, + "grad_norm": 0.7893314957618713, + "learning_rate": 0.0001177935127972606, + "loss": 0.9764, + "step": 9966 + }, + { + "epoch": 1.7745726495726495, + "grad_norm": 0.6098896265029907, + "learning_rate": 0.00011777973851940826, + "loss": 0.9407, + "step": 9967 + }, + { + "epoch": 1.7747507122507122, + "grad_norm": 0.6420868039131165, + "learning_rate": 0.0001177659638931989, + "loss": 1.1328, + "step": 9968 + }, + { + "epoch": 1.774928774928775, + "grad_norm": 0.7732378244400024, + "learning_rate": 0.00011775218891890234, + "loss": 1.1236, + "step": 9969 + }, + { + "epoch": 1.7751068376068377, + "grad_norm": 0.6591582894325256, + "learning_rate": 0.00011773841359678855, + "loss": 1.1523, + "step": 9970 + }, + { + "epoch": 1.7752849002849003, + "grad_norm": 0.6337170004844666, + "learning_rate": 0.00011772463792712738, + "loss": 1.1998, + "step": 9971 + }, + { + "epoch": 1.7754629629629628, + "grad_norm": 0.6400532126426697, + "learning_rate": 0.00011771086191018874, + "loss": 0.9543, + "step": 9972 + }, + { + "epoch": 1.7756410256410255, + "grad_norm": 0.6431527733802795, + "learning_rate": 0.00011769708554624257, + "loss": 0.8164, + "step": 9973 + }, + { + "epoch": 1.7758190883190883, + "grad_norm": 0.7303599119186401, + "learning_rate": 0.00011768330883555876, + "loss": 0.9553, + "step": 9974 + }, + { + "epoch": 1.775997150997151, + "grad_norm": 0.7838605642318726, + "learning_rate": 0.00011766953177840725, + "loss": 0.9759, + "step": 9975 + }, + { + "epoch": 1.7761752136752138, + "grad_norm": 0.6505265831947327, + "learning_rate": 0.00011765575437505796, + "loss": 0.8527, + "step": 9976 + }, + { + "epoch": 1.7763532763532763, + "grad_norm": 0.7336180806159973, + "learning_rate": 0.00011764197662578086, + "loss": 1.1098, + "step": 9977 + }, + { + "epoch": 1.776531339031339, + "grad_norm": 0.7040138244628906, + "learning_rate": 0.00011762819853084586, + "loss": 1.1289, + "step": 9978 + }, + { + "epoch": 1.7767094017094016, + "grad_norm": 0.6414867043495178, + "learning_rate": 0.00011761442009052293, + "loss": 1.0826, + "step": 9979 + }, + { + "epoch": 1.7768874643874644, + "grad_norm": 0.6760666370391846, + "learning_rate": 0.00011760064130508204, + "loss": 1.0188, + "step": 9980 + }, + { + "epoch": 1.7770655270655271, + "grad_norm": 0.7864978909492493, + "learning_rate": 0.00011758686217479316, + "loss": 1.1938, + "step": 9981 + }, + { + "epoch": 1.7772435897435899, + "grad_norm": 0.7964870929718018, + "learning_rate": 0.00011757308269992622, + "loss": 0.9876, + "step": 9982 + }, + { + "epoch": 1.7774216524216524, + "grad_norm": 0.5158692002296448, + "learning_rate": 0.00011755930288075123, + "loss": 0.6508, + "step": 9983 + }, + { + "epoch": 1.7775997150997151, + "grad_norm": 0.7208606600761414, + "learning_rate": 0.00011754552271753819, + "loss": 1.0738, + "step": 9984 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 0.6811334490776062, + "learning_rate": 0.00011753174221055705, + "loss": 1.1216, + "step": 9985 + }, + { + "epoch": 1.7779558404558404, + "grad_norm": 0.6389986276626587, + "learning_rate": 0.00011751796136007787, + "loss": 0.9664, + "step": 9986 + }, + { + "epoch": 1.7781339031339032, + "grad_norm": 0.7081875205039978, + "learning_rate": 0.00011750418016637064, + "loss": 0.9365, + "step": 9987 + }, + { + "epoch": 1.778311965811966, + "grad_norm": 0.7291778326034546, + "learning_rate": 0.00011749039862970535, + "loss": 1.3222, + "step": 9988 + }, + { + "epoch": 1.7784900284900285, + "grad_norm": 0.6790453791618347, + "learning_rate": 0.000117476616750352, + "loss": 0.9537, + "step": 9989 + }, + { + "epoch": 1.7786680911680912, + "grad_norm": 0.6271076202392578, + "learning_rate": 0.00011746283452858069, + "loss": 0.9842, + "step": 9990 + }, + { + "epoch": 1.7788461538461537, + "grad_norm": 0.675628662109375, + "learning_rate": 0.00011744905196466138, + "loss": 0.8675, + "step": 9991 + }, + { + "epoch": 1.7790242165242165, + "grad_norm": 0.7328314185142517, + "learning_rate": 0.00011743526905886417, + "loss": 0.9793, + "step": 9992 + }, + { + "epoch": 1.7792022792022792, + "grad_norm": 0.698764979839325, + "learning_rate": 0.00011742148581145908, + "loss": 0.9527, + "step": 9993 + }, + { + "epoch": 1.779380341880342, + "grad_norm": 0.6911364793777466, + "learning_rate": 0.00011740770222271616, + "loss": 1.1069, + "step": 9994 + }, + { + "epoch": 1.7795584045584045, + "grad_norm": 0.6990836262702942, + "learning_rate": 0.00011739391829290547, + "loss": 0.9132, + "step": 9995 + }, + { + "epoch": 1.7797364672364673, + "grad_norm": 0.7056801319122314, + "learning_rate": 0.0001173801340222971, + "loss": 1.053, + "step": 9996 + }, + { + "epoch": 1.7799145299145298, + "grad_norm": 0.7453791499137878, + "learning_rate": 0.0001173663494111611, + "loss": 0.8806, + "step": 9997 + }, + { + "epoch": 1.7800925925925926, + "grad_norm": 0.7211771011352539, + "learning_rate": 0.00011735256445976757, + "loss": 0.9968, + "step": 9998 + }, + { + "epoch": 1.7802706552706553, + "grad_norm": 0.7259734272956848, + "learning_rate": 0.00011733877916838656, + "loss": 1.167, + "step": 9999 + }, + { + "epoch": 1.780448717948718, + "grad_norm": 0.6931926012039185, + "learning_rate": 0.00011732499353728821, + "loss": 1.0634, + "step": 10000 + }, + { + "epoch": 1.7806267806267806, + "grad_norm": 0.6900074481964111, + "learning_rate": 0.00011731120756674259, + "loss": 0.9718, + "step": 10001 + }, + { + "epoch": 1.7808048433048433, + "grad_norm": 0.6817582845687866, + "learning_rate": 0.00011729742125701984, + "loss": 1.0896, + "step": 10002 + }, + { + "epoch": 1.7809829059829059, + "grad_norm": 0.6901891231536865, + "learning_rate": 0.00011728363460839003, + "loss": 1.0163, + "step": 10003 + }, + { + "epoch": 1.7811609686609686, + "grad_norm": 0.9138323664665222, + "learning_rate": 0.00011726984762112328, + "loss": 1.1713, + "step": 10004 + }, + { + "epoch": 1.7813390313390314, + "grad_norm": 0.6105810403823853, + "learning_rate": 0.00011725606029548977, + "loss": 0.9331, + "step": 10005 + }, + { + "epoch": 1.7815170940170941, + "grad_norm": 0.5605259537696838, + "learning_rate": 0.0001172422726317596, + "loss": 0.7154, + "step": 10006 + }, + { + "epoch": 1.7816951566951567, + "grad_norm": 0.6950963735580444, + "learning_rate": 0.00011722848463020292, + "loss": 1.0093, + "step": 10007 + }, + { + "epoch": 1.7818732193732194, + "grad_norm": 0.6806309819221497, + "learning_rate": 0.00011721469629108988, + "loss": 0.8662, + "step": 10008 + }, + { + "epoch": 1.782051282051282, + "grad_norm": 0.7528520226478577, + "learning_rate": 0.00011720090761469063, + "loss": 0.8567, + "step": 10009 + }, + { + "epoch": 1.7822293447293447, + "grad_norm": 0.6617229580879211, + "learning_rate": 0.00011718711860127529, + "loss": 1.0378, + "step": 10010 + }, + { + "epoch": 1.7824074074074074, + "grad_norm": 0.6468376517295837, + "learning_rate": 0.00011717332925111411, + "loss": 1.0658, + "step": 10011 + }, + { + "epoch": 1.7825854700854702, + "grad_norm": 0.7141897082328796, + "learning_rate": 0.00011715953956447721, + "loss": 1.023, + "step": 10012 + }, + { + "epoch": 1.7827635327635327, + "grad_norm": 0.5777570605278015, + "learning_rate": 0.00011714574954163475, + "loss": 0.9154, + "step": 10013 + }, + { + "epoch": 1.7829415954415955, + "grad_norm": 0.7536137700080872, + "learning_rate": 0.00011713195918285695, + "loss": 0.9651, + "step": 10014 + }, + { + "epoch": 1.783119658119658, + "grad_norm": 0.6977683305740356, + "learning_rate": 0.00011711816848841402, + "loss": 0.7977, + "step": 10015 + }, + { + "epoch": 1.7832977207977208, + "grad_norm": 0.6522472500801086, + "learning_rate": 0.00011710437745857614, + "loss": 0.8834, + "step": 10016 + }, + { + "epoch": 1.7834757834757835, + "grad_norm": 0.6263057589530945, + "learning_rate": 0.0001170905860936135, + "loss": 1.0576, + "step": 10017 + }, + { + "epoch": 1.7836538461538463, + "grad_norm": 0.6470699310302734, + "learning_rate": 0.00011707679439379635, + "loss": 0.9412, + "step": 10018 + }, + { + "epoch": 1.7838319088319088, + "grad_norm": Infinity, + "learning_rate": 0.00011707679439379635, + "loss": 1.1746, + "step": 10019 + }, + { + "epoch": 1.7840099715099715, + "grad_norm": 0.6022017002105713, + "learning_rate": 0.00011706300235939485, + "loss": 0.8945, + "step": 10020 + }, + { + "epoch": 1.784188034188034, + "grad_norm": 0.637208104133606, + "learning_rate": 0.00011704920999067927, + "loss": 1.0215, + "step": 10021 + }, + { + "epoch": 1.7843660968660968, + "grad_norm": 0.7467851042747498, + "learning_rate": 0.00011703541728791987, + "loss": 1.0341, + "step": 10022 + }, + { + "epoch": 1.7845441595441596, + "grad_norm": 0.7562711238861084, + "learning_rate": 0.00011702162425138683, + "loss": 0.9748, + "step": 10023 + }, + { + "epoch": 1.7847222222222223, + "grad_norm": 0.6480089426040649, + "learning_rate": 0.00011700783088135043, + "loss": 1.05, + "step": 10024 + }, + { + "epoch": 1.7849002849002849, + "grad_norm": 0.6293981671333313, + "learning_rate": 0.00011699403717808091, + "loss": 1.0376, + "step": 10025 + }, + { + "epoch": 1.7850783475783476, + "grad_norm": 0.6821253895759583, + "learning_rate": 0.00011698024314184853, + "loss": 1.0542, + "step": 10026 + }, + { + "epoch": 1.7852564102564101, + "grad_norm": 0.6681216359138489, + "learning_rate": 0.00011696644877292356, + "loss": 1.0018, + "step": 10027 + }, + { + "epoch": 1.7854344729344729, + "grad_norm": 0.6788804531097412, + "learning_rate": 0.00011695265407157628, + "loss": 1.1823, + "step": 10028 + }, + { + "epoch": 1.7856125356125356, + "grad_norm": 0.6147881150245667, + "learning_rate": 0.00011693885903807697, + "loss": 0.9246, + "step": 10029 + }, + { + "epoch": 1.7857905982905984, + "grad_norm": 0.7952296137809753, + "learning_rate": 0.00011692506367269588, + "loss": 1.0528, + "step": 10030 + }, + { + "epoch": 1.785968660968661, + "grad_norm": 0.6985954642295837, + "learning_rate": 0.00011691126797570333, + "loss": 0.9173, + "step": 10031 + }, + { + "epoch": 1.7861467236467237, + "grad_norm": 0.6211223602294922, + "learning_rate": 0.00011689747194736961, + "loss": 0.7527, + "step": 10032 + }, + { + "epoch": 1.7863247863247862, + "grad_norm": 0.7531208992004395, + "learning_rate": 0.00011688367558796507, + "loss": 1.1087, + "step": 10033 + }, + { + "epoch": 1.786502849002849, + "grad_norm": 0.7742924690246582, + "learning_rate": 0.00011686987889775996, + "loss": 1.1512, + "step": 10034 + }, + { + "epoch": 1.7866809116809117, + "grad_norm": 0.7046231627464294, + "learning_rate": 0.00011685608187702459, + "loss": 1.0516, + "step": 10035 + }, + { + "epoch": 1.7868589743589745, + "grad_norm": 0.6264076232910156, + "learning_rate": 0.00011684228452602933, + "loss": 0.8938, + "step": 10036 + }, + { + "epoch": 1.7870370370370372, + "grad_norm": 0.6342145800590515, + "learning_rate": 0.00011682848684504448, + "loss": 0.8177, + "step": 10037 + }, + { + "epoch": 1.7872150997150997, + "grad_norm": 0.6609861254692078, + "learning_rate": 0.00011681468883434041, + "loss": 0.9692, + "step": 10038 + }, + { + "epoch": 1.7873931623931623, + "grad_norm": 0.7918622493743896, + "learning_rate": 0.00011680089049418743, + "loss": 0.8246, + "step": 10039 + }, + { + "epoch": 1.787571225071225, + "grad_norm": 0.697712779045105, + "learning_rate": 0.00011678709182485592, + "loss": 0.8981, + "step": 10040 + }, + { + "epoch": 1.7877492877492878, + "grad_norm": 0.6747658252716064, + "learning_rate": 0.00011677329282661617, + "loss": 1.1243, + "step": 10041 + }, + { + "epoch": 1.7879273504273505, + "grad_norm": 0.6525771617889404, + "learning_rate": 0.00011675949349973863, + "loss": 0.852, + "step": 10042 + }, + { + "epoch": 1.7881054131054133, + "grad_norm": 0.7062464952468872, + "learning_rate": 0.00011674569384449363, + "loss": 1.2582, + "step": 10043 + }, + { + "epoch": 1.7882834757834758, + "grad_norm": 0.6453786492347717, + "learning_rate": 0.00011673189386115154, + "loss": 0.868, + "step": 10044 + }, + { + "epoch": 1.7884615384615383, + "grad_norm": 0.7939708232879639, + "learning_rate": 0.00011671809354998273, + "loss": 0.7553, + "step": 10045 + }, + { + "epoch": 1.788639601139601, + "grad_norm": 0.6466066837310791, + "learning_rate": 0.00011670429291125761, + "loss": 0.942, + "step": 10046 + }, + { + "epoch": 1.7888176638176638, + "grad_norm": 0.7380510568618774, + "learning_rate": 0.00011669049194524657, + "loss": 1.044, + "step": 10047 + }, + { + "epoch": 1.7889957264957266, + "grad_norm": 0.6719707250595093, + "learning_rate": 0.00011667669065222002, + "loss": 1.1624, + "step": 10048 + }, + { + "epoch": 1.7891737891737893, + "grad_norm": 0.6996603012084961, + "learning_rate": 0.00011666288903244837, + "loss": 1.001, + "step": 10049 + }, + { + "epoch": 1.7893518518518519, + "grad_norm": 0.696590006351471, + "learning_rate": 0.00011664908708620202, + "loss": 1.17, + "step": 10050 + }, + { + "epoch": 1.7895299145299144, + "grad_norm": 0.7226764559745789, + "learning_rate": 0.00011663528481375137, + "loss": 1.0762, + "step": 10051 + }, + { + "epoch": 1.7897079772079771, + "grad_norm": 0.6117866635322571, + "learning_rate": 0.00011662148221536689, + "loss": 0.9199, + "step": 10052 + }, + { + "epoch": 1.78988603988604, + "grad_norm": 0.6424985527992249, + "learning_rate": 0.000116607679291319, + "loss": 1.1672, + "step": 10053 + }, + { + "epoch": 1.7900641025641026, + "grad_norm": 0.6390290856361389, + "learning_rate": 0.00011659387604187813, + "loss": 1.1895, + "step": 10054 + }, + { + "epoch": 1.7902421652421654, + "grad_norm": 0.6553205251693726, + "learning_rate": 0.00011658007246731473, + "loss": 1.0967, + "step": 10055 + }, + { + "epoch": 1.790420227920228, + "grad_norm": 0.7737570405006409, + "learning_rate": 0.00011656626856789922, + "loss": 0.9637, + "step": 10056 + }, + { + "epoch": 1.7905982905982905, + "grad_norm": 0.644296407699585, + "learning_rate": 0.00011655246434390212, + "loss": 0.9933, + "step": 10057 + }, + { + "epoch": 1.7907763532763532, + "grad_norm": 0.8154410123825073, + "learning_rate": 0.00011653865979559388, + "loss": 0.9623, + "step": 10058 + }, + { + "epoch": 1.790954415954416, + "grad_norm": 0.7181384563446045, + "learning_rate": 0.00011652485492324495, + "loss": 0.9113, + "step": 10059 + }, + { + "epoch": 1.7911324786324787, + "grad_norm": 0.7835097908973694, + "learning_rate": 0.00011651104972712582, + "loss": 1.0804, + "step": 10060 + }, + { + "epoch": 1.7913105413105415, + "grad_norm": 0.6843693852424622, + "learning_rate": 0.00011649724420750691, + "loss": 1.0242, + "step": 10061 + }, + { + "epoch": 1.791488603988604, + "grad_norm": 0.8364703059196472, + "learning_rate": 0.00011648343836465885, + "loss": 0.8445, + "step": 10062 + }, + { + "epoch": 1.7916666666666665, + "grad_norm": 0.7122092843055725, + "learning_rate": 0.00011646963219885201, + "loss": 1.0453, + "step": 10063 + }, + { + "epoch": 1.7918447293447293, + "grad_norm": 0.7018755078315735, + "learning_rate": 0.00011645582571035696, + "loss": 0.9753, + "step": 10064 + }, + { + "epoch": 1.792022792022792, + "grad_norm": 0.6522594094276428, + "learning_rate": 0.00011644201889944419, + "loss": 1.0328, + "step": 10065 + }, + { + "epoch": 1.7922008547008548, + "grad_norm": 0.70301353931427, + "learning_rate": 0.00011642821176638419, + "loss": 0.9143, + "step": 10066 + }, + { + "epoch": 1.7923789173789175, + "grad_norm": 0.6255469918251038, + "learning_rate": 0.0001164144043114475, + "loss": 0.9527, + "step": 10067 + }, + { + "epoch": 1.79255698005698, + "grad_norm": 0.6780602931976318, + "learning_rate": 0.0001164005965349047, + "loss": 0.9192, + "step": 10068 + }, + { + "epoch": 1.7927350427350426, + "grad_norm": 0.6025984287261963, + "learning_rate": 0.00011638678843702626, + "loss": 0.9055, + "step": 10069 + }, + { + "epoch": 1.7929131054131053, + "grad_norm": 0.6430829763412476, + "learning_rate": 0.00011637298001808275, + "loss": 0.9359, + "step": 10070 + }, + { + "epoch": 1.793091168091168, + "grad_norm": 0.6388106942176819, + "learning_rate": 0.0001163591712783447, + "loss": 0.8847, + "step": 10071 + }, + { + "epoch": 1.7932692307692308, + "grad_norm": 0.706347644329071, + "learning_rate": 0.00011634536221808265, + "loss": 0.9055, + "step": 10072 + }, + { + "epoch": 1.7934472934472936, + "grad_norm": 0.661226749420166, + "learning_rate": 0.00011633155283756721, + "loss": 1.118, + "step": 10073 + }, + { + "epoch": 1.7936253561253561, + "grad_norm": 0.543207049369812, + "learning_rate": 0.00011631774313706891, + "loss": 0.8856, + "step": 10074 + }, + { + "epoch": 1.7938034188034186, + "grad_norm": 0.6514154672622681, + "learning_rate": 0.00011630393311685835, + "loss": 0.8967, + "step": 10075 + }, + { + "epoch": 1.7939814814814814, + "grad_norm": 0.8669198155403137, + "learning_rate": 0.00011629012277720607, + "loss": 1.0362, + "step": 10076 + }, + { + "epoch": 1.7941595441595442, + "grad_norm": 0.7256068587303162, + "learning_rate": 0.00011627631211838266, + "loss": 1.1948, + "step": 10077 + }, + { + "epoch": 1.794337606837607, + "grad_norm": 0.6504935622215271, + "learning_rate": 0.00011626250114065875, + "loss": 0.8309, + "step": 10078 + }, + { + "epoch": 1.7945156695156697, + "grad_norm": 0.6964160799980164, + "learning_rate": 0.0001162486898443049, + "loss": 0.9593, + "step": 10079 + }, + { + "epoch": 1.7946937321937322, + "grad_norm": 0.668727695941925, + "learning_rate": 0.00011623487822959174, + "loss": 0.8897, + "step": 10080 + }, + { + "epoch": 1.7948717948717947, + "grad_norm": 0.6907223463058472, + "learning_rate": 0.00011622106629678986, + "loss": 0.897, + "step": 10081 + }, + { + "epoch": 1.7950498575498575, + "grad_norm": 0.6652865409851074, + "learning_rate": 0.00011620725404616985, + "loss": 0.9321, + "step": 10082 + }, + { + "epoch": 1.7952279202279202, + "grad_norm": 0.6523811221122742, + "learning_rate": 0.00011619344147800239, + "loss": 0.8991, + "step": 10083 + }, + { + "epoch": 1.795405982905983, + "grad_norm": 0.6162952184677124, + "learning_rate": 0.0001161796285925581, + "loss": 0.8061, + "step": 10084 + }, + { + "epoch": 1.7955840455840457, + "grad_norm": 0.670606791973114, + "learning_rate": 0.0001161658153901076, + "loss": 0.9341, + "step": 10085 + }, + { + "epoch": 1.7957621082621082, + "grad_norm": 0.6372489333152771, + "learning_rate": 0.00011615200187092148, + "loss": 1.1049, + "step": 10086 + }, + { + "epoch": 1.7959401709401708, + "grad_norm": 0.7311037182807922, + "learning_rate": 0.00011613818803527045, + "loss": 1.0881, + "step": 10087 + }, + { + "epoch": 1.7961182336182335, + "grad_norm": 0.7440751194953918, + "learning_rate": 0.00011612437388342518, + "loss": 0.9487, + "step": 10088 + }, + { + "epoch": 1.7962962962962963, + "grad_norm": 0.6605934500694275, + "learning_rate": 0.00011611055941565629, + "loss": 0.8757, + "step": 10089 + }, + { + "epoch": 1.796474358974359, + "grad_norm": 0.7546001076698303, + "learning_rate": 0.00011609674463223446, + "loss": 0.9368, + "step": 10090 + }, + { + "epoch": 1.7966524216524218, + "grad_norm": 0.7001389861106873, + "learning_rate": 0.00011608292953343036, + "loss": 0.9098, + "step": 10091 + }, + { + "epoch": 1.7968304843304843, + "grad_norm": 0.6898102760314941, + "learning_rate": 0.00011606911411951462, + "loss": 0.8821, + "step": 10092 + }, + { + "epoch": 1.797008547008547, + "grad_norm": 0.7020773887634277, + "learning_rate": 0.00011605529839075801, + "loss": 1.2775, + "step": 10093 + }, + { + "epoch": 1.7971866096866096, + "grad_norm": 0.6061446070671082, + "learning_rate": 0.0001160414823474312, + "loss": 1.0156, + "step": 10094 + }, + { + "epoch": 1.7973646723646723, + "grad_norm": 0.6746069192886353, + "learning_rate": 0.00011602766598980484, + "loss": 0.8223, + "step": 10095 + }, + { + "epoch": 1.797542735042735, + "grad_norm": 0.655829131603241, + "learning_rate": 0.00011601384931814967, + "loss": 0.9482, + "step": 10096 + }, + { + "epoch": 1.7977207977207978, + "grad_norm": 0.6762703061103821, + "learning_rate": 0.00011600003233273636, + "loss": 1.0191, + "step": 10097 + }, + { + "epoch": 1.7978988603988604, + "grad_norm": 0.7610527276992798, + "learning_rate": 0.00011598621503383566, + "loss": 1.0771, + "step": 10098 + }, + { + "epoch": 1.7980769230769231, + "grad_norm": 0.6857240200042725, + "learning_rate": 0.0001159723974217183, + "loss": 0.8325, + "step": 10099 + }, + { + "epoch": 1.7982549857549857, + "grad_norm": 0.6897954940795898, + "learning_rate": 0.00011595857949665501, + "loss": 1.0064, + "step": 10100 + }, + { + "epoch": 1.7984330484330484, + "grad_norm": 0.7023211717605591, + "learning_rate": 0.00011594476125891649, + "loss": 1.1346, + "step": 10101 + }, + { + "epoch": 1.7986111111111112, + "grad_norm": 0.8131003975868225, + "learning_rate": 0.00011593094270877347, + "loss": 1.0384, + "step": 10102 + }, + { + "epoch": 1.798789173789174, + "grad_norm": 0.6504445672035217, + "learning_rate": 0.00011591712384649676, + "loss": 0.8172, + "step": 10103 + }, + { + "epoch": 1.7989672364672364, + "grad_norm": 0.7379748821258545, + "learning_rate": 0.00011590330467235704, + "loss": 1.0118, + "step": 10104 + }, + { + "epoch": 1.7991452991452992, + "grad_norm": 0.8867329955101013, + "learning_rate": 0.0001158894851866251, + "loss": 1.023, + "step": 10105 + }, + { + "epoch": 1.7993233618233617, + "grad_norm": 0.7057412266731262, + "learning_rate": 0.00011587566538957173, + "loss": 0.8415, + "step": 10106 + }, + { + "epoch": 1.7995014245014245, + "grad_norm": 0.7479654550552368, + "learning_rate": 0.00011586184528146769, + "loss": 0.9663, + "step": 10107 + }, + { + "epoch": 1.7996794871794872, + "grad_norm": 0.6280845403671265, + "learning_rate": 0.00011584802486258368, + "loss": 0.973, + "step": 10108 + }, + { + "epoch": 1.79985754985755, + "grad_norm": 0.6735749840736389, + "learning_rate": 0.00011583420413319059, + "loss": 0.8631, + "step": 10109 + }, + { + "epoch": 1.8000356125356125, + "grad_norm": 0.5940406918525696, + "learning_rate": 0.00011582038309355918, + "loss": 0.8533, + "step": 10110 + }, + { + "epoch": 1.8002136752136753, + "grad_norm": 0.6923874020576477, + "learning_rate": 0.00011580656174396021, + "loss": 1.1105, + "step": 10111 + }, + { + "epoch": 1.8003917378917378, + "grad_norm": 0.6996715664863586, + "learning_rate": 0.00011579274008466447, + "loss": 0.9952, + "step": 10112 + }, + { + "epoch": 1.8005698005698005, + "grad_norm": 0.656561553478241, + "learning_rate": 0.00011577891811594281, + "loss": 0.9621, + "step": 10113 + }, + { + "epoch": 1.8007478632478633, + "grad_norm": 0.7121242880821228, + "learning_rate": 0.00011576509583806605, + "loss": 0.8658, + "step": 10114 + }, + { + "epoch": 1.800925925925926, + "grad_norm": 0.7864459753036499, + "learning_rate": 0.00011575127325130498, + "loss": 0.9867, + "step": 10115 + }, + { + "epoch": 1.8011039886039886, + "grad_norm": 0.6086452007293701, + "learning_rate": 0.00011573745035593042, + "loss": 0.8625, + "step": 10116 + }, + { + "epoch": 1.8012820512820513, + "grad_norm": 0.6553642749786377, + "learning_rate": 0.00011572362715221321, + "loss": 0.8475, + "step": 10117 + }, + { + "epoch": 1.8014601139601139, + "grad_norm": 0.6677348017692566, + "learning_rate": 0.00011570980364042419, + "loss": 0.9672, + "step": 10118 + }, + { + "epoch": 1.8016381766381766, + "grad_norm": 0.6275015473365784, + "learning_rate": 0.0001156959798208342, + "loss": 0.8663, + "step": 10119 + }, + { + "epoch": 1.8018162393162394, + "grad_norm": 0.787568211555481, + "learning_rate": 0.0001156821556937141, + "loss": 1.0188, + "step": 10120 + }, + { + "epoch": 1.801994301994302, + "grad_norm": 0.6983163356781006, + "learning_rate": 0.00011566833125933473, + "loss": 1.0767, + "step": 10121 + }, + { + "epoch": 1.8021723646723646, + "grad_norm": 0.7008936405181885, + "learning_rate": 0.00011565450651796695, + "loss": 1.0116, + "step": 10122 + }, + { + "epoch": 1.8023504273504274, + "grad_norm": 0.7694976925849915, + "learning_rate": 0.00011564068146988163, + "loss": 1.0227, + "step": 10123 + }, + { + "epoch": 1.80252849002849, + "grad_norm": 0.9530014991760254, + "learning_rate": 0.00011562685611534967, + "loss": 0.907, + "step": 10124 + }, + { + "epoch": 1.8027065527065527, + "grad_norm": 0.6714984178543091, + "learning_rate": 0.00011561303045464189, + "loss": 0.9501, + "step": 10125 + }, + { + "epoch": 1.8028846153846154, + "grad_norm": 0.7233797311782837, + "learning_rate": 0.00011559920448802925, + "loss": 1.021, + "step": 10126 + }, + { + "epoch": 1.8030626780626782, + "grad_norm": 0.7600540518760681, + "learning_rate": 0.0001155853782157826, + "loss": 1.1056, + "step": 10127 + }, + { + "epoch": 1.8032407407407407, + "grad_norm": 0.7836297750473022, + "learning_rate": 0.00011557155163817281, + "loss": 0.9906, + "step": 10128 + }, + { + "epoch": 1.8034188034188035, + "grad_norm": 0.7161104083061218, + "learning_rate": 0.00011555772475547084, + "loss": 0.9541, + "step": 10129 + }, + { + "epoch": 1.803596866096866, + "grad_norm": 0.6613732576370239, + "learning_rate": 0.00011554389756794757, + "loss": 0.9188, + "step": 10130 + }, + { + "epoch": 1.8037749287749287, + "grad_norm": 0.6415915489196777, + "learning_rate": 0.00011553007007587391, + "loss": 0.9928, + "step": 10131 + }, + { + "epoch": 1.8039529914529915, + "grad_norm": 0.7730516195297241, + "learning_rate": 0.0001155162422795208, + "loss": 1.0654, + "step": 10132 + }, + { + "epoch": 1.8041310541310542, + "grad_norm": 0.6769654750823975, + "learning_rate": 0.00011550241417915913, + "loss": 1.0678, + "step": 10133 + }, + { + "epoch": 1.8043091168091168, + "grad_norm": 0.6542425751686096, + "learning_rate": 0.00011548858577505988, + "loss": 0.9796, + "step": 10134 + }, + { + "epoch": 1.8044871794871795, + "grad_norm": 0.7282404899597168, + "learning_rate": 0.00011547475706749395, + "loss": 1.0314, + "step": 10135 + }, + { + "epoch": 1.804665242165242, + "grad_norm": 0.6450245976448059, + "learning_rate": 0.00011546092805673232, + "loss": 0.9564, + "step": 10136 + }, + { + "epoch": 1.8048433048433048, + "grad_norm": 0.65577632188797, + "learning_rate": 0.0001154470987430459, + "loss": 1.0219, + "step": 10137 + }, + { + "epoch": 1.8050213675213675, + "grad_norm": 0.7151737809181213, + "learning_rate": 0.00011543326912670567, + "loss": 0.9245, + "step": 10138 + }, + { + "epoch": 1.8051994301994303, + "grad_norm": 0.6695905327796936, + "learning_rate": 0.00011541943920798259, + "loss": 0.9535, + "step": 10139 + }, + { + "epoch": 1.8053774928774928, + "grad_norm": 0.7443813681602478, + "learning_rate": 0.00011540560898714767, + "loss": 1.1697, + "step": 10140 + }, + { + "epoch": 1.8055555555555556, + "grad_norm": 0.5701992511749268, + "learning_rate": 0.0001153917784644718, + "loss": 0.7868, + "step": 10141 + }, + { + "epoch": 1.805733618233618, + "grad_norm": 0.6992354989051819, + "learning_rate": 0.00011537794764022605, + "loss": 0.9856, + "step": 10142 + }, + { + "epoch": 1.8059116809116809, + "grad_norm": 0.6354477405548096, + "learning_rate": 0.00011536411651468131, + "loss": 0.8752, + "step": 10143 + }, + { + "epoch": 1.8060897435897436, + "grad_norm": 0.6952932476997375, + "learning_rate": 0.00011535028508810864, + "loss": 0.9446, + "step": 10144 + }, + { + "epoch": 1.8062678062678064, + "grad_norm": 0.5527541637420654, + "learning_rate": 0.00011533645336077901, + "loss": 0.5486, + "step": 10145 + }, + { + "epoch": 1.806445868945869, + "grad_norm": 0.685046374797821, + "learning_rate": 0.00011532262133296345, + "loss": 0.9529, + "step": 10146 + }, + { + "epoch": 1.8066239316239316, + "grad_norm": 0.6927558779716492, + "learning_rate": 0.00011530878900493296, + "loss": 1.1758, + "step": 10147 + }, + { + "epoch": 1.8068019943019942, + "grad_norm": 0.6758309602737427, + "learning_rate": 0.00011529495637695855, + "loss": 1.0076, + "step": 10148 + }, + { + "epoch": 1.806980056980057, + "grad_norm": 0.6739441156387329, + "learning_rate": 0.00011528112344931121, + "loss": 1.1914, + "step": 10149 + }, + { + "epoch": 1.8071581196581197, + "grad_norm": 0.7031944394111633, + "learning_rate": 0.00011526729022226204, + "loss": 0.783, + "step": 10150 + }, + { + "epoch": 1.8073361823361824, + "grad_norm": 0.6476930975914001, + "learning_rate": 0.00011525345669608202, + "loss": 0.9595, + "step": 10151 + }, + { + "epoch": 1.8075142450142452, + "grad_norm": 0.710498571395874, + "learning_rate": 0.00011523962287104222, + "loss": 0.8821, + "step": 10152 + }, + { + "epoch": 1.8076923076923077, + "grad_norm": 0.6664412617683411, + "learning_rate": 0.00011522578874741365, + "loss": 1.0182, + "step": 10153 + }, + { + "epoch": 1.8078703703703702, + "grad_norm": 0.8374263048171997, + "learning_rate": 0.00011521195432546737, + "loss": 0.9394, + "step": 10154 + }, + { + "epoch": 1.808048433048433, + "grad_norm": 0.6770764589309692, + "learning_rate": 0.00011519811960547447, + "loss": 1.0568, + "step": 10155 + }, + { + "epoch": 1.8082264957264957, + "grad_norm": 0.7014045715332031, + "learning_rate": 0.00011518428458770595, + "loss": 1.1705, + "step": 10156 + }, + { + "epoch": 1.8084045584045585, + "grad_norm": 0.6590061187744141, + "learning_rate": 0.00011517044927243295, + "loss": 1.1233, + "step": 10157 + }, + { + "epoch": 1.8085826210826212, + "grad_norm": 0.6093801856040955, + "learning_rate": 0.00011515661365992647, + "loss": 0.953, + "step": 10158 + }, + { + "epoch": 1.8087606837606838, + "grad_norm": 0.6197089552879333, + "learning_rate": 0.00011514277775045768, + "loss": 0.9414, + "step": 10159 + }, + { + "epoch": 1.8089387464387463, + "grad_norm": 0.7530463337898254, + "learning_rate": 0.00011512894154429759, + "loss": 0.9168, + "step": 10160 + }, + { + "epoch": 1.809116809116809, + "grad_norm": 0.6051347851753235, + "learning_rate": 0.00011511510504171735, + "loss": 0.9132, + "step": 10161 + }, + { + "epoch": 1.8092948717948718, + "grad_norm": 0.6388311982154846, + "learning_rate": 0.000115101268242988, + "loss": 0.6551, + "step": 10162 + }, + { + "epoch": 1.8094729344729346, + "grad_norm": 0.7040972709655762, + "learning_rate": 0.00011508743114838063, + "loss": 0.9409, + "step": 10163 + }, + { + "epoch": 1.8096509971509973, + "grad_norm": 0.7669548392295837, + "learning_rate": 0.00011507359375816644, + "loss": 1.0376, + "step": 10164 + }, + { + "epoch": 1.8098290598290598, + "grad_norm": 0.7309662699699402, + "learning_rate": 0.00011505975607261646, + "loss": 0.9071, + "step": 10165 + }, + { + "epoch": 1.8100071225071224, + "grad_norm": 0.6624547839164734, + "learning_rate": 0.00011504591809200187, + "loss": 1.0765, + "step": 10166 + }, + { + "epoch": 1.8101851851851851, + "grad_norm": 0.7719045281410217, + "learning_rate": 0.00011503207981659376, + "loss": 0.9244, + "step": 10167 + }, + { + "epoch": 1.8103632478632479, + "grad_norm": 0.6701484322547913, + "learning_rate": 0.0001150182412466633, + "loss": 0.9475, + "step": 10168 + }, + { + "epoch": 1.8105413105413106, + "grad_norm": 0.5604981184005737, + "learning_rate": 0.00011500440238248154, + "loss": 0.6268, + "step": 10169 + }, + { + "epoch": 1.8107193732193734, + "grad_norm": 0.6736510992050171, + "learning_rate": 0.00011499056322431973, + "loss": 0.9088, + "step": 10170 + }, + { + "epoch": 1.810897435897436, + "grad_norm": 0.7428455948829651, + "learning_rate": 0.00011497672377244897, + "loss": 0.9298, + "step": 10171 + }, + { + "epoch": 1.8110754985754984, + "grad_norm": 0.6543142795562744, + "learning_rate": 0.00011496288402714042, + "loss": 0.8863, + "step": 10172 + }, + { + "epoch": 1.8112535612535612, + "grad_norm": 0.6809250712394714, + "learning_rate": 0.00011494904398866524, + "loss": 0.977, + "step": 10173 + }, + { + "epoch": 1.811431623931624, + "grad_norm": 0.8105120062828064, + "learning_rate": 0.00011493520365729456, + "loss": 1.2115, + "step": 10174 + }, + { + "epoch": 1.8116096866096867, + "grad_norm": 0.6985095143318176, + "learning_rate": 0.00011492136303329964, + "loss": 0.8233, + "step": 10175 + }, + { + "epoch": 1.8117877492877494, + "grad_norm": 0.7198361754417419, + "learning_rate": 0.00011490752211695158, + "loss": 1.0552, + "step": 10176 + }, + { + "epoch": 1.811965811965812, + "grad_norm": 0.7077036499977112, + "learning_rate": 0.0001148936809085216, + "loss": 0.9171, + "step": 10177 + }, + { + "epoch": 1.8121438746438745, + "grad_norm": 0.9362925887107849, + "learning_rate": 0.00011487983940828089, + "loss": 0.9042, + "step": 10178 + }, + { + "epoch": 1.8123219373219372, + "grad_norm": 0.6732819676399231, + "learning_rate": 0.0001148659976165006, + "loss": 1.1033, + "step": 10179 + }, + { + "epoch": 1.8125, + "grad_norm": 0.747702419757843, + "learning_rate": 0.00011485215553345201, + "loss": 1.0692, + "step": 10180 + }, + { + "epoch": 1.8126780626780628, + "grad_norm": 0.7011259198188782, + "learning_rate": 0.00011483831315940627, + "loss": 0.9278, + "step": 10181 + }, + { + "epoch": 1.8128561253561255, + "grad_norm": 0.8542702198028564, + "learning_rate": 0.00011482447049463462, + "loss": 0.9476, + "step": 10182 + }, + { + "epoch": 1.813034188034188, + "grad_norm": 0.6975166201591492, + "learning_rate": 0.00011481062753940825, + "loss": 0.9486, + "step": 10183 + }, + { + "epoch": 1.8132122507122506, + "grad_norm": 0.8239036798477173, + "learning_rate": 0.0001147967842939984, + "loss": 1.0518, + "step": 10184 + }, + { + "epoch": 1.8133903133903133, + "grad_norm": 0.7559717297554016, + "learning_rate": 0.00011478294075867628, + "loss": 1.1877, + "step": 10185 + }, + { + "epoch": 1.813568376068376, + "grad_norm": 0.6755532026290894, + "learning_rate": 0.00011476909693371318, + "loss": 0.9287, + "step": 10186 + }, + { + "epoch": 1.8137464387464388, + "grad_norm": 0.6561332941055298, + "learning_rate": 0.0001147552528193803, + "loss": 0.83, + "step": 10187 + }, + { + "epoch": 1.8139245014245016, + "grad_norm": 0.7223508954048157, + "learning_rate": 0.00011474140841594887, + "loss": 1.1259, + "step": 10188 + }, + { + "epoch": 1.814102564102564, + "grad_norm": 0.7920593023300171, + "learning_rate": 0.0001147275637236902, + "loss": 1.0925, + "step": 10189 + }, + { + "epoch": 1.8142806267806266, + "grad_norm": 0.6896616816520691, + "learning_rate": 0.00011471371874287546, + "loss": 1.0204, + "step": 10190 + }, + { + "epoch": 1.8144586894586894, + "grad_norm": 0.6149865388870239, + "learning_rate": 0.00011469987347377602, + "loss": 1.1249, + "step": 10191 + }, + { + "epoch": 1.8146367521367521, + "grad_norm": 0.6650002598762512, + "learning_rate": 0.00011468602791666307, + "loss": 0.9723, + "step": 10192 + }, + { + "epoch": 1.8148148148148149, + "grad_norm": 0.7298738956451416, + "learning_rate": 0.00011467218207180792, + "loss": 1.0225, + "step": 10193 + }, + { + "epoch": 1.8149928774928776, + "grad_norm": 0.8075628876686096, + "learning_rate": 0.00011465833593948183, + "loss": 1.0429, + "step": 10194 + }, + { + "epoch": 1.8151709401709402, + "grad_norm": 0.8196593523025513, + "learning_rate": 0.0001146444895199561, + "loss": 0.9148, + "step": 10195 + }, + { + "epoch": 1.8153490028490027, + "grad_norm": 0.6394698023796082, + "learning_rate": 0.00011463064281350204, + "loss": 0.9781, + "step": 10196 + }, + { + "epoch": 1.8155270655270654, + "grad_norm": 0.7302836775779724, + "learning_rate": 0.00011461679582039091, + "loss": 1.0394, + "step": 10197 + }, + { + "epoch": 1.8157051282051282, + "grad_norm": 0.7066670060157776, + "learning_rate": 0.00011460294854089404, + "loss": 1.1153, + "step": 10198 + }, + { + "epoch": 1.815883190883191, + "grad_norm": 0.6471068263053894, + "learning_rate": 0.0001145891009752827, + "loss": 1.1533, + "step": 10199 + }, + { + "epoch": 1.8160612535612537, + "grad_norm": 0.6842355132102966, + "learning_rate": 0.00011457525312382826, + "loss": 0.953, + "step": 10200 + }, + { + "epoch": 1.8162393162393162, + "grad_norm": 0.6720319986343384, + "learning_rate": 0.00011456140498680202, + "loss": 1.003, + "step": 10201 + }, + { + "epoch": 1.8164173789173788, + "grad_norm": 0.632017970085144, + "learning_rate": 0.00011454755656447527, + "loss": 0.8148, + "step": 10202 + }, + { + "epoch": 1.8165954415954415, + "grad_norm": 0.7193828225135803, + "learning_rate": 0.00011453370785711939, + "loss": 1.0098, + "step": 10203 + }, + { + "epoch": 1.8167735042735043, + "grad_norm": 0.7098045349121094, + "learning_rate": 0.00011451985886500566, + "loss": 1.1276, + "step": 10204 + }, + { + "epoch": 1.816951566951567, + "grad_norm": 0.7076733708381653, + "learning_rate": 0.00011450600958840547, + "loss": 1.1216, + "step": 10205 + }, + { + "epoch": 1.8171296296296298, + "grad_norm": 0.6864610314369202, + "learning_rate": 0.00011449216002759018, + "loss": 0.9896, + "step": 10206 + }, + { + "epoch": 1.8173076923076923, + "grad_norm": 0.737727701663971, + "learning_rate": 0.0001144783101828311, + "loss": 0.9447, + "step": 10207 + }, + { + "epoch": 1.8174857549857548, + "grad_norm": 0.6562525033950806, + "learning_rate": 0.00011446446005439964, + "loss": 1.1208, + "step": 10208 + }, + { + "epoch": 1.8176638176638176, + "grad_norm": 0.7203826308250427, + "learning_rate": 0.0001144506096425671, + "loss": 1.1339, + "step": 10209 + }, + { + "epoch": 1.8178418803418803, + "grad_norm": 0.6657233834266663, + "learning_rate": 0.00011443675894760489, + "loss": 0.8307, + "step": 10210 + }, + { + "epoch": 1.818019943019943, + "grad_norm": 0.7032586932182312, + "learning_rate": 0.00011442290796978437, + "loss": 0.8546, + "step": 10211 + }, + { + "epoch": 1.8181980056980058, + "grad_norm": 0.6989460587501526, + "learning_rate": 0.00011440905670937696, + "loss": 1.0749, + "step": 10212 + }, + { + "epoch": 1.8183760683760684, + "grad_norm": 0.6461085677146912, + "learning_rate": 0.00011439520516665399, + "loss": 0.984, + "step": 10213 + }, + { + "epoch": 1.818554131054131, + "grad_norm": 0.7077372670173645, + "learning_rate": 0.00011438135334188689, + "loss": 1.0813, + "step": 10214 + }, + { + "epoch": 1.8187321937321936, + "grad_norm": 0.6724075675010681, + "learning_rate": 0.00011436750123534704, + "loss": 0.9975, + "step": 10215 + }, + { + "epoch": 1.8189102564102564, + "grad_norm": 0.6205753684043884, + "learning_rate": 0.00011435364884730583, + "loss": 0.7414, + "step": 10216 + }, + { + "epoch": 1.8190883190883191, + "grad_norm": 0.6416093707084656, + "learning_rate": 0.00011433979617803472, + "loss": 1.0024, + "step": 10217 + }, + { + "epoch": 1.819266381766382, + "grad_norm": 0.7817183136940002, + "learning_rate": 0.00011432594322780508, + "loss": 1.0577, + "step": 10218 + }, + { + "epoch": 1.8194444444444444, + "grad_norm": 0.688220202922821, + "learning_rate": 0.00011431208999688835, + "loss": 1.0301, + "step": 10219 + }, + { + "epoch": 1.8196225071225072, + "grad_norm": 0.6464754343032837, + "learning_rate": 0.0001142982364855559, + "loss": 1.0608, + "step": 10220 + }, + { + "epoch": 1.8198005698005697, + "grad_norm": 0.6607306599617004, + "learning_rate": 0.00011428438269407926, + "loss": 1.1203, + "step": 10221 + }, + { + "epoch": 1.8199786324786325, + "grad_norm": 0.5779942870140076, + "learning_rate": 0.00011427052862272982, + "loss": 0.7895, + "step": 10222 + }, + { + "epoch": 1.8201566951566952, + "grad_norm": 0.7599068880081177, + "learning_rate": 0.000114256674271779, + "loss": 0.883, + "step": 10223 + }, + { + "epoch": 1.820334757834758, + "grad_norm": 0.6578865051269531, + "learning_rate": 0.00011424281964149824, + "loss": 1.101, + "step": 10224 + }, + { + "epoch": 1.8205128205128205, + "grad_norm": 0.7090746760368347, + "learning_rate": 0.00011422896473215905, + "loss": 0.9514, + "step": 10225 + }, + { + "epoch": 1.8206908831908832, + "grad_norm": 0.7537758946418762, + "learning_rate": 0.00011421510954403281, + "loss": 1.2193, + "step": 10226 + }, + { + "epoch": 1.8208689458689458, + "grad_norm": 0.670183002948761, + "learning_rate": 0.00011420125407739106, + "loss": 1.1408, + "step": 10227 + }, + { + "epoch": 1.8210470085470085, + "grad_norm": 0.742520809173584, + "learning_rate": 0.00011418739833250524, + "loss": 0.8826, + "step": 10228 + }, + { + "epoch": 1.8212250712250713, + "grad_norm": 0.6542800664901733, + "learning_rate": 0.00011417354230964683, + "loss": 1.0039, + "step": 10229 + }, + { + "epoch": 1.821403133903134, + "grad_norm": 0.6713709235191345, + "learning_rate": 0.00011415968600908727, + "loss": 0.9351, + "step": 10230 + }, + { + "epoch": 1.8215811965811965, + "grad_norm": 0.6794951558113098, + "learning_rate": 0.0001141458294310981, + "loss": 0.9491, + "step": 10231 + }, + { + "epoch": 1.8217592592592593, + "grad_norm": 0.6921972632408142, + "learning_rate": 0.00011413197257595079, + "loss": 1.1342, + "step": 10232 + }, + { + "epoch": 1.8219373219373218, + "grad_norm": 0.702586829662323, + "learning_rate": 0.00011411811544391682, + "loss": 0.9992, + "step": 10233 + }, + { + "epoch": 1.8221153846153846, + "grad_norm": 0.8147975206375122, + "learning_rate": 0.00011410425803526772, + "loss": 1.0507, + "step": 10234 + }, + { + "epoch": 1.8222934472934473, + "grad_norm": 0.66419517993927, + "learning_rate": 0.00011409040035027496, + "loss": 1.0426, + "step": 10235 + }, + { + "epoch": 1.82247150997151, + "grad_norm": 0.6132485866546631, + "learning_rate": 0.00011407654238921011, + "loss": 0.9859, + "step": 10236 + }, + { + "epoch": 1.8226495726495726, + "grad_norm": 0.7522366046905518, + "learning_rate": 0.00011406268415234462, + "loss": 0.9379, + "step": 10237 + }, + { + "epoch": 1.8228276353276354, + "grad_norm": 0.6335554122924805, + "learning_rate": 0.00011404882563995007, + "loss": 0.9322, + "step": 10238 + }, + { + "epoch": 1.823005698005698, + "grad_norm": 0.7577497363090515, + "learning_rate": 0.00011403496685229797, + "loss": 1.1383, + "step": 10239 + }, + { + "epoch": 1.8231837606837606, + "grad_norm": 0.6796886920928955, + "learning_rate": 0.00011402110778965982, + "loss": 1.0092, + "step": 10240 + }, + { + "epoch": 1.8233618233618234, + "grad_norm": 0.7676617503166199, + "learning_rate": 0.0001140072484523072, + "loss": 1.0137, + "step": 10241 + }, + { + "epoch": 1.8235398860398861, + "grad_norm": 0.7807821035385132, + "learning_rate": 0.00011399338884051165, + "loss": 0.8987, + "step": 10242 + }, + { + "epoch": 1.8237179487179487, + "grad_norm": 0.7169568538665771, + "learning_rate": 0.00011397952895454473, + "loss": 0.8984, + "step": 10243 + }, + { + "epoch": 1.8238960113960114, + "grad_norm": 0.6564654111862183, + "learning_rate": 0.00011396566879467793, + "loss": 1.0255, + "step": 10244 + }, + { + "epoch": 1.824074074074074, + "grad_norm": 0.7290034294128418, + "learning_rate": 0.00011395180836118292, + "loss": 0.9962, + "step": 10245 + }, + { + "epoch": 1.8242521367521367, + "grad_norm": 0.6610758900642395, + "learning_rate": 0.00011393794765433115, + "loss": 1.102, + "step": 10246 + }, + { + "epoch": 1.8244301994301995, + "grad_norm": 0.6875932216644287, + "learning_rate": 0.0001139240866743943, + "loss": 0.9963, + "step": 10247 + }, + { + "epoch": 1.8246082621082622, + "grad_norm": 0.7595645189285278, + "learning_rate": 0.00011391022542164387, + "loss": 1.1285, + "step": 10248 + }, + { + "epoch": 1.8247863247863247, + "grad_norm": 0.6752721667289734, + "learning_rate": 0.0001138963638963515, + "loss": 0.9447, + "step": 10249 + }, + { + "epoch": 1.8249643874643875, + "grad_norm": 0.6697955131530762, + "learning_rate": 0.00011388250209878873, + "loss": 1.0804, + "step": 10250 + }, + { + "epoch": 1.82514245014245, + "grad_norm": 0.6546956896781921, + "learning_rate": 0.00011386864002922713, + "loss": 0.9626, + "step": 10251 + }, + { + "epoch": 1.8253205128205128, + "grad_norm": 0.8002896904945374, + "learning_rate": 0.00011385477768793838, + "loss": 1.1933, + "step": 10252 + }, + { + "epoch": 1.8254985754985755, + "grad_norm": 0.6566781401634216, + "learning_rate": 0.00011384091507519403, + "loss": 0.9802, + "step": 10253 + }, + { + "epoch": 1.8256766381766383, + "grad_norm": 0.617420494556427, + "learning_rate": 0.00011382705219126572, + "loss": 1.1098, + "step": 10254 + }, + { + "epoch": 1.8258547008547008, + "grad_norm": 0.6558036208152771, + "learning_rate": 0.00011381318903642504, + "loss": 1.0291, + "step": 10255 + }, + { + "epoch": 1.8260327635327636, + "grad_norm": 0.6295637488365173, + "learning_rate": 0.00011379932561094358, + "loss": 1.0792, + "step": 10256 + }, + { + "epoch": 1.826210826210826, + "grad_norm": 0.7475154399871826, + "learning_rate": 0.00011378546191509303, + "loss": 1.1362, + "step": 10257 + }, + { + "epoch": 1.8263888888888888, + "grad_norm": 0.6814939379692078, + "learning_rate": 0.00011377159794914498, + "loss": 0.9131, + "step": 10258 + }, + { + "epoch": 1.8265669515669516, + "grad_norm": 0.6726876497268677, + "learning_rate": 0.00011375773371337111, + "loss": 0.9147, + "step": 10259 + }, + { + "epoch": 1.8267450142450143, + "grad_norm": 0.785943865776062, + "learning_rate": 0.00011374386920804298, + "loss": 1.0137, + "step": 10260 + }, + { + "epoch": 1.8269230769230769, + "grad_norm": 0.7614478468894958, + "learning_rate": 0.0001137300044334323, + "loss": 1.2118, + "step": 10261 + }, + { + "epoch": 1.8271011396011396, + "grad_norm": 0.7317564487457275, + "learning_rate": 0.00011371613938981072, + "loss": 1.0602, + "step": 10262 + }, + { + "epoch": 1.8272792022792022, + "grad_norm": 0.6716432571411133, + "learning_rate": 0.00011370227407744987, + "loss": 0.952, + "step": 10263 + }, + { + "epoch": 1.827457264957265, + "grad_norm": 0.6946425437927246, + "learning_rate": 0.00011368840849662139, + "loss": 1.0554, + "step": 10264 + }, + { + "epoch": 1.8276353276353277, + "grad_norm": 0.6692264080047607, + "learning_rate": 0.00011367454264759703, + "loss": 0.8944, + "step": 10265 + }, + { + "epoch": 1.8278133903133904, + "grad_norm": 0.6931505799293518, + "learning_rate": 0.00011366067653064838, + "loss": 0.9045, + "step": 10266 + }, + { + "epoch": 1.827991452991453, + "grad_norm": 0.7233194708824158, + "learning_rate": 0.00011364681014604716, + "loss": 0.9441, + "step": 10267 + }, + { + "epoch": 1.8281695156695157, + "grad_norm": 0.6451242566108704, + "learning_rate": 0.00011363294349406506, + "loss": 0.9948, + "step": 10268 + }, + { + "epoch": 1.8283475783475782, + "grad_norm": 0.6993351578712463, + "learning_rate": 0.00011361907657497375, + "loss": 1.1057, + "step": 10269 + }, + { + "epoch": 1.828525641025641, + "grad_norm": 0.7241137623786926, + "learning_rate": 0.00011360520938904493, + "loss": 0.974, + "step": 10270 + }, + { + "epoch": 1.8287037037037037, + "grad_norm": 0.6349480152130127, + "learning_rate": 0.00011359134193655027, + "loss": 0.9026, + "step": 10271 + }, + { + "epoch": 1.8288817663817665, + "grad_norm": 0.6916826963424683, + "learning_rate": 0.00011357747421776151, + "loss": 0.9153, + "step": 10272 + }, + { + "epoch": 1.8290598290598292, + "grad_norm": 0.879770040512085, + "learning_rate": 0.00011356360623295037, + "loss": 1.0818, + "step": 10273 + }, + { + "epoch": 1.8292378917378918, + "grad_norm": 0.6293807029724121, + "learning_rate": 0.00011354973798238853, + "loss": 1.1164, + "step": 10274 + }, + { + "epoch": 1.8294159544159543, + "grad_norm": 0.7070622444152832, + "learning_rate": 0.0001135358694663477, + "loss": 0.8795, + "step": 10275 + }, + { + "epoch": 1.829594017094017, + "grad_norm": 0.6847673654556274, + "learning_rate": 0.00011352200068509962, + "loss": 0.9173, + "step": 10276 + }, + { + "epoch": 1.8297720797720798, + "grad_norm": 0.6552146077156067, + "learning_rate": 0.00011350813163891605, + "loss": 1.0425, + "step": 10277 + }, + { + "epoch": 1.8299501424501425, + "grad_norm": 0.6432808041572571, + "learning_rate": 0.0001134942623280687, + "loss": 0.9418, + "step": 10278 + }, + { + "epoch": 1.8301282051282053, + "grad_norm": 0.7412393093109131, + "learning_rate": 0.00011348039275282931, + "loss": 1.1212, + "step": 10279 + }, + { + "epoch": 1.8303062678062678, + "grad_norm": 0.6543423533439636, + "learning_rate": 0.00011346652291346965, + "loss": 1.0553, + "step": 10280 + }, + { + "epoch": 1.8304843304843303, + "grad_norm": 0.7159286141395569, + "learning_rate": 0.00011345265281026138, + "loss": 1.0582, + "step": 10281 + }, + { + "epoch": 1.830662393162393, + "grad_norm": 0.6443323493003845, + "learning_rate": 0.00011343878244347639, + "loss": 0.9462, + "step": 10282 + }, + { + "epoch": 1.8308404558404558, + "grad_norm": 0.7592014074325562, + "learning_rate": 0.00011342491181338634, + "loss": 1.2718, + "step": 10283 + }, + { + "epoch": 1.8310185185185186, + "grad_norm": 0.627109944820404, + "learning_rate": 0.00011341104092026302, + "loss": 1.0177, + "step": 10284 + }, + { + "epoch": 1.8311965811965814, + "grad_norm": 0.8061598539352417, + "learning_rate": 0.00011339716976437827, + "loss": 0.9416, + "step": 10285 + }, + { + "epoch": 1.8313746438746439, + "grad_norm": 0.6584261059761047, + "learning_rate": 0.00011338329834600377, + "loss": 0.8297, + "step": 10286 + }, + { + "epoch": 1.8315527065527064, + "grad_norm": 0.6329470276832581, + "learning_rate": 0.00011336942666541133, + "loss": 0.8386, + "step": 10287 + }, + { + "epoch": 1.8317307692307692, + "grad_norm": 0.6833979487419128, + "learning_rate": 0.00011335555472287275, + "loss": 0.9407, + "step": 10288 + }, + { + "epoch": 1.831908831908832, + "grad_norm": 0.7663840651512146, + "learning_rate": 0.00011334168251865985, + "loss": 1.0018, + "step": 10289 + }, + { + "epoch": 1.8320868945868947, + "grad_norm": 0.7751262784004211, + "learning_rate": 0.00011332781005304436, + "loss": 1.0576, + "step": 10290 + }, + { + "epoch": 1.8322649572649574, + "grad_norm": 0.6857370138168335, + "learning_rate": 0.00011331393732629814, + "loss": 0.9888, + "step": 10291 + }, + { + "epoch": 1.83244301994302, + "grad_norm": 0.7534535527229309, + "learning_rate": 0.00011330006433869296, + "loss": 1.0834, + "step": 10292 + }, + { + "epoch": 1.8326210826210825, + "grad_norm": 0.6785250306129456, + "learning_rate": 0.00011328619109050065, + "loss": 1.0471, + "step": 10293 + }, + { + "epoch": 1.8327991452991452, + "grad_norm": 0.7023689150810242, + "learning_rate": 0.00011327231758199303, + "loss": 1.0652, + "step": 10294 + }, + { + "epoch": 1.832977207977208, + "grad_norm": 0.6776610612869263, + "learning_rate": 0.00011325844381344192, + "loss": 0.9504, + "step": 10295 + }, + { + "epoch": 1.8331552706552707, + "grad_norm": 0.7704112529754639, + "learning_rate": 0.00011324456978511917, + "loss": 0.9712, + "step": 10296 + }, + { + "epoch": 1.8333333333333335, + "grad_norm": 0.601502537727356, + "learning_rate": 0.00011323069549729654, + "loss": 1.075, + "step": 10297 + }, + { + "epoch": 1.833511396011396, + "grad_norm": 0.6282439231872559, + "learning_rate": 0.00011321682095024596, + "loss": 0.9238, + "step": 10298 + }, + { + "epoch": 1.8336894586894585, + "grad_norm": 0.6873499155044556, + "learning_rate": 0.00011320294614423921, + "loss": 1.0464, + "step": 10299 + }, + { + "epoch": 1.8338675213675213, + "grad_norm": 0.6063792705535889, + "learning_rate": 0.00011318907107954815, + "loss": 0.9732, + "step": 10300 + }, + { + "epoch": 1.834045584045584, + "grad_norm": 0.5830921530723572, + "learning_rate": 0.00011317519575644464, + "loss": 0.7568, + "step": 10301 + }, + { + "epoch": 1.8342236467236468, + "grad_norm": 0.6394222378730774, + "learning_rate": 0.00011316132017520053, + "loss": 0.9958, + "step": 10302 + }, + { + "epoch": 1.8344017094017095, + "grad_norm": 0.7052412033081055, + "learning_rate": 0.00011314744433608773, + "loss": 0.9129, + "step": 10303 + }, + { + "epoch": 1.834579772079772, + "grad_norm": 0.7287624478340149, + "learning_rate": 0.00011313356823937801, + "loss": 0.8608, + "step": 10304 + }, + { + "epoch": 1.8347578347578346, + "grad_norm": 0.702937662601471, + "learning_rate": 0.00011311969188534334, + "loss": 1.3074, + "step": 10305 + }, + { + "epoch": 1.8349358974358974, + "grad_norm": 0.6693850159645081, + "learning_rate": 0.00011310581527425557, + "loss": 0.928, + "step": 10306 + }, + { + "epoch": 1.83511396011396, + "grad_norm": 0.8153932094573975, + "learning_rate": 0.00011309193840638654, + "loss": 1.1771, + "step": 10307 + }, + { + "epoch": 1.8352920227920229, + "grad_norm": 0.6517418622970581, + "learning_rate": 0.00011307806128200821, + "loss": 0.9634, + "step": 10308 + }, + { + "epoch": 1.8354700854700856, + "grad_norm": 0.6626226305961609, + "learning_rate": 0.00011306418390139245, + "loss": 0.9371, + "step": 10309 + }, + { + "epoch": 1.8356481481481481, + "grad_norm": 0.7397477030754089, + "learning_rate": 0.0001130503062648111, + "loss": 0.9398, + "step": 10310 + }, + { + "epoch": 1.8358262108262107, + "grad_norm": 0.6790265440940857, + "learning_rate": 0.00011303642837253614, + "loss": 0.9728, + "step": 10311 + }, + { + "epoch": 1.8360042735042734, + "grad_norm": 0.6266449093818665, + "learning_rate": 0.00011302255022483941, + "loss": 0.847, + "step": 10312 + }, + { + "epoch": 1.8361823361823362, + "grad_norm": 0.791657030582428, + "learning_rate": 0.00011300867182199288, + "loss": 0.8342, + "step": 10313 + }, + { + "epoch": 1.836360398860399, + "grad_norm": 0.7128583788871765, + "learning_rate": 0.00011299479316426846, + "loss": 0.9591, + "step": 10314 + }, + { + "epoch": 1.8365384615384617, + "grad_norm": 0.659928023815155, + "learning_rate": 0.00011298091425193806, + "loss": 1.0282, + "step": 10315 + }, + { + "epoch": 1.8367165242165242, + "grad_norm": 0.6641396284103394, + "learning_rate": 0.00011296703508527363, + "loss": 1.0161, + "step": 10316 + }, + { + "epoch": 1.8368945868945867, + "grad_norm": 0.7921316027641296, + "learning_rate": 0.00011295315566454702, + "loss": 0.8897, + "step": 10317 + }, + { + "epoch": 1.8370726495726495, + "grad_norm": 0.6900694966316223, + "learning_rate": 0.00011293927599003029, + "loss": 1.0094, + "step": 10318 + }, + { + "epoch": 1.8372507122507122, + "grad_norm": 0.8054366707801819, + "learning_rate": 0.0001129253960619953, + "loss": 0.9489, + "step": 10319 + }, + { + "epoch": 1.837428774928775, + "grad_norm": 0.6623767018318176, + "learning_rate": 0.00011291151588071405, + "loss": 0.92, + "step": 10320 + }, + { + "epoch": 1.8376068376068377, + "grad_norm": 0.6143901348114014, + "learning_rate": 0.00011289763544645846, + "loss": 0.8093, + "step": 10321 + }, + { + "epoch": 1.8377849002849003, + "grad_norm": 0.8207027316093445, + "learning_rate": 0.00011288375475950046, + "loss": 1.2402, + "step": 10322 + }, + { + "epoch": 1.8379629629629628, + "grad_norm": 0.6759985685348511, + "learning_rate": 0.00011286987382011209, + "loss": 0.9179, + "step": 10323 + }, + { + "epoch": 1.8381410256410255, + "grad_norm": 0.745439887046814, + "learning_rate": 0.00011285599262856523, + "loss": 0.8157, + "step": 10324 + }, + { + "epoch": 1.8383190883190883, + "grad_norm": 0.6873317360877991, + "learning_rate": 0.00011284211118513194, + "loss": 0.8681, + "step": 10325 + }, + { + "epoch": 1.838497150997151, + "grad_norm": 0.7060160040855408, + "learning_rate": 0.00011282822949008416, + "loss": 1.0833, + "step": 10326 + }, + { + "epoch": 1.8386752136752138, + "grad_norm": 0.8079642653465271, + "learning_rate": 0.00011281434754369389, + "loss": 0.8639, + "step": 10327 + }, + { + "epoch": 1.8388532763532763, + "grad_norm": 0.6434001922607422, + "learning_rate": 0.00011280046534623303, + "loss": 0.9269, + "step": 10328 + }, + { + "epoch": 1.839031339031339, + "grad_norm": 0.7005292773246765, + "learning_rate": 0.0001127865828979737, + "loss": 1.1475, + "step": 10329 + }, + { + "epoch": 1.8392094017094016, + "grad_norm": 0.7004852890968323, + "learning_rate": 0.00011277270019918784, + "loss": 0.9467, + "step": 10330 + }, + { + "epoch": 1.8393874643874644, + "grad_norm": 0.7542549967765808, + "learning_rate": 0.00011275881725014743, + "loss": 1.0371, + "step": 10331 + }, + { + "epoch": 1.8395655270655271, + "grad_norm": 0.674051821231842, + "learning_rate": 0.00011274493405112452, + "loss": 1.1097, + "step": 10332 + }, + { + "epoch": 1.8397435897435899, + "grad_norm": 0.8136405348777771, + "learning_rate": 0.00011273105060239107, + "loss": 0.9718, + "step": 10333 + }, + { + "epoch": 1.8399216524216524, + "grad_norm": 0.6524073481559753, + "learning_rate": 0.00011271716690421916, + "loss": 0.9953, + "step": 10334 + }, + { + "epoch": 1.8400997150997151, + "grad_norm": 0.7436625957489014, + "learning_rate": 0.00011270328295688077, + "loss": 1.0722, + "step": 10335 + }, + { + "epoch": 1.8402777777777777, + "grad_norm": 0.6815723180770874, + "learning_rate": 0.00011268939876064795, + "loss": 1.0924, + "step": 10336 + }, + { + "epoch": 1.8404558404558404, + "grad_norm": 0.6923388242721558, + "learning_rate": 0.0001126755143157927, + "loss": 0.921, + "step": 10337 + }, + { + "epoch": 1.8406339031339032, + "grad_norm": 0.7464849948883057, + "learning_rate": 0.00011266162962258708, + "loss": 1.0549, + "step": 10338 + }, + { + "epoch": 1.840811965811966, + "grad_norm": 0.6621805429458618, + "learning_rate": 0.00011264774468130315, + "loss": 1.0764, + "step": 10339 + }, + { + "epoch": 1.8409900284900285, + "grad_norm": 0.7370132803916931, + "learning_rate": 0.00011263385949221295, + "loss": 0.7818, + "step": 10340 + }, + { + "epoch": 1.8411680911680912, + "grad_norm": 0.673100471496582, + "learning_rate": 0.00011261997405558848, + "loss": 1.04, + "step": 10341 + }, + { + "epoch": 1.8413461538461537, + "grad_norm": 0.5978201031684875, + "learning_rate": 0.00011260608837170183, + "loss": 0.9644, + "step": 10342 + }, + { + "epoch": 1.8415242165242165, + "grad_norm": 0.6868628263473511, + "learning_rate": 0.00011259220244082507, + "loss": 0.9533, + "step": 10343 + }, + { + "epoch": 1.8417022792022792, + "grad_norm": 0.6580314636230469, + "learning_rate": 0.0001125783162632303, + "loss": 0.9506, + "step": 10344 + }, + { + "epoch": 1.841880341880342, + "grad_norm": 0.7238291501998901, + "learning_rate": 0.00011256442983918951, + "loss": 0.8663, + "step": 10345 + }, + { + "epoch": 1.8420584045584045, + "grad_norm": 0.5838520526885986, + "learning_rate": 0.00011255054316897484, + "loss": 0.9606, + "step": 10346 + }, + { + "epoch": 1.8422364672364673, + "grad_norm": 0.7102842926979065, + "learning_rate": 0.00011253665625285836, + "loss": 0.801, + "step": 10347 + }, + { + "epoch": 1.8424145299145298, + "grad_norm": 0.6449147462844849, + "learning_rate": 0.0001125227690911121, + "loss": 1.0827, + "step": 10348 + }, + { + "epoch": 1.8425925925925926, + "grad_norm": 0.6355304718017578, + "learning_rate": 0.00011250888168400823, + "loss": 1.0369, + "step": 10349 + }, + { + "epoch": 1.8427706552706553, + "grad_norm": 0.678977906703949, + "learning_rate": 0.0001124949940318188, + "loss": 0.9491, + "step": 10350 + }, + { + "epoch": 1.842948717948718, + "grad_norm": 0.6366633772850037, + "learning_rate": 0.00011248110613481592, + "loss": 0.7272, + "step": 10351 + }, + { + "epoch": 1.8431267806267806, + "grad_norm": 0.6639098525047302, + "learning_rate": 0.00011246721799327171, + "loss": 1.0313, + "step": 10352 + }, + { + "epoch": 1.8433048433048433, + "grad_norm": 0.6034720540046692, + "learning_rate": 0.00011245332960745822, + "loss": 0.7141, + "step": 10353 + }, + { + "epoch": 1.8434829059829059, + "grad_norm": 0.8118346333503723, + "learning_rate": 0.00011243944097764763, + "loss": 1.171, + "step": 10354 + }, + { + "epoch": 1.8436609686609686, + "grad_norm": 0.6706618070602417, + "learning_rate": 0.00011242555210411203, + "loss": 0.9578, + "step": 10355 + }, + { + "epoch": 1.8438390313390314, + "grad_norm": 0.619562029838562, + "learning_rate": 0.00011241166298712355, + "loss": 0.9883, + "step": 10356 + }, + { + "epoch": 1.8440170940170941, + "grad_norm": 0.6471936106681824, + "learning_rate": 0.00011239777362695434, + "loss": 0.8897, + "step": 10357 + }, + { + "epoch": 1.8441951566951567, + "grad_norm": 0.7179005742073059, + "learning_rate": 0.00011238388402387645, + "loss": 0.9646, + "step": 10358 + }, + { + "epoch": 1.8443732193732194, + "grad_norm": 0.7726966738700867, + "learning_rate": 0.00011236999417816214, + "loss": 0.8855, + "step": 10359 + }, + { + "epoch": 1.844551282051282, + "grad_norm": 0.6733565330505371, + "learning_rate": 0.00011235610409008346, + "loss": 1.0379, + "step": 10360 + }, + { + "epoch": 1.8447293447293447, + "grad_norm": 0.7317814826965332, + "learning_rate": 0.0001123422137599126, + "loss": 0.8528, + "step": 10361 + }, + { + "epoch": 1.8449074074074074, + "grad_norm": 0.6727005839347839, + "learning_rate": 0.0001123283231879217, + "loss": 0.9612, + "step": 10362 + }, + { + "epoch": 1.8450854700854702, + "grad_norm": 0.6350542306900024, + "learning_rate": 0.00011231443237438289, + "loss": 0.9939, + "step": 10363 + }, + { + "epoch": 1.8452635327635327, + "grad_norm": 0.693148672580719, + "learning_rate": 0.00011230054131956836, + "loss": 1.0149, + "step": 10364 + }, + { + "epoch": 1.8454415954415955, + "grad_norm": 0.7263579368591309, + "learning_rate": 0.0001122866500237503, + "loss": 1.1044, + "step": 10365 + }, + { + "epoch": 1.845619658119658, + "grad_norm": 0.7044230699539185, + "learning_rate": 0.00011227275848720085, + "loss": 1.0677, + "step": 10366 + }, + { + "epoch": 1.8457977207977208, + "grad_norm": 0.6895326972007751, + "learning_rate": 0.00011225886671019219, + "loss": 1.1025, + "step": 10367 + }, + { + "epoch": 1.8459757834757835, + "grad_norm": 0.6045145988464355, + "learning_rate": 0.00011224497469299651, + "loss": 0.8079, + "step": 10368 + }, + { + "epoch": 1.8461538461538463, + "grad_norm": 0.6613210439682007, + "learning_rate": 0.00011223108243588599, + "loss": 1.0345, + "step": 10369 + }, + { + "epoch": 1.8463319088319088, + "grad_norm": 0.6288960576057434, + "learning_rate": 0.0001122171899391328, + "loss": 1.0166, + "step": 10370 + }, + { + "epoch": 1.8465099715099715, + "grad_norm": 0.6158748865127563, + "learning_rate": 0.00011220329720300917, + "loss": 0.895, + "step": 10371 + }, + { + "epoch": 1.846688034188034, + "grad_norm": 0.6583057641983032, + "learning_rate": 0.00011218940422778728, + "loss": 0.8059, + "step": 10372 + }, + { + "epoch": 1.8468660968660968, + "grad_norm": 0.6761550903320312, + "learning_rate": 0.00011217551101373932, + "loss": 0.9253, + "step": 10373 + }, + { + "epoch": 1.8470441595441596, + "grad_norm": 0.5969263315200806, + "learning_rate": 0.0001121616175611375, + "loss": 0.8549, + "step": 10374 + }, + { + "epoch": 1.8472222222222223, + "grad_norm": 0.7994722723960876, + "learning_rate": 0.00011214772387025407, + "loss": 0.9918, + "step": 10375 + }, + { + "epoch": 1.8474002849002849, + "grad_norm": 0.6949167847633362, + "learning_rate": 0.00011213382994136123, + "loss": 1.1853, + "step": 10376 + }, + { + "epoch": 1.8475783475783476, + "grad_norm": 0.7356176376342773, + "learning_rate": 0.00011211993577473121, + "loss": 0.8809, + "step": 10377 + }, + { + "epoch": 1.8477564102564101, + "grad_norm": 0.7110268473625183, + "learning_rate": 0.0001121060413706362, + "loss": 0.9805, + "step": 10378 + }, + { + "epoch": 1.8479344729344729, + "grad_norm": 0.6509962677955627, + "learning_rate": 0.00011209214672934846, + "loss": 0.8899, + "step": 10379 + }, + { + "epoch": 1.8481125356125356, + "grad_norm": 0.6103082299232483, + "learning_rate": 0.00011207825185114025, + "loss": 0.8576, + "step": 10380 + }, + { + "epoch": 1.8482905982905984, + "grad_norm": 0.6261070966720581, + "learning_rate": 0.00011206435673628377, + "loss": 0.8884, + "step": 10381 + }, + { + "epoch": 1.848468660968661, + "grad_norm": 0.7629222273826599, + "learning_rate": 0.00011205046138505126, + "loss": 1.1714, + "step": 10382 + }, + { + "epoch": 1.8486467236467237, + "grad_norm": 0.617957353591919, + "learning_rate": 0.000112036565797715, + "loss": 0.9546, + "step": 10383 + }, + { + "epoch": 1.8488247863247862, + "grad_norm": 0.6926987171173096, + "learning_rate": 0.00011202266997454724, + "loss": 0.8842, + "step": 10384 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 0.602758526802063, + "learning_rate": 0.00011200877391582025, + "loss": 0.9782, + "step": 10385 + }, + { + "epoch": 1.8491809116809117, + "grad_norm": 0.706731915473938, + "learning_rate": 0.00011199487762180627, + "loss": 0.8176, + "step": 10386 + }, + { + "epoch": 1.8493589743589745, + "grad_norm": 0.7135118842124939, + "learning_rate": 0.0001119809810927776, + "loss": 0.9277, + "step": 10387 + }, + { + "epoch": 1.8495370370370372, + "grad_norm": 0.7484592199325562, + "learning_rate": 0.00011196708432900647, + "loss": 1.0733, + "step": 10388 + }, + { + "epoch": 1.8497150997150997, + "grad_norm": 0.7087157964706421, + "learning_rate": 0.00011195318733076519, + "loss": 0.9443, + "step": 10389 + }, + { + "epoch": 1.8498931623931623, + "grad_norm": 0.6511468291282654, + "learning_rate": 0.00011193929009832602, + "loss": 0.955, + "step": 10390 + }, + { + "epoch": 1.850071225071225, + "grad_norm": 0.6386628746986389, + "learning_rate": 0.0001119253926319613, + "loss": 1.0357, + "step": 10391 + }, + { + "epoch": 1.8502492877492878, + "grad_norm": 0.6400021314620972, + "learning_rate": 0.00011191149493194327, + "loss": 0.8094, + "step": 10392 + }, + { + "epoch": 1.8504273504273505, + "grad_norm": 0.7942537069320679, + "learning_rate": 0.00011189759699854423, + "loss": 0.9717, + "step": 10393 + }, + { + "epoch": 1.8506054131054133, + "grad_norm": 0.7230474948883057, + "learning_rate": 0.00011188369883203647, + "loss": 0.9043, + "step": 10394 + }, + { + "epoch": 1.8507834757834758, + "grad_norm": 0.8837162852287292, + "learning_rate": 0.00011186980043269235, + "loss": 1.2821, + "step": 10395 + }, + { + "epoch": 1.8509615384615383, + "grad_norm": 0.7260291576385498, + "learning_rate": 0.00011185590180078413, + "loss": 1.1672, + "step": 10396 + }, + { + "epoch": 1.851139601139601, + "grad_norm": 0.6290066242218018, + "learning_rate": 0.00011184200293658415, + "loss": 0.8942, + "step": 10397 + }, + { + "epoch": 1.8513176638176638, + "grad_norm": 0.6571013331413269, + "learning_rate": 0.00011182810384036475, + "loss": 1.0753, + "step": 10398 + }, + { + "epoch": 1.8514957264957266, + "grad_norm": 0.6494737267494202, + "learning_rate": 0.00011181420451239817, + "loss": 0.8833, + "step": 10399 + }, + { + "epoch": 1.8516737891737893, + "grad_norm": 0.7383694648742676, + "learning_rate": 0.00011180030495295684, + "loss": 1.0094, + "step": 10400 + }, + { + "epoch": 1.8518518518518519, + "grad_norm": 0.6713876724243164, + "learning_rate": 0.00011178640516231302, + "loss": 0.975, + "step": 10401 + }, + { + "epoch": 1.8520299145299144, + "grad_norm": 0.8041042685508728, + "learning_rate": 0.00011177250514073912, + "loss": 1.1419, + "step": 10402 + }, + { + "epoch": 1.8522079772079771, + "grad_norm": 0.7035061120986938, + "learning_rate": 0.00011175860488850738, + "loss": 1.0921, + "step": 10403 + }, + { + "epoch": 1.85238603988604, + "grad_norm": 0.6135673522949219, + "learning_rate": 0.00011174470440589022, + "loss": 0.9611, + "step": 10404 + }, + { + "epoch": 1.8525641025641026, + "grad_norm": 0.7868386507034302, + "learning_rate": 0.00011173080369315999, + "loss": 0.8561, + "step": 10405 + }, + { + "epoch": 1.8527421652421654, + "grad_norm": 0.6575735211372375, + "learning_rate": 0.00011171690275058902, + "loss": 1.0256, + "step": 10406 + }, + { + "epoch": 1.852920227920228, + "grad_norm": 0.7514392137527466, + "learning_rate": 0.00011170300157844969, + "loss": 1.0868, + "step": 10407 + }, + { + "epoch": 1.8530982905982905, + "grad_norm": 0.6915257573127747, + "learning_rate": 0.00011168910017701436, + "loss": 1.1223, + "step": 10408 + }, + { + "epoch": 1.8532763532763532, + "grad_norm": 0.7406772971153259, + "learning_rate": 0.00011167519854655535, + "loss": 1.0922, + "step": 10409 + }, + { + "epoch": 1.853454415954416, + "grad_norm": 0.6632742881774902, + "learning_rate": 0.0001116612966873451, + "loss": 0.9082, + "step": 10410 + }, + { + "epoch": 1.8536324786324787, + "grad_norm": 0.8154461979866028, + "learning_rate": 0.00011164739459965598, + "loss": 1.1126, + "step": 10411 + }, + { + "epoch": 1.8538105413105415, + "grad_norm": 0.895764172077179, + "learning_rate": 0.00011163349228376037, + "loss": 1.0589, + "step": 10412 + }, + { + "epoch": 1.853988603988604, + "grad_norm": 0.6746504902839661, + "learning_rate": 0.00011161958973993063, + "loss": 1.0184, + "step": 10413 + }, + { + "epoch": 1.8541666666666665, + "grad_norm": 0.7271263003349304, + "learning_rate": 0.00011160568696843916, + "loss": 0.9989, + "step": 10414 + }, + { + "epoch": 1.8543447293447293, + "grad_norm": 0.7503132820129395, + "learning_rate": 0.00011159178396955836, + "loss": 1.0783, + "step": 10415 + }, + { + "epoch": 1.854522792022792, + "grad_norm": 0.6768177151679993, + "learning_rate": 0.00011157788074356066, + "loss": 0.9916, + "step": 10416 + }, + { + "epoch": 1.8547008547008548, + "grad_norm": 0.6804978251457214, + "learning_rate": 0.00011156397729071842, + "loss": 0.9534, + "step": 10417 + }, + { + "epoch": 1.8548789173789175, + "grad_norm": 0.7144617438316345, + "learning_rate": 0.00011155007361130408, + "loss": 0.991, + "step": 10418 + }, + { + "epoch": 1.85505698005698, + "grad_norm": 0.6816750168800354, + "learning_rate": 0.00011153616970559, + "loss": 0.9551, + "step": 10419 + }, + { + "epoch": 1.8552350427350426, + "grad_norm": 0.6620030999183655, + "learning_rate": 0.00011152226557384866, + "loss": 0.8854, + "step": 10420 + }, + { + "epoch": 1.8554131054131053, + "grad_norm": 0.8400058746337891, + "learning_rate": 0.00011150836121635249, + "loss": 1.1593, + "step": 10421 + }, + { + "epoch": 1.855591168091168, + "grad_norm": 0.6666815280914307, + "learning_rate": 0.00011149445663337385, + "loss": 1.2112, + "step": 10422 + }, + { + "epoch": 1.8557692307692308, + "grad_norm": 0.7298431396484375, + "learning_rate": 0.00011148055182518522, + "loss": 0.9721, + "step": 10423 + }, + { + "epoch": 1.8559472934472936, + "grad_norm": 0.66816645860672, + "learning_rate": 0.00011146664679205903, + "loss": 1.0945, + "step": 10424 + }, + { + "epoch": 1.8561253561253561, + "grad_norm": 0.5979483127593994, + "learning_rate": 0.00011145274153426771, + "loss": 1.0176, + "step": 10425 + }, + { + "epoch": 1.8563034188034186, + "grad_norm": 0.6579445600509644, + "learning_rate": 0.00011143883605208372, + "loss": 0.9143, + "step": 10426 + }, + { + "epoch": 1.8564814814814814, + "grad_norm": 0.6871697902679443, + "learning_rate": 0.0001114249303457795, + "loss": 1.071, + "step": 10427 + }, + { + "epoch": 1.8566595441595442, + "grad_norm": 0.6683333516120911, + "learning_rate": 0.0001114110244156275, + "loss": 0.7809, + "step": 10428 + }, + { + "epoch": 1.856837606837607, + "grad_norm": 0.6122907996177673, + "learning_rate": 0.0001113971182619002, + "loss": 0.8329, + "step": 10429 + }, + { + "epoch": 1.8570156695156697, + "grad_norm": 0.6510575413703918, + "learning_rate": 0.00011138321188487, + "loss": 1.0068, + "step": 10430 + }, + { + "epoch": 1.8571937321937322, + "grad_norm": 0.6417793035507202, + "learning_rate": 0.00011136930528480945, + "loss": 1.0093, + "step": 10431 + }, + { + "epoch": 1.8573717948717947, + "grad_norm": 0.595824658870697, + "learning_rate": 0.00011135539846199096, + "loss": 0.9856, + "step": 10432 + }, + { + "epoch": 1.8575498575498575, + "grad_norm": 0.7594470381736755, + "learning_rate": 0.00011134149141668704, + "loss": 0.8173, + "step": 10433 + }, + { + "epoch": 1.8577279202279202, + "grad_norm": 0.7078324556350708, + "learning_rate": 0.00011132758414917016, + "loss": 1.0236, + "step": 10434 + }, + { + "epoch": 1.857905982905983, + "grad_norm": 0.6830437779426575, + "learning_rate": 0.00011131367665971275, + "loss": 0.8483, + "step": 10435 + }, + { + "epoch": 1.8580840455840457, + "grad_norm": 0.6856399774551392, + "learning_rate": 0.0001112997689485874, + "loss": 0.8729, + "step": 10436 + }, + { + "epoch": 1.8582621082621082, + "grad_norm": 0.6530426144599915, + "learning_rate": 0.00011128586101606653, + "loss": 0.8616, + "step": 10437 + }, + { + "epoch": 1.8584401709401708, + "grad_norm": 0.6341808438301086, + "learning_rate": 0.00011127195286242267, + "loss": 0.896, + "step": 10438 + }, + { + "epoch": 1.8586182336182335, + "grad_norm": 0.6278257966041565, + "learning_rate": 0.00011125804448792831, + "loss": 0.8309, + "step": 10439 + }, + { + "epoch": 1.8587962962962963, + "grad_norm": 0.708705723285675, + "learning_rate": 0.00011124413589285594, + "loss": 1.1065, + "step": 10440 + }, + { + "epoch": 1.858974358974359, + "grad_norm": 0.6845232248306274, + "learning_rate": 0.00011123022707747808, + "loss": 0.9292, + "step": 10441 + }, + { + "epoch": 1.8591524216524218, + "grad_norm": 0.749204695224762, + "learning_rate": 0.00011121631804206726, + "loss": 1.0487, + "step": 10442 + }, + { + "epoch": 1.8593304843304843, + "grad_norm": 0.7123128771781921, + "learning_rate": 0.00011120240878689599, + "loss": 0.9138, + "step": 10443 + }, + { + "epoch": 1.859508547008547, + "grad_norm": 0.6862115263938904, + "learning_rate": 0.00011118849931223679, + "loss": 1.0675, + "step": 10444 + }, + { + "epoch": 1.8596866096866096, + "grad_norm": 0.7245760560035706, + "learning_rate": 0.00011117458961836215, + "loss": 0.9643, + "step": 10445 + }, + { + "epoch": 1.8598646723646723, + "grad_norm": 0.701574444770813, + "learning_rate": 0.0001111606797055447, + "loss": 1.0022, + "step": 10446 + }, + { + "epoch": 1.860042735042735, + "grad_norm": 0.7292088270187378, + "learning_rate": 0.0001111467695740569, + "loss": 0.9465, + "step": 10447 + }, + { + "epoch": 1.8602207977207978, + "grad_norm": 0.7045044302940369, + "learning_rate": 0.0001111328592241713, + "loss": 1.0942, + "step": 10448 + }, + { + "epoch": 1.8603988603988604, + "grad_norm": 0.7181426286697388, + "learning_rate": 0.00011111894865616046, + "loss": 1.2108, + "step": 10449 + }, + { + "epoch": 1.8605769230769231, + "grad_norm": 0.6083306074142456, + "learning_rate": 0.00011110503787029689, + "loss": 0.929, + "step": 10450 + }, + { + "epoch": 1.8607549857549857, + "grad_norm": 0.6847347617149353, + "learning_rate": 0.00011109112686685319, + "loss": 1.0911, + "step": 10451 + }, + { + "epoch": 1.8609330484330484, + "grad_norm": 0.7131744027137756, + "learning_rate": 0.0001110772156461019, + "loss": 0.9649, + "step": 10452 + }, + { + "epoch": 1.8611111111111112, + "grad_norm": 0.7920312881469727, + "learning_rate": 0.00011106330420831559, + "loss": 0.9965, + "step": 10453 + }, + { + "epoch": 1.861289173789174, + "grad_norm": 0.6640987992286682, + "learning_rate": 0.00011104939255376681, + "loss": 1.2346, + "step": 10454 + }, + { + "epoch": 1.8614672364672364, + "grad_norm": 0.5878208875656128, + "learning_rate": 0.00011103548068272811, + "loss": 0.8565, + "step": 10455 + }, + { + "epoch": 1.8616452991452992, + "grad_norm": 0.6636882424354553, + "learning_rate": 0.0001110215685954721, + "loss": 0.8556, + "step": 10456 + }, + { + "epoch": 1.8618233618233617, + "grad_norm": 0.5985570549964905, + "learning_rate": 0.00011100765629227137, + "loss": 1.0291, + "step": 10457 + }, + { + "epoch": 1.8620014245014245, + "grad_norm": 0.7546643614768982, + "learning_rate": 0.00011099374377339846, + "loss": 1.0199, + "step": 10458 + }, + { + "epoch": 1.8621794871794872, + "grad_norm": 0.6529727578163147, + "learning_rate": 0.00011097983103912602, + "loss": 1.0826, + "step": 10459 + }, + { + "epoch": 1.86235754985755, + "grad_norm": 0.6394338607788086, + "learning_rate": 0.00011096591808972654, + "loss": 0.9896, + "step": 10460 + }, + { + "epoch": 1.8625356125356125, + "grad_norm": 0.6508805751800537, + "learning_rate": 0.00011095200492547271, + "loss": 0.9659, + "step": 10461 + }, + { + "epoch": 1.8627136752136753, + "grad_norm": 0.7085812091827393, + "learning_rate": 0.00011093809154663705, + "loss": 0.9998, + "step": 10462 + }, + { + "epoch": 1.8628917378917378, + "grad_norm": 0.6488457322120667, + "learning_rate": 0.00011092417795349226, + "loss": 0.9757, + "step": 10463 + }, + { + "epoch": 1.8630698005698005, + "grad_norm": 0.6405763626098633, + "learning_rate": 0.0001109102641463109, + "loss": 0.8188, + "step": 10464 + }, + { + "epoch": 1.8632478632478633, + "grad_norm": 0.713361918926239, + "learning_rate": 0.00011089635012536554, + "loss": 0.886, + "step": 10465 + }, + { + "epoch": 1.863425925925926, + "grad_norm": 0.5752255916595459, + "learning_rate": 0.00011088243589092886, + "loss": 1.0223, + "step": 10466 + }, + { + "epoch": 1.8636039886039886, + "grad_norm": 0.6722734570503235, + "learning_rate": 0.00011086852144327344, + "loss": 0.9499, + "step": 10467 + }, + { + "epoch": 1.8637820512820513, + "grad_norm": 0.5516420006752014, + "learning_rate": 0.00011085460678267194, + "loss": 0.7767, + "step": 10468 + }, + { + "epoch": 1.8639601139601139, + "grad_norm": 0.731257438659668, + "learning_rate": 0.00011084069190939697, + "loss": 1.2299, + "step": 10469 + }, + { + "epoch": 1.8641381766381766, + "grad_norm": 0.7977055907249451, + "learning_rate": 0.00011082677682372114, + "loss": 0.9109, + "step": 10470 + }, + { + "epoch": 1.8643162393162394, + "grad_norm": 0.679900586605072, + "learning_rate": 0.0001108128615259171, + "loss": 0.9319, + "step": 10471 + }, + { + "epoch": 1.864494301994302, + "grad_norm": 0.7428545951843262, + "learning_rate": 0.00011079894601625754, + "loss": 0.8585, + "step": 10472 + }, + { + "epoch": 1.8646723646723646, + "grad_norm": 0.6560967564582825, + "learning_rate": 0.00011078503029501504, + "loss": 1.0069, + "step": 10473 + }, + { + "epoch": 1.8648504273504274, + "grad_norm": 0.636202871799469, + "learning_rate": 0.00011077111436246228, + "loss": 1.0329, + "step": 10474 + }, + { + "epoch": 1.86502849002849, + "grad_norm": 0.6666205525398254, + "learning_rate": 0.00011075719821887191, + "loss": 1.0123, + "step": 10475 + }, + { + "epoch": 1.8652065527065527, + "grad_norm": 0.7089471220970154, + "learning_rate": 0.00011074328186451657, + "loss": 0.7851, + "step": 10476 + }, + { + "epoch": 1.8653846153846154, + "grad_norm": 0.6054788827896118, + "learning_rate": 0.00011072936529966895, + "loss": 0.8224, + "step": 10477 + }, + { + "epoch": 1.8655626780626782, + "grad_norm": 0.6009029150009155, + "learning_rate": 0.00011071544852460172, + "loss": 0.865, + "step": 10478 + }, + { + "epoch": 1.8657407407407407, + "grad_norm": 0.6238716244697571, + "learning_rate": 0.00011070153153958753, + "loss": 0.8685, + "step": 10479 + }, + { + "epoch": 1.8659188034188035, + "grad_norm": 0.719985842704773, + "learning_rate": 0.00011068761434489903, + "loss": 1.2204, + "step": 10480 + }, + { + "epoch": 1.866096866096866, + "grad_norm": 0.72972172498703, + "learning_rate": 0.00011067369694080895, + "loss": 1.0454, + "step": 10481 + }, + { + "epoch": 1.8662749287749287, + "grad_norm": 0.6741998791694641, + "learning_rate": 0.00011065977932758995, + "loss": 0.9992, + "step": 10482 + }, + { + "epoch": 1.8664529914529915, + "grad_norm": 0.6150268912315369, + "learning_rate": 0.00011064586150551472, + "loss": 0.8866, + "step": 10483 + }, + { + "epoch": 1.8666310541310542, + "grad_norm": 0.8253782391548157, + "learning_rate": 0.00011063194347485597, + "loss": 1.1173, + "step": 10484 + }, + { + "epoch": 1.8668091168091168, + "grad_norm": 0.7176247835159302, + "learning_rate": 0.00011061802523588636, + "loss": 1.0414, + "step": 10485 + }, + { + "epoch": 1.8669871794871795, + "grad_norm": 0.6372736096382141, + "learning_rate": 0.00011060410678887858, + "loss": 1.0548, + "step": 10486 + }, + { + "epoch": 1.867165242165242, + "grad_norm": 0.7107454538345337, + "learning_rate": 0.00011059018813410538, + "loss": 1.2298, + "step": 10487 + }, + { + "epoch": 1.8673433048433048, + "grad_norm": 0.7113911509513855, + "learning_rate": 0.00011057626927183944, + "loss": 0.9598, + "step": 10488 + }, + { + "epoch": 1.8675213675213675, + "grad_norm": 0.6734410524368286, + "learning_rate": 0.00011056235020235346, + "loss": 0.9475, + "step": 10489 + }, + { + "epoch": 1.8676994301994303, + "grad_norm": 0.6875202655792236, + "learning_rate": 0.0001105484309259202, + "loss": 1.0735, + "step": 10490 + }, + { + "epoch": 1.8678774928774928, + "grad_norm": 0.6908353567123413, + "learning_rate": 0.0001105345114428123, + "loss": 1.0558, + "step": 10491 + }, + { + "epoch": 1.8680555555555556, + "grad_norm": 0.6283324360847473, + "learning_rate": 0.00011052059175330256, + "loss": 0.8872, + "step": 10492 + }, + { + "epoch": 1.868233618233618, + "grad_norm": 0.6422587633132935, + "learning_rate": 0.00011050667185766368, + "loss": 1.1022, + "step": 10493 + }, + { + "epoch": 1.8684116809116809, + "grad_norm": 0.7075859904289246, + "learning_rate": 0.0001104927517561684, + "loss": 1.1389, + "step": 10494 + }, + { + "epoch": 1.8685897435897436, + "grad_norm": 0.5896905064582825, + "learning_rate": 0.00011047883144908944, + "loss": 0.7732, + "step": 10495 + }, + { + "epoch": 1.8687678062678064, + "grad_norm": 0.7647629976272583, + "learning_rate": 0.00011046491093669953, + "loss": 0.9983, + "step": 10496 + }, + { + "epoch": 1.868945868945869, + "grad_norm": 0.5864735841751099, + "learning_rate": 0.00011045099021927144, + "loss": 0.8427, + "step": 10497 + }, + { + "epoch": 1.8691239316239316, + "grad_norm": 0.6766837239265442, + "learning_rate": 0.00011043706929707791, + "loss": 0.9595, + "step": 10498 + }, + { + "epoch": 1.8693019943019942, + "grad_norm": 0.5480074286460876, + "learning_rate": 0.00011042314817039168, + "loss": 0.691, + "step": 10499 + }, + { + "epoch": 1.869480056980057, + "grad_norm": 0.6259615421295166, + "learning_rate": 0.00011040922683948553, + "loss": 0.9991, + "step": 10500 + }, + { + "epoch": 1.8696581196581197, + "grad_norm": 0.5950598120689392, + "learning_rate": 0.00011039530530463218, + "loss": 0.7413, + "step": 10501 + }, + { + "epoch": 1.8698361823361824, + "grad_norm": 0.8099377751350403, + "learning_rate": 0.00011038138356610441, + "loss": 1.1351, + "step": 10502 + }, + { + "epoch": 1.8700142450142452, + "grad_norm": 0.6716185212135315, + "learning_rate": 0.00011036746162417501, + "loss": 1.1057, + "step": 10503 + }, + { + "epoch": 1.8701923076923077, + "grad_norm": 0.7993219494819641, + "learning_rate": 0.00011035353947911675, + "loss": 1.2095, + "step": 10504 + }, + { + "epoch": 1.8703703703703702, + "grad_norm": 0.6381276249885559, + "learning_rate": 0.00011033961713120237, + "loss": 1.0261, + "step": 10505 + }, + { + "epoch": 1.870548433048433, + "grad_norm": 0.6326032280921936, + "learning_rate": 0.00011032569458070469, + "loss": 0.8664, + "step": 10506 + }, + { + "epoch": 1.8707264957264957, + "grad_norm": 0.6864820718765259, + "learning_rate": 0.00011031177182789644, + "loss": 0.9959, + "step": 10507 + }, + { + "epoch": 1.8709045584045585, + "grad_norm": 0.6341838240623474, + "learning_rate": 0.00011029784887305048, + "loss": 0.8029, + "step": 10508 + }, + { + "epoch": 1.8710826210826212, + "grad_norm": 0.6559172868728638, + "learning_rate": 0.00011028392571643957, + "loss": 0.9282, + "step": 10509 + }, + { + "epoch": 1.8712606837606838, + "grad_norm": 0.6976849436759949, + "learning_rate": 0.0001102700023583365, + "loss": 1.0198, + "step": 10510 + }, + { + "epoch": 1.8714387464387463, + "grad_norm": 0.7159395217895508, + "learning_rate": 0.00011025607879901402, + "loss": 1.1585, + "step": 10511 + }, + { + "epoch": 1.871616809116809, + "grad_norm": 0.7168624997138977, + "learning_rate": 0.000110242155038745, + "loss": 1.0558, + "step": 10512 + }, + { + "epoch": 1.8717948717948718, + "grad_norm": 0.5784319043159485, + "learning_rate": 0.00011022823107780224, + "loss": 0.9481, + "step": 10513 + }, + { + "epoch": 1.8719729344729346, + "grad_norm": 0.6602259874343872, + "learning_rate": 0.00011021430691645856, + "loss": 1.0538, + "step": 10514 + }, + { + "epoch": 1.8721509971509973, + "grad_norm": 0.6874588131904602, + "learning_rate": 0.00011020038255498672, + "loss": 1.1396, + "step": 10515 + }, + { + "epoch": 1.8723290598290598, + "grad_norm": 0.7311663031578064, + "learning_rate": 0.00011018645799365956, + "loss": 1.084, + "step": 10516 + }, + { + "epoch": 1.8725071225071224, + "grad_norm": 0.7097118496894836, + "learning_rate": 0.00011017253323274996, + "loss": 0.9872, + "step": 10517 + }, + { + "epoch": 1.8726851851851851, + "grad_norm": 0.6667875051498413, + "learning_rate": 0.00011015860827253068, + "loss": 1.105, + "step": 10518 + }, + { + "epoch": 1.8728632478632479, + "grad_norm": 0.6807677745819092, + "learning_rate": 0.0001101446831132746, + "loss": 0.9093, + "step": 10519 + }, + { + "epoch": 1.8730413105413106, + "grad_norm": 0.6885797381401062, + "learning_rate": 0.0001101307577552545, + "loss": 0.8479, + "step": 10520 + }, + { + "epoch": 1.8732193732193734, + "grad_norm": 0.6269213557243347, + "learning_rate": 0.00011011683219874323, + "loss": 0.9457, + "step": 10521 + }, + { + "epoch": 1.873397435897436, + "grad_norm": 0.7096766829490662, + "learning_rate": 0.00011010290644401364, + "loss": 1.0971, + "step": 10522 + }, + { + "epoch": 1.8735754985754984, + "grad_norm": 0.6909209489822388, + "learning_rate": 0.00011008898049133863, + "loss": 0.9928, + "step": 10523 + }, + { + "epoch": 1.8737535612535612, + "grad_norm": 0.6586211323738098, + "learning_rate": 0.000110075054340991, + "loss": 0.818, + "step": 10524 + }, + { + "epoch": 1.873931623931624, + "grad_norm": 0.5934817790985107, + "learning_rate": 0.0001100611279932436, + "loss": 0.7698, + "step": 10525 + }, + { + "epoch": 1.8741096866096867, + "grad_norm": 0.6361709237098694, + "learning_rate": 0.00011004720144836931, + "loss": 0.9465, + "step": 10526 + }, + { + "epoch": 1.8742877492877494, + "grad_norm": 0.6742212176322937, + "learning_rate": 0.00011003327470664095, + "loss": 1.0998, + "step": 10527 + }, + { + "epoch": 1.874465811965812, + "grad_norm": 0.6634946465492249, + "learning_rate": 0.00011001934776833143, + "loss": 0.8328, + "step": 10528 + }, + { + "epoch": 1.8746438746438745, + "grad_norm": 0.6754063963890076, + "learning_rate": 0.0001100054206337136, + "loss": 1.147, + "step": 10529 + }, + { + "epoch": 1.8748219373219372, + "grad_norm": 0.5951135158538818, + "learning_rate": 0.00010999149330306036, + "loss": 0.8956, + "step": 10530 + }, + { + "epoch": 1.875, + "grad_norm": 0.6140317320823669, + "learning_rate": 0.00010997756577664455, + "loss": 0.9368, + "step": 10531 + }, + { + "epoch": 1.8751780626780628, + "grad_norm": 0.6419258713722229, + "learning_rate": 0.00010996363805473904, + "loss": 0.9817, + "step": 10532 + }, + { + "epoch": 1.8753561253561255, + "grad_norm": 0.7173396348953247, + "learning_rate": 0.00010994971013761677, + "loss": 0.9638, + "step": 10533 + }, + { + "epoch": 1.875534188034188, + "grad_norm": 0.8125925660133362, + "learning_rate": 0.0001099357820255506, + "loss": 1.0996, + "step": 10534 + }, + { + "epoch": 1.8757122507122506, + "grad_norm": 0.6191564798355103, + "learning_rate": 0.00010992185371881341, + "loss": 0.8266, + "step": 10535 + }, + { + "epoch": 1.8758903133903133, + "grad_norm": 0.6632885336875916, + "learning_rate": 0.0001099079252176781, + "loss": 1.1884, + "step": 10536 + }, + { + "epoch": 1.876068376068376, + "grad_norm": 0.7323372960090637, + "learning_rate": 0.00010989399652241759, + "loss": 1.0842, + "step": 10537 + }, + { + "epoch": 1.8762464387464388, + "grad_norm": 0.7553854584693909, + "learning_rate": 0.00010988006763330476, + "loss": 0.9948, + "step": 10538 + }, + { + "epoch": 1.8764245014245016, + "grad_norm": 0.5887658596038818, + "learning_rate": 0.00010986613855061255, + "loss": 0.7653, + "step": 10539 + }, + { + "epoch": 1.876602564102564, + "grad_norm": 0.6849574446678162, + "learning_rate": 0.00010985220927461384, + "loss": 1.152, + "step": 10540 + }, + { + "epoch": 1.8767806267806266, + "grad_norm": 0.6985000371932983, + "learning_rate": 0.00010983827980558155, + "loss": 0.9869, + "step": 10541 + }, + { + "epoch": 1.8769586894586894, + "grad_norm": 0.6885373592376709, + "learning_rate": 0.00010982435014378858, + "loss": 1.1803, + "step": 10542 + }, + { + "epoch": 1.8771367521367521, + "grad_norm": 0.7610142827033997, + "learning_rate": 0.00010981042028950788, + "loss": 0.9219, + "step": 10543 + }, + { + "epoch": 1.8773148148148149, + "grad_norm": 0.6545612215995789, + "learning_rate": 0.00010979649024301242, + "loss": 1.0337, + "step": 10544 + }, + { + "epoch": 1.8774928774928776, + "grad_norm": 0.7307698130607605, + "learning_rate": 0.00010978256000457505, + "loss": 0.9726, + "step": 10545 + }, + { + "epoch": 1.8776709401709402, + "grad_norm": 0.68310546875, + "learning_rate": 0.00010976862957446877, + "loss": 1.161, + "step": 10546 + }, + { + "epoch": 1.8778490028490027, + "grad_norm": 0.6114758253097534, + "learning_rate": 0.00010975469895296646, + "loss": 0.8863, + "step": 10547 + }, + { + "epoch": 1.8780270655270654, + "grad_norm": 0.732390820980072, + "learning_rate": 0.00010974076814034106, + "loss": 1.0339, + "step": 10548 + }, + { + "epoch": 1.8782051282051282, + "grad_norm": 0.6741712689399719, + "learning_rate": 0.0001097268371368656, + "loss": 1.0024, + "step": 10549 + }, + { + "epoch": 1.878383190883191, + "grad_norm": 0.6374897360801697, + "learning_rate": 0.00010971290594281294, + "loss": 0.91, + "step": 10550 + }, + { + "epoch": 1.8785612535612537, + "grad_norm": 0.6434261202812195, + "learning_rate": 0.00010969897455845608, + "loss": 1.0048, + "step": 10551 + }, + { + "epoch": 1.8787393162393162, + "grad_norm": 0.6573047041893005, + "learning_rate": 0.00010968504298406794, + "loss": 1.118, + "step": 10552 + }, + { + "epoch": 1.8789173789173788, + "grad_norm": 0.6686552166938782, + "learning_rate": 0.00010967111121992152, + "loss": 1.089, + "step": 10553 + }, + { + "epoch": 1.8790954415954415, + "grad_norm": 0.7899606823921204, + "learning_rate": 0.00010965717926628976, + "loss": 1.059, + "step": 10554 + }, + { + "epoch": 1.8792735042735043, + "grad_norm": 0.5808879733085632, + "learning_rate": 0.00010964324712344564, + "loss": 0.9369, + "step": 10555 + }, + { + "epoch": 1.879451566951567, + "grad_norm": 0.6322834491729736, + "learning_rate": 0.00010962931479166211, + "loss": 0.8783, + "step": 10556 + }, + { + "epoch": 1.8796296296296298, + "grad_norm": 0.647002637386322, + "learning_rate": 0.00010961538227121218, + "loss": 0.9468, + "step": 10557 + }, + { + "epoch": 1.8798076923076923, + "grad_norm": 0.6581854820251465, + "learning_rate": 0.0001096014495623688, + "loss": 1.0077, + "step": 10558 + }, + { + "epoch": 1.8799857549857548, + "grad_norm": 0.6879259943962097, + "learning_rate": 0.00010958751666540496, + "loss": 0.976, + "step": 10559 + }, + { + "epoch": 1.8801638176638176, + "grad_norm": 0.7055090665817261, + "learning_rate": 0.00010957358358059364, + "loss": 0.8903, + "step": 10560 + }, + { + "epoch": 1.8803418803418803, + "grad_norm": 0.6865016222000122, + "learning_rate": 0.00010955965030820782, + "loss": 0.9872, + "step": 10561 + }, + { + "epoch": 1.880519943019943, + "grad_norm": 0.663436770439148, + "learning_rate": 0.00010954571684852055, + "loss": 1.0485, + "step": 10562 + }, + { + "epoch": 1.8806980056980058, + "grad_norm": 0.6861656904220581, + "learning_rate": 0.00010953178320180475, + "loss": 1.0691, + "step": 10563 + }, + { + "epoch": 1.8808760683760684, + "grad_norm": 0.8045449256896973, + "learning_rate": 0.0001095178493683335, + "loss": 1.1534, + "step": 10564 + }, + { + "epoch": 1.881054131054131, + "grad_norm": 0.6493151187896729, + "learning_rate": 0.00010950391534837973, + "loss": 0.8756, + "step": 10565 + }, + { + "epoch": 1.8812321937321936, + "grad_norm": 0.7057121992111206, + "learning_rate": 0.00010948998114221651, + "loss": 1.1709, + "step": 10566 + }, + { + "epoch": 1.8814102564102564, + "grad_norm": 0.7708197236061096, + "learning_rate": 0.0001094760467501168, + "loss": 1.0037, + "step": 10567 + }, + { + "epoch": 1.8815883190883191, + "grad_norm": 0.7234642505645752, + "learning_rate": 0.00010946211217235364, + "loss": 1.0757, + "step": 10568 + }, + { + "epoch": 1.881766381766382, + "grad_norm": 0.6964395642280579, + "learning_rate": 0.00010944817740920006, + "loss": 1.0769, + "step": 10569 + }, + { + "epoch": 1.8819444444444444, + "grad_norm": 0.7465848922729492, + "learning_rate": 0.00010943424246092906, + "loss": 0.9772, + "step": 10570 + }, + { + "epoch": 1.8821225071225072, + "grad_norm": 0.7145788073539734, + "learning_rate": 0.0001094203073278137, + "loss": 0.9638, + "step": 10571 + }, + { + "epoch": 1.8823005698005697, + "grad_norm": 0.7421764135360718, + "learning_rate": 0.00010940637201012698, + "loss": 1.0324, + "step": 10572 + }, + { + "epoch": 1.8824786324786325, + "grad_norm": 0.7373253107070923, + "learning_rate": 0.0001093924365081419, + "loss": 1.1554, + "step": 10573 + }, + { + "epoch": 1.8826566951566952, + "grad_norm": 0.6861984729766846, + "learning_rate": 0.00010937850082213156, + "loss": 0.9899, + "step": 10574 + }, + { + "epoch": 1.882834757834758, + "grad_norm": 0.6173393130302429, + "learning_rate": 0.000109364564952369, + "loss": 0.8495, + "step": 10575 + }, + { + "epoch": 1.8830128205128205, + "grad_norm": 0.6871610879898071, + "learning_rate": 0.00010935062889912723, + "loss": 1.2164, + "step": 10576 + }, + { + "epoch": 1.8831908831908832, + "grad_norm": 0.7062903642654419, + "learning_rate": 0.00010933669266267931, + "loss": 1.1077, + "step": 10577 + }, + { + "epoch": 1.8833689458689458, + "grad_norm": 0.6574689745903015, + "learning_rate": 0.00010932275624329828, + "loss": 0.9326, + "step": 10578 + }, + { + "epoch": 1.8835470085470085, + "grad_norm": 0.636385440826416, + "learning_rate": 0.00010930881964125723, + "loss": 1.0581, + "step": 10579 + }, + { + "epoch": 1.8837250712250713, + "grad_norm": 0.6178432106971741, + "learning_rate": 0.0001092948828568292, + "loss": 1.1288, + "step": 10580 + }, + { + "epoch": 1.883903133903134, + "grad_norm": 0.6509431600570679, + "learning_rate": 0.00010928094589028721, + "loss": 1.0113, + "step": 10581 + }, + { + "epoch": 1.8840811965811965, + "grad_norm": 0.6543706059455872, + "learning_rate": 0.00010926700874190441, + "loss": 1.0041, + "step": 10582 + }, + { + "epoch": 1.8842592592592593, + "grad_norm": 0.6815463304519653, + "learning_rate": 0.0001092530714119538, + "loss": 1.0892, + "step": 10583 + }, + { + "epoch": 1.8844373219373218, + "grad_norm": 0.6787421107292175, + "learning_rate": 0.00010923913390070846, + "loss": 1.2693, + "step": 10584 + }, + { + "epoch": 1.8846153846153846, + "grad_norm": 0.6953850984573364, + "learning_rate": 0.00010922519620844151, + "loss": 0.9848, + "step": 10585 + }, + { + "epoch": 1.8847934472934473, + "grad_norm": 0.7061360478401184, + "learning_rate": 0.000109211258335426, + "loss": 0.949, + "step": 10586 + }, + { + "epoch": 1.88497150997151, + "grad_norm": 0.6845372915267944, + "learning_rate": 0.00010919732028193504, + "loss": 0.9554, + "step": 10587 + }, + { + "epoch": 1.8851495726495726, + "grad_norm": 0.6524720788002014, + "learning_rate": 0.00010918338204824165, + "loss": 1.1037, + "step": 10588 + }, + { + "epoch": 1.8853276353276354, + "grad_norm": 0.6410523653030396, + "learning_rate": 0.00010916944363461899, + "loss": 0.9085, + "step": 10589 + }, + { + "epoch": 1.885505698005698, + "grad_norm": 0.7109059691429138, + "learning_rate": 0.00010915550504134014, + "loss": 1.0526, + "step": 10590 + }, + { + "epoch": 1.8856837606837606, + "grad_norm": 0.7781991362571716, + "learning_rate": 0.00010914156626867818, + "loss": 0.9737, + "step": 10591 + }, + { + "epoch": 1.8858618233618234, + "grad_norm": 0.7173767685890198, + "learning_rate": 0.00010912762731690623, + "loss": 0.8862, + "step": 10592 + }, + { + "epoch": 1.8860398860398861, + "grad_norm": 0.7650504112243652, + "learning_rate": 0.00010911368818629732, + "loss": 1.2175, + "step": 10593 + }, + { + "epoch": 1.8862179487179487, + "grad_norm": 0.6316116452217102, + "learning_rate": 0.00010909974887712468, + "loss": 0.8332, + "step": 10594 + }, + { + "epoch": 1.8863960113960114, + "grad_norm": 0.6504800319671631, + "learning_rate": 0.00010908580938966138, + "loss": 0.8864, + "step": 10595 + }, + { + "epoch": 1.886574074074074, + "grad_norm": 0.675507128238678, + "learning_rate": 0.00010907186972418049, + "loss": 0.8523, + "step": 10596 + }, + { + "epoch": 1.8867521367521367, + "grad_norm": 0.6535763144493103, + "learning_rate": 0.00010905792988095515, + "loss": 1.0786, + "step": 10597 + }, + { + "epoch": 1.8869301994301995, + "grad_norm": 0.7071853280067444, + "learning_rate": 0.0001090439898602585, + "loss": 0.9319, + "step": 10598 + }, + { + "epoch": 1.8871082621082622, + "grad_norm": 0.699466347694397, + "learning_rate": 0.00010903004966236365, + "loss": 0.9573, + "step": 10599 + }, + { + "epoch": 1.8872863247863247, + "grad_norm": 0.7099201083183289, + "learning_rate": 0.00010901610928754375, + "loss": 0.9447, + "step": 10600 + }, + { + "epoch": 1.8874643874643875, + "grad_norm": 0.6140450835227966, + "learning_rate": 0.00010900216873607189, + "loss": 1.0227, + "step": 10601 + }, + { + "epoch": 1.88764245014245, + "grad_norm": 0.6613629460334778, + "learning_rate": 0.00010898822800822127, + "loss": 1.0152, + "step": 10602 + }, + { + "epoch": 1.8878205128205128, + "grad_norm": 0.7334819436073303, + "learning_rate": 0.00010897428710426498, + "loss": 1.1452, + "step": 10603 + }, + { + "epoch": 1.8879985754985755, + "grad_norm": 0.6819368004798889, + "learning_rate": 0.00010896034602447616, + "loss": 1.0504, + "step": 10604 + }, + { + "epoch": 1.8881766381766383, + "grad_norm": 0.6781361103057861, + "learning_rate": 0.00010894640476912799, + "loss": 0.8719, + "step": 10605 + }, + { + "epoch": 1.8883547008547008, + "grad_norm": 0.621960461139679, + "learning_rate": 0.00010893246333849361, + "loss": 0.9264, + "step": 10606 + }, + { + "epoch": 1.8885327635327636, + "grad_norm": 0.6350592374801636, + "learning_rate": 0.00010891852173284615, + "loss": 1.0042, + "step": 10607 + }, + { + "epoch": 1.888710826210826, + "grad_norm": 0.6650694012641907, + "learning_rate": 0.00010890457995245879, + "loss": 1.1387, + "step": 10608 + }, + { + "epoch": 1.8888888888888888, + "grad_norm": 0.6515723466873169, + "learning_rate": 0.00010889063799760468, + "loss": 0.9508, + "step": 10609 + }, + { + "epoch": 1.8890669515669516, + "grad_norm": 0.6368890404701233, + "learning_rate": 0.000108876695868557, + "loss": 0.8051, + "step": 10610 + }, + { + "epoch": 1.8892450142450143, + "grad_norm": 0.7971013188362122, + "learning_rate": 0.00010886275356558888, + "loss": 0.8629, + "step": 10611 + }, + { + "epoch": 1.8894230769230769, + "grad_norm": 0.6739095449447632, + "learning_rate": 0.00010884881108897353, + "loss": 0.9606, + "step": 10612 + }, + { + "epoch": 1.8896011396011396, + "grad_norm": 0.7754076719284058, + "learning_rate": 0.00010883486843898412, + "loss": 1.0751, + "step": 10613 + }, + { + "epoch": 1.8897792022792022, + "grad_norm": 0.6538285613059998, + "learning_rate": 0.00010882092561589379, + "loss": 0.9288, + "step": 10614 + }, + { + "epoch": 1.889957264957265, + "grad_norm": 0.7373257875442505, + "learning_rate": 0.00010880698261997577, + "loss": 0.9884, + "step": 10615 + }, + { + "epoch": 1.8901353276353277, + "grad_norm": 0.6575660109519958, + "learning_rate": 0.00010879303945150321, + "loss": 1.0307, + "step": 10616 + }, + { + "epoch": 1.8903133903133904, + "grad_norm": 0.7500179409980774, + "learning_rate": 0.00010877909611074932, + "loss": 1.0812, + "step": 10617 + }, + { + "epoch": 1.890491452991453, + "grad_norm": 0.7607308030128479, + "learning_rate": 0.00010876515259798727, + "loss": 0.9746, + "step": 10618 + }, + { + "epoch": 1.8906695156695157, + "grad_norm": 0.7930253744125366, + "learning_rate": 0.00010875120891349024, + "loss": 0.7911, + "step": 10619 + }, + { + "epoch": 1.8908475783475782, + "grad_norm": 0.635254979133606, + "learning_rate": 0.00010873726505753148, + "loss": 1.0468, + "step": 10620 + }, + { + "epoch": 1.891025641025641, + "grad_norm": 0.7579759359359741, + "learning_rate": 0.00010872332103038414, + "loss": 0.9558, + "step": 10621 + }, + { + "epoch": 1.8912037037037037, + "grad_norm": 0.5841903686523438, + "learning_rate": 0.00010870937683232146, + "loss": 0.913, + "step": 10622 + }, + { + "epoch": 1.8913817663817665, + "grad_norm": 0.7088860273361206, + "learning_rate": 0.00010869543246361664, + "loss": 1.0814, + "step": 10623 + }, + { + "epoch": 1.8915598290598292, + "grad_norm": 0.6713772416114807, + "learning_rate": 0.00010868148792454285, + "loss": 0.9972, + "step": 10624 + }, + { + "epoch": 1.8917378917378918, + "grad_norm": 0.6733243465423584, + "learning_rate": 0.00010866754321537338, + "loss": 0.9596, + "step": 10625 + }, + { + "epoch": 1.8919159544159543, + "grad_norm": 0.7747747898101807, + "learning_rate": 0.00010865359833638138, + "loss": 1.0871, + "step": 10626 + }, + { + "epoch": 1.892094017094017, + "grad_norm": 0.677175760269165, + "learning_rate": 0.00010863965328784011, + "loss": 0.9939, + "step": 10627 + }, + { + "epoch": 1.8922720797720798, + "grad_norm": 0.7883930206298828, + "learning_rate": 0.00010862570807002279, + "loss": 1.0708, + "step": 10628 + }, + { + "epoch": 1.8924501424501425, + "grad_norm": 0.7003030180931091, + "learning_rate": 0.00010861176268320261, + "loss": 0.9791, + "step": 10629 + }, + { + "epoch": 1.8926282051282053, + "grad_norm": 0.7450358271598816, + "learning_rate": 0.00010859781712765284, + "loss": 0.9672, + "step": 10630 + }, + { + "epoch": 1.8928062678062678, + "grad_norm": 0.7776696085929871, + "learning_rate": 0.00010858387140364672, + "loss": 1.1037, + "step": 10631 + }, + { + "epoch": 1.8929843304843303, + "grad_norm": 0.6896173357963562, + "learning_rate": 0.00010856992551145745, + "loss": 1.0048, + "step": 10632 + }, + { + "epoch": 1.893162393162393, + "grad_norm": 0.5997697710990906, + "learning_rate": 0.00010855597945135834, + "loss": 0.8025, + "step": 10633 + }, + { + "epoch": 1.8933404558404558, + "grad_norm": 0.8781484365463257, + "learning_rate": 0.00010854203322362251, + "loss": 1.0014, + "step": 10634 + }, + { + "epoch": 1.8935185185185186, + "grad_norm": 0.6348843574523926, + "learning_rate": 0.00010852808682852334, + "loss": 0.9857, + "step": 10635 + }, + { + "epoch": 1.8936965811965814, + "grad_norm": 0.9704267978668213, + "learning_rate": 0.000108514140266334, + "loss": 1.0522, + "step": 10636 + }, + { + "epoch": 1.8938746438746439, + "grad_norm": 0.70372074842453, + "learning_rate": 0.00010850019353732779, + "loss": 1.1044, + "step": 10637 + }, + { + "epoch": 1.8940527065527064, + "grad_norm": 0.6528043150901794, + "learning_rate": 0.00010848624664177793, + "loss": 0.9328, + "step": 10638 + }, + { + "epoch": 1.8942307692307692, + "grad_norm": 0.6299768090248108, + "learning_rate": 0.00010847229957995768, + "loss": 1.0099, + "step": 10639 + }, + { + "epoch": 1.894408831908832, + "grad_norm": 0.6347038149833679, + "learning_rate": 0.00010845835235214034, + "loss": 1.1354, + "step": 10640 + }, + { + "epoch": 1.8945868945868947, + "grad_norm": 0.7087811827659607, + "learning_rate": 0.00010844440495859913, + "loss": 1.0543, + "step": 10641 + }, + { + "epoch": 1.8947649572649574, + "grad_norm": 0.7386305332183838, + "learning_rate": 0.00010843045739960738, + "loss": 0.9192, + "step": 10642 + }, + { + "epoch": 1.89494301994302, + "grad_norm": 0.6047097444534302, + "learning_rate": 0.00010841650967543833, + "loss": 0.8668, + "step": 10643 + }, + { + "epoch": 1.8951210826210825, + "grad_norm": 0.6779503226280212, + "learning_rate": 0.00010840256178636523, + "loss": 0.9263, + "step": 10644 + }, + { + "epoch": 1.8952991452991452, + "grad_norm": 0.7398194670677185, + "learning_rate": 0.00010838861373266138, + "loss": 0.9534, + "step": 10645 + }, + { + "epoch": 1.895477207977208, + "grad_norm": 0.8138558864593506, + "learning_rate": 0.00010837466551460011, + "loss": 0.9835, + "step": 10646 + }, + { + "epoch": 1.8956552706552707, + "grad_norm": 0.8847818374633789, + "learning_rate": 0.00010836071713245466, + "loss": 0.9769, + "step": 10647 + }, + { + "epoch": 1.8958333333333335, + "grad_norm": 0.6824164390563965, + "learning_rate": 0.0001083467685864983, + "loss": 0.9901, + "step": 10648 + }, + { + "epoch": 1.896011396011396, + "grad_norm": 0.6318182945251465, + "learning_rate": 0.00010833281987700436, + "loss": 0.7677, + "step": 10649 + }, + { + "epoch": 1.8961894586894585, + "grad_norm": 0.7372074127197266, + "learning_rate": 0.00010831887100424612, + "loss": 0.9858, + "step": 10650 + }, + { + "epoch": 1.8963675213675213, + "grad_norm": 0.7246516346931458, + "learning_rate": 0.00010830492196849688, + "loss": 0.9644, + "step": 10651 + }, + { + "epoch": 1.896545584045584, + "grad_norm": 0.6517095565795898, + "learning_rate": 0.00010829097277002997, + "loss": 1.1733, + "step": 10652 + }, + { + "epoch": 1.8967236467236468, + "grad_norm": 0.6931695342063904, + "learning_rate": 0.00010827702340911867, + "loss": 0.9923, + "step": 10653 + }, + { + "epoch": 1.8969017094017095, + "grad_norm": 0.6210272312164307, + "learning_rate": 0.00010826307388603628, + "loss": 0.8757, + "step": 10654 + }, + { + "epoch": 1.897079772079772, + "grad_norm": 0.7011165618896484, + "learning_rate": 0.00010824912420105611, + "loss": 1.0011, + "step": 10655 + }, + { + "epoch": 1.8972578347578346, + "grad_norm": 0.7431246638298035, + "learning_rate": 0.0001082351743544515, + "loss": 1.1498, + "step": 10656 + }, + { + "epoch": 1.8974358974358974, + "grad_norm": 0.7099978923797607, + "learning_rate": 0.00010822122434649576, + "loss": 1.0673, + "step": 10657 + }, + { + "epoch": 1.89761396011396, + "grad_norm": 0.7375551462173462, + "learning_rate": 0.00010820727417746219, + "loss": 1.0157, + "step": 10658 + }, + { + "epoch": 1.8977920227920229, + "grad_norm": 0.8155642151832581, + "learning_rate": 0.00010819332384762413, + "loss": 1.229, + "step": 10659 + }, + { + "epoch": 1.8979700854700856, + "grad_norm": 0.6917914748191833, + "learning_rate": 0.00010817937335725493, + "loss": 0.9701, + "step": 10660 + }, + { + "epoch": 1.8981481481481481, + "grad_norm": 0.8498218059539795, + "learning_rate": 0.00010816542270662786, + "loss": 1.0123, + "step": 10661 + }, + { + "epoch": 1.8983262108262107, + "grad_norm": 0.7234359979629517, + "learning_rate": 0.00010815147189601634, + "loss": 1.0755, + "step": 10662 + }, + { + "epoch": 1.8985042735042734, + "grad_norm": 0.6997553110122681, + "learning_rate": 0.00010813752092569365, + "loss": 1.1594, + "step": 10663 + }, + { + "epoch": 1.8986823361823362, + "grad_norm": 0.6519457101821899, + "learning_rate": 0.00010812356979593314, + "loss": 0.9609, + "step": 10664 + }, + { + "epoch": 1.898860398860399, + "grad_norm": 0.7215374708175659, + "learning_rate": 0.00010810961850700813, + "loss": 1.1392, + "step": 10665 + }, + { + "epoch": 1.8990384615384617, + "grad_norm": 0.7766093611717224, + "learning_rate": 0.00010809566705919202, + "loss": 1.0256, + "step": 10666 + }, + { + "epoch": 1.8992165242165242, + "grad_norm": 0.6520358920097351, + "learning_rate": 0.00010808171545275814, + "loss": 1.0434, + "step": 10667 + }, + { + "epoch": 1.8993945868945867, + "grad_norm": 0.7454953193664551, + "learning_rate": 0.00010806776368797982, + "loss": 1.2323, + "step": 10668 + }, + { + "epoch": 1.8995726495726495, + "grad_norm": 0.6891530752182007, + "learning_rate": 0.00010805381176513043, + "loss": 1.1104, + "step": 10669 + }, + { + "epoch": 1.8997507122507122, + "grad_norm": 0.6609626412391663, + "learning_rate": 0.00010803985968448331, + "loss": 0.8565, + "step": 10670 + }, + { + "epoch": 1.899928774928775, + "grad_norm": 0.6650999188423157, + "learning_rate": 0.00010802590744631187, + "loss": 1.1003, + "step": 10671 + }, + { + "epoch": 1.9001068376068377, + "grad_norm": 0.5794292092323303, + "learning_rate": 0.00010801195505088945, + "loss": 0.528, + "step": 10672 + }, + { + "epoch": 1.9002849002849003, + "grad_norm": 1.0802743434906006, + "learning_rate": 0.00010799800249848939, + "loss": 0.8861, + "step": 10673 + }, + { + "epoch": 1.9004629629629628, + "grad_norm": 0.650833249092102, + "learning_rate": 0.00010798404978938513, + "loss": 0.9962, + "step": 10674 + }, + { + "epoch": 1.9006410256410255, + "grad_norm": 0.7290451526641846, + "learning_rate": 0.00010797009692384994, + "loss": 1.0764, + "step": 10675 + }, + { + "epoch": 1.9008190883190883, + "grad_norm": 0.6273928880691528, + "learning_rate": 0.00010795614390215727, + "loss": 0.9478, + "step": 10676 + }, + { + "epoch": 1.900997150997151, + "grad_norm": 0.6939455270767212, + "learning_rate": 0.00010794219072458052, + "loss": 0.8991, + "step": 10677 + }, + { + "epoch": 1.9011752136752138, + "grad_norm": 0.7455828189849854, + "learning_rate": 0.00010792823739139302, + "loss": 0.8902, + "step": 10678 + }, + { + "epoch": 1.9013532763532763, + "grad_norm": 0.6894607543945312, + "learning_rate": 0.00010791428390286817, + "loss": 0.9355, + "step": 10679 + }, + { + "epoch": 1.901531339031339, + "grad_norm": 0.6844658851623535, + "learning_rate": 0.00010790033025927936, + "loss": 0.9835, + "step": 10680 + }, + { + "epoch": 1.9017094017094016, + "grad_norm": 0.6646730899810791, + "learning_rate": 0.00010788637646090001, + "loss": 0.9376, + "step": 10681 + }, + { + "epoch": 1.9018874643874644, + "grad_norm": 0.6494864225387573, + "learning_rate": 0.00010787242250800349, + "loss": 0.8533, + "step": 10682 + }, + { + "epoch": 1.9020655270655271, + "grad_norm": 0.686198353767395, + "learning_rate": 0.0001078584684008632, + "loss": 0.8075, + "step": 10683 + }, + { + "epoch": 1.9022435897435899, + "grad_norm": 0.7014855742454529, + "learning_rate": 0.00010784451413975256, + "loss": 1.0805, + "step": 10684 + }, + { + "epoch": 1.9024216524216524, + "grad_norm": 0.7191864252090454, + "learning_rate": 0.00010783055972494496, + "loss": 0.9375, + "step": 10685 + }, + { + "epoch": 1.9025997150997151, + "grad_norm": 0.8114212155342102, + "learning_rate": 0.00010781660515671379, + "loss": 0.9716, + "step": 10686 + }, + { + "epoch": 1.9027777777777777, + "grad_norm": 0.7423529028892517, + "learning_rate": 0.0001078026504353325, + "loss": 0.9066, + "step": 10687 + }, + { + "epoch": 1.9029558404558404, + "grad_norm": 0.6517882347106934, + "learning_rate": 0.00010778869556107447, + "loss": 0.9908, + "step": 10688 + }, + { + "epoch": 1.9031339031339032, + "grad_norm": 0.6983367800712585, + "learning_rate": 0.00010777474053421315, + "loss": 1.1048, + "step": 10689 + }, + { + "epoch": 1.903311965811966, + "grad_norm": 0.597766101360321, + "learning_rate": 0.00010776078535502193, + "loss": 0.84, + "step": 10690 + }, + { + "epoch": 1.9034900284900285, + "grad_norm": 0.7335455417633057, + "learning_rate": 0.00010774683002377422, + "loss": 1.0387, + "step": 10691 + }, + { + "epoch": 1.9036680911680912, + "grad_norm": 0.6742176413536072, + "learning_rate": 0.0001077328745407435, + "loss": 0.9743, + "step": 10692 + }, + { + "epoch": 1.9038461538461537, + "grad_norm": 0.7954961657524109, + "learning_rate": 0.00010771891890620316, + "loss": 1.1025, + "step": 10693 + }, + { + "epoch": 1.9040242165242165, + "grad_norm": 0.733351469039917, + "learning_rate": 0.00010770496312042664, + "loss": 1.028, + "step": 10694 + }, + { + "epoch": 1.9042022792022792, + "grad_norm": 0.7059772610664368, + "learning_rate": 0.00010769100718368734, + "loss": 1.0103, + "step": 10695 + }, + { + "epoch": 1.904380341880342, + "grad_norm": 0.6234813332557678, + "learning_rate": 0.00010767705109625877, + "loss": 0.6893, + "step": 10696 + }, + { + "epoch": 1.9045584045584045, + "grad_norm": 0.6670311689376831, + "learning_rate": 0.0001076630948584143, + "loss": 1.1386, + "step": 10697 + }, + { + "epoch": 1.9047364672364673, + "grad_norm": 0.7444894909858704, + "learning_rate": 0.00010764913847042744, + "loss": 0.8524, + "step": 10698 + }, + { + "epoch": 1.9049145299145298, + "grad_norm": 0.6252964735031128, + "learning_rate": 0.00010763518193257158, + "loss": 0.9407, + "step": 10699 + }, + { + "epoch": 1.9050925925925926, + "grad_norm": 0.7794382572174072, + "learning_rate": 0.0001076212252451202, + "loss": 1.05, + "step": 10700 + }, + { + "epoch": 1.9052706552706553, + "grad_norm": 0.6313693523406982, + "learning_rate": 0.00010760726840834671, + "loss": 0.8667, + "step": 10701 + }, + { + "epoch": 1.905448717948718, + "grad_norm": 0.6766461730003357, + "learning_rate": 0.00010759331142252462, + "loss": 0.9675, + "step": 10702 + }, + { + "epoch": 1.9056267806267806, + "grad_norm": 0.7457365393638611, + "learning_rate": 0.00010757935428792739, + "loss": 0.9177, + "step": 10703 + }, + { + "epoch": 1.9058048433048433, + "grad_norm": 0.6649872064590454, + "learning_rate": 0.00010756539700482844, + "loss": 0.8703, + "step": 10704 + }, + { + "epoch": 1.9059829059829059, + "grad_norm": 0.8418740034103394, + "learning_rate": 0.00010755143957350127, + "loss": 0.8993, + "step": 10705 + }, + { + "epoch": 1.9061609686609686, + "grad_norm": 0.6767167448997498, + "learning_rate": 0.00010753748199421929, + "loss": 1.0063, + "step": 10706 + }, + { + "epoch": 1.9063390313390314, + "grad_norm": 0.6959242820739746, + "learning_rate": 0.00010752352426725603, + "loss": 1.0516, + "step": 10707 + }, + { + "epoch": 1.9065170940170941, + "grad_norm": 0.7106529474258423, + "learning_rate": 0.00010750956639288493, + "loss": 0.9596, + "step": 10708 + }, + { + "epoch": 1.9066951566951567, + "grad_norm": 0.7611243724822998, + "learning_rate": 0.00010749560837137949, + "loss": 1.0739, + "step": 10709 + }, + { + "epoch": 1.9068732193732194, + "grad_norm": 0.6684338450431824, + "learning_rate": 0.00010748165020301317, + "loss": 1.1437, + "step": 10710 + }, + { + "epoch": 1.907051282051282, + "grad_norm": 0.5957385897636414, + "learning_rate": 0.00010746769188805945, + "loss": 0.8802, + "step": 10711 + }, + { + "epoch": 1.9072293447293447, + "grad_norm": 0.69919353723526, + "learning_rate": 0.00010745373342679184, + "loss": 1.1891, + "step": 10712 + }, + { + "epoch": 1.9074074074074074, + "grad_norm": 0.7562127709388733, + "learning_rate": 0.0001074397748194838, + "loss": 0.8717, + "step": 10713 + }, + { + "epoch": 1.9075854700854702, + "grad_norm": 0.6420038938522339, + "learning_rate": 0.00010742581606640882, + "loss": 1.1196, + "step": 10714 + }, + { + "epoch": 1.9077635327635327, + "grad_norm": 0.7545611262321472, + "learning_rate": 0.00010741185716784039, + "loss": 1.161, + "step": 10715 + }, + { + "epoch": 1.9079415954415955, + "grad_norm": 0.6467727422714233, + "learning_rate": 0.000107397898124052, + "loss": 0.8029, + "step": 10716 + }, + { + "epoch": 1.908119658119658, + "grad_norm": 0.6129235625267029, + "learning_rate": 0.00010738393893531722, + "loss": 0.8802, + "step": 10717 + }, + { + "epoch": 1.9082977207977208, + "grad_norm": 0.6416113376617432, + "learning_rate": 0.00010736997960190946, + "loss": 0.8465, + "step": 10718 + }, + { + "epoch": 1.9084757834757835, + "grad_norm": 0.6609050631523132, + "learning_rate": 0.00010735602012410229, + "loss": 0.9484, + "step": 10719 + }, + { + "epoch": 1.9086538461538463, + "grad_norm": 0.6302639842033386, + "learning_rate": 0.00010734206050216913, + "loss": 0.898, + "step": 10720 + }, + { + "epoch": 1.9088319088319088, + "grad_norm": 0.7291215658187866, + "learning_rate": 0.00010732810073638358, + "loss": 0.9544, + "step": 10721 + }, + { + "epoch": 1.9090099715099715, + "grad_norm": 0.6436966061592102, + "learning_rate": 0.0001073141408270191, + "loss": 0.956, + "step": 10722 + }, + { + "epoch": 1.909188034188034, + "grad_norm": 0.6247875094413757, + "learning_rate": 0.00010730018077434924, + "loss": 0.8704, + "step": 10723 + }, + { + "epoch": 1.9093660968660968, + "grad_norm": 0.7599029541015625, + "learning_rate": 0.00010728622057864753, + "loss": 1.2024, + "step": 10724 + }, + { + "epoch": 1.9095441595441596, + "grad_norm": 0.6894544959068298, + "learning_rate": 0.00010727226024018744, + "loss": 1.1226, + "step": 10725 + }, + { + "epoch": 1.9097222222222223, + "grad_norm": 0.6920733451843262, + "learning_rate": 0.0001072582997592425, + "loss": 0.7682, + "step": 10726 + }, + { + "epoch": 1.9099002849002849, + "grad_norm": 0.6013005375862122, + "learning_rate": 0.00010724433913608627, + "loss": 0.9462, + "step": 10727 + }, + { + "epoch": 1.9100783475783476, + "grad_norm": 0.7466302514076233, + "learning_rate": 0.00010723037837099225, + "loss": 0.9507, + "step": 10728 + }, + { + "epoch": 1.9102564102564101, + "grad_norm": 0.7070091962814331, + "learning_rate": 0.00010721641746423401, + "loss": 1.0704, + "step": 10729 + }, + { + "epoch": 1.9104344729344729, + "grad_norm": 0.6747950315475464, + "learning_rate": 0.00010720245641608506, + "loss": 0.7899, + "step": 10730 + }, + { + "epoch": 1.9106125356125356, + "grad_norm": 0.7338371276855469, + "learning_rate": 0.00010718849522681891, + "loss": 0.9574, + "step": 10731 + }, + { + "epoch": 1.9107905982905984, + "grad_norm": 0.6923216581344604, + "learning_rate": 0.00010717453389670915, + "loss": 1.0725, + "step": 10732 + }, + { + "epoch": 1.910968660968661, + "grad_norm": 0.6050783395767212, + "learning_rate": 0.0001071605724260293, + "loss": 0.9224, + "step": 10733 + }, + { + "epoch": 1.9111467236467237, + "grad_norm": 0.6854597330093384, + "learning_rate": 0.00010714661081505291, + "loss": 0.9749, + "step": 10734 + }, + { + "epoch": 1.9113247863247862, + "grad_norm": 0.7661508321762085, + "learning_rate": 0.00010713264906405351, + "loss": 1.1564, + "step": 10735 + }, + { + "epoch": 1.911502849002849, + "grad_norm": 0.6389622688293457, + "learning_rate": 0.00010711868717330467, + "loss": 0.8148, + "step": 10736 + }, + { + "epoch": 1.9116809116809117, + "grad_norm": 0.6318161487579346, + "learning_rate": 0.00010710472514307996, + "loss": 0.7833, + "step": 10737 + }, + { + "epoch": 1.9118589743589745, + "grad_norm": 0.8646727800369263, + "learning_rate": 0.00010709076297365292, + "loss": 1.2682, + "step": 10738 + }, + { + "epoch": 1.9120370370370372, + "grad_norm": 0.6085501909255981, + "learning_rate": 0.0001070768006652971, + "loss": 0.8706, + "step": 10739 + }, + { + "epoch": 1.9122150997150997, + "grad_norm": 0.8259731531143188, + "learning_rate": 0.00010706283821828607, + "loss": 0.9014, + "step": 10740 + }, + { + "epoch": 1.9123931623931623, + "grad_norm": 0.6509148478507996, + "learning_rate": 0.0001070488756328934, + "loss": 0.8814, + "step": 10741 + }, + { + "epoch": 1.912571225071225, + "grad_norm": 0.7241966128349304, + "learning_rate": 0.00010703491290939264, + "loss": 0.9925, + "step": 10742 + }, + { + "epoch": 1.9127492877492878, + "grad_norm": 0.7736822366714478, + "learning_rate": 0.00010702095004805738, + "loss": 1.0881, + "step": 10743 + }, + { + "epoch": 1.9129273504273505, + "grad_norm": 0.6912824511528015, + "learning_rate": 0.00010700698704916123, + "loss": 1.2334, + "step": 10744 + }, + { + "epoch": 1.9131054131054133, + "grad_norm": 0.825065553188324, + "learning_rate": 0.0001069930239129777, + "loss": 0.9783, + "step": 10745 + }, + { + "epoch": 1.9132834757834758, + "grad_norm": 0.7650560140609741, + "learning_rate": 0.00010697906063978038, + "loss": 0.9788, + "step": 10746 + }, + { + "epoch": 1.9134615384615383, + "grad_norm": 0.7368232607841492, + "learning_rate": 0.00010696509722984287, + "loss": 0.8704, + "step": 10747 + }, + { + "epoch": 1.913639601139601, + "grad_norm": 0.6630628108978271, + "learning_rate": 0.00010695113368343875, + "loss": 1.1993, + "step": 10748 + }, + { + "epoch": 1.9138176638176638, + "grad_norm": 0.6842190027236938, + "learning_rate": 0.0001069371700008416, + "loss": 0.9128, + "step": 10749 + }, + { + "epoch": 1.9139957264957266, + "grad_norm": 0.591655969619751, + "learning_rate": 0.00010692320618232503, + "loss": 1.0607, + "step": 10750 + }, + { + "epoch": 1.9141737891737893, + "grad_norm": 0.74644535779953, + "learning_rate": 0.0001069092422281626, + "loss": 1.0937, + "step": 10751 + }, + { + "epoch": 1.9143518518518519, + "grad_norm": 0.7123813629150391, + "learning_rate": 0.00010689527813862792, + "loss": 0.9043, + "step": 10752 + }, + { + "epoch": 1.9145299145299144, + "grad_norm": 0.6850089430809021, + "learning_rate": 0.0001068813139139946, + "loss": 1.0908, + "step": 10753 + }, + { + "epoch": 1.9147079772079771, + "grad_norm": 0.5882078409194946, + "learning_rate": 0.00010686734955453623, + "loss": 0.829, + "step": 10754 + }, + { + "epoch": 1.91488603988604, + "grad_norm": 0.6741717457771301, + "learning_rate": 0.00010685338506052642, + "loss": 0.9197, + "step": 10755 + }, + { + "epoch": 1.9150641025641026, + "grad_norm": 0.6597354412078857, + "learning_rate": 0.00010683942043223876, + "loss": 0.8778, + "step": 10756 + }, + { + "epoch": 1.9152421652421654, + "grad_norm": 0.6682151556015015, + "learning_rate": 0.00010682545566994684, + "loss": 0.9305, + "step": 10757 + }, + { + "epoch": 1.915420227920228, + "grad_norm": 0.8283176422119141, + "learning_rate": 0.00010681149077392431, + "loss": 1.0164, + "step": 10758 + }, + { + "epoch": 1.9155982905982905, + "grad_norm": 0.648845374584198, + "learning_rate": 0.00010679752574444477, + "loss": 1.0114, + "step": 10759 + }, + { + "epoch": 1.9157763532763532, + "grad_norm": 0.755913496017456, + "learning_rate": 0.00010678356058178182, + "loss": 1.1142, + "step": 10760 + }, + { + "epoch": 1.915954415954416, + "grad_norm": 0.7334780097007751, + "learning_rate": 0.00010676959528620911, + "loss": 0.8758, + "step": 10761 + }, + { + "epoch": 1.9161324786324787, + "grad_norm": 0.9132041335105896, + "learning_rate": 0.00010675562985800025, + "loss": 0.995, + "step": 10762 + }, + { + "epoch": 1.9163105413105415, + "grad_norm": 0.7070860266685486, + "learning_rate": 0.00010674166429742882, + "loss": 0.9856, + "step": 10763 + }, + { + "epoch": 1.916488603988604, + "grad_norm": 0.7143638134002686, + "learning_rate": 0.00010672769860476853, + "loss": 1.0612, + "step": 10764 + }, + { + "epoch": 1.9166666666666665, + "grad_norm": 0.815717339515686, + "learning_rate": 0.00010671373278029293, + "loss": 1.1539, + "step": 10765 + }, + { + "epoch": 1.9168447293447293, + "grad_norm": 0.6379499435424805, + "learning_rate": 0.0001066997668242757, + "loss": 0.8295, + "step": 10766 + }, + { + "epoch": 1.917022792022792, + "grad_norm": 0.6482511758804321, + "learning_rate": 0.00010668580073699044, + "loss": 1.0079, + "step": 10767 + }, + { + "epoch": 1.9172008547008548, + "grad_norm": 0.7382873296737671, + "learning_rate": 0.00010667183451871082, + "loss": 0.8973, + "step": 10768 + }, + { + "epoch": 1.9173789173789175, + "grad_norm": 0.7818579077720642, + "learning_rate": 0.00010665786816971044, + "loss": 1.2131, + "step": 10769 + }, + { + "epoch": 1.91755698005698, + "grad_norm": 0.6960901021957397, + "learning_rate": 0.000106643901690263, + "loss": 1.1466, + "step": 10770 + }, + { + "epoch": 1.9177350427350426, + "grad_norm": 0.696966826915741, + "learning_rate": 0.00010662993508064208, + "loss": 0.854, + "step": 10771 + }, + { + "epoch": 1.9179131054131053, + "grad_norm": 0.6745442152023315, + "learning_rate": 0.00010661596834112133, + "loss": 0.9559, + "step": 10772 + }, + { + "epoch": 1.918091168091168, + "grad_norm": 0.7436230778694153, + "learning_rate": 0.00010660200147197447, + "loss": 1.1367, + "step": 10773 + }, + { + "epoch": 1.9182692307692308, + "grad_norm": 0.6051676869392395, + "learning_rate": 0.00010658803447347509, + "loss": 1.05, + "step": 10774 + }, + { + "epoch": 1.9184472934472936, + "grad_norm": 0.5662530660629272, + "learning_rate": 0.00010657406734589686, + "loss": 0.8697, + "step": 10775 + }, + { + "epoch": 1.9186253561253561, + "grad_norm": 0.6640757322311401, + "learning_rate": 0.00010656010008951344, + "loss": 1.0636, + "step": 10776 + }, + { + "epoch": 1.9188034188034186, + "grad_norm": 0.6994011998176575, + "learning_rate": 0.00010654613270459848, + "loss": 0.9326, + "step": 10777 + }, + { + "epoch": 1.9189814814814814, + "grad_norm": 0.6827420592308044, + "learning_rate": 0.00010653216519142563, + "loss": 0.8667, + "step": 10778 + }, + { + "epoch": 1.9191595441595442, + "grad_norm": 0.6814691424369812, + "learning_rate": 0.00010651819755026862, + "loss": 0.828, + "step": 10779 + }, + { + "epoch": 1.919337606837607, + "grad_norm": 0.7033611536026001, + "learning_rate": 0.00010650422978140103, + "loss": 1.0427, + "step": 10780 + }, + { + "epoch": 1.9195156695156697, + "grad_norm": 0.7098833322525024, + "learning_rate": 0.00010649026188509657, + "loss": 1.1723, + "step": 10781 + }, + { + "epoch": 1.9196937321937322, + "grad_norm": 0.7184767723083496, + "learning_rate": 0.00010647629386162893, + "loss": 0.852, + "step": 10782 + }, + { + "epoch": 1.9198717948717947, + "grad_norm": 0.6682565808296204, + "learning_rate": 0.00010646232571127175, + "loss": 0.8827, + "step": 10783 + }, + { + "epoch": 1.9200498575498575, + "grad_norm": 0.6699280142784119, + "learning_rate": 0.00010644835743429873, + "loss": 0.8346, + "step": 10784 + }, + { + "epoch": 1.9202279202279202, + "grad_norm": 0.8041857481002808, + "learning_rate": 0.00010643438903098355, + "loss": 0.9622, + "step": 10785 + }, + { + "epoch": 1.920405982905983, + "grad_norm": 0.7315110564231873, + "learning_rate": 0.00010642042050159986, + "loss": 1.0443, + "step": 10786 + }, + { + "epoch": 1.9205840455840457, + "grad_norm": 0.5850204229354858, + "learning_rate": 0.0001064064518464214, + "loss": 1.0155, + "step": 10787 + }, + { + "epoch": 1.9207621082621082, + "grad_norm": 0.7320640683174133, + "learning_rate": 0.00010639248306572178, + "loss": 1.1556, + "step": 10788 + }, + { + "epoch": 1.9209401709401708, + "grad_norm": 0.689804196357727, + "learning_rate": 0.00010637851415977478, + "loss": 1.1058, + "step": 10789 + }, + { + "epoch": 1.9211182336182335, + "grad_norm": 0.6433262228965759, + "learning_rate": 0.000106364545128854, + "loss": 1.0916, + "step": 10790 + }, + { + "epoch": 1.9212962962962963, + "grad_norm": 0.6802626252174377, + "learning_rate": 0.00010635057597323323, + "loss": 1.126, + "step": 10791 + }, + { + "epoch": 1.921474358974359, + "grad_norm": 0.7503384351730347, + "learning_rate": 0.00010633660669318608, + "loss": 0.9354, + "step": 10792 + }, + { + "epoch": 1.9216524216524218, + "grad_norm": 0.6370253562927246, + "learning_rate": 0.00010632263728898629, + "loss": 0.9976, + "step": 10793 + }, + { + "epoch": 1.9218304843304843, + "grad_norm": 0.7566042542457581, + "learning_rate": 0.00010630866776090755, + "loss": 1.0311, + "step": 10794 + }, + { + "epoch": 1.922008547008547, + "grad_norm": 0.7011943459510803, + "learning_rate": 0.0001062946981092236, + "loss": 0.8777, + "step": 10795 + }, + { + "epoch": 1.9221866096866096, + "grad_norm": 0.6621114015579224, + "learning_rate": 0.00010628072833420811, + "loss": 0.9615, + "step": 10796 + }, + { + "epoch": 1.9223646723646723, + "grad_norm": 0.6863150000572205, + "learning_rate": 0.00010626675843613478, + "loss": 1.071, + "step": 10797 + }, + { + "epoch": 1.922542735042735, + "grad_norm": 0.597970724105835, + "learning_rate": 0.00010625278841527733, + "loss": 0.8661, + "step": 10798 + }, + { + "epoch": 1.9227207977207978, + "grad_norm": 0.5958755612373352, + "learning_rate": 0.00010623881827190947, + "loss": 0.9075, + "step": 10799 + }, + { + "epoch": 1.9228988603988604, + "grad_norm": 0.7764523029327393, + "learning_rate": 0.00010622484800630494, + "loss": 1.0576, + "step": 10800 + }, + { + "epoch": 1.9230769230769231, + "grad_norm": 0.774156391620636, + "learning_rate": 0.00010621087761873748, + "loss": 0.9273, + "step": 10801 + }, + { + "epoch": 1.9232549857549857, + "grad_norm": 0.6321687698364258, + "learning_rate": 0.00010619690710948074, + "loss": 0.8805, + "step": 10802 + }, + { + "epoch": 1.9234330484330484, + "grad_norm": 0.659538984298706, + "learning_rate": 0.00010618293647880846, + "loss": 0.9845, + "step": 10803 + }, + { + "epoch": 1.9236111111111112, + "grad_norm": 0.6931299567222595, + "learning_rate": 0.00010616896572699442, + "loss": 1.2005, + "step": 10804 + }, + { + "epoch": 1.923789173789174, + "grad_norm": 0.6054762005805969, + "learning_rate": 0.00010615499485431228, + "loss": 0.825, + "step": 10805 + }, + { + "epoch": 1.9239672364672364, + "grad_norm": 0.6631526947021484, + "learning_rate": 0.00010614102386103584, + "loss": 0.9149, + "step": 10806 + }, + { + "epoch": 1.9241452991452992, + "grad_norm": 0.6667893528938293, + "learning_rate": 0.00010612705274743878, + "loss": 1.014, + "step": 10807 + }, + { + "epoch": 1.9243233618233617, + "grad_norm": 0.861302375793457, + "learning_rate": 0.00010611308151379482, + "loss": 1.0809, + "step": 10808 + }, + { + "epoch": 1.9245014245014245, + "grad_norm": 0.6997994780540466, + "learning_rate": 0.00010609911016037777, + "loss": 0.8897, + "step": 10809 + }, + { + "epoch": 1.9246794871794872, + "grad_norm": 0.5689206123352051, + "learning_rate": 0.00010608513868746131, + "loss": 0.7517, + "step": 10810 + }, + { + "epoch": 1.92485754985755, + "grad_norm": 0.5972287654876709, + "learning_rate": 0.00010607116709531918, + "loss": 0.9015, + "step": 10811 + }, + { + "epoch": 1.9250356125356125, + "grad_norm": 0.7115643620491028, + "learning_rate": 0.00010605719538422519, + "loss": 0.6974, + "step": 10812 + }, + { + "epoch": 1.9252136752136753, + "grad_norm": 0.6548098921775818, + "learning_rate": 0.00010604322355445297, + "loss": 0.7075, + "step": 10813 + }, + { + "epoch": 1.9253917378917378, + "grad_norm": 0.6666337847709656, + "learning_rate": 0.00010602925160627639, + "loss": 1.0389, + "step": 10814 + }, + { + "epoch": 1.9255698005698005, + "grad_norm": 0.7754444479942322, + "learning_rate": 0.00010601527953996913, + "loss": 1.0674, + "step": 10815 + }, + { + "epoch": 1.9257478632478633, + "grad_norm": 0.6602712869644165, + "learning_rate": 0.00010600130735580498, + "loss": 1.2622, + "step": 10816 + }, + { + "epoch": 1.925925925925926, + "grad_norm": 0.6974020004272461, + "learning_rate": 0.00010598733505405767, + "loss": 0.9748, + "step": 10817 + }, + { + "epoch": 1.9261039886039886, + "grad_norm": 0.6236271858215332, + "learning_rate": 0.00010597336263500095, + "loss": 0.9463, + "step": 10818 + }, + { + "epoch": 1.9262820512820513, + "grad_norm": 0.6856079697608948, + "learning_rate": 0.00010595939009890859, + "loss": 0.9484, + "step": 10819 + }, + { + "epoch": 1.9264601139601139, + "grad_norm": 0.7300925850868225, + "learning_rate": 0.00010594541744605437, + "loss": 0.9702, + "step": 10820 + }, + { + "epoch": 1.9266381766381766, + "grad_norm": 0.6546478867530823, + "learning_rate": 0.00010593144467671208, + "loss": 0.8235, + "step": 10821 + }, + { + "epoch": 1.9268162393162394, + "grad_norm": 0.7215169072151184, + "learning_rate": 0.00010591747179115543, + "loss": 0.9986, + "step": 10822 + }, + { + "epoch": 1.926994301994302, + "grad_norm": 0.7304712533950806, + "learning_rate": 0.00010590349878965822, + "loss": 1.099, + "step": 10823 + }, + { + "epoch": 1.9271723646723646, + "grad_norm": 0.5883305668830872, + "learning_rate": 0.0001058895256724942, + "loss": 1.0647, + "step": 10824 + }, + { + "epoch": 1.9273504273504274, + "grad_norm": 0.8067272305488586, + "learning_rate": 0.00010587555243993716, + "loss": 1.0295, + "step": 10825 + }, + { + "epoch": 1.92752849002849, + "grad_norm": 0.6607550978660583, + "learning_rate": 0.00010586157909226089, + "loss": 0.8669, + "step": 10826 + }, + { + "epoch": 1.9277065527065527, + "grad_norm": 0.7256106734275818, + "learning_rate": 0.00010584760562973914, + "loss": 1.1674, + "step": 10827 + }, + { + "epoch": 1.9278846153846154, + "grad_norm": 0.6584621071815491, + "learning_rate": 0.00010583363205264574, + "loss": 0.8901, + "step": 10828 + }, + { + "epoch": 1.9280626780626782, + "grad_norm": 0.7200617790222168, + "learning_rate": 0.00010581965836125439, + "loss": 1.0463, + "step": 10829 + }, + { + "epoch": 1.9282407407407407, + "grad_norm": 0.7244223952293396, + "learning_rate": 0.00010580568455583894, + "loss": 1.0973, + "step": 10830 + }, + { + "epoch": 1.9284188034188035, + "grad_norm": 0.7678009867668152, + "learning_rate": 0.00010579171063667317, + "loss": 1.1753, + "step": 10831 + }, + { + "epoch": 1.928596866096866, + "grad_norm": 0.6455881595611572, + "learning_rate": 0.00010577773660403085, + "loss": 0.8988, + "step": 10832 + }, + { + "epoch": 1.9287749287749287, + "grad_norm": 0.6804864406585693, + "learning_rate": 0.0001057637624581858, + "loss": 0.8156, + "step": 10833 + }, + { + "epoch": 1.9289529914529915, + "grad_norm": 0.7874828577041626, + "learning_rate": 0.00010574978819941176, + "loss": 1.1876, + "step": 10834 + }, + { + "epoch": 1.9291310541310542, + "grad_norm": 0.7396490573883057, + "learning_rate": 0.00010573581382798261, + "loss": 0.8709, + "step": 10835 + }, + { + "epoch": 1.9293091168091168, + "grad_norm": 0.6800381541252136, + "learning_rate": 0.00010572183934417209, + "loss": 0.9906, + "step": 10836 + }, + { + "epoch": 1.9294871794871795, + "grad_norm": 0.7077754139900208, + "learning_rate": 0.000105707864748254, + "loss": 0.9785, + "step": 10837 + }, + { + "epoch": 1.929665242165242, + "grad_norm": 0.693249523639679, + "learning_rate": 0.00010569389004050216, + "loss": 0.9515, + "step": 10838 + }, + { + "epoch": 1.9298433048433048, + "grad_norm": 0.706924319267273, + "learning_rate": 0.00010567991522119037, + "loss": 1.074, + "step": 10839 + }, + { + "epoch": 1.9300213675213675, + "grad_norm": 0.6504101157188416, + "learning_rate": 0.00010566594029059244, + "loss": 1.0635, + "step": 10840 + }, + { + "epoch": 1.9301994301994303, + "grad_norm": 0.7620238661766052, + "learning_rate": 0.00010565196524898219, + "loss": 0.944, + "step": 10841 + }, + { + "epoch": 1.9303774928774928, + "grad_norm": 0.6713484525680542, + "learning_rate": 0.00010563799009663344, + "loss": 0.749, + "step": 10842 + }, + { + "epoch": 1.9305555555555556, + "grad_norm": 0.9279242157936096, + "learning_rate": 0.00010562401483381997, + "loss": 0.961, + "step": 10843 + }, + { + "epoch": 1.930733618233618, + "grad_norm": 0.6710723638534546, + "learning_rate": 0.00010561003946081558, + "loss": 1.1288, + "step": 10844 + }, + { + "epoch": 1.9309116809116809, + "grad_norm": 0.7751701474189758, + "learning_rate": 0.00010559606397789416, + "loss": 0.9435, + "step": 10845 + }, + { + "epoch": 1.9310897435897436, + "grad_norm": 0.6741766929626465, + "learning_rate": 0.00010558208838532948, + "loss": 1.0299, + "step": 10846 + }, + { + "epoch": 1.9312678062678064, + "grad_norm": 0.6988041400909424, + "learning_rate": 0.00010556811268339539, + "loss": 1.0236, + "step": 10847 + }, + { + "epoch": 1.931445868945869, + "grad_norm": 0.6353505253791809, + "learning_rate": 0.00010555413687236568, + "loss": 1.0361, + "step": 10848 + }, + { + "epoch": 1.9316239316239316, + "grad_norm": 0.7162703275680542, + "learning_rate": 0.0001055401609525142, + "loss": 1.0931, + "step": 10849 + }, + { + "epoch": 1.9318019943019942, + "grad_norm": 0.61545330286026, + "learning_rate": 0.00010552618492411476, + "loss": 0.8829, + "step": 10850 + }, + { + "epoch": 1.931980056980057, + "grad_norm": 0.6304612159729004, + "learning_rate": 0.00010551220878744124, + "loss": 0.8574, + "step": 10851 + }, + { + "epoch": 1.9321581196581197, + "grad_norm": 0.6372067928314209, + "learning_rate": 0.00010549823254276743, + "loss": 1.0949, + "step": 10852 + }, + { + "epoch": 1.9323361823361824, + "grad_norm": 0.6952856779098511, + "learning_rate": 0.00010548425619036715, + "loss": 0.9232, + "step": 10853 + }, + { + "epoch": 1.9325142450142452, + "grad_norm": 0.6510106325149536, + "learning_rate": 0.00010547027973051427, + "loss": 1.0753, + "step": 10854 + }, + { + "epoch": 1.9326923076923077, + "grad_norm": 0.6377716064453125, + "learning_rate": 0.00010545630316348263, + "loss": 0.8466, + "step": 10855 + }, + { + "epoch": 1.9328703703703702, + "grad_norm": 0.7366968393325806, + "learning_rate": 0.00010544232648954606, + "loss": 0.9351, + "step": 10856 + }, + { + "epoch": 1.933048433048433, + "grad_norm": 0.703652024269104, + "learning_rate": 0.00010542834970897843, + "loss": 1.0032, + "step": 10857 + }, + { + "epoch": 1.9332264957264957, + "grad_norm": 0.6685494780540466, + "learning_rate": 0.00010541437282205355, + "loss": 0.8818, + "step": 10858 + }, + { + "epoch": 1.9334045584045585, + "grad_norm": 0.6594362854957581, + "learning_rate": 0.00010540039582904527, + "loss": 0.9535, + "step": 10859 + }, + { + "epoch": 1.9335826210826212, + "grad_norm": 0.8003259301185608, + "learning_rate": 0.00010538641873022744, + "loss": 0.8852, + "step": 10860 + }, + { + "epoch": 1.9337606837606838, + "grad_norm": 0.6567012071609497, + "learning_rate": 0.00010537244152587393, + "loss": 1.0832, + "step": 10861 + }, + { + "epoch": 1.9339387464387463, + "grad_norm": 0.6714941263198853, + "learning_rate": 0.00010535846421625862, + "loss": 1.1047, + "step": 10862 + }, + { + "epoch": 1.934116809116809, + "grad_norm": 0.6998924612998962, + "learning_rate": 0.00010534448680165531, + "loss": 0.8827, + "step": 10863 + }, + { + "epoch": 1.9342948717948718, + "grad_norm": 0.6065765619277954, + "learning_rate": 0.0001053305092823379, + "loss": 0.5773, + "step": 10864 + }, + { + "epoch": 1.9344729344729346, + "grad_norm": 0.7678273916244507, + "learning_rate": 0.0001053165316585802, + "loss": 0.9199, + "step": 10865 + }, + { + "epoch": 1.9346509971509973, + "grad_norm": 0.7071540951728821, + "learning_rate": 0.00010530255393065613, + "loss": 1.0292, + "step": 10866 + }, + { + "epoch": 1.9348290598290598, + "grad_norm": 0.6329835057258606, + "learning_rate": 0.00010528857609883956, + "loss": 0.9915, + "step": 10867 + }, + { + "epoch": 1.9350071225071224, + "grad_norm": 0.6274038553237915, + "learning_rate": 0.00010527459816340427, + "loss": 0.8499, + "step": 10868 + }, + { + "epoch": 1.9351851851851851, + "grad_norm": 0.6564371585845947, + "learning_rate": 0.00010526062012462424, + "loss": 1.1707, + "step": 10869 + }, + { + "epoch": 1.9353632478632479, + "grad_norm": 0.8561269044876099, + "learning_rate": 0.00010524664198277326, + "loss": 1.148, + "step": 10870 + }, + { + "epoch": 1.9355413105413106, + "grad_norm": 0.6322671175003052, + "learning_rate": 0.00010523266373812521, + "loss": 0.9165, + "step": 10871 + }, + { + "epoch": 1.9357193732193734, + "grad_norm": 0.7602947354316711, + "learning_rate": 0.00010521868539095403, + "loss": 0.9647, + "step": 10872 + }, + { + "epoch": 1.935897435897436, + "grad_norm": 0.5962168574333191, + "learning_rate": 0.00010520470694153353, + "loss": 0.8585, + "step": 10873 + }, + { + "epoch": 1.9360754985754984, + "grad_norm": 0.7498637437820435, + "learning_rate": 0.00010519072839013757, + "loss": 0.9828, + "step": 10874 + }, + { + "epoch": 1.9362535612535612, + "grad_norm": 0.6841256022453308, + "learning_rate": 0.00010517674973704012, + "loss": 0.9991, + "step": 10875 + }, + { + "epoch": 1.936431623931624, + "grad_norm": 0.8281826972961426, + "learning_rate": 0.00010516277098251499, + "loss": 1.028, + "step": 10876 + }, + { + "epoch": 1.9366096866096867, + "grad_norm": 0.6673563718795776, + "learning_rate": 0.0001051487921268361, + "loss": 1.1594, + "step": 10877 + }, + { + "epoch": 1.9367877492877494, + "grad_norm": 0.7833667993545532, + "learning_rate": 0.00010513481317027733, + "loss": 0.7675, + "step": 10878 + }, + { + "epoch": 1.936965811965812, + "grad_norm": 0.6087225675582886, + "learning_rate": 0.00010512083411311253, + "loss": 0.7803, + "step": 10879 + }, + { + "epoch": 1.9371438746438745, + "grad_norm": 0.6758120656013489, + "learning_rate": 0.00010510685495561563, + "loss": 1.0621, + "step": 10880 + }, + { + "epoch": 1.9373219373219372, + "grad_norm": 0.6720096468925476, + "learning_rate": 0.00010509287569806055, + "loss": 0.8502, + "step": 10881 + }, + { + "epoch": 1.9375, + "grad_norm": 0.6233887672424316, + "learning_rate": 0.00010507889634072113, + "loss": 1.0127, + "step": 10882 + }, + { + "epoch": 1.9376780626780628, + "grad_norm": 0.667742908000946, + "learning_rate": 0.00010506491688387127, + "loss": 0.9086, + "step": 10883 + }, + { + "epoch": 1.9378561253561255, + "grad_norm": 0.6533677577972412, + "learning_rate": 0.00010505093732778492, + "loss": 0.9724, + "step": 10884 + }, + { + "epoch": 1.938034188034188, + "grad_norm": 0.7171359062194824, + "learning_rate": 0.00010503695767273591, + "loss": 0.9915, + "step": 10885 + }, + { + "epoch": 1.9382122507122506, + "grad_norm": 0.723655641078949, + "learning_rate": 0.0001050229779189982, + "loss": 0.8981, + "step": 10886 + }, + { + "epoch": 1.9383903133903133, + "grad_norm": 0.6863494515419006, + "learning_rate": 0.00010500899806684568, + "loss": 1.2577, + "step": 10887 + }, + { + "epoch": 1.938568376068376, + "grad_norm": 0.8174706697463989, + "learning_rate": 0.00010499501811655224, + "loss": 0.9848, + "step": 10888 + }, + { + "epoch": 1.9387464387464388, + "grad_norm": 0.6378024220466614, + "learning_rate": 0.00010498103806839179, + "loss": 0.8499, + "step": 10889 + }, + { + "epoch": 1.9389245014245016, + "grad_norm": 0.6734544634819031, + "learning_rate": 0.00010496705792263823, + "loss": 0.8446, + "step": 10890 + }, + { + "epoch": 1.939102564102564, + "grad_norm": 0.6802361607551575, + "learning_rate": 0.00010495307767956551, + "loss": 0.9285, + "step": 10891 + }, + { + "epoch": 1.9392806267806266, + "grad_norm": 0.7821299433708191, + "learning_rate": 0.00010493909733944752, + "loss": 1.08, + "step": 10892 + }, + { + "epoch": 1.9394586894586894, + "grad_norm": 0.6204990148544312, + "learning_rate": 0.00010492511690255818, + "loss": 0.7861, + "step": 10893 + }, + { + "epoch": 1.9396367521367521, + "grad_norm": 0.6386391520500183, + "learning_rate": 0.0001049111363691714, + "loss": 0.9162, + "step": 10894 + }, + { + "epoch": 1.9398148148148149, + "grad_norm": 0.6885092854499817, + "learning_rate": 0.0001048971557395611, + "loss": 1.0026, + "step": 10895 + }, + { + "epoch": 1.9399928774928776, + "grad_norm": 0.6962558627128601, + "learning_rate": 0.00010488317501400122, + "loss": 1.146, + "step": 10896 + }, + { + "epoch": 1.9401709401709402, + "grad_norm": 0.6283716559410095, + "learning_rate": 0.00010486919419276566, + "loss": 1.0268, + "step": 10897 + }, + { + "epoch": 1.9403490028490027, + "grad_norm": 0.7183622717857361, + "learning_rate": 0.00010485521327612835, + "loss": 1.0123, + "step": 10898 + }, + { + "epoch": 1.9405270655270654, + "grad_norm": 0.6354197263717651, + "learning_rate": 0.00010484123226436321, + "loss": 0.871, + "step": 10899 + }, + { + "epoch": 1.9407051282051282, + "grad_norm": 0.804358184337616, + "learning_rate": 0.00010482725115774421, + "loss": 1.1001, + "step": 10900 + }, + { + "epoch": 1.940883190883191, + "grad_norm": 0.6896754503250122, + "learning_rate": 0.00010481326995654524, + "loss": 1.0976, + "step": 10901 + }, + { + "epoch": 1.9410612535612537, + "grad_norm": 0.9108015894889832, + "learning_rate": 0.00010479928866104023, + "loss": 0.8785, + "step": 10902 + }, + { + "epoch": 1.9412393162393162, + "grad_norm": 0.6963121294975281, + "learning_rate": 0.00010478530727150316, + "loss": 1.0458, + "step": 10903 + }, + { + "epoch": 1.9414173789173788, + "grad_norm": 0.6657114624977112, + "learning_rate": 0.00010477132578820792, + "loss": 0.8188, + "step": 10904 + }, + { + "epoch": 1.9415954415954415, + "grad_norm": 0.671716034412384, + "learning_rate": 0.00010475734421142847, + "loss": 1.0915, + "step": 10905 + }, + { + "epoch": 1.9417735042735043, + "grad_norm": 0.6790717244148254, + "learning_rate": 0.0001047433625414387, + "loss": 0.9688, + "step": 10906 + }, + { + "epoch": 1.941951566951567, + "grad_norm": 0.6411764621734619, + "learning_rate": 0.00010472938077851264, + "loss": 1.0387, + "step": 10907 + }, + { + "epoch": 1.9421296296296298, + "grad_norm": 0.8579615950584412, + "learning_rate": 0.00010471539892292417, + "loss": 1.1635, + "step": 10908 + }, + { + "epoch": 1.9423076923076923, + "grad_norm": 0.7031029462814331, + "learning_rate": 0.00010470141697494726, + "loss": 0.9813, + "step": 10909 + }, + { + "epoch": 1.9424857549857548, + "grad_norm": 0.6657388806343079, + "learning_rate": 0.00010468743493485584, + "loss": 0.7947, + "step": 10910 + }, + { + "epoch": 1.9426638176638176, + "grad_norm": 0.6364194750785828, + "learning_rate": 0.00010467345280292389, + "loss": 0.8554, + "step": 10911 + }, + { + "epoch": 1.9428418803418803, + "grad_norm": 0.7394127249717712, + "learning_rate": 0.00010465947057942534, + "loss": 0.822, + "step": 10912 + }, + { + "epoch": 1.943019943019943, + "grad_norm": 0.6557473540306091, + "learning_rate": 0.00010464548826463411, + "loss": 1.0025, + "step": 10913 + }, + { + "epoch": 1.9431980056980058, + "grad_norm": 0.6530601382255554, + "learning_rate": 0.00010463150585882422, + "loss": 1.0828, + "step": 10914 + }, + { + "epoch": 1.9433760683760684, + "grad_norm": 0.7376404404640198, + "learning_rate": 0.00010461752336226957, + "loss": 0.9413, + "step": 10915 + }, + { + "epoch": 1.943554131054131, + "grad_norm": 0.7110656499862671, + "learning_rate": 0.00010460354077524417, + "loss": 0.9162, + "step": 10916 + }, + { + "epoch": 1.9437321937321936, + "grad_norm": 0.6515666246414185, + "learning_rate": 0.00010458955809802194, + "loss": 0.9211, + "step": 10917 + }, + { + "epoch": 1.9439102564102564, + "grad_norm": 0.6888720989227295, + "learning_rate": 0.00010457557533087683, + "loss": 1.0632, + "step": 10918 + }, + { + "epoch": 1.9440883190883191, + "grad_norm": 0.7246627807617188, + "learning_rate": 0.00010456159247408286, + "loss": 0.9807, + "step": 10919 + }, + { + "epoch": 1.944266381766382, + "grad_norm": 0.727834165096283, + "learning_rate": 0.00010454760952791394, + "loss": 1.0793, + "step": 10920 + }, + { + "epoch": 1.9444444444444444, + "grad_norm": 0.6365306377410889, + "learning_rate": 0.00010453362649264407, + "loss": 1.0415, + "step": 10921 + }, + { + "epoch": 1.9446225071225072, + "grad_norm": 0.7187839150428772, + "learning_rate": 0.0001045196433685472, + "loss": 1.007, + "step": 10922 + }, + { + "epoch": 1.9448005698005697, + "grad_norm": 0.5905138254165649, + "learning_rate": 0.00010450566015589732, + "loss": 0.9818, + "step": 10923 + }, + { + "epoch": 1.9449786324786325, + "grad_norm": 0.7008894085884094, + "learning_rate": 0.00010449167685496837, + "loss": 0.8444, + "step": 10924 + }, + { + "epoch": 1.9451566951566952, + "grad_norm": 0.6126312017440796, + "learning_rate": 0.00010447769346603435, + "loss": 0.7207, + "step": 10925 + }, + { + "epoch": 1.945334757834758, + "grad_norm": 0.7513176202774048, + "learning_rate": 0.00010446370998936922, + "loss": 0.8693, + "step": 10926 + }, + { + "epoch": 1.9455128205128205, + "grad_norm": 0.6382531523704529, + "learning_rate": 0.00010444972642524697, + "loss": 0.8379, + "step": 10927 + }, + { + "epoch": 1.9456908831908832, + "grad_norm": 0.7062170505523682, + "learning_rate": 0.0001044357427739416, + "loss": 1.0525, + "step": 10928 + }, + { + "epoch": 1.9458689458689458, + "grad_norm": 0.6954067349433899, + "learning_rate": 0.00010442175903572703, + "loss": 1.0238, + "step": 10929 + }, + { + "epoch": 1.9460470085470085, + "grad_norm": 0.7257117033004761, + "learning_rate": 0.00010440777521087731, + "loss": 1.1413, + "step": 10930 + }, + { + "epoch": 1.9462250712250713, + "grad_norm": 0.6617701053619385, + "learning_rate": 0.00010439379129966635, + "loss": 1.0089, + "step": 10931 + }, + { + "epoch": 1.946403133903134, + "grad_norm": 0.6860800385475159, + "learning_rate": 0.00010437980730236821, + "loss": 1.1778, + "step": 10932 + }, + { + "epoch": 1.9465811965811965, + "grad_norm": 0.846235454082489, + "learning_rate": 0.00010436582321925684, + "loss": 0.9851, + "step": 10933 + }, + { + "epoch": 1.9467592592592593, + "grad_norm": 0.6385617852210999, + "learning_rate": 0.00010435183905060623, + "loss": 0.9542, + "step": 10934 + }, + { + "epoch": 1.9469373219373218, + "grad_norm": 0.7137401700019836, + "learning_rate": 0.00010433785479669038, + "loss": 1.0499, + "step": 10935 + }, + { + "epoch": 1.9471153846153846, + "grad_norm": 0.6269308924674988, + "learning_rate": 0.00010432387045778324, + "loss": 0.8929, + "step": 10936 + }, + { + "epoch": 1.9472934472934473, + "grad_norm": 0.7903163433074951, + "learning_rate": 0.00010430988603415888, + "loss": 0.9812, + "step": 10937 + }, + { + "epoch": 1.94747150997151, + "grad_norm": 0.6006736159324646, + "learning_rate": 0.00010429590152609121, + "loss": 0.7959, + "step": 10938 + }, + { + "epoch": 1.9476495726495726, + "grad_norm": 0.6061521768569946, + "learning_rate": 0.00010428191693385431, + "loss": 0.8748, + "step": 10939 + }, + { + "epoch": 1.9478276353276354, + "grad_norm": 0.6637623906135559, + "learning_rate": 0.00010426793225772216, + "loss": 0.7047, + "step": 10940 + }, + { + "epoch": 1.948005698005698, + "grad_norm": 0.7650586366653442, + "learning_rate": 0.00010425394749796874, + "loss": 1.0018, + "step": 10941 + }, + { + "epoch": 1.9481837606837606, + "grad_norm": 0.6575125455856323, + "learning_rate": 0.000104239962654868, + "loss": 0.8915, + "step": 10942 + }, + { + "epoch": 1.9483618233618234, + "grad_norm": 0.6315393447875977, + "learning_rate": 0.00010422597772869404, + "loss": 1.1884, + "step": 10943 + }, + { + "epoch": 1.9485398860398861, + "grad_norm": 0.7607148885726929, + "learning_rate": 0.00010421199271972083, + "loss": 0.9341, + "step": 10944 + }, + { + "epoch": 1.9487179487179487, + "grad_norm": 0.6491827964782715, + "learning_rate": 0.00010419800762822239, + "loss": 0.9991, + "step": 10945 + }, + { + "epoch": 1.9488960113960114, + "grad_norm": 0.6294243335723877, + "learning_rate": 0.00010418402245447265, + "loss": 0.9253, + "step": 10946 + }, + { + "epoch": 1.949074074074074, + "grad_norm": 0.6472215056419373, + "learning_rate": 0.00010417003719874571, + "loss": 1.0402, + "step": 10947 + }, + { + "epoch": 1.9492521367521367, + "grad_norm": 0.7377899885177612, + "learning_rate": 0.00010415605186131559, + "loss": 1.046, + "step": 10948 + }, + { + "epoch": 1.9494301994301995, + "grad_norm": 0.6391907334327698, + "learning_rate": 0.00010414206644245623, + "loss": 0.8529, + "step": 10949 + }, + { + "epoch": 1.9496082621082622, + "grad_norm": 0.7101355195045471, + "learning_rate": 0.0001041280809424417, + "loss": 0.925, + "step": 10950 + }, + { + "epoch": 1.9497863247863247, + "grad_norm": 0.7891978025436401, + "learning_rate": 0.00010411409536154597, + "loss": 1.0691, + "step": 10951 + }, + { + "epoch": 1.9499643874643875, + "grad_norm": 0.7225242853164673, + "learning_rate": 0.00010410010970004311, + "loss": 1.158, + "step": 10952 + }, + { + "epoch": 1.95014245014245, + "grad_norm": 0.6073256731033325, + "learning_rate": 0.00010408612395820714, + "loss": 0.9977, + "step": 10953 + }, + { + "epoch": 1.9503205128205128, + "grad_norm": 0.6373769044876099, + "learning_rate": 0.00010407213813631203, + "loss": 1.019, + "step": 10954 + }, + { + "epoch": 1.9504985754985755, + "grad_norm": 0.7451884746551514, + "learning_rate": 0.00010405815223463184, + "loss": 0.9497, + "step": 10955 + }, + { + "epoch": 1.9506766381766383, + "grad_norm": 0.7760418057441711, + "learning_rate": 0.00010404416625344058, + "loss": 1.0378, + "step": 10956 + }, + { + "epoch": 1.9508547008547008, + "grad_norm": 0.7057808041572571, + "learning_rate": 0.00010403018019301228, + "loss": 0.8953, + "step": 10957 + }, + { + "epoch": 1.9510327635327636, + "grad_norm": 0.6599584817886353, + "learning_rate": 0.00010401619405362095, + "loss": 0.8859, + "step": 10958 + }, + { + "epoch": 1.951210826210826, + "grad_norm": 0.6977253556251526, + "learning_rate": 0.00010400220783554069, + "loss": 0.9038, + "step": 10959 + }, + { + "epoch": 1.9513888888888888, + "grad_norm": 0.6930267810821533, + "learning_rate": 0.00010398822153904546, + "loss": 1.1547, + "step": 10960 + }, + { + "epoch": 1.9515669515669516, + "grad_norm": 0.6301694512367249, + "learning_rate": 0.00010397423516440931, + "loss": 0.8875, + "step": 10961 + }, + { + "epoch": 1.9517450142450143, + "grad_norm": 0.7447484135627747, + "learning_rate": 0.00010396024871190628, + "loss": 1.0454, + "step": 10962 + }, + { + "epoch": 1.9519230769230769, + "grad_norm": 0.8666765093803406, + "learning_rate": 0.00010394626218181041, + "loss": 1.2211, + "step": 10963 + }, + { + "epoch": 1.9521011396011396, + "grad_norm": 0.599354088306427, + "learning_rate": 0.00010393227557439573, + "loss": 1.0419, + "step": 10964 + }, + { + "epoch": 1.9522792022792022, + "grad_norm": 0.6991702914237976, + "learning_rate": 0.00010391828888993627, + "loss": 0.8217, + "step": 10965 + }, + { + "epoch": 1.952457264957265, + "grad_norm": 0.7467028498649597, + "learning_rate": 0.0001039043021287061, + "loss": 0.8708, + "step": 10966 + }, + { + "epoch": 1.9526353276353277, + "grad_norm": 0.6806215047836304, + "learning_rate": 0.0001038903152909792, + "loss": 1.218, + "step": 10967 + }, + { + "epoch": 1.9528133903133904, + "grad_norm": 0.6704212427139282, + "learning_rate": 0.00010387632837702968, + "loss": 0.8428, + "step": 10968 + }, + { + "epoch": 1.952991452991453, + "grad_norm": 0.6843154430389404, + "learning_rate": 0.00010386234138713155, + "loss": 0.9729, + "step": 10969 + }, + { + "epoch": 1.9531695156695157, + "grad_norm": 0.6619821190834045, + "learning_rate": 0.00010384835432155888, + "loss": 1.021, + "step": 10970 + }, + { + "epoch": 1.9533475783475782, + "grad_norm": 0.6249803900718689, + "learning_rate": 0.0001038343671805857, + "loss": 0.9321, + "step": 10971 + }, + { + "epoch": 1.953525641025641, + "grad_norm": 0.7361689805984497, + "learning_rate": 0.00010382037996448604, + "loss": 0.9451, + "step": 10972 + }, + { + "epoch": 1.9537037037037037, + "grad_norm": 0.6464847922325134, + "learning_rate": 0.00010380639267353398, + "loss": 1.0188, + "step": 10973 + }, + { + "epoch": 1.9538817663817665, + "grad_norm": 0.5975635647773743, + "learning_rate": 0.00010379240530800356, + "loss": 0.9025, + "step": 10974 + }, + { + "epoch": 1.9540598290598292, + "grad_norm": 0.6734475493431091, + "learning_rate": 0.00010377841786816884, + "loss": 1.0742, + "step": 10975 + }, + { + "epoch": 1.9542378917378918, + "grad_norm": 0.7318592667579651, + "learning_rate": 0.00010376443035430386, + "loss": 1.1082, + "step": 10976 + }, + { + "epoch": 1.9544159544159543, + "grad_norm": 0.7696142792701721, + "learning_rate": 0.00010375044276668271, + "loss": 0.8421, + "step": 10977 + }, + { + "epoch": 1.954594017094017, + "grad_norm": 0.68442302942276, + "learning_rate": 0.00010373645510557939, + "loss": 1.0794, + "step": 10978 + }, + { + "epoch": 1.9547720797720798, + "grad_norm": 0.7582547068595886, + "learning_rate": 0.00010372246737126801, + "loss": 1.0332, + "step": 10979 + }, + { + "epoch": 1.9549501424501425, + "grad_norm": 0.6529998183250427, + "learning_rate": 0.00010370847956402262, + "loss": 1.1833, + "step": 10980 + }, + { + "epoch": 1.9551282051282053, + "grad_norm": 0.7565605044364929, + "learning_rate": 0.00010369449168411729, + "loss": 1.0494, + "step": 10981 + }, + { + "epoch": 1.9553062678062678, + "grad_norm": 0.6346915364265442, + "learning_rate": 0.00010368050373182605, + "loss": 1.0052, + "step": 10982 + }, + { + "epoch": 1.9554843304843303, + "grad_norm": 0.7021830081939697, + "learning_rate": 0.00010366651570742298, + "loss": 0.9716, + "step": 10983 + }, + { + "epoch": 1.955662393162393, + "grad_norm": 0.6464530825614929, + "learning_rate": 0.00010365252761118218, + "loss": 0.9802, + "step": 10984 + }, + { + "epoch": 1.9558404558404558, + "grad_norm": 0.6845090985298157, + "learning_rate": 0.00010363853944337768, + "loss": 0.9529, + "step": 10985 + }, + { + "epoch": 1.9560185185185186, + "grad_norm": 0.7178115248680115, + "learning_rate": 0.00010362455120428356, + "loss": 0.9968, + "step": 10986 + }, + { + "epoch": 1.9561965811965814, + "grad_norm": 0.6131038069725037, + "learning_rate": 0.00010361056289417385, + "loss": 1.0559, + "step": 10987 + }, + { + "epoch": 1.9563746438746439, + "grad_norm": 0.6946909427642822, + "learning_rate": 0.0001035965745133227, + "loss": 1.0457, + "step": 10988 + }, + { + "epoch": 1.9565527065527064, + "grad_norm": 0.7376706600189209, + "learning_rate": 0.00010358258606200413, + "loss": 0.7775, + "step": 10989 + }, + { + "epoch": 1.9567307692307692, + "grad_norm": 0.6864920854568481, + "learning_rate": 0.00010356859754049225, + "loss": 0.8798, + "step": 10990 + }, + { + "epoch": 1.956908831908832, + "grad_norm": 0.6301153302192688, + "learning_rate": 0.0001035546089490611, + "loss": 0.8757, + "step": 10991 + }, + { + "epoch": 1.9570868945868947, + "grad_norm": 0.7184807062149048, + "learning_rate": 0.00010354062028798474, + "loss": 1.0783, + "step": 10992 + }, + { + "epoch": 1.9572649572649574, + "grad_norm": 0.7138563394546509, + "learning_rate": 0.00010352663155753732, + "loss": 1.0328, + "step": 10993 + }, + { + "epoch": 1.95744301994302, + "grad_norm": 0.6565547585487366, + "learning_rate": 0.00010351264275799286, + "loss": 1.1312, + "step": 10994 + }, + { + "epoch": 1.9576210826210825, + "grad_norm": 0.7055862545967102, + "learning_rate": 0.00010349865388962547, + "loss": 1.0787, + "step": 10995 + }, + { + "epoch": 1.9577991452991452, + "grad_norm": 0.6184022426605225, + "learning_rate": 0.00010348466495270926, + "loss": 0.9635, + "step": 10996 + }, + { + "epoch": 1.957977207977208, + "grad_norm": 0.6563652753829956, + "learning_rate": 0.0001034706759475182, + "loss": 0.772, + "step": 10997 + }, + { + "epoch": 1.9581552706552707, + "grad_norm": 0.6103591322898865, + "learning_rate": 0.00010345668687432651, + "loss": 0.8113, + "step": 10998 + }, + { + "epoch": 1.9583333333333335, + "grad_norm": 0.6715512275695801, + "learning_rate": 0.0001034426977334082, + "loss": 1.1841, + "step": 10999 + }, + { + "epoch": 1.958511396011396, + "grad_norm": 0.680092453956604, + "learning_rate": 0.00010342870852503739, + "loss": 0.9992, + "step": 11000 + }, + { + "epoch": 1.9586894586894585, + "grad_norm": 0.828472375869751, + "learning_rate": 0.00010341471924948816, + "loss": 1.0975, + "step": 11001 + }, + { + "epoch": 1.9588675213675213, + "grad_norm": 0.758441686630249, + "learning_rate": 0.00010340072990703463, + "loss": 1.0632, + "step": 11002 + }, + { + "epoch": 1.959045584045584, + "grad_norm": 0.6847560405731201, + "learning_rate": 0.00010338674049795079, + "loss": 1.0054, + "step": 11003 + }, + { + "epoch": 1.9592236467236468, + "grad_norm": 0.707626223564148, + "learning_rate": 0.00010337275102251085, + "loss": 0.9427, + "step": 11004 + }, + { + "epoch": 1.9594017094017095, + "grad_norm": 0.769036591053009, + "learning_rate": 0.00010335876148098887, + "loss": 1.0424, + "step": 11005 + }, + { + "epoch": 1.959579772079772, + "grad_norm": 0.822695791721344, + "learning_rate": 0.00010334477187365892, + "loss": 1.1573, + "step": 11006 + }, + { + "epoch": 1.9597578347578346, + "grad_norm": 0.6290286183357239, + "learning_rate": 0.00010333078220079513, + "loss": 0.936, + "step": 11007 + }, + { + "epoch": 1.9599358974358974, + "grad_norm": 0.6802252531051636, + "learning_rate": 0.00010331679246267155, + "loss": 0.8049, + "step": 11008 + }, + { + "epoch": 1.96011396011396, + "grad_norm": 0.6652607321739197, + "learning_rate": 0.00010330280265956232, + "loss": 0.926, + "step": 11009 + }, + { + "epoch": 1.9602920227920229, + "grad_norm": 0.7057216763496399, + "learning_rate": 0.00010328881279174154, + "loss": 0.9464, + "step": 11010 + }, + { + "epoch": 1.9604700854700856, + "grad_norm": 0.6951601505279541, + "learning_rate": 0.00010327482285948331, + "loss": 0.9882, + "step": 11011 + }, + { + "epoch": 1.9606481481481481, + "grad_norm": 0.6537632942199707, + "learning_rate": 0.00010326083286306174, + "loss": 0.8663, + "step": 11012 + }, + { + "epoch": 1.9608262108262107, + "grad_norm": 0.7252047657966614, + "learning_rate": 0.0001032468428027509, + "loss": 1.1377, + "step": 11013 + }, + { + "epoch": 1.9610042735042734, + "grad_norm": 0.6494104266166687, + "learning_rate": 0.00010323285267882492, + "loss": 0.8072, + "step": 11014 + }, + { + "epoch": 1.9611823361823362, + "grad_norm": 0.8463460206985474, + "learning_rate": 0.00010321886249155792, + "loss": 1.22, + "step": 11015 + }, + { + "epoch": 1.961360398860399, + "grad_norm": 0.6071396470069885, + "learning_rate": 0.00010320487224122401, + "loss": 0.7975, + "step": 11016 + }, + { + "epoch": 1.9615384615384617, + "grad_norm": 0.6546960473060608, + "learning_rate": 0.00010319088192809725, + "loss": 1.1729, + "step": 11017 + }, + { + "epoch": 1.9617165242165242, + "grad_norm": 0.7399442791938782, + "learning_rate": 0.00010317689155245178, + "loss": 1.092, + "step": 11018 + }, + { + "epoch": 1.9618945868945867, + "grad_norm": 0.7103837728500366, + "learning_rate": 0.00010316290111456175, + "loss": 0.8436, + "step": 11019 + }, + { + "epoch": 1.9620726495726495, + "grad_norm": 0.6990065574645996, + "learning_rate": 0.00010314891061470125, + "loss": 0.9003, + "step": 11020 + }, + { + "epoch": 1.9622507122507122, + "grad_norm": 0.7945666313171387, + "learning_rate": 0.00010313492005314438, + "loss": 0.8812, + "step": 11021 + }, + { + "epoch": 1.962428774928775, + "grad_norm": 0.6177538633346558, + "learning_rate": 0.00010312092943016527, + "loss": 1.0091, + "step": 11022 + }, + { + "epoch": 1.9626068376068377, + "grad_norm": 0.7260771989822388, + "learning_rate": 0.000103106938746038, + "loss": 0.9376, + "step": 11023 + }, + { + "epoch": 1.9627849002849003, + "grad_norm": 0.6726518273353577, + "learning_rate": 0.00010309294800103674, + "loss": 0.8048, + "step": 11024 + }, + { + "epoch": 1.9629629629629628, + "grad_norm": 0.8759992122650146, + "learning_rate": 0.00010307895719543562, + "loss": 1.0248, + "step": 11025 + }, + { + "epoch": 1.9631410256410255, + "grad_norm": 0.683437168598175, + "learning_rate": 0.00010306496632950868, + "loss": 1.0314, + "step": 11026 + }, + { + "epoch": 1.9633190883190883, + "grad_norm": 0.7255756258964539, + "learning_rate": 0.00010305097540353012, + "loss": 0.9828, + "step": 11027 + }, + { + "epoch": 1.963497150997151, + "grad_norm": 0.6904804706573486, + "learning_rate": 0.000103036984417774, + "loss": 0.9054, + "step": 11028 + }, + { + "epoch": 1.9636752136752138, + "grad_norm": 0.6906846761703491, + "learning_rate": 0.00010302299337251451, + "loss": 1.0287, + "step": 11029 + }, + { + "epoch": 1.9638532763532763, + "grad_norm": 0.6677078008651733, + "learning_rate": 0.00010300900226802575, + "loss": 0.8742, + "step": 11030 + }, + { + "epoch": 1.964031339031339, + "grad_norm": 0.6144888997077942, + "learning_rate": 0.00010299501110458183, + "loss": 0.6942, + "step": 11031 + }, + { + "epoch": 1.9642094017094016, + "grad_norm": 0.753010094165802, + "learning_rate": 0.0001029810198824569, + "loss": 0.9018, + "step": 11032 + }, + { + "epoch": 1.9643874643874644, + "grad_norm": 0.6872276663780212, + "learning_rate": 0.00010296702860192505, + "loss": 1.1647, + "step": 11033 + }, + { + "epoch": 1.9645655270655271, + "grad_norm": 0.709000289440155, + "learning_rate": 0.00010295303726326047, + "loss": 0.9143, + "step": 11034 + }, + { + "epoch": 1.9647435897435899, + "grad_norm": 0.6507021188735962, + "learning_rate": 0.00010293904586673723, + "loss": 1.006, + "step": 11035 + }, + { + "epoch": 1.9649216524216524, + "grad_norm": 0.6789946556091309, + "learning_rate": 0.00010292505441262952, + "loss": 0.9049, + "step": 11036 + }, + { + "epoch": 1.9650997150997151, + "grad_norm": 0.7156081795692444, + "learning_rate": 0.00010291106290121143, + "loss": 0.9195, + "step": 11037 + }, + { + "epoch": 1.9652777777777777, + "grad_norm": 0.6770932078361511, + "learning_rate": 0.0001028970713327571, + "loss": 0.9524, + "step": 11038 + }, + { + "epoch": 1.9654558404558404, + "grad_norm": 0.7304288148880005, + "learning_rate": 0.00010288307970754067, + "loss": 0.9276, + "step": 11039 + }, + { + "epoch": 1.9656339031339032, + "grad_norm": 0.7603645324707031, + "learning_rate": 0.0001028690880258363, + "loss": 1.2157, + "step": 11040 + }, + { + "epoch": 1.965811965811966, + "grad_norm": 0.6875246167182922, + "learning_rate": 0.00010285509628791811, + "loss": 1.0269, + "step": 11041 + }, + { + "epoch": 1.9659900284900285, + "grad_norm": 0.7234818935394287, + "learning_rate": 0.00010284110449406026, + "loss": 0.9695, + "step": 11042 + }, + { + "epoch": 1.9661680911680912, + "grad_norm": 0.7322804927825928, + "learning_rate": 0.00010282711264453684, + "loss": 0.9752, + "step": 11043 + }, + { + "epoch": 1.9663461538461537, + "grad_norm": 0.7524822950363159, + "learning_rate": 0.00010281312073962202, + "loss": 1.2144, + "step": 11044 + }, + { + "epoch": 1.9665242165242165, + "grad_norm": 0.6623101234436035, + "learning_rate": 0.00010279912877958995, + "loss": 1.1334, + "step": 11045 + }, + { + "epoch": 1.9667022792022792, + "grad_norm": 0.7814893126487732, + "learning_rate": 0.00010278513676471477, + "loss": 1.266, + "step": 11046 + }, + { + "epoch": 1.966880341880342, + "grad_norm": 0.7129884362220764, + "learning_rate": 0.00010277114469527063, + "loss": 1.0918, + "step": 11047 + }, + { + "epoch": 1.9670584045584045, + "grad_norm": 0.6996828317642212, + "learning_rate": 0.00010275715257153164, + "loss": 0.9269, + "step": 11048 + }, + { + "epoch": 1.9672364672364673, + "grad_norm": 0.6439059972763062, + "learning_rate": 0.00010274316039377198, + "loss": 1.1998, + "step": 11049 + }, + { + "epoch": 1.9674145299145298, + "grad_norm": 0.6837672591209412, + "learning_rate": 0.00010272916816226581, + "loss": 0.8899, + "step": 11050 + }, + { + "epoch": 1.9675925925925926, + "grad_norm": 0.702583909034729, + "learning_rate": 0.00010271517587728726, + "loss": 1.1862, + "step": 11051 + }, + { + "epoch": 1.9677706552706553, + "grad_norm": 0.6627798676490784, + "learning_rate": 0.00010270118353911047, + "loss": 0.898, + "step": 11052 + }, + { + "epoch": 1.967948717948718, + "grad_norm": 0.7628579139709473, + "learning_rate": 0.00010268719114800957, + "loss": 1.006, + "step": 11053 + }, + { + "epoch": 1.9681267806267806, + "grad_norm": 0.6425395607948303, + "learning_rate": 0.00010267319870425877, + "loss": 0.962, + "step": 11054 + }, + { + "epoch": 1.9683048433048433, + "grad_norm": 0.7462666630744934, + "learning_rate": 0.00010265920620813219, + "loss": 1.0703, + "step": 11055 + }, + { + "epoch": 1.9684829059829059, + "grad_norm": 0.67641681432724, + "learning_rate": 0.00010264521365990401, + "loss": 1.1077, + "step": 11056 + }, + { + "epoch": 1.9686609686609686, + "grad_norm": 0.6716381311416626, + "learning_rate": 0.0001026312210598483, + "loss": 1.1048, + "step": 11057 + }, + { + "epoch": 1.9688390313390314, + "grad_norm": 0.7207448482513428, + "learning_rate": 0.00010261722840823935, + "loss": 0.9236, + "step": 11058 + }, + { + "epoch": 1.9690170940170941, + "grad_norm": 0.7208544015884399, + "learning_rate": 0.0001026032357053512, + "loss": 1.0814, + "step": 11059 + }, + { + "epoch": 1.9691951566951567, + "grad_norm": 0.6076363325119019, + "learning_rate": 0.00010258924295145807, + "loss": 0.9388, + "step": 11060 + }, + { + "epoch": 1.9693732193732194, + "grad_norm": 0.6460439562797546, + "learning_rate": 0.00010257525014683411, + "loss": 0.9506, + "step": 11061 + }, + { + "epoch": 1.969551282051282, + "grad_norm": 0.7449939250946045, + "learning_rate": 0.00010256125729175348, + "loss": 1.0209, + "step": 11062 + }, + { + "epoch": 1.9697293447293447, + "grad_norm": 0.640885055065155, + "learning_rate": 0.00010254726438649031, + "loss": 1.0235, + "step": 11063 + }, + { + "epoch": 1.9699074074074074, + "grad_norm": 0.6872261166572571, + "learning_rate": 0.00010253327143131879, + "loss": 0.9217, + "step": 11064 + }, + { + "epoch": 1.9700854700854702, + "grad_norm": 0.6213285326957703, + "learning_rate": 0.0001025192784265131, + "loss": 0.8204, + "step": 11065 + }, + { + "epoch": 1.9702635327635327, + "grad_norm": 0.6594449281692505, + "learning_rate": 0.00010250528537234736, + "loss": 0.9789, + "step": 11066 + }, + { + "epoch": 1.9704415954415955, + "grad_norm": 0.7098729610443115, + "learning_rate": 0.00010249129226909577, + "loss": 1.2551, + "step": 11067 + }, + { + "epoch": 1.970619658119658, + "grad_norm": 0.7455953359603882, + "learning_rate": 0.0001024772991170325, + "loss": 1.0281, + "step": 11068 + }, + { + "epoch": 1.9707977207977208, + "grad_norm": 0.6657416224479675, + "learning_rate": 0.00010246330591643166, + "loss": 0.9421, + "step": 11069 + }, + { + "epoch": 1.9709757834757835, + "grad_norm": 0.6480659246444702, + "learning_rate": 0.00010244931266756748, + "loss": 0.9424, + "step": 11070 + }, + { + "epoch": 1.9711538461538463, + "grad_norm": 0.6440510749816895, + "learning_rate": 0.00010243531937071411, + "loss": 0.9651, + "step": 11071 + }, + { + "epoch": 1.9713319088319088, + "grad_norm": 0.6329794526100159, + "learning_rate": 0.00010242132602614571, + "loss": 0.9233, + "step": 11072 + }, + { + "epoch": 1.9715099715099715, + "grad_norm": 0.6694819927215576, + "learning_rate": 0.00010240733263413646, + "loss": 0.884, + "step": 11073 + }, + { + "epoch": 1.971688034188034, + "grad_norm": 0.7702556848526001, + "learning_rate": 0.0001023933391949605, + "loss": 1.216, + "step": 11074 + }, + { + "epoch": 1.9718660968660968, + "grad_norm": 0.6587536931037903, + "learning_rate": 0.00010237934570889207, + "loss": 0.9324, + "step": 11075 + }, + { + "epoch": 1.9720441595441596, + "grad_norm": 0.7919837832450867, + "learning_rate": 0.00010236535217620529, + "loss": 1.0011, + "step": 11076 + }, + { + "epoch": 1.9722222222222223, + "grad_norm": 0.6604606509208679, + "learning_rate": 0.00010235135859717433, + "loss": 0.929, + "step": 11077 + }, + { + "epoch": 1.9724002849002849, + "grad_norm": 0.7158446907997131, + "learning_rate": 0.0001023373649720734, + "loss": 0.8912, + "step": 11078 + }, + { + "epoch": 1.9725783475783476, + "grad_norm": 0.7450904846191406, + "learning_rate": 0.00010232337130117666, + "loss": 1.0782, + "step": 11079 + }, + { + "epoch": 1.9727564102564101, + "grad_norm": 0.6687077283859253, + "learning_rate": 0.00010230937758475827, + "loss": 1.0662, + "step": 11080 + }, + { + "epoch": 1.9729344729344729, + "grad_norm": 0.7188364267349243, + "learning_rate": 0.00010229538382309245, + "loss": 1.024, + "step": 11081 + }, + { + "epoch": 1.9731125356125356, + "grad_norm": 0.6787814497947693, + "learning_rate": 0.00010228139001645334, + "loss": 0.9559, + "step": 11082 + }, + { + "epoch": 1.9732905982905984, + "grad_norm": 0.6834072470664978, + "learning_rate": 0.00010226739616511513, + "loss": 0.8143, + "step": 11083 + }, + { + "epoch": 1.973468660968661, + "grad_norm": 0.6651090979576111, + "learning_rate": 0.00010225340226935201, + "loss": 1.05, + "step": 11084 + }, + { + "epoch": 1.9736467236467237, + "grad_norm": 0.7125018835067749, + "learning_rate": 0.00010223940832943813, + "loss": 1.0275, + "step": 11085 + }, + { + "epoch": 1.9738247863247862, + "grad_norm": 0.6886870861053467, + "learning_rate": 0.00010222541434564772, + "loss": 1.0972, + "step": 11086 + }, + { + "epoch": 1.974002849002849, + "grad_norm": 0.7068913578987122, + "learning_rate": 0.00010221142031825492, + "loss": 0.9248, + "step": 11087 + }, + { + "epoch": 1.9741809116809117, + "grad_norm": 0.7752319574356079, + "learning_rate": 0.00010219742624753397, + "loss": 0.9754, + "step": 11088 + }, + { + "epoch": 1.9743589743589745, + "grad_norm": 0.7915459871292114, + "learning_rate": 0.00010218343213375896, + "loss": 1.2589, + "step": 11089 + }, + { + "epoch": 1.9745370370370372, + "grad_norm": 0.6597068309783936, + "learning_rate": 0.00010216943797720418, + "loss": 1.0004, + "step": 11090 + }, + { + "epoch": 1.9747150997150997, + "grad_norm": 0.7060620188713074, + "learning_rate": 0.00010215544377814375, + "loss": 0.9968, + "step": 11091 + }, + { + "epoch": 1.9748931623931623, + "grad_norm": 0.6815677881240845, + "learning_rate": 0.0001021414495368519, + "loss": 0.8889, + "step": 11092 + }, + { + "epoch": 1.975071225071225, + "grad_norm": 0.6872935891151428, + "learning_rate": 0.00010212745525360277, + "loss": 1.1582, + "step": 11093 + }, + { + "epoch": 1.9752492877492878, + "grad_norm": 0.6781140565872192, + "learning_rate": 0.00010211346092867056, + "loss": 0.9988, + "step": 11094 + }, + { + "epoch": 1.9754273504273505, + "grad_norm": 0.6959224343299866, + "learning_rate": 0.00010209946656232949, + "loss": 1.1097, + "step": 11095 + }, + { + "epoch": 1.9756054131054133, + "grad_norm": 0.7205058336257935, + "learning_rate": 0.00010208547215485376, + "loss": 0.9951, + "step": 11096 + }, + { + "epoch": 1.9757834757834758, + "grad_norm": 0.6968751549720764, + "learning_rate": 0.00010207147770651748, + "loss": 0.9313, + "step": 11097 + }, + { + "epoch": 1.9759615384615383, + "grad_norm": 0.6688823103904724, + "learning_rate": 0.00010205748321759494, + "loss": 0.9439, + "step": 11098 + }, + { + "epoch": 1.976139601139601, + "grad_norm": 0.6169568300247192, + "learning_rate": 0.00010204348868836028, + "loss": 1.123, + "step": 11099 + }, + { + "epoch": 1.9763176638176638, + "grad_norm": 0.6995537281036377, + "learning_rate": 0.00010202949411908768, + "loss": 1.1928, + "step": 11100 + }, + { + "epoch": 1.9764957264957266, + "grad_norm": 0.7102637887001038, + "learning_rate": 0.00010201549951005138, + "loss": 1.0265, + "step": 11101 + }, + { + "epoch": 1.9766737891737893, + "grad_norm": 0.6820045113563538, + "learning_rate": 0.00010200150486152558, + "loss": 0.9309, + "step": 11102 + }, + { + "epoch": 1.9768518518518519, + "grad_norm": 0.7050938010215759, + "learning_rate": 0.00010198751017378443, + "loss": 1.0047, + "step": 11103 + }, + { + "epoch": 1.9770299145299144, + "grad_norm": 0.6418201923370361, + "learning_rate": 0.00010197351544710214, + "loss": 1.1172, + "step": 11104 + }, + { + "epoch": 1.9772079772079771, + "grad_norm": 0.6681215763092041, + "learning_rate": 0.0001019595206817529, + "loss": 1.0621, + "step": 11105 + }, + { + "epoch": 1.97738603988604, + "grad_norm": 0.7725709676742554, + "learning_rate": 0.00010194552587801094, + "loss": 1.0044, + "step": 11106 + }, + { + "epoch": 1.9775641025641026, + "grad_norm": 0.6870455741882324, + "learning_rate": 0.00010193153103615045, + "loss": 1.2652, + "step": 11107 + }, + { + "epoch": 1.9777421652421654, + "grad_norm": 0.6352108120918274, + "learning_rate": 0.00010191753615644561, + "loss": 1.1081, + "step": 11108 + }, + { + "epoch": 1.977920227920228, + "grad_norm": 0.7322626113891602, + "learning_rate": 0.00010190354123917066, + "loss": 1.0003, + "step": 11109 + }, + { + "epoch": 1.9780982905982905, + "grad_norm": 0.6240935921669006, + "learning_rate": 0.00010188954628459972, + "loss": 0.8925, + "step": 11110 + }, + { + "epoch": 1.9782763532763532, + "grad_norm": 0.6648945212364197, + "learning_rate": 0.00010187555129300708, + "loss": 1.0882, + "step": 11111 + }, + { + "epoch": 1.978454415954416, + "grad_norm": 0.6704208850860596, + "learning_rate": 0.00010186155626466692, + "loss": 0.8873, + "step": 11112 + }, + { + "epoch": 1.9786324786324787, + "grad_norm": 0.6716459393501282, + "learning_rate": 0.00010184756119985341, + "loss": 1.0045, + "step": 11113 + }, + { + "epoch": 1.9788105413105415, + "grad_norm": 0.81277996301651, + "learning_rate": 0.0001018335660988408, + "loss": 0.8867, + "step": 11114 + }, + { + "epoch": 1.978988603988604, + "grad_norm": 0.7008311748504639, + "learning_rate": 0.00010181957096190323, + "loss": 0.9391, + "step": 11115 + }, + { + "epoch": 1.9791666666666665, + "grad_norm": 0.727676272392273, + "learning_rate": 0.00010180557578931498, + "loss": 1.0157, + "step": 11116 + }, + { + "epoch": 1.9793447293447293, + "grad_norm": 0.7058015465736389, + "learning_rate": 0.00010179158058135018, + "loss": 1.0, + "step": 11117 + }, + { + "epoch": 1.979522792022792, + "grad_norm": 0.7770412564277649, + "learning_rate": 0.00010177758533828312, + "loss": 1.0428, + "step": 11118 + }, + { + "epoch": 1.9797008547008548, + "grad_norm": 0.6557414531707764, + "learning_rate": 0.00010176359006038798, + "loss": 0.8557, + "step": 11119 + }, + { + "epoch": 1.9798789173789175, + "grad_norm": 0.7681090235710144, + "learning_rate": 0.00010174959474793894, + "loss": 0.867, + "step": 11120 + }, + { + "epoch": 1.98005698005698, + "grad_norm": 0.7915860414505005, + "learning_rate": 0.0001017355994012102, + "loss": 0.9961, + "step": 11121 + }, + { + "epoch": 1.9802350427350426, + "grad_norm": 0.8039166927337646, + "learning_rate": 0.00010172160402047604, + "loss": 1.1378, + "step": 11122 + }, + { + "epoch": 1.9804131054131053, + "grad_norm": 0.6641189455986023, + "learning_rate": 0.0001017076086060106, + "loss": 0.8914, + "step": 11123 + }, + { + "epoch": 1.980591168091168, + "grad_norm": 0.7673811316490173, + "learning_rate": 0.00010169361315808812, + "loss": 1.018, + "step": 11124 + }, + { + "epoch": 1.9807692307692308, + "grad_norm": 0.7320558428764343, + "learning_rate": 0.00010167961767698279, + "loss": 1.0515, + "step": 11125 + }, + { + "epoch": 1.9809472934472936, + "grad_norm": 0.5717357993125916, + "learning_rate": 0.00010166562216296886, + "loss": 0.7619, + "step": 11126 + }, + { + "epoch": 1.9811253561253561, + "grad_norm": 0.6638465523719788, + "learning_rate": 0.00010165162661632052, + "loss": 1.0161, + "step": 11127 + }, + { + "epoch": 1.9813034188034186, + "grad_norm": 0.7293243408203125, + "learning_rate": 0.00010163763103731201, + "loss": 1.063, + "step": 11128 + }, + { + "epoch": 1.9814814814814814, + "grad_norm": 0.634694516658783, + "learning_rate": 0.00010162363542621752, + "loss": 0.8945, + "step": 11129 + }, + { + "epoch": 1.9816595441595442, + "grad_norm": 0.7086902856826782, + "learning_rate": 0.00010160963978331122, + "loss": 1.0542, + "step": 11130 + }, + { + "epoch": 1.981837606837607, + "grad_norm": 0.5939825773239136, + "learning_rate": 0.00010159564410886742, + "loss": 0.7822, + "step": 11131 + }, + { + "epoch": 1.9820156695156697, + "grad_norm": 0.722183346748352, + "learning_rate": 0.00010158164840316027, + "loss": 1.0252, + "step": 11132 + }, + { + "epoch": 1.9821937321937322, + "grad_norm": 0.7300103306770325, + "learning_rate": 0.000101567652666464, + "loss": 0.9099, + "step": 11133 + }, + { + "epoch": 1.9823717948717947, + "grad_norm": 0.7148736119270325, + "learning_rate": 0.00010155365689905285, + "loss": 1.0149, + "step": 11134 + }, + { + "epoch": 1.9825498575498575, + "grad_norm": 0.8214462995529175, + "learning_rate": 0.000101539661101201, + "loss": 1.0127, + "step": 11135 + }, + { + "epoch": 1.9827279202279202, + "grad_norm": 0.7111126780509949, + "learning_rate": 0.00010152566527318265, + "loss": 1.045, + "step": 11136 + }, + { + "epoch": 1.982905982905983, + "grad_norm": 0.6640021800994873, + "learning_rate": 0.00010151166941527213, + "loss": 0.9618, + "step": 11137 + }, + { + "epoch": 1.9830840455840457, + "grad_norm": 0.7177722454071045, + "learning_rate": 0.00010149767352774358, + "loss": 1.0373, + "step": 11138 + }, + { + "epoch": 1.9832621082621082, + "grad_norm": 0.6728883981704712, + "learning_rate": 0.00010148367761087121, + "loss": 0.9886, + "step": 11139 + }, + { + "epoch": 1.9834401709401708, + "grad_norm": 0.7060428857803345, + "learning_rate": 0.00010146968166492926, + "loss": 1.042, + "step": 11140 + }, + { + "epoch": 1.9836182336182335, + "grad_norm": 0.706253707408905, + "learning_rate": 0.00010145568569019192, + "loss": 1.2249, + "step": 11141 + }, + { + "epoch": 1.9837962962962963, + "grad_norm": 0.618221640586853, + "learning_rate": 0.00010144168968693348, + "loss": 0.9223, + "step": 11142 + }, + { + "epoch": 1.983974358974359, + "grad_norm": 0.7005748748779297, + "learning_rate": 0.00010142769365542814, + "loss": 1.2735, + "step": 11143 + }, + { + "epoch": 1.9841524216524218, + "grad_norm": 0.6059799194335938, + "learning_rate": 0.0001014136975959501, + "loss": 0.7216, + "step": 11144 + }, + { + "epoch": 1.9843304843304843, + "grad_norm": 0.7169116735458374, + "learning_rate": 0.00010139970150877358, + "loss": 0.9541, + "step": 11145 + }, + { + "epoch": 1.984508547008547, + "grad_norm": 0.7402058839797974, + "learning_rate": 0.00010138570539417281, + "loss": 1.1268, + "step": 11146 + }, + { + "epoch": 1.9846866096866096, + "grad_norm": 0.7204117178916931, + "learning_rate": 0.00010137170925242201, + "loss": 1.1557, + "step": 11147 + }, + { + "epoch": 1.9848646723646723, + "grad_norm": 0.589163064956665, + "learning_rate": 0.00010135771308379545, + "loss": 0.9863, + "step": 11148 + }, + { + "epoch": 1.985042735042735, + "grad_norm": 0.6342785358428955, + "learning_rate": 0.00010134371688856732, + "loss": 0.9294, + "step": 11149 + }, + { + "epoch": 1.9852207977207978, + "grad_norm": 0.7144256234169006, + "learning_rate": 0.00010132972066701183, + "loss": 0.9428, + "step": 11150 + }, + { + "epoch": 1.9853988603988604, + "grad_norm": 0.658032238483429, + "learning_rate": 0.00010131572441940322, + "loss": 0.9749, + "step": 11151 + }, + { + "epoch": 1.9855769230769231, + "grad_norm": 0.7609163522720337, + "learning_rate": 0.00010130172814601576, + "loss": 1.1771, + "step": 11152 + }, + { + "epoch": 1.9857549857549857, + "grad_norm": 0.6531760692596436, + "learning_rate": 0.00010128773184712361, + "loss": 0.8529, + "step": 11153 + }, + { + "epoch": 1.9859330484330484, + "grad_norm": 0.6983599066734314, + "learning_rate": 0.00010127373552300103, + "loss": 1.0307, + "step": 11154 + }, + { + "epoch": 1.9861111111111112, + "grad_norm": 0.7121559381484985, + "learning_rate": 0.00010125973917392224, + "loss": 0.9426, + "step": 11155 + }, + { + "epoch": 1.986289173789174, + "grad_norm": 0.6282170414924622, + "learning_rate": 0.0001012457428001615, + "loss": 0.8983, + "step": 11156 + }, + { + "epoch": 1.9864672364672364, + "grad_norm": 0.6960387825965881, + "learning_rate": 0.000101231746401993, + "loss": 0.9001, + "step": 11157 + }, + { + "epoch": 1.9866452991452992, + "grad_norm": 0.7523152232170105, + "learning_rate": 0.000101217749979691, + "loss": 1.3462, + "step": 11158 + }, + { + "epoch": 1.9868233618233617, + "grad_norm": 0.71713787317276, + "learning_rate": 0.00010120375353352971, + "loss": 1.0147, + "step": 11159 + }, + { + "epoch": 1.9870014245014245, + "grad_norm": 0.7304390072822571, + "learning_rate": 0.00010118975706378339, + "loss": 0.8436, + "step": 11160 + }, + { + "epoch": 1.9871794871794872, + "grad_norm": 0.789968729019165, + "learning_rate": 0.00010117576057072622, + "loss": 1.1162, + "step": 11161 + }, + { + "epoch": 1.98735754985755, + "grad_norm": 0.6752170920372009, + "learning_rate": 0.00010116176405463249, + "loss": 1.0619, + "step": 11162 + }, + { + "epoch": 1.9875356125356125, + "grad_norm": 0.681398868560791, + "learning_rate": 0.0001011477675157764, + "loss": 0.8981, + "step": 11163 + }, + { + "epoch": 1.9877136752136753, + "grad_norm": 0.61469566822052, + "learning_rate": 0.0001011337709544322, + "loss": 1.0139, + "step": 11164 + }, + { + "epoch": 1.9878917378917378, + "grad_norm": 0.7524265050888062, + "learning_rate": 0.0001011197743708741, + "loss": 1.1571, + "step": 11165 + }, + { + "epoch": 1.9880698005698005, + "grad_norm": 0.6289594173431396, + "learning_rate": 0.00010110577776537633, + "loss": 0.93, + "step": 11166 + }, + { + "epoch": 1.9882478632478633, + "grad_norm": 0.6991903781890869, + "learning_rate": 0.00010109178113821318, + "loss": 1.1176, + "step": 11167 + }, + { + "epoch": 1.988425925925926, + "grad_norm": 0.7604053020477295, + "learning_rate": 0.00010107778448965883, + "loss": 1.0497, + "step": 11168 + }, + { + "epoch": 1.9886039886039886, + "grad_norm": 0.7166453003883362, + "learning_rate": 0.00010106378781998753, + "loss": 1.1237, + "step": 11169 + }, + { + "epoch": 1.9887820512820513, + "grad_norm": 0.6071686744689941, + "learning_rate": 0.00010104979112947352, + "loss": 0.8934, + "step": 11170 + }, + { + "epoch": 1.9889601139601139, + "grad_norm": 0.6618169546127319, + "learning_rate": 0.00010103579441839101, + "loss": 1.0596, + "step": 11171 + }, + { + "epoch": 1.9891381766381766, + "grad_norm": 0.6838458776473999, + "learning_rate": 0.0001010217976870143, + "loss": 1.0167, + "step": 11172 + }, + { + "epoch": 1.9893162393162394, + "grad_norm": 0.6369979381561279, + "learning_rate": 0.00010100780093561757, + "loss": 0.9001, + "step": 11173 + }, + { + "epoch": 1.989494301994302, + "grad_norm": 0.661313533782959, + "learning_rate": 0.00010099380416447508, + "loss": 0.8952, + "step": 11174 + }, + { + "epoch": 1.9896723646723646, + "grad_norm": 0.6991600394248962, + "learning_rate": 0.00010097980737386106, + "loss": 1.0083, + "step": 11175 + }, + { + "epoch": 1.9898504273504274, + "grad_norm": 0.618748664855957, + "learning_rate": 0.00010096581056404972, + "loss": 0.8797, + "step": 11176 + }, + { + "epoch": 1.99002849002849, + "grad_norm": 0.7039223909378052, + "learning_rate": 0.00010095181373531535, + "loss": 1.0385, + "step": 11177 + }, + { + "epoch": 1.9902065527065527, + "grad_norm": 0.7598999738693237, + "learning_rate": 0.00010093781688793216, + "loss": 0.9205, + "step": 11178 + }, + { + "epoch": 1.9903846153846154, + "grad_norm": 0.6355955600738525, + "learning_rate": 0.00010092382002217441, + "loss": 0.8646, + "step": 11179 + }, + { + "epoch": 1.9905626780626782, + "grad_norm": 0.8024569153785706, + "learning_rate": 0.00010090982313831634, + "loss": 1.1678, + "step": 11180 + }, + { + "epoch": 1.9907407407407407, + "grad_norm": 0.5960529446601868, + "learning_rate": 0.00010089582623663216, + "loss": 0.8277, + "step": 11181 + }, + { + "epoch": 1.9909188034188035, + "grad_norm": 0.6323728561401367, + "learning_rate": 0.00010088182931739609, + "loss": 0.948, + "step": 11182 + }, + { + "epoch": 1.991096866096866, + "grad_norm": 0.7532381415367126, + "learning_rate": 0.00010086783238088244, + "loss": 1.2948, + "step": 11183 + }, + { + "epoch": 1.9912749287749287, + "grad_norm": 0.5740166306495667, + "learning_rate": 0.00010085383542736543, + "loss": 0.7019, + "step": 11184 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 0.616985559463501, + "learning_rate": 0.00010083983845711929, + "loss": 1.0802, + "step": 11185 + }, + { + "epoch": 1.9916310541310542, + "grad_norm": 0.7505929470062256, + "learning_rate": 0.00010082584147041824, + "loss": 1.0523, + "step": 11186 + }, + { + "epoch": 1.9918091168091168, + "grad_norm": 0.7147656679153442, + "learning_rate": 0.00010081184446753653, + "loss": 1.0019, + "step": 11187 + }, + { + "epoch": 1.9919871794871795, + "grad_norm": 0.7301992774009705, + "learning_rate": 0.00010079784744874845, + "loss": 1.0329, + "step": 11188 + }, + { + "epoch": 1.992165242165242, + "grad_norm": 0.6847206354141235, + "learning_rate": 0.00010078385041432819, + "loss": 1.0367, + "step": 11189 + }, + { + "epoch": 1.9923433048433048, + "grad_norm": 0.7310990691184998, + "learning_rate": 0.00010076985336455, + "loss": 1.1675, + "step": 11190 + }, + { + "epoch": 1.9925213675213675, + "grad_norm": 0.6916858553886414, + "learning_rate": 0.00010075585629968813, + "loss": 0.8615, + "step": 11191 + }, + { + "epoch": 1.9926994301994303, + "grad_norm": 0.6519390344619751, + "learning_rate": 0.00010074185922001685, + "loss": 0.8105, + "step": 11192 + }, + { + "epoch": 1.9928774928774928, + "grad_norm": 0.7437400817871094, + "learning_rate": 0.00010072786212581036, + "loss": 0.9993, + "step": 11193 + }, + { + "epoch": 1.9930555555555556, + "grad_norm": 0.5048928260803223, + "learning_rate": 0.00010071386501734292, + "loss": 0.7912, + "step": 11194 + }, + { + "epoch": 1.993233618233618, + "grad_norm": 0.8042343258857727, + "learning_rate": 0.00010069986789488882, + "loss": 0.9156, + "step": 11195 + }, + { + "epoch": 1.9934116809116809, + "grad_norm": 0.7188669443130493, + "learning_rate": 0.0001006858707587222, + "loss": 1.0474, + "step": 11196 + }, + { + "epoch": 1.9935897435897436, + "grad_norm": 0.7377660870552063, + "learning_rate": 0.00010067187360911738, + "loss": 0.7013, + "step": 11197 + }, + { + "epoch": 1.9937678062678064, + "grad_norm": 0.6684696078300476, + "learning_rate": 0.00010065787644634861, + "loss": 0.9199, + "step": 11198 + }, + { + "epoch": 1.993945868945869, + "grad_norm": 0.7341524958610535, + "learning_rate": 0.00010064387927069012, + "loss": 1.0925, + "step": 11199 + }, + { + "epoch": 1.9941239316239316, + "grad_norm": 0.685745120048523, + "learning_rate": 0.00010062988208241614, + "loss": 1.083, + "step": 11200 + }, + { + "epoch": 1.9943019943019942, + "grad_norm": 0.6923556327819824, + "learning_rate": 0.00010061588488180096, + "loss": 1.2728, + "step": 11201 + }, + { + "epoch": 1.994480056980057, + "grad_norm": 0.6663293242454529, + "learning_rate": 0.00010060188766911876, + "loss": 1.0937, + "step": 11202 + }, + { + "epoch": 1.9946581196581197, + "grad_norm": 0.7963639497756958, + "learning_rate": 0.00010058789044464383, + "loss": 1.0592, + "step": 11203 + }, + { + "epoch": 1.9948361823361824, + "grad_norm": 0.6362990140914917, + "learning_rate": 0.00010057389320865042, + "loss": 0.8872, + "step": 11204 + }, + { + "epoch": 1.9950142450142452, + "grad_norm": 0.7752974033355713, + "learning_rate": 0.00010055989596141278, + "loss": 1.043, + "step": 11205 + }, + { + "epoch": 1.9951923076923077, + "grad_norm": 0.7125133275985718, + "learning_rate": 0.00010054589870320512, + "loss": 1.0015, + "step": 11206 + }, + { + "epoch": 1.9953703703703702, + "grad_norm": 0.7102736830711365, + "learning_rate": 0.00010053190143430169, + "loss": 1.0052, + "step": 11207 + }, + { + "epoch": 1.995548433048433, + "grad_norm": 0.8628628849983215, + "learning_rate": 0.00010051790415497677, + "loss": 1.2351, + "step": 11208 + }, + { + "epoch": 1.9957264957264957, + "grad_norm": 0.7233129739761353, + "learning_rate": 0.00010050390686550462, + "loss": 1.0848, + "step": 11209 + }, + { + "epoch": 1.9959045584045585, + "grad_norm": 0.5936228036880493, + "learning_rate": 0.00010048990956615944, + "loss": 0.7998, + "step": 11210 + }, + { + "epoch": 1.9960826210826212, + "grad_norm": 0.7345388531684875, + "learning_rate": 0.0001004759122572155, + "loss": 1.0329, + "step": 11211 + }, + { + "epoch": 1.9962606837606838, + "grad_norm": 0.7344130873680115, + "learning_rate": 0.00010046191493894703, + "loss": 1.1563, + "step": 11212 + }, + { + "epoch": 1.9964387464387463, + "grad_norm": 0.6979942321777344, + "learning_rate": 0.00010044791761162833, + "loss": 0.9269, + "step": 11213 + }, + { + "epoch": 1.996616809116809, + "grad_norm": 0.67514967918396, + "learning_rate": 0.0001004339202755336, + "loss": 0.9028, + "step": 11214 + }, + { + "epoch": 1.9967948717948718, + "grad_norm": 0.6379111409187317, + "learning_rate": 0.00010041992293093712, + "loss": 0.7816, + "step": 11215 + }, + { + "epoch": 1.9969729344729346, + "grad_norm": 0.693976104259491, + "learning_rate": 0.00010040592557811308, + "loss": 0.8411, + "step": 11216 + }, + { + "epoch": 1.9971509971509973, + "grad_norm": 0.5952646732330322, + "learning_rate": 0.0001003919282173358, + "loss": 0.8681, + "step": 11217 + }, + { + "epoch": 1.9973290598290598, + "grad_norm": 0.7452160716056824, + "learning_rate": 0.00010037793084887948, + "loss": 1.0198, + "step": 11218 + }, + { + "epoch": 1.9975071225071224, + "grad_norm": 0.6683938503265381, + "learning_rate": 0.00010036393347301841, + "loss": 0.8162, + "step": 11219 + }, + { + "epoch": 1.9976851851851851, + "grad_norm": 0.6849120855331421, + "learning_rate": 0.00010034993609002683, + "loss": 1.0668, + "step": 11220 + }, + { + "epoch": 1.9978632478632479, + "grad_norm": 0.8782517910003662, + "learning_rate": 0.00010033593870017897, + "loss": 1.222, + "step": 11221 + }, + { + "epoch": 1.9980413105413106, + "grad_norm": 0.6482772827148438, + "learning_rate": 0.00010032194130374908, + "loss": 0.7722, + "step": 11222 + }, + { + "epoch": 1.9982193732193734, + "grad_norm": 0.8595399260520935, + "learning_rate": 0.00010030794390101142, + "loss": 1.3004, + "step": 11223 + }, + { + "epoch": 1.998397435897436, + "grad_norm": 0.7258931994438171, + "learning_rate": 0.00010029394649224024, + "loss": 0.8825, + "step": 11224 + }, + { + "epoch": 1.9985754985754984, + "grad_norm": 0.6291348934173584, + "learning_rate": 0.00010027994907770981, + "loss": 0.8681, + "step": 11225 + }, + { + "epoch": 1.9987535612535612, + "grad_norm": 0.7528844475746155, + "learning_rate": 0.00010026595165769434, + "loss": 1.1443, + "step": 11226 + }, + { + "epoch": 1.998931623931624, + "grad_norm": 0.654017984867096, + "learning_rate": 0.0001002519542324681, + "loss": 0.8585, + "step": 11227 + }, + { + "epoch": 1.9991096866096867, + "grad_norm": 0.6812533736228943, + "learning_rate": 0.00010023795680230532, + "loss": 0.8757, + "step": 11228 + }, + { + "epoch": 1.9992877492877494, + "grad_norm": 0.7120179533958435, + "learning_rate": 0.0001002239593674803, + "loss": 1.0159, + "step": 11229 + }, + { + "epoch": 1.999465811965812, + "grad_norm": 0.6943802237510681, + "learning_rate": 0.00010020996192826725, + "loss": 1.0193, + "step": 11230 + }, + { + "epoch": 1.9996438746438745, + "grad_norm": 0.7227906584739685, + "learning_rate": 0.00010019596448494047, + "loss": 1.1536, + "step": 11231 + }, + { + "epoch": 1.9998219373219372, + "grad_norm": 0.6233312487602234, + "learning_rate": 0.00010018196703777411, + "loss": 0.9117, + "step": 11232 + }, + { + "epoch": 1.9998219373219372, + "eval_loss": 1.0963108539581299, + "eval_runtime": 24.4478, + "eval_samples_per_second": 42.58, + "eval_steps_per_second": 21.311, + "step": 11232 + }, + { + "epoch": 2.0, + "grad_norm": 0.67911696434021, + "learning_rate": 0.00010016796958704254, + "loss": 0.9516, + "step": 11233 + }, + { + "epoch": 2.0001780626780628, + "grad_norm": 0.7372198700904846, + "learning_rate": 0.00010015397213301992, + "loss": 1.3066, + "step": 11234 + }, + { + "epoch": 2.0001780626780628, + "grad_norm": 0.7573498487472534, + "learning_rate": 0.00010013997467598055, + "loss": 1.0204, + "step": 11235 + }, + { + "epoch": 2.0003561253561255, + "grad_norm": 0.6862889528274536, + "learning_rate": 0.00010012597721619863, + "loss": 0.9447, + "step": 11236 + }, + { + "epoch": 2.0005341880341883, + "grad_norm": 0.5977628231048584, + "learning_rate": 0.00010011197975394851, + "loss": 0.9849, + "step": 11237 + }, + { + "epoch": 2.0007122507122506, + "grad_norm": 0.620206892490387, + "learning_rate": 0.00010009798228950431, + "loss": 0.7498, + "step": 11238 + }, + { + "epoch": 2.0008903133903133, + "grad_norm": 0.5694536566734314, + "learning_rate": 0.0001000839848231404, + "loss": 0.7092, + "step": 11239 + }, + { + "epoch": 2.001068376068376, + "grad_norm": 0.5880212783813477, + "learning_rate": 0.00010006998735513098, + "loss": 0.9057, + "step": 11240 + }, + { + "epoch": 2.001246438746439, + "grad_norm": 0.6152323484420776, + "learning_rate": 0.00010005598988575029, + "loss": 0.8356, + "step": 11241 + }, + { + "epoch": 2.0014245014245016, + "grad_norm": 0.6827659010887146, + "learning_rate": 0.00010004199241527261, + "loss": 0.8302, + "step": 11242 + }, + { + "epoch": 2.0016025641025643, + "grad_norm": 0.5883491635322571, + "learning_rate": 0.00010002799494397215, + "loss": 0.8616, + "step": 11243 + }, + { + "epoch": 2.0017806267806266, + "grad_norm": 0.7649462819099426, + "learning_rate": 0.00010001399747212322, + "loss": 1.1643, + "step": 11244 + }, + { + "epoch": 2.0019586894586894, + "grad_norm": 0.6435316205024719, + "learning_rate": 0.0001, + "loss": 0.9239, + "step": 11245 + }, + { + "epoch": 2.002136752136752, + "grad_norm": 0.5662951469421387, + "learning_rate": 9.99860025278768e-05, + "loss": 0.9111, + "step": 11246 + }, + { + "epoch": 2.002314814814815, + "grad_norm": 0.6234064102172852, + "learning_rate": 9.997200505602787e-05, + "loss": 0.4852, + "step": 11247 + }, + { + "epoch": 2.0024928774928776, + "grad_norm": 0.6322146058082581, + "learning_rate": 9.995800758472741e-05, + "loss": 0.8599, + "step": 11248 + }, + { + "epoch": 2.0026709401709404, + "grad_norm": 0.6131469011306763, + "learning_rate": 9.994401011424972e-05, + "loss": 0.8504, + "step": 11249 + }, + { + "epoch": 2.0028490028490027, + "grad_norm": 0.6809168457984924, + "learning_rate": 9.993001264486903e-05, + "loss": 0.761, + "step": 11250 + }, + { + "epoch": 2.0030270655270654, + "grad_norm": 0.6721677184104919, + "learning_rate": 9.991601517685962e-05, + "loss": 0.9146, + "step": 11251 + }, + { + "epoch": 2.003205128205128, + "grad_norm": 0.6395483016967773, + "learning_rate": 9.990201771049569e-05, + "loss": 0.8583, + "step": 11252 + }, + { + "epoch": 2.003383190883191, + "grad_norm": 0.8524805903434753, + "learning_rate": 9.988802024605153e-05, + "loss": 0.859, + "step": 11253 + }, + { + "epoch": 2.0035612535612537, + "grad_norm": 0.6186681389808655, + "learning_rate": 9.987402278380136e-05, + "loss": 0.6695, + "step": 11254 + }, + { + "epoch": 2.0037393162393164, + "grad_norm": 0.593245267868042, + "learning_rate": 9.98600253240195e-05, + "loss": 0.7104, + "step": 11255 + }, + { + "epoch": 2.0039173789173788, + "grad_norm": 0.6806482672691345, + "learning_rate": 9.98460278669801e-05, + "loss": 0.6208, + "step": 11256 + }, + { + "epoch": 2.0040954415954415, + "grad_norm": 0.7329097390174866, + "learning_rate": 9.983203041295753e-05, + "loss": 0.8264, + "step": 11257 + }, + { + "epoch": 2.0042735042735043, + "grad_norm": 0.7579078078269958, + "learning_rate": 9.981803296222591e-05, + "loss": 0.7572, + "step": 11258 + }, + { + "epoch": 2.004451566951567, + "grad_norm": 0.7945193648338318, + "learning_rate": 9.980403551505958e-05, + "loss": 0.7916, + "step": 11259 + }, + { + "epoch": 2.0046296296296298, + "grad_norm": 0.5528121590614319, + "learning_rate": 9.979003807173276e-05, + "loss": 0.5609, + "step": 11260 + }, + { + "epoch": 2.0048076923076925, + "grad_norm": 0.7321668863296509, + "learning_rate": 9.977604063251973e-05, + "loss": 0.9041, + "step": 11261 + }, + { + "epoch": 2.004985754985755, + "grad_norm": 0.6553691029548645, + "learning_rate": 9.976204319769469e-05, + "loss": 0.853, + "step": 11262 + }, + { + "epoch": 2.0051638176638176, + "grad_norm": 0.789152979850769, + "learning_rate": 9.974804576753194e-05, + "loss": 0.9909, + "step": 11263 + }, + { + "epoch": 2.0053418803418803, + "grad_norm": 0.6342231631278992, + "learning_rate": 9.973404834230568e-05, + "loss": 0.8841, + "step": 11264 + }, + { + "epoch": 2.005519943019943, + "grad_norm": 0.671882688999176, + "learning_rate": 9.97200509222902e-05, + "loss": 0.8147, + "step": 11265 + }, + { + "epoch": 2.005698005698006, + "grad_norm": 0.8409315347671509, + "learning_rate": 9.970605350775978e-05, + "loss": 1.0466, + "step": 11266 + }, + { + "epoch": 2.0058760683760686, + "grad_norm": 0.6155081987380981, + "learning_rate": 9.969205609898858e-05, + "loss": 0.9461, + "step": 11267 + }, + { + "epoch": 2.006054131054131, + "grad_norm": 0.656370997428894, + "learning_rate": 9.967805869625093e-05, + "loss": 0.6152, + "step": 11268 + }, + { + "epoch": 2.0062321937321936, + "grad_norm": 0.6441524624824524, + "learning_rate": 9.966406129982103e-05, + "loss": 0.9023, + "step": 11269 + }, + { + "epoch": 2.0064102564102564, + "grad_norm": 0.7976031303405762, + "learning_rate": 9.96500639099732e-05, + "loss": 0.8886, + "step": 11270 + }, + { + "epoch": 2.006588319088319, + "grad_norm": 0.6888235807418823, + "learning_rate": 9.963606652698159e-05, + "loss": 0.7216, + "step": 11271 + }, + { + "epoch": 2.006766381766382, + "grad_norm": 0.8439735770225525, + "learning_rate": 9.962206915112054e-05, + "loss": 1.0161, + "step": 11272 + }, + { + "epoch": 2.0069444444444446, + "grad_norm": 0.6425265669822693, + "learning_rate": 9.960807178266423e-05, + "loss": 0.8718, + "step": 11273 + }, + { + "epoch": 2.007122507122507, + "grad_norm": 0.7393937110900879, + "learning_rate": 9.959407442188696e-05, + "loss": 0.7615, + "step": 11274 + }, + { + "epoch": 2.0073005698005697, + "grad_norm": 0.5919229984283447, + "learning_rate": 9.958007706906292e-05, + "loss": 0.6761, + "step": 11275 + }, + { + "epoch": 2.0074786324786325, + "grad_norm": 0.765935480594635, + "learning_rate": 9.956607972446644e-05, + "loss": 0.9057, + "step": 11276 + }, + { + "epoch": 2.007656695156695, + "grad_norm": 0.7091122269630432, + "learning_rate": 9.955208238837169e-05, + "loss": 0.8322, + "step": 11277 + }, + { + "epoch": 2.007834757834758, + "grad_norm": 0.850652813911438, + "learning_rate": 9.953808506105299e-05, + "loss": 0.9942, + "step": 11278 + }, + { + "epoch": 2.0080128205128207, + "grad_norm": 0.7341200113296509, + "learning_rate": 9.952408774278452e-05, + "loss": 0.7826, + "step": 11279 + }, + { + "epoch": 2.008190883190883, + "grad_norm": 0.6891999840736389, + "learning_rate": 9.95100904338406e-05, + "loss": 0.8939, + "step": 11280 + }, + { + "epoch": 2.0083689458689458, + "grad_norm": 0.800881028175354, + "learning_rate": 9.94960931344954e-05, + "loss": 0.8036, + "step": 11281 + }, + { + "epoch": 2.0085470085470085, + "grad_norm": 0.7483115792274475, + "learning_rate": 9.948209584502328e-05, + "loss": 0.7203, + "step": 11282 + }, + { + "epoch": 2.0087250712250713, + "grad_norm": 0.7314630150794983, + "learning_rate": 9.946809856569833e-05, + "loss": 0.8907, + "step": 11283 + }, + { + "epoch": 2.008903133903134, + "grad_norm": 0.7317429184913635, + "learning_rate": 9.945410129679493e-05, + "loss": 0.8971, + "step": 11284 + }, + { + "epoch": 2.0090811965811968, + "grad_norm": 0.6968898177146912, + "learning_rate": 9.944010403858726e-05, + "loss": 0.8638, + "step": 11285 + }, + { + "epoch": 2.009259259259259, + "grad_norm": 0.6680058240890503, + "learning_rate": 9.942610679134957e-05, + "loss": 0.7524, + "step": 11286 + }, + { + "epoch": 2.009437321937322, + "grad_norm": 0.6863839030265808, + "learning_rate": 9.941210955535618e-05, + "loss": 0.9647, + "step": 11287 + }, + { + "epoch": 2.0096153846153846, + "grad_norm": 0.7137607336044312, + "learning_rate": 9.939811233088125e-05, + "loss": 0.7089, + "step": 11288 + }, + { + "epoch": 2.0097934472934473, + "grad_norm": 0.8341759443283081, + "learning_rate": 9.938411511819907e-05, + "loss": 0.9461, + "step": 11289 + }, + { + "epoch": 2.00997150997151, + "grad_norm": 0.7326228022575378, + "learning_rate": 9.937011791758384e-05, + "loss": 0.8795, + "step": 11290 + }, + { + "epoch": 2.010149572649573, + "grad_norm": 0.6795905232429504, + "learning_rate": 9.935612072930989e-05, + "loss": 0.8298, + "step": 11291 + }, + { + "epoch": 2.010327635327635, + "grad_norm": 0.7060360312461853, + "learning_rate": 9.934212355365139e-05, + "loss": 0.8483, + "step": 11292 + }, + { + "epoch": 2.010505698005698, + "grad_norm": 0.7532246112823486, + "learning_rate": 9.932812639088265e-05, + "loss": 0.9061, + "step": 11293 + }, + { + "epoch": 2.0106837606837606, + "grad_norm": 0.6563972234725952, + "learning_rate": 9.931412924127781e-05, + "loss": 0.8511, + "step": 11294 + }, + { + "epoch": 2.0108618233618234, + "grad_norm": 0.6672948002815247, + "learning_rate": 9.930013210511125e-05, + "loss": 0.7875, + "step": 11295 + }, + { + "epoch": 2.011039886039886, + "grad_norm": 0.7173593640327454, + "learning_rate": 9.928613498265709e-05, + "loss": 0.8602, + "step": 11296 + }, + { + "epoch": 2.011217948717949, + "grad_norm": 0.7399459481239319, + "learning_rate": 9.927213787418968e-05, + "loss": 0.8711, + "step": 11297 + }, + { + "epoch": 2.011396011396011, + "grad_norm": 0.7693262696266174, + "learning_rate": 9.925814077998317e-05, + "loss": 0.9927, + "step": 11298 + }, + { + "epoch": 2.011574074074074, + "grad_norm": 0.7998616695404053, + "learning_rate": 9.92441437003119e-05, + "loss": 1.1585, + "step": 11299 + }, + { + "epoch": 2.0117521367521367, + "grad_norm": 0.7239874005317688, + "learning_rate": 9.923014663545002e-05, + "loss": 0.8736, + "step": 11300 + }, + { + "epoch": 2.0119301994301995, + "grad_norm": 0.8565806150436401, + "learning_rate": 9.921614958567186e-05, + "loss": 0.9768, + "step": 11301 + }, + { + "epoch": 2.012108262108262, + "grad_norm": 0.6341429948806763, + "learning_rate": 9.920215255125158e-05, + "loss": 0.6553, + "step": 11302 + }, + { + "epoch": 2.012286324786325, + "grad_norm": 0.824182391166687, + "learning_rate": 9.91881555324635e-05, + "loss": 1.0138, + "step": 11303 + }, + { + "epoch": 2.0124643874643873, + "grad_norm": 0.6309344172477722, + "learning_rate": 9.917415852958178e-05, + "loss": 0.619, + "step": 11304 + }, + { + "epoch": 2.01264245014245, + "grad_norm": 0.7469239830970764, + "learning_rate": 9.916016154288071e-05, + "loss": 0.8537, + "step": 11305 + }, + { + "epoch": 2.0128205128205128, + "grad_norm": 0.7433663606643677, + "learning_rate": 9.914616457263459e-05, + "loss": 0.8518, + "step": 11306 + }, + { + "epoch": 2.0129985754985755, + "grad_norm": 0.6550318002700806, + "learning_rate": 9.913216761911755e-05, + "loss": 0.8021, + "step": 11307 + }, + { + "epoch": 2.0131766381766383, + "grad_norm": 0.7360837459564209, + "learning_rate": 9.911817068260392e-05, + "loss": 0.7002, + "step": 11308 + }, + { + "epoch": 2.013354700854701, + "grad_norm": 0.7208407521247864, + "learning_rate": 9.910417376336786e-05, + "loss": 0.8633, + "step": 11309 + }, + { + "epoch": 2.0135327635327633, + "grad_norm": 0.7758026719093323, + "learning_rate": 9.909017686168369e-05, + "loss": 0.764, + "step": 11310 + }, + { + "epoch": 2.013710826210826, + "grad_norm": 0.8215547204017639, + "learning_rate": 9.90761799778256e-05, + "loss": 0.7062, + "step": 11311 + }, + { + "epoch": 2.013888888888889, + "grad_norm": 0.6731052994728088, + "learning_rate": 9.906218311206786e-05, + "loss": 0.902, + "step": 11312 + }, + { + "epoch": 2.0140669515669516, + "grad_norm": 0.74113929271698, + "learning_rate": 9.904818626468466e-05, + "loss": 0.7229, + "step": 11313 + }, + { + "epoch": 2.0142450142450143, + "grad_norm": 0.6673575639724731, + "learning_rate": 9.90341894359503e-05, + "loss": 0.7299, + "step": 11314 + }, + { + "epoch": 2.014423076923077, + "grad_norm": 0.7665545344352722, + "learning_rate": 9.902019262613897e-05, + "loss": 0.6993, + "step": 11315 + }, + { + "epoch": 2.0146011396011394, + "grad_norm": 0.6423895359039307, + "learning_rate": 9.900619583552497e-05, + "loss": 0.7344, + "step": 11316 + }, + { + "epoch": 2.014779202279202, + "grad_norm": 0.7071038484573364, + "learning_rate": 9.899219906438245e-05, + "loss": 0.6951, + "step": 11317 + }, + { + "epoch": 2.014957264957265, + "grad_norm": 0.689984142780304, + "learning_rate": 9.897820231298574e-05, + "loss": 0.8496, + "step": 11318 + }, + { + "epoch": 2.0151353276353277, + "grad_norm": 0.8747256398200989, + "learning_rate": 9.896420558160901e-05, + "loss": 0.9752, + "step": 11319 + }, + { + "epoch": 2.0153133903133904, + "grad_norm": 0.6828433275222778, + "learning_rate": 9.895020887052651e-05, + "loss": 0.8369, + "step": 11320 + }, + { + "epoch": 2.015491452991453, + "grad_norm": 0.7334261536598206, + "learning_rate": 9.89362121800125e-05, + "loss": 0.7744, + "step": 11321 + }, + { + "epoch": 2.0156695156695155, + "grad_norm": 0.7896139621734619, + "learning_rate": 9.892221551034122e-05, + "loss": 0.8353, + "step": 11322 + }, + { + "epoch": 2.015847578347578, + "grad_norm": 0.6673476099967957, + "learning_rate": 9.890821886178684e-05, + "loss": 0.8644, + "step": 11323 + }, + { + "epoch": 2.016025641025641, + "grad_norm": 0.7475691437721252, + "learning_rate": 9.889422223462368e-05, + "loss": 0.8034, + "step": 11324 + }, + { + "epoch": 2.0162037037037037, + "grad_norm": 0.9086315631866455, + "learning_rate": 9.888022562912593e-05, + "loss": 1.1878, + "step": 11325 + }, + { + "epoch": 2.0163817663817665, + "grad_norm": 0.6634678244590759, + "learning_rate": 9.88662290455678e-05, + "loss": 0.9655, + "step": 11326 + }, + { + "epoch": 2.0165598290598292, + "grad_norm": 0.7184932827949524, + "learning_rate": 9.885223248422361e-05, + "loss": 0.5964, + "step": 11327 + }, + { + "epoch": 2.0167378917378915, + "grad_norm": 0.6319148540496826, + "learning_rate": 9.883823594536751e-05, + "loss": 0.5692, + "step": 11328 + }, + { + "epoch": 2.0169159544159543, + "grad_norm": 0.6232550144195557, + "learning_rate": 9.88242394292738e-05, + "loss": 0.6492, + "step": 11329 + }, + { + "epoch": 2.017094017094017, + "grad_norm": 0.7149667143821716, + "learning_rate": 9.881024293621663e-05, + "loss": 0.7023, + "step": 11330 + }, + { + "epoch": 2.01727207977208, + "grad_norm": 0.8871679902076721, + "learning_rate": 9.879624646647031e-05, + "loss": 0.954, + "step": 11331 + }, + { + "epoch": 2.0174501424501425, + "grad_norm": 0.6905941367149353, + "learning_rate": 9.878225002030901e-05, + "loss": 0.8534, + "step": 11332 + }, + { + "epoch": 2.0176282051282053, + "grad_norm": 0.8891478776931763, + "learning_rate": 9.876825359800703e-05, + "loss": 0.8324, + "step": 11333 + }, + { + "epoch": 2.0178062678062676, + "grad_norm": 0.8125092387199402, + "learning_rate": 9.875425719983852e-05, + "loss": 0.9604, + "step": 11334 + }, + { + "epoch": 2.0179843304843303, + "grad_norm": 0.7362027764320374, + "learning_rate": 9.874026082607778e-05, + "loss": 0.7879, + "step": 11335 + }, + { + "epoch": 2.018162393162393, + "grad_norm": 0.6763492226600647, + "learning_rate": 9.872626447699899e-05, + "loss": 0.8839, + "step": 11336 + }, + { + "epoch": 2.018340455840456, + "grad_norm": 0.7350467443466187, + "learning_rate": 9.871226815287644e-05, + "loss": 0.834, + "step": 11337 + }, + { + "epoch": 2.0185185185185186, + "grad_norm": 0.7768327593803406, + "learning_rate": 9.869827185398428e-05, + "loss": 1.1123, + "step": 11338 + }, + { + "epoch": 2.0186965811965814, + "grad_norm": 0.9218043088912964, + "learning_rate": 9.868427558059681e-05, + "loss": 0.9439, + "step": 11339 + }, + { + "epoch": 2.0188746438746437, + "grad_norm": 0.6613419651985168, + "learning_rate": 9.867027933298819e-05, + "loss": 0.836, + "step": 11340 + }, + { + "epoch": 2.0190527065527064, + "grad_norm": 0.7251055240631104, + "learning_rate": 9.865628311143273e-05, + "loss": 0.973, + "step": 11341 + }, + { + "epoch": 2.019230769230769, + "grad_norm": 0.6571859121322632, + "learning_rate": 9.864228691620458e-05, + "loss": 0.8811, + "step": 11342 + }, + { + "epoch": 2.019408831908832, + "grad_norm": 0.7552264928817749, + "learning_rate": 9.862829074757802e-05, + "loss": 0.9128, + "step": 11343 + }, + { + "epoch": 2.0195868945868947, + "grad_norm": 0.6724083423614502, + "learning_rate": 9.861429460582723e-05, + "loss": 0.8894, + "step": 11344 + }, + { + "epoch": 2.0197649572649574, + "grad_norm": 0.8309593200683594, + "learning_rate": 9.860029849122644e-05, + "loss": 0.9374, + "step": 11345 + }, + { + "epoch": 2.0199430199430197, + "grad_norm": 0.7709865570068359, + "learning_rate": 9.858630240404993e-05, + "loss": 0.8195, + "step": 11346 + }, + { + "epoch": 2.0201210826210825, + "grad_norm": 0.8163080811500549, + "learning_rate": 9.857230634457187e-05, + "loss": 0.9329, + "step": 11347 + }, + { + "epoch": 2.0202991452991452, + "grad_norm": 0.8424021005630493, + "learning_rate": 9.855831031306653e-05, + "loss": 0.8732, + "step": 11348 + }, + { + "epoch": 2.020477207977208, + "grad_norm": 0.7816365361213684, + "learning_rate": 9.854431430980808e-05, + "loss": 0.8858, + "step": 11349 + }, + { + "epoch": 2.0206552706552707, + "grad_norm": 0.7559000253677368, + "learning_rate": 9.853031833507075e-05, + "loss": 0.7146, + "step": 11350 + }, + { + "epoch": 2.0208333333333335, + "grad_norm": 0.6723140478134155, + "learning_rate": 9.85163223891288e-05, + "loss": 0.813, + "step": 11351 + }, + { + "epoch": 2.021011396011396, + "grad_norm": 0.757641077041626, + "learning_rate": 9.850232647225646e-05, + "loss": 0.794, + "step": 11352 + }, + { + "epoch": 2.0211894586894585, + "grad_norm": 0.8217115998268127, + "learning_rate": 9.848833058472787e-05, + "loss": 1.0407, + "step": 11353 + }, + { + "epoch": 2.0213675213675213, + "grad_norm": 0.8016467690467834, + "learning_rate": 9.847433472681736e-05, + "loss": 0.8967, + "step": 11354 + }, + { + "epoch": 2.021545584045584, + "grad_norm": 0.7703533172607422, + "learning_rate": 9.846033889879903e-05, + "loss": 0.9669, + "step": 11355 + }, + { + "epoch": 2.021723646723647, + "grad_norm": 0.7372044920921326, + "learning_rate": 9.84463431009472e-05, + "loss": 0.8581, + "step": 11356 + }, + { + "epoch": 2.0219017094017095, + "grad_norm": 0.7676188945770264, + "learning_rate": 9.8432347333536e-05, + "loss": 0.8498, + "step": 11357 + }, + { + "epoch": 2.0220797720797723, + "grad_norm": 0.7485190629959106, + "learning_rate": 9.841835159683977e-05, + "loss": 0.8492, + "step": 11358 + }, + { + "epoch": 2.0222578347578346, + "grad_norm": 0.7287883758544922, + "learning_rate": 9.840435589113262e-05, + "loss": 0.9072, + "step": 11359 + }, + { + "epoch": 2.0224358974358974, + "grad_norm": 0.7719354033470154, + "learning_rate": 9.83903602166888e-05, + "loss": 0.7657, + "step": 11360 + }, + { + "epoch": 2.02261396011396, + "grad_norm": 0.7679458260536194, + "learning_rate": 9.837636457378251e-05, + "loss": 0.7098, + "step": 11361 + }, + { + "epoch": 2.022792022792023, + "grad_norm": 0.7496665120124817, + "learning_rate": 9.836236896268803e-05, + "loss": 0.8459, + "step": 11362 + }, + { + "epoch": 2.0229700854700856, + "grad_norm": 0.8511863350868225, + "learning_rate": 9.834837338367949e-05, + "loss": 0.9782, + "step": 11363 + }, + { + "epoch": 2.0231481481481484, + "grad_norm": 0.5752342343330383, + "learning_rate": 9.833437783703114e-05, + "loss": 0.4539, + "step": 11364 + }, + { + "epoch": 2.0233262108262107, + "grad_norm": 0.6654593348503113, + "learning_rate": 9.832038232301722e-05, + "loss": 0.8009, + "step": 11365 + }, + { + "epoch": 2.0235042735042734, + "grad_norm": 0.7296777963638306, + "learning_rate": 9.83063868419119e-05, + "loss": 0.7841, + "step": 11366 + }, + { + "epoch": 2.023682336182336, + "grad_norm": 0.8404465913772583, + "learning_rate": 9.829239139398943e-05, + "loss": 0.9152, + "step": 11367 + }, + { + "epoch": 2.023860398860399, + "grad_norm": 0.6407002806663513, + "learning_rate": 9.827839597952397e-05, + "loss": 0.6953, + "step": 11368 + }, + { + "epoch": 2.0240384615384617, + "grad_norm": 0.8107042908668518, + "learning_rate": 9.826440059878982e-05, + "loss": 0.8726, + "step": 11369 + }, + { + "epoch": 2.0242165242165244, + "grad_norm": 0.803804874420166, + "learning_rate": 9.825040525206108e-05, + "loss": 0.8906, + "step": 11370 + }, + { + "epoch": 2.0243945868945867, + "grad_norm": 0.7625358700752258, + "learning_rate": 9.823640993961205e-05, + "loss": 0.8938, + "step": 11371 + }, + { + "epoch": 2.0245726495726495, + "grad_norm": 0.690793514251709, + "learning_rate": 9.822241466171686e-05, + "loss": 0.7926, + "step": 11372 + }, + { + "epoch": 2.0247507122507122, + "grad_norm": 0.7006554007530212, + "learning_rate": 9.820841941864983e-05, + "loss": 0.793, + "step": 11373 + }, + { + "epoch": 2.024928774928775, + "grad_norm": 0.8029078841209412, + "learning_rate": 9.819442421068504e-05, + "loss": 0.867, + "step": 11374 + }, + { + "epoch": 2.0251068376068377, + "grad_norm": 0.6999112367630005, + "learning_rate": 9.818042903809678e-05, + "loss": 0.688, + "step": 11375 + }, + { + "epoch": 2.0252849002849005, + "grad_norm": 0.6848462224006653, + "learning_rate": 9.816643390115923e-05, + "loss": 0.7337, + "step": 11376 + }, + { + "epoch": 2.025462962962963, + "grad_norm": 0.7698155641555786, + "learning_rate": 9.815243880014663e-05, + "loss": 0.9712, + "step": 11377 + }, + { + "epoch": 2.0256410256410255, + "grad_norm": 0.8449836373329163, + "learning_rate": 9.81384437353331e-05, + "loss": 0.9144, + "step": 11378 + }, + { + "epoch": 2.0258190883190883, + "grad_norm": 0.6340110301971436, + "learning_rate": 9.812444870699296e-05, + "loss": 0.6365, + "step": 11379 + }, + { + "epoch": 2.025997150997151, + "grad_norm": 0.7104073762893677, + "learning_rate": 9.81104537154003e-05, + "loss": 0.7781, + "step": 11380 + }, + { + "epoch": 2.026175213675214, + "grad_norm": 0.7287606000900269, + "learning_rate": 9.809645876082939e-05, + "loss": 0.9351, + "step": 11381 + }, + { + "epoch": 2.0263532763532766, + "grad_norm": 0.9640787243843079, + "learning_rate": 9.80824638435544e-05, + "loss": 0.8745, + "step": 11382 + }, + { + "epoch": 2.026531339031339, + "grad_norm": 0.5718010067939758, + "learning_rate": 9.806846896384959e-05, + "loss": 0.4711, + "step": 11383 + }, + { + "epoch": 2.0267094017094016, + "grad_norm": 0.7903527021408081, + "learning_rate": 9.805447412198907e-05, + "loss": 0.8241, + "step": 11384 + }, + { + "epoch": 2.0268874643874644, + "grad_norm": 0.8579357862472534, + "learning_rate": 9.80404793182471e-05, + "loss": 0.8621, + "step": 11385 + }, + { + "epoch": 2.027065527065527, + "grad_norm": 0.8466464877128601, + "learning_rate": 9.802648455289787e-05, + "loss": 0.8772, + "step": 11386 + }, + { + "epoch": 2.02724358974359, + "grad_norm": 0.7888286709785461, + "learning_rate": 9.801248982621557e-05, + "loss": 0.8352, + "step": 11387 + }, + { + "epoch": 2.0274216524216526, + "grad_norm": 0.6967005133628845, + "learning_rate": 9.799849513847444e-05, + "loss": 0.7936, + "step": 11388 + }, + { + "epoch": 2.027599715099715, + "grad_norm": 0.6987027525901794, + "learning_rate": 9.79845004899486e-05, + "loss": 0.7778, + "step": 11389 + }, + { + "epoch": 2.0277777777777777, + "grad_norm": 0.7414312362670898, + "learning_rate": 9.797050588091233e-05, + "loss": 0.9017, + "step": 11390 + }, + { + "epoch": 2.0279558404558404, + "grad_norm": 0.7932028770446777, + "learning_rate": 9.795651131163974e-05, + "loss": 0.8662, + "step": 11391 + }, + { + "epoch": 2.028133903133903, + "grad_norm": 0.8166332244873047, + "learning_rate": 9.79425167824051e-05, + "loss": 1.0489, + "step": 11392 + }, + { + "epoch": 2.028311965811966, + "grad_norm": 0.7265253663063049, + "learning_rate": 9.792852229348251e-05, + "loss": 0.9458, + "step": 11393 + }, + { + "epoch": 2.0284900284900287, + "grad_norm": 0.7374703288078308, + "learning_rate": 9.791452784514629e-05, + "loss": 0.9203, + "step": 11394 + }, + { + "epoch": 2.028668091168091, + "grad_norm": 0.6912441253662109, + "learning_rate": 9.790053343767052e-05, + "loss": 0.8986, + "step": 11395 + }, + { + "epoch": 2.0288461538461537, + "grad_norm": 0.871231734752655, + "learning_rate": 9.788653907132946e-05, + "loss": 0.6811, + "step": 11396 + }, + { + "epoch": 2.0290242165242165, + "grad_norm": 0.7361812591552734, + "learning_rate": 9.787254474639726e-05, + "loss": 0.7868, + "step": 11397 + }, + { + "epoch": 2.0292022792022792, + "grad_norm": 0.6828895211219788, + "learning_rate": 9.785855046314815e-05, + "loss": 0.7739, + "step": 11398 + }, + { + "epoch": 2.029380341880342, + "grad_norm": 0.7203328609466553, + "learning_rate": 9.784455622185626e-05, + "loss": 0.6474, + "step": 11399 + }, + { + "epoch": 2.0295584045584047, + "grad_norm": 0.774886429309845, + "learning_rate": 9.783056202279587e-05, + "loss": 0.8073, + "step": 11400 + }, + { + "epoch": 2.029736467236467, + "grad_norm": 0.6479005813598633, + "learning_rate": 9.781656786624106e-05, + "loss": 0.7237, + "step": 11401 + }, + { + "epoch": 2.02991452991453, + "grad_norm": 0.7269866466522217, + "learning_rate": 9.78025737524661e-05, + "loss": 0.9089, + "step": 11402 + }, + { + "epoch": 2.0300925925925926, + "grad_norm": 0.7265415191650391, + "learning_rate": 9.778857968174509e-05, + "loss": 0.827, + "step": 11403 + }, + { + "epoch": 2.0302706552706553, + "grad_norm": 0.8174277544021606, + "learning_rate": 9.777458565435227e-05, + "loss": 0.6752, + "step": 11404 + }, + { + "epoch": 2.030448717948718, + "grad_norm": 0.9333333969116211, + "learning_rate": 9.77605916705619e-05, + "loss": 0.9542, + "step": 11405 + }, + { + "epoch": 2.030626780626781, + "grad_norm": 0.6854027509689331, + "learning_rate": 9.774659773064801e-05, + "loss": 0.8526, + "step": 11406 + }, + { + "epoch": 2.030804843304843, + "grad_norm": 0.7711043357849121, + "learning_rate": 9.773260383488489e-05, + "loss": 0.9009, + "step": 11407 + }, + { + "epoch": 2.030982905982906, + "grad_norm": 0.6915287971496582, + "learning_rate": 9.771860998354667e-05, + "loss": 0.9635, + "step": 11408 + }, + { + "epoch": 2.0311609686609686, + "grad_norm": 0.7978841066360474, + "learning_rate": 9.770461617690758e-05, + "loss": 0.7563, + "step": 11409 + }, + { + "epoch": 2.0313390313390314, + "grad_norm": 0.6686414480209351, + "learning_rate": 9.769062241524172e-05, + "loss": 0.8282, + "step": 11410 + }, + { + "epoch": 2.031517094017094, + "grad_norm": 0.7024029493331909, + "learning_rate": 9.767662869882335e-05, + "loss": 0.9176, + "step": 11411 + }, + { + "epoch": 2.031695156695157, + "grad_norm": 0.6945844292640686, + "learning_rate": 9.766263502792659e-05, + "loss": 0.86, + "step": 11412 + }, + { + "epoch": 2.031873219373219, + "grad_norm": 0.7351676821708679, + "learning_rate": 9.764864140282569e-05, + "loss": 0.865, + "step": 11413 + }, + { + "epoch": 2.032051282051282, + "grad_norm": 0.7663825750350952, + "learning_rate": 9.763464782379472e-05, + "loss": 0.9309, + "step": 11414 + }, + { + "epoch": 2.0322293447293447, + "grad_norm": 0.7552894949913025, + "learning_rate": 9.762065429110798e-05, + "loss": 0.8366, + "step": 11415 + }, + { + "epoch": 2.0324074074074074, + "grad_norm": 0.6852208971977234, + "learning_rate": 9.760666080503951e-05, + "loss": 0.9095, + "step": 11416 + }, + { + "epoch": 2.03258547008547, + "grad_norm": 0.7759820222854614, + "learning_rate": 9.759266736586358e-05, + "loss": 0.7461, + "step": 11417 + }, + { + "epoch": 2.032763532763533, + "grad_norm": 0.6514183878898621, + "learning_rate": 9.757867397385431e-05, + "loss": 0.5479, + "step": 11418 + }, + { + "epoch": 2.0329415954415953, + "grad_norm": 0.7703103423118591, + "learning_rate": 9.756468062928593e-05, + "loss": 0.9588, + "step": 11419 + }, + { + "epoch": 2.033119658119658, + "grad_norm": 0.6937198638916016, + "learning_rate": 9.755068733243255e-05, + "loss": 0.8661, + "step": 11420 + }, + { + "epoch": 2.0332977207977208, + "grad_norm": 0.6675645112991333, + "learning_rate": 9.753669408356835e-05, + "loss": 0.7484, + "step": 11421 + }, + { + "epoch": 2.0334757834757835, + "grad_norm": 0.6653266549110413, + "learning_rate": 9.752270088296753e-05, + "loss": 0.7217, + "step": 11422 + }, + { + "epoch": 2.0336538461538463, + "grad_norm": 0.7893908023834229, + "learning_rate": 9.750870773090425e-05, + "loss": 0.6346, + "step": 11423 + }, + { + "epoch": 2.033831908831909, + "grad_norm": 0.7442745566368103, + "learning_rate": 9.749471462765265e-05, + "loss": 0.9452, + "step": 11424 + }, + { + "epoch": 2.0340099715099713, + "grad_norm": 0.8270035982131958, + "learning_rate": 9.748072157348691e-05, + "loss": 0.9906, + "step": 11425 + }, + { + "epoch": 2.034188034188034, + "grad_norm": 0.7195143699645996, + "learning_rate": 9.746672856868123e-05, + "loss": 0.9564, + "step": 11426 + }, + { + "epoch": 2.034366096866097, + "grad_norm": 0.75486820936203, + "learning_rate": 9.745273561350971e-05, + "loss": 0.9052, + "step": 11427 + }, + { + "epoch": 2.0345441595441596, + "grad_norm": 0.6710293889045715, + "learning_rate": 9.743874270824655e-05, + "loss": 0.6801, + "step": 11428 + }, + { + "epoch": 2.0347222222222223, + "grad_norm": 0.704175591468811, + "learning_rate": 9.742474985316588e-05, + "loss": 0.8619, + "step": 11429 + }, + { + "epoch": 2.034900284900285, + "grad_norm": 0.7941717505455017, + "learning_rate": 9.741075704854196e-05, + "loss": 0.8318, + "step": 11430 + }, + { + "epoch": 2.0350783475783474, + "grad_norm": 0.8592050671577454, + "learning_rate": 9.739676429464881e-05, + "loss": 0.8203, + "step": 11431 + }, + { + "epoch": 2.03525641025641, + "grad_norm": 0.9149407148361206, + "learning_rate": 9.738277159176068e-05, + "loss": 0.87, + "step": 11432 + }, + { + "epoch": 2.035434472934473, + "grad_norm": 0.780890941619873, + "learning_rate": 9.736877894015169e-05, + "loss": 0.6971, + "step": 11433 + }, + { + "epoch": 2.0356125356125356, + "grad_norm": 0.7540209293365479, + "learning_rate": 9.735478634009605e-05, + "loss": 0.8927, + "step": 11434 + }, + { + "epoch": 2.0357905982905984, + "grad_norm": 0.8556281924247742, + "learning_rate": 9.734079379186782e-05, + "loss": 0.7498, + "step": 11435 + }, + { + "epoch": 2.035968660968661, + "grad_norm": 0.8710931539535522, + "learning_rate": 9.732680129574128e-05, + "loss": 0.6009, + "step": 11436 + }, + { + "epoch": 2.0361467236467234, + "grad_norm": 0.6873082518577576, + "learning_rate": 9.731280885199045e-05, + "loss": 0.8441, + "step": 11437 + }, + { + "epoch": 2.036324786324786, + "grad_norm": 0.8333037495613098, + "learning_rate": 9.729881646088958e-05, + "loss": 0.888, + "step": 11438 + }, + { + "epoch": 2.036502849002849, + "grad_norm": 0.859365701675415, + "learning_rate": 9.728482412271277e-05, + "loss": 1.0272, + "step": 11439 + }, + { + "epoch": 2.0366809116809117, + "grad_norm": 0.7239334583282471, + "learning_rate": 9.727083183773423e-05, + "loss": 0.9428, + "step": 11440 + }, + { + "epoch": 2.0368589743589745, + "grad_norm": 0.8341524004936218, + "learning_rate": 9.725683960622804e-05, + "loss": 0.9275, + "step": 11441 + }, + { + "epoch": 2.037037037037037, + "grad_norm": 0.6992602348327637, + "learning_rate": 9.724284742846838e-05, + "loss": 0.8492, + "step": 11442 + }, + { + "epoch": 2.0372150997150995, + "grad_norm": 0.7429133057594299, + "learning_rate": 9.72288553047294e-05, + "loss": 0.8246, + "step": 11443 + }, + { + "epoch": 2.0373931623931623, + "grad_norm": 0.7765250205993652, + "learning_rate": 9.721486323528522e-05, + "loss": 0.8624, + "step": 11444 + }, + { + "epoch": 2.037571225071225, + "grad_norm": 0.9104889631271362, + "learning_rate": 9.720087122041007e-05, + "loss": 0.8369, + "step": 11445 + }, + { + "epoch": 2.0377492877492878, + "grad_norm": 0.6483191251754761, + "learning_rate": 9.718687926037798e-05, + "loss": 0.7347, + "step": 11446 + }, + { + "epoch": 2.0379273504273505, + "grad_norm": 0.7816178202629089, + "learning_rate": 9.717288735546317e-05, + "loss": 0.8607, + "step": 11447 + }, + { + "epoch": 2.0381054131054133, + "grad_norm": 0.6909009218215942, + "learning_rate": 9.715889550593975e-05, + "loss": 0.8764, + "step": 11448 + }, + { + "epoch": 2.0382834757834756, + "grad_norm": 0.8101255297660828, + "learning_rate": 9.71449037120819e-05, + "loss": 0.8858, + "step": 11449 + }, + { + "epoch": 2.0384615384615383, + "grad_norm": 0.7476511001586914, + "learning_rate": 9.71309119741637e-05, + "loss": 0.8765, + "step": 11450 + }, + { + "epoch": 2.038639601139601, + "grad_norm": 0.7514875531196594, + "learning_rate": 9.711692029245934e-05, + "loss": 0.942, + "step": 11451 + }, + { + "epoch": 2.038817663817664, + "grad_norm": 0.7400087118148804, + "learning_rate": 9.710292866724292e-05, + "loss": 0.8327, + "step": 11452 + }, + { + "epoch": 2.0389957264957266, + "grad_norm": 0.832979142665863, + "learning_rate": 9.70889370987886e-05, + "loss": 0.9714, + "step": 11453 + }, + { + "epoch": 2.0391737891737893, + "grad_norm": 0.6918326616287231, + "learning_rate": 9.70749455873705e-05, + "loss": 0.7765, + "step": 11454 + }, + { + "epoch": 2.0393518518518516, + "grad_norm": 0.8286036849021912, + "learning_rate": 9.70609541332628e-05, + "loss": 0.9138, + "step": 11455 + }, + { + "epoch": 2.0395299145299144, + "grad_norm": 0.6436729431152344, + "learning_rate": 9.704696273673955e-05, + "loss": 0.738, + "step": 11456 + }, + { + "epoch": 2.039707977207977, + "grad_norm": 0.7057681679725647, + "learning_rate": 9.703297139807496e-05, + "loss": 0.8107, + "step": 11457 + }, + { + "epoch": 2.03988603988604, + "grad_norm": 0.7444550395011902, + "learning_rate": 9.701898011754313e-05, + "loss": 0.8188, + "step": 11458 + }, + { + "epoch": 2.0400641025641026, + "grad_norm": 0.7622130513191223, + "learning_rate": 9.70049888954182e-05, + "loss": 0.8451, + "step": 11459 + }, + { + "epoch": 2.0402421652421654, + "grad_norm": 0.8166092038154602, + "learning_rate": 9.699099773197426e-05, + "loss": 0.9399, + "step": 11460 + }, + { + "epoch": 2.0404202279202277, + "grad_norm": 0.7235924601554871, + "learning_rate": 9.697700662748552e-05, + "loss": 0.7863, + "step": 11461 + }, + { + "epoch": 2.0405982905982905, + "grad_norm": 0.7150312662124634, + "learning_rate": 9.696301558222601e-05, + "loss": 0.8288, + "step": 11462 + }, + { + "epoch": 2.040776353276353, + "grad_norm": 0.8007016777992249, + "learning_rate": 9.694902459646993e-05, + "loss": 0.9203, + "step": 11463 + }, + { + "epoch": 2.040954415954416, + "grad_norm": 0.7665491700172424, + "learning_rate": 9.693503367049134e-05, + "loss": 0.7956, + "step": 11464 + }, + { + "epoch": 2.0411324786324787, + "grad_norm": 0.7499460577964783, + "learning_rate": 9.692104280456439e-05, + "loss": 0.9973, + "step": 11465 + }, + { + "epoch": 2.0413105413105415, + "grad_norm": 0.7598159909248352, + "learning_rate": 9.690705199896327e-05, + "loss": 0.95, + "step": 11466 + }, + { + "epoch": 2.041488603988604, + "grad_norm": 0.7699945569038391, + "learning_rate": 9.689306125396201e-05, + "loss": 0.8731, + "step": 11467 + }, + { + "epoch": 2.0416666666666665, + "grad_norm": 0.6724731922149658, + "learning_rate": 9.687907056983476e-05, + "loss": 0.906, + "step": 11468 + }, + { + "epoch": 2.0418447293447293, + "grad_norm": 0.9238275289535522, + "learning_rate": 9.686507994685562e-05, + "loss": 0.8397, + "step": 11469 + }, + { + "epoch": 2.042022792022792, + "grad_norm": 0.744969367980957, + "learning_rate": 9.685108938529876e-05, + "loss": 0.8436, + "step": 11470 + }, + { + "epoch": 2.0422008547008548, + "grad_norm": 0.6983298063278198, + "learning_rate": 9.683709888543824e-05, + "loss": 0.8235, + "step": 11471 + }, + { + "epoch": 2.0423789173789175, + "grad_norm": 0.7098708748817444, + "learning_rate": 9.682310844754824e-05, + "loss": 0.8235, + "step": 11472 + }, + { + "epoch": 2.04255698005698, + "grad_norm": 0.7492793798446655, + "learning_rate": 9.680911807190277e-05, + "loss": 0.7988, + "step": 11473 + }, + { + "epoch": 2.0427350427350426, + "grad_norm": 0.6952250003814697, + "learning_rate": 9.679512775877604e-05, + "loss": 0.7928, + "step": 11474 + }, + { + "epoch": 2.0429131054131053, + "grad_norm": 0.6442983150482178, + "learning_rate": 9.678113750844209e-05, + "loss": 0.8206, + "step": 11475 + }, + { + "epoch": 2.043091168091168, + "grad_norm": 0.7408245205879211, + "learning_rate": 9.67671473211751e-05, + "loss": 0.6941, + "step": 11476 + }, + { + "epoch": 2.043269230769231, + "grad_norm": 0.8277738094329834, + "learning_rate": 9.675315719724913e-05, + "loss": 1.3153, + "step": 11477 + }, + { + "epoch": 2.0434472934472936, + "grad_norm": 0.7535714507102966, + "learning_rate": 9.67391671369383e-05, + "loss": 0.9238, + "step": 11478 + }, + { + "epoch": 2.0436253561253563, + "grad_norm": 0.8341996073722839, + "learning_rate": 9.67251771405167e-05, + "loss": 0.8149, + "step": 11479 + }, + { + "epoch": 2.0438034188034186, + "grad_norm": 0.7365956902503967, + "learning_rate": 9.671118720825849e-05, + "loss": 0.799, + "step": 11480 + }, + { + "epoch": 2.0439814814814814, + "grad_norm": 0.7630738615989685, + "learning_rate": 9.669719734043769e-05, + "loss": 0.9284, + "step": 11481 + }, + { + "epoch": 2.044159544159544, + "grad_norm": 0.659172773361206, + "learning_rate": 9.668320753732848e-05, + "loss": 0.7594, + "step": 11482 + }, + { + "epoch": 2.044337606837607, + "grad_norm": 0.7724705934524536, + "learning_rate": 9.66692177992049e-05, + "loss": 0.8623, + "step": 11483 + }, + { + "epoch": 2.0445156695156697, + "grad_norm": 0.7140040993690491, + "learning_rate": 9.665522812634108e-05, + "loss": 0.851, + "step": 11484 + }, + { + "epoch": 2.0446937321937324, + "grad_norm": 0.9072890877723694, + "learning_rate": 9.664123851901115e-05, + "loss": 0.9459, + "step": 11485 + }, + { + "epoch": 2.0448717948717947, + "grad_norm": 0.8145443201065063, + "learning_rate": 9.662724897748915e-05, + "loss": 0.9067, + "step": 11486 + }, + { + "epoch": 2.0450498575498575, + "grad_norm": 0.8471246957778931, + "learning_rate": 9.661325950204922e-05, + "loss": 0.7194, + "step": 11487 + }, + { + "epoch": 2.04522792022792, + "grad_norm": 0.8465375304222107, + "learning_rate": 9.659927009296541e-05, + "loss": 0.9495, + "step": 11488 + }, + { + "epoch": 2.045405982905983, + "grad_norm": 0.7597832083702087, + "learning_rate": 9.658528075051185e-05, + "loss": 0.7526, + "step": 11489 + }, + { + "epoch": 2.0455840455840457, + "grad_norm": 0.8013564944267273, + "learning_rate": 9.657129147496261e-05, + "loss": 1.0514, + "step": 11490 + }, + { + "epoch": 2.0457621082621085, + "grad_norm": 0.8695764541625977, + "learning_rate": 9.655730226659182e-05, + "loss": 0.9925, + "step": 11491 + }, + { + "epoch": 2.0459401709401708, + "grad_norm": 0.7295607328414917, + "learning_rate": 9.65433131256735e-05, + "loss": 0.8652, + "step": 11492 + }, + { + "epoch": 2.0461182336182335, + "grad_norm": 0.7819971442222595, + "learning_rate": 9.652932405248181e-05, + "loss": 0.8601, + "step": 11493 + }, + { + "epoch": 2.0462962962962963, + "grad_norm": 0.7244205474853516, + "learning_rate": 9.651533504729078e-05, + "loss": 0.752, + "step": 11494 + }, + { + "epoch": 2.046474358974359, + "grad_norm": 0.7774363160133362, + "learning_rate": 9.650134611037456e-05, + "loss": 0.8638, + "step": 11495 + }, + { + "epoch": 2.046652421652422, + "grad_norm": 0.7955372929573059, + "learning_rate": 9.648735724200715e-05, + "loss": 0.9662, + "step": 11496 + }, + { + "epoch": 2.0468304843304845, + "grad_norm": 0.7114127278327942, + "learning_rate": 9.647336844246273e-05, + "loss": 0.9523, + "step": 11497 + }, + { + "epoch": 2.047008547008547, + "grad_norm": 0.7449100017547607, + "learning_rate": 9.645937971201527e-05, + "loss": 0.7898, + "step": 11498 + }, + { + "epoch": 2.0471866096866096, + "grad_norm": 0.7541512846946716, + "learning_rate": 9.644539105093895e-05, + "loss": 0.9286, + "step": 11499 + }, + { + "epoch": 2.0473646723646723, + "grad_norm": 0.6816682815551758, + "learning_rate": 9.643140245950778e-05, + "loss": 0.7757, + "step": 11500 + }, + { + "epoch": 2.047542735042735, + "grad_norm": 0.7222850918769836, + "learning_rate": 9.641741393799591e-05, + "loss": 0.8415, + "step": 11501 + }, + { + "epoch": 2.047720797720798, + "grad_norm": 0.7605552077293396, + "learning_rate": 9.640342548667732e-05, + "loss": 0.8875, + "step": 11502 + }, + { + "epoch": 2.0478988603988606, + "grad_norm": 0.7442240118980408, + "learning_rate": 9.638943710582615e-05, + "loss": 0.8755, + "step": 11503 + }, + { + "epoch": 2.048076923076923, + "grad_norm": 0.7065736651420593, + "learning_rate": 9.637544879571648e-05, + "loss": 0.6885, + "step": 11504 + }, + { + "epoch": 2.0482549857549857, + "grad_norm": 0.6400303244590759, + "learning_rate": 9.636146055662232e-05, + "loss": 0.5775, + "step": 11505 + }, + { + "epoch": 2.0484330484330484, + "grad_norm": 0.7955389022827148, + "learning_rate": 9.634747238881783e-05, + "loss": 1.0182, + "step": 11506 + }, + { + "epoch": 2.048611111111111, + "grad_norm": 0.8283255696296692, + "learning_rate": 9.6333484292577e-05, + "loss": 0.9247, + "step": 11507 + }, + { + "epoch": 2.048789173789174, + "grad_norm": 0.7619521617889404, + "learning_rate": 9.631949626817399e-05, + "loss": 0.8355, + "step": 11508 + }, + { + "epoch": 2.0489672364672367, + "grad_norm": 0.7204191088676453, + "learning_rate": 9.630550831588273e-05, + "loss": 0.8571, + "step": 11509 + }, + { + "epoch": 2.049145299145299, + "grad_norm": 0.7568399310112, + "learning_rate": 9.629152043597738e-05, + "loss": 0.7349, + "step": 11510 + }, + { + "epoch": 2.0493233618233617, + "grad_norm": 0.8594959378242493, + "learning_rate": 9.627753262873199e-05, + "loss": 0.6918, + "step": 11511 + }, + { + "epoch": 2.0495014245014245, + "grad_norm": 0.6345391869544983, + "learning_rate": 9.626354489442064e-05, + "loss": 0.871, + "step": 11512 + }, + { + "epoch": 2.0496794871794872, + "grad_norm": 0.7671827673912048, + "learning_rate": 9.624955723331732e-05, + "loss": 0.8518, + "step": 11513 + }, + { + "epoch": 2.04985754985755, + "grad_norm": 0.7182049751281738, + "learning_rate": 9.623556964569616e-05, + "loss": 0.889, + "step": 11514 + }, + { + "epoch": 2.0500356125356127, + "grad_norm": 0.8342016339302063, + "learning_rate": 9.622158213183118e-05, + "loss": 0.9712, + "step": 11515 + }, + { + "epoch": 2.050213675213675, + "grad_norm": 0.7007761001586914, + "learning_rate": 9.620759469199649e-05, + "loss": 0.8113, + "step": 11516 + }, + { + "epoch": 2.050391737891738, + "grad_norm": 0.7129531502723694, + "learning_rate": 9.619360732646605e-05, + "loss": 0.8666, + "step": 11517 + }, + { + "epoch": 2.0505698005698005, + "grad_norm": 0.7505812048912048, + "learning_rate": 9.6179620035514e-05, + "loss": 0.7321, + "step": 11518 + }, + { + "epoch": 2.0507478632478633, + "grad_norm": 0.7407607436180115, + "learning_rate": 9.616563281941433e-05, + "loss": 0.9275, + "step": 11519 + }, + { + "epoch": 2.050925925925926, + "grad_norm": 0.769345223903656, + "learning_rate": 9.615164567844116e-05, + "loss": 0.9731, + "step": 11520 + }, + { + "epoch": 2.051103988603989, + "grad_norm": 0.7782812118530273, + "learning_rate": 9.613765861286846e-05, + "loss": 0.9702, + "step": 11521 + }, + { + "epoch": 2.051282051282051, + "grad_norm": 0.7071413993835449, + "learning_rate": 9.612367162297037e-05, + "loss": 0.8451, + "step": 11522 + }, + { + "epoch": 2.051460113960114, + "grad_norm": 0.7598503232002258, + "learning_rate": 9.610968470902082e-05, + "loss": 0.8641, + "step": 11523 + }, + { + "epoch": 2.0516381766381766, + "grad_norm": 0.7951003313064575, + "learning_rate": 9.609569787129394e-05, + "loss": 0.9131, + "step": 11524 + }, + { + "epoch": 2.0518162393162394, + "grad_norm": 0.8029175996780396, + "learning_rate": 9.608171111006374e-05, + "loss": 0.8618, + "step": 11525 + }, + { + "epoch": 2.051994301994302, + "grad_norm": 0.6993120908737183, + "learning_rate": 9.606772442560428e-05, + "loss": 0.8487, + "step": 11526 + }, + { + "epoch": 2.052172364672365, + "grad_norm": 0.8039231896400452, + "learning_rate": 9.605373781818961e-05, + "loss": 1.0102, + "step": 11527 + }, + { + "epoch": 2.052350427350427, + "grad_norm": 0.714849054813385, + "learning_rate": 9.603975128809373e-05, + "loss": 0.8977, + "step": 11528 + }, + { + "epoch": 2.05252849002849, + "grad_norm": 0.8728037476539612, + "learning_rate": 9.60257648355907e-05, + "loss": 0.8004, + "step": 11529 + }, + { + "epoch": 2.0527065527065527, + "grad_norm": 0.764776885509491, + "learning_rate": 9.601177846095454e-05, + "loss": 0.9205, + "step": 11530 + }, + { + "epoch": 2.0528846153846154, + "grad_norm": 0.6948725581169128, + "learning_rate": 9.599779216445934e-05, + "loss": 0.7864, + "step": 11531 + }, + { + "epoch": 2.053062678062678, + "grad_norm": 0.7663996815681458, + "learning_rate": 9.598380594637903e-05, + "loss": 0.8877, + "step": 11532 + }, + { + "epoch": 2.053240740740741, + "grad_norm": 0.7584146857261658, + "learning_rate": 9.596981980698776e-05, + "loss": 1.1328, + "step": 11533 + }, + { + "epoch": 2.0534188034188032, + "grad_norm": 0.7701094150543213, + "learning_rate": 9.595583374655945e-05, + "loss": 0.7551, + "step": 11534 + }, + { + "epoch": 2.053596866096866, + "grad_norm": 0.7745714783668518, + "learning_rate": 9.594184776536821e-05, + "loss": 0.8862, + "step": 11535 + }, + { + "epoch": 2.0537749287749287, + "grad_norm": 0.7832430005073547, + "learning_rate": 9.5927861863688e-05, + "loss": 0.8736, + "step": 11536 + }, + { + "epoch": 2.0539529914529915, + "grad_norm": 0.7354840040206909, + "learning_rate": 9.591387604179291e-05, + "loss": 0.8183, + "step": 11537 + }, + { + "epoch": 2.0541310541310542, + "grad_norm": 0.7516480684280396, + "learning_rate": 9.589989029995691e-05, + "loss": 0.924, + "step": 11538 + }, + { + "epoch": 2.054309116809117, + "grad_norm": 0.7942310571670532, + "learning_rate": 9.588590463845405e-05, + "loss": 1.0283, + "step": 11539 + }, + { + "epoch": 2.0544871794871793, + "grad_norm": 0.7716572880744934, + "learning_rate": 9.587191905755832e-05, + "loss": 0.8686, + "step": 11540 + }, + { + "epoch": 2.054665242165242, + "grad_norm": 0.8075140118598938, + "learning_rate": 9.585793355754381e-05, + "loss": 0.8731, + "step": 11541 + }, + { + "epoch": 2.054843304843305, + "grad_norm": 0.8119283318519592, + "learning_rate": 9.584394813868444e-05, + "loss": 0.9543, + "step": 11542 + }, + { + "epoch": 2.0550213675213675, + "grad_norm": 0.6476314067840576, + "learning_rate": 9.582996280125427e-05, + "loss": 0.6943, + "step": 11543 + }, + { + "epoch": 2.0551994301994303, + "grad_norm": 0.7617185711860657, + "learning_rate": 9.581597754552737e-05, + "loss": 0.6942, + "step": 11544 + }, + { + "epoch": 2.055377492877493, + "grad_norm": 0.879355788230896, + "learning_rate": 9.580199237177765e-05, + "loss": 0.825, + "step": 11545 + }, + { + "epoch": 2.0555555555555554, + "grad_norm": 0.8229055404663086, + "learning_rate": 9.578800728027919e-05, + "loss": 0.9973, + "step": 11546 + }, + { + "epoch": 2.055733618233618, + "grad_norm": 0.7808930277824402, + "learning_rate": 9.577402227130596e-05, + "loss": 0.9525, + "step": 11547 + }, + { + "epoch": 2.055911680911681, + "grad_norm": 0.870499849319458, + "learning_rate": 9.576003734513201e-05, + "loss": 0.8874, + "step": 11548 + }, + { + "epoch": 2.0560897435897436, + "grad_norm": 0.8254318833351135, + "learning_rate": 9.57460525020313e-05, + "loss": 0.6656, + "step": 11549 + }, + { + "epoch": 2.0562678062678064, + "grad_norm": 0.8358132243156433, + "learning_rate": 9.573206774227786e-05, + "loss": 0.7946, + "step": 11550 + }, + { + "epoch": 2.056445868945869, + "grad_norm": 0.636366605758667, + "learning_rate": 9.571808306614568e-05, + "loss": 0.6757, + "step": 11551 + }, + { + "epoch": 2.0566239316239314, + "grad_norm": 0.8884546160697937, + "learning_rate": 9.57040984739088e-05, + "loss": 0.6775, + "step": 11552 + }, + { + "epoch": 2.056801994301994, + "grad_norm": 0.7240797877311707, + "learning_rate": 9.569011396584115e-05, + "loss": 0.8033, + "step": 11553 + }, + { + "epoch": 2.056980056980057, + "grad_norm": 0.8730767965316772, + "learning_rate": 9.567612954221678e-05, + "loss": 0.9577, + "step": 11554 + }, + { + "epoch": 2.0571581196581197, + "grad_norm": 0.6785064339637756, + "learning_rate": 9.566214520330966e-05, + "loss": 0.6241, + "step": 11555 + }, + { + "epoch": 2.0573361823361824, + "grad_norm": 0.7757805585861206, + "learning_rate": 9.564816094939382e-05, + "loss": 0.7926, + "step": 11556 + }, + { + "epoch": 2.057514245014245, + "grad_norm": 0.7630164623260498, + "learning_rate": 9.563417678074319e-05, + "loss": 0.8547, + "step": 11557 + }, + { + "epoch": 2.0576923076923075, + "grad_norm": 0.7690725922584534, + "learning_rate": 9.562019269763184e-05, + "loss": 0.9172, + "step": 11558 + }, + { + "epoch": 2.0578703703703702, + "grad_norm": 0.81644207239151, + "learning_rate": 9.560620870033367e-05, + "loss": 0.811, + "step": 11559 + }, + { + "epoch": 2.058048433048433, + "grad_norm": 0.8240723013877869, + "learning_rate": 9.559222478912273e-05, + "loss": 0.9094, + "step": 11560 + }, + { + "epoch": 2.0582264957264957, + "grad_norm": 0.7168204188346863, + "learning_rate": 9.557824096427297e-05, + "loss": 1.0617, + "step": 11561 + }, + { + "epoch": 2.0584045584045585, + "grad_norm": 0.6648391485214233, + "learning_rate": 9.556425722605846e-05, + "loss": 0.6556, + "step": 11562 + }, + { + "epoch": 2.0585826210826212, + "grad_norm": 0.7291145324707031, + "learning_rate": 9.555027357475305e-05, + "loss": 0.784, + "step": 11563 + }, + { + "epoch": 2.0587606837606836, + "grad_norm": 0.6910824775695801, + "learning_rate": 9.553629001063079e-05, + "loss": 0.9332, + "step": 11564 + }, + { + "epoch": 2.0589387464387463, + "grad_norm": 0.757247805595398, + "learning_rate": 9.552230653396566e-05, + "loss": 0.6598, + "step": 11565 + }, + { + "epoch": 2.059116809116809, + "grad_norm": 0.7778435349464417, + "learning_rate": 9.550832314503163e-05, + "loss": 0.8899, + "step": 11566 + }, + { + "epoch": 2.059294871794872, + "grad_norm": 0.7827669978141785, + "learning_rate": 9.54943398441027e-05, + "loss": 0.8036, + "step": 11567 + }, + { + "epoch": 2.0594729344729346, + "grad_norm": 0.7462462186813354, + "learning_rate": 9.54803566314528e-05, + "loss": 0.9306, + "step": 11568 + }, + { + "epoch": 2.0596509971509973, + "grad_norm": 0.8088639974594116, + "learning_rate": 9.546637350735597e-05, + "loss": 0.8766, + "step": 11569 + }, + { + "epoch": 2.0598290598290596, + "grad_norm": 0.6477743983268738, + "learning_rate": 9.545239047208607e-05, + "loss": 0.7239, + "step": 11570 + }, + { + "epoch": 2.0600071225071224, + "grad_norm": 0.7535004615783691, + "learning_rate": 9.543840752591718e-05, + "loss": 0.8891, + "step": 11571 + }, + { + "epoch": 2.060185185185185, + "grad_norm": 0.7085242867469788, + "learning_rate": 9.542442466912316e-05, + "loss": 0.8105, + "step": 11572 + }, + { + "epoch": 2.060363247863248, + "grad_norm": 0.8129137754440308, + "learning_rate": 9.541044190197811e-05, + "loss": 0.6955, + "step": 11573 + }, + { + "epoch": 2.0605413105413106, + "grad_norm": 0.7160677909851074, + "learning_rate": 9.539645922475586e-05, + "loss": 0.833, + "step": 11574 + }, + { + "epoch": 2.0607193732193734, + "grad_norm": 0.6983035206794739, + "learning_rate": 9.538247663773044e-05, + "loss": 0.6439, + "step": 11575 + }, + { + "epoch": 2.0608974358974357, + "grad_norm": 0.8732622861862183, + "learning_rate": 9.536849414117578e-05, + "loss": 0.7763, + "step": 11576 + }, + { + "epoch": 2.0610754985754984, + "grad_norm": 0.7745480537414551, + "learning_rate": 9.535451173536591e-05, + "loss": 0.8272, + "step": 11577 + }, + { + "epoch": 2.061253561253561, + "grad_norm": 0.8210037350654602, + "learning_rate": 9.53405294205747e-05, + "loss": 0.9539, + "step": 11578 + }, + { + "epoch": 2.061431623931624, + "grad_norm": 0.6742323637008667, + "learning_rate": 9.532654719707617e-05, + "loss": 0.6525, + "step": 11579 + }, + { + "epoch": 2.0616096866096867, + "grad_norm": 0.8312603831291199, + "learning_rate": 9.531256506514418e-05, + "loss": 0.7776, + "step": 11580 + }, + { + "epoch": 2.0617877492877494, + "grad_norm": 0.7817347049713135, + "learning_rate": 9.529858302505278e-05, + "loss": 1.0148, + "step": 11581 + }, + { + "epoch": 2.0619658119658117, + "grad_norm": 0.751153290271759, + "learning_rate": 9.528460107707584e-05, + "loss": 0.8064, + "step": 11582 + }, + { + "epoch": 2.0621438746438745, + "grad_norm": 0.7483627200126648, + "learning_rate": 9.527061922148737e-05, + "loss": 0.8706, + "step": 11583 + }, + { + "epoch": 2.0623219373219372, + "grad_norm": 0.7044979929924011, + "learning_rate": 9.525663745856132e-05, + "loss": 0.8008, + "step": 11584 + }, + { + "epoch": 2.0625, + "grad_norm": 0.8249054551124573, + "learning_rate": 9.524265578857157e-05, + "loss": 0.9339, + "step": 11585 + }, + { + "epoch": 2.0626780626780628, + "grad_norm": 0.7184668183326721, + "learning_rate": 9.522867421179211e-05, + "loss": 0.9191, + "step": 11586 + }, + { + "epoch": 2.0628561253561255, + "grad_norm": 0.8135001063346863, + "learning_rate": 9.521469272849685e-05, + "loss": 0.966, + "step": 11587 + }, + { + "epoch": 2.0630341880341883, + "grad_norm": 0.8151242733001709, + "learning_rate": 9.520071133895978e-05, + "loss": 0.7947, + "step": 11588 + }, + { + "epoch": 2.0632122507122506, + "grad_norm": 0.8044771552085876, + "learning_rate": 9.518673004345477e-05, + "loss": 0.893, + "step": 11589 + }, + { + "epoch": 2.0633903133903133, + "grad_norm": 0.730505645275116, + "learning_rate": 9.517274884225581e-05, + "loss": 0.8691, + "step": 11590 + }, + { + "epoch": 2.063568376068376, + "grad_norm": 0.7419933676719666, + "learning_rate": 9.515876773563678e-05, + "loss": 0.809, + "step": 11591 + }, + { + "epoch": 2.063746438746439, + "grad_norm": 0.7809683084487915, + "learning_rate": 9.514478672387169e-05, + "loss": 0.6926, + "step": 11592 + }, + { + "epoch": 2.0639245014245016, + "grad_norm": 1.0065315961837769, + "learning_rate": 9.513080580723435e-05, + "loss": 0.7506, + "step": 11593 + }, + { + "epoch": 2.064102564102564, + "grad_norm": 0.7424543499946594, + "learning_rate": 9.511682498599883e-05, + "loss": 0.9158, + "step": 11594 + }, + { + "epoch": 2.0642806267806266, + "grad_norm": 0.6907097697257996, + "learning_rate": 9.510284426043893e-05, + "loss": 0.8795, + "step": 11595 + }, + { + "epoch": 2.0644586894586894, + "grad_norm": 0.7849169969558716, + "learning_rate": 9.508886363082864e-05, + "loss": 1.0065, + "step": 11596 + }, + { + "epoch": 2.064636752136752, + "grad_norm": 0.7421438694000244, + "learning_rate": 9.507488309744183e-05, + "loss": 0.7574, + "step": 11597 + }, + { + "epoch": 2.064814814814815, + "grad_norm": 0.6636283993721008, + "learning_rate": 9.506090266055252e-05, + "loss": 0.7489, + "step": 11598 + }, + { + "epoch": 2.0649928774928776, + "grad_norm": 0.7133244872093201, + "learning_rate": 9.504692232043452e-05, + "loss": 0.7652, + "step": 11599 + }, + { + "epoch": 2.0651709401709404, + "grad_norm": 0.7891597151756287, + "learning_rate": 9.50329420773618e-05, + "loss": 0.7268, + "step": 11600 + }, + { + "epoch": 2.0653490028490027, + "grad_norm": 0.8578699827194214, + "learning_rate": 9.501896193160822e-05, + "loss": 0.9872, + "step": 11601 + }, + { + "epoch": 2.0655270655270654, + "grad_norm": 0.7071980834007263, + "learning_rate": 9.500498188344777e-05, + "loss": 0.7278, + "step": 11602 + }, + { + "epoch": 2.065705128205128, + "grad_norm": 0.8434318900108337, + "learning_rate": 9.499100193315436e-05, + "loss": 0.8811, + "step": 11603 + }, + { + "epoch": 2.065883190883191, + "grad_norm": 0.7429414391517639, + "learning_rate": 9.49770220810018e-05, + "loss": 0.8256, + "step": 11604 + }, + { + "epoch": 2.0660612535612537, + "grad_norm": 0.7059712409973145, + "learning_rate": 9.496304232726412e-05, + "loss": 0.8012, + "step": 11605 + }, + { + "epoch": 2.0662393162393164, + "grad_norm": 0.7095850706100464, + "learning_rate": 9.49490626722151e-05, + "loss": 0.8767, + "step": 11606 + }, + { + "epoch": 2.0664173789173788, + "grad_norm": 0.8135038018226624, + "learning_rate": 9.493508311612874e-05, + "loss": 1.1402, + "step": 11607 + }, + { + "epoch": 2.0665954415954415, + "grad_norm": 0.9023036360740662, + "learning_rate": 9.492110365927888e-05, + "loss": 0.7307, + "step": 11608 + }, + { + "epoch": 2.0667735042735043, + "grad_norm": 0.6990833282470703, + "learning_rate": 9.490712430193949e-05, + "loss": 0.7996, + "step": 11609 + }, + { + "epoch": 2.066951566951567, + "grad_norm": 0.7765957713127136, + "learning_rate": 9.489314504438437e-05, + "loss": 0.9721, + "step": 11610 + }, + { + "epoch": 2.0671296296296298, + "grad_norm": 0.883575975894928, + "learning_rate": 9.487916588688749e-05, + "loss": 0.8778, + "step": 11611 + }, + { + "epoch": 2.0673076923076925, + "grad_norm": 0.7226536870002747, + "learning_rate": 9.48651868297227e-05, + "loss": 1.0438, + "step": 11612 + }, + { + "epoch": 2.067485754985755, + "grad_norm": 0.7399018406867981, + "learning_rate": 9.485120787316394e-05, + "loss": 0.8154, + "step": 11613 + }, + { + "epoch": 2.0676638176638176, + "grad_norm": 1.0130186080932617, + "learning_rate": 9.483722901748502e-05, + "loss": 0.9832, + "step": 11614 + }, + { + "epoch": 2.0678418803418803, + "grad_norm": 0.8163331151008606, + "learning_rate": 9.482325026295993e-05, + "loss": 0.8854, + "step": 11615 + }, + { + "epoch": 2.068019943019943, + "grad_norm": 0.7165096998214722, + "learning_rate": 9.480927160986244e-05, + "loss": 0.7636, + "step": 11616 + }, + { + "epoch": 2.068198005698006, + "grad_norm": 0.8579450845718384, + "learning_rate": 9.479529305846652e-05, + "loss": 0.9105, + "step": 11617 + }, + { + "epoch": 2.0683760683760686, + "grad_norm": 0.7062679529190063, + "learning_rate": 9.4781314609046e-05, + "loss": 0.7601, + "step": 11618 + }, + { + "epoch": 2.068554131054131, + "grad_norm": 0.8178739547729492, + "learning_rate": 9.476733626187483e-05, + "loss": 0.8393, + "step": 11619 + }, + { + "epoch": 2.0687321937321936, + "grad_norm": 0.6667241454124451, + "learning_rate": 9.475335801722678e-05, + "loss": 0.8791, + "step": 11620 + }, + { + "epoch": 2.0689102564102564, + "grad_norm": 0.6603145599365234, + "learning_rate": 9.47393798753758e-05, + "loss": 0.7073, + "step": 11621 + }, + { + "epoch": 2.069088319088319, + "grad_norm": 0.7719821333885193, + "learning_rate": 9.472540183659573e-05, + "loss": 0.9195, + "step": 11622 + }, + { + "epoch": 2.069266381766382, + "grad_norm": 0.8059320449829102, + "learning_rate": 9.471142390116045e-05, + "loss": 0.927, + "step": 11623 + }, + { + "epoch": 2.0694444444444446, + "grad_norm": 0.6513992547988892, + "learning_rate": 9.469744606934388e-05, + "loss": 0.857, + "step": 11624 + }, + { + "epoch": 2.069622507122507, + "grad_norm": 0.6948497295379639, + "learning_rate": 9.468346834141979e-05, + "loss": 0.65, + "step": 11625 + }, + { + "epoch": 2.0698005698005697, + "grad_norm": 0.8086618781089783, + "learning_rate": 9.466949071766213e-05, + "loss": 0.8328, + "step": 11626 + }, + { + "epoch": 2.0699786324786325, + "grad_norm": 0.794731616973877, + "learning_rate": 9.465551319834468e-05, + "loss": 0.8641, + "step": 11627 + }, + { + "epoch": 2.070156695156695, + "grad_norm": 0.7312739491462708, + "learning_rate": 9.46415357837414e-05, + "loss": 0.7878, + "step": 11628 + }, + { + "epoch": 2.070334757834758, + "grad_norm": 0.8025211691856384, + "learning_rate": 9.462755847412606e-05, + "loss": 0.8624, + "step": 11629 + }, + { + "epoch": 2.0705128205128207, + "grad_norm": 0.7296801209449768, + "learning_rate": 9.461358126977259e-05, + "loss": 0.7299, + "step": 11630 + }, + { + "epoch": 2.070690883190883, + "grad_norm": 0.7176340222358704, + "learning_rate": 9.459960417095477e-05, + "loss": 0.7374, + "step": 11631 + }, + { + "epoch": 2.0708689458689458, + "grad_norm": 0.7656565308570862, + "learning_rate": 9.45856271779465e-05, + "loss": 0.7791, + "step": 11632 + }, + { + "epoch": 2.0710470085470085, + "grad_norm": 0.7232711315155029, + "learning_rate": 9.457165029102159e-05, + "loss": 0.85, + "step": 11633 + }, + { + "epoch": 2.0712250712250713, + "grad_norm": 0.7342440485954285, + "learning_rate": 9.455767351045397e-05, + "loss": 0.8423, + "step": 11634 + }, + { + "epoch": 2.071403133903134, + "grad_norm": 0.7844834923744202, + "learning_rate": 9.45436968365174e-05, + "loss": 0.9179, + "step": 11635 + }, + { + "epoch": 2.0715811965811968, + "grad_norm": 0.8880203366279602, + "learning_rate": 9.452972026948575e-05, + "loss": 0.7139, + "step": 11636 + }, + { + "epoch": 2.071759259259259, + "grad_norm": 0.7611206769943237, + "learning_rate": 9.451574380963286e-05, + "loss": 0.8915, + "step": 11637 + }, + { + "epoch": 2.071937321937322, + "grad_norm": 0.8123503923416138, + "learning_rate": 9.450176745723262e-05, + "loss": 0.7011, + "step": 11638 + }, + { + "epoch": 2.0721153846153846, + "grad_norm": 0.7703253030776978, + "learning_rate": 9.448779121255879e-05, + "loss": 0.8493, + "step": 11639 + }, + { + "epoch": 2.0722934472934473, + "grad_norm": 0.755836009979248, + "learning_rate": 9.447381507588527e-05, + "loss": 0.9145, + "step": 11640 + }, + { + "epoch": 2.07247150997151, + "grad_norm": 0.7879568338394165, + "learning_rate": 9.445983904748583e-05, + "loss": 0.7761, + "step": 11641 + }, + { + "epoch": 2.072649572649573, + "grad_norm": 0.6695574522018433, + "learning_rate": 9.444586312763434e-05, + "loss": 0.8594, + "step": 11642 + }, + { + "epoch": 2.072827635327635, + "grad_norm": 0.6734640002250671, + "learning_rate": 9.443188731660462e-05, + "loss": 0.7324, + "step": 11643 + }, + { + "epoch": 2.073005698005698, + "grad_norm": 0.7823841571807861, + "learning_rate": 9.441791161467051e-05, + "loss": 1.0223, + "step": 11644 + }, + { + "epoch": 2.0731837606837606, + "grad_norm": 0.8152045011520386, + "learning_rate": 9.440393602210585e-05, + "loss": 1.0364, + "step": 11645 + }, + { + "epoch": 2.0733618233618234, + "grad_norm": 0.8664864897727966, + "learning_rate": 9.438996053918441e-05, + "loss": 0.7607, + "step": 11646 + }, + { + "epoch": 2.073539886039886, + "grad_norm": 0.7949544787406921, + "learning_rate": 9.437598516618006e-05, + "loss": 0.7644, + "step": 11647 + }, + { + "epoch": 2.073717948717949, + "grad_norm": 0.767045259475708, + "learning_rate": 9.436200990336657e-05, + "loss": 0.7563, + "step": 11648 + }, + { + "epoch": 2.073896011396011, + "grad_norm": 0.669129490852356, + "learning_rate": 9.434803475101782e-05, + "loss": 0.7644, + "step": 11649 + }, + { + "epoch": 2.074074074074074, + "grad_norm": 0.7969587445259094, + "learning_rate": 9.433405970940755e-05, + "loss": 1.0249, + "step": 11650 + }, + { + "epoch": 2.0742521367521367, + "grad_norm": 0.6744855642318726, + "learning_rate": 9.432008477880966e-05, + "loss": 0.7478, + "step": 11651 + }, + { + "epoch": 2.0744301994301995, + "grad_norm": 0.6236920356750488, + "learning_rate": 9.430610995949786e-05, + "loss": 0.7309, + "step": 11652 + }, + { + "epoch": 2.074608262108262, + "grad_norm": 0.7952008843421936, + "learning_rate": 9.429213525174603e-05, + "loss": 0.927, + "step": 11653 + }, + { + "epoch": 2.074786324786325, + "grad_norm": 0.7075965404510498, + "learning_rate": 9.427816065582792e-05, + "loss": 0.8494, + "step": 11654 + }, + { + "epoch": 2.0749643874643873, + "grad_norm": 0.8018102049827576, + "learning_rate": 9.426418617201744e-05, + "loss": 0.9261, + "step": 11655 + }, + { + "epoch": 2.07514245014245, + "grad_norm": 0.7155446410179138, + "learning_rate": 9.425021180058824e-05, + "loss": 0.8296, + "step": 11656 + }, + { + "epoch": 2.0753205128205128, + "grad_norm": 0.6611294150352478, + "learning_rate": 9.423623754181425e-05, + "loss": 0.687, + "step": 11657 + }, + { + "epoch": 2.0754985754985755, + "grad_norm": 0.706280529499054, + "learning_rate": 9.422226339596917e-05, + "loss": 0.8664, + "step": 11658 + }, + { + "epoch": 2.0756766381766383, + "grad_norm": 0.7512072324752808, + "learning_rate": 9.420828936332687e-05, + "loss": 0.9074, + "step": 11659 + }, + { + "epoch": 2.075854700854701, + "grad_norm": 0.8833743333816528, + "learning_rate": 9.419431544416108e-05, + "loss": 1.0541, + "step": 11660 + }, + { + "epoch": 2.0760327635327633, + "grad_norm": 0.6991413235664368, + "learning_rate": 9.418034163874564e-05, + "loss": 0.7543, + "step": 11661 + }, + { + "epoch": 2.076210826210826, + "grad_norm": 0.784294605255127, + "learning_rate": 9.41663679473543e-05, + "loss": 0.8156, + "step": 11662 + }, + { + "epoch": 2.076388888888889, + "grad_norm": 0.7716241478919983, + "learning_rate": 9.415239437026086e-05, + "loss": 0.9613, + "step": 11663 + }, + { + "epoch": 2.0765669515669516, + "grad_norm": 0.8247698545455933, + "learning_rate": 9.413842090773914e-05, + "loss": 0.8811, + "step": 11664 + }, + { + "epoch": 2.0767450142450143, + "grad_norm": 0.7988204956054688, + "learning_rate": 9.412444756006283e-05, + "loss": 1.017, + "step": 11665 + }, + { + "epoch": 2.076923076923077, + "grad_norm": 0.8069472908973694, + "learning_rate": 9.411047432750583e-05, + "loss": 0.8292, + "step": 11666 + }, + { + "epoch": 2.0771011396011394, + "grad_norm": 0.8177345991134644, + "learning_rate": 9.40965012103418e-05, + "loss": 0.8004, + "step": 11667 + }, + { + "epoch": 2.077279202279202, + "grad_norm": 0.8589172959327698, + "learning_rate": 9.40825282088446e-05, + "loss": 0.8274, + "step": 11668 + }, + { + "epoch": 2.077457264957265, + "grad_norm": 0.677379846572876, + "learning_rate": 9.406855532328792e-05, + "loss": 0.7044, + "step": 11669 + }, + { + "epoch": 2.0776353276353277, + "grad_norm": 0.9417888522148132, + "learning_rate": 9.405458255394564e-05, + "loss": 0.8418, + "step": 11670 + }, + { + "epoch": 2.0778133903133904, + "grad_norm": 0.7226679921150208, + "learning_rate": 9.404060990109141e-05, + "loss": 0.8496, + "step": 11671 + }, + { + "epoch": 2.077991452991453, + "grad_norm": 0.7451614737510681, + "learning_rate": 9.402663736499909e-05, + "loss": 0.7569, + "step": 11672 + }, + { + "epoch": 2.0781695156695155, + "grad_norm": 0.6516944169998169, + "learning_rate": 9.401266494594235e-05, + "loss": 0.5591, + "step": 11673 + }, + { + "epoch": 2.078347578347578, + "grad_norm": 0.7473219633102417, + "learning_rate": 9.399869264419507e-05, + "loss": 1.0098, + "step": 11674 + }, + { + "epoch": 2.078525641025641, + "grad_norm": 0.8346691131591797, + "learning_rate": 9.398472046003088e-05, + "loss": 1.2654, + "step": 11675 + }, + { + "epoch": 2.0787037037037037, + "grad_norm": 0.8611979484558105, + "learning_rate": 9.397074839372366e-05, + "loss": 0.571, + "step": 11676 + }, + { + "epoch": 2.0788817663817665, + "grad_norm": 0.8093259334564209, + "learning_rate": 9.395677644554705e-05, + "loss": 1.0383, + "step": 11677 + }, + { + "epoch": 2.0790598290598292, + "grad_norm": 0.7954222559928894, + "learning_rate": 9.394280461577488e-05, + "loss": 0.8078, + "step": 11678 + }, + { + "epoch": 2.0792378917378915, + "grad_norm": 0.8380635380744934, + "learning_rate": 9.392883290468083e-05, + "loss": 1.0368, + "step": 11679 + }, + { + "epoch": 2.0794159544159543, + "grad_norm": 0.8427146077156067, + "learning_rate": 9.391486131253874e-05, + "loss": 0.8638, + "step": 11680 + }, + { + "epoch": 2.079594017094017, + "grad_norm": 0.7211564779281616, + "learning_rate": 9.390088983962227e-05, + "loss": 0.8211, + "step": 11681 + }, + { + "epoch": 2.07977207977208, + "grad_norm": 0.7480773329734802, + "learning_rate": 9.388691848620517e-05, + "loss": 0.9313, + "step": 11682 + }, + { + "epoch": 2.0799501424501425, + "grad_norm": 0.8421902060508728, + "learning_rate": 9.387294725256123e-05, + "loss": 0.8808, + "step": 11683 + }, + { + "epoch": 2.0801282051282053, + "grad_norm": 0.7753815650939941, + "learning_rate": 9.385897613896416e-05, + "loss": 0.7738, + "step": 11684 + }, + { + "epoch": 2.0803062678062676, + "grad_norm": 0.8053030967712402, + "learning_rate": 9.384500514568773e-05, + "loss": 0.9194, + "step": 11685 + }, + { + "epoch": 2.0804843304843303, + "grad_norm": 0.7628602981567383, + "learning_rate": 9.383103427300559e-05, + "loss": 0.7866, + "step": 11686 + }, + { + "epoch": 2.080662393162393, + "grad_norm": 0.7087932825088501, + "learning_rate": 9.381706352119156e-05, + "loss": 0.8817, + "step": 11687 + }, + { + "epoch": 2.080840455840456, + "grad_norm": 0.8687152862548828, + "learning_rate": 9.380309289051929e-05, + "loss": 0.8902, + "step": 11688 + }, + { + "epoch": 2.0810185185185186, + "grad_norm": 0.8181152939796448, + "learning_rate": 9.378912238126256e-05, + "loss": 0.9244, + "step": 11689 + }, + { + "epoch": 2.0811965811965814, + "grad_norm": 0.7961983680725098, + "learning_rate": 9.377515199369506e-05, + "loss": 0.714, + "step": 11690 + }, + { + "epoch": 2.0813746438746437, + "grad_norm": 0.8307793736457825, + "learning_rate": 9.376118172809056e-05, + "loss": 0.9573, + "step": 11691 + }, + { + "epoch": 2.0815527065527064, + "grad_norm": 0.7349256277084351, + "learning_rate": 9.374721158472269e-05, + "loss": 0.7533, + "step": 11692 + }, + { + "epoch": 2.081730769230769, + "grad_norm": 0.7625117897987366, + "learning_rate": 9.373324156386526e-05, + "loss": 0.8387, + "step": 11693 + }, + { + "epoch": 2.081908831908832, + "grad_norm": 0.9537683129310608, + "learning_rate": 9.371927166579191e-05, + "loss": 0.9444, + "step": 11694 + }, + { + "epoch": 2.0820868945868947, + "grad_norm": 0.7170497179031372, + "learning_rate": 9.370530189077644e-05, + "loss": 0.9132, + "step": 11695 + }, + { + "epoch": 2.0822649572649574, + "grad_norm": 0.7750041484832764, + "learning_rate": 9.369133223909246e-05, + "loss": 0.6635, + "step": 11696 + }, + { + "epoch": 2.08244301994302, + "grad_norm": 0.8990386128425598, + "learning_rate": 9.367736271101373e-05, + "loss": 0.8692, + "step": 11697 + }, + { + "epoch": 2.0826210826210825, + "grad_norm": 0.5909343361854553, + "learning_rate": 9.366339330681393e-05, + "loss": 0.6811, + "step": 11698 + }, + { + "epoch": 2.0827991452991452, + "grad_norm": 0.7783302068710327, + "learning_rate": 9.364942402676682e-05, + "loss": 1.1024, + "step": 11699 + }, + { + "epoch": 2.082977207977208, + "grad_norm": 0.8926466703414917, + "learning_rate": 9.3635454871146e-05, + "loss": 0.971, + "step": 11700 + }, + { + "epoch": 2.0831552706552707, + "grad_norm": 0.7374816536903381, + "learning_rate": 9.362148584022527e-05, + "loss": 0.7151, + "step": 11701 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 0.7491161227226257, + "learning_rate": 9.360751693427823e-05, + "loss": 0.9213, + "step": 11702 + }, + { + "epoch": 2.083511396011396, + "grad_norm": 0.726859986782074, + "learning_rate": 9.359354815357862e-05, + "loss": 0.8412, + "step": 11703 + }, + { + "epoch": 2.0836894586894585, + "grad_norm": 0.756703794002533, + "learning_rate": 9.357957949840015e-05, + "loss": 0.8074, + "step": 11704 + }, + { + "epoch": 2.0838675213675213, + "grad_norm": 0.8475984334945679, + "learning_rate": 9.356561096901646e-05, + "loss": 0.8926, + "step": 11705 + }, + { + "epoch": 2.084045584045584, + "grad_norm": 0.9776971936225891, + "learning_rate": 9.355164256570129e-05, + "loss": 0.8543, + "step": 11706 + }, + { + "epoch": 2.084223646723647, + "grad_norm": 0.7185834646224976, + "learning_rate": 9.353767428872826e-05, + "loss": 0.6946, + "step": 11707 + }, + { + "epoch": 2.0844017094017095, + "grad_norm": 0.7075535655021667, + "learning_rate": 9.352370613837109e-05, + "loss": 0.7171, + "step": 11708 + }, + { + "epoch": 2.0845797720797723, + "grad_norm": 0.8549726009368896, + "learning_rate": 9.350973811490343e-05, + "loss": 0.9028, + "step": 11709 + }, + { + "epoch": 2.0847578347578346, + "grad_norm": 0.731235682964325, + "learning_rate": 9.3495770218599e-05, + "loss": 0.7703, + "step": 11710 + }, + { + "epoch": 2.0849358974358974, + "grad_norm": 0.8660612106323242, + "learning_rate": 9.34818024497314e-05, + "loss": 0.8464, + "step": 11711 + }, + { + "epoch": 2.08511396011396, + "grad_norm": 0.7687711715698242, + "learning_rate": 9.346783480857439e-05, + "loss": 0.8199, + "step": 11712 + }, + { + "epoch": 2.085292022792023, + "grad_norm": 0.6802884936332703, + "learning_rate": 9.345386729540155e-05, + "loss": 0.7537, + "step": 11713 + }, + { + "epoch": 2.0854700854700856, + "grad_norm": 0.7688863277435303, + "learning_rate": 9.34398999104866e-05, + "loss": 0.8374, + "step": 11714 + }, + { + "epoch": 2.0856481481481484, + "grad_norm": 0.7872602939605713, + "learning_rate": 9.342593265410315e-05, + "loss": 0.8786, + "step": 11715 + }, + { + "epoch": 2.0858262108262107, + "grad_norm": 0.9752106666564941, + "learning_rate": 9.341196552652496e-05, + "loss": 0.9572, + "step": 11716 + }, + { + "epoch": 2.0860042735042734, + "grad_norm": 0.7023422718048096, + "learning_rate": 9.339799852802555e-05, + "loss": 0.7613, + "step": 11717 + }, + { + "epoch": 2.086182336182336, + "grad_norm": 0.8366875052452087, + "learning_rate": 9.338403165887868e-05, + "loss": 0.8206, + "step": 11718 + }, + { + "epoch": 2.086360398860399, + "grad_norm": 0.8534985184669495, + "learning_rate": 9.337006491935794e-05, + "loss": 0.8549, + "step": 11719 + }, + { + "epoch": 2.0865384615384617, + "grad_norm": 0.7902935743331909, + "learning_rate": 9.335609830973707e-05, + "loss": 0.8399, + "step": 11720 + }, + { + "epoch": 2.0867165242165244, + "grad_norm": 0.8064647316932678, + "learning_rate": 9.334213183028958e-05, + "loss": 0.7978, + "step": 11721 + }, + { + "epoch": 2.0868945868945867, + "grad_norm": 0.816412627696991, + "learning_rate": 9.332816548128919e-05, + "loss": 0.8814, + "step": 11722 + }, + { + "epoch": 2.0870726495726495, + "grad_norm": 0.7778908610343933, + "learning_rate": 9.33141992630096e-05, + "loss": 0.9916, + "step": 11723 + }, + { + "epoch": 2.0872507122507122, + "grad_norm": 0.7899400591850281, + "learning_rate": 9.330023317572433e-05, + "loss": 0.5682, + "step": 11724 + }, + { + "epoch": 2.087428774928775, + "grad_norm": 0.6770033836364746, + "learning_rate": 9.32862672197071e-05, + "loss": 0.7327, + "step": 11725 + }, + { + "epoch": 2.0876068376068377, + "grad_norm": 0.8385946750640869, + "learning_rate": 9.327230139523148e-05, + "loss": 0.7793, + "step": 11726 + }, + { + "epoch": 2.0877849002849005, + "grad_norm": 0.708091139793396, + "learning_rate": 9.32583357025712e-05, + "loss": 0.6199, + "step": 11727 + }, + { + "epoch": 2.087962962962963, + "grad_norm": 0.8172122836112976, + "learning_rate": 9.324437014199978e-05, + "loss": 0.749, + "step": 11728 + }, + { + "epoch": 2.0881410256410255, + "grad_norm": 0.818324625492096, + "learning_rate": 9.323040471379091e-05, + "loss": 0.6605, + "step": 11729 + }, + { + "epoch": 2.0883190883190883, + "grad_norm": 0.9074803590774536, + "learning_rate": 9.321643941821819e-05, + "loss": 0.9231, + "step": 11730 + }, + { + "epoch": 2.088497150997151, + "grad_norm": 0.7559560537338257, + "learning_rate": 9.320247425555527e-05, + "loss": 0.7291, + "step": 11731 + }, + { + "epoch": 2.088675213675214, + "grad_norm": 0.8001563549041748, + "learning_rate": 9.318850922607571e-05, + "loss": 0.7781, + "step": 11732 + }, + { + "epoch": 2.0888532763532766, + "grad_norm": 0.7365888953208923, + "learning_rate": 9.31745443300532e-05, + "loss": 0.8243, + "step": 11733 + }, + { + "epoch": 2.089031339031339, + "grad_norm": 0.7861692309379578, + "learning_rate": 9.316057956776126e-05, + "loss": 0.7568, + "step": 11734 + }, + { + "epoch": 2.0892094017094016, + "grad_norm": 0.8399034738540649, + "learning_rate": 9.314661493947363e-05, + "loss": 0.8019, + "step": 11735 + }, + { + "epoch": 2.0893874643874644, + "grad_norm": 0.7718507051467896, + "learning_rate": 9.313265044546378e-05, + "loss": 0.909, + "step": 11736 + }, + { + "epoch": 2.089565527065527, + "grad_norm": 0.8940733671188354, + "learning_rate": 9.311868608600543e-05, + "loss": 0.7154, + "step": 11737 + }, + { + "epoch": 2.08974358974359, + "grad_norm": 0.8506718873977661, + "learning_rate": 9.31047218613721e-05, + "loss": 0.8367, + "step": 11738 + }, + { + "epoch": 2.0899216524216526, + "grad_norm": 0.8431367874145508, + "learning_rate": 9.309075777183743e-05, + "loss": 0.9532, + "step": 11739 + }, + { + "epoch": 2.090099715099715, + "grad_norm": 0.7683414220809937, + "learning_rate": 9.307679381767499e-05, + "loss": 0.9301, + "step": 11740 + }, + { + "epoch": 2.0902777777777777, + "grad_norm": 0.7601380348205566, + "learning_rate": 9.306282999915839e-05, + "loss": 0.8462, + "step": 11741 + }, + { + "epoch": 2.0904558404558404, + "grad_norm": 0.7531782388687134, + "learning_rate": 9.304886631656127e-05, + "loss": 0.8012, + "step": 11742 + }, + { + "epoch": 2.090633903133903, + "grad_norm": 0.7869617938995361, + "learning_rate": 9.303490277015714e-05, + "loss": 0.6645, + "step": 11743 + }, + { + "epoch": 2.090811965811966, + "grad_norm": 0.8042751550674438, + "learning_rate": 9.302093936021964e-05, + "loss": 1.1078, + "step": 11744 + }, + { + "epoch": 2.0909900284900287, + "grad_norm": 0.750350296497345, + "learning_rate": 9.300697608702231e-05, + "loss": 0.8552, + "step": 11745 + }, + { + "epoch": 2.091168091168091, + "grad_norm": 0.7624406814575195, + "learning_rate": 9.29930129508388e-05, + "loss": 0.861, + "step": 11746 + }, + { + "epoch": 2.0913461538461537, + "grad_norm": 0.7634474635124207, + "learning_rate": 9.29790499519426e-05, + "loss": 0.9483, + "step": 11747 + }, + { + "epoch": 2.0915242165242165, + "grad_norm": 0.7312899231910706, + "learning_rate": 9.296508709060738e-05, + "loss": 0.655, + "step": 11748 + }, + { + "epoch": 2.0917022792022792, + "grad_norm": 0.8181857466697693, + "learning_rate": 9.295112436710662e-05, + "loss": 0.7912, + "step": 11749 + }, + { + "epoch": 2.091880341880342, + "grad_norm": 0.6349542737007141, + "learning_rate": 9.293716178171396e-05, + "loss": 0.6268, + "step": 11750 + }, + { + "epoch": 2.0920584045584047, + "grad_norm": 0.8832548260688782, + "learning_rate": 9.292319933470291e-05, + "loss": 0.7805, + "step": 11751 + }, + { + "epoch": 2.092236467236467, + "grad_norm": 0.7251408100128174, + "learning_rate": 9.290923702634712e-05, + "loss": 0.7553, + "step": 11752 + }, + { + "epoch": 2.09241452991453, + "grad_norm": 0.8794457912445068, + "learning_rate": 9.289527485692006e-05, + "loss": 0.9187, + "step": 11753 + }, + { + "epoch": 2.0925925925925926, + "grad_norm": 0.7768839597702026, + "learning_rate": 9.288131282669534e-05, + "loss": 0.9267, + "step": 11754 + }, + { + "epoch": 2.0927706552706553, + "grad_norm": 0.744144856929779, + "learning_rate": 9.28673509359465e-05, + "loss": 0.714, + "step": 11755 + }, + { + "epoch": 2.092948717948718, + "grad_norm": 0.9117433428764343, + "learning_rate": 9.285338918494714e-05, + "loss": 0.9965, + "step": 11756 + }, + { + "epoch": 2.093126780626781, + "grad_norm": 0.8105267286300659, + "learning_rate": 9.283942757397073e-05, + "loss": 0.7517, + "step": 11757 + }, + { + "epoch": 2.093304843304843, + "grad_norm": 0.7348153591156006, + "learning_rate": 9.28254661032909e-05, + "loss": 0.7101, + "step": 11758 + }, + { + "epoch": 2.093482905982906, + "grad_norm": 0.7625702023506165, + "learning_rate": 9.281150477318113e-05, + "loss": 0.6863, + "step": 11759 + }, + { + "epoch": 2.0936609686609686, + "grad_norm": 0.7987569570541382, + "learning_rate": 9.2797543583915e-05, + "loss": 0.8848, + "step": 11760 + }, + { + "epoch": 2.0938390313390314, + "grad_norm": 0.706235408782959, + "learning_rate": 9.278358253576601e-05, + "loss": 0.7375, + "step": 11761 + }, + { + "epoch": 2.094017094017094, + "grad_norm": 0.9716742038726807, + "learning_rate": 9.276962162900774e-05, + "loss": 0.8602, + "step": 11762 + }, + { + "epoch": 2.094195156695157, + "grad_norm": 0.7711777687072754, + "learning_rate": 9.275566086391377e-05, + "loss": 0.8553, + "step": 11763 + }, + { + "epoch": 2.094373219373219, + "grad_norm": 0.8542511463165283, + "learning_rate": 9.274170024075751e-05, + "loss": 0.8412, + "step": 11764 + }, + { + "epoch": 2.094551282051282, + "grad_norm": 0.8255360126495361, + "learning_rate": 9.272773975981259e-05, + "loss": 1.0245, + "step": 11765 + }, + { + "epoch": 2.0947293447293447, + "grad_norm": 0.7416045665740967, + "learning_rate": 9.271377942135248e-05, + "loss": 0.57, + "step": 11766 + }, + { + "epoch": 2.0949074074074074, + "grad_norm": 0.8805620670318604, + "learning_rate": 9.269981922565078e-05, + "loss": 1.0262, + "step": 11767 + }, + { + "epoch": 2.09508547008547, + "grad_norm": 0.7293491363525391, + "learning_rate": 9.26858591729809e-05, + "loss": 0.7945, + "step": 11768 + }, + { + "epoch": 2.095263532763533, + "grad_norm": 0.7949206233024597, + "learning_rate": 9.267189926361643e-05, + "loss": 0.7071, + "step": 11769 + }, + { + "epoch": 2.0954415954415953, + "grad_norm": 0.771806538105011, + "learning_rate": 9.265793949783087e-05, + "loss": 0.8125, + "step": 11770 + }, + { + "epoch": 2.095619658119658, + "grad_norm": 0.7256866693496704, + "learning_rate": 9.264397987589776e-05, + "loss": 0.7607, + "step": 11771 + }, + { + "epoch": 2.0957977207977208, + "grad_norm": 0.8175343871116638, + "learning_rate": 9.263002039809055e-05, + "loss": 0.8486, + "step": 11772 + }, + { + "epoch": 2.0959757834757835, + "grad_norm": 0.7618881464004517, + "learning_rate": 9.261606106468282e-05, + "loss": 0.8182, + "step": 11773 + }, + { + "epoch": 2.0961538461538463, + "grad_norm": 0.7574927806854248, + "learning_rate": 9.2602101875948e-05, + "loss": 0.8703, + "step": 11774 + }, + { + "epoch": 2.096331908831909, + "grad_norm": 0.8639108538627625, + "learning_rate": 9.258814283215964e-05, + "loss": 0.9044, + "step": 11775 + }, + { + "epoch": 2.0965099715099713, + "grad_norm": 0.7221997380256653, + "learning_rate": 9.25741839335912e-05, + "loss": 0.7599, + "step": 11776 + }, + { + "epoch": 2.096688034188034, + "grad_norm": 0.9379764795303345, + "learning_rate": 9.256022518051626e-05, + "loss": 1.0002, + "step": 11777 + }, + { + "epoch": 2.096866096866097, + "grad_norm": 0.8430935740470886, + "learning_rate": 9.25462665732082e-05, + "loss": 0.7711, + "step": 11778 + }, + { + "epoch": 2.0970441595441596, + "grad_norm": 0.8371061086654663, + "learning_rate": 9.253230811194057e-05, + "loss": 0.9028, + "step": 11779 + }, + { + "epoch": 2.0972222222222223, + "grad_norm": 0.6960258483886719, + "learning_rate": 9.251834979698684e-05, + "loss": 0.7491, + "step": 11780 + }, + { + "epoch": 2.097400284900285, + "grad_norm": 0.7736398577690125, + "learning_rate": 9.25043916286205e-05, + "loss": 0.8985, + "step": 11781 + }, + { + "epoch": 2.0975783475783474, + "grad_norm": 0.6901512145996094, + "learning_rate": 9.249043360711509e-05, + "loss": 0.5881, + "step": 11782 + }, + { + "epoch": 2.09775641025641, + "grad_norm": 0.6741603016853333, + "learning_rate": 9.247647573274397e-05, + "loss": 0.7641, + "step": 11783 + }, + { + "epoch": 2.097934472934473, + "grad_norm": 0.736657440662384, + "learning_rate": 9.246251800578074e-05, + "loss": 0.8286, + "step": 11784 + }, + { + "epoch": 2.0981125356125356, + "grad_norm": 0.8235752582550049, + "learning_rate": 9.244856042649877e-05, + "loss": 0.8835, + "step": 11785 + }, + { + "epoch": 2.0982905982905984, + "grad_norm": 0.8083409667015076, + "learning_rate": 9.243460299517158e-05, + "loss": 0.9032, + "step": 11786 + }, + { + "epoch": 2.098468660968661, + "grad_norm": 0.7650952339172363, + "learning_rate": 9.242064571207262e-05, + "loss": 0.775, + "step": 11787 + }, + { + "epoch": 2.0986467236467234, + "grad_norm": 0.7961280345916748, + "learning_rate": 9.24066885774754e-05, + "loss": 0.6308, + "step": 11788 + }, + { + "epoch": 2.098824786324786, + "grad_norm": 0.8032481670379639, + "learning_rate": 9.23927315916533e-05, + "loss": 0.7544, + "step": 11789 + }, + { + "epoch": 2.099002849002849, + "grad_norm": 0.7452995777130127, + "learning_rate": 9.237877475487984e-05, + "loss": 0.8573, + "step": 11790 + }, + { + "epoch": 2.0991809116809117, + "grad_norm": 0.8141751289367676, + "learning_rate": 9.236481806742844e-05, + "loss": 0.9055, + "step": 11791 + }, + { + "epoch": 2.0993589743589745, + "grad_norm": 0.7862252593040466, + "learning_rate": 9.235086152957261e-05, + "loss": 0.6967, + "step": 11792 + }, + { + "epoch": 2.099537037037037, + "grad_norm": 0.771587073802948, + "learning_rate": 9.233690514158571e-05, + "loss": 0.7544, + "step": 11793 + }, + { + "epoch": 2.0997150997150995, + "grad_norm": 0.851445198059082, + "learning_rate": 9.23229489037413e-05, + "loss": 0.9249, + "step": 11794 + }, + { + "epoch": 2.0998931623931623, + "grad_norm": 0.7483612895011902, + "learning_rate": 9.23089928163127e-05, + "loss": 0.747, + "step": 11795 + }, + { + "epoch": 2.100071225071225, + "grad_norm": 0.8493219017982483, + "learning_rate": 9.229503687957342e-05, + "loss": 0.8898, + "step": 11796 + }, + { + "epoch": 2.1002492877492878, + "grad_norm": 0.8331718444824219, + "learning_rate": 9.228108109379687e-05, + "loss": 0.8943, + "step": 11797 + }, + { + "epoch": 2.1004273504273505, + "grad_norm": 0.7756054997444153, + "learning_rate": 9.226712545925655e-05, + "loss": 0.8586, + "step": 11798 + }, + { + "epoch": 2.1006054131054133, + "grad_norm": 0.7292607426643372, + "learning_rate": 9.225316997622579e-05, + "loss": 0.7591, + "step": 11799 + }, + { + "epoch": 2.1007834757834756, + "grad_norm": 0.8575723767280579, + "learning_rate": 9.223921464497811e-05, + "loss": 1.0147, + "step": 11800 + }, + { + "epoch": 2.1009615384615383, + "grad_norm": 0.7882707118988037, + "learning_rate": 9.222525946578687e-05, + "loss": 0.8297, + "step": 11801 + }, + { + "epoch": 2.101139601139601, + "grad_norm": 0.7982630729675293, + "learning_rate": 9.221130443892551e-05, + "loss": 0.9308, + "step": 11802 + }, + { + "epoch": 2.101317663817664, + "grad_norm": 0.7577962279319763, + "learning_rate": 9.219734956466752e-05, + "loss": 0.8474, + "step": 11803 + }, + { + "epoch": 2.1014957264957266, + "grad_norm": 0.7103776335716248, + "learning_rate": 9.218339484328621e-05, + "loss": 0.7863, + "step": 11804 + }, + { + "epoch": 2.1016737891737893, + "grad_norm": 0.8307296633720398, + "learning_rate": 9.216944027505505e-05, + "loss": 0.7633, + "step": 11805 + }, + { + "epoch": 2.1018518518518516, + "grad_norm": 0.8197653293609619, + "learning_rate": 9.215548586024743e-05, + "loss": 0.8987, + "step": 11806 + }, + { + "epoch": 2.1020299145299144, + "grad_norm": 0.9192719459533691, + "learning_rate": 9.21415315991368e-05, + "loss": 0.7829, + "step": 11807 + }, + { + "epoch": 2.102207977207977, + "grad_norm": 0.7249892354011536, + "learning_rate": 9.21275774919965e-05, + "loss": 0.9143, + "step": 11808 + }, + { + "epoch": 2.10238603988604, + "grad_norm": 0.7942582368850708, + "learning_rate": 9.211362353910002e-05, + "loss": 0.8634, + "step": 11809 + }, + { + "epoch": 2.1025641025641026, + "grad_norm": 0.7773341536521912, + "learning_rate": 9.209966974072065e-05, + "loss": 0.7865, + "step": 11810 + }, + { + "epoch": 2.1027421652421654, + "grad_norm": 0.802175760269165, + "learning_rate": 9.208571609713185e-05, + "loss": 0.7473, + "step": 11811 + }, + { + "epoch": 2.1029202279202277, + "grad_norm": 1.0248547792434692, + "learning_rate": 9.207176260860701e-05, + "loss": 1.0097, + "step": 11812 + }, + { + "epoch": 2.1030982905982905, + "grad_norm": 0.5781275629997253, + "learning_rate": 9.205780927541954e-05, + "loss": 0.5813, + "step": 11813 + }, + { + "epoch": 2.103276353276353, + "grad_norm": 0.7252389192581177, + "learning_rate": 9.204385609784274e-05, + "loss": 0.7978, + "step": 11814 + }, + { + "epoch": 2.103454415954416, + "grad_norm": 0.8497771620750427, + "learning_rate": 9.20299030761501e-05, + "loss": 0.95, + "step": 11815 + }, + { + "epoch": 2.1036324786324787, + "grad_norm": 0.8420650362968445, + "learning_rate": 9.201595021061491e-05, + "loss": 0.9693, + "step": 11816 + }, + { + "epoch": 2.1038105413105415, + "grad_norm": 0.8286302089691162, + "learning_rate": 9.200199750151063e-05, + "loss": 0.9457, + "step": 11817 + }, + { + "epoch": 2.103988603988604, + "grad_norm": 0.877740204334259, + "learning_rate": 9.198804494911057e-05, + "loss": 0.9082, + "step": 11818 + }, + { + "epoch": 2.1041666666666665, + "grad_norm": 0.7579863667488098, + "learning_rate": 9.197409255368817e-05, + "loss": 0.7681, + "step": 11819 + }, + { + "epoch": 2.1043447293447293, + "grad_norm": 0.7141458988189697, + "learning_rate": 9.19601403155167e-05, + "loss": 0.659, + "step": 11820 + }, + { + "epoch": 2.104522792022792, + "grad_norm": 0.8493850827217102, + "learning_rate": 9.194618823486958e-05, + "loss": 0.8197, + "step": 11821 + }, + { + "epoch": 2.1047008547008548, + "grad_norm": 0.8319337368011475, + "learning_rate": 9.193223631202019e-05, + "loss": 0.8955, + "step": 11822 + }, + { + "epoch": 2.1048789173789175, + "grad_norm": 0.7180153727531433, + "learning_rate": 9.191828454724186e-05, + "loss": 0.8068, + "step": 11823 + }, + { + "epoch": 2.10505698005698, + "grad_norm": 0.6748450398445129, + "learning_rate": 9.190433294080799e-05, + "loss": 0.7469, + "step": 11824 + }, + { + "epoch": 2.1052350427350426, + "grad_norm": 0.7750198841094971, + "learning_rate": 9.189038149299186e-05, + "loss": 0.932, + "step": 11825 + }, + { + "epoch": 2.1054131054131053, + "grad_norm": 0.7763389945030212, + "learning_rate": 9.187643020406688e-05, + "loss": 0.8027, + "step": 11826 + }, + { + "epoch": 2.105591168091168, + "grad_norm": 0.8382455110549927, + "learning_rate": 9.186247907430636e-05, + "loss": 0.8288, + "step": 11827 + }, + { + "epoch": 2.105769230769231, + "grad_norm": 0.6744221448898315, + "learning_rate": 9.184852810398367e-05, + "loss": 0.6807, + "step": 11828 + }, + { + "epoch": 2.1059472934472936, + "grad_norm": 0.7798452377319336, + "learning_rate": 9.183457729337212e-05, + "loss": 0.9853, + "step": 11829 + }, + { + "epoch": 2.1061253561253563, + "grad_norm": 0.7377058863639832, + "learning_rate": 9.182062664274513e-05, + "loss": 0.9043, + "step": 11830 + }, + { + "epoch": 2.1063034188034186, + "grad_norm": 0.8190791010856628, + "learning_rate": 9.180667615237589e-05, + "loss": 0.9786, + "step": 11831 + }, + { + "epoch": 2.1064814814814814, + "grad_norm": 0.7629963755607605, + "learning_rate": 9.179272582253785e-05, + "loss": 0.9168, + "step": 11832 + }, + { + "epoch": 2.106659544159544, + "grad_norm": 0.7753663063049316, + "learning_rate": 9.177877565350426e-05, + "loss": 0.963, + "step": 11833 + }, + { + "epoch": 2.106837606837607, + "grad_norm": 0.7842921614646912, + "learning_rate": 9.176482564554855e-05, + "loss": 0.8194, + "step": 11834 + }, + { + "epoch": 2.1070156695156697, + "grad_norm": 0.6640288233757019, + "learning_rate": 9.175087579894393e-05, + "loss": 0.6227, + "step": 11835 + }, + { + "epoch": 2.1071937321937324, + "grad_norm": 0.8474540710449219, + "learning_rate": 9.173692611396376e-05, + "loss": 0.7817, + "step": 11836 + }, + { + "epoch": 2.1073717948717947, + "grad_norm": 0.7123007774353027, + "learning_rate": 9.172297659088135e-05, + "loss": 0.9508, + "step": 11837 + }, + { + "epoch": 2.1075498575498575, + "grad_norm": 0.7418060898780823, + "learning_rate": 9.170902722997007e-05, + "loss": 0.8832, + "step": 11838 + }, + { + "epoch": 2.10772792022792, + "grad_norm": 0.7899464964866638, + "learning_rate": 9.169507803150313e-05, + "loss": 0.8474, + "step": 11839 + }, + { + "epoch": 2.107905982905983, + "grad_norm": 0.7543701529502869, + "learning_rate": 9.168112899575388e-05, + "loss": 0.8113, + "step": 11840 + }, + { + "epoch": 2.1080840455840457, + "grad_norm": 0.8057922720909119, + "learning_rate": 9.166718012299565e-05, + "loss": 0.998, + "step": 11841 + }, + { + "epoch": 2.1082621082621085, + "grad_norm": 0.7879176139831543, + "learning_rate": 9.16532314135017e-05, + "loss": 1.0509, + "step": 11842 + }, + { + "epoch": 2.1084401709401708, + "grad_norm": 0.8796642422676086, + "learning_rate": 9.163928286754537e-05, + "loss": 1.0481, + "step": 11843 + }, + { + "epoch": 2.1086182336182335, + "grad_norm": 0.7158889174461365, + "learning_rate": 9.16253344853999e-05, + "loss": 0.796, + "step": 11844 + }, + { + "epoch": 2.1087962962962963, + "grad_norm": 0.8020899295806885, + "learning_rate": 9.161138626733863e-05, + "loss": 0.822, + "step": 11845 + }, + { + "epoch": 2.108974358974359, + "grad_norm": 0.7217469215393066, + "learning_rate": 9.159743821363478e-05, + "loss": 1.0037, + "step": 11846 + }, + { + "epoch": 2.109152421652422, + "grad_norm": 0.762450098991394, + "learning_rate": 9.158349032456171e-05, + "loss": 1.0047, + "step": 11847 + }, + { + "epoch": 2.1093304843304845, + "grad_norm": 0.7227019667625427, + "learning_rate": 9.156954260039263e-05, + "loss": 0.8034, + "step": 11848 + }, + { + "epoch": 2.109508547008547, + "grad_norm": 0.7358957529067993, + "learning_rate": 9.155559504140089e-05, + "loss": 0.9483, + "step": 11849 + }, + { + "epoch": 2.1096866096866096, + "grad_norm": 0.7039931416511536, + "learning_rate": 9.154164764785968e-05, + "loss": 0.9255, + "step": 11850 + }, + { + "epoch": 2.1098646723646723, + "grad_norm": 0.8479618430137634, + "learning_rate": 9.152770042004234e-05, + "loss": 0.7379, + "step": 11851 + }, + { + "epoch": 2.110042735042735, + "grad_norm": 0.8320785164833069, + "learning_rate": 9.151375335822208e-05, + "loss": 0.944, + "step": 11852 + }, + { + "epoch": 2.110220797720798, + "grad_norm": 0.8186322450637817, + "learning_rate": 9.149980646267225e-05, + "loss": 0.7757, + "step": 11853 + }, + { + "epoch": 2.1103988603988606, + "grad_norm": 0.7816671133041382, + "learning_rate": 9.148585973366601e-05, + "loss": 0.8592, + "step": 11854 + }, + { + "epoch": 2.110576923076923, + "grad_norm": 0.8747152090072632, + "learning_rate": 9.147191317147671e-05, + "loss": 1.0852, + "step": 11855 + }, + { + "epoch": 2.1107549857549857, + "grad_norm": 0.7762712240219116, + "learning_rate": 9.14579667763775e-05, + "loss": 0.8466, + "step": 11856 + }, + { + "epoch": 2.1109330484330484, + "grad_norm": 0.8426344394683838, + "learning_rate": 9.144402054864171e-05, + "loss": 0.9949, + "step": 11857 + }, + { + "epoch": 2.111111111111111, + "grad_norm": 0.7581121921539307, + "learning_rate": 9.143007448854256e-05, + "loss": 0.748, + "step": 11858 + }, + { + "epoch": 2.111289173789174, + "grad_norm": 0.837939977645874, + "learning_rate": 9.141612859635333e-05, + "loss": 0.9479, + "step": 11859 + }, + { + "epoch": 2.1114672364672367, + "grad_norm": 0.7402070760726929, + "learning_rate": 9.140218287234718e-05, + "loss": 0.7829, + "step": 11860 + }, + { + "epoch": 2.111645299145299, + "grad_norm": 0.7125605344772339, + "learning_rate": 9.13882373167974e-05, + "loss": 1.0175, + "step": 11861 + }, + { + "epoch": 2.1118233618233617, + "grad_norm": 0.8021374344825745, + "learning_rate": 9.137429192997723e-05, + "loss": 0.9258, + "step": 11862 + }, + { + "epoch": 2.1120014245014245, + "grad_norm": 0.7860891222953796, + "learning_rate": 9.136034671215988e-05, + "loss": 0.7351, + "step": 11863 + }, + { + "epoch": 2.1121794871794872, + "grad_norm": 0.8324207067489624, + "learning_rate": 9.134640166361864e-05, + "loss": 0.8933, + "step": 11864 + }, + { + "epoch": 2.11235754985755, + "grad_norm": 0.8209179639816284, + "learning_rate": 9.133245678462663e-05, + "loss": 0.6983, + "step": 11865 + }, + { + "epoch": 2.1125356125356127, + "grad_norm": 0.7071694731712341, + "learning_rate": 9.131851207545716e-05, + "loss": 0.7796, + "step": 11866 + }, + { + "epoch": 2.112713675213675, + "grad_norm": 0.8126310110092163, + "learning_rate": 9.130456753638339e-05, + "loss": 0.8887, + "step": 11867 + }, + { + "epoch": 2.112891737891738, + "grad_norm": 0.7713829874992371, + "learning_rate": 9.129062316767855e-05, + "loss": 0.8169, + "step": 11868 + }, + { + "epoch": 2.1130698005698005, + "grad_norm": 0.8065944314002991, + "learning_rate": 9.127667896961585e-05, + "loss": 0.9295, + "step": 11869 + }, + { + "epoch": 2.1132478632478633, + "grad_norm": 0.7433435320854187, + "learning_rate": 9.126273494246856e-05, + "loss": 1.089, + "step": 11870 + }, + { + "epoch": 2.113425925925926, + "grad_norm": 0.8168141841888428, + "learning_rate": 9.124879108650978e-05, + "loss": 0.7914, + "step": 11871 + }, + { + "epoch": 2.113603988603989, + "grad_norm": 0.7703335285186768, + "learning_rate": 9.123484740201276e-05, + "loss": 1.0599, + "step": 11872 + }, + { + "epoch": 2.113782051282051, + "grad_norm": 0.810584545135498, + "learning_rate": 9.12209038892507e-05, + "loss": 0.886, + "step": 11873 + }, + { + "epoch": 2.113960113960114, + "grad_norm": 0.8441819548606873, + "learning_rate": 9.120696054849683e-05, + "loss": 0.9069, + "step": 11874 + }, + { + "epoch": 2.1141381766381766, + "grad_norm": 0.816067636013031, + "learning_rate": 9.119301738002425e-05, + "loss": 0.9084, + "step": 11875 + }, + { + "epoch": 2.1143162393162394, + "grad_norm": 0.8595525622367859, + "learning_rate": 9.117907438410622e-05, + "loss": 0.916, + "step": 11876 + }, + { + "epoch": 2.114494301994302, + "grad_norm": 0.8604792356491089, + "learning_rate": 9.116513156101589e-05, + "loss": 1.1207, + "step": 11877 + }, + { + "epoch": 2.114672364672365, + "grad_norm": 0.673664927482605, + "learning_rate": 9.115118891102649e-05, + "loss": 0.9767, + "step": 11878 + }, + { + "epoch": 2.114850427350427, + "grad_norm": 0.7064382433891296, + "learning_rate": 9.113724643441113e-05, + "loss": 0.91, + "step": 11879 + }, + { + "epoch": 2.11502849002849, + "grad_norm": 0.7256918549537659, + "learning_rate": 9.112330413144301e-05, + "loss": 0.9061, + "step": 11880 + }, + { + "epoch": 2.1152065527065527, + "grad_norm": 0.7914155721664429, + "learning_rate": 9.110936200239534e-05, + "loss": 0.6652, + "step": 11881 + }, + { + "epoch": 2.1153846153846154, + "grad_norm": 0.7484595775604248, + "learning_rate": 9.109542004754122e-05, + "loss": 0.8049, + "step": 11882 + }, + { + "epoch": 2.115562678062678, + "grad_norm": 0.8062677979469299, + "learning_rate": 9.108147826715387e-05, + "loss": 0.8671, + "step": 11883 + }, + { + "epoch": 2.115740740740741, + "grad_norm": 0.9595313668251038, + "learning_rate": 9.10675366615064e-05, + "loss": 0.995, + "step": 11884 + }, + { + "epoch": 2.1159188034188032, + "grad_norm": 0.7263179421424866, + "learning_rate": 9.105359523087203e-05, + "loss": 0.9177, + "step": 11885 + }, + { + "epoch": 2.116096866096866, + "grad_norm": 0.900650680065155, + "learning_rate": 9.103965397552385e-05, + "loss": 0.8599, + "step": 11886 + }, + { + "epoch": 2.1162749287749287, + "grad_norm": 0.7682752013206482, + "learning_rate": 9.102571289573506e-05, + "loss": 0.8942, + "step": 11887 + }, + { + "epoch": 2.1164529914529915, + "grad_norm": 0.7076446413993835, + "learning_rate": 9.101177199177874e-05, + "loss": 0.7498, + "step": 11888 + }, + { + "epoch": 2.1166310541310542, + "grad_norm": 0.711475133895874, + "learning_rate": 9.099783126392813e-05, + "loss": 0.7035, + "step": 11889 + }, + { + "epoch": 2.116809116809117, + "grad_norm": 0.6720870137214661, + "learning_rate": 9.098389071245627e-05, + "loss": 0.7315, + "step": 11890 + }, + { + "epoch": 2.1169871794871793, + "grad_norm": 0.8207699656486511, + "learning_rate": 9.096995033763639e-05, + "loss": 0.7465, + "step": 11891 + }, + { + "epoch": 2.117165242165242, + "grad_norm": 0.9032317996025085, + "learning_rate": 9.095601013974153e-05, + "loss": 0.9209, + "step": 11892 + }, + { + "epoch": 2.117343304843305, + "grad_norm": 0.886545717716217, + "learning_rate": 9.094207011904489e-05, + "loss": 0.9411, + "step": 11893 + }, + { + "epoch": 2.1175213675213675, + "grad_norm": 0.8235130906105042, + "learning_rate": 9.092813027581953e-05, + "loss": 0.9264, + "step": 11894 + }, + { + "epoch": 2.1176994301994303, + "grad_norm": 0.7530205845832825, + "learning_rate": 9.091419061033867e-05, + "loss": 0.8926, + "step": 11895 + }, + { + "epoch": 2.117877492877493, + "grad_norm": 0.8329548835754395, + "learning_rate": 9.090025112287533e-05, + "loss": 0.9615, + "step": 11896 + }, + { + "epoch": 2.1180555555555554, + "grad_norm": 0.8184738755226135, + "learning_rate": 9.088631181370269e-05, + "loss": 0.9069, + "step": 11897 + }, + { + "epoch": 2.118233618233618, + "grad_norm": 0.8071370720863342, + "learning_rate": 9.087237268309381e-05, + "loss": 0.8721, + "step": 11898 + }, + { + "epoch": 2.118411680911681, + "grad_norm": 0.8995245695114136, + "learning_rate": 9.085843373132187e-05, + "loss": 0.8815, + "step": 11899 + }, + { + "epoch": 2.1185897435897436, + "grad_norm": 0.7601714730262756, + "learning_rate": 9.084449495865989e-05, + "loss": 0.6824, + "step": 11900 + }, + { + "epoch": 2.1187678062678064, + "grad_norm": 0.8499618172645569, + "learning_rate": 9.083055636538101e-05, + "loss": 0.9868, + "step": 11901 + }, + { + "epoch": 2.118945868945869, + "grad_norm": 0.8190310001373291, + "learning_rate": 9.081661795175837e-05, + "loss": 0.8156, + "step": 11902 + }, + { + "epoch": 2.1191239316239314, + "grad_norm": 0.8340418934822083, + "learning_rate": 9.080267971806498e-05, + "loss": 1.0153, + "step": 11903 + }, + { + "epoch": 2.119301994301994, + "grad_norm": 0.8460756540298462, + "learning_rate": 9.0788741664574e-05, + "loss": 0.8752, + "step": 11904 + }, + { + "epoch": 2.119480056980057, + "grad_norm": 0.7457373738288879, + "learning_rate": 9.077480379155848e-05, + "loss": 0.9105, + "step": 11905 + }, + { + "epoch": 2.1196581196581197, + "grad_norm": 0.7883822917938232, + "learning_rate": 9.076086609929155e-05, + "loss": 0.8782, + "step": 11906 + }, + { + "epoch": 2.1198361823361824, + "grad_norm": 0.912143886089325, + "learning_rate": 9.074692858804622e-05, + "loss": 0.9898, + "step": 11907 + }, + { + "epoch": 2.120014245014245, + "grad_norm": 0.7801905274391174, + "learning_rate": 9.073299125809562e-05, + "loss": 1.091, + "step": 11908 + }, + { + "epoch": 2.1201923076923075, + "grad_norm": 0.6836256384849548, + "learning_rate": 9.071905410971279e-05, + "loss": 0.7967, + "step": 11909 + }, + { + "epoch": 2.1203703703703702, + "grad_norm": 0.7656795382499695, + "learning_rate": 9.070511714317085e-05, + "loss": 0.9696, + "step": 11910 + }, + { + "epoch": 2.120548433048433, + "grad_norm": 0.7010015249252319, + "learning_rate": 9.06911803587428e-05, + "loss": 0.6501, + "step": 11911 + }, + { + "epoch": 2.1207264957264957, + "grad_norm": 0.6673064827919006, + "learning_rate": 9.067724375670174e-05, + "loss": 0.5663, + "step": 11912 + }, + { + "epoch": 2.1209045584045585, + "grad_norm": 0.8683220148086548, + "learning_rate": 9.06633073373207e-05, + "loss": 0.9722, + "step": 11913 + }, + { + "epoch": 2.1210826210826212, + "grad_norm": 0.7793976068496704, + "learning_rate": 9.06493711008728e-05, + "loss": 0.7595, + "step": 11914 + }, + { + "epoch": 2.1212606837606836, + "grad_norm": 0.7803528308868408, + "learning_rate": 9.0635435047631e-05, + "loss": 0.9262, + "step": 11915 + }, + { + "epoch": 2.1214387464387463, + "grad_norm": 0.8067244291305542, + "learning_rate": 9.062149917786846e-05, + "loss": 0.9376, + "step": 11916 + }, + { + "epoch": 2.121616809116809, + "grad_norm": 0.7389153838157654, + "learning_rate": 9.060756349185812e-05, + "loss": 0.7414, + "step": 11917 + }, + { + "epoch": 2.121794871794872, + "grad_norm": 0.7717151045799255, + "learning_rate": 9.059362798987308e-05, + "loss": 0.7261, + "step": 11918 + }, + { + "epoch": 2.1219729344729346, + "grad_norm": 0.7668650150299072, + "learning_rate": 9.057969267218632e-05, + "loss": 0.8145, + "step": 11919 + }, + { + "epoch": 2.1221509971509973, + "grad_norm": 1.0015910863876343, + "learning_rate": 9.056575753907093e-05, + "loss": 0.8997, + "step": 11920 + }, + { + "epoch": 2.1223290598290596, + "grad_norm": 0.8731024861335754, + "learning_rate": 9.055182259079997e-05, + "loss": 1.0101, + "step": 11921 + }, + { + "epoch": 2.1225071225071224, + "grad_norm": 0.7662718892097473, + "learning_rate": 9.053788782764637e-05, + "loss": 0.697, + "step": 11922 + }, + { + "epoch": 2.122685185185185, + "grad_norm": 0.7783135771751404, + "learning_rate": 9.05239532498832e-05, + "loss": 0.8506, + "step": 11923 + }, + { + "epoch": 2.122863247863248, + "grad_norm": 0.8667652606964111, + "learning_rate": 9.05100188577835e-05, + "loss": 0.9851, + "step": 11924 + }, + { + "epoch": 2.1230413105413106, + "grad_norm": 0.7785412073135376, + "learning_rate": 9.049608465162028e-05, + "loss": 0.5924, + "step": 11925 + }, + { + "epoch": 2.1232193732193734, + "grad_norm": 0.7968559861183167, + "learning_rate": 9.04821506316665e-05, + "loss": 0.8114, + "step": 11926 + }, + { + "epoch": 2.123397435897436, + "grad_norm": 0.8065921068191528, + "learning_rate": 9.046821679819527e-05, + "loss": 0.9045, + "step": 11927 + }, + { + "epoch": 2.1235754985754984, + "grad_norm": 0.7509779930114746, + "learning_rate": 9.045428315147948e-05, + "loss": 0.7337, + "step": 11928 + }, + { + "epoch": 2.123753561253561, + "grad_norm": 0.8174976110458374, + "learning_rate": 9.044034969179219e-05, + "loss": 1.0113, + "step": 11929 + }, + { + "epoch": 2.123931623931624, + "grad_norm": 0.8723294734954834, + "learning_rate": 9.042641641940638e-05, + "loss": 0.9657, + "step": 11930 + }, + { + "epoch": 2.1241096866096867, + "grad_norm": 0.7412081360816956, + "learning_rate": 9.041248333459509e-05, + "loss": 0.9311, + "step": 11931 + }, + { + "epoch": 2.1242877492877494, + "grad_norm": 0.7376424670219421, + "learning_rate": 9.039855043763124e-05, + "loss": 0.7039, + "step": 11932 + }, + { + "epoch": 2.1244658119658117, + "grad_norm": 0.8002118468284607, + "learning_rate": 9.038461772878786e-05, + "loss": 0.9555, + "step": 11933 + }, + { + "epoch": 2.1246438746438745, + "grad_norm": 0.7221434712409973, + "learning_rate": 9.03706852083379e-05, + "loss": 0.8462, + "step": 11934 + }, + { + "epoch": 2.1248219373219372, + "grad_norm": 0.8506385684013367, + "learning_rate": 9.035675287655441e-05, + "loss": 0.7977, + "step": 11935 + }, + { + "epoch": 2.125, + "grad_norm": 0.8088411688804626, + "learning_rate": 9.034282073371025e-05, + "loss": 1.0146, + "step": 11936 + }, + { + "epoch": 2.1251780626780628, + "grad_norm": 0.9231638312339783, + "learning_rate": 9.032888878007853e-05, + "loss": 0.7017, + "step": 11937 + }, + { + "epoch": 2.1253561253561255, + "grad_norm": 0.721066951751709, + "learning_rate": 9.03149570159321e-05, + "loss": 0.7662, + "step": 11938 + }, + { + "epoch": 2.1255341880341883, + "grad_norm": 0.7804762721061707, + "learning_rate": 9.030102544154395e-05, + "loss": 0.6835, + "step": 11939 + }, + { + "epoch": 2.1257122507122506, + "grad_norm": 0.9728445410728455, + "learning_rate": 9.028709405718707e-05, + "loss": 0.9161, + "step": 11940 + }, + { + "epoch": 2.1258903133903133, + "grad_norm": 0.8209855556488037, + "learning_rate": 9.02731628631344e-05, + "loss": 0.7492, + "step": 11941 + }, + { + "epoch": 2.126068376068376, + "grad_norm": 0.7054622769355774, + "learning_rate": 9.025923185965896e-05, + "loss": 0.7908, + "step": 11942 + }, + { + "epoch": 2.126246438746439, + "grad_norm": 0.73018878698349, + "learning_rate": 9.024530104703358e-05, + "loss": 0.7902, + "step": 11943 + }, + { + "epoch": 2.1264245014245016, + "grad_norm": 0.73788982629776, + "learning_rate": 9.023137042553127e-05, + "loss": 0.8473, + "step": 11944 + }, + { + "epoch": 2.126602564102564, + "grad_norm": 0.7733396291732788, + "learning_rate": 9.021743999542495e-05, + "loss": 0.9595, + "step": 11945 + }, + { + "epoch": 2.1267806267806266, + "grad_norm": 0.9066760540008545, + "learning_rate": 9.020350975698761e-05, + "loss": 0.8517, + "step": 11946 + }, + { + "epoch": 2.1269586894586894, + "grad_norm": 0.7552717328071594, + "learning_rate": 9.018957971049211e-05, + "loss": 0.6802, + "step": 11947 + }, + { + "epoch": 2.127136752136752, + "grad_norm": 0.7437541484832764, + "learning_rate": 9.017564985621144e-05, + "loss": 0.9365, + "step": 11948 + }, + { + "epoch": 2.127314814814815, + "grad_norm": 0.8216256499290466, + "learning_rate": 9.016172019441847e-05, + "loss": 0.9019, + "step": 11949 + }, + { + "epoch": 2.1274928774928776, + "grad_norm": 0.752247154712677, + "learning_rate": 9.014779072538621e-05, + "loss": 0.7771, + "step": 11950 + }, + { + "epoch": 2.1276709401709404, + "grad_norm": 0.7714348435401917, + "learning_rate": 9.013386144938748e-05, + "loss": 0.8495, + "step": 11951 + }, + { + "epoch": 2.1278490028490027, + "grad_norm": 0.8347537517547607, + "learning_rate": 9.011993236669529e-05, + "loss": 0.861, + "step": 11952 + }, + { + "epoch": 2.1280270655270654, + "grad_norm": 0.8180193901062012, + "learning_rate": 9.010600347758245e-05, + "loss": 0.9059, + "step": 11953 + }, + { + "epoch": 2.128205128205128, + "grad_norm": 0.7328528761863708, + "learning_rate": 9.009207478232193e-05, + "loss": 0.9144, + "step": 11954 + }, + { + "epoch": 2.128383190883191, + "grad_norm": 0.7590839862823486, + "learning_rate": 9.007814628118661e-05, + "loss": 0.8642, + "step": 11955 + }, + { + "epoch": 2.1285612535612537, + "grad_norm": 0.7962782382965088, + "learning_rate": 9.006421797444945e-05, + "loss": 0.8958, + "step": 11956 + }, + { + "epoch": 2.128739316239316, + "grad_norm": 0.7302426695823669, + "learning_rate": 9.005028986238325e-05, + "loss": 0.9419, + "step": 11957 + }, + { + "epoch": 2.1289173789173788, + "grad_norm": 0.9223780632019043, + "learning_rate": 9.003636194526098e-05, + "loss": 0.7631, + "step": 11958 + }, + { + "epoch": 2.1290954415954415, + "grad_norm": 0.728225588798523, + "learning_rate": 9.002243422335547e-05, + "loss": 0.7705, + "step": 11959 + }, + { + "epoch": 2.1292735042735043, + "grad_norm": 0.8519338369369507, + "learning_rate": 9.000850669693964e-05, + "loss": 0.8962, + "step": 11960 + }, + { + "epoch": 2.129451566951567, + "grad_norm": 0.8920532464981079, + "learning_rate": 8.999457936628641e-05, + "loss": 0.618, + "step": 11961 + }, + { + "epoch": 2.1296296296296298, + "grad_norm": 0.9719427824020386, + "learning_rate": 8.998065223166857e-05, + "loss": 0.9142, + "step": 11962 + }, + { + "epoch": 2.1298076923076925, + "grad_norm": 0.8130887150764465, + "learning_rate": 8.996672529335908e-05, + "loss": 1.0246, + "step": 11963 + }, + { + "epoch": 2.129985754985755, + "grad_norm": 0.7682677507400513, + "learning_rate": 8.995279855163073e-05, + "loss": 0.7964, + "step": 11964 + }, + { + "epoch": 2.1301638176638176, + "grad_norm": 0.8507778644561768, + "learning_rate": 8.993887200675641e-05, + "loss": 0.756, + "step": 11965 + }, + { + "epoch": 2.1303418803418803, + "grad_norm": 0.815487802028656, + "learning_rate": 8.992494565900901e-05, + "loss": 0.7596, + "step": 11966 + }, + { + "epoch": 2.130519943019943, + "grad_norm": 0.8560892939567566, + "learning_rate": 8.991101950866138e-05, + "loss": 0.9939, + "step": 11967 + }, + { + "epoch": 2.130698005698006, + "grad_norm": 0.8737899661064148, + "learning_rate": 8.989709355598635e-05, + "loss": 0.9235, + "step": 11968 + }, + { + "epoch": 2.1308760683760686, + "grad_norm": 0.8434267640113831, + "learning_rate": 8.98831678012568e-05, + "loss": 0.7832, + "step": 11969 + }, + { + "epoch": 2.131054131054131, + "grad_norm": 0.8286582827568054, + "learning_rate": 8.986924224474553e-05, + "loss": 1.0591, + "step": 11970 + }, + { + "epoch": 2.1312321937321936, + "grad_norm": 0.8023663759231567, + "learning_rate": 8.985531688672546e-05, + "loss": 0.935, + "step": 11971 + }, + { + "epoch": 2.1314102564102564, + "grad_norm": 0.6504420042037964, + "learning_rate": 8.984139172746933e-05, + "loss": 0.79, + "step": 11972 + }, + { + "epoch": 2.131588319088319, + "grad_norm": 0.8969349265098572, + "learning_rate": 8.982746676725009e-05, + "loss": 1.0531, + "step": 11973 + }, + { + "epoch": 2.131766381766382, + "grad_norm": 0.802094042301178, + "learning_rate": 8.981354200634046e-05, + "loss": 0.8873, + "step": 11974 + }, + { + "epoch": 2.1319444444444446, + "grad_norm": 0.7630797624588013, + "learning_rate": 8.979961744501332e-05, + "loss": 0.9299, + "step": 11975 + }, + { + "epoch": 2.132122507122507, + "grad_norm": 0.8395546674728394, + "learning_rate": 8.978569308354148e-05, + "loss": 0.922, + "step": 11976 + }, + { + "epoch": 2.1323005698005697, + "grad_norm": 0.9325534701347351, + "learning_rate": 8.97717689221978e-05, + "loss": 0.9156, + "step": 11977 + }, + { + "epoch": 2.1324786324786325, + "grad_norm": 0.8139503002166748, + "learning_rate": 8.975784496125502e-05, + "loss": 0.8882, + "step": 11978 + }, + { + "epoch": 2.132656695156695, + "grad_norm": 1.0311007499694824, + "learning_rate": 8.974392120098599e-05, + "loss": 1.0068, + "step": 11979 + }, + { + "epoch": 2.132834757834758, + "grad_norm": 0.9328663945198059, + "learning_rate": 8.972999764166354e-05, + "loss": 0.8313, + "step": 11980 + }, + { + "epoch": 2.1330128205128207, + "grad_norm": 0.747276782989502, + "learning_rate": 8.971607428356044e-05, + "loss": 0.9302, + "step": 11981 + }, + { + "epoch": 2.133190883190883, + "grad_norm": 0.7572789788246155, + "learning_rate": 8.970215112694953e-05, + "loss": 0.8016, + "step": 11982 + }, + { + "epoch": 2.1333689458689458, + "grad_norm": 0.8988085389137268, + "learning_rate": 8.968822817210354e-05, + "loss": 0.9307, + "step": 11983 + }, + { + "epoch": 2.1335470085470085, + "grad_norm": 0.7537818551063538, + "learning_rate": 8.967430541929532e-05, + "loss": 0.6423, + "step": 11984 + }, + { + "epoch": 2.1337250712250713, + "grad_norm": 0.7470884323120117, + "learning_rate": 8.966038286879763e-05, + "loss": 0.7753, + "step": 11985 + }, + { + "epoch": 2.133903133903134, + "grad_norm": 0.8670676946640015, + "learning_rate": 8.964646052088328e-05, + "loss": 1.0407, + "step": 11986 + }, + { + "epoch": 2.1340811965811968, + "grad_norm": 0.8322215676307678, + "learning_rate": 8.9632538375825e-05, + "loss": 0.6498, + "step": 11987 + }, + { + "epoch": 2.134259259259259, + "grad_norm": 0.7089048027992249, + "learning_rate": 8.961861643389562e-05, + "loss": 0.8778, + "step": 11988 + }, + { + "epoch": 2.134437321937322, + "grad_norm": 0.7980125546455383, + "learning_rate": 8.960469469536786e-05, + "loss": 0.7797, + "step": 11989 + }, + { + "epoch": 2.1346153846153846, + "grad_norm": 0.9979715943336487, + "learning_rate": 8.959077316051452e-05, + "loss": 0.7388, + "step": 11990 + }, + { + "epoch": 2.1347934472934473, + "grad_norm": 1.0040662288665771, + "learning_rate": 8.957685182960833e-05, + "loss": 0.954, + "step": 11991 + }, + { + "epoch": 2.13497150997151, + "grad_norm": 0.7885099053382874, + "learning_rate": 8.956293070292214e-05, + "loss": 0.9232, + "step": 11992 + }, + { + "epoch": 2.135149572649573, + "grad_norm": 0.7242771983146667, + "learning_rate": 8.954900978072859e-05, + "loss": 0.8614, + "step": 11993 + }, + { + "epoch": 2.135327635327635, + "grad_norm": 0.7970352172851562, + "learning_rate": 8.95350890633005e-05, + "loss": 0.8959, + "step": 11994 + }, + { + "epoch": 2.135505698005698, + "grad_norm": 0.8587128520011902, + "learning_rate": 8.952116855091059e-05, + "loss": 0.9981, + "step": 11995 + }, + { + "epoch": 2.1356837606837606, + "grad_norm": 0.8206220269203186, + "learning_rate": 8.950724824383164e-05, + "loss": 1.0271, + "step": 11996 + }, + { + "epoch": 2.1358618233618234, + "grad_norm": 0.8085001707077026, + "learning_rate": 8.949332814233635e-05, + "loss": 0.925, + "step": 11997 + }, + { + "epoch": 2.136039886039886, + "grad_norm": 0.8361417651176453, + "learning_rate": 8.947940824669748e-05, + "loss": 0.8744, + "step": 11998 + }, + { + "epoch": 2.136217948717949, + "grad_norm": 0.7548407316207886, + "learning_rate": 8.946548855718773e-05, + "loss": 0.7365, + "step": 11999 + }, + { + "epoch": 2.136396011396011, + "grad_norm": 0.8671223521232605, + "learning_rate": 8.945156907407983e-05, + "loss": 0.8958, + "step": 12000 + }, + { + "epoch": 2.136574074074074, + "grad_norm": 0.8007429838180542, + "learning_rate": 8.943764979764656e-05, + "loss": 0.955, + "step": 12001 + }, + { + "epoch": 2.1367521367521367, + "grad_norm": 0.7834315299987793, + "learning_rate": 8.942373072816057e-05, + "loss": 0.9226, + "step": 12002 + }, + { + "epoch": 2.1369301994301995, + "grad_norm": 0.896920919418335, + "learning_rate": 8.940981186589466e-05, + "loss": 0.8779, + "step": 12003 + }, + { + "epoch": 2.137108262108262, + "grad_norm": 0.7473411560058594, + "learning_rate": 8.939589321112143e-05, + "loss": 0.8993, + "step": 12004 + }, + { + "epoch": 2.137286324786325, + "grad_norm": 0.8071674704551697, + "learning_rate": 8.938197476411367e-05, + "loss": 0.9998, + "step": 12005 + }, + { + "epoch": 2.1374643874643873, + "grad_norm": 0.839290976524353, + "learning_rate": 8.936805652514404e-05, + "loss": 0.8311, + "step": 12006 + }, + { + "epoch": 2.13764245014245, + "grad_norm": 0.7217035293579102, + "learning_rate": 8.93541384944853e-05, + "loss": 0.8009, + "step": 12007 + }, + { + "epoch": 2.1378205128205128, + "grad_norm": 0.7392259836196899, + "learning_rate": 8.934022067241004e-05, + "loss": 0.9854, + "step": 12008 + }, + { + "epoch": 2.1379985754985755, + "grad_norm": 0.7470507621765137, + "learning_rate": 8.932630305919107e-05, + "loss": 0.8111, + "step": 12009 + }, + { + "epoch": 2.1381766381766383, + "grad_norm": 0.7988318204879761, + "learning_rate": 8.931238565510098e-05, + "loss": 0.8492, + "step": 12010 + }, + { + "epoch": 2.138354700854701, + "grad_norm": 0.9267526268959045, + "learning_rate": 8.929846846041251e-05, + "loss": 1.2238, + "step": 12011 + }, + { + "epoch": 2.1385327635327633, + "grad_norm": 0.8036465644836426, + "learning_rate": 8.92845514753983e-05, + "loss": 0.8837, + "step": 12012 + }, + { + "epoch": 2.138710826210826, + "grad_norm": 0.809256911277771, + "learning_rate": 8.927063470033109e-05, + "loss": 0.8836, + "step": 12013 + }, + { + "epoch": 2.138888888888889, + "grad_norm": 0.754692792892456, + "learning_rate": 8.925671813548345e-05, + "loss": 0.9469, + "step": 12014 + }, + { + "epoch": 2.1390669515669516, + "grad_norm": 0.9183036088943481, + "learning_rate": 8.924280178112814e-05, + "loss": 0.7654, + "step": 12015 + }, + { + "epoch": 2.1392450142450143, + "grad_norm": 0.82411128282547, + "learning_rate": 8.922888563753775e-05, + "loss": 0.9132, + "step": 12016 + }, + { + "epoch": 2.139423076923077, + "grad_norm": 0.8455918431282043, + "learning_rate": 8.9214969704985e-05, + "loss": 1.0041, + "step": 12017 + }, + { + "epoch": 2.1396011396011394, + "grad_norm": 0.9235896468162537, + "learning_rate": 8.92010539837425e-05, + "loss": 0.7842, + "step": 12018 + }, + { + "epoch": 2.139779202279202, + "grad_norm": 0.8965059518814087, + "learning_rate": 8.918713847408289e-05, + "loss": 1.0665, + "step": 12019 + }, + { + "epoch": 2.139957264957265, + "grad_norm": Infinity, + "learning_rate": 8.918713847408289e-05, + "loss": 1.0773, + "step": 12020 + }, + { + "epoch": 2.1401353276353277, + "grad_norm": 0.8859738707542419, + "learning_rate": 8.917322317627887e-05, + "loss": 0.9175, + "step": 12021 + }, + { + "epoch": 2.1403133903133904, + "grad_norm": 0.7828214764595032, + "learning_rate": 8.915930809060304e-05, + "loss": 0.8433, + "step": 12022 + }, + { + "epoch": 2.140491452991453, + "grad_norm": 0.7705734372138977, + "learning_rate": 8.914539321732808e-05, + "loss": 0.8696, + "step": 12023 + }, + { + "epoch": 2.1406695156695155, + "grad_norm": 0.7999989986419678, + "learning_rate": 8.913147855672655e-05, + "loss": 1.0531, + "step": 12024 + }, + { + "epoch": 2.140847578347578, + "grad_norm": 0.7210655212402344, + "learning_rate": 8.911756410907118e-05, + "loss": 0.6703, + "step": 12025 + }, + { + "epoch": 2.141025641025641, + "grad_norm": 0.7153459191322327, + "learning_rate": 8.910364987463447e-05, + "loss": 0.7166, + "step": 12026 + }, + { + "epoch": 2.1412037037037037, + "grad_norm": 0.771530032157898, + "learning_rate": 8.908973585368913e-05, + "loss": 0.6881, + "step": 12027 + }, + { + "epoch": 2.1413817663817665, + "grad_norm": 0.9988116025924683, + "learning_rate": 8.907582204650774e-05, + "loss": 0.8329, + "step": 12028 + }, + { + "epoch": 2.1415598290598292, + "grad_norm": 0.6992440819740295, + "learning_rate": 8.906190845336296e-05, + "loss": 0.6262, + "step": 12029 + }, + { + "epoch": 2.1417378917378915, + "grad_norm": 0.8061181902885437, + "learning_rate": 8.904799507452731e-05, + "loss": 0.8325, + "step": 12030 + }, + { + "epoch": 2.1419159544159543, + "grad_norm": 0.8372871279716492, + "learning_rate": 8.903408191027349e-05, + "loss": 0.8894, + "step": 12031 + }, + { + "epoch": 2.142094017094017, + "grad_norm": 0.803719162940979, + "learning_rate": 8.902016896087402e-05, + "loss": 0.9031, + "step": 12032 + }, + { + "epoch": 2.14227207977208, + "grad_norm": 0.8168890476226807, + "learning_rate": 8.900625622660158e-05, + "loss": 0.8174, + "step": 12033 + }, + { + "epoch": 2.1424501424501425, + "grad_norm": 0.8011388182640076, + "learning_rate": 8.899234370772865e-05, + "loss": 0.8267, + "step": 12034 + }, + { + "epoch": 2.1426282051282053, + "grad_norm": 0.8209220767021179, + "learning_rate": 8.897843140452795e-05, + "loss": 0.9303, + "step": 12035 + }, + { + "epoch": 2.142806267806268, + "grad_norm": 0.773525059223175, + "learning_rate": 8.896451931727192e-05, + "loss": 0.7037, + "step": 12036 + }, + { + "epoch": 2.1429843304843303, + "grad_norm": 0.7568892240524292, + "learning_rate": 8.895060744623324e-05, + "loss": 0.8568, + "step": 12037 + }, + { + "epoch": 2.143162393162393, + "grad_norm": 0.713636040687561, + "learning_rate": 8.893669579168444e-05, + "loss": 0.7838, + "step": 12038 + }, + { + "epoch": 2.143340455840456, + "grad_norm": 0.7462167739868164, + "learning_rate": 8.892278435389814e-05, + "loss": 0.6311, + "step": 12039 + }, + { + "epoch": 2.1435185185185186, + "grad_norm": 0.7164530158042908, + "learning_rate": 8.890887313314685e-05, + "loss": 1.0228, + "step": 12040 + }, + { + "epoch": 2.1436965811965814, + "grad_norm": 0.7540927529335022, + "learning_rate": 8.889496212970312e-05, + "loss": 0.8958, + "step": 12041 + }, + { + "epoch": 2.1438746438746437, + "grad_norm": 0.8119065761566162, + "learning_rate": 8.888105134383957e-05, + "loss": 0.8925, + "step": 12042 + }, + { + "epoch": 2.1440527065527064, + "grad_norm": 0.7905679941177368, + "learning_rate": 8.88671407758287e-05, + "loss": 0.7579, + "step": 12043 + }, + { + "epoch": 2.144230769230769, + "grad_norm": 0.8901177048683167, + "learning_rate": 8.885323042594312e-05, + "loss": 0.8849, + "step": 12044 + }, + { + "epoch": 2.144408831908832, + "grad_norm": 0.6958974599838257, + "learning_rate": 8.88393202944553e-05, + "loss": 0.8072, + "step": 12045 + }, + { + "epoch": 2.1445868945868947, + "grad_norm": 0.790036141872406, + "learning_rate": 8.882541038163786e-05, + "loss": 0.796, + "step": 12046 + }, + { + "epoch": 2.1447649572649574, + "grad_norm": 0.757655680179596, + "learning_rate": 8.881150068776324e-05, + "loss": 0.8094, + "step": 12047 + }, + { + "epoch": 2.14494301994302, + "grad_norm": 0.7525215148925781, + "learning_rate": 8.879759121310404e-05, + "loss": 0.6746, + "step": 12048 + }, + { + "epoch": 2.1451210826210825, + "grad_norm": 0.740566074848175, + "learning_rate": 8.878368195793276e-05, + "loss": 0.688, + "step": 12049 + }, + { + "epoch": 2.1452991452991452, + "grad_norm": 0.7771985530853271, + "learning_rate": 8.876977292252196e-05, + "loss": 1.0013, + "step": 12050 + }, + { + "epoch": 2.145477207977208, + "grad_norm": 0.8582369685173035, + "learning_rate": 8.875586410714409e-05, + "loss": 0.9185, + "step": 12051 + }, + { + "epoch": 2.1456552706552707, + "grad_norm": 0.7992526292800903, + "learning_rate": 8.874195551207174e-05, + "loss": 0.8388, + "step": 12052 + }, + { + "epoch": 2.1458333333333335, + "grad_norm": 0.795129120349884, + "learning_rate": 8.872804713757735e-05, + "loss": 0.88, + "step": 12053 + }, + { + "epoch": 2.146011396011396, + "grad_norm": 0.7467540502548218, + "learning_rate": 8.871413898393351e-05, + "loss": 0.8092, + "step": 12054 + }, + { + "epoch": 2.1461894586894585, + "grad_norm": 0.9468266367912292, + "learning_rate": 8.870023105141264e-05, + "loss": 0.7759, + "step": 12055 + }, + { + "epoch": 2.1463675213675213, + "grad_norm": 0.7893772721290588, + "learning_rate": 8.868632334028727e-05, + "loss": 0.8508, + "step": 12056 + }, + { + "epoch": 2.146545584045584, + "grad_norm": 0.6931375861167908, + "learning_rate": 8.867241585082988e-05, + "loss": 0.5013, + "step": 12057 + }, + { + "epoch": 2.146723646723647, + "grad_norm": 0.8978447318077087, + "learning_rate": 8.865850858331301e-05, + "loss": 0.9518, + "step": 12058 + }, + { + "epoch": 2.1469017094017095, + "grad_norm": 0.7293453812599182, + "learning_rate": 8.864460153800906e-05, + "loss": 0.731, + "step": 12059 + }, + { + "epoch": 2.1470797720797723, + "grad_norm": 0.8537824749946594, + "learning_rate": 8.863069471519056e-05, + "loss": 0.7935, + "step": 12060 + }, + { + "epoch": 2.1472578347578346, + "grad_norm": 0.6527614593505859, + "learning_rate": 8.861678811513002e-05, + "loss": 0.6579, + "step": 12061 + }, + { + "epoch": 2.1474358974358974, + "grad_norm": 0.9407904148101807, + "learning_rate": 8.860288173809983e-05, + "loss": 0.9057, + "step": 12062 + }, + { + "epoch": 2.14761396011396, + "grad_norm": 0.9314194321632385, + "learning_rate": 8.858897558437251e-05, + "loss": 1.0826, + "step": 12063 + }, + { + "epoch": 2.147792022792023, + "grad_norm": 0.7872337102890015, + "learning_rate": 8.85750696542205e-05, + "loss": 0.8919, + "step": 12064 + }, + { + "epoch": 2.1479700854700856, + "grad_norm": 0.8379341959953308, + "learning_rate": 8.85611639479163e-05, + "loss": 0.8193, + "step": 12065 + }, + { + "epoch": 2.148148148148148, + "grad_norm": 0.801295280456543, + "learning_rate": 8.85472584657323e-05, + "loss": 0.9305, + "step": 12066 + }, + { + "epoch": 2.1483262108262107, + "grad_norm": 0.7625086903572083, + "learning_rate": 8.853335320794098e-05, + "loss": 0.8379, + "step": 12067 + }, + { + "epoch": 2.1485042735042734, + "grad_norm": 0.8256231546401978, + "learning_rate": 8.851944817481478e-05, + "loss": 0.8901, + "step": 12068 + }, + { + "epoch": 2.148682336182336, + "grad_norm": 0.6940581202507019, + "learning_rate": 8.850554336662618e-05, + "loss": 0.6706, + "step": 12069 + }, + { + "epoch": 2.148860398860399, + "grad_norm": 0.910836398601532, + "learning_rate": 8.849163878364755e-05, + "loss": 0.9326, + "step": 12070 + }, + { + "epoch": 2.1490384615384617, + "grad_norm": 0.8550460934638977, + "learning_rate": 8.847773442615138e-05, + "loss": 0.8474, + "step": 12071 + }, + { + "epoch": 2.1492165242165244, + "grad_norm": 0.8178627490997314, + "learning_rate": 8.846383029441002e-05, + "loss": 0.8331, + "step": 12072 + }, + { + "epoch": 2.1493945868945867, + "grad_norm": 0.7606281638145447, + "learning_rate": 8.844992638869599e-05, + "loss": 0.6571, + "step": 12073 + }, + { + "epoch": 2.1495726495726495, + "grad_norm": 0.7166888117790222, + "learning_rate": 8.84360227092816e-05, + "loss": 0.8592, + "step": 12074 + }, + { + "epoch": 2.1497507122507122, + "grad_norm": 0.7688186764717102, + "learning_rate": 8.84221192564394e-05, + "loss": 0.691, + "step": 12075 + }, + { + "epoch": 2.149928774928775, + "grad_norm": 0.876740038394928, + "learning_rate": 8.840821603044166e-05, + "loss": 0.9962, + "step": 12076 + }, + { + "epoch": 2.1501068376068377, + "grad_norm": 0.7910363078117371, + "learning_rate": 8.839431303156087e-05, + "loss": 0.8061, + "step": 12077 + }, + { + "epoch": 2.1502849002849005, + "grad_norm": 0.6880493760108948, + "learning_rate": 8.83804102600694e-05, + "loss": 0.8078, + "step": 12078 + }, + { + "epoch": 2.150462962962963, + "grad_norm": 0.7795937061309814, + "learning_rate": 8.836650771623963e-05, + "loss": 0.8504, + "step": 12079 + }, + { + "epoch": 2.1506410256410255, + "grad_norm": 0.7761844992637634, + "learning_rate": 8.835260540034403e-05, + "loss": 0.7253, + "step": 12080 + }, + { + "epoch": 2.1508190883190883, + "grad_norm": 0.7070515751838684, + "learning_rate": 8.83387033126549e-05, + "loss": 0.7156, + "step": 12081 + }, + { + "epoch": 2.150997150997151, + "grad_norm": 0.7666274309158325, + "learning_rate": 8.832480145344467e-05, + "loss": 0.813, + "step": 12082 + }, + { + "epoch": 2.151175213675214, + "grad_norm": 0.9145975708961487, + "learning_rate": 8.831089982298568e-05, + "loss": 0.8889, + "step": 12083 + }, + { + "epoch": 2.1513532763532766, + "grad_norm": 0.7735843062400818, + "learning_rate": 8.829699842155035e-05, + "loss": 0.7152, + "step": 12084 + }, + { + "epoch": 2.151531339031339, + "grad_norm": 0.7625414729118347, + "learning_rate": 8.828309724941099e-05, + "loss": 0.8752, + "step": 12085 + }, + { + "epoch": 2.1517094017094016, + "grad_norm": 0.8874264359474182, + "learning_rate": 8.826919630684005e-05, + "loss": 0.8175, + "step": 12086 + }, + { + "epoch": 2.1518874643874644, + "grad_norm": 0.7425693273544312, + "learning_rate": 8.82552955941098e-05, + "loss": 0.7505, + "step": 12087 + }, + { + "epoch": 2.152065527065527, + "grad_norm": 0.7098270058631897, + "learning_rate": 8.824139511149265e-05, + "loss": 0.7129, + "step": 12088 + }, + { + "epoch": 2.15224358974359, + "grad_norm": 0.8470510840415955, + "learning_rate": 8.822749485926092e-05, + "loss": 0.9656, + "step": 12089 + }, + { + "epoch": 2.1524216524216526, + "grad_norm": 0.7690402865409851, + "learning_rate": 8.8213594837687e-05, + "loss": 0.9436, + "step": 12090 + }, + { + "epoch": 2.152599715099715, + "grad_norm": 0.77431321144104, + "learning_rate": 8.819969504704318e-05, + "loss": 0.9912, + "step": 12091 + }, + { + "epoch": 2.1527777777777777, + "grad_norm": 0.7590892314910889, + "learning_rate": 8.818579548760184e-05, + "loss": 0.7412, + "step": 12092 + }, + { + "epoch": 2.1529558404558404, + "grad_norm": 0.870966374874115, + "learning_rate": 8.817189615963528e-05, + "loss": 1.0248, + "step": 12093 + }, + { + "epoch": 2.153133903133903, + "grad_norm": 0.7989356517791748, + "learning_rate": 8.815799706341587e-05, + "loss": 0.7104, + "step": 12094 + }, + { + "epoch": 2.153311965811966, + "grad_norm": 0.9615582227706909, + "learning_rate": 8.814409819921589e-05, + "loss": 1.0191, + "step": 12095 + }, + { + "epoch": 2.1534900284900287, + "grad_norm": 0.7063159346580505, + "learning_rate": 8.81301995673077e-05, + "loss": 0.8209, + "step": 12096 + }, + { + "epoch": 2.153668091168091, + "grad_norm": 0.8179874420166016, + "learning_rate": 8.811630116796356e-05, + "loss": 0.9457, + "step": 12097 + }, + { + "epoch": 2.1538461538461537, + "grad_norm": 0.7227353453636169, + "learning_rate": 8.810240300145582e-05, + "loss": 0.8112, + "step": 12098 + }, + { + "epoch": 2.1540242165242165, + "grad_norm": 0.7480359077453613, + "learning_rate": 8.808850506805677e-05, + "loss": 0.6293, + "step": 12099 + }, + { + "epoch": 2.1542022792022792, + "grad_norm": 0.7610893845558167, + "learning_rate": 8.807460736803871e-05, + "loss": 0.911, + "step": 12100 + }, + { + "epoch": 2.154380341880342, + "grad_norm": 0.774640679359436, + "learning_rate": 8.806070990167399e-05, + "loss": 0.8144, + "step": 12101 + }, + { + "epoch": 2.1545584045584047, + "grad_norm": 0.7785552144050598, + "learning_rate": 8.804681266923482e-05, + "loss": 0.8841, + "step": 12102 + }, + { + "epoch": 2.154736467236467, + "grad_norm": 0.843715488910675, + "learning_rate": 8.803291567099354e-05, + "loss": 0.9056, + "step": 12103 + }, + { + "epoch": 2.15491452991453, + "grad_norm": 0.7996346354484558, + "learning_rate": 8.801901890722241e-05, + "loss": 0.8916, + "step": 12104 + }, + { + "epoch": 2.1550925925925926, + "grad_norm": 0.9159125685691833, + "learning_rate": 8.800512237819376e-05, + "loss": 0.8225, + "step": 12105 + }, + { + "epoch": 2.1552706552706553, + "grad_norm": 0.8341643810272217, + "learning_rate": 8.799122608417976e-05, + "loss": 0.8702, + "step": 12106 + }, + { + "epoch": 2.155448717948718, + "grad_norm": 0.8075932264328003, + "learning_rate": 8.797733002545278e-05, + "loss": 0.9167, + "step": 12107 + }, + { + "epoch": 2.155626780626781, + "grad_norm": 0.8370183706283569, + "learning_rate": 8.7963434202285e-05, + "loss": 0.9213, + "step": 12108 + }, + { + "epoch": 2.155804843304843, + "grad_norm": 0.7500374913215637, + "learning_rate": 8.794953861494877e-05, + "loss": 0.7702, + "step": 12109 + }, + { + "epoch": 2.155982905982906, + "grad_norm": 0.7347766160964966, + "learning_rate": 8.793564326371626e-05, + "loss": 0.7057, + "step": 12110 + }, + { + "epoch": 2.1561609686609686, + "grad_norm": 0.754917562007904, + "learning_rate": 8.79217481488598e-05, + "loss": 0.8725, + "step": 12111 + }, + { + "epoch": 2.1563390313390314, + "grad_norm": 0.6942774057388306, + "learning_rate": 8.790785327065155e-05, + "loss": 0.85, + "step": 12112 + }, + { + "epoch": 2.156517094017094, + "grad_norm": 0.8082157969474792, + "learning_rate": 8.789395862936383e-05, + "loss": 1.1462, + "step": 12113 + }, + { + "epoch": 2.156695156695157, + "grad_norm": 0.898435652256012, + "learning_rate": 8.788006422526881e-05, + "loss": 0.8044, + "step": 12114 + }, + { + "epoch": 2.156873219373219, + "grad_norm": 0.9474737048149109, + "learning_rate": 8.786617005863879e-05, + "loss": 0.9089, + "step": 12115 + }, + { + "epoch": 2.157051282051282, + "grad_norm": 0.7898718118667603, + "learning_rate": 8.785227612974594e-05, + "loss": 0.9758, + "step": 12116 + }, + { + "epoch": 2.1572293447293447, + "grad_norm": 0.6734052300453186, + "learning_rate": 8.783838243886253e-05, + "loss": 0.7835, + "step": 12117 + }, + { + "epoch": 2.1574074074074074, + "grad_norm": 0.9381069540977478, + "learning_rate": 8.782448898626072e-05, + "loss": 0.7666, + "step": 12118 + }, + { + "epoch": 2.15758547008547, + "grad_norm": 0.8677506446838379, + "learning_rate": 8.781059577221276e-05, + "loss": 0.7442, + "step": 12119 + }, + { + "epoch": 2.157763532763533, + "grad_norm": 0.8244445323944092, + "learning_rate": 8.779670279699086e-05, + "loss": 0.8104, + "step": 12120 + }, + { + "epoch": 2.1579415954415953, + "grad_norm": 0.7984805703163147, + "learning_rate": 8.77828100608672e-05, + "loss": 1.029, + "step": 12121 + }, + { + "epoch": 2.158119658119658, + "grad_norm": 0.7817366123199463, + "learning_rate": 8.776891756411405e-05, + "loss": 0.6797, + "step": 12122 + }, + { + "epoch": 2.1582977207977208, + "grad_norm": 0.7084082365036011, + "learning_rate": 8.77550253070035e-05, + "loss": 0.688, + "step": 12123 + }, + { + "epoch": 2.1584757834757835, + "grad_norm": 0.7659782767295837, + "learning_rate": 8.774113328980782e-05, + "loss": 0.8691, + "step": 12124 + }, + { + "epoch": 2.1586538461538463, + "grad_norm": 0.7010130286216736, + "learning_rate": 8.772724151279913e-05, + "loss": 0.7587, + "step": 12125 + }, + { + "epoch": 2.158831908831909, + "grad_norm": 0.8183525800704956, + "learning_rate": 8.771334997624973e-05, + "loss": 0.8696, + "step": 12126 + }, + { + "epoch": 2.1590099715099713, + "grad_norm": 0.7944908142089844, + "learning_rate": 8.769945868043164e-05, + "loss": 0.8625, + "step": 12127 + }, + { + "epoch": 2.159188034188034, + "grad_norm": 0.7710323333740234, + "learning_rate": 8.768556762561713e-05, + "loss": 0.7765, + "step": 12128 + }, + { + "epoch": 2.159366096866097, + "grad_norm": 0.7416872382164001, + "learning_rate": 8.767167681207833e-05, + "loss": 0.9151, + "step": 12129 + }, + { + "epoch": 2.1595441595441596, + "grad_norm": 0.9230012893676758, + "learning_rate": 8.765778624008744e-05, + "loss": 0.914, + "step": 12130 + }, + { + "epoch": 2.1597222222222223, + "grad_norm": 0.7468557357788086, + "learning_rate": 8.764389590991657e-05, + "loss": 0.8624, + "step": 12131 + }, + { + "epoch": 2.159900284900285, + "grad_norm": 0.7746220827102661, + "learning_rate": 8.763000582183791e-05, + "loss": 0.9683, + "step": 12132 + }, + { + "epoch": 2.1600783475783474, + "grad_norm": 0.8429577350616455, + "learning_rate": 8.761611597612356e-05, + "loss": 0.8808, + "step": 12133 + }, + { + "epoch": 2.16025641025641, + "grad_norm": 0.8117298483848572, + "learning_rate": 8.760222637304572e-05, + "loss": 0.7067, + "step": 12134 + }, + { + "epoch": 2.160434472934473, + "grad_norm": 0.7717329859733582, + "learning_rate": 8.758833701287647e-05, + "loss": 1.0001, + "step": 12135 + }, + { + "epoch": 2.1606125356125356, + "grad_norm": 0.8493856191635132, + "learning_rate": 8.7574447895888e-05, + "loss": 0.9883, + "step": 12136 + }, + { + "epoch": 2.1607905982905984, + "grad_norm": 0.8592587113380432, + "learning_rate": 8.75605590223524e-05, + "loss": 0.7437, + "step": 12137 + }, + { + "epoch": 2.160968660968661, + "grad_norm": 0.6487032771110535, + "learning_rate": 8.75466703925418e-05, + "loss": 0.6841, + "step": 12138 + }, + { + "epoch": 2.1611467236467234, + "grad_norm": 0.8449310660362244, + "learning_rate": 8.753278200672832e-05, + "loss": 0.9221, + "step": 12139 + }, + { + "epoch": 2.161324786324786, + "grad_norm": 0.9603136777877808, + "learning_rate": 8.751889386518407e-05, + "loss": 0.8664, + "step": 12140 + }, + { + "epoch": 2.161502849002849, + "grad_norm": 0.7288493514060974, + "learning_rate": 8.750500596818121e-05, + "loss": 0.5745, + "step": 12141 + }, + { + "epoch": 2.1616809116809117, + "grad_norm": 0.8626441955566406, + "learning_rate": 8.749111831599178e-05, + "loss": 0.8346, + "step": 12142 + }, + { + "epoch": 2.1618589743589745, + "grad_norm": 0.7634188532829285, + "learning_rate": 8.74772309088879e-05, + "loss": 0.7503, + "step": 12143 + }, + { + "epoch": 2.162037037037037, + "grad_norm": 0.8641456365585327, + "learning_rate": 8.746334374714167e-05, + "loss": 0.9033, + "step": 12144 + }, + { + "epoch": 2.1622150997150995, + "grad_norm": 0.8103315234184265, + "learning_rate": 8.744945683102517e-05, + "loss": 0.8181, + "step": 12145 + }, + { + "epoch": 2.1623931623931623, + "grad_norm": 1.2493078708648682, + "learning_rate": 8.743557016081047e-05, + "loss": 1.0308, + "step": 12146 + }, + { + "epoch": 2.162571225071225, + "grad_norm": 0.6447771191596985, + "learning_rate": 8.742168373676973e-05, + "loss": 0.6886, + "step": 12147 + }, + { + "epoch": 2.1627492877492878, + "grad_norm": 0.90229332447052, + "learning_rate": 8.740779755917492e-05, + "loss": 1.0361, + "step": 12148 + }, + { + "epoch": 2.1629273504273505, + "grad_norm": 0.7414017915725708, + "learning_rate": 8.739391162829818e-05, + "loss": 0.781, + "step": 12149 + }, + { + "epoch": 2.1631054131054133, + "grad_norm": 0.8897294998168945, + "learning_rate": 8.738002594441154e-05, + "loss": 0.7712, + "step": 12150 + }, + { + "epoch": 2.1632834757834756, + "grad_norm": 0.8515656590461731, + "learning_rate": 8.73661405077871e-05, + "loss": 0.8843, + "step": 12151 + }, + { + "epoch": 2.1634615384615383, + "grad_norm": 0.7901699542999268, + "learning_rate": 8.735225531869686e-05, + "loss": 0.8588, + "step": 12152 + }, + { + "epoch": 2.163639601139601, + "grad_norm": 0.7262305021286011, + "learning_rate": 8.733837037741295e-05, + "loss": 0.9257, + "step": 12153 + }, + { + "epoch": 2.163817663817664, + "grad_norm": 1.1076871156692505, + "learning_rate": 8.732448568420732e-05, + "loss": 0.9817, + "step": 12154 + }, + { + "epoch": 2.1639957264957266, + "grad_norm": 0.8384785652160645, + "learning_rate": 8.731060123935209e-05, + "loss": 0.8024, + "step": 12155 + }, + { + "epoch": 2.1641737891737893, + "grad_norm": 0.8376259803771973, + "learning_rate": 8.729671704311924e-05, + "loss": 1.0299, + "step": 12156 + }, + { + "epoch": 2.164351851851852, + "grad_norm": 0.8248558044433594, + "learning_rate": 8.728283309578089e-05, + "loss": 0.9557, + "step": 12157 + }, + { + "epoch": 2.1645299145299144, + "grad_norm": 0.7452875375747681, + "learning_rate": 8.726894939760894e-05, + "loss": 0.8267, + "step": 12158 + }, + { + "epoch": 2.164707977207977, + "grad_norm": 0.8329267501831055, + "learning_rate": 8.72550659488755e-05, + "loss": 0.9288, + "step": 12159 + }, + { + "epoch": 2.16488603988604, + "grad_norm": 0.8748268485069275, + "learning_rate": 8.724118274985259e-05, + "loss": 0.8663, + "step": 12160 + }, + { + "epoch": 2.1650641025641026, + "grad_norm": 0.6839116811752319, + "learning_rate": 8.722729980081217e-05, + "loss": 0.6067, + "step": 12161 + }, + { + "epoch": 2.1652421652421654, + "grad_norm": 0.8343674540519714, + "learning_rate": 8.721341710202632e-05, + "loss": 0.9611, + "step": 12162 + }, + { + "epoch": 2.1654202279202277, + "grad_norm": 0.7783843874931335, + "learning_rate": 8.719953465376695e-05, + "loss": 0.8921, + "step": 12163 + }, + { + "epoch": 2.1655982905982905, + "grad_norm": 0.8357030749320984, + "learning_rate": 8.718565245630615e-05, + "loss": 0.9189, + "step": 12164 + }, + { + "epoch": 2.165776353276353, + "grad_norm": 0.8150131702423096, + "learning_rate": 8.717177050991582e-05, + "loss": 0.7486, + "step": 12165 + }, + { + "epoch": 2.165954415954416, + "grad_norm": 0.7282506823539734, + "learning_rate": 8.715788881486807e-05, + "loss": 0.8894, + "step": 12166 + }, + { + "epoch": 2.1661324786324787, + "grad_norm": 0.9958226680755615, + "learning_rate": 8.714400737143475e-05, + "loss": 1.0359, + "step": 12167 + }, + { + "epoch": 2.1663105413105415, + "grad_norm": 0.7162553071975708, + "learning_rate": 8.713012617988796e-05, + "loss": 0.7728, + "step": 12168 + }, + { + "epoch": 2.166488603988604, + "grad_norm": 0.6364821791648865, + "learning_rate": 8.711624524049955e-05, + "loss": 0.5881, + "step": 12169 + }, + { + "epoch": 2.1666666666666665, + "grad_norm": 0.9431148767471313, + "learning_rate": 8.710236455354159e-05, + "loss": 0.8804, + "step": 12170 + }, + { + "epoch": 2.1668447293447293, + "grad_norm": 0.7328855395317078, + "learning_rate": 8.708848411928598e-05, + "loss": 0.7762, + "step": 12171 + }, + { + "epoch": 2.167022792022792, + "grad_norm": 0.7855633497238159, + "learning_rate": 8.707460393800472e-05, + "loss": 0.7687, + "step": 12172 + }, + { + "epoch": 2.1672008547008548, + "grad_norm": 0.8694273233413696, + "learning_rate": 8.706072400996973e-05, + "loss": 0.7153, + "step": 12173 + }, + { + "epoch": 2.1673789173789175, + "grad_norm": 0.7371255159378052, + "learning_rate": 8.704684433545299e-05, + "loss": 0.7901, + "step": 12174 + }, + { + "epoch": 2.16755698005698, + "grad_norm": 0.7719849944114685, + "learning_rate": 8.70329649147264e-05, + "loss": 0.7569, + "step": 12175 + }, + { + "epoch": 2.1677350427350426, + "grad_norm": 0.883618175983429, + "learning_rate": 8.701908574806197e-05, + "loss": 0.8941, + "step": 12176 + }, + { + "epoch": 2.1679131054131053, + "grad_norm": 0.9455791711807251, + "learning_rate": 8.700520683573155e-05, + "loss": 0.8596, + "step": 12177 + }, + { + "epoch": 2.168091168091168, + "grad_norm": 0.7487229108810425, + "learning_rate": 8.69913281780071e-05, + "loss": 0.7353, + "step": 12178 + }, + { + "epoch": 2.168269230769231, + "grad_norm": 0.8050364255905151, + "learning_rate": 8.697744977516062e-05, + "loss": 0.8564, + "step": 12179 + }, + { + "epoch": 2.1684472934472936, + "grad_norm": 0.759355902671814, + "learning_rate": 8.69635716274639e-05, + "loss": 0.7128, + "step": 12180 + }, + { + "epoch": 2.1686253561253563, + "grad_norm": 0.8730760216712952, + "learning_rate": 8.694969373518892e-05, + "loss": 0.9944, + "step": 12181 + }, + { + "epoch": 2.1688034188034186, + "grad_norm": 0.7761130332946777, + "learning_rate": 8.693581609860756e-05, + "loss": 0.6845, + "step": 12182 + }, + { + "epoch": 2.1689814814814814, + "grad_norm": 0.8118788003921509, + "learning_rate": 8.692193871799181e-05, + "loss": 0.798, + "step": 12183 + }, + { + "epoch": 2.169159544159544, + "grad_norm": 0.8340219855308533, + "learning_rate": 8.690806159361344e-05, + "loss": 0.9754, + "step": 12184 + }, + { + "epoch": 2.169337606837607, + "grad_norm": 0.7515831589698792, + "learning_rate": 8.689418472574444e-05, + "loss": 0.998, + "step": 12185 + }, + { + "epoch": 2.1695156695156697, + "grad_norm": 0.7781083583831787, + "learning_rate": 8.688030811465665e-05, + "loss": 1.0152, + "step": 12186 + }, + { + "epoch": 2.169693732193732, + "grad_norm": 0.775097131729126, + "learning_rate": 8.6866431760622e-05, + "loss": 0.808, + "step": 12187 + }, + { + "epoch": 2.1698717948717947, + "grad_norm": 0.8288158178329468, + "learning_rate": 8.68525556639123e-05, + "loss": 0.9172, + "step": 12188 + }, + { + "epoch": 2.1700498575498575, + "grad_norm": 0.7754917740821838, + "learning_rate": 8.68386798247995e-05, + "loss": 0.719, + "step": 12189 + }, + { + "epoch": 2.17022792022792, + "grad_norm": 0.786685585975647, + "learning_rate": 8.682480424355539e-05, + "loss": 0.8883, + "step": 12190 + }, + { + "epoch": 2.170405982905983, + "grad_norm": 0.9970952272415161, + "learning_rate": 8.681092892045189e-05, + "loss": 0.9258, + "step": 12191 + }, + { + "epoch": 2.1705840455840457, + "grad_norm": 0.9732664227485657, + "learning_rate": 8.679705385576082e-05, + "loss": 0.8916, + "step": 12192 + }, + { + "epoch": 2.1707621082621085, + "grad_norm": 0.8557142615318298, + "learning_rate": 8.67831790497541e-05, + "loss": 0.8908, + "step": 12193 + }, + { + "epoch": 2.1709401709401708, + "grad_norm": 0.8564930558204651, + "learning_rate": 8.676930450270347e-05, + "loss": 1.054, + "step": 12194 + }, + { + "epoch": 2.1711182336182335, + "grad_norm": 0.785732090473175, + "learning_rate": 8.675543021488087e-05, + "loss": 0.7459, + "step": 12195 + }, + { + "epoch": 2.1712962962962963, + "grad_norm": 0.8739910125732422, + "learning_rate": 8.674155618655809e-05, + "loss": 0.8464, + "step": 12196 + }, + { + "epoch": 2.171474358974359, + "grad_norm": 0.8624834418296814, + "learning_rate": 8.672768241800699e-05, + "loss": 0.9405, + "step": 12197 + }, + { + "epoch": 2.171652421652422, + "grad_norm": 0.7948583364486694, + "learning_rate": 8.671380890949936e-05, + "loss": 1.0271, + "step": 12198 + }, + { + "epoch": 2.1718304843304845, + "grad_norm": 0.8078029155731201, + "learning_rate": 8.669993566130704e-05, + "loss": 0.6845, + "step": 12199 + }, + { + "epoch": 2.172008547008547, + "grad_norm": 0.7599586844444275, + "learning_rate": 8.668606267370187e-05, + "loss": 0.8438, + "step": 12200 + }, + { + "epoch": 2.1721866096866096, + "grad_norm": 0.8085161447525024, + "learning_rate": 8.667218994695562e-05, + "loss": 0.8398, + "step": 12201 + }, + { + "epoch": 2.1723646723646723, + "grad_norm": 0.9033090472221375, + "learning_rate": 8.665831748134019e-05, + "loss": 0.8591, + "step": 12202 + }, + { + "epoch": 2.172542735042735, + "grad_norm": 0.8638277649879456, + "learning_rate": 8.664444527712726e-05, + "loss": 0.8276, + "step": 12203 + }, + { + "epoch": 2.172720797720798, + "grad_norm": 0.8230745792388916, + "learning_rate": 8.663057333458871e-05, + "loss": 0.8663, + "step": 12204 + }, + { + "epoch": 2.1728988603988606, + "grad_norm": 0.8588439226150513, + "learning_rate": 8.661670165399626e-05, + "loss": 0.8543, + "step": 12205 + }, + { + "epoch": 2.173076923076923, + "grad_norm": 0.8461976051330566, + "learning_rate": 8.660283023562177e-05, + "loss": 0.8973, + "step": 12206 + }, + { + "epoch": 2.1732549857549857, + "grad_norm": 0.7768828868865967, + "learning_rate": 8.658895907973697e-05, + "loss": 0.7285, + "step": 12207 + }, + { + "epoch": 2.1734330484330484, + "grad_norm": 0.7384130954742432, + "learning_rate": 8.65750881866137e-05, + "loss": 0.7654, + "step": 12208 + }, + { + "epoch": 2.173611111111111, + "grad_norm": 0.8700957298278809, + "learning_rate": 8.656121755652365e-05, + "loss": 1.067, + "step": 12209 + }, + { + "epoch": 2.173789173789174, + "grad_norm": 0.8067826628684998, + "learning_rate": 8.654734718973863e-05, + "loss": 0.9863, + "step": 12210 + }, + { + "epoch": 2.1739672364672367, + "grad_norm": 0.7515989542007446, + "learning_rate": 8.653347708653039e-05, + "loss": 0.8434, + "step": 12211 + }, + { + "epoch": 2.174145299145299, + "grad_norm": 0.8280966877937317, + "learning_rate": 8.651960724717072e-05, + "loss": 1.0065, + "step": 12212 + }, + { + "epoch": 2.1743233618233617, + "grad_norm": 0.7988734841346741, + "learning_rate": 8.650573767193132e-05, + "loss": 0.7892, + "step": 12213 + }, + { + "epoch": 2.1745014245014245, + "grad_norm": 0.785323977470398, + "learning_rate": 8.649186836108399e-05, + "loss": 0.7563, + "step": 12214 + }, + { + "epoch": 2.1746794871794872, + "grad_norm": 0.7884892821311951, + "learning_rate": 8.64779993149004e-05, + "loss": 0.7225, + "step": 12215 + }, + { + "epoch": 2.17485754985755, + "grad_norm": 0.9376154541969299, + "learning_rate": 8.646413053365235e-05, + "loss": 0.8103, + "step": 12216 + }, + { + "epoch": 2.1750356125356127, + "grad_norm": 0.9680297374725342, + "learning_rate": 8.64502620176115e-05, + "loss": 0.8924, + "step": 12217 + }, + { + "epoch": 2.175213675213675, + "grad_norm": 1.016848087310791, + "learning_rate": 8.643639376704964e-05, + "loss": 0.9017, + "step": 12218 + }, + { + "epoch": 2.175391737891738, + "grad_norm": 0.790868878364563, + "learning_rate": 8.64225257822385e-05, + "loss": 0.7597, + "step": 12219 + }, + { + "epoch": 2.1755698005698005, + "grad_norm": 0.7539415955543518, + "learning_rate": 8.640865806344974e-05, + "loss": 0.7801, + "step": 12220 + }, + { + "epoch": 2.1757478632478633, + "grad_norm": 0.7288404703140259, + "learning_rate": 8.63947906109551e-05, + "loss": 0.6753, + "step": 12221 + }, + { + "epoch": 2.175925925925926, + "grad_norm": 0.8449869155883789, + "learning_rate": 8.638092342502623e-05, + "loss": 0.7165, + "step": 12222 + }, + { + "epoch": 2.176103988603989, + "grad_norm": 0.8210735321044922, + "learning_rate": 8.636705650593495e-05, + "loss": 0.8677, + "step": 12223 + }, + { + "epoch": 2.176282051282051, + "grad_norm": 0.7431774735450745, + "learning_rate": 8.635318985395284e-05, + "loss": 0.7914, + "step": 12224 + }, + { + "epoch": 2.176460113960114, + "grad_norm": 0.8223997354507446, + "learning_rate": 8.633932346935165e-05, + "loss": 0.7243, + "step": 12225 + }, + { + "epoch": 2.1766381766381766, + "grad_norm": 1.0101778507232666, + "learning_rate": 8.632545735240299e-05, + "loss": 0.8608, + "step": 12226 + }, + { + "epoch": 2.1768162393162394, + "grad_norm": 0.7270255088806152, + "learning_rate": 8.631159150337862e-05, + "loss": 0.8699, + "step": 12227 + }, + { + "epoch": 2.176994301994302, + "grad_norm": 0.8687323331832886, + "learning_rate": 8.629772592255016e-05, + "loss": 0.9168, + "step": 12228 + }, + { + "epoch": 2.177172364672365, + "grad_norm": 0.7623698115348816, + "learning_rate": 8.628386061018934e-05, + "loss": 0.9012, + "step": 12229 + }, + { + "epoch": 2.177350427350427, + "grad_norm": 0.7458708882331848, + "learning_rate": 8.626999556656771e-05, + "loss": 0.9068, + "step": 12230 + }, + { + "epoch": 2.17752849002849, + "grad_norm": 0.8262876868247986, + "learning_rate": 8.625613079195704e-05, + "loss": 0.9425, + "step": 12231 + }, + { + "epoch": 2.1777065527065527, + "grad_norm": 0.8737035989761353, + "learning_rate": 8.624226628662893e-05, + "loss": 0.9943, + "step": 12232 + }, + { + "epoch": 2.1778846153846154, + "grad_norm": 0.8250965476036072, + "learning_rate": 8.622840205085505e-05, + "loss": 0.9237, + "step": 12233 + }, + { + "epoch": 2.178062678062678, + "grad_norm": 0.8689019680023193, + "learning_rate": 8.621453808490699e-05, + "loss": 0.8544, + "step": 12234 + }, + { + "epoch": 2.178240740740741, + "grad_norm": 0.8672708868980408, + "learning_rate": 8.620067438905643e-05, + "loss": 0.7623, + "step": 12235 + }, + { + "epoch": 2.1784188034188032, + "grad_norm": 0.7077436447143555, + "learning_rate": 8.6186810963575e-05, + "loss": 0.7468, + "step": 12236 + }, + { + "epoch": 2.178596866096866, + "grad_norm": 0.838474452495575, + "learning_rate": 8.617294780873433e-05, + "loss": 0.9207, + "step": 12237 + }, + { + "epoch": 2.1787749287749287, + "grad_norm": 0.7240039110183716, + "learning_rate": 8.615908492480598e-05, + "loss": 0.8981, + "step": 12238 + }, + { + "epoch": 2.1789529914529915, + "grad_norm": 0.6995998620986938, + "learning_rate": 8.614522231206162e-05, + "loss": 0.7131, + "step": 12239 + }, + { + "epoch": 2.1791310541310542, + "grad_norm": 0.7011054158210754, + "learning_rate": 8.613135997077288e-05, + "loss": 0.8138, + "step": 12240 + }, + { + "epoch": 2.179309116809117, + "grad_norm": 0.9815019369125366, + "learning_rate": 8.611749790121131e-05, + "loss": 0.9637, + "step": 12241 + }, + { + "epoch": 2.1794871794871793, + "grad_norm": 0.7523870468139648, + "learning_rate": 8.610363610364853e-05, + "loss": 0.7555, + "step": 12242 + }, + { + "epoch": 2.179665242165242, + "grad_norm": 0.8286668658256531, + "learning_rate": 8.608977457835612e-05, + "loss": 0.7911, + "step": 12243 + }, + { + "epoch": 2.179843304843305, + "grad_norm": 0.8183441758155823, + "learning_rate": 8.607591332560573e-05, + "loss": 0.793, + "step": 12244 + }, + { + "epoch": 2.1800213675213675, + "grad_norm": 0.7104299068450928, + "learning_rate": 8.606205234566885e-05, + "loss": 0.6856, + "step": 12245 + }, + { + "epoch": 2.1801994301994303, + "grad_norm": 0.871588945388794, + "learning_rate": 8.60481916388171e-05, + "loss": 0.8342, + "step": 12246 + }, + { + "epoch": 2.180377492877493, + "grad_norm": 0.8002356290817261, + "learning_rate": 8.603433120532206e-05, + "loss": 0.9451, + "step": 12247 + }, + { + "epoch": 2.1805555555555554, + "grad_norm": 0.8223865032196045, + "learning_rate": 8.602047104545532e-05, + "loss": 0.8446, + "step": 12248 + }, + { + "epoch": 2.180733618233618, + "grad_norm": 0.7381762266159058, + "learning_rate": 8.600661115948836e-05, + "loss": 0.7575, + "step": 12249 + }, + { + "epoch": 2.180911680911681, + "grad_norm": 0.8717563152313232, + "learning_rate": 8.599275154769284e-05, + "loss": 0.9615, + "step": 12250 + }, + { + "epoch": 2.1810897435897436, + "grad_norm": 0.7935179471969604, + "learning_rate": 8.597889221034022e-05, + "loss": 0.8603, + "step": 12251 + }, + { + "epoch": 2.1812678062678064, + "grad_norm": 0.6350329518318176, + "learning_rate": 8.596503314770208e-05, + "loss": 0.4981, + "step": 12252 + }, + { + "epoch": 2.181445868945869, + "grad_norm": 0.8739648461341858, + "learning_rate": 8.595117436004995e-05, + "loss": 0.7872, + "step": 12253 + }, + { + "epoch": 2.1816239316239314, + "grad_norm": 0.8199412822723389, + "learning_rate": 8.593731584765542e-05, + "loss": 0.7905, + "step": 12254 + }, + { + "epoch": 2.181801994301994, + "grad_norm": 0.7289649844169617, + "learning_rate": 8.592345761078993e-05, + "loss": 0.6981, + "step": 12255 + }, + { + "epoch": 2.181980056980057, + "grad_norm": 0.8234626650810242, + "learning_rate": 8.590959964972506e-05, + "loss": 1.0442, + "step": 12256 + }, + { + "epoch": 2.1821581196581197, + "grad_norm": 0.7804498076438904, + "learning_rate": 8.589574196473229e-05, + "loss": 0.9386, + "step": 12257 + }, + { + "epoch": 2.1823361823361824, + "grad_norm": 0.9459218382835388, + "learning_rate": 8.588188455608317e-05, + "loss": 0.8075, + "step": 12258 + }, + { + "epoch": 2.182514245014245, + "grad_norm": 0.8133191466331482, + "learning_rate": 8.586802742404924e-05, + "loss": 1.0275, + "step": 12259 + }, + { + "epoch": 2.1826923076923075, + "grad_norm": 0.8302663564682007, + "learning_rate": 8.58541705689019e-05, + "loss": 0.9887, + "step": 12260 + }, + { + "epoch": 2.1828703703703702, + "grad_norm": 0.7839202284812927, + "learning_rate": 8.584031399091274e-05, + "loss": 1.0256, + "step": 12261 + }, + { + "epoch": 2.183048433048433, + "grad_norm": 0.8050578236579895, + "learning_rate": 8.582645769035319e-05, + "loss": 0.843, + "step": 12262 + }, + { + "epoch": 2.1832264957264957, + "grad_norm": 0.749110221862793, + "learning_rate": 8.581260166749477e-05, + "loss": 0.7683, + "step": 12263 + }, + { + "epoch": 2.1834045584045585, + "grad_norm": 0.7982701659202576, + "learning_rate": 8.579874592260894e-05, + "loss": 0.942, + "step": 12264 + }, + { + "epoch": 2.1835826210826212, + "grad_norm": 0.7571866512298584, + "learning_rate": 8.57848904559672e-05, + "loss": 0.8828, + "step": 12265 + }, + { + "epoch": 2.183760683760684, + "grad_norm": 0.7445113658905029, + "learning_rate": 8.577103526784098e-05, + "loss": 0.8869, + "step": 12266 + }, + { + "epoch": 2.1839387464387463, + "grad_norm": 0.7999380230903625, + "learning_rate": 8.575718035850177e-05, + "loss": 0.9476, + "step": 12267 + }, + { + "epoch": 2.184116809116809, + "grad_norm": 0.7188777923583984, + "learning_rate": 8.574332572822103e-05, + "loss": 0.7961, + "step": 12268 + }, + { + "epoch": 2.184294871794872, + "grad_norm": 0.9545742869377136, + "learning_rate": 8.572947137727023e-05, + "loss": 0.8629, + "step": 12269 + }, + { + "epoch": 2.1844729344729346, + "grad_norm": 0.8066838979721069, + "learning_rate": 8.571561730592075e-05, + "loss": 0.7728, + "step": 12270 + }, + { + "epoch": 2.1846509971509973, + "grad_norm": 0.7819525003433228, + "learning_rate": 8.57017635144441e-05, + "loss": 0.9897, + "step": 12271 + }, + { + "epoch": 2.1848290598290596, + "grad_norm": 0.9249349236488342, + "learning_rate": 8.568791000311166e-05, + "loss": 0.8562, + "step": 12272 + }, + { + "epoch": 2.1850071225071224, + "grad_norm": 0.8118993043899536, + "learning_rate": 8.567405677219497e-05, + "loss": 0.819, + "step": 12273 + }, + { + "epoch": 2.185185185185185, + "grad_norm": 0.7858524322509766, + "learning_rate": 8.566020382196532e-05, + "loss": 1.0708, + "step": 12274 + }, + { + "epoch": 2.185363247863248, + "grad_norm": 1.0223300457000732, + "learning_rate": 8.564635115269422e-05, + "loss": 0.9929, + "step": 12275 + }, + { + "epoch": 2.1855413105413106, + "grad_norm": 0.7749526500701904, + "learning_rate": 8.5632498764653e-05, + "loss": 0.7555, + "step": 12276 + }, + { + "epoch": 2.1857193732193734, + "grad_norm": 0.8443665504455566, + "learning_rate": 8.561864665811313e-05, + "loss": 0.8488, + "step": 12277 + }, + { + "epoch": 2.185897435897436, + "grad_norm": 0.7482786178588867, + "learning_rate": 8.560479483334603e-05, + "loss": 0.6535, + "step": 12278 + }, + { + "epoch": 2.1860754985754984, + "grad_norm": 0.7981070876121521, + "learning_rate": 8.559094329062305e-05, + "loss": 0.7698, + "step": 12279 + }, + { + "epoch": 2.186253561253561, + "grad_norm": 0.7612428665161133, + "learning_rate": 8.557709203021564e-05, + "loss": 0.9086, + "step": 12280 + }, + { + "epoch": 2.186431623931624, + "grad_norm": 0.8246445059776306, + "learning_rate": 8.556324105239512e-05, + "loss": 0.9927, + "step": 12281 + }, + { + "epoch": 2.1866096866096867, + "grad_norm": 0.8902820348739624, + "learning_rate": 8.554939035743292e-05, + "loss": 0.8474, + "step": 12282 + }, + { + "epoch": 2.1867877492877494, + "grad_norm": 0.9992623329162598, + "learning_rate": 8.553553994560037e-05, + "loss": 1.0898, + "step": 12283 + }, + { + "epoch": 2.1869658119658117, + "grad_norm": 0.9124125838279724, + "learning_rate": 8.552168981716892e-05, + "loss": 0.9542, + "step": 12284 + }, + { + "epoch": 2.1871438746438745, + "grad_norm": 0.6818730235099792, + "learning_rate": 8.550783997240983e-05, + "loss": 0.678, + "step": 12285 + }, + { + "epoch": 2.1873219373219372, + "grad_norm": 0.8302112817764282, + "learning_rate": 8.549399041159455e-05, + "loss": 0.9955, + "step": 12286 + }, + { + "epoch": 2.1875, + "grad_norm": 0.840419352054596, + "learning_rate": 8.548014113499436e-05, + "loss": 0.9299, + "step": 12287 + }, + { + "epoch": 2.1876780626780628, + "grad_norm": 0.8317474722862244, + "learning_rate": 8.546629214288067e-05, + "loss": 0.754, + "step": 12288 + }, + { + "epoch": 2.1878561253561255, + "grad_norm": 0.879148542881012, + "learning_rate": 8.545244343552476e-05, + "loss": 0.9238, + "step": 12289 + }, + { + "epoch": 2.1880341880341883, + "grad_norm": 0.7899607419967651, + "learning_rate": 8.543859501319805e-05, + "loss": 0.8277, + "step": 12290 + }, + { + "epoch": 2.1882122507122506, + "grad_norm": 0.837785542011261, + "learning_rate": 8.542474687617176e-05, + "loss": 0.8801, + "step": 12291 + }, + { + "epoch": 2.1883903133903133, + "grad_norm": 0.856213390827179, + "learning_rate": 8.541089902471733e-05, + "loss": 0.7589, + "step": 12292 + }, + { + "epoch": 2.188568376068376, + "grad_norm": 0.7915818095207214, + "learning_rate": 8.539705145910599e-05, + "loss": 0.8071, + "step": 12293 + }, + { + "epoch": 2.188746438746439, + "grad_norm": 0.8266519904136658, + "learning_rate": 8.538320417960914e-05, + "loss": 0.9482, + "step": 12294 + }, + { + "epoch": 2.1889245014245016, + "grad_norm": 0.8505687713623047, + "learning_rate": 8.536935718649799e-05, + "loss": 0.9272, + "step": 12295 + }, + { + "epoch": 2.189102564102564, + "grad_norm": 0.7530698776245117, + "learning_rate": 8.535551048004394e-05, + "loss": 0.7908, + "step": 12296 + }, + { + "epoch": 2.1892806267806266, + "grad_norm": 0.7904362678527832, + "learning_rate": 8.534166406051818e-05, + "loss": 0.9771, + "step": 12297 + }, + { + "epoch": 2.1894586894586894, + "grad_norm": 0.7860299944877625, + "learning_rate": 8.532781792819209e-05, + "loss": 0.605, + "step": 12298 + }, + { + "epoch": 2.189636752136752, + "grad_norm": 0.7718655467033386, + "learning_rate": 8.531397208333695e-05, + "loss": 0.8844, + "step": 12299 + }, + { + "epoch": 2.189814814814815, + "grad_norm": 0.8069637417793274, + "learning_rate": 8.530012652622397e-05, + "loss": 0.8571, + "step": 12300 + }, + { + "epoch": 2.1899928774928776, + "grad_norm": 0.8557140231132507, + "learning_rate": 8.528628125712455e-05, + "loss": 0.7396, + "step": 12301 + }, + { + "epoch": 2.1901709401709404, + "grad_norm": 0.8547600507736206, + "learning_rate": 8.527243627630983e-05, + "loss": 0.8073, + "step": 12302 + }, + { + "epoch": 2.1903490028490027, + "grad_norm": 0.8217329382896423, + "learning_rate": 8.525859158405114e-05, + "loss": 0.9723, + "step": 12303 + }, + { + "epoch": 2.1905270655270654, + "grad_norm": 0.896946132183075, + "learning_rate": 8.524474718061972e-05, + "loss": 0.7896, + "step": 12304 + }, + { + "epoch": 2.190705128205128, + "grad_norm": 1.0846823453903198, + "learning_rate": 8.523090306628685e-05, + "loss": 0.7689, + "step": 12305 + }, + { + "epoch": 2.190883190883191, + "grad_norm": 0.7265166640281677, + "learning_rate": 8.521705924132373e-05, + "loss": 0.8451, + "step": 12306 + }, + { + "epoch": 2.1910612535612537, + "grad_norm": 0.8806917071342468, + "learning_rate": 8.520321570600162e-05, + "loss": 0.8769, + "step": 12307 + }, + { + "epoch": 2.191239316239316, + "grad_norm": 0.7528414130210876, + "learning_rate": 8.518937246059176e-05, + "loss": 0.7137, + "step": 12308 + }, + { + "epoch": 2.1914173789173788, + "grad_norm": 0.9313900470733643, + "learning_rate": 8.517552950536543e-05, + "loss": 0.933, + "step": 12309 + }, + { + "epoch": 2.1915954415954415, + "grad_norm": 0.8363727331161499, + "learning_rate": 8.516168684059375e-05, + "loss": 0.899, + "step": 12310 + }, + { + "epoch": 2.1917735042735043, + "grad_norm": 0.7939122915267944, + "learning_rate": 8.514784446654803e-05, + "loss": 1.0323, + "step": 12311 + }, + { + "epoch": 2.191951566951567, + "grad_norm": 0.8744710087776184, + "learning_rate": 8.51340023834994e-05, + "loss": 0.8738, + "step": 12312 + }, + { + "epoch": 2.1921296296296298, + "grad_norm": 0.779353678226471, + "learning_rate": 8.512016059171916e-05, + "loss": 0.7692, + "step": 12313 + }, + { + "epoch": 2.1923076923076925, + "grad_norm": 0.8578362464904785, + "learning_rate": 8.510631909147841e-05, + "loss": 1.0636, + "step": 12314 + }, + { + "epoch": 2.192485754985755, + "grad_norm": 0.7210206985473633, + "learning_rate": 8.509247788304846e-05, + "loss": 0.6342, + "step": 12315 + }, + { + "epoch": 2.1926638176638176, + "grad_norm": 0.7221980690956116, + "learning_rate": 8.50786369667004e-05, + "loss": 0.7022, + "step": 12316 + }, + { + "epoch": 2.1928418803418803, + "grad_norm": 0.7871465086936951, + "learning_rate": 8.506479634270544e-05, + "loss": 0.9349, + "step": 12317 + }, + { + "epoch": 2.193019943019943, + "grad_norm": 0.7396262884140015, + "learning_rate": 8.505095601133479e-05, + "loss": 0.8644, + "step": 12318 + }, + { + "epoch": 2.193198005698006, + "grad_norm": 0.7513349652290344, + "learning_rate": 8.503711597285959e-05, + "loss": 0.7881, + "step": 12319 + }, + { + "epoch": 2.1933760683760686, + "grad_norm": 0.7280148863792419, + "learning_rate": 8.502327622755106e-05, + "loss": 0.7115, + "step": 12320 + }, + { + "epoch": 2.193554131054131, + "grad_norm": 0.792238712310791, + "learning_rate": 8.500943677568028e-05, + "loss": 0.8193, + "step": 12321 + }, + { + "epoch": 2.1937321937321936, + "grad_norm": 0.8709526062011719, + "learning_rate": 8.499559761751847e-05, + "loss": 0.8957, + "step": 12322 + }, + { + "epoch": 2.1939102564102564, + "grad_norm": 0.6865217685699463, + "learning_rate": 8.498175875333674e-05, + "loss": 0.6853, + "step": 12323 + }, + { + "epoch": 2.194088319088319, + "grad_norm": 0.7797526121139526, + "learning_rate": 8.496792018340625e-05, + "loss": 0.8885, + "step": 12324 + }, + { + "epoch": 2.194266381766382, + "grad_norm": 0.8806295394897461, + "learning_rate": 8.495408190799814e-05, + "loss": 0.9322, + "step": 12325 + }, + { + "epoch": 2.1944444444444446, + "grad_norm": 0.8566734790802002, + "learning_rate": 8.494024392738355e-05, + "loss": 0.9363, + "step": 12326 + }, + { + "epoch": 2.194622507122507, + "grad_norm": 0.8058465123176575, + "learning_rate": 8.49264062418336e-05, + "loss": 0.9007, + "step": 12327 + }, + { + "epoch": 2.1948005698005697, + "grad_norm": 0.7895804643630981, + "learning_rate": 8.491256885161938e-05, + "loss": 0.8486, + "step": 12328 + }, + { + "epoch": 2.1949786324786325, + "grad_norm": 0.7626506686210632, + "learning_rate": 8.489873175701204e-05, + "loss": 0.8208, + "step": 12329 + }, + { + "epoch": 2.195156695156695, + "grad_norm": 0.8917649388313293, + "learning_rate": 8.488489495828272e-05, + "loss": 1.1036, + "step": 12330 + }, + { + "epoch": 2.195334757834758, + "grad_norm": 0.7614438533782959, + "learning_rate": 8.487105845570242e-05, + "loss": 0.7124, + "step": 12331 + }, + { + "epoch": 2.1955128205128207, + "grad_norm": 0.7697421312332153, + "learning_rate": 8.485722224954237e-05, + "loss": 0.8831, + "step": 12332 + }, + { + "epoch": 2.195690883190883, + "grad_norm": 0.7449761629104614, + "learning_rate": 8.484338634007354e-05, + "loss": 0.9115, + "step": 12333 + }, + { + "epoch": 2.1958689458689458, + "grad_norm": 0.7099741101264954, + "learning_rate": 8.482955072756709e-05, + "loss": 0.6907, + "step": 12334 + }, + { + "epoch": 2.1960470085470085, + "grad_norm": 0.7856435775756836, + "learning_rate": 8.481571541229406e-05, + "loss": 1.0565, + "step": 12335 + }, + { + "epoch": 2.1962250712250713, + "grad_norm": 0.8374622464179993, + "learning_rate": 8.48018803945256e-05, + "loss": 0.9191, + "step": 12336 + }, + { + "epoch": 2.196403133903134, + "grad_norm": 0.7530848383903503, + "learning_rate": 8.478804567453265e-05, + "loss": 0.6576, + "step": 12337 + }, + { + "epoch": 2.1965811965811968, + "grad_norm": 0.774861216545105, + "learning_rate": 8.477421125258637e-05, + "loss": 1.0258, + "step": 12338 + }, + { + "epoch": 2.196759259259259, + "grad_norm": 0.9623909592628479, + "learning_rate": 8.47603771289578e-05, + "loss": 1.0192, + "step": 12339 + }, + { + "epoch": 2.196937321937322, + "grad_norm": 0.8253501653671265, + "learning_rate": 8.474654330391797e-05, + "loss": 0.7823, + "step": 12340 + }, + { + "epoch": 2.1971153846153846, + "grad_norm": 0.8683596849441528, + "learning_rate": 8.473270977773797e-05, + "loss": 0.8002, + "step": 12341 + }, + { + "epoch": 2.1972934472934473, + "grad_norm": 0.9093332886695862, + "learning_rate": 8.471887655068877e-05, + "loss": 1.0315, + "step": 12342 + }, + { + "epoch": 2.19747150997151, + "grad_norm": 0.7313206791877747, + "learning_rate": 8.470504362304147e-05, + "loss": 0.8238, + "step": 12343 + }, + { + "epoch": 2.197649572649573, + "grad_norm": 0.8464672565460205, + "learning_rate": 8.469121099506703e-05, + "loss": 0.8104, + "step": 12344 + }, + { + "epoch": 2.197827635327635, + "grad_norm": 0.9213936924934387, + "learning_rate": 8.467737866703657e-05, + "loss": 0.9963, + "step": 12345 + }, + { + "epoch": 2.198005698005698, + "grad_norm": 0.8033352494239807, + "learning_rate": 8.466354663922099e-05, + "loss": 0.9788, + "step": 12346 + }, + { + "epoch": 2.1981837606837606, + "grad_norm": 0.7210986018180847, + "learning_rate": 8.464971491189141e-05, + "loss": 0.7597, + "step": 12347 + }, + { + "epoch": 2.1983618233618234, + "grad_norm": 0.8128374814987183, + "learning_rate": 8.463588348531872e-05, + "loss": 0.9575, + "step": 12348 + }, + { + "epoch": 2.198539886039886, + "grad_norm": 0.7276061773300171, + "learning_rate": 8.4622052359774e-05, + "loss": 0.815, + "step": 12349 + }, + { + "epoch": 2.198717948717949, + "grad_norm": 0.7463665008544922, + "learning_rate": 8.46082215355282e-05, + "loss": 0.9782, + "step": 12350 + }, + { + "epoch": 2.198896011396011, + "grad_norm": 0.8288317918777466, + "learning_rate": 8.459439101285238e-05, + "loss": 0.8206, + "step": 12351 + }, + { + "epoch": 2.199074074074074, + "grad_norm": 0.8286055326461792, + "learning_rate": 8.458056079201742e-05, + "loss": 0.9819, + "step": 12352 + }, + { + "epoch": 2.1992521367521367, + "grad_norm": 0.8138381242752075, + "learning_rate": 8.456673087329436e-05, + "loss": 0.8565, + "step": 12353 + }, + { + "epoch": 2.1994301994301995, + "grad_norm": 0.9059311747550964, + "learning_rate": 8.455290125695412e-05, + "loss": 0.8727, + "step": 12354 + }, + { + "epoch": 2.199608262108262, + "grad_norm": 0.6138933300971985, + "learning_rate": 8.453907194326773e-05, + "loss": 0.5635, + "step": 12355 + }, + { + "epoch": 2.199786324786325, + "grad_norm": 0.870585560798645, + "learning_rate": 8.452524293250608e-05, + "loss": 0.7401, + "step": 12356 + }, + { + "epoch": 2.1999643874643873, + "grad_norm": 0.8393024802207947, + "learning_rate": 8.451141422494013e-05, + "loss": 1.0083, + "step": 12357 + }, + { + "epoch": 2.20014245014245, + "grad_norm": 0.7667146325111389, + "learning_rate": 8.449758582084091e-05, + "loss": 0.8915, + "step": 12358 + }, + { + "epoch": 2.2003205128205128, + "grad_norm": 1.0229144096374512, + "learning_rate": 8.448375772047923e-05, + "loss": 0.8879, + "step": 12359 + }, + { + "epoch": 2.2004985754985755, + "grad_norm": 0.7670294046401978, + "learning_rate": 8.446992992412611e-05, + "loss": 0.8233, + "step": 12360 + }, + { + "epoch": 2.2006766381766383, + "grad_norm": 0.7110083103179932, + "learning_rate": 8.445610243205244e-05, + "loss": 0.6315, + "step": 12361 + }, + { + "epoch": 2.200854700854701, + "grad_norm": 0.7801400423049927, + "learning_rate": 8.444227524452918e-05, + "loss": 0.7758, + "step": 12362 + }, + { + "epoch": 2.2010327635327633, + "grad_norm": 0.8762022852897644, + "learning_rate": 8.44284483618272e-05, + "loss": 0.9308, + "step": 12363 + }, + { + "epoch": 2.201210826210826, + "grad_norm": 0.811890184879303, + "learning_rate": 8.441462178421742e-05, + "loss": 1.0322, + "step": 12364 + }, + { + "epoch": 2.201388888888889, + "grad_norm": 0.8128690719604492, + "learning_rate": 8.440079551197076e-05, + "loss": 1.0669, + "step": 12365 + }, + { + "epoch": 2.2015669515669516, + "grad_norm": 0.8925766348838806, + "learning_rate": 8.438696954535812e-05, + "loss": 0.8848, + "step": 12366 + }, + { + "epoch": 2.2017450142450143, + "grad_norm": 0.9104064106941223, + "learning_rate": 8.437314388465036e-05, + "loss": 0.8227, + "step": 12367 + }, + { + "epoch": 2.201923076923077, + "grad_norm": 0.7956777215003967, + "learning_rate": 8.43593185301184e-05, + "loss": 0.7616, + "step": 12368 + }, + { + "epoch": 2.2021011396011394, + "grad_norm": 0.7658423185348511, + "learning_rate": 8.434549348203309e-05, + "loss": 0.9406, + "step": 12369 + }, + { + "epoch": 2.202279202279202, + "grad_norm": 0.7650682926177979, + "learning_rate": 8.433166874066532e-05, + "loss": 0.9031, + "step": 12370 + }, + { + "epoch": 2.202457264957265, + "grad_norm": 0.8613301515579224, + "learning_rate": 8.431784430628594e-05, + "loss": 0.9184, + "step": 12371 + }, + { + "epoch": 2.2026353276353277, + "grad_norm": 0.8446599245071411, + "learning_rate": 8.430402017916586e-05, + "loss": 0.8639, + "step": 12372 + }, + { + "epoch": 2.2028133903133904, + "grad_norm": 0.8082340955734253, + "learning_rate": 8.429019635957585e-05, + "loss": 0.7365, + "step": 12373 + }, + { + "epoch": 2.202991452991453, + "grad_norm": 0.8843092918395996, + "learning_rate": 8.427637284778683e-05, + "loss": 0.8679, + "step": 12374 + }, + { + "epoch": 2.2031695156695155, + "grad_norm": 0.8475705981254578, + "learning_rate": 8.426254964406961e-05, + "loss": 0.6614, + "step": 12375 + }, + { + "epoch": 2.203347578347578, + "grad_norm": 0.9980667233467102, + "learning_rate": 8.424872674869507e-05, + "loss": 0.9103, + "step": 12376 + }, + { + "epoch": 2.203525641025641, + "grad_norm": 0.8033170104026794, + "learning_rate": 8.423490416193398e-05, + "loss": 0.7668, + "step": 12377 + }, + { + "epoch": 2.2037037037037037, + "grad_norm": 0.8275265097618103, + "learning_rate": 8.422108188405718e-05, + "loss": 0.7448, + "step": 12378 + }, + { + "epoch": 2.2038817663817665, + "grad_norm": 0.7622979283332825, + "learning_rate": 8.420725991533554e-05, + "loss": 0.8121, + "step": 12379 + }, + { + "epoch": 2.2040598290598292, + "grad_norm": 0.8580977320671082, + "learning_rate": 8.41934382560398e-05, + "loss": 0.8437, + "step": 12380 + }, + { + "epoch": 2.2042378917378915, + "grad_norm": 0.8443751931190491, + "learning_rate": 8.417961690644086e-05, + "loss": 0.971, + "step": 12381 + }, + { + "epoch": 2.2044159544159543, + "grad_norm": 0.782430112361908, + "learning_rate": 8.416579586680939e-05, + "loss": 0.8367, + "step": 12382 + }, + { + "epoch": 2.204594017094017, + "grad_norm": 0.8664544820785522, + "learning_rate": 8.415197513741633e-05, + "loss": 0.8288, + "step": 12383 + }, + { + "epoch": 2.20477207977208, + "grad_norm": 0.7207586169242859, + "learning_rate": 8.413815471853235e-05, + "loss": 0.8038, + "step": 12384 + }, + { + "epoch": 2.2049501424501425, + "grad_norm": 0.743195652961731, + "learning_rate": 8.412433461042828e-05, + "loss": 0.705, + "step": 12385 + }, + { + "epoch": 2.2051282051282053, + "grad_norm": 0.7891412377357483, + "learning_rate": 8.411051481337488e-05, + "loss": 0.9729, + "step": 12386 + }, + { + "epoch": 2.205306267806268, + "grad_norm": 0.838847815990448, + "learning_rate": 8.4096695327643e-05, + "loss": 0.9053, + "step": 12387 + }, + { + "epoch": 2.2054843304843303, + "grad_norm": 0.7717056274414062, + "learning_rate": 8.408287615350328e-05, + "loss": 0.7388, + "step": 12388 + }, + { + "epoch": 2.205662393162393, + "grad_norm": 0.7209389209747314, + "learning_rate": 8.406905729122654e-05, + "loss": 0.6411, + "step": 12389 + }, + { + "epoch": 2.205840455840456, + "grad_norm": 0.822475790977478, + "learning_rate": 8.405523874108354e-05, + "loss": 0.9574, + "step": 12390 + }, + { + "epoch": 2.2060185185185186, + "grad_norm": 0.9401286840438843, + "learning_rate": 8.404142050334504e-05, + "loss": 0.8915, + "step": 12391 + }, + { + "epoch": 2.2061965811965814, + "grad_norm": 0.8247103691101074, + "learning_rate": 8.40276025782817e-05, + "loss": 0.8369, + "step": 12392 + }, + { + "epoch": 2.2063746438746437, + "grad_norm": 0.8082301020622253, + "learning_rate": 8.401378496616437e-05, + "loss": 0.9321, + "step": 12393 + }, + { + "epoch": 2.2065527065527064, + "grad_norm": 0.8156028389930725, + "learning_rate": 8.399996766726367e-05, + "loss": 0.7599, + "step": 12394 + }, + { + "epoch": 2.206730769230769, + "grad_norm": 0.7941898107528687, + "learning_rate": 8.398615068185038e-05, + "loss": 0.812, + "step": 12395 + }, + { + "epoch": 2.206908831908832, + "grad_norm": 0.7013470530509949, + "learning_rate": 8.397233401019518e-05, + "loss": 0.7914, + "step": 12396 + }, + { + "epoch": 2.2070868945868947, + "grad_norm": 0.6028649210929871, + "learning_rate": 8.395851765256881e-05, + "loss": 0.5787, + "step": 12397 + }, + { + "epoch": 2.2072649572649574, + "grad_norm": 0.9031504392623901, + "learning_rate": 8.3944701609242e-05, + "loss": 0.8677, + "step": 12398 + }, + { + "epoch": 2.20744301994302, + "grad_norm": 0.7370864748954773, + "learning_rate": 8.393088588048536e-05, + "loss": 0.9025, + "step": 12399 + }, + { + "epoch": 2.2076210826210825, + "grad_norm": 0.7764220237731934, + "learning_rate": 8.391707046656968e-05, + "loss": 0.8805, + "step": 12400 + }, + { + "epoch": 2.2077991452991452, + "grad_norm": 0.7456721663475037, + "learning_rate": 8.390325536776553e-05, + "loss": 0.7739, + "step": 12401 + }, + { + "epoch": 2.207977207977208, + "grad_norm": 0.8032360076904297, + "learning_rate": 8.388944058434373e-05, + "loss": 0.9765, + "step": 12402 + }, + { + "epoch": 2.2081552706552707, + "grad_norm": 0.8502830266952515, + "learning_rate": 8.387562611657483e-05, + "loss": 0.9356, + "step": 12403 + }, + { + "epoch": 2.2083333333333335, + "grad_norm": 0.812216579914093, + "learning_rate": 8.386181196472956e-05, + "loss": 0.8846, + "step": 12404 + }, + { + "epoch": 2.208511396011396, + "grad_norm": 0.6996115446090698, + "learning_rate": 8.384799812907853e-05, + "loss": 0.7035, + "step": 12405 + }, + { + "epoch": 2.2086894586894585, + "grad_norm": 0.7909261584281921, + "learning_rate": 8.383418460989245e-05, + "loss": 0.8025, + "step": 12406 + }, + { + "epoch": 2.2088675213675213, + "grad_norm": 0.8278310894966125, + "learning_rate": 8.382037140744192e-05, + "loss": 0.7982, + "step": 12407 + }, + { + "epoch": 2.209045584045584, + "grad_norm": 0.7558199167251587, + "learning_rate": 8.380655852199763e-05, + "loss": 0.854, + "step": 12408 + }, + { + "epoch": 2.209223646723647, + "grad_norm": 0.8516034483909607, + "learning_rate": 8.379274595383016e-05, + "loss": 0.7497, + "step": 12409 + }, + { + "epoch": 2.2094017094017095, + "grad_norm": 0.777004599571228, + "learning_rate": 8.377893370321018e-05, + "loss": 0.797, + "step": 12410 + }, + { + "epoch": 2.2095797720797723, + "grad_norm": 0.8820251822471619, + "learning_rate": 8.376512177040829e-05, + "loss": 0.9229, + "step": 12411 + }, + { + "epoch": 2.2097578347578346, + "grad_norm": 0.8623200058937073, + "learning_rate": 8.375131015569514e-05, + "loss": 1.011, + "step": 12412 + }, + { + "epoch": 2.2099358974358974, + "grad_norm": 0.9192054271697998, + "learning_rate": 8.373749885934127e-05, + "loss": 0.8711, + "step": 12413 + }, + { + "epoch": 2.21011396011396, + "grad_norm": 0.7627860903739929, + "learning_rate": 8.372368788161736e-05, + "loss": 0.5937, + "step": 12414 + }, + { + "epoch": 2.210292022792023, + "grad_norm": 0.74603670835495, + "learning_rate": 8.370987722279395e-05, + "loss": 0.8238, + "step": 12415 + }, + { + "epoch": 2.2104700854700856, + "grad_norm": 0.884469211101532, + "learning_rate": 8.369606688314165e-05, + "loss": 1.1957, + "step": 12416 + }, + { + "epoch": 2.210648148148148, + "grad_norm": 0.8145224452018738, + "learning_rate": 8.36822568629311e-05, + "loss": 0.8517, + "step": 12417 + }, + { + "epoch": 2.2108262108262107, + "grad_norm": 0.8167604207992554, + "learning_rate": 8.366844716243279e-05, + "loss": 0.9701, + "step": 12418 + }, + { + "epoch": 2.2110042735042734, + "grad_norm": 0.7668562531471252, + "learning_rate": 8.365463778191736e-05, + "loss": 1.0281, + "step": 12419 + }, + { + "epoch": 2.211182336182336, + "grad_norm": 0.8455148339271545, + "learning_rate": 8.364082872165532e-05, + "loss": 0.7812, + "step": 12420 + }, + { + "epoch": 2.211360398860399, + "grad_norm": 0.8756504654884338, + "learning_rate": 8.362701998191728e-05, + "loss": 0.779, + "step": 12421 + }, + { + "epoch": 2.2115384615384617, + "grad_norm": 0.8239594101905823, + "learning_rate": 8.361321156297374e-05, + "loss": 0.8581, + "step": 12422 + }, + { + "epoch": 2.2117165242165244, + "grad_norm": 0.7719405889511108, + "learning_rate": 8.359940346509533e-05, + "loss": 0.7593, + "step": 12423 + }, + { + "epoch": 2.2118945868945867, + "grad_norm": 0.8607308268547058, + "learning_rate": 8.358559568855249e-05, + "loss": 1.0618, + "step": 12424 + }, + { + "epoch": 2.2120726495726495, + "grad_norm": 0.750431478023529, + "learning_rate": 8.357178823361582e-05, + "loss": 0.7779, + "step": 12425 + }, + { + "epoch": 2.2122507122507122, + "grad_norm": 0.7770674824714661, + "learning_rate": 8.355798110055583e-05, + "loss": 0.6837, + "step": 12426 + }, + { + "epoch": 2.212428774928775, + "grad_norm": 0.7924200296401978, + "learning_rate": 8.354417428964307e-05, + "loss": 0.8092, + "step": 12427 + }, + { + "epoch": 2.2126068376068377, + "grad_norm": 0.7784677743911743, + "learning_rate": 8.3530367801148e-05, + "loss": 0.7168, + "step": 12428 + }, + { + "epoch": 2.2127849002849005, + "grad_norm": 1.0548151731491089, + "learning_rate": 8.351656163534121e-05, + "loss": 0.9286, + "step": 12429 + }, + { + "epoch": 2.212962962962963, + "grad_norm": 0.8983006477355957, + "learning_rate": 8.35027557924931e-05, + "loss": 0.878, + "step": 12430 + }, + { + "epoch": 2.2131410256410255, + "grad_norm": 0.8136780261993408, + "learning_rate": 8.348895027287424e-05, + "loss": 0.7901, + "step": 12431 + }, + { + "epoch": 2.2133190883190883, + "grad_norm": 0.8186678290367126, + "learning_rate": 8.347514507675508e-05, + "loss": 0.8994, + "step": 12432 + }, + { + "epoch": 2.213497150997151, + "grad_norm": 0.880790650844574, + "learning_rate": 8.346134020440617e-05, + "loss": 1.0681, + "step": 12433 + }, + { + "epoch": 2.213675213675214, + "grad_norm": 0.8061994910240173, + "learning_rate": 8.344753565609789e-05, + "loss": 0.8466, + "step": 12434 + }, + { + "epoch": 2.2138532763532766, + "grad_norm": 0.8041423559188843, + "learning_rate": 8.34337314321008e-05, + "loss": 0.897, + "step": 12435 + }, + { + "epoch": 2.214031339031339, + "grad_norm": 0.5797891616821289, + "learning_rate": 8.34199275326853e-05, + "loss": 0.4827, + "step": 12436 + }, + { + "epoch": 2.2142094017094016, + "grad_norm": 0.7373392581939697, + "learning_rate": 8.340612395812188e-05, + "loss": 0.779, + "step": 12437 + }, + { + "epoch": 2.2143874643874644, + "grad_norm": 0.7852202653884888, + "learning_rate": 8.339232070868102e-05, + "loss": 0.8001, + "step": 12438 + }, + { + "epoch": 2.214565527065527, + "grad_norm": 0.8209689259529114, + "learning_rate": 8.337851778463311e-05, + "loss": 0.7492, + "step": 12439 + }, + { + "epoch": 2.21474358974359, + "grad_norm": 0.9393492937088013, + "learning_rate": 8.336471518624867e-05, + "loss": 0.884, + "step": 12440 + }, + { + "epoch": 2.2149216524216526, + "grad_norm": 0.6966122984886169, + "learning_rate": 8.3350912913798e-05, + "loss": 0.7364, + "step": 12441 + }, + { + "epoch": 2.215099715099715, + "grad_norm": 0.7379066944122314, + "learning_rate": 8.333711096755165e-05, + "loss": 0.7345, + "step": 12442 + }, + { + "epoch": 2.2152777777777777, + "grad_norm": 0.9011021256446838, + "learning_rate": 8.332330934777999e-05, + "loss": 0.8392, + "step": 12443 + }, + { + "epoch": 2.2154558404558404, + "grad_norm": 0.7718381285667419, + "learning_rate": 8.330950805475346e-05, + "loss": 0.9062, + "step": 12444 + }, + { + "epoch": 2.215633903133903, + "grad_norm": 0.8584564328193665, + "learning_rate": 8.329570708874241e-05, + "loss": 0.9612, + "step": 12445 + }, + { + "epoch": 2.215811965811966, + "grad_norm": 0.7711616158485413, + "learning_rate": 8.32819064500173e-05, + "loss": 0.731, + "step": 12446 + }, + { + "epoch": 2.2159900284900287, + "grad_norm": 0.8014609217643738, + "learning_rate": 8.326810613884849e-05, + "loss": 1.0128, + "step": 12447 + }, + { + "epoch": 2.216168091168091, + "grad_norm": 0.7837486863136292, + "learning_rate": 8.325430615550642e-05, + "loss": 0.8271, + "step": 12448 + }, + { + "epoch": 2.2163461538461537, + "grad_norm": 0.9399738907814026, + "learning_rate": 8.324050650026139e-05, + "loss": 1.0433, + "step": 12449 + }, + { + "epoch": 2.2165242165242165, + "grad_norm": 0.8302193284034729, + "learning_rate": 8.322670717338385e-05, + "loss": 1.0259, + "step": 12450 + }, + { + "epoch": 2.2167022792022792, + "grad_norm": 0.7707721590995789, + "learning_rate": 8.321290817514411e-05, + "loss": 0.6972, + "step": 12451 + }, + { + "epoch": 2.216880341880342, + "grad_norm": 0.5814536809921265, + "learning_rate": 8.319910950581261e-05, + "loss": 0.5846, + "step": 12452 + }, + { + "epoch": 2.2170584045584047, + "grad_norm": 0.8249124884605408, + "learning_rate": 8.318531116565962e-05, + "loss": 0.7417, + "step": 12453 + }, + { + "epoch": 2.217236467236467, + "grad_norm": 0.7116015553474426, + "learning_rate": 8.317151315495556e-05, + "loss": 0.8698, + "step": 12454 + }, + { + "epoch": 2.21741452991453, + "grad_norm": 0.8025332689285278, + "learning_rate": 8.31577154739707e-05, + "loss": 0.825, + "step": 12455 + }, + { + "epoch": 2.2175925925925926, + "grad_norm": 0.8962773680686951, + "learning_rate": 8.314391812297542e-05, + "loss": 0.9987, + "step": 12456 + }, + { + "epoch": 2.2177706552706553, + "grad_norm": 0.8446899652481079, + "learning_rate": 8.313012110224008e-05, + "loss": 0.8554, + "step": 12457 + }, + { + "epoch": 2.217948717948718, + "grad_norm": 0.7759326696395874, + "learning_rate": 8.311632441203494e-05, + "loss": 0.8206, + "step": 12458 + }, + { + "epoch": 2.218126780626781, + "grad_norm": 0.9782015085220337, + "learning_rate": 8.31025280526304e-05, + "loss": 0.8183, + "step": 12459 + }, + { + "epoch": 2.218304843304843, + "grad_norm": 0.7445226907730103, + "learning_rate": 8.308873202429666e-05, + "loss": 0.6819, + "step": 12460 + }, + { + "epoch": 2.218482905982906, + "grad_norm": 0.7613980770111084, + "learning_rate": 8.307493632730413e-05, + "loss": 0.6283, + "step": 12461 + }, + { + "epoch": 2.2186609686609686, + "grad_norm": 0.7437549829483032, + "learning_rate": 8.306114096192304e-05, + "loss": 0.7511, + "step": 12462 + }, + { + "epoch": 2.2188390313390314, + "grad_norm": 0.7600140571594238, + "learning_rate": 8.304734592842373e-05, + "loss": 0.8784, + "step": 12463 + }, + { + "epoch": 2.219017094017094, + "grad_norm": 0.9086898565292358, + "learning_rate": 8.303355122707644e-05, + "loss": 1.0818, + "step": 12464 + }, + { + "epoch": 2.219195156695157, + "grad_norm": 0.8674180507659912, + "learning_rate": 8.30197568581515e-05, + "loss": 0.8925, + "step": 12465 + }, + { + "epoch": 2.219373219373219, + "grad_norm": 0.893606960773468, + "learning_rate": 8.300596282191911e-05, + "loss": 0.9382, + "step": 12466 + }, + { + "epoch": 2.219551282051282, + "grad_norm": 0.7664543390274048, + "learning_rate": 8.29921691186496e-05, + "loss": 0.7893, + "step": 12467 + }, + { + "epoch": 2.2197293447293447, + "grad_norm": 0.8730209469795227, + "learning_rate": 8.297837574861318e-05, + "loss": 1.0509, + "step": 12468 + }, + { + "epoch": 2.2199074074074074, + "grad_norm": 0.8138112425804138, + "learning_rate": 8.296458271208018e-05, + "loss": 0.784, + "step": 12469 + }, + { + "epoch": 2.22008547008547, + "grad_norm": 0.8362413644790649, + "learning_rate": 8.295079000932073e-05, + "loss": 1.0236, + "step": 12470 + }, + { + "epoch": 2.220263532763533, + "grad_norm": 0.8422487378120422, + "learning_rate": 8.293699764060518e-05, + "loss": 0.9677, + "step": 12471 + }, + { + "epoch": 2.2204415954415953, + "grad_norm": 0.7290427088737488, + "learning_rate": 8.292320560620369e-05, + "loss": 0.7514, + "step": 12472 + }, + { + "epoch": 2.220619658119658, + "grad_norm": 0.8083370923995972, + "learning_rate": 8.290941390638653e-05, + "loss": 0.8136, + "step": 12473 + }, + { + "epoch": 2.2207977207977208, + "grad_norm": 0.8045510053634644, + "learning_rate": 8.289562254142389e-05, + "loss": 0.6753, + "step": 12474 + }, + { + "epoch": 2.2209757834757835, + "grad_norm": 0.8019934892654419, + "learning_rate": 8.288183151158602e-05, + "loss": 0.8147, + "step": 12475 + }, + { + "epoch": 2.2211538461538463, + "grad_norm": 0.8129584193229675, + "learning_rate": 8.286804081714306e-05, + "loss": 0.9137, + "step": 12476 + }, + { + "epoch": 2.221331908831909, + "grad_norm": 0.9729450345039368, + "learning_rate": 8.285425045836526e-05, + "loss": 0.7884, + "step": 12477 + }, + { + "epoch": 2.2215099715099713, + "grad_norm": 0.755081295967102, + "learning_rate": 8.284046043552282e-05, + "loss": 0.8496, + "step": 12478 + }, + { + "epoch": 2.221688034188034, + "grad_norm": 0.725267767906189, + "learning_rate": 8.282667074888589e-05, + "loss": 0.7054, + "step": 12479 + }, + { + "epoch": 2.221866096866097, + "grad_norm": 0.832098662853241, + "learning_rate": 8.281288139872472e-05, + "loss": 0.8729, + "step": 12480 + }, + { + "epoch": 2.2220441595441596, + "grad_norm": 0.9908086657524109, + "learning_rate": 8.27990923853094e-05, + "loss": 0.9106, + "step": 12481 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 0.8001172542572021, + "learning_rate": 8.278530370891013e-05, + "loss": 0.906, + "step": 12482 + }, + { + "epoch": 2.222400284900285, + "grad_norm": 0.7607424259185791, + "learning_rate": 8.277151536979709e-05, + "loss": 0.8125, + "step": 12483 + }, + { + "epoch": 2.2225783475783474, + "grad_norm": 0.7850996255874634, + "learning_rate": 8.275772736824042e-05, + "loss": 0.7017, + "step": 12484 + }, + { + "epoch": 2.22275641025641, + "grad_norm": 0.8376613855361938, + "learning_rate": 8.274393970451024e-05, + "loss": 1.0453, + "step": 12485 + }, + { + "epoch": 2.222934472934473, + "grad_norm": 0.7973353266716003, + "learning_rate": 8.273015237887673e-05, + "loss": 0.8337, + "step": 12486 + }, + { + "epoch": 2.2231125356125356, + "grad_norm": 0.7622607350349426, + "learning_rate": 8.271636539161e-05, + "loss": 0.8574, + "step": 12487 + }, + { + "epoch": 2.2232905982905984, + "grad_norm": 0.7839400768280029, + "learning_rate": 8.270257874298022e-05, + "loss": 0.8857, + "step": 12488 + }, + { + "epoch": 2.223468660968661, + "grad_norm": 0.7730473875999451, + "learning_rate": 8.268879243325743e-05, + "loss": 0.9578, + "step": 12489 + }, + { + "epoch": 2.2236467236467234, + "grad_norm": 0.7811899185180664, + "learning_rate": 8.267500646271184e-05, + "loss": 0.9469, + "step": 12490 + }, + { + "epoch": 2.223824786324786, + "grad_norm": 0.8570041060447693, + "learning_rate": 8.266122083161347e-05, + "loss": 0.8853, + "step": 12491 + }, + { + "epoch": 2.224002849002849, + "grad_norm": 0.7989770174026489, + "learning_rate": 8.264743554023248e-05, + "loss": 0.7467, + "step": 12492 + }, + { + "epoch": 2.2241809116809117, + "grad_norm": 0.8287475109100342, + "learning_rate": 8.263365058883891e-05, + "loss": 0.9987, + "step": 12493 + }, + { + "epoch": 2.2243589743589745, + "grad_norm": 0.8879026174545288, + "learning_rate": 8.261986597770295e-05, + "loss": 0.9503, + "step": 12494 + }, + { + "epoch": 2.224537037037037, + "grad_norm": 0.8153596520423889, + "learning_rate": 8.260608170709456e-05, + "loss": 0.9715, + "step": 12495 + }, + { + "epoch": 2.2247150997150995, + "grad_norm": 0.8294584155082703, + "learning_rate": 8.259229777728384e-05, + "loss": 0.958, + "step": 12496 + }, + { + "epoch": 2.2248931623931623, + "grad_norm": 0.76850426197052, + "learning_rate": 8.257851418854093e-05, + "loss": 0.7666, + "step": 12497 + }, + { + "epoch": 2.225071225071225, + "grad_norm": 0.743966817855835, + "learning_rate": 8.256473094113582e-05, + "loss": 0.8893, + "step": 12498 + }, + { + "epoch": 2.2252492877492878, + "grad_norm": 0.7339308857917786, + "learning_rate": 8.255094803533863e-05, + "loss": 0.7317, + "step": 12499 + }, + { + "epoch": 2.2254273504273505, + "grad_norm": 1.0800104141235352, + "learning_rate": 8.253716547141932e-05, + "loss": 1.0147, + "step": 12500 + }, + { + "epoch": 2.2256054131054133, + "grad_norm": 0.8518815636634827, + "learning_rate": 8.252338324964802e-05, + "loss": 0.9695, + "step": 12501 + }, + { + "epoch": 2.2257834757834756, + "grad_norm": 0.8706745505332947, + "learning_rate": 8.250960137029469e-05, + "loss": 0.7735, + "step": 12502 + }, + { + "epoch": 2.2259615384615383, + "grad_norm": 1.0482546091079712, + "learning_rate": 8.24958198336294e-05, + "loss": 1.0882, + "step": 12503 + }, + { + "epoch": 2.226139601139601, + "grad_norm": 0.8025278449058533, + "learning_rate": 8.248203863992213e-05, + "loss": 0.8573, + "step": 12504 + }, + { + "epoch": 2.226317663817664, + "grad_norm": 0.8267400860786438, + "learning_rate": 8.246825778944297e-05, + "loss": 0.8609, + "step": 12505 + }, + { + "epoch": 2.2264957264957266, + "grad_norm": 0.703681230545044, + "learning_rate": 8.245447728246184e-05, + "loss": 0.6934, + "step": 12506 + }, + { + "epoch": 2.2266737891737893, + "grad_norm": 0.807736873626709, + "learning_rate": 8.24406971192488e-05, + "loss": 0.7258, + "step": 12507 + }, + { + "epoch": 2.226851851851852, + "grad_norm": 0.7663748860359192, + "learning_rate": 8.24269173000738e-05, + "loss": 0.7825, + "step": 12508 + }, + { + "epoch": 2.2270299145299144, + "grad_norm": 0.7799240946769714, + "learning_rate": 8.24131378252069e-05, + "loss": 0.7868, + "step": 12509 + }, + { + "epoch": 2.227207977207977, + "grad_norm": 0.8309668302536011, + "learning_rate": 8.239935869491799e-05, + "loss": 0.7697, + "step": 12510 + }, + { + "epoch": 2.22738603988604, + "grad_norm": 0.7257094979286194, + "learning_rate": 8.23855799094771e-05, + "loss": 0.8168, + "step": 12511 + }, + { + "epoch": 2.2275641025641026, + "grad_norm": 0.8902100920677185, + "learning_rate": 8.237180146915416e-05, + "loss": 0.8606, + "step": 12512 + }, + { + "epoch": 2.2277421652421654, + "grad_norm": 0.8100315928459167, + "learning_rate": 8.235802337421919e-05, + "loss": 0.9225, + "step": 12513 + }, + { + "epoch": 2.2279202279202277, + "grad_norm": 0.6804848909378052, + "learning_rate": 8.234424562494205e-05, + "loss": 0.7047, + "step": 12514 + }, + { + "epoch": 2.2280982905982905, + "grad_norm": 0.8664964437484741, + "learning_rate": 8.233046822159276e-05, + "loss": 1.0255, + "step": 12515 + }, + { + "epoch": 2.228276353276353, + "grad_norm": 0.836857795715332, + "learning_rate": 8.231669116444128e-05, + "loss": 0.9818, + "step": 12516 + }, + { + "epoch": 2.228454415954416, + "grad_norm": 0.6999024748802185, + "learning_rate": 8.230291445375744e-05, + "loss": 0.7298, + "step": 12517 + }, + { + "epoch": 2.2286324786324787, + "grad_norm": 0.8676811456680298, + "learning_rate": 8.228913808981127e-05, + "loss": 0.9592, + "step": 12518 + }, + { + "epoch": 2.2288105413105415, + "grad_norm": 0.8088808655738831, + "learning_rate": 8.227536207287263e-05, + "loss": 1.0021, + "step": 12519 + }, + { + "epoch": 2.228988603988604, + "grad_norm": 0.7802120447158813, + "learning_rate": 8.226158640321149e-05, + "loss": 0.8519, + "step": 12520 + }, + { + "epoch": 2.2291666666666665, + "grad_norm": 0.7560334801673889, + "learning_rate": 8.224781108109766e-05, + "loss": 0.7676, + "step": 12521 + }, + { + "epoch": 2.2293447293447293, + "grad_norm": 0.7806954383850098, + "learning_rate": 8.223403610680113e-05, + "loss": 0.9151, + "step": 12522 + }, + { + "epoch": 2.229522792022792, + "grad_norm": 0.7972870469093323, + "learning_rate": 8.222026148059173e-05, + "loss": 0.8785, + "step": 12523 + }, + { + "epoch": 2.2297008547008548, + "grad_norm": 0.7868863344192505, + "learning_rate": 8.220648720273941e-05, + "loss": 0.8981, + "step": 12524 + }, + { + "epoch": 2.2298789173789175, + "grad_norm": 0.7388648390769958, + "learning_rate": 8.219271327351397e-05, + "loss": 0.7361, + "step": 12525 + }, + { + "epoch": 2.23005698005698, + "grad_norm": 0.7367138862609863, + "learning_rate": 8.217893969318538e-05, + "loss": 0.7357, + "step": 12526 + }, + { + "epoch": 2.2302350427350426, + "grad_norm": 0.8345077037811279, + "learning_rate": 8.216516646202339e-05, + "loss": 0.9671, + "step": 12527 + }, + { + "epoch": 2.2304131054131053, + "grad_norm": 0.7875744104385376, + "learning_rate": 8.215139358029793e-05, + "loss": 0.7991, + "step": 12528 + }, + { + "epoch": 2.230591168091168, + "grad_norm": 0.7444638609886169, + "learning_rate": 8.213762104827882e-05, + "loss": 0.6524, + "step": 12529 + }, + { + "epoch": 2.230769230769231, + "grad_norm": 0.6670697927474976, + "learning_rate": 8.212384886623597e-05, + "loss": 0.639, + "step": 12530 + }, + { + "epoch": 2.2309472934472936, + "grad_norm": 0.8348705172538757, + "learning_rate": 8.211007703443913e-05, + "loss": 0.8904, + "step": 12531 + }, + { + "epoch": 2.2311253561253563, + "grad_norm": 0.8458212614059448, + "learning_rate": 8.209630555315817e-05, + "loss": 0.8398, + "step": 12532 + }, + { + "epoch": 2.2313034188034186, + "grad_norm": 0.9043961763381958, + "learning_rate": 8.20825344226629e-05, + "loss": 0.904, + "step": 12533 + }, + { + "epoch": 2.2314814814814814, + "grad_norm": 0.8207734227180481, + "learning_rate": 8.206876364322319e-05, + "loss": 0.853, + "step": 12534 + }, + { + "epoch": 2.231659544159544, + "grad_norm": 0.9311240911483765, + "learning_rate": 8.205499321510876e-05, + "loss": 0.9807, + "step": 12535 + }, + { + "epoch": 2.231837606837607, + "grad_norm": 0.8379791378974915, + "learning_rate": 8.204122313858946e-05, + "loss": 0.8318, + "step": 12536 + }, + { + "epoch": 2.2320156695156697, + "grad_norm": 0.8078454732894897, + "learning_rate": 8.202745341393515e-05, + "loss": 0.8692, + "step": 12537 + }, + { + "epoch": 2.232193732193732, + "grad_norm": 0.7555927038192749, + "learning_rate": 8.201368404141547e-05, + "loss": 0.8514, + "step": 12538 + }, + { + "epoch": 2.2323717948717947, + "grad_norm": 0.7724241018295288, + "learning_rate": 8.199991502130035e-05, + "loss": 0.5758, + "step": 12539 + }, + { + "epoch": 2.2325498575498575, + "grad_norm": 0.7388870120048523, + "learning_rate": 8.198614635385946e-05, + "loss": 0.6265, + "step": 12540 + }, + { + "epoch": 2.23272792022792, + "grad_norm": 0.9006723761558533, + "learning_rate": 8.197237803936267e-05, + "loss": 0.8238, + "step": 12541 + }, + { + "epoch": 2.232905982905983, + "grad_norm": 0.917884349822998, + "learning_rate": 8.195861007807962e-05, + "loss": 0.9447, + "step": 12542 + }, + { + "epoch": 2.2330840455840457, + "grad_norm": 0.81849205493927, + "learning_rate": 8.194484247028016e-05, + "loss": 0.9071, + "step": 12543 + }, + { + "epoch": 2.2332621082621085, + "grad_norm": 0.8572089076042175, + "learning_rate": 8.193107521623398e-05, + "loss": 0.9068, + "step": 12544 + }, + { + "epoch": 2.2334401709401708, + "grad_norm": 0.7870976328849792, + "learning_rate": 8.19173083162109e-05, + "loss": 0.7595, + "step": 12545 + }, + { + "epoch": 2.2336182336182335, + "grad_norm": 0.8728759288787842, + "learning_rate": 8.190354177048055e-05, + "loss": 1.0974, + "step": 12546 + }, + { + "epoch": 2.2337962962962963, + "grad_norm": 0.7679606080055237, + "learning_rate": 8.188977557931274e-05, + "loss": 0.7068, + "step": 12547 + }, + { + "epoch": 2.233974358974359, + "grad_norm": 0.7753520011901855, + "learning_rate": 8.187600974297714e-05, + "loss": 0.8008, + "step": 12548 + }, + { + "epoch": 2.234152421652422, + "grad_norm": 0.7785305976867676, + "learning_rate": 8.186224426174348e-05, + "loss": 0.8528, + "step": 12549 + }, + { + "epoch": 2.2343304843304845, + "grad_norm": 0.7762976288795471, + "learning_rate": 8.184847913588145e-05, + "loss": 0.9264, + "step": 12550 + }, + { + "epoch": 2.234508547008547, + "grad_norm": 1.0543726682662964, + "learning_rate": 8.18347143656608e-05, + "loss": 0.9201, + "step": 12551 + }, + { + "epoch": 2.2346866096866096, + "grad_norm": 0.815389096736908, + "learning_rate": 8.182094995135116e-05, + "loss": 0.8834, + "step": 12552 + }, + { + "epoch": 2.2348646723646723, + "grad_norm": 0.774773895740509, + "learning_rate": 8.180718589322225e-05, + "loss": 0.8864, + "step": 12553 + }, + { + "epoch": 2.235042735042735, + "grad_norm": 0.8139658570289612, + "learning_rate": 8.179342219154372e-05, + "loss": 0.8696, + "step": 12554 + }, + { + "epoch": 2.235220797720798, + "grad_norm": 0.7804924249649048, + "learning_rate": 8.177965884658527e-05, + "loss": 0.8854, + "step": 12555 + }, + { + "epoch": 2.2353988603988606, + "grad_norm": 0.8601226210594177, + "learning_rate": 8.176589585861659e-05, + "loss": 0.9115, + "step": 12556 + }, + { + "epoch": 2.235576923076923, + "grad_norm": 0.7518162727355957, + "learning_rate": 8.175213322790726e-05, + "loss": 0.7871, + "step": 12557 + }, + { + "epoch": 2.2357549857549857, + "grad_norm": 0.7595868110656738, + "learning_rate": 8.1738370954727e-05, + "loss": 0.7597, + "step": 12558 + }, + { + "epoch": 2.2359330484330484, + "grad_norm": 0.8191643357276917, + "learning_rate": 8.17246090393454e-05, + "loss": 0.9443, + "step": 12559 + }, + { + "epoch": 2.236111111111111, + "grad_norm": 0.7854904532432556, + "learning_rate": 8.171084748203217e-05, + "loss": 0.8547, + "step": 12560 + }, + { + "epoch": 2.236289173789174, + "grad_norm": 0.8610023260116577, + "learning_rate": 8.169708628305684e-05, + "loss": 0.7846, + "step": 12561 + }, + { + "epoch": 2.2364672364672367, + "grad_norm": 0.8254715204238892, + "learning_rate": 8.168332544268914e-05, + "loss": 0.7493, + "step": 12562 + }, + { + "epoch": 2.236645299145299, + "grad_norm": 0.8390897512435913, + "learning_rate": 8.166956496119857e-05, + "loss": 0.9867, + "step": 12563 + }, + { + "epoch": 2.2368233618233617, + "grad_norm": 0.8179677128791809, + "learning_rate": 8.165580483885483e-05, + "loss": 0.8039, + "step": 12564 + }, + { + "epoch": 2.2370014245014245, + "grad_norm": 0.6722155809402466, + "learning_rate": 8.164204507592745e-05, + "loss": 0.695, + "step": 12565 + }, + { + "epoch": 2.2371794871794872, + "grad_norm": 0.8228170871734619, + "learning_rate": 8.162828567268612e-05, + "loss": 1.0414, + "step": 12566 + }, + { + "epoch": 2.23735754985755, + "grad_norm": 0.8676900267601013, + "learning_rate": 8.161452662940032e-05, + "loss": 1.0157, + "step": 12567 + }, + { + "epoch": 2.2375356125356127, + "grad_norm": 0.8174694180488586, + "learning_rate": 8.16007679463397e-05, + "loss": 0.691, + "step": 12568 + }, + { + "epoch": 2.237713675213675, + "grad_norm": 0.8137148022651672, + "learning_rate": 8.158700962377379e-05, + "loss": 1.0022, + "step": 12569 + }, + { + "epoch": 2.237891737891738, + "grad_norm": 0.970250129699707, + "learning_rate": 8.157325166197221e-05, + "loss": 0.7946, + "step": 12570 + }, + { + "epoch": 2.2380698005698005, + "grad_norm": 0.7366915941238403, + "learning_rate": 8.155949406120446e-05, + "loss": 0.9039, + "step": 12571 + }, + { + "epoch": 2.2382478632478633, + "grad_norm": 0.878358781337738, + "learning_rate": 8.154573682174014e-05, + "loss": 0.8172, + "step": 12572 + }, + { + "epoch": 2.238425925925926, + "grad_norm": 0.7552989721298218, + "learning_rate": 8.153197994384875e-05, + "loss": 0.9955, + "step": 12573 + }, + { + "epoch": 2.238603988603989, + "grad_norm": 0.8198257684707642, + "learning_rate": 8.151822342779985e-05, + "loss": 0.8677, + "step": 12574 + }, + { + "epoch": 2.238782051282051, + "grad_norm": 0.9128977656364441, + "learning_rate": 8.150446727386297e-05, + "loss": 0.9531, + "step": 12575 + }, + { + "epoch": 2.238960113960114, + "grad_norm": 0.867671549320221, + "learning_rate": 8.149071148230762e-05, + "loss": 1.0226, + "step": 12576 + }, + { + "epoch": 2.2391381766381766, + "grad_norm": 0.8640758395195007, + "learning_rate": 8.147695605340337e-05, + "loss": 1.1284, + "step": 12577 + }, + { + "epoch": 2.2393162393162394, + "grad_norm": 0.7453210353851318, + "learning_rate": 8.146320098741964e-05, + "loss": 0.7812, + "step": 12578 + }, + { + "epoch": 2.239494301994302, + "grad_norm": 0.9207521080970764, + "learning_rate": 8.144944628462602e-05, + "loss": 0.9955, + "step": 12579 + }, + { + "epoch": 2.239672364672365, + "grad_norm": 0.751732349395752, + "learning_rate": 8.143569194529193e-05, + "loss": 0.7858, + "step": 12580 + }, + { + "epoch": 2.239850427350427, + "grad_norm": 0.7955539226531982, + "learning_rate": 8.142193796968694e-05, + "loss": 0.8482, + "step": 12581 + }, + { + "epoch": 2.24002849002849, + "grad_norm": 0.8020164370536804, + "learning_rate": 8.140818435808043e-05, + "loss": 0.8069, + "step": 12582 + }, + { + "epoch": 2.2402065527065527, + "grad_norm": 0.7460235357284546, + "learning_rate": 8.139443111074198e-05, + "loss": 0.6478, + "step": 12583 + }, + { + "epoch": 2.2403846153846154, + "grad_norm": 0.7504379153251648, + "learning_rate": 8.138067822794096e-05, + "loss": 0.726, + "step": 12584 + }, + { + "epoch": 2.240562678062678, + "grad_norm": 0.8214267492294312, + "learning_rate": 8.136692570994688e-05, + "loss": 1.0114, + "step": 12585 + }, + { + "epoch": 2.240740740740741, + "grad_norm": 0.9436941742897034, + "learning_rate": 8.135317355702917e-05, + "loss": 0.873, + "step": 12586 + }, + { + "epoch": 2.2409188034188032, + "grad_norm": 0.7541804909706116, + "learning_rate": 8.133942176945733e-05, + "loss": 0.8013, + "step": 12587 + }, + { + "epoch": 2.241096866096866, + "grad_norm": 0.8725557327270508, + "learning_rate": 8.132567034750073e-05, + "loss": 0.8506, + "step": 12588 + }, + { + "epoch": 2.2412749287749287, + "grad_norm": 0.7766169905662537, + "learning_rate": 8.131191929142882e-05, + "loss": 0.9076, + "step": 12589 + }, + { + "epoch": 2.2414529914529915, + "grad_norm": 0.8852736353874207, + "learning_rate": 8.129816860151104e-05, + "loss": 0.9278, + "step": 12590 + }, + { + "epoch": 2.2416310541310542, + "grad_norm": 0.6939527988433838, + "learning_rate": 8.128441827801681e-05, + "loss": 0.753, + "step": 12591 + }, + { + "epoch": 2.241809116809117, + "grad_norm": 0.8932832479476929, + "learning_rate": 8.127066832121551e-05, + "loss": 0.8089, + "step": 12592 + }, + { + "epoch": 2.2419871794871793, + "grad_norm": 0.7399743795394897, + "learning_rate": 8.125691873137656e-05, + "loss": 0.6905, + "step": 12593 + }, + { + "epoch": 2.242165242165242, + "grad_norm": 0.7664098143577576, + "learning_rate": 8.124316950876933e-05, + "loss": 0.8698, + "step": 12594 + }, + { + "epoch": 2.242343304843305, + "grad_norm": 0.8222574591636658, + "learning_rate": 8.122942065366323e-05, + "loss": 0.8922, + "step": 12595 + }, + { + "epoch": 2.2425213675213675, + "grad_norm": 0.8072433471679688, + "learning_rate": 8.121567216632771e-05, + "loss": 0.7613, + "step": 12596 + }, + { + "epoch": 2.2426994301994303, + "grad_norm": 0.7647300362586975, + "learning_rate": 8.120192404703199e-05, + "loss": 0.8736, + "step": 12597 + }, + { + "epoch": 2.242877492877493, + "grad_norm": 0.7536396980285645, + "learning_rate": 8.118817629604559e-05, + "loss": 0.7697, + "step": 12598 + }, + { + "epoch": 2.2430555555555554, + "grad_norm": 0.7295291423797607, + "learning_rate": 8.117442891363774e-05, + "loss": 0.8477, + "step": 12599 + }, + { + "epoch": 2.243233618233618, + "grad_norm": 0.7677894830703735, + "learning_rate": 8.116068190007787e-05, + "loss": 0.8113, + "step": 12600 + }, + { + "epoch": 2.243411680911681, + "grad_norm": 0.825614869594574, + "learning_rate": 8.114693525563529e-05, + "loss": 0.915, + "step": 12601 + }, + { + "epoch": 2.2435897435897436, + "grad_norm": 0.7841798663139343, + "learning_rate": 8.113318898057939e-05, + "loss": 0.7028, + "step": 12602 + }, + { + "epoch": 2.2437678062678064, + "grad_norm": 1.085337519645691, + "learning_rate": 8.111944307517942e-05, + "loss": 0.8354, + "step": 12603 + }, + { + "epoch": 2.243945868945869, + "grad_norm": 0.7831527590751648, + "learning_rate": 8.110569753970475e-05, + "loss": 1.0275, + "step": 12604 + }, + { + "epoch": 2.2441239316239314, + "grad_norm": 0.800504744052887, + "learning_rate": 8.109195237442467e-05, + "loss": 0.7006, + "step": 12605 + }, + { + "epoch": 2.244301994301994, + "grad_norm": 0.8189738392829895, + "learning_rate": 8.107820757960856e-05, + "loss": 0.8036, + "step": 12606 + }, + { + "epoch": 2.244480056980057, + "grad_norm": 0.8892425298690796, + "learning_rate": 8.106446315552562e-05, + "loss": 0.8274, + "step": 12607 + }, + { + "epoch": 2.2446581196581197, + "grad_norm": 0.8144643306732178, + "learning_rate": 8.105071910244521e-05, + "loss": 1.0648, + "step": 12608 + }, + { + "epoch": 2.2448361823361824, + "grad_norm": 0.914513111114502, + "learning_rate": 8.103697542063657e-05, + "loss": 0.8999, + "step": 12609 + }, + { + "epoch": 2.245014245014245, + "grad_norm": 0.8273763656616211, + "learning_rate": 8.102323211036904e-05, + "loss": 0.8554, + "step": 12610 + }, + { + "epoch": 2.2451923076923075, + "grad_norm": 0.9459149837493896, + "learning_rate": 8.100948917191181e-05, + "loss": 1.2345, + "step": 12611 + }, + { + "epoch": 2.2453703703703702, + "grad_norm": 0.8377025723457336, + "learning_rate": 8.099574660553425e-05, + "loss": 0.8096, + "step": 12612 + }, + { + "epoch": 2.245548433048433, + "grad_norm": 0.8639607429504395, + "learning_rate": 8.098200441150551e-05, + "loss": 0.8238, + "step": 12613 + }, + { + "epoch": 2.2457264957264957, + "grad_norm": 1.0107637643814087, + "learning_rate": 8.09682625900949e-05, + "loss": 0.8747, + "step": 12614 + }, + { + "epoch": 2.2459045584045585, + "grad_norm": 0.8153043985366821, + "learning_rate": 8.095452114157164e-05, + "loss": 0.9357, + "step": 12615 + }, + { + "epoch": 2.2460826210826212, + "grad_norm": 0.8948562741279602, + "learning_rate": 8.094078006620497e-05, + "loss": 0.8245, + "step": 12616 + }, + { + "epoch": 2.246260683760684, + "grad_norm": 0.7983259558677673, + "learning_rate": 8.092703936426416e-05, + "loss": 0.8936, + "step": 12617 + }, + { + "epoch": 2.2464387464387463, + "grad_norm": 0.9016979336738586, + "learning_rate": 8.091329903601835e-05, + "loss": 1.0685, + "step": 12618 + }, + { + "epoch": 2.246616809116809, + "grad_norm": 0.7192493677139282, + "learning_rate": 8.089955908173685e-05, + "loss": 0.8622, + "step": 12619 + }, + { + "epoch": 2.246794871794872, + "grad_norm": 0.78288334608078, + "learning_rate": 8.088581950168877e-05, + "loss": 0.7874, + "step": 12620 + }, + { + "epoch": 2.2469729344729346, + "grad_norm": 0.8438683152198792, + "learning_rate": 8.087208029614336e-05, + "loss": 0.9262, + "step": 12621 + }, + { + "epoch": 2.2471509971509973, + "grad_norm": 0.8384907245635986, + "learning_rate": 8.085834146536978e-05, + "loss": 0.9069, + "step": 12622 + }, + { + "epoch": 2.2473290598290596, + "grad_norm": 0.8209545016288757, + "learning_rate": 8.084460300963729e-05, + "loss": 0.9457, + "step": 12623 + }, + { + "epoch": 2.2475071225071224, + "grad_norm": 0.8220782279968262, + "learning_rate": 8.083086492921496e-05, + "loss": 0.9224, + "step": 12624 + }, + { + "epoch": 2.247685185185185, + "grad_norm": 0.8927256464958191, + "learning_rate": 8.081712722437204e-05, + "loss": 0.7091, + "step": 12625 + }, + { + "epoch": 2.247863247863248, + "grad_norm": 0.8878564238548279, + "learning_rate": 8.080338989537764e-05, + "loss": 0.8879, + "step": 12626 + }, + { + "epoch": 2.2480413105413106, + "grad_norm": 0.8380948305130005, + "learning_rate": 8.078965294250097e-05, + "loss": 0.8504, + "step": 12627 + }, + { + "epoch": 2.2482193732193734, + "grad_norm": 0.8005350828170776, + "learning_rate": 8.07759163660111e-05, + "loss": 1.2119, + "step": 12628 + }, + { + "epoch": 2.248397435897436, + "grad_norm": 0.7990152835845947, + "learning_rate": 8.076218016617726e-05, + "loss": 0.72, + "step": 12629 + }, + { + "epoch": 2.2485754985754984, + "grad_norm": 0.9264963269233704, + "learning_rate": 8.07484443432685e-05, + "loss": 0.8398, + "step": 12630 + }, + { + "epoch": 2.248753561253561, + "grad_norm": 0.9103235602378845, + "learning_rate": 8.073470889755402e-05, + "loss": 0.9122, + "step": 12631 + }, + { + "epoch": 2.248931623931624, + "grad_norm": 0.8042106032371521, + "learning_rate": 8.072097382930285e-05, + "loss": 0.8065, + "step": 12632 + }, + { + "epoch": 2.2491096866096867, + "grad_norm": 0.8464857935905457, + "learning_rate": 8.070723913878421e-05, + "loss": 0.9117, + "step": 12633 + }, + { + "epoch": 2.2492877492877494, + "grad_norm": 0.7476474642753601, + "learning_rate": 8.06935048262671e-05, + "loss": 1.0252, + "step": 12634 + }, + { + "epoch": 2.2494658119658117, + "grad_norm": 0.8098256587982178, + "learning_rate": 8.067977089202065e-05, + "loss": 0.911, + "step": 12635 + }, + { + "epoch": 2.2496438746438745, + "grad_norm": 0.9311509728431702, + "learning_rate": 8.066603733631398e-05, + "loss": 0.9594, + "step": 12636 + }, + { + "epoch": 2.2496438746438745, + "eval_loss": 1.1335573196411133, + "eval_runtime": 24.2688, + "eval_samples_per_second": 42.895, + "eval_steps_per_second": 21.468, + "step": 12636 + }, + { + "epoch": 2.2498219373219372, + "grad_norm": 0.7744980454444885, + "learning_rate": 8.065230415941612e-05, + "loss": 0.8983, + "step": 12637 + }, + { + "epoch": 2.25, + "grad_norm": 0.9464056491851807, + "learning_rate": 8.06385713615962e-05, + "loss": 0.8646, + "step": 12638 + }, + { + "epoch": 2.2501780626780628, + "grad_norm": 0.8263896107673645, + "learning_rate": 8.062483894312323e-05, + "loss": 0.8557, + "step": 12639 + }, + { + "epoch": 2.2503561253561255, + "grad_norm": 0.8827885389328003, + "learning_rate": 8.06111069042663e-05, + "loss": 0.7632, + "step": 12640 + }, + { + "epoch": 2.2505341880341883, + "grad_norm": 0.8537881374359131, + "learning_rate": 8.059737524529443e-05, + "loss": 0.8004, + "step": 12641 + }, + { + "epoch": 2.2507122507122506, + "grad_norm": 0.8397842049598694, + "learning_rate": 8.058364396647674e-05, + "loss": 0.9487, + "step": 12642 + }, + { + "epoch": 2.2508903133903133, + "grad_norm": 1.071976661682129, + "learning_rate": 8.056991306808217e-05, + "loss": 1.0699, + "step": 12643 + }, + { + "epoch": 2.251068376068376, + "grad_norm": 0.8712023496627808, + "learning_rate": 8.055618255037983e-05, + "loss": 0.6518, + "step": 12644 + }, + { + "epoch": 2.251246438746439, + "grad_norm": 0.7885438799858093, + "learning_rate": 8.054245241363866e-05, + "loss": 0.8458, + "step": 12645 + }, + { + "epoch": 2.2514245014245016, + "grad_norm": 0.947169840335846, + "learning_rate": 8.052872265812774e-05, + "loss": 0.6631, + "step": 12646 + }, + { + "epoch": 2.251602564102564, + "grad_norm": 0.8554182052612305, + "learning_rate": 8.051499328411603e-05, + "loss": 0.8945, + "step": 12647 + }, + { + "epoch": 2.2517806267806266, + "grad_norm": 0.8081278800964355, + "learning_rate": 8.050126429187259e-05, + "loss": 0.8969, + "step": 12648 + }, + { + "epoch": 2.2519586894586894, + "grad_norm": 0.7826179265975952, + "learning_rate": 8.048753568166633e-05, + "loss": 0.6965, + "step": 12649 + }, + { + "epoch": 2.252136752136752, + "grad_norm": 0.9688517451286316, + "learning_rate": 8.04738074537663e-05, + "loss": 1.0044, + "step": 12650 + }, + { + "epoch": 2.252314814814815, + "grad_norm": 0.7780970931053162, + "learning_rate": 8.04600796084414e-05, + "loss": 0.8712, + "step": 12651 + }, + { + "epoch": 2.2524928774928776, + "grad_norm": 0.8360016942024231, + "learning_rate": 8.044635214596073e-05, + "loss": 0.9522, + "step": 12652 + }, + { + "epoch": 2.2526709401709404, + "grad_norm": 0.8137710094451904, + "learning_rate": 8.043262506659311e-05, + "loss": 0.7953, + "step": 12653 + }, + { + "epoch": 2.2528490028490027, + "grad_norm": 0.8394312858581543, + "learning_rate": 8.041889837060755e-05, + "loss": 0.77, + "step": 12654 + }, + { + "epoch": 2.2530270655270654, + "grad_norm": 0.7245169878005981, + "learning_rate": 8.040517205827307e-05, + "loss": 0.7657, + "step": 12655 + }, + { + "epoch": 2.253205128205128, + "grad_norm": 0.8018792271614075, + "learning_rate": 8.039144612985846e-05, + "loss": 0.8974, + "step": 12656 + }, + { + "epoch": 2.253383190883191, + "grad_norm": 0.8204617500305176, + "learning_rate": 8.037772058563278e-05, + "loss": 0.8635, + "step": 12657 + }, + { + "epoch": 2.2535612535612537, + "grad_norm": 0.906288743019104, + "learning_rate": 8.036399542586485e-05, + "loss": 1.0498, + "step": 12658 + }, + { + "epoch": 2.253739316239316, + "grad_norm": 0.8674196600914001, + "learning_rate": 8.035027065082371e-05, + "loss": 0.8621, + "step": 12659 + }, + { + "epoch": 2.2539173789173788, + "grad_norm": 0.8112890124320984, + "learning_rate": 8.033654626077816e-05, + "loss": 0.9937, + "step": 12660 + }, + { + "epoch": 2.2540954415954415, + "grad_norm": 0.8072839975357056, + "learning_rate": 8.032282225599714e-05, + "loss": 0.8555, + "step": 12661 + }, + { + "epoch": 2.2542735042735043, + "grad_norm": 0.7853979468345642, + "learning_rate": 8.030909863674952e-05, + "loss": 0.8698, + "step": 12662 + }, + { + "epoch": 2.254451566951567, + "grad_norm": 0.7456761598587036, + "learning_rate": 8.029537540330426e-05, + "loss": 0.6214, + "step": 12663 + }, + { + "epoch": 2.2546296296296298, + "grad_norm": 0.7207663059234619, + "learning_rate": 8.028165255593015e-05, + "loss": 0.7641, + "step": 12664 + }, + { + "epoch": 2.2548076923076925, + "grad_norm": 0.6541373133659363, + "learning_rate": 8.02679300948961e-05, + "loss": 0.7438, + "step": 12665 + }, + { + "epoch": 2.254985754985755, + "grad_norm": 0.7535310983657837, + "learning_rate": 8.025420802047096e-05, + "loss": 0.9417, + "step": 12666 + }, + { + "epoch": 2.2551638176638176, + "grad_norm": 0.88471919298172, + "learning_rate": 8.024048633292364e-05, + "loss": 0.9122, + "step": 12667 + }, + { + "epoch": 2.2553418803418803, + "grad_norm": 0.8621570467948914, + "learning_rate": 8.02267650325229e-05, + "loss": 0.7026, + "step": 12668 + }, + { + "epoch": 2.255519943019943, + "grad_norm": 0.8574202060699463, + "learning_rate": 8.021304411953767e-05, + "loss": 0.8997, + "step": 12669 + }, + { + "epoch": 2.255698005698006, + "grad_norm": 0.8038806915283203, + "learning_rate": 8.019932359423667e-05, + "loss": 0.9386, + "step": 12670 + }, + { + "epoch": 2.255876068376068, + "grad_norm": 0.7760711908340454, + "learning_rate": 8.018560345688883e-05, + "loss": 0.7777, + "step": 12671 + }, + { + "epoch": 2.256054131054131, + "grad_norm": 0.7433146834373474, + "learning_rate": 8.017188370776292e-05, + "loss": 0.7245, + "step": 12672 + }, + { + "epoch": 2.2562321937321936, + "grad_norm": 0.8710882067680359, + "learning_rate": 8.01581643471278e-05, + "loss": 0.9478, + "step": 12673 + }, + { + "epoch": 2.2564102564102564, + "grad_norm": 0.7726423740386963, + "learning_rate": 8.014444537525218e-05, + "loss": 0.8388, + "step": 12674 + }, + { + "epoch": 2.256588319088319, + "grad_norm": 0.7967063188552856, + "learning_rate": 8.01307267924049e-05, + "loss": 0.8226, + "step": 12675 + }, + { + "epoch": 2.256766381766382, + "grad_norm": 0.7524598836898804, + "learning_rate": 8.011700859885479e-05, + "loss": 0.7285, + "step": 12676 + }, + { + "epoch": 2.2569444444444446, + "grad_norm": 0.808729887008667, + "learning_rate": 8.010329079487055e-05, + "loss": 0.7498, + "step": 12677 + }, + { + "epoch": 2.257122507122507, + "grad_norm": 0.7842788100242615, + "learning_rate": 8.008957338072106e-05, + "loss": 0.9216, + "step": 12678 + }, + { + "epoch": 2.2573005698005697, + "grad_norm": 0.8905709981918335, + "learning_rate": 8.007585635667497e-05, + "loss": 0.9254, + "step": 12679 + }, + { + "epoch": 2.2574786324786325, + "grad_norm": 0.7495295405387878, + "learning_rate": 8.006213972300112e-05, + "loss": 0.8407, + "step": 12680 + }, + { + "epoch": 2.257656695156695, + "grad_norm": 0.7425774335861206, + "learning_rate": 8.004842347996819e-05, + "loss": 0.7893, + "step": 12681 + }, + { + "epoch": 2.257834757834758, + "grad_norm": 0.8028583526611328, + "learning_rate": 8.003470762784498e-05, + "loss": 0.8106, + "step": 12682 + }, + { + "epoch": 2.2580128205128207, + "grad_norm": 0.8874917030334473, + "learning_rate": 8.002099216690017e-05, + "loss": 0.97, + "step": 12683 + }, + { + "epoch": 2.258190883190883, + "grad_norm": 0.8830558061599731, + "learning_rate": 8.000727709740257e-05, + "loss": 1.028, + "step": 12684 + }, + { + "epoch": 2.2583689458689458, + "grad_norm": 0.8720272779464722, + "learning_rate": 7.99935624196208e-05, + "loss": 0.9401, + "step": 12685 + }, + { + "epoch": 2.2585470085470085, + "grad_norm": 0.736709713935852, + "learning_rate": 7.997984813382362e-05, + "loss": 0.8479, + "step": 12686 + }, + { + "epoch": 2.2587250712250713, + "grad_norm": 0.8028469085693359, + "learning_rate": 7.996613424027973e-05, + "loss": 0.9291, + "step": 12687 + }, + { + "epoch": 2.258903133903134, + "grad_norm": 0.777618944644928, + "learning_rate": 7.995242073925784e-05, + "loss": 0.7021, + "step": 12688 + }, + { + "epoch": 2.2590811965811968, + "grad_norm": 0.8371155261993408, + "learning_rate": 7.993870763102659e-05, + "loss": 0.8309, + "step": 12689 + }, + { + "epoch": 2.259259259259259, + "grad_norm": 0.8853654861450195, + "learning_rate": 7.992499491585473e-05, + "loss": 0.762, + "step": 12690 + }, + { + "epoch": 2.259437321937322, + "grad_norm": 0.742594301700592, + "learning_rate": 7.991128259401086e-05, + "loss": 0.8025, + "step": 12691 + }, + { + "epoch": 2.2596153846153846, + "grad_norm": 1.0678842067718506, + "learning_rate": 7.989757066576369e-05, + "loss": 0.9127, + "step": 12692 + }, + { + "epoch": 2.2597934472934473, + "grad_norm": 0.7917565703392029, + "learning_rate": 7.988385913138183e-05, + "loss": 0.8078, + "step": 12693 + }, + { + "epoch": 2.25997150997151, + "grad_norm": 0.6907288432121277, + "learning_rate": 7.987014799113397e-05, + "loss": 0.6313, + "step": 12694 + }, + { + "epoch": 2.260149572649573, + "grad_norm": 0.9007455706596375, + "learning_rate": 7.98564372452888e-05, + "loss": 1.0734, + "step": 12695 + }, + { + "epoch": 2.260327635327635, + "grad_norm": 0.7732774615287781, + "learning_rate": 7.984272689411484e-05, + "loss": 0.9925, + "step": 12696 + }, + { + "epoch": 2.260505698005698, + "grad_norm": 0.7470823526382446, + "learning_rate": 7.982901693788082e-05, + "loss": 0.8518, + "step": 12697 + }, + { + "epoch": 2.2606837606837606, + "grad_norm": 0.8018864989280701, + "learning_rate": 7.981530737685526e-05, + "loss": 0.8668, + "step": 12698 + }, + { + "epoch": 2.2608618233618234, + "grad_norm": 0.8459745049476624, + "learning_rate": 7.980159821130688e-05, + "loss": 0.8972, + "step": 12699 + }, + { + "epoch": 2.261039886039886, + "grad_norm": 0.8255595564842224, + "learning_rate": 7.978788944150419e-05, + "loss": 0.9562, + "step": 12700 + }, + { + "epoch": 2.261217948717949, + "grad_norm": 0.8243128061294556, + "learning_rate": 7.977418106771582e-05, + "loss": 0.6634, + "step": 12701 + }, + { + "epoch": 2.261396011396011, + "grad_norm": 0.802949845790863, + "learning_rate": 7.976047309021034e-05, + "loss": 0.8155, + "step": 12702 + }, + { + "epoch": 2.261574074074074, + "grad_norm": 0.8480857014656067, + "learning_rate": 7.97467655092564e-05, + "loss": 0.8568, + "step": 12703 + }, + { + "epoch": 2.2617521367521367, + "grad_norm": 0.8777545690536499, + "learning_rate": 7.973305832512247e-05, + "loss": 0.8688, + "step": 12704 + }, + { + "epoch": 2.2619301994301995, + "grad_norm": 0.8334060907363892, + "learning_rate": 7.971935153807719e-05, + "loss": 0.932, + "step": 12705 + }, + { + "epoch": 2.262108262108262, + "grad_norm": 0.836976170539856, + "learning_rate": 7.970564514838907e-05, + "loss": 0.8205, + "step": 12706 + }, + { + "epoch": 2.262286324786325, + "grad_norm": 0.782866895198822, + "learning_rate": 7.969193915632667e-05, + "loss": 0.8362, + "step": 12707 + }, + { + "epoch": 2.2624643874643873, + "grad_norm": 0.9018504619598389, + "learning_rate": 7.967823356215854e-05, + "loss": 0.8354, + "step": 12708 + }, + { + "epoch": 2.26264245014245, + "grad_norm": 0.7974916696548462, + "learning_rate": 7.966452836615324e-05, + "loss": 0.8035, + "step": 12709 + }, + { + "epoch": 2.2628205128205128, + "grad_norm": 0.8745712637901306, + "learning_rate": 7.965082356857922e-05, + "loss": 0.8803, + "step": 12710 + }, + { + "epoch": 2.2629985754985755, + "grad_norm": 0.8667176365852356, + "learning_rate": 7.963711916970505e-05, + "loss": 0.8005, + "step": 12711 + }, + { + "epoch": 2.2631766381766383, + "grad_norm": 0.849998950958252, + "learning_rate": 7.962341516979922e-05, + "loss": 0.8208, + "step": 12712 + }, + { + "epoch": 2.263354700854701, + "grad_norm": 0.803727388381958, + "learning_rate": 7.960971156913028e-05, + "loss": 0.7232, + "step": 12713 + }, + { + "epoch": 2.263532763532764, + "grad_norm": 0.842913031578064, + "learning_rate": 7.959600836796664e-05, + "loss": 0.8182, + "step": 12714 + }, + { + "epoch": 2.263710826210826, + "grad_norm": 0.8191903829574585, + "learning_rate": 7.958230556657684e-05, + "loss": 0.8353, + "step": 12715 + }, + { + "epoch": 2.263888888888889, + "grad_norm": 0.8525017499923706, + "learning_rate": 7.95686031652294e-05, + "loss": 0.9824, + "step": 12716 + }, + { + "epoch": 2.2640669515669516, + "grad_norm": 0.7176641225814819, + "learning_rate": 7.955490116419267e-05, + "loss": 0.7722, + "step": 12717 + }, + { + "epoch": 2.2642450142450143, + "grad_norm": 0.8740555047988892, + "learning_rate": 7.954119956373521e-05, + "loss": 0.8286, + "step": 12718 + }, + { + "epoch": 2.264423076923077, + "grad_norm": 0.7928949594497681, + "learning_rate": 7.952749836412543e-05, + "loss": 0.9183, + "step": 12719 + }, + { + "epoch": 2.2646011396011394, + "grad_norm": 0.787661612033844, + "learning_rate": 7.951379756563185e-05, + "loss": 0.7741, + "step": 12720 + }, + { + "epoch": 2.264779202279202, + "grad_norm": 0.8369856476783752, + "learning_rate": 7.950009716852277e-05, + "loss": 0.911, + "step": 12721 + }, + { + "epoch": 2.264957264957265, + "grad_norm": 0.7838568687438965, + "learning_rate": 7.948639717306675e-05, + "loss": 0.8532, + "step": 12722 + }, + { + "epoch": 2.2651353276353277, + "grad_norm": 0.8287179470062256, + "learning_rate": 7.947269757953213e-05, + "loss": 0.893, + "step": 12723 + }, + { + "epoch": 2.2653133903133904, + "grad_norm": 0.7754728198051453, + "learning_rate": 7.945899838818741e-05, + "loss": 0.9516, + "step": 12724 + }, + { + "epoch": 2.265491452991453, + "grad_norm": 0.7088906764984131, + "learning_rate": 7.94452995993009e-05, + "loss": 0.6797, + "step": 12725 + }, + { + "epoch": 2.265669515669516, + "grad_norm": 0.8004380464553833, + "learning_rate": 7.94316012131411e-05, + "loss": 0.8583, + "step": 12726 + }, + { + "epoch": 2.265847578347578, + "grad_norm": 0.8221408128738403, + "learning_rate": 7.941790322997629e-05, + "loss": 0.9575, + "step": 12727 + }, + { + "epoch": 2.266025641025641, + "grad_norm": 0.8640061020851135, + "learning_rate": 7.940420565007492e-05, + "loss": 0.9471, + "step": 12728 + }, + { + "epoch": 2.2662037037037037, + "grad_norm": 0.8151915669441223, + "learning_rate": 7.939050847370536e-05, + "loss": 0.7841, + "step": 12729 + }, + { + "epoch": 2.2663817663817665, + "grad_norm": 0.7910612225532532, + "learning_rate": 7.9376811701136e-05, + "loss": 0.8826, + "step": 12730 + }, + { + "epoch": 2.2665598290598292, + "grad_norm": 0.7158875465393066, + "learning_rate": 7.936311533263514e-05, + "loss": 0.8598, + "step": 12731 + }, + { + "epoch": 2.2667378917378915, + "grad_norm": 0.6968050003051758, + "learning_rate": 7.934941936847119e-05, + "loss": 0.742, + "step": 12732 + }, + { + "epoch": 2.2669159544159543, + "grad_norm": 0.8630516529083252, + "learning_rate": 7.933572380891245e-05, + "loss": 0.789, + "step": 12733 + }, + { + "epoch": 2.267094017094017, + "grad_norm": 0.8060851097106934, + "learning_rate": 7.932202865422726e-05, + "loss": 0.8447, + "step": 12734 + }, + { + "epoch": 2.26727207977208, + "grad_norm": 0.9570813775062561, + "learning_rate": 7.930833390468402e-05, + "loss": 0.8948, + "step": 12735 + }, + { + "epoch": 2.2674501424501425, + "grad_norm": 0.7649935483932495, + "learning_rate": 7.929463956055094e-05, + "loss": 0.905, + "step": 12736 + }, + { + "epoch": 2.2676282051282053, + "grad_norm": 0.7498226165771484, + "learning_rate": 7.928094562209641e-05, + "loss": 0.9168, + "step": 12737 + }, + { + "epoch": 2.267806267806268, + "grad_norm": 0.7915979027748108, + "learning_rate": 7.926725208958869e-05, + "loss": 0.8628, + "step": 12738 + }, + { + "epoch": 2.2679843304843303, + "grad_norm": 0.7620252966880798, + "learning_rate": 7.925355896329615e-05, + "loss": 0.8768, + "step": 12739 + }, + { + "epoch": 2.268162393162393, + "grad_norm": 0.9785344004631042, + "learning_rate": 7.923986624348697e-05, + "loss": 0.9579, + "step": 12740 + }, + { + "epoch": 2.268340455840456, + "grad_norm": 0.9146337509155273, + "learning_rate": 7.922617393042954e-05, + "loss": 1.2241, + "step": 12741 + }, + { + "epoch": 2.2685185185185186, + "grad_norm": 0.7815660238265991, + "learning_rate": 7.921248202439203e-05, + "loss": 0.7367, + "step": 12742 + }, + { + "epoch": 2.2686965811965814, + "grad_norm": 0.8466008305549622, + "learning_rate": 7.919879052564276e-05, + "loss": 0.923, + "step": 12743 + }, + { + "epoch": 2.2688746438746437, + "grad_norm": 0.742203950881958, + "learning_rate": 7.918509943444998e-05, + "loss": 0.8282, + "step": 12744 + }, + { + "epoch": 2.2690527065527064, + "grad_norm": 0.785446286201477, + "learning_rate": 7.917140875108196e-05, + "loss": 0.7689, + "step": 12745 + }, + { + "epoch": 2.269230769230769, + "grad_norm": 0.912765383720398, + "learning_rate": 7.915771847580689e-05, + "loss": 0.8259, + "step": 12746 + }, + { + "epoch": 2.269408831908832, + "grad_norm": 0.7319221496582031, + "learning_rate": 7.914402860889305e-05, + "loss": 0.8861, + "step": 12747 + }, + { + "epoch": 2.2695868945868947, + "grad_norm": 1.0215193033218384, + "learning_rate": 7.913033915060861e-05, + "loss": 0.8556, + "step": 12748 + }, + { + "epoch": 2.2697649572649574, + "grad_norm": 0.9348630905151367, + "learning_rate": 7.911665010122188e-05, + "loss": 1.0269, + "step": 12749 + }, + { + "epoch": 2.26994301994302, + "grad_norm": 0.7521753311157227, + "learning_rate": 7.910296146100096e-05, + "loss": 0.8492, + "step": 12750 + }, + { + "epoch": 2.2701210826210825, + "grad_norm": 0.7274978756904602, + "learning_rate": 7.908927323021414e-05, + "loss": 0.7029, + "step": 12751 + }, + { + "epoch": 2.2702991452991452, + "grad_norm": 0.8103266954421997, + "learning_rate": 7.907558540912954e-05, + "loss": 0.5268, + "step": 12752 + }, + { + "epoch": 2.270477207977208, + "grad_norm": 0.8645551800727844, + "learning_rate": 7.906189799801538e-05, + "loss": 0.8172, + "step": 12753 + }, + { + "epoch": 2.2706552706552707, + "grad_norm": 0.8652981519699097, + "learning_rate": 7.904821099713984e-05, + "loss": 0.8711, + "step": 12754 + }, + { + "epoch": 2.2708333333333335, + "grad_norm": 0.7020241618156433, + "learning_rate": 7.903452440677106e-05, + "loss": 0.7202, + "step": 12755 + }, + { + "epoch": 2.271011396011396, + "grad_norm": 0.9812583923339844, + "learning_rate": 7.902083822717727e-05, + "loss": 0.8274, + "step": 12756 + }, + { + "epoch": 2.2711894586894585, + "grad_norm": 0.9119269847869873, + "learning_rate": 7.900715245862655e-05, + "loss": 0.8695, + "step": 12757 + }, + { + "epoch": 2.2713675213675213, + "grad_norm": 0.7336047291755676, + "learning_rate": 7.899346710138706e-05, + "loss": 0.8138, + "step": 12758 + }, + { + "epoch": 2.271545584045584, + "grad_norm": 0.900337278842926, + "learning_rate": 7.897978215572695e-05, + "loss": 0.8346, + "step": 12759 + }, + { + "epoch": 2.271723646723647, + "grad_norm": 0.736018717288971, + "learning_rate": 7.896609762191437e-05, + "loss": 0.7045, + "step": 12760 + }, + { + "epoch": 2.2719017094017095, + "grad_norm": 0.8484935760498047, + "learning_rate": 7.895241350021737e-05, + "loss": 1.05, + "step": 12761 + }, + { + "epoch": 2.2720797720797723, + "grad_norm": 0.8032337427139282, + "learning_rate": 7.893872979090415e-05, + "loss": 0.8024, + "step": 12762 + }, + { + "epoch": 2.2722578347578346, + "grad_norm": 0.8957629203796387, + "learning_rate": 7.892504649424272e-05, + "loss": 0.9593, + "step": 12763 + }, + { + "epoch": 2.2724358974358974, + "grad_norm": 0.9227191805839539, + "learning_rate": 7.891136361050126e-05, + "loss": 0.9978, + "step": 12764 + }, + { + "epoch": 2.27261396011396, + "grad_norm": 0.8649391531944275, + "learning_rate": 7.88976811399478e-05, + "loss": 0.8525, + "step": 12765 + }, + { + "epoch": 2.272792022792023, + "grad_norm": 0.8762859106063843, + "learning_rate": 7.888399908285046e-05, + "loss": 0.9526, + "step": 12766 + }, + { + "epoch": 2.2729700854700856, + "grad_norm": 0.8566350340843201, + "learning_rate": 7.887031743947729e-05, + "loss": 0.7886, + "step": 12767 + }, + { + "epoch": 2.273148148148148, + "grad_norm": 0.9285386800765991, + "learning_rate": 7.885663621009634e-05, + "loss": 1.013, + "step": 12768 + }, + { + "epoch": 2.2733262108262107, + "grad_norm": 0.9326284527778625, + "learning_rate": 7.884295539497569e-05, + "loss": 0.9908, + "step": 12769 + }, + { + "epoch": 2.2735042735042734, + "grad_norm": 0.8035810589790344, + "learning_rate": 7.882927499438341e-05, + "loss": 0.7452, + "step": 12770 + }, + { + "epoch": 2.273682336182336, + "grad_norm": 0.831741988658905, + "learning_rate": 7.881559500858747e-05, + "loss": 0.8782, + "step": 12771 + }, + { + "epoch": 2.273860398860399, + "grad_norm": 0.7790034413337708, + "learning_rate": 7.880191543785594e-05, + "loss": 0.9494, + "step": 12772 + }, + { + "epoch": 2.2740384615384617, + "grad_norm": 0.7070899605751038, + "learning_rate": 7.878823628245684e-05, + "loss": 0.7007, + "step": 12773 + }, + { + "epoch": 2.2742165242165244, + "grad_norm": 0.739573061466217, + "learning_rate": 7.877455754265818e-05, + "loss": 0.758, + "step": 12774 + }, + { + "epoch": 2.2743945868945867, + "grad_norm": 1.091391921043396, + "learning_rate": 7.876087921872803e-05, + "loss": 1.1333, + "step": 12775 + }, + { + "epoch": 2.2745726495726495, + "grad_norm": 0.623710036277771, + "learning_rate": 7.874720131093427e-05, + "loss": 0.7068, + "step": 12776 + }, + { + "epoch": 2.2747507122507122, + "grad_norm": 0.7989393472671509, + "learning_rate": 7.8733523819545e-05, + "loss": 0.77, + "step": 12777 + }, + { + "epoch": 2.274928774928775, + "grad_norm": 0.8401352167129517, + "learning_rate": 7.87198467448281e-05, + "loss": 0.8192, + "step": 12778 + }, + { + "epoch": 2.2751068376068377, + "grad_norm": 0.7962843179702759, + "learning_rate": 7.870617008705164e-05, + "loss": 0.8071, + "step": 12779 + }, + { + "epoch": 2.2752849002849, + "grad_norm": 0.9518889784812927, + "learning_rate": 7.869249384648351e-05, + "loss": 0.8956, + "step": 12780 + }, + { + "epoch": 2.275462962962963, + "grad_norm": 0.7469878792762756, + "learning_rate": 7.867881802339175e-05, + "loss": 0.6816, + "step": 12781 + }, + { + "epoch": 2.2756410256410255, + "grad_norm": 0.8888431191444397, + "learning_rate": 7.866514261804421e-05, + "loss": 0.7906, + "step": 12782 + }, + { + "epoch": 2.2758190883190883, + "grad_norm": 0.9856036305427551, + "learning_rate": 7.86514676307089e-05, + "loss": 0.9371, + "step": 12783 + }, + { + "epoch": 2.275997150997151, + "grad_norm": 0.9144912958145142, + "learning_rate": 7.863779306165371e-05, + "loss": 0.9613, + "step": 12784 + }, + { + "epoch": 2.276175213675214, + "grad_norm": 0.7898108959197998, + "learning_rate": 7.862411891114665e-05, + "loss": 0.8631, + "step": 12785 + }, + { + "epoch": 2.2763532763532766, + "grad_norm": 0.8524056077003479, + "learning_rate": 7.861044517945552e-05, + "loss": 0.7011, + "step": 12786 + }, + { + "epoch": 2.276531339031339, + "grad_norm": 0.8811307549476624, + "learning_rate": 7.859677186684831e-05, + "loss": 0.8138, + "step": 12787 + }, + { + "epoch": 2.2767094017094016, + "grad_norm": 0.9247646331787109, + "learning_rate": 7.858309897359289e-05, + "loss": 0.8971, + "step": 12788 + }, + { + "epoch": 2.2768874643874644, + "grad_norm": 0.8655884861946106, + "learning_rate": 7.856942649995715e-05, + "loss": 0.885, + "step": 12789 + }, + { + "epoch": 2.277065527065527, + "grad_norm": 0.9330910444259644, + "learning_rate": 7.855575444620897e-05, + "loss": 0.8493, + "step": 12790 + }, + { + "epoch": 2.27724358974359, + "grad_norm": 0.746694028377533, + "learning_rate": 7.854208281261626e-05, + "loss": 0.7169, + "step": 12791 + }, + { + "epoch": 2.277421652421652, + "grad_norm": 0.9785143136978149, + "learning_rate": 7.852841159944685e-05, + "loss": 1.023, + "step": 12792 + }, + { + "epoch": 2.277599715099715, + "grad_norm": 0.6107021570205688, + "learning_rate": 7.851474080696859e-05, + "loss": 0.652, + "step": 12793 + }, + { + "epoch": 2.2777777777777777, + "grad_norm": 0.9269224405288696, + "learning_rate": 7.850107043544937e-05, + "loss": 0.8024, + "step": 12794 + }, + { + "epoch": 2.2779558404558404, + "grad_norm": 0.8488328456878662, + "learning_rate": 7.8487400485157e-05, + "loss": 0.8626, + "step": 12795 + }, + { + "epoch": 2.278133903133903, + "grad_norm": 0.7187852263450623, + "learning_rate": 7.847373095635937e-05, + "loss": 0.7416, + "step": 12796 + }, + { + "epoch": 2.278311965811966, + "grad_norm": 1.00519859790802, + "learning_rate": 7.846006184932422e-05, + "loss": 1.0577, + "step": 12797 + }, + { + "epoch": 2.2784900284900287, + "grad_norm": 0.8175216913223267, + "learning_rate": 7.844639316431945e-05, + "loss": 0.9685, + "step": 12798 + }, + { + "epoch": 2.278668091168091, + "grad_norm": 0.8239067792892456, + "learning_rate": 7.843272490161281e-05, + "loss": 0.8714, + "step": 12799 + }, + { + "epoch": 2.2788461538461537, + "grad_norm": 0.8089447617530823, + "learning_rate": 7.841905706147212e-05, + "loss": 0.8397, + "step": 12800 + }, + { + "epoch": 2.2790242165242165, + "grad_norm": 0.8505867719650269, + "learning_rate": 7.840538964416518e-05, + "loss": 0.6872, + "step": 12801 + }, + { + "epoch": 2.2792022792022792, + "grad_norm": 0.8512473702430725, + "learning_rate": 7.83917226499598e-05, + "loss": 1.0422, + "step": 12802 + }, + { + "epoch": 2.279380341880342, + "grad_norm": 0.8935198187828064, + "learning_rate": 7.837805607912369e-05, + "loss": 0.9874, + "step": 12803 + }, + { + "epoch": 2.2795584045584047, + "grad_norm": 0.6903477907180786, + "learning_rate": 7.836438993192466e-05, + "loss": 0.7301, + "step": 12804 + }, + { + "epoch": 2.279736467236467, + "grad_norm": 0.7140037417411804, + "learning_rate": 7.835072420863046e-05, + "loss": 0.8323, + "step": 12805 + }, + { + "epoch": 2.27991452991453, + "grad_norm": 0.7974498867988586, + "learning_rate": 7.833705890950888e-05, + "loss": 0.7784, + "step": 12806 + }, + { + "epoch": 2.2800925925925926, + "grad_norm": 0.8191199898719788, + "learning_rate": 7.83233940348276e-05, + "loss": 0.8584, + "step": 12807 + }, + { + "epoch": 2.2802706552706553, + "grad_norm": 0.843112587928772, + "learning_rate": 7.83097295848544e-05, + "loss": 1.043, + "step": 12808 + }, + { + "epoch": 2.280448717948718, + "grad_norm": 0.8029288053512573, + "learning_rate": 7.829606555985698e-05, + "loss": 0.6806, + "step": 12809 + }, + { + "epoch": 2.280626780626781, + "grad_norm": 0.712228536605835, + "learning_rate": 7.828240196010311e-05, + "loss": 0.7748, + "step": 12810 + }, + { + "epoch": 2.280804843304843, + "grad_norm": 0.801659882068634, + "learning_rate": 7.82687387858604e-05, + "loss": 0.9374, + "step": 12811 + }, + { + "epoch": 2.280982905982906, + "grad_norm": 0.8457205891609192, + "learning_rate": 7.825507603739666e-05, + "loss": 0.9453, + "step": 12812 + }, + { + "epoch": 2.2811609686609686, + "grad_norm": 0.9129060506820679, + "learning_rate": 7.824141371497948e-05, + "loss": 0.9324, + "step": 12813 + }, + { + "epoch": 2.2813390313390314, + "grad_norm": 0.947914183139801, + "learning_rate": 7.822775181887663e-05, + "loss": 0.8275, + "step": 12814 + }, + { + "epoch": 2.281517094017094, + "grad_norm": 0.7204358577728271, + "learning_rate": 7.821409034935576e-05, + "loss": 0.6538, + "step": 12815 + }, + { + "epoch": 2.281695156695157, + "grad_norm": 0.8021003603935242, + "learning_rate": 7.82004293066845e-05, + "loss": 1.0464, + "step": 12816 + }, + { + "epoch": 2.281873219373219, + "grad_norm": 0.9530314803123474, + "learning_rate": 7.818676869113059e-05, + "loss": 0.8854, + "step": 12817 + }, + { + "epoch": 2.282051282051282, + "grad_norm": 0.7932098507881165, + "learning_rate": 7.81731085029616e-05, + "loss": 0.8219, + "step": 12818 + }, + { + "epoch": 2.2822293447293447, + "grad_norm": 0.7955925464630127, + "learning_rate": 7.815944874244523e-05, + "loss": 0.801, + "step": 12819 + }, + { + "epoch": 2.2824074074074074, + "grad_norm": 0.8490158915519714, + "learning_rate": 7.814578940984907e-05, + "loss": 0.8666, + "step": 12820 + }, + { + "epoch": 2.28258547008547, + "grad_norm": 0.7325232028961182, + "learning_rate": 7.813213050544081e-05, + "loss": 0.9579, + "step": 12821 + }, + { + "epoch": 2.282763532763533, + "grad_norm": 0.9203488230705261, + "learning_rate": 7.811847202948798e-05, + "loss": 1.0581, + "step": 12822 + }, + { + "epoch": 2.2829415954415953, + "grad_norm": 0.8207429647445679, + "learning_rate": 7.810481398225827e-05, + "loss": 0.8613, + "step": 12823 + }, + { + "epoch": 2.283119658119658, + "grad_norm": 0.872207522392273, + "learning_rate": 7.809115636401921e-05, + "loss": 0.9155, + "step": 12824 + }, + { + "epoch": 2.2832977207977208, + "grad_norm": 0.8032099604606628, + "learning_rate": 7.807749917503845e-05, + "loss": 0.8294, + "step": 12825 + }, + { + "epoch": 2.2834757834757835, + "grad_norm": 0.8824980854988098, + "learning_rate": 7.806384241558354e-05, + "loss": 0.8618, + "step": 12826 + }, + { + "epoch": 2.2836538461538463, + "grad_norm": 0.9057566523551941, + "learning_rate": 7.805018608592212e-05, + "loss": 0.826, + "step": 12827 + }, + { + "epoch": 2.283831908831909, + "grad_norm": 0.8092000484466553, + "learning_rate": 7.803653018632164e-05, + "loss": 0.8091, + "step": 12828 + }, + { + "epoch": 2.2840099715099713, + "grad_norm": 0.8372754454612732, + "learning_rate": 7.802287471704976e-05, + "loss": 1.108, + "step": 12829 + }, + { + "epoch": 2.284188034188034, + "grad_norm": 0.8702181577682495, + "learning_rate": 7.800921967837398e-05, + "loss": 0.9654, + "step": 12830 + }, + { + "epoch": 2.284366096866097, + "grad_norm": 0.9543859958648682, + "learning_rate": 7.79955650705619e-05, + "loss": 0.9268, + "step": 12831 + }, + { + "epoch": 2.2845441595441596, + "grad_norm": 0.7992038726806641, + "learning_rate": 7.798191089388096e-05, + "loss": 0.8299, + "step": 12832 + }, + { + "epoch": 2.2847222222222223, + "grad_norm": 0.8655165433883667, + "learning_rate": 7.796825714859874e-05, + "loss": 0.9656, + "step": 12833 + }, + { + "epoch": 2.284900284900285, + "grad_norm": 0.9013311862945557, + "learning_rate": 7.795460383498281e-05, + "loss": 0.9373, + "step": 12834 + }, + { + "epoch": 2.285078347578348, + "grad_norm": 0.8453806638717651, + "learning_rate": 7.794095095330058e-05, + "loss": 0.7711, + "step": 12835 + }, + { + "epoch": 2.28525641025641, + "grad_norm": 0.8016965985298157, + "learning_rate": 7.792729850381959e-05, + "loss": 0.7492, + "step": 12836 + }, + { + "epoch": 2.285434472934473, + "grad_norm": 0.7191343307495117, + "learning_rate": 7.791364648680734e-05, + "loss": 0.7541, + "step": 12837 + }, + { + "epoch": 2.2856125356125356, + "grad_norm": 0.8210958242416382, + "learning_rate": 7.789999490253133e-05, + "loss": 0.7448, + "step": 12838 + }, + { + "epoch": 2.2857905982905984, + "grad_norm": 0.904022216796875, + "learning_rate": 7.788634375125898e-05, + "loss": 1.0329, + "step": 12839 + }, + { + "epoch": 2.285968660968661, + "grad_norm": 0.8934714794158936, + "learning_rate": 7.787269303325779e-05, + "loss": 0.8982, + "step": 12840 + }, + { + "epoch": 2.2861467236467234, + "grad_norm": 0.9424307942390442, + "learning_rate": 7.785904274879521e-05, + "loss": 1.0298, + "step": 12841 + }, + { + "epoch": 2.286324786324786, + "grad_norm": 0.7753969430923462, + "learning_rate": 7.784539289813873e-05, + "loss": 0.7811, + "step": 12842 + }, + { + "epoch": 2.286502849002849, + "grad_norm": 0.7851901054382324, + "learning_rate": 7.78317434815557e-05, + "loss": 0.8395, + "step": 12843 + }, + { + "epoch": 2.2866809116809117, + "grad_norm": 0.7734000086784363, + "learning_rate": 7.781809449931365e-05, + "loss": 0.6572, + "step": 12844 + }, + { + "epoch": 2.2868589743589745, + "grad_norm": 0.8322952389717102, + "learning_rate": 7.780444595167992e-05, + "loss": 0.9707, + "step": 12845 + }, + { + "epoch": 2.287037037037037, + "grad_norm": 0.8243176341056824, + "learning_rate": 7.779079783892203e-05, + "loss": 0.8413, + "step": 12846 + }, + { + "epoch": 2.2872150997151, + "grad_norm": 0.8600375056266785, + "learning_rate": 7.777715016130727e-05, + "loss": 0.8471, + "step": 12847 + }, + { + "epoch": 2.2873931623931623, + "grad_norm": 0.9846388101577759, + "learning_rate": 7.776350291910311e-05, + "loss": 1.0187, + "step": 12848 + }, + { + "epoch": 2.287571225071225, + "grad_norm": 0.8445034623146057, + "learning_rate": 7.774985611257688e-05, + "loss": 0.9113, + "step": 12849 + }, + { + "epoch": 2.2877492877492878, + "grad_norm": 0.804595947265625, + "learning_rate": 7.773620974199604e-05, + "loss": 0.8331, + "step": 12850 + }, + { + "epoch": 2.2879273504273505, + "grad_norm": 0.7600802779197693, + "learning_rate": 7.772256380762789e-05, + "loss": 0.8448, + "step": 12851 + }, + { + "epoch": 2.2881054131054133, + "grad_norm": 0.7406377792358398, + "learning_rate": 7.770891830973984e-05, + "loss": 0.7904, + "step": 12852 + }, + { + "epoch": 2.2882834757834756, + "grad_norm": 0.7294487357139587, + "learning_rate": 7.769527324859924e-05, + "loss": 0.8799, + "step": 12853 + }, + { + "epoch": 2.2884615384615383, + "grad_norm": 0.8864750266075134, + "learning_rate": 7.768162862447342e-05, + "loss": 0.9038, + "step": 12854 + }, + { + "epoch": 2.288639601139601, + "grad_norm": 0.8933553099632263, + "learning_rate": 7.766798443762972e-05, + "loss": 0.929, + "step": 12855 + }, + { + "epoch": 2.288817663817664, + "grad_norm": 0.8065192103385925, + "learning_rate": 7.765434068833545e-05, + "loss": 0.9335, + "step": 12856 + }, + { + "epoch": 2.2889957264957266, + "grad_norm": 0.8644578456878662, + "learning_rate": 7.764069737685802e-05, + "loss": 0.7717, + "step": 12857 + }, + { + "epoch": 2.2891737891737893, + "grad_norm": 0.8957899212837219, + "learning_rate": 7.762705450346462e-05, + "loss": 0.8625, + "step": 12858 + }, + { + "epoch": 2.289351851851852, + "grad_norm": 0.7164827585220337, + "learning_rate": 7.761341206842265e-05, + "loss": 0.8018, + "step": 12859 + }, + { + "epoch": 2.2895299145299144, + "grad_norm": 0.8752971291542053, + "learning_rate": 7.759977007199933e-05, + "loss": 0.8517, + "step": 12860 + }, + { + "epoch": 2.289707977207977, + "grad_norm": 0.8448139429092407, + "learning_rate": 7.758612851446201e-05, + "loss": 1.0449, + "step": 12861 + }, + { + "epoch": 2.28988603988604, + "grad_norm": 0.81675785779953, + "learning_rate": 7.75724873960779e-05, + "loss": 1.0952, + "step": 12862 + }, + { + "epoch": 2.2900641025641026, + "grad_norm": 0.8215656876564026, + "learning_rate": 7.755884671711437e-05, + "loss": 0.8419, + "step": 12863 + }, + { + "epoch": 2.2902421652421654, + "grad_norm": 0.8270167708396912, + "learning_rate": 7.754520647783859e-05, + "loss": 0.9065, + "step": 12864 + }, + { + "epoch": 2.2904202279202277, + "grad_norm": 0.8222723603248596, + "learning_rate": 7.753156667851784e-05, + "loss": 0.8536, + "step": 12865 + }, + { + "epoch": 2.2905982905982905, + "grad_norm": 0.8383764028549194, + "learning_rate": 7.751792731941936e-05, + "loss": 0.8829, + "step": 12866 + }, + { + "epoch": 2.290776353276353, + "grad_norm": 0.8115772008895874, + "learning_rate": 7.750428840081043e-05, + "loss": 0.8969, + "step": 12867 + }, + { + "epoch": 2.290954415954416, + "grad_norm": 0.8721897602081299, + "learning_rate": 7.74906499229582e-05, + "loss": 1.031, + "step": 12868 + }, + { + "epoch": 2.2911324786324787, + "grad_norm": 0.6958467364311218, + "learning_rate": 7.747701188612996e-05, + "loss": 0.7528, + "step": 12869 + }, + { + "epoch": 2.2913105413105415, + "grad_norm": 0.8352338671684265, + "learning_rate": 7.746337429059285e-05, + "loss": 0.9297, + "step": 12870 + }, + { + "epoch": 2.291488603988604, + "grad_norm": 0.8407408595085144, + "learning_rate": 7.744973713661411e-05, + "loss": 0.8209, + "step": 12871 + }, + { + "epoch": 2.2916666666666665, + "grad_norm": 0.9509777426719666, + "learning_rate": 7.743610042446092e-05, + "loss": 0.9408, + "step": 12872 + }, + { + "epoch": 2.2918447293447293, + "grad_norm": 0.7913112640380859, + "learning_rate": 7.742246415440048e-05, + "loss": 0.9063, + "step": 12873 + }, + { + "epoch": 2.292022792022792, + "grad_norm": 0.90866619348526, + "learning_rate": 7.740882832669998e-05, + "loss": 1.0178, + "step": 12874 + }, + { + "epoch": 2.2922008547008548, + "grad_norm": 0.5832980871200562, + "learning_rate": 7.739519294162652e-05, + "loss": 0.4138, + "step": 12875 + }, + { + "epoch": 2.2923789173789175, + "grad_norm": 0.717993974685669, + "learning_rate": 7.738155799944732e-05, + "loss": 0.7303, + "step": 12876 + }, + { + "epoch": 2.29255698005698, + "grad_norm": 0.7821396589279175, + "learning_rate": 7.736792350042948e-05, + "loss": 0.829, + "step": 12877 + }, + { + "epoch": 2.2927350427350426, + "grad_norm": 0.8877809047698975, + "learning_rate": 7.735428944484021e-05, + "loss": 0.8883, + "step": 12878 + }, + { + "epoch": 2.2929131054131053, + "grad_norm": 0.7754776477813721, + "learning_rate": 7.734065583294656e-05, + "loss": 0.807, + "step": 12879 + }, + { + "epoch": 2.293091168091168, + "grad_norm": 0.851157009601593, + "learning_rate": 7.73270226650157e-05, + "loss": 0.8859, + "step": 12880 + }, + { + "epoch": 2.293269230769231, + "grad_norm": 0.7635365128517151, + "learning_rate": 7.731338994131472e-05, + "loss": 0.9796, + "step": 12881 + }, + { + "epoch": 2.2934472934472936, + "grad_norm": 0.8386050462722778, + "learning_rate": 7.729975766211078e-05, + "loss": 0.788, + "step": 12882 + }, + { + "epoch": 2.2936253561253563, + "grad_norm": 0.7092825174331665, + "learning_rate": 7.728612582767088e-05, + "loss": 0.6855, + "step": 12883 + }, + { + "epoch": 2.2938034188034186, + "grad_norm": 0.8651222586631775, + "learning_rate": 7.72724944382622e-05, + "loss": 0.8875, + "step": 12884 + }, + { + "epoch": 2.2939814814814814, + "grad_norm": 0.89743572473526, + "learning_rate": 7.725886349415175e-05, + "loss": 0.9256, + "step": 12885 + }, + { + "epoch": 2.294159544159544, + "grad_norm": 0.8257600665092468, + "learning_rate": 7.724523299560664e-05, + "loss": 0.6703, + "step": 12886 + }, + { + "epoch": 2.294337606837607, + "grad_norm": 0.8133751153945923, + "learning_rate": 7.72316029428939e-05, + "loss": 0.8991, + "step": 12887 + }, + { + "epoch": 2.2945156695156697, + "grad_norm": 0.7874962687492371, + "learning_rate": 7.721797333628065e-05, + "loss": 0.8679, + "step": 12888 + }, + { + "epoch": 2.294693732193732, + "grad_norm": 0.8284404277801514, + "learning_rate": 7.720434417603384e-05, + "loss": 0.873, + "step": 12889 + }, + { + "epoch": 2.2948717948717947, + "grad_norm": 0.8751698136329651, + "learning_rate": 7.719071546242058e-05, + "loss": 1.0671, + "step": 12890 + }, + { + "epoch": 2.2950498575498575, + "grad_norm": 0.9355120062828064, + "learning_rate": 7.717708719570784e-05, + "loss": 0.93, + "step": 12891 + }, + { + "epoch": 2.29522792022792, + "grad_norm": 0.8643141984939575, + "learning_rate": 7.716345937616267e-05, + "loss": 0.7635, + "step": 12892 + }, + { + "epoch": 2.295405982905983, + "grad_norm": 0.9343852996826172, + "learning_rate": 7.714983200405212e-05, + "loss": 1.0624, + "step": 12893 + }, + { + "epoch": 2.2955840455840457, + "grad_norm": 0.893825352191925, + "learning_rate": 7.71362050796431e-05, + "loss": 1.0843, + "step": 12894 + }, + { + "epoch": 2.2957621082621085, + "grad_norm": 0.920723021030426, + "learning_rate": 7.712257860320269e-05, + "loss": 0.9681, + "step": 12895 + }, + { + "epoch": 2.2959401709401708, + "grad_norm": 0.9275181293487549, + "learning_rate": 7.710895257499778e-05, + "loss": 0.8904, + "step": 12896 + }, + { + "epoch": 2.2961182336182335, + "grad_norm": 0.9343428611755371, + "learning_rate": 7.709532699529543e-05, + "loss": 0.9338, + "step": 12897 + }, + { + "epoch": 2.2962962962962963, + "grad_norm": 0.7457774877548218, + "learning_rate": 7.708170186436252e-05, + "loss": 0.6521, + "step": 12898 + }, + { + "epoch": 2.296474358974359, + "grad_norm": 0.7977834343910217, + "learning_rate": 7.706807718246611e-05, + "loss": 0.887, + "step": 12899 + }, + { + "epoch": 2.296652421652422, + "grad_norm": 0.774459719657898, + "learning_rate": 7.705445294987304e-05, + "loss": 0.914, + "step": 12900 + }, + { + "epoch": 2.296830484330484, + "grad_norm": 0.8464851379394531, + "learning_rate": 7.704082916685034e-05, + "loss": 1.0116, + "step": 12901 + }, + { + "epoch": 2.297008547008547, + "grad_norm": 0.8497290015220642, + "learning_rate": 7.702720583366486e-05, + "loss": 0.9242, + "step": 12902 + }, + { + "epoch": 2.2971866096866096, + "grad_norm": 0.8673670291900635, + "learning_rate": 7.70135829505836e-05, + "loss": 0.8172, + "step": 12903 + }, + { + "epoch": 2.2973646723646723, + "grad_norm": 0.786389172077179, + "learning_rate": 7.699996051787341e-05, + "loss": 0.6713, + "step": 12904 + }, + { + "epoch": 2.297542735042735, + "grad_norm": 0.8441919088363647, + "learning_rate": 7.698633853580124e-05, + "loss": 0.7835, + "step": 12905 + }, + { + "epoch": 2.297720797720798, + "grad_norm": 0.8806493878364563, + "learning_rate": 7.697271700463392e-05, + "loss": 0.9103, + "step": 12906 + }, + { + "epoch": 2.2978988603988606, + "grad_norm": 0.7418580651283264, + "learning_rate": 7.69590959246384e-05, + "loss": 0.9052, + "step": 12907 + }, + { + "epoch": 2.298076923076923, + "grad_norm": 0.7883853316307068, + "learning_rate": 7.694547529608152e-05, + "loss": 0.7689, + "step": 12908 + }, + { + "epoch": 2.2982549857549857, + "grad_norm": 0.7842690944671631, + "learning_rate": 7.693185511923017e-05, + "loss": 0.9587, + "step": 12909 + }, + { + "epoch": 2.2984330484330484, + "grad_norm": 0.884484052658081, + "learning_rate": 7.691823539435119e-05, + "loss": 0.9562, + "step": 12910 + }, + { + "epoch": 2.298611111111111, + "grad_norm": 0.8152852058410645, + "learning_rate": 7.690461612171145e-05, + "loss": 0.9857, + "step": 12911 + }, + { + "epoch": 2.298789173789174, + "grad_norm": 0.8502064943313599, + "learning_rate": 7.689099730157776e-05, + "loss": 0.7806, + "step": 12912 + }, + { + "epoch": 2.298967236467236, + "grad_norm": 0.9655177593231201, + "learning_rate": 7.687737893421697e-05, + "loss": 0.9693, + "step": 12913 + }, + { + "epoch": 2.299145299145299, + "grad_norm": 0.7759003639221191, + "learning_rate": 7.686376101989596e-05, + "loss": 0.9137, + "step": 12914 + }, + { + "epoch": 2.2993233618233617, + "grad_norm": 0.6987054944038391, + "learning_rate": 7.685014355888143e-05, + "loss": 0.9026, + "step": 12915 + }, + { + "epoch": 2.2995014245014245, + "grad_norm": 0.762819230556488, + "learning_rate": 7.683652655144027e-05, + "loss": 0.8358, + "step": 12916 + }, + { + "epoch": 2.2996794871794872, + "grad_norm": 0.8233383893966675, + "learning_rate": 7.682290999783924e-05, + "loss": 0.8468, + "step": 12917 + }, + { + "epoch": 2.29985754985755, + "grad_norm": 0.8558689951896667, + "learning_rate": 7.68092938983452e-05, + "loss": 0.9018, + "step": 12918 + }, + { + "epoch": 2.3000356125356127, + "grad_norm": 0.741760790348053, + "learning_rate": 7.67956782532248e-05, + "loss": 0.7968, + "step": 12919 + }, + { + "epoch": 2.300213675213675, + "grad_norm": 0.9132583737373352, + "learning_rate": 7.678206306274495e-05, + "loss": 0.9952, + "step": 12920 + }, + { + "epoch": 2.300391737891738, + "grad_norm": 0.7656551003456116, + "learning_rate": 7.67684483271723e-05, + "loss": 0.8772, + "step": 12921 + }, + { + "epoch": 2.3005698005698005, + "grad_norm": 0.7407111525535583, + "learning_rate": 7.675483404677364e-05, + "loss": 0.8199, + "step": 12922 + }, + { + "epoch": 2.3007478632478633, + "grad_norm": 0.9602083563804626, + "learning_rate": 7.674122022181571e-05, + "loss": 1.0837, + "step": 12923 + }, + { + "epoch": 2.300925925925926, + "grad_norm": 0.7562392354011536, + "learning_rate": 7.672760685256531e-05, + "loss": 0.8148, + "step": 12924 + }, + { + "epoch": 2.301103988603989, + "grad_norm": 0.9260091185569763, + "learning_rate": 7.671399393928906e-05, + "loss": 0.9508, + "step": 12925 + }, + { + "epoch": 2.301282051282051, + "grad_norm": 0.8745924234390259, + "learning_rate": 7.670038148225374e-05, + "loss": 0.8688, + "step": 12926 + }, + { + "epoch": 2.301460113960114, + "grad_norm": 0.7802116274833679, + "learning_rate": 7.668676948172602e-05, + "loss": 0.7698, + "step": 12927 + }, + { + "epoch": 2.3016381766381766, + "grad_norm": 0.7701709866523743, + "learning_rate": 7.667315793797268e-05, + "loss": 0.7633, + "step": 12928 + }, + { + "epoch": 2.3018162393162394, + "grad_norm": 0.8084021806716919, + "learning_rate": 7.66595468512603e-05, + "loss": 0.8502, + "step": 12929 + }, + { + "epoch": 2.301994301994302, + "grad_norm": 1.0485330820083618, + "learning_rate": 7.664593622185568e-05, + "loss": 0.8049, + "step": 12930 + }, + { + "epoch": 2.302172364672365, + "grad_norm": 0.7852743864059448, + "learning_rate": 7.663232605002535e-05, + "loss": 0.882, + "step": 12931 + }, + { + "epoch": 2.302350427350427, + "grad_norm": 0.7795702815055847, + "learning_rate": 7.661871633603607e-05, + "loss": 0.7841, + "step": 12932 + }, + { + "epoch": 2.30252849002849, + "grad_norm": 0.8882975578308105, + "learning_rate": 7.660510708015448e-05, + "loss": 1.117, + "step": 12933 + }, + { + "epoch": 2.3027065527065527, + "grad_norm": 0.7987662553787231, + "learning_rate": 7.65914982826472e-05, + "loss": 0.8552, + "step": 12934 + }, + { + "epoch": 2.3028846153846154, + "grad_norm": 0.8141679167747498, + "learning_rate": 7.657788994378095e-05, + "loss": 0.8288, + "step": 12935 + }, + { + "epoch": 2.303062678062678, + "grad_norm": 0.8506320118904114, + "learning_rate": 7.656428206382222e-05, + "loss": 0.7521, + "step": 12936 + }, + { + "epoch": 2.303240740740741, + "grad_norm": 0.7666227221488953, + "learning_rate": 7.655067464303773e-05, + "loss": 0.8394, + "step": 12937 + }, + { + "epoch": 2.3034188034188032, + "grad_norm": 0.8018062710762024, + "learning_rate": 7.653706768169405e-05, + "loss": 0.7566, + "step": 12938 + }, + { + "epoch": 2.303596866096866, + "grad_norm": 0.8054059743881226, + "learning_rate": 7.652346118005779e-05, + "loss": 0.8749, + "step": 12939 + }, + { + "epoch": 2.3037749287749287, + "grad_norm": 0.8663263320922852, + "learning_rate": 7.650985513839554e-05, + "loss": 0.7799, + "step": 12940 + }, + { + "epoch": 2.3039529914529915, + "grad_norm": 0.7591161727905273, + "learning_rate": 7.64962495569739e-05, + "loss": 0.7378, + "step": 12941 + }, + { + "epoch": 2.3041310541310542, + "grad_norm": 0.8118969202041626, + "learning_rate": 7.64826444360594e-05, + "loss": 0.7948, + "step": 12942 + }, + { + "epoch": 2.304309116809117, + "grad_norm": 0.6880847811698914, + "learning_rate": 7.646903977591865e-05, + "loss": 0.9164, + "step": 12943 + }, + { + "epoch": 2.3044871794871793, + "grad_norm": 0.814386248588562, + "learning_rate": 7.645543557681816e-05, + "loss": 0.7998, + "step": 12944 + }, + { + "epoch": 2.304665242165242, + "grad_norm": 0.8295530676841736, + "learning_rate": 7.644183183902454e-05, + "loss": 0.812, + "step": 12945 + }, + { + "epoch": 2.304843304843305, + "grad_norm": 0.7872505187988281, + "learning_rate": 7.642822856280424e-05, + "loss": 0.9073, + "step": 12946 + }, + { + "epoch": 2.3050213675213675, + "grad_norm": 0.9217497110366821, + "learning_rate": 7.641462574842387e-05, + "loss": 0.7762, + "step": 12947 + }, + { + "epoch": 2.3051994301994303, + "grad_norm": 0.7502169609069824, + "learning_rate": 7.640102339614987e-05, + "loss": 0.9374, + "step": 12948 + }, + { + "epoch": 2.305377492877493, + "grad_norm": 0.8262767195701599, + "learning_rate": 7.638742150624886e-05, + "loss": 0.5363, + "step": 12949 + }, + { + "epoch": 2.3055555555555554, + "grad_norm": 0.7571384310722351, + "learning_rate": 7.637382007898722e-05, + "loss": 0.9548, + "step": 12950 + }, + { + "epoch": 2.305733618233618, + "grad_norm": 0.7899317145347595, + "learning_rate": 7.636021911463152e-05, + "loss": 0.7718, + "step": 12951 + }, + { + "epoch": 2.305911680911681, + "grad_norm": 0.7772458791732788, + "learning_rate": 7.634661861344819e-05, + "loss": 0.7158, + "step": 12952 + }, + { + "epoch": 2.3060897435897436, + "grad_norm": 0.8279168009757996, + "learning_rate": 7.633301857570374e-05, + "loss": 0.7835, + "step": 12953 + }, + { + "epoch": 2.3062678062678064, + "grad_norm": 0.751268208026886, + "learning_rate": 7.631941900166468e-05, + "loss": 0.8609, + "step": 12954 + }, + { + "epoch": 2.306445868945869, + "grad_norm": 0.8101294636726379, + "learning_rate": 7.630581989159736e-05, + "loss": 1.0242, + "step": 12955 + }, + { + "epoch": 2.306623931623932, + "grad_norm": 0.7707645297050476, + "learning_rate": 7.629222124576831e-05, + "loss": 0.7969, + "step": 12956 + }, + { + "epoch": 2.306801994301994, + "grad_norm": 0.6519944667816162, + "learning_rate": 7.627862306444391e-05, + "loss": 0.5459, + "step": 12957 + }, + { + "epoch": 2.306980056980057, + "grad_norm": 0.7738897800445557, + "learning_rate": 7.626502534789063e-05, + "loss": 0.957, + "step": 12958 + }, + { + "epoch": 2.3071581196581197, + "grad_norm": 0.7059842944145203, + "learning_rate": 7.625142809637485e-05, + "loss": 0.6316, + "step": 12959 + }, + { + "epoch": 2.3073361823361824, + "grad_norm": 0.8380797505378723, + "learning_rate": 7.623783131016305e-05, + "loss": 0.9685, + "step": 12960 + }, + { + "epoch": 2.307514245014245, + "grad_norm": 0.8272121548652649, + "learning_rate": 7.622423498952154e-05, + "loss": 0.9425, + "step": 12961 + }, + { + "epoch": 2.3076923076923075, + "grad_norm": 0.763522744178772, + "learning_rate": 7.621063913471678e-05, + "loss": 0.7778, + "step": 12962 + }, + { + "epoch": 2.3078703703703702, + "grad_norm": 0.8345584273338318, + "learning_rate": 7.61970437460151e-05, + "loss": 0.9652, + "step": 12963 + }, + { + "epoch": 2.308048433048433, + "grad_norm": 0.943286657333374, + "learning_rate": 7.618344882368294e-05, + "loss": 0.9088, + "step": 12964 + }, + { + "epoch": 2.3082264957264957, + "grad_norm": 0.8568450212478638, + "learning_rate": 7.616985436798659e-05, + "loss": 0.7535, + "step": 12965 + }, + { + "epoch": 2.3084045584045585, + "grad_norm": 0.8722548484802246, + "learning_rate": 7.615626037919248e-05, + "loss": 0.9802, + "step": 12966 + }, + { + "epoch": 2.3085826210826212, + "grad_norm": 1.0332363843917847, + "learning_rate": 7.614266685756688e-05, + "loss": 0.9105, + "step": 12967 + }, + { + "epoch": 2.308760683760684, + "grad_norm": 0.7503480315208435, + "learning_rate": 7.612907380337619e-05, + "loss": 0.7345, + "step": 12968 + }, + { + "epoch": 2.3089387464387463, + "grad_norm": 0.7406014204025269, + "learning_rate": 7.611548121688668e-05, + "loss": 0.9222, + "step": 12969 + }, + { + "epoch": 2.309116809116809, + "grad_norm": 0.7574487328529358, + "learning_rate": 7.610188909836474e-05, + "loss": 0.7709, + "step": 12970 + }, + { + "epoch": 2.309294871794872, + "grad_norm": 0.8669037818908691, + "learning_rate": 7.608829744807661e-05, + "loss": 0.838, + "step": 12971 + }, + { + "epoch": 2.3094729344729346, + "grad_norm": 0.7544569373130798, + "learning_rate": 7.607470626628861e-05, + "loss": 0.6966, + "step": 12972 + }, + { + "epoch": 2.3096509971509973, + "grad_norm": 0.8201249241828918, + "learning_rate": 7.606111555326706e-05, + "loss": 0.9322, + "step": 12973 + }, + { + "epoch": 2.3098290598290596, + "grad_norm": 0.7935477495193481, + "learning_rate": 7.60475253092782e-05, + "loss": 0.7981, + "step": 12974 + }, + { + "epoch": 2.3100071225071224, + "grad_norm": 0.8775026798248291, + "learning_rate": 7.603393553458838e-05, + "loss": 0.8352, + "step": 12975 + }, + { + "epoch": 2.310185185185185, + "grad_norm": 0.8422450423240662, + "learning_rate": 7.602034622946374e-05, + "loss": 0.577, + "step": 12976 + }, + { + "epoch": 2.310363247863248, + "grad_norm": 0.8584204316139221, + "learning_rate": 7.600675739417067e-05, + "loss": 0.9767, + "step": 12977 + }, + { + "epoch": 2.3105413105413106, + "grad_norm": 0.7818547487258911, + "learning_rate": 7.599316902897528e-05, + "loss": 0.7675, + "step": 12978 + }, + { + "epoch": 2.3107193732193734, + "grad_norm": 0.93815016746521, + "learning_rate": 7.597958113414391e-05, + "loss": 0.8517, + "step": 12979 + }, + { + "epoch": 2.310897435897436, + "grad_norm": 0.8092408776283264, + "learning_rate": 7.596599370994272e-05, + "loss": 0.7266, + "step": 12980 + }, + { + "epoch": 2.3110754985754984, + "grad_norm": 0.8577243089675903, + "learning_rate": 7.595240675663802e-05, + "loss": 0.9138, + "step": 12981 + }, + { + "epoch": 2.311253561253561, + "grad_norm": 0.8600401878356934, + "learning_rate": 7.59388202744959e-05, + "loss": 0.8348, + "step": 12982 + }, + { + "epoch": 2.311431623931624, + "grad_norm": 0.6399564743041992, + "learning_rate": 7.592523426378264e-05, + "loss": 0.6649, + "step": 12983 + }, + { + "epoch": 2.3116096866096867, + "grad_norm": 0.7916820049285889, + "learning_rate": 7.591164872476438e-05, + "loss": 0.8048, + "step": 12984 + }, + { + "epoch": 2.3117877492877494, + "grad_norm": 0.7748355269432068, + "learning_rate": 7.589806365770736e-05, + "loss": 1.0101, + "step": 12985 + }, + { + "epoch": 2.3119658119658117, + "grad_norm": 0.8463436365127563, + "learning_rate": 7.588447906287767e-05, + "loss": 0.7547, + "step": 12986 + }, + { + "epoch": 2.3121438746438745, + "grad_norm": 0.8257808685302734, + "learning_rate": 7.587089494054155e-05, + "loss": 0.8093, + "step": 12987 + }, + { + "epoch": 2.3123219373219372, + "grad_norm": 0.843781054019928, + "learning_rate": 7.58573112909651e-05, + "loss": 0.8379, + "step": 12988 + }, + { + "epoch": 2.3125, + "grad_norm": 0.8782341480255127, + "learning_rate": 7.584372811441452e-05, + "loss": 0.9372, + "step": 12989 + }, + { + "epoch": 2.3126780626780628, + "grad_norm": 0.8465158343315125, + "learning_rate": 7.583014541115585e-05, + "loss": 0.8427, + "step": 12990 + }, + { + "epoch": 2.3128561253561255, + "grad_norm": 0.7140238285064697, + "learning_rate": 7.58165631814553e-05, + "loss": 0.7896, + "step": 12991 + }, + { + "epoch": 2.3130341880341883, + "grad_norm": 0.9414699077606201, + "learning_rate": 7.580298142557898e-05, + "loss": 1.0464, + "step": 12992 + }, + { + "epoch": 2.3132122507122506, + "grad_norm": 0.7970326542854309, + "learning_rate": 7.578940014379293e-05, + "loss": 0.696, + "step": 12993 + }, + { + "epoch": 2.3133903133903133, + "grad_norm": 0.7377375960350037, + "learning_rate": 7.577581933636332e-05, + "loss": 0.7205, + "step": 12994 + }, + { + "epoch": 2.313568376068376, + "grad_norm": 0.7740316987037659, + "learning_rate": 7.576223900355619e-05, + "loss": 0.7448, + "step": 12995 + }, + { + "epoch": 2.313746438746439, + "grad_norm": 0.7820385098457336, + "learning_rate": 7.574865914563767e-05, + "loss": 0.7289, + "step": 12996 + }, + { + "epoch": 2.3139245014245016, + "grad_norm": 0.8180822134017944, + "learning_rate": 7.573507976287376e-05, + "loss": 0.8709, + "step": 12997 + }, + { + "epoch": 2.314102564102564, + "grad_norm": 0.9008440971374512, + "learning_rate": 7.572150085553058e-05, + "loss": 0.7938, + "step": 12998 + }, + { + "epoch": 2.3142806267806266, + "grad_norm": 0.786400318145752, + "learning_rate": 7.570792242387414e-05, + "loss": 0.9866, + "step": 12999 + }, + { + "epoch": 2.3144586894586894, + "grad_norm": 0.872160792350769, + "learning_rate": 7.569434446817052e-05, + "loss": 0.7319, + "step": 13000 + }, + { + "epoch": 2.314636752136752, + "grad_norm": 0.7858988642692566, + "learning_rate": 7.56807669886857e-05, + "loss": 0.8786, + "step": 13001 + }, + { + "epoch": 2.314814814814815, + "grad_norm": 0.7090579271316528, + "learning_rate": 7.566718998568579e-05, + "loss": 0.7092, + "step": 13002 + }, + { + "epoch": 2.3149928774928776, + "grad_norm": 0.7881498336791992, + "learning_rate": 7.565361345943668e-05, + "loss": 0.876, + "step": 13003 + }, + { + "epoch": 2.3151709401709404, + "grad_norm": 0.8768819570541382, + "learning_rate": 7.564003741020447e-05, + "loss": 1.0374, + "step": 13004 + }, + { + "epoch": 2.3153490028490027, + "grad_norm": 0.7608295679092407, + "learning_rate": 7.56264618382551e-05, + "loss": 0.7551, + "step": 13005 + }, + { + "epoch": 2.3155270655270654, + "grad_norm": 0.6947942972183228, + "learning_rate": 7.561288674385462e-05, + "loss": 0.8132, + "step": 13006 + }, + { + "epoch": 2.315705128205128, + "grad_norm": 0.8722706437110901, + "learning_rate": 7.559931212726892e-05, + "loss": 0.7486, + "step": 13007 + }, + { + "epoch": 2.315883190883191, + "grad_norm": 0.9804681539535522, + "learning_rate": 7.558573798876404e-05, + "loss": 0.899, + "step": 13008 + }, + { + "epoch": 2.3160612535612537, + "grad_norm": 0.9246195554733276, + "learning_rate": 7.557216432860587e-05, + "loss": 0.7742, + "step": 13009 + }, + { + "epoch": 2.316239316239316, + "grad_norm": 0.8792895078659058, + "learning_rate": 7.555859114706046e-05, + "loss": 0.8299, + "step": 13010 + }, + { + "epoch": 2.3164173789173788, + "grad_norm": 0.8280500769615173, + "learning_rate": 7.554501844439362e-05, + "loss": 0.8708, + "step": 13011 + }, + { + "epoch": 2.3165954415954415, + "grad_norm": 0.8560570478439331, + "learning_rate": 7.553144622087136e-05, + "loss": 0.9571, + "step": 13012 + }, + { + "epoch": 2.3167735042735043, + "grad_norm": 0.8504697680473328, + "learning_rate": 7.551787447675962e-05, + "loss": 0.7609, + "step": 13013 + }, + { + "epoch": 2.316951566951567, + "grad_norm": 0.8199480772018433, + "learning_rate": 7.550430321232423e-05, + "loss": 0.7077, + "step": 13014 + }, + { + "epoch": 2.3171296296296298, + "grad_norm": 0.854341447353363, + "learning_rate": 7.549073242783115e-05, + "loss": 0.9602, + "step": 13015 + }, + { + "epoch": 2.3173076923076925, + "grad_norm": 0.7619566917419434, + "learning_rate": 7.547716212354623e-05, + "loss": 0.9967, + "step": 13016 + }, + { + "epoch": 2.317485754985755, + "grad_norm": 0.7371547222137451, + "learning_rate": 7.546359229973543e-05, + "loss": 0.7239, + "step": 13017 + }, + { + "epoch": 2.3176638176638176, + "grad_norm": 0.7797415852546692, + "learning_rate": 7.545002295666453e-05, + "loss": 0.7472, + "step": 13018 + }, + { + "epoch": 2.3178418803418803, + "grad_norm": 0.8549608588218689, + "learning_rate": 7.543645409459943e-05, + "loss": 0.8968, + "step": 13019 + }, + { + "epoch": 2.318019943019943, + "grad_norm": 0.7931239008903503, + "learning_rate": 7.542288571380598e-05, + "loss": 0.9853, + "step": 13020 + }, + { + "epoch": 2.318198005698006, + "grad_norm": 0.797726035118103, + "learning_rate": 7.540931781455008e-05, + "loss": 0.9366, + "step": 13021 + }, + { + "epoch": 2.318376068376068, + "grad_norm": 0.7382092475891113, + "learning_rate": 7.539575039709747e-05, + "loss": 0.6484, + "step": 13022 + }, + { + "epoch": 2.318554131054131, + "grad_norm": 0.83231121301651, + "learning_rate": 7.538218346171403e-05, + "loss": 1.0184, + "step": 13023 + }, + { + "epoch": 2.3187321937321936, + "grad_norm": 0.8613845109939575, + "learning_rate": 7.536861700866554e-05, + "loss": 0.8019, + "step": 13024 + }, + { + "epoch": 2.3189102564102564, + "grad_norm": 0.7736538648605347, + "learning_rate": 7.53550510382179e-05, + "loss": 0.8228, + "step": 13025 + }, + { + "epoch": 2.319088319088319, + "grad_norm": 0.7894346714019775, + "learning_rate": 7.534148555063678e-05, + "loss": 0.8189, + "step": 13026 + }, + { + "epoch": 2.319266381766382, + "grad_norm": 0.7333146333694458, + "learning_rate": 7.532792054618807e-05, + "loss": 0.8456, + "step": 13027 + }, + { + "epoch": 2.3194444444444446, + "grad_norm": 1.0321780443191528, + "learning_rate": 7.531435602513745e-05, + "loss": 0.8594, + "step": 13028 + }, + { + "epoch": 2.319622507122507, + "grad_norm": 0.8658601641654968, + "learning_rate": 7.530079198775079e-05, + "loss": 0.9638, + "step": 13029 + }, + { + "epoch": 2.3198005698005697, + "grad_norm": 0.7287920713424683, + "learning_rate": 7.528722843429376e-05, + "loss": 0.7001, + "step": 13030 + }, + { + "epoch": 2.3199786324786325, + "grad_norm": 0.7398431301116943, + "learning_rate": 7.527366536503218e-05, + "loss": 0.8595, + "step": 13031 + }, + { + "epoch": 2.320156695156695, + "grad_norm": 0.8127739429473877, + "learning_rate": 7.526010278023178e-05, + "loss": 0.7641, + "step": 13032 + }, + { + "epoch": 2.320334757834758, + "grad_norm": 0.776497483253479, + "learning_rate": 7.524654068015824e-05, + "loss": 0.8666, + "step": 13033 + }, + { + "epoch": 2.3205128205128207, + "grad_norm": 0.8524185419082642, + "learning_rate": 7.523297906507733e-05, + "loss": 0.9815, + "step": 13034 + }, + { + "epoch": 2.320690883190883, + "grad_norm": 0.7745016813278198, + "learning_rate": 7.521941793525474e-05, + "loss": 0.8527, + "step": 13035 + }, + { + "epoch": 2.3208689458689458, + "grad_norm": 0.8695911169052124, + "learning_rate": 7.52058572909562e-05, + "loss": 0.8881, + "step": 13036 + }, + { + "epoch": 2.3210470085470085, + "grad_norm": 0.788969099521637, + "learning_rate": 7.519229713244736e-05, + "loss": 0.7886, + "step": 13037 + }, + { + "epoch": 2.3212250712250713, + "grad_norm": 0.776520311832428, + "learning_rate": 7.517873745999394e-05, + "loss": 0.5986, + "step": 13038 + }, + { + "epoch": 2.321403133903134, + "grad_norm": 0.8118561506271362, + "learning_rate": 7.516517827386158e-05, + "loss": 0.8805, + "step": 13039 + }, + { + "epoch": 2.3215811965811968, + "grad_norm": 0.8859134912490845, + "learning_rate": 7.515161957431596e-05, + "loss": 0.8861, + "step": 13040 + }, + { + "epoch": 2.321759259259259, + "grad_norm": 0.8181297779083252, + "learning_rate": 7.513806136162273e-05, + "loss": 0.9015, + "step": 13041 + }, + { + "epoch": 2.321937321937322, + "grad_norm": 0.8488339185714722, + "learning_rate": 7.512450363604759e-05, + "loss": 1.0423, + "step": 13042 + }, + { + "epoch": 2.3221153846153846, + "grad_norm": 0.7755734920501709, + "learning_rate": 7.511094639785607e-05, + "loss": 0.7595, + "step": 13043 + }, + { + "epoch": 2.3222934472934473, + "grad_norm": 0.8437283635139465, + "learning_rate": 7.509738964731389e-05, + "loss": 0.9011, + "step": 13044 + }, + { + "epoch": 2.32247150997151, + "grad_norm": 0.7508310675621033, + "learning_rate": 7.508383338468659e-05, + "loss": 0.8335, + "step": 13045 + }, + { + "epoch": 2.322649572649573, + "grad_norm": 0.8001464605331421, + "learning_rate": 7.507027761023987e-05, + "loss": 0.9785, + "step": 13046 + }, + { + "epoch": 2.322827635327635, + "grad_norm": 0.8142531514167786, + "learning_rate": 7.505672232423923e-05, + "loss": 0.8577, + "step": 13047 + }, + { + "epoch": 2.323005698005698, + "grad_norm": 0.7852125763893127, + "learning_rate": 7.504316752695035e-05, + "loss": 0.798, + "step": 13048 + }, + { + "epoch": 2.3231837606837606, + "grad_norm": 0.8998631238937378, + "learning_rate": 7.502961321863871e-05, + "loss": 0.9291, + "step": 13049 + }, + { + "epoch": 2.3233618233618234, + "grad_norm": 0.8850175738334656, + "learning_rate": 7.501605939956995e-05, + "loss": 0.9536, + "step": 13050 + }, + { + "epoch": 2.323539886039886, + "grad_norm": 0.8305020928382874, + "learning_rate": 7.500250607000959e-05, + "loss": 0.8695, + "step": 13051 + }, + { + "epoch": 2.323717948717949, + "grad_norm": 0.8073359727859497, + "learning_rate": 7.498895323022317e-05, + "loss": 0.6831, + "step": 13052 + }, + { + "epoch": 2.323896011396011, + "grad_norm": 0.8435724973678589, + "learning_rate": 7.497540088047632e-05, + "loss": 0.9419, + "step": 13053 + }, + { + "epoch": 2.324074074074074, + "grad_norm": 0.927147388458252, + "learning_rate": 7.496184902103446e-05, + "loss": 0.957, + "step": 13054 + }, + { + "epoch": 2.3242521367521367, + "grad_norm": 0.7923009395599365, + "learning_rate": 7.494829765216319e-05, + "loss": 0.839, + "step": 13055 + }, + { + "epoch": 2.3244301994301995, + "grad_norm": 0.7830277681350708, + "learning_rate": 7.493474677412794e-05, + "loss": 1.0236, + "step": 13056 + }, + { + "epoch": 2.324608262108262, + "grad_norm": 0.8470967411994934, + "learning_rate": 7.492119638719432e-05, + "loss": 0.9144, + "step": 13057 + }, + { + "epoch": 2.324786324786325, + "grad_norm": 0.7469272613525391, + "learning_rate": 7.490764649162771e-05, + "loss": 0.7101, + "step": 13058 + }, + { + "epoch": 2.3249643874643873, + "grad_norm": 0.9236082434654236, + "learning_rate": 7.489409708769366e-05, + "loss": 0.6658, + "step": 13059 + }, + { + "epoch": 2.32514245014245, + "grad_norm": 0.8271692395210266, + "learning_rate": 7.48805481756576e-05, + "loss": 0.775, + "step": 13060 + }, + { + "epoch": 2.3253205128205128, + "grad_norm": 0.9878279566764832, + "learning_rate": 7.486699975578507e-05, + "loss": 0.6881, + "step": 13061 + }, + { + "epoch": 2.3254985754985755, + "grad_norm": 0.7332003712654114, + "learning_rate": 7.485345182834142e-05, + "loss": 0.8384, + "step": 13062 + }, + { + "epoch": 2.3256766381766383, + "grad_norm": 0.9525214433670044, + "learning_rate": 7.483990439359221e-05, + "loss": 0.9892, + "step": 13063 + }, + { + "epoch": 2.325854700854701, + "grad_norm": 0.7413233518600464, + "learning_rate": 7.482635745180273e-05, + "loss": 0.8749, + "step": 13064 + }, + { + "epoch": 2.326032763532764, + "grad_norm": 0.8286891579627991, + "learning_rate": 7.481281100323854e-05, + "loss": 1.0313, + "step": 13065 + }, + { + "epoch": 2.326210826210826, + "grad_norm": 0.868653416633606, + "learning_rate": 7.479926504816495e-05, + "loss": 0.7407, + "step": 13066 + }, + { + "epoch": 2.326388888888889, + "grad_norm": 0.790052056312561, + "learning_rate": 7.478571958684746e-05, + "loss": 0.7156, + "step": 13067 + }, + { + "epoch": 2.3265669515669516, + "grad_norm": 0.8799049854278564, + "learning_rate": 7.477217461955137e-05, + "loss": 0.7706, + "step": 13068 + }, + { + "epoch": 2.3267450142450143, + "grad_norm": 0.8246361017227173, + "learning_rate": 7.475863014654214e-05, + "loss": 0.9402, + "step": 13069 + }, + { + "epoch": 2.326923076923077, + "grad_norm": 0.870917022228241, + "learning_rate": 7.474508616808508e-05, + "loss": 0.962, + "step": 13070 + }, + { + "epoch": 2.3271011396011394, + "grad_norm": 0.8706079125404358, + "learning_rate": 7.473154268444563e-05, + "loss": 0.9094, + "step": 13071 + }, + { + "epoch": 2.327279202279202, + "grad_norm": 0.9031453728675842, + "learning_rate": 7.471799969588912e-05, + "loss": 0.8447, + "step": 13072 + }, + { + "epoch": 2.327457264957265, + "grad_norm": 0.9153435230255127, + "learning_rate": 7.470445720268086e-05, + "loss": 0.9935, + "step": 13073 + }, + { + "epoch": 2.3276353276353277, + "grad_norm": 0.8236302733421326, + "learning_rate": 7.469091520508624e-05, + "loss": 0.7911, + "step": 13074 + }, + { + "epoch": 2.3278133903133904, + "grad_norm": 0.7344710826873779, + "learning_rate": 7.467737370337054e-05, + "loss": 0.6565, + "step": 13075 + }, + { + "epoch": 2.327991452991453, + "grad_norm": 0.8711966276168823, + "learning_rate": 7.466383269779911e-05, + "loss": 0.8332, + "step": 13076 + }, + { + "epoch": 2.328169515669516, + "grad_norm": 0.836825966835022, + "learning_rate": 7.465029218863723e-05, + "loss": 0.9359, + "step": 13077 + }, + { + "epoch": 2.328347578347578, + "grad_norm": 0.9791260361671448, + "learning_rate": 7.463675217615024e-05, + "loss": 0.8938, + "step": 13078 + }, + { + "epoch": 2.328525641025641, + "grad_norm": 0.7260454893112183, + "learning_rate": 7.46232126606034e-05, + "loss": 0.6871, + "step": 13079 + }, + { + "epoch": 2.3287037037037037, + "grad_norm": 0.7887428998947144, + "learning_rate": 7.460967364226197e-05, + "loss": 0.8098, + "step": 13080 + }, + { + "epoch": 2.3288817663817665, + "grad_norm": 0.8303743004798889, + "learning_rate": 7.459613512139124e-05, + "loss": 0.8897, + "step": 13081 + }, + { + "epoch": 2.3290598290598292, + "grad_norm": 0.7933324575424194, + "learning_rate": 7.458259709825652e-05, + "loss": 0.754, + "step": 13082 + }, + { + "epoch": 2.3292378917378915, + "grad_norm": 0.8998779058456421, + "learning_rate": 7.456905957312296e-05, + "loss": 0.866, + "step": 13083 + }, + { + "epoch": 2.3294159544159543, + "grad_norm": 0.8205044269561768, + "learning_rate": 7.455552254625588e-05, + "loss": 0.7455, + "step": 13084 + }, + { + "epoch": 2.329594017094017, + "grad_norm": 0.8731769323348999, + "learning_rate": 7.454198601792046e-05, + "loss": 0.8876, + "step": 13085 + }, + { + "epoch": 2.32977207977208, + "grad_norm": 0.9183599352836609, + "learning_rate": 7.452844998838194e-05, + "loss": 1.0991, + "step": 13086 + }, + { + "epoch": 2.3299501424501425, + "grad_norm": 0.8820931315422058, + "learning_rate": 7.451491445790553e-05, + "loss": 0.7591, + "step": 13087 + }, + { + "epoch": 2.3301282051282053, + "grad_norm": 0.7837240099906921, + "learning_rate": 7.450137942675646e-05, + "loss": 0.792, + "step": 13088 + }, + { + "epoch": 2.330306267806268, + "grad_norm": 0.8960266709327698, + "learning_rate": 7.448784489519984e-05, + "loss": 0.9725, + "step": 13089 + }, + { + "epoch": 2.3304843304843303, + "grad_norm": 0.8010196685791016, + "learning_rate": 7.447431086350092e-05, + "loss": 0.8418, + "step": 13090 + }, + { + "epoch": 2.330662393162393, + "grad_norm": 0.8652680516242981, + "learning_rate": 7.446077733192486e-05, + "loss": 0.93, + "step": 13091 + }, + { + "epoch": 2.330840455840456, + "grad_norm": 0.9385902285575867, + "learning_rate": 7.44472443007368e-05, + "loss": 0.968, + "step": 13092 + }, + { + "epoch": 2.3310185185185186, + "grad_norm": 0.8097951412200928, + "learning_rate": 7.443371177020195e-05, + "loss": 0.8715, + "step": 13093 + }, + { + "epoch": 2.3311965811965814, + "grad_norm": 0.7931473255157471, + "learning_rate": 7.442017974058537e-05, + "loss": 0.865, + "step": 13094 + }, + { + "epoch": 2.3313746438746437, + "grad_norm": 0.7680486440658569, + "learning_rate": 7.440664821215224e-05, + "loss": 0.9155, + "step": 13095 + }, + { + "epoch": 2.3315527065527064, + "grad_norm": 0.8128345012664795, + "learning_rate": 7.439311718516766e-05, + "loss": 0.8707, + "step": 13096 + }, + { + "epoch": 2.331730769230769, + "grad_norm": 0.9534463286399841, + "learning_rate": 7.43795866598968e-05, + "loss": 1.1102, + "step": 13097 + }, + { + "epoch": 2.331908831908832, + "grad_norm": 0.9140331745147705, + "learning_rate": 7.436605663660468e-05, + "loss": 0.7744, + "step": 13098 + }, + { + "epoch": 2.3320868945868947, + "grad_norm": 0.8316463828086853, + "learning_rate": 7.435252711555645e-05, + "loss": 0.7201, + "step": 13099 + }, + { + "epoch": 2.3322649572649574, + "grad_norm": 0.7714298963546753, + "learning_rate": 7.433899809701714e-05, + "loss": 0.9111, + "step": 13100 + }, + { + "epoch": 2.33244301994302, + "grad_norm": 0.999081552028656, + "learning_rate": 7.432546958125188e-05, + "loss": 0.8839, + "step": 13101 + }, + { + "epoch": 2.3326210826210825, + "grad_norm": 0.8515602350234985, + "learning_rate": 7.43119415685257e-05, + "loss": 0.7245, + "step": 13102 + }, + { + "epoch": 2.3327991452991452, + "grad_norm": 0.9441094398498535, + "learning_rate": 7.42984140591037e-05, + "loss": 0.8678, + "step": 13103 + }, + { + "epoch": 2.332977207977208, + "grad_norm": 0.8220996856689453, + "learning_rate": 7.428488705325084e-05, + "loss": 0.9737, + "step": 13104 + }, + { + "epoch": 2.3331552706552707, + "grad_norm": 0.8164090514183044, + "learning_rate": 7.427136055123222e-05, + "loss": 0.9138, + "step": 13105 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.9672707319259644, + "learning_rate": 7.425783455331281e-05, + "loss": 0.7723, + "step": 13106 + }, + { + "epoch": 2.333511396011396, + "grad_norm": 0.7953858971595764, + "learning_rate": 7.424430905975773e-05, + "loss": 0.8876, + "step": 13107 + }, + { + "epoch": 2.3336894586894585, + "grad_norm": 0.7809541821479797, + "learning_rate": 7.423078407083183e-05, + "loss": 1.0268, + "step": 13108 + }, + { + "epoch": 2.3338675213675213, + "grad_norm": 0.805270254611969, + "learning_rate": 7.421725958680025e-05, + "loss": 0.7515, + "step": 13109 + }, + { + "epoch": 2.334045584045584, + "grad_norm": 0.8066652417182922, + "learning_rate": 7.420373560792788e-05, + "loss": 1.004, + "step": 13110 + }, + { + "epoch": 2.334223646723647, + "grad_norm": 0.9382686018943787, + "learning_rate": 7.41902121344797e-05, + "loss": 0.8769, + "step": 13111 + }, + { + "epoch": 2.3344017094017095, + "grad_norm": 0.7908356785774231, + "learning_rate": 7.417668916672074e-05, + "loss": 0.8491, + "step": 13112 + }, + { + "epoch": 2.3345797720797723, + "grad_norm": 0.7188867330551147, + "learning_rate": 7.416316670491588e-05, + "loss": 0.7138, + "step": 13113 + }, + { + "epoch": 2.3347578347578346, + "grad_norm": 0.8477714657783508, + "learning_rate": 7.414964474933012e-05, + "loss": 0.9715, + "step": 13114 + }, + { + "epoch": 2.3349358974358974, + "grad_norm": 0.8769845366477966, + "learning_rate": 7.413612330022835e-05, + "loss": 0.9029, + "step": 13115 + }, + { + "epoch": 2.33511396011396, + "grad_norm": 0.9011028409004211, + "learning_rate": 7.412260235787554e-05, + "loss": 1.026, + "step": 13116 + }, + { + "epoch": 2.335292022792023, + "grad_norm": 0.7775689363479614, + "learning_rate": 7.410908192253656e-05, + "loss": 0.8492, + "step": 13117 + }, + { + "epoch": 2.3354700854700856, + "grad_norm": 0.9587660431861877, + "learning_rate": 7.409556199447637e-05, + "loss": 0.8731, + "step": 13118 + }, + { + "epoch": 2.335648148148148, + "grad_norm": 0.8117266297340393, + "learning_rate": 7.408204257395979e-05, + "loss": 0.8827, + "step": 13119 + }, + { + "epoch": 2.3358262108262107, + "grad_norm": 0.7382497787475586, + "learning_rate": 7.40685236612518e-05, + "loss": 0.6617, + "step": 13120 + }, + { + "epoch": 2.3360042735042734, + "grad_norm": 0.8630974888801575, + "learning_rate": 7.405500525661717e-05, + "loss": 0.843, + "step": 13121 + }, + { + "epoch": 2.336182336182336, + "grad_norm": 0.7496539950370789, + "learning_rate": 7.404148736032083e-05, + "loss": 1.0354, + "step": 13122 + }, + { + "epoch": 2.336360398860399, + "grad_norm": 0.8409397602081299, + "learning_rate": 7.402796997262761e-05, + "loss": 0.9848, + "step": 13123 + }, + { + "epoch": 2.3365384615384617, + "grad_norm": 0.8018865585327148, + "learning_rate": 7.40144530938024e-05, + "loss": 0.8929, + "step": 13124 + }, + { + "epoch": 2.3367165242165244, + "grad_norm": 0.7378625869750977, + "learning_rate": 7.400093672410996e-05, + "loss": 0.7749, + "step": 13125 + }, + { + "epoch": 2.3368945868945867, + "grad_norm": 0.8906251192092896, + "learning_rate": 7.398742086381519e-05, + "loss": 1.007, + "step": 13126 + }, + { + "epoch": 2.3370726495726495, + "grad_norm": 0.8324725031852722, + "learning_rate": 7.397390551318283e-05, + "loss": 0.8493, + "step": 13127 + }, + { + "epoch": 2.3372507122507122, + "grad_norm": 0.781080961227417, + "learning_rate": 7.39603906724778e-05, + "loss": 0.696, + "step": 13128 + }, + { + "epoch": 2.337428774928775, + "grad_norm": 0.8068976402282715, + "learning_rate": 7.394687634196476e-05, + "loss": 0.7196, + "step": 13129 + }, + { + "epoch": 2.3376068376068377, + "grad_norm": 0.7588358521461487, + "learning_rate": 7.393336252190854e-05, + "loss": 0.9179, + "step": 13130 + }, + { + "epoch": 2.3377849002849, + "grad_norm": 0.8334088325500488, + "learning_rate": 7.391984921257398e-05, + "loss": 0.94, + "step": 13131 + }, + { + "epoch": 2.337962962962963, + "grad_norm": 0.9485353231430054, + "learning_rate": 7.390633641422578e-05, + "loss": 0.9253, + "step": 13132 + }, + { + "epoch": 2.3381410256410255, + "grad_norm": 0.9447978734970093, + "learning_rate": 7.389282412712874e-05, + "loss": 0.9112, + "step": 13133 + }, + { + "epoch": 2.3383190883190883, + "grad_norm": 0.7348376512527466, + "learning_rate": 7.387931235154754e-05, + "loss": 0.7817, + "step": 13134 + }, + { + "epoch": 2.338497150997151, + "grad_norm": 0.8610092401504517, + "learning_rate": 7.386580108774699e-05, + "loss": 0.8231, + "step": 13135 + }, + { + "epoch": 2.338675213675214, + "grad_norm": 0.8314286470413208, + "learning_rate": 7.385229033599175e-05, + "loss": 0.7323, + "step": 13136 + }, + { + "epoch": 2.3388532763532766, + "grad_norm": 0.7775855660438538, + "learning_rate": 7.383878009654657e-05, + "loss": 0.9897, + "step": 13137 + }, + { + "epoch": 2.339031339031339, + "grad_norm": 0.8140097260475159, + "learning_rate": 7.382527036967614e-05, + "loss": 0.7815, + "step": 13138 + }, + { + "epoch": 2.3392094017094016, + "grad_norm": 0.8154003620147705, + "learning_rate": 7.38117611556452e-05, + "loss": 0.7646, + "step": 13139 + }, + { + "epoch": 2.3393874643874644, + "grad_norm": 0.7705643177032471, + "learning_rate": 7.379825245471836e-05, + "loss": 0.7633, + "step": 13140 + }, + { + "epoch": 2.339565527065527, + "grad_norm": 0.7856985330581665, + "learning_rate": 7.378474426716035e-05, + "loss": 0.6803, + "step": 13141 + }, + { + "epoch": 2.33974358974359, + "grad_norm": 0.8384547233581543, + "learning_rate": 7.377123659323579e-05, + "loss": 0.9092, + "step": 13142 + }, + { + "epoch": 2.339921652421652, + "grad_norm": 0.7456032633781433, + "learning_rate": 7.375772943320942e-05, + "loss": 0.8393, + "step": 13143 + }, + { + "epoch": 2.340099715099715, + "grad_norm": 0.9527342319488525, + "learning_rate": 7.374422278734579e-05, + "loss": 1.0272, + "step": 13144 + }, + { + "epoch": 2.3402777777777777, + "grad_norm": 0.8976300954818726, + "learning_rate": 7.37307166559096e-05, + "loss": 0.7184, + "step": 13145 + }, + { + "epoch": 2.3404558404558404, + "grad_norm": 0.7698291540145874, + "learning_rate": 7.371721103916542e-05, + "loss": 0.6783, + "step": 13146 + }, + { + "epoch": 2.340633903133903, + "grad_norm": 0.8646810054779053, + "learning_rate": 7.37037059373779e-05, + "loss": 0.8559, + "step": 13147 + }, + { + "epoch": 2.340811965811966, + "grad_norm": 0.7534750699996948, + "learning_rate": 7.369020135081161e-05, + "loss": 0.8087, + "step": 13148 + }, + { + "epoch": 2.3409900284900287, + "grad_norm": 0.7408546209335327, + "learning_rate": 7.367669727973123e-05, + "loss": 0.675, + "step": 13149 + }, + { + "epoch": 2.341168091168091, + "grad_norm": 0.8753145933151245, + "learning_rate": 7.366319372440124e-05, + "loss": 0.8163, + "step": 13150 + }, + { + "epoch": 2.3413461538461537, + "grad_norm": 0.7065265774726868, + "learning_rate": 7.364969068508624e-05, + "loss": 0.5786, + "step": 13151 + }, + { + "epoch": 2.3415242165242165, + "grad_norm": 0.7976117730140686, + "learning_rate": 7.363618816205087e-05, + "loss": 0.9053, + "step": 13152 + }, + { + "epoch": 2.3417022792022792, + "grad_norm": 0.7261707782745361, + "learning_rate": 7.362268615555958e-05, + "loss": 0.7677, + "step": 13153 + }, + { + "epoch": 2.341880341880342, + "grad_norm": 0.7868889570236206, + "learning_rate": 7.360918466587701e-05, + "loss": 0.8648, + "step": 13154 + }, + { + "epoch": 2.3420584045584047, + "grad_norm": 0.8473666310310364, + "learning_rate": 7.35956836932676e-05, + "loss": 0.8533, + "step": 13155 + }, + { + "epoch": 2.342236467236467, + "grad_norm": 0.7456569671630859, + "learning_rate": 7.358218323799594e-05, + "loss": 0.7617, + "step": 13156 + }, + { + "epoch": 2.34241452991453, + "grad_norm": 0.8130928874015808, + "learning_rate": 7.356868330032652e-05, + "loss": 0.8667, + "step": 13157 + }, + { + "epoch": 2.3425925925925926, + "grad_norm": 0.8743309378623962, + "learning_rate": 7.355518388052384e-05, + "loss": 0.9196, + "step": 13158 + }, + { + "epoch": 2.3427706552706553, + "grad_norm": 0.8228809237480164, + "learning_rate": 7.354168497885237e-05, + "loss": 0.6509, + "step": 13159 + }, + { + "epoch": 2.342948717948718, + "grad_norm": 0.6998807191848755, + "learning_rate": 7.352818659557668e-05, + "loss": 0.5762, + "step": 13160 + }, + { + "epoch": 2.343126780626781, + "grad_norm": 0.8757675290107727, + "learning_rate": 7.351468873096114e-05, + "loss": 0.9094, + "step": 13161 + }, + { + "epoch": 2.343304843304843, + "grad_norm": 0.7495744824409485, + "learning_rate": 7.350119138527026e-05, + "loss": 0.653, + "step": 13162 + }, + { + "epoch": 2.343482905982906, + "grad_norm": 0.8229764103889465, + "learning_rate": 7.348769455876849e-05, + "loss": 0.8146, + "step": 13163 + }, + { + "epoch": 2.3436609686609686, + "grad_norm": 0.8317791819572449, + "learning_rate": 7.347419825172029e-05, + "loss": 0.7754, + "step": 13164 + }, + { + "epoch": 2.3438390313390314, + "grad_norm": 0.8210344910621643, + "learning_rate": 7.346070246439005e-05, + "loss": 0.876, + "step": 13165 + }, + { + "epoch": 2.344017094017094, + "grad_norm": 0.7711526155471802, + "learning_rate": 7.344720719704223e-05, + "loss": 0.7426, + "step": 13166 + }, + { + "epoch": 2.344195156695157, + "grad_norm": 0.8231741189956665, + "learning_rate": 7.343371244994119e-05, + "loss": 0.8992, + "step": 13167 + }, + { + "epoch": 2.344373219373219, + "grad_norm": 0.7145521640777588, + "learning_rate": 7.342021822335143e-05, + "loss": 0.8787, + "step": 13168 + }, + { + "epoch": 2.344551282051282, + "grad_norm": 0.8323171734809875, + "learning_rate": 7.340672451753723e-05, + "loss": 0.7035, + "step": 13169 + }, + { + "epoch": 2.3447293447293447, + "grad_norm": 0.7061881422996521, + "learning_rate": 7.339323133276301e-05, + "loss": 0.8077, + "step": 13170 + }, + { + "epoch": 2.3449074074074074, + "grad_norm": 0.8705938458442688, + "learning_rate": 7.33797386692932e-05, + "loss": 0.8616, + "step": 13171 + }, + { + "epoch": 2.34508547008547, + "grad_norm": 0.8777729868888855, + "learning_rate": 7.336624652739208e-05, + "loss": 0.9524, + "step": 13172 + }, + { + "epoch": 2.345263532763533, + "grad_norm": 0.9099276065826416, + "learning_rate": 7.335275490732406e-05, + "loss": 0.8248, + "step": 13173 + }, + { + "epoch": 2.3454415954415953, + "grad_norm": 0.7963444590568542, + "learning_rate": 7.333926380935341e-05, + "loss": 0.78, + "step": 13174 + }, + { + "epoch": 2.345619658119658, + "grad_norm": 0.9400636553764343, + "learning_rate": 7.332577323374454e-05, + "loss": 1.0062, + "step": 13175 + }, + { + "epoch": 2.3457977207977208, + "grad_norm": 0.7794054746627808, + "learning_rate": 7.331228318076171e-05, + "loss": 0.8564, + "step": 13176 + }, + { + "epoch": 2.3459757834757835, + "grad_norm": 0.7767263054847717, + "learning_rate": 7.329879365066927e-05, + "loss": 0.8191, + "step": 13177 + }, + { + "epoch": 2.3461538461538463, + "grad_norm": 0.7170942425727844, + "learning_rate": 7.328530464373148e-05, + "loss": 0.7018, + "step": 13178 + }, + { + "epoch": 2.346331908831909, + "grad_norm": 0.8246886134147644, + "learning_rate": 7.327181616021268e-05, + "loss": 0.8498, + "step": 13179 + }, + { + "epoch": 2.3465099715099713, + "grad_norm": 0.9531362652778625, + "learning_rate": 7.325832820037711e-05, + "loss": 0.7031, + "step": 13180 + }, + { + "epoch": 2.346688034188034, + "grad_norm": 0.8561878204345703, + "learning_rate": 7.324484076448905e-05, + "loss": 0.8627, + "step": 13181 + }, + { + "epoch": 2.346866096866097, + "grad_norm": 0.7890949845314026, + "learning_rate": 7.323135385281274e-05, + "loss": 0.7675, + "step": 13182 + }, + { + "epoch": 2.3470441595441596, + "grad_norm": 0.72523033618927, + "learning_rate": 7.321786746561246e-05, + "loss": 0.8847, + "step": 13183 + }, + { + "epoch": 2.3472222222222223, + "grad_norm": 0.7866469025611877, + "learning_rate": 7.32043816031524e-05, + "loss": 0.9657, + "step": 13184 + }, + { + "epoch": 2.347400284900285, + "grad_norm": 0.8669828176498413, + "learning_rate": 7.319089626569687e-05, + "loss": 0.9098, + "step": 13185 + }, + { + "epoch": 2.347578347578348, + "grad_norm": 0.7874458432197571, + "learning_rate": 7.317741145351e-05, + "loss": 0.9545, + "step": 13186 + }, + { + "epoch": 2.34775641025641, + "grad_norm": 0.7924689054489136, + "learning_rate": 7.316392716685604e-05, + "loss": 0.8577, + "step": 13187 + }, + { + "epoch": 2.347934472934473, + "grad_norm": 0.731119692325592, + "learning_rate": 7.315044340599918e-05, + "loss": 0.9251, + "step": 13188 + }, + { + "epoch": 2.3481125356125356, + "grad_norm": 0.914900004863739, + "learning_rate": 7.313696017120361e-05, + "loss": 0.9224, + "step": 13189 + }, + { + "epoch": 2.3482905982905984, + "grad_norm": 0.7616490125656128, + "learning_rate": 7.312347746273349e-05, + "loss": 0.7263, + "step": 13190 + }, + { + "epoch": 2.348468660968661, + "grad_norm": 0.8357210159301758, + "learning_rate": 7.310999528085301e-05, + "loss": 0.8572, + "step": 13191 + }, + { + "epoch": 2.3486467236467234, + "grad_norm": 0.8404232263565063, + "learning_rate": 7.309651362582633e-05, + "loss": 0.6822, + "step": 13192 + }, + { + "epoch": 2.348824786324786, + "grad_norm": 0.8992070555686951, + "learning_rate": 7.308303249791754e-05, + "loss": 0.91, + "step": 13193 + }, + { + "epoch": 2.349002849002849, + "grad_norm": 0.8150524497032166, + "learning_rate": 7.306955189739084e-05, + "loss": 0.984, + "step": 13194 + }, + { + "epoch": 2.3491809116809117, + "grad_norm": 0.9042861461639404, + "learning_rate": 7.305607182451031e-05, + "loss": 1.0111, + "step": 13195 + }, + { + "epoch": 2.3493589743589745, + "grad_norm": 0.8402968049049377, + "learning_rate": 7.30425922795401e-05, + "loss": 0.801, + "step": 13196 + }, + { + "epoch": 2.349537037037037, + "grad_norm": 0.7742997407913208, + "learning_rate": 7.302911326274428e-05, + "loss": 0.659, + "step": 13197 + }, + { + "epoch": 2.3497150997151, + "grad_norm": 0.8005271553993225, + "learning_rate": 7.301563477438698e-05, + "loss": 0.8549, + "step": 13198 + }, + { + "epoch": 2.3498931623931623, + "grad_norm": 0.8253805637359619, + "learning_rate": 7.300215681473224e-05, + "loss": 0.9049, + "step": 13199 + }, + { + "epoch": 2.350071225071225, + "grad_norm": 0.8539033532142639, + "learning_rate": 7.29886793840442e-05, + "loss": 0.9359, + "step": 13200 + }, + { + "epoch": 2.3502492877492878, + "grad_norm": 0.827608048915863, + "learning_rate": 7.297520248258681e-05, + "loss": 1.0105, + "step": 13201 + }, + { + "epoch": 2.3504273504273505, + "grad_norm": 0.8418487310409546, + "learning_rate": 7.296172611062422e-05, + "loss": 0.8138, + "step": 13202 + }, + { + "epoch": 2.3506054131054133, + "grad_norm": 0.7853255867958069, + "learning_rate": 7.294825026842042e-05, + "loss": 0.9279, + "step": 13203 + }, + { + "epoch": 2.3507834757834756, + "grad_norm": 0.8454880714416504, + "learning_rate": 7.293477495623951e-05, + "loss": 0.7687, + "step": 13204 + }, + { + "epoch": 2.3509615384615383, + "grad_norm": 0.7620453238487244, + "learning_rate": 7.29213001743454e-05, + "loss": 0.7567, + "step": 13205 + }, + { + "epoch": 2.351139601139601, + "grad_norm": 0.8993792533874512, + "learning_rate": 7.290782592300223e-05, + "loss": 0.9716, + "step": 13206 + }, + { + "epoch": 2.351317663817664, + "grad_norm": 1.1063668727874756, + "learning_rate": 7.289435220247387e-05, + "loss": 0.9763, + "step": 13207 + }, + { + "epoch": 2.3514957264957266, + "grad_norm": 0.8205364346504211, + "learning_rate": 7.288087901302439e-05, + "loss": 0.9395, + "step": 13208 + }, + { + "epoch": 2.3516737891737893, + "grad_norm": 0.680487871170044, + "learning_rate": 7.286740635491774e-05, + "loss": 0.6252, + "step": 13209 + }, + { + "epoch": 2.351851851851852, + "grad_norm": 0.8450767397880554, + "learning_rate": 7.285393422841791e-05, + "loss": 0.8707, + "step": 13210 + }, + { + "epoch": 2.3520299145299144, + "grad_norm": 0.6871187686920166, + "learning_rate": 7.284046263378888e-05, + "loss": 0.5695, + "step": 13211 + }, + { + "epoch": 2.352207977207977, + "grad_norm": 0.7968555688858032, + "learning_rate": 7.282699157129451e-05, + "loss": 0.7014, + "step": 13212 + }, + { + "epoch": 2.35238603988604, + "grad_norm": 0.863798201084137, + "learning_rate": 7.281352104119883e-05, + "loss": 0.9241, + "step": 13213 + }, + { + "epoch": 2.3525641025641026, + "grad_norm": 0.8848825693130493, + "learning_rate": 7.28000510437657e-05, + "loss": 0.8252, + "step": 13214 + }, + { + "epoch": 2.3527421652421654, + "grad_norm": 0.7528855800628662, + "learning_rate": 7.278658157925912e-05, + "loss": 0.7428, + "step": 13215 + }, + { + "epoch": 2.3529202279202277, + "grad_norm": 0.7636159062385559, + "learning_rate": 7.277311264794288e-05, + "loss": 0.8952, + "step": 13216 + }, + { + "epoch": 2.3530982905982905, + "grad_norm": 1.0585514307022095, + "learning_rate": 7.2759644250081e-05, + "loss": 0.9153, + "step": 13217 + }, + { + "epoch": 2.353276353276353, + "grad_norm": 0.7691277265548706, + "learning_rate": 7.274617638593725e-05, + "loss": 1.057, + "step": 13218 + }, + { + "epoch": 2.353454415954416, + "grad_norm": 0.8324813842773438, + "learning_rate": 7.273270905577561e-05, + "loss": 0.9253, + "step": 13219 + }, + { + "epoch": 2.3536324786324787, + "grad_norm": 0.835491418838501, + "learning_rate": 7.271924225985984e-05, + "loss": 1.0103, + "step": 13220 + }, + { + "epoch": 2.3538105413105415, + "grad_norm": 0.8318347930908203, + "learning_rate": 7.270577599845389e-05, + "loss": 0.8896, + "step": 13221 + }, + { + "epoch": 2.353988603988604, + "grad_norm": 0.7801460027694702, + "learning_rate": 7.269231027182153e-05, + "loss": 0.9274, + "step": 13222 + }, + { + "epoch": 2.3541666666666665, + "grad_norm": 0.8195397257804871, + "learning_rate": 7.267884508022665e-05, + "loss": 0.8126, + "step": 13223 + }, + { + "epoch": 2.3543447293447293, + "grad_norm": 0.7978246212005615, + "learning_rate": 7.2665380423933e-05, + "loss": 0.8426, + "step": 13224 + }, + { + "epoch": 2.354522792022792, + "grad_norm": 0.7614684104919434, + "learning_rate": 7.265191630320452e-05, + "loss": 0.7277, + "step": 13225 + }, + { + "epoch": 2.3547008547008548, + "grad_norm": 0.8684967756271362, + "learning_rate": 7.263845271830485e-05, + "loss": 0.7054, + "step": 13226 + }, + { + "epoch": 2.3548789173789175, + "grad_norm": 0.878842830657959, + "learning_rate": 7.262498966949791e-05, + "loss": 1.0478, + "step": 13227 + }, + { + "epoch": 2.35505698005698, + "grad_norm": 0.8321235179901123, + "learning_rate": 7.26115271570474e-05, + "loss": 0.8119, + "step": 13228 + }, + { + "epoch": 2.3552350427350426, + "grad_norm": 0.9144030213356018, + "learning_rate": 7.259806518121713e-05, + "loss": 0.8626, + "step": 13229 + }, + { + "epoch": 2.3554131054131053, + "grad_norm": 0.8437082767486572, + "learning_rate": 7.258460374227085e-05, + "loss": 0.8439, + "step": 13230 + }, + { + "epoch": 2.355591168091168, + "grad_norm": 0.8405697345733643, + "learning_rate": 7.257114284047229e-05, + "loss": 0.863, + "step": 13231 + }, + { + "epoch": 2.355769230769231, + "grad_norm": 0.8674731850624084, + "learning_rate": 7.255768247608525e-05, + "loss": 0.9823, + "step": 13232 + }, + { + "epoch": 2.3559472934472936, + "grad_norm": 0.8844531178474426, + "learning_rate": 7.254422264937337e-05, + "loss": 0.9018, + "step": 13233 + }, + { + "epoch": 2.3561253561253563, + "grad_norm": 0.8394746780395508, + "learning_rate": 7.253076336060045e-05, + "loss": 0.8407, + "step": 13234 + }, + { + "epoch": 2.3563034188034186, + "grad_norm": 0.8759872317314148, + "learning_rate": 7.251730461003012e-05, + "loss": 0.841, + "step": 13235 + }, + { + "epoch": 2.3564814814814814, + "grad_norm": 0.7240089774131775, + "learning_rate": 7.250384639792617e-05, + "loss": 0.7918, + "step": 13236 + }, + { + "epoch": 2.356659544159544, + "grad_norm": 0.8619599342346191, + "learning_rate": 7.24903887245522e-05, + "loss": 0.761, + "step": 13237 + }, + { + "epoch": 2.356837606837607, + "grad_norm": 0.7291443943977356, + "learning_rate": 7.247693159017192e-05, + "loss": 0.8189, + "step": 13238 + }, + { + "epoch": 2.3570156695156697, + "grad_norm": 0.8006066083908081, + "learning_rate": 7.246347499504898e-05, + "loss": 0.8924, + "step": 13239 + }, + { + "epoch": 2.357193732193732, + "grad_norm": 0.7774627208709717, + "learning_rate": 7.245001893944707e-05, + "loss": 0.946, + "step": 13240 + }, + { + "epoch": 2.3573717948717947, + "grad_norm": 0.7643784284591675, + "learning_rate": 7.243656342362978e-05, + "loss": 0.8717, + "step": 13241 + }, + { + "epoch": 2.3575498575498575, + "grad_norm": 0.7197792530059814, + "learning_rate": 7.242310844786082e-05, + "loss": 0.6792, + "step": 13242 + }, + { + "epoch": 2.35772792022792, + "grad_norm": 0.9124938249588013, + "learning_rate": 7.240965401240371e-05, + "loss": 0.841, + "step": 13243 + }, + { + "epoch": 2.357905982905983, + "grad_norm": 0.7350388765335083, + "learning_rate": 7.239620011752215e-05, + "loss": 0.8294, + "step": 13244 + }, + { + "epoch": 2.3580840455840457, + "grad_norm": 0.8814936280250549, + "learning_rate": 7.238274676347967e-05, + "loss": 0.9732, + "step": 13245 + }, + { + "epoch": 2.3582621082621085, + "grad_norm": 0.8379302024841309, + "learning_rate": 7.236929395053995e-05, + "loss": 0.8896, + "step": 13246 + }, + { + "epoch": 2.3584401709401708, + "grad_norm": 0.8200546503067017, + "learning_rate": 7.235584167896648e-05, + "loss": 0.7991, + "step": 13247 + }, + { + "epoch": 2.3586182336182335, + "grad_norm": 0.7842608690261841, + "learning_rate": 7.234238994902287e-05, + "loss": 0.695, + "step": 13248 + }, + { + "epoch": 2.3587962962962963, + "grad_norm": 0.8872218132019043, + "learning_rate": 7.232893876097266e-05, + "loss": 0.8611, + "step": 13249 + }, + { + "epoch": 2.358974358974359, + "grad_norm": 0.8358500599861145, + "learning_rate": 7.231548811507942e-05, + "loss": 0.7829, + "step": 13250 + }, + { + "epoch": 2.359152421652422, + "grad_norm": 0.8269400000572205, + "learning_rate": 7.23020380116067e-05, + "loss": 0.6904, + "step": 13251 + }, + { + "epoch": 2.359330484330484, + "grad_norm": 0.8693541288375854, + "learning_rate": 7.2288588450818e-05, + "loss": 0.8659, + "step": 13252 + }, + { + "epoch": 2.359508547008547, + "grad_norm": 0.858076810836792, + "learning_rate": 7.227513943297688e-05, + "loss": 0.8824, + "step": 13253 + }, + { + "epoch": 2.3596866096866096, + "grad_norm": 0.796541154384613, + "learning_rate": 7.226169095834675e-05, + "loss": 0.8999, + "step": 13254 + }, + { + "epoch": 2.3598646723646723, + "grad_norm": 0.7692779898643494, + "learning_rate": 7.22482430271912e-05, + "loss": 0.9492, + "step": 13255 + }, + { + "epoch": 2.360042735042735, + "grad_norm": 0.9259434342384338, + "learning_rate": 7.223479563977364e-05, + "loss": 0.9115, + "step": 13256 + }, + { + "epoch": 2.360220797720798, + "grad_norm": 0.9048989415168762, + "learning_rate": 7.222134879635764e-05, + "loss": 0.8057, + "step": 13257 + }, + { + "epoch": 2.3603988603988606, + "grad_norm": 0.9342616200447083, + "learning_rate": 7.220790249720656e-05, + "loss": 0.8554, + "step": 13258 + }, + { + "epoch": 2.360576923076923, + "grad_norm": 0.7747787237167358, + "learning_rate": 7.219445674258392e-05, + "loss": 0.7555, + "step": 13259 + }, + { + "epoch": 2.3607549857549857, + "grad_norm": 0.805437445640564, + "learning_rate": 7.218101153275311e-05, + "loss": 0.7442, + "step": 13260 + }, + { + "epoch": 2.3609330484330484, + "grad_norm": 0.9797805547714233, + "learning_rate": 7.216756686797764e-05, + "loss": 1.0975, + "step": 13261 + }, + { + "epoch": 2.361111111111111, + "grad_norm": 0.7361458539962769, + "learning_rate": 7.215412274852083e-05, + "loss": 0.6597, + "step": 13262 + }, + { + "epoch": 2.361289173789174, + "grad_norm": 0.8041569590568542, + "learning_rate": 7.21406791746462e-05, + "loss": 0.8343, + "step": 13263 + }, + { + "epoch": 2.361467236467236, + "grad_norm": 0.8364384770393372, + "learning_rate": 7.212723614661703e-05, + "loss": 0.9486, + "step": 13264 + }, + { + "epoch": 2.361645299145299, + "grad_norm": 0.714241623878479, + "learning_rate": 7.21137936646968e-05, + "loss": 0.5978, + "step": 13265 + }, + { + "epoch": 2.3618233618233617, + "grad_norm": 0.8830710053443909, + "learning_rate": 7.210035172914882e-05, + "loss": 0.9584, + "step": 13266 + }, + { + "epoch": 2.3620014245014245, + "grad_norm": 0.714112401008606, + "learning_rate": 7.208691034023653e-05, + "loss": 0.8878, + "step": 13267 + }, + { + "epoch": 2.3621794871794872, + "grad_norm": 0.7654083371162415, + "learning_rate": 7.207346949822322e-05, + "loss": 0.822, + "step": 13268 + }, + { + "epoch": 2.36235754985755, + "grad_norm": 0.772693395614624, + "learning_rate": 7.206002920337225e-05, + "loss": 0.7993, + "step": 13269 + }, + { + "epoch": 2.3625356125356127, + "grad_norm": 0.9678596258163452, + "learning_rate": 7.2046589455947e-05, + "loss": 0.948, + "step": 13270 + }, + { + "epoch": 2.362713675213675, + "grad_norm": 0.8254278302192688, + "learning_rate": 7.203315025621073e-05, + "loss": 0.8654, + "step": 13271 + }, + { + "epoch": 2.362891737891738, + "grad_norm": 0.7527315020561218, + "learning_rate": 7.201971160442685e-05, + "loss": 0.6881, + "step": 13272 + }, + { + "epoch": 2.3630698005698005, + "grad_norm": 0.7658267021179199, + "learning_rate": 7.200627350085853e-05, + "loss": 0.7332, + "step": 13273 + }, + { + "epoch": 2.3632478632478633, + "grad_norm": 0.8590806126594543, + "learning_rate": 7.199283594576916e-05, + "loss": 0.879, + "step": 13274 + }, + { + "epoch": 2.363425925925926, + "grad_norm": 0.7533347606658936, + "learning_rate": 7.197939893942197e-05, + "loss": 0.8738, + "step": 13275 + }, + { + "epoch": 2.363603988603989, + "grad_norm": Infinity, + "learning_rate": 7.197939893942197e-05, + "loss": 0.7641, + "step": 13276 + }, + { + "epoch": 2.363782051282051, + "grad_norm": 0.6873685121536255, + "learning_rate": 7.196596248208029e-05, + "loss": 0.4708, + "step": 13277 + }, + { + "epoch": 2.363960113960114, + "grad_norm": 0.7659112215042114, + "learning_rate": 7.195252657400729e-05, + "loss": 0.839, + "step": 13278 + }, + { + "epoch": 2.3641381766381766, + "grad_norm": 0.8355028629302979, + "learning_rate": 7.193909121546631e-05, + "loss": 0.9792, + "step": 13279 + }, + { + "epoch": 2.3643162393162394, + "grad_norm": 0.9633997678756714, + "learning_rate": 7.192565640672052e-05, + "loss": 0.9891, + "step": 13280 + }, + { + "epoch": 2.364494301994302, + "grad_norm": 0.7984298467636108, + "learning_rate": 7.191222214803318e-05, + "loss": 0.8343, + "step": 13281 + }, + { + "epoch": 2.364672364672365, + "grad_norm": 0.8239994645118713, + "learning_rate": 7.189878843966749e-05, + "loss": 0.8586, + "step": 13282 + }, + { + "epoch": 2.364850427350427, + "grad_norm": 0.8695420026779175, + "learning_rate": 7.188535528188671e-05, + "loss": 0.9161, + "step": 13283 + }, + { + "epoch": 2.36502849002849, + "grad_norm": 0.8272924423217773, + "learning_rate": 7.187192267495393e-05, + "loss": 0.8158, + "step": 13284 + }, + { + "epoch": 2.3652065527065527, + "grad_norm": 0.8217222690582275, + "learning_rate": 7.185849061913243e-05, + "loss": 0.892, + "step": 13285 + }, + { + "epoch": 2.3653846153846154, + "grad_norm": 0.9041243195533752, + "learning_rate": 7.184505911468532e-05, + "loss": 0.9093, + "step": 13286 + }, + { + "epoch": 2.365562678062678, + "grad_norm": 0.8325521349906921, + "learning_rate": 7.183162816187582e-05, + "loss": 0.7546, + "step": 13287 + }, + { + "epoch": 2.365740740740741, + "grad_norm": 0.9160267114639282, + "learning_rate": 7.181819776096704e-05, + "loss": 0.9662, + "step": 13288 + }, + { + "epoch": 2.3659188034188032, + "grad_norm": 0.8771381974220276, + "learning_rate": 7.180476791222215e-05, + "loss": 1.0083, + "step": 13289 + }, + { + "epoch": 2.366096866096866, + "grad_norm": 0.8251327872276306, + "learning_rate": 7.179133861590421e-05, + "loss": 0.8209, + "step": 13290 + }, + { + "epoch": 2.3662749287749287, + "grad_norm": 0.8760706186294556, + "learning_rate": 7.177790987227641e-05, + "loss": 0.7479, + "step": 13291 + }, + { + "epoch": 2.3664529914529915, + "grad_norm": 0.7857288122177124, + "learning_rate": 7.176448168160187e-05, + "loss": 0.6511, + "step": 13292 + }, + { + "epoch": 2.3666310541310542, + "grad_norm": 0.9548102021217346, + "learning_rate": 7.175105404414362e-05, + "loss": 0.731, + "step": 13293 + }, + { + "epoch": 2.366809116809117, + "grad_norm": 0.7604304552078247, + "learning_rate": 7.173762696016484e-05, + "loss": 0.8212, + "step": 13294 + }, + { + "epoch": 2.3669871794871793, + "grad_norm": 0.9121061563491821, + "learning_rate": 7.172420042992849e-05, + "loss": 0.939, + "step": 13295 + }, + { + "epoch": 2.367165242165242, + "grad_norm": 0.8128613233566284, + "learning_rate": 7.171077445369772e-05, + "loss": 0.8908, + "step": 13296 + }, + { + "epoch": 2.367343304843305, + "grad_norm": 0.9184401035308838, + "learning_rate": 7.169734903173555e-05, + "loss": 0.957, + "step": 13297 + }, + { + "epoch": 2.3675213675213675, + "grad_norm": 0.9234427809715271, + "learning_rate": 7.168392416430507e-05, + "loss": 0.8403, + "step": 13298 + }, + { + "epoch": 2.3676994301994303, + "grad_norm": 0.8810806274414062, + "learning_rate": 7.167049985166922e-05, + "loss": 0.9754, + "step": 13299 + }, + { + "epoch": 2.367877492877493, + "grad_norm": 0.8208937048912048, + "learning_rate": 7.165707609409113e-05, + "loss": 0.9418, + "step": 13300 + }, + { + "epoch": 2.3680555555555554, + "grad_norm": 0.8666219711303711, + "learning_rate": 7.164365289183371e-05, + "loss": 0.8936, + "step": 13301 + }, + { + "epoch": 2.368233618233618, + "grad_norm": 0.9385154843330383, + "learning_rate": 7.163023024516002e-05, + "loss": 0.8158, + "step": 13302 + }, + { + "epoch": 2.368411680911681, + "grad_norm": 1.0415911674499512, + "learning_rate": 7.161680815433303e-05, + "loss": 1.0445, + "step": 13303 + }, + { + "epoch": 2.3685897435897436, + "grad_norm": 0.6882192492485046, + "learning_rate": 7.160338661961577e-05, + "loss": 0.4929, + "step": 13304 + }, + { + "epoch": 2.3687678062678064, + "grad_norm": 0.8695144653320312, + "learning_rate": 7.15899656412711e-05, + "loss": 0.8991, + "step": 13305 + }, + { + "epoch": 2.368945868945869, + "grad_norm": 0.8973569273948669, + "learning_rate": 7.157654521956206e-05, + "loss": 0.8423, + "step": 13306 + }, + { + "epoch": 2.369123931623932, + "grad_norm": 0.7656881213188171, + "learning_rate": 7.156312535475155e-05, + "loss": 0.7351, + "step": 13307 + }, + { + "epoch": 2.369301994301994, + "grad_norm": 0.8023402690887451, + "learning_rate": 7.154970604710258e-05, + "loss": 0.9943, + "step": 13308 + }, + { + "epoch": 2.369480056980057, + "grad_norm": 0.916946530342102, + "learning_rate": 7.153628729687797e-05, + "loss": 0.8649, + "step": 13309 + }, + { + "epoch": 2.3696581196581197, + "grad_norm": 0.8764750361442566, + "learning_rate": 7.152286910434068e-05, + "loss": 0.9799, + "step": 13310 + }, + { + "epoch": 2.3698361823361824, + "grad_norm": 0.8732671737670898, + "learning_rate": 7.150945146975364e-05, + "loss": 1.0431, + "step": 13311 + }, + { + "epoch": 2.370014245014245, + "grad_norm": 0.8447144031524658, + "learning_rate": 7.149603439337969e-05, + "loss": 0.7805, + "step": 13312 + }, + { + "epoch": 2.3701923076923075, + "grad_norm": 0.9017399549484253, + "learning_rate": 7.148261787548178e-05, + "loss": 0.8102, + "step": 13313 + }, + { + "epoch": 2.3703703703703702, + "grad_norm": 0.7187124490737915, + "learning_rate": 7.14692019163227e-05, + "loss": 0.7327, + "step": 13314 + }, + { + "epoch": 2.370548433048433, + "grad_norm": 0.8579949736595154, + "learning_rate": 7.145578651616536e-05, + "loss": 0.8685, + "step": 13315 + }, + { + "epoch": 2.3707264957264957, + "grad_norm": 0.6088887453079224, + "learning_rate": 7.144237167527256e-05, + "loss": 0.7004, + "step": 13316 + }, + { + "epoch": 2.3709045584045585, + "grad_norm": 0.6400231719017029, + "learning_rate": 7.142895739390718e-05, + "loss": 0.7273, + "step": 13317 + }, + { + "epoch": 2.3710826210826212, + "grad_norm": 0.8680049180984497, + "learning_rate": 7.141554367233201e-05, + "loss": 0.7886, + "step": 13318 + }, + { + "epoch": 2.371260683760684, + "grad_norm": 0.8894832134246826, + "learning_rate": 7.140213051080991e-05, + "loss": 1.0597, + "step": 13319 + }, + { + "epoch": 2.3714387464387463, + "grad_norm": 0.7371698021888733, + "learning_rate": 7.138871790960365e-05, + "loss": 0.8344, + "step": 13320 + }, + { + "epoch": 2.371616809116809, + "grad_norm": 0.7396906018257141, + "learning_rate": 7.137530586897601e-05, + "loss": 0.7185, + "step": 13321 + }, + { + "epoch": 2.371794871794872, + "grad_norm": 0.7884365320205688, + "learning_rate": 7.136189438918978e-05, + "loss": 0.8311, + "step": 13322 + }, + { + "epoch": 2.3719729344729346, + "grad_norm": 0.8064826130867004, + "learning_rate": 7.13484834705078e-05, + "loss": 0.6933, + "step": 13323 + }, + { + "epoch": 2.3721509971509973, + "grad_norm": 0.8865584135055542, + "learning_rate": 7.13350731131927e-05, + "loss": 0.979, + "step": 13324 + }, + { + "epoch": 2.3723290598290596, + "grad_norm": 0.7782325148582458, + "learning_rate": 7.132166331750736e-05, + "loss": 0.8147, + "step": 13325 + }, + { + "epoch": 2.3725071225071224, + "grad_norm": 0.8515480160713196, + "learning_rate": 7.13082540837144e-05, + "loss": 0.7571, + "step": 13326 + }, + { + "epoch": 2.372685185185185, + "grad_norm": 0.8665108680725098, + "learning_rate": 7.129484541207662e-05, + "loss": 0.8171, + "step": 13327 + }, + { + "epoch": 2.372863247863248, + "grad_norm": 0.7640653252601624, + "learning_rate": 7.128143730285668e-05, + "loss": 0.7118, + "step": 13328 + }, + { + "epoch": 2.3730413105413106, + "grad_norm": 0.844083309173584, + "learning_rate": 7.126802975631735e-05, + "loss": 0.8394, + "step": 13329 + }, + { + "epoch": 2.3732193732193734, + "grad_norm": 0.8718371391296387, + "learning_rate": 7.12546227727213e-05, + "loss": 0.8729, + "step": 13330 + }, + { + "epoch": 2.373397435897436, + "grad_norm": 0.7254782319068909, + "learning_rate": 7.124121635233118e-05, + "loss": 0.8178, + "step": 13331 + }, + { + "epoch": 2.3735754985754984, + "grad_norm": 0.7211804389953613, + "learning_rate": 7.12278104954097e-05, + "loss": 0.9415, + "step": 13332 + }, + { + "epoch": 2.373753561253561, + "grad_norm": 0.8538317680358887, + "learning_rate": 7.121440520221949e-05, + "loss": 0.8614, + "step": 13333 + }, + { + "epoch": 2.373931623931624, + "grad_norm": 0.8942680358886719, + "learning_rate": 7.120100047302324e-05, + "loss": 0.985, + "step": 13334 + }, + { + "epoch": 2.3741096866096867, + "grad_norm": 0.8282434344291687, + "learning_rate": 7.118759630808354e-05, + "loss": 0.94, + "step": 13335 + }, + { + "epoch": 2.3742877492877494, + "grad_norm": 0.8036409616470337, + "learning_rate": 7.117419270766308e-05, + "loss": 0.7145, + "step": 13336 + }, + { + "epoch": 2.3744658119658117, + "grad_norm": 0.9169675707817078, + "learning_rate": 7.116078967202437e-05, + "loss": 1.1078, + "step": 13337 + }, + { + "epoch": 2.3746438746438745, + "grad_norm": 0.7805418372154236, + "learning_rate": 7.114738720143011e-05, + "loss": 0.8216, + "step": 13338 + }, + { + "epoch": 2.3748219373219372, + "grad_norm": 1.0444506406784058, + "learning_rate": 7.113398529614285e-05, + "loss": 0.8153, + "step": 13339 + }, + { + "epoch": 2.375, + "grad_norm": 0.8254665732383728, + "learning_rate": 7.112058395642522e-05, + "loss": 0.8127, + "step": 13340 + }, + { + "epoch": 2.3751780626780628, + "grad_norm": 0.8327687382698059, + "learning_rate": 7.11071831825397e-05, + "loss": 0.7014, + "step": 13341 + }, + { + "epoch": 2.3753561253561255, + "grad_norm": 0.7473437786102295, + "learning_rate": 7.109378297474894e-05, + "loss": 0.7621, + "step": 13342 + }, + { + "epoch": 2.3755341880341883, + "grad_norm": 0.8537931442260742, + "learning_rate": 7.108038333331544e-05, + "loss": 0.9302, + "step": 13343 + }, + { + "epoch": 2.3757122507122506, + "grad_norm": 0.81959468126297, + "learning_rate": 7.106698425850178e-05, + "loss": 0.9157, + "step": 13344 + }, + { + "epoch": 2.3758903133903133, + "grad_norm": 0.769257128238678, + "learning_rate": 7.105358575057043e-05, + "loss": 0.8739, + "step": 13345 + }, + { + "epoch": 2.376068376068376, + "grad_norm": 0.7428072690963745, + "learning_rate": 7.104018780978394e-05, + "loss": 0.7001, + "step": 13346 + }, + { + "epoch": 2.376246438746439, + "grad_norm": 0.8152543306350708, + "learning_rate": 7.102679043640481e-05, + "loss": 0.9866, + "step": 13347 + }, + { + "epoch": 2.3764245014245016, + "grad_norm": 0.8732424974441528, + "learning_rate": 7.101339363069556e-05, + "loss": 1.0207, + "step": 13348 + }, + { + "epoch": 2.376602564102564, + "grad_norm": 0.759279191493988, + "learning_rate": 7.099999739291862e-05, + "loss": 0.8703, + "step": 13349 + }, + { + "epoch": 2.3767806267806266, + "grad_norm": 0.8751664161682129, + "learning_rate": 7.098660172333648e-05, + "loss": 0.9805, + "step": 13350 + }, + { + "epoch": 2.3769586894586894, + "grad_norm": 0.9646390080451965, + "learning_rate": 7.097320662221168e-05, + "loss": 0.8623, + "step": 13351 + }, + { + "epoch": 2.377136752136752, + "grad_norm": 0.8626869320869446, + "learning_rate": 7.095981208980652e-05, + "loss": 0.7175, + "step": 13352 + }, + { + "epoch": 2.377314814814815, + "grad_norm": 0.8075738549232483, + "learning_rate": 7.094641812638354e-05, + "loss": 0.7741, + "step": 13353 + }, + { + "epoch": 2.3774928774928776, + "grad_norm": 0.7733559608459473, + "learning_rate": 7.093302473220513e-05, + "loss": 0.8553, + "step": 13354 + }, + { + "epoch": 2.3776709401709404, + "grad_norm": 0.7372797727584839, + "learning_rate": 7.091963190753376e-05, + "loss": 0.8554, + "step": 13355 + }, + { + "epoch": 2.3778490028490027, + "grad_norm": 0.804649293422699, + "learning_rate": 7.090623965263177e-05, + "loss": 0.8704, + "step": 13356 + }, + { + "epoch": 2.3780270655270654, + "grad_norm": 0.8370727300643921, + "learning_rate": 7.089284796776157e-05, + "loss": 0.9786, + "step": 13357 + }, + { + "epoch": 2.378205128205128, + "grad_norm": 0.7565299272537231, + "learning_rate": 7.087945685318554e-05, + "loss": 0.8096, + "step": 13358 + }, + { + "epoch": 2.378383190883191, + "grad_norm": 0.9046086072921753, + "learning_rate": 7.086606630916611e-05, + "loss": 0.8108, + "step": 13359 + }, + { + "epoch": 2.3785612535612537, + "grad_norm": 0.8453067541122437, + "learning_rate": 7.085267633596552e-05, + "loss": 0.8226, + "step": 13360 + }, + { + "epoch": 2.378739316239316, + "grad_norm": 0.8499273061752319, + "learning_rate": 7.083928693384628e-05, + "loss": 1.001, + "step": 13361 + }, + { + "epoch": 2.3789173789173788, + "grad_norm": 0.8358726501464844, + "learning_rate": 7.082589810307055e-05, + "loss": 0.7891, + "step": 13362 + }, + { + "epoch": 2.3790954415954415, + "grad_norm": 0.9156573414802551, + "learning_rate": 7.081250984390078e-05, + "loss": 0.9381, + "step": 13363 + }, + { + "epoch": 2.3792735042735043, + "grad_norm": 0.8704338669776917, + "learning_rate": 7.079912215659923e-05, + "loss": 0.9004, + "step": 13364 + }, + { + "epoch": 2.379451566951567, + "grad_norm": 0.8201949000358582, + "learning_rate": 7.078573504142824e-05, + "loss": 0.7501, + "step": 13365 + }, + { + "epoch": 2.3796296296296298, + "grad_norm": 0.9453420639038086, + "learning_rate": 7.077234849865008e-05, + "loss": 0.9658, + "step": 13366 + }, + { + "epoch": 2.3798076923076925, + "grad_norm": 0.8556796908378601, + "learning_rate": 7.075896252852703e-05, + "loss": 0.8054, + "step": 13367 + }, + { + "epoch": 2.379985754985755, + "grad_norm": 0.7961027026176453, + "learning_rate": 7.074557713132136e-05, + "loss": 0.8065, + "step": 13368 + }, + { + "epoch": 2.3801638176638176, + "grad_norm": 0.8777903318405151, + "learning_rate": 7.073219230729533e-05, + "loss": 0.9399, + "step": 13369 + }, + { + "epoch": 2.3803418803418803, + "grad_norm": 0.8569813370704651, + "learning_rate": 7.071880805671123e-05, + "loss": 0.9424, + "step": 13370 + }, + { + "epoch": 2.380519943019943, + "grad_norm": 0.8810455203056335, + "learning_rate": 7.070542437983123e-05, + "loss": 1.1313, + "step": 13371 + }, + { + "epoch": 2.380698005698006, + "grad_norm": 0.8691363334655762, + "learning_rate": 7.069204127691761e-05, + "loss": 0.9114, + "step": 13372 + }, + { + "epoch": 2.380876068376068, + "grad_norm": 0.7922945618629456, + "learning_rate": 7.067865874823253e-05, + "loss": 0.9158, + "step": 13373 + }, + { + "epoch": 2.381054131054131, + "grad_norm": 0.7465389370918274, + "learning_rate": 7.066527679403825e-05, + "loss": 0.6597, + "step": 13374 + }, + { + "epoch": 2.3812321937321936, + "grad_norm": 0.8386009931564331, + "learning_rate": 7.065189541459689e-05, + "loss": 0.7194, + "step": 13375 + }, + { + "epoch": 2.3814102564102564, + "grad_norm": 0.8633689880371094, + "learning_rate": 7.063851461017073e-05, + "loss": 0.8877, + "step": 13376 + }, + { + "epoch": 2.381588319088319, + "grad_norm": 0.8689528107643127, + "learning_rate": 7.062513438102184e-05, + "loss": 0.8384, + "step": 13377 + }, + { + "epoch": 2.381766381766382, + "grad_norm": 0.7648544311523438, + "learning_rate": 7.061175472741243e-05, + "loss": 0.7669, + "step": 13378 + }, + { + "epoch": 2.3819444444444446, + "grad_norm": 0.8502510786056519, + "learning_rate": 7.059837564960465e-05, + "loss": 0.9379, + "step": 13379 + }, + { + "epoch": 2.382122507122507, + "grad_norm": 0.8277843594551086, + "learning_rate": 7.058499714786063e-05, + "loss": 0.7372, + "step": 13380 + }, + { + "epoch": 2.3823005698005697, + "grad_norm": 0.7394976615905762, + "learning_rate": 7.057161922244246e-05, + "loss": 0.7628, + "step": 13381 + }, + { + "epoch": 2.3824786324786325, + "grad_norm": 0.7906123399734497, + "learning_rate": 7.05582418736123e-05, + "loss": 0.7645, + "step": 13382 + }, + { + "epoch": 2.382656695156695, + "grad_norm": 0.7889885902404785, + "learning_rate": 7.054486510163221e-05, + "loss": 0.8316, + "step": 13383 + }, + { + "epoch": 2.382834757834758, + "grad_norm": 0.7983359098434448, + "learning_rate": 7.053148890676434e-05, + "loss": 0.7925, + "step": 13384 + }, + { + "epoch": 2.3830128205128207, + "grad_norm": 0.9067932963371277, + "learning_rate": 7.051811328927067e-05, + "loss": 0.9385, + "step": 13385 + }, + { + "epoch": 2.383190883190883, + "grad_norm": 0.7210679650306702, + "learning_rate": 7.05047382494134e-05, + "loss": 0.591, + "step": 13386 + }, + { + "epoch": 2.3833689458689458, + "grad_norm": 0.9977821707725525, + "learning_rate": 7.049136378745445e-05, + "loss": 0.8362, + "step": 13387 + }, + { + "epoch": 2.3835470085470085, + "grad_norm": 0.9260198473930359, + "learning_rate": 7.047798990365595e-05, + "loss": 1.0051, + "step": 13388 + }, + { + "epoch": 2.3837250712250713, + "grad_norm": 0.8903454542160034, + "learning_rate": 7.04646165982799e-05, + "loss": 0.7055, + "step": 13389 + }, + { + "epoch": 2.383903133903134, + "grad_norm": 0.9634504914283752, + "learning_rate": 7.045124387158832e-05, + "loss": 0.7681, + "step": 13390 + }, + { + "epoch": 2.3840811965811968, + "grad_norm": 0.8645864129066467, + "learning_rate": 7.043787172384329e-05, + "loss": 0.9271, + "step": 13391 + }, + { + "epoch": 2.384259259259259, + "grad_norm": 0.8738446235656738, + "learning_rate": 7.04245001553067e-05, + "loss": 0.9, + "step": 13392 + }, + { + "epoch": 2.384437321937322, + "grad_norm": 0.7869822382926941, + "learning_rate": 7.041112916624062e-05, + "loss": 0.8639, + "step": 13393 + }, + { + "epoch": 2.3846153846153846, + "grad_norm": 0.8728111386299133, + "learning_rate": 7.039775875690698e-05, + "loss": 1.0367, + "step": 13394 + }, + { + "epoch": 2.3847934472934473, + "grad_norm": 0.7883852124214172, + "learning_rate": 7.03843889275678e-05, + "loss": 0.8338, + "step": 13395 + }, + { + "epoch": 2.38497150997151, + "grad_norm": 0.9267113208770752, + "learning_rate": 7.037101967848496e-05, + "loss": 0.8931, + "step": 13396 + }, + { + "epoch": 2.385149572649573, + "grad_norm": 0.8940320611000061, + "learning_rate": 7.035765100992048e-05, + "loss": 0.8071, + "step": 13397 + }, + { + "epoch": 2.385327635327635, + "grad_norm": 0.8109263777732849, + "learning_rate": 7.03442829221362e-05, + "loss": 0.8083, + "step": 13398 + }, + { + "epoch": 2.385505698005698, + "grad_norm": 0.8223438262939453, + "learning_rate": 7.033091541539413e-05, + "loss": 0.9296, + "step": 13399 + }, + { + "epoch": 2.3856837606837606, + "grad_norm": 0.817894697189331, + "learning_rate": 7.031754848995612e-05, + "loss": 0.9168, + "step": 13400 + }, + { + "epoch": 2.3858618233618234, + "grad_norm": 0.831462562084198, + "learning_rate": 7.030418214608411e-05, + "loss": 0.8613, + "step": 13401 + }, + { + "epoch": 2.386039886039886, + "grad_norm": 0.8388770818710327, + "learning_rate": 7.029081638403994e-05, + "loss": 0.7477, + "step": 13402 + }, + { + "epoch": 2.386217948717949, + "grad_norm": 0.9557843804359436, + "learning_rate": 7.02774512040855e-05, + "loss": 0.8932, + "step": 13403 + }, + { + "epoch": 2.386396011396011, + "grad_norm": 0.8249707221984863, + "learning_rate": 7.026408660648268e-05, + "loss": 1.0301, + "step": 13404 + }, + { + "epoch": 2.386574074074074, + "grad_norm": 0.8355069160461426, + "learning_rate": 7.025072259149333e-05, + "loss": 0.8081, + "step": 13405 + }, + { + "epoch": 2.3867521367521367, + "grad_norm": 0.8373300433158875, + "learning_rate": 7.023735915937924e-05, + "loss": 0.9911, + "step": 13406 + }, + { + "epoch": 2.3869301994301995, + "grad_norm": 0.7177539467811584, + "learning_rate": 7.022399631040228e-05, + "loss": 0.6397, + "step": 13407 + }, + { + "epoch": 2.387108262108262, + "grad_norm": 0.7371904253959656, + "learning_rate": 7.021063404482426e-05, + "loss": 0.8634, + "step": 13408 + }, + { + "epoch": 2.387286324786325, + "grad_norm": 0.8919385671615601, + "learning_rate": 7.019727236290696e-05, + "loss": 0.9514, + "step": 13409 + }, + { + "epoch": 2.3874643874643873, + "grad_norm": 0.7673050761222839, + "learning_rate": 7.018391126491225e-05, + "loss": 0.8957, + "step": 13410 + }, + { + "epoch": 2.38764245014245, + "grad_norm": 0.8401889801025391, + "learning_rate": 7.01705507511018e-05, + "loss": 0.908, + "step": 13411 + }, + { + "epoch": 2.3878205128205128, + "grad_norm": 0.822903037071228, + "learning_rate": 7.01571908217375e-05, + "loss": 0.911, + "step": 13412 + }, + { + "epoch": 2.3879985754985755, + "grad_norm": 0.9824740290641785, + "learning_rate": 7.014383147708102e-05, + "loss": 0.8314, + "step": 13413 + }, + { + "epoch": 2.3881766381766383, + "grad_norm": 0.9485064148902893, + "learning_rate": 7.013047271739414e-05, + "loss": 0.9819, + "step": 13414 + }, + { + "epoch": 2.388354700854701, + "grad_norm": 0.7565387487411499, + "learning_rate": 7.01171145429386e-05, + "loss": 0.9702, + "step": 13415 + }, + { + "epoch": 2.388532763532764, + "grad_norm": 0.8159620761871338, + "learning_rate": 7.010375695397615e-05, + "loss": 0.7302, + "step": 13416 + }, + { + "epoch": 2.388710826210826, + "grad_norm": 0.7818536162376404, + "learning_rate": 7.009039995076844e-05, + "loss": 0.6821, + "step": 13417 + }, + { + "epoch": 2.388888888888889, + "grad_norm": 0.7958348989486694, + "learning_rate": 7.007704353357724e-05, + "loss": 0.7996, + "step": 13418 + }, + { + "epoch": 2.3890669515669516, + "grad_norm": 0.8097305297851562, + "learning_rate": 7.006368770266421e-05, + "loss": 0.8177, + "step": 13419 + }, + { + "epoch": 2.3892450142450143, + "grad_norm": 0.9326507449150085, + "learning_rate": 7.005033245829105e-05, + "loss": 0.9307, + "step": 13420 + }, + { + "epoch": 2.389423076923077, + "grad_norm": 0.8954049944877625, + "learning_rate": 7.003697780071936e-05, + "loss": 0.8527, + "step": 13421 + }, + { + "epoch": 2.3896011396011394, + "grad_norm": 0.890548586845398, + "learning_rate": 7.00236237302109e-05, + "loss": 0.8203, + "step": 13422 + }, + { + "epoch": 2.389779202279202, + "grad_norm": 0.7508596181869507, + "learning_rate": 7.001027024702722e-05, + "loss": 0.7056, + "step": 13423 + }, + { + "epoch": 2.389957264957265, + "grad_norm": 0.9403550624847412, + "learning_rate": 6.999691735143002e-05, + "loss": 0.7336, + "step": 13424 + }, + { + "epoch": 2.3901353276353277, + "grad_norm": 0.8187662959098816, + "learning_rate": 6.998356504368087e-05, + "loss": 0.6897, + "step": 13425 + }, + { + "epoch": 2.3903133903133904, + "grad_norm": 0.8584417104721069, + "learning_rate": 6.997021332404145e-05, + "loss": 0.9143, + "step": 13426 + }, + { + "epoch": 2.390491452991453, + "grad_norm": 0.8739892840385437, + "learning_rate": 6.995686219277329e-05, + "loss": 0.8028, + "step": 13427 + }, + { + "epoch": 2.390669515669516, + "grad_norm": 0.9291013479232788, + "learning_rate": 6.994351165013799e-05, + "loss": 1.0305, + "step": 13428 + }, + { + "epoch": 2.390847578347578, + "grad_norm": 0.7937391400337219, + "learning_rate": 6.993016169639719e-05, + "loss": 0.8326, + "step": 13429 + }, + { + "epoch": 2.391025641025641, + "grad_norm": 0.655261754989624, + "learning_rate": 6.991681233181236e-05, + "loss": 0.7939, + "step": 13430 + }, + { + "epoch": 2.3912037037037037, + "grad_norm": 0.9606142640113831, + "learning_rate": 6.990346355664515e-05, + "loss": 1.1344, + "step": 13431 + }, + { + "epoch": 2.3913817663817665, + "grad_norm": 0.8111617565155029, + "learning_rate": 6.9890115371157e-05, + "loss": 0.8398, + "step": 13432 + }, + { + "epoch": 2.3915598290598292, + "grad_norm": 0.8111898899078369, + "learning_rate": 6.987676777560955e-05, + "loss": 0.9189, + "step": 13433 + }, + { + "epoch": 2.3917378917378915, + "grad_norm": 0.7850473523139954, + "learning_rate": 6.98634207702642e-05, + "loss": 0.9563, + "step": 13434 + }, + { + "epoch": 2.3919159544159543, + "grad_norm": 0.7740257978439331, + "learning_rate": 6.985007435538256e-05, + "loss": 0.7446, + "step": 13435 + }, + { + "epoch": 2.392094017094017, + "grad_norm": 0.9354606866836548, + "learning_rate": 6.983672853122604e-05, + "loss": 0.879, + "step": 13436 + }, + { + "epoch": 2.39227207977208, + "grad_norm": 0.8909385800361633, + "learning_rate": 6.982338329805622e-05, + "loss": 0.9381, + "step": 13437 + }, + { + "epoch": 2.3924501424501425, + "grad_norm": 0.7748416066169739, + "learning_rate": 6.981003865613448e-05, + "loss": 0.7169, + "step": 13438 + }, + { + "epoch": 2.3926282051282053, + "grad_norm": 0.7357833981513977, + "learning_rate": 6.979669460572234e-05, + "loss": 0.669, + "step": 13439 + }, + { + "epoch": 2.392806267806268, + "grad_norm": 0.8370460271835327, + "learning_rate": 6.978335114708119e-05, + "loss": 0.6215, + "step": 13440 + }, + { + "epoch": 2.3929843304843303, + "grad_norm": 0.7578476071357727, + "learning_rate": 6.977000828047256e-05, + "loss": 0.871, + "step": 13441 + }, + { + "epoch": 2.393162393162393, + "grad_norm": 0.8111903071403503, + "learning_rate": 6.975666600615776e-05, + "loss": 0.7888, + "step": 13442 + }, + { + "epoch": 2.393340455840456, + "grad_norm": 0.9584433436393738, + "learning_rate": 6.974332432439831e-05, + "loss": 1.0011, + "step": 13443 + }, + { + "epoch": 2.3935185185185186, + "grad_norm": 0.9105294942855835, + "learning_rate": 6.972998323545555e-05, + "loss": 1.0832, + "step": 13444 + }, + { + "epoch": 2.3936965811965814, + "grad_norm": 0.7990328669548035, + "learning_rate": 6.971664273959089e-05, + "loss": 0.9561, + "step": 13445 + }, + { + "epoch": 2.3938746438746437, + "grad_norm": 0.8575631976127625, + "learning_rate": 6.970330283706569e-05, + "loss": 0.7965, + "step": 13446 + }, + { + "epoch": 2.3940527065527064, + "grad_norm": 0.8147784471511841, + "learning_rate": 6.968996352814139e-05, + "loss": 0.806, + "step": 13447 + }, + { + "epoch": 2.394230769230769, + "grad_norm": 0.8284323215484619, + "learning_rate": 6.967662481307923e-05, + "loss": 0.942, + "step": 13448 + }, + { + "epoch": 2.394408831908832, + "grad_norm": 0.8238104581832886, + "learning_rate": 6.966328669214062e-05, + "loss": 0.9163, + "step": 13449 + }, + { + "epoch": 2.3945868945868947, + "grad_norm": 0.8855763673782349, + "learning_rate": 6.964994916558692e-05, + "loss": 0.7683, + "step": 13450 + }, + { + "epoch": 2.3947649572649574, + "grad_norm": 1.02780020236969, + "learning_rate": 6.963661223367937e-05, + "loss": 0.904, + "step": 13451 + }, + { + "epoch": 2.39494301994302, + "grad_norm": 0.8001773953437805, + "learning_rate": 6.96232758966794e-05, + "loss": 0.8459, + "step": 13452 + }, + { + "epoch": 2.3951210826210825, + "grad_norm": 0.755388617515564, + "learning_rate": 6.960994015484818e-05, + "loss": 0.7759, + "step": 13453 + }, + { + "epoch": 2.3952991452991452, + "grad_norm": 0.7774340510368347, + "learning_rate": 6.959660500844708e-05, + "loss": 0.7353, + "step": 13454 + }, + { + "epoch": 2.395477207977208, + "grad_norm": 0.8696026802062988, + "learning_rate": 6.958327045773733e-05, + "loss": 0.635, + "step": 13455 + }, + { + "epoch": 2.3956552706552707, + "grad_norm": 0.8419780731201172, + "learning_rate": 6.956993650298025e-05, + "loss": 0.8515, + "step": 13456 + }, + { + "epoch": 2.3958333333333335, + "grad_norm": 0.9125590324401855, + "learning_rate": 6.955660314443699e-05, + "loss": 0.9099, + "step": 13457 + }, + { + "epoch": 2.396011396011396, + "grad_norm": 0.6847489476203918, + "learning_rate": 6.954327038236891e-05, + "loss": 0.6652, + "step": 13458 + }, + { + "epoch": 2.3961894586894585, + "grad_norm": 0.8674905896186829, + "learning_rate": 6.952993821703713e-05, + "loss": 0.7049, + "step": 13459 + }, + { + "epoch": 2.3963675213675213, + "grad_norm": 0.7777035236358643, + "learning_rate": 6.951660664870296e-05, + "loss": 0.818, + "step": 13460 + }, + { + "epoch": 2.396545584045584, + "grad_norm": 0.8349783420562744, + "learning_rate": 6.950327567762751e-05, + "loss": 0.9203, + "step": 13461 + }, + { + "epoch": 2.396723646723647, + "grad_norm": 0.7589834332466125, + "learning_rate": 6.948994530407206e-05, + "loss": 1.015, + "step": 13462 + }, + { + "epoch": 2.3969017094017095, + "grad_norm": 0.9340610504150391, + "learning_rate": 6.947661552829773e-05, + "loss": 1.0575, + "step": 13463 + }, + { + "epoch": 2.3970797720797723, + "grad_norm": 0.9100959300994873, + "learning_rate": 6.946328635056573e-05, + "loss": 0.8824, + "step": 13464 + }, + { + "epoch": 2.3972578347578346, + "grad_norm": 0.8255945444107056, + "learning_rate": 6.944995777113717e-05, + "loss": 0.7701, + "step": 13465 + }, + { + "epoch": 2.3974358974358974, + "grad_norm": 0.8572675585746765, + "learning_rate": 6.943662979027328e-05, + "loss": 0.9425, + "step": 13466 + }, + { + "epoch": 2.39761396011396, + "grad_norm": 0.8219536542892456, + "learning_rate": 6.94233024082351e-05, + "loss": 0.8184, + "step": 13467 + }, + { + "epoch": 2.397792022792023, + "grad_norm": 0.8260995149612427, + "learning_rate": 6.940997562528377e-05, + "loss": 0.8324, + "step": 13468 + }, + { + "epoch": 2.3979700854700856, + "grad_norm": 0.9707075357437134, + "learning_rate": 6.939664944168047e-05, + "loss": 0.9865, + "step": 13469 + }, + { + "epoch": 2.398148148148148, + "grad_norm": 0.9030438661575317, + "learning_rate": 6.938332385768622e-05, + "loss": 1.0244, + "step": 13470 + }, + { + "epoch": 2.3983262108262107, + "grad_norm": 0.8425108194351196, + "learning_rate": 6.936999887356214e-05, + "loss": 0.7053, + "step": 13471 + }, + { + "epoch": 2.3985042735042734, + "grad_norm": 1.0073270797729492, + "learning_rate": 6.93566744895693e-05, + "loss": 0.9324, + "step": 13472 + }, + { + "epoch": 2.398682336182336, + "grad_norm": 0.7647563219070435, + "learning_rate": 6.93433507059688e-05, + "loss": 0.7233, + "step": 13473 + }, + { + "epoch": 2.398860398860399, + "grad_norm": 0.7632454633712769, + "learning_rate": 6.933002752302162e-05, + "loss": 0.8678, + "step": 13474 + }, + { + "epoch": 2.3990384615384617, + "grad_norm": 0.7943702936172485, + "learning_rate": 6.931670494098887e-05, + "loss": 0.8805, + "step": 13475 + }, + { + "epoch": 2.3992165242165244, + "grad_norm": 0.9440419673919678, + "learning_rate": 6.930338296013153e-05, + "loss": 1.0103, + "step": 13476 + }, + { + "epoch": 2.3993945868945867, + "grad_norm": 0.9119253754615784, + "learning_rate": 6.929006158071065e-05, + "loss": 1.0235, + "step": 13477 + }, + { + "epoch": 2.3995726495726495, + "grad_norm": 0.7750248908996582, + "learning_rate": 6.927674080298721e-05, + "loss": 0.957, + "step": 13478 + }, + { + "epoch": 2.3997507122507122, + "grad_norm": 0.8847192525863647, + "learning_rate": 6.926342062722223e-05, + "loss": 0.9066, + "step": 13479 + }, + { + "epoch": 2.399928774928775, + "grad_norm": 0.814396321773529, + "learning_rate": 6.925010105367665e-05, + "loss": 1.0001, + "step": 13480 + }, + { + "epoch": 2.4001068376068377, + "grad_norm": 0.8323664665222168, + "learning_rate": 6.923678208261147e-05, + "loss": 1.0027, + "step": 13481 + }, + { + "epoch": 2.4002849002849, + "grad_norm": 0.8351104259490967, + "learning_rate": 6.92234637142876e-05, + "loss": 0.8629, + "step": 13482 + }, + { + "epoch": 2.400462962962963, + "grad_norm": 0.9298360347747803, + "learning_rate": 6.92101459489661e-05, + "loss": 0.9161, + "step": 13483 + }, + { + "epoch": 2.4006410256410255, + "grad_norm": 0.9423344135284424, + "learning_rate": 6.919682878690777e-05, + "loss": 1.4416, + "step": 13484 + }, + { + "epoch": 2.4008190883190883, + "grad_norm": 0.8340599536895752, + "learning_rate": 6.918351222837363e-05, + "loss": 0.8696, + "step": 13485 + }, + { + "epoch": 2.400997150997151, + "grad_norm": 0.8533751368522644, + "learning_rate": 6.917019627362451e-05, + "loss": 1.1383, + "step": 13486 + }, + { + "epoch": 2.401175213675214, + "grad_norm": 0.8060563206672668, + "learning_rate": 6.91568809229214e-05, + "loss": 0.8544, + "step": 13487 + }, + { + "epoch": 2.4013532763532766, + "grad_norm": 0.865485668182373, + "learning_rate": 6.914356617652511e-05, + "loss": 0.9286, + "step": 13488 + }, + { + "epoch": 2.401531339031339, + "grad_norm": 0.8785045742988586, + "learning_rate": 6.913025203469652e-05, + "loss": 0.7339, + "step": 13489 + }, + { + "epoch": 2.4017094017094016, + "grad_norm": 0.7718466520309448, + "learning_rate": 6.911693849769654e-05, + "loss": 0.8821, + "step": 13490 + }, + { + "epoch": 2.4018874643874644, + "grad_norm": 0.7274343371391296, + "learning_rate": 6.910362556578599e-05, + "loss": 0.6179, + "step": 13491 + }, + { + "epoch": 2.402065527065527, + "grad_norm": 0.8848530054092407, + "learning_rate": 6.909031323922574e-05, + "loss": 0.7848, + "step": 13492 + }, + { + "epoch": 2.40224358974359, + "grad_norm": 0.7384527325630188, + "learning_rate": 6.907700151827657e-05, + "loss": 0.5, + "step": 13493 + }, + { + "epoch": 2.402421652421652, + "grad_norm": 0.865505576133728, + "learning_rate": 6.906369040319936e-05, + "loss": 0.7127, + "step": 13494 + }, + { + "epoch": 2.402599715099715, + "grad_norm": 0.8588849902153015, + "learning_rate": 6.90503798942548e-05, + "loss": 0.8833, + "step": 13495 + }, + { + "epoch": 2.4027777777777777, + "grad_norm": 0.8570847511291504, + "learning_rate": 6.903706999170381e-05, + "loss": 0.9765, + "step": 13496 + }, + { + "epoch": 2.4029558404558404, + "grad_norm": 0.9193849563598633, + "learning_rate": 6.902376069580706e-05, + "loss": 0.8654, + "step": 13497 + }, + { + "epoch": 2.403133903133903, + "grad_norm": 0.8181582093238831, + "learning_rate": 6.901045200682545e-05, + "loss": 0.8815, + "step": 13498 + }, + { + "epoch": 2.403311965811966, + "grad_norm": 0.783163845539093, + "learning_rate": 6.89971439250196e-05, + "loss": 0.8383, + "step": 13499 + }, + { + "epoch": 2.4034900284900287, + "grad_norm": 1.0679216384887695, + "learning_rate": 6.898383645065032e-05, + "loss": 1.0525, + "step": 13500 + }, + { + "epoch": 2.403668091168091, + "grad_norm": 0.7945899367332458, + "learning_rate": 6.897052958397831e-05, + "loss": 1.0091, + "step": 13501 + }, + { + "epoch": 2.4038461538461537, + "grad_norm": 0.8310369253158569, + "learning_rate": 6.895722332526438e-05, + "loss": 0.8909, + "step": 13502 + }, + { + "epoch": 2.4040242165242165, + "grad_norm": 0.8811371922492981, + "learning_rate": 6.894391767476911e-05, + "loss": 0.8354, + "step": 13503 + }, + { + "epoch": 2.4042022792022792, + "grad_norm": 1.011495590209961, + "learning_rate": 6.893061263275332e-05, + "loss": 0.8846, + "step": 13504 + }, + { + "epoch": 2.404380341880342, + "grad_norm": 0.7587227821350098, + "learning_rate": 6.891730819947758e-05, + "loss": 0.8886, + "step": 13505 + }, + { + "epoch": 2.4045584045584047, + "grad_norm": 0.8367353677749634, + "learning_rate": 6.890400437520265e-05, + "loss": 1.008, + "step": 13506 + }, + { + "epoch": 2.404736467236467, + "grad_norm": 0.7200010418891907, + "learning_rate": 6.889070116018911e-05, + "loss": 0.8405, + "step": 13507 + }, + { + "epoch": 2.40491452991453, + "grad_norm": 0.9391907453536987, + "learning_rate": 6.887739855469769e-05, + "loss": 0.8904, + "step": 13508 + }, + { + "epoch": 2.4050925925925926, + "grad_norm": 0.8687568306922913, + "learning_rate": 6.886409655898902e-05, + "loss": 0.7145, + "step": 13509 + }, + { + "epoch": 2.4052706552706553, + "grad_norm": 0.7382767796516418, + "learning_rate": 6.885079517332366e-05, + "loss": 0.7639, + "step": 13510 + }, + { + "epoch": 2.405448717948718, + "grad_norm": 0.8322962522506714, + "learning_rate": 6.883749439796227e-05, + "loss": 1.0002, + "step": 13511 + }, + { + "epoch": 2.405626780626781, + "grad_norm": 0.815183162689209, + "learning_rate": 6.882419423316544e-05, + "loss": 0.8628, + "step": 13512 + }, + { + "epoch": 2.405804843304843, + "grad_norm": 0.9304860234260559, + "learning_rate": 6.881089467919381e-05, + "loss": 0.9489, + "step": 13513 + }, + { + "epoch": 2.405982905982906, + "grad_norm": 0.9071274995803833, + "learning_rate": 6.879759573630784e-05, + "loss": 0.8117, + "step": 13514 + }, + { + "epoch": 2.4061609686609686, + "grad_norm": 0.9378795027732849, + "learning_rate": 6.878429740476822e-05, + "loss": 1.22, + "step": 13515 + }, + { + "epoch": 2.4063390313390314, + "grad_norm": 0.7354511618614197, + "learning_rate": 6.877099968483541e-05, + "loss": 0.6696, + "step": 13516 + }, + { + "epoch": 2.406517094017094, + "grad_norm": 0.8701893091201782, + "learning_rate": 6.875770257677002e-05, + "loss": 0.8691, + "step": 13517 + }, + { + "epoch": 2.406695156695157, + "grad_norm": 0.8819001913070679, + "learning_rate": 6.87444060808325e-05, + "loss": 0.7428, + "step": 13518 + }, + { + "epoch": 2.406873219373219, + "grad_norm": 0.7339609265327454, + "learning_rate": 6.873111019728347e-05, + "loss": 0.7959, + "step": 13519 + }, + { + "epoch": 2.407051282051282, + "grad_norm": 0.8365123867988586, + "learning_rate": 6.871781492638335e-05, + "loss": 0.8199, + "step": 13520 + }, + { + "epoch": 2.4072293447293447, + "grad_norm": 0.9667043685913086, + "learning_rate": 6.870452026839266e-05, + "loss": 0.8261, + "step": 13521 + }, + { + "epoch": 2.4074074074074074, + "grad_norm": 0.6979679465293884, + "learning_rate": 6.869122622357187e-05, + "loss": 0.5909, + "step": 13522 + }, + { + "epoch": 2.40758547008547, + "grad_norm": 0.7326778769493103, + "learning_rate": 6.867793279218152e-05, + "loss": 0.9297, + "step": 13523 + }, + { + "epoch": 2.407763532763533, + "grad_norm": 0.8808563351631165, + "learning_rate": 6.866463997448196e-05, + "loss": 0.7481, + "step": 13524 + }, + { + "epoch": 2.4079415954415953, + "grad_norm": 0.7830268740653992, + "learning_rate": 6.86513477707337e-05, + "loss": 0.7902, + "step": 13525 + }, + { + "epoch": 2.408119658119658, + "grad_norm": 0.9482602477073669, + "learning_rate": 6.863805618119713e-05, + "loss": 1.1541, + "step": 13526 + }, + { + "epoch": 2.4082977207977208, + "grad_norm": 0.8369114995002747, + "learning_rate": 6.862476520613276e-05, + "loss": 0.874, + "step": 13527 + }, + { + "epoch": 2.4084757834757835, + "grad_norm": 0.9107078909873962, + "learning_rate": 6.86114748458009e-05, + "loss": 0.9412, + "step": 13528 + }, + { + "epoch": 2.4086538461538463, + "grad_norm": 0.8086137771606445, + "learning_rate": 6.859818510046199e-05, + "loss": 0.8495, + "step": 13529 + }, + { + "epoch": 2.408831908831909, + "grad_norm": 0.8824704885482788, + "learning_rate": 6.858489597037646e-05, + "loss": 0.8967, + "step": 13530 + }, + { + "epoch": 2.4090099715099713, + "grad_norm": 0.8514662384986877, + "learning_rate": 6.857160745580455e-05, + "loss": 0.9171, + "step": 13531 + }, + { + "epoch": 2.409188034188034, + "grad_norm": 0.7788167595863342, + "learning_rate": 6.855831955700675e-05, + "loss": 0.904, + "step": 13532 + }, + { + "epoch": 2.409366096866097, + "grad_norm": 0.913113534450531, + "learning_rate": 6.854503227424337e-05, + "loss": 0.8696, + "step": 13533 + }, + { + "epoch": 2.4095441595441596, + "grad_norm": 0.8424487113952637, + "learning_rate": 6.853174560777475e-05, + "loss": 0.8388, + "step": 13534 + }, + { + "epoch": 2.4097222222222223, + "grad_norm": 0.8609711527824402, + "learning_rate": 6.851845955786116e-05, + "loss": 0.7142, + "step": 13535 + }, + { + "epoch": 2.409900284900285, + "grad_norm": 0.8141375184059143, + "learning_rate": 6.850517412476301e-05, + "loss": 0.7198, + "step": 13536 + }, + { + "epoch": 2.410078347578348, + "grad_norm": 0.8615440130233765, + "learning_rate": 6.84918893087405e-05, + "loss": 0.958, + "step": 13537 + }, + { + "epoch": 2.41025641025641, + "grad_norm": 0.7733060717582703, + "learning_rate": 6.847860511005401e-05, + "loss": 0.7639, + "step": 13538 + }, + { + "epoch": 2.410434472934473, + "grad_norm": 0.9519185423851013, + "learning_rate": 6.846532152896375e-05, + "loss": 0.8239, + "step": 13539 + }, + { + "epoch": 2.4106125356125356, + "grad_norm": 0.774053692817688, + "learning_rate": 6.845203856573002e-05, + "loss": 0.891, + "step": 13540 + }, + { + "epoch": 2.4107905982905984, + "grad_norm": 0.8791571259498596, + "learning_rate": 6.843875622061304e-05, + "loss": 1.0107, + "step": 13541 + }, + { + "epoch": 2.410968660968661, + "grad_norm": 0.9431949853897095, + "learning_rate": 6.842547449387309e-05, + "loss": 0.8575, + "step": 13542 + }, + { + "epoch": 2.4111467236467234, + "grad_norm": 1.0521612167358398, + "learning_rate": 6.841219338577034e-05, + "loss": 0.9446, + "step": 13543 + }, + { + "epoch": 2.411324786324786, + "grad_norm": 0.7592857480049133, + "learning_rate": 6.83989128965651e-05, + "loss": 1.0595, + "step": 13544 + }, + { + "epoch": 2.411502849002849, + "grad_norm": 0.9002043604850769, + "learning_rate": 6.838563302651747e-05, + "loss": 0.9067, + "step": 13545 + }, + { + "epoch": 2.4116809116809117, + "grad_norm": 0.7144047021865845, + "learning_rate": 6.83723537758877e-05, + "loss": 0.6699, + "step": 13546 + }, + { + "epoch": 2.4118589743589745, + "grad_norm": 0.8226693868637085, + "learning_rate": 6.835907514493594e-05, + "loss": 0.987, + "step": 13547 + }, + { + "epoch": 2.412037037037037, + "grad_norm": 0.8507830500602722, + "learning_rate": 6.834579713392237e-05, + "loss": 0.7803, + "step": 13548 + }, + { + "epoch": 2.4122150997151, + "grad_norm": 0.727870762348175, + "learning_rate": 6.83325197431072e-05, + "loss": 0.8071, + "step": 13549 + }, + { + "epoch": 2.4123931623931623, + "grad_norm": 0.7601624727249146, + "learning_rate": 6.831924297275049e-05, + "loss": 0.7627, + "step": 13550 + }, + { + "epoch": 2.412571225071225, + "grad_norm": 0.8519877791404724, + "learning_rate": 6.830596682311243e-05, + "loss": 0.9271, + "step": 13551 + }, + { + "epoch": 2.4127492877492878, + "grad_norm": 1.0122307538986206, + "learning_rate": 6.829269129445307e-05, + "loss": 0.6424, + "step": 13552 + }, + { + "epoch": 2.4129273504273505, + "grad_norm": 0.8992687463760376, + "learning_rate": 6.827941638703258e-05, + "loss": 0.8034, + "step": 13553 + }, + { + "epoch": 2.4131054131054133, + "grad_norm": 0.7740746140480042, + "learning_rate": 6.826614210111102e-05, + "loss": 0.955, + "step": 13554 + }, + { + "epoch": 2.4132834757834756, + "grad_norm": 0.8176493048667908, + "learning_rate": 6.825286843694852e-05, + "loss": 0.7844, + "step": 13555 + }, + { + "epoch": 2.4134615384615383, + "grad_norm": 0.8112488985061646, + "learning_rate": 6.823959539480507e-05, + "loss": 0.8495, + "step": 13556 + }, + { + "epoch": 2.413639601139601, + "grad_norm": 0.8186960220336914, + "learning_rate": 6.822632297494078e-05, + "loss": 0.8922, + "step": 13557 + }, + { + "epoch": 2.413817663817664, + "grad_norm": 0.9498438835144043, + "learning_rate": 6.821305117761569e-05, + "loss": 0.8862, + "step": 13558 + }, + { + "epoch": 2.4139957264957266, + "grad_norm": 0.8591099381446838, + "learning_rate": 6.819978000308987e-05, + "loss": 0.837, + "step": 13559 + }, + { + "epoch": 2.4141737891737893, + "grad_norm": 0.8130860328674316, + "learning_rate": 6.818650945162324e-05, + "loss": 1.0723, + "step": 13560 + }, + { + "epoch": 2.414351851851852, + "grad_norm": 0.8800109624862671, + "learning_rate": 6.81732395234759e-05, + "loss": 0.8067, + "step": 13561 + }, + { + "epoch": 2.4145299145299144, + "grad_norm": 0.7786064147949219, + "learning_rate": 6.81599702189078e-05, + "loss": 0.9, + "step": 13562 + }, + { + "epoch": 2.414707977207977, + "grad_norm": 0.8343027234077454, + "learning_rate": 6.814670153817898e-05, + "loss": 0.7487, + "step": 13563 + }, + { + "epoch": 2.41488603988604, + "grad_norm": 0.7904187440872192, + "learning_rate": 6.813343348154934e-05, + "loss": 0.7904, + "step": 13564 + }, + { + "epoch": 2.4150641025641026, + "grad_norm": 0.7609010934829712, + "learning_rate": 6.81201660492789e-05, + "loss": 0.6734, + "step": 13565 + }, + { + "epoch": 2.4152421652421654, + "grad_norm": 0.8402243256568909, + "learning_rate": 6.810689924162756e-05, + "loss": 0.9581, + "step": 13566 + }, + { + "epoch": 2.4154202279202277, + "grad_norm": 0.8557454943656921, + "learning_rate": 6.809363305885527e-05, + "loss": 0.8387, + "step": 13567 + }, + { + "epoch": 2.4155982905982905, + "grad_norm": 0.8983132243156433, + "learning_rate": 6.808036750122197e-05, + "loss": 0.832, + "step": 13568 + }, + { + "epoch": 2.415776353276353, + "grad_norm": 0.8552190065383911, + "learning_rate": 6.806710256898755e-05, + "loss": 0.9257, + "step": 13569 + }, + { + "epoch": 2.415954415954416, + "grad_norm": 1.0639078617095947, + "learning_rate": 6.805383826241197e-05, + "loss": 0.9743, + "step": 13570 + }, + { + "epoch": 2.4161324786324787, + "grad_norm": 0.7951667904853821, + "learning_rate": 6.804057458175501e-05, + "loss": 0.9326, + "step": 13571 + }, + { + "epoch": 2.4163105413105415, + "grad_norm": 0.8652639985084534, + "learning_rate": 6.802731152727664e-05, + "loss": 1.0114, + "step": 13572 + }, + { + "epoch": 2.416488603988604, + "grad_norm": 0.8777487874031067, + "learning_rate": 6.801404909923664e-05, + "loss": 0.759, + "step": 13573 + }, + { + "epoch": 2.4166666666666665, + "grad_norm": 0.7922869920730591, + "learning_rate": 6.800078729789497e-05, + "loss": 0.8392, + "step": 13574 + }, + { + "epoch": 2.4168447293447293, + "grad_norm": 0.9189477562904358, + "learning_rate": 6.798752612351133e-05, + "loss": 0.9485, + "step": 13575 + }, + { + "epoch": 2.417022792022792, + "grad_norm": 0.8752175569534302, + "learning_rate": 6.797426557634567e-05, + "loss": 0.725, + "step": 13576 + }, + { + "epoch": 2.4172008547008548, + "grad_norm": 0.8646897077560425, + "learning_rate": 6.79610056566577e-05, + "loss": 0.9791, + "step": 13577 + }, + { + "epoch": 2.4173789173789175, + "grad_norm": 0.8749415278434753, + "learning_rate": 6.794774636470731e-05, + "loss": 0.9059, + "step": 13578 + }, + { + "epoch": 2.41755698005698, + "grad_norm": 0.9642252922058105, + "learning_rate": 6.793448770075422e-05, + "loss": 0.9972, + "step": 13579 + }, + { + "epoch": 2.4177350427350426, + "grad_norm": 0.8430541157722473, + "learning_rate": 6.792122966505827e-05, + "loss": 0.7126, + "step": 13580 + }, + { + "epoch": 2.4179131054131053, + "grad_norm": 0.8478374481201172, + "learning_rate": 6.790797225787913e-05, + "loss": 0.8995, + "step": 13581 + }, + { + "epoch": 2.418091168091168, + "grad_norm": 0.7008727788925171, + "learning_rate": 6.789471547947665e-05, + "loss": 0.6694, + "step": 13582 + }, + { + "epoch": 2.418269230769231, + "grad_norm": 0.873543918132782, + "learning_rate": 6.78814593301105e-05, + "loss": 1.0418, + "step": 13583 + }, + { + "epoch": 2.4184472934472936, + "grad_norm": 0.7208766341209412, + "learning_rate": 6.786820381004047e-05, + "loss": 0.8095, + "step": 13584 + }, + { + "epoch": 2.4186253561253563, + "grad_norm": 0.7272628545761108, + "learning_rate": 6.78549489195262e-05, + "loss": 0.5801, + "step": 13585 + }, + { + "epoch": 2.4188034188034186, + "grad_norm": 0.7155343294143677, + "learning_rate": 6.784169465882747e-05, + "loss": 0.748, + "step": 13586 + }, + { + "epoch": 2.4189814814814814, + "grad_norm": 0.928404688835144, + "learning_rate": 6.78284410282039e-05, + "loss": 0.9175, + "step": 13587 + }, + { + "epoch": 2.419159544159544, + "grad_norm": 0.7239044308662415, + "learning_rate": 6.781518802791519e-05, + "loss": 0.5855, + "step": 13588 + }, + { + "epoch": 2.419337606837607, + "grad_norm": 0.8126311302185059, + "learning_rate": 6.780193565822104e-05, + "loss": 0.832, + "step": 13589 + }, + { + "epoch": 2.4195156695156697, + "grad_norm": 0.7470774054527283, + "learning_rate": 6.778868391938103e-05, + "loss": 0.6202, + "step": 13590 + }, + { + "epoch": 2.419693732193732, + "grad_norm": 0.9161462187767029, + "learning_rate": 6.77754328116549e-05, + "loss": 0.8674, + "step": 13591 + }, + { + "epoch": 2.4198717948717947, + "grad_norm": 0.7225745320320129, + "learning_rate": 6.77621823353022e-05, + "loss": 0.8745, + "step": 13592 + }, + { + "epoch": 2.4200498575498575, + "grad_norm": 0.8380082845687866, + "learning_rate": 6.774893249058257e-05, + "loss": 0.5501, + "step": 13593 + }, + { + "epoch": 2.42022792022792, + "grad_norm": 0.8031942844390869, + "learning_rate": 6.77356832777556e-05, + "loss": 0.9925, + "step": 13594 + }, + { + "epoch": 2.420405982905983, + "grad_norm": 0.8278502821922302, + "learning_rate": 6.772243469708093e-05, + "loss": 0.6411, + "step": 13595 + }, + { + "epoch": 2.4205840455840457, + "grad_norm": 0.7655481100082397, + "learning_rate": 6.770918674881805e-05, + "loss": 0.7896, + "step": 13596 + }, + { + "epoch": 2.4207621082621085, + "grad_norm": 0.8260186314582825, + "learning_rate": 6.769593943322661e-05, + "loss": 0.8531, + "step": 13597 + }, + { + "epoch": 2.4209401709401708, + "grad_norm": 0.8293251395225525, + "learning_rate": 6.76826927505661e-05, + "loss": 0.8193, + "step": 13598 + }, + { + "epoch": 2.4211182336182335, + "grad_norm": 0.8868293762207031, + "learning_rate": 6.766944670109616e-05, + "loss": 0.8453, + "step": 13599 + }, + { + "epoch": 2.4212962962962963, + "grad_norm": 0.769124448299408, + "learning_rate": 6.765620128507619e-05, + "loss": 0.7412, + "step": 13600 + }, + { + "epoch": 2.421474358974359, + "grad_norm": 0.7727167010307312, + "learning_rate": 6.764295650276581e-05, + "loss": 0.8721, + "step": 13601 + }, + { + "epoch": 2.421652421652422, + "grad_norm": 0.9975818395614624, + "learning_rate": 6.762971235442444e-05, + "loss": 0.8128, + "step": 13602 + }, + { + "epoch": 2.421830484330484, + "grad_norm": 0.8000788688659668, + "learning_rate": 6.761646884031164e-05, + "loss": 0.8328, + "step": 13603 + }, + { + "epoch": 2.422008547008547, + "grad_norm": 0.7196731567382812, + "learning_rate": 6.760322596068684e-05, + "loss": 0.8912, + "step": 13604 + }, + { + "epoch": 2.4221866096866096, + "grad_norm": 0.8092321753501892, + "learning_rate": 6.758998371580955e-05, + "loss": 0.9066, + "step": 13605 + }, + { + "epoch": 2.4223646723646723, + "grad_norm": 0.7664031982421875, + "learning_rate": 6.757674210593918e-05, + "loss": 0.9854, + "step": 13606 + }, + { + "epoch": 2.422542735042735, + "grad_norm": 0.794507622718811, + "learning_rate": 6.75635011313352e-05, + "loss": 0.9296, + "step": 13607 + }, + { + "epoch": 2.422720797720798, + "grad_norm": 0.9127107858657837, + "learning_rate": 6.755026079225705e-05, + "loss": 0.9516, + "step": 13608 + }, + { + "epoch": 2.4228988603988606, + "grad_norm": 0.8025720715522766, + "learning_rate": 6.753702108896411e-05, + "loss": 1.0664, + "step": 13609 + }, + { + "epoch": 2.423076923076923, + "grad_norm": 0.7304871678352356, + "learning_rate": 6.752378202171585e-05, + "loss": 0.9071, + "step": 13610 + }, + { + "epoch": 2.4232549857549857, + "grad_norm": 0.9048241972923279, + "learning_rate": 6.751054359077157e-05, + "loss": 0.8649, + "step": 13611 + }, + { + "epoch": 2.4234330484330484, + "grad_norm": 0.8589995503425598, + "learning_rate": 6.749730579639074e-05, + "loss": 0.8895, + "step": 13612 + }, + { + "epoch": 2.423611111111111, + "grad_norm": 0.8098960518836975, + "learning_rate": 6.748406863883265e-05, + "loss": 0.6651, + "step": 13613 + }, + { + "epoch": 2.423789173789174, + "grad_norm": 0.9875120520591736, + "learning_rate": 6.74708321183567e-05, + "loss": 0.9605, + "step": 13614 + }, + { + "epoch": 2.423967236467236, + "grad_norm": 0.8211431503295898, + "learning_rate": 6.74575962352222e-05, + "loss": 0.8721, + "step": 13615 + }, + { + "epoch": 2.424145299145299, + "grad_norm": 0.9732884168624878, + "learning_rate": 6.744436098968855e-05, + "loss": 0.7501, + "step": 13616 + }, + { + "epoch": 2.4243233618233617, + "grad_norm": 0.9813733696937561, + "learning_rate": 6.743112638201496e-05, + "loss": 0.9823, + "step": 13617 + }, + { + "epoch": 2.4245014245014245, + "grad_norm": 0.8075012564659119, + "learning_rate": 6.741789241246083e-05, + "loss": 0.7018, + "step": 13618 + }, + { + "epoch": 2.4246794871794872, + "grad_norm": 0.7845864295959473, + "learning_rate": 6.740465908128539e-05, + "loss": 0.7423, + "step": 13619 + }, + { + "epoch": 2.42485754985755, + "grad_norm": 0.7754862308502197, + "learning_rate": 6.739142638874799e-05, + "loss": 0.8735, + "step": 13620 + }, + { + "epoch": 2.4250356125356127, + "grad_norm": 0.7971537709236145, + "learning_rate": 6.737819433510781e-05, + "loss": 0.7663, + "step": 13621 + }, + { + "epoch": 2.425213675213675, + "grad_norm": 0.8043563365936279, + "learning_rate": 6.736496292062416e-05, + "loss": 0.9311, + "step": 13622 + }, + { + "epoch": 2.425391737891738, + "grad_norm": 0.8150136470794678, + "learning_rate": 6.735173214555628e-05, + "loss": 0.9164, + "step": 13623 + }, + { + "epoch": 2.4255698005698005, + "grad_norm": 0.9853758811950684, + "learning_rate": 6.733850201016338e-05, + "loss": 0.6253, + "step": 13624 + }, + { + "epoch": 2.4257478632478633, + "grad_norm": 1.2138506174087524, + "learning_rate": 6.732527251470465e-05, + "loss": 1.0536, + "step": 13625 + }, + { + "epoch": 2.425925925925926, + "grad_norm": 0.9306546449661255, + "learning_rate": 6.73120436594394e-05, + "loss": 0.8607, + "step": 13626 + }, + { + "epoch": 2.426103988603989, + "grad_norm": 0.8536837697029114, + "learning_rate": 6.729881544462668e-05, + "loss": 0.9418, + "step": 13627 + }, + { + "epoch": 2.426282051282051, + "grad_norm": 0.8561417460441589, + "learning_rate": 6.728558787052574e-05, + "loss": 0.9556, + "step": 13628 + }, + { + "epoch": 2.426460113960114, + "grad_norm": 0.7499847412109375, + "learning_rate": 6.727236093739579e-05, + "loss": 0.7795, + "step": 13629 + }, + { + "epoch": 2.4266381766381766, + "grad_norm": 0.8541018962860107, + "learning_rate": 6.725913464549591e-05, + "loss": 1.0322, + "step": 13630 + }, + { + "epoch": 2.4268162393162394, + "grad_norm": 0.9659489989280701, + "learning_rate": 6.724590899508532e-05, + "loss": 1.1907, + "step": 13631 + }, + { + "epoch": 2.426994301994302, + "grad_norm": 0.9548102617263794, + "learning_rate": 6.723268398642307e-05, + "loss": 1.0545, + "step": 13632 + }, + { + "epoch": 2.427172364672365, + "grad_norm": 0.8543868660926819, + "learning_rate": 6.72194596197683e-05, + "loss": 0.8041, + "step": 13633 + }, + { + "epoch": 2.427350427350427, + "grad_norm": 0.838178277015686, + "learning_rate": 6.720623589538013e-05, + "loss": 0.7772, + "step": 13634 + }, + { + "epoch": 2.42752849002849, + "grad_norm": 0.8207933306694031, + "learning_rate": 6.719301281351768e-05, + "loss": 0.9222, + "step": 13635 + }, + { + "epoch": 2.4277065527065527, + "grad_norm": 0.7705093026161194, + "learning_rate": 6.717979037443996e-05, + "loss": 0.8989, + "step": 13636 + }, + { + "epoch": 2.4278846153846154, + "grad_norm": 0.8627061247825623, + "learning_rate": 6.716656857840609e-05, + "loss": 0.8834, + "step": 13637 + }, + { + "epoch": 2.428062678062678, + "grad_norm": 0.8404269218444824, + "learning_rate": 6.715334742567507e-05, + "loss": 0.8087, + "step": 13638 + }, + { + "epoch": 2.428240740740741, + "grad_norm": 0.9098958373069763, + "learning_rate": 6.7140126916506e-05, + "loss": 0.803, + "step": 13639 + }, + { + "epoch": 2.4284188034188032, + "grad_norm": 0.7482922673225403, + "learning_rate": 6.712690705115785e-05, + "loss": 0.8254, + "step": 13640 + }, + { + "epoch": 2.428596866096866, + "grad_norm": 0.8636375665664673, + "learning_rate": 6.711368782988972e-05, + "loss": 0.8788, + "step": 13641 + }, + { + "epoch": 2.4287749287749287, + "grad_norm": 0.8261808753013611, + "learning_rate": 6.710046925296052e-05, + "loss": 0.9135, + "step": 13642 + }, + { + "epoch": 2.4289529914529915, + "grad_norm": 1.060263752937317, + "learning_rate": 6.70872513206293e-05, + "loss": 0.9589, + "step": 13643 + }, + { + "epoch": 2.4291310541310542, + "grad_norm": 0.8128657341003418, + "learning_rate": 6.7074034033155e-05, + "loss": 0.7735, + "step": 13644 + }, + { + "epoch": 2.429309116809117, + "grad_norm": 0.9948938488960266, + "learning_rate": 6.706081739079663e-05, + "loss": 0.7242, + "step": 13645 + }, + { + "epoch": 2.4294871794871793, + "grad_norm": 0.8850025534629822, + "learning_rate": 6.704760139381311e-05, + "loss": 0.9393, + "step": 13646 + }, + { + "epoch": 2.429665242165242, + "grad_norm": 0.833534300327301, + "learning_rate": 6.703438604246337e-05, + "loss": 0.7824, + "step": 13647 + }, + { + "epoch": 2.429843304843305, + "grad_norm": 0.7362738251686096, + "learning_rate": 6.70211713370064e-05, + "loss": 0.7594, + "step": 13648 + }, + { + "epoch": 2.4300213675213675, + "grad_norm": 0.97635817527771, + "learning_rate": 6.700795727770101e-05, + "loss": 1.0097, + "step": 13649 + }, + { + "epoch": 2.4301994301994303, + "grad_norm": 0.8434939980506897, + "learning_rate": 6.699474386480622e-05, + "loss": 0.6639, + "step": 13650 + }, + { + "epoch": 2.430377492877493, + "grad_norm": 0.7960709929466248, + "learning_rate": 6.69815310985808e-05, + "loss": 0.9557, + "step": 13651 + }, + { + "epoch": 2.4305555555555554, + "grad_norm": 0.8336359262466431, + "learning_rate": 6.696831897928376e-05, + "loss": 1.0112, + "step": 13652 + }, + { + "epoch": 2.430733618233618, + "grad_norm": 0.8353996872901917, + "learning_rate": 6.695510750717384e-05, + "loss": 0.8578, + "step": 13653 + }, + { + "epoch": 2.430911680911681, + "grad_norm": 0.8968163132667542, + "learning_rate": 6.694189668250996e-05, + "loss": 1.0412, + "step": 13654 + }, + { + "epoch": 2.4310897435897436, + "grad_norm": 0.8091850876808167, + "learning_rate": 6.692868650555093e-05, + "loss": 0.761, + "step": 13655 + }, + { + "epoch": 2.4312678062678064, + "grad_norm": 0.7735705375671387, + "learning_rate": 6.691547697655563e-05, + "loss": 1.0507, + "step": 13656 + }, + { + "epoch": 2.431445868945869, + "grad_norm": 0.7707101702690125, + "learning_rate": 6.690226809578279e-05, + "loss": 0.8883, + "step": 13657 + }, + { + "epoch": 2.431623931623932, + "grad_norm": 0.8384747505187988, + "learning_rate": 6.688905986349127e-05, + "loss": 0.7772, + "step": 13658 + }, + { + "epoch": 2.431801994301994, + "grad_norm": 0.7676185369491577, + "learning_rate": 6.687585227993985e-05, + "loss": 0.844, + "step": 13659 + }, + { + "epoch": 2.431980056980057, + "grad_norm": 0.8745819926261902, + "learning_rate": 6.686264534538726e-05, + "loss": 0.9996, + "step": 13660 + }, + { + "epoch": 2.4321581196581197, + "grad_norm": 0.7455142736434937, + "learning_rate": 6.684943906009232e-05, + "loss": 0.7133, + "step": 13661 + }, + { + "epoch": 2.4323361823361824, + "grad_norm": 0.8742238879203796, + "learning_rate": 6.683623342431378e-05, + "loss": 0.8155, + "step": 13662 + }, + { + "epoch": 2.432514245014245, + "grad_norm": 0.7863791584968567, + "learning_rate": 6.68230284383103e-05, + "loss": 1.0017, + "step": 13663 + }, + { + "epoch": 2.4326923076923075, + "grad_norm": 0.9469232559204102, + "learning_rate": 6.68098241023407e-05, + "loss": 0.8172, + "step": 13664 + }, + { + "epoch": 2.4328703703703702, + "grad_norm": 0.808024525642395, + "learning_rate": 6.679662041666362e-05, + "loss": 0.8813, + "step": 13665 + }, + { + "epoch": 2.433048433048433, + "grad_norm": 0.834863543510437, + "learning_rate": 6.67834173815378e-05, + "loss": 0.9765, + "step": 13666 + }, + { + "epoch": 2.4332264957264957, + "grad_norm": 0.8903583288192749, + "learning_rate": 6.677021499722193e-05, + "loss": 0.9393, + "step": 13667 + }, + { + "epoch": 2.4334045584045585, + "grad_norm": 0.8341929912567139, + "learning_rate": 6.675701326397466e-05, + "loss": 0.8681, + "step": 13668 + }, + { + "epoch": 2.4335826210826212, + "grad_norm": 0.9348049163818359, + "learning_rate": 6.674381218205465e-05, + "loss": 0.7396, + "step": 13669 + }, + { + "epoch": 2.433760683760684, + "grad_norm": 0.8898159861564636, + "learning_rate": 6.673061175172055e-05, + "loss": 0.8638, + "step": 13670 + }, + { + "epoch": 2.4339387464387463, + "grad_norm": 0.8101391792297363, + "learning_rate": 6.671741197323105e-05, + "loss": 0.8064, + "step": 13671 + }, + { + "epoch": 2.434116809116809, + "grad_norm": 0.8756688237190247, + "learning_rate": 6.670421284684467e-05, + "loss": 0.7718, + "step": 13672 + }, + { + "epoch": 2.434294871794872, + "grad_norm": 0.8060923218727112, + "learning_rate": 6.669101437282012e-05, + "loss": 0.8137, + "step": 13673 + }, + { + "epoch": 2.4344729344729346, + "grad_norm": 0.792891800403595, + "learning_rate": 6.667781655141589e-05, + "loss": 0.9104, + "step": 13674 + }, + { + "epoch": 2.4346509971509973, + "grad_norm": 0.8590527772903442, + "learning_rate": 6.666461938289068e-05, + "loss": 0.9578, + "step": 13675 + }, + { + "epoch": 2.4348290598290596, + "grad_norm": 0.8593253493309021, + "learning_rate": 6.665142286750297e-05, + "loss": 0.8083, + "step": 13676 + }, + { + "epoch": 2.4350071225071224, + "grad_norm": 0.8237900733947754, + "learning_rate": 6.663822700551137e-05, + "loss": 0.8096, + "step": 13677 + }, + { + "epoch": 2.435185185185185, + "grad_norm": 0.9017227292060852, + "learning_rate": 6.66250317971744e-05, + "loss": 0.9599, + "step": 13678 + }, + { + "epoch": 2.435363247863248, + "grad_norm": 0.7811765670776367, + "learning_rate": 6.661183724275061e-05, + "loss": 0.8392, + "step": 13679 + }, + { + "epoch": 2.4355413105413106, + "grad_norm": 0.8113176822662354, + "learning_rate": 6.659864334249848e-05, + "loss": 0.6788, + "step": 13680 + }, + { + "epoch": 2.4357193732193734, + "grad_norm": 0.795261561870575, + "learning_rate": 6.65854500966766e-05, + "loss": 0.783, + "step": 13681 + }, + { + "epoch": 2.435897435897436, + "grad_norm": 0.7738518714904785, + "learning_rate": 6.657225750554338e-05, + "loss": 0.676, + "step": 13682 + }, + { + "epoch": 2.4360754985754984, + "grad_norm": 0.8513518571853638, + "learning_rate": 6.655906556935737e-05, + "loss": 0.8743, + "step": 13683 + }, + { + "epoch": 2.436253561253561, + "grad_norm": 0.9595896005630493, + "learning_rate": 6.654587428837696e-05, + "loss": 0.9333, + "step": 13684 + }, + { + "epoch": 2.436431623931624, + "grad_norm": 0.7720373272895813, + "learning_rate": 6.653268366286066e-05, + "loss": 0.77, + "step": 13685 + }, + { + "epoch": 2.4366096866096867, + "grad_norm": 0.9022032022476196, + "learning_rate": 6.651949369306689e-05, + "loss": 0.8786, + "step": 13686 + }, + { + "epoch": 2.4367877492877494, + "grad_norm": 0.795092761516571, + "learning_rate": 6.650630437925409e-05, + "loss": 0.8557, + "step": 13687 + }, + { + "epoch": 2.4369658119658117, + "grad_norm": 0.8517789244651794, + "learning_rate": 6.649311572168072e-05, + "loss": 0.9781, + "step": 13688 + }, + { + "epoch": 2.4371438746438745, + "grad_norm": 1.0523463487625122, + "learning_rate": 6.64799277206051e-05, + "loss": 0.9046, + "step": 13689 + }, + { + "epoch": 2.4373219373219372, + "grad_norm": 0.9401832818984985, + "learning_rate": 6.646674037628568e-05, + "loss": 1.0966, + "step": 13690 + }, + { + "epoch": 2.4375, + "grad_norm": 0.9895738959312439, + "learning_rate": 6.645355368898082e-05, + "loss": 0.714, + "step": 13691 + }, + { + "epoch": 2.4376780626780628, + "grad_norm": 0.8458610773086548, + "learning_rate": 6.644036765894892e-05, + "loss": 0.6265, + "step": 13692 + }, + { + "epoch": 2.4378561253561255, + "grad_norm": 0.7908345460891724, + "learning_rate": 6.642718228644826e-05, + "loss": 0.7943, + "step": 13693 + }, + { + "epoch": 2.4380341880341883, + "grad_norm": 0.8119938373565674, + "learning_rate": 6.641399757173725e-05, + "loss": 0.5948, + "step": 13694 + }, + { + "epoch": 2.4382122507122506, + "grad_norm": 0.8175633549690247, + "learning_rate": 6.640081351507417e-05, + "loss": 0.9098, + "step": 13695 + }, + { + "epoch": 2.4383903133903133, + "grad_norm": 0.8546686768531799, + "learning_rate": 6.638763011671736e-05, + "loss": 1.0347, + "step": 13696 + }, + { + "epoch": 2.438568376068376, + "grad_norm": 0.812406599521637, + "learning_rate": 6.637444737692508e-05, + "loss": 0.8469, + "step": 13697 + }, + { + "epoch": 2.438746438746439, + "grad_norm": 0.7802549004554749, + "learning_rate": 6.636126529595572e-05, + "loss": 0.7024, + "step": 13698 + }, + { + "epoch": 2.4389245014245016, + "grad_norm": 0.8046648502349854, + "learning_rate": 6.634808387406744e-05, + "loss": 0.8292, + "step": 13699 + }, + { + "epoch": 2.439102564102564, + "grad_norm": 0.8544600009918213, + "learning_rate": 6.633490311151857e-05, + "loss": 0.8033, + "step": 13700 + }, + { + "epoch": 2.4392806267806266, + "grad_norm": 0.8327271938323975, + "learning_rate": 6.632172300856731e-05, + "loss": 0.8641, + "step": 13701 + }, + { + "epoch": 2.4394586894586894, + "grad_norm": 0.9563352465629578, + "learning_rate": 6.630854356547199e-05, + "loss": 0.8144, + "step": 13702 + }, + { + "epoch": 2.439636752136752, + "grad_norm": 0.8993256092071533, + "learning_rate": 6.629536478249071e-05, + "loss": 0.8688, + "step": 13703 + }, + { + "epoch": 2.439814814814815, + "grad_norm": 0.8741861581802368, + "learning_rate": 6.628218665988178e-05, + "loss": 0.9868, + "step": 13704 + }, + { + "epoch": 2.4399928774928776, + "grad_norm": 0.7898648381233215, + "learning_rate": 6.626900919790332e-05, + "loss": 0.7802, + "step": 13705 + }, + { + "epoch": 2.4401709401709404, + "grad_norm": 0.7651925086975098, + "learning_rate": 6.625583239681357e-05, + "loss": 0.8131, + "step": 13706 + }, + { + "epoch": 2.4403490028490027, + "grad_norm": 0.7917741537094116, + "learning_rate": 6.624265625687071e-05, + "loss": 0.8581, + "step": 13707 + }, + { + "epoch": 2.4405270655270654, + "grad_norm": 0.7631075978279114, + "learning_rate": 6.622948077833284e-05, + "loss": 0.6069, + "step": 13708 + }, + { + "epoch": 2.440705128205128, + "grad_norm": 0.920765221118927, + "learning_rate": 6.621630596145819e-05, + "loss": 0.6846, + "step": 13709 + }, + { + "epoch": 2.440883190883191, + "grad_norm": 0.822335422039032, + "learning_rate": 6.62031318065048e-05, + "loss": 1.0309, + "step": 13710 + }, + { + "epoch": 2.4410612535612537, + "grad_norm": 0.7978029251098633, + "learning_rate": 6.618995831373086e-05, + "loss": 0.9593, + "step": 13711 + }, + { + "epoch": 2.441239316239316, + "grad_norm": 0.8908950686454773, + "learning_rate": 6.617678548339443e-05, + "loss": 0.7147, + "step": 13712 + }, + { + "epoch": 2.4414173789173788, + "grad_norm": 0.7772884368896484, + "learning_rate": 6.616361331575368e-05, + "loss": 0.8839, + "step": 13713 + }, + { + "epoch": 2.4415954415954415, + "grad_norm": 0.8437771797180176, + "learning_rate": 6.615044181106658e-05, + "loss": 0.8388, + "step": 13714 + }, + { + "epoch": 2.4417735042735043, + "grad_norm": 0.8549850583076477, + "learning_rate": 6.613727096959128e-05, + "loss": 0.9251, + "step": 13715 + }, + { + "epoch": 2.441951566951567, + "grad_norm": 0.8402581810951233, + "learning_rate": 6.612410079158579e-05, + "loss": 0.8607, + "step": 13716 + }, + { + "epoch": 2.4421296296296298, + "grad_norm": 0.8181160688400269, + "learning_rate": 6.611093127730821e-05, + "loss": 0.6082, + "step": 13717 + }, + { + "epoch": 2.4423076923076925, + "grad_norm": 0.9006236791610718, + "learning_rate": 6.609776242701651e-05, + "loss": 0.9091, + "step": 13718 + }, + { + "epoch": 2.442485754985755, + "grad_norm": 0.7759920358657837, + "learning_rate": 6.608459424096876e-05, + "loss": 0.842, + "step": 13719 + }, + { + "epoch": 2.4426638176638176, + "grad_norm": 0.825701117515564, + "learning_rate": 6.60714267194229e-05, + "loss": 0.9325, + "step": 13720 + }, + { + "epoch": 2.4428418803418803, + "grad_norm": 0.7646961212158203, + "learning_rate": 6.605825986263697e-05, + "loss": 0.8124, + "step": 13721 + }, + { + "epoch": 2.443019943019943, + "grad_norm": 0.896112322807312, + "learning_rate": 6.604509367086888e-05, + "loss": 1.0962, + "step": 13722 + }, + { + "epoch": 2.443198005698006, + "grad_norm": 0.8079821467399597, + "learning_rate": 6.603192814437672e-05, + "loss": 0.8195, + "step": 13723 + }, + { + "epoch": 2.443376068376068, + "grad_norm": 0.8901529908180237, + "learning_rate": 6.601876328341831e-05, + "loss": 0.7886, + "step": 13724 + }, + { + "epoch": 2.443554131054131, + "grad_norm": 1.0454550981521606, + "learning_rate": 6.600559908825168e-05, + "loss": 0.9642, + "step": 13725 + }, + { + "epoch": 2.4437321937321936, + "grad_norm": 0.7995026707649231, + "learning_rate": 6.599243555913469e-05, + "loss": 0.6927, + "step": 13726 + }, + { + "epoch": 2.4439102564102564, + "grad_norm": 0.9235756397247314, + "learning_rate": 6.597927269632526e-05, + "loss": 0.8986, + "step": 13727 + }, + { + "epoch": 2.444088319088319, + "grad_norm": 0.7869365215301514, + "learning_rate": 6.596611050008137e-05, + "loss": 0.7592, + "step": 13728 + }, + { + "epoch": 2.444266381766382, + "grad_norm": 0.8172873258590698, + "learning_rate": 6.595294897066081e-05, + "loss": 0.8048, + "step": 13729 + }, + { + "epoch": 2.4444444444444446, + "grad_norm": 0.8021790981292725, + "learning_rate": 6.593978810832152e-05, + "loss": 0.767, + "step": 13730 + }, + { + "epoch": 2.444622507122507, + "grad_norm": 0.7781784534454346, + "learning_rate": 6.592662791332129e-05, + "loss": 0.8364, + "step": 13731 + }, + { + "epoch": 2.4448005698005697, + "grad_norm": 0.8227871656417847, + "learning_rate": 6.591346838591803e-05, + "loss": 0.8967, + "step": 13732 + }, + { + "epoch": 2.4449786324786325, + "grad_norm": 0.8349295854568481, + "learning_rate": 6.590030952636952e-05, + "loss": 0.8593, + "step": 13733 + }, + { + "epoch": 2.445156695156695, + "grad_norm": 1.0261762142181396, + "learning_rate": 6.588715133493365e-05, + "loss": 0.701, + "step": 13734 + }, + { + "epoch": 2.445334757834758, + "grad_norm": 0.8612635731697083, + "learning_rate": 6.587399381186814e-05, + "loss": 0.9803, + "step": 13735 + }, + { + "epoch": 2.4455128205128207, + "grad_norm": 0.7890039682388306, + "learning_rate": 6.586083695743086e-05, + "loss": 0.8001, + "step": 13736 + }, + { + "epoch": 2.445690883190883, + "grad_norm": 0.9934018850326538, + "learning_rate": 6.584768077187955e-05, + "loss": 1.0089, + "step": 13737 + }, + { + "epoch": 2.4458689458689458, + "grad_norm": 0.8232909440994263, + "learning_rate": 6.583452525547202e-05, + "loss": 0.8088, + "step": 13738 + }, + { + "epoch": 2.4460470085470085, + "grad_norm": 0.8635872006416321, + "learning_rate": 6.582137040846595e-05, + "loss": 1.0007, + "step": 13739 + }, + { + "epoch": 2.4462250712250713, + "grad_norm": 0.905575692653656, + "learning_rate": 6.580821623111914e-05, + "loss": 0.7577, + "step": 13740 + }, + { + "epoch": 2.446403133903134, + "grad_norm": 0.8264324069023132, + "learning_rate": 6.579506272368931e-05, + "loss": 0.9337, + "step": 13741 + }, + { + "epoch": 2.4465811965811968, + "grad_norm": 0.9100900292396545, + "learning_rate": 6.57819098864342e-05, + "loss": 0.8115, + "step": 13742 + }, + { + "epoch": 2.446759259259259, + "grad_norm": 0.8536351919174194, + "learning_rate": 6.576875771961145e-05, + "loss": 0.8612, + "step": 13743 + }, + { + "epoch": 2.446937321937322, + "grad_norm": 0.8968019485473633, + "learning_rate": 6.57556062234788e-05, + "loss": 0.8798, + "step": 13744 + }, + { + "epoch": 2.4471153846153846, + "grad_norm": 0.8745046854019165, + "learning_rate": 6.574245539829389e-05, + "loss": 0.8992, + "step": 13745 + }, + { + "epoch": 2.4472934472934473, + "grad_norm": 0.8336703777313232, + "learning_rate": 6.57293052443144e-05, + "loss": 0.7947, + "step": 13746 + }, + { + "epoch": 2.44747150997151, + "grad_norm": 0.8544902801513672, + "learning_rate": 6.571615576179801e-05, + "loss": 1.0315, + "step": 13747 + }, + { + "epoch": 2.447649572649573, + "grad_norm": 0.848242461681366, + "learning_rate": 6.570300695100229e-05, + "loss": 0.853, + "step": 13748 + }, + { + "epoch": 2.447827635327635, + "grad_norm": 0.7753778100013733, + "learning_rate": 6.568985881218496e-05, + "loss": 0.934, + "step": 13749 + }, + { + "epoch": 2.448005698005698, + "grad_norm": 0.8294853568077087, + "learning_rate": 6.567671134560351e-05, + "loss": 0.9328, + "step": 13750 + }, + { + "epoch": 2.4481837606837606, + "grad_norm": 0.8720992803573608, + "learning_rate": 6.566356455151565e-05, + "loss": 0.875, + "step": 13751 + }, + { + "epoch": 2.4483618233618234, + "grad_norm": 0.8204464316368103, + "learning_rate": 6.565041843017888e-05, + "loss": 0.923, + "step": 13752 + }, + { + "epoch": 2.448539886039886, + "grad_norm": 0.940037190914154, + "learning_rate": 6.563727298185085e-05, + "loss": 1.1596, + "step": 13753 + }, + { + "epoch": 2.448717948717949, + "grad_norm": 0.8390263915061951, + "learning_rate": 6.562412820678902e-05, + "loss": 0.8256, + "step": 13754 + }, + { + "epoch": 2.448896011396011, + "grad_norm": 0.8572748303413391, + "learning_rate": 6.561098410525106e-05, + "loss": 0.7833, + "step": 13755 + }, + { + "epoch": 2.449074074074074, + "grad_norm": 0.7981020212173462, + "learning_rate": 6.559784067749436e-05, + "loss": 0.7609, + "step": 13756 + }, + { + "epoch": 2.4492521367521367, + "grad_norm": 0.8356930613517761, + "learning_rate": 6.558469792377653e-05, + "loss": 0.7542, + "step": 13757 + }, + { + "epoch": 2.4494301994301995, + "grad_norm": 0.9340906739234924, + "learning_rate": 6.557155584435504e-05, + "loss": 0.9898, + "step": 13758 + }, + { + "epoch": 2.449608262108262, + "grad_norm": 1.0551100969314575, + "learning_rate": 6.555841443948743e-05, + "loss": 0.7189, + "step": 13759 + }, + { + "epoch": 2.449786324786325, + "grad_norm": 0.9572125673294067, + "learning_rate": 6.554527370943111e-05, + "loss": 0.8878, + "step": 13760 + }, + { + "epoch": 2.4499643874643873, + "grad_norm": 0.8760324716567993, + "learning_rate": 6.55321336544436e-05, + "loss": 0.75, + "step": 13761 + }, + { + "epoch": 2.45014245014245, + "grad_norm": 0.7599226236343384, + "learning_rate": 6.55189942747823e-05, + "loss": 0.9222, + "step": 13762 + }, + { + "epoch": 2.4503205128205128, + "grad_norm": 0.7307319045066833, + "learning_rate": 6.550585557070473e-05, + "loss": 0.833, + "step": 13763 + }, + { + "epoch": 2.4504985754985755, + "grad_norm": 0.8022613525390625, + "learning_rate": 6.549271754246822e-05, + "loss": 0.9439, + "step": 13764 + }, + { + "epoch": 2.4506766381766383, + "grad_norm": 0.7447740435600281, + "learning_rate": 6.547958019033024e-05, + "loss": 0.7803, + "step": 13765 + }, + { + "epoch": 2.450854700854701, + "grad_norm": 0.9021183252334595, + "learning_rate": 6.546644351454818e-05, + "loss": 0.8373, + "step": 13766 + }, + { + "epoch": 2.451032763532764, + "grad_norm": 0.8230152726173401, + "learning_rate": 6.545330751537941e-05, + "loss": 0.7023, + "step": 13767 + }, + { + "epoch": 2.451210826210826, + "grad_norm": 0.9581316113471985, + "learning_rate": 6.544017219308132e-05, + "loss": 1.0024, + "step": 13768 + }, + { + "epoch": 2.451388888888889, + "grad_norm": 0.7969945073127747, + "learning_rate": 6.542703754791127e-05, + "loss": 0.8996, + "step": 13769 + }, + { + "epoch": 2.4515669515669516, + "grad_norm": 0.864604115486145, + "learning_rate": 6.54139035801266e-05, + "loss": 0.9325, + "step": 13770 + }, + { + "epoch": 2.4517450142450143, + "grad_norm": 0.8156671524047852, + "learning_rate": 6.540077028998463e-05, + "loss": 0.7926, + "step": 13771 + }, + { + "epoch": 2.451923076923077, + "grad_norm": 0.8704202175140381, + "learning_rate": 6.538763767774272e-05, + "loss": 0.8855, + "step": 13772 + }, + { + "epoch": 2.4521011396011394, + "grad_norm": 0.7533015012741089, + "learning_rate": 6.537450574365811e-05, + "loss": 0.9322, + "step": 13773 + }, + { + "epoch": 2.452279202279202, + "grad_norm": 0.8272553086280823, + "learning_rate": 6.536137448798819e-05, + "loss": 0.7474, + "step": 13774 + }, + { + "epoch": 2.452457264957265, + "grad_norm": 0.7788257598876953, + "learning_rate": 6.534824391099013e-05, + "loss": 0.7163, + "step": 13775 + }, + { + "epoch": 2.4526353276353277, + "grad_norm": 0.8309275507926941, + "learning_rate": 6.533511401292125e-05, + "loss": 1.1595, + "step": 13776 + }, + { + "epoch": 2.4528133903133904, + "grad_norm": 0.8369085788726807, + "learning_rate": 6.53219847940388e-05, + "loss": 0.7211, + "step": 13777 + }, + { + "epoch": 2.452991452991453, + "grad_norm": 0.8571248054504395, + "learning_rate": 6.530885625460007e-05, + "loss": 0.729, + "step": 13778 + }, + { + "epoch": 2.453169515669516, + "grad_norm": 0.7579928040504456, + "learning_rate": 6.529572839486217e-05, + "loss": 0.8799, + "step": 13779 + }, + { + "epoch": 2.453347578347578, + "grad_norm": 0.822463870048523, + "learning_rate": 6.528260121508245e-05, + "loss": 0.7948, + "step": 13780 + }, + { + "epoch": 2.453525641025641, + "grad_norm": 0.7910317778587341, + "learning_rate": 6.526947471551798e-05, + "loss": 0.7727, + "step": 13781 + }, + { + "epoch": 2.4537037037037037, + "grad_norm": 0.9321692585945129, + "learning_rate": 6.525634889642605e-05, + "loss": 0.8754, + "step": 13782 + }, + { + "epoch": 2.4538817663817665, + "grad_norm": 1.0130813121795654, + "learning_rate": 6.524322375806374e-05, + "loss": 0.7845, + "step": 13783 + }, + { + "epoch": 2.4540598290598292, + "grad_norm": 0.7254214882850647, + "learning_rate": 6.52300993006883e-05, + "loss": 0.4665, + "step": 13784 + }, + { + "epoch": 2.4542378917378915, + "grad_norm": 0.7874964475631714, + "learning_rate": 6.521697552455683e-05, + "loss": 0.8535, + "step": 13785 + }, + { + "epoch": 2.4544159544159543, + "grad_norm": 0.8275010585784912, + "learning_rate": 6.520385242992644e-05, + "loss": 0.7744, + "step": 13786 + }, + { + "epoch": 2.454594017094017, + "grad_norm": 0.7972453236579895, + "learning_rate": 6.519073001705431e-05, + "loss": 0.9494, + "step": 13787 + }, + { + "epoch": 2.45477207977208, + "grad_norm": 0.8763988018035889, + "learning_rate": 6.517760828619748e-05, + "loss": 0.8043, + "step": 13788 + }, + { + "epoch": 2.4549501424501425, + "grad_norm": 0.7948910593986511, + "learning_rate": 6.516448723761315e-05, + "loss": 0.7218, + "step": 13789 + }, + { + "epoch": 2.4551282051282053, + "grad_norm": 0.9416671395301819, + "learning_rate": 6.515136687155825e-05, + "loss": 0.7866, + "step": 13790 + }, + { + "epoch": 2.455306267806268, + "grad_norm": 0.8702704906463623, + "learning_rate": 6.513824718828999e-05, + "loss": 1.1579, + "step": 13791 + }, + { + "epoch": 2.4554843304843303, + "grad_norm": 0.8148752450942993, + "learning_rate": 6.51251281880653e-05, + "loss": 0.897, + "step": 13792 + }, + { + "epoch": 2.455662393162393, + "grad_norm": 0.8088299036026001, + "learning_rate": 6.511200987114132e-05, + "loss": 0.939, + "step": 13793 + }, + { + "epoch": 2.455840455840456, + "grad_norm": 0.9836809635162354, + "learning_rate": 6.509889223777499e-05, + "loss": 0.8841, + "step": 13794 + }, + { + "epoch": 2.4560185185185186, + "grad_norm": 0.7677251696586609, + "learning_rate": 6.508577528822342e-05, + "loss": 0.7816, + "step": 13795 + }, + { + "epoch": 2.4561965811965814, + "grad_norm": 0.835421085357666, + "learning_rate": 6.507265902274351e-05, + "loss": 0.9302, + "step": 13796 + }, + { + "epoch": 2.4563746438746437, + "grad_norm": 0.8892473578453064, + "learning_rate": 6.50595434415923e-05, + "loss": 0.8281, + "step": 13797 + }, + { + "epoch": 2.4565527065527064, + "grad_norm": 0.810459315776825, + "learning_rate": 6.504642854502676e-05, + "loss": 0.81, + "step": 13798 + }, + { + "epoch": 2.456730769230769, + "grad_norm": 0.9277065992355347, + "learning_rate": 6.503331433330386e-05, + "loss": 0.7956, + "step": 13799 + }, + { + "epoch": 2.456908831908832, + "grad_norm": 0.861725389957428, + "learning_rate": 6.502020080668051e-05, + "loss": 0.7788, + "step": 13800 + }, + { + "epoch": 2.4570868945868947, + "grad_norm": 1.0000818967819214, + "learning_rate": 6.500708796541366e-05, + "loss": 0.9197, + "step": 13801 + }, + { + "epoch": 2.4572649572649574, + "grad_norm": 0.920998215675354, + "learning_rate": 6.499397580976024e-05, + "loss": 0.7816, + "step": 13802 + }, + { + "epoch": 2.45744301994302, + "grad_norm": 0.7574821710586548, + "learning_rate": 6.498086433997715e-05, + "loss": 0.8982, + "step": 13803 + }, + { + "epoch": 2.4576210826210825, + "grad_norm": 1.026700496673584, + "learning_rate": 6.496775355632125e-05, + "loss": 1.131, + "step": 13804 + }, + { + "epoch": 2.4577991452991452, + "grad_norm": 0.7532633543014526, + "learning_rate": 6.495464345904945e-05, + "loss": 0.7998, + "step": 13805 + }, + { + "epoch": 2.457977207977208, + "grad_norm": 0.7380105257034302, + "learning_rate": 6.494153404841865e-05, + "loss": 0.7656, + "step": 13806 + }, + { + "epoch": 2.4581552706552707, + "grad_norm": 0.7933080792427063, + "learning_rate": 6.492842532468561e-05, + "loss": 0.7419, + "step": 13807 + }, + { + "epoch": 2.4583333333333335, + "grad_norm": 0.7731907963752747, + "learning_rate": 6.491531728810724e-05, + "loss": 0.8334, + "step": 13808 + }, + { + "epoch": 2.458511396011396, + "grad_norm": 0.7368177771568298, + "learning_rate": 6.490220993894035e-05, + "loss": 0.6184, + "step": 13809 + }, + { + "epoch": 2.4586894586894585, + "grad_norm": 0.8381120562553406, + "learning_rate": 6.488910327744178e-05, + "loss": 0.7875, + "step": 13810 + }, + { + "epoch": 2.4588675213675213, + "grad_norm": 0.910142183303833, + "learning_rate": 6.487599730386824e-05, + "loss": 0.8216, + "step": 13811 + }, + { + "epoch": 2.459045584045584, + "grad_norm": 0.9005017876625061, + "learning_rate": 6.48628920184766e-05, + "loss": 0.8928, + "step": 13812 + }, + { + "epoch": 2.459223646723647, + "grad_norm": 0.8437321782112122, + "learning_rate": 6.484978742152358e-05, + "loss": 0.9243, + "step": 13813 + }, + { + "epoch": 2.4594017094017095, + "grad_norm": 0.9145610928535461, + "learning_rate": 6.483668351326599e-05, + "loss": 0.9759, + "step": 13814 + }, + { + "epoch": 2.4595797720797723, + "grad_norm": 0.8391930460929871, + "learning_rate": 6.48235802939605e-05, + "loss": 0.8021, + "step": 13815 + }, + { + "epoch": 2.4597578347578346, + "grad_norm": 0.8035653233528137, + "learning_rate": 6.481047776386394e-05, + "loss": 0.8622, + "step": 13816 + }, + { + "epoch": 2.4599358974358974, + "grad_norm": 0.7238573431968689, + "learning_rate": 6.479737592323291e-05, + "loss": 0.6864, + "step": 13817 + }, + { + "epoch": 2.46011396011396, + "grad_norm": 0.8629193902015686, + "learning_rate": 6.47842747723242e-05, + "loss": 0.6107, + "step": 13818 + }, + { + "epoch": 2.460292022792023, + "grad_norm": 0.9015333652496338, + "learning_rate": 6.477117431139444e-05, + "loss": 0.845, + "step": 13819 + }, + { + "epoch": 2.4604700854700856, + "grad_norm": 0.8544989824295044, + "learning_rate": 6.47580745407004e-05, + "loss": 0.7103, + "step": 13820 + }, + { + "epoch": 2.460648148148148, + "grad_norm": 0.7665401101112366, + "learning_rate": 6.474497546049862e-05, + "loss": 0.6864, + "step": 13821 + }, + { + "epoch": 2.4608262108262107, + "grad_norm": 0.8640002012252808, + "learning_rate": 6.473187707104584e-05, + "loss": 0.953, + "step": 13822 + }, + { + "epoch": 2.4610042735042734, + "grad_norm": 0.8563477396965027, + "learning_rate": 6.471877937259864e-05, + "loss": 0.7776, + "step": 13823 + }, + { + "epoch": 2.461182336182336, + "grad_norm": 0.8089157938957214, + "learning_rate": 6.470568236541371e-05, + "loss": 0.8273, + "step": 13824 + }, + { + "epoch": 2.461360398860399, + "grad_norm": 0.8710005283355713, + "learning_rate": 6.469258604974757e-05, + "loss": 0.8827, + "step": 13825 + }, + { + "epoch": 2.4615384615384617, + "grad_norm": 0.7780489325523376, + "learning_rate": 6.467949042585688e-05, + "loss": 0.8035, + "step": 13826 + }, + { + "epoch": 2.4617165242165244, + "grad_norm": 1.010976791381836, + "learning_rate": 6.466639549399822e-05, + "loss": 0.7442, + "step": 13827 + }, + { + "epoch": 2.4618945868945867, + "grad_norm": 0.714077889919281, + "learning_rate": 6.465330125442812e-05, + "loss": 0.694, + "step": 13828 + }, + { + "epoch": 2.4620726495726495, + "grad_norm": 0.7408512234687805, + "learning_rate": 6.464020770740316e-05, + "loss": 0.6709, + "step": 13829 + }, + { + "epoch": 2.4622507122507122, + "grad_norm": 0.8433945178985596, + "learning_rate": 6.462711485317987e-05, + "loss": 0.9127, + "step": 13830 + }, + { + "epoch": 2.462428774928775, + "grad_norm": 1.031745195388794, + "learning_rate": 6.461402269201481e-05, + "loss": 0.9105, + "step": 13831 + }, + { + "epoch": 2.4626068376068377, + "grad_norm": 0.8884360790252686, + "learning_rate": 6.460093122416444e-05, + "loss": 0.9354, + "step": 13832 + }, + { + "epoch": 2.4627849002849, + "grad_norm": 0.8466372489929199, + "learning_rate": 6.45878404498853e-05, + "loss": 0.9554, + "step": 13833 + }, + { + "epoch": 2.462962962962963, + "grad_norm": 0.9026118516921997, + "learning_rate": 6.457475036943386e-05, + "loss": 0.928, + "step": 13834 + }, + { + "epoch": 2.4631410256410255, + "grad_norm": 0.9034590721130371, + "learning_rate": 6.456166098306661e-05, + "loss": 0.7694, + "step": 13835 + }, + { + "epoch": 2.4633190883190883, + "grad_norm": 0.8369483947753906, + "learning_rate": 6.454857229103998e-05, + "loss": 0.928, + "step": 13836 + }, + { + "epoch": 2.463497150997151, + "grad_norm": 0.8670645356178284, + "learning_rate": 6.453548429361045e-05, + "loss": 0.971, + "step": 13837 + }, + { + "epoch": 2.463675213675214, + "grad_norm": 0.8415539860725403, + "learning_rate": 6.452239699103442e-05, + "loss": 0.8461, + "step": 13838 + }, + { + "epoch": 2.4638532763532766, + "grad_norm": 0.7434490323066711, + "learning_rate": 6.450931038356834e-05, + "loss": 0.6677, + "step": 13839 + }, + { + "epoch": 2.464031339031339, + "grad_norm": 0.8113850355148315, + "learning_rate": 6.449622447146855e-05, + "loss": 0.8644, + "step": 13840 + }, + { + "epoch": 2.4642094017094016, + "grad_norm": 0.7424083352088928, + "learning_rate": 6.448313925499154e-05, + "loss": 0.7469, + "step": 13841 + }, + { + "epoch": 2.4643874643874644, + "grad_norm": 1.006949782371521, + "learning_rate": 6.44700547343936e-05, + "loss": 0.8426, + "step": 13842 + }, + { + "epoch": 2.464565527065527, + "grad_norm": 1.0643857717514038, + "learning_rate": 6.445697090993117e-05, + "loss": 0.7793, + "step": 13843 + }, + { + "epoch": 2.46474358974359, + "grad_norm": 0.8716835975646973, + "learning_rate": 6.444388778186051e-05, + "loss": 0.9684, + "step": 13844 + }, + { + "epoch": 2.464921652421652, + "grad_norm": 0.8677120804786682, + "learning_rate": 6.443080535043802e-05, + "loss": 0.9772, + "step": 13845 + }, + { + "epoch": 2.465099715099715, + "grad_norm": 0.8955141305923462, + "learning_rate": 6.441772361592005e-05, + "loss": 0.881, + "step": 13846 + }, + { + "epoch": 2.4652777777777777, + "grad_norm": 0.806794285774231, + "learning_rate": 6.440464257856283e-05, + "loss": 0.6393, + "step": 13847 + }, + { + "epoch": 2.4654558404558404, + "grad_norm": 0.8438352346420288, + "learning_rate": 6.439156223862272e-05, + "loss": 0.8305, + "step": 13848 + }, + { + "epoch": 2.465633903133903, + "grad_norm": 0.828960120677948, + "learning_rate": 6.437848259635594e-05, + "loss": 0.8564, + "step": 13849 + }, + { + "epoch": 2.465811965811966, + "grad_norm": 0.790199339389801, + "learning_rate": 6.436540365201886e-05, + "loss": 0.6573, + "step": 13850 + }, + { + "epoch": 2.4659900284900287, + "grad_norm": 0.8476296663284302, + "learning_rate": 6.435232540586763e-05, + "loss": 0.6979, + "step": 13851 + }, + { + "epoch": 2.466168091168091, + "grad_norm": 0.6880464553833008, + "learning_rate": 6.433924785815857e-05, + "loss": 0.6925, + "step": 13852 + }, + { + "epoch": 2.4663461538461537, + "grad_norm": 0.9369434118270874, + "learning_rate": 6.432617100914782e-05, + "loss": 0.6891, + "step": 13853 + }, + { + "epoch": 2.4665242165242165, + "grad_norm": 0.8522159457206726, + "learning_rate": 6.431309485909166e-05, + "loss": 1.0786, + "step": 13854 + }, + { + "epoch": 2.4667022792022792, + "grad_norm": 0.8479002714157104, + "learning_rate": 6.430001940824625e-05, + "loss": 0.9293, + "step": 13855 + }, + { + "epoch": 2.466880341880342, + "grad_norm": 0.8382098078727722, + "learning_rate": 6.428694465686787e-05, + "loss": 1.1279, + "step": 13856 + }, + { + "epoch": 2.4670584045584047, + "grad_norm": 0.9641128182411194, + "learning_rate": 6.427387060521255e-05, + "loss": 0.7938, + "step": 13857 + }, + { + "epoch": 2.467236467236467, + "grad_norm": 0.8458924293518066, + "learning_rate": 6.426079725353656e-05, + "loss": 0.7804, + "step": 13858 + }, + { + "epoch": 2.46741452991453, + "grad_norm": 0.8023849725723267, + "learning_rate": 6.424772460209597e-05, + "loss": 0.718, + "step": 13859 + }, + { + "epoch": 2.4675925925925926, + "grad_norm": 0.800864040851593, + "learning_rate": 6.423465265114699e-05, + "loss": 0.7627, + "step": 13860 + }, + { + "epoch": 2.4677706552706553, + "grad_norm": 0.9159586429595947, + "learning_rate": 6.422158140094566e-05, + "loss": 0.8348, + "step": 13861 + }, + { + "epoch": 2.467948717948718, + "grad_norm": 0.7982872724533081, + "learning_rate": 6.420851085174817e-05, + "loss": 0.8092, + "step": 13862 + }, + { + "epoch": 2.468126780626781, + "grad_norm": 0.8847397565841675, + "learning_rate": 6.41954410038105e-05, + "loss": 0.9165, + "step": 13863 + }, + { + "epoch": 2.468304843304843, + "grad_norm": 0.7885190844535828, + "learning_rate": 6.418237185738882e-05, + "loss": 0.733, + "step": 13864 + }, + { + "epoch": 2.468482905982906, + "grad_norm": 0.902428150177002, + "learning_rate": 6.416930341273914e-05, + "loss": 0.8345, + "step": 13865 + }, + { + "epoch": 2.4686609686609686, + "grad_norm": 0.9344130158424377, + "learning_rate": 6.415623567011751e-05, + "loss": 0.8651, + "step": 13866 + }, + { + "epoch": 2.4688390313390314, + "grad_norm": 0.840679407119751, + "learning_rate": 6.414316862978003e-05, + "loss": 0.7534, + "step": 13867 + }, + { + "epoch": 2.469017094017094, + "grad_norm": 0.8799613118171692, + "learning_rate": 6.413010229198263e-05, + "loss": 0.9335, + "step": 13868 + }, + { + "epoch": 2.469195156695157, + "grad_norm": 1.1401816606521606, + "learning_rate": 6.411703665698142e-05, + "loss": 0.8902, + "step": 13869 + }, + { + "epoch": 2.469373219373219, + "grad_norm": 0.9733933210372925, + "learning_rate": 6.410397172503227e-05, + "loss": 0.9288, + "step": 13870 + }, + { + "epoch": 2.469551282051282, + "grad_norm": 0.9260223507881165, + "learning_rate": 6.409090749639128e-05, + "loss": 0.9664, + "step": 13871 + }, + { + "epoch": 2.4697293447293447, + "grad_norm": 1.0200423002243042, + "learning_rate": 6.407784397131433e-05, + "loss": 0.9095, + "step": 13872 + }, + { + "epoch": 2.4699074074074074, + "grad_norm": 0.8015561103820801, + "learning_rate": 6.406478115005743e-05, + "loss": 0.8015, + "step": 13873 + }, + { + "epoch": 2.47008547008547, + "grad_norm": 0.8035915493965149, + "learning_rate": 6.40517190328765e-05, + "loss": 0.9275, + "step": 13874 + }, + { + "epoch": 2.470263532763533, + "grad_norm": 0.8248090147972107, + "learning_rate": 6.403865762002743e-05, + "loss": 0.8736, + "step": 13875 + }, + { + "epoch": 2.4704415954415953, + "grad_norm": 0.9310920834541321, + "learning_rate": 6.402559691176616e-05, + "loss": 0.794, + "step": 13876 + }, + { + "epoch": 2.470619658119658, + "grad_norm": 0.7796428799629211, + "learning_rate": 6.401253690834863e-05, + "loss": 0.832, + "step": 13877 + }, + { + "epoch": 2.4707977207977208, + "grad_norm": 0.9046199321746826, + "learning_rate": 6.399947761003063e-05, + "loss": 0.9988, + "step": 13878 + }, + { + "epoch": 2.4709757834757835, + "grad_norm": 0.8970019221305847, + "learning_rate": 6.398641901706812e-05, + "loss": 0.6805, + "step": 13879 + }, + { + "epoch": 2.4711538461538463, + "grad_norm": 0.935786247253418, + "learning_rate": 6.397336112971688e-05, + "loss": 0.9826, + "step": 13880 + }, + { + "epoch": 2.471331908831909, + "grad_norm": 0.8497617244720459, + "learning_rate": 6.396030394823285e-05, + "loss": 0.8842, + "step": 13881 + }, + { + "epoch": 2.4715099715099713, + "grad_norm": 0.9159898161888123, + "learning_rate": 6.394724747287173e-05, + "loss": 0.8592, + "step": 13882 + }, + { + "epoch": 2.471688034188034, + "grad_norm": 0.891951858997345, + "learning_rate": 6.393419170388943e-05, + "loss": 0.8197, + "step": 13883 + }, + { + "epoch": 2.471866096866097, + "grad_norm": 0.9038097858428955, + "learning_rate": 6.392113664154172e-05, + "loss": 0.7535, + "step": 13884 + }, + { + "epoch": 2.4720441595441596, + "grad_norm": 0.8502489924430847, + "learning_rate": 6.390808228608438e-05, + "loss": 0.9183, + "step": 13885 + }, + { + "epoch": 2.4722222222222223, + "grad_norm": 0.7442654371261597, + "learning_rate": 6.389502863777323e-05, + "loss": 0.7741, + "step": 13886 + }, + { + "epoch": 2.472400284900285, + "grad_norm": 0.8811324238777161, + "learning_rate": 6.388197569686395e-05, + "loss": 0.7261, + "step": 13887 + }, + { + "epoch": 2.472578347578348, + "grad_norm": 0.9487552046775818, + "learning_rate": 6.386892346361239e-05, + "loss": 0.779, + "step": 13888 + }, + { + "epoch": 2.47275641025641, + "grad_norm": 0.8138917088508606, + "learning_rate": 6.385587193827416e-05, + "loss": 0.915, + "step": 13889 + }, + { + "epoch": 2.472934472934473, + "grad_norm": 0.7842695713043213, + "learning_rate": 6.384282112110506e-05, + "loss": 0.9458, + "step": 13890 + }, + { + "epoch": 2.4731125356125356, + "grad_norm": 0.7852116823196411, + "learning_rate": 6.382977101236074e-05, + "loss": 0.9515, + "step": 13891 + }, + { + "epoch": 2.4732905982905984, + "grad_norm": 0.8429296016693115, + "learning_rate": 6.381672161229698e-05, + "loss": 0.9466, + "step": 13892 + }, + { + "epoch": 2.473468660968661, + "grad_norm": 0.8713327050209045, + "learning_rate": 6.380367292116933e-05, + "loss": 0.7552, + "step": 13893 + }, + { + "epoch": 2.4736467236467234, + "grad_norm": 0.8153441548347473, + "learning_rate": 6.379062493923355e-05, + "loss": 0.7833, + "step": 13894 + }, + { + "epoch": 2.473824786324786, + "grad_norm": 0.8283601999282837, + "learning_rate": 6.377757766674526e-05, + "loss": 0.8986, + "step": 13895 + }, + { + "epoch": 2.474002849002849, + "grad_norm": 0.8116408586502075, + "learning_rate": 6.37645311039601e-05, + "loss": 0.8549, + "step": 13896 + }, + { + "epoch": 2.4741809116809117, + "grad_norm": 0.7999116778373718, + "learning_rate": 6.375148525113365e-05, + "loss": 0.8082, + "step": 13897 + }, + { + "epoch": 2.4743589743589745, + "grad_norm": 0.7307565808296204, + "learning_rate": 6.373844010852159e-05, + "loss": 0.7946, + "step": 13898 + }, + { + "epoch": 2.474537037037037, + "grad_norm": 0.7519806027412415, + "learning_rate": 6.372539567637941e-05, + "loss": 0.6511, + "step": 13899 + }, + { + "epoch": 2.4747150997151, + "grad_norm": 0.8571820259094238, + "learning_rate": 6.371235195496279e-05, + "loss": 0.8266, + "step": 13900 + }, + { + "epoch": 2.4748931623931623, + "grad_norm": 0.8118062019348145, + "learning_rate": 6.369930894452723e-05, + "loss": 0.8573, + "step": 13901 + }, + { + "epoch": 2.475071225071225, + "grad_norm": 0.8729892373085022, + "learning_rate": 6.368626664532833e-05, + "loss": 0.812, + "step": 13902 + }, + { + "epoch": 2.4752492877492878, + "grad_norm": 0.7663209438323975, + "learning_rate": 6.367322505762157e-05, + "loss": 0.6648, + "step": 13903 + }, + { + "epoch": 2.4754273504273505, + "grad_norm": 0.7913058996200562, + "learning_rate": 6.366018418166251e-05, + "loss": 0.7486, + "step": 13904 + }, + { + "epoch": 2.4756054131054133, + "grad_norm": 0.7714928984642029, + "learning_rate": 6.364714401770666e-05, + "loss": 0.9134, + "step": 13905 + }, + { + "epoch": 2.4757834757834756, + "grad_norm": 0.8226378560066223, + "learning_rate": 6.363410456600949e-05, + "loss": 0.903, + "step": 13906 + }, + { + "epoch": 2.4759615384615383, + "grad_norm": 0.8643919825553894, + "learning_rate": 6.362106582682653e-05, + "loss": 0.9068, + "step": 13907 + }, + { + "epoch": 2.476139601139601, + "grad_norm": 0.8390868306159973, + "learning_rate": 6.360802780041317e-05, + "loss": 0.8938, + "step": 13908 + }, + { + "epoch": 2.476317663817664, + "grad_norm": 0.9549261927604675, + "learning_rate": 6.359499048702495e-05, + "loss": 0.8431, + "step": 13909 + }, + { + "epoch": 2.4764957264957266, + "grad_norm": 0.7509152889251709, + "learning_rate": 6.358195388691726e-05, + "loss": 0.8661, + "step": 13910 + }, + { + "epoch": 2.4766737891737893, + "grad_norm": 0.9447416067123413, + "learning_rate": 6.356891800034552e-05, + "loss": 0.8766, + "step": 13911 + }, + { + "epoch": 2.476851851851852, + "grad_norm": 0.8972395062446594, + "learning_rate": 6.355588282756515e-05, + "loss": 0.8517, + "step": 13912 + }, + { + "epoch": 2.4770299145299144, + "grad_norm": 0.8645047545433044, + "learning_rate": 6.354284836883156e-05, + "loss": 0.8188, + "step": 13913 + }, + { + "epoch": 2.477207977207977, + "grad_norm": 0.7939230799674988, + "learning_rate": 6.35298146244001e-05, + "loss": 0.778, + "step": 13914 + }, + { + "epoch": 2.47738603988604, + "grad_norm": 0.7714613676071167, + "learning_rate": 6.351678159452618e-05, + "loss": 0.7964, + "step": 13915 + }, + { + "epoch": 2.4775641025641026, + "grad_norm": 0.9293754696846008, + "learning_rate": 6.350374927946512e-05, + "loss": 0.7965, + "step": 13916 + }, + { + "epoch": 2.4777421652421654, + "grad_norm": 0.8025050759315491, + "learning_rate": 6.349071767947233e-05, + "loss": 0.8222, + "step": 13917 + }, + { + "epoch": 2.4779202279202277, + "grad_norm": 0.7790399193763733, + "learning_rate": 6.347768679480304e-05, + "loss": 1.0294, + "step": 13918 + }, + { + "epoch": 2.4780982905982905, + "grad_norm": 1.0063512325286865, + "learning_rate": 6.346465662571261e-05, + "loss": 0.953, + "step": 13919 + }, + { + "epoch": 2.478276353276353, + "grad_norm": 0.8742708563804626, + "learning_rate": 6.345162717245634e-05, + "loss": 0.7119, + "step": 13920 + }, + { + "epoch": 2.478454415954416, + "grad_norm": 0.760497510433197, + "learning_rate": 6.343859843528955e-05, + "loss": 0.7446, + "step": 13921 + }, + { + "epoch": 2.4786324786324787, + "grad_norm": 0.7635362148284912, + "learning_rate": 6.342557041446743e-05, + "loss": 0.8286, + "step": 13922 + }, + { + "epoch": 2.4788105413105415, + "grad_norm": 0.7876720428466797, + "learning_rate": 6.341254311024532e-05, + "loss": 0.8007, + "step": 13923 + }, + { + "epoch": 2.478988603988604, + "grad_norm": 0.8144401907920837, + "learning_rate": 6.339951652287839e-05, + "loss": 0.9177, + "step": 13924 + }, + { + "epoch": 2.4791666666666665, + "grad_norm": 0.7400189638137817, + "learning_rate": 6.338649065262189e-05, + "loss": 0.8935, + "step": 13925 + }, + { + "epoch": 2.4793447293447293, + "grad_norm": 0.8412175178527832, + "learning_rate": 6.337346549973106e-05, + "loss": 0.6929, + "step": 13926 + }, + { + "epoch": 2.479522792022792, + "grad_norm": 1.0156967639923096, + "learning_rate": 6.336044106446108e-05, + "loss": 0.9843, + "step": 13927 + }, + { + "epoch": 2.4797008547008548, + "grad_norm": 0.8556809425354004, + "learning_rate": 6.33474173470672e-05, + "loss": 0.81, + "step": 13928 + }, + { + "epoch": 2.4798789173789175, + "grad_norm": 0.8103616237640381, + "learning_rate": 6.333439434780448e-05, + "loss": 0.925, + "step": 13929 + }, + { + "epoch": 2.48005698005698, + "grad_norm": 0.9460168480873108, + "learning_rate": 6.332137206692817e-05, + "loss": 0.7059, + "step": 13930 + }, + { + "epoch": 2.4802350427350426, + "grad_norm": 0.847226619720459, + "learning_rate": 6.330835050469334e-05, + "loss": 1.0139, + "step": 13931 + }, + { + "epoch": 2.4804131054131053, + "grad_norm": 0.7639240622520447, + "learning_rate": 6.329532966135523e-05, + "loss": 0.8141, + "step": 13932 + }, + { + "epoch": 2.480591168091168, + "grad_norm": 0.9273494482040405, + "learning_rate": 6.328230953716883e-05, + "loss": 0.9864, + "step": 13933 + }, + { + "epoch": 2.480769230769231, + "grad_norm": 0.7709840536117554, + "learning_rate": 6.326929013238934e-05, + "loss": 0.7275, + "step": 13934 + }, + { + "epoch": 2.4809472934472936, + "grad_norm": 0.755933940410614, + "learning_rate": 6.325627144727177e-05, + "loss": 0.7785, + "step": 13935 + }, + { + "epoch": 2.4811253561253563, + "grad_norm": 0.9058536291122437, + "learning_rate": 6.324325348207125e-05, + "loss": 0.9694, + "step": 13936 + }, + { + "epoch": 2.4813034188034186, + "grad_norm": 0.8490056395530701, + "learning_rate": 6.323023623704282e-05, + "loss": 0.8882, + "step": 13937 + }, + { + "epoch": 2.4814814814814814, + "grad_norm": 0.9559429883956909, + "learning_rate": 6.321721971244155e-05, + "loss": 0.9243, + "step": 13938 + }, + { + "epoch": 2.481659544159544, + "grad_norm": 0.8607096076011658, + "learning_rate": 6.320420390852242e-05, + "loss": 0.9281, + "step": 13939 + }, + { + "epoch": 2.481837606837607, + "grad_norm": 1.1263439655303955, + "learning_rate": 6.319118882554049e-05, + "loss": 0.8772, + "step": 13940 + }, + { + "epoch": 2.4820156695156697, + "grad_norm": 0.9691354632377625, + "learning_rate": 6.317817446375074e-05, + "loss": 0.9349, + "step": 13941 + }, + { + "epoch": 2.482193732193732, + "grad_norm": 0.8636828064918518, + "learning_rate": 6.31651608234082e-05, + "loss": 0.9021, + "step": 13942 + }, + { + "epoch": 2.4823717948717947, + "grad_norm": 0.8405864238739014, + "learning_rate": 6.315214790476777e-05, + "loss": 0.8917, + "step": 13943 + }, + { + "epoch": 2.4825498575498575, + "grad_norm": 0.8082821369171143, + "learning_rate": 6.313913570808448e-05, + "loss": 0.914, + "step": 13944 + }, + { + "epoch": 2.48272792022792, + "grad_norm": 0.8734335899353027, + "learning_rate": 6.312612423361328e-05, + "loss": 0.6885, + "step": 13945 + }, + { + "epoch": 2.482905982905983, + "grad_norm": 0.943190336227417, + "learning_rate": 6.311311348160904e-05, + "loss": 0.894, + "step": 13946 + }, + { + "epoch": 2.4830840455840457, + "grad_norm": 0.9425446391105652, + "learning_rate": 6.310010345232673e-05, + "loss": 0.7893, + "step": 13947 + }, + { + "epoch": 2.4832621082621085, + "grad_norm": 0.9668664932250977, + "learning_rate": 6.308709414602123e-05, + "loss": 0.9715, + "step": 13948 + }, + { + "epoch": 2.4834401709401708, + "grad_norm": 0.7771579027175903, + "learning_rate": 6.307408556294747e-05, + "loss": 0.872, + "step": 13949 + }, + { + "epoch": 2.4836182336182335, + "grad_norm": 0.8183084726333618, + "learning_rate": 6.306107770336025e-05, + "loss": 1.0008, + "step": 13950 + }, + { + "epoch": 2.4837962962962963, + "grad_norm": 0.8284399509429932, + "learning_rate": 6.304807056751452e-05, + "loss": 0.687, + "step": 13951 + }, + { + "epoch": 2.483974358974359, + "grad_norm": 0.8663082718849182, + "learning_rate": 6.303506415566504e-05, + "loss": 0.8792, + "step": 13952 + }, + { + "epoch": 2.484152421652422, + "grad_norm": 0.8272001147270203, + "learning_rate": 6.302205846806675e-05, + "loss": 0.9255, + "step": 13953 + }, + { + "epoch": 2.484330484330484, + "grad_norm": 0.9398671388626099, + "learning_rate": 6.300905350497437e-05, + "loss": 0.9605, + "step": 13954 + }, + { + "epoch": 2.484508547008547, + "grad_norm": 0.7368931770324707, + "learning_rate": 6.299604926664276e-05, + "loss": 0.6734, + "step": 13955 + }, + { + "epoch": 2.4846866096866096, + "grad_norm": 0.8315541744232178, + "learning_rate": 6.298304575332668e-05, + "loss": 0.9578, + "step": 13956 + }, + { + "epoch": 2.4848646723646723, + "grad_norm": 0.8656954169273376, + "learning_rate": 6.297004296528095e-05, + "loss": 0.8897, + "step": 13957 + }, + { + "epoch": 2.485042735042735, + "grad_norm": 0.9047118425369263, + "learning_rate": 6.295704090276026e-05, + "loss": 1.0308, + "step": 13958 + }, + { + "epoch": 2.485220797720798, + "grad_norm": 0.8771422505378723, + "learning_rate": 6.294403956601946e-05, + "loss": 0.9505, + "step": 13959 + }, + { + "epoch": 2.4853988603988606, + "grad_norm": 0.8935427069664001, + "learning_rate": 6.293103895531319e-05, + "loss": 1.0485, + "step": 13960 + }, + { + "epoch": 2.485576923076923, + "grad_norm": 0.7345624566078186, + "learning_rate": 6.291803907089621e-05, + "loss": 0.7485, + "step": 13961 + }, + { + "epoch": 2.4857549857549857, + "grad_norm": 0.6415224075317383, + "learning_rate": 6.290503991302324e-05, + "loss": 0.6237, + "step": 13962 + }, + { + "epoch": 2.4859330484330484, + "grad_norm": 0.8547754883766174, + "learning_rate": 6.289204148194896e-05, + "loss": 0.9152, + "step": 13963 + }, + { + "epoch": 2.486111111111111, + "grad_norm": 0.7434722185134888, + "learning_rate": 6.2879043777928e-05, + "loss": 0.8499, + "step": 13964 + }, + { + "epoch": 2.486289173789174, + "grad_norm": 0.8609980940818787, + "learning_rate": 6.286604680121509e-05, + "loss": 0.8101, + "step": 13965 + }, + { + "epoch": 2.486467236467236, + "grad_norm": 0.8709290027618408, + "learning_rate": 6.285305055206486e-05, + "loss": 0.8335, + "step": 13966 + }, + { + "epoch": 2.486645299145299, + "grad_norm": 0.7758293747901917, + "learning_rate": 6.284005503073191e-05, + "loss": 0.8408, + "step": 13967 + }, + { + "epoch": 2.4868233618233617, + "grad_norm": 0.9778353571891785, + "learning_rate": 6.282706023747094e-05, + "loss": 0.8595, + "step": 13968 + }, + { + "epoch": 2.4870014245014245, + "grad_norm": 0.8438369631767273, + "learning_rate": 6.281406617253646e-05, + "loss": 0.8843, + "step": 13969 + }, + { + "epoch": 2.4871794871794872, + "grad_norm": 0.9704681634902954, + "learning_rate": 6.280107283618315e-05, + "loss": 0.62, + "step": 13970 + }, + { + "epoch": 2.48735754985755, + "grad_norm": 0.7774441242218018, + "learning_rate": 6.278808022866549e-05, + "loss": 0.9087, + "step": 13971 + }, + { + "epoch": 2.4875356125356127, + "grad_norm": 0.8387142419815063, + "learning_rate": 6.277508835023813e-05, + "loss": 0.854, + "step": 13972 + }, + { + "epoch": 2.487713675213675, + "grad_norm": 0.8483029007911682, + "learning_rate": 6.276209720115556e-05, + "loss": 0.8665, + "step": 13973 + }, + { + "epoch": 2.487891737891738, + "grad_norm": 0.8251432180404663, + "learning_rate": 6.274910678167239e-05, + "loss": 0.7816, + "step": 13974 + }, + { + "epoch": 2.4880698005698005, + "grad_norm": 0.8503836989402771, + "learning_rate": 6.273611709204304e-05, + "loss": 0.939, + "step": 13975 + }, + { + "epoch": 2.4882478632478633, + "grad_norm": 0.6545158624649048, + "learning_rate": 6.27231281325221e-05, + "loss": 0.564, + "step": 13976 + }, + { + "epoch": 2.488425925925926, + "grad_norm": 0.7353499531745911, + "learning_rate": 6.2710139903364e-05, + "loss": 0.7103, + "step": 13977 + }, + { + "epoch": 2.488603988603989, + "grad_norm": 0.9032405614852905, + "learning_rate": 6.269715240482327e-05, + "loss": 0.8802, + "step": 13978 + }, + { + "epoch": 2.488782051282051, + "grad_norm": 0.8141019940376282, + "learning_rate": 6.268416563715434e-05, + "loss": 0.9836, + "step": 13979 + }, + { + "epoch": 2.488960113960114, + "grad_norm": 0.9087637066841125, + "learning_rate": 6.267117960061167e-05, + "loss": 0.8488, + "step": 13980 + }, + { + "epoch": 2.4891381766381766, + "grad_norm": 0.6649556756019592, + "learning_rate": 6.265819429544969e-05, + "loss": 0.604, + "step": 13981 + }, + { + "epoch": 2.4893162393162394, + "grad_norm": 0.7872918248176575, + "learning_rate": 6.264520972192283e-05, + "loss": 0.8537, + "step": 13982 + }, + { + "epoch": 2.489494301994302, + "grad_norm": 0.8821072578430176, + "learning_rate": 6.263222588028546e-05, + "loss": 0.7485, + "step": 13983 + }, + { + "epoch": 2.489672364672365, + "grad_norm": 0.8077933192253113, + "learning_rate": 6.2619242770792e-05, + "loss": 0.9174, + "step": 13984 + }, + { + "epoch": 2.489850427350427, + "grad_norm": 0.8087183237075806, + "learning_rate": 6.260626039369686e-05, + "loss": 0.9184, + "step": 13985 + }, + { + "epoch": 2.49002849002849, + "grad_norm": 0.8849205374717712, + "learning_rate": 6.259327874925434e-05, + "loss": 0.9095, + "step": 13986 + }, + { + "epoch": 2.4902065527065527, + "grad_norm": 0.7899976968765259, + "learning_rate": 6.258029783771884e-05, + "loss": 0.7286, + "step": 13987 + }, + { + "epoch": 2.4903846153846154, + "grad_norm": 0.7907543182373047, + "learning_rate": 6.256731765934464e-05, + "loss": 0.8433, + "step": 13988 + }, + { + "epoch": 2.490562678062678, + "grad_norm": 0.7794694304466248, + "learning_rate": 6.255433821438614e-05, + "loss": 0.7197, + "step": 13989 + }, + { + "epoch": 2.490740740740741, + "grad_norm": 0.8443161249160767, + "learning_rate": 6.254135950309753e-05, + "loss": 0.7982, + "step": 13990 + }, + { + "epoch": 2.4909188034188032, + "grad_norm": 0.972024142742157, + "learning_rate": 6.252838152573323e-05, + "loss": 0.9212, + "step": 13991 + }, + { + "epoch": 2.491096866096866, + "grad_norm": 0.8017764687538147, + "learning_rate": 6.25154042825474e-05, + "loss": 0.9544, + "step": 13992 + }, + { + "epoch": 2.4912749287749287, + "grad_norm": 0.8622884154319763, + "learning_rate": 6.250242777379442e-05, + "loss": 0.8411, + "step": 13993 + }, + { + "epoch": 2.4914529914529915, + "grad_norm": 0.7384446263313293, + "learning_rate": 6.248945199972842e-05, + "loss": 0.8357, + "step": 13994 + }, + { + "epoch": 2.4916310541310542, + "grad_norm": 0.7748960256576538, + "learning_rate": 6.247647696060372e-05, + "loss": 0.7739, + "step": 13995 + }, + { + "epoch": 2.491809116809117, + "grad_norm": 0.8295742273330688, + "learning_rate": 6.246350265667448e-05, + "loss": 0.8032, + "step": 13996 + }, + { + "epoch": 2.4919871794871793, + "grad_norm": 0.8604934811592102, + "learning_rate": 6.245052908819494e-05, + "loss": 0.8738, + "step": 13997 + }, + { + "epoch": 2.492165242165242, + "grad_norm": 0.8381406664848328, + "learning_rate": 6.243755625541926e-05, + "loss": 0.8351, + "step": 13998 + }, + { + "epoch": 2.492343304843305, + "grad_norm": 0.9238134026527405, + "learning_rate": 6.242458415860168e-05, + "loss": 0.9529, + "step": 13999 + }, + { + "epoch": 2.4925213675213675, + "grad_norm": 0.9234444499015808, + "learning_rate": 6.241161279799628e-05, + "loss": 1.1086, + "step": 14000 + }, + { + "epoch": 2.4926994301994303, + "grad_norm": 0.8056737780570984, + "learning_rate": 6.239864217385727e-05, + "loss": 0.7957, + "step": 14001 + }, + { + "epoch": 2.492877492877493, + "grad_norm": 0.7877696752548218, + "learning_rate": 6.238567228643872e-05, + "loss": 0.9577, + "step": 14002 + }, + { + "epoch": 2.4930555555555554, + "grad_norm": 0.7437340021133423, + "learning_rate": 6.237270313599479e-05, + "loss": 0.6171, + "step": 14003 + }, + { + "epoch": 2.493233618233618, + "grad_norm": 0.8503403067588806, + "learning_rate": 6.235973472277962e-05, + "loss": 0.7608, + "step": 14004 + }, + { + "epoch": 2.493411680911681, + "grad_norm": 0.8557562232017517, + "learning_rate": 6.234676704704722e-05, + "loss": 0.8414, + "step": 14005 + }, + { + "epoch": 2.4935897435897436, + "grad_norm": 0.9188289046287537, + "learning_rate": 6.233380010905174e-05, + "loss": 0.89, + "step": 14006 + }, + { + "epoch": 2.4937678062678064, + "grad_norm": 0.9433556199073792, + "learning_rate": 6.232083390904716e-05, + "loss": 0.9316, + "step": 14007 + }, + { + "epoch": 2.493945868945869, + "grad_norm": 0.9278882145881653, + "learning_rate": 6.230786844728759e-05, + "loss": 0.9211, + "step": 14008 + }, + { + "epoch": 2.494123931623932, + "grad_norm": 0.8365640640258789, + "learning_rate": 6.229490372402702e-05, + "loss": 0.9578, + "step": 14009 + }, + { + "epoch": 2.494301994301994, + "grad_norm": 0.7987647652626038, + "learning_rate": 6.228193973951953e-05, + "loss": 0.9279, + "step": 14010 + }, + { + "epoch": 2.494480056980057, + "grad_norm": 0.7707502841949463, + "learning_rate": 6.226897649401902e-05, + "loss": 0.8879, + "step": 14011 + }, + { + "epoch": 2.4946581196581197, + "grad_norm": 0.8623191118240356, + "learning_rate": 6.225601398777957e-05, + "loss": 0.7427, + "step": 14012 + }, + { + "epoch": 2.4948361823361824, + "grad_norm": 0.8470782041549683, + "learning_rate": 6.22430522210551e-05, + "loss": 0.9425, + "step": 14013 + }, + { + "epoch": 2.495014245014245, + "grad_norm": 0.9169524908065796, + "learning_rate": 6.223009119409963e-05, + "loss": 0.9595, + "step": 14014 + }, + { + "epoch": 2.4951923076923075, + "grad_norm": 0.8541738986968994, + "learning_rate": 6.221713090716701e-05, + "loss": 1.0726, + "step": 14015 + }, + { + "epoch": 2.4953703703703702, + "grad_norm": 0.8801444172859192, + "learning_rate": 6.220417136051126e-05, + "loss": 0.8323, + "step": 14016 + }, + { + "epoch": 2.495548433048433, + "grad_norm": 0.845448911190033, + "learning_rate": 6.219121255438624e-05, + "loss": 0.835, + "step": 14017 + }, + { + "epoch": 2.4957264957264957, + "grad_norm": 0.7653858661651611, + "learning_rate": 6.217825448904588e-05, + "loss": 0.7027, + "step": 14018 + }, + { + "epoch": 2.4959045584045585, + "grad_norm": 0.7779282927513123, + "learning_rate": 6.216529716474404e-05, + "loss": 0.7881, + "step": 14019 + }, + { + "epoch": 2.4960826210826212, + "grad_norm": 0.8739959597587585, + "learning_rate": 6.215234058173465e-05, + "loss": 0.9738, + "step": 14020 + }, + { + "epoch": 2.496260683760684, + "grad_norm": 0.8388087749481201, + "learning_rate": 6.213938474027148e-05, + "loss": 1.0128, + "step": 14021 + }, + { + "epoch": 2.4964387464387463, + "grad_norm": 0.8963341116905212, + "learning_rate": 6.212642964060843e-05, + "loss": 0.9669, + "step": 14022 + }, + { + "epoch": 2.496616809116809, + "grad_norm": 0.8959031701087952, + "learning_rate": 6.211347528299928e-05, + "loss": 0.9558, + "step": 14023 + }, + { + "epoch": 2.496794871794872, + "grad_norm": 0.8463472127914429, + "learning_rate": 6.210052166769791e-05, + "loss": 0.9835, + "step": 14024 + }, + { + "epoch": 2.4969729344729346, + "grad_norm": 0.7827564477920532, + "learning_rate": 6.208756879495812e-05, + "loss": 0.8411, + "step": 14025 + }, + { + "epoch": 2.4971509971509973, + "grad_norm": 0.8851028084754944, + "learning_rate": 6.207461666503363e-05, + "loss": 1.0409, + "step": 14026 + }, + { + "epoch": 2.4973290598290596, + "grad_norm": 1.0151652097702026, + "learning_rate": 6.206166527817825e-05, + "loss": 0.6671, + "step": 14027 + }, + { + "epoch": 2.4975071225071224, + "grad_norm": 0.7924346923828125, + "learning_rate": 6.204871463464572e-05, + "loss": 0.9971, + "step": 14028 + }, + { + "epoch": 2.497685185185185, + "grad_norm": 0.8524144887924194, + "learning_rate": 6.203576473468981e-05, + "loss": 0.9228, + "step": 14029 + }, + { + "epoch": 2.497863247863248, + "grad_norm": 0.7936401963233948, + "learning_rate": 6.20228155785642e-05, + "loss": 0.8128, + "step": 14030 + }, + { + "epoch": 2.4980413105413106, + "grad_norm": 1.0074050426483154, + "learning_rate": 6.200986716652267e-05, + "loss": 0.7846, + "step": 14031 + }, + { + "epoch": 2.4982193732193734, + "grad_norm": 0.7972239851951599, + "learning_rate": 6.199691949881882e-05, + "loss": 0.8689, + "step": 14032 + }, + { + "epoch": 2.498397435897436, + "grad_norm": 0.8810364007949829, + "learning_rate": 6.198397257570643e-05, + "loss": 0.7775, + "step": 14033 + }, + { + "epoch": 2.4985754985754984, + "grad_norm": 0.8819566965103149, + "learning_rate": 6.19710263974391e-05, + "loss": 0.8852, + "step": 14034 + }, + { + "epoch": 2.498753561253561, + "grad_norm": 0.8020595908164978, + "learning_rate": 6.195808096427054e-05, + "loss": 0.9691, + "step": 14035 + }, + { + "epoch": 2.498931623931624, + "grad_norm": 0.83958899974823, + "learning_rate": 6.194513627645433e-05, + "loss": 0.8072, + "step": 14036 + }, + { + "epoch": 2.4991096866096867, + "grad_norm": 0.7525333166122437, + "learning_rate": 6.193219233424414e-05, + "loss": 0.7839, + "step": 14037 + }, + { + "epoch": 2.4992877492877494, + "grad_norm": 0.8687964677810669, + "learning_rate": 6.191924913789353e-05, + "loss": 0.9512, + "step": 14038 + }, + { + "epoch": 2.4994658119658117, + "grad_norm": 0.9080697298049927, + "learning_rate": 6.190630668765617e-05, + "loss": 0.8635, + "step": 14039 + }, + { + "epoch": 2.4996438746438745, + "grad_norm": 0.8174137473106384, + "learning_rate": 6.189336498378557e-05, + "loss": 0.9034, + "step": 14040 + }, + { + "epoch": 2.4996438746438745, + "eval_loss": 1.1338438987731934, + "eval_runtime": 24.4013, + "eval_samples_per_second": 42.662, + "eval_steps_per_second": 21.351, + "step": 14040 + }, + { + "epoch": 2.4998219373219372, + "grad_norm": 0.9711320996284485, + "learning_rate": 6.188042402653536e-05, + "loss": 0.9892, + "step": 14041 + }, + { + "epoch": 2.5, + "grad_norm": 0.8726856112480164, + "learning_rate": 6.1867483816159e-05, + "loss": 0.7482, + "step": 14042 + }, + { + "epoch": 2.5001780626780628, + "grad_norm": 0.875801682472229, + "learning_rate": 6.18545443529101e-05, + "loss": 0.7599, + "step": 14043 + }, + { + "epoch": 2.5003561253561255, + "grad_norm": 0.8867987990379333, + "learning_rate": 6.184160563704218e-05, + "loss": 0.805, + "step": 14044 + }, + { + "epoch": 2.5005341880341883, + "grad_norm": 0.8766322135925293, + "learning_rate": 6.18286676688087e-05, + "loss": 0.7343, + "step": 14045 + }, + { + "epoch": 2.5007122507122506, + "grad_norm": 0.8096646070480347, + "learning_rate": 6.181573044846323e-05, + "loss": 0.7957, + "step": 14046 + }, + { + "epoch": 2.5008903133903133, + "grad_norm": 1.0121821165084839, + "learning_rate": 6.180279397625917e-05, + "loss": 0.8775, + "step": 14047 + }, + { + "epoch": 2.501068376068376, + "grad_norm": 0.79291170835495, + "learning_rate": 6.178985825245003e-05, + "loss": 0.91, + "step": 14048 + }, + { + "epoch": 2.501246438746439, + "grad_norm": 0.83204185962677, + "learning_rate": 6.177692327728922e-05, + "loss": 0.799, + "step": 14049 + }, + { + "epoch": 2.5014245014245016, + "grad_norm": 0.8746328949928284, + "learning_rate": 6.176398905103023e-05, + "loss": 0.9595, + "step": 14050 + }, + { + "epoch": 2.501602564102564, + "grad_norm": 0.7665601968765259, + "learning_rate": 6.17510555739264e-05, + "loss": 0.7935, + "step": 14051 + }, + { + "epoch": 2.5017806267806266, + "grad_norm": 0.8761195540428162, + "learning_rate": 6.173812284623122e-05, + "loss": 1.017, + "step": 14052 + }, + { + "epoch": 2.5019586894586894, + "grad_norm": 0.8847656846046448, + "learning_rate": 6.172519086819802e-05, + "loss": 0.7684, + "step": 14053 + }, + { + "epoch": 2.502136752136752, + "grad_norm": 0.8320107460021973, + "learning_rate": 6.171225964008021e-05, + "loss": 0.8828, + "step": 14054 + }, + { + "epoch": 2.502314814814815, + "grad_norm": 1.0184354782104492, + "learning_rate": 6.169932916213111e-05, + "loss": 0.887, + "step": 14055 + }, + { + "epoch": 2.5024928774928776, + "grad_norm": 0.7870062589645386, + "learning_rate": 6.168639943460415e-05, + "loss": 0.8595, + "step": 14056 + }, + { + "epoch": 2.5026709401709404, + "grad_norm": 0.8314430117607117, + "learning_rate": 6.167347045775254e-05, + "loss": 0.7822, + "step": 14057 + }, + { + "epoch": 2.5028490028490027, + "grad_norm": 0.7521854043006897, + "learning_rate": 6.166054223182968e-05, + "loss": 0.916, + "step": 14058 + }, + { + "epoch": 2.5030270655270654, + "grad_norm": 0.8240202069282532, + "learning_rate": 6.164761475708885e-05, + "loss": 0.9404, + "step": 14059 + }, + { + "epoch": 2.503205128205128, + "grad_norm": 0.8467113375663757, + "learning_rate": 6.163468803378338e-05, + "loss": 0.7554, + "step": 14060 + }, + { + "epoch": 2.503383190883191, + "grad_norm": 0.7639012336730957, + "learning_rate": 6.162176206216645e-05, + "loss": 0.8141, + "step": 14061 + }, + { + "epoch": 2.5035612535612537, + "grad_norm": 0.8578195571899414, + "learning_rate": 6.160883684249138e-05, + "loss": 0.821, + "step": 14062 + }, + { + "epoch": 2.503739316239316, + "grad_norm": 0.7478210926055908, + "learning_rate": 6.159591237501139e-05, + "loss": 0.5878, + "step": 14063 + }, + { + "epoch": 2.5039173789173788, + "grad_norm": 0.7936450242996216, + "learning_rate": 6.158298865997972e-05, + "loss": 0.9616, + "step": 14064 + }, + { + "epoch": 2.5040954415954415, + "grad_norm": 0.9196288585662842, + "learning_rate": 6.157006569764963e-05, + "loss": 0.8147, + "step": 14065 + }, + { + "epoch": 2.5042735042735043, + "grad_norm": 1.0488382577896118, + "learning_rate": 6.155714348827422e-05, + "loss": 0.7941, + "step": 14066 + }, + { + "epoch": 2.504451566951567, + "grad_norm": 0.9195658564567566, + "learning_rate": 6.154422203210676e-05, + "loss": 1.0186, + "step": 14067 + }, + { + "epoch": 2.5046296296296298, + "grad_norm": 0.9088640213012695, + "learning_rate": 6.153130132940037e-05, + "loss": 0.7611, + "step": 14068 + }, + { + "epoch": 2.5048076923076925, + "grad_norm": 0.8168773651123047, + "learning_rate": 6.151838138040821e-05, + "loss": 0.8466, + "step": 14069 + }, + { + "epoch": 2.504985754985755, + "grad_norm": 0.9976982474327087, + "learning_rate": 6.150546218538342e-05, + "loss": 0.9438, + "step": 14070 + }, + { + "epoch": 2.5051638176638176, + "grad_norm": 0.9469537138938904, + "learning_rate": 6.149254374457917e-05, + "loss": 0.9485, + "step": 14071 + }, + { + "epoch": 2.5053418803418803, + "grad_norm": 0.8861194849014282, + "learning_rate": 6.147962605824851e-05, + "loss": 0.9676, + "step": 14072 + }, + { + "epoch": 2.505519943019943, + "grad_norm": 0.9008424878120422, + "learning_rate": 6.146670912664457e-05, + "loss": 0.7343, + "step": 14073 + }, + { + "epoch": 2.505698005698006, + "grad_norm": 0.8957796096801758, + "learning_rate": 6.145379295002038e-05, + "loss": 0.8538, + "step": 14074 + }, + { + "epoch": 2.505876068376068, + "grad_norm": 0.8739160895347595, + "learning_rate": 6.14408775286291e-05, + "loss": 0.9029, + "step": 14075 + }, + { + "epoch": 2.506054131054131, + "grad_norm": 0.7713274955749512, + "learning_rate": 6.142796286272368e-05, + "loss": 0.6962, + "step": 14076 + }, + { + "epoch": 2.5062321937321936, + "grad_norm": 0.8545170426368713, + "learning_rate": 6.141504895255725e-05, + "loss": 0.9208, + "step": 14077 + }, + { + "epoch": 2.5064102564102564, + "grad_norm": 0.8102772235870361, + "learning_rate": 6.140213579838274e-05, + "loss": 0.7785, + "step": 14078 + }, + { + "epoch": 2.506588319088319, + "grad_norm": 0.8055099844932556, + "learning_rate": 6.138922340045321e-05, + "loss": 0.8502, + "step": 14079 + }, + { + "epoch": 2.506766381766382, + "grad_norm": 0.8132893443107605, + "learning_rate": 6.137631175902164e-05, + "loss": 0.7559, + "step": 14080 + }, + { + "epoch": 2.5069444444444446, + "grad_norm": 0.8608863949775696, + "learning_rate": 6.136340087434102e-05, + "loss": 0.838, + "step": 14081 + }, + { + "epoch": 2.5071225071225074, + "grad_norm": 0.8480643630027771, + "learning_rate": 6.135049074666428e-05, + "loss": 0.9062, + "step": 14082 + }, + { + "epoch": 2.5073005698005697, + "grad_norm": 0.7107672691345215, + "learning_rate": 6.133758137624437e-05, + "loss": 0.7494, + "step": 14083 + }, + { + "epoch": 2.5074786324786325, + "grad_norm": 0.812416672706604, + "learning_rate": 6.132467276333427e-05, + "loss": 0.6428, + "step": 14084 + }, + { + "epoch": 2.507656695156695, + "grad_norm": 0.8304431438446045, + "learning_rate": 6.131176490818684e-05, + "loss": 0.9931, + "step": 14085 + }, + { + "epoch": 2.507834757834758, + "grad_norm": 0.8344886302947998, + "learning_rate": 6.129885781105507e-05, + "loss": 0.921, + "step": 14086 + }, + { + "epoch": 2.5080128205128203, + "grad_norm": 0.8137457966804504, + "learning_rate": 6.128595147219172e-05, + "loss": 0.9113, + "step": 14087 + }, + { + "epoch": 2.508190883190883, + "grad_norm": 0.7404686212539673, + "learning_rate": 6.127304589184976e-05, + "loss": 0.7625, + "step": 14088 + }, + { + "epoch": 2.5083689458689458, + "grad_norm": 0.8179733157157898, + "learning_rate": 6.126014107028202e-05, + "loss": 0.9049, + "step": 14089 + }, + { + "epoch": 2.5085470085470085, + "grad_norm": 0.7788520455360413, + "learning_rate": 6.124723700774133e-05, + "loss": 0.7391, + "step": 14090 + }, + { + "epoch": 2.5087250712250713, + "grad_norm": 0.8127198219299316, + "learning_rate": 6.123433370448052e-05, + "loss": 0.8551, + "step": 14091 + }, + { + "epoch": 2.508903133903134, + "grad_norm": 0.8134245276451111, + "learning_rate": 6.122143116075245e-05, + "loss": 0.7422, + "step": 14092 + }, + { + "epoch": 2.5090811965811968, + "grad_norm": 0.9117823243141174, + "learning_rate": 6.120852937680983e-05, + "loss": 0.8649, + "step": 14093 + }, + { + "epoch": 2.5092592592592595, + "grad_norm": 0.8417702913284302, + "learning_rate": 6.119562835290553e-05, + "loss": 0.6902, + "step": 14094 + }, + { + "epoch": 2.509437321937322, + "grad_norm": 0.8655431866645813, + "learning_rate": 6.118272808929225e-05, + "loss": 1.0778, + "step": 14095 + }, + { + "epoch": 2.5096153846153846, + "grad_norm": 0.9228867888450623, + "learning_rate": 6.116982858622282e-05, + "loss": 0.9991, + "step": 14096 + }, + { + "epoch": 2.5097934472934473, + "grad_norm": 0.819505512714386, + "learning_rate": 6.115692984394992e-05, + "loss": 1.0351, + "step": 14097 + }, + { + "epoch": 2.50997150997151, + "grad_norm": 0.8419737219810486, + "learning_rate": 6.114403186272628e-05, + "loss": 0.92, + "step": 14098 + }, + { + "epoch": 2.5101495726495724, + "grad_norm": 0.8294256329536438, + "learning_rate": 6.11311346428046e-05, + "loss": 0.792, + "step": 14099 + }, + { + "epoch": 2.510327635327635, + "grad_norm": 0.8549113273620605, + "learning_rate": 6.111823818443765e-05, + "loss": 0.9453, + "step": 14100 + }, + { + "epoch": 2.510505698005698, + "grad_norm": 1.0194092988967896, + "learning_rate": 6.1105342487878e-05, + "loss": 0.9747, + "step": 14101 + }, + { + "epoch": 2.5106837606837606, + "grad_norm": 0.7633654475212097, + "learning_rate": 6.109244755337842e-05, + "loss": 0.8227, + "step": 14102 + }, + { + "epoch": 2.5108618233618234, + "grad_norm": 0.9679104685783386, + "learning_rate": 6.107955338119147e-05, + "loss": 1.0407, + "step": 14103 + }, + { + "epoch": 2.511039886039886, + "grad_norm": 0.8342793583869934, + "learning_rate": 6.10666599715698e-05, + "loss": 0.7868, + "step": 14104 + }, + { + "epoch": 2.511217948717949, + "grad_norm": 0.9264410734176636, + "learning_rate": 6.105376732476609e-05, + "loss": 0.802, + "step": 14105 + }, + { + "epoch": 2.5113960113960117, + "grad_norm": 0.7511885762214661, + "learning_rate": 6.104087544103287e-05, + "loss": 0.7561, + "step": 14106 + }, + { + "epoch": 2.511574074074074, + "grad_norm": 0.8330591320991516, + "learning_rate": 6.102798432062282e-05, + "loss": 0.8511, + "step": 14107 + }, + { + "epoch": 2.5117521367521367, + "grad_norm": 0.8971241116523743, + "learning_rate": 6.1015093963788415e-05, + "loss": 0.8854, + "step": 14108 + }, + { + "epoch": 2.5119301994301995, + "grad_norm": 0.7926762700080872, + "learning_rate": 6.100220437078228e-05, + "loss": 0.6803, + "step": 14109 + }, + { + "epoch": 2.512108262108262, + "grad_norm": 0.7384431958198547, + "learning_rate": 6.098931554185692e-05, + "loss": 0.6162, + "step": 14110 + }, + { + "epoch": 2.5122863247863245, + "grad_norm": 0.9201281070709229, + "learning_rate": 6.097642747726491e-05, + "loss": 1.1095, + "step": 14111 + }, + { + "epoch": 2.5124643874643873, + "grad_norm": 0.8822020888328552, + "learning_rate": 6.0963540177258716e-05, + "loss": 0.87, + "step": 14112 + }, + { + "epoch": 2.51264245014245, + "grad_norm": 0.8243268728256226, + "learning_rate": 6.09506536420909e-05, + "loss": 1.0899, + "step": 14113 + }, + { + "epoch": 2.5128205128205128, + "grad_norm": 0.8657538294792175, + "learning_rate": 6.093776787201386e-05, + "loss": 0.8218, + "step": 14114 + }, + { + "epoch": 2.5129985754985755, + "grad_norm": 0.8651030659675598, + "learning_rate": 6.092488286728013e-05, + "loss": 0.7903, + "step": 14115 + }, + { + "epoch": 2.5131766381766383, + "grad_norm": 0.8341799378395081, + "learning_rate": 6.091199862814214e-05, + "loss": 0.8612, + "step": 14116 + }, + { + "epoch": 2.513354700854701, + "grad_norm": 0.7693229913711548, + "learning_rate": 6.0899115154852384e-05, + "loss": 0.819, + "step": 14117 + }, + { + "epoch": 2.513532763532764, + "grad_norm": 0.8883055448532104, + "learning_rate": 6.088623244766318e-05, + "loss": 0.9026, + "step": 14118 + }, + { + "epoch": 2.513710826210826, + "grad_norm": 0.7761621475219727, + "learning_rate": 6.087335050682703e-05, + "loss": 0.7505, + "step": 14119 + }, + { + "epoch": 2.513888888888889, + "grad_norm": 0.8152571320533752, + "learning_rate": 6.086046933259628e-05, + "loss": 0.7637, + "step": 14120 + }, + { + "epoch": 2.5140669515669516, + "grad_norm": 0.7990148663520813, + "learning_rate": 6.0847588925223376e-05, + "loss": 0.8615, + "step": 14121 + }, + { + "epoch": 2.5142450142450143, + "grad_norm": 0.844756007194519, + "learning_rate": 6.083470928496058e-05, + "loss": 0.8696, + "step": 14122 + }, + { + "epoch": 2.5144230769230766, + "grad_norm": 0.7533631324768066, + "learning_rate": 6.082183041206031e-05, + "loss": 0.7674, + "step": 14123 + }, + { + "epoch": 2.5146011396011394, + "grad_norm": 0.7914009690284729, + "learning_rate": 6.0808952306774905e-05, + "loss": 0.7577, + "step": 14124 + }, + { + "epoch": 2.514779202279202, + "grad_norm": 0.8341572284698486, + "learning_rate": 6.079607496935666e-05, + "loss": 0.8899, + "step": 14125 + }, + { + "epoch": 2.514957264957265, + "grad_norm": 0.9185548424720764, + "learning_rate": 6.078319840005788e-05, + "loss": 0.9486, + "step": 14126 + }, + { + "epoch": 2.5151353276353277, + "grad_norm": 0.8611742854118347, + "learning_rate": 6.0770322599130856e-05, + "loss": 0.8267, + "step": 14127 + }, + { + "epoch": 2.5153133903133904, + "grad_norm": 0.899135410785675, + "learning_rate": 6.0757447566827906e-05, + "loss": 1.0829, + "step": 14128 + }, + { + "epoch": 2.515491452991453, + "grad_norm": 0.8016429543495178, + "learning_rate": 6.074457330340122e-05, + "loss": 0.8582, + "step": 14129 + }, + { + "epoch": 2.515669515669516, + "grad_norm": 0.7781331539154053, + "learning_rate": 6.073169980910307e-05, + "loss": 0.8435, + "step": 14130 + }, + { + "epoch": 2.515847578347578, + "grad_norm": 0.7605105042457581, + "learning_rate": 6.071882708418568e-05, + "loss": 0.6961, + "step": 14131 + }, + { + "epoch": 2.516025641025641, + "grad_norm": 0.9337655901908875, + "learning_rate": 6.0705955128901326e-05, + "loss": 0.8673, + "step": 14132 + }, + { + "epoch": 2.5162037037037037, + "grad_norm": 0.7868272662162781, + "learning_rate": 6.06930839435021e-05, + "loss": 0.7526, + "step": 14133 + }, + { + "epoch": 2.5163817663817665, + "grad_norm": 0.8722387552261353, + "learning_rate": 6.068021352824027e-05, + "loss": 0.9541, + "step": 14134 + }, + { + "epoch": 2.5165598290598292, + "grad_norm": 0.7682648301124573, + "learning_rate": 6.066734388336794e-05, + "loss": 0.7191, + "step": 14135 + }, + { + "epoch": 2.5167378917378915, + "grad_norm": 0.9540650844573975, + "learning_rate": 6.065447500913737e-05, + "loss": 1.0638, + "step": 14136 + }, + { + "epoch": 2.5169159544159543, + "grad_norm": 0.8276218175888062, + "learning_rate": 6.064160690580056e-05, + "loss": 0.7967, + "step": 14137 + }, + { + "epoch": 2.517094017094017, + "grad_norm": 0.7966098785400391, + "learning_rate": 6.062873957360976e-05, + "loss": 0.8913, + "step": 14138 + }, + { + "epoch": 2.51727207977208, + "grad_norm": 0.9670028686523438, + "learning_rate": 6.0615873012816974e-05, + "loss": 0.8846, + "step": 14139 + }, + { + "epoch": 2.5174501424501425, + "grad_norm": 0.819952666759491, + "learning_rate": 6.0603007223674366e-05, + "loss": 0.8409, + "step": 14140 + }, + { + "epoch": 2.5176282051282053, + "grad_norm": 0.7746681571006775, + "learning_rate": 6.0590142206433973e-05, + "loss": 0.7382, + "step": 14141 + }, + { + "epoch": 2.517806267806268, + "grad_norm": 0.9452744722366333, + "learning_rate": 6.057727796134787e-05, + "loss": 0.8878, + "step": 14142 + }, + { + "epoch": 2.5179843304843303, + "grad_norm": 0.7940170168876648, + "learning_rate": 6.0564414488668165e-05, + "loss": 0.8289, + "step": 14143 + }, + { + "epoch": 2.518162393162393, + "grad_norm": 0.9046176671981812, + "learning_rate": 6.0551551788646774e-05, + "loss": 0.8596, + "step": 14144 + }, + { + "epoch": 2.518340455840456, + "grad_norm": 0.8460658192634583, + "learning_rate": 6.053868986153581e-05, + "loss": 0.7678, + "step": 14145 + }, + { + "epoch": 2.5185185185185186, + "grad_norm": 0.9131760597229004, + "learning_rate": 6.052582870758723e-05, + "loss": 0.8845, + "step": 14146 + }, + { + "epoch": 2.5186965811965814, + "grad_norm": 0.8375167846679688, + "learning_rate": 6.0512968327053076e-05, + "loss": 1.0082, + "step": 14147 + }, + { + "epoch": 2.5188746438746437, + "grad_norm": 0.8587140440940857, + "learning_rate": 6.050010872018523e-05, + "loss": 0.8745, + "step": 14148 + }, + { + "epoch": 2.5190527065527064, + "grad_norm": 0.8347265124320984, + "learning_rate": 6.048724988723575e-05, + "loss": 0.8345, + "step": 14149 + }, + { + "epoch": 2.519230769230769, + "grad_norm": 1.0271183252334595, + "learning_rate": 6.047439182845649e-05, + "loss": 0.9863, + "step": 14150 + }, + { + "epoch": 2.519408831908832, + "grad_norm": 0.6951111555099487, + "learning_rate": 6.046153454409943e-05, + "loss": 0.6162, + "step": 14151 + }, + { + "epoch": 2.5195868945868947, + "grad_norm": 0.7702959179878235, + "learning_rate": 6.044867803441645e-05, + "loss": 0.8127, + "step": 14152 + }, + { + "epoch": 2.5197649572649574, + "grad_norm": 0.7997276186943054, + "learning_rate": 6.0435822299659496e-05, + "loss": 0.6777, + "step": 14153 + }, + { + "epoch": 2.51994301994302, + "grad_norm": 0.8006166815757751, + "learning_rate": 6.0422967340080385e-05, + "loss": 0.9122, + "step": 14154 + }, + { + "epoch": 2.5201210826210825, + "grad_norm": 0.888225793838501, + "learning_rate": 6.041011315593102e-05, + "loss": 0.7621, + "step": 14155 + }, + { + "epoch": 2.5202991452991452, + "grad_norm": 0.928814172744751, + "learning_rate": 6.039725974746324e-05, + "loss": 1.0245, + "step": 14156 + }, + { + "epoch": 2.520477207977208, + "grad_norm": 0.7914403676986694, + "learning_rate": 6.038440711492892e-05, + "loss": 0.6585, + "step": 14157 + }, + { + "epoch": 2.5206552706552707, + "grad_norm": 0.82389897108078, + "learning_rate": 6.0371555258579826e-05, + "loss": 0.7862, + "step": 14158 + }, + { + "epoch": 2.5208333333333335, + "grad_norm": 0.952135443687439, + "learning_rate": 6.035870417866778e-05, + "loss": 0.8952, + "step": 14159 + }, + { + "epoch": 2.521011396011396, + "grad_norm": 0.8626661896705627, + "learning_rate": 6.034585387544458e-05, + "loss": 0.9166, + "step": 14160 + }, + { + "epoch": 2.5211894586894585, + "grad_norm": 0.9641584157943726, + "learning_rate": 6.033300434916203e-05, + "loss": 0.8481, + "step": 14161 + }, + { + "epoch": 2.5213675213675213, + "grad_norm": 0.949110209941864, + "learning_rate": 6.0320155600071814e-05, + "loss": 0.9628, + "step": 14162 + }, + { + "epoch": 2.521545584045584, + "grad_norm": 0.8198522329330444, + "learning_rate": 6.030730762842573e-05, + "loss": 0.817, + "step": 14163 + }, + { + "epoch": 2.521723646723647, + "grad_norm": 0.9209866523742676, + "learning_rate": 6.029446043447553e-05, + "loss": 0.925, + "step": 14164 + }, + { + "epoch": 2.5219017094017095, + "grad_norm": 0.8604369163513184, + "learning_rate": 6.0281614018472854e-05, + "loss": 0.7846, + "step": 14165 + }, + { + "epoch": 2.5220797720797723, + "grad_norm": 0.882255494594574, + "learning_rate": 6.026876838066948e-05, + "loss": 0.8715, + "step": 14166 + }, + { + "epoch": 2.5222578347578346, + "grad_norm": 0.8609021306037903, + "learning_rate": 6.0255923521317015e-05, + "loss": 0.8627, + "step": 14167 + }, + { + "epoch": 2.5224358974358974, + "grad_norm": 0.9782202243804932, + "learning_rate": 6.0243079440667226e-05, + "loss": 0.8499, + "step": 14168 + }, + { + "epoch": 2.52261396011396, + "grad_norm": 0.7932701706886292, + "learning_rate": 6.023023613897165e-05, + "loss": 0.9174, + "step": 14169 + }, + { + "epoch": 2.522792022792023, + "grad_norm": 0.8827422261238098, + "learning_rate": 6.021739361648202e-05, + "loss": 0.9384, + "step": 14170 + }, + { + "epoch": 2.5229700854700856, + "grad_norm": 0.9764171838760376, + "learning_rate": 6.020455187344989e-05, + "loss": 0.8806, + "step": 14171 + }, + { + "epoch": 2.523148148148148, + "grad_norm": 0.7635362148284912, + "learning_rate": 6.019171091012694e-05, + "loss": 0.7519, + "step": 14172 + }, + { + "epoch": 2.5233262108262107, + "grad_norm": 0.9925556182861328, + "learning_rate": 6.017887072676468e-05, + "loss": 0.8467, + "step": 14173 + }, + { + "epoch": 2.5235042735042734, + "grad_norm": 0.9624950289726257, + "learning_rate": 6.016603132361477e-05, + "loss": 0.9492, + "step": 14174 + }, + { + "epoch": 2.523682336182336, + "grad_norm": 0.7960891127586365, + "learning_rate": 6.0153192700928685e-05, + "loss": 1.0111, + "step": 14175 + }, + { + "epoch": 2.523860398860399, + "grad_norm": 0.8387307524681091, + "learning_rate": 6.014035485895804e-05, + "loss": 0.8013, + "step": 14176 + }, + { + "epoch": 2.5240384615384617, + "grad_norm": 0.8488287925720215, + "learning_rate": 6.0127517797954316e-05, + "loss": 0.8508, + "step": 14177 + }, + { + "epoch": 2.5242165242165244, + "grad_norm": 0.7339358329772949, + "learning_rate": 6.011468151816908e-05, + "loss": 0.7225, + "step": 14178 + }, + { + "epoch": 2.5243945868945867, + "grad_norm": 0.9265308976173401, + "learning_rate": 6.010184601985378e-05, + "loss": 0.7993, + "step": 14179 + }, + { + "epoch": 2.5245726495726495, + "grad_norm": 0.7752045392990112, + "learning_rate": 6.0089011303259944e-05, + "loss": 0.8315, + "step": 14180 + }, + { + "epoch": 2.5247507122507122, + "grad_norm": 0.7794929146766663, + "learning_rate": 6.007617736863901e-05, + "loss": 0.9174, + "step": 14181 + }, + { + "epoch": 2.524928774928775, + "grad_norm": 0.9099361896514893, + "learning_rate": 6.0063344216242434e-05, + "loss": 0.8948, + "step": 14182 + }, + { + "epoch": 2.5251068376068377, + "grad_norm": 0.8161521553993225, + "learning_rate": 6.005051184632171e-05, + "loss": 0.8018, + "step": 14183 + }, + { + "epoch": 2.5252849002849, + "grad_norm": 0.9279208183288574, + "learning_rate": 6.003768025912819e-05, + "loss": 0.9032, + "step": 14184 + }, + { + "epoch": 2.525462962962963, + "grad_norm": 0.9689664840698242, + "learning_rate": 6.002484945491333e-05, + "loss": 0.9463, + "step": 14185 + }, + { + "epoch": 2.5256410256410255, + "grad_norm": 0.8367486596107483, + "learning_rate": 6.001201943392848e-05, + "loss": 0.7866, + "step": 14186 + }, + { + "epoch": 2.5258190883190883, + "grad_norm": 0.8383589386940002, + "learning_rate": 5.9999190196425056e-05, + "loss": 0.7642, + "step": 14187 + }, + { + "epoch": 2.525997150997151, + "grad_norm": 0.9113569855690002, + "learning_rate": 5.99863617426544e-05, + "loss": 1.0451, + "step": 14188 + }, + { + "epoch": 2.526175213675214, + "grad_norm": 0.896575391292572, + "learning_rate": 5.997353407286788e-05, + "loss": 0.8559, + "step": 14189 + }, + { + "epoch": 2.5263532763532766, + "grad_norm": 0.899214506149292, + "learning_rate": 5.996070718731679e-05, + "loss": 1.0006, + "step": 14190 + }, + { + "epoch": 2.5265313390313393, + "grad_norm": 0.9739418625831604, + "learning_rate": 5.994788108625247e-05, + "loss": 0.9412, + "step": 14191 + }, + { + "epoch": 2.5267094017094016, + "grad_norm": 0.7940781712532043, + "learning_rate": 5.9935055769926215e-05, + "loss": 0.8802, + "step": 14192 + }, + { + "epoch": 2.5268874643874644, + "grad_norm": 0.802066445350647, + "learning_rate": 5.9922231238589346e-05, + "loss": 0.8459, + "step": 14193 + }, + { + "epoch": 2.527065527065527, + "grad_norm": 0.827560544013977, + "learning_rate": 5.990940749249306e-05, + "loss": 0.8732, + "step": 14194 + }, + { + "epoch": 2.52724358974359, + "grad_norm": 0.7782348394393921, + "learning_rate": 5.989658453188869e-05, + "loss": 0.6742, + "step": 14195 + }, + { + "epoch": 2.527421652421652, + "grad_norm": 0.8418310284614563, + "learning_rate": 5.9883762357027416e-05, + "loss": 0.8699, + "step": 14196 + }, + { + "epoch": 2.527599715099715, + "grad_norm": 0.7925812602043152, + "learning_rate": 5.987094096816051e-05, + "loss": 0.6807, + "step": 14197 + }, + { + "epoch": 2.5277777777777777, + "grad_norm": 0.8200794458389282, + "learning_rate": 5.9858120365539105e-05, + "loss": 0.8249, + "step": 14198 + }, + { + "epoch": 2.5279558404558404, + "grad_norm": 0.7137587070465088, + "learning_rate": 5.9845300549414505e-05, + "loss": 0.7882, + "step": 14199 + }, + { + "epoch": 2.528133903133903, + "grad_norm": 0.8084787726402283, + "learning_rate": 5.983248152003778e-05, + "loss": 1.0161, + "step": 14200 + }, + { + "epoch": 2.528311965811966, + "grad_norm": 0.7717064023017883, + "learning_rate": 5.9819663277660156e-05, + "loss": 0.798, + "step": 14201 + }, + { + "epoch": 2.5284900284900287, + "grad_norm": 0.7722328305244446, + "learning_rate": 5.980684582253275e-05, + "loss": 0.8324, + "step": 14202 + }, + { + "epoch": 2.5286680911680914, + "grad_norm": 0.8357635140419006, + "learning_rate": 5.9794029154906696e-05, + "loss": 0.9224, + "step": 14203 + }, + { + "epoch": 2.5288461538461537, + "grad_norm": 0.8159863352775574, + "learning_rate": 5.978121327503317e-05, + "loss": 0.7529, + "step": 14204 + }, + { + "epoch": 2.5290242165242165, + "grad_norm": 0.8255389332771301, + "learning_rate": 5.976839818316317e-05, + "loss": 0.9674, + "step": 14205 + }, + { + "epoch": 2.5292022792022792, + "grad_norm": 0.8204228281974792, + "learning_rate": 5.975558387954787e-05, + "loss": 0.9138, + "step": 14206 + }, + { + "epoch": 2.529380341880342, + "grad_norm": 0.8232463598251343, + "learning_rate": 5.9742770364438275e-05, + "loss": 0.7949, + "step": 14207 + }, + { + "epoch": 2.5295584045584043, + "grad_norm": 0.8164107203483582, + "learning_rate": 5.972995763808551e-05, + "loss": 0.7087, + "step": 14208 + }, + { + "epoch": 2.529736467236467, + "grad_norm": 0.8100822567939758, + "learning_rate": 5.971714570074052e-05, + "loss": 0.9187, + "step": 14209 + }, + { + "epoch": 2.52991452991453, + "grad_norm": 0.7401103377342224, + "learning_rate": 5.970433455265443e-05, + "loss": 0.798, + "step": 14210 + }, + { + "epoch": 2.5300925925925926, + "grad_norm": 0.798327624797821, + "learning_rate": 5.9691524194078154e-05, + "loss": 0.7312, + "step": 14211 + }, + { + "epoch": 2.5302706552706553, + "grad_norm": 0.8566045165061951, + "learning_rate": 5.9678714625262754e-05, + "loss": 0.8555, + "step": 14212 + }, + { + "epoch": 2.530448717948718, + "grad_norm": 0.8005902767181396, + "learning_rate": 5.9665905846459155e-05, + "loss": 0.7979, + "step": 14213 + }, + { + "epoch": 2.530626780626781, + "grad_norm": 0.815990686416626, + "learning_rate": 5.9653097857918396e-05, + "loss": 0.8739, + "step": 14214 + }, + { + "epoch": 2.5308048433048436, + "grad_norm": 0.7694230079650879, + "learning_rate": 5.9640290659891316e-05, + "loss": 0.7249, + "step": 14215 + }, + { + "epoch": 2.530982905982906, + "grad_norm": 0.8469253182411194, + "learning_rate": 5.962748425262892e-05, + "loss": 0.8505, + "step": 14216 + }, + { + "epoch": 2.5311609686609686, + "grad_norm": 0.8061797022819519, + "learning_rate": 5.961467863638209e-05, + "loss": 0.8979, + "step": 14217 + }, + { + "epoch": 2.5313390313390314, + "grad_norm": 1.0380569696426392, + "learning_rate": 5.960187381140179e-05, + "loss": 0.8664, + "step": 14218 + }, + { + "epoch": 2.531517094017094, + "grad_norm": 0.9435166716575623, + "learning_rate": 5.9589069777938786e-05, + "loss": 0.7566, + "step": 14219 + }, + { + "epoch": 2.5316951566951564, + "grad_norm": 0.8882613182067871, + "learning_rate": 5.957626653624407e-05, + "loss": 0.6999, + "step": 14220 + }, + { + "epoch": 2.531873219373219, + "grad_norm": 0.8544003963470459, + "learning_rate": 5.95634640865684e-05, + "loss": 0.8028, + "step": 14221 + }, + { + "epoch": 2.532051282051282, + "grad_norm": 0.8407679796218872, + "learning_rate": 5.9550662429162655e-05, + "loss": 0.6868, + "step": 14222 + }, + { + "epoch": 2.5322293447293447, + "grad_norm": 0.9049725532531738, + "learning_rate": 5.9537861564277654e-05, + "loss": 0.8177, + "step": 14223 + }, + { + "epoch": 2.5324074074074074, + "grad_norm": 0.938050389289856, + "learning_rate": 5.952506149216419e-05, + "loss": 0.9187, + "step": 14224 + }, + { + "epoch": 2.53258547008547, + "grad_norm": 0.9515482783317566, + "learning_rate": 5.951226221307312e-05, + "loss": 0.6735, + "step": 14225 + }, + { + "epoch": 2.532763532763533, + "grad_norm": 0.8545815348625183, + "learning_rate": 5.949946372725512e-05, + "loss": 0.7643, + "step": 14226 + }, + { + "epoch": 2.5329415954415957, + "grad_norm": 0.8388620615005493, + "learning_rate": 5.9486666034961e-05, + "loss": 0.9437, + "step": 14227 + }, + { + "epoch": 2.533119658119658, + "grad_norm": 0.7202512621879578, + "learning_rate": 5.9473869136441506e-05, + "loss": 0.718, + "step": 14228 + }, + { + "epoch": 2.5332977207977208, + "grad_norm": 0.8375558853149414, + "learning_rate": 5.946107303194739e-05, + "loss": 0.852, + "step": 14229 + }, + { + "epoch": 2.5334757834757835, + "grad_norm": 0.8980572819709778, + "learning_rate": 5.94482777217293e-05, + "loss": 0.9684, + "step": 14230 + }, + { + "epoch": 2.5336538461538463, + "grad_norm": 0.7374732494354248, + "learning_rate": 5.9435483206037977e-05, + "loss": 0.6498, + "step": 14231 + }, + { + "epoch": 2.5338319088319086, + "grad_norm": 1.073758602142334, + "learning_rate": 5.942268948512409e-05, + "loss": 1.0315, + "step": 14232 + }, + { + "epoch": 2.5340099715099713, + "grad_norm": 0.9503611326217651, + "learning_rate": 5.940989655923832e-05, + "loss": 1.0644, + "step": 14233 + }, + { + "epoch": 2.534188034188034, + "grad_norm": 0.870490550994873, + "learning_rate": 5.939710442863129e-05, + "loss": 1.1, + "step": 14234 + }, + { + "epoch": 2.534366096866097, + "grad_norm": 0.8019965887069702, + "learning_rate": 5.93843130935537e-05, + "loss": 0.9169, + "step": 14235 + }, + { + "epoch": 2.5345441595441596, + "grad_norm": 0.8333065509796143, + "learning_rate": 5.9371522554256076e-05, + "loss": 0.7848, + "step": 14236 + }, + { + "epoch": 2.5347222222222223, + "grad_norm": 0.8606435656547546, + "learning_rate": 5.935873281098909e-05, + "loss": 0.9941, + "step": 14237 + }, + { + "epoch": 2.534900284900285, + "grad_norm": 0.7711295485496521, + "learning_rate": 5.934594386400328e-05, + "loss": 0.8495, + "step": 14238 + }, + { + "epoch": 2.535078347578348, + "grad_norm": 0.871533215045929, + "learning_rate": 5.93331557135493e-05, + "loss": 0.9071, + "step": 14239 + }, + { + "epoch": 2.53525641025641, + "grad_norm": 0.9828163981437683, + "learning_rate": 5.932036835987762e-05, + "loss": 0.9561, + "step": 14240 + }, + { + "epoch": 2.535434472934473, + "grad_norm": 0.8485092520713806, + "learning_rate": 5.930758180323881e-05, + "loss": 0.7278, + "step": 14241 + }, + { + "epoch": 2.5356125356125356, + "grad_norm": 0.7608986496925354, + "learning_rate": 5.929479604388342e-05, + "loss": 1.0449, + "step": 14242 + }, + { + "epoch": 2.5357905982905984, + "grad_norm": 0.7852896451950073, + "learning_rate": 5.928201108206193e-05, + "loss": 0.8844, + "step": 14243 + }, + { + "epoch": 2.5359686609686607, + "grad_norm": 0.7636764645576477, + "learning_rate": 5.9269226918024875e-05, + "loss": 0.8259, + "step": 14244 + }, + { + "epoch": 2.5361467236467234, + "grad_norm": 0.9067455530166626, + "learning_rate": 5.925644355202269e-05, + "loss": 0.8742, + "step": 14245 + }, + { + "epoch": 2.536324786324786, + "grad_norm": 0.7911350727081299, + "learning_rate": 5.924366098430588e-05, + "loss": 0.8586, + "step": 14246 + }, + { + "epoch": 2.536502849002849, + "grad_norm": 0.8010593056678772, + "learning_rate": 5.923087921512483e-05, + "loss": 0.8524, + "step": 14247 + }, + { + "epoch": 2.5366809116809117, + "grad_norm": 1.232219934463501, + "learning_rate": 5.9218098244730034e-05, + "loss": 0.8302, + "step": 14248 + }, + { + "epoch": 2.5368589743589745, + "grad_norm": 0.8717244267463684, + "learning_rate": 5.9205318073371874e-05, + "loss": 0.8692, + "step": 14249 + }, + { + "epoch": 2.537037037037037, + "grad_norm": 0.9757453799247742, + "learning_rate": 5.919253870130079e-05, + "loss": 0.7986, + "step": 14250 + }, + { + "epoch": 2.5372150997151, + "grad_norm": 0.8183274865150452, + "learning_rate": 5.917976012876712e-05, + "loss": 0.8277, + "step": 14251 + }, + { + "epoch": 2.5373931623931623, + "grad_norm": 0.823930025100708, + "learning_rate": 5.916698235602125e-05, + "loss": 0.7972, + "step": 14252 + }, + { + "epoch": 2.537571225071225, + "grad_norm": 0.8480231761932373, + "learning_rate": 5.915420538331353e-05, + "loss": 0.8234, + "step": 14253 + }, + { + "epoch": 2.5377492877492878, + "grad_norm": 0.6718716621398926, + "learning_rate": 5.914142921089434e-05, + "loss": 0.5984, + "step": 14254 + }, + { + "epoch": 2.5379273504273505, + "grad_norm": 0.8506333827972412, + "learning_rate": 5.912865383901394e-05, + "loss": 0.8004, + "step": 14255 + }, + { + "epoch": 2.5381054131054133, + "grad_norm": 0.755740225315094, + "learning_rate": 5.911587926792269e-05, + "loss": 0.7465, + "step": 14256 + }, + { + "epoch": 2.5382834757834756, + "grad_norm": 0.8908467888832092, + "learning_rate": 5.9103105497870815e-05, + "loss": 0.7822, + "step": 14257 + }, + { + "epoch": 2.5384615384615383, + "grad_norm": 0.9572851061820984, + "learning_rate": 5.909033252910867e-05, + "loss": 0.7891, + "step": 14258 + }, + { + "epoch": 2.538639601139601, + "grad_norm": 0.8606489896774292, + "learning_rate": 5.907756036188644e-05, + "loss": 0.9071, + "step": 14259 + }, + { + "epoch": 2.538817663817664, + "grad_norm": 0.837658166885376, + "learning_rate": 5.906478899645444e-05, + "loss": 0.829, + "step": 14260 + }, + { + "epoch": 2.5389957264957266, + "grad_norm": 0.8964337706565857, + "learning_rate": 5.905201843306285e-05, + "loss": 1.0385, + "step": 14261 + }, + { + "epoch": 2.5391737891737893, + "grad_norm": 0.7854750156402588, + "learning_rate": 5.903924867196189e-05, + "loss": 0.8905, + "step": 14262 + }, + { + "epoch": 2.539351851851852, + "grad_norm": 0.8828065991401672, + "learning_rate": 5.902647971340176e-05, + "loss": 0.8541, + "step": 14263 + }, + { + "epoch": 2.5395299145299144, + "grad_norm": 0.804121196269989, + "learning_rate": 5.9013711557632645e-05, + "loss": 0.8333, + "step": 14264 + }, + { + "epoch": 2.539707977207977, + "grad_norm": 0.8868918418884277, + "learning_rate": 5.900094420490475e-05, + "loss": 0.7959, + "step": 14265 + }, + { + "epoch": 2.53988603988604, + "grad_norm": 0.9231327176094055, + "learning_rate": 5.8988177655468134e-05, + "loss": 0.8007, + "step": 14266 + }, + { + "epoch": 2.5400641025641026, + "grad_norm": 0.9918177127838135, + "learning_rate": 5.897541190957301e-05, + "loss": 0.7766, + "step": 14267 + }, + { + "epoch": 2.5402421652421654, + "grad_norm": 0.8468625545501709, + "learning_rate": 5.896264696746947e-05, + "loss": 0.9209, + "step": 14268 + }, + { + "epoch": 2.5404202279202277, + "grad_norm": 0.9160833358764648, + "learning_rate": 5.894988282940761e-05, + "loss": 0.8994, + "step": 14269 + }, + { + "epoch": 2.5405982905982905, + "grad_norm": 0.8029152154922485, + "learning_rate": 5.8937119495637515e-05, + "loss": 0.7936, + "step": 14270 + }, + { + "epoch": 2.540776353276353, + "grad_norm": 0.8829928636550903, + "learning_rate": 5.8924356966409286e-05, + "loss": 0.7368, + "step": 14271 + }, + { + "epoch": 2.540954415954416, + "grad_norm": 0.9698056578636169, + "learning_rate": 5.8911595241972925e-05, + "loss": 0.789, + "step": 14272 + }, + { + "epoch": 2.5411324786324787, + "grad_norm": 0.7949244379997253, + "learning_rate": 5.8898834322578524e-05, + "loss": 0.8885, + "step": 14273 + }, + { + "epoch": 2.5413105413105415, + "grad_norm": 1.2430917024612427, + "learning_rate": 5.888607420847605e-05, + "loss": 0.861, + "step": 14274 + }, + { + "epoch": 2.541488603988604, + "grad_norm": 0.7476705312728882, + "learning_rate": 5.887331489991559e-05, + "loss": 0.7942, + "step": 14275 + }, + { + "epoch": 2.5416666666666665, + "grad_norm": 0.9204338192939758, + "learning_rate": 5.886055639714706e-05, + "loss": 0.8633, + "step": 14276 + }, + { + "epoch": 2.5418447293447293, + "grad_norm": 0.8812162280082703, + "learning_rate": 5.884779870042047e-05, + "loss": 0.7162, + "step": 14277 + }, + { + "epoch": 2.542022792022792, + "grad_norm": 0.7859770655632019, + "learning_rate": 5.883504180998578e-05, + "loss": 0.7965, + "step": 14278 + }, + { + "epoch": 2.5422008547008548, + "grad_norm": 0.7732986211776733, + "learning_rate": 5.882228572609296e-05, + "loss": 0.9671, + "step": 14279 + }, + { + "epoch": 2.5423789173789175, + "grad_norm": 0.8555598855018616, + "learning_rate": 5.880953044899189e-05, + "loss": 0.8993, + "step": 14280 + }, + { + "epoch": 2.54255698005698, + "grad_norm": 0.7980908155441284, + "learning_rate": 5.879677597893248e-05, + "loss": 0.873, + "step": 14281 + }, + { + "epoch": 2.5427350427350426, + "grad_norm": 0.9244991540908813, + "learning_rate": 5.878402231616471e-05, + "loss": 0.934, + "step": 14282 + }, + { + "epoch": 2.5429131054131053, + "grad_norm": 1.0128331184387207, + "learning_rate": 5.877126946093835e-05, + "loss": 0.9607, + "step": 14283 + }, + { + "epoch": 2.543091168091168, + "grad_norm": 0.7916569709777832, + "learning_rate": 5.875851741350334e-05, + "loss": 0.831, + "step": 14284 + }, + { + "epoch": 2.543269230769231, + "grad_norm": 0.705007791519165, + "learning_rate": 5.8745766174109495e-05, + "loss": 0.6399, + "step": 14285 + }, + { + "epoch": 2.5434472934472936, + "grad_norm": 0.8785403966903687, + "learning_rate": 5.873301574300671e-05, + "loss": 0.9336, + "step": 14286 + }, + { + "epoch": 2.5436253561253563, + "grad_norm": 0.8225776553153992, + "learning_rate": 5.872026612044471e-05, + "loss": 0.8252, + "step": 14287 + }, + { + "epoch": 2.5438034188034186, + "grad_norm": 0.9629518985748291, + "learning_rate": 5.870751730667337e-05, + "loss": 1.0213, + "step": 14288 + }, + { + "epoch": 2.5439814814814814, + "grad_norm": 0.8242672681808472, + "learning_rate": 5.869476930194242e-05, + "loss": 0.9642, + "step": 14289 + }, + { + "epoch": 2.544159544159544, + "grad_norm": 0.5798216462135315, + "learning_rate": 5.868202210650171e-05, + "loss": 0.4366, + "step": 14290 + }, + { + "epoch": 2.544337606837607, + "grad_norm": 0.7945725917816162, + "learning_rate": 5.86692757206009e-05, + "loss": 0.9252, + "step": 14291 + }, + { + "epoch": 2.5445156695156697, + "grad_norm": 0.9078665375709534, + "learning_rate": 5.865653014448982e-05, + "loss": 1.0551, + "step": 14292 + }, + { + "epoch": 2.544693732193732, + "grad_norm": 0.8044732809066772, + "learning_rate": 5.86437853784181e-05, + "loss": 0.7778, + "step": 14293 + }, + { + "epoch": 2.5448717948717947, + "grad_norm": 0.8317133784294128, + "learning_rate": 5.863104142263553e-05, + "loss": 1.0047, + "step": 14294 + }, + { + "epoch": 2.5450498575498575, + "grad_norm": 0.8330327272415161, + "learning_rate": 5.861829827739174e-05, + "loss": 0.8074, + "step": 14295 + }, + { + "epoch": 2.54522792022792, + "grad_norm": 0.8731801509857178, + "learning_rate": 5.8605555942936474e-05, + "loss": 0.9311, + "step": 14296 + }, + { + "epoch": 2.545405982905983, + "grad_norm": 0.8906812071800232, + "learning_rate": 5.85928144195193e-05, + "loss": 0.9084, + "step": 14297 + }, + { + "epoch": 2.5455840455840457, + "grad_norm": 0.948535144329071, + "learning_rate": 5.8580073707389935e-05, + "loss": 0.923, + "step": 14298 + }, + { + "epoch": 2.5457621082621085, + "grad_norm": 1.0418797731399536, + "learning_rate": 5.8567333806797975e-05, + "loss": 0.9786, + "step": 14299 + }, + { + "epoch": 2.5459401709401708, + "grad_norm": 0.8591430187225342, + "learning_rate": 5.8554594717993075e-05, + "loss": 0.8706, + "step": 14300 + }, + { + "epoch": 2.5461182336182335, + "grad_norm": 1.1056550741195679, + "learning_rate": 5.854185644122475e-05, + "loss": 0.891, + "step": 14301 + }, + { + "epoch": 2.5462962962962963, + "grad_norm": 0.8945133090019226, + "learning_rate": 5.8529118976742624e-05, + "loss": 0.9584, + "step": 14302 + }, + { + "epoch": 2.546474358974359, + "grad_norm": 0.8568279147148132, + "learning_rate": 5.851638232479629e-05, + "loss": 0.7462, + "step": 14303 + }, + { + "epoch": 2.546652421652422, + "grad_norm": 0.934648871421814, + "learning_rate": 5.850364648563527e-05, + "loss": 0.8977, + "step": 14304 + }, + { + "epoch": 2.546830484330484, + "grad_norm": 0.8074216842651367, + "learning_rate": 5.849091145950909e-05, + "loss": 0.8779, + "step": 14305 + }, + { + "epoch": 2.547008547008547, + "grad_norm": 0.8781399726867676, + "learning_rate": 5.8478177246667266e-05, + "loss": 0.8715, + "step": 14306 + }, + { + "epoch": 2.5471866096866096, + "grad_norm": 0.8237441182136536, + "learning_rate": 5.846544384735933e-05, + "loss": 0.8806, + "step": 14307 + }, + { + "epoch": 2.5473646723646723, + "grad_norm": 0.933709442615509, + "learning_rate": 5.8452711261834717e-05, + "loss": 0.8555, + "step": 14308 + }, + { + "epoch": 2.547542735042735, + "grad_norm": 0.9045436978340149, + "learning_rate": 5.843997949034292e-05, + "loss": 1.0105, + "step": 14309 + }, + { + "epoch": 2.547720797720798, + "grad_norm": 0.8088112473487854, + "learning_rate": 5.842724853313337e-05, + "loss": 0.8667, + "step": 14310 + }, + { + "epoch": 2.5478988603988606, + "grad_norm": 1.1022162437438965, + "learning_rate": 5.841451839045559e-05, + "loss": 0.9818, + "step": 14311 + }, + { + "epoch": 2.5480769230769234, + "grad_norm": 0.8974189162254333, + "learning_rate": 5.8401789062558876e-05, + "loss": 0.843, + "step": 14312 + }, + { + "epoch": 2.5482549857549857, + "grad_norm": 0.7816309928894043, + "learning_rate": 5.838906054969272e-05, + "loss": 0.8665, + "step": 14313 + }, + { + "epoch": 2.5484330484330484, + "grad_norm": 0.8243623971939087, + "learning_rate": 5.8376332852106485e-05, + "loss": 0.7291, + "step": 14314 + }, + { + "epoch": 2.548611111111111, + "grad_norm": 0.8475931286811829, + "learning_rate": 5.8363605970049526e-05, + "loss": 0.7551, + "step": 14315 + }, + { + "epoch": 2.548789173789174, + "grad_norm": 0.8949251770973206, + "learning_rate": 5.835087990377124e-05, + "loss": 0.9138, + "step": 14316 + }, + { + "epoch": 2.548967236467236, + "grad_norm": 1.0444703102111816, + "learning_rate": 5.833815465352093e-05, + "loss": 0.8663, + "step": 14317 + }, + { + "epoch": 2.549145299145299, + "grad_norm": 0.8611619472503662, + "learning_rate": 5.8325430219547895e-05, + "loss": 0.795, + "step": 14318 + }, + { + "epoch": 2.5493233618233617, + "grad_norm": 0.7808047533035278, + "learning_rate": 5.8312706602101564e-05, + "loss": 0.905, + "step": 14319 + }, + { + "epoch": 2.5495014245014245, + "grad_norm": 0.9137473106384277, + "learning_rate": 5.8299983801431066e-05, + "loss": 0.8763, + "step": 14320 + }, + { + "epoch": 2.5496794871794872, + "grad_norm": 0.9511715769767761, + "learning_rate": 5.828726181778581e-05, + "loss": 0.8385, + "step": 14321 + }, + { + "epoch": 2.54985754985755, + "grad_norm": 0.9250940084457397, + "learning_rate": 5.8274540651415e-05, + "loss": 0.7026, + "step": 14322 + }, + { + "epoch": 2.5500356125356127, + "grad_norm": 1.001017689704895, + "learning_rate": 5.826182030256786e-05, + "loss": 0.7952, + "step": 14323 + }, + { + "epoch": 2.5502136752136755, + "grad_norm": 0.7638011574745178, + "learning_rate": 5.824910077149371e-05, + "loss": 0.7178, + "step": 14324 + }, + { + "epoch": 2.550391737891738, + "grad_norm": 0.9289371967315674, + "learning_rate": 5.823638205844164e-05, + "loss": 0.8492, + "step": 14325 + }, + { + "epoch": 2.5505698005698005, + "grad_norm": 0.8494341969490051, + "learning_rate": 5.822366416366093e-05, + "loss": 0.9095, + "step": 14326 + }, + { + "epoch": 2.5507478632478633, + "grad_norm": 0.8686699867248535, + "learning_rate": 5.8210947087400746e-05, + "loss": 0.9548, + "step": 14327 + }, + { + "epoch": 2.550925925925926, + "grad_norm": 1.1318142414093018, + "learning_rate": 5.819823082991025e-05, + "loss": 1.0554, + "step": 14328 + }, + { + "epoch": 2.5511039886039883, + "grad_norm": 0.8405448198318481, + "learning_rate": 5.818551539143857e-05, + "loss": 0.704, + "step": 14329 + }, + { + "epoch": 2.551282051282051, + "grad_norm": 0.9133256673812866, + "learning_rate": 5.8172800772234856e-05, + "loss": 0.922, + "step": 14330 + }, + { + "epoch": 2.551460113960114, + "grad_norm": 0.8526531457901001, + "learning_rate": 5.816008697254824e-05, + "loss": 0.6452, + "step": 14331 + }, + { + "epoch": 2.5516381766381766, + "grad_norm": 0.7987905144691467, + "learning_rate": 5.81473739926278e-05, + "loss": 0.8159, + "step": 14332 + }, + { + "epoch": 2.5518162393162394, + "grad_norm": 0.8217538595199585, + "learning_rate": 5.813466183272257e-05, + "loss": 0.6703, + "step": 14333 + }, + { + "epoch": 2.551994301994302, + "grad_norm": 0.7654905915260315, + "learning_rate": 5.8121950493081765e-05, + "loss": 0.8711, + "step": 14334 + }, + { + "epoch": 2.552172364672365, + "grad_norm": 0.872327983379364, + "learning_rate": 5.8109239973954264e-05, + "loss": 1.0079, + "step": 14335 + }, + { + "epoch": 2.5523504273504276, + "grad_norm": 0.7675468325614929, + "learning_rate": 5.809653027558922e-05, + "loss": 0.7541, + "step": 14336 + }, + { + "epoch": 2.55252849002849, + "grad_norm": 0.8367551565170288, + "learning_rate": 5.808382139823563e-05, + "loss": 0.9325, + "step": 14337 + }, + { + "epoch": 2.5527065527065527, + "grad_norm": 0.7946585416793823, + "learning_rate": 5.807111334214248e-05, + "loss": 0.6703, + "step": 14338 + }, + { + "epoch": 2.5528846153846154, + "grad_norm": 0.8752394318580627, + "learning_rate": 5.805840610755876e-05, + "loss": 0.82, + "step": 14339 + }, + { + "epoch": 2.553062678062678, + "grad_norm": 0.9394813776016235, + "learning_rate": 5.804569969473341e-05, + "loss": 1.0094, + "step": 14340 + }, + { + "epoch": 2.5532407407407405, + "grad_norm": 0.7763680219650269, + "learning_rate": 5.803299410391551e-05, + "loss": 0.8424, + "step": 14341 + }, + { + "epoch": 2.5534188034188032, + "grad_norm": 0.9148688912391663, + "learning_rate": 5.8020289335353816e-05, + "loss": 0.9344, + "step": 14342 + }, + { + "epoch": 2.553596866096866, + "grad_norm": 0.8666844367980957, + "learning_rate": 5.80075853892974e-05, + "loss": 0.8651, + "step": 14343 + }, + { + "epoch": 2.5537749287749287, + "grad_norm": 0.7240473628044128, + "learning_rate": 5.799488226599511e-05, + "loss": 0.6913, + "step": 14344 + }, + { + "epoch": 2.5539529914529915, + "grad_norm": 0.8949013352394104, + "learning_rate": 5.798217996569585e-05, + "loss": 0.7419, + "step": 14345 + }, + { + "epoch": 2.5541310541310542, + "grad_norm": 0.7760846614837646, + "learning_rate": 5.796947848864849e-05, + "loss": 0.8292, + "step": 14346 + }, + { + "epoch": 2.554309116809117, + "grad_norm": 0.8448507785797119, + "learning_rate": 5.795677783510187e-05, + "loss": 0.9953, + "step": 14347 + }, + { + "epoch": 2.5544871794871797, + "grad_norm": 0.834007203578949, + "learning_rate": 5.794407800530484e-05, + "loss": 0.8135, + "step": 14348 + }, + { + "epoch": 2.554665242165242, + "grad_norm": 0.8247915506362915, + "learning_rate": 5.793137899950629e-05, + "loss": 0.8607, + "step": 14349 + }, + { + "epoch": 2.554843304843305, + "grad_norm": 0.8796288967132568, + "learning_rate": 5.7918680817954906e-05, + "loss": 1.0479, + "step": 14350 + }, + { + "epoch": 2.5550213675213675, + "grad_norm": 0.8384763598442078, + "learning_rate": 5.790598346089964e-05, + "loss": 0.98, + "step": 14351 + }, + { + "epoch": 2.5551994301994303, + "grad_norm": 0.9394076466560364, + "learning_rate": 5.7893286928589107e-05, + "loss": 0.922, + "step": 14352 + }, + { + "epoch": 2.5553774928774926, + "grad_norm": 0.9548128843307495, + "learning_rate": 5.7880591221272184e-05, + "loss": 0.9877, + "step": 14353 + }, + { + "epoch": 2.5555555555555554, + "grad_norm": 0.7609717845916748, + "learning_rate": 5.786789633919758e-05, + "loss": 0.8115, + "step": 14354 + }, + { + "epoch": 2.555733618233618, + "grad_norm": 0.7415568232536316, + "learning_rate": 5.785520228261403e-05, + "loss": 0.6336, + "step": 14355 + }, + { + "epoch": 2.555911680911681, + "grad_norm": 0.8595952391624451, + "learning_rate": 5.7842509051770246e-05, + "loss": 0.8065, + "step": 14356 + }, + { + "epoch": 2.5560897435897436, + "grad_norm": 1.0075218677520752, + "learning_rate": 5.782981664691491e-05, + "loss": 0.9967, + "step": 14357 + }, + { + "epoch": 2.5562678062678064, + "grad_norm": 0.8405288457870483, + "learning_rate": 5.781712506829669e-05, + "loss": 0.7953, + "step": 14358 + }, + { + "epoch": 2.556445868945869, + "grad_norm": 0.8259321451187134, + "learning_rate": 5.780443431616435e-05, + "loss": 0.925, + "step": 14359 + }, + { + "epoch": 2.556623931623932, + "grad_norm": 0.8155162334442139, + "learning_rate": 5.7791744390766376e-05, + "loss": 0.9658, + "step": 14360 + }, + { + "epoch": 2.556801994301994, + "grad_norm": 0.8670404553413391, + "learning_rate": 5.7779055292351545e-05, + "loss": 1.0029, + "step": 14361 + }, + { + "epoch": 2.556980056980057, + "grad_norm": 0.8574714660644531, + "learning_rate": 5.7766367021168423e-05, + "loss": 0.9208, + "step": 14362 + }, + { + "epoch": 2.5571581196581197, + "grad_norm": 1.0231248140335083, + "learning_rate": 5.775367957746556e-05, + "loss": 1.0422, + "step": 14363 + }, + { + "epoch": 2.5573361823361824, + "grad_norm": 0.8403676152229309, + "learning_rate": 5.7740992961491655e-05, + "loss": 0.8068, + "step": 14364 + }, + { + "epoch": 2.557514245014245, + "grad_norm": 0.8792767524719238, + "learning_rate": 5.7728307173495136e-05, + "loss": 1.0405, + "step": 14365 + }, + { + "epoch": 2.5576923076923075, + "grad_norm": 0.8546510934829712, + "learning_rate": 5.771562221372471e-05, + "loss": 0.8246, + "step": 14366 + }, + { + "epoch": 2.5578703703703702, + "grad_norm": 0.7620588541030884, + "learning_rate": 5.770293808242875e-05, + "loss": 0.7588, + "step": 14367 + }, + { + "epoch": 2.558048433048433, + "grad_norm": 0.8154500722885132, + "learning_rate": 5.769025477985588e-05, + "loss": 0.8217, + "step": 14368 + }, + { + "epoch": 2.5582264957264957, + "grad_norm": 0.8630158305168152, + "learning_rate": 5.767757230625459e-05, + "loss": 0.8486, + "step": 14369 + }, + { + "epoch": 2.5584045584045585, + "grad_norm": 0.8991047143936157, + "learning_rate": 5.766489066187335e-05, + "loss": 0.9012, + "step": 14370 + }, + { + "epoch": 2.5585826210826212, + "grad_norm": 1.056725263595581, + "learning_rate": 5.7652209846960626e-05, + "loss": 0.8764, + "step": 14371 + }, + { + "epoch": 2.558760683760684, + "grad_norm": 0.7467330694198608, + "learning_rate": 5.7639529861764885e-05, + "loss": 0.6614, + "step": 14372 + }, + { + "epoch": 2.5589387464387463, + "grad_norm": 0.7930710315704346, + "learning_rate": 5.762685070653453e-05, + "loss": 0.6866, + "step": 14373 + }, + { + "epoch": 2.559116809116809, + "grad_norm": 0.9234277606010437, + "learning_rate": 5.7614172381518085e-05, + "loss": 0.9158, + "step": 14374 + }, + { + "epoch": 2.559294871794872, + "grad_norm": 1.0100786685943604, + "learning_rate": 5.7601494886963806e-05, + "loss": 0.9061, + "step": 14375 + }, + { + "epoch": 2.5594729344729346, + "grad_norm": 0.9864867925643921, + "learning_rate": 5.758881822312023e-05, + "loss": 0.9955, + "step": 14376 + }, + { + "epoch": 2.5596509971509973, + "grad_norm": 0.7328418493270874, + "learning_rate": 5.757614239023559e-05, + "loss": 0.874, + "step": 14377 + }, + { + "epoch": 2.5598290598290596, + "grad_norm": 0.8538700938224792, + "learning_rate": 5.7563467388558355e-05, + "loss": 0.8251, + "step": 14378 + }, + { + "epoch": 2.5600071225071224, + "grad_norm": 0.7603667378425598, + "learning_rate": 5.755079321833681e-05, + "loss": 0.8466, + "step": 14379 + }, + { + "epoch": 2.560185185185185, + "grad_norm": 0.8983954787254333, + "learning_rate": 5.753811987981925e-05, + "loss": 0.754, + "step": 14380 + }, + { + "epoch": 2.560363247863248, + "grad_norm": 0.8304823040962219, + "learning_rate": 5.752544737325411e-05, + "loss": 0.7057, + "step": 14381 + }, + { + "epoch": 2.5605413105413106, + "grad_norm": 0.8694877028465271, + "learning_rate": 5.751277569888952e-05, + "loss": 0.843, + "step": 14382 + }, + { + "epoch": 2.5607193732193734, + "grad_norm": 0.7965344786643982, + "learning_rate": 5.750010485697387e-05, + "loss": 0.7679, + "step": 14383 + }, + { + "epoch": 2.560897435897436, + "grad_norm": 0.8181809782981873, + "learning_rate": 5.7487434847755386e-05, + "loss": 0.8408, + "step": 14384 + }, + { + "epoch": 2.5610754985754984, + "grad_norm": 0.8492250442504883, + "learning_rate": 5.747476567148229e-05, + "loss": 0.6768, + "step": 14385 + }, + { + "epoch": 2.561253561253561, + "grad_norm": 0.9129379987716675, + "learning_rate": 5.746209732840282e-05, + "loss": 0.8804, + "step": 14386 + }, + { + "epoch": 2.561431623931624, + "grad_norm": 0.8701111078262329, + "learning_rate": 5.74494298187652e-05, + "loss": 1.0908, + "step": 14387 + }, + { + "epoch": 2.5616096866096867, + "grad_norm": 0.9152243733406067, + "learning_rate": 5.7436763142817606e-05, + "loss": 0.8197, + "step": 14388 + }, + { + "epoch": 2.5617877492877494, + "grad_norm": 0.8663429617881775, + "learning_rate": 5.742409730080822e-05, + "loss": 0.8065, + "step": 14389 + }, + { + "epoch": 2.5619658119658117, + "grad_norm": 0.8722090721130371, + "learning_rate": 5.741143229298516e-05, + "loss": 0.8634, + "step": 14390 + }, + { + "epoch": 2.5621438746438745, + "grad_norm": 0.8126732707023621, + "learning_rate": 5.7398768119596704e-05, + "loss": 0.831, + "step": 14391 + }, + { + "epoch": 2.5623219373219372, + "grad_norm": 0.9060684442520142, + "learning_rate": 5.7386104780890794e-05, + "loss": 0.8757, + "step": 14392 + }, + { + "epoch": 2.5625, + "grad_norm": 0.947692334651947, + "learning_rate": 5.7373442277115696e-05, + "loss": 0.8606, + "step": 14393 + }, + { + "epoch": 2.5626780626780628, + "grad_norm": 0.8826618790626526, + "learning_rate": 5.736078060851944e-05, + "loss": 0.9942, + "step": 14394 + }, + { + "epoch": 2.5628561253561255, + "grad_norm": 0.915372908115387, + "learning_rate": 5.734811977535011e-05, + "loss": 0.7871, + "step": 14395 + }, + { + "epoch": 2.5630341880341883, + "grad_norm": 0.8202184438705444, + "learning_rate": 5.733545977785577e-05, + "loss": 0.8889, + "step": 14396 + }, + { + "epoch": 2.5632122507122506, + "grad_norm": 1.0160186290740967, + "learning_rate": 5.7322800616284475e-05, + "loss": 0.8973, + "step": 14397 + }, + { + "epoch": 2.5633903133903133, + "grad_norm": 0.848753809928894, + "learning_rate": 5.7310142290884206e-05, + "loss": 0.9517, + "step": 14398 + }, + { + "epoch": 2.563568376068376, + "grad_norm": 0.7473777532577515, + "learning_rate": 5.72974848019031e-05, + "loss": 0.6549, + "step": 14399 + }, + { + "epoch": 2.563746438746439, + "grad_norm": 0.7396529316902161, + "learning_rate": 5.728482814958899e-05, + "loss": 0.7814, + "step": 14400 + }, + { + "epoch": 2.5639245014245016, + "grad_norm": 0.8617672920227051, + "learning_rate": 5.727217233418998e-05, + "loss": 0.885, + "step": 14401 + }, + { + "epoch": 2.564102564102564, + "grad_norm": 0.920477569103241, + "learning_rate": 5.7259517355953984e-05, + "loss": 1.0269, + "step": 14402 + }, + { + "epoch": 2.5642806267806266, + "grad_norm": 0.8460386991500854, + "learning_rate": 5.7246863215128975e-05, + "loss": 0.7418, + "step": 14403 + }, + { + "epoch": 2.5644586894586894, + "grad_norm": 0.8857694268226624, + "learning_rate": 5.723420991196287e-05, + "loss": 0.7771, + "step": 14404 + }, + { + "epoch": 2.564636752136752, + "grad_norm": 0.9912863969802856, + "learning_rate": 5.722155744670352e-05, + "loss": 1.101, + "step": 14405 + }, + { + "epoch": 2.564814814814815, + "grad_norm": 0.8062789440155029, + "learning_rate": 5.720890581959899e-05, + "loss": 0.8602, + "step": 14406 + }, + { + "epoch": 2.5649928774928776, + "grad_norm": 0.8620314598083496, + "learning_rate": 5.719625503089698e-05, + "loss": 0.9433, + "step": 14407 + }, + { + "epoch": 2.5651709401709404, + "grad_norm": 0.8119623064994812, + "learning_rate": 5.718360508084546e-05, + "loss": 0.884, + "step": 14408 + }, + { + "epoch": 2.5653490028490027, + "grad_norm": 0.7872169613838196, + "learning_rate": 5.7170955969692265e-05, + "loss": 0.8247, + "step": 14409 + }, + { + "epoch": 2.5655270655270654, + "grad_norm": 0.8314040303230286, + "learning_rate": 5.715830769768522e-05, + "loss": 0.8643, + "step": 14410 + }, + { + "epoch": 2.565705128205128, + "grad_norm": 0.9003102779388428, + "learning_rate": 5.7145660265072145e-05, + "loss": 0.9426, + "step": 14411 + }, + { + "epoch": 2.565883190883191, + "grad_norm": 0.6572127938270569, + "learning_rate": 5.713301367210082e-05, + "loss": 0.4852, + "step": 14412 + }, + { + "epoch": 2.5660612535612537, + "grad_norm": 0.9557960629463196, + "learning_rate": 5.7120367919019044e-05, + "loss": 0.9281, + "step": 14413 + }, + { + "epoch": 2.566239316239316, + "grad_norm": 0.9009736180305481, + "learning_rate": 5.71077230060746e-05, + "loss": 1.0154, + "step": 14414 + }, + { + "epoch": 2.5664173789173788, + "grad_norm": 0.8672121167182922, + "learning_rate": 5.7095078933515175e-05, + "loss": 0.805, + "step": 14415 + }, + { + "epoch": 2.5665954415954415, + "grad_norm": 0.9077832698822021, + "learning_rate": 5.708243570158862e-05, + "loss": 0.7446, + "step": 14416 + }, + { + "epoch": 2.5667735042735043, + "grad_norm": 0.850246787071228, + "learning_rate": 5.706979331054252e-05, + "loss": 0.7773, + "step": 14417 + }, + { + "epoch": 2.566951566951567, + "grad_norm": 0.803983211517334, + "learning_rate": 5.705715176062467e-05, + "loss": 0.9361, + "step": 14418 + }, + { + "epoch": 2.5671296296296298, + "grad_norm": 0.8956922888755798, + "learning_rate": 5.704451105208273e-05, + "loss": 0.8962, + "step": 14419 + }, + { + "epoch": 2.5673076923076925, + "grad_norm": 0.8994067907333374, + "learning_rate": 5.703187118516433e-05, + "loss": 0.9902, + "step": 14420 + }, + { + "epoch": 2.567485754985755, + "grad_norm": 0.7383418679237366, + "learning_rate": 5.701923216011722e-05, + "loss": 0.8188, + "step": 14421 + }, + { + "epoch": 2.5676638176638176, + "grad_norm": 0.8397318720817566, + "learning_rate": 5.70065939771889e-05, + "loss": 0.8557, + "step": 14422 + }, + { + "epoch": 2.5678418803418803, + "grad_norm": 0.8804301023483276, + "learning_rate": 5.699395663662714e-05, + "loss": 0.7248, + "step": 14423 + }, + { + "epoch": 2.568019943019943, + "grad_norm": 0.8391412496566772, + "learning_rate": 5.698132013867938e-05, + "loss": 0.6986, + "step": 14424 + }, + { + "epoch": 2.568198005698006, + "grad_norm": 0.7337331771850586, + "learning_rate": 5.6968684483593334e-05, + "loss": 0.7911, + "step": 14425 + }, + { + "epoch": 2.568376068376068, + "grad_norm": 1.006412386894226, + "learning_rate": 5.695604967161652e-05, + "loss": 1.0131, + "step": 14426 + }, + { + "epoch": 2.568554131054131, + "grad_norm": 0.7777771353721619, + "learning_rate": 5.6943415702996494e-05, + "loss": 0.67, + "step": 14427 + }, + { + "epoch": 2.5687321937321936, + "grad_norm": 0.8864775896072388, + "learning_rate": 5.6930782577980803e-05, + "loss": 0.9513, + "step": 14428 + }, + { + "epoch": 2.5689102564102564, + "grad_norm": 0.8505052328109741, + "learning_rate": 5.691815029681695e-05, + "loss": 0.7213, + "step": 14429 + }, + { + "epoch": 2.569088319088319, + "grad_norm": 0.705781877040863, + "learning_rate": 5.6905518859752416e-05, + "loss": 0.8273, + "step": 14430 + }, + { + "epoch": 2.569266381766382, + "grad_norm": 0.7157384753227234, + "learning_rate": 5.689288826703479e-05, + "loss": 0.6854, + "step": 14431 + }, + { + "epoch": 2.5694444444444446, + "grad_norm": 0.871244490146637, + "learning_rate": 5.68802585189114e-05, + "loss": 0.8786, + "step": 14432 + }, + { + "epoch": 2.5696225071225074, + "grad_norm": 0.8742622137069702, + "learning_rate": 5.686762961562981e-05, + "loss": 0.7253, + "step": 14433 + }, + { + "epoch": 2.5698005698005697, + "grad_norm": 0.8194206357002258, + "learning_rate": 5.685500155743742e-05, + "loss": 0.9028, + "step": 14434 + }, + { + "epoch": 2.5699786324786325, + "grad_norm": 0.7505850195884705, + "learning_rate": 5.684237434458164e-05, + "loss": 0.7711, + "step": 14435 + }, + { + "epoch": 2.570156695156695, + "grad_norm": 0.9128859639167786, + "learning_rate": 5.6829747977309885e-05, + "loss": 0.8495, + "step": 14436 + }, + { + "epoch": 2.570334757834758, + "grad_norm": 0.6996384263038635, + "learning_rate": 5.681712245586954e-05, + "loss": 0.6938, + "step": 14437 + }, + { + "epoch": 2.5705128205128203, + "grad_norm": 0.8720461130142212, + "learning_rate": 5.680449778050798e-05, + "loss": 1.0547, + "step": 14438 + }, + { + "epoch": 2.570690883190883, + "grad_norm": 0.7767693996429443, + "learning_rate": 5.6791873951472544e-05, + "loss": 0.8718, + "step": 14439 + }, + { + "epoch": 2.5708689458689458, + "grad_norm": 0.8596739768981934, + "learning_rate": 5.6779250969010554e-05, + "loss": 0.792, + "step": 14440 + }, + { + "epoch": 2.5710470085470085, + "grad_norm": 1.0065197944641113, + "learning_rate": 5.676662883336939e-05, + "loss": 0.9199, + "step": 14441 + }, + { + "epoch": 2.5712250712250713, + "grad_norm": 0.8707680702209473, + "learning_rate": 5.6754007544796316e-05, + "loss": 0.9696, + "step": 14442 + }, + { + "epoch": 2.571403133903134, + "grad_norm": 0.8060235977172852, + "learning_rate": 5.674138710353865e-05, + "loss": 0.6626, + "step": 14443 + }, + { + "epoch": 2.5715811965811968, + "grad_norm": 0.7640239596366882, + "learning_rate": 5.6728767509843627e-05, + "loss": 0.8177, + "step": 14444 + }, + { + "epoch": 2.5717592592592595, + "grad_norm": 1.0821335315704346, + "learning_rate": 5.671614876395848e-05, + "loss": 1.0084, + "step": 14445 + }, + { + "epoch": 2.571937321937322, + "grad_norm": 0.874721884727478, + "learning_rate": 5.670353086613056e-05, + "loss": 0.9508, + "step": 14446 + }, + { + "epoch": 2.5721153846153846, + "grad_norm": 0.7837753891944885, + "learning_rate": 5.669091381660694e-05, + "loss": 0.6546, + "step": 14447 + }, + { + "epoch": 2.5722934472934473, + "grad_norm": 0.832924485206604, + "learning_rate": 5.6678297615634965e-05, + "loss": 0.9055, + "step": 14448 + }, + { + "epoch": 2.57247150997151, + "grad_norm": 0.8463562726974487, + "learning_rate": 5.6665682263461696e-05, + "loss": 0.8234, + "step": 14449 + }, + { + "epoch": 2.5726495726495724, + "grad_norm": 0.8785214424133301, + "learning_rate": 5.6653067760334386e-05, + "loss": 0.8478, + "step": 14450 + }, + { + "epoch": 2.572827635327635, + "grad_norm": 0.7375151515007019, + "learning_rate": 5.664045410650017e-05, + "loss": 0.8629, + "step": 14451 + }, + { + "epoch": 2.573005698005698, + "grad_norm": 0.7428547143936157, + "learning_rate": 5.6627841302206196e-05, + "loss": 0.9198, + "step": 14452 + }, + { + "epoch": 2.5731837606837606, + "grad_norm": 0.7373468279838562, + "learning_rate": 5.661522934769956e-05, + "loss": 0.6931, + "step": 14453 + }, + { + "epoch": 2.5733618233618234, + "grad_norm": 0.9162034392356873, + "learning_rate": 5.660261824322739e-05, + "loss": 0.9971, + "step": 14454 + }, + { + "epoch": 2.573539886039886, + "grad_norm": 0.7816632390022278, + "learning_rate": 5.659000798903672e-05, + "loss": 1.0481, + "step": 14455 + }, + { + "epoch": 2.573717948717949, + "grad_norm": 0.8594158291816711, + "learning_rate": 5.657739858537474e-05, + "loss": 1.1846, + "step": 14456 + }, + { + "epoch": 2.5738960113960117, + "grad_norm": 0.8171747922897339, + "learning_rate": 5.656479003248836e-05, + "loss": 0.8435, + "step": 14457 + }, + { + "epoch": 2.574074074074074, + "grad_norm": 0.8568267822265625, + "learning_rate": 5.6552182330624784e-05, + "loss": 1.031, + "step": 14458 + }, + { + "epoch": 2.5742521367521367, + "grad_norm": 0.8238523602485657, + "learning_rate": 5.653957548003084e-05, + "loss": 0.8917, + "step": 14459 + }, + { + "epoch": 2.5744301994301995, + "grad_norm": 0.7226746082305908, + "learning_rate": 5.652696948095369e-05, + "loss": 0.832, + "step": 14460 + }, + { + "epoch": 2.574608262108262, + "grad_norm": 0.9448554515838623, + "learning_rate": 5.651436433364024e-05, + "loss": 0.9696, + "step": 14461 + }, + { + "epoch": 2.5747863247863245, + "grad_norm": 0.9404924511909485, + "learning_rate": 5.650176003833747e-05, + "loss": 0.9813, + "step": 14462 + }, + { + "epoch": 2.5749643874643873, + "grad_norm": 0.9445366859436035, + "learning_rate": 5.648915659529241e-05, + "loss": 0.7205, + "step": 14463 + }, + { + "epoch": 2.57514245014245, + "grad_norm": 0.9205772876739502, + "learning_rate": 5.647655400475189e-05, + "loss": 0.958, + "step": 14464 + }, + { + "epoch": 2.5753205128205128, + "grad_norm": 0.9025790691375732, + "learning_rate": 5.646395226696291e-05, + "loss": 0.9107, + "step": 14465 + }, + { + "epoch": 2.5754985754985755, + "grad_norm": 0.9562451839447021, + "learning_rate": 5.645135138217235e-05, + "loss": 0.7618, + "step": 14466 + }, + { + "epoch": 2.5756766381766383, + "grad_norm": 0.8896244764328003, + "learning_rate": 5.6438751350627085e-05, + "loss": 0.9696, + "step": 14467 + }, + { + "epoch": 2.575854700854701, + "grad_norm": 0.9051744937896729, + "learning_rate": 5.6426152172574e-05, + "loss": 0.9537, + "step": 14468 + }, + { + "epoch": 2.576032763532764, + "grad_norm": 0.844556450843811, + "learning_rate": 5.641355384825995e-05, + "loss": 0.8686, + "step": 14469 + }, + { + "epoch": 2.576210826210826, + "grad_norm": 0.7751742601394653, + "learning_rate": 5.6400956377931726e-05, + "loss": 0.8373, + "step": 14470 + }, + { + "epoch": 2.576388888888889, + "grad_norm": 0.8988052010536194, + "learning_rate": 5.638835976183627e-05, + "loss": 0.8661, + "step": 14471 + }, + { + "epoch": 2.5765669515669516, + "grad_norm": 0.9114456176757812, + "learning_rate": 5.637576400022023e-05, + "loss": 1.0583, + "step": 14472 + }, + { + "epoch": 2.5767450142450143, + "grad_norm": 0.8742861151695251, + "learning_rate": 5.636316909333056e-05, + "loss": 0.8392, + "step": 14473 + }, + { + "epoch": 2.5769230769230766, + "grad_norm": 0.8418447375297546, + "learning_rate": 5.6350575041413854e-05, + "loss": 0.7494, + "step": 14474 + }, + { + "epoch": 2.5771011396011394, + "grad_norm": 0.9942673444747925, + "learning_rate": 5.633798184471701e-05, + "loss": 1.0183, + "step": 14475 + }, + { + "epoch": 2.577279202279202, + "grad_norm": 0.7663289308547974, + "learning_rate": 5.63253895034867e-05, + "loss": 0.7551, + "step": 14476 + }, + { + "epoch": 2.577457264957265, + "grad_norm": 0.8866778016090393, + "learning_rate": 5.631279801796966e-05, + "loss": 0.8623, + "step": 14477 + }, + { + "epoch": 2.5776353276353277, + "grad_norm": 0.9198449850082397, + "learning_rate": 5.6300207388412595e-05, + "loss": 1.0388, + "step": 14478 + }, + { + "epoch": 2.5778133903133904, + "grad_norm": 0.8202611804008484, + "learning_rate": 5.628761761506214e-05, + "loss": 0.7556, + "step": 14479 + }, + { + "epoch": 2.577991452991453, + "grad_norm": 0.751899003982544, + "learning_rate": 5.627502869816505e-05, + "loss": 0.8231, + "step": 14480 + }, + { + "epoch": 2.578169515669516, + "grad_norm": 1.0094623565673828, + "learning_rate": 5.626244063796795e-05, + "loss": 0.9778, + "step": 14481 + }, + { + "epoch": 2.578347578347578, + "grad_norm": 0.8163259625434875, + "learning_rate": 5.624985343471747e-05, + "loss": 0.8355, + "step": 14482 + }, + { + "epoch": 2.578525641025641, + "grad_norm": 0.8190516829490662, + "learning_rate": 5.623726708866023e-05, + "loss": 0.7736, + "step": 14483 + }, + { + "epoch": 2.5787037037037037, + "grad_norm": 0.884303629398346, + "learning_rate": 5.622468160004283e-05, + "loss": 0.8618, + "step": 14484 + }, + { + "epoch": 2.5788817663817665, + "grad_norm": 0.8564121723175049, + "learning_rate": 5.621209696911185e-05, + "loss": 0.9691, + "step": 14485 + }, + { + "epoch": 2.5790598290598292, + "grad_norm": 0.8122418522834778, + "learning_rate": 5.619951319611388e-05, + "loss": 0.7539, + "step": 14486 + }, + { + "epoch": 2.5792378917378915, + "grad_norm": 0.764470636844635, + "learning_rate": 5.6186930281295425e-05, + "loss": 0.7065, + "step": 14487 + }, + { + "epoch": 2.5794159544159543, + "grad_norm": 0.7477477192878723, + "learning_rate": 5.617434822490313e-05, + "loss": 0.7456, + "step": 14488 + }, + { + "epoch": 2.579594017094017, + "grad_norm": 0.9460917711257935, + "learning_rate": 5.616176702718335e-05, + "loss": 0.8427, + "step": 14489 + }, + { + "epoch": 2.57977207977208, + "grad_norm": 0.858561098575592, + "learning_rate": 5.614918668838274e-05, + "loss": 0.8913, + "step": 14490 + }, + { + "epoch": 2.5799501424501425, + "grad_norm": 0.8664894104003906, + "learning_rate": 5.613660720874772e-05, + "loss": 0.9211, + "step": 14491 + }, + { + "epoch": 2.5801282051282053, + "grad_norm": 0.8475569486618042, + "learning_rate": 5.612402858852475e-05, + "loss": 0.8149, + "step": 14492 + }, + { + "epoch": 2.580306267806268, + "grad_norm": 0.9543033838272095, + "learning_rate": 5.6111450827960296e-05, + "loss": 0.679, + "step": 14493 + }, + { + "epoch": 2.5804843304843303, + "grad_norm": 0.8219496011734009, + "learning_rate": 5.60988739273008e-05, + "loss": 0.9586, + "step": 14494 + }, + { + "epoch": 2.580662393162393, + "grad_norm": 0.8432445526123047, + "learning_rate": 5.6086297886792684e-05, + "loss": 0.8939, + "step": 14495 + }, + { + "epoch": 2.580840455840456, + "grad_norm": 0.9026654362678528, + "learning_rate": 5.607372270668232e-05, + "loss": 0.7422, + "step": 14496 + }, + { + "epoch": 2.5810185185185186, + "grad_norm": 1.0681802034378052, + "learning_rate": 5.606114838721608e-05, + "loss": 0.8208, + "step": 14497 + }, + { + "epoch": 2.5811965811965814, + "grad_norm": 0.8807427883148193, + "learning_rate": 5.604857492864044e-05, + "loss": 0.8463, + "step": 14498 + }, + { + "epoch": 2.5813746438746437, + "grad_norm": 0.7520862221717834, + "learning_rate": 5.603600233120159e-05, + "loss": 0.6691, + "step": 14499 + }, + { + "epoch": 2.5815527065527064, + "grad_norm": 0.8214079737663269, + "learning_rate": 5.602343059514599e-05, + "loss": 0.8416, + "step": 14500 + }, + { + "epoch": 2.581730769230769, + "grad_norm": 0.9263389110565186, + "learning_rate": 5.601085972071991e-05, + "loss": 1.1466, + "step": 14501 + }, + { + "epoch": 2.581908831908832, + "grad_norm": 0.8501101136207581, + "learning_rate": 5.5998289708169626e-05, + "loss": 0.855, + "step": 14502 + }, + { + "epoch": 2.5820868945868947, + "grad_norm": 0.8312939405441284, + "learning_rate": 5.598572055774152e-05, + "loss": 0.9843, + "step": 14503 + }, + { + "epoch": 2.5822649572649574, + "grad_norm": 0.7309035658836365, + "learning_rate": 5.5973152269681714e-05, + "loss": 0.813, + "step": 14504 + }, + { + "epoch": 2.58244301994302, + "grad_norm": 0.8962578177452087, + "learning_rate": 5.596058484423656e-05, + "loss": 0.7619, + "step": 14505 + }, + { + "epoch": 2.5826210826210825, + "grad_norm": 0.7805112600326538, + "learning_rate": 5.594801828165228e-05, + "loss": 1.1011, + "step": 14506 + }, + { + "epoch": 2.5827991452991452, + "grad_norm": 1.224509358406067, + "learning_rate": 5.593545258217505e-05, + "loss": 0.9764, + "step": 14507 + }, + { + "epoch": 2.582977207977208, + "grad_norm": 0.8085877895355225, + "learning_rate": 5.59228877460511e-05, + "loss": 0.9324, + "step": 14508 + }, + { + "epoch": 2.5831552706552707, + "grad_norm": 0.7962629199028015, + "learning_rate": 5.591032377352661e-05, + "loss": 0.6294, + "step": 14509 + }, + { + "epoch": 2.5833333333333335, + "grad_norm": 0.8638611435890198, + "learning_rate": 5.589776066484773e-05, + "loss": 0.7355, + "step": 14510 + }, + { + "epoch": 2.583511396011396, + "grad_norm": 0.8975821733474731, + "learning_rate": 5.588519842026061e-05, + "loss": 1.0264, + "step": 14511 + }, + { + "epoch": 2.5836894586894585, + "grad_norm": 0.8327218890190125, + "learning_rate": 5.5872637040011355e-05, + "loss": 0.8864, + "step": 14512 + }, + { + "epoch": 2.5838675213675213, + "grad_norm": 0.8141334652900696, + "learning_rate": 5.5860076524346197e-05, + "loss": 1.0277, + "step": 14513 + }, + { + "epoch": 2.584045584045584, + "grad_norm": 0.8557519316673279, + "learning_rate": 5.584751687351105e-05, + "loss": 0.9215, + "step": 14514 + }, + { + "epoch": 2.584223646723647, + "grad_norm": 0.902601957321167, + "learning_rate": 5.583495808775214e-05, + "loss": 0.8527, + "step": 14515 + }, + { + "epoch": 2.5844017094017095, + "grad_norm": 0.826359212398529, + "learning_rate": 5.582240016731548e-05, + "loss": 0.8524, + "step": 14516 + }, + { + "epoch": 2.5845797720797723, + "grad_norm": 0.7099179029464722, + "learning_rate": 5.580984311244713e-05, + "loss": 0.6923, + "step": 14517 + }, + { + "epoch": 2.5847578347578346, + "grad_norm": 0.829795777797699, + "learning_rate": 5.5797286923393086e-05, + "loss": 0.7211, + "step": 14518 + }, + { + "epoch": 2.5849358974358974, + "grad_norm": 0.8006768226623535, + "learning_rate": 5.5784731600399355e-05, + "loss": 0.7237, + "step": 14519 + }, + { + "epoch": 2.58511396011396, + "grad_norm": 0.7596119046211243, + "learning_rate": 5.577217714371203e-05, + "loss": 0.7651, + "step": 14520 + }, + { + "epoch": 2.585292022792023, + "grad_norm": 0.7901585102081299, + "learning_rate": 5.575962355357694e-05, + "loss": 0.7672, + "step": 14521 + }, + { + "epoch": 2.5854700854700856, + "grad_norm": 0.8586403131484985, + "learning_rate": 5.574707083024018e-05, + "loss": 1.084, + "step": 14522 + }, + { + "epoch": 2.585648148148148, + "grad_norm": 0.7670607566833496, + "learning_rate": 5.5734518973947616e-05, + "loss": 0.7929, + "step": 14523 + }, + { + "epoch": 2.5858262108262107, + "grad_norm": 0.8114384412765503, + "learning_rate": 5.572196798494522e-05, + "loss": 0.8154, + "step": 14524 + }, + { + "epoch": 2.5860042735042734, + "grad_norm": 0.8050188422203064, + "learning_rate": 5.570941786347888e-05, + "loss": 0.7969, + "step": 14525 + }, + { + "epoch": 2.586182336182336, + "grad_norm": 0.8641461133956909, + "learning_rate": 5.569686860979447e-05, + "loss": 0.8469, + "step": 14526 + }, + { + "epoch": 2.586360398860399, + "grad_norm": 0.7644940614700317, + "learning_rate": 5.568432022413787e-05, + "loss": 0.563, + "step": 14527 + }, + { + "epoch": 2.5865384615384617, + "grad_norm": 0.7620565891265869, + "learning_rate": 5.567177270675503e-05, + "loss": 0.657, + "step": 14528 + }, + { + "epoch": 2.5867165242165244, + "grad_norm": 0.8371306657791138, + "learning_rate": 5.5659226057891634e-05, + "loss": 0.8862, + "step": 14529 + }, + { + "epoch": 2.5868945868945867, + "grad_norm": 0.8996389508247375, + "learning_rate": 5.564668027779367e-05, + "loss": 0.6031, + "step": 14530 + }, + { + "epoch": 2.5870726495726495, + "grad_norm": 0.8691734671592712, + "learning_rate": 5.5634135366706806e-05, + "loss": 0.9198, + "step": 14531 + }, + { + "epoch": 2.5872507122507122, + "grad_norm": 0.8926620483398438, + "learning_rate": 5.562159132487693e-05, + "loss": 0.8691, + "step": 14532 + }, + { + "epoch": 2.587428774928775, + "grad_norm": 1.0852068662643433, + "learning_rate": 5.5609048152549794e-05, + "loss": 1.2338, + "step": 14533 + }, + { + "epoch": 2.5876068376068377, + "grad_norm": 0.7894790172576904, + "learning_rate": 5.5596505849971124e-05, + "loss": 0.907, + "step": 14534 + }, + { + "epoch": 2.5877849002849, + "grad_norm": 0.8084964156150818, + "learning_rate": 5.558396441738669e-05, + "loss": 0.9082, + "step": 14535 + }, + { + "epoch": 2.587962962962963, + "grad_norm": 1.0563920736312866, + "learning_rate": 5.557142385504222e-05, + "loss": 1.0364, + "step": 14536 + }, + { + "epoch": 2.5881410256410255, + "grad_norm": 0.7996996641159058, + "learning_rate": 5.5558884163183354e-05, + "loss": 0.925, + "step": 14537 + }, + { + "epoch": 2.5883190883190883, + "grad_norm": 0.7493244409561157, + "learning_rate": 5.5546345342055916e-05, + "loss": 0.9516, + "step": 14538 + }, + { + "epoch": 2.588497150997151, + "grad_norm": 0.8916776776313782, + "learning_rate": 5.553380739190541e-05, + "loss": 0.8164, + "step": 14539 + }, + { + "epoch": 2.588675213675214, + "grad_norm": 0.8178156614303589, + "learning_rate": 5.552127031297762e-05, + "loss": 0.905, + "step": 14540 + }, + { + "epoch": 2.5888532763532766, + "grad_norm": 0.8305806517601013, + "learning_rate": 5.550873410551816e-05, + "loss": 0.789, + "step": 14541 + }, + { + "epoch": 2.5890313390313393, + "grad_norm": 0.9307064414024353, + "learning_rate": 5.549619876977258e-05, + "loss": 0.8529, + "step": 14542 + }, + { + "epoch": 2.5892094017094016, + "grad_norm": 0.8526419401168823, + "learning_rate": 5.5483664305986614e-05, + "loss": 0.8314, + "step": 14543 + }, + { + "epoch": 2.5893874643874644, + "grad_norm": 0.884918212890625, + "learning_rate": 5.547113071440568e-05, + "loss": 0.7957, + "step": 14544 + }, + { + "epoch": 2.589565527065527, + "grad_norm": 0.7517948746681213, + "learning_rate": 5.5458597995275554e-05, + "loss": 0.7012, + "step": 14545 + }, + { + "epoch": 2.58974358974359, + "grad_norm": 0.8321232199668884, + "learning_rate": 5.5446066148841556e-05, + "loss": 1.0017, + "step": 14546 + }, + { + "epoch": 2.589921652421652, + "grad_norm": 0.8279885053634644, + "learning_rate": 5.543353517534939e-05, + "loss": 0.987, + "step": 14547 + }, + { + "epoch": 2.590099715099715, + "grad_norm": 0.8651175498962402, + "learning_rate": 5.542100507504454e-05, + "loss": 0.8929, + "step": 14548 + }, + { + "epoch": 2.5902777777777777, + "grad_norm": 0.9273492097854614, + "learning_rate": 5.540847584817248e-05, + "loss": 0.9503, + "step": 14549 + }, + { + "epoch": 2.5904558404558404, + "grad_norm": 0.8779071569442749, + "learning_rate": 5.5395947494978696e-05, + "loss": 0.9099, + "step": 14550 + }, + { + "epoch": 2.590633903133903, + "grad_norm": 0.8860164880752563, + "learning_rate": 5.538342001570868e-05, + "loss": 0.9559, + "step": 14551 + }, + { + "epoch": 2.590811965811966, + "grad_norm": 0.9232339859008789, + "learning_rate": 5.5370893410607816e-05, + "loss": 0.9495, + "step": 14552 + }, + { + "epoch": 2.5909900284900287, + "grad_norm": 0.8176831007003784, + "learning_rate": 5.5358367679921666e-05, + "loss": 0.8897, + "step": 14553 + }, + { + "epoch": 2.5911680911680914, + "grad_norm": 0.7926605939865112, + "learning_rate": 5.5345842823895486e-05, + "loss": 0.7609, + "step": 14554 + }, + { + "epoch": 2.5913461538461537, + "grad_norm": 0.9837173819541931, + "learning_rate": 5.533331884277484e-05, + "loss": 0.7842, + "step": 14555 + }, + { + "epoch": 2.5915242165242165, + "grad_norm": 0.7303726673126221, + "learning_rate": 5.5320795736804945e-05, + "loss": 0.824, + "step": 14556 + }, + { + "epoch": 2.5917022792022792, + "grad_norm": 0.8379296660423279, + "learning_rate": 5.530827350623128e-05, + "loss": 0.8005, + "step": 14557 + }, + { + "epoch": 2.591880341880342, + "grad_norm": 0.8562047481536865, + "learning_rate": 5.529575215129916e-05, + "loss": 1.0048, + "step": 14558 + }, + { + "epoch": 2.5920584045584043, + "grad_norm": 0.7543022632598877, + "learning_rate": 5.528323167225386e-05, + "loss": 0.7543, + "step": 14559 + }, + { + "epoch": 2.592236467236467, + "grad_norm": 0.8205977082252502, + "learning_rate": 5.5270712069340847e-05, + "loss": 0.997, + "step": 14560 + }, + { + "epoch": 2.59241452991453, + "grad_norm": 0.8566918969154358, + "learning_rate": 5.525819334280522e-05, + "loss": 0.9222, + "step": 14561 + }, + { + "epoch": 2.5925925925925926, + "grad_norm": 0.8513971567153931, + "learning_rate": 5.524567549289239e-05, + "loss": 0.7007, + "step": 14562 + }, + { + "epoch": 2.5927706552706553, + "grad_norm": 0.8939194679260254, + "learning_rate": 5.523315851984758e-05, + "loss": 0.8597, + "step": 14563 + }, + { + "epoch": 2.592948717948718, + "grad_norm": 0.7597625851631165, + "learning_rate": 5.5220642423916035e-05, + "loss": 0.7122, + "step": 14564 + }, + { + "epoch": 2.593126780626781, + "grad_norm": 0.9511955976486206, + "learning_rate": 5.5208127205342983e-05, + "loss": 1.0905, + "step": 14565 + }, + { + "epoch": 2.5933048433048436, + "grad_norm": 0.8359304070472717, + "learning_rate": 5.5195612864373626e-05, + "loss": 0.7132, + "step": 14566 + }, + { + "epoch": 2.593482905982906, + "grad_norm": 0.8302733302116394, + "learning_rate": 5.518309940125317e-05, + "loss": 0.9123, + "step": 14567 + }, + { + "epoch": 2.5936609686609686, + "grad_norm": 0.7923629283905029, + "learning_rate": 5.517058681622678e-05, + "loss": 0.8384, + "step": 14568 + }, + { + "epoch": 2.5938390313390314, + "grad_norm": 1.0625137090682983, + "learning_rate": 5.515807510953956e-05, + "loss": 1.0262, + "step": 14569 + }, + { + "epoch": 2.594017094017094, + "grad_norm": 1.0595879554748535, + "learning_rate": 5.5145564281436804e-05, + "loss": 0.9112, + "step": 14570 + }, + { + "epoch": 2.5941951566951564, + "grad_norm": 0.7307499647140503, + "learning_rate": 5.513305433216346e-05, + "loss": 0.8273, + "step": 14571 + }, + { + "epoch": 2.594373219373219, + "grad_norm": 0.9221912026405334, + "learning_rate": 5.512054526196475e-05, + "loss": 1.0679, + "step": 14572 + }, + { + "epoch": 2.594551282051282, + "grad_norm": 0.8098722100257874, + "learning_rate": 5.5108037071085725e-05, + "loss": 0.922, + "step": 14573 + }, + { + "epoch": 2.5947293447293447, + "grad_norm": 0.984785258769989, + "learning_rate": 5.509552975977146e-05, + "loss": 0.7525, + "step": 14574 + }, + { + "epoch": 2.5949074074074074, + "grad_norm": 0.8076850771903992, + "learning_rate": 5.5083023328267006e-05, + "loss": 1.008, + "step": 14575 + }, + { + "epoch": 2.59508547008547, + "grad_norm": 0.8375436067581177, + "learning_rate": 5.507051777681741e-05, + "loss": 0.8822, + "step": 14576 + }, + { + "epoch": 2.595263532763533, + "grad_norm": 0.779228687286377, + "learning_rate": 5.505801310566764e-05, + "loss": 0.8072, + "step": 14577 + }, + { + "epoch": 2.5954415954415957, + "grad_norm": 0.7347875833511353, + "learning_rate": 5.504550931506278e-05, + "loss": 0.7796, + "step": 14578 + }, + { + "epoch": 2.595619658119658, + "grad_norm": 0.8229580521583557, + "learning_rate": 5.503300640524779e-05, + "loss": 0.9337, + "step": 14579 + }, + { + "epoch": 2.5957977207977208, + "grad_norm": 0.8643096089363098, + "learning_rate": 5.502050437646762e-05, + "loss": 0.9101, + "step": 14580 + }, + { + "epoch": 2.5959757834757835, + "grad_norm": 0.769158661365509, + "learning_rate": 5.500800322896723e-05, + "loss": 0.8417, + "step": 14581 + }, + { + "epoch": 2.5961538461538463, + "grad_norm": 0.7792086005210876, + "learning_rate": 5.4995502962991566e-05, + "loss": 0.6965, + "step": 14582 + }, + { + "epoch": 2.5963319088319086, + "grad_norm": 0.7833219170570374, + "learning_rate": 5.498300357878552e-05, + "loss": 0.641, + "step": 14583 + }, + { + "epoch": 2.5965099715099713, + "grad_norm": 0.9491978287696838, + "learning_rate": 5.4970505076593956e-05, + "loss": 0.9229, + "step": 14584 + }, + { + "epoch": 2.596688034188034, + "grad_norm": 0.9128090739250183, + "learning_rate": 5.495800745666191e-05, + "loss": 0.8047, + "step": 14585 + }, + { + "epoch": 2.596866096866097, + "grad_norm": 0.9235281944274902, + "learning_rate": 5.494551071923404e-05, + "loss": 0.961, + "step": 14586 + }, + { + "epoch": 2.5970441595441596, + "grad_norm": 0.8582631349563599, + "learning_rate": 5.493301486455536e-05, + "loss": 0.7203, + "step": 14587 + }, + { + "epoch": 2.5972222222222223, + "grad_norm": 0.9605505466461182, + "learning_rate": 5.4920519892870605e-05, + "loss": 0.8315, + "step": 14588 + }, + { + "epoch": 2.597400284900285, + "grad_norm": 0.9344304203987122, + "learning_rate": 5.490802580442462e-05, + "loss": 0.9031, + "step": 14589 + }, + { + "epoch": 2.597578347578348, + "grad_norm": 1.0027791261672974, + "learning_rate": 5.4895532599462216e-05, + "loss": 1.0361, + "step": 14590 + }, + { + "epoch": 2.59775641025641, + "grad_norm": 0.8774647116661072, + "learning_rate": 5.488304027822815e-05, + "loss": 0.9533, + "step": 14591 + }, + { + "epoch": 2.597934472934473, + "grad_norm": 0.886246919631958, + "learning_rate": 5.487054884096718e-05, + "loss": 0.8588, + "step": 14592 + }, + { + "epoch": 2.5981125356125356, + "grad_norm": 0.8963425755500793, + "learning_rate": 5.485805828792408e-05, + "loss": 0.8685, + "step": 14593 + }, + { + "epoch": 2.5982905982905984, + "grad_norm": 0.7650768756866455, + "learning_rate": 5.484556861934349e-05, + "loss": 0.7441, + "step": 14594 + }, + { + "epoch": 2.5984686609686607, + "grad_norm": 0.8266916871070862, + "learning_rate": 5.483307983547026e-05, + "loss": 0.9625, + "step": 14595 + }, + { + "epoch": 2.5986467236467234, + "grad_norm": 0.8243923783302307, + "learning_rate": 5.482059193654894e-05, + "loss": 0.8553, + "step": 14596 + }, + { + "epoch": 2.598824786324786, + "grad_norm": 0.8200470209121704, + "learning_rate": 5.48081049228243e-05, + "loss": 0.6682, + "step": 14597 + }, + { + "epoch": 2.599002849002849, + "grad_norm": 0.8360442519187927, + "learning_rate": 5.479561879454097e-05, + "loss": 0.8996, + "step": 14598 + }, + { + "epoch": 2.5991809116809117, + "grad_norm": 0.8326625227928162, + "learning_rate": 5.4783133551943546e-05, + "loss": 0.6532, + "step": 14599 + }, + { + "epoch": 2.5993589743589745, + "grad_norm": 0.8162251114845276, + "learning_rate": 5.4770649195276766e-05, + "loss": 1.0514, + "step": 14600 + }, + { + "epoch": 2.599537037037037, + "grad_norm": 1.0407251119613647, + "learning_rate": 5.4758165724785084e-05, + "loss": 0.7991, + "step": 14601 + }, + { + "epoch": 2.5997150997151, + "grad_norm": 0.9161550998687744, + "learning_rate": 5.474568314071323e-05, + "loss": 0.8623, + "step": 14602 + }, + { + "epoch": 2.5998931623931623, + "grad_norm": 0.8405734896659851, + "learning_rate": 5.4733201443305646e-05, + "loss": 0.8406, + "step": 14603 + }, + { + "epoch": 2.600071225071225, + "grad_norm": 0.937198281288147, + "learning_rate": 5.472072063280698e-05, + "loss": 1.0887, + "step": 14604 + }, + { + "epoch": 2.6002492877492878, + "grad_norm": 0.8800520896911621, + "learning_rate": 5.470824070946172e-05, + "loss": 0.8738, + "step": 14605 + }, + { + "epoch": 2.6004273504273505, + "grad_norm": 0.9473027586936951, + "learning_rate": 5.4695761673514425e-05, + "loss": 0.8188, + "step": 14606 + }, + { + "epoch": 2.6006054131054133, + "grad_norm": 0.8547683954238892, + "learning_rate": 5.468328352520955e-05, + "loss": 0.7619, + "step": 14607 + }, + { + "epoch": 2.6007834757834756, + "grad_norm": 1.0138040781021118, + "learning_rate": 5.4670806264791595e-05, + "loss": 1.0805, + "step": 14608 + }, + { + "epoch": 2.6009615384615383, + "grad_norm": 0.8458215594291687, + "learning_rate": 5.465832989250499e-05, + "loss": 0.8386, + "step": 14609 + }, + { + "epoch": 2.601139601139601, + "grad_norm": 0.811152458190918, + "learning_rate": 5.464585440859431e-05, + "loss": 0.8158, + "step": 14610 + }, + { + "epoch": 2.601317663817664, + "grad_norm": 0.9584031701087952, + "learning_rate": 5.463337981330381e-05, + "loss": 0.8537, + "step": 14611 + }, + { + "epoch": 2.6014957264957266, + "grad_norm": 0.8734773397445679, + "learning_rate": 5.462090610687802e-05, + "loss": 1.0246, + "step": 14612 + }, + { + "epoch": 2.6016737891737893, + "grad_norm": 0.8463562726974487, + "learning_rate": 5.460843328956133e-05, + "loss": 0.8763, + "step": 14613 + }, + { + "epoch": 2.601851851851852, + "grad_norm": 0.8010903000831604, + "learning_rate": 5.459596136159808e-05, + "loss": 0.8438, + "step": 14614 + }, + { + "epoch": 2.6020299145299144, + "grad_norm": 0.7927500009536743, + "learning_rate": 5.458349032323267e-05, + "loss": 0.7388, + "step": 14615 + }, + { + "epoch": 2.602207977207977, + "grad_norm": 0.784017026424408, + "learning_rate": 5.4571020174709407e-05, + "loss": 0.6981, + "step": 14616 + }, + { + "epoch": 2.60238603988604, + "grad_norm": 0.8732004761695862, + "learning_rate": 5.455855091627263e-05, + "loss": 1.043, + "step": 14617 + }, + { + "epoch": 2.6025641025641026, + "grad_norm": 0.7947654128074646, + "learning_rate": 5.454608254816662e-05, + "loss": 0.9487, + "step": 14618 + }, + { + "epoch": 2.6027421652421654, + "grad_norm": 0.8809077739715576, + "learning_rate": 5.4533615070635734e-05, + "loss": 0.8499, + "step": 14619 + }, + { + "epoch": 2.6029202279202277, + "grad_norm": 0.9094803333282471, + "learning_rate": 5.452114848392422e-05, + "loss": 0.9522, + "step": 14620 + }, + { + "epoch": 2.6030982905982905, + "grad_norm": 0.8943446278572083, + "learning_rate": 5.4508682788276324e-05, + "loss": 0.7328, + "step": 14621 + }, + { + "epoch": 2.603276353276353, + "grad_norm": 0.856849730014801, + "learning_rate": 5.449621798393628e-05, + "loss": 0.7536, + "step": 14622 + }, + { + "epoch": 2.603454415954416, + "grad_norm": 0.8199608325958252, + "learning_rate": 5.448375407114833e-05, + "loss": 0.6377, + "step": 14623 + }, + { + "epoch": 2.6036324786324787, + "grad_norm": 0.8981915712356567, + "learning_rate": 5.4471291050156626e-05, + "loss": 1.0372, + "step": 14624 + }, + { + "epoch": 2.6038105413105415, + "grad_norm": 0.8449446558952332, + "learning_rate": 5.4458828921205465e-05, + "loss": 0.9948, + "step": 14625 + }, + { + "epoch": 2.603988603988604, + "grad_norm": 0.8807474970817566, + "learning_rate": 5.444636768453888e-05, + "loss": 0.9752, + "step": 14626 + }, + { + "epoch": 2.6041666666666665, + "grad_norm": 0.8212316036224365, + "learning_rate": 5.443390734040117e-05, + "loss": 0.9221, + "step": 14627 + }, + { + "epoch": 2.6043447293447293, + "grad_norm": 0.8049453496932983, + "learning_rate": 5.4421447889036304e-05, + "loss": 0.7726, + "step": 14628 + }, + { + "epoch": 2.604522792022792, + "grad_norm": 0.8091840744018555, + "learning_rate": 5.440898933068853e-05, + "loss": 0.9152, + "step": 14629 + }, + { + "epoch": 2.6047008547008548, + "grad_norm": 0.8409022688865662, + "learning_rate": 5.43965316656019e-05, + "loss": 0.8672, + "step": 14630 + }, + { + "epoch": 2.6048789173789175, + "grad_norm": 0.7622308731079102, + "learning_rate": 5.4384074894020496e-05, + "loss": 0.9021, + "step": 14631 + }, + { + "epoch": 2.60505698005698, + "grad_norm": 0.8272425532341003, + "learning_rate": 5.437161901618839e-05, + "loss": 0.7729, + "step": 14632 + }, + { + "epoch": 2.6052350427350426, + "grad_norm": 0.8699020743370056, + "learning_rate": 5.435916403234963e-05, + "loss": 0.8211, + "step": 14633 + }, + { + "epoch": 2.6054131054131053, + "grad_norm": 0.8145751357078552, + "learning_rate": 5.4346709942748196e-05, + "loss": 0.8996, + "step": 14634 + }, + { + "epoch": 2.605591168091168, + "grad_norm": 0.9398832321166992, + "learning_rate": 5.433425674762822e-05, + "loss": 0.8116, + "step": 14635 + }, + { + "epoch": 2.605769230769231, + "grad_norm": 0.9191767573356628, + "learning_rate": 5.4321804447233535e-05, + "loss": 0.8933, + "step": 14636 + }, + { + "epoch": 2.6059472934472936, + "grad_norm": 0.7511529326438904, + "learning_rate": 5.430935304180831e-05, + "loss": 0.7595, + "step": 14637 + }, + { + "epoch": 2.6061253561253563, + "grad_norm": 0.9087170362472534, + "learning_rate": 5.4296902531596296e-05, + "loss": 0.9781, + "step": 14638 + }, + { + "epoch": 2.6063034188034186, + "grad_norm": 0.8496448397636414, + "learning_rate": 5.4284452916841575e-05, + "loss": 0.9852, + "step": 14639 + }, + { + "epoch": 2.6064814814814814, + "grad_norm": 0.868609607219696, + "learning_rate": 5.427200419778804e-05, + "loss": 0.781, + "step": 14640 + }, + { + "epoch": 2.606659544159544, + "grad_norm": 0.7752132415771484, + "learning_rate": 5.4259556374679553e-05, + "loss": 0.7319, + "step": 14641 + }, + { + "epoch": 2.606837606837607, + "grad_norm": 0.8950543999671936, + "learning_rate": 5.4247109447760124e-05, + "loss": 0.7637, + "step": 14642 + }, + { + "epoch": 2.6070156695156697, + "grad_norm": 0.892699658870697, + "learning_rate": 5.423466341727346e-05, + "loss": 0.8274, + "step": 14643 + }, + { + "epoch": 2.607193732193732, + "grad_norm": 0.9283786416053772, + "learning_rate": 5.422221828346352e-05, + "loss": 1.1009, + "step": 14644 + }, + { + "epoch": 2.6073717948717947, + "grad_norm": 0.7551446557044983, + "learning_rate": 5.420977404657413e-05, + "loss": 0.8105, + "step": 14645 + }, + { + "epoch": 2.6075498575498575, + "grad_norm": 0.8014101386070251, + "learning_rate": 5.41973307068491e-05, + "loss": 0.838, + "step": 14646 + }, + { + "epoch": 2.60772792022792, + "grad_norm": 0.8941731452941895, + "learning_rate": 5.418488826453223e-05, + "loss": 0.9557, + "step": 14647 + }, + { + "epoch": 2.607905982905983, + "grad_norm": 0.7990903258323669, + "learning_rate": 5.41724467198673e-05, + "loss": 0.9634, + "step": 14648 + }, + { + "epoch": 2.6080840455840457, + "grad_norm": 1.0688040256500244, + "learning_rate": 5.4160006073098035e-05, + "loss": 0.8976, + "step": 14649 + }, + { + "epoch": 2.6082621082621085, + "grad_norm": 0.8451266884803772, + "learning_rate": 5.4147566324468313e-05, + "loss": 0.8703, + "step": 14650 + }, + { + "epoch": 2.6084401709401708, + "grad_norm": 0.8196333050727844, + "learning_rate": 5.413512747422169e-05, + "loss": 0.9423, + "step": 14651 + }, + { + "epoch": 2.6086182336182335, + "grad_norm": 0.7639298439025879, + "learning_rate": 5.412268952260204e-05, + "loss": 0.9092, + "step": 14652 + }, + { + "epoch": 2.6087962962962963, + "grad_norm": 0.88963782787323, + "learning_rate": 5.411025246985293e-05, + "loss": 1.2503, + "step": 14653 + }, + { + "epoch": 2.608974358974359, + "grad_norm": 0.831516683101654, + "learning_rate": 5.409781631621812e-05, + "loss": 0.8643, + "step": 14654 + }, + { + "epoch": 2.609152421652422, + "grad_norm": 0.7729721069335938, + "learning_rate": 5.408538106194125e-05, + "loss": 0.8289, + "step": 14655 + }, + { + "epoch": 2.609330484330484, + "grad_norm": 0.8360101580619812, + "learning_rate": 5.407294670726596e-05, + "loss": 0.8619, + "step": 14656 + }, + { + "epoch": 2.609508547008547, + "grad_norm": 0.7525733709335327, + "learning_rate": 5.406051325243586e-05, + "loss": 0.8353, + "step": 14657 + }, + { + "epoch": 2.6096866096866096, + "grad_norm": 0.8943357467651367, + "learning_rate": 5.404808069769456e-05, + "loss": 0.9291, + "step": 14658 + }, + { + "epoch": 2.6098646723646723, + "grad_norm": 1.024953007698059, + "learning_rate": 5.403564904328568e-05, + "loss": 1.0414, + "step": 14659 + }, + { + "epoch": 2.610042735042735, + "grad_norm": 0.8671780228614807, + "learning_rate": 5.402321828945278e-05, + "loss": 0.9309, + "step": 14660 + }, + { + "epoch": 2.610220797720798, + "grad_norm": 0.9765334725379944, + "learning_rate": 5.4010788436439406e-05, + "loss": 0.9399, + "step": 14661 + }, + { + "epoch": 2.6103988603988606, + "grad_norm": 0.8996732234954834, + "learning_rate": 5.3998359484489106e-05, + "loss": 0.9868, + "step": 14662 + }, + { + "epoch": 2.6105769230769234, + "grad_norm": 0.8597404956817627, + "learning_rate": 5.398593143384538e-05, + "loss": 1.0328, + "step": 14663 + }, + { + "epoch": 2.6107549857549857, + "grad_norm": 0.8909318447113037, + "learning_rate": 5.397350428475176e-05, + "loss": 0.9362, + "step": 14664 + }, + { + "epoch": 2.6109330484330484, + "grad_norm": 0.8874006867408752, + "learning_rate": 5.39610780374517e-05, + "loss": 0.9254, + "step": 14665 + }, + { + "epoch": 2.611111111111111, + "grad_norm": 0.8325822949409485, + "learning_rate": 5.3948652692188626e-05, + "loss": 0.8495, + "step": 14666 + }, + { + "epoch": 2.611289173789174, + "grad_norm": 0.847998857498169, + "learning_rate": 5.393622824920614e-05, + "loss": 0.8372, + "step": 14667 + }, + { + "epoch": 2.611467236467236, + "grad_norm": 0.8439756631851196, + "learning_rate": 5.392380470874749e-05, + "loss": 0.8934, + "step": 14668 + }, + { + "epoch": 2.611645299145299, + "grad_norm": 0.9563834071159363, + "learning_rate": 5.39113820710562e-05, + "loss": 1.1213, + "step": 14669 + }, + { + "epoch": 2.6118233618233617, + "grad_norm": 0.7761119604110718, + "learning_rate": 5.3898960336375646e-05, + "loss": 0.6104, + "step": 14670 + }, + { + "epoch": 2.6120014245014245, + "grad_norm": 0.8661524653434753, + "learning_rate": 5.38865395049492e-05, + "loss": 0.7562, + "step": 14671 + }, + { + "epoch": 2.6121794871794872, + "grad_norm": 0.854347825050354, + "learning_rate": 5.387411957702021e-05, + "loss": 0.8613, + "step": 14672 + }, + { + "epoch": 2.61235754985755, + "grad_norm": 0.7728402614593506, + "learning_rate": 5.386170055283204e-05, + "loss": 0.7879, + "step": 14673 + }, + { + "epoch": 2.6125356125356127, + "grad_norm": 0.8647109270095825, + "learning_rate": 5.384928243262799e-05, + "loss": 0.835, + "step": 14674 + }, + { + "epoch": 2.6127136752136755, + "grad_norm": 0.6764749884605408, + "learning_rate": 5.383686521665139e-05, + "loss": 0.7233, + "step": 14675 + }, + { + "epoch": 2.612891737891738, + "grad_norm": 0.8431640863418579, + "learning_rate": 5.382444890514548e-05, + "loss": 1.1699, + "step": 14676 + }, + { + "epoch": 2.6130698005698005, + "grad_norm": 0.9196193814277649, + "learning_rate": 5.381203349835364e-05, + "loss": 0.8668, + "step": 14677 + }, + { + "epoch": 2.6132478632478633, + "grad_norm": 0.9449048638343811, + "learning_rate": 5.3799618996519e-05, + "loss": 0.8353, + "step": 14678 + }, + { + "epoch": 2.613425925925926, + "grad_norm": 0.9835928678512573, + "learning_rate": 5.378720539988488e-05, + "loss": 1.1129, + "step": 14679 + }, + { + "epoch": 2.6136039886039883, + "grad_norm": 0.763592004776001, + "learning_rate": 5.377479270869448e-05, + "loss": 0.7929, + "step": 14680 + }, + { + "epoch": 2.613782051282051, + "grad_norm": 0.8119748830795288, + "learning_rate": 5.376238092319094e-05, + "loss": 1.0257, + "step": 14681 + }, + { + "epoch": 2.613960113960114, + "grad_norm": 0.7605236172676086, + "learning_rate": 5.374997004361757e-05, + "loss": 0.7005, + "step": 14682 + }, + { + "epoch": 2.6141381766381766, + "grad_norm": 0.9077369570732117, + "learning_rate": 5.3737560070217394e-05, + "loss": 0.9208, + "step": 14683 + }, + { + "epoch": 2.6143162393162394, + "grad_norm": 0.9089310765266418, + "learning_rate": 5.3725151003233665e-05, + "loss": 0.6855, + "step": 14684 + }, + { + "epoch": 2.614494301994302, + "grad_norm": 0.8387685418128967, + "learning_rate": 5.371274284290947e-05, + "loss": 0.8682, + "step": 14685 + }, + { + "epoch": 2.614672364672365, + "grad_norm": 0.7626301050186157, + "learning_rate": 5.3700335589487925e-05, + "loss": 0.6928, + "step": 14686 + }, + { + "epoch": 2.6148504273504276, + "grad_norm": 1.2667319774627686, + "learning_rate": 5.368792924321213e-05, + "loss": 0.9288, + "step": 14687 + }, + { + "epoch": 2.61502849002849, + "grad_norm": 0.8570333123207092, + "learning_rate": 5.3675523804325154e-05, + "loss": 0.9916, + "step": 14688 + }, + { + "epoch": 2.6152065527065527, + "grad_norm": 0.9050240516662598, + "learning_rate": 5.366311927307006e-05, + "loss": 0.7734, + "step": 14689 + }, + { + "epoch": 2.6153846153846154, + "grad_norm": 1.000036358833313, + "learning_rate": 5.365071564968989e-05, + "loss": 0.7932, + "step": 14690 + }, + { + "epoch": 2.615562678062678, + "grad_norm": 0.8147441744804382, + "learning_rate": 5.363831293442763e-05, + "loss": 0.8867, + "step": 14691 + }, + { + "epoch": 2.6157407407407405, + "grad_norm": 0.8662015795707703, + "learning_rate": 5.3625911127526375e-05, + "loss": 0.6742, + "step": 14692 + }, + { + "epoch": 2.6159188034188032, + "grad_norm": 0.8576271533966064, + "learning_rate": 5.3613510229229e-05, + "loss": 0.8161, + "step": 14693 + }, + { + "epoch": 2.616096866096866, + "grad_norm": 0.8862481713294983, + "learning_rate": 5.360111023977856e-05, + "loss": 0.8774, + "step": 14694 + }, + { + "epoch": 2.6162749287749287, + "grad_norm": 0.8384450674057007, + "learning_rate": 5.358871115941799e-05, + "loss": 0.9149, + "step": 14695 + }, + { + "epoch": 2.6164529914529915, + "grad_norm": 0.9055412411689758, + "learning_rate": 5.357631298839021e-05, + "loss": 0.8197, + "step": 14696 + }, + { + "epoch": 2.6166310541310542, + "grad_norm": 0.937764585018158, + "learning_rate": 5.356391572693813e-05, + "loss": 1.0392, + "step": 14697 + }, + { + "epoch": 2.616809116809117, + "grad_norm": 0.8917306661605835, + "learning_rate": 5.355151937530463e-05, + "loss": 0.868, + "step": 14698 + }, + { + "epoch": 2.6169871794871797, + "grad_norm": 0.7353024482727051, + "learning_rate": 5.3539123933732705e-05, + "loss": 0.7788, + "step": 14699 + }, + { + "epoch": 2.617165242165242, + "grad_norm": 0.8607454299926758, + "learning_rate": 5.352672940246504e-05, + "loss": 0.7746, + "step": 14700 + }, + { + "epoch": 2.617343304843305, + "grad_norm": 0.9775658249855042, + "learning_rate": 5.3514335781744616e-05, + "loss": 0.9438, + "step": 14701 + }, + { + "epoch": 2.6175213675213675, + "grad_norm": 0.9416237473487854, + "learning_rate": 5.350194307181422e-05, + "loss": 0.9581, + "step": 14702 + }, + { + "epoch": 2.6176994301994303, + "grad_norm": 0.8378105163574219, + "learning_rate": 5.348955127291666e-05, + "loss": 1.0038, + "step": 14703 + }, + { + "epoch": 2.6178774928774926, + "grad_norm": 0.8199161887168884, + "learning_rate": 5.347716038529471e-05, + "loss": 0.9492, + "step": 14704 + }, + { + "epoch": 2.6180555555555554, + "grad_norm": 0.9511042833328247, + "learning_rate": 5.3464770409191176e-05, + "loss": 1.2101, + "step": 14705 + }, + { + "epoch": 2.618233618233618, + "grad_norm": 0.8017105460166931, + "learning_rate": 5.3452381344848754e-05, + "loss": 0.9524, + "step": 14706 + }, + { + "epoch": 2.618411680911681, + "grad_norm": 0.8174898624420166, + "learning_rate": 5.34399931925103e-05, + "loss": 0.911, + "step": 14707 + }, + { + "epoch": 2.6185897435897436, + "grad_norm": 0.8134239315986633, + "learning_rate": 5.342760595241838e-05, + "loss": 0.8971, + "step": 14708 + }, + { + "epoch": 2.6187678062678064, + "grad_norm": 0.817252516746521, + "learning_rate": 5.341521962481586e-05, + "loss": 0.8472, + "step": 14709 + }, + { + "epoch": 2.618945868945869, + "grad_norm": 0.8675270080566406, + "learning_rate": 5.3402834209945264e-05, + "loss": 0.9607, + "step": 14710 + }, + { + "epoch": 2.619123931623932, + "grad_norm": 1.0281410217285156, + "learning_rate": 5.339044970804936e-05, + "loss": 1.0487, + "step": 14711 + }, + { + "epoch": 2.619301994301994, + "grad_norm": 0.9276307225227356, + "learning_rate": 5.33780661193708e-05, + "loss": 0.8915, + "step": 14712 + }, + { + "epoch": 2.619480056980057, + "grad_norm": 0.8479217290878296, + "learning_rate": 5.336568344415216e-05, + "loss": 0.929, + "step": 14713 + }, + { + "epoch": 2.6196581196581197, + "grad_norm": 0.8695724010467529, + "learning_rate": 5.335330168263608e-05, + "loss": 0.8651, + "step": 14714 + }, + { + "epoch": 2.6198361823361824, + "grad_norm": 0.7740936875343323, + "learning_rate": 5.3340920835065155e-05, + "loss": 0.8572, + "step": 14715 + }, + { + "epoch": 2.620014245014245, + "grad_norm": 0.8619815111160278, + "learning_rate": 5.332854090168192e-05, + "loss": 0.6934, + "step": 14716 + }, + { + "epoch": 2.6201923076923075, + "grad_norm": 0.8866271376609802, + "learning_rate": 5.331616188272902e-05, + "loss": 1.038, + "step": 14717 + }, + { + "epoch": 2.6203703703703702, + "grad_norm": 0.7526047825813293, + "learning_rate": 5.330378377844896e-05, + "loss": 0.8534, + "step": 14718 + }, + { + "epoch": 2.620548433048433, + "grad_norm": 0.6914070248603821, + "learning_rate": 5.329140658908423e-05, + "loss": 0.5355, + "step": 14719 + }, + { + "epoch": 2.6207264957264957, + "grad_norm": 0.886074423789978, + "learning_rate": 5.3279030314877374e-05, + "loss": 0.8277, + "step": 14720 + }, + { + "epoch": 2.6209045584045585, + "grad_norm": 0.9101460576057434, + "learning_rate": 5.326665495607082e-05, + "loss": 0.8711, + "step": 14721 + }, + { + "epoch": 2.6210826210826212, + "grad_norm": 0.9744461178779602, + "learning_rate": 5.3254280512907175e-05, + "loss": 1.2376, + "step": 14722 + }, + { + "epoch": 2.621260683760684, + "grad_norm": 1.013480544090271, + "learning_rate": 5.32419069856287e-05, + "loss": 0.8946, + "step": 14723 + }, + { + "epoch": 2.6214387464387463, + "grad_norm": 0.82442706823349, + "learning_rate": 5.3229534374478005e-05, + "loss": 0.732, + "step": 14724 + }, + { + "epoch": 2.621616809116809, + "grad_norm": 0.7960239052772522, + "learning_rate": 5.3217162679697366e-05, + "loss": 0.7633, + "step": 14725 + }, + { + "epoch": 2.621794871794872, + "grad_norm": 0.819844126701355, + "learning_rate": 5.320479190152926e-05, + "loss": 0.974, + "step": 14726 + }, + { + "epoch": 2.6219729344729346, + "grad_norm": 0.8245221376419067, + "learning_rate": 5.319242204021606e-05, + "loss": 0.9122, + "step": 14727 + }, + { + "epoch": 2.6221509971509973, + "grad_norm": 0.7574561834335327, + "learning_rate": 5.318005309600011e-05, + "loss": 0.8427, + "step": 14728 + }, + { + "epoch": 2.6223290598290596, + "grad_norm": 1.0385704040527344, + "learning_rate": 5.316768506912377e-05, + "loss": 0.8214, + "step": 14729 + }, + { + "epoch": 2.6225071225071224, + "grad_norm": 0.8616722822189331, + "learning_rate": 5.3155317959829346e-05, + "loss": 0.8469, + "step": 14730 + }, + { + "epoch": 2.622685185185185, + "grad_norm": 0.909667432308197, + "learning_rate": 5.314295176835912e-05, + "loss": 0.9156, + "step": 14731 + }, + { + "epoch": 2.622863247863248, + "grad_norm": 0.9016293883323669, + "learning_rate": 5.3130586494955494e-05, + "loss": 0.9183, + "step": 14732 + }, + { + "epoch": 2.6230413105413106, + "grad_norm": 0.8828284740447998, + "learning_rate": 5.311822213986057e-05, + "loss": 0.8338, + "step": 14733 + }, + { + "epoch": 2.6232193732193734, + "grad_norm": 0.8159047365188599, + "learning_rate": 5.3105858703316794e-05, + "loss": 0.7055, + "step": 14734 + }, + { + "epoch": 2.623397435897436, + "grad_norm": 0.9240905046463013, + "learning_rate": 5.309349618556623e-05, + "loss": 0.9078, + "step": 14735 + }, + { + "epoch": 2.6235754985754984, + "grad_norm": 0.8881595134735107, + "learning_rate": 5.308113458685118e-05, + "loss": 0.9946, + "step": 14736 + }, + { + "epoch": 2.623753561253561, + "grad_norm": 0.8781841397285461, + "learning_rate": 5.306877390741385e-05, + "loss": 0.8252, + "step": 14737 + }, + { + "epoch": 2.623931623931624, + "grad_norm": 0.8348106741905212, + "learning_rate": 5.3056414147496355e-05, + "loss": 0.8653, + "step": 14738 + }, + { + "epoch": 2.6241096866096867, + "grad_norm": 0.9692304134368896, + "learning_rate": 5.3044055307341e-05, + "loss": 0.7814, + "step": 14739 + }, + { + "epoch": 2.6242877492877494, + "grad_norm": 0.866179347038269, + "learning_rate": 5.303169738718976e-05, + "loss": 0.9255, + "step": 14740 + }, + { + "epoch": 2.6244658119658117, + "grad_norm": 0.9306690692901611, + "learning_rate": 5.301934038728487e-05, + "loss": 0.9123, + "step": 14741 + }, + { + "epoch": 2.6246438746438745, + "grad_norm": 0.949357271194458, + "learning_rate": 5.3006984307868415e-05, + "loss": 0.8452, + "step": 14742 + }, + { + "epoch": 2.6248219373219372, + "grad_norm": 0.8638128042221069, + "learning_rate": 5.299462914918249e-05, + "loss": 0.8026, + "step": 14743 + }, + { + "epoch": 2.625, + "grad_norm": 0.9075117707252502, + "learning_rate": 5.2982274911469154e-05, + "loss": 1.0644, + "step": 14744 + }, + { + "epoch": 2.6251780626780628, + "grad_norm": 0.8146225810050964, + "learning_rate": 5.296992159497047e-05, + "loss": 0.8494, + "step": 14745 + }, + { + "epoch": 2.6253561253561255, + "grad_norm": 0.8887025713920593, + "learning_rate": 5.295756919992847e-05, + "loss": 0.8143, + "step": 14746 + }, + { + "epoch": 2.6255341880341883, + "grad_norm": 0.8262654542922974, + "learning_rate": 5.29452177265852e-05, + "loss": 0.7559, + "step": 14747 + }, + { + "epoch": 2.6257122507122506, + "grad_norm": 0.8126912117004395, + "learning_rate": 5.2932867175182574e-05, + "loss": 0.8528, + "step": 14748 + }, + { + "epoch": 2.6258903133903133, + "grad_norm": 0.8970595598220825, + "learning_rate": 5.2920517545962746e-05, + "loss": 0.8584, + "step": 14749 + }, + { + "epoch": 2.626068376068376, + "grad_norm": 0.8678651452064514, + "learning_rate": 5.290816883916748e-05, + "loss": 0.8686, + "step": 14750 + }, + { + "epoch": 2.626246438746439, + "grad_norm": 0.8069576621055603, + "learning_rate": 5.289582105503887e-05, + "loss": 0.868, + "step": 14751 + }, + { + "epoch": 2.6264245014245016, + "grad_norm": 1.0322144031524658, + "learning_rate": 5.28834741938188e-05, + "loss": 1.1537, + "step": 14752 + }, + { + "epoch": 2.626602564102564, + "grad_norm": 0.8274349570274353, + "learning_rate": 5.287112825574917e-05, + "loss": 1.0126, + "step": 14753 + }, + { + "epoch": 2.6267806267806266, + "grad_norm": 0.8820709586143494, + "learning_rate": 5.2858783241071875e-05, + "loss": 0.893, + "step": 14754 + }, + { + "epoch": 2.6269586894586894, + "grad_norm": 1.0102146863937378, + "learning_rate": 5.28464391500288e-05, + "loss": 0.8524, + "step": 14755 + }, + { + "epoch": 2.627136752136752, + "grad_norm": 0.875468373298645, + "learning_rate": 5.2834095982861764e-05, + "loss": 1.0991, + "step": 14756 + }, + { + "epoch": 2.627314814814815, + "grad_norm": 0.8155242800712585, + "learning_rate": 5.282175373981267e-05, + "loss": 0.666, + "step": 14757 + }, + { + "epoch": 2.6274928774928776, + "grad_norm": 0.8777057528495789, + "learning_rate": 5.280941242112332e-05, + "loss": 0.892, + "step": 14758 + }, + { + "epoch": 2.6276709401709404, + "grad_norm": 0.8357667922973633, + "learning_rate": 5.279707202703549e-05, + "loss": 0.8118, + "step": 14759 + }, + { + "epoch": 2.6278490028490027, + "grad_norm": 0.7862337827682495, + "learning_rate": 5.278473255779097e-05, + "loss": 0.7287, + "step": 14760 + }, + { + "epoch": 2.6280270655270654, + "grad_norm": 0.8340336084365845, + "learning_rate": 5.277239401363155e-05, + "loss": 0.7697, + "step": 14761 + }, + { + "epoch": 2.628205128205128, + "grad_norm": 0.7986457943916321, + "learning_rate": 5.276005639479896e-05, + "loss": 0.9358, + "step": 14762 + }, + { + "epoch": 2.628383190883191, + "grad_norm": 0.7377769947052002, + "learning_rate": 5.2747719701534895e-05, + "loss": 0.8091, + "step": 14763 + }, + { + "epoch": 2.6285612535612537, + "grad_norm": 0.9749723672866821, + "learning_rate": 5.273538393408117e-05, + "loss": 0.8163, + "step": 14764 + }, + { + "epoch": 2.628739316239316, + "grad_norm": 0.8718321323394775, + "learning_rate": 5.2723049092679354e-05, + "loss": 1.1587, + "step": 14765 + }, + { + "epoch": 2.6289173789173788, + "grad_norm": 0.9394767880439758, + "learning_rate": 5.27107151775712e-05, + "loss": 0.9409, + "step": 14766 + }, + { + "epoch": 2.6290954415954415, + "grad_norm": 0.9763813614845276, + "learning_rate": 5.269838218899836e-05, + "loss": 1.0171, + "step": 14767 + }, + { + "epoch": 2.6292735042735043, + "grad_norm": 0.878968358039856, + "learning_rate": 5.268605012720247e-05, + "loss": 0.9117, + "step": 14768 + }, + { + "epoch": 2.629451566951567, + "grad_norm": 0.8240547776222229, + "learning_rate": 5.267371899242512e-05, + "loss": 0.9351, + "step": 14769 + }, + { + "epoch": 2.6296296296296298, + "grad_norm": 0.8048275709152222, + "learning_rate": 5.266138878490795e-05, + "loss": 0.9331, + "step": 14770 + }, + { + "epoch": 2.6298076923076925, + "grad_norm": 0.7176041007041931, + "learning_rate": 5.264905950489252e-05, + "loss": 0.6424, + "step": 14771 + }, + { + "epoch": 2.629985754985755, + "grad_norm": 0.973258912563324, + "learning_rate": 5.263673115262041e-05, + "loss": 0.9295, + "step": 14772 + }, + { + "epoch": 2.6301638176638176, + "grad_norm": 0.8955824971199036, + "learning_rate": 5.262440372833313e-05, + "loss": 0.9306, + "step": 14773 + }, + { + "epoch": 2.6303418803418803, + "grad_norm": 0.8430632948875427, + "learning_rate": 5.2612077232272305e-05, + "loss": 0.9343, + "step": 14774 + }, + { + "epoch": 2.630519943019943, + "grad_norm": 1.0231794118881226, + "learning_rate": 5.2599751664679334e-05, + "loss": 0.941, + "step": 14775 + }, + { + "epoch": 2.630698005698006, + "grad_norm": 0.9726024866104126, + "learning_rate": 5.258742702579579e-05, + "loss": 1.1726, + "step": 14776 + }, + { + "epoch": 2.630876068376068, + "grad_norm": 0.8575723171234131, + "learning_rate": 5.257510331586312e-05, + "loss": 0.5644, + "step": 14777 + }, + { + "epoch": 2.631054131054131, + "grad_norm": 0.853165864944458, + "learning_rate": 5.2562780535122744e-05, + "loss": 0.8555, + "step": 14778 + }, + { + "epoch": 2.6312321937321936, + "grad_norm": 0.861574649810791, + "learning_rate": 5.255045868381623e-05, + "loss": 0.8298, + "step": 14779 + }, + { + "epoch": 2.6314102564102564, + "grad_norm": 0.8744526505470276, + "learning_rate": 5.2538137762184816e-05, + "loss": 0.9889, + "step": 14780 + }, + { + "epoch": 2.631588319088319, + "grad_norm": 0.7891412973403931, + "learning_rate": 5.2525817770470084e-05, + "loss": 0.9765, + "step": 14781 + }, + { + "epoch": 2.631766381766382, + "grad_norm": 0.9155156016349792, + "learning_rate": 5.251349870891327e-05, + "loss": 0.8927, + "step": 14782 + }, + { + "epoch": 2.6319444444444446, + "grad_norm": 0.8547508120536804, + "learning_rate": 5.250118057775582e-05, + "loss": 0.8479, + "step": 14783 + }, + { + "epoch": 2.6321225071225074, + "grad_norm": 0.7606263756752014, + "learning_rate": 5.248886337723908e-05, + "loss": 0.7557, + "step": 14784 + }, + { + "epoch": 2.6323005698005697, + "grad_norm": 0.855315625667572, + "learning_rate": 5.247654710760437e-05, + "loss": 0.8527, + "step": 14785 + }, + { + "epoch": 2.6324786324786325, + "grad_norm": 0.7656288743019104, + "learning_rate": 5.246423176909298e-05, + "loss": 0.8881, + "step": 14786 + }, + { + "epoch": 2.632656695156695, + "grad_norm": 0.817034125328064, + "learning_rate": 5.2451917361946236e-05, + "loss": 1.042, + "step": 14787 + }, + { + "epoch": 2.632834757834758, + "grad_norm": 0.8473303318023682, + "learning_rate": 5.2439603886405356e-05, + "loss": 0.8804, + "step": 14788 + }, + { + "epoch": 2.6330128205128203, + "grad_norm": 0.9563126564025879, + "learning_rate": 5.242729134271171e-05, + "loss": 0.8463, + "step": 14789 + }, + { + "epoch": 2.633190883190883, + "grad_norm": 0.8297066688537598, + "learning_rate": 5.241497973110641e-05, + "loss": 0.7776, + "step": 14790 + }, + { + "epoch": 2.6333689458689458, + "grad_norm": 0.8433563709259033, + "learning_rate": 5.240266905183075e-05, + "loss": 0.8712, + "step": 14791 + }, + { + "epoch": 2.6335470085470085, + "grad_norm": 0.814725935459137, + "learning_rate": 5.239035930512593e-05, + "loss": 0.9819, + "step": 14792 + }, + { + "epoch": 2.6337250712250713, + "grad_norm": 0.844292163848877, + "learning_rate": 5.23780504912331e-05, + "loss": 0.8693, + "step": 14793 + }, + { + "epoch": 2.633903133903134, + "grad_norm": 0.8194862008094788, + "learning_rate": 5.2365742610393464e-05, + "loss": 0.7878, + "step": 14794 + }, + { + "epoch": 2.6340811965811968, + "grad_norm": 0.8570502400398254, + "learning_rate": 5.2353435662848135e-05, + "loss": 0.815, + "step": 14795 + }, + { + "epoch": 2.6342592592592595, + "grad_norm": 0.9301772713661194, + "learning_rate": 5.2341129648838275e-05, + "loss": 0.9092, + "step": 14796 + }, + { + "epoch": 2.634437321937322, + "grad_norm": 0.7605858445167542, + "learning_rate": 5.232882456860493e-05, + "loss": 0.8753, + "step": 14797 + }, + { + "epoch": 2.6346153846153846, + "grad_norm": 0.8265452980995178, + "learning_rate": 5.231652042238927e-05, + "loss": 0.9134, + "step": 14798 + }, + { + "epoch": 2.6347934472934473, + "grad_norm": 0.7440468072891235, + "learning_rate": 5.230421721043235e-05, + "loss": 0.7471, + "step": 14799 + }, + { + "epoch": 2.63497150997151, + "grad_norm": 0.9172230958938599, + "learning_rate": 5.2291914932975205e-05, + "loss": 1.0155, + "step": 14800 + }, + { + "epoch": 2.6351495726495724, + "grad_norm": 0.8364499807357788, + "learning_rate": 5.227961359025888e-05, + "loss": 0.9561, + "step": 14801 + }, + { + "epoch": 2.635327635327635, + "grad_norm": 0.7756382822990417, + "learning_rate": 5.22673131825244e-05, + "loss": 0.6893, + "step": 14802 + }, + { + "epoch": 2.635505698005698, + "grad_norm": 0.9042136669158936, + "learning_rate": 5.225501371001273e-05, + "loss": 0.7613, + "step": 14803 + }, + { + "epoch": 2.6356837606837606, + "grad_norm": 0.8989379405975342, + "learning_rate": 5.224271517296495e-05, + "loss": 0.8092, + "step": 14804 + }, + { + "epoch": 2.6358618233618234, + "grad_norm": 0.7999827265739441, + "learning_rate": 5.2230417571621906e-05, + "loss": 0.8115, + "step": 14805 + }, + { + "epoch": 2.636039886039886, + "grad_norm": 0.9071131348609924, + "learning_rate": 5.221812090622464e-05, + "loss": 0.9072, + "step": 14806 + }, + { + "epoch": 2.636217948717949, + "grad_norm": 0.7227704524993896, + "learning_rate": 5.220582517701398e-05, + "loss": 0.7598, + "step": 14807 + }, + { + "epoch": 2.6363960113960117, + "grad_norm": 0.8520537614822388, + "learning_rate": 5.219353038423094e-05, + "loss": 1.1072, + "step": 14808 + }, + { + "epoch": 2.636574074074074, + "grad_norm": 0.8690574765205383, + "learning_rate": 5.218123652811634e-05, + "loss": 0.773, + "step": 14809 + }, + { + "epoch": 2.6367521367521367, + "grad_norm": 0.7897602319717407, + "learning_rate": 5.216894360891109e-05, + "loss": 0.792, + "step": 14810 + }, + { + "epoch": 2.6369301994301995, + "grad_norm": 0.8746532201766968, + "learning_rate": 5.215665162685601e-05, + "loss": 0.8853, + "step": 14811 + }, + { + "epoch": 2.637108262108262, + "grad_norm": 0.8525128364562988, + "learning_rate": 5.214436058219199e-05, + "loss": 0.7293, + "step": 14812 + }, + { + "epoch": 2.6372863247863245, + "grad_norm": 0.979969322681427, + "learning_rate": 5.213207047515975e-05, + "loss": 0.8485, + "step": 14813 + }, + { + "epoch": 2.6374643874643873, + "grad_norm": 0.8439529538154602, + "learning_rate": 5.211978130600024e-05, + "loss": 0.7492, + "step": 14814 + }, + { + "epoch": 2.63764245014245, + "grad_norm": 0.8356610536575317, + "learning_rate": 5.2107493074954064e-05, + "loss": 0.8255, + "step": 14815 + }, + { + "epoch": 2.6378205128205128, + "grad_norm": 0.7857736349105835, + "learning_rate": 5.2095205782262116e-05, + "loss": 0.766, + "step": 14816 + }, + { + "epoch": 2.6379985754985755, + "grad_norm": 0.919058084487915, + "learning_rate": 5.20829194281651e-05, + "loss": 1.0661, + "step": 14817 + }, + { + "epoch": 2.6381766381766383, + "grad_norm": 0.8793047070503235, + "learning_rate": 5.207063401290373e-05, + "loss": 0.8297, + "step": 14818 + }, + { + "epoch": 2.638354700854701, + "grad_norm": 0.7848390340805054, + "learning_rate": 5.205834953671873e-05, + "loss": 0.8051, + "step": 14819 + }, + { + "epoch": 2.638532763532764, + "grad_norm": 0.8391907215118408, + "learning_rate": 5.2046065999850736e-05, + "loss": 0.8444, + "step": 14820 + }, + { + "epoch": 2.638710826210826, + "grad_norm": 0.8137226700782776, + "learning_rate": 5.2033783402540546e-05, + "loss": 0.7908, + "step": 14821 + }, + { + "epoch": 2.638888888888889, + "grad_norm": 0.8440108299255371, + "learning_rate": 5.2021501745028645e-05, + "loss": 0.7985, + "step": 14822 + }, + { + "epoch": 2.6390669515669516, + "grad_norm": 0.7432600855827332, + "learning_rate": 5.200922102755581e-05, + "loss": 0.7816, + "step": 14823 + }, + { + "epoch": 2.6392450142450143, + "grad_norm": 0.9003379344940186, + "learning_rate": 5.199694125036257e-05, + "loss": 0.9171, + "step": 14824 + }, + { + "epoch": 2.6394230769230766, + "grad_norm": 0.8994988203048706, + "learning_rate": 5.198466241368957e-05, + "loss": 0.8333, + "step": 14825 + }, + { + "epoch": 2.6396011396011394, + "grad_norm": 0.9042859077453613, + "learning_rate": 5.197238451777735e-05, + "loss": 0.7491, + "step": 14826 + }, + { + "epoch": 2.639779202279202, + "grad_norm": 0.8024145364761353, + "learning_rate": 5.196010756286649e-05, + "loss": 0.882, + "step": 14827 + }, + { + "epoch": 2.639957264957265, + "grad_norm": 0.73011714220047, + "learning_rate": 5.1947831549197504e-05, + "loss": 0.865, + "step": 14828 + }, + { + "epoch": 2.6401353276353277, + "grad_norm": 0.845160186290741, + "learning_rate": 5.1935556477011006e-05, + "loss": 0.8912, + "step": 14829 + }, + { + "epoch": 2.6403133903133904, + "grad_norm": 0.8264908194541931, + "learning_rate": 5.192328234654735e-05, + "loss": 0.8597, + "step": 14830 + }, + { + "epoch": 2.640491452991453, + "grad_norm": 0.8400609493255615, + "learning_rate": 5.191100915804718e-05, + "loss": 0.5906, + "step": 14831 + }, + { + "epoch": 2.640669515669516, + "grad_norm": 0.8633815050125122, + "learning_rate": 5.189873691175082e-05, + "loss": 1.031, + "step": 14832 + }, + { + "epoch": 2.640847578347578, + "grad_norm": 0.9047896862030029, + "learning_rate": 5.188646560789884e-05, + "loss": 0.6929, + "step": 14833 + }, + { + "epoch": 2.641025641025641, + "grad_norm": 0.8293144106864929, + "learning_rate": 5.18741952467316e-05, + "loss": 0.9608, + "step": 14834 + }, + { + "epoch": 2.6412037037037037, + "grad_norm": 0.7980968356132507, + "learning_rate": 5.186192582848955e-05, + "loss": 0.6021, + "step": 14835 + }, + { + "epoch": 2.6413817663817665, + "grad_norm": 0.7945372462272644, + "learning_rate": 5.184965735341305e-05, + "loss": 0.7069, + "step": 14836 + }, + { + "epoch": 2.6415598290598292, + "grad_norm": 0.8388827443122864, + "learning_rate": 5.183738982174246e-05, + "loss": 1.2404, + "step": 14837 + }, + { + "epoch": 2.6417378917378915, + "grad_norm": 0.8332177400588989, + "learning_rate": 5.18251232337182e-05, + "loss": 0.9353, + "step": 14838 + }, + { + "epoch": 2.6419159544159543, + "grad_norm": 0.9658130407333374, + "learning_rate": 5.1812857589580565e-05, + "loss": 0.8, + "step": 14839 + }, + { + "epoch": 2.642094017094017, + "grad_norm": 0.9074252247810364, + "learning_rate": 5.180059288956991e-05, + "loss": 0.7567, + "step": 14840 + }, + { + "epoch": 2.64227207977208, + "grad_norm": 0.8543582558631897, + "learning_rate": 5.178832913392649e-05, + "loss": 0.9754, + "step": 14841 + }, + { + "epoch": 2.6424501424501425, + "grad_norm": 0.8235877156257629, + "learning_rate": 5.177606632289063e-05, + "loss": 0.825, + "step": 14842 + }, + { + "epoch": 2.6426282051282053, + "grad_norm": 0.8550012111663818, + "learning_rate": 5.1763804456702545e-05, + "loss": 1.0286, + "step": 14843 + }, + { + "epoch": 2.642806267806268, + "grad_norm": 0.8879600763320923, + "learning_rate": 5.175154353560254e-05, + "loss": 0.8935, + "step": 14844 + }, + { + "epoch": 2.6429843304843303, + "grad_norm": 0.8822683095932007, + "learning_rate": 5.1739283559830754e-05, + "loss": 0.8659, + "step": 14845 + }, + { + "epoch": 2.643162393162393, + "grad_norm": 1.0260087251663208, + "learning_rate": 5.1727024529627544e-05, + "loss": 0.8952, + "step": 14846 + }, + { + "epoch": 2.643340455840456, + "grad_norm": 0.8105470538139343, + "learning_rate": 5.171476644523292e-05, + "loss": 0.7987, + "step": 14847 + }, + { + "epoch": 2.6435185185185186, + "grad_norm": 0.8861166834831238, + "learning_rate": 5.170250930688719e-05, + "loss": 0.8476, + "step": 14848 + }, + { + "epoch": 2.6436965811965814, + "grad_norm": 0.8035899996757507, + "learning_rate": 5.169025311483047e-05, + "loss": 0.7366, + "step": 14849 + }, + { + "epoch": 2.6438746438746437, + "grad_norm": 0.8359752297401428, + "learning_rate": 5.1677997869302874e-05, + "loss": 0.8931, + "step": 14850 + }, + { + "epoch": 2.6440527065527064, + "grad_norm": 0.8483668565750122, + "learning_rate": 5.166574357054452e-05, + "loss": 0.7662, + "step": 14851 + }, + { + "epoch": 2.644230769230769, + "grad_norm": 0.9865937829017639, + "learning_rate": 5.165349021879553e-05, + "loss": 1.101, + "step": 14852 + }, + { + "epoch": 2.644408831908832, + "grad_norm": 0.8491073250770569, + "learning_rate": 5.164123781429596e-05, + "loss": 0.9576, + "step": 14853 + }, + { + "epoch": 2.6445868945868947, + "grad_norm": 0.8185597062110901, + "learning_rate": 5.162898635728588e-05, + "loss": 0.6353, + "step": 14854 + }, + { + "epoch": 2.6447649572649574, + "grad_norm": 0.8583887815475464, + "learning_rate": 5.1616735848005306e-05, + "loss": 0.8715, + "step": 14855 + }, + { + "epoch": 2.64494301994302, + "grad_norm": 0.8107531666755676, + "learning_rate": 5.16044862866943e-05, + "loss": 0.7111, + "step": 14856 + }, + { + "epoch": 2.6451210826210825, + "grad_norm": 0.7675925493240356, + "learning_rate": 5.1592237673592867e-05, + "loss": 0.8145, + "step": 14857 + }, + { + "epoch": 2.6452991452991452, + "grad_norm": 0.9418326020240784, + "learning_rate": 5.157999000894098e-05, + "loss": 0.8454, + "step": 14858 + }, + { + "epoch": 2.645477207977208, + "grad_norm": 0.8420053720474243, + "learning_rate": 5.15677432929786e-05, + "loss": 0.7343, + "step": 14859 + }, + { + "epoch": 2.6456552706552707, + "grad_norm": 0.9815202951431274, + "learning_rate": 5.155549752594564e-05, + "loss": 0.9252, + "step": 14860 + }, + { + "epoch": 2.6458333333333335, + "grad_norm": 0.8282185792922974, + "learning_rate": 5.1543252708082146e-05, + "loss": 0.9935, + "step": 14861 + }, + { + "epoch": 2.646011396011396, + "grad_norm": 0.7398781180381775, + "learning_rate": 5.153100883962788e-05, + "loss": 0.5024, + "step": 14862 + }, + { + "epoch": 2.6461894586894585, + "grad_norm": 1.0273998975753784, + "learning_rate": 5.1518765920822856e-05, + "loss": 0.9023, + "step": 14863 + }, + { + "epoch": 2.6463675213675213, + "grad_norm": 0.8017948269844055, + "learning_rate": 5.150652395190689e-05, + "loss": 0.6755, + "step": 14864 + }, + { + "epoch": 2.646545584045584, + "grad_norm": 0.7470258474349976, + "learning_rate": 5.1494282933119864e-05, + "loss": 0.5408, + "step": 14865 + }, + { + "epoch": 2.646723646723647, + "grad_norm": 0.8118627071380615, + "learning_rate": 5.1482042864701595e-05, + "loss": 0.8032, + "step": 14866 + }, + { + "epoch": 2.6469017094017095, + "grad_norm": 0.8302956223487854, + "learning_rate": 5.146980374689192e-05, + "loss": 0.7428, + "step": 14867 + }, + { + "epoch": 2.6470797720797723, + "grad_norm": 0.8660209774971008, + "learning_rate": 5.145756557993061e-05, + "loss": 0.8284, + "step": 14868 + }, + { + "epoch": 2.6472578347578346, + "grad_norm": 1.0153858661651611, + "learning_rate": 5.1445328364057475e-05, + "loss": 0.9766, + "step": 14869 + }, + { + "epoch": 2.6474358974358974, + "grad_norm": 0.9047706127166748, + "learning_rate": 5.143309209951223e-05, + "loss": 1.0099, + "step": 14870 + }, + { + "epoch": 2.64761396011396, + "grad_norm": 0.7924295663833618, + "learning_rate": 5.1420856786534724e-05, + "loss": 0.8385, + "step": 14871 + }, + { + "epoch": 2.647792022792023, + "grad_norm": 0.8885742425918579, + "learning_rate": 5.140862242536455e-05, + "loss": 1.0259, + "step": 14872 + }, + { + "epoch": 2.6479700854700856, + "grad_norm": 0.8826889991760254, + "learning_rate": 5.139638901624151e-05, + "loss": 1.0755, + "step": 14873 + }, + { + "epoch": 2.648148148148148, + "grad_norm": 0.7793754935264587, + "learning_rate": 5.138415655940526e-05, + "loss": 0.8289, + "step": 14874 + }, + { + "epoch": 2.6483262108262107, + "grad_norm": 0.8587870597839355, + "learning_rate": 5.137192505509547e-05, + "loss": 0.9185, + "step": 14875 + }, + { + "epoch": 2.6485042735042734, + "grad_norm": 0.8799259066581726, + "learning_rate": 5.13596945035518e-05, + "loss": 0.9493, + "step": 14876 + }, + { + "epoch": 2.648682336182336, + "grad_norm": 0.8108882308006287, + "learning_rate": 5.1347464905013834e-05, + "loss": 0.8892, + "step": 14877 + }, + { + "epoch": 2.648860398860399, + "grad_norm": 0.8387644290924072, + "learning_rate": 5.1335236259721296e-05, + "loss": 0.8723, + "step": 14878 + }, + { + "epoch": 2.6490384615384617, + "grad_norm": 0.8750926852226257, + "learning_rate": 5.1323008567913655e-05, + "loss": 0.6978, + "step": 14879 + }, + { + "epoch": 2.6492165242165244, + "grad_norm": 0.7837518453598022, + "learning_rate": 5.131078182983055e-05, + "loss": 0.8747, + "step": 14880 + }, + { + "epoch": 2.6493945868945867, + "grad_norm": 0.8998439311981201, + "learning_rate": 5.1298556045711566e-05, + "loss": 0.9903, + "step": 14881 + }, + { + "epoch": 2.6495726495726495, + "grad_norm": 0.8012915253639221, + "learning_rate": 5.128633121579619e-05, + "loss": 0.9767, + "step": 14882 + }, + { + "epoch": 2.6497507122507122, + "grad_norm": 0.9051218032836914, + "learning_rate": 5.1274107340323964e-05, + "loss": 0.7454, + "step": 14883 + }, + { + "epoch": 2.649928774928775, + "grad_norm": 0.8373401761054993, + "learning_rate": 5.1261884419534376e-05, + "loss": 0.821, + "step": 14884 + }, + { + "epoch": 2.6501068376068377, + "grad_norm": 0.7482876181602478, + "learning_rate": 5.124966245366689e-05, + "loss": 0.7051, + "step": 14885 + }, + { + "epoch": 2.6502849002849, + "grad_norm": 0.8445764183998108, + "learning_rate": 5.1237441442961074e-05, + "loss": 0.7416, + "step": 14886 + }, + { + "epoch": 2.650462962962963, + "grad_norm": 0.887598991394043, + "learning_rate": 5.122522138765622e-05, + "loss": 0.9027, + "step": 14887 + }, + { + "epoch": 2.6506410256410255, + "grad_norm": 0.8089238405227661, + "learning_rate": 5.1213002287991905e-05, + "loss": 0.9294, + "step": 14888 + }, + { + "epoch": 2.6508190883190883, + "grad_norm": 0.8614209890365601, + "learning_rate": 5.120078414420739e-05, + "loss": 0.7716, + "step": 14889 + }, + { + "epoch": 2.650997150997151, + "grad_norm": 0.6805269718170166, + "learning_rate": 5.118856695654217e-05, + "loss": 0.6183, + "step": 14890 + }, + { + "epoch": 2.651175213675214, + "grad_norm": 0.9024596214294434, + "learning_rate": 5.117635072523559e-05, + "loss": 0.9516, + "step": 14891 + }, + { + "epoch": 2.6513532763532766, + "grad_norm": 0.906373143196106, + "learning_rate": 5.116413545052701e-05, + "loss": 0.7522, + "step": 14892 + }, + { + "epoch": 2.6515313390313393, + "grad_norm": 0.827235996723175, + "learning_rate": 5.1151921132655725e-05, + "loss": 0.6776, + "step": 14893 + }, + { + "epoch": 2.6517094017094016, + "grad_norm": 0.7769291996955872, + "learning_rate": 5.113970777186108e-05, + "loss": 0.6682, + "step": 14894 + }, + { + "epoch": 2.6518874643874644, + "grad_norm": 0.8420324921607971, + "learning_rate": 5.112749536838233e-05, + "loss": 0.8303, + "step": 14895 + }, + { + "epoch": 2.652065527065527, + "grad_norm": 0.789368748664856, + "learning_rate": 5.1115283922458814e-05, + "loss": 0.773, + "step": 14896 + }, + { + "epoch": 2.65224358974359, + "grad_norm": 0.9156190752983093, + "learning_rate": 5.1103073434329766e-05, + "loss": 1.0318, + "step": 14897 + }, + { + "epoch": 2.652421652421652, + "grad_norm": 1.0411027669906616, + "learning_rate": 5.109086390423441e-05, + "loss": 0.81, + "step": 14898 + }, + { + "epoch": 2.652599715099715, + "grad_norm": 0.9908538460731506, + "learning_rate": 5.107865533241198e-05, + "loss": 0.9386, + "step": 14899 + }, + { + "epoch": 2.6527777777777777, + "grad_norm": 0.7364035844802856, + "learning_rate": 5.106644771910165e-05, + "loss": 0.675, + "step": 14900 + }, + { + "epoch": 2.6529558404558404, + "grad_norm": 0.8409245014190674, + "learning_rate": 5.1054241064542686e-05, + "loss": 0.9446, + "step": 14901 + }, + { + "epoch": 2.653133903133903, + "grad_norm": 0.7731066942214966, + "learning_rate": 5.104203536897412e-05, + "loss": 0.4684, + "step": 14902 + }, + { + "epoch": 2.653311965811966, + "grad_norm": 0.9114529490470886, + "learning_rate": 5.102983063263525e-05, + "loss": 0.9551, + "step": 14903 + }, + { + "epoch": 2.6534900284900287, + "grad_norm": 0.7949321269989014, + "learning_rate": 5.101762685576503e-05, + "loss": 0.989, + "step": 14904 + }, + { + "epoch": 2.6536680911680914, + "grad_norm": 0.940191924571991, + "learning_rate": 5.1005424038602724e-05, + "loss": 1.0377, + "step": 14905 + }, + { + "epoch": 2.6538461538461537, + "grad_norm": 0.7629654407501221, + "learning_rate": 5.0993222181387334e-05, + "loss": 0.7908, + "step": 14906 + }, + { + "epoch": 2.6540242165242165, + "grad_norm": 0.9712302684783936, + "learning_rate": 5.098102128435797e-05, + "loss": 1.1486, + "step": 14907 + }, + { + "epoch": 2.6542022792022792, + "grad_norm": 0.9054526686668396, + "learning_rate": 5.096882134775365e-05, + "loss": 0.8078, + "step": 14908 + }, + { + "epoch": 2.654380341880342, + "grad_norm": 0.824647068977356, + "learning_rate": 5.095662237181343e-05, + "loss": 0.9095, + "step": 14909 + }, + { + "epoch": 2.6545584045584043, + "grad_norm": 0.8760488033294678, + "learning_rate": 5.0944424356776287e-05, + "loss": 0.8538, + "step": 14910 + }, + { + "epoch": 2.654736467236467, + "grad_norm": 0.8012890219688416, + "learning_rate": 5.093222730288131e-05, + "loss": 0.7972, + "step": 14911 + }, + { + "epoch": 2.65491452991453, + "grad_norm": 0.9025147557258606, + "learning_rate": 5.0920031210367326e-05, + "loss": 0.8485, + "step": 14912 + }, + { + "epoch": 2.6550925925925926, + "grad_norm": 0.8621100783348083, + "learning_rate": 5.090783607947347e-05, + "loss": 1.1856, + "step": 14913 + }, + { + "epoch": 2.6552706552706553, + "grad_norm": 0.7914317846298218, + "learning_rate": 5.08956419104385e-05, + "loss": 0.78, + "step": 14914 + }, + { + "epoch": 2.655448717948718, + "grad_norm": 0.8691070675849915, + "learning_rate": 5.088344870350146e-05, + "loss": 0.8406, + "step": 14915 + }, + { + "epoch": 2.655626780626781, + "grad_norm": 0.8521141409873962, + "learning_rate": 5.087125645890121e-05, + "loss": 1.0077, + "step": 14916 + }, + { + "epoch": 2.6558048433048436, + "grad_norm": 0.7918437123298645, + "learning_rate": 5.08590651768766e-05, + "loss": 0.8367, + "step": 14917 + }, + { + "epoch": 2.655982905982906, + "grad_norm": 0.8580697178840637, + "learning_rate": 5.084687485766659e-05, + "loss": 0.921, + "step": 14918 + }, + { + "epoch": 2.6561609686609686, + "grad_norm": 0.7943900227546692, + "learning_rate": 5.0834685501509894e-05, + "loss": 0.7934, + "step": 14919 + }, + { + "epoch": 2.6563390313390314, + "grad_norm": 0.7467655539512634, + "learning_rate": 5.082249710864544e-05, + "loss": 0.8625, + "step": 14920 + }, + { + "epoch": 2.656517094017094, + "grad_norm": 0.7654036283493042, + "learning_rate": 5.0810309679311996e-05, + "loss": 0.888, + "step": 14921 + }, + { + "epoch": 2.6566951566951564, + "grad_norm": 0.8428319692611694, + "learning_rate": 5.079812321374836e-05, + "loss": 0.858, + "step": 14922 + }, + { + "epoch": 2.656873219373219, + "grad_norm": 0.8273693323135376, + "learning_rate": 5.078593771219329e-05, + "loss": 0.8982, + "step": 14923 + }, + { + "epoch": 2.657051282051282, + "grad_norm": 0.9037185311317444, + "learning_rate": 5.077375317488553e-05, + "loss": 0.7022, + "step": 14924 + }, + { + "epoch": 2.6572293447293447, + "grad_norm": 0.916585385799408, + "learning_rate": 5.0761569602063816e-05, + "loss": 0.8058, + "step": 14925 + }, + { + "epoch": 2.6574074074074074, + "grad_norm": 0.8697561621665955, + "learning_rate": 5.074938699396687e-05, + "loss": 0.8142, + "step": 14926 + }, + { + "epoch": 2.65758547008547, + "grad_norm": 1.024512529373169, + "learning_rate": 5.073720535083334e-05, + "loss": 0.7462, + "step": 14927 + }, + { + "epoch": 2.657763532763533, + "grad_norm": 0.8258776664733887, + "learning_rate": 5.072502467290201e-05, + "loss": 0.7467, + "step": 14928 + }, + { + "epoch": 2.6579415954415957, + "grad_norm": 0.8279047012329102, + "learning_rate": 5.071284496041138e-05, + "loss": 0.9148, + "step": 14929 + }, + { + "epoch": 2.658119658119658, + "grad_norm": 0.8176717758178711, + "learning_rate": 5.070066621360021e-05, + "loss": 1.0971, + "step": 14930 + }, + { + "epoch": 2.6582977207977208, + "grad_norm": 0.7482925057411194, + "learning_rate": 5.0688488432707074e-05, + "loss": 0.8666, + "step": 14931 + }, + { + "epoch": 2.6584757834757835, + "grad_norm": 0.9302734136581421, + "learning_rate": 5.067631161797057e-05, + "loss": 0.9994, + "step": 14932 + }, + { + "epoch": 2.6586538461538463, + "grad_norm": 0.7811494469642639, + "learning_rate": 5.066413576962927e-05, + "loss": 0.5959, + "step": 14933 + }, + { + "epoch": 2.6588319088319086, + "grad_norm": 0.8109773993492126, + "learning_rate": 5.065196088792177e-05, + "loss": 0.7342, + "step": 14934 + }, + { + "epoch": 2.6590099715099713, + "grad_norm": 0.8351961374282837, + "learning_rate": 5.0639786973086525e-05, + "loss": 0.775, + "step": 14935 + }, + { + "epoch": 2.659188034188034, + "grad_norm": 0.8558792471885681, + "learning_rate": 5.062761402536216e-05, + "loss": 0.8819, + "step": 14936 + }, + { + "epoch": 2.659366096866097, + "grad_norm": 0.7928652167320251, + "learning_rate": 5.061544204498714e-05, + "loss": 0.8313, + "step": 14937 + }, + { + "epoch": 2.6595441595441596, + "grad_norm": 0.8388734459877014, + "learning_rate": 5.060327103219993e-05, + "loss": 0.7208, + "step": 14938 + }, + { + "epoch": 2.6597222222222223, + "grad_norm": 0.8921391367912292, + "learning_rate": 5.059110098723903e-05, + "loss": 0.8974, + "step": 14939 + }, + { + "epoch": 2.659900284900285, + "grad_norm": 0.8111342787742615, + "learning_rate": 5.057893191034286e-05, + "loss": 0.6879, + "step": 14940 + }, + { + "epoch": 2.660078347578348, + "grad_norm": 0.8677322864532471, + "learning_rate": 5.056676380174985e-05, + "loss": 0.8643, + "step": 14941 + }, + { + "epoch": 2.66025641025641, + "grad_norm": 0.7969355583190918, + "learning_rate": 5.055459666169839e-05, + "loss": 0.8462, + "step": 14942 + }, + { + "epoch": 2.660434472934473, + "grad_norm": 0.9927026629447937, + "learning_rate": 5.0542430490426975e-05, + "loss": 0.7954, + "step": 14943 + }, + { + "epoch": 2.6606125356125356, + "grad_norm": 1.0181084871292114, + "learning_rate": 5.053026528817379e-05, + "loss": 0.9597, + "step": 14944 + }, + { + "epoch": 2.6607905982905984, + "grad_norm": 1.0274122953414917, + "learning_rate": 5.0518101055177355e-05, + "loss": 0.7321, + "step": 14945 + }, + { + "epoch": 2.6609686609686607, + "grad_norm": 1.056132197380066, + "learning_rate": 5.050593779167594e-05, + "loss": 0.8405, + "step": 14946 + }, + { + "epoch": 2.6611467236467234, + "grad_norm": 0.8586339950561523, + "learning_rate": 5.0493775497907846e-05, + "loss": 1.0238, + "step": 14947 + }, + { + "epoch": 2.661324786324786, + "grad_norm": 0.8103144764900208, + "learning_rate": 5.048161417411139e-05, + "loss": 0.5885, + "step": 14948 + }, + { + "epoch": 2.661502849002849, + "grad_norm": 0.7321345210075378, + "learning_rate": 5.0469453820524834e-05, + "loss": 0.7987, + "step": 14949 + }, + { + "epoch": 2.6616809116809117, + "grad_norm": 0.8244233727455139, + "learning_rate": 5.045729443738645e-05, + "loss": 0.8855, + "step": 14950 + }, + { + "epoch": 2.6618589743589745, + "grad_norm": 0.7888374924659729, + "learning_rate": 5.0445136024934456e-05, + "loss": 0.9192, + "step": 14951 + }, + { + "epoch": 2.662037037037037, + "grad_norm": 0.8414669036865234, + "learning_rate": 5.0432978583407044e-05, + "loss": 0.8152, + "step": 14952 + }, + { + "epoch": 2.6622150997151, + "grad_norm": 0.9176363348960876, + "learning_rate": 5.042082211304252e-05, + "loss": 0.8836, + "step": 14953 + }, + { + "epoch": 2.6623931623931623, + "grad_norm": 0.9827163219451904, + "learning_rate": 5.040866661407893e-05, + "loss": 0.963, + "step": 14954 + }, + { + "epoch": 2.662571225071225, + "grad_norm": 0.8765084743499756, + "learning_rate": 5.0396512086754535e-05, + "loss": 1.022, + "step": 14955 + }, + { + "epoch": 2.6627492877492878, + "grad_norm": 0.9236209392547607, + "learning_rate": 5.038435853130743e-05, + "loss": 0.9152, + "step": 14956 + }, + { + "epoch": 2.6629273504273505, + "grad_norm": 0.8300418853759766, + "learning_rate": 5.037220594797574e-05, + "loss": 0.8063, + "step": 14957 + }, + { + "epoch": 2.6631054131054133, + "grad_norm": 0.9248050451278687, + "learning_rate": 5.036005433699764e-05, + "loss": 0.8799, + "step": 14958 + }, + { + "epoch": 2.6632834757834756, + "grad_norm": 0.9670597910881042, + "learning_rate": 5.0347903698611085e-05, + "loss": 0.9068, + "step": 14959 + }, + { + "epoch": 2.6634615384615383, + "grad_norm": 0.851403534412384, + "learning_rate": 5.033575403305428e-05, + "loss": 0.8058, + "step": 14960 + }, + { + "epoch": 2.663639601139601, + "grad_norm": 0.9643952250480652, + "learning_rate": 5.032360534056515e-05, + "loss": 1.076, + "step": 14961 + }, + { + "epoch": 2.663817663817664, + "grad_norm": 0.8473731279373169, + "learning_rate": 5.031145762138181e-05, + "loss": 0.9585, + "step": 14962 + }, + { + "epoch": 2.6639957264957266, + "grad_norm": 0.8265015482902527, + "learning_rate": 5.029931087574222e-05, + "loss": 0.8602, + "step": 14963 + }, + { + "epoch": 2.6641737891737893, + "grad_norm": 0.8004183173179626, + "learning_rate": 5.0287165103884416e-05, + "loss": 0.7293, + "step": 14964 + }, + { + "epoch": 2.664351851851852, + "grad_norm": 0.8410465121269226, + "learning_rate": 5.027502030604633e-05, + "loss": 0.9479, + "step": 14965 + }, + { + "epoch": 2.6645299145299144, + "grad_norm": 0.8365132808685303, + "learning_rate": 5.0262876482465925e-05, + "loss": 0.7373, + "step": 14966 + }, + { + "epoch": 2.664707977207977, + "grad_norm": 0.9017055630683899, + "learning_rate": 5.025073363338111e-05, + "loss": 0.9463, + "step": 14967 + }, + { + "epoch": 2.66488603988604, + "grad_norm": 0.7985300421714783, + "learning_rate": 5.023859175902988e-05, + "loss": 0.7074, + "step": 14968 + }, + { + "epoch": 2.6650641025641026, + "grad_norm": 0.8032601475715637, + "learning_rate": 5.022645085965001e-05, + "loss": 0.6796, + "step": 14969 + }, + { + "epoch": 2.6652421652421654, + "grad_norm": 0.7785899639129639, + "learning_rate": 5.021431093547948e-05, + "loss": 0.7256, + "step": 14970 + }, + { + "epoch": 2.6654202279202277, + "grad_norm": 0.8083044290542603, + "learning_rate": 5.02021719867561e-05, + "loss": 0.9254, + "step": 14971 + }, + { + "epoch": 2.6655982905982905, + "grad_norm": 0.8896783590316772, + "learning_rate": 5.019003401371771e-05, + "loss": 0.9231, + "step": 14972 + }, + { + "epoch": 2.665776353276353, + "grad_norm": 0.9304720163345337, + "learning_rate": 5.017789701660215e-05, + "loss": 0.8915, + "step": 14973 + }, + { + "epoch": 2.665954415954416, + "grad_norm": 0.8683121204376221, + "learning_rate": 5.016576099564718e-05, + "loss": 0.8654, + "step": 14974 + }, + { + "epoch": 2.6661324786324787, + "grad_norm": 1.1082890033721924, + "learning_rate": 5.015362595109062e-05, + "loss": 1.0669, + "step": 14975 + }, + { + "epoch": 2.6663105413105415, + "grad_norm": 1.1696041822433472, + "learning_rate": 5.014149188317017e-05, + "loss": 0.9273, + "step": 14976 + }, + { + "epoch": 2.666488603988604, + "grad_norm": 0.8726202845573425, + "learning_rate": 5.0129358792123637e-05, + "loss": 0.6615, + "step": 14977 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.8246448636054993, + "learning_rate": 5.011722667818875e-05, + "loss": 0.8263, + "step": 14978 + }, + { + "epoch": 2.6668447293447293, + "grad_norm": 0.7201130390167236, + "learning_rate": 5.010509554160316e-05, + "loss": 0.7122, + "step": 14979 + }, + { + "epoch": 2.667022792022792, + "grad_norm": 0.8296586275100708, + "learning_rate": 5.009296538260457e-05, + "loss": 0.8816, + "step": 14980 + }, + { + "epoch": 2.6672008547008548, + "grad_norm": 0.8647085428237915, + "learning_rate": 5.008083620143067e-05, + "loss": 1.0, + "step": 14981 + }, + { + "epoch": 2.6673789173789175, + "grad_norm": 0.8175796270370483, + "learning_rate": 5.0068707998319045e-05, + "loss": 0.6727, + "step": 14982 + }, + { + "epoch": 2.66755698005698, + "grad_norm": 0.8537090420722961, + "learning_rate": 5.0056580773507434e-05, + "loss": 0.8034, + "step": 14983 + }, + { + "epoch": 2.6677350427350426, + "grad_norm": 0.7980232238769531, + "learning_rate": 5.00444545272333e-05, + "loss": 0.83, + "step": 14984 + }, + { + "epoch": 2.6679131054131053, + "grad_norm": 0.8231784701347351, + "learning_rate": 5.003232925973438e-05, + "loss": 0.6292, + "step": 14985 + }, + { + "epoch": 2.668091168091168, + "grad_norm": 0.9140519499778748, + "learning_rate": 5.0020204971248096e-05, + "loss": 0.893, + "step": 14986 + }, + { + "epoch": 2.668269230769231, + "grad_norm": 0.7462875247001648, + "learning_rate": 5.000808166201212e-05, + "loss": 0.7335, + "step": 14987 + }, + { + "epoch": 2.6684472934472936, + "grad_norm": 0.8201214671134949, + "learning_rate": 4.999595933226392e-05, + "loss": 0.7888, + "step": 14988 + }, + { + "epoch": 2.6686253561253563, + "grad_norm": 0.9165699481964111, + "learning_rate": 4.9983837982241024e-05, + "loss": 0.8808, + "step": 14989 + }, + { + "epoch": 2.6688034188034186, + "grad_norm": 0.9286229610443115, + "learning_rate": 4.997171761218092e-05, + "loss": 0.969, + "step": 14990 + }, + { + "epoch": 2.6689814814814814, + "grad_norm": 0.6710283160209656, + "learning_rate": 4.995959822232109e-05, + "loss": 0.6046, + "step": 14991 + }, + { + "epoch": 2.669159544159544, + "grad_norm": 0.9091618061065674, + "learning_rate": 4.994747981289895e-05, + "loss": 0.9747, + "step": 14992 + }, + { + "epoch": 2.669337606837607, + "grad_norm": 0.7992748618125916, + "learning_rate": 4.993536238415204e-05, + "loss": 0.8441, + "step": 14993 + }, + { + "epoch": 2.6695156695156697, + "grad_norm": 0.926811695098877, + "learning_rate": 4.992324593631762e-05, + "loss": 1.0308, + "step": 14994 + }, + { + "epoch": 2.669693732193732, + "grad_norm": 0.8966291546821594, + "learning_rate": 4.9911130469633216e-05, + "loss": 1.0689, + "step": 14995 + }, + { + "epoch": 2.6698717948717947, + "grad_norm": 0.8300046324729919, + "learning_rate": 4.989901598433616e-05, + "loss": 0.8539, + "step": 14996 + }, + { + "epoch": 2.6700498575498575, + "grad_norm": 0.9567606449127197, + "learning_rate": 4.988690248066381e-05, + "loss": 0.7707, + "step": 14997 + }, + { + "epoch": 2.67022792022792, + "grad_norm": 0.7993598580360413, + "learning_rate": 4.987478995885351e-05, + "loss": 0.9241, + "step": 14998 + }, + { + "epoch": 2.670405982905983, + "grad_norm": 0.9573900103569031, + "learning_rate": 4.986267841914253e-05, + "loss": 0.8051, + "step": 14999 + }, + { + "epoch": 2.6705840455840457, + "grad_norm": 0.8562188148498535, + "learning_rate": 4.985056786176828e-05, + "loss": 0.8818, + "step": 15000 + }, + { + "epoch": 2.6707621082621085, + "grad_norm": 0.7997880578041077, + "learning_rate": 4.983845828696792e-05, + "loss": 0.8749, + "step": 15001 + }, + { + "epoch": 2.6709401709401708, + "grad_norm": 0.8442137837409973, + "learning_rate": 4.982634969497879e-05, + "loss": 1.0239, + "step": 15002 + }, + { + "epoch": 2.6711182336182335, + "grad_norm": 0.775762140750885, + "learning_rate": 4.981424208603812e-05, + "loss": 0.728, + "step": 15003 + }, + { + "epoch": 2.6712962962962963, + "grad_norm": 0.7570006251335144, + "learning_rate": 4.9802135460383126e-05, + "loss": 0.6964, + "step": 15004 + }, + { + "epoch": 2.671474358974359, + "grad_norm": 0.8406931161880493, + "learning_rate": 4.979002981825101e-05, + "loss": 0.783, + "step": 15005 + }, + { + "epoch": 2.671652421652422, + "grad_norm": 0.827357828617096, + "learning_rate": 4.977792515987896e-05, + "loss": 0.9294, + "step": 15006 + }, + { + "epoch": 2.671830484330484, + "grad_norm": 0.9244057536125183, + "learning_rate": 4.9765821485504094e-05, + "loss": 0.8993, + "step": 15007 + }, + { + "epoch": 2.672008547008547, + "grad_norm": 0.7569696307182312, + "learning_rate": 4.975371879536368e-05, + "loss": 0.8305, + "step": 15008 + }, + { + "epoch": 2.6721866096866096, + "grad_norm": 0.8337959051132202, + "learning_rate": 4.9741617089694695e-05, + "loss": 0.8793, + "step": 15009 + }, + { + "epoch": 2.6723646723646723, + "grad_norm": 0.7254770994186401, + "learning_rate": 4.97295163687344e-05, + "loss": 0.9325, + "step": 15010 + }, + { + "epoch": 2.672542735042735, + "grad_norm": 0.7988013029098511, + "learning_rate": 4.971741663271972e-05, + "loss": 0.9787, + "step": 15011 + }, + { + "epoch": 2.672720797720798, + "grad_norm": 0.8326970338821411, + "learning_rate": 4.9705317881887845e-05, + "loss": 0.9164, + "step": 15012 + }, + { + "epoch": 2.6728988603988606, + "grad_norm": 0.7416687607765198, + "learning_rate": 4.96932201164758e-05, + "loss": 0.9041, + "step": 15013 + }, + { + "epoch": 2.6730769230769234, + "grad_norm": 0.868765652179718, + "learning_rate": 4.968112333672059e-05, + "loss": 0.646, + "step": 15014 + }, + { + "epoch": 2.6732549857549857, + "grad_norm": 0.7440044283866882, + "learning_rate": 4.966902754285925e-05, + "loss": 0.9147, + "step": 15015 + }, + { + "epoch": 2.6734330484330484, + "grad_norm": 0.8410077691078186, + "learning_rate": 4.9656932735128724e-05, + "loss": 0.8044, + "step": 15016 + }, + { + "epoch": 2.673611111111111, + "grad_norm": 0.8185286521911621, + "learning_rate": 4.964483891376606e-05, + "loss": 0.8057, + "step": 15017 + }, + { + "epoch": 2.673789173789174, + "grad_norm": 0.8550063967704773, + "learning_rate": 4.9632746079008166e-05, + "loss": 0.7841, + "step": 15018 + }, + { + "epoch": 2.673967236467236, + "grad_norm": 0.9171682000160217, + "learning_rate": 4.962065423109199e-05, + "loss": 0.8731, + "step": 15019 + }, + { + "epoch": 2.674145299145299, + "grad_norm": 0.8567686676979065, + "learning_rate": 4.9608563370254436e-05, + "loss": 0.9284, + "step": 15020 + }, + { + "epoch": 2.6743233618233617, + "grad_norm": 0.8641629219055176, + "learning_rate": 4.959647349673241e-05, + "loss": 1.0165, + "step": 15021 + }, + { + "epoch": 2.6745014245014245, + "grad_norm": 0.8058172464370728, + "learning_rate": 4.958438461076277e-05, + "loss": 0.9737, + "step": 15022 + }, + { + "epoch": 2.6746794871794872, + "grad_norm": 0.8329246640205383, + "learning_rate": 4.95722967125824e-05, + "loss": 0.7943, + "step": 15023 + }, + { + "epoch": 2.67485754985755, + "grad_norm": 0.9603211879730225, + "learning_rate": 4.956020980242807e-05, + "loss": 0.9453, + "step": 15024 + }, + { + "epoch": 2.6750356125356127, + "grad_norm": 0.8635705709457397, + "learning_rate": 4.9548123880536736e-05, + "loss": 0.9028, + "step": 15025 + }, + { + "epoch": 2.6752136752136755, + "grad_norm": 0.8909839987754822, + "learning_rate": 4.9536038947145024e-05, + "loss": 1.0376, + "step": 15026 + }, + { + "epoch": 2.675391737891738, + "grad_norm": 0.7507481575012207, + "learning_rate": 4.952395500248984e-05, + "loss": 0.7151, + "step": 15027 + }, + { + "epoch": 2.6755698005698005, + "grad_norm": 0.9425675272941589, + "learning_rate": 4.951187204680791e-05, + "loss": 0.839, + "step": 15028 + }, + { + "epoch": 2.6757478632478633, + "grad_norm": 0.8826829195022583, + "learning_rate": 4.949979008033596e-05, + "loss": 1.0107, + "step": 15029 + }, + { + "epoch": 2.675925925925926, + "grad_norm": 0.9209766387939453, + "learning_rate": 4.948770910331072e-05, + "loss": 0.8685, + "step": 15030 + }, + { + "epoch": 2.6761039886039883, + "grad_norm": 0.8018497824668884, + "learning_rate": 4.947562911596889e-05, + "loss": 0.7417, + "step": 15031 + }, + { + "epoch": 2.676282051282051, + "grad_norm": 0.7865417003631592, + "learning_rate": 4.9463550118547155e-05, + "loss": 0.9332, + "step": 15032 + }, + { + "epoch": 2.676460113960114, + "grad_norm": 0.8146806955337524, + "learning_rate": 4.945147211128216e-05, + "loss": 0.8658, + "step": 15033 + }, + { + "epoch": 2.6766381766381766, + "grad_norm": 0.8176286816596985, + "learning_rate": 4.943939509441054e-05, + "loss": 1.0603, + "step": 15034 + }, + { + "epoch": 2.6768162393162394, + "grad_norm": 0.8441028594970703, + "learning_rate": 4.942731906816897e-05, + "loss": 0.8699, + "step": 15035 + }, + { + "epoch": 2.676994301994302, + "grad_norm": 1.0035977363586426, + "learning_rate": 4.941524403279405e-05, + "loss": 0.8149, + "step": 15036 + }, + { + "epoch": 2.677172364672365, + "grad_norm": 0.8316586017608643, + "learning_rate": 4.9403169988522324e-05, + "loss": 0.9674, + "step": 15037 + }, + { + "epoch": 2.6773504273504276, + "grad_norm": 0.7379693388938904, + "learning_rate": 4.9391096935590375e-05, + "loss": 0.7097, + "step": 15038 + }, + { + "epoch": 2.67752849002849, + "grad_norm": 0.8861358165740967, + "learning_rate": 4.937902487423473e-05, + "loss": 0.9145, + "step": 15039 + }, + { + "epoch": 2.6777065527065527, + "grad_norm": 0.8769996166229248, + "learning_rate": 4.9366953804691994e-05, + "loss": 0.92, + "step": 15040 + }, + { + "epoch": 2.6778846153846154, + "grad_norm": 0.891703724861145, + "learning_rate": 4.9354883727198545e-05, + "loss": 0.8898, + "step": 15041 + }, + { + "epoch": 2.678062678062678, + "grad_norm": 0.8371208310127258, + "learning_rate": 4.934281464199099e-05, + "loss": 0.8868, + "step": 15042 + }, + { + "epoch": 2.6782407407407405, + "grad_norm": 0.8618297576904297, + "learning_rate": 4.933074654930574e-05, + "loss": 0.8577, + "step": 15043 + }, + { + "epoch": 2.6784188034188032, + "grad_norm": 0.7748361229896545, + "learning_rate": 4.931867944937926e-05, + "loss": 0.7273, + "step": 15044 + }, + { + "epoch": 2.678596866096866, + "grad_norm": 0.8320143222808838, + "learning_rate": 4.930661334244797e-05, + "loss": 0.8654, + "step": 15045 + }, + { + "epoch": 2.6787749287749287, + "grad_norm": 0.8370615243911743, + "learning_rate": 4.929454822874829e-05, + "loss": 0.751, + "step": 15046 + }, + { + "epoch": 2.6789529914529915, + "grad_norm": 0.9115342497825623, + "learning_rate": 4.9282484108516614e-05, + "loss": 0.823, + "step": 15047 + }, + { + "epoch": 2.6791310541310542, + "grad_norm": 0.9542914032936096, + "learning_rate": 4.9270420981989294e-05, + "loss": 0.9271, + "step": 15048 + }, + { + "epoch": 2.679309116809117, + "grad_norm": 0.765336275100708, + "learning_rate": 4.9258358849402655e-05, + "loss": 0.5523, + "step": 15049 + }, + { + "epoch": 2.6794871794871797, + "grad_norm": 0.8169335722923279, + "learning_rate": 4.924629771099315e-05, + "loss": 0.7437, + "step": 15050 + }, + { + "epoch": 2.679665242165242, + "grad_norm": 0.8192304968833923, + "learning_rate": 4.9234237566996935e-05, + "loss": 0.8888, + "step": 15051 + }, + { + "epoch": 2.679843304843305, + "grad_norm": 0.8657594919204712, + "learning_rate": 4.922217841765041e-05, + "loss": 0.9858, + "step": 15052 + }, + { + "epoch": 2.6800213675213675, + "grad_norm": 0.9291723370552063, + "learning_rate": 4.921012026318982e-05, + "loss": 0.9731, + "step": 15053 + }, + { + "epoch": 2.6801994301994303, + "grad_norm": 0.7988953590393066, + "learning_rate": 4.919806310385138e-05, + "loss": 0.8467, + "step": 15054 + }, + { + "epoch": 2.6803774928774926, + "grad_norm": 0.8022913336753845, + "learning_rate": 4.9186006939871434e-05, + "loss": 0.9009, + "step": 15055 + }, + { + "epoch": 2.6805555555555554, + "grad_norm": 0.8444825410842896, + "learning_rate": 4.917395177148605e-05, + "loss": 0.8851, + "step": 15056 + }, + { + "epoch": 2.680733618233618, + "grad_norm": 0.8054760694503784, + "learning_rate": 4.9161897598931575e-05, + "loss": 0.8679, + "step": 15057 + }, + { + "epoch": 2.680911680911681, + "grad_norm": 0.8291507959365845, + "learning_rate": 4.9149844422444023e-05, + "loss": 0.7229, + "step": 15058 + }, + { + "epoch": 2.6810897435897436, + "grad_norm": 0.9225491285324097, + "learning_rate": 4.91377922422597e-05, + "loss": 0.7584, + "step": 15059 + }, + { + "epoch": 2.6812678062678064, + "grad_norm": 0.9598490595817566, + "learning_rate": 4.912574105861466e-05, + "loss": 1.0548, + "step": 15060 + }, + { + "epoch": 2.681445868945869, + "grad_norm": 0.7480899691581726, + "learning_rate": 4.911369087174504e-05, + "loss": 0.8389, + "step": 15061 + }, + { + "epoch": 2.681623931623932, + "grad_norm": 1.0396811962127686, + "learning_rate": 4.910164168188696e-05, + "loss": 0.8776, + "step": 15062 + }, + { + "epoch": 2.681801994301994, + "grad_norm": 0.8191503882408142, + "learning_rate": 4.9089593489276465e-05, + "loss": 0.7601, + "step": 15063 + }, + { + "epoch": 2.681980056980057, + "grad_norm": 0.8405289053916931, + "learning_rate": 4.907754629414959e-05, + "loss": 1.0859, + "step": 15064 + }, + { + "epoch": 2.6821581196581197, + "grad_norm": 0.8369600176811218, + "learning_rate": 4.90655000967425e-05, + "loss": 0.9159, + "step": 15065 + }, + { + "epoch": 2.6823361823361824, + "grad_norm": 0.8304924368858337, + "learning_rate": 4.905345489729104e-05, + "loss": 0.743, + "step": 15066 + }, + { + "epoch": 2.682514245014245, + "grad_norm": 0.7378702163696289, + "learning_rate": 4.904141069603139e-05, + "loss": 0.9386, + "step": 15067 + }, + { + "epoch": 2.6826923076923075, + "grad_norm": 0.9135075807571411, + "learning_rate": 4.902936749319935e-05, + "loss": 0.7341, + "step": 15068 + }, + { + "epoch": 2.6828703703703702, + "grad_norm": 0.77586430311203, + "learning_rate": 4.901732528903101e-05, + "loss": 0.5586, + "step": 15069 + }, + { + "epoch": 2.683048433048433, + "grad_norm": 0.8733307719230652, + "learning_rate": 4.900528408376228e-05, + "loss": 0.8173, + "step": 15070 + }, + { + "epoch": 2.6832264957264957, + "grad_norm": 0.7499578595161438, + "learning_rate": 4.8993243877629066e-05, + "loss": 0.7355, + "step": 15071 + }, + { + "epoch": 2.6834045584045585, + "grad_norm": 0.8372282385826111, + "learning_rate": 4.8981204670867295e-05, + "loss": 0.8169, + "step": 15072 + }, + { + "epoch": 2.6835826210826212, + "grad_norm": 0.7705212235450745, + "learning_rate": 4.8969166463712834e-05, + "loss": 0.7382, + "step": 15073 + }, + { + "epoch": 2.683760683760684, + "grad_norm": 0.8367058038711548, + "learning_rate": 4.89571292564015e-05, + "loss": 0.7268, + "step": 15074 + }, + { + "epoch": 2.6839387464387463, + "grad_norm": 0.8421934843063354, + "learning_rate": 4.8945093049169233e-05, + "loss": 0.8319, + "step": 15075 + }, + { + "epoch": 2.684116809116809, + "grad_norm": 0.8927276730537415, + "learning_rate": 4.893305784225181e-05, + "loss": 0.8669, + "step": 15076 + }, + { + "epoch": 2.684294871794872, + "grad_norm": 0.8147335052490234, + "learning_rate": 4.892102363588503e-05, + "loss": 0.7722, + "step": 15077 + }, + { + "epoch": 2.6844729344729346, + "grad_norm": 0.9491320848464966, + "learning_rate": 4.890899043030469e-05, + "loss": 1.0213, + "step": 15078 + }, + { + "epoch": 2.6846509971509973, + "grad_norm": 0.8635398745536804, + "learning_rate": 4.889695822574651e-05, + "loss": 0.797, + "step": 15079 + }, + { + "epoch": 2.6848290598290596, + "grad_norm": 0.7290985584259033, + "learning_rate": 4.888492702244636e-05, + "loss": 1.0142, + "step": 15080 + }, + { + "epoch": 2.6850071225071224, + "grad_norm": 0.7667058110237122, + "learning_rate": 4.8872896820639794e-05, + "loss": 0.7547, + "step": 15081 + }, + { + "epoch": 2.685185185185185, + "grad_norm": 0.9096128344535828, + "learning_rate": 4.886086762056269e-05, + "loss": 0.7972, + "step": 15082 + }, + { + "epoch": 2.685363247863248, + "grad_norm": 0.7461803555488586, + "learning_rate": 4.884883942245057e-05, + "loss": 0.8994, + "step": 15083 + }, + { + "epoch": 2.6855413105413106, + "grad_norm": 0.7640016674995422, + "learning_rate": 4.883681222653923e-05, + "loss": 0.7607, + "step": 15084 + }, + { + "epoch": 2.6857193732193734, + "grad_norm": 0.7481253743171692, + "learning_rate": 4.882478603306427e-05, + "loss": 0.7089, + "step": 15085 + }, + { + "epoch": 2.685897435897436, + "grad_norm": 0.825998842716217, + "learning_rate": 4.881276084226132e-05, + "loss": 0.6617, + "step": 15086 + }, + { + "epoch": 2.6860754985754984, + "grad_norm": 0.9775291085243225, + "learning_rate": 4.8800736654365986e-05, + "loss": 0.9345, + "step": 15087 + }, + { + "epoch": 2.686253561253561, + "grad_norm": 0.8158339262008667, + "learning_rate": 4.878871346961387e-05, + "loss": 0.8198, + "step": 15088 + }, + { + "epoch": 2.686431623931624, + "grad_norm": 0.8778133988380432, + "learning_rate": 4.8776691288240486e-05, + "loss": 0.8323, + "step": 15089 + }, + { + "epoch": 2.6866096866096867, + "grad_norm": 0.9657309055328369, + "learning_rate": 4.8764670110481505e-05, + "loss": 0.907, + "step": 15090 + }, + { + "epoch": 2.6867877492877494, + "grad_norm": 1.0467438697814941, + "learning_rate": 4.8752649936572304e-05, + "loss": 1.0128, + "step": 15091 + }, + { + "epoch": 2.6869658119658117, + "grad_norm": 0.7682142853736877, + "learning_rate": 4.874063076674854e-05, + "loss": 1.0164, + "step": 15092 + }, + { + "epoch": 2.6871438746438745, + "grad_norm": 0.8184331059455872, + "learning_rate": 4.8728612601245574e-05, + "loss": 0.6614, + "step": 15093 + }, + { + "epoch": 2.6873219373219372, + "grad_norm": 0.8372936248779297, + "learning_rate": 4.871659544029896e-05, + "loss": 0.9011, + "step": 15094 + }, + { + "epoch": 2.6875, + "grad_norm": 0.7872710824012756, + "learning_rate": 4.870457928414414e-05, + "loss": 0.6986, + "step": 15095 + }, + { + "epoch": 2.6876780626780628, + "grad_norm": 0.7297250628471375, + "learning_rate": 4.8692564133016485e-05, + "loss": 0.5399, + "step": 15096 + }, + { + "epoch": 2.6878561253561255, + "grad_norm": 0.8855645060539246, + "learning_rate": 4.868054998715153e-05, + "loss": 0.8992, + "step": 15097 + }, + { + "epoch": 2.6880341880341883, + "grad_norm": 0.9055765271186829, + "learning_rate": 4.866853684678452e-05, + "loss": 0.888, + "step": 15098 + }, + { + "epoch": 2.6882122507122506, + "grad_norm": 1.0414996147155762, + "learning_rate": 4.865652471215093e-05, + "loss": 1.0375, + "step": 15099 + }, + { + "epoch": 2.6883903133903133, + "grad_norm": 0.8606446385383606, + "learning_rate": 4.8644513583486086e-05, + "loss": 0.8906, + "step": 15100 + }, + { + "epoch": 2.688568376068376, + "grad_norm": 0.9065528512001038, + "learning_rate": 4.8632503461025316e-05, + "loss": 0.848, + "step": 15101 + }, + { + "epoch": 2.688746438746439, + "grad_norm": 0.7832834720611572, + "learning_rate": 4.862049434500393e-05, + "loss": 0.7028, + "step": 15102 + }, + { + "epoch": 2.6889245014245016, + "grad_norm": 0.7107385396957397, + "learning_rate": 4.860848623565723e-05, + "loss": 0.7249, + "step": 15103 + }, + { + "epoch": 2.689102564102564, + "grad_norm": 0.8936449289321899, + "learning_rate": 4.8596479133220485e-05, + "loss": 0.9651, + "step": 15104 + }, + { + "epoch": 2.6892806267806266, + "grad_norm": 0.9019163846969604, + "learning_rate": 4.8584473037928944e-05, + "loss": 0.7165, + "step": 15105 + }, + { + "epoch": 2.6894586894586894, + "grad_norm": 0.8838223218917847, + "learning_rate": 4.857246795001782e-05, + "loss": 0.8148, + "step": 15106 + }, + { + "epoch": 2.689636752136752, + "grad_norm": 0.8004612922668457, + "learning_rate": 4.856046386972243e-05, + "loss": 0.9109, + "step": 15107 + }, + { + "epoch": 2.689814814814815, + "grad_norm": 0.9337486028671265, + "learning_rate": 4.854846079727781e-05, + "loss": 1.0952, + "step": 15108 + }, + { + "epoch": 2.6899928774928776, + "grad_norm": 0.6513102650642395, + "learning_rate": 4.853645873291926e-05, + "loss": 0.5435, + "step": 15109 + }, + { + "epoch": 2.6901709401709404, + "grad_norm": 0.8750485181808472, + "learning_rate": 4.85244576768819e-05, + "loss": 0.8783, + "step": 15110 + }, + { + "epoch": 2.6903490028490027, + "grad_norm": 0.9513342976570129, + "learning_rate": 4.851245762940085e-05, + "loss": 0.8822, + "step": 15111 + }, + { + "epoch": 2.6905270655270654, + "grad_norm": 0.8832191824913025, + "learning_rate": 4.850045859071125e-05, + "loss": 0.9216, + "step": 15112 + }, + { + "epoch": 2.690705128205128, + "grad_norm": 0.875396728515625, + "learning_rate": 4.8488460561048175e-05, + "loss": 0.998, + "step": 15113 + }, + { + "epoch": 2.690883190883191, + "grad_norm": 0.8847890496253967, + "learning_rate": 4.847646354064668e-05, + "loss": 1.0916, + "step": 15114 + }, + { + "epoch": 2.6910612535612537, + "grad_norm": 0.8235226273536682, + "learning_rate": 4.846446752974187e-05, + "loss": 0.8154, + "step": 15115 + }, + { + "epoch": 2.691239316239316, + "grad_norm": 0.8099366426467896, + "learning_rate": 4.845247252856878e-05, + "loss": 0.9392, + "step": 15116 + }, + { + "epoch": 2.6914173789173788, + "grad_norm": 0.8525599837303162, + "learning_rate": 4.84404785373624e-05, + "loss": 0.6619, + "step": 15117 + }, + { + "epoch": 2.6915954415954415, + "grad_norm": 1.0223274230957031, + "learning_rate": 4.842848555635775e-05, + "loss": 0.9479, + "step": 15118 + }, + { + "epoch": 2.6917735042735043, + "grad_norm": 0.7834655046463013, + "learning_rate": 4.841649358578978e-05, + "loss": 0.6962, + "step": 15119 + }, + { + "epoch": 2.691951566951567, + "grad_norm": 0.787391185760498, + "learning_rate": 4.8404502625893474e-05, + "loss": 0.8598, + "step": 15120 + }, + { + "epoch": 2.6921296296296298, + "grad_norm": 0.907228410243988, + "learning_rate": 4.839251267690371e-05, + "loss": 0.9913, + "step": 15121 + }, + { + "epoch": 2.6923076923076925, + "grad_norm": 0.8313533663749695, + "learning_rate": 4.838052373905554e-05, + "loss": 0.9542, + "step": 15122 + }, + { + "epoch": 2.692485754985755, + "grad_norm": 0.8444675207138062, + "learning_rate": 4.83685358125837e-05, + "loss": 0.7437, + "step": 15123 + }, + { + "epoch": 2.6926638176638176, + "grad_norm": 0.8656189441680908, + "learning_rate": 4.835654889772319e-05, + "loss": 1.104, + "step": 15124 + }, + { + "epoch": 2.6928418803418803, + "grad_norm": 0.9181584715843201, + "learning_rate": 4.8344562994708805e-05, + "loss": 0.8533, + "step": 15125 + }, + { + "epoch": 2.693019943019943, + "grad_norm": 0.5977702140808105, + "learning_rate": 4.833257810377542e-05, + "loss": 0.495, + "step": 15126 + }, + { + "epoch": 2.693198005698006, + "grad_norm": 0.8839932084083557, + "learning_rate": 4.8320594225157834e-05, + "loss": 0.8026, + "step": 15127 + }, + { + "epoch": 2.693376068376068, + "grad_norm": 0.876559853553772, + "learning_rate": 4.8308611359090846e-05, + "loss": 0.893, + "step": 15128 + }, + { + "epoch": 2.693554131054131, + "grad_norm": 0.7847880721092224, + "learning_rate": 4.829662950580924e-05, + "loss": 0.7794, + "step": 15129 + }, + { + "epoch": 2.6937321937321936, + "grad_norm": 0.8713442087173462, + "learning_rate": 4.828464866554778e-05, + "loss": 1.0394, + "step": 15130 + }, + { + "epoch": 2.6939102564102564, + "grad_norm": 0.9720988869667053, + "learning_rate": 4.827266883854116e-05, + "loss": 0.7844, + "step": 15131 + }, + { + "epoch": 2.694088319088319, + "grad_norm": 0.8163195252418518, + "learning_rate": 4.82606900250242e-05, + "loss": 0.711, + "step": 15132 + }, + { + "epoch": 2.694266381766382, + "grad_norm": 0.7119855880737305, + "learning_rate": 4.8248712225231486e-05, + "loss": 0.6224, + "step": 15133 + }, + { + "epoch": 2.6944444444444446, + "grad_norm": 0.8176950812339783, + "learning_rate": 4.823673543939777e-05, + "loss": 0.8695, + "step": 15134 + }, + { + "epoch": 2.6946225071225074, + "grad_norm": 0.8138632774353027, + "learning_rate": 4.822475966775771e-05, + "loss": 0.7331, + "step": 15135 + }, + { + "epoch": 2.6948005698005697, + "grad_norm": 0.9323116540908813, + "learning_rate": 4.821278491054589e-05, + "loss": 0.8275, + "step": 15136 + }, + { + "epoch": 2.6949786324786325, + "grad_norm": 0.7593950033187866, + "learning_rate": 4.820081116799704e-05, + "loss": 0.9571, + "step": 15137 + }, + { + "epoch": 2.695156695156695, + "grad_norm": 0.9058876037597656, + "learning_rate": 4.818883844034563e-05, + "loss": 0.7676, + "step": 15138 + }, + { + "epoch": 2.695334757834758, + "grad_norm": 1.0943962335586548, + "learning_rate": 4.8176866727826365e-05, + "loss": 0.7542, + "step": 15139 + }, + { + "epoch": 2.6955128205128203, + "grad_norm": 0.9133912324905396, + "learning_rate": 4.8164896030673664e-05, + "loss": 0.8419, + "step": 15140 + }, + { + "epoch": 2.695690883190883, + "grad_norm": 0.8556821942329407, + "learning_rate": 4.8152926349122195e-05, + "loss": 0.8234, + "step": 15141 + }, + { + "epoch": 2.6958689458689458, + "grad_norm": 1.0329471826553345, + "learning_rate": 4.814095768340643e-05, + "loss": 0.8181, + "step": 15142 + }, + { + "epoch": 2.6960470085470085, + "grad_norm": 0.89934903383255, + "learning_rate": 4.812899003376087e-05, + "loss": 0.8392, + "step": 15143 + }, + { + "epoch": 2.6962250712250713, + "grad_norm": 0.7836576104164124, + "learning_rate": 4.811702340042e-05, + "loss": 0.9491, + "step": 15144 + }, + { + "epoch": 2.696403133903134, + "grad_norm": 0.9841184020042419, + "learning_rate": 4.810505778361828e-05, + "loss": 1.0763, + "step": 15145 + }, + { + "epoch": 2.6965811965811968, + "grad_norm": 1.0479893684387207, + "learning_rate": 4.80930931835901e-05, + "loss": 1.054, + "step": 15146 + }, + { + "epoch": 2.6967592592592595, + "grad_norm": 0.895803689956665, + "learning_rate": 4.808112960057002e-05, + "loss": 0.8769, + "step": 15147 + }, + { + "epoch": 2.696937321937322, + "grad_norm": 0.8467312455177307, + "learning_rate": 4.806916703479227e-05, + "loss": 0.8036, + "step": 15148 + }, + { + "epoch": 2.6971153846153846, + "grad_norm": 0.7371073365211487, + "learning_rate": 4.8057205486491366e-05, + "loss": 0.72, + "step": 15149 + }, + { + "epoch": 2.6972934472934473, + "grad_norm": 0.9631866812705994, + "learning_rate": 4.80452449559016e-05, + "loss": 0.8661, + "step": 15150 + }, + { + "epoch": 2.69747150997151, + "grad_norm": 0.8467531204223633, + "learning_rate": 4.803328544325735e-05, + "loss": 0.9359, + "step": 15151 + }, + { + "epoch": 2.6976495726495724, + "grad_norm": 0.8170605897903442, + "learning_rate": 4.802132694879291e-05, + "loss": 0.9086, + "step": 15152 + }, + { + "epoch": 2.697827635327635, + "grad_norm": 0.8378857970237732, + "learning_rate": 4.800936947274255e-05, + "loss": 0.6255, + "step": 15153 + }, + { + "epoch": 2.698005698005698, + "grad_norm": 0.8074176907539368, + "learning_rate": 4.799741301534067e-05, + "loss": 0.9129, + "step": 15154 + }, + { + "epoch": 2.6981837606837606, + "grad_norm": 0.862147331237793, + "learning_rate": 4.798545757682139e-05, + "loss": 0.8298, + "step": 15155 + }, + { + "epoch": 2.6983618233618234, + "grad_norm": 0.8020915985107422, + "learning_rate": 4.797350315741905e-05, + "loss": 0.8364, + "step": 15156 + }, + { + "epoch": 2.698539886039886, + "grad_norm": 0.7929054498672485, + "learning_rate": 4.7961549757367854e-05, + "loss": 1.0302, + "step": 15157 + }, + { + "epoch": 2.698717948717949, + "grad_norm": 0.8528931140899658, + "learning_rate": 4.7949597376901964e-05, + "loss": 0.7891, + "step": 15158 + }, + { + "epoch": 2.6988960113960117, + "grad_norm": 0.8090588450431824, + "learning_rate": 4.793764601625561e-05, + "loss": 0.7905, + "step": 15159 + }, + { + "epoch": 2.699074074074074, + "grad_norm": 0.8221202492713928, + "learning_rate": 4.7925695675662916e-05, + "loss": 0.8156, + "step": 15160 + }, + { + "epoch": 2.6992521367521367, + "grad_norm": 0.8121498823165894, + "learning_rate": 4.791374635535802e-05, + "loss": 0.865, + "step": 15161 + }, + { + "epoch": 2.6994301994301995, + "grad_norm": 0.7626228928565979, + "learning_rate": 4.790179805557513e-05, + "loss": 0.8033, + "step": 15162 + }, + { + "epoch": 2.699608262108262, + "grad_norm": 0.8483169078826904, + "learning_rate": 4.7889850776548205e-05, + "loss": 0.9239, + "step": 15163 + }, + { + "epoch": 2.6997863247863245, + "grad_norm": 0.8302589058876038, + "learning_rate": 4.7877904518511485e-05, + "loss": 0.8445, + "step": 15164 + }, + { + "epoch": 2.6999643874643873, + "grad_norm": 0.9140453338623047, + "learning_rate": 4.786595928169887e-05, + "loss": 1.0492, + "step": 15165 + }, + { + "epoch": 2.70014245014245, + "grad_norm": 0.8046873807907104, + "learning_rate": 4.785401506634453e-05, + "loss": 1.0009, + "step": 15166 + }, + { + "epoch": 2.7003205128205128, + "grad_norm": 0.8879752159118652, + "learning_rate": 4.7842071872682434e-05, + "loss": 0.7788, + "step": 15167 + }, + { + "epoch": 2.7004985754985755, + "grad_norm": 0.8190163969993591, + "learning_rate": 4.783012970094659e-05, + "loss": 0.9063, + "step": 15168 + }, + { + "epoch": 2.7006766381766383, + "grad_norm": 0.9363130331039429, + "learning_rate": 4.781818855137099e-05, + "loss": 0.9723, + "step": 15169 + }, + { + "epoch": 2.700854700854701, + "grad_norm": 0.8428171873092651, + "learning_rate": 4.780624842418958e-05, + "loss": 0.9173, + "step": 15170 + }, + { + "epoch": 2.701032763532764, + "grad_norm": 0.8089821934700012, + "learning_rate": 4.779430931963627e-05, + "loss": 0.8996, + "step": 15171 + }, + { + "epoch": 2.701210826210826, + "grad_norm": 0.8893290758132935, + "learning_rate": 4.77823712379451e-05, + "loss": 0.9483, + "step": 15172 + }, + { + "epoch": 2.701388888888889, + "grad_norm": 0.8589824438095093, + "learning_rate": 4.777043417934981e-05, + "loss": 0.8765, + "step": 15173 + }, + { + "epoch": 2.7015669515669516, + "grad_norm": 0.8665438294410706, + "learning_rate": 4.7758498144084405e-05, + "loss": 0.8546, + "step": 15174 + }, + { + "epoch": 2.7017450142450143, + "grad_norm": 0.743841826915741, + "learning_rate": 4.774656313238272e-05, + "loss": 0.6866, + "step": 15175 + }, + { + "epoch": 2.7019230769230766, + "grad_norm": 0.9317346811294556, + "learning_rate": 4.7734629144478574e-05, + "loss": 0.8004, + "step": 15176 + }, + { + "epoch": 2.7021011396011394, + "grad_norm": 0.8244655132293701, + "learning_rate": 4.77226961806058e-05, + "loss": 0.9302, + "step": 15177 + }, + { + "epoch": 2.702279202279202, + "grad_norm": 1.0759600400924683, + "learning_rate": 4.771076424099815e-05, + "loss": 0.9073, + "step": 15178 + }, + { + "epoch": 2.702457264957265, + "grad_norm": 0.8852303624153137, + "learning_rate": 4.769883332588954e-05, + "loss": 0.8084, + "step": 15179 + }, + { + "epoch": 2.7026353276353277, + "grad_norm": 0.8642051815986633, + "learning_rate": 4.7686903435513564e-05, + "loss": 1.0018, + "step": 15180 + }, + { + "epoch": 2.7028133903133904, + "grad_norm": 0.9442928433418274, + "learning_rate": 4.767497457010408e-05, + "loss": 0.8099, + "step": 15181 + }, + { + "epoch": 2.702991452991453, + "grad_norm": 0.8357751965522766, + "learning_rate": 4.7663046729894776e-05, + "loss": 0.8594, + "step": 15182 + }, + { + "epoch": 2.703169515669516, + "grad_norm": 1.0791765451431274, + "learning_rate": 4.765111991511936e-05, + "loss": 1.1203, + "step": 15183 + }, + { + "epoch": 2.703347578347578, + "grad_norm": 0.7855654954910278, + "learning_rate": 4.7639194126011485e-05, + "loss": 0.7218, + "step": 15184 + }, + { + "epoch": 2.703525641025641, + "grad_norm": 0.8058420419692993, + "learning_rate": 4.762726936280485e-05, + "loss": 0.7885, + "step": 15185 + }, + { + "epoch": 2.7037037037037037, + "grad_norm": 0.7701787352561951, + "learning_rate": 4.761534562573302e-05, + "loss": 0.6378, + "step": 15186 + }, + { + "epoch": 2.7038817663817665, + "grad_norm": 0.9011744856834412, + "learning_rate": 4.760342291502976e-05, + "loss": 0.9106, + "step": 15187 + }, + { + "epoch": 2.7040598290598292, + "grad_norm": 0.7268012762069702, + "learning_rate": 4.759150123092851e-05, + "loss": 0.6303, + "step": 15188 + }, + { + "epoch": 2.7042378917378915, + "grad_norm": 0.8369283676147461, + "learning_rate": 4.7579580573663e-05, + "loss": 0.8013, + "step": 15189 + }, + { + "epoch": 2.7044159544159543, + "grad_norm": 0.9511098861694336, + "learning_rate": 4.756766094346663e-05, + "loss": 1.0211, + "step": 15190 + }, + { + "epoch": 2.704594017094017, + "grad_norm": 0.8408896327018738, + "learning_rate": 4.7555742340573074e-05, + "loss": 1.1018, + "step": 15191 + }, + { + "epoch": 2.70477207977208, + "grad_norm": 0.9166504740715027, + "learning_rate": 4.7543824765215795e-05, + "loss": 0.9222, + "step": 15192 + }, + { + "epoch": 2.7049501424501425, + "grad_norm": 0.8373738527297974, + "learning_rate": 4.753190821762826e-05, + "loss": 0.9735, + "step": 15193 + }, + { + "epoch": 2.7051282051282053, + "grad_norm": 0.8610605597496033, + "learning_rate": 4.751999269804408e-05, + "loss": 0.7942, + "step": 15194 + }, + { + "epoch": 2.705306267806268, + "grad_norm": 0.8778019547462463, + "learning_rate": 4.750807820669654e-05, + "loss": 0.8055, + "step": 15195 + }, + { + "epoch": 2.7054843304843303, + "grad_norm": 0.9997664093971252, + "learning_rate": 4.749616474381921e-05, + "loss": 0.8461, + "step": 15196 + }, + { + "epoch": 2.705662393162393, + "grad_norm": 0.8362101316452026, + "learning_rate": 4.748425230964545e-05, + "loss": 1.0008, + "step": 15197 + }, + { + "epoch": 2.705840455840456, + "grad_norm": 0.870482861995697, + "learning_rate": 4.747234090440869e-05, + "loss": 0.9547, + "step": 15198 + }, + { + "epoch": 2.7060185185185186, + "grad_norm": 0.867431104183197, + "learning_rate": 4.746043052834228e-05, + "loss": 0.8533, + "step": 15199 + }, + { + "epoch": 2.7061965811965814, + "grad_norm": 0.842071533203125, + "learning_rate": 4.7448521181679604e-05, + "loss": 0.8919, + "step": 15200 + }, + { + "epoch": 2.7063746438746437, + "grad_norm": 0.9487791657447815, + "learning_rate": 4.743661286465398e-05, + "loss": 0.8072, + "step": 15201 + }, + { + "epoch": 2.7065527065527064, + "grad_norm": 0.8469042181968689, + "learning_rate": 4.742470557749874e-05, + "loss": 0.8792, + "step": 15202 + }, + { + "epoch": 2.706730769230769, + "grad_norm": 0.86415696144104, + "learning_rate": 4.7412799320447145e-05, + "loss": 0.9725, + "step": 15203 + }, + { + "epoch": 2.706908831908832, + "grad_norm": 0.9035004377365112, + "learning_rate": 4.740089409373257e-05, + "loss": 0.9915, + "step": 15204 + }, + { + "epoch": 2.7070868945868947, + "grad_norm": 0.8122807741165161, + "learning_rate": 4.7388989897588156e-05, + "loss": 0.946, + "step": 15205 + }, + { + "epoch": 2.7072649572649574, + "grad_norm": 0.9801422357559204, + "learning_rate": 4.737708673224721e-05, + "loss": 0.9357, + "step": 15206 + }, + { + "epoch": 2.70744301994302, + "grad_norm": 1.0265265703201294, + "learning_rate": 4.736518459794295e-05, + "loss": 0.7982, + "step": 15207 + }, + { + "epoch": 2.7076210826210825, + "grad_norm": 0.828814685344696, + "learning_rate": 4.735328349490855e-05, + "loss": 0.6864, + "step": 15208 + }, + { + "epoch": 2.7077991452991452, + "grad_norm": 0.7948212623596191, + "learning_rate": 4.7341383423377195e-05, + "loss": 0.8661, + "step": 15209 + }, + { + "epoch": 2.707977207977208, + "grad_norm": 0.8372616767883301, + "learning_rate": 4.7329484383582046e-05, + "loss": 0.8818, + "step": 15210 + }, + { + "epoch": 2.7081552706552707, + "grad_norm": 0.8000285029411316, + "learning_rate": 4.731758637575624e-05, + "loss": 0.8006, + "step": 15211 + }, + { + "epoch": 2.7083333333333335, + "grad_norm": 0.7860875725746155, + "learning_rate": 4.730568940013289e-05, + "loss": 0.926, + "step": 15212 + }, + { + "epoch": 2.708511396011396, + "grad_norm": 0.9157412052154541, + "learning_rate": 4.7293793456945054e-05, + "loss": 0.7042, + "step": 15213 + }, + { + "epoch": 2.7086894586894585, + "grad_norm": 0.8802906274795532, + "learning_rate": 4.728189854642589e-05, + "loss": 0.8639, + "step": 15214 + }, + { + "epoch": 2.7088675213675213, + "grad_norm": 0.8047248721122742, + "learning_rate": 4.7270004668808397e-05, + "loss": 0.7603, + "step": 15215 + }, + { + "epoch": 2.709045584045584, + "grad_norm": 0.9848080277442932, + "learning_rate": 4.725811182432564e-05, + "loss": 0.8213, + "step": 15216 + }, + { + "epoch": 2.709223646723647, + "grad_norm": 0.8568090200424194, + "learning_rate": 4.724622001321062e-05, + "loss": 0.7663, + "step": 15217 + }, + { + "epoch": 2.7094017094017095, + "grad_norm": 0.7926214337348938, + "learning_rate": 4.7234329235696284e-05, + "loss": 0.874, + "step": 15218 + }, + { + "epoch": 2.7095797720797723, + "grad_norm": 0.8389978408813477, + "learning_rate": 4.7222439492015734e-05, + "loss": 0.623, + "step": 15219 + }, + { + "epoch": 2.7097578347578346, + "grad_norm": 0.8635174036026001, + "learning_rate": 4.7210550782401773e-05, + "loss": 0.822, + "step": 15220 + }, + { + "epoch": 2.7099358974358974, + "grad_norm": 0.8381666541099548, + "learning_rate": 4.7198663107087446e-05, + "loss": 1.0864, + "step": 15221 + }, + { + "epoch": 2.71011396011396, + "grad_norm": 1.0722376108169556, + "learning_rate": 4.718677646630564e-05, + "loss": 0.8527, + "step": 15222 + }, + { + "epoch": 2.710292022792023, + "grad_norm": 0.9505516290664673, + "learning_rate": 4.7174890860289224e-05, + "loss": 1.0645, + "step": 15223 + }, + { + "epoch": 2.7104700854700856, + "grad_norm": 0.7757406234741211, + "learning_rate": 4.7163006289271095e-05, + "loss": 0.5924, + "step": 15224 + }, + { + "epoch": 2.710648148148148, + "grad_norm": 0.816387414932251, + "learning_rate": 4.71511227534841e-05, + "loss": 0.8337, + "step": 15225 + }, + { + "epoch": 2.7108262108262107, + "grad_norm": 0.7817156910896301, + "learning_rate": 4.7139240253161065e-05, + "loss": 0.8315, + "step": 15226 + }, + { + "epoch": 2.7110042735042734, + "grad_norm": 0.9753041863441467, + "learning_rate": 4.7127358788534816e-05, + "loss": 0.851, + "step": 15227 + }, + { + "epoch": 2.711182336182336, + "grad_norm": 0.7564638257026672, + "learning_rate": 4.7115478359838095e-05, + "loss": 0.8132, + "step": 15228 + }, + { + "epoch": 2.711360398860399, + "grad_norm": 0.8709259629249573, + "learning_rate": 4.710359896730379e-05, + "loss": 1.0277, + "step": 15229 + }, + { + "epoch": 2.7115384615384617, + "grad_norm": 0.9849836230278015, + "learning_rate": 4.7091720611164504e-05, + "loss": 0.9778, + "step": 15230 + }, + { + "epoch": 2.7117165242165244, + "grad_norm": 0.8330100178718567, + "learning_rate": 4.707984329165309e-05, + "loss": 0.7138, + "step": 15231 + }, + { + "epoch": 2.7118945868945867, + "grad_norm": 1.005644679069519, + "learning_rate": 4.706796700900221e-05, + "loss": 1.0089, + "step": 15232 + }, + { + "epoch": 2.7120726495726495, + "grad_norm": 0.8292263746261597, + "learning_rate": 4.705609176344452e-05, + "loss": 0.8323, + "step": 15233 + }, + { + "epoch": 2.7122507122507122, + "grad_norm": 0.860713005065918, + "learning_rate": 4.704421755521281e-05, + "loss": 0.821, + "step": 15234 + }, + { + "epoch": 2.712428774928775, + "grad_norm": 0.8316803574562073, + "learning_rate": 4.703234438453958e-05, + "loss": 0.9181, + "step": 15235 + }, + { + "epoch": 2.7126068376068377, + "grad_norm": 0.7368014454841614, + "learning_rate": 4.70204722516576e-05, + "loss": 0.8206, + "step": 15236 + }, + { + "epoch": 2.7127849002849, + "grad_norm": 1.0202926397323608, + "learning_rate": 4.7008601156799336e-05, + "loss": 0.8101, + "step": 15237 + }, + { + "epoch": 2.712962962962963, + "grad_norm": 0.8069320917129517, + "learning_rate": 4.69967311001975e-05, + "loss": 0.9042, + "step": 15238 + }, + { + "epoch": 2.7131410256410255, + "grad_norm": 0.8426684737205505, + "learning_rate": 4.69848620820846e-05, + "loss": 0.7318, + "step": 15239 + }, + { + "epoch": 2.7133190883190883, + "grad_norm": 0.8863842487335205, + "learning_rate": 4.69729941026932e-05, + "loss": 1.0172, + "step": 15240 + }, + { + "epoch": 2.713497150997151, + "grad_norm": 0.7984182834625244, + "learning_rate": 4.696112716225582e-05, + "loss": 0.8298, + "step": 15241 + }, + { + "epoch": 2.713675213675214, + "grad_norm": 0.8328375220298767, + "learning_rate": 4.6949261261005e-05, + "loss": 0.7663, + "step": 15242 + }, + { + "epoch": 2.7138532763532766, + "grad_norm": 0.9197641015052795, + "learning_rate": 4.693739639917314e-05, + "loss": 0.8951, + "step": 15243 + }, + { + "epoch": 2.7140313390313393, + "grad_norm": 0.7421545386314392, + "learning_rate": 4.692553257699286e-05, + "loss": 0.7235, + "step": 15244 + }, + { + "epoch": 2.7142094017094016, + "grad_norm": 0.8033188581466675, + "learning_rate": 4.691366979469642e-05, + "loss": 0.9693, + "step": 15245 + }, + { + "epoch": 2.7143874643874644, + "grad_norm": 0.8765473365783691, + "learning_rate": 4.6901808052516436e-05, + "loss": 0.8851, + "step": 15246 + }, + { + "epoch": 2.714565527065527, + "grad_norm": 0.8351873755455017, + "learning_rate": 4.688994735068515e-05, + "loss": 1.0156, + "step": 15247 + }, + { + "epoch": 2.71474358974359, + "grad_norm": 0.8569470643997192, + "learning_rate": 4.6878087689435046e-05, + "loss": 0.7149, + "step": 15248 + }, + { + "epoch": 2.714921652421652, + "grad_norm": 0.8334367871284485, + "learning_rate": 4.686622906899847e-05, + "loss": 0.9218, + "step": 15249 + }, + { + "epoch": 2.715099715099715, + "grad_norm": 0.8889651298522949, + "learning_rate": 4.685437148960775e-05, + "loss": 0.8987, + "step": 15250 + }, + { + "epoch": 2.7152777777777777, + "grad_norm": 0.9381657838821411, + "learning_rate": 4.684251495149522e-05, + "loss": 0.7798, + "step": 15251 + }, + { + "epoch": 2.7154558404558404, + "grad_norm": 0.7698730826377869, + "learning_rate": 4.68306594548932e-05, + "loss": 0.8248, + "step": 15252 + }, + { + "epoch": 2.715633903133903, + "grad_norm": 0.8980026245117188, + "learning_rate": 4.681880500003391e-05, + "loss": 1.0156, + "step": 15253 + }, + { + "epoch": 2.715811965811966, + "grad_norm": 0.7872338891029358, + "learning_rate": 4.6806951587149694e-05, + "loss": 0.6389, + "step": 15254 + }, + { + "epoch": 2.7159900284900287, + "grad_norm": 0.8155974745750427, + "learning_rate": 4.6795099216472774e-05, + "loss": 0.9081, + "step": 15255 + }, + { + "epoch": 2.7161680911680914, + "grad_norm": 0.7678217887878418, + "learning_rate": 4.678324788823535e-05, + "loss": 0.6193, + "step": 15256 + }, + { + "epoch": 2.7163461538461537, + "grad_norm": 0.75429767370224, + "learning_rate": 4.6771397602669643e-05, + "loss": 0.9384, + "step": 15257 + }, + { + "epoch": 2.7165242165242165, + "grad_norm": 0.8755250573158264, + "learning_rate": 4.675954836000779e-05, + "loss": 0.8563, + "step": 15258 + }, + { + "epoch": 2.7167022792022792, + "grad_norm": 0.8393009305000305, + "learning_rate": 4.6747700160482053e-05, + "loss": 0.9407, + "step": 15259 + }, + { + "epoch": 2.716880341880342, + "grad_norm": 0.8478221297264099, + "learning_rate": 4.673585300432445e-05, + "loss": 0.7562, + "step": 15260 + }, + { + "epoch": 2.7170584045584043, + "grad_norm": 0.7497259974479675, + "learning_rate": 4.672400689176722e-05, + "loss": 0.8406, + "step": 15261 + }, + { + "epoch": 2.717236467236467, + "grad_norm": 0.9695250391960144, + "learning_rate": 4.671216182304234e-05, + "loss": 0.9505, + "step": 15262 + }, + { + "epoch": 2.71741452991453, + "grad_norm": 0.9375512599945068, + "learning_rate": 4.6700317798382e-05, + "loss": 0.9024, + "step": 15263 + }, + { + "epoch": 2.7175925925925926, + "grad_norm": 0.7930737137794495, + "learning_rate": 4.6688474818018194e-05, + "loss": 0.8416, + "step": 15264 + }, + { + "epoch": 2.7177706552706553, + "grad_norm": 0.9707022309303284, + "learning_rate": 4.667663288218298e-05, + "loss": 1.1172, + "step": 15265 + }, + { + "epoch": 2.717948717948718, + "grad_norm": 0.7616816759109497, + "learning_rate": 4.666479199110838e-05, + "loss": 0.8557, + "step": 15266 + }, + { + "epoch": 2.718126780626781, + "grad_norm": 0.7836055159568787, + "learning_rate": 4.66529521450264e-05, + "loss": 0.7299, + "step": 15267 + }, + { + "epoch": 2.7183048433048436, + "grad_norm": 0.8313519954681396, + "learning_rate": 4.664111334416894e-05, + "loss": 0.8268, + "step": 15268 + }, + { + "epoch": 2.718482905982906, + "grad_norm": 0.9130576252937317, + "learning_rate": 4.662927558876812e-05, + "loss": 0.8913, + "step": 15269 + }, + { + "epoch": 2.7186609686609686, + "grad_norm": 0.8552213907241821, + "learning_rate": 4.661743887905569e-05, + "loss": 0.9396, + "step": 15270 + }, + { + "epoch": 2.7188390313390314, + "grad_norm": 0.7953839898109436, + "learning_rate": 4.660560321526373e-05, + "loss": 0.74, + "step": 15271 + }, + { + "epoch": 2.719017094017094, + "grad_norm": 0.9148657321929932, + "learning_rate": 4.6593768597623974e-05, + "loss": 0.7821, + "step": 15272 + }, + { + "epoch": 2.7191951566951564, + "grad_norm": 0.8587655425071716, + "learning_rate": 4.658193502636843e-05, + "loss": 0.9495, + "step": 15273 + }, + { + "epoch": 2.719373219373219, + "grad_norm": 0.8915669322013855, + "learning_rate": 4.6570102501728896e-05, + "loss": 0.8612, + "step": 15274 + }, + { + "epoch": 2.719551282051282, + "grad_norm": 0.957039475440979, + "learning_rate": 4.655827102393717e-05, + "loss": 0.8506, + "step": 15275 + }, + { + "epoch": 2.7197293447293447, + "grad_norm": 0.7784267067909241, + "learning_rate": 4.654644059322519e-05, + "loss": 0.6864, + "step": 15276 + }, + { + "epoch": 2.7199074074074074, + "grad_norm": 0.9508241415023804, + "learning_rate": 4.65346112098246e-05, + "loss": 1.0097, + "step": 15277 + }, + { + "epoch": 2.72008547008547, + "grad_norm": 0.8316742777824402, + "learning_rate": 4.6522782873967265e-05, + "loss": 0.7444, + "step": 15278 + }, + { + "epoch": 2.720263532763533, + "grad_norm": 0.8781944513320923, + "learning_rate": 4.651095558588491e-05, + "loss": 0.8725, + "step": 15279 + }, + { + "epoch": 2.7204415954415957, + "grad_norm": 0.9407825469970703, + "learning_rate": 4.649912934580927e-05, + "loss": 0.9788, + "step": 15280 + }, + { + "epoch": 2.720619658119658, + "grad_norm": 0.9863289594650269, + "learning_rate": 4.6487304153972045e-05, + "loss": 0.7777, + "step": 15281 + }, + { + "epoch": 2.7207977207977208, + "grad_norm": 0.7580869793891907, + "learning_rate": 4.6475480010604945e-05, + "loss": 0.5835, + "step": 15282 + }, + { + "epoch": 2.7209757834757835, + "grad_norm": 0.7973836660385132, + "learning_rate": 4.646365691593961e-05, + "loss": 0.7633, + "step": 15283 + }, + { + "epoch": 2.7211538461538463, + "grad_norm": 0.8107978701591492, + "learning_rate": 4.645183487020772e-05, + "loss": 0.7149, + "step": 15284 + }, + { + "epoch": 2.7213319088319086, + "grad_norm": 0.8944578170776367, + "learning_rate": 4.644001387364084e-05, + "loss": 0.9227, + "step": 15285 + }, + { + "epoch": 2.7215099715099713, + "grad_norm": 0.7592978477478027, + "learning_rate": 4.642819392647071e-05, + "loss": 0.464, + "step": 15286 + }, + { + "epoch": 2.721688034188034, + "grad_norm": 0.8484344482421875, + "learning_rate": 4.641637502892876e-05, + "loss": 1.0439, + "step": 15287 + }, + { + "epoch": 2.721866096866097, + "grad_norm": 0.8766823410987854, + "learning_rate": 4.640455718124667e-05, + "loss": 0.7561, + "step": 15288 + }, + { + "epoch": 2.7220441595441596, + "grad_norm": 0.8039024472236633, + "learning_rate": 4.639274038365594e-05, + "loss": 0.6774, + "step": 15289 + }, + { + "epoch": 2.7222222222222223, + "grad_norm": 0.8199611902236938, + "learning_rate": 4.63809246363881e-05, + "loss": 0.7721, + "step": 15290 + }, + { + "epoch": 2.722400284900285, + "grad_norm": 0.8209745287895203, + "learning_rate": 4.636910993967467e-05, + "loss": 0.7017, + "step": 15291 + }, + { + "epoch": 2.722578347578348, + "grad_norm": 0.8822476267814636, + "learning_rate": 4.6357296293747075e-05, + "loss": 0.8742, + "step": 15292 + }, + { + "epoch": 2.72275641025641, + "grad_norm": 0.8172603249549866, + "learning_rate": 4.634548369883687e-05, + "loss": 0.8165, + "step": 15293 + }, + { + "epoch": 2.722934472934473, + "grad_norm": 0.8601866960525513, + "learning_rate": 4.633367215517546e-05, + "loss": 0.7961, + "step": 15294 + }, + { + "epoch": 2.7231125356125356, + "grad_norm": 0.9346174001693726, + "learning_rate": 4.632186166299425e-05, + "loss": 0.9229, + "step": 15295 + }, + { + "epoch": 2.7232905982905984, + "grad_norm": 0.8956635594367981, + "learning_rate": 4.631005222252465e-05, + "loss": 0.7886, + "step": 15296 + }, + { + "epoch": 2.7234686609686607, + "grad_norm": 0.8453384637832642, + "learning_rate": 4.629824383399805e-05, + "loss": 0.8513, + "step": 15297 + }, + { + "epoch": 2.7236467236467234, + "grad_norm": 0.8931429982185364, + "learning_rate": 4.628643649764581e-05, + "loss": 1.0195, + "step": 15298 + }, + { + "epoch": 2.723824786324786, + "grad_norm": 0.7326723337173462, + "learning_rate": 4.6274630213699265e-05, + "loss": 0.7616, + "step": 15299 + }, + { + "epoch": 2.724002849002849, + "grad_norm": 0.8572023510932922, + "learning_rate": 4.6262824982389706e-05, + "loss": 0.8266, + "step": 15300 + }, + { + "epoch": 2.7241809116809117, + "grad_norm": 0.7753783464431763, + "learning_rate": 4.625102080394853e-05, + "loss": 0.6907, + "step": 15301 + }, + { + "epoch": 2.7243589743589745, + "grad_norm": 0.8758052587509155, + "learning_rate": 4.623921767860687e-05, + "loss": 0.6369, + "step": 15302 + }, + { + "epoch": 2.724537037037037, + "grad_norm": 0.8508220314979553, + "learning_rate": 4.6227415606596104e-05, + "loss": 0.933, + "step": 15303 + }, + { + "epoch": 2.7247150997151, + "grad_norm": 0.7440072298049927, + "learning_rate": 4.621561458814743e-05, + "loss": 0.7172, + "step": 15304 + }, + { + "epoch": 2.7248931623931623, + "grad_norm": 0.9081870317459106, + "learning_rate": 4.6203814623492046e-05, + "loss": 0.8964, + "step": 15305 + }, + { + "epoch": 2.725071225071225, + "grad_norm": 0.9127907156944275, + "learning_rate": 4.619201571286117e-05, + "loss": 1.0081, + "step": 15306 + }, + { + "epoch": 2.7252492877492878, + "grad_norm": 0.9508554935455322, + "learning_rate": 4.618021785648597e-05, + "loss": 0.94, + "step": 15307 + }, + { + "epoch": 2.7254273504273505, + "grad_norm": 0.8726735711097717, + "learning_rate": 4.616842105459761e-05, + "loss": 0.9284, + "step": 15308 + }, + { + "epoch": 2.7256054131054133, + "grad_norm": 0.9266753792762756, + "learning_rate": 4.6156625307427206e-05, + "loss": 0.9069, + "step": 15309 + }, + { + "epoch": 2.7257834757834756, + "grad_norm": 0.952553391456604, + "learning_rate": 4.614483061520584e-05, + "loss": 0.8604, + "step": 15310 + }, + { + "epoch": 2.7259615384615383, + "grad_norm": 0.7702621817588806, + "learning_rate": 4.613303697816471e-05, + "loss": 0.808, + "step": 15311 + }, + { + "epoch": 2.726139601139601, + "grad_norm": 0.8052653670310974, + "learning_rate": 4.612124439653477e-05, + "loss": 0.8696, + "step": 15312 + }, + { + "epoch": 2.726317663817664, + "grad_norm": 0.8808547854423523, + "learning_rate": 4.610945287054714e-05, + "loss": 0.9595, + "step": 15313 + }, + { + "epoch": 2.7264957264957266, + "grad_norm": 1.0233266353607178, + "learning_rate": 4.609766240043284e-05, + "loss": 1.0691, + "step": 15314 + }, + { + "epoch": 2.7266737891737893, + "grad_norm": 0.8129898309707642, + "learning_rate": 4.6085872986422826e-05, + "loss": 0.9269, + "step": 15315 + }, + { + "epoch": 2.726851851851852, + "grad_norm": 1.2745141983032227, + "learning_rate": 4.607408462874823e-05, + "loss": 0.937, + "step": 15316 + }, + { + "epoch": 2.7270299145299144, + "grad_norm": 0.808274507522583, + "learning_rate": 4.606229732763984e-05, + "loss": 0.789, + "step": 15317 + }, + { + "epoch": 2.727207977207977, + "grad_norm": 0.8849375247955322, + "learning_rate": 4.605051108332875e-05, + "loss": 0.8993, + "step": 15318 + }, + { + "epoch": 2.72738603988604, + "grad_norm": 0.8251593112945557, + "learning_rate": 4.603872589604576e-05, + "loss": 0.9057, + "step": 15319 + }, + { + "epoch": 2.7275641025641026, + "grad_norm": 0.8271582126617432, + "learning_rate": 4.602694176602188e-05, + "loss": 0.8378, + "step": 15320 + }, + { + "epoch": 2.7277421652421654, + "grad_norm": 0.8139070868492126, + "learning_rate": 4.6015158693487956e-05, + "loss": 0.8014, + "step": 15321 + }, + { + "epoch": 2.7279202279202277, + "grad_norm": 0.8873880505561829, + "learning_rate": 4.600337667867486e-05, + "loss": 0.8707, + "step": 15322 + }, + { + "epoch": 2.7280982905982905, + "grad_norm": 0.8616414666175842, + "learning_rate": 4.599159572181342e-05, + "loss": 0.8538, + "step": 15323 + }, + { + "epoch": 2.728276353276353, + "grad_norm": 0.8280995488166809, + "learning_rate": 4.5979815823134466e-05, + "loss": 0.8444, + "step": 15324 + }, + { + "epoch": 2.728454415954416, + "grad_norm": 0.8684375882148743, + "learning_rate": 4.596803698286878e-05, + "loss": 0.7562, + "step": 15325 + }, + { + "epoch": 2.7286324786324787, + "grad_norm": 0.8113002181053162, + "learning_rate": 4.595625920124723e-05, + "loss": 0.8331, + "step": 15326 + }, + { + "epoch": 2.7288105413105415, + "grad_norm": 0.8675588965415955, + "learning_rate": 4.5944482478500436e-05, + "loss": 1.1016, + "step": 15327 + }, + { + "epoch": 2.728988603988604, + "grad_norm": 0.9015034437179565, + "learning_rate": 4.593270681485927e-05, + "loss": 0.9002, + "step": 15328 + }, + { + "epoch": 2.7291666666666665, + "grad_norm": 0.9215324521064758, + "learning_rate": 4.592093221055439e-05, + "loss": 0.9491, + "step": 15329 + }, + { + "epoch": 2.7293447293447293, + "grad_norm": 0.8969921469688416, + "learning_rate": 4.590915866581651e-05, + "loss": 0.8791, + "step": 15330 + }, + { + "epoch": 2.729522792022792, + "grad_norm": 0.9012344479560852, + "learning_rate": 4.5897386180876304e-05, + "loss": 0.9114, + "step": 15331 + }, + { + "epoch": 2.7297008547008548, + "grad_norm": 1.0024429559707642, + "learning_rate": 4.588561475596438e-05, + "loss": 0.8782, + "step": 15332 + }, + { + "epoch": 2.7298789173789175, + "grad_norm": 0.9079484343528748, + "learning_rate": 4.5873844391311496e-05, + "loss": 1.0012, + "step": 15333 + }, + { + "epoch": 2.73005698005698, + "grad_norm": 0.709800660610199, + "learning_rate": 4.5862075087148124e-05, + "loss": 0.7473, + "step": 15334 + }, + { + "epoch": 2.7302350427350426, + "grad_norm": 0.9776272773742676, + "learning_rate": 4.585030684370497e-05, + "loss": 1.1927, + "step": 15335 + }, + { + "epoch": 2.7304131054131053, + "grad_norm": 0.8624512553215027, + "learning_rate": 4.5838539661212565e-05, + "loss": 0.6661, + "step": 15336 + }, + { + "epoch": 2.730591168091168, + "grad_norm": 0.7901379466056824, + "learning_rate": 4.5826773539901456e-05, + "loss": 0.8244, + "step": 15337 + }, + { + "epoch": 2.730769230769231, + "grad_norm": 0.8546316027641296, + "learning_rate": 4.58150084800022e-05, + "loss": 0.8232, + "step": 15338 + }, + { + "epoch": 2.7309472934472936, + "grad_norm": 1.0038648843765259, + "learning_rate": 4.5803244481745275e-05, + "loss": 0.8363, + "step": 15339 + }, + { + "epoch": 2.7311253561253563, + "grad_norm": 0.7757763266563416, + "learning_rate": 4.579148154536117e-05, + "loss": 0.6935, + "step": 15340 + }, + { + "epoch": 2.7313034188034186, + "grad_norm": 0.8671833276748657, + "learning_rate": 4.5779719671080436e-05, + "loss": 0.8453, + "step": 15341 + }, + { + "epoch": 2.7314814814814814, + "grad_norm": 0.8507152795791626, + "learning_rate": 4.57679588591334e-05, + "loss": 0.8142, + "step": 15342 + }, + { + "epoch": 2.731659544159544, + "grad_norm": 0.8205499053001404, + "learning_rate": 4.575619910975062e-05, + "loss": 0.8442, + "step": 15343 + }, + { + "epoch": 2.731837606837607, + "grad_norm": 0.8809645771980286, + "learning_rate": 4.574444042316236e-05, + "loss": 1.024, + "step": 15344 + }, + { + "epoch": 2.7320156695156697, + "grad_norm": 0.825038492679596, + "learning_rate": 4.573268279959912e-05, + "loss": 0.9089, + "step": 15345 + }, + { + "epoch": 2.732193732193732, + "grad_norm": 0.7646815776824951, + "learning_rate": 4.572092623929124e-05, + "loss": 0.9239, + "step": 15346 + }, + { + "epoch": 2.7323717948717947, + "grad_norm": 0.8372252583503723, + "learning_rate": 4.570917074246905e-05, + "loss": 0.8687, + "step": 15347 + }, + { + "epoch": 2.7325498575498575, + "grad_norm": 0.9108861088752747, + "learning_rate": 4.5697416309362885e-05, + "loss": 0.9451, + "step": 15348 + }, + { + "epoch": 2.73272792022792, + "grad_norm": 0.8537670969963074, + "learning_rate": 4.568566294020303e-05, + "loss": 0.7437, + "step": 15349 + }, + { + "epoch": 2.732905982905983, + "grad_norm": 0.8660921454429626, + "learning_rate": 4.5673910635219766e-05, + "loss": 1.0617, + "step": 15350 + }, + { + "epoch": 2.7330840455840457, + "grad_norm": 0.8499423265457153, + "learning_rate": 4.5662159394643424e-05, + "loss": 0.9355, + "step": 15351 + }, + { + "epoch": 2.7332621082621085, + "grad_norm": 0.7776598930358887, + "learning_rate": 4.565040921870413e-05, + "loss": 0.8132, + "step": 15352 + }, + { + "epoch": 2.7334401709401708, + "grad_norm": 0.9197307229042053, + "learning_rate": 4.563866010763219e-05, + "loss": 0.8648, + "step": 15353 + }, + { + "epoch": 2.7336182336182335, + "grad_norm": 0.8543015122413635, + "learning_rate": 4.5626912061657786e-05, + "loss": 0.8453, + "step": 15354 + }, + { + "epoch": 2.7337962962962963, + "grad_norm": 0.9448479413986206, + "learning_rate": 4.561516508101109e-05, + "loss": 0.9515, + "step": 15355 + }, + { + "epoch": 2.733974358974359, + "grad_norm": 0.7448729276657104, + "learning_rate": 4.5603419165922265e-05, + "loss": 0.64, + "step": 15356 + }, + { + "epoch": 2.734152421652422, + "grad_norm": 0.8229237198829651, + "learning_rate": 4.5591674316621405e-05, + "loss": 0.7936, + "step": 15357 + }, + { + "epoch": 2.734330484330484, + "grad_norm": 0.8518769145011902, + "learning_rate": 4.557993053333873e-05, + "loss": 1.1976, + "step": 15358 + }, + { + "epoch": 2.734508547008547, + "grad_norm": 0.8680224418640137, + "learning_rate": 4.55681878163042e-05, + "loss": 0.8223, + "step": 15359 + }, + { + "epoch": 2.7346866096866096, + "grad_norm": 0.8199124336242676, + "learning_rate": 4.555644616574799e-05, + "loss": 0.634, + "step": 15360 + }, + { + "epoch": 2.7348646723646723, + "grad_norm": 0.8262977004051208, + "learning_rate": 4.554470558190013e-05, + "loss": 0.6373, + "step": 15361 + }, + { + "epoch": 2.735042735042735, + "grad_norm": 0.8114070296287537, + "learning_rate": 4.553296606499062e-05, + "loss": 0.6624, + "step": 15362 + }, + { + "epoch": 2.735220797720798, + "grad_norm": 0.9944671392440796, + "learning_rate": 4.552122761524952e-05, + "loss": 0.8246, + "step": 15363 + }, + { + "epoch": 2.7353988603988606, + "grad_norm": 0.8174465298652649, + "learning_rate": 4.550949023290678e-05, + "loss": 0.8431, + "step": 15364 + }, + { + "epoch": 2.7355769230769234, + "grad_norm": 0.8303970694541931, + "learning_rate": 4.5497753918192356e-05, + "loss": 0.417, + "step": 15365 + }, + { + "epoch": 2.7357549857549857, + "grad_norm": 0.8428391218185425, + "learning_rate": 4.548601867133629e-05, + "loss": 0.751, + "step": 15366 + }, + { + "epoch": 2.7359330484330484, + "grad_norm": 0.8769099116325378, + "learning_rate": 4.5474284492568384e-05, + "loss": 0.8984, + "step": 15367 + }, + { + "epoch": 2.736111111111111, + "grad_norm": 0.8389245867729187, + "learning_rate": 4.546255138211867e-05, + "loss": 0.8503, + "step": 15368 + }, + { + "epoch": 2.736289173789174, + "grad_norm": 0.8404824137687683, + "learning_rate": 4.5450819340216896e-05, + "loss": 0.836, + "step": 15369 + }, + { + "epoch": 2.736467236467236, + "grad_norm": 1.0007327795028687, + "learning_rate": 4.543908836709304e-05, + "loss": 0.9746, + "step": 15370 + }, + { + "epoch": 2.736645299145299, + "grad_norm": 0.7373863458633423, + "learning_rate": 4.542735846297691e-05, + "loss": 0.6752, + "step": 15371 + }, + { + "epoch": 2.7368233618233617, + "grad_norm": 0.8973239660263062, + "learning_rate": 4.541562962809829e-05, + "loss": 0.8988, + "step": 15372 + }, + { + "epoch": 2.7370014245014245, + "grad_norm": 0.8576705455780029, + "learning_rate": 4.5403901862687095e-05, + "loss": 0.9279, + "step": 15373 + }, + { + "epoch": 2.7371794871794872, + "grad_norm": 0.7462539076805115, + "learning_rate": 4.539217516697295e-05, + "loss": 0.7228, + "step": 15374 + }, + { + "epoch": 2.73735754985755, + "grad_norm": 0.8082219362258911, + "learning_rate": 4.538044954118573e-05, + "loss": 0.9562, + "step": 15375 + }, + { + "epoch": 2.7375356125356127, + "grad_norm": 0.7067760825157166, + "learning_rate": 4.5368724985555134e-05, + "loss": 0.5306, + "step": 15376 + }, + { + "epoch": 2.7377136752136755, + "grad_norm": 0.9178285598754883, + "learning_rate": 4.535700150031089e-05, + "loss": 0.8087, + "step": 15377 + }, + { + "epoch": 2.737891737891738, + "grad_norm": 0.803240180015564, + "learning_rate": 4.53452790856827e-05, + "loss": 0.7315, + "step": 15378 + }, + { + "epoch": 2.7380698005698005, + "grad_norm": 0.7200242877006531, + "learning_rate": 4.5333557741900226e-05, + "loss": 0.6329, + "step": 15379 + }, + { + "epoch": 2.7382478632478633, + "grad_norm": 0.8744874596595764, + "learning_rate": 4.5321837469193117e-05, + "loss": 0.8279, + "step": 15380 + }, + { + "epoch": 2.738425925925926, + "grad_norm": 0.7736984491348267, + "learning_rate": 4.531011826779103e-05, + "loss": 0.7267, + "step": 15381 + }, + { + "epoch": 2.7386039886039883, + "grad_norm": 0.893189549446106, + "learning_rate": 4.5298400137923527e-05, + "loss": 0.754, + "step": 15382 + }, + { + "epoch": 2.738782051282051, + "grad_norm": 0.7637171149253845, + "learning_rate": 4.5286683079820314e-05, + "loss": 0.6201, + "step": 15383 + }, + { + "epoch": 2.738960113960114, + "grad_norm": 0.8324360847473145, + "learning_rate": 4.527496709371082e-05, + "loss": 0.7617, + "step": 15384 + }, + { + "epoch": 2.7391381766381766, + "grad_norm": 0.7570679783821106, + "learning_rate": 4.52632521798247e-05, + "loss": 0.7275, + "step": 15385 + }, + { + "epoch": 2.7393162393162394, + "grad_norm": 0.7802938222885132, + "learning_rate": 4.525153833839144e-05, + "loss": 0.7423, + "step": 15386 + }, + { + "epoch": 2.739494301994302, + "grad_norm": 0.8645743727684021, + "learning_rate": 4.523982556964056e-05, + "loss": 0.8066, + "step": 15387 + }, + { + "epoch": 2.739672364672365, + "grad_norm": 0.8080064654350281, + "learning_rate": 4.522811387380155e-05, + "loss": 0.7418, + "step": 15388 + }, + { + "epoch": 2.7398504273504276, + "grad_norm": 0.925401508808136, + "learning_rate": 4.521640325110387e-05, + "loss": 0.9622, + "step": 15389 + }, + { + "epoch": 2.74002849002849, + "grad_norm": 0.8898165822029114, + "learning_rate": 4.520469370177696e-05, + "loss": 0.8543, + "step": 15390 + }, + { + "epoch": 2.7402065527065527, + "grad_norm": 0.8610122799873352, + "learning_rate": 4.519298522605021e-05, + "loss": 0.9502, + "step": 15391 + }, + { + "epoch": 2.7403846153846154, + "grad_norm": 0.8111294507980347, + "learning_rate": 4.51812778241531e-05, + "loss": 0.7526, + "step": 15392 + }, + { + "epoch": 2.740562678062678, + "grad_norm": 0.8465895056724548, + "learning_rate": 4.516957149631498e-05, + "loss": 0.9076, + "step": 15393 + }, + { + "epoch": 2.7407407407407405, + "grad_norm": 0.8541668057441711, + "learning_rate": 4.51578662427652e-05, + "loss": 0.8996, + "step": 15394 + }, + { + "epoch": 2.7409188034188032, + "grad_norm": 0.9113210439682007, + "learning_rate": 4.514616206373311e-05, + "loss": 0.9129, + "step": 15395 + }, + { + "epoch": 2.741096866096866, + "grad_norm": 0.7553523182868958, + "learning_rate": 4.513445895944802e-05, + "loss": 0.7265, + "step": 15396 + }, + { + "epoch": 2.7412749287749287, + "grad_norm": 0.8949921131134033, + "learning_rate": 4.5122756930139206e-05, + "loss": 0.9176, + "step": 15397 + }, + { + "epoch": 2.7414529914529915, + "grad_norm": 0.7957020401954651, + "learning_rate": 4.5111055976036044e-05, + "loss": 0.9384, + "step": 15398 + }, + { + "epoch": 2.7416310541310542, + "grad_norm": 0.759608805179596, + "learning_rate": 4.509935609736764e-05, + "loss": 0.7791, + "step": 15399 + }, + { + "epoch": 2.741809116809117, + "grad_norm": 0.927768886089325, + "learning_rate": 4.508765729436335e-05, + "loss": 0.9113, + "step": 15400 + }, + { + "epoch": 2.7419871794871797, + "grad_norm": 0.910513162612915, + "learning_rate": 4.5075959567252335e-05, + "loss": 0.9334, + "step": 15401 + }, + { + "epoch": 2.742165242165242, + "grad_norm": 0.9029644727706909, + "learning_rate": 4.5064262916263814e-05, + "loss": 0.9487, + "step": 15402 + }, + { + "epoch": 2.742343304843305, + "grad_norm": 0.8001708984375, + "learning_rate": 4.505256734162693e-05, + "loss": 0.8447, + "step": 15403 + }, + { + "epoch": 2.7425213675213675, + "grad_norm": 0.8000209927558899, + "learning_rate": 4.504087284357085e-05, + "loss": 0.6764, + "step": 15404 + }, + { + "epoch": 2.7426994301994303, + "grad_norm": 0.7368536591529846, + "learning_rate": 4.5029179422324686e-05, + "loss": 0.563, + "step": 15405 + }, + { + "epoch": 2.7428774928774926, + "grad_norm": 0.9035481214523315, + "learning_rate": 4.501748707811757e-05, + "loss": 0.8165, + "step": 15406 + }, + { + "epoch": 2.7430555555555554, + "grad_norm": 0.7985709309577942, + "learning_rate": 4.500579581117854e-05, + "loss": 1.0773, + "step": 15407 + }, + { + "epoch": 2.743233618233618, + "grad_norm": 0.7867546677589417, + "learning_rate": 4.499410562173678e-05, + "loss": 0.7416, + "step": 15408 + }, + { + "epoch": 2.743411680911681, + "grad_norm": 0.8863609433174133, + "learning_rate": 4.498241651002117e-05, + "loss": 0.8609, + "step": 15409 + }, + { + "epoch": 2.7435897435897436, + "grad_norm": 0.8197270631790161, + "learning_rate": 4.497072847626087e-05, + "loss": 1.0664, + "step": 15410 + }, + { + "epoch": 2.7437678062678064, + "grad_norm": 0.843718409538269, + "learning_rate": 4.495904152068483e-05, + "loss": 0.9831, + "step": 15411 + }, + { + "epoch": 2.743945868945869, + "grad_norm": 0.8311102986335754, + "learning_rate": 4.4947355643521985e-05, + "loss": 0.8035, + "step": 15412 + }, + { + "epoch": 2.744123931623932, + "grad_norm": 0.8396357297897339, + "learning_rate": 4.493567084500143e-05, + "loss": 1.0015, + "step": 15413 + }, + { + "epoch": 2.744301994301994, + "grad_norm": 0.7959007620811462, + "learning_rate": 4.492398712535194e-05, + "loss": 0.8414, + "step": 15414 + }, + { + "epoch": 2.744480056980057, + "grad_norm": 0.7720336318016052, + "learning_rate": 4.491230448480258e-05, + "loss": 0.8185, + "step": 15415 + }, + { + "epoch": 2.7446581196581197, + "grad_norm": 0.7999769449234009, + "learning_rate": 4.4900622923582115e-05, + "loss": 0.6807, + "step": 15416 + }, + { + "epoch": 2.7448361823361824, + "grad_norm": 0.9882165789604187, + "learning_rate": 4.488894244191951e-05, + "loss": 0.975, + "step": 15417 + }, + { + "epoch": 2.745014245014245, + "grad_norm": 0.8275474309921265, + "learning_rate": 4.48772630400436e-05, + "loss": 0.8459, + "step": 15418 + }, + { + "epoch": 2.7451923076923075, + "grad_norm": 0.8468943238258362, + "learning_rate": 4.486558471818322e-05, + "loss": 0.8217, + "step": 15419 + }, + { + "epoch": 2.7453703703703702, + "grad_norm": 0.8845008015632629, + "learning_rate": 4.485390747656717e-05, + "loss": 0.9811, + "step": 15420 + }, + { + "epoch": 2.745548433048433, + "grad_norm": 1.0010331869125366, + "learning_rate": 4.4842231315424255e-05, + "loss": 0.9437, + "step": 15421 + }, + { + "epoch": 2.7457264957264957, + "grad_norm": 0.7468565106391907, + "learning_rate": 4.483055623498319e-05, + "loss": 0.7205, + "step": 15422 + }, + { + "epoch": 2.7459045584045585, + "grad_norm": 0.9002050757408142, + "learning_rate": 4.4818882235472845e-05, + "loss": 0.8812, + "step": 15423 + }, + { + "epoch": 2.7460826210826212, + "grad_norm": 0.8684462904930115, + "learning_rate": 4.48072093171218e-05, + "loss": 0.7929, + "step": 15424 + }, + { + "epoch": 2.746260683760684, + "grad_norm": 0.8685877323150635, + "learning_rate": 4.479553748015891e-05, + "loss": 0.805, + "step": 15425 + }, + { + "epoch": 2.7464387464387463, + "grad_norm": 0.8292124271392822, + "learning_rate": 4.478386672481272e-05, + "loss": 0.9622, + "step": 15426 + }, + { + "epoch": 2.746616809116809, + "grad_norm": 0.8269517421722412, + "learning_rate": 4.477219705131199e-05, + "loss": 0.8011, + "step": 15427 + }, + { + "epoch": 2.746794871794872, + "grad_norm": 0.8913753628730774, + "learning_rate": 4.4760528459885334e-05, + "loss": 0.9794, + "step": 15428 + }, + { + "epoch": 2.7469729344729346, + "grad_norm": 0.8017858266830444, + "learning_rate": 4.474886095076137e-05, + "loss": 0.8593, + "step": 15429 + }, + { + "epoch": 2.7471509971509973, + "grad_norm": 1.0657325983047485, + "learning_rate": 4.47371945241687e-05, + "loss": 0.8043, + "step": 15430 + }, + { + "epoch": 2.7473290598290596, + "grad_norm": 0.8358477354049683, + "learning_rate": 4.472552918033588e-05, + "loss": 0.818, + "step": 15431 + }, + { + "epoch": 2.7475071225071224, + "grad_norm": 1.0436886548995972, + "learning_rate": 4.4713864919491514e-05, + "loss": 0.9246, + "step": 15432 + }, + { + "epoch": 2.747685185185185, + "grad_norm": 0.9838647246360779, + "learning_rate": 4.470220174186413e-05, + "loss": 1.0345, + "step": 15433 + }, + { + "epoch": 2.747863247863248, + "grad_norm": 0.8583347201347351, + "learning_rate": 4.469053964768222e-05, + "loss": 0.6242, + "step": 15434 + }, + { + "epoch": 2.7480413105413106, + "grad_norm": 0.832467794418335, + "learning_rate": 4.4678878637174304e-05, + "loss": 0.7988, + "step": 15435 + }, + { + "epoch": 2.7482193732193734, + "grad_norm": 0.7854242324829102, + "learning_rate": 4.4667218710568825e-05, + "loss": 0.9386, + "step": 15436 + }, + { + "epoch": 2.748397435897436, + "grad_norm": 0.7748091816902161, + "learning_rate": 4.465555986809423e-05, + "loss": 0.7322, + "step": 15437 + }, + { + "epoch": 2.7485754985754984, + "grad_norm": 0.818305492401123, + "learning_rate": 4.464390210997904e-05, + "loss": 0.822, + "step": 15438 + }, + { + "epoch": 2.748753561253561, + "grad_norm": 0.8253993391990662, + "learning_rate": 4.463224543645151e-05, + "loss": 0.7631, + "step": 15439 + }, + { + "epoch": 2.748931623931624, + "grad_norm": 0.804768443107605, + "learning_rate": 4.46205898477402e-05, + "loss": 0.8783, + "step": 15440 + }, + { + "epoch": 2.7491096866096867, + "grad_norm": 0.8612813949584961, + "learning_rate": 4.460893534407332e-05, + "loss": 0.944, + "step": 15441 + }, + { + "epoch": 2.7492877492877494, + "grad_norm": 0.8149600625038147, + "learning_rate": 4.459728192567932e-05, + "loss": 0.8592, + "step": 15442 + }, + { + "epoch": 2.7494658119658117, + "grad_norm": 0.996081531047821, + "learning_rate": 4.4585629592786496e-05, + "loss": 0.8648, + "step": 15443 + }, + { + "epoch": 2.7496438746438745, + "grad_norm": 0.8563137650489807, + "learning_rate": 4.457397834562314e-05, + "loss": 0.6645, + "step": 15444 + }, + { + "epoch": 2.7496438746438745, + "eval_loss": 1.1326396465301514, + "eval_runtime": 24.3757, + "eval_samples_per_second": 42.707, + "eval_steps_per_second": 21.374, + "step": 15444 + }, + { + "epoch": 2.7498219373219372, + "grad_norm": 0.7415599226951599, + "learning_rate": 4.4562328184417547e-05, + "loss": 0.7481, + "step": 15445 + }, + { + "epoch": 2.75, + "grad_norm": 0.8192741274833679, + "learning_rate": 4.455067910939796e-05, + "loss": 0.8367, + "step": 15446 + }, + { + "epoch": 2.7501780626780628, + "grad_norm": 0.8514624834060669, + "learning_rate": 4.4539031120792604e-05, + "loss": 0.7107, + "step": 15447 + }, + { + "epoch": 2.7503561253561255, + "grad_norm": 0.8594211339950562, + "learning_rate": 4.4527384218829796e-05, + "loss": 0.9332, + "step": 15448 + }, + { + "epoch": 2.7505341880341883, + "grad_norm": 0.7828420996665955, + "learning_rate": 4.4515738403737585e-05, + "loss": 0.719, + "step": 15449 + }, + { + "epoch": 2.7507122507122506, + "grad_norm": 0.9195737242698669, + "learning_rate": 4.4504093675744285e-05, + "loss": 0.8581, + "step": 15450 + }, + { + "epoch": 2.7508903133903133, + "grad_norm": 0.8090249300003052, + "learning_rate": 4.449245003507793e-05, + "loss": 0.8012, + "step": 15451 + }, + { + "epoch": 2.751068376068376, + "grad_norm": 0.9179023504257202, + "learning_rate": 4.4480807481966736e-05, + "loss": 0.8968, + "step": 15452 + }, + { + "epoch": 2.751246438746439, + "grad_norm": 0.8140867352485657, + "learning_rate": 4.446916601663879e-05, + "loss": 0.791, + "step": 15453 + }, + { + "epoch": 2.7514245014245016, + "grad_norm": 0.8313645720481873, + "learning_rate": 4.445752563932214e-05, + "loss": 0.8658, + "step": 15454 + }, + { + "epoch": 2.751602564102564, + "grad_norm": 1.0235247611999512, + "learning_rate": 4.444588635024497e-05, + "loss": 0.9383, + "step": 15455 + }, + { + "epoch": 2.7517806267806266, + "grad_norm": 0.8270257711410522, + "learning_rate": 4.443424814963518e-05, + "loss": 0.7666, + "step": 15456 + }, + { + "epoch": 2.7519586894586894, + "grad_norm": 0.8711290955543518, + "learning_rate": 4.442261103772092e-05, + "loss": 1.0819, + "step": 15457 + }, + { + "epoch": 2.752136752136752, + "grad_norm": 0.7851848602294922, + "learning_rate": 4.441097501473013e-05, + "loss": 0.8788, + "step": 15458 + }, + { + "epoch": 2.752314814814815, + "grad_norm": 0.9703593850135803, + "learning_rate": 4.4399340080890816e-05, + "loss": 0.8661, + "step": 15459 + }, + { + "epoch": 2.7524928774928776, + "grad_norm": 0.8575040102005005, + "learning_rate": 4.438770623643093e-05, + "loss": 0.8318, + "step": 15460 + }, + { + "epoch": 2.7526709401709404, + "grad_norm": 0.9393935799598694, + "learning_rate": 4.43760734815784e-05, + "loss": 0.8899, + "step": 15461 + }, + { + "epoch": 2.7528490028490027, + "grad_norm": 0.9310712814331055, + "learning_rate": 4.4364441816561185e-05, + "loss": 0.9519, + "step": 15462 + }, + { + "epoch": 2.7530270655270654, + "grad_norm": 0.8066901564598083, + "learning_rate": 4.435281124160715e-05, + "loss": 0.8102, + "step": 15463 + }, + { + "epoch": 2.753205128205128, + "grad_norm": 0.8681934475898743, + "learning_rate": 4.434118175694415e-05, + "loss": 0.7745, + "step": 15464 + }, + { + "epoch": 2.753383190883191, + "grad_norm": 0.7921330332756042, + "learning_rate": 4.432955336280014e-05, + "loss": 0.8396, + "step": 15465 + }, + { + "epoch": 2.7535612535612537, + "grad_norm": 0.8818981051445007, + "learning_rate": 4.4317926059402816e-05, + "loss": 0.9268, + "step": 15466 + }, + { + "epoch": 2.753739316239316, + "grad_norm": 0.8018338084220886, + "learning_rate": 4.4306299846980096e-05, + "loss": 0.9018, + "step": 15467 + }, + { + "epoch": 2.7539173789173788, + "grad_norm": 0.8704143762588501, + "learning_rate": 4.4294674725759734e-05, + "loss": 0.9517, + "step": 15468 + }, + { + "epoch": 2.7540954415954415, + "grad_norm": 1.1460380554199219, + "learning_rate": 4.4283050695969506e-05, + "loss": 0.7642, + "step": 15469 + }, + { + "epoch": 2.7542735042735043, + "grad_norm": 0.8134510517120361, + "learning_rate": 4.427142775783716e-05, + "loss": 1.0405, + "step": 15470 + }, + { + "epoch": 2.754451566951567, + "grad_norm": 0.8054876327514648, + "learning_rate": 4.425980591159038e-05, + "loss": 0.8929, + "step": 15471 + }, + { + "epoch": 2.7546296296296298, + "grad_norm": 0.8607433438301086, + "learning_rate": 4.4248185157456953e-05, + "loss": 0.9187, + "step": 15472 + }, + { + "epoch": 2.7548076923076925, + "grad_norm": 0.7448357939720154, + "learning_rate": 4.423656549566453e-05, + "loss": 0.7298, + "step": 15473 + }, + { + "epoch": 2.754985754985755, + "grad_norm": 0.9228075742721558, + "learning_rate": 4.422494692644076e-05, + "loss": 0.8704, + "step": 15474 + }, + { + "epoch": 2.7551638176638176, + "grad_norm": 0.8781694173812866, + "learning_rate": 4.421332945001329e-05, + "loss": 0.9555, + "step": 15475 + }, + { + "epoch": 2.7553418803418803, + "grad_norm": 0.8632338643074036, + "learning_rate": 4.420171306660975e-05, + "loss": 0.7322, + "step": 15476 + }, + { + "epoch": 2.755519943019943, + "grad_norm": 0.9808199405670166, + "learning_rate": 4.4190097776457716e-05, + "loss": 0.619, + "step": 15477 + }, + { + "epoch": 2.755698005698006, + "grad_norm": 1.025109052658081, + "learning_rate": 4.41784835797848e-05, + "loss": 1.0617, + "step": 15478 + }, + { + "epoch": 2.755876068376068, + "grad_norm": 0.8132767677307129, + "learning_rate": 4.416687047681849e-05, + "loss": 1.0045, + "step": 15479 + }, + { + "epoch": 2.756054131054131, + "grad_norm": 0.9630453586578369, + "learning_rate": 4.415525846778645e-05, + "loss": 0.9576, + "step": 15480 + }, + { + "epoch": 2.7562321937321936, + "grad_norm": 0.9891922473907471, + "learning_rate": 4.4143647552916034e-05, + "loss": 0.9333, + "step": 15481 + }, + { + "epoch": 2.7564102564102564, + "grad_norm": 0.7974509000778198, + "learning_rate": 4.413203773243486e-05, + "loss": 0.7809, + "step": 15482 + }, + { + "epoch": 2.756588319088319, + "grad_norm": 0.8173473477363586, + "learning_rate": 4.412042900657034e-05, + "loss": 1.1023, + "step": 15483 + }, + { + "epoch": 2.756766381766382, + "grad_norm": 0.8502877950668335, + "learning_rate": 4.410882137554994e-05, + "loss": 0.8705, + "step": 15484 + }, + { + "epoch": 2.7569444444444446, + "grad_norm": 0.8519158959388733, + "learning_rate": 4.4097214839601074e-05, + "loss": 0.8901, + "step": 15485 + }, + { + "epoch": 2.7571225071225074, + "grad_norm": 0.7851125001907349, + "learning_rate": 4.4085609398951164e-05, + "loss": 1.0612, + "step": 15486 + }, + { + "epoch": 2.7573005698005697, + "grad_norm": 0.9585029482841492, + "learning_rate": 4.407400505382758e-05, + "loss": 0.9229, + "step": 15487 + }, + { + "epoch": 2.7574786324786325, + "grad_norm": 0.775071918964386, + "learning_rate": 4.4062401804457686e-05, + "loss": 0.8246, + "step": 15488 + }, + { + "epoch": 2.757656695156695, + "grad_norm": 0.8049488067626953, + "learning_rate": 4.405079965106881e-05, + "loss": 0.9681, + "step": 15489 + }, + { + "epoch": 2.757834757834758, + "grad_norm": 0.9452522993087769, + "learning_rate": 4.4039198593888306e-05, + "loss": 0.7288, + "step": 15490 + }, + { + "epoch": 2.7580128205128203, + "grad_norm": 0.8296085596084595, + "learning_rate": 4.402759863314346e-05, + "loss": 0.8053, + "step": 15491 + }, + { + "epoch": 2.758190883190883, + "grad_norm": 0.8086248636245728, + "learning_rate": 4.4015999769061556e-05, + "loss": 0.8692, + "step": 15492 + }, + { + "epoch": 2.7583689458689458, + "grad_norm": 0.8784860372543335, + "learning_rate": 4.4004402001869836e-05, + "loss": 1.0503, + "step": 15493 + }, + { + "epoch": 2.7585470085470085, + "grad_norm": 0.82901930809021, + "learning_rate": 4.399280533179551e-05, + "loss": 0.8479, + "step": 15494 + }, + { + "epoch": 2.7587250712250713, + "grad_norm": 0.7654509544372559, + "learning_rate": 4.3981209759065875e-05, + "loss": 0.743, + "step": 15495 + }, + { + "epoch": 2.758903133903134, + "grad_norm": 0.8240879774093628, + "learning_rate": 4.3969615283908e-05, + "loss": 0.8303, + "step": 15496 + }, + { + "epoch": 2.7590811965811968, + "grad_norm": 0.9411282539367676, + "learning_rate": 4.3958021906549195e-05, + "loss": 0.8217, + "step": 15497 + }, + { + "epoch": 2.7592592592592595, + "grad_norm": 0.8222329616546631, + "learning_rate": 4.394642962721647e-05, + "loss": 0.9596, + "step": 15498 + }, + { + "epoch": 2.759437321937322, + "grad_norm": 0.8462044596672058, + "learning_rate": 4.393483844613704e-05, + "loss": 0.8029, + "step": 15499 + }, + { + "epoch": 2.7596153846153846, + "grad_norm": 1.0385619401931763, + "learning_rate": 4.392324836353798e-05, + "loss": 1.0352, + "step": 15500 + }, + { + "epoch": 2.7597934472934473, + "grad_norm": 0.9049911499023438, + "learning_rate": 4.3911659379646384e-05, + "loss": 1.0761, + "step": 15501 + }, + { + "epoch": 2.75997150997151, + "grad_norm": 0.8253830671310425, + "learning_rate": 4.390007149468932e-05, + "loss": 0.7693, + "step": 15502 + }, + { + "epoch": 2.7601495726495724, + "grad_norm": 0.7939008474349976, + "learning_rate": 4.388848470889381e-05, + "loss": 0.8847, + "step": 15503 + }, + { + "epoch": 2.760327635327635, + "grad_norm": 1.048941969871521, + "learning_rate": 4.387689902248684e-05, + "loss": 0.7012, + "step": 15504 + }, + { + "epoch": 2.760505698005698, + "grad_norm": 0.8834842443466187, + "learning_rate": 4.386531443569553e-05, + "loss": 0.9561, + "step": 15505 + }, + { + "epoch": 2.7606837606837606, + "grad_norm": 0.9147583842277527, + "learning_rate": 4.385373094874669e-05, + "loss": 0.9736, + "step": 15506 + }, + { + "epoch": 2.7608618233618234, + "grad_norm": 0.6820386648178101, + "learning_rate": 4.38421485618674e-05, + "loss": 0.4449, + "step": 15507 + }, + { + "epoch": 2.761039886039886, + "grad_norm": 0.9519942402839661, + "learning_rate": 4.383056727528455e-05, + "loss": 1.0385, + "step": 15508 + }, + { + "epoch": 2.761217948717949, + "grad_norm": 0.8701474070549011, + "learning_rate": 4.381898708922505e-05, + "loss": 0.9896, + "step": 15509 + }, + { + "epoch": 2.7613960113960117, + "grad_norm": 0.8756018877029419, + "learning_rate": 4.38074080039158e-05, + "loss": 0.9378, + "step": 15510 + }, + { + "epoch": 2.761574074074074, + "grad_norm": 0.8670514225959778, + "learning_rate": 4.379583001958362e-05, + "loss": 0.9175, + "step": 15511 + }, + { + "epoch": 2.7617521367521367, + "grad_norm": 0.8227131366729736, + "learning_rate": 4.378425313645547e-05, + "loss": 0.7864, + "step": 15512 + }, + { + "epoch": 2.7619301994301995, + "grad_norm": 0.9121497273445129, + "learning_rate": 4.377267735475802e-05, + "loss": 0.877, + "step": 15513 + }, + { + "epoch": 2.762108262108262, + "grad_norm": 0.8347102999687195, + "learning_rate": 4.3761102674718205e-05, + "loss": 0.8223, + "step": 15514 + }, + { + "epoch": 2.7622863247863245, + "grad_norm": 0.8657951951026917, + "learning_rate": 4.374952909656275e-05, + "loss": 0.7117, + "step": 15515 + }, + { + "epoch": 2.7624643874643873, + "grad_norm": 0.8934728503227234, + "learning_rate": 4.3737956620518414e-05, + "loss": 0.84, + "step": 15516 + }, + { + "epoch": 2.76264245014245, + "grad_norm": 0.7592045068740845, + "learning_rate": 4.3726385246811964e-05, + "loss": 0.7856, + "step": 15517 + }, + { + "epoch": 2.7628205128205128, + "grad_norm": 0.8480674028396606, + "learning_rate": 4.371481497567008e-05, + "loss": 0.7805, + "step": 15518 + }, + { + "epoch": 2.7629985754985755, + "grad_norm": 1.0231767892837524, + "learning_rate": 4.3703245807319437e-05, + "loss": 1.1517, + "step": 15519 + }, + { + "epoch": 2.7631766381766383, + "grad_norm": 0.9852092862129211, + "learning_rate": 4.369167774198684e-05, + "loss": 0.8735, + "step": 15520 + }, + { + "epoch": 2.763354700854701, + "grad_norm": 0.8751610517501831, + "learning_rate": 4.368011077989875e-05, + "loss": 0.9975, + "step": 15521 + }, + { + "epoch": 2.763532763532764, + "grad_norm": 0.8397828340530396, + "learning_rate": 4.3668544921281976e-05, + "loss": 1.0675, + "step": 15522 + }, + { + "epoch": 2.763710826210826, + "grad_norm": 0.7040372490882874, + "learning_rate": 4.3656980166362974e-05, + "loss": 0.7123, + "step": 15523 + }, + { + "epoch": 2.763888888888889, + "grad_norm": 1.0610599517822266, + "learning_rate": 4.364541651536844e-05, + "loss": 1.0854, + "step": 15524 + }, + { + "epoch": 2.7640669515669516, + "grad_norm": 0.78865647315979, + "learning_rate": 4.363385396852491e-05, + "loss": 0.924, + "step": 15525 + }, + { + "epoch": 2.7642450142450143, + "grad_norm": 0.87164705991745, + "learning_rate": 4.362229252605891e-05, + "loss": 0.9739, + "step": 15526 + }, + { + "epoch": 2.7644230769230766, + "grad_norm": 0.9362281560897827, + "learning_rate": 4.361073218819698e-05, + "loss": 0.7751, + "step": 15527 + }, + { + "epoch": 2.7646011396011394, + "grad_norm": 0.7944566011428833, + "learning_rate": 4.3599172955165605e-05, + "loss": 0.8913, + "step": 15528 + }, + { + "epoch": 2.764779202279202, + "grad_norm": 0.9346068501472473, + "learning_rate": 4.358761482719125e-05, + "loss": 0.8286, + "step": 15529 + }, + { + "epoch": 2.764957264957265, + "grad_norm": 0.8570913076400757, + "learning_rate": 4.3576057804500414e-05, + "loss": 1.0334, + "step": 15530 + }, + { + "epoch": 2.7651353276353277, + "grad_norm": 0.801908552646637, + "learning_rate": 4.356450188731953e-05, + "loss": 0.9021, + "step": 15531 + }, + { + "epoch": 2.7653133903133904, + "grad_norm": 0.848849892616272, + "learning_rate": 4.355294707587499e-05, + "loss": 0.9132, + "step": 15532 + }, + { + "epoch": 2.765491452991453, + "grad_norm": 0.7961751818656921, + "learning_rate": 4.35413933703932e-05, + "loss": 0.841, + "step": 15533 + }, + { + "epoch": 2.765669515669516, + "grad_norm": 0.8609708547592163, + "learning_rate": 4.352984077110052e-05, + "loss": 0.8176, + "step": 15534 + }, + { + "epoch": 2.765847578347578, + "grad_norm": 0.8779369592666626, + "learning_rate": 4.35182892782233e-05, + "loss": 0.8252, + "step": 15535 + }, + { + "epoch": 2.766025641025641, + "grad_norm": 0.7878577709197998, + "learning_rate": 4.3506738891987844e-05, + "loss": 0.7498, + "step": 15536 + }, + { + "epoch": 2.7662037037037037, + "grad_norm": 0.9531580805778503, + "learning_rate": 4.3495189612620557e-05, + "loss": 0.8438, + "step": 15537 + }, + { + "epoch": 2.7663817663817665, + "grad_norm": 0.7791294455528259, + "learning_rate": 4.3483641440347564e-05, + "loss": 0.9188, + "step": 15538 + }, + { + "epoch": 2.7665598290598292, + "grad_norm": 0.8683488965034485, + "learning_rate": 4.347209437539527e-05, + "loss": 1.0422, + "step": 15539 + }, + { + "epoch": 2.7667378917378915, + "grad_norm": 0.8904309272766113, + "learning_rate": 4.346054841798984e-05, + "loss": 0.8072, + "step": 15540 + }, + { + "epoch": 2.7669159544159543, + "grad_norm": 0.7409844398498535, + "learning_rate": 4.344900356835753e-05, + "loss": 0.7179, + "step": 15541 + }, + { + "epoch": 2.767094017094017, + "grad_norm": 0.9663724899291992, + "learning_rate": 4.343745982672451e-05, + "loss": 1.0568, + "step": 15542 + }, + { + "epoch": 2.76727207977208, + "grad_norm": 0.8481591939926147, + "learning_rate": 4.342591719331698e-05, + "loss": 0.8678, + "step": 15543 + }, + { + "epoch": 2.7674501424501425, + "grad_norm": 0.7301938533782959, + "learning_rate": 4.341437566836103e-05, + "loss": 0.6357, + "step": 15544 + }, + { + "epoch": 2.7676282051282053, + "grad_norm": 0.8628479242324829, + "learning_rate": 4.340283525208292e-05, + "loss": 0.9622, + "step": 15545 + }, + { + "epoch": 2.767806267806268, + "grad_norm": 0.953744113445282, + "learning_rate": 4.339129594470861e-05, + "loss": 0.683, + "step": 15546 + }, + { + "epoch": 2.7679843304843303, + "grad_norm": 0.7589353322982788, + "learning_rate": 4.3379757746464336e-05, + "loss": 0.8343, + "step": 15547 + }, + { + "epoch": 2.768162393162393, + "grad_norm": 0.8304651379585266, + "learning_rate": 4.336822065757601e-05, + "loss": 0.9084, + "step": 15548 + }, + { + "epoch": 2.768340455840456, + "grad_norm": 0.8092817068099976, + "learning_rate": 4.33566846782698e-05, + "loss": 0.8371, + "step": 15549 + }, + { + "epoch": 2.7685185185185186, + "grad_norm": 0.8983978033065796, + "learning_rate": 4.334514980877169e-05, + "loss": 0.9631, + "step": 15550 + }, + { + "epoch": 2.7686965811965814, + "grad_norm": 0.766621470451355, + "learning_rate": 4.3333616049307636e-05, + "loss": 0.768, + "step": 15551 + }, + { + "epoch": 2.7688746438746437, + "grad_norm": 0.8753345012664795, + "learning_rate": 4.332208340010374e-05, + "loss": 0.8854, + "step": 15552 + }, + { + "epoch": 2.7690527065527064, + "grad_norm": 0.8396589756011963, + "learning_rate": 4.331055186138581e-05, + "loss": 0.8322, + "step": 15553 + }, + { + "epoch": 2.769230769230769, + "grad_norm": 0.8134872317314148, + "learning_rate": 4.3299021433379885e-05, + "loss": 0.945, + "step": 15554 + }, + { + "epoch": 2.769408831908832, + "grad_norm": 0.8712667226791382, + "learning_rate": 4.3287492116311854e-05, + "loss": 0.8487, + "step": 15555 + }, + { + "epoch": 2.7695868945868947, + "grad_norm": 0.8938018083572388, + "learning_rate": 4.32759639104076e-05, + "loss": 0.854, + "step": 15556 + }, + { + "epoch": 2.7697649572649574, + "grad_norm": 0.8213987946510315, + "learning_rate": 4.3264436815893005e-05, + "loss": 0.9055, + "step": 15557 + }, + { + "epoch": 2.76994301994302, + "grad_norm": 0.9587214589118958, + "learning_rate": 4.3252910832993906e-05, + "loss": 0.9664, + "step": 15558 + }, + { + "epoch": 2.7701210826210825, + "grad_norm": 0.8746159076690674, + "learning_rate": 4.3241385961936146e-05, + "loss": 0.8394, + "step": 15559 + }, + { + "epoch": 2.7702991452991452, + "grad_norm": 0.8395819067955017, + "learning_rate": 4.3229862202945517e-05, + "loss": 0.7991, + "step": 15560 + }, + { + "epoch": 2.770477207977208, + "grad_norm": 0.856765627861023, + "learning_rate": 4.321833955624777e-05, + "loss": 0.8106, + "step": 15561 + }, + { + "epoch": 2.7706552706552707, + "grad_norm": 0.8630124926567078, + "learning_rate": 4.3206818022068776e-05, + "loss": 0.7293, + "step": 15562 + }, + { + "epoch": 2.7708333333333335, + "grad_norm": 0.8196776509284973, + "learning_rate": 4.319529760063414e-05, + "loss": 0.8831, + "step": 15563 + }, + { + "epoch": 2.771011396011396, + "grad_norm": 0.8283860683441162, + "learning_rate": 4.3183778292169674e-05, + "loss": 0.9249, + "step": 15564 + }, + { + "epoch": 2.7711894586894585, + "grad_norm": 0.8983619809150696, + "learning_rate": 4.3172260096901054e-05, + "loss": 1.2334, + "step": 15565 + }, + { + "epoch": 2.7713675213675213, + "grad_norm": 0.8437079191207886, + "learning_rate": 4.316074301505395e-05, + "loss": 0.771, + "step": 15566 + }, + { + "epoch": 2.771545584045584, + "grad_norm": 0.9565808773040771, + "learning_rate": 4.314922704685401e-05, + "loss": 0.8927, + "step": 15567 + }, + { + "epoch": 2.771723646723647, + "grad_norm": 0.7943497896194458, + "learning_rate": 4.313771219252687e-05, + "loss": 0.656, + "step": 15568 + }, + { + "epoch": 2.7719017094017095, + "grad_norm": 0.862404465675354, + "learning_rate": 4.3126198452298126e-05, + "loss": 0.8783, + "step": 15569 + }, + { + "epoch": 2.7720797720797723, + "grad_norm": 0.7928122878074646, + "learning_rate": 4.3114685826393365e-05, + "loss": 0.9799, + "step": 15570 + }, + { + "epoch": 2.7722578347578346, + "grad_norm": 0.8270733952522278, + "learning_rate": 4.3103174315038184e-05, + "loss": 0.7878, + "step": 15571 + }, + { + "epoch": 2.7724358974358974, + "grad_norm": 0.8223987817764282, + "learning_rate": 4.309166391845811e-05, + "loss": 0.7222, + "step": 15572 + }, + { + "epoch": 2.77261396011396, + "grad_norm": 0.8159852027893066, + "learning_rate": 4.3080154636878675e-05, + "loss": 0.8082, + "step": 15573 + }, + { + "epoch": 2.772792022792023, + "grad_norm": 0.882792055606842, + "learning_rate": 4.306864647052537e-05, + "loss": 1.0659, + "step": 15574 + }, + { + "epoch": 2.7729700854700856, + "grad_norm": 0.8734562993049622, + "learning_rate": 4.305713941962366e-05, + "loss": 0.9301, + "step": 15575 + }, + { + "epoch": 2.773148148148148, + "grad_norm": 0.8544983267784119, + "learning_rate": 4.304563348439898e-05, + "loss": 0.7442, + "step": 15576 + }, + { + "epoch": 2.7733262108262107, + "grad_norm": 0.9045799374580383, + "learning_rate": 4.303412866507689e-05, + "loss": 1.1023, + "step": 15577 + }, + { + "epoch": 2.7735042735042734, + "grad_norm": 0.8132993578910828, + "learning_rate": 4.3022624961882615e-05, + "loss": 0.7032, + "step": 15578 + }, + { + "epoch": 2.773682336182336, + "grad_norm": 0.7072446942329407, + "learning_rate": 4.30111223750417e-05, + "loss": 0.7176, + "step": 15579 + }, + { + "epoch": 2.773860398860399, + "grad_norm": 0.8212466239929199, + "learning_rate": 4.299962090477945e-05, + "loss": 0.7664, + "step": 15580 + }, + { + "epoch": 2.7740384615384617, + "grad_norm": 0.7781338095664978, + "learning_rate": 4.298812055132122e-05, + "loss": 0.7439, + "step": 15581 + }, + { + "epoch": 2.7742165242165244, + "grad_norm": 0.9289973378181458, + "learning_rate": 4.297662131489234e-05, + "loss": 0.9504, + "step": 15582 + }, + { + "epoch": 2.7743945868945867, + "grad_norm": 0.8571373224258423, + "learning_rate": 4.2965123195718105e-05, + "loss": 0.9959, + "step": 15583 + }, + { + "epoch": 2.7745726495726495, + "grad_norm": 0.8670883774757385, + "learning_rate": 4.29536261940238e-05, + "loss": 0.8207, + "step": 15584 + }, + { + "epoch": 2.7747507122507122, + "grad_norm": 0.8684807419776917, + "learning_rate": 4.294213031003469e-05, + "loss": 0.7508, + "step": 15585 + }, + { + "epoch": 2.774928774928775, + "grad_norm": 0.7746252417564392, + "learning_rate": 4.293063554397597e-05, + "loss": 0.909, + "step": 15586 + }, + { + "epoch": 2.7751068376068377, + "grad_norm": 0.8363521099090576, + "learning_rate": 4.291914189607297e-05, + "loss": 0.8564, + "step": 15587 + }, + { + "epoch": 2.7752849002849, + "grad_norm": 0.8843217492103577, + "learning_rate": 4.2907649366550726e-05, + "loss": 0.8187, + "step": 15588 + }, + { + "epoch": 2.775462962962963, + "grad_norm": 0.9330897331237793, + "learning_rate": 4.2896157955634545e-05, + "loss": 0.8179, + "step": 15589 + }, + { + "epoch": 2.7756410256410255, + "grad_norm": 0.7950356602668762, + "learning_rate": 4.288466766354953e-05, + "loss": 0.7091, + "step": 15590 + }, + { + "epoch": 2.7758190883190883, + "grad_norm": 0.9085933566093445, + "learning_rate": 4.287317849052075e-05, + "loss": 0.8015, + "step": 15591 + }, + { + "epoch": 2.775997150997151, + "grad_norm": 0.9285191893577576, + "learning_rate": 4.286169043677345e-05, + "loss": 0.9967, + "step": 15592 + }, + { + "epoch": 2.776175213675214, + "grad_norm": 0.8202041387557983, + "learning_rate": 4.285020350253256e-05, + "loss": 0.9286, + "step": 15593 + }, + { + "epoch": 2.7763532763532766, + "grad_norm": 1.0619434118270874, + "learning_rate": 4.283871768802328e-05, + "loss": 0.7863, + "step": 15594 + }, + { + "epoch": 2.7765313390313393, + "grad_norm": 0.8250051140785217, + "learning_rate": 4.282723299347052e-05, + "loss": 0.8531, + "step": 15595 + }, + { + "epoch": 2.7767094017094016, + "grad_norm": 0.8794218897819519, + "learning_rate": 4.281574941909939e-05, + "loss": 0.906, + "step": 15596 + }, + { + "epoch": 2.7768874643874644, + "grad_norm": 0.7725922465324402, + "learning_rate": 4.2804266965134866e-05, + "loss": 0.9084, + "step": 15597 + }, + { + "epoch": 2.777065527065527, + "grad_norm": 0.7845144867897034, + "learning_rate": 4.279278563180192e-05, + "loss": 0.9768, + "step": 15598 + }, + { + "epoch": 2.77724358974359, + "grad_norm": 0.9395498633384705, + "learning_rate": 4.27813054193255e-05, + "loss": 0.9055, + "step": 15599 + }, + { + "epoch": 2.777421652421652, + "grad_norm": 0.8043427467346191, + "learning_rate": 4.276982632793054e-05, + "loss": 0.8244, + "step": 15600 + }, + { + "epoch": 2.777599715099715, + "grad_norm": 0.7874096632003784, + "learning_rate": 4.27583483578419e-05, + "loss": 0.8861, + "step": 15601 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 0.8874611258506775, + "learning_rate": 4.27468715092846e-05, + "loss": 1.0457, + "step": 15602 + }, + { + "epoch": 2.7779558404558404, + "grad_norm": 1.0025757551193237, + "learning_rate": 4.273539578248334e-05, + "loss": 1.1114, + "step": 15603 + }, + { + "epoch": 2.778133903133903, + "grad_norm": 0.9982876777648926, + "learning_rate": 4.272392117766313e-05, + "loss": 0.9142, + "step": 15604 + }, + { + "epoch": 2.778311965811966, + "grad_norm": 0.8762221932411194, + "learning_rate": 4.2712447695048616e-05, + "loss": 1.0114, + "step": 15605 + }, + { + "epoch": 2.7784900284900287, + "grad_norm": 0.9136927723884583, + "learning_rate": 4.2700975334864726e-05, + "loss": 0.8224, + "step": 15606 + }, + { + "epoch": 2.7786680911680914, + "grad_norm": 0.8845604062080383, + "learning_rate": 4.2689504097336184e-05, + "loss": 0.8135, + "step": 15607 + }, + { + "epoch": 2.7788461538461537, + "grad_norm": 0.8584510087966919, + "learning_rate": 4.267803398268777e-05, + "loss": 0.8928, + "step": 15608 + }, + { + "epoch": 2.7790242165242165, + "grad_norm": 0.711402177810669, + "learning_rate": 4.266656499114421e-05, + "loss": 0.792, + "step": 15609 + }, + { + "epoch": 2.7792022792022792, + "grad_norm": 0.9480760097503662, + "learning_rate": 4.2655097122930165e-05, + "loss": 0.8707, + "step": 15610 + }, + { + "epoch": 2.779380341880342, + "grad_norm": 0.863855242729187, + "learning_rate": 4.264363037827041e-05, + "loss": 1.0114, + "step": 15611 + }, + { + "epoch": 2.7795584045584043, + "grad_norm": 0.8010865449905396, + "learning_rate": 4.2632164757389556e-05, + "loss": 0.9221, + "step": 15612 + }, + { + "epoch": 2.779736467236467, + "grad_norm": 0.7950930595397949, + "learning_rate": 4.262070026051227e-05, + "loss": 0.6951, + "step": 15613 + }, + { + "epoch": 2.77991452991453, + "grad_norm": 0.8252870440483093, + "learning_rate": 4.260923688786317e-05, + "loss": 0.6976, + "step": 15614 + }, + { + "epoch": 2.7800925925925926, + "grad_norm": 0.7855920791625977, + "learning_rate": 4.259777463966686e-05, + "loss": 0.8277, + "step": 15615 + }, + { + "epoch": 2.7802706552706553, + "grad_norm": 0.8783130645751953, + "learning_rate": 4.258631351614786e-05, + "loss": 0.6995, + "step": 15616 + }, + { + "epoch": 2.780448717948718, + "grad_norm": 0.8904485106468201, + "learning_rate": 4.257485351753085e-05, + "loss": 0.8226, + "step": 15617 + }, + { + "epoch": 2.780626780626781, + "grad_norm": 0.8761011958122253, + "learning_rate": 4.2563394644040244e-05, + "loss": 0.9187, + "step": 15618 + }, + { + "epoch": 2.7808048433048436, + "grad_norm": 0.897404670715332, + "learning_rate": 4.255193689590067e-05, + "loss": 1.0234, + "step": 15619 + }, + { + "epoch": 2.780982905982906, + "grad_norm": 0.8966960906982422, + "learning_rate": 4.254048027333648e-05, + "loss": 0.9, + "step": 15620 + }, + { + "epoch": 2.7811609686609686, + "grad_norm": 0.7506237030029297, + "learning_rate": 4.2529024776572245e-05, + "loss": 0.8939, + "step": 15621 + }, + { + "epoch": 2.7813390313390314, + "grad_norm": 0.8073886036872864, + "learning_rate": 4.2517570405832396e-05, + "loss": 0.7779, + "step": 15622 + }, + { + "epoch": 2.781517094017094, + "grad_norm": 0.7928911447525024, + "learning_rate": 4.250611716134134e-05, + "loss": 0.8278, + "step": 15623 + }, + { + "epoch": 2.7816951566951564, + "grad_norm": 0.7301982045173645, + "learning_rate": 4.249466504332349e-05, + "loss": 0.7515, + "step": 15624 + }, + { + "epoch": 2.781873219373219, + "grad_norm": 0.8215289115905762, + "learning_rate": 4.248321405200322e-05, + "loss": 0.9752, + "step": 15625 + }, + { + "epoch": 2.782051282051282, + "grad_norm": 0.8281431198120117, + "learning_rate": 4.247176418760486e-05, + "loss": 0.9625, + "step": 15626 + }, + { + "epoch": 2.7822293447293447, + "grad_norm": 0.9202759265899658, + "learning_rate": 4.246031545035283e-05, + "loss": 0.8757, + "step": 15627 + }, + { + "epoch": 2.7824074074074074, + "grad_norm": 0.8628471493721008, + "learning_rate": 4.244886784047133e-05, + "loss": 0.7626, + "step": 15628 + }, + { + "epoch": 2.78258547008547, + "grad_norm": 0.9345491528511047, + "learning_rate": 4.2437421358184747e-05, + "loss": 0.8714, + "step": 15629 + }, + { + "epoch": 2.782763532763533, + "grad_norm": 0.893713116645813, + "learning_rate": 4.2425976003717314e-05, + "loss": 0.9953, + "step": 15630 + }, + { + "epoch": 2.7829415954415957, + "grad_norm": 0.8794371485710144, + "learning_rate": 4.2414531777293286e-05, + "loss": 0.7899, + "step": 15631 + }, + { + "epoch": 2.783119658119658, + "grad_norm": 0.9003345370292664, + "learning_rate": 4.240308867913688e-05, + "loss": 0.9607, + "step": 15632 + }, + { + "epoch": 2.7832977207977208, + "grad_norm": 0.8352270126342773, + "learning_rate": 4.239164670947228e-05, + "loss": 1.0315, + "step": 15633 + }, + { + "epoch": 2.7834757834757835, + "grad_norm": 0.825252890586853, + "learning_rate": 4.238020586852375e-05, + "loss": 1.0493, + "step": 15634 + }, + { + "epoch": 2.7836538461538463, + "grad_norm": 1.1185758113861084, + "learning_rate": 4.2368766156515324e-05, + "loss": 0.8962, + "step": 15635 + }, + { + "epoch": 2.7838319088319086, + "grad_norm": 0.840336799621582, + "learning_rate": 4.235732757367125e-05, + "loss": 0.8289, + "step": 15636 + }, + { + "epoch": 2.7840099715099713, + "grad_norm": 0.9251887202262878, + "learning_rate": 4.2345890120215595e-05, + "loss": 0.9306, + "step": 15637 + }, + { + "epoch": 2.784188034188034, + "grad_norm": 0.9645969867706299, + "learning_rate": 4.233445379637244e-05, + "loss": 0.8453, + "step": 15638 + }, + { + "epoch": 2.784366096866097, + "grad_norm": 0.9010009765625, + "learning_rate": 4.232301860236589e-05, + "loss": 0.9796, + "step": 15639 + }, + { + "epoch": 2.7845441595441596, + "grad_norm": 0.920427143573761, + "learning_rate": 4.231158453841998e-05, + "loss": 0.7905, + "step": 15640 + }, + { + "epoch": 2.7847222222222223, + "grad_norm": 0.8292316794395447, + "learning_rate": 4.2300151604758734e-05, + "loss": 0.938, + "step": 15641 + }, + { + "epoch": 2.784900284900285, + "grad_norm": 0.8550885319709778, + "learning_rate": 4.228871980160615e-05, + "loss": 0.7728, + "step": 15642 + }, + { + "epoch": 2.785078347578348, + "grad_norm": 0.8785567283630371, + "learning_rate": 4.227728912918617e-05, + "loss": 1.0367, + "step": 15643 + }, + { + "epoch": 2.78525641025641, + "grad_norm": 0.8732814788818359, + "learning_rate": 4.226585958772289e-05, + "loss": 0.9914, + "step": 15644 + }, + { + "epoch": 2.785434472934473, + "grad_norm": 0.9473167061805725, + "learning_rate": 4.225443117744008e-05, + "loss": 1.1311, + "step": 15645 + }, + { + "epoch": 2.7856125356125356, + "grad_norm": 0.8819913864135742, + "learning_rate": 4.224300389856177e-05, + "loss": 0.8798, + "step": 15646 + }, + { + "epoch": 2.7857905982905984, + "grad_norm": 0.835367739200592, + "learning_rate": 4.223157775131182e-05, + "loss": 0.6977, + "step": 15647 + }, + { + "epoch": 2.7859686609686607, + "grad_norm": 0.8122659921646118, + "learning_rate": 4.222015273591411e-05, + "loss": 0.9656, + "step": 15648 + }, + { + "epoch": 2.7861467236467234, + "grad_norm": 0.8085313439369202, + "learning_rate": 4.220872885259247e-05, + "loss": 0.7456, + "step": 15649 + }, + { + "epoch": 2.786324786324786, + "grad_norm": 0.681515097618103, + "learning_rate": 4.21973061015707e-05, + "loss": 0.5008, + "step": 15650 + }, + { + "epoch": 2.786502849002849, + "grad_norm": 0.8021831512451172, + "learning_rate": 4.2185884483072676e-05, + "loss": 0.8954, + "step": 15651 + }, + { + "epoch": 2.7866809116809117, + "grad_norm": 0.9254723787307739, + "learning_rate": 4.217446399732216e-05, + "loss": 0.7855, + "step": 15652 + }, + { + "epoch": 2.7868589743589745, + "grad_norm": 0.8415037989616394, + "learning_rate": 4.2163044644542894e-05, + "loss": 0.8835, + "step": 15653 + }, + { + "epoch": 2.787037037037037, + "grad_norm": 0.9031959772109985, + "learning_rate": 4.2151626424958614e-05, + "loss": 1.0048, + "step": 15654 + }, + { + "epoch": 2.7872150997151, + "grad_norm": Infinity, + "learning_rate": 4.2151626424958614e-05, + "loss": 0.9344, + "step": 15655 + }, + { + "epoch": 2.7873931623931623, + "grad_norm": 0.815680742263794, + "learning_rate": 4.214020933879306e-05, + "loss": 0.6897, + "step": 15656 + }, + { + "epoch": 2.787571225071225, + "grad_norm": 0.9080044627189636, + "learning_rate": 4.212879338626989e-05, + "loss": 1.0366, + "step": 15657 + }, + { + "epoch": 2.7877492877492878, + "grad_norm": 0.8387414813041687, + "learning_rate": 4.211737856761281e-05, + "loss": 0.9255, + "step": 15658 + }, + { + "epoch": 2.7879273504273505, + "grad_norm": 0.9269571304321289, + "learning_rate": 4.210596488304542e-05, + "loss": 0.8971, + "step": 15659 + }, + { + "epoch": 2.7881054131054133, + "grad_norm": 0.7987017035484314, + "learning_rate": 4.2094552332791456e-05, + "loss": 0.8293, + "step": 15660 + }, + { + "epoch": 2.7882834757834756, + "grad_norm": 0.8481683731079102, + "learning_rate": 4.208314091707437e-05, + "loss": 0.9159, + "step": 15661 + }, + { + "epoch": 2.7884615384615383, + "grad_norm": 0.944736897945404, + "learning_rate": 4.207173063611788e-05, + "loss": 0.9398, + "step": 15662 + }, + { + "epoch": 2.788639601139601, + "grad_norm": 0.8471882343292236, + "learning_rate": 4.206032149014547e-05, + "loss": 0.7534, + "step": 15663 + }, + { + "epoch": 2.788817663817664, + "grad_norm": 0.8500807881355286, + "learning_rate": 4.2048913479380714e-05, + "loss": 0.8874, + "step": 15664 + }, + { + "epoch": 2.7889957264957266, + "grad_norm": 0.7949451804161072, + "learning_rate": 4.2037506604047115e-05, + "loss": 0.8691, + "step": 15665 + }, + { + "epoch": 2.7891737891737893, + "grad_norm": 0.8587945103645325, + "learning_rate": 4.202610086436817e-05, + "loss": 0.8288, + "step": 15666 + }, + { + "epoch": 2.789351851851852, + "grad_norm": 0.9155020117759705, + "learning_rate": 4.201469626056734e-05, + "loss": 0.8384, + "step": 15667 + }, + { + "epoch": 2.7895299145299144, + "grad_norm": 0.9402222037315369, + "learning_rate": 4.200329279286809e-05, + "loss": 0.8239, + "step": 15668 + }, + { + "epoch": 2.789707977207977, + "grad_norm": 0.9111437201499939, + "learning_rate": 4.19918904614938e-05, + "loss": 0.9251, + "step": 15669 + }, + { + "epoch": 2.78988603988604, + "grad_norm": 0.9434856176376343, + "learning_rate": 4.198048926666795e-05, + "loss": 1.0517, + "step": 15670 + }, + { + "epoch": 2.7900641025641026, + "grad_norm": 0.9518313407897949, + "learning_rate": 4.1969089208613896e-05, + "loss": 0.8893, + "step": 15671 + }, + { + "epoch": 2.7902421652421654, + "grad_norm": 0.8107752799987793, + "learning_rate": 4.1957690287554986e-05, + "loss": 0.6548, + "step": 15672 + }, + { + "epoch": 2.7904202279202277, + "grad_norm": 0.8361678719520569, + "learning_rate": 4.1946292503714556e-05, + "loss": 0.9224, + "step": 15673 + }, + { + "epoch": 2.7905982905982905, + "grad_norm": 0.7812657952308655, + "learning_rate": 4.1934895857315904e-05, + "loss": 0.7126, + "step": 15674 + }, + { + "epoch": 2.790776353276353, + "grad_norm": 0.9054265022277832, + "learning_rate": 4.192350034858241e-05, + "loss": 0.6891, + "step": 15675 + }, + { + "epoch": 2.790954415954416, + "grad_norm": 0.9675585627555847, + "learning_rate": 4.1912105977737214e-05, + "loss": 0.8429, + "step": 15676 + }, + { + "epoch": 2.7911324786324787, + "grad_norm": 0.9077114462852478, + "learning_rate": 4.19007127450037e-05, + "loss": 0.8864, + "step": 15677 + }, + { + "epoch": 2.7913105413105415, + "grad_norm": 0.9230541586875916, + "learning_rate": 4.188932065060497e-05, + "loss": 1.0065, + "step": 15678 + }, + { + "epoch": 2.791488603988604, + "grad_norm": 0.8667981028556824, + "learning_rate": 4.1877929694764315e-05, + "loss": 0.7584, + "step": 15679 + }, + { + "epoch": 2.7916666666666665, + "grad_norm": 0.8986212015151978, + "learning_rate": 4.1866539877704894e-05, + "loss": 0.9205, + "step": 15680 + }, + { + "epoch": 2.7918447293447293, + "grad_norm": 0.8524685502052307, + "learning_rate": 4.185515119964986e-05, + "loss": 0.8516, + "step": 15681 + }, + { + "epoch": 2.792022792022792, + "grad_norm": 0.8247089385986328, + "learning_rate": 4.184376366082234e-05, + "loss": 0.8733, + "step": 15682 + }, + { + "epoch": 2.7922008547008548, + "grad_norm": 0.8236528635025024, + "learning_rate": 4.183237726144549e-05, + "loss": 0.8715, + "step": 15683 + }, + { + "epoch": 2.7923789173789175, + "grad_norm": 0.8853272199630737, + "learning_rate": 4.182099200174232e-05, + "loss": 0.7741, + "step": 15684 + }, + { + "epoch": 2.79255698005698, + "grad_norm": 0.8243789672851562, + "learning_rate": 4.180960788193603e-05, + "loss": 0.9196, + "step": 15685 + }, + { + "epoch": 2.7927350427350426, + "grad_norm": 0.9670386910438538, + "learning_rate": 4.1798224902249515e-05, + "loss": 0.828, + "step": 15686 + }, + { + "epoch": 2.7929131054131053, + "grad_norm": 0.7831283211708069, + "learning_rate": 4.178684306290592e-05, + "loss": 0.8389, + "step": 15687 + }, + { + "epoch": 2.793091168091168, + "grad_norm": 0.9372588396072388, + "learning_rate": 4.177546236412822e-05, + "loss": 1.2267, + "step": 15688 + }, + { + "epoch": 2.793269230769231, + "grad_norm": 0.9065600633621216, + "learning_rate": 4.176408280613937e-05, + "loss": 0.9674, + "step": 15689 + }, + { + "epoch": 2.7934472934472936, + "grad_norm": 0.8220530152320862, + "learning_rate": 4.1752704389162344e-05, + "loss": 0.8717, + "step": 15690 + }, + { + "epoch": 2.7936253561253563, + "grad_norm": 0.8952174782752991, + "learning_rate": 4.174132711342005e-05, + "loss": 0.8904, + "step": 15691 + }, + { + "epoch": 2.7938034188034186, + "grad_norm": 0.8454076647758484, + "learning_rate": 4.172995097913549e-05, + "loss": 0.9784, + "step": 15692 + }, + { + "epoch": 2.7939814814814814, + "grad_norm": 0.8697866797447205, + "learning_rate": 4.171857598653143e-05, + "loss": 1.0042, + "step": 15693 + }, + { + "epoch": 2.794159544159544, + "grad_norm": 0.8736211657524109, + "learning_rate": 4.170720213583084e-05, + "loss": 0.8787, + "step": 15694 + }, + { + "epoch": 2.794337606837607, + "grad_norm": 1.0082578659057617, + "learning_rate": 4.1695829427256525e-05, + "loss": 1.2508, + "step": 15695 + }, + { + "epoch": 2.7945156695156697, + "grad_norm": 0.8092042207717896, + "learning_rate": 4.1684457861031325e-05, + "loss": 0.8968, + "step": 15696 + }, + { + "epoch": 2.794693732193732, + "grad_norm": 0.847034752368927, + "learning_rate": 4.167308743737802e-05, + "loss": 0.8019, + "step": 15697 + }, + { + "epoch": 2.7948717948717947, + "grad_norm": 0.9059078097343445, + "learning_rate": 4.1661718156519414e-05, + "loss": 1.0393, + "step": 15698 + }, + { + "epoch": 2.7950498575498575, + "grad_norm": 0.8907228112220764, + "learning_rate": 4.165035001867822e-05, + "loss": 0.7388, + "step": 15699 + }, + { + "epoch": 2.79522792022792, + "grad_norm": 0.8089052438735962, + "learning_rate": 4.163898302407727e-05, + "loss": 0.8052, + "step": 15700 + }, + { + "epoch": 2.795405982905983, + "grad_norm": 1.1158883571624756, + "learning_rate": 4.162761717293915e-05, + "loss": 0.6923, + "step": 15701 + }, + { + "epoch": 2.7955840455840457, + "grad_norm": 0.8157755732536316, + "learning_rate": 4.1616252465486684e-05, + "loss": 0.606, + "step": 15702 + }, + { + "epoch": 2.7957621082621085, + "grad_norm": 0.8905386328697205, + "learning_rate": 4.1604888901942386e-05, + "loss": 0.7798, + "step": 15703 + }, + { + "epoch": 2.7959401709401708, + "grad_norm": 0.7655990719795227, + "learning_rate": 4.1593526482529034e-05, + "loss": 0.7435, + "step": 15704 + }, + { + "epoch": 2.7961182336182335, + "grad_norm": 0.900643527507782, + "learning_rate": 4.1582165207469195e-05, + "loss": 0.8314, + "step": 15705 + }, + { + "epoch": 2.7962962962962963, + "grad_norm": 0.7737550735473633, + "learning_rate": 4.1570805076985475e-05, + "loss": 0.8281, + "step": 15706 + }, + { + "epoch": 2.796474358974359, + "grad_norm": 0.8385021090507507, + "learning_rate": 4.1559446091300455e-05, + "loss": 0.896, + "step": 15707 + }, + { + "epoch": 2.796652421652422, + "grad_norm": 0.8830214142799377, + "learning_rate": 4.1548088250636687e-05, + "loss": 0.8856, + "step": 15708 + }, + { + "epoch": 2.796830484330484, + "grad_norm": 0.9748533368110657, + "learning_rate": 4.1536731555216676e-05, + "loss": 0.8768, + "step": 15709 + }, + { + "epoch": 2.797008547008547, + "grad_norm": 0.8918380737304688, + "learning_rate": 4.1525376005263e-05, + "loss": 0.8332, + "step": 15710 + }, + { + "epoch": 2.7971866096866096, + "grad_norm": 0.9205654263496399, + "learning_rate": 4.15140216009981e-05, + "loss": 0.7698, + "step": 15711 + }, + { + "epoch": 2.7973646723646723, + "grad_norm": 0.9631472229957581, + "learning_rate": 4.1502668342644455e-05, + "loss": 0.9604, + "step": 15712 + }, + { + "epoch": 2.797542735042735, + "grad_norm": 0.8770546913146973, + "learning_rate": 4.1491316230424516e-05, + "loss": 0.7661, + "step": 15713 + }, + { + "epoch": 2.797720797720798, + "grad_norm": 0.8872628808021545, + "learning_rate": 4.147996526456069e-05, + "loss": 1.0847, + "step": 15714 + }, + { + "epoch": 2.7978988603988606, + "grad_norm": 0.8924010396003723, + "learning_rate": 4.146861544527538e-05, + "loss": 0.8159, + "step": 15715 + }, + { + "epoch": 2.7980769230769234, + "grad_norm": 0.9251703023910522, + "learning_rate": 4.1457266772790923e-05, + "loss": 0.7204, + "step": 15716 + }, + { + "epoch": 2.7982549857549857, + "grad_norm": 0.8891414403915405, + "learning_rate": 4.144591924732979e-05, + "loss": 0.8576, + "step": 15717 + }, + { + "epoch": 2.7984330484330484, + "grad_norm": 0.9676079154014587, + "learning_rate": 4.143457286911415e-05, + "loss": 1.1912, + "step": 15718 + }, + { + "epoch": 2.798611111111111, + "grad_norm": 0.8125061392784119, + "learning_rate": 4.142322763836645e-05, + "loss": 0.8111, + "step": 15719 + }, + { + "epoch": 2.798789173789174, + "grad_norm": 0.8612900972366333, + "learning_rate": 4.141188355530891e-05, + "loss": 0.7537, + "step": 15720 + }, + { + "epoch": 2.798967236467236, + "grad_norm": 1.0774086713790894, + "learning_rate": 4.14005406201638e-05, + "loss": 0.9364, + "step": 15721 + }, + { + "epoch": 2.799145299145299, + "grad_norm": 0.8296873569488525, + "learning_rate": 4.138919883315338e-05, + "loss": 0.9329, + "step": 15722 + }, + { + "epoch": 2.7993233618233617, + "grad_norm": 0.757978618144989, + "learning_rate": 4.137785819449984e-05, + "loss": 1.0353, + "step": 15723 + }, + { + "epoch": 2.7995014245014245, + "grad_norm": 0.8584328293800354, + "learning_rate": 4.136651870442536e-05, + "loss": 1.0577, + "step": 15724 + }, + { + "epoch": 2.7996794871794872, + "grad_norm": 0.7919153571128845, + "learning_rate": 4.135518036315222e-05, + "loss": 0.8013, + "step": 15725 + }, + { + "epoch": 2.79985754985755, + "grad_norm": 0.968519926071167, + "learning_rate": 4.134384317090243e-05, + "loss": 0.7684, + "step": 15726 + }, + { + "epoch": 2.8000356125356127, + "grad_norm": 0.8565614819526672, + "learning_rate": 4.133250712789826e-05, + "loss": 0.9177, + "step": 15727 + }, + { + "epoch": 2.8002136752136755, + "grad_norm": 0.8614934086799622, + "learning_rate": 4.1321172234361647e-05, + "loss": 0.9613, + "step": 15728 + }, + { + "epoch": 2.800391737891738, + "grad_norm": 0.8621053099632263, + "learning_rate": 4.130983849051483e-05, + "loss": 0.8254, + "step": 15729 + }, + { + "epoch": 2.8005698005698005, + "grad_norm": 0.8108318448066711, + "learning_rate": 4.12985058965798e-05, + "loss": 0.7577, + "step": 15730 + }, + { + "epoch": 2.8007478632478633, + "grad_norm": 0.9211961627006531, + "learning_rate": 4.1287174452778564e-05, + "loss": 0.8204, + "step": 15731 + }, + { + "epoch": 2.800925925925926, + "grad_norm": 0.8582359552383423, + "learning_rate": 4.127584415933326e-05, + "loss": 0.8915, + "step": 15732 + }, + { + "epoch": 2.8011039886039883, + "grad_norm": 0.9122742414474487, + "learning_rate": 4.126451501646573e-05, + "loss": 1.0896, + "step": 15733 + }, + { + "epoch": 2.801282051282051, + "grad_norm": 0.7392016649246216, + "learning_rate": 4.125318702439804e-05, + "loss": 0.6354, + "step": 15734 + }, + { + "epoch": 2.801460113960114, + "grad_norm": 0.8227471113204956, + "learning_rate": 4.124186018335213e-05, + "loss": 0.7903, + "step": 15735 + }, + { + "epoch": 2.8016381766381766, + "grad_norm": 0.8843638300895691, + "learning_rate": 4.12305344935499e-05, + "loss": 0.7545, + "step": 15736 + }, + { + "epoch": 2.8018162393162394, + "grad_norm": 0.9978471994400024, + "learning_rate": 4.121920995521327e-05, + "loss": 1.0127, + "step": 15737 + }, + { + "epoch": 2.801994301994302, + "grad_norm": 0.9969626069068909, + "learning_rate": 4.12078865685641e-05, + "loss": 0.8137, + "step": 15738 + }, + { + "epoch": 2.802172364672365, + "grad_norm": 1.0768957138061523, + "learning_rate": 4.119656433382428e-05, + "loss": 0.6866, + "step": 15739 + }, + { + "epoch": 2.8023504273504276, + "grad_norm": 0.7401831746101379, + "learning_rate": 4.1185243251215624e-05, + "loss": 0.9103, + "step": 15740 + }, + { + "epoch": 2.80252849002849, + "grad_norm": 0.9753470420837402, + "learning_rate": 4.1173923320959905e-05, + "loss": 0.9499, + "step": 15741 + }, + { + "epoch": 2.8027065527065527, + "grad_norm": 0.9174960851669312, + "learning_rate": 4.116260454327904e-05, + "loss": 1.0355, + "step": 15742 + }, + { + "epoch": 2.8028846153846154, + "grad_norm": 0.8292258381843567, + "learning_rate": 4.115128691839464e-05, + "loss": 0.8806, + "step": 15743 + }, + { + "epoch": 2.803062678062678, + "grad_norm": 0.9542452096939087, + "learning_rate": 4.1139970446528564e-05, + "loss": 0.8378, + "step": 15744 + }, + { + "epoch": 2.8032407407407405, + "grad_norm": 0.848686933517456, + "learning_rate": 4.1128655127902485e-05, + "loss": 0.7939, + "step": 15745 + }, + { + "epoch": 2.8034188034188032, + "grad_norm": 0.8277645111083984, + "learning_rate": 4.1117340962738125e-05, + "loss": 0.9277, + "step": 15746 + }, + { + "epoch": 2.803596866096866, + "grad_norm": 0.8613318204879761, + "learning_rate": 4.110602795125714e-05, + "loss": 0.7622, + "step": 15747 + }, + { + "epoch": 2.8037749287749287, + "grad_norm": 0.7106199860572815, + "learning_rate": 4.109471609368121e-05, + "loss": 0.787, + "step": 15748 + }, + { + "epoch": 2.8039529914529915, + "grad_norm": 0.8933543562889099, + "learning_rate": 4.108340539023194e-05, + "loss": 0.857, + "step": 15749 + }, + { + "epoch": 2.8041310541310542, + "grad_norm": 0.8682022094726562, + "learning_rate": 4.107209584113092e-05, + "loss": 0.7931, + "step": 15750 + }, + { + "epoch": 2.804309116809117, + "grad_norm": 0.828279435634613, + "learning_rate": 4.106078744659981e-05, + "loss": 0.8404, + "step": 15751 + }, + { + "epoch": 2.8044871794871797, + "grad_norm": 0.9503956437110901, + "learning_rate": 4.1049480206860136e-05, + "loss": 0.8588, + "step": 15752 + }, + { + "epoch": 2.804665242165242, + "grad_norm": 0.810714602470398, + "learning_rate": 4.1038174122133435e-05, + "loss": 0.8592, + "step": 15753 + }, + { + "epoch": 2.804843304843305, + "grad_norm": 0.9458156824111938, + "learning_rate": 4.1026869192641225e-05, + "loss": 0.7441, + "step": 15754 + }, + { + "epoch": 2.8050213675213675, + "grad_norm": 0.8267046809196472, + "learning_rate": 4.1015565418605016e-05, + "loss": 0.8574, + "step": 15755 + }, + { + "epoch": 2.8051994301994303, + "grad_norm": 0.8413352370262146, + "learning_rate": 4.100426280024623e-05, + "loss": 0.7598, + "step": 15756 + }, + { + "epoch": 2.8053774928774926, + "grad_norm": 0.9205049872398376, + "learning_rate": 4.099296133778644e-05, + "loss": 0.8346, + "step": 15757 + }, + { + "epoch": 2.8055555555555554, + "grad_norm": 0.7986966967582703, + "learning_rate": 4.098166103144692e-05, + "loss": 0.7607, + "step": 15758 + }, + { + "epoch": 2.805733618233618, + "grad_norm": 0.9400181174278259, + "learning_rate": 4.097036188144918e-05, + "loss": 0.8947, + "step": 15759 + }, + { + "epoch": 2.805911680911681, + "grad_norm": 0.9014864563941956, + "learning_rate": 4.0959063888014594e-05, + "loss": 0.7781, + "step": 15760 + }, + { + "epoch": 2.8060897435897436, + "grad_norm": 0.8163666725158691, + "learning_rate": 4.094776705136448e-05, + "loss": 0.9042, + "step": 15761 + }, + { + "epoch": 2.8062678062678064, + "grad_norm": 0.8435617685317993, + "learning_rate": 4.0936471371720217e-05, + "loss": 0.9229, + "step": 15762 + }, + { + "epoch": 2.806445868945869, + "grad_norm": 0.7995414137840271, + "learning_rate": 4.0925176849303104e-05, + "loss": 0.7294, + "step": 15763 + }, + { + "epoch": 2.806623931623932, + "grad_norm": 0.9147883057594299, + "learning_rate": 4.091388348433442e-05, + "loss": 0.955, + "step": 15764 + }, + { + "epoch": 2.806801994301994, + "grad_norm": 0.7326688170433044, + "learning_rate": 4.0902591277035474e-05, + "loss": 0.5885, + "step": 15765 + }, + { + "epoch": 2.806980056980057, + "grad_norm": 0.8756957054138184, + "learning_rate": 4.0891300227627425e-05, + "loss": 0.9404, + "step": 15766 + }, + { + "epoch": 2.8071581196581197, + "grad_norm": 0.8897661566734314, + "learning_rate": 4.088001033633165e-05, + "loss": 1.0214, + "step": 15767 + }, + { + "epoch": 2.8073361823361824, + "grad_norm": 0.8007500171661377, + "learning_rate": 4.086872160336919e-05, + "loss": 0.6553, + "step": 15768 + }, + { + "epoch": 2.807514245014245, + "grad_norm": 0.8175814747810364, + "learning_rate": 4.0857434028961316e-05, + "loss": 0.8572, + "step": 15769 + }, + { + "epoch": 2.8076923076923075, + "grad_norm": 0.8290224671363831, + "learning_rate": 4.084614761332917e-05, + "loss": 0.9196, + "step": 15770 + }, + { + "epoch": 2.8078703703703702, + "grad_norm": 0.9355306625366211, + "learning_rate": 4.083486235669385e-05, + "loss": 0.8198, + "step": 15771 + }, + { + "epoch": 2.808048433048433, + "grad_norm": 0.942641019821167, + "learning_rate": 4.082357825927656e-05, + "loss": 0.7817, + "step": 15772 + }, + { + "epoch": 2.8082264957264957, + "grad_norm": 0.9115625023841858, + "learning_rate": 4.081229532129827e-05, + "loss": 1.046, + "step": 15773 + }, + { + "epoch": 2.8084045584045585, + "grad_norm": 0.8559226393699646, + "learning_rate": 4.080101354298016e-05, + "loss": 0.7085, + "step": 15774 + }, + { + "epoch": 2.8085826210826212, + "grad_norm": 0.8223599195480347, + "learning_rate": 4.0789732924543144e-05, + "loss": 0.9019, + "step": 15775 + }, + { + "epoch": 2.808760683760684, + "grad_norm": 0.8511637449264526, + "learning_rate": 4.0778453466208344e-05, + "loss": 0.6845, + "step": 15776 + }, + { + "epoch": 2.8089387464387463, + "grad_norm": 0.9633409976959229, + "learning_rate": 4.076717516819674e-05, + "loss": 1.0478, + "step": 15777 + }, + { + "epoch": 2.809116809116809, + "grad_norm": 0.8657141327857971, + "learning_rate": 4.075589803072928e-05, + "loss": 0.8694, + "step": 15778 + }, + { + "epoch": 2.809294871794872, + "grad_norm": 0.8126863241195679, + "learning_rate": 4.0744622054026936e-05, + "loss": 0.8529, + "step": 15779 + }, + { + "epoch": 2.8094729344729346, + "grad_norm": 0.8671838045120239, + "learning_rate": 4.0733347238310626e-05, + "loss": 0.81, + "step": 15780 + }, + { + "epoch": 2.8096509971509973, + "grad_norm": 0.8336054086685181, + "learning_rate": 4.0722073583801223e-05, + "loss": 0.7103, + "step": 15781 + }, + { + "epoch": 2.8098290598290596, + "grad_norm": 0.7833893299102783, + "learning_rate": 4.071080109071973e-05, + "loss": 0.875, + "step": 15782 + }, + { + "epoch": 2.8100071225071224, + "grad_norm": 0.9142106771469116, + "learning_rate": 4.0699529759286844e-05, + "loss": 0.9063, + "step": 15783 + }, + { + "epoch": 2.810185185185185, + "grad_norm": 0.7767373919487, + "learning_rate": 4.0688259589723565e-05, + "loss": 0.841, + "step": 15784 + }, + { + "epoch": 2.810363247863248, + "grad_norm": 0.8679327368736267, + "learning_rate": 4.067699058225056e-05, + "loss": 0.8581, + "step": 15785 + }, + { + "epoch": 2.8105413105413106, + "grad_norm": 0.9434911608695984, + "learning_rate": 4.066572273708873e-05, + "loss": 1.0166, + "step": 15786 + }, + { + "epoch": 2.8107193732193734, + "grad_norm": 0.8053399920463562, + "learning_rate": 4.06544560544588e-05, + "loss": 0.874, + "step": 15787 + }, + { + "epoch": 2.810897435897436, + "grad_norm": 1.0207599401474, + "learning_rate": 4.0643190534581524e-05, + "loss": 0.8296, + "step": 15788 + }, + { + "epoch": 2.8110754985754984, + "grad_norm": 0.8879590034484863, + "learning_rate": 4.0631926177677626e-05, + "loss": 0.8787, + "step": 15789 + }, + { + "epoch": 2.811253561253561, + "grad_norm": 1.0816758871078491, + "learning_rate": 4.062066298396778e-05, + "loss": 0.8129, + "step": 15790 + }, + { + "epoch": 2.811431623931624, + "grad_norm": 0.8332961797714233, + "learning_rate": 4.060940095367272e-05, + "loss": 0.8855, + "step": 15791 + }, + { + "epoch": 2.8116096866096867, + "grad_norm": 0.98028963804245, + "learning_rate": 4.059814008701308e-05, + "loss": 1.0065, + "step": 15792 + }, + { + "epoch": 2.8117877492877494, + "grad_norm": 0.8983020782470703, + "learning_rate": 4.058688038420949e-05, + "loss": 0.8259, + "step": 15793 + }, + { + "epoch": 2.8119658119658117, + "grad_norm": 0.8029065728187561, + "learning_rate": 4.057562184548255e-05, + "loss": 0.7639, + "step": 15794 + }, + { + "epoch": 2.8121438746438745, + "grad_norm": 0.8188722729682922, + "learning_rate": 4.056436447105286e-05, + "loss": 0.7179, + "step": 15795 + }, + { + "epoch": 2.8123219373219372, + "grad_norm": 0.8493495583534241, + "learning_rate": 4.055310826114095e-05, + "loss": 0.9479, + "step": 15796 + }, + { + "epoch": 2.8125, + "grad_norm": 0.7770833373069763, + "learning_rate": 4.0541853215967466e-05, + "loss": 0.6915, + "step": 15797 + }, + { + "epoch": 2.8126780626780628, + "grad_norm": 0.8238499760627747, + "learning_rate": 4.053059933575277e-05, + "loss": 1.0257, + "step": 15798 + }, + { + "epoch": 2.8128561253561255, + "grad_norm": 0.8537020683288574, + "learning_rate": 4.051934662071754e-05, + "loss": 0.7037, + "step": 15799 + }, + { + "epoch": 2.8130341880341883, + "grad_norm": 0.8120609521865845, + "learning_rate": 4.0508095071082055e-05, + "loss": 0.9533, + "step": 15800 + }, + { + "epoch": 2.8132122507122506, + "grad_norm": 0.8701691627502441, + "learning_rate": 4.0496844687066915e-05, + "loss": 0.7129, + "step": 15801 + }, + { + "epoch": 2.8133903133903133, + "grad_norm": 0.9007154107093811, + "learning_rate": 4.048559546889249e-05, + "loss": 0.8879, + "step": 15802 + }, + { + "epoch": 2.813568376068376, + "grad_norm": 0.8858364820480347, + "learning_rate": 4.047434741677919e-05, + "loss": 0.9391, + "step": 15803 + }, + { + "epoch": 2.813746438746439, + "grad_norm": 0.8597678542137146, + "learning_rate": 4.046310053094739e-05, + "loss": 0.7897, + "step": 15804 + }, + { + "epoch": 2.8139245014245016, + "grad_norm": 0.8493016362190247, + "learning_rate": 4.0451854811617475e-05, + "loss": 0.914, + "step": 15805 + }, + { + "epoch": 2.814102564102564, + "grad_norm": 0.8347373008728027, + "learning_rate": 4.044061025900973e-05, + "loss": 0.671, + "step": 15806 + }, + { + "epoch": 2.8142806267806266, + "grad_norm": 0.7406242489814758, + "learning_rate": 4.042936687334455e-05, + "loss": 0.6592, + "step": 15807 + }, + { + "epoch": 2.8144586894586894, + "grad_norm": 0.93736732006073, + "learning_rate": 4.041812465484214e-05, + "loss": 0.8301, + "step": 15808 + }, + { + "epoch": 2.814636752136752, + "grad_norm": 0.8744384050369263, + "learning_rate": 4.040688360372282e-05, + "loss": 0.9104, + "step": 15809 + }, + { + "epoch": 2.814814814814815, + "grad_norm": 0.7417266964912415, + "learning_rate": 4.0395643720206834e-05, + "loss": 0.7698, + "step": 15810 + }, + { + "epoch": 2.8149928774928776, + "grad_norm": 0.8601716160774231, + "learning_rate": 4.038440500451438e-05, + "loss": 0.8459, + "step": 15811 + }, + { + "epoch": 2.8151709401709404, + "grad_norm": 0.9801309108734131, + "learning_rate": 4.037316745686568e-05, + "loss": 0.9555, + "step": 15812 + }, + { + "epoch": 2.8153490028490027, + "grad_norm": 0.8559770584106445, + "learning_rate": 4.036193107748085e-05, + "loss": 0.9912, + "step": 15813 + }, + { + "epoch": 2.8155270655270654, + "grad_norm": 0.8155802488327026, + "learning_rate": 4.035069586658018e-05, + "loss": 0.8471, + "step": 15814 + }, + { + "epoch": 2.815705128205128, + "grad_norm": 0.96893310546875, + "learning_rate": 4.033946182438364e-05, + "loss": 0.9329, + "step": 15815 + }, + { + "epoch": 2.815883190883191, + "grad_norm": 0.9981120228767395, + "learning_rate": 4.032822895111144e-05, + "loss": 1.0333, + "step": 15816 + }, + { + "epoch": 2.8160612535612537, + "grad_norm": 0.9491816163063049, + "learning_rate": 4.031699724698363e-05, + "loss": 0.8838, + "step": 15817 + }, + { + "epoch": 2.816239316239316, + "grad_norm": 0.7149206399917603, + "learning_rate": 4.030576671222029e-05, + "loss": 0.8248, + "step": 15818 + }, + { + "epoch": 2.8164173789173788, + "grad_norm": 0.8366861343383789, + "learning_rate": 4.029453734704145e-05, + "loss": 0.8041, + "step": 15819 + }, + { + "epoch": 2.8165954415954415, + "grad_norm": 0.8147984147071838, + "learning_rate": 4.0283309151667116e-05, + "loss": 0.8702, + "step": 15820 + }, + { + "epoch": 2.8167735042735043, + "grad_norm": 0.8462722897529602, + "learning_rate": 4.027208212631729e-05, + "loss": 0.9899, + "step": 15821 + }, + { + "epoch": 2.816951566951567, + "grad_norm": 0.9278651475906372, + "learning_rate": 4.0260856271211946e-05, + "loss": 0.8684, + "step": 15822 + }, + { + "epoch": 2.8171296296296298, + "grad_norm": 0.8845569491386414, + "learning_rate": 4.0249631586570993e-05, + "loss": 0.6956, + "step": 15823 + }, + { + "epoch": 2.8173076923076925, + "grad_norm": 0.7803958654403687, + "learning_rate": 4.0238408072614453e-05, + "loss": 0.758, + "step": 15824 + }, + { + "epoch": 2.817485754985755, + "grad_norm": 0.8920331597328186, + "learning_rate": 4.022718572956209e-05, + "loss": 0.9131, + "step": 15825 + }, + { + "epoch": 2.8176638176638176, + "grad_norm": 0.9914098978042603, + "learning_rate": 4.021596455763389e-05, + "loss": 0.973, + "step": 15826 + }, + { + "epoch": 2.8178418803418803, + "grad_norm": 0.8329834938049316, + "learning_rate": 4.020474455704968e-05, + "loss": 0.8746, + "step": 15827 + }, + { + "epoch": 2.818019943019943, + "grad_norm": 0.7789189219474792, + "learning_rate": 4.019352572802928e-05, + "loss": 0.7359, + "step": 15828 + }, + { + "epoch": 2.818198005698006, + "grad_norm": 0.8405792713165283, + "learning_rate": 4.0182308070792505e-05, + "loss": 0.7897, + "step": 15829 + }, + { + "epoch": 2.818376068376068, + "grad_norm": 0.9215472936630249, + "learning_rate": 4.0171091585559116e-05, + "loss": 0.9017, + "step": 15830 + }, + { + "epoch": 2.818554131054131, + "grad_norm": 0.8310655355453491, + "learning_rate": 4.0159876272548933e-05, + "loss": 0.7499, + "step": 15831 + }, + { + "epoch": 2.8187321937321936, + "grad_norm": 0.8645792603492737, + "learning_rate": 4.014866213198167e-05, + "loss": 0.9009, + "step": 15832 + }, + { + "epoch": 2.8189102564102564, + "grad_norm": 0.8059788942337036, + "learning_rate": 4.013744916407703e-05, + "loss": 0.7367, + "step": 15833 + }, + { + "epoch": 2.819088319088319, + "grad_norm": 0.7990314960479736, + "learning_rate": 4.0126237369054745e-05, + "loss": 1.0172, + "step": 15834 + }, + { + "epoch": 2.819266381766382, + "grad_norm": 0.9321257472038269, + "learning_rate": 4.0115026747134446e-05, + "loss": 1.1224, + "step": 15835 + }, + { + "epoch": 2.8194444444444446, + "grad_norm": 0.9674378037452698, + "learning_rate": 4.0103817298535794e-05, + "loss": 0.9881, + "step": 15836 + }, + { + "epoch": 2.8196225071225074, + "grad_norm": 0.8573030829429626, + "learning_rate": 4.009260902347842e-05, + "loss": 0.7515, + "step": 15837 + }, + { + "epoch": 2.8198005698005697, + "grad_norm": 0.9248984456062317, + "learning_rate": 4.00814019221819e-05, + "loss": 0.9265, + "step": 15838 + }, + { + "epoch": 2.8199786324786325, + "grad_norm": 0.8749010562896729, + "learning_rate": 4.00701959948659e-05, + "loss": 0.8659, + "step": 15839 + }, + { + "epoch": 2.820156695156695, + "grad_norm": 0.8070803880691528, + "learning_rate": 4.005899124174986e-05, + "loss": 0.8813, + "step": 15840 + }, + { + "epoch": 2.820334757834758, + "grad_norm": 1.0711981058120728, + "learning_rate": 4.004778766305339e-05, + "loss": 1.019, + "step": 15841 + }, + { + "epoch": 2.8205128205128203, + "grad_norm": 0.7646795511245728, + "learning_rate": 4.0036585258995985e-05, + "loss": 0.8326, + "step": 15842 + }, + { + "epoch": 2.820690883190883, + "grad_norm": 0.6720184683799744, + "learning_rate": 4.002538402979713e-05, + "loss": 0.6642, + "step": 15843 + }, + { + "epoch": 2.8208689458689458, + "grad_norm": 0.8062998056411743, + "learning_rate": 4.001418397567629e-05, + "loss": 0.9585, + "step": 15844 + }, + { + "epoch": 2.8210470085470085, + "grad_norm": 0.835515558719635, + "learning_rate": 4.0002985096852893e-05, + "loss": 0.7992, + "step": 15845 + }, + { + "epoch": 2.8212250712250713, + "grad_norm": 0.8308731913566589, + "learning_rate": 3.9991787393546385e-05, + "loss": 0.9982, + "step": 15846 + }, + { + "epoch": 2.821403133903134, + "grad_norm": 0.9056837558746338, + "learning_rate": 3.998059086597614e-05, + "loss": 0.8283, + "step": 15847 + }, + { + "epoch": 2.8215811965811968, + "grad_norm": 0.853070080280304, + "learning_rate": 3.9969395514361506e-05, + "loss": 1.033, + "step": 15848 + }, + { + "epoch": 2.8217592592592595, + "grad_norm": 0.9179061055183411, + "learning_rate": 3.99582013389219e-05, + "loss": 0.929, + "step": 15849 + }, + { + "epoch": 2.821937321937322, + "grad_norm": 0.8702627420425415, + "learning_rate": 3.9947008339876616e-05, + "loss": 0.9994, + "step": 15850 + }, + { + "epoch": 2.8221153846153846, + "grad_norm": 0.9594024419784546, + "learning_rate": 3.9935816517444935e-05, + "loss": 0.88, + "step": 15851 + }, + { + "epoch": 2.8222934472934473, + "grad_norm": 0.8474575877189636, + "learning_rate": 3.992462587184618e-05, + "loss": 0.7817, + "step": 15852 + }, + { + "epoch": 2.82247150997151, + "grad_norm": 0.7588878870010376, + "learning_rate": 3.9913436403299533e-05, + "loss": 0.7911, + "step": 15853 + }, + { + "epoch": 2.8226495726495724, + "grad_norm": 0.8467457890510559, + "learning_rate": 3.9902248112024366e-05, + "loss": 0.7433, + "step": 15854 + }, + { + "epoch": 2.822827635327635, + "grad_norm": 0.839137077331543, + "learning_rate": 3.989106099823972e-05, + "loss": 0.7303, + "step": 15855 + }, + { + "epoch": 2.823005698005698, + "grad_norm": 0.8945586085319519, + "learning_rate": 3.987987506216495e-05, + "loss": 0.6697, + "step": 15856 + }, + { + "epoch": 2.8231837606837606, + "grad_norm": 0.749971330165863, + "learning_rate": 3.9868690304019064e-05, + "loss": 0.65, + "step": 15857 + }, + { + "epoch": 2.8233618233618234, + "grad_norm": 0.9841105341911316, + "learning_rate": 3.985750672402131e-05, + "loss": 0.8382, + "step": 15858 + }, + { + "epoch": 2.823539886039886, + "grad_norm": 0.8696077466011047, + "learning_rate": 3.984632432239078e-05, + "loss": 0.7895, + "step": 15859 + }, + { + "epoch": 2.823717948717949, + "grad_norm": 0.7845979928970337, + "learning_rate": 3.9835143099346575e-05, + "loss": 0.8673, + "step": 15860 + }, + { + "epoch": 2.8238960113960117, + "grad_norm": 0.8145211338996887, + "learning_rate": 3.982396305510775e-05, + "loss": 0.905, + "step": 15861 + }, + { + "epoch": 2.824074074074074, + "grad_norm": 0.8533337116241455, + "learning_rate": 3.981278418989336e-05, + "loss": 0.9597, + "step": 15862 + }, + { + "epoch": 2.8242521367521367, + "grad_norm": 0.9430350065231323, + "learning_rate": 3.980160650392241e-05, + "loss": 0.9043, + "step": 15863 + }, + { + "epoch": 2.8244301994301995, + "grad_norm": 0.7469115257263184, + "learning_rate": 3.9790429997414e-05, + "loss": 0.9074, + "step": 15864 + }, + { + "epoch": 2.824608262108262, + "grad_norm": 0.8558746576309204, + "learning_rate": 3.977925467058696e-05, + "loss": 0.8958, + "step": 15865 + }, + { + "epoch": 2.8247863247863245, + "grad_norm": 1.1557669639587402, + "learning_rate": 3.976808052366037e-05, + "loss": 0.7884, + "step": 15866 + }, + { + "epoch": 2.8249643874643873, + "grad_norm": 0.8448477983474731, + "learning_rate": 3.975690755685312e-05, + "loss": 0.9542, + "step": 15867 + }, + { + "epoch": 2.82514245014245, + "grad_norm": 0.833256721496582, + "learning_rate": 3.9745735770384086e-05, + "loss": 0.9196, + "step": 15868 + }, + { + "epoch": 2.8253205128205128, + "grad_norm": 0.9318852424621582, + "learning_rate": 3.973456516447226e-05, + "loss": 0.7792, + "step": 15869 + }, + { + "epoch": 2.8254985754985755, + "grad_norm": 0.832655131816864, + "learning_rate": 3.972339573933638e-05, + "loss": 0.9526, + "step": 15870 + }, + { + "epoch": 2.8256766381766383, + "grad_norm": 0.7546842694282532, + "learning_rate": 3.9712227495195406e-05, + "loss": 0.5969, + "step": 15871 + }, + { + "epoch": 2.825854700854701, + "grad_norm": 0.8538267016410828, + "learning_rate": 3.970106043226802e-05, + "loss": 0.8369, + "step": 15872 + }, + { + "epoch": 2.826032763532764, + "grad_norm": 0.8023465871810913, + "learning_rate": 3.968989455077314e-05, + "loss": 0.7257, + "step": 15873 + }, + { + "epoch": 2.826210826210826, + "grad_norm": 0.7905409932136536, + "learning_rate": 3.96787298509295e-05, + "loss": 0.7965, + "step": 15874 + }, + { + "epoch": 2.826388888888889, + "grad_norm": 0.8316642642021179, + "learning_rate": 3.966756633295583e-05, + "loss": 0.808, + "step": 15875 + }, + { + "epoch": 2.8265669515669516, + "grad_norm": 0.9130233526229858, + "learning_rate": 3.965640399707088e-05, + "loss": 0.882, + "step": 15876 + }, + { + "epoch": 2.8267450142450143, + "grad_norm": 0.873147189617157, + "learning_rate": 3.9645242843493325e-05, + "loss": 0.8347, + "step": 15877 + }, + { + "epoch": 2.8269230769230766, + "grad_norm": 0.8806825280189514, + "learning_rate": 3.963408287244183e-05, + "loss": 1.0496, + "step": 15878 + }, + { + "epoch": 2.8271011396011394, + "grad_norm": 0.8899962306022644, + "learning_rate": 3.962292408413516e-05, + "loss": 1.0014, + "step": 15879 + }, + { + "epoch": 2.827279202279202, + "grad_norm": 0.900303065776825, + "learning_rate": 3.961176647879179e-05, + "loss": 0.7345, + "step": 15880 + }, + { + "epoch": 2.827457264957265, + "grad_norm": 1.1055848598480225, + "learning_rate": 3.960061005663049e-05, + "loss": 0.9697, + "step": 15881 + }, + { + "epoch": 2.8276353276353277, + "grad_norm": 0.891404390335083, + "learning_rate": 3.958945481786969e-05, + "loss": 0.8878, + "step": 15882 + }, + { + "epoch": 2.8278133903133904, + "grad_norm": 0.9403249025344849, + "learning_rate": 3.957830076272807e-05, + "loss": 0.9536, + "step": 15883 + }, + { + "epoch": 2.827991452991453, + "grad_norm": 0.8735896944999695, + "learning_rate": 3.9567147891424126e-05, + "loss": 0.9113, + "step": 15884 + }, + { + "epoch": 2.828169515669516, + "grad_norm": 0.7758442759513855, + "learning_rate": 3.9555996204176385e-05, + "loss": 0.7336, + "step": 15885 + }, + { + "epoch": 2.828347578347578, + "grad_norm": 0.8632972836494446, + "learning_rate": 3.9544845701203335e-05, + "loss": 0.6883, + "step": 15886 + }, + { + "epoch": 2.828525641025641, + "grad_norm": 0.8639035224914551, + "learning_rate": 3.953369638272345e-05, + "loss": 0.8836, + "step": 15887 + }, + { + "epoch": 2.8287037037037037, + "grad_norm": 0.8131114840507507, + "learning_rate": 3.952254824895514e-05, + "loss": 0.8175, + "step": 15888 + }, + { + "epoch": 2.8288817663817665, + "grad_norm": 0.7421914935112, + "learning_rate": 3.9511401300116904e-05, + "loss": 0.7721, + "step": 15889 + }, + { + "epoch": 2.8290598290598292, + "grad_norm": 0.8358926177024841, + "learning_rate": 3.95002555364271e-05, + "loss": 0.7327, + "step": 15890 + }, + { + "epoch": 2.8292378917378915, + "grad_norm": 0.8913134932518005, + "learning_rate": 3.9489110958104115e-05, + "loss": 1.0415, + "step": 15891 + }, + { + "epoch": 2.8294159544159543, + "grad_norm": 0.9675887227058411, + "learning_rate": 3.94779675653663e-05, + "loss": 0.8295, + "step": 15892 + }, + { + "epoch": 2.829594017094017, + "grad_norm": 0.8618438839912415, + "learning_rate": 3.946682535843199e-05, + "loss": 1.0754, + "step": 15893 + }, + { + "epoch": 2.82977207977208, + "grad_norm": 0.820209801197052, + "learning_rate": 3.945568433751948e-05, + "loss": 0.9484, + "step": 15894 + }, + { + "epoch": 2.8299501424501425, + "grad_norm": 0.8641984462738037, + "learning_rate": 3.944454450284705e-05, + "loss": 0.964, + "step": 15895 + }, + { + "epoch": 2.8301282051282053, + "grad_norm": 0.8229194283485413, + "learning_rate": 3.943340585463303e-05, + "loss": 0.7069, + "step": 15896 + }, + { + "epoch": 2.830306267806268, + "grad_norm": 0.7874621748924255, + "learning_rate": 3.942226839309554e-05, + "loss": 0.7815, + "step": 15897 + }, + { + "epoch": 2.8304843304843303, + "grad_norm": 0.8581945896148682, + "learning_rate": 3.9411132118452896e-05, + "loss": 0.9119, + "step": 15898 + }, + { + "epoch": 2.830662393162393, + "grad_norm": 0.9327018857002258, + "learning_rate": 3.939999703092326e-05, + "loss": 1.1002, + "step": 15899 + }, + { + "epoch": 2.830840455840456, + "grad_norm": 0.7793048024177551, + "learning_rate": 3.9388863130724794e-05, + "loss": 0.6529, + "step": 15900 + }, + { + "epoch": 2.8310185185185186, + "grad_norm": 0.9133790135383606, + "learning_rate": 3.9377730418075645e-05, + "loss": 0.7354, + "step": 15901 + }, + { + "epoch": 2.8311965811965814, + "grad_norm": 0.7800240516662598, + "learning_rate": 3.936659889319394e-05, + "loss": 0.8541, + "step": 15902 + }, + { + "epoch": 2.8313746438746437, + "grad_norm": 0.782433271408081, + "learning_rate": 3.9355468556297737e-05, + "loss": 0.9084, + "step": 15903 + }, + { + "epoch": 2.8315527065527064, + "grad_norm": 0.8926814198493958, + "learning_rate": 3.9344339407605226e-05, + "loss": 1.0252, + "step": 15904 + }, + { + "epoch": 2.831730769230769, + "grad_norm": 0.92144376039505, + "learning_rate": 3.93332114473343e-05, + "loss": 0.7859, + "step": 15905 + }, + { + "epoch": 2.831908831908832, + "grad_norm": 0.7403308749198914, + "learning_rate": 3.932208467570315e-05, + "loss": 0.796, + "step": 15906 + }, + { + "epoch": 2.8320868945868947, + "grad_norm": 0.939708411693573, + "learning_rate": 3.9310959092929636e-05, + "loss": 0.9622, + "step": 15907 + }, + { + "epoch": 2.8322649572649574, + "grad_norm": 0.7546647787094116, + "learning_rate": 3.929983469923184e-05, + "loss": 0.7551, + "step": 15908 + }, + { + "epoch": 2.83244301994302, + "grad_norm": 0.8301447033882141, + "learning_rate": 3.928871149482768e-05, + "loss": 0.9377, + "step": 15909 + }, + { + "epoch": 2.8326210826210825, + "grad_norm": 0.8961447477340698, + "learning_rate": 3.927758947993508e-05, + "loss": 0.8245, + "step": 15910 + }, + { + "epoch": 2.8327991452991452, + "grad_norm": 0.7845988869667053, + "learning_rate": 3.926646865477204e-05, + "loss": 0.8662, + "step": 15911 + }, + { + "epoch": 2.832977207977208, + "grad_norm": 0.9339789152145386, + "learning_rate": 3.925534901955631e-05, + "loss": 0.8533, + "step": 15912 + }, + { + "epoch": 2.8331552706552707, + "grad_norm": 0.936855137348175, + "learning_rate": 3.924423057450587e-05, + "loss": 0.8809, + "step": 15913 + }, + { + "epoch": 2.8333333333333335, + "grad_norm": 0.972535252571106, + "learning_rate": 3.923311331983852e-05, + "loss": 1.0103, + "step": 15914 + }, + { + "epoch": 2.833511396011396, + "grad_norm": 0.9115430116653442, + "learning_rate": 3.922199725577208e-05, + "loss": 1.0041, + "step": 15915 + }, + { + "epoch": 2.8336894586894585, + "grad_norm": 0.8378027081489563, + "learning_rate": 3.921088238252435e-05, + "loss": 0.8475, + "step": 15916 + }, + { + "epoch": 2.8338675213675213, + "grad_norm": 0.9383054375648499, + "learning_rate": 3.91997687003131e-05, + "loss": 0.7305, + "step": 15917 + }, + { + "epoch": 2.834045584045584, + "grad_norm": 0.7996332049369812, + "learning_rate": 3.918865620935609e-05, + "loss": 0.8749, + "step": 15918 + }, + { + "epoch": 2.834223646723647, + "grad_norm": 0.8876177668571472, + "learning_rate": 3.917754490987103e-05, + "loss": 0.8836, + "step": 15919 + }, + { + "epoch": 2.8344017094017095, + "grad_norm": 0.8004130125045776, + "learning_rate": 3.9166434802075594e-05, + "loss": 1.0383, + "step": 15920 + }, + { + "epoch": 2.8345797720797723, + "grad_norm": 0.76146000623703, + "learning_rate": 3.915532588618756e-05, + "loss": 0.9616, + "step": 15921 + }, + { + "epoch": 2.8347578347578346, + "grad_norm": 0.9333193898200989, + "learning_rate": 3.914421816242446e-05, + "loss": 0.7386, + "step": 15922 + }, + { + "epoch": 2.8349358974358974, + "grad_norm": 0.9340601563453674, + "learning_rate": 3.913311163100403e-05, + "loss": 0.8894, + "step": 15923 + }, + { + "epoch": 2.83511396011396, + "grad_norm": 0.8401036858558655, + "learning_rate": 3.912200629214383e-05, + "loss": 0.8366, + "step": 15924 + }, + { + "epoch": 2.835292022792023, + "grad_norm": 0.9298731088638306, + "learning_rate": 3.911090214606146e-05, + "loss": 0.899, + "step": 15925 + }, + { + "epoch": 2.8354700854700856, + "grad_norm": 0.8085874915122986, + "learning_rate": 3.909979919297446e-05, + "loss": 0.7597, + "step": 15926 + }, + { + "epoch": 2.835648148148148, + "grad_norm": 0.7841027975082397, + "learning_rate": 3.9088697433100396e-05, + "loss": 0.897, + "step": 15927 + }, + { + "epoch": 2.8358262108262107, + "grad_norm": 1.0678621530532837, + "learning_rate": 3.907759686665677e-05, + "loss": 0.976, + "step": 15928 + }, + { + "epoch": 2.8360042735042734, + "grad_norm": 0.7748154997825623, + "learning_rate": 3.906649749386106e-05, + "loss": 0.6834, + "step": 15929 + }, + { + "epoch": 2.836182336182336, + "grad_norm": 0.8119567632675171, + "learning_rate": 3.905539931493076e-05, + "loss": 0.8076, + "step": 15930 + }, + { + "epoch": 2.836360398860399, + "grad_norm": 0.8723282814025879, + "learning_rate": 3.9044302330083326e-05, + "loss": 0.9057, + "step": 15931 + }, + { + "epoch": 2.8365384615384617, + "grad_norm": 0.7785065174102783, + "learning_rate": 3.903320653953616e-05, + "loss": 0.8899, + "step": 15932 + }, + { + "epoch": 2.8367165242165244, + "grad_norm": 0.9053105115890503, + "learning_rate": 3.902211194350667e-05, + "loss": 0.9038, + "step": 15933 + }, + { + "epoch": 2.8368945868945867, + "grad_norm": 0.8937689065933228, + "learning_rate": 3.9011018542212216e-05, + "loss": 0.736, + "step": 15934 + }, + { + "epoch": 2.8370726495726495, + "grad_norm": 0.7169269323348999, + "learning_rate": 3.899992633587014e-05, + "loss": 0.6632, + "step": 15935 + }, + { + "epoch": 2.8372507122507122, + "grad_norm": 0.8168412446975708, + "learning_rate": 3.898883532469785e-05, + "loss": 0.8482, + "step": 15936 + }, + { + "epoch": 2.837428774928775, + "grad_norm": 0.7374065518379211, + "learning_rate": 3.897774550891252e-05, + "loss": 0.8141, + "step": 15937 + }, + { + "epoch": 2.8376068376068377, + "grad_norm": 0.8844853043556213, + "learning_rate": 3.8966656888731546e-05, + "loss": 0.8204, + "step": 15938 + }, + { + "epoch": 2.8377849002849, + "grad_norm": 0.9031739234924316, + "learning_rate": 3.895556946437213e-05, + "loss": 0.862, + "step": 15939 + }, + { + "epoch": 2.837962962962963, + "grad_norm": 0.8141549229621887, + "learning_rate": 3.894448323605154e-05, + "loss": 0.6687, + "step": 15940 + }, + { + "epoch": 2.8381410256410255, + "grad_norm": 0.796144962310791, + "learning_rate": 3.893339820398696e-05, + "loss": 0.9021, + "step": 15941 + }, + { + "epoch": 2.8383190883190883, + "grad_norm": 0.8840420246124268, + "learning_rate": 3.8922314368395584e-05, + "loss": 0.9608, + "step": 15942 + }, + { + "epoch": 2.838497150997151, + "grad_norm": 0.8297450542449951, + "learning_rate": 3.891123172949459e-05, + "loss": 0.9442, + "step": 15943 + }, + { + "epoch": 2.838675213675214, + "grad_norm": 0.7875503301620483, + "learning_rate": 3.89001502875011e-05, + "loss": 0.9269, + "step": 15944 + }, + { + "epoch": 2.8388532763532766, + "grad_norm": 0.9460122585296631, + "learning_rate": 3.8889070042632217e-05, + "loss": 0.9459, + "step": 15945 + }, + { + "epoch": 2.8390313390313393, + "grad_norm": 0.8144980669021606, + "learning_rate": 3.887799099510512e-05, + "loss": 0.8409, + "step": 15946 + }, + { + "epoch": 2.8392094017094016, + "grad_norm": 0.8182117342948914, + "learning_rate": 3.886691314513675e-05, + "loss": 0.8093, + "step": 15947 + }, + { + "epoch": 2.8393874643874644, + "grad_norm": 0.8287648558616638, + "learning_rate": 3.885583649294426e-05, + "loss": 0.7792, + "step": 15948 + }, + { + "epoch": 2.839565527065527, + "grad_norm": 0.9165690541267395, + "learning_rate": 3.884476103874464e-05, + "loss": 0.9865, + "step": 15949 + }, + { + "epoch": 2.83974358974359, + "grad_norm": 0.7819885015487671, + "learning_rate": 3.883368678275485e-05, + "loss": 0.7245, + "step": 15950 + }, + { + "epoch": 2.839921652421652, + "grad_norm": 0.8354606628417969, + "learning_rate": 3.882261372519198e-05, + "loss": 1.0513, + "step": 15951 + }, + { + "epoch": 2.840099715099715, + "grad_norm": 0.7606815099716187, + "learning_rate": 3.881154186627284e-05, + "loss": 0.9357, + "step": 15952 + }, + { + "epoch": 2.8402777777777777, + "grad_norm": 0.9649691581726074, + "learning_rate": 3.88004712062145e-05, + "loss": 0.8756, + "step": 15953 + }, + { + "epoch": 2.8404558404558404, + "grad_norm": 0.8770344853401184, + "learning_rate": 3.878940174523371e-05, + "loss": 0.868, + "step": 15954 + }, + { + "epoch": 2.840633903133903, + "grad_norm": 0.898287832736969, + "learning_rate": 3.877833348354749e-05, + "loss": 0.8734, + "step": 15955 + }, + { + "epoch": 2.840811965811966, + "grad_norm": 0.84062260389328, + "learning_rate": 3.876726642137264e-05, + "loss": 0.9362, + "step": 15956 + }, + { + "epoch": 2.8409900284900287, + "grad_norm": 0.7898240685462952, + "learning_rate": 3.8756200558926013e-05, + "loss": 0.7788, + "step": 15957 + }, + { + "epoch": 2.8411680911680914, + "grad_norm": 0.7237298488616943, + "learning_rate": 3.874513589642441e-05, + "loss": 0.8426, + "step": 15958 + }, + { + "epoch": 2.8413461538461537, + "grad_norm": 0.9025090932846069, + "learning_rate": 3.873407243408462e-05, + "loss": 0.7135, + "step": 15959 + }, + { + "epoch": 2.8415242165242165, + "grad_norm": 0.807295560836792, + "learning_rate": 3.872301017212337e-05, + "loss": 0.6889, + "step": 15960 + }, + { + "epoch": 2.8417022792022792, + "grad_norm": 0.8537244200706482, + "learning_rate": 3.8711949110757525e-05, + "loss": 0.798, + "step": 15961 + }, + { + "epoch": 2.841880341880342, + "grad_norm": 0.8148910999298096, + "learning_rate": 3.870088925020366e-05, + "loss": 0.8783, + "step": 15962 + }, + { + "epoch": 2.8420584045584043, + "grad_norm": 0.8254446983337402, + "learning_rate": 3.868983059067859e-05, + "loss": 0.7043, + "step": 15963 + }, + { + "epoch": 2.842236467236467, + "grad_norm": 0.8392706513404846, + "learning_rate": 3.867877313239886e-05, + "loss": 0.9942, + "step": 15964 + }, + { + "epoch": 2.84241452991453, + "grad_norm": 0.8974948525428772, + "learning_rate": 3.8667716875581217e-05, + "loss": 0.7646, + "step": 15965 + }, + { + "epoch": 2.8425925925925926, + "grad_norm": 0.9764110445976257, + "learning_rate": 3.8656661820442264e-05, + "loss": 0.8803, + "step": 15966 + }, + { + "epoch": 2.8427706552706553, + "grad_norm": 0.9663669466972351, + "learning_rate": 3.864560796719855e-05, + "loss": 0.8764, + "step": 15967 + }, + { + "epoch": 2.842948717948718, + "grad_norm": 0.837733268737793, + "learning_rate": 3.863455531606677e-05, + "loss": 0.8992, + "step": 15968 + }, + { + "epoch": 2.843126780626781, + "grad_norm": 0.8458481431007385, + "learning_rate": 3.8623503867263335e-05, + "loss": 0.9025, + "step": 15969 + }, + { + "epoch": 2.8433048433048436, + "grad_norm": 0.901089072227478, + "learning_rate": 3.861245362100488e-05, + "loss": 0.8025, + "step": 15970 + }, + { + "epoch": 2.843482905982906, + "grad_norm": 0.9032089114189148, + "learning_rate": 3.860140457750786e-05, + "loss": 0.7217, + "step": 15971 + }, + { + "epoch": 2.8436609686609686, + "grad_norm": 0.7998839616775513, + "learning_rate": 3.859035673698879e-05, + "loss": 0.9127, + "step": 15972 + }, + { + "epoch": 2.8438390313390314, + "grad_norm": 0.8568583726882935, + "learning_rate": 3.85793100996641e-05, + "loss": 0.8847, + "step": 15973 + }, + { + "epoch": 2.844017094017094, + "grad_norm": 0.8720089793205261, + "learning_rate": 3.856826466575024e-05, + "loss": 0.7822, + "step": 15974 + }, + { + "epoch": 2.8441951566951564, + "grad_norm": 0.8872382640838623, + "learning_rate": 3.8557220435463594e-05, + "loss": 0.9601, + "step": 15975 + }, + { + "epoch": 2.844373219373219, + "grad_norm": 1.1950596570968628, + "learning_rate": 3.8546177409020634e-05, + "loss": 1.075, + "step": 15976 + }, + { + "epoch": 2.844551282051282, + "grad_norm": 0.9111549854278564, + "learning_rate": 3.85351355866376e-05, + "loss": 1.0103, + "step": 15977 + }, + { + "epoch": 2.8447293447293447, + "grad_norm": 0.9310214519500732, + "learning_rate": 3.852409496853099e-05, + "loss": 1.0163, + "step": 15978 + }, + { + "epoch": 2.8449074074074074, + "grad_norm": 0.8177474737167358, + "learning_rate": 3.851305555491695e-05, + "loss": 0.7488, + "step": 15979 + }, + { + "epoch": 2.84508547008547, + "grad_norm": 0.9321249127388, + "learning_rate": 3.85020173460119e-05, + "loss": 0.9914, + "step": 15980 + }, + { + "epoch": 2.845263532763533, + "grad_norm": 0.7649266719818115, + "learning_rate": 3.849098034203206e-05, + "loss": 0.692, + "step": 15981 + }, + { + "epoch": 2.8454415954415957, + "grad_norm": 0.7714266777038574, + "learning_rate": 3.847994454319369e-05, + "loss": 0.8859, + "step": 15982 + }, + { + "epoch": 2.845619658119658, + "grad_norm": 0.9535303711891174, + "learning_rate": 3.846890994971302e-05, + "loss": 0.8992, + "step": 15983 + }, + { + "epoch": 2.8457977207977208, + "grad_norm": 0.8171879649162292, + "learning_rate": 3.845787656180623e-05, + "loss": 0.7125, + "step": 15984 + }, + { + "epoch": 2.8459757834757835, + "grad_norm": 0.8546884655952454, + "learning_rate": 3.8446844379689464e-05, + "loss": 0.9895, + "step": 15985 + }, + { + "epoch": 2.8461538461538463, + "grad_norm": 0.9110364317893982, + "learning_rate": 3.843581340357899e-05, + "loss": 1.0702, + "step": 15986 + }, + { + "epoch": 2.8463319088319086, + "grad_norm": 0.8862065672874451, + "learning_rate": 3.84247836336908e-05, + "loss": 0.9138, + "step": 15987 + }, + { + "epoch": 2.8465099715099713, + "grad_norm": 0.8485249876976013, + "learning_rate": 3.84137550702411e-05, + "loss": 1.2831, + "step": 15988 + }, + { + "epoch": 2.846688034188034, + "grad_norm": 0.8271495699882507, + "learning_rate": 3.840272771344593e-05, + "loss": 0.9497, + "step": 15989 + }, + { + "epoch": 2.846866096866097, + "grad_norm": 0.7829293608665466, + "learning_rate": 3.839170156352135e-05, + "loss": 0.6503, + "step": 15990 + }, + { + "epoch": 2.8470441595441596, + "grad_norm": 0.9366582036018372, + "learning_rate": 3.838067662068341e-05, + "loss": 0.7805, + "step": 15991 + }, + { + "epoch": 2.8472222222222223, + "grad_norm": 0.8666117787361145, + "learning_rate": 3.836965288514807e-05, + "loss": 0.7721, + "step": 15992 + }, + { + "epoch": 2.847400284900285, + "grad_norm": 0.7855546474456787, + "learning_rate": 3.835863035713142e-05, + "loss": 0.7457, + "step": 15993 + }, + { + "epoch": 2.847578347578348, + "grad_norm": 0.8234511017799377, + "learning_rate": 3.8347609036849284e-05, + "loss": 0.8937, + "step": 15994 + }, + { + "epoch": 2.84775641025641, + "grad_norm": 0.8896345496177673, + "learning_rate": 3.833658892451773e-05, + "loss": 0.9146, + "step": 15995 + }, + { + "epoch": 2.847934472934473, + "grad_norm": 0.8099349737167358, + "learning_rate": 3.83255700203526e-05, + "loss": 0.8353, + "step": 15996 + }, + { + "epoch": 2.8481125356125356, + "grad_norm": 0.874100387096405, + "learning_rate": 3.831455232456982e-05, + "loss": 0.6829, + "step": 15997 + }, + { + "epoch": 2.8482905982905984, + "grad_norm": 0.9338345527648926, + "learning_rate": 3.830353583738524e-05, + "loss": 1.0345, + "step": 15998 + }, + { + "epoch": 2.8484686609686607, + "grad_norm": 0.7876978516578674, + "learning_rate": 3.829252055901472e-05, + "loss": 0.6703, + "step": 15999 + }, + { + "epoch": 2.8486467236467234, + "grad_norm": 0.8565872311592102, + "learning_rate": 3.828150648967408e-05, + "loss": 0.8227, + "step": 16000 + }, + { + "epoch": 2.848824786324786, + "grad_norm": 1.0180596113204956, + "learning_rate": 3.82704936295791e-05, + "loss": 0.8933, + "step": 16001 + }, + { + "epoch": 2.849002849002849, + "grad_norm": 0.8450096845626831, + "learning_rate": 3.825948197894553e-05, + "loss": 0.851, + "step": 16002 + }, + { + "epoch": 2.8491809116809117, + "grad_norm": 0.7936033010482788, + "learning_rate": 3.824847153798923e-05, + "loss": 0.8573, + "step": 16003 + }, + { + "epoch": 2.8493589743589745, + "grad_norm": 0.9499372839927673, + "learning_rate": 3.8237462306925774e-05, + "loss": 0.8269, + "step": 16004 + }, + { + "epoch": 2.849537037037037, + "grad_norm": 0.874855101108551, + "learning_rate": 3.822645428597099e-05, + "loss": 0.9657, + "step": 16005 + }, + { + "epoch": 2.8497150997151, + "grad_norm": 0.8966119885444641, + "learning_rate": 3.8215447475340506e-05, + "loss": 0.9239, + "step": 16006 + }, + { + "epoch": 2.8498931623931623, + "grad_norm": 0.8341490030288696, + "learning_rate": 3.820444187524994e-05, + "loss": 1.0051, + "step": 16007 + }, + { + "epoch": 2.850071225071225, + "grad_norm": 0.7965613007545471, + "learning_rate": 3.8193437485915054e-05, + "loss": 0.8591, + "step": 16008 + }, + { + "epoch": 2.8502492877492878, + "grad_norm": 0.7846593856811523, + "learning_rate": 3.818243430755128e-05, + "loss": 0.8095, + "step": 16009 + }, + { + "epoch": 2.8504273504273505, + "grad_norm": 0.9422695636749268, + "learning_rate": 3.8171432340374334e-05, + "loss": 0.8902, + "step": 16010 + }, + { + "epoch": 2.8506054131054133, + "grad_norm": 0.8810960650444031, + "learning_rate": 3.8160431584599744e-05, + "loss": 0.8483, + "step": 16011 + }, + { + "epoch": 2.8507834757834756, + "grad_norm": 0.8513348698616028, + "learning_rate": 3.814943204044302e-05, + "loss": 0.832, + "step": 16012 + }, + { + "epoch": 2.8509615384615383, + "grad_norm": 0.8906846046447754, + "learning_rate": 3.8138433708119704e-05, + "loss": 0.9702, + "step": 16013 + }, + { + "epoch": 2.851139601139601, + "grad_norm": 0.9517511129379272, + "learning_rate": 3.812743658784526e-05, + "loss": 0.8138, + "step": 16014 + }, + { + "epoch": 2.851317663817664, + "grad_norm": 0.7989702820777893, + "learning_rate": 3.811644067983517e-05, + "loss": 0.8653, + "step": 16015 + }, + { + "epoch": 2.8514957264957266, + "grad_norm": 0.8255589008331299, + "learning_rate": 3.8105445984304874e-05, + "loss": 0.9456, + "step": 16016 + }, + { + "epoch": 2.8516737891737893, + "grad_norm": 0.7919938564300537, + "learning_rate": 3.809445250146977e-05, + "loss": 0.6261, + "step": 16017 + }, + { + "epoch": 2.851851851851852, + "grad_norm": 0.866316020488739, + "learning_rate": 3.808346023154532e-05, + "loss": 0.8171, + "step": 16018 + }, + { + "epoch": 2.8520299145299144, + "grad_norm": 1.0050057172775269, + "learning_rate": 3.8072469174746794e-05, + "loss": 0.9094, + "step": 16019 + }, + { + "epoch": 2.852207977207977, + "grad_norm": 0.8405657410621643, + "learning_rate": 3.806147933128962e-05, + "loss": 0.7737, + "step": 16020 + }, + { + "epoch": 2.85238603988604, + "grad_norm": 0.8127378821372986, + "learning_rate": 3.8050490701389085e-05, + "loss": 0.9102, + "step": 16021 + }, + { + "epoch": 2.8525641025641026, + "grad_norm": 0.8622255921363831, + "learning_rate": 3.8039503285260506e-05, + "loss": 0.8815, + "step": 16022 + }, + { + "epoch": 2.8527421652421654, + "grad_norm": 0.8802367448806763, + "learning_rate": 3.802851708311913e-05, + "loss": 1.0123, + "step": 16023 + }, + { + "epoch": 2.8529202279202277, + "grad_norm": 0.908149778842926, + "learning_rate": 3.801753209518024e-05, + "loss": 0.6808, + "step": 16024 + }, + { + "epoch": 2.8530982905982905, + "grad_norm": 0.9346339702606201, + "learning_rate": 3.8006548321659055e-05, + "loss": 1.1107, + "step": 16025 + }, + { + "epoch": 2.853276353276353, + "grad_norm": 0.945125937461853, + "learning_rate": 3.799556576277077e-05, + "loss": 0.6578, + "step": 16026 + }, + { + "epoch": 2.853454415954416, + "grad_norm": 0.8294890522956848, + "learning_rate": 3.798458441873054e-05, + "loss": 0.869, + "step": 16027 + }, + { + "epoch": 2.8536324786324787, + "grad_norm": 0.7922961115837097, + "learning_rate": 3.797360428975358e-05, + "loss": 0.718, + "step": 16028 + }, + { + "epoch": 2.8538105413105415, + "grad_norm": 1.0540844202041626, + "learning_rate": 3.7962625376055005e-05, + "loss": 0.8287, + "step": 16029 + }, + { + "epoch": 2.853988603988604, + "grad_norm": 0.9409742951393127, + "learning_rate": 3.795164767784991e-05, + "loss": 1.0484, + "step": 16030 + }, + { + "epoch": 2.8541666666666665, + "grad_norm": 0.7328341603279114, + "learning_rate": 3.7940671195353385e-05, + "loss": 0.7603, + "step": 16031 + }, + { + "epoch": 2.8543447293447293, + "grad_norm": 0.9151208996772766, + "learning_rate": 3.792969592878045e-05, + "loss": 0.7523, + "step": 16032 + }, + { + "epoch": 2.854522792022792, + "grad_norm": 0.7935783267021179, + "learning_rate": 3.791872187834626e-05, + "loss": 0.7559, + "step": 16033 + }, + { + "epoch": 2.8547008547008548, + "grad_norm": 0.8030906915664673, + "learning_rate": 3.790774904426568e-05, + "loss": 0.7905, + "step": 16034 + }, + { + "epoch": 2.8548789173789175, + "grad_norm": 0.8756175637245178, + "learning_rate": 3.789677742675384e-05, + "loss": 0.9067, + "step": 16035 + }, + { + "epoch": 2.85505698005698, + "grad_norm": 0.7602807283401489, + "learning_rate": 3.788580702602558e-05, + "loss": 0.9192, + "step": 16036 + }, + { + "epoch": 2.8552350427350426, + "grad_norm": 0.9411010146141052, + "learning_rate": 3.787483784229592e-05, + "loss": 0.8192, + "step": 16037 + }, + { + "epoch": 2.8554131054131053, + "grad_norm": 0.9473391175270081, + "learning_rate": 3.786386987577976e-05, + "loss": 0.7845, + "step": 16038 + }, + { + "epoch": 2.855591168091168, + "grad_norm": 0.9226218461990356, + "learning_rate": 3.7852903126692e-05, + "loss": 0.8712, + "step": 16039 + }, + { + "epoch": 2.855769230769231, + "grad_norm": 0.9519350528717041, + "learning_rate": 3.78419375952475e-05, + "loss": 0.9584, + "step": 16040 + }, + { + "epoch": 2.8559472934472936, + "grad_norm": 0.825547456741333, + "learning_rate": 3.783097328166111e-05, + "loss": 0.9279, + "step": 16041 + }, + { + "epoch": 2.8561253561253563, + "grad_norm": 0.8645279407501221, + "learning_rate": 3.782001018614763e-05, + "loss": 0.83, + "step": 16042 + }, + { + "epoch": 2.8563034188034186, + "grad_norm": 0.827126145362854, + "learning_rate": 3.7809048308921936e-05, + "loss": 0.7661, + "step": 16043 + }, + { + "epoch": 2.8564814814814814, + "grad_norm": 0.9441137909889221, + "learning_rate": 3.779808765019869e-05, + "loss": 0.8745, + "step": 16044 + }, + { + "epoch": 2.856659544159544, + "grad_norm": 0.8505343794822693, + "learning_rate": 3.7787128210192736e-05, + "loss": 0.8176, + "step": 16045 + }, + { + "epoch": 2.856837606837607, + "grad_norm": 0.8797150254249573, + "learning_rate": 3.777616998911876e-05, + "loss": 0.7018, + "step": 16046 + }, + { + "epoch": 2.8570156695156697, + "grad_norm": 0.8386834263801575, + "learning_rate": 3.776521298719144e-05, + "loss": 0.9805, + "step": 16047 + }, + { + "epoch": 2.857193732193732, + "grad_norm": 0.818373441696167, + "learning_rate": 3.775425720462558e-05, + "loss": 0.7782, + "step": 16048 + }, + { + "epoch": 2.8573717948717947, + "grad_norm": 0.760405957698822, + "learning_rate": 3.774330264163566e-05, + "loss": 0.7283, + "step": 16049 + }, + { + "epoch": 2.8575498575498575, + "grad_norm": 0.9082552194595337, + "learning_rate": 3.7732349298436465e-05, + "loss": 0.7594, + "step": 16050 + }, + { + "epoch": 2.85772792022792, + "grad_norm": 0.859868586063385, + "learning_rate": 3.7721397175242477e-05, + "loss": 0.7841, + "step": 16051 + }, + { + "epoch": 2.857905982905983, + "grad_norm": 1.031545877456665, + "learning_rate": 3.771044627226836e-05, + "loss": 1.2157, + "step": 16052 + }, + { + "epoch": 2.8580840455840457, + "grad_norm": 0.859491765499115, + "learning_rate": 3.769949658972867e-05, + "loss": 0.7942, + "step": 16053 + }, + { + "epoch": 2.8582621082621085, + "grad_norm": 0.775382936000824, + "learning_rate": 3.768854812783791e-05, + "loss": 0.7321, + "step": 16054 + }, + { + "epoch": 2.8584401709401708, + "grad_norm": 0.9268868565559387, + "learning_rate": 3.767760088681062e-05, + "loss": 1.0104, + "step": 16055 + }, + { + "epoch": 2.8586182336182335, + "grad_norm": 0.8408828973770142, + "learning_rate": 3.7666654866861274e-05, + "loss": 0.5298, + "step": 16056 + }, + { + "epoch": 2.8587962962962963, + "grad_norm": 0.8417157530784607, + "learning_rate": 3.76557100682043e-05, + "loss": 0.9123, + "step": 16057 + }, + { + "epoch": 2.858974358974359, + "grad_norm": 0.8079593777656555, + "learning_rate": 3.764476649105425e-05, + "loss": 0.794, + "step": 16058 + }, + { + "epoch": 2.859152421652422, + "grad_norm": 0.9449031352996826, + "learning_rate": 3.763382413562541e-05, + "loss": 1.1281, + "step": 16059 + }, + { + "epoch": 2.859330484330484, + "grad_norm": 0.8985004425048828, + "learning_rate": 3.762288300213228e-05, + "loss": 0.929, + "step": 16060 + }, + { + "epoch": 2.859508547008547, + "grad_norm": 0.9850391149520874, + "learning_rate": 3.761194309078913e-05, + "loss": 0.9796, + "step": 16061 + }, + { + "epoch": 2.8596866096866096, + "grad_norm": 0.9231089949607849, + "learning_rate": 3.760100440181038e-05, + "loss": 0.6981, + "step": 16062 + }, + { + "epoch": 2.8598646723646723, + "grad_norm": 0.8458681702613831, + "learning_rate": 3.759006693541033e-05, + "loss": 0.9502, + "step": 16063 + }, + { + "epoch": 2.860042735042735, + "grad_norm": 0.8494541645050049, + "learning_rate": 3.7579130691803266e-05, + "loss": 0.7477, + "step": 16064 + }, + { + "epoch": 2.860220797720798, + "grad_norm": 0.879878580570221, + "learning_rate": 3.756819567120348e-05, + "loss": 0.8426, + "step": 16065 + }, + { + "epoch": 2.8603988603988606, + "grad_norm": 0.8161541223526001, + "learning_rate": 3.7557261873825155e-05, + "loss": 0.7411, + "step": 16066 + }, + { + "epoch": 2.8605769230769234, + "grad_norm": 0.9438506364822388, + "learning_rate": 3.754632929988262e-05, + "loss": 0.8494, + "step": 16067 + }, + { + "epoch": 2.8607549857549857, + "grad_norm": 0.8552418351173401, + "learning_rate": 3.753539794959002e-05, + "loss": 0.766, + "step": 16068 + }, + { + "epoch": 2.8609330484330484, + "grad_norm": 0.8670600056648254, + "learning_rate": 3.7524467823161546e-05, + "loss": 0.7462, + "step": 16069 + }, + { + "epoch": 2.861111111111111, + "grad_norm": 0.7906678318977356, + "learning_rate": 3.751353892081134e-05, + "loss": 0.791, + "step": 16070 + }, + { + "epoch": 2.861289173789174, + "grad_norm": 0.8461915254592896, + "learning_rate": 3.7502611242753536e-05, + "loss": 0.6979, + "step": 16071 + }, + { + "epoch": 2.861467236467236, + "grad_norm": 0.8197309970855713, + "learning_rate": 3.749168478920223e-05, + "loss": 0.8095, + "step": 16072 + }, + { + "epoch": 2.861645299145299, + "grad_norm": 0.9489047527313232, + "learning_rate": 3.7480759560371516e-05, + "loss": 0.9462, + "step": 16073 + }, + { + "epoch": 2.8618233618233617, + "grad_norm": 0.8539329767227173, + "learning_rate": 3.7469835556475405e-05, + "loss": 0.853, + "step": 16074 + }, + { + "epoch": 2.8620014245014245, + "grad_norm": 0.9104743003845215, + "learning_rate": 3.745891277772805e-05, + "loss": 0.757, + "step": 16075 + }, + { + "epoch": 2.8621794871794872, + "grad_norm": 0.8277523517608643, + "learning_rate": 3.744799122434332e-05, + "loss": 0.8073, + "step": 16076 + }, + { + "epoch": 2.86235754985755, + "grad_norm": 0.866422176361084, + "learning_rate": 3.743707089653527e-05, + "loss": 0.7201, + "step": 16077 + }, + { + "epoch": 2.8625356125356127, + "grad_norm": 0.8722748160362244, + "learning_rate": 3.742615179451787e-05, + "loss": 0.7901, + "step": 16078 + }, + { + "epoch": 2.8627136752136755, + "grad_norm": 0.74676513671875, + "learning_rate": 3.741523391850504e-05, + "loss": 0.7558, + "step": 16079 + }, + { + "epoch": 2.862891737891738, + "grad_norm": 0.7674166560173035, + "learning_rate": 3.740431726871069e-05, + "loss": 0.8699, + "step": 16080 + }, + { + "epoch": 2.8630698005698005, + "grad_norm": 0.9028998613357544, + "learning_rate": 3.739340184534871e-05, + "loss": 1.136, + "step": 16081 + }, + { + "epoch": 2.8632478632478633, + "grad_norm": 0.8240773677825928, + "learning_rate": 3.7382487648632936e-05, + "loss": 0.9357, + "step": 16082 + }, + { + "epoch": 2.863425925925926, + "grad_norm": 0.8877659440040588, + "learning_rate": 3.737157467877731e-05, + "loss": 0.9167, + "step": 16083 + }, + { + "epoch": 2.8636039886039883, + "grad_norm": 0.9677366614341736, + "learning_rate": 3.73606629359955e-05, + "loss": 0.8342, + "step": 16084 + }, + { + "epoch": 2.863782051282051, + "grad_norm": 0.8721164464950562, + "learning_rate": 3.734975242050146e-05, + "loss": 0.9195, + "step": 16085 + }, + { + "epoch": 2.863960113960114, + "grad_norm": 0.9791151881217957, + "learning_rate": 3.733884313250879e-05, + "loss": 1.0011, + "step": 16086 + }, + { + "epoch": 2.8641381766381766, + "grad_norm": 0.7869369983673096, + "learning_rate": 3.7327935072231366e-05, + "loss": 0.7998, + "step": 16087 + }, + { + "epoch": 2.8643162393162394, + "grad_norm": 0.891656756401062, + "learning_rate": 3.731702823988287e-05, + "loss": 0.7287, + "step": 16088 + }, + { + "epoch": 2.864494301994302, + "grad_norm": 0.8720460534095764, + "learning_rate": 3.7306122635676955e-05, + "loss": 0.9492, + "step": 16089 + }, + { + "epoch": 2.864672364672365, + "grad_norm": 0.7878959774971008, + "learning_rate": 3.72952182598274e-05, + "loss": 0.7652, + "step": 16090 + }, + { + "epoch": 2.8648504273504276, + "grad_norm": 0.9350453019142151, + "learning_rate": 3.728431511254772e-05, + "loss": 0.8661, + "step": 16091 + }, + { + "epoch": 2.86502849002849, + "grad_norm": 0.7575289011001587, + "learning_rate": 3.727341319405163e-05, + "loss": 0.7943, + "step": 16092 + }, + { + "epoch": 2.8652065527065527, + "grad_norm": 0.8256776928901672, + "learning_rate": 3.7262512504552716e-05, + "loss": 1.0025, + "step": 16093 + }, + { + "epoch": 2.8653846153846154, + "grad_norm": 0.7274962067604065, + "learning_rate": 3.7251613044264536e-05, + "loss": 0.7628, + "step": 16094 + }, + { + "epoch": 2.865562678062678, + "grad_norm": 0.867734968662262, + "learning_rate": 3.7240714813400646e-05, + "loss": 0.8931, + "step": 16095 + }, + { + "epoch": 2.8657407407407405, + "grad_norm": 0.8225845098495483, + "learning_rate": 3.722981781217458e-05, + "loss": 0.7934, + "step": 16096 + }, + { + "epoch": 2.8659188034188032, + "grad_norm": 0.8180573582649231, + "learning_rate": 3.721892204079985e-05, + "loss": 0.8119, + "step": 16097 + }, + { + "epoch": 2.866096866096866, + "grad_norm": 1.0235565900802612, + "learning_rate": 3.720802749948993e-05, + "loss": 0.8281, + "step": 16098 + }, + { + "epoch": 2.8662749287749287, + "grad_norm": 0.7290656566619873, + "learning_rate": 3.719713418845823e-05, + "loss": 0.7099, + "step": 16099 + }, + { + "epoch": 2.8664529914529915, + "grad_norm": 0.8408772349357605, + "learning_rate": 3.718624210791828e-05, + "loss": 0.8631, + "step": 16100 + }, + { + "epoch": 2.8666310541310542, + "grad_norm": 0.8182529807090759, + "learning_rate": 3.717535125808338e-05, + "loss": 0.7584, + "step": 16101 + }, + { + "epoch": 2.866809116809117, + "grad_norm": 0.8381599187850952, + "learning_rate": 3.716446163916699e-05, + "loss": 0.8735, + "step": 16102 + }, + { + "epoch": 2.8669871794871797, + "grad_norm": 0.8975555300712585, + "learning_rate": 3.715357325138245e-05, + "loss": 0.7564, + "step": 16103 + }, + { + "epoch": 2.867165242165242, + "grad_norm": 0.9531118869781494, + "learning_rate": 3.714268609494309e-05, + "loss": 0.7627, + "step": 16104 + }, + { + "epoch": 2.867343304843305, + "grad_norm": 0.853065550327301, + "learning_rate": 3.7131800170062216e-05, + "loss": 0.8001, + "step": 16105 + }, + { + "epoch": 2.8675213675213675, + "grad_norm": 0.788351833820343, + "learning_rate": 3.7120915476953085e-05, + "loss": 0.7935, + "step": 16106 + }, + { + "epoch": 2.8676994301994303, + "grad_norm": 0.9121149778366089, + "learning_rate": 3.711003201582908e-05, + "loss": 0.9212, + "step": 16107 + }, + { + "epoch": 2.8678774928774926, + "grad_norm": 0.8156226277351379, + "learning_rate": 3.7099149786903263e-05, + "loss": 0.86, + "step": 16108 + }, + { + "epoch": 2.8680555555555554, + "grad_norm": 0.8555662035942078, + "learning_rate": 3.708826879038899e-05, + "loss": 0.8872, + "step": 16109 + }, + { + "epoch": 2.868233618233618, + "grad_norm": 1.0395163297653198, + "learning_rate": 3.70773890264994e-05, + "loss": 0.9794, + "step": 16110 + }, + { + "epoch": 2.868411680911681, + "grad_norm": 0.7535551190376282, + "learning_rate": 3.706651049544766e-05, + "loss": 0.91, + "step": 16111 + }, + { + "epoch": 2.8685897435897436, + "grad_norm": 1.0145034790039062, + "learning_rate": 3.705563319744691e-05, + "loss": 0.9386, + "step": 16112 + }, + { + "epoch": 2.8687678062678064, + "grad_norm": 0.8577025532722473, + "learning_rate": 3.704475713271029e-05, + "loss": 1.0491, + "step": 16113 + }, + { + "epoch": 2.868945868945869, + "grad_norm": 0.8291150331497192, + "learning_rate": 3.7033882301450815e-05, + "loss": 0.8418, + "step": 16114 + }, + { + "epoch": 2.869123931623932, + "grad_norm": 0.7628613710403442, + "learning_rate": 3.70230087038817e-05, + "loss": 0.681, + "step": 16115 + }, + { + "epoch": 2.869301994301994, + "grad_norm": 0.8664639592170715, + "learning_rate": 3.701213634021583e-05, + "loss": 1.0247, + "step": 16116 + }, + { + "epoch": 2.869480056980057, + "grad_norm": 0.9613258838653564, + "learning_rate": 3.700126521066635e-05, + "loss": 1.1058, + "step": 16117 + }, + { + "epoch": 2.8696581196581197, + "grad_norm": 0.8279051780700684, + "learning_rate": 3.699039531544619e-05, + "loss": 0.859, + "step": 16118 + }, + { + "epoch": 2.8698361823361824, + "grad_norm": 0.8285593390464783, + "learning_rate": 3.697952665476836e-05, + "loss": 0.9008, + "step": 16119 + }, + { + "epoch": 2.870014245014245, + "grad_norm": 0.9056670069694519, + "learning_rate": 3.696865922884578e-05, + "loss": 0.965, + "step": 16120 + }, + { + "epoch": 2.8701923076923075, + "grad_norm": 0.7092664837837219, + "learning_rate": 3.69577930378914e-05, + "loss": 0.7417, + "step": 16121 + }, + { + "epoch": 2.8703703703703702, + "grad_norm": 0.8726393580436707, + "learning_rate": 3.6946928082118096e-05, + "loss": 0.8582, + "step": 16122 + }, + { + "epoch": 2.870548433048433, + "grad_norm": 1.0046098232269287, + "learning_rate": 3.693606436173875e-05, + "loss": 0.9616, + "step": 16123 + }, + { + "epoch": 2.8707264957264957, + "grad_norm": 0.7739760875701904, + "learning_rate": 3.69252018769662e-05, + "loss": 0.8775, + "step": 16124 + }, + { + "epoch": 2.8709045584045585, + "grad_norm": 0.9054580926895142, + "learning_rate": 3.6914340628013344e-05, + "loss": 0.9205, + "step": 16125 + }, + { + "epoch": 2.8710826210826212, + "grad_norm": 0.8324142694473267, + "learning_rate": 3.690348061509288e-05, + "loss": 0.8712, + "step": 16126 + }, + { + "epoch": 2.871260683760684, + "grad_norm": 0.9162326455116272, + "learning_rate": 3.6892621838417664e-05, + "loss": 0.9102, + "step": 16127 + }, + { + "epoch": 2.8714387464387463, + "grad_norm": 0.8579963445663452, + "learning_rate": 3.688176429820044e-05, + "loss": 0.8337, + "step": 16128 + }, + { + "epoch": 2.871616809116809, + "grad_norm": 0.7649274468421936, + "learning_rate": 3.687090799465388e-05, + "loss": 0.6982, + "step": 16129 + }, + { + "epoch": 2.871794871794872, + "grad_norm": 0.9612696766853333, + "learning_rate": 3.6860052927990816e-05, + "loss": 0.8779, + "step": 16130 + }, + { + "epoch": 2.8719729344729346, + "grad_norm": 0.982455313205719, + "learning_rate": 3.6849199098423795e-05, + "loss": 0.8145, + "step": 16131 + }, + { + "epoch": 2.8721509971509973, + "grad_norm": 0.7292434573173523, + "learning_rate": 3.6838346506165587e-05, + "loss": 0.6711, + "step": 16132 + }, + { + "epoch": 2.8723290598290596, + "grad_norm": 0.877310574054718, + "learning_rate": 3.68274951514287e-05, + "loss": 0.7682, + "step": 16133 + }, + { + "epoch": 2.8725071225071224, + "grad_norm": 0.9633384943008423, + "learning_rate": 3.681664503442586e-05, + "loss": 1.0046, + "step": 16134 + }, + { + "epoch": 2.872685185185185, + "grad_norm": 0.940661609172821, + "learning_rate": 3.680579615536961e-05, + "loss": 0.897, + "step": 16135 + }, + { + "epoch": 2.872863247863248, + "grad_norm": 1.0335214138031006, + "learning_rate": 3.6794948514472505e-05, + "loss": 0.8836, + "step": 16136 + }, + { + "epoch": 2.8730413105413106, + "grad_norm": 0.8682044148445129, + "learning_rate": 3.6784102111947084e-05, + "loss": 0.7733, + "step": 16137 + }, + { + "epoch": 2.8732193732193734, + "grad_norm": 0.8767847418785095, + "learning_rate": 3.677325694800586e-05, + "loss": 0.8828, + "step": 16138 + }, + { + "epoch": 2.873397435897436, + "grad_norm": 0.954585075378418, + "learning_rate": 3.6762413022861305e-05, + "loss": 0.9294, + "step": 16139 + }, + { + "epoch": 2.8735754985754984, + "grad_norm": 0.8497310876846313, + "learning_rate": 3.675157033672596e-05, + "loss": 0.9396, + "step": 16140 + }, + { + "epoch": 2.873753561253561, + "grad_norm": 0.7619023323059082, + "learning_rate": 3.674072888981214e-05, + "loss": 0.7467, + "step": 16141 + }, + { + "epoch": 2.873931623931624, + "grad_norm": 0.9939205646514893, + "learning_rate": 3.6729888682332394e-05, + "loss": 0.9122, + "step": 16142 + }, + { + "epoch": 2.8741096866096867, + "grad_norm": 0.943192183971405, + "learning_rate": 3.671904971449899e-05, + "loss": 0.8878, + "step": 16143 + }, + { + "epoch": 2.8742877492877494, + "grad_norm": 0.8002169728279114, + "learning_rate": 3.6708211986524365e-05, + "loss": 0.8337, + "step": 16144 + }, + { + "epoch": 2.8744658119658117, + "grad_norm": 0.7558008432388306, + "learning_rate": 3.669737549862087e-05, + "loss": 0.6592, + "step": 16145 + }, + { + "epoch": 2.8746438746438745, + "grad_norm": 0.8732983469963074, + "learning_rate": 3.6686540251000756e-05, + "loss": 0.9166, + "step": 16146 + }, + { + "epoch": 2.8748219373219372, + "grad_norm": 0.8272808194160461, + "learning_rate": 3.667570624387643e-05, + "loss": 0.6342, + "step": 16147 + }, + { + "epoch": 2.875, + "grad_norm": 0.8012139797210693, + "learning_rate": 3.666487347746004e-05, + "loss": 0.8436, + "step": 16148 + }, + { + "epoch": 2.8751780626780628, + "grad_norm": 0.8791360259056091, + "learning_rate": 3.66540419519639e-05, + "loss": 0.8634, + "step": 16149 + }, + { + "epoch": 2.8753561253561255, + "grad_norm": 0.8274601101875305, + "learning_rate": 3.6643211667600244e-05, + "loss": 0.7488, + "step": 16150 + }, + { + "epoch": 2.8755341880341883, + "grad_norm": 0.8390287756919861, + "learning_rate": 3.663238262458122e-05, + "loss": 0.7084, + "step": 16151 + }, + { + "epoch": 2.8757122507122506, + "grad_norm": 0.921089768409729, + "learning_rate": 3.662155482311903e-05, + "loss": 0.8909, + "step": 16152 + }, + { + "epoch": 2.8758903133903133, + "grad_norm": 0.8790102601051331, + "learning_rate": 3.661072826342583e-05, + "loss": 0.8235, + "step": 16153 + }, + { + "epoch": 2.876068376068376, + "grad_norm": 0.8030495643615723, + "learning_rate": 3.659990294571368e-05, + "loss": 0.874, + "step": 16154 + }, + { + "epoch": 2.876246438746439, + "grad_norm": 0.9690510034561157, + "learning_rate": 3.6589078870194804e-05, + "loss": 1.0926, + "step": 16155 + }, + { + "epoch": 2.8764245014245016, + "grad_norm": 0.8150941133499146, + "learning_rate": 3.657825603708114e-05, + "loss": 0.861, + "step": 16156 + }, + { + "epoch": 2.876602564102564, + "grad_norm": 0.8865286111831665, + "learning_rate": 3.656743444658486e-05, + "loss": 0.8219, + "step": 16157 + }, + { + "epoch": 2.8767806267806266, + "grad_norm": 0.8591124415397644, + "learning_rate": 3.655661409891786e-05, + "loss": 0.889, + "step": 16158 + }, + { + "epoch": 2.8769586894586894, + "grad_norm": 0.8625560402870178, + "learning_rate": 3.6545794994292256e-05, + "loss": 0.9581, + "step": 16159 + }, + { + "epoch": 2.877136752136752, + "grad_norm": 0.8699239492416382, + "learning_rate": 3.653497713291999e-05, + "loss": 0.7275, + "step": 16160 + }, + { + "epoch": 2.877314814814815, + "grad_norm": 0.9005762338638306, + "learning_rate": 3.652416051501301e-05, + "loss": 0.8894, + "step": 16161 + }, + { + "epoch": 2.8774928774928776, + "grad_norm": 0.7541293501853943, + "learning_rate": 3.651334514078323e-05, + "loss": 0.7265, + "step": 16162 + }, + { + "epoch": 2.8776709401709404, + "grad_norm": 0.9799004793167114, + "learning_rate": 3.650253101044258e-05, + "loss": 0.8817, + "step": 16163 + }, + { + "epoch": 2.8778490028490027, + "grad_norm": 0.7796139717102051, + "learning_rate": 3.64917181242029e-05, + "loss": 0.7105, + "step": 16164 + }, + { + "epoch": 2.8780270655270654, + "grad_norm": 0.8818691968917847, + "learning_rate": 3.648090648227613e-05, + "loss": 0.8868, + "step": 16165 + }, + { + "epoch": 2.878205128205128, + "grad_norm": 0.7982428073883057, + "learning_rate": 3.647009608487399e-05, + "loss": 0.9269, + "step": 16166 + }, + { + "epoch": 2.878383190883191, + "grad_norm": 0.9602195620536804, + "learning_rate": 3.645928693220838e-05, + "loss": 0.8264, + "step": 16167 + }, + { + "epoch": 2.8785612535612537, + "grad_norm": 0.8941618800163269, + "learning_rate": 3.6448479024491054e-05, + "loss": 0.8009, + "step": 16168 + }, + { + "epoch": 2.878739316239316, + "grad_norm": 0.7777221202850342, + "learning_rate": 3.643767236193375e-05, + "loss": 0.7734, + "step": 16169 + }, + { + "epoch": 2.8789173789173788, + "grad_norm": 0.8050239086151123, + "learning_rate": 3.642686694474823e-05, + "loss": 0.9732, + "step": 16170 + }, + { + "epoch": 2.8790954415954415, + "grad_norm": 0.8437817096710205, + "learning_rate": 3.6416062773146156e-05, + "loss": 0.8161, + "step": 16171 + }, + { + "epoch": 2.8792735042735043, + "grad_norm": 0.7887414693832397, + "learning_rate": 3.64052598473393e-05, + "loss": 0.9165, + "step": 16172 + }, + { + "epoch": 2.879451566951567, + "grad_norm": 0.8671287894248962, + "learning_rate": 3.639445816753921e-05, + "loss": 0.9519, + "step": 16173 + }, + { + "epoch": 2.8796296296296298, + "grad_norm": 0.8444932103157043, + "learning_rate": 3.638365773395763e-05, + "loss": 0.784, + "step": 16174 + }, + { + "epoch": 2.8798076923076925, + "grad_norm": 0.8580447435379028, + "learning_rate": 3.637285854680612e-05, + "loss": 0.797, + "step": 16175 + }, + { + "epoch": 2.879985754985755, + "grad_norm": 0.7977848052978516, + "learning_rate": 3.636206060629627e-05, + "loss": 0.8552, + "step": 16176 + }, + { + "epoch": 2.8801638176638176, + "grad_norm": 0.8350155353546143, + "learning_rate": 3.6351263912639644e-05, + "loss": 0.9481, + "step": 16177 + }, + { + "epoch": 2.8803418803418803, + "grad_norm": 0.8759897351264954, + "learning_rate": 3.634046846604778e-05, + "loss": 0.9972, + "step": 16178 + }, + { + "epoch": 2.880519943019943, + "grad_norm": 0.8259425163269043, + "learning_rate": 3.6329674266732194e-05, + "loss": 0.8947, + "step": 16179 + }, + { + "epoch": 2.880698005698006, + "grad_norm": 0.7992371320724487, + "learning_rate": 3.631888131490438e-05, + "loss": 0.8346, + "step": 16180 + }, + { + "epoch": 2.880876068376068, + "grad_norm": 0.9702637195587158, + "learning_rate": 3.6308089610775775e-05, + "loss": 0.9543, + "step": 16181 + }, + { + "epoch": 2.881054131054131, + "grad_norm": 0.8494347929954529, + "learning_rate": 3.6297299154557905e-05, + "loss": 0.8672, + "step": 16182 + }, + { + "epoch": 2.8812321937321936, + "grad_norm": 0.8098355531692505, + "learning_rate": 3.628650994646207e-05, + "loss": 0.8356, + "step": 16183 + }, + { + "epoch": 2.8814102564102564, + "grad_norm": 0.7736916542053223, + "learning_rate": 3.627572198669974e-05, + "loss": 0.849, + "step": 16184 + }, + { + "epoch": 2.881588319088319, + "grad_norm": 0.806710422039032, + "learning_rate": 3.626493527548226e-05, + "loss": 0.9479, + "step": 16185 + }, + { + "epoch": 2.881766381766382, + "grad_norm": 1.0013649463653564, + "learning_rate": 3.625414981302095e-05, + "loss": 0.967, + "step": 16186 + }, + { + "epoch": 2.8819444444444446, + "grad_norm": 0.8116905093193054, + "learning_rate": 3.624336559952723e-05, + "loss": 0.8608, + "step": 16187 + }, + { + "epoch": 2.8821225071225074, + "grad_norm": 0.7848439812660217, + "learning_rate": 3.6232582635212233e-05, + "loss": 0.9224, + "step": 16188 + }, + { + "epoch": 2.8823005698005697, + "grad_norm": 0.9881102442741394, + "learning_rate": 3.622180092028735e-05, + "loss": 0.8089, + "step": 16189 + }, + { + "epoch": 2.8824786324786325, + "grad_norm": 0.790452778339386, + "learning_rate": 3.6211020454963796e-05, + "loss": 0.8312, + "step": 16190 + }, + { + "epoch": 2.882656695156695, + "grad_norm": 0.8228929042816162, + "learning_rate": 3.6200241239452784e-05, + "loss": 0.8277, + "step": 16191 + }, + { + "epoch": 2.882834757834758, + "grad_norm": 0.871133029460907, + "learning_rate": 3.618946327396552e-05, + "loss": 0.8861, + "step": 16192 + }, + { + "epoch": 2.8830128205128203, + "grad_norm": 0.8964230418205261, + "learning_rate": 3.617868655871316e-05, + "loss": 0.9987, + "step": 16193 + }, + { + "epoch": 2.883190883190883, + "grad_norm": 0.8292636275291443, + "learning_rate": 3.6167911093906856e-05, + "loss": 0.9823, + "step": 16194 + }, + { + "epoch": 2.8833689458689458, + "grad_norm": 0.9594070315361023, + "learning_rate": 3.615713687975774e-05, + "loss": 0.7963, + "step": 16195 + }, + { + "epoch": 2.8835470085470085, + "grad_norm": 1.014891266822815, + "learning_rate": 3.6146363916476864e-05, + "loss": 0.9203, + "step": 16196 + }, + { + "epoch": 2.8837250712250713, + "grad_norm": 0.8205485343933105, + "learning_rate": 3.6135592204275424e-05, + "loss": 0.894, + "step": 16197 + }, + { + "epoch": 2.883903133903134, + "grad_norm": 0.9057072997093201, + "learning_rate": 3.6124821743364315e-05, + "loss": 0.9131, + "step": 16198 + }, + { + "epoch": 2.8840811965811968, + "grad_norm": 0.8080529570579529, + "learning_rate": 3.6114052533954665e-05, + "loss": 0.9995, + "step": 16199 + }, + { + "epoch": 2.8842592592592595, + "grad_norm": 0.762264609336853, + "learning_rate": 3.6103284576257446e-05, + "loss": 0.915, + "step": 16200 + }, + { + "epoch": 2.884437321937322, + "grad_norm": 0.8386275172233582, + "learning_rate": 3.609251787048363e-05, + "loss": 0.9016, + "step": 16201 + }, + { + "epoch": 2.8846153846153846, + "grad_norm": 0.9676657319068909, + "learning_rate": 3.608175241684417e-05, + "loss": 1.1004, + "step": 16202 + }, + { + "epoch": 2.8847934472934473, + "grad_norm": 0.8155630826950073, + "learning_rate": 3.607098821554999e-05, + "loss": 0.8579, + "step": 16203 + }, + { + "epoch": 2.88497150997151, + "grad_norm": 0.8426685333251953, + "learning_rate": 3.606022526681201e-05, + "loss": 0.8722, + "step": 16204 + }, + { + "epoch": 2.8851495726495724, + "grad_norm": 0.7646408081054688, + "learning_rate": 3.604946357084105e-05, + "loss": 0.7806, + "step": 16205 + }, + { + "epoch": 2.885327635327635, + "grad_norm": 0.808560848236084, + "learning_rate": 3.603870312784803e-05, + "loss": 0.8419, + "step": 16206 + }, + { + "epoch": 2.885505698005698, + "grad_norm": 0.7197920083999634, + "learning_rate": 3.602794393804376e-05, + "loss": 0.7356, + "step": 16207 + }, + { + "epoch": 2.8856837606837606, + "grad_norm": 0.9405228495597839, + "learning_rate": 3.6017186001639036e-05, + "loss": 0.7279, + "step": 16208 + }, + { + "epoch": 2.8858618233618234, + "grad_norm": 0.8910958170890808, + "learning_rate": 3.600642931884465e-05, + "loss": 0.7653, + "step": 16209 + }, + { + "epoch": 2.886039886039886, + "grad_norm": 0.7354677319526672, + "learning_rate": 3.599567388987134e-05, + "loss": 0.6884, + "step": 16210 + }, + { + "epoch": 2.886217948717949, + "grad_norm": 0.75583815574646, + "learning_rate": 3.598491971492981e-05, + "loss": 0.7555, + "step": 16211 + }, + { + "epoch": 2.8863960113960117, + "grad_norm": 0.839042603969574, + "learning_rate": 3.597416679423086e-05, + "loss": 1.0522, + "step": 16212 + }, + { + "epoch": 2.886574074074074, + "grad_norm": 0.7965270280838013, + "learning_rate": 3.596341512798505e-05, + "loss": 0.9101, + "step": 16213 + }, + { + "epoch": 2.8867521367521367, + "grad_norm": 0.9170811772346497, + "learning_rate": 3.5952664716403154e-05, + "loss": 0.9078, + "step": 16214 + }, + { + "epoch": 2.8869301994301995, + "grad_norm": 0.8612267374992371, + "learning_rate": 3.5941915559695685e-05, + "loss": 0.8615, + "step": 16215 + }, + { + "epoch": 2.887108262108262, + "grad_norm": 0.9182412028312683, + "learning_rate": 3.593116765807335e-05, + "loss": 0.8918, + "step": 16216 + }, + { + "epoch": 2.8872863247863245, + "grad_norm": 0.9452505707740784, + "learning_rate": 3.592042101174669e-05, + "loss": 1.0386, + "step": 16217 + }, + { + "epoch": 2.8874643874643873, + "grad_norm": 0.8544769287109375, + "learning_rate": 3.5909675620926255e-05, + "loss": 1.0305, + "step": 16218 + }, + { + "epoch": 2.88764245014245, + "grad_norm": 0.8184592127799988, + "learning_rate": 3.5898931485822605e-05, + "loss": 0.6815, + "step": 16219 + }, + { + "epoch": 2.8878205128205128, + "grad_norm": 0.8263654708862305, + "learning_rate": 3.5888188606646224e-05, + "loss": 1.0097, + "step": 16220 + }, + { + "epoch": 2.8879985754985755, + "grad_norm": 0.8290582299232483, + "learning_rate": 3.587744698360758e-05, + "loss": 0.6539, + "step": 16221 + }, + { + "epoch": 2.8881766381766383, + "grad_norm": 0.7936849594116211, + "learning_rate": 3.5866706616917226e-05, + "loss": 0.9524, + "step": 16222 + }, + { + "epoch": 2.888354700854701, + "grad_norm": 0.9449033737182617, + "learning_rate": 3.585596750678546e-05, + "loss": 0.9243, + "step": 16223 + }, + { + "epoch": 2.888532763532764, + "grad_norm": 0.7599559426307678, + "learning_rate": 3.58452296534228e-05, + "loss": 0.8597, + "step": 16224 + }, + { + "epoch": 2.888710826210826, + "grad_norm": 0.8485760688781738, + "learning_rate": 3.583449305703959e-05, + "loss": 0.7236, + "step": 16225 + }, + { + "epoch": 2.888888888888889, + "grad_norm": 0.8510624170303345, + "learning_rate": 3.582375771784616e-05, + "loss": 0.9081, + "step": 16226 + }, + { + "epoch": 2.8890669515669516, + "grad_norm": 0.815827488899231, + "learning_rate": 3.581302363605296e-05, + "loss": 0.8259, + "step": 16227 + }, + { + "epoch": 2.8892450142450143, + "grad_norm": 0.7588803768157959, + "learning_rate": 3.580229081187016e-05, + "loss": 0.8722, + "step": 16228 + }, + { + "epoch": 2.8894230769230766, + "grad_norm": 1.0699365139007568, + "learning_rate": 3.579155924550817e-05, + "loss": 1.0261, + "step": 16229 + }, + { + "epoch": 2.8896011396011394, + "grad_norm": 0.9127700924873352, + "learning_rate": 3.5780828937177126e-05, + "loss": 0.8993, + "step": 16230 + }, + { + "epoch": 2.889779202279202, + "grad_norm": 0.8101344108581543, + "learning_rate": 3.577009988708737e-05, + "loss": 1.0002, + "step": 16231 + }, + { + "epoch": 2.889957264957265, + "grad_norm": 0.8259516358375549, + "learning_rate": 3.5759372095449085e-05, + "loss": 0.9223, + "step": 16232 + }, + { + "epoch": 2.8901353276353277, + "grad_norm": 0.8366032242774963, + "learning_rate": 3.574864556247246e-05, + "loss": 0.9635, + "step": 16233 + }, + { + "epoch": 2.8903133903133904, + "grad_norm": 0.8667864203453064, + "learning_rate": 3.573792028836764e-05, + "loss": 0.9923, + "step": 16234 + }, + { + "epoch": 2.890491452991453, + "grad_norm": 0.8982881903648376, + "learning_rate": 3.5727196273344784e-05, + "loss": 0.8735, + "step": 16235 + }, + { + "epoch": 2.890669515669516, + "grad_norm": 0.8208550214767456, + "learning_rate": 3.571647351761398e-05, + "loss": 0.809, + "step": 16236 + }, + { + "epoch": 2.890847578347578, + "grad_norm": 1.0483866930007935, + "learning_rate": 3.5705752021385395e-05, + "loss": 0.9106, + "step": 16237 + }, + { + "epoch": 2.891025641025641, + "grad_norm": 0.8783283829689026, + "learning_rate": 3.5695031784868984e-05, + "loss": 0.8061, + "step": 16238 + }, + { + "epoch": 2.8912037037037037, + "grad_norm": 0.7530997395515442, + "learning_rate": 3.56843128082749e-05, + "loss": 0.58, + "step": 16239 + }, + { + "epoch": 2.8913817663817665, + "grad_norm": 1.0263420343399048, + "learning_rate": 3.567359509181304e-05, + "loss": 0.9946, + "step": 16240 + }, + { + "epoch": 2.8915598290598292, + "grad_norm": 0.7237655520439148, + "learning_rate": 3.5662878635693484e-05, + "loss": 0.6941, + "step": 16241 + }, + { + "epoch": 2.8917378917378915, + "grad_norm": 0.8415992856025696, + "learning_rate": 3.565216344012618e-05, + "loss": 0.7908, + "step": 16242 + }, + { + "epoch": 2.8919159544159543, + "grad_norm": 0.8285344243049622, + "learning_rate": 3.564144950532107e-05, + "loss": 0.8044, + "step": 16243 + }, + { + "epoch": 2.892094017094017, + "grad_norm": 0.9715448617935181, + "learning_rate": 3.5630736831488046e-05, + "loss": 0.8236, + "step": 16244 + }, + { + "epoch": 2.89227207977208, + "grad_norm": 0.8461307287216187, + "learning_rate": 3.5620025418836985e-05, + "loss": 0.9222, + "step": 16245 + }, + { + "epoch": 2.8924501424501425, + "grad_norm": 0.7191013097763062, + "learning_rate": 3.5609315267577836e-05, + "loss": 0.6089, + "step": 16246 + }, + { + "epoch": 2.8926282051282053, + "grad_norm": 0.8537501692771912, + "learning_rate": 3.559860637792038e-05, + "loss": 0.8771, + "step": 16247 + }, + { + "epoch": 2.892806267806268, + "grad_norm": 0.8684942126274109, + "learning_rate": 3.558789875007447e-05, + "loss": 0.9009, + "step": 16248 + }, + { + "epoch": 2.8929843304843303, + "grad_norm": 1.0619043111801147, + "learning_rate": 3.5577192384249856e-05, + "loss": 0.7283, + "step": 16249 + }, + { + "epoch": 2.893162393162393, + "grad_norm": 0.8889201879501343, + "learning_rate": 3.556648728065635e-05, + "loss": 0.7538, + "step": 16250 + }, + { + "epoch": 2.893340455840456, + "grad_norm": 0.8162542581558228, + "learning_rate": 3.555578343950367e-05, + "loss": 0.8338, + "step": 16251 + }, + { + "epoch": 2.8935185185185186, + "grad_norm": 0.8750036954879761, + "learning_rate": 3.5545080861001535e-05, + "loss": 0.6956, + "step": 16252 + }, + { + "epoch": 2.8936965811965814, + "grad_norm": 0.911232054233551, + "learning_rate": 3.553437954535962e-05, + "loss": 0.7559, + "step": 16253 + }, + { + "epoch": 2.8938746438746437, + "grad_norm": 0.889566957950592, + "learning_rate": 3.5523679492787685e-05, + "loss": 0.969, + "step": 16254 + }, + { + "epoch": 2.8940527065527064, + "grad_norm": 0.933595597743988, + "learning_rate": 3.551298070349525e-05, + "loss": 0.9767, + "step": 16255 + }, + { + "epoch": 2.894230769230769, + "grad_norm": 0.8633596897125244, + "learning_rate": 3.550228317769203e-05, + "loss": 0.7823, + "step": 16256 + }, + { + "epoch": 2.894408831908832, + "grad_norm": 0.8595561981201172, + "learning_rate": 3.5491586915587585e-05, + "loss": 0.9583, + "step": 16257 + }, + { + "epoch": 2.8945868945868947, + "grad_norm": 0.7595796585083008, + "learning_rate": 3.548089191739149e-05, + "loss": 0.7167, + "step": 16258 + }, + { + "epoch": 2.8947649572649574, + "grad_norm": 0.8662711381912231, + "learning_rate": 3.54701981833133e-05, + "loss": 0.9439, + "step": 16259 + }, + { + "epoch": 2.89494301994302, + "grad_norm": 0.8028330206871033, + "learning_rate": 3.5459505713562525e-05, + "loss": 0.7442, + "step": 16260 + }, + { + "epoch": 2.8951210826210825, + "grad_norm": 0.9413794279098511, + "learning_rate": 3.5448814508348616e-05, + "loss": 0.97, + "step": 16261 + }, + { + "epoch": 2.8952991452991452, + "grad_norm": 0.9300761222839355, + "learning_rate": 3.543812456788118e-05, + "loss": 0.9294, + "step": 16262 + }, + { + "epoch": 2.895477207977208, + "grad_norm": 0.7364256381988525, + "learning_rate": 3.54274358923695e-05, + "loss": 0.5352, + "step": 16263 + }, + { + "epoch": 2.8956552706552707, + "grad_norm": 0.8452964425086975, + "learning_rate": 3.541674848202314e-05, + "loss": 0.7362, + "step": 16264 + }, + { + "epoch": 2.8958333333333335, + "grad_norm": 0.8745927214622498, + "learning_rate": 3.540606233705137e-05, + "loss": 1.0222, + "step": 16265 + }, + { + "epoch": 2.896011396011396, + "grad_norm": 0.9202282428741455, + "learning_rate": 3.539537745766367e-05, + "loss": 1.0853, + "step": 16266 + }, + { + "epoch": 2.8961894586894585, + "grad_norm": 0.8450053930282593, + "learning_rate": 3.538469384406933e-05, + "loss": 0.9295, + "step": 16267 + }, + { + "epoch": 2.8963675213675213, + "grad_norm": 0.8761671781539917, + "learning_rate": 3.5374011496477656e-05, + "loss": 0.9273, + "step": 16268 + }, + { + "epoch": 2.896545584045584, + "grad_norm": 0.8562198281288147, + "learning_rate": 3.536333041509805e-05, + "loss": 0.7654, + "step": 16269 + }, + { + "epoch": 2.896723646723647, + "grad_norm": 0.9013510942459106, + "learning_rate": 3.535265060013965e-05, + "loss": 0.7432, + "step": 16270 + }, + { + "epoch": 2.8969017094017095, + "grad_norm": 1.1163274049758911, + "learning_rate": 3.534197205181179e-05, + "loss": 1.1223, + "step": 16271 + }, + { + "epoch": 2.8970797720797723, + "grad_norm": 0.869686484336853, + "learning_rate": 3.5331294770323674e-05, + "loss": 0.9255, + "step": 16272 + }, + { + "epoch": 2.8972578347578346, + "grad_norm": 0.8521125316619873, + "learning_rate": 3.53206187558845e-05, + "loss": 1.0221, + "step": 16273 + }, + { + "epoch": 2.8974358974358974, + "grad_norm": 0.9807026982307434, + "learning_rate": 3.530994400870345e-05, + "loss": 1.0133, + "step": 16274 + }, + { + "epoch": 2.89761396011396, + "grad_norm": 0.9236428141593933, + "learning_rate": 3.529927052898967e-05, + "loss": 0.8711, + "step": 16275 + }, + { + "epoch": 2.897792022792023, + "grad_norm": 0.8108885884284973, + "learning_rate": 3.528859831695227e-05, + "loss": 0.917, + "step": 16276 + }, + { + "epoch": 2.8979700854700856, + "grad_norm": 0.7522720098495483, + "learning_rate": 3.527792737280036e-05, + "loss": 0.6776, + "step": 16277 + }, + { + "epoch": 2.898148148148148, + "grad_norm": 0.9261712431907654, + "learning_rate": 3.526725769674297e-05, + "loss": 1.0728, + "step": 16278 + }, + { + "epoch": 2.8983262108262107, + "grad_norm": 0.849559485912323, + "learning_rate": 3.5256589288989285e-05, + "loss": 0.9886, + "step": 16279 + }, + { + "epoch": 2.8985042735042734, + "grad_norm": 0.9024273157119751, + "learning_rate": 3.5245922149748155e-05, + "loss": 0.9296, + "step": 16280 + }, + { + "epoch": 2.898682336182336, + "grad_norm": 0.8285173773765564, + "learning_rate": 3.52352562792287e-05, + "loss": 0.8776, + "step": 16281 + }, + { + "epoch": 2.898860398860399, + "grad_norm": 0.895517885684967, + "learning_rate": 3.522459167763987e-05, + "loss": 0.8975, + "step": 16282 + }, + { + "epoch": 2.8990384615384617, + "grad_norm": 0.8449265956878662, + "learning_rate": 3.521392834519061e-05, + "loss": 0.6736, + "step": 16283 + }, + { + "epoch": 2.8992165242165244, + "grad_norm": 0.8652997612953186, + "learning_rate": 3.520326628208983e-05, + "loss": 0.7834, + "step": 16284 + }, + { + "epoch": 2.8993945868945867, + "grad_norm": 0.9012393951416016, + "learning_rate": 3.519260548854642e-05, + "loss": 0.8082, + "step": 16285 + }, + { + "epoch": 2.8995726495726495, + "grad_norm": 0.9048463106155396, + "learning_rate": 3.5181945964769333e-05, + "loss": 0.8078, + "step": 16286 + }, + { + "epoch": 2.8997507122507122, + "grad_norm": 0.8788473010063171, + "learning_rate": 3.5171287710967314e-05, + "loss": 0.8022, + "step": 16287 + }, + { + "epoch": 2.899928774928775, + "grad_norm": 0.8322813510894775, + "learning_rate": 3.516063072734928e-05, + "loss": 0.8855, + "step": 16288 + }, + { + "epoch": 2.9001068376068377, + "grad_norm": 0.8762373328208923, + "learning_rate": 3.514997501412398e-05, + "loss": 0.7858, + "step": 16289 + }, + { + "epoch": 2.9002849002849, + "grad_norm": 0.7718746066093445, + "learning_rate": 3.513932057150021e-05, + "loss": 0.6881, + "step": 16290 + }, + { + "epoch": 2.900462962962963, + "grad_norm": 1.0138204097747803, + "learning_rate": 3.5128667399686724e-05, + "loss": 0.9378, + "step": 16291 + }, + { + "epoch": 2.9006410256410255, + "grad_norm": 0.6968120336532593, + "learning_rate": 3.5118015498892234e-05, + "loss": 0.7135, + "step": 16292 + }, + { + "epoch": 2.9008190883190883, + "grad_norm": 0.7925532460212708, + "learning_rate": 3.510736486932542e-05, + "loss": 0.7414, + "step": 16293 + }, + { + "epoch": 2.900997150997151, + "grad_norm": 1.0432425737380981, + "learning_rate": 3.5096715511195056e-05, + "loss": 0.8957, + "step": 16294 + }, + { + "epoch": 2.901175213675214, + "grad_norm": 0.8664390444755554, + "learning_rate": 3.508606742470966e-05, + "loss": 0.7832, + "step": 16295 + }, + { + "epoch": 2.9013532763532766, + "grad_norm": 0.8470353484153748, + "learning_rate": 3.507542061007795e-05, + "loss": 0.8133, + "step": 16296 + }, + { + "epoch": 2.9015313390313393, + "grad_norm": 0.8339848518371582, + "learning_rate": 3.5064775067508514e-05, + "loss": 0.9275, + "step": 16297 + }, + { + "epoch": 2.9017094017094016, + "grad_norm": 0.8686776757240295, + "learning_rate": 3.5054130797209916e-05, + "loss": 0.9238, + "step": 16298 + }, + { + "epoch": 2.9018874643874644, + "grad_norm": 0.8178901076316833, + "learning_rate": 3.504348779939071e-05, + "loss": 1.0776, + "step": 16299 + }, + { + "epoch": 2.902065527065527, + "grad_norm": 0.8446379899978638, + "learning_rate": 3.5032846074259426e-05, + "loss": 0.715, + "step": 16300 + }, + { + "epoch": 2.90224358974359, + "grad_norm": 0.8320762515068054, + "learning_rate": 3.502220562202457e-05, + "loss": 0.8543, + "step": 16301 + }, + { + "epoch": 2.902421652421652, + "grad_norm": 0.9112939238548279, + "learning_rate": 3.501156644289462e-05, + "loss": 0.8239, + "step": 16302 + }, + { + "epoch": 2.902599715099715, + "grad_norm": 0.8749213218688965, + "learning_rate": 3.500092853707797e-05, + "loss": 0.8057, + "step": 16303 + }, + { + "epoch": 2.9027777777777777, + "grad_norm": 0.8245106339454651, + "learning_rate": 3.4990291904783143e-05, + "loss": 0.8955, + "step": 16304 + }, + { + "epoch": 2.9029558404558404, + "grad_norm": 0.7982145547866821, + "learning_rate": 3.4979656546218506e-05, + "loss": 0.6641, + "step": 16305 + }, + { + "epoch": 2.903133903133903, + "grad_norm": 0.8777986168861389, + "learning_rate": 3.496902246159244e-05, + "loss": 0.7365, + "step": 16306 + }, + { + "epoch": 2.903311965811966, + "grad_norm": 0.8463431596755981, + "learning_rate": 3.4958389651113275e-05, + "loss": 1.0434, + "step": 16307 + }, + { + "epoch": 2.9034900284900287, + "grad_norm": 0.849039614200592, + "learning_rate": 3.494775811498931e-05, + "loss": 0.8813, + "step": 16308 + }, + { + "epoch": 2.9036680911680914, + "grad_norm": 0.7352656126022339, + "learning_rate": 3.4937127853428976e-05, + "loss": 0.7936, + "step": 16309 + }, + { + "epoch": 2.9038461538461537, + "grad_norm": 0.764543354511261, + "learning_rate": 3.49264988666404e-05, + "loss": 0.7253, + "step": 16310 + }, + { + "epoch": 2.9040242165242165, + "grad_norm": 0.7932603359222412, + "learning_rate": 3.491587115483196e-05, + "loss": 0.9217, + "step": 16311 + }, + { + "epoch": 2.9042022792022792, + "grad_norm": 0.9001819491386414, + "learning_rate": 3.490524471821175e-05, + "loss": 0.8628, + "step": 16312 + }, + { + "epoch": 2.904380341880342, + "grad_norm": 0.7983253002166748, + "learning_rate": 3.4894619556988085e-05, + "loss": 0.9858, + "step": 16313 + }, + { + "epoch": 2.9045584045584043, + "grad_norm": 0.747611403465271, + "learning_rate": 3.488399567136911e-05, + "loss": 0.8656, + "step": 16314 + }, + { + "epoch": 2.904736467236467, + "grad_norm": 0.8770463466644287, + "learning_rate": 3.487337306156296e-05, + "loss": 1.0154, + "step": 16315 + }, + { + "epoch": 2.90491452991453, + "grad_norm": 0.8757193088531494, + "learning_rate": 3.4862751727777797e-05, + "loss": 0.7931, + "step": 16316 + }, + { + "epoch": 2.9050925925925926, + "grad_norm": 0.8232926726341248, + "learning_rate": 3.485213167022169e-05, + "loss": 0.8477, + "step": 16317 + }, + { + "epoch": 2.9052706552706553, + "grad_norm": 0.8445250391960144, + "learning_rate": 3.48415128891027e-05, + "loss": 1.0618, + "step": 16318 + }, + { + "epoch": 2.905448717948718, + "grad_norm": 0.8172810673713684, + "learning_rate": 3.483089538462897e-05, + "loss": 0.7882, + "step": 16319 + }, + { + "epoch": 2.905626780626781, + "grad_norm": 0.910757303237915, + "learning_rate": 3.4820279157008404e-05, + "loss": 0.9383, + "step": 16320 + }, + { + "epoch": 2.9058048433048436, + "grad_norm": 0.8837474584579468, + "learning_rate": 3.480966420644911e-05, + "loss": 1.1006, + "step": 16321 + }, + { + "epoch": 2.905982905982906, + "grad_norm": 0.7739782333374023, + "learning_rate": 3.4799050533159014e-05, + "loss": 0.6885, + "step": 16322 + }, + { + "epoch": 2.9061609686609686, + "grad_norm": 0.911738932132721, + "learning_rate": 3.478843813734609e-05, + "loss": 0.993, + "step": 16323 + }, + { + "epoch": 2.9063390313390314, + "grad_norm": 0.8834345936775208, + "learning_rate": 3.477782701921825e-05, + "loss": 0.8724, + "step": 16324 + }, + { + "epoch": 2.906517094017094, + "grad_norm": 0.8103434443473816, + "learning_rate": 3.476721717898337e-05, + "loss": 1.0517, + "step": 16325 + }, + { + "epoch": 2.9066951566951564, + "grad_norm": 0.8648924827575684, + "learning_rate": 3.475660861684943e-05, + "loss": 0.7223, + "step": 16326 + }, + { + "epoch": 2.906873219373219, + "grad_norm": 0.8762979507446289, + "learning_rate": 3.4746001333024134e-05, + "loss": 0.8421, + "step": 16327 + }, + { + "epoch": 2.907051282051282, + "grad_norm": 0.9596083164215088, + "learning_rate": 3.4735395327715434e-05, + "loss": 1.0206, + "step": 16328 + }, + { + "epoch": 2.9072293447293447, + "grad_norm": 0.8210311532020569, + "learning_rate": 3.472479060113107e-05, + "loss": 0.836, + "step": 16329 + }, + { + "epoch": 2.9074074074074074, + "grad_norm": 0.8436611890792847, + "learning_rate": 3.471418715347886e-05, + "loss": 0.8459, + "step": 16330 + }, + { + "epoch": 2.90758547008547, + "grad_norm": 0.9176212549209595, + "learning_rate": 3.470358498496652e-05, + "loss": 0.8227, + "step": 16331 + }, + { + "epoch": 2.907763532763533, + "grad_norm": 0.7709631323814392, + "learning_rate": 3.4692984095801796e-05, + "loss": 0.8096, + "step": 16332 + }, + { + "epoch": 2.9079415954415957, + "grad_norm": 0.7727495431900024, + "learning_rate": 3.4682384486192346e-05, + "loss": 0.6843, + "step": 16333 + }, + { + "epoch": 2.908119658119658, + "grad_norm": 0.9743461608886719, + "learning_rate": 3.4671786156345955e-05, + "loss": 0.7542, + "step": 16334 + }, + { + "epoch": 2.9082977207977208, + "grad_norm": 0.9035171270370483, + "learning_rate": 3.466118910647014e-05, + "loss": 0.8827, + "step": 16335 + }, + { + "epoch": 2.9084757834757835, + "grad_norm": 1.1434134244918823, + "learning_rate": 3.465059333677266e-05, + "loss": 0.8026, + "step": 16336 + }, + { + "epoch": 2.9086538461538463, + "grad_norm": 0.8229905962944031, + "learning_rate": 3.4639998847461e-05, + "loss": 0.7415, + "step": 16337 + }, + { + "epoch": 2.9088319088319086, + "grad_norm": 0.9193732738494873, + "learning_rate": 3.462940563874281e-05, + "loss": 0.9773, + "step": 16338 + }, + { + "epoch": 2.9090099715099713, + "grad_norm": 0.8461189270019531, + "learning_rate": 3.4618813710825614e-05, + "loss": 0.9169, + "step": 16339 + }, + { + "epoch": 2.909188034188034, + "grad_norm": 0.9471017718315125, + "learning_rate": 3.460822306391696e-05, + "loss": 0.9941, + "step": 16340 + }, + { + "epoch": 2.909366096866097, + "grad_norm": 0.8515542149543762, + "learning_rate": 3.459763369822432e-05, + "loss": 0.8209, + "step": 16341 + }, + { + "epoch": 2.9095441595441596, + "grad_norm": 0.8520704507827759, + "learning_rate": 3.458704561395519e-05, + "loss": 0.8443, + "step": 16342 + }, + { + "epoch": 2.9097222222222223, + "grad_norm": 0.9236885905265808, + "learning_rate": 3.457645881131699e-05, + "loss": 0.8407, + "step": 16343 + }, + { + "epoch": 2.909900284900285, + "grad_norm": 0.9255889654159546, + "learning_rate": 3.4565873290517203e-05, + "loss": 0.9754, + "step": 16344 + }, + { + "epoch": 2.910078347578348, + "grad_norm": 0.7904002666473389, + "learning_rate": 3.455528905176321e-05, + "loss": 0.7123, + "step": 16345 + }, + { + "epoch": 2.91025641025641, + "grad_norm": 0.821877658367157, + "learning_rate": 3.454470609526237e-05, + "loss": 0.7024, + "step": 16346 + }, + { + "epoch": 2.910434472934473, + "grad_norm": 0.84690260887146, + "learning_rate": 3.453412442122205e-05, + "loss": 0.7551, + "step": 16347 + }, + { + "epoch": 2.9106125356125356, + "grad_norm": 1.0308923721313477, + "learning_rate": 3.452354402984955e-05, + "loss": 1.0316, + "step": 16348 + }, + { + "epoch": 2.9107905982905984, + "grad_norm": 0.7016192674636841, + "learning_rate": 3.451296492135221e-05, + "loss": 0.6816, + "step": 16349 + }, + { + "epoch": 2.9109686609686607, + "grad_norm": 0.942915141582489, + "learning_rate": 3.4502387095937237e-05, + "loss": 0.9029, + "step": 16350 + }, + { + "epoch": 2.9111467236467234, + "grad_norm": 0.7511810660362244, + "learning_rate": 3.449181055381201e-05, + "loss": 0.641, + "step": 16351 + }, + { + "epoch": 2.911324786324786, + "grad_norm": 0.8904419541358948, + "learning_rate": 3.44812352951836e-05, + "loss": 0.8176, + "step": 16352 + }, + { + "epoch": 2.911502849002849, + "grad_norm": 0.7585623264312744, + "learning_rate": 3.447066132025931e-05, + "loss": 0.7815, + "step": 16353 + }, + { + "epoch": 2.9116809116809117, + "grad_norm": 0.8587543964385986, + "learning_rate": 3.446008862924629e-05, + "loss": 0.951, + "step": 16354 + }, + { + "epoch": 2.9118589743589745, + "grad_norm": 0.7638232707977295, + "learning_rate": 3.444951722235169e-05, + "loss": 0.6784, + "step": 16355 + }, + { + "epoch": 2.912037037037037, + "grad_norm": 0.7942266464233398, + "learning_rate": 3.4438947099782624e-05, + "loss": 0.7211, + "step": 16356 + }, + { + "epoch": 2.9122150997151, + "grad_norm": 0.7207983732223511, + "learning_rate": 3.4428378261746195e-05, + "loss": 0.7515, + "step": 16357 + }, + { + "epoch": 2.9123931623931623, + "grad_norm": 0.8683337569236755, + "learning_rate": 3.4417810708449495e-05, + "loss": 0.8724, + "step": 16358 + }, + { + "epoch": 2.912571225071225, + "grad_norm": 0.8052859902381897, + "learning_rate": 3.440724444009955e-05, + "loss": 0.8331, + "step": 16359 + }, + { + "epoch": 2.9127492877492878, + "grad_norm": 0.7425459027290344, + "learning_rate": 3.439667945690336e-05, + "loss": 0.5711, + "step": 16360 + }, + { + "epoch": 2.9129273504273505, + "grad_norm": 0.8537404537200928, + "learning_rate": 3.438611575906803e-05, + "loss": 0.8283, + "step": 16361 + }, + { + "epoch": 2.9131054131054133, + "grad_norm": 0.794684648513794, + "learning_rate": 3.437555334680038e-05, + "loss": 0.679, + "step": 16362 + }, + { + "epoch": 2.9132834757834756, + "grad_norm": 0.8330501914024353, + "learning_rate": 3.436499222030748e-05, + "loss": 0.6892, + "step": 16363 + }, + { + "epoch": 2.9134615384615383, + "grad_norm": 0.9036495089530945, + "learning_rate": 3.435443237979621e-05, + "loss": 0.8349, + "step": 16364 + }, + { + "epoch": 2.913639601139601, + "grad_norm": 0.776745080947876, + "learning_rate": 3.434387382547344e-05, + "loss": 0.6691, + "step": 16365 + }, + { + "epoch": 2.913817663817664, + "grad_norm": 0.8921083807945251, + "learning_rate": 3.4333316557546145e-05, + "loss": 0.9497, + "step": 16366 + }, + { + "epoch": 2.9139957264957266, + "grad_norm": 0.8435728549957275, + "learning_rate": 3.4322760576221023e-05, + "loss": 0.787, + "step": 16367 + }, + { + "epoch": 2.9141737891737893, + "grad_norm": 0.7281554341316223, + "learning_rate": 3.4312205881705015e-05, + "loss": 0.7604, + "step": 16368 + }, + { + "epoch": 2.914351851851852, + "grad_norm": 0.8474677801132202, + "learning_rate": 3.430165247420488e-05, + "loss": 0.9052, + "step": 16369 + }, + { + "epoch": 2.9145299145299144, + "grad_norm": 0.9927265048027039, + "learning_rate": 3.42911003539274e-05, + "loss": 0.7486, + "step": 16370 + }, + { + "epoch": 2.914707977207977, + "grad_norm": 0.7754966020584106, + "learning_rate": 3.4280549521079286e-05, + "loss": 0.8368, + "step": 16371 + }, + { + "epoch": 2.91488603988604, + "grad_norm": 0.7774887084960938, + "learning_rate": 3.4269999975867295e-05, + "loss": 0.7014, + "step": 16372 + }, + { + "epoch": 2.9150641025641026, + "grad_norm": 0.9286267161369324, + "learning_rate": 3.4259451718498115e-05, + "loss": 1.0414, + "step": 16373 + }, + { + "epoch": 2.9152421652421654, + "grad_norm": 0.8269815444946289, + "learning_rate": 3.424890474917841e-05, + "loss": 0.9731, + "step": 16374 + }, + { + "epoch": 2.9154202279202277, + "grad_norm": 0.8319926857948303, + "learning_rate": 3.42383590681148e-05, + "loss": 0.9047, + "step": 16375 + }, + { + "epoch": 2.9155982905982905, + "grad_norm": 0.8976882696151733, + "learning_rate": 3.4227814675514e-05, + "loss": 0.8608, + "step": 16376 + }, + { + "epoch": 2.915776353276353, + "grad_norm": 0.80831378698349, + "learning_rate": 3.421727157158248e-05, + "loss": 0.9213, + "step": 16377 + }, + { + "epoch": 2.915954415954416, + "grad_norm": 0.881201982498169, + "learning_rate": 3.42067297565269e-05, + "loss": 0.9015, + "step": 16378 + }, + { + "epoch": 2.9161324786324787, + "grad_norm": 0.7797132134437561, + "learning_rate": 3.4196189230553775e-05, + "loss": 0.9002, + "step": 16379 + }, + { + "epoch": 2.9163105413105415, + "grad_norm": 1.0163915157318115, + "learning_rate": 3.4185649993869626e-05, + "loss": 1.0185, + "step": 16380 + }, + { + "epoch": 2.916488603988604, + "grad_norm": 0.8868339657783508, + "learning_rate": 3.417511204668096e-05, + "loss": 0.7316, + "step": 16381 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 0.8215135931968689, + "learning_rate": 3.416457538919422e-05, + "loss": 0.8693, + "step": 16382 + }, + { + "epoch": 2.9168447293447293, + "grad_norm": 0.8070623278617859, + "learning_rate": 3.4154040021615876e-05, + "loss": 0.8532, + "step": 16383 + }, + { + "epoch": 2.917022792022792, + "grad_norm": 0.8857349157333374, + "learning_rate": 3.4143505944152284e-05, + "loss": 0.961, + "step": 16384 + }, + { + "epoch": 2.9172008547008548, + "grad_norm": 0.8146688342094421, + "learning_rate": 3.4132973157009936e-05, + "loss": 0.8597, + "step": 16385 + }, + { + "epoch": 2.9173789173789175, + "grad_norm": 0.8757439851760864, + "learning_rate": 3.4122441660395156e-05, + "loss": 0.7683, + "step": 16386 + }, + { + "epoch": 2.91755698005698, + "grad_norm": 0.8319665789604187, + "learning_rate": 3.4111911454514266e-05, + "loss": 0.8034, + "step": 16387 + }, + { + "epoch": 2.9177350427350426, + "grad_norm": 0.8103782534599304, + "learning_rate": 3.410138253957361e-05, + "loss": 0.8469, + "step": 16388 + }, + { + "epoch": 2.9179131054131053, + "grad_norm": 0.8783053755760193, + "learning_rate": 3.4090854915779466e-05, + "loss": 0.7997, + "step": 16389 + }, + { + "epoch": 2.918091168091168, + "grad_norm": 0.9793184995651245, + "learning_rate": 3.408032858333808e-05, + "loss": 0.8934, + "step": 16390 + }, + { + "epoch": 2.918269230769231, + "grad_norm": 0.8603301048278809, + "learning_rate": 3.406980354245578e-05, + "loss": 0.847, + "step": 16391 + }, + { + "epoch": 2.9184472934472936, + "grad_norm": 0.9763472080230713, + "learning_rate": 3.405927979333866e-05, + "loss": 0.9491, + "step": 16392 + }, + { + "epoch": 2.9186253561253563, + "grad_norm": 0.9340085983276367, + "learning_rate": 3.4048757336193036e-05, + "loss": 0.8913, + "step": 16393 + }, + { + "epoch": 2.9188034188034186, + "grad_norm": 1.051066517829895, + "learning_rate": 3.4038236171224946e-05, + "loss": 0.9903, + "step": 16394 + }, + { + "epoch": 2.9189814814814814, + "grad_norm": 0.8462432026863098, + "learning_rate": 3.402771629864062e-05, + "loss": 0.868, + "step": 16395 + }, + { + "epoch": 2.919159544159544, + "grad_norm": 0.9149147868156433, + "learning_rate": 3.401719771864615e-05, + "loss": 0.9215, + "step": 16396 + }, + { + "epoch": 2.919337606837607, + "grad_norm": 0.7974846959114075, + "learning_rate": 3.4006680431447636e-05, + "loss": 0.9577, + "step": 16397 + }, + { + "epoch": 2.9195156695156697, + "grad_norm": 0.9298149943351746, + "learning_rate": 3.399616443725112e-05, + "loss": 0.8977, + "step": 16398 + }, + { + "epoch": 2.919693732193732, + "grad_norm": 0.8799013495445251, + "learning_rate": 3.398564973626265e-05, + "loss": 1.0161, + "step": 16399 + }, + { + "epoch": 2.9198717948717947, + "grad_norm": 0.8730150461196899, + "learning_rate": 3.3975136328688206e-05, + "loss": 0.7953, + "step": 16400 + }, + { + "epoch": 2.9200498575498575, + "grad_norm": 1.0036572217941284, + "learning_rate": 3.396462421473387e-05, + "loss": 0.5941, + "step": 16401 + }, + { + "epoch": 2.92022792022792, + "grad_norm": 0.8903291821479797, + "learning_rate": 3.39541133946055e-05, + "loss": 0.7715, + "step": 16402 + }, + { + "epoch": 2.920405982905983, + "grad_norm": 0.8254680633544922, + "learning_rate": 3.39436038685091e-05, + "loss": 0.7268, + "step": 16403 + }, + { + "epoch": 2.9205840455840457, + "grad_norm": 0.8490110039710999, + "learning_rate": 3.393309563665057e-05, + "loss": 0.8684, + "step": 16404 + }, + { + "epoch": 2.9207621082621085, + "grad_norm": 0.8137654066085815, + "learning_rate": 3.392258869923575e-05, + "loss": 0.7068, + "step": 16405 + }, + { + "epoch": 2.9209401709401708, + "grad_norm": 0.8084680438041687, + "learning_rate": 3.39120830564706e-05, + "loss": 0.8214, + "step": 16406 + }, + { + "epoch": 2.9211182336182335, + "grad_norm": 0.9359310269355774, + "learning_rate": 3.3901578708560835e-05, + "loss": 1.0247, + "step": 16407 + }, + { + "epoch": 2.9212962962962963, + "grad_norm": 0.826842725276947, + "learning_rate": 3.38910756557124e-05, + "loss": 0.8018, + "step": 16408 + }, + { + "epoch": 2.921474358974359, + "grad_norm": 0.9049538373947144, + "learning_rate": 3.388057389813093e-05, + "loss": 0.8825, + "step": 16409 + }, + { + "epoch": 2.921652421652422, + "grad_norm": 0.7966105937957764, + "learning_rate": 3.387007343602231e-05, + "loss": 0.8716, + "step": 16410 + }, + { + "epoch": 2.921830484330484, + "grad_norm": 0.8647517561912537, + "learning_rate": 3.385957426959222e-05, + "loss": 0.9947, + "step": 16411 + }, + { + "epoch": 2.922008547008547, + "grad_norm": 0.8983362317085266, + "learning_rate": 3.384907639904638e-05, + "loss": 0.8701, + "step": 16412 + }, + { + "epoch": 2.9221866096866096, + "grad_norm": 0.7970262765884399, + "learning_rate": 3.3838579824590465e-05, + "loss": 0.8921, + "step": 16413 + }, + { + "epoch": 2.9223646723646723, + "grad_norm": 0.8115224242210388, + "learning_rate": 3.382808454643015e-05, + "loss": 0.5948, + "step": 16414 + }, + { + "epoch": 2.922542735042735, + "grad_norm": 0.832318902015686, + "learning_rate": 3.381759056477102e-05, + "loss": 0.8399, + "step": 16415 + }, + { + "epoch": 2.922720797720798, + "grad_norm": 0.9396497011184692, + "learning_rate": 3.380709787981878e-05, + "loss": 0.7322, + "step": 16416 + }, + { + "epoch": 2.9228988603988606, + "grad_norm": 0.7956545352935791, + "learning_rate": 3.3796606491778904e-05, + "loss": 0.8566, + "step": 16417 + }, + { + "epoch": 2.9230769230769234, + "grad_norm": 0.8257092833518982, + "learning_rate": 3.378611640085705e-05, + "loss": 0.8682, + "step": 16418 + }, + { + "epoch": 2.9232549857549857, + "grad_norm": 0.7565430402755737, + "learning_rate": 3.377562760725863e-05, + "loss": 0.8513, + "step": 16419 + }, + { + "epoch": 2.9234330484330484, + "grad_norm": 0.769944965839386, + "learning_rate": 3.3765140111189265e-05, + "loss": 0.8869, + "step": 16420 + }, + { + "epoch": 2.923611111111111, + "grad_norm": 0.8117483854293823, + "learning_rate": 3.375465391285438e-05, + "loss": 0.8718, + "step": 16421 + }, + { + "epoch": 2.923789173789174, + "grad_norm": 0.8011773824691772, + "learning_rate": 3.374416901245944e-05, + "loss": 0.9326, + "step": 16422 + }, + { + "epoch": 2.923967236467236, + "grad_norm": 0.8096779584884644, + "learning_rate": 3.373368541020987e-05, + "loss": 0.9317, + "step": 16423 + }, + { + "epoch": 2.924145299145299, + "grad_norm": 0.8383152484893799, + "learning_rate": 3.3723203106311055e-05, + "loss": 0.9399, + "step": 16424 + }, + { + "epoch": 2.9243233618233617, + "grad_norm": 0.9268670678138733, + "learning_rate": 3.371272210096842e-05, + "loss": 0.6867, + "step": 16425 + }, + { + "epoch": 2.9245014245014245, + "grad_norm": 0.941338837146759, + "learning_rate": 3.3702242394387294e-05, + "loss": 0.9164, + "step": 16426 + }, + { + "epoch": 2.9246794871794872, + "grad_norm": 0.7108882665634155, + "learning_rate": 3.3691763986773014e-05, + "loss": 0.6956, + "step": 16427 + }, + { + "epoch": 2.92485754985755, + "grad_norm": 0.9144331812858582, + "learning_rate": 3.3681286878330876e-05, + "loss": 0.8148, + "step": 16428 + }, + { + "epoch": 2.9250356125356127, + "grad_norm": 0.882851243019104, + "learning_rate": 3.367081106926615e-05, + "loss": 0.846, + "step": 16429 + }, + { + "epoch": 2.9252136752136755, + "grad_norm": 0.9508523941040039, + "learning_rate": 3.366033655978409e-05, + "loss": 0.9637, + "step": 16430 + }, + { + "epoch": 2.925391737891738, + "grad_norm": 0.7770804166793823, + "learning_rate": 3.3649863350089935e-05, + "loss": 0.6673, + "step": 16431 + }, + { + "epoch": 2.9255698005698005, + "grad_norm": 0.8342770934104919, + "learning_rate": 3.3639391440388845e-05, + "loss": 1.0366, + "step": 16432 + }, + { + "epoch": 2.9257478632478633, + "grad_norm": 0.8854461908340454, + "learning_rate": 3.362892083088609e-05, + "loss": 0.7885, + "step": 16433 + }, + { + "epoch": 2.925925925925926, + "grad_norm": 0.8054807186126709, + "learning_rate": 3.36184515217867e-05, + "loss": 0.7705, + "step": 16434 + }, + { + "epoch": 2.9261039886039883, + "grad_norm": 0.9037294983863831, + "learning_rate": 3.360798351329587e-05, + "loss": 1.077, + "step": 16435 + }, + { + "epoch": 2.926282051282051, + "grad_norm": 0.8743478059768677, + "learning_rate": 3.35975168056187e-05, + "loss": 0.8633, + "step": 16436 + }, + { + "epoch": 2.926460113960114, + "grad_norm": 0.8879369497299194, + "learning_rate": 3.3587051398960245e-05, + "loss": 0.8059, + "step": 16437 + }, + { + "epoch": 2.9266381766381766, + "grad_norm": 0.9445768594741821, + "learning_rate": 3.3576587293525564e-05, + "loss": 0.861, + "step": 16438 + }, + { + "epoch": 2.9268162393162394, + "grad_norm": 0.8131009340286255, + "learning_rate": 3.356612448951967e-05, + "loss": 0.8299, + "step": 16439 + }, + { + "epoch": 2.926994301994302, + "grad_norm": 0.8781847953796387, + "learning_rate": 3.3555662987147515e-05, + "loss": 0.8447, + "step": 16440 + }, + { + "epoch": 2.927172364672365, + "grad_norm": 0.9173591732978821, + "learning_rate": 3.3545202786614206e-05, + "loss": 0.9297, + "step": 16441 + }, + { + "epoch": 2.9273504273504276, + "grad_norm": 0.9596586227416992, + "learning_rate": 3.353474388812452e-05, + "loss": 0.9109, + "step": 16442 + }, + { + "epoch": 2.92752849002849, + "grad_norm": 0.8325486183166504, + "learning_rate": 3.352428629188349e-05, + "loss": 0.6822, + "step": 16443 + }, + { + "epoch": 2.9277065527065527, + "grad_norm": 0.8758354187011719, + "learning_rate": 3.351382999809599e-05, + "loss": 0.9522, + "step": 16444 + }, + { + "epoch": 2.9278846153846154, + "grad_norm": 0.909718930721283, + "learning_rate": 3.3503375006966866e-05, + "loss": 0.8226, + "step": 16445 + }, + { + "epoch": 2.928062678062678, + "grad_norm": 0.9259094595909119, + "learning_rate": 3.3492921318700974e-05, + "loss": 0.9895, + "step": 16446 + }, + { + "epoch": 2.9282407407407405, + "grad_norm": 0.7815300226211548, + "learning_rate": 3.348246893350311e-05, + "loss": 0.8854, + "step": 16447 + }, + { + "epoch": 2.9284188034188032, + "grad_norm": 0.8576910495758057, + "learning_rate": 3.3472017851578154e-05, + "loss": 0.9379, + "step": 16448 + }, + { + "epoch": 2.928596866096866, + "grad_norm": 0.9139176607131958, + "learning_rate": 3.3461568073130735e-05, + "loss": 0.9372, + "step": 16449 + }, + { + "epoch": 2.9287749287749287, + "grad_norm": 0.8304639458656311, + "learning_rate": 3.34511195983657e-05, + "loss": 0.7105, + "step": 16450 + }, + { + "epoch": 2.9289529914529915, + "grad_norm": 0.8689056634902954, + "learning_rate": 3.344067242748774e-05, + "loss": 0.9518, + "step": 16451 + }, + { + "epoch": 2.9291310541310542, + "grad_norm": 0.9180546998977661, + "learning_rate": 3.343022656070154e-05, + "loss": 0.982, + "step": 16452 + }, + { + "epoch": 2.929309116809117, + "grad_norm": 0.9291700124740601, + "learning_rate": 3.341978199821175e-05, + "loss": 1.0003, + "step": 16453 + }, + { + "epoch": 2.9294871794871797, + "grad_norm": 0.970675528049469, + "learning_rate": 3.340933874022304e-05, + "loss": 0.8941, + "step": 16454 + }, + { + "epoch": 2.929665242165242, + "grad_norm": 0.8425672650337219, + "learning_rate": 3.339889678693999e-05, + "loss": 0.7173, + "step": 16455 + }, + { + "epoch": 2.929843304843305, + "grad_norm": 0.9666314721107483, + "learning_rate": 3.3388456138567225e-05, + "loss": 1.0403, + "step": 16456 + }, + { + "epoch": 2.9300213675213675, + "grad_norm": 0.9586226344108582, + "learning_rate": 3.337801679530924e-05, + "loss": 0.7691, + "step": 16457 + }, + { + "epoch": 2.9301994301994303, + "grad_norm": 0.888327419757843, + "learning_rate": 3.33675787573707e-05, + "loss": 0.7019, + "step": 16458 + }, + { + "epoch": 2.9303774928774926, + "grad_norm": 0.7410684823989868, + "learning_rate": 3.335714202495596e-05, + "loss": 0.9935, + "step": 16459 + }, + { + "epoch": 2.9305555555555554, + "grad_norm": 1.0408822298049927, + "learning_rate": 3.3346706598269617e-05, + "loss": 0.9454, + "step": 16460 + }, + { + "epoch": 2.930733618233618, + "grad_norm": 0.9065089821815491, + "learning_rate": 3.333627247751611e-05, + "loss": 0.9665, + "step": 16461 + }, + { + "epoch": 2.930911680911681, + "grad_norm": 0.8904961943626404, + "learning_rate": 3.332583966289985e-05, + "loss": 0.9402, + "step": 16462 + }, + { + "epoch": 2.9310897435897436, + "grad_norm": 0.920364260673523, + "learning_rate": 3.331540815462526e-05, + "loss": 0.8524, + "step": 16463 + }, + { + "epoch": 2.9312678062678064, + "grad_norm": 0.9185073375701904, + "learning_rate": 3.330497795289669e-05, + "loss": 0.8982, + "step": 16464 + }, + { + "epoch": 2.931445868945869, + "grad_norm": 0.9365581274032593, + "learning_rate": 3.32945490579186e-05, + "loss": 0.9272, + "step": 16465 + }, + { + "epoch": 2.931623931623932, + "grad_norm": 0.9139016270637512, + "learning_rate": 3.328412146989518e-05, + "loss": 1.0019, + "step": 16466 + }, + { + "epoch": 2.931801994301994, + "grad_norm": 0.9021140336990356, + "learning_rate": 3.327369518903085e-05, + "loss": 0.866, + "step": 16467 + }, + { + "epoch": 2.931980056980057, + "grad_norm": 0.8053449988365173, + "learning_rate": 3.326327021552984e-05, + "loss": 0.8915, + "step": 16468 + }, + { + "epoch": 2.9321581196581197, + "grad_norm": 1.0013985633850098, + "learning_rate": 3.325284654959643e-05, + "loss": 0.9063, + "step": 16469 + }, + { + "epoch": 2.9323361823361824, + "grad_norm": 0.8236168622970581, + "learning_rate": 3.324242419143483e-05, + "loss": 0.7113, + "step": 16470 + }, + { + "epoch": 2.932514245014245, + "grad_norm": 0.7256088852882385, + "learning_rate": 3.323200314124925e-05, + "loss": 0.6009, + "step": 16471 + }, + { + "epoch": 2.9326923076923075, + "grad_norm": 0.9991055727005005, + "learning_rate": 3.322158339924384e-05, + "loss": 1.039, + "step": 16472 + }, + { + "epoch": 2.9328703703703702, + "grad_norm": 0.8769686818122864, + "learning_rate": 3.3211164965622845e-05, + "loss": 0.989, + "step": 16473 + }, + { + "epoch": 2.933048433048433, + "grad_norm": 0.8226488828659058, + "learning_rate": 3.320074784059026e-05, + "loss": 0.9269, + "step": 16474 + }, + { + "epoch": 2.9332264957264957, + "grad_norm": 0.7763178944587708, + "learning_rate": 3.31903320243503e-05, + "loss": 0.9245, + "step": 16475 + }, + { + "epoch": 2.9334045584045585, + "grad_norm": 0.9204390645027161, + "learning_rate": 3.3179917517107e-05, + "loss": 1.2374, + "step": 16476 + }, + { + "epoch": 2.9335826210826212, + "grad_norm": 0.9970325827598572, + "learning_rate": 3.31695043190644e-05, + "loss": 0.7862, + "step": 16477 + }, + { + "epoch": 2.933760683760684, + "grad_norm": 0.8539462685585022, + "learning_rate": 3.315909243042654e-05, + "loss": 0.8015, + "step": 16478 + }, + { + "epoch": 2.9339387464387463, + "grad_norm": 0.8208832144737244, + "learning_rate": 3.314868185139742e-05, + "loss": 0.8571, + "step": 16479 + }, + { + "epoch": 2.934116809116809, + "grad_norm": 0.8628700375556946, + "learning_rate": 3.313827258218101e-05, + "loss": 0.8274, + "step": 16480 + }, + { + "epoch": 2.934294871794872, + "grad_norm": 1.0472661256790161, + "learning_rate": 3.312786462298124e-05, + "loss": 0.9252, + "step": 16481 + }, + { + "epoch": 2.9344729344729346, + "grad_norm": 0.8292158842086792, + "learning_rate": 3.311745797400202e-05, + "loss": 0.8548, + "step": 16482 + }, + { + "epoch": 2.9346509971509973, + "grad_norm": 0.85657799243927, + "learning_rate": 3.310705263544731e-05, + "loss": 0.7657, + "step": 16483 + }, + { + "epoch": 2.9348290598290596, + "grad_norm": 0.7560283541679382, + "learning_rate": 3.309664860752095e-05, + "loss": 0.8063, + "step": 16484 + }, + { + "epoch": 2.9350071225071224, + "grad_norm": 0.7933059334754944, + "learning_rate": 3.308624589042677e-05, + "loss": 0.7102, + "step": 16485 + }, + { + "epoch": 2.935185185185185, + "grad_norm": 0.8430653810501099, + "learning_rate": 3.3075844484368615e-05, + "loss": 0.8815, + "step": 16486 + }, + { + "epoch": 2.935363247863248, + "grad_norm": 0.8113032579421997, + "learning_rate": 3.306544438955021e-05, + "loss": 0.8144, + "step": 16487 + }, + { + "epoch": 2.9355413105413106, + "grad_norm": 0.8344797492027283, + "learning_rate": 3.3055045606175474e-05, + "loss": 0.7164, + "step": 16488 + }, + { + "epoch": 2.9357193732193734, + "grad_norm": 1.0109050273895264, + "learning_rate": 3.3044648134447964e-05, + "loss": 0.6619, + "step": 16489 + }, + { + "epoch": 2.935897435897436, + "grad_norm": 0.8746094703674316, + "learning_rate": 3.303425197457156e-05, + "loss": 0.73, + "step": 16490 + }, + { + "epoch": 2.9360754985754984, + "grad_norm": 0.919924795627594, + "learning_rate": 3.302385712674981e-05, + "loss": 1.1008, + "step": 16491 + }, + { + "epoch": 2.936253561253561, + "grad_norm": 0.8407595753669739, + "learning_rate": 3.301346359118648e-05, + "loss": 1.0963, + "step": 16492 + }, + { + "epoch": 2.936431623931624, + "grad_norm": 0.8145756721496582, + "learning_rate": 3.3003071368085184e-05, + "loss": 0.805, + "step": 16493 + }, + { + "epoch": 2.9366096866096867, + "grad_norm": 0.8298826813697815, + "learning_rate": 3.299268045764953e-05, + "loss": 0.7075, + "step": 16494 + }, + { + "epoch": 2.9367877492877494, + "grad_norm": 1.1027754545211792, + "learning_rate": 3.2982290860083106e-05, + "loss": 1.0381, + "step": 16495 + }, + { + "epoch": 2.9369658119658117, + "grad_norm": 0.8431075215339661, + "learning_rate": 3.2971902575589476e-05, + "loss": 0.9013, + "step": 16496 + }, + { + "epoch": 2.9371438746438745, + "grad_norm": 1.0045086145401, + "learning_rate": 3.296151560437214e-05, + "loss": 0.9054, + "step": 16497 + }, + { + "epoch": 2.9373219373219372, + "grad_norm": 0.8290889263153076, + "learning_rate": 3.295112994663471e-05, + "loss": 0.8054, + "step": 16498 + }, + { + "epoch": 2.9375, + "grad_norm": 0.7854097485542297, + "learning_rate": 3.2940745602580544e-05, + "loss": 0.7585, + "step": 16499 + }, + { + "epoch": 2.9376780626780628, + "grad_norm": 0.7470735907554626, + "learning_rate": 3.29303625724132e-05, + "loss": 0.7122, + "step": 16500 + }, + { + "epoch": 2.9378561253561255, + "grad_norm": 0.847463071346283, + "learning_rate": 3.2919980856336075e-05, + "loss": 0.7949, + "step": 16501 + }, + { + "epoch": 2.9380341880341883, + "grad_norm": 0.9595068097114563, + "learning_rate": 3.290960045455257e-05, + "loss": 0.8335, + "step": 16502 + }, + { + "epoch": 2.9382122507122506, + "grad_norm": 0.9636897444725037, + "learning_rate": 3.289922136726609e-05, + "loss": 0.7454, + "step": 16503 + }, + { + "epoch": 2.9383903133903133, + "grad_norm": 0.9030486941337585, + "learning_rate": 3.288884359467993e-05, + "loss": 0.8346, + "step": 16504 + }, + { + "epoch": 2.938568376068376, + "grad_norm": 0.8722931146621704, + "learning_rate": 3.287846713699755e-05, + "loss": 1.0182, + "step": 16505 + }, + { + "epoch": 2.938746438746439, + "grad_norm": 0.7612178325653076, + "learning_rate": 3.286809199442209e-05, + "loss": 0.8759, + "step": 16506 + }, + { + "epoch": 2.9389245014245016, + "grad_norm": 0.951334536075592, + "learning_rate": 3.2857718167156956e-05, + "loss": 0.8035, + "step": 16507 + }, + { + "epoch": 2.939102564102564, + "grad_norm": 0.8043029308319092, + "learning_rate": 3.284734565540536e-05, + "loss": 0.7652, + "step": 16508 + }, + { + "epoch": 2.9392806267806266, + "grad_norm": 0.8762648701667786, + "learning_rate": 3.283697445937053e-05, + "loss": 0.9362, + "step": 16509 + }, + { + "epoch": 2.9394586894586894, + "grad_norm": 0.8046880960464478, + "learning_rate": 3.282660457925566e-05, + "loss": 0.7757, + "step": 16510 + }, + { + "epoch": 2.939636752136752, + "grad_norm": 0.7703250050544739, + "learning_rate": 3.281623601526394e-05, + "loss": 0.6699, + "step": 16511 + }, + { + "epoch": 2.939814814814815, + "grad_norm": 0.9165888428688049, + "learning_rate": 3.280586876759847e-05, + "loss": 0.8321, + "step": 16512 + }, + { + "epoch": 2.9399928774928776, + "grad_norm": 0.7389699816703796, + "learning_rate": 3.279550283646249e-05, + "loss": 0.7621, + "step": 16513 + }, + { + "epoch": 2.9401709401709404, + "grad_norm": 0.9256302714347839, + "learning_rate": 3.278513822205897e-05, + "loss": 0.8634, + "step": 16514 + }, + { + "epoch": 2.9403490028490027, + "grad_norm": 0.8384902477264404, + "learning_rate": 3.277477492459109e-05, + "loss": 0.7645, + "step": 16515 + }, + { + "epoch": 2.9405270655270654, + "grad_norm": 0.9446337819099426, + "learning_rate": 3.276441294426178e-05, + "loss": 0.9729, + "step": 16516 + }, + { + "epoch": 2.940705128205128, + "grad_norm": 0.920237123966217, + "learning_rate": 3.275405228127417e-05, + "loss": 0.7834, + "step": 16517 + }, + { + "epoch": 2.940883190883191, + "grad_norm": 0.8432943224906921, + "learning_rate": 3.274369293583121e-05, + "loss": 0.8246, + "step": 16518 + }, + { + "epoch": 2.9410612535612537, + "grad_norm": 0.8046762943267822, + "learning_rate": 3.2733334908135885e-05, + "loss": 0.9363, + "step": 16519 + }, + { + "epoch": 2.941239316239316, + "grad_norm": 0.9555963277816772, + "learning_rate": 3.2722978198391106e-05, + "loss": 0.8699, + "step": 16520 + }, + { + "epoch": 2.9414173789173788, + "grad_norm": 0.8619177937507629, + "learning_rate": 3.2712622806799834e-05, + "loss": 0.8482, + "step": 16521 + }, + { + "epoch": 2.9415954415954415, + "grad_norm": 0.8801655769348145, + "learning_rate": 3.27022687335649e-05, + "loss": 0.9601, + "step": 16522 + }, + { + "epoch": 2.9417735042735043, + "grad_norm": 0.9054547548294067, + "learning_rate": 3.2691915978889244e-05, + "loss": 0.8752, + "step": 16523 + }, + { + "epoch": 2.941951566951567, + "grad_norm": 0.9078481197357178, + "learning_rate": 3.2681564542975675e-05, + "loss": 0.8225, + "step": 16524 + }, + { + "epoch": 2.9421296296296298, + "grad_norm": 0.9574032425880432, + "learning_rate": 3.267121442602701e-05, + "loss": 0.9133, + "step": 16525 + }, + { + "epoch": 2.9423076923076925, + "grad_norm": 0.8808075189590454, + "learning_rate": 3.2660865628246026e-05, + "loss": 0.8544, + "step": 16526 + }, + { + "epoch": 2.942485754985755, + "grad_norm": 0.8007816076278687, + "learning_rate": 3.26505181498355e-05, + "loss": 0.7846, + "step": 16527 + }, + { + "epoch": 2.9426638176638176, + "grad_norm": 0.8623418211936951, + "learning_rate": 3.264017199099816e-05, + "loss": 0.9056, + "step": 16528 + }, + { + "epoch": 2.9428418803418803, + "grad_norm": 0.870961606502533, + "learning_rate": 3.2629827151936695e-05, + "loss": 0.9883, + "step": 16529 + }, + { + "epoch": 2.943019943019943, + "grad_norm": 0.9122142791748047, + "learning_rate": 3.2619483632853885e-05, + "loss": 0.8012, + "step": 16530 + }, + { + "epoch": 2.943198005698006, + "grad_norm": 0.9072979688644409, + "learning_rate": 3.260914143395225e-05, + "loss": 0.9999, + "step": 16531 + }, + { + "epoch": 2.943376068376068, + "grad_norm": 0.8746095299720764, + "learning_rate": 3.259880055543454e-05, + "loss": 0.8022, + "step": 16532 + }, + { + "epoch": 2.943554131054131, + "grad_norm": 1.0012668371200562, + "learning_rate": 3.2588460997503314e-05, + "loss": 0.8909, + "step": 16533 + }, + { + "epoch": 2.9437321937321936, + "grad_norm": 0.913070023059845, + "learning_rate": 3.2578122760361154e-05, + "loss": 0.7729, + "step": 16534 + }, + { + "epoch": 2.9439102564102564, + "grad_norm": 0.791753888130188, + "learning_rate": 3.2567785844210616e-05, + "loss": 0.9731, + "step": 16535 + }, + { + "epoch": 2.944088319088319, + "grad_norm": 0.9673776030540466, + "learning_rate": 3.255745024925425e-05, + "loss": 0.8008, + "step": 16536 + }, + { + "epoch": 2.944266381766382, + "grad_norm": 0.9382752180099487, + "learning_rate": 3.254711597569454e-05, + "loss": 0.9611, + "step": 16537 + }, + { + "epoch": 2.9444444444444446, + "grad_norm": 0.8816630840301514, + "learning_rate": 3.2536783023733975e-05, + "loss": 0.7565, + "step": 16538 + }, + { + "epoch": 2.9446225071225074, + "grad_norm": 0.8474457859992981, + "learning_rate": 3.2526451393574964e-05, + "loss": 0.7766, + "step": 16539 + }, + { + "epoch": 2.9448005698005697, + "grad_norm": 0.9510074257850647, + "learning_rate": 3.251612108542005e-05, + "loss": 0.9899, + "step": 16540 + }, + { + "epoch": 2.9449786324786325, + "grad_norm": 0.776924192905426, + "learning_rate": 3.250579209947149e-05, + "loss": 0.7845, + "step": 16541 + }, + { + "epoch": 2.945156695156695, + "grad_norm": 0.9543585181236267, + "learning_rate": 3.2495464435931756e-05, + "loss": 1.1313, + "step": 16542 + }, + { + "epoch": 2.945334757834758, + "grad_norm": 0.9087918400764465, + "learning_rate": 3.2485138095003164e-05, + "loss": 0.774, + "step": 16543 + }, + { + "epoch": 2.9455128205128203, + "grad_norm": 0.7719675302505493, + "learning_rate": 3.247481307688801e-05, + "loss": 0.7755, + "step": 16544 + }, + { + "epoch": 2.945690883190883, + "grad_norm": 0.8550716638565063, + "learning_rate": 3.246448938178869e-05, + "loss": 0.8455, + "step": 16545 + }, + { + "epoch": 2.9458689458689458, + "grad_norm": 0.8585572838783264, + "learning_rate": 3.2454167009907346e-05, + "loss": 0.7048, + "step": 16546 + }, + { + "epoch": 2.9460470085470085, + "grad_norm": 0.819939136505127, + "learning_rate": 3.2443845961446315e-05, + "loss": 0.7671, + "step": 16547 + }, + { + "epoch": 2.9462250712250713, + "grad_norm": 0.8811594247817993, + "learning_rate": 3.243352623660778e-05, + "loss": 0.9244, + "step": 16548 + }, + { + "epoch": 2.946403133903134, + "grad_norm": 0.8128607869148254, + "learning_rate": 3.242320783559395e-05, + "loss": 0.8103, + "step": 16549 + }, + { + "epoch": 2.9465811965811968, + "grad_norm": 0.778759241104126, + "learning_rate": 3.2412890758606985e-05, + "loss": 0.6261, + "step": 16550 + }, + { + "epoch": 2.9467592592592595, + "grad_norm": 1.0277715921401978, + "learning_rate": 3.240257500584901e-05, + "loss": 0.8728, + "step": 16551 + }, + { + "epoch": 2.946937321937322, + "grad_norm": 0.7647507786750793, + "learning_rate": 3.239226057752217e-05, + "loss": 0.7142, + "step": 16552 + }, + { + "epoch": 2.9471153846153846, + "grad_norm": 0.8381546139717102, + "learning_rate": 3.238194747382855e-05, + "loss": 0.737, + "step": 16553 + }, + { + "epoch": 2.9472934472934473, + "grad_norm": 0.7928317189216614, + "learning_rate": 3.237163569497016e-05, + "loss": 0.8926, + "step": 16554 + }, + { + "epoch": 2.94747150997151, + "grad_norm": 0.7086058259010315, + "learning_rate": 3.236132524114914e-05, + "loss": 0.6845, + "step": 16555 + }, + { + "epoch": 2.9476495726495724, + "grad_norm": 0.83002769947052, + "learning_rate": 3.235101611256739e-05, + "loss": 0.7365, + "step": 16556 + }, + { + "epoch": 2.947827635327635, + "grad_norm": 0.9012778401374817, + "learning_rate": 3.234070830942698e-05, + "loss": 0.9999, + "step": 16557 + }, + { + "epoch": 2.948005698005698, + "grad_norm": 0.8554810881614685, + "learning_rate": 3.233040183192985e-05, + "loss": 0.6699, + "step": 16558 + }, + { + "epoch": 2.9481837606837606, + "grad_norm": 0.9322055578231812, + "learning_rate": 3.2320096680277915e-05, + "loss": 0.8969, + "step": 16559 + }, + { + "epoch": 2.9483618233618234, + "grad_norm": 0.8755966424942017, + "learning_rate": 3.2309792854673095e-05, + "loss": 0.8398, + "step": 16560 + }, + { + "epoch": 2.948539886039886, + "grad_norm": 0.8203766345977783, + "learning_rate": 3.229949035531726e-05, + "loss": 0.8354, + "step": 16561 + }, + { + "epoch": 2.948717948717949, + "grad_norm": 0.8970799446105957, + "learning_rate": 3.228918918241229e-05, + "loss": 0.9074, + "step": 16562 + }, + { + "epoch": 2.9488960113960117, + "grad_norm": 0.8263736963272095, + "learning_rate": 3.227888933615997e-05, + "loss": 0.8896, + "step": 16563 + }, + { + "epoch": 2.949074074074074, + "grad_norm": 1.0277043581008911, + "learning_rate": 3.2268590816762155e-05, + "loss": 0.8706, + "step": 16564 + }, + { + "epoch": 2.9492521367521367, + "grad_norm": 0.8965407013893127, + "learning_rate": 3.225829362442061e-05, + "loss": 0.6796, + "step": 16565 + }, + { + "epoch": 2.9494301994301995, + "grad_norm": 0.8175839185714722, + "learning_rate": 3.224799775933708e-05, + "loss": 0.9243, + "step": 16566 + }, + { + "epoch": 2.949608262108262, + "grad_norm": 0.7979576587677002, + "learning_rate": 3.2237703221713286e-05, + "loss": 0.8008, + "step": 16567 + }, + { + "epoch": 2.9497863247863245, + "grad_norm": 1.054843783378601, + "learning_rate": 3.2227410011750945e-05, + "loss": 1.0279, + "step": 16568 + }, + { + "epoch": 2.9499643874643873, + "grad_norm": 0.7947831749916077, + "learning_rate": 3.221711812965168e-05, + "loss": 0.6767, + "step": 16569 + }, + { + "epoch": 2.95014245014245, + "grad_norm": 0.8702623844146729, + "learning_rate": 3.220682757561725e-05, + "loss": 0.9844, + "step": 16570 + }, + { + "epoch": 2.9503205128205128, + "grad_norm": 0.8653056621551514, + "learning_rate": 3.2196538349849123e-05, + "loss": 1.0747, + "step": 16571 + }, + { + "epoch": 2.9504985754985755, + "grad_norm": 0.9718163013458252, + "learning_rate": 3.2186250452549026e-05, + "loss": 0.7793, + "step": 16572 + }, + { + "epoch": 2.9506766381766383, + "grad_norm": 0.8626788854598999, + "learning_rate": 3.217596388391848e-05, + "loss": 0.9145, + "step": 16573 + }, + { + "epoch": 2.950854700854701, + "grad_norm": 0.9753466844558716, + "learning_rate": 3.2165678644159025e-05, + "loss": 0.9784, + "step": 16574 + }, + { + "epoch": 2.951032763532764, + "grad_norm": 0.795011043548584, + "learning_rate": 3.2155394733472186e-05, + "loss": 0.7457, + "step": 16575 + }, + { + "epoch": 2.951210826210826, + "grad_norm": 0.842991828918457, + "learning_rate": 3.2145112152059454e-05, + "loss": 0.9126, + "step": 16576 + }, + { + "epoch": 2.951388888888889, + "grad_norm": 0.8642476797103882, + "learning_rate": 3.213483090012228e-05, + "loss": 0.7213, + "step": 16577 + }, + { + "epoch": 2.9515669515669516, + "grad_norm": 0.9114034175872803, + "learning_rate": 3.212455097786214e-05, + "loss": 0.651, + "step": 16578 + }, + { + "epoch": 2.9517450142450143, + "grad_norm": 1.0575958490371704, + "learning_rate": 3.211427238548037e-05, + "loss": 0.9727, + "step": 16579 + }, + { + "epoch": 2.9519230769230766, + "grad_norm": 0.8618924617767334, + "learning_rate": 3.210399512317849e-05, + "loss": 0.8593, + "step": 16580 + }, + { + "epoch": 2.9521011396011394, + "grad_norm": 0.768313467502594, + "learning_rate": 3.209371919115771e-05, + "loss": 0.8305, + "step": 16581 + }, + { + "epoch": 2.952279202279202, + "grad_norm": 0.8160355091094971, + "learning_rate": 3.208344458961947e-05, + "loss": 0.7266, + "step": 16582 + }, + { + "epoch": 2.952457264957265, + "grad_norm": 0.908545196056366, + "learning_rate": 3.207317131876506e-05, + "loss": 0.8022, + "step": 16583 + }, + { + "epoch": 2.9526353276353277, + "grad_norm": 1.2593516111373901, + "learning_rate": 3.206289937879571e-05, + "loss": 1.0274, + "step": 16584 + }, + { + "epoch": 2.9528133903133904, + "grad_norm": 0.7789214849472046, + "learning_rate": 3.2052628769912795e-05, + "loss": 0.8993, + "step": 16585 + }, + { + "epoch": 2.952991452991453, + "grad_norm": 0.8475270867347717, + "learning_rate": 3.20423594923174e-05, + "loss": 1.0145, + "step": 16586 + }, + { + "epoch": 2.953169515669516, + "grad_norm": 1.0083874464035034, + "learning_rate": 3.203209154621086e-05, + "loss": 1.02, + "step": 16587 + }, + { + "epoch": 2.953347578347578, + "grad_norm": 0.7013131380081177, + "learning_rate": 3.2021824931794245e-05, + "loss": 0.5449, + "step": 16588 + }, + { + "epoch": 2.953525641025641, + "grad_norm": 0.8298764824867249, + "learning_rate": 3.201155964926878e-05, + "loss": 0.9914, + "step": 16589 + }, + { + "epoch": 2.9537037037037037, + "grad_norm": 0.9371963143348694, + "learning_rate": 3.200129569883556e-05, + "loss": 0.7349, + "step": 16590 + }, + { + "epoch": 2.9538817663817665, + "grad_norm": 0.7932438850402832, + "learning_rate": 3.199103308069571e-05, + "loss": 0.7911, + "step": 16591 + }, + { + "epoch": 2.9540598290598292, + "grad_norm": 0.8415567278862, + "learning_rate": 3.198077179505029e-05, + "loss": 0.766, + "step": 16592 + }, + { + "epoch": 2.9542378917378915, + "grad_norm": 0.9155336618423462, + "learning_rate": 3.197051184210035e-05, + "loss": 0.9622, + "step": 16593 + }, + { + "epoch": 2.9544159544159543, + "grad_norm": 0.8972458839416504, + "learning_rate": 3.196025322204688e-05, + "loss": 1.0202, + "step": 16594 + }, + { + "epoch": 2.954594017094017, + "grad_norm": 0.9620199799537659, + "learning_rate": 3.194999593509096e-05, + "loss": 0.8114, + "step": 16595 + }, + { + "epoch": 2.95477207977208, + "grad_norm": 0.819244921207428, + "learning_rate": 3.1939739981433456e-05, + "loss": 0.7858, + "step": 16596 + }, + { + "epoch": 2.9549501424501425, + "grad_norm": 0.8560270667076111, + "learning_rate": 3.192948536127542e-05, + "loss": 0.9163, + "step": 16597 + }, + { + "epoch": 2.9551282051282053, + "grad_norm": 0.9105932116508484, + "learning_rate": 3.191923207481765e-05, + "loss": 0.7372, + "step": 16598 + }, + { + "epoch": 2.955306267806268, + "grad_norm": 0.888846218585968, + "learning_rate": 3.190898012226114e-05, + "loss": 0.8624, + "step": 16599 + }, + { + "epoch": 2.9554843304843303, + "grad_norm": 0.8116851449012756, + "learning_rate": 3.1898729503806726e-05, + "loss": 0.8549, + "step": 16600 + }, + { + "epoch": 2.955662393162393, + "grad_norm": 0.8970577120780945, + "learning_rate": 3.188848021965522e-05, + "loss": 0.8228, + "step": 16601 + }, + { + "epoch": 2.955840455840456, + "grad_norm": 0.8820711374282837, + "learning_rate": 3.187823227000747e-05, + "loss": 0.7976, + "step": 16602 + }, + { + "epoch": 2.9560185185185186, + "grad_norm": 0.9349139928817749, + "learning_rate": 3.1867985655064205e-05, + "loss": 0.8965, + "step": 16603 + }, + { + "epoch": 2.9561965811965814, + "grad_norm": 0.939132571220398, + "learning_rate": 3.185774037502627e-05, + "loss": 0.9152, + "step": 16604 + }, + { + "epoch": 2.9563746438746437, + "grad_norm": 1.029159665107727, + "learning_rate": 3.184749643009435e-05, + "loss": 1.0904, + "step": 16605 + }, + { + "epoch": 2.9565527065527064, + "grad_norm": 0.966548502445221, + "learning_rate": 3.183725382046917e-05, + "loss": 0.9508, + "step": 16606 + }, + { + "epoch": 2.956730769230769, + "grad_norm": 0.8782771229743958, + "learning_rate": 3.18270125463514e-05, + "loss": 0.9929, + "step": 16607 + }, + { + "epoch": 2.956908831908832, + "grad_norm": 0.9193231463432312, + "learning_rate": 3.18167726079417e-05, + "loss": 0.9942, + "step": 16608 + }, + { + "epoch": 2.9570868945868947, + "grad_norm": 0.8940062522888184, + "learning_rate": 3.180653400544071e-05, + "loss": 1.0594, + "step": 16609 + }, + { + "epoch": 2.9572649572649574, + "grad_norm": 0.8237268328666687, + "learning_rate": 3.179629673904903e-05, + "loss": 0.8211, + "step": 16610 + }, + { + "epoch": 2.95744301994302, + "grad_norm": 0.92745041847229, + "learning_rate": 3.17860608089672e-05, + "loss": 0.8287, + "step": 16611 + }, + { + "epoch": 2.9576210826210825, + "grad_norm": 0.8203856348991394, + "learning_rate": 3.177582621539586e-05, + "loss": 0.9457, + "step": 16612 + }, + { + "epoch": 2.9577991452991452, + "grad_norm": 0.8719314932823181, + "learning_rate": 3.176559295853543e-05, + "loss": 0.8022, + "step": 16613 + }, + { + "epoch": 2.957977207977208, + "grad_norm": 0.8742199540138245, + "learning_rate": 3.175536103858648e-05, + "loss": 0.976, + "step": 16614 + }, + { + "epoch": 2.9581552706552707, + "grad_norm": 0.8677577972412109, + "learning_rate": 3.174513045574947e-05, + "loss": 0.9513, + "step": 16615 + }, + { + "epoch": 2.9583333333333335, + "grad_norm": 0.809138298034668, + "learning_rate": 3.173490121022485e-05, + "loss": 0.7001, + "step": 16616 + }, + { + "epoch": 2.958511396011396, + "grad_norm": 0.9672527313232422, + "learning_rate": 3.1724673302213025e-05, + "loss": 0.9932, + "step": 16617 + }, + { + "epoch": 2.9586894586894585, + "grad_norm": 0.8101853728294373, + "learning_rate": 3.17144467319144e-05, + "loss": 0.8236, + "step": 16618 + }, + { + "epoch": 2.9588675213675213, + "grad_norm": 0.8965981006622314, + "learning_rate": 3.170422149952931e-05, + "loss": 0.9353, + "step": 16619 + }, + { + "epoch": 2.959045584045584, + "grad_norm": 0.8267533183097839, + "learning_rate": 3.1693997605258184e-05, + "loss": 0.6445, + "step": 16620 + }, + { + "epoch": 2.959223646723647, + "grad_norm": 0.9084979891777039, + "learning_rate": 3.168377504930122e-05, + "loss": 0.9311, + "step": 16621 + }, + { + "epoch": 2.9594017094017095, + "grad_norm": 0.8414687514305115, + "learning_rate": 3.1673553831858805e-05, + "loss": 0.7663, + "step": 16622 + }, + { + "epoch": 2.9595797720797723, + "grad_norm": 0.8619266748428345, + "learning_rate": 3.166333395313116e-05, + "loss": 0.8524, + "step": 16623 + }, + { + "epoch": 2.9597578347578346, + "grad_norm": 0.9963071346282959, + "learning_rate": 3.1653115413318534e-05, + "loss": 0.813, + "step": 16624 + }, + { + "epoch": 2.9599358974358974, + "grad_norm": 0.8264978528022766, + "learning_rate": 3.164289821262113e-05, + "loss": 0.9716, + "step": 16625 + }, + { + "epoch": 2.96011396011396, + "grad_norm": 0.8166584372520447, + "learning_rate": 3.163268235123911e-05, + "loss": 0.7492, + "step": 16626 + }, + { + "epoch": 2.960292022792023, + "grad_norm": 0.809241533279419, + "learning_rate": 3.1622467829372724e-05, + "loss": 0.9033, + "step": 16627 + }, + { + "epoch": 2.9604700854700856, + "grad_norm": 1.0441617965698242, + "learning_rate": 3.161225464722197e-05, + "loss": 0.7305, + "step": 16628 + }, + { + "epoch": 2.960648148148148, + "grad_norm": 0.9226490259170532, + "learning_rate": 3.160204280498705e-05, + "loss": 0.6896, + "step": 16629 + }, + { + "epoch": 2.9608262108262107, + "grad_norm": 0.7890266180038452, + "learning_rate": 3.159183230286803e-05, + "loss": 0.873, + "step": 16630 + }, + { + "epoch": 2.9610042735042734, + "grad_norm": 0.8491777181625366, + "learning_rate": 3.1581623141064934e-05, + "loss": 0.8828, + "step": 16631 + }, + { + "epoch": 2.961182336182336, + "grad_norm": 0.7522078156471252, + "learning_rate": 3.157141531977782e-05, + "loss": 0.7717, + "step": 16632 + }, + { + "epoch": 2.961360398860399, + "grad_norm": 0.9260183572769165, + "learning_rate": 3.156120883920667e-05, + "loss": 0.9715, + "step": 16633 + }, + { + "epoch": 2.9615384615384617, + "grad_norm": 0.6620128154754639, + "learning_rate": 3.1551003699551465e-05, + "loss": 0.5624, + "step": 16634 + }, + { + "epoch": 2.9617165242165244, + "grad_norm": 0.7571594715118408, + "learning_rate": 3.154079990101214e-05, + "loss": 0.6911, + "step": 16635 + }, + { + "epoch": 2.9618945868945867, + "grad_norm": 0.8571279644966125, + "learning_rate": 3.153059744378861e-05, + "loss": 0.9057, + "step": 16636 + }, + { + "epoch": 2.9620726495726495, + "grad_norm": 0.8895478248596191, + "learning_rate": 3.152039632808085e-05, + "loss": 0.7097, + "step": 16637 + }, + { + "epoch": 2.9622507122507122, + "grad_norm": 0.8340024352073669, + "learning_rate": 3.15101965540886e-05, + "loss": 0.7719, + "step": 16638 + }, + { + "epoch": 2.962428774928775, + "grad_norm": 0.8263829946517944, + "learning_rate": 3.149999812201182e-05, + "loss": 0.8561, + "step": 16639 + }, + { + "epoch": 2.9626068376068377, + "grad_norm": 0.9083819389343262, + "learning_rate": 3.148980103205027e-05, + "loss": 0.9319, + "step": 16640 + }, + { + "epoch": 2.9627849002849, + "grad_norm": 0.9346078038215637, + "learning_rate": 3.147960528440372e-05, + "loss": 0.8036, + "step": 16641 + }, + { + "epoch": 2.962962962962963, + "grad_norm": 0.907319188117981, + "learning_rate": 3.146941087927203e-05, + "loss": 0.9228, + "step": 16642 + }, + { + "epoch": 2.9631410256410255, + "grad_norm": 0.7912126183509827, + "learning_rate": 3.1459217816854815e-05, + "loss": 0.7996, + "step": 16643 + }, + { + "epoch": 2.9633190883190883, + "grad_norm": 0.8844919800758362, + "learning_rate": 3.1449026097351896e-05, + "loss": 1.0235, + "step": 16644 + }, + { + "epoch": 2.963497150997151, + "grad_norm": 0.7468230128288269, + "learning_rate": 3.143883572096286e-05, + "loss": 0.7292, + "step": 16645 + }, + { + "epoch": 2.963675213675214, + "grad_norm": 0.8521941900253296, + "learning_rate": 3.142864668788744e-05, + "loss": 0.9012, + "step": 16646 + }, + { + "epoch": 2.9638532763532766, + "grad_norm": 0.9340695738792419, + "learning_rate": 3.141845899832524e-05, + "loss": 0.8736, + "step": 16647 + }, + { + "epoch": 2.9640313390313393, + "grad_norm": 0.859395444393158, + "learning_rate": 3.140827265247588e-05, + "loss": 0.796, + "step": 16648 + }, + { + "epoch": 2.9642094017094016, + "grad_norm": 0.8320850729942322, + "learning_rate": 3.139808765053892e-05, + "loss": 0.86, + "step": 16649 + }, + { + "epoch": 2.9643874643874644, + "grad_norm": 0.8911257386207581, + "learning_rate": 3.138790399271393e-05, + "loss": 0.966, + "step": 16650 + }, + { + "epoch": 2.964565527065527, + "grad_norm": 0.8817025423049927, + "learning_rate": 3.13777216792004e-05, + "loss": 0.983, + "step": 16651 + }, + { + "epoch": 2.96474358974359, + "grad_norm": 0.7765538692474365, + "learning_rate": 3.136754071019793e-05, + "loss": 0.643, + "step": 16652 + }, + { + "epoch": 2.964921652421652, + "grad_norm": 0.7961843609809875, + "learning_rate": 3.135736108590586e-05, + "loss": 0.733, + "step": 16653 + }, + { + "epoch": 2.965099715099715, + "grad_norm": 0.7910877466201782, + "learning_rate": 3.134718280652373e-05, + "loss": 0.7291, + "step": 16654 + }, + { + "epoch": 2.9652777777777777, + "grad_norm": 0.9326547384262085, + "learning_rate": 3.1337005872250956e-05, + "loss": 0.8834, + "step": 16655 + }, + { + "epoch": 2.9654558404558404, + "grad_norm": 0.9362995624542236, + "learning_rate": 3.132683028328691e-05, + "loss": 0.9693, + "step": 16656 + }, + { + "epoch": 2.965633903133903, + "grad_norm": 0.8322434425354004, + "learning_rate": 3.131665603983096e-05, + "loss": 0.9782, + "step": 16657 + }, + { + "epoch": 2.965811965811966, + "grad_norm": 0.8336097598075867, + "learning_rate": 3.130648314208247e-05, + "loss": 0.9995, + "step": 16658 + }, + { + "epoch": 2.9659900284900287, + "grad_norm": 0.8637044429779053, + "learning_rate": 3.129631159024074e-05, + "loss": 0.8291, + "step": 16659 + }, + { + "epoch": 2.9661680911680914, + "grad_norm": 0.7853943109512329, + "learning_rate": 3.128614138450506e-05, + "loss": 0.9016, + "step": 16660 + }, + { + "epoch": 2.9663461538461537, + "grad_norm": 0.9506492614746094, + "learning_rate": 3.1275972525074674e-05, + "loss": 0.9762, + "step": 16661 + }, + { + "epoch": 2.9665242165242165, + "grad_norm": 0.9375113248825073, + "learning_rate": 3.126580501214887e-05, + "loss": 0.8981, + "step": 16662 + }, + { + "epoch": 2.9667022792022792, + "grad_norm": 0.9404717683792114, + "learning_rate": 3.125563884592684e-05, + "loss": 0.7506, + "step": 16663 + }, + { + "epoch": 2.966880341880342, + "grad_norm": 0.9678782820701599, + "learning_rate": 3.124547402660776e-05, + "loss": 0.896, + "step": 16664 + }, + { + "epoch": 2.9670584045584043, + "grad_norm": 0.8530639410018921, + "learning_rate": 3.12353105543908e-05, + "loss": 0.8792, + "step": 16665 + }, + { + "epoch": 2.967236467236467, + "grad_norm": 0.8015365600585938, + "learning_rate": 3.122514842947504e-05, + "loss": 0.7052, + "step": 16666 + }, + { + "epoch": 2.96741452991453, + "grad_norm": 0.8753054141998291, + "learning_rate": 3.121498765205969e-05, + "loss": 0.9032, + "step": 16667 + }, + { + "epoch": 2.9675925925925926, + "grad_norm": 1.0053389072418213, + "learning_rate": 3.1204828222343716e-05, + "loss": 1.0236, + "step": 16668 + }, + { + "epoch": 2.9677706552706553, + "grad_norm": 0.8506449460983276, + "learning_rate": 3.119467014052628e-05, + "loss": 0.6762, + "step": 16669 + }, + { + "epoch": 2.967948717948718, + "grad_norm": 0.9376404881477356, + "learning_rate": 3.118451340680629e-05, + "loss": 0.9391, + "step": 16670 + }, + { + "epoch": 2.968126780626781, + "grad_norm": 0.8381097912788391, + "learning_rate": 3.117435802138284e-05, + "loss": 0.7343, + "step": 16671 + }, + { + "epoch": 2.9683048433048436, + "grad_norm": 0.8525682091712952, + "learning_rate": 3.116420398445488e-05, + "loss": 0.7928, + "step": 16672 + }, + { + "epoch": 2.968482905982906, + "grad_norm": 0.9399489760398865, + "learning_rate": 3.115405129622133e-05, + "loss": 0.975, + "step": 16673 + }, + { + "epoch": 2.9686609686609686, + "grad_norm": 0.8394346237182617, + "learning_rate": 3.114389995688114e-05, + "loss": 0.7591, + "step": 16674 + }, + { + "epoch": 2.9688390313390314, + "grad_norm": 0.8935068845748901, + "learning_rate": 3.11337499666332e-05, + "loss": 0.8899, + "step": 16675 + }, + { + "epoch": 2.969017094017094, + "grad_norm": 0.8111040592193604, + "learning_rate": 3.112360132567633e-05, + "loss": 0.795, + "step": 16676 + }, + { + "epoch": 2.9691951566951564, + "grad_norm": 0.8854177594184875, + "learning_rate": 3.1113454034209486e-05, + "loss": 1.0677, + "step": 16677 + }, + { + "epoch": 2.969373219373219, + "grad_norm": 0.9821479916572571, + "learning_rate": 3.110330809243134e-05, + "loss": 0.7859, + "step": 16678 + }, + { + "epoch": 2.969551282051282, + "grad_norm": 0.9066275954246521, + "learning_rate": 3.109316350054079e-05, + "loss": 1.1727, + "step": 16679 + }, + { + "epoch": 2.9697293447293447, + "grad_norm": 0.981238603591919, + "learning_rate": 3.108302025873656e-05, + "loss": 1.0036, + "step": 16680 + }, + { + "epoch": 2.9699074074074074, + "grad_norm": 0.8290690779685974, + "learning_rate": 3.107287836721737e-05, + "loss": 0.911, + "step": 16681 + }, + { + "epoch": 2.97008547008547, + "grad_norm": 0.8419190049171448, + "learning_rate": 3.106273782618196e-05, + "loss": 0.688, + "step": 16682 + }, + { + "epoch": 2.970263532763533, + "grad_norm": 0.9250679612159729, + "learning_rate": 3.1052598635828964e-05, + "loss": 0.8506, + "step": 16683 + }, + { + "epoch": 2.9704415954415957, + "grad_norm": 0.9414278864860535, + "learning_rate": 3.104246079635713e-05, + "loss": 0.8501, + "step": 16684 + }, + { + "epoch": 2.970619658119658, + "grad_norm": 0.9107208847999573, + "learning_rate": 3.1032324307964974e-05, + "loss": 0.8234, + "step": 16685 + }, + { + "epoch": 2.9707977207977208, + "grad_norm": 0.8691245317459106, + "learning_rate": 3.102218917085119e-05, + "loss": 0.9341, + "step": 16686 + }, + { + "epoch": 2.9709757834757835, + "grad_norm": 0.8816282153129578, + "learning_rate": 3.101205538521431e-05, + "loss": 0.9412, + "step": 16687 + }, + { + "epoch": 2.9711538461538463, + "grad_norm": 0.8969736695289612, + "learning_rate": 3.100192295125289e-05, + "loss": 0.9468, + "step": 16688 + }, + { + "epoch": 2.9713319088319086, + "grad_norm": 1.0070735216140747, + "learning_rate": 3.099179186916548e-05, + "loss": 0.8587, + "step": 16689 + }, + { + "epoch": 2.9715099715099713, + "grad_norm": 0.9801154136657715, + "learning_rate": 3.0981662139150537e-05, + "loss": 0.8999, + "step": 16690 + }, + { + "epoch": 2.971688034188034, + "grad_norm": 0.9949473142623901, + "learning_rate": 3.0971533761406526e-05, + "loss": 1.0552, + "step": 16691 + }, + { + "epoch": 2.971866096866097, + "grad_norm": 0.9654440879821777, + "learning_rate": 3.096140673613198e-05, + "loss": 0.8618, + "step": 16692 + }, + { + "epoch": 2.9720441595441596, + "grad_norm": 0.9247317910194397, + "learning_rate": 3.0951281063525185e-05, + "loss": 0.8915, + "step": 16693 + }, + { + "epoch": 2.9722222222222223, + "grad_norm": 1.0254271030426025, + "learning_rate": 3.094115674378467e-05, + "loss": 0.9098, + "step": 16694 + }, + { + "epoch": 2.972400284900285, + "grad_norm": 0.7822396159172058, + "learning_rate": 3.0931033777108666e-05, + "loss": 0.8497, + "step": 16695 + }, + { + "epoch": 2.972578347578348, + "grad_norm": 0.9368909001350403, + "learning_rate": 3.092091216369561e-05, + "loss": 0.8978, + "step": 16696 + }, + { + "epoch": 2.97275641025641, + "grad_norm": 1.0181694030761719, + "learning_rate": 3.0910791903743786e-05, + "loss": 1.0605, + "step": 16697 + }, + { + "epoch": 2.972934472934473, + "grad_norm": 0.8718426823616028, + "learning_rate": 3.090067299745146e-05, + "loss": 0.9844, + "step": 16698 + }, + { + "epoch": 2.9731125356125356, + "grad_norm": 0.8327153325080872, + "learning_rate": 3.089055544501692e-05, + "loss": 0.8795, + "step": 16699 + }, + { + "epoch": 2.9732905982905984, + "grad_norm": 0.7970293164253235, + "learning_rate": 3.088043924663838e-05, + "loss": 0.794, + "step": 16700 + }, + { + "epoch": 2.9734686609686607, + "grad_norm": 0.8885688185691833, + "learning_rate": 3.0870324402514014e-05, + "loss": 0.984, + "step": 16701 + }, + { + "epoch": 2.9736467236467234, + "grad_norm": 0.9126071333885193, + "learning_rate": 3.086021091284207e-05, + "loss": 0.8962, + "step": 16702 + }, + { + "epoch": 2.973824786324786, + "grad_norm": 0.9220738410949707, + "learning_rate": 3.085009877782067e-05, + "loss": 0.9129, + "step": 16703 + }, + { + "epoch": 2.974002849002849, + "grad_norm": 0.8738197684288025, + "learning_rate": 3.0839987997647935e-05, + "loss": 0.871, + "step": 16704 + }, + { + "epoch": 2.9741809116809117, + "grad_norm": 0.7505079507827759, + "learning_rate": 3.0829878572521964e-05, + "loss": 0.7934, + "step": 16705 + }, + { + "epoch": 2.9743589743589745, + "grad_norm": 0.8328925371170044, + "learning_rate": 3.081977050264084e-05, + "loss": 0.8431, + "step": 16706 + }, + { + "epoch": 2.974537037037037, + "grad_norm": 0.8248029947280884, + "learning_rate": 3.08096637882026e-05, + "loss": 0.7877, + "step": 16707 + }, + { + "epoch": 2.9747150997151, + "grad_norm": 1.1086161136627197, + "learning_rate": 3.079955842940524e-05, + "loss": 1.1196, + "step": 16708 + }, + { + "epoch": 2.9748931623931623, + "grad_norm": 0.8271856307983398, + "learning_rate": 3.078945442644684e-05, + "loss": 0.946, + "step": 16709 + }, + { + "epoch": 2.975071225071225, + "grad_norm": 0.9062432050704956, + "learning_rate": 3.0779351779525246e-05, + "loss": 0.9369, + "step": 16710 + }, + { + "epoch": 2.9752492877492878, + "grad_norm": 0.9849454760551453, + "learning_rate": 3.0769250488838496e-05, + "loss": 0.7748, + "step": 16711 + }, + { + "epoch": 2.9754273504273505, + "grad_norm": 0.8420920372009277, + "learning_rate": 3.075915055458447e-05, + "loss": 0.9954, + "step": 16712 + }, + { + "epoch": 2.9756054131054133, + "grad_norm": 1.027130126953125, + "learning_rate": 3.074905197696104e-05, + "loss": 1.1053, + "step": 16713 + }, + { + "epoch": 2.9757834757834756, + "grad_norm": 0.8968556523323059, + "learning_rate": 3.073895475616609e-05, + "loss": 0.7487, + "step": 16714 + }, + { + "epoch": 2.9759615384615383, + "grad_norm": 0.9237984418869019, + "learning_rate": 3.072885889239745e-05, + "loss": 1.2853, + "step": 16715 + }, + { + "epoch": 2.976139601139601, + "grad_norm": 0.7972691059112549, + "learning_rate": 3.071876438585292e-05, + "loss": 0.6724, + "step": 16716 + }, + { + "epoch": 2.976317663817664, + "grad_norm": 0.8158389925956726, + "learning_rate": 3.0708671236730266e-05, + "loss": 0.8302, + "step": 16717 + }, + { + "epoch": 2.9764957264957266, + "grad_norm": 0.7947682738304138, + "learning_rate": 3.0698579445227236e-05, + "loss": 0.7817, + "step": 16718 + }, + { + "epoch": 2.9766737891737893, + "grad_norm": 0.847747802734375, + "learning_rate": 3.068848901154166e-05, + "loss": 0.8391, + "step": 16719 + }, + { + "epoch": 2.976851851851852, + "grad_norm": 0.7906867265701294, + "learning_rate": 3.067839993587107e-05, + "loss": 0.8935, + "step": 16720 + }, + { + "epoch": 2.9770299145299144, + "grad_norm": 0.9028356075286865, + "learning_rate": 3.066831221841328e-05, + "loss": 0.821, + "step": 16721 + }, + { + "epoch": 2.977207977207977, + "grad_norm": 0.8177092671394348, + "learning_rate": 3.065822585936589e-05, + "loss": 0.7956, + "step": 16722 + }, + { + "epoch": 2.97738603988604, + "grad_norm": 0.771271824836731, + "learning_rate": 3.064814085892647e-05, + "loss": 0.8675, + "step": 16723 + }, + { + "epoch": 2.9775641025641026, + "grad_norm": 0.8575150966644287, + "learning_rate": 3.063805721729274e-05, + "loss": 0.6999, + "step": 16724 + }, + { + "epoch": 2.9777421652421654, + "grad_norm": 0.8909936547279358, + "learning_rate": 3.062797493466212e-05, + "loss": 0.9325, + "step": 16725 + }, + { + "epoch": 2.9779202279202277, + "grad_norm": 0.836875319480896, + "learning_rate": 3.061789401123226e-05, + "loss": 0.9276, + "step": 16726 + }, + { + "epoch": 2.9780982905982905, + "grad_norm": 0.8965426683425903, + "learning_rate": 3.060781444720065e-05, + "loss": 0.7728, + "step": 16727 + }, + { + "epoch": 2.978276353276353, + "grad_norm": 0.8298100829124451, + "learning_rate": 3.059773624276475e-05, + "loss": 0.926, + "step": 16728 + }, + { + "epoch": 2.978454415954416, + "grad_norm": 0.9369875192642212, + "learning_rate": 3.058765939812204e-05, + "loss": 0.8598, + "step": 16729 + }, + { + "epoch": 2.9786324786324787, + "grad_norm": 1.019013524055481, + "learning_rate": 3.057758391346995e-05, + "loss": 0.9075, + "step": 16730 + }, + { + "epoch": 2.9788105413105415, + "grad_norm": 0.8597549200057983, + "learning_rate": 3.05675097890059e-05, + "loss": 0.8279, + "step": 16731 + }, + { + "epoch": 2.978988603988604, + "grad_norm": 0.8331323862075806, + "learning_rate": 3.055743702492726e-05, + "loss": 0.8366, + "step": 16732 + }, + { + "epoch": 2.9791666666666665, + "grad_norm": 0.7987640500068665, + "learning_rate": 3.054736562143135e-05, + "loss": 0.7618, + "step": 16733 + }, + { + "epoch": 2.9793447293447293, + "grad_norm": 0.84430330991745, + "learning_rate": 3.0537295578715606e-05, + "loss": 0.9791, + "step": 16734 + }, + { + "epoch": 2.979522792022792, + "grad_norm": 0.7811259627342224, + "learning_rate": 3.052722689697719e-05, + "loss": 0.7342, + "step": 16735 + }, + { + "epoch": 2.9797008547008548, + "grad_norm": 0.8311488032341003, + "learning_rate": 3.0517159576413477e-05, + "loss": 0.64, + "step": 16736 + }, + { + "epoch": 2.9798789173789175, + "grad_norm": 0.8865106105804443, + "learning_rate": 3.0507093617221683e-05, + "loss": 0.9196, + "step": 16737 + }, + { + "epoch": 2.98005698005698, + "grad_norm": 0.7360345125198364, + "learning_rate": 3.0497029019599033e-05, + "loss": 0.8709, + "step": 16738 + }, + { + "epoch": 2.9802350427350426, + "grad_norm": 0.8705546855926514, + "learning_rate": 3.0486965783742726e-05, + "loss": 0.7567, + "step": 16739 + }, + { + "epoch": 2.9804131054131053, + "grad_norm": 0.8622585535049438, + "learning_rate": 3.0476903909849908e-05, + "loss": 0.7909, + "step": 16740 + }, + { + "epoch": 2.980591168091168, + "grad_norm": 0.7877208590507507, + "learning_rate": 3.046684339811775e-05, + "loss": 0.7046, + "step": 16741 + }, + { + "epoch": 2.980769230769231, + "grad_norm": 0.915133535861969, + "learning_rate": 3.045678424874332e-05, + "loss": 0.9142, + "step": 16742 + }, + { + "epoch": 2.9809472934472936, + "grad_norm": 0.8539748191833496, + "learning_rate": 3.044672646192377e-05, + "loss": 0.9333, + "step": 16743 + }, + { + "epoch": 2.9811253561253563, + "grad_norm": 1.1402504444122314, + "learning_rate": 3.043667003785612e-05, + "loss": 0.8041, + "step": 16744 + }, + { + "epoch": 2.9813034188034186, + "grad_norm": 0.7355749607086182, + "learning_rate": 3.042661497673742e-05, + "loss": 0.8057, + "step": 16745 + }, + { + "epoch": 2.9814814814814814, + "grad_norm": 0.8524863719940186, + "learning_rate": 3.041656127876469e-05, + "loss": 0.7118, + "step": 16746 + }, + { + "epoch": 2.981659544159544, + "grad_norm": 0.9434519410133362, + "learning_rate": 3.040650894413487e-05, + "loss": 0.8104, + "step": 16747 + }, + { + "epoch": 2.981837606837607, + "grad_norm": 0.8716524839401245, + "learning_rate": 3.0396457973044923e-05, + "loss": 0.7502, + "step": 16748 + }, + { + "epoch": 2.9820156695156697, + "grad_norm": 0.942481279373169, + "learning_rate": 3.0386408365691855e-05, + "loss": 0.8506, + "step": 16749 + }, + { + "epoch": 2.982193732193732, + "grad_norm": 0.7921879291534424, + "learning_rate": 3.037636012227244e-05, + "loss": 0.7005, + "step": 16750 + }, + { + "epoch": 2.9823717948717947, + "grad_norm": 0.8415821194648743, + "learning_rate": 3.0366313242983645e-05, + "loss": 0.987, + "step": 16751 + }, + { + "epoch": 2.9825498575498575, + "grad_norm": 0.9556678533554077, + "learning_rate": 3.0356267728022293e-05, + "loss": 0.9198, + "step": 16752 + }, + { + "epoch": 2.98272792022792, + "grad_norm": 0.856810450553894, + "learning_rate": 3.0346223577585208e-05, + "loss": 0.9634, + "step": 16753 + }, + { + "epoch": 2.982905982905983, + "grad_norm": 0.8425500392913818, + "learning_rate": 3.0336180791869172e-05, + "loss": 0.8725, + "step": 16754 + }, + { + "epoch": 2.9830840455840457, + "grad_norm": 0.773210883140564, + "learning_rate": 3.032613937107096e-05, + "loss": 0.8363, + "step": 16755 + }, + { + "epoch": 2.9832621082621085, + "grad_norm": 0.7771822214126587, + "learning_rate": 3.0316099315387324e-05, + "loss": 0.8586, + "step": 16756 + }, + { + "epoch": 2.9834401709401708, + "grad_norm": 0.8261004686355591, + "learning_rate": 3.0306060625014954e-05, + "loss": 0.8185, + "step": 16757 + }, + { + "epoch": 2.9836182336182335, + "grad_norm": 1.074456810951233, + "learning_rate": 3.029602330015052e-05, + "loss": 0.9008, + "step": 16758 + }, + { + "epoch": 2.9837962962962963, + "grad_norm": 0.8160609006881714, + "learning_rate": 3.0285987340990774e-05, + "loss": 0.8494, + "step": 16759 + }, + { + "epoch": 2.983974358974359, + "grad_norm": 0.8296660780906677, + "learning_rate": 3.0275952747732227e-05, + "loss": 0.9089, + "step": 16760 + }, + { + "epoch": 2.984152421652422, + "grad_norm": 0.815726637840271, + "learning_rate": 3.0265919520571572e-05, + "loss": 0.8605, + "step": 16761 + }, + { + "epoch": 2.984330484330484, + "grad_norm": 0.7600420117378235, + "learning_rate": 3.0255887659705374e-05, + "loss": 0.6833, + "step": 16762 + }, + { + "epoch": 2.984508547008547, + "grad_norm": 0.8039379119873047, + "learning_rate": 3.024585716533014e-05, + "loss": 0.8156, + "step": 16763 + }, + { + "epoch": 2.9846866096866096, + "grad_norm": 0.8797104358673096, + "learning_rate": 3.02358280376425e-05, + "loss": 0.9965, + "step": 16764 + }, + { + "epoch": 2.9848646723646723, + "grad_norm": 0.8681818246841431, + "learning_rate": 3.0225800276838822e-05, + "loss": 0.8967, + "step": 16765 + }, + { + "epoch": 2.985042735042735, + "grad_norm": 0.804710328578949, + "learning_rate": 3.0215773883115706e-05, + "loss": 0.7424, + "step": 16766 + }, + { + "epoch": 2.985220797720798, + "grad_norm": 0.7346279621124268, + "learning_rate": 3.0205748856669467e-05, + "loss": 0.6986, + "step": 16767 + }, + { + "epoch": 2.9853988603988606, + "grad_norm": 0.8896076083183289, + "learning_rate": 3.0195725197696634e-05, + "loss": 0.8472, + "step": 16768 + }, + { + "epoch": 2.9855769230769234, + "grad_norm": 0.8281870484352112, + "learning_rate": 3.0185702906393555e-05, + "loss": 0.9873, + "step": 16769 + }, + { + "epoch": 2.9857549857549857, + "grad_norm": 0.7376837134361267, + "learning_rate": 3.0175681982956607e-05, + "loss": 1.0171, + "step": 16770 + }, + { + "epoch": 2.9859330484330484, + "grad_norm": 0.9346923828125, + "learning_rate": 3.016566242758212e-05, + "loss": 0.8742, + "step": 16771 + }, + { + "epoch": 2.986111111111111, + "grad_norm": 0.918557345867157, + "learning_rate": 3.0155644240466406e-05, + "loss": 1.0504, + "step": 16772 + }, + { + "epoch": 2.986289173789174, + "grad_norm": 0.8885583281517029, + "learning_rate": 3.0145627421805733e-05, + "loss": 0.7462, + "step": 16773 + }, + { + "epoch": 2.986467236467236, + "grad_norm": 0.8475548028945923, + "learning_rate": 3.013561197179644e-05, + "loss": 0.7353, + "step": 16774 + }, + { + "epoch": 2.986645299145299, + "grad_norm": 0.8925435543060303, + "learning_rate": 3.0125597890634626e-05, + "loss": 0.8462, + "step": 16775 + }, + { + "epoch": 2.9868233618233617, + "grad_norm": 0.923281729221344, + "learning_rate": 3.0115585178516648e-05, + "loss": 0.9725, + "step": 16776 + }, + { + "epoch": 2.9870014245014245, + "grad_norm": 0.9134986996650696, + "learning_rate": 3.0105573835638533e-05, + "loss": 0.9428, + "step": 16777 + }, + { + "epoch": 2.9871794871794872, + "grad_norm": 0.8284963369369507, + "learning_rate": 3.009556386219654e-05, + "loss": 0.9134, + "step": 16778 + }, + { + "epoch": 2.98735754985755, + "grad_norm": 0.9571327567100525, + "learning_rate": 3.0085555258386756e-05, + "loss": 0.8407, + "step": 16779 + }, + { + "epoch": 2.9875356125356127, + "grad_norm": 0.8699741363525391, + "learning_rate": 3.0075548024405254e-05, + "loss": 0.862, + "step": 16780 + }, + { + "epoch": 2.9877136752136755, + "grad_norm": 0.9462196826934814, + "learning_rate": 3.006554216044819e-05, + "loss": 0.8936, + "step": 16781 + }, + { + "epoch": 2.987891737891738, + "grad_norm": 0.9415904879570007, + "learning_rate": 3.0055537666711496e-05, + "loss": 1.0089, + "step": 16782 + }, + { + "epoch": 2.9880698005698005, + "grad_norm": 0.8529258370399475, + "learning_rate": 3.0045534543391275e-05, + "loss": 0.7896, + "step": 16783 + }, + { + "epoch": 2.9882478632478633, + "grad_norm": 0.7967036962509155, + "learning_rate": 3.0035532790683486e-05, + "loss": 0.7661, + "step": 16784 + }, + { + "epoch": 2.988425925925926, + "grad_norm": 1.087546944618225, + "learning_rate": 3.0025532408784097e-05, + "loss": 0.7232, + "step": 16785 + }, + { + "epoch": 2.9886039886039883, + "grad_norm": 1.0032312870025635, + "learning_rate": 3.001553339788903e-05, + "loss": 0.8462, + "step": 16786 + }, + { + "epoch": 2.988782051282051, + "grad_norm": 0.7726916074752808, + "learning_rate": 3.0005535758194216e-05, + "loss": 0.8318, + "step": 16787 + }, + { + "epoch": 2.988960113960114, + "grad_norm": 0.8814677000045776, + "learning_rate": 2.999553948989552e-05, + "loss": 0.8058, + "step": 16788 + }, + { + "epoch": 2.9891381766381766, + "grad_norm": 0.8866276741027832, + "learning_rate": 2.9985544593188818e-05, + "loss": 0.8576, + "step": 16789 + }, + { + "epoch": 2.9893162393162394, + "grad_norm": 0.9877329468727112, + "learning_rate": 2.997555106826988e-05, + "loss": 0.8675, + "step": 16790 + }, + { + "epoch": 2.989494301994302, + "grad_norm": 0.842682421207428, + "learning_rate": 2.9965558915334636e-05, + "loss": 1.094, + "step": 16791 + }, + { + "epoch": 2.989672364672365, + "grad_norm": 0.8740942478179932, + "learning_rate": 2.9955568134578703e-05, + "loss": 0.904, + "step": 16792 + }, + { + "epoch": 2.9898504273504276, + "grad_norm": 0.9166956543922424, + "learning_rate": 2.9945578726197944e-05, + "loss": 1.0695, + "step": 16793 + }, + { + "epoch": 2.99002849002849, + "grad_norm": 0.7124375104904175, + "learning_rate": 2.993559069038805e-05, + "loss": 0.6606, + "step": 16794 + }, + { + "epoch": 2.9902065527065527, + "grad_norm": 0.8711931109428406, + "learning_rate": 2.99256040273447e-05, + "loss": 0.9049, + "step": 16795 + }, + { + "epoch": 2.9903846153846154, + "grad_norm": 0.8475961089134216, + "learning_rate": 2.9915618737263584e-05, + "loss": 0.8487, + "step": 16796 + }, + { + "epoch": 2.990562678062678, + "grad_norm": 0.8962427377700806, + "learning_rate": 2.9905634820340324e-05, + "loss": 0.9318, + "step": 16797 + }, + { + "epoch": 2.9907407407407405, + "grad_norm": 0.9120275378227234, + "learning_rate": 2.9895652276770514e-05, + "loss": 0.8787, + "step": 16798 + }, + { + "epoch": 2.9909188034188032, + "grad_norm": 0.8273578882217407, + "learning_rate": 2.9885671106749822e-05, + "loss": 0.8566, + "step": 16799 + }, + { + "epoch": 2.991096866096866, + "grad_norm": 0.8050937056541443, + "learning_rate": 2.9875691310473697e-05, + "loss": 1.1276, + "step": 16800 + }, + { + "epoch": 2.9912749287749287, + "grad_norm": 0.8434747457504272, + "learning_rate": 2.9865712888137766e-05, + "loss": 0.6286, + "step": 16801 + }, + { + "epoch": 2.9914529914529915, + "grad_norm": 0.7851191759109497, + "learning_rate": 2.9855735839937493e-05, + "loss": 0.8116, + "step": 16802 + }, + { + "epoch": 2.9916310541310542, + "grad_norm": 0.8858240842819214, + "learning_rate": 2.984576016606837e-05, + "loss": 0.8663, + "step": 16803 + }, + { + "epoch": 2.991809116809117, + "grad_norm": 0.8688439726829529, + "learning_rate": 2.9835785866725842e-05, + "loss": 0.9229, + "step": 16804 + }, + { + "epoch": 2.9919871794871797, + "grad_norm": 0.7873746752738953, + "learning_rate": 2.9825812942105302e-05, + "loss": 0.779, + "step": 16805 + }, + { + "epoch": 2.992165242165242, + "grad_norm": 0.9577028751373291, + "learning_rate": 2.9815841392402255e-05, + "loss": 0.7068, + "step": 16806 + }, + { + "epoch": 2.992343304843305, + "grad_norm": 0.8219005465507507, + "learning_rate": 2.980587121781193e-05, + "loss": 0.7712, + "step": 16807 + }, + { + "epoch": 2.9925213675213675, + "grad_norm": 0.9321123957633972, + "learning_rate": 2.9795902418529776e-05, + "loss": 0.9309, + "step": 16808 + }, + { + "epoch": 2.9926994301994303, + "grad_norm": 0.8010317087173462, + "learning_rate": 2.9785934994751074e-05, + "loss": 0.8973, + "step": 16809 + }, + { + "epoch": 2.9928774928774926, + "grad_norm": 0.8819838762283325, + "learning_rate": 2.9775968946671117e-05, + "loss": 0.6899, + "step": 16810 + }, + { + "epoch": 2.9930555555555554, + "grad_norm": 0.8337511420249939, + "learning_rate": 2.976600427448518e-05, + "loss": 0.8005, + "step": 16811 + }, + { + "epoch": 2.993233618233618, + "grad_norm": 1.0560593605041504, + "learning_rate": 2.975604097838849e-05, + "loss": 0.8804, + "step": 16812 + }, + { + "epoch": 2.993411680911681, + "grad_norm": 0.8482444882392883, + "learning_rate": 2.9746079058576247e-05, + "loss": 0.8606, + "step": 16813 + }, + { + "epoch": 2.9935897435897436, + "grad_norm": 0.8289902210235596, + "learning_rate": 2.973611851524366e-05, + "loss": 0.8163, + "step": 16814 + }, + { + "epoch": 2.9937678062678064, + "grad_norm": 0.780939519405365, + "learning_rate": 2.9726159348585826e-05, + "loss": 0.7041, + "step": 16815 + }, + { + "epoch": 2.993945868945869, + "grad_norm": 0.8907291889190674, + "learning_rate": 2.9716201558797984e-05, + "loss": 0.6484, + "step": 16816 + }, + { + "epoch": 2.994123931623932, + "grad_norm": 0.9351929426193237, + "learning_rate": 2.9706245146075118e-05, + "loss": 0.7232, + "step": 16817 + }, + { + "epoch": 2.994301994301994, + "grad_norm": 0.934117317199707, + "learning_rate": 2.969629011061238e-05, + "loss": 0.9189, + "step": 16818 + }, + { + "epoch": 2.994480056980057, + "grad_norm": 0.8751780390739441, + "learning_rate": 2.968633645260479e-05, + "loss": 0.9695, + "step": 16819 + }, + { + "epoch": 2.9946581196581197, + "grad_norm": 0.921024739742279, + "learning_rate": 2.967638417224734e-05, + "loss": 0.9011, + "step": 16820 + }, + { + "epoch": 2.9948361823361824, + "grad_norm": 0.7881745100021362, + "learning_rate": 2.9666433269735126e-05, + "loss": 0.8001, + "step": 16821 + }, + { + "epoch": 2.995014245014245, + "grad_norm": 0.8107823729515076, + "learning_rate": 2.9656483745262985e-05, + "loss": 0.9239, + "step": 16822 + }, + { + "epoch": 2.9951923076923075, + "grad_norm": 0.8491915464401245, + "learning_rate": 2.964653559902595e-05, + "loss": 0.8193, + "step": 16823 + }, + { + "epoch": 2.9953703703703702, + "grad_norm": 0.9521864056587219, + "learning_rate": 2.9636588831218904e-05, + "loss": 1.0745, + "step": 16824 + }, + { + "epoch": 2.995548433048433, + "grad_norm": 0.9784126281738281, + "learning_rate": 2.9626643442036727e-05, + "loss": 0.9899, + "step": 16825 + }, + { + "epoch": 2.9957264957264957, + "grad_norm": 0.8035039305686951, + "learning_rate": 2.961669943167429e-05, + "loss": 0.9172, + "step": 16826 + }, + { + "epoch": 2.9959045584045585, + "grad_norm": 0.9292920827865601, + "learning_rate": 2.9606756800326408e-05, + "loss": 0.9228, + "step": 16827 + }, + { + "epoch": 2.9960826210826212, + "grad_norm": 0.9243139028549194, + "learning_rate": 2.9596815548187908e-05, + "loss": 0.8698, + "step": 16828 + }, + { + "epoch": 2.996260683760684, + "grad_norm": 0.9847014546394348, + "learning_rate": 2.958687567545355e-05, + "loss": 0.8885, + "step": 16829 + }, + { + "epoch": 2.9964387464387463, + "grad_norm": 0.9667131900787354, + "learning_rate": 2.9576937182318078e-05, + "loss": 0.7951, + "step": 16830 + }, + { + "epoch": 2.996616809116809, + "grad_norm": 0.8431822061538696, + "learning_rate": 2.956700006897628e-05, + "loss": 1.0645, + "step": 16831 + }, + { + "epoch": 2.996794871794872, + "grad_norm": 0.7381225228309631, + "learning_rate": 2.955706433562274e-05, + "loss": 0.5631, + "step": 16832 + }, + { + "epoch": 2.9969729344729346, + "grad_norm": 0.7975440621376038, + "learning_rate": 2.9547129982452228e-05, + "loss": 0.8324, + "step": 16833 + }, + { + "epoch": 2.9971509971509973, + "grad_norm": 0.9004024863243103, + "learning_rate": 2.9537197009659344e-05, + "loss": 1.045, + "step": 16834 + }, + { + "epoch": 2.9973290598290596, + "grad_norm": 0.786651074886322, + "learning_rate": 2.952726541743871e-05, + "loss": 0.7569, + "step": 16835 + }, + { + "epoch": 2.9975071225071224, + "grad_norm": 0.8053869605064392, + "learning_rate": 2.951733520598492e-05, + "loss": 0.8327, + "step": 16836 + }, + { + "epoch": 2.997685185185185, + "grad_norm": 0.8725607991218567, + "learning_rate": 2.9507406375492543e-05, + "loss": 0.7702, + "step": 16837 + }, + { + "epoch": 2.997863247863248, + "grad_norm": 0.9187145233154297, + "learning_rate": 2.9497478926156087e-05, + "loss": 0.8946, + "step": 16838 + }, + { + "epoch": 2.9980413105413106, + "grad_norm": 0.9324785470962524, + "learning_rate": 2.9487552858170076e-05, + "loss": 0.8669, + "step": 16839 + }, + { + "epoch": 2.9982193732193734, + "grad_norm": 0.8439409136772156, + "learning_rate": 2.9477628171728964e-05, + "loss": 0.7187, + "step": 16840 + }, + { + "epoch": 2.998397435897436, + "grad_norm": 0.7890669703483582, + "learning_rate": 2.9467704867027258e-05, + "loss": 0.7856, + "step": 16841 + }, + { + "epoch": 2.9985754985754984, + "grad_norm": 0.7931693196296692, + "learning_rate": 2.9457782944259362e-05, + "loss": 0.7212, + "step": 16842 + }, + { + "epoch": 2.998753561253561, + "grad_norm": 0.8256354331970215, + "learning_rate": 2.9447862403619665e-05, + "loss": 0.8255, + "step": 16843 + }, + { + "epoch": 2.998931623931624, + "grad_norm": 0.7902492880821228, + "learning_rate": 2.9437943245302547e-05, + "loss": 0.8298, + "step": 16844 + }, + { + "epoch": 2.9991096866096867, + "grad_norm": 1.0743845701217651, + "learning_rate": 2.9428025469502318e-05, + "loss": 0.92, + "step": 16845 + }, + { + "epoch": 2.9992877492877494, + "grad_norm": 0.9262487292289734, + "learning_rate": 2.9418109076413404e-05, + "loss": 0.9108, + "step": 16846 + }, + { + "epoch": 2.9994658119658117, + "grad_norm": 0.855722963809967, + "learning_rate": 2.9408194066229944e-05, + "loss": 0.8596, + "step": 16847 + }, + { + "epoch": 2.9996438746438745, + "grad_norm": 0.8279969692230225, + "learning_rate": 2.939828043914634e-05, + "loss": 0.8913, + "step": 16848 + }, + { + "epoch": 2.9996438746438745, + "eval_loss": 1.1308976411819458, + "eval_runtime": 24.7916, + "eval_samples_per_second": 41.99, + "eval_steps_per_second": 21.015, + "step": 16848 + } + ], + "logging_steps": 1, + "max_steps": 22464, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 5616, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 6.50817528379392e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-16848/training_args.bin b/checkpoint-16848/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..1245f6a2afbe9a6eefbb6d141231d555e0b0bf84 --- /dev/null +++ b/checkpoint-16848/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86de370014ed2be86ea27c820b434ceec5e097da2b5f9b08d0eac9aa564d8961 +size 6200 diff --git a/checkpoint-22464/README.md b/checkpoint-22464/README.md new file mode 100644 index 0000000000000000000000000000000000000000..719b4726992f7d0707a4253e9123dec35e4de390 --- /dev/null +++ b/checkpoint-22464/README.md @@ -0,0 +1,202 @@ +--- +base_model: openlm-research/open_llama_3b_v2 +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.13.2 \ No newline at end of file diff --git a/checkpoint-22464/adapter_config.json b/checkpoint-22464/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6f20a570fc808390da3f2e001093ac1e56c1da --- /dev/null +++ b/checkpoint-22464/adapter_config.json @@ -0,0 +1,34 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "openlm-research/open_llama_3b_v2", + "bias": "none", + "fan_in_fan_out": null, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.0, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "v_proj", + "down_proj", + "o_proj", + "q_proj", + "up_proj", + "k_proj" + ], + "task_type": "CAUSAL_LM", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/checkpoint-22464/adapter_model.safetensors b/checkpoint-22464/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d0a93a30f0bf40bec09a58b6ccc07491c3553cfe --- /dev/null +++ b/checkpoint-22464/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:930ff2e666a5de6fb8ff8ade97c1215a559ff9fb63379555eedaffda38295a26 +size 50899792 diff --git a/checkpoint-22464/optimizer.pt b/checkpoint-22464/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..4cef3c803b0c82cbbc48b1f436cdf6413d841409 --- /dev/null +++ b/checkpoint-22464/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f1587d9d9345add3c41e800393eca55316e9950b5cf74334903865c353653cc +size 26231684 diff --git a/checkpoint-22464/rng_state.pth b/checkpoint-22464/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..d726c03142c02ea5fd2cae01a48760edaf4acd07 --- /dev/null +++ b/checkpoint-22464/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68c0a7bb5c807ad5bc1c4dff28401d8ba88a8d5d80ef889d2be2ca17beb56b13 +size 14244 diff --git a/checkpoint-22464/scheduler.pt b/checkpoint-22464/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..a7dc01f8cd8f408ff40f1bf6a8019e6ed38eec82 --- /dev/null +++ b/checkpoint-22464/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90ea528e0afe11e8605def0efb5c9acd5db48044014557508bbd87d9badb7309 +size 1064 diff --git a/checkpoint-22464/special_tokens_map.json b/checkpoint-22464/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/checkpoint-22464/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoint-22464/tokenizer.model b/checkpoint-22464/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..98866ff8ae3631f331c57923c921a0c9ad22b97d --- /dev/null +++ b/checkpoint-22464/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b289e85fa20fd375d8b33dc12f77616f18abc6359804471d1fafcb425fecb8 +size 511574 diff --git a/checkpoint-22464/tokenizer_config.json b/checkpoint-22464/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/checkpoint-22464/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/checkpoint-22464/trainer_state.json b/checkpoint-22464/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d921e7eefc054473118d8e98b9155471da13c5d8 --- /dev/null +++ b/checkpoint-22464/trainer_state.json @@ -0,0 +1,157417 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.9994658119658117, + "eval_steps": 1404, + "global_step": 22464, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00017806267806267807, + "grad_norm": 0.2854898273944855, + "learning_rate": 1e-05, + "loss": 1.1997, + "step": 1 + }, + { + "epoch": 0.00017806267806267807, + "eval_loss": 1.3698358535766602, + "eval_runtime": 24.1591, + "eval_samples_per_second": 43.089, + "eval_steps_per_second": 21.565, + "step": 1 + }, + { + "epoch": 0.00035612535612535614, + "grad_norm": 0.3508087396621704, + "learning_rate": 2e-05, + "loss": 1.4134, + "step": 2 + }, + { + "epoch": 0.0005341880341880342, + "grad_norm": 0.27050870656967163, + "learning_rate": 3e-05, + "loss": 1.3447, + "step": 3 + }, + { + "epoch": 0.0007122507122507123, + "grad_norm": 0.27706292271614075, + "learning_rate": 4e-05, + "loss": 1.0354, + "step": 4 + }, + { + "epoch": 0.0008903133903133903, + "grad_norm": 0.30398961901664734, + "learning_rate": 5e-05, + "loss": 1.1441, + "step": 5 + }, + { + "epoch": 0.0010683760683760685, + "grad_norm": 0.3103881776332855, + "learning_rate": 6e-05, + "loss": 1.341, + "step": 6 + }, + { + "epoch": 0.0012464387464387464, + "grad_norm": 0.5191189646720886, + "learning_rate": 7e-05, + "loss": 1.3457, + "step": 7 + }, + { + "epoch": 0.0014245014245014246, + "grad_norm": 0.4449467360973358, + "learning_rate": 8e-05, + "loss": 1.5051, + "step": 8 + }, + { + "epoch": 0.0016025641025641025, + "grad_norm": 0.3914581537246704, + "learning_rate": 9e-05, + "loss": 1.5525, + "step": 9 + }, + { + "epoch": 0.0017806267806267807, + "grad_norm": 0.37746086716651917, + "learning_rate": 0.0001, + "loss": 1.3266, + "step": 10 + }, + { + "epoch": 0.001958689458689459, + "grad_norm": 0.35226109623908997, + "learning_rate": 0.00011000000000000002, + "loss": 1.5416, + "step": 11 + }, + { + "epoch": 0.002136752136752137, + "grad_norm": 0.3343672454357147, + "learning_rate": 0.00012, + "loss": 1.3221, + "step": 12 + }, + { + "epoch": 0.0023148148148148147, + "grad_norm": 0.47298333048820496, + "learning_rate": 0.00013000000000000002, + "loss": 1.2999, + "step": 13 + }, + { + "epoch": 0.002492877492877493, + "grad_norm": 0.377814918756485, + "learning_rate": 0.00014, + "loss": 1.1688, + "step": 14 + }, + { + "epoch": 0.002670940170940171, + "grad_norm": 0.46344801783561707, + "learning_rate": 0.00015000000000000001, + "loss": 1.3565, + "step": 15 + }, + { + "epoch": 0.002849002849002849, + "grad_norm": 0.49615249037742615, + "learning_rate": 0.00016, + "loss": 1.5692, + "step": 16 + }, + { + "epoch": 0.003027065527065527, + "grad_norm": 0.5109946131706238, + "learning_rate": 0.00017, + "loss": 1.2991, + "step": 17 + }, + { + "epoch": 0.003205128205128205, + "grad_norm": 0.5125070214271545, + "learning_rate": 0.00018, + "loss": 1.3309, + "step": 18 + }, + { + "epoch": 0.003383190883190883, + "grad_norm": 0.4517767131328583, + "learning_rate": 0.00019, + "loss": 1.357, + "step": 19 + }, + { + "epoch": 0.0035612535612535613, + "grad_norm": 0.47267794609069824, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 20 + }, + { + "epoch": 0.0037393162393162395, + "grad_norm": 0.46823424100875854, + "learning_rate": 0.00019999999902035388, + "loss": 1.1195, + "step": 21 + }, + { + "epoch": 0.003917378917378918, + "grad_norm": 0.440036803483963, + "learning_rate": 0.00019999999608141548, + "loss": 1.2822, + "step": 22 + }, + { + "epoch": 0.004095441595441595, + "grad_norm": 0.371101975440979, + "learning_rate": 0.00019999999118318492, + "loss": 1.132, + "step": 23 + }, + { + "epoch": 0.004273504273504274, + "grad_norm": 0.44691094756126404, + "learning_rate": 0.00019999998432566226, + "loss": 1.2968, + "step": 24 + }, + { + "epoch": 0.004451566951566952, + "grad_norm": 0.5462725162506104, + "learning_rate": 0.0001999999755088476, + "loss": 1.1714, + "step": 25 + }, + { + "epoch": 0.004629629629629629, + "grad_norm": 0.39860013127326965, + "learning_rate": 0.0001999999647327412, + "loss": 1.0407, + "step": 26 + }, + { + "epoch": 0.004807692307692308, + "grad_norm": 0.5031934380531311, + "learning_rate": 0.0001999999519973432, + "loss": 1.2773, + "step": 27 + }, + { + "epoch": 0.004985754985754986, + "grad_norm": 0.42162764072418213, + "learning_rate": 0.0001999999373026539, + "loss": 1.2824, + "step": 28 + }, + { + "epoch": 0.005163817663817663, + "grad_norm": 0.40964868664741516, + "learning_rate": 0.00019999992064867353, + "loss": 1.226, + "step": 29 + }, + { + "epoch": 0.005341880341880342, + "grad_norm": 0.41650915145874023, + "learning_rate": 0.00019999990203540245, + "loss": 1.2677, + "step": 30 + }, + { + "epoch": 0.00551994301994302, + "grad_norm": 0.40052226185798645, + "learning_rate": 0.00019999988146284103, + "loss": 0.9443, + "step": 31 + }, + { + "epoch": 0.005698005698005698, + "grad_norm": 0.5198387503623962, + "learning_rate": 0.00019999985893098964, + "loss": 1.3043, + "step": 32 + }, + { + "epoch": 0.005876068376068376, + "grad_norm": 0.50941002368927, + "learning_rate": 0.00019999983443984878, + "loss": 1.2002, + "step": 33 + }, + { + "epoch": 0.006054131054131054, + "grad_norm": 0.30082932114601135, + "learning_rate": 0.00019999980798941888, + "loss": 0.9904, + "step": 34 + }, + { + "epoch": 0.006232193732193732, + "grad_norm": 0.4228935241699219, + "learning_rate": 0.00019999977957970048, + "loss": 1.1137, + "step": 35 + }, + { + "epoch": 0.00641025641025641, + "grad_norm": 0.41294750571250916, + "learning_rate": 0.0001999997492106941, + "loss": 1.3385, + "step": 36 + }, + { + "epoch": 0.006588319088319089, + "grad_norm": 0.4415493905544281, + "learning_rate": 0.00019999971688240041, + "loss": 1.1695, + "step": 37 + }, + { + "epoch": 0.006766381766381766, + "grad_norm": 0.3726460933685303, + "learning_rate": 0.00019999968259482, + "loss": 1.1734, + "step": 38 + }, + { + "epoch": 0.006944444444444444, + "grad_norm": 0.3969627320766449, + "learning_rate": 0.0001999996463479535, + "loss": 1.1209, + "step": 39 + }, + { + "epoch": 0.007122507122507123, + "grad_norm": 0.3779667913913727, + "learning_rate": 0.0001999996081418017, + "loss": 1.1635, + "step": 40 + }, + { + "epoch": 0.0073005698005698, + "grad_norm": 0.3933636546134949, + "learning_rate": 0.0001999995679763653, + "loss": 1.1514, + "step": 41 + }, + { + "epoch": 0.007478632478632479, + "grad_norm": 0.3567957282066345, + "learning_rate": 0.00019999952585164507, + "loss": 1.2488, + "step": 42 + }, + { + "epoch": 0.007656695156695157, + "grad_norm": 0.32506081461906433, + "learning_rate": 0.00019999948176764186, + "loss": 1.149, + "step": 43 + }, + { + "epoch": 0.007834757834757835, + "grad_norm": 0.46588361263275146, + "learning_rate": 0.0001999994357243566, + "loss": 1.4263, + "step": 44 + }, + { + "epoch": 0.008012820512820512, + "grad_norm": 0.5070307850837708, + "learning_rate": 0.00019999938772179005, + "loss": 1.0698, + "step": 45 + }, + { + "epoch": 0.00819088319088319, + "grad_norm": 0.38199326395988464, + "learning_rate": 0.00019999933775994327, + "loss": 0.9907, + "step": 46 + }, + { + "epoch": 0.00836894586894587, + "grad_norm": 0.43684661388397217, + "learning_rate": 0.0001999992858388172, + "loss": 1.2905, + "step": 47 + }, + { + "epoch": 0.008547008547008548, + "grad_norm": 0.44482162594795227, + "learning_rate": 0.00019999923195841284, + "loss": 1.2153, + "step": 48 + }, + { + "epoch": 0.008725071225071225, + "grad_norm": 0.4259667694568634, + "learning_rate": 0.0001999991761187313, + "loss": 1.1582, + "step": 49 + }, + { + "epoch": 0.008903133903133903, + "grad_norm": 0.41649091243743896, + "learning_rate": 0.00019999911831977357, + "loss": 1.0185, + "step": 50 + }, + { + "epoch": 0.009081196581196582, + "grad_norm": 0.4179716110229492, + "learning_rate": 0.0001999990585615409, + "loss": 1.3579, + "step": 51 + }, + { + "epoch": 0.009259259259259259, + "grad_norm": 0.3372558355331421, + "learning_rate": 0.00019999899684403438, + "loss": 1.0638, + "step": 52 + }, + { + "epoch": 0.009437321937321937, + "grad_norm": 0.41294020414352417, + "learning_rate": 0.00019999893316725525, + "loss": 1.1932, + "step": 53 + }, + { + "epoch": 0.009615384615384616, + "grad_norm": 0.4407919645309448, + "learning_rate": 0.00019999886753120473, + "loss": 1.4129, + "step": 54 + }, + { + "epoch": 0.009793447293447293, + "grad_norm": 0.47948843240737915, + "learning_rate": 0.00019999879993588414, + "loss": 1.2424, + "step": 55 + }, + { + "epoch": 0.009971509971509971, + "grad_norm": 0.3535355031490326, + "learning_rate": 0.00019999873038129484, + "loss": 1.0145, + "step": 56 + }, + { + "epoch": 0.01014957264957265, + "grad_norm": 0.5067078471183777, + "learning_rate": 0.00019999865886743813, + "loss": 1.4708, + "step": 57 + }, + { + "epoch": 0.010327635327635327, + "grad_norm": 0.42862898111343384, + "learning_rate": 0.0001999985853943154, + "loss": 1.0399, + "step": 58 + }, + { + "epoch": 0.010505698005698005, + "grad_norm": 0.4769059419631958, + "learning_rate": 0.00019999850996192816, + "loss": 1.1258, + "step": 59 + }, + { + "epoch": 0.010683760683760684, + "grad_norm": 0.4065442383289337, + "learning_rate": 0.0001999984325702778, + "loss": 1.2077, + "step": 60 + }, + { + "epoch": 0.010861823361823363, + "grad_norm": 0.5318329930305481, + "learning_rate": 0.0001999983532193659, + "loss": 1.2298, + "step": 61 + }, + { + "epoch": 0.01103988603988604, + "grad_norm": 0.4777173101902008, + "learning_rate": 0.000199998271909194, + "loss": 1.3195, + "step": 62 + }, + { + "epoch": 0.011217948717948718, + "grad_norm": 0.37553808093070984, + "learning_rate": 0.0001999981886397637, + "loss": 1.1188, + "step": 63 + }, + { + "epoch": 0.011396011396011397, + "grad_norm": 0.3920556902885437, + "learning_rate": 0.0001999981034110766, + "loss": 1.1448, + "step": 64 + }, + { + "epoch": 0.011574074074074073, + "grad_norm": 0.454272598028183, + "learning_rate": 0.0001999980162231344, + "loss": 1.0812, + "step": 65 + }, + { + "epoch": 0.011752136752136752, + "grad_norm": 0.4354456663131714, + "learning_rate": 0.00019999792707593882, + "loss": 1.1174, + "step": 66 + }, + { + "epoch": 0.01193019943019943, + "grad_norm": 0.5030252933502197, + "learning_rate": 0.00019999783596949156, + "loss": 1.2925, + "step": 67 + }, + { + "epoch": 0.012108262108262107, + "grad_norm": 0.5141571164131165, + "learning_rate": 0.00019999774290379446, + "loss": 1.6193, + "step": 68 + }, + { + "epoch": 0.012286324786324786, + "grad_norm": 0.417298287153244, + "learning_rate": 0.0001999976478788493, + "loss": 1.1875, + "step": 69 + }, + { + "epoch": 0.012464387464387465, + "grad_norm": 0.4642415940761566, + "learning_rate": 0.00019999755089465795, + "loss": 1.4138, + "step": 70 + }, + { + "epoch": 0.012642450142450143, + "grad_norm": 0.43184754252433777, + "learning_rate": 0.0001999974519512223, + "loss": 1.0697, + "step": 71 + }, + { + "epoch": 0.01282051282051282, + "grad_norm": 0.46698349714279175, + "learning_rate": 0.00019999735104854436, + "loss": 0.709, + "step": 72 + }, + { + "epoch": 0.012998575498575499, + "grad_norm": 0.37253814935684204, + "learning_rate": 0.000199997248186626, + "loss": 1.2084, + "step": 73 + }, + { + "epoch": 0.013176638176638177, + "grad_norm": 0.3851388692855835, + "learning_rate": 0.0001999971433654693, + "loss": 1.0548, + "step": 74 + }, + { + "epoch": 0.013354700854700854, + "grad_norm": 0.4434688985347748, + "learning_rate": 0.00019999703658507635, + "loss": 1.4084, + "step": 75 + }, + { + "epoch": 0.013532763532763533, + "grad_norm": 0.43164482712745667, + "learning_rate": 0.00019999692784544913, + "loss": 1.4872, + "step": 76 + }, + { + "epoch": 0.013710826210826211, + "grad_norm": 0.4224303364753723, + "learning_rate": 0.00019999681714658984, + "loss": 1.2221, + "step": 77 + }, + { + "epoch": 0.013888888888888888, + "grad_norm": 0.35588955879211426, + "learning_rate": 0.00019999670448850069, + "loss": 0.84, + "step": 78 + }, + { + "epoch": 0.014066951566951567, + "grad_norm": 0.3970590829849243, + "learning_rate": 0.0001999965898711838, + "loss": 1.1886, + "step": 79 + }, + { + "epoch": 0.014245014245014245, + "grad_norm": 0.4331924319267273, + "learning_rate": 0.00019999647329464146, + "loss": 1.179, + "step": 80 + }, + { + "epoch": 0.014423076923076924, + "grad_norm": 0.4226946234703064, + "learning_rate": 0.00019999635475887598, + "loss": 1.1496, + "step": 81 + }, + { + "epoch": 0.0146011396011396, + "grad_norm": 0.381592720746994, + "learning_rate": 0.00019999623426388962, + "loss": 1.1774, + "step": 82 + }, + { + "epoch": 0.01477920227920228, + "grad_norm": 0.4190855622291565, + "learning_rate": 0.00019999611180968478, + "loss": 1.1491, + "step": 83 + }, + { + "epoch": 0.014957264957264958, + "grad_norm": 0.3904292583465576, + "learning_rate": 0.00019999598739626389, + "loss": 1.1275, + "step": 84 + }, + { + "epoch": 0.015135327635327635, + "grad_norm": 0.4515478014945984, + "learning_rate": 0.0001999958610236293, + "loss": 1.2404, + "step": 85 + }, + { + "epoch": 0.015313390313390313, + "grad_norm": 0.48341724276542664, + "learning_rate": 0.00019999573269178359, + "loss": 1.3572, + "step": 86 + }, + { + "epoch": 0.015491452991452992, + "grad_norm": 0.42150333523750305, + "learning_rate": 0.00019999560240072914, + "loss": 1.0203, + "step": 87 + }, + { + "epoch": 0.01566951566951567, + "grad_norm": 0.45445525646209717, + "learning_rate": 0.00019999547015046867, + "loss": 1.0677, + "step": 88 + }, + { + "epoch": 0.01584757834757835, + "grad_norm": 0.3581015467643738, + "learning_rate": 0.00019999533594100463, + "loss": 1.0693, + "step": 89 + }, + { + "epoch": 0.016025641025641024, + "grad_norm": 0.4430878758430481, + "learning_rate": 0.00019999519977233971, + "loss": 1.1591, + "step": 90 + }, + { + "epoch": 0.016203703703703703, + "grad_norm": 0.3940352201461792, + "learning_rate": 0.0001999950616444766, + "loss": 1.1325, + "step": 91 + }, + { + "epoch": 0.01638176638176638, + "grad_norm": 0.4521673321723938, + "learning_rate": 0.00019999492155741794, + "loss": 1.3288, + "step": 92 + }, + { + "epoch": 0.01655982905982906, + "grad_norm": 0.3988296687602997, + "learning_rate": 0.00019999477951116658, + "loss": 1.0023, + "step": 93 + }, + { + "epoch": 0.01673789173789174, + "grad_norm": 0.38709723949432373, + "learning_rate": 0.00019999463550572516, + "loss": 1.2623, + "step": 94 + }, + { + "epoch": 0.016915954415954417, + "grad_norm": 0.35376182198524475, + "learning_rate": 0.00019999448954109662, + "loss": 1.0643, + "step": 95 + }, + { + "epoch": 0.017094017094017096, + "grad_norm": 0.49547120928764343, + "learning_rate": 0.00019999434161728377, + "loss": 1.2121, + "step": 96 + }, + { + "epoch": 0.01727207977207977, + "grad_norm": 0.49593672156333923, + "learning_rate": 0.00019999419173428952, + "loss": 1.1635, + "step": 97 + }, + { + "epoch": 0.01745014245014245, + "grad_norm": 0.4146541953086853, + "learning_rate": 0.0001999940398921168, + "loss": 1.1452, + "step": 98 + }, + { + "epoch": 0.017628205128205128, + "grad_norm": 0.5177254676818848, + "learning_rate": 0.00019999388609076858, + "loss": 1.2178, + "step": 99 + }, + { + "epoch": 0.017806267806267807, + "grad_norm": 0.4012768864631653, + "learning_rate": 0.0001999937303302479, + "loss": 0.9222, + "step": 100 + }, + { + "epoch": 0.017984330484330485, + "grad_norm": 0.4597131907939911, + "learning_rate": 0.00019999357261055777, + "loss": 0.979, + "step": 101 + }, + { + "epoch": 0.018162393162393164, + "grad_norm": 0.6190966963768005, + "learning_rate": 0.00019999341293170132, + "loss": 1.3909, + "step": 102 + }, + { + "epoch": 0.01834045584045584, + "grad_norm": 0.4576462209224701, + "learning_rate": 0.00019999325129368164, + "loss": 1.073, + "step": 103 + }, + { + "epoch": 0.018518518518518517, + "grad_norm": 0.4036749005317688, + "learning_rate": 0.00019999308769650192, + "loss": 1.1354, + "step": 104 + }, + { + "epoch": 0.018696581196581196, + "grad_norm": 0.4722452759742737, + "learning_rate": 0.00019999292214016538, + "loss": 1.2039, + "step": 105 + }, + { + "epoch": 0.018874643874643875, + "grad_norm": 0.5338274240493774, + "learning_rate": 0.00019999275462467527, + "loss": 1.225, + "step": 106 + }, + { + "epoch": 0.019052706552706553, + "grad_norm": 0.4301491677761078, + "learning_rate": 0.00019999258515003484, + "loss": 1.0601, + "step": 107 + }, + { + "epoch": 0.019230769230769232, + "grad_norm": 0.33271175622940063, + "learning_rate": 0.0001999924137162474, + "loss": 0.8441, + "step": 108 + }, + { + "epoch": 0.01940883190883191, + "grad_norm": 0.4648784399032593, + "learning_rate": 0.0001999922403233163, + "loss": 1.2038, + "step": 109 + }, + { + "epoch": 0.019586894586894586, + "grad_norm": 0.37915176153182983, + "learning_rate": 0.00019999206497124504, + "loss": 1.0923, + "step": 110 + }, + { + "epoch": 0.019764957264957264, + "grad_norm": 0.3865506052970886, + "learning_rate": 0.00019999188766003695, + "loss": 0.9535, + "step": 111 + }, + { + "epoch": 0.019943019943019943, + "grad_norm": 0.35739636421203613, + "learning_rate": 0.0001999917083896955, + "loss": 1.2688, + "step": 112 + }, + { + "epoch": 0.02012108262108262, + "grad_norm": 0.3943796157836914, + "learning_rate": 0.0001999915271602243, + "loss": 1.1097, + "step": 113 + }, + { + "epoch": 0.0202991452991453, + "grad_norm": 0.44758161902427673, + "learning_rate": 0.0001999913439716268, + "loss": 1.2698, + "step": 114 + }, + { + "epoch": 0.02047720797720798, + "grad_norm": 0.3749747574329376, + "learning_rate": 0.00019999115882390664, + "loss": 1.1091, + "step": 115 + }, + { + "epoch": 0.020655270655270654, + "grad_norm": 0.3479487895965576, + "learning_rate": 0.00019999097171706745, + "loss": 1.0049, + "step": 116 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 0.4491243064403534, + "learning_rate": 0.00019999078265111285, + "loss": 1.1857, + "step": 117 + }, + { + "epoch": 0.02101139601139601, + "grad_norm": 0.345289021730423, + "learning_rate": 0.00019999059162604662, + "loss": 1.1397, + "step": 118 + }, + { + "epoch": 0.02118945868945869, + "grad_norm": 0.5467649698257446, + "learning_rate": 0.00019999039864187243, + "loss": 1.2196, + "step": 119 + }, + { + "epoch": 0.021367521367521368, + "grad_norm": 0.36446481943130493, + "learning_rate": 0.00019999020369859409, + "loss": 0.796, + "step": 120 + }, + { + "epoch": 0.021545584045584047, + "grad_norm": 0.4225841760635376, + "learning_rate": 0.00019999000679621543, + "loss": 0.9684, + "step": 121 + }, + { + "epoch": 0.021723646723646725, + "grad_norm": 0.4205594062805176, + "learning_rate": 0.0001999898079347403, + "loss": 1.2762, + "step": 122 + }, + { + "epoch": 0.0219017094017094, + "grad_norm": 0.43773892521858215, + "learning_rate": 0.00019998960711417257, + "loss": 1.117, + "step": 123 + }, + { + "epoch": 0.02207977207977208, + "grad_norm": 0.41279685497283936, + "learning_rate": 0.00019998940433451623, + "loss": 1.1502, + "step": 124 + }, + { + "epoch": 0.022257834757834757, + "grad_norm": 0.4090803563594818, + "learning_rate": 0.0001999891995957752, + "loss": 1.2591, + "step": 125 + }, + { + "epoch": 0.022435897435897436, + "grad_norm": 0.6000410914421082, + "learning_rate": 0.0001999889928979535, + "loss": 1.4321, + "step": 126 + }, + { + "epoch": 0.022613960113960115, + "grad_norm": 0.524264395236969, + "learning_rate": 0.00019998878424105524, + "loss": 1.1849, + "step": 127 + }, + { + "epoch": 0.022792022792022793, + "grad_norm": 0.4581047296524048, + "learning_rate": 0.00019998857362508443, + "loss": 1.0598, + "step": 128 + }, + { + "epoch": 0.022970085470085472, + "grad_norm": 0.42663446068763733, + "learning_rate": 0.00019998836105004526, + "loss": 1.1909, + "step": 129 + }, + { + "epoch": 0.023148148148148147, + "grad_norm": 0.45709118247032166, + "learning_rate": 0.00019998814651594183, + "loss": 1.2104, + "step": 130 + }, + { + "epoch": 0.023326210826210825, + "grad_norm": 0.39528369903564453, + "learning_rate": 0.0001999879300227784, + "loss": 1.3073, + "step": 131 + }, + { + "epoch": 0.023504273504273504, + "grad_norm": 0.46896448731422424, + "learning_rate": 0.00019998771157055914, + "loss": 1.3202, + "step": 132 + }, + { + "epoch": 0.023682336182336183, + "grad_norm": 0.4386129677295685, + "learning_rate": 0.00019998749115928842, + "loss": 1.2196, + "step": 133 + }, + { + "epoch": 0.02386039886039886, + "grad_norm": 0.45920488238334656, + "learning_rate": 0.00019998726878897051, + "loss": 1.3668, + "step": 134 + }, + { + "epoch": 0.02403846153846154, + "grad_norm": 0.4115797281265259, + "learning_rate": 0.0001999870444596098, + "loss": 1.1052, + "step": 135 + }, + { + "epoch": 0.024216524216524215, + "grad_norm": 0.3860839903354645, + "learning_rate": 0.0001999868181712106, + "loss": 1.0344, + "step": 136 + }, + { + "epoch": 0.024394586894586893, + "grad_norm": 0.42514732480049133, + "learning_rate": 0.00019998658992377742, + "loss": 1.1979, + "step": 137 + }, + { + "epoch": 0.024572649572649572, + "grad_norm": 0.36001840233802795, + "learning_rate": 0.00019998635971731475, + "loss": 1.4536, + "step": 138 + }, + { + "epoch": 0.02475071225071225, + "grad_norm": 0.3739112317562103, + "learning_rate": 0.00019998612755182707, + "loss": 1.0097, + "step": 139 + }, + { + "epoch": 0.02492877492877493, + "grad_norm": 0.37545472383499146, + "learning_rate": 0.00019998589342731888, + "loss": 0.829, + "step": 140 + }, + { + "epoch": 0.025106837606837608, + "grad_norm": 0.38660728931427, + "learning_rate": 0.0001999856573437948, + "loss": 1.1324, + "step": 141 + }, + { + "epoch": 0.025284900284900286, + "grad_norm": 0.3741356432437897, + "learning_rate": 0.00019998541930125953, + "loss": 1.0934, + "step": 142 + }, + { + "epoch": 0.02546296296296296, + "grad_norm": 0.41900336742401123, + "learning_rate": 0.00019998517929971764, + "loss": 1.0336, + "step": 143 + }, + { + "epoch": 0.02564102564102564, + "grad_norm": 0.4167572259902954, + "learning_rate": 0.00019998493733917384, + "loss": 1.2571, + "step": 144 + }, + { + "epoch": 0.02581908831908832, + "grad_norm": 0.39437636733055115, + "learning_rate": 0.0001999846934196329, + "loss": 1.2283, + "step": 145 + }, + { + "epoch": 0.025997150997150997, + "grad_norm": 0.39129480719566345, + "learning_rate": 0.00019998444754109964, + "loss": 0.9893, + "step": 146 + }, + { + "epoch": 0.026175213675213676, + "grad_norm": 0.45533549785614014, + "learning_rate": 0.0001999841997035788, + "loss": 1.0793, + "step": 147 + }, + { + "epoch": 0.026353276353276354, + "grad_norm": 0.3741768002510071, + "learning_rate": 0.00019998394990707524, + "loss": 1.2179, + "step": 148 + }, + { + "epoch": 0.026531339031339033, + "grad_norm": 0.4066533148288727, + "learning_rate": 0.0001999836981515939, + "loss": 1.1443, + "step": 149 + }, + { + "epoch": 0.026709401709401708, + "grad_norm": 0.4851688742637634, + "learning_rate": 0.0001999834444371397, + "loss": 1.1668, + "step": 150 + }, + { + "epoch": 0.026887464387464387, + "grad_norm": 0.428091436624527, + "learning_rate": 0.0001999831887637176, + "loss": 1.2676, + "step": 151 + }, + { + "epoch": 0.027065527065527065, + "grad_norm": 0.4024655222892761, + "learning_rate": 0.0001999829311313326, + "loss": 1.3115, + "step": 152 + }, + { + "epoch": 0.027243589743589744, + "grad_norm": 0.43983033299446106, + "learning_rate": 0.00019998267153998976, + "loss": 1.1019, + "step": 153 + }, + { + "epoch": 0.027421652421652423, + "grad_norm": 0.4317505359649658, + "learning_rate": 0.0001999824099896942, + "loss": 1.3129, + "step": 154 + }, + { + "epoch": 0.0275997150997151, + "grad_norm": 0.43107882142066956, + "learning_rate": 0.000199982146480451, + "loss": 1.2134, + "step": 155 + }, + { + "epoch": 0.027777777777777776, + "grad_norm": 0.3939448297023773, + "learning_rate": 0.00019998188101226532, + "loss": 1.0321, + "step": 156 + }, + { + "epoch": 0.027955840455840455, + "grad_norm": 0.4641847610473633, + "learning_rate": 0.00019998161358514237, + "loss": 1.2369, + "step": 157 + }, + { + "epoch": 0.028133903133903133, + "grad_norm": 0.3538529872894287, + "learning_rate": 0.0001999813441990874, + "loss": 1.2061, + "step": 158 + }, + { + "epoch": 0.028311965811965812, + "grad_norm": 0.3277950584888458, + "learning_rate": 0.0001999810728541057, + "loss": 0.9419, + "step": 159 + }, + { + "epoch": 0.02849002849002849, + "grad_norm": 0.424710750579834, + "learning_rate": 0.00019998079955020254, + "loss": 1.3302, + "step": 160 + }, + { + "epoch": 0.02866809116809117, + "grad_norm": 0.4120834469795227, + "learning_rate": 0.00019998052428738333, + "loss": 1.079, + "step": 161 + }, + { + "epoch": 0.028846153846153848, + "grad_norm": 0.45811930298805237, + "learning_rate": 0.00019998024706565346, + "loss": 1.1259, + "step": 162 + }, + { + "epoch": 0.029024216524216523, + "grad_norm": 0.3873266875743866, + "learning_rate": 0.0001999799678850183, + "loss": 1.2124, + "step": 163 + }, + { + "epoch": 0.0292022792022792, + "grad_norm": 0.5806412696838379, + "learning_rate": 0.00019997968674548337, + "loss": 1.3467, + "step": 164 + }, + { + "epoch": 0.02938034188034188, + "grad_norm": 0.3906802833080292, + "learning_rate": 0.00019997940364705418, + "loss": 1.1438, + "step": 165 + }, + { + "epoch": 0.02955840455840456, + "grad_norm": 0.45201995968818665, + "learning_rate": 0.00019997911858973626, + "loss": 1.1469, + "step": 166 + }, + { + "epoch": 0.029736467236467237, + "grad_norm": 0.4965892732143402, + "learning_rate": 0.0001999788315735352, + "loss": 1.0829, + "step": 167 + }, + { + "epoch": 0.029914529914529916, + "grad_norm": 0.32578057050704956, + "learning_rate": 0.0001999785425984566, + "loss": 1.0432, + "step": 168 + }, + { + "epoch": 0.03009259259259259, + "grad_norm": 0.4146028161048889, + "learning_rate": 0.00019997825166450617, + "loss": 1.1657, + "step": 169 + }, + { + "epoch": 0.03027065527065527, + "grad_norm": 0.4342964291572571, + "learning_rate": 0.0001999779587716896, + "loss": 1.2038, + "step": 170 + }, + { + "epoch": 0.030448717948717948, + "grad_norm": 0.40128546953201294, + "learning_rate": 0.00019997766392001258, + "loss": 1.3044, + "step": 171 + }, + { + "epoch": 0.030626780626780627, + "grad_norm": 0.4357539117336273, + "learning_rate": 0.00019997736710948094, + "loss": 1.2143, + "step": 172 + }, + { + "epoch": 0.030804843304843305, + "grad_norm": 0.4821035861968994, + "learning_rate": 0.00019997706834010045, + "loss": 1.0469, + "step": 173 + }, + { + "epoch": 0.030982905982905984, + "grad_norm": 0.3966675102710724, + "learning_rate": 0.000199976767611877, + "loss": 1.2122, + "step": 174 + }, + { + "epoch": 0.031160968660968662, + "grad_norm": 0.4265064299106598, + "learning_rate": 0.00019997646492481648, + "loss": 1.0871, + "step": 175 + }, + { + "epoch": 0.03133903133903134, + "grad_norm": 0.3445652723312378, + "learning_rate": 0.00019997616027892485, + "loss": 1.0412, + "step": 176 + }, + { + "epoch": 0.031517094017094016, + "grad_norm": 0.47187718749046326, + "learning_rate": 0.000199975853674208, + "loss": 1.0822, + "step": 177 + }, + { + "epoch": 0.0316951566951567, + "grad_norm": 0.37751707434654236, + "learning_rate": 0.000199975545110672, + "loss": 1.1439, + "step": 178 + }, + { + "epoch": 0.03187321937321937, + "grad_norm": 0.38792455196380615, + "learning_rate": 0.00019997523458832286, + "loss": 0.8604, + "step": 179 + }, + { + "epoch": 0.03205128205128205, + "grad_norm": 0.35199594497680664, + "learning_rate": 0.00019997492210716667, + "loss": 1.0819, + "step": 180 + }, + { + "epoch": 0.03222934472934473, + "grad_norm": 0.4828922748565674, + "learning_rate": 0.00019997460766720958, + "loss": 1.1879, + "step": 181 + }, + { + "epoch": 0.032407407407407406, + "grad_norm": 0.46153363585472107, + "learning_rate": 0.00019997429126845774, + "loss": 1.1592, + "step": 182 + }, + { + "epoch": 0.03258547008547009, + "grad_norm": 0.4844890832901001, + "learning_rate": 0.0001999739729109173, + "loss": 1.1334, + "step": 183 + }, + { + "epoch": 0.03276353276353276, + "grad_norm": 0.414617121219635, + "learning_rate": 0.00019997365259459457, + "loss": 1.0547, + "step": 184 + }, + { + "epoch": 0.032941595441595445, + "grad_norm": 0.46544626355171204, + "learning_rate": 0.00019997333031949581, + "loss": 1.4067, + "step": 185 + }, + { + "epoch": 0.03311965811965812, + "grad_norm": 0.48489415645599365, + "learning_rate": 0.0001999730060856273, + "loss": 1.4027, + "step": 186 + }, + { + "epoch": 0.033297720797720795, + "grad_norm": 0.3963346481323242, + "learning_rate": 0.0001999726798929954, + "loss": 1.1327, + "step": 187 + }, + { + "epoch": 0.03347578347578348, + "grad_norm": 0.3809385895729065, + "learning_rate": 0.00019997235174160652, + "loss": 1.3475, + "step": 188 + }, + { + "epoch": 0.03365384615384615, + "grad_norm": 0.3866960406303406, + "learning_rate": 0.0001999720216314671, + "loss": 1.1576, + "step": 189 + }, + { + "epoch": 0.033831908831908834, + "grad_norm": 0.34976935386657715, + "learning_rate": 0.00019997168956258356, + "loss": 0.9361, + "step": 190 + }, + { + "epoch": 0.03400997150997151, + "grad_norm": 0.38681939244270325, + "learning_rate": 0.00019997135553496243, + "loss": 1.1796, + "step": 191 + }, + { + "epoch": 0.03418803418803419, + "grad_norm": 0.41905197501182556, + "learning_rate": 0.0001999710195486103, + "loss": 1.1714, + "step": 192 + }, + { + "epoch": 0.03436609686609687, + "grad_norm": 0.42356589436531067, + "learning_rate": 0.0001999706816035337, + "loss": 1.0022, + "step": 193 + }, + { + "epoch": 0.03454415954415954, + "grad_norm": 0.3929740786552429, + "learning_rate": 0.00019997034169973925, + "loss": 1.3769, + "step": 194 + }, + { + "epoch": 0.034722222222222224, + "grad_norm": 0.4325186312198639, + "learning_rate": 0.00019996999983723366, + "loss": 1.3057, + "step": 195 + }, + { + "epoch": 0.0349002849002849, + "grad_norm": 0.3954029381275177, + "learning_rate": 0.00019996965601602355, + "loss": 1.1958, + "step": 196 + }, + { + "epoch": 0.03507834757834758, + "grad_norm": 0.34454262256622314, + "learning_rate": 0.00019996931023611572, + "loss": 1.0972, + "step": 197 + }, + { + "epoch": 0.035256410256410256, + "grad_norm": 0.48900291323661804, + "learning_rate": 0.0001999689624975169, + "loss": 1.213, + "step": 198 + }, + { + "epoch": 0.03543447293447293, + "grad_norm": 0.35214388370513916, + "learning_rate": 0.00019996861280023397, + "loss": 1.0285, + "step": 199 + }, + { + "epoch": 0.03561253561253561, + "grad_norm": 0.49393126368522644, + "learning_rate": 0.00019996826114427373, + "loss": 1.2313, + "step": 200 + }, + { + "epoch": 0.03579059829059829, + "grad_norm": 0.3994458019733429, + "learning_rate": 0.00019996790752964305, + "loss": 1.0474, + "step": 201 + }, + { + "epoch": 0.03596866096866097, + "grad_norm": 0.5387318730354309, + "learning_rate": 0.0001999675519563489, + "loss": 1.3067, + "step": 202 + }, + { + "epoch": 0.036146723646723646, + "grad_norm": 0.4976751208305359, + "learning_rate": 0.00019996719442439824, + "loss": 1.2593, + "step": 203 + }, + { + "epoch": 0.03632478632478633, + "grad_norm": 0.47052907943725586, + "learning_rate": 0.0001999668349337981, + "loss": 1.1036, + "step": 204 + }, + { + "epoch": 0.036502849002849, + "grad_norm": 0.39616644382476807, + "learning_rate": 0.00019996647348455543, + "loss": 1.0481, + "step": 205 + }, + { + "epoch": 0.03668091168091168, + "grad_norm": 0.42987677454948425, + "learning_rate": 0.00019996611007667742, + "loss": 1.0923, + "step": 206 + }, + { + "epoch": 0.03685897435897436, + "grad_norm": 0.47065848112106323, + "learning_rate": 0.00019996574471017113, + "loss": 1.1403, + "step": 207 + }, + { + "epoch": 0.037037037037037035, + "grad_norm": 0.4363015592098236, + "learning_rate": 0.00019996537738504373, + "loss": 1.253, + "step": 208 + }, + { + "epoch": 0.03721509971509972, + "grad_norm": 0.4038296937942505, + "learning_rate": 0.00019996500810130243, + "loss": 1.1679, + "step": 209 + }, + { + "epoch": 0.03739316239316239, + "grad_norm": 0.5038532018661499, + "learning_rate": 0.00019996463685895445, + "loss": 1.1182, + "step": 210 + }, + { + "epoch": 0.037571225071225074, + "grad_norm": 0.37740692496299744, + "learning_rate": 0.00019996426365800706, + "loss": 1.0465, + "step": 211 + }, + { + "epoch": 0.03774928774928775, + "grad_norm": 0.47794604301452637, + "learning_rate": 0.00019996388849846759, + "loss": 1.2836, + "step": 212 + }, + { + "epoch": 0.037927350427350424, + "grad_norm": 0.38460609316825867, + "learning_rate": 0.0001999635113803434, + "loss": 1.2099, + "step": 213 + }, + { + "epoch": 0.038105413105413107, + "grad_norm": 0.42016157507896423, + "learning_rate": 0.0001999631323036418, + "loss": 1.152, + "step": 214 + }, + { + "epoch": 0.03828347578347578, + "grad_norm": 0.4024946391582489, + "learning_rate": 0.00019996275126837033, + "loss": 1.1534, + "step": 215 + }, + { + "epoch": 0.038461538461538464, + "grad_norm": 0.4573793411254883, + "learning_rate": 0.00019996236827453642, + "loss": 1.2019, + "step": 216 + }, + { + "epoch": 0.03863960113960114, + "grad_norm": 0.3642503321170807, + "learning_rate": 0.0001999619833221475, + "loss": 1.0541, + "step": 217 + }, + { + "epoch": 0.03881766381766382, + "grad_norm": 0.38492897152900696, + "learning_rate": 0.0001999615964112112, + "loss": 1.1269, + "step": 218 + }, + { + "epoch": 0.038995726495726496, + "grad_norm": 0.427219420671463, + "learning_rate": 0.0001999612075417351, + "loss": 1.1126, + "step": 219 + }, + { + "epoch": 0.03917378917378917, + "grad_norm": 0.40781742334365845, + "learning_rate": 0.00019996081671372676, + "loss": 1.2207, + "step": 220 + }, + { + "epoch": 0.03935185185185185, + "grad_norm": 0.39229512214660645, + "learning_rate": 0.00019996042392719386, + "loss": 1.0403, + "step": 221 + }, + { + "epoch": 0.03952991452991453, + "grad_norm": 0.42038577795028687, + "learning_rate": 0.0001999600291821441, + "loss": 1.2157, + "step": 222 + }, + { + "epoch": 0.03970797720797721, + "grad_norm": 0.3963491916656494, + "learning_rate": 0.00019995963247858525, + "loss": 1.0532, + "step": 223 + }, + { + "epoch": 0.039886039886039885, + "grad_norm": 0.4389874041080475, + "learning_rate": 0.00019995923381652502, + "loss": 1.4279, + "step": 224 + }, + { + "epoch": 0.04006410256410257, + "grad_norm": 0.357312947511673, + "learning_rate": 0.00019995883319597123, + "loss": 0.9871, + "step": 225 + }, + { + "epoch": 0.04024216524216524, + "grad_norm": 0.3644427955150604, + "learning_rate": 0.00019995843061693181, + "loss": 1.0879, + "step": 226 + }, + { + "epoch": 0.04042022792022792, + "grad_norm": 0.4074651002883911, + "learning_rate": 0.00019995802607941453, + "loss": 1.2138, + "step": 227 + }, + { + "epoch": 0.0405982905982906, + "grad_norm": 0.40709465742111206, + "learning_rate": 0.0001999576195834274, + "loss": 1.1905, + "step": 228 + }, + { + "epoch": 0.040776353276353275, + "grad_norm": 0.4280182719230652, + "learning_rate": 0.00019995721112897838, + "loss": 1.2331, + "step": 229 + }, + { + "epoch": 0.04095441595441596, + "grad_norm": 0.37846076488494873, + "learning_rate": 0.00019995680071607544, + "loss": 1.078, + "step": 230 + }, + { + "epoch": 0.04113247863247863, + "grad_norm": 0.3877260088920593, + "learning_rate": 0.0001999563883447266, + "loss": 1.0309, + "step": 231 + }, + { + "epoch": 0.04131054131054131, + "grad_norm": 0.42886826395988464, + "learning_rate": 0.00019995597401494, + "loss": 1.0403, + "step": 232 + }, + { + "epoch": 0.04148860398860399, + "grad_norm": 0.4316534101963043, + "learning_rate": 0.00019995555772672372, + "loss": 1.2418, + "step": 233 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 0.45768865942955017, + "learning_rate": 0.00019995513948008593, + "loss": 1.233, + "step": 234 + }, + { + "epoch": 0.041844729344729346, + "grad_norm": 0.5647913813591003, + "learning_rate": 0.00019995471927503481, + "loss": 1.1346, + "step": 235 + }, + { + "epoch": 0.04202279202279202, + "grad_norm": 0.3797492980957031, + "learning_rate": 0.00019995429711157863, + "loss": 1.1574, + "step": 236 + }, + { + "epoch": 0.042200854700854704, + "grad_norm": 0.4392767548561096, + "learning_rate": 0.00019995387298972562, + "loss": 0.8988, + "step": 237 + }, + { + "epoch": 0.04237891737891738, + "grad_norm": 0.37331557273864746, + "learning_rate": 0.0001999534469094841, + "loss": 1.0439, + "step": 238 + }, + { + "epoch": 0.042556980056980054, + "grad_norm": 0.3785935938358307, + "learning_rate": 0.00019995301887086245, + "loss": 0.9839, + "step": 239 + }, + { + "epoch": 0.042735042735042736, + "grad_norm": 0.4351862668991089, + "learning_rate": 0.00019995258887386898, + "loss": 1.2653, + "step": 240 + }, + { + "epoch": 0.04291310541310541, + "grad_norm": 0.399475634098053, + "learning_rate": 0.0001999521569185122, + "loss": 0.9877, + "step": 241 + }, + { + "epoch": 0.04309116809116809, + "grad_norm": 0.42332810163497925, + "learning_rate": 0.00019995172300480053, + "loss": 1.2403, + "step": 242 + }, + { + "epoch": 0.04326923076923077, + "grad_norm": 0.4397708475589752, + "learning_rate": 0.00019995128713274247, + "loss": 0.9316, + "step": 243 + }, + { + "epoch": 0.04344729344729345, + "grad_norm": 0.3614110052585602, + "learning_rate": 0.00019995084930234658, + "loss": 1.1088, + "step": 244 + }, + { + "epoch": 0.043625356125356125, + "grad_norm": 0.39433717727661133, + "learning_rate": 0.0001999504095136214, + "loss": 1.2002, + "step": 245 + }, + { + "epoch": 0.0438034188034188, + "grad_norm": 0.33088216185569763, + "learning_rate": 0.0001999499677665756, + "loss": 0.8796, + "step": 246 + }, + { + "epoch": 0.04398148148148148, + "grad_norm": 0.5239143967628479, + "learning_rate": 0.00019994952406121784, + "loss": 1.2808, + "step": 247 + }, + { + "epoch": 0.04415954415954416, + "grad_norm": 0.42156723141670227, + "learning_rate": 0.00019994907839755675, + "loss": 1.1775, + "step": 248 + }, + { + "epoch": 0.04433760683760684, + "grad_norm": 0.42569902539253235, + "learning_rate": 0.0001999486307756011, + "loss": 1.001, + "step": 249 + }, + { + "epoch": 0.044515669515669515, + "grad_norm": 0.38241544365882874, + "learning_rate": 0.00019994818119535964, + "loss": 1.1064, + "step": 250 + }, + { + "epoch": 0.0446937321937322, + "grad_norm": 0.4185071885585785, + "learning_rate": 0.0001999477296568412, + "loss": 1.2109, + "step": 251 + }, + { + "epoch": 0.04487179487179487, + "grad_norm": 0.4189644157886505, + "learning_rate": 0.00019994727616005464, + "loss": 1.2902, + "step": 252 + }, + { + "epoch": 0.04504985754985755, + "grad_norm": 0.34671884775161743, + "learning_rate": 0.0001999468207050088, + "loss": 0.9429, + "step": 253 + }, + { + "epoch": 0.04522792022792023, + "grad_norm": 0.42391687631607056, + "learning_rate": 0.00019994636329171266, + "loss": 0.7179, + "step": 254 + }, + { + "epoch": 0.045405982905982904, + "grad_norm": 0.3803195655345917, + "learning_rate": 0.00019994590392017513, + "loss": 1.0318, + "step": 255 + }, + { + "epoch": 0.045584045584045586, + "grad_norm": 0.3389956057071686, + "learning_rate": 0.00019994544259040525, + "loss": 1.0485, + "step": 256 + }, + { + "epoch": 0.04576210826210826, + "grad_norm": 0.4927038550376892, + "learning_rate": 0.000199944979302412, + "loss": 1.3426, + "step": 257 + }, + { + "epoch": 0.045940170940170943, + "grad_norm": 0.33200421929359436, + "learning_rate": 0.00019994451405620453, + "loss": 1.0071, + "step": 258 + }, + { + "epoch": 0.04611823361823362, + "grad_norm": 0.38028615713119507, + "learning_rate": 0.00019994404685179195, + "loss": 1.0985, + "step": 259 + }, + { + "epoch": 0.046296296296296294, + "grad_norm": 0.3752151429653168, + "learning_rate": 0.00019994357768918333, + "loss": 0.9209, + "step": 260 + }, + { + "epoch": 0.046474358974358976, + "grad_norm": 0.43030866980552673, + "learning_rate": 0.00019994310656838796, + "loss": 0.9921, + "step": 261 + }, + { + "epoch": 0.04665242165242165, + "grad_norm": 0.4402460753917694, + "learning_rate": 0.00019994263348941502, + "loss": 1.1051, + "step": 262 + }, + { + "epoch": 0.04683048433048433, + "grad_norm": 0.43012720346450806, + "learning_rate": 0.0001999421584522738, + "loss": 1.1839, + "step": 263 + }, + { + "epoch": 0.04700854700854701, + "grad_norm": 0.4195305407047272, + "learning_rate": 0.0001999416814569736, + "loss": 1.1749, + "step": 264 + }, + { + "epoch": 0.04718660968660968, + "grad_norm": 0.45623287558555603, + "learning_rate": 0.00019994120250352372, + "loss": 1.2433, + "step": 265 + }, + { + "epoch": 0.047364672364672365, + "grad_norm": 0.4736156761646271, + "learning_rate": 0.00019994072159193363, + "loss": 1.2882, + "step": 266 + }, + { + "epoch": 0.04754273504273504, + "grad_norm": 0.36698561906814575, + "learning_rate": 0.0001999402387222127, + "loss": 1.1486, + "step": 267 + }, + { + "epoch": 0.04772079772079772, + "grad_norm": 0.3854144215583801, + "learning_rate": 0.00019993975389437038, + "loss": 0.8115, + "step": 268 + }, + { + "epoch": 0.0478988603988604, + "grad_norm": 0.41512808203697205, + "learning_rate": 0.0001999392671084162, + "loss": 1.0959, + "step": 269 + }, + { + "epoch": 0.04807692307692308, + "grad_norm": 0.3869563341140747, + "learning_rate": 0.0001999387783643597, + "loss": 1.087, + "step": 270 + }, + { + "epoch": 0.048254985754985755, + "grad_norm": 0.4649744927883148, + "learning_rate": 0.00019993828766221044, + "loss": 1.0011, + "step": 271 + }, + { + "epoch": 0.04843304843304843, + "grad_norm": 0.40331923961639404, + "learning_rate": 0.00019993779500197803, + "loss": 1.1463, + "step": 272 + }, + { + "epoch": 0.04861111111111111, + "grad_norm": 0.3826279938220978, + "learning_rate": 0.0001999373003836721, + "loss": 1.1491, + "step": 273 + }, + { + "epoch": 0.04878917378917379, + "grad_norm": 0.3967166543006897, + "learning_rate": 0.00019993680380730243, + "loss": 1.1462, + "step": 274 + }, + { + "epoch": 0.04896723646723647, + "grad_norm": 0.4298507869243622, + "learning_rate": 0.00019993630527287865, + "loss": 1.2471, + "step": 275 + }, + { + "epoch": 0.049145299145299144, + "grad_norm": 0.41486215591430664, + "learning_rate": 0.0001999358047804106, + "loss": 1.287, + "step": 276 + }, + { + "epoch": 0.049323361823361826, + "grad_norm": 0.3914124369621277, + "learning_rate": 0.00019993530232990803, + "loss": 1.0935, + "step": 277 + }, + { + "epoch": 0.0495014245014245, + "grad_norm": 0.39888378977775574, + "learning_rate": 0.00019993479792138082, + "loss": 1.2347, + "step": 278 + }, + { + "epoch": 0.049679487179487176, + "grad_norm": 0.3911665678024292, + "learning_rate": 0.00019993429155483884, + "loss": 1.0917, + "step": 279 + }, + { + "epoch": 0.04985754985754986, + "grad_norm": 0.42871445417404175, + "learning_rate": 0.00019993378323029197, + "loss": 1.0277, + "step": 280 + }, + { + "epoch": 0.050035612535612534, + "grad_norm": 0.35397860407829285, + "learning_rate": 0.00019993327294775027, + "loss": 0.9549, + "step": 281 + }, + { + "epoch": 0.050213675213675216, + "grad_norm": 0.4528059959411621, + "learning_rate": 0.00019993276070722364, + "loss": 1.2338, + "step": 282 + }, + { + "epoch": 0.05039173789173789, + "grad_norm": 0.354735791683197, + "learning_rate": 0.00019993224650872218, + "loss": 1.1892, + "step": 283 + }, + { + "epoch": 0.05056980056980057, + "grad_norm": 0.44407567381858826, + "learning_rate": 0.00019993173035225592, + "loss": 1.1621, + "step": 284 + }, + { + "epoch": 0.05074786324786325, + "grad_norm": 0.4177244305610657, + "learning_rate": 0.000199931212237835, + "loss": 1.1184, + "step": 285 + }, + { + "epoch": 0.05092592592592592, + "grad_norm": 0.5627759695053101, + "learning_rate": 0.0001999306921654696, + "loss": 1.0755, + "step": 286 + }, + { + "epoch": 0.051103988603988605, + "grad_norm": 0.46767523884773254, + "learning_rate": 0.00019993017013516986, + "loss": 1.2654, + "step": 287 + }, + { + "epoch": 0.05128205128205128, + "grad_norm": 0.4163128733634949, + "learning_rate": 0.000199929646146946, + "loss": 1.1307, + "step": 288 + }, + { + "epoch": 0.05146011396011396, + "grad_norm": 0.36954161524772644, + "learning_rate": 0.00019992912020080832, + "loss": 0.8274, + "step": 289 + }, + { + "epoch": 0.05163817663817664, + "grad_norm": 0.4770594835281372, + "learning_rate": 0.00019992859229676712, + "loss": 1.2235, + "step": 290 + }, + { + "epoch": 0.05181623931623932, + "grad_norm": 0.4174608290195465, + "learning_rate": 0.00019992806243483274, + "loss": 1.2893, + "step": 291 + }, + { + "epoch": 0.051994301994301995, + "grad_norm": 0.3794898986816406, + "learning_rate": 0.00019992753061501555, + "loss": 1.104, + "step": 292 + }, + { + "epoch": 0.05217236467236467, + "grad_norm": 0.3912592828273773, + "learning_rate": 0.000199926996837326, + "loss": 1.0043, + "step": 293 + }, + { + "epoch": 0.05235042735042735, + "grad_norm": 0.39641159772872925, + "learning_rate": 0.00019992646110177448, + "loss": 1.083, + "step": 294 + }, + { + "epoch": 0.05252849002849003, + "grad_norm": 0.3518857955932617, + "learning_rate": 0.00019992592340837157, + "loss": 0.9275, + "step": 295 + }, + { + "epoch": 0.05270655270655271, + "grad_norm": 0.3955721855163574, + "learning_rate": 0.00019992538375712777, + "loss": 1.0153, + "step": 296 + }, + { + "epoch": 0.052884615384615384, + "grad_norm": 0.3837333023548126, + "learning_rate": 0.00019992484214805364, + "loss": 1.1664, + "step": 297 + }, + { + "epoch": 0.053062678062678066, + "grad_norm": 0.39400920271873474, + "learning_rate": 0.0001999242985811598, + "loss": 1.0532, + "step": 298 + }, + { + "epoch": 0.05324074074074074, + "grad_norm": 0.39258649945259094, + "learning_rate": 0.00019992375305645692, + "loss": 1.0081, + "step": 299 + }, + { + "epoch": 0.053418803418803416, + "grad_norm": 0.49768248200416565, + "learning_rate": 0.00019992320557395566, + "loss": 1.2553, + "step": 300 + }, + { + "epoch": 0.0535968660968661, + "grad_norm": 0.364776074886322, + "learning_rate": 0.00019992265613366677, + "loss": 1.0582, + "step": 301 + }, + { + "epoch": 0.053774928774928774, + "grad_norm": 0.47317907214164734, + "learning_rate": 0.00019992210473560097, + "loss": 1.3114, + "step": 302 + }, + { + "epoch": 0.053952991452991456, + "grad_norm": 0.3706119656562805, + "learning_rate": 0.00019992155137976917, + "loss": 0.9554, + "step": 303 + }, + { + "epoch": 0.05413105413105413, + "grad_norm": 0.42809563875198364, + "learning_rate": 0.0001999209960661821, + "loss": 1.306, + "step": 304 + }, + { + "epoch": 0.054309116809116806, + "grad_norm": 0.4514487385749817, + "learning_rate": 0.00019992043879485066, + "loss": 1.0147, + "step": 305 + }, + { + "epoch": 0.05448717948717949, + "grad_norm": 0.36672836542129517, + "learning_rate": 0.0001999198795657858, + "loss": 1.1392, + "step": 306 + }, + { + "epoch": 0.05466524216524216, + "grad_norm": 0.4206554889678955, + "learning_rate": 0.00019991931837899847, + "loss": 1.2405, + "step": 307 + }, + { + "epoch": 0.054843304843304845, + "grad_norm": 0.46168261766433716, + "learning_rate": 0.00019991875523449966, + "loss": 1.2707, + "step": 308 + }, + { + "epoch": 0.05502136752136752, + "grad_norm": 0.39503365755081177, + "learning_rate": 0.00019991819013230039, + "loss": 1.0776, + "step": 309 + }, + { + "epoch": 0.0551994301994302, + "grad_norm": 0.35244834423065186, + "learning_rate": 0.00019991762307241178, + "loss": 1.0864, + "step": 310 + }, + { + "epoch": 0.05537749287749288, + "grad_norm": 0.3865319490432739, + "learning_rate": 0.0001999170540548449, + "loss": 1.3659, + "step": 311 + }, + { + "epoch": 0.05555555555555555, + "grad_norm": 0.3666876554489136, + "learning_rate": 0.0001999164830796109, + "loss": 0.9884, + "step": 312 + }, + { + "epoch": 0.055733618233618235, + "grad_norm": 0.4278281629085541, + "learning_rate": 0.00019991591014672096, + "loss": 1.1522, + "step": 313 + }, + { + "epoch": 0.05591168091168091, + "grad_norm": 0.4172627031803131, + "learning_rate": 0.0001999153352561863, + "loss": 1.2527, + "step": 314 + }, + { + "epoch": 0.05608974358974359, + "grad_norm": 0.38872212171554565, + "learning_rate": 0.00019991475840801823, + "loss": 1.2985, + "step": 315 + }, + { + "epoch": 0.05626780626780627, + "grad_norm": 0.4160458445549011, + "learning_rate": 0.00019991417960222804, + "loss": 1.1347, + "step": 316 + }, + { + "epoch": 0.05644586894586895, + "grad_norm": 0.5169723033905029, + "learning_rate": 0.00019991359883882705, + "loss": 1.0819, + "step": 317 + }, + { + "epoch": 0.056623931623931624, + "grad_norm": 0.42306259274482727, + "learning_rate": 0.0001999130161178266, + "loss": 1.3139, + "step": 318 + }, + { + "epoch": 0.0568019943019943, + "grad_norm": 0.41975873708724976, + "learning_rate": 0.00019991243143923816, + "loss": 1.2277, + "step": 319 + }, + { + "epoch": 0.05698005698005698, + "grad_norm": 0.3873472511768341, + "learning_rate": 0.00019991184480307324, + "loss": 1.156, + "step": 320 + }, + { + "epoch": 0.057158119658119656, + "grad_norm": 0.43656104803085327, + "learning_rate": 0.0001999112562093432, + "loss": 1.2344, + "step": 321 + }, + { + "epoch": 0.05733618233618234, + "grad_norm": 0.3738791048526764, + "learning_rate": 0.00019991066565805968, + "loss": 0.9573, + "step": 322 + }, + { + "epoch": 0.05751424501424501, + "grad_norm": 0.3838156461715698, + "learning_rate": 0.00019991007314923418, + "loss": 0.9274, + "step": 323 + }, + { + "epoch": 0.057692307692307696, + "grad_norm": 0.4564770758152008, + "learning_rate": 0.00019990947868287837, + "loss": 1.0756, + "step": 324 + }, + { + "epoch": 0.05787037037037037, + "grad_norm": 0.4560079872608185, + "learning_rate": 0.00019990888225900386, + "loss": 1.1508, + "step": 325 + }, + { + "epoch": 0.058048433048433046, + "grad_norm": 0.44356057047843933, + "learning_rate": 0.00019990828387762236, + "loss": 1.2323, + "step": 326 + }, + { + "epoch": 0.05822649572649573, + "grad_norm": 0.46390119194984436, + "learning_rate": 0.00019990768353874553, + "loss": 1.0031, + "step": 327 + }, + { + "epoch": 0.0584045584045584, + "grad_norm": 0.4502357244491577, + "learning_rate": 0.00019990708124238525, + "loss": 1.3454, + "step": 328 + }, + { + "epoch": 0.058582621082621085, + "grad_norm": 0.3979945182800293, + "learning_rate": 0.0001999064769885532, + "loss": 1.2833, + "step": 329 + }, + { + "epoch": 0.05876068376068376, + "grad_norm": 0.3899286687374115, + "learning_rate": 0.00019990587077726128, + "loss": 1.0175, + "step": 330 + }, + { + "epoch": 0.05893874643874644, + "grad_norm": 0.41422948241233826, + "learning_rate": 0.00019990526260852139, + "loss": 1.1151, + "step": 331 + }, + { + "epoch": 0.05911680911680912, + "grad_norm": 0.4266608953475952, + "learning_rate": 0.0001999046524823454, + "loss": 1.1119, + "step": 332 + }, + { + "epoch": 0.05929487179487179, + "grad_norm": 0.46563324332237244, + "learning_rate": 0.00019990404039874524, + "loss": 1.2358, + "step": 333 + }, + { + "epoch": 0.059472934472934474, + "grad_norm": 0.4404347240924835, + "learning_rate": 0.00019990342635773297, + "loss": 1.1748, + "step": 334 + }, + { + "epoch": 0.05965099715099715, + "grad_norm": 0.5133237838745117, + "learning_rate": 0.00019990281035932062, + "loss": 1.1649, + "step": 335 + }, + { + "epoch": 0.05982905982905983, + "grad_norm": 0.3593895435333252, + "learning_rate": 0.00019990219240352018, + "loss": 1.0318, + "step": 336 + }, + { + "epoch": 0.06000712250712251, + "grad_norm": 0.40554583072662354, + "learning_rate": 0.00019990157249034384, + "loss": 1.1202, + "step": 337 + }, + { + "epoch": 0.06018518518518518, + "grad_norm": 0.3770706057548523, + "learning_rate": 0.00019990095061980372, + "loss": 0.9908, + "step": 338 + }, + { + "epoch": 0.060363247863247864, + "grad_norm": 0.39676955342292786, + "learning_rate": 0.000199900326791912, + "loss": 0.8176, + "step": 339 + }, + { + "epoch": 0.06054131054131054, + "grad_norm": 0.41448578238487244, + "learning_rate": 0.00019989970100668086, + "loss": 1.2877, + "step": 340 + }, + { + "epoch": 0.06071937321937322, + "grad_norm": 0.4200015068054199, + "learning_rate": 0.00019989907326412265, + "loss": 1.2293, + "step": 341 + }, + { + "epoch": 0.060897435897435896, + "grad_norm": 0.47350621223449707, + "learning_rate": 0.0001998984435642496, + "loss": 1.2331, + "step": 342 + }, + { + "epoch": 0.06107549857549858, + "grad_norm": 0.47050634026527405, + "learning_rate": 0.00019989781190707406, + "loss": 0.8888, + "step": 343 + }, + { + "epoch": 0.06125356125356125, + "grad_norm": 0.4994896948337555, + "learning_rate": 0.00019989717829260842, + "loss": 1.0921, + "step": 344 + }, + { + "epoch": 0.06143162393162393, + "grad_norm": 0.36340200901031494, + "learning_rate": 0.0001998965427208651, + "loss": 0.9777, + "step": 345 + }, + { + "epoch": 0.06160968660968661, + "grad_norm": 0.3538152873516083, + "learning_rate": 0.00019989590519185654, + "loss": 1.0055, + "step": 346 + }, + { + "epoch": 0.061787749287749286, + "grad_norm": 0.5388944149017334, + "learning_rate": 0.00019989526570559526, + "loss": 1.1001, + "step": 347 + }, + { + "epoch": 0.06196581196581197, + "grad_norm": 0.4411574602127075, + "learning_rate": 0.00019989462426209373, + "loss": 1.0038, + "step": 348 + }, + { + "epoch": 0.06214387464387464, + "grad_norm": 0.3930876851081848, + "learning_rate": 0.00019989398086136455, + "loss": 1.1534, + "step": 349 + }, + { + "epoch": 0.062321937321937325, + "grad_norm": 0.47357070446014404, + "learning_rate": 0.00019989333550342033, + "loss": 1.2687, + "step": 350 + }, + { + "epoch": 0.0625, + "grad_norm": 0.40302303433418274, + "learning_rate": 0.00019989268818827372, + "loss": 1.1894, + "step": 351 + }, + { + "epoch": 0.06267806267806268, + "grad_norm": 0.4470510184764862, + "learning_rate": 0.00019989203891593738, + "loss": 1.2207, + "step": 352 + }, + { + "epoch": 0.06285612535612535, + "grad_norm": 0.42235100269317627, + "learning_rate": 0.00019989138768642406, + "loss": 1.2086, + "step": 353 + }, + { + "epoch": 0.06303418803418803, + "grad_norm": 0.38305309414863586, + "learning_rate": 0.0001998907344997465, + "loss": 1.0473, + "step": 354 + }, + { + "epoch": 0.06321225071225071, + "grad_norm": 0.3893027901649475, + "learning_rate": 0.0001998900793559175, + "loss": 1.1746, + "step": 355 + }, + { + "epoch": 0.0633903133903134, + "grad_norm": 0.41206735372543335, + "learning_rate": 0.0001998894222549499, + "loss": 1.188, + "step": 356 + }, + { + "epoch": 0.06356837606837606, + "grad_norm": 0.3700513243675232, + "learning_rate": 0.00019988876319685658, + "loss": 0.9862, + "step": 357 + }, + { + "epoch": 0.06374643874643875, + "grad_norm": 0.3708794116973877, + "learning_rate": 0.0001998881021816504, + "loss": 1.2003, + "step": 358 + }, + { + "epoch": 0.06392450142450143, + "grad_norm": 0.4058014154434204, + "learning_rate": 0.00019988743920934442, + "loss": 1.2311, + "step": 359 + }, + { + "epoch": 0.0641025641025641, + "grad_norm": 0.39134132862091064, + "learning_rate": 0.00019988677427995155, + "loss": 1.001, + "step": 360 + }, + { + "epoch": 0.06428062678062678, + "grad_norm": 0.3853437602519989, + "learning_rate": 0.00019988610739348484, + "loss": 1.0725, + "step": 361 + }, + { + "epoch": 0.06445868945868946, + "grad_norm": 0.47114330530166626, + "learning_rate": 0.00019988543854995735, + "loss": 1.2196, + "step": 362 + }, + { + "epoch": 0.06463675213675214, + "grad_norm": 0.40465688705444336, + "learning_rate": 0.00019988476774938216, + "loss": 1.1869, + "step": 363 + }, + { + "epoch": 0.06481481481481481, + "grad_norm": 0.40301886200904846, + "learning_rate": 0.00019988409499177245, + "loss": 1.1765, + "step": 364 + }, + { + "epoch": 0.0649928774928775, + "grad_norm": 0.43443185091018677, + "learning_rate": 0.0001998834202771414, + "loss": 1.2022, + "step": 365 + }, + { + "epoch": 0.06517094017094018, + "grad_norm": 0.4712986350059509, + "learning_rate": 0.00019988274360550217, + "loss": 1.156, + "step": 366 + }, + { + "epoch": 0.06534900284900284, + "grad_norm": 0.4524450898170471, + "learning_rate": 0.00019988206497686815, + "loss": 1.2917, + "step": 367 + }, + { + "epoch": 0.06552706552706553, + "grad_norm": 0.40302205085754395, + "learning_rate": 0.0001998813843912525, + "loss": 0.9993, + "step": 368 + }, + { + "epoch": 0.06570512820512821, + "grad_norm": 0.39435216784477234, + "learning_rate": 0.00019988070184866864, + "loss": 1.0914, + "step": 369 + }, + { + "epoch": 0.06588319088319089, + "grad_norm": 0.39267390966415405, + "learning_rate": 0.00019988001734912988, + "loss": 1.3138, + "step": 370 + }, + { + "epoch": 0.06606125356125356, + "grad_norm": 0.38351675868034363, + "learning_rate": 0.00019987933089264968, + "loss": 1.0997, + "step": 371 + }, + { + "epoch": 0.06623931623931624, + "grad_norm": 0.3294839859008789, + "learning_rate": 0.00019987864247924145, + "loss": 0.9656, + "step": 372 + }, + { + "epoch": 0.06641737891737892, + "grad_norm": 0.45333364605903625, + "learning_rate": 0.00019987795210891872, + "loss": 1.095, + "step": 373 + }, + { + "epoch": 0.06659544159544159, + "grad_norm": 0.4362282454967499, + "learning_rate": 0.00019987725978169501, + "loss": 1.2103, + "step": 374 + }, + { + "epoch": 0.06677350427350427, + "grad_norm": 0.41314780712127686, + "learning_rate": 0.00019987656549758385, + "loss": 1.2115, + "step": 375 + }, + { + "epoch": 0.06695156695156695, + "grad_norm": 0.4230864644050598, + "learning_rate": 0.00019987586925659888, + "loss": 1.17, + "step": 376 + }, + { + "epoch": 0.06712962962962964, + "grad_norm": 0.4703855812549591, + "learning_rate": 0.00019987517105875372, + "loss": 1.367, + "step": 377 + }, + { + "epoch": 0.0673076923076923, + "grad_norm": 0.4671297073364258, + "learning_rate": 0.00019987447090406206, + "loss": 1.2543, + "step": 378 + }, + { + "epoch": 0.06748575498575499, + "grad_norm": 0.43746981024742126, + "learning_rate": 0.0001998737687925376, + "loss": 1.214, + "step": 379 + }, + { + "epoch": 0.06766381766381767, + "grad_norm": 0.40889596939086914, + "learning_rate": 0.00019987306472419412, + "loss": 1.0496, + "step": 380 + }, + { + "epoch": 0.06784188034188034, + "grad_norm": 0.3677358627319336, + "learning_rate": 0.0001998723586990454, + "loss": 1.1242, + "step": 381 + }, + { + "epoch": 0.06801994301994302, + "grad_norm": 0.3892628848552704, + "learning_rate": 0.00019987165071710527, + "loss": 1.0246, + "step": 382 + }, + { + "epoch": 0.0681980056980057, + "grad_norm": 0.4281293749809265, + "learning_rate": 0.00019987094077838764, + "loss": 1.2817, + "step": 383 + }, + { + "epoch": 0.06837606837606838, + "grad_norm": 0.45030340552330017, + "learning_rate": 0.00019987022888290636, + "loss": 1.159, + "step": 384 + }, + { + "epoch": 0.06855413105413105, + "grad_norm": 0.6327905058860779, + "learning_rate": 0.00019986951503067545, + "loss": 0.9577, + "step": 385 + }, + { + "epoch": 0.06873219373219373, + "grad_norm": 0.40339627861976624, + "learning_rate": 0.0001998687992217088, + "loss": 1.138, + "step": 386 + }, + { + "epoch": 0.06891025641025642, + "grad_norm": 0.4018291234970093, + "learning_rate": 0.00019986808145602052, + "loss": 0.9109, + "step": 387 + }, + { + "epoch": 0.06908831908831908, + "grad_norm": 0.41566264629364014, + "learning_rate": 0.00019986736173362464, + "loss": 1.1516, + "step": 388 + }, + { + "epoch": 0.06926638176638177, + "grad_norm": 0.3569067418575287, + "learning_rate": 0.00019986664005453527, + "loss": 1.2329, + "step": 389 + }, + { + "epoch": 0.06944444444444445, + "grad_norm": 0.3959648907184601, + "learning_rate": 0.0001998659164187665, + "loss": 1.1041, + "step": 390 + }, + { + "epoch": 0.06962250712250712, + "grad_norm": 0.42853206396102905, + "learning_rate": 0.00019986519082633257, + "loss": 1.0859, + "step": 391 + }, + { + "epoch": 0.0698005698005698, + "grad_norm": 0.42005518078804016, + "learning_rate": 0.0001998644632772477, + "loss": 1.2017, + "step": 392 + }, + { + "epoch": 0.06997863247863248, + "grad_norm": 0.4296947419643402, + "learning_rate": 0.00019986373377152612, + "loss": 1.1464, + "step": 393 + }, + { + "epoch": 0.07015669515669516, + "grad_norm": 0.394747793674469, + "learning_rate": 0.0001998630023091821, + "loss": 1.0316, + "step": 394 + }, + { + "epoch": 0.07033475783475783, + "grad_norm": 0.3779357969760895, + "learning_rate": 0.00019986226889023002, + "loss": 1.1081, + "step": 395 + }, + { + "epoch": 0.07051282051282051, + "grad_norm": 0.4271804690361023, + "learning_rate": 0.00019986153351468424, + "loss": 0.985, + "step": 396 + }, + { + "epoch": 0.0706908831908832, + "grad_norm": 0.49412235617637634, + "learning_rate": 0.00019986079618255912, + "loss": 1.2606, + "step": 397 + }, + { + "epoch": 0.07086894586894586, + "grad_norm": 0.43657439947128296, + "learning_rate": 0.00019986005689386915, + "loss": 1.2266, + "step": 398 + }, + { + "epoch": 0.07104700854700854, + "grad_norm": 0.4060729444026947, + "learning_rate": 0.0001998593156486288, + "loss": 1.1787, + "step": 399 + }, + { + "epoch": 0.07122507122507123, + "grad_norm": 0.387046217918396, + "learning_rate": 0.00019985857244685264, + "loss": 0.9411, + "step": 400 + }, + { + "epoch": 0.07140313390313391, + "grad_norm": 0.4243999123573303, + "learning_rate": 0.00019985782728855516, + "loss": 1.2024, + "step": 401 + }, + { + "epoch": 0.07158119658119658, + "grad_norm": 0.43113812804222107, + "learning_rate": 0.000199857080173751, + "loss": 1.1246, + "step": 402 + }, + { + "epoch": 0.07175925925925926, + "grad_norm": 0.4653271436691284, + "learning_rate": 0.0001998563311024548, + "loss": 1.2343, + "step": 403 + }, + { + "epoch": 0.07193732193732194, + "grad_norm": 0.43260812759399414, + "learning_rate": 0.0001998555800746812, + "loss": 0.9543, + "step": 404 + }, + { + "epoch": 0.07211538461538461, + "grad_norm": 0.4635484516620636, + "learning_rate": 0.00019985482709044495, + "loss": 1.1091, + "step": 405 + }, + { + "epoch": 0.07229344729344729, + "grad_norm": 0.38362643122673035, + "learning_rate": 0.00019985407214976076, + "loss": 1.2584, + "step": 406 + }, + { + "epoch": 0.07247150997150997, + "grad_norm": 0.4068310558795929, + "learning_rate": 0.00019985331525264351, + "loss": 1.1944, + "step": 407 + }, + { + "epoch": 0.07264957264957266, + "grad_norm": 0.43909943103790283, + "learning_rate": 0.00019985255639910795, + "loss": 1.3748, + "step": 408 + }, + { + "epoch": 0.07282763532763532, + "grad_norm": 0.48674601316452026, + "learning_rate": 0.000199851795589169, + "loss": 1.2684, + "step": 409 + }, + { + "epoch": 0.073005698005698, + "grad_norm": 0.4218580722808838, + "learning_rate": 0.0001998510328228415, + "loss": 1.168, + "step": 410 + }, + { + "epoch": 0.07318376068376069, + "grad_norm": 0.4688236117362976, + "learning_rate": 0.00019985026810014046, + "loss": 1.3088, + "step": 411 + }, + { + "epoch": 0.07336182336182336, + "grad_norm": 0.3863612711429596, + "learning_rate": 0.00019984950142108083, + "loss": 1.0261, + "step": 412 + }, + { + "epoch": 0.07353988603988604, + "grad_norm": 0.4177640378475189, + "learning_rate": 0.00019984873278567765, + "loss": 1.1985, + "step": 413 + }, + { + "epoch": 0.07371794871794872, + "grad_norm": 0.4645586311817169, + "learning_rate": 0.00019984796219394592, + "loss": 1.2463, + "step": 414 + }, + { + "epoch": 0.0738960113960114, + "grad_norm": 0.5051766633987427, + "learning_rate": 0.00019984718964590083, + "loss": 1.3031, + "step": 415 + }, + { + "epoch": 0.07407407407407407, + "grad_norm": 0.4200040400028229, + "learning_rate": 0.0001998464151415575, + "loss": 1.0842, + "step": 416 + }, + { + "epoch": 0.07425213675213675, + "grad_norm": 0.34211036562919617, + "learning_rate": 0.000199845638680931, + "loss": 0.9659, + "step": 417 + }, + { + "epoch": 0.07443019943019943, + "grad_norm": 0.3553323447704315, + "learning_rate": 0.00019984486026403668, + "loss": 1.0102, + "step": 418 + }, + { + "epoch": 0.0746082621082621, + "grad_norm": 0.4967300295829773, + "learning_rate": 0.00019984407989088974, + "loss": 1.3125, + "step": 419 + }, + { + "epoch": 0.07478632478632478, + "grad_norm": 0.41649797558784485, + "learning_rate": 0.00019984329756150544, + "loss": 1.3092, + "step": 420 + }, + { + "epoch": 0.07496438746438747, + "grad_norm": 0.43825802206993103, + "learning_rate": 0.00019984251327589912, + "loss": 1.3678, + "step": 421 + }, + { + "epoch": 0.07514245014245015, + "grad_norm": 0.363394170999527, + "learning_rate": 0.00019984172703408617, + "loss": 1.305, + "step": 422 + }, + { + "epoch": 0.07532051282051282, + "grad_norm": 0.411563903093338, + "learning_rate": 0.000199840938836082, + "loss": 1.4248, + "step": 423 + }, + { + "epoch": 0.0754985754985755, + "grad_norm": 0.40548190474510193, + "learning_rate": 0.000199840148681902, + "loss": 1.1081, + "step": 424 + }, + { + "epoch": 0.07567663817663818, + "grad_norm": 0.3781099021434784, + "learning_rate": 0.00019983935657156171, + "loss": 1.185, + "step": 425 + }, + { + "epoch": 0.07585470085470085, + "grad_norm": 0.46597573161125183, + "learning_rate": 0.00019983856250507662, + "loss": 1.119, + "step": 426 + }, + { + "epoch": 0.07603276353276353, + "grad_norm": 0.3988197147846222, + "learning_rate": 0.00019983776648246232, + "loss": 1.206, + "step": 427 + }, + { + "epoch": 0.07621082621082621, + "grad_norm": 0.41210901737213135, + "learning_rate": 0.00019983696850373433, + "loss": 1.1843, + "step": 428 + }, + { + "epoch": 0.0763888888888889, + "grad_norm": 0.41870948672294617, + "learning_rate": 0.00019983616856890837, + "loss": 1.2248, + "step": 429 + }, + { + "epoch": 0.07656695156695156, + "grad_norm": 0.4320056140422821, + "learning_rate": 0.00019983536667800007, + "loss": 0.9743, + "step": 430 + }, + { + "epoch": 0.07674501424501425, + "grad_norm": 0.48455503582954407, + "learning_rate": 0.00019983456283102517, + "loss": 1.0438, + "step": 431 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 0.38712427020072937, + "learning_rate": 0.00019983375702799935, + "loss": 1.2041, + "step": 432 + }, + { + "epoch": 0.0771011396011396, + "grad_norm": 0.3578857481479645, + "learning_rate": 0.0001998329492689385, + "loss": 1.1623, + "step": 433 + }, + { + "epoch": 0.07727920227920228, + "grad_norm": 0.43065932393074036, + "learning_rate": 0.00019983213955385834, + "loss": 1.3033, + "step": 434 + }, + { + "epoch": 0.07745726495726496, + "grad_norm": 0.4882095754146576, + "learning_rate": 0.00019983132788277484, + "loss": 1.1635, + "step": 435 + }, + { + "epoch": 0.07763532763532764, + "grad_norm": 0.3429015874862671, + "learning_rate": 0.00019983051425570382, + "loss": 0.7289, + "step": 436 + }, + { + "epoch": 0.07781339031339031, + "grad_norm": 0.4320310056209564, + "learning_rate": 0.00019982969867266128, + "loss": 1.3685, + "step": 437 + }, + { + "epoch": 0.07799145299145299, + "grad_norm": 0.39891982078552246, + "learning_rate": 0.00019982888113366314, + "loss": 1.0444, + "step": 438 + }, + { + "epoch": 0.07816951566951567, + "grad_norm": 0.3675695061683655, + "learning_rate": 0.00019982806163872547, + "loss": 1.0527, + "step": 439 + }, + { + "epoch": 0.07834757834757834, + "grad_norm": 0.42824694514274597, + "learning_rate": 0.0001998272401878643, + "loss": 1.166, + "step": 440 + }, + { + "epoch": 0.07852564102564102, + "grad_norm": 0.3721694350242615, + "learning_rate": 0.00019982641678109575, + "loss": 1.1328, + "step": 441 + }, + { + "epoch": 0.0787037037037037, + "grad_norm": 0.33899208903312683, + "learning_rate": 0.00019982559141843592, + "loss": 1.016, + "step": 442 + }, + { + "epoch": 0.07888176638176639, + "grad_norm": 0.4029340147972107, + "learning_rate": 0.000199824764099901, + "loss": 1.0076, + "step": 443 + }, + { + "epoch": 0.07905982905982906, + "grad_norm": 0.4169132113456726, + "learning_rate": 0.0001998239348255072, + "loss": 1.208, + "step": 444 + }, + { + "epoch": 0.07923789173789174, + "grad_norm": 0.3865824043750763, + "learning_rate": 0.00019982310359527075, + "loss": 1.067, + "step": 445 + }, + { + "epoch": 0.07941595441595442, + "grad_norm": 0.4218919277191162, + "learning_rate": 0.00019982227040920796, + "loss": 1.195, + "step": 446 + }, + { + "epoch": 0.07959401709401709, + "grad_norm": 0.40504586696624756, + "learning_rate": 0.00019982143526733512, + "loss": 1.0188, + "step": 447 + }, + { + "epoch": 0.07977207977207977, + "grad_norm": 0.38330578804016113, + "learning_rate": 0.00019982059816966863, + "loss": 1.0484, + "step": 448 + }, + { + "epoch": 0.07995014245014245, + "grad_norm": 0.43731689453125, + "learning_rate": 0.00019981975911622488, + "loss": 1.074, + "step": 449 + }, + { + "epoch": 0.08012820512820513, + "grad_norm": 0.40858447551727295, + "learning_rate": 0.00019981891810702033, + "loss": 1.0008, + "step": 450 + }, + { + "epoch": 0.0803062678062678, + "grad_norm": 0.4031754732131958, + "learning_rate": 0.00019981807514207143, + "loss": 1.2179, + "step": 451 + }, + { + "epoch": 0.08048433048433049, + "grad_norm": 0.41920867562294006, + "learning_rate": 0.00019981723022139466, + "loss": 1.1406, + "step": 452 + }, + { + "epoch": 0.08066239316239317, + "grad_norm": 0.40305474400520325, + "learning_rate": 0.00019981638334500668, + "loss": 1.098, + "step": 453 + }, + { + "epoch": 0.08084045584045584, + "grad_norm": 0.4564182460308075, + "learning_rate": 0.00019981553451292396, + "loss": 1.419, + "step": 454 + }, + { + "epoch": 0.08101851851851852, + "grad_norm": 0.3832945227622986, + "learning_rate": 0.00019981468372516322, + "loss": 1.0919, + "step": 455 + }, + { + "epoch": 0.0811965811965812, + "grad_norm": 0.43062624335289, + "learning_rate": 0.0001998138309817411, + "loss": 1.0458, + "step": 456 + }, + { + "epoch": 0.08137464387464387, + "grad_norm": 0.3871173560619354, + "learning_rate": 0.0001998129762826743, + "loss": 1.1391, + "step": 457 + }, + { + "epoch": 0.08155270655270655, + "grad_norm": 0.43423157930374146, + "learning_rate": 0.0001998121196279796, + "loss": 1.1132, + "step": 458 + }, + { + "epoch": 0.08173076923076923, + "grad_norm": 0.4341012239456177, + "learning_rate": 0.00019981126101767372, + "loss": 1.113, + "step": 459 + }, + { + "epoch": 0.08190883190883191, + "grad_norm": 0.36748576164245605, + "learning_rate": 0.00019981040045177352, + "loss": 0.8108, + "step": 460 + }, + { + "epoch": 0.08208689458689458, + "grad_norm": 0.43133220076560974, + "learning_rate": 0.00019980953793029586, + "loss": 1.1861, + "step": 461 + }, + { + "epoch": 0.08226495726495726, + "grad_norm": 0.37204909324645996, + "learning_rate": 0.00019980867345325767, + "loss": 0.9222, + "step": 462 + }, + { + "epoch": 0.08244301994301995, + "grad_norm": 0.43370047211647034, + "learning_rate": 0.00019980780702067582, + "loss": 1.2984, + "step": 463 + }, + { + "epoch": 0.08262108262108261, + "grad_norm": 0.4991510808467865, + "learning_rate": 0.00019980693863256736, + "loss": 1.2222, + "step": 464 + }, + { + "epoch": 0.0827991452991453, + "grad_norm": 0.44318175315856934, + "learning_rate": 0.00019980606828894927, + "loss": 1.2262, + "step": 465 + }, + { + "epoch": 0.08297720797720798, + "grad_norm": 0.380231648683548, + "learning_rate": 0.0001998051959898386, + "loss": 1.0274, + "step": 466 + }, + { + "epoch": 0.08315527065527066, + "grad_norm": 0.39519667625427246, + "learning_rate": 0.0001998043217352524, + "loss": 1.2499, + "step": 467 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.457499235868454, + "learning_rate": 0.0001998034455252079, + "loss": 1.0751, + "step": 468 + }, + { + "epoch": 0.08351139601139601, + "grad_norm": 0.368522584438324, + "learning_rate": 0.00019980256735972215, + "loss": 1.0776, + "step": 469 + }, + { + "epoch": 0.08368945868945869, + "grad_norm": 0.3768427073955536, + "learning_rate": 0.00019980168723881243, + "loss": 1.2198, + "step": 470 + }, + { + "epoch": 0.08386752136752136, + "grad_norm": 0.37045565247535706, + "learning_rate": 0.000199800805162496, + "loss": 1.1816, + "step": 471 + }, + { + "epoch": 0.08404558404558404, + "grad_norm": 0.4219281077384949, + "learning_rate": 0.0001997999211307901, + "loss": 1.0515, + "step": 472 + }, + { + "epoch": 0.08422364672364673, + "grad_norm": 0.3815271258354187, + "learning_rate": 0.00019979903514371207, + "loss": 1.1709, + "step": 473 + }, + { + "epoch": 0.08440170940170941, + "grad_norm": 0.4566493630409241, + "learning_rate": 0.00019979814720127924, + "loss": 1.3063, + "step": 474 + }, + { + "epoch": 0.08457977207977208, + "grad_norm": 0.4043879806995392, + "learning_rate": 0.000199797257303509, + "loss": 1.0549, + "step": 475 + }, + { + "epoch": 0.08475783475783476, + "grad_norm": 0.3897830545902252, + "learning_rate": 0.00019979636545041886, + "loss": 1.1483, + "step": 476 + }, + { + "epoch": 0.08493589743589744, + "grad_norm": 0.36097025871276855, + "learning_rate": 0.00019979547164202622, + "loss": 1.1196, + "step": 477 + }, + { + "epoch": 0.08511396011396011, + "grad_norm": 0.3766986131668091, + "learning_rate": 0.00019979457587834863, + "loss": 1.0131, + "step": 478 + }, + { + "epoch": 0.08529202279202279, + "grad_norm": 0.39460286498069763, + "learning_rate": 0.00019979367815940364, + "loss": 1.1729, + "step": 479 + }, + { + "epoch": 0.08547008547008547, + "grad_norm": 0.4137469232082367, + "learning_rate": 0.00019979277848520885, + "loss": 1.2569, + "step": 480 + }, + { + "epoch": 0.08564814814814815, + "grad_norm": 0.464688777923584, + "learning_rate": 0.00019979187685578183, + "loss": 1.2064, + "step": 481 + }, + { + "epoch": 0.08582621082621082, + "grad_norm": 0.4245518147945404, + "learning_rate": 0.0001997909732711403, + "loss": 0.9812, + "step": 482 + }, + { + "epoch": 0.0860042735042735, + "grad_norm": 0.43368837237358093, + "learning_rate": 0.00019979006773130197, + "loss": 1.2822, + "step": 483 + }, + { + "epoch": 0.08618233618233619, + "grad_norm": 0.4232824444770813, + "learning_rate": 0.00019978916023628452, + "loss": 1.1446, + "step": 484 + }, + { + "epoch": 0.08636039886039885, + "grad_norm": 0.4183506369590759, + "learning_rate": 0.00019978825078610578, + "loss": 1.2605, + "step": 485 + }, + { + "epoch": 0.08653846153846154, + "grad_norm": 0.4391268491744995, + "learning_rate": 0.00019978733938078356, + "loss": 1.2165, + "step": 486 + }, + { + "epoch": 0.08671652421652422, + "grad_norm": 0.4139612317085266, + "learning_rate": 0.0001997864260203357, + "loss": 0.9389, + "step": 487 + }, + { + "epoch": 0.0868945868945869, + "grad_norm": 0.4058656096458435, + "learning_rate": 0.00019978551070478013, + "loss": 1.0652, + "step": 488 + }, + { + "epoch": 0.08707264957264957, + "grad_norm": 0.42333099246025085, + "learning_rate": 0.00019978459343413473, + "loss": 1.119, + "step": 489 + }, + { + "epoch": 0.08725071225071225, + "grad_norm": 0.4573031961917877, + "learning_rate": 0.00019978367420841754, + "loss": 1.1546, + "step": 490 + }, + { + "epoch": 0.08742877492877493, + "grad_norm": 0.4161617159843445, + "learning_rate": 0.00019978275302764655, + "loss": 1.0836, + "step": 491 + }, + { + "epoch": 0.0876068376068376, + "grad_norm": 0.422145277261734, + "learning_rate": 0.00019978182989183977, + "loss": 1.1908, + "step": 492 + }, + { + "epoch": 0.08778490028490028, + "grad_norm": 0.4588126838207245, + "learning_rate": 0.00019978090480101532, + "loss": 1.1758, + "step": 493 + }, + { + "epoch": 0.08796296296296297, + "grad_norm": 0.4425722062587738, + "learning_rate": 0.00019977997775519132, + "loss": 1.088, + "step": 494 + }, + { + "epoch": 0.08814102564102565, + "grad_norm": 0.37860307097435, + "learning_rate": 0.00019977904875438594, + "loss": 1.1532, + "step": 495 + }, + { + "epoch": 0.08831908831908832, + "grad_norm": 0.40435823798179626, + "learning_rate": 0.00019977811779861733, + "loss": 1.1271, + "step": 496 + }, + { + "epoch": 0.088497150997151, + "grad_norm": 0.42578884959220886, + "learning_rate": 0.0001997771848879038, + "loss": 0.9889, + "step": 497 + }, + { + "epoch": 0.08867521367521368, + "grad_norm": 0.3439478874206543, + "learning_rate": 0.00019977625002226361, + "loss": 1.1273, + "step": 498 + }, + { + "epoch": 0.08885327635327635, + "grad_norm": 0.362341970205307, + "learning_rate": 0.00019977531320171504, + "loss": 1.0214, + "step": 499 + }, + { + "epoch": 0.08903133903133903, + "grad_norm": 0.4305768609046936, + "learning_rate": 0.0001997743744262765, + "loss": 1.2648, + "step": 500 + }, + { + "epoch": 0.08920940170940171, + "grad_norm": 0.35900023579597473, + "learning_rate": 0.00019977343369596636, + "loss": 1.0274, + "step": 501 + }, + { + "epoch": 0.0893874643874644, + "grad_norm": 0.4950818717479706, + "learning_rate": 0.00019977249101080306, + "loss": 1.1483, + "step": 502 + }, + { + "epoch": 0.08956552706552706, + "grad_norm": 0.3800346553325653, + "learning_rate": 0.00019977154637080503, + "loss": 1.0636, + "step": 503 + }, + { + "epoch": 0.08974358974358974, + "grad_norm": 0.46202352643013, + "learning_rate": 0.0001997705997759908, + "loss": 1.1544, + "step": 504 + }, + { + "epoch": 0.08992165242165243, + "grad_norm": 0.36818403005599976, + "learning_rate": 0.00019976965122637895, + "loss": 0.9824, + "step": 505 + }, + { + "epoch": 0.0900997150997151, + "grad_norm": 0.40248095989227295, + "learning_rate": 0.00019976870072198805, + "loss": 1.1002, + "step": 506 + }, + { + "epoch": 0.09027777777777778, + "grad_norm": 0.3841850459575653, + "learning_rate": 0.00019976774826283667, + "loss": 1.2433, + "step": 507 + }, + { + "epoch": 0.09045584045584046, + "grad_norm": 0.46892330050468445, + "learning_rate": 0.0001997667938489435, + "loss": 1.3194, + "step": 508 + }, + { + "epoch": 0.09063390313390314, + "grad_norm": 0.39059561491012573, + "learning_rate": 0.0001997658374803273, + "loss": 1.1778, + "step": 509 + }, + { + "epoch": 0.09081196581196581, + "grad_norm": 0.3793235421180725, + "learning_rate": 0.00019976487915700672, + "loss": 1.0659, + "step": 510 + }, + { + "epoch": 0.09099002849002849, + "grad_norm": 0.39067742228507996, + "learning_rate": 0.00019976391887900058, + "loss": 1.107, + "step": 511 + }, + { + "epoch": 0.09116809116809117, + "grad_norm": 0.40121713280677795, + "learning_rate": 0.00019976295664632772, + "loss": 1.102, + "step": 512 + }, + { + "epoch": 0.09134615384615384, + "grad_norm": 0.49830010533332825, + "learning_rate": 0.00019976199245900697, + "loss": 1.1701, + "step": 513 + }, + { + "epoch": 0.09152421652421652, + "grad_norm": 0.4536968171596527, + "learning_rate": 0.0001997610263170572, + "loss": 1.1067, + "step": 514 + }, + { + "epoch": 0.0917022792022792, + "grad_norm": 0.3832971453666687, + "learning_rate": 0.00019976005822049735, + "loss": 1.0991, + "step": 515 + }, + { + "epoch": 0.09188034188034189, + "grad_norm": 0.4093509614467621, + "learning_rate": 0.0001997590881693464, + "loss": 1.0565, + "step": 516 + }, + { + "epoch": 0.09205840455840456, + "grad_norm": 0.46073687076568604, + "learning_rate": 0.0001997581161636233, + "loss": 1.0057, + "step": 517 + }, + { + "epoch": 0.09223646723646724, + "grad_norm": 0.5001922845840454, + "learning_rate": 0.0001997571422033472, + "loss": 1.2639, + "step": 518 + }, + { + "epoch": 0.09241452991452992, + "grad_norm": 0.4620618224143982, + "learning_rate": 0.00019975616628853713, + "loss": 1.0966, + "step": 519 + }, + { + "epoch": 0.09259259259259259, + "grad_norm": 0.3788183927536011, + "learning_rate": 0.0001997551884192122, + "loss": 0.9783, + "step": 520 + }, + { + "epoch": 0.09277065527065527, + "grad_norm": 0.45589539408683777, + "learning_rate": 0.00019975420859539154, + "loss": 1.2194, + "step": 521 + }, + { + "epoch": 0.09294871794871795, + "grad_norm": 0.40747523307800293, + "learning_rate": 0.00019975322681709443, + "loss": 1.0349, + "step": 522 + }, + { + "epoch": 0.09312678062678063, + "grad_norm": 0.5045142769813538, + "learning_rate": 0.00019975224308434002, + "loss": 1.1373, + "step": 523 + }, + { + "epoch": 0.0933048433048433, + "grad_norm": 0.40352702140808105, + "learning_rate": 0.00019975125739714767, + "loss": 1.1236, + "step": 524 + }, + { + "epoch": 0.09348290598290598, + "grad_norm": 0.4301735758781433, + "learning_rate": 0.0001997502697555366, + "loss": 1.2932, + "step": 525 + }, + { + "epoch": 0.09366096866096867, + "grad_norm": 0.36800238490104675, + "learning_rate": 0.00019974928015952624, + "loss": 1.0734, + "step": 526 + }, + { + "epoch": 0.09383903133903133, + "grad_norm": 0.4027230143547058, + "learning_rate": 0.00019974828860913594, + "loss": 1.2776, + "step": 527 + }, + { + "epoch": 0.09401709401709402, + "grad_norm": 0.42497140169143677, + "learning_rate": 0.0001997472951043851, + "loss": 1.248, + "step": 528 + }, + { + "epoch": 0.0941951566951567, + "grad_norm": 0.3888593018054962, + "learning_rate": 0.00019974629964529325, + "loss": 1.0231, + "step": 529 + }, + { + "epoch": 0.09437321937321937, + "grad_norm": 0.3761361241340637, + "learning_rate": 0.00019974530223187986, + "loss": 1.0216, + "step": 530 + }, + { + "epoch": 0.09455128205128205, + "grad_norm": 0.42192980647087097, + "learning_rate": 0.00019974430286416448, + "loss": 1.0731, + "step": 531 + }, + { + "epoch": 0.09472934472934473, + "grad_norm": 0.44244512915611267, + "learning_rate": 0.00019974330154216667, + "loss": 1.2793, + "step": 532 + }, + { + "epoch": 0.09490740740740741, + "grad_norm": 0.378252774477005, + "learning_rate": 0.0001997422982659061, + "loss": 1.0462, + "step": 533 + }, + { + "epoch": 0.09508547008547008, + "grad_norm": 0.45589110255241394, + "learning_rate": 0.00019974129303540236, + "loss": 1.1884, + "step": 534 + }, + { + "epoch": 0.09526353276353276, + "grad_norm": 0.33930808305740356, + "learning_rate": 0.0001997402858506752, + "loss": 0.8381, + "step": 535 + }, + { + "epoch": 0.09544159544159544, + "grad_norm": 0.45408427715301514, + "learning_rate": 0.0001997392767117443, + "loss": 1.2379, + "step": 536 + }, + { + "epoch": 0.09561965811965811, + "grad_norm": 0.44125741720199585, + "learning_rate": 0.0001997382656186295, + "loss": 1.1941, + "step": 537 + }, + { + "epoch": 0.0957977207977208, + "grad_norm": 0.4075697660446167, + "learning_rate": 0.00019973725257135054, + "loss": 1.0142, + "step": 538 + }, + { + "epoch": 0.09597578347578348, + "grad_norm": 0.4258415102958679, + "learning_rate": 0.00019973623756992733, + "loss": 1.0447, + "step": 539 + }, + { + "epoch": 0.09615384615384616, + "grad_norm": 0.2738485038280487, + "learning_rate": 0.0001997352206143797, + "loss": 0.5521, + "step": 540 + }, + { + "epoch": 0.09633190883190883, + "grad_norm": 0.38815587759017944, + "learning_rate": 0.00019973420170472762, + "loss": 1.1052, + "step": 541 + }, + { + "epoch": 0.09650997150997151, + "grad_norm": 0.3909834027290344, + "learning_rate": 0.00019973318084099106, + "loss": 1.0494, + "step": 542 + }, + { + "epoch": 0.09668803418803419, + "grad_norm": 0.4517597258090973, + "learning_rate": 0.00019973215802318996, + "loss": 1.0611, + "step": 543 + }, + { + "epoch": 0.09686609686609686, + "grad_norm": 0.48659002780914307, + "learning_rate": 0.00019973113325134442, + "loss": 0.9967, + "step": 544 + }, + { + "epoch": 0.09704415954415954, + "grad_norm": 0.4039791524410248, + "learning_rate": 0.0001997301065254745, + "loss": 1.251, + "step": 545 + }, + { + "epoch": 0.09722222222222222, + "grad_norm": 0.3985383212566376, + "learning_rate": 0.0001997290778456003, + "loss": 1.2263, + "step": 546 + }, + { + "epoch": 0.0974002849002849, + "grad_norm": 0.4540637731552124, + "learning_rate": 0.00019972804721174199, + "loss": 1.2084, + "step": 547 + }, + { + "epoch": 0.09757834757834757, + "grad_norm": 0.36867982149124146, + "learning_rate": 0.00019972701462391977, + "loss": 0.9704, + "step": 548 + }, + { + "epoch": 0.09775641025641026, + "grad_norm": 0.40199780464172363, + "learning_rate": 0.00019972598008215385, + "loss": 1.1121, + "step": 549 + }, + { + "epoch": 0.09793447293447294, + "grad_norm": 0.42728984355926514, + "learning_rate": 0.00019972494358646455, + "loss": 1.1606, + "step": 550 + }, + { + "epoch": 0.0981125356125356, + "grad_norm": 0.4212374687194824, + "learning_rate": 0.0001997239051368721, + "loss": 1.3093, + "step": 551 + }, + { + "epoch": 0.09829059829059829, + "grad_norm": 0.3972226083278656, + "learning_rate": 0.0001997228647333969, + "loss": 1.1218, + "step": 552 + }, + { + "epoch": 0.09846866096866097, + "grad_norm": 0.43649932742118835, + "learning_rate": 0.00019972182237605935, + "loss": 1.2532, + "step": 553 + }, + { + "epoch": 0.09864672364672365, + "grad_norm": 0.3812280595302582, + "learning_rate": 0.0001997207780648798, + "loss": 1.0409, + "step": 554 + }, + { + "epoch": 0.09882478632478632, + "grad_norm": 0.41684821248054504, + "learning_rate": 0.00019971973179987878, + "loss": 0.9569, + "step": 555 + }, + { + "epoch": 0.099002849002849, + "grad_norm": 0.38081470131874084, + "learning_rate": 0.00019971868358107674, + "loss": 1.1615, + "step": 556 + }, + { + "epoch": 0.09918091168091168, + "grad_norm": 0.3702073097229004, + "learning_rate": 0.0001997176334084943, + "loss": 1.3907, + "step": 557 + }, + { + "epoch": 0.09935897435897435, + "grad_norm": 0.3625728189945221, + "learning_rate": 0.00019971658128215193, + "loss": 1.1897, + "step": 558 + }, + { + "epoch": 0.09953703703703703, + "grad_norm": 0.3815405070781708, + "learning_rate": 0.0001997155272020703, + "loss": 1.1473, + "step": 559 + }, + { + "epoch": 0.09971509971509972, + "grad_norm": 0.48664286732673645, + "learning_rate": 0.00019971447116827004, + "loss": 1.2462, + "step": 560 + }, + { + "epoch": 0.0998931623931624, + "grad_norm": 0.3708696663379669, + "learning_rate": 0.0001997134131807719, + "loss": 1.0979, + "step": 561 + }, + { + "epoch": 0.10007122507122507, + "grad_norm": 0.44511324167251587, + "learning_rate": 0.00019971235323959654, + "loss": 1.2313, + "step": 562 + }, + { + "epoch": 0.10024928774928775, + "grad_norm": 0.3687448799610138, + "learning_rate": 0.00019971129134476473, + "loss": 1.1526, + "step": 563 + }, + { + "epoch": 0.10042735042735043, + "grad_norm": 0.4506866931915283, + "learning_rate": 0.00019971022749629735, + "loss": 1.0003, + "step": 564 + }, + { + "epoch": 0.1006054131054131, + "grad_norm": 0.41910406947135925, + "learning_rate": 0.00019970916169421515, + "loss": 1.013, + "step": 565 + }, + { + "epoch": 0.10078347578347578, + "grad_norm": 0.39728936553001404, + "learning_rate": 0.0001997080939385391, + "loss": 1.0501, + "step": 566 + }, + { + "epoch": 0.10096153846153846, + "grad_norm": 0.41415902972221375, + "learning_rate": 0.00019970702422929005, + "loss": 1.0791, + "step": 567 + }, + { + "epoch": 0.10113960113960115, + "grad_norm": 0.45630788803100586, + "learning_rate": 0.00019970595256648896, + "loss": 1.2884, + "step": 568 + }, + { + "epoch": 0.10131766381766381, + "grad_norm": 0.4371698796749115, + "learning_rate": 0.00019970487895015686, + "loss": 1.0684, + "step": 569 + }, + { + "epoch": 0.1014957264957265, + "grad_norm": 0.4350591003894806, + "learning_rate": 0.00019970380338031477, + "loss": 1.2415, + "step": 570 + }, + { + "epoch": 0.10167378917378918, + "grad_norm": 0.4232708215713501, + "learning_rate": 0.00019970272585698382, + "loss": 1.2656, + "step": 571 + }, + { + "epoch": 0.10185185185185185, + "grad_norm": 0.3917689919471741, + "learning_rate": 0.00019970164638018502, + "loss": 1.0178, + "step": 572 + }, + { + "epoch": 0.10202991452991453, + "grad_norm": 0.4262804388999939, + "learning_rate": 0.0001997005649499396, + "loss": 1.1805, + "step": 573 + }, + { + "epoch": 0.10220797720797721, + "grad_norm": 0.5217884182929993, + "learning_rate": 0.0001996994815662687, + "loss": 1.2392, + "step": 574 + }, + { + "epoch": 0.10238603988603989, + "grad_norm": 0.4273875057697296, + "learning_rate": 0.00019969839622919358, + "loss": 1.0844, + "step": 575 + }, + { + "epoch": 0.10256410256410256, + "grad_norm": 0.41588085889816284, + "learning_rate": 0.00019969730893873547, + "loss": 1.2437, + "step": 576 + }, + { + "epoch": 0.10274216524216524, + "grad_norm": 0.41617709398269653, + "learning_rate": 0.0001996962196949157, + "loss": 0.9519, + "step": 577 + }, + { + "epoch": 0.10292022792022792, + "grad_norm": 0.4832979142665863, + "learning_rate": 0.00019969512849775565, + "loss": 1.1889, + "step": 578 + }, + { + "epoch": 0.10309829059829059, + "grad_norm": 0.3936060965061188, + "learning_rate": 0.0001996940353472766, + "loss": 0.9888, + "step": 579 + }, + { + "epoch": 0.10327635327635327, + "grad_norm": 0.4147680997848511, + "learning_rate": 0.00019969294024350004, + "loss": 1.0733, + "step": 580 + }, + { + "epoch": 0.10345441595441596, + "grad_norm": 0.37791356444358826, + "learning_rate": 0.00019969184318644742, + "loss": 1.212, + "step": 581 + }, + { + "epoch": 0.10363247863247864, + "grad_norm": 0.44297221302986145, + "learning_rate": 0.00019969074417614023, + "loss": 1.0535, + "step": 582 + }, + { + "epoch": 0.10381054131054131, + "grad_norm": 0.4032835066318512, + "learning_rate": 0.0001996896432126, + "loss": 1.1869, + "step": 583 + }, + { + "epoch": 0.10398860398860399, + "grad_norm": 0.49271953105926514, + "learning_rate": 0.00019968854029584827, + "loss": 1.1661, + "step": 584 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 0.362699031829834, + "learning_rate": 0.0001996874354259067, + "loss": 0.868, + "step": 585 + }, + { + "epoch": 0.10434472934472934, + "grad_norm": 0.401795357465744, + "learning_rate": 0.0001996863286027969, + "loss": 1.1045, + "step": 586 + }, + { + "epoch": 0.10452279202279202, + "grad_norm": 0.45380479097366333, + "learning_rate": 0.00019968521982654058, + "loss": 0.8503, + "step": 587 + }, + { + "epoch": 0.1047008547008547, + "grad_norm": 0.49759066104888916, + "learning_rate": 0.00019968410909715947, + "loss": 1.4073, + "step": 588 + }, + { + "epoch": 0.10487891737891739, + "grad_norm": 0.4421198070049286, + "learning_rate": 0.0001996829964146753, + "loss": 1.1512, + "step": 589 + }, + { + "epoch": 0.10505698005698005, + "grad_norm": 0.46675658226013184, + "learning_rate": 0.00019968188177910988, + "loss": 1.0132, + "step": 590 + }, + { + "epoch": 0.10523504273504274, + "grad_norm": 0.5710657238960266, + "learning_rate": 0.00019968076519048507, + "loss": 1.267, + "step": 591 + }, + { + "epoch": 0.10541310541310542, + "grad_norm": 0.4655563235282898, + "learning_rate": 0.00019967964664882276, + "loss": 1.1204, + "step": 592 + }, + { + "epoch": 0.10559116809116809, + "grad_norm": 0.3895256519317627, + "learning_rate": 0.00019967852615414478, + "loss": 1.0814, + "step": 593 + }, + { + "epoch": 0.10576923076923077, + "grad_norm": 0.424216091632843, + "learning_rate": 0.00019967740370647322, + "loss": 1.1663, + "step": 594 + }, + { + "epoch": 0.10594729344729345, + "grad_norm": 0.3978985846042633, + "learning_rate": 0.00019967627930582996, + "loss": 0.909, + "step": 595 + }, + { + "epoch": 0.10612535612535613, + "grad_norm": 0.47064995765686035, + "learning_rate": 0.00019967515295223705, + "loss": 1.2351, + "step": 596 + }, + { + "epoch": 0.1063034188034188, + "grad_norm": 0.42449644207954407, + "learning_rate": 0.0001996740246457166, + "loss": 0.9739, + "step": 597 + }, + { + "epoch": 0.10648148148148148, + "grad_norm": 0.39033401012420654, + "learning_rate": 0.00019967289438629066, + "loss": 1.0933, + "step": 598 + }, + { + "epoch": 0.10665954415954416, + "grad_norm": 0.4398612678050995, + "learning_rate": 0.00019967176217398143, + "loss": 1.2479, + "step": 599 + }, + { + "epoch": 0.10683760683760683, + "grad_norm": 0.3946632742881775, + "learning_rate": 0.00019967062800881107, + "loss": 1.0417, + "step": 600 + }, + { + "epoch": 0.10701566951566951, + "grad_norm": 0.5083445906639099, + "learning_rate": 0.0001996694918908018, + "loss": 1.1109, + "step": 601 + }, + { + "epoch": 0.1071937321937322, + "grad_norm": 0.477724552154541, + "learning_rate": 0.00019966835381997585, + "loss": 1.2891, + "step": 602 + }, + { + "epoch": 0.10737179487179487, + "grad_norm": 0.4110167920589447, + "learning_rate": 0.0001996672137963556, + "loss": 1.0555, + "step": 603 + }, + { + "epoch": 0.10754985754985755, + "grad_norm": 0.44078320264816284, + "learning_rate": 0.00019966607181996334, + "loss": 0.9188, + "step": 604 + }, + { + "epoch": 0.10772792022792023, + "grad_norm": 0.41251105070114136, + "learning_rate": 0.00019966492789082142, + "loss": 1.2592, + "step": 605 + }, + { + "epoch": 0.10790598290598291, + "grad_norm": 0.37701505422592163, + "learning_rate": 0.00019966378200895227, + "loss": 1.0233, + "step": 606 + }, + { + "epoch": 0.10808404558404558, + "grad_norm": 0.44624966382980347, + "learning_rate": 0.00019966263417437835, + "loss": 1.2273, + "step": 607 + }, + { + "epoch": 0.10826210826210826, + "grad_norm": 0.3618549108505249, + "learning_rate": 0.00019966148438712214, + "loss": 0.9101, + "step": 608 + }, + { + "epoch": 0.10844017094017094, + "grad_norm": 0.384574294090271, + "learning_rate": 0.00019966033264720616, + "loss": 1.1769, + "step": 609 + }, + { + "epoch": 0.10861823361823361, + "grad_norm": 0.50872403383255, + "learning_rate": 0.000199659178954653, + "loss": 1.1213, + "step": 610 + }, + { + "epoch": 0.1087962962962963, + "grad_norm": 0.39736685156822205, + "learning_rate": 0.00019965802330948527, + "loss": 1.275, + "step": 611 + }, + { + "epoch": 0.10897435897435898, + "grad_norm": 0.484660267829895, + "learning_rate": 0.00019965686571172557, + "loss": 1.1671, + "step": 612 + }, + { + "epoch": 0.10915242165242166, + "grad_norm": 0.41420218348503113, + "learning_rate": 0.0001996557061613966, + "loss": 0.9541, + "step": 613 + }, + { + "epoch": 0.10933048433048433, + "grad_norm": 0.4057196080684662, + "learning_rate": 0.00019965454465852112, + "loss": 1.0145, + "step": 614 + }, + { + "epoch": 0.10950854700854701, + "grad_norm": 0.4559510052204132, + "learning_rate": 0.00019965338120312182, + "loss": 1.0889, + "step": 615 + }, + { + "epoch": 0.10968660968660969, + "grad_norm": 0.40960055589675903, + "learning_rate": 0.00019965221579522154, + "loss": 1.1447, + "step": 616 + }, + { + "epoch": 0.10986467236467236, + "grad_norm": 0.4701732099056244, + "learning_rate": 0.0001996510484348431, + "loss": 1.2871, + "step": 617 + }, + { + "epoch": 0.11004273504273504, + "grad_norm": 0.38420796394348145, + "learning_rate": 0.0001996498791220094, + "loss": 1.058, + "step": 618 + }, + { + "epoch": 0.11022079772079772, + "grad_norm": 0.4014730453491211, + "learning_rate": 0.00019964870785674327, + "loss": 1.023, + "step": 619 + }, + { + "epoch": 0.1103988603988604, + "grad_norm": 0.38846179842948914, + "learning_rate": 0.00019964753463906773, + "loss": 0.9834, + "step": 620 + }, + { + "epoch": 0.11057692307692307, + "grad_norm": 0.5120236277580261, + "learning_rate": 0.00019964635946900577, + "loss": 1.2347, + "step": 621 + }, + { + "epoch": 0.11075498575498575, + "grad_norm": 0.40483301877975464, + "learning_rate": 0.00019964518234658038, + "loss": 1.131, + "step": 622 + }, + { + "epoch": 0.11093304843304844, + "grad_norm": 0.445782870054245, + "learning_rate": 0.00019964400327181464, + "loss": 0.9349, + "step": 623 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 0.490460604429245, + "learning_rate": 0.00019964282224473165, + "loss": 1.0257, + "step": 624 + }, + { + "epoch": 0.11128917378917379, + "grad_norm": 0.37585243582725525, + "learning_rate": 0.00019964163926535454, + "loss": 0.9724, + "step": 625 + }, + { + "epoch": 0.11146723646723647, + "grad_norm": 0.4160473346710205, + "learning_rate": 0.00019964045433370651, + "loss": 0.874, + "step": 626 + }, + { + "epoch": 0.11164529914529915, + "grad_norm": 0.442425012588501, + "learning_rate": 0.00019963926744981074, + "loss": 1.064, + "step": 627 + }, + { + "epoch": 0.11182336182336182, + "grad_norm": 0.4451471269130707, + "learning_rate": 0.00019963807861369054, + "loss": 1.2343, + "step": 628 + }, + { + "epoch": 0.1120014245014245, + "grad_norm": 0.5018183588981628, + "learning_rate": 0.00019963688782536913, + "loss": 1.1226, + "step": 629 + }, + { + "epoch": 0.11217948717948718, + "grad_norm": 0.43723925948143005, + "learning_rate": 0.0001996356950848699, + "loss": 1.0178, + "step": 630 + }, + { + "epoch": 0.11235754985754985, + "grad_norm": 0.4794611930847168, + "learning_rate": 0.0001996345003922162, + "loss": 0.9695, + "step": 631 + }, + { + "epoch": 0.11253561253561253, + "grad_norm": 0.5021790266036987, + "learning_rate": 0.00019963330374743143, + "loss": 1.1748, + "step": 632 + }, + { + "epoch": 0.11271367521367522, + "grad_norm": 0.47228625416755676, + "learning_rate": 0.00019963210515053906, + "loss": 1.2138, + "step": 633 + }, + { + "epoch": 0.1128917378917379, + "grad_norm": 0.4261155128479004, + "learning_rate": 0.00019963090460156256, + "loss": 0.9428, + "step": 634 + }, + { + "epoch": 0.11306980056980057, + "grad_norm": 0.3279525339603424, + "learning_rate": 0.00019962970210052542, + "loss": 0.7803, + "step": 635 + }, + { + "epoch": 0.11324786324786325, + "grad_norm": 0.5106086730957031, + "learning_rate": 0.00019962849764745125, + "loss": 1.113, + "step": 636 + }, + { + "epoch": 0.11342592592592593, + "grad_norm": 0.38272222876548767, + "learning_rate": 0.00019962729124236363, + "loss": 0.896, + "step": 637 + }, + { + "epoch": 0.1136039886039886, + "grad_norm": 0.39532098174095154, + "learning_rate": 0.0001996260828852862, + "loss": 0.9308, + "step": 638 + }, + { + "epoch": 0.11378205128205128, + "grad_norm": 0.44947221875190735, + "learning_rate": 0.00019962487257624262, + "loss": 1.207, + "step": 639 + }, + { + "epoch": 0.11396011396011396, + "grad_norm": 0.40684598684310913, + "learning_rate": 0.00019962366031525664, + "loss": 1.11, + "step": 640 + }, + { + "epoch": 0.11413817663817664, + "grad_norm": 0.4296625852584839, + "learning_rate": 0.00019962244610235194, + "loss": 1.2784, + "step": 641 + }, + { + "epoch": 0.11431623931623931, + "grad_norm": 0.4560794532299042, + "learning_rate": 0.0001996212299375524, + "loss": 1.1191, + "step": 642 + }, + { + "epoch": 0.114494301994302, + "grad_norm": 0.40246087312698364, + "learning_rate": 0.00019962001182088177, + "loss": 1.1401, + "step": 643 + }, + { + "epoch": 0.11467236467236468, + "grad_norm": 0.3938910663127899, + "learning_rate": 0.000199618791752364, + "loss": 1.0959, + "step": 644 + }, + { + "epoch": 0.11485042735042734, + "grad_norm": 0.4123380184173584, + "learning_rate": 0.00019961756973202287, + "loss": 1.2824, + "step": 645 + }, + { + "epoch": 0.11502849002849003, + "grad_norm": 0.41085442900657654, + "learning_rate": 0.00019961634575988243, + "loss": 1.1137, + "step": 646 + }, + { + "epoch": 0.11520655270655271, + "grad_norm": 0.38276201486587524, + "learning_rate": 0.0001996151198359667, + "loss": 1.0747, + "step": 647 + }, + { + "epoch": 0.11538461538461539, + "grad_norm": 0.49269407987594604, + "learning_rate": 0.00019961389196029953, + "loss": 1.1731, + "step": 648 + }, + { + "epoch": 0.11556267806267806, + "grad_norm": 0.5152469277381897, + "learning_rate": 0.00019961266213290512, + "loss": 1.3574, + "step": 649 + }, + { + "epoch": 0.11574074074074074, + "grad_norm": 0.4835714101791382, + "learning_rate": 0.0001996114303538075, + "loss": 1.2859, + "step": 650 + }, + { + "epoch": 0.11591880341880342, + "grad_norm": 0.4284524917602539, + "learning_rate": 0.00019961019662303087, + "loss": 1.1103, + "step": 651 + }, + { + "epoch": 0.11609686609686609, + "grad_norm": 0.3933276832103729, + "learning_rate": 0.00019960896094059933, + "loss": 1.2647, + "step": 652 + }, + { + "epoch": 0.11627492877492877, + "grad_norm": 0.33749741315841675, + "learning_rate": 0.00019960772330653712, + "loss": 0.819, + "step": 653 + }, + { + "epoch": 0.11645299145299146, + "grad_norm": 0.48122069239616394, + "learning_rate": 0.00019960648372086852, + "loss": 1.2781, + "step": 654 + }, + { + "epoch": 0.11663105413105414, + "grad_norm": 0.4681607186794281, + "learning_rate": 0.00019960524218361775, + "loss": 0.9723, + "step": 655 + }, + { + "epoch": 0.1168091168091168, + "grad_norm": 0.3974960148334503, + "learning_rate": 0.0001996039986948092, + "loss": 1.0302, + "step": 656 + }, + { + "epoch": 0.11698717948717949, + "grad_norm": 0.43180662393569946, + "learning_rate": 0.0001996027532544672, + "loss": 1.3265, + "step": 657 + }, + { + "epoch": 0.11716524216524217, + "grad_norm": 0.4481917917728424, + "learning_rate": 0.00019960150586261613, + "loss": 1.136, + "step": 658 + }, + { + "epoch": 0.11734330484330484, + "grad_norm": 0.43428945541381836, + "learning_rate": 0.00019960025651928045, + "loss": 1.2412, + "step": 659 + }, + { + "epoch": 0.11752136752136752, + "grad_norm": 0.36211395263671875, + "learning_rate": 0.00019959900522448467, + "loss": 0.9563, + "step": 660 + }, + { + "epoch": 0.1176994301994302, + "grad_norm": 0.43585848808288574, + "learning_rate": 0.0001995977519782533, + "loss": 1.1677, + "step": 661 + }, + { + "epoch": 0.11787749287749288, + "grad_norm": 0.4232597351074219, + "learning_rate": 0.00019959649678061086, + "loss": 1.1187, + "step": 662 + }, + { + "epoch": 0.11805555555555555, + "grad_norm": 0.3304753303527832, + "learning_rate": 0.00019959523963158194, + "loss": 0.8473, + "step": 663 + }, + { + "epoch": 0.11823361823361823, + "grad_norm": 0.37600061297416687, + "learning_rate": 0.0001995939805311912, + "loss": 1.1227, + "step": 664 + }, + { + "epoch": 0.11841168091168092, + "grad_norm": 0.33417847752571106, + "learning_rate": 0.0001995927194794633, + "loss": 1.0315, + "step": 665 + }, + { + "epoch": 0.11858974358974358, + "grad_norm": 0.46799129247665405, + "learning_rate": 0.00019959145647642298, + "loss": 1.135, + "step": 666 + }, + { + "epoch": 0.11876780626780627, + "grad_norm": 0.4141576886177063, + "learning_rate": 0.0001995901915220949, + "loss": 1.0956, + "step": 667 + }, + { + "epoch": 0.11894586894586895, + "grad_norm": 0.3824596405029297, + "learning_rate": 0.0001995889246165039, + "loss": 1.1782, + "step": 668 + }, + { + "epoch": 0.11912393162393162, + "grad_norm": 0.4087786376476288, + "learning_rate": 0.00019958765575967484, + "loss": 0.9704, + "step": 669 + }, + { + "epoch": 0.1193019943019943, + "grad_norm": 0.5161317586898804, + "learning_rate": 0.00019958638495163252, + "loss": 1.2207, + "step": 670 + }, + { + "epoch": 0.11948005698005698, + "grad_norm": 0.4782274067401886, + "learning_rate": 0.0001995851121924019, + "loss": 1.1257, + "step": 671 + }, + { + "epoch": 0.11965811965811966, + "grad_norm": 0.40617331862449646, + "learning_rate": 0.00019958383748200782, + "loss": 1.1153, + "step": 672 + }, + { + "epoch": 0.11983618233618233, + "grad_norm": 0.40149980783462524, + "learning_rate": 0.00019958256082047533, + "loss": 0.9785, + "step": 673 + }, + { + "epoch": 0.12001424501424501, + "grad_norm": 0.4378886818885803, + "learning_rate": 0.00019958128220782942, + "loss": 1.1355, + "step": 674 + }, + { + "epoch": 0.1201923076923077, + "grad_norm": 0.4449596703052521, + "learning_rate": 0.0001995800016440952, + "loss": 1.0325, + "step": 675 + }, + { + "epoch": 0.12037037037037036, + "grad_norm": 0.4268079698085785, + "learning_rate": 0.00019957871912929765, + "loss": 1.1901, + "step": 676 + }, + { + "epoch": 0.12054843304843305, + "grad_norm": 0.4250091016292572, + "learning_rate": 0.00019957743466346198, + "loss": 1.0084, + "step": 677 + }, + { + "epoch": 0.12072649572649573, + "grad_norm": 0.40724286437034607, + "learning_rate": 0.0001995761482466133, + "loss": 1.0866, + "step": 678 + }, + { + "epoch": 0.12090455840455841, + "grad_norm": 0.42478349804878235, + "learning_rate": 0.00019957485987877688, + "loss": 1.1909, + "step": 679 + }, + { + "epoch": 0.12108262108262108, + "grad_norm": 0.371362566947937, + "learning_rate": 0.0001995735695599779, + "loss": 1.083, + "step": 680 + }, + { + "epoch": 0.12126068376068376, + "grad_norm": 0.4715283513069153, + "learning_rate": 0.0001995722772902417, + "loss": 1.2942, + "step": 681 + }, + { + "epoch": 0.12143874643874644, + "grad_norm": 0.3611983060836792, + "learning_rate": 0.00019957098306959355, + "loss": 0.9878, + "step": 682 + }, + { + "epoch": 0.12161680911680911, + "grad_norm": 0.4764883816242218, + "learning_rate": 0.00019956968689805883, + "loss": 1.0082, + "step": 683 + }, + { + "epoch": 0.12179487179487179, + "grad_norm": 0.33170604705810547, + "learning_rate": 0.00019956838877566293, + "loss": 0.8529, + "step": 684 + }, + { + "epoch": 0.12197293447293447, + "grad_norm": 0.46896886825561523, + "learning_rate": 0.00019956708870243133, + "loss": 1.0745, + "step": 685 + }, + { + "epoch": 0.12215099715099716, + "grad_norm": 0.4120674431324005, + "learning_rate": 0.00019956578667838941, + "loss": 1.1828, + "step": 686 + }, + { + "epoch": 0.12232905982905982, + "grad_norm": 0.45671191811561584, + "learning_rate": 0.00019956448270356275, + "loss": 1.3484, + "step": 687 + }, + { + "epoch": 0.1225071225071225, + "grad_norm": 0.4023838937282562, + "learning_rate": 0.00019956317677797687, + "loss": 0.9623, + "step": 688 + }, + { + "epoch": 0.12268518518518519, + "grad_norm": 0.5205856561660767, + "learning_rate": 0.00019956186890165737, + "loss": 1.2221, + "step": 689 + }, + { + "epoch": 0.12286324786324786, + "grad_norm": 0.43956050276756287, + "learning_rate": 0.00019956055907462987, + "loss": 1.1051, + "step": 690 + }, + { + "epoch": 0.12304131054131054, + "grad_norm": 0.4341758191585541, + "learning_rate": 0.00019955924729692003, + "loss": 0.8972, + "step": 691 + }, + { + "epoch": 0.12321937321937322, + "grad_norm": 0.42025020718574524, + "learning_rate": 0.00019955793356855357, + "loss": 1.1137, + "step": 692 + }, + { + "epoch": 0.1233974358974359, + "grad_norm": 0.44375079870224, + "learning_rate": 0.0001995566178895562, + "loss": 1.2783, + "step": 693 + }, + { + "epoch": 0.12357549857549857, + "grad_norm": 0.4703320264816284, + "learning_rate": 0.00019955530025995372, + "loss": 1.1991, + "step": 694 + }, + { + "epoch": 0.12375356125356125, + "grad_norm": 0.43781620264053345, + "learning_rate": 0.00019955398067977195, + "loss": 1.2316, + "step": 695 + }, + { + "epoch": 0.12393162393162394, + "grad_norm": 0.4362877607345581, + "learning_rate": 0.0001995526591490367, + "loss": 1.1374, + "step": 696 + }, + { + "epoch": 0.1241096866096866, + "grad_norm": 0.4434499442577362, + "learning_rate": 0.00019955133566777392, + "loss": 1.1034, + "step": 697 + }, + { + "epoch": 0.12428774928774929, + "grad_norm": 0.46613508462905884, + "learning_rate": 0.00019955001023600955, + "loss": 1.2252, + "step": 698 + }, + { + "epoch": 0.12446581196581197, + "grad_norm": 0.46226736903190613, + "learning_rate": 0.00019954868285376945, + "loss": 1.0296, + "step": 699 + }, + { + "epoch": 0.12464387464387465, + "grad_norm": 0.4460904002189636, + "learning_rate": 0.00019954735352107977, + "loss": 1.0553, + "step": 700 + }, + { + "epoch": 0.12482193732193732, + "grad_norm": 0.36708924174308777, + "learning_rate": 0.00019954602223796648, + "loss": 0.9384, + "step": 701 + }, + { + "epoch": 0.125, + "grad_norm": 0.3780093491077423, + "learning_rate": 0.00019954468900445566, + "loss": 0.9062, + "step": 702 + }, + { + "epoch": 0.12517806267806267, + "grad_norm": 0.41797417402267456, + "learning_rate": 0.00019954335382057345, + "loss": 1.0344, + "step": 703 + }, + { + "epoch": 0.12535612535612536, + "grad_norm": 0.43710798025131226, + "learning_rate": 0.00019954201668634597, + "loss": 1.1324, + "step": 704 + }, + { + "epoch": 0.12553418803418803, + "grad_norm": 0.4732789695262909, + "learning_rate": 0.00019954067760179952, + "loss": 1.1419, + "step": 705 + }, + { + "epoch": 0.1257122507122507, + "grad_norm": 0.43248575925827026, + "learning_rate": 0.00019953933656696022, + "loss": 1.5112, + "step": 706 + }, + { + "epoch": 0.1258903133903134, + "grad_norm": 0.4074753522872925, + "learning_rate": 0.00019953799358185442, + "loss": 0.9751, + "step": 707 + }, + { + "epoch": 0.12606837606837606, + "grad_norm": 0.4586823880672455, + "learning_rate": 0.0001995366486465084, + "loss": 1.267, + "step": 708 + }, + { + "epoch": 0.12624643874643873, + "grad_norm": 0.4716857075691223, + "learning_rate": 0.0001995353017609485, + "loss": 1.1636, + "step": 709 + }, + { + "epoch": 0.12642450142450143, + "grad_norm": 0.5214398503303528, + "learning_rate": 0.00019953395292520115, + "loss": 1.2317, + "step": 710 + }, + { + "epoch": 0.1266025641025641, + "grad_norm": 0.42961129546165466, + "learning_rate": 0.00019953260213929276, + "loss": 1.0271, + "step": 711 + }, + { + "epoch": 0.1267806267806268, + "grad_norm": 0.4764653444290161, + "learning_rate": 0.00019953124940324979, + "loss": 1.1747, + "step": 712 + }, + { + "epoch": 0.12695868945868946, + "grad_norm": 0.4420304000377655, + "learning_rate": 0.00019952989471709874, + "loss": 0.9783, + "step": 713 + }, + { + "epoch": 0.12713675213675213, + "grad_norm": 0.44114625453948975, + "learning_rate": 0.00019952853808086616, + "loss": 1.1953, + "step": 714 + }, + { + "epoch": 0.12731481481481483, + "grad_norm": 0.501923143863678, + "learning_rate": 0.0001995271794945786, + "loss": 0.9886, + "step": 715 + }, + { + "epoch": 0.1274928774928775, + "grad_norm": 0.42266538739204407, + "learning_rate": 0.00019952581895826276, + "loss": 1.2033, + "step": 716 + }, + { + "epoch": 0.12767094017094016, + "grad_norm": 0.37770554423332214, + "learning_rate": 0.00019952445647194523, + "loss": 1.0164, + "step": 717 + }, + { + "epoch": 0.12784900284900286, + "grad_norm": 0.369266152381897, + "learning_rate": 0.00019952309203565268, + "loss": 0.9186, + "step": 718 + }, + { + "epoch": 0.12802706552706553, + "grad_norm": 0.40446221828460693, + "learning_rate": 0.00019952172564941193, + "loss": 1.1576, + "step": 719 + }, + { + "epoch": 0.1282051282051282, + "grad_norm": 0.504172146320343, + "learning_rate": 0.00019952035731324967, + "loss": 1.2695, + "step": 720 + }, + { + "epoch": 0.1283831908831909, + "grad_norm": 0.37284108996391296, + "learning_rate": 0.0001995189870271928, + "loss": 1.0288, + "step": 721 + }, + { + "epoch": 0.12856125356125356, + "grad_norm": 0.41811618208885193, + "learning_rate": 0.00019951761479126805, + "loss": 1.2241, + "step": 722 + }, + { + "epoch": 0.12873931623931623, + "grad_norm": 0.44706249237060547, + "learning_rate": 0.0001995162406055024, + "loss": 1.0831, + "step": 723 + }, + { + "epoch": 0.12891737891737892, + "grad_norm": 0.426572322845459, + "learning_rate": 0.00019951486446992273, + "loss": 1.0047, + "step": 724 + }, + { + "epoch": 0.1290954415954416, + "grad_norm": 0.4446277618408203, + "learning_rate": 0.00019951348638455602, + "loss": 1.0827, + "step": 725 + }, + { + "epoch": 0.12927350427350429, + "grad_norm": 0.3934919834136963, + "learning_rate": 0.00019951210634942926, + "loss": 0.9808, + "step": 726 + }, + { + "epoch": 0.12945156695156695, + "grad_norm": 0.4316558241844177, + "learning_rate": 0.0001995107243645695, + "loss": 1.3341, + "step": 727 + }, + { + "epoch": 0.12962962962962962, + "grad_norm": 0.43074217438697815, + "learning_rate": 0.00019950934043000382, + "loss": 1.007, + "step": 728 + }, + { + "epoch": 0.12980769230769232, + "grad_norm": 0.5212171673774719, + "learning_rate": 0.0001995079545457593, + "loss": 1.1822, + "step": 729 + }, + { + "epoch": 0.129985754985755, + "grad_norm": 0.3749600946903229, + "learning_rate": 0.00019950656671186313, + "loss": 0.9657, + "step": 730 + }, + { + "epoch": 0.13016381766381765, + "grad_norm": 0.36626043915748596, + "learning_rate": 0.00019950517692834252, + "loss": 1.1274, + "step": 731 + }, + { + "epoch": 0.13034188034188035, + "grad_norm": 0.4635467529296875, + "learning_rate": 0.00019950378519522467, + "loss": 1.2305, + "step": 732 + }, + { + "epoch": 0.13051994301994302, + "grad_norm": 0.4077455699443817, + "learning_rate": 0.00019950239151253683, + "loss": 0.9485, + "step": 733 + }, + { + "epoch": 0.1306980056980057, + "grad_norm": 0.4222758114337921, + "learning_rate": 0.0001995009958803063, + "loss": 1.0376, + "step": 734 + }, + { + "epoch": 0.13087606837606838, + "grad_norm": 0.4330402612686157, + "learning_rate": 0.0001994995982985605, + "loss": 1.1774, + "step": 735 + }, + { + "epoch": 0.13105413105413105, + "grad_norm": 0.42275673151016235, + "learning_rate": 0.00019949819876732673, + "loss": 1.1238, + "step": 736 + }, + { + "epoch": 0.13123219373219372, + "grad_norm": 0.45576968789100647, + "learning_rate": 0.00019949679728663246, + "loss": 1.0428, + "step": 737 + }, + { + "epoch": 0.13141025641025642, + "grad_norm": 0.5508752465248108, + "learning_rate": 0.00019949539385650514, + "loss": 1.3221, + "step": 738 + }, + { + "epoch": 0.13158831908831908, + "grad_norm": 0.4115872383117676, + "learning_rate": 0.00019949398847697225, + "loss": 1.0301, + "step": 739 + }, + { + "epoch": 0.13176638176638178, + "grad_norm": 0.4662442207336426, + "learning_rate": 0.00019949258114806132, + "loss": 1.3263, + "step": 740 + }, + { + "epoch": 0.13194444444444445, + "grad_norm": 0.6077266931533813, + "learning_rate": 0.00019949117186979999, + "loss": 1.0269, + "step": 741 + }, + { + "epoch": 0.13212250712250712, + "grad_norm": 0.47039318084716797, + "learning_rate": 0.00019948976064221579, + "loss": 1.3782, + "step": 742 + }, + { + "epoch": 0.1323005698005698, + "grad_norm": 0.4773450493812561, + "learning_rate": 0.0001994883474653364, + "loss": 1.289, + "step": 743 + }, + { + "epoch": 0.13247863247863248, + "grad_norm": 0.40180155634880066, + "learning_rate": 0.00019948693233918952, + "loss": 0.8691, + "step": 744 + }, + { + "epoch": 0.13265669515669515, + "grad_norm": 0.45216289162635803, + "learning_rate": 0.00019948551526380288, + "loss": 1.071, + "step": 745 + }, + { + "epoch": 0.13283475783475784, + "grad_norm": 0.4289272427558899, + "learning_rate": 0.0001994840962392042, + "loss": 1.0422, + "step": 746 + }, + { + "epoch": 0.1330128205128205, + "grad_norm": 0.4617730379104614, + "learning_rate": 0.00019948267526542134, + "loss": 1.0835, + "step": 747 + }, + { + "epoch": 0.13319088319088318, + "grad_norm": 0.42710617184638977, + "learning_rate": 0.00019948125234248208, + "loss": 1.0535, + "step": 748 + }, + { + "epoch": 0.13336894586894588, + "grad_norm": 0.43433234095573425, + "learning_rate": 0.0001994798274704144, + "loss": 0.9313, + "step": 749 + }, + { + "epoch": 0.13354700854700854, + "grad_norm": 0.46270284056663513, + "learning_rate": 0.0001994784006492461, + "loss": 1.0903, + "step": 750 + }, + { + "epoch": 0.1337250712250712, + "grad_norm": 0.5319814682006836, + "learning_rate": 0.00019947697187900517, + "loss": 1.2329, + "step": 751 + }, + { + "epoch": 0.1339031339031339, + "grad_norm": 0.3511372208595276, + "learning_rate": 0.00019947554115971967, + "loss": 0.7116, + "step": 752 + }, + { + "epoch": 0.13408119658119658, + "grad_norm": 0.4103890359401703, + "learning_rate": 0.00019947410849141756, + "loss": 1.1527, + "step": 753 + }, + { + "epoch": 0.13425925925925927, + "grad_norm": 0.5390757322311401, + "learning_rate": 0.00019947267387412695, + "loss": 1.1682, + "step": 754 + }, + { + "epoch": 0.13443732193732194, + "grad_norm": 0.29939723014831543, + "learning_rate": 0.0001994712373078759, + "loss": 0.5848, + "step": 755 + }, + { + "epoch": 0.1346153846153846, + "grad_norm": 0.4605920612812042, + "learning_rate": 0.0001994697987926926, + "loss": 0.9448, + "step": 756 + }, + { + "epoch": 0.1347934472934473, + "grad_norm": 0.426213800907135, + "learning_rate": 0.00019946835832860527, + "loss": 1.0487, + "step": 757 + }, + { + "epoch": 0.13497150997150997, + "grad_norm": 0.4209515154361725, + "learning_rate": 0.00019946691591564203, + "loss": 1.0951, + "step": 758 + }, + { + "epoch": 0.13514957264957264, + "grad_norm": 0.39555591344833374, + "learning_rate": 0.0001994654715538312, + "loss": 0.8754, + "step": 759 + }, + { + "epoch": 0.13532763532763534, + "grad_norm": 0.4065483510494232, + "learning_rate": 0.0001994640252432011, + "loss": 0.9451, + "step": 760 + }, + { + "epoch": 0.135505698005698, + "grad_norm": 0.4489104151725769, + "learning_rate": 0.00019946257698378003, + "loss": 1.2031, + "step": 761 + }, + { + "epoch": 0.13568376068376067, + "grad_norm": 0.39928409457206726, + "learning_rate": 0.0001994611267755964, + "loss": 1.1124, + "step": 762 + }, + { + "epoch": 0.13586182336182337, + "grad_norm": 0.4145409166812897, + "learning_rate": 0.00019945967461867858, + "loss": 1.083, + "step": 763 + }, + { + "epoch": 0.13603988603988604, + "grad_norm": 0.43508613109588623, + "learning_rate": 0.00019945822051305507, + "loss": 1.1119, + "step": 764 + }, + { + "epoch": 0.1362179487179487, + "grad_norm": 0.5186598300933838, + "learning_rate": 0.0001994567644587543, + "loss": 1.3256, + "step": 765 + }, + { + "epoch": 0.1363960113960114, + "grad_norm": 0.4615778625011444, + "learning_rate": 0.00019945530645580487, + "loss": 1.3906, + "step": 766 + }, + { + "epoch": 0.13657407407407407, + "grad_norm": 0.4838152527809143, + "learning_rate": 0.00019945384650423532, + "loss": 0.8169, + "step": 767 + }, + { + "epoch": 0.13675213675213677, + "grad_norm": 0.49253368377685547, + "learning_rate": 0.0001994523846040742, + "loss": 1.1613, + "step": 768 + }, + { + "epoch": 0.13693019943019943, + "grad_norm": 0.4697009325027466, + "learning_rate": 0.00019945092075535024, + "loss": 1.1722, + "step": 769 + }, + { + "epoch": 0.1371082621082621, + "grad_norm": 0.47162383794784546, + "learning_rate": 0.00019944945495809204, + "loss": 1.054, + "step": 770 + }, + { + "epoch": 0.1372863247863248, + "grad_norm": 0.4653547704219818, + "learning_rate": 0.00019944798721232835, + "loss": 1.1791, + "step": 771 + }, + { + "epoch": 0.13746438746438747, + "grad_norm": 0.4244011640548706, + "learning_rate": 0.000199446517518088, + "loss": 1.1557, + "step": 772 + }, + { + "epoch": 0.13764245014245013, + "grad_norm": 0.43812859058380127, + "learning_rate": 0.00019944504587539967, + "loss": 1.1567, + "step": 773 + }, + { + "epoch": 0.13782051282051283, + "grad_norm": 0.3984275162220001, + "learning_rate": 0.00019944357228429227, + "loss": 1.0715, + "step": 774 + }, + { + "epoch": 0.1379985754985755, + "grad_norm": 0.3794248104095459, + "learning_rate": 0.0001994420967447946, + "loss": 0.9377, + "step": 775 + }, + { + "epoch": 0.13817663817663817, + "grad_norm": 0.4214578866958618, + "learning_rate": 0.00019944061925693566, + "loss": 1.0112, + "step": 776 + }, + { + "epoch": 0.13835470085470086, + "grad_norm": 0.4738999605178833, + "learning_rate": 0.00019943913982074435, + "loss": 0.8718, + "step": 777 + }, + { + "epoch": 0.13853276353276353, + "grad_norm": 0.43455326557159424, + "learning_rate": 0.00019943765843624965, + "loss": 1.1343, + "step": 778 + }, + { + "epoch": 0.1387108262108262, + "grad_norm": 0.44973456859588623, + "learning_rate": 0.00019943617510348062, + "loss": 1.0487, + "step": 779 + }, + { + "epoch": 0.1388888888888889, + "grad_norm": 0.4216597080230713, + "learning_rate": 0.00019943468982246628, + "loss": 1.0765, + "step": 780 + }, + { + "epoch": 0.13906695156695156, + "grad_norm": 0.5089883208274841, + "learning_rate": 0.00019943320259323578, + "loss": 1.3137, + "step": 781 + }, + { + "epoch": 0.13924501424501423, + "grad_norm": 0.4358222782611847, + "learning_rate": 0.00019943171341581822, + "loss": 1.1891, + "step": 782 + }, + { + "epoch": 0.13942307692307693, + "grad_norm": 0.40918609499931335, + "learning_rate": 0.00019943022229024275, + "loss": 1.279, + "step": 783 + }, + { + "epoch": 0.1396011396011396, + "grad_norm": 0.4614863395690918, + "learning_rate": 0.00019942872921653866, + "loss": 1.2477, + "step": 784 + }, + { + "epoch": 0.1397792022792023, + "grad_norm": 0.4141528904438019, + "learning_rate": 0.00019942723419473515, + "loss": 0.9622, + "step": 785 + }, + { + "epoch": 0.13995726495726496, + "grad_norm": 0.536139726638794, + "learning_rate": 0.00019942573722486154, + "loss": 1.2127, + "step": 786 + }, + { + "epoch": 0.14013532763532763, + "grad_norm": 0.4968845546245575, + "learning_rate": 0.0001994242383069471, + "loss": 1.2965, + "step": 787 + }, + { + "epoch": 0.14031339031339032, + "grad_norm": 0.3897174894809723, + "learning_rate": 0.00019942273744102132, + "loss": 0.9907, + "step": 788 + }, + { + "epoch": 0.140491452991453, + "grad_norm": 0.466307669878006, + "learning_rate": 0.0001994212346271135, + "loss": 1.2021, + "step": 789 + }, + { + "epoch": 0.14066951566951566, + "grad_norm": 0.49283576011657715, + "learning_rate": 0.0001994197298652531, + "loss": 1.0969, + "step": 790 + }, + { + "epoch": 0.14084757834757836, + "grad_norm": 0.4686102271080017, + "learning_rate": 0.00019941822315546964, + "loss": 1.0125, + "step": 791 + }, + { + "epoch": 0.14102564102564102, + "grad_norm": 0.4389997124671936, + "learning_rate": 0.0001994167144977926, + "loss": 1.1294, + "step": 792 + }, + { + "epoch": 0.1412037037037037, + "grad_norm": 0.38539355993270874, + "learning_rate": 0.00019941520389225162, + "loss": 1.1231, + "step": 793 + }, + { + "epoch": 0.1413817663817664, + "grad_norm": 0.4860847592353821, + "learning_rate": 0.00019941369133887618, + "loss": 1.2268, + "step": 794 + }, + { + "epoch": 0.14155982905982906, + "grad_norm": 0.4567467272281647, + "learning_rate": 0.00019941217683769598, + "loss": 1.1482, + "step": 795 + }, + { + "epoch": 0.14173789173789172, + "grad_norm": 0.5549420714378357, + "learning_rate": 0.00019941066038874067, + "loss": 1.1899, + "step": 796 + }, + { + "epoch": 0.14191595441595442, + "grad_norm": 0.3950003385543823, + "learning_rate": 0.00019940914199204, + "loss": 0.96, + "step": 797 + }, + { + "epoch": 0.1420940170940171, + "grad_norm": 0.43845999240875244, + "learning_rate": 0.00019940762164762373, + "loss": 1.0338, + "step": 798 + }, + { + "epoch": 0.14227207977207978, + "grad_norm": 0.468537300825119, + "learning_rate": 0.00019940609935552157, + "loss": 1.2416, + "step": 799 + }, + { + "epoch": 0.14245014245014245, + "grad_norm": 0.4292038679122925, + "learning_rate": 0.0001994045751157634, + "loss": 1.1397, + "step": 800 + }, + { + "epoch": 0.14262820512820512, + "grad_norm": 0.3800995647907257, + "learning_rate": 0.00019940304892837908, + "loss": 0.939, + "step": 801 + }, + { + "epoch": 0.14280626780626782, + "grad_norm": 0.38004353642463684, + "learning_rate": 0.00019940152079339852, + "loss": 1.0485, + "step": 802 + }, + { + "epoch": 0.14298433048433049, + "grad_norm": 0.4658142924308777, + "learning_rate": 0.00019939999071085163, + "loss": 1.1561, + "step": 803 + }, + { + "epoch": 0.14316239316239315, + "grad_norm": 0.4235048294067383, + "learning_rate": 0.0001993984586807684, + "loss": 1.0516, + "step": 804 + }, + { + "epoch": 0.14334045584045585, + "grad_norm": 0.42925819754600525, + "learning_rate": 0.00019939692470317887, + "loss": 1.2238, + "step": 805 + }, + { + "epoch": 0.14351851851851852, + "grad_norm": 0.43701639771461487, + "learning_rate": 0.00019939538877811308, + "loss": 1.0129, + "step": 806 + }, + { + "epoch": 0.14369658119658119, + "grad_norm": 0.42786353826522827, + "learning_rate": 0.00019939385090560113, + "loss": 1.1355, + "step": 807 + }, + { + "epoch": 0.14387464387464388, + "grad_norm": 0.371218740940094, + "learning_rate": 0.00019939231108567312, + "loss": 0.9712, + "step": 808 + }, + { + "epoch": 0.14405270655270655, + "grad_norm": 0.4834294617176056, + "learning_rate": 0.00019939076931835926, + "loss": 1.1375, + "step": 809 + }, + { + "epoch": 0.14423076923076922, + "grad_norm": 0.4700150191783905, + "learning_rate": 0.00019938922560368974, + "loss": 1.1943, + "step": 810 + }, + { + "epoch": 0.14440883190883191, + "grad_norm": 0.4430996775627136, + "learning_rate": 0.0001993876799416948, + "loss": 1.1976, + "step": 811 + }, + { + "epoch": 0.14458689458689458, + "grad_norm": 0.4161672592163086, + "learning_rate": 0.00019938613233240476, + "loss": 1.0291, + "step": 812 + }, + { + "epoch": 0.14476495726495728, + "grad_norm": 0.39838850498199463, + "learning_rate": 0.0001993845827758499, + "loss": 1.2103, + "step": 813 + }, + { + "epoch": 0.14494301994301995, + "grad_norm": 0.429198294878006, + "learning_rate": 0.00019938303127206057, + "loss": 0.9971, + "step": 814 + }, + { + "epoch": 0.14512108262108261, + "grad_norm": 0.4589254856109619, + "learning_rate": 0.00019938147782106719, + "loss": 1.2392, + "step": 815 + }, + { + "epoch": 0.1452991452991453, + "grad_norm": 0.42506635189056396, + "learning_rate": 0.00019937992242290023, + "loss": 1.0827, + "step": 816 + }, + { + "epoch": 0.14547720797720798, + "grad_norm": 0.3778113126754761, + "learning_rate": 0.00019937836507759012, + "loss": 1.021, + "step": 817 + }, + { + "epoch": 0.14565527065527065, + "grad_norm": 0.43071216344833374, + "learning_rate": 0.0001993768057851674, + "loss": 1.273, + "step": 818 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 0.4944681227207184, + "learning_rate": 0.00019937524454566262, + "loss": 1.3037, + "step": 819 + }, + { + "epoch": 0.146011396011396, + "grad_norm": 0.4438824951648712, + "learning_rate": 0.00019937368135910632, + "loss": 1.1383, + "step": 820 + }, + { + "epoch": 0.14618945868945868, + "grad_norm": 0.400215744972229, + "learning_rate": 0.0001993721162255292, + "loss": 1.0669, + "step": 821 + }, + { + "epoch": 0.14636752136752137, + "grad_norm": 0.4341452121734619, + "learning_rate": 0.00019937054914496185, + "loss": 1.1431, + "step": 822 + }, + { + "epoch": 0.14654558404558404, + "grad_norm": 0.3941744267940521, + "learning_rate": 0.00019936898011743503, + "loss": 1.1593, + "step": 823 + }, + { + "epoch": 0.1467236467236467, + "grad_norm": 0.4318541884422302, + "learning_rate": 0.00019936740914297947, + "loss": 1.2814, + "step": 824 + }, + { + "epoch": 0.1469017094017094, + "grad_norm": 0.44488632678985596, + "learning_rate": 0.00019936583622162595, + "loss": 1.1054, + "step": 825 + }, + { + "epoch": 0.14707977207977208, + "grad_norm": 0.38701096177101135, + "learning_rate": 0.00019936426135340528, + "loss": 1.1086, + "step": 826 + }, + { + "epoch": 0.14725783475783477, + "grad_norm": 0.45794424414634705, + "learning_rate": 0.0001993626845383483, + "loss": 1.2395, + "step": 827 + }, + { + "epoch": 0.14743589743589744, + "grad_norm": 0.49237680435180664, + "learning_rate": 0.00019936110577648596, + "loss": 1.3483, + "step": 828 + }, + { + "epoch": 0.1476139601139601, + "grad_norm": 0.481666624546051, + "learning_rate": 0.00019935952506784914, + "loss": 1.1848, + "step": 829 + }, + { + "epoch": 0.1477920227920228, + "grad_norm": 0.4015209376811981, + "learning_rate": 0.00019935794241246883, + "loss": 1.0624, + "step": 830 + }, + { + "epoch": 0.14797008547008547, + "grad_norm": 0.47975999116897583, + "learning_rate": 0.00019935635781037606, + "loss": 1.1595, + "step": 831 + }, + { + "epoch": 0.14814814814814814, + "grad_norm": 0.4440356492996216, + "learning_rate": 0.00019935477126160181, + "loss": 1.1325, + "step": 832 + }, + { + "epoch": 0.14832621082621084, + "grad_norm": 0.4167410731315613, + "learning_rate": 0.00019935318276617723, + "loss": 1.0662, + "step": 833 + }, + { + "epoch": 0.1485042735042735, + "grad_norm": 0.4107447862625122, + "learning_rate": 0.0001993515923241334, + "loss": 0.8816, + "step": 834 + }, + { + "epoch": 0.14868233618233617, + "grad_norm": 0.4020158648490906, + "learning_rate": 0.00019934999993550154, + "loss": 0.9797, + "step": 835 + }, + { + "epoch": 0.14886039886039887, + "grad_norm": 0.4186473786830902, + "learning_rate": 0.0001993484056003128, + "loss": 1.1243, + "step": 836 + }, + { + "epoch": 0.14903846153846154, + "grad_norm": 0.5534794926643372, + "learning_rate": 0.00019934680931859842, + "loss": 1.1189, + "step": 837 + }, + { + "epoch": 0.1492165242165242, + "grad_norm": 0.37901270389556885, + "learning_rate": 0.0001993452110903897, + "loss": 0.9241, + "step": 838 + }, + { + "epoch": 0.1493945868945869, + "grad_norm": 0.41773587465286255, + "learning_rate": 0.00019934361091571793, + "loss": 0.9467, + "step": 839 + }, + { + "epoch": 0.14957264957264957, + "grad_norm": 0.4962073564529419, + "learning_rate": 0.00019934200879461448, + "loss": 1.2423, + "step": 840 + }, + { + "epoch": 0.14975071225071226, + "grad_norm": 0.38565897941589355, + "learning_rate": 0.00019934040472711074, + "loss": 1.1545, + "step": 841 + }, + { + "epoch": 0.14992877492877493, + "grad_norm": 0.4295346736907959, + "learning_rate": 0.0001993387987132381, + "loss": 1.2482, + "step": 842 + }, + { + "epoch": 0.1501068376068376, + "grad_norm": 0.4279189705848694, + "learning_rate": 0.0001993371907530281, + "loss": 1.1135, + "step": 843 + }, + { + "epoch": 0.1502849002849003, + "grad_norm": 0.44649168848991394, + "learning_rate": 0.0001993355808465122, + "loss": 1.0734, + "step": 844 + }, + { + "epoch": 0.15046296296296297, + "grad_norm": 0.453707218170166, + "learning_rate": 0.0001993339689937219, + "loss": 1.0992, + "step": 845 + }, + { + "epoch": 0.15064102564102563, + "grad_norm": 0.5113263726234436, + "learning_rate": 0.00019933235519468886, + "loss": 1.1792, + "step": 846 + }, + { + "epoch": 0.15081908831908833, + "grad_norm": 0.5822970271110535, + "learning_rate": 0.00019933073944944466, + "loss": 1.367, + "step": 847 + }, + { + "epoch": 0.150997150997151, + "grad_norm": 0.3946528732776642, + "learning_rate": 0.00019932912175802097, + "loss": 0.9781, + "step": 848 + }, + { + "epoch": 0.15117521367521367, + "grad_norm": 0.5429860949516296, + "learning_rate": 0.00019932750212044945, + "loss": 0.9783, + "step": 849 + }, + { + "epoch": 0.15135327635327636, + "grad_norm": 0.45847952365875244, + "learning_rate": 0.0001993258805367619, + "loss": 1.1352, + "step": 850 + }, + { + "epoch": 0.15153133903133903, + "grad_norm": 0.42770692706108093, + "learning_rate": 0.00019932425700699004, + "loss": 1.2365, + "step": 851 + }, + { + "epoch": 0.1517094017094017, + "grad_norm": 0.41845405101776123, + "learning_rate": 0.00019932263153116565, + "loss": 1.2642, + "step": 852 + }, + { + "epoch": 0.1518874643874644, + "grad_norm": 0.4641731083393097, + "learning_rate": 0.00019932100410932066, + "loss": 1.2009, + "step": 853 + }, + { + "epoch": 0.15206552706552706, + "grad_norm": 0.4128672778606415, + "learning_rate": 0.00019931937474148689, + "loss": 1.1981, + "step": 854 + }, + { + "epoch": 0.15224358974358973, + "grad_norm": 0.4730764925479889, + "learning_rate": 0.00019931774342769632, + "loss": 1.2145, + "step": 855 + }, + { + "epoch": 0.15242165242165243, + "grad_norm": 0.36611825227737427, + "learning_rate": 0.00019931611016798089, + "loss": 0.8504, + "step": 856 + }, + { + "epoch": 0.1525997150997151, + "grad_norm": 0.40944692492485046, + "learning_rate": 0.00019931447496237254, + "loss": 1.2853, + "step": 857 + }, + { + "epoch": 0.1527777777777778, + "grad_norm": 0.4521993398666382, + "learning_rate": 0.0001993128378109034, + "loss": 1.0198, + "step": 858 + }, + { + "epoch": 0.15295584045584046, + "grad_norm": 0.42113015055656433, + "learning_rate": 0.0001993111987136055, + "loss": 1.1284, + "step": 859 + }, + { + "epoch": 0.15313390313390313, + "grad_norm": 0.4117624759674072, + "learning_rate": 0.00019930955767051098, + "loss": 1.0445, + "step": 860 + }, + { + "epoch": 0.15331196581196582, + "grad_norm": 0.4807964265346527, + "learning_rate": 0.00019930791468165197, + "loss": 1.1378, + "step": 861 + }, + { + "epoch": 0.1534900284900285, + "grad_norm": 0.4186483323574066, + "learning_rate": 0.00019930626974706063, + "loss": 1.1636, + "step": 862 + }, + { + "epoch": 0.15366809116809116, + "grad_norm": 0.3764737844467163, + "learning_rate": 0.00019930462286676926, + "loss": 0.9523, + "step": 863 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.4283556044101715, + "learning_rate": 0.00019930297404081008, + "loss": 1.1008, + "step": 864 + }, + { + "epoch": 0.15402421652421652, + "grad_norm": 0.4485796093940735, + "learning_rate": 0.00019930132326921541, + "loss": 1.0834, + "step": 865 + }, + { + "epoch": 0.1542022792022792, + "grad_norm": 0.3882720172405243, + "learning_rate": 0.0001992996705520176, + "loss": 1.1086, + "step": 866 + }, + { + "epoch": 0.1543803418803419, + "grad_norm": 0.44698455929756165, + "learning_rate": 0.00019929801588924902, + "loss": 1.1437, + "step": 867 + }, + { + "epoch": 0.15455840455840456, + "grad_norm": 0.46978411078453064, + "learning_rate": 0.00019929635928094208, + "loss": 1.091, + "step": 868 + }, + { + "epoch": 0.15473646723646722, + "grad_norm": 0.4717854857444763, + "learning_rate": 0.00019929470072712927, + "loss": 1.1959, + "step": 869 + }, + { + "epoch": 0.15491452991452992, + "grad_norm": 0.4324854016304016, + "learning_rate": 0.00019929304022784305, + "loss": 1.2062, + "step": 870 + }, + { + "epoch": 0.1550925925925926, + "grad_norm": 0.3948180675506592, + "learning_rate": 0.00019929137778311597, + "loss": 1.1101, + "step": 871 + }, + { + "epoch": 0.15527065527065528, + "grad_norm": 0.40345287322998047, + "learning_rate": 0.0001992897133929806, + "loss": 0.8894, + "step": 872 + }, + { + "epoch": 0.15544871794871795, + "grad_norm": 0.44931963086128235, + "learning_rate": 0.00019928804705746957, + "loss": 0.9389, + "step": 873 + }, + { + "epoch": 0.15562678062678062, + "grad_norm": 0.529196560382843, + "learning_rate": 0.0001992863787766155, + "loss": 1.3362, + "step": 874 + }, + { + "epoch": 0.15580484330484332, + "grad_norm": 0.41218671202659607, + "learning_rate": 0.0001992847085504511, + "loss": 1.0727, + "step": 875 + }, + { + "epoch": 0.15598290598290598, + "grad_norm": 0.44074541330337524, + "learning_rate": 0.00019928303637900907, + "loss": 1.1091, + "step": 876 + }, + { + "epoch": 0.15616096866096865, + "grad_norm": 0.5264310240745544, + "learning_rate": 0.00019928136226232218, + "loss": 1.201, + "step": 877 + }, + { + "epoch": 0.15633903133903135, + "grad_norm": 0.4255099594593048, + "learning_rate": 0.00019927968620042324, + "loss": 1.2514, + "step": 878 + }, + { + "epoch": 0.15651709401709402, + "grad_norm": 0.4030280113220215, + "learning_rate": 0.0001992780081933451, + "loss": 1.0422, + "step": 879 + }, + { + "epoch": 0.15669515669515668, + "grad_norm": 0.5270203948020935, + "learning_rate": 0.00019927632824112058, + "loss": 1.2476, + "step": 880 + }, + { + "epoch": 0.15687321937321938, + "grad_norm": 0.37767237424850464, + "learning_rate": 0.00019927464634378268, + "loss": 1.0768, + "step": 881 + }, + { + "epoch": 0.15705128205128205, + "grad_norm": 0.4535936415195465, + "learning_rate": 0.0001992729625013643, + "loss": 1.2097, + "step": 882 + }, + { + "epoch": 0.15722934472934472, + "grad_norm": 0.4282119870185852, + "learning_rate": 0.00019927127671389843, + "loss": 1.0904, + "step": 883 + }, + { + "epoch": 0.1574074074074074, + "grad_norm": 0.3924157917499542, + "learning_rate": 0.0001992695889814181, + "loss": 0.9692, + "step": 884 + }, + { + "epoch": 0.15758547008547008, + "grad_norm": 0.525075376033783, + "learning_rate": 0.0001992678993039564, + "loss": 1.0292, + "step": 885 + }, + { + "epoch": 0.15776353276353278, + "grad_norm": 0.4388505518436432, + "learning_rate": 0.00019926620768154644, + "loss": 1.1944, + "step": 886 + }, + { + "epoch": 0.15794159544159544, + "grad_norm": 0.4362235963344574, + "learning_rate": 0.00019926451411422132, + "loss": 0.97, + "step": 887 + }, + { + "epoch": 0.1581196581196581, + "grad_norm": 0.4265296459197998, + "learning_rate": 0.0001992628186020143, + "loss": 0.9196, + "step": 888 + }, + { + "epoch": 0.1582977207977208, + "grad_norm": 0.4019876718521118, + "learning_rate": 0.0001992611211449585, + "loss": 1.1368, + "step": 889 + }, + { + "epoch": 0.15847578347578348, + "grad_norm": 0.5003397464752197, + "learning_rate": 0.00019925942174308726, + "loss": 1.2582, + "step": 890 + }, + { + "epoch": 0.15865384615384615, + "grad_norm": 0.4774404466152191, + "learning_rate": 0.00019925772039643382, + "loss": 1.2277, + "step": 891 + }, + { + "epoch": 0.15883190883190884, + "grad_norm": 0.4590449333190918, + "learning_rate": 0.00019925601710503153, + "loss": 1.1679, + "step": 892 + }, + { + "epoch": 0.1590099715099715, + "grad_norm": 0.4221442639827728, + "learning_rate": 0.0001992543118689138, + "loss": 1.1626, + "step": 893 + }, + { + "epoch": 0.15918803418803418, + "grad_norm": 0.47613003849983215, + "learning_rate": 0.00019925260468811403, + "loss": 1.1509, + "step": 894 + }, + { + "epoch": 0.15936609686609687, + "grad_norm": 0.41706812381744385, + "learning_rate": 0.0001992508955626656, + "loss": 1.0366, + "step": 895 + }, + { + "epoch": 0.15954415954415954, + "grad_norm": 0.5064654350280762, + "learning_rate": 0.00019924918449260205, + "loss": 1.0729, + "step": 896 + }, + { + "epoch": 0.1597222222222222, + "grad_norm": 0.5019610524177551, + "learning_rate": 0.00019924747147795696, + "loss": 1.0642, + "step": 897 + }, + { + "epoch": 0.1599002849002849, + "grad_norm": 0.4345671534538269, + "learning_rate": 0.00019924575651876378, + "loss": 1.1747, + "step": 898 + }, + { + "epoch": 0.16007834757834757, + "grad_norm": 0.4397568702697754, + "learning_rate": 0.0001992440396150562, + "loss": 1.282, + "step": 899 + }, + { + "epoch": 0.16025641025641027, + "grad_norm": 0.520187497138977, + "learning_rate": 0.0001992423207668678, + "loss": 0.976, + "step": 900 + }, + { + "epoch": 0.16043447293447294, + "grad_norm": 0.39329993724823, + "learning_rate": 0.0001992405999742323, + "loss": 0.9829, + "step": 901 + }, + { + "epoch": 0.1606125356125356, + "grad_norm": 0.42361345887184143, + "learning_rate": 0.00019923887723718339, + "loss": 1.139, + "step": 902 + }, + { + "epoch": 0.1607905982905983, + "grad_norm": 0.3846314251422882, + "learning_rate": 0.00019923715255575482, + "loss": 0.8262, + "step": 903 + }, + { + "epoch": 0.16096866096866097, + "grad_norm": 0.39258381724357605, + "learning_rate": 0.0001992354259299804, + "loss": 0.9638, + "step": 904 + }, + { + "epoch": 0.16114672364672364, + "grad_norm": 0.4000850319862366, + "learning_rate": 0.00019923369735989397, + "loss": 0.91, + "step": 905 + }, + { + "epoch": 0.16132478632478633, + "grad_norm": 0.46303513646125793, + "learning_rate": 0.00019923196684552936, + "loss": 1.1447, + "step": 906 + }, + { + "epoch": 0.161502849002849, + "grad_norm": 0.38437438011169434, + "learning_rate": 0.0001992302343869205, + "loss": 1.0212, + "step": 907 + }, + { + "epoch": 0.16168091168091167, + "grad_norm": 0.44585472345352173, + "learning_rate": 0.00019922849998410135, + "loss": 1.1964, + "step": 908 + }, + { + "epoch": 0.16185897435897437, + "grad_norm": 0.41959813237190247, + "learning_rate": 0.00019922676363710583, + "loss": 0.9925, + "step": 909 + }, + { + "epoch": 0.16203703703703703, + "grad_norm": 0.47442761063575745, + "learning_rate": 0.00019922502534596803, + "loss": 0.9237, + "step": 910 + }, + { + "epoch": 0.1622150997150997, + "grad_norm": 0.5065128207206726, + "learning_rate": 0.00019922328511072198, + "loss": 1.2573, + "step": 911 + }, + { + "epoch": 0.1623931623931624, + "grad_norm": 0.4739879369735718, + "learning_rate": 0.0001992215429314018, + "loss": 1.4416, + "step": 912 + }, + { + "epoch": 0.16257122507122507, + "grad_norm": 0.48763832449913025, + "learning_rate": 0.00019921979880804157, + "loss": 1.0408, + "step": 913 + }, + { + "epoch": 0.16274928774928774, + "grad_norm": 0.4841614067554474, + "learning_rate": 0.0001992180527406755, + "loss": 1.1826, + "step": 914 + }, + { + "epoch": 0.16292735042735043, + "grad_norm": 0.49433308839797974, + "learning_rate": 0.0001992163047293378, + "loss": 1.3552, + "step": 915 + }, + { + "epoch": 0.1631054131054131, + "grad_norm": 0.4985002875328064, + "learning_rate": 0.0001992145547740627, + "loss": 1.2639, + "step": 916 + }, + { + "epoch": 0.1632834757834758, + "grad_norm": 0.40348032116889954, + "learning_rate": 0.00019921280287488448, + "loss": 1.1731, + "step": 917 + }, + { + "epoch": 0.16346153846153846, + "grad_norm": 0.5166002511978149, + "learning_rate": 0.0001992110490318375, + "loss": 1.0692, + "step": 918 + }, + { + "epoch": 0.16363960113960113, + "grad_norm": 0.44233468174934387, + "learning_rate": 0.00019920929324495615, + "loss": 1.0488, + "step": 919 + }, + { + "epoch": 0.16381766381766383, + "grad_norm": 0.43709903955459595, + "learning_rate": 0.00019920753551427476, + "loss": 0.8884, + "step": 920 + }, + { + "epoch": 0.1639957264957265, + "grad_norm": 0.4054167568683624, + "learning_rate": 0.00019920577583982778, + "loss": 0.9872, + "step": 921 + }, + { + "epoch": 0.16417378917378916, + "grad_norm": 0.4657362997531891, + "learning_rate": 0.0001992040142216497, + "loss": 1.4402, + "step": 922 + }, + { + "epoch": 0.16435185185185186, + "grad_norm": 0.42550426721572876, + "learning_rate": 0.0001992022506597751, + "loss": 1.0456, + "step": 923 + }, + { + "epoch": 0.16452991452991453, + "grad_norm": 0.49346762895584106, + "learning_rate": 0.00019920048515423842, + "loss": 1.527, + "step": 924 + }, + { + "epoch": 0.1647079772079772, + "grad_norm": 0.3970337510108948, + "learning_rate": 0.0001991987177050743, + "loss": 1.0363, + "step": 925 + }, + { + "epoch": 0.1648860398860399, + "grad_norm": 0.4027378559112549, + "learning_rate": 0.0001991969483123174, + "loss": 0.8416, + "step": 926 + }, + { + "epoch": 0.16506410256410256, + "grad_norm": 0.4181644916534424, + "learning_rate": 0.00019919517697600237, + "loss": 1.2253, + "step": 927 + }, + { + "epoch": 0.16524216524216523, + "grad_norm": 0.43686383962631226, + "learning_rate": 0.0001991934036961639, + "loss": 1.0808, + "step": 928 + }, + { + "epoch": 0.16542022792022792, + "grad_norm": 0.4242876172065735, + "learning_rate": 0.0001991916284728367, + "loss": 0.9483, + "step": 929 + }, + { + "epoch": 0.1655982905982906, + "grad_norm": 0.3690609037876129, + "learning_rate": 0.00019918985130605563, + "loss": 0.9495, + "step": 930 + }, + { + "epoch": 0.1657763532763533, + "grad_norm": 0.42184555530548096, + "learning_rate": 0.00019918807219585546, + "loss": 1.0966, + "step": 931 + }, + { + "epoch": 0.16595441595441596, + "grad_norm": 0.4342746138572693, + "learning_rate": 0.00019918629114227106, + "loss": 1.0875, + "step": 932 + }, + { + "epoch": 0.16613247863247863, + "grad_norm": 0.4191494286060333, + "learning_rate": 0.00019918450814533737, + "loss": 1.0777, + "step": 933 + }, + { + "epoch": 0.16631054131054132, + "grad_norm": 0.37124550342559814, + "learning_rate": 0.00019918272320508922, + "loss": 1.0131, + "step": 934 + }, + { + "epoch": 0.166488603988604, + "grad_norm": 0.4475722014904022, + "learning_rate": 0.00019918093632156168, + "loss": 1.1185, + "step": 935 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.4629058241844177, + "learning_rate": 0.0001991791474947897, + "loss": 1.0353, + "step": 936 + }, + { + "epoch": 0.16684472934472935, + "grad_norm": 0.48192909359931946, + "learning_rate": 0.00019917735672480834, + "loss": 1.1628, + "step": 937 + }, + { + "epoch": 0.16702279202279202, + "grad_norm": 0.5542252063751221, + "learning_rate": 0.00019917556401165273, + "loss": 1.3133, + "step": 938 + }, + { + "epoch": 0.1672008547008547, + "grad_norm": 0.4172651171684265, + "learning_rate": 0.00019917376935535796, + "loss": 1.1733, + "step": 939 + }, + { + "epoch": 0.16737891737891739, + "grad_norm": 0.4424920380115509, + "learning_rate": 0.0001991719727559592, + "loss": 1.0262, + "step": 940 + }, + { + "epoch": 0.16755698005698005, + "grad_norm": 0.4551742970943451, + "learning_rate": 0.00019917017421349162, + "loss": 1.0883, + "step": 941 + }, + { + "epoch": 0.16773504273504272, + "grad_norm": 0.45929640531539917, + "learning_rate": 0.00019916837372799048, + "loss": 1.1836, + "step": 942 + }, + { + "epoch": 0.16791310541310542, + "grad_norm": 0.4609353542327881, + "learning_rate": 0.0001991665712994911, + "loss": 1.0682, + "step": 943 + }, + { + "epoch": 0.16809116809116809, + "grad_norm": 0.42617303133010864, + "learning_rate": 0.00019916476692802873, + "loss": 1.074, + "step": 944 + }, + { + "epoch": 0.16826923076923078, + "grad_norm": 0.41919493675231934, + "learning_rate": 0.00019916296061363875, + "loss": 1.0969, + "step": 945 + }, + { + "epoch": 0.16844729344729345, + "grad_norm": 0.450979083776474, + "learning_rate": 0.00019916115235635656, + "loss": 1.1686, + "step": 946 + }, + { + "epoch": 0.16862535612535612, + "grad_norm": 0.42166751623153687, + "learning_rate": 0.00019915934215621758, + "loss": 0.9273, + "step": 947 + }, + { + "epoch": 0.16880341880341881, + "grad_norm": 0.4404160976409912, + "learning_rate": 0.00019915753001325729, + "loss": 1.1663, + "step": 948 + }, + { + "epoch": 0.16898148148148148, + "grad_norm": 0.42025226354599, + "learning_rate": 0.0001991557159275111, + "loss": 0.9433, + "step": 949 + }, + { + "epoch": 0.16915954415954415, + "grad_norm": 0.4277796745300293, + "learning_rate": 0.00019915389989901474, + "loss": 0.8475, + "step": 950 + }, + { + "epoch": 0.16933760683760685, + "grad_norm": 0.5162755250930786, + "learning_rate": 0.00019915208192780365, + "loss": 1.1155, + "step": 951 + }, + { + "epoch": 0.16951566951566951, + "grad_norm": 0.4214856028556824, + "learning_rate": 0.00019915026201391346, + "loss": 1.173, + "step": 952 + }, + { + "epoch": 0.16969373219373218, + "grad_norm": 0.4713292419910431, + "learning_rate": 0.00019914844015737985, + "loss": 1.1615, + "step": 953 + }, + { + "epoch": 0.16987179487179488, + "grad_norm": 0.461179256439209, + "learning_rate": 0.00019914661635823854, + "loss": 1.1169, + "step": 954 + }, + { + "epoch": 0.17004985754985755, + "grad_norm": 0.46200552582740784, + "learning_rate": 0.00019914479061652527, + "loss": 1.0274, + "step": 955 + }, + { + "epoch": 0.17022792022792022, + "grad_norm": 0.40968334674835205, + "learning_rate": 0.00019914296293227572, + "loss": 1.066, + "step": 956 + }, + { + "epoch": 0.1704059829059829, + "grad_norm": 0.40877434611320496, + "learning_rate": 0.0001991411333055258, + "loss": 1.1595, + "step": 957 + }, + { + "epoch": 0.17058404558404558, + "grad_norm": 0.42940187454223633, + "learning_rate": 0.00019913930173631132, + "loss": 1.0364, + "step": 958 + }, + { + "epoch": 0.17076210826210828, + "grad_norm": 0.49648910760879517, + "learning_rate": 0.00019913746822466819, + "loss": 1.0763, + "step": 959 + }, + { + "epoch": 0.17094017094017094, + "grad_norm": 0.4353426396846771, + "learning_rate": 0.00019913563277063228, + "loss": 0.9698, + "step": 960 + }, + { + "epoch": 0.1711182336182336, + "grad_norm": 0.45079681277275085, + "learning_rate": 0.00019913379537423958, + "loss": 1.2244, + "step": 961 + }, + { + "epoch": 0.1712962962962963, + "grad_norm": 0.4276828467845917, + "learning_rate": 0.00019913195603552607, + "loss": 0.9976, + "step": 962 + }, + { + "epoch": 0.17147435897435898, + "grad_norm": 0.41122403740882874, + "learning_rate": 0.00019913011475452785, + "loss": 1.0077, + "step": 963 + }, + { + "epoch": 0.17165242165242164, + "grad_norm": 0.43170276284217834, + "learning_rate": 0.00019912827153128096, + "loss": 1.1402, + "step": 964 + }, + { + "epoch": 0.17183048433048434, + "grad_norm": 0.37950268387794495, + "learning_rate": 0.0001991264263658215, + "loss": 0.9818, + "step": 965 + }, + { + "epoch": 0.172008547008547, + "grad_norm": 0.477333128452301, + "learning_rate": 0.00019912457925818562, + "loss": 1.1756, + "step": 966 + }, + { + "epoch": 0.17218660968660968, + "grad_norm": 0.4326401352882385, + "learning_rate": 0.00019912273020840954, + "loss": 1.3718, + "step": 967 + }, + { + "epoch": 0.17236467236467237, + "grad_norm": 0.37711042165756226, + "learning_rate": 0.00019912087921652945, + "loss": 0.9011, + "step": 968 + }, + { + "epoch": 0.17254273504273504, + "grad_norm": 0.50013667345047, + "learning_rate": 0.00019911902628258162, + "loss": 1.1163, + "step": 969 + }, + { + "epoch": 0.1727207977207977, + "grad_norm": 0.41913339495658875, + "learning_rate": 0.0001991171714066024, + "loss": 1.2614, + "step": 970 + }, + { + "epoch": 0.1728988603988604, + "grad_norm": 0.4075855612754822, + "learning_rate": 0.00019911531458862813, + "loss": 0.8984, + "step": 971 + }, + { + "epoch": 0.17307692307692307, + "grad_norm": 0.40277954936027527, + "learning_rate": 0.00019911345582869513, + "loss": 1.0851, + "step": 972 + }, + { + "epoch": 0.17325498575498577, + "grad_norm": 0.4312847852706909, + "learning_rate": 0.00019911159512683987, + "loss": 1.1273, + "step": 973 + }, + { + "epoch": 0.17343304843304844, + "grad_norm": 0.40303611755371094, + "learning_rate": 0.0001991097324830988, + "loss": 0.9645, + "step": 974 + }, + { + "epoch": 0.1736111111111111, + "grad_norm": 0.45560577511787415, + "learning_rate": 0.00019910786789750838, + "loss": 1.0864, + "step": 975 + }, + { + "epoch": 0.1737891737891738, + "grad_norm": 0.43775680661201477, + "learning_rate": 0.00019910600137010517, + "loss": 1.028, + "step": 976 + }, + { + "epoch": 0.17396723646723647, + "grad_norm": 0.3917224407196045, + "learning_rate": 0.00019910413290092572, + "loss": 1.0491, + "step": 977 + }, + { + "epoch": 0.17414529914529914, + "grad_norm": 0.4068751037120819, + "learning_rate": 0.0001991022624900067, + "loss": 1.0476, + "step": 978 + }, + { + "epoch": 0.17432336182336183, + "grad_norm": 0.4463370144367218, + "learning_rate": 0.0001991003901373847, + "loss": 1.0612, + "step": 979 + }, + { + "epoch": 0.1745014245014245, + "grad_norm": 0.46949052810668945, + "learning_rate": 0.0001990985158430964, + "loss": 1.3099, + "step": 980 + }, + { + "epoch": 0.17467948717948717, + "grad_norm": 0.4250012934207916, + "learning_rate": 0.00019909663960717856, + "loss": 0.9903, + "step": 981 + }, + { + "epoch": 0.17485754985754987, + "grad_norm": 0.5293903946876526, + "learning_rate": 0.0001990947614296679, + "loss": 0.9908, + "step": 982 + }, + { + "epoch": 0.17503561253561253, + "grad_norm": 0.3838284909725189, + "learning_rate": 0.0001990928813106013, + "loss": 0.716, + "step": 983 + }, + { + "epoch": 0.1752136752136752, + "grad_norm": 0.4597751200199127, + "learning_rate": 0.0001990909992500155, + "loss": 1.0126, + "step": 984 + }, + { + "epoch": 0.1753917378917379, + "grad_norm": 0.4844081699848175, + "learning_rate": 0.0001990891152479474, + "loss": 1.1043, + "step": 985 + }, + { + "epoch": 0.17556980056980057, + "grad_norm": 0.4763399660587311, + "learning_rate": 0.00019908722930443392, + "loss": 1.019, + "step": 986 + }, + { + "epoch": 0.17574786324786323, + "grad_norm": 0.4670077860355377, + "learning_rate": 0.00019908534141951204, + "loss": 1.1382, + "step": 987 + }, + { + "epoch": 0.17592592592592593, + "grad_norm": 0.39372730255126953, + "learning_rate": 0.00019908345159321873, + "loss": 1.1219, + "step": 988 + }, + { + "epoch": 0.1761039886039886, + "grad_norm": 0.41869843006134033, + "learning_rate": 0.00019908155982559098, + "loss": 0.9461, + "step": 989 + }, + { + "epoch": 0.1762820512820513, + "grad_norm": 0.4398406147956848, + "learning_rate": 0.00019907966611666593, + "loss": 1.1328, + "step": 990 + }, + { + "epoch": 0.17646011396011396, + "grad_norm": 0.4315733015537262, + "learning_rate": 0.0001990777704664806, + "loss": 1.0974, + "step": 991 + }, + { + "epoch": 0.17663817663817663, + "grad_norm": 0.42859575152397156, + "learning_rate": 0.00019907587287507222, + "loss": 1.2637, + "step": 992 + }, + { + "epoch": 0.17681623931623933, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.0001990739733424779, + "loss": 1.0699, + "step": 993 + }, + { + "epoch": 0.176994301994302, + "grad_norm": 0.4443826973438263, + "learning_rate": 0.00019907207186873488, + "loss": 1.0547, + "step": 994 + }, + { + "epoch": 0.17717236467236466, + "grad_norm": 0.4108099937438965, + "learning_rate": 0.00019907016845388043, + "loss": 1.1401, + "step": 995 + }, + { + "epoch": 0.17735042735042736, + "grad_norm": 0.4474675953388214, + "learning_rate": 0.00019906826309795182, + "loss": 1.0712, + "step": 996 + }, + { + "epoch": 0.17752849002849003, + "grad_norm": 0.4149756133556366, + "learning_rate": 0.00019906635580098638, + "loss": 0.9585, + "step": 997 + }, + { + "epoch": 0.1777065527065527, + "grad_norm": 0.4875968098640442, + "learning_rate": 0.00019906444656302152, + "loss": 1.0659, + "step": 998 + }, + { + "epoch": 0.1778846153846154, + "grad_norm": 0.5494784116744995, + "learning_rate": 0.0001990625353840946, + "loss": 1.2858, + "step": 999 + }, + { + "epoch": 0.17806267806267806, + "grad_norm": 0.425062358379364, + "learning_rate": 0.0001990606222642431, + "loss": 1.1826, + "step": 1000 + }, + { + "epoch": 0.17824074074074073, + "grad_norm": 0.3890725374221802, + "learning_rate": 0.00019905870720350445, + "loss": 0.9568, + "step": 1001 + }, + { + "epoch": 0.17841880341880342, + "grad_norm": 0.3884070813655853, + "learning_rate": 0.00019905679020191624, + "loss": 0.9674, + "step": 1002 + }, + { + "epoch": 0.1785968660968661, + "grad_norm": 0.49496129155158997, + "learning_rate": 0.00019905487125951597, + "loss": 0.9143, + "step": 1003 + }, + { + "epoch": 0.1787749287749288, + "grad_norm": 0.43448135256767273, + "learning_rate": 0.00019905295037634128, + "loss": 1.2677, + "step": 1004 + }, + { + "epoch": 0.17895299145299146, + "grad_norm": 0.47327905893325806, + "learning_rate": 0.00019905102755242982, + "loss": 0.9089, + "step": 1005 + }, + { + "epoch": 0.17913105413105412, + "grad_norm": 0.4962378442287445, + "learning_rate": 0.00019904910278781922, + "loss": 1.1748, + "step": 1006 + }, + { + "epoch": 0.17930911680911682, + "grad_norm": 0.4343934655189514, + "learning_rate": 0.0001990471760825472, + "loss": 1.2176, + "step": 1007 + }, + { + "epoch": 0.1794871794871795, + "grad_norm": 0.4695793092250824, + "learning_rate": 0.0001990452474366515, + "loss": 1.1822, + "step": 1008 + }, + { + "epoch": 0.17966524216524216, + "grad_norm": 0.4156060516834259, + "learning_rate": 0.00019904331685016995, + "loss": 0.8231, + "step": 1009 + }, + { + "epoch": 0.17984330484330485, + "grad_norm": 0.5068191885948181, + "learning_rate": 0.00019904138432314035, + "loss": 1.1363, + "step": 1010 + }, + { + "epoch": 0.18002136752136752, + "grad_norm": 0.5189786553382874, + "learning_rate": 0.00019903944985560058, + "loss": 1.3131, + "step": 1011 + }, + { + "epoch": 0.1801994301994302, + "grad_norm": 0.5126828551292419, + "learning_rate": 0.00019903751344758848, + "loss": 1.0305, + "step": 1012 + }, + { + "epoch": 0.18037749287749288, + "grad_norm": 0.41045933961868286, + "learning_rate": 0.00019903557509914205, + "loss": 1.2726, + "step": 1013 + }, + { + "epoch": 0.18055555555555555, + "grad_norm": 0.4141713082790375, + "learning_rate": 0.0001990336348102993, + "loss": 0.9606, + "step": 1014 + }, + { + "epoch": 0.18073361823361822, + "grad_norm": 0.42652079463005066, + "learning_rate": 0.00019903169258109812, + "loss": 1.0235, + "step": 1015 + }, + { + "epoch": 0.18091168091168092, + "grad_norm": 0.42098379135131836, + "learning_rate": 0.0001990297484115767, + "loss": 1.0602, + "step": 1016 + }, + { + "epoch": 0.18108974358974358, + "grad_norm": 0.49920013546943665, + "learning_rate": 0.0001990278023017731, + "loss": 1.3322, + "step": 1017 + }, + { + "epoch": 0.18126780626780628, + "grad_norm": 0.412304550409317, + "learning_rate": 0.00019902585425172537, + "loss": 1.1011, + "step": 1018 + }, + { + "epoch": 0.18144586894586895, + "grad_norm": 0.44226935505867004, + "learning_rate": 0.00019902390426147177, + "loss": 0.9777, + "step": 1019 + }, + { + "epoch": 0.18162393162393162, + "grad_norm": 0.4685269594192505, + "learning_rate": 0.00019902195233105046, + "loss": 1.3587, + "step": 1020 + }, + { + "epoch": 0.1818019943019943, + "grad_norm": 0.4500584304332733, + "learning_rate": 0.00019901999846049968, + "loss": 0.9888, + "step": 1021 + }, + { + "epoch": 0.18198005698005698, + "grad_norm": 0.48566994071006775, + "learning_rate": 0.00019901804264985774, + "loss": 1.2364, + "step": 1022 + }, + { + "epoch": 0.18215811965811965, + "grad_norm": 0.4063156247138977, + "learning_rate": 0.00019901608489916294, + "loss": 1.2224, + "step": 1023 + }, + { + "epoch": 0.18233618233618235, + "grad_norm": 0.471276193857193, + "learning_rate": 0.00019901412520845367, + "loss": 0.9926, + "step": 1024 + }, + { + "epoch": 0.182514245014245, + "grad_norm": 0.5165421366691589, + "learning_rate": 0.00019901216357776829, + "loss": 0.9595, + "step": 1025 + }, + { + "epoch": 0.18269230769230768, + "grad_norm": 0.4746754467487335, + "learning_rate": 0.0001990102000071452, + "loss": 1.2057, + "step": 1026 + }, + { + "epoch": 0.18287037037037038, + "grad_norm": 0.44803035259246826, + "learning_rate": 0.00019900823449662297, + "loss": 1.2114, + "step": 1027 + }, + { + "epoch": 0.18304843304843305, + "grad_norm": 0.47256240248680115, + "learning_rate": 0.00019900626704624005, + "loss": 1.112, + "step": 1028 + }, + { + "epoch": 0.18322649572649571, + "grad_norm": 0.4253387153148651, + "learning_rate": 0.000199004297656035, + "loss": 0.9899, + "step": 1029 + }, + { + "epoch": 0.1834045584045584, + "grad_norm": 0.44958099722862244, + "learning_rate": 0.00019900232632604636, + "loss": 1.1445, + "step": 1030 + }, + { + "epoch": 0.18358262108262108, + "grad_norm": 0.5296537280082703, + "learning_rate": 0.00019900035305631285, + "loss": 1.2502, + "step": 1031 + }, + { + "epoch": 0.18376068376068377, + "grad_norm": 0.5057148933410645, + "learning_rate": 0.00019899837784687302, + "loss": 1.1426, + "step": 1032 + }, + { + "epoch": 0.18393874643874644, + "grad_norm": 0.41463762521743774, + "learning_rate": 0.00019899640069776566, + "loss": 1.1854, + "step": 1033 + }, + { + "epoch": 0.1841168091168091, + "grad_norm": 0.45800045132637024, + "learning_rate": 0.00019899442160902945, + "loss": 1.2438, + "step": 1034 + }, + { + "epoch": 0.1842948717948718, + "grad_norm": 0.43450453877449036, + "learning_rate": 0.00019899244058070324, + "loss": 1.0598, + "step": 1035 + }, + { + "epoch": 0.18447293447293447, + "grad_norm": 0.4141148626804352, + "learning_rate": 0.00019899045761282577, + "loss": 1.0465, + "step": 1036 + }, + { + "epoch": 0.18465099715099714, + "grad_norm": 0.3938458263874054, + "learning_rate": 0.0001989884727054359, + "loss": 1.0142, + "step": 1037 + }, + { + "epoch": 0.18482905982905984, + "grad_norm": 0.43898263573646545, + "learning_rate": 0.00019898648585857257, + "loss": 0.9212, + "step": 1038 + }, + { + "epoch": 0.1850071225071225, + "grad_norm": 0.4425487816333771, + "learning_rate": 0.00019898449707227465, + "loss": 1.2987, + "step": 1039 + }, + { + "epoch": 0.18518518518518517, + "grad_norm": 0.4537975490093231, + "learning_rate": 0.00019898250634658115, + "loss": 1.2023, + "step": 1040 + }, + { + "epoch": 0.18536324786324787, + "grad_norm": 0.4107198119163513, + "learning_rate": 0.00019898051368153104, + "loss": 0.8443, + "step": 1041 + }, + { + "epoch": 0.18554131054131054, + "grad_norm": 0.4389404058456421, + "learning_rate": 0.0001989785190771634, + "loss": 1.0502, + "step": 1042 + }, + { + "epoch": 0.1857193732193732, + "grad_norm": 0.4288824796676636, + "learning_rate": 0.00019897652253351726, + "loss": 1.01, + "step": 1043 + }, + { + "epoch": 0.1858974358974359, + "grad_norm": 0.50815349817276, + "learning_rate": 0.00019897452405063178, + "loss": 1.0308, + "step": 1044 + }, + { + "epoch": 0.18607549857549857, + "grad_norm": 0.45252710580825806, + "learning_rate": 0.0001989725236285461, + "loss": 1.0967, + "step": 1045 + }, + { + "epoch": 0.18625356125356127, + "grad_norm": 0.45049402117729187, + "learning_rate": 0.00019897052126729943, + "loss": 1.0141, + "step": 1046 + }, + { + "epoch": 0.18643162393162394, + "grad_norm": 0.49637508392333984, + "learning_rate": 0.00019896851696693098, + "loss": 1.0997, + "step": 1047 + }, + { + "epoch": 0.1866096866096866, + "grad_norm": 0.4465886056423187, + "learning_rate": 0.00019896651072748005, + "loss": 1.1415, + "step": 1048 + }, + { + "epoch": 0.1867877492877493, + "grad_norm": 0.5309500694274902, + "learning_rate": 0.00019896450254898592, + "loss": 1.1028, + "step": 1049 + }, + { + "epoch": 0.18696581196581197, + "grad_norm": 0.3516653776168823, + "learning_rate": 0.00019896249243148793, + "loss": 0.9841, + "step": 1050 + }, + { + "epoch": 0.18714387464387464, + "grad_norm": 0.4529176950454712, + "learning_rate": 0.0001989604803750255, + "loss": 1.1335, + "step": 1051 + }, + { + "epoch": 0.18732193732193733, + "grad_norm": 0.47694942355155945, + "learning_rate": 0.000198958466379638, + "loss": 1.2383, + "step": 1052 + }, + { + "epoch": 0.1875, + "grad_norm": 0.5524206757545471, + "learning_rate": 0.0001989564504453649, + "loss": 1.3668, + "step": 1053 + }, + { + "epoch": 0.18767806267806267, + "grad_norm": 0.39203691482543945, + "learning_rate": 0.00019895443257224576, + "loss": 1.2203, + "step": 1054 + }, + { + "epoch": 0.18785612535612536, + "grad_norm": 0.4164120852947235, + "learning_rate": 0.00019895241276032005, + "loss": 0.8954, + "step": 1055 + }, + { + "epoch": 0.18803418803418803, + "grad_norm": 0.41217970848083496, + "learning_rate": 0.0001989503910096274, + "loss": 1.0238, + "step": 1056 + }, + { + "epoch": 0.1882122507122507, + "grad_norm": 0.44038307666778564, + "learning_rate": 0.00019894836732020735, + "loss": 0.8159, + "step": 1057 + }, + { + "epoch": 0.1883903133903134, + "grad_norm": 0.45780670642852783, + "learning_rate": 0.0001989463416920996, + "loss": 1.2864, + "step": 1058 + }, + { + "epoch": 0.18856837606837606, + "grad_norm": 0.5197559595108032, + "learning_rate": 0.00019894431412534384, + "loss": 1.0756, + "step": 1059 + }, + { + "epoch": 0.18874643874643873, + "grad_norm": 0.43283385038375854, + "learning_rate": 0.00019894228461997979, + "loss": 1.0642, + "step": 1060 + }, + { + "epoch": 0.18892450142450143, + "grad_norm": 0.4657376706600189, + "learning_rate": 0.00019894025317604717, + "loss": 1.1159, + "step": 1061 + }, + { + "epoch": 0.1891025641025641, + "grad_norm": 0.4474908113479614, + "learning_rate": 0.00019893821979358588, + "loss": 1.2006, + "step": 1062 + }, + { + "epoch": 0.1892806267806268, + "grad_norm": 0.43878164887428284, + "learning_rate": 0.00019893618447263566, + "loss": 1.1599, + "step": 1063 + }, + { + "epoch": 0.18945868945868946, + "grad_norm": 0.4598735272884369, + "learning_rate": 0.00019893414721323645, + "loss": 1.3346, + "step": 1064 + }, + { + "epoch": 0.18963675213675213, + "grad_norm": 0.3947420120239258, + "learning_rate": 0.00019893210801542812, + "loss": 1.1201, + "step": 1065 + }, + { + "epoch": 0.18981481481481483, + "grad_norm": 0.3401558995246887, + "learning_rate": 0.00019893006687925064, + "loss": 0.7568, + "step": 1066 + }, + { + "epoch": 0.1899928774928775, + "grad_norm": 0.4400341808795929, + "learning_rate": 0.00019892802380474405, + "loss": 1.1706, + "step": 1067 + }, + { + "epoch": 0.19017094017094016, + "grad_norm": 0.42394164204597473, + "learning_rate": 0.00019892597879194829, + "loss": 1.0163, + "step": 1068 + }, + { + "epoch": 0.19034900284900286, + "grad_norm": 0.42904096841812134, + "learning_rate": 0.00019892393184090353, + "loss": 0.9193, + "step": 1069 + }, + { + "epoch": 0.19052706552706553, + "grad_norm": 0.497601181268692, + "learning_rate": 0.00019892188295164977, + "loss": 1.0377, + "step": 1070 + }, + { + "epoch": 0.1907051282051282, + "grad_norm": 0.4536020755767822, + "learning_rate": 0.00019891983212422723, + "loss": 1.0946, + "step": 1071 + }, + { + "epoch": 0.1908831908831909, + "grad_norm": 0.44916942715644836, + "learning_rate": 0.00019891777935867607, + "loss": 1.0563, + "step": 1072 + }, + { + "epoch": 0.19106125356125356, + "grad_norm": 0.4256889820098877, + "learning_rate": 0.0001989157246550365, + "loss": 1.0988, + "step": 1073 + }, + { + "epoch": 0.19123931623931623, + "grad_norm": 0.5559163689613342, + "learning_rate": 0.0001989136680133488, + "loss": 0.9155, + "step": 1074 + }, + { + "epoch": 0.19141737891737892, + "grad_norm": 0.391804963350296, + "learning_rate": 0.00019891160943365322, + "loss": 0.9314, + "step": 1075 + }, + { + "epoch": 0.1915954415954416, + "grad_norm": 0.4535716474056244, + "learning_rate": 0.00019890954891599015, + "loss": 1.0768, + "step": 1076 + }, + { + "epoch": 0.19177350427350429, + "grad_norm": 0.46770521998405457, + "learning_rate": 0.00019890748646039991, + "loss": 0.8406, + "step": 1077 + }, + { + "epoch": 0.19195156695156695, + "grad_norm": 0.4875394403934479, + "learning_rate": 0.00019890542206692295, + "loss": 1.1055, + "step": 1078 + }, + { + "epoch": 0.19212962962962962, + "grad_norm": 0.5072727203369141, + "learning_rate": 0.0001989033557355997, + "loss": 1.3093, + "step": 1079 + }, + { + "epoch": 0.19230769230769232, + "grad_norm": 0.4419287443161011, + "learning_rate": 0.00019890128746647068, + "loss": 1.1916, + "step": 1080 + }, + { + "epoch": 0.192485754985755, + "grad_norm": 0.45803651213645935, + "learning_rate": 0.00019889921725957637, + "loss": 1.2579, + "step": 1081 + }, + { + "epoch": 0.19266381766381765, + "grad_norm": 0.4832262098789215, + "learning_rate": 0.0001988971451149573, + "loss": 1.3217, + "step": 1082 + }, + { + "epoch": 0.19284188034188035, + "grad_norm": 0.4819786250591278, + "learning_rate": 0.00019889507103265416, + "loss": 1.0979, + "step": 1083 + }, + { + "epoch": 0.19301994301994302, + "grad_norm": 0.49360713362693787, + "learning_rate": 0.0001988929950127075, + "loss": 1.0987, + "step": 1084 + }, + { + "epoch": 0.1931980056980057, + "grad_norm": 0.44209200143814087, + "learning_rate": 0.00019889091705515806, + "loss": 1.2616, + "step": 1085 + }, + { + "epoch": 0.19337606837606838, + "grad_norm": 0.41626206040382385, + "learning_rate": 0.00019888883716004654, + "loss": 1.0922, + "step": 1086 + }, + { + "epoch": 0.19355413105413105, + "grad_norm": 0.4916635751724243, + "learning_rate": 0.00019888675532741366, + "loss": 0.9331, + "step": 1087 + }, + { + "epoch": 0.19373219373219372, + "grad_norm": 0.4493125379085541, + "learning_rate": 0.00019888467155730025, + "loss": 1.1261, + "step": 1088 + }, + { + "epoch": 0.19391025641025642, + "grad_norm": 0.3755671977996826, + "learning_rate": 0.00019888258584974708, + "loss": 0.9821, + "step": 1089 + }, + { + "epoch": 0.19408831908831908, + "grad_norm": 0.41917556524276733, + "learning_rate": 0.00019888049820479507, + "loss": 1.251, + "step": 1090 + }, + { + "epoch": 0.19426638176638178, + "grad_norm": 0.46184420585632324, + "learning_rate": 0.0001988784086224851, + "loss": 1.1731, + "step": 1091 + }, + { + "epoch": 0.19444444444444445, + "grad_norm": 0.4783691465854645, + "learning_rate": 0.00019887631710285812, + "loss": 1.1635, + "step": 1092 + }, + { + "epoch": 0.19462250712250712, + "grad_norm": 0.4710482060909271, + "learning_rate": 0.00019887422364595512, + "loss": 1.0229, + "step": 1093 + }, + { + "epoch": 0.1948005698005698, + "grad_norm": 0.4738706648349762, + "learning_rate": 0.00019887212825181707, + "loss": 1.128, + "step": 1094 + }, + { + "epoch": 0.19497863247863248, + "grad_norm": 0.45665010809898376, + "learning_rate": 0.00019887003092048508, + "loss": 1.0425, + "step": 1095 + }, + { + "epoch": 0.19515669515669515, + "grad_norm": 0.42740485072135925, + "learning_rate": 0.0001988679316520002, + "loss": 1.0738, + "step": 1096 + }, + { + "epoch": 0.19533475783475784, + "grad_norm": 0.5977092385292053, + "learning_rate": 0.0001988658304464036, + "loss": 1.2687, + "step": 1097 + }, + { + "epoch": 0.1955128205128205, + "grad_norm": 0.4411074221134186, + "learning_rate": 0.0001988637273037364, + "loss": 1.287, + "step": 1098 + }, + { + "epoch": 0.19569088319088318, + "grad_norm": 0.4409518539905548, + "learning_rate": 0.00019886162222403986, + "loss": 1.0515, + "step": 1099 + }, + { + "epoch": 0.19586894586894588, + "grad_norm": 0.4926736652851105, + "learning_rate": 0.0001988595152073552, + "loss": 1.1388, + "step": 1100 + }, + { + "epoch": 0.19604700854700854, + "grad_norm": 0.4607115387916565, + "learning_rate": 0.00019885740625372368, + "loss": 0.9803, + "step": 1101 + }, + { + "epoch": 0.1962250712250712, + "grad_norm": 0.4725342094898224, + "learning_rate": 0.0001988552953631867, + "loss": 1.199, + "step": 1102 + }, + { + "epoch": 0.1964031339031339, + "grad_norm": 0.48014503717422485, + "learning_rate": 0.00019885318253578548, + "loss": 1.1868, + "step": 1103 + }, + { + "epoch": 0.19658119658119658, + "grad_norm": 0.3872644603252411, + "learning_rate": 0.00019885106777156155, + "loss": 0.9182, + "step": 1104 + }, + { + "epoch": 0.19675925925925927, + "grad_norm": 0.4737720787525177, + "learning_rate": 0.00019884895107055627, + "loss": 1.1513, + "step": 1105 + }, + { + "epoch": 0.19693732193732194, + "grad_norm": 0.4144562780857086, + "learning_rate": 0.00019884683243281116, + "loss": 1.1711, + "step": 1106 + }, + { + "epoch": 0.1971153846153846, + "grad_norm": 0.4672079384326935, + "learning_rate": 0.00019884471185836769, + "loss": 1.0386, + "step": 1107 + }, + { + "epoch": 0.1972934472934473, + "grad_norm": 0.4558824598789215, + "learning_rate": 0.0001988425893472674, + "loss": 1.0535, + "step": 1108 + }, + { + "epoch": 0.19747150997150997, + "grad_norm": 0.5149834752082825, + "learning_rate": 0.00019884046489955192, + "loss": 1.0296, + "step": 1109 + }, + { + "epoch": 0.19764957264957264, + "grad_norm": 0.43444496393203735, + "learning_rate": 0.00019883833851526287, + "loss": 1.1475, + "step": 1110 + }, + { + "epoch": 0.19782763532763534, + "grad_norm": 0.46062374114990234, + "learning_rate": 0.00019883621019444188, + "loss": 1.183, + "step": 1111 + }, + { + "epoch": 0.198005698005698, + "grad_norm": 0.4893282949924469, + "learning_rate": 0.00019883407993713065, + "loss": 1.3733, + "step": 1112 + }, + { + "epoch": 0.19818376068376067, + "grad_norm": 0.5434843897819519, + "learning_rate": 0.00019883194774337096, + "loss": 1.2505, + "step": 1113 + }, + { + "epoch": 0.19836182336182337, + "grad_norm": 0.4698035418987274, + "learning_rate": 0.00019882981361320456, + "loss": 1.0152, + "step": 1114 + }, + { + "epoch": 0.19853988603988604, + "grad_norm": 0.4582163989543915, + "learning_rate": 0.00019882767754667325, + "loss": 1.1718, + "step": 1115 + }, + { + "epoch": 0.1987179487179487, + "grad_norm": 0.48744696378707886, + "learning_rate": 0.0001988255395438189, + "loss": 1.2923, + "step": 1116 + }, + { + "epoch": 0.1988960113960114, + "grad_norm": 0.4172030985355377, + "learning_rate": 0.0001988233996046834, + "loss": 0.8098, + "step": 1117 + }, + { + "epoch": 0.19907407407407407, + "grad_norm": 0.4556557834148407, + "learning_rate": 0.00019882125772930867, + "loss": 0.9654, + "step": 1118 + }, + { + "epoch": 0.19925213675213677, + "grad_norm": 0.4363219141960144, + "learning_rate": 0.00019881911391773666, + "loss": 1.0333, + "step": 1119 + }, + { + "epoch": 0.19943019943019943, + "grad_norm": 0.4336536228656769, + "learning_rate": 0.0001988169681700094, + "loss": 1.091, + "step": 1120 + }, + { + "epoch": 0.1996082621082621, + "grad_norm": 0.42073166370391846, + "learning_rate": 0.00019881482048616893, + "loss": 0.9687, + "step": 1121 + }, + { + "epoch": 0.1997863247863248, + "grad_norm": 0.4330587685108185, + "learning_rate": 0.00019881267086625733, + "loss": 1.0512, + "step": 1122 + }, + { + "epoch": 0.19996438746438747, + "grad_norm": 0.4602276682853699, + "learning_rate": 0.0001988105193103167, + "loss": 1.1806, + "step": 1123 + }, + { + "epoch": 0.20014245014245013, + "grad_norm": 0.4271257817745209, + "learning_rate": 0.0001988083658183892, + "loss": 1.1079, + "step": 1124 + }, + { + "epoch": 0.20032051282051283, + "grad_norm": 0.35446426272392273, + "learning_rate": 0.00019880621039051707, + "loss": 0.6769, + "step": 1125 + }, + { + "epoch": 0.2004985754985755, + "grad_norm": 0.413753479719162, + "learning_rate": 0.00019880405302674244, + "loss": 1.1088, + "step": 1126 + }, + { + "epoch": 0.20067663817663817, + "grad_norm": 0.4423675835132599, + "learning_rate": 0.00019880189372710767, + "loss": 1.1371, + "step": 1127 + }, + { + "epoch": 0.20085470085470086, + "grad_norm": 0.41865605115890503, + "learning_rate": 0.00019879973249165502, + "loss": 1.0027, + "step": 1128 + }, + { + "epoch": 0.20103276353276353, + "grad_norm": 0.4109594225883484, + "learning_rate": 0.00019879756932042686, + "loss": 0.8734, + "step": 1129 + }, + { + "epoch": 0.2012108262108262, + "grad_norm": 0.42326363921165466, + "learning_rate": 0.00019879540421346555, + "loss": 0.9722, + "step": 1130 + }, + { + "epoch": 0.2013888888888889, + "grad_norm": 0.4601542055606842, + "learning_rate": 0.00019879323717081354, + "loss": 1.1251, + "step": 1131 + }, + { + "epoch": 0.20156695156695156, + "grad_norm": 0.4704367518424988, + "learning_rate": 0.00019879106819251327, + "loss": 0.9457, + "step": 1132 + }, + { + "epoch": 0.20174501424501423, + "grad_norm": 0.465023934841156, + "learning_rate": 0.00019878889727860724, + "loss": 0.9633, + "step": 1133 + }, + { + "epoch": 0.20192307692307693, + "grad_norm": 0.4572450518608093, + "learning_rate": 0.00019878672442913796, + "loss": 1.1965, + "step": 1134 + }, + { + "epoch": 0.2021011396011396, + "grad_norm": 0.4323410391807556, + "learning_rate": 0.00019878454964414807, + "loss": 1.1296, + "step": 1135 + }, + { + "epoch": 0.2022792022792023, + "grad_norm": 0.4513751268386841, + "learning_rate": 0.00019878237292368013, + "loss": 1.0571, + "step": 1136 + }, + { + "epoch": 0.20245726495726496, + "grad_norm": 0.45504096150398254, + "learning_rate": 0.00019878019426777677, + "loss": 1.0316, + "step": 1137 + }, + { + "epoch": 0.20263532763532763, + "grad_norm": 0.45715275406837463, + "learning_rate": 0.0001987780136764807, + "loss": 1.0528, + "step": 1138 + }, + { + "epoch": 0.20281339031339032, + "grad_norm": 0.4934465289115906, + "learning_rate": 0.00019877583114983466, + "loss": 1.3238, + "step": 1139 + }, + { + "epoch": 0.202991452991453, + "grad_norm": 0.4304082989692688, + "learning_rate": 0.0001987736466878814, + "loss": 1.1774, + "step": 1140 + }, + { + "epoch": 0.20316951566951566, + "grad_norm": 0.49721968173980713, + "learning_rate": 0.00019877146029066372, + "loss": 1.1767, + "step": 1141 + }, + { + "epoch": 0.20334757834757836, + "grad_norm": 0.3629468083381653, + "learning_rate": 0.00019876927195822445, + "loss": 0.8588, + "step": 1142 + }, + { + "epoch": 0.20352564102564102, + "grad_norm": 0.49310383200645447, + "learning_rate": 0.00019876708169060648, + "loss": 1.0588, + "step": 1143 + }, + { + "epoch": 0.2037037037037037, + "grad_norm": 0.4270328879356384, + "learning_rate": 0.00019876488948785271, + "loss": 1.1523, + "step": 1144 + }, + { + "epoch": 0.2038817663817664, + "grad_norm": 0.4559730887413025, + "learning_rate": 0.0001987626953500061, + "loss": 1.1736, + "step": 1145 + }, + { + "epoch": 0.20405982905982906, + "grad_norm": 0.5335259437561035, + "learning_rate": 0.00019876049927710962, + "loss": 0.991, + "step": 1146 + }, + { + "epoch": 0.20423789173789172, + "grad_norm": 0.43500083684921265, + "learning_rate": 0.0001987583012692063, + "loss": 1.0631, + "step": 1147 + }, + { + "epoch": 0.20441595441595442, + "grad_norm": 0.4135417938232422, + "learning_rate": 0.00019875610132633927, + "loss": 1.0896, + "step": 1148 + }, + { + "epoch": 0.2045940170940171, + "grad_norm": 0.4078896641731262, + "learning_rate": 0.00019875389944855153, + "loss": 1.0395, + "step": 1149 + }, + { + "epoch": 0.20477207977207978, + "grad_norm": 0.46612194180488586, + "learning_rate": 0.00019875169563588632, + "loss": 1.0541, + "step": 1150 + }, + { + "epoch": 0.20495014245014245, + "grad_norm": 0.5093224048614502, + "learning_rate": 0.00019874948988838674, + "loss": 1.1486, + "step": 1151 + }, + { + "epoch": 0.20512820512820512, + "grad_norm": 0.5079755187034607, + "learning_rate": 0.00019874728220609607, + "loss": 1.2614, + "step": 1152 + }, + { + "epoch": 0.20530626780626782, + "grad_norm": 0.43663498759269714, + "learning_rate": 0.0001987450725890575, + "loss": 1.0683, + "step": 1153 + }, + { + "epoch": 0.20548433048433049, + "grad_norm": 0.5029327273368835, + "learning_rate": 0.00019874286103731435, + "loss": 1.1934, + "step": 1154 + }, + { + "epoch": 0.20566239316239315, + "grad_norm": 0.48770397901535034, + "learning_rate": 0.00019874064755090999, + "loss": 1.1634, + "step": 1155 + }, + { + "epoch": 0.20584045584045585, + "grad_norm": 0.46826690435409546, + "learning_rate": 0.00019873843212988776, + "loss": 1.0621, + "step": 1156 + }, + { + "epoch": 0.20601851851851852, + "grad_norm": 0.4810047149658203, + "learning_rate": 0.00019873621477429105, + "loss": 1.0879, + "step": 1157 + }, + { + "epoch": 0.20619658119658119, + "grad_norm": 0.4769522249698639, + "learning_rate": 0.00019873399548416335, + "loss": 1.1365, + "step": 1158 + }, + { + "epoch": 0.20637464387464388, + "grad_norm": 0.4221782982349396, + "learning_rate": 0.00019873177425954806, + "loss": 1.1168, + "step": 1159 + }, + { + "epoch": 0.20655270655270655, + "grad_norm": 0.4084923565387726, + "learning_rate": 0.00019872955110048876, + "loss": 1.2364, + "step": 1160 + }, + { + "epoch": 0.20673076923076922, + "grad_norm": 0.4781704545021057, + "learning_rate": 0.00019872732600702904, + "loss": 1.19, + "step": 1161 + }, + { + "epoch": 0.20690883190883191, + "grad_norm": 0.3984242081642151, + "learning_rate": 0.0001987250989792124, + "loss": 1.0568, + "step": 1162 + }, + { + "epoch": 0.20708689458689458, + "grad_norm": 0.4601972997188568, + "learning_rate": 0.00019872287001708257, + "loss": 1.1625, + "step": 1163 + }, + { + "epoch": 0.20726495726495728, + "grad_norm": 0.4853581190109253, + "learning_rate": 0.00019872063912068316, + "loss": 1.2304, + "step": 1164 + }, + { + "epoch": 0.20744301994301995, + "grad_norm": 0.41779839992523193, + "learning_rate": 0.0001987184062900579, + "loss": 0.9807, + "step": 1165 + }, + { + "epoch": 0.20762108262108261, + "grad_norm": 0.4945356249809265, + "learning_rate": 0.00019871617152525056, + "loss": 1.1861, + "step": 1166 + }, + { + "epoch": 0.2077991452991453, + "grad_norm": 0.47432294487953186, + "learning_rate": 0.00019871393482630487, + "loss": 1.1448, + "step": 1167 + }, + { + "epoch": 0.20797720797720798, + "grad_norm": 0.44647398591041565, + "learning_rate": 0.00019871169619326473, + "loss": 1.096, + "step": 1168 + }, + { + "epoch": 0.20815527065527065, + "grad_norm": 0.4643072783946991, + "learning_rate": 0.00019870945562617393, + "loss": 1.1561, + "step": 1169 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 0.4544340968132019, + "learning_rate": 0.0001987072131250764, + "loss": 1.0764, + "step": 1170 + }, + { + "epoch": 0.208511396011396, + "grad_norm": 0.6036561727523804, + "learning_rate": 0.00019870496869001607, + "loss": 1.3961, + "step": 1171 + }, + { + "epoch": 0.20868945868945868, + "grad_norm": 0.41348758339881897, + "learning_rate": 0.00019870272232103695, + "loss": 1.2219, + "step": 1172 + }, + { + "epoch": 0.20886752136752137, + "grad_norm": 0.4184056222438812, + "learning_rate": 0.000198700474018183, + "loss": 1.1115, + "step": 1173 + }, + { + "epoch": 0.20904558404558404, + "grad_norm": 0.41920599341392517, + "learning_rate": 0.0001986982237814983, + "loss": 0.9207, + "step": 1174 + }, + { + "epoch": 0.2092236467236467, + "grad_norm": 0.4710249602794647, + "learning_rate": 0.00019869597161102694, + "loss": 1.1342, + "step": 1175 + }, + { + "epoch": 0.2094017094017094, + "grad_norm": 0.46897777915000916, + "learning_rate": 0.000198693717506813, + "loss": 0.983, + "step": 1176 + }, + { + "epoch": 0.20957977207977208, + "grad_norm": 0.4817039370536804, + "learning_rate": 0.00019869146146890074, + "loss": 1.0923, + "step": 1177 + }, + { + "epoch": 0.20975783475783477, + "grad_norm": 0.4806751012802124, + "learning_rate": 0.00019868920349733427, + "loss": 1.2296, + "step": 1178 + }, + { + "epoch": 0.20993589743589744, + "grad_norm": 0.44182994961738586, + "learning_rate": 0.0001986869435921579, + "loss": 1.1856, + "step": 1179 + }, + { + "epoch": 0.2101139601139601, + "grad_norm": 0.4282805621623993, + "learning_rate": 0.00019868468175341584, + "loss": 1.0046, + "step": 1180 + }, + { + "epoch": 0.2102920227920228, + "grad_norm": 0.5011838674545288, + "learning_rate": 0.00019868241798115242, + "loss": 1.2401, + "step": 1181 + }, + { + "epoch": 0.21047008547008547, + "grad_norm": 0.4282447397708893, + "learning_rate": 0.00019868015227541208, + "loss": 0.9338, + "step": 1182 + }, + { + "epoch": 0.21064814814814814, + "grad_norm": 0.4348810911178589, + "learning_rate": 0.00019867788463623912, + "loss": 0.926, + "step": 1183 + }, + { + "epoch": 0.21082621082621084, + "grad_norm": 0.41518425941467285, + "learning_rate": 0.00019867561506367799, + "loss": 1.2723, + "step": 1184 + }, + { + "epoch": 0.2110042735042735, + "grad_norm": 0.47346001863479614, + "learning_rate": 0.00019867334355777315, + "loss": 1.1931, + "step": 1185 + }, + { + "epoch": 0.21118233618233617, + "grad_norm": 0.4071715474128723, + "learning_rate": 0.00019867107011856914, + "loss": 0.9619, + "step": 1186 + }, + { + "epoch": 0.21136039886039887, + "grad_norm": 0.4803447425365448, + "learning_rate": 0.00019866879474611046, + "loss": 1.2, + "step": 1187 + }, + { + "epoch": 0.21153846153846154, + "grad_norm": 0.4827699661254883, + "learning_rate": 0.00019866651744044172, + "loss": 1.0938, + "step": 1188 + }, + { + "epoch": 0.2117165242165242, + "grad_norm": 0.4528424143791199, + "learning_rate": 0.00019866423820160756, + "loss": 0.9721, + "step": 1189 + }, + { + "epoch": 0.2118945868945869, + "grad_norm": 0.43566834926605225, + "learning_rate": 0.0001986619570296526, + "loss": 1.0352, + "step": 1190 + }, + { + "epoch": 0.21207264957264957, + "grad_norm": 0.4516540467739105, + "learning_rate": 0.0001986596739246215, + "loss": 1.1333, + "step": 1191 + }, + { + "epoch": 0.21225071225071226, + "grad_norm": 0.4456641376018524, + "learning_rate": 0.00019865738888655908, + "loss": 1.2813, + "step": 1192 + }, + { + "epoch": 0.21242877492877493, + "grad_norm": 0.47048309445381165, + "learning_rate": 0.00019865510191551008, + "loss": 1.1067, + "step": 1193 + }, + { + "epoch": 0.2126068376068376, + "grad_norm": 0.4604061543941498, + "learning_rate": 0.00019865281301151928, + "loss": 0.925, + "step": 1194 + }, + { + "epoch": 0.2127849002849003, + "grad_norm": 0.49341437220573425, + "learning_rate": 0.00019865052217463153, + "loss": 1.2319, + "step": 1195 + }, + { + "epoch": 0.21296296296296297, + "grad_norm": 0.5099014639854431, + "learning_rate": 0.00019864822940489173, + "loss": 1.139, + "step": 1196 + }, + { + "epoch": 0.21314102564102563, + "grad_norm": 0.41396936774253845, + "learning_rate": 0.0001986459347023448, + "loss": 1.0594, + "step": 1197 + }, + { + "epoch": 0.21331908831908833, + "grad_norm": 0.46071869134902954, + "learning_rate": 0.0001986436380670357, + "loss": 1.0815, + "step": 1198 + }, + { + "epoch": 0.213497150997151, + "grad_norm": 0.507882297039032, + "learning_rate": 0.00019864133949900942, + "loss": 1.3841, + "step": 1199 + }, + { + "epoch": 0.21367521367521367, + "grad_norm": 0.45680439472198486, + "learning_rate": 0.00019863903899831103, + "loss": 1.0945, + "step": 1200 + }, + { + "epoch": 0.21385327635327636, + "grad_norm": 0.44277429580688477, + "learning_rate": 0.00019863673656498555, + "loss": 1.1655, + "step": 1201 + }, + { + "epoch": 0.21403133903133903, + "grad_norm": 0.43890756368637085, + "learning_rate": 0.00019863443219907812, + "loss": 1.1186, + "step": 1202 + }, + { + "epoch": 0.2142094017094017, + "grad_norm": 0.3910178542137146, + "learning_rate": 0.0001986321259006339, + "loss": 1.0817, + "step": 1203 + }, + { + "epoch": 0.2143874643874644, + "grad_norm": 0.3803878128528595, + "learning_rate": 0.00019862981766969803, + "loss": 0.8022, + "step": 1204 + }, + { + "epoch": 0.21456552706552706, + "grad_norm": 0.4495108425617218, + "learning_rate": 0.0001986275075063158, + "loss": 1.2212, + "step": 1205 + }, + { + "epoch": 0.21474358974358973, + "grad_norm": 0.5211976766586304, + "learning_rate": 0.00019862519541053244, + "loss": 1.2771, + "step": 1206 + }, + { + "epoch": 0.21492165242165243, + "grad_norm": 0.4313061535358429, + "learning_rate": 0.00019862288138239325, + "loss": 1.1205, + "step": 1207 + }, + { + "epoch": 0.2150997150997151, + "grad_norm": 0.47110888361930847, + "learning_rate": 0.00019862056542194355, + "loss": 1.1835, + "step": 1208 + }, + { + "epoch": 0.2152777777777778, + "grad_norm": 0.5129403471946716, + "learning_rate": 0.00019861824752922876, + "loss": 1.1655, + "step": 1209 + }, + { + "epoch": 0.21545584045584046, + "grad_norm": 0.4353938102722168, + "learning_rate": 0.00019861592770429427, + "loss": 1.2794, + "step": 1210 + }, + { + "epoch": 0.21563390313390313, + "grad_norm": 0.48590636253356934, + "learning_rate": 0.0001986136059471855, + "loss": 1.2003, + "step": 1211 + }, + { + "epoch": 0.21581196581196582, + "grad_norm": 0.4738406836986542, + "learning_rate": 0.00019861128225794804, + "loss": 1.2271, + "step": 1212 + }, + { + "epoch": 0.2159900284900285, + "grad_norm": 0.45983126759529114, + "learning_rate": 0.0001986089566366273, + "loss": 1.1896, + "step": 1213 + }, + { + "epoch": 0.21616809116809116, + "grad_norm": 0.37296006083488464, + "learning_rate": 0.00019860662908326892, + "loss": 1.079, + "step": 1214 + }, + { + "epoch": 0.21634615384615385, + "grad_norm": 0.4442676305770874, + "learning_rate": 0.00019860429959791845, + "loss": 1.1754, + "step": 1215 + }, + { + "epoch": 0.21652421652421652, + "grad_norm": 0.4950128495693207, + "learning_rate": 0.0001986019681806216, + "loss": 1.1571, + "step": 1216 + }, + { + "epoch": 0.2167022792022792, + "grad_norm": 0.4374556541442871, + "learning_rate": 0.000198599634831424, + "loss": 1.1003, + "step": 1217 + }, + { + "epoch": 0.2168803418803419, + "grad_norm": 0.47301414608955383, + "learning_rate": 0.00019859729955037136, + "loss": 1.1426, + "step": 1218 + }, + { + "epoch": 0.21705840455840456, + "grad_norm": 0.41213178634643555, + "learning_rate": 0.00019859496233750947, + "loss": 1.0659, + "step": 1219 + }, + { + "epoch": 0.21723646723646722, + "grad_norm": 0.41601964831352234, + "learning_rate": 0.0001985926231928841, + "loss": 1.0248, + "step": 1220 + }, + { + "epoch": 0.21741452991452992, + "grad_norm": 0.46328839659690857, + "learning_rate": 0.0001985902821165411, + "loss": 1.0405, + "step": 1221 + }, + { + "epoch": 0.2175925925925926, + "grad_norm": 0.43287959694862366, + "learning_rate": 0.0001985879391085263, + "loss": 0.9202, + "step": 1222 + }, + { + "epoch": 0.21777065527065528, + "grad_norm": 0.4770444631576538, + "learning_rate": 0.00019858559416888568, + "loss": 1.0911, + "step": 1223 + }, + { + "epoch": 0.21794871794871795, + "grad_norm": 0.4756585955619812, + "learning_rate": 0.00019858324729766507, + "loss": 1.1566, + "step": 1224 + }, + { + "epoch": 0.21812678062678062, + "grad_norm": 0.4337233006954193, + "learning_rate": 0.00019858089849491054, + "loss": 0.9084, + "step": 1225 + }, + { + "epoch": 0.21830484330484332, + "grad_norm": 0.5165579319000244, + "learning_rate": 0.00019857854776066813, + "loss": 1.4154, + "step": 1226 + }, + { + "epoch": 0.21848290598290598, + "grad_norm": 0.4280378520488739, + "learning_rate": 0.00019857619509498382, + "loss": 1.1291, + "step": 1227 + }, + { + "epoch": 0.21866096866096865, + "grad_norm": 0.5375089049339294, + "learning_rate": 0.00019857384049790376, + "loss": 1.2985, + "step": 1228 + }, + { + "epoch": 0.21883903133903135, + "grad_norm": 0.4708811640739441, + "learning_rate": 0.00019857148396947401, + "loss": 1.0589, + "step": 1229 + }, + { + "epoch": 0.21901709401709402, + "grad_norm": 0.4744570255279541, + "learning_rate": 0.00019856912550974084, + "loss": 1.1269, + "step": 1230 + }, + { + "epoch": 0.21919515669515668, + "grad_norm": 0.5355265736579895, + "learning_rate": 0.00019856676511875043, + "loss": 1.1441, + "step": 1231 + }, + { + "epoch": 0.21937321937321938, + "grad_norm": 0.42718183994293213, + "learning_rate": 0.00019856440279654897, + "loss": 1.0244, + "step": 1232 + }, + { + "epoch": 0.21955128205128205, + "grad_norm": 0.5162127614021301, + "learning_rate": 0.00019856203854318283, + "loss": 1.2674, + "step": 1233 + }, + { + "epoch": 0.21972934472934472, + "grad_norm": 0.5180695652961731, + "learning_rate": 0.00019855967235869827, + "loss": 1.2472, + "step": 1234 + }, + { + "epoch": 0.2199074074074074, + "grad_norm": 0.4290023744106293, + "learning_rate": 0.00019855730424314167, + "loss": 1.0502, + "step": 1235 + }, + { + "epoch": 0.22008547008547008, + "grad_norm": 0.4418254792690277, + "learning_rate": 0.00019855493419655945, + "loss": 1.0589, + "step": 1236 + }, + { + "epoch": 0.22026353276353278, + "grad_norm": 0.4074663817882538, + "learning_rate": 0.000198552562218998, + "loss": 0.9197, + "step": 1237 + }, + { + "epoch": 0.22044159544159544, + "grad_norm": 0.4526660740375519, + "learning_rate": 0.00019855018831050383, + "loss": 1.2578, + "step": 1238 + }, + { + "epoch": 0.2206196581196581, + "grad_norm": 0.4747827649116516, + "learning_rate": 0.00019854781247112343, + "loss": 1.0841, + "step": 1239 + }, + { + "epoch": 0.2207977207977208, + "grad_norm": 0.41567128896713257, + "learning_rate": 0.00019854543470090334, + "loss": 1.0737, + "step": 1240 + }, + { + "epoch": 0.22097578347578348, + "grad_norm": 0.4793100953102112, + "learning_rate": 0.00019854305499989022, + "loss": 1.1972, + "step": 1241 + }, + { + "epoch": 0.22115384615384615, + "grad_norm": 0.41755473613739014, + "learning_rate": 0.00019854067336813058, + "loss": 1.2529, + "step": 1242 + }, + { + "epoch": 0.22133190883190884, + "grad_norm": 0.40421152114868164, + "learning_rate": 0.0001985382898056712, + "loss": 1.0549, + "step": 1243 + }, + { + "epoch": 0.2215099715099715, + "grad_norm": 0.45779645442962646, + "learning_rate": 0.0001985359043125587, + "loss": 1.1586, + "step": 1244 + }, + { + "epoch": 0.22168803418803418, + "grad_norm": 0.4380546808242798, + "learning_rate": 0.00019853351688883987, + "loss": 1.1024, + "step": 1245 + }, + { + "epoch": 0.22186609686609687, + "grad_norm": 0.39917269349098206, + "learning_rate": 0.00019853112753456142, + "loss": 0.9823, + "step": 1246 + }, + { + "epoch": 0.22204415954415954, + "grad_norm": 0.4228038489818573, + "learning_rate": 0.00019852873624977022, + "loss": 1.1684, + "step": 1247 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 0.4462146759033203, + "learning_rate": 0.00019852634303451315, + "loss": 0.9027, + "step": 1248 + }, + { + "epoch": 0.2224002849002849, + "grad_norm": 0.5682163834571838, + "learning_rate": 0.000198523947888837, + "loss": 1.141, + "step": 1249 + }, + { + "epoch": 0.22257834757834757, + "grad_norm": 0.44866830110549927, + "learning_rate": 0.0001985215508127888, + "loss": 1.0759, + "step": 1250 + }, + { + "epoch": 0.22275641025641027, + "grad_norm": 0.4034106135368347, + "learning_rate": 0.00019851915180641548, + "loss": 1.0675, + "step": 1251 + }, + { + "epoch": 0.22293447293447294, + "grad_norm": 0.4780726432800293, + "learning_rate": 0.00019851675086976397, + "loss": 1.0283, + "step": 1252 + }, + { + "epoch": 0.2231125356125356, + "grad_norm": 0.48892372846603394, + "learning_rate": 0.00019851434800288145, + "loss": 1.1159, + "step": 1253 + }, + { + "epoch": 0.2232905982905983, + "grad_norm": 0.42629215121269226, + "learning_rate": 0.0001985119432058149, + "loss": 1.0292, + "step": 1254 + }, + { + "epoch": 0.22346866096866097, + "grad_norm": 0.4496444761753082, + "learning_rate": 0.00019850953647861146, + "loss": 1.0252, + "step": 1255 + }, + { + "epoch": 0.22364672364672364, + "grad_norm": 0.4371408224105835, + "learning_rate": 0.00019850712782131828, + "loss": 1.1104, + "step": 1256 + }, + { + "epoch": 0.22382478632478633, + "grad_norm": 0.4910794496536255, + "learning_rate": 0.00019850471723398258, + "loss": 1.1928, + "step": 1257 + }, + { + "epoch": 0.224002849002849, + "grad_norm": 0.41235068440437317, + "learning_rate": 0.00019850230471665157, + "loss": 1.1261, + "step": 1258 + }, + { + "epoch": 0.22418091168091167, + "grad_norm": 0.4507700502872467, + "learning_rate": 0.0001984998902693725, + "loss": 1.0602, + "step": 1259 + }, + { + "epoch": 0.22435897435897437, + "grad_norm": 0.4654198884963989, + "learning_rate": 0.00019849747389219272, + "loss": 1.1258, + "step": 1260 + }, + { + "epoch": 0.22453703703703703, + "grad_norm": 0.439807653427124, + "learning_rate": 0.00019849505558515952, + "loss": 1.2312, + "step": 1261 + }, + { + "epoch": 0.2247150997150997, + "grad_norm": 0.4309258759021759, + "learning_rate": 0.00019849263534832035, + "loss": 1.0083, + "step": 1262 + }, + { + "epoch": 0.2248931623931624, + "grad_norm": 0.4920141100883484, + "learning_rate": 0.00019849021318172255, + "loss": 1.0254, + "step": 1263 + }, + { + "epoch": 0.22507122507122507, + "grad_norm": 0.5333457589149475, + "learning_rate": 0.00019848778908541367, + "loss": 1.3017, + "step": 1264 + }, + { + "epoch": 0.22524928774928774, + "grad_norm": 0.4096757769584656, + "learning_rate": 0.0001984853630594411, + "loss": 0.9531, + "step": 1265 + }, + { + "epoch": 0.22542735042735043, + "grad_norm": 0.5744075775146484, + "learning_rate": 0.00019848293510385244, + "loss": 1.1414, + "step": 1266 + }, + { + "epoch": 0.2256054131054131, + "grad_norm": 0.44707193970680237, + "learning_rate": 0.00019848050521869529, + "loss": 1.1926, + "step": 1267 + }, + { + "epoch": 0.2257834757834758, + "grad_norm": 0.4162999391555786, + "learning_rate": 0.00019847807340401716, + "loss": 1.1354, + "step": 1268 + }, + { + "epoch": 0.22596153846153846, + "grad_norm": 0.4273204207420349, + "learning_rate": 0.0001984756396598658, + "loss": 0.9956, + "step": 1269 + }, + { + "epoch": 0.22613960113960113, + "grad_norm": 0.5670466423034668, + "learning_rate": 0.00019847320398628878, + "loss": 1.2384, + "step": 1270 + }, + { + "epoch": 0.22631766381766383, + "grad_norm": 0.424544095993042, + "learning_rate": 0.00019847076638333395, + "loss": 0.9963, + "step": 1271 + }, + { + "epoch": 0.2264957264957265, + "grad_norm": 0.3716120719909668, + "learning_rate": 0.000198468326851049, + "loss": 0.865, + "step": 1272 + }, + { + "epoch": 0.22667378917378916, + "grad_norm": 0.4472847282886505, + "learning_rate": 0.00019846588538948172, + "loss": 1.174, + "step": 1273 + }, + { + "epoch": 0.22685185185185186, + "grad_norm": 0.4599195718765259, + "learning_rate": 0.00019846344199867994, + "loss": 1.289, + "step": 1274 + }, + { + "epoch": 0.22702991452991453, + "grad_norm": 0.4303213357925415, + "learning_rate": 0.0001984609966786916, + "loss": 1.1606, + "step": 1275 + }, + { + "epoch": 0.2272079772079772, + "grad_norm": 0.44893527030944824, + "learning_rate": 0.00019845854942956455, + "loss": 1.1043, + "step": 1276 + }, + { + "epoch": 0.2273860398860399, + "grad_norm": 0.40033379197120667, + "learning_rate": 0.00019845610025134676, + "loss": 1.1434, + "step": 1277 + }, + { + "epoch": 0.22756410256410256, + "grad_norm": 0.4385402202606201, + "learning_rate": 0.00019845364914408616, + "loss": 0.9943, + "step": 1278 + }, + { + "epoch": 0.22774216524216523, + "grad_norm": 0.42123618721961975, + "learning_rate": 0.0001984511961078309, + "loss": 1.0911, + "step": 1279 + }, + { + "epoch": 0.22792022792022792, + "grad_norm": 0.5558577179908752, + "learning_rate": 0.00019844874114262893, + "loss": 1.3893, + "step": 1280 + }, + { + "epoch": 0.2280982905982906, + "grad_norm": 0.3996453583240509, + "learning_rate": 0.00019844628424852835, + "loss": 0.8951, + "step": 1281 + }, + { + "epoch": 0.2282763532763533, + "grad_norm": 0.3943425714969635, + "learning_rate": 0.0001984438254255774, + "loss": 1.0595, + "step": 1282 + }, + { + "epoch": 0.22845441595441596, + "grad_norm": 0.4429021179676056, + "learning_rate": 0.00019844136467382414, + "loss": 1.0853, + "step": 1283 + }, + { + "epoch": 0.22863247863247863, + "grad_norm": 0.4515686631202698, + "learning_rate": 0.00019843890199331687, + "loss": 1.0829, + "step": 1284 + }, + { + "epoch": 0.22881054131054132, + "grad_norm": 0.5157768726348877, + "learning_rate": 0.00019843643738410378, + "loss": 1.334, + "step": 1285 + }, + { + "epoch": 0.228988603988604, + "grad_norm": 0.45833173394203186, + "learning_rate": 0.0001984339708462332, + "loss": 1.1353, + "step": 1286 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 0.46610337495803833, + "learning_rate": 0.00019843150237975344, + "loss": 1.1338, + "step": 1287 + }, + { + "epoch": 0.22934472934472935, + "grad_norm": 0.5076978802680969, + "learning_rate": 0.00019842903198471286, + "loss": 1.1811, + "step": 1288 + }, + { + "epoch": 0.22952279202279202, + "grad_norm": 0.4297824800014496, + "learning_rate": 0.00019842655966115986, + "loss": 1.1799, + "step": 1289 + }, + { + "epoch": 0.2297008547008547, + "grad_norm": 0.5304586291313171, + "learning_rate": 0.0001984240854091429, + "loss": 1.1315, + "step": 1290 + }, + { + "epoch": 0.22987891737891739, + "grad_norm": 0.45359212160110474, + "learning_rate": 0.00019842160922871042, + "loss": 1.1037, + "step": 1291 + }, + { + "epoch": 0.23005698005698005, + "grad_norm": 0.4416881203651428, + "learning_rate": 0.00019841913111991096, + "loss": 1.122, + "step": 1292 + }, + { + "epoch": 0.23023504273504272, + "grad_norm": 0.46682995557785034, + "learning_rate": 0.0001984166510827931, + "loss": 0.9808, + "step": 1293 + }, + { + "epoch": 0.23041310541310542, + "grad_norm": 0.44172337651252747, + "learning_rate": 0.00019841416911740538, + "loss": 0.9167, + "step": 1294 + }, + { + "epoch": 0.23059116809116809, + "grad_norm": 0.40562742948532104, + "learning_rate": 0.0001984116852237965, + "loss": 0.9547, + "step": 1295 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.4040384888648987, + "learning_rate": 0.00019840919940201503, + "loss": 1.1039, + "step": 1296 + }, + { + "epoch": 0.23094729344729345, + "grad_norm": 0.5094077587127686, + "learning_rate": 0.00019840671165210973, + "loss": 1.2283, + "step": 1297 + }, + { + "epoch": 0.23112535612535612, + "grad_norm": 0.48553213477134705, + "learning_rate": 0.00019840422197412938, + "loss": 1.0927, + "step": 1298 + }, + { + "epoch": 0.23130341880341881, + "grad_norm": 0.5197509527206421, + "learning_rate": 0.00019840173036812266, + "loss": 1.2154, + "step": 1299 + }, + { + "epoch": 0.23148148148148148, + "grad_norm": 0.42069005966186523, + "learning_rate": 0.0001983992368341385, + "loss": 1.0076, + "step": 1300 + }, + { + "epoch": 0.23165954415954415, + "grad_norm": 0.475204735994339, + "learning_rate": 0.00019839674137222567, + "loss": 1.1682, + "step": 1301 + }, + { + "epoch": 0.23183760683760685, + "grad_norm": 0.55730140209198, + "learning_rate": 0.0001983942439824331, + "loss": 1.2948, + "step": 1302 + }, + { + "epoch": 0.23201566951566951, + "grad_norm": 0.4533313512802124, + "learning_rate": 0.00019839174466480973, + "loss": 1.2691, + "step": 1303 + }, + { + "epoch": 0.23219373219373218, + "grad_norm": 0.4733520746231079, + "learning_rate": 0.0001983892434194045, + "loss": 1.2232, + "step": 1304 + }, + { + "epoch": 0.23237179487179488, + "grad_norm": 0.5085756182670593, + "learning_rate": 0.00019838674024626643, + "loss": 1.1347, + "step": 1305 + }, + { + "epoch": 0.23254985754985755, + "grad_norm": 0.4679976999759674, + "learning_rate": 0.00019838423514544456, + "loss": 1.0018, + "step": 1306 + }, + { + "epoch": 0.23272792022792022, + "grad_norm": 0.4234481751918793, + "learning_rate": 0.00019838172811698795, + "loss": 1.0472, + "step": 1307 + }, + { + "epoch": 0.2329059829059829, + "grad_norm": 0.5749204158782959, + "learning_rate": 0.00019837921916094579, + "loss": 1.2239, + "step": 1308 + }, + { + "epoch": 0.23308404558404558, + "grad_norm": 0.46715882420539856, + "learning_rate": 0.0001983767082773672, + "loss": 1.1924, + "step": 1309 + }, + { + "epoch": 0.23326210826210828, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.00019837419546630137, + "loss": 1.1086, + "step": 1310 + }, + { + "epoch": 0.23344017094017094, + "grad_norm": 0.4419243037700653, + "learning_rate": 0.0001983716807277975, + "loss": 1.1911, + "step": 1311 + }, + { + "epoch": 0.2336182336182336, + "grad_norm": 0.5107570290565491, + "learning_rate": 0.00019836916406190493, + "loss": 1.1071, + "step": 1312 + }, + { + "epoch": 0.2337962962962963, + "grad_norm": 0.5295659303665161, + "learning_rate": 0.00019836664546867293, + "loss": 1.2905, + "step": 1313 + }, + { + "epoch": 0.23397435897435898, + "grad_norm": 0.4844837784767151, + "learning_rate": 0.00019836412494815084, + "loss": 1.3507, + "step": 1314 + }, + { + "epoch": 0.23415242165242164, + "grad_norm": 0.6166049242019653, + "learning_rate": 0.00019836160250038808, + "loss": 1.2822, + "step": 1315 + }, + { + "epoch": 0.23433048433048434, + "grad_norm": 0.3229198753833771, + "learning_rate": 0.00019835907812543402, + "loss": 0.4959, + "step": 1316 + }, + { + "epoch": 0.234508547008547, + "grad_norm": 0.5788772702217102, + "learning_rate": 0.00019835655182333815, + "loss": 1.0832, + "step": 1317 + }, + { + "epoch": 0.23468660968660968, + "grad_norm": 0.525705099105835, + "learning_rate": 0.00019835402359414997, + "loss": 1.0968, + "step": 1318 + }, + { + "epoch": 0.23486467236467237, + "grad_norm": 0.5007779002189636, + "learning_rate": 0.000198351493437919, + "loss": 1.2788, + "step": 1319 + }, + { + "epoch": 0.23504273504273504, + "grad_norm": 0.4276871383190155, + "learning_rate": 0.00019834896135469484, + "loss": 1.0419, + "step": 1320 + }, + { + "epoch": 0.2352207977207977, + "grad_norm": 0.5359070301055908, + "learning_rate": 0.00019834642734452708, + "loss": 1.1308, + "step": 1321 + }, + { + "epoch": 0.2353988603988604, + "grad_norm": 0.4854908883571625, + "learning_rate": 0.0001983438914074654, + "loss": 1.1211, + "step": 1322 + }, + { + "epoch": 0.23557692307692307, + "grad_norm": 0.4913707375526428, + "learning_rate": 0.0001983413535435594, + "loss": 1.2392, + "step": 1323 + }, + { + "epoch": 0.23575498575498577, + "grad_norm": 0.46755748987197876, + "learning_rate": 0.0001983388137528589, + "loss": 0.9348, + "step": 1324 + }, + { + "epoch": 0.23593304843304844, + "grad_norm": 0.4592570960521698, + "learning_rate": 0.0001983362720354136, + "loss": 1.1339, + "step": 1325 + }, + { + "epoch": 0.2361111111111111, + "grad_norm": 0.5121711492538452, + "learning_rate": 0.00019833372839127335, + "loss": 1.2973, + "step": 1326 + }, + { + "epoch": 0.2362891737891738, + "grad_norm": 0.4809017479419708, + "learning_rate": 0.000198331182820488, + "loss": 0.9849, + "step": 1327 + }, + { + "epoch": 0.23646723646723647, + "grad_norm": 0.42340895533561707, + "learning_rate": 0.00019832863532310733, + "loss": 1.0731, + "step": 1328 + }, + { + "epoch": 0.23664529914529914, + "grad_norm": 0.5388045310974121, + "learning_rate": 0.00019832608589918135, + "loss": 1.0729, + "step": 1329 + }, + { + "epoch": 0.23682336182336183, + "grad_norm": 0.43075770139694214, + "learning_rate": 0.00019832353454875992, + "loss": 1.1684, + "step": 1330 + }, + { + "epoch": 0.2370014245014245, + "grad_norm": 0.554927408695221, + "learning_rate": 0.00019832098127189313, + "loss": 1.0842, + "step": 1331 + }, + { + "epoch": 0.23717948717948717, + "grad_norm": 0.5359260439872742, + "learning_rate": 0.0001983184260686309, + "loss": 1.2399, + "step": 1332 + }, + { + "epoch": 0.23735754985754987, + "grad_norm": 0.5141251087188721, + "learning_rate": 0.0001983158689390234, + "loss": 1.3752, + "step": 1333 + }, + { + "epoch": 0.23753561253561253, + "grad_norm": 0.4578750431537628, + "learning_rate": 0.00019831330988312067, + "loss": 1.0965, + "step": 1334 + }, + { + "epoch": 0.2377136752136752, + "grad_norm": 0.47974497079849243, + "learning_rate": 0.00019831074890097286, + "loss": 1.3379, + "step": 1335 + }, + { + "epoch": 0.2378917378917379, + "grad_norm": 0.4618176817893982, + "learning_rate": 0.00019830818599263014, + "loss": 1.274, + "step": 1336 + }, + { + "epoch": 0.23806980056980057, + "grad_norm": 0.4279816448688507, + "learning_rate": 0.00019830562115814276, + "loss": 0.996, + "step": 1337 + }, + { + "epoch": 0.23824786324786323, + "grad_norm": 0.4255026876926422, + "learning_rate": 0.0001983030543975609, + "loss": 0.969, + "step": 1338 + }, + { + "epoch": 0.23842592592592593, + "grad_norm": 0.4551412761211395, + "learning_rate": 0.00019830048571093493, + "loss": 1.0204, + "step": 1339 + }, + { + "epoch": 0.2386039886039886, + "grad_norm": 0.4747903048992157, + "learning_rate": 0.00019829791509831513, + "loss": 1.1816, + "step": 1340 + }, + { + "epoch": 0.2387820512820513, + "grad_norm": 0.47187140583992004, + "learning_rate": 0.00019829534255975188, + "loss": 1.1205, + "step": 1341 + }, + { + "epoch": 0.23896011396011396, + "grad_norm": 0.49332180619239807, + "learning_rate": 0.0001982927680952956, + "loss": 1.2657, + "step": 1342 + }, + { + "epoch": 0.23913817663817663, + "grad_norm": 0.5162837505340576, + "learning_rate": 0.0001982901917049967, + "loss": 1.2247, + "step": 1343 + }, + { + "epoch": 0.23931623931623933, + "grad_norm": 0.43407055735588074, + "learning_rate": 0.0001982876133889057, + "loss": 1.0038, + "step": 1344 + }, + { + "epoch": 0.239494301994302, + "grad_norm": 0.5132251977920532, + "learning_rate": 0.00019828503314707306, + "loss": 1.0678, + "step": 1345 + }, + { + "epoch": 0.23967236467236466, + "grad_norm": 0.46295464038848877, + "learning_rate": 0.00019828245097954937, + "loss": 1.1802, + "step": 1346 + }, + { + "epoch": 0.23985042735042736, + "grad_norm": 0.4682658314704895, + "learning_rate": 0.00019827986688638523, + "loss": 1.0249, + "step": 1347 + }, + { + "epoch": 0.24002849002849003, + "grad_norm": 0.49990561604499817, + "learning_rate": 0.00019827728086763125, + "loss": 1.0691, + "step": 1348 + }, + { + "epoch": 0.2402065527065527, + "grad_norm": 0.39090847969055176, + "learning_rate": 0.00019827469292333806, + "loss": 0.8367, + "step": 1349 + }, + { + "epoch": 0.2403846153846154, + "grad_norm": 0.5023905634880066, + "learning_rate": 0.00019827210305355645, + "loss": 1.0675, + "step": 1350 + }, + { + "epoch": 0.24056267806267806, + "grad_norm": 0.4744076430797577, + "learning_rate": 0.00019826951125833715, + "loss": 1.3166, + "step": 1351 + }, + { + "epoch": 0.24074074074074073, + "grad_norm": 0.44914689660072327, + "learning_rate": 0.00019826691753773088, + "loss": 0.9818, + "step": 1352 + }, + { + "epoch": 0.24091880341880342, + "grad_norm": 0.44391971826553345, + "learning_rate": 0.00019826432189178853, + "loss": 1.0448, + "step": 1353 + }, + { + "epoch": 0.2410968660968661, + "grad_norm": 0.46102839708328247, + "learning_rate": 0.00019826172432056086, + "loss": 0.9952, + "step": 1354 + }, + { + "epoch": 0.2412749287749288, + "grad_norm": 0.4796878695487976, + "learning_rate": 0.00019825912482409884, + "loss": 1.0977, + "step": 1355 + }, + { + "epoch": 0.24145299145299146, + "grad_norm": 0.5003768801689148, + "learning_rate": 0.0001982565234024534, + "loss": 1.3149, + "step": 1356 + }, + { + "epoch": 0.24163105413105412, + "grad_norm": 0.43475663661956787, + "learning_rate": 0.00019825392005567551, + "loss": 1.0527, + "step": 1357 + }, + { + "epoch": 0.24180911680911682, + "grad_norm": 0.46120527386665344, + "learning_rate": 0.00019825131478381613, + "loss": 1.2333, + "step": 1358 + }, + { + "epoch": 0.2419871794871795, + "grad_norm": 0.43748101592063904, + "learning_rate": 0.00019824870758692638, + "loss": 0.9788, + "step": 1359 + }, + { + "epoch": 0.24216524216524216, + "grad_norm": 0.5275192856788635, + "learning_rate": 0.00019824609846505727, + "loss": 1.1473, + "step": 1360 + }, + { + "epoch": 0.24234330484330485, + "grad_norm": 0.346463143825531, + "learning_rate": 0.00019824348741825993, + "loss": 0.6824, + "step": 1361 + }, + { + "epoch": 0.24252136752136752, + "grad_norm": 0.5004115700721741, + "learning_rate": 0.00019824087444658556, + "loss": 1.1853, + "step": 1362 + }, + { + "epoch": 0.2426994301994302, + "grad_norm": 0.42746666073799133, + "learning_rate": 0.00019823825955008533, + "loss": 0.9355, + "step": 1363 + }, + { + "epoch": 0.24287749287749288, + "grad_norm": 0.4099743068218231, + "learning_rate": 0.00019823564272881047, + "loss": 1.0753, + "step": 1364 + }, + { + "epoch": 0.24305555555555555, + "grad_norm": 0.5262967944145203, + "learning_rate": 0.00019823302398281226, + "loss": 1.2324, + "step": 1365 + }, + { + "epoch": 0.24323361823361822, + "grad_norm": 0.436069518327713, + "learning_rate": 0.000198230403312142, + "loss": 1.1887, + "step": 1366 + }, + { + "epoch": 0.24341168091168092, + "grad_norm": 0.38252368569374084, + "learning_rate": 0.00019822778071685107, + "loss": 1.0211, + "step": 1367 + }, + { + "epoch": 0.24358974358974358, + "grad_norm": 0.48024141788482666, + "learning_rate": 0.00019822515619699081, + "loss": 1.065, + "step": 1368 + }, + { + "epoch": 0.24376780626780628, + "grad_norm": 0.47421589493751526, + "learning_rate": 0.00019822252975261267, + "loss": 1.0433, + "step": 1369 + }, + { + "epoch": 0.24394586894586895, + "grad_norm": 0.46094807982444763, + "learning_rate": 0.00019821990138376808, + "loss": 1.1427, + "step": 1370 + }, + { + "epoch": 0.24412393162393162, + "grad_norm": 0.5093680620193481, + "learning_rate": 0.00019821727109050856, + "loss": 1.1086, + "step": 1371 + }, + { + "epoch": 0.2443019943019943, + "grad_norm": 0.41084879636764526, + "learning_rate": 0.00019821463887288566, + "loss": 1.0068, + "step": 1372 + }, + { + "epoch": 0.24448005698005698, + "grad_norm": 0.4991084635257721, + "learning_rate": 0.0001982120047309509, + "loss": 1.1884, + "step": 1373 + }, + { + "epoch": 0.24465811965811965, + "grad_norm": 0.39198383688926697, + "learning_rate": 0.00019820936866475595, + "loss": 0.9776, + "step": 1374 + }, + { + "epoch": 0.24483618233618235, + "grad_norm": 0.4517424702644348, + "learning_rate": 0.00019820673067435244, + "loss": 1.1491, + "step": 1375 + }, + { + "epoch": 0.245014245014245, + "grad_norm": 0.45881983637809753, + "learning_rate": 0.00019820409075979202, + "loss": 1.1198, + "step": 1376 + }, + { + "epoch": 0.24519230769230768, + "grad_norm": 0.4498792290687561, + "learning_rate": 0.00019820144892112646, + "loss": 1.0897, + "step": 1377 + }, + { + "epoch": 0.24537037037037038, + "grad_norm": 0.4128037393093109, + "learning_rate": 0.00019819880515840752, + "loss": 0.9415, + "step": 1378 + }, + { + "epoch": 0.24554843304843305, + "grad_norm": 0.4340885281562805, + "learning_rate": 0.00019819615947168698, + "loss": 1.201, + "step": 1379 + }, + { + "epoch": 0.24572649572649571, + "grad_norm": 0.43814027309417725, + "learning_rate": 0.00019819351186101667, + "loss": 1.1039, + "step": 1380 + }, + { + "epoch": 0.2459045584045584, + "grad_norm": 0.40115082263946533, + "learning_rate": 0.00019819086232644845, + "loss": 1.2599, + "step": 1381 + }, + { + "epoch": 0.24608262108262108, + "grad_norm": 0.4947351813316345, + "learning_rate": 0.00019818821086803426, + "loss": 1.252, + "step": 1382 + }, + { + "epoch": 0.24626068376068377, + "grad_norm": 0.45179441571235657, + "learning_rate": 0.0001981855574858261, + "loss": 1.1323, + "step": 1383 + }, + { + "epoch": 0.24643874643874644, + "grad_norm": 0.47159844636917114, + "learning_rate": 0.00019818290217987587, + "loss": 1.2053, + "step": 1384 + }, + { + "epoch": 0.2466168091168091, + "grad_norm": 0.4358448386192322, + "learning_rate": 0.0001981802449502356, + "loss": 1.1174, + "step": 1385 + }, + { + "epoch": 0.2467948717948718, + "grad_norm": 0.4588233530521393, + "learning_rate": 0.00019817758579695745, + "loss": 1.1098, + "step": 1386 + }, + { + "epoch": 0.24697293447293447, + "grad_norm": 0.4955112636089325, + "learning_rate": 0.00019817492472009338, + "loss": 1.258, + "step": 1387 + }, + { + "epoch": 0.24715099715099714, + "grad_norm": 0.4226941764354706, + "learning_rate": 0.00019817226171969565, + "loss": 1.0976, + "step": 1388 + }, + { + "epoch": 0.24732905982905984, + "grad_norm": 0.4076840579509735, + "learning_rate": 0.00019816959679581637, + "loss": 1.0121, + "step": 1389 + }, + { + "epoch": 0.2475071225071225, + "grad_norm": 0.4395063519477844, + "learning_rate": 0.0001981669299485078, + "loss": 1.3153, + "step": 1390 + }, + { + "epoch": 0.24768518518518517, + "grad_norm": 0.41010400652885437, + "learning_rate": 0.0001981642611778221, + "loss": 1.0717, + "step": 1391 + }, + { + "epoch": 0.24786324786324787, + "grad_norm": 0.43459352850914, + "learning_rate": 0.00019816159048381167, + "loss": 1.1077, + "step": 1392 + }, + { + "epoch": 0.24804131054131054, + "grad_norm": 0.46291449666023254, + "learning_rate": 0.00019815891786652875, + "loss": 1.0257, + "step": 1393 + }, + { + "epoch": 0.2482193732193732, + "grad_norm": 0.46408146619796753, + "learning_rate": 0.00019815624332602578, + "loss": 0.7899, + "step": 1394 + }, + { + "epoch": 0.2483974358974359, + "grad_norm": 0.4763357937335968, + "learning_rate": 0.00019815356686235508, + "loss": 0.9857, + "step": 1395 + }, + { + "epoch": 0.24857549857549857, + "grad_norm": 0.4766457676887512, + "learning_rate": 0.00019815088847556918, + "loss": 1.0589, + "step": 1396 + }, + { + "epoch": 0.24875356125356127, + "grad_norm": 0.4486583173274994, + "learning_rate": 0.0001981482081657205, + "loss": 1.2572, + "step": 1397 + }, + { + "epoch": 0.24893162393162394, + "grad_norm": 0.468878835439682, + "learning_rate": 0.00019814552593286155, + "loss": 1.101, + "step": 1398 + }, + { + "epoch": 0.2491096866096866, + "grad_norm": 0.4230278730392456, + "learning_rate": 0.0001981428417770449, + "loss": 0.9457, + "step": 1399 + }, + { + "epoch": 0.2492877492877493, + "grad_norm": 0.45630761981010437, + "learning_rate": 0.00019814015569832315, + "loss": 1.0665, + "step": 1400 + }, + { + "epoch": 0.24946581196581197, + "grad_norm": 0.5780113935470581, + "learning_rate": 0.00019813746769674893, + "loss": 1.1064, + "step": 1401 + }, + { + "epoch": 0.24964387464387464, + "grad_norm": 0.4343436658382416, + "learning_rate": 0.0001981347777723749, + "loss": 1.1132, + "step": 1402 + }, + { + "epoch": 0.24982193732193733, + "grad_norm": 0.4879056513309479, + "learning_rate": 0.0001981320859252537, + "loss": 1.1301, + "step": 1403 + }, + { + "epoch": 0.25, + "grad_norm": 0.5248328447341919, + "learning_rate": 0.00019812939215543818, + "loss": 1.1468, + "step": 1404 + }, + { + "epoch": 0.25, + "eval_loss": 1.115895390510559, + "eval_runtime": 25.0474, + "eval_samples_per_second": 41.561, + "eval_steps_per_second": 20.801, + "step": 1404 + }, + { + "epoch": 0.2501780626780627, + "grad_norm": 0.5076769590377808, + "learning_rate": 0.00019812669646298106, + "loss": 1.1428, + "step": 1405 + }, + { + "epoch": 0.25035612535612534, + "grad_norm": 0.5510252714157104, + "learning_rate": 0.00019812399884793514, + "loss": 1.3383, + "step": 1406 + }, + { + "epoch": 0.25053418803418803, + "grad_norm": 0.48918986320495605, + "learning_rate": 0.0001981212993103533, + "loss": 1.1507, + "step": 1407 + }, + { + "epoch": 0.25071225071225073, + "grad_norm": 0.4678935110569, + "learning_rate": 0.00019811859785028846, + "loss": 1.13, + "step": 1408 + }, + { + "epoch": 0.25089031339031337, + "grad_norm": 0.5155254602432251, + "learning_rate": 0.0001981158944677935, + "loss": 1.1194, + "step": 1409 + }, + { + "epoch": 0.25106837606837606, + "grad_norm": 0.4533839523792267, + "learning_rate": 0.00019811318916292142, + "loss": 0.9464, + "step": 1410 + }, + { + "epoch": 0.25124643874643876, + "grad_norm": 0.5142433047294617, + "learning_rate": 0.00019811048193572517, + "loss": 1.0837, + "step": 1411 + }, + { + "epoch": 0.2514245014245014, + "grad_norm": 0.4330446124076843, + "learning_rate": 0.00019810777278625788, + "loss": 0.9117, + "step": 1412 + }, + { + "epoch": 0.2516025641025641, + "grad_norm": 0.44806256890296936, + "learning_rate": 0.00019810506171457254, + "loss": 1.1643, + "step": 1413 + }, + { + "epoch": 0.2517806267806268, + "grad_norm": 0.43526285886764526, + "learning_rate": 0.00019810234872072235, + "loss": 0.9776, + "step": 1414 + }, + { + "epoch": 0.25195868945868943, + "grad_norm": 0.47394511103630066, + "learning_rate": 0.00019809963380476039, + "loss": 1.0935, + "step": 1415 + }, + { + "epoch": 0.25213675213675213, + "grad_norm": 0.48961278796195984, + "learning_rate": 0.00019809691696673993, + "loss": 1.179, + "step": 1416 + }, + { + "epoch": 0.2523148148148148, + "grad_norm": 0.43153589963912964, + "learning_rate": 0.00019809419820671412, + "loss": 0.906, + "step": 1417 + }, + { + "epoch": 0.25249287749287747, + "grad_norm": 0.41187527775764465, + "learning_rate": 0.00019809147752473632, + "loss": 0.899, + "step": 1418 + }, + { + "epoch": 0.25267094017094016, + "grad_norm": 0.5003183484077454, + "learning_rate": 0.00019808875492085973, + "loss": 1.0606, + "step": 1419 + }, + { + "epoch": 0.25284900284900286, + "grad_norm": 0.4430316984653473, + "learning_rate": 0.00019808603039513778, + "loss": 0.9167, + "step": 1420 + }, + { + "epoch": 0.25302706552706555, + "grad_norm": 0.4577699601650238, + "learning_rate": 0.00019808330394762382, + "loss": 1.1184, + "step": 1421 + }, + { + "epoch": 0.2532051282051282, + "grad_norm": 0.42656826972961426, + "learning_rate": 0.0001980805755783713, + "loss": 0.9335, + "step": 1422 + }, + { + "epoch": 0.2533831908831909, + "grad_norm": 0.40980881452560425, + "learning_rate": 0.0001980778452874336, + "loss": 0.9756, + "step": 1423 + }, + { + "epoch": 0.2535612535612536, + "grad_norm": 0.5752090811729431, + "learning_rate": 0.00019807511307486423, + "loss": 1.1694, + "step": 1424 + }, + { + "epoch": 0.2537393162393162, + "grad_norm": 0.5000349283218384, + "learning_rate": 0.00019807237894071681, + "loss": 0.9515, + "step": 1425 + }, + { + "epoch": 0.2539173789173789, + "grad_norm": 0.5159069299697876, + "learning_rate": 0.00019806964288504483, + "loss": 1.4014, + "step": 1426 + }, + { + "epoch": 0.2540954415954416, + "grad_norm": 0.5377941131591797, + "learning_rate": 0.00019806690490790194, + "loss": 1.2832, + "step": 1427 + }, + { + "epoch": 0.25427350427350426, + "grad_norm": 0.4565938711166382, + "learning_rate": 0.00019806416500934174, + "loss": 1.0629, + "step": 1428 + }, + { + "epoch": 0.25445156695156695, + "grad_norm": 0.49867144227027893, + "learning_rate": 0.00019806142318941797, + "loss": 1.2011, + "step": 1429 + }, + { + "epoch": 0.25462962962962965, + "grad_norm": 0.5111994743347168, + "learning_rate": 0.00019805867944818427, + "loss": 0.8925, + "step": 1430 + }, + { + "epoch": 0.2548076923076923, + "grad_norm": 0.5204268097877502, + "learning_rate": 0.00019805593378569448, + "loss": 1.2956, + "step": 1431 + }, + { + "epoch": 0.254985754985755, + "grad_norm": 0.3889026939868927, + "learning_rate": 0.00019805318620200234, + "loss": 1.0355, + "step": 1432 + }, + { + "epoch": 0.2551638176638177, + "grad_norm": 0.46825656294822693, + "learning_rate": 0.00019805043669716174, + "loss": 1.0444, + "step": 1433 + }, + { + "epoch": 0.2553418803418803, + "grad_norm": 0.4509420394897461, + "learning_rate": 0.00019804768527122648, + "loss": 1.0423, + "step": 1434 + }, + { + "epoch": 0.255519943019943, + "grad_norm": 0.4514774978160858, + "learning_rate": 0.0001980449319242505, + "loss": 1.1588, + "step": 1435 + }, + { + "epoch": 0.2556980056980057, + "grad_norm": 0.43019044399261475, + "learning_rate": 0.0001980421766562878, + "loss": 0.9939, + "step": 1436 + }, + { + "epoch": 0.25587606837606836, + "grad_norm": 0.5056091547012329, + "learning_rate": 0.00019803941946739228, + "loss": 1.1238, + "step": 1437 + }, + { + "epoch": 0.25605413105413105, + "grad_norm": 0.48664605617523193, + "learning_rate": 0.000198036660357618, + "loss": 1.0702, + "step": 1438 + }, + { + "epoch": 0.25623219373219375, + "grad_norm": 0.4500972032546997, + "learning_rate": 0.000198033899327019, + "loss": 0.9365, + "step": 1439 + }, + { + "epoch": 0.2564102564102564, + "grad_norm": 0.4800589382648468, + "learning_rate": 0.0001980311363756494, + "loss": 1.1159, + "step": 1440 + }, + { + "epoch": 0.2565883190883191, + "grad_norm": 0.3486495316028595, + "learning_rate": 0.0001980283715035633, + "loss": 0.6029, + "step": 1441 + }, + { + "epoch": 0.2567663817663818, + "grad_norm": 0.46258702874183655, + "learning_rate": 0.00019802560471081493, + "loss": 1.025, + "step": 1442 + }, + { + "epoch": 0.2569444444444444, + "grad_norm": 0.4846673607826233, + "learning_rate": 0.00019802283599745844, + "loss": 1.1105, + "step": 1443 + }, + { + "epoch": 0.2571225071225071, + "grad_norm": 0.4586990475654602, + "learning_rate": 0.00019802006536354813, + "loss": 0.9897, + "step": 1444 + }, + { + "epoch": 0.2573005698005698, + "grad_norm": 0.5177786350250244, + "learning_rate": 0.00019801729280913825, + "loss": 1.2558, + "step": 1445 + }, + { + "epoch": 0.25747863247863245, + "grad_norm": 0.43213751912117004, + "learning_rate": 0.00019801451833428312, + "loss": 1.0961, + "step": 1446 + }, + { + "epoch": 0.25765669515669515, + "grad_norm": 0.42974478006362915, + "learning_rate": 0.00019801174193903714, + "loss": 1.0659, + "step": 1447 + }, + { + "epoch": 0.25783475783475784, + "grad_norm": 0.4424504339694977, + "learning_rate": 0.00019800896362345464, + "loss": 0.9805, + "step": 1448 + }, + { + "epoch": 0.25801282051282054, + "grad_norm": 0.4734833836555481, + "learning_rate": 0.0001980061833875901, + "loss": 1.255, + "step": 1449 + }, + { + "epoch": 0.2581908831908832, + "grad_norm": 0.41024845838546753, + "learning_rate": 0.000198003401231498, + "loss": 1.0908, + "step": 1450 + }, + { + "epoch": 0.2583689458689459, + "grad_norm": 0.43603816628456116, + "learning_rate": 0.00019800061715523283, + "loss": 1.0611, + "step": 1451 + }, + { + "epoch": 0.25854700854700857, + "grad_norm": 0.4871339499950409, + "learning_rate": 0.00019799783115884915, + "loss": 1.1851, + "step": 1452 + }, + { + "epoch": 0.2587250712250712, + "grad_norm": 0.49758270382881165, + "learning_rate": 0.00019799504324240157, + "loss": 1.1936, + "step": 1453 + }, + { + "epoch": 0.2589031339031339, + "grad_norm": 0.4201010763645172, + "learning_rate": 0.00019799225340594466, + "loss": 1.1567, + "step": 1454 + }, + { + "epoch": 0.2590811965811966, + "grad_norm": 0.4200313091278076, + "learning_rate": 0.00019798946164953309, + "loss": 0.9666, + "step": 1455 + }, + { + "epoch": 0.25925925925925924, + "grad_norm": 0.43001702427864075, + "learning_rate": 0.0001979866679732216, + "loss": 1.0104, + "step": 1456 + }, + { + "epoch": 0.25943732193732194, + "grad_norm": 0.46733465790748596, + "learning_rate": 0.0001979838723770649, + "loss": 1.0927, + "step": 1457 + }, + { + "epoch": 0.25961538461538464, + "grad_norm": 0.4513280391693115, + "learning_rate": 0.00019798107486111773, + "loss": 1.0282, + "step": 1458 + }, + { + "epoch": 0.2597934472934473, + "grad_norm": 0.40411749482154846, + "learning_rate": 0.00019797827542543495, + "loss": 1.0789, + "step": 1459 + }, + { + "epoch": 0.25997150997151, + "grad_norm": 0.4359099268913269, + "learning_rate": 0.0001979754740700714, + "loss": 1.0616, + "step": 1460 + }, + { + "epoch": 0.26014957264957267, + "grad_norm": 0.4979047477245331, + "learning_rate": 0.00019797267079508198, + "loss": 1.2948, + "step": 1461 + }, + { + "epoch": 0.2603276353276353, + "grad_norm": 0.44698619842529297, + "learning_rate": 0.0001979698656005216, + "loss": 0.9198, + "step": 1462 + }, + { + "epoch": 0.260505698005698, + "grad_norm": 0.48437631130218506, + "learning_rate": 0.00019796705848644516, + "loss": 1.3207, + "step": 1463 + }, + { + "epoch": 0.2606837606837607, + "grad_norm": 0.4382587671279907, + "learning_rate": 0.00019796424945290778, + "loss": 1.1315, + "step": 1464 + }, + { + "epoch": 0.26086182336182334, + "grad_norm": 0.4565944969654083, + "learning_rate": 0.0001979614384999644, + "loss": 1.1893, + "step": 1465 + }, + { + "epoch": 0.26103988603988604, + "grad_norm": 0.4705163836479187, + "learning_rate": 0.00019795862562767017, + "loss": 1.1132, + "step": 1466 + }, + { + "epoch": 0.26121794871794873, + "grad_norm": 0.525184690952301, + "learning_rate": 0.00019795581083608012, + "loss": 1.2111, + "step": 1467 + }, + { + "epoch": 0.2613960113960114, + "grad_norm": 0.45215457677841187, + "learning_rate": 0.00019795299412524945, + "loss": 1.1851, + "step": 1468 + }, + { + "epoch": 0.26157407407407407, + "grad_norm": 0.4336663484573364, + "learning_rate": 0.00019795017549523335, + "loss": 1.0147, + "step": 1469 + }, + { + "epoch": 0.26175213675213677, + "grad_norm": 0.5327649712562561, + "learning_rate": 0.00019794735494608703, + "loss": 1.1743, + "step": 1470 + }, + { + "epoch": 0.2619301994301994, + "grad_norm": 0.49972307682037354, + "learning_rate": 0.00019794453247786578, + "loss": 1.1624, + "step": 1471 + }, + { + "epoch": 0.2621082621082621, + "grad_norm": 0.43475785851478577, + "learning_rate": 0.00019794170809062485, + "loss": 0.9888, + "step": 1472 + }, + { + "epoch": 0.2622863247863248, + "grad_norm": 0.428838849067688, + "learning_rate": 0.0001979388817844196, + "loss": 0.9154, + "step": 1473 + }, + { + "epoch": 0.26246438746438744, + "grad_norm": 0.508568286895752, + "learning_rate": 0.00019793605355930544, + "loss": 1.1679, + "step": 1474 + }, + { + "epoch": 0.26264245014245013, + "grad_norm": 0.47791770100593567, + "learning_rate": 0.00019793322341533776, + "loss": 1.1375, + "step": 1475 + }, + { + "epoch": 0.26282051282051283, + "grad_norm": 0.41909220814704895, + "learning_rate": 0.00019793039135257196, + "loss": 1.0235, + "step": 1476 + }, + { + "epoch": 0.26299857549857547, + "grad_norm": 0.5564408302307129, + "learning_rate": 0.00019792755737106361, + "loss": 1.0756, + "step": 1477 + }, + { + "epoch": 0.26317663817663817, + "grad_norm": 0.42813625931739807, + "learning_rate": 0.0001979247214708682, + "loss": 0.8213, + "step": 1478 + }, + { + "epoch": 0.26335470085470086, + "grad_norm": 0.44495970010757446, + "learning_rate": 0.00019792188365204126, + "loss": 0.9654, + "step": 1479 + }, + { + "epoch": 0.26353276353276356, + "grad_norm": 0.47473424673080444, + "learning_rate": 0.00019791904391463846, + "loss": 1.1643, + "step": 1480 + }, + { + "epoch": 0.2637108262108262, + "grad_norm": 0.40189051628112793, + "learning_rate": 0.0001979162022587154, + "loss": 0.8687, + "step": 1481 + }, + { + "epoch": 0.2638888888888889, + "grad_norm": 0.44629937410354614, + "learning_rate": 0.00019791335868432776, + "loss": 1.0284, + "step": 1482 + }, + { + "epoch": 0.2640669515669516, + "grad_norm": 0.511275053024292, + "learning_rate": 0.00019791051319153124, + "loss": 1.2217, + "step": 1483 + }, + { + "epoch": 0.26424501424501423, + "grad_norm": 0.5136445164680481, + "learning_rate": 0.00019790766578038163, + "loss": 1.1129, + "step": 1484 + }, + { + "epoch": 0.2644230769230769, + "grad_norm": 0.4450451135635376, + "learning_rate": 0.00019790481645093469, + "loss": 0.9912, + "step": 1485 + }, + { + "epoch": 0.2646011396011396, + "grad_norm": 0.39455199241638184, + "learning_rate": 0.00019790196520324621, + "loss": 1.0887, + "step": 1486 + }, + { + "epoch": 0.26477920227920226, + "grad_norm": 0.4444045126438141, + "learning_rate": 0.00019789911203737216, + "loss": 1.1559, + "step": 1487 + }, + { + "epoch": 0.26495726495726496, + "grad_norm": 0.4769677221775055, + "learning_rate": 0.0001978962569533683, + "loss": 1.147, + "step": 1488 + }, + { + "epoch": 0.26513532763532766, + "grad_norm": 0.40226617455482483, + "learning_rate": 0.0001978933999512907, + "loss": 1.0966, + "step": 1489 + }, + { + "epoch": 0.2653133903133903, + "grad_norm": 0.4640974700450897, + "learning_rate": 0.00019789054103119526, + "loss": 1.1002, + "step": 1490 + }, + { + "epoch": 0.265491452991453, + "grad_norm": 0.48251107335090637, + "learning_rate": 0.00019788768019313806, + "loss": 1.07, + "step": 1491 + }, + { + "epoch": 0.2656695156695157, + "grad_norm": 0.4836949408054352, + "learning_rate": 0.00019788481743717506, + "loss": 1.2992, + "step": 1492 + }, + { + "epoch": 0.26584757834757833, + "grad_norm": 0.4253857135772705, + "learning_rate": 0.00019788195276336244, + "loss": 1.1326, + "step": 1493 + }, + { + "epoch": 0.266025641025641, + "grad_norm": 0.5161862373352051, + "learning_rate": 0.0001978790861717563, + "loss": 1.2131, + "step": 1494 + }, + { + "epoch": 0.2662037037037037, + "grad_norm": 0.5223346948623657, + "learning_rate": 0.00019787621766241274, + "loss": 1.0933, + "step": 1495 + }, + { + "epoch": 0.26638176638176636, + "grad_norm": 0.37622541189193726, + "learning_rate": 0.000197873347235388, + "loss": 0.8919, + "step": 1496 + }, + { + "epoch": 0.26655982905982906, + "grad_norm": 0.4425419569015503, + "learning_rate": 0.0001978704748907384, + "loss": 1.0411, + "step": 1497 + }, + { + "epoch": 0.26673789173789175, + "grad_norm": 0.4536985456943512, + "learning_rate": 0.00019786760062852015, + "loss": 1.2747, + "step": 1498 + }, + { + "epoch": 0.2669159544159544, + "grad_norm": 0.4998049736022949, + "learning_rate": 0.00019786472444878955, + "loss": 1.3214, + "step": 1499 + }, + { + "epoch": 0.2670940170940171, + "grad_norm": 0.42104312777519226, + "learning_rate": 0.00019786184635160295, + "loss": 0.7878, + "step": 1500 + }, + { + "epoch": 0.2672720797720798, + "grad_norm": 0.5354288220405579, + "learning_rate": 0.00019785896633701678, + "loss": 1.0642, + "step": 1501 + }, + { + "epoch": 0.2674501424501424, + "grad_norm": 0.4681485891342163, + "learning_rate": 0.00019785608440508744, + "loss": 1.1737, + "step": 1502 + }, + { + "epoch": 0.2676282051282051, + "grad_norm": 0.49107062816619873, + "learning_rate": 0.0001978532005558714, + "loss": 1.1507, + "step": 1503 + }, + { + "epoch": 0.2678062678062678, + "grad_norm": 0.4173283576965332, + "learning_rate": 0.0001978503147894252, + "loss": 1.0538, + "step": 1504 + }, + { + "epoch": 0.26798433048433046, + "grad_norm": 0.49354055523872375, + "learning_rate": 0.0001978474271058053, + "loss": 1.1043, + "step": 1505 + }, + { + "epoch": 0.26816239316239315, + "grad_norm": 0.5787215232849121, + "learning_rate": 0.00019784453750506834, + "loss": 0.9245, + "step": 1506 + }, + { + "epoch": 0.26834045584045585, + "grad_norm": 0.48982590436935425, + "learning_rate": 0.00019784164598727095, + "loss": 1.2007, + "step": 1507 + }, + { + "epoch": 0.26851851851851855, + "grad_norm": 0.4971007704734802, + "learning_rate": 0.00019783875255246973, + "loss": 1.1174, + "step": 1508 + }, + { + "epoch": 0.2686965811965812, + "grad_norm": 0.5200340151786804, + "learning_rate": 0.00019783585720072142, + "loss": 1.1967, + "step": 1509 + }, + { + "epoch": 0.2688746438746439, + "grad_norm": 0.47911885380744934, + "learning_rate": 0.00019783295993208271, + "loss": 1.162, + "step": 1510 + }, + { + "epoch": 0.2690527065527066, + "grad_norm": 0.4764275848865509, + "learning_rate": 0.00019783006074661037, + "loss": 1.1358, + "step": 1511 + }, + { + "epoch": 0.2692307692307692, + "grad_norm": 0.478545606136322, + "learning_rate": 0.00019782715964436124, + "loss": 1.0096, + "step": 1512 + }, + { + "epoch": 0.2694088319088319, + "grad_norm": 0.5512787699699402, + "learning_rate": 0.00019782425662539212, + "loss": 1.1799, + "step": 1513 + }, + { + "epoch": 0.2695868945868946, + "grad_norm": 0.5495108962059021, + "learning_rate": 0.00019782135168975988, + "loss": 1.0959, + "step": 1514 + }, + { + "epoch": 0.26976495726495725, + "grad_norm": 0.42052868008613586, + "learning_rate": 0.0001978184448375215, + "loss": 1.1872, + "step": 1515 + }, + { + "epoch": 0.26994301994301995, + "grad_norm": 0.4994426965713501, + "learning_rate": 0.0001978155360687339, + "loss": 1.0568, + "step": 1516 + }, + { + "epoch": 0.27012108262108264, + "grad_norm": 0.459577351808548, + "learning_rate": 0.00019781262538345402, + "loss": 1.0315, + "step": 1517 + }, + { + "epoch": 0.2702991452991453, + "grad_norm": 0.4792841374874115, + "learning_rate": 0.00019780971278173895, + "loss": 1.2055, + "step": 1518 + }, + { + "epoch": 0.270477207977208, + "grad_norm": 0.5017708539962769, + "learning_rate": 0.00019780679826364575, + "loss": 1.157, + "step": 1519 + }, + { + "epoch": 0.2706552706552707, + "grad_norm": 0.5197349786758423, + "learning_rate": 0.00019780388182923152, + "loss": 0.9101, + "step": 1520 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 0.4226742684841156, + "learning_rate": 0.00019780096347855338, + "loss": 1.0525, + "step": 1521 + }, + { + "epoch": 0.271011396011396, + "grad_norm": 0.5058164596557617, + "learning_rate": 0.00019779804321166852, + "loss": 0.931, + "step": 1522 + }, + { + "epoch": 0.2711894586894587, + "grad_norm": 0.44492244720458984, + "learning_rate": 0.00019779512102863418, + "loss": 1.0641, + "step": 1523 + }, + { + "epoch": 0.27136752136752135, + "grad_norm": 0.5348989963531494, + "learning_rate": 0.00019779219692950758, + "loss": 1.1692, + "step": 1524 + }, + { + "epoch": 0.27154558404558404, + "grad_norm": 0.4631774425506592, + "learning_rate": 0.00019778927091434602, + "loss": 1.0876, + "step": 1525 + }, + { + "epoch": 0.27172364672364674, + "grad_norm": 0.45957499742507935, + "learning_rate": 0.00019778634298320684, + "loss": 0.9527, + "step": 1526 + }, + { + "epoch": 0.2719017094017094, + "grad_norm": 0.4506755769252777, + "learning_rate": 0.00019778341313614743, + "loss": 1.086, + "step": 1527 + }, + { + "epoch": 0.2720797720797721, + "grad_norm": 0.4900587797164917, + "learning_rate": 0.00019778048137322513, + "loss": 0.9911, + "step": 1528 + }, + { + "epoch": 0.27225783475783477, + "grad_norm": 0.478127658367157, + "learning_rate": 0.00019777754769449745, + "loss": 1.2083, + "step": 1529 + }, + { + "epoch": 0.2724358974358974, + "grad_norm": 0.47220897674560547, + "learning_rate": 0.00019777461210002183, + "loss": 1.0313, + "step": 1530 + }, + { + "epoch": 0.2726139601139601, + "grad_norm": 0.4526277184486389, + "learning_rate": 0.0001977716745898558, + "loss": 1.2648, + "step": 1531 + }, + { + "epoch": 0.2727920227920228, + "grad_norm": 0.42907601594924927, + "learning_rate": 0.00019776873516405688, + "loss": 0.8645, + "step": 1532 + }, + { + "epoch": 0.27297008547008544, + "grad_norm": 0.43440163135528564, + "learning_rate": 0.00019776579382268272, + "loss": 0.9702, + "step": 1533 + }, + { + "epoch": 0.27314814814814814, + "grad_norm": 0.48213550448417664, + "learning_rate": 0.0001977628505657909, + "loss": 0.998, + "step": 1534 + }, + { + "epoch": 0.27332621082621084, + "grad_norm": 0.43385565280914307, + "learning_rate": 0.00019775990539343914, + "loss": 1.0575, + "step": 1535 + }, + { + "epoch": 0.27350427350427353, + "grad_norm": 0.45706847310066223, + "learning_rate": 0.00019775695830568507, + "loss": 1.3024, + "step": 1536 + }, + { + "epoch": 0.27368233618233617, + "grad_norm": 0.45769137144088745, + "learning_rate": 0.00019775400930258652, + "loss": 1.0987, + "step": 1537 + }, + { + "epoch": 0.27386039886039887, + "grad_norm": 0.44682395458221436, + "learning_rate": 0.00019775105838420117, + "loss": 1.1327, + "step": 1538 + }, + { + "epoch": 0.27403846153846156, + "grad_norm": 0.5923072099685669, + "learning_rate": 0.00019774810555058694, + "loss": 1.4766, + "step": 1539 + }, + { + "epoch": 0.2742165242165242, + "grad_norm": 0.4327206015586853, + "learning_rate": 0.0001977451508018016, + "loss": 1.1175, + "step": 1540 + }, + { + "epoch": 0.2743945868945869, + "grad_norm": 0.48036691546440125, + "learning_rate": 0.00019774219413790315, + "loss": 1.1189, + "step": 1541 + }, + { + "epoch": 0.2745726495726496, + "grad_norm": 0.41371914744377136, + "learning_rate": 0.00019773923555894935, + "loss": 1.1366, + "step": 1542 + }, + { + "epoch": 0.27475071225071224, + "grad_norm": 0.4452378749847412, + "learning_rate": 0.00019773627506499832, + "loss": 0.9517, + "step": 1543 + }, + { + "epoch": 0.27492877492877493, + "grad_norm": 0.469098299741745, + "learning_rate": 0.00019773331265610802, + "loss": 1.0848, + "step": 1544 + }, + { + "epoch": 0.27510683760683763, + "grad_norm": 0.5390294790267944, + "learning_rate": 0.00019773034833233646, + "loss": 0.8589, + "step": 1545 + }, + { + "epoch": 0.27528490028490027, + "grad_norm": 0.5368238091468811, + "learning_rate": 0.00019772738209374174, + "loss": 1.2954, + "step": 1546 + }, + { + "epoch": 0.27546296296296297, + "grad_norm": 0.4705318510532379, + "learning_rate": 0.00019772441394038198, + "loss": 1.2252, + "step": 1547 + }, + { + "epoch": 0.27564102564102566, + "grad_norm": 0.4682813286781311, + "learning_rate": 0.00019772144387231533, + "loss": 1.0855, + "step": 1548 + }, + { + "epoch": 0.2758190883190883, + "grad_norm": 0.46876460313796997, + "learning_rate": 0.0001977184718896, + "loss": 1.1959, + "step": 1549 + }, + { + "epoch": 0.275997150997151, + "grad_norm": 0.4172806441783905, + "learning_rate": 0.00019771549799229416, + "loss": 1.2166, + "step": 1550 + }, + { + "epoch": 0.2761752136752137, + "grad_norm": 0.5088075399398804, + "learning_rate": 0.0001977125221804562, + "loss": 1.1285, + "step": 1551 + }, + { + "epoch": 0.27635327635327633, + "grad_norm": 0.4728628396987915, + "learning_rate": 0.0001977095444541443, + "loss": 1.2985, + "step": 1552 + }, + { + "epoch": 0.27653133903133903, + "grad_norm": 0.4431236684322357, + "learning_rate": 0.00019770656481341684, + "loss": 1.1298, + "step": 1553 + }, + { + "epoch": 0.2767094017094017, + "grad_norm": 0.474065363407135, + "learning_rate": 0.00019770358325833223, + "loss": 1.1915, + "step": 1554 + }, + { + "epoch": 0.27688746438746437, + "grad_norm": 0.45718875527381897, + "learning_rate": 0.00019770059978894885, + "loss": 1.0626, + "step": 1555 + }, + { + "epoch": 0.27706552706552706, + "grad_norm": 0.49300211668014526, + "learning_rate": 0.00019769761440532522, + "loss": 1.0134, + "step": 1556 + }, + { + "epoch": 0.27724358974358976, + "grad_norm": 0.4389498829841614, + "learning_rate": 0.00019769462710751974, + "loss": 1.0292, + "step": 1557 + }, + { + "epoch": 0.2774216524216524, + "grad_norm": 0.47330448031425476, + "learning_rate": 0.000197691637895591, + "loss": 1.1273, + "step": 1558 + }, + { + "epoch": 0.2775997150997151, + "grad_norm": 0.5322058200836182, + "learning_rate": 0.00019768864676959755, + "loss": 1.059, + "step": 1559 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 0.4714536964893341, + "learning_rate": 0.000197685653729598, + "loss": 1.1987, + "step": 1560 + }, + { + "epoch": 0.27795584045584043, + "grad_norm": 0.48687809705734253, + "learning_rate": 0.00019768265877565097, + "loss": 1.3206, + "step": 1561 + }, + { + "epoch": 0.2781339031339031, + "grad_norm": 0.46066713333129883, + "learning_rate": 0.00019767966190781518, + "loss": 1.0845, + "step": 1562 + }, + { + "epoch": 0.2783119658119658, + "grad_norm": 0.44372090697288513, + "learning_rate": 0.00019767666312614935, + "loss": 1.0942, + "step": 1563 + }, + { + "epoch": 0.27849002849002846, + "grad_norm": 0.4615907073020935, + "learning_rate": 0.00019767366243071216, + "loss": 1.071, + "step": 1564 + }, + { + "epoch": 0.27866809116809116, + "grad_norm": 0.502097487449646, + "learning_rate": 0.0001976706598215625, + "loss": 1.1164, + "step": 1565 + }, + { + "epoch": 0.27884615384615385, + "grad_norm": 0.4371815621852875, + "learning_rate": 0.00019766765529875913, + "loss": 1.0252, + "step": 1566 + }, + { + "epoch": 0.27902421652421655, + "grad_norm": 0.43035808205604553, + "learning_rate": 0.00019766464886236093, + "loss": 1.073, + "step": 1567 + }, + { + "epoch": 0.2792022792022792, + "grad_norm": 0.49721601605415344, + "learning_rate": 0.00019766164051242683, + "loss": 1.0316, + "step": 1568 + }, + { + "epoch": 0.2793803418803419, + "grad_norm": 0.44866231083869934, + "learning_rate": 0.00019765863024901576, + "loss": 1.0951, + "step": 1569 + }, + { + "epoch": 0.2795584045584046, + "grad_norm": 0.46318337321281433, + "learning_rate": 0.0001976556180721867, + "loss": 0.9836, + "step": 1570 + }, + { + "epoch": 0.2797364672364672, + "grad_norm": 0.4227696657180786, + "learning_rate": 0.00019765260398199868, + "loss": 1.0414, + "step": 1571 + }, + { + "epoch": 0.2799145299145299, + "grad_norm": 0.6062980890274048, + "learning_rate": 0.00019764958797851073, + "loss": 1.137, + "step": 1572 + }, + { + "epoch": 0.2800925925925926, + "grad_norm": 0.4856833219528198, + "learning_rate": 0.00019764657006178196, + "loss": 1.1361, + "step": 1573 + }, + { + "epoch": 0.28027065527065526, + "grad_norm": 0.45612895488739014, + "learning_rate": 0.00019764355023187146, + "loss": 1.0005, + "step": 1574 + }, + { + "epoch": 0.28044871794871795, + "grad_norm": 0.4143696129322052, + "learning_rate": 0.00019764052848883845, + "loss": 1.051, + "step": 1575 + }, + { + "epoch": 0.28062678062678065, + "grad_norm": 0.4532071352005005, + "learning_rate": 0.00019763750483274212, + "loss": 1.0595, + "step": 1576 + }, + { + "epoch": 0.2808048433048433, + "grad_norm": 0.4940357208251953, + "learning_rate": 0.0001976344792636417, + "loss": 1.0983, + "step": 1577 + }, + { + "epoch": 0.280982905982906, + "grad_norm": 0.44405099749565125, + "learning_rate": 0.0001976314517815965, + "loss": 1.0846, + "step": 1578 + }, + { + "epoch": 0.2811609686609687, + "grad_norm": 0.5508625507354736, + "learning_rate": 0.00019762842238666578, + "loss": 1.1722, + "step": 1579 + }, + { + "epoch": 0.2813390313390313, + "grad_norm": 0.5241084694862366, + "learning_rate": 0.00019762539107890894, + "loss": 1.351, + "step": 1580 + }, + { + "epoch": 0.281517094017094, + "grad_norm": 0.5307353734970093, + "learning_rate": 0.00019762235785838537, + "loss": 1.1868, + "step": 1581 + }, + { + "epoch": 0.2816951566951567, + "grad_norm": 0.45697924494743347, + "learning_rate": 0.00019761932272515447, + "loss": 1.1982, + "step": 1582 + }, + { + "epoch": 0.28187321937321935, + "grad_norm": 0.412483811378479, + "learning_rate": 0.00019761628567927574, + "loss": 1.0433, + "step": 1583 + }, + { + "epoch": 0.28205128205128205, + "grad_norm": 0.4614165425300598, + "learning_rate": 0.00019761324672080868, + "loss": 1.104, + "step": 1584 + }, + { + "epoch": 0.28222934472934474, + "grad_norm": 0.47644901275634766, + "learning_rate": 0.00019761020584981284, + "loss": 1.1037, + "step": 1585 + }, + { + "epoch": 0.2824074074074074, + "grad_norm": 0.4985184669494629, + "learning_rate": 0.00019760716306634773, + "loss": 1.2213, + "step": 1586 + }, + { + "epoch": 0.2825854700854701, + "grad_norm": 0.508301317691803, + "learning_rate": 0.00019760411837047305, + "loss": 1.1315, + "step": 1587 + }, + { + "epoch": 0.2827635327635328, + "grad_norm": 0.5346587300300598, + "learning_rate": 0.00019760107176224845, + "loss": 1.2281, + "step": 1588 + }, + { + "epoch": 0.2829415954415954, + "grad_norm": 0.5106825232505798, + "learning_rate": 0.00019759802324173357, + "loss": 1.2904, + "step": 1589 + }, + { + "epoch": 0.2831196581196581, + "grad_norm": 0.46458688378334045, + "learning_rate": 0.00019759497280898817, + "loss": 1.0861, + "step": 1590 + }, + { + "epoch": 0.2832977207977208, + "grad_norm": 0.49115365743637085, + "learning_rate": 0.00019759192046407201, + "loss": 1.0529, + "step": 1591 + }, + { + "epoch": 0.28347578347578345, + "grad_norm": 0.5114167332649231, + "learning_rate": 0.0001975888662070449, + "loss": 1.2555, + "step": 1592 + }, + { + "epoch": 0.28365384615384615, + "grad_norm": 0.45844775438308716, + "learning_rate": 0.0001975858100379667, + "loss": 1.0662, + "step": 1593 + }, + { + "epoch": 0.28383190883190884, + "grad_norm": 0.4684161841869354, + "learning_rate": 0.00019758275195689727, + "loss": 1.0537, + "step": 1594 + }, + { + "epoch": 0.28400997150997154, + "grad_norm": 0.4816220998764038, + "learning_rate": 0.0001975796919638965, + "loss": 1.126, + "step": 1595 + }, + { + "epoch": 0.2841880341880342, + "grad_norm": 0.46578118205070496, + "learning_rate": 0.0001975766300590244, + "loss": 0.9651, + "step": 1596 + }, + { + "epoch": 0.2843660968660969, + "grad_norm": 0.4181675612926483, + "learning_rate": 0.0001975735662423409, + "loss": 1.0888, + "step": 1597 + }, + { + "epoch": 0.28454415954415957, + "grad_norm": 0.49417954683303833, + "learning_rate": 0.00019757050051390609, + "loss": 1.1878, + "step": 1598 + }, + { + "epoch": 0.2847222222222222, + "grad_norm": 0.47264960408210754, + "learning_rate": 0.00019756743287377998, + "loss": 1.027, + "step": 1599 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 0.47686338424682617, + "learning_rate": 0.0001975643633220227, + "loss": 1.1307, + "step": 1600 + }, + { + "epoch": 0.2850783475783476, + "grad_norm": 0.5571266412734985, + "learning_rate": 0.00019756129185869443, + "loss": 0.984, + "step": 1601 + }, + { + "epoch": 0.28525641025641024, + "grad_norm": 0.46942809224128723, + "learning_rate": 0.00019755821848385527, + "loss": 1.0397, + "step": 1602 + }, + { + "epoch": 0.28543447293447294, + "grad_norm": 0.6325890421867371, + "learning_rate": 0.00019755514319756551, + "loss": 1.0918, + "step": 1603 + }, + { + "epoch": 0.28561253561253563, + "grad_norm": 0.5297608375549316, + "learning_rate": 0.00019755206599988533, + "loss": 0.9911, + "step": 1604 + }, + { + "epoch": 0.2857905982905983, + "grad_norm": 0.4736945331096649, + "learning_rate": 0.00019754898689087512, + "loss": 1.0786, + "step": 1605 + }, + { + "epoch": 0.28596866096866097, + "grad_norm": 0.5048685669898987, + "learning_rate": 0.00019754590587059512, + "loss": 0.9834, + "step": 1606 + }, + { + "epoch": 0.28614672364672367, + "grad_norm": 0.3823149502277374, + "learning_rate": 0.00019754282293910574, + "loss": 0.8341, + "step": 1607 + }, + { + "epoch": 0.2863247863247863, + "grad_norm": 0.44071945548057556, + "learning_rate": 0.00019753973809646738, + "loss": 1.131, + "step": 1608 + }, + { + "epoch": 0.286502849002849, + "grad_norm": 0.44182759523391724, + "learning_rate": 0.00019753665134274043, + "loss": 1.0321, + "step": 1609 + }, + { + "epoch": 0.2866809116809117, + "grad_norm": 0.4486250877380371, + "learning_rate": 0.00019753356267798546, + "loss": 0.9941, + "step": 1610 + }, + { + "epoch": 0.28685897435897434, + "grad_norm": 0.42796584963798523, + "learning_rate": 0.00019753047210226292, + "loss": 1.0235, + "step": 1611 + }, + { + "epoch": 0.28703703703703703, + "grad_norm": 0.47294023633003235, + "learning_rate": 0.00019752737961563336, + "loss": 1.11, + "step": 1612 + }, + { + "epoch": 0.28721509971509973, + "grad_norm": 0.44550734758377075, + "learning_rate": 0.00019752428521815742, + "loss": 1.0849, + "step": 1613 + }, + { + "epoch": 0.28739316239316237, + "grad_norm": 0.44189929962158203, + "learning_rate": 0.0001975211889098957, + "loss": 0.8904, + "step": 1614 + }, + { + "epoch": 0.28757122507122507, + "grad_norm": 0.5302733182907104, + "learning_rate": 0.00019751809069090885, + "loss": 1.2348, + "step": 1615 + }, + { + "epoch": 0.28774928774928776, + "grad_norm": 0.5951390862464905, + "learning_rate": 0.00019751499056125762, + "loss": 1.3035, + "step": 1616 + }, + { + "epoch": 0.2879273504273504, + "grad_norm": 0.5431534647941589, + "learning_rate": 0.0001975118885210027, + "loss": 1.0016, + "step": 1617 + }, + { + "epoch": 0.2881054131054131, + "grad_norm": 0.47301986813545227, + "learning_rate": 0.00019750878457020489, + "loss": 1.2245, + "step": 1618 + }, + { + "epoch": 0.2882834757834758, + "grad_norm": 0.44785359501838684, + "learning_rate": 0.00019750567870892497, + "loss": 1.122, + "step": 1619 + }, + { + "epoch": 0.28846153846153844, + "grad_norm": 0.49494361877441406, + "learning_rate": 0.00019750257093722383, + "loss": 0.9421, + "step": 1620 + }, + { + "epoch": 0.28863960113960113, + "grad_norm": 0.4484521150588989, + "learning_rate": 0.00019749946125516242, + "loss": 1.2146, + "step": 1621 + }, + { + "epoch": 0.28881766381766383, + "grad_norm": 0.4635269343852997, + "learning_rate": 0.00019749634966280156, + "loss": 0.976, + "step": 1622 + }, + { + "epoch": 0.28899572649572647, + "grad_norm": 0.5532249808311462, + "learning_rate": 0.00019749323616020226, + "loss": 1.1818, + "step": 1623 + }, + { + "epoch": 0.28917378917378916, + "grad_norm": 0.4730629622936249, + "learning_rate": 0.00019749012074742552, + "loss": 1.0321, + "step": 1624 + }, + { + "epoch": 0.28935185185185186, + "grad_norm": 0.47437289357185364, + "learning_rate": 0.0001974870034245324, + "loss": 1.1572, + "step": 1625 + }, + { + "epoch": 0.28952991452991456, + "grad_norm": 0.4796304404735565, + "learning_rate": 0.00019748388419158394, + "loss": 1.1667, + "step": 1626 + }, + { + "epoch": 0.2897079772079772, + "grad_norm": 0.42686304450035095, + "learning_rate": 0.0001974807630486413, + "loss": 0.9824, + "step": 1627 + }, + { + "epoch": 0.2898860398860399, + "grad_norm": 0.4444865584373474, + "learning_rate": 0.00019747763999576558, + "loss": 1.2789, + "step": 1628 + }, + { + "epoch": 0.2900641025641026, + "grad_norm": 0.5039985179901123, + "learning_rate": 0.000197474515033018, + "loss": 1.1488, + "step": 1629 + }, + { + "epoch": 0.29024216524216523, + "grad_norm": 0.581479549407959, + "learning_rate": 0.00019747138816045978, + "loss": 1.1232, + "step": 1630 + }, + { + "epoch": 0.2904202279202279, + "grad_norm": 0.5415821075439453, + "learning_rate": 0.00019746825937815222, + "loss": 1.2326, + "step": 1631 + }, + { + "epoch": 0.2905982905982906, + "grad_norm": 0.45528364181518555, + "learning_rate": 0.00019746512868615656, + "loss": 1.0246, + "step": 1632 + }, + { + "epoch": 0.29077635327635326, + "grad_norm": 0.5255574584007263, + "learning_rate": 0.00019746199608453418, + "loss": 1.0592, + "step": 1633 + }, + { + "epoch": 0.29095441595441596, + "grad_norm": 0.5064096450805664, + "learning_rate": 0.00019745886157334646, + "loss": 1.3439, + "step": 1634 + }, + { + "epoch": 0.29113247863247865, + "grad_norm": 0.500848650932312, + "learning_rate": 0.00019745572515265475, + "loss": 1.1212, + "step": 1635 + }, + { + "epoch": 0.2913105413105413, + "grad_norm": 0.5229088068008423, + "learning_rate": 0.00019745258682252062, + "loss": 1.1019, + "step": 1636 + }, + { + "epoch": 0.291488603988604, + "grad_norm": 0.4494398832321167, + "learning_rate": 0.00019744944658300545, + "loss": 1.1298, + "step": 1637 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 0.48383277654647827, + "learning_rate": 0.00019744630443417082, + "loss": 1.206, + "step": 1638 + }, + { + "epoch": 0.2918447293447293, + "grad_norm": 0.4870131313800812, + "learning_rate": 0.00019744316037607828, + "loss": 1.2096, + "step": 1639 + }, + { + "epoch": 0.292022792022792, + "grad_norm": 0.4153090715408325, + "learning_rate": 0.00019744001440878944, + "loss": 1.0478, + "step": 1640 + }, + { + "epoch": 0.2922008547008547, + "grad_norm": 0.4262249171733856, + "learning_rate": 0.0001974368665323659, + "loss": 1.0393, + "step": 1641 + }, + { + "epoch": 0.29237891737891736, + "grad_norm": 0.46131134033203125, + "learning_rate": 0.00019743371674686938, + "loss": 1.0908, + "step": 1642 + }, + { + "epoch": 0.29255698005698005, + "grad_norm": 0.44877463579177856, + "learning_rate": 0.0001974305650523616, + "loss": 1.1906, + "step": 1643 + }, + { + "epoch": 0.29273504273504275, + "grad_norm": 0.5199326276779175, + "learning_rate": 0.00019742741144890432, + "loss": 1.1147, + "step": 1644 + }, + { + "epoch": 0.2929131054131054, + "grad_norm": 0.48142504692077637, + "learning_rate": 0.00019742425593655924, + "loss": 1.1951, + "step": 1645 + }, + { + "epoch": 0.2930911680911681, + "grad_norm": 0.5672988891601562, + "learning_rate": 0.0001974210985153883, + "loss": 1.1817, + "step": 1646 + }, + { + "epoch": 0.2932692307692308, + "grad_norm": 0.38135233521461487, + "learning_rate": 0.00019741793918545326, + "loss": 0.8567, + "step": 1647 + }, + { + "epoch": 0.2934472934472934, + "grad_norm": 0.6153588891029358, + "learning_rate": 0.0001974147779468161, + "loss": 1.0593, + "step": 1648 + }, + { + "epoch": 0.2936253561253561, + "grad_norm": 0.38935527205467224, + "learning_rate": 0.0001974116147995387, + "loss": 0.9907, + "step": 1649 + }, + { + "epoch": 0.2938034188034188, + "grad_norm": 0.467351496219635, + "learning_rate": 0.0001974084497436831, + "loss": 1.091, + "step": 1650 + }, + { + "epoch": 0.29398148148148145, + "grad_norm": 0.45613420009613037, + "learning_rate": 0.00019740528277931128, + "loss": 0.6789, + "step": 1651 + }, + { + "epoch": 0.29415954415954415, + "grad_norm": 0.4045158326625824, + "learning_rate": 0.00019740211390648524, + "loss": 1.0727, + "step": 1652 + }, + { + "epoch": 0.29433760683760685, + "grad_norm": 0.5122803449630737, + "learning_rate": 0.00019739894312526714, + "loss": 1.2297, + "step": 1653 + }, + { + "epoch": 0.29451566951566954, + "grad_norm": 0.44304123520851135, + "learning_rate": 0.00019739577043571908, + "loss": 0.9562, + "step": 1654 + }, + { + "epoch": 0.2946937321937322, + "grad_norm": 0.6070618629455566, + "learning_rate": 0.00019739259583790322, + "loss": 1.2745, + "step": 1655 + }, + { + "epoch": 0.2948717948717949, + "grad_norm": 0.48815637826919556, + "learning_rate": 0.00019738941933188176, + "loss": 1.0574, + "step": 1656 + }, + { + "epoch": 0.2950498575498576, + "grad_norm": 0.5067802667617798, + "learning_rate": 0.00019738624091771693, + "loss": 1.1874, + "step": 1657 + }, + { + "epoch": 0.2952279202279202, + "grad_norm": 0.4956928491592407, + "learning_rate": 0.000197383060595471, + "loss": 1.1085, + "step": 1658 + }, + { + "epoch": 0.2954059829059829, + "grad_norm": 0.46313008666038513, + "learning_rate": 0.00019737987836520633, + "loss": 1.0548, + "step": 1659 + }, + { + "epoch": 0.2955840455840456, + "grad_norm": 0.49944064021110535, + "learning_rate": 0.0001973766942269852, + "loss": 1.1485, + "step": 1660 + }, + { + "epoch": 0.29576210826210825, + "grad_norm": 0.4743517339229584, + "learning_rate": 0.00019737350818087003, + "loss": 0.9279, + "step": 1661 + }, + { + "epoch": 0.29594017094017094, + "grad_norm": 0.45935431122779846, + "learning_rate": 0.00019737032022692326, + "loss": 0.9574, + "step": 1662 + }, + { + "epoch": 0.29611823361823364, + "grad_norm": 0.4550873637199402, + "learning_rate": 0.00019736713036520734, + "loss": 1.1642, + "step": 1663 + }, + { + "epoch": 0.2962962962962963, + "grad_norm": 0.45252951979637146, + "learning_rate": 0.00019736393859578474, + "loss": 1.0113, + "step": 1664 + }, + { + "epoch": 0.296474358974359, + "grad_norm": 0.5147238969802856, + "learning_rate": 0.00019736074491871804, + "loss": 1.1604, + "step": 1665 + }, + { + "epoch": 0.29665242165242167, + "grad_norm": 0.5122934579849243, + "learning_rate": 0.00019735754933406977, + "loss": 0.9525, + "step": 1666 + }, + { + "epoch": 0.2968304843304843, + "grad_norm": 0.438620001077652, + "learning_rate": 0.00019735435184190257, + "loss": 1.0728, + "step": 1667 + }, + { + "epoch": 0.297008547008547, + "grad_norm": 0.41970670223236084, + "learning_rate": 0.00019735115244227908, + "loss": 0.9782, + "step": 1668 + }, + { + "epoch": 0.2971866096866097, + "grad_norm": 0.5447152256965637, + "learning_rate": 0.000197347951135262, + "loss": 1.0633, + "step": 1669 + }, + { + "epoch": 0.29736467236467234, + "grad_norm": 0.4846996068954468, + "learning_rate": 0.00019734474792091407, + "loss": 0.9019, + "step": 1670 + }, + { + "epoch": 0.29754273504273504, + "grad_norm": 0.4721437990665436, + "learning_rate": 0.00019734154279929796, + "loss": 1.1793, + "step": 1671 + }, + { + "epoch": 0.29772079772079774, + "grad_norm": 0.4659852385520935, + "learning_rate": 0.00019733833577047655, + "loss": 1.1503, + "step": 1672 + }, + { + "epoch": 0.2978988603988604, + "grad_norm": 0.3733183443546295, + "learning_rate": 0.00019733512683451268, + "loss": 0.7763, + "step": 1673 + }, + { + "epoch": 0.2980769230769231, + "grad_norm": 0.4898292124271393, + "learning_rate": 0.0001973319159914692, + "loss": 1.3146, + "step": 1674 + }, + { + "epoch": 0.29825498575498577, + "grad_norm": 0.41774725914001465, + "learning_rate": 0.00019732870324140899, + "loss": 1.2069, + "step": 1675 + }, + { + "epoch": 0.2984330484330484, + "grad_norm": 0.4607912003993988, + "learning_rate": 0.000197325488584395, + "loss": 1.2255, + "step": 1676 + }, + { + "epoch": 0.2986111111111111, + "grad_norm": 0.4692424237728119, + "learning_rate": 0.00019732227202049025, + "loss": 1.0793, + "step": 1677 + }, + { + "epoch": 0.2987891737891738, + "grad_norm": 0.5925022959709167, + "learning_rate": 0.00019731905354975778, + "loss": 1.0297, + "step": 1678 + }, + { + "epoch": 0.29896723646723644, + "grad_norm": 0.44047990441322327, + "learning_rate": 0.00019731583317226056, + "loss": 1.0982, + "step": 1679 + }, + { + "epoch": 0.29914529914529914, + "grad_norm": 0.5863066911697388, + "learning_rate": 0.0001973126108880618, + "loss": 1.0284, + "step": 1680 + }, + { + "epoch": 0.29932336182336183, + "grad_norm": 0.48962152004241943, + "learning_rate": 0.00019730938669722457, + "loss": 1.1861, + "step": 1681 + }, + { + "epoch": 0.29950142450142453, + "grad_norm": 0.5445577502250671, + "learning_rate": 0.00019730616059981205, + "loss": 1.2574, + "step": 1682 + }, + { + "epoch": 0.29967948717948717, + "grad_norm": 0.49327564239501953, + "learning_rate": 0.00019730293259588743, + "loss": 0.9578, + "step": 1683 + }, + { + "epoch": 0.29985754985754987, + "grad_norm": 0.4252840578556061, + "learning_rate": 0.00019729970268551398, + "loss": 1.0083, + "step": 1684 + }, + { + "epoch": 0.30003561253561256, + "grad_norm": 0.5140926241874695, + "learning_rate": 0.000197296470868755, + "loss": 1.3263, + "step": 1685 + }, + { + "epoch": 0.3002136752136752, + "grad_norm": 0.5143948197364807, + "learning_rate": 0.00019729323714567375, + "loss": 1.0424, + "step": 1686 + }, + { + "epoch": 0.3003917378917379, + "grad_norm": 0.3811354339122772, + "learning_rate": 0.00019729000151633367, + "loss": 0.6319, + "step": 1687 + }, + { + "epoch": 0.3005698005698006, + "grad_norm": 0.5249716639518738, + "learning_rate": 0.0001972867639807981, + "loss": 1.0173, + "step": 1688 + }, + { + "epoch": 0.30074786324786323, + "grad_norm": 0.41832098364830017, + "learning_rate": 0.00019728352453913048, + "loss": 1.0503, + "step": 1689 + }, + { + "epoch": 0.30092592592592593, + "grad_norm": 0.5961149334907532, + "learning_rate": 0.00019728028319139428, + "loss": 1.1843, + "step": 1690 + }, + { + "epoch": 0.3011039886039886, + "grad_norm": 0.44083690643310547, + "learning_rate": 0.00019727703993765303, + "loss": 1.1311, + "step": 1691 + }, + { + "epoch": 0.30128205128205127, + "grad_norm": 0.4368111491203308, + "learning_rate": 0.00019727379477797022, + "loss": 0.9463, + "step": 1692 + }, + { + "epoch": 0.30146011396011396, + "grad_norm": 0.5289376974105835, + "learning_rate": 0.00019727054771240954, + "loss": 0.9836, + "step": 1693 + }, + { + "epoch": 0.30163817663817666, + "grad_norm": 0.4132843613624573, + "learning_rate": 0.00019726729874103448, + "loss": 1.1052, + "step": 1694 + }, + { + "epoch": 0.3018162393162393, + "grad_norm": 0.4919086992740631, + "learning_rate": 0.00019726404786390877, + "loss": 1.2219, + "step": 1695 + }, + { + "epoch": 0.301994301994302, + "grad_norm": 0.42561691999435425, + "learning_rate": 0.0001972607950810961, + "loss": 1.0756, + "step": 1696 + }, + { + "epoch": 0.3021723646723647, + "grad_norm": 0.5030396580696106, + "learning_rate": 0.0001972575403926602, + "loss": 1.2207, + "step": 1697 + }, + { + "epoch": 0.30235042735042733, + "grad_norm": 0.4779801666736603, + "learning_rate": 0.0001972542837986648, + "loss": 1.194, + "step": 1698 + }, + { + "epoch": 0.30252849002849, + "grad_norm": 0.45395568013191223, + "learning_rate": 0.00019725102529917377, + "loss": 1.0775, + "step": 1699 + }, + { + "epoch": 0.3027065527065527, + "grad_norm": 0.6540699005126953, + "learning_rate": 0.0001972477648942509, + "loss": 1.181, + "step": 1700 + }, + { + "epoch": 0.30288461538461536, + "grad_norm": 0.46281275153160095, + "learning_rate": 0.00019724450258396008, + "loss": 0.629, + "step": 1701 + }, + { + "epoch": 0.30306267806267806, + "grad_norm": 0.3452845811843872, + "learning_rate": 0.00019724123836836527, + "loss": 0.51, + "step": 1702 + }, + { + "epoch": 0.30324074074074076, + "grad_norm": 0.4507991671562195, + "learning_rate": 0.00019723797224753038, + "loss": 1.0258, + "step": 1703 + }, + { + "epoch": 0.3034188034188034, + "grad_norm": 0.5385412573814392, + "learning_rate": 0.0001972347042215194, + "loss": 1.0232, + "step": 1704 + }, + { + "epoch": 0.3035968660968661, + "grad_norm": 0.4460466504096985, + "learning_rate": 0.00019723143429039642, + "loss": 1.1307, + "step": 1705 + }, + { + "epoch": 0.3037749287749288, + "grad_norm": 0.5229718685150146, + "learning_rate": 0.00019722816245422545, + "loss": 1.0964, + "step": 1706 + }, + { + "epoch": 0.30395299145299143, + "grad_norm": 0.4776979088783264, + "learning_rate": 0.00019722488871307058, + "loss": 1.2678, + "step": 1707 + }, + { + "epoch": 0.3041310541310541, + "grad_norm": 0.5371831655502319, + "learning_rate": 0.00019722161306699601, + "loss": 1.2808, + "step": 1708 + }, + { + "epoch": 0.3043091168091168, + "grad_norm": 0.45322108268737793, + "learning_rate": 0.0001972183355160659, + "loss": 1.0775, + "step": 1709 + }, + { + "epoch": 0.30448717948717946, + "grad_norm": 0.5036569833755493, + "learning_rate": 0.00019721505606034448, + "loss": 1.1859, + "step": 1710 + }, + { + "epoch": 0.30466524216524216, + "grad_norm": 0.5425969958305359, + "learning_rate": 0.00019721177469989593, + "loss": 1.0173, + "step": 1711 + }, + { + "epoch": 0.30484330484330485, + "grad_norm": 0.5638980269432068, + "learning_rate": 0.00019720849143478462, + "loss": 1.182, + "step": 1712 + }, + { + "epoch": 0.30502136752136755, + "grad_norm": 0.5160546898841858, + "learning_rate": 0.00019720520626507486, + "loss": 0.9853, + "step": 1713 + }, + { + "epoch": 0.3051994301994302, + "grad_norm": 0.5079004168510437, + "learning_rate": 0.000197201919190831, + "loss": 1.3154, + "step": 1714 + }, + { + "epoch": 0.3053774928774929, + "grad_norm": 0.4590355455875397, + "learning_rate": 0.00019719863021211745, + "loss": 1.007, + "step": 1715 + }, + { + "epoch": 0.3055555555555556, + "grad_norm": 0.49656423926353455, + "learning_rate": 0.00019719533932899865, + "loss": 1.2187, + "step": 1716 + }, + { + "epoch": 0.3057336182336182, + "grad_norm": 0.46426209807395935, + "learning_rate": 0.0001971920465415391, + "loss": 1.3007, + "step": 1717 + }, + { + "epoch": 0.3059116809116809, + "grad_norm": 0.5211917757987976, + "learning_rate": 0.00019718875184980328, + "loss": 1.2256, + "step": 1718 + }, + { + "epoch": 0.3060897435897436, + "grad_norm": 0.42953309416770935, + "learning_rate": 0.00019718545525385578, + "loss": 1.2838, + "step": 1719 + }, + { + "epoch": 0.30626780626780625, + "grad_norm": 0.4893105924129486, + "learning_rate": 0.00019718215675376116, + "loss": 1.052, + "step": 1720 + }, + { + "epoch": 0.30644586894586895, + "grad_norm": 0.4833602011203766, + "learning_rate": 0.00019717885634958405, + "loss": 1.069, + "step": 1721 + }, + { + "epoch": 0.30662393162393164, + "grad_norm": 0.502176821231842, + "learning_rate": 0.0001971755540413891, + "loss": 1.1659, + "step": 1722 + }, + { + "epoch": 0.3068019943019943, + "grad_norm": 0.4648856818675995, + "learning_rate": 0.00019717224982924108, + "loss": 1.1873, + "step": 1723 + }, + { + "epoch": 0.306980056980057, + "grad_norm": 0.405429869890213, + "learning_rate": 0.00019716894371320465, + "loss": 0.99, + "step": 1724 + }, + { + "epoch": 0.3071581196581197, + "grad_norm": 0.4306945204734802, + "learning_rate": 0.00019716563569334463, + "loss": 0.8751, + "step": 1725 + }, + { + "epoch": 0.3073361823361823, + "grad_norm": 0.49424824118614197, + "learning_rate": 0.00019716232576972583, + "loss": 0.9205, + "step": 1726 + }, + { + "epoch": 0.307514245014245, + "grad_norm": 0.5044034123420715, + "learning_rate": 0.00019715901394241306, + "loss": 1.2042, + "step": 1727 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 0.512180507183075, + "learning_rate": 0.00019715570021147126, + "loss": 1.1644, + "step": 1728 + }, + { + "epoch": 0.30787037037037035, + "grad_norm": 0.4377981126308441, + "learning_rate": 0.00019715238457696538, + "loss": 1.1625, + "step": 1729 + }, + { + "epoch": 0.30804843304843305, + "grad_norm": 0.49107855558395386, + "learning_rate": 0.00019714906703896027, + "loss": 1.1037, + "step": 1730 + }, + { + "epoch": 0.30822649572649574, + "grad_norm": 0.47342559695243835, + "learning_rate": 0.00019714574759752105, + "loss": 1.3186, + "step": 1731 + }, + { + "epoch": 0.3084045584045584, + "grad_norm": 0.487177312374115, + "learning_rate": 0.0001971424262527127, + "loss": 1.1196, + "step": 1732 + }, + { + "epoch": 0.3085826210826211, + "grad_norm": 0.5290025472640991, + "learning_rate": 0.0001971391030046003, + "loss": 1.2103, + "step": 1733 + }, + { + "epoch": 0.3087606837606838, + "grad_norm": 0.4587760269641876, + "learning_rate": 0.00019713577785324896, + "loss": 1.1017, + "step": 1734 + }, + { + "epoch": 0.3089387464387464, + "grad_norm": 0.45323294401168823, + "learning_rate": 0.00019713245079872388, + "loss": 1.0, + "step": 1735 + }, + { + "epoch": 0.3091168091168091, + "grad_norm": 0.43414804339408875, + "learning_rate": 0.00019712912184109013, + "loss": 1.0341, + "step": 1736 + }, + { + "epoch": 0.3092948717948718, + "grad_norm": 0.49604663252830505, + "learning_rate": 0.00019712579098041304, + "loss": 0.9437, + "step": 1737 + }, + { + "epoch": 0.30947293447293445, + "grad_norm": 0.48580703139305115, + "learning_rate": 0.00019712245821675785, + "loss": 1.2622, + "step": 1738 + }, + { + "epoch": 0.30965099715099714, + "grad_norm": 0.45333603024482727, + "learning_rate": 0.00019711912355018982, + "loss": 1.2063, + "step": 1739 + }, + { + "epoch": 0.30982905982905984, + "grad_norm": 0.5990764498710632, + "learning_rate": 0.00019711578698077432, + "loss": 1.5097, + "step": 1740 + }, + { + "epoch": 0.31000712250712253, + "grad_norm": 0.4386102259159088, + "learning_rate": 0.0001971124485085767, + "loss": 1.1283, + "step": 1741 + }, + { + "epoch": 0.3101851851851852, + "grad_norm": 0.4476035237312317, + "learning_rate": 0.00019710910813366242, + "loss": 0.8922, + "step": 1742 + }, + { + "epoch": 0.31036324786324787, + "grad_norm": 0.5276228785514832, + "learning_rate": 0.00019710576585609685, + "loss": 1.2373, + "step": 1743 + }, + { + "epoch": 0.31054131054131057, + "grad_norm": 0.4885637164115906, + "learning_rate": 0.00019710242167594557, + "loss": 1.0881, + "step": 1744 + }, + { + "epoch": 0.3107193732193732, + "grad_norm": 0.421132355928421, + "learning_rate": 0.000197099075593274, + "loss": 1.0544, + "step": 1745 + }, + { + "epoch": 0.3108974358974359, + "grad_norm": 0.5257927179336548, + "learning_rate": 0.00019709572760814777, + "loss": 1.265, + "step": 1746 + }, + { + "epoch": 0.3110754985754986, + "grad_norm": 0.5164850950241089, + "learning_rate": 0.00019709237772063247, + "loss": 0.9593, + "step": 1747 + }, + { + "epoch": 0.31125356125356124, + "grad_norm": 0.5176383256912231, + "learning_rate": 0.00019708902593079374, + "loss": 1.0194, + "step": 1748 + }, + { + "epoch": 0.31143162393162394, + "grad_norm": 0.4620790481567383, + "learning_rate": 0.00019708567223869716, + "loss": 0.9241, + "step": 1749 + }, + { + "epoch": 0.31160968660968663, + "grad_norm": 0.48307979106903076, + "learning_rate": 0.00019708231664440854, + "loss": 1.2314, + "step": 1750 + }, + { + "epoch": 0.31178774928774927, + "grad_norm": 0.4931468069553375, + "learning_rate": 0.00019707895914799364, + "loss": 1.2065, + "step": 1751 + }, + { + "epoch": 0.31196581196581197, + "grad_norm": 0.5035979747772217, + "learning_rate": 0.00019707559974951818, + "loss": 1.1867, + "step": 1752 + }, + { + "epoch": 0.31214387464387466, + "grad_norm": 0.47543632984161377, + "learning_rate": 0.00019707223844904795, + "loss": 1.0603, + "step": 1753 + }, + { + "epoch": 0.3123219373219373, + "grad_norm": 0.49929797649383545, + "learning_rate": 0.00019706887524664892, + "loss": 1.0597, + "step": 1754 + }, + { + "epoch": 0.3125, + "grad_norm": 0.5075222253799438, + "learning_rate": 0.00019706551014238687, + "loss": 1.1398, + "step": 1755 + }, + { + "epoch": 0.3126780626780627, + "grad_norm": 0.5096884369850159, + "learning_rate": 0.00019706214313632784, + "loss": 1.1382, + "step": 1756 + }, + { + "epoch": 0.31285612535612534, + "grad_norm": 0.4629988372325897, + "learning_rate": 0.0001970587742285377, + "loss": 1.0009, + "step": 1757 + }, + { + "epoch": 0.31303418803418803, + "grad_norm": 0.5244084596633911, + "learning_rate": 0.00019705540341908253, + "loss": 1.047, + "step": 1758 + }, + { + "epoch": 0.31321225071225073, + "grad_norm": 0.5136716961860657, + "learning_rate": 0.00019705203070802832, + "loss": 1.29, + "step": 1759 + }, + { + "epoch": 0.31339031339031337, + "grad_norm": 0.43991541862487793, + "learning_rate": 0.0001970486560954412, + "loss": 0.9605, + "step": 1760 + }, + { + "epoch": 0.31356837606837606, + "grad_norm": 0.4633477032184601, + "learning_rate": 0.00019704527958138725, + "loss": 1.1507, + "step": 1761 + }, + { + "epoch": 0.31374643874643876, + "grad_norm": 0.4419999420642853, + "learning_rate": 0.00019704190116593266, + "loss": 0.9262, + "step": 1762 + }, + { + "epoch": 0.3139245014245014, + "grad_norm": 0.49359434843063354, + "learning_rate": 0.00019703852084914357, + "loss": 0.9348, + "step": 1763 + }, + { + "epoch": 0.3141025641025641, + "grad_norm": 0.5072139501571655, + "learning_rate": 0.00019703513863108627, + "loss": 1.1592, + "step": 1764 + }, + { + "epoch": 0.3142806267806268, + "grad_norm": 0.45969831943511963, + "learning_rate": 0.00019703175451182698, + "loss": 1.1519, + "step": 1765 + }, + { + "epoch": 0.31445868945868943, + "grad_norm": 0.5148758292198181, + "learning_rate": 0.00019702836849143208, + "loss": 1.1673, + "step": 1766 + }, + { + "epoch": 0.31463675213675213, + "grad_norm": 0.43033209443092346, + "learning_rate": 0.0001970249805699678, + "loss": 0.9256, + "step": 1767 + }, + { + "epoch": 0.3148148148148148, + "grad_norm": 0.48143425583839417, + "learning_rate": 0.00019702159074750058, + "loss": 1.08, + "step": 1768 + }, + { + "epoch": 0.31499287749287747, + "grad_norm": 0.4780619740486145, + "learning_rate": 0.00019701819902409685, + "loss": 1.1198, + "step": 1769 + }, + { + "epoch": 0.31517094017094016, + "grad_norm": 0.4662075936794281, + "learning_rate": 0.00019701480539982305, + "loss": 0.8424, + "step": 1770 + }, + { + "epoch": 0.31534900284900286, + "grad_norm": 0.503901481628418, + "learning_rate": 0.00019701140987474566, + "loss": 1.1026, + "step": 1771 + }, + { + "epoch": 0.31552706552706555, + "grad_norm": 0.5197132229804993, + "learning_rate": 0.00019700801244893124, + "loss": 1.2148, + "step": 1772 + }, + { + "epoch": 0.3157051282051282, + "grad_norm": 0.4746309220790863, + "learning_rate": 0.00019700461312244634, + "loss": 1.0906, + "step": 1773 + }, + { + "epoch": 0.3158831908831909, + "grad_norm": 0.5277339816093445, + "learning_rate": 0.00019700121189535752, + "loss": 1.0588, + "step": 1774 + }, + { + "epoch": 0.3160612535612536, + "grad_norm": 0.436002254486084, + "learning_rate": 0.00019699780876773147, + "loss": 1.0341, + "step": 1775 + }, + { + "epoch": 0.3162393162393162, + "grad_norm": 0.5171145796775818, + "learning_rate": 0.00019699440373963486, + "loss": 1.282, + "step": 1776 + }, + { + "epoch": 0.3164173789173789, + "grad_norm": 0.38382846117019653, + "learning_rate": 0.00019699099681113436, + "loss": 0.8908, + "step": 1777 + }, + { + "epoch": 0.3165954415954416, + "grad_norm": 0.4621630609035492, + "learning_rate": 0.0001969875879822968, + "loss": 1.1074, + "step": 1778 + }, + { + "epoch": 0.31677350427350426, + "grad_norm": 0.5543130040168762, + "learning_rate": 0.00019698417725318892, + "loss": 0.9682, + "step": 1779 + }, + { + "epoch": 0.31695156695156695, + "grad_norm": 0.49534836411476135, + "learning_rate": 0.00019698076462387753, + "loss": 1.107, + "step": 1780 + }, + { + "epoch": 0.31712962962962965, + "grad_norm": 0.48844948410987854, + "learning_rate": 0.00019697735009442956, + "loss": 1.1295, + "step": 1781 + }, + { + "epoch": 0.3173076923076923, + "grad_norm": 0.5070686936378479, + "learning_rate": 0.00019697393366491185, + "loss": 1.083, + "step": 1782 + }, + { + "epoch": 0.317485754985755, + "grad_norm": 0.47817620635032654, + "learning_rate": 0.00019697051533539134, + "loss": 1.3014, + "step": 1783 + }, + { + "epoch": 0.3176638176638177, + "grad_norm": 0.538488507270813, + "learning_rate": 0.00019696709510593502, + "loss": 1.0354, + "step": 1784 + }, + { + "epoch": 0.3178418803418803, + "grad_norm": 0.5141439437866211, + "learning_rate": 0.0001969636729766099, + "loss": 1.2912, + "step": 1785 + }, + { + "epoch": 0.318019943019943, + "grad_norm": 0.5009665489196777, + "learning_rate": 0.00019696024894748306, + "loss": 0.9014, + "step": 1786 + }, + { + "epoch": 0.3181980056980057, + "grad_norm": 0.46199744939804077, + "learning_rate": 0.00019695682301862155, + "loss": 1.0532, + "step": 1787 + }, + { + "epoch": 0.31837606837606836, + "grad_norm": 0.4649423062801361, + "learning_rate": 0.0001969533951900925, + "loss": 0.8608, + "step": 1788 + }, + { + "epoch": 0.31855413105413105, + "grad_norm": 0.516909658908844, + "learning_rate": 0.0001969499654619631, + "loss": 1.1385, + "step": 1789 + }, + { + "epoch": 0.31873219373219375, + "grad_norm": 0.46016669273376465, + "learning_rate": 0.00019694653383430048, + "loss": 0.9168, + "step": 1790 + }, + { + "epoch": 0.3189102564102564, + "grad_norm": 0.4794938564300537, + "learning_rate": 0.00019694310030717193, + "loss": 1.0244, + "step": 1791 + }, + { + "epoch": 0.3190883190883191, + "grad_norm": 0.46577662229537964, + "learning_rate": 0.00019693966488064471, + "loss": 1.0954, + "step": 1792 + }, + { + "epoch": 0.3192663817663818, + "grad_norm": 0.4866746962070465, + "learning_rate": 0.00019693622755478614, + "loss": 1.2925, + "step": 1793 + }, + { + "epoch": 0.3194444444444444, + "grad_norm": 0.4841702878475189, + "learning_rate": 0.00019693278832966357, + "loss": 1.119, + "step": 1794 + }, + { + "epoch": 0.3196225071225071, + "grad_norm": 0.4835243821144104, + "learning_rate": 0.00019692934720534435, + "loss": 1.1702, + "step": 1795 + }, + { + "epoch": 0.3198005698005698, + "grad_norm": 0.5200608968734741, + "learning_rate": 0.00019692590418189594, + "loss": 1.1989, + "step": 1796 + }, + { + "epoch": 0.31997863247863245, + "grad_norm": 0.5147821307182312, + "learning_rate": 0.00019692245925938577, + "loss": 1.1417, + "step": 1797 + }, + { + "epoch": 0.32015669515669515, + "grad_norm": 0.5145614743232727, + "learning_rate": 0.00019691901243788136, + "loss": 1.0571, + "step": 1798 + }, + { + "epoch": 0.32033475783475784, + "grad_norm": 0.5416026711463928, + "learning_rate": 0.00019691556371745022, + "loss": 1.188, + "step": 1799 + }, + { + "epoch": 0.32051282051282054, + "grad_norm": 0.5140644311904907, + "learning_rate": 0.00019691211309815995, + "loss": 1.1795, + "step": 1800 + }, + { + "epoch": 0.3206908831908832, + "grad_norm": 0.44219106435775757, + "learning_rate": 0.00019690866058007817, + "loss": 0.9215, + "step": 1801 + }, + { + "epoch": 0.3208689458689459, + "grad_norm": 0.49523603916168213, + "learning_rate": 0.00019690520616327245, + "loss": 1.1117, + "step": 1802 + }, + { + "epoch": 0.32104700854700857, + "grad_norm": 0.5818293690681458, + "learning_rate": 0.0001969017498478105, + "loss": 1.16, + "step": 1803 + }, + { + "epoch": 0.3212250712250712, + "grad_norm": 0.5175749659538269, + "learning_rate": 0.0001968982916337601, + "loss": 1.1999, + "step": 1804 + }, + { + "epoch": 0.3214031339031339, + "grad_norm": 0.49916017055511475, + "learning_rate": 0.00019689483152118898, + "loss": 0.9505, + "step": 1805 + }, + { + "epoch": 0.3215811965811966, + "grad_norm": 0.46849536895751953, + "learning_rate": 0.00019689136951016488, + "loss": 0.9627, + "step": 1806 + }, + { + "epoch": 0.32175925925925924, + "grad_norm": 0.4226818382740021, + "learning_rate": 0.00019688790560075568, + "loss": 1.037, + "step": 1807 + }, + { + "epoch": 0.32193732193732194, + "grad_norm": 0.4697103798389435, + "learning_rate": 0.00019688443979302923, + "loss": 1.1431, + "step": 1808 + }, + { + "epoch": 0.32211538461538464, + "grad_norm": 0.4999365508556366, + "learning_rate": 0.00019688097208705343, + "loss": 1.171, + "step": 1809 + }, + { + "epoch": 0.3222934472934473, + "grad_norm": 0.5229731798171997, + "learning_rate": 0.00019687750248289625, + "loss": 1.3395, + "step": 1810 + }, + { + "epoch": 0.32247150997151, + "grad_norm": 0.512525737285614, + "learning_rate": 0.00019687403098062566, + "loss": 1.1438, + "step": 1811 + }, + { + "epoch": 0.32264957264957267, + "grad_norm": 0.4558548927307129, + "learning_rate": 0.00019687055758030967, + "loss": 1.0012, + "step": 1812 + }, + { + "epoch": 0.3228276353276353, + "grad_norm": 0.45195743441581726, + "learning_rate": 0.00019686708228201636, + "loss": 1.0222, + "step": 1813 + }, + { + "epoch": 0.323005698005698, + "grad_norm": 0.5023126602172852, + "learning_rate": 0.00019686360508581373, + "loss": 1.2128, + "step": 1814 + }, + { + "epoch": 0.3231837606837607, + "grad_norm": 0.46516045928001404, + "learning_rate": 0.00019686012599177003, + "loss": 0.989, + "step": 1815 + }, + { + "epoch": 0.32336182336182334, + "grad_norm": 0.4142672121524811, + "learning_rate": 0.00019685664499995338, + "loss": 1.0144, + "step": 1816 + }, + { + "epoch": 0.32353988603988604, + "grad_norm": 0.4511009752750397, + "learning_rate": 0.0001968531621104319, + "loss": 0.885, + "step": 1817 + }, + { + "epoch": 0.32371794871794873, + "grad_norm": 0.49583545327186584, + "learning_rate": 0.00019684967732327396, + "loss": 1.0986, + "step": 1818 + }, + { + "epoch": 0.3238960113960114, + "grad_norm": 0.5872161388397217, + "learning_rate": 0.0001968461906385478, + "loss": 1.1482, + "step": 1819 + }, + { + "epoch": 0.32407407407407407, + "grad_norm": 0.4509563148021698, + "learning_rate": 0.00019684270205632168, + "loss": 1.0578, + "step": 1820 + }, + { + "epoch": 0.32425213675213677, + "grad_norm": 0.501345157623291, + "learning_rate": 0.00019683921157666402, + "loss": 1.1792, + "step": 1821 + }, + { + "epoch": 0.3244301994301994, + "grad_norm": 0.48257577419281006, + "learning_rate": 0.00019683571919964314, + "loss": 1.0448, + "step": 1822 + }, + { + "epoch": 0.3246082621082621, + "grad_norm": 0.5399422645568848, + "learning_rate": 0.00019683222492532752, + "loss": 1.0579, + "step": 1823 + }, + { + "epoch": 0.3247863247863248, + "grad_norm": 0.4382506012916565, + "learning_rate": 0.0001968287287537856, + "loss": 1.0246, + "step": 1824 + }, + { + "epoch": 0.32496438746438744, + "grad_norm": 0.49247491359710693, + "learning_rate": 0.00019682523068508586, + "loss": 1.318, + "step": 1825 + }, + { + "epoch": 0.32514245014245013, + "grad_norm": 0.49067625403404236, + "learning_rate": 0.0001968217307192969, + "loss": 1.1028, + "step": 1826 + }, + { + "epoch": 0.32532051282051283, + "grad_norm": 0.4832286238670349, + "learning_rate": 0.00019681822885648723, + "loss": 1.0996, + "step": 1827 + }, + { + "epoch": 0.32549857549857547, + "grad_norm": 0.47144386172294617, + "learning_rate": 0.0001968147250967255, + "loss": 1.0707, + "step": 1828 + }, + { + "epoch": 0.32567663817663817, + "grad_norm": 0.46299225091934204, + "learning_rate": 0.0001968112194400803, + "loss": 1.0461, + "step": 1829 + }, + { + "epoch": 0.32585470085470086, + "grad_norm": 0.4880816340446472, + "learning_rate": 0.00019680771188662044, + "loss": 1.1198, + "step": 1830 + }, + { + "epoch": 0.32603276353276356, + "grad_norm": 0.43837276101112366, + "learning_rate": 0.00019680420243641452, + "loss": 1.0599, + "step": 1831 + }, + { + "epoch": 0.3262108262108262, + "grad_norm": 0.453168660402298, + "learning_rate": 0.0001968006910895314, + "loss": 1.0327, + "step": 1832 + }, + { + "epoch": 0.3263888888888889, + "grad_norm": 0.45183828473091125, + "learning_rate": 0.00019679717784603975, + "loss": 1.1381, + "step": 1833 + }, + { + "epoch": 0.3265669515669516, + "grad_norm": 0.5326765775680542, + "learning_rate": 0.00019679366270600852, + "loss": 1.3169, + "step": 1834 + }, + { + "epoch": 0.32674501424501423, + "grad_norm": 0.47468429803848267, + "learning_rate": 0.00019679014566950653, + "loss": 1.1816, + "step": 1835 + }, + { + "epoch": 0.3269230769230769, + "grad_norm": 0.5096879005432129, + "learning_rate": 0.0001967866267366027, + "loss": 1.1162, + "step": 1836 + }, + { + "epoch": 0.3271011396011396, + "grad_norm": 0.491514652967453, + "learning_rate": 0.00019678310590736598, + "loss": 1.2793, + "step": 1837 + }, + { + "epoch": 0.32727920227920226, + "grad_norm": 0.601439356803894, + "learning_rate": 0.00019677958318186533, + "loss": 0.9851, + "step": 1838 + }, + { + "epoch": 0.32745726495726496, + "grad_norm": 0.45270970463752747, + "learning_rate": 0.0001967760585601698, + "loss": 1.0042, + "step": 1839 + }, + { + "epoch": 0.32763532763532766, + "grad_norm": 0.48864325881004333, + "learning_rate": 0.00019677253204234847, + "loss": 1.0835, + "step": 1840 + }, + { + "epoch": 0.3278133903133903, + "grad_norm": 0.5855685472488403, + "learning_rate": 0.00019676900362847037, + "loss": 1.193, + "step": 1841 + }, + { + "epoch": 0.327991452991453, + "grad_norm": 0.7181013822555542, + "learning_rate": 0.00019676547331860466, + "loss": 1.2028, + "step": 1842 + }, + { + "epoch": 0.3281695156695157, + "grad_norm": 0.4517378807067871, + "learning_rate": 0.00019676194111282054, + "loss": 1.013, + "step": 1843 + }, + { + "epoch": 0.32834757834757833, + "grad_norm": 0.5477756857872009, + "learning_rate": 0.00019675840701118718, + "loss": 1.2311, + "step": 1844 + }, + { + "epoch": 0.328525641025641, + "grad_norm": 0.5194997191429138, + "learning_rate": 0.00019675487101377382, + "loss": 1.0953, + "step": 1845 + }, + { + "epoch": 0.3287037037037037, + "grad_norm": 0.44454067945480347, + "learning_rate": 0.00019675133312064977, + "loss": 0.8505, + "step": 1846 + }, + { + "epoch": 0.32888176638176636, + "grad_norm": 0.3938713073730469, + "learning_rate": 0.00019674779333188428, + "loss": 0.8525, + "step": 1847 + }, + { + "epoch": 0.32905982905982906, + "grad_norm": 0.4927884340286255, + "learning_rate": 0.00019674425164754682, + "loss": 1.2477, + "step": 1848 + }, + { + "epoch": 0.32923789173789175, + "grad_norm": 0.4516635239124298, + "learning_rate": 0.0001967407080677067, + "loss": 0.8333, + "step": 1849 + }, + { + "epoch": 0.3294159544159544, + "grad_norm": 0.47105780243873596, + "learning_rate": 0.00019673716259243336, + "loss": 1.0989, + "step": 1850 + }, + { + "epoch": 0.3295940170940171, + "grad_norm": 0.5192127823829651, + "learning_rate": 0.00019673361522179627, + "loss": 1.1164, + "step": 1851 + }, + { + "epoch": 0.3297720797720798, + "grad_norm": 0.5222696661949158, + "learning_rate": 0.00019673006595586495, + "loss": 1.3191, + "step": 1852 + }, + { + "epoch": 0.3299501424501424, + "grad_norm": 0.6046679019927979, + "learning_rate": 0.0001967265147947089, + "loss": 0.9782, + "step": 1853 + }, + { + "epoch": 0.3301282051282051, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.00019672296173839775, + "loss": 1.2247, + "step": 1854 + }, + { + "epoch": 0.3303062678062678, + "grad_norm": 0.5435982346534729, + "learning_rate": 0.00019671940678700107, + "loss": 1.1647, + "step": 1855 + }, + { + "epoch": 0.33048433048433046, + "grad_norm": 0.46878984570503235, + "learning_rate": 0.00019671584994058856, + "loss": 1.132, + "step": 1856 + }, + { + "epoch": 0.33066239316239315, + "grad_norm": 0.5336877107620239, + "learning_rate": 0.00019671229119922986, + "loss": 1.0583, + "step": 1857 + }, + { + "epoch": 0.33084045584045585, + "grad_norm": 0.4811093807220459, + "learning_rate": 0.0001967087305629947, + "loss": 1.0089, + "step": 1858 + }, + { + "epoch": 0.33101851851851855, + "grad_norm": 0.5140184760093689, + "learning_rate": 0.0001967051680319529, + "loss": 1.2335, + "step": 1859 + }, + { + "epoch": 0.3311965811965812, + "grad_norm": 0.5855883955955505, + "learning_rate": 0.00019670160360617418, + "loss": 1.1107, + "step": 1860 + }, + { + "epoch": 0.3313746438746439, + "grad_norm": 0.5081531405448914, + "learning_rate": 0.00019669803728572844, + "loss": 1.0669, + "step": 1861 + }, + { + "epoch": 0.3315527065527066, + "grad_norm": 0.48749417066574097, + "learning_rate": 0.0001966944690706855, + "loss": 1.1465, + "step": 1862 + }, + { + "epoch": 0.3317307692307692, + "grad_norm": 0.5175687670707703, + "learning_rate": 0.00019669089896111536, + "loss": 1.254, + "step": 1863 + }, + { + "epoch": 0.3319088319088319, + "grad_norm": 0.4198860824108124, + "learning_rate": 0.0001966873269570879, + "loss": 0.9811, + "step": 1864 + }, + { + "epoch": 0.3320868945868946, + "grad_norm": 0.5220273733139038, + "learning_rate": 0.0001966837530586731, + "loss": 1.277, + "step": 1865 + }, + { + "epoch": 0.33226495726495725, + "grad_norm": 0.551954448223114, + "learning_rate": 0.00019668017726594101, + "loss": 1.0627, + "step": 1866 + }, + { + "epoch": 0.33244301994301995, + "grad_norm": 0.5289301872253418, + "learning_rate": 0.00019667659957896166, + "loss": 1.4525, + "step": 1867 + }, + { + "epoch": 0.33262108262108264, + "grad_norm": 0.5190161466598511, + "learning_rate": 0.00019667301999780522, + "loss": 1.1064, + "step": 1868 + }, + { + "epoch": 0.3327991452991453, + "grad_norm": 0.437637060880661, + "learning_rate": 0.00019666943852254172, + "loss": 1.1304, + "step": 1869 + }, + { + "epoch": 0.332977207977208, + "grad_norm": 0.4801286458969116, + "learning_rate": 0.00019666585515324138, + "loss": 1.032, + "step": 1870 + }, + { + "epoch": 0.3331552706552707, + "grad_norm": 0.5041908621788025, + "learning_rate": 0.00019666226988997445, + "loss": 1.2611, + "step": 1871 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.4529375731945038, + "learning_rate": 0.00019665868273281115, + "loss": 1.1346, + "step": 1872 + }, + { + "epoch": 0.333511396011396, + "grad_norm": 0.4797019064426422, + "learning_rate": 0.00019665509368182172, + "loss": 1.1716, + "step": 1873 + }, + { + "epoch": 0.3336894586894587, + "grad_norm": 0.5505055785179138, + "learning_rate": 0.00019665150273707652, + "loss": 0.9729, + "step": 1874 + }, + { + "epoch": 0.33386752136752135, + "grad_norm": 0.4228051006793976, + "learning_rate": 0.00019664790989864592, + "loss": 0.9023, + "step": 1875 + }, + { + "epoch": 0.33404558404558404, + "grad_norm": 0.4926959276199341, + "learning_rate": 0.00019664431516660028, + "loss": 1.0999, + "step": 1876 + }, + { + "epoch": 0.33422364672364674, + "grad_norm": 0.4273219704627991, + "learning_rate": 0.00019664071854101005, + "loss": 1.1039, + "step": 1877 + }, + { + "epoch": 0.3344017094017094, + "grad_norm": 0.48438936471939087, + "learning_rate": 0.00019663712002194566, + "loss": 1.1308, + "step": 1878 + }, + { + "epoch": 0.3345797720797721, + "grad_norm": 0.5102053284645081, + "learning_rate": 0.0001966335196094777, + "loss": 1.0618, + "step": 1879 + }, + { + "epoch": 0.33475783475783477, + "grad_norm": 0.4357300400733948, + "learning_rate": 0.00019662991730367663, + "loss": 1.0521, + "step": 1880 + }, + { + "epoch": 0.3349358974358974, + "grad_norm": 0.5052695870399475, + "learning_rate": 0.00019662631310461308, + "loss": 0.9579, + "step": 1881 + }, + { + "epoch": 0.3351139601139601, + "grad_norm": 0.4889117181301117, + "learning_rate": 0.00019662270701235762, + "loss": 1.0304, + "step": 1882 + }, + { + "epoch": 0.3352920227920228, + "grad_norm": 0.4671195149421692, + "learning_rate": 0.000196619099026981, + "loss": 1.2228, + "step": 1883 + }, + { + "epoch": 0.33547008547008544, + "grad_norm": 0.4700174331665039, + "learning_rate": 0.0001966154891485538, + "loss": 0.9634, + "step": 1884 + }, + { + "epoch": 0.33564814814814814, + "grad_norm": 0.488817423582077, + "learning_rate": 0.00019661187737714676, + "loss": 1.2499, + "step": 1885 + }, + { + "epoch": 0.33582621082621084, + "grad_norm": 0.5336169600486755, + "learning_rate": 0.00019660826371283073, + "loss": 1.251, + "step": 1886 + }, + { + "epoch": 0.33600427350427353, + "grad_norm": 0.5054540038108826, + "learning_rate": 0.00019660464815567642, + "loss": 1.221, + "step": 1887 + }, + { + "epoch": 0.33618233618233617, + "grad_norm": 0.5078747868537903, + "learning_rate": 0.00019660103070575472, + "loss": 0.9792, + "step": 1888 + }, + { + "epoch": 0.33636039886039887, + "grad_norm": 0.498571515083313, + "learning_rate": 0.0001965974113631365, + "loss": 1.1682, + "step": 1889 + }, + { + "epoch": 0.33653846153846156, + "grad_norm": 0.49969518184661865, + "learning_rate": 0.00019659379012789264, + "loss": 1.0012, + "step": 1890 + }, + { + "epoch": 0.3367165242165242, + "grad_norm": 0.4238094687461853, + "learning_rate": 0.00019659016700009416, + "loss": 1.0455, + "step": 1891 + }, + { + "epoch": 0.3368945868945869, + "grad_norm": 0.5139104723930359, + "learning_rate": 0.000196586541979812, + "loss": 0.9979, + "step": 1892 + }, + { + "epoch": 0.3370726495726496, + "grad_norm": 0.5446547269821167, + "learning_rate": 0.00019658291506711715, + "loss": 0.9271, + "step": 1893 + }, + { + "epoch": 0.33725071225071224, + "grad_norm": 0.5284572839736938, + "learning_rate": 0.00019657928626208077, + "loss": 1.0356, + "step": 1894 + }, + { + "epoch": 0.33742877492877493, + "grad_norm": 0.49936217069625854, + "learning_rate": 0.00019657565556477387, + "loss": 0.9785, + "step": 1895 + }, + { + "epoch": 0.33760683760683763, + "grad_norm": 0.4678729772567749, + "learning_rate": 0.00019657202297526763, + "loss": 1.2135, + "step": 1896 + }, + { + "epoch": 0.33778490028490027, + "grad_norm": 0.46844249963760376, + "learning_rate": 0.0001965683884936332, + "loss": 0.9369, + "step": 1897 + }, + { + "epoch": 0.33796296296296297, + "grad_norm": 0.4307389557361603, + "learning_rate": 0.0001965647521199418, + "loss": 0.9301, + "step": 1898 + }, + { + "epoch": 0.33814102564102566, + "grad_norm": 0.48227834701538086, + "learning_rate": 0.00019656111385426468, + "loss": 1.3169, + "step": 1899 + }, + { + "epoch": 0.3383190883190883, + "grad_norm": 0.45860713720321655, + "learning_rate": 0.00019655747369667315, + "loss": 0.9835, + "step": 1900 + }, + { + "epoch": 0.338497150997151, + "grad_norm": 0.5522414445877075, + "learning_rate": 0.00019655383164723846, + "loss": 1.363, + "step": 1901 + }, + { + "epoch": 0.3386752136752137, + "grad_norm": 0.5283710360527039, + "learning_rate": 0.000196550187706032, + "loss": 1.1499, + "step": 1902 + }, + { + "epoch": 0.33885327635327633, + "grad_norm": 0.4419134259223938, + "learning_rate": 0.00019654654187312525, + "loss": 1.2039, + "step": 1903 + }, + { + "epoch": 0.33903133903133903, + "grad_norm": 0.49066096544265747, + "learning_rate": 0.00019654289414858952, + "loss": 0.9707, + "step": 1904 + }, + { + "epoch": 0.3392094017094017, + "grad_norm": 0.4619338810443878, + "learning_rate": 0.00019653924453249633, + "loss": 1.0849, + "step": 1905 + }, + { + "epoch": 0.33938746438746437, + "grad_norm": 0.5191119313240051, + "learning_rate": 0.0001965355930249172, + "loss": 1.1387, + "step": 1906 + }, + { + "epoch": 0.33956552706552706, + "grad_norm": 0.5245711207389832, + "learning_rate": 0.00019653193962592368, + "loss": 1.3435, + "step": 1907 + }, + { + "epoch": 0.33974358974358976, + "grad_norm": 0.49562904238700867, + "learning_rate": 0.0001965282843355873, + "loss": 1.2781, + "step": 1908 + }, + { + "epoch": 0.3399216524216524, + "grad_norm": 0.4661353826522827, + "learning_rate": 0.0001965246271539797, + "loss": 0.9317, + "step": 1909 + }, + { + "epoch": 0.3400997150997151, + "grad_norm": 0.4723222851753235, + "learning_rate": 0.00019652096808117254, + "loss": 1.0733, + "step": 1910 + }, + { + "epoch": 0.3402777777777778, + "grad_norm": 0.4358505308628082, + "learning_rate": 0.00019651730711723754, + "loss": 1.1461, + "step": 1911 + }, + { + "epoch": 0.34045584045584043, + "grad_norm": 0.462422251701355, + "learning_rate": 0.00019651364426224638, + "loss": 1.0914, + "step": 1912 + }, + { + "epoch": 0.3406339031339031, + "grad_norm": 0.47952914237976074, + "learning_rate": 0.0001965099795162709, + "loss": 1.0392, + "step": 1913 + }, + { + "epoch": 0.3408119658119658, + "grad_norm": 0.5036373734474182, + "learning_rate": 0.00019650631287938282, + "loss": 1.4002, + "step": 1914 + }, + { + "epoch": 0.34099002849002846, + "grad_norm": 0.5130090713500977, + "learning_rate": 0.000196502644351654, + "loss": 1.3499, + "step": 1915 + }, + { + "epoch": 0.34116809116809116, + "grad_norm": 0.4426332414150238, + "learning_rate": 0.00019649897393315635, + "loss": 1.0726, + "step": 1916 + }, + { + "epoch": 0.34134615384615385, + "grad_norm": 0.5580727458000183, + "learning_rate": 0.00019649530162396176, + "loss": 1.1164, + "step": 1917 + }, + { + "epoch": 0.34152421652421655, + "grad_norm": 0.545001745223999, + "learning_rate": 0.00019649162742414218, + "loss": 0.962, + "step": 1918 + }, + { + "epoch": 0.3417022792022792, + "grad_norm": 0.5225808024406433, + "learning_rate": 0.00019648795133376962, + "loss": 1.1415, + "step": 1919 + }, + { + "epoch": 0.3418803418803419, + "grad_norm": 0.48210129141807556, + "learning_rate": 0.0001964842733529161, + "loss": 1.1188, + "step": 1920 + }, + { + "epoch": 0.3420584045584046, + "grad_norm": 0.4515395164489746, + "learning_rate": 0.00019648059348165365, + "loss": 1.0828, + "step": 1921 + }, + { + "epoch": 0.3422364672364672, + "grad_norm": 0.5802633166313171, + "learning_rate": 0.0001964769117200544, + "loss": 1.3137, + "step": 1922 + }, + { + "epoch": 0.3424145299145299, + "grad_norm": 0.4432032108306885, + "learning_rate": 0.00019647322806819046, + "loss": 1.0523, + "step": 1923 + }, + { + "epoch": 0.3425925925925926, + "grad_norm": 0.4697614908218384, + "learning_rate": 0.00019646954252613402, + "loss": 0.8426, + "step": 1924 + }, + { + "epoch": 0.34277065527065526, + "grad_norm": 0.4610968232154846, + "learning_rate": 0.0001964658550939573, + "loss": 0.9826, + "step": 1925 + }, + { + "epoch": 0.34294871794871795, + "grad_norm": 0.5278257727622986, + "learning_rate": 0.00019646216577173258, + "loss": 1.1064, + "step": 1926 + }, + { + "epoch": 0.34312678062678065, + "grad_norm": 0.5686144232749939, + "learning_rate": 0.00019645847455953205, + "loss": 0.9138, + "step": 1927 + }, + { + "epoch": 0.3433048433048433, + "grad_norm": 0.42894792556762695, + "learning_rate": 0.0001964547814574281, + "loss": 1.0461, + "step": 1928 + }, + { + "epoch": 0.343482905982906, + "grad_norm": 0.5567317605018616, + "learning_rate": 0.0001964510864654931, + "loss": 0.8787, + "step": 1929 + }, + { + "epoch": 0.3436609686609687, + "grad_norm": 0.5015586614608765, + "learning_rate": 0.0001964473895837994, + "loss": 1.1406, + "step": 1930 + }, + { + "epoch": 0.3438390313390313, + "grad_norm": 0.47391530871391296, + "learning_rate": 0.00019644369081241948, + "loss": 1.0685, + "step": 1931 + }, + { + "epoch": 0.344017094017094, + "grad_norm": 0.546037495136261, + "learning_rate": 0.00019643999015142574, + "loss": 1.2349, + "step": 1932 + }, + { + "epoch": 0.3441951566951567, + "grad_norm": 0.4724953770637512, + "learning_rate": 0.00019643628760089078, + "loss": 1.0621, + "step": 1933 + }, + { + "epoch": 0.34437321937321935, + "grad_norm": 0.5644593834877014, + "learning_rate": 0.00019643258316088703, + "loss": 1.2559, + "step": 1934 + }, + { + "epoch": 0.34455128205128205, + "grad_norm": 0.500815749168396, + "learning_rate": 0.00019642887683148718, + "loss": 1.0439, + "step": 1935 + }, + { + "epoch": 0.34472934472934474, + "grad_norm": 0.4932316541671753, + "learning_rate": 0.0001964251686127638, + "loss": 1.0404, + "step": 1936 + }, + { + "epoch": 0.3449074074074074, + "grad_norm": 0.48494651913642883, + "learning_rate": 0.00019642145850478954, + "loss": 0.9951, + "step": 1937 + }, + { + "epoch": 0.3450854700854701, + "grad_norm": 0.5191963315010071, + "learning_rate": 0.00019641774650763706, + "loss": 1.1258, + "step": 1938 + }, + { + "epoch": 0.3452635327635328, + "grad_norm": 0.4439312815666199, + "learning_rate": 0.00019641403262137918, + "loss": 1.1158, + "step": 1939 + }, + { + "epoch": 0.3454415954415954, + "grad_norm": 0.4829137921333313, + "learning_rate": 0.0001964103168460886, + "loss": 1.0531, + "step": 1940 + }, + { + "epoch": 0.3456196581196581, + "grad_norm": 0.49433329701423645, + "learning_rate": 0.00019640659918183811, + "loss": 1.1295, + "step": 1941 + }, + { + "epoch": 0.3457977207977208, + "grad_norm": 0.5351347923278809, + "learning_rate": 0.00019640287962870062, + "loss": 1.2379, + "step": 1942 + }, + { + "epoch": 0.34597578347578345, + "grad_norm": 0.4845680892467499, + "learning_rate": 0.00019639915818674895, + "loss": 1.0197, + "step": 1943 + }, + { + "epoch": 0.34615384615384615, + "grad_norm": 0.5312514901161194, + "learning_rate": 0.00019639543485605604, + "loss": 0.9734, + "step": 1944 + }, + { + "epoch": 0.34633190883190884, + "grad_norm": 0.4571874737739563, + "learning_rate": 0.00019639170963669478, + "loss": 1.1012, + "step": 1945 + }, + { + "epoch": 0.34650997150997154, + "grad_norm": 0.4449031949043274, + "learning_rate": 0.00019638798252873824, + "loss": 1.1393, + "step": 1946 + }, + { + "epoch": 0.3466880341880342, + "grad_norm": 0.47470834851264954, + "learning_rate": 0.0001963842535322594, + "loss": 0.981, + "step": 1947 + }, + { + "epoch": 0.3468660968660969, + "grad_norm": 0.5386981964111328, + "learning_rate": 0.00019638052264733132, + "loss": 1.1247, + "step": 1948 + }, + { + "epoch": 0.34704415954415957, + "grad_norm": 0.535589873790741, + "learning_rate": 0.00019637678987402714, + "loss": 1.3157, + "step": 1949 + }, + { + "epoch": 0.3472222222222222, + "grad_norm": 0.49338245391845703, + "learning_rate": 0.00019637305521242, + "loss": 1.1066, + "step": 1950 + }, + { + "epoch": 0.3474002849002849, + "grad_norm": 0.4247688353061676, + "learning_rate": 0.00019636931866258298, + "loss": 1.0039, + "step": 1951 + }, + { + "epoch": 0.3475783475783476, + "grad_norm": 0.5351517200469971, + "learning_rate": 0.00019636558022458934, + "loss": 1.0344, + "step": 1952 + }, + { + "epoch": 0.34775641025641024, + "grad_norm": 0.4633362889289856, + "learning_rate": 0.00019636183989851238, + "loss": 1.1383, + "step": 1953 + }, + { + "epoch": 0.34793447293447294, + "grad_norm": 0.553709089756012, + "learning_rate": 0.00019635809768442535, + "loss": 1.0389, + "step": 1954 + }, + { + "epoch": 0.34811253561253563, + "grad_norm": 0.479374498128891, + "learning_rate": 0.00019635435358240154, + "loss": 1.1774, + "step": 1955 + }, + { + "epoch": 0.3482905982905983, + "grad_norm": 0.5274081230163574, + "learning_rate": 0.0001963506075925143, + "loss": 1.1809, + "step": 1956 + }, + { + "epoch": 0.34846866096866097, + "grad_norm": 0.45398542284965515, + "learning_rate": 0.0001963468597148371, + "loss": 1.0502, + "step": 1957 + }, + { + "epoch": 0.34864672364672367, + "grad_norm": 0.48201611638069153, + "learning_rate": 0.00019634310994944332, + "loss": 1.0557, + "step": 1958 + }, + { + "epoch": 0.3488247863247863, + "grad_norm": 0.6407544016838074, + "learning_rate": 0.00019633935829640642, + "loss": 1.2138, + "step": 1959 + }, + { + "epoch": 0.349002849002849, + "grad_norm": 0.5385687351226807, + "learning_rate": 0.00019633560475579995, + "loss": 1.3496, + "step": 1960 + }, + { + "epoch": 0.3491809116809117, + "grad_norm": 0.5260964035987854, + "learning_rate": 0.0001963318493276974, + "loss": 1.0253, + "step": 1961 + }, + { + "epoch": 0.34935897435897434, + "grad_norm": 0.48478585481643677, + "learning_rate": 0.00019632809201217238, + "loss": 1.137, + "step": 1962 + }, + { + "epoch": 0.34953703703703703, + "grad_norm": 0.620033860206604, + "learning_rate": 0.0001963243328092985, + "loss": 1.3445, + "step": 1963 + }, + { + "epoch": 0.34971509971509973, + "grad_norm": 0.5149700045585632, + "learning_rate": 0.00019632057171914942, + "loss": 1.1042, + "step": 1964 + }, + { + "epoch": 0.34989316239316237, + "grad_norm": 0.42695048451423645, + "learning_rate": 0.0001963168087417988, + "loss": 0.8789, + "step": 1965 + }, + { + "epoch": 0.35007122507122507, + "grad_norm": 0.5281283855438232, + "learning_rate": 0.00019631304387732044, + "loss": 1.1155, + "step": 1966 + }, + { + "epoch": 0.35024928774928776, + "grad_norm": 0.4994089901447296, + "learning_rate": 0.00019630927712578804, + "loss": 1.1226, + "step": 1967 + }, + { + "epoch": 0.3504273504273504, + "grad_norm": 0.4433288276195526, + "learning_rate": 0.0001963055084872754, + "loss": 1.0262, + "step": 1968 + }, + { + "epoch": 0.3506054131054131, + "grad_norm": 0.46541857719421387, + "learning_rate": 0.0001963017379618564, + "loss": 1.1438, + "step": 1969 + }, + { + "epoch": 0.3507834757834758, + "grad_norm": 0.5097604393959045, + "learning_rate": 0.00019629796554960488, + "loss": 0.9641, + "step": 1970 + }, + { + "epoch": 0.35096153846153844, + "grad_norm": 0.49461981654167175, + "learning_rate": 0.00019629419125059478, + "loss": 1.1765, + "step": 1971 + }, + { + "epoch": 0.35113960113960113, + "grad_norm": 0.4763339161872864, + "learning_rate": 0.00019629041506490005, + "loss": 1.0527, + "step": 1972 + }, + { + "epoch": 0.35131766381766383, + "grad_norm": 0.4528443217277527, + "learning_rate": 0.00019628663699259463, + "loss": 1.1409, + "step": 1973 + }, + { + "epoch": 0.35149572649572647, + "grad_norm": 0.4436309039592743, + "learning_rate": 0.00019628285703375258, + "loss": 1.0459, + "step": 1974 + }, + { + "epoch": 0.35167378917378916, + "grad_norm": 0.5146129727363586, + "learning_rate": 0.00019627907518844797, + "loss": 1.2527, + "step": 1975 + }, + { + "epoch": 0.35185185185185186, + "grad_norm": 0.5202171802520752, + "learning_rate": 0.0001962752914567549, + "loss": 1.226, + "step": 1976 + }, + { + "epoch": 0.35202991452991456, + "grad_norm": 0.5267411470413208, + "learning_rate": 0.00019627150583874747, + "loss": 1.0898, + "step": 1977 + }, + { + "epoch": 0.3522079772079772, + "grad_norm": 0.546840250492096, + "learning_rate": 0.00019626771833449987, + "loss": 1.1716, + "step": 1978 + }, + { + "epoch": 0.3523860398860399, + "grad_norm": 0.5525290966033936, + "learning_rate": 0.0001962639289440863, + "loss": 1.1762, + "step": 1979 + }, + { + "epoch": 0.3525641025641026, + "grad_norm": 0.48967215418815613, + "learning_rate": 0.000196260137667581, + "loss": 1.1884, + "step": 1980 + }, + { + "epoch": 0.35274216524216523, + "grad_norm": 0.5908235907554626, + "learning_rate": 0.0001962563445050583, + "loss": 1.1887, + "step": 1981 + }, + { + "epoch": 0.3529202279202279, + "grad_norm": 0.46708086133003235, + "learning_rate": 0.00019625254945659245, + "loss": 0.8842, + "step": 1982 + }, + { + "epoch": 0.3530982905982906, + "grad_norm": 0.41652458906173706, + "learning_rate": 0.00019624875252225788, + "loss": 1.0268, + "step": 1983 + }, + { + "epoch": 0.35327635327635326, + "grad_norm": 0.5084529519081116, + "learning_rate": 0.00019624495370212892, + "loss": 1.0547, + "step": 1984 + }, + { + "epoch": 0.35345441595441596, + "grad_norm": 0.5667507648468018, + "learning_rate": 0.00019624115299628003, + "loss": 1.0656, + "step": 1985 + }, + { + "epoch": 0.35363247863247865, + "grad_norm": 0.5022873282432556, + "learning_rate": 0.00019623735040478568, + "loss": 1.0627, + "step": 1986 + }, + { + "epoch": 0.3538105413105413, + "grad_norm": 0.48342058062553406, + "learning_rate": 0.00019623354592772035, + "loss": 1.0976, + "step": 1987 + }, + { + "epoch": 0.353988603988604, + "grad_norm": 0.48117366433143616, + "learning_rate": 0.0001962297395651586, + "loss": 1.0515, + "step": 1988 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 0.492564857006073, + "learning_rate": 0.000196225931317175, + "loss": 1.1957, + "step": 1989 + }, + { + "epoch": 0.3543447293447293, + "grad_norm": 0.4756208658218384, + "learning_rate": 0.00019622212118384417, + "loss": 1.007, + "step": 1990 + }, + { + "epoch": 0.354522792022792, + "grad_norm": 0.581930935382843, + "learning_rate": 0.00019621830916524076, + "loss": 1.232, + "step": 1991 + }, + { + "epoch": 0.3547008547008547, + "grad_norm": 0.480064332485199, + "learning_rate": 0.00019621449526143947, + "loss": 1.2693, + "step": 1992 + }, + { + "epoch": 0.35487891737891736, + "grad_norm": 0.5679123401641846, + "learning_rate": 0.000196210679472515, + "loss": 1.2985, + "step": 1993 + }, + { + "epoch": 0.35505698005698005, + "grad_norm": 0.43757280707359314, + "learning_rate": 0.00019620686179854213, + "loss": 1.1387, + "step": 1994 + }, + { + "epoch": 0.35523504273504275, + "grad_norm": 0.4950634837150574, + "learning_rate": 0.00019620304223959566, + "loss": 1.1809, + "step": 1995 + }, + { + "epoch": 0.3554131054131054, + "grad_norm": 0.5574113726615906, + "learning_rate": 0.00019619922079575043, + "loss": 1.2434, + "step": 1996 + }, + { + "epoch": 0.3555911680911681, + "grad_norm": 0.5154930949211121, + "learning_rate": 0.00019619539746708128, + "loss": 1.1747, + "step": 1997 + }, + { + "epoch": 0.3557692307692308, + "grad_norm": 0.4377825856208801, + "learning_rate": 0.00019619157225366315, + "loss": 0.9547, + "step": 1998 + }, + { + "epoch": 0.3559472934472934, + "grad_norm": 0.530714213848114, + "learning_rate": 0.00019618774515557097, + "loss": 1.2057, + "step": 1999 + }, + { + "epoch": 0.3561253561253561, + "grad_norm": 0.5703464150428772, + "learning_rate": 0.00019618391617287978, + "loss": 1.3068, + "step": 2000 + }, + { + "epoch": 0.3563034188034188, + "grad_norm": 0.4862228333950043, + "learning_rate": 0.0001961800853056645, + "loss": 1.0077, + "step": 2001 + }, + { + "epoch": 0.35648148148148145, + "grad_norm": 0.5575395822525024, + "learning_rate": 0.00019617625255400028, + "loss": 1.03, + "step": 2002 + }, + { + "epoch": 0.35665954415954415, + "grad_norm": 0.4826279580593109, + "learning_rate": 0.0001961724179179622, + "loss": 1.268, + "step": 2003 + }, + { + "epoch": 0.35683760683760685, + "grad_norm": 0.49423274397850037, + "learning_rate": 0.00019616858139762534, + "loss": 1.1305, + "step": 2004 + }, + { + "epoch": 0.35701566951566954, + "grad_norm": 0.5208541750907898, + "learning_rate": 0.00019616474299306491, + "loss": 1.1651, + "step": 2005 + }, + { + "epoch": 0.3571937321937322, + "grad_norm": 0.5324164032936096, + "learning_rate": 0.0001961609027043561, + "loss": 1.1406, + "step": 2006 + }, + { + "epoch": 0.3573717948717949, + "grad_norm": 0.45385462045669556, + "learning_rate": 0.00019615706053157416, + "loss": 1.0716, + "step": 2007 + }, + { + "epoch": 0.3575498575498576, + "grad_norm": 0.5016173720359802, + "learning_rate": 0.00019615321647479438, + "loss": 1.0878, + "step": 2008 + }, + { + "epoch": 0.3577279202279202, + "grad_norm": 0.5073097348213196, + "learning_rate": 0.00019614937053409205, + "loss": 1.237, + "step": 2009 + }, + { + "epoch": 0.3579059829059829, + "grad_norm": 0.48880141973495483, + "learning_rate": 0.00019614552270954256, + "loss": 0.8794, + "step": 2010 + }, + { + "epoch": 0.3580840455840456, + "grad_norm": 0.43902209401130676, + "learning_rate": 0.00019614167300122126, + "loss": 0.912, + "step": 2011 + }, + { + "epoch": 0.35826210826210825, + "grad_norm": 0.42809322476387024, + "learning_rate": 0.0001961378214092036, + "loss": 0.7804, + "step": 2012 + }, + { + "epoch": 0.35844017094017094, + "grad_norm": 0.4464281499385834, + "learning_rate": 0.00019613396793356503, + "loss": 1.0004, + "step": 2013 + }, + { + "epoch": 0.35861823361823364, + "grad_norm": 0.49085676670074463, + "learning_rate": 0.00019613011257438109, + "loss": 1.1087, + "step": 2014 + }, + { + "epoch": 0.3587962962962963, + "grad_norm": 0.4997732937335968, + "learning_rate": 0.00019612625533172725, + "loss": 0.9591, + "step": 2015 + }, + { + "epoch": 0.358974358974359, + "grad_norm": 0.48442545533180237, + "learning_rate": 0.00019612239620567912, + "loss": 0.9744, + "step": 2016 + }, + { + "epoch": 0.35915242165242167, + "grad_norm": 0.4989205002784729, + "learning_rate": 0.00019611853519631233, + "loss": 0.9844, + "step": 2017 + }, + { + "epoch": 0.3593304843304843, + "grad_norm": 0.6107521653175354, + "learning_rate": 0.00019611467230370248, + "loss": 1.147, + "step": 2018 + }, + { + "epoch": 0.359508547008547, + "grad_norm": 0.5594844818115234, + "learning_rate": 0.00019611080752792535, + "loss": 1.3195, + "step": 2019 + }, + { + "epoch": 0.3596866096866097, + "grad_norm": 0.4786946475505829, + "learning_rate": 0.00019610694086905656, + "loss": 1.2108, + "step": 2020 + }, + { + "epoch": 0.35986467236467234, + "grad_norm": 0.5186030268669128, + "learning_rate": 0.0001961030723271719, + "loss": 1.0008, + "step": 2021 + }, + { + "epoch": 0.36004273504273504, + "grad_norm": 0.4520573318004608, + "learning_rate": 0.0001960992019023472, + "loss": 1.1307, + "step": 2022 + }, + { + "epoch": 0.36022079772079774, + "grad_norm": 0.4983210563659668, + "learning_rate": 0.00019609532959465823, + "loss": 1.1486, + "step": 2023 + }, + { + "epoch": 0.3603988603988604, + "grad_norm": 0.6209200024604797, + "learning_rate": 0.00019609145540418094, + "loss": 1.2566, + "step": 2024 + }, + { + "epoch": 0.3605769230769231, + "grad_norm": 0.47047603130340576, + "learning_rate": 0.00019608757933099117, + "loss": 1.1588, + "step": 2025 + }, + { + "epoch": 0.36075498575498577, + "grad_norm": 0.5147389769554138, + "learning_rate": 0.0001960837013751649, + "loss": 1.2113, + "step": 2026 + }, + { + "epoch": 0.3609330484330484, + "grad_norm": 0.45826098322868347, + "learning_rate": 0.00019607982153677808, + "loss": 1.13, + "step": 2027 + }, + { + "epoch": 0.3611111111111111, + "grad_norm": 0.5699561834335327, + "learning_rate": 0.00019607593981590675, + "loss": 1.2476, + "step": 2028 + }, + { + "epoch": 0.3612891737891738, + "grad_norm": 0.5349239110946655, + "learning_rate": 0.000196072056212627, + "loss": 1.2295, + "step": 2029 + }, + { + "epoch": 0.36146723646723644, + "grad_norm": 0.6212165355682373, + "learning_rate": 0.00019606817072701484, + "loss": 1.1965, + "step": 2030 + }, + { + "epoch": 0.36164529914529914, + "grad_norm": 0.4870990216732025, + "learning_rate": 0.00019606428335914645, + "loss": 1.4464, + "step": 2031 + }, + { + "epoch": 0.36182336182336183, + "grad_norm": 0.42427828907966614, + "learning_rate": 0.00019606039410909797, + "loss": 1.1546, + "step": 2032 + }, + { + "epoch": 0.36200142450142453, + "grad_norm": 0.5081788301467896, + "learning_rate": 0.0001960565029769456, + "loss": 1.1867, + "step": 2033 + }, + { + "epoch": 0.36217948717948717, + "grad_norm": 0.4813104271888733, + "learning_rate": 0.00019605260996276565, + "loss": 1.3726, + "step": 2034 + }, + { + "epoch": 0.36235754985754987, + "grad_norm": 0.4648851156234741, + "learning_rate": 0.0001960487150666343, + "loss": 1.2434, + "step": 2035 + }, + { + "epoch": 0.36253561253561256, + "grad_norm": 0.484161913394928, + "learning_rate": 0.00019604481828862792, + "loss": 1.1309, + "step": 2036 + }, + { + "epoch": 0.3627136752136752, + "grad_norm": 0.4929439127445221, + "learning_rate": 0.00019604091962882283, + "loss": 1.1007, + "step": 2037 + }, + { + "epoch": 0.3628917378917379, + "grad_norm": 0.45599642395973206, + "learning_rate": 0.00019603701908729544, + "loss": 1.2628, + "step": 2038 + }, + { + "epoch": 0.3630698005698006, + "grad_norm": 0.45295149087905884, + "learning_rate": 0.00019603311666412213, + "loss": 0.9808, + "step": 2039 + }, + { + "epoch": 0.36324786324786323, + "grad_norm": 0.48681163787841797, + "learning_rate": 0.00019602921235937942, + "loss": 1.0574, + "step": 2040 + }, + { + "epoch": 0.36342592592592593, + "grad_norm": 0.41232365369796753, + "learning_rate": 0.00019602530617314378, + "loss": 1.0454, + "step": 2041 + }, + { + "epoch": 0.3636039886039886, + "grad_norm": 0.46214723587036133, + "learning_rate": 0.00019602139810549174, + "loss": 0.9985, + "step": 2042 + }, + { + "epoch": 0.36378205128205127, + "grad_norm": 0.44307878613471985, + "learning_rate": 0.00019601748815649989, + "loss": 0.9683, + "step": 2043 + }, + { + "epoch": 0.36396011396011396, + "grad_norm": 0.4809451401233673, + "learning_rate": 0.00019601357632624477, + "loss": 1.028, + "step": 2044 + }, + { + "epoch": 0.36413817663817666, + "grad_norm": 0.4638497531414032, + "learning_rate": 0.0001960096626148031, + "loss": 0.9851, + "step": 2045 + }, + { + "epoch": 0.3643162393162393, + "grad_norm": 0.5942164063453674, + "learning_rate": 0.00019600574702225153, + "loss": 1.1606, + "step": 2046 + }, + { + "epoch": 0.364494301994302, + "grad_norm": 0.5171293616294861, + "learning_rate": 0.00019600182954866675, + "loss": 1.2335, + "step": 2047 + }, + { + "epoch": 0.3646723646723647, + "grad_norm": 0.5294404625892639, + "learning_rate": 0.00019599791019412558, + "loss": 1.0966, + "step": 2048 + }, + { + "epoch": 0.36485042735042733, + "grad_norm": 0.46117448806762695, + "learning_rate": 0.00019599398895870477, + "loss": 1.0565, + "step": 2049 + }, + { + "epoch": 0.36502849002849, + "grad_norm": 0.5385118126869202, + "learning_rate": 0.00019599006584248118, + "loss": 1.0076, + "step": 2050 + }, + { + "epoch": 0.3652065527065527, + "grad_norm": 0.4915166199207306, + "learning_rate": 0.00019598614084553165, + "loss": 0.9686, + "step": 2051 + }, + { + "epoch": 0.36538461538461536, + "grad_norm": 0.46769094467163086, + "learning_rate": 0.00019598221396793303, + "loss": 1.1217, + "step": 2052 + }, + { + "epoch": 0.36556267806267806, + "grad_norm": 0.5440493822097778, + "learning_rate": 0.00019597828520976236, + "loss": 1.2344, + "step": 2053 + }, + { + "epoch": 0.36574074074074076, + "grad_norm": 0.616727352142334, + "learning_rate": 0.00019597435457109657, + "loss": 1.2953, + "step": 2054 + }, + { + "epoch": 0.3659188034188034, + "grad_norm": 0.4859183430671692, + "learning_rate": 0.00019597042205201265, + "loss": 1.16, + "step": 2055 + }, + { + "epoch": 0.3660968660968661, + "grad_norm": 0.47056329250335693, + "learning_rate": 0.0001959664876525877, + "loss": 0.9982, + "step": 2056 + }, + { + "epoch": 0.3662749287749288, + "grad_norm": 0.48347967863082886, + "learning_rate": 0.00019596255137289875, + "loss": 1.0966, + "step": 2057 + }, + { + "epoch": 0.36645299145299143, + "grad_norm": 0.5068454742431641, + "learning_rate": 0.00019595861321302296, + "loss": 1.2891, + "step": 2058 + }, + { + "epoch": 0.3666310541310541, + "grad_norm": 0.5702359080314636, + "learning_rate": 0.00019595467317303747, + "loss": 1.1394, + "step": 2059 + }, + { + "epoch": 0.3668091168091168, + "grad_norm": 0.5028812885284424, + "learning_rate": 0.0001959507312530195, + "loss": 1.2324, + "step": 2060 + }, + { + "epoch": 0.36698717948717946, + "grad_norm": 0.4672880172729492, + "learning_rate": 0.00019594678745304628, + "loss": 1.0581, + "step": 2061 + }, + { + "epoch": 0.36716524216524216, + "grad_norm": 0.5233900547027588, + "learning_rate": 0.00019594284177319504, + "loss": 1.138, + "step": 2062 + }, + { + "epoch": 0.36734330484330485, + "grad_norm": 0.46871712803840637, + "learning_rate": 0.00019593889421354316, + "loss": 1.2159, + "step": 2063 + }, + { + "epoch": 0.36752136752136755, + "grad_norm": 0.5180533528327942, + "learning_rate": 0.00019593494477416793, + "loss": 1.1116, + "step": 2064 + }, + { + "epoch": 0.3676994301994302, + "grad_norm": 0.5398494005203247, + "learning_rate": 0.0001959309934551467, + "loss": 1.2038, + "step": 2065 + }, + { + "epoch": 0.3678774928774929, + "grad_norm": 0.4850373864173889, + "learning_rate": 0.000195927040256557, + "loss": 1.4315, + "step": 2066 + }, + { + "epoch": 0.3680555555555556, + "grad_norm": 0.49190905690193176, + "learning_rate": 0.0001959230851784762, + "loss": 0.9993, + "step": 2067 + }, + { + "epoch": 0.3682336182336182, + "grad_norm": 0.4546903073787689, + "learning_rate": 0.00019591912822098178, + "loss": 1.0979, + "step": 2068 + }, + { + "epoch": 0.3684116809116809, + "grad_norm": 0.4726468622684479, + "learning_rate": 0.00019591516938415133, + "loss": 1.1629, + "step": 2069 + }, + { + "epoch": 0.3685897435897436, + "grad_norm": 0.47856009006500244, + "learning_rate": 0.00019591120866806235, + "loss": 1.2048, + "step": 2070 + }, + { + "epoch": 0.36876780626780625, + "grad_norm": 0.46847718954086304, + "learning_rate": 0.0001959072460727925, + "loss": 1.0958, + "step": 2071 + }, + { + "epoch": 0.36894586894586895, + "grad_norm": 0.47164350748062134, + "learning_rate": 0.0001959032815984194, + "loss": 1.1912, + "step": 2072 + }, + { + "epoch": 0.36912393162393164, + "grad_norm": 0.4838213324546814, + "learning_rate": 0.0001958993152450207, + "loss": 1.1466, + "step": 2073 + }, + { + "epoch": 0.3693019943019943, + "grad_norm": 0.47234636545181274, + "learning_rate": 0.00019589534701267412, + "loss": 0.9475, + "step": 2074 + }, + { + "epoch": 0.369480056980057, + "grad_norm": 0.4913126826286316, + "learning_rate": 0.00019589137690145746, + "loss": 1.1571, + "step": 2075 + }, + { + "epoch": 0.3696581196581197, + "grad_norm": 0.4696233570575714, + "learning_rate": 0.00019588740491144842, + "loss": 0.9797, + "step": 2076 + }, + { + "epoch": 0.3698361823361823, + "grad_norm": 0.46146106719970703, + "learning_rate": 0.00019588343104272492, + "loss": 1.027, + "step": 2077 + }, + { + "epoch": 0.370014245014245, + "grad_norm": 0.4920627176761627, + "learning_rate": 0.00019587945529536474, + "loss": 1.1008, + "step": 2078 + }, + { + "epoch": 0.3701923076923077, + "grad_norm": 0.4854249954223633, + "learning_rate": 0.0001958754776694458, + "loss": 1.0759, + "step": 2079 + }, + { + "epoch": 0.37037037037037035, + "grad_norm": 0.4884897768497467, + "learning_rate": 0.00019587149816504608, + "loss": 1.1403, + "step": 2080 + }, + { + "epoch": 0.37054843304843305, + "grad_norm": 0.5062584280967712, + "learning_rate": 0.00019586751678224345, + "loss": 1.0185, + "step": 2081 + }, + { + "epoch": 0.37072649572649574, + "grad_norm": 0.44697675108909607, + "learning_rate": 0.000195863533521116, + "loss": 1.0462, + "step": 2082 + }, + { + "epoch": 0.3709045584045584, + "grad_norm": 0.5122885704040527, + "learning_rate": 0.00019585954838174176, + "loss": 1.108, + "step": 2083 + }, + { + "epoch": 0.3710826210826211, + "grad_norm": 0.486650288105011, + "learning_rate": 0.0001958555613641988, + "loss": 1.126, + "step": 2084 + }, + { + "epoch": 0.3712606837606838, + "grad_norm": 0.5296297669410706, + "learning_rate": 0.00019585157246856523, + "loss": 1.1757, + "step": 2085 + }, + { + "epoch": 0.3714387464387464, + "grad_norm": 0.4935721457004547, + "learning_rate": 0.0001958475816949192, + "loss": 1.1654, + "step": 2086 + }, + { + "epoch": 0.3716168091168091, + "grad_norm": 0.6226509213447571, + "learning_rate": 0.00019584358904333891, + "loss": 1.1981, + "step": 2087 + }, + { + "epoch": 0.3717948717948718, + "grad_norm": 0.44094228744506836, + "learning_rate": 0.0001958395945139026, + "loss": 0.8468, + "step": 2088 + }, + { + "epoch": 0.37197293447293445, + "grad_norm": 0.5335884690284729, + "learning_rate": 0.00019583559810668858, + "loss": 1.1597, + "step": 2089 + }, + { + "epoch": 0.37215099715099714, + "grad_norm": 0.4585414528846741, + "learning_rate": 0.000195831599821775, + "loss": 0.9343, + "step": 2090 + }, + { + "epoch": 0.37232905982905984, + "grad_norm": 0.533087432384491, + "learning_rate": 0.00019582759965924035, + "loss": 1.1209, + "step": 2091 + }, + { + "epoch": 0.37250712250712253, + "grad_norm": 0.5302683711051941, + "learning_rate": 0.00019582359761916295, + "loss": 1.236, + "step": 2092 + }, + { + "epoch": 0.3726851851851852, + "grad_norm": 0.4522508382797241, + "learning_rate": 0.00019581959370162122, + "loss": 1.0196, + "step": 2093 + }, + { + "epoch": 0.37286324786324787, + "grad_norm": 0.52391517162323, + "learning_rate": 0.00019581558790669358, + "loss": 1.0077, + "step": 2094 + }, + { + "epoch": 0.37304131054131057, + "grad_norm": 0.47144797444343567, + "learning_rate": 0.00019581158023445854, + "loss": 1.0956, + "step": 2095 + }, + { + "epoch": 0.3732193732193732, + "grad_norm": 0.4486723244190216, + "learning_rate": 0.00019580757068499459, + "loss": 0.8697, + "step": 2096 + }, + { + "epoch": 0.3733974358974359, + "grad_norm": 0.4626580476760864, + "learning_rate": 0.00019580355925838034, + "loss": 0.8489, + "step": 2097 + }, + { + "epoch": 0.3735754985754986, + "grad_norm": 0.5647920370101929, + "learning_rate": 0.00019579954595469438, + "loss": 1.1458, + "step": 2098 + }, + { + "epoch": 0.37375356125356124, + "grad_norm": 0.4734349846839905, + "learning_rate": 0.00019579553077401528, + "loss": 1.1036, + "step": 2099 + }, + { + "epoch": 0.37393162393162394, + "grad_norm": 0.5624295473098755, + "learning_rate": 0.00019579151371642176, + "loss": 0.9793, + "step": 2100 + }, + { + "epoch": 0.37410968660968663, + "grad_norm": 0.47507283091545105, + "learning_rate": 0.00019578749478199256, + "loss": 1.0371, + "step": 2101 + }, + { + "epoch": 0.37428774928774927, + "grad_norm": 0.550865113735199, + "learning_rate": 0.00019578347397080633, + "loss": 1.046, + "step": 2102 + }, + { + "epoch": 0.37446581196581197, + "grad_norm": 0.5249403715133667, + "learning_rate": 0.00019577945128294193, + "loss": 1.3185, + "step": 2103 + }, + { + "epoch": 0.37464387464387466, + "grad_norm": 0.4921024739742279, + "learning_rate": 0.00019577542671847815, + "loss": 1.0758, + "step": 2104 + }, + { + "epoch": 0.3748219373219373, + "grad_norm": 0.5351784825325012, + "learning_rate": 0.00019577140027749384, + "loss": 1.067, + "step": 2105 + }, + { + "epoch": 0.375, + "grad_norm": 0.44420507550239563, + "learning_rate": 0.00019576737196006787, + "loss": 1.1065, + "step": 2106 + }, + { + "epoch": 0.3751780626780627, + "grad_norm": 0.531384289264679, + "learning_rate": 0.0001957633417662792, + "loss": 1.1634, + "step": 2107 + }, + { + "epoch": 0.37535612535612534, + "grad_norm": 0.5167618989944458, + "learning_rate": 0.00019575930969620677, + "loss": 1.1646, + "step": 2108 + }, + { + "epoch": 0.37553418803418803, + "grad_norm": 0.41487228870391846, + "learning_rate": 0.0001957552757499296, + "loss": 0.793, + "step": 2109 + }, + { + "epoch": 0.37571225071225073, + "grad_norm": 0.5110787153244019, + "learning_rate": 0.00019575123992752672, + "loss": 1.1752, + "step": 2110 + }, + { + "epoch": 0.37589031339031337, + "grad_norm": 0.4422051012516022, + "learning_rate": 0.00019574720222907717, + "loss": 1.0102, + "step": 2111 + }, + { + "epoch": 0.37606837606837606, + "grad_norm": 0.4757538139820099, + "learning_rate": 0.0001957431626546601, + "loss": 1.0467, + "step": 2112 + }, + { + "epoch": 0.37624643874643876, + "grad_norm": 0.4736764430999756, + "learning_rate": 0.00019573912120435466, + "loss": 1.3048, + "step": 2113 + }, + { + "epoch": 0.3764245014245014, + "grad_norm": 0.49894335865974426, + "learning_rate": 0.00019573507787824004, + "loss": 1.0502, + "step": 2114 + }, + { + "epoch": 0.3766025641025641, + "grad_norm": 0.48120981454849243, + "learning_rate": 0.00019573103267639543, + "loss": 1.2405, + "step": 2115 + }, + { + "epoch": 0.3767806267806268, + "grad_norm": 0.4826737642288208, + "learning_rate": 0.0001957269855989001, + "loss": 1.1189, + "step": 2116 + }, + { + "epoch": 0.37695868945868943, + "grad_norm": 0.4736921489238739, + "learning_rate": 0.0001957229366458333, + "loss": 1.2862, + "step": 2117 + }, + { + "epoch": 0.37713675213675213, + "grad_norm": 0.3895208537578583, + "learning_rate": 0.00019571888581727446, + "loss": 1.0573, + "step": 2118 + }, + { + "epoch": 0.3773148148148148, + "grad_norm": 0.5107510089874268, + "learning_rate": 0.00019571483311330284, + "loss": 1.2913, + "step": 2119 + }, + { + "epoch": 0.37749287749287747, + "grad_norm": 0.4543241262435913, + "learning_rate": 0.00019571077853399794, + "loss": 0.949, + "step": 2120 + }, + { + "epoch": 0.37767094017094016, + "grad_norm": 0.46897491812705994, + "learning_rate": 0.00019570672207943913, + "loss": 1.2235, + "step": 2121 + }, + { + "epoch": 0.37784900284900286, + "grad_norm": 0.4812130630016327, + "learning_rate": 0.0001957026637497059, + "loss": 0.8857, + "step": 2122 + }, + { + "epoch": 0.37802706552706555, + "grad_norm": 0.47452476620674133, + "learning_rate": 0.00019569860354487782, + "loss": 1.0549, + "step": 2123 + }, + { + "epoch": 0.3782051282051282, + "grad_norm": 0.49879950284957886, + "learning_rate": 0.00019569454146503438, + "loss": 1.0475, + "step": 2124 + }, + { + "epoch": 0.3783831908831909, + "grad_norm": 0.4246445894241333, + "learning_rate": 0.00019569047751025518, + "loss": 0.8788, + "step": 2125 + }, + { + "epoch": 0.3785612535612536, + "grad_norm": 0.4868565499782562, + "learning_rate": 0.00019568641168061986, + "loss": 1.1801, + "step": 2126 + }, + { + "epoch": 0.3787393162393162, + "grad_norm": 0.46723654866218567, + "learning_rate": 0.0001956823439762081, + "loss": 1.1661, + "step": 2127 + }, + { + "epoch": 0.3789173789173789, + "grad_norm": 0.4989059269428253, + "learning_rate": 0.00019567827439709954, + "loss": 1.3037, + "step": 2128 + }, + { + "epoch": 0.3790954415954416, + "grad_norm": 0.441307932138443, + "learning_rate": 0.00019567420294337395, + "loss": 1.0197, + "step": 2129 + }, + { + "epoch": 0.37927350427350426, + "grad_norm": 0.5200160145759583, + "learning_rate": 0.0001956701296151111, + "loss": 1.3366, + "step": 2130 + }, + { + "epoch": 0.37945156695156695, + "grad_norm": 0.43610256910324097, + "learning_rate": 0.00019566605441239082, + "loss": 1.0148, + "step": 2131 + }, + { + "epoch": 0.37962962962962965, + "grad_norm": 0.4160982370376587, + "learning_rate": 0.00019566197733529293, + "loss": 1.0758, + "step": 2132 + }, + { + "epoch": 0.3798076923076923, + "grad_norm": 0.5007950663566589, + "learning_rate": 0.00019565789838389726, + "loss": 1.1937, + "step": 2133 + }, + { + "epoch": 0.379985754985755, + "grad_norm": 0.4991525113582611, + "learning_rate": 0.00019565381755828385, + "loss": 1.1788, + "step": 2134 + }, + { + "epoch": 0.3801638176638177, + "grad_norm": 0.6313113570213318, + "learning_rate": 0.00019564973485853258, + "loss": 1.1241, + "step": 2135 + }, + { + "epoch": 0.3803418803418803, + "grad_norm": 0.49736538529396057, + "learning_rate": 0.0001956456502847234, + "loss": 1.0299, + "step": 2136 + }, + { + "epoch": 0.380519943019943, + "grad_norm": 0.4384380578994751, + "learning_rate": 0.00019564156383693643, + "loss": 1.132, + "step": 2137 + }, + { + "epoch": 0.3806980056980057, + "grad_norm": 0.4696183502674103, + "learning_rate": 0.00019563747551525168, + "loss": 1.1145, + "step": 2138 + }, + { + "epoch": 0.38087606837606836, + "grad_norm": 0.42039749026298523, + "learning_rate": 0.0001956333853197493, + "loss": 0.9549, + "step": 2139 + }, + { + "epoch": 0.38105413105413105, + "grad_norm": 0.5547221899032593, + "learning_rate": 0.00019562929325050936, + "loss": 1.0476, + "step": 2140 + }, + { + "epoch": 0.38123219373219375, + "grad_norm": 0.4803301692008972, + "learning_rate": 0.0001956251993076121, + "loss": 1.1285, + "step": 2141 + }, + { + "epoch": 0.3814102564102564, + "grad_norm": 0.609501838684082, + "learning_rate": 0.00019562110349113766, + "loss": 1.2375, + "step": 2142 + }, + { + "epoch": 0.3815883190883191, + "grad_norm": 0.5134759545326233, + "learning_rate": 0.00019561700580116639, + "loss": 1.0895, + "step": 2143 + }, + { + "epoch": 0.3817663817663818, + "grad_norm": 0.5086711049079895, + "learning_rate": 0.00019561290623777846, + "loss": 1.1139, + "step": 2144 + }, + { + "epoch": 0.3819444444444444, + "grad_norm": 0.5371596813201904, + "learning_rate": 0.00019560880480105428, + "loss": 0.9302, + "step": 2145 + }, + { + "epoch": 0.3821225071225071, + "grad_norm": 0.4966319799423218, + "learning_rate": 0.00019560470149107418, + "loss": 1.2485, + "step": 2146 + }, + { + "epoch": 0.3823005698005698, + "grad_norm": 0.5296950340270996, + "learning_rate": 0.00019560059630791855, + "loss": 1.4449, + "step": 2147 + }, + { + "epoch": 0.38247863247863245, + "grad_norm": 0.5564194321632385, + "learning_rate": 0.00019559648925166783, + "loss": 1.0817, + "step": 2148 + }, + { + "epoch": 0.38265669515669515, + "grad_norm": 0.5763841867446899, + "learning_rate": 0.0001955923803224025, + "loss": 1.1915, + "step": 2149 + }, + { + "epoch": 0.38283475783475784, + "grad_norm": 0.4782295823097229, + "learning_rate": 0.00019558826952020304, + "loss": 1.1317, + "step": 2150 + }, + { + "epoch": 0.38301282051282054, + "grad_norm": 0.4876856207847595, + "learning_rate": 0.00019558415684515002, + "loss": 1.2113, + "step": 2151 + }, + { + "epoch": 0.3831908831908832, + "grad_norm": 0.4894421398639679, + "learning_rate": 0.00019558004229732398, + "loss": 1.0761, + "step": 2152 + }, + { + "epoch": 0.3833689458689459, + "grad_norm": 0.47914227843284607, + "learning_rate": 0.0001955759258768056, + "loss": 1.0869, + "step": 2153 + }, + { + "epoch": 0.38354700854700857, + "grad_norm": 0.43933629989624023, + "learning_rate": 0.00019557180758367543, + "loss": 1.0581, + "step": 2154 + }, + { + "epoch": 0.3837250712250712, + "grad_norm": 0.4078103005886078, + "learning_rate": 0.00019556768741801428, + "loss": 1.065, + "step": 2155 + }, + { + "epoch": 0.3839031339031339, + "grad_norm": 0.5112793445587158, + "learning_rate": 0.00019556356537990278, + "loss": 1.2023, + "step": 2156 + }, + { + "epoch": 0.3840811965811966, + "grad_norm": 0.4699678122997284, + "learning_rate": 0.00019555944146942177, + "loss": 1.2459, + "step": 2157 + }, + { + "epoch": 0.38425925925925924, + "grad_norm": 0.4723528027534485, + "learning_rate": 0.00019555531568665198, + "loss": 1.2204, + "step": 2158 + }, + { + "epoch": 0.38443732193732194, + "grad_norm": 0.4648225009441376, + "learning_rate": 0.00019555118803167432, + "loss": 1.1355, + "step": 2159 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 0.49861815571784973, + "learning_rate": 0.00019554705850456961, + "loss": 1.1301, + "step": 2160 + }, + { + "epoch": 0.3847934472934473, + "grad_norm": 0.4076344966888428, + "learning_rate": 0.00019554292710541874, + "loss": 0.8997, + "step": 2161 + }, + { + "epoch": 0.38497150997151, + "grad_norm": 0.5510796308517456, + "learning_rate": 0.00019553879383430272, + "loss": 1.0594, + "step": 2162 + }, + { + "epoch": 0.38514957264957267, + "grad_norm": 0.55793696641922, + "learning_rate": 0.00019553465869130249, + "loss": 1.1284, + "step": 2163 + }, + { + "epoch": 0.3853276353276353, + "grad_norm": 0.5096491575241089, + "learning_rate": 0.00019553052167649906, + "loss": 1.0419, + "step": 2164 + }, + { + "epoch": 0.385505698005698, + "grad_norm": 0.49077361822128296, + "learning_rate": 0.0001955263827899735, + "loss": 1.1632, + "step": 2165 + }, + { + "epoch": 0.3856837606837607, + "grad_norm": 0.5546894073486328, + "learning_rate": 0.00019552224203180693, + "loss": 1.1487, + "step": 2166 + }, + { + "epoch": 0.38586182336182334, + "grad_norm": 0.4930037260055542, + "learning_rate": 0.00019551809940208047, + "loss": 1.2668, + "step": 2167 + }, + { + "epoch": 0.38603988603988604, + "grad_norm": 0.5600671172142029, + "learning_rate": 0.00019551395490087525, + "loss": 1.3988, + "step": 2168 + }, + { + "epoch": 0.38621794871794873, + "grad_norm": 0.45897629857063293, + "learning_rate": 0.0001955098085282725, + "loss": 0.7792, + "step": 2169 + }, + { + "epoch": 0.3863960113960114, + "grad_norm": 0.46138936281204224, + "learning_rate": 0.00019550566028435346, + "loss": 1.1749, + "step": 2170 + }, + { + "epoch": 0.38657407407407407, + "grad_norm": 0.5136167407035828, + "learning_rate": 0.0001955015101691994, + "loss": 1.0153, + "step": 2171 + }, + { + "epoch": 0.38675213675213677, + "grad_norm": 0.4886440336704254, + "learning_rate": 0.00019549735818289165, + "loss": 1.0006, + "step": 2172 + }, + { + "epoch": 0.3869301994301994, + "grad_norm": 0.4339776635169983, + "learning_rate": 0.00019549320432551154, + "loss": 1.0109, + "step": 2173 + }, + { + "epoch": 0.3871082621082621, + "grad_norm": 0.48729443550109863, + "learning_rate": 0.00019548904859714044, + "loss": 1.2016, + "step": 2174 + }, + { + "epoch": 0.3872863247863248, + "grad_norm": 0.5128757357597351, + "learning_rate": 0.0001954848909978598, + "loss": 1.085, + "step": 2175 + }, + { + "epoch": 0.38746438746438744, + "grad_norm": 0.49636292457580566, + "learning_rate": 0.0001954807315277511, + "loss": 1.0671, + "step": 2176 + }, + { + "epoch": 0.38764245014245013, + "grad_norm": 0.4946988821029663, + "learning_rate": 0.00019547657018689578, + "loss": 1.2091, + "step": 2177 + }, + { + "epoch": 0.38782051282051283, + "grad_norm": 0.49004554748535156, + "learning_rate": 0.00019547240697537544, + "loss": 1.0241, + "step": 2178 + }, + { + "epoch": 0.38799857549857547, + "grad_norm": 0.48750075697898865, + "learning_rate": 0.00019546824189327157, + "loss": 1.1082, + "step": 2179 + }, + { + "epoch": 0.38817663817663817, + "grad_norm": 0.47726166248321533, + "learning_rate": 0.00019546407494066585, + "loss": 1.1275, + "step": 2180 + }, + { + "epoch": 0.38835470085470086, + "grad_norm": 0.5253444910049438, + "learning_rate": 0.00019545990611763986, + "loss": 1.0164, + "step": 2181 + }, + { + "epoch": 0.38853276353276356, + "grad_norm": 0.4470371603965759, + "learning_rate": 0.00019545573542427533, + "loss": 1.0138, + "step": 2182 + }, + { + "epoch": 0.3887108262108262, + "grad_norm": 0.6645087599754333, + "learning_rate": 0.00019545156286065397, + "loss": 1.0884, + "step": 2183 + }, + { + "epoch": 0.3888888888888889, + "grad_norm": 0.498775839805603, + "learning_rate": 0.0001954473884268575, + "loss": 1.1035, + "step": 2184 + }, + { + "epoch": 0.3890669515669516, + "grad_norm": 0.5830566883087158, + "learning_rate": 0.00019544321212296772, + "loss": 1.1665, + "step": 2185 + }, + { + "epoch": 0.38924501424501423, + "grad_norm": 0.48162809014320374, + "learning_rate": 0.00019543903394906646, + "loss": 1.1035, + "step": 2186 + }, + { + "epoch": 0.3894230769230769, + "grad_norm": 0.46334075927734375, + "learning_rate": 0.0001954348539052356, + "loss": 0.9764, + "step": 2187 + }, + { + "epoch": 0.3896011396011396, + "grad_norm": 0.6343515515327454, + "learning_rate": 0.00019543067199155704, + "loss": 0.9474, + "step": 2188 + }, + { + "epoch": 0.38977920227920226, + "grad_norm": 0.4867806136608124, + "learning_rate": 0.0001954264882081127, + "loss": 1.1161, + "step": 2189 + }, + { + "epoch": 0.38995726495726496, + "grad_norm": 0.49305734038352966, + "learning_rate": 0.00019542230255498454, + "loss": 1.1825, + "step": 2190 + }, + { + "epoch": 0.39013532763532766, + "grad_norm": 0.518465518951416, + "learning_rate": 0.00019541811503225457, + "loss": 1.0695, + "step": 2191 + }, + { + "epoch": 0.3903133903133903, + "grad_norm": 0.4892457127571106, + "learning_rate": 0.00019541392564000488, + "loss": 1.3113, + "step": 2192 + }, + { + "epoch": 0.390491452991453, + "grad_norm": 0.5150920152664185, + "learning_rate": 0.00019540973437831753, + "loss": 1.0735, + "step": 2193 + }, + { + "epoch": 0.3906695156695157, + "grad_norm": 0.5414708256721497, + "learning_rate": 0.00019540554124727462, + "loss": 1.0773, + "step": 2194 + }, + { + "epoch": 0.39084757834757833, + "grad_norm": 0.49826398491859436, + "learning_rate": 0.0001954013462469583, + "loss": 1.0542, + "step": 2195 + }, + { + "epoch": 0.391025641025641, + "grad_norm": 0.5203596949577332, + "learning_rate": 0.0001953971493774508, + "loss": 1.178, + "step": 2196 + }, + { + "epoch": 0.3912037037037037, + "grad_norm": 0.45095738768577576, + "learning_rate": 0.00019539295063883432, + "loss": 1.1254, + "step": 2197 + }, + { + "epoch": 0.39138176638176636, + "grad_norm": 0.4938857853412628, + "learning_rate": 0.00019538875003119113, + "loss": 1.1061, + "step": 2198 + }, + { + "epoch": 0.39155982905982906, + "grad_norm": 0.5260919332504272, + "learning_rate": 0.00019538454755460354, + "loss": 1.3292, + "step": 2199 + }, + { + "epoch": 0.39173789173789175, + "grad_norm": 0.46527108550071716, + "learning_rate": 0.00019538034320915388, + "loss": 1.2074, + "step": 2200 + }, + { + "epoch": 0.3919159544159544, + "grad_norm": 0.5608304738998413, + "learning_rate": 0.00019537613699492453, + "loss": 1.0385, + "step": 2201 + }, + { + "epoch": 0.3920940170940171, + "grad_norm": 0.5056684613227844, + "learning_rate": 0.00019537192891199792, + "loss": 1.1513, + "step": 2202 + }, + { + "epoch": 0.3922720797720798, + "grad_norm": 0.3764426112174988, + "learning_rate": 0.00019536771896045644, + "loss": 0.8966, + "step": 2203 + }, + { + "epoch": 0.3924501424501424, + "grad_norm": 0.4983638823032379, + "learning_rate": 0.0001953635071403827, + "loss": 1.097, + "step": 2204 + }, + { + "epoch": 0.3926282051282051, + "grad_norm": 0.5733919739723206, + "learning_rate": 0.00019535929345185904, + "loss": 1.4992, + "step": 2205 + }, + { + "epoch": 0.3928062678062678, + "grad_norm": 0.632064163684845, + "learning_rate": 0.00019535507789496817, + "loss": 1.0611, + "step": 2206 + }, + { + "epoch": 0.39298433048433046, + "grad_norm": 0.409978449344635, + "learning_rate": 0.00019535086046979262, + "loss": 0.7172, + "step": 2207 + }, + { + "epoch": 0.39316239316239315, + "grad_norm": 0.40910813212394714, + "learning_rate": 0.00019534664117641502, + "loss": 0.8803, + "step": 2208 + }, + { + "epoch": 0.39334045584045585, + "grad_norm": 0.4696179926395416, + "learning_rate": 0.00019534242001491807, + "loss": 1.1551, + "step": 2209 + }, + { + "epoch": 0.39351851851851855, + "grad_norm": 0.538425862789154, + "learning_rate": 0.00019533819698538444, + "loss": 1.1296, + "step": 2210 + }, + { + "epoch": 0.3936965811965812, + "grad_norm": 0.5913630723953247, + "learning_rate": 0.00019533397208789692, + "loss": 0.9757, + "step": 2211 + }, + { + "epoch": 0.3938746438746439, + "grad_norm": 0.5649870038032532, + "learning_rate": 0.00019532974532253822, + "loss": 0.9976, + "step": 2212 + }, + { + "epoch": 0.3940527065527066, + "grad_norm": 0.5012063980102539, + "learning_rate": 0.00019532551668939121, + "loss": 0.9969, + "step": 2213 + }, + { + "epoch": 0.3942307692307692, + "grad_norm": 0.5098594427108765, + "learning_rate": 0.00019532128618853872, + "loss": 1.1229, + "step": 2214 + }, + { + "epoch": 0.3944088319088319, + "grad_norm": 0.4753342568874359, + "learning_rate": 0.0001953170538200636, + "loss": 1.0808, + "step": 2215 + }, + { + "epoch": 0.3945868945868946, + "grad_norm": 0.4770098626613617, + "learning_rate": 0.00019531281958404888, + "loss": 1.0656, + "step": 2216 + }, + { + "epoch": 0.39476495726495725, + "grad_norm": 0.6007979512214661, + "learning_rate": 0.00019530858348057746, + "loss": 1.0093, + "step": 2217 + }, + { + "epoch": 0.39494301994301995, + "grad_norm": 0.4501650929450989, + "learning_rate": 0.00019530434550973227, + "loss": 0.8557, + "step": 2218 + }, + { + "epoch": 0.39512108262108264, + "grad_norm": 0.5123980641365051, + "learning_rate": 0.00019530010567159645, + "loss": 0.9833, + "step": 2219 + }, + { + "epoch": 0.3952991452991453, + "grad_norm": 0.4623969495296478, + "learning_rate": 0.000195295863966253, + "loss": 0.913, + "step": 2220 + }, + { + "epoch": 0.395477207977208, + "grad_norm": 0.4341880679130554, + "learning_rate": 0.0001952916203937851, + "loss": 1.0234, + "step": 2221 + }, + { + "epoch": 0.3956552706552707, + "grad_norm": 0.5935006141662598, + "learning_rate": 0.00019528737495427581, + "loss": 1.061, + "step": 2222 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 0.44835174083709717, + "learning_rate": 0.00019528312764780837, + "loss": 1.1567, + "step": 2223 + }, + { + "epoch": 0.396011396011396, + "grad_norm": 0.5476976633071899, + "learning_rate": 0.00019527887847446595, + "loss": 1.2304, + "step": 2224 + }, + { + "epoch": 0.3961894586894587, + "grad_norm": 0.4487939774990082, + "learning_rate": 0.00019527462743433187, + "loss": 1.1813, + "step": 2225 + }, + { + "epoch": 0.39636752136752135, + "grad_norm": 0.4053241014480591, + "learning_rate": 0.00019527037452748936, + "loss": 0.7899, + "step": 2226 + }, + { + "epoch": 0.39654558404558404, + "grad_norm": 0.534570574760437, + "learning_rate": 0.00019526611975402176, + "loss": 1.0681, + "step": 2227 + }, + { + "epoch": 0.39672364672364674, + "grad_norm": 0.46096158027648926, + "learning_rate": 0.00019526186311401246, + "loss": 0.9234, + "step": 2228 + }, + { + "epoch": 0.3969017094017094, + "grad_norm": 0.47363516688346863, + "learning_rate": 0.00019525760460754483, + "loss": 1.0197, + "step": 2229 + }, + { + "epoch": 0.3970797720797721, + "grad_norm": 0.46317258477211, + "learning_rate": 0.00019525334423470234, + "loss": 1.2103, + "step": 2230 + }, + { + "epoch": 0.39725783475783477, + "grad_norm": 0.4924237132072449, + "learning_rate": 0.0001952490819955684, + "loss": 1.3299, + "step": 2231 + }, + { + "epoch": 0.3974358974358974, + "grad_norm": 0.5419978499412537, + "learning_rate": 0.0001952448178902266, + "loss": 1.2526, + "step": 2232 + }, + { + "epoch": 0.3976139601139601, + "grad_norm": 0.5003267526626587, + "learning_rate": 0.00019524055191876043, + "loss": 1.1073, + "step": 2233 + }, + { + "epoch": 0.3977920227920228, + "grad_norm": 0.621789276599884, + "learning_rate": 0.00019523628408125347, + "loss": 1.3409, + "step": 2234 + }, + { + "epoch": 0.39797008547008544, + "grad_norm": 0.44235602021217346, + "learning_rate": 0.0001952320143777894, + "loss": 0.9799, + "step": 2235 + }, + { + "epoch": 0.39814814814814814, + "grad_norm": 0.49954718351364136, + "learning_rate": 0.0001952277428084518, + "loss": 1.2227, + "step": 2236 + }, + { + "epoch": 0.39832621082621084, + "grad_norm": 0.5113739967346191, + "learning_rate": 0.00019522346937332443, + "loss": 1.1644, + "step": 2237 + }, + { + "epoch": 0.39850427350427353, + "grad_norm": 0.5026139616966248, + "learning_rate": 0.00019521919407249096, + "loss": 1.0823, + "step": 2238 + }, + { + "epoch": 0.39868233618233617, + "grad_norm": 0.4943205714225769, + "learning_rate": 0.0001952149169060352, + "loss": 1.0961, + "step": 2239 + }, + { + "epoch": 0.39886039886039887, + "grad_norm": 0.4680631458759308, + "learning_rate": 0.00019521063787404094, + "loss": 0.9787, + "step": 2240 + }, + { + "epoch": 0.39903846153846156, + "grad_norm": 0.5511566400527954, + "learning_rate": 0.00019520635697659202, + "loss": 1.2543, + "step": 2241 + }, + { + "epoch": 0.3992165242165242, + "grad_norm": 0.5494263172149658, + "learning_rate": 0.00019520207421377229, + "loss": 1.1978, + "step": 2242 + }, + { + "epoch": 0.3993945868945869, + "grad_norm": 0.4850340485572815, + "learning_rate": 0.00019519778958566568, + "loss": 0.8531, + "step": 2243 + }, + { + "epoch": 0.3995726495726496, + "grad_norm": 0.47168150544166565, + "learning_rate": 0.00019519350309235613, + "loss": 1.0746, + "step": 2244 + }, + { + "epoch": 0.39975071225071224, + "grad_norm": 0.571133553981781, + "learning_rate": 0.00019518921473392765, + "loss": 1.2984, + "step": 2245 + }, + { + "epoch": 0.39992877492877493, + "grad_norm": 0.4636089503765106, + "learning_rate": 0.00019518492451046427, + "loss": 1.019, + "step": 2246 + }, + { + "epoch": 0.40010683760683763, + "grad_norm": 0.4573518931865692, + "learning_rate": 0.00019518063242205, + "loss": 1.1042, + "step": 2247 + }, + { + "epoch": 0.40028490028490027, + "grad_norm": 0.49098989367485046, + "learning_rate": 0.00019517633846876894, + "loss": 1.1224, + "step": 2248 + }, + { + "epoch": 0.40046296296296297, + "grad_norm": 0.5475491881370544, + "learning_rate": 0.00019517204265070523, + "loss": 1.0984, + "step": 2249 + }, + { + "epoch": 0.40064102564102566, + "grad_norm": 0.45498281717300415, + "learning_rate": 0.00019516774496794307, + "loss": 0.8883, + "step": 2250 + }, + { + "epoch": 0.4008190883190883, + "grad_norm": 0.4908423125743866, + "learning_rate": 0.00019516344542056666, + "loss": 1.328, + "step": 2251 + }, + { + "epoch": 0.400997150997151, + "grad_norm": 0.5474920272827148, + "learning_rate": 0.0001951591440086602, + "loss": 1.3825, + "step": 2252 + }, + { + "epoch": 0.4011752136752137, + "grad_norm": 0.5165615081787109, + "learning_rate": 0.000195154840732308, + "loss": 1.33, + "step": 2253 + }, + { + "epoch": 0.40135327635327633, + "grad_norm": 0.5185585021972656, + "learning_rate": 0.00019515053559159435, + "loss": 1.1689, + "step": 2254 + }, + { + "epoch": 0.40153133903133903, + "grad_norm": 0.5468854904174805, + "learning_rate": 0.00019514622858660363, + "loss": 1.2708, + "step": 2255 + }, + { + "epoch": 0.4017094017094017, + "grad_norm": 0.47556906938552856, + "learning_rate": 0.0001951419197174202, + "loss": 1.0488, + "step": 2256 + }, + { + "epoch": 0.40188746438746437, + "grad_norm": 0.5521323084831238, + "learning_rate": 0.0001951376089841285, + "loss": 1.0868, + "step": 2257 + }, + { + "epoch": 0.40206552706552706, + "grad_norm": 0.6029638051986694, + "learning_rate": 0.00019513329638681296, + "loss": 1.1735, + "step": 2258 + }, + { + "epoch": 0.40224358974358976, + "grad_norm": 0.4897766411304474, + "learning_rate": 0.00019512898192555812, + "loss": 1.1687, + "step": 2259 + }, + { + "epoch": 0.4024216524216524, + "grad_norm": 0.45527184009552, + "learning_rate": 0.00019512466560044848, + "loss": 1.0352, + "step": 2260 + }, + { + "epoch": 0.4025997150997151, + "grad_norm": 0.5025625824928284, + "learning_rate": 0.00019512034741156863, + "loss": 1.2503, + "step": 2261 + }, + { + "epoch": 0.4027777777777778, + "grad_norm": 0.46415451169013977, + "learning_rate": 0.00019511602735900317, + "loss": 1.032, + "step": 2262 + }, + { + "epoch": 0.40295584045584043, + "grad_norm": 0.4812934398651123, + "learning_rate": 0.00019511170544283678, + "loss": 1.0523, + "step": 2263 + }, + { + "epoch": 0.4031339031339031, + "grad_norm": 0.49937039613723755, + "learning_rate": 0.00019510738166315404, + "loss": 1.2238, + "step": 2264 + }, + { + "epoch": 0.4033119658119658, + "grad_norm": 0.5428698062896729, + "learning_rate": 0.00019510305602003975, + "loss": 1.0361, + "step": 2265 + }, + { + "epoch": 0.40349002849002846, + "grad_norm": 0.44836854934692383, + "learning_rate": 0.0001950987285135786, + "loss": 1.169, + "step": 2266 + }, + { + "epoch": 0.40366809116809116, + "grad_norm": 0.5071489214897156, + "learning_rate": 0.00019509439914385549, + "loss": 1.1567, + "step": 2267 + }, + { + "epoch": 0.40384615384615385, + "grad_norm": 0.5204613208770752, + "learning_rate": 0.00019509006791095513, + "loss": 0.9949, + "step": 2268 + }, + { + "epoch": 0.40402421652421655, + "grad_norm": 0.4583234488964081, + "learning_rate": 0.00019508573481496238, + "loss": 0.9051, + "step": 2269 + }, + { + "epoch": 0.4042022792022792, + "grad_norm": 0.5436791181564331, + "learning_rate": 0.00019508139985596222, + "loss": 1.3239, + "step": 2270 + }, + { + "epoch": 0.4043803418803419, + "grad_norm": 0.48774269223213196, + "learning_rate": 0.00019507706303403954, + "loss": 1.2102, + "step": 2271 + }, + { + "epoch": 0.4045584045584046, + "grad_norm": 0.4742540717124939, + "learning_rate": 0.00019507272434927933, + "loss": 1.1137, + "step": 2272 + }, + { + "epoch": 0.4047364672364672, + "grad_norm": 0.531148374080658, + "learning_rate": 0.00019506838380176658, + "loss": 1.3162, + "step": 2273 + }, + { + "epoch": 0.4049145299145299, + "grad_norm": 0.5002314448356628, + "learning_rate": 0.0001950640413915863, + "loss": 1.0743, + "step": 2274 + }, + { + "epoch": 0.4050925925925926, + "grad_norm": 0.39826446771621704, + "learning_rate": 0.00019505969711882366, + "loss": 0.7698, + "step": 2275 + }, + { + "epoch": 0.40527065527065526, + "grad_norm": 0.5177471041679382, + "learning_rate": 0.00019505535098356371, + "loss": 1.1821, + "step": 2276 + }, + { + "epoch": 0.40544871794871795, + "grad_norm": 0.467241108417511, + "learning_rate": 0.00019505100298589158, + "loss": 0.8036, + "step": 2277 + }, + { + "epoch": 0.40562678062678065, + "grad_norm": 0.43711844086647034, + "learning_rate": 0.00019504665312589255, + "loss": 0.8667, + "step": 2278 + }, + { + "epoch": 0.4058048433048433, + "grad_norm": 0.4929116368293762, + "learning_rate": 0.00019504230140365177, + "loss": 1.1279, + "step": 2279 + }, + { + "epoch": 0.405982905982906, + "grad_norm": 0.5279183983802795, + "learning_rate": 0.00019503794781925452, + "loss": 1.1318, + "step": 2280 + }, + { + "epoch": 0.4061609686609687, + "grad_norm": 0.549217939376831, + "learning_rate": 0.00019503359237278608, + "loss": 1.2007, + "step": 2281 + }, + { + "epoch": 0.4063390313390313, + "grad_norm": 0.5485880374908447, + "learning_rate": 0.00019502923506433187, + "loss": 1.1079, + "step": 2282 + }, + { + "epoch": 0.406517094017094, + "grad_norm": 0.48379644751548767, + "learning_rate": 0.0001950248758939772, + "loss": 0.9978, + "step": 2283 + }, + { + "epoch": 0.4066951566951567, + "grad_norm": 0.5943657755851746, + "learning_rate": 0.00019502051486180744, + "loss": 1.0466, + "step": 2284 + }, + { + "epoch": 0.40687321937321935, + "grad_norm": 0.5721273422241211, + "learning_rate": 0.00019501615196790812, + "loss": 1.2674, + "step": 2285 + }, + { + "epoch": 0.40705128205128205, + "grad_norm": 0.47624221444129944, + "learning_rate": 0.00019501178721236464, + "loss": 1.089, + "step": 2286 + }, + { + "epoch": 0.40722934472934474, + "grad_norm": 0.5091297030448914, + "learning_rate": 0.0001950074205952626, + "loss": 1.2035, + "step": 2287 + }, + { + "epoch": 0.4074074074074074, + "grad_norm": 0.45206236839294434, + "learning_rate": 0.0001950030521166875, + "loss": 0.9188, + "step": 2288 + }, + { + "epoch": 0.4075854700854701, + "grad_norm": 0.5563844442367554, + "learning_rate": 0.00019499868177672497, + "loss": 1.3444, + "step": 2289 + }, + { + "epoch": 0.4077635327635328, + "grad_norm": 0.4971138536930084, + "learning_rate": 0.00019499430957546055, + "loss": 1.1615, + "step": 2290 + }, + { + "epoch": 0.4079415954415954, + "grad_norm": 0.49355944991111755, + "learning_rate": 0.00019498993551298, + "loss": 1.1528, + "step": 2291 + }, + { + "epoch": 0.4081196581196581, + "grad_norm": 0.534705638885498, + "learning_rate": 0.000194985559589369, + "loss": 1.197, + "step": 2292 + }, + { + "epoch": 0.4082977207977208, + "grad_norm": 0.5113020539283752, + "learning_rate": 0.0001949811818047133, + "loss": 1.109, + "step": 2293 + }, + { + "epoch": 0.40847578347578345, + "grad_norm": 0.4823366701602936, + "learning_rate": 0.00019497680215909858, + "loss": 1.168, + "step": 2294 + }, + { + "epoch": 0.40865384615384615, + "grad_norm": 0.500792920589447, + "learning_rate": 0.00019497242065261077, + "loss": 1.1567, + "step": 2295 + }, + { + "epoch": 0.40883190883190884, + "grad_norm": 0.5047918558120728, + "learning_rate": 0.00019496803728533566, + "loss": 1.0515, + "step": 2296 + }, + { + "epoch": 0.40900997150997154, + "grad_norm": 0.474624365568161, + "learning_rate": 0.00019496365205735913, + "loss": 1.1747, + "step": 2297 + }, + { + "epoch": 0.4091880341880342, + "grad_norm": 0.5522183179855347, + "learning_rate": 0.0001949592649687671, + "loss": 1.1506, + "step": 2298 + }, + { + "epoch": 0.4093660968660969, + "grad_norm": 0.4526083767414093, + "learning_rate": 0.00019495487601964553, + "loss": 0.9968, + "step": 2299 + }, + { + "epoch": 0.40954415954415957, + "grad_norm": 0.545845091342926, + "learning_rate": 0.00019495048521008044, + "loss": 1.146, + "step": 2300 + }, + { + "epoch": 0.4097222222222222, + "grad_norm": 0.5475544333457947, + "learning_rate": 0.00019494609254015784, + "loss": 1.0101, + "step": 2301 + }, + { + "epoch": 0.4099002849002849, + "grad_norm": 0.43419042229652405, + "learning_rate": 0.00019494169800996373, + "loss": 1.065, + "step": 2302 + }, + { + "epoch": 0.4100783475783476, + "grad_norm": 0.44998374581336975, + "learning_rate": 0.00019493730161958435, + "loss": 0.9948, + "step": 2303 + }, + { + "epoch": 0.41025641025641024, + "grad_norm": 0.5401661992073059, + "learning_rate": 0.0001949329033691057, + "loss": 1.0473, + "step": 2304 + }, + { + "epoch": 0.41043447293447294, + "grad_norm": 0.48064103722572327, + "learning_rate": 0.00019492850325861404, + "loss": 1.0486, + "step": 2305 + }, + { + "epoch": 0.41061253561253563, + "grad_norm": 0.5398300290107727, + "learning_rate": 0.00019492410128819557, + "loss": 1.0314, + "step": 2306 + }, + { + "epoch": 0.4107905982905983, + "grad_norm": 0.4771125912666321, + "learning_rate": 0.0001949196974579365, + "loss": 0.9855, + "step": 2307 + }, + { + "epoch": 0.41096866096866097, + "grad_norm": 0.5375809669494629, + "learning_rate": 0.00019491529176792315, + "loss": 1.0777, + "step": 2308 + }, + { + "epoch": 0.41114672364672367, + "grad_norm": 0.48424094915390015, + "learning_rate": 0.00019491088421824183, + "loss": 1.0751, + "step": 2309 + }, + { + "epoch": 0.4113247863247863, + "grad_norm": 0.5054880380630493, + "learning_rate": 0.00019490647480897887, + "loss": 1.2457, + "step": 2310 + }, + { + "epoch": 0.411502849002849, + "grad_norm": 0.47118356823921204, + "learning_rate": 0.0001949020635402207, + "loss": 1.0445, + "step": 2311 + }, + { + "epoch": 0.4116809116809117, + "grad_norm": 0.47171851992607117, + "learning_rate": 0.00019489765041205375, + "loss": 1.0062, + "step": 2312 + }, + { + "epoch": 0.41185897435897434, + "grad_norm": 0.5703238844871521, + "learning_rate": 0.00019489323542456447, + "loss": 1.5639, + "step": 2313 + }, + { + "epoch": 0.41203703703703703, + "grad_norm": 0.5045075416564941, + "learning_rate": 0.00019488881857783935, + "loss": 1.1665, + "step": 2314 + }, + { + "epoch": 0.41221509971509973, + "grad_norm": 0.46835362911224365, + "learning_rate": 0.00019488439987196495, + "loss": 1.2078, + "step": 2315 + }, + { + "epoch": 0.41239316239316237, + "grad_norm": 0.5187196731567383, + "learning_rate": 0.00019487997930702785, + "loss": 1.1049, + "step": 2316 + }, + { + "epoch": 0.41257122507122507, + "grad_norm": 0.5190554857254028, + "learning_rate": 0.00019487555688311463, + "loss": 1.331, + "step": 2317 + }, + { + "epoch": 0.41274928774928776, + "grad_norm": 0.7394969463348389, + "learning_rate": 0.00019487113260031197, + "loss": 0.9646, + "step": 2318 + }, + { + "epoch": 0.4129273504273504, + "grad_norm": 0.532982349395752, + "learning_rate": 0.00019486670645870656, + "loss": 1.166, + "step": 2319 + }, + { + "epoch": 0.4131054131054131, + "grad_norm": 0.48659515380859375, + "learning_rate": 0.00019486227845838509, + "loss": 1.0016, + "step": 2320 + }, + { + "epoch": 0.4132834757834758, + "grad_norm": 0.5364453196525574, + "learning_rate": 0.00019485784859943434, + "loss": 1.3877, + "step": 2321 + }, + { + "epoch": 0.41346153846153844, + "grad_norm": 0.49788740277290344, + "learning_rate": 0.0001948534168819411, + "loss": 1.2949, + "step": 2322 + }, + { + "epoch": 0.41363960113960113, + "grad_norm": 0.5125377774238586, + "learning_rate": 0.00019484898330599217, + "loss": 0.9769, + "step": 2323 + }, + { + "epoch": 0.41381766381766383, + "grad_norm": 0.5434861779212952, + "learning_rate": 0.00019484454787167447, + "loss": 1.254, + "step": 2324 + }, + { + "epoch": 0.41399572649572647, + "grad_norm": 0.5324583053588867, + "learning_rate": 0.00019484011057907487, + "loss": 0.9788, + "step": 2325 + }, + { + "epoch": 0.41417378917378916, + "grad_norm": 0.4806961715221405, + "learning_rate": 0.00019483567142828033, + "loss": 1.0089, + "step": 2326 + }, + { + "epoch": 0.41435185185185186, + "grad_norm": 0.5152947306632996, + "learning_rate": 0.0001948312304193778, + "loss": 1.15, + "step": 2327 + }, + { + "epoch": 0.41452991452991456, + "grad_norm": 0.6030138731002808, + "learning_rate": 0.0001948267875524543, + "loss": 1.196, + "step": 2328 + }, + { + "epoch": 0.4147079772079772, + "grad_norm": 0.4504946768283844, + "learning_rate": 0.0001948223428275969, + "loss": 0.8742, + "step": 2329 + }, + { + "epoch": 0.4148860398860399, + "grad_norm": 0.5195745825767517, + "learning_rate": 0.00019481789624489263, + "loss": 1.0104, + "step": 2330 + }, + { + "epoch": 0.4150641025641026, + "grad_norm": 0.5269250869750977, + "learning_rate": 0.0001948134478044287, + "loss": 1.2284, + "step": 2331 + }, + { + "epoch": 0.41524216524216523, + "grad_norm": 0.5302315354347229, + "learning_rate": 0.00019480899750629218, + "loss": 1.1374, + "step": 2332 + }, + { + "epoch": 0.4154202279202279, + "grad_norm": 0.5501471161842346, + "learning_rate": 0.0001948045453505703, + "loss": 1.214, + "step": 2333 + }, + { + "epoch": 0.4155982905982906, + "grad_norm": 0.4674588739871979, + "learning_rate": 0.0001948000913373503, + "loss": 1.0568, + "step": 2334 + }, + { + "epoch": 0.41577635327635326, + "grad_norm": 0.5262266993522644, + "learning_rate": 0.0001947956354667195, + "loss": 1.111, + "step": 2335 + }, + { + "epoch": 0.41595441595441596, + "grad_norm": 0.4549071788787842, + "learning_rate": 0.00019479117773876507, + "loss": 1.2655, + "step": 2336 + }, + { + "epoch": 0.41613247863247865, + "grad_norm": 0.48897311091423035, + "learning_rate": 0.00019478671815357447, + "loss": 1.0543, + "step": 2337 + }, + { + "epoch": 0.4163105413105413, + "grad_norm": 0.5544867515563965, + "learning_rate": 0.000194782256711235, + "loss": 1.2276, + "step": 2338 + }, + { + "epoch": 0.416488603988604, + "grad_norm": 0.5050773024559021, + "learning_rate": 0.0001947777934118341, + "loss": 0.9781, + "step": 2339 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 0.4831899106502533, + "learning_rate": 0.00019477332825545925, + "loss": 1.0213, + "step": 2340 + }, + { + "epoch": 0.4168447293447293, + "grad_norm": 0.5392552614212036, + "learning_rate": 0.0001947688612421979, + "loss": 1.3251, + "step": 2341 + }, + { + "epoch": 0.417022792022792, + "grad_norm": 0.5003608465194702, + "learning_rate": 0.00019476439237213754, + "loss": 1.0714, + "step": 2342 + }, + { + "epoch": 0.4172008547008547, + "grad_norm": 0.5016986727714539, + "learning_rate": 0.00019475992164536582, + "loss": 1.0656, + "step": 2343 + }, + { + "epoch": 0.41737891737891736, + "grad_norm": 0.5139234066009521, + "learning_rate": 0.00019475544906197024, + "loss": 1.1317, + "step": 2344 + }, + { + "epoch": 0.41755698005698005, + "grad_norm": 0.582478940486908, + "learning_rate": 0.00019475097462203847, + "loss": 1.4209, + "step": 2345 + }, + { + "epoch": 0.41773504273504275, + "grad_norm": 0.5248767137527466, + "learning_rate": 0.00019474649832565823, + "loss": 1.2965, + "step": 2346 + }, + { + "epoch": 0.4179131054131054, + "grad_norm": 0.4977390170097351, + "learning_rate": 0.00019474202017291713, + "loss": 1.3319, + "step": 2347 + }, + { + "epoch": 0.4180911680911681, + "grad_norm": 0.4868984818458557, + "learning_rate": 0.00019473754016390298, + "loss": 1.0595, + "step": 2348 + }, + { + "epoch": 0.4182692307692308, + "grad_norm": 0.5965346693992615, + "learning_rate": 0.00019473305829870353, + "loss": 1.2289, + "step": 2349 + }, + { + "epoch": 0.4184472934472934, + "grad_norm": 0.46590209007263184, + "learning_rate": 0.0001947285745774066, + "loss": 1.0468, + "step": 2350 + }, + { + "epoch": 0.4186253561253561, + "grad_norm": 0.497811883687973, + "learning_rate": 0.0001947240890001, + "loss": 1.1247, + "step": 2351 + }, + { + "epoch": 0.4188034188034188, + "grad_norm": 0.5348289012908936, + "learning_rate": 0.0001947196015668717, + "loss": 0.9496, + "step": 2352 + }, + { + "epoch": 0.41898148148148145, + "grad_norm": 0.5086174607276917, + "learning_rate": 0.0001947151122778095, + "loss": 0.8869, + "step": 2353 + }, + { + "epoch": 0.41915954415954415, + "grad_norm": 0.4844677150249481, + "learning_rate": 0.00019471062113300146, + "loss": 0.847, + "step": 2354 + }, + { + "epoch": 0.41933760683760685, + "grad_norm": 0.5395866632461548, + "learning_rate": 0.00019470612813253556, + "loss": 0.9684, + "step": 2355 + }, + { + "epoch": 0.41951566951566954, + "grad_norm": 0.479403018951416, + "learning_rate": 0.0001947016332764998, + "loss": 1.0532, + "step": 2356 + }, + { + "epoch": 0.4196937321937322, + "grad_norm": 0.5499961376190186, + "learning_rate": 0.00019469713656498227, + "loss": 1.2565, + "step": 2357 + }, + { + "epoch": 0.4198717948717949, + "grad_norm": 0.5865352153778076, + "learning_rate": 0.00019469263799807104, + "loss": 1.1349, + "step": 2358 + }, + { + "epoch": 0.4200498575498576, + "grad_norm": 0.4454309046268463, + "learning_rate": 0.00019468813757585432, + "loss": 0.9631, + "step": 2359 + }, + { + "epoch": 0.4202279202279202, + "grad_norm": 0.48426875472068787, + "learning_rate": 0.00019468363529842023, + "loss": 0.9795, + "step": 2360 + }, + { + "epoch": 0.4204059829059829, + "grad_norm": 0.47428226470947266, + "learning_rate": 0.00019467913116585697, + "loss": 0.9316, + "step": 2361 + }, + { + "epoch": 0.4205840455840456, + "grad_norm": 0.5193758010864258, + "learning_rate": 0.00019467462517825282, + "loss": 1.235, + "step": 2362 + }, + { + "epoch": 0.42076210826210825, + "grad_norm": 0.49845513701438904, + "learning_rate": 0.00019467011733569607, + "loss": 1.2413, + "step": 2363 + }, + { + "epoch": 0.42094017094017094, + "grad_norm": 0.45483845472335815, + "learning_rate": 0.00019466560763827502, + "loss": 1.2817, + "step": 2364 + }, + { + "epoch": 0.42111823361823364, + "grad_norm": 0.43345287442207336, + "learning_rate": 0.00019466109608607806, + "loss": 0.8568, + "step": 2365 + }, + { + "epoch": 0.4212962962962963, + "grad_norm": 0.4467088282108307, + "learning_rate": 0.00019465658267919352, + "loss": 1.1408, + "step": 2366 + }, + { + "epoch": 0.421474358974359, + "grad_norm": 0.6705610156059265, + "learning_rate": 0.00019465206741770992, + "loss": 1.445, + "step": 2367 + }, + { + "epoch": 0.42165242165242167, + "grad_norm": 0.5037859678268433, + "learning_rate": 0.00019464755030171565, + "loss": 0.8682, + "step": 2368 + }, + { + "epoch": 0.4218304843304843, + "grad_norm": 0.49576324224472046, + "learning_rate": 0.00019464303133129928, + "loss": 0.8387, + "step": 2369 + }, + { + "epoch": 0.422008547008547, + "grad_norm": 0.5222806334495544, + "learning_rate": 0.00019463851050654927, + "loss": 1.1443, + "step": 2370 + }, + { + "epoch": 0.4221866096866097, + "grad_norm": 0.4966863989830017, + "learning_rate": 0.00019463398782755426, + "loss": 1.1555, + "step": 2371 + }, + { + "epoch": 0.42236467236467234, + "grad_norm": 0.6140168309211731, + "learning_rate": 0.00019462946329440285, + "loss": 1.2264, + "step": 2372 + }, + { + "epoch": 0.42254273504273504, + "grad_norm": 0.4906651973724365, + "learning_rate": 0.0001946249369071837, + "loss": 1.2459, + "step": 2373 + }, + { + "epoch": 0.42272079772079774, + "grad_norm": 0.5956700444221497, + "learning_rate": 0.00019462040866598544, + "loss": 1.1521, + "step": 2374 + }, + { + "epoch": 0.4228988603988604, + "grad_norm": 0.46044886112213135, + "learning_rate": 0.00019461587857089687, + "loss": 1.2084, + "step": 2375 + }, + { + "epoch": 0.4230769230769231, + "grad_norm": 0.5109430551528931, + "learning_rate": 0.00019461134662200668, + "loss": 1.2684, + "step": 2376 + }, + { + "epoch": 0.42325498575498577, + "grad_norm": 0.4373733103275299, + "learning_rate": 0.0001946068128194037, + "loss": 1.0451, + "step": 2377 + }, + { + "epoch": 0.4234330484330484, + "grad_norm": 0.553817868232727, + "learning_rate": 0.00019460227716317673, + "loss": 1.1052, + "step": 2378 + }, + { + "epoch": 0.4236111111111111, + "grad_norm": 0.5742647647857666, + "learning_rate": 0.00019459773965341468, + "loss": 1.1647, + "step": 2379 + }, + { + "epoch": 0.4237891737891738, + "grad_norm": 0.5461940169334412, + "learning_rate": 0.00019459320029020642, + "loss": 1.0953, + "step": 2380 + }, + { + "epoch": 0.42396723646723644, + "grad_norm": 0.5837802290916443, + "learning_rate": 0.0001945886590736409, + "loss": 1.1303, + "step": 2381 + }, + { + "epoch": 0.42414529914529914, + "grad_norm": 0.5316985249519348, + "learning_rate": 0.0001945841160038071, + "loss": 1.1204, + "step": 2382 + }, + { + "epoch": 0.42432336182336183, + "grad_norm": 0.5846191048622131, + "learning_rate": 0.00019457957108079404, + "loss": 1.2622, + "step": 2383 + }, + { + "epoch": 0.42450142450142453, + "grad_norm": 0.43266957998275757, + "learning_rate": 0.00019457502430469075, + "loss": 0.9834, + "step": 2384 + }, + { + "epoch": 0.42467948717948717, + "grad_norm": 0.514081597328186, + "learning_rate": 0.00019457047567558632, + "loss": 0.8413, + "step": 2385 + }, + { + "epoch": 0.42485754985754987, + "grad_norm": 0.4831700325012207, + "learning_rate": 0.00019456592519356987, + "loss": 0.9244, + "step": 2386 + }, + { + "epoch": 0.42503561253561256, + "grad_norm": 0.5612850785255432, + "learning_rate": 0.00019456137285873057, + "loss": 0.9438, + "step": 2387 + }, + { + "epoch": 0.4252136752136752, + "grad_norm": 0.5197352766990662, + "learning_rate": 0.00019455681867115758, + "loss": 1.1095, + "step": 2388 + }, + { + "epoch": 0.4253917378917379, + "grad_norm": 0.5045261979103088, + "learning_rate": 0.00019455226263094018, + "loss": 1.0007, + "step": 2389 + }, + { + "epoch": 0.4255698005698006, + "grad_norm": 0.5167570114135742, + "learning_rate": 0.00019454770473816758, + "loss": 1.1335, + "step": 2390 + }, + { + "epoch": 0.42574786324786323, + "grad_norm": 0.49262070655822754, + "learning_rate": 0.00019454314499292913, + "loss": 1.0436, + "step": 2391 + }, + { + "epoch": 0.42592592592592593, + "grad_norm": 0.4489207863807678, + "learning_rate": 0.00019453858339531417, + "loss": 1.0138, + "step": 2392 + }, + { + "epoch": 0.4261039886039886, + "grad_norm": 0.6024920344352722, + "learning_rate": 0.00019453401994541203, + "loss": 1.1921, + "step": 2393 + }, + { + "epoch": 0.42628205128205127, + "grad_norm": 0.46807861328125, + "learning_rate": 0.00019452945464331215, + "loss": 1.0947, + "step": 2394 + }, + { + "epoch": 0.42646011396011396, + "grad_norm": 0.48776543140411377, + "learning_rate": 0.00019452488748910397, + "loss": 1.0029, + "step": 2395 + }, + { + "epoch": 0.42663817663817666, + "grad_norm": 0.4798663556575775, + "learning_rate": 0.000194520318482877, + "loss": 0.7863, + "step": 2396 + }, + { + "epoch": 0.4268162393162393, + "grad_norm": 0.5067816972732544, + "learning_rate": 0.0001945157476247207, + "loss": 1.0049, + "step": 2397 + }, + { + "epoch": 0.426994301994302, + "grad_norm": 0.5179638266563416, + "learning_rate": 0.00019451117491472468, + "loss": 1.1851, + "step": 2398 + }, + { + "epoch": 0.4271723646723647, + "grad_norm": 0.4782430827617645, + "learning_rate": 0.00019450660035297854, + "loss": 1.125, + "step": 2399 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 0.560077965259552, + "learning_rate": 0.00019450202393957186, + "loss": 1.1843, + "step": 2400 + }, + { + "epoch": 0.42752849002849, + "grad_norm": 0.5247970223426819, + "learning_rate": 0.00019449744567459436, + "loss": 1.1576, + "step": 2401 + }, + { + "epoch": 0.4277065527065527, + "grad_norm": 0.6414062976837158, + "learning_rate": 0.00019449286555813568, + "loss": 1.1833, + "step": 2402 + }, + { + "epoch": 0.42788461538461536, + "grad_norm": 0.5006586909294128, + "learning_rate": 0.00019448828359028563, + "loss": 1.1778, + "step": 2403 + }, + { + "epoch": 0.42806267806267806, + "grad_norm": 0.4946450889110565, + "learning_rate": 0.0001944836997711339, + "loss": 1.1611, + "step": 2404 + }, + { + "epoch": 0.42824074074074076, + "grad_norm": 0.4601200222969055, + "learning_rate": 0.00019447911410077037, + "loss": 1.2456, + "step": 2405 + }, + { + "epoch": 0.4284188034188034, + "grad_norm": 0.4653947651386261, + "learning_rate": 0.00019447452657928485, + "loss": 1.0941, + "step": 2406 + }, + { + "epoch": 0.4285968660968661, + "grad_norm": 0.5015713572502136, + "learning_rate": 0.00019446993720676726, + "loss": 1.3113, + "step": 2407 + }, + { + "epoch": 0.4287749287749288, + "grad_norm": 0.5803143978118896, + "learning_rate": 0.0001944653459833075, + "loss": 1.0568, + "step": 2408 + }, + { + "epoch": 0.42895299145299143, + "grad_norm": 0.5259647965431213, + "learning_rate": 0.0001944607529089955, + "loss": 1.1243, + "step": 2409 + }, + { + "epoch": 0.4291310541310541, + "grad_norm": 0.5150414109230042, + "learning_rate": 0.00019445615798392124, + "loss": 1.0676, + "step": 2410 + }, + { + "epoch": 0.4293091168091168, + "grad_norm": 0.5848649740219116, + "learning_rate": 0.0001944515612081748, + "loss": 1.0671, + "step": 2411 + }, + { + "epoch": 0.42948717948717946, + "grad_norm": 0.5696990489959717, + "learning_rate": 0.00019444696258184626, + "loss": 1.3323, + "step": 2412 + }, + { + "epoch": 0.42966524216524216, + "grad_norm": 0.49822330474853516, + "learning_rate": 0.00019444236210502567, + "loss": 1.1004, + "step": 2413 + }, + { + "epoch": 0.42984330484330485, + "grad_norm": 0.4683490991592407, + "learning_rate": 0.00019443775977780317, + "loss": 0.9768, + "step": 2414 + }, + { + "epoch": 0.43002136752136755, + "grad_norm": 0.5703811049461365, + "learning_rate": 0.00019443315560026893, + "loss": 1.154, + "step": 2415 + }, + { + "epoch": 0.4301994301994302, + "grad_norm": 0.5121861100196838, + "learning_rate": 0.0001944285495725132, + "loss": 1.1388, + "step": 2416 + }, + { + "epoch": 0.4303774928774929, + "grad_norm": 0.4864094853401184, + "learning_rate": 0.00019442394169462619, + "loss": 0.9214, + "step": 2417 + }, + { + "epoch": 0.4305555555555556, + "grad_norm": 0.5234864354133606, + "learning_rate": 0.0001944193319666982, + "loss": 1.2787, + "step": 2418 + }, + { + "epoch": 0.4307336182336182, + "grad_norm": 0.5137650370597839, + "learning_rate": 0.00019441472038881955, + "loss": 1.1406, + "step": 2419 + }, + { + "epoch": 0.4309116809116809, + "grad_norm": 0.49687784910202026, + "learning_rate": 0.00019441010696108054, + "loss": 0.93, + "step": 2420 + }, + { + "epoch": 0.4310897435897436, + "grad_norm": 0.5078722834587097, + "learning_rate": 0.00019440549168357163, + "loss": 1.1417, + "step": 2421 + }, + { + "epoch": 0.43126780626780625, + "grad_norm": 0.4483391046524048, + "learning_rate": 0.00019440087455638324, + "loss": 0.9016, + "step": 2422 + }, + { + "epoch": 0.43144586894586895, + "grad_norm": 0.5963045954704285, + "learning_rate": 0.00019439625557960576, + "loss": 1.1567, + "step": 2423 + }, + { + "epoch": 0.43162393162393164, + "grad_norm": 0.5534471273422241, + "learning_rate": 0.0001943916347533298, + "loss": 1.1409, + "step": 2424 + }, + { + "epoch": 0.4318019943019943, + "grad_norm": 0.6400241851806641, + "learning_rate": 0.0001943870120776458, + "loss": 1.2041, + "step": 2425 + }, + { + "epoch": 0.431980056980057, + "grad_norm": 0.4599420726299286, + "learning_rate": 0.0001943823875526444, + "loss": 1.023, + "step": 2426 + }, + { + "epoch": 0.4321581196581197, + "grad_norm": 0.4799708425998688, + "learning_rate": 0.00019437776117841614, + "loss": 1.0872, + "step": 2427 + }, + { + "epoch": 0.4323361823361823, + "grad_norm": 0.5138532519340515, + "learning_rate": 0.00019437313295505172, + "loss": 1.1175, + "step": 2428 + }, + { + "epoch": 0.432514245014245, + "grad_norm": 0.538223147392273, + "learning_rate": 0.00019436850288264183, + "loss": 1.1203, + "step": 2429 + }, + { + "epoch": 0.4326923076923077, + "grad_norm": 0.458044171333313, + "learning_rate": 0.00019436387096127713, + "loss": 1.0383, + "step": 2430 + }, + { + "epoch": 0.43287037037037035, + "grad_norm": 0.5928303599357605, + "learning_rate": 0.00019435923719104842, + "loss": 1.1191, + "step": 2431 + }, + { + "epoch": 0.43304843304843305, + "grad_norm": 0.5818437933921814, + "learning_rate": 0.00019435460157204645, + "loss": 1.0352, + "step": 2432 + }, + { + "epoch": 0.43322649572649574, + "grad_norm": 0.487341046333313, + "learning_rate": 0.0001943499641043621, + "loss": 1.2608, + "step": 2433 + }, + { + "epoch": 0.4334045584045584, + "grad_norm": 0.4737292230129242, + "learning_rate": 0.0001943453247880862, + "loss": 1.0084, + "step": 2434 + }, + { + "epoch": 0.4335826210826211, + "grad_norm": 0.4251207709312439, + "learning_rate": 0.0001943406836233096, + "loss": 0.9163, + "step": 2435 + }, + { + "epoch": 0.4337606837606838, + "grad_norm": 0.49468478560447693, + "learning_rate": 0.00019433604061012331, + "loss": 1.0293, + "step": 2436 + }, + { + "epoch": 0.4339387464387464, + "grad_norm": 0.47120022773742676, + "learning_rate": 0.00019433139574861826, + "loss": 1.0097, + "step": 2437 + }, + { + "epoch": 0.4341168091168091, + "grad_norm": 0.5060358047485352, + "learning_rate": 0.00019432674903888548, + "loss": 1.0683, + "step": 2438 + }, + { + "epoch": 0.4342948717948718, + "grad_norm": 0.5455917119979858, + "learning_rate": 0.00019432210048101598, + "loss": 0.8886, + "step": 2439 + }, + { + "epoch": 0.43447293447293445, + "grad_norm": 0.7960546612739563, + "learning_rate": 0.00019431745007510086, + "loss": 0.8648, + "step": 2440 + }, + { + "epoch": 0.43465099715099714, + "grad_norm": 0.5069689154624939, + "learning_rate": 0.00019431279782123126, + "loss": 1.1315, + "step": 2441 + }, + { + "epoch": 0.43482905982905984, + "grad_norm": 0.5597776174545288, + "learning_rate": 0.0001943081437194983, + "loss": 1.2281, + "step": 2442 + }, + { + "epoch": 0.43500712250712253, + "grad_norm": 0.4527420997619629, + "learning_rate": 0.00019430348776999315, + "loss": 0.7576, + "step": 2443 + }, + { + "epoch": 0.4351851851851852, + "grad_norm": 0.5625936388969421, + "learning_rate": 0.00019429882997280706, + "loss": 1.0302, + "step": 2444 + }, + { + "epoch": 0.43536324786324787, + "grad_norm": 0.5173513293266296, + "learning_rate": 0.0001942941703280313, + "loss": 1.2255, + "step": 2445 + }, + { + "epoch": 0.43554131054131057, + "grad_norm": 0.45889151096343994, + "learning_rate": 0.00019428950883575714, + "loss": 0.9322, + "step": 2446 + }, + { + "epoch": 0.4357193732193732, + "grad_norm": 0.5288477540016174, + "learning_rate": 0.00019428484549607593, + "loss": 1.0572, + "step": 2447 + }, + { + "epoch": 0.4358974358974359, + "grad_norm": 0.48328033089637756, + "learning_rate": 0.00019428018030907902, + "loss": 1.1213, + "step": 2448 + }, + { + "epoch": 0.4360754985754986, + "grad_norm": 0.5146737098693848, + "learning_rate": 0.00019427551327485786, + "loss": 0.9633, + "step": 2449 + }, + { + "epoch": 0.43625356125356124, + "grad_norm": 0.5138360261917114, + "learning_rate": 0.00019427084439350382, + "loss": 1.0561, + "step": 2450 + }, + { + "epoch": 0.43643162393162394, + "grad_norm": 0.5192533135414124, + "learning_rate": 0.00019426617366510843, + "loss": 1.1704, + "step": 2451 + }, + { + "epoch": 0.43660968660968663, + "grad_norm": 0.4819495379924774, + "learning_rate": 0.00019426150108976318, + "loss": 1.0958, + "step": 2452 + }, + { + "epoch": 0.43678774928774927, + "grad_norm": 0.4626680910587311, + "learning_rate": 0.00019425682666755965, + "loss": 1.1872, + "step": 2453 + }, + { + "epoch": 0.43696581196581197, + "grad_norm": 0.5773931741714478, + "learning_rate": 0.00019425215039858937, + "loss": 1.0722, + "step": 2454 + }, + { + "epoch": 0.43714387464387466, + "grad_norm": 0.5003872513771057, + "learning_rate": 0.00019424747228294402, + "loss": 1.0561, + "step": 2455 + }, + { + "epoch": 0.4373219373219373, + "grad_norm": 0.47370314598083496, + "learning_rate": 0.0001942427923207152, + "loss": 1.1619, + "step": 2456 + }, + { + "epoch": 0.4375, + "grad_norm": 0.466421514749527, + "learning_rate": 0.00019423811051199466, + "loss": 1.1311, + "step": 2457 + }, + { + "epoch": 0.4376780626780627, + "grad_norm": 0.44564682245254517, + "learning_rate": 0.00019423342685687413, + "loss": 1.1889, + "step": 2458 + }, + { + "epoch": 0.43785612535612534, + "grad_norm": 0.40986698865890503, + "learning_rate": 0.00019422874135544533, + "loss": 0.7312, + "step": 2459 + }, + { + "epoch": 0.43803418803418803, + "grad_norm": 0.4714358448982239, + "learning_rate": 0.0001942240540078001, + "loss": 0.9273, + "step": 2460 + }, + { + "epoch": 0.43821225071225073, + "grad_norm": 0.5298398733139038, + "learning_rate": 0.00019421936481403025, + "loss": 1.3377, + "step": 2461 + }, + { + "epoch": 0.43839031339031337, + "grad_norm": 0.6326695680618286, + "learning_rate": 0.0001942146737742277, + "loss": 1.0258, + "step": 2462 + }, + { + "epoch": 0.43856837606837606, + "grad_norm": 0.5087653994560242, + "learning_rate": 0.00019420998088848427, + "loss": 1.0007, + "step": 2463 + }, + { + "epoch": 0.43874643874643876, + "grad_norm": 0.4895429313182831, + "learning_rate": 0.00019420528615689202, + "loss": 1.0032, + "step": 2464 + }, + { + "epoch": 0.4389245014245014, + "grad_norm": 0.5029937028884888, + "learning_rate": 0.00019420058957954285, + "loss": 1.2877, + "step": 2465 + }, + { + "epoch": 0.4391025641025641, + "grad_norm": 0.4953192174434662, + "learning_rate": 0.00019419589115652884, + "loss": 1.0759, + "step": 2466 + }, + { + "epoch": 0.4392806267806268, + "grad_norm": 0.5081778168678284, + "learning_rate": 0.000194191190887942, + "loss": 0.8816, + "step": 2467 + }, + { + "epoch": 0.43945868945868943, + "grad_norm": 0.5065913200378418, + "learning_rate": 0.00019418648877387446, + "loss": 1.0362, + "step": 2468 + }, + { + "epoch": 0.43963675213675213, + "grad_norm": 0.540600061416626, + "learning_rate": 0.00019418178481441832, + "loss": 1.0911, + "step": 2469 + }, + { + "epoch": 0.4398148148148148, + "grad_norm": 0.5122954845428467, + "learning_rate": 0.00019417707900966572, + "loss": 0.9866, + "step": 2470 + }, + { + "epoch": 0.43999287749287747, + "grad_norm": 0.5380190014839172, + "learning_rate": 0.00019417237135970893, + "loss": 1.2775, + "step": 2471 + }, + { + "epoch": 0.44017094017094016, + "grad_norm": 1.2977570295333862, + "learning_rate": 0.00019416766186464016, + "loss": 1.3993, + "step": 2472 + }, + { + "epoch": 0.44034900284900286, + "grad_norm": 0.48105308413505554, + "learning_rate": 0.00019416295052455165, + "loss": 0.9369, + "step": 2473 + }, + { + "epoch": 0.44052706552706555, + "grad_norm": 0.4742157459259033, + "learning_rate": 0.00019415823733953574, + "loss": 1.101, + "step": 2474 + }, + { + "epoch": 0.4407051282051282, + "grad_norm": 0.4958631694316864, + "learning_rate": 0.00019415352230968473, + "loss": 0.9906, + "step": 2475 + }, + { + "epoch": 0.4408831908831909, + "grad_norm": 0.5808146595954895, + "learning_rate": 0.00019414880543509107, + "loss": 1.2315, + "step": 2476 + }, + { + "epoch": 0.4410612535612536, + "grad_norm": 0.4294755160808563, + "learning_rate": 0.00019414408671584714, + "loss": 0.8275, + "step": 2477 + }, + { + "epoch": 0.4412393162393162, + "grad_norm": 0.5346055626869202, + "learning_rate": 0.0001941393661520454, + "loss": 1.2432, + "step": 2478 + }, + { + "epoch": 0.4414173789173789, + "grad_norm": 0.5827590227127075, + "learning_rate": 0.00019413464374377833, + "loss": 1.3204, + "step": 2479 + }, + { + "epoch": 0.4415954415954416, + "grad_norm": 0.45688143372535706, + "learning_rate": 0.00019412991949113847, + "loss": 0.9307, + "step": 2480 + }, + { + "epoch": 0.44177350427350426, + "grad_norm": 0.512999415397644, + "learning_rate": 0.0001941251933942184, + "loss": 1.2808, + "step": 2481 + }, + { + "epoch": 0.44195156695156695, + "grad_norm": 0.4546334445476532, + "learning_rate": 0.00019412046545311064, + "loss": 1.0156, + "step": 2482 + }, + { + "epoch": 0.44212962962962965, + "grad_norm": 0.48552581667900085, + "learning_rate": 0.00019411573566790793, + "loss": 1.3798, + "step": 2483 + }, + { + "epoch": 0.4423076923076923, + "grad_norm": 0.511970579624176, + "learning_rate": 0.00019411100403870287, + "loss": 1.065, + "step": 2484 + }, + { + "epoch": 0.442485754985755, + "grad_norm": 0.6367824077606201, + "learning_rate": 0.00019410627056558815, + "loss": 1.3242, + "step": 2485 + }, + { + "epoch": 0.4426638176638177, + "grad_norm": 0.48913368582725525, + "learning_rate": 0.00019410153524865659, + "loss": 0.9761, + "step": 2486 + }, + { + "epoch": 0.4428418803418803, + "grad_norm": 0.5077710151672363, + "learning_rate": 0.0001940967980880009, + "loss": 1.1023, + "step": 2487 + }, + { + "epoch": 0.443019943019943, + "grad_norm": 0.4956335723400116, + "learning_rate": 0.00019409205908371395, + "loss": 1.1788, + "step": 2488 + }, + { + "epoch": 0.4431980056980057, + "grad_norm": 0.4726616442203522, + "learning_rate": 0.00019408731823588853, + "loss": 1.1445, + "step": 2489 + }, + { + "epoch": 0.44337606837606836, + "grad_norm": 0.5676438212394714, + "learning_rate": 0.00019408257554461757, + "loss": 1.0344, + "step": 2490 + }, + { + "epoch": 0.44355413105413105, + "grad_norm": 0.537656843662262, + "learning_rate": 0.000194077831009994, + "loss": 0.9876, + "step": 2491 + }, + { + "epoch": 0.44373219373219375, + "grad_norm": 0.517905592918396, + "learning_rate": 0.00019407308463211074, + "loss": 1.1389, + "step": 2492 + }, + { + "epoch": 0.4439102564102564, + "grad_norm": 0.49227026104927063, + "learning_rate": 0.0001940683364110608, + "loss": 1.0351, + "step": 2493 + }, + { + "epoch": 0.4440883190883191, + "grad_norm": 0.5131173729896545, + "learning_rate": 0.00019406358634693725, + "loss": 1.0351, + "step": 2494 + }, + { + "epoch": 0.4442663817663818, + "grad_norm": 0.5064495205879211, + "learning_rate": 0.0001940588344398331, + "loss": 1.0248, + "step": 2495 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 0.44107526540756226, + "learning_rate": 0.00019405408068984148, + "loss": 0.8068, + "step": 2496 + }, + { + "epoch": 0.4446225071225071, + "grad_norm": 0.6711848378181458, + "learning_rate": 0.00019404932509705554, + "loss": 1.059, + "step": 2497 + }, + { + "epoch": 0.4448005698005698, + "grad_norm": 0.5862596035003662, + "learning_rate": 0.00019404456766156845, + "loss": 1.2012, + "step": 2498 + }, + { + "epoch": 0.44497863247863245, + "grad_norm": 0.5528512001037598, + "learning_rate": 0.0001940398083834734, + "loss": 1.1121, + "step": 2499 + }, + { + "epoch": 0.44515669515669515, + "grad_norm": 0.5326655507087708, + "learning_rate": 0.0001940350472628637, + "loss": 1.166, + "step": 2500 + }, + { + "epoch": 0.44533475783475784, + "grad_norm": 0.5384873747825623, + "learning_rate": 0.00019403028429983252, + "loss": 1.4111, + "step": 2501 + }, + { + "epoch": 0.44551282051282054, + "grad_norm": 0.5142310857772827, + "learning_rate": 0.0001940255194944733, + "loss": 1.3353, + "step": 2502 + }, + { + "epoch": 0.4456908831908832, + "grad_norm": 0.49124231934547424, + "learning_rate": 0.0001940207528468793, + "loss": 1.1443, + "step": 2503 + }, + { + "epoch": 0.4458689458689459, + "grad_norm": 0.509713888168335, + "learning_rate": 0.000194015984357144, + "loss": 1.1857, + "step": 2504 + }, + { + "epoch": 0.44604700854700857, + "grad_norm": 0.5211275219917297, + "learning_rate": 0.00019401121402536078, + "loss": 0.9911, + "step": 2505 + }, + { + "epoch": 0.4462250712250712, + "grad_norm": 0.480340838432312, + "learning_rate": 0.00019400644185162312, + "loss": 1.1018, + "step": 2506 + }, + { + "epoch": 0.4464031339031339, + "grad_norm": 0.4212559163570404, + "learning_rate": 0.00019400166783602448, + "loss": 0.7501, + "step": 2507 + }, + { + "epoch": 0.4465811965811966, + "grad_norm": 0.5110511183738708, + "learning_rate": 0.00019399689197865846, + "loss": 1.1244, + "step": 2508 + }, + { + "epoch": 0.44675925925925924, + "grad_norm": 0.5604230165481567, + "learning_rate": 0.0001939921142796186, + "loss": 1.1066, + "step": 2509 + }, + { + "epoch": 0.44693732193732194, + "grad_norm": 0.5578675270080566, + "learning_rate": 0.0001939873347389985, + "loss": 1.0514, + "step": 2510 + }, + { + "epoch": 0.44711538461538464, + "grad_norm": 0.520908772945404, + "learning_rate": 0.00019398255335689184, + "loss": 1.1217, + "step": 2511 + }, + { + "epoch": 0.4472934472934473, + "grad_norm": 0.4405131936073303, + "learning_rate": 0.00019397777013339224, + "loss": 1.043, + "step": 2512 + }, + { + "epoch": 0.44747150997151, + "grad_norm": 0.5217751860618591, + "learning_rate": 0.0001939729850685935, + "loss": 1.1301, + "step": 2513 + }, + { + "epoch": 0.44764957264957267, + "grad_norm": 0.6151493191719055, + "learning_rate": 0.00019396819816258932, + "loss": 1.3498, + "step": 2514 + }, + { + "epoch": 0.4478276353276353, + "grad_norm": 0.5622836947441101, + "learning_rate": 0.0001939634094154735, + "loss": 1.146, + "step": 2515 + }, + { + "epoch": 0.448005698005698, + "grad_norm": 0.4671688973903656, + "learning_rate": 0.00019395861882733984, + "loss": 0.9456, + "step": 2516 + }, + { + "epoch": 0.4481837606837607, + "grad_norm": 0.453951358795166, + "learning_rate": 0.00019395382639828223, + "loss": 1.0042, + "step": 2517 + }, + { + "epoch": 0.44836182336182334, + "grad_norm": 0.5150699615478516, + "learning_rate": 0.0001939490321283946, + "loss": 1.166, + "step": 2518 + }, + { + "epoch": 0.44853988603988604, + "grad_norm": 0.5718298554420471, + "learning_rate": 0.0001939442360177708, + "loss": 1.2033, + "step": 2519 + }, + { + "epoch": 0.44871794871794873, + "grad_norm": 0.5306782126426697, + "learning_rate": 0.00019393943806650488, + "loss": 1.0765, + "step": 2520 + }, + { + "epoch": 0.4488960113960114, + "grad_norm": 0.47633033990859985, + "learning_rate": 0.0001939346382746908, + "loss": 0.9957, + "step": 2521 + }, + { + "epoch": 0.44907407407407407, + "grad_norm": 0.496441513299942, + "learning_rate": 0.00019392983664242262, + "loss": 1.2016, + "step": 2522 + }, + { + "epoch": 0.44925213675213677, + "grad_norm": 0.45956477522850037, + "learning_rate": 0.00019392503316979442, + "loss": 1.026, + "step": 2523 + }, + { + "epoch": 0.4494301994301994, + "grad_norm": 0.5400575995445251, + "learning_rate": 0.0001939202278569003, + "loss": 1.0785, + "step": 2524 + }, + { + "epoch": 0.4496082621082621, + "grad_norm": 0.4847868084907532, + "learning_rate": 0.00019391542070383442, + "loss": 1.013, + "step": 2525 + }, + { + "epoch": 0.4497863247863248, + "grad_norm": 0.4694063663482666, + "learning_rate": 0.00019391061171069094, + "loss": 0.8793, + "step": 2526 + }, + { + "epoch": 0.44996438746438744, + "grad_norm": 0.5158169269561768, + "learning_rate": 0.00019390580087756413, + "loss": 0.9602, + "step": 2527 + }, + { + "epoch": 0.45014245014245013, + "grad_norm": 0.5404585003852844, + "learning_rate": 0.00019390098820454822, + "loss": 1.2247, + "step": 2528 + }, + { + "epoch": 0.45032051282051283, + "grad_norm": 0.5302738547325134, + "learning_rate": 0.00019389617369173752, + "loss": 0.918, + "step": 2529 + }, + { + "epoch": 0.45049857549857547, + "grad_norm": 0.5065485835075378, + "learning_rate": 0.00019389135733922634, + "loss": 1.0934, + "step": 2530 + }, + { + "epoch": 0.45067663817663817, + "grad_norm": 0.5491471886634827, + "learning_rate": 0.00019388653914710903, + "loss": 1.0736, + "step": 2531 + }, + { + "epoch": 0.45085470085470086, + "grad_norm": 0.4850206971168518, + "learning_rate": 0.00019388171911548005, + "loss": 1.2401, + "step": 2532 + }, + { + "epoch": 0.45103276353276356, + "grad_norm": 0.5419789552688599, + "learning_rate": 0.0001938768972444338, + "loss": 1.269, + "step": 2533 + }, + { + "epoch": 0.4512108262108262, + "grad_norm": 0.4209023714065552, + "learning_rate": 0.00019387207353406476, + "loss": 1.0544, + "step": 2534 + }, + { + "epoch": 0.4513888888888889, + "grad_norm": 0.578588604927063, + "learning_rate": 0.00019386724798446743, + "loss": 1.0564, + "step": 2535 + }, + { + "epoch": 0.4515669515669516, + "grad_norm": 0.5277524590492249, + "learning_rate": 0.00019386242059573638, + "loss": 1.1497, + "step": 2536 + }, + { + "epoch": 0.45174501424501423, + "grad_norm": 0.5536073446273804, + "learning_rate": 0.0001938575913679662, + "loss": 1.2213, + "step": 2537 + }, + { + "epoch": 0.4519230769230769, + "grad_norm": 0.5572254657745361, + "learning_rate": 0.00019385276030125143, + "loss": 1.0231, + "step": 2538 + }, + { + "epoch": 0.4521011396011396, + "grad_norm": 0.493847131729126, + "learning_rate": 0.00019384792739568686, + "loss": 0.9385, + "step": 2539 + }, + { + "epoch": 0.45227920227920226, + "grad_norm": 0.4641396403312683, + "learning_rate": 0.00019384309265136707, + "loss": 0.9332, + "step": 2540 + }, + { + "epoch": 0.45245726495726496, + "grad_norm": 0.5439442992210388, + "learning_rate": 0.00019383825606838681, + "loss": 1.317, + "step": 2541 + }, + { + "epoch": 0.45263532763532766, + "grad_norm": 0.7050970792770386, + "learning_rate": 0.00019383341764684086, + "loss": 0.9508, + "step": 2542 + }, + { + "epoch": 0.4528133903133903, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.000193828577386824, + "loss": 1.2704, + "step": 2543 + }, + { + "epoch": 0.452991452991453, + "grad_norm": 0.47641924023628235, + "learning_rate": 0.0001938237352884311, + "loss": 1.0101, + "step": 2544 + }, + { + "epoch": 0.4531695156695157, + "grad_norm": 0.5223637819290161, + "learning_rate": 0.000193818891351757, + "loss": 1.0548, + "step": 2545 + }, + { + "epoch": 0.45334757834757833, + "grad_norm": 0.49065667390823364, + "learning_rate": 0.0001938140455768966, + "loss": 1.0927, + "step": 2546 + }, + { + "epoch": 0.453525641025641, + "grad_norm": 0.4808312654495239, + "learning_rate": 0.0001938091979639449, + "loss": 1.0599, + "step": 2547 + }, + { + "epoch": 0.4537037037037037, + "grad_norm": 0.5157489776611328, + "learning_rate": 0.0001938043485129968, + "loss": 1.2596, + "step": 2548 + }, + { + "epoch": 0.45388176638176636, + "grad_norm": 0.5983387231826782, + "learning_rate": 0.0001937994972241474, + "loss": 1.2276, + "step": 2549 + }, + { + "epoch": 0.45405982905982906, + "grad_norm": 0.49776506423950195, + "learning_rate": 0.00019379464409749163, + "loss": 1.3666, + "step": 2550 + }, + { + "epoch": 0.45423789173789175, + "grad_norm": 0.4693490266799927, + "learning_rate": 0.00019378978913312471, + "loss": 1.087, + "step": 2551 + }, + { + "epoch": 0.4544159544159544, + "grad_norm": 0.4754335880279541, + "learning_rate": 0.00019378493233114167, + "loss": 1.1282, + "step": 2552 + }, + { + "epoch": 0.4545940170940171, + "grad_norm": 0.5852862000465393, + "learning_rate": 0.00019378007369163776, + "loss": 1.1113, + "step": 2553 + }, + { + "epoch": 0.4547720797720798, + "grad_norm": 0.47442635893821716, + "learning_rate": 0.00019377521321470805, + "loss": 0.983, + "step": 2554 + }, + { + "epoch": 0.4549501424501424, + "grad_norm": 0.47432273626327515, + "learning_rate": 0.00019377035090044787, + "loss": 1.0169, + "step": 2555 + }, + { + "epoch": 0.4551282051282051, + "grad_norm": 0.4929196834564209, + "learning_rate": 0.00019376548674895246, + "loss": 1.0182, + "step": 2556 + }, + { + "epoch": 0.4553062678062678, + "grad_norm": 0.5433184504508972, + "learning_rate": 0.00019376062076031708, + "loss": 1.1339, + "step": 2557 + }, + { + "epoch": 0.45548433048433046, + "grad_norm": 0.47430408000946045, + "learning_rate": 0.00019375575293463715, + "loss": 1.1589, + "step": 2558 + }, + { + "epoch": 0.45566239316239315, + "grad_norm": 0.46641045808792114, + "learning_rate": 0.000193750883272008, + "loss": 1.029, + "step": 2559 + }, + { + "epoch": 0.45584045584045585, + "grad_norm": 0.44476228952407837, + "learning_rate": 0.00019374601177252502, + "loss": 0.8494, + "step": 2560 + }, + { + "epoch": 0.45601851851851855, + "grad_norm": 0.4886183440685272, + "learning_rate": 0.00019374113843628366, + "loss": 1.1374, + "step": 2561 + }, + { + "epoch": 0.4561965811965812, + "grad_norm": 0.4786703288555145, + "learning_rate": 0.00019373626326337946, + "loss": 1.2861, + "step": 2562 + }, + { + "epoch": 0.4563746438746439, + "grad_norm": 0.5752716660499573, + "learning_rate": 0.0001937313862539079, + "loss": 1.2365, + "step": 2563 + }, + { + "epoch": 0.4565527065527066, + "grad_norm": 0.519176185131073, + "learning_rate": 0.00019372650740796452, + "loss": 1.2264, + "step": 2564 + }, + { + "epoch": 0.4567307692307692, + "grad_norm": 0.5927292704582214, + "learning_rate": 0.00019372162672564493, + "loss": 0.8979, + "step": 2565 + }, + { + "epoch": 0.4569088319088319, + "grad_norm": 0.5467435121536255, + "learning_rate": 0.00019371674420704478, + "loss": 1.1016, + "step": 2566 + }, + { + "epoch": 0.4570868945868946, + "grad_norm": 0.49593284726142883, + "learning_rate": 0.00019371185985225968, + "loss": 0.982, + "step": 2567 + }, + { + "epoch": 0.45726495726495725, + "grad_norm": 0.5696587562561035, + "learning_rate": 0.00019370697366138538, + "loss": 0.979, + "step": 2568 + }, + { + "epoch": 0.45744301994301995, + "grad_norm": 0.4455752968788147, + "learning_rate": 0.00019370208563451757, + "loss": 0.8832, + "step": 2569 + }, + { + "epoch": 0.45762108262108264, + "grad_norm": 0.5072923302650452, + "learning_rate": 0.00019369719577175203, + "loss": 1.1046, + "step": 2570 + }, + { + "epoch": 0.4577991452991453, + "grad_norm": 0.45119982957839966, + "learning_rate": 0.0001936923040731846, + "loss": 1.0083, + "step": 2571 + }, + { + "epoch": 0.457977207977208, + "grad_norm": 0.5062251091003418, + "learning_rate": 0.00019368741053891108, + "loss": 1.2771, + "step": 2572 + }, + { + "epoch": 0.4581552706552707, + "grad_norm": 0.5511104464530945, + "learning_rate": 0.0001936825151690274, + "loss": 1.0039, + "step": 2573 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 0.4721006453037262, + "learning_rate": 0.0001936776179636294, + "loss": 1.3246, + "step": 2574 + }, + { + "epoch": 0.458511396011396, + "grad_norm": 0.5021488666534424, + "learning_rate": 0.0001936727189228131, + "loss": 1.1733, + "step": 2575 + }, + { + "epoch": 0.4586894586894587, + "grad_norm": 0.5755292177200317, + "learning_rate": 0.0001936678180466745, + "loss": 1.2241, + "step": 2576 + }, + { + "epoch": 0.45886752136752135, + "grad_norm": 0.4501610994338989, + "learning_rate": 0.00019366291533530952, + "loss": 1.0503, + "step": 2577 + }, + { + "epoch": 0.45904558404558404, + "grad_norm": 0.4067458212375641, + "learning_rate": 0.00019365801078881432, + "loss": 0.8259, + "step": 2578 + }, + { + "epoch": 0.45922364672364674, + "grad_norm": 0.539730429649353, + "learning_rate": 0.0001936531044072849, + "loss": 1.1964, + "step": 2579 + }, + { + "epoch": 0.4594017094017094, + "grad_norm": 0.5624797344207764, + "learning_rate": 0.0001936481961908175, + "loss": 1.2059, + "step": 2580 + }, + { + "epoch": 0.4595797720797721, + "grad_norm": 0.43679240345954895, + "learning_rate": 0.00019364328613950824, + "loss": 1.1371, + "step": 2581 + }, + { + "epoch": 0.45975783475783477, + "grad_norm": 0.5214769244194031, + "learning_rate": 0.00019363837425345328, + "loss": 1.109, + "step": 2582 + }, + { + "epoch": 0.4599358974358974, + "grad_norm": 0.4522894024848938, + "learning_rate": 0.00019363346053274892, + "loss": 1.0532, + "step": 2583 + }, + { + "epoch": 0.4601139601139601, + "grad_norm": 0.44980281591415405, + "learning_rate": 0.0001936285449774914, + "loss": 0.9352, + "step": 2584 + }, + { + "epoch": 0.4602920227920228, + "grad_norm": 0.5697414875030518, + "learning_rate": 0.00019362362758777705, + "loss": 1.2171, + "step": 2585 + }, + { + "epoch": 0.46047008547008544, + "grad_norm": 0.4636315107345581, + "learning_rate": 0.00019361870836370217, + "loss": 1.0662, + "step": 2586 + }, + { + "epoch": 0.46064814814814814, + "grad_norm": 0.5144017338752747, + "learning_rate": 0.00019361378730536321, + "loss": 1.0681, + "step": 2587 + }, + { + "epoch": 0.46082621082621084, + "grad_norm": 0.5007636547088623, + "learning_rate": 0.00019360886441285654, + "loss": 1.2058, + "step": 2588 + }, + { + "epoch": 0.46100427350427353, + "grad_norm": 0.5024117231369019, + "learning_rate": 0.00019360393968627864, + "loss": 1.065, + "step": 2589 + }, + { + "epoch": 0.46118233618233617, + "grad_norm": 0.48105588555336, + "learning_rate": 0.00019359901312572596, + "loss": 1.0887, + "step": 2590 + }, + { + "epoch": 0.46136039886039887, + "grad_norm": 0.5381982326507568, + "learning_rate": 0.00019359408473129506, + "loss": 1.2754, + "step": 2591 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 0.5051333904266357, + "learning_rate": 0.0001935891545030825, + "loss": 0.9334, + "step": 2592 + }, + { + "epoch": 0.4617165242165242, + "grad_norm": 0.43818601965904236, + "learning_rate": 0.0001935842224411849, + "loss": 1.0967, + "step": 2593 + }, + { + "epoch": 0.4618945868945869, + "grad_norm": 0.4727257490158081, + "learning_rate": 0.0001935792885456988, + "loss": 0.8136, + "step": 2594 + }, + { + "epoch": 0.4620726495726496, + "grad_norm": 0.5505291223526001, + "learning_rate": 0.00019357435281672098, + "loss": 1.3113, + "step": 2595 + }, + { + "epoch": 0.46225071225071224, + "grad_norm": 0.4705682396888733, + "learning_rate": 0.0001935694152543481, + "loss": 0.9863, + "step": 2596 + }, + { + "epoch": 0.46242877492877493, + "grad_norm": 0.49653419852256775, + "learning_rate": 0.0001935644758586769, + "loss": 1.035, + "step": 2597 + }, + { + "epoch": 0.46260683760683763, + "grad_norm": 0.4788367748260498, + "learning_rate": 0.00019355953462980415, + "loss": 1.1253, + "step": 2598 + }, + { + "epoch": 0.46278490028490027, + "grad_norm": 0.5295125842094421, + "learning_rate": 0.00019355459156782668, + "loss": 1.0853, + "step": 2599 + }, + { + "epoch": 0.46296296296296297, + "grad_norm": 0.4878056049346924, + "learning_rate": 0.00019354964667284133, + "loss": 1.1381, + "step": 2600 + }, + { + "epoch": 0.46314102564102566, + "grad_norm": 0.5442031025886536, + "learning_rate": 0.00019354469994494497, + "loss": 1.1349, + "step": 2601 + }, + { + "epoch": 0.4633190883190883, + "grad_norm": 0.4845225214958191, + "learning_rate": 0.00019353975138423457, + "loss": 1.0538, + "step": 2602 + }, + { + "epoch": 0.463497150997151, + "grad_norm": 0.4957871437072754, + "learning_rate": 0.00019353480099080703, + "loss": 1.2765, + "step": 2603 + }, + { + "epoch": 0.4636752136752137, + "grad_norm": 0.5414339303970337, + "learning_rate": 0.00019352984876475936, + "loss": 1.1015, + "step": 2604 + }, + { + "epoch": 0.46385327635327633, + "grad_norm": 0.5171043872833252, + "learning_rate": 0.0001935248947061886, + "loss": 0.9995, + "step": 2605 + }, + { + "epoch": 0.46403133903133903, + "grad_norm": 0.46040529012680054, + "learning_rate": 0.0001935199388151918, + "loss": 1.1126, + "step": 2606 + }, + { + "epoch": 0.4642094017094017, + "grad_norm": 0.5327033400535583, + "learning_rate": 0.00019351498109186613, + "loss": 1.1983, + "step": 2607 + }, + { + "epoch": 0.46438746438746437, + "grad_norm": 0.4451361298561096, + "learning_rate": 0.0001935100215363086, + "loss": 0.9689, + "step": 2608 + }, + { + "epoch": 0.46456552706552706, + "grad_norm": 0.5462809801101685, + "learning_rate": 0.00019350506014861646, + "loss": 1.036, + "step": 2609 + }, + { + "epoch": 0.46474358974358976, + "grad_norm": 0.4907000958919525, + "learning_rate": 0.00019350009692888694, + "loss": 1.0724, + "step": 2610 + }, + { + "epoch": 0.4649216524216524, + "grad_norm": 0.47523510456085205, + "learning_rate": 0.00019349513187721723, + "loss": 0.9214, + "step": 2611 + }, + { + "epoch": 0.4650997150997151, + "grad_norm": 0.539732813835144, + "learning_rate": 0.0001934901649937046, + "loss": 1.1166, + "step": 2612 + }, + { + "epoch": 0.4652777777777778, + "grad_norm": 0.4827860891819, + "learning_rate": 0.00019348519627844643, + "loss": 1.1613, + "step": 2613 + }, + { + "epoch": 0.46545584045584043, + "grad_norm": 0.5385223031044006, + "learning_rate": 0.00019348022573154, + "loss": 1.0105, + "step": 2614 + }, + { + "epoch": 0.4656339031339031, + "grad_norm": 0.4629383087158203, + "learning_rate": 0.0001934752533530828, + "loss": 1.0298, + "step": 2615 + }, + { + "epoch": 0.4658119658119658, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00019347027914317212, + "loss": 1.3158, + "step": 2616 + }, + { + "epoch": 0.46599002849002846, + "grad_norm": 0.5954698324203491, + "learning_rate": 0.00019346530310190553, + "loss": 1.1882, + "step": 2617 + }, + { + "epoch": 0.46616809116809116, + "grad_norm": 0.49185171723365784, + "learning_rate": 0.00019346032522938046, + "loss": 1.0977, + "step": 2618 + }, + { + "epoch": 0.46634615384615385, + "grad_norm": 0.5145422220230103, + "learning_rate": 0.0001934553455256945, + "loss": 0.9948, + "step": 2619 + }, + { + "epoch": 0.46652421652421655, + "grad_norm": 0.6809412837028503, + "learning_rate": 0.00019345036399094517, + "loss": 1.5798, + "step": 2620 + }, + { + "epoch": 0.4667022792022792, + "grad_norm": 0.4606841206550598, + "learning_rate": 0.00019344538062523005, + "loss": 0.7357, + "step": 2621 + }, + { + "epoch": 0.4668803418803419, + "grad_norm": 0.49036628007888794, + "learning_rate": 0.00019344039542864685, + "loss": 1.1518, + "step": 2622 + }, + { + "epoch": 0.4670584045584046, + "grad_norm": 0.47904539108276367, + "learning_rate": 0.0001934354084012932, + "loss": 0.9929, + "step": 2623 + }, + { + "epoch": 0.4672364672364672, + "grad_norm": 0.5224666595458984, + "learning_rate": 0.0001934304195432668, + "loss": 1.2544, + "step": 2624 + }, + { + "epoch": 0.4674145299145299, + "grad_norm": 0.4902483820915222, + "learning_rate": 0.00019342542885466543, + "loss": 1.0301, + "step": 2625 + }, + { + "epoch": 0.4675925925925926, + "grad_norm": 0.46824702620506287, + "learning_rate": 0.00019342043633558683, + "loss": 0.9364, + "step": 2626 + }, + { + "epoch": 0.46777065527065526, + "grad_norm": 0.46272051334381104, + "learning_rate": 0.00019341544198612888, + "loss": 1.056, + "step": 2627 + }, + { + "epoch": 0.46794871794871795, + "grad_norm": 0.6216606497764587, + "learning_rate": 0.0001934104458063894, + "loss": 1.0825, + "step": 2628 + }, + { + "epoch": 0.46812678062678065, + "grad_norm": 0.5024014115333557, + "learning_rate": 0.00019340544779646623, + "loss": 1.1832, + "step": 2629 + }, + { + "epoch": 0.4683048433048433, + "grad_norm": 0.5547130107879639, + "learning_rate": 0.00019340044795645737, + "loss": 1.1335, + "step": 2630 + }, + { + "epoch": 0.468482905982906, + "grad_norm": 0.5439161658287048, + "learning_rate": 0.0001933954462864608, + "loss": 1.0229, + "step": 2631 + }, + { + "epoch": 0.4686609686609687, + "grad_norm": 0.4782990515232086, + "learning_rate": 0.0001933904427865744, + "loss": 1.2318, + "step": 2632 + }, + { + "epoch": 0.4688390313390313, + "grad_norm": 0.5872140526771545, + "learning_rate": 0.00019338543745689633, + "loss": 1.0132, + "step": 2633 + }, + { + "epoch": 0.469017094017094, + "grad_norm": 0.44163307547569275, + "learning_rate": 0.00019338043029752458, + "loss": 1.0091, + "step": 2634 + }, + { + "epoch": 0.4691951566951567, + "grad_norm": 0.541081428527832, + "learning_rate": 0.0001933754213085573, + "loss": 1.2155, + "step": 2635 + }, + { + "epoch": 0.46937321937321935, + "grad_norm": 0.4761527478694916, + "learning_rate": 0.00019337041049009255, + "loss": 1.1138, + "step": 2636 + }, + { + "epoch": 0.46955128205128205, + "grad_norm": 0.46414369344711304, + "learning_rate": 0.0001933653978422286, + "loss": 0.9903, + "step": 2637 + }, + { + "epoch": 0.46972934472934474, + "grad_norm": 0.5337086915969849, + "learning_rate": 0.00019336038336506363, + "loss": 1.2873, + "step": 2638 + }, + { + "epoch": 0.4699074074074074, + "grad_norm": 0.5065379738807678, + "learning_rate": 0.00019335536705869592, + "loss": 1.1436, + "step": 2639 + }, + { + "epoch": 0.4700854700854701, + "grad_norm": 0.5539217591285706, + "learning_rate": 0.0001933503489232237, + "loss": 1.2881, + "step": 2640 + }, + { + "epoch": 0.4702635327635328, + "grad_norm": 0.48303213715553284, + "learning_rate": 0.0001933453289587453, + "loss": 1.0209, + "step": 2641 + }, + { + "epoch": 0.4704415954415954, + "grad_norm": 0.6986871957778931, + "learning_rate": 0.00019334030716535908, + "loss": 1.1979, + "step": 2642 + }, + { + "epoch": 0.4706196581196581, + "grad_norm": 0.46137234568595886, + "learning_rate": 0.00019333528354316347, + "loss": 1.0682, + "step": 2643 + }, + { + "epoch": 0.4707977207977208, + "grad_norm": 0.4726654291152954, + "learning_rate": 0.00019333025809225684, + "loss": 1.1712, + "step": 2644 + }, + { + "epoch": 0.47097578347578345, + "grad_norm": 0.46188637614250183, + "learning_rate": 0.0001933252308127377, + "loss": 1.0183, + "step": 2645 + }, + { + "epoch": 0.47115384615384615, + "grad_norm": 0.5323259830474854, + "learning_rate": 0.0001933202017047045, + "loss": 0.935, + "step": 2646 + }, + { + "epoch": 0.47133190883190884, + "grad_norm": 0.5004189014434814, + "learning_rate": 0.00019331517076825582, + "loss": 1.1331, + "step": 2647 + }, + { + "epoch": 0.47150997150997154, + "grad_norm": 0.5443634986877441, + "learning_rate": 0.0001933101380034902, + "loss": 1.0514, + "step": 2648 + }, + { + "epoch": 0.4716880341880342, + "grad_norm": 0.504180371761322, + "learning_rate": 0.0001933051034105063, + "loss": 1.3099, + "step": 2649 + }, + { + "epoch": 0.4718660968660969, + "grad_norm": 0.5092344284057617, + "learning_rate": 0.0001933000669894027, + "loss": 1.0716, + "step": 2650 + }, + { + "epoch": 0.47204415954415957, + "grad_norm": 0.5236422419548035, + "learning_rate": 0.0001932950287402781, + "loss": 1.0981, + "step": 2651 + }, + { + "epoch": 0.4722222222222222, + "grad_norm": 0.6228063702583313, + "learning_rate": 0.0001932899886632312, + "loss": 1.3398, + "step": 2652 + }, + { + "epoch": 0.4724002849002849, + "grad_norm": 0.5112748146057129, + "learning_rate": 0.00019328494675836078, + "loss": 1.0151, + "step": 2653 + }, + { + "epoch": 0.4725783475783476, + "grad_norm": 0.5554201602935791, + "learning_rate": 0.00019327990302576563, + "loss": 1.404, + "step": 2654 + }, + { + "epoch": 0.47275641025641024, + "grad_norm": 0.5050725340843201, + "learning_rate": 0.0001932748574655445, + "loss": 0.951, + "step": 2655 + }, + { + "epoch": 0.47293447293447294, + "grad_norm": 0.5161749720573425, + "learning_rate": 0.00019326981007779636, + "loss": 1.2425, + "step": 2656 + }, + { + "epoch": 0.47311253561253563, + "grad_norm": 0.4865442216396332, + "learning_rate": 0.00019326476086262002, + "loss": 1.1175, + "step": 2657 + }, + { + "epoch": 0.4732905982905983, + "grad_norm": 0.5276186466217041, + "learning_rate": 0.0001932597098201144, + "loss": 1.3687, + "step": 2658 + }, + { + "epoch": 0.47346866096866097, + "grad_norm": 0.509139358997345, + "learning_rate": 0.00019325465695037855, + "loss": 1.0546, + "step": 2659 + }, + { + "epoch": 0.47364672364672367, + "grad_norm": 0.49815434217453003, + "learning_rate": 0.00019324960225351138, + "loss": 1.0807, + "step": 2660 + }, + { + "epoch": 0.4738247863247863, + "grad_norm": 0.5059618353843689, + "learning_rate": 0.00019324454572961197, + "loss": 1.0827, + "step": 2661 + }, + { + "epoch": 0.474002849002849, + "grad_norm": 0.5698565244674683, + "learning_rate": 0.00019323948737877942, + "loss": 1.2019, + "step": 2662 + }, + { + "epoch": 0.4741809116809117, + "grad_norm": 0.49661511182785034, + "learning_rate": 0.00019323442720111276, + "loss": 1.1447, + "step": 2663 + }, + { + "epoch": 0.47435897435897434, + "grad_norm": 0.46442747116088867, + "learning_rate": 0.0001932293651967112, + "loss": 0.8796, + "step": 2664 + }, + { + "epoch": 0.47453703703703703, + "grad_norm": 0.48306044936180115, + "learning_rate": 0.00019322430136567388, + "loss": 1.1358, + "step": 2665 + }, + { + "epoch": 0.47471509971509973, + "grad_norm": 0.5677350759506226, + "learning_rate": 0.00019321923570810005, + "loss": 1.1026, + "step": 2666 + }, + { + "epoch": 0.47489316239316237, + "grad_norm": 0.3700144588947296, + "learning_rate": 0.0001932141682240889, + "loss": 0.7514, + "step": 2667 + }, + { + "epoch": 0.47507122507122507, + "grad_norm": 0.6003054976463318, + "learning_rate": 0.0001932090989137398, + "loss": 1.1591, + "step": 2668 + }, + { + "epoch": 0.47524928774928776, + "grad_norm": 0.520298421382904, + "learning_rate": 0.00019320402777715204, + "loss": 1.339, + "step": 2669 + }, + { + "epoch": 0.4754273504273504, + "grad_norm": 0.46453598141670227, + "learning_rate": 0.00019319895481442493, + "loss": 0.9879, + "step": 2670 + }, + { + "epoch": 0.4756054131054131, + "grad_norm": 0.5247363448143005, + "learning_rate": 0.00019319388002565793, + "loss": 0.9862, + "step": 2671 + }, + { + "epoch": 0.4757834757834758, + "grad_norm": 0.5498613715171814, + "learning_rate": 0.00019318880341095046, + "loss": 1.2224, + "step": 2672 + }, + { + "epoch": 0.47596153846153844, + "grad_norm": 0.565838098526001, + "learning_rate": 0.00019318372497040192, + "loss": 1.0712, + "step": 2673 + }, + { + "epoch": 0.47613960113960113, + "grad_norm": 0.5797489881515503, + "learning_rate": 0.00019317864470411191, + "loss": 1.0176, + "step": 2674 + }, + { + "epoch": 0.47631766381766383, + "grad_norm": 0.5114326477050781, + "learning_rate": 0.0001931735626121799, + "loss": 1.1027, + "step": 2675 + }, + { + "epoch": 0.47649572649572647, + "grad_norm": 0.5396515727043152, + "learning_rate": 0.00019316847869470547, + "loss": 1.1782, + "step": 2676 + }, + { + "epoch": 0.47667378917378916, + "grad_norm": 0.4812076985836029, + "learning_rate": 0.00019316339295178824, + "loss": 1.1196, + "step": 2677 + }, + { + "epoch": 0.47685185185185186, + "grad_norm": 0.4875647723674774, + "learning_rate": 0.00019315830538352787, + "loss": 1.1407, + "step": 2678 + }, + { + "epoch": 0.47702991452991456, + "grad_norm": 0.5036377906799316, + "learning_rate": 0.00019315321599002404, + "loss": 0.9842, + "step": 2679 + }, + { + "epoch": 0.4772079772079772, + "grad_norm": 0.5054177641868591, + "learning_rate": 0.00019314812477137645, + "loss": 0.8196, + "step": 2680 + }, + { + "epoch": 0.4773860398860399, + "grad_norm": 0.5050665736198425, + "learning_rate": 0.00019314303172768483, + "loss": 0.8463, + "step": 2681 + }, + { + "epoch": 0.4775641025641026, + "grad_norm": 0.5179004669189453, + "learning_rate": 0.000193137936859049, + "loss": 1.2485, + "step": 2682 + }, + { + "epoch": 0.47774216524216523, + "grad_norm": 0.44986143708229065, + "learning_rate": 0.00019313284016556876, + "loss": 0.9855, + "step": 2683 + }, + { + "epoch": 0.4779202279202279, + "grad_norm": 0.5594347715377808, + "learning_rate": 0.00019312774164734398, + "loss": 1.0987, + "step": 2684 + }, + { + "epoch": 0.4780982905982906, + "grad_norm": 0.4837244749069214, + "learning_rate": 0.0001931226413044746, + "loss": 1.1119, + "step": 2685 + }, + { + "epoch": 0.47827635327635326, + "grad_norm": 0.489145427942276, + "learning_rate": 0.0001931175391370605, + "loss": 1.1962, + "step": 2686 + }, + { + "epoch": 0.47845441595441596, + "grad_norm": 0.503568708896637, + "learning_rate": 0.00019311243514520164, + "loss": 0.9668, + "step": 2687 + }, + { + "epoch": 0.47863247863247865, + "grad_norm": 0.5401005744934082, + "learning_rate": 0.00019310732932899805, + "loss": 1.3072, + "step": 2688 + }, + { + "epoch": 0.4788105413105413, + "grad_norm": 0.526523768901825, + "learning_rate": 0.00019310222168854971, + "loss": 1.1387, + "step": 2689 + }, + { + "epoch": 0.478988603988604, + "grad_norm": 0.5223183631896973, + "learning_rate": 0.00019309711222395678, + "loss": 1.1391, + "step": 2690 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 0.5840879082679749, + "learning_rate": 0.00019309200093531933, + "loss": 1.1543, + "step": 2691 + }, + { + "epoch": 0.4793447293447293, + "grad_norm": 0.5173699259757996, + "learning_rate": 0.00019308688782273753, + "loss": 1.1889, + "step": 2692 + }, + { + "epoch": 0.479522792022792, + "grad_norm": 0.5417894124984741, + "learning_rate": 0.00019308177288631146, + "loss": 1.299, + "step": 2693 + }, + { + "epoch": 0.4797008547008547, + "grad_norm": 0.4890797734260559, + "learning_rate": 0.0001930766561261415, + "loss": 1.1516, + "step": 2694 + }, + { + "epoch": 0.47987891737891736, + "grad_norm": 0.5422119498252869, + "learning_rate": 0.00019307153754232772, + "loss": 1.0301, + "step": 2695 + }, + { + "epoch": 0.48005698005698005, + "grad_norm": 0.5838702917098999, + "learning_rate": 0.00019306641713497057, + "loss": 1.265, + "step": 2696 + }, + { + "epoch": 0.48023504273504275, + "grad_norm": 0.5020943284034729, + "learning_rate": 0.00019306129490417027, + "loss": 1.1119, + "step": 2697 + }, + { + "epoch": 0.4804131054131054, + "grad_norm": 0.412993460893631, + "learning_rate": 0.00019305617085002723, + "loss": 0.8083, + "step": 2698 + }, + { + "epoch": 0.4805911680911681, + "grad_norm": 0.6270101070404053, + "learning_rate": 0.00019305104497264184, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 0.4807692307692308, + "grad_norm": 0.45256730914115906, + "learning_rate": 0.0001930459172721145, + "loss": 1.0368, + "step": 2700 + }, + { + "epoch": 0.4809472934472934, + "grad_norm": 0.5351749658584595, + "learning_rate": 0.0001930407877485457, + "loss": 1.135, + "step": 2701 + }, + { + "epoch": 0.4811253561253561, + "grad_norm": 0.49324163794517517, + "learning_rate": 0.00019303565640203593, + "loss": 0.9383, + "step": 2702 + }, + { + "epoch": 0.4813034188034188, + "grad_norm": 0.5434361100196838, + "learning_rate": 0.00019303052323268576, + "loss": 1.2605, + "step": 2703 + }, + { + "epoch": 0.48148148148148145, + "grad_norm": 0.5858064889907837, + "learning_rate": 0.00019302538824059572, + "loss": 1.0846, + "step": 2704 + }, + { + "epoch": 0.48165954415954415, + "grad_norm": 0.5753700733184814, + "learning_rate": 0.00019302025142586647, + "loss": 1.0371, + "step": 2705 + }, + { + "epoch": 0.48183760683760685, + "grad_norm": 0.43102699518203735, + "learning_rate": 0.00019301511278859858, + "loss": 0.9189, + "step": 2706 + }, + { + "epoch": 0.48201566951566954, + "grad_norm": 0.4731025993824005, + "learning_rate": 0.0001930099723288928, + "loss": 1.1291, + "step": 2707 + }, + { + "epoch": 0.4821937321937322, + "grad_norm": 0.5685615539550781, + "learning_rate": 0.00019300483004684987, + "loss": 1.1006, + "step": 2708 + }, + { + "epoch": 0.4823717948717949, + "grad_norm": 0.4368155896663666, + "learning_rate": 0.00019299968594257044, + "loss": 0.9959, + "step": 2709 + }, + { + "epoch": 0.4825498575498576, + "grad_norm": 0.5594738125801086, + "learning_rate": 0.00019299454001615537, + "loss": 1.0826, + "step": 2710 + }, + { + "epoch": 0.4827279202279202, + "grad_norm": 0.48876598477363586, + "learning_rate": 0.00019298939226770548, + "loss": 1.1556, + "step": 2711 + }, + { + "epoch": 0.4829059829059829, + "grad_norm": 0.548039436340332, + "learning_rate": 0.00019298424269732157, + "loss": 1.158, + "step": 2712 + }, + { + "epoch": 0.4830840455840456, + "grad_norm": 0.4957645535469055, + "learning_rate": 0.00019297909130510464, + "loss": 0.9824, + "step": 2713 + }, + { + "epoch": 0.48326210826210825, + "grad_norm": 0.5197011232376099, + "learning_rate": 0.00019297393809115555, + "loss": 1.1074, + "step": 2714 + }, + { + "epoch": 0.48344017094017094, + "grad_norm": 0.5742064118385315, + "learning_rate": 0.00019296878305557526, + "loss": 1.0431, + "step": 2715 + }, + { + "epoch": 0.48361823361823364, + "grad_norm": 0.5698413252830505, + "learning_rate": 0.0001929636261984648, + "loss": 1.0713, + "step": 2716 + }, + { + "epoch": 0.4837962962962963, + "grad_norm": 0.48126333951950073, + "learning_rate": 0.0001929584675199252, + "loss": 0.9274, + "step": 2717 + }, + { + "epoch": 0.483974358974359, + "grad_norm": 0.49299830198287964, + "learning_rate": 0.00019295330702005754, + "loss": 0.9392, + "step": 2718 + }, + { + "epoch": 0.48415242165242167, + "grad_norm": 0.4780774414539337, + "learning_rate": 0.0001929481446989629, + "loss": 1.1459, + "step": 2719 + }, + { + "epoch": 0.4843304843304843, + "grad_norm": 0.5462654829025269, + "learning_rate": 0.00019294298055674248, + "loss": 1.0635, + "step": 2720 + }, + { + "epoch": 0.484508547008547, + "grad_norm": 0.5371061563491821, + "learning_rate": 0.00019293781459349743, + "loss": 1.3578, + "step": 2721 + }, + { + "epoch": 0.4846866096866097, + "grad_norm": 0.46308520436286926, + "learning_rate": 0.00019293264680932893, + "loss": 0.9001, + "step": 2722 + }, + { + "epoch": 0.48486467236467234, + "grad_norm": 0.5149807929992676, + "learning_rate": 0.0001929274772043383, + "loss": 0.6908, + "step": 2723 + }, + { + "epoch": 0.48504273504273504, + "grad_norm": 0.5435031056404114, + "learning_rate": 0.00019292230577862678, + "loss": 1.2143, + "step": 2724 + }, + { + "epoch": 0.48522079772079774, + "grad_norm": 0.44217726588249207, + "learning_rate": 0.00019291713253229568, + "loss": 0.9303, + "step": 2725 + }, + { + "epoch": 0.4853988603988604, + "grad_norm": 0.6120226383209229, + "learning_rate": 0.00019291195746544643, + "loss": 1.3801, + "step": 2726 + }, + { + "epoch": 0.4855769230769231, + "grad_norm": 0.5014316439628601, + "learning_rate": 0.00019290678057818037, + "loss": 1.0631, + "step": 2727 + }, + { + "epoch": 0.48575498575498577, + "grad_norm": 0.5667829513549805, + "learning_rate": 0.00019290160187059895, + "loss": 1.3166, + "step": 2728 + }, + { + "epoch": 0.4859330484330484, + "grad_norm": 0.5011509656906128, + "learning_rate": 0.0001928964213428036, + "loss": 1.1887, + "step": 2729 + }, + { + "epoch": 0.4861111111111111, + "grad_norm": 0.48317405581474304, + "learning_rate": 0.00019289123899489586, + "loss": 1.1125, + "step": 2730 + }, + { + "epoch": 0.4862891737891738, + "grad_norm": 0.4669005870819092, + "learning_rate": 0.00019288605482697726, + "loss": 1.0091, + "step": 2731 + }, + { + "epoch": 0.48646723646723644, + "grad_norm": 0.4330739974975586, + "learning_rate": 0.00019288086883914937, + "loss": 0.9789, + "step": 2732 + }, + { + "epoch": 0.48664529914529914, + "grad_norm": 0.48482781648635864, + "learning_rate": 0.0001928756810315138, + "loss": 1.1922, + "step": 2733 + }, + { + "epoch": 0.48682336182336183, + "grad_norm": 0.5781838297843933, + "learning_rate": 0.0001928704914041722, + "loss": 1.1793, + "step": 2734 + }, + { + "epoch": 0.48700142450142453, + "grad_norm": 0.5955413579940796, + "learning_rate": 0.00019286529995722623, + "loss": 1.1001, + "step": 2735 + }, + { + "epoch": 0.48717948717948717, + "grad_norm": 0.49204322695732117, + "learning_rate": 0.00019286010669077763, + "loss": 0.9219, + "step": 2736 + }, + { + "epoch": 0.48735754985754987, + "grad_norm": 0.5853500962257385, + "learning_rate": 0.00019285491160492813, + "loss": 1.1133, + "step": 2737 + }, + { + "epoch": 0.48753561253561256, + "grad_norm": 0.5555846095085144, + "learning_rate": 0.0001928497146997795, + "loss": 1.0915, + "step": 2738 + }, + { + "epoch": 0.4877136752136752, + "grad_norm": 0.5166759490966797, + "learning_rate": 0.00019284451597543364, + "loss": 0.9349, + "step": 2739 + }, + { + "epoch": 0.4878917378917379, + "grad_norm": 0.47816506028175354, + "learning_rate": 0.00019283931543199234, + "loss": 0.8978, + "step": 2740 + }, + { + "epoch": 0.4880698005698006, + "grad_norm": 0.5632442831993103, + "learning_rate": 0.0001928341130695575, + "loss": 1.0491, + "step": 2741 + }, + { + "epoch": 0.48824786324786323, + "grad_norm": 0.6532769799232483, + "learning_rate": 0.00019282890888823107, + "loss": 1.2779, + "step": 2742 + }, + { + "epoch": 0.48842592592592593, + "grad_norm": 0.5733640789985657, + "learning_rate": 0.000192823702888115, + "loss": 1.4127, + "step": 2743 + }, + { + "epoch": 0.4886039886039886, + "grad_norm": 0.5701746344566345, + "learning_rate": 0.00019281849506931132, + "loss": 1.138, + "step": 2744 + }, + { + "epoch": 0.48878205128205127, + "grad_norm": 0.5227449536323547, + "learning_rate": 0.000192813285431922, + "loss": 1.1831, + "step": 2745 + }, + { + "epoch": 0.48896011396011396, + "grad_norm": 0.48457080125808716, + "learning_rate": 0.00019280807397604915, + "loss": 1.2468, + "step": 2746 + }, + { + "epoch": 0.48913817663817666, + "grad_norm": 0.4596176743507385, + "learning_rate": 0.0001928028607017949, + "loss": 1.1098, + "step": 2747 + }, + { + "epoch": 0.4893162393162393, + "grad_norm": 0.5204966068267822, + "learning_rate": 0.00019279764560926142, + "loss": 1.1501, + "step": 2748 + }, + { + "epoch": 0.489494301994302, + "grad_norm": 0.5179490447044373, + "learning_rate": 0.0001927924286985508, + "loss": 1.2601, + "step": 2749 + }, + { + "epoch": 0.4896723646723647, + "grad_norm": 0.4563423693180084, + "learning_rate": 0.00019278720996976533, + "loss": 1.081, + "step": 2750 + }, + { + "epoch": 0.48985042735042733, + "grad_norm": 0.4906339943408966, + "learning_rate": 0.00019278198942300717, + "loss": 1.157, + "step": 2751 + }, + { + "epoch": 0.49002849002849, + "grad_norm": 0.42241403460502625, + "learning_rate": 0.00019277676705837873, + "loss": 1.0333, + "step": 2752 + }, + { + "epoch": 0.4902065527065527, + "grad_norm": 0.6310175657272339, + "learning_rate": 0.00019277154287598226, + "loss": 1.1225, + "step": 2753 + }, + { + "epoch": 0.49038461538461536, + "grad_norm": 0.5109034776687622, + "learning_rate": 0.0001927663168759201, + "loss": 1.1619, + "step": 2754 + }, + { + "epoch": 0.49056267806267806, + "grad_norm": 0.4809598922729492, + "learning_rate": 0.00019276108905829465, + "loss": 1.0423, + "step": 2755 + }, + { + "epoch": 0.49074074074074076, + "grad_norm": 0.557502806186676, + "learning_rate": 0.00019275585942320837, + "loss": 0.8783, + "step": 2756 + }, + { + "epoch": 0.4909188034188034, + "grad_norm": 0.5434393882751465, + "learning_rate": 0.0001927506279707637, + "loss": 1.1701, + "step": 2757 + }, + { + "epoch": 0.4910968660968661, + "grad_norm": 0.49278944730758667, + "learning_rate": 0.00019274539470106317, + "loss": 1.0447, + "step": 2758 + }, + { + "epoch": 0.4912749287749288, + "grad_norm": 0.5634264349937439, + "learning_rate": 0.00019274015961420927, + "loss": 1.0639, + "step": 2759 + }, + { + "epoch": 0.49145299145299143, + "grad_norm": 0.5632645487785339, + "learning_rate": 0.00019273492271030464, + "loss": 0.9223, + "step": 2760 + }, + { + "epoch": 0.4916310541310541, + "grad_norm": 0.5949172377586365, + "learning_rate": 0.00019272968398945177, + "loss": 0.894, + "step": 2761 + }, + { + "epoch": 0.4918091168091168, + "grad_norm": 0.5375374555587769, + "learning_rate": 0.00019272444345175342, + "loss": 1.0311, + "step": 2762 + }, + { + "epoch": 0.49198717948717946, + "grad_norm": 0.5211305022239685, + "learning_rate": 0.00019271920109731222, + "loss": 1.1531, + "step": 2763 + }, + { + "epoch": 0.49216524216524216, + "grad_norm": 0.44022253155708313, + "learning_rate": 0.00019271395692623084, + "loss": 0.9147, + "step": 2764 + }, + { + "epoch": 0.49234330484330485, + "grad_norm": 0.4682174623012543, + "learning_rate": 0.0001927087109386121, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 0.49252136752136755, + "grad_norm": 0.4971517324447632, + "learning_rate": 0.0001927034631345588, + "loss": 1.1017, + "step": 2766 + }, + { + "epoch": 0.4926994301994302, + "grad_norm": 0.5015294551849365, + "learning_rate": 0.00019269821351417364, + "loss": 1.1093, + "step": 2767 + }, + { + "epoch": 0.4928774928774929, + "grad_norm": 0.5512694716453552, + "learning_rate": 0.00019269296207755958, + "loss": 0.9657, + "step": 2768 + }, + { + "epoch": 0.4930555555555556, + "grad_norm": 0.4914868474006653, + "learning_rate": 0.00019268770882481948, + "loss": 1.0379, + "step": 2769 + }, + { + "epoch": 0.4932336182336182, + "grad_norm": 0.567337691783905, + "learning_rate": 0.00019268245375605626, + "loss": 1.004, + "step": 2770 + }, + { + "epoch": 0.4934116809116809, + "grad_norm": 0.518489420413971, + "learning_rate": 0.0001926771968713729, + "loss": 1.0734, + "step": 2771 + }, + { + "epoch": 0.4935897435897436, + "grad_norm": 0.567742109298706, + "learning_rate": 0.00019267193817087237, + "loss": 1.1276, + "step": 2772 + }, + { + "epoch": 0.49376780626780625, + "grad_norm": 0.5287964344024658, + "learning_rate": 0.00019266667765465773, + "loss": 1.1429, + "step": 2773 + }, + { + "epoch": 0.49394586894586895, + "grad_norm": 0.5302085876464844, + "learning_rate": 0.00019266141532283207, + "loss": 1.0934, + "step": 2774 + }, + { + "epoch": 0.49412393162393164, + "grad_norm": 0.5569987297058105, + "learning_rate": 0.00019265615117549842, + "loss": 1.1453, + "step": 2775 + }, + { + "epoch": 0.4943019943019943, + "grad_norm": 0.519695520401001, + "learning_rate": 0.00019265088521275997, + "loss": 1.1255, + "step": 2776 + }, + { + "epoch": 0.494480056980057, + "grad_norm": 0.5073211193084717, + "learning_rate": 0.0001926456174347199, + "loss": 1.0609, + "step": 2777 + }, + { + "epoch": 0.4946581196581197, + "grad_norm": 0.45028239488601685, + "learning_rate": 0.00019264034784148142, + "loss": 0.9098, + "step": 2778 + }, + { + "epoch": 0.4948361823361823, + "grad_norm": 0.6641215682029724, + "learning_rate": 0.00019263507643314776, + "loss": 0.8903, + "step": 2779 + }, + { + "epoch": 0.495014245014245, + "grad_norm": 0.5281413793563843, + "learning_rate": 0.00019262980320982224, + "loss": 1.2906, + "step": 2780 + }, + { + "epoch": 0.4951923076923077, + "grad_norm": 0.6256437301635742, + "learning_rate": 0.0001926245281716081, + "loss": 1.4142, + "step": 2781 + }, + { + "epoch": 0.49537037037037035, + "grad_norm": 0.5422517657279968, + "learning_rate": 0.00019261925131860877, + "loss": 1.1606, + "step": 2782 + }, + { + "epoch": 0.49554843304843305, + "grad_norm": 0.46938949823379517, + "learning_rate": 0.0001926139726509276, + "loss": 1.0333, + "step": 2783 + }, + { + "epoch": 0.49572649572649574, + "grad_norm": 0.5799683928489685, + "learning_rate": 0.000192608692168668, + "loss": 1.0333, + "step": 2784 + }, + { + "epoch": 0.4959045584045584, + "grad_norm": 0.5231602787971497, + "learning_rate": 0.0001926034098719335, + "loss": 1.1847, + "step": 2785 + }, + { + "epoch": 0.4960826210826211, + "grad_norm": 0.477845698595047, + "learning_rate": 0.00019259812576082752, + "loss": 1.0746, + "step": 2786 + }, + { + "epoch": 0.4962606837606838, + "grad_norm": 0.5490350723266602, + "learning_rate": 0.00019259283983545365, + "loss": 1.2462, + "step": 2787 + }, + { + "epoch": 0.4964387464387464, + "grad_norm": 0.5788847208023071, + "learning_rate": 0.0001925875520959154, + "loss": 1.3485, + "step": 2788 + }, + { + "epoch": 0.4966168091168091, + "grad_norm": 0.46184736490249634, + "learning_rate": 0.00019258226254231643, + "loss": 0.8673, + "step": 2789 + }, + { + "epoch": 0.4967948717948718, + "grad_norm": 0.4890633225440979, + "learning_rate": 0.0001925769711747603, + "loss": 0.9474, + "step": 2790 + }, + { + "epoch": 0.49697293447293445, + "grad_norm": 0.5719282627105713, + "learning_rate": 0.00019257167799335078, + "loss": 1.2532, + "step": 2791 + }, + { + "epoch": 0.49715099715099714, + "grad_norm": 0.5385584235191345, + "learning_rate": 0.0001925663829981915, + "loss": 1.1326, + "step": 2792 + }, + { + "epoch": 0.49732905982905984, + "grad_norm": 0.5339545011520386, + "learning_rate": 0.00019256108618938625, + "loss": 1.1362, + "step": 2793 + }, + { + "epoch": 0.49750712250712253, + "grad_norm": 0.5017803907394409, + "learning_rate": 0.00019255578756703878, + "loss": 1.0449, + "step": 2794 + }, + { + "epoch": 0.4976851851851852, + "grad_norm": 0.6004226803779602, + "learning_rate": 0.00019255048713125294, + "loss": 0.9346, + "step": 2795 + }, + { + "epoch": 0.49786324786324787, + "grad_norm": 0.44581490755081177, + "learning_rate": 0.00019254518488213255, + "loss": 1.038, + "step": 2796 + }, + { + "epoch": 0.49804131054131057, + "grad_norm": 0.5180951356887817, + "learning_rate": 0.00019253988081978151, + "loss": 1.0479, + "step": 2797 + }, + { + "epoch": 0.4982193732193732, + "grad_norm": 0.53944993019104, + "learning_rate": 0.00019253457494430376, + "loss": 1.2598, + "step": 2798 + }, + { + "epoch": 0.4983974358974359, + "grad_norm": 0.5633010268211365, + "learning_rate": 0.00019252926725580322, + "loss": 1.205, + "step": 2799 + }, + { + "epoch": 0.4985754985754986, + "grad_norm": 0.6653175950050354, + "learning_rate": 0.0001925239577543839, + "loss": 1.2383, + "step": 2800 + }, + { + "epoch": 0.49875356125356124, + "grad_norm": 0.5083333849906921, + "learning_rate": 0.00019251864644014984, + "loss": 1.0649, + "step": 2801 + }, + { + "epoch": 0.49893162393162394, + "grad_norm": 0.4842020571231842, + "learning_rate": 0.00019251333331320506, + "loss": 1.1991, + "step": 2802 + }, + { + "epoch": 0.49910968660968663, + "grad_norm": 0.47987112402915955, + "learning_rate": 0.00019250801837365373, + "loss": 1.1686, + "step": 2803 + }, + { + "epoch": 0.49928774928774927, + "grad_norm": 0.5316333770751953, + "learning_rate": 0.00019250270162159992, + "loss": 1.1759, + "step": 2804 + }, + { + "epoch": 0.49946581196581197, + "grad_norm": 0.5015079379081726, + "learning_rate": 0.00019249738305714787, + "loss": 0.9424, + "step": 2805 + }, + { + "epoch": 0.49964387464387466, + "grad_norm": 0.6488274931907654, + "learning_rate": 0.00019249206268040172, + "loss": 1.066, + "step": 2806 + }, + { + "epoch": 0.4998219373219373, + "grad_norm": 0.40364864468574524, + "learning_rate": 0.00019248674049146574, + "loss": 0.6998, + "step": 2807 + }, + { + "epoch": 0.5, + "grad_norm": 0.5535672903060913, + "learning_rate": 0.00019248141649044423, + "loss": 1.2207, + "step": 2808 + }, + { + "epoch": 0.5, + "eval_loss": 1.1072274446487427, + "eval_runtime": 28.6913, + "eval_samples_per_second": 36.283, + "eval_steps_per_second": 18.159, + "step": 2808 + }, + { + "epoch": 0.5001780626780626, + "grad_norm": 0.4834389090538025, + "learning_rate": 0.00019247609067744143, + "loss": 1.1686, + "step": 2809 + }, + { + "epoch": 0.5003561253561254, + "grad_norm": 0.5007249712944031, + "learning_rate": 0.00019247076305256176, + "loss": 1.1343, + "step": 2810 + }, + { + "epoch": 0.500534188034188, + "grad_norm": 0.4773348271846771, + "learning_rate": 0.00019246543361590957, + "loss": 0.9324, + "step": 2811 + }, + { + "epoch": 0.5007122507122507, + "grad_norm": 0.47324609756469727, + "learning_rate": 0.0001924601023675893, + "loss": 1.0223, + "step": 2812 + }, + { + "epoch": 0.5008903133903134, + "grad_norm": 0.5583845973014832, + "learning_rate": 0.00019245476930770537, + "loss": 1.1328, + "step": 2813 + }, + { + "epoch": 0.5010683760683761, + "grad_norm": 0.4814579486846924, + "learning_rate": 0.00019244943443636232, + "loss": 1.0528, + "step": 2814 + }, + { + "epoch": 0.5012464387464387, + "grad_norm": 0.4996104836463928, + "learning_rate": 0.00019244409775366465, + "loss": 1.2482, + "step": 2815 + }, + { + "epoch": 0.5014245014245015, + "grad_norm": 0.47870904207229614, + "learning_rate": 0.0001924387592597169, + "loss": 0.9452, + "step": 2816 + }, + { + "epoch": 0.5016025641025641, + "grad_norm": 0.5617441534996033, + "learning_rate": 0.0001924334189546237, + "loss": 1.378, + "step": 2817 + }, + { + "epoch": 0.5017806267806267, + "grad_norm": 0.4872083365917206, + "learning_rate": 0.00019242807683848967, + "loss": 1.1571, + "step": 2818 + }, + { + "epoch": 0.5019586894586895, + "grad_norm": 0.5147804021835327, + "learning_rate": 0.00019242273291141947, + "loss": 1.1086, + "step": 2819 + }, + { + "epoch": 0.5021367521367521, + "grad_norm": 0.4698995351791382, + "learning_rate": 0.00019241738717351784, + "loss": 1.1579, + "step": 2820 + }, + { + "epoch": 0.5023148148148148, + "grad_norm": 0.5158926844596863, + "learning_rate": 0.00019241203962488946, + "loss": 1.2763, + "step": 2821 + }, + { + "epoch": 0.5024928774928775, + "grad_norm": 0.5218976736068726, + "learning_rate": 0.00019240669026563914, + "loss": 1.0633, + "step": 2822 + }, + { + "epoch": 0.5026709401709402, + "grad_norm": 0.5511452555656433, + "learning_rate": 0.0001924013390958717, + "loss": 0.9939, + "step": 2823 + }, + { + "epoch": 0.5028490028490028, + "grad_norm": 0.5227555632591248, + "learning_rate": 0.00019239598611569191, + "loss": 1.2478, + "step": 2824 + }, + { + "epoch": 0.5030270655270656, + "grad_norm": 0.5444719791412354, + "learning_rate": 0.00019239063132520475, + "loss": 1.1574, + "step": 2825 + }, + { + "epoch": 0.5032051282051282, + "grad_norm": 0.4752781093120575, + "learning_rate": 0.0001923852747245151, + "loss": 0.9034, + "step": 2826 + }, + { + "epoch": 0.5033831908831908, + "grad_norm": 0.5286496877670288, + "learning_rate": 0.00019237991631372792, + "loss": 1.1391, + "step": 2827 + }, + { + "epoch": 0.5035612535612536, + "grad_norm": 0.5009933710098267, + "learning_rate": 0.00019237455609294815, + "loss": 1.2178, + "step": 2828 + }, + { + "epoch": 0.5037393162393162, + "grad_norm": 0.5012276768684387, + "learning_rate": 0.00019236919406228085, + "loss": 0.9877, + "step": 2829 + }, + { + "epoch": 0.5039173789173789, + "grad_norm": 0.576508104801178, + "learning_rate": 0.00019236383022183106, + "loss": 1.1299, + "step": 2830 + }, + { + "epoch": 0.5040954415954416, + "grad_norm": 0.4716590642929077, + "learning_rate": 0.0001923584645717039, + "loss": 1.0451, + "step": 2831 + }, + { + "epoch": 0.5042735042735043, + "grad_norm": 0.5817418098449707, + "learning_rate": 0.00019235309711200448, + "loss": 1.0911, + "step": 2832 + }, + { + "epoch": 0.5044515669515669, + "grad_norm": 0.5695745944976807, + "learning_rate": 0.000192347727842838, + "loss": 1.0229, + "step": 2833 + }, + { + "epoch": 0.5046296296296297, + "grad_norm": 0.49127066135406494, + "learning_rate": 0.00019234235676430958, + "loss": 1.1377, + "step": 2834 + }, + { + "epoch": 0.5048076923076923, + "grad_norm": 0.5426172614097595, + "learning_rate": 0.00019233698387652453, + "loss": 1.2427, + "step": 2835 + }, + { + "epoch": 0.5049857549857549, + "grad_norm": 0.5342385172843933, + "learning_rate": 0.0001923316091795881, + "loss": 1.1427, + "step": 2836 + }, + { + "epoch": 0.5051638176638177, + "grad_norm": 0.5480486750602722, + "learning_rate": 0.00019232623267360558, + "loss": 1.0647, + "step": 2837 + }, + { + "epoch": 0.5053418803418803, + "grad_norm": 0.4584530293941498, + "learning_rate": 0.00019232085435868235, + "loss": 1.0461, + "step": 2838 + }, + { + "epoch": 0.5055199430199431, + "grad_norm": 0.5992119908332825, + "learning_rate": 0.00019231547423492371, + "loss": 1.1456, + "step": 2839 + }, + { + "epoch": 0.5056980056980057, + "grad_norm": 0.514018177986145, + "learning_rate": 0.00019231009230243515, + "loss": 1.2559, + "step": 2840 + }, + { + "epoch": 0.5058760683760684, + "grad_norm": 0.5392283797264099, + "learning_rate": 0.0001923047085613221, + "loss": 1.044, + "step": 2841 + }, + { + "epoch": 0.5060541310541311, + "grad_norm": 0.4486566483974457, + "learning_rate": 0.00019229932301169, + "loss": 1.0679, + "step": 2842 + }, + { + "epoch": 0.5062321937321937, + "grad_norm": 0.4523460566997528, + "learning_rate": 0.00019229393565364442, + "loss": 1.1651, + "step": 2843 + }, + { + "epoch": 0.5064102564102564, + "grad_norm": 0.6032688021659851, + "learning_rate": 0.0001922885464872909, + "loss": 1.15, + "step": 2844 + }, + { + "epoch": 0.5065883190883191, + "grad_norm": 0.5883688926696777, + "learning_rate": 0.000192283155512735, + "loss": 1.2179, + "step": 2845 + }, + { + "epoch": 0.5067663817663818, + "grad_norm": 0.5534378886222839, + "learning_rate": 0.00019227776273008238, + "loss": 1.0387, + "step": 2846 + }, + { + "epoch": 0.5069444444444444, + "grad_norm": 0.5899033546447754, + "learning_rate": 0.00019227236813943872, + "loss": 1.0812, + "step": 2847 + }, + { + "epoch": 0.5071225071225072, + "grad_norm": 0.5718855261802673, + "learning_rate": 0.00019226697174090965, + "loss": 1.1375, + "step": 2848 + }, + { + "epoch": 0.5073005698005698, + "grad_norm": 0.5080967545509338, + "learning_rate": 0.00019226157353460094, + "loss": 1.1421, + "step": 2849 + }, + { + "epoch": 0.5074786324786325, + "grad_norm": 0.5253677368164062, + "learning_rate": 0.0001922561735206184, + "loss": 1.0166, + "step": 2850 + }, + { + "epoch": 0.5076566951566952, + "grad_norm": 0.47797444462776184, + "learning_rate": 0.00019225077169906772, + "loss": 1.0504, + "step": 2851 + }, + { + "epoch": 0.5078347578347578, + "grad_norm": 0.4911690652370453, + "learning_rate": 0.0001922453680700548, + "loss": 1.0629, + "step": 2852 + }, + { + "epoch": 0.5080128205128205, + "grad_norm": 0.49678200483322144, + "learning_rate": 0.00019223996263368557, + "loss": 1.1672, + "step": 2853 + }, + { + "epoch": 0.5081908831908832, + "grad_norm": 0.5451810359954834, + "learning_rate": 0.00019223455539006586, + "loss": 1.3031, + "step": 2854 + }, + { + "epoch": 0.5083689458689459, + "grad_norm": 0.5708984136581421, + "learning_rate": 0.00019222914633930166, + "loss": 1.0986, + "step": 2855 + }, + { + "epoch": 0.5085470085470085, + "grad_norm": 0.47232356667518616, + "learning_rate": 0.00019222373548149888, + "loss": 1.0449, + "step": 2856 + }, + { + "epoch": 0.5087250712250713, + "grad_norm": 0.6027610898017883, + "learning_rate": 0.0001922183228167636, + "loss": 0.862, + "step": 2857 + }, + { + "epoch": 0.5089031339031339, + "grad_norm": 0.5211802124977112, + "learning_rate": 0.00019221290834520188, + "loss": 1.1048, + "step": 2858 + }, + { + "epoch": 0.5090811965811965, + "grad_norm": 0.45101237297058105, + "learning_rate": 0.00019220749206691972, + "loss": 1.0046, + "step": 2859 + }, + { + "epoch": 0.5092592592592593, + "grad_norm": 0.5526158213615417, + "learning_rate": 0.00019220207398202335, + "loss": 1.2275, + "step": 2860 + }, + { + "epoch": 0.5094373219373219, + "grad_norm": 0.48322010040283203, + "learning_rate": 0.00019219665409061885, + "loss": 0.9974, + "step": 2861 + }, + { + "epoch": 0.5096153846153846, + "grad_norm": 0.4775219261646271, + "learning_rate": 0.00019219123239281244, + "loss": 1.1852, + "step": 2862 + }, + { + "epoch": 0.5097934472934473, + "grad_norm": 0.46184200048446655, + "learning_rate": 0.00019218580888871034, + "loss": 0.9393, + "step": 2863 + }, + { + "epoch": 0.50997150997151, + "grad_norm": 0.47495174407958984, + "learning_rate": 0.00019218038357841883, + "loss": 0.9631, + "step": 2864 + }, + { + "epoch": 0.5101495726495726, + "grad_norm": 0.48600029945373535, + "learning_rate": 0.00019217495646204418, + "loss": 1.0498, + "step": 2865 + }, + { + "epoch": 0.5103276353276354, + "grad_norm": 0.5801547169685364, + "learning_rate": 0.00019216952753969274, + "loss": 1.2181, + "step": 2866 + }, + { + "epoch": 0.510505698005698, + "grad_norm": 0.5082106590270996, + "learning_rate": 0.00019216409681147085, + "loss": 1.2009, + "step": 2867 + }, + { + "epoch": 0.5106837606837606, + "grad_norm": 0.4184330701828003, + "learning_rate": 0.00019215866427748493, + "loss": 0.8462, + "step": 2868 + }, + { + "epoch": 0.5108618233618234, + "grad_norm": 0.518099844455719, + "learning_rate": 0.00019215322993784147, + "loss": 1.2091, + "step": 2869 + }, + { + "epoch": 0.511039886039886, + "grad_norm": 0.569464921951294, + "learning_rate": 0.0001921477937926469, + "loss": 1.0264, + "step": 2870 + }, + { + "epoch": 0.5112179487179487, + "grad_norm": 0.526767909526825, + "learning_rate": 0.00019214235584200768, + "loss": 1.1192, + "step": 2871 + }, + { + "epoch": 0.5113960113960114, + "grad_norm": 0.6511057019233704, + "learning_rate": 0.00019213691608603047, + "loss": 1.3193, + "step": 2872 + }, + { + "epoch": 0.5115740740740741, + "grad_norm": 0.48536401987075806, + "learning_rate": 0.00019213147452482173, + "loss": 1.1671, + "step": 2873 + }, + { + "epoch": 0.5117521367521367, + "grad_norm": 0.7972469329833984, + "learning_rate": 0.00019212603115848818, + "loss": 1.1393, + "step": 2874 + }, + { + "epoch": 0.5119301994301995, + "grad_norm": 0.5543264746665955, + "learning_rate": 0.00019212058598713642, + "loss": 1.1436, + "step": 2875 + }, + { + "epoch": 0.5121082621082621, + "grad_norm": 0.49688720703125, + "learning_rate": 0.0001921151390108731, + "loss": 1.0897, + "step": 2876 + }, + { + "epoch": 0.5122863247863247, + "grad_norm": 0.4928736090660095, + "learning_rate": 0.000192109690229805, + "loss": 1.2426, + "step": 2877 + }, + { + "epoch": 0.5124643874643875, + "grad_norm": 0.4917896091938019, + "learning_rate": 0.0001921042396440389, + "loss": 1.0047, + "step": 2878 + }, + { + "epoch": 0.5126424501424501, + "grad_norm": 0.5485204458236694, + "learning_rate": 0.00019209878725368152, + "loss": 1.2615, + "step": 2879 + }, + { + "epoch": 0.5128205128205128, + "grad_norm": 0.5229470133781433, + "learning_rate": 0.0001920933330588397, + "loss": 1.3249, + "step": 2880 + }, + { + "epoch": 0.5129985754985755, + "grad_norm": 0.4783077538013458, + "learning_rate": 0.00019208787705962037, + "loss": 1.2004, + "step": 2881 + }, + { + "epoch": 0.5131766381766382, + "grad_norm": 0.5106910467147827, + "learning_rate": 0.00019208241925613035, + "loss": 1.1745, + "step": 2882 + }, + { + "epoch": 0.5133547008547008, + "grad_norm": 0.5308730006217957, + "learning_rate": 0.00019207695964847666, + "loss": 0.9706, + "step": 2883 + }, + { + "epoch": 0.5135327635327636, + "grad_norm": 0.5489775538444519, + "learning_rate": 0.00019207149823676617, + "loss": 1.0073, + "step": 2884 + }, + { + "epoch": 0.5137108262108262, + "grad_norm": 0.4992835521697998, + "learning_rate": 0.00019206603502110596, + "loss": 1.1053, + "step": 2885 + }, + { + "epoch": 0.5138888888888888, + "grad_norm": 0.5304922461509705, + "learning_rate": 0.00019206057000160302, + "loss": 1.0565, + "step": 2886 + }, + { + "epoch": 0.5140669515669516, + "grad_norm": 0.46411609649658203, + "learning_rate": 0.00019205510317836448, + "loss": 0.9202, + "step": 2887 + }, + { + "epoch": 0.5142450142450142, + "grad_norm": 0.5236835479736328, + "learning_rate": 0.0001920496345514974, + "loss": 0.9075, + "step": 2888 + }, + { + "epoch": 0.5144230769230769, + "grad_norm": 0.4416964054107666, + "learning_rate": 0.00019204416412110895, + "loss": 0.9225, + "step": 2889 + }, + { + "epoch": 0.5146011396011396, + "grad_norm": 0.5470940470695496, + "learning_rate": 0.00019203869188730633, + "loss": 1.2195, + "step": 2890 + }, + { + "epoch": 0.5147792022792023, + "grad_norm": 0.5380414128303528, + "learning_rate": 0.0001920332178501967, + "loss": 1.0731, + "step": 2891 + }, + { + "epoch": 0.5149572649572649, + "grad_norm": 0.4405716359615326, + "learning_rate": 0.00019202774200988737, + "loss": 0.8739, + "step": 2892 + }, + { + "epoch": 0.5151353276353277, + "grad_norm": 0.5222984552383423, + "learning_rate": 0.0001920222643664856, + "loss": 1.1806, + "step": 2893 + }, + { + "epoch": 0.5153133903133903, + "grad_norm": 0.48545539379119873, + "learning_rate": 0.0001920167849200987, + "loss": 0.9939, + "step": 2894 + }, + { + "epoch": 0.5154914529914529, + "grad_norm": 0.45078009366989136, + "learning_rate": 0.0001920113036708341, + "loss": 1.0085, + "step": 2895 + }, + { + "epoch": 0.5156695156695157, + "grad_norm": 0.5029830932617188, + "learning_rate": 0.00019200582061879913, + "loss": 1.1095, + "step": 2896 + }, + { + "epoch": 0.5158475783475783, + "grad_norm": 0.5316143035888672, + "learning_rate": 0.00019200033576410118, + "loss": 0.9883, + "step": 2897 + }, + { + "epoch": 0.5160256410256411, + "grad_norm": 0.5282100439071655, + "learning_rate": 0.0001919948491068478, + "loss": 1.1441, + "step": 2898 + }, + { + "epoch": 0.5162037037037037, + "grad_norm": 0.5145367980003357, + "learning_rate": 0.00019198936064714647, + "loss": 1.1999, + "step": 2899 + }, + { + "epoch": 0.5163817663817664, + "grad_norm": 0.5385651588439941, + "learning_rate": 0.00019198387038510468, + "loss": 1.1831, + "step": 2900 + }, + { + "epoch": 0.5165598290598291, + "grad_norm": 0.4971916377544403, + "learning_rate": 0.00019197837832083002, + "loss": 1.2518, + "step": 2901 + }, + { + "epoch": 0.5167378917378918, + "grad_norm": 0.5253807306289673, + "learning_rate": 0.00019197288445443016, + "loss": 1.0788, + "step": 2902 + }, + { + "epoch": 0.5169159544159544, + "grad_norm": 0.49724945425987244, + "learning_rate": 0.00019196738878601263, + "loss": 1.0985, + "step": 2903 + }, + { + "epoch": 0.5170940170940171, + "grad_norm": 0.5327325463294983, + "learning_rate": 0.0001919618913156852, + "loss": 1.2862, + "step": 2904 + }, + { + "epoch": 0.5172720797720798, + "grad_norm": 0.639999270439148, + "learning_rate": 0.00019195639204355554, + "loss": 1.2052, + "step": 2905 + }, + { + "epoch": 0.5174501424501424, + "grad_norm": 0.4630785584449768, + "learning_rate": 0.0001919508909697314, + "loss": 1.1157, + "step": 2906 + }, + { + "epoch": 0.5176282051282052, + "grad_norm": 0.513949990272522, + "learning_rate": 0.00019194538809432055, + "loss": 1.0047, + "step": 2907 + }, + { + "epoch": 0.5178062678062678, + "grad_norm": 0.488034725189209, + "learning_rate": 0.0001919398834174308, + "loss": 0.9008, + "step": 2908 + }, + { + "epoch": 0.5179843304843305, + "grad_norm": 0.4892788529396057, + "learning_rate": 0.00019193437693917006, + "loss": 1.1024, + "step": 2909 + }, + { + "epoch": 0.5181623931623932, + "grad_norm": 0.5503842830657959, + "learning_rate": 0.00019192886865964618, + "loss": 1.2283, + "step": 2910 + }, + { + "epoch": 0.5183404558404558, + "grad_norm": 0.48885393142700195, + "learning_rate": 0.00019192335857896707, + "loss": 0.9522, + "step": 2911 + }, + { + "epoch": 0.5185185185185185, + "grad_norm": 0.5479527115821838, + "learning_rate": 0.00019191784669724072, + "loss": 1.1616, + "step": 2912 + }, + { + "epoch": 0.5186965811965812, + "grad_norm": 0.42701148986816406, + "learning_rate": 0.00019191233301457506, + "loss": 0.8434, + "step": 2913 + }, + { + "epoch": 0.5188746438746439, + "grad_norm": 0.4273422658443451, + "learning_rate": 0.00019190681753107822, + "loss": 0.8316, + "step": 2914 + }, + { + "epoch": 0.5190527065527065, + "grad_norm": 0.5047736763954163, + "learning_rate": 0.00019190130024685818, + "loss": 1.171, + "step": 2915 + }, + { + "epoch": 0.5192307692307693, + "grad_norm": 0.5221177935600281, + "learning_rate": 0.00019189578116202307, + "loss": 1.0256, + "step": 2916 + }, + { + "epoch": 0.5194088319088319, + "grad_norm": 0.4782322943210602, + "learning_rate": 0.00019189026027668105, + "loss": 0.8598, + "step": 2917 + }, + { + "epoch": 0.5195868945868946, + "grad_norm": 0.5627185702323914, + "learning_rate": 0.00019188473759094022, + "loss": 1.1825, + "step": 2918 + }, + { + "epoch": 0.5197649572649573, + "grad_norm": 0.5036423206329346, + "learning_rate": 0.00019187921310490888, + "loss": 1.0881, + "step": 2919 + }, + { + "epoch": 0.51994301994302, + "grad_norm": 0.4271143972873688, + "learning_rate": 0.0001918736868186952, + "loss": 0.9265, + "step": 2920 + }, + { + "epoch": 0.5201210826210826, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00019186815873240747, + "loss": 1.196, + "step": 2921 + }, + { + "epoch": 0.5202991452991453, + "grad_norm": 0.5494198203086853, + "learning_rate": 0.00019186262884615402, + "loss": 1.1207, + "step": 2922 + }, + { + "epoch": 0.520477207977208, + "grad_norm": 0.5305119752883911, + "learning_rate": 0.0001918570971600432, + "loss": 1.0393, + "step": 2923 + }, + { + "epoch": 0.5206552706552706, + "grad_norm": 0.46713170409202576, + "learning_rate": 0.00019185156367418333, + "loss": 0.9583, + "step": 2924 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 0.597776472568512, + "learning_rate": 0.00019184602838868292, + "loss": 1.2978, + "step": 2925 + }, + { + "epoch": 0.521011396011396, + "grad_norm": 0.520976722240448, + "learning_rate": 0.00019184049130365036, + "loss": 1.0515, + "step": 2926 + }, + { + "epoch": 0.5211894586894587, + "grad_norm": 0.5266290307044983, + "learning_rate": 0.00019183495241919415, + "loss": 1.0437, + "step": 2927 + }, + { + "epoch": 0.5213675213675214, + "grad_norm": 0.50911545753479, + "learning_rate": 0.00019182941173542285, + "loss": 0.9977, + "step": 2928 + }, + { + "epoch": 0.521545584045584, + "grad_norm": 0.4924670457839966, + "learning_rate": 0.00019182386925244496, + "loss": 0.9309, + "step": 2929 + }, + { + "epoch": 0.5217236467236467, + "grad_norm": 0.4979301393032074, + "learning_rate": 0.00019181832497036912, + "loss": 0.87, + "step": 2930 + }, + { + "epoch": 0.5219017094017094, + "grad_norm": 0.6307916045188904, + "learning_rate": 0.0001918127788893039, + "loss": 1.2159, + "step": 2931 + }, + { + "epoch": 0.5220797720797721, + "grad_norm": 0.4915660619735718, + "learning_rate": 0.00019180723100935802, + "loss": 1.0828, + "step": 2932 + }, + { + "epoch": 0.5222578347578347, + "grad_norm": 0.4312742352485657, + "learning_rate": 0.00019180168133064017, + "loss": 1.0496, + "step": 2933 + }, + { + "epoch": 0.5224358974358975, + "grad_norm": 0.6006124019622803, + "learning_rate": 0.00019179612985325908, + "loss": 1.0751, + "step": 2934 + }, + { + "epoch": 0.5226139601139601, + "grad_norm": 0.5332220196723938, + "learning_rate": 0.0001917905765773235, + "loss": 1.2601, + "step": 2935 + }, + { + "epoch": 0.5227920227920227, + "grad_norm": 0.4877954423427582, + "learning_rate": 0.00019178502150294223, + "loss": 1.2279, + "step": 2936 + }, + { + "epoch": 0.5229700854700855, + "grad_norm": 0.5975968837738037, + "learning_rate": 0.00019177946463022418, + "loss": 1.3371, + "step": 2937 + }, + { + "epoch": 0.5231481481481481, + "grad_norm": 0.5363923907279968, + "learning_rate": 0.00019177390595927815, + "loss": 1.0705, + "step": 2938 + }, + { + "epoch": 0.5233262108262108, + "grad_norm": 0.4314909875392914, + "learning_rate": 0.0001917683454902131, + "loss": 0.9172, + "step": 2939 + }, + { + "epoch": 0.5235042735042735, + "grad_norm": 0.46187883615493774, + "learning_rate": 0.0001917627832231379, + "loss": 1.1201, + "step": 2940 + }, + { + "epoch": 0.5236823361823362, + "grad_norm": 0.4648260772228241, + "learning_rate": 0.00019175721915816162, + "loss": 1.1307, + "step": 2941 + }, + { + "epoch": 0.5238603988603988, + "grad_norm": 0.4427165687084198, + "learning_rate": 0.00019175165329539325, + "loss": 0.9459, + "step": 2942 + }, + { + "epoch": 0.5240384615384616, + "grad_norm": 0.4645056128501892, + "learning_rate": 0.0001917460856349418, + "loss": 0.9176, + "step": 2943 + }, + { + "epoch": 0.5242165242165242, + "grad_norm": 0.4939568042755127, + "learning_rate": 0.0001917405161769164, + "loss": 1.1056, + "step": 2944 + }, + { + "epoch": 0.5243945868945868, + "grad_norm": 0.6057310104370117, + "learning_rate": 0.00019173494492142617, + "loss": 1.2714, + "step": 2945 + }, + { + "epoch": 0.5245726495726496, + "grad_norm": 0.5038546323776245, + "learning_rate": 0.00019172937186858025, + "loss": 0.911, + "step": 2946 + }, + { + "epoch": 0.5247507122507122, + "grad_norm": 0.5521321296691895, + "learning_rate": 0.00019172379701848784, + "loss": 1.0781, + "step": 2947 + }, + { + "epoch": 0.5249287749287749, + "grad_norm": 0.516979455947876, + "learning_rate": 0.00019171822037125817, + "loss": 1.1051, + "step": 2948 + }, + { + "epoch": 0.5251068376068376, + "grad_norm": 0.5443150997161865, + "learning_rate": 0.0001917126419270005, + "loss": 1.0802, + "step": 2949 + }, + { + "epoch": 0.5252849002849003, + "grad_norm": 0.5373311042785645, + "learning_rate": 0.00019170706168582412, + "loss": 0.9313, + "step": 2950 + }, + { + "epoch": 0.5254629629629629, + "grad_norm": 0.7511917948722839, + "learning_rate": 0.0001917014796478384, + "loss": 1.1958, + "step": 2951 + }, + { + "epoch": 0.5256410256410257, + "grad_norm": 0.49893468618392944, + "learning_rate": 0.00019169589581315263, + "loss": 0.9387, + "step": 2952 + }, + { + "epoch": 0.5258190883190883, + "grad_norm": 0.48010289669036865, + "learning_rate": 0.00019169031018187628, + "loss": 1.2459, + "step": 2953 + }, + { + "epoch": 0.5259971509971509, + "grad_norm": 0.48768678307533264, + "learning_rate": 0.0001916847227541188, + "loss": 1.0127, + "step": 2954 + }, + { + "epoch": 0.5261752136752137, + "grad_norm": 0.5973068475723267, + "learning_rate": 0.00019167913352998963, + "loss": 1.1685, + "step": 2955 + }, + { + "epoch": 0.5263532763532763, + "grad_norm": 0.5567806959152222, + "learning_rate": 0.00019167354250959826, + "loss": 1.142, + "step": 2956 + }, + { + "epoch": 0.5265313390313391, + "grad_norm": 0.47819700837135315, + "learning_rate": 0.00019166794969305428, + "loss": 0.712, + "step": 2957 + }, + { + "epoch": 0.5267094017094017, + "grad_norm": 0.5191744565963745, + "learning_rate": 0.00019166235508046725, + "loss": 1.2208, + "step": 2958 + }, + { + "epoch": 0.5268874643874644, + "grad_norm": 0.4987856149673462, + "learning_rate": 0.00019165675867194675, + "loss": 1.0466, + "step": 2959 + }, + { + "epoch": 0.5270655270655271, + "grad_norm": 0.5017665028572083, + "learning_rate": 0.0001916511604676025, + "loss": 1.1236, + "step": 2960 + }, + { + "epoch": 0.5272435897435898, + "grad_norm": 0.5115348696708679, + "learning_rate": 0.00019164556046754415, + "loss": 1.1497, + "step": 2961 + }, + { + "epoch": 0.5274216524216524, + "grad_norm": 0.4934345781803131, + "learning_rate": 0.0001916399586718814, + "loss": 1.0183, + "step": 2962 + }, + { + "epoch": 0.5275997150997151, + "grad_norm": 0.5033719539642334, + "learning_rate": 0.00019163435508072404, + "loss": 1.0256, + "step": 2963 + }, + { + "epoch": 0.5277777777777778, + "grad_norm": 0.5325372219085693, + "learning_rate": 0.00019162874969418184, + "loss": 1.1384, + "step": 2964 + }, + { + "epoch": 0.5279558404558404, + "grad_norm": 0.4901772141456604, + "learning_rate": 0.00019162314251236465, + "loss": 1.0831, + "step": 2965 + }, + { + "epoch": 0.5281339031339032, + "grad_norm": 0.4743805229663849, + "learning_rate": 0.0001916175335353823, + "loss": 1.1894, + "step": 2966 + }, + { + "epoch": 0.5283119658119658, + "grad_norm": 0.5439450740814209, + "learning_rate": 0.00019161192276334466, + "loss": 1.2066, + "step": 2967 + }, + { + "epoch": 0.5284900284900285, + "grad_norm": 0.5123090744018555, + "learning_rate": 0.00019160631019636174, + "loss": 1.1829, + "step": 2968 + }, + { + "epoch": 0.5286680911680912, + "grad_norm": 0.5995343923568726, + "learning_rate": 0.00019160069583454346, + "loss": 1.4872, + "step": 2969 + }, + { + "epoch": 0.5288461538461539, + "grad_norm": 0.4596657156944275, + "learning_rate": 0.00019159507967799985, + "loss": 0.8948, + "step": 2970 + }, + { + "epoch": 0.5290242165242165, + "grad_norm": 0.5533682107925415, + "learning_rate": 0.0001915894617268409, + "loss": 1.1779, + "step": 2971 + }, + { + "epoch": 0.5292022792022792, + "grad_norm": 0.3860718309879303, + "learning_rate": 0.00019158384198117673, + "loss": 0.6424, + "step": 2972 + }, + { + "epoch": 0.5293803418803419, + "grad_norm": 0.47424063086509705, + "learning_rate": 0.0001915782204411174, + "loss": 1.1592, + "step": 2973 + }, + { + "epoch": 0.5295584045584045, + "grad_norm": 0.5050228834152222, + "learning_rate": 0.00019157259710677309, + "loss": 1.1971, + "step": 2974 + }, + { + "epoch": 0.5297364672364673, + "grad_norm": 0.6080113649368286, + "learning_rate": 0.00019156697197825396, + "loss": 1.1511, + "step": 2975 + }, + { + "epoch": 0.5299145299145299, + "grad_norm": 0.4805932641029358, + "learning_rate": 0.00019156134505567024, + "loss": 1.1033, + "step": 2976 + }, + { + "epoch": 0.5300925925925926, + "grad_norm": 0.4835345447063446, + "learning_rate": 0.00019155571633913215, + "loss": 1.1832, + "step": 2977 + }, + { + "epoch": 0.5302706552706553, + "grad_norm": 0.5183725953102112, + "learning_rate": 0.00019155008582875, + "loss": 0.9221, + "step": 2978 + }, + { + "epoch": 0.530448717948718, + "grad_norm": 0.48015761375427246, + "learning_rate": 0.00019154445352463412, + "loss": 1.045, + "step": 2979 + }, + { + "epoch": 0.5306267806267806, + "grad_norm": 0.4670043885707855, + "learning_rate": 0.0001915388194268948, + "loss": 0.9025, + "step": 2980 + }, + { + "epoch": 0.5308048433048433, + "grad_norm": 0.5048824548721313, + "learning_rate": 0.0001915331835356425, + "loss": 1.0681, + "step": 2981 + }, + { + "epoch": 0.530982905982906, + "grad_norm": 0.4785633981227875, + "learning_rate": 0.00019152754585098758, + "loss": 1.0097, + "step": 2982 + }, + { + "epoch": 0.5311609686609686, + "grad_norm": 0.4829573333263397, + "learning_rate": 0.00019152190637304056, + "loss": 1.0856, + "step": 2983 + }, + { + "epoch": 0.5313390313390314, + "grad_norm": 0.5425563454627991, + "learning_rate": 0.00019151626510191189, + "loss": 1.2313, + "step": 2984 + }, + { + "epoch": 0.531517094017094, + "grad_norm": 0.5532251596450806, + "learning_rate": 0.0001915106220377121, + "loss": 1.0328, + "step": 2985 + }, + { + "epoch": 0.5316951566951567, + "grad_norm": 0.47016972303390503, + "learning_rate": 0.0001915049771805518, + "loss": 1.2003, + "step": 2986 + }, + { + "epoch": 0.5318732193732194, + "grad_norm": 0.5241743326187134, + "learning_rate": 0.00019149933053054153, + "loss": 1.046, + "step": 2987 + }, + { + "epoch": 0.532051282051282, + "grad_norm": 0.5043526887893677, + "learning_rate": 0.00019149368208779197, + "loss": 1.0022, + "step": 2988 + }, + { + "epoch": 0.5322293447293447, + "grad_norm": 0.5563312768936157, + "learning_rate": 0.00019148803185241374, + "loss": 1.1017, + "step": 2989 + }, + { + "epoch": 0.5324074074074074, + "grad_norm": 0.5414231419563293, + "learning_rate": 0.00019148237982451763, + "loss": 0.9649, + "step": 2990 + }, + { + "epoch": 0.5325854700854701, + "grad_norm": 0.5452231764793396, + "learning_rate": 0.0001914767260042143, + "loss": 1.2281, + "step": 2991 + }, + { + "epoch": 0.5327635327635327, + "grad_norm": 0.5500698685646057, + "learning_rate": 0.00019147107039161454, + "loss": 1.2865, + "step": 2992 + }, + { + "epoch": 0.5329415954415955, + "grad_norm": 0.49747416377067566, + "learning_rate": 0.00019146541298682918, + "loss": 1.1296, + "step": 2993 + }, + { + "epoch": 0.5331196581196581, + "grad_norm": 0.5684167742729187, + "learning_rate": 0.00019145975378996903, + "loss": 1.0685, + "step": 2994 + }, + { + "epoch": 0.5332977207977208, + "grad_norm": 0.5411235690116882, + "learning_rate": 0.00019145409280114502, + "loss": 1.1372, + "step": 2995 + }, + { + "epoch": 0.5334757834757835, + "grad_norm": 0.5006675720214844, + "learning_rate": 0.00019144843002046806, + "loss": 1.0688, + "step": 2996 + }, + { + "epoch": 0.5336538461538461, + "grad_norm": 0.4591315686702728, + "learning_rate": 0.00019144276544804908, + "loss": 1.1071, + "step": 2997 + }, + { + "epoch": 0.5338319088319088, + "grad_norm": 0.5615306496620178, + "learning_rate": 0.000191437099083999, + "loss": 1.1033, + "step": 2998 + }, + { + "epoch": 0.5340099715099715, + "grad_norm": 0.4986817240715027, + "learning_rate": 0.00019143143092842897, + "loss": 1.176, + "step": 2999 + }, + { + "epoch": 0.5341880341880342, + "grad_norm": 0.5017120242118835, + "learning_rate": 0.00019142576098144995, + "loss": 1.0174, + "step": 3000 + }, + { + "epoch": 0.5343660968660968, + "grad_norm": 0.508298397064209, + "learning_rate": 0.0001914200892431731, + "loss": 1.164, + "step": 3001 + }, + { + "epoch": 0.5345441595441596, + "grad_norm": 0.48068809509277344, + "learning_rate": 0.0001914144157137095, + "loss": 0.7959, + "step": 3002 + }, + { + "epoch": 0.5347222222222222, + "grad_norm": 0.6347028017044067, + "learning_rate": 0.0001914087403931703, + "loss": 1.1727, + "step": 3003 + }, + { + "epoch": 0.5349002849002849, + "grad_norm": 0.5558401942253113, + "learning_rate": 0.00019140306328166676, + "loss": 1.2282, + "step": 3004 + }, + { + "epoch": 0.5350783475783476, + "grad_norm": 0.5093596577644348, + "learning_rate": 0.00019139738437931004, + "loss": 1.3258, + "step": 3005 + }, + { + "epoch": 0.5352564102564102, + "grad_norm": 0.4653106927871704, + "learning_rate": 0.0001913917036862114, + "loss": 1.1062, + "step": 3006 + }, + { + "epoch": 0.5354344729344729, + "grad_norm": 0.48085781931877136, + "learning_rate": 0.00019138602120248222, + "loss": 0.9019, + "step": 3007 + }, + { + "epoch": 0.5356125356125356, + "grad_norm": 0.5174745321273804, + "learning_rate": 0.0001913803369282338, + "loss": 1.044, + "step": 3008 + }, + { + "epoch": 0.5357905982905983, + "grad_norm": 0.5359669327735901, + "learning_rate": 0.00019137465086357746, + "loss": 1.0723, + "step": 3009 + }, + { + "epoch": 0.5359686609686609, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.00019136896300862467, + "loss": 1.2192, + "step": 3010 + }, + { + "epoch": 0.5361467236467237, + "grad_norm": 0.4905693829059601, + "learning_rate": 0.00019136327336348688, + "loss": 1.2372, + "step": 3011 + }, + { + "epoch": 0.5363247863247863, + "grad_norm": 0.5741264820098877, + "learning_rate": 0.0001913575819282755, + "loss": 1.1703, + "step": 3012 + }, + { + "epoch": 0.5365028490028491, + "grad_norm": 0.577033281326294, + "learning_rate": 0.0001913518887031021, + "loss": 1.1555, + "step": 3013 + }, + { + "epoch": 0.5366809116809117, + "grad_norm": 0.46795153617858887, + "learning_rate": 0.00019134619368807822, + "loss": 0.8583, + "step": 3014 + }, + { + "epoch": 0.5368589743589743, + "grad_norm": 0.5973345637321472, + "learning_rate": 0.0001913404968833154, + "loss": 1.1509, + "step": 3015 + }, + { + "epoch": 0.5370370370370371, + "grad_norm": 0.62020343542099, + "learning_rate": 0.00019133479828892531, + "loss": 1.0781, + "step": 3016 + }, + { + "epoch": 0.5372150997150997, + "grad_norm": 0.5342286229133606, + "learning_rate": 0.00019132909790501958, + "loss": 1.1556, + "step": 3017 + }, + { + "epoch": 0.5373931623931624, + "grad_norm": 0.49612846970558167, + "learning_rate": 0.0001913233957317099, + "loss": 0.9027, + "step": 3018 + }, + { + "epoch": 0.5375712250712251, + "grad_norm": 0.5403908491134644, + "learning_rate": 0.00019131769176910796, + "loss": 1.1125, + "step": 3019 + }, + { + "epoch": 0.5377492877492878, + "grad_norm": 0.4952050447463989, + "learning_rate": 0.0001913119860173256, + "loss": 1.2329, + "step": 3020 + }, + { + "epoch": 0.5379273504273504, + "grad_norm": 0.5877819657325745, + "learning_rate": 0.0001913062784764745, + "loss": 1.2855, + "step": 3021 + }, + { + "epoch": 0.5381054131054132, + "grad_norm": 0.49312907457351685, + "learning_rate": 0.00019130056914666655, + "loss": 1.0212, + "step": 3022 + }, + { + "epoch": 0.5382834757834758, + "grad_norm": 0.45544490218162537, + "learning_rate": 0.00019129485802801366, + "loss": 0.9748, + "step": 3023 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.5535242557525635, + "learning_rate": 0.00019128914512062762, + "loss": 1.2134, + "step": 3024 + }, + { + "epoch": 0.5386396011396012, + "grad_norm": 0.45369696617126465, + "learning_rate": 0.00019128343042462044, + "loss": 0.9964, + "step": 3025 + }, + { + "epoch": 0.5388176638176638, + "grad_norm": 0.6240725517272949, + "learning_rate": 0.00019127771394010406, + "loss": 1.425, + "step": 3026 + }, + { + "epoch": 0.5389957264957265, + "grad_norm": 0.4859573245048523, + "learning_rate": 0.0001912719956671905, + "loss": 1.087, + "step": 3027 + }, + { + "epoch": 0.5391737891737892, + "grad_norm": 0.47529762983322144, + "learning_rate": 0.0001912662756059918, + "loss": 0.9517, + "step": 3028 + }, + { + "epoch": 0.5393518518518519, + "grad_norm": 0.5317288637161255, + "learning_rate": 0.00019126055375661997, + "loss": 1.0945, + "step": 3029 + }, + { + "epoch": 0.5395299145299145, + "grad_norm": 0.55974280834198, + "learning_rate": 0.00019125483011918722, + "loss": 1.0794, + "step": 3030 + }, + { + "epoch": 0.5397079772079773, + "grad_norm": 0.48579123616218567, + "learning_rate": 0.0001912491046938056, + "loss": 1.1421, + "step": 3031 + }, + { + "epoch": 0.5398860398860399, + "grad_norm": 0.4917181134223938, + "learning_rate": 0.00019124337748058733, + "loss": 0.9708, + "step": 3032 + }, + { + "epoch": 0.5400641025641025, + "grad_norm": 0.525291383266449, + "learning_rate": 0.00019123764847964466, + "loss": 1.064, + "step": 3033 + }, + { + "epoch": 0.5402421652421653, + "grad_norm": 0.5733301639556885, + "learning_rate": 0.00019123191769108977, + "loss": 1.2142, + "step": 3034 + }, + { + "epoch": 0.5404202279202279, + "grad_norm": 0.5400987863540649, + "learning_rate": 0.00019122618511503494, + "loss": 1.1309, + "step": 3035 + }, + { + "epoch": 0.5405982905982906, + "grad_norm": 0.6261051893234253, + "learning_rate": 0.00019122045075159257, + "loss": 1.2112, + "step": 3036 + }, + { + "epoch": 0.5407763532763533, + "grad_norm": 0.5483576059341431, + "learning_rate": 0.0001912147146008749, + "loss": 1.2705, + "step": 3037 + }, + { + "epoch": 0.540954415954416, + "grad_norm": 0.5442137122154236, + "learning_rate": 0.00019120897666299443, + "loss": 1.2512, + "step": 3038 + }, + { + "epoch": 0.5411324786324786, + "grad_norm": 0.5680811405181885, + "learning_rate": 0.00019120323693806355, + "loss": 1.392, + "step": 3039 + }, + { + "epoch": 0.5413105413105413, + "grad_norm": 0.5237287878990173, + "learning_rate": 0.00019119749542619466, + "loss": 1.1599, + "step": 3040 + }, + { + "epoch": 0.541488603988604, + "grad_norm": 0.48119300603866577, + "learning_rate": 0.00019119175212750032, + "loss": 1.0976, + "step": 3041 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 0.507033109664917, + "learning_rate": 0.00019118600704209302, + "loss": 1.0181, + "step": 3042 + }, + { + "epoch": 0.5418447293447294, + "grad_norm": 0.484672874212265, + "learning_rate": 0.00019118026017008531, + "loss": 1.1636, + "step": 3043 + }, + { + "epoch": 0.542022792022792, + "grad_norm": 0.4923502206802368, + "learning_rate": 0.00019117451151158985, + "loss": 1.0388, + "step": 3044 + }, + { + "epoch": 0.5422008547008547, + "grad_norm": 0.4882057309150696, + "learning_rate": 0.00019116876106671922, + "loss": 1.131, + "step": 3045 + }, + { + "epoch": 0.5423789173789174, + "grad_norm": 0.6068355441093445, + "learning_rate": 0.0001911630088355861, + "loss": 1.3218, + "step": 3046 + }, + { + "epoch": 0.54255698005698, + "grad_norm": 0.5012881755828857, + "learning_rate": 0.0001911572548183032, + "loss": 1.0514, + "step": 3047 + }, + { + "epoch": 0.5427350427350427, + "grad_norm": 0.49849793314933777, + "learning_rate": 0.00019115149901498328, + "loss": 1.0003, + "step": 3048 + }, + { + "epoch": 0.5429131054131054, + "grad_norm": 0.4934251010417938, + "learning_rate": 0.00019114574142573904, + "loss": 1.0319, + "step": 3049 + }, + { + "epoch": 0.5430911680911681, + "grad_norm": 0.4947762191295624, + "learning_rate": 0.00019113998205068334, + "loss": 1.0906, + "step": 3050 + }, + { + "epoch": 0.5432692307692307, + "grad_norm": 0.5449416041374207, + "learning_rate": 0.00019113422088992907, + "loss": 0.9093, + "step": 3051 + }, + { + "epoch": 0.5434472934472935, + "grad_norm": 0.49395284056663513, + "learning_rate": 0.00019112845794358902, + "loss": 1.0071, + "step": 3052 + }, + { + "epoch": 0.5436253561253561, + "grad_norm": 0.5478728413581848, + "learning_rate": 0.00019112269321177613, + "loss": 1.2124, + "step": 3053 + }, + { + "epoch": 0.5438034188034188, + "grad_norm": 0.6205173134803772, + "learning_rate": 0.0001911169266946034, + "loss": 1.021, + "step": 3054 + }, + { + "epoch": 0.5439814814814815, + "grad_norm": 0.4777783751487732, + "learning_rate": 0.00019111115839218372, + "loss": 0.9192, + "step": 3055 + }, + { + "epoch": 0.5441595441595442, + "grad_norm": 0.5541689991950989, + "learning_rate": 0.00019110538830463018, + "loss": 1.1248, + "step": 3056 + }, + { + "epoch": 0.5443376068376068, + "grad_norm": 0.4750942289829254, + "learning_rate": 0.0001910996164320558, + "loss": 1.3147, + "step": 3057 + }, + { + "epoch": 0.5445156695156695, + "grad_norm": 0.6283948421478271, + "learning_rate": 0.0001910938427745737, + "loss": 1.0919, + "step": 3058 + }, + { + "epoch": 0.5446937321937322, + "grad_norm": 0.552725076675415, + "learning_rate": 0.00019108806733229698, + "loss": 1.3807, + "step": 3059 + }, + { + "epoch": 0.5448717948717948, + "grad_norm": 0.4832848310470581, + "learning_rate": 0.0001910822901053388, + "loss": 1.0705, + "step": 3060 + }, + { + "epoch": 0.5450498575498576, + "grad_norm": 0.6468375325202942, + "learning_rate": 0.00019107651109381233, + "loss": 1.0766, + "step": 3061 + }, + { + "epoch": 0.5452279202279202, + "grad_norm": 0.5464920401573181, + "learning_rate": 0.00019107073029783083, + "loss": 1.0453, + "step": 3062 + }, + { + "epoch": 0.5454059829059829, + "grad_norm": 0.5321210026741028, + "learning_rate": 0.0001910649477175076, + "loss": 1.2326, + "step": 3063 + }, + { + "epoch": 0.5455840455840456, + "grad_norm": 0.5572962164878845, + "learning_rate": 0.00019105916335295582, + "loss": 1.0673, + "step": 3064 + }, + { + "epoch": 0.5457621082621082, + "grad_norm": 0.5239177942276001, + "learning_rate": 0.00019105337720428894, + "loss": 1.04, + "step": 3065 + }, + { + "epoch": 0.5459401709401709, + "grad_norm": 0.5633319616317749, + "learning_rate": 0.00019104758927162023, + "loss": 0.9606, + "step": 3066 + }, + { + "epoch": 0.5461182336182336, + "grad_norm": 0.5317914485931396, + "learning_rate": 0.0001910417995550632, + "loss": 1.0651, + "step": 3067 + }, + { + "epoch": 0.5462962962962963, + "grad_norm": 0.5126453638076782, + "learning_rate": 0.00019103600805473118, + "loss": 1.0316, + "step": 3068 + }, + { + "epoch": 0.5464743589743589, + "grad_norm": 0.5262107253074646, + "learning_rate": 0.00019103021477073773, + "loss": 1.0752, + "step": 3069 + }, + { + "epoch": 0.5466524216524217, + "grad_norm": 0.5384877324104309, + "learning_rate": 0.0001910244197031963, + "loss": 1.1731, + "step": 3070 + }, + { + "epoch": 0.5468304843304843, + "grad_norm": 0.5126553773880005, + "learning_rate": 0.00019101862285222048, + "loss": 1.2229, + "step": 3071 + }, + { + "epoch": 0.5470085470085471, + "grad_norm": 0.4841194450855255, + "learning_rate": 0.0001910128242179238, + "loss": 0.9955, + "step": 3072 + }, + { + "epoch": 0.5471866096866097, + "grad_norm": 0.526546061038971, + "learning_rate": 0.00019100702380041987, + "loss": 1.2436, + "step": 3073 + }, + { + "epoch": 0.5473646723646723, + "grad_norm": 0.5085833072662354, + "learning_rate": 0.0001910012215998224, + "loss": 1.011, + "step": 3074 + }, + { + "epoch": 0.5475427350427351, + "grad_norm": 0.5149994492530823, + "learning_rate": 0.000190995417616245, + "loss": 0.8632, + "step": 3075 + }, + { + "epoch": 0.5477207977207977, + "grad_norm": 0.48079630732536316, + "learning_rate": 0.00019098961184980145, + "loss": 1.1115, + "step": 3076 + }, + { + "epoch": 0.5478988603988604, + "grad_norm": 0.5769477486610413, + "learning_rate": 0.00019098380430060546, + "loss": 0.9544, + "step": 3077 + }, + { + "epoch": 0.5480769230769231, + "grad_norm": 0.5260093808174133, + "learning_rate": 0.0001909779949687708, + "loss": 1.2354, + "step": 3078 + }, + { + "epoch": 0.5482549857549858, + "grad_norm": 0.5518734455108643, + "learning_rate": 0.00019097218385441135, + "loss": 1.1944, + "step": 3079 + }, + { + "epoch": 0.5484330484330484, + "grad_norm": 0.5436808466911316, + "learning_rate": 0.00019096637095764095, + "loss": 1.0717, + "step": 3080 + }, + { + "epoch": 0.5486111111111112, + "grad_norm": 0.4749584197998047, + "learning_rate": 0.00019096055627857344, + "loss": 1.0417, + "step": 3081 + }, + { + "epoch": 0.5487891737891738, + "grad_norm": 0.5485591292381287, + "learning_rate": 0.0001909547398173228, + "loss": 1.2515, + "step": 3082 + }, + { + "epoch": 0.5489672364672364, + "grad_norm": 0.5751016736030579, + "learning_rate": 0.00019094892157400296, + "loss": 1.2112, + "step": 3083 + }, + { + "epoch": 0.5491452991452992, + "grad_norm": 0.5404475331306458, + "learning_rate": 0.00019094310154872795, + "loss": 0.4334, + "step": 3084 + }, + { + "epoch": 0.5493233618233618, + "grad_norm": 0.5198020935058594, + "learning_rate": 0.00019093727974161178, + "loss": 0.9759, + "step": 3085 + }, + { + "epoch": 0.5495014245014245, + "grad_norm": 0.4893439710140228, + "learning_rate": 0.0001909314561527685, + "loss": 1.1287, + "step": 3086 + }, + { + "epoch": 0.5496794871794872, + "grad_norm": 0.5675956606864929, + "learning_rate": 0.00019092563078231228, + "loss": 1.234, + "step": 3087 + }, + { + "epoch": 0.5498575498575499, + "grad_norm": 0.5539132356643677, + "learning_rate": 0.00019091980363035714, + "loss": 1.2378, + "step": 3088 + }, + { + "epoch": 0.5500356125356125, + "grad_norm": 0.5194353461265564, + "learning_rate": 0.00019091397469701735, + "loss": 1.1338, + "step": 3089 + }, + { + "epoch": 0.5502136752136753, + "grad_norm": 0.5143756866455078, + "learning_rate": 0.0001909081439824071, + "loss": 0.9118, + "step": 3090 + }, + { + "epoch": 0.5503917378917379, + "grad_norm": 0.5624327659606934, + "learning_rate": 0.0001909023114866406, + "loss": 1.035, + "step": 3091 + }, + { + "epoch": 0.5505698005698005, + "grad_norm": 0.5285067558288574, + "learning_rate": 0.0001908964772098321, + "loss": 1.0451, + "step": 3092 + }, + { + "epoch": 0.5507478632478633, + "grad_norm": 0.5730587244033813, + "learning_rate": 0.000190890641152096, + "loss": 1.0672, + "step": 3093 + }, + { + "epoch": 0.5509259259259259, + "grad_norm": 0.5822951197624207, + "learning_rate": 0.0001908848033135466, + "loss": 1.1791, + "step": 3094 + }, + { + "epoch": 0.5511039886039886, + "grad_norm": 0.596161961555481, + "learning_rate": 0.00019087896369429826, + "loss": 1.0954, + "step": 3095 + }, + { + "epoch": 0.5512820512820513, + "grad_norm": 0.5138190984725952, + "learning_rate": 0.00019087312229446542, + "loss": 0.896, + "step": 3096 + }, + { + "epoch": 0.551460113960114, + "grad_norm": 0.5061872601509094, + "learning_rate": 0.0001908672791141625, + "loss": 1.1017, + "step": 3097 + }, + { + "epoch": 0.5516381766381766, + "grad_norm": 0.5189547538757324, + "learning_rate": 0.00019086143415350404, + "loss": 1.2906, + "step": 3098 + }, + { + "epoch": 0.5518162393162394, + "grad_norm": 0.5640039443969727, + "learning_rate": 0.00019085558741260448, + "loss": 1.1001, + "step": 3099 + }, + { + "epoch": 0.551994301994302, + "grad_norm": 0.453867107629776, + "learning_rate": 0.00019084973889157844, + "loss": 0.9731, + "step": 3100 + }, + { + "epoch": 0.5521723646723646, + "grad_norm": 0.5431303977966309, + "learning_rate": 0.0001908438885905405, + "loss": 1.3511, + "step": 3101 + }, + { + "epoch": 0.5523504273504274, + "grad_norm": 0.47693368792533875, + "learning_rate": 0.00019083803650960527, + "loss": 1.0426, + "step": 3102 + }, + { + "epoch": 0.55252849002849, + "grad_norm": 0.4663422703742981, + "learning_rate": 0.00019083218264888743, + "loss": 1.05, + "step": 3103 + }, + { + "epoch": 0.5527065527065527, + "grad_norm": 0.561354398727417, + "learning_rate": 0.00019082632700850164, + "loss": 0.9608, + "step": 3104 + }, + { + "epoch": 0.5528846153846154, + "grad_norm": 0.4981916844844818, + "learning_rate": 0.00019082046958856266, + "loss": 1.1935, + "step": 3105 + }, + { + "epoch": 0.5530626780626781, + "grad_norm": 0.5301326513290405, + "learning_rate": 0.0001908146103891852, + "loss": 1.0646, + "step": 3106 + }, + { + "epoch": 0.5532407407407407, + "grad_norm": 0.5023610591888428, + "learning_rate": 0.00019080874941048416, + "loss": 1.127, + "step": 3107 + }, + { + "epoch": 0.5534188034188035, + "grad_norm": 0.5172514319419861, + "learning_rate": 0.00019080288665257426, + "loss": 1.0435, + "step": 3108 + }, + { + "epoch": 0.5535968660968661, + "grad_norm": 0.6340598464012146, + "learning_rate": 0.00019079702211557048, + "loss": 1.3528, + "step": 3109 + }, + { + "epoch": 0.5537749287749287, + "grad_norm": 0.46882256865501404, + "learning_rate": 0.0001907911557995876, + "loss": 1.1361, + "step": 3110 + }, + { + "epoch": 0.5539529914529915, + "grad_norm": 0.6401382088661194, + "learning_rate": 0.00019078528770474068, + "loss": 1.2415, + "step": 3111 + }, + { + "epoch": 0.5541310541310541, + "grad_norm": 0.5141328573226929, + "learning_rate": 0.00019077941783114463, + "loss": 1.0505, + "step": 3112 + }, + { + "epoch": 0.5543091168091168, + "grad_norm": 0.522318959236145, + "learning_rate": 0.00019077354617891444, + "loss": 1.0964, + "step": 3113 + }, + { + "epoch": 0.5544871794871795, + "grad_norm": 0.539551854133606, + "learning_rate": 0.00019076767274816517, + "loss": 1.0735, + "step": 3114 + }, + { + "epoch": 0.5546652421652422, + "grad_norm": 0.495320200920105, + "learning_rate": 0.00019076179753901195, + "loss": 0.9754, + "step": 3115 + }, + { + "epoch": 0.5548433048433048, + "grad_norm": 0.5499199628829956, + "learning_rate": 0.00019075592055156984, + "loss": 1.0043, + "step": 3116 + }, + { + "epoch": 0.5550213675213675, + "grad_norm": 0.5352509617805481, + "learning_rate": 0.00019075004178595396, + "loss": 1.1701, + "step": 3117 + }, + { + "epoch": 0.5551994301994302, + "grad_norm": 0.5392300486564636, + "learning_rate": 0.00019074416124227953, + "loss": 1.1612, + "step": 3118 + }, + { + "epoch": 0.5553774928774928, + "grad_norm": 0.5195050835609436, + "learning_rate": 0.0001907382789206618, + "loss": 1.0934, + "step": 3119 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 0.5276884436607361, + "learning_rate": 0.000190732394821216, + "loss": 0.9011, + "step": 3120 + }, + { + "epoch": 0.5557336182336182, + "grad_norm": 0.6115903258323669, + "learning_rate": 0.00019072650894405734, + "loss": 1.3065, + "step": 3121 + }, + { + "epoch": 0.5559116809116809, + "grad_norm": 0.5752483010292053, + "learning_rate": 0.00019072062128930127, + "loss": 1.0063, + "step": 3122 + }, + { + "epoch": 0.5560897435897436, + "grad_norm": 0.5508273243904114, + "learning_rate": 0.00019071473185706302, + "loss": 1.2598, + "step": 3123 + }, + { + "epoch": 0.5562678062678063, + "grad_norm": 0.49712198972702026, + "learning_rate": 0.00019070884064745808, + "loss": 0.924, + "step": 3124 + }, + { + "epoch": 0.5564458689458689, + "grad_norm": 0.572849452495575, + "learning_rate": 0.00019070294766060185, + "loss": 0.9683, + "step": 3125 + }, + { + "epoch": 0.5566239316239316, + "grad_norm": 0.4807920753955841, + "learning_rate": 0.00019069705289660976, + "loss": 1.0998, + "step": 3126 + }, + { + "epoch": 0.5568019943019943, + "grad_norm": 0.5543031096458435, + "learning_rate": 0.0001906911563555973, + "loss": 1.0878, + "step": 3127 + }, + { + "epoch": 0.5569800569800569, + "grad_norm": 0.5710418820381165, + "learning_rate": 0.00019068525803768007, + "loss": 1.0381, + "step": 3128 + }, + { + "epoch": 0.5571581196581197, + "grad_norm": 0.5169163346290588, + "learning_rate": 0.00019067935794297357, + "loss": 1.1149, + "step": 3129 + }, + { + "epoch": 0.5573361823361823, + "grad_norm": 0.6474376916885376, + "learning_rate": 0.00019067345607159345, + "loss": 0.9828, + "step": 3130 + }, + { + "epoch": 0.5575142450142451, + "grad_norm": 0.5029847621917725, + "learning_rate": 0.0001906675524236553, + "loss": 0.797, + "step": 3131 + }, + { + "epoch": 0.5576923076923077, + "grad_norm": 0.5681431293487549, + "learning_rate": 0.00019066164699927478, + "loss": 1.1565, + "step": 3132 + }, + { + "epoch": 0.5578703703703703, + "grad_norm": 0.5654549598693848, + "learning_rate": 0.00019065573979856764, + "loss": 1.2488, + "step": 3133 + }, + { + "epoch": 0.5580484330484331, + "grad_norm": 0.47653043270111084, + "learning_rate": 0.0001906498308216496, + "loss": 1.0428, + "step": 3134 + }, + { + "epoch": 0.5582264957264957, + "grad_norm": 0.5068467259407043, + "learning_rate": 0.00019064392006863643, + "loss": 0.9659, + "step": 3135 + }, + { + "epoch": 0.5584045584045584, + "grad_norm": 0.7076661586761475, + "learning_rate": 0.00019063800753964393, + "loss": 1.1289, + "step": 3136 + }, + { + "epoch": 0.5585826210826211, + "grad_norm": 0.551456868648529, + "learning_rate": 0.000190632093234788, + "loss": 1.1925, + "step": 3137 + }, + { + "epoch": 0.5587606837606838, + "grad_norm": 0.518276035785675, + "learning_rate": 0.00019062617715418442, + "loss": 0.8681, + "step": 3138 + }, + { + "epoch": 0.5589387464387464, + "grad_norm": 0.5272278785705566, + "learning_rate": 0.0001906202592979492, + "loss": 1.0865, + "step": 3139 + }, + { + "epoch": 0.5591168091168092, + "grad_norm": 0.5344942212104797, + "learning_rate": 0.00019061433966619822, + "loss": 1.1647, + "step": 3140 + }, + { + "epoch": 0.5592948717948718, + "grad_norm": 0.5833460092544556, + "learning_rate": 0.00019060841825904753, + "loss": 1.3403, + "step": 3141 + }, + { + "epoch": 0.5594729344729344, + "grad_norm": 0.5707054734230042, + "learning_rate": 0.00019060249507661306, + "loss": 1.1236, + "step": 3142 + }, + { + "epoch": 0.5596509971509972, + "grad_norm": 0.5446065664291382, + "learning_rate": 0.00019059657011901094, + "loss": 1.017, + "step": 3143 + }, + { + "epoch": 0.5598290598290598, + "grad_norm": 0.5285109281539917, + "learning_rate": 0.0001905906433863572, + "loss": 1.3186, + "step": 3144 + }, + { + "epoch": 0.5600071225071225, + "grad_norm": 0.5308659672737122, + "learning_rate": 0.00019058471487876802, + "loss": 0.8464, + "step": 3145 + }, + { + "epoch": 0.5601851851851852, + "grad_norm": 0.5218054056167603, + "learning_rate": 0.00019057878459635948, + "loss": 1.0219, + "step": 3146 + }, + { + "epoch": 0.5603632478632479, + "grad_norm": 0.45067787170410156, + "learning_rate": 0.00019057285253924785, + "loss": 1.0364, + "step": 3147 + }, + { + "epoch": 0.5605413105413105, + "grad_norm": 0.4856041669845581, + "learning_rate": 0.0001905669187075493, + "loss": 1.1928, + "step": 3148 + }, + { + "epoch": 0.5607193732193733, + "grad_norm": 0.506912112236023, + "learning_rate": 0.00019056098310138016, + "loss": 1.119, + "step": 3149 + }, + { + "epoch": 0.5608974358974359, + "grad_norm": 0.49049463868141174, + "learning_rate": 0.00019055504572085662, + "loss": 1.2165, + "step": 3150 + }, + { + "epoch": 0.5610754985754985, + "grad_norm": 0.5250293612480164, + "learning_rate": 0.0001905491065660951, + "loss": 1.1427, + "step": 3151 + }, + { + "epoch": 0.5612535612535613, + "grad_norm": 0.43438446521759033, + "learning_rate": 0.00019054316563721195, + "loss": 0.884, + "step": 3152 + }, + { + "epoch": 0.5614316239316239, + "grad_norm": 0.5386807918548584, + "learning_rate": 0.00019053722293432354, + "loss": 1.1494, + "step": 3153 + }, + { + "epoch": 0.5616096866096866, + "grad_norm": 0.5403809547424316, + "learning_rate": 0.00019053127845754632, + "loss": 1.1743, + "step": 3154 + }, + { + "epoch": 0.5617877492877493, + "grad_norm": 0.4759823977947235, + "learning_rate": 0.00019052533220699678, + "loss": 1.0716, + "step": 3155 + }, + { + "epoch": 0.561965811965812, + "grad_norm": 0.45332327485084534, + "learning_rate": 0.0001905193841827914, + "loss": 0.8405, + "step": 3156 + }, + { + "epoch": 0.5621438746438746, + "grad_norm": 0.5617053508758545, + "learning_rate": 0.00019051343438504671, + "loss": 1.0422, + "step": 3157 + }, + { + "epoch": 0.5623219373219374, + "grad_norm": 0.5088049173355103, + "learning_rate": 0.00019050748281387931, + "loss": 1.0067, + "step": 3158 + }, + { + "epoch": 0.5625, + "grad_norm": 0.5174484848976135, + "learning_rate": 0.00019050152946940578, + "loss": 1.0623, + "step": 3159 + }, + { + "epoch": 0.5626780626780626, + "grad_norm": 0.6093568801879883, + "learning_rate": 0.0001904955743517428, + "loss": 1.24, + "step": 3160 + }, + { + "epoch": 0.5628561253561254, + "grad_norm": 0.49063584208488464, + "learning_rate": 0.00019048961746100703, + "loss": 0.8563, + "step": 3161 + }, + { + "epoch": 0.563034188034188, + "grad_norm": 0.583940863609314, + "learning_rate": 0.00019048365879731517, + "loss": 1.0695, + "step": 3162 + }, + { + "epoch": 0.5632122507122507, + "grad_norm": 0.4943268597126007, + "learning_rate": 0.000190477698360784, + "loss": 0.8606, + "step": 3163 + }, + { + "epoch": 0.5633903133903134, + "grad_norm": 0.5050932168960571, + "learning_rate": 0.00019047173615153028, + "loss": 1.1591, + "step": 3164 + }, + { + "epoch": 0.5635683760683761, + "grad_norm": 0.5445677638053894, + "learning_rate": 0.0001904657721696708, + "loss": 1.262, + "step": 3165 + }, + { + "epoch": 0.5637464387464387, + "grad_norm": 0.5445297360420227, + "learning_rate": 0.00019045980641532246, + "loss": 1.223, + "step": 3166 + }, + { + "epoch": 0.5639245014245015, + "grad_norm": 0.5098413228988647, + "learning_rate": 0.00019045383888860213, + "loss": 1.0829, + "step": 3167 + }, + { + "epoch": 0.5641025641025641, + "grad_norm": 0.484998881816864, + "learning_rate": 0.0001904478695896267, + "loss": 1.0711, + "step": 3168 + }, + { + "epoch": 0.5642806267806267, + "grad_norm": 0.5515334010124207, + "learning_rate": 0.0001904418985185132, + "loss": 1.1583, + "step": 3169 + }, + { + "epoch": 0.5644586894586895, + "grad_norm": 0.545460045337677, + "learning_rate": 0.00019043592567537853, + "loss": 1.2321, + "step": 3170 + }, + { + "epoch": 0.5646367521367521, + "grad_norm": 0.5463964343070984, + "learning_rate": 0.0001904299510603398, + "loss": 1.1019, + "step": 3171 + }, + { + "epoch": 0.5648148148148148, + "grad_norm": 0.5619220733642578, + "learning_rate": 0.000190423974673514, + "loss": 1.1001, + "step": 3172 + }, + { + "epoch": 0.5649928774928775, + "grad_norm": 0.4448916018009186, + "learning_rate": 0.00019041799651501825, + "loss": 1.057, + "step": 3173 + }, + { + "epoch": 0.5651709401709402, + "grad_norm": 0.6073006987571716, + "learning_rate": 0.00019041201658496975, + "loss": 1.0306, + "step": 3174 + }, + { + "epoch": 0.5653490028490028, + "grad_norm": 0.5342072248458862, + "learning_rate": 0.0001904060348834855, + "loss": 0.9231, + "step": 3175 + }, + { + "epoch": 0.5655270655270656, + "grad_norm": 0.4505697786808014, + "learning_rate": 0.0001904000514106829, + "loss": 1.1134, + "step": 3176 + }, + { + "epoch": 0.5657051282051282, + "grad_norm": 0.5627852082252502, + "learning_rate": 0.00019039406616667902, + "loss": 1.2138, + "step": 3177 + }, + { + "epoch": 0.5658831908831908, + "grad_norm": 0.499734103679657, + "learning_rate": 0.0001903880791515912, + "loss": 1.1074, + "step": 3178 + }, + { + "epoch": 0.5660612535612536, + "grad_norm": 0.4768189489841461, + "learning_rate": 0.00019038209036553676, + "loss": 0.9442, + "step": 3179 + }, + { + "epoch": 0.5662393162393162, + "grad_norm": 0.5265373587608337, + "learning_rate": 0.00019037609980863298, + "loss": 1.0907, + "step": 3180 + }, + { + "epoch": 0.5664173789173789, + "grad_norm": 0.5506128072738647, + "learning_rate": 0.00019037010748099728, + "loss": 1.2541, + "step": 3181 + }, + { + "epoch": 0.5665954415954416, + "grad_norm": 0.44860872626304626, + "learning_rate": 0.00019036411338274703, + "loss": 0.893, + "step": 3182 + }, + { + "epoch": 0.5667735042735043, + "grad_norm": 0.4901522994041443, + "learning_rate": 0.00019035811751399973, + "loss": 1.0469, + "step": 3183 + }, + { + "epoch": 0.5669515669515669, + "grad_norm": 0.500868022441864, + "learning_rate": 0.0001903521198748728, + "loss": 1.0527, + "step": 3184 + }, + { + "epoch": 0.5671296296296297, + "grad_norm": 0.5508102774620056, + "learning_rate": 0.00019034612046548376, + "loss": 1.283, + "step": 3185 + }, + { + "epoch": 0.5673076923076923, + "grad_norm": 0.5079495906829834, + "learning_rate": 0.0001903401192859502, + "loss": 1.0808, + "step": 3186 + }, + { + "epoch": 0.5674857549857549, + "grad_norm": 0.5758788585662842, + "learning_rate": 0.00019033411633638964, + "loss": 1.1301, + "step": 3187 + }, + { + "epoch": 0.5676638176638177, + "grad_norm": 0.46557924151420593, + "learning_rate": 0.00019032811161691972, + "loss": 1.0205, + "step": 3188 + }, + { + "epoch": 0.5678418803418803, + "grad_norm": 0.5665056109428406, + "learning_rate": 0.0001903221051276581, + "loss": 1.1926, + "step": 3189 + }, + { + "epoch": 0.5680199430199431, + "grad_norm": 0.5948992967605591, + "learning_rate": 0.00019031609686872246, + "loss": 1.2724, + "step": 3190 + }, + { + "epoch": 0.5681980056980057, + "grad_norm": 0.6189367771148682, + "learning_rate": 0.00019031008684023055, + "loss": 1.2762, + "step": 3191 + }, + { + "epoch": 0.5683760683760684, + "grad_norm": 0.49511992931365967, + "learning_rate": 0.00019030407504230006, + "loss": 1.0117, + "step": 3192 + }, + { + "epoch": 0.5685541310541311, + "grad_norm": 0.5358837842941284, + "learning_rate": 0.00019029806147504878, + "loss": 0.944, + "step": 3193 + }, + { + "epoch": 0.5687321937321937, + "grad_norm": 0.458636999130249, + "learning_rate": 0.00019029204613859463, + "loss": 0.8174, + "step": 3194 + }, + { + "epoch": 0.5689102564102564, + "grad_norm": 0.5168304443359375, + "learning_rate": 0.00019028602903305535, + "loss": 1.1533, + "step": 3195 + }, + { + "epoch": 0.5690883190883191, + "grad_norm": 0.5334134697914124, + "learning_rate": 0.00019028001015854892, + "loss": 1.1868, + "step": 3196 + }, + { + "epoch": 0.5692663817663818, + "grad_norm": 0.5649123191833496, + "learning_rate": 0.0001902739895151932, + "loss": 0.9876, + "step": 3197 + }, + { + "epoch": 0.5694444444444444, + "grad_norm": 0.5647651553153992, + "learning_rate": 0.0001902679671031062, + "loss": 1.0805, + "step": 3198 + }, + { + "epoch": 0.5696225071225072, + "grad_norm": 0.5251876711845398, + "learning_rate": 0.00019026194292240587, + "loss": 1.2335, + "step": 3199 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 0.5268014669418335, + "learning_rate": 0.0001902559169732103, + "loss": 1.19, + "step": 3200 + }, + { + "epoch": 0.5699786324786325, + "grad_norm": 0.5301041007041931, + "learning_rate": 0.00019024988925563752, + "loss": 1.1173, + "step": 3201 + }, + { + "epoch": 0.5701566951566952, + "grad_norm": 0.4531562030315399, + "learning_rate": 0.00019024385976980566, + "loss": 0.7576, + "step": 3202 + }, + { + "epoch": 0.5703347578347578, + "grad_norm": 0.5779716372489929, + "learning_rate": 0.00019023782851583282, + "loss": 1.1719, + "step": 3203 + }, + { + "epoch": 0.5705128205128205, + "grad_norm": 0.4886093735694885, + "learning_rate": 0.00019023179549383716, + "loss": 1.085, + "step": 3204 + }, + { + "epoch": 0.5706908831908832, + "grad_norm": 0.510117769241333, + "learning_rate": 0.0001902257607039369, + "loss": 0.8931, + "step": 3205 + }, + { + "epoch": 0.5708689458689459, + "grad_norm": 0.5195479393005371, + "learning_rate": 0.00019021972414625036, + "loss": 0.9922, + "step": 3206 + }, + { + "epoch": 0.5710470085470085, + "grad_norm": 0.5791407227516174, + "learning_rate": 0.00019021368582089568, + "loss": 1.112, + "step": 3207 + }, + { + "epoch": 0.5712250712250713, + "grad_norm": 0.5056005716323853, + "learning_rate": 0.00019020764572799122, + "loss": 0.8474, + "step": 3208 + }, + { + "epoch": 0.5714031339031339, + "grad_norm": 0.5060068964958191, + "learning_rate": 0.00019020160386765537, + "loss": 1.071, + "step": 3209 + }, + { + "epoch": 0.5715811965811965, + "grad_norm": 0.5396568775177002, + "learning_rate": 0.00019019556024000648, + "loss": 1.0436, + "step": 3210 + }, + { + "epoch": 0.5717592592592593, + "grad_norm": 0.6552190780639648, + "learning_rate": 0.0001901895148451629, + "loss": 0.9869, + "step": 3211 + }, + { + "epoch": 0.5719373219373219, + "grad_norm": 0.5177004337310791, + "learning_rate": 0.00019018346768324314, + "loss": 1.0193, + "step": 3212 + }, + { + "epoch": 0.5721153846153846, + "grad_norm": 0.5192117094993591, + "learning_rate": 0.0001901774187543657, + "loss": 1.1263, + "step": 3213 + }, + { + "epoch": 0.5722934472934473, + "grad_norm": 0.4857729971408844, + "learning_rate": 0.00019017136805864906, + "loss": 0.9808, + "step": 3214 + }, + { + "epoch": 0.57247150997151, + "grad_norm": 0.5800918936729431, + "learning_rate": 0.00019016531559621177, + "loss": 1.2334, + "step": 3215 + }, + { + "epoch": 0.5726495726495726, + "grad_norm": 0.4812086522579193, + "learning_rate": 0.00019015926136717242, + "loss": 1.2409, + "step": 3216 + }, + { + "epoch": 0.5728276353276354, + "grad_norm": 0.5128398537635803, + "learning_rate": 0.00019015320537164963, + "loss": 0.9036, + "step": 3217 + }, + { + "epoch": 0.573005698005698, + "grad_norm": 0.4761141538619995, + "learning_rate": 0.00019014714760976205, + "loss": 1.1058, + "step": 3218 + }, + { + "epoch": 0.5731837606837606, + "grad_norm": 0.5850459933280945, + "learning_rate": 0.0001901410880816284, + "loss": 1.1011, + "step": 3219 + }, + { + "epoch": 0.5733618233618234, + "grad_norm": 0.5648714303970337, + "learning_rate": 0.00019013502678736738, + "loss": 1.0479, + "step": 3220 + }, + { + "epoch": 0.573539886039886, + "grad_norm": 0.5835902094841003, + "learning_rate": 0.00019012896372709774, + "loss": 1.0555, + "step": 3221 + }, + { + "epoch": 0.5737179487179487, + "grad_norm": 0.5155113935470581, + "learning_rate": 0.00019012289890093828, + "loss": 0.9488, + "step": 3222 + }, + { + "epoch": 0.5738960113960114, + "grad_norm": 0.5064889788627625, + "learning_rate": 0.00019011683230900784, + "loss": 0.9144, + "step": 3223 + }, + { + "epoch": 0.5740740740740741, + "grad_norm": 0.53825843334198, + "learning_rate": 0.00019011076395142527, + "loss": 1.0713, + "step": 3224 + }, + { + "epoch": 0.5742521367521367, + "grad_norm": 0.5341386198997498, + "learning_rate": 0.00019010469382830947, + "loss": 1.1438, + "step": 3225 + }, + { + "epoch": 0.5744301994301995, + "grad_norm": 0.5300050973892212, + "learning_rate": 0.00019009862193977936, + "loss": 1.0114, + "step": 3226 + }, + { + "epoch": 0.5746082621082621, + "grad_norm": 0.6033682823181152, + "learning_rate": 0.0001900925482859539, + "loss": 1.0458, + "step": 3227 + }, + { + "epoch": 0.5747863247863247, + "grad_norm": 0.5108983516693115, + "learning_rate": 0.00019008647286695215, + "loss": 1.1211, + "step": 3228 + }, + { + "epoch": 0.5749643874643875, + "grad_norm": 0.5263782739639282, + "learning_rate": 0.00019008039568289308, + "loss": 0.8647, + "step": 3229 + }, + { + "epoch": 0.5751424501424501, + "grad_norm": 0.47119566798210144, + "learning_rate": 0.0001900743167338958, + "loss": 1.019, + "step": 3230 + }, + { + "epoch": 0.5753205128205128, + "grad_norm": 0.56391841173172, + "learning_rate": 0.00019006823602007937, + "loss": 0.9791, + "step": 3231 + }, + { + "epoch": 0.5754985754985755, + "grad_norm": 0.5364985466003418, + "learning_rate": 0.000190062153541563, + "loss": 1.1355, + "step": 3232 + }, + { + "epoch": 0.5756766381766382, + "grad_norm": 0.5098565220832825, + "learning_rate": 0.00019005606929846578, + "loss": 0.987, + "step": 3233 + }, + { + "epoch": 0.5758547008547008, + "grad_norm": 0.6640968918800354, + "learning_rate": 0.00019004998329090692, + "loss": 1.1165, + "step": 3234 + }, + { + "epoch": 0.5760327635327636, + "grad_norm": 0.5044721961021423, + "learning_rate": 0.00019004389551900578, + "loss": 0.8643, + "step": 3235 + }, + { + "epoch": 0.5762108262108262, + "grad_norm": 0.4822785258293152, + "learning_rate": 0.00019003780598288153, + "loss": 1.0735, + "step": 3236 + }, + { + "epoch": 0.5763888888888888, + "grad_norm": 0.505261242389679, + "learning_rate": 0.00019003171468265348, + "loss": 1.0001, + "step": 3237 + }, + { + "epoch": 0.5765669515669516, + "grad_norm": 0.5020412802696228, + "learning_rate": 0.00019002562161844102, + "loss": 0.9601, + "step": 3238 + }, + { + "epoch": 0.5767450142450142, + "grad_norm": 0.4920475482940674, + "learning_rate": 0.00019001952679036354, + "loss": 1.0111, + "step": 3239 + }, + { + "epoch": 0.5769230769230769, + "grad_norm": 0.5638813376426697, + "learning_rate": 0.00019001343019854042, + "loss": 1.1456, + "step": 3240 + }, + { + "epoch": 0.5771011396011396, + "grad_norm": 0.5519235134124756, + "learning_rate": 0.0001900073318430911, + "loss": 0.9258, + "step": 3241 + }, + { + "epoch": 0.5772792022792023, + "grad_norm": 0.5207770466804504, + "learning_rate": 0.0001900012317241351, + "loss": 0.9859, + "step": 3242 + }, + { + "epoch": 0.5774572649572649, + "grad_norm": 0.5493707656860352, + "learning_rate": 0.00018999512984179195, + "loss": 1.1183, + "step": 3243 + }, + { + "epoch": 0.5776353276353277, + "grad_norm": 0.4504764676094055, + "learning_rate": 0.00018998902619618116, + "loss": 0.9363, + "step": 3244 + }, + { + "epoch": 0.5778133903133903, + "grad_norm": 0.5232836604118347, + "learning_rate": 0.00018998292078742233, + "loss": 1.1887, + "step": 3245 + }, + { + "epoch": 0.5779914529914529, + "grad_norm": 0.5715088248252869, + "learning_rate": 0.0001899768136156351, + "loss": 1.4524, + "step": 3246 + }, + { + "epoch": 0.5781695156695157, + "grad_norm": 0.59555584192276, + "learning_rate": 0.0001899707046809391, + "loss": 1.0922, + "step": 3247 + }, + { + "epoch": 0.5783475783475783, + "grad_norm": 0.4500894546508789, + "learning_rate": 0.00018996459398345404, + "loss": 1.0087, + "step": 3248 + }, + { + "epoch": 0.5785256410256411, + "grad_norm": 0.49126625061035156, + "learning_rate": 0.00018995848152329967, + "loss": 1.1512, + "step": 3249 + }, + { + "epoch": 0.5787037037037037, + "grad_norm": 0.4096335172653198, + "learning_rate": 0.00018995236730059574, + "loss": 0.7633, + "step": 3250 + }, + { + "epoch": 0.5788817663817664, + "grad_norm": 0.5364313721656799, + "learning_rate": 0.00018994625131546199, + "loss": 1.295, + "step": 3251 + }, + { + "epoch": 0.5790598290598291, + "grad_norm": 0.4897502660751343, + "learning_rate": 0.00018994013356801834, + "loss": 1.2197, + "step": 3252 + }, + { + "epoch": 0.5792378917378918, + "grad_norm": 0.5101368427276611, + "learning_rate": 0.00018993401405838456, + "loss": 1.1129, + "step": 3253 + }, + { + "epoch": 0.5794159544159544, + "grad_norm": 0.5426377654075623, + "learning_rate": 0.00018992789278668063, + "loss": 1.188, + "step": 3254 + }, + { + "epoch": 0.5795940170940171, + "grad_norm": 0.5066362023353577, + "learning_rate": 0.00018992176975302644, + "loss": 1.2802, + "step": 3255 + }, + { + "epoch": 0.5797720797720798, + "grad_norm": 0.5418947339057922, + "learning_rate": 0.00018991564495754196, + "loss": 1.1675, + "step": 3256 + }, + { + "epoch": 0.5799501424501424, + "grad_norm": 0.5139963626861572, + "learning_rate": 0.0001899095184003472, + "loss": 0.9717, + "step": 3257 + }, + { + "epoch": 0.5801282051282052, + "grad_norm": 0.5167285799980164, + "learning_rate": 0.00018990339008156219, + "loss": 1.1529, + "step": 3258 + }, + { + "epoch": 0.5803062678062678, + "grad_norm": 0.53471440076828, + "learning_rate": 0.00018989726000130704, + "loss": 1.0711, + "step": 3259 + }, + { + "epoch": 0.5804843304843305, + "grad_norm": 0.49875229597091675, + "learning_rate": 0.0001898911281597018, + "loss": 1.1095, + "step": 3260 + }, + { + "epoch": 0.5806623931623932, + "grad_norm": 0.4473155438899994, + "learning_rate": 0.00018988499455686663, + "loss": 0.836, + "step": 3261 + }, + { + "epoch": 0.5808404558404558, + "grad_norm": 0.6181996464729309, + "learning_rate": 0.00018987885919292174, + "loss": 1.2787, + "step": 3262 + }, + { + "epoch": 0.5810185185185185, + "grad_norm": 0.4996899664402008, + "learning_rate": 0.00018987272206798733, + "loss": 1.2132, + "step": 3263 + }, + { + "epoch": 0.5811965811965812, + "grad_norm": 0.49979713559150696, + "learning_rate": 0.00018986658318218358, + "loss": 0.8388, + "step": 3264 + }, + { + "epoch": 0.5813746438746439, + "grad_norm": 0.5288876295089722, + "learning_rate": 0.00018986044253563084, + "loss": 1.1871, + "step": 3265 + }, + { + "epoch": 0.5815527065527065, + "grad_norm": 0.534063458442688, + "learning_rate": 0.00018985430012844937, + "loss": 0.96, + "step": 3266 + }, + { + "epoch": 0.5817307692307693, + "grad_norm": 0.5081285834312439, + "learning_rate": 0.00018984815596075953, + "loss": 1.1577, + "step": 3267 + }, + { + "epoch": 0.5819088319088319, + "grad_norm": 0.5648202896118164, + "learning_rate": 0.00018984201003268176, + "loss": 1.2235, + "step": 3268 + }, + { + "epoch": 0.5820868945868946, + "grad_norm": 0.495061993598938, + "learning_rate": 0.00018983586234433642, + "loss": 1.056, + "step": 3269 + }, + { + "epoch": 0.5822649572649573, + "grad_norm": 0.47149857878685, + "learning_rate": 0.000189829712895844, + "loss": 1.0844, + "step": 3270 + }, + { + "epoch": 0.58244301994302, + "grad_norm": 0.6107062697410583, + "learning_rate": 0.00018982356168732492, + "loss": 0.9868, + "step": 3271 + }, + { + "epoch": 0.5826210826210826, + "grad_norm": 0.7355940341949463, + "learning_rate": 0.00018981740871889974, + "loss": 1.1448, + "step": 3272 + }, + { + "epoch": 0.5827991452991453, + "grad_norm": 0.5950441956520081, + "learning_rate": 0.00018981125399068907, + "loss": 0.9618, + "step": 3273 + }, + { + "epoch": 0.582977207977208, + "grad_norm": 0.47607290744781494, + "learning_rate": 0.0001898050975028134, + "loss": 0.957, + "step": 3274 + }, + { + "epoch": 0.5831552706552706, + "grad_norm": 0.541164755821228, + "learning_rate": 0.00018979893925539338, + "loss": 1.1426, + "step": 3275 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 0.5240640044212341, + "learning_rate": 0.00018979277924854974, + "loss": 1.1421, + "step": 3276 + }, + { + "epoch": 0.583511396011396, + "grad_norm": 0.48155727982521057, + "learning_rate": 0.00018978661748240307, + "loss": 1.0069, + "step": 3277 + }, + { + "epoch": 0.5836894586894587, + "grad_norm": 0.5559938549995422, + "learning_rate": 0.00018978045395707418, + "loss": 1.1227, + "step": 3278 + }, + { + "epoch": 0.5838675213675214, + "grad_norm": 0.5244291424751282, + "learning_rate": 0.0001897742886726838, + "loss": 1.1103, + "step": 3279 + }, + { + "epoch": 0.584045584045584, + "grad_norm": 0.5277758240699768, + "learning_rate": 0.00018976812162935268, + "loss": 1.2125, + "step": 3280 + }, + { + "epoch": 0.5842236467236467, + "grad_norm": 0.5415039658546448, + "learning_rate": 0.00018976195282720173, + "loss": 1.146, + "step": 3281 + }, + { + "epoch": 0.5844017094017094, + "grad_norm": 0.5152051448822021, + "learning_rate": 0.00018975578226635177, + "loss": 1.0092, + "step": 3282 + }, + { + "epoch": 0.5845797720797721, + "grad_norm": 0.5489452481269836, + "learning_rate": 0.00018974960994692371, + "loss": 1.2425, + "step": 3283 + }, + { + "epoch": 0.5847578347578347, + "grad_norm": 0.491274356842041, + "learning_rate": 0.00018974343586903848, + "loss": 0.9559, + "step": 3284 + }, + { + "epoch": 0.5849358974358975, + "grad_norm": 0.5783739686012268, + "learning_rate": 0.00018973726003281707, + "loss": 1.1971, + "step": 3285 + }, + { + "epoch": 0.5851139601139601, + "grad_norm": 0.5056472420692444, + "learning_rate": 0.00018973108243838045, + "loss": 1.0313, + "step": 3286 + }, + { + "epoch": 0.5852920227920227, + "grad_norm": 0.4939729571342468, + "learning_rate": 0.00018972490308584962, + "loss": 1.1061, + "step": 3287 + }, + { + "epoch": 0.5854700854700855, + "grad_norm": 0.4889580011367798, + "learning_rate": 0.00018971872197534576, + "loss": 0.9157, + "step": 3288 + }, + { + "epoch": 0.5856481481481481, + "grad_norm": 0.40889349579811096, + "learning_rate": 0.00018971253910698993, + "loss": 0.8083, + "step": 3289 + }, + { + "epoch": 0.5858262108262108, + "grad_norm": 0.5221503973007202, + "learning_rate": 0.00018970635448090322, + "loss": 0.9995, + "step": 3290 + }, + { + "epoch": 0.5860042735042735, + "grad_norm": 0.47060561180114746, + "learning_rate": 0.00018970016809720687, + "loss": 0.9738, + "step": 3291 + }, + { + "epoch": 0.5861823361823362, + "grad_norm": 0.6083170771598816, + "learning_rate": 0.000189693979956022, + "loss": 1.188, + "step": 3292 + }, + { + "epoch": 0.5863603988603988, + "grad_norm": 0.4696751534938812, + "learning_rate": 0.00018968779005746998, + "loss": 1.089, + "step": 3293 + }, + { + "epoch": 0.5865384615384616, + "grad_norm": 0.5081014633178711, + "learning_rate": 0.00018968159840167202, + "loss": 1.1869, + "step": 3294 + }, + { + "epoch": 0.5867165242165242, + "grad_norm": 0.48042431473731995, + "learning_rate": 0.0001896754049887494, + "loss": 0.964, + "step": 3295 + }, + { + "epoch": 0.5868945868945868, + "grad_norm": 0.5075193643569946, + "learning_rate": 0.00018966920981882353, + "loss": 1.1884, + "step": 3296 + }, + { + "epoch": 0.5870726495726496, + "grad_norm": 0.5734842419624329, + "learning_rate": 0.00018966301289201576, + "loss": 1.1475, + "step": 3297 + }, + { + "epoch": 0.5872507122507122, + "grad_norm": 0.5525311231613159, + "learning_rate": 0.00018965681420844753, + "loss": 1.241, + "step": 3298 + }, + { + "epoch": 0.5874287749287749, + "grad_norm": 0.48142680525779724, + "learning_rate": 0.00018965061376824025, + "loss": 1.0871, + "step": 3299 + }, + { + "epoch": 0.5876068376068376, + "grad_norm": 0.5360350608825684, + "learning_rate": 0.00018964441157151544, + "loss": 1.1895, + "step": 3300 + }, + { + "epoch": 0.5877849002849003, + "grad_norm": 0.5207685232162476, + "learning_rate": 0.00018963820761839457, + "loss": 0.9323, + "step": 3301 + }, + { + "epoch": 0.5879629629629629, + "grad_norm": 0.453620970249176, + "learning_rate": 0.00018963200190899926, + "loss": 0.802, + "step": 3302 + }, + { + "epoch": 0.5881410256410257, + "grad_norm": 0.5198796391487122, + "learning_rate": 0.00018962579444345106, + "loss": 1.0243, + "step": 3303 + }, + { + "epoch": 0.5883190883190883, + "grad_norm": 0.5597525835037231, + "learning_rate": 0.0001896195852218716, + "loss": 0.9351, + "step": 3304 + }, + { + "epoch": 0.5884971509971509, + "grad_norm": 0.5738299489021301, + "learning_rate": 0.00018961337424438254, + "loss": 1.3737, + "step": 3305 + }, + { + "epoch": 0.5886752136752137, + "grad_norm": 0.5569949150085449, + "learning_rate": 0.00018960716151110554, + "loss": 1.0469, + "step": 3306 + }, + { + "epoch": 0.5888532763532763, + "grad_norm": 0.5088010430335999, + "learning_rate": 0.00018960094702216238, + "loss": 1.0982, + "step": 3307 + }, + { + "epoch": 0.5890313390313391, + "grad_norm": 0.5127636790275574, + "learning_rate": 0.0001895947307776748, + "loss": 0.9986, + "step": 3308 + }, + { + "epoch": 0.5892094017094017, + "grad_norm": 0.5160682797431946, + "learning_rate": 0.00018958851277776456, + "loss": 1.0219, + "step": 3309 + }, + { + "epoch": 0.5893874643874644, + "grad_norm": 0.5380711555480957, + "learning_rate": 0.00018958229302255356, + "loss": 1.118, + "step": 3310 + }, + { + "epoch": 0.5895655270655271, + "grad_norm": 0.5571228861808777, + "learning_rate": 0.0001895760715121636, + "loss": 1.0302, + "step": 3311 + }, + { + "epoch": 0.5897435897435898, + "grad_norm": 0.542266309261322, + "learning_rate": 0.00018956984824671657, + "loss": 1.0372, + "step": 3312 + }, + { + "epoch": 0.5899216524216524, + "grad_norm": 0.48350459337234497, + "learning_rate": 0.00018956362322633446, + "loss": 1.2, + "step": 3313 + }, + { + "epoch": 0.5900997150997151, + "grad_norm": 0.5001645088195801, + "learning_rate": 0.0001895573964511392, + "loss": 0.9749, + "step": 3314 + }, + { + "epoch": 0.5902777777777778, + "grad_norm": 0.5227531790733337, + "learning_rate": 0.00018955116792125276, + "loss": 1.025, + "step": 3315 + }, + { + "epoch": 0.5904558404558404, + "grad_norm": 0.522251546382904, + "learning_rate": 0.00018954493763679727, + "loss": 1.0821, + "step": 3316 + }, + { + "epoch": 0.5906339031339032, + "grad_norm": 0.5423251390457153, + "learning_rate": 0.00018953870559789467, + "loss": 1.0961, + "step": 3317 + }, + { + "epoch": 0.5908119658119658, + "grad_norm": 0.5615720748901367, + "learning_rate": 0.0001895324718046672, + "loss": 1.1209, + "step": 3318 + }, + { + "epoch": 0.5909900284900285, + "grad_norm": 0.44746771454811096, + "learning_rate": 0.00018952623625723692, + "loss": 0.9935, + "step": 3319 + }, + { + "epoch": 0.5911680911680912, + "grad_norm": 0.5993229150772095, + "learning_rate": 0.00018951999895572597, + "loss": 1.1409, + "step": 3320 + }, + { + "epoch": 0.5913461538461539, + "grad_norm": 0.4969801902770996, + "learning_rate": 0.00018951375990025666, + "loss": 1.1568, + "step": 3321 + }, + { + "epoch": 0.5915242165242165, + "grad_norm": 0.6001267433166504, + "learning_rate": 0.00018950751909095116, + "loss": 1.1135, + "step": 3322 + }, + { + "epoch": 0.5917022792022792, + "grad_norm": 0.5386021733283997, + "learning_rate": 0.00018950127652793172, + "loss": 0.947, + "step": 3323 + }, + { + "epoch": 0.5918803418803419, + "grad_norm": 0.49043843150138855, + "learning_rate": 0.00018949503221132074, + "loss": 0.9581, + "step": 3324 + }, + { + "epoch": 0.5920584045584045, + "grad_norm": 0.5241141319274902, + "learning_rate": 0.00018948878614124048, + "loss": 1.0797, + "step": 3325 + }, + { + "epoch": 0.5922364672364673, + "grad_norm": 0.5755026340484619, + "learning_rate": 0.00018948253831781338, + "loss": 1.1046, + "step": 3326 + }, + { + "epoch": 0.5924145299145299, + "grad_norm": 0.5004449486732483, + "learning_rate": 0.00018947628874116179, + "loss": 1.1416, + "step": 3327 + }, + { + "epoch": 0.5925925925925926, + "grad_norm": 0.53347247838974, + "learning_rate": 0.00018947003741140821, + "loss": 1.2718, + "step": 3328 + }, + { + "epoch": 0.5927706552706553, + "grad_norm": 0.6473469138145447, + "learning_rate": 0.0001894637843286751, + "loss": 1.2255, + "step": 3329 + }, + { + "epoch": 0.592948717948718, + "grad_norm": 0.4750518798828125, + "learning_rate": 0.00018945752949308498, + "loss": 1.0537, + "step": 3330 + }, + { + "epoch": 0.5931267806267806, + "grad_norm": 0.5636306405067444, + "learning_rate": 0.00018945127290476043, + "loss": 0.9906, + "step": 3331 + }, + { + "epoch": 0.5933048433048433, + "grad_norm": 0.4871736466884613, + "learning_rate": 0.00018944501456382397, + "loss": 1.0549, + "step": 3332 + }, + { + "epoch": 0.593482905982906, + "grad_norm": 0.5554637312889099, + "learning_rate": 0.0001894387544703983, + "loss": 1.1587, + "step": 3333 + }, + { + "epoch": 0.5936609686609686, + "grad_norm": 0.5385799407958984, + "learning_rate": 0.000189432492624606, + "loss": 0.9565, + "step": 3334 + }, + { + "epoch": 0.5938390313390314, + "grad_norm": 0.4996553063392639, + "learning_rate": 0.00018942622902656976, + "loss": 1.0456, + "step": 3335 + }, + { + "epoch": 0.594017094017094, + "grad_norm": 0.46810707449913025, + "learning_rate": 0.00018941996367641237, + "loss": 1.119, + "step": 3336 + }, + { + "epoch": 0.5941951566951567, + "grad_norm": 0.5672653913497925, + "learning_rate": 0.0001894136965742565, + "loss": 1.1317, + "step": 3337 + }, + { + "epoch": 0.5943732193732194, + "grad_norm": 0.4790053367614746, + "learning_rate": 0.00018940742772022504, + "loss": 1.0967, + "step": 3338 + }, + { + "epoch": 0.594551282051282, + "grad_norm": 0.5935906171798706, + "learning_rate": 0.00018940115711444072, + "loss": 1.3044, + "step": 3339 + }, + { + "epoch": 0.5947293447293447, + "grad_norm": 0.4790516793727875, + "learning_rate": 0.00018939488475702647, + "loss": 1.074, + "step": 3340 + }, + { + "epoch": 0.5949074074074074, + "grad_norm": 0.474588006734848, + "learning_rate": 0.00018938861064810516, + "loss": 1.1476, + "step": 3341 + }, + { + "epoch": 0.5950854700854701, + "grad_norm": 0.4908665120601654, + "learning_rate": 0.0001893823347877997, + "loss": 1.216, + "step": 3342 + }, + { + "epoch": 0.5952635327635327, + "grad_norm": 0.531650960445404, + "learning_rate": 0.00018937605717623307, + "loss": 1.1057, + "step": 3343 + }, + { + "epoch": 0.5954415954415955, + "grad_norm": 0.5581082105636597, + "learning_rate": 0.00018936977781352823, + "loss": 0.7972, + "step": 3344 + }, + { + "epoch": 0.5956196581196581, + "grad_norm": 0.42370662093162537, + "learning_rate": 0.00018936349669980827, + "loss": 0.8888, + "step": 3345 + }, + { + "epoch": 0.5957977207977208, + "grad_norm": 0.5817318558692932, + "learning_rate": 0.00018935721383519624, + "loss": 1.2801, + "step": 3346 + }, + { + "epoch": 0.5959757834757835, + "grad_norm": 0.4766376316547394, + "learning_rate": 0.00018935092921981524, + "loss": 1.0918, + "step": 3347 + }, + { + "epoch": 0.5961538461538461, + "grad_norm": 0.5567346811294556, + "learning_rate": 0.00018934464285378836, + "loss": 1.0269, + "step": 3348 + }, + { + "epoch": 0.5963319088319088, + "grad_norm": 0.5285565257072449, + "learning_rate": 0.0001893383547372388, + "loss": 1.1887, + "step": 3349 + }, + { + "epoch": 0.5965099715099715, + "grad_norm": 0.49052694439888, + "learning_rate": 0.00018933206487028979, + "loss": 1.0773, + "step": 3350 + }, + { + "epoch": 0.5966880341880342, + "grad_norm": 0.6175199151039124, + "learning_rate": 0.0001893257732530645, + "loss": 1.0192, + "step": 3351 + }, + { + "epoch": 0.5968660968660968, + "grad_norm": 0.56049644947052, + "learning_rate": 0.00018931947988568628, + "loss": 0.9516, + "step": 3352 + }, + { + "epoch": 0.5970441595441596, + "grad_norm": 0.47873660922050476, + "learning_rate": 0.00018931318476827838, + "loss": 0.8174, + "step": 3353 + }, + { + "epoch": 0.5972222222222222, + "grad_norm": 0.4748854339122772, + "learning_rate": 0.00018930688790096416, + "loss": 1.0238, + "step": 3354 + }, + { + "epoch": 0.5974002849002849, + "grad_norm": 0.5382232666015625, + "learning_rate": 0.00018930058928386698, + "loss": 1.0815, + "step": 3355 + }, + { + "epoch": 0.5975783475783476, + "grad_norm": 0.5038299560546875, + "learning_rate": 0.00018929428891711027, + "loss": 1.0472, + "step": 3356 + }, + { + "epoch": 0.5977564102564102, + "grad_norm": 0.5185908079147339, + "learning_rate": 0.00018928798680081744, + "loss": 1.0435, + "step": 3357 + }, + { + "epoch": 0.5979344729344729, + "grad_norm": 0.5169877409934998, + "learning_rate": 0.00018928168293511202, + "loss": 1.0437, + "step": 3358 + }, + { + "epoch": 0.5981125356125356, + "grad_norm": 0.5218369960784912, + "learning_rate": 0.00018927537732011749, + "loss": 1.082, + "step": 3359 + }, + { + "epoch": 0.5982905982905983, + "grad_norm": 0.5358219742774963, + "learning_rate": 0.0001892690699559574, + "loss": 1.2523, + "step": 3360 + }, + { + "epoch": 0.5984686609686609, + "grad_norm": 0.47716647386550903, + "learning_rate": 0.0001892627608427553, + "loss": 1.2069, + "step": 3361 + }, + { + "epoch": 0.5986467236467237, + "grad_norm": 0.5484169125556946, + "learning_rate": 0.00018925644998063482, + "loss": 1.2016, + "step": 3362 + }, + { + "epoch": 0.5988247863247863, + "grad_norm": 0.46814846992492676, + "learning_rate": 0.00018925013736971965, + "loss": 0.7989, + "step": 3363 + }, + { + "epoch": 0.5990028490028491, + "grad_norm": 0.5391258001327515, + "learning_rate": 0.0001892438230101334, + "loss": 1.224, + "step": 3364 + }, + { + "epoch": 0.5991809116809117, + "grad_norm": 0.5248384475708008, + "learning_rate": 0.00018923750690199987, + "loss": 1.1532, + "step": 3365 + }, + { + "epoch": 0.5993589743589743, + "grad_norm": 0.5074637532234192, + "learning_rate": 0.00018923118904544273, + "loss": 1.0968, + "step": 3366 + }, + { + "epoch": 0.5995370370370371, + "grad_norm": 0.5260029435157776, + "learning_rate": 0.00018922486944058581, + "loss": 1.1311, + "step": 3367 + }, + { + "epoch": 0.5997150997150997, + "grad_norm": 0.48497965931892395, + "learning_rate": 0.00018921854808755294, + "loss": 1.1208, + "step": 3368 + }, + { + "epoch": 0.5998931623931624, + "grad_norm": 0.5108651518821716, + "learning_rate": 0.00018921222498646792, + "loss": 1.147, + "step": 3369 + }, + { + "epoch": 0.6000712250712251, + "grad_norm": 0.5243437886238098, + "learning_rate": 0.00018920590013745471, + "loss": 0.9614, + "step": 3370 + }, + { + "epoch": 0.6002492877492878, + "grad_norm": 0.47022634744644165, + "learning_rate": 0.00018919957354063719, + "loss": 1.0579, + "step": 3371 + }, + { + "epoch": 0.6004273504273504, + "grad_norm": 0.6461413502693176, + "learning_rate": 0.00018919324519613931, + "loss": 1.2126, + "step": 3372 + }, + { + "epoch": 0.6006054131054132, + "grad_norm": 0.4654616713523865, + "learning_rate": 0.00018918691510408508, + "loss": 1.1476, + "step": 3373 + }, + { + "epoch": 0.6007834757834758, + "grad_norm": 0.48571303486824036, + "learning_rate": 0.00018918058326459854, + "loss": 1.2093, + "step": 3374 + }, + { + "epoch": 0.6009615384615384, + "grad_norm": 0.5255016684532166, + "learning_rate": 0.00018917424967780368, + "loss": 1.1538, + "step": 3375 + }, + { + "epoch": 0.6011396011396012, + "grad_norm": 0.5059894323348999, + "learning_rate": 0.00018916791434382468, + "loss": 1.0556, + "step": 3376 + }, + { + "epoch": 0.6013176638176638, + "grad_norm": 0.4581229090690613, + "learning_rate": 0.00018916157726278561, + "loss": 1.1468, + "step": 3377 + }, + { + "epoch": 0.6014957264957265, + "grad_norm": 0.5701818466186523, + "learning_rate": 0.00018915523843481067, + "loss": 1.3641, + "step": 3378 + }, + { + "epoch": 0.6016737891737892, + "grad_norm": 0.5007243752479553, + "learning_rate": 0.00018914889786002403, + "loss": 1.2705, + "step": 3379 + }, + { + "epoch": 0.6018518518518519, + "grad_norm": 0.5192995071411133, + "learning_rate": 0.0001891425555385499, + "loss": 0.9922, + "step": 3380 + }, + { + "epoch": 0.6020299145299145, + "grad_norm": 0.5880612134933472, + "learning_rate": 0.00018913621147051258, + "loss": 0.8783, + "step": 3381 + }, + { + "epoch": 0.6022079772079773, + "grad_norm": 0.5161563158035278, + "learning_rate": 0.0001891298656560364, + "loss": 0.9634, + "step": 3382 + }, + { + "epoch": 0.6023860398860399, + "grad_norm": 0.48450782895088196, + "learning_rate": 0.00018912351809524563, + "loss": 0.809, + "step": 3383 + }, + { + "epoch": 0.6025641025641025, + "grad_norm": 0.621537983417511, + "learning_rate": 0.00018911716878826465, + "loss": 1.2031, + "step": 3384 + }, + { + "epoch": 0.6027421652421653, + "grad_norm": 0.6014544367790222, + "learning_rate": 0.00018911081773521787, + "loss": 1.1552, + "step": 3385 + }, + { + "epoch": 0.6029202279202279, + "grad_norm": 0.49995481967926025, + "learning_rate": 0.00018910446493622976, + "loss": 0.8569, + "step": 3386 + }, + { + "epoch": 0.6030982905982906, + "grad_norm": 0.5157307386398315, + "learning_rate": 0.00018909811039142472, + "loss": 0.9515, + "step": 3387 + }, + { + "epoch": 0.6032763532763533, + "grad_norm": 0.5164140462875366, + "learning_rate": 0.0001890917541009273, + "loss": 0.9803, + "step": 3388 + }, + { + "epoch": 0.603454415954416, + "grad_norm": 0.5555596947669983, + "learning_rate": 0.00018908539606486206, + "loss": 1.2994, + "step": 3389 + }, + { + "epoch": 0.6036324786324786, + "grad_norm": 0.605697512626648, + "learning_rate": 0.00018907903628335353, + "loss": 1.2865, + "step": 3390 + }, + { + "epoch": 0.6038105413105413, + "grad_norm": 0.5700713992118835, + "learning_rate": 0.0001890726747565263, + "loss": 1.2493, + "step": 3391 + }, + { + "epoch": 0.603988603988604, + "grad_norm": 0.5516746044158936, + "learning_rate": 0.0001890663114845051, + "loss": 1.2743, + "step": 3392 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 0.5233162641525269, + "learning_rate": 0.0001890599464674145, + "loss": 0.9237, + "step": 3393 + }, + { + "epoch": 0.6043447293447294, + "grad_norm": 0.5709942579269409, + "learning_rate": 0.00018905357970537925, + "loss": 0.9922, + "step": 3394 + }, + { + "epoch": 0.604522792022792, + "grad_norm": 0.48403796553611755, + "learning_rate": 0.0001890472111985241, + "loss": 1.1255, + "step": 3395 + }, + { + "epoch": 0.6047008547008547, + "grad_norm": 0.628718376159668, + "learning_rate": 0.00018904084094697386, + "loss": 1.1458, + "step": 3396 + }, + { + "epoch": 0.6048789173789174, + "grad_norm": 0.46822869777679443, + "learning_rate": 0.00018903446895085328, + "loss": 0.8727, + "step": 3397 + }, + { + "epoch": 0.60505698005698, + "grad_norm": 0.505584180355072, + "learning_rate": 0.00018902809521028724, + "loss": 1.1595, + "step": 3398 + }, + { + "epoch": 0.6052350427350427, + "grad_norm": 0.4494974911212921, + "learning_rate": 0.00018902171972540058, + "loss": 0.6685, + "step": 3399 + }, + { + "epoch": 0.6054131054131054, + "grad_norm": 0.5101519227027893, + "learning_rate": 0.0001890153424963183, + "loss": 0.9313, + "step": 3400 + }, + { + "epoch": 0.6055911680911681, + "grad_norm": 0.5081079602241516, + "learning_rate": 0.00018900896352316528, + "loss": 1.2588, + "step": 3401 + }, + { + "epoch": 0.6057692307692307, + "grad_norm": 0.5784309506416321, + "learning_rate": 0.00018900258280606653, + "loss": 1.2077, + "step": 3402 + }, + { + "epoch": 0.6059472934472935, + "grad_norm": 0.4506312608718872, + "learning_rate": 0.00018899620034514705, + "loss": 1.05, + "step": 3403 + }, + { + "epoch": 0.6061253561253561, + "grad_norm": 0.5243048071861267, + "learning_rate": 0.0001889898161405319, + "loss": 1.2295, + "step": 3404 + }, + { + "epoch": 0.6063034188034188, + "grad_norm": 0.5447196364402771, + "learning_rate": 0.00018898343019234615, + "loss": 1.1476, + "step": 3405 + }, + { + "epoch": 0.6064814814814815, + "grad_norm": 0.46813663840293884, + "learning_rate": 0.00018897704250071492, + "loss": 1.2113, + "step": 3406 + }, + { + "epoch": 0.6066595441595442, + "grad_norm": 0.5340631604194641, + "learning_rate": 0.00018897065306576342, + "loss": 1.1656, + "step": 3407 + }, + { + "epoch": 0.6068376068376068, + "grad_norm": 0.513708233833313, + "learning_rate": 0.00018896426188761675, + "loss": 1.1616, + "step": 3408 + }, + { + "epoch": 0.6070156695156695, + "grad_norm": 0.594601035118103, + "learning_rate": 0.00018895786896640023, + "loss": 1.2564, + "step": 3409 + }, + { + "epoch": 0.6071937321937322, + "grad_norm": 0.45067599415779114, + "learning_rate": 0.000188951474302239, + "loss": 1.0107, + "step": 3410 + }, + { + "epoch": 0.6073717948717948, + "grad_norm": 0.5394250750541687, + "learning_rate": 0.00018894507789525843, + "loss": 1.4081, + "step": 3411 + }, + { + "epoch": 0.6075498575498576, + "grad_norm": 0.5612049102783203, + "learning_rate": 0.00018893867974558383, + "loss": 1.1015, + "step": 3412 + }, + { + "epoch": 0.6077279202279202, + "grad_norm": 0.4794061779975891, + "learning_rate": 0.00018893227985334056, + "loss": 1.2103, + "step": 3413 + }, + { + "epoch": 0.6079059829059829, + "grad_norm": 0.6060562133789062, + "learning_rate": 0.00018892587821865402, + "loss": 1.3693, + "step": 3414 + }, + { + "epoch": 0.6080840455840456, + "grad_norm": 0.44624534249305725, + "learning_rate": 0.00018891947484164963, + "loss": 0.8209, + "step": 3415 + }, + { + "epoch": 0.6082621082621082, + "grad_norm": 0.49297213554382324, + "learning_rate": 0.0001889130697224528, + "loss": 1.2027, + "step": 3416 + }, + { + "epoch": 0.6084401709401709, + "grad_norm": 0.4431746304035187, + "learning_rate": 0.0001889066628611891, + "loss": 1.0347, + "step": 3417 + }, + { + "epoch": 0.6086182336182336, + "grad_norm": 0.5425933599472046, + "learning_rate": 0.00018890025425798404, + "loss": 1.0556, + "step": 3418 + }, + { + "epoch": 0.6087962962962963, + "grad_norm": 0.5502763390541077, + "learning_rate": 0.00018889384391296315, + "loss": 1.2362, + "step": 3419 + }, + { + "epoch": 0.6089743589743589, + "grad_norm": 0.5442292094230652, + "learning_rate": 0.00018888743182625203, + "loss": 1.1306, + "step": 3420 + }, + { + "epoch": 0.6091524216524217, + "grad_norm": 0.4651123583316803, + "learning_rate": 0.00018888101799797636, + "loss": 0.9305, + "step": 3421 + }, + { + "epoch": 0.6093304843304843, + "grad_norm": 0.4713892340660095, + "learning_rate": 0.00018887460242826177, + "loss": 1.0789, + "step": 3422 + }, + { + "epoch": 0.6095085470085471, + "grad_norm": 0.5283244848251343, + "learning_rate": 0.00018886818511723398, + "loss": 1.345, + "step": 3423 + }, + { + "epoch": 0.6096866096866097, + "grad_norm": 0.5527324080467224, + "learning_rate": 0.0001888617660650187, + "loss": 1.1297, + "step": 3424 + }, + { + "epoch": 0.6098646723646723, + "grad_norm": 0.5412901043891907, + "learning_rate": 0.00018885534527174168, + "loss": 1.1213, + "step": 3425 + }, + { + "epoch": 0.6100427350427351, + "grad_norm": 0.5295354127883911, + "learning_rate": 0.00018884892273752878, + "loss": 1.1217, + "step": 3426 + }, + { + "epoch": 0.6102207977207977, + "grad_norm": 0.461900532245636, + "learning_rate": 0.0001888424984625058, + "loss": 0.827, + "step": 3427 + }, + { + "epoch": 0.6103988603988604, + "grad_norm": 0.4922671616077423, + "learning_rate": 0.00018883607244679865, + "loss": 1.2216, + "step": 3428 + }, + { + "epoch": 0.6105769230769231, + "grad_norm": 0.5080927014350891, + "learning_rate": 0.00018882964469053317, + "loss": 1.2446, + "step": 3429 + }, + { + "epoch": 0.6107549857549858, + "grad_norm": 0.5523943901062012, + "learning_rate": 0.00018882321519383534, + "loss": 1.3346, + "step": 3430 + }, + { + "epoch": 0.6109330484330484, + "grad_norm": 0.5105271935462952, + "learning_rate": 0.0001888167839568311, + "loss": 1.1311, + "step": 3431 + }, + { + "epoch": 0.6111111111111112, + "grad_norm": 0.5635872483253479, + "learning_rate": 0.0001888103509796465, + "loss": 1.1875, + "step": 3432 + }, + { + "epoch": 0.6112891737891738, + "grad_norm": 0.4619547426700592, + "learning_rate": 0.00018880391626240755, + "loss": 0.9176, + "step": 3433 + }, + { + "epoch": 0.6114672364672364, + "grad_norm": 0.5896356105804443, + "learning_rate": 0.00018879747980524034, + "loss": 1.0251, + "step": 3434 + }, + { + "epoch": 0.6116452991452992, + "grad_norm": 0.49062737822532654, + "learning_rate": 0.000188791041608271, + "loss": 1.1598, + "step": 3435 + }, + { + "epoch": 0.6118233618233618, + "grad_norm": 0.45717164874076843, + "learning_rate": 0.00018878460167162558, + "loss": 0.8647, + "step": 3436 + }, + { + "epoch": 0.6120014245014245, + "grad_norm": 0.5903525352478027, + "learning_rate": 0.00018877815999543038, + "loss": 0.9671, + "step": 3437 + }, + { + "epoch": 0.6121794871794872, + "grad_norm": 0.5315384268760681, + "learning_rate": 0.00018877171657981153, + "loss": 1.1759, + "step": 3438 + }, + { + "epoch": 0.6123575498575499, + "grad_norm": 0.5650150775909424, + "learning_rate": 0.0001887652714248953, + "loss": 1.0128, + "step": 3439 + }, + { + "epoch": 0.6125356125356125, + "grad_norm": 0.49841752648353577, + "learning_rate": 0.000188758824530808, + "loss": 1.1259, + "step": 3440 + }, + { + "epoch": 0.6127136752136753, + "grad_norm": 0.4985620975494385, + "learning_rate": 0.00018875237589767593, + "loss": 1.0158, + "step": 3441 + }, + { + "epoch": 0.6128917378917379, + "grad_norm": 0.45266565680503845, + "learning_rate": 0.00018874592552562536, + "loss": 0.93, + "step": 3442 + }, + { + "epoch": 0.6130698005698005, + "grad_norm": 0.5696130990982056, + "learning_rate": 0.00018873947341478274, + "loss": 1.1432, + "step": 3443 + }, + { + "epoch": 0.6132478632478633, + "grad_norm": 0.5211645364761353, + "learning_rate": 0.00018873301956527451, + "loss": 1.1317, + "step": 3444 + }, + { + "epoch": 0.6134259259259259, + "grad_norm": 0.4991866946220398, + "learning_rate": 0.00018872656397722707, + "loss": 1.0362, + "step": 3445 + }, + { + "epoch": 0.6136039886039886, + "grad_norm": 0.5109508037567139, + "learning_rate": 0.00018872010665076694, + "loss": 1.2728, + "step": 3446 + }, + { + "epoch": 0.6137820512820513, + "grad_norm": 0.5838373899459839, + "learning_rate": 0.00018871364758602058, + "loss": 1.1131, + "step": 3447 + }, + { + "epoch": 0.613960113960114, + "grad_norm": 0.5139824151992798, + "learning_rate": 0.00018870718678311462, + "loss": 1.238, + "step": 3448 + }, + { + "epoch": 0.6141381766381766, + "grad_norm": 0.4852082431316376, + "learning_rate": 0.00018870072424217562, + "loss": 1.0677, + "step": 3449 + }, + { + "epoch": 0.6143162393162394, + "grad_norm": 0.5312315225601196, + "learning_rate": 0.00018869425996333018, + "loss": 1.178, + "step": 3450 + }, + { + "epoch": 0.614494301994302, + "grad_norm": 0.6343565583229065, + "learning_rate": 0.00018868779394670492, + "loss": 0.8839, + "step": 3451 + }, + { + "epoch": 0.6146723646723646, + "grad_norm": 0.6029773950576782, + "learning_rate": 0.00018868132619242662, + "loss": 1.1188, + "step": 3452 + }, + { + "epoch": 0.6148504273504274, + "grad_norm": 0.5246016383171082, + "learning_rate": 0.00018867485670062193, + "loss": 1.0797, + "step": 3453 + }, + { + "epoch": 0.61502849002849, + "grad_norm": 0.49307698011398315, + "learning_rate": 0.00018866838547141763, + "loss": 0.9749, + "step": 3454 + }, + { + "epoch": 0.6152065527065527, + "grad_norm": 0.5232903361320496, + "learning_rate": 0.00018866191250494052, + "loss": 1.0785, + "step": 3455 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 0.5545645356178284, + "learning_rate": 0.0001886554378013174, + "loss": 1.0496, + "step": 3456 + }, + { + "epoch": 0.6155626780626781, + "grad_norm": 0.493945837020874, + "learning_rate": 0.00018864896136067515, + "loss": 0.9248, + "step": 3457 + }, + { + "epoch": 0.6157407407407407, + "grad_norm": 0.5223548412322998, + "learning_rate": 0.00018864248318314065, + "loss": 1.0617, + "step": 3458 + }, + { + "epoch": 0.6159188034188035, + "grad_norm": 0.5666514039039612, + "learning_rate": 0.00018863600326884082, + "loss": 0.9981, + "step": 3459 + }, + { + "epoch": 0.6160968660968661, + "grad_norm": 0.4648127257823944, + "learning_rate": 0.00018862952161790265, + "loss": 0.917, + "step": 3460 + }, + { + "epoch": 0.6162749287749287, + "grad_norm": 0.590326189994812, + "learning_rate": 0.0001886230382304531, + "loss": 1.044, + "step": 3461 + }, + { + "epoch": 0.6164529914529915, + "grad_norm": 0.5511625409126282, + "learning_rate": 0.00018861655310661925, + "loss": 1.0988, + "step": 3462 + }, + { + "epoch": 0.6166310541310541, + "grad_norm": 0.567182183265686, + "learning_rate": 0.0001886100662465281, + "loss": 1.3017, + "step": 3463 + }, + { + "epoch": 0.6168091168091168, + "grad_norm": 0.5708897709846497, + "learning_rate": 0.0001886035776503068, + "loss": 0.9123, + "step": 3464 + }, + { + "epoch": 0.6169871794871795, + "grad_norm": 0.4945180416107178, + "learning_rate": 0.0001885970873180824, + "loss": 1.1645, + "step": 3465 + }, + { + "epoch": 0.6171652421652422, + "grad_norm": 0.4713336229324341, + "learning_rate": 0.00018859059524998215, + "loss": 1.0546, + "step": 3466 + }, + { + "epoch": 0.6173433048433048, + "grad_norm": 0.532859206199646, + "learning_rate": 0.0001885841014461332, + "loss": 1.0795, + "step": 3467 + }, + { + "epoch": 0.6175213675213675, + "grad_norm": 0.5165733695030212, + "learning_rate": 0.00018857760590666284, + "loss": 1.1284, + "step": 3468 + }, + { + "epoch": 0.6176994301994302, + "grad_norm": 0.48623126745224, + "learning_rate": 0.00018857110863169826, + "loss": 0.8618, + "step": 3469 + }, + { + "epoch": 0.6178774928774928, + "grad_norm": 0.628559947013855, + "learning_rate": 0.0001885646096213668, + "loss": 1.1089, + "step": 3470 + }, + { + "epoch": 0.6180555555555556, + "grad_norm": 0.503545880317688, + "learning_rate": 0.0001885581088757958, + "loss": 1.2311, + "step": 3471 + }, + { + "epoch": 0.6182336182336182, + "grad_norm": 0.6172101497650146, + "learning_rate": 0.00018855160639511264, + "loss": 1.2651, + "step": 3472 + }, + { + "epoch": 0.6184116809116809, + "grad_norm": 0.49572527408599854, + "learning_rate": 0.00018854510217944465, + "loss": 1.1026, + "step": 3473 + }, + { + "epoch": 0.6185897435897436, + "grad_norm": 0.5373549461364746, + "learning_rate": 0.00018853859622891938, + "loss": 1.2562, + "step": 3474 + }, + { + "epoch": 0.6187678062678063, + "grad_norm": 0.5272396206855774, + "learning_rate": 0.0001885320885436642, + "loss": 1.1763, + "step": 3475 + }, + { + "epoch": 0.6189458689458689, + "grad_norm": 0.46584269404411316, + "learning_rate": 0.00018852557912380665, + "loss": 1.1762, + "step": 3476 + }, + { + "epoch": 0.6191239316239316, + "grad_norm": 0.4798245131969452, + "learning_rate": 0.0001885190679694743, + "loss": 0.9229, + "step": 3477 + }, + { + "epoch": 0.6193019943019943, + "grad_norm": 0.5221366286277771, + "learning_rate": 0.0001885125550807947, + "loss": 1.1078, + "step": 3478 + }, + { + "epoch": 0.6194800569800569, + "grad_norm": 0.5051897168159485, + "learning_rate": 0.0001885060404578954, + "loss": 1.0055, + "step": 3479 + }, + { + "epoch": 0.6196581196581197, + "grad_norm": 0.492662250995636, + "learning_rate": 0.00018849952410090413, + "loss": 1.1172, + "step": 3480 + }, + { + "epoch": 0.6198361823361823, + "grad_norm": 0.4906775951385498, + "learning_rate": 0.00018849300600994853, + "loss": 1.1223, + "step": 3481 + }, + { + "epoch": 0.6200142450142451, + "grad_norm": 0.5032641291618347, + "learning_rate": 0.0001884864861851563, + "loss": 0.9541, + "step": 3482 + }, + { + "epoch": 0.6201923076923077, + "grad_norm": 0.5262296795845032, + "learning_rate": 0.00018847996462665521, + "loss": 1.021, + "step": 3483 + }, + { + "epoch": 0.6203703703703703, + "grad_norm": 0.5253522992134094, + "learning_rate": 0.00018847344133457295, + "loss": 0.9075, + "step": 3484 + }, + { + "epoch": 0.6205484330484331, + "grad_norm": 0.4204299747943878, + "learning_rate": 0.00018846691630903744, + "loss": 0.895, + "step": 3485 + }, + { + "epoch": 0.6207264957264957, + "grad_norm": 0.557604193687439, + "learning_rate": 0.0001884603895501765, + "loss": 1.1758, + "step": 3486 + }, + { + "epoch": 0.6209045584045584, + "grad_norm": 0.5981321930885315, + "learning_rate": 0.00018845386105811795, + "loss": 1.1087, + "step": 3487 + }, + { + "epoch": 0.6210826210826211, + "grad_norm": 0.5285581946372986, + "learning_rate": 0.00018844733083298975, + "loss": 1.0692, + "step": 3488 + }, + { + "epoch": 0.6212606837606838, + "grad_norm": 0.5403170585632324, + "learning_rate": 0.00018844079887491986, + "loss": 1.1998, + "step": 3489 + }, + { + "epoch": 0.6214387464387464, + "grad_norm": 0.5471615791320801, + "learning_rate": 0.0001884342651840362, + "loss": 0.9556, + "step": 3490 + }, + { + "epoch": 0.6216168091168092, + "grad_norm": 0.6126871705055237, + "learning_rate": 0.00018842772976046686, + "loss": 1.2629, + "step": 3491 + }, + { + "epoch": 0.6217948717948718, + "grad_norm": 0.45669353008270264, + "learning_rate": 0.00018842119260433982, + "loss": 1.0203, + "step": 3492 + }, + { + "epoch": 0.6219729344729344, + "grad_norm": 0.4998520612716675, + "learning_rate": 0.0001884146537157832, + "loss": 1.0271, + "step": 3493 + }, + { + "epoch": 0.6221509971509972, + "grad_norm": 0.5820242166519165, + "learning_rate": 0.00018840811309492507, + "loss": 1.0321, + "step": 3494 + }, + { + "epoch": 0.6223290598290598, + "grad_norm": 0.581676185131073, + "learning_rate": 0.00018840157074189367, + "loss": 0.9219, + "step": 3495 + }, + { + "epoch": 0.6225071225071225, + "grad_norm": 0.6044120788574219, + "learning_rate": 0.0001883950266568171, + "loss": 1.1621, + "step": 3496 + }, + { + "epoch": 0.6226851851851852, + "grad_norm": 0.5448858737945557, + "learning_rate": 0.0001883884808398236, + "loss": 1.0686, + "step": 3497 + }, + { + "epoch": 0.6228632478632479, + "grad_norm": 0.4921551048755646, + "learning_rate": 0.00018838193329104143, + "loss": 1.2259, + "step": 3498 + }, + { + "epoch": 0.6230413105413105, + "grad_norm": 0.5374335646629333, + "learning_rate": 0.00018837538401059888, + "loss": 1.2608, + "step": 3499 + }, + { + "epoch": 0.6232193732193733, + "grad_norm": 0.5123008489608765, + "learning_rate": 0.0001883688329986243, + "loss": 0.8682, + "step": 3500 + }, + { + "epoch": 0.6233974358974359, + "grad_norm": 0.566145122051239, + "learning_rate": 0.00018836228025524595, + "loss": 1.1807, + "step": 3501 + }, + { + "epoch": 0.6235754985754985, + "grad_norm": 0.6658587455749512, + "learning_rate": 0.00018835572578059233, + "loss": 1.1641, + "step": 3502 + }, + { + "epoch": 0.6237535612535613, + "grad_norm": 0.4992465078830719, + "learning_rate": 0.00018834916957479177, + "loss": 0.9125, + "step": 3503 + }, + { + "epoch": 0.6239316239316239, + "grad_norm": 0.5081812739372253, + "learning_rate": 0.00018834261163797278, + "loss": 1.0939, + "step": 3504 + }, + { + "epoch": 0.6241096866096866, + "grad_norm": 0.5168607234954834, + "learning_rate": 0.0001883360519702638, + "loss": 1.2382, + "step": 3505 + }, + { + "epoch": 0.6242877492877493, + "grad_norm": 0.5517697334289551, + "learning_rate": 0.00018832949057179344, + "loss": 1.206, + "step": 3506 + }, + { + "epoch": 0.624465811965812, + "grad_norm": 0.4505497217178345, + "learning_rate": 0.00018832292744269013, + "loss": 0.8485, + "step": 3507 + }, + { + "epoch": 0.6246438746438746, + "grad_norm": 0.5230690240859985, + "learning_rate": 0.0001883163625830826, + "loss": 1.1701, + "step": 3508 + }, + { + "epoch": 0.6248219373219374, + "grad_norm": 0.5062205195426941, + "learning_rate": 0.00018830979599309937, + "loss": 1.0602, + "step": 3509 + }, + { + "epoch": 0.625, + "grad_norm": 0.49922460317611694, + "learning_rate": 0.00018830322767286913, + "loss": 1.1937, + "step": 3510 + }, + { + "epoch": 0.6251780626780626, + "grad_norm": 0.4637366831302643, + "learning_rate": 0.0001882966576225206, + "loss": 1.038, + "step": 3511 + }, + { + "epoch": 0.6253561253561254, + "grad_norm": 0.5330080389976501, + "learning_rate": 0.00018829008584218246, + "loss": 0.9308, + "step": 3512 + }, + { + "epoch": 0.625534188034188, + "grad_norm": 0.5443428754806519, + "learning_rate": 0.0001882835123319835, + "loss": 1.0006, + "step": 3513 + }, + { + "epoch": 0.6257122507122507, + "grad_norm": 0.5534018874168396, + "learning_rate": 0.00018827693709205253, + "loss": 1.2383, + "step": 3514 + }, + { + "epoch": 0.6258903133903134, + "grad_norm": 0.49207547307014465, + "learning_rate": 0.00018827036012251832, + "loss": 0.9804, + "step": 3515 + }, + { + "epoch": 0.6260683760683761, + "grad_norm": 0.4900086224079132, + "learning_rate": 0.0001882637814235098, + "loss": 1.012, + "step": 3516 + }, + { + "epoch": 0.6262464387464387, + "grad_norm": 0.5267475247383118, + "learning_rate": 0.00018825720099515585, + "loss": 1.1104, + "step": 3517 + }, + { + "epoch": 0.6264245014245015, + "grad_norm": 0.5711902379989624, + "learning_rate": 0.00018825061883758534, + "loss": 1.0616, + "step": 3518 + }, + { + "epoch": 0.6266025641025641, + "grad_norm": 0.5007771849632263, + "learning_rate": 0.0001882440349509273, + "loss": 0.9578, + "step": 3519 + }, + { + "epoch": 0.6267806267806267, + "grad_norm": 0.5657192468643188, + "learning_rate": 0.00018823744933531075, + "loss": 1.2768, + "step": 3520 + }, + { + "epoch": 0.6269586894586895, + "grad_norm": 0.6077173352241516, + "learning_rate": 0.00018823086199086462, + "loss": 1.147, + "step": 3521 + }, + { + "epoch": 0.6271367521367521, + "grad_norm": 0.5114718079566956, + "learning_rate": 0.000188224272917718, + "loss": 1.1176, + "step": 3522 + }, + { + "epoch": 0.6273148148148148, + "grad_norm": 0.4831676185131073, + "learning_rate": 0.0001882176821160001, + "loss": 0.8021, + "step": 3523 + }, + { + "epoch": 0.6274928774928775, + "grad_norm": 0.6327390670776367, + "learning_rate": 0.00018821108958583994, + "loss": 0.9449, + "step": 3524 + }, + { + "epoch": 0.6276709401709402, + "grad_norm": 0.5541796684265137, + "learning_rate": 0.00018820449532736672, + "loss": 1.2018, + "step": 3525 + }, + { + "epoch": 0.6278490028490028, + "grad_norm": 0.5224639773368835, + "learning_rate": 0.00018819789934070968, + "loss": 1.0138, + "step": 3526 + }, + { + "epoch": 0.6280270655270656, + "grad_norm": 0.49359360337257385, + "learning_rate": 0.00018819130162599798, + "loss": 1.0768, + "step": 3527 + }, + { + "epoch": 0.6282051282051282, + "grad_norm": 0.5525050759315491, + "learning_rate": 0.00018818470218336092, + "loss": 1.0883, + "step": 3528 + }, + { + "epoch": 0.6283831908831908, + "grad_norm": 0.5563427209854126, + "learning_rate": 0.00018817810101292787, + "loss": 1.1491, + "step": 3529 + }, + { + "epoch": 0.6285612535612536, + "grad_norm": 0.49363306164741516, + "learning_rate": 0.00018817149811482803, + "loss": 1.1409, + "step": 3530 + }, + { + "epoch": 0.6287393162393162, + "grad_norm": 0.5102340579032898, + "learning_rate": 0.00018816489348919086, + "loss": 1.1914, + "step": 3531 + }, + { + "epoch": 0.6289173789173789, + "grad_norm": 0.5173332691192627, + "learning_rate": 0.00018815828713614576, + "loss": 0.9308, + "step": 3532 + }, + { + "epoch": 0.6290954415954416, + "grad_norm": 0.5093010067939758, + "learning_rate": 0.00018815167905582216, + "loss": 0.9429, + "step": 3533 + }, + { + "epoch": 0.6292735042735043, + "grad_norm": 0.5453153848648071, + "learning_rate": 0.00018814506924834954, + "loss": 1.0147, + "step": 3534 + }, + { + "epoch": 0.6294515669515669, + "grad_norm": 0.5850773453712463, + "learning_rate": 0.00018813845771385737, + "loss": 1.3372, + "step": 3535 + }, + { + "epoch": 0.6296296296296297, + "grad_norm": 0.5095621943473816, + "learning_rate": 0.00018813184445247525, + "loss": 1.0515, + "step": 3536 + }, + { + "epoch": 0.6298076923076923, + "grad_norm": 0.6216054558753967, + "learning_rate": 0.00018812522946433266, + "loss": 0.8703, + "step": 3537 + }, + { + "epoch": 0.6299857549857549, + "grad_norm": 0.4945531189441681, + "learning_rate": 0.00018811861274955932, + "loss": 1.1485, + "step": 3538 + }, + { + "epoch": 0.6301638176638177, + "grad_norm": 0.47882601618766785, + "learning_rate": 0.00018811199430828477, + "loss": 1.1107, + "step": 3539 + }, + { + "epoch": 0.6303418803418803, + "grad_norm": 0.5005326867103577, + "learning_rate": 0.00018810537414063876, + "loss": 1.0237, + "step": 3540 + }, + { + "epoch": 0.6305199430199431, + "grad_norm": 0.5382370352745056, + "learning_rate": 0.00018809875224675093, + "loss": 0.9965, + "step": 3541 + }, + { + "epoch": 0.6306980056980057, + "grad_norm": 0.47002625465393066, + "learning_rate": 0.0001880921286267511, + "loss": 1.065, + "step": 3542 + }, + { + "epoch": 0.6308760683760684, + "grad_norm": 0.4519105851650238, + "learning_rate": 0.00018808550328076897, + "loss": 0.9312, + "step": 3543 + }, + { + "epoch": 0.6310541310541311, + "grad_norm": 0.45360881090164185, + "learning_rate": 0.0001880788762089344, + "loss": 1.0739, + "step": 3544 + }, + { + "epoch": 0.6312321937321937, + "grad_norm": 0.5578218698501587, + "learning_rate": 0.00018807224741137723, + "loss": 1.2478, + "step": 3545 + }, + { + "epoch": 0.6314102564102564, + "grad_norm": 0.4838615655899048, + "learning_rate": 0.0001880656168882273, + "loss": 1.0221, + "step": 3546 + }, + { + "epoch": 0.6315883190883191, + "grad_norm": 0.5733556747436523, + "learning_rate": 0.0001880589846396146, + "loss": 1.1249, + "step": 3547 + }, + { + "epoch": 0.6317663817663818, + "grad_norm": 0.4939686954021454, + "learning_rate": 0.00018805235066566894, + "loss": 0.8559, + "step": 3548 + }, + { + "epoch": 0.6319444444444444, + "grad_norm": 0.5072234869003296, + "learning_rate": 0.00018804571496652044, + "loss": 1.0842, + "step": 3549 + }, + { + "epoch": 0.6321225071225072, + "grad_norm": 0.4640493392944336, + "learning_rate": 0.00018803907754229903, + "loss": 1.0728, + "step": 3550 + }, + { + "epoch": 0.6323005698005698, + "grad_norm": 0.5314788818359375, + "learning_rate": 0.00018803243839313481, + "loss": 1.0752, + "step": 3551 + }, + { + "epoch": 0.6324786324786325, + "grad_norm": 0.5511462092399597, + "learning_rate": 0.0001880257975191578, + "loss": 1.0238, + "step": 3552 + }, + { + "epoch": 0.6326566951566952, + "grad_norm": 0.4980711042881012, + "learning_rate": 0.00018801915492049816, + "loss": 1.0981, + "step": 3553 + }, + { + "epoch": 0.6328347578347578, + "grad_norm": 0.7746123671531677, + "learning_rate": 0.00018801251059728604, + "loss": 1.0968, + "step": 3554 + }, + { + "epoch": 0.6330128205128205, + "grad_norm": 0.5006106495857239, + "learning_rate": 0.00018800586454965155, + "loss": 1.1802, + "step": 3555 + }, + { + "epoch": 0.6331908831908832, + "grad_norm": 0.49427780508995056, + "learning_rate": 0.000187999216777725, + "loss": 1.1257, + "step": 3556 + }, + { + "epoch": 0.6333689458689459, + "grad_norm": 0.5484146475791931, + "learning_rate": 0.00018799256728163662, + "loss": 1.1344, + "step": 3557 + }, + { + "epoch": 0.6335470085470085, + "grad_norm": 0.5007877349853516, + "learning_rate": 0.00018798591606151662, + "loss": 1.1328, + "step": 3558 + }, + { + "epoch": 0.6337250712250713, + "grad_norm": 0.5068148970603943, + "learning_rate": 0.00018797926311749544, + "loss": 0.976, + "step": 3559 + }, + { + "epoch": 0.6339031339031339, + "grad_norm": 0.44936859607696533, + "learning_rate": 0.00018797260844970334, + "loss": 0.9735, + "step": 3560 + }, + { + "epoch": 0.6340811965811965, + "grad_norm": 0.4592931866645813, + "learning_rate": 0.0001879659520582707, + "loss": 1.1306, + "step": 3561 + }, + { + "epoch": 0.6342592592592593, + "grad_norm": 0.4664020836353302, + "learning_rate": 0.00018795929394332795, + "loss": 1.0577, + "step": 3562 + }, + { + "epoch": 0.6344373219373219, + "grad_norm": 0.5638116002082825, + "learning_rate": 0.00018795263410500556, + "loss": 1.1747, + "step": 3563 + }, + { + "epoch": 0.6346153846153846, + "grad_norm": 0.524736225605011, + "learning_rate": 0.00018794597254343401, + "loss": 0.8964, + "step": 3564 + }, + { + "epoch": 0.6347934472934473, + "grad_norm": 0.4645404517650604, + "learning_rate": 0.00018793930925874386, + "loss": 0.8673, + "step": 3565 + }, + { + "epoch": 0.63497150997151, + "grad_norm": 0.4800064265727997, + "learning_rate": 0.00018793264425106558, + "loss": 1.0334, + "step": 3566 + }, + { + "epoch": 0.6351495726495726, + "grad_norm": 0.6202501058578491, + "learning_rate": 0.0001879259775205298, + "loss": 1.1061, + "step": 3567 + }, + { + "epoch": 0.6353276353276354, + "grad_norm": 0.503383457660675, + "learning_rate": 0.00018791930906726718, + "loss": 0.8545, + "step": 3568 + }, + { + "epoch": 0.635505698005698, + "grad_norm": 0.5256780982017517, + "learning_rate": 0.00018791263889140832, + "loss": 1.0785, + "step": 3569 + }, + { + "epoch": 0.6356837606837606, + "grad_norm": 0.47562023997306824, + "learning_rate": 0.00018790596699308392, + "loss": 1.0041, + "step": 3570 + }, + { + "epoch": 0.6358618233618234, + "grad_norm": 0.5103238224983215, + "learning_rate": 0.00018789929337242469, + "loss": 1.1488, + "step": 3571 + }, + { + "epoch": 0.636039886039886, + "grad_norm": 0.5023695826530457, + "learning_rate": 0.0001878926180295614, + "loss": 1.0696, + "step": 3572 + }, + { + "epoch": 0.6362179487179487, + "grad_norm": 0.5302290916442871, + "learning_rate": 0.00018788594096462487, + "loss": 1.0554, + "step": 3573 + }, + { + "epoch": 0.6363960113960114, + "grad_norm": 0.4798361361026764, + "learning_rate": 0.00018787926217774588, + "loss": 0.8872, + "step": 3574 + }, + { + "epoch": 0.6365740740740741, + "grad_norm": 0.5529209971427917, + "learning_rate": 0.00018787258166905527, + "loss": 1.0976, + "step": 3575 + }, + { + "epoch": 0.6367521367521367, + "grad_norm": 0.49757125973701477, + "learning_rate": 0.00018786589943868402, + "loss": 1.0049, + "step": 3576 + }, + { + "epoch": 0.6369301994301995, + "grad_norm": 0.5497848391532898, + "learning_rate": 0.00018785921548676295, + "loss": 1.2272, + "step": 3577 + }, + { + "epoch": 0.6371082621082621, + "grad_norm": 0.5061752200126648, + "learning_rate": 0.0001878525298134231, + "loss": 1.0307, + "step": 3578 + }, + { + "epoch": 0.6372863247863247, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00018784584241879538, + "loss": 1.1064, + "step": 3579 + }, + { + "epoch": 0.6374643874643875, + "grad_norm": 0.48312774300575256, + "learning_rate": 0.0001878391533030109, + "loss": 1.078, + "step": 3580 + }, + { + "epoch": 0.6376424501424501, + "grad_norm": 0.5059898495674133, + "learning_rate": 0.00018783246246620067, + "loss": 1.0922, + "step": 3581 + }, + { + "epoch": 0.6378205128205128, + "grad_norm": 0.5144124031066895, + "learning_rate": 0.00018782576990849581, + "loss": 1.0909, + "step": 3582 + }, + { + "epoch": 0.6379985754985755, + "grad_norm": 0.5535032153129578, + "learning_rate": 0.0001878190756300274, + "loss": 1.2579, + "step": 3583 + }, + { + "epoch": 0.6381766381766382, + "grad_norm": 0.49145692586898804, + "learning_rate": 0.00018781237963092667, + "loss": 1.0823, + "step": 3584 + }, + { + "epoch": 0.6383547008547008, + "grad_norm": 0.5245576500892639, + "learning_rate": 0.00018780568191132472, + "loss": 0.9595, + "step": 3585 + }, + { + "epoch": 0.6385327635327636, + "grad_norm": 0.5026637315750122, + "learning_rate": 0.00018779898247135287, + "loss": 1.153, + "step": 3586 + }, + { + "epoch": 0.6387108262108262, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018779228131114234, + "loss": 1.0661, + "step": 3587 + }, + { + "epoch": 0.6388888888888888, + "grad_norm": 0.517387330532074, + "learning_rate": 0.00018778557843082444, + "loss": 1.0113, + "step": 3588 + }, + { + "epoch": 0.6390669515669516, + "grad_norm": 0.5149948000907898, + "learning_rate": 0.00018777887383053047, + "loss": 0.9483, + "step": 3589 + }, + { + "epoch": 0.6392450142450142, + "grad_norm": 0.4854544997215271, + "learning_rate": 0.00018777216751039185, + "loss": 1.22, + "step": 3590 + }, + { + "epoch": 0.6394230769230769, + "grad_norm": 0.5317271947860718, + "learning_rate": 0.0001877654594705399, + "loss": 1.2483, + "step": 3591 + }, + { + "epoch": 0.6396011396011396, + "grad_norm": 0.4554755687713623, + "learning_rate": 0.0001877587497111061, + "loss": 0.9864, + "step": 3592 + }, + { + "epoch": 0.6397792022792023, + "grad_norm": 0.4833736717700958, + "learning_rate": 0.0001877520382322219, + "loss": 0.8895, + "step": 3593 + }, + { + "epoch": 0.6399572649572649, + "grad_norm": 0.5018072724342346, + "learning_rate": 0.00018774532503401878, + "loss": 1.2523, + "step": 3594 + }, + { + "epoch": 0.6401353276353277, + "grad_norm": 0.4478762447834015, + "learning_rate": 0.00018773861011662832, + "loss": 0.8833, + "step": 3595 + }, + { + "epoch": 0.6403133903133903, + "grad_norm": 0.5686985850334167, + "learning_rate": 0.00018773189348018205, + "loss": 0.9934, + "step": 3596 + }, + { + "epoch": 0.6404914529914529, + "grad_norm": 0.5144175291061401, + "learning_rate": 0.00018772517512481157, + "loss": 0.8149, + "step": 3597 + }, + { + "epoch": 0.6406695156695157, + "grad_norm": 0.5359936356544495, + "learning_rate": 0.00018771845505064852, + "loss": 1.1822, + "step": 3598 + }, + { + "epoch": 0.6408475783475783, + "grad_norm": 0.532573938369751, + "learning_rate": 0.00018771173325782457, + "loss": 1.0361, + "step": 3599 + }, + { + "epoch": 0.6410256410256411, + "grad_norm": 0.46121537685394287, + "learning_rate": 0.00018770500974647138, + "loss": 1.0792, + "step": 3600 + }, + { + "epoch": 0.6412037037037037, + "grad_norm": 0.4804821312427521, + "learning_rate": 0.00018769828451672076, + "loss": 1.1119, + "step": 3601 + }, + { + "epoch": 0.6413817663817664, + "grad_norm": 0.4955114722251892, + "learning_rate": 0.00018769155756870443, + "loss": 0.9312, + "step": 3602 + }, + { + "epoch": 0.6415598290598291, + "grad_norm": 0.4987298250198364, + "learning_rate": 0.00018768482890255415, + "loss": 1.2326, + "step": 3603 + }, + { + "epoch": 0.6417378917378918, + "grad_norm": 0.47216179966926575, + "learning_rate": 0.0001876780985184018, + "loss": 1.0114, + "step": 3604 + }, + { + "epoch": 0.6419159544159544, + "grad_norm": 0.5891931653022766, + "learning_rate": 0.0001876713664163793, + "loss": 1.2963, + "step": 3605 + }, + { + "epoch": 0.6420940170940171, + "grad_norm": 0.4645081162452698, + "learning_rate": 0.00018766463259661846, + "loss": 1.0874, + "step": 3606 + }, + { + "epoch": 0.6422720797720798, + "grad_norm": 0.5275476574897766, + "learning_rate": 0.00018765789705925125, + "loss": 0.9453, + "step": 3607 + }, + { + "epoch": 0.6424501424501424, + "grad_norm": 0.5884957313537598, + "learning_rate": 0.00018765115980440964, + "loss": 1.0796, + "step": 3608 + }, + { + "epoch": 0.6426282051282052, + "grad_norm": 0.4843178987503052, + "learning_rate": 0.00018764442083222567, + "loss": 1.1657, + "step": 3609 + }, + { + "epoch": 0.6428062678062678, + "grad_norm": 0.5188381671905518, + "learning_rate": 0.00018763768014283126, + "loss": 1.1109, + "step": 3610 + }, + { + "epoch": 0.6429843304843305, + "grad_norm": 0.4101468324661255, + "learning_rate": 0.00018763093773635863, + "loss": 0.895, + "step": 3611 + }, + { + "epoch": 0.6431623931623932, + "grad_norm": 0.4552084505558014, + "learning_rate": 0.00018762419361293979, + "loss": 0.9418, + "step": 3612 + }, + { + "epoch": 0.6433404558404558, + "grad_norm": 0.5924661159515381, + "learning_rate": 0.0001876174477727069, + "loss": 1.2562, + "step": 3613 + }, + { + "epoch": 0.6435185185185185, + "grad_norm": 0.5072348713874817, + "learning_rate": 0.00018761070021579212, + "loss": 1.1501, + "step": 3614 + }, + { + "epoch": 0.6436965811965812, + "grad_norm": 0.5312697887420654, + "learning_rate": 0.0001876039509423277, + "loss": 1.0751, + "step": 3615 + }, + { + "epoch": 0.6438746438746439, + "grad_norm": 0.6046462059020996, + "learning_rate": 0.0001875971999524458, + "loss": 1.0927, + "step": 3616 + }, + { + "epoch": 0.6440527065527065, + "grad_norm": 0.4992375373840332, + "learning_rate": 0.00018759044724627876, + "loss": 0.96, + "step": 3617 + }, + { + "epoch": 0.6442307692307693, + "grad_norm": 0.4983134865760803, + "learning_rate": 0.00018758369282395886, + "loss": 1.0599, + "step": 3618 + }, + { + "epoch": 0.6444088319088319, + "grad_norm": 0.5655683279037476, + "learning_rate": 0.00018757693668561843, + "loss": 1.2372, + "step": 3619 + }, + { + "epoch": 0.6445868945868946, + "grad_norm": 0.4968827962875366, + "learning_rate": 0.00018757017883138985, + "loss": 1.1639, + "step": 3620 + }, + { + "epoch": 0.6447649572649573, + "grad_norm": 0.5831420421600342, + "learning_rate": 0.00018756341926140553, + "loss": 0.9002, + "step": 3621 + }, + { + "epoch": 0.64494301994302, + "grad_norm": 0.4828467071056366, + "learning_rate": 0.0001875566579757979, + "loss": 0.9201, + "step": 3622 + }, + { + "epoch": 0.6451210826210826, + "grad_norm": 0.5067087411880493, + "learning_rate": 0.00018754989497469943, + "loss": 0.9874, + "step": 3623 + }, + { + "epoch": 0.6452991452991453, + "grad_norm": 0.5182318091392517, + "learning_rate": 0.00018754313025824267, + "loss": 1.1291, + "step": 3624 + }, + { + "epoch": 0.645477207977208, + "grad_norm": 0.472200483083725, + "learning_rate": 0.0001875363638265601, + "loss": 1.0286, + "step": 3625 + }, + { + "epoch": 0.6456552706552706, + "grad_norm": 0.4597308039665222, + "learning_rate": 0.0001875295956797843, + "loss": 0.7517, + "step": 3626 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 0.5358221530914307, + "learning_rate": 0.00018752282581804798, + "loss": 1.2264, + "step": 3627 + }, + { + "epoch": 0.646011396011396, + "grad_norm": 0.5268992781639099, + "learning_rate": 0.00018751605424148363, + "loss": 1.0801, + "step": 3628 + }, + { + "epoch": 0.6461894586894587, + "grad_norm": 0.5917379260063171, + "learning_rate": 0.00018750928095022403, + "loss": 0.9538, + "step": 3629 + }, + { + "epoch": 0.6463675213675214, + "grad_norm": 0.44506707787513733, + "learning_rate": 0.00018750250594440183, + "loss": 0.9818, + "step": 3630 + }, + { + "epoch": 0.646545584045584, + "grad_norm": 0.5578880906105042, + "learning_rate": 0.00018749572922414982, + "loss": 0.9958, + "step": 3631 + }, + { + "epoch": 0.6467236467236467, + "grad_norm": 0.5155318975448608, + "learning_rate": 0.00018748895078960076, + "loss": 1.2888, + "step": 3632 + }, + { + "epoch": 0.6469017094017094, + "grad_norm": 0.5117297768592834, + "learning_rate": 0.0001874821706408874, + "loss": 1.0452, + "step": 3633 + }, + { + "epoch": 0.6470797720797721, + "grad_norm": 0.5169841647148132, + "learning_rate": 0.00018747538877814267, + "loss": 1.1649, + "step": 3634 + }, + { + "epoch": 0.6472578347578347, + "grad_norm": 0.5001181960105896, + "learning_rate": 0.00018746860520149942, + "loss": 1.1472, + "step": 3635 + }, + { + "epoch": 0.6474358974358975, + "grad_norm": 0.6289856433868408, + "learning_rate": 0.00018746181991109056, + "loss": 1.0351, + "step": 3636 + }, + { + "epoch": 0.6476139601139601, + "grad_norm": 0.5490612983703613, + "learning_rate": 0.00018745503290704897, + "loss": 0.8938, + "step": 3637 + }, + { + "epoch": 0.6477920227920227, + "grad_norm": 0.47378283739089966, + "learning_rate": 0.00018744824418950775, + "loss": 0.937, + "step": 3638 + }, + { + "epoch": 0.6479700854700855, + "grad_norm": 0.6079059839248657, + "learning_rate": 0.0001874414537585998, + "loss": 1.0486, + "step": 3639 + }, + { + "epoch": 0.6481481481481481, + "grad_norm": 0.5351769924163818, + "learning_rate": 0.00018743466161445823, + "loss": 1.0316, + "step": 3640 + }, + { + "epoch": 0.6483262108262108, + "grad_norm": 0.5516425967216492, + "learning_rate": 0.0001874278677572161, + "loss": 1.1552, + "step": 3641 + }, + { + "epoch": 0.6485042735042735, + "grad_norm": 0.5027523636817932, + "learning_rate": 0.0001874210721870065, + "loss": 1.0491, + "step": 3642 + }, + { + "epoch": 0.6486823361823362, + "grad_norm": 0.5596168041229248, + "learning_rate": 0.00018741427490396258, + "loss": 1.0256, + "step": 3643 + }, + { + "epoch": 0.6488603988603988, + "grad_norm": 0.5601046681404114, + "learning_rate": 0.00018740747590821751, + "loss": 1.1604, + "step": 3644 + }, + { + "epoch": 0.6490384615384616, + "grad_norm": 0.49749523401260376, + "learning_rate": 0.0001874006751999046, + "loss": 1.0532, + "step": 3645 + }, + { + "epoch": 0.6492165242165242, + "grad_norm": 0.6226113438606262, + "learning_rate": 0.00018739387277915697, + "loss": 1.1402, + "step": 3646 + }, + { + "epoch": 0.6493945868945868, + "grad_norm": 0.6142009496688843, + "learning_rate": 0.00018738706864610794, + "loss": 1.2437, + "step": 3647 + }, + { + "epoch": 0.6495726495726496, + "grad_norm": 0.48814916610717773, + "learning_rate": 0.00018738026280089084, + "loss": 0.8429, + "step": 3648 + }, + { + "epoch": 0.6497507122507122, + "grad_norm": 0.5717982053756714, + "learning_rate": 0.00018737345524363902, + "loss": 1.1095, + "step": 3649 + }, + { + "epoch": 0.6499287749287749, + "grad_norm": 0.5150009989738464, + "learning_rate": 0.00018736664597448582, + "loss": 1.199, + "step": 3650 + }, + { + "epoch": 0.6501068376068376, + "grad_norm": 0.58461594581604, + "learning_rate": 0.00018735983499356472, + "loss": 1.0704, + "step": 3651 + }, + { + "epoch": 0.6502849002849003, + "grad_norm": 0.5108643770217896, + "learning_rate": 0.0001873530223010091, + "loss": 1.2039, + "step": 3652 + }, + { + "epoch": 0.6504629629629629, + "grad_norm": 0.513306736946106, + "learning_rate": 0.00018734620789695247, + "loss": 1.1448, + "step": 3653 + }, + { + "epoch": 0.6506410256410257, + "grad_norm": 0.5139986872673035, + "learning_rate": 0.00018733939178152835, + "loss": 1.0023, + "step": 3654 + }, + { + "epoch": 0.6508190883190883, + "grad_norm": 0.5187703967094421, + "learning_rate": 0.00018733257395487027, + "loss": 1.1304, + "step": 3655 + }, + { + "epoch": 0.6509971509971509, + "grad_norm": 0.5470501184463501, + "learning_rate": 0.00018732575441711183, + "loss": 1.0272, + "step": 3656 + }, + { + "epoch": 0.6511752136752137, + "grad_norm": 0.537309467792511, + "learning_rate": 0.00018731893316838665, + "loss": 1.0806, + "step": 3657 + }, + { + "epoch": 0.6513532763532763, + "grad_norm": 0.5187864899635315, + "learning_rate": 0.00018731211020882836, + "loss": 1.0154, + "step": 3658 + }, + { + "epoch": 0.6515313390313391, + "grad_norm": 0.48373252153396606, + "learning_rate": 0.00018730528553857062, + "loss": 1.0135, + "step": 3659 + }, + { + "epoch": 0.6517094017094017, + "grad_norm": 0.5645000338554382, + "learning_rate": 0.00018729845915774716, + "loss": 0.8924, + "step": 3660 + }, + { + "epoch": 0.6518874643874644, + "grad_norm": 0.5722129940986633, + "learning_rate": 0.00018729163106649178, + "loss": 1.2416, + "step": 3661 + }, + { + "epoch": 0.6520655270655271, + "grad_norm": 0.5904877185821533, + "learning_rate": 0.00018728480126493823, + "loss": 0.9792, + "step": 3662 + }, + { + "epoch": 0.6522435897435898, + "grad_norm": 0.5224713087081909, + "learning_rate": 0.00018727796975322026, + "loss": 1.079, + "step": 3663 + }, + { + "epoch": 0.6524216524216524, + "grad_norm": 0.5667217969894409, + "learning_rate": 0.00018727113653147184, + "loss": 1.1397, + "step": 3664 + }, + { + "epoch": 0.6525997150997151, + "grad_norm": 0.5274622440338135, + "learning_rate": 0.00018726430159982677, + "loss": 1.0569, + "step": 3665 + }, + { + "epoch": 0.6527777777777778, + "grad_norm": 0.5745310187339783, + "learning_rate": 0.00018725746495841896, + "loss": 1.2129, + "step": 3666 + }, + { + "epoch": 0.6529558404558404, + "grad_norm": 0.6123398542404175, + "learning_rate": 0.0001872506266073824, + "loss": 1.186, + "step": 3667 + }, + { + "epoch": 0.6531339031339032, + "grad_norm": 0.4983387291431427, + "learning_rate": 0.00018724378654685106, + "loss": 1.1957, + "step": 3668 + }, + { + "epoch": 0.6533119658119658, + "grad_norm": 0.5584192276000977, + "learning_rate": 0.00018723694477695897, + "loss": 1.0939, + "step": 3669 + }, + { + "epoch": 0.6534900284900285, + "grad_norm": 0.5318745374679565, + "learning_rate": 0.00018723010129784016, + "loss": 1.1869, + "step": 3670 + }, + { + "epoch": 0.6536680911680912, + "grad_norm": 0.4607617259025574, + "learning_rate": 0.0001872232561096287, + "loss": 0.8447, + "step": 3671 + }, + { + "epoch": 0.6538461538461539, + "grad_norm": 0.5312213897705078, + "learning_rate": 0.00018721640921245874, + "loss": 1.0623, + "step": 3672 + }, + { + "epoch": 0.6540242165242165, + "grad_norm": 0.5099136233329773, + "learning_rate": 0.0001872095606064644, + "loss": 0.7174, + "step": 3673 + }, + { + "epoch": 0.6542022792022792, + "grad_norm": 0.6894404888153076, + "learning_rate": 0.0001872027102917799, + "loss": 1.0251, + "step": 3674 + }, + { + "epoch": 0.6543803418803419, + "grad_norm": 0.5758535861968994, + "learning_rate": 0.00018719585826853944, + "loss": 1.1655, + "step": 3675 + }, + { + "epoch": 0.6545584045584045, + "grad_norm": 0.521824061870575, + "learning_rate": 0.0001871890045368773, + "loss": 1.1653, + "step": 3676 + }, + { + "epoch": 0.6547364672364673, + "grad_norm": 0.5370712280273438, + "learning_rate": 0.00018718214909692771, + "loss": 1.3152, + "step": 3677 + }, + { + "epoch": 0.6549145299145299, + "grad_norm": 0.4459827244281769, + "learning_rate": 0.000187175291948825, + "loss": 1.0953, + "step": 3678 + }, + { + "epoch": 0.6550925925925926, + "grad_norm": 0.44131460785865784, + "learning_rate": 0.00018716843309270353, + "loss": 0.8568, + "step": 3679 + }, + { + "epoch": 0.6552706552706553, + "grad_norm": 0.5529624819755554, + "learning_rate": 0.00018716157252869772, + "loss": 1.2085, + "step": 3680 + }, + { + "epoch": 0.655448717948718, + "grad_norm": 0.44604751467704773, + "learning_rate": 0.00018715471025694194, + "loss": 0.9605, + "step": 3681 + }, + { + "epoch": 0.6556267806267806, + "grad_norm": 0.4662449359893799, + "learning_rate": 0.0001871478462775707, + "loss": 1.2092, + "step": 3682 + }, + { + "epoch": 0.6558048433048433, + "grad_norm": 0.42632922530174255, + "learning_rate": 0.0001871409805907184, + "loss": 0.9141, + "step": 3683 + }, + { + "epoch": 0.655982905982906, + "grad_norm": 0.534009575843811, + "learning_rate": 0.00018713411319651958, + "loss": 1.0147, + "step": 3684 + }, + { + "epoch": 0.6561609686609686, + "grad_norm": 0.5433241724967957, + "learning_rate": 0.00018712724409510888, + "loss": 1.1998, + "step": 3685 + }, + { + "epoch": 0.6563390313390314, + "grad_norm": 0.4771319627761841, + "learning_rate": 0.0001871203732866208, + "loss": 1.0384, + "step": 3686 + }, + { + "epoch": 0.656517094017094, + "grad_norm": 0.507641077041626, + "learning_rate": 0.00018711350077119, + "loss": 0.9608, + "step": 3687 + }, + { + "epoch": 0.6566951566951567, + "grad_norm": 0.5069413185119629, + "learning_rate": 0.00018710662654895108, + "loss": 1.055, + "step": 3688 + }, + { + "epoch": 0.6568732193732194, + "grad_norm": 0.512340247631073, + "learning_rate": 0.00018709975062003876, + "loss": 0.9506, + "step": 3689 + }, + { + "epoch": 0.657051282051282, + "grad_norm": 0.5156390070915222, + "learning_rate": 0.00018709287298458778, + "loss": 1.0089, + "step": 3690 + }, + { + "epoch": 0.6572293447293447, + "grad_norm": 0.5101696252822876, + "learning_rate": 0.0001870859936427329, + "loss": 1.0441, + "step": 3691 + }, + { + "epoch": 0.6574074074074074, + "grad_norm": 0.4394689202308655, + "learning_rate": 0.00018707911259460884, + "loss": 0.9124, + "step": 3692 + }, + { + "epoch": 0.6575854700854701, + "grad_norm": 0.4842554032802582, + "learning_rate": 0.00018707222984035043, + "loss": 1.0051, + "step": 3693 + }, + { + "epoch": 0.6577635327635327, + "grad_norm": 0.6418108344078064, + "learning_rate": 0.00018706534538009262, + "loss": 1.1165, + "step": 3694 + }, + { + "epoch": 0.6579415954415955, + "grad_norm": 0.5596832036972046, + "learning_rate": 0.00018705845921397022, + "loss": 1.1127, + "step": 3695 + }, + { + "epoch": 0.6581196581196581, + "grad_norm": 0.6692909002304077, + "learning_rate": 0.00018705157134211813, + "loss": 1.2403, + "step": 3696 + }, + { + "epoch": 0.6582977207977208, + "grad_norm": 0.5046468377113342, + "learning_rate": 0.00018704468176467134, + "loss": 1.1016, + "step": 3697 + }, + { + "epoch": 0.6584757834757835, + "grad_norm": 0.6723586320877075, + "learning_rate": 0.00018703779048176485, + "loss": 1.1777, + "step": 3698 + }, + { + "epoch": 0.6586538461538461, + "grad_norm": 0.5269754528999329, + "learning_rate": 0.00018703089749353365, + "loss": 1.1441, + "step": 3699 + }, + { + "epoch": 0.6588319088319088, + "grad_norm": 0.5303323268890381, + "learning_rate": 0.0001870240028001128, + "loss": 1.07, + "step": 3700 + }, + { + "epoch": 0.6590099715099715, + "grad_norm": 0.4795511066913605, + "learning_rate": 0.00018701710640163738, + "loss": 1.0189, + "step": 3701 + }, + { + "epoch": 0.6591880341880342, + "grad_norm": 0.514659583568573, + "learning_rate": 0.00018701020829824255, + "loss": 1.0792, + "step": 3702 + }, + { + "epoch": 0.6593660968660968, + "grad_norm": 0.5407463312149048, + "learning_rate": 0.0001870033084900634, + "loss": 0.9346, + "step": 3703 + }, + { + "epoch": 0.6595441595441596, + "grad_norm": 0.5358424186706543, + "learning_rate": 0.0001869964069772352, + "loss": 1.1242, + "step": 3704 + }, + { + "epoch": 0.6597222222222222, + "grad_norm": 0.470825731754303, + "learning_rate": 0.00018698950375989307, + "loss": 0.9952, + "step": 3705 + }, + { + "epoch": 0.6599002849002849, + "grad_norm": 0.5711592435836792, + "learning_rate": 0.00018698259883817236, + "loss": 1.1678, + "step": 3706 + }, + { + "epoch": 0.6600783475783476, + "grad_norm": 0.5298995971679688, + "learning_rate": 0.00018697569221220832, + "loss": 0.869, + "step": 3707 + }, + { + "epoch": 0.6602564102564102, + "grad_norm": 0.5453875064849854, + "learning_rate": 0.00018696878388213626, + "loss": 0.9706, + "step": 3708 + }, + { + "epoch": 0.6604344729344729, + "grad_norm": 0.6219926476478577, + "learning_rate": 0.00018696187384809154, + "loss": 1.1902, + "step": 3709 + }, + { + "epoch": 0.6606125356125356, + "grad_norm": 0.5972491502761841, + "learning_rate": 0.00018695496211020953, + "loss": 1.2054, + "step": 3710 + }, + { + "epoch": 0.6607905982905983, + "grad_norm": 0.5048904418945312, + "learning_rate": 0.0001869480486686257, + "loss": 1.0405, + "step": 3711 + }, + { + "epoch": 0.6609686609686609, + "grad_norm": 0.5474200248718262, + "learning_rate": 0.00018694113352347546, + "loss": 1.09, + "step": 3712 + }, + { + "epoch": 0.6611467236467237, + "grad_norm": 0.5073318481445312, + "learning_rate": 0.00018693421667489432, + "loss": 1.0698, + "step": 3713 + }, + { + "epoch": 0.6613247863247863, + "grad_norm": 0.5693208575248718, + "learning_rate": 0.0001869272981230178, + "loss": 0.9664, + "step": 3714 + }, + { + "epoch": 0.6615028490028491, + "grad_norm": 0.5678503513336182, + "learning_rate": 0.00018692037786798143, + "loss": 1.0895, + "step": 3715 + }, + { + "epoch": 0.6616809116809117, + "grad_norm": 0.4950976073741913, + "learning_rate": 0.00018691345590992082, + "loss": 0.9584, + "step": 3716 + }, + { + "epoch": 0.6618589743589743, + "grad_norm": 0.4944666624069214, + "learning_rate": 0.0001869065322489716, + "loss": 0.8607, + "step": 3717 + }, + { + "epoch": 0.6620370370370371, + "grad_norm": 0.5197804570198059, + "learning_rate": 0.0001868996068852694, + "loss": 1.2335, + "step": 3718 + }, + { + "epoch": 0.6622150997150997, + "grad_norm": 0.6550365686416626, + "learning_rate": 0.00018689267981894994, + "loss": 1.0441, + "step": 3719 + }, + { + "epoch": 0.6623931623931624, + "grad_norm": 0.5331503748893738, + "learning_rate": 0.00018688575105014888, + "loss": 1.1696, + "step": 3720 + }, + { + "epoch": 0.6625712250712251, + "grad_norm": 0.47304239869117737, + "learning_rate": 0.00018687882057900207, + "loss": 0.9695, + "step": 3721 + }, + { + "epoch": 0.6627492877492878, + "grad_norm": 0.5653772354125977, + "learning_rate": 0.00018687188840564524, + "loss": 1.2082, + "step": 3722 + }, + { + "epoch": 0.6629273504273504, + "grad_norm": 0.5323491096496582, + "learning_rate": 0.00018686495453021417, + "loss": 0.9106, + "step": 3723 + }, + { + "epoch": 0.6631054131054132, + "grad_norm": 0.5612817406654358, + "learning_rate": 0.00018685801895284483, + "loss": 1.1302, + "step": 3724 + }, + { + "epoch": 0.6632834757834758, + "grad_norm": 0.4562164545059204, + "learning_rate": 0.000186851081673673, + "loss": 0.8886, + "step": 3725 + }, + { + "epoch": 0.6634615384615384, + "grad_norm": 0.5006430745124817, + "learning_rate": 0.00018684414269283463, + "loss": 0.9128, + "step": 3726 + }, + { + "epoch": 0.6636396011396012, + "grad_norm": 0.5305442810058594, + "learning_rate": 0.0001868372020104657, + "loss": 1.1766, + "step": 3727 + }, + { + "epoch": 0.6638176638176638, + "grad_norm": 0.6129274368286133, + "learning_rate": 0.0001868302596267022, + "loss": 1.04, + "step": 3728 + }, + { + "epoch": 0.6639957264957265, + "grad_norm": 0.5530399084091187, + "learning_rate": 0.00018682331554168013, + "loss": 1.4114, + "step": 3729 + }, + { + "epoch": 0.6641737891737892, + "grad_norm": 0.5397193431854248, + "learning_rate": 0.00018681636975553557, + "loss": 1.1945, + "step": 3730 + }, + { + "epoch": 0.6643518518518519, + "grad_norm": 0.5510205030441284, + "learning_rate": 0.00018680942226840456, + "loss": 1.0489, + "step": 3731 + }, + { + "epoch": 0.6645299145299145, + "grad_norm": 0.5519221425056458, + "learning_rate": 0.00018680247308042324, + "loss": 1.1633, + "step": 3732 + }, + { + "epoch": 0.6647079772079773, + "grad_norm": 0.4848768711090088, + "learning_rate": 0.00018679552219172784, + "loss": 0.8716, + "step": 3733 + }, + { + "epoch": 0.6648860398860399, + "grad_norm": 0.5490246415138245, + "learning_rate": 0.0001867885696024544, + "loss": 1.1347, + "step": 3734 + }, + { + "epoch": 0.6650641025641025, + "grad_norm": 0.5281458497047424, + "learning_rate": 0.00018678161531273928, + "loss": 1.0987, + "step": 3735 + }, + { + "epoch": 0.6652421652421653, + "grad_norm": 0.5313079953193665, + "learning_rate": 0.00018677465932271867, + "loss": 0.9705, + "step": 3736 + }, + { + "epoch": 0.6654202279202279, + "grad_norm": 0.5425750017166138, + "learning_rate": 0.0001867677016325289, + "loss": 1.1847, + "step": 3737 + }, + { + "epoch": 0.6655982905982906, + "grad_norm": 0.5796298980712891, + "learning_rate": 0.0001867607422423062, + "loss": 1.2639, + "step": 3738 + }, + { + "epoch": 0.6657763532763533, + "grad_norm": 0.49738675355911255, + "learning_rate": 0.00018675378115218702, + "loss": 1.0536, + "step": 3739 + }, + { + "epoch": 0.665954415954416, + "grad_norm": 0.665250301361084, + "learning_rate": 0.0001867468183623077, + "loss": 1.2836, + "step": 3740 + }, + { + "epoch": 0.6661324786324786, + "grad_norm": 0.5184717178344727, + "learning_rate": 0.00018673985387280469, + "loss": 1.0497, + "step": 3741 + }, + { + "epoch": 0.6663105413105413, + "grad_norm": 0.5129656791687012, + "learning_rate": 0.00018673288768381442, + "loss": 1.2041, + "step": 3742 + }, + { + "epoch": 0.666488603988604, + "grad_norm": 0.5308768153190613, + "learning_rate": 0.00018672591979547337, + "loss": 1.2092, + "step": 3743 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.5059141516685486, + "learning_rate": 0.00018671895020791812, + "loss": 1.1929, + "step": 3744 + }, + { + "epoch": 0.6668447293447294, + "grad_norm": 0.5237857103347778, + "learning_rate": 0.00018671197892128517, + "loss": 1.2538, + "step": 3745 + }, + { + "epoch": 0.667022792022792, + "grad_norm": 0.450000137090683, + "learning_rate": 0.0001867050059357111, + "loss": 0.7138, + "step": 3746 + }, + { + "epoch": 0.6672008547008547, + "grad_norm": 0.5413795709609985, + "learning_rate": 0.00018669803125133258, + "loss": 1.1383, + "step": 3747 + }, + { + "epoch": 0.6673789173789174, + "grad_norm": 0.4657825529575348, + "learning_rate": 0.00018669105486828622, + "loss": 1.0518, + "step": 3748 + }, + { + "epoch": 0.66755698005698, + "grad_norm": 0.6198551654815674, + "learning_rate": 0.00018668407678670875, + "loss": 1.2697, + "step": 3749 + }, + { + "epoch": 0.6677350427350427, + "grad_norm": 0.5112186074256897, + "learning_rate": 0.00018667709700673685, + "loss": 0.9907, + "step": 3750 + }, + { + "epoch": 0.6679131054131054, + "grad_norm": 0.5446593761444092, + "learning_rate": 0.00018667011552850728, + "loss": 1.0708, + "step": 3751 + }, + { + "epoch": 0.6680911680911681, + "grad_norm": 0.5673866271972656, + "learning_rate": 0.00018666313235215682, + "loss": 1.05, + "step": 3752 + }, + { + "epoch": 0.6682692307692307, + "grad_norm": 0.4821988046169281, + "learning_rate": 0.00018665614747782235, + "loss": 1.0543, + "step": 3753 + }, + { + "epoch": 0.6684472934472935, + "grad_norm": 0.5158842206001282, + "learning_rate": 0.00018664916090564067, + "loss": 1.0331, + "step": 3754 + }, + { + "epoch": 0.6686253561253561, + "grad_norm": 0.45486921072006226, + "learning_rate": 0.00018664217263574865, + "loss": 0.9262, + "step": 3755 + }, + { + "epoch": 0.6688034188034188, + "grad_norm": 0.46193036437034607, + "learning_rate": 0.00018663518266828327, + "loss": 0.9858, + "step": 3756 + }, + { + "epoch": 0.6689814814814815, + "grad_norm": 0.5144094824790955, + "learning_rate": 0.00018662819100338148, + "loss": 1.0302, + "step": 3757 + }, + { + "epoch": 0.6691595441595442, + "grad_norm": 0.5246134400367737, + "learning_rate": 0.0001866211976411802, + "loss": 1.064, + "step": 3758 + }, + { + "epoch": 0.6693376068376068, + "grad_norm": 0.4853166937828064, + "learning_rate": 0.0001866142025818165, + "loss": 0.9481, + "step": 3759 + }, + { + "epoch": 0.6695156695156695, + "grad_norm": 0.5029586553573608, + "learning_rate": 0.00018660720582542743, + "loss": 0.9443, + "step": 3760 + }, + { + "epoch": 0.6696937321937322, + "grad_norm": 0.5373172163963318, + "learning_rate": 0.0001866002073721501, + "loss": 1.1401, + "step": 3761 + }, + { + "epoch": 0.6698717948717948, + "grad_norm": 0.6236287951469421, + "learning_rate": 0.00018659320722212158, + "loss": 1.1255, + "step": 3762 + }, + { + "epoch": 0.6700498575498576, + "grad_norm": 0.5470684766769409, + "learning_rate": 0.00018658620537547903, + "loss": 1.0622, + "step": 3763 + }, + { + "epoch": 0.6702279202279202, + "grad_norm": 0.63177090883255, + "learning_rate": 0.00018657920183235964, + "loss": 0.9736, + "step": 3764 + }, + { + "epoch": 0.6704059829059829, + "grad_norm": 0.5456309914588928, + "learning_rate": 0.00018657219659290068, + "loss": 1.027, + "step": 3765 + }, + { + "epoch": 0.6705840455840456, + "grad_norm": 0.4816138744354248, + "learning_rate": 0.00018656518965723935, + "loss": 0.7801, + "step": 3766 + }, + { + "epoch": 0.6707621082621082, + "grad_norm": 0.4811640679836273, + "learning_rate": 0.00018655818102551294, + "loss": 1.0535, + "step": 3767 + }, + { + "epoch": 0.6709401709401709, + "grad_norm": 0.4677673280239105, + "learning_rate": 0.00018655117069785884, + "loss": 1.1043, + "step": 3768 + }, + { + "epoch": 0.6711182336182336, + "grad_norm": 0.5628635883331299, + "learning_rate": 0.0001865441586744143, + "loss": 1.0392, + "step": 3769 + }, + { + "epoch": 0.6712962962962963, + "grad_norm": 0.5484504103660583, + "learning_rate": 0.00018653714495531673, + "loss": 1.1533, + "step": 3770 + }, + { + "epoch": 0.6714743589743589, + "grad_norm": 0.5830571055412292, + "learning_rate": 0.0001865301295407036, + "loss": 1.2479, + "step": 3771 + }, + { + "epoch": 0.6716524216524217, + "grad_norm": 0.5516841411590576, + "learning_rate": 0.00018652311243071235, + "loss": 1.2152, + "step": 3772 + }, + { + "epoch": 0.6718304843304843, + "grad_norm": 0.6360766291618347, + "learning_rate": 0.0001865160936254804, + "loss": 1.0752, + "step": 3773 + }, + { + "epoch": 0.6720085470085471, + "grad_norm": 0.6038610935211182, + "learning_rate": 0.00018650907312514533, + "loss": 1.2425, + "step": 3774 + }, + { + "epoch": 0.6721866096866097, + "grad_norm": 0.49572908878326416, + "learning_rate": 0.0001865020509298447, + "loss": 1.0057, + "step": 3775 + }, + { + "epoch": 0.6723646723646723, + "grad_norm": 0.4551616311073303, + "learning_rate": 0.00018649502703971607, + "loss": 1.0763, + "step": 3776 + }, + { + "epoch": 0.6725427350427351, + "grad_norm": 0.6621482372283936, + "learning_rate": 0.00018648800145489706, + "loss": 1.0306, + "step": 3777 + }, + { + "epoch": 0.6727207977207977, + "grad_norm": 0.5523806810379028, + "learning_rate": 0.0001864809741755253, + "loss": 0.9906, + "step": 3778 + }, + { + "epoch": 0.6728988603988604, + "grad_norm": 0.5527048110961914, + "learning_rate": 0.00018647394520173856, + "loss": 1.0734, + "step": 3779 + }, + { + "epoch": 0.6730769230769231, + "grad_norm": 0.573573887348175, + "learning_rate": 0.00018646691453367444, + "loss": 1.1409, + "step": 3780 + }, + { + "epoch": 0.6732549857549858, + "grad_norm": 0.6273239254951477, + "learning_rate": 0.00018645988217147079, + "loss": 0.9682, + "step": 3781 + }, + { + "epoch": 0.6734330484330484, + "grad_norm": 0.4917762279510498, + "learning_rate": 0.00018645284811526534, + "loss": 0.9681, + "step": 3782 + }, + { + "epoch": 0.6736111111111112, + "grad_norm": 0.4901154339313507, + "learning_rate": 0.0001864458123651959, + "loss": 1.1828, + "step": 3783 + }, + { + "epoch": 0.6737891737891738, + "grad_norm": 0.6292546391487122, + "learning_rate": 0.00018643877492140036, + "loss": 1.1987, + "step": 3784 + }, + { + "epoch": 0.6739672364672364, + "grad_norm": 0.5334137678146362, + "learning_rate": 0.0001864317357840166, + "loss": 1.0347, + "step": 3785 + }, + { + "epoch": 0.6741452991452992, + "grad_norm": 0.6064338684082031, + "learning_rate": 0.0001864246949531825, + "loss": 1.4154, + "step": 3786 + }, + { + "epoch": 0.6743233618233618, + "grad_norm": 0.5442034602165222, + "learning_rate": 0.000186417652429036, + "loss": 1.2604, + "step": 3787 + }, + { + "epoch": 0.6745014245014245, + "grad_norm": 0.490858793258667, + "learning_rate": 0.00018641060821171518, + "loss": 1.1511, + "step": 3788 + }, + { + "epoch": 0.6746794871794872, + "grad_norm": 0.571116030216217, + "learning_rate": 0.00018640356230135798, + "loss": 1.1479, + "step": 3789 + }, + { + "epoch": 0.6748575498575499, + "grad_norm": 0.4857785105705261, + "learning_rate": 0.00018639651469810247, + "loss": 0.9, + "step": 3790 + }, + { + "epoch": 0.6750356125356125, + "grad_norm": 0.5320703983306885, + "learning_rate": 0.0001863894654020867, + "loss": 1.2284, + "step": 3791 + }, + { + "epoch": 0.6752136752136753, + "grad_norm": 0.5586925745010376, + "learning_rate": 0.0001863824144134488, + "loss": 1.1183, + "step": 3792 + }, + { + "epoch": 0.6753917378917379, + "grad_norm": 0.47740885615348816, + "learning_rate": 0.000186375361732327, + "loss": 1.1512, + "step": 3793 + }, + { + "epoch": 0.6755698005698005, + "grad_norm": 0.5867732167243958, + "learning_rate": 0.00018636830735885935, + "loss": 1.1903, + "step": 3794 + }, + { + "epoch": 0.6757478632478633, + "grad_norm": 0.5013887882232666, + "learning_rate": 0.0001863612512931842, + "loss": 0.8581, + "step": 3795 + }, + { + "epoch": 0.6759259259259259, + "grad_norm": 0.6026871204376221, + "learning_rate": 0.0001863541935354397, + "loss": 0.9581, + "step": 3796 + }, + { + "epoch": 0.6761039886039886, + "grad_norm": 0.5238468647003174, + "learning_rate": 0.00018634713408576415, + "loss": 1.0949, + "step": 3797 + }, + { + "epoch": 0.6762820512820513, + "grad_norm": 0.5128598213195801, + "learning_rate": 0.00018634007294429585, + "loss": 0.8992, + "step": 3798 + }, + { + "epoch": 0.676460113960114, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018633301011117324, + "loss": 1.0793, + "step": 3799 + }, + { + "epoch": 0.6766381766381766, + "grad_norm": 0.592566728591919, + "learning_rate": 0.00018632594558653457, + "loss": 1.3242, + "step": 3800 + }, + { + "epoch": 0.6768162393162394, + "grad_norm": 0.4953067898750305, + "learning_rate": 0.0001863188793705184, + "loss": 0.9925, + "step": 3801 + }, + { + "epoch": 0.676994301994302, + "grad_norm": 0.4989747107028961, + "learning_rate": 0.00018631181146326305, + "loss": 1.0677, + "step": 3802 + }, + { + "epoch": 0.6771723646723646, + "grad_norm": 0.5375261902809143, + "learning_rate": 0.00018630474186490705, + "loss": 1.0556, + "step": 3803 + }, + { + "epoch": 0.6773504273504274, + "grad_norm": 0.6512624025344849, + "learning_rate": 0.00018629767057558894, + "loss": 1.2041, + "step": 3804 + }, + { + "epoch": 0.67752849002849, + "grad_norm": 0.5428260564804077, + "learning_rate": 0.00018629059759544723, + "loss": 0.9645, + "step": 3805 + }, + { + "epoch": 0.6777065527065527, + "grad_norm": 0.5598662495613098, + "learning_rate": 0.00018628352292462052, + "loss": 1.1683, + "step": 3806 + }, + { + "epoch": 0.6778846153846154, + "grad_norm": 0.49351340532302856, + "learning_rate": 0.0001862764465632474, + "loss": 1.1622, + "step": 3807 + }, + { + "epoch": 0.6780626780626781, + "grad_norm": 0.4796701669692993, + "learning_rate": 0.00018626936851146657, + "loss": 1.0017, + "step": 3808 + }, + { + "epoch": 0.6782407407407407, + "grad_norm": 0.444533109664917, + "learning_rate": 0.00018626228876941664, + "loss": 0.9145, + "step": 3809 + }, + { + "epoch": 0.6784188034188035, + "grad_norm": 0.5197392702102661, + "learning_rate": 0.00018625520733723635, + "loss": 1.283, + "step": 3810 + }, + { + "epoch": 0.6785968660968661, + "grad_norm": 0.48785829544067383, + "learning_rate": 0.00018624812421506447, + "loss": 1.1084, + "step": 3811 + }, + { + "epoch": 0.6787749287749287, + "grad_norm": 0.5083680152893066, + "learning_rate": 0.00018624103940303974, + "loss": 0.9071, + "step": 3812 + }, + { + "epoch": 0.6789529914529915, + "grad_norm": 0.553819477558136, + "learning_rate": 0.00018623395290130103, + "loss": 0.9986, + "step": 3813 + }, + { + "epoch": 0.6791310541310541, + "grad_norm": 0.5347508788108826, + "learning_rate": 0.00018622686470998713, + "loss": 1.0148, + "step": 3814 + }, + { + "epoch": 0.6793091168091168, + "grad_norm": 0.5080769062042236, + "learning_rate": 0.00018621977482923693, + "loss": 1.0169, + "step": 3815 + }, + { + "epoch": 0.6794871794871795, + "grad_norm": 0.5444077849388123, + "learning_rate": 0.00018621268325918938, + "loss": 1.172, + "step": 3816 + }, + { + "epoch": 0.6796652421652422, + "grad_norm": 0.521946132183075, + "learning_rate": 0.00018620558999998335, + "loss": 1.0247, + "step": 3817 + }, + { + "epoch": 0.6798433048433048, + "grad_norm": 0.5257413983345032, + "learning_rate": 0.00018619849505175786, + "loss": 1.1574, + "step": 3818 + }, + { + "epoch": 0.6800213675213675, + "grad_norm": 0.5473007559776306, + "learning_rate": 0.00018619139841465193, + "loss": 1.1254, + "step": 3819 + }, + { + "epoch": 0.6801994301994302, + "grad_norm": 0.5479872226715088, + "learning_rate": 0.00018618430008880463, + "loss": 1.0196, + "step": 3820 + }, + { + "epoch": 0.6803774928774928, + "grad_norm": 0.5918973088264465, + "learning_rate": 0.00018617720007435497, + "loss": 1.082, + "step": 3821 + }, + { + "epoch": 0.6805555555555556, + "grad_norm": 0.5411791801452637, + "learning_rate": 0.0001861700983714421, + "loss": 0.7723, + "step": 3822 + }, + { + "epoch": 0.6807336182336182, + "grad_norm": 0.5466326475143433, + "learning_rate": 0.00018616299498020516, + "loss": 1.0979, + "step": 3823 + }, + { + "epoch": 0.6809116809116809, + "grad_norm": 0.5405182838439941, + "learning_rate": 0.00018615588990078332, + "loss": 0.8891, + "step": 3824 + }, + { + "epoch": 0.6810897435897436, + "grad_norm": 0.5415780544281006, + "learning_rate": 0.00018614878313331579, + "loss": 1.0927, + "step": 3825 + }, + { + "epoch": 0.6812678062678063, + "grad_norm": 0.5284909605979919, + "learning_rate": 0.00018614167467794182, + "loss": 1.0684, + "step": 3826 + }, + { + "epoch": 0.6814458689458689, + "grad_norm": 0.4873995780944824, + "learning_rate": 0.00018613456453480062, + "loss": 1.1653, + "step": 3827 + }, + { + "epoch": 0.6816239316239316, + "grad_norm": 0.5506551265716553, + "learning_rate": 0.0001861274527040316, + "loss": 0.9876, + "step": 3828 + }, + { + "epoch": 0.6818019943019943, + "grad_norm": 0.5031297206878662, + "learning_rate": 0.0001861203391857741, + "loss": 1.067, + "step": 3829 + }, + { + "epoch": 0.6819800569800569, + "grad_norm": 0.622346043586731, + "learning_rate": 0.0001861132239801674, + "loss": 1.1514, + "step": 3830 + }, + { + "epoch": 0.6821581196581197, + "grad_norm": 0.47706183791160583, + "learning_rate": 0.000186106107087351, + "loss": 0.9857, + "step": 3831 + }, + { + "epoch": 0.6823361823361823, + "grad_norm": 0.5082845091819763, + "learning_rate": 0.00018609898850746424, + "loss": 1.123, + "step": 3832 + }, + { + "epoch": 0.6825142450142451, + "grad_norm": 0.5119805932044983, + "learning_rate": 0.00018609186824064671, + "loss": 1.1386, + "step": 3833 + }, + { + "epoch": 0.6826923076923077, + "grad_norm": 0.5247541069984436, + "learning_rate": 0.00018608474628703788, + "loss": 0.9433, + "step": 3834 + }, + { + "epoch": 0.6828703703703703, + "grad_norm": 0.4618282616138458, + "learning_rate": 0.00018607762264677722, + "loss": 0.8727, + "step": 3835 + }, + { + "epoch": 0.6830484330484331, + "grad_norm": 0.6014040112495422, + "learning_rate": 0.00018607049732000436, + "loss": 1.1823, + "step": 3836 + }, + { + "epoch": 0.6832264957264957, + "grad_norm": 0.6489043831825256, + "learning_rate": 0.00018606337030685892, + "loss": 1.1466, + "step": 3837 + }, + { + "epoch": 0.6834045584045584, + "grad_norm": 0.5527763366699219, + "learning_rate": 0.00018605624160748053, + "loss": 1.3015, + "step": 3838 + }, + { + "epoch": 0.6835826210826211, + "grad_norm": 0.5628284215927124, + "learning_rate": 0.0001860491112220088, + "loss": 1.1504, + "step": 3839 + }, + { + "epoch": 0.6837606837606838, + "grad_norm": 0.5414566993713379, + "learning_rate": 0.00018604197915058355, + "loss": 1.0155, + "step": 3840 + }, + { + "epoch": 0.6839387464387464, + "grad_norm": 0.5378929376602173, + "learning_rate": 0.00018603484539334443, + "loss": 0.8917, + "step": 3841 + }, + { + "epoch": 0.6841168091168092, + "grad_norm": 0.5953748822212219, + "learning_rate": 0.00018602770995043125, + "loss": 1.1971, + "step": 3842 + }, + { + "epoch": 0.6842948717948718, + "grad_norm": 0.511813759803772, + "learning_rate": 0.00018602057282198376, + "loss": 1.1345, + "step": 3843 + }, + { + "epoch": 0.6844729344729344, + "grad_norm": 0.5145484209060669, + "learning_rate": 0.00018601343400814185, + "loss": 1.0786, + "step": 3844 + }, + { + "epoch": 0.6846509971509972, + "grad_norm": 0.5199604034423828, + "learning_rate": 0.00018600629350904542, + "loss": 1.2063, + "step": 3845 + }, + { + "epoch": 0.6848290598290598, + "grad_norm": 0.5653825998306274, + "learning_rate": 0.0001859991513248343, + "loss": 1.0314, + "step": 3846 + }, + { + "epoch": 0.6850071225071225, + "grad_norm": 0.5660843849182129, + "learning_rate": 0.00018599200745564843, + "loss": 1.2754, + "step": 3847 + }, + { + "epoch": 0.6851851851851852, + "grad_norm": 0.5225719809532166, + "learning_rate": 0.00018598486190162788, + "loss": 1.0837, + "step": 3848 + }, + { + "epoch": 0.6853632478632479, + "grad_norm": 0.5011669397354126, + "learning_rate": 0.00018597771466291252, + "loss": 1.1, + "step": 3849 + }, + { + "epoch": 0.6855413105413105, + "grad_norm": 0.5923115015029907, + "learning_rate": 0.00018597056573964245, + "loss": 1.1875, + "step": 3850 + }, + { + "epoch": 0.6857193732193733, + "grad_norm": 0.5666482448577881, + "learning_rate": 0.00018596341513195776, + "loss": 1.1663, + "step": 3851 + }, + { + "epoch": 0.6858974358974359, + "grad_norm": 0.5396790504455566, + "learning_rate": 0.0001859562628399985, + "loss": 1.1179, + "step": 3852 + }, + { + "epoch": 0.6860754985754985, + "grad_norm": 0.5709532499313354, + "learning_rate": 0.00018594910886390485, + "loss": 1.0369, + "step": 3853 + }, + { + "epoch": 0.6862535612535613, + "grad_norm": 0.45524322986602783, + "learning_rate": 0.00018594195320381692, + "loss": 1.0171, + "step": 3854 + }, + { + "epoch": 0.6864316239316239, + "grad_norm": 0.6130724549293518, + "learning_rate": 0.00018593479585987498, + "loss": 1.1944, + "step": 3855 + }, + { + "epoch": 0.6866096866096866, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.0001859276368322192, + "loss": 1.2567, + "step": 3856 + }, + { + "epoch": 0.6867877492877493, + "grad_norm": 0.49919846653938293, + "learning_rate": 0.00018592047612098992, + "loss": 0.9459, + "step": 3857 + }, + { + "epoch": 0.686965811965812, + "grad_norm": 0.5776857733726501, + "learning_rate": 0.00018591331372632734, + "loss": 1.2456, + "step": 3858 + }, + { + "epoch": 0.6871438746438746, + "grad_norm": 0.4740692377090454, + "learning_rate": 0.00018590614964837188, + "loss": 1.0401, + "step": 3859 + }, + { + "epoch": 0.6873219373219374, + "grad_norm": 0.5015742182731628, + "learning_rate": 0.00018589898388726389, + "loss": 1.2052, + "step": 3860 + }, + { + "epoch": 0.6875, + "grad_norm": 0.4819730818271637, + "learning_rate": 0.0001858918164431437, + "loss": 1.007, + "step": 3861 + }, + { + "epoch": 0.6876780626780626, + "grad_norm": 0.5510426163673401, + "learning_rate": 0.00018588464731615184, + "loss": 1.0123, + "step": 3862 + }, + { + "epoch": 0.6878561253561254, + "grad_norm": 0.4950829744338989, + "learning_rate": 0.00018587747650642867, + "loss": 1.033, + "step": 3863 + }, + { + "epoch": 0.688034188034188, + "grad_norm": 0.5278680920600891, + "learning_rate": 0.0001858703040141148, + "loss": 1.0912, + "step": 3864 + }, + { + "epoch": 0.6882122507122507, + "grad_norm": 0.6359158158302307, + "learning_rate": 0.00018586312983935068, + "loss": 1.2868, + "step": 3865 + }, + { + "epoch": 0.6883903133903134, + "grad_norm": 0.5098239183425903, + "learning_rate": 0.0001858559539822769, + "loss": 0.8364, + "step": 3866 + }, + { + "epoch": 0.6885683760683761, + "grad_norm": 0.5651038289070129, + "learning_rate": 0.000185848776443034, + "loss": 1.1983, + "step": 3867 + }, + { + "epoch": 0.6887464387464387, + "grad_norm": 0.5305678248405457, + "learning_rate": 0.00018584159722176272, + "loss": 1.32, + "step": 3868 + }, + { + "epoch": 0.6889245014245015, + "grad_norm": 0.5481845140457153, + "learning_rate": 0.00018583441631860368, + "loss": 1.013, + "step": 3869 + }, + { + "epoch": 0.6891025641025641, + "grad_norm": 0.5214795470237732, + "learning_rate": 0.00018582723373369753, + "loss": 1.172, + "step": 3870 + }, + { + "epoch": 0.6892806267806267, + "grad_norm": 0.6282780766487122, + "learning_rate": 0.00018582004946718502, + "loss": 1.7304, + "step": 3871 + }, + { + "epoch": 0.6894586894586895, + "grad_norm": 0.5266988277435303, + "learning_rate": 0.0001858128635192069, + "loss": 1.1418, + "step": 3872 + }, + { + "epoch": 0.6896367521367521, + "grad_norm": 0.4761001467704773, + "learning_rate": 0.000185805675889904, + "loss": 0.8585, + "step": 3873 + }, + { + "epoch": 0.6898148148148148, + "grad_norm": 0.528779923915863, + "learning_rate": 0.00018579848657941715, + "loss": 1.0036, + "step": 3874 + }, + { + "epoch": 0.6899928774928775, + "grad_norm": 0.5427684783935547, + "learning_rate": 0.00018579129558788716, + "loss": 0.9769, + "step": 3875 + }, + { + "epoch": 0.6901709401709402, + "grad_norm": 0.6229544281959534, + "learning_rate": 0.00018578410291545495, + "loss": 1.2848, + "step": 3876 + }, + { + "epoch": 0.6903490028490028, + "grad_norm": 0.6602693200111389, + "learning_rate": 0.00018577690856226147, + "loss": 1.2713, + "step": 3877 + }, + { + "epoch": 0.6905270655270656, + "grad_norm": 0.45884042978286743, + "learning_rate": 0.0001857697125284476, + "loss": 0.9143, + "step": 3878 + }, + { + "epoch": 0.6907051282051282, + "grad_norm": 0.4956444203853607, + "learning_rate": 0.00018576251481415443, + "loss": 0.9646, + "step": 3879 + }, + { + "epoch": 0.6908831908831908, + "grad_norm": 0.473561555147171, + "learning_rate": 0.00018575531541952292, + "loss": 0.843, + "step": 3880 + }, + { + "epoch": 0.6910612535612536, + "grad_norm": 0.4676312506198883, + "learning_rate": 0.00018574811434469415, + "loss": 0.9464, + "step": 3881 + }, + { + "epoch": 0.6912393162393162, + "grad_norm": 0.5452045202255249, + "learning_rate": 0.00018574091158980922, + "loss": 0.985, + "step": 3882 + }, + { + "epoch": 0.6914173789173789, + "grad_norm": 0.6274946331977844, + "learning_rate": 0.0001857337071550092, + "loss": 1.0357, + "step": 3883 + }, + { + "epoch": 0.6915954415954416, + "grad_norm": 0.5533788800239563, + "learning_rate": 0.00018572650104043531, + "loss": 1.2636, + "step": 3884 + }, + { + "epoch": 0.6917735042735043, + "grad_norm": 0.48312318325042725, + "learning_rate": 0.00018571929324622872, + "loss": 1.2402, + "step": 3885 + }, + { + "epoch": 0.6919515669515669, + "grad_norm": 0.6087453961372375, + "learning_rate": 0.00018571208377253062, + "loss": 1.2961, + "step": 3886 + }, + { + "epoch": 0.6921296296296297, + "grad_norm": 0.49156486988067627, + "learning_rate": 0.00018570487261948234, + "loss": 0.9585, + "step": 3887 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 0.5200015902519226, + "learning_rate": 0.0001856976597872251, + "loss": 0.9274, + "step": 3888 + }, + { + "epoch": 0.6924857549857549, + "grad_norm": 0.5185118913650513, + "learning_rate": 0.0001856904452759002, + "loss": 1.0015, + "step": 3889 + }, + { + "epoch": 0.6926638176638177, + "grad_norm": 0.5859049558639526, + "learning_rate": 0.00018568322908564904, + "loss": 1.0959, + "step": 3890 + }, + { + "epoch": 0.6928418803418803, + "grad_norm": 0.5882301926612854, + "learning_rate": 0.00018567601121661302, + "loss": 1.3214, + "step": 3891 + }, + { + "epoch": 0.6930199430199431, + "grad_norm": 0.6475503444671631, + "learning_rate": 0.0001856687916689335, + "loss": 1.3265, + "step": 3892 + }, + { + "epoch": 0.6931980056980057, + "grad_norm": 0.46175432205200195, + "learning_rate": 0.000185661570442752, + "loss": 0.8547, + "step": 3893 + }, + { + "epoch": 0.6933760683760684, + "grad_norm": 0.5362716913223267, + "learning_rate": 0.00018565434753820998, + "loss": 0.974, + "step": 3894 + }, + { + "epoch": 0.6935541310541311, + "grad_norm": 0.4317963719367981, + "learning_rate": 0.00018564712295544896, + "loss": 0.7653, + "step": 3895 + }, + { + "epoch": 0.6937321937321937, + "grad_norm": 0.5679717659950256, + "learning_rate": 0.00018563989669461047, + "loss": 1.0691, + "step": 3896 + }, + { + "epoch": 0.6939102564102564, + "grad_norm": 0.5058363676071167, + "learning_rate": 0.00018563266875583608, + "loss": 1.0665, + "step": 3897 + }, + { + "epoch": 0.6940883190883191, + "grad_norm": 0.5365496277809143, + "learning_rate": 0.00018562543913926746, + "loss": 0.9963, + "step": 3898 + }, + { + "epoch": 0.6942663817663818, + "grad_norm": 0.49945300817489624, + "learning_rate": 0.0001856182078450462, + "loss": 0.8668, + "step": 3899 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 0.5869430899620056, + "learning_rate": 0.00018561097487331405, + "loss": 1.1942, + "step": 3900 + }, + { + "epoch": 0.6946225071225072, + "grad_norm": 0.5188950300216675, + "learning_rate": 0.0001856037402242127, + "loss": 0.9493, + "step": 3901 + }, + { + "epoch": 0.6948005698005698, + "grad_norm": 0.510788083076477, + "learning_rate": 0.00018559650389788384, + "loss": 0.9989, + "step": 3902 + }, + { + "epoch": 0.6949786324786325, + "grad_norm": 0.5360601544380188, + "learning_rate": 0.0001855892658944693, + "loss": 1.2766, + "step": 3903 + }, + { + "epoch": 0.6951566951566952, + "grad_norm": 0.522502601146698, + "learning_rate": 0.00018558202621411093, + "loss": 0.8774, + "step": 3904 + }, + { + "epoch": 0.6953347578347578, + "grad_norm": 0.5330635905265808, + "learning_rate": 0.00018557478485695052, + "loss": 0.972, + "step": 3905 + }, + { + "epoch": 0.6955128205128205, + "grad_norm": 0.5387479066848755, + "learning_rate": 0.00018556754182312996, + "loss": 1.0574, + "step": 3906 + }, + { + "epoch": 0.6956908831908832, + "grad_norm": 0.5357984900474548, + "learning_rate": 0.00018556029711279116, + "loss": 1.396, + "step": 3907 + }, + { + "epoch": 0.6958689458689459, + "grad_norm": 0.5647178292274475, + "learning_rate": 0.00018555305072607612, + "loss": 1.3304, + "step": 3908 + }, + { + "epoch": 0.6960470085470085, + "grad_norm": 0.46460914611816406, + "learning_rate": 0.00018554580266312673, + "loss": 0.9574, + "step": 3909 + }, + { + "epoch": 0.6962250712250713, + "grad_norm": 0.6206206679344177, + "learning_rate": 0.00018553855292408503, + "loss": 1.1637, + "step": 3910 + }, + { + "epoch": 0.6964031339031339, + "grad_norm": 0.5899842977523804, + "learning_rate": 0.00018553130150909312, + "loss": 1.1067, + "step": 3911 + }, + { + "epoch": 0.6965811965811965, + "grad_norm": 0.47294262051582336, + "learning_rate": 0.000185524048418293, + "loss": 1.1516, + "step": 3912 + }, + { + "epoch": 0.6967592592592593, + "grad_norm": 0.5791197419166565, + "learning_rate": 0.00018551679365182684, + "loss": 1.0007, + "step": 3913 + }, + { + "epoch": 0.6969373219373219, + "grad_norm": 0.5678651332855225, + "learning_rate": 0.00018550953720983672, + "loss": 1.2698, + "step": 3914 + }, + { + "epoch": 0.6971153846153846, + "grad_norm": 0.6509683728218079, + "learning_rate": 0.0001855022790924649, + "loss": 1.0354, + "step": 3915 + }, + { + "epoch": 0.6972934472934473, + "grad_norm": 0.5176648497581482, + "learning_rate": 0.0001854950192998535, + "loss": 1.1243, + "step": 3916 + }, + { + "epoch": 0.69747150997151, + "grad_norm": 0.520631730556488, + "learning_rate": 0.00018548775783214477, + "loss": 1.1371, + "step": 3917 + }, + { + "epoch": 0.6976495726495726, + "grad_norm": 0.5408333539962769, + "learning_rate": 0.00018548049468948108, + "loss": 1.1185, + "step": 3918 + }, + { + "epoch": 0.6978276353276354, + "grad_norm": 0.5423790216445923, + "learning_rate": 0.00018547322987200461, + "loss": 1.1539, + "step": 3919 + }, + { + "epoch": 0.698005698005698, + "grad_norm": 0.5422113537788391, + "learning_rate": 0.0001854659633798578, + "loss": 1.171, + "step": 3920 + }, + { + "epoch": 0.6981837606837606, + "grad_norm": 0.5113416314125061, + "learning_rate": 0.00018545869521318292, + "loss": 1.0597, + "step": 3921 + }, + { + "epoch": 0.6983618233618234, + "grad_norm": 0.49901214241981506, + "learning_rate": 0.00018545142537212248, + "loss": 1.1043, + "step": 3922 + }, + { + "epoch": 0.698539886039886, + "grad_norm": 0.6606622338294983, + "learning_rate": 0.00018544415385681885, + "loss": 1.1797, + "step": 3923 + }, + { + "epoch": 0.6987179487179487, + "grad_norm": 0.4786234498023987, + "learning_rate": 0.00018543688066741454, + "loss": 0.9532, + "step": 3924 + }, + { + "epoch": 0.6988960113960114, + "grad_norm": 0.5900700688362122, + "learning_rate": 0.00018542960580405203, + "loss": 1.1171, + "step": 3925 + }, + { + "epoch": 0.6990740740740741, + "grad_norm": 0.53485506772995, + "learning_rate": 0.00018542232926687383, + "loss": 1.1535, + "step": 3926 + }, + { + "epoch": 0.6992521367521367, + "grad_norm": 0.5269177556037903, + "learning_rate": 0.00018541505105602255, + "loss": 1.0287, + "step": 3927 + }, + { + "epoch": 0.6994301994301995, + "grad_norm": 0.5185505151748657, + "learning_rate": 0.0001854077711716408, + "loss": 1.2526, + "step": 3928 + }, + { + "epoch": 0.6996082621082621, + "grad_norm": 0.5615512132644653, + "learning_rate": 0.00018540048961387115, + "loss": 1.0189, + "step": 3929 + }, + { + "epoch": 0.6997863247863247, + "grad_norm": 0.4492493271827698, + "learning_rate": 0.00018539320638285637, + "loss": 0.8917, + "step": 3930 + }, + { + "epoch": 0.6999643874643875, + "grad_norm": 0.5062302947044373, + "learning_rate": 0.00018538592147873906, + "loss": 1.053, + "step": 3931 + }, + { + "epoch": 0.7001424501424501, + "grad_norm": 0.5508798956871033, + "learning_rate": 0.000185378634901662, + "loss": 0.9638, + "step": 3932 + }, + { + "epoch": 0.7003205128205128, + "grad_norm": 0.463980108499527, + "learning_rate": 0.00018537134665176793, + "loss": 1.0945, + "step": 3933 + }, + { + "epoch": 0.7004985754985755, + "grad_norm": 0.5027088522911072, + "learning_rate": 0.0001853640567291997, + "loss": 1.1745, + "step": 3934 + }, + { + "epoch": 0.7006766381766382, + "grad_norm": 0.5006551146507263, + "learning_rate": 0.00018535676513410009, + "loss": 0.8521, + "step": 3935 + }, + { + "epoch": 0.7008547008547008, + "grad_norm": 0.5870724320411682, + "learning_rate": 0.000185349471866612, + "loss": 0.9197, + "step": 3936 + }, + { + "epoch": 0.7010327635327636, + "grad_norm": 0.5030696392059326, + "learning_rate": 0.00018534217692687825, + "loss": 1.1049, + "step": 3937 + }, + { + "epoch": 0.7012108262108262, + "grad_norm": 0.5212681889533997, + "learning_rate": 0.00018533488031504186, + "loss": 1.3397, + "step": 3938 + }, + { + "epoch": 0.7013888888888888, + "grad_norm": 0.5649709105491638, + "learning_rate": 0.0001853275820312458, + "loss": 1.1994, + "step": 3939 + }, + { + "epoch": 0.7015669515669516, + "grad_norm": 0.4892779290676117, + "learning_rate": 0.00018532028207563297, + "loss": 1.1511, + "step": 3940 + }, + { + "epoch": 0.7017450142450142, + "grad_norm": 0.4929407835006714, + "learning_rate": 0.00018531298044834643, + "loss": 1.0792, + "step": 3941 + }, + { + "epoch": 0.7019230769230769, + "grad_norm": 0.5645940899848938, + "learning_rate": 0.00018530567714952932, + "loss": 1.0937, + "step": 3942 + }, + { + "epoch": 0.7021011396011396, + "grad_norm": 0.5471178293228149, + "learning_rate": 0.00018529837217932466, + "loss": 1.193, + "step": 3943 + }, + { + "epoch": 0.7022792022792023, + "grad_norm": 0.576627790927887, + "learning_rate": 0.00018529106553787558, + "loss": 1.1032, + "step": 3944 + }, + { + "epoch": 0.7024572649572649, + "grad_norm": 0.5015735626220703, + "learning_rate": 0.00018528375722532526, + "loss": 1.066, + "step": 3945 + }, + { + "epoch": 0.7026353276353277, + "grad_norm": 0.5315404534339905, + "learning_rate": 0.00018527644724181683, + "loss": 1.2059, + "step": 3946 + }, + { + "epoch": 0.7028133903133903, + "grad_norm": 0.5516065955162048, + "learning_rate": 0.0001852691355874936, + "loss": 1.161, + "step": 3947 + }, + { + "epoch": 0.7029914529914529, + "grad_norm": 0.5026212930679321, + "learning_rate": 0.0001852618222624988, + "loss": 1.2616, + "step": 3948 + }, + { + "epoch": 0.7031695156695157, + "grad_norm": 0.49874603748321533, + "learning_rate": 0.0001852545072669757, + "loss": 0.805, + "step": 3949 + }, + { + "epoch": 0.7033475783475783, + "grad_norm": 0.47698748111724854, + "learning_rate": 0.00018524719060106763, + "loss": 1.2321, + "step": 3950 + }, + { + "epoch": 0.7035256410256411, + "grad_norm": 0.5201322436332703, + "learning_rate": 0.00018523987226491792, + "loss": 1.1577, + "step": 3951 + }, + { + "epoch": 0.7037037037037037, + "grad_norm": 0.5506543517112732, + "learning_rate": 0.00018523255225867002, + "loss": 1.2289, + "step": 3952 + }, + { + "epoch": 0.7038817663817664, + "grad_norm": 0.5691256523132324, + "learning_rate": 0.0001852252305824673, + "loss": 1.1945, + "step": 3953 + }, + { + "epoch": 0.7040598290598291, + "grad_norm": 0.5324838757514954, + "learning_rate": 0.00018521790723645322, + "loss": 1.1037, + "step": 3954 + }, + { + "epoch": 0.7042378917378918, + "grad_norm": 0.5238786339759827, + "learning_rate": 0.00018521058222077127, + "loss": 1.2075, + "step": 3955 + }, + { + "epoch": 0.7044159544159544, + "grad_norm": 0.4936453402042389, + "learning_rate": 0.00018520325553556498, + "loss": 1.0537, + "step": 3956 + }, + { + "epoch": 0.7045940170940171, + "grad_norm": 0.6198282837867737, + "learning_rate": 0.00018519592718097791, + "loss": 1.0728, + "step": 3957 + }, + { + "epoch": 0.7047720797720798, + "grad_norm": 0.44729140400886536, + "learning_rate": 0.0001851885971571536, + "loss": 0.8432, + "step": 3958 + }, + { + "epoch": 0.7049501424501424, + "grad_norm": 0.5884211659431458, + "learning_rate": 0.00018518126546423572, + "loss": 0.9515, + "step": 3959 + }, + { + "epoch": 0.7051282051282052, + "grad_norm": 0.5293807983398438, + "learning_rate": 0.00018517393210236788, + "loss": 1.1178, + "step": 3960 + }, + { + "epoch": 0.7053062678062678, + "grad_norm": 0.6036825180053711, + "learning_rate": 0.00018516659707169374, + "loss": 1.0408, + "step": 3961 + }, + { + "epoch": 0.7054843304843305, + "grad_norm": 0.5157122015953064, + "learning_rate": 0.0001851592603723571, + "loss": 1.2136, + "step": 3962 + }, + { + "epoch": 0.7056623931623932, + "grad_norm": 0.5354781150817871, + "learning_rate": 0.00018515192200450163, + "loss": 0.7165, + "step": 3963 + }, + { + "epoch": 0.7058404558404558, + "grad_norm": 0.6073734760284424, + "learning_rate": 0.00018514458196827111, + "loss": 1.3079, + "step": 3964 + }, + { + "epoch": 0.7060185185185185, + "grad_norm": 0.4324839413166046, + "learning_rate": 0.0001851372402638094, + "loss": 0.7903, + "step": 3965 + }, + { + "epoch": 0.7061965811965812, + "grad_norm": 0.6530333161354065, + "learning_rate": 0.00018512989689126034, + "loss": 1.3179, + "step": 3966 + }, + { + "epoch": 0.7063746438746439, + "grad_norm": 0.5500404238700867, + "learning_rate": 0.00018512255185076782, + "loss": 1.0624, + "step": 3967 + }, + { + "epoch": 0.7065527065527065, + "grad_norm": 0.6277863383293152, + "learning_rate": 0.00018511520514247567, + "loss": 1.1056, + "step": 3968 + }, + { + "epoch": 0.7067307692307693, + "grad_norm": 0.580544650554657, + "learning_rate": 0.0001851078567665279, + "loss": 0.9849, + "step": 3969 + }, + { + "epoch": 0.7069088319088319, + "grad_norm": 0.4880999028682709, + "learning_rate": 0.00018510050672306848, + "loss": 1.0185, + "step": 3970 + }, + { + "epoch": 0.7070868945868946, + "grad_norm": 0.4919959306716919, + "learning_rate": 0.0001850931550122414, + "loss": 1.0334, + "step": 3971 + }, + { + "epoch": 0.7072649572649573, + "grad_norm": 0.6001213192939758, + "learning_rate": 0.0001850858016341907, + "loss": 1.0729, + "step": 3972 + }, + { + "epoch": 0.70744301994302, + "grad_norm": 0.538690447807312, + "learning_rate": 0.00018507844658906052, + "loss": 1.0733, + "step": 3973 + }, + { + "epoch": 0.7076210826210826, + "grad_norm": 0.5427643656730652, + "learning_rate": 0.00018507108987699487, + "loss": 1.1207, + "step": 3974 + }, + { + "epoch": 0.7077991452991453, + "grad_norm": 0.43014347553253174, + "learning_rate": 0.00018506373149813795, + "loss": 0.7958, + "step": 3975 + }, + { + "epoch": 0.707977207977208, + "grad_norm": 0.56591796875, + "learning_rate": 0.00018505637145263394, + "loss": 1.2199, + "step": 3976 + }, + { + "epoch": 0.7081552706552706, + "grad_norm": 0.59147047996521, + "learning_rate": 0.000185049009740627, + "loss": 1.2354, + "step": 3977 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 0.5078346133232117, + "learning_rate": 0.00018504164636226137, + "loss": 0.976, + "step": 3978 + }, + { + "epoch": 0.708511396011396, + "grad_norm": 0.533302366733551, + "learning_rate": 0.00018503428131768135, + "loss": 0.9653, + "step": 3979 + }, + { + "epoch": 0.7086894586894587, + "grad_norm": 0.4985341727733612, + "learning_rate": 0.00018502691460703122, + "loss": 1.1485, + "step": 3980 + }, + { + "epoch": 0.7088675213675214, + "grad_norm": 0.5143141150474548, + "learning_rate": 0.00018501954623045532, + "loss": 1.148, + "step": 3981 + }, + { + "epoch": 0.709045584045584, + "grad_norm": 0.507189154624939, + "learning_rate": 0.00018501217618809804, + "loss": 0.9306, + "step": 3982 + }, + { + "epoch": 0.7092236467236467, + "grad_norm": 0.5246604084968567, + "learning_rate": 0.00018500480448010377, + "loss": 0.9116, + "step": 3983 + }, + { + "epoch": 0.7094017094017094, + "grad_norm": 0.5321049094200134, + "learning_rate": 0.00018499743110661693, + "loss": 0.9607, + "step": 3984 + }, + { + "epoch": 0.7095797720797721, + "grad_norm": 0.62645423412323, + "learning_rate": 0.000184990056067782, + "loss": 1.5834, + "step": 3985 + }, + { + "epoch": 0.7097578347578347, + "grad_norm": 0.486557275056839, + "learning_rate": 0.0001849826793637435, + "loss": 1.0598, + "step": 3986 + }, + { + "epoch": 0.7099358974358975, + "grad_norm": 0.5122783184051514, + "learning_rate": 0.0001849753009946459, + "loss": 1.2213, + "step": 3987 + }, + { + "epoch": 0.7101139601139601, + "grad_norm": 0.4864068627357483, + "learning_rate": 0.0001849679209606338, + "loss": 1.2708, + "step": 3988 + }, + { + "epoch": 0.7102920227920227, + "grad_norm": 0.5860990881919861, + "learning_rate": 0.00018496053926185183, + "loss": 1.2421, + "step": 3989 + }, + { + "epoch": 0.7104700854700855, + "grad_norm": 0.471194326877594, + "learning_rate": 0.00018495315589844453, + "loss": 0.879, + "step": 3990 + }, + { + "epoch": 0.7106481481481481, + "grad_norm": 0.5626323819160461, + "learning_rate": 0.00018494577087055662, + "loss": 1.1297, + "step": 3991 + }, + { + "epoch": 0.7108262108262108, + "grad_norm": 0.4706762135028839, + "learning_rate": 0.0001849383841783328, + "loss": 1.0444, + "step": 3992 + }, + { + "epoch": 0.7110042735042735, + "grad_norm": 0.5776444673538208, + "learning_rate": 0.00018493099582191783, + "loss": 1.1773, + "step": 3993 + }, + { + "epoch": 0.7111823361823362, + "grad_norm": 0.5493253469467163, + "learning_rate": 0.00018492360580145637, + "loss": 1.0354, + "step": 3994 + }, + { + "epoch": 0.7113603988603988, + "grad_norm": 0.5328514575958252, + "learning_rate": 0.0001849162141170933, + "loss": 0.9251, + "step": 3995 + }, + { + "epoch": 0.7115384615384616, + "grad_norm": 0.5814893841743469, + "learning_rate": 0.0001849088207689734, + "loss": 1.1066, + "step": 3996 + }, + { + "epoch": 0.7117165242165242, + "grad_norm": 0.5476071834564209, + "learning_rate": 0.00018490142575724154, + "loss": 1.1613, + "step": 3997 + }, + { + "epoch": 0.7118945868945868, + "grad_norm": 0.5216463208198547, + "learning_rate": 0.00018489402908204258, + "loss": 1.2574, + "step": 3998 + }, + { + "epoch": 0.7120726495726496, + "grad_norm": 0.5110020637512207, + "learning_rate": 0.00018488663074352153, + "loss": 1.0663, + "step": 3999 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 0.448090523481369, + "learning_rate": 0.00018487923074182326, + "loss": 0.6687, + "step": 4000 + }, + { + "epoch": 0.7124287749287749, + "grad_norm": 0.4980565011501312, + "learning_rate": 0.00018487182907709279, + "loss": 1.2365, + "step": 4001 + }, + { + "epoch": 0.7126068376068376, + "grad_norm": 0.485831081867218, + "learning_rate": 0.00018486442574947511, + "loss": 1.0941, + "step": 4002 + }, + { + "epoch": 0.7127849002849003, + "grad_norm": 0.4955040216445923, + "learning_rate": 0.00018485702075911534, + "loss": 1.248, + "step": 4003 + }, + { + "epoch": 0.7129629629629629, + "grad_norm": 0.5168375968933105, + "learning_rate": 0.00018484961410615845, + "loss": 1.1118, + "step": 4004 + }, + { + "epoch": 0.7131410256410257, + "grad_norm": 0.5255687832832336, + "learning_rate": 0.00018484220579074968, + "loss": 1.0558, + "step": 4005 + }, + { + "epoch": 0.7133190883190883, + "grad_norm": 0.5502219796180725, + "learning_rate": 0.00018483479581303416, + "loss": 1.1604, + "step": 4006 + }, + { + "epoch": 0.7134971509971509, + "grad_norm": 0.5155881643295288, + "learning_rate": 0.000184827384173157, + "loss": 0.8246, + "step": 4007 + }, + { + "epoch": 0.7136752136752137, + "grad_norm": 0.5321542024612427, + "learning_rate": 0.0001848199708712635, + "loss": 1.2058, + "step": 4008 + }, + { + "epoch": 0.7138532763532763, + "grad_norm": 0.4929848313331604, + "learning_rate": 0.00018481255590749884, + "loss": 1.4023, + "step": 4009 + }, + { + "epoch": 0.7140313390313391, + "grad_norm": 0.5070937871932983, + "learning_rate": 0.00018480513928200836, + "loss": 1.0561, + "step": 4010 + }, + { + "epoch": 0.7142094017094017, + "grad_norm": 0.5750083327293396, + "learning_rate": 0.00018479772099493728, + "loss": 1.0276, + "step": 4011 + }, + { + "epoch": 0.7143874643874644, + "grad_norm": 0.5265933275222778, + "learning_rate": 0.00018479030104643108, + "loss": 1.0295, + "step": 4012 + }, + { + "epoch": 0.7145655270655271, + "grad_norm": 0.526830792427063, + "learning_rate": 0.00018478287943663504, + "loss": 1.0157, + "step": 4013 + }, + { + "epoch": 0.7147435897435898, + "grad_norm": 0.5344091653823853, + "learning_rate": 0.00018477545616569458, + "loss": 1.1997, + "step": 4014 + }, + { + "epoch": 0.7149216524216524, + "grad_norm": 0.4935445189476013, + "learning_rate": 0.0001847680312337552, + "loss": 1.1858, + "step": 4015 + }, + { + "epoch": 0.7150997150997151, + "grad_norm": 0.5291212797164917, + "learning_rate": 0.0001847606046409623, + "loss": 0.926, + "step": 4016 + }, + { + "epoch": 0.7152777777777778, + "grad_norm": 0.559050977230072, + "learning_rate": 0.00018475317638746142, + "loss": 1.0947, + "step": 4017 + }, + { + "epoch": 0.7154558404558404, + "grad_norm": 0.4566570222377777, + "learning_rate": 0.00018474574647339814, + "loss": 1.0334, + "step": 4018 + }, + { + "epoch": 0.7156339031339032, + "grad_norm": 0.5156155824661255, + "learning_rate": 0.000184738314898918, + "loss": 1.0076, + "step": 4019 + }, + { + "epoch": 0.7158119658119658, + "grad_norm": 0.5008716583251953, + "learning_rate": 0.00018473088166416662, + "loss": 1.0378, + "step": 4020 + }, + { + "epoch": 0.7159900284900285, + "grad_norm": 0.49556368589401245, + "learning_rate": 0.0001847234467692896, + "loss": 1.15, + "step": 4021 + }, + { + "epoch": 0.7161680911680912, + "grad_norm": 0.5464680790901184, + "learning_rate": 0.00018471601021443265, + "loss": 1.2975, + "step": 4022 + }, + { + "epoch": 0.7163461538461539, + "grad_norm": 0.6291980147361755, + "learning_rate": 0.00018470857199974144, + "loss": 1.05, + "step": 4023 + }, + { + "epoch": 0.7165242165242165, + "grad_norm": 0.5566631555557251, + "learning_rate": 0.00018470113212536176, + "loss": 1.1296, + "step": 4024 + }, + { + "epoch": 0.7167022792022792, + "grad_norm": 0.5569562911987305, + "learning_rate": 0.00018469369059143933, + "loss": 1.2484, + "step": 4025 + }, + { + "epoch": 0.7168803418803419, + "grad_norm": 0.5804716944694519, + "learning_rate": 0.00018468624739812, + "loss": 1.0547, + "step": 4026 + }, + { + "epoch": 0.7170584045584045, + "grad_norm": 0.6316802501678467, + "learning_rate": 0.00018467880254554952, + "loss": 1.1188, + "step": 4027 + }, + { + "epoch": 0.7172364672364673, + "grad_norm": 0.6131419539451599, + "learning_rate": 0.00018467135603387385, + "loss": 1.1662, + "step": 4028 + }, + { + "epoch": 0.7174145299145299, + "grad_norm": 0.4703124761581421, + "learning_rate": 0.00018466390786323883, + "loss": 1.038, + "step": 4029 + }, + { + "epoch": 0.7175925925925926, + "grad_norm": 0.5718469023704529, + "learning_rate": 0.0001846564580337904, + "loss": 1.0786, + "step": 4030 + }, + { + "epoch": 0.7177706552706553, + "grad_norm": 0.5227612853050232, + "learning_rate": 0.00018464900654567457, + "loss": 1.0561, + "step": 4031 + }, + { + "epoch": 0.717948717948718, + "grad_norm": 0.5800358057022095, + "learning_rate": 0.00018464155339903727, + "loss": 1.0944, + "step": 4032 + }, + { + "epoch": 0.7181267806267806, + "grad_norm": 0.5562314987182617, + "learning_rate": 0.00018463409859402455, + "loss": 0.8573, + "step": 4033 + }, + { + "epoch": 0.7183048433048433, + "grad_norm": 0.6420153379440308, + "learning_rate": 0.0001846266421307825, + "loss": 1.088, + "step": 4034 + }, + { + "epoch": 0.718482905982906, + "grad_norm": 0.4745902717113495, + "learning_rate": 0.00018461918400945718, + "loss": 1.1679, + "step": 4035 + }, + { + "epoch": 0.7186609686609686, + "grad_norm": 0.5070300102233887, + "learning_rate": 0.00018461172423019475, + "loss": 1.1984, + "step": 4036 + }, + { + "epoch": 0.7188390313390314, + "grad_norm": 0.5339375138282776, + "learning_rate": 0.00018460426279314133, + "loss": 1.3038, + "step": 4037 + }, + { + "epoch": 0.719017094017094, + "grad_norm": 0.5947147607803345, + "learning_rate": 0.00018459679969844313, + "loss": 1.0103, + "step": 4038 + }, + { + "epoch": 0.7191951566951567, + "grad_norm": 0.5493791699409485, + "learning_rate": 0.00018458933494624642, + "loss": 1.1001, + "step": 4039 + }, + { + "epoch": 0.7193732193732194, + "grad_norm": 0.5700310468673706, + "learning_rate": 0.00018458186853669736, + "loss": 0.9006, + "step": 4040 + }, + { + "epoch": 0.719551282051282, + "grad_norm": 0.60371994972229, + "learning_rate": 0.0001845744004699423, + "loss": 1.3001, + "step": 4041 + }, + { + "epoch": 0.7197293447293447, + "grad_norm": 0.5469261407852173, + "learning_rate": 0.00018456693074612757, + "loss": 1.1745, + "step": 4042 + }, + { + "epoch": 0.7199074074074074, + "grad_norm": 0.5179165601730347, + "learning_rate": 0.00018455945936539947, + "loss": 0.9883, + "step": 4043 + }, + { + "epoch": 0.7200854700854701, + "grad_norm": 0.5396696329116821, + "learning_rate": 0.00018455198632790447, + "loss": 1.1277, + "step": 4044 + }, + { + "epoch": 0.7202635327635327, + "grad_norm": 0.4559909403324127, + "learning_rate": 0.00018454451163378888, + "loss": 0.9644, + "step": 4045 + }, + { + "epoch": 0.7204415954415955, + "grad_norm": 0.49863892793655396, + "learning_rate": 0.00018453703528319927, + "loss": 1.1276, + "step": 4046 + }, + { + "epoch": 0.7206196581196581, + "grad_norm": 0.4790710508823395, + "learning_rate": 0.000184529557276282, + "loss": 0.9443, + "step": 4047 + }, + { + "epoch": 0.7207977207977208, + "grad_norm": 0.541999876499176, + "learning_rate": 0.0001845220776131837, + "loss": 1.0681, + "step": 4048 + }, + { + "epoch": 0.7209757834757835, + "grad_norm": 0.5119109153747559, + "learning_rate": 0.00018451459629405088, + "loss": 1.2078, + "step": 4049 + }, + { + "epoch": 0.7211538461538461, + "grad_norm": 0.6141307353973389, + "learning_rate": 0.00018450711331903006, + "loss": 1.1071, + "step": 4050 + }, + { + "epoch": 0.7213319088319088, + "grad_norm": 0.48679864406585693, + "learning_rate": 0.00018449962868826795, + "loss": 0.9713, + "step": 4051 + }, + { + "epoch": 0.7215099715099715, + "grad_norm": 0.5548661947250366, + "learning_rate": 0.0001844921424019111, + "loss": 1.2099, + "step": 4052 + }, + { + "epoch": 0.7216880341880342, + "grad_norm": 0.5000107884407043, + "learning_rate": 0.00018448465446010626, + "loss": 1.0184, + "step": 4053 + }, + { + "epoch": 0.7218660968660968, + "grad_norm": 0.6131454110145569, + "learning_rate": 0.00018447716486300013, + "loss": 1.2581, + "step": 4054 + }, + { + "epoch": 0.7220441595441596, + "grad_norm": 0.5145987868309021, + "learning_rate": 0.0001844696736107394, + "loss": 1.1646, + "step": 4055 + }, + { + "epoch": 0.7222222222222222, + "grad_norm": 0.4361337125301361, + "learning_rate": 0.00018446218070347094, + "loss": 0.8239, + "step": 4056 + }, + { + "epoch": 0.7224002849002849, + "grad_norm": 0.5549173355102539, + "learning_rate": 0.00018445468614134146, + "loss": 1.1935, + "step": 4057 + }, + { + "epoch": 0.7225783475783476, + "grad_norm": 0.5569297671318054, + "learning_rate": 0.00018444718992449789, + "loss": 1.0137, + "step": 4058 + }, + { + "epoch": 0.7227564102564102, + "grad_norm": 0.44866305589675903, + "learning_rate": 0.00018443969205308704, + "loss": 0.987, + "step": 4059 + }, + { + "epoch": 0.7229344729344729, + "grad_norm": 0.5142943263053894, + "learning_rate": 0.0001844321925272558, + "loss": 1.0837, + "step": 4060 + }, + { + "epoch": 0.7231125356125356, + "grad_norm": 0.4922119379043579, + "learning_rate": 0.0001844246913471512, + "loss": 0.8477, + "step": 4061 + }, + { + "epoch": 0.7232905982905983, + "grad_norm": 0.5245375633239746, + "learning_rate": 0.0001844171885129201, + "loss": 0.9985, + "step": 4062 + }, + { + "epoch": 0.7234686609686609, + "grad_norm": 0.45562678575515747, + "learning_rate": 0.00018440968402470956, + "loss": 0.8678, + "step": 4063 + }, + { + "epoch": 0.7236467236467237, + "grad_norm": 0.5388376712799072, + "learning_rate": 0.0001844021778826666, + "loss": 1.0586, + "step": 4064 + }, + { + "epoch": 0.7238247863247863, + "grad_norm": 0.48945263028144836, + "learning_rate": 0.00018439467008693833, + "loss": 1.0547, + "step": 4065 + }, + { + "epoch": 0.7240028490028491, + "grad_norm": 0.5202330350875854, + "learning_rate": 0.00018438716063767178, + "loss": 1.3142, + "step": 4066 + }, + { + "epoch": 0.7241809116809117, + "grad_norm": 0.5432567000389099, + "learning_rate": 0.00018437964953501413, + "loss": 1.0192, + "step": 4067 + }, + { + "epoch": 0.7243589743589743, + "grad_norm": 0.5220325589179993, + "learning_rate": 0.00018437213677911253, + "loss": 1.0904, + "step": 4068 + }, + { + "epoch": 0.7245370370370371, + "grad_norm": 0.45711690187454224, + "learning_rate": 0.00018436462237011417, + "loss": 1.0417, + "step": 4069 + }, + { + "epoch": 0.7247150997150997, + "grad_norm": 0.560778021812439, + "learning_rate": 0.0001843571063081663, + "loss": 1.2316, + "step": 4070 + }, + { + "epoch": 0.7248931623931624, + "grad_norm": 0.591533362865448, + "learning_rate": 0.0001843495885934162, + "loss": 1.0294, + "step": 4071 + }, + { + "epoch": 0.7250712250712251, + "grad_norm": 0.5550443530082703, + "learning_rate": 0.00018434206922601106, + "loss": 1.0162, + "step": 4072 + }, + { + "epoch": 0.7252492877492878, + "grad_norm": 0.5744053721427917, + "learning_rate": 0.00018433454820609833, + "loss": 1.2774, + "step": 4073 + }, + { + "epoch": 0.7254273504273504, + "grad_norm": 0.6210703253746033, + "learning_rate": 0.0001843270255338253, + "loss": 1.2526, + "step": 4074 + }, + { + "epoch": 0.7256054131054132, + "grad_norm": 0.49684277176856995, + "learning_rate": 0.0001843195012093394, + "loss": 1.0786, + "step": 4075 + }, + { + "epoch": 0.7257834757834758, + "grad_norm": 0.5851606130599976, + "learning_rate": 0.00018431197523278802, + "loss": 1.14, + "step": 4076 + }, + { + "epoch": 0.7259615384615384, + "grad_norm": 0.5494425296783447, + "learning_rate": 0.00018430444760431862, + "loss": 1.211, + "step": 4077 + }, + { + "epoch": 0.7261396011396012, + "grad_norm": 0.5247658491134644, + "learning_rate": 0.00018429691832407867, + "loss": 0.8031, + "step": 4078 + }, + { + "epoch": 0.7263176638176638, + "grad_norm": 0.5012249946594238, + "learning_rate": 0.00018428938739221574, + "loss": 1.1258, + "step": 4079 + }, + { + "epoch": 0.7264957264957265, + "grad_norm": 0.5226427912712097, + "learning_rate": 0.0001842818548088774, + "loss": 1.0029, + "step": 4080 + }, + { + "epoch": 0.7266737891737892, + "grad_norm": 0.45008543133735657, + "learning_rate": 0.00018427432057421114, + "loss": 1.0681, + "step": 4081 + }, + { + "epoch": 0.7268518518518519, + "grad_norm": 0.5127285122871399, + "learning_rate": 0.00018426678468836467, + "loss": 1.1069, + "step": 4082 + }, + { + "epoch": 0.7270299145299145, + "grad_norm": 0.5406150221824646, + "learning_rate": 0.0001842592471514856, + "loss": 1.052, + "step": 4083 + }, + { + "epoch": 0.7272079772079773, + "grad_norm": 0.5001157522201538, + "learning_rate": 0.0001842517079637216, + "loss": 0.9157, + "step": 4084 + }, + { + "epoch": 0.7273860398860399, + "grad_norm": 0.6169779300689697, + "learning_rate": 0.00018424416712522042, + "loss": 1.3133, + "step": 4085 + }, + { + "epoch": 0.7275641025641025, + "grad_norm": 0.4891316890716553, + "learning_rate": 0.00018423662463612974, + "loss": 0.9505, + "step": 4086 + }, + { + "epoch": 0.7277421652421653, + "grad_norm": 0.5883708596229553, + "learning_rate": 0.00018422908049659743, + "loss": 1.2797, + "step": 4087 + }, + { + "epoch": 0.7279202279202279, + "grad_norm": 0.6679072976112366, + "learning_rate": 0.00018422153470677125, + "loss": 1.1096, + "step": 4088 + }, + { + "epoch": 0.7280982905982906, + "grad_norm": 0.5178479552268982, + "learning_rate": 0.00018421398726679904, + "loss": 1.0299, + "step": 4089 + }, + { + "epoch": 0.7282763532763533, + "grad_norm": 0.6343900561332703, + "learning_rate": 0.0001842064381768287, + "loss": 1.2983, + "step": 4090 + }, + { + "epoch": 0.728454415954416, + "grad_norm": 0.43816515803337097, + "learning_rate": 0.0001841988874370081, + "loss": 0.9452, + "step": 4091 + }, + { + "epoch": 0.7286324786324786, + "grad_norm": 0.579790472984314, + "learning_rate": 0.00018419133504748528, + "loss": 1.1037, + "step": 4092 + }, + { + "epoch": 0.7288105413105413, + "grad_norm": 0.571374773979187, + "learning_rate": 0.00018418378100840807, + "loss": 1.1655, + "step": 4093 + }, + { + "epoch": 0.728988603988604, + "grad_norm": 0.5163514018058777, + "learning_rate": 0.0001841762253199246, + "loss": 1.1579, + "step": 4094 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 0.6553022265434265, + "learning_rate": 0.0001841686679821828, + "loss": 0.9664, + "step": 4095 + }, + { + "epoch": 0.7293447293447294, + "grad_norm": 0.5072969198226929, + "learning_rate": 0.00018416110899533084, + "loss": 0.9416, + "step": 4096 + }, + { + "epoch": 0.729522792022792, + "grad_norm": 0.5103251338005066, + "learning_rate": 0.00018415354835951675, + "loss": 1.0715, + "step": 4097 + }, + { + "epoch": 0.7297008547008547, + "grad_norm": 0.49752289056777954, + "learning_rate": 0.00018414598607488874, + "loss": 1.1848, + "step": 4098 + }, + { + "epoch": 0.7298789173789174, + "grad_norm": 0.5361882448196411, + "learning_rate": 0.00018413842214159488, + "loss": 1.1035, + "step": 4099 + }, + { + "epoch": 0.73005698005698, + "grad_norm": 0.5167670249938965, + "learning_rate": 0.00018413085655978343, + "loss": 1.0015, + "step": 4100 + }, + { + "epoch": 0.7302350427350427, + "grad_norm": 0.5930629372596741, + "learning_rate": 0.00018412328932960263, + "loss": 0.9766, + "step": 4101 + }, + { + "epoch": 0.7304131054131054, + "grad_norm": 0.5234778523445129, + "learning_rate": 0.00018411572045120073, + "loss": 1.0317, + "step": 4102 + }, + { + "epoch": 0.7305911680911681, + "grad_norm": 0.5361374020576477, + "learning_rate": 0.000184108149924726, + "loss": 1.1228, + "step": 4103 + }, + { + "epoch": 0.7307692307692307, + "grad_norm": 0.5845770239830017, + "learning_rate": 0.0001841005777503268, + "loss": 0.9541, + "step": 4104 + }, + { + "epoch": 0.7309472934472935, + "grad_norm": 0.49320483207702637, + "learning_rate": 0.0001840930039281515, + "loss": 0.9445, + "step": 4105 + }, + { + "epoch": 0.7311253561253561, + "grad_norm": 0.5391250252723694, + "learning_rate": 0.00018408542845834845, + "loss": 1.1983, + "step": 4106 + }, + { + "epoch": 0.7313034188034188, + "grad_norm": 0.4890393316745758, + "learning_rate": 0.00018407785134106613, + "loss": 0.8353, + "step": 4107 + }, + { + "epoch": 0.7314814814814815, + "grad_norm": 0.5839747190475464, + "learning_rate": 0.00018407027257645296, + "loss": 1.4074, + "step": 4108 + }, + { + "epoch": 0.7316595441595442, + "grad_norm": 0.5957708358764648, + "learning_rate": 0.0001840626921646574, + "loss": 1.1032, + "step": 4109 + }, + { + "epoch": 0.7318376068376068, + "grad_norm": 0.5029017925262451, + "learning_rate": 0.00018405511010582805, + "loss": 1.095, + "step": 4110 + }, + { + "epoch": 0.7320156695156695, + "grad_norm": 0.6054347157478333, + "learning_rate": 0.00018404752640011345, + "loss": 1.0366, + "step": 4111 + }, + { + "epoch": 0.7321937321937322, + "grad_norm": 0.5476830005645752, + "learning_rate": 0.00018403994104766212, + "loss": 1.0976, + "step": 4112 + }, + { + "epoch": 0.7323717948717948, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00018403235404862277, + "loss": 1.0809, + "step": 4113 + }, + { + "epoch": 0.7325498575498576, + "grad_norm": 0.5119251012802124, + "learning_rate": 0.00018402476540314394, + "loss": 1.0176, + "step": 4114 + }, + { + "epoch": 0.7327279202279202, + "grad_norm": 0.5825830698013306, + "learning_rate": 0.00018401717511137445, + "loss": 1.2357, + "step": 4115 + }, + { + "epoch": 0.7329059829059829, + "grad_norm": 0.5702941417694092, + "learning_rate": 0.0001840095831734629, + "loss": 1.1549, + "step": 4116 + }, + { + "epoch": 0.7330840455840456, + "grad_norm": 0.5660699605941772, + "learning_rate": 0.00018400198958955807, + "loss": 1.1778, + "step": 4117 + }, + { + "epoch": 0.7332621082621082, + "grad_norm": 0.5241161584854126, + "learning_rate": 0.0001839943943598088, + "loss": 0.8587, + "step": 4118 + }, + { + "epoch": 0.7334401709401709, + "grad_norm": 0.581194281578064, + "learning_rate": 0.0001839867974843638, + "loss": 1.2169, + "step": 4119 + }, + { + "epoch": 0.7336182336182336, + "grad_norm": 0.4342379570007324, + "learning_rate": 0.00018397919896337198, + "loss": 0.9182, + "step": 4120 + }, + { + "epoch": 0.7337962962962963, + "grad_norm": 0.5708567500114441, + "learning_rate": 0.00018397159879698224, + "loss": 1.1781, + "step": 4121 + }, + { + "epoch": 0.7339743589743589, + "grad_norm": 0.5827265977859497, + "learning_rate": 0.00018396399698534344, + "loss": 1.2905, + "step": 4122 + }, + { + "epoch": 0.7341524216524217, + "grad_norm": 0.5274056792259216, + "learning_rate": 0.00018395639352860457, + "loss": 1.1786, + "step": 4123 + }, + { + "epoch": 0.7343304843304843, + "grad_norm": 0.5094266533851624, + "learning_rate": 0.00018394878842691452, + "loss": 1.2016, + "step": 4124 + }, + { + "epoch": 0.7345085470085471, + "grad_norm": 0.48779475688934326, + "learning_rate": 0.0001839411816804224, + "loss": 1.0562, + "step": 4125 + }, + { + "epoch": 0.7346866096866097, + "grad_norm": 0.5805709958076477, + "learning_rate": 0.00018393357328927716, + "loss": 1.1705, + "step": 4126 + }, + { + "epoch": 0.7348646723646723, + "grad_norm": 0.4910700023174286, + "learning_rate": 0.00018392596325362791, + "loss": 1.0682, + "step": 4127 + }, + { + "epoch": 0.7350427350427351, + "grad_norm": 0.5297428369522095, + "learning_rate": 0.0001839183515736238, + "loss": 0.9505, + "step": 4128 + }, + { + "epoch": 0.7352207977207977, + "grad_norm": 0.45442086458206177, + "learning_rate": 0.00018391073824941385, + "loss": 0.9548, + "step": 4129 + }, + { + "epoch": 0.7353988603988604, + "grad_norm": 0.49299946427345276, + "learning_rate": 0.00018390312328114733, + "loss": 1.0868, + "step": 4130 + }, + { + "epoch": 0.7355769230769231, + "grad_norm": 0.4839940369129181, + "learning_rate": 0.0001838955066689734, + "loss": 0.9565, + "step": 4131 + }, + { + "epoch": 0.7357549857549858, + "grad_norm": 0.48600608110427856, + "learning_rate": 0.00018388788841304128, + "loss": 1.2353, + "step": 4132 + }, + { + "epoch": 0.7359330484330484, + "grad_norm": 0.4893583357334137, + "learning_rate": 0.0001838802685135003, + "loss": 0.9595, + "step": 4133 + }, + { + "epoch": 0.7361111111111112, + "grad_norm": 0.4587398171424866, + "learning_rate": 0.00018387264697049963, + "loss": 1.1222, + "step": 4134 + }, + { + "epoch": 0.7362891737891738, + "grad_norm": 0.5361055731773376, + "learning_rate": 0.00018386502378418872, + "loss": 1.3304, + "step": 4135 + }, + { + "epoch": 0.7364672364672364, + "grad_norm": 0.5556629300117493, + "learning_rate": 0.00018385739895471686, + "loss": 1.0358, + "step": 4136 + }, + { + "epoch": 0.7366452991452992, + "grad_norm": 0.45555856823921204, + "learning_rate": 0.00018384977248223346, + "loss": 1.0081, + "step": 4137 + }, + { + "epoch": 0.7368233618233618, + "grad_norm": 0.5606052875518799, + "learning_rate": 0.00018384214436688797, + "loss": 0.9367, + "step": 4138 + }, + { + "epoch": 0.7370014245014245, + "grad_norm": 0.5428356528282166, + "learning_rate": 0.00018383451460882982, + "loss": 1.1391, + "step": 4139 + }, + { + "epoch": 0.7371794871794872, + "grad_norm": 0.4891330897808075, + "learning_rate": 0.00018382688320820853, + "loss": 0.9805, + "step": 4140 + }, + { + "epoch": 0.7373575498575499, + "grad_norm": 0.5407996773719788, + "learning_rate": 0.0001838192501651736, + "loss": 1.0532, + "step": 4141 + }, + { + "epoch": 0.7375356125356125, + "grad_norm": 0.5241971611976624, + "learning_rate": 0.00018381161547987454, + "loss": 0.9509, + "step": 4142 + }, + { + "epoch": 0.7377136752136753, + "grad_norm": 0.5370210409164429, + "learning_rate": 0.000183803979152461, + "loss": 1.2342, + "step": 4143 + }, + { + "epoch": 0.7378917378917379, + "grad_norm": 0.5470060706138611, + "learning_rate": 0.00018379634118308259, + "loss": 0.9621, + "step": 4144 + }, + { + "epoch": 0.7380698005698005, + "grad_norm": 0.546313464641571, + "learning_rate": 0.00018378870157188893, + "loss": 1.1253, + "step": 4145 + }, + { + "epoch": 0.7382478632478633, + "grad_norm": 0.502027153968811, + "learning_rate": 0.00018378106031902974, + "loss": 1.1919, + "step": 4146 + }, + { + "epoch": 0.7384259259259259, + "grad_norm": 0.5282283425331116, + "learning_rate": 0.0001837734174246547, + "loss": 1.0088, + "step": 4147 + }, + { + "epoch": 0.7386039886039886, + "grad_norm": 0.5152897238731384, + "learning_rate": 0.00018376577288891355, + "loss": 1.0813, + "step": 4148 + }, + { + "epoch": 0.7387820512820513, + "grad_norm": 0.5002804398536682, + "learning_rate": 0.0001837581267119561, + "loss": 0.9797, + "step": 4149 + }, + { + "epoch": 0.738960113960114, + "grad_norm": 0.5698176026344299, + "learning_rate": 0.00018375047889393215, + "loss": 1.1099, + "step": 4150 + }, + { + "epoch": 0.7391381766381766, + "grad_norm": 0.5384604930877686, + "learning_rate": 0.00018374282943499156, + "loss": 1.1944, + "step": 4151 + }, + { + "epoch": 0.7393162393162394, + "grad_norm": 0.5483044385910034, + "learning_rate": 0.00018373517833528418, + "loss": 1.1734, + "step": 4152 + }, + { + "epoch": 0.739494301994302, + "grad_norm": 0.4824066162109375, + "learning_rate": 0.0001837275255949599, + "loss": 0.9515, + "step": 4153 + }, + { + "epoch": 0.7396723646723646, + "grad_norm": 0.45413634181022644, + "learning_rate": 0.00018371987121416873, + "loss": 0.7534, + "step": 4154 + }, + { + "epoch": 0.7398504273504274, + "grad_norm": 0.5874246954917908, + "learning_rate": 0.00018371221519306055, + "loss": 0.9464, + "step": 4155 + }, + { + "epoch": 0.74002849002849, + "grad_norm": 0.5219913125038147, + "learning_rate": 0.00018370455753178544, + "loss": 1.0494, + "step": 4156 + }, + { + "epoch": 0.7402065527065527, + "grad_norm": 0.5937709212303162, + "learning_rate": 0.00018369689823049341, + "loss": 1.0529, + "step": 4157 + }, + { + "epoch": 0.7403846153846154, + "grad_norm": 0.5204295516014099, + "learning_rate": 0.00018368923728933449, + "loss": 1.0602, + "step": 4158 + }, + { + "epoch": 0.7405626780626781, + "grad_norm": 0.5422890186309814, + "learning_rate": 0.00018368157470845885, + "loss": 0.9261, + "step": 4159 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.6163852214813232, + "learning_rate": 0.00018367391048801655, + "loss": 1.2771, + "step": 4160 + }, + { + "epoch": 0.7409188034188035, + "grad_norm": 0.5070751309394836, + "learning_rate": 0.00018366624462815785, + "loss": 1.0401, + "step": 4161 + }, + { + "epoch": 0.7410968660968661, + "grad_norm": 0.4477100968360901, + "learning_rate": 0.00018365857712903283, + "loss": 1.1463, + "step": 4162 + }, + { + "epoch": 0.7412749287749287, + "grad_norm": 0.5421462655067444, + "learning_rate": 0.0001836509079907918, + "loss": 0.9373, + "step": 4163 + }, + { + "epoch": 0.7414529914529915, + "grad_norm": 0.6162141561508179, + "learning_rate": 0.000183643237213585, + "loss": 1.1827, + "step": 4164 + }, + { + "epoch": 0.7416310541310541, + "grad_norm": 0.5653836131095886, + "learning_rate": 0.00018363556479756272, + "loss": 1.0689, + "step": 4165 + }, + { + "epoch": 0.7418091168091168, + "grad_norm": 0.57053542137146, + "learning_rate": 0.00018362789074287527, + "loss": 1.0289, + "step": 4166 + }, + { + "epoch": 0.7419871794871795, + "grad_norm": 0.5603055953979492, + "learning_rate": 0.00018362021504967304, + "loss": 1.1926, + "step": 4167 + }, + { + "epoch": 0.7421652421652422, + "grad_norm": 0.5460166335105896, + "learning_rate": 0.0001836125377181064, + "loss": 1.1488, + "step": 4168 + }, + { + "epoch": 0.7423433048433048, + "grad_norm": 0.5097107887268066, + "learning_rate": 0.00018360485874832579, + "loss": 1.0781, + "step": 4169 + }, + { + "epoch": 0.7425213675213675, + "grad_norm": 0.6280624270439148, + "learning_rate": 0.00018359717814048164, + "loss": 1.3625, + "step": 4170 + }, + { + "epoch": 0.7426994301994302, + "grad_norm": 0.4528210759162903, + "learning_rate": 0.0001835894958947244, + "loss": 0.8417, + "step": 4171 + }, + { + "epoch": 0.7428774928774928, + "grad_norm": 0.48735132813453674, + "learning_rate": 0.00018358181201120468, + "loss": 0.9544, + "step": 4172 + }, + { + "epoch": 0.7430555555555556, + "grad_norm": 0.48388174176216125, + "learning_rate": 0.00018357412649007296, + "loss": 1.0663, + "step": 4173 + }, + { + "epoch": 0.7432336182336182, + "grad_norm": 0.5435357689857483, + "learning_rate": 0.00018356643933147986, + "loss": 1.2074, + "step": 4174 + }, + { + "epoch": 0.7434116809116809, + "grad_norm": 0.49890074133872986, + "learning_rate": 0.00018355875053557594, + "loss": 1.1322, + "step": 4175 + }, + { + "epoch": 0.7435897435897436, + "grad_norm": 0.5680708885192871, + "learning_rate": 0.0001835510601025119, + "loss": 1.1964, + "step": 4176 + }, + { + "epoch": 0.7437678062678063, + "grad_norm": 0.5002360939979553, + "learning_rate": 0.00018354336803243842, + "loss": 1.1396, + "step": 4177 + }, + { + "epoch": 0.7439458689458689, + "grad_norm": 0.5202965140342712, + "learning_rate": 0.00018353567432550616, + "loss": 1.1498, + "step": 4178 + }, + { + "epoch": 0.7441239316239316, + "grad_norm": 0.514492928981781, + "learning_rate": 0.00018352797898186588, + "loss": 1.0959, + "step": 4179 + }, + { + "epoch": 0.7443019943019943, + "grad_norm": 0.6395383477210999, + "learning_rate": 0.0001835202820016684, + "loss": 1.2867, + "step": 4180 + }, + { + "epoch": 0.7444800569800569, + "grad_norm": 0.5489062070846558, + "learning_rate": 0.00018351258338506447, + "loss": 1.1638, + "step": 4181 + }, + { + "epoch": 0.7446581196581197, + "grad_norm": 0.5705671906471252, + "learning_rate": 0.00018350488313220498, + "loss": 0.9493, + "step": 4182 + }, + { + "epoch": 0.7448361823361823, + "grad_norm": 0.5404297709465027, + "learning_rate": 0.00018349718124324076, + "loss": 0.9876, + "step": 4183 + }, + { + "epoch": 0.7450142450142451, + "grad_norm": 0.5841003060340881, + "learning_rate": 0.0001834894777183227, + "loss": 1.1225, + "step": 4184 + }, + { + "epoch": 0.7451923076923077, + "grad_norm": 0.49774688482284546, + "learning_rate": 0.00018348177255760178, + "loss": 1.1442, + "step": 4185 + }, + { + "epoch": 0.7453703703703703, + "grad_norm": 0.5212422609329224, + "learning_rate": 0.00018347406576122894, + "loss": 1.101, + "step": 4186 + }, + { + "epoch": 0.7455484330484331, + "grad_norm": 0.615024983882904, + "learning_rate": 0.00018346635732935517, + "loss": 1.4188, + "step": 4187 + }, + { + "epoch": 0.7457264957264957, + "grad_norm": 0.46818843483924866, + "learning_rate": 0.00018345864726213154, + "loss": 1.0071, + "step": 4188 + }, + { + "epoch": 0.7459045584045584, + "grad_norm": 0.4921121895313263, + "learning_rate": 0.00018345093555970906, + "loss": 1.015, + "step": 4189 + }, + { + "epoch": 0.7460826210826211, + "grad_norm": 0.5042136311531067, + "learning_rate": 0.00018344322222223889, + "loss": 0.9974, + "step": 4190 + }, + { + "epoch": 0.7462606837606838, + "grad_norm": 0.5872490406036377, + "learning_rate": 0.0001834355072498721, + "loss": 1.3166, + "step": 4191 + }, + { + "epoch": 0.7464387464387464, + "grad_norm": 0.559117317199707, + "learning_rate": 0.00018342779064275984, + "loss": 1.2227, + "step": 4192 + }, + { + "epoch": 0.7466168091168092, + "grad_norm": 0.5269635319709778, + "learning_rate": 0.00018342007240105336, + "loss": 1.0281, + "step": 4193 + }, + { + "epoch": 0.7467948717948718, + "grad_norm": 0.4608335793018341, + "learning_rate": 0.00018341235252490387, + "loss": 0.98, + "step": 4194 + }, + { + "epoch": 0.7469729344729344, + "grad_norm": 0.5818259119987488, + "learning_rate": 0.00018340463101446255, + "loss": 1.1544, + "step": 4195 + }, + { + "epoch": 0.7471509971509972, + "grad_norm": 0.5577529668807983, + "learning_rate": 0.00018339690786988079, + "loss": 1.3059, + "step": 4196 + }, + { + "epoch": 0.7473290598290598, + "grad_norm": 0.5430468320846558, + "learning_rate": 0.00018338918309130983, + "loss": 1.2766, + "step": 4197 + }, + { + "epoch": 0.7475071225071225, + "grad_norm": 0.4941701591014862, + "learning_rate": 0.0001833814566789011, + "loss": 1.193, + "step": 4198 + }, + { + "epoch": 0.7476851851851852, + "grad_norm": 0.5471884608268738, + "learning_rate": 0.00018337372863280589, + "loss": 1.2261, + "step": 4199 + }, + { + "epoch": 0.7478632478632479, + "grad_norm": 0.4641438126564026, + "learning_rate": 0.0001833659989531757, + "loss": 0.7953, + "step": 4200 + }, + { + "epoch": 0.7480413105413105, + "grad_norm": 0.5244714617729187, + "learning_rate": 0.0001833582676401619, + "loss": 0.9344, + "step": 4201 + }, + { + "epoch": 0.7482193732193733, + "grad_norm": 0.5964360237121582, + "learning_rate": 0.00018335053469391603, + "loss": 1.2072, + "step": 4202 + }, + { + "epoch": 0.7483974358974359, + "grad_norm": 0.4929158091545105, + "learning_rate": 0.00018334280011458954, + "loss": 1.2183, + "step": 4203 + }, + { + "epoch": 0.7485754985754985, + "grad_norm": 0.46221864223480225, + "learning_rate": 0.00018333506390233405, + "loss": 1.1957, + "step": 4204 + }, + { + "epoch": 0.7487535612535613, + "grad_norm": 0.6301732659339905, + "learning_rate": 0.0001833273260573011, + "loss": 1.0582, + "step": 4205 + }, + { + "epoch": 0.7489316239316239, + "grad_norm": 0.5606021881103516, + "learning_rate": 0.0001833195865796423, + "loss": 1.4034, + "step": 4206 + }, + { + "epoch": 0.7491096866096866, + "grad_norm": 0.44856077432632446, + "learning_rate": 0.00018331184546950926, + "loss": 0.8421, + "step": 4207 + }, + { + "epoch": 0.7492877492877493, + "grad_norm": 0.5487226247787476, + "learning_rate": 0.00018330410272705366, + "loss": 1.238, + "step": 4208 + }, + { + "epoch": 0.749465811965812, + "grad_norm": 0.6043636798858643, + "learning_rate": 0.00018329635835242724, + "loss": 1.1215, + "step": 4209 + }, + { + "epoch": 0.7496438746438746, + "grad_norm": 0.5145319104194641, + "learning_rate": 0.00018328861234578173, + "loss": 1.1002, + "step": 4210 + }, + { + "epoch": 0.7498219373219374, + "grad_norm": 0.5667078495025635, + "learning_rate": 0.00018328086470726884, + "loss": 1.2994, + "step": 4211 + }, + { + "epoch": 0.75, + "grad_norm": 0.5117634534835815, + "learning_rate": 0.00018327311543704043, + "loss": 0.9448, + "step": 4212 + }, + { + "epoch": 0.75, + "eval_loss": 1.0982474088668823, + "eval_runtime": 24.6617, + "eval_samples_per_second": 42.211, + "eval_steps_per_second": 21.126, + "step": 4212 + }, + { + "epoch": 0.7501780626780626, + "grad_norm": 0.5451585054397583, + "learning_rate": 0.00018326536453524826, + "loss": 0.9023, + "step": 4213 + }, + { + "epoch": 0.7503561253561254, + "grad_norm": 0.6585208773612976, + "learning_rate": 0.0001832576120020443, + "loss": 1.2798, + "step": 4214 + }, + { + "epoch": 0.750534188034188, + "grad_norm": 0.6444812417030334, + "learning_rate": 0.00018324985783758037, + "loss": 1.3999, + "step": 4215 + }, + { + "epoch": 0.7507122507122507, + "grad_norm": 0.6178330779075623, + "learning_rate": 0.0001832421020420084, + "loss": 1.1846, + "step": 4216 + }, + { + "epoch": 0.7508903133903134, + "grad_norm": 0.509969174861908, + "learning_rate": 0.00018323434461548036, + "loss": 1.1831, + "step": 4217 + }, + { + "epoch": 0.7510683760683761, + "grad_norm": 0.5558911561965942, + "learning_rate": 0.00018322658555814826, + "loss": 1.1599, + "step": 4218 + }, + { + "epoch": 0.7512464387464387, + "grad_norm": 0.5714917778968811, + "learning_rate": 0.0001832188248701641, + "loss": 0.9702, + "step": 4219 + }, + { + "epoch": 0.7514245014245015, + "grad_norm": 0.6136442422866821, + "learning_rate": 0.00018321106255167995, + "loss": 0.9376, + "step": 4220 + }, + { + "epoch": 0.7516025641025641, + "grad_norm": 0.5832077264785767, + "learning_rate": 0.00018320329860284785, + "loss": 1.2564, + "step": 4221 + }, + { + "epoch": 0.7517806267806267, + "grad_norm": 0.45330923795700073, + "learning_rate": 0.00018319553302381997, + "loss": 0.9321, + "step": 4222 + }, + { + "epoch": 0.7519586894586895, + "grad_norm": 0.5278468132019043, + "learning_rate": 0.00018318776581474847, + "loss": 1.1334, + "step": 4223 + }, + { + "epoch": 0.7521367521367521, + "grad_norm": 0.49267473816871643, + "learning_rate": 0.00018317999697578549, + "loss": 1.1577, + "step": 4224 + }, + { + "epoch": 0.7523148148148148, + "grad_norm": 0.5372124314308167, + "learning_rate": 0.00018317222650708325, + "loss": 1.037, + "step": 4225 + }, + { + "epoch": 0.7524928774928775, + "grad_norm": 0.5879829525947571, + "learning_rate": 0.000183164454408794, + "loss": 1.1312, + "step": 4226 + }, + { + "epoch": 0.7526709401709402, + "grad_norm": 0.5363932251930237, + "learning_rate": 0.00018315668068107004, + "loss": 1.174, + "step": 4227 + }, + { + "epoch": 0.7528490028490028, + "grad_norm": 0.5585991740226746, + "learning_rate": 0.00018314890532406366, + "loss": 1.2106, + "step": 4228 + }, + { + "epoch": 0.7530270655270656, + "grad_norm": 0.49395787715911865, + "learning_rate": 0.0001831411283379272, + "loss": 1.1163, + "step": 4229 + }, + { + "epoch": 0.7532051282051282, + "grad_norm": 0.5081066489219666, + "learning_rate": 0.00018313334972281306, + "loss": 1.184, + "step": 4230 + }, + { + "epoch": 0.7533831908831908, + "grad_norm": 0.40304034948349, + "learning_rate": 0.0001831255694788736, + "loss": 0.7548, + "step": 4231 + }, + { + "epoch": 0.7535612535612536, + "grad_norm": 0.4999815821647644, + "learning_rate": 0.0001831177876062613, + "loss": 1.0092, + "step": 4232 + }, + { + "epoch": 0.7537393162393162, + "grad_norm": 0.48917025327682495, + "learning_rate": 0.00018311000410512862, + "loss": 1.0354, + "step": 4233 + }, + { + "epoch": 0.7539173789173789, + "grad_norm": 0.475606769323349, + "learning_rate": 0.00018310221897562806, + "loss": 0.8728, + "step": 4234 + }, + { + "epoch": 0.7540954415954416, + "grad_norm": 0.630439817905426, + "learning_rate": 0.00018309443221791214, + "loss": 1.1436, + "step": 4235 + }, + { + "epoch": 0.7542735042735043, + "grad_norm": 0.524740993976593, + "learning_rate": 0.00018308664383213344, + "loss": 1.0487, + "step": 4236 + }, + { + "epoch": 0.7544515669515669, + "grad_norm": 0.4734523892402649, + "learning_rate": 0.0001830788538184445, + "loss": 1.0681, + "step": 4237 + }, + { + "epoch": 0.7546296296296297, + "grad_norm": 0.5767266750335693, + "learning_rate": 0.00018307106217699807, + "loss": 1.0599, + "step": 4238 + }, + { + "epoch": 0.7548076923076923, + "grad_norm": 0.6276642084121704, + "learning_rate": 0.0001830632689079467, + "loss": 1.2837, + "step": 4239 + }, + { + "epoch": 0.7549857549857549, + "grad_norm": 0.5539988279342651, + "learning_rate": 0.00018305547401144316, + "loss": 0.9072, + "step": 4240 + }, + { + "epoch": 0.7551638176638177, + "grad_norm": 0.4551292061805725, + "learning_rate": 0.00018304767748764014, + "loss": 1.0204, + "step": 4241 + }, + { + "epoch": 0.7553418803418803, + "grad_norm": 0.47344550490379333, + "learning_rate": 0.00018303987933669034, + "loss": 1.0473, + "step": 4242 + }, + { + "epoch": 0.7555199430199431, + "grad_norm": 0.6050213575363159, + "learning_rate": 0.00018303207955874665, + "loss": 1.1552, + "step": 4243 + }, + { + "epoch": 0.7556980056980057, + "grad_norm": 0.48943889141082764, + "learning_rate": 0.00018302427815396186, + "loss": 1.0002, + "step": 4244 + }, + { + "epoch": 0.7558760683760684, + "grad_norm": 0.5664682984352112, + "learning_rate": 0.00018301647512248878, + "loss": 1.1865, + "step": 4245 + }, + { + "epoch": 0.7560541310541311, + "grad_norm": 0.5702242255210876, + "learning_rate": 0.00018300867046448034, + "loss": 1.3029, + "step": 4246 + }, + { + "epoch": 0.7562321937321937, + "grad_norm": 0.593207836151123, + "learning_rate": 0.00018300086418008942, + "loss": 1.109, + "step": 4247 + }, + { + "epoch": 0.7564102564102564, + "grad_norm": 0.5887887477874756, + "learning_rate": 0.000182993056269469, + "loss": 1.3022, + "step": 4248 + }, + { + "epoch": 0.7565883190883191, + "grad_norm": 0.5277966260910034, + "learning_rate": 0.00018298524673277203, + "loss": 1.1738, + "step": 4249 + }, + { + "epoch": 0.7567663817663818, + "grad_norm": 0.589347779750824, + "learning_rate": 0.00018297743557015155, + "loss": 1.0185, + "step": 4250 + }, + { + "epoch": 0.7569444444444444, + "grad_norm": 0.49920859932899475, + "learning_rate": 0.0001829696227817606, + "loss": 1.118, + "step": 4251 + }, + { + "epoch": 0.7571225071225072, + "grad_norm": 0.502565324306488, + "learning_rate": 0.0001829618083677522, + "loss": 1.1856, + "step": 4252 + }, + { + "epoch": 0.7573005698005698, + "grad_norm": 0.49814435839653015, + "learning_rate": 0.00018295399232827955, + "loss": 1.0432, + "step": 4253 + }, + { + "epoch": 0.7574786324786325, + "grad_norm": 0.5087502598762512, + "learning_rate": 0.00018294617466349574, + "loss": 1.2325, + "step": 4254 + }, + { + "epoch": 0.7576566951566952, + "grad_norm": 0.5107288956642151, + "learning_rate": 0.00018293835537355394, + "loss": 1.0487, + "step": 4255 + }, + { + "epoch": 0.7578347578347578, + "grad_norm": 0.524725615978241, + "learning_rate": 0.00018293053445860732, + "loss": 1.1821, + "step": 4256 + }, + { + "epoch": 0.7580128205128205, + "grad_norm": 0.5234082937240601, + "learning_rate": 0.0001829227119188092, + "loss": 0.8896, + "step": 4257 + }, + { + "epoch": 0.7581908831908832, + "grad_norm": 0.5102918744087219, + "learning_rate": 0.00018291488775431275, + "loss": 1.0246, + "step": 4258 + }, + { + "epoch": 0.7583689458689459, + "grad_norm": 0.5552714467048645, + "learning_rate": 0.00018290706196527135, + "loss": 1.0193, + "step": 4259 + }, + { + "epoch": 0.7585470085470085, + "grad_norm": 0.5395022630691528, + "learning_rate": 0.00018289923455183825, + "loss": 1.3203, + "step": 4260 + }, + { + "epoch": 0.7587250712250713, + "grad_norm": 0.7474865913391113, + "learning_rate": 0.00018289140551416692, + "loss": 1.182, + "step": 4261 + }, + { + "epoch": 0.7589031339031339, + "grad_norm": 0.4892016649246216, + "learning_rate": 0.00018288357485241066, + "loss": 0.968, + "step": 4262 + }, + { + "epoch": 0.7590811965811965, + "grad_norm": 0.4627816081047058, + "learning_rate": 0.00018287574256672291, + "loss": 0.6895, + "step": 4263 + }, + { + "epoch": 0.7592592592592593, + "grad_norm": 0.6221280097961426, + "learning_rate": 0.00018286790865725715, + "loss": 0.9691, + "step": 4264 + }, + { + "epoch": 0.7594373219373219, + "grad_norm": 0.5542295575141907, + "learning_rate": 0.0001828600731241669, + "loss": 0.9996, + "step": 4265 + }, + { + "epoch": 0.7596153846153846, + "grad_norm": 0.5570770502090454, + "learning_rate": 0.00018285223596760562, + "loss": 1.1996, + "step": 4266 + }, + { + "epoch": 0.7597934472934473, + "grad_norm": 0.5495262742042542, + "learning_rate": 0.00018284439718772687, + "loss": 1.1572, + "step": 4267 + }, + { + "epoch": 0.75997150997151, + "grad_norm": 0.5006741881370544, + "learning_rate": 0.00018283655678468427, + "loss": 1.1215, + "step": 4268 + }, + { + "epoch": 0.7601495726495726, + "grad_norm": 0.4682157635688782, + "learning_rate": 0.00018282871475863144, + "loss": 1.0547, + "step": 4269 + }, + { + "epoch": 0.7603276353276354, + "grad_norm": 0.6275840997695923, + "learning_rate": 0.00018282087110972197, + "loss": 1.3855, + "step": 4270 + }, + { + "epoch": 0.760505698005698, + "grad_norm": 0.5341474413871765, + "learning_rate": 0.0001828130258381096, + "loss": 1.2024, + "step": 4271 + }, + { + "epoch": 0.7606837606837606, + "grad_norm": 0.4330833852291107, + "learning_rate": 0.000182805178943948, + "loss": 1.0508, + "step": 4272 + }, + { + "epoch": 0.7608618233618234, + "grad_norm": 0.6276537179946899, + "learning_rate": 0.00018279733042739094, + "loss": 1.1635, + "step": 4273 + }, + { + "epoch": 0.761039886039886, + "grad_norm": 0.5370199084281921, + "learning_rate": 0.00018278948028859217, + "loss": 1.0579, + "step": 4274 + }, + { + "epoch": 0.7612179487179487, + "grad_norm": 0.524959921836853, + "learning_rate": 0.00018278162852770552, + "loss": 1.0972, + "step": 4275 + }, + { + "epoch": 0.7613960113960114, + "grad_norm": 0.5029389262199402, + "learning_rate": 0.00018277377514488486, + "loss": 0.959, + "step": 4276 + }, + { + "epoch": 0.7615740740740741, + "grad_norm": 0.49772894382476807, + "learning_rate": 0.00018276592014028397, + "loss": 1.2773, + "step": 4277 + }, + { + "epoch": 0.7617521367521367, + "grad_norm": 0.5195719003677368, + "learning_rate": 0.00018275806351405685, + "loss": 1.0676, + "step": 4278 + }, + { + "epoch": 0.7619301994301995, + "grad_norm": 0.5167942643165588, + "learning_rate": 0.00018275020526635735, + "loss": 1.0615, + "step": 4279 + }, + { + "epoch": 0.7621082621082621, + "grad_norm": 0.4958035945892334, + "learning_rate": 0.0001827423453973395, + "loss": 0.9605, + "step": 4280 + }, + { + "epoch": 0.7622863247863247, + "grad_norm": 0.6256808042526245, + "learning_rate": 0.00018273448390715728, + "loss": 1.2526, + "step": 4281 + }, + { + "epoch": 0.7624643874643875, + "grad_norm": 0.5062580108642578, + "learning_rate": 0.0001827266207959647, + "loss": 1.0604, + "step": 4282 + }, + { + "epoch": 0.7626424501424501, + "grad_norm": 0.5080778002738953, + "learning_rate": 0.00018271875606391583, + "loss": 1.1246, + "step": 4283 + }, + { + "epoch": 0.7628205128205128, + "grad_norm": 0.5069389939308167, + "learning_rate": 0.00018271088971116479, + "loss": 1.3158, + "step": 4284 + }, + { + "epoch": 0.7629985754985755, + "grad_norm": 0.7280121445655823, + "learning_rate": 0.00018270302173786567, + "loss": 1.2066, + "step": 4285 + }, + { + "epoch": 0.7631766381766382, + "grad_norm": 0.6523470282554626, + "learning_rate": 0.00018269515214417267, + "loss": 1.3236, + "step": 4286 + }, + { + "epoch": 0.7633547008547008, + "grad_norm": 0.5799322724342346, + "learning_rate": 0.00018268728093023988, + "loss": 0.9786, + "step": 4287 + }, + { + "epoch": 0.7635327635327636, + "grad_norm": 0.46675166487693787, + "learning_rate": 0.00018267940809622163, + "loss": 0.8131, + "step": 4288 + }, + { + "epoch": 0.7637108262108262, + "grad_norm": 0.5566182732582092, + "learning_rate": 0.00018267153364227214, + "loss": 1.0565, + "step": 4289 + }, + { + "epoch": 0.7638888888888888, + "grad_norm": 0.532028079032898, + "learning_rate": 0.00018266365756854566, + "loss": 0.952, + "step": 4290 + }, + { + "epoch": 0.7640669515669516, + "grad_norm": 0.5082666873931885, + "learning_rate": 0.00018265577987519653, + "loss": 1.0704, + "step": 4291 + }, + { + "epoch": 0.7642450142450142, + "grad_norm": 0.5223562717437744, + "learning_rate": 0.00018264790056237912, + "loss": 1.1161, + "step": 4292 + }, + { + "epoch": 0.7644230769230769, + "grad_norm": 0.48472318053245544, + "learning_rate": 0.00018264001963024778, + "loss": 0.8784, + "step": 4293 + }, + { + "epoch": 0.7646011396011396, + "grad_norm": 0.5901281833648682, + "learning_rate": 0.0001826321370789569, + "loss": 1.1031, + "step": 4294 + }, + { + "epoch": 0.7647792022792023, + "grad_norm": 0.570350706577301, + "learning_rate": 0.000182624252908661, + "loss": 0.9047, + "step": 4295 + }, + { + "epoch": 0.7649572649572649, + "grad_norm": 0.568373441696167, + "learning_rate": 0.00018261636711951445, + "loss": 1.0106, + "step": 4296 + }, + { + "epoch": 0.7651353276353277, + "grad_norm": 0.6175880432128906, + "learning_rate": 0.00018260847971167182, + "loss": 1.3531, + "step": 4297 + }, + { + "epoch": 0.7653133903133903, + "grad_norm": 0.5682594776153564, + "learning_rate": 0.00018260059068528762, + "loss": 1.1261, + "step": 4298 + }, + { + "epoch": 0.7654914529914529, + "grad_norm": 0.5050225257873535, + "learning_rate": 0.00018259270004051644, + "loss": 1.0921, + "step": 4299 + }, + { + "epoch": 0.7656695156695157, + "grad_norm": 0.5416565537452698, + "learning_rate": 0.0001825848077775129, + "loss": 1.0881, + "step": 4300 + }, + { + "epoch": 0.7658475783475783, + "grad_norm": 0.5418867468833923, + "learning_rate": 0.0001825769138964316, + "loss": 1.2069, + "step": 4301 + }, + { + "epoch": 0.7660256410256411, + "grad_norm": 0.5447866320610046, + "learning_rate": 0.00018256901839742718, + "loss": 1.1827, + "step": 4302 + }, + { + "epoch": 0.7662037037037037, + "grad_norm": 0.5482802987098694, + "learning_rate": 0.00018256112128065439, + "loss": 1.0492, + "step": 4303 + }, + { + "epoch": 0.7663817663817664, + "grad_norm": 0.5059601664543152, + "learning_rate": 0.0001825532225462679, + "loss": 1.0996, + "step": 4304 + }, + { + "epoch": 0.7665598290598291, + "grad_norm": 0.5153701901435852, + "learning_rate": 0.00018254532219442258, + "loss": 1.3237, + "step": 4305 + }, + { + "epoch": 0.7667378917378918, + "grad_norm": 0.5370768904685974, + "learning_rate": 0.0001825374202252731, + "loss": 0.9925, + "step": 4306 + }, + { + "epoch": 0.7669159544159544, + "grad_norm": 0.4516580402851105, + "learning_rate": 0.00018252951663897432, + "loss": 1.0749, + "step": 4307 + }, + { + "epoch": 0.7670940170940171, + "grad_norm": 0.5565171837806702, + "learning_rate": 0.0001825216114356811, + "loss": 1.1617, + "step": 4308 + }, + { + "epoch": 0.7672720797720798, + "grad_norm": 0.5212662220001221, + "learning_rate": 0.00018251370461554834, + "loss": 1.1108, + "step": 4309 + }, + { + "epoch": 0.7674501424501424, + "grad_norm": 0.49061715602874756, + "learning_rate": 0.00018250579617873095, + "loss": 1.0881, + "step": 4310 + }, + { + "epoch": 0.7676282051282052, + "grad_norm": 0.5535751581192017, + "learning_rate": 0.00018249788612538387, + "loss": 0.9341, + "step": 4311 + }, + { + "epoch": 0.7678062678062678, + "grad_norm": 0.5425209403038025, + "learning_rate": 0.00018248997445566208, + "loss": 1.1858, + "step": 4312 + }, + { + "epoch": 0.7679843304843305, + "grad_norm": 0.6224395036697388, + "learning_rate": 0.0001824820611697206, + "loss": 1.0836, + "step": 4313 + }, + { + "epoch": 0.7681623931623932, + "grad_norm": 0.4895690977573395, + "learning_rate": 0.00018247414626771445, + "loss": 0.8598, + "step": 4314 + }, + { + "epoch": 0.7683404558404558, + "grad_norm": 0.5279615521430969, + "learning_rate": 0.00018246622974979877, + "loss": 1.1742, + "step": 4315 + }, + { + "epoch": 0.7685185185185185, + "grad_norm": 0.45300471782684326, + "learning_rate": 0.0001824583116161286, + "loss": 0.8872, + "step": 4316 + }, + { + "epoch": 0.7686965811965812, + "grad_norm": 0.6499692797660828, + "learning_rate": 0.00018245039186685916, + "loss": 1.2495, + "step": 4317 + }, + { + "epoch": 0.7688746438746439, + "grad_norm": 0.48151278495788574, + "learning_rate": 0.00018244247050214552, + "loss": 1.2382, + "step": 4318 + }, + { + "epoch": 0.7690527065527065, + "grad_norm": 0.6597028374671936, + "learning_rate": 0.0001824345475221429, + "loss": 1.3453, + "step": 4319 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.4536992609500885, + "learning_rate": 0.0001824266229270066, + "loss": 1.1141, + "step": 4320 + }, + { + "epoch": 0.7694088319088319, + "grad_norm": 0.5489405393600464, + "learning_rate": 0.00018241869671689184, + "loss": 1.0333, + "step": 4321 + }, + { + "epoch": 0.7695868945868946, + "grad_norm": 0.5741586089134216, + "learning_rate": 0.00018241076889195394, + "loss": 0.9939, + "step": 4322 + }, + { + "epoch": 0.7697649572649573, + "grad_norm": 0.47170960903167725, + "learning_rate": 0.00018240283945234823, + "loss": 0.9878, + "step": 4323 + }, + { + "epoch": 0.76994301994302, + "grad_norm": 0.4729093313217163, + "learning_rate": 0.00018239490839823004, + "loss": 1.0087, + "step": 4324 + }, + { + "epoch": 0.7701210826210826, + "grad_norm": 0.49869823455810547, + "learning_rate": 0.0001823869757297548, + "loss": 1.169, + "step": 4325 + }, + { + "epoch": 0.7702991452991453, + "grad_norm": 0.5118468403816223, + "learning_rate": 0.0001823790414470779, + "loss": 1.1092, + "step": 4326 + }, + { + "epoch": 0.770477207977208, + "grad_norm": 0.5076048970222473, + "learning_rate": 0.0001823711055503548, + "loss": 1.1028, + "step": 4327 + }, + { + "epoch": 0.7706552706552706, + "grad_norm": 0.5661569237709045, + "learning_rate": 0.00018236316803974098, + "loss": 1.1114, + "step": 4328 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 0.5542354583740234, + "learning_rate": 0.000182355228915392, + "loss": 1.0931, + "step": 4329 + }, + { + "epoch": 0.771011396011396, + "grad_norm": 0.5476680994033813, + "learning_rate": 0.0001823472881774634, + "loss": 1.036, + "step": 4330 + }, + { + "epoch": 0.7711894586894587, + "grad_norm": 0.5449798703193665, + "learning_rate": 0.00018233934582611073, + "loss": 1.0682, + "step": 4331 + }, + { + "epoch": 0.7713675213675214, + "grad_norm": 0.61089026927948, + "learning_rate": 0.00018233140186148963, + "loss": 1.0748, + "step": 4332 + }, + { + "epoch": 0.771545584045584, + "grad_norm": 0.5015206336975098, + "learning_rate": 0.00018232345628375576, + "loss": 1.2032, + "step": 4333 + }, + { + "epoch": 0.7717236467236467, + "grad_norm": 0.579289972782135, + "learning_rate": 0.00018231550909306475, + "loss": 1.0764, + "step": 4334 + }, + { + "epoch": 0.7719017094017094, + "grad_norm": 0.5889299511909485, + "learning_rate": 0.00018230756028957235, + "loss": 1.1768, + "step": 4335 + }, + { + "epoch": 0.7720797720797721, + "grad_norm": 0.5328249335289001, + "learning_rate": 0.00018229960987343428, + "loss": 1.0055, + "step": 4336 + }, + { + "epoch": 0.7722578347578347, + "grad_norm": 0.5766382217407227, + "learning_rate": 0.0001822916578448063, + "loss": 0.9923, + "step": 4337 + }, + { + "epoch": 0.7724358974358975, + "grad_norm": 0.6448187828063965, + "learning_rate": 0.00018228370420384423, + "loss": 1.1135, + "step": 4338 + }, + { + "epoch": 0.7726139601139601, + "grad_norm": 0.5505210757255554, + "learning_rate": 0.00018227574895070394, + "loss": 1.2048, + "step": 4339 + }, + { + "epoch": 0.7727920227920227, + "grad_norm": 0.6278925538063049, + "learning_rate": 0.00018226779208554126, + "loss": 1.1045, + "step": 4340 + }, + { + "epoch": 0.7729700854700855, + "grad_norm": 0.5345009565353394, + "learning_rate": 0.00018225983360851207, + "loss": 1.0102, + "step": 4341 + }, + { + "epoch": 0.7731481481481481, + "grad_norm": 0.566633403301239, + "learning_rate": 0.00018225187351977233, + "loss": 1.0038, + "step": 4342 + }, + { + "epoch": 0.7733262108262108, + "grad_norm": 0.5066078901290894, + "learning_rate": 0.000182243911819478, + "loss": 1.0339, + "step": 4343 + }, + { + "epoch": 0.7735042735042735, + "grad_norm": 0.5614920258522034, + "learning_rate": 0.00018223594850778503, + "loss": 1.1021, + "step": 4344 + }, + { + "epoch": 0.7736823361823362, + "grad_norm": 0.7747337818145752, + "learning_rate": 0.0001822279835848495, + "loss": 1.1129, + "step": 4345 + }, + { + "epoch": 0.7738603988603988, + "grad_norm": 0.7066529989242554, + "learning_rate": 0.00018222001705082744, + "loss": 1.3234, + "step": 4346 + }, + { + "epoch": 0.7740384615384616, + "grad_norm": 0.6340884566307068, + "learning_rate": 0.00018221204890587497, + "loss": 1.0726, + "step": 4347 + }, + { + "epoch": 0.7742165242165242, + "grad_norm": 0.5401145815849304, + "learning_rate": 0.00018220407915014818, + "loss": 0.9904, + "step": 4348 + }, + { + "epoch": 0.7743945868945868, + "grad_norm": 0.5069159269332886, + "learning_rate": 0.00018219610778380315, + "loss": 1.0654, + "step": 4349 + }, + { + "epoch": 0.7745726495726496, + "grad_norm": 0.5422839522361755, + "learning_rate": 0.00018218813480699623, + "loss": 1.1741, + "step": 4350 + }, + { + "epoch": 0.7747507122507122, + "grad_norm": 0.5550300478935242, + "learning_rate": 0.0001821801602198835, + "loss": 1.0033, + "step": 4351 + }, + { + "epoch": 0.7749287749287749, + "grad_norm": 0.5987736582756042, + "learning_rate": 0.00018217218402262123, + "loss": 0.935, + "step": 4352 + }, + { + "epoch": 0.7751068376068376, + "grad_norm": 0.6137008666992188, + "learning_rate": 0.00018216420621536573, + "loss": 1.17, + "step": 4353 + }, + { + "epoch": 0.7752849002849003, + "grad_norm": 0.47124359011650085, + "learning_rate": 0.0001821562267982733, + "loss": 0.8316, + "step": 4354 + }, + { + "epoch": 0.7754629629629629, + "grad_norm": 0.5057868361473083, + "learning_rate": 0.00018214824577150024, + "loss": 1.0246, + "step": 4355 + }, + { + "epoch": 0.7756410256410257, + "grad_norm": 0.604055643081665, + "learning_rate": 0.00018214026313520299, + "loss": 1.1272, + "step": 4356 + }, + { + "epoch": 0.7758190883190883, + "grad_norm": 0.6690384149551392, + "learning_rate": 0.0001821322788895379, + "loss": 1.0464, + "step": 4357 + }, + { + "epoch": 0.7759971509971509, + "grad_norm": 0.5458958745002747, + "learning_rate": 0.0001821242930346614, + "loss": 1.1712, + "step": 4358 + }, + { + "epoch": 0.7761752136752137, + "grad_norm": 0.6448663473129272, + "learning_rate": 0.00018211630557073, + "loss": 1.1125, + "step": 4359 + }, + { + "epoch": 0.7763532763532763, + "grad_norm": 0.49889448285102844, + "learning_rate": 0.00018210831649790018, + "loss": 1.097, + "step": 4360 + }, + { + "epoch": 0.7765313390313391, + "grad_norm": 0.5118046998977661, + "learning_rate": 0.00018210032581632843, + "loss": 1.009, + "step": 4361 + }, + { + "epoch": 0.7767094017094017, + "grad_norm": 0.5450068116188049, + "learning_rate": 0.00018209233352617135, + "loss": 1.1138, + "step": 4362 + }, + { + "epoch": 0.7768874643874644, + "grad_norm": 0.6147481203079224, + "learning_rate": 0.00018208433962758558, + "loss": 1.212, + "step": 4363 + }, + { + "epoch": 0.7770655270655271, + "grad_norm": 0.554176926612854, + "learning_rate": 0.00018207634412072764, + "loss": 1.1271, + "step": 4364 + }, + { + "epoch": 0.7772435897435898, + "grad_norm": 0.5872851014137268, + "learning_rate": 0.00018206834700575426, + "loss": 1.2793, + "step": 4365 + }, + { + "epoch": 0.7774216524216524, + "grad_norm": 0.5135685205459595, + "learning_rate": 0.00018206034828282207, + "loss": 0.9642, + "step": 4366 + }, + { + "epoch": 0.7775997150997151, + "grad_norm": 0.5699490308761597, + "learning_rate": 0.00018205234795208786, + "loss": 0.9086, + "step": 4367 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 0.5908057689666748, + "learning_rate": 0.00018204434601370832, + "loss": 1.1973, + "step": 4368 + }, + { + "epoch": 0.7779558404558404, + "grad_norm": 0.5777581334114075, + "learning_rate": 0.00018203634246784025, + "loss": 1.0447, + "step": 4369 + }, + { + "epoch": 0.7781339031339032, + "grad_norm": 0.4822927713394165, + "learning_rate": 0.00018202833731464048, + "loss": 0.814, + "step": 4370 + }, + { + "epoch": 0.7783119658119658, + "grad_norm": 0.5343610644340515, + "learning_rate": 0.0001820203305542658, + "loss": 1.2785, + "step": 4371 + }, + { + "epoch": 0.7784900284900285, + "grad_norm": 0.5462222695350647, + "learning_rate": 0.00018201232218687316, + "loss": 1.1785, + "step": 4372 + }, + { + "epoch": 0.7786680911680912, + "grad_norm": 0.5177609324455261, + "learning_rate": 0.00018200431221261943, + "loss": 1.111, + "step": 4373 + }, + { + "epoch": 0.7788461538461539, + "grad_norm": 0.5324625968933105, + "learning_rate": 0.00018199630063166157, + "loss": 1.0738, + "step": 4374 + }, + { + "epoch": 0.7790242165242165, + "grad_norm": 0.6392876505851746, + "learning_rate": 0.0001819882874441565, + "loss": 1.1758, + "step": 4375 + }, + { + "epoch": 0.7792022792022792, + "grad_norm": 0.49964696168899536, + "learning_rate": 0.00018198027265026127, + "loss": 1.0556, + "step": 4376 + }, + { + "epoch": 0.7793803418803419, + "grad_norm": 0.6090660691261292, + "learning_rate": 0.00018197225625013287, + "loss": 1.0102, + "step": 4377 + }, + { + "epoch": 0.7795584045584045, + "grad_norm": 0.5242345929145813, + "learning_rate": 0.00018196423824392842, + "loss": 0.8335, + "step": 4378 + }, + { + "epoch": 0.7797364672364673, + "grad_norm": 0.5265036225318909, + "learning_rate": 0.00018195621863180498, + "loss": 1.0781, + "step": 4379 + }, + { + "epoch": 0.7799145299145299, + "grad_norm": 0.5115378499031067, + "learning_rate": 0.0001819481974139197, + "loss": 1.1658, + "step": 4380 + }, + { + "epoch": 0.7800925925925926, + "grad_norm": 0.6489549875259399, + "learning_rate": 0.00018194017459042972, + "loss": 1.0572, + "step": 4381 + }, + { + "epoch": 0.7802706552706553, + "grad_norm": 0.5800202488899231, + "learning_rate": 0.0001819321501614922, + "loss": 0.9593, + "step": 4382 + }, + { + "epoch": 0.780448717948718, + "grad_norm": 0.5608528256416321, + "learning_rate": 0.00018192412412726443, + "loss": 1.0324, + "step": 4383 + }, + { + "epoch": 0.7806267806267806, + "grad_norm": 0.5596401691436768, + "learning_rate": 0.00018191609648790362, + "loss": 1.071, + "step": 4384 + }, + { + "epoch": 0.7808048433048433, + "grad_norm": 0.5712903141975403, + "learning_rate": 0.00018190806724356707, + "loss": 0.9011, + "step": 4385 + }, + { + "epoch": 0.780982905982906, + "grad_norm": 0.5079438090324402, + "learning_rate": 0.0001819000363944121, + "loss": 1.1194, + "step": 4386 + }, + { + "epoch": 0.7811609686609686, + "grad_norm": 0.5785079598426819, + "learning_rate": 0.00018189200394059602, + "loss": 1.1703, + "step": 4387 + }, + { + "epoch": 0.7813390313390314, + "grad_norm": 0.6901816129684448, + "learning_rate": 0.00018188396988227625, + "loss": 1.6689, + "step": 4388 + }, + { + "epoch": 0.781517094017094, + "grad_norm": 0.48107922077178955, + "learning_rate": 0.00018187593421961022, + "loss": 1.0116, + "step": 4389 + }, + { + "epoch": 0.7816951566951567, + "grad_norm": 0.5843084454536438, + "learning_rate": 0.0001818678969527553, + "loss": 1.1172, + "step": 4390 + }, + { + "epoch": 0.7818732193732194, + "grad_norm": 0.479034423828125, + "learning_rate": 0.00018185985808186902, + "loss": 0.811, + "step": 4391 + }, + { + "epoch": 0.782051282051282, + "grad_norm": 0.5864158272743225, + "learning_rate": 0.00018185181760710888, + "loss": 0.9522, + "step": 4392 + }, + { + "epoch": 0.7822293447293447, + "grad_norm": 0.4824625551700592, + "learning_rate": 0.00018184377552863242, + "loss": 0.9039, + "step": 4393 + }, + { + "epoch": 0.7824074074074074, + "grad_norm": 0.580102801322937, + "learning_rate": 0.00018183573184659717, + "loss": 1.2382, + "step": 4394 + }, + { + "epoch": 0.7825854700854701, + "grad_norm": 0.5300056338310242, + "learning_rate": 0.00018182768656116073, + "loss": 1.2268, + "step": 4395 + }, + { + "epoch": 0.7827635327635327, + "grad_norm": 0.5548123121261597, + "learning_rate": 0.00018181963967248078, + "loss": 1.0628, + "step": 4396 + }, + { + "epoch": 0.7829415954415955, + "grad_norm": 0.5485070943832397, + "learning_rate": 0.00018181159118071496, + "loss": 0.9628, + "step": 4397 + }, + { + "epoch": 0.7831196581196581, + "grad_norm": 0.47405415773391724, + "learning_rate": 0.00018180354108602095, + "loss": 1.1413, + "step": 4398 + }, + { + "epoch": 0.7832977207977208, + "grad_norm": 0.5545752644538879, + "learning_rate": 0.0001817954893885565, + "loss": 1.3807, + "step": 4399 + }, + { + "epoch": 0.7834757834757835, + "grad_norm": 0.5339497327804565, + "learning_rate": 0.00018178743608847933, + "loss": 0.9978, + "step": 4400 + }, + { + "epoch": 0.7836538461538461, + "grad_norm": 0.5006352663040161, + "learning_rate": 0.00018177938118594725, + "loss": 0.8873, + "step": 4401 + }, + { + "epoch": 0.7838319088319088, + "grad_norm": 0.4845179319381714, + "learning_rate": 0.00018177132468111812, + "loss": 0.8866, + "step": 4402 + }, + { + "epoch": 0.7840099715099715, + "grad_norm": 0.5240967869758606, + "learning_rate": 0.0001817632665741497, + "loss": 1.0347, + "step": 4403 + }, + { + "epoch": 0.7841880341880342, + "grad_norm": 0.5311884880065918, + "learning_rate": 0.00018175520686519993, + "loss": 1.2065, + "step": 4404 + }, + { + "epoch": 0.7843660968660968, + "grad_norm": 0.5562815070152283, + "learning_rate": 0.00018174714555442673, + "loss": 1.1272, + "step": 4405 + }, + { + "epoch": 0.7845441595441596, + "grad_norm": 0.5524366497993469, + "learning_rate": 0.00018173908264198802, + "loss": 1.2337, + "step": 4406 + }, + { + "epoch": 0.7847222222222222, + "grad_norm": 0.5612216591835022, + "learning_rate": 0.0001817310181280418, + "loss": 1.1809, + "step": 4407 + }, + { + "epoch": 0.7849002849002849, + "grad_norm": 0.5315343737602234, + "learning_rate": 0.000181722952012746, + "loss": 1.0491, + "step": 4408 + }, + { + "epoch": 0.7850783475783476, + "grad_norm": 0.5233435034751892, + "learning_rate": 0.00018171488429625878, + "loss": 1.0457, + "step": 4409 + }, + { + "epoch": 0.7852564102564102, + "grad_norm": 0.7809093594551086, + "learning_rate": 0.00018170681497873813, + "loss": 1.1578, + "step": 4410 + }, + { + "epoch": 0.7854344729344729, + "grad_norm": 0.49659839272499084, + "learning_rate": 0.00018169874406034217, + "loss": 1.0815, + "step": 4411 + }, + { + "epoch": 0.7856125356125356, + "grad_norm": 0.5020765066146851, + "learning_rate": 0.00018169067154122904, + "loss": 1.1985, + "step": 4412 + }, + { + "epoch": 0.7857905982905983, + "grad_norm": 0.6408432126045227, + "learning_rate": 0.0001816825974215569, + "loss": 1.2272, + "step": 4413 + }, + { + "epoch": 0.7859686609686609, + "grad_norm": 0.5062605142593384, + "learning_rate": 0.00018167452170148396, + "loss": 0.9663, + "step": 4414 + }, + { + "epoch": 0.7861467236467237, + "grad_norm": 0.5100119113922119, + "learning_rate": 0.0001816664443811684, + "loss": 1.0256, + "step": 4415 + }, + { + "epoch": 0.7863247863247863, + "grad_norm": 0.5277643799781799, + "learning_rate": 0.00018165836546076854, + "loss": 1.2885, + "step": 4416 + }, + { + "epoch": 0.7865028490028491, + "grad_norm": 0.5568150281906128, + "learning_rate": 0.0001816502849404426, + "loss": 1.2673, + "step": 4417 + }, + { + "epoch": 0.7866809116809117, + "grad_norm": 0.5061392188072205, + "learning_rate": 0.00018164220282034896, + "loss": 1.072, + "step": 4418 + }, + { + "epoch": 0.7868589743589743, + "grad_norm": 0.5383077263832092, + "learning_rate": 0.00018163411910064597, + "loss": 1.0621, + "step": 4419 + }, + { + "epoch": 0.7870370370370371, + "grad_norm": 0.5167948007583618, + "learning_rate": 0.00018162603378149198, + "loss": 1.099, + "step": 4420 + }, + { + "epoch": 0.7872150997150997, + "grad_norm": 0.5084534287452698, + "learning_rate": 0.0001816179468630454, + "loss": 1.3984, + "step": 4421 + }, + { + "epoch": 0.7873931623931624, + "grad_norm": 0.608762264251709, + "learning_rate": 0.00018160985834546475, + "loss": 1.3553, + "step": 4422 + }, + { + "epoch": 0.7875712250712251, + "grad_norm": 0.4900866746902466, + "learning_rate": 0.00018160176822890842, + "loss": 1.0009, + "step": 4423 + }, + { + "epoch": 0.7877492877492878, + "grad_norm": 0.5928917527198792, + "learning_rate": 0.00018159367651353496, + "loss": 1.0523, + "step": 4424 + }, + { + "epoch": 0.7879273504273504, + "grad_norm": 0.624422013759613, + "learning_rate": 0.0001815855831995029, + "loss": 1.0519, + "step": 4425 + }, + { + "epoch": 0.7881054131054132, + "grad_norm": 0.5140150785446167, + "learning_rate": 0.00018157748828697082, + "loss": 1.048, + "step": 4426 + }, + { + "epoch": 0.7882834757834758, + "grad_norm": 0.47006943821907043, + "learning_rate": 0.00018156939177609732, + "loss": 1.0067, + "step": 4427 + }, + { + "epoch": 0.7884615384615384, + "grad_norm": 0.5178864002227783, + "learning_rate": 0.00018156129366704105, + "loss": 1.0583, + "step": 4428 + }, + { + "epoch": 0.7886396011396012, + "grad_norm": 0.5279985666275024, + "learning_rate": 0.00018155319395996066, + "loss": 1.3023, + "step": 4429 + }, + { + "epoch": 0.7888176638176638, + "grad_norm": 0.5238787531852722, + "learning_rate": 0.00018154509265501482, + "loss": 1.0851, + "step": 4430 + }, + { + "epoch": 0.7889957264957265, + "grad_norm": 0.5914917588233948, + "learning_rate": 0.00018153698975236228, + "loss": 0.9291, + "step": 4431 + }, + { + "epoch": 0.7891737891737892, + "grad_norm": 0.5046082735061646, + "learning_rate": 0.00018152888525216183, + "loss": 0.9951, + "step": 4432 + }, + { + "epoch": 0.7893518518518519, + "grad_norm": 0.5042256116867065, + "learning_rate": 0.00018152077915457225, + "loss": 1.0243, + "step": 4433 + }, + { + "epoch": 0.7895299145299145, + "grad_norm": 0.5950339436531067, + "learning_rate": 0.0001815126714597523, + "loss": 0.9803, + "step": 4434 + }, + { + "epoch": 0.7897079772079773, + "grad_norm": 0.5163764953613281, + "learning_rate": 0.0001815045621678609, + "loss": 1.0353, + "step": 4435 + }, + { + "epoch": 0.7898860398860399, + "grad_norm": 0.5166211128234863, + "learning_rate": 0.00018149645127905691, + "loss": 0.9649, + "step": 4436 + }, + { + "epoch": 0.7900641025641025, + "grad_norm": 0.5239769220352173, + "learning_rate": 0.00018148833879349927, + "loss": 0.9747, + "step": 4437 + }, + { + "epoch": 0.7902421652421653, + "grad_norm": 0.5803237557411194, + "learning_rate": 0.00018148022471134692, + "loss": 1.315, + "step": 4438 + }, + { + "epoch": 0.7904202279202279, + "grad_norm": 0.5141370296478271, + "learning_rate": 0.00018147210903275877, + "loss": 1.0547, + "step": 4439 + }, + { + "epoch": 0.7905982905982906, + "grad_norm": 0.545788586139679, + "learning_rate": 0.00018146399175789394, + "loss": 1.0797, + "step": 4440 + }, + { + "epoch": 0.7907763532763533, + "grad_norm": 0.5273314714431763, + "learning_rate": 0.0001814558728869114, + "loss": 0.7928, + "step": 4441 + }, + { + "epoch": 0.790954415954416, + "grad_norm": 0.4614652693271637, + "learning_rate": 0.00018144775241997024, + "loss": 0.8826, + "step": 4442 + }, + { + "epoch": 0.7911324786324786, + "grad_norm": 0.6203590631484985, + "learning_rate": 0.00018143963035722958, + "loss": 1.2891, + "step": 4443 + }, + { + "epoch": 0.7913105413105413, + "grad_norm": 0.4870408773422241, + "learning_rate": 0.0001814315066988485, + "loss": 1.0717, + "step": 4444 + }, + { + "epoch": 0.791488603988604, + "grad_norm": 0.6468982696533203, + "learning_rate": 0.00018142338144498625, + "loss": 1.3398, + "step": 4445 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 0.4727918207645416, + "learning_rate": 0.00018141525459580197, + "loss": 1.0195, + "step": 4446 + }, + { + "epoch": 0.7918447293447294, + "grad_norm": 0.5080479979515076, + "learning_rate": 0.0001814071261514549, + "loss": 1.0163, + "step": 4447 + }, + { + "epoch": 0.792022792022792, + "grad_norm": 0.5380908250808716, + "learning_rate": 0.0001813989961121043, + "loss": 1.1673, + "step": 4448 + }, + { + "epoch": 0.7922008547008547, + "grad_norm": 0.5020384192466736, + "learning_rate": 0.00018139086447790945, + "loss": 0.8591, + "step": 4449 + }, + { + "epoch": 0.7923789173789174, + "grad_norm": 0.5279949903488159, + "learning_rate": 0.0001813827312490297, + "loss": 1.1221, + "step": 4450 + }, + { + "epoch": 0.79255698005698, + "grad_norm": 0.6739233732223511, + "learning_rate": 0.00018137459642562437, + "loss": 1.2704, + "step": 4451 + }, + { + "epoch": 0.7927350427350427, + "grad_norm": 0.5112259984016418, + "learning_rate": 0.00018136646000785288, + "loss": 1.1161, + "step": 4452 + }, + { + "epoch": 0.7929131054131054, + "grad_norm": 0.5244031548500061, + "learning_rate": 0.00018135832199587463, + "loss": 0.7866, + "step": 4453 + }, + { + "epoch": 0.7930911680911681, + "grad_norm": 0.5803347229957581, + "learning_rate": 0.0001813501823898491, + "loss": 0.994, + "step": 4454 + }, + { + "epoch": 0.7932692307692307, + "grad_norm": 0.6191152930259705, + "learning_rate": 0.00018134204118993568, + "loss": 1.0725, + "step": 4455 + }, + { + "epoch": 0.7934472934472935, + "grad_norm": 0.549735963344574, + "learning_rate": 0.00018133389839629396, + "loss": 0.9915, + "step": 4456 + }, + { + "epoch": 0.7936253561253561, + "grad_norm": 0.4940381646156311, + "learning_rate": 0.00018132575400908347, + "loss": 1.1815, + "step": 4457 + }, + { + "epoch": 0.7938034188034188, + "grad_norm": 0.5009099245071411, + "learning_rate": 0.00018131760802846377, + "loss": 1.0833, + "step": 4458 + }, + { + "epoch": 0.7939814814814815, + "grad_norm": 0.595853865146637, + "learning_rate": 0.00018130946045459445, + "loss": 1.2774, + "step": 4459 + }, + { + "epoch": 0.7941595441595442, + "grad_norm": 0.534794807434082, + "learning_rate": 0.00018130131128763513, + "loss": 1.0891, + "step": 4460 + }, + { + "epoch": 0.7943376068376068, + "grad_norm": 0.5828582048416138, + "learning_rate": 0.00018129316052774557, + "loss": 1.0786, + "step": 4461 + }, + { + "epoch": 0.7945156695156695, + "grad_norm": 0.4750654697418213, + "learning_rate": 0.00018128500817508533, + "loss": 1.0818, + "step": 4462 + }, + { + "epoch": 0.7946937321937322, + "grad_norm": 0.5626576542854309, + "learning_rate": 0.00018127685422981426, + "loss": 1.0807, + "step": 4463 + }, + { + "epoch": 0.7948717948717948, + "grad_norm": 0.6434760093688965, + "learning_rate": 0.00018126869869209203, + "loss": 1.0908, + "step": 4464 + }, + { + "epoch": 0.7950498575498576, + "grad_norm": 0.5577414631843567, + "learning_rate": 0.00018126054156207853, + "loss": 1.0281, + "step": 4465 + }, + { + "epoch": 0.7952279202279202, + "grad_norm": 0.5001249313354492, + "learning_rate": 0.00018125238283993347, + "loss": 0.9083, + "step": 4466 + }, + { + "epoch": 0.7954059829059829, + "grad_norm": 0.5298314690589905, + "learning_rate": 0.00018124422252581676, + "loss": 0.971, + "step": 4467 + }, + { + "epoch": 0.7955840455840456, + "grad_norm": 0.4872737228870392, + "learning_rate": 0.00018123606061988832, + "loss": 1.0515, + "step": 4468 + }, + { + "epoch": 0.7957621082621082, + "grad_norm": 0.5895398259162903, + "learning_rate": 0.00018122789712230798, + "loss": 1.0771, + "step": 4469 + }, + { + "epoch": 0.7959401709401709, + "grad_norm": 0.5212514996528625, + "learning_rate": 0.00018121973203323577, + "loss": 1.0365, + "step": 4470 + }, + { + "epoch": 0.7961182336182336, + "grad_norm": 0.4679451584815979, + "learning_rate": 0.0001812115653528316, + "loss": 0.9445, + "step": 4471 + }, + { + "epoch": 0.7962962962962963, + "grad_norm": 0.5852653980255127, + "learning_rate": 0.00018120339708125552, + "loss": 1.1781, + "step": 4472 + }, + { + "epoch": 0.7964743589743589, + "grad_norm": 0.6081342697143555, + "learning_rate": 0.00018119522721866756, + "loss": 1.3881, + "step": 4473 + }, + { + "epoch": 0.7966524216524217, + "grad_norm": 0.5254155993461609, + "learning_rate": 0.00018118705576522777, + "loss": 1.2198, + "step": 4474 + }, + { + "epoch": 0.7968304843304843, + "grad_norm": 0.5959419012069702, + "learning_rate": 0.00018117888272109632, + "loss": 1.0922, + "step": 4475 + }, + { + "epoch": 0.7970085470085471, + "grad_norm": 0.6243147253990173, + "learning_rate": 0.0001811707080864333, + "loss": 1.1782, + "step": 4476 + }, + { + "epoch": 0.7971866096866097, + "grad_norm": 0.5336906909942627, + "learning_rate": 0.0001811625318613988, + "loss": 1.167, + "step": 4477 + }, + { + "epoch": 0.7973646723646723, + "grad_norm": 0.5287907719612122, + "learning_rate": 0.00018115435404615315, + "loss": 0.9923, + "step": 4478 + }, + { + "epoch": 0.7975427350427351, + "grad_norm": 0.48941442370414734, + "learning_rate": 0.0001811461746408565, + "loss": 0.863, + "step": 4479 + }, + { + "epoch": 0.7977207977207977, + "grad_norm": 0.48465651273727417, + "learning_rate": 0.0001811379936456691, + "loss": 1.147, + "step": 4480 + }, + { + "epoch": 0.7978988603988604, + "grad_norm": 0.5676067471504211, + "learning_rate": 0.0001811298110607513, + "loss": 1.3121, + "step": 4481 + }, + { + "epoch": 0.7980769230769231, + "grad_norm": 0.4894018769264221, + "learning_rate": 0.00018112162688626337, + "loss": 1.1831, + "step": 4482 + }, + { + "epoch": 0.7982549857549858, + "grad_norm": 0.5626382827758789, + "learning_rate": 0.0001811134411223657, + "loss": 1.1977, + "step": 4483 + }, + { + "epoch": 0.7984330484330484, + "grad_norm": 0.564119815826416, + "learning_rate": 0.00018110525376921862, + "loss": 1.2686, + "step": 4484 + }, + { + "epoch": 0.7986111111111112, + "grad_norm": 0.6385740041732788, + "learning_rate": 0.00018109706482698256, + "loss": 1.2418, + "step": 4485 + }, + { + "epoch": 0.7987891737891738, + "grad_norm": 0.5550164580345154, + "learning_rate": 0.00018108887429581802, + "loss": 1.081, + "step": 4486 + }, + { + "epoch": 0.7989672364672364, + "grad_norm": 0.5583973526954651, + "learning_rate": 0.00018108068217588544, + "loss": 1.1757, + "step": 4487 + }, + { + "epoch": 0.7991452991452992, + "grad_norm": 0.5533342957496643, + "learning_rate": 0.00018107248846734527, + "loss": 1.1947, + "step": 4488 + }, + { + "epoch": 0.7993233618233618, + "grad_norm": 0.5291479229927063, + "learning_rate": 0.00018106429317035815, + "loss": 1.2769, + "step": 4489 + }, + { + "epoch": 0.7995014245014245, + "grad_norm": 0.4680160582065582, + "learning_rate": 0.00018105609628508458, + "loss": 0.7059, + "step": 4490 + }, + { + "epoch": 0.7996794871794872, + "grad_norm": 0.5364881157875061, + "learning_rate": 0.00018104789781168517, + "loss": 1.0566, + "step": 4491 + }, + { + "epoch": 0.7998575498575499, + "grad_norm": 0.5917307734489441, + "learning_rate": 0.0001810396977503206, + "loss": 1.2263, + "step": 4492 + }, + { + "epoch": 0.8000356125356125, + "grad_norm": 0.6013199090957642, + "learning_rate": 0.0001810314961011515, + "loss": 1.2053, + "step": 4493 + }, + { + "epoch": 0.8002136752136753, + "grad_norm": 0.6005663275718689, + "learning_rate": 0.0001810232928643385, + "loss": 1.2241, + "step": 4494 + }, + { + "epoch": 0.8003917378917379, + "grad_norm": 0.49207603931427, + "learning_rate": 0.00018101508804004246, + "loss": 1.0661, + "step": 4495 + }, + { + "epoch": 0.8005698005698005, + "grad_norm": 0.4834063947200775, + "learning_rate": 0.00018100688162842401, + "loss": 1.1745, + "step": 4496 + }, + { + "epoch": 0.8007478632478633, + "grad_norm": 0.5347156524658203, + "learning_rate": 0.000180998673629644, + "loss": 1.0679, + "step": 4497 + }, + { + "epoch": 0.8009259259259259, + "grad_norm": 0.5815600156784058, + "learning_rate": 0.00018099046404386327, + "loss": 1.2652, + "step": 4498 + }, + { + "epoch": 0.8011039886039886, + "grad_norm": 0.5291135311126709, + "learning_rate": 0.00018098225287124263, + "loss": 1.2072, + "step": 4499 + }, + { + "epoch": 0.8012820512820513, + "grad_norm": 0.5779497027397156, + "learning_rate": 0.000180974040111943, + "loss": 1.3277, + "step": 4500 + }, + { + "epoch": 0.801460113960114, + "grad_norm": 0.44566696882247925, + "learning_rate": 0.0001809658257661252, + "loss": 0.7702, + "step": 4501 + }, + { + "epoch": 0.8016381766381766, + "grad_norm": 0.5407577753067017, + "learning_rate": 0.00018095760983395027, + "loss": 1.2894, + "step": 4502 + }, + { + "epoch": 0.8018162393162394, + "grad_norm": 0.4771903455257416, + "learning_rate": 0.00018094939231557916, + "loss": 1.045, + "step": 4503 + }, + { + "epoch": 0.801994301994302, + "grad_norm": 0.5970945358276367, + "learning_rate": 0.00018094117321117286, + "loss": 1.2059, + "step": 4504 + }, + { + "epoch": 0.8021723646723646, + "grad_norm": 0.4959338903427124, + "learning_rate": 0.0001809329525208924, + "loss": 1.155, + "step": 4505 + }, + { + "epoch": 0.8023504273504274, + "grad_norm": 0.5142548084259033, + "learning_rate": 0.00018092473024489887, + "loss": 0.9413, + "step": 4506 + }, + { + "epoch": 0.80252849002849, + "grad_norm": 0.5336433053016663, + "learning_rate": 0.00018091650638335334, + "loss": 1.0699, + "step": 4507 + }, + { + "epoch": 0.8027065527065527, + "grad_norm": 0.47770628333091736, + "learning_rate": 0.00018090828093641698, + "loss": 1.1515, + "step": 4508 + }, + { + "epoch": 0.8028846153846154, + "grad_norm": 0.5443438291549683, + "learning_rate": 0.00018090005390425091, + "loss": 1.189, + "step": 4509 + }, + { + "epoch": 0.8030626780626781, + "grad_norm": 0.523179829120636, + "learning_rate": 0.00018089182528701632, + "loss": 1.1272, + "step": 4510 + }, + { + "epoch": 0.8032407407407407, + "grad_norm": 0.49628451466560364, + "learning_rate": 0.00018088359508487448, + "loss": 0.9754, + "step": 4511 + }, + { + "epoch": 0.8034188034188035, + "grad_norm": 0.5933086276054382, + "learning_rate": 0.00018087536329798663, + "loss": 1.2111, + "step": 4512 + }, + { + "epoch": 0.8035968660968661, + "grad_norm": 0.4565310776233673, + "learning_rate": 0.00018086712992651402, + "loss": 0.7729, + "step": 4513 + }, + { + "epoch": 0.8037749287749287, + "grad_norm": 0.5013461112976074, + "learning_rate": 0.00018085889497061798, + "loss": 1.2178, + "step": 4514 + }, + { + "epoch": 0.8039529914529915, + "grad_norm": 0.5170024633407593, + "learning_rate": 0.00018085065843045987, + "loss": 0.9181, + "step": 4515 + }, + { + "epoch": 0.8041310541310541, + "grad_norm": 0.583363950252533, + "learning_rate": 0.00018084242030620104, + "loss": 1.1542, + "step": 4516 + }, + { + "epoch": 0.8043091168091168, + "grad_norm": 0.46835777163505554, + "learning_rate": 0.00018083418059800297, + "loss": 0.8954, + "step": 4517 + }, + { + "epoch": 0.8044871794871795, + "grad_norm": 0.5145657062530518, + "learning_rate": 0.000180825939306027, + "loss": 1.0417, + "step": 4518 + }, + { + "epoch": 0.8046652421652422, + "grad_norm": 0.47216105461120605, + "learning_rate": 0.00018081769643043467, + "loss": 0.9516, + "step": 4519 + }, + { + "epoch": 0.8048433048433048, + "grad_norm": 0.5059915781021118, + "learning_rate": 0.0001808094519713875, + "loss": 1.1643, + "step": 4520 + }, + { + "epoch": 0.8050213675213675, + "grad_norm": 0.5406439900398254, + "learning_rate": 0.00018080120592904692, + "loss": 1.2038, + "step": 4521 + }, + { + "epoch": 0.8051994301994302, + "grad_norm": 0.6123420000076294, + "learning_rate": 0.0001807929583035746, + "loss": 1.4004, + "step": 4522 + }, + { + "epoch": 0.8053774928774928, + "grad_norm": 0.49699845910072327, + "learning_rate": 0.00018078470909513208, + "loss": 1.0347, + "step": 4523 + }, + { + "epoch": 0.8055555555555556, + "grad_norm": 0.5369421243667603, + "learning_rate": 0.000180776458303881, + "loss": 1.0418, + "step": 4524 + }, + { + "epoch": 0.8057336182336182, + "grad_norm": 0.5407396554946899, + "learning_rate": 0.00018076820592998301, + "loss": 0.9546, + "step": 4525 + }, + { + "epoch": 0.8059116809116809, + "grad_norm": 0.5749752521514893, + "learning_rate": 0.00018075995197359984, + "loss": 1.1438, + "step": 4526 + }, + { + "epoch": 0.8060897435897436, + "grad_norm": 0.5523102283477783, + "learning_rate": 0.00018075169643489317, + "loss": 1.1312, + "step": 4527 + }, + { + "epoch": 0.8062678062678063, + "grad_norm": 0.5767508149147034, + "learning_rate": 0.00018074343931402472, + "loss": 1.1951, + "step": 4528 + }, + { + "epoch": 0.8064458689458689, + "grad_norm": 0.5262924432754517, + "learning_rate": 0.00018073518061115633, + "loss": 1.1985, + "step": 4529 + }, + { + "epoch": 0.8066239316239316, + "grad_norm": 0.4742378294467926, + "learning_rate": 0.0001807269203264498, + "loss": 1.0126, + "step": 4530 + }, + { + "epoch": 0.8068019943019943, + "grad_norm": 0.5190158486366272, + "learning_rate": 0.00018071865846006692, + "loss": 0.9985, + "step": 4531 + }, + { + "epoch": 0.8069800569800569, + "grad_norm": 0.5910618305206299, + "learning_rate": 0.00018071039501216964, + "loss": 1.2776, + "step": 4532 + }, + { + "epoch": 0.8071581196581197, + "grad_norm": 0.5363098382949829, + "learning_rate": 0.00018070212998291983, + "loss": 1.3346, + "step": 4533 + }, + { + "epoch": 0.8073361823361823, + "grad_norm": 0.47711408138275146, + "learning_rate": 0.0001806938633724794, + "loss": 1.04, + "step": 4534 + }, + { + "epoch": 0.8075142450142451, + "grad_norm": 0.5092964172363281, + "learning_rate": 0.0001806855951810104, + "loss": 1.1409, + "step": 4535 + }, + { + "epoch": 0.8076923076923077, + "grad_norm": 0.5828777551651001, + "learning_rate": 0.00018067732540867472, + "loss": 1.3048, + "step": 4536 + }, + { + "epoch": 0.8078703703703703, + "grad_norm": 0.5779826045036316, + "learning_rate": 0.00018066905405563445, + "loss": 1.1599, + "step": 4537 + }, + { + "epoch": 0.8080484330484331, + "grad_norm": 0.49908435344696045, + "learning_rate": 0.00018066078112205167, + "loss": 1.1502, + "step": 4538 + }, + { + "epoch": 0.8082264957264957, + "grad_norm": 0.4772704839706421, + "learning_rate": 0.0001806525066080884, + "loss": 0.7925, + "step": 4539 + }, + { + "epoch": 0.8084045584045584, + "grad_norm": 0.4298383295536041, + "learning_rate": 0.00018064423051390683, + "loss": 0.7322, + "step": 4540 + }, + { + "epoch": 0.8085826210826211, + "grad_norm": 0.49349579215049744, + "learning_rate": 0.0001806359528396691, + "loss": 1.0021, + "step": 4541 + }, + { + "epoch": 0.8087606837606838, + "grad_norm": 0.4698609411716461, + "learning_rate": 0.00018062767358553735, + "loss": 0.9751, + "step": 4542 + }, + { + "epoch": 0.8089387464387464, + "grad_norm": 0.4949014186859131, + "learning_rate": 0.00018061939275167385, + "loss": 0.9553, + "step": 4543 + }, + { + "epoch": 0.8091168091168092, + "grad_norm": 0.5604463815689087, + "learning_rate": 0.0001806111103382408, + "loss": 0.9894, + "step": 4544 + }, + { + "epoch": 0.8092948717948718, + "grad_norm": 0.5761561989784241, + "learning_rate": 0.00018060282634540053, + "loss": 1.258, + "step": 4545 + }, + { + "epoch": 0.8094729344729344, + "grad_norm": 0.5239115357398987, + "learning_rate": 0.00018059454077331527, + "loss": 0.9189, + "step": 4546 + }, + { + "epoch": 0.8096509971509972, + "grad_norm": 0.47902220487594604, + "learning_rate": 0.00018058625362214742, + "loss": 1.0389, + "step": 4547 + }, + { + "epoch": 0.8098290598290598, + "grad_norm": 0.6274173259735107, + "learning_rate": 0.00018057796489205936, + "loss": 1.3368, + "step": 4548 + }, + { + "epoch": 0.8100071225071225, + "grad_norm": 0.5789401531219482, + "learning_rate": 0.00018056967458321345, + "loss": 1.1473, + "step": 4549 + }, + { + "epoch": 0.8101851851851852, + "grad_norm": 0.5850043296813965, + "learning_rate": 0.0001805613826957721, + "loss": 1.2224, + "step": 4550 + }, + { + "epoch": 0.8103632478632479, + "grad_norm": 0.6310738921165466, + "learning_rate": 0.00018055308922989788, + "loss": 1.0707, + "step": 4551 + }, + { + "epoch": 0.8105413105413105, + "grad_norm": 0.5198429822921753, + "learning_rate": 0.00018054479418575317, + "loss": 0.8984, + "step": 4552 + }, + { + "epoch": 0.8107193732193733, + "grad_norm": 0.5757743120193481, + "learning_rate": 0.00018053649756350054, + "loss": 1.2007, + "step": 4553 + }, + { + "epoch": 0.8108974358974359, + "grad_norm": 0.5109567642211914, + "learning_rate": 0.0001805281993633025, + "loss": 1.0696, + "step": 4554 + }, + { + "epoch": 0.8110754985754985, + "grad_norm": 0.5030225515365601, + "learning_rate": 0.00018051989958532173, + "loss": 0.9667, + "step": 4555 + }, + { + "epoch": 0.8112535612535613, + "grad_norm": 0.5291743874549866, + "learning_rate": 0.00018051159822972079, + "loss": 1.0219, + "step": 4556 + }, + { + "epoch": 0.8114316239316239, + "grad_norm": 0.5874896049499512, + "learning_rate": 0.00018050329529666233, + "loss": 0.8589, + "step": 4557 + }, + { + "epoch": 0.8116096866096866, + "grad_norm": 0.673284113407135, + "learning_rate": 0.000180494990786309, + "loss": 1.1902, + "step": 4558 + }, + { + "epoch": 0.8117877492877493, + "grad_norm": 0.4742524027824402, + "learning_rate": 0.00018048668469882354, + "loss": 1.0578, + "step": 4559 + }, + { + "epoch": 0.811965811965812, + "grad_norm": 0.5519167184829712, + "learning_rate": 0.0001804783770343687, + "loss": 1.083, + "step": 4560 + }, + { + "epoch": 0.8121438746438746, + "grad_norm": 0.5669941306114197, + "learning_rate": 0.00018047006779310727, + "loss": 1.0784, + "step": 4561 + }, + { + "epoch": 0.8123219373219374, + "grad_norm": 0.512759804725647, + "learning_rate": 0.000180461756975202, + "loss": 1.0361, + "step": 4562 + }, + { + "epoch": 0.8125, + "grad_norm": 0.5721749067306519, + "learning_rate": 0.00018045344458081575, + "loss": 1.0246, + "step": 4563 + }, + { + "epoch": 0.8126780626780626, + "grad_norm": 0.566430389881134, + "learning_rate": 0.00018044513061011137, + "loss": 1.1452, + "step": 4564 + }, + { + "epoch": 0.8128561253561254, + "grad_norm": 0.49391916394233704, + "learning_rate": 0.00018043681506325177, + "loss": 0.89, + "step": 4565 + }, + { + "epoch": 0.813034188034188, + "grad_norm": 0.5379437804222107, + "learning_rate": 0.00018042849794039988, + "loss": 1.1289, + "step": 4566 + }, + { + "epoch": 0.8132122507122507, + "grad_norm": 0.5667982697486877, + "learning_rate": 0.00018042017924171865, + "loss": 1.1596, + "step": 4567 + }, + { + "epoch": 0.8133903133903134, + "grad_norm": 0.6214209794998169, + "learning_rate": 0.00018041185896737109, + "loss": 1.0622, + "step": 4568 + }, + { + "epoch": 0.8135683760683761, + "grad_norm": 0.5442491173744202, + "learning_rate": 0.00018040353711752015, + "loss": 1.0536, + "step": 4569 + }, + { + "epoch": 0.8137464387464387, + "grad_norm": 0.5266172885894775, + "learning_rate": 0.00018039521369232894, + "loss": 1.0576, + "step": 4570 + }, + { + "epoch": 0.8139245014245015, + "grad_norm": 0.6057912111282349, + "learning_rate": 0.00018038688869196053, + "loss": 1.3067, + "step": 4571 + }, + { + "epoch": 0.8141025641025641, + "grad_norm": 0.489869087934494, + "learning_rate": 0.00018037856211657803, + "loss": 1.0279, + "step": 4572 + }, + { + "epoch": 0.8142806267806267, + "grad_norm": 0.5497978329658508, + "learning_rate": 0.00018037023396634457, + "loss": 1.1568, + "step": 4573 + }, + { + "epoch": 0.8144586894586895, + "grad_norm": 0.5243251919746399, + "learning_rate": 0.0001803619042414233, + "loss": 0.9767, + "step": 4574 + }, + { + "epoch": 0.8146367521367521, + "grad_norm": 0.503032922744751, + "learning_rate": 0.0001803535729419775, + "loss": 1.065, + "step": 4575 + }, + { + "epoch": 0.8148148148148148, + "grad_norm": 0.49955418705940247, + "learning_rate": 0.00018034524006817034, + "loss": 1.2752, + "step": 4576 + }, + { + "epoch": 0.8149928774928775, + "grad_norm": 0.5746406316757202, + "learning_rate": 0.00018033690562016508, + "loss": 1.098, + "step": 4577 + }, + { + "epoch": 0.8151709401709402, + "grad_norm": 0.5224192142486572, + "learning_rate": 0.00018032856959812507, + "loss": 1.1284, + "step": 4578 + }, + { + "epoch": 0.8153490028490028, + "grad_norm": 0.5484535694122314, + "learning_rate": 0.00018032023200221362, + "loss": 0.9182, + "step": 4579 + }, + { + "epoch": 0.8155270655270656, + "grad_norm": 0.5003355741500854, + "learning_rate": 0.00018031189283259405, + "loss": 1.136, + "step": 4580 + }, + { + "epoch": 0.8157051282051282, + "grad_norm": 0.5395768284797668, + "learning_rate": 0.00018030355208942977, + "loss": 1.2349, + "step": 4581 + }, + { + "epoch": 0.8158831908831908, + "grad_norm": 0.561966598033905, + "learning_rate": 0.0001802952097728842, + "loss": 0.999, + "step": 4582 + }, + { + "epoch": 0.8160612535612536, + "grad_norm": 0.4886479675769806, + "learning_rate": 0.00018028686588312083, + "loss": 0.9165, + "step": 4583 + }, + { + "epoch": 0.8162393162393162, + "grad_norm": 0.4769509732723236, + "learning_rate": 0.00018027852042030307, + "loss": 1.1377, + "step": 4584 + }, + { + "epoch": 0.8164173789173789, + "grad_norm": 0.4723633825778961, + "learning_rate": 0.00018027017338459448, + "loss": 1.0274, + "step": 4585 + }, + { + "epoch": 0.8165954415954416, + "grad_norm": 0.5773285627365112, + "learning_rate": 0.00018026182477615859, + "loss": 1.1468, + "step": 4586 + }, + { + "epoch": 0.8167735042735043, + "grad_norm": 0.5529203414916992, + "learning_rate": 0.00018025347459515895, + "loss": 1.0815, + "step": 4587 + }, + { + "epoch": 0.8169515669515669, + "grad_norm": 0.5449469685554504, + "learning_rate": 0.00018024512284175922, + "loss": 1.1637, + "step": 4588 + }, + { + "epoch": 0.8171296296296297, + "grad_norm": 0.5155341625213623, + "learning_rate": 0.00018023676951612298, + "loss": 1.1842, + "step": 4589 + }, + { + "epoch": 0.8173076923076923, + "grad_norm": 0.5569564700126648, + "learning_rate": 0.00018022841461841393, + "loss": 0.9254, + "step": 4590 + }, + { + "epoch": 0.8174857549857549, + "grad_norm": 0.45203131437301636, + "learning_rate": 0.00018022005814879573, + "loss": 0.9561, + "step": 4591 + }, + { + "epoch": 0.8176638176638177, + "grad_norm": 0.5735056400299072, + "learning_rate": 0.00018021170010743218, + "loss": 1.1402, + "step": 4592 + }, + { + "epoch": 0.8178418803418803, + "grad_norm": 0.6075260043144226, + "learning_rate": 0.00018020334049448697, + "loss": 0.8601, + "step": 4593 + }, + { + "epoch": 0.8180199430199431, + "grad_norm": 0.522682785987854, + "learning_rate": 0.0001801949793101239, + "loss": 1.0088, + "step": 4594 + }, + { + "epoch": 0.8181980056980057, + "grad_norm": 0.5648437142372131, + "learning_rate": 0.00018018661655450682, + "loss": 0.8359, + "step": 4595 + }, + { + "epoch": 0.8183760683760684, + "grad_norm": 0.5406472086906433, + "learning_rate": 0.00018017825222779954, + "loss": 1.1553, + "step": 4596 + }, + { + "epoch": 0.8185541310541311, + "grad_norm": 0.4917788803577423, + "learning_rate": 0.000180169886330166, + "loss": 1.2198, + "step": 4597 + }, + { + "epoch": 0.8187321937321937, + "grad_norm": 0.6293069124221802, + "learning_rate": 0.00018016151886177004, + "loss": 1.0245, + "step": 4598 + }, + { + "epoch": 0.8189102564102564, + "grad_norm": 0.47277843952178955, + "learning_rate": 0.00018015314982277564, + "loss": 1.1141, + "step": 4599 + }, + { + "epoch": 0.8190883190883191, + "grad_norm": 0.6132395267486572, + "learning_rate": 0.0001801447792133468, + "loss": 1.1227, + "step": 4600 + }, + { + "epoch": 0.8192663817663818, + "grad_norm": 0.46839597821235657, + "learning_rate": 0.00018013640703364747, + "loss": 0.9239, + "step": 4601 + }, + { + "epoch": 0.8194444444444444, + "grad_norm": 0.5055009722709656, + "learning_rate": 0.00018012803328384171, + "loss": 0.8486, + "step": 4602 + }, + { + "epoch": 0.8196225071225072, + "grad_norm": 0.5094841718673706, + "learning_rate": 0.00018011965796409362, + "loss": 0.9969, + "step": 4603 + }, + { + "epoch": 0.8198005698005698, + "grad_norm": 0.6177363395690918, + "learning_rate": 0.00018011128107456726, + "loss": 1.242, + "step": 4604 + }, + { + "epoch": 0.8199786324786325, + "grad_norm": 0.5280042290687561, + "learning_rate": 0.00018010290261542676, + "loss": 1.1569, + "step": 4605 + }, + { + "epoch": 0.8201566951566952, + "grad_norm": 0.5259367227554321, + "learning_rate": 0.00018009452258683625, + "loss": 0.9993, + "step": 4606 + }, + { + "epoch": 0.8203347578347578, + "grad_norm": 0.464469850063324, + "learning_rate": 0.00018008614098896, + "loss": 1.0288, + "step": 4607 + }, + { + "epoch": 0.8205128205128205, + "grad_norm": 0.6136324405670166, + "learning_rate": 0.00018007775782196214, + "loss": 1.1541, + "step": 4608 + }, + { + "epoch": 0.8206908831908832, + "grad_norm": 0.5376590490341187, + "learning_rate": 0.000180069373086007, + "loss": 1.0624, + "step": 4609 + }, + { + "epoch": 0.8208689458689459, + "grad_norm": 0.662916362285614, + "learning_rate": 0.0001800609867812588, + "loss": 1.1502, + "step": 4610 + }, + { + "epoch": 0.8210470085470085, + "grad_norm": 0.5153383612632751, + "learning_rate": 0.00018005259890788188, + "loss": 0.9789, + "step": 4611 + }, + { + "epoch": 0.8212250712250713, + "grad_norm": 0.5042359232902527, + "learning_rate": 0.00018004420946604057, + "loss": 0.9585, + "step": 4612 + }, + { + "epoch": 0.8214031339031339, + "grad_norm": 0.5395993590354919, + "learning_rate": 0.00018003581845589927, + "loss": 1.159, + "step": 4613 + }, + { + "epoch": 0.8215811965811965, + "grad_norm": 0.5561928749084473, + "learning_rate": 0.00018002742587762237, + "loss": 1.1604, + "step": 4614 + }, + { + "epoch": 0.8217592592592593, + "grad_norm": 0.5602710843086243, + "learning_rate": 0.00018001903173137432, + "loss": 0.9922, + "step": 4615 + }, + { + "epoch": 0.8219373219373219, + "grad_norm": 0.5529088377952576, + "learning_rate": 0.00018001063601731955, + "loss": 1.0943, + "step": 4616 + }, + { + "epoch": 0.8221153846153846, + "grad_norm": 0.5156456828117371, + "learning_rate": 0.00018000223873562254, + "loss": 1.1399, + "step": 4617 + }, + { + "epoch": 0.8222934472934473, + "grad_norm": 0.4868306517601013, + "learning_rate": 0.0001799938398864479, + "loss": 1.0692, + "step": 4618 + }, + { + "epoch": 0.82247150997151, + "grad_norm": 0.5372915267944336, + "learning_rate": 0.0001799854394699601, + "loss": 1.2675, + "step": 4619 + }, + { + "epoch": 0.8226495726495726, + "grad_norm": 0.6101839542388916, + "learning_rate": 0.0001799770374863238, + "loss": 0.9586, + "step": 4620 + }, + { + "epoch": 0.8228276353276354, + "grad_norm": 0.5034586787223816, + "learning_rate": 0.00017996863393570357, + "loss": 1.0885, + "step": 4621 + }, + { + "epoch": 0.823005698005698, + "grad_norm": 0.5608823299407959, + "learning_rate": 0.0001799602288182641, + "loss": 1.0002, + "step": 4622 + }, + { + "epoch": 0.8231837606837606, + "grad_norm": 0.5700048208236694, + "learning_rate": 0.00017995182213417, + "loss": 1.1484, + "step": 4623 + }, + { + "epoch": 0.8233618233618234, + "grad_norm": 0.5283229351043701, + "learning_rate": 0.00017994341388358608, + "loss": 1.0744, + "step": 4624 + }, + { + "epoch": 0.823539886039886, + "grad_norm": 0.5215758681297302, + "learning_rate": 0.00017993500406667703, + "loss": 1.2686, + "step": 4625 + }, + { + "epoch": 0.8237179487179487, + "grad_norm": 0.528883159160614, + "learning_rate": 0.0001799265926836076, + "loss": 1.1393, + "step": 4626 + }, + { + "epoch": 0.8238960113960114, + "grad_norm": 0.5589834451675415, + "learning_rate": 0.00017991817973454265, + "loss": 1.1744, + "step": 4627 + }, + { + "epoch": 0.8240740740740741, + "grad_norm": 0.49817174673080444, + "learning_rate": 0.00017990976521964697, + "loss": 1.0544, + "step": 4628 + }, + { + "epoch": 0.8242521367521367, + "grad_norm": 0.613961398601532, + "learning_rate": 0.00017990134913908542, + "loss": 1.0951, + "step": 4629 + }, + { + "epoch": 0.8244301994301995, + "grad_norm": 0.47278255224227905, + "learning_rate": 0.00017989293149302295, + "loss": 0.9742, + "step": 4630 + }, + { + "epoch": 0.8246082621082621, + "grad_norm": 0.49807092547416687, + "learning_rate": 0.00017988451228162443, + "loss": 1.0985, + "step": 4631 + }, + { + "epoch": 0.8247863247863247, + "grad_norm": 0.5624374747276306, + "learning_rate": 0.00017987609150505485, + "loss": 1.2446, + "step": 4632 + }, + { + "epoch": 0.8249643874643875, + "grad_norm": 0.4863535761833191, + "learning_rate": 0.00017986766916347916, + "loss": 1.0239, + "step": 4633 + }, + { + "epoch": 0.8251424501424501, + "grad_norm": 0.679585874080658, + "learning_rate": 0.00017985924525706245, + "loss": 1.1698, + "step": 4634 + }, + { + "epoch": 0.8253205128205128, + "grad_norm": 0.5545455813407898, + "learning_rate": 0.00017985081978596967, + "loss": 1.0926, + "step": 4635 + }, + { + "epoch": 0.8254985754985755, + "grad_norm": 0.5303109288215637, + "learning_rate": 0.000179842392750366, + "loss": 1.0978, + "step": 4636 + }, + { + "epoch": 0.8256766381766382, + "grad_norm": 0.6053299307823181, + "learning_rate": 0.00017983396415041644, + "loss": 1.0596, + "step": 4637 + }, + { + "epoch": 0.8258547008547008, + "grad_norm": 0.5241885185241699, + "learning_rate": 0.00017982553398628625, + "loss": 0.8541, + "step": 4638 + }, + { + "epoch": 0.8260327635327636, + "grad_norm": 0.5934443473815918, + "learning_rate": 0.00017981710225814052, + "loss": 1.145, + "step": 4639 + }, + { + "epoch": 0.8262108262108262, + "grad_norm": 0.5341619849205017, + "learning_rate": 0.00017980866896614447, + "loss": 1.0745, + "step": 4640 + }, + { + "epoch": 0.8263888888888888, + "grad_norm": 0.6732913851737976, + "learning_rate": 0.00017980023411046336, + "loss": 1.0775, + "step": 4641 + }, + { + "epoch": 0.8265669515669516, + "grad_norm": 0.5134359002113342, + "learning_rate": 0.0001797917976912624, + "loss": 1.0298, + "step": 4642 + }, + { + "epoch": 0.8267450142450142, + "grad_norm": 0.5234783887863159, + "learning_rate": 0.00017978335970870698, + "loss": 1.1069, + "step": 4643 + }, + { + "epoch": 0.8269230769230769, + "grad_norm": 0.4776439964771271, + "learning_rate": 0.00017977492016296232, + "loss": 0.6367, + "step": 4644 + }, + { + "epoch": 0.8271011396011396, + "grad_norm": 0.53763347864151, + "learning_rate": 0.0001797664790541938, + "loss": 1.1356, + "step": 4645 + }, + { + "epoch": 0.8272792022792023, + "grad_norm": 0.5082212686538696, + "learning_rate": 0.00017975803638256682, + "loss": 0.7873, + "step": 4646 + }, + { + "epoch": 0.8274572649572649, + "grad_norm": 0.5156424641609192, + "learning_rate": 0.00017974959214824685, + "loss": 1.084, + "step": 4647 + }, + { + "epoch": 0.8276353276353277, + "grad_norm": 0.5275198817253113, + "learning_rate": 0.00017974114635139926, + "loss": 1.1219, + "step": 4648 + }, + { + "epoch": 0.8278133903133903, + "grad_norm": 0.5548223257064819, + "learning_rate": 0.00017973269899218956, + "loss": 1.0808, + "step": 4649 + }, + { + "epoch": 0.8279914529914529, + "grad_norm": 0.535347580909729, + "learning_rate": 0.00017972425007078323, + "loss": 1.1211, + "step": 4650 + }, + { + "epoch": 0.8281695156695157, + "grad_norm": 0.5299580693244934, + "learning_rate": 0.00017971579958734587, + "loss": 0.9911, + "step": 4651 + }, + { + "epoch": 0.8283475783475783, + "grad_norm": 0.4863550066947937, + "learning_rate": 0.000179707347542043, + "loss": 0.9122, + "step": 4652 + }, + { + "epoch": 0.8285256410256411, + "grad_norm": 0.5284972190856934, + "learning_rate": 0.00017969889393504022, + "loss": 1.0424, + "step": 4653 + }, + { + "epoch": 0.8287037037037037, + "grad_norm": 0.5305661559104919, + "learning_rate": 0.00017969043876650317, + "loss": 1.1122, + "step": 4654 + }, + { + "epoch": 0.8288817663817664, + "grad_norm": 0.5645657777786255, + "learning_rate": 0.00017968198203659755, + "loss": 1.2195, + "step": 4655 + }, + { + "epoch": 0.8290598290598291, + "grad_norm": 0.521649181842804, + "learning_rate": 0.000179673523745489, + "loss": 1.2684, + "step": 4656 + }, + { + "epoch": 0.8292378917378918, + "grad_norm": 0.5984422564506531, + "learning_rate": 0.00017966506389334322, + "loss": 0.9894, + "step": 4657 + }, + { + "epoch": 0.8294159544159544, + "grad_norm": 0.5318729281425476, + "learning_rate": 0.00017965660248032603, + "loss": 1.2929, + "step": 4658 + }, + { + "epoch": 0.8295940170940171, + "grad_norm": 0.4666081368923187, + "learning_rate": 0.0001796481395066032, + "loss": 0.9646, + "step": 4659 + }, + { + "epoch": 0.8297720797720798, + "grad_norm": 0.5780388116836548, + "learning_rate": 0.00017963967497234054, + "loss": 1.1043, + "step": 4660 + }, + { + "epoch": 0.8299501424501424, + "grad_norm": 0.44089245796203613, + "learning_rate": 0.00017963120887770387, + "loss": 0.8932, + "step": 4661 + }, + { + "epoch": 0.8301282051282052, + "grad_norm": 0.5198349356651306, + "learning_rate": 0.0001796227412228591, + "loss": 0.9378, + "step": 4662 + }, + { + "epoch": 0.8303062678062678, + "grad_norm": 0.5298343896865845, + "learning_rate": 0.00017961427200797206, + "loss": 1.0272, + "step": 4663 + }, + { + "epoch": 0.8304843304843305, + "grad_norm": 0.5087099671363831, + "learning_rate": 0.0001796058012332088, + "loss": 0.989, + "step": 4664 + }, + { + "epoch": 0.8306623931623932, + "grad_norm": 0.504228949546814, + "learning_rate": 0.0001795973288987352, + "loss": 1.0134, + "step": 4665 + }, + { + "epoch": 0.8308404558404558, + "grad_norm": 0.6788033843040466, + "learning_rate": 0.00017958885500471728, + "loss": 0.8856, + "step": 4666 + }, + { + "epoch": 0.8310185185185185, + "grad_norm": 0.5166172385215759, + "learning_rate": 0.00017958037955132113, + "loss": 0.8711, + "step": 4667 + }, + { + "epoch": 0.8311965811965812, + "grad_norm": 0.5712400078773499, + "learning_rate": 0.00017957190253871272, + "loss": 1.0418, + "step": 4668 + }, + { + "epoch": 0.8313746438746439, + "grad_norm": 0.5531231164932251, + "learning_rate": 0.0001795634239670582, + "loss": 0.9021, + "step": 4669 + }, + { + "epoch": 0.8315527065527065, + "grad_norm": 0.6165615916252136, + "learning_rate": 0.00017955494383652365, + "loss": 1.0927, + "step": 4670 + }, + { + "epoch": 0.8317307692307693, + "grad_norm": 0.5920368432998657, + "learning_rate": 0.00017954646214727525, + "loss": 1.231, + "step": 4671 + }, + { + "epoch": 0.8319088319088319, + "grad_norm": 0.5037244558334351, + "learning_rate": 0.00017953797889947915, + "loss": 0.85, + "step": 4672 + }, + { + "epoch": 0.8320868945868946, + "grad_norm": 0.5618211627006531, + "learning_rate": 0.0001795294940933016, + "loss": 1.145, + "step": 4673 + }, + { + "epoch": 0.8322649572649573, + "grad_norm": 0.6275593042373657, + "learning_rate": 0.00017952100772890877, + "loss": 0.9061, + "step": 4674 + }, + { + "epoch": 0.83244301994302, + "grad_norm": 0.5376096367835999, + "learning_rate": 0.00017951251980646702, + "loss": 1.1948, + "step": 4675 + }, + { + "epoch": 0.8326210826210826, + "grad_norm": 0.5162268877029419, + "learning_rate": 0.0001795040303261426, + "loss": 1.2158, + "step": 4676 + }, + { + "epoch": 0.8327991452991453, + "grad_norm": 0.5730512142181396, + "learning_rate": 0.0001794955392881019, + "loss": 0.9962, + "step": 4677 + }, + { + "epoch": 0.832977207977208, + "grad_norm": 0.5128712058067322, + "learning_rate": 0.00017948704669251122, + "loss": 1.2797, + "step": 4678 + }, + { + "epoch": 0.8331552706552706, + "grad_norm": 0.5173979997634888, + "learning_rate": 0.00017947855253953697, + "loss": 1.1093, + "step": 4679 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 0.504646897315979, + "learning_rate": 0.0001794700568293456, + "loss": 1.3171, + "step": 4680 + }, + { + "epoch": 0.833511396011396, + "grad_norm": 0.5638105869293213, + "learning_rate": 0.00017946155956210356, + "loss": 0.9224, + "step": 4681 + }, + { + "epoch": 0.8336894586894587, + "grad_norm": 0.5289680361747742, + "learning_rate": 0.00017945306073797733, + "loss": 0.8919, + "step": 4682 + }, + { + "epoch": 0.8338675213675214, + "grad_norm": 0.5224629044532776, + "learning_rate": 0.0001794445603571334, + "loss": 1.0345, + "step": 4683 + }, + { + "epoch": 0.834045584045584, + "grad_norm": 0.5342282056808472, + "learning_rate": 0.00017943605841973836, + "loss": 1.2305, + "step": 4684 + }, + { + "epoch": 0.8342236467236467, + "grad_norm": 0.6118032336235046, + "learning_rate": 0.00017942755492595874, + "loss": 1.0316, + "step": 4685 + }, + { + "epoch": 0.8344017094017094, + "grad_norm": 0.49112311005592346, + "learning_rate": 0.00017941904987596121, + "loss": 0.9809, + "step": 4686 + }, + { + "epoch": 0.8345797720797721, + "grad_norm": 0.5044063925743103, + "learning_rate": 0.0001794105432699124, + "loss": 0.834, + "step": 4687 + }, + { + "epoch": 0.8347578347578347, + "grad_norm": 0.4849987328052521, + "learning_rate": 0.00017940203510797892, + "loss": 0.9971, + "step": 4688 + }, + { + "epoch": 0.8349358974358975, + "grad_norm": 0.5539469122886658, + "learning_rate": 0.00017939352539032748, + "loss": 1.1599, + "step": 4689 + }, + { + "epoch": 0.8351139601139601, + "grad_norm": 0.5474258065223694, + "learning_rate": 0.00017938501411712485, + "loss": 1.25, + "step": 4690 + }, + { + "epoch": 0.8352920227920227, + "grad_norm": 0.4880213737487793, + "learning_rate": 0.0001793765012885378, + "loss": 1.1471, + "step": 4691 + }, + { + "epoch": 0.8354700854700855, + "grad_norm": 0.5602759718894958, + "learning_rate": 0.00017936798690473309, + "loss": 1.0723, + "step": 4692 + }, + { + "epoch": 0.8356481481481481, + "grad_norm": 0.627775251865387, + "learning_rate": 0.00017935947096587755, + "loss": 1.3768, + "step": 4693 + }, + { + "epoch": 0.8358262108262108, + "grad_norm": 0.5324847102165222, + "learning_rate": 0.00017935095347213804, + "loss": 0.9945, + "step": 4694 + }, + { + "epoch": 0.8360042735042735, + "grad_norm": 0.5244048237800598, + "learning_rate": 0.0001793424344236814, + "loss": 1.1725, + "step": 4695 + }, + { + "epoch": 0.8361823361823362, + "grad_norm": 0.5420708656311035, + "learning_rate": 0.00017933391382067462, + "loss": 1.1267, + "step": 4696 + }, + { + "epoch": 0.8363603988603988, + "grad_norm": 0.5285456776618958, + "learning_rate": 0.00017932539166328458, + "loss": 1.0368, + "step": 4697 + }, + { + "epoch": 0.8365384615384616, + "grad_norm": 0.5330373048782349, + "learning_rate": 0.00017931686795167825, + "loss": 1.1082, + "step": 4698 + }, + { + "epoch": 0.8367165242165242, + "grad_norm": 0.5516682267189026, + "learning_rate": 0.0001793083426860227, + "loss": 1.1833, + "step": 4699 + }, + { + "epoch": 0.8368945868945868, + "grad_norm": 0.5229935646057129, + "learning_rate": 0.0001792998158664849, + "loss": 0.8527, + "step": 4700 + }, + { + "epoch": 0.8370726495726496, + "grad_norm": 0.4821490943431854, + "learning_rate": 0.00017929128749323195, + "loss": 1.1201, + "step": 4701 + }, + { + "epoch": 0.8372507122507122, + "grad_norm": 0.6276404857635498, + "learning_rate": 0.0001792827575664309, + "loss": 1.0986, + "step": 4702 + }, + { + "epoch": 0.8374287749287749, + "grad_norm": 0.5681334733963013, + "learning_rate": 0.00017927422608624897, + "loss": 1.3821, + "step": 4703 + }, + { + "epoch": 0.8376068376068376, + "grad_norm": 0.5257087349891663, + "learning_rate": 0.00017926569305285324, + "loss": 1.1033, + "step": 4704 + }, + { + "epoch": 0.8377849002849003, + "grad_norm": 0.5665168166160583, + "learning_rate": 0.0001792571584664109, + "loss": 1.104, + "step": 4705 + }, + { + "epoch": 0.8379629629629629, + "grad_norm": 0.5202076435089111, + "learning_rate": 0.00017924862232708918, + "loss": 1.052, + "step": 4706 + }, + { + "epoch": 0.8381410256410257, + "grad_norm": 0.5103010535240173, + "learning_rate": 0.00017924008463505534, + "loss": 1.1348, + "step": 4707 + }, + { + "epoch": 0.8383190883190883, + "grad_norm": 0.6811865568161011, + "learning_rate": 0.00017923154539047667, + "loss": 1.2804, + "step": 4708 + }, + { + "epoch": 0.8384971509971509, + "grad_norm": 0.46808311343193054, + "learning_rate": 0.00017922300459352042, + "loss": 0.9302, + "step": 4709 + }, + { + "epoch": 0.8386752136752137, + "grad_norm": 0.47713059186935425, + "learning_rate": 0.00017921446224435398, + "loss": 0.78, + "step": 4710 + }, + { + "epoch": 0.8388532763532763, + "grad_norm": 0.7579890489578247, + "learning_rate": 0.0001792059183431447, + "loss": 1.4776, + "step": 4711 + }, + { + "epoch": 0.8390313390313391, + "grad_norm": 0.6009423136711121, + "learning_rate": 0.00017919737289006, + "loss": 1.2679, + "step": 4712 + }, + { + "epoch": 0.8392094017094017, + "grad_norm": 0.56390780210495, + "learning_rate": 0.00017918882588526729, + "loss": 1.0402, + "step": 4713 + }, + { + "epoch": 0.8393874643874644, + "grad_norm": 0.5698862075805664, + "learning_rate": 0.00017918027732893404, + "loss": 1.2336, + "step": 4714 + }, + { + "epoch": 0.8395655270655271, + "grad_norm": 0.5016305446624756, + "learning_rate": 0.0001791717272212277, + "loss": 1.0373, + "step": 4715 + }, + { + "epoch": 0.8397435897435898, + "grad_norm": 0.5886971950531006, + "learning_rate": 0.0001791631755623159, + "loss": 1.1062, + "step": 4716 + }, + { + "epoch": 0.8399216524216524, + "grad_norm": 0.647833526134491, + "learning_rate": 0.00017915462235236607, + "loss": 1.0464, + "step": 4717 + }, + { + "epoch": 0.8400997150997151, + "grad_norm": 0.4961194396018982, + "learning_rate": 0.00017914606759154587, + "loss": 1.0763, + "step": 4718 + }, + { + "epoch": 0.8402777777777778, + "grad_norm": 0.47041359543800354, + "learning_rate": 0.00017913751128002288, + "loss": 1.0685, + "step": 4719 + }, + { + "epoch": 0.8404558404558404, + "grad_norm": 0.5752858519554138, + "learning_rate": 0.00017912895341796475, + "loss": 1.0577, + "step": 4720 + }, + { + "epoch": 0.8406339031339032, + "grad_norm": 0.5233224034309387, + "learning_rate": 0.00017912039400553914, + "loss": 1.1484, + "step": 4721 + }, + { + "epoch": 0.8408119658119658, + "grad_norm": 0.5327485203742981, + "learning_rate": 0.00017911183304291378, + "loss": 1.0028, + "step": 4722 + }, + { + "epoch": 0.8409900284900285, + "grad_norm": 0.5320752263069153, + "learning_rate": 0.00017910327053025638, + "loss": 1.1247, + "step": 4723 + }, + { + "epoch": 0.8411680911680912, + "grad_norm": 0.529617965221405, + "learning_rate": 0.00017909470646773477, + "loss": 1.1698, + "step": 4724 + }, + { + "epoch": 0.8413461538461539, + "grad_norm": 0.5055609345436096, + "learning_rate": 0.00017908614085551664, + "loss": 1.0925, + "step": 4725 + }, + { + "epoch": 0.8415242165242165, + "grad_norm": 0.5356255769729614, + "learning_rate": 0.00017907757369376985, + "loss": 1.0354, + "step": 4726 + }, + { + "epoch": 0.8417022792022792, + "grad_norm": 0.582834780216217, + "learning_rate": 0.00017906900498266233, + "loss": 1.1248, + "step": 4727 + }, + { + "epoch": 0.8418803418803419, + "grad_norm": 0.5750834941864014, + "learning_rate": 0.00017906043472236188, + "loss": 1.0119, + "step": 4728 + }, + { + "epoch": 0.8420584045584045, + "grad_norm": 0.5923320055007935, + "learning_rate": 0.00017905186291303644, + "loss": 1.0662, + "step": 4729 + }, + { + "epoch": 0.8422364672364673, + "grad_norm": 0.4767811894416809, + "learning_rate": 0.00017904328955485396, + "loss": 1.0911, + "step": 4730 + }, + { + "epoch": 0.8424145299145299, + "grad_norm": 0.5294556021690369, + "learning_rate": 0.00017903471464798245, + "loss": 1.2861, + "step": 4731 + }, + { + "epoch": 0.8425925925925926, + "grad_norm": 0.599117636680603, + "learning_rate": 0.00017902613819258985, + "loss": 1.1707, + "step": 4732 + }, + { + "epoch": 0.8427706552706553, + "grad_norm": 0.5912977457046509, + "learning_rate": 0.00017901756018884424, + "loss": 1.1884, + "step": 4733 + }, + { + "epoch": 0.842948717948718, + "grad_norm": 0.587676465511322, + "learning_rate": 0.0001790089806369137, + "loss": 1.1054, + "step": 4734 + }, + { + "epoch": 0.8431267806267806, + "grad_norm": 0.6271800398826599, + "learning_rate": 0.0001790003995369663, + "loss": 1.2094, + "step": 4735 + }, + { + "epoch": 0.8433048433048433, + "grad_norm": 0.47198590636253357, + "learning_rate": 0.00017899181688917017, + "loss": 0.9561, + "step": 4736 + }, + { + "epoch": 0.843482905982906, + "grad_norm": 0.690732479095459, + "learning_rate": 0.00017898323269369351, + "loss": 1.1629, + "step": 4737 + }, + { + "epoch": 0.8436609686609686, + "grad_norm": 0.4926888048648834, + "learning_rate": 0.00017897464695070445, + "loss": 1.1097, + "step": 4738 + }, + { + "epoch": 0.8438390313390314, + "grad_norm": 0.7071278691291809, + "learning_rate": 0.00017896605966037128, + "loss": 1.195, + "step": 4739 + }, + { + "epoch": 0.844017094017094, + "grad_norm": 0.5650486350059509, + "learning_rate": 0.00017895747082286216, + "loss": 1.0107, + "step": 4740 + }, + { + "epoch": 0.8441951566951567, + "grad_norm": 0.5291931629180908, + "learning_rate": 0.00017894888043834545, + "loss": 1.0104, + "step": 4741 + }, + { + "epoch": 0.8443732193732194, + "grad_norm": 0.5751241445541382, + "learning_rate": 0.00017894028850698942, + "loss": 1.2482, + "step": 4742 + }, + { + "epoch": 0.844551282051282, + "grad_norm": 0.5833632349967957, + "learning_rate": 0.0001789316950289624, + "loss": 1.0552, + "step": 4743 + }, + { + "epoch": 0.8447293447293447, + "grad_norm": 0.543729841709137, + "learning_rate": 0.00017892310000443282, + "loss": 1.1453, + "step": 4744 + }, + { + "epoch": 0.8449074074074074, + "grad_norm": 0.5674204230308533, + "learning_rate": 0.00017891450343356902, + "loss": 1.0757, + "step": 4745 + }, + { + "epoch": 0.8450854700854701, + "grad_norm": 0.5161892771720886, + "learning_rate": 0.00017890590531653946, + "loss": 1.1163, + "step": 4746 + }, + { + "epoch": 0.8452635327635327, + "grad_norm": 0.49907612800598145, + "learning_rate": 0.00017889730565351258, + "loss": 1.0356, + "step": 4747 + }, + { + "epoch": 0.8454415954415955, + "grad_norm": 0.4994732439517975, + "learning_rate": 0.00017888870444465692, + "loss": 1.026, + "step": 4748 + }, + { + "epoch": 0.8456196581196581, + "grad_norm": 0.6397520303726196, + "learning_rate": 0.00017888010169014095, + "loss": 0.957, + "step": 4749 + }, + { + "epoch": 0.8457977207977208, + "grad_norm": 0.5379729270935059, + "learning_rate": 0.00017887149739013327, + "loss": 1.1664, + "step": 4750 + }, + { + "epoch": 0.8459757834757835, + "grad_norm": 0.4487382769584656, + "learning_rate": 0.00017886289154480246, + "loss": 0.9377, + "step": 4751 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.5645943880081177, + "learning_rate": 0.00017885428415431707, + "loss": 1.273, + "step": 4752 + }, + { + "epoch": 0.8463319088319088, + "grad_norm": 0.5535289645195007, + "learning_rate": 0.00017884567521884577, + "loss": 1.1779, + "step": 4753 + }, + { + "epoch": 0.8465099715099715, + "grad_norm": 0.5039721131324768, + "learning_rate": 0.0001788370647385573, + "loss": 1.0237, + "step": 4754 + }, + { + "epoch": 0.8466880341880342, + "grad_norm": 0.4543854892253876, + "learning_rate": 0.00017882845271362032, + "loss": 0.8149, + "step": 4755 + }, + { + "epoch": 0.8468660968660968, + "grad_norm": 0.5095639824867249, + "learning_rate": 0.00017881983914420352, + "loss": 1.0141, + "step": 4756 + }, + { + "epoch": 0.8470441595441596, + "grad_norm": 0.5341798663139343, + "learning_rate": 0.00017881122403047575, + "loss": 1.1885, + "step": 4757 + }, + { + "epoch": 0.8472222222222222, + "grad_norm": 0.5595062971115112, + "learning_rate": 0.00017880260737260573, + "loss": 0.8939, + "step": 4758 + }, + { + "epoch": 0.8474002849002849, + "grad_norm": 0.5355880260467529, + "learning_rate": 0.00017879398917076232, + "loss": 1.2434, + "step": 4759 + }, + { + "epoch": 0.8475783475783476, + "grad_norm": 0.49477261304855347, + "learning_rate": 0.0001787853694251144, + "loss": 0.979, + "step": 4760 + }, + { + "epoch": 0.8477564102564102, + "grad_norm": 0.5154359340667725, + "learning_rate": 0.00017877674813583078, + "loss": 1.0957, + "step": 4761 + }, + { + "epoch": 0.8479344729344729, + "grad_norm": 0.5651070475578308, + "learning_rate": 0.00017876812530308046, + "loss": 1.1884, + "step": 4762 + }, + { + "epoch": 0.8481125356125356, + "grad_norm": 0.537277340888977, + "learning_rate": 0.00017875950092703232, + "loss": 1.0272, + "step": 4763 + }, + { + "epoch": 0.8482905982905983, + "grad_norm": 0.5259691476821899, + "learning_rate": 0.00017875087500785538, + "loss": 1.1493, + "step": 4764 + }, + { + "epoch": 0.8484686609686609, + "grad_norm": 0.5491300225257874, + "learning_rate": 0.00017874224754571867, + "loss": 0.8316, + "step": 4765 + }, + { + "epoch": 0.8486467236467237, + "grad_norm": 0.5493744611740112, + "learning_rate": 0.00017873361854079116, + "loss": 1.2328, + "step": 4766 + }, + { + "epoch": 0.8488247863247863, + "grad_norm": 0.571002185344696, + "learning_rate": 0.00017872498799324197, + "loss": 1.1384, + "step": 4767 + }, + { + "epoch": 0.8490028490028491, + "grad_norm": 0.538152813911438, + "learning_rate": 0.00017871635590324013, + "loss": 1.0581, + "step": 4768 + }, + { + "epoch": 0.8491809116809117, + "grad_norm": 0.5214923620223999, + "learning_rate": 0.00017870772227095486, + "loss": 1.0612, + "step": 4769 + }, + { + "epoch": 0.8493589743589743, + "grad_norm": 0.5714883804321289, + "learning_rate": 0.0001786990870965553, + "loss": 0.9076, + "step": 4770 + }, + { + "epoch": 0.8495370370370371, + "grad_norm": 0.4181775450706482, + "learning_rate": 0.00017869045038021054, + "loss": 0.8366, + "step": 4771 + }, + { + "epoch": 0.8497150997150997, + "grad_norm": 0.6266027688980103, + "learning_rate": 0.00017868181212208993, + "loss": 1.2047, + "step": 4772 + }, + { + "epoch": 0.8498931623931624, + "grad_norm": 0.5423732399940491, + "learning_rate": 0.0001786731723223626, + "loss": 1.3878, + "step": 4773 + }, + { + "epoch": 0.8500712250712251, + "grad_norm": 0.5512300133705139, + "learning_rate": 0.00017866453098119793, + "loss": 1.1132, + "step": 4774 + }, + { + "epoch": 0.8502492877492878, + "grad_norm": 0.5767185688018799, + "learning_rate": 0.00017865588809876519, + "loss": 0.97, + "step": 4775 + }, + { + "epoch": 0.8504273504273504, + "grad_norm": 0.5305790305137634, + "learning_rate": 0.00017864724367523368, + "loss": 1.1158, + "step": 4776 + }, + { + "epoch": 0.8506054131054132, + "grad_norm": 0.49702391028404236, + "learning_rate": 0.00017863859771077284, + "loss": 0.9669, + "step": 4777 + }, + { + "epoch": 0.8507834757834758, + "grad_norm": 0.5490063428878784, + "learning_rate": 0.00017862995020555205, + "loss": 1.0646, + "step": 4778 + }, + { + "epoch": 0.8509615384615384, + "grad_norm": 0.5308689475059509, + "learning_rate": 0.00017862130115974068, + "loss": 0.8922, + "step": 4779 + }, + { + "epoch": 0.8511396011396012, + "grad_norm": 0.5412983894348145, + "learning_rate": 0.00017861265057350826, + "loss": 1.1444, + "step": 4780 + }, + { + "epoch": 0.8513176638176638, + "grad_norm": 0.5857377052307129, + "learning_rate": 0.00017860399844702425, + "loss": 1.1643, + "step": 4781 + }, + { + "epoch": 0.8514957264957265, + "grad_norm": 0.599273681640625, + "learning_rate": 0.00017859534478045815, + "loss": 1.169, + "step": 4782 + }, + { + "epoch": 0.8516737891737892, + "grad_norm": 0.5677087903022766, + "learning_rate": 0.00017858668957397957, + "loss": 1.0793, + "step": 4783 + }, + { + "epoch": 0.8518518518518519, + "grad_norm": 0.5648362636566162, + "learning_rate": 0.00017857803282775807, + "loss": 1.1932, + "step": 4784 + }, + { + "epoch": 0.8520299145299145, + "grad_norm": 0.5138826966285706, + "learning_rate": 0.00017856937454196323, + "loss": 1.0011, + "step": 4785 + }, + { + "epoch": 0.8522079772079773, + "grad_norm": 0.5951429009437561, + "learning_rate": 0.0001785607147167647, + "loss": 1.3198, + "step": 4786 + }, + { + "epoch": 0.8523860398860399, + "grad_norm": 0.5341953039169312, + "learning_rate": 0.00017855205335233216, + "loss": 0.9094, + "step": 4787 + }, + { + "epoch": 0.8525641025641025, + "grad_norm": 0.5193579196929932, + "learning_rate": 0.00017854339044883535, + "loss": 0.892, + "step": 4788 + }, + { + "epoch": 0.8527421652421653, + "grad_norm": 0.5053097009658813, + "learning_rate": 0.00017853472600644392, + "loss": 1.0589, + "step": 4789 + }, + { + "epoch": 0.8529202279202279, + "grad_norm": 0.5819617509841919, + "learning_rate": 0.0001785260600253277, + "loss": 1.2646, + "step": 4790 + }, + { + "epoch": 0.8530982905982906, + "grad_norm": 0.5327470302581787, + "learning_rate": 0.00017851739250565645, + "loss": 1.056, + "step": 4791 + }, + { + "epoch": 0.8532763532763533, + "grad_norm": 0.5131269097328186, + "learning_rate": 0.0001785087234476, + "loss": 1.1192, + "step": 4792 + }, + { + "epoch": 0.853454415954416, + "grad_norm": 0.4698086977005005, + "learning_rate": 0.00017850005285132821, + "loss": 0.9849, + "step": 4793 + }, + { + "epoch": 0.8536324786324786, + "grad_norm": 0.5503947734832764, + "learning_rate": 0.00017849138071701092, + "loss": 1.1139, + "step": 4794 + }, + { + "epoch": 0.8538105413105413, + "grad_norm": 0.5120903849601746, + "learning_rate": 0.0001784827070448181, + "loss": 0.9801, + "step": 4795 + }, + { + "epoch": 0.853988603988604, + "grad_norm": 0.47650405764579773, + "learning_rate": 0.00017847403183491968, + "loss": 1.0268, + "step": 4796 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 0.5773387551307678, + "learning_rate": 0.0001784653550874856, + "loss": 1.0336, + "step": 4797 + }, + { + "epoch": 0.8543447293447294, + "grad_norm": 0.545531153678894, + "learning_rate": 0.00017845667680268593, + "loss": 1.0532, + "step": 4798 + }, + { + "epoch": 0.854522792022792, + "grad_norm": 0.533161461353302, + "learning_rate": 0.0001784479969806906, + "loss": 1.1964, + "step": 4799 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 0.5880789160728455, + "learning_rate": 0.00017843931562166977, + "loss": 1.1588, + "step": 4800 + }, + { + "epoch": 0.8548789173789174, + "grad_norm": 0.5381524562835693, + "learning_rate": 0.00017843063272579346, + "loss": 1.1533, + "step": 4801 + }, + { + "epoch": 0.85505698005698, + "grad_norm": 0.6280176639556885, + "learning_rate": 0.00017842194829323187, + "loss": 1.0084, + "step": 4802 + }, + { + "epoch": 0.8552350427350427, + "grad_norm": 0.5098552703857422, + "learning_rate": 0.0001784132623241551, + "loss": 1.0804, + "step": 4803 + }, + { + "epoch": 0.8554131054131054, + "grad_norm": 0.5406526923179626, + "learning_rate": 0.00017840457481873328, + "loss": 1.2571, + "step": 4804 + }, + { + "epoch": 0.8555911680911681, + "grad_norm": 0.5859003663063049, + "learning_rate": 0.00017839588577713678, + "loss": 1.2462, + "step": 4805 + }, + { + "epoch": 0.8557692307692307, + "grad_norm": 0.6209002137184143, + "learning_rate": 0.00017838719519953572, + "loss": 1.307, + "step": 4806 + }, + { + "epoch": 0.8559472934472935, + "grad_norm": 0.525753915309906, + "learning_rate": 0.00017837850308610037, + "loss": 1.2957, + "step": 4807 + }, + { + "epoch": 0.8561253561253561, + "grad_norm": 0.5096195340156555, + "learning_rate": 0.0001783698094370011, + "loss": 1.1433, + "step": 4808 + }, + { + "epoch": 0.8563034188034188, + "grad_norm": 0.5873076915740967, + "learning_rate": 0.0001783611142524082, + "loss": 1.2271, + "step": 4809 + }, + { + "epoch": 0.8564814814814815, + "grad_norm": 0.5093944668769836, + "learning_rate": 0.0001783524175324921, + "loss": 0.8788, + "step": 4810 + }, + { + "epoch": 0.8566595441595442, + "grad_norm": 0.5485084652900696, + "learning_rate": 0.00017834371927742307, + "loss": 1.256, + "step": 4811 + }, + { + "epoch": 0.8568376068376068, + "grad_norm": 0.5808873772621155, + "learning_rate": 0.00017833501948737163, + "loss": 0.9287, + "step": 4812 + }, + { + "epoch": 0.8570156695156695, + "grad_norm": 0.5113978385925293, + "learning_rate": 0.00017832631816250822, + "loss": 1.0372, + "step": 4813 + }, + { + "epoch": 0.8571937321937322, + "grad_norm": 0.5877016186714172, + "learning_rate": 0.0001783176153030033, + "loss": 1.3023, + "step": 4814 + }, + { + "epoch": 0.8573717948717948, + "grad_norm": 0.534328043460846, + "learning_rate": 0.00017830891090902742, + "loss": 1.1023, + "step": 4815 + }, + { + "epoch": 0.8575498575498576, + "grad_norm": 0.5781638026237488, + "learning_rate": 0.0001783002049807511, + "loss": 0.9562, + "step": 4816 + }, + { + "epoch": 0.8577279202279202, + "grad_norm": 0.5760263204574585, + "learning_rate": 0.00017829149751834487, + "loss": 0.8733, + "step": 4817 + }, + { + "epoch": 0.8579059829059829, + "grad_norm": 0.3887255787849426, + "learning_rate": 0.00017828278852197944, + "loss": 0.5949, + "step": 4818 + }, + { + "epoch": 0.8580840455840456, + "grad_norm": 0.47814446687698364, + "learning_rate": 0.00017827407799182537, + "loss": 1.0698, + "step": 4819 + }, + { + "epoch": 0.8582621082621082, + "grad_norm": 0.5520272254943848, + "learning_rate": 0.00017826536592805334, + "loss": 1.1314, + "step": 4820 + }, + { + "epoch": 0.8584401709401709, + "grad_norm": 0.5285319685935974, + "learning_rate": 0.00017825665233083405, + "loss": 1.1618, + "step": 4821 + }, + { + "epoch": 0.8586182336182336, + "grad_norm": 0.6080102324485779, + "learning_rate": 0.0001782479372003382, + "loss": 1.3817, + "step": 4822 + }, + { + "epoch": 0.8587962962962963, + "grad_norm": 0.7474410533905029, + "learning_rate": 0.00017823922053673662, + "loss": 1.1321, + "step": 4823 + }, + { + "epoch": 0.8589743589743589, + "grad_norm": 0.559283435344696, + "learning_rate": 0.0001782305023402, + "loss": 1.1894, + "step": 4824 + }, + { + "epoch": 0.8591524216524217, + "grad_norm": 0.5620571374893188, + "learning_rate": 0.00017822178261089918, + "loss": 1.134, + "step": 4825 + }, + { + "epoch": 0.8593304843304843, + "grad_norm": 0.5553044676780701, + "learning_rate": 0.00017821306134900504, + "loss": 1.3222, + "step": 4826 + }, + { + "epoch": 0.8595085470085471, + "grad_norm": 0.6177778244018555, + "learning_rate": 0.00017820433855468846, + "loss": 1.2545, + "step": 4827 + }, + { + "epoch": 0.8596866096866097, + "grad_norm": 0.656233012676239, + "learning_rate": 0.0001781956142281203, + "loss": 1.1346, + "step": 4828 + }, + { + "epoch": 0.8598646723646723, + "grad_norm": 0.6710973381996155, + "learning_rate": 0.0001781868883694715, + "loss": 1.1361, + "step": 4829 + }, + { + "epoch": 0.8600427350427351, + "grad_norm": 0.5093601942062378, + "learning_rate": 0.0001781781609789131, + "loss": 1.0509, + "step": 4830 + }, + { + "epoch": 0.8602207977207977, + "grad_norm": 0.5707578063011169, + "learning_rate": 0.00017816943205661598, + "loss": 1.0964, + "step": 4831 + }, + { + "epoch": 0.8603988603988604, + "grad_norm": 0.6159597635269165, + "learning_rate": 0.00017816070160275125, + "loss": 1.0322, + "step": 4832 + }, + { + "epoch": 0.8605769230769231, + "grad_norm": 0.5430580377578735, + "learning_rate": 0.0001781519696174899, + "loss": 1.2464, + "step": 4833 + }, + { + "epoch": 0.8607549857549858, + "grad_norm": 0.48104700446128845, + "learning_rate": 0.0001781432361010031, + "loss": 1.1031, + "step": 4834 + }, + { + "epoch": 0.8609330484330484, + "grad_norm": 0.5304946303367615, + "learning_rate": 0.0001781345010534619, + "loss": 1.0281, + "step": 4835 + }, + { + "epoch": 0.8611111111111112, + "grad_norm": 0.5230711698532104, + "learning_rate": 0.00017812576447503742, + "loss": 0.9499, + "step": 4836 + }, + { + "epoch": 0.8612891737891738, + "grad_norm": 0.5363606214523315, + "learning_rate": 0.00017811702636590093, + "loss": 1.1358, + "step": 4837 + }, + { + "epoch": 0.8614672364672364, + "grad_norm": 0.5880044102668762, + "learning_rate": 0.00017810828672622358, + "loss": 1.1765, + "step": 4838 + }, + { + "epoch": 0.8616452991452992, + "grad_norm": 0.5194395184516907, + "learning_rate": 0.0001780995455561766, + "loss": 1.1622, + "step": 4839 + }, + { + "epoch": 0.8618233618233618, + "grad_norm": 0.5114264488220215, + "learning_rate": 0.00017809080285593126, + "loss": 1.0081, + "step": 4840 + }, + { + "epoch": 0.8620014245014245, + "grad_norm": 0.6174240112304688, + "learning_rate": 0.00017808205862565886, + "loss": 1.0745, + "step": 4841 + }, + { + "epoch": 0.8621794871794872, + "grad_norm": 0.5662630200386047, + "learning_rate": 0.0001780733128655307, + "loss": 1.3369, + "step": 4842 + }, + { + "epoch": 0.8623575498575499, + "grad_norm": 0.5917882919311523, + "learning_rate": 0.00017806456557571817, + "loss": 1.1631, + "step": 4843 + }, + { + "epoch": 0.8625356125356125, + "grad_norm": 0.5305736660957336, + "learning_rate": 0.00017805581675639265, + "loss": 0.9875, + "step": 4844 + }, + { + "epoch": 0.8627136752136753, + "grad_norm": 0.5181219577789307, + "learning_rate": 0.00017804706640772556, + "loss": 0.9918, + "step": 4845 + }, + { + "epoch": 0.8628917378917379, + "grad_norm": 0.5467997789382935, + "learning_rate": 0.00017803831452988832, + "loss": 1.1395, + "step": 4846 + }, + { + "epoch": 0.8630698005698005, + "grad_norm": 0.5494031310081482, + "learning_rate": 0.00017802956112305241, + "loss": 1.0312, + "step": 4847 + }, + { + "epoch": 0.8632478632478633, + "grad_norm": 0.5804065465927124, + "learning_rate": 0.00017802080618738931, + "loss": 1.1555, + "step": 4848 + }, + { + "epoch": 0.8634259259259259, + "grad_norm": 0.5424801111221313, + "learning_rate": 0.00017801204972307067, + "loss": 1.0215, + "step": 4849 + }, + { + "epoch": 0.8636039886039886, + "grad_norm": 0.5321891903877258, + "learning_rate": 0.0001780032917302679, + "loss": 1.0187, + "step": 4850 + }, + { + "epoch": 0.8637820512820513, + "grad_norm": 0.5543400049209595, + "learning_rate": 0.0001779945322091527, + "loss": 1.1972, + "step": 4851 + }, + { + "epoch": 0.863960113960114, + "grad_norm": 0.566649317741394, + "learning_rate": 0.00017798577115989668, + "loss": 1.0758, + "step": 4852 + }, + { + "epoch": 0.8641381766381766, + "grad_norm": 0.5538444519042969, + "learning_rate": 0.00017797700858267145, + "loss": 1.1338, + "step": 4853 + }, + { + "epoch": 0.8643162393162394, + "grad_norm": 0.5641313791275024, + "learning_rate": 0.0001779682444776487, + "loss": 1.256, + "step": 4854 + }, + { + "epoch": 0.864494301994302, + "grad_norm": 0.6377350091934204, + "learning_rate": 0.00017795947884500016, + "loss": 1.144, + "step": 4855 + }, + { + "epoch": 0.8646723646723646, + "grad_norm": 0.5581876039505005, + "learning_rate": 0.0001779507116848976, + "loss": 1.3163, + "step": 4856 + }, + { + "epoch": 0.8648504273504274, + "grad_norm": 0.5416772365570068, + "learning_rate": 0.0001779419429975128, + "loss": 1.0219, + "step": 4857 + }, + { + "epoch": 0.86502849002849, + "grad_norm": 0.5450608730316162, + "learning_rate": 0.0001779331727830175, + "loss": 1.0093, + "step": 4858 + }, + { + "epoch": 0.8652065527065527, + "grad_norm": 0.5151242017745972, + "learning_rate": 0.00017792440104158358, + "loss": 1.067, + "step": 4859 + }, + { + "epoch": 0.8653846153846154, + "grad_norm": 0.5225046873092651, + "learning_rate": 0.0001779156277733829, + "loss": 1.0432, + "step": 4860 + }, + { + "epoch": 0.8655626780626781, + "grad_norm": 0.5168602466583252, + "learning_rate": 0.00017790685297858737, + "loss": 0.9665, + "step": 4861 + }, + { + "epoch": 0.8657407407407407, + "grad_norm": 0.5749059319496155, + "learning_rate": 0.00017789807665736889, + "loss": 1.1607, + "step": 4862 + }, + { + "epoch": 0.8659188034188035, + "grad_norm": 0.45656394958496094, + "learning_rate": 0.00017788929880989938, + "loss": 0.8362, + "step": 4863 + }, + { + "epoch": 0.8660968660968661, + "grad_norm": 0.5090615749359131, + "learning_rate": 0.00017788051943635086, + "loss": 0.9553, + "step": 4864 + }, + { + "epoch": 0.8662749287749287, + "grad_norm": 0.5381240248680115, + "learning_rate": 0.0001778717385368954, + "loss": 1.1391, + "step": 4865 + }, + { + "epoch": 0.8664529914529915, + "grad_norm": 0.522720456123352, + "learning_rate": 0.00017786295611170493, + "loss": 1.1869, + "step": 4866 + }, + { + "epoch": 0.8666310541310541, + "grad_norm": 0.530986487865448, + "learning_rate": 0.0001778541721609516, + "loss": 1.1046, + "step": 4867 + }, + { + "epoch": 0.8668091168091168, + "grad_norm": 0.5065864324569702, + "learning_rate": 0.0001778453866848075, + "loss": 1.008, + "step": 4868 + }, + { + "epoch": 0.8669871794871795, + "grad_norm": 0.5541394352912903, + "learning_rate": 0.00017783659968344476, + "loss": 1.0004, + "step": 4869 + }, + { + "epoch": 0.8671652421652422, + "grad_norm": 0.5059576630592346, + "learning_rate": 0.00017782781115703556, + "loss": 1.128, + "step": 4870 + }, + { + "epoch": 0.8673433048433048, + "grad_norm": 0.5052187442779541, + "learning_rate": 0.00017781902110575203, + "loss": 0.8544, + "step": 4871 + }, + { + "epoch": 0.8675213675213675, + "grad_norm": 0.5383397340774536, + "learning_rate": 0.00017781022952976646, + "loss": 1.1411, + "step": 4872 + }, + { + "epoch": 0.8676994301994302, + "grad_norm": 0.4760429859161377, + "learning_rate": 0.00017780143642925106, + "loss": 0.8246, + "step": 4873 + }, + { + "epoch": 0.8678774928774928, + "grad_norm": 0.5480535626411438, + "learning_rate": 0.00017779264180437817, + "loss": 1.013, + "step": 4874 + }, + { + "epoch": 0.8680555555555556, + "grad_norm": 0.5303317904472351, + "learning_rate": 0.00017778384565532004, + "loss": 1.0201, + "step": 4875 + }, + { + "epoch": 0.8682336182336182, + "grad_norm": 0.5365355014801025, + "learning_rate": 0.00017777504798224903, + "loss": 1.1107, + "step": 4876 + }, + { + "epoch": 0.8684116809116809, + "grad_norm": 0.5173360705375671, + "learning_rate": 0.00017776624878533754, + "loss": 1.0808, + "step": 4877 + }, + { + "epoch": 0.8685897435897436, + "grad_norm": 0.5088842511177063, + "learning_rate": 0.00017775744806475792, + "loss": 0.995, + "step": 4878 + }, + { + "epoch": 0.8687678062678063, + "grad_norm": 0.5796698927879333, + "learning_rate": 0.00017774864582068264, + "loss": 1.1485, + "step": 4879 + }, + { + "epoch": 0.8689458689458689, + "grad_norm": 0.5719375610351562, + "learning_rate": 0.00017773984205328417, + "loss": 1.0133, + "step": 4880 + }, + { + "epoch": 0.8691239316239316, + "grad_norm": 0.6396418213844299, + "learning_rate": 0.00017773103676273498, + "loss": 1.0932, + "step": 4881 + }, + { + "epoch": 0.8693019943019943, + "grad_norm": 0.5602468252182007, + "learning_rate": 0.00017772222994920763, + "loss": 0.9702, + "step": 4882 + }, + { + "epoch": 0.8694800569800569, + "grad_norm": 0.5167748332023621, + "learning_rate": 0.00017771342161287457, + "loss": 1.0528, + "step": 4883 + }, + { + "epoch": 0.8696581196581197, + "grad_norm": 0.5572916865348816, + "learning_rate": 0.00017770461175390848, + "loss": 1.1341, + "step": 4884 + }, + { + "epoch": 0.8698361823361823, + "grad_norm": 0.6666276454925537, + "learning_rate": 0.00017769580037248195, + "loss": 1.1948, + "step": 4885 + }, + { + "epoch": 0.8700142450142451, + "grad_norm": 0.5348601937294006, + "learning_rate": 0.0001776869874687676, + "loss": 1.0562, + "step": 4886 + }, + { + "epoch": 0.8701923076923077, + "grad_norm": 0.5449648499488831, + "learning_rate": 0.00017767817304293812, + "loss": 0.988, + "step": 4887 + }, + { + "epoch": 0.8703703703703703, + "grad_norm": 0.5995045304298401, + "learning_rate": 0.0001776693570951662, + "loss": 1.2526, + "step": 4888 + }, + { + "epoch": 0.8705484330484331, + "grad_norm": 0.6575320959091187, + "learning_rate": 0.00017766053962562457, + "loss": 1.1717, + "step": 4889 + }, + { + "epoch": 0.8707264957264957, + "grad_norm": 0.5882139801979065, + "learning_rate": 0.00017765172063448597, + "loss": 1.238, + "step": 4890 + }, + { + "epoch": 0.8709045584045584, + "grad_norm": 0.5908389091491699, + "learning_rate": 0.00017764290012192325, + "loss": 1.0606, + "step": 4891 + }, + { + "epoch": 0.8710826210826211, + "grad_norm": 0.6169339418411255, + "learning_rate": 0.00017763407808810917, + "loss": 1.1456, + "step": 4892 + }, + { + "epoch": 0.8712606837606838, + "grad_norm": 0.5916035771369934, + "learning_rate": 0.0001776252545332166, + "loss": 1.0026, + "step": 4893 + }, + { + "epoch": 0.8714387464387464, + "grad_norm": 0.539995551109314, + "learning_rate": 0.00017761642945741843, + "loss": 1.2397, + "step": 4894 + }, + { + "epoch": 0.8716168091168092, + "grad_norm": 0.5346137881278992, + "learning_rate": 0.00017760760286088755, + "loss": 1.1232, + "step": 4895 + }, + { + "epoch": 0.8717948717948718, + "grad_norm": 0.570202112197876, + "learning_rate": 0.00017759877474379692, + "loss": 1.0708, + "step": 4896 + }, + { + "epoch": 0.8719729344729344, + "grad_norm": 0.5023398399353027, + "learning_rate": 0.00017758994510631948, + "loss": 1.1056, + "step": 4897 + }, + { + "epoch": 0.8721509971509972, + "grad_norm": 0.5447137951850891, + "learning_rate": 0.00017758111394862826, + "loss": 0.8776, + "step": 4898 + }, + { + "epoch": 0.8723290598290598, + "grad_norm": 0.5193906426429749, + "learning_rate": 0.00017757228127089625, + "loss": 0.9959, + "step": 4899 + }, + { + "epoch": 0.8725071225071225, + "grad_norm": 0.5958787798881531, + "learning_rate": 0.00017756344707329656, + "loss": 1.092, + "step": 4900 + }, + { + "epoch": 0.8726851851851852, + "grad_norm": 0.521045982837677, + "learning_rate": 0.00017755461135600221, + "loss": 0.9864, + "step": 4901 + }, + { + "epoch": 0.8728632478632479, + "grad_norm": 0.5257635116577148, + "learning_rate": 0.00017754577411918638, + "loss": 1.216, + "step": 4902 + }, + { + "epoch": 0.8730413105413105, + "grad_norm": 0.5425964593887329, + "learning_rate": 0.0001775369353630222, + "loss": 1.1432, + "step": 4903 + }, + { + "epoch": 0.8732193732193733, + "grad_norm": 0.47995322942733765, + "learning_rate": 0.00017752809508768286, + "loss": 1.0227, + "step": 4904 + }, + { + "epoch": 0.8733974358974359, + "grad_norm": 0.5747429728507996, + "learning_rate": 0.0001775192532933415, + "loss": 0.9984, + "step": 4905 + }, + { + "epoch": 0.8735754985754985, + "grad_norm": 0.5745723247528076, + "learning_rate": 0.00017751040998017142, + "loss": 1.2559, + "step": 4906 + }, + { + "epoch": 0.8737535612535613, + "grad_norm": 0.6114141941070557, + "learning_rate": 0.0001775015651483459, + "loss": 1.3224, + "step": 4907 + }, + { + "epoch": 0.8739316239316239, + "grad_norm": 0.4757187068462372, + "learning_rate": 0.00017749271879803817, + "loss": 1.0352, + "step": 4908 + }, + { + "epoch": 0.8741096866096866, + "grad_norm": 0.48644450306892395, + "learning_rate": 0.0001774838709294216, + "loss": 1.0876, + "step": 4909 + }, + { + "epoch": 0.8742877492877493, + "grad_norm": 0.5652037262916565, + "learning_rate": 0.00017747502154266955, + "loss": 0.9189, + "step": 4910 + }, + { + "epoch": 0.874465811965812, + "grad_norm": 0.5289644002914429, + "learning_rate": 0.00017746617063795538, + "loss": 0.9431, + "step": 4911 + }, + { + "epoch": 0.8746438746438746, + "grad_norm": 0.594656229019165, + "learning_rate": 0.00017745731821545253, + "loss": 1.2408, + "step": 4912 + }, + { + "epoch": 0.8748219373219374, + "grad_norm": 0.5693240165710449, + "learning_rate": 0.0001774484642753344, + "loss": 1.347, + "step": 4913 + }, + { + "epoch": 0.875, + "grad_norm": 0.5291008949279785, + "learning_rate": 0.00017743960881777456, + "loss": 1.161, + "step": 4914 + }, + { + "epoch": 0.8751780626780626, + "grad_norm": 0.5958300232887268, + "learning_rate": 0.00017743075184294642, + "loss": 1.2058, + "step": 4915 + }, + { + "epoch": 0.8753561253561254, + "grad_norm": 0.513884425163269, + "learning_rate": 0.00017742189335102354, + "loss": 1.0952, + "step": 4916 + }, + { + "epoch": 0.875534188034188, + "grad_norm": 0.5860681533813477, + "learning_rate": 0.00017741303334217948, + "loss": 1.1801, + "step": 4917 + }, + { + "epoch": 0.8757122507122507, + "grad_norm": 0.47962820529937744, + "learning_rate": 0.00017740417181658788, + "loss": 1.0785, + "step": 4918 + }, + { + "epoch": 0.8758903133903134, + "grad_norm": 0.5110440254211426, + "learning_rate": 0.00017739530877442227, + "loss": 1.1385, + "step": 4919 + }, + { + "epoch": 0.8760683760683761, + "grad_norm": 0.5106285214424133, + "learning_rate": 0.00017738644421585643, + "loss": 1.1204, + "step": 4920 + }, + { + "epoch": 0.8762464387464387, + "grad_norm": 0.5709205865859985, + "learning_rate": 0.00017737757814106393, + "loss": 1.0108, + "step": 4921 + }, + { + "epoch": 0.8764245014245015, + "grad_norm": 0.5850250124931335, + "learning_rate": 0.0001773687105502185, + "loss": 1.0059, + "step": 4922 + }, + { + "epoch": 0.8766025641025641, + "grad_norm": 0.5194727778434753, + "learning_rate": 0.00017735984144349396, + "loss": 0.9466, + "step": 4923 + }, + { + "epoch": 0.8767806267806267, + "grad_norm": 0.5246787667274475, + "learning_rate": 0.000177350970821064, + "loss": 1.1336, + "step": 4924 + }, + { + "epoch": 0.8769586894586895, + "grad_norm": 0.5798323154449463, + "learning_rate": 0.00017734209868310244, + "loss": 1.1641, + "step": 4925 + }, + { + "epoch": 0.8771367521367521, + "grad_norm": 0.5188565850257874, + "learning_rate": 0.00017733322502978314, + "loss": 0.9959, + "step": 4926 + }, + { + "epoch": 0.8773148148148148, + "grad_norm": 0.5969653725624084, + "learning_rate": 0.00017732434986127995, + "loss": 1.2162, + "step": 4927 + }, + { + "epoch": 0.8774928774928775, + "grad_norm": 0.5520089268684387, + "learning_rate": 0.00017731547317776674, + "loss": 1.0163, + "step": 4928 + }, + { + "epoch": 0.8776709401709402, + "grad_norm": 0.48789507150650024, + "learning_rate": 0.00017730659497941745, + "loss": 0.9757, + "step": 4929 + }, + { + "epoch": 0.8778490028490028, + "grad_norm": 0.6034960746765137, + "learning_rate": 0.000177297715266406, + "loss": 1.1278, + "step": 4930 + }, + { + "epoch": 0.8780270655270656, + "grad_norm": 0.53016597032547, + "learning_rate": 0.00017728883403890638, + "loss": 1.0637, + "step": 4931 + }, + { + "epoch": 0.8782051282051282, + "grad_norm": 0.5073726177215576, + "learning_rate": 0.00017727995129709266, + "loss": 1.1491, + "step": 4932 + }, + { + "epoch": 0.8783831908831908, + "grad_norm": 0.540605366230011, + "learning_rate": 0.00017727106704113878, + "loss": 1.0133, + "step": 4933 + }, + { + "epoch": 0.8785612535612536, + "grad_norm": 0.5346775054931641, + "learning_rate": 0.0001772621812712189, + "loss": 1.1781, + "step": 4934 + }, + { + "epoch": 0.8787393162393162, + "grad_norm": 0.5659036040306091, + "learning_rate": 0.00017725329398750702, + "loss": 1.1023, + "step": 4935 + }, + { + "epoch": 0.8789173789173789, + "grad_norm": 0.591063380241394, + "learning_rate": 0.00017724440519017738, + "loss": 1.0298, + "step": 4936 + }, + { + "epoch": 0.8790954415954416, + "grad_norm": 0.5173781514167786, + "learning_rate": 0.0001772355148794041, + "loss": 1.0483, + "step": 4937 + }, + { + "epoch": 0.8792735042735043, + "grad_norm": 0.5405352711677551, + "learning_rate": 0.0001772266230553613, + "loss": 1.0716, + "step": 4938 + }, + { + "epoch": 0.8794515669515669, + "grad_norm": 0.518442690372467, + "learning_rate": 0.00017721772971822323, + "loss": 1.1373, + "step": 4939 + }, + { + "epoch": 0.8796296296296297, + "grad_norm": 0.533673107624054, + "learning_rate": 0.0001772088348681642, + "loss": 1.0489, + "step": 4940 + }, + { + "epoch": 0.8798076923076923, + "grad_norm": 0.46117857098579407, + "learning_rate": 0.0001771999385053584, + "loss": 1.0297, + "step": 4941 + }, + { + "epoch": 0.8799857549857549, + "grad_norm": 0.4687997102737427, + "learning_rate": 0.0001771910406299802, + "loss": 1.071, + "step": 4942 + }, + { + "epoch": 0.8801638176638177, + "grad_norm": 0.5064153075218201, + "learning_rate": 0.0001771821412422039, + "loss": 0.9518, + "step": 4943 + }, + { + "epoch": 0.8803418803418803, + "grad_norm": 0.6561978459358215, + "learning_rate": 0.00017717324034220385, + "loss": 1.11, + "step": 4944 + }, + { + "epoch": 0.8805199430199431, + "grad_norm": 0.5551498532295227, + "learning_rate": 0.00017716433793015454, + "loss": 0.9719, + "step": 4945 + }, + { + "epoch": 0.8806980056980057, + "grad_norm": 0.47059500217437744, + "learning_rate": 0.00017715543400623025, + "loss": 0.8891, + "step": 4946 + }, + { + "epoch": 0.8808760683760684, + "grad_norm": 0.5035740733146667, + "learning_rate": 0.00017714652857060554, + "loss": 0.9671, + "step": 4947 + }, + { + "epoch": 0.8810541310541311, + "grad_norm": 0.4599960446357727, + "learning_rate": 0.00017713762162345487, + "loss": 0.9588, + "step": 4948 + }, + { + "epoch": 0.8812321937321937, + "grad_norm": 0.5087231397628784, + "learning_rate": 0.0001771287131649527, + "loss": 1.1433, + "step": 4949 + }, + { + "epoch": 0.8814102564102564, + "grad_norm": 0.5609854459762573, + "learning_rate": 0.00017711980319527366, + "loss": 1.2022, + "step": 4950 + }, + { + "epoch": 0.8815883190883191, + "grad_norm": 0.49460700154304504, + "learning_rate": 0.00017711089171459227, + "loss": 1.019, + "step": 4951 + }, + { + "epoch": 0.8817663817663818, + "grad_norm": 0.5047259330749512, + "learning_rate": 0.00017710197872308314, + "loss": 0.8301, + "step": 4952 + }, + { + "epoch": 0.8819444444444444, + "grad_norm": 0.5784406065940857, + "learning_rate": 0.0001770930642209209, + "loss": 0.9336, + "step": 4953 + }, + { + "epoch": 0.8821225071225072, + "grad_norm": 0.5037121772766113, + "learning_rate": 0.00017708414820828022, + "loss": 1.0199, + "step": 4954 + }, + { + "epoch": 0.8823005698005698, + "grad_norm": 0.5683804750442505, + "learning_rate": 0.00017707523068533575, + "loss": 0.9758, + "step": 4955 + }, + { + "epoch": 0.8824786324786325, + "grad_norm": 0.5167922973632812, + "learning_rate": 0.0001770663116522623, + "loss": 1.0389, + "step": 4956 + }, + { + "epoch": 0.8826566951566952, + "grad_norm": 0.5813606381416321, + "learning_rate": 0.0001770573911092345, + "loss": 1.3998, + "step": 4957 + }, + { + "epoch": 0.8828347578347578, + "grad_norm": 0.5280475616455078, + "learning_rate": 0.00017704846905642723, + "loss": 1.0545, + "step": 4958 + }, + { + "epoch": 0.8830128205128205, + "grad_norm": 0.5421732068061829, + "learning_rate": 0.00017703954549401528, + "loss": 0.899, + "step": 4959 + }, + { + "epoch": 0.8831908831908832, + "grad_norm": 0.5177720189094543, + "learning_rate": 0.00017703062042217344, + "loss": 0.975, + "step": 4960 + }, + { + "epoch": 0.8833689458689459, + "grad_norm": 0.639327883720398, + "learning_rate": 0.00017702169384107666, + "loss": 1.1936, + "step": 4961 + }, + { + "epoch": 0.8835470085470085, + "grad_norm": 0.5201572179794312, + "learning_rate": 0.00017701276575089975, + "loss": 0.9891, + "step": 4962 + }, + { + "epoch": 0.8837250712250713, + "grad_norm": 0.5304145216941833, + "learning_rate": 0.00017700383615181767, + "loss": 1.0569, + "step": 4963 + }, + { + "epoch": 0.8839031339031339, + "grad_norm": 0.6068132519721985, + "learning_rate": 0.00017699490504400538, + "loss": 1.2653, + "step": 4964 + }, + { + "epoch": 0.8840811965811965, + "grad_norm": 0.597895085811615, + "learning_rate": 0.00017698597242763787, + "loss": 1.2577, + "step": 4965 + }, + { + "epoch": 0.8842592592592593, + "grad_norm": 0.5356902480125427, + "learning_rate": 0.00017697703830289017, + "loss": 1.1056, + "step": 4966 + }, + { + "epoch": 0.8844373219373219, + "grad_norm": 0.5429540872573853, + "learning_rate": 0.0001769681026699373, + "loss": 1.0951, + "step": 4967 + }, + { + "epoch": 0.8846153846153846, + "grad_norm": 0.5789309144020081, + "learning_rate": 0.00017695916552895436, + "loss": 1.0786, + "step": 4968 + }, + { + "epoch": 0.8847934472934473, + "grad_norm": 0.5621341466903687, + "learning_rate": 0.0001769502268801164, + "loss": 1.0645, + "step": 4969 + }, + { + "epoch": 0.88497150997151, + "grad_norm": 0.5879453420639038, + "learning_rate": 0.00017694128672359865, + "loss": 1.2171, + "step": 4970 + }, + { + "epoch": 0.8851495726495726, + "grad_norm": 0.5005951523780823, + "learning_rate": 0.0001769323450595762, + "loss": 1.0725, + "step": 4971 + }, + { + "epoch": 0.8853276353276354, + "grad_norm": 0.5439660549163818, + "learning_rate": 0.00017692340188822425, + "loss": 1.162, + "step": 4972 + }, + { + "epoch": 0.885505698005698, + "grad_norm": 0.6309837698936462, + "learning_rate": 0.00017691445720971802, + "loss": 1.2861, + "step": 4973 + }, + { + "epoch": 0.8856837606837606, + "grad_norm": 0.4997463822364807, + "learning_rate": 0.00017690551102423282, + "loss": 1.1887, + "step": 4974 + }, + { + "epoch": 0.8858618233618234, + "grad_norm": 0.5430852174758911, + "learning_rate": 0.00017689656333194385, + "loss": 1.1231, + "step": 4975 + }, + { + "epoch": 0.886039886039886, + "grad_norm": 0.5414215922355652, + "learning_rate": 0.00017688761413302644, + "loss": 1.2345, + "step": 4976 + }, + { + "epoch": 0.8862179487179487, + "grad_norm": 0.5594443082809448, + "learning_rate": 0.00017687866342765601, + "loss": 1.0775, + "step": 4977 + }, + { + "epoch": 0.8863960113960114, + "grad_norm": 0.5827134847640991, + "learning_rate": 0.00017686971121600787, + "loss": 1.0609, + "step": 4978 + }, + { + "epoch": 0.8865740740740741, + "grad_norm": 0.5075414776802063, + "learning_rate": 0.00017686075749825738, + "loss": 0.796, + "step": 4979 + }, + { + "epoch": 0.8867521367521367, + "grad_norm": 0.6007544994354248, + "learning_rate": 0.00017685180227458003, + "loss": 1.1716, + "step": 4980 + }, + { + "epoch": 0.8869301994301995, + "grad_norm": 0.6458030343055725, + "learning_rate": 0.00017684284554515128, + "loss": 1.1945, + "step": 4981 + }, + { + "epoch": 0.8871082621082621, + "grad_norm": 0.5519212484359741, + "learning_rate": 0.00017683388731014657, + "loss": 1.2571, + "step": 4982 + }, + { + "epoch": 0.8872863247863247, + "grad_norm": 0.5079960227012634, + "learning_rate": 0.00017682492756974146, + "loss": 1.1186, + "step": 4983 + }, + { + "epoch": 0.8874643874643875, + "grad_norm": 0.63576740026474, + "learning_rate": 0.00017681596632411147, + "loss": 1.389, + "step": 4984 + }, + { + "epoch": 0.8876424501424501, + "grad_norm": 0.43325698375701904, + "learning_rate": 0.0001768070035734322, + "loss": 0.7757, + "step": 4985 + }, + { + "epoch": 0.8878205128205128, + "grad_norm": 0.49492064118385315, + "learning_rate": 0.00017679803931787923, + "loss": 1.0096, + "step": 4986 + }, + { + "epoch": 0.8879985754985755, + "grad_norm": 0.5561224222183228, + "learning_rate": 0.00017678907355762825, + "loss": 0.952, + "step": 4987 + }, + { + "epoch": 0.8881766381766382, + "grad_norm": 0.5392457246780396, + "learning_rate": 0.00017678010629285486, + "loss": 1.0442, + "step": 4988 + }, + { + "epoch": 0.8883547008547008, + "grad_norm": 0.4659234881401062, + "learning_rate": 0.00017677113752373482, + "loss": 0.8668, + "step": 4989 + }, + { + "epoch": 0.8885327635327636, + "grad_norm": 0.5139175057411194, + "learning_rate": 0.0001767621672504438, + "loss": 0.8386, + "step": 4990 + }, + { + "epoch": 0.8887108262108262, + "grad_norm": 0.5395823121070862, + "learning_rate": 0.00017675319547315755, + "loss": 0.9754, + "step": 4991 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 0.4751867949962616, + "learning_rate": 0.0001767442221920519, + "loss": 0.8775, + "step": 4992 + }, + { + "epoch": 0.8890669515669516, + "grad_norm": 0.5728281736373901, + "learning_rate": 0.00017673524740730265, + "loss": 1.2807, + "step": 4993 + }, + { + "epoch": 0.8892450142450142, + "grad_norm": 0.5545622110366821, + "learning_rate": 0.00017672627111908558, + "loss": 1.0039, + "step": 4994 + }, + { + "epoch": 0.8894230769230769, + "grad_norm": 0.5127374529838562, + "learning_rate": 0.00017671729332757665, + "loss": 1.0505, + "step": 4995 + }, + { + "epoch": 0.8896011396011396, + "grad_norm": 0.5238714218139648, + "learning_rate": 0.00017670831403295175, + "loss": 1.1775, + "step": 4996 + }, + { + "epoch": 0.8897792022792023, + "grad_norm": 0.5610160827636719, + "learning_rate": 0.00017669933323538674, + "loss": 1.0555, + "step": 4997 + }, + { + "epoch": 0.8899572649572649, + "grad_norm": 0.5481634736061096, + "learning_rate": 0.00017669035093505762, + "loss": 1.0802, + "step": 4998 + }, + { + "epoch": 0.8901353276353277, + "grad_norm": 0.4725174307823181, + "learning_rate": 0.0001766813671321404, + "loss": 0.9611, + "step": 4999 + }, + { + "epoch": 0.8903133903133903, + "grad_norm": 0.5184635519981384, + "learning_rate": 0.0001766723818268111, + "loss": 1.1659, + "step": 5000 + }, + { + "epoch": 0.8904914529914529, + "grad_norm": 0.5503578186035156, + "learning_rate": 0.00017666339501924575, + "loss": 1.2165, + "step": 5001 + }, + { + "epoch": 0.8906695156695157, + "grad_norm": 0.5299594402313232, + "learning_rate": 0.0001766544067096204, + "loss": 1.0196, + "step": 5002 + }, + { + "epoch": 0.8908475783475783, + "grad_norm": 0.5673944354057312, + "learning_rate": 0.00017664541689811118, + "loss": 1.2058, + "step": 5003 + }, + { + "epoch": 0.8910256410256411, + "grad_norm": 0.6057320833206177, + "learning_rate": 0.00017663642558489426, + "loss": 1.0136, + "step": 5004 + }, + { + "epoch": 0.8912037037037037, + "grad_norm": 0.4767026901245117, + "learning_rate": 0.00017662743277014578, + "loss": 0.8522, + "step": 5005 + }, + { + "epoch": 0.8913817663817664, + "grad_norm": 0.5346270203590393, + "learning_rate": 0.00017661843845404192, + "loss": 1.1568, + "step": 5006 + }, + { + "epoch": 0.8915598290598291, + "grad_norm": 0.5365738868713379, + "learning_rate": 0.00017660944263675891, + "loss": 1.0488, + "step": 5007 + }, + { + "epoch": 0.8917378917378918, + "grad_norm": 0.5536269545555115, + "learning_rate": 0.00017660044531847305, + "loss": 1.1216, + "step": 5008 + }, + { + "epoch": 0.8919159544159544, + "grad_norm": 0.6325978636741638, + "learning_rate": 0.00017659144649936055, + "loss": 1.2843, + "step": 5009 + }, + { + "epoch": 0.8920940170940171, + "grad_norm": 0.5890641212463379, + "learning_rate": 0.00017658244617959777, + "loss": 1.1976, + "step": 5010 + }, + { + "epoch": 0.8922720797720798, + "grad_norm": 0.604870080947876, + "learning_rate": 0.00017657344435936107, + "loss": 1.2881, + "step": 5011 + }, + { + "epoch": 0.8924501424501424, + "grad_norm": 0.49805206060409546, + "learning_rate": 0.00017656444103882676, + "loss": 0.8998, + "step": 5012 + }, + { + "epoch": 0.8926282051282052, + "grad_norm": 0.506926953792572, + "learning_rate": 0.0001765554362181713, + "loss": 1.0731, + "step": 5013 + }, + { + "epoch": 0.8928062678062678, + "grad_norm": 0.5353260636329651, + "learning_rate": 0.0001765464298975711, + "loss": 1.0676, + "step": 5014 + }, + { + "epoch": 0.8929843304843305, + "grad_norm": 0.5641853213310242, + "learning_rate": 0.0001765374220772026, + "loss": 0.9606, + "step": 5015 + }, + { + "epoch": 0.8931623931623932, + "grad_norm": 0.5049327611923218, + "learning_rate": 0.00017652841275724233, + "loss": 1.009, + "step": 5016 + }, + { + "epoch": 0.8933404558404558, + "grad_norm": 0.6255155205726624, + "learning_rate": 0.0001765194019378668, + "loss": 1.138, + "step": 5017 + }, + { + "epoch": 0.8935185185185185, + "grad_norm": 0.5816851854324341, + "learning_rate": 0.00017651038961925247, + "loss": 1.3398, + "step": 5018 + }, + { + "epoch": 0.8936965811965812, + "grad_norm": 0.5188020467758179, + "learning_rate": 0.00017650137580157605, + "loss": 1.0126, + "step": 5019 + }, + { + "epoch": 0.8938746438746439, + "grad_norm": 0.5231554508209229, + "learning_rate": 0.00017649236048501406, + "loss": 1.0328, + "step": 5020 + }, + { + "epoch": 0.8940527065527065, + "grad_norm": 0.7638634443283081, + "learning_rate": 0.0001764833436697432, + "loss": 1.3016, + "step": 5021 + }, + { + "epoch": 0.8942307692307693, + "grad_norm": 0.5354094505310059, + "learning_rate": 0.00017647432535594008, + "loss": 1.0646, + "step": 5022 + }, + { + "epoch": 0.8944088319088319, + "grad_norm": 0.6938086748123169, + "learning_rate": 0.0001764653055437814, + "loss": 1.2051, + "step": 5023 + }, + { + "epoch": 0.8945868945868946, + "grad_norm": 0.5546849370002747, + "learning_rate": 0.00017645628423344393, + "loss": 1.0671, + "step": 5024 + }, + { + "epoch": 0.8947649572649573, + "grad_norm": 0.49294665455818176, + "learning_rate": 0.0001764472614251044, + "loss": 1.0328, + "step": 5025 + }, + { + "epoch": 0.89494301994302, + "grad_norm": 0.5965796113014221, + "learning_rate": 0.00017643823711893956, + "loss": 1.0741, + "step": 5026 + }, + { + "epoch": 0.8951210826210826, + "grad_norm": 0.4846448302268982, + "learning_rate": 0.00017642921131512626, + "loss": 1.0409, + "step": 5027 + }, + { + "epoch": 0.8952991452991453, + "grad_norm": 0.5767390131950378, + "learning_rate": 0.00017642018401384135, + "loss": 1.018, + "step": 5028 + }, + { + "epoch": 0.895477207977208, + "grad_norm": 0.503027617931366, + "learning_rate": 0.00017641115521526167, + "loss": 1.0002, + "step": 5029 + }, + { + "epoch": 0.8956552706552706, + "grad_norm": 0.6668619513511658, + "learning_rate": 0.00017640212491956412, + "loss": 1.2154, + "step": 5030 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 0.5544148683547974, + "learning_rate": 0.00017639309312692566, + "loss": 1.2701, + "step": 5031 + }, + { + "epoch": 0.896011396011396, + "grad_norm": 0.6026872992515564, + "learning_rate": 0.00017638405983752323, + "loss": 0.9335, + "step": 5032 + }, + { + "epoch": 0.8961894586894587, + "grad_norm": 0.6288694143295288, + "learning_rate": 0.00017637502505153384, + "loss": 0.9075, + "step": 5033 + }, + { + "epoch": 0.8963675213675214, + "grad_norm": 0.4890204966068268, + "learning_rate": 0.00017636598876913446, + "loss": 0.8492, + "step": 5034 + }, + { + "epoch": 0.896545584045584, + "grad_norm": 0.5746598243713379, + "learning_rate": 0.00017635695099050218, + "loss": 1.1557, + "step": 5035 + }, + { + "epoch": 0.8967236467236467, + "grad_norm": 0.5165683031082153, + "learning_rate": 0.00017634791171581405, + "loss": 1.0899, + "step": 5036 + }, + { + "epoch": 0.8969017094017094, + "grad_norm": 0.4621037244796753, + "learning_rate": 0.0001763388709452472, + "loss": 1.0457, + "step": 5037 + }, + { + "epoch": 0.8970797720797721, + "grad_norm": 0.532358705997467, + "learning_rate": 0.00017632982867897876, + "loss": 1.139, + "step": 5038 + }, + { + "epoch": 0.8972578347578347, + "grad_norm": 0.5794399976730347, + "learning_rate": 0.00017632078491718587, + "loss": 1.031, + "step": 5039 + }, + { + "epoch": 0.8974358974358975, + "grad_norm": 0.5031905174255371, + "learning_rate": 0.00017631173966004576, + "loss": 0.9508, + "step": 5040 + }, + { + "epoch": 0.8976139601139601, + "grad_norm": 0.6528840065002441, + "learning_rate": 0.00017630269290773564, + "loss": 0.9974, + "step": 5041 + }, + { + "epoch": 0.8977920227920227, + "grad_norm": 0.6007558703422546, + "learning_rate": 0.00017629364466043273, + "loss": 1.0993, + "step": 5042 + }, + { + "epoch": 0.8979700854700855, + "grad_norm": 0.5104095339775085, + "learning_rate": 0.00017628459491831437, + "loss": 0.9175, + "step": 5043 + }, + { + "epoch": 0.8981481481481481, + "grad_norm": 0.5285516977310181, + "learning_rate": 0.00017627554368155782, + "loss": 0.998, + "step": 5044 + }, + { + "epoch": 0.8983262108262108, + "grad_norm": 0.5629046559333801, + "learning_rate": 0.00017626649095034045, + "loss": 1.2021, + "step": 5045 + }, + { + "epoch": 0.8985042735042735, + "grad_norm": 0.57548987865448, + "learning_rate": 0.00017625743672483962, + "loss": 1.2076, + "step": 5046 + }, + { + "epoch": 0.8986823361823362, + "grad_norm": 0.4883024990558624, + "learning_rate": 0.0001762483810052327, + "loss": 0.9761, + "step": 5047 + }, + { + "epoch": 0.8988603988603988, + "grad_norm": 0.6378034949302673, + "learning_rate": 0.0001762393237916972, + "loss": 1.2266, + "step": 5048 + }, + { + "epoch": 0.8990384615384616, + "grad_norm": 0.5201624035835266, + "learning_rate": 0.0001762302650844105, + "loss": 1.247, + "step": 5049 + }, + { + "epoch": 0.8992165242165242, + "grad_norm": 0.5438048243522644, + "learning_rate": 0.0001762212048835501, + "loss": 0.993, + "step": 5050 + }, + { + "epoch": 0.8993945868945868, + "grad_norm": 0.5928253531455994, + "learning_rate": 0.00017621214318929354, + "loss": 1.0469, + "step": 5051 + }, + { + "epoch": 0.8995726495726496, + "grad_norm": 0.6437996625900269, + "learning_rate": 0.00017620308000181831, + "loss": 1.3136, + "step": 5052 + }, + { + "epoch": 0.8997507122507122, + "grad_norm": 0.5961456298828125, + "learning_rate": 0.00017619401532130208, + "loss": 1.1495, + "step": 5053 + }, + { + "epoch": 0.8999287749287749, + "grad_norm": 0.497388631105423, + "learning_rate": 0.0001761849491479224, + "loss": 0.7783, + "step": 5054 + }, + { + "epoch": 0.9001068376068376, + "grad_norm": 0.5984451174736023, + "learning_rate": 0.00017617588148185687, + "loss": 1.3115, + "step": 5055 + }, + { + "epoch": 0.9002849002849003, + "grad_norm": 0.549163818359375, + "learning_rate": 0.0001761668123232832, + "loss": 1.1649, + "step": 5056 + }, + { + "epoch": 0.9004629629629629, + "grad_norm": 0.5831968188285828, + "learning_rate": 0.00017615774167237903, + "loss": 1.1749, + "step": 5057 + }, + { + "epoch": 0.9006410256410257, + "grad_norm": 0.5111076235771179, + "learning_rate": 0.00017614866952932214, + "loss": 0.8936, + "step": 5058 + }, + { + "epoch": 0.9008190883190883, + "grad_norm": 0.5740947723388672, + "learning_rate": 0.00017613959589429028, + "loss": 1.2606, + "step": 5059 + }, + { + "epoch": 0.9009971509971509, + "grad_norm": 0.5881099700927734, + "learning_rate": 0.0001761305207674612, + "loss": 1.3682, + "step": 5060 + }, + { + "epoch": 0.9011752136752137, + "grad_norm": 0.5007091760635376, + "learning_rate": 0.00017612144414901268, + "loss": 0.7788, + "step": 5061 + }, + { + "epoch": 0.9013532763532763, + "grad_norm": 0.5127760171890259, + "learning_rate": 0.00017611236603912262, + "loss": 1.0519, + "step": 5062 + }, + { + "epoch": 0.9015313390313391, + "grad_norm": 0.6185184121131897, + "learning_rate": 0.00017610328643796882, + "loss": 1.1672, + "step": 5063 + }, + { + "epoch": 0.9017094017094017, + "grad_norm": 0.49707287549972534, + "learning_rate": 0.00017609420534572926, + "loss": 1.1865, + "step": 5064 + }, + { + "epoch": 0.9018874643874644, + "grad_norm": 0.5667552351951599, + "learning_rate": 0.0001760851227625818, + "loss": 1.1388, + "step": 5065 + }, + { + "epoch": 0.9020655270655271, + "grad_norm": 0.50298011302948, + "learning_rate": 0.00017607603868870442, + "loss": 0.9552, + "step": 5066 + }, + { + "epoch": 0.9022435897435898, + "grad_norm": 0.5709219574928284, + "learning_rate": 0.0001760669531242751, + "loss": 1.2636, + "step": 5067 + }, + { + "epoch": 0.9024216524216524, + "grad_norm": 0.4943496286869049, + "learning_rate": 0.0001760578660694718, + "loss": 0.8951, + "step": 5068 + }, + { + "epoch": 0.9025997150997151, + "grad_norm": 0.5475931167602539, + "learning_rate": 0.00017604877752447267, + "loss": 1.1442, + "step": 5069 + }, + { + "epoch": 0.9027777777777778, + "grad_norm": 0.5280239582061768, + "learning_rate": 0.0001760396874894557, + "loss": 0.9537, + "step": 5070 + }, + { + "epoch": 0.9029558404558404, + "grad_norm": 0.5480797290802002, + "learning_rate": 0.000176030595964599, + "loss": 1.1557, + "step": 5071 + }, + { + "epoch": 0.9031339031339032, + "grad_norm": 0.5232734680175781, + "learning_rate": 0.00017602150295008073, + "loss": 1.0219, + "step": 5072 + }, + { + "epoch": 0.9033119658119658, + "grad_norm": 0.5448359251022339, + "learning_rate": 0.000176012408446079, + "loss": 1.1964, + "step": 5073 + }, + { + "epoch": 0.9034900284900285, + "grad_norm": 0.4841914474964142, + "learning_rate": 0.00017600331245277206, + "loss": 1.0667, + "step": 5074 + }, + { + "epoch": 0.9036680911680912, + "grad_norm": 0.5407083630561829, + "learning_rate": 0.0001759942149703381, + "loss": 1.1895, + "step": 5075 + }, + { + "epoch": 0.9038461538461539, + "grad_norm": 0.5140416026115417, + "learning_rate": 0.00017598511599895534, + "loss": 0.9402, + "step": 5076 + }, + { + "epoch": 0.9040242165242165, + "grad_norm": 0.6333765983581543, + "learning_rate": 0.00017597601553880207, + "loss": 1.239, + "step": 5077 + }, + { + "epoch": 0.9042022792022792, + "grad_norm": 0.4996028244495392, + "learning_rate": 0.00017596691359005664, + "loss": 1.0259, + "step": 5078 + }, + { + "epoch": 0.9043803418803419, + "grad_norm": 0.591892421245575, + "learning_rate": 0.00017595781015289732, + "loss": 1.2148, + "step": 5079 + }, + { + "epoch": 0.9045584045584045, + "grad_norm": 0.736499011516571, + "learning_rate": 0.0001759487052275025, + "loss": 1.1373, + "step": 5080 + }, + { + "epoch": 0.9047364672364673, + "grad_norm": 0.5951572060585022, + "learning_rate": 0.00017593959881405057, + "loss": 1.1833, + "step": 5081 + }, + { + "epoch": 0.9049145299145299, + "grad_norm": 0.5092006325721741, + "learning_rate": 0.00017593049091271996, + "loss": 0.8841, + "step": 5082 + }, + { + "epoch": 0.9050925925925926, + "grad_norm": 0.5679013729095459, + "learning_rate": 0.0001759213815236891, + "loss": 1.1056, + "step": 5083 + }, + { + "epoch": 0.9052706552706553, + "grad_norm": 0.5708174109458923, + "learning_rate": 0.0001759122706471365, + "loss": 1.1952, + "step": 5084 + }, + { + "epoch": 0.905448717948718, + "grad_norm": 0.5726733803749084, + "learning_rate": 0.00017590315828324067, + "loss": 1.1013, + "step": 5085 + }, + { + "epoch": 0.9056267806267806, + "grad_norm": 0.5821273326873779, + "learning_rate": 0.00017589404443218008, + "loss": 1.2323, + "step": 5086 + }, + { + "epoch": 0.9058048433048433, + "grad_norm": 0.5811445713043213, + "learning_rate": 0.00017588492909413337, + "loss": 1.2241, + "step": 5087 + }, + { + "epoch": 0.905982905982906, + "grad_norm": 0.5377545952796936, + "learning_rate": 0.0001758758122692791, + "loss": 0.9777, + "step": 5088 + }, + { + "epoch": 0.9061609686609686, + "grad_norm": 0.5985640287399292, + "learning_rate": 0.0001758666939577959, + "loss": 0.9737, + "step": 5089 + }, + { + "epoch": 0.9063390313390314, + "grad_norm": 0.6038222908973694, + "learning_rate": 0.00017585757415986247, + "loss": 1.2116, + "step": 5090 + }, + { + "epoch": 0.906517094017094, + "grad_norm": 0.6752246022224426, + "learning_rate": 0.00017584845287565743, + "loss": 1.1975, + "step": 5091 + }, + { + "epoch": 0.9066951566951567, + "grad_norm": 0.5400625467300415, + "learning_rate": 0.0001758393301053595, + "loss": 0.9669, + "step": 5092 + }, + { + "epoch": 0.9068732193732194, + "grad_norm": 0.5637784004211426, + "learning_rate": 0.00017583020584914746, + "loss": 1.2672, + "step": 5093 + }, + { + "epoch": 0.907051282051282, + "grad_norm": 0.4825877249240875, + "learning_rate": 0.00017582108010720006, + "loss": 0.9719, + "step": 5094 + }, + { + "epoch": 0.9072293447293447, + "grad_norm": 0.49902790784835815, + "learning_rate": 0.00017581195287969613, + "loss": 0.7941, + "step": 5095 + }, + { + "epoch": 0.9074074074074074, + "grad_norm": 0.5991541743278503, + "learning_rate": 0.0001758028241668144, + "loss": 1.049, + "step": 5096 + }, + { + "epoch": 0.9075854700854701, + "grad_norm": 0.5788859724998474, + "learning_rate": 0.00017579369396873384, + "loss": 1.0318, + "step": 5097 + }, + { + "epoch": 0.9077635327635327, + "grad_norm": 0.5914160013198853, + "learning_rate": 0.0001757845622856333, + "loss": 1.1007, + "step": 5098 + }, + { + "epoch": 0.9079415954415955, + "grad_norm": 0.5361711382865906, + "learning_rate": 0.00017577542911769166, + "loss": 1.0694, + "step": 5099 + }, + { + "epoch": 0.9081196581196581, + "grad_norm": 0.5752849578857422, + "learning_rate": 0.00017576629446508792, + "loss": 1.1184, + "step": 5100 + }, + { + "epoch": 0.9082977207977208, + "grad_norm": 0.6042249798774719, + "learning_rate": 0.000175757158328001, + "loss": 1.2808, + "step": 5101 + }, + { + "epoch": 0.9084757834757835, + "grad_norm": 0.508352518081665, + "learning_rate": 0.00017574802070661, + "loss": 1.0038, + "step": 5102 + }, + { + "epoch": 0.9086538461538461, + "grad_norm": 0.5667358040809631, + "learning_rate": 0.00017573888160109385, + "loss": 1.0208, + "step": 5103 + }, + { + "epoch": 0.9088319088319088, + "grad_norm": 0.653619647026062, + "learning_rate": 0.00017572974101163165, + "loss": 1.2053, + "step": 5104 + }, + { + "epoch": 0.9090099715099715, + "grad_norm": 0.5069597363471985, + "learning_rate": 0.00017572059893840246, + "loss": 0.8634, + "step": 5105 + }, + { + "epoch": 0.9091880341880342, + "grad_norm": 0.6160602569580078, + "learning_rate": 0.00017571145538158547, + "loss": 1.2626, + "step": 5106 + }, + { + "epoch": 0.9093660968660968, + "grad_norm": 0.6335833668708801, + "learning_rate": 0.00017570231034135978, + "loss": 1.3381, + "step": 5107 + }, + { + "epoch": 0.9095441595441596, + "grad_norm": 0.5140398740768433, + "learning_rate": 0.00017569316381790454, + "loss": 1.1258, + "step": 5108 + }, + { + "epoch": 0.9097222222222222, + "grad_norm": 0.5682975649833679, + "learning_rate": 0.00017568401581139905, + "loss": 1.3367, + "step": 5109 + }, + { + "epoch": 0.9099002849002849, + "grad_norm": 0.49765729904174805, + "learning_rate": 0.00017567486632202246, + "loss": 1.1891, + "step": 5110 + }, + { + "epoch": 0.9100783475783476, + "grad_norm": 0.5139224529266357, + "learning_rate": 0.00017566571534995406, + "loss": 0.9768, + "step": 5111 + }, + { + "epoch": 0.9102564102564102, + "grad_norm": 0.5510922074317932, + "learning_rate": 0.00017565656289537316, + "loss": 1.1552, + "step": 5112 + }, + { + "epoch": 0.9104344729344729, + "grad_norm": 0.6243364810943604, + "learning_rate": 0.00017564740895845908, + "loss": 1.1341, + "step": 5113 + }, + { + "epoch": 0.9106125356125356, + "grad_norm": 0.5334977507591248, + "learning_rate": 0.00017563825353939116, + "loss": 1.0894, + "step": 5114 + }, + { + "epoch": 0.9107905982905983, + "grad_norm": 0.5195826292037964, + "learning_rate": 0.00017562909663834878, + "loss": 1.1011, + "step": 5115 + }, + { + "epoch": 0.9109686609686609, + "grad_norm": 0.5298168063163757, + "learning_rate": 0.00017561993825551138, + "loss": 1.0079, + "step": 5116 + }, + { + "epoch": 0.9111467236467237, + "grad_norm": 0.5858965516090393, + "learning_rate": 0.00017561077839105835, + "loss": 1.2746, + "step": 5117 + }, + { + "epoch": 0.9113247863247863, + "grad_norm": 0.5572476387023926, + "learning_rate": 0.0001756016170451692, + "loss": 0.8169, + "step": 5118 + }, + { + "epoch": 0.9115028490028491, + "grad_norm": 0.5247095823287964, + "learning_rate": 0.0001755924542180234, + "loss": 1.1206, + "step": 5119 + }, + { + "epoch": 0.9116809116809117, + "grad_norm": 0.5605118274688721, + "learning_rate": 0.0001755832899098005, + "loss": 1.371, + "step": 5120 + }, + { + "epoch": 0.9118589743589743, + "grad_norm": 0.5732316970825195, + "learning_rate": 0.00017557412412068005, + "loss": 1.1248, + "step": 5121 + }, + { + "epoch": 0.9120370370370371, + "grad_norm": 0.6167279481887817, + "learning_rate": 0.0001755649568508416, + "loss": 0.94, + "step": 5122 + }, + { + "epoch": 0.9122150997150997, + "grad_norm": 0.5497499108314514, + "learning_rate": 0.00017555578810046483, + "loss": 1.0112, + "step": 5123 + }, + { + "epoch": 0.9123931623931624, + "grad_norm": 0.540762186050415, + "learning_rate": 0.00017554661786972931, + "loss": 1.1058, + "step": 5124 + }, + { + "epoch": 0.9125712250712251, + "grad_norm": 0.5943556427955627, + "learning_rate": 0.0001755374461588148, + "loss": 0.9086, + "step": 5125 + }, + { + "epoch": 0.9127492877492878, + "grad_norm": 0.5300756692886353, + "learning_rate": 0.0001755282729679009, + "loss": 1.1566, + "step": 5126 + }, + { + "epoch": 0.9129273504273504, + "grad_norm": 0.5390434861183167, + "learning_rate": 0.00017551909829716743, + "loss": 1.1395, + "step": 5127 + }, + { + "epoch": 0.9131054131054132, + "grad_norm": 0.627434492111206, + "learning_rate": 0.00017550992214679405, + "loss": 1.1537, + "step": 5128 + }, + { + "epoch": 0.9132834757834758, + "grad_norm": 0.4806903302669525, + "learning_rate": 0.00017550074451696063, + "loss": 0.7905, + "step": 5129 + }, + { + "epoch": 0.9134615384615384, + "grad_norm": 0.5714817047119141, + "learning_rate": 0.00017549156540784696, + "loss": 1.1042, + "step": 5130 + }, + { + "epoch": 0.9136396011396012, + "grad_norm": 0.5839236378669739, + "learning_rate": 0.0001754823848196329, + "loss": 1.0383, + "step": 5131 + }, + { + "epoch": 0.9138176638176638, + "grad_norm": 0.6089872717857361, + "learning_rate": 0.0001754732027524983, + "loss": 0.9399, + "step": 5132 + }, + { + "epoch": 0.9139957264957265, + "grad_norm": 0.4937956631183624, + "learning_rate": 0.00017546401920662307, + "loss": 0.7382, + "step": 5133 + }, + { + "epoch": 0.9141737891737892, + "grad_norm": 0.5918676257133484, + "learning_rate": 0.00017545483418218716, + "loss": 1.2207, + "step": 5134 + }, + { + "epoch": 0.9143518518518519, + "grad_norm": 0.5825346112251282, + "learning_rate": 0.0001754456476793705, + "loss": 0.9669, + "step": 5135 + }, + { + "epoch": 0.9145299145299145, + "grad_norm": 0.49829617142677307, + "learning_rate": 0.0001754364596983531, + "loss": 1.2247, + "step": 5136 + }, + { + "epoch": 0.9147079772079773, + "grad_norm": 0.5128271579742432, + "learning_rate": 0.00017542727023931497, + "loss": 0.9563, + "step": 5137 + }, + { + "epoch": 0.9148860398860399, + "grad_norm": 0.5789414644241333, + "learning_rate": 0.00017541807930243622, + "loss": 1.22, + "step": 5138 + }, + { + "epoch": 0.9150641025641025, + "grad_norm": 0.44155433773994446, + "learning_rate": 0.00017540888688789683, + "loss": 0.9897, + "step": 5139 + }, + { + "epoch": 0.9152421652421653, + "grad_norm": 0.550464391708374, + "learning_rate": 0.00017539969299587696, + "loss": 1.0624, + "step": 5140 + }, + { + "epoch": 0.9154202279202279, + "grad_norm": 0.5019831657409668, + "learning_rate": 0.0001753904976265567, + "loss": 0.9045, + "step": 5141 + }, + { + "epoch": 0.9155982905982906, + "grad_norm": 0.589658796787262, + "learning_rate": 0.0001753813007801163, + "loss": 1.0454, + "step": 5142 + }, + { + "epoch": 0.9157763532763533, + "grad_norm": 0.5945459008216858, + "learning_rate": 0.00017537210245673586, + "loss": 1.0042, + "step": 5143 + }, + { + "epoch": 0.915954415954416, + "grad_norm": 0.5409809947013855, + "learning_rate": 0.00017536290265659566, + "loss": 1.0609, + "step": 5144 + }, + { + "epoch": 0.9161324786324786, + "grad_norm": 0.5302975177764893, + "learning_rate": 0.00017535370137987597, + "loss": 1.1394, + "step": 5145 + }, + { + "epoch": 0.9163105413105413, + "grad_norm": 0.5253351330757141, + "learning_rate": 0.00017534449862675698, + "loss": 1.2249, + "step": 5146 + }, + { + "epoch": 0.916488603988604, + "grad_norm": 0.6363829970359802, + "learning_rate": 0.00017533529439741908, + "loss": 1.1333, + "step": 5147 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.4703354835510254, + "learning_rate": 0.0001753260886920426, + "loss": 0.9971, + "step": 5148 + }, + { + "epoch": 0.9168447293447294, + "grad_norm": 0.6394907236099243, + "learning_rate": 0.00017531688151080786, + "loss": 1.5942, + "step": 5149 + }, + { + "epoch": 0.917022792022792, + "grad_norm": 0.5573459267616272, + "learning_rate": 0.00017530767285389527, + "loss": 0.9669, + "step": 5150 + }, + { + "epoch": 0.9172008547008547, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00017529846272148532, + "loss": 1.2151, + "step": 5151 + }, + { + "epoch": 0.9173789173789174, + "grad_norm": 0.5550395846366882, + "learning_rate": 0.0001752892511137584, + "loss": 1.1765, + "step": 5152 + }, + { + "epoch": 0.91755698005698, + "grad_norm": 0.5461394786834717, + "learning_rate": 0.00017528003803089496, + "loss": 1.1136, + "step": 5153 + }, + { + "epoch": 0.9177350427350427, + "grad_norm": 0.5512672662734985, + "learning_rate": 0.00017527082347307558, + "loss": 1.1727, + "step": 5154 + }, + { + "epoch": 0.9179131054131054, + "grad_norm": 0.5210778713226318, + "learning_rate": 0.0001752616074404808, + "loss": 1.09, + "step": 5155 + }, + { + "epoch": 0.9180911680911681, + "grad_norm": 0.5214943289756775, + "learning_rate": 0.00017525238993329115, + "loss": 0.9654, + "step": 5156 + }, + { + "epoch": 0.9182692307692307, + "grad_norm": 0.5822862386703491, + "learning_rate": 0.00017524317095168724, + "loss": 1.0951, + "step": 5157 + }, + { + "epoch": 0.9184472934472935, + "grad_norm": 0.43948012590408325, + "learning_rate": 0.0001752339504958497, + "loss": 0.6984, + "step": 5158 + }, + { + "epoch": 0.9186253561253561, + "grad_norm": 0.5024449229240417, + "learning_rate": 0.00017522472856595916, + "loss": 0.983, + "step": 5159 + }, + { + "epoch": 0.9188034188034188, + "grad_norm": 0.5815144181251526, + "learning_rate": 0.00017521550516219636, + "loss": 0.9784, + "step": 5160 + }, + { + "epoch": 0.9189814814814815, + "grad_norm": 0.5519825220108032, + "learning_rate": 0.00017520628028474197, + "loss": 1.064, + "step": 5161 + }, + { + "epoch": 0.9191595441595442, + "grad_norm": 0.5615749955177307, + "learning_rate": 0.00017519705393377675, + "loss": 1.1284, + "step": 5162 + }, + { + "epoch": 0.9193376068376068, + "grad_norm": 0.5929917693138123, + "learning_rate": 0.00017518782610948148, + "loss": 1.1221, + "step": 5163 + }, + { + "epoch": 0.9195156695156695, + "grad_norm": 0.7116361856460571, + "learning_rate": 0.00017517859681203692, + "loss": 1.0188, + "step": 5164 + }, + { + "epoch": 0.9196937321937322, + "grad_norm": 0.5095893740653992, + "learning_rate": 0.00017516936604162396, + "loss": 1.0724, + "step": 5165 + }, + { + "epoch": 0.9198717948717948, + "grad_norm": 0.5701385736465454, + "learning_rate": 0.00017516013379842337, + "loss": 1.0572, + "step": 5166 + }, + { + "epoch": 0.9200498575498576, + "grad_norm": 0.518412709236145, + "learning_rate": 0.00017515090008261613, + "loss": 1.0514, + "step": 5167 + }, + { + "epoch": 0.9202279202279202, + "grad_norm": 0.5324261784553528, + "learning_rate": 0.00017514166489438312, + "loss": 1.1708, + "step": 5168 + }, + { + "epoch": 0.9204059829059829, + "grad_norm": 0.5640990138053894, + "learning_rate": 0.00017513242823390525, + "loss": 1.2846, + "step": 5169 + }, + { + "epoch": 0.9205840455840456, + "grad_norm": 0.510352373123169, + "learning_rate": 0.00017512319010136356, + "loss": 1.0763, + "step": 5170 + }, + { + "epoch": 0.9207621082621082, + "grad_norm": 0.4994175136089325, + "learning_rate": 0.00017511395049693898, + "loss": 0.9665, + "step": 5171 + }, + { + "epoch": 0.9209401709401709, + "grad_norm": 0.43196994066238403, + "learning_rate": 0.00017510470942081258, + "loss": 0.761, + "step": 5172 + }, + { + "epoch": 0.9211182336182336, + "grad_norm": 0.558977484703064, + "learning_rate": 0.00017509546687316543, + "loss": 1.0758, + "step": 5173 + }, + { + "epoch": 0.9212962962962963, + "grad_norm": 0.573302149772644, + "learning_rate": 0.0001750862228541786, + "loss": 0.9635, + "step": 5174 + }, + { + "epoch": 0.9214743589743589, + "grad_norm": 0.5083786845207214, + "learning_rate": 0.00017507697736403321, + "loss": 1.0311, + "step": 5175 + }, + { + "epoch": 0.9216524216524217, + "grad_norm": 0.5478954911231995, + "learning_rate": 0.00017506773040291043, + "loss": 1.074, + "step": 5176 + }, + { + "epoch": 0.9218304843304843, + "grad_norm": 0.522376537322998, + "learning_rate": 0.00017505848197099137, + "loss": 1.1162, + "step": 5177 + }, + { + "epoch": 0.9220085470085471, + "grad_norm": 0.5946292281150818, + "learning_rate": 0.0001750492320684573, + "loss": 0.9494, + "step": 5178 + }, + { + "epoch": 0.9221866096866097, + "grad_norm": 0.5423247814178467, + "learning_rate": 0.00017503998069548943, + "loss": 1.0558, + "step": 5179 + }, + { + "epoch": 0.9223646723646723, + "grad_norm": 0.49960651993751526, + "learning_rate": 0.000175030727852269, + "loss": 1.0748, + "step": 5180 + }, + { + "epoch": 0.9225427350427351, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00017502147353897732, + "loss": 1.2066, + "step": 5181 + }, + { + "epoch": 0.9227207977207977, + "grad_norm": 0.57244473695755, + "learning_rate": 0.00017501221775579576, + "loss": 1.048, + "step": 5182 + }, + { + "epoch": 0.9228988603988604, + "grad_norm": 0.512464165687561, + "learning_rate": 0.00017500296050290557, + "loss": 1.1405, + "step": 5183 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.5380734801292419, + "learning_rate": 0.00017499370178048818, + "loss": 1.0641, + "step": 5184 + }, + { + "epoch": 0.9232549857549858, + "grad_norm": 0.47102874517440796, + "learning_rate": 0.000174984441588725, + "loss": 0.7948, + "step": 5185 + }, + { + "epoch": 0.9234330484330484, + "grad_norm": 0.6702211499214172, + "learning_rate": 0.00017497517992779747, + "loss": 1.3009, + "step": 5186 + }, + { + "epoch": 0.9236111111111112, + "grad_norm": 0.4685834050178528, + "learning_rate": 0.000174965916797887, + "loss": 0.8136, + "step": 5187 + }, + { + "epoch": 0.9237891737891738, + "grad_norm": 0.5414277911186218, + "learning_rate": 0.00017495665219917513, + "loss": 0.9708, + "step": 5188 + }, + { + "epoch": 0.9239672364672364, + "grad_norm": 0.5253050923347473, + "learning_rate": 0.0001749473861318434, + "loss": 1.0691, + "step": 5189 + }, + { + "epoch": 0.9241452991452992, + "grad_norm": 0.6009906530380249, + "learning_rate": 0.00017493811859607328, + "loss": 1.2023, + "step": 5190 + }, + { + "epoch": 0.9243233618233618, + "grad_norm": 0.5519336462020874, + "learning_rate": 0.00017492884959204643, + "loss": 1.189, + "step": 5191 + }, + { + "epoch": 0.9245014245014245, + "grad_norm": 0.5024857521057129, + "learning_rate": 0.0001749195791199444, + "loss": 0.8685, + "step": 5192 + }, + { + "epoch": 0.9246794871794872, + "grad_norm": 0.5735679864883423, + "learning_rate": 0.00017491030717994887, + "loss": 1.1903, + "step": 5193 + }, + { + "epoch": 0.9248575498575499, + "grad_norm": 0.5338658094406128, + "learning_rate": 0.00017490103377224147, + "loss": 1.0442, + "step": 5194 + }, + { + "epoch": 0.9250356125356125, + "grad_norm": 0.46669119596481323, + "learning_rate": 0.0001748917588970039, + "loss": 0.6343, + "step": 5195 + }, + { + "epoch": 0.9252136752136753, + "grad_norm": 0.510910153388977, + "learning_rate": 0.00017488248255441793, + "loss": 0.9334, + "step": 5196 + }, + { + "epoch": 0.9253917378917379, + "grad_norm": 0.5732216238975525, + "learning_rate": 0.00017487320474466524, + "loss": 1.0483, + "step": 5197 + }, + { + "epoch": 0.9255698005698005, + "grad_norm": 0.5864318609237671, + "learning_rate": 0.00017486392546792762, + "loss": 1.0669, + "step": 5198 + }, + { + "epoch": 0.9257478632478633, + "grad_norm": 0.5074281096458435, + "learning_rate": 0.00017485464472438692, + "loss": 1.0636, + "step": 5199 + }, + { + "epoch": 0.9259259259259259, + "grad_norm": 0.5833215117454529, + "learning_rate": 0.00017484536251422496, + "loss": 1.2005, + "step": 5200 + }, + { + "epoch": 0.9261039886039886, + "grad_norm": 0.5624990463256836, + "learning_rate": 0.0001748360788376236, + "loss": 1.1623, + "step": 5201 + }, + { + "epoch": 0.9262820512820513, + "grad_norm": 0.5618230104446411, + "learning_rate": 0.00017482679369476472, + "loss": 1.0495, + "step": 5202 + }, + { + "epoch": 0.926460113960114, + "grad_norm": 0.6254985332489014, + "learning_rate": 0.00017481750708583024, + "loss": 0.9521, + "step": 5203 + }, + { + "epoch": 0.9266381766381766, + "grad_norm": 0.5488203763961792, + "learning_rate": 0.00017480821901100216, + "loss": 1.0689, + "step": 5204 + }, + { + "epoch": 0.9268162393162394, + "grad_norm": 0.6157993674278259, + "learning_rate": 0.00017479892947046245, + "loss": 1.2852, + "step": 5205 + }, + { + "epoch": 0.926994301994302, + "grad_norm": 0.49653390049934387, + "learning_rate": 0.00017478963846439305, + "loss": 0.8616, + "step": 5206 + }, + { + "epoch": 0.9271723646723646, + "grad_norm": 0.5079081058502197, + "learning_rate": 0.00017478034599297603, + "loss": 1.0192, + "step": 5207 + }, + { + "epoch": 0.9273504273504274, + "grad_norm": 0.5392495393753052, + "learning_rate": 0.00017477105205639354, + "loss": 1.115, + "step": 5208 + }, + { + "epoch": 0.92752849002849, + "grad_norm": 0.5336191654205322, + "learning_rate": 0.00017476175665482756, + "loss": 1.1892, + "step": 5209 + }, + { + "epoch": 0.9277065527065527, + "grad_norm": 0.631712019443512, + "learning_rate": 0.00017475245978846026, + "loss": 0.9619, + "step": 5210 + }, + { + "epoch": 0.9278846153846154, + "grad_norm": 0.5123951435089111, + "learning_rate": 0.0001747431614574738, + "loss": 1.1477, + "step": 5211 + }, + { + "epoch": 0.9280626780626781, + "grad_norm": 0.5045743584632874, + "learning_rate": 0.00017473386166205038, + "loss": 0.9749, + "step": 5212 + }, + { + "epoch": 0.9282407407407407, + "grad_norm": 0.5296525359153748, + "learning_rate": 0.00017472456040237217, + "loss": 1.0736, + "step": 5213 + }, + { + "epoch": 0.9284188034188035, + "grad_norm": 0.6304933428764343, + "learning_rate": 0.00017471525767862145, + "loss": 1.2444, + "step": 5214 + }, + { + "epoch": 0.9285968660968661, + "grad_norm": 0.4851958155632019, + "learning_rate": 0.00017470595349098044, + "loss": 0.9049, + "step": 5215 + }, + { + "epoch": 0.9287749287749287, + "grad_norm": 0.5730679631233215, + "learning_rate": 0.00017469664783963148, + "loss": 1.0773, + "step": 5216 + }, + { + "epoch": 0.9289529914529915, + "grad_norm": 0.6020415425300598, + "learning_rate": 0.00017468734072475684, + "loss": 1.3247, + "step": 5217 + }, + { + "epoch": 0.9291310541310541, + "grad_norm": 0.47981077432632446, + "learning_rate": 0.00017467803214653893, + "loss": 1.0009, + "step": 5218 + }, + { + "epoch": 0.9293091168091168, + "grad_norm": 0.5787527561187744, + "learning_rate": 0.0001746687221051601, + "loss": 1.2523, + "step": 5219 + }, + { + "epoch": 0.9294871794871795, + "grad_norm": 0.4495891332626343, + "learning_rate": 0.00017465941060080278, + "loss": 0.7364, + "step": 5220 + }, + { + "epoch": 0.9296652421652422, + "grad_norm": 0.5721768140792847, + "learning_rate": 0.0001746500976336494, + "loss": 1.015, + "step": 5221 + }, + { + "epoch": 0.9298433048433048, + "grad_norm": 0.5500208735466003, + "learning_rate": 0.0001746407832038824, + "loss": 1.053, + "step": 5222 + }, + { + "epoch": 0.9300213675213675, + "grad_norm": 0.5784386992454529, + "learning_rate": 0.00017463146731168437, + "loss": 0.9784, + "step": 5223 + }, + { + "epoch": 0.9301994301994302, + "grad_norm": 0.4960322082042694, + "learning_rate": 0.00017462214995723772, + "loss": 0.8674, + "step": 5224 + }, + { + "epoch": 0.9303774928774928, + "grad_norm": 0.5005537271499634, + "learning_rate": 0.00017461283114072508, + "loss": 1.0486, + "step": 5225 + }, + { + "epoch": 0.9305555555555556, + "grad_norm": 0.5064167380332947, + "learning_rate": 0.000174603510862329, + "loss": 0.9722, + "step": 5226 + }, + { + "epoch": 0.9307336182336182, + "grad_norm": 0.583558976650238, + "learning_rate": 0.0001745941891222321, + "loss": 0.9957, + "step": 5227 + }, + { + "epoch": 0.9309116809116809, + "grad_norm": 0.4982515871524811, + "learning_rate": 0.00017458486592061704, + "loss": 0.958, + "step": 5228 + }, + { + "epoch": 0.9310897435897436, + "grad_norm": 0.526549756526947, + "learning_rate": 0.0001745755412576664, + "loss": 1.1172, + "step": 5229 + }, + { + "epoch": 0.9312678062678063, + "grad_norm": 0.6129719018936157, + "learning_rate": 0.000174566215133563, + "loss": 1.2524, + "step": 5230 + }, + { + "epoch": 0.9314458689458689, + "grad_norm": 0.5385653972625732, + "learning_rate": 0.00017455688754848948, + "loss": 1.1655, + "step": 5231 + }, + { + "epoch": 0.9316239316239316, + "grad_norm": 0.5646410584449768, + "learning_rate": 0.0001745475585026287, + "loss": 0.9026, + "step": 5232 + }, + { + "epoch": 0.9318019943019943, + "grad_norm": 0.549223780632019, + "learning_rate": 0.0001745382279961633, + "loss": 0.804, + "step": 5233 + }, + { + "epoch": 0.9319800569800569, + "grad_norm": 0.48547953367233276, + "learning_rate": 0.0001745288960292762, + "loss": 1.0224, + "step": 5234 + }, + { + "epoch": 0.9321581196581197, + "grad_norm": 0.5260967016220093, + "learning_rate": 0.00017451956260215016, + "loss": 0.9688, + "step": 5235 + }, + { + "epoch": 0.9323361823361823, + "grad_norm": 0.6261999011039734, + "learning_rate": 0.00017451022771496812, + "loss": 1.2539, + "step": 5236 + }, + { + "epoch": 0.9325142450142451, + "grad_norm": 0.5801421999931335, + "learning_rate": 0.00017450089136791298, + "loss": 1.11, + "step": 5237 + }, + { + "epoch": 0.9326923076923077, + "grad_norm": 0.5833573937416077, + "learning_rate": 0.0001744915535611676, + "loss": 0.9328, + "step": 5238 + }, + { + "epoch": 0.9328703703703703, + "grad_norm": 0.5422634482383728, + "learning_rate": 0.00017448221429491496, + "loss": 1.034, + "step": 5239 + }, + { + "epoch": 0.9330484330484331, + "grad_norm": 0.5105658769607544, + "learning_rate": 0.00017447287356933808, + "loss": 0.8924, + "step": 5240 + }, + { + "epoch": 0.9332264957264957, + "grad_norm": 0.5114831924438477, + "learning_rate": 0.00017446353138461995, + "loss": 0.9328, + "step": 5241 + }, + { + "epoch": 0.9334045584045584, + "grad_norm": 0.5105039477348328, + "learning_rate": 0.00017445418774094358, + "loss": 1.0468, + "step": 5242 + }, + { + "epoch": 0.9335826210826211, + "grad_norm": 0.593250036239624, + "learning_rate": 0.00017444484263849208, + "loss": 1.0603, + "step": 5243 + }, + { + "epoch": 0.9337606837606838, + "grad_norm": 0.600788414478302, + "learning_rate": 0.00017443549607744853, + "loss": 1.1506, + "step": 5244 + }, + { + "epoch": 0.9339387464387464, + "grad_norm": 0.5394418239593506, + "learning_rate": 0.00017442614805799605, + "loss": 1.038, + "step": 5245 + }, + { + "epoch": 0.9341168091168092, + "grad_norm": 0.5446375608444214, + "learning_rate": 0.00017441679858031786, + "loss": 1.079, + "step": 5246 + }, + { + "epoch": 0.9342948717948718, + "grad_norm": 0.5859794616699219, + "learning_rate": 0.00017440744764459702, + "loss": 1.1453, + "step": 5247 + }, + { + "epoch": 0.9344729344729344, + "grad_norm": 0.4899081289768219, + "learning_rate": 0.00017439809525101688, + "loss": 1.163, + "step": 5248 + }, + { + "epoch": 0.9346509971509972, + "grad_norm": 0.652846097946167, + "learning_rate": 0.00017438874139976055, + "loss": 1.1819, + "step": 5249 + }, + { + "epoch": 0.9348290598290598, + "grad_norm": 0.5402514934539795, + "learning_rate": 0.00017437938609101138, + "loss": 1.0159, + "step": 5250 + }, + { + "epoch": 0.9350071225071225, + "grad_norm": 0.565864086151123, + "learning_rate": 0.00017437002932495265, + "loss": 1.1121, + "step": 5251 + }, + { + "epoch": 0.9351851851851852, + "grad_norm": 0.611786425113678, + "learning_rate": 0.0001743606711017677, + "loss": 1.2511, + "step": 5252 + }, + { + "epoch": 0.9353632478632479, + "grad_norm": 0.5706882476806641, + "learning_rate": 0.00017435131142163988, + "loss": 1.128, + "step": 5253 + }, + { + "epoch": 0.9355413105413105, + "grad_norm": 0.5369367003440857, + "learning_rate": 0.00017434195028475253, + "loss": 1.0562, + "step": 5254 + }, + { + "epoch": 0.9357193732193733, + "grad_norm": 0.49957552552223206, + "learning_rate": 0.0001743325876912891, + "loss": 1.0568, + "step": 5255 + }, + { + "epoch": 0.9358974358974359, + "grad_norm": 0.5398106575012207, + "learning_rate": 0.00017432322364143305, + "loss": 1.1502, + "step": 5256 + }, + { + "epoch": 0.9360754985754985, + "grad_norm": 0.6522027254104614, + "learning_rate": 0.00017431385813536783, + "loss": 1.0591, + "step": 5257 + }, + { + "epoch": 0.9362535612535613, + "grad_norm": 0.5872012972831726, + "learning_rate": 0.00017430449117327693, + "loss": 1.3737, + "step": 5258 + }, + { + "epoch": 0.9364316239316239, + "grad_norm": 0.5124474167823792, + "learning_rate": 0.00017429512275534382, + "loss": 1.0727, + "step": 5259 + }, + { + "epoch": 0.9366096866096866, + "grad_norm": 0.5103365778923035, + "learning_rate": 0.00017428575288175218, + "loss": 1.0339, + "step": 5260 + }, + { + "epoch": 0.9367877492877493, + "grad_norm": 0.585483729839325, + "learning_rate": 0.0001742763815526855, + "loss": 1.1844, + "step": 5261 + }, + { + "epoch": 0.936965811965812, + "grad_norm": 0.5855562090873718, + "learning_rate": 0.00017426700876832746, + "loss": 1.3234, + "step": 5262 + }, + { + "epoch": 0.9371438746438746, + "grad_norm": 0.5774588584899902, + "learning_rate": 0.00017425763452886162, + "loss": 1.0937, + "step": 5263 + }, + { + "epoch": 0.9373219373219374, + "grad_norm": 0.5718343257904053, + "learning_rate": 0.00017424825883447168, + "loss": 1.0783, + "step": 5264 + }, + { + "epoch": 0.9375, + "grad_norm": 0.5414558053016663, + "learning_rate": 0.00017423888168534136, + "loss": 1.1244, + "step": 5265 + }, + { + "epoch": 0.9376780626780626, + "grad_norm": 0.5818275809288025, + "learning_rate": 0.00017422950308165438, + "loss": 1.247, + "step": 5266 + }, + { + "epoch": 0.9378561253561254, + "grad_norm": 0.586398184299469, + "learning_rate": 0.00017422012302359448, + "loss": 1.0515, + "step": 5267 + }, + { + "epoch": 0.938034188034188, + "grad_norm": 0.5236606001853943, + "learning_rate": 0.00017421074151134544, + "loss": 1.1907, + "step": 5268 + }, + { + "epoch": 0.9382122507122507, + "grad_norm": 0.5108010172843933, + "learning_rate": 0.0001742013585450911, + "loss": 1.1125, + "step": 5269 + }, + { + "epoch": 0.9383903133903134, + "grad_norm": 0.4956454038619995, + "learning_rate": 0.00017419197412501527, + "loss": 1.0305, + "step": 5270 + }, + { + "epoch": 0.9385683760683761, + "grad_norm": 0.5432302951812744, + "learning_rate": 0.0001741825882513018, + "loss": 1.1946, + "step": 5271 + }, + { + "epoch": 0.9387464387464387, + "grad_norm": 0.5119295716285706, + "learning_rate": 0.00017417320092413463, + "loss": 0.875, + "step": 5272 + }, + { + "epoch": 0.9389245014245015, + "grad_norm": 0.49740248918533325, + "learning_rate": 0.0001741638121436977, + "loss": 1.1093, + "step": 5273 + }, + { + "epoch": 0.9391025641025641, + "grad_norm": 0.5069027543067932, + "learning_rate": 0.00017415442191017491, + "loss": 1.2498, + "step": 5274 + }, + { + "epoch": 0.9392806267806267, + "grad_norm": 0.570264995098114, + "learning_rate": 0.00017414503022375027, + "loss": 1.0192, + "step": 5275 + }, + { + "epoch": 0.9394586894586895, + "grad_norm": 0.48129352927207947, + "learning_rate": 0.00017413563708460776, + "loss": 0.8467, + "step": 5276 + }, + { + "epoch": 0.9396367521367521, + "grad_norm": 0.5214534401893616, + "learning_rate": 0.00017412624249293148, + "loss": 0.9723, + "step": 5277 + }, + { + "epoch": 0.9398148148148148, + "grad_norm": 0.5150161385536194, + "learning_rate": 0.00017411684644890544, + "loss": 1.0906, + "step": 5278 + }, + { + "epoch": 0.9399928774928775, + "grad_norm": 0.5695852637290955, + "learning_rate": 0.00017410744895271377, + "loss": 1.2891, + "step": 5279 + }, + { + "epoch": 0.9401709401709402, + "grad_norm": 0.5613594651222229, + "learning_rate": 0.00017409805000454055, + "loss": 1.1373, + "step": 5280 + }, + { + "epoch": 0.9403490028490028, + "grad_norm": 0.5134239196777344, + "learning_rate": 0.00017408864960457004, + "loss": 1.1081, + "step": 5281 + }, + { + "epoch": 0.9405270655270656, + "grad_norm": 0.5256397724151611, + "learning_rate": 0.00017407924775298628, + "loss": 1.058, + "step": 5282 + }, + { + "epoch": 0.9407051282051282, + "grad_norm": 0.5145402550697327, + "learning_rate": 0.00017406984444997357, + "loss": 1.0667, + "step": 5283 + }, + { + "epoch": 0.9408831908831908, + "grad_norm": 0.5435704588890076, + "learning_rate": 0.0001740604396957161, + "loss": 1.2275, + "step": 5284 + }, + { + "epoch": 0.9410612535612536, + "grad_norm": 0.5798762440681458, + "learning_rate": 0.0001740510334903982, + "loss": 1.2061, + "step": 5285 + }, + { + "epoch": 0.9412393162393162, + "grad_norm": 0.5461057424545288, + "learning_rate": 0.00017404162583420414, + "loss": 1.1585, + "step": 5286 + }, + { + "epoch": 0.9414173789173789, + "grad_norm": 0.5090487003326416, + "learning_rate": 0.00017403221672731818, + "loss": 1.2496, + "step": 5287 + }, + { + "epoch": 0.9415954415954416, + "grad_norm": 0.5171035528182983, + "learning_rate": 0.00017402280616992476, + "loss": 1.1947, + "step": 5288 + }, + { + "epoch": 0.9417735042735043, + "grad_norm": 0.5292364358901978, + "learning_rate": 0.00017401339416220818, + "loss": 1.0182, + "step": 5289 + }, + { + "epoch": 0.9419515669515669, + "grad_norm": 0.5011499524116516, + "learning_rate": 0.00017400398070435293, + "loss": 1.3363, + "step": 5290 + }, + { + "epoch": 0.9421296296296297, + "grad_norm": 0.4821554720401764, + "learning_rate": 0.0001739945657965434, + "loss": 0.9077, + "step": 5291 + }, + { + "epoch": 0.9423076923076923, + "grad_norm": 0.5849515199661255, + "learning_rate": 0.00017398514943896403, + "loss": 1.1582, + "step": 5292 + }, + { + "epoch": 0.9424857549857549, + "grad_norm": 0.49826139211654663, + "learning_rate": 0.00017397573163179937, + "loss": 1.1025, + "step": 5293 + }, + { + "epoch": 0.9426638176638177, + "grad_norm": 0.6031842827796936, + "learning_rate": 0.00017396631237523392, + "loss": 1.1932, + "step": 5294 + }, + { + "epoch": 0.9428418803418803, + "grad_norm": 0.6013330221176147, + "learning_rate": 0.00017395689166945224, + "loss": 1.2078, + "step": 5295 + }, + { + "epoch": 0.9430199430199431, + "grad_norm": 0.5147021412849426, + "learning_rate": 0.00017394746951463893, + "loss": 0.9988, + "step": 5296 + }, + { + "epoch": 0.9431980056980057, + "grad_norm": 0.5721762776374817, + "learning_rate": 0.0001739380459109785, + "loss": 1.1442, + "step": 5297 + }, + { + "epoch": 0.9433760683760684, + "grad_norm": 0.49272531270980835, + "learning_rate": 0.0001739286208586557, + "loss": 1.0481, + "step": 5298 + }, + { + "epoch": 0.9435541310541311, + "grad_norm": 0.6545688509941101, + "learning_rate": 0.00017391919435785514, + "loss": 1.1393, + "step": 5299 + }, + { + "epoch": 0.9437321937321937, + "grad_norm": 0.617756724357605, + "learning_rate": 0.00017390976640876152, + "loss": 1.1108, + "step": 5300 + }, + { + "epoch": 0.9439102564102564, + "grad_norm": 0.4870470464229584, + "learning_rate": 0.00017390033701155955, + "loss": 0.9028, + "step": 5301 + }, + { + "epoch": 0.9440883190883191, + "grad_norm": 0.5250138640403748, + "learning_rate": 0.000173890906166434, + "loss": 1.0326, + "step": 5302 + }, + { + "epoch": 0.9442663817663818, + "grad_norm": 0.5879467129707336, + "learning_rate": 0.00017388147387356964, + "loss": 1.1569, + "step": 5303 + }, + { + "epoch": 0.9444444444444444, + "grad_norm": 0.4790486991405487, + "learning_rate": 0.00017387204013315127, + "loss": 0.967, + "step": 5304 + }, + { + "epoch": 0.9446225071225072, + "grad_norm": 0.5884372591972351, + "learning_rate": 0.0001738626049453637, + "loss": 1.1342, + "step": 5305 + }, + { + "epoch": 0.9448005698005698, + "grad_norm": 0.4633975028991699, + "learning_rate": 0.00017385316831039187, + "loss": 0.8942, + "step": 5306 + }, + { + "epoch": 0.9449786324786325, + "grad_norm": 0.5301823019981384, + "learning_rate": 0.0001738437302284206, + "loss": 1.1683, + "step": 5307 + }, + { + "epoch": 0.9451566951566952, + "grad_norm": 0.5476770997047424, + "learning_rate": 0.00017383429069963484, + "loss": 1.1574, + "step": 5308 + }, + { + "epoch": 0.9453347578347578, + "grad_norm": 0.47689101099967957, + "learning_rate": 0.00017382484972421953, + "loss": 1.0792, + "step": 5309 + }, + { + "epoch": 0.9455128205128205, + "grad_norm": 0.526063084602356, + "learning_rate": 0.00017381540730235963, + "loss": 0.9012, + "step": 5310 + }, + { + "epoch": 0.9456908831908832, + "grad_norm": 0.5667058229446411, + "learning_rate": 0.0001738059634342402, + "loss": 1.0908, + "step": 5311 + }, + { + "epoch": 0.9458689458689459, + "grad_norm": 0.5402196645736694, + "learning_rate": 0.00017379651812004623, + "loss": 0.943, + "step": 5312 + }, + { + "epoch": 0.9460470085470085, + "grad_norm": 0.5288932919502258, + "learning_rate": 0.00017378707135996276, + "loss": 1.0055, + "step": 5313 + }, + { + "epoch": 0.9462250712250713, + "grad_norm": 0.5607456564903259, + "learning_rate": 0.00017377762315417492, + "loss": 1.2073, + "step": 5314 + }, + { + "epoch": 0.9464031339031339, + "grad_norm": 0.5737698674201965, + "learning_rate": 0.00017376817350286781, + "loss": 1.0001, + "step": 5315 + }, + { + "epoch": 0.9465811965811965, + "grad_norm": 0.6562079787254333, + "learning_rate": 0.00017375872240622657, + "loss": 1.1503, + "step": 5316 + }, + { + "epoch": 0.9467592592592593, + "grad_norm": 0.5407183170318604, + "learning_rate": 0.0001737492698644364, + "loss": 1.1169, + "step": 5317 + }, + { + "epoch": 0.9469373219373219, + "grad_norm": 0.5504152178764343, + "learning_rate": 0.00017373981587768248, + "loss": 1.0468, + "step": 5318 + }, + { + "epoch": 0.9471153846153846, + "grad_norm": 0.4813530743122101, + "learning_rate": 0.00017373036044615006, + "loss": 0.9707, + "step": 5319 + }, + { + "epoch": 0.9472934472934473, + "grad_norm": 0.5810509920120239, + "learning_rate": 0.00017372090357002437, + "loss": 1.4949, + "step": 5320 + }, + { + "epoch": 0.94747150997151, + "grad_norm": 0.5250222086906433, + "learning_rate": 0.00017371144524949074, + "loss": 1.0818, + "step": 5321 + }, + { + "epoch": 0.9476495726495726, + "grad_norm": 0.4852280914783478, + "learning_rate": 0.00017370198548473444, + "loss": 1.1793, + "step": 5322 + }, + { + "epoch": 0.9478276353276354, + "grad_norm": 0.5392420291900635, + "learning_rate": 0.00017369252427594086, + "loss": 1.153, + "step": 5323 + }, + { + "epoch": 0.948005698005698, + "grad_norm": 0.521294116973877, + "learning_rate": 0.00017368306162329533, + "loss": 0.8572, + "step": 5324 + }, + { + "epoch": 0.9481837606837606, + "grad_norm": 0.5579673647880554, + "learning_rate": 0.0001736735975269833, + "loss": 1.0452, + "step": 5325 + }, + { + "epoch": 0.9483618233618234, + "grad_norm": 0.6027318835258484, + "learning_rate": 0.0001736641319871901, + "loss": 1.3475, + "step": 5326 + }, + { + "epoch": 0.948539886039886, + "grad_norm": 0.5600738525390625, + "learning_rate": 0.00017365466500410132, + "loss": 1.0338, + "step": 5327 + }, + { + "epoch": 0.9487179487179487, + "grad_norm": 0.5691532492637634, + "learning_rate": 0.00017364519657790236, + "loss": 1.129, + "step": 5328 + }, + { + "epoch": 0.9488960113960114, + "grad_norm": 0.5161463022232056, + "learning_rate": 0.0001736357267087788, + "loss": 1.0438, + "step": 5329 + }, + { + "epoch": 0.9490740740740741, + "grad_norm": 0.5049656629562378, + "learning_rate": 0.0001736262553969161, + "loss": 0.9484, + "step": 5330 + }, + { + "epoch": 0.9492521367521367, + "grad_norm": 0.5477150678634644, + "learning_rate": 0.00017361678264249988, + "loss": 0.8995, + "step": 5331 + }, + { + "epoch": 0.9494301994301995, + "grad_norm": 0.5679608583450317, + "learning_rate": 0.0001736073084457157, + "loss": 1.241, + "step": 5332 + }, + { + "epoch": 0.9496082621082621, + "grad_norm": 0.5748196840286255, + "learning_rate": 0.00017359783280674926, + "loss": 1.0046, + "step": 5333 + }, + { + "epoch": 0.9497863247863247, + "grad_norm": 0.5677094459533691, + "learning_rate": 0.00017358835572578617, + "loss": 1.2913, + "step": 5334 + }, + { + "epoch": 0.9499643874643875, + "grad_norm": 0.49663659930229187, + "learning_rate": 0.0001735788772030121, + "loss": 1.0388, + "step": 5335 + }, + { + "epoch": 0.9501424501424501, + "grad_norm": 0.5687218904495239, + "learning_rate": 0.0001735693972386128, + "loss": 1.1631, + "step": 5336 + }, + { + "epoch": 0.9503205128205128, + "grad_norm": 0.520708441734314, + "learning_rate": 0.00017355991583277395, + "loss": 1.0744, + "step": 5337 + }, + { + "epoch": 0.9504985754985755, + "grad_norm": 0.5738952159881592, + "learning_rate": 0.00017355043298568137, + "loss": 1.318, + "step": 5338 + }, + { + "epoch": 0.9506766381766382, + "grad_norm": 0.5378455519676208, + "learning_rate": 0.00017354094869752085, + "loss": 0.9827, + "step": 5339 + }, + { + "epoch": 0.9508547008547008, + "grad_norm": 0.5047366619110107, + "learning_rate": 0.0001735314629684782, + "loss": 1.0966, + "step": 5340 + }, + { + "epoch": 0.9510327635327636, + "grad_norm": 0.5526043772697449, + "learning_rate": 0.0001735219757987393, + "loss": 1.059, + "step": 5341 + }, + { + "epoch": 0.9512108262108262, + "grad_norm": 0.5741400718688965, + "learning_rate": 0.00017351248718849003, + "loss": 1.1232, + "step": 5342 + }, + { + "epoch": 0.9513888888888888, + "grad_norm": 0.5421118140220642, + "learning_rate": 0.00017350299713791626, + "loss": 1.0427, + "step": 5343 + }, + { + "epoch": 0.9515669515669516, + "grad_norm": 0.4857081472873688, + "learning_rate": 0.00017349350564720392, + "loss": 0.8663, + "step": 5344 + }, + { + "epoch": 0.9517450142450142, + "grad_norm": 0.5411618947982788, + "learning_rate": 0.00017348401271653904, + "loss": 1.0317, + "step": 5345 + }, + { + "epoch": 0.9519230769230769, + "grad_norm": 0.5246246457099915, + "learning_rate": 0.00017347451834610756, + "loss": 1.0076, + "step": 5346 + }, + { + "epoch": 0.9521011396011396, + "grad_norm": 0.5278927683830261, + "learning_rate": 0.00017346502253609556, + "loss": 0.931, + "step": 5347 + }, + { + "epoch": 0.9522792022792023, + "grad_norm": 0.5934548377990723, + "learning_rate": 0.00017345552528668902, + "loss": 1.3205, + "step": 5348 + }, + { + "epoch": 0.9524572649572649, + "grad_norm": 0.5466100573539734, + "learning_rate": 0.00017344602659807406, + "loss": 0.8725, + "step": 5349 + }, + { + "epoch": 0.9526353276353277, + "grad_norm": 0.5220118761062622, + "learning_rate": 0.00017343652647043678, + "loss": 1.1642, + "step": 5350 + }, + { + "epoch": 0.9528133903133903, + "grad_norm": 0.6166301965713501, + "learning_rate": 0.0001734270249039633, + "loss": 0.8152, + "step": 5351 + }, + { + "epoch": 0.9529914529914529, + "grad_norm": 0.5173428058624268, + "learning_rate": 0.00017341752189883983, + "loss": 0.9296, + "step": 5352 + }, + { + "epoch": 0.9531695156695157, + "grad_norm": 0.5363461375236511, + "learning_rate": 0.0001734080174552525, + "loss": 1.3546, + "step": 5353 + }, + { + "epoch": 0.9533475783475783, + "grad_norm": 0.5333831906318665, + "learning_rate": 0.0001733985115733876, + "loss": 1.0401, + "step": 5354 + }, + { + "epoch": 0.9535256410256411, + "grad_norm": 0.5179334878921509, + "learning_rate": 0.00017338900425343132, + "loss": 1.1254, + "step": 5355 + }, + { + "epoch": 0.9537037037037037, + "grad_norm": 0.5171303153038025, + "learning_rate": 0.00017337949549556993, + "loss": 1.0518, + "step": 5356 + }, + { + "epoch": 0.9538817663817664, + "grad_norm": 0.5164596438407898, + "learning_rate": 0.00017336998529998978, + "loss": 0.8732, + "step": 5357 + }, + { + "epoch": 0.9540598290598291, + "grad_norm": 0.5555717349052429, + "learning_rate": 0.00017336047366687719, + "loss": 1.2312, + "step": 5358 + }, + { + "epoch": 0.9542378917378918, + "grad_norm": 0.45685622096061707, + "learning_rate": 0.00017335096059641847, + "loss": 0.8882, + "step": 5359 + }, + { + "epoch": 0.9544159544159544, + "grad_norm": 0.5260133743286133, + "learning_rate": 0.0001733414460888001, + "loss": 1.0952, + "step": 5360 + }, + { + "epoch": 0.9545940170940171, + "grad_norm": 0.4597703814506531, + "learning_rate": 0.0001733319301442084, + "loss": 1.0835, + "step": 5361 + }, + { + "epoch": 0.9547720797720798, + "grad_norm": 0.5279495120048523, + "learning_rate": 0.0001733224127628299, + "loss": 1.0295, + "step": 5362 + }, + { + "epoch": 0.9549501424501424, + "grad_norm": 0.48919400572776794, + "learning_rate": 0.00017331289394485104, + "loss": 0.9693, + "step": 5363 + }, + { + "epoch": 0.9551282051282052, + "grad_norm": 0.5639515519142151, + "learning_rate": 0.0001733033736904583, + "loss": 1.0893, + "step": 5364 + }, + { + "epoch": 0.9553062678062678, + "grad_norm": 0.49761319160461426, + "learning_rate": 0.00017329385199983823, + "loss": 1.038, + "step": 5365 + }, + { + "epoch": 0.9554843304843305, + "grad_norm": 0.5503305792808533, + "learning_rate": 0.0001732843288731774, + "loss": 0.9976, + "step": 5366 + }, + { + "epoch": 0.9556623931623932, + "grad_norm": 0.5633028745651245, + "learning_rate": 0.00017327480431066235, + "loss": 1.0602, + "step": 5367 + }, + { + "epoch": 0.9558404558404558, + "grad_norm": 0.48074454069137573, + "learning_rate": 0.00017326527831247973, + "loss": 1.0286, + "step": 5368 + }, + { + "epoch": 0.9560185185185185, + "grad_norm": 0.506597638130188, + "learning_rate": 0.0001732557508788162, + "loss": 0.9061, + "step": 5369 + }, + { + "epoch": 0.9561965811965812, + "grad_norm": 0.6570749282836914, + "learning_rate": 0.0001732462220098584, + "loss": 1.0852, + "step": 5370 + }, + { + "epoch": 0.9563746438746439, + "grad_norm": 0.5607653856277466, + "learning_rate": 0.00017323669170579302, + "loss": 1.0486, + "step": 5371 + }, + { + "epoch": 0.9565527065527065, + "grad_norm": 0.6047050356864929, + "learning_rate": 0.0001732271599668068, + "loss": 1.2175, + "step": 5372 + }, + { + "epoch": 0.9567307692307693, + "grad_norm": 0.5506869554519653, + "learning_rate": 0.00017321762679308651, + "loss": 1.0114, + "step": 5373 + }, + { + "epoch": 0.9569088319088319, + "grad_norm": 0.5868638157844543, + "learning_rate": 0.00017320809218481891, + "loss": 1.2983, + "step": 5374 + }, + { + "epoch": 0.9570868945868946, + "grad_norm": 0.539619505405426, + "learning_rate": 0.00017319855614219084, + "loss": 1.2361, + "step": 5375 + }, + { + "epoch": 0.9572649572649573, + "grad_norm": 0.5525495409965515, + "learning_rate": 0.0001731890186653891, + "loss": 1.1316, + "step": 5376 + }, + { + "epoch": 0.95744301994302, + "grad_norm": 0.5549767017364502, + "learning_rate": 0.0001731794797546006, + "loss": 1.0547, + "step": 5377 + }, + { + "epoch": 0.9576210826210826, + "grad_norm": 0.5356076955795288, + "learning_rate": 0.00017316993941001222, + "loss": 0.9942, + "step": 5378 + }, + { + "epoch": 0.9577991452991453, + "grad_norm": 0.5365784168243408, + "learning_rate": 0.00017316039763181084, + "loss": 1.226, + "step": 5379 + }, + { + "epoch": 0.957977207977208, + "grad_norm": 0.5190927386283875, + "learning_rate": 0.00017315085442018343, + "loss": 1.1704, + "step": 5380 + }, + { + "epoch": 0.9581552706552706, + "grad_norm": 0.526658833026886, + "learning_rate": 0.00017314130977531705, + "loss": 1.109, + "step": 5381 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 0.5373684763908386, + "learning_rate": 0.0001731317636973986, + "loss": 1.0018, + "step": 5382 + }, + { + "epoch": 0.958511396011396, + "grad_norm": 0.5714904069900513, + "learning_rate": 0.00017312221618661516, + "loss": 1.1855, + "step": 5383 + }, + { + "epoch": 0.9586894586894587, + "grad_norm": 0.5707863569259644, + "learning_rate": 0.00017311266724315377, + "loss": 0.9482, + "step": 5384 + }, + { + "epoch": 0.9588675213675214, + "grad_norm": 0.5856872797012329, + "learning_rate": 0.00017310311686720157, + "loss": 0.9543, + "step": 5385 + }, + { + "epoch": 0.959045584045584, + "grad_norm": 0.5041963458061218, + "learning_rate": 0.00017309356505894568, + "loss": 1.1427, + "step": 5386 + }, + { + "epoch": 0.9592236467236467, + "grad_norm": 0.5409179925918579, + "learning_rate": 0.00017308401181857316, + "loss": 0.8432, + "step": 5387 + }, + { + "epoch": 0.9594017094017094, + "grad_norm": 0.5248702764511108, + "learning_rate": 0.00017307445714627128, + "loss": 1.1403, + "step": 5388 + }, + { + "epoch": 0.9595797720797721, + "grad_norm": 0.50718092918396, + "learning_rate": 0.00017306490104222722, + "loss": 0.9066, + "step": 5389 + }, + { + "epoch": 0.9597578347578347, + "grad_norm": 0.5563821196556091, + "learning_rate": 0.0001730553435066282, + "loss": 1.0204, + "step": 5390 + }, + { + "epoch": 0.9599358974358975, + "grad_norm": 0.5696987509727478, + "learning_rate": 0.00017304578453966146, + "loss": 1.1405, + "step": 5391 + }, + { + "epoch": 0.9601139601139601, + "grad_norm": 0.5927395224571228, + "learning_rate": 0.00017303622414151435, + "loss": 1.0398, + "step": 5392 + }, + { + "epoch": 0.9602920227920227, + "grad_norm": 0.5375707745552063, + "learning_rate": 0.0001730266623123741, + "loss": 0.9519, + "step": 5393 + }, + { + "epoch": 0.9604700854700855, + "grad_norm": 0.457998126745224, + "learning_rate": 0.00017301709905242815, + "loss": 0.8743, + "step": 5394 + }, + { + "epoch": 0.9606481481481481, + "grad_norm": 0.5427796244621277, + "learning_rate": 0.00017300753436186382, + "loss": 1.078, + "step": 5395 + }, + { + "epoch": 0.9608262108262108, + "grad_norm": 0.5458595752716064, + "learning_rate": 0.0001729979682408685, + "loss": 1.1081, + "step": 5396 + }, + { + "epoch": 0.9610042735042735, + "grad_norm": 0.5495280027389526, + "learning_rate": 0.00017298840068962962, + "loss": 1.0141, + "step": 5397 + }, + { + "epoch": 0.9611823361823362, + "grad_norm": 0.5878560543060303, + "learning_rate": 0.00017297883170833465, + "loss": 1.302, + "step": 5398 + }, + { + "epoch": 0.9613603988603988, + "grad_norm": 0.5452881455421448, + "learning_rate": 0.00017296926129717108, + "loss": 0.9929, + "step": 5399 + }, + { + "epoch": 0.9615384615384616, + "grad_norm": 0.6021811366081238, + "learning_rate": 0.0001729596894563264, + "loss": 1.2629, + "step": 5400 + }, + { + "epoch": 0.9617165242165242, + "grad_norm": 0.5820204615592957, + "learning_rate": 0.0001729501161859882, + "loss": 1.0662, + "step": 5401 + }, + { + "epoch": 0.9618945868945868, + "grad_norm": 0.4953218102455139, + "learning_rate": 0.000172940541486344, + "loss": 1.047, + "step": 5402 + }, + { + "epoch": 0.9620726495726496, + "grad_norm": 0.5409793853759766, + "learning_rate": 0.00017293096535758143, + "loss": 1.1993, + "step": 5403 + }, + { + "epoch": 0.9622507122507122, + "grad_norm": 0.49702873826026917, + "learning_rate": 0.00017292138779988805, + "loss": 1.2471, + "step": 5404 + }, + { + "epoch": 0.9624287749287749, + "grad_norm": 0.5743489861488342, + "learning_rate": 0.00017291180881345158, + "loss": 1.0816, + "step": 5405 + }, + { + "epoch": 0.9626068376068376, + "grad_norm": 0.5747945308685303, + "learning_rate": 0.00017290222839845968, + "loss": 1.3548, + "step": 5406 + }, + { + "epoch": 0.9627849002849003, + "grad_norm": 0.5341345071792603, + "learning_rate": 0.00017289264655510005, + "loss": 1.0435, + "step": 5407 + }, + { + "epoch": 0.9629629629629629, + "grad_norm": 0.5719689130783081, + "learning_rate": 0.00017288306328356044, + "loss": 1.2319, + "step": 5408 + }, + { + "epoch": 0.9631410256410257, + "grad_norm": 0.4783279597759247, + "learning_rate": 0.0001728734785840286, + "loss": 0.9397, + "step": 5409 + }, + { + "epoch": 0.9633190883190883, + "grad_norm": 0.4730507731437683, + "learning_rate": 0.00017286389245669233, + "loss": 0.9384, + "step": 5410 + }, + { + "epoch": 0.9634971509971509, + "grad_norm": 0.5309939384460449, + "learning_rate": 0.00017285430490173944, + "loss": 1.098, + "step": 5411 + }, + { + "epoch": 0.9636752136752137, + "grad_norm": 0.5177853107452393, + "learning_rate": 0.0001728447159193578, + "loss": 1.2777, + "step": 5412 + }, + { + "epoch": 0.9638532763532763, + "grad_norm": 0.6437913775444031, + "learning_rate": 0.00017283512550973526, + "loss": 1.2661, + "step": 5413 + }, + { + "epoch": 0.9640313390313391, + "grad_norm": 0.6096072196960449, + "learning_rate": 0.00017282553367305975, + "loss": 0.9569, + "step": 5414 + }, + { + "epoch": 0.9642094017094017, + "grad_norm": 0.5104934573173523, + "learning_rate": 0.00017281594040951918, + "loss": 0.9666, + "step": 5415 + }, + { + "epoch": 0.9643874643874644, + "grad_norm": 0.6178240776062012, + "learning_rate": 0.00017280634571930153, + "loss": 1.1277, + "step": 5416 + }, + { + "epoch": 0.9645655270655271, + "grad_norm": 0.5749034881591797, + "learning_rate": 0.0001727967496025948, + "loss": 1.245, + "step": 5417 + }, + { + "epoch": 0.9647435897435898, + "grad_norm": 0.5036978721618652, + "learning_rate": 0.00017278715205958694, + "loss": 1.3049, + "step": 5418 + }, + { + "epoch": 0.9649216524216524, + "grad_norm": 0.5593041777610779, + "learning_rate": 0.00017277755309046605, + "loss": 1.2304, + "step": 5419 + }, + { + "epoch": 0.9650997150997151, + "grad_norm": 0.5446555614471436, + "learning_rate": 0.0001727679526954202, + "loss": 0.732, + "step": 5420 + }, + { + "epoch": 0.9652777777777778, + "grad_norm": 0.6063070297241211, + "learning_rate": 0.00017275835087463747, + "loss": 1.3723, + "step": 5421 + }, + { + "epoch": 0.9654558404558404, + "grad_norm": 0.4994211792945862, + "learning_rate": 0.00017274874762830602, + "loss": 1.0505, + "step": 5422 + }, + { + "epoch": 0.9656339031339032, + "grad_norm": 0.49396973848342896, + "learning_rate": 0.00017273914295661395, + "loss": 0.8691, + "step": 5423 + }, + { + "epoch": 0.9658119658119658, + "grad_norm": 0.5067027807235718, + "learning_rate": 0.0001727295368597495, + "loss": 0.9744, + "step": 5424 + }, + { + "epoch": 0.9659900284900285, + "grad_norm": 0.6720643043518066, + "learning_rate": 0.00017271992933790085, + "loss": 1.1513, + "step": 5425 + }, + { + "epoch": 0.9661680911680912, + "grad_norm": 0.5494341254234314, + "learning_rate": 0.00017271032039125624, + "loss": 0.8295, + "step": 5426 + }, + { + "epoch": 0.9663461538461539, + "grad_norm": 0.644332230091095, + "learning_rate": 0.00017270071002000394, + "loss": 1.0043, + "step": 5427 + }, + { + "epoch": 0.9665242165242165, + "grad_norm": 0.5658500790596008, + "learning_rate": 0.00017269109822433225, + "loss": 1.2575, + "step": 5428 + }, + { + "epoch": 0.9667022792022792, + "grad_norm": 0.5163155794143677, + "learning_rate": 0.00017268148500442952, + "loss": 1.1391, + "step": 5429 + }, + { + "epoch": 0.9668803418803419, + "grad_norm": 0.5113703608512878, + "learning_rate": 0.00017267187036048404, + "loss": 1.0819, + "step": 5430 + }, + { + "epoch": 0.9670584045584045, + "grad_norm": 0.6339422464370728, + "learning_rate": 0.00017266225429268426, + "loss": 1.0733, + "step": 5431 + }, + { + "epoch": 0.9672364672364673, + "grad_norm": 0.5158288478851318, + "learning_rate": 0.0001726526368012185, + "loss": 0.9518, + "step": 5432 + }, + { + "epoch": 0.9674145299145299, + "grad_norm": 0.593717634677887, + "learning_rate": 0.00017264301788627527, + "loss": 0.9416, + "step": 5433 + }, + { + "epoch": 0.9675925925925926, + "grad_norm": 0.49593186378479004, + "learning_rate": 0.00017263339754804301, + "loss": 1.0307, + "step": 5434 + }, + { + "epoch": 0.9677706552706553, + "grad_norm": 0.44032949209213257, + "learning_rate": 0.00017262377578671024, + "loss": 0.7884, + "step": 5435 + }, + { + "epoch": 0.967948717948718, + "grad_norm": 0.513073742389679, + "learning_rate": 0.00017261415260246538, + "loss": 0.9797, + "step": 5436 + }, + { + "epoch": 0.9681267806267806, + "grad_norm": 0.5737422108650208, + "learning_rate": 0.0001726045279954971, + "loss": 1.0487, + "step": 5437 + }, + { + "epoch": 0.9683048433048433, + "grad_norm": 0.5385867953300476, + "learning_rate": 0.0001725949019659939, + "loss": 1.4166, + "step": 5438 + }, + { + "epoch": 0.968482905982906, + "grad_norm": 0.5224326848983765, + "learning_rate": 0.00017258527451414438, + "loss": 1.195, + "step": 5439 + }, + { + "epoch": 0.9686609686609686, + "grad_norm": 0.5305148363113403, + "learning_rate": 0.0001725756456401372, + "loss": 1.0301, + "step": 5440 + }, + { + "epoch": 0.9688390313390314, + "grad_norm": 0.532588005065918, + "learning_rate": 0.000172566015344161, + "loss": 1.1269, + "step": 5441 + }, + { + "epoch": 0.969017094017094, + "grad_norm": 0.5812515020370483, + "learning_rate": 0.0001725563836264045, + "loss": 1.1787, + "step": 5442 + }, + { + "epoch": 0.9691951566951567, + "grad_norm": 0.4962109327316284, + "learning_rate": 0.00017254675048705638, + "loss": 1.0639, + "step": 5443 + }, + { + "epoch": 0.9693732193732194, + "grad_norm": 0.5094883441925049, + "learning_rate": 0.00017253711592630534, + "loss": 1.0922, + "step": 5444 + }, + { + "epoch": 0.969551282051282, + "grad_norm": 0.5728049874305725, + "learning_rate": 0.00017252747994434025, + "loss": 1.1237, + "step": 5445 + }, + { + "epoch": 0.9697293447293447, + "grad_norm": 0.5406180620193481, + "learning_rate": 0.00017251784254134983, + "loss": 1.1161, + "step": 5446 + }, + { + "epoch": 0.9699074074074074, + "grad_norm": 0.5724552869796753, + "learning_rate": 0.00017250820371752292, + "loss": 1.2205, + "step": 5447 + }, + { + "epoch": 0.9700854700854701, + "grad_norm": 0.5698846578598022, + "learning_rate": 0.0001724985634730484, + "loss": 1.1472, + "step": 5448 + }, + { + "epoch": 0.9702635327635327, + "grad_norm": 0.5315805673599243, + "learning_rate": 0.0001724889218081151, + "loss": 1.0253, + "step": 5449 + }, + { + "epoch": 0.9704415954415955, + "grad_norm": 0.5970377326011658, + "learning_rate": 0.000172479278722912, + "loss": 1.3033, + "step": 5450 + }, + { + "epoch": 0.9706196581196581, + "grad_norm": 0.6149488687515259, + "learning_rate": 0.00017246963421762798, + "loss": 1.0689, + "step": 5451 + }, + { + "epoch": 0.9707977207977208, + "grad_norm": 0.4848574995994568, + "learning_rate": 0.00017245998829245202, + "loss": 0.8829, + "step": 5452 + }, + { + "epoch": 0.9709757834757835, + "grad_norm": 0.6073294281959534, + "learning_rate": 0.00017245034094757312, + "loss": 1.2378, + "step": 5453 + }, + { + "epoch": 0.9711538461538461, + "grad_norm": 0.6362034678459167, + "learning_rate": 0.00017244069218318026, + "loss": 1.3606, + "step": 5454 + }, + { + "epoch": 0.9713319088319088, + "grad_norm": 0.5353880524635315, + "learning_rate": 0.00017243104199946257, + "loss": 1.1288, + "step": 5455 + }, + { + "epoch": 0.9715099715099715, + "grad_norm": 0.5096352100372314, + "learning_rate": 0.00017242139039660902, + "loss": 1.0056, + "step": 5456 + }, + { + "epoch": 0.9716880341880342, + "grad_norm": 0.5086682438850403, + "learning_rate": 0.00017241173737480884, + "loss": 1.091, + "step": 5457 + }, + { + "epoch": 0.9718660968660968, + "grad_norm": 0.5034295320510864, + "learning_rate": 0.000172402082934251, + "loss": 0.9749, + "step": 5458 + }, + { + "epoch": 0.9720441595441596, + "grad_norm": 0.5205379724502563, + "learning_rate": 0.0001723924270751248, + "loss": 1.1068, + "step": 5459 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 0.5904826521873474, + "learning_rate": 0.00017238276979761937, + "loss": 1.0613, + "step": 5460 + }, + { + "epoch": 0.9724002849002849, + "grad_norm": 0.6415045261383057, + "learning_rate": 0.0001723731111019239, + "loss": 1.2126, + "step": 5461 + }, + { + "epoch": 0.9725783475783476, + "grad_norm": 0.5769147872924805, + "learning_rate": 0.0001723634509882277, + "loss": 1.337, + "step": 5462 + }, + { + "epoch": 0.9727564102564102, + "grad_norm": 0.5585111975669861, + "learning_rate": 0.00017235378945671998, + "loss": 1.3922, + "step": 5463 + }, + { + "epoch": 0.9729344729344729, + "grad_norm": 0.5788411498069763, + "learning_rate": 0.00017234412650759008, + "loss": 0.8532, + "step": 5464 + }, + { + "epoch": 0.9731125356125356, + "grad_norm": 0.5617673397064209, + "learning_rate": 0.00017233446214102728, + "loss": 1.2575, + "step": 5465 + }, + { + "epoch": 0.9732905982905983, + "grad_norm": 0.4227815568447113, + "learning_rate": 0.00017232479635722093, + "loss": 1.0618, + "step": 5466 + }, + { + "epoch": 0.9734686609686609, + "grad_norm": 0.49751797318458557, + "learning_rate": 0.00017231512915636047, + "loss": 0.7714, + "step": 5467 + }, + { + "epoch": 0.9736467236467237, + "grad_norm": 0.5983800292015076, + "learning_rate": 0.0001723054605386353, + "loss": 1.2297, + "step": 5468 + }, + { + "epoch": 0.9738247863247863, + "grad_norm": 0.543394923210144, + "learning_rate": 0.0001722957905042348, + "loss": 1.0078, + "step": 5469 + }, + { + "epoch": 0.9740028490028491, + "grad_norm": 0.5633566975593567, + "learning_rate": 0.00017228611905334846, + "loss": 1.0938, + "step": 5470 + }, + { + "epoch": 0.9741809116809117, + "grad_norm": 0.49377235770225525, + "learning_rate": 0.00017227644618616578, + "loss": 1.096, + "step": 5471 + }, + { + "epoch": 0.9743589743589743, + "grad_norm": 0.4963362216949463, + "learning_rate": 0.00017226677190287627, + "loss": 1.0003, + "step": 5472 + }, + { + "epoch": 0.9745370370370371, + "grad_norm": 0.4483006000518799, + "learning_rate": 0.00017225709620366953, + "loss": 0.8623, + "step": 5473 + }, + { + "epoch": 0.9747150997150997, + "grad_norm": 0.5429352521896362, + "learning_rate": 0.00017224741908873506, + "loss": 1.1383, + "step": 5474 + }, + { + "epoch": 0.9748931623931624, + "grad_norm": 0.5871657729148865, + "learning_rate": 0.0001722377405582625, + "loss": 1.2005, + "step": 5475 + }, + { + "epoch": 0.9750712250712251, + "grad_norm": 0.6002383828163147, + "learning_rate": 0.0001722280606124415, + "loss": 1.0696, + "step": 5476 + }, + { + "epoch": 0.9752492877492878, + "grad_norm": 0.5351617336273193, + "learning_rate": 0.00017221837925146164, + "loss": 1.243, + "step": 5477 + }, + { + "epoch": 0.9754273504273504, + "grad_norm": 0.46613118052482605, + "learning_rate": 0.00017220869647551268, + "loss": 1.0344, + "step": 5478 + }, + { + "epoch": 0.9756054131054132, + "grad_norm": 0.6015593409538269, + "learning_rate": 0.00017219901228478432, + "loss": 1.082, + "step": 5479 + }, + { + "epoch": 0.9757834757834758, + "grad_norm": 0.5829521417617798, + "learning_rate": 0.0001721893266794663, + "loss": 0.8683, + "step": 5480 + }, + { + "epoch": 0.9759615384615384, + "grad_norm": 0.6344960927963257, + "learning_rate": 0.00017217963965974838, + "loss": 1.1048, + "step": 5481 + }, + { + "epoch": 0.9761396011396012, + "grad_norm": 0.5586308240890503, + "learning_rate": 0.00017216995122582034, + "loss": 0.9657, + "step": 5482 + }, + { + "epoch": 0.9763176638176638, + "grad_norm": 0.48625239729881287, + "learning_rate": 0.00017216026137787204, + "loss": 1.1026, + "step": 5483 + }, + { + "epoch": 0.9764957264957265, + "grad_norm": 0.5625223517417908, + "learning_rate": 0.00017215057011609332, + "loss": 1.1579, + "step": 5484 + }, + { + "epoch": 0.9766737891737892, + "grad_norm": 0.6016653776168823, + "learning_rate": 0.0001721408774406741, + "loss": 1.1777, + "step": 5485 + }, + { + "epoch": 0.9768518518518519, + "grad_norm": 0.5444921851158142, + "learning_rate": 0.00017213118335180418, + "loss": 1.119, + "step": 5486 + }, + { + "epoch": 0.9770299145299145, + "grad_norm": 0.5574755668640137, + "learning_rate": 0.0001721214878496736, + "loss": 1.1128, + "step": 5487 + }, + { + "epoch": 0.9772079772079773, + "grad_norm": 0.5486113429069519, + "learning_rate": 0.00017211179093447226, + "loss": 1.1673, + "step": 5488 + }, + { + "epoch": 0.9773860398860399, + "grad_norm": 0.5545483231544495, + "learning_rate": 0.00017210209260639018, + "loss": 1.1748, + "step": 5489 + }, + { + "epoch": 0.9775641025641025, + "grad_norm": 0.5756667256355286, + "learning_rate": 0.0001720923928656174, + "loss": 1.2377, + "step": 5490 + }, + { + "epoch": 0.9777421652421653, + "grad_norm": 0.5744972229003906, + "learning_rate": 0.00017208269171234392, + "loss": 1.1242, + "step": 5491 + }, + { + "epoch": 0.9779202279202279, + "grad_norm": 0.6109468340873718, + "learning_rate": 0.00017207298914675984, + "loss": 1.1948, + "step": 5492 + }, + { + "epoch": 0.9780982905982906, + "grad_norm": 0.5195167660713196, + "learning_rate": 0.00017206328516905525, + "loss": 1.0941, + "step": 5493 + }, + { + "epoch": 0.9782763532763533, + "grad_norm": 0.5549042224884033, + "learning_rate": 0.0001720535797794203, + "loss": 1.1503, + "step": 5494 + }, + { + "epoch": 0.978454415954416, + "grad_norm": 0.6317743062973022, + "learning_rate": 0.0001720438729780451, + "loss": 1.3468, + "step": 5495 + }, + { + "epoch": 0.9786324786324786, + "grad_norm": 0.5932528972625732, + "learning_rate": 0.0001720341647651199, + "loss": 1.105, + "step": 5496 + }, + { + "epoch": 0.9788105413105413, + "grad_norm": 0.607880175113678, + "learning_rate": 0.00017202445514083488, + "loss": 1.1465, + "step": 5497 + }, + { + "epoch": 0.978988603988604, + "grad_norm": 0.49227309226989746, + "learning_rate": 0.00017201474410538027, + "loss": 0.9075, + "step": 5498 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 0.5059443116188049, + "learning_rate": 0.00017200503165894636, + "loss": 1.0483, + "step": 5499 + }, + { + "epoch": 0.9793447293447294, + "grad_norm": 0.5792799592018127, + "learning_rate": 0.0001719953178017234, + "loss": 1.0987, + "step": 5500 + }, + { + "epoch": 0.979522792022792, + "grad_norm": 0.5010457038879395, + "learning_rate": 0.00017198560253390177, + "loss": 1.1051, + "step": 5501 + }, + { + "epoch": 0.9797008547008547, + "grad_norm": 0.5866543054580688, + "learning_rate": 0.0001719758858556718, + "loss": 1.2824, + "step": 5502 + }, + { + "epoch": 0.9798789173789174, + "grad_norm": 0.5392137169837952, + "learning_rate": 0.00017196616776722382, + "loss": 0.886, + "step": 5503 + }, + { + "epoch": 0.98005698005698, + "grad_norm": 0.5200899839401245, + "learning_rate": 0.00017195644826874834, + "loss": 1.1504, + "step": 5504 + }, + { + "epoch": 0.9802350427350427, + "grad_norm": 0.533159077167511, + "learning_rate": 0.00017194672736043569, + "loss": 1.1216, + "step": 5505 + }, + { + "epoch": 0.9804131054131054, + "grad_norm": 0.5543524622917175, + "learning_rate": 0.0001719370050424764, + "loss": 1.0161, + "step": 5506 + }, + { + "epoch": 0.9805911680911681, + "grad_norm": 0.5315365195274353, + "learning_rate": 0.00017192728131506092, + "loss": 1.0509, + "step": 5507 + }, + { + "epoch": 0.9807692307692307, + "grad_norm": 0.5406147837638855, + "learning_rate": 0.00017191755617837977, + "loss": 1.0695, + "step": 5508 + }, + { + "epoch": 0.9809472934472935, + "grad_norm": 0.4563386142253876, + "learning_rate": 0.00017190782963262354, + "loss": 0.995, + "step": 5509 + }, + { + "epoch": 0.9811253561253561, + "grad_norm": 0.5456405282020569, + "learning_rate": 0.00017189810167798274, + "loss": 1.0546, + "step": 5510 + }, + { + "epoch": 0.9813034188034188, + "grad_norm": 0.6275575160980225, + "learning_rate": 0.00017188837231464795, + "loss": 1.0432, + "step": 5511 + }, + { + "epoch": 0.9814814814814815, + "grad_norm": 0.49735602736473083, + "learning_rate": 0.0001718786415428099, + "loss": 1.035, + "step": 5512 + }, + { + "epoch": 0.9816595441595442, + "grad_norm": 0.5234259963035583, + "learning_rate": 0.00017186890936265916, + "loss": 1.0918, + "step": 5513 + }, + { + "epoch": 0.9818376068376068, + "grad_norm": 0.5091170072555542, + "learning_rate": 0.00017185917577438643, + "loss": 1.0239, + "step": 5514 + }, + { + "epoch": 0.9820156695156695, + "grad_norm": 0.6155703067779541, + "learning_rate": 0.00017184944077818244, + "loss": 1.2366, + "step": 5515 + }, + { + "epoch": 0.9821937321937322, + "grad_norm": 0.5074070692062378, + "learning_rate": 0.0001718397043742379, + "loss": 1.0318, + "step": 5516 + }, + { + "epoch": 0.9823717948717948, + "grad_norm": 0.5234423279762268, + "learning_rate": 0.0001718299665627436, + "loss": 1.0322, + "step": 5517 + }, + { + "epoch": 0.9825498575498576, + "grad_norm": 0.5783474445343018, + "learning_rate": 0.0001718202273438903, + "loss": 0.9486, + "step": 5518 + }, + { + "epoch": 0.9827279202279202, + "grad_norm": 0.5708683133125305, + "learning_rate": 0.00017181048671786886, + "loss": 1.0785, + "step": 5519 + }, + { + "epoch": 0.9829059829059829, + "grad_norm": 0.5985961556434631, + "learning_rate": 0.00017180074468487009, + "loss": 1.198, + "step": 5520 + }, + { + "epoch": 0.9830840455840456, + "grad_norm": 0.5711352229118347, + "learning_rate": 0.0001717910012450849, + "loss": 1.0386, + "step": 5521 + }, + { + "epoch": 0.9832621082621082, + "grad_norm": 0.5338063836097717, + "learning_rate": 0.00017178125639870416, + "loss": 1.1594, + "step": 5522 + }, + { + "epoch": 0.9834401709401709, + "grad_norm": 0.6144943237304688, + "learning_rate": 0.00017177151014591881, + "loss": 1.1083, + "step": 5523 + }, + { + "epoch": 0.9836182336182336, + "grad_norm": 0.547285795211792, + "learning_rate": 0.00017176176248691983, + "loss": 1.1507, + "step": 5524 + }, + { + "epoch": 0.9837962962962963, + "grad_norm": 0.5807644724845886, + "learning_rate": 0.00017175201342189817, + "loss": 1.3044, + "step": 5525 + }, + { + "epoch": 0.9839743589743589, + "grad_norm": 0.5229477882385254, + "learning_rate": 0.00017174226295104485, + "loss": 1.2622, + "step": 5526 + }, + { + "epoch": 0.9841524216524217, + "grad_norm": 0.6100695133209229, + "learning_rate": 0.00017173251107455094, + "loss": 1.2026, + "step": 5527 + }, + { + "epoch": 0.9843304843304843, + "grad_norm": 0.5410884618759155, + "learning_rate": 0.00017172275779260744, + "loss": 1.2964, + "step": 5528 + }, + { + "epoch": 0.9845085470085471, + "grad_norm": 0.5937406420707703, + "learning_rate": 0.00017171300310540554, + "loss": 1.1435, + "step": 5529 + }, + { + "epoch": 0.9846866096866097, + "grad_norm": 0.56817227602005, + "learning_rate": 0.00017170324701313634, + "loss": 1.0099, + "step": 5530 + }, + { + "epoch": 0.9848646723646723, + "grad_norm": 0.5776323080062866, + "learning_rate": 0.00017169348951599092, + "loss": 1.3539, + "step": 5531 + }, + { + "epoch": 0.9850427350427351, + "grad_norm": 0.5208535194396973, + "learning_rate": 0.0001716837306141605, + "loss": 1.2306, + "step": 5532 + }, + { + "epoch": 0.9852207977207977, + "grad_norm": 0.552173376083374, + "learning_rate": 0.0001716739703078363, + "loss": 1.0551, + "step": 5533 + }, + { + "epoch": 0.9853988603988604, + "grad_norm": 0.5327515602111816, + "learning_rate": 0.00017166420859720955, + "loss": 1.2443, + "step": 5534 + }, + { + "epoch": 0.9855769230769231, + "grad_norm": 0.5255244374275208, + "learning_rate": 0.0001716544454824715, + "loss": 1.005, + "step": 5535 + }, + { + "epoch": 0.9857549857549858, + "grad_norm": 0.4753847122192383, + "learning_rate": 0.00017164468096381343, + "loss": 1.0081, + "step": 5536 + }, + { + "epoch": 0.9859330484330484, + "grad_norm": 0.5261829495429993, + "learning_rate": 0.00017163491504142665, + "loss": 1.2249, + "step": 5537 + }, + { + "epoch": 0.9861111111111112, + "grad_norm": 0.46499499678611755, + "learning_rate": 0.00017162514771550255, + "loss": 0.8759, + "step": 5538 + }, + { + "epoch": 0.9862891737891738, + "grad_norm": 0.5233004689216614, + "learning_rate": 0.00017161537898623247, + "loss": 1.0474, + "step": 5539 + }, + { + "epoch": 0.9864672364672364, + "grad_norm": 0.46905553340911865, + "learning_rate": 0.00017160560885380778, + "loss": 0.9033, + "step": 5540 + }, + { + "epoch": 0.9866452991452992, + "grad_norm": 0.5816231369972229, + "learning_rate": 0.00017159583731841998, + "loss": 1.0628, + "step": 5541 + }, + { + "epoch": 0.9868233618233618, + "grad_norm": 0.4575413167476654, + "learning_rate": 0.00017158606438026045, + "loss": 1.0446, + "step": 5542 + }, + { + "epoch": 0.9870014245014245, + "grad_norm": 0.5968109965324402, + "learning_rate": 0.00017157629003952067, + "loss": 1.032, + "step": 5543 + }, + { + "epoch": 0.9871794871794872, + "grad_norm": 0.5316148400306702, + "learning_rate": 0.00017156651429639218, + "loss": 0.9167, + "step": 5544 + }, + { + "epoch": 0.9873575498575499, + "grad_norm": 0.5185125470161438, + "learning_rate": 0.00017155673715106651, + "loss": 1.1527, + "step": 5545 + }, + { + "epoch": 0.9875356125356125, + "grad_norm": 0.5167772769927979, + "learning_rate": 0.00017154695860373525, + "loss": 0.9954, + "step": 5546 + }, + { + "epoch": 0.9877136752136753, + "grad_norm": 0.6406680345535278, + "learning_rate": 0.00017153717865458994, + "loss": 1.2758, + "step": 5547 + }, + { + "epoch": 0.9878917378917379, + "grad_norm": 0.5223956108093262, + "learning_rate": 0.00017152739730382223, + "loss": 1.1526, + "step": 5548 + }, + { + "epoch": 0.9880698005698005, + "grad_norm": 0.6131790280342102, + "learning_rate": 0.00017151761455162375, + "loss": 1.1024, + "step": 5549 + }, + { + "epoch": 0.9882478632478633, + "grad_norm": 0.5574753880500793, + "learning_rate": 0.00017150783039818616, + "loss": 0.9733, + "step": 5550 + }, + { + "epoch": 0.9884259259259259, + "grad_norm": 0.5417882800102234, + "learning_rate": 0.0001714980448437012, + "loss": 1.2244, + "step": 5551 + }, + { + "epoch": 0.9886039886039886, + "grad_norm": 0.6217474341392517, + "learning_rate": 0.0001714882578883606, + "loss": 0.9224, + "step": 5552 + }, + { + "epoch": 0.9887820512820513, + "grad_norm": 0.5846285223960876, + "learning_rate": 0.00017147846953235606, + "loss": 1.2429, + "step": 5553 + }, + { + "epoch": 0.988960113960114, + "grad_norm": 0.5924782752990723, + "learning_rate": 0.00017146867977587936, + "loss": 0.9907, + "step": 5554 + }, + { + "epoch": 0.9891381766381766, + "grad_norm": 0.5756853818893433, + "learning_rate": 0.00017145888861912242, + "loss": 1.1266, + "step": 5555 + }, + { + "epoch": 0.9893162393162394, + "grad_norm": 0.5277376770973206, + "learning_rate": 0.00017144909606227693, + "loss": 1.1676, + "step": 5556 + }, + { + "epoch": 0.989494301994302, + "grad_norm": 0.5138902068138123, + "learning_rate": 0.00017143930210553485, + "loss": 0.9864, + "step": 5557 + }, + { + "epoch": 0.9896723646723646, + "grad_norm": 0.8072507977485657, + "learning_rate": 0.00017142950674908805, + "loss": 1.111, + "step": 5558 + }, + { + "epoch": 0.9898504273504274, + "grad_norm": 0.5641721487045288, + "learning_rate": 0.00017141970999312844, + "loss": 0.9106, + "step": 5559 + }, + { + "epoch": 0.99002849002849, + "grad_norm": 0.5260798931121826, + "learning_rate": 0.000171409911837848, + "loss": 1.1609, + "step": 5560 + }, + { + "epoch": 0.9902065527065527, + "grad_norm": 0.5398530960083008, + "learning_rate": 0.00017140011228343864, + "loss": 1.0368, + "step": 5561 + }, + { + "epoch": 0.9903846153846154, + "grad_norm": 0.6011313199996948, + "learning_rate": 0.00017139031133009245, + "loss": 1.1314, + "step": 5562 + }, + { + "epoch": 0.9905626780626781, + "grad_norm": 0.6194971203804016, + "learning_rate": 0.00017138050897800135, + "loss": 1.3493, + "step": 5563 + }, + { + "epoch": 0.9907407407407407, + "grad_norm": 0.5779356956481934, + "learning_rate": 0.0001713707052273575, + "loss": 0.943, + "step": 5564 + }, + { + "epoch": 0.9909188034188035, + "grad_norm": 0.5321127772331238, + "learning_rate": 0.00017136090007835293, + "loss": 0.7914, + "step": 5565 + }, + { + "epoch": 0.9910968660968661, + "grad_norm": 0.5470426678657532, + "learning_rate": 0.00017135109353117977, + "loss": 1.2113, + "step": 5566 + }, + { + "epoch": 0.9912749287749287, + "grad_norm": 0.5551436543464661, + "learning_rate": 0.00017134128558603012, + "loss": 0.8932, + "step": 5567 + }, + { + "epoch": 0.9914529914529915, + "grad_norm": 0.45770928263664246, + "learning_rate": 0.0001713314762430962, + "loss": 1.0061, + "step": 5568 + }, + { + "epoch": 0.9916310541310541, + "grad_norm": 0.5578967332839966, + "learning_rate": 0.00017132166550257017, + "loss": 1.148, + "step": 5569 + }, + { + "epoch": 0.9918091168091168, + "grad_norm": 0.5086452960968018, + "learning_rate": 0.0001713118533646443, + "loss": 0.9803, + "step": 5570 + }, + { + "epoch": 0.9919871794871795, + "grad_norm": 0.4714745879173279, + "learning_rate": 0.00017130203982951078, + "loss": 1.0176, + "step": 5571 + }, + { + "epoch": 0.9921652421652422, + "grad_norm": 0.6254406571388245, + "learning_rate": 0.0001712922248973619, + "loss": 1.0932, + "step": 5572 + }, + { + "epoch": 0.9923433048433048, + "grad_norm": 0.5005003809928894, + "learning_rate": 0.00017128240856838998, + "loss": 1.0783, + "step": 5573 + }, + { + "epoch": 0.9925213675213675, + "grad_norm": 0.5668206214904785, + "learning_rate": 0.00017127259084278733, + "loss": 1.0404, + "step": 5574 + }, + { + "epoch": 0.9926994301994302, + "grad_norm": 0.4976036250591278, + "learning_rate": 0.00017126277172074632, + "loss": 1.1437, + "step": 5575 + }, + { + "epoch": 0.9928774928774928, + "grad_norm": 0.567546546459198, + "learning_rate": 0.00017125295120245935, + "loss": 1.2188, + "step": 5576 + }, + { + "epoch": 0.9930555555555556, + "grad_norm": 0.5614372491836548, + "learning_rate": 0.0001712431292881188, + "loss": 0.9187, + "step": 5577 + }, + { + "epoch": 0.9932336182336182, + "grad_norm": 0.6117973327636719, + "learning_rate": 0.00017123330597791712, + "loss": 1.1285, + "step": 5578 + }, + { + "epoch": 0.9934116809116809, + "grad_norm": 0.6000342965126038, + "learning_rate": 0.00017122348127204676, + "loss": 0.9837, + "step": 5579 + }, + { + "epoch": 0.9935897435897436, + "grad_norm": 0.5453050136566162, + "learning_rate": 0.0001712136551707003, + "loss": 0.8771, + "step": 5580 + }, + { + "epoch": 0.9937678062678063, + "grad_norm": 0.49603891372680664, + "learning_rate": 0.00017120382767407018, + "loss": 1.0754, + "step": 5581 + }, + { + "epoch": 0.9939458689458689, + "grad_norm": 0.48031488060951233, + "learning_rate": 0.00017119399878234894, + "loss": 0.6933, + "step": 5582 + }, + { + "epoch": 0.9941239316239316, + "grad_norm": 0.6048742532730103, + "learning_rate": 0.0001711841684957292, + "loss": 0.9696, + "step": 5583 + }, + { + "epoch": 0.9943019943019943, + "grad_norm": 0.5183123350143433, + "learning_rate": 0.00017117433681440355, + "loss": 1.1313, + "step": 5584 + }, + { + "epoch": 0.9944800569800569, + "grad_norm": 0.504916250705719, + "learning_rate": 0.00017116450373856466, + "loss": 1.0273, + "step": 5585 + }, + { + "epoch": 0.9946581196581197, + "grad_norm": 0.5804886817932129, + "learning_rate": 0.0001711546692684051, + "loss": 1.1162, + "step": 5586 + }, + { + "epoch": 0.9948361823361823, + "grad_norm": 0.5531938672065735, + "learning_rate": 0.0001711448334041176, + "loss": 1.2893, + "step": 5587 + }, + { + "epoch": 0.9950142450142451, + "grad_norm": 0.5079928636550903, + "learning_rate": 0.00017113499614589492, + "loss": 1.0393, + "step": 5588 + }, + { + "epoch": 0.9951923076923077, + "grad_norm": 0.5421964526176453, + "learning_rate": 0.00017112515749392973, + "loss": 0.8844, + "step": 5589 + }, + { + "epoch": 0.9953703703703703, + "grad_norm": 0.4834558367729187, + "learning_rate": 0.00017111531744841486, + "loss": 1.0187, + "step": 5590 + }, + { + "epoch": 0.9955484330484331, + "grad_norm": 0.6704340577125549, + "learning_rate": 0.00017110547600954307, + "loss": 0.8524, + "step": 5591 + }, + { + "epoch": 0.9957264957264957, + "grad_norm": 0.4578927159309387, + "learning_rate": 0.00017109563317750718, + "loss": 1.059, + "step": 5592 + }, + { + "epoch": 0.9959045584045584, + "grad_norm": 0.5563494563102722, + "learning_rate": 0.00017108578895250006, + "loss": 1.1211, + "step": 5593 + }, + { + "epoch": 0.9960826210826211, + "grad_norm": 0.5272170901298523, + "learning_rate": 0.00017107594333471454, + "loss": 0.9224, + "step": 5594 + }, + { + "epoch": 0.9962606837606838, + "grad_norm": 0.5697501301765442, + "learning_rate": 0.00017106609632434357, + "loss": 1.2223, + "step": 5595 + }, + { + "epoch": 0.9964387464387464, + "grad_norm": 0.5385653376579285, + "learning_rate": 0.00017105624792158007, + "loss": 1.0809, + "step": 5596 + }, + { + "epoch": 0.9966168091168092, + "grad_norm": 0.5608006119728088, + "learning_rate": 0.000171046398126617, + "loss": 1.3936, + "step": 5597 + }, + { + "epoch": 0.9967948717948718, + "grad_norm": 0.5063132643699646, + "learning_rate": 0.00017103654693964736, + "loss": 1.2086, + "step": 5598 + }, + { + "epoch": 0.9969729344729344, + "grad_norm": 0.6014235019683838, + "learning_rate": 0.00017102669436086415, + "loss": 1.1231, + "step": 5599 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 0.49549567699432373, + "learning_rate": 0.00017101684039046036, + "loss": 1.0013, + "step": 5600 + }, + { + "epoch": 0.9973290598290598, + "grad_norm": 0.517464816570282, + "learning_rate": 0.00017100698502862916, + "loss": 1.1143, + "step": 5601 + }, + { + "epoch": 0.9975071225071225, + "grad_norm": 0.514281153678894, + "learning_rate": 0.00017099712827556358, + "loss": 1.0336, + "step": 5602 + }, + { + "epoch": 0.9976851851851852, + "grad_norm": 0.5378567576408386, + "learning_rate": 0.00017098727013145672, + "loss": 0.8278, + "step": 5603 + }, + { + "epoch": 0.9978632478632479, + "grad_norm": 0.5098404884338379, + "learning_rate": 0.0001709774105965018, + "loss": 0.9902, + "step": 5604 + }, + { + "epoch": 0.9980413105413105, + "grad_norm": 0.6231759190559387, + "learning_rate": 0.00017096754967089198, + "loss": 1.0564, + "step": 5605 + }, + { + "epoch": 0.9982193732193733, + "grad_norm": 0.47434380650520325, + "learning_rate": 0.00017095768735482042, + "loss": 0.7457, + "step": 5606 + }, + { + "epoch": 0.9983974358974359, + "grad_norm": 0.5771013498306274, + "learning_rate": 0.00017094782364848035, + "loss": 1.1191, + "step": 5607 + }, + { + "epoch": 0.9985754985754985, + "grad_norm": 0.5617234706878662, + "learning_rate": 0.00017093795855206508, + "loss": 1.0779, + "step": 5608 + }, + { + "epoch": 0.9987535612535613, + "grad_norm": 0.6573554873466492, + "learning_rate": 0.00017092809206576792, + "loss": 1.0191, + "step": 5609 + }, + { + "epoch": 0.9989316239316239, + "grad_norm": 0.482834130525589, + "learning_rate": 0.00017091822418978207, + "loss": 1.0119, + "step": 5610 + }, + { + "epoch": 0.9991096866096866, + "grad_norm": 0.47496405243873596, + "learning_rate": 0.000170908354924301, + "loss": 0.8297, + "step": 5611 + }, + { + "epoch": 0.9992877492877493, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.00017089848426951796, + "loss": 1.1511, + "step": 5612 + }, + { + "epoch": 0.999465811965812, + "grad_norm": 0.5402522683143616, + "learning_rate": 0.00017088861222562643, + "loss": 1.1401, + "step": 5613 + }, + { + "epoch": 0.9996438746438746, + "grad_norm": 0.546302318572998, + "learning_rate": 0.00017087873879281977, + "loss": 0.8611, + "step": 5614 + }, + { + "epoch": 0.9998219373219374, + "grad_norm": 0.44279807806015015, + "learning_rate": 0.0001708688639712915, + "loss": 0.79, + "step": 5615 + }, + { + "epoch": 1.0, + "grad_norm": 0.5514659285545349, + "learning_rate": 0.00017085898776123502, + "loss": 1.0709, + "step": 5616 + }, + { + "epoch": 1.0, + "eval_loss": 1.093075156211853, + "eval_runtime": 24.6155, + "eval_samples_per_second": 42.29, + "eval_steps_per_second": 21.166, + "step": 5616 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.6290156841278076, + "learning_rate": 0.0001708491101628439, + "loss": 1.1786, + "step": 5617 + }, + { + "epoch": 1.0001780626780628, + "grad_norm": 0.4703841209411621, + "learning_rate": 0.00017083923117631162, + "loss": 0.9548, + "step": 5618 + }, + { + "epoch": 1.0003561253561253, + "grad_norm": 0.4518105089664459, + "learning_rate": 0.0001708293508018318, + "loss": 1.0089, + "step": 5619 + }, + { + "epoch": 1.000534188034188, + "grad_norm": 0.5658619403839111, + "learning_rate": 0.00017081946903959794, + "loss": 0.9466, + "step": 5620 + }, + { + "epoch": 1.0007122507122508, + "grad_norm": 0.6153838634490967, + "learning_rate": 0.00017080958588980372, + "loss": 1.2898, + "step": 5621 + }, + { + "epoch": 1.0008903133903133, + "grad_norm": 0.5245628952980042, + "learning_rate": 0.00017079970135264275, + "loss": 1.1702, + "step": 5622 + }, + { + "epoch": 1.001068376068376, + "grad_norm": 0.5291880965232849, + "learning_rate": 0.00017078981542830875, + "loss": 1.0779, + "step": 5623 + }, + { + "epoch": 1.0012464387464388, + "grad_norm": 0.500579297542572, + "learning_rate": 0.0001707799281169953, + "loss": 0.9587, + "step": 5624 + }, + { + "epoch": 1.0014245014245013, + "grad_norm": 0.45739707350730896, + "learning_rate": 0.00017077003941889625, + "loss": 0.9373, + "step": 5625 + }, + { + "epoch": 1.001602564102564, + "grad_norm": 0.5513401031494141, + "learning_rate": 0.00017076014933420526, + "loss": 1.0368, + "step": 5626 + }, + { + "epoch": 1.0017806267806268, + "grad_norm": 0.46513232588768005, + "learning_rate": 0.00017075025786311612, + "loss": 0.9422, + "step": 5627 + }, + { + "epoch": 1.0019586894586894, + "grad_norm": 0.4530394673347473, + "learning_rate": 0.00017074036500582267, + "loss": 0.8211, + "step": 5628 + }, + { + "epoch": 1.0021367521367521, + "grad_norm": 0.5612013339996338, + "learning_rate": 0.00017073047076251872, + "loss": 0.9466, + "step": 5629 + }, + { + "epoch": 1.0023148148148149, + "grad_norm": 0.4976879954338074, + "learning_rate": 0.00017072057513339812, + "loss": 0.8059, + "step": 5630 + }, + { + "epoch": 1.0024928774928774, + "grad_norm": 0.4842833876609802, + "learning_rate": 0.00017071067811865476, + "loss": 0.6554, + "step": 5631 + }, + { + "epoch": 1.0026709401709402, + "grad_norm": 0.5446373224258423, + "learning_rate": 0.00017070077971848257, + "loss": 1.1001, + "step": 5632 + }, + { + "epoch": 1.002849002849003, + "grad_norm": 0.5996584892272949, + "learning_rate": 0.00017069087993307544, + "loss": 1.0317, + "step": 5633 + }, + { + "epoch": 1.0030270655270654, + "grad_norm": 0.5369443297386169, + "learning_rate": 0.00017068097876262738, + "loss": 0.8019, + "step": 5634 + }, + { + "epoch": 1.0032051282051282, + "grad_norm": 0.4985966682434082, + "learning_rate": 0.00017067107620733236, + "loss": 1.0121, + "step": 5635 + }, + { + "epoch": 1.003383190883191, + "grad_norm": 0.5262824892997742, + "learning_rate": 0.0001706611722673844, + "loss": 1.0157, + "step": 5636 + }, + { + "epoch": 1.0035612535612535, + "grad_norm": 0.5912795066833496, + "learning_rate": 0.00017065126694297756, + "loss": 1.0327, + "step": 5637 + }, + { + "epoch": 1.0037393162393162, + "grad_norm": 0.5866343379020691, + "learning_rate": 0.00017064136023430595, + "loss": 1.1194, + "step": 5638 + }, + { + "epoch": 1.003917378917379, + "grad_norm": 0.5009918808937073, + "learning_rate": 0.0001706314521415636, + "loss": 1.0467, + "step": 5639 + }, + { + "epoch": 1.0040954415954415, + "grad_norm": 0.5455304384231567, + "learning_rate": 0.00017062154266494464, + "loss": 0.8749, + "step": 5640 + }, + { + "epoch": 1.0042735042735043, + "grad_norm": 0.5648258328437805, + "learning_rate": 0.00017061163180464328, + "loss": 0.9408, + "step": 5641 + }, + { + "epoch": 1.004451566951567, + "grad_norm": 0.5276365876197815, + "learning_rate": 0.00017060171956085368, + "loss": 0.9681, + "step": 5642 + }, + { + "epoch": 1.0046296296296295, + "grad_norm": 0.5212745070457458, + "learning_rate": 0.00017059180593377007, + "loss": 0.9188, + "step": 5643 + }, + { + "epoch": 1.0048076923076923, + "grad_norm": 0.540626585483551, + "learning_rate": 0.00017058189092358664, + "loss": 1.0809, + "step": 5644 + }, + { + "epoch": 1.004985754985755, + "grad_norm": 0.5592377781867981, + "learning_rate": 0.00017057197453049767, + "loss": 0.8589, + "step": 5645 + }, + { + "epoch": 1.0051638176638176, + "grad_norm": 0.5115051865577698, + "learning_rate": 0.00017056205675469746, + "loss": 0.8006, + "step": 5646 + }, + { + "epoch": 1.0053418803418803, + "grad_norm": 0.5031117796897888, + "learning_rate": 0.00017055213759638034, + "loss": 0.9242, + "step": 5647 + }, + { + "epoch": 1.005519943019943, + "grad_norm": 0.5342774987220764, + "learning_rate": 0.00017054221705574066, + "loss": 0.8268, + "step": 5648 + }, + { + "epoch": 1.0056980056980056, + "grad_norm": 0.44480493664741516, + "learning_rate": 0.00017053229513297276, + "loss": 0.6892, + "step": 5649 + }, + { + "epoch": 1.0058760683760684, + "grad_norm": 0.5032621622085571, + "learning_rate": 0.00017052237182827105, + "loss": 0.971, + "step": 5650 + }, + { + "epoch": 1.006054131054131, + "grad_norm": 0.5611015558242798, + "learning_rate": 0.00017051244714182996, + "loss": 0.9403, + "step": 5651 + }, + { + "epoch": 1.0062321937321936, + "grad_norm": 0.5064613223075867, + "learning_rate": 0.00017050252107384393, + "loss": 0.9718, + "step": 5652 + }, + { + "epoch": 1.0064102564102564, + "grad_norm": 0.6458395719528198, + "learning_rate": 0.0001704925936245075, + "loss": 1.1161, + "step": 5653 + }, + { + "epoch": 1.0065883190883191, + "grad_norm": 0.527418315410614, + "learning_rate": 0.00017048266479401512, + "loss": 0.9315, + "step": 5654 + }, + { + "epoch": 1.0067663817663817, + "grad_norm": 0.5127941370010376, + "learning_rate": 0.00017047273458256133, + "loss": 0.8206, + "step": 5655 + }, + { + "epoch": 1.0069444444444444, + "grad_norm": 0.6257100105285645, + "learning_rate": 0.00017046280299034067, + "loss": 0.9854, + "step": 5656 + }, + { + "epoch": 1.0071225071225072, + "grad_norm": 0.5081700682640076, + "learning_rate": 0.0001704528700175478, + "loss": 0.9478, + "step": 5657 + }, + { + "epoch": 1.0073005698005697, + "grad_norm": 0.598127543926239, + "learning_rate": 0.00017044293566437725, + "loss": 1.0721, + "step": 5658 + }, + { + "epoch": 1.0074786324786325, + "grad_norm": 0.5429877638816833, + "learning_rate": 0.00017043299993102376, + "loss": 0.9732, + "step": 5659 + }, + { + "epoch": 1.0076566951566952, + "grad_norm": 0.6006619334220886, + "learning_rate": 0.00017042306281768194, + "loss": 1.1262, + "step": 5660 + }, + { + "epoch": 1.0078347578347577, + "grad_norm": 0.48933324217796326, + "learning_rate": 0.00017041312432454646, + "loss": 0.8596, + "step": 5661 + }, + { + "epoch": 1.0080128205128205, + "grad_norm": 0.5902166366577148, + "learning_rate": 0.0001704031844518121, + "loss": 1.1035, + "step": 5662 + }, + { + "epoch": 1.0081908831908832, + "grad_norm": 0.523597776889801, + "learning_rate": 0.0001703932431996736, + "loss": 0.7117, + "step": 5663 + }, + { + "epoch": 1.0083689458689458, + "grad_norm": 0.6313928365707397, + "learning_rate": 0.00017038330056832573, + "loss": 1.0204, + "step": 5664 + }, + { + "epoch": 1.0085470085470085, + "grad_norm": 0.5627471804618835, + "learning_rate": 0.00017037335655796328, + "loss": 0.7648, + "step": 5665 + }, + { + "epoch": 1.0087250712250713, + "grad_norm": 0.5817851424217224, + "learning_rate": 0.0001703634111687811, + "loss": 1.0452, + "step": 5666 + }, + { + "epoch": 1.0089031339031338, + "grad_norm": 0.5143535137176514, + "learning_rate": 0.00017035346440097407, + "loss": 0.9788, + "step": 5667 + }, + { + "epoch": 1.0090811965811965, + "grad_norm": 0.5331187844276428, + "learning_rate": 0.000170343516254737, + "loss": 0.7584, + "step": 5668 + }, + { + "epoch": 1.0092592592592593, + "grad_norm": 0.5723634362220764, + "learning_rate": 0.00017033356673026487, + "loss": 0.9435, + "step": 5669 + }, + { + "epoch": 1.0094373219373218, + "grad_norm": 0.6012297868728638, + "learning_rate": 0.00017032361582775265, + "loss": 1.142, + "step": 5670 + }, + { + "epoch": 1.0096153846153846, + "grad_norm": 0.6161282658576965, + "learning_rate": 0.00017031366354739523, + "loss": 1.2823, + "step": 5671 + }, + { + "epoch": 1.0097934472934473, + "grad_norm": 0.5088054537773132, + "learning_rate": 0.00017030370988938763, + "loss": 0.9743, + "step": 5672 + }, + { + "epoch": 1.0099715099715099, + "grad_norm": 0.512003481388092, + "learning_rate": 0.0001702937548539249, + "loss": 0.9112, + "step": 5673 + }, + { + "epoch": 1.0101495726495726, + "grad_norm": 0.5565149784088135, + "learning_rate": 0.00017028379844120207, + "loss": 1.0074, + "step": 5674 + }, + { + "epoch": 1.0103276353276354, + "grad_norm": 0.6463099718093872, + "learning_rate": 0.00017027384065141418, + "loss": 1.175, + "step": 5675 + }, + { + "epoch": 1.010505698005698, + "grad_norm": 0.46999064087867737, + "learning_rate": 0.00017026388148475637, + "loss": 0.8429, + "step": 5676 + }, + { + "epoch": 1.0106837606837606, + "grad_norm": 0.5617384910583496, + "learning_rate": 0.00017025392094142377, + "loss": 1.045, + "step": 5677 + }, + { + "epoch": 1.0108618233618234, + "grad_norm": 0.5156623721122742, + "learning_rate": 0.00017024395902161154, + "loss": 1.016, + "step": 5678 + }, + { + "epoch": 1.0110398860398861, + "grad_norm": 0.5693390369415283, + "learning_rate": 0.00017023399572551484, + "loss": 0.8616, + "step": 5679 + }, + { + "epoch": 1.0112179487179487, + "grad_norm": 0.5234879851341248, + "learning_rate": 0.00017022403105332892, + "loss": 0.9244, + "step": 5680 + }, + { + "epoch": 1.0113960113960114, + "grad_norm": 0.6513097286224365, + "learning_rate": 0.00017021406500524893, + "loss": 0.9565, + "step": 5681 + }, + { + "epoch": 1.0115740740740742, + "grad_norm": 0.5788878202438354, + "learning_rate": 0.00017020409758147022, + "loss": 0.8994, + "step": 5682 + }, + { + "epoch": 1.0117521367521367, + "grad_norm": 0.5495247840881348, + "learning_rate": 0.00017019412878218807, + "loss": 0.9371, + "step": 5683 + }, + { + "epoch": 1.0119301994301995, + "grad_norm": 0.639045238494873, + "learning_rate": 0.00017018415860759777, + "loss": 1.0297, + "step": 5684 + }, + { + "epoch": 1.0121082621082622, + "grad_norm": 0.5167784690856934, + "learning_rate": 0.0001701741870578947, + "loss": 0.8974, + "step": 5685 + }, + { + "epoch": 1.0122863247863247, + "grad_norm": 0.6131011247634888, + "learning_rate": 0.00017016421413327417, + "loss": 1.13, + "step": 5686 + }, + { + "epoch": 1.0124643874643875, + "grad_norm": 0.4804688096046448, + "learning_rate": 0.00017015423983393166, + "loss": 1.0098, + "step": 5687 + }, + { + "epoch": 1.0126424501424502, + "grad_norm": 0.6605221629142761, + "learning_rate": 0.00017014426416006253, + "loss": 1.1123, + "step": 5688 + }, + { + "epoch": 1.0128205128205128, + "grad_norm": 0.5523666739463806, + "learning_rate": 0.00017013428711186226, + "loss": 0.8226, + "step": 5689 + }, + { + "epoch": 1.0129985754985755, + "grad_norm": 0.6012941598892212, + "learning_rate": 0.00017012430868952632, + "loss": 0.8915, + "step": 5690 + }, + { + "epoch": 1.0131766381766383, + "grad_norm": 0.5830875039100647, + "learning_rate": 0.00017011432889325022, + "loss": 1.021, + "step": 5691 + }, + { + "epoch": 1.0133547008547008, + "grad_norm": 0.5546056032180786, + "learning_rate": 0.0001701043477232295, + "loss": 0.7656, + "step": 5692 + }, + { + "epoch": 1.0135327635327636, + "grad_norm": 0.5592601299285889, + "learning_rate": 0.0001700943651796597, + "loss": 1.0172, + "step": 5693 + }, + { + "epoch": 1.0137108262108263, + "grad_norm": 0.5708866715431213, + "learning_rate": 0.00017008438126273645, + "loss": 1.0012, + "step": 5694 + }, + { + "epoch": 1.0138888888888888, + "grad_norm": 0.6856338381767273, + "learning_rate": 0.0001700743959726553, + "loss": 1.1278, + "step": 5695 + }, + { + "epoch": 1.0140669515669516, + "grad_norm": 0.6523802876472473, + "learning_rate": 0.000170064409309612, + "loss": 1.0406, + "step": 5696 + }, + { + "epoch": 1.0142450142450143, + "grad_norm": 0.6653079986572266, + "learning_rate": 0.00017005442127380208, + "loss": 1.1086, + "step": 5697 + }, + { + "epoch": 1.0144230769230769, + "grad_norm": 0.5841104388237, + "learning_rate": 0.00017004443186542133, + "loss": 0.9335, + "step": 5698 + }, + { + "epoch": 1.0146011396011396, + "grad_norm": 0.5696784257888794, + "learning_rate": 0.0001700344410846654, + "loss": 1.0247, + "step": 5699 + }, + { + "epoch": 1.0147792022792024, + "grad_norm": 0.7135653495788574, + "learning_rate": 0.00017002444893173013, + "loss": 1.0259, + "step": 5700 + }, + { + "epoch": 1.014957264957265, + "grad_norm": 0.5806999802589417, + "learning_rate": 0.00017001445540681124, + "loss": 1.0053, + "step": 5701 + }, + { + "epoch": 1.0151353276353277, + "grad_norm": 0.5298715829849243, + "learning_rate": 0.0001700044605101045, + "loss": 0.9415, + "step": 5702 + }, + { + "epoch": 1.0153133903133904, + "grad_norm": 0.5817379951477051, + "learning_rate": 0.0001699944642418058, + "loss": 1.0906, + "step": 5703 + }, + { + "epoch": 1.015491452991453, + "grad_norm": 0.6564923524856567, + "learning_rate": 0.00016998446660211098, + "loss": 0.9933, + "step": 5704 + }, + { + "epoch": 1.0156695156695157, + "grad_norm": 0.6547308564186096, + "learning_rate": 0.00016997446759121592, + "loss": 1.0045, + "step": 5705 + }, + { + "epoch": 1.0158475783475784, + "grad_norm": 0.5763013958930969, + "learning_rate": 0.00016996446720931652, + "loss": 1.0898, + "step": 5706 + }, + { + "epoch": 1.016025641025641, + "grad_norm": 0.6118074059486389, + "learning_rate": 0.00016995446545660871, + "loss": 0.9398, + "step": 5707 + }, + { + "epoch": 1.0162037037037037, + "grad_norm": 0.6810526251792908, + "learning_rate": 0.0001699444623332885, + "loss": 1.0968, + "step": 5708 + }, + { + "epoch": 1.0163817663817665, + "grad_norm": 0.5292752981185913, + "learning_rate": 0.00016993445783955184, + "loss": 0.7549, + "step": 5709 + }, + { + "epoch": 1.016559829059829, + "grad_norm": 0.6014277935028076, + "learning_rate": 0.00016992445197559474, + "loss": 1.1711, + "step": 5710 + }, + { + "epoch": 1.0167378917378918, + "grad_norm": 0.5089772343635559, + "learning_rate": 0.00016991444474161326, + "loss": 0.9188, + "step": 5711 + }, + { + "epoch": 1.0169159544159545, + "grad_norm": 0.567193865776062, + "learning_rate": 0.0001699044361378035, + "loss": 0.7462, + "step": 5712 + }, + { + "epoch": 1.017094017094017, + "grad_norm": 0.5638598799705505, + "learning_rate": 0.00016989442616436147, + "loss": 0.9643, + "step": 5713 + }, + { + "epoch": 1.0172720797720798, + "grad_norm": 0.5634039640426636, + "learning_rate": 0.0001698844148214834, + "loss": 1.0141, + "step": 5714 + }, + { + "epoch": 1.0174501424501425, + "grad_norm": 0.5326652526855469, + "learning_rate": 0.00016987440210936537, + "loss": 0.865, + "step": 5715 + }, + { + "epoch": 1.017628205128205, + "grad_norm": 0.5858046412467957, + "learning_rate": 0.0001698643880282036, + "loss": 0.9561, + "step": 5716 + }, + { + "epoch": 1.0178062678062678, + "grad_norm": 0.6424698829650879, + "learning_rate": 0.00016985437257819428, + "loss": 1.0169, + "step": 5717 + }, + { + "epoch": 1.0179843304843306, + "grad_norm": 0.6294280290603638, + "learning_rate": 0.00016984435575953364, + "loss": 1.0438, + "step": 5718 + }, + { + "epoch": 1.018162393162393, + "grad_norm": 0.5533088445663452, + "learning_rate": 0.00016983433757241788, + "loss": 0.8901, + "step": 5719 + }, + { + "epoch": 1.0183404558404558, + "grad_norm": 0.5148718953132629, + "learning_rate": 0.00016982431801704342, + "loss": 0.9201, + "step": 5720 + }, + { + "epoch": 1.0185185185185186, + "grad_norm": 0.5609371662139893, + "learning_rate": 0.00016981429709360645, + "loss": 0.9347, + "step": 5721 + }, + { + "epoch": 1.0186965811965811, + "grad_norm": 0.5502731204032898, + "learning_rate": 0.00016980427480230338, + "loss": 1.0508, + "step": 5722 + }, + { + "epoch": 1.0188746438746439, + "grad_norm": 0.5880394577980042, + "learning_rate": 0.00016979425114333055, + "loss": 1.1258, + "step": 5723 + }, + { + "epoch": 1.0190527065527066, + "grad_norm": 0.5569866895675659, + "learning_rate": 0.0001697842261168843, + "loss": 0.9186, + "step": 5724 + }, + { + "epoch": 1.0192307692307692, + "grad_norm": 0.7468093037605286, + "learning_rate": 0.00016977419972316116, + "loss": 1.2066, + "step": 5725 + }, + { + "epoch": 1.019408831908832, + "grad_norm": 0.6041515469551086, + "learning_rate": 0.00016976417196235753, + "loss": 0.939, + "step": 5726 + }, + { + "epoch": 1.0195868945868947, + "grad_norm": 0.6102641224861145, + "learning_rate": 0.00016975414283466983, + "loss": 0.8334, + "step": 5727 + }, + { + "epoch": 1.0197649572649572, + "grad_norm": 0.5418640375137329, + "learning_rate": 0.00016974411234029467, + "loss": 0.8072, + "step": 5728 + }, + { + "epoch": 1.01994301994302, + "grad_norm": 0.6569705605506897, + "learning_rate": 0.00016973408047942843, + "loss": 1.103, + "step": 5729 + }, + { + "epoch": 1.0201210826210827, + "grad_norm": 0.5778102278709412, + "learning_rate": 0.00016972404725226778, + "loss": 0.9353, + "step": 5730 + }, + { + "epoch": 1.0202991452991452, + "grad_norm": 0.5474382638931274, + "learning_rate": 0.0001697140126590093, + "loss": 1.0009, + "step": 5731 + }, + { + "epoch": 1.020477207977208, + "grad_norm": 0.5869506597518921, + "learning_rate": 0.00016970397669984947, + "loss": 1.0027, + "step": 5732 + }, + { + "epoch": 1.0206552706552707, + "grad_norm": 0.5078117251396179, + "learning_rate": 0.00016969393937498508, + "loss": 0.8316, + "step": 5733 + }, + { + "epoch": 1.0208333333333333, + "grad_norm": 0.5488452911376953, + "learning_rate": 0.0001696839006846127, + "loss": 0.8438, + "step": 5734 + }, + { + "epoch": 1.021011396011396, + "grad_norm": 0.5921052098274231, + "learning_rate": 0.00016967386062892908, + "loss": 0.9147, + "step": 5735 + }, + { + "epoch": 1.0211894586894588, + "grad_norm": 0.5486881136894226, + "learning_rate": 0.00016966381920813085, + "loss": 0.7619, + "step": 5736 + }, + { + "epoch": 1.0213675213675213, + "grad_norm": 0.5250689387321472, + "learning_rate": 0.00016965377642241483, + "loss": 0.9192, + "step": 5737 + }, + { + "epoch": 1.021545584045584, + "grad_norm": 0.5355087518692017, + "learning_rate": 0.00016964373227197773, + "loss": 0.954, + "step": 5738 + }, + { + "epoch": 1.0217236467236468, + "grad_norm": 0.6758780479431152, + "learning_rate": 0.0001696336867570164, + "loss": 1.1257, + "step": 5739 + }, + { + "epoch": 1.0219017094017093, + "grad_norm": 0.6361044049263, + "learning_rate": 0.00016962363987772756, + "loss": 1.0889, + "step": 5740 + }, + { + "epoch": 1.022079772079772, + "grad_norm": 0.5802326798439026, + "learning_rate": 0.00016961359163430819, + "loss": 0.8966, + "step": 5741 + }, + { + "epoch": 1.0222578347578348, + "grad_norm": 0.5535712242126465, + "learning_rate": 0.00016960354202695508, + "loss": 1.0007, + "step": 5742 + }, + { + "epoch": 1.0224358974358974, + "grad_norm": 0.5469220280647278, + "learning_rate": 0.00016959349105586516, + "loss": 0.8202, + "step": 5743 + }, + { + "epoch": 1.02261396011396, + "grad_norm": 0.5533008575439453, + "learning_rate": 0.00016958343872123534, + "loss": 0.9576, + "step": 5744 + }, + { + "epoch": 1.0227920227920229, + "grad_norm": 0.615132749080658, + "learning_rate": 0.00016957338502326258, + "loss": 0.8719, + "step": 5745 + }, + { + "epoch": 1.0229700854700854, + "grad_norm": 0.519075334072113, + "learning_rate": 0.0001695633299621439, + "loss": 0.8309, + "step": 5746 + }, + { + "epoch": 1.0231481481481481, + "grad_norm": 0.6249759197235107, + "learning_rate": 0.00016955327353807624, + "loss": 1.151, + "step": 5747 + }, + { + "epoch": 1.023326210826211, + "grad_norm": 0.560299277305603, + "learning_rate": 0.00016954321575125668, + "loss": 0.7889, + "step": 5748 + }, + { + "epoch": 1.0235042735042734, + "grad_norm": 0.5735262036323547, + "learning_rate": 0.0001695331566018823, + "loss": 0.8794, + "step": 5749 + }, + { + "epoch": 1.0236823361823362, + "grad_norm": 0.5893994569778442, + "learning_rate": 0.00016952309609015012, + "loss": 0.9696, + "step": 5750 + }, + { + "epoch": 1.023860398860399, + "grad_norm": 0.6064512133598328, + "learning_rate": 0.0001695130342162573, + "loss": 0.9771, + "step": 5751 + }, + { + "epoch": 1.0240384615384615, + "grad_norm": 0.5833427309989929, + "learning_rate": 0.00016950297098040099, + "loss": 1.1768, + "step": 5752 + }, + { + "epoch": 1.0242165242165242, + "grad_norm": 0.5940282344818115, + "learning_rate": 0.00016949290638277833, + "loss": 1.0758, + "step": 5753 + }, + { + "epoch": 1.024394586894587, + "grad_norm": 0.5267124772071838, + "learning_rate": 0.00016948284042358656, + "loss": 0.772, + "step": 5754 + }, + { + "epoch": 1.0245726495726495, + "grad_norm": 0.6217982172966003, + "learning_rate": 0.00016947277310302284, + "loss": 0.8583, + "step": 5755 + }, + { + "epoch": 1.0247507122507122, + "grad_norm": 0.6192215085029602, + "learning_rate": 0.00016946270442128443, + "loss": 0.9148, + "step": 5756 + }, + { + "epoch": 1.024928774928775, + "grad_norm": 0.5337123870849609, + "learning_rate": 0.00016945263437856867, + "loss": 1.0054, + "step": 5757 + }, + { + "epoch": 1.0251068376068375, + "grad_norm": 0.5462040901184082, + "learning_rate": 0.00016944256297507276, + "loss": 1.1097, + "step": 5758 + }, + { + "epoch": 1.0252849002849003, + "grad_norm": 0.5606170892715454, + "learning_rate": 0.00016943249021099415, + "loss": 1.0192, + "step": 5759 + }, + { + "epoch": 1.025462962962963, + "grad_norm": 0.636974573135376, + "learning_rate": 0.00016942241608653008, + "loss": 1.0241, + "step": 5760 + }, + { + "epoch": 1.0256410256410255, + "grad_norm": 0.4895164966583252, + "learning_rate": 0.00016941234060187797, + "loss": 0.9057, + "step": 5761 + }, + { + "epoch": 1.0258190883190883, + "grad_norm": 0.5810303092002869, + "learning_rate": 0.00016940226375723527, + "loss": 1.0809, + "step": 5762 + }, + { + "epoch": 1.025997150997151, + "grad_norm": 0.6043853163719177, + "learning_rate": 0.00016939218555279937, + "loss": 1.0685, + "step": 5763 + }, + { + "epoch": 1.0261752136752136, + "grad_norm": 0.5827188491821289, + "learning_rate": 0.00016938210598876774, + "loss": 1.0236, + "step": 5764 + }, + { + "epoch": 1.0263532763532763, + "grad_norm": 0.6677887439727783, + "learning_rate": 0.0001693720250653379, + "loss": 1.0586, + "step": 5765 + }, + { + "epoch": 1.026531339031339, + "grad_norm": 0.558051347732544, + "learning_rate": 0.0001693619427827073, + "loss": 0.745, + "step": 5766 + }, + { + "epoch": 1.0267094017094016, + "grad_norm": 0.6336706280708313, + "learning_rate": 0.0001693518591410735, + "loss": 1.0658, + "step": 5767 + }, + { + "epoch": 1.0268874643874644, + "grad_norm": 0.7077126502990723, + "learning_rate": 0.00016934177414063416, + "loss": 1.18, + "step": 5768 + }, + { + "epoch": 1.0270655270655271, + "grad_norm": 0.5342326760292053, + "learning_rate": 0.00016933168778158675, + "loss": 0.8347, + "step": 5769 + }, + { + "epoch": 1.0272435897435896, + "grad_norm": 0.6116416454315186, + "learning_rate": 0.00016932160006412895, + "loss": 1.0648, + "step": 5770 + }, + { + "epoch": 1.0274216524216524, + "grad_norm": 0.5411320924758911, + "learning_rate": 0.0001693115109884584, + "loss": 1.0756, + "step": 5771 + }, + { + "epoch": 1.0275997150997151, + "grad_norm": 0.5549847483634949, + "learning_rate": 0.00016930142055477277, + "loss": 0.7259, + "step": 5772 + }, + { + "epoch": 1.0277777777777777, + "grad_norm": 0.549010694026947, + "learning_rate": 0.00016929132876326977, + "loss": 0.9488, + "step": 5773 + }, + { + "epoch": 1.0279558404558404, + "grad_norm": 0.6302017569541931, + "learning_rate": 0.00016928123561414714, + "loss": 0.8851, + "step": 5774 + }, + { + "epoch": 1.0281339031339032, + "grad_norm": 0.5831273198127747, + "learning_rate": 0.00016927114110760257, + "loss": 0.7841, + "step": 5775 + }, + { + "epoch": 1.0283119658119657, + "grad_norm": 0.5528474450111389, + "learning_rate": 0.00016926104524383394, + "loss": 1.0108, + "step": 5776 + }, + { + "epoch": 1.0284900284900285, + "grad_norm": 0.6279126405715942, + "learning_rate": 0.00016925094802303897, + "loss": 0.8632, + "step": 5777 + }, + { + "epoch": 1.0286680911680912, + "grad_norm": 0.6783218383789062, + "learning_rate": 0.00016924084944541554, + "loss": 1.0746, + "step": 5778 + }, + { + "epoch": 1.0288461538461537, + "grad_norm": 0.5823925137519836, + "learning_rate": 0.00016923074951116153, + "loss": 1.0486, + "step": 5779 + }, + { + "epoch": 1.0290242165242165, + "grad_norm": 0.6095981597900391, + "learning_rate": 0.00016922064822047473, + "loss": 0.8113, + "step": 5780 + }, + { + "epoch": 1.0292022792022792, + "grad_norm": 0.7887664437294006, + "learning_rate": 0.00016921054557355317, + "loss": 1.2411, + "step": 5781 + }, + { + "epoch": 1.0293803418803418, + "grad_norm": 0.6511263251304626, + "learning_rate": 0.00016920044157059475, + "loss": 0.924, + "step": 5782 + }, + { + "epoch": 1.0295584045584045, + "grad_norm": 0.6045661568641663, + "learning_rate": 0.00016919033621179744, + "loss": 0.8373, + "step": 5783 + }, + { + "epoch": 1.0297364672364673, + "grad_norm": 0.6914188861846924, + "learning_rate": 0.0001691802294973592, + "loss": 0.9589, + "step": 5784 + }, + { + "epoch": 1.0299145299145298, + "grad_norm": 0.6483730673789978, + "learning_rate": 0.00016917012142747805, + "loss": 0.9871, + "step": 5785 + }, + { + "epoch": 1.0300925925925926, + "grad_norm": 0.5775033235549927, + "learning_rate": 0.0001691600120023521, + "loss": 1.0591, + "step": 5786 + }, + { + "epoch": 1.0302706552706553, + "grad_norm": 0.6206814646720886, + "learning_rate": 0.00016914990122217932, + "loss": 0.9126, + "step": 5787 + }, + { + "epoch": 1.0304487179487178, + "grad_norm": 0.5422028303146362, + "learning_rate": 0.00016913978908715796, + "loss": 0.8227, + "step": 5788 + }, + { + "epoch": 1.0306267806267806, + "grad_norm": 0.5824416875839233, + "learning_rate": 0.000169129675597486, + "loss": 1.111, + "step": 5789 + }, + { + "epoch": 1.0308048433048433, + "grad_norm": 0.5419015884399414, + "learning_rate": 0.00016911956075336165, + "loss": 0.8941, + "step": 5790 + }, + { + "epoch": 1.0309829059829059, + "grad_norm": 0.6171557903289795, + "learning_rate": 0.0001691094445549831, + "loss": 0.8679, + "step": 5791 + }, + { + "epoch": 1.0311609686609686, + "grad_norm": 0.6136980056762695, + "learning_rate": 0.00016909932700254855, + "loss": 0.9266, + "step": 5792 + }, + { + "epoch": 1.0313390313390314, + "grad_norm": 0.6275020241737366, + "learning_rate": 0.00016908920809625624, + "loss": 1.0828, + "step": 5793 + }, + { + "epoch": 1.0315170940170941, + "grad_norm": 0.6538251638412476, + "learning_rate": 0.0001690790878363044, + "loss": 0.8413, + "step": 5794 + }, + { + "epoch": 1.0316951566951567, + "grad_norm": 0.5981295108795166, + "learning_rate": 0.00016906896622289136, + "loss": 0.9845, + "step": 5795 + }, + { + "epoch": 1.0318732193732194, + "grad_norm": 0.5390967130661011, + "learning_rate": 0.00016905884325621538, + "loss": 0.8755, + "step": 5796 + }, + { + "epoch": 1.032051282051282, + "grad_norm": 0.5534448623657227, + "learning_rate": 0.00016904871893647482, + "loss": 1.1868, + "step": 5797 + }, + { + "epoch": 1.0322293447293447, + "grad_norm": 0.664556086063385, + "learning_rate": 0.00016903859326386806, + "loss": 1.1418, + "step": 5798 + }, + { + "epoch": 1.0324074074074074, + "grad_norm": 0.5737143158912659, + "learning_rate": 0.00016902846623859346, + "loss": 1.124, + "step": 5799 + }, + { + "epoch": 1.0325854700854702, + "grad_norm": 0.6499935388565063, + "learning_rate": 0.0001690183378608495, + "loss": 1.0331, + "step": 5800 + }, + { + "epoch": 1.0327635327635327, + "grad_norm": 0.5721518993377686, + "learning_rate": 0.00016900820813083454, + "loss": 0.8664, + "step": 5801 + }, + { + "epoch": 1.0329415954415955, + "grad_norm": 0.5651140809059143, + "learning_rate": 0.0001689980770487471, + "loss": 1.1661, + "step": 5802 + }, + { + "epoch": 1.0331196581196582, + "grad_norm": 0.5935871005058289, + "learning_rate": 0.0001689879446147857, + "loss": 0.8722, + "step": 5803 + }, + { + "epoch": 1.0332977207977208, + "grad_norm": 0.5627842545509338, + "learning_rate": 0.00016897781082914884, + "loss": 1.0036, + "step": 5804 + }, + { + "epoch": 1.0334757834757835, + "grad_norm": 0.5866895914077759, + "learning_rate": 0.00016896767569203502, + "loss": 0.9739, + "step": 5805 + }, + { + "epoch": 1.0336538461538463, + "grad_norm": 0.5568059682846069, + "learning_rate": 0.0001689575392036429, + "loss": 0.7081, + "step": 5806 + }, + { + "epoch": 1.0338319088319088, + "grad_norm": 0.6054235100746155, + "learning_rate": 0.00016894740136417103, + "loss": 1.1168, + "step": 5807 + }, + { + "epoch": 1.0340099715099715, + "grad_norm": 0.5215454697608948, + "learning_rate": 0.00016893726217381805, + "loss": 0.9172, + "step": 5808 + }, + { + "epoch": 1.0341880341880343, + "grad_norm": 0.5415732860565186, + "learning_rate": 0.00016892712163278263, + "loss": 0.7812, + "step": 5809 + }, + { + "epoch": 1.0343660968660968, + "grad_norm": 0.6341692805290222, + "learning_rate": 0.00016891697974126345, + "loss": 1.0658, + "step": 5810 + }, + { + "epoch": 1.0345441595441596, + "grad_norm": 0.6326245665550232, + "learning_rate": 0.00016890683649945922, + "loss": 1.0134, + "step": 5811 + }, + { + "epoch": 1.0347222222222223, + "grad_norm": 0.5729571580886841, + "learning_rate": 0.00016889669190756868, + "loss": 0.9139, + "step": 5812 + }, + { + "epoch": 1.0349002849002849, + "grad_norm": 0.5912853479385376, + "learning_rate": 0.00016888654596579054, + "loss": 1.122, + "step": 5813 + }, + { + "epoch": 1.0350783475783476, + "grad_norm": 0.8410450220108032, + "learning_rate": 0.00016887639867432368, + "loss": 1.3009, + "step": 5814 + }, + { + "epoch": 1.0352564102564104, + "grad_norm": 0.5416620969772339, + "learning_rate": 0.00016886625003336683, + "loss": 0.8751, + "step": 5815 + }, + { + "epoch": 1.0354344729344729, + "grad_norm": 0.6367851495742798, + "learning_rate": 0.0001688561000431189, + "loss": 0.956, + "step": 5816 + }, + { + "epoch": 1.0356125356125356, + "grad_norm": 0.4618827700614929, + "learning_rate": 0.0001688459487037787, + "loss": 0.5313, + "step": 5817 + }, + { + "epoch": 1.0357905982905984, + "grad_norm": 0.7139244079589844, + "learning_rate": 0.00016883579601554516, + "loss": 1.0787, + "step": 5818 + }, + { + "epoch": 1.035968660968661, + "grad_norm": 0.6896135210990906, + "learning_rate": 0.00016882564197861715, + "loss": 0.932, + "step": 5819 + }, + { + "epoch": 1.0361467236467237, + "grad_norm": 0.5889739394187927, + "learning_rate": 0.00016881548659319372, + "loss": 0.8852, + "step": 5820 + }, + { + "epoch": 1.0363247863247864, + "grad_norm": 0.5954701900482178, + "learning_rate": 0.00016880532985947375, + "loss": 0.8192, + "step": 5821 + }, + { + "epoch": 1.036502849002849, + "grad_norm": 0.6665091514587402, + "learning_rate": 0.00016879517177765627, + "loss": 0.9578, + "step": 5822 + }, + { + "epoch": 1.0366809116809117, + "grad_norm": 0.5990539789199829, + "learning_rate": 0.00016878501234794034, + "loss": 0.9797, + "step": 5823 + }, + { + "epoch": 1.0368589743589745, + "grad_norm": 0.596755862236023, + "learning_rate": 0.00016877485157052496, + "loss": 1.173, + "step": 5824 + }, + { + "epoch": 1.037037037037037, + "grad_norm": 0.544658362865448, + "learning_rate": 0.00016876468944560923, + "loss": 1.0742, + "step": 5825 + }, + { + "epoch": 1.0372150997150997, + "grad_norm": 0.5841910243034363, + "learning_rate": 0.00016875452597339225, + "loss": 1.029, + "step": 5826 + }, + { + "epoch": 1.0373931623931625, + "grad_norm": 0.6508592963218689, + "learning_rate": 0.00016874436115407317, + "loss": 0.9883, + "step": 5827 + }, + { + "epoch": 1.037571225071225, + "grad_norm": 0.590050458908081, + "learning_rate": 0.00016873419498785114, + "loss": 1.0713, + "step": 5828 + }, + { + "epoch": 1.0377492877492878, + "grad_norm": 0.5386307239532471, + "learning_rate": 0.00016872402747492534, + "loss": 1.0159, + "step": 5829 + }, + { + "epoch": 1.0379273504273505, + "grad_norm": 0.6173896193504333, + "learning_rate": 0.00016871385861549497, + "loss": 1.0056, + "step": 5830 + }, + { + "epoch": 1.038105413105413, + "grad_norm": 0.5377787351608276, + "learning_rate": 0.0001687036884097593, + "loss": 0.8708, + "step": 5831 + }, + { + "epoch": 1.0382834757834758, + "grad_norm": 0.5753569006919861, + "learning_rate": 0.00016869351685791756, + "loss": 1.0529, + "step": 5832 + }, + { + "epoch": 1.0384615384615385, + "grad_norm": 0.6085895299911499, + "learning_rate": 0.00016868334396016906, + "loss": 1.1017, + "step": 5833 + }, + { + "epoch": 1.038639601139601, + "grad_norm": 0.6320509910583496, + "learning_rate": 0.0001686731697167131, + "loss": 1.0543, + "step": 5834 + }, + { + "epoch": 1.0388176638176638, + "grad_norm": 0.5691760778427124, + "learning_rate": 0.00016866299412774907, + "loss": 0.9975, + "step": 5835 + }, + { + "epoch": 1.0389957264957266, + "grad_norm": 0.5990765690803528, + "learning_rate": 0.0001686528171934763, + "loss": 0.8776, + "step": 5836 + }, + { + "epoch": 1.039173789173789, + "grad_norm": 0.6650477647781372, + "learning_rate": 0.00016864263891409415, + "loss": 1.0652, + "step": 5837 + }, + { + "epoch": 1.0393518518518519, + "grad_norm": 0.6050353646278381, + "learning_rate": 0.00016863245928980212, + "loss": 0.9313, + "step": 5838 + }, + { + "epoch": 1.0395299145299146, + "grad_norm": 0.587505578994751, + "learning_rate": 0.0001686222783207996, + "loss": 0.9892, + "step": 5839 + }, + { + "epoch": 1.0397079772079771, + "grad_norm": 0.6310170292854309, + "learning_rate": 0.00016861209600728608, + "loss": 1.1045, + "step": 5840 + }, + { + "epoch": 1.03988603988604, + "grad_norm": 0.5683430433273315, + "learning_rate": 0.0001686019123494611, + "loss": 1.0507, + "step": 5841 + }, + { + "epoch": 1.0400641025641026, + "grad_norm": 0.6621488332748413, + "learning_rate": 0.00016859172734752414, + "loss": 0.9255, + "step": 5842 + }, + { + "epoch": 1.0402421652421652, + "grad_norm": 0.6197706460952759, + "learning_rate": 0.00016858154100167475, + "loss": 1.0031, + "step": 5843 + }, + { + "epoch": 1.040420227920228, + "grad_norm": 0.6805898547172546, + "learning_rate": 0.00016857135331211257, + "loss": 0.9901, + "step": 5844 + }, + { + "epoch": 1.0405982905982907, + "grad_norm": 0.5512405633926392, + "learning_rate": 0.00016856116427903714, + "loss": 1.0033, + "step": 5845 + }, + { + "epoch": 1.0407763532763532, + "grad_norm": 0.5643384456634521, + "learning_rate": 0.00016855097390264815, + "loss": 0.9136, + "step": 5846 + }, + { + "epoch": 1.040954415954416, + "grad_norm": 0.48351922631263733, + "learning_rate": 0.0001685407821831452, + "loss": 0.6163, + "step": 5847 + }, + { + "epoch": 1.0411324786324787, + "grad_norm": 0.6256039142608643, + "learning_rate": 0.00016853058912072802, + "loss": 0.9409, + "step": 5848 + }, + { + "epoch": 1.0413105413105412, + "grad_norm": 0.6539996862411499, + "learning_rate": 0.00016852039471559627, + "loss": 0.9367, + "step": 5849 + }, + { + "epoch": 1.041488603988604, + "grad_norm": 0.6192609667778015, + "learning_rate": 0.00016851019896794975, + "loss": 0.9631, + "step": 5850 + }, + { + "epoch": 1.0416666666666667, + "grad_norm": 0.613563060760498, + "learning_rate": 0.0001685000018779882, + "loss": 0.9132, + "step": 5851 + }, + { + "epoch": 1.0418447293447293, + "grad_norm": 0.6004200577735901, + "learning_rate": 0.0001684898034459114, + "loss": 1.1313, + "step": 5852 + }, + { + "epoch": 1.042022792022792, + "grad_norm": 0.6158567070960999, + "learning_rate": 0.0001684796036719192, + "loss": 1.0253, + "step": 5853 + }, + { + "epoch": 1.0422008547008548, + "grad_norm": 0.6362335085868835, + "learning_rate": 0.00016846940255621143, + "loss": 0.93, + "step": 5854 + }, + { + "epoch": 1.0423789173789173, + "grad_norm": 0.6148427128791809, + "learning_rate": 0.00016845920009898787, + "loss": 0.9122, + "step": 5855 + }, + { + "epoch": 1.04255698005698, + "grad_norm": 0.5119984149932861, + "learning_rate": 0.00016844899630044858, + "loss": 0.7954, + "step": 5856 + }, + { + "epoch": 1.0427350427350428, + "grad_norm": 0.571849524974823, + "learning_rate": 0.00016843879116079338, + "loss": 0.8588, + "step": 5857 + }, + { + "epoch": 1.0429131054131053, + "grad_norm": 0.6173384785652161, + "learning_rate": 0.00016842858468022221, + "loss": 1.0475, + "step": 5858 + }, + { + "epoch": 1.043091168091168, + "grad_norm": 0.566114068031311, + "learning_rate": 0.0001684183768589351, + "loss": 0.8485, + "step": 5859 + }, + { + "epoch": 1.0432692307692308, + "grad_norm": 0.653134286403656, + "learning_rate": 0.000168408167697132, + "loss": 0.9976, + "step": 5860 + }, + { + "epoch": 1.0434472934472934, + "grad_norm": 0.63815838098526, + "learning_rate": 0.00016839795719501296, + "loss": 0.7091, + "step": 5861 + }, + { + "epoch": 1.0436253561253561, + "grad_norm": 0.5109001994132996, + "learning_rate": 0.00016838774535277805, + "loss": 0.7668, + "step": 5862 + }, + { + "epoch": 1.0438034188034189, + "grad_norm": 0.6741907596588135, + "learning_rate": 0.0001683775321706273, + "loss": 1.0493, + "step": 5863 + }, + { + "epoch": 1.0439814814814814, + "grad_norm": 0.6006115674972534, + "learning_rate": 0.0001683673176487609, + "loss": 0.9784, + "step": 5864 + }, + { + "epoch": 1.0441595441595442, + "grad_norm": 0.5504778027534485, + "learning_rate": 0.0001683571017873789, + "loss": 0.9718, + "step": 5865 + }, + { + "epoch": 1.044337606837607, + "grad_norm": 0.5713102221488953, + "learning_rate": 0.00016834688458668148, + "loss": 1.12, + "step": 5866 + }, + { + "epoch": 1.0445156695156694, + "grad_norm": 0.7878454923629761, + "learning_rate": 0.00016833666604686886, + "loss": 1.1803, + "step": 5867 + }, + { + "epoch": 1.0446937321937322, + "grad_norm": 0.582697331905365, + "learning_rate": 0.00016832644616814122, + "loss": 0.943, + "step": 5868 + }, + { + "epoch": 1.044871794871795, + "grad_norm": 0.5300645232200623, + "learning_rate": 0.00016831622495069878, + "loss": 0.9087, + "step": 5869 + }, + { + "epoch": 1.0450498575498575, + "grad_norm": 0.5627666115760803, + "learning_rate": 0.00016830600239474186, + "loss": 1.081, + "step": 5870 + }, + { + "epoch": 1.0452279202279202, + "grad_norm": 0.6760496497154236, + "learning_rate": 0.0001682957785004707, + "loss": 1.1098, + "step": 5871 + }, + { + "epoch": 1.045405982905983, + "grad_norm": 0.6424084901809692, + "learning_rate": 0.00016828555326808565, + "loss": 0.9657, + "step": 5872 + }, + { + "epoch": 1.0455840455840455, + "grad_norm": 0.5523313283920288, + "learning_rate": 0.000168275326697787, + "loss": 1.0163, + "step": 5873 + }, + { + "epoch": 1.0457621082621082, + "grad_norm": 0.5582337975502014, + "learning_rate": 0.00016826509878977518, + "loss": 0.8825, + "step": 5874 + }, + { + "epoch": 1.045940170940171, + "grad_norm": 0.5603214502334595, + "learning_rate": 0.00016825486954425055, + "loss": 0.9032, + "step": 5875 + }, + { + "epoch": 1.0461182336182335, + "grad_norm": 0.5944222807884216, + "learning_rate": 0.00016824463896141355, + "loss": 0.9384, + "step": 5876 + }, + { + "epoch": 1.0462962962962963, + "grad_norm": 0.6220229268074036, + "learning_rate": 0.00016823440704146457, + "loss": 0.8962, + "step": 5877 + }, + { + "epoch": 1.046474358974359, + "grad_norm": 0.5607972145080566, + "learning_rate": 0.0001682241737846042, + "loss": 0.9385, + "step": 5878 + }, + { + "epoch": 1.0466524216524216, + "grad_norm": 0.6206870079040527, + "learning_rate": 0.00016821393919103282, + "loss": 1.0597, + "step": 5879 + }, + { + "epoch": 1.0468304843304843, + "grad_norm": 0.5126399993896484, + "learning_rate": 0.000168203703260951, + "loss": 0.9403, + "step": 5880 + }, + { + "epoch": 1.047008547008547, + "grad_norm": 0.6569282412528992, + "learning_rate": 0.00016819346599455929, + "loss": 0.8124, + "step": 5881 + }, + { + "epoch": 1.0471866096866096, + "grad_norm": 0.6670137047767639, + "learning_rate": 0.0001681832273920583, + "loss": 1.1927, + "step": 5882 + }, + { + "epoch": 1.0473646723646723, + "grad_norm": 0.5403243899345398, + "learning_rate": 0.00016817298745364862, + "loss": 0.8539, + "step": 5883 + }, + { + "epoch": 1.047542735042735, + "grad_norm": 0.5500505566596985, + "learning_rate": 0.00016816274617953086, + "loss": 1.1064, + "step": 5884 + }, + { + "epoch": 1.0477207977207976, + "grad_norm": 0.5482703447341919, + "learning_rate": 0.00016815250356990566, + "loss": 0.7276, + "step": 5885 + }, + { + "epoch": 1.0478988603988604, + "grad_norm": 0.6290771961212158, + "learning_rate": 0.00016814225962497373, + "loss": 0.9018, + "step": 5886 + }, + { + "epoch": 1.0480769230769231, + "grad_norm": 0.6404094696044922, + "learning_rate": 0.00016813201434493578, + "loss": 1.0638, + "step": 5887 + }, + { + "epoch": 1.0482549857549857, + "grad_norm": 0.5484994053840637, + "learning_rate": 0.0001681217677299926, + "loss": 1.0033, + "step": 5888 + }, + { + "epoch": 1.0484330484330484, + "grad_norm": 0.6474852561950684, + "learning_rate": 0.0001681115197803448, + "loss": 1.1017, + "step": 5889 + }, + { + "epoch": 1.0486111111111112, + "grad_norm": 0.6186243295669556, + "learning_rate": 0.0001681012704961933, + "loss": 0.9978, + "step": 5890 + }, + { + "epoch": 1.0487891737891737, + "grad_norm": 0.6244034767150879, + "learning_rate": 0.00016809101987773887, + "loss": 0.9906, + "step": 5891 + }, + { + "epoch": 1.0489672364672364, + "grad_norm": 0.5893426537513733, + "learning_rate": 0.00016808076792518235, + "loss": 0.9345, + "step": 5892 + }, + { + "epoch": 1.0491452991452992, + "grad_norm": 0.6283876299858093, + "learning_rate": 0.0001680705146387246, + "loss": 1.0041, + "step": 5893 + }, + { + "epoch": 1.0493233618233617, + "grad_norm": 0.6075255870819092, + "learning_rate": 0.00016806026001856656, + "loss": 1.0661, + "step": 5894 + }, + { + "epoch": 1.0495014245014245, + "grad_norm": 0.5350496768951416, + "learning_rate": 0.00016805000406490907, + "loss": 0.6789, + "step": 5895 + }, + { + "epoch": 1.0496794871794872, + "grad_norm": 0.5380373597145081, + "learning_rate": 0.00016803974677795312, + "loss": 0.8889, + "step": 5896 + }, + { + "epoch": 1.0498575498575498, + "grad_norm": 0.6145668029785156, + "learning_rate": 0.0001680294881578997, + "loss": 0.8952, + "step": 5897 + }, + { + "epoch": 1.0500356125356125, + "grad_norm": 0.5666532516479492, + "learning_rate": 0.00016801922820494972, + "loss": 0.9697, + "step": 5898 + }, + { + "epoch": 1.0502136752136753, + "grad_norm": 0.5352747440338135, + "learning_rate": 0.0001680089669193043, + "loss": 0.9619, + "step": 5899 + }, + { + "epoch": 1.0503917378917378, + "grad_norm": 0.5405527949333191, + "learning_rate": 0.00016799870430116444, + "loss": 0.8733, + "step": 5900 + }, + { + "epoch": 1.0505698005698005, + "grad_norm": 0.5936748385429382, + "learning_rate": 0.00016798844035073124, + "loss": 0.8746, + "step": 5901 + }, + { + "epoch": 1.0507478632478633, + "grad_norm": 0.539652943611145, + "learning_rate": 0.00016797817506820578, + "loss": 0.8743, + "step": 5902 + }, + { + "epoch": 1.0509259259259258, + "grad_norm": 0.644528865814209, + "learning_rate": 0.00016796790845378915, + "loss": 0.9251, + "step": 5903 + }, + { + "epoch": 1.0511039886039886, + "grad_norm": 0.5429201126098633, + "learning_rate": 0.00016795764050768258, + "loss": 0.747, + "step": 5904 + }, + { + "epoch": 1.0512820512820513, + "grad_norm": 0.6432006359100342, + "learning_rate": 0.00016794737123008725, + "loss": 0.9166, + "step": 5905 + }, + { + "epoch": 1.0514601139601139, + "grad_norm": 0.6084117293357849, + "learning_rate": 0.00016793710062120427, + "loss": 1.0778, + "step": 5906 + }, + { + "epoch": 1.0516381766381766, + "grad_norm": 0.5351580381393433, + "learning_rate": 0.00016792682868123495, + "loss": 0.9124, + "step": 5907 + }, + { + "epoch": 1.0518162393162394, + "grad_norm": 0.7078854441642761, + "learning_rate": 0.00016791655541038053, + "loss": 1.1209, + "step": 5908 + }, + { + "epoch": 1.051994301994302, + "grad_norm": 0.5943832993507385, + "learning_rate": 0.0001679062808088423, + "loss": 0.9077, + "step": 5909 + }, + { + "epoch": 1.0521723646723646, + "grad_norm": 0.5216894745826721, + "learning_rate": 0.00016789600487682156, + "loss": 0.9866, + "step": 5910 + }, + { + "epoch": 1.0523504273504274, + "grad_norm": 0.738451361656189, + "learning_rate": 0.00016788572761451963, + "loss": 1.1611, + "step": 5911 + }, + { + "epoch": 1.05252849002849, + "grad_norm": 0.6411251425743103, + "learning_rate": 0.00016787544902213791, + "loss": 1.1481, + "step": 5912 + }, + { + "epoch": 1.0527065527065527, + "grad_norm": 0.6768319010734558, + "learning_rate": 0.00016786516909987774, + "loss": 0.8614, + "step": 5913 + }, + { + "epoch": 1.0528846153846154, + "grad_norm": 0.5838070511817932, + "learning_rate": 0.0001678548878479406, + "loss": 0.9719, + "step": 5914 + }, + { + "epoch": 1.0530626780626782, + "grad_norm": 0.541522741317749, + "learning_rate": 0.00016784460526652784, + "loss": 0.767, + "step": 5915 + }, + { + "epoch": 1.0532407407407407, + "grad_norm": 0.6064762473106384, + "learning_rate": 0.000167834321355841, + "loss": 1.0792, + "step": 5916 + }, + { + "epoch": 1.0534188034188035, + "grad_norm": 0.5515492558479309, + "learning_rate": 0.00016782403611608152, + "loss": 0.7897, + "step": 5917 + }, + { + "epoch": 1.0535968660968662, + "grad_norm": 0.6326262950897217, + "learning_rate": 0.000167813749547451, + "loss": 0.9279, + "step": 5918 + }, + { + "epoch": 1.0537749287749287, + "grad_norm": 0.6262009739875793, + "learning_rate": 0.0001678034616501509, + "loss": 0.9752, + "step": 5919 + }, + { + "epoch": 1.0539529914529915, + "grad_norm": 0.6049023270606995, + "learning_rate": 0.00016779317242438278, + "loss": 0.9167, + "step": 5920 + }, + { + "epoch": 1.0541310541310542, + "grad_norm": 0.6286031007766724, + "learning_rate": 0.0001677828818703483, + "loss": 1.1277, + "step": 5921 + }, + { + "epoch": 1.0543091168091168, + "grad_norm": 0.662086009979248, + "learning_rate": 0.00016777258998824907, + "loss": 1.0824, + "step": 5922 + }, + { + "epoch": 1.0544871794871795, + "grad_norm": 0.5358783006668091, + "learning_rate": 0.00016776229677828672, + "loss": 0.825, + "step": 5923 + }, + { + "epoch": 1.0546652421652423, + "grad_norm": 0.490326464176178, + "learning_rate": 0.00016775200224066294, + "loss": 0.7916, + "step": 5924 + }, + { + "epoch": 1.0548433048433048, + "grad_norm": 0.5940443277359009, + "learning_rate": 0.0001677417063755794, + "loss": 1.0121, + "step": 5925 + }, + { + "epoch": 1.0550213675213675, + "grad_norm": 0.5974507927894592, + "learning_rate": 0.00016773140918323787, + "loss": 0.7629, + "step": 5926 + }, + { + "epoch": 1.0551994301994303, + "grad_norm": 0.5747174024581909, + "learning_rate": 0.00016772111066384003, + "loss": 0.9373, + "step": 5927 + }, + { + "epoch": 1.0553774928774928, + "grad_norm": 0.5998024940490723, + "learning_rate": 0.00016771081081758772, + "loss": 0.8543, + "step": 5928 + }, + { + "epoch": 1.0555555555555556, + "grad_norm": 0.5771155953407288, + "learning_rate": 0.00016770050964468275, + "loss": 0.9108, + "step": 5929 + }, + { + "epoch": 1.0557336182336183, + "grad_norm": 0.5695661306381226, + "learning_rate": 0.00016769020714532692, + "loss": 0.8055, + "step": 5930 + }, + { + "epoch": 1.0559116809116809, + "grad_norm": 0.6164212226867676, + "learning_rate": 0.0001676799033197221, + "loss": 1.0917, + "step": 5931 + }, + { + "epoch": 1.0560897435897436, + "grad_norm": 0.6092487573623657, + "learning_rate": 0.00016766959816807018, + "loss": 0.9276, + "step": 5932 + }, + { + "epoch": 1.0562678062678064, + "grad_norm": 0.5595401525497437, + "learning_rate": 0.00016765929169057305, + "loss": 0.9435, + "step": 5933 + }, + { + "epoch": 1.056445868945869, + "grad_norm": 0.5875109434127808, + "learning_rate": 0.00016764898388743263, + "loss": 0.959, + "step": 5934 + }, + { + "epoch": 1.0566239316239316, + "grad_norm": 0.6045668721199036, + "learning_rate": 0.00016763867475885088, + "loss": 0.8636, + "step": 5935 + }, + { + "epoch": 1.0568019943019944, + "grad_norm": 0.6088171005249023, + "learning_rate": 0.00016762836430502987, + "loss": 0.6807, + "step": 5936 + }, + { + "epoch": 1.056980056980057, + "grad_norm": 0.6293274760246277, + "learning_rate": 0.00016761805252617148, + "loss": 1.042, + "step": 5937 + }, + { + "epoch": 1.0571581196581197, + "grad_norm": 0.588472843170166, + "learning_rate": 0.00016760773942247785, + "loss": 0.8896, + "step": 5938 + }, + { + "epoch": 1.0573361823361824, + "grad_norm": 0.4412326216697693, + "learning_rate": 0.000167597424994151, + "loss": 0.6727, + "step": 5939 + }, + { + "epoch": 1.057514245014245, + "grad_norm": 0.6086825132369995, + "learning_rate": 0.00016758710924139302, + "loss": 0.9908, + "step": 5940 + }, + { + "epoch": 1.0576923076923077, + "grad_norm": 0.6424705386161804, + "learning_rate": 0.00016757679216440608, + "loss": 1.0182, + "step": 5941 + }, + { + "epoch": 1.0578703703703705, + "grad_norm": 0.6610676050186157, + "learning_rate": 0.00016756647376339222, + "loss": 0.9645, + "step": 5942 + }, + { + "epoch": 1.058048433048433, + "grad_norm": 0.598292887210846, + "learning_rate": 0.0001675561540385537, + "loss": 0.9694, + "step": 5943 + }, + { + "epoch": 1.0582264957264957, + "grad_norm": 0.6941167116165161, + "learning_rate": 0.00016754583299009266, + "loss": 1.0786, + "step": 5944 + }, + { + "epoch": 1.0584045584045585, + "grad_norm": 0.6543232798576355, + "learning_rate": 0.00016753551061821133, + "loss": 1.0488, + "step": 5945 + }, + { + "epoch": 1.058582621082621, + "grad_norm": 0.606159508228302, + "learning_rate": 0.000167525186923112, + "loss": 0.9448, + "step": 5946 + }, + { + "epoch": 1.0587606837606838, + "grad_norm": 0.5051791071891785, + "learning_rate": 0.00016751486190499685, + "loss": 0.7485, + "step": 5947 + }, + { + "epoch": 1.0589387464387465, + "grad_norm": 0.6459367275238037, + "learning_rate": 0.00016750453556406826, + "loss": 1.0055, + "step": 5948 + }, + { + "epoch": 1.059116809116809, + "grad_norm": 0.551591157913208, + "learning_rate": 0.00016749420790052852, + "loss": 0.9717, + "step": 5949 + }, + { + "epoch": 1.0592948717948718, + "grad_norm": 0.5899214148521423, + "learning_rate": 0.00016748387891458, + "loss": 0.7774, + "step": 5950 + }, + { + "epoch": 1.0594729344729346, + "grad_norm": 0.582379162311554, + "learning_rate": 0.00016747354860642503, + "loss": 0.953, + "step": 5951 + }, + { + "epoch": 1.059650997150997, + "grad_norm": 0.6035816073417664, + "learning_rate": 0.00016746321697626605, + "loss": 1.1175, + "step": 5952 + }, + { + "epoch": 1.0598290598290598, + "grad_norm": 0.6476401686668396, + "learning_rate": 0.00016745288402430548, + "loss": 0.9448, + "step": 5953 + }, + { + "epoch": 1.0600071225071226, + "grad_norm": 0.6126405596733093, + "learning_rate": 0.00016744254975074578, + "loss": 0.882, + "step": 5954 + }, + { + "epoch": 1.0601851851851851, + "grad_norm": 0.5333579182624817, + "learning_rate": 0.0001674322141557894, + "loss": 0.9539, + "step": 5955 + }, + { + "epoch": 1.0603632478632479, + "grad_norm": 0.6085022687911987, + "learning_rate": 0.0001674218772396389, + "loss": 1.0028, + "step": 5956 + }, + { + "epoch": 1.0605413105413106, + "grad_norm": 0.5809528827667236, + "learning_rate": 0.0001674115390024967, + "loss": 0.84, + "step": 5957 + }, + { + "epoch": 1.0607193732193732, + "grad_norm": 0.5820229649543762, + "learning_rate": 0.00016740119944456548, + "loss": 0.9563, + "step": 5958 + }, + { + "epoch": 1.060897435897436, + "grad_norm": 0.6349015831947327, + "learning_rate": 0.00016739085856604775, + "loss": 0.9739, + "step": 5959 + }, + { + "epoch": 1.0610754985754987, + "grad_norm": 0.6346020102500916, + "learning_rate": 0.00016738051636714616, + "loss": 0.907, + "step": 5960 + }, + { + "epoch": 1.0612535612535612, + "grad_norm": 0.5850573778152466, + "learning_rate": 0.0001673701728480633, + "loss": 1.0688, + "step": 5961 + }, + { + "epoch": 1.061431623931624, + "grad_norm": 0.6258122324943542, + "learning_rate": 0.00016735982800900184, + "loss": 0.9997, + "step": 5962 + }, + { + "epoch": 1.0616096866096867, + "grad_norm": 0.6744239330291748, + "learning_rate": 0.00016734948185016452, + "loss": 0.9431, + "step": 5963 + }, + { + "epoch": 1.0617877492877492, + "grad_norm": 0.5769457817077637, + "learning_rate": 0.000167339134371754, + "loss": 0.9658, + "step": 5964 + }, + { + "epoch": 1.061965811965812, + "grad_norm": 0.6385112404823303, + "learning_rate": 0.000167328785573973, + "loss": 1.0199, + "step": 5965 + }, + { + "epoch": 1.0621438746438747, + "grad_norm": 0.536522388458252, + "learning_rate": 0.00016731843545702435, + "loss": 0.8496, + "step": 5966 + }, + { + "epoch": 1.0623219373219372, + "grad_norm": 0.5978497862815857, + "learning_rate": 0.00016730808402111075, + "loss": 0.8536, + "step": 5967 + }, + { + "epoch": 1.0625, + "grad_norm": 0.6091681122779846, + "learning_rate": 0.0001672977312664351, + "loss": 1.0241, + "step": 5968 + }, + { + "epoch": 1.0626780626780628, + "grad_norm": 0.5807273387908936, + "learning_rate": 0.0001672873771932002, + "loss": 1.0522, + "step": 5969 + }, + { + "epoch": 1.0628561253561253, + "grad_norm": 0.6511965990066528, + "learning_rate": 0.0001672770218016089, + "loss": 0.8908, + "step": 5970 + }, + { + "epoch": 1.063034188034188, + "grad_norm": 0.6241721510887146, + "learning_rate": 0.00016726666509186416, + "loss": 0.9854, + "step": 5971 + }, + { + "epoch": 1.0632122507122508, + "grad_norm": 0.6112468242645264, + "learning_rate": 0.0001672563070641688, + "loss": 1.0091, + "step": 5972 + }, + { + "epoch": 1.0633903133903133, + "grad_norm": 0.6135509014129639, + "learning_rate": 0.00016724594771872587, + "loss": 0.8891, + "step": 5973 + }, + { + "epoch": 1.063568376068376, + "grad_norm": 0.608384370803833, + "learning_rate": 0.00016723558705573823, + "loss": 1.017, + "step": 5974 + }, + { + "epoch": 1.0637464387464388, + "grad_norm": 0.6578485369682312, + "learning_rate": 0.00016722522507540895, + "loss": 0.9165, + "step": 5975 + }, + { + "epoch": 1.0639245014245013, + "grad_norm": 0.562588095664978, + "learning_rate": 0.00016721486177794106, + "loss": 0.7989, + "step": 5976 + }, + { + "epoch": 1.064102564102564, + "grad_norm": 0.5541409254074097, + "learning_rate": 0.00016720449716353753, + "loss": 0.8917, + "step": 5977 + }, + { + "epoch": 1.0642806267806268, + "grad_norm": 0.551167905330658, + "learning_rate": 0.0001671941312324015, + "loss": 0.824, + "step": 5978 + }, + { + "epoch": 1.0644586894586894, + "grad_norm": 0.6280582547187805, + "learning_rate": 0.0001671837639847361, + "loss": 0.9708, + "step": 5979 + }, + { + "epoch": 1.0646367521367521, + "grad_norm": 0.6389226913452148, + "learning_rate": 0.00016717339542074436, + "loss": 1.0081, + "step": 5980 + }, + { + "epoch": 1.0648148148148149, + "grad_norm": 0.6677889823913574, + "learning_rate": 0.0001671630255406295, + "loss": 1.2709, + "step": 5981 + }, + { + "epoch": 1.0649928774928774, + "grad_norm": 0.5748161673545837, + "learning_rate": 0.00016715265434459465, + "loss": 0.9157, + "step": 5982 + }, + { + "epoch": 1.0651709401709402, + "grad_norm": 0.6677651405334473, + "learning_rate": 0.00016714228183284304, + "loss": 1.1097, + "step": 5983 + }, + { + "epoch": 1.065349002849003, + "grad_norm": 0.6253604292869568, + "learning_rate": 0.0001671319080055779, + "loss": 0.9819, + "step": 5984 + }, + { + "epoch": 1.0655270655270654, + "grad_norm": 0.5548844337463379, + "learning_rate": 0.0001671215328630025, + "loss": 0.9324, + "step": 5985 + }, + { + "epoch": 1.0657051282051282, + "grad_norm": 0.622062623500824, + "learning_rate": 0.00016711115640532004, + "loss": 0.8749, + "step": 5986 + }, + { + "epoch": 1.065883190883191, + "grad_norm": 0.6496043801307678, + "learning_rate": 0.00016710077863273394, + "loss": 1.0642, + "step": 5987 + }, + { + "epoch": 1.0660612535612535, + "grad_norm": 0.6140534281730652, + "learning_rate": 0.00016709039954544746, + "loss": 0.8928, + "step": 5988 + }, + { + "epoch": 1.0662393162393162, + "grad_norm": 0.6387218236923218, + "learning_rate": 0.00016708001914366393, + "loss": 0.9525, + "step": 5989 + }, + { + "epoch": 1.066417378917379, + "grad_norm": 0.6119858026504517, + "learning_rate": 0.0001670696374275868, + "loss": 0.8663, + "step": 5990 + }, + { + "epoch": 1.0665954415954415, + "grad_norm": 0.6722040772438049, + "learning_rate": 0.00016705925439741947, + "loss": 1.1173, + "step": 5991 + }, + { + "epoch": 1.0667735042735043, + "grad_norm": 0.8226081132888794, + "learning_rate": 0.00016704887005336534, + "loss": 1.0572, + "step": 5992 + }, + { + "epoch": 1.066951566951567, + "grad_norm": 0.7248596549034119, + "learning_rate": 0.00016703848439562785, + "loss": 1.0493, + "step": 5993 + }, + { + "epoch": 1.0671296296296295, + "grad_norm": 0.7185787558555603, + "learning_rate": 0.00016702809742441058, + "loss": 1.1366, + "step": 5994 + }, + { + "epoch": 1.0673076923076923, + "grad_norm": 0.6118780970573425, + "learning_rate": 0.00016701770913991694, + "loss": 0.9557, + "step": 5995 + }, + { + "epoch": 1.067485754985755, + "grad_norm": 0.6472596526145935, + "learning_rate": 0.0001670073195423505, + "loss": 0.9977, + "step": 5996 + }, + { + "epoch": 1.0676638176638176, + "grad_norm": 0.7110133767127991, + "learning_rate": 0.00016699692863191484, + "loss": 1.1932, + "step": 5997 + }, + { + "epoch": 1.0678418803418803, + "grad_norm": 0.5827305912971497, + "learning_rate": 0.00016698653640881354, + "loss": 0.7641, + "step": 5998 + }, + { + "epoch": 1.068019943019943, + "grad_norm": 0.527208149433136, + "learning_rate": 0.00016697614287325017, + "loss": 0.7683, + "step": 5999 + }, + { + "epoch": 1.0681980056980056, + "grad_norm": 0.6680626273155212, + "learning_rate": 0.00016696574802542848, + "loss": 1.1748, + "step": 6000 + }, + { + "epoch": 1.0683760683760684, + "grad_norm": 0.5947227478027344, + "learning_rate": 0.00016695535186555204, + "loss": 1.0894, + "step": 6001 + }, + { + "epoch": 1.068554131054131, + "grad_norm": 0.5828250646591187, + "learning_rate": 0.00016694495439382456, + "loss": 0.9895, + "step": 6002 + }, + { + "epoch": 1.0687321937321936, + "grad_norm": 0.5897728204727173, + "learning_rate": 0.00016693455561044978, + "loss": 0.9686, + "step": 6003 + }, + { + "epoch": 1.0689102564102564, + "grad_norm": 0.5441751480102539, + "learning_rate": 0.0001669241555156314, + "loss": 0.8948, + "step": 6004 + }, + { + "epoch": 1.0690883190883191, + "grad_norm": 0.694199800491333, + "learning_rate": 0.00016691375410957324, + "loss": 1.0824, + "step": 6005 + }, + { + "epoch": 1.0692663817663817, + "grad_norm": 0.6077630519866943, + "learning_rate": 0.00016690335139247906, + "loss": 1.0931, + "step": 6006 + }, + { + "epoch": 1.0694444444444444, + "grad_norm": 0.6558539867401123, + "learning_rate": 0.0001668929473645527, + "loss": 1.0099, + "step": 6007 + }, + { + "epoch": 1.0696225071225072, + "grad_norm": 0.5722812414169312, + "learning_rate": 0.00016688254202599798, + "loss": 0.7999, + "step": 6008 + }, + { + "epoch": 1.0698005698005697, + "grad_norm": 0.5915400981903076, + "learning_rate": 0.0001668721353770188, + "loss": 0.7866, + "step": 6009 + }, + { + "epoch": 1.0699786324786325, + "grad_norm": 0.5290952324867249, + "learning_rate": 0.00016686172741781901, + "loss": 0.793, + "step": 6010 + }, + { + "epoch": 1.0701566951566952, + "grad_norm": 0.5501774549484253, + "learning_rate": 0.00016685131814860263, + "loss": 0.8775, + "step": 6011 + }, + { + "epoch": 1.0703347578347577, + "grad_norm": 0.6192594766616821, + "learning_rate": 0.00016684090756957347, + "loss": 1.1686, + "step": 6012 + }, + { + "epoch": 1.0705128205128205, + "grad_norm": 0.6640267968177795, + "learning_rate": 0.00016683049568093561, + "loss": 1.1789, + "step": 6013 + }, + { + "epoch": 1.0706908831908832, + "grad_norm": 0.552893877029419, + "learning_rate": 0.00016682008248289303, + "loss": 0.7957, + "step": 6014 + }, + { + "epoch": 1.0708689458689458, + "grad_norm": 0.6406302452087402, + "learning_rate": 0.00016680966797564972, + "loss": 1.1174, + "step": 6015 + }, + { + "epoch": 1.0710470085470085, + "grad_norm": Infinity, + "learning_rate": 0.00016680966797564972, + "loss": 0.9168, + "step": 6016 + }, + { + "epoch": 1.0712250712250713, + "grad_norm": 0.6384762525558472, + "learning_rate": 0.00016679925215940975, + "loss": 0.9831, + "step": 6017 + }, + { + "epoch": 1.071403133903134, + "grad_norm": 0.5906224846839905, + "learning_rate": 0.0001667888350343772, + "loss": 0.9167, + "step": 6018 + }, + { + "epoch": 1.0715811965811965, + "grad_norm": 0.658044695854187, + "learning_rate": 0.00016677841660075617, + "loss": 1.0075, + "step": 6019 + }, + { + "epoch": 1.0717592592592593, + "grad_norm": 0.6313242316246033, + "learning_rate": 0.00016676799685875078, + "loss": 0.8551, + "step": 6020 + }, + { + "epoch": 1.0719373219373218, + "grad_norm": 0.5891841053962708, + "learning_rate": 0.00016675757580856518, + "loss": 0.8475, + "step": 6021 + }, + { + "epoch": 1.0721153846153846, + "grad_norm": 0.581317126750946, + "learning_rate": 0.00016674715345040358, + "loss": 0.9308, + "step": 6022 + }, + { + "epoch": 1.0722934472934473, + "grad_norm": 0.5952537655830383, + "learning_rate": 0.00016673672978447017, + "loss": 0.9104, + "step": 6023 + }, + { + "epoch": 1.07247150997151, + "grad_norm": 0.5934227705001831, + "learning_rate": 0.00016672630481096915, + "loss": 0.9882, + "step": 6024 + }, + { + "epoch": 1.0726495726495726, + "grad_norm": 0.5867539048194885, + "learning_rate": 0.00016671587853010482, + "loss": 1.0186, + "step": 6025 + }, + { + "epoch": 1.0728276353276354, + "grad_norm": 0.6002280116081238, + "learning_rate": 0.00016670545094208143, + "loss": 0.92, + "step": 6026 + }, + { + "epoch": 1.073005698005698, + "grad_norm": 0.6261683702468872, + "learning_rate": 0.0001666950220471033, + "loss": 0.9293, + "step": 6027 + }, + { + "epoch": 1.0731837606837606, + "grad_norm": 0.6128147840499878, + "learning_rate": 0.00016668459184537477, + "loss": 1.0787, + "step": 6028 + }, + { + "epoch": 1.0733618233618234, + "grad_norm": 0.62148118019104, + "learning_rate": 0.00016667416033710016, + "loss": 0.8843, + "step": 6029 + }, + { + "epoch": 1.0735398860398861, + "grad_norm": 0.7166166305541992, + "learning_rate": 0.0001666637275224839, + "loss": 0.8877, + "step": 6030 + }, + { + "epoch": 1.0737179487179487, + "grad_norm": 0.5275574922561646, + "learning_rate": 0.0001666532934017304, + "loss": 0.9604, + "step": 6031 + }, + { + "epoch": 1.0738960113960114, + "grad_norm": 0.8132784962654114, + "learning_rate": 0.00016664285797504406, + "loss": 1.0203, + "step": 6032 + }, + { + "epoch": 1.074074074074074, + "grad_norm": 0.5887695550918579, + "learning_rate": 0.00016663242124262935, + "loss": 0.8819, + "step": 6033 + }, + { + "epoch": 1.0742521367521367, + "grad_norm": 0.5552900433540344, + "learning_rate": 0.00016662198320469078, + "loss": 0.7542, + "step": 6034 + }, + { + "epoch": 1.0744301994301995, + "grad_norm": 0.6228970885276794, + "learning_rate": 0.0001666115438614328, + "loss": 1.0362, + "step": 6035 + }, + { + "epoch": 1.0746082621082622, + "grad_norm": 0.7193471789360046, + "learning_rate": 0.00016660110321306003, + "loss": 1.3073, + "step": 6036 + }, + { + "epoch": 1.0747863247863247, + "grad_norm": 0.6167412996292114, + "learning_rate": 0.000166590661259777, + "loss": 0.941, + "step": 6037 + }, + { + "epoch": 1.0749643874643875, + "grad_norm": 0.5716922879219055, + "learning_rate": 0.00016658021800178827, + "loss": 0.83, + "step": 6038 + }, + { + "epoch": 1.0751424501424502, + "grad_norm": 0.6404047012329102, + "learning_rate": 0.00016656977343929848, + "loss": 1.0617, + "step": 6039 + }, + { + "epoch": 1.0753205128205128, + "grad_norm": 0.531395435333252, + "learning_rate": 0.00016655932757251226, + "loss": 0.7785, + "step": 6040 + }, + { + "epoch": 1.0754985754985755, + "grad_norm": 0.6468462347984314, + "learning_rate": 0.0001665488804016343, + "loss": 0.7893, + "step": 6041 + }, + { + "epoch": 1.0756766381766383, + "grad_norm": 0.6539653539657593, + "learning_rate": 0.00016653843192686925, + "loss": 1.1011, + "step": 6042 + }, + { + "epoch": 1.0758547008547008, + "grad_norm": 0.630107045173645, + "learning_rate": 0.0001665279821484219, + "loss": 0.9262, + "step": 6043 + }, + { + "epoch": 1.0760327635327636, + "grad_norm": 0.5875992774963379, + "learning_rate": 0.00016651753106649688, + "loss": 1.0501, + "step": 6044 + }, + { + "epoch": 1.0762108262108263, + "grad_norm": 0.573428750038147, + "learning_rate": 0.00016650707868129904, + "loss": 1.0672, + "step": 6045 + }, + { + "epoch": 1.0763888888888888, + "grad_norm": 0.6215469241142273, + "learning_rate": 0.00016649662499303316, + "loss": 0.868, + "step": 6046 + }, + { + "epoch": 1.0765669515669516, + "grad_norm": 0.6666893362998962, + "learning_rate": 0.00016648617000190402, + "loss": 1.0965, + "step": 6047 + }, + { + "epoch": 1.0767450142450143, + "grad_norm": 0.8343498706817627, + "learning_rate": 0.00016647571370811653, + "loss": 1.2302, + "step": 6048 + }, + { + "epoch": 1.0769230769230769, + "grad_norm": 0.591147780418396, + "learning_rate": 0.0001664652561118755, + "loss": 0.9698, + "step": 6049 + }, + { + "epoch": 1.0771011396011396, + "grad_norm": 0.573375940322876, + "learning_rate": 0.00016645479721338584, + "loss": 0.8798, + "step": 6050 + }, + { + "epoch": 1.0772792022792024, + "grad_norm": 0.4956737160682678, + "learning_rate": 0.00016644433701285246, + "loss": 0.6523, + "step": 6051 + }, + { + "epoch": 1.077457264957265, + "grad_norm": 0.6896619200706482, + "learning_rate": 0.00016643387551048034, + "loss": 0.8911, + "step": 6052 + }, + { + "epoch": 1.0776353276353277, + "grad_norm": 0.5820416808128357, + "learning_rate": 0.00016642341270647445, + "loss": 1.1486, + "step": 6053 + }, + { + "epoch": 1.0778133903133904, + "grad_norm": 0.611132025718689, + "learning_rate": 0.00016641294860103976, + "loss": 1.0705, + "step": 6054 + }, + { + "epoch": 1.077991452991453, + "grad_norm": 0.6705698370933533, + "learning_rate": 0.00016640248319438133, + "loss": 0.9826, + "step": 6055 + }, + { + "epoch": 1.0781695156695157, + "grad_norm": 0.5987013578414917, + "learning_rate": 0.00016639201648670416, + "loss": 1.0409, + "step": 6056 + }, + { + "epoch": 1.0783475783475784, + "grad_norm": 0.6707149744033813, + "learning_rate": 0.00016638154847821332, + "loss": 1.1332, + "step": 6057 + }, + { + "epoch": 1.078525641025641, + "grad_norm": 0.6400678157806396, + "learning_rate": 0.00016637107916911393, + "loss": 1.2559, + "step": 6058 + }, + { + "epoch": 1.0787037037037037, + "grad_norm": 0.6370311379432678, + "learning_rate": 0.00016636060855961115, + "loss": 0.9752, + "step": 6059 + }, + { + "epoch": 1.0788817663817665, + "grad_norm": 0.6116052269935608, + "learning_rate": 0.00016635013664991012, + "loss": 0.8364, + "step": 6060 + }, + { + "epoch": 1.079059829059829, + "grad_norm": 0.7932127714157104, + "learning_rate": 0.00016633966344021593, + "loss": 0.939, + "step": 6061 + }, + { + "epoch": 1.0792378917378918, + "grad_norm": 0.576249897480011, + "learning_rate": 0.00016632918893073385, + "loss": 0.8911, + "step": 6062 + }, + { + "epoch": 1.0794159544159545, + "grad_norm": 0.5456888675689697, + "learning_rate": 0.00016631871312166915, + "loss": 0.8646, + "step": 6063 + }, + { + "epoch": 1.079594017094017, + "grad_norm": 0.717522919178009, + "learning_rate": 0.000166308236013227, + "loss": 1.0814, + "step": 6064 + }, + { + "epoch": 1.0797720797720798, + "grad_norm": 0.6637256145477295, + "learning_rate": 0.0001662977576056127, + "loss": 1.22, + "step": 6065 + }, + { + "epoch": 1.0799501424501425, + "grad_norm": 0.5846666693687439, + "learning_rate": 0.0001662872778990316, + "loss": 1.1745, + "step": 6066 + }, + { + "epoch": 1.080128205128205, + "grad_norm": 0.6611326336860657, + "learning_rate": 0.00016627679689368895, + "loss": 1.1262, + "step": 6067 + }, + { + "epoch": 1.0803062678062678, + "grad_norm": 0.6022892594337463, + "learning_rate": 0.00016626631458979015, + "loss": 0.9741, + "step": 6068 + }, + { + "epoch": 1.0804843304843306, + "grad_norm": 0.5862685441970825, + "learning_rate": 0.00016625583098754058, + "loss": 0.914, + "step": 6069 + }, + { + "epoch": 1.080662393162393, + "grad_norm": 0.7089241147041321, + "learning_rate": 0.00016624534608714563, + "loss": 1.0614, + "step": 6070 + }, + { + "epoch": 1.0808404558404558, + "grad_norm": 0.5286028981208801, + "learning_rate": 0.00016623485988881076, + "loss": 0.8756, + "step": 6071 + }, + { + "epoch": 1.0810185185185186, + "grad_norm": 0.6437101364135742, + "learning_rate": 0.00016622437239274137, + "loss": 0.7222, + "step": 6072 + }, + { + "epoch": 1.0811965811965811, + "grad_norm": 0.6197740435600281, + "learning_rate": 0.000166213883599143, + "loss": 0.7876, + "step": 6073 + }, + { + "epoch": 1.0813746438746439, + "grad_norm": 0.5889328122138977, + "learning_rate": 0.0001662033935082211, + "loss": 0.9587, + "step": 6074 + }, + { + "epoch": 1.0815527065527066, + "grad_norm": 0.5353847742080688, + "learning_rate": 0.00016619290212018125, + "loss": 0.8664, + "step": 6075 + }, + { + "epoch": 1.0817307692307692, + "grad_norm": 0.7202061414718628, + "learning_rate": 0.00016618240943522898, + "loss": 1.0429, + "step": 6076 + }, + { + "epoch": 1.081908831908832, + "grad_norm": 0.5831515192985535, + "learning_rate": 0.0001661719154535699, + "loss": 1.0323, + "step": 6077 + }, + { + "epoch": 1.0820868945868947, + "grad_norm": 0.6270500421524048, + "learning_rate": 0.00016616142017540953, + "loss": 0.9272, + "step": 6078 + }, + { + "epoch": 1.0822649572649572, + "grad_norm": 0.6064695119857788, + "learning_rate": 0.00016615092360095364, + "loss": 1.0629, + "step": 6079 + }, + { + "epoch": 1.08244301994302, + "grad_norm": 0.5578122138977051, + "learning_rate": 0.00016614042573040777, + "loss": 0.8601, + "step": 6080 + }, + { + "epoch": 1.0826210826210827, + "grad_norm": 0.5920688509941101, + "learning_rate": 0.0001661299265639777, + "loss": 1.0082, + "step": 6081 + }, + { + "epoch": 1.0827991452991452, + "grad_norm": 0.6191682815551758, + "learning_rate": 0.0001661194261018691, + "loss": 0.9645, + "step": 6082 + }, + { + "epoch": 1.082977207977208, + "grad_norm": 0.6403279304504395, + "learning_rate": 0.00016610892434428765, + "loss": 0.9263, + "step": 6083 + }, + { + "epoch": 1.0831552706552707, + "grad_norm": 0.579502284526825, + "learning_rate": 0.00016609842129143915, + "loss": 0.8997, + "step": 6084 + }, + { + "epoch": 1.0833333333333333, + "grad_norm": 0.5831437706947327, + "learning_rate": 0.00016608791694352944, + "loss": 1.0703, + "step": 6085 + }, + { + "epoch": 1.083511396011396, + "grad_norm": 0.6188452243804932, + "learning_rate": 0.00016607741130076424, + "loss": 0.8856, + "step": 6086 + }, + { + "epoch": 1.0836894586894588, + "grad_norm": 0.7413692474365234, + "learning_rate": 0.00016606690436334946, + "loss": 1.1995, + "step": 6087 + }, + { + "epoch": 1.0838675213675213, + "grad_norm": 0.5552099347114563, + "learning_rate": 0.00016605639613149093, + "loss": 0.8514, + "step": 6088 + }, + { + "epoch": 1.084045584045584, + "grad_norm": 0.5906503200531006, + "learning_rate": 0.00016604588660539452, + "loss": 0.9431, + "step": 6089 + }, + { + "epoch": 1.0842236467236468, + "grad_norm": 0.5326111316680908, + "learning_rate": 0.0001660353757852662, + "loss": 0.8306, + "step": 6090 + }, + { + "epoch": 1.0844017094017093, + "grad_norm": 0.7273091673851013, + "learning_rate": 0.0001660248636713118, + "loss": 1.1109, + "step": 6091 + }, + { + "epoch": 1.084579772079772, + "grad_norm": 0.66513592004776, + "learning_rate": 0.00016601435026373737, + "loss": 1.0621, + "step": 6092 + }, + { + "epoch": 1.0847578347578348, + "grad_norm": 0.6470831632614136, + "learning_rate": 0.00016600383556274892, + "loss": 1.1075, + "step": 6093 + }, + { + "epoch": 1.0849358974358974, + "grad_norm": 0.6308658719062805, + "learning_rate": 0.0001659933195685524, + "loss": 0.9832, + "step": 6094 + }, + { + "epoch": 1.08511396011396, + "grad_norm": 0.6569336652755737, + "learning_rate": 0.00016598280228135388, + "loss": 0.9754, + "step": 6095 + }, + { + "epoch": 1.0852920227920229, + "grad_norm": 0.5672318339347839, + "learning_rate": 0.0001659722837013594, + "loss": 0.9075, + "step": 6096 + }, + { + "epoch": 1.0854700854700854, + "grad_norm": 0.6397247314453125, + "learning_rate": 0.00016596176382877506, + "loss": 1.0358, + "step": 6097 + }, + { + "epoch": 1.0856481481481481, + "grad_norm": 0.6046154499053955, + "learning_rate": 0.000165951242663807, + "loss": 0.9036, + "step": 6098 + }, + { + "epoch": 1.085826210826211, + "grad_norm": 0.7190790176391602, + "learning_rate": 0.00016594072020666134, + "loss": 1.05, + "step": 6099 + }, + { + "epoch": 1.0860042735042734, + "grad_norm": 0.636986255645752, + "learning_rate": 0.00016593019645754425, + "loss": 1.0648, + "step": 6100 + }, + { + "epoch": 1.0861823361823362, + "grad_norm": 0.7239426374435425, + "learning_rate": 0.00016591967141666193, + "loss": 1.3332, + "step": 6101 + }, + { + "epoch": 1.086360398860399, + "grad_norm": 0.5623281002044678, + "learning_rate": 0.00016590914508422054, + "loss": 0.997, + "step": 6102 + }, + { + "epoch": 1.0865384615384615, + "grad_norm": 0.5559574365615845, + "learning_rate": 0.00016589861746042642, + "loss": 0.9309, + "step": 6103 + }, + { + "epoch": 1.0867165242165242, + "grad_norm": 0.6056998372077942, + "learning_rate": 0.00016588808854548574, + "loss": 1.05, + "step": 6104 + }, + { + "epoch": 1.086894586894587, + "grad_norm": 0.6419603228569031, + "learning_rate": 0.00016587755833960487, + "loss": 0.8933, + "step": 6105 + }, + { + "epoch": 1.0870726495726495, + "grad_norm": 0.5236496329307556, + "learning_rate": 0.00016586702684299006, + "loss": 1.0061, + "step": 6106 + }, + { + "epoch": 1.0872507122507122, + "grad_norm": 0.5764613747596741, + "learning_rate": 0.0001658564940558477, + "loss": 1.0218, + "step": 6107 + }, + { + "epoch": 1.087428774928775, + "grad_norm": 0.6049391627311707, + "learning_rate": 0.00016584595997838416, + "loss": 0.8157, + "step": 6108 + }, + { + "epoch": 1.0876068376068375, + "grad_norm": 0.585422933101654, + "learning_rate": 0.0001658354246108058, + "loss": 1.2761, + "step": 6109 + }, + { + "epoch": 1.0877849002849003, + "grad_norm": 0.6420125365257263, + "learning_rate": 0.00016582488795331907, + "loss": 1.1978, + "step": 6110 + }, + { + "epoch": 1.087962962962963, + "grad_norm": 0.646091878414154, + "learning_rate": 0.00016581435000613038, + "loss": 0.8946, + "step": 6111 + }, + { + "epoch": 1.0881410256410255, + "grad_norm": 0.6563934087753296, + "learning_rate": 0.00016580381076944625, + "loss": 1.0625, + "step": 6112 + }, + { + "epoch": 1.0883190883190883, + "grad_norm": 0.6796613931655884, + "learning_rate": 0.0001657932702434731, + "loss": 0.9401, + "step": 6113 + }, + { + "epoch": 1.088497150997151, + "grad_norm": 0.6248648762702942, + "learning_rate": 0.00016578272842841753, + "loss": 0.8558, + "step": 6114 + }, + { + "epoch": 1.0886752136752136, + "grad_norm": 0.5136269330978394, + "learning_rate": 0.00016577218532448605, + "loss": 0.6424, + "step": 6115 + }, + { + "epoch": 1.0888532763532763, + "grad_norm": 0.5581641793251038, + "learning_rate": 0.00016576164093188523, + "loss": 0.7923, + "step": 6116 + }, + { + "epoch": 1.089031339031339, + "grad_norm": 0.630352258682251, + "learning_rate": 0.0001657510952508216, + "loss": 0.9115, + "step": 6117 + }, + { + "epoch": 1.0892094017094016, + "grad_norm": 0.6167593002319336, + "learning_rate": 0.0001657405482815019, + "loss": 1.1112, + "step": 6118 + }, + { + "epoch": 1.0893874643874644, + "grad_norm": 0.5908578634262085, + "learning_rate": 0.00016573000002413271, + "loss": 1.0359, + "step": 6119 + }, + { + "epoch": 1.0895655270655271, + "grad_norm": 0.6326140761375427, + "learning_rate": 0.00016571945047892073, + "loss": 1.0459, + "step": 6120 + }, + { + "epoch": 1.0897435897435896, + "grad_norm": 0.7273572683334351, + "learning_rate": 0.00016570889964607262, + "loss": 1.0901, + "step": 6121 + }, + { + "epoch": 1.0899216524216524, + "grad_norm": 0.6168062090873718, + "learning_rate": 0.00016569834752579513, + "loss": 0.8739, + "step": 6122 + }, + { + "epoch": 1.0900997150997151, + "grad_norm": 0.5620378255844116, + "learning_rate": 0.00016568779411829497, + "loss": 0.9614, + "step": 6123 + }, + { + "epoch": 1.0902777777777777, + "grad_norm": 0.6319156885147095, + "learning_rate": 0.00016567723942377899, + "loss": 1.1031, + "step": 6124 + }, + { + "epoch": 1.0904558404558404, + "grad_norm": 0.6590072512626648, + "learning_rate": 0.00016566668344245388, + "loss": 1.0086, + "step": 6125 + }, + { + "epoch": 1.0906339031339032, + "grad_norm": 0.5823387503623962, + "learning_rate": 0.00016565612617452656, + "loss": 0.8886, + "step": 6126 + }, + { + "epoch": 1.0908119658119657, + "grad_norm": 0.5795989632606506, + "learning_rate": 0.00016564556762020381, + "loss": 0.7683, + "step": 6127 + }, + { + "epoch": 1.0909900284900285, + "grad_norm": 0.5940101742744446, + "learning_rate": 0.00016563500777969255, + "loss": 0.8873, + "step": 6128 + }, + { + "epoch": 1.0911680911680912, + "grad_norm": 0.5708247423171997, + "learning_rate": 0.00016562444665319963, + "loss": 0.7382, + "step": 6129 + }, + { + "epoch": 1.0913461538461537, + "grad_norm": 0.6339239478111267, + "learning_rate": 0.00016561388424093202, + "loss": 0.9323, + "step": 6130 + }, + { + "epoch": 1.0915242165242165, + "grad_norm": 0.720000147819519, + "learning_rate": 0.00016560332054309663, + "loss": 1.0437, + "step": 6131 + }, + { + "epoch": 1.0917022792022792, + "grad_norm": 0.686580240726471, + "learning_rate": 0.00016559275555990048, + "loss": 0.9841, + "step": 6132 + }, + { + "epoch": 1.091880341880342, + "grad_norm": 0.6067900061607361, + "learning_rate": 0.00016558218929155053, + "loss": 1.0862, + "step": 6133 + }, + { + "epoch": 1.0920584045584045, + "grad_norm": 0.6678896546363831, + "learning_rate": 0.00016557162173825384, + "loss": 0.8509, + "step": 6134 + }, + { + "epoch": 1.0922364672364673, + "grad_norm": 0.53044193983078, + "learning_rate": 0.0001655610529002174, + "loss": 0.9227, + "step": 6135 + }, + { + "epoch": 1.0924145299145298, + "grad_norm": 0.6499412655830383, + "learning_rate": 0.00016555048277764836, + "loss": 1.0867, + "step": 6136 + }, + { + "epoch": 1.0925925925925926, + "grad_norm": 0.6543099284172058, + "learning_rate": 0.00016553991137075374, + "loss": 0.849, + "step": 6137 + }, + { + "epoch": 1.0927706552706553, + "grad_norm": 0.5772737860679626, + "learning_rate": 0.0001655293386797407, + "loss": 0.8475, + "step": 6138 + }, + { + "epoch": 1.092948717948718, + "grad_norm": 0.616348385810852, + "learning_rate": 0.00016551876470481642, + "loss": 0.9205, + "step": 6139 + }, + { + "epoch": 1.0931267806267806, + "grad_norm": 0.7151142954826355, + "learning_rate": 0.00016550818944618801, + "loss": 1.1389, + "step": 6140 + }, + { + "epoch": 1.0933048433048433, + "grad_norm": 0.6566469669342041, + "learning_rate": 0.00016549761290406275, + "loss": 0.8216, + "step": 6141 + }, + { + "epoch": 1.0934829059829059, + "grad_norm": 0.7075428366661072, + "learning_rate": 0.00016548703507864783, + "loss": 1.065, + "step": 6142 + }, + { + "epoch": 1.0936609686609686, + "grad_norm": 0.6589360237121582, + "learning_rate": 0.00016547645597015046, + "loss": 0.9899, + "step": 6143 + }, + { + "epoch": 1.0938390313390314, + "grad_norm": 0.6445585489273071, + "learning_rate": 0.00016546587557877797, + "loss": 1.1629, + "step": 6144 + }, + { + "epoch": 1.0940170940170941, + "grad_norm": 0.6216462850570679, + "learning_rate": 0.00016545529390473763, + "loss": 0.9685, + "step": 6145 + }, + { + "epoch": 1.0941951566951567, + "grad_norm": 0.6195303797721863, + "learning_rate": 0.0001654447109482368, + "loss": 1.144, + "step": 6146 + }, + { + "epoch": 1.0943732193732194, + "grad_norm": 0.6625444293022156, + "learning_rate": 0.0001654341267094828, + "loss": 0.9886, + "step": 6147 + }, + { + "epoch": 1.094551282051282, + "grad_norm": 0.6449851393699646, + "learning_rate": 0.000165423541188683, + "loss": 0.9568, + "step": 6148 + }, + { + "epoch": 1.0947293447293447, + "grad_norm": 0.6490375995635986, + "learning_rate": 0.00016541295438604484, + "loss": 1.1304, + "step": 6149 + }, + { + "epoch": 1.0949074074074074, + "grad_norm": 0.6771987676620483, + "learning_rate": 0.00016540236630177574, + "loss": 1.0426, + "step": 6150 + }, + { + "epoch": 1.0950854700854702, + "grad_norm": 0.5214568376541138, + "learning_rate": 0.00016539177693608307, + "loss": 0.6742, + "step": 6151 + }, + { + "epoch": 1.0952635327635327, + "grad_norm": 0.6005097031593323, + "learning_rate": 0.00016538118628917442, + "loss": 0.9901, + "step": 6152 + }, + { + "epoch": 1.0954415954415955, + "grad_norm": 0.6449539065361023, + "learning_rate": 0.0001653705943612572, + "loss": 0.9654, + "step": 6153 + }, + { + "epoch": 1.095619658119658, + "grad_norm": 0.6443646550178528, + "learning_rate": 0.00016536000115253903, + "loss": 0.9084, + "step": 6154 + }, + { + "epoch": 1.0957977207977208, + "grad_norm": 0.6072495579719543, + "learning_rate": 0.0001653494066632274, + "loss": 0.6308, + "step": 6155 + }, + { + "epoch": 1.0959757834757835, + "grad_norm": 0.5751157999038696, + "learning_rate": 0.00016533881089352988, + "loss": 0.96, + "step": 6156 + }, + { + "epoch": 1.0961538461538463, + "grad_norm": 0.6310713291168213, + "learning_rate": 0.0001653282138436541, + "loss": 1.0997, + "step": 6157 + }, + { + "epoch": 1.0963319088319088, + "grad_norm": 0.5573651790618896, + "learning_rate": 0.00016531761551380765, + "loss": 0.9738, + "step": 6158 + }, + { + "epoch": 1.0965099715099715, + "grad_norm": 0.5615308880805969, + "learning_rate": 0.00016530701590419824, + "loss": 0.9658, + "step": 6159 + }, + { + "epoch": 1.0966880341880343, + "grad_norm": 0.6471942663192749, + "learning_rate": 0.0001652964150150335, + "loss": 1.0763, + "step": 6160 + }, + { + "epoch": 1.0968660968660968, + "grad_norm": 0.6305427551269531, + "learning_rate": 0.00016528581284652117, + "loss": 1.112, + "step": 6161 + }, + { + "epoch": 1.0970441595441596, + "grad_norm": 0.6881145238876343, + "learning_rate": 0.00016527520939886892, + "loss": 0.8476, + "step": 6162 + }, + { + "epoch": 1.0972222222222223, + "grad_norm": 0.6507891416549683, + "learning_rate": 0.00016526460467228458, + "loss": 1.1097, + "step": 6163 + }, + { + "epoch": 1.0974002849002849, + "grad_norm": 0.5960137844085693, + "learning_rate": 0.00016525399866697586, + "loss": 0.9934, + "step": 6164 + }, + { + "epoch": 1.0975783475783476, + "grad_norm": 0.6001808643341064, + "learning_rate": 0.0001652433913831506, + "loss": 1.0782, + "step": 6165 + }, + { + "epoch": 1.0977564102564104, + "grad_norm": 0.5639005303382874, + "learning_rate": 0.00016523278282101663, + "loss": 1.0929, + "step": 6166 + }, + { + "epoch": 1.0979344729344729, + "grad_norm": 0.5962058305740356, + "learning_rate": 0.00016522217298078177, + "loss": 1.0315, + "step": 6167 + }, + { + "epoch": 1.0981125356125356, + "grad_norm": 0.6920329928398132, + "learning_rate": 0.0001652115618626539, + "loss": 0.9176, + "step": 6168 + }, + { + "epoch": 1.0982905982905984, + "grad_norm": 0.6963527202606201, + "learning_rate": 0.00016520094946684098, + "loss": 1.2136, + "step": 6169 + }, + { + "epoch": 1.098468660968661, + "grad_norm": 0.5855711102485657, + "learning_rate": 0.00016519033579355093, + "loss": 0.8453, + "step": 6170 + }, + { + "epoch": 1.0986467236467237, + "grad_norm": 0.6454927325248718, + "learning_rate": 0.0001651797208429916, + "loss": 1.0747, + "step": 6171 + }, + { + "epoch": 1.0988247863247864, + "grad_norm": 0.644585907459259, + "learning_rate": 0.00016516910461537108, + "loss": 0.8165, + "step": 6172 + }, + { + "epoch": 1.099002849002849, + "grad_norm": 0.6488069891929626, + "learning_rate": 0.00016515848711089732, + "loss": 1.1048, + "step": 6173 + }, + { + "epoch": 1.0991809116809117, + "grad_norm": 0.5867953896522522, + "learning_rate": 0.00016514786832977834, + "loss": 0.63, + "step": 6174 + }, + { + "epoch": 1.0993589743589745, + "grad_norm": 0.560591459274292, + "learning_rate": 0.00016513724827222227, + "loss": 0.9255, + "step": 6175 + }, + { + "epoch": 1.099537037037037, + "grad_norm": 0.675262451171875, + "learning_rate": 0.00016512662693843707, + "loss": 0.7637, + "step": 6176 + }, + { + "epoch": 1.0997150997150997, + "grad_norm": 0.6515669822692871, + "learning_rate": 0.00016511600432863091, + "loss": 0.7579, + "step": 6177 + }, + { + "epoch": 1.0998931623931625, + "grad_norm": 0.683409571647644, + "learning_rate": 0.00016510538044301192, + "loss": 0.9183, + "step": 6178 + }, + { + "epoch": 1.100071225071225, + "grad_norm": 0.6194507479667664, + "learning_rate": 0.00016509475528178827, + "loss": 1.16, + "step": 6179 + }, + { + "epoch": 1.1002492877492878, + "grad_norm": 0.6192209720611572, + "learning_rate": 0.0001650841288451681, + "loss": 1.1392, + "step": 6180 + }, + { + "epoch": 1.1004273504273505, + "grad_norm": 0.6029189825057983, + "learning_rate": 0.0001650735011333596, + "loss": 1.1453, + "step": 6181 + }, + { + "epoch": 1.100605413105413, + "grad_norm": 0.7040731310844421, + "learning_rate": 0.00016506287214657105, + "loss": 0.9367, + "step": 6182 + }, + { + "epoch": 1.1007834757834758, + "grad_norm": 0.5909842252731323, + "learning_rate": 0.00016505224188501067, + "loss": 0.6463, + "step": 6183 + }, + { + "epoch": 1.1009615384615385, + "grad_norm": 0.6129698157310486, + "learning_rate": 0.00016504161034888674, + "loss": 0.9432, + "step": 6184 + }, + { + "epoch": 1.101139601139601, + "grad_norm": 0.6181607842445374, + "learning_rate": 0.00016503097753840757, + "loss": 0.9934, + "step": 6185 + }, + { + "epoch": 1.1013176638176638, + "grad_norm": 0.6463226675987244, + "learning_rate": 0.0001650203434537815, + "loss": 0.8471, + "step": 6186 + }, + { + "epoch": 1.1014957264957266, + "grad_norm": 0.5999348163604736, + "learning_rate": 0.00016500970809521688, + "loss": 0.9418, + "step": 6187 + }, + { + "epoch": 1.101673789173789, + "grad_norm": 0.629504919052124, + "learning_rate": 0.00016499907146292204, + "loss": 0.9699, + "step": 6188 + }, + { + "epoch": 1.1018518518518519, + "grad_norm": 0.694767951965332, + "learning_rate": 0.00016498843355710542, + "loss": 0.8793, + "step": 6189 + }, + { + "epoch": 1.1020299145299146, + "grad_norm": 0.6205509901046753, + "learning_rate": 0.00016497779437797547, + "loss": 0.8384, + "step": 6190 + }, + { + "epoch": 1.1022079772079771, + "grad_norm": 0.6256579756736755, + "learning_rate": 0.0001649671539257406, + "loss": 0.9275, + "step": 6191 + }, + { + "epoch": 1.10238603988604, + "grad_norm": 0.6593793034553528, + "learning_rate": 0.00016495651220060933, + "loss": 1.0495, + "step": 6192 + }, + { + "epoch": 1.1025641025641026, + "grad_norm": 0.7809221148490906, + "learning_rate": 0.00016494586920279012, + "loss": 1.0485, + "step": 6193 + }, + { + "epoch": 1.1027421652421652, + "grad_norm": 0.6147717833518982, + "learning_rate": 0.0001649352249324915, + "loss": 0.8739, + "step": 6194 + }, + { + "epoch": 1.102920227920228, + "grad_norm": 0.565411388874054, + "learning_rate": 0.00016492457938992208, + "loss": 0.9759, + "step": 6195 + }, + { + "epoch": 1.1030982905982907, + "grad_norm": 0.596370279788971, + "learning_rate": 0.00016491393257529036, + "loss": 0.9658, + "step": 6196 + }, + { + "epoch": 1.1032763532763532, + "grad_norm": 0.6334326863288879, + "learning_rate": 0.00016490328448880498, + "loss": 0.8785, + "step": 6197 + }, + { + "epoch": 1.103454415954416, + "grad_norm": 0.5538334846496582, + "learning_rate": 0.0001648926351306746, + "loss": 0.7174, + "step": 6198 + }, + { + "epoch": 1.1036324786324787, + "grad_norm": 0.6249658465385437, + "learning_rate": 0.00016488198450110778, + "loss": 0.8579, + "step": 6199 + }, + { + "epoch": 1.1038105413105412, + "grad_norm": 0.6128895878791809, + "learning_rate": 0.00016487133260031329, + "loss": 0.8538, + "step": 6200 + }, + { + "epoch": 1.103988603988604, + "grad_norm": 0.5808702707290649, + "learning_rate": 0.0001648606794284998, + "loss": 0.8143, + "step": 6201 + }, + { + "epoch": 1.1041666666666667, + "grad_norm": 0.671419084072113, + "learning_rate": 0.00016485002498587602, + "loss": 1.1268, + "step": 6202 + }, + { + "epoch": 1.1043447293447293, + "grad_norm": 0.5706788897514343, + "learning_rate": 0.00016483936927265075, + "loss": 0.9558, + "step": 6203 + }, + { + "epoch": 1.104522792022792, + "grad_norm": 0.5700307488441467, + "learning_rate": 0.00016482871228903266, + "loss": 0.9616, + "step": 6204 + }, + { + "epoch": 1.1047008547008548, + "grad_norm": 0.5764816403388977, + "learning_rate": 0.0001648180540352307, + "loss": 0.8692, + "step": 6205 + }, + { + "epoch": 1.1048789173789173, + "grad_norm": 0.5786563754081726, + "learning_rate": 0.00016480739451145358, + "loss": 0.9406, + "step": 6206 + }, + { + "epoch": 1.10505698005698, + "grad_norm": 0.6112591624259949, + "learning_rate": 0.0001647967337179102, + "loss": 0.8999, + "step": 6207 + }, + { + "epoch": 1.1052350427350428, + "grad_norm": 0.5708907246589661, + "learning_rate": 0.00016478607165480944, + "loss": 0.9236, + "step": 6208 + }, + { + "epoch": 1.1054131054131053, + "grad_norm": 0.6742013692855835, + "learning_rate": 0.00016477540832236014, + "loss": 1.0911, + "step": 6209 + }, + { + "epoch": 1.105591168091168, + "grad_norm": 0.6382617354393005, + "learning_rate": 0.0001647647437207713, + "loss": 0.7901, + "step": 6210 + }, + { + "epoch": 1.1057692307692308, + "grad_norm": 0.6241547465324402, + "learning_rate": 0.00016475407785025188, + "loss": 1.0048, + "step": 6211 + }, + { + "epoch": 1.1059472934472934, + "grad_norm": 0.6452877521514893, + "learning_rate": 0.00016474341071101077, + "loss": 0.8902, + "step": 6212 + }, + { + "epoch": 1.1061253561253561, + "grad_norm": 0.6212326288223267, + "learning_rate": 0.00016473274230325704, + "loss": 1.078, + "step": 6213 + }, + { + "epoch": 1.1063034188034189, + "grad_norm": 0.6870912909507751, + "learning_rate": 0.00016472207262719968, + "loss": 0.9127, + "step": 6214 + }, + { + "epoch": 1.1064814814814814, + "grad_norm": 0.6286750435829163, + "learning_rate": 0.00016471140168304777, + "loss": 1.0271, + "step": 6215 + }, + { + "epoch": 1.1066595441595442, + "grad_norm": 0.645806074142456, + "learning_rate": 0.00016470072947101036, + "loss": 1.1514, + "step": 6216 + }, + { + "epoch": 1.106837606837607, + "grad_norm": 0.6800320148468018, + "learning_rate": 0.00016469005599129653, + "loss": 0.9322, + "step": 6217 + }, + { + "epoch": 1.1070156695156694, + "grad_norm": 0.5898309946060181, + "learning_rate": 0.0001646793812441155, + "loss": 1.065, + "step": 6218 + }, + { + "epoch": 1.1071937321937322, + "grad_norm": 0.6000019907951355, + "learning_rate": 0.00016466870522967634, + "loss": 0.911, + "step": 6219 + }, + { + "epoch": 1.107371794871795, + "grad_norm": 0.6164331436157227, + "learning_rate": 0.0001646580279481882, + "loss": 0.8421, + "step": 6220 + }, + { + "epoch": 1.1075498575498575, + "grad_norm": 0.6410242319107056, + "learning_rate": 0.00016464734939986036, + "loss": 0.9688, + "step": 6221 + }, + { + "epoch": 1.1077279202279202, + "grad_norm": 0.7153300046920776, + "learning_rate": 0.00016463666958490197, + "loss": 1.0722, + "step": 6222 + }, + { + "epoch": 1.107905982905983, + "grad_norm": 0.6977026462554932, + "learning_rate": 0.00016462598850352234, + "loss": 1.0192, + "step": 6223 + }, + { + "epoch": 1.1080840455840455, + "grad_norm": 0.6379461884498596, + "learning_rate": 0.0001646153061559307, + "loss": 1.0474, + "step": 6224 + }, + { + "epoch": 1.1082621082621082, + "grad_norm": 0.6135090589523315, + "learning_rate": 0.00016460462254233634, + "loss": 1.0082, + "step": 6225 + }, + { + "epoch": 1.108440170940171, + "grad_norm": 0.6326230764389038, + "learning_rate": 0.00016459393766294866, + "loss": 1.1097, + "step": 6226 + }, + { + "epoch": 1.1086182336182335, + "grad_norm": 0.6636839509010315, + "learning_rate": 0.0001645832515179769, + "loss": 0.9689, + "step": 6227 + }, + { + "epoch": 1.1087962962962963, + "grad_norm": 0.5713129043579102, + "learning_rate": 0.00016457256410763052, + "loss": 0.8642, + "step": 6228 + }, + { + "epoch": 1.108974358974359, + "grad_norm": 0.584204912185669, + "learning_rate": 0.00016456187543211888, + "loss": 0.9957, + "step": 6229 + }, + { + "epoch": 1.1091524216524216, + "grad_norm": 0.5920230746269226, + "learning_rate": 0.0001645511854916514, + "loss": 0.7297, + "step": 6230 + }, + { + "epoch": 1.1093304843304843, + "grad_norm": 0.6207385063171387, + "learning_rate": 0.0001645404942864375, + "loss": 0.868, + "step": 6231 + }, + { + "epoch": 1.109508547008547, + "grad_norm": 0.7267234921455383, + "learning_rate": 0.00016452980181668673, + "loss": 1.0248, + "step": 6232 + }, + { + "epoch": 1.1096866096866096, + "grad_norm": 0.5925650596618652, + "learning_rate": 0.00016451910808260852, + "loss": 1.1075, + "step": 6233 + }, + { + "epoch": 1.1098646723646723, + "grad_norm": 0.5632196664810181, + "learning_rate": 0.00016450841308441244, + "loss": 0.9865, + "step": 6234 + }, + { + "epoch": 1.110042735042735, + "grad_norm": 0.6115161180496216, + "learning_rate": 0.000164497716822308, + "loss": 1.1343, + "step": 6235 + }, + { + "epoch": 1.1102207977207976, + "grad_norm": 0.634398341178894, + "learning_rate": 0.00016448701929650477, + "loss": 1.1039, + "step": 6236 + }, + { + "epoch": 1.1103988603988604, + "grad_norm": 0.5843468308448792, + "learning_rate": 0.00016447632050721237, + "loss": 0.8462, + "step": 6237 + }, + { + "epoch": 1.1105769230769231, + "grad_norm": 0.799375593662262, + "learning_rate": 0.0001644656204546404, + "loss": 0.9861, + "step": 6238 + }, + { + "epoch": 1.1107549857549857, + "grad_norm": 0.600289523601532, + "learning_rate": 0.0001644549191389985, + "loss": 1.0323, + "step": 6239 + }, + { + "epoch": 1.1109330484330484, + "grad_norm": 0.6154919266700745, + "learning_rate": 0.00016444421656049637, + "loss": 0.9158, + "step": 6240 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 0.6685689687728882, + "learning_rate": 0.00016443351271934367, + "loss": 1.0429, + "step": 6241 + }, + { + "epoch": 1.1112891737891737, + "grad_norm": 0.699978232383728, + "learning_rate": 0.00016442280761575016, + "loss": 1.072, + "step": 6242 + }, + { + "epoch": 1.1114672364672364, + "grad_norm": 0.6461396217346191, + "learning_rate": 0.00016441210124992556, + "loss": 0.9758, + "step": 6243 + }, + { + "epoch": 1.1116452991452992, + "grad_norm": 0.6463284492492676, + "learning_rate": 0.00016440139362207962, + "loss": 0.9205, + "step": 6244 + }, + { + "epoch": 1.1118233618233617, + "grad_norm": 0.6587556004524231, + "learning_rate": 0.00016439068473242217, + "loss": 1.0027, + "step": 6245 + }, + { + "epoch": 1.1120014245014245, + "grad_norm": 0.6896520256996155, + "learning_rate": 0.000164379974581163, + "loss": 0.9788, + "step": 6246 + }, + { + "epoch": 1.1121794871794872, + "grad_norm": 0.6766142845153809, + "learning_rate": 0.000164369263168512, + "loss": 0.9647, + "step": 6247 + }, + { + "epoch": 1.1123575498575498, + "grad_norm": 0.7024297118186951, + "learning_rate": 0.00016435855049467898, + "loss": 1.1163, + "step": 6248 + }, + { + "epoch": 1.1125356125356125, + "grad_norm": 0.6654963493347168, + "learning_rate": 0.00016434783655987385, + "loss": 0.9302, + "step": 6249 + }, + { + "epoch": 1.1127136752136753, + "grad_norm": 0.6973692774772644, + "learning_rate": 0.0001643371213643065, + "loss": 0.9585, + "step": 6250 + }, + { + "epoch": 1.1128917378917378, + "grad_norm": 0.7153545022010803, + "learning_rate": 0.000164326404908187, + "loss": 1.0485, + "step": 6251 + }, + { + "epoch": 1.1130698005698005, + "grad_norm": 0.6114685535430908, + "learning_rate": 0.00016431568719172516, + "loss": 0.8881, + "step": 6252 + }, + { + "epoch": 1.1132478632478633, + "grad_norm": 0.6500731706619263, + "learning_rate": 0.00016430496821513103, + "loss": 1.0658, + "step": 6253 + }, + { + "epoch": 1.113425925925926, + "grad_norm": 0.5800092220306396, + "learning_rate": 0.00016429424797861466, + "loss": 0.9158, + "step": 6254 + }, + { + "epoch": 1.1136039886039886, + "grad_norm": 0.6653759479522705, + "learning_rate": 0.00016428352648238602, + "loss": 0.9762, + "step": 6255 + }, + { + "epoch": 1.1137820512820513, + "grad_norm": 0.649208128452301, + "learning_rate": 0.00016427280372665525, + "loss": 1.1184, + "step": 6256 + }, + { + "epoch": 1.1139601139601139, + "grad_norm": 0.6665199398994446, + "learning_rate": 0.00016426207971163238, + "loss": 0.9417, + "step": 6257 + }, + { + "epoch": 1.1141381766381766, + "grad_norm": 0.6110978126525879, + "learning_rate": 0.00016425135443752758, + "loss": 1.1531, + "step": 6258 + }, + { + "epoch": 1.1143162393162394, + "grad_norm": 0.6517077088356018, + "learning_rate": 0.00016424062790455093, + "loss": 0.9055, + "step": 6259 + }, + { + "epoch": 1.114494301994302, + "grad_norm": 0.6278966665267944, + "learning_rate": 0.00016422990011291265, + "loss": 1.0087, + "step": 6260 + }, + { + "epoch": 1.1146723646723646, + "grad_norm": 0.5818809270858765, + "learning_rate": 0.00016421917106282288, + "loss": 1.0202, + "step": 6261 + }, + { + "epoch": 1.1148504273504274, + "grad_norm": 0.5670005679130554, + "learning_rate": 0.00016420844075449187, + "loss": 0.841, + "step": 6262 + }, + { + "epoch": 1.11502849002849, + "grad_norm": 0.6584762334823608, + "learning_rate": 0.00016419770918812984, + "loss": 1.0322, + "step": 6263 + }, + { + "epoch": 1.1152065527065527, + "grad_norm": 0.6023790836334229, + "learning_rate": 0.00016418697636394705, + "loss": 0.9152, + "step": 6264 + }, + { + "epoch": 1.1153846153846154, + "grad_norm": 0.6234691739082336, + "learning_rate": 0.00016417624228215382, + "loss": 0.9555, + "step": 6265 + }, + { + "epoch": 1.1155626780626782, + "grad_norm": 0.6690816879272461, + "learning_rate": 0.00016416550694296045, + "loss": 0.9341, + "step": 6266 + }, + { + "epoch": 1.1157407407407407, + "grad_norm": 0.6030237078666687, + "learning_rate": 0.00016415477034657723, + "loss": 1.0442, + "step": 6267 + }, + { + "epoch": 1.1159188034188035, + "grad_norm": 0.5954633951187134, + "learning_rate": 0.00016414403249321455, + "loss": 0.9132, + "step": 6268 + }, + { + "epoch": 1.116096866096866, + "grad_norm": 0.7876830101013184, + "learning_rate": 0.0001641332933830828, + "loss": 0.9456, + "step": 6269 + }, + { + "epoch": 1.1162749287749287, + "grad_norm": 0.6776009798049927, + "learning_rate": 0.00016412255301639244, + "loss": 0.9022, + "step": 6270 + }, + { + "epoch": 1.1164529914529915, + "grad_norm": 0.6094426512718201, + "learning_rate": 0.0001641118113933538, + "loss": 0.9629, + "step": 6271 + }, + { + "epoch": 1.1166310541310542, + "grad_norm": 0.5818213820457458, + "learning_rate": 0.00016410106851417742, + "loss": 0.9049, + "step": 6272 + }, + { + "epoch": 1.1168091168091168, + "grad_norm": 0.5668078064918518, + "learning_rate": 0.00016409032437907377, + "loss": 1.0011, + "step": 6273 + }, + { + "epoch": 1.1169871794871795, + "grad_norm": 0.6984922289848328, + "learning_rate": 0.00016407957898825334, + "loss": 0.9454, + "step": 6274 + }, + { + "epoch": 1.1171652421652423, + "grad_norm": 0.5509830117225647, + "learning_rate": 0.00016406883234192668, + "loss": 0.9132, + "step": 6275 + }, + { + "epoch": 1.1173433048433048, + "grad_norm": 0.5117461681365967, + "learning_rate": 0.00016405808444030435, + "loss": 0.7675, + "step": 6276 + }, + { + "epoch": 1.1175213675213675, + "grad_norm": 0.6358339786529541, + "learning_rate": 0.00016404733528359688, + "loss": 0.9777, + "step": 6277 + }, + { + "epoch": 1.1176994301994303, + "grad_norm": 0.5870591402053833, + "learning_rate": 0.00016403658487201494, + "loss": 0.8576, + "step": 6278 + }, + { + "epoch": 1.1178774928774928, + "grad_norm": 0.6532407999038696, + "learning_rate": 0.00016402583320576915, + "loss": 1.1787, + "step": 6279 + }, + { + "epoch": 1.1180555555555556, + "grad_norm": 0.6374639272689819, + "learning_rate": 0.00016401508028507017, + "loss": 0.9298, + "step": 6280 + }, + { + "epoch": 1.1182336182336183, + "grad_norm": 0.7280316352844238, + "learning_rate": 0.00016400432611012869, + "loss": 1.1081, + "step": 6281 + }, + { + "epoch": 1.1184116809116809, + "grad_norm": 0.6070699095726013, + "learning_rate": 0.00016399357068115538, + "loss": 0.9107, + "step": 6282 + }, + { + "epoch": 1.1185897435897436, + "grad_norm": 0.6701489686965942, + "learning_rate": 0.00016398281399836097, + "loss": 1.0879, + "step": 6283 + }, + { + "epoch": 1.1187678062678064, + "grad_norm": 0.6343162655830383, + "learning_rate": 0.00016397205606195626, + "loss": 0.8552, + "step": 6284 + }, + { + "epoch": 1.118945868945869, + "grad_norm": 0.6450608968734741, + "learning_rate": 0.00016396129687215198, + "loss": 1.1119, + "step": 6285 + }, + { + "epoch": 1.1191239316239316, + "grad_norm": 0.7219904661178589, + "learning_rate": 0.00016395053642915896, + "loss": 0.9081, + "step": 6286 + }, + { + "epoch": 1.1193019943019944, + "grad_norm": 0.6189733147621155, + "learning_rate": 0.00016393977473318802, + "loss": 0.9818, + "step": 6287 + }, + { + "epoch": 1.119480056980057, + "grad_norm": 0.6310907602310181, + "learning_rate": 0.00016392901178445004, + "loss": 1.0334, + "step": 6288 + }, + { + "epoch": 1.1196581196581197, + "grad_norm": 0.6556720733642578, + "learning_rate": 0.00016391824758315587, + "loss": 1.0452, + "step": 6289 + }, + { + "epoch": 1.1198361823361824, + "grad_norm": 0.6697782278060913, + "learning_rate": 0.00016390748212951638, + "loss": 0.9627, + "step": 6290 + }, + { + "epoch": 1.120014245014245, + "grad_norm": 0.6341549754142761, + "learning_rate": 0.00016389671542374256, + "loss": 1.112, + "step": 6291 + }, + { + "epoch": 1.1201923076923077, + "grad_norm": 0.6913946270942688, + "learning_rate": 0.00016388594746604535, + "loss": 0.9622, + "step": 6292 + }, + { + "epoch": 1.1203703703703705, + "grad_norm": 0.695488691329956, + "learning_rate": 0.0001638751782566357, + "loss": 1.0951, + "step": 6293 + }, + { + "epoch": 1.120548433048433, + "grad_norm": 0.6965359449386597, + "learning_rate": 0.00016386440779572463, + "loss": 1.1742, + "step": 6294 + }, + { + "epoch": 1.1207264957264957, + "grad_norm": 0.624679684638977, + "learning_rate": 0.00016385363608352314, + "loss": 0.9756, + "step": 6295 + }, + { + "epoch": 1.1209045584045585, + "grad_norm": 0.7511318922042847, + "learning_rate": 0.0001638428631202423, + "loss": 0.907, + "step": 6296 + }, + { + "epoch": 1.121082621082621, + "grad_norm": 0.5334641337394714, + "learning_rate": 0.00016383208890609317, + "loss": 0.7932, + "step": 6297 + }, + { + "epoch": 1.1212606837606838, + "grad_norm": 0.7518552541732788, + "learning_rate": 0.00016382131344128687, + "loss": 1.1556, + "step": 6298 + }, + { + "epoch": 1.1214387464387465, + "grad_norm": 0.618618369102478, + "learning_rate": 0.00016381053672603449, + "loss": 1.1027, + "step": 6299 + }, + { + "epoch": 1.121616809116809, + "grad_norm": 0.638956606388092, + "learning_rate": 0.00016379975876054724, + "loss": 1.0377, + "step": 6300 + }, + { + "epoch": 1.1217948717948718, + "grad_norm": 0.8031370639801025, + "learning_rate": 0.0001637889795450362, + "loss": 1.0821, + "step": 6301 + }, + { + "epoch": 1.1219729344729346, + "grad_norm": 0.6710168123245239, + "learning_rate": 0.00016377819907971265, + "loss": 1.2896, + "step": 6302 + }, + { + "epoch": 1.122150997150997, + "grad_norm": 0.5850739479064941, + "learning_rate": 0.00016376741736478777, + "loss": 1.0836, + "step": 6303 + }, + { + "epoch": 1.1223290598290598, + "grad_norm": 0.6410611271858215, + "learning_rate": 0.0001637566344004728, + "loss": 1.0395, + "step": 6304 + }, + { + "epoch": 1.1225071225071226, + "grad_norm": 0.6884660720825195, + "learning_rate": 0.00016374585018697903, + "loss": 0.871, + "step": 6305 + }, + { + "epoch": 1.1226851851851851, + "grad_norm": 0.622207522392273, + "learning_rate": 0.00016373506472451777, + "loss": 0.9897, + "step": 6306 + }, + { + "epoch": 1.1228632478632479, + "grad_norm": 0.6018275618553162, + "learning_rate": 0.00016372427801330028, + "loss": 0.8398, + "step": 6307 + }, + { + "epoch": 1.1230413105413106, + "grad_norm": 0.6451539993286133, + "learning_rate": 0.00016371349005353796, + "loss": 0.9878, + "step": 6308 + }, + { + "epoch": 1.1232193732193732, + "grad_norm": 0.5549424886703491, + "learning_rate": 0.00016370270084544215, + "loss": 0.844, + "step": 6309 + }, + { + "epoch": 1.123397435897436, + "grad_norm": 0.6082940697669983, + "learning_rate": 0.00016369191038922423, + "loss": 1.0704, + "step": 6310 + }, + { + "epoch": 1.1235754985754987, + "grad_norm": 0.6423100829124451, + "learning_rate": 0.00016368111868509563, + "loss": 1.0639, + "step": 6311 + }, + { + "epoch": 1.1237535612535612, + "grad_norm": 0.6274200081825256, + "learning_rate": 0.00016367032573326784, + "loss": 0.9996, + "step": 6312 + }, + { + "epoch": 1.123931623931624, + "grad_norm": 0.6618558168411255, + "learning_rate": 0.00016365953153395227, + "loss": 0.8074, + "step": 6313 + }, + { + "epoch": 1.1241096866096867, + "grad_norm": 0.7624069452285767, + "learning_rate": 0.00016364873608736038, + "loss": 0.9741, + "step": 6314 + }, + { + "epoch": 1.1242877492877492, + "grad_norm": 0.5391361117362976, + "learning_rate": 0.00016363793939370375, + "loss": 0.6992, + "step": 6315 + }, + { + "epoch": 1.124465811965812, + "grad_norm": 0.7564396858215332, + "learning_rate": 0.0001636271414531939, + "loss": 1.1971, + "step": 6316 + }, + { + "epoch": 1.1246438746438747, + "grad_norm": 0.6584066152572632, + "learning_rate": 0.00016361634226604239, + "loss": 1.0842, + "step": 6317 + }, + { + "epoch": 1.1248219373219372, + "grad_norm": 0.6851227283477783, + "learning_rate": 0.00016360554183246078, + "loss": 1.0879, + "step": 6318 + }, + { + "epoch": 1.125, + "grad_norm": 0.5699417591094971, + "learning_rate": 0.00016359474015266074, + "loss": 0.782, + "step": 6319 + }, + { + "epoch": 1.1251780626780628, + "grad_norm": 0.5495570302009583, + "learning_rate": 0.00016358393722685385, + "loss": 1.076, + "step": 6320 + }, + { + "epoch": 1.1253561253561253, + "grad_norm": 0.5872206091880798, + "learning_rate": 0.0001635731330552518, + "loss": 0.8601, + "step": 6321 + }, + { + "epoch": 1.125534188034188, + "grad_norm": 0.7012827396392822, + "learning_rate": 0.00016356232763806627, + "loss": 1.0443, + "step": 6322 + }, + { + "epoch": 1.1257122507122508, + "grad_norm": 0.6645881533622742, + "learning_rate": 0.00016355152097550897, + "loss": 1.0027, + "step": 6323 + }, + { + "epoch": 1.1258903133903133, + "grad_norm": 0.7376120090484619, + "learning_rate": 0.00016354071306779163, + "loss": 1.1941, + "step": 6324 + }, + { + "epoch": 1.126068376068376, + "grad_norm": 0.648932695388794, + "learning_rate": 0.000163529903915126, + "loss": 1.096, + "step": 6325 + }, + { + "epoch": 1.1262464387464388, + "grad_norm": 0.6186314821243286, + "learning_rate": 0.0001635190935177239, + "loss": 1.011, + "step": 6326 + }, + { + "epoch": 1.1264245014245013, + "grad_norm": 0.5964710116386414, + "learning_rate": 0.0001635082818757971, + "loss": 0.8893, + "step": 6327 + }, + { + "epoch": 1.126602564102564, + "grad_norm": 0.5264934301376343, + "learning_rate": 0.00016349746898955747, + "loss": 0.7325, + "step": 6328 + }, + { + "epoch": 1.1267806267806268, + "grad_norm": 0.6523048877716064, + "learning_rate": 0.00016348665485921678, + "loss": 1.0488, + "step": 6329 + }, + { + "epoch": 1.1269586894586894, + "grad_norm": 0.6878600120544434, + "learning_rate": 0.00016347583948498703, + "loss": 1.0926, + "step": 6330 + }, + { + "epoch": 1.1271367521367521, + "grad_norm": 0.592656672000885, + "learning_rate": 0.00016346502286708004, + "loss": 0.978, + "step": 6331 + }, + { + "epoch": 1.1273148148148149, + "grad_norm": 0.6338315606117249, + "learning_rate": 0.00016345420500570777, + "loss": 1.1048, + "step": 6332 + }, + { + "epoch": 1.1274928774928774, + "grad_norm": 0.5955204367637634, + "learning_rate": 0.00016344338590108218, + "loss": 0.88, + "step": 6333 + }, + { + "epoch": 1.1276709401709402, + "grad_norm": 0.690448522567749, + "learning_rate": 0.0001634325655534152, + "loss": 1.0564, + "step": 6334 + }, + { + "epoch": 1.127849002849003, + "grad_norm": 0.6125795841217041, + "learning_rate": 0.00016342174396291888, + "loss": 1.0608, + "step": 6335 + }, + { + "epoch": 1.1280270655270654, + "grad_norm": 0.6387807726860046, + "learning_rate": 0.00016341092112980523, + "loss": 0.9581, + "step": 6336 + }, + { + "epoch": 1.1282051282051282, + "grad_norm": 0.6247823238372803, + "learning_rate": 0.0001634000970542863, + "loss": 0.932, + "step": 6337 + }, + { + "epoch": 1.128383190883191, + "grad_norm": 0.5928077697753906, + "learning_rate": 0.0001633892717365742, + "loss": 0.8963, + "step": 6338 + }, + { + "epoch": 1.1285612535612535, + "grad_norm": 0.5922074913978577, + "learning_rate": 0.000163378445176881, + "loss": 0.9772, + "step": 6339 + }, + { + "epoch": 1.1287393162393162, + "grad_norm": 0.6573056578636169, + "learning_rate": 0.00016336761737541878, + "loss": 0.8233, + "step": 6340 + }, + { + "epoch": 1.128917378917379, + "grad_norm": 0.627772867679596, + "learning_rate": 0.0001633567883323998, + "loss": 0.9618, + "step": 6341 + }, + { + "epoch": 1.1290954415954415, + "grad_norm": 0.6066579818725586, + "learning_rate": 0.0001633459580480361, + "loss": 0.9066, + "step": 6342 + }, + { + "epoch": 1.1292735042735043, + "grad_norm": 0.670295000076294, + "learning_rate": 0.00016333512652253997, + "loss": 0.8003, + "step": 6343 + }, + { + "epoch": 1.129451566951567, + "grad_norm": 0.6402488946914673, + "learning_rate": 0.0001633242937561236, + "loss": 0.998, + "step": 6344 + }, + { + "epoch": 1.1296296296296295, + "grad_norm": 0.7224995493888855, + "learning_rate": 0.00016331345974899923, + "loss": 1.0308, + "step": 6345 + }, + { + "epoch": 1.1298076923076923, + "grad_norm": 0.5019716620445251, + "learning_rate": 0.00016330262450137917, + "loss": 0.6874, + "step": 6346 + }, + { + "epoch": 1.129985754985755, + "grad_norm": 0.5774167776107788, + "learning_rate": 0.00016329178801347566, + "loss": 0.8287, + "step": 6347 + }, + { + "epoch": 1.1301638176638176, + "grad_norm": 0.7797795534133911, + "learning_rate": 0.00016328095028550103, + "loss": 1.2145, + "step": 6348 + }, + { + "epoch": 1.1303418803418803, + "grad_norm": 0.5384017825126648, + "learning_rate": 0.00016327011131766765, + "loss": 0.8022, + "step": 6349 + }, + { + "epoch": 1.130519943019943, + "grad_norm": 0.6350888609886169, + "learning_rate": 0.00016325927111018786, + "loss": 1.1178, + "step": 6350 + }, + { + "epoch": 1.1306980056980056, + "grad_norm": 0.6386831998825073, + "learning_rate": 0.0001632484296632741, + "loss": 0.967, + "step": 6351 + }, + { + "epoch": 1.1308760683760684, + "grad_norm": 0.6214167475700378, + "learning_rate": 0.0001632375869771387, + "loss": 0.9416, + "step": 6352 + }, + { + "epoch": 1.131054131054131, + "grad_norm": 0.6145567297935486, + "learning_rate": 0.00016322674305199416, + "loss": 0.9175, + "step": 6353 + }, + { + "epoch": 1.1312321937321936, + "grad_norm": 0.7027857303619385, + "learning_rate": 0.00016321589788805297, + "loss": 1.0063, + "step": 6354 + }, + { + "epoch": 1.1314102564102564, + "grad_norm": 0.6942669153213501, + "learning_rate": 0.00016320505148552755, + "loss": 0.9191, + "step": 6355 + }, + { + "epoch": 1.1315883190883191, + "grad_norm": 0.6388658285140991, + "learning_rate": 0.0001631942038446304, + "loss": 0.993, + "step": 6356 + }, + { + "epoch": 1.131766381766382, + "grad_norm": 0.6627292633056641, + "learning_rate": 0.00016318335496557415, + "loss": 1.0055, + "step": 6357 + }, + { + "epoch": 1.1319444444444444, + "grad_norm": 0.7997342944145203, + "learning_rate": 0.0001631725048485713, + "loss": 0.9019, + "step": 6358 + }, + { + "epoch": 1.1321225071225072, + "grad_norm": 0.8817830681800842, + "learning_rate": 0.00016316165349383445, + "loss": 0.9793, + "step": 6359 + }, + { + "epoch": 1.1323005698005697, + "grad_norm": 0.5629408955574036, + "learning_rate": 0.00016315080090157621, + "loss": 0.6139, + "step": 6360 + }, + { + "epoch": 1.1324786324786325, + "grad_norm": 0.647220253944397, + "learning_rate": 0.0001631399470720092, + "loss": 0.9776, + "step": 6361 + }, + { + "epoch": 1.1326566951566952, + "grad_norm": 0.6762630939483643, + "learning_rate": 0.0001631290920053461, + "loss": 1.1027, + "step": 6362 + }, + { + "epoch": 1.132834757834758, + "grad_norm": 0.5862727761268616, + "learning_rate": 0.00016311823570179957, + "loss": 1.1359, + "step": 6363 + }, + { + "epoch": 1.1330128205128205, + "grad_norm": 0.7042981386184692, + "learning_rate": 0.00016310737816158235, + "loss": 1.142, + "step": 6364 + }, + { + "epoch": 1.1331908831908832, + "grad_norm": 0.5990639328956604, + "learning_rate": 0.00016309651938490712, + "loss": 0.9306, + "step": 6365 + }, + { + "epoch": 1.1333689458689458, + "grad_norm": 0.5894871950149536, + "learning_rate": 0.00016308565937198669, + "loss": 0.8343, + "step": 6366 + }, + { + "epoch": 1.1335470085470085, + "grad_norm": 0.6863628029823303, + "learning_rate": 0.0001630747981230338, + "loss": 0.9552, + "step": 6367 + }, + { + "epoch": 1.1337250712250713, + "grad_norm": 0.7438958287239075, + "learning_rate": 0.00016306393563826128, + "loss": 1.0422, + "step": 6368 + }, + { + "epoch": 1.133903133903134, + "grad_norm": 0.5695775747299194, + "learning_rate": 0.00016305307191788194, + "loss": 0.8633, + "step": 6369 + }, + { + "epoch": 1.1340811965811965, + "grad_norm": 0.6257741451263428, + "learning_rate": 0.00016304220696210863, + "loss": 1.0333, + "step": 6370 + }, + { + "epoch": 1.1342592592592593, + "grad_norm": 0.6366072297096252, + "learning_rate": 0.00016303134077115425, + "loss": 1.1452, + "step": 6371 + }, + { + "epoch": 1.1344373219373218, + "grad_norm": 0.624569296836853, + "learning_rate": 0.00016302047334523168, + "loss": 1.0569, + "step": 6372 + }, + { + "epoch": 1.1346153846153846, + "grad_norm": 0.5585938096046448, + "learning_rate": 0.00016300960468455382, + "loss": 0.9612, + "step": 6373 + }, + { + "epoch": 1.1347934472934473, + "grad_norm": 0.5738831162452698, + "learning_rate": 0.00016299873478933368, + "loss": 0.9206, + "step": 6374 + }, + { + "epoch": 1.13497150997151, + "grad_norm": 0.6797143220901489, + "learning_rate": 0.00016298786365978417, + "loss": 1.0748, + "step": 6375 + }, + { + "epoch": 1.1351495726495726, + "grad_norm": 0.6341326832771301, + "learning_rate": 0.00016297699129611833, + "loss": 0.9901, + "step": 6376 + }, + { + "epoch": 1.1353276353276354, + "grad_norm": 0.6568490862846375, + "learning_rate": 0.00016296611769854916, + "loss": 1.0598, + "step": 6377 + }, + { + "epoch": 1.135505698005698, + "grad_norm": 0.6151928901672363, + "learning_rate": 0.00016295524286728973, + "loss": 0.8352, + "step": 6378 + }, + { + "epoch": 1.1356837606837606, + "grad_norm": 0.7209593057632446, + "learning_rate": 0.0001629443668025531, + "loss": 0.9945, + "step": 6379 + }, + { + "epoch": 1.1358618233618234, + "grad_norm": 0.6600689888000488, + "learning_rate": 0.00016293348950455235, + "loss": 1.0572, + "step": 6380 + }, + { + "epoch": 1.1360398860398861, + "grad_norm": 0.5587523579597473, + "learning_rate": 0.0001629226109735006, + "loss": 0.8526, + "step": 6381 + }, + { + "epoch": 1.1362179487179487, + "grad_norm": 0.6184542775154114, + "learning_rate": 0.00016291173120961102, + "loss": 0.8246, + "step": 6382 + }, + { + "epoch": 1.1363960113960114, + "grad_norm": 0.6604713797569275, + "learning_rate": 0.00016290085021309673, + "loss": 1.0349, + "step": 6383 + }, + { + "epoch": 1.136574074074074, + "grad_norm": 0.5880835056304932, + "learning_rate": 0.00016288996798417097, + "loss": 0.8726, + "step": 6384 + }, + { + "epoch": 1.1367521367521367, + "grad_norm": 0.5770880579948425, + "learning_rate": 0.00016287908452304692, + "loss": 0.7639, + "step": 6385 + }, + { + "epoch": 1.1369301994301995, + "grad_norm": 0.5719713568687439, + "learning_rate": 0.00016286819982993782, + "loss": 0.9717, + "step": 6386 + }, + { + "epoch": 1.1371082621082622, + "grad_norm": 0.7028461694717407, + "learning_rate": 0.00016285731390505695, + "loss": 1.0147, + "step": 6387 + }, + { + "epoch": 1.1372863247863247, + "grad_norm": 0.5396828651428223, + "learning_rate": 0.00016284642674861756, + "loss": 0.8119, + "step": 6388 + }, + { + "epoch": 1.1374643874643875, + "grad_norm": 0.592580258846283, + "learning_rate": 0.00016283553836083303, + "loss": 1.0914, + "step": 6389 + }, + { + "epoch": 1.13764245014245, + "grad_norm": 0.634596586227417, + "learning_rate": 0.00016282464874191663, + "loss": 1.1037, + "step": 6390 + }, + { + "epoch": 1.1378205128205128, + "grad_norm": 0.6462705731391907, + "learning_rate": 0.00016281375789208176, + "loss": 1.1523, + "step": 6391 + }, + { + "epoch": 1.1379985754985755, + "grad_norm": 0.6527917385101318, + "learning_rate": 0.0001628028658115418, + "loss": 1.0415, + "step": 6392 + }, + { + "epoch": 1.1381766381766383, + "grad_norm": 0.6309964060783386, + "learning_rate": 0.00016279197250051013, + "loss": 0.9747, + "step": 6393 + }, + { + "epoch": 1.1383547008547008, + "grad_norm": 0.6342993974685669, + "learning_rate": 0.00016278107795920018, + "loss": 0.9897, + "step": 6394 + }, + { + "epoch": 1.1385327635327636, + "grad_norm": 0.7149887084960938, + "learning_rate": 0.00016277018218782544, + "loss": 0.9659, + "step": 6395 + }, + { + "epoch": 1.138710826210826, + "grad_norm": 0.7219462394714355, + "learning_rate": 0.00016275928518659938, + "loss": 0.9301, + "step": 6396 + }, + { + "epoch": 1.1388888888888888, + "grad_norm": 0.6649485230445862, + "learning_rate": 0.0001627483869557355, + "loss": 0.9012, + "step": 6397 + }, + { + "epoch": 1.1390669515669516, + "grad_norm": 0.6910027861595154, + "learning_rate": 0.00016273748749544731, + "loss": 0.956, + "step": 6398 + }, + { + "epoch": 1.1392450142450143, + "grad_norm": 0.6369016766548157, + "learning_rate": 0.00016272658680594837, + "loss": 0.8027, + "step": 6399 + }, + { + "epoch": 1.1394230769230769, + "grad_norm": 0.6540524959564209, + "learning_rate": 0.00016271568488745227, + "loss": 1.2397, + "step": 6400 + }, + { + "epoch": 1.1396011396011396, + "grad_norm": 0.5912376046180725, + "learning_rate": 0.00016270478174017263, + "loss": 0.8453, + "step": 6401 + }, + { + "epoch": 1.1397792022792024, + "grad_norm": 0.6847240924835205, + "learning_rate": 0.00016269387736432303, + "loss": 0.9776, + "step": 6402 + }, + { + "epoch": 1.139957264957265, + "grad_norm": 0.6465024352073669, + "learning_rate": 0.00016268297176011716, + "loss": 0.8971, + "step": 6403 + }, + { + "epoch": 1.1401353276353277, + "grad_norm": 0.6639063954353333, + "learning_rate": 0.00016267206492776866, + "loss": 0.9756, + "step": 6404 + }, + { + "epoch": 1.1403133903133904, + "grad_norm": 0.6343763470649719, + "learning_rate": 0.00016266115686749123, + "loss": 0.9368, + "step": 6405 + }, + { + "epoch": 1.140491452991453, + "grad_norm": 0.7144993543624878, + "learning_rate": 0.0001626502475794986, + "loss": 0.9285, + "step": 6406 + }, + { + "epoch": 1.1406695156695157, + "grad_norm": 0.6217414736747742, + "learning_rate": 0.00016263933706400451, + "loss": 0.8867, + "step": 6407 + }, + { + "epoch": 1.1408475783475784, + "grad_norm": 0.6843730807304382, + "learning_rate": 0.00016262842532122274, + "loss": 0.9863, + "step": 6408 + }, + { + "epoch": 1.141025641025641, + "grad_norm": 0.6866166591644287, + "learning_rate": 0.00016261751235136705, + "loss": 1.0517, + "step": 6409 + }, + { + "epoch": 1.1412037037037037, + "grad_norm": 0.6650584936141968, + "learning_rate": 0.0001626065981546513, + "loss": 1.0629, + "step": 6410 + }, + { + "epoch": 1.1413817663817665, + "grad_norm": 0.5805012583732605, + "learning_rate": 0.00016259568273128933, + "loss": 0.8175, + "step": 6411 + }, + { + "epoch": 1.141559829059829, + "grad_norm": 0.7005903124809265, + "learning_rate": 0.00016258476608149497, + "loss": 1.0267, + "step": 6412 + }, + { + "epoch": 1.1417378917378918, + "grad_norm": 0.6293461322784424, + "learning_rate": 0.00016257384820548217, + "loss": 1.1034, + "step": 6413 + }, + { + "epoch": 1.1419159544159545, + "grad_norm": 0.6281774640083313, + "learning_rate": 0.00016256292910346476, + "loss": 1.0775, + "step": 6414 + }, + { + "epoch": 1.142094017094017, + "grad_norm": 0.5912862420082092, + "learning_rate": 0.0001625520087756567, + "loss": 0.9589, + "step": 6415 + }, + { + "epoch": 1.1422720797720798, + "grad_norm": 0.5813978314399719, + "learning_rate": 0.00016254108722227198, + "loss": 0.9195, + "step": 6416 + }, + { + "epoch": 1.1424501424501425, + "grad_norm": 0.650805652141571, + "learning_rate": 0.00016253016444352458, + "loss": 1.0207, + "step": 6417 + }, + { + "epoch": 1.142628205128205, + "grad_norm": 0.6909520030021667, + "learning_rate": 0.00016251924043962851, + "loss": 0.9854, + "step": 6418 + }, + { + "epoch": 1.1428062678062678, + "grad_norm": 0.6054595112800598, + "learning_rate": 0.0001625083152107978, + "loss": 0.852, + "step": 6419 + }, + { + "epoch": 1.1429843304843306, + "grad_norm": 0.601078987121582, + "learning_rate": 0.00016249738875724647, + "loss": 0.9609, + "step": 6420 + }, + { + "epoch": 1.143162393162393, + "grad_norm": 0.5340180397033691, + "learning_rate": 0.00016248646107918868, + "loss": 0.8364, + "step": 6421 + }, + { + "epoch": 1.1433404558404558, + "grad_norm": 0.6687821745872498, + "learning_rate": 0.00016247553217683846, + "loss": 1.005, + "step": 6422 + }, + { + "epoch": 1.1435185185185186, + "grad_norm": 0.6347902417182922, + "learning_rate": 0.00016246460205040998, + "loss": 1.026, + "step": 6423 + }, + { + "epoch": 1.1436965811965811, + "grad_norm": 0.6136734485626221, + "learning_rate": 0.00016245367070011736, + "loss": 0.7811, + "step": 6424 + }, + { + "epoch": 1.1438746438746439, + "grad_norm": 0.6591334342956543, + "learning_rate": 0.00016244273812617482, + "loss": 0.991, + "step": 6425 + }, + { + "epoch": 1.1440527065527066, + "grad_norm": 0.6062475442886353, + "learning_rate": 0.00016243180432879656, + "loss": 0.9879, + "step": 6426 + }, + { + "epoch": 1.1442307692307692, + "grad_norm": 0.5941380858421326, + "learning_rate": 0.00016242086930819678, + "loss": 0.9771, + "step": 6427 + }, + { + "epoch": 1.144408831908832, + "grad_norm": 0.7320533990859985, + "learning_rate": 0.00016240993306458973, + "loss": 1.0919, + "step": 6428 + }, + { + "epoch": 1.1445868945868947, + "grad_norm": 0.6998075246810913, + "learning_rate": 0.00016239899559818962, + "loss": 1.0721, + "step": 6429 + }, + { + "epoch": 1.1447649572649572, + "grad_norm": 0.847931444644928, + "learning_rate": 0.0001623880569092109, + "loss": 0.8759, + "step": 6430 + }, + { + "epoch": 1.14494301994302, + "grad_norm": 0.6670104265213013, + "learning_rate": 0.00016237711699786775, + "loss": 1.0515, + "step": 6431 + }, + { + "epoch": 1.1451210826210827, + "grad_norm": 0.601759672164917, + "learning_rate": 0.00016236617586437463, + "loss": 0.7298, + "step": 6432 + }, + { + "epoch": 1.1452991452991452, + "grad_norm": 0.6411594152450562, + "learning_rate": 0.00016235523350894578, + "loss": 0.9336, + "step": 6433 + }, + { + "epoch": 1.145477207977208, + "grad_norm": 0.6485120058059692, + "learning_rate": 0.0001623442899317957, + "loss": 1.1215, + "step": 6434 + }, + { + "epoch": 1.1456552706552707, + "grad_norm": 0.6041508316993713, + "learning_rate": 0.00016233334513313875, + "loss": 0.8917, + "step": 6435 + }, + { + "epoch": 1.1458333333333333, + "grad_norm": 0.6292745471000671, + "learning_rate": 0.0001623223991131894, + "loss": 0.9976, + "step": 6436 + }, + { + "epoch": 1.146011396011396, + "grad_norm": 0.5442200303077698, + "learning_rate": 0.0001623114518721621, + "loss": 0.8072, + "step": 6437 + }, + { + "epoch": 1.1461894586894588, + "grad_norm": 0.6668170094490051, + "learning_rate": 0.00016230050341027136, + "loss": 0.9641, + "step": 6438 + }, + { + "epoch": 1.1463675213675213, + "grad_norm": 0.644186794757843, + "learning_rate": 0.00016228955372773164, + "loss": 0.9248, + "step": 6439 + }, + { + "epoch": 1.146545584045584, + "grad_norm": 0.6661991477012634, + "learning_rate": 0.00016227860282475753, + "loss": 0.8719, + "step": 6440 + }, + { + "epoch": 1.1467236467236468, + "grad_norm": 0.5232062935829163, + "learning_rate": 0.00016226765070156355, + "loss": 0.5418, + "step": 6441 + }, + { + "epoch": 1.1469017094017093, + "grad_norm": 0.573176383972168, + "learning_rate": 0.00016225669735836436, + "loss": 1.0858, + "step": 6442 + }, + { + "epoch": 1.147079772079772, + "grad_norm": 0.6137439608573914, + "learning_rate": 0.00016224574279537446, + "loss": 1.1205, + "step": 6443 + }, + { + "epoch": 1.1472578347578348, + "grad_norm": 0.6328136920928955, + "learning_rate": 0.00016223478701280855, + "loss": 0.8957, + "step": 6444 + }, + { + "epoch": 1.1474358974358974, + "grad_norm": 0.6687374114990234, + "learning_rate": 0.00016222383001088126, + "loss": 1.0318, + "step": 6445 + }, + { + "epoch": 1.14761396011396, + "grad_norm": 0.6057115793228149, + "learning_rate": 0.0001622128717898073, + "loss": 0.9575, + "step": 6446 + }, + { + "epoch": 1.1477920227920229, + "grad_norm": 0.6758735775947571, + "learning_rate": 0.0001622019123498013, + "loss": 1.2273, + "step": 6447 + }, + { + "epoch": 1.1479700854700854, + "grad_norm": 0.6233550310134888, + "learning_rate": 0.0001621909516910781, + "loss": 0.7875, + "step": 6448 + }, + { + "epoch": 1.1481481481481481, + "grad_norm": 0.6371827721595764, + "learning_rate": 0.0001621799898138524, + "loss": 1.0488, + "step": 6449 + }, + { + "epoch": 1.148326210826211, + "grad_norm": 0.6179831624031067, + "learning_rate": 0.00016216902671833892, + "loss": 0.9792, + "step": 6450 + }, + { + "epoch": 1.1485042735042734, + "grad_norm": 0.6234193444252014, + "learning_rate": 0.00016215806240475256, + "loss": 0.927, + "step": 6451 + }, + { + "epoch": 1.1486823361823362, + "grad_norm": 0.6940563917160034, + "learning_rate": 0.00016214709687330803, + "loss": 1.047, + "step": 6452 + }, + { + "epoch": 1.148860398860399, + "grad_norm": 0.6567606925964355, + "learning_rate": 0.00016213613012422027, + "loss": 0.9695, + "step": 6453 + }, + { + "epoch": 1.1490384615384615, + "grad_norm": 0.7374183535575867, + "learning_rate": 0.0001621251621577041, + "loss": 1.0443, + "step": 6454 + }, + { + "epoch": 1.1492165242165242, + "grad_norm": 0.6789869666099548, + "learning_rate": 0.00016211419297397443, + "loss": 1.0319, + "step": 6455 + }, + { + "epoch": 1.149394586894587, + "grad_norm": 0.6225521564483643, + "learning_rate": 0.00016210322257324619, + "loss": 1.0529, + "step": 6456 + }, + { + "epoch": 1.1495726495726495, + "grad_norm": 0.619701623916626, + "learning_rate": 0.00016209225095573432, + "loss": 0.962, + "step": 6457 + }, + { + "epoch": 1.1497507122507122, + "grad_norm": 0.6132834553718567, + "learning_rate": 0.00016208127812165375, + "loss": 0.9588, + "step": 6458 + }, + { + "epoch": 1.149928774928775, + "grad_norm": 0.6005367040634155, + "learning_rate": 0.00016207030407121954, + "loss": 0.9497, + "step": 6459 + }, + { + "epoch": 1.1501068376068375, + "grad_norm": 0.575309157371521, + "learning_rate": 0.00016205932880464664, + "loss": 1.0035, + "step": 6460 + }, + { + "epoch": 1.1502849002849003, + "grad_norm": 0.5958710312843323, + "learning_rate": 0.0001620483523221501, + "loss": 1.0004, + "step": 6461 + }, + { + "epoch": 1.150462962962963, + "grad_norm": 0.5934719443321228, + "learning_rate": 0.000162037374623945, + "loss": 0.8694, + "step": 6462 + }, + { + "epoch": 1.1506410256410255, + "grad_norm": 0.6042510271072388, + "learning_rate": 0.00016202639571024643, + "loss": 0.8598, + "step": 6463 + }, + { + "epoch": 1.1508190883190883, + "grad_norm": 0.6206158399581909, + "learning_rate": 0.00016201541558126946, + "loss": 0.961, + "step": 6464 + }, + { + "epoch": 1.150997150997151, + "grad_norm": 0.5997715592384338, + "learning_rate": 0.00016200443423722925, + "loss": 0.8686, + "step": 6465 + }, + { + "epoch": 1.1511752136752136, + "grad_norm": 0.742457926273346, + "learning_rate": 0.00016199345167834098, + "loss": 1.1113, + "step": 6466 + }, + { + "epoch": 1.1513532763532763, + "grad_norm": 0.6772766709327698, + "learning_rate": 0.00016198246790481976, + "loss": 1.0717, + "step": 6467 + }, + { + "epoch": 1.151531339031339, + "grad_norm": 0.6127712726593018, + "learning_rate": 0.0001619714829168809, + "loss": 0.8887, + "step": 6468 + }, + { + "epoch": 1.1517094017094016, + "grad_norm": 0.5585067272186279, + "learning_rate": 0.00016196049671473954, + "loss": 1.0144, + "step": 6469 + }, + { + "epoch": 1.1518874643874644, + "grad_norm": 0.6269431710243225, + "learning_rate": 0.00016194950929861092, + "loss": 1.0206, + "step": 6470 + }, + { + "epoch": 1.1520655270655271, + "grad_norm": 0.6270785331726074, + "learning_rate": 0.0001619385206687104, + "loss": 1.0517, + "step": 6471 + }, + { + "epoch": 1.1522435897435896, + "grad_norm": 0.744712233543396, + "learning_rate": 0.00016192753082525322, + "loss": 1.0699, + "step": 6472 + }, + { + "epoch": 1.1524216524216524, + "grad_norm": 0.7025929689407349, + "learning_rate": 0.00016191653976845474, + "loss": 0.951, + "step": 6473 + }, + { + "epoch": 1.1525997150997151, + "grad_norm": 0.6175379753112793, + "learning_rate": 0.00016190554749853024, + "loss": 1.2153, + "step": 6474 + }, + { + "epoch": 1.1527777777777777, + "grad_norm": 0.6212149858474731, + "learning_rate": 0.00016189455401569513, + "loss": 1.0428, + "step": 6475 + }, + { + "epoch": 1.1529558404558404, + "grad_norm": 0.6716817617416382, + "learning_rate": 0.00016188355932016484, + "loss": 1.179, + "step": 6476 + }, + { + "epoch": 1.1531339031339032, + "grad_norm": 0.6247739791870117, + "learning_rate": 0.00016187256341215476, + "loss": 0.9451, + "step": 6477 + }, + { + "epoch": 1.153311965811966, + "grad_norm": 0.6223008036613464, + "learning_rate": 0.00016186156629188032, + "loss": 0.9915, + "step": 6478 + }, + { + "epoch": 1.1534900284900285, + "grad_norm": 0.5610866546630859, + "learning_rate": 0.000161850567959557, + "loss": 0.7741, + "step": 6479 + }, + { + "epoch": 1.1536680911680912, + "grad_norm": 0.6241226196289062, + "learning_rate": 0.0001618395684154003, + "loss": 1.2193, + "step": 6480 + }, + { + "epoch": 1.1538461538461537, + "grad_norm": 0.703789472579956, + "learning_rate": 0.00016182856765962567, + "loss": 1.0725, + "step": 6481 + }, + { + "epoch": 1.1540242165242165, + "grad_norm": 0.6802006959915161, + "learning_rate": 0.00016181756569244872, + "loss": 1.0908, + "step": 6482 + }, + { + "epoch": 1.1542022792022792, + "grad_norm": 0.6504136919975281, + "learning_rate": 0.000161806562514085, + "loss": 0.9706, + "step": 6483 + }, + { + "epoch": 1.154380341880342, + "grad_norm": 0.7217034101486206, + "learning_rate": 0.00016179555812475003, + "loss": 0.9084, + "step": 6484 + }, + { + "epoch": 1.1545584045584045, + "grad_norm": 0.5919039249420166, + "learning_rate": 0.0001617845525246595, + "loss": 0.949, + "step": 6485 + }, + { + "epoch": 1.1547364672364673, + "grad_norm": 0.6160184741020203, + "learning_rate": 0.00016177354571402902, + "loss": 0.8144, + "step": 6486 + }, + { + "epoch": 1.1549145299145298, + "grad_norm": 0.7323806285858154, + "learning_rate": 0.00016176253769307426, + "loss": 1.0528, + "step": 6487 + }, + { + "epoch": 1.1550925925925926, + "grad_norm": 0.6051317453384399, + "learning_rate": 0.0001617515284620108, + "loss": 0.9558, + "step": 6488 + }, + { + "epoch": 1.1552706552706553, + "grad_norm": 0.6418905258178711, + "learning_rate": 0.00016174051802105447, + "loss": 1.062, + "step": 6489 + }, + { + "epoch": 1.155448717948718, + "grad_norm": 0.6914883852005005, + "learning_rate": 0.00016172950637042096, + "loss": 0.9999, + "step": 6490 + }, + { + "epoch": 1.1556267806267806, + "grad_norm": 0.5558316707611084, + "learning_rate": 0.000161718493510326, + "loss": 0.9561, + "step": 6491 + }, + { + "epoch": 1.1558048433048433, + "grad_norm": 0.6632496118545532, + "learning_rate": 0.00016170747944098531, + "loss": 1.0133, + "step": 6492 + }, + { + "epoch": 1.1559829059829059, + "grad_norm": 0.6407149434089661, + "learning_rate": 0.00016169646416261478, + "loss": 1.0563, + "step": 6493 + }, + { + "epoch": 1.1561609686609686, + "grad_norm": 0.8128494024276733, + "learning_rate": 0.0001616854476754302, + "loss": 1.1559, + "step": 6494 + }, + { + "epoch": 1.1563390313390314, + "grad_norm": 0.6403429508209229, + "learning_rate": 0.00016167442997964742, + "loss": 1.0983, + "step": 6495 + }, + { + "epoch": 1.1565170940170941, + "grad_norm": 0.76612788438797, + "learning_rate": 0.0001616634110754823, + "loss": 0.973, + "step": 6496 + }, + { + "epoch": 1.1566951566951567, + "grad_norm": 0.6914355754852295, + "learning_rate": 0.0001616523909631507, + "loss": 0.9307, + "step": 6497 + }, + { + "epoch": 1.1568732193732194, + "grad_norm": 0.546602725982666, + "learning_rate": 0.00016164136964286863, + "loss": 1.0328, + "step": 6498 + }, + { + "epoch": 1.157051282051282, + "grad_norm": 0.5695818662643433, + "learning_rate": 0.00016163034711485193, + "loss": 0.9607, + "step": 6499 + }, + { + "epoch": 1.1572293447293447, + "grad_norm": 0.5649738311767578, + "learning_rate": 0.00016161932337931662, + "loss": 1.1521, + "step": 6500 + }, + { + "epoch": 1.1574074074074074, + "grad_norm": 0.6437582969665527, + "learning_rate": 0.00016160829843647867, + "loss": 0.9613, + "step": 6501 + }, + { + "epoch": 1.1575854700854702, + "grad_norm": 0.5841929316520691, + "learning_rate": 0.0001615972722865541, + "loss": 0.8187, + "step": 6502 + }, + { + "epoch": 1.1577635327635327, + "grad_norm": 0.6481246948242188, + "learning_rate": 0.00016158624492975892, + "loss": 1.0447, + "step": 6503 + }, + { + "epoch": 1.1579415954415955, + "grad_norm": 0.629804790019989, + "learning_rate": 0.0001615752163663092, + "loss": 0.9034, + "step": 6504 + }, + { + "epoch": 1.158119658119658, + "grad_norm": 0.5797054171562195, + "learning_rate": 0.00016156418659642104, + "loss": 0.8168, + "step": 6505 + }, + { + "epoch": 1.1582977207977208, + "grad_norm": 0.588424563407898, + "learning_rate": 0.00016155315562031052, + "loss": 0.828, + "step": 6506 + }, + { + "epoch": 1.1584757834757835, + "grad_norm": 0.7120068669319153, + "learning_rate": 0.0001615421234381938, + "loss": 1.0637, + "step": 6507 + }, + { + "epoch": 1.1586538461538463, + "grad_norm": 0.6635081768035889, + "learning_rate": 0.00016153109005028702, + "loss": 0.9838, + "step": 6508 + }, + { + "epoch": 1.1588319088319088, + "grad_norm": 0.6080414056777954, + "learning_rate": 0.00016152005545680634, + "loss": 0.983, + "step": 6509 + }, + { + "epoch": 1.1590099715099715, + "grad_norm": 0.7131237983703613, + "learning_rate": 0.00016150901965796796, + "loss": 1.1053, + "step": 6510 + }, + { + "epoch": 1.159188034188034, + "grad_norm": 0.6051005125045776, + "learning_rate": 0.00016149798265398813, + "loss": 0.9903, + "step": 6511 + }, + { + "epoch": 1.1593660968660968, + "grad_norm": 0.6193733811378479, + "learning_rate": 0.00016148694444508306, + "loss": 1.0478, + "step": 6512 + }, + { + "epoch": 1.1595441595441596, + "grad_norm": 0.567888081073761, + "learning_rate": 0.00016147590503146905, + "loss": 0.7995, + "step": 6513 + }, + { + "epoch": 1.1597222222222223, + "grad_norm": 0.6889783143997192, + "learning_rate": 0.00016146486441336242, + "loss": 0.9684, + "step": 6514 + }, + { + "epoch": 1.1599002849002849, + "grad_norm": 0.6470308303833008, + "learning_rate": 0.0001614538225909794, + "loss": 0.9824, + "step": 6515 + }, + { + "epoch": 1.1600783475783476, + "grad_norm": 0.6833886504173279, + "learning_rate": 0.00016144277956453638, + "loss": 0.9845, + "step": 6516 + }, + { + "epoch": 1.1602564102564104, + "grad_norm": 0.5827815532684326, + "learning_rate": 0.00016143173533424978, + "loss": 0.9476, + "step": 6517 + }, + { + "epoch": 1.1604344729344729, + "grad_norm": 0.6701242327690125, + "learning_rate": 0.00016142068990033593, + "loss": 1.0839, + "step": 6518 + }, + { + "epoch": 1.1606125356125356, + "grad_norm": 0.5844996571540833, + "learning_rate": 0.00016140964326301122, + "loss": 0.8861, + "step": 6519 + }, + { + "epoch": 1.1607905982905984, + "grad_norm": 0.5831994414329529, + "learning_rate": 0.00016139859542249214, + "loss": 0.9817, + "step": 6520 + }, + { + "epoch": 1.160968660968661, + "grad_norm": 0.6830124855041504, + "learning_rate": 0.0001613875463789951, + "loss": 0.8749, + "step": 6521 + }, + { + "epoch": 1.1611467236467237, + "grad_norm": 0.6003018021583557, + "learning_rate": 0.00016137649613273667, + "loss": 0.9593, + "step": 6522 + }, + { + "epoch": 1.1613247863247864, + "grad_norm": 0.5973994731903076, + "learning_rate": 0.00016136544468393327, + "loss": 1.0384, + "step": 6523 + }, + { + "epoch": 1.161502849002849, + "grad_norm": 0.6702523827552795, + "learning_rate": 0.00016135439203280143, + "loss": 1.0431, + "step": 6524 + }, + { + "epoch": 1.1616809116809117, + "grad_norm": 0.6160697937011719, + "learning_rate": 0.00016134333817955775, + "loss": 1.0339, + "step": 6525 + }, + { + "epoch": 1.1618589743589745, + "grad_norm": 0.7078264355659485, + "learning_rate": 0.0001613322831244188, + "loss": 1.0285, + "step": 6526 + }, + { + "epoch": 1.162037037037037, + "grad_norm": 0.5744216442108154, + "learning_rate": 0.00016132122686760117, + "loss": 0.6589, + "step": 6527 + }, + { + "epoch": 1.1622150997150997, + "grad_norm": 0.6802098155021667, + "learning_rate": 0.00016131016940932146, + "loss": 0.9532, + "step": 6528 + }, + { + "epoch": 1.1623931623931625, + "grad_norm": 0.6523237228393555, + "learning_rate": 0.00016129911074979635, + "loss": 0.9409, + "step": 6529 + }, + { + "epoch": 1.162571225071225, + "grad_norm": 0.710307776927948, + "learning_rate": 0.00016128805088924252, + "loss": 1.2536, + "step": 6530 + }, + { + "epoch": 1.1627492877492878, + "grad_norm": 0.6349819898605347, + "learning_rate": 0.0001612769898278766, + "loss": 1.0857, + "step": 6531 + }, + { + "epoch": 1.1629273504273505, + "grad_norm": 0.5348139405250549, + "learning_rate": 0.00016126592756591542, + "loss": 0.5969, + "step": 6532 + }, + { + "epoch": 1.163105413105413, + "grad_norm": 0.635619580745697, + "learning_rate": 0.00016125486410357564, + "loss": 0.9885, + "step": 6533 + }, + { + "epoch": 1.1632834757834758, + "grad_norm": 0.6434559226036072, + "learning_rate": 0.000161243799441074, + "loss": 0.8377, + "step": 6534 + }, + { + "epoch": 1.1634615384615385, + "grad_norm": 0.6509647369384766, + "learning_rate": 0.00016123273357862737, + "loss": 0.8393, + "step": 6535 + }, + { + "epoch": 1.163639601139601, + "grad_norm": 0.6179081797599792, + "learning_rate": 0.0001612216665164525, + "loss": 0.9143, + "step": 6536 + }, + { + "epoch": 1.1638176638176638, + "grad_norm": 0.5923223495483398, + "learning_rate": 0.0001612105982547663, + "loss": 1.0185, + "step": 6537 + }, + { + "epoch": 1.1639957264957266, + "grad_norm": 0.702150285243988, + "learning_rate": 0.00016119952879378556, + "loss": 0.863, + "step": 6538 + }, + { + "epoch": 1.164173789173789, + "grad_norm": 0.6596643328666687, + "learning_rate": 0.00016118845813372715, + "loss": 1.0089, + "step": 6539 + }, + { + "epoch": 1.1643518518518519, + "grad_norm": 0.7675769329071045, + "learning_rate": 0.00016117738627480804, + "loss": 1.0179, + "step": 6540 + }, + { + "epoch": 1.1645299145299146, + "grad_norm": 0.6742541193962097, + "learning_rate": 0.00016116631321724513, + "loss": 1.0663, + "step": 6541 + }, + { + "epoch": 1.1647079772079771, + "grad_norm": 0.7379785776138306, + "learning_rate": 0.0001611552389612554, + "loss": 1.0162, + "step": 6542 + }, + { + "epoch": 1.16488603988604, + "grad_norm": 0.5729365944862366, + "learning_rate": 0.00016114416350705577, + "loss": 0.8146, + "step": 6543 + }, + { + "epoch": 1.1650641025641026, + "grad_norm": 0.6481349468231201, + "learning_rate": 0.00016113308685486327, + "loss": 1.0748, + "step": 6544 + }, + { + "epoch": 1.1652421652421652, + "grad_norm": 0.5588181018829346, + "learning_rate": 0.00016112200900489493, + "loss": 0.7511, + "step": 6545 + }, + { + "epoch": 1.165420227920228, + "grad_norm": 0.674363911151886, + "learning_rate": 0.0001611109299573678, + "loss": 0.9852, + "step": 6546 + }, + { + "epoch": 1.1655982905982907, + "grad_norm": 0.6712620854377747, + "learning_rate": 0.00016109984971249893, + "loss": 0.9558, + "step": 6547 + }, + { + "epoch": 1.1657763532763532, + "grad_norm": 0.5260626077651978, + "learning_rate": 0.00016108876827050544, + "loss": 0.7008, + "step": 6548 + }, + { + "epoch": 1.165954415954416, + "grad_norm": 0.6056292057037354, + "learning_rate": 0.00016107768563160445, + "loss": 0.7756, + "step": 6549 + }, + { + "epoch": 1.1661324786324787, + "grad_norm": 0.5725821256637573, + "learning_rate": 0.00016106660179601308, + "loss": 0.8228, + "step": 6550 + }, + { + "epoch": 1.1663105413105412, + "grad_norm": 0.6708397269248962, + "learning_rate": 0.00016105551676394848, + "loss": 1.0711, + "step": 6551 + }, + { + "epoch": 1.166488603988604, + "grad_norm": 0.645453155040741, + "learning_rate": 0.00016104443053562787, + "loss": 0.9299, + "step": 6552 + }, + { + "epoch": 1.1666666666666667, + "grad_norm": 0.6743524074554443, + "learning_rate": 0.00016103334311126847, + "loss": 0.8977, + "step": 6553 + }, + { + "epoch": 1.1668447293447293, + "grad_norm": 0.7248545289039612, + "learning_rate": 0.0001610222544910875, + "loss": 1.2135, + "step": 6554 + }, + { + "epoch": 1.167022792022792, + "grad_norm": 0.5798853635787964, + "learning_rate": 0.00016101116467530217, + "loss": 0.857, + "step": 6555 + }, + { + "epoch": 1.1672008547008548, + "grad_norm": 0.6828082799911499, + "learning_rate": 0.00016100007366412985, + "loss": 0.9405, + "step": 6556 + }, + { + "epoch": 1.1673789173789173, + "grad_norm": 0.6820163130760193, + "learning_rate": 0.0001609889814577878, + "loss": 0.9144, + "step": 6557 + }, + { + "epoch": 1.16755698005698, + "grad_norm": 0.6482275128364563, + "learning_rate": 0.00016097788805649333, + "loss": 0.8586, + "step": 6558 + }, + { + "epoch": 1.1677350427350428, + "grad_norm": 0.6404715180397034, + "learning_rate": 0.00016096679346046385, + "loss": 0.7018, + "step": 6559 + }, + { + "epoch": 1.1679131054131053, + "grad_norm": 0.6315203309059143, + "learning_rate": 0.0001609556976699167, + "loss": 0.9602, + "step": 6560 + }, + { + "epoch": 1.168091168091168, + "grad_norm": 0.5521387457847595, + "learning_rate": 0.00016094460068506925, + "loss": 0.9294, + "step": 6561 + }, + { + "epoch": 1.1682692307692308, + "grad_norm": 0.583372175693512, + "learning_rate": 0.00016093350250613895, + "loss": 1.077, + "step": 6562 + }, + { + "epoch": 1.1684472934472934, + "grad_norm": 0.5990512371063232, + "learning_rate": 0.00016092240313334325, + "loss": 1.0102, + "step": 6563 + }, + { + "epoch": 1.1686253561253561, + "grad_norm": 0.675128161907196, + "learning_rate": 0.00016091130256689964, + "loss": 1.0407, + "step": 6564 + }, + { + "epoch": 1.1688034188034189, + "grad_norm": 0.48797324299812317, + "learning_rate": 0.00016090020080702556, + "loss": 0.7821, + "step": 6565 + }, + { + "epoch": 1.1689814814814814, + "grad_norm": 0.7487484216690063, + "learning_rate": 0.00016088909785393857, + "loss": 1.0444, + "step": 6566 + }, + { + "epoch": 1.1691595441595442, + "grad_norm": 0.6288858652114868, + "learning_rate": 0.00016087799370785618, + "loss": 1.1854, + "step": 6567 + }, + { + "epoch": 1.169337606837607, + "grad_norm": 0.6639021635055542, + "learning_rate": 0.000160866888368996, + "loss": 0.9632, + "step": 6568 + }, + { + "epoch": 1.1695156695156694, + "grad_norm": 0.6553738713264465, + "learning_rate": 0.00016085578183757556, + "loss": 1.2765, + "step": 6569 + }, + { + "epoch": 1.1696937321937322, + "grad_norm": 0.7489066123962402, + "learning_rate": 0.00016084467411381248, + "loss": 1.0705, + "step": 6570 + }, + { + "epoch": 1.169871794871795, + "grad_norm": 0.7079828381538391, + "learning_rate": 0.00016083356519792444, + "loss": 0.8256, + "step": 6571 + }, + { + "epoch": 1.1700498575498575, + "grad_norm": 0.7065926790237427, + "learning_rate": 0.00016082245509012902, + "loss": 1.0439, + "step": 6572 + }, + { + "epoch": 1.1702279202279202, + "grad_norm": 0.6113346815109253, + "learning_rate": 0.00016081134379064395, + "loss": 0.9153, + "step": 6573 + }, + { + "epoch": 1.170405982905983, + "grad_norm": 0.6094171404838562, + "learning_rate": 0.0001608002312996869, + "loss": 0.9723, + "step": 6574 + }, + { + "epoch": 1.1705840455840455, + "grad_norm": 0.6208072900772095, + "learning_rate": 0.00016078911761747565, + "loss": 0.948, + "step": 6575 + }, + { + "epoch": 1.1707621082621082, + "grad_norm": 0.5736680626869202, + "learning_rate": 0.00016077800274422792, + "loss": 0.9155, + "step": 6576 + }, + { + "epoch": 1.170940170940171, + "grad_norm": 0.6793957948684692, + "learning_rate": 0.0001607668866801615, + "loss": 0.9574, + "step": 6577 + }, + { + "epoch": 1.1711182336182335, + "grad_norm": 0.6251805424690247, + "learning_rate": 0.00016075576942549413, + "loss": 1.0319, + "step": 6578 + }, + { + "epoch": 1.1712962962962963, + "grad_norm": 0.628882110118866, + "learning_rate": 0.0001607446509804437, + "loss": 0.9336, + "step": 6579 + }, + { + "epoch": 1.171474358974359, + "grad_norm": 0.6712356805801392, + "learning_rate": 0.000160733531345228, + "loss": 1.0958, + "step": 6580 + }, + { + "epoch": 1.1716524216524216, + "grad_norm": 0.599365770816803, + "learning_rate": 0.0001607224105200649, + "loss": 0.9814, + "step": 6581 + }, + { + "epoch": 1.1718304843304843, + "grad_norm": 0.5798245668411255, + "learning_rate": 0.00016071128850517235, + "loss": 1.0355, + "step": 6582 + }, + { + "epoch": 1.172008547008547, + "grad_norm": 0.7646229863166809, + "learning_rate": 0.00016070016530076817, + "loss": 0.9976, + "step": 6583 + }, + { + "epoch": 1.1721866096866096, + "grad_norm": 0.6371127367019653, + "learning_rate": 0.0001606890409070704, + "loss": 0.9588, + "step": 6584 + }, + { + "epoch": 1.1723646723646723, + "grad_norm": 0.6497066617012024, + "learning_rate": 0.0001606779153242969, + "loss": 0.8817, + "step": 6585 + }, + { + "epoch": 1.172542735042735, + "grad_norm": 0.7255781888961792, + "learning_rate": 0.0001606667885526657, + "loss": 1.1319, + "step": 6586 + }, + { + "epoch": 1.1727207977207976, + "grad_norm": 0.67711341381073, + "learning_rate": 0.00016065566059239483, + "loss": 1.0755, + "step": 6587 + }, + { + "epoch": 1.1728988603988604, + "grad_norm": 0.6159650087356567, + "learning_rate": 0.00016064453144370227, + "loss": 0.9892, + "step": 6588 + }, + { + "epoch": 1.1730769230769231, + "grad_norm": 0.658938467502594, + "learning_rate": 0.00016063340110680609, + "loss": 0.9131, + "step": 6589 + }, + { + "epoch": 1.1732549857549857, + "grad_norm": 0.6754795908927917, + "learning_rate": 0.00016062226958192438, + "loss": 1.0119, + "step": 6590 + }, + { + "epoch": 1.1734330484330484, + "grad_norm": 0.6453405022621155, + "learning_rate": 0.00016061113686927523, + "loss": 0.997, + "step": 6591 + }, + { + "epoch": 1.1736111111111112, + "grad_norm": 0.6580284237861633, + "learning_rate": 0.00016060000296907675, + "loss": 0.8432, + "step": 6592 + }, + { + "epoch": 1.173789173789174, + "grad_norm": 0.6588153839111328, + "learning_rate": 0.00016058886788154712, + "loss": 1.0725, + "step": 6593 + }, + { + "epoch": 1.1739672364672364, + "grad_norm": 0.6247910857200623, + "learning_rate": 0.00016057773160690447, + "loss": 0.8736, + "step": 6594 + }, + { + "epoch": 1.1741452991452992, + "grad_norm": 0.579594075679779, + "learning_rate": 0.000160566594145367, + "loss": 0.8809, + "step": 6595 + }, + { + "epoch": 1.1743233618233617, + "grad_norm": 0.6738116145133972, + "learning_rate": 0.00016055545549715293, + "loss": 0.825, + "step": 6596 + }, + { + "epoch": 1.1745014245014245, + "grad_norm": 0.6658982634544373, + "learning_rate": 0.00016054431566248054, + "loss": 1.0809, + "step": 6597 + }, + { + "epoch": 1.1746794871794872, + "grad_norm": 0.5367915630340576, + "learning_rate": 0.00016053317464156803, + "loss": 0.9005, + "step": 6598 + }, + { + "epoch": 1.17485754985755, + "grad_norm": 0.7243228554725647, + "learning_rate": 0.00016052203243463372, + "loss": 1.0573, + "step": 6599 + }, + { + "epoch": 1.1750356125356125, + "grad_norm": 0.6359432935714722, + "learning_rate": 0.0001605108890418959, + "loss": 0.8569, + "step": 6600 + }, + { + "epoch": 1.1752136752136753, + "grad_norm": 0.6565225720405579, + "learning_rate": 0.0001604997444635729, + "loss": 0.9748, + "step": 6601 + }, + { + "epoch": 1.1753917378917378, + "grad_norm": 0.7124663591384888, + "learning_rate": 0.0001604885986998831, + "loss": 1.0271, + "step": 6602 + }, + { + "epoch": 1.1755698005698005, + "grad_norm": 0.659766435623169, + "learning_rate": 0.00016047745175104487, + "loss": 1.0635, + "step": 6603 + }, + { + "epoch": 1.1757478632478633, + "grad_norm": 0.5874318480491638, + "learning_rate": 0.00016046630361727656, + "loss": 0.9257, + "step": 6604 + }, + { + "epoch": 1.175925925925926, + "grad_norm": 0.587345540523529, + "learning_rate": 0.0001604551542987967, + "loss": 1.0759, + "step": 6605 + }, + { + "epoch": 1.1761039886039886, + "grad_norm": 0.733567476272583, + "learning_rate": 0.00016044400379582364, + "loss": 0.9877, + "step": 6606 + }, + { + "epoch": 1.1762820512820513, + "grad_norm": 0.6538317203521729, + "learning_rate": 0.0001604328521085759, + "loss": 1.0094, + "step": 6607 + }, + { + "epoch": 1.1764601139601139, + "grad_norm": 0.6279696822166443, + "learning_rate": 0.00016042169923727195, + "loss": 1.1049, + "step": 6608 + }, + { + "epoch": 1.1766381766381766, + "grad_norm": 0.6949752569198608, + "learning_rate": 0.00016041054518213033, + "loss": 1.1418, + "step": 6609 + }, + { + "epoch": 1.1768162393162394, + "grad_norm": 0.6144010424613953, + "learning_rate": 0.00016039938994336957, + "loss": 1.0306, + "step": 6610 + }, + { + "epoch": 1.176994301994302, + "grad_norm": 0.5868683457374573, + "learning_rate": 0.00016038823352120823, + "loss": 0.9894, + "step": 6611 + }, + { + "epoch": 1.1771723646723646, + "grad_norm": 0.7181115746498108, + "learning_rate": 0.0001603770759158649, + "loss": 1.1674, + "step": 6612 + }, + { + "epoch": 1.1773504273504274, + "grad_norm": 0.6271308064460754, + "learning_rate": 0.00016036591712755818, + "loss": 0.9726, + "step": 6613 + }, + { + "epoch": 1.17752849002849, + "grad_norm": 0.6922675371170044, + "learning_rate": 0.00016035475715650668, + "loss": 0.9142, + "step": 6614 + }, + { + "epoch": 1.1777065527065527, + "grad_norm": 0.6838833689689636, + "learning_rate": 0.00016034359600292913, + "loss": 1.1627, + "step": 6615 + }, + { + "epoch": 1.1778846153846154, + "grad_norm": 0.6628252267837524, + "learning_rate": 0.00016033243366704418, + "loss": 0.739, + "step": 6616 + }, + { + "epoch": 1.1780626780626782, + "grad_norm": 0.6367576122283936, + "learning_rate": 0.0001603212701490705, + "loss": 0.9015, + "step": 6617 + }, + { + "epoch": 1.1782407407407407, + "grad_norm": 0.6498967409133911, + "learning_rate": 0.00016031010544922687, + "loss": 0.9645, + "step": 6618 + }, + { + "epoch": 1.1784188034188035, + "grad_norm": 0.468795508146286, + "learning_rate": 0.00016029893956773198, + "loss": 0.7305, + "step": 6619 + }, + { + "epoch": 1.178596866096866, + "grad_norm": 0.6355500817298889, + "learning_rate": 0.00016028777250480465, + "loss": 0.9183, + "step": 6620 + }, + { + "epoch": 1.1787749287749287, + "grad_norm": 0.7582615613937378, + "learning_rate": 0.0001602766042606636, + "loss": 1.1641, + "step": 6621 + }, + { + "epoch": 1.1789529914529915, + "grad_norm": 0.580035924911499, + "learning_rate": 0.00016026543483552776, + "loss": 0.9164, + "step": 6622 + }, + { + "epoch": 1.1791310541310542, + "grad_norm": 0.6198559999465942, + "learning_rate": 0.00016025426422961592, + "loss": 0.9803, + "step": 6623 + }, + { + "epoch": 1.1793091168091168, + "grad_norm": 0.59112149477005, + "learning_rate": 0.0001602430924431469, + "loss": 0.8645, + "step": 6624 + }, + { + "epoch": 1.1794871794871795, + "grad_norm": 0.6200533509254456, + "learning_rate": 0.00016023191947633965, + "loss": 1.068, + "step": 6625 + }, + { + "epoch": 1.179665242165242, + "grad_norm": 0.6077516078948975, + "learning_rate": 0.00016022074532941305, + "loss": 1.0017, + "step": 6626 + }, + { + "epoch": 1.1798433048433048, + "grad_norm": 0.6770145893096924, + "learning_rate": 0.00016020957000258606, + "loss": 0.9022, + "step": 6627 + }, + { + "epoch": 1.1800213675213675, + "grad_norm": 0.6478054523468018, + "learning_rate": 0.0001601983934960776, + "loss": 0.8615, + "step": 6628 + }, + { + "epoch": 1.1801994301994303, + "grad_norm": 0.6528988480567932, + "learning_rate": 0.00016018721581010666, + "loss": 1.0015, + "step": 6629 + }, + { + "epoch": 1.1803774928774928, + "grad_norm": 0.6160712242126465, + "learning_rate": 0.0001601760369448923, + "loss": 0.9382, + "step": 6630 + }, + { + "epoch": 1.1805555555555556, + "grad_norm": 0.5755789875984192, + "learning_rate": 0.00016016485690065345, + "loss": 1.0551, + "step": 6631 + }, + { + "epoch": 1.180733618233618, + "grad_norm": 0.8495022654533386, + "learning_rate": 0.00016015367567760925, + "loss": 0.9295, + "step": 6632 + }, + { + "epoch": 1.1809116809116809, + "grad_norm": 0.6010929346084595, + "learning_rate": 0.0001601424932759787, + "loss": 1.0413, + "step": 6633 + }, + { + "epoch": 1.1810897435897436, + "grad_norm": 0.6953579187393188, + "learning_rate": 0.00016013130969598093, + "loss": 1.0149, + "step": 6634 + }, + { + "epoch": 1.1812678062678064, + "grad_norm": 0.6949529647827148, + "learning_rate": 0.0001601201249378351, + "loss": 0.9992, + "step": 6635 + }, + { + "epoch": 1.181445868945869, + "grad_norm": 0.6471893787384033, + "learning_rate": 0.00016010893900176028, + "loss": 0.7985, + "step": 6636 + }, + { + "epoch": 1.1816239316239316, + "grad_norm": 0.6524858474731445, + "learning_rate": 0.00016009775188797568, + "loss": 0.9517, + "step": 6637 + }, + { + "epoch": 1.1818019943019944, + "grad_norm": 0.639214038848877, + "learning_rate": 0.00016008656359670046, + "loss": 1.0357, + "step": 6638 + }, + { + "epoch": 1.181980056980057, + "grad_norm": 0.6039628386497498, + "learning_rate": 0.00016007537412815386, + "loss": 1.0536, + "step": 6639 + }, + { + "epoch": 1.1821581196581197, + "grad_norm": 0.653540313243866, + "learning_rate": 0.00016006418348255507, + "loss": 0.9414, + "step": 6640 + }, + { + "epoch": 1.1823361823361824, + "grad_norm": 0.6331741809844971, + "learning_rate": 0.0001600529916601234, + "loss": 1.0352, + "step": 6641 + }, + { + "epoch": 1.182514245014245, + "grad_norm": 0.7552719712257385, + "learning_rate": 0.00016004179866107812, + "loss": 1.1103, + "step": 6642 + }, + { + "epoch": 1.1826923076923077, + "grad_norm": 0.6795875430107117, + "learning_rate": 0.00016003060448563852, + "loss": 1.1246, + "step": 6643 + }, + { + "epoch": 1.1828703703703705, + "grad_norm": 0.6308842301368713, + "learning_rate": 0.0001600194091340239, + "loss": 0.9532, + "step": 6644 + }, + { + "epoch": 1.183048433048433, + "grad_norm": 0.5640553832054138, + "learning_rate": 0.00016000821260645366, + "loss": 0.7491, + "step": 6645 + }, + { + "epoch": 1.1832264957264957, + "grad_norm": 0.5611832141876221, + "learning_rate": 0.00015999701490314712, + "loss": 0.9239, + "step": 6646 + }, + { + "epoch": 1.1834045584045585, + "grad_norm": 0.5881187915802002, + "learning_rate": 0.00015998581602432374, + "loss": 0.9246, + "step": 6647 + }, + { + "epoch": 1.183582621082621, + "grad_norm": 0.7291010022163391, + "learning_rate": 0.00015997461597020291, + "loss": 1.0314, + "step": 6648 + }, + { + "epoch": 1.1837606837606838, + "grad_norm": 0.6784794926643372, + "learning_rate": 0.00015996341474100402, + "loss": 1.0011, + "step": 6649 + }, + { + "epoch": 1.1839387464387465, + "grad_norm": 0.7083746194839478, + "learning_rate": 0.00015995221233694663, + "loss": 1.0336, + "step": 6650 + }, + { + "epoch": 1.184116809116809, + "grad_norm": 0.7081790566444397, + "learning_rate": 0.00015994100875825015, + "loss": 1.2386, + "step": 6651 + }, + { + "epoch": 1.1842948717948718, + "grad_norm": 0.5938812494277954, + "learning_rate": 0.00015992980400513415, + "loss": 0.7549, + "step": 6652 + }, + { + "epoch": 1.1844729344729346, + "grad_norm": 0.7084267139434814, + "learning_rate": 0.00015991859807781811, + "loss": 1.1194, + "step": 6653 + }, + { + "epoch": 1.184650997150997, + "grad_norm": 0.6391362547874451, + "learning_rate": 0.0001599073909765216, + "loss": 1.0857, + "step": 6654 + }, + { + "epoch": 1.1848290598290598, + "grad_norm": 0.8074106574058533, + "learning_rate": 0.00015989618270146423, + "loss": 1.1715, + "step": 6655 + }, + { + "epoch": 1.1850071225071226, + "grad_norm": 0.5778565406799316, + "learning_rate": 0.0001598849732528656, + "loss": 0.8843, + "step": 6656 + }, + { + "epoch": 1.1851851851851851, + "grad_norm": 0.6955079436302185, + "learning_rate": 0.00015987376263094526, + "loss": 1.0281, + "step": 6657 + }, + { + "epoch": 1.1853632478632479, + "grad_norm": 0.6789296269416809, + "learning_rate": 0.00015986255083592297, + "loss": 0.9739, + "step": 6658 + }, + { + "epoch": 1.1855413105413106, + "grad_norm": 0.6294292211532593, + "learning_rate": 0.00015985133786801834, + "loss": 1.0692, + "step": 6659 + }, + { + "epoch": 1.1857193732193732, + "grad_norm": 0.5604581832885742, + "learning_rate": 0.00015984012372745107, + "loss": 0.9059, + "step": 6660 + }, + { + "epoch": 1.185897435897436, + "grad_norm": 0.6727550625801086, + "learning_rate": 0.00015982890841444088, + "loss": 1.049, + "step": 6661 + }, + { + "epoch": 1.1860754985754987, + "grad_norm": 0.620914101600647, + "learning_rate": 0.0001598176919292075, + "loss": 1.1021, + "step": 6662 + }, + { + "epoch": 1.1862535612535612, + "grad_norm": 0.6696683168411255, + "learning_rate": 0.00015980647427197076, + "loss": 0.9053, + "step": 6663 + }, + { + "epoch": 1.186431623931624, + "grad_norm": 0.6713385581970215, + "learning_rate": 0.00015979525544295036, + "loss": 0.9596, + "step": 6664 + }, + { + "epoch": 1.1866096866096867, + "grad_norm": 0.7643477320671082, + "learning_rate": 0.00015978403544236614, + "loss": 0.882, + "step": 6665 + }, + { + "epoch": 1.1867877492877492, + "grad_norm": 0.5890966057777405, + "learning_rate": 0.00015977281427043794, + "loss": 1.0215, + "step": 6666 + }, + { + "epoch": 1.186965811965812, + "grad_norm": 0.7287502288818359, + "learning_rate": 0.0001597615919273856, + "loss": 1.0111, + "step": 6667 + }, + { + "epoch": 1.1871438746438747, + "grad_norm": 0.5713803172111511, + "learning_rate": 0.00015975036841342903, + "loss": 1.0068, + "step": 6668 + }, + { + "epoch": 1.1873219373219372, + "grad_norm": 0.5113094449043274, + "learning_rate": 0.0001597391437287881, + "loss": 0.9018, + "step": 6669 + }, + { + "epoch": 1.1875, + "grad_norm": 0.585640013217926, + "learning_rate": 0.00015972791787368276, + "loss": 1.0375, + "step": 6670 + }, + { + "epoch": 1.1876780626780628, + "grad_norm": 0.5778326392173767, + "learning_rate": 0.00015971669084833293, + "loss": 0.9975, + "step": 6671 + }, + { + "epoch": 1.1878561253561253, + "grad_norm": 0.6707763075828552, + "learning_rate": 0.0001597054626529586, + "loss": 1.0048, + "step": 6672 + }, + { + "epoch": 1.188034188034188, + "grad_norm": 0.6113292574882507, + "learning_rate": 0.00015969423328777974, + "loss": 1.1447, + "step": 6673 + }, + { + "epoch": 1.1882122507122508, + "grad_norm": 0.6075651049613953, + "learning_rate": 0.00015968300275301638, + "loss": 0.9212, + "step": 6674 + }, + { + "epoch": 1.1883903133903133, + "grad_norm": 0.6990494132041931, + "learning_rate": 0.00015967177104888857, + "loss": 0.9952, + "step": 6675 + }, + { + "epoch": 1.188568376068376, + "grad_norm": 0.6228706240653992, + "learning_rate": 0.00015966053817561638, + "loss": 1.0187, + "step": 6676 + }, + { + "epoch": 1.1887464387464388, + "grad_norm": 0.6387844085693359, + "learning_rate": 0.00015964930413341985, + "loss": 1.1614, + "step": 6677 + }, + { + "epoch": 1.1889245014245013, + "grad_norm": 0.6501925587654114, + "learning_rate": 0.00015963806892251915, + "loss": 1.0366, + "step": 6678 + }, + { + "epoch": 1.189102564102564, + "grad_norm": 0.6923739910125732, + "learning_rate": 0.00015962683254313435, + "loss": 1.1992, + "step": 6679 + }, + { + "epoch": 1.1892806267806268, + "grad_norm": 0.6640275120735168, + "learning_rate": 0.00015961559499548563, + "loss": 0.8883, + "step": 6680 + }, + { + "epoch": 1.1894586894586894, + "grad_norm": 0.6493857502937317, + "learning_rate": 0.00015960435627979317, + "loss": 1.1368, + "step": 6681 + }, + { + "epoch": 1.1896367521367521, + "grad_norm": 0.6357189416885376, + "learning_rate": 0.0001595931163962772, + "loss": 1.0502, + "step": 6682 + }, + { + "epoch": 1.1898148148148149, + "grad_norm": 0.5756343007087708, + "learning_rate": 0.0001595818753451579, + "loss": 0.9871, + "step": 6683 + }, + { + "epoch": 1.1899928774928774, + "grad_norm": 0.7369210124015808, + "learning_rate": 0.0001595706331266555, + "loss": 1.3229, + "step": 6684 + }, + { + "epoch": 1.1901709401709402, + "grad_norm": 0.7140820622444153, + "learning_rate": 0.0001595593897409903, + "loss": 1.1154, + "step": 6685 + }, + { + "epoch": 1.190349002849003, + "grad_norm": 0.696973443031311, + "learning_rate": 0.00015954814518838255, + "loss": 0.9806, + "step": 6686 + }, + { + "epoch": 1.1905270655270654, + "grad_norm": 0.5299260020256042, + "learning_rate": 0.00015953689946905262, + "loss": 0.771, + "step": 6687 + }, + { + "epoch": 1.1907051282051282, + "grad_norm": 0.6814879775047302, + "learning_rate": 0.00015952565258322085, + "loss": 0.8444, + "step": 6688 + }, + { + "epoch": 1.190883190883191, + "grad_norm": 0.6215870976448059, + "learning_rate": 0.00015951440453110754, + "loss": 1.0743, + "step": 6689 + }, + { + "epoch": 1.1910612535612535, + "grad_norm": 0.7017203569412231, + "learning_rate": 0.00015950315531293308, + "loss": 1.185, + "step": 6690 + }, + { + "epoch": 1.1912393162393162, + "grad_norm": 0.7147250175476074, + "learning_rate": 0.00015949190492891795, + "loss": 1.0646, + "step": 6691 + }, + { + "epoch": 1.191417378917379, + "grad_norm": 0.5867117047309875, + "learning_rate": 0.00015948065337928252, + "loss": 1.0554, + "step": 6692 + }, + { + "epoch": 1.1915954415954415, + "grad_norm": 0.6813527345657349, + "learning_rate": 0.0001594694006642472, + "loss": 1.1451, + "step": 6693 + }, + { + "epoch": 1.1917735042735043, + "grad_norm": 0.5192593932151794, + "learning_rate": 0.00015945814678403256, + "loss": 0.7886, + "step": 6694 + }, + { + "epoch": 1.191951566951567, + "grad_norm": 0.6537744402885437, + "learning_rate": 0.00015944689173885904, + "loss": 0.9905, + "step": 6695 + }, + { + "epoch": 1.1921296296296295, + "grad_norm": 0.7350276112556458, + "learning_rate": 0.00015943563552894716, + "loss": 0.9009, + "step": 6696 + }, + { + "epoch": 1.1923076923076923, + "grad_norm": 0.7086381316184998, + "learning_rate": 0.00015942437815451746, + "loss": 0.9117, + "step": 6697 + }, + { + "epoch": 1.192485754985755, + "grad_norm": 0.6774969696998596, + "learning_rate": 0.00015941311961579054, + "loss": 1.1172, + "step": 6698 + }, + { + "epoch": 1.1926638176638176, + "grad_norm": 0.7034362554550171, + "learning_rate": 0.00015940185991298694, + "loss": 0.8054, + "step": 6699 + }, + { + "epoch": 1.1928418803418803, + "grad_norm": 0.66145920753479, + "learning_rate": 0.00015939059904632728, + "loss": 0.7417, + "step": 6700 + }, + { + "epoch": 1.193019943019943, + "grad_norm": 0.6590890884399414, + "learning_rate": 0.00015937933701603223, + "loss": 0.9169, + "step": 6701 + }, + { + "epoch": 1.1931980056980056, + "grad_norm": 0.7492850422859192, + "learning_rate": 0.0001593680738223224, + "loss": 1.0529, + "step": 6702 + }, + { + "epoch": 1.1933760683760684, + "grad_norm": 0.7103236317634583, + "learning_rate": 0.00015935680946541848, + "loss": 1.1377, + "step": 6703 + }, + { + "epoch": 1.193554131054131, + "grad_norm": 0.6164175868034363, + "learning_rate": 0.00015934554394554122, + "loss": 0.8636, + "step": 6704 + }, + { + "epoch": 1.1937321937321936, + "grad_norm": 0.6667410135269165, + "learning_rate": 0.0001593342772629113, + "loss": 1.0073, + "step": 6705 + }, + { + "epoch": 1.1939102564102564, + "grad_norm": 0.6785695552825928, + "learning_rate": 0.00015932300941774944, + "loss": 1.0752, + "step": 6706 + }, + { + "epoch": 1.1940883190883191, + "grad_norm": 0.6446872353553772, + "learning_rate": 0.0001593117404102765, + "loss": 0.9509, + "step": 6707 + }, + { + "epoch": 1.194266381766382, + "grad_norm": 0.6607686877250671, + "learning_rate": 0.00015930047024071317, + "loss": 1.0902, + "step": 6708 + }, + { + "epoch": 1.1944444444444444, + "grad_norm": 0.664804995059967, + "learning_rate": 0.0001592891989092803, + "loss": 0.9783, + "step": 6709 + }, + { + "epoch": 1.1946225071225072, + "grad_norm": 0.7147907018661499, + "learning_rate": 0.00015927792641619876, + "loss": 1.0558, + "step": 6710 + }, + { + "epoch": 1.1948005698005697, + "grad_norm": 0.6858944296836853, + "learning_rate": 0.0001592666527616894, + "loss": 1.0514, + "step": 6711 + }, + { + "epoch": 1.1949786324786325, + "grad_norm": 0.598463773727417, + "learning_rate": 0.0001592553779459731, + "loss": 0.8927, + "step": 6712 + }, + { + "epoch": 1.1951566951566952, + "grad_norm": 0.6872668862342834, + "learning_rate": 0.00015924410196927076, + "loss": 1.016, + "step": 6713 + }, + { + "epoch": 1.195334757834758, + "grad_norm": 0.6547996401786804, + "learning_rate": 0.00015923282483180326, + "loss": 1.1573, + "step": 6714 + }, + { + "epoch": 1.1955128205128205, + "grad_norm": 0.6254705786705017, + "learning_rate": 0.00015922154653379167, + "loss": 1.0179, + "step": 6715 + }, + { + "epoch": 1.1956908831908832, + "grad_norm": 0.6049207448959351, + "learning_rate": 0.00015921026707545684, + "loss": 1.0713, + "step": 6716 + }, + { + "epoch": 1.1958689458689458, + "grad_norm": 0.6042858958244324, + "learning_rate": 0.0001591989864570199, + "loss": 0.919, + "step": 6717 + }, + { + "epoch": 1.1960470085470085, + "grad_norm": 0.6521187424659729, + "learning_rate": 0.0001591877046787017, + "loss": 1.0112, + "step": 6718 + }, + { + "epoch": 1.1962250712250713, + "grad_norm": 0.766260027885437, + "learning_rate": 0.00015917642174072348, + "loss": 0.9774, + "step": 6719 + }, + { + "epoch": 1.196403133903134, + "grad_norm": 0.7066532373428345, + "learning_rate": 0.00015916513764330613, + "loss": 1.1112, + "step": 6720 + }, + { + "epoch": 1.1965811965811965, + "grad_norm": 0.7351508140563965, + "learning_rate": 0.00015915385238667083, + "loss": 0.9841, + "step": 6721 + }, + { + "epoch": 1.1967592592592593, + "grad_norm": 0.6133812069892883, + "learning_rate": 0.0001591425659710387, + "loss": 0.8629, + "step": 6722 + }, + { + "epoch": 1.1969373219373218, + "grad_norm": 0.7244157791137695, + "learning_rate": 0.00015913127839663083, + "loss": 1.1584, + "step": 6723 + }, + { + "epoch": 1.1971153846153846, + "grad_norm": 0.5986210107803345, + "learning_rate": 0.00015911998966366842, + "loss": 0.8507, + "step": 6724 + }, + { + "epoch": 1.1972934472934473, + "grad_norm": 0.6087439060211182, + "learning_rate": 0.00015910869977237257, + "loss": 0.884, + "step": 6725 + }, + { + "epoch": 1.19747150997151, + "grad_norm": 0.7546007633209229, + "learning_rate": 0.00015909740872296457, + "loss": 1.1449, + "step": 6726 + }, + { + "epoch": 1.1976495726495726, + "grad_norm": 0.6437731385231018, + "learning_rate": 0.0001590861165156656, + "loss": 0.7845, + "step": 6727 + }, + { + "epoch": 1.1978276353276354, + "grad_norm": 0.6281737089157104, + "learning_rate": 0.00015907482315069693, + "loss": 0.8969, + "step": 6728 + }, + { + "epoch": 1.198005698005698, + "grad_norm": 0.6196113228797913, + "learning_rate": 0.00015906352862827983, + "loss": 1.0264, + "step": 6729 + }, + { + "epoch": 1.1981837606837606, + "grad_norm": 0.5990965962409973, + "learning_rate": 0.00015905223294863553, + "loss": 1.0017, + "step": 6730 + }, + { + "epoch": 1.1983618233618234, + "grad_norm": 0.6509191393852234, + "learning_rate": 0.00015904093611198542, + "loss": 1.1066, + "step": 6731 + }, + { + "epoch": 1.1985398860398861, + "grad_norm": 0.6648043990135193, + "learning_rate": 0.00015902963811855085, + "loss": 1.077, + "step": 6732 + }, + { + "epoch": 1.1987179487179487, + "grad_norm": 0.7071963548660278, + "learning_rate": 0.00015901833896855307, + "loss": 1.1346, + "step": 6733 + }, + { + "epoch": 1.1988960113960114, + "grad_norm": 0.5889959335327148, + "learning_rate": 0.0001590070386622136, + "loss": 0.9525, + "step": 6734 + }, + { + "epoch": 1.199074074074074, + "grad_norm": 0.6233037710189819, + "learning_rate": 0.00015899573719975376, + "loss": 1.0513, + "step": 6735 + }, + { + "epoch": 1.1992521367521367, + "grad_norm": 0.7912302613258362, + "learning_rate": 0.000158984434581395, + "loss": 0.8749, + "step": 6736 + }, + { + "epoch": 1.1994301994301995, + "grad_norm": 0.5783160924911499, + "learning_rate": 0.0001589731308073588, + "loss": 0.7173, + "step": 6737 + }, + { + "epoch": 1.1996082621082622, + "grad_norm": 0.718950092792511, + "learning_rate": 0.00015896182587786658, + "loss": 1.0815, + "step": 6738 + }, + { + "epoch": 1.1997863247863247, + "grad_norm": 0.6700926423072815, + "learning_rate": 0.0001589505197931399, + "loss": 1.0817, + "step": 6739 + }, + { + "epoch": 1.1999643874643875, + "grad_norm": 0.7614455223083496, + "learning_rate": 0.0001589392125534002, + "loss": 0.9707, + "step": 6740 + }, + { + "epoch": 1.20014245014245, + "grad_norm": 0.6998619437217712, + "learning_rate": 0.00015892790415886906, + "loss": 1.0541, + "step": 6741 + }, + { + "epoch": 1.2003205128205128, + "grad_norm": 0.6127668619155884, + "learning_rate": 0.0001589165946097681, + "loss": 0.9147, + "step": 6742 + }, + { + "epoch": 1.2004985754985755, + "grad_norm": 0.7112005352973938, + "learning_rate": 0.00015890528390631885, + "loss": 0.868, + "step": 6743 + }, + { + "epoch": 1.2006766381766383, + "grad_norm": 0.6631024479866028, + "learning_rate": 0.0001588939720487429, + "loss": 0.9277, + "step": 6744 + }, + { + "epoch": 1.2008547008547008, + "grad_norm": 0.6106321215629578, + "learning_rate": 0.00015888265903726188, + "loss": 1.0223, + "step": 6745 + }, + { + "epoch": 1.2010327635327636, + "grad_norm": 0.6400851607322693, + "learning_rate": 0.00015887134487209753, + "loss": 1.1279, + "step": 6746 + }, + { + "epoch": 1.201210826210826, + "grad_norm": 0.6298650503158569, + "learning_rate": 0.00015886002955347147, + "loss": 0.9481, + "step": 6747 + }, + { + "epoch": 1.2013888888888888, + "grad_norm": 0.647974967956543, + "learning_rate": 0.00015884871308160538, + "loss": 1.1513, + "step": 6748 + }, + { + "epoch": 1.2015669515669516, + "grad_norm": 0.6770651936531067, + "learning_rate": 0.000158837395456721, + "loss": 0.9914, + "step": 6749 + }, + { + "epoch": 1.2017450142450143, + "grad_norm": 0.6708947420120239, + "learning_rate": 0.0001588260766790401, + "loss": 1.1848, + "step": 6750 + }, + { + "epoch": 1.2019230769230769, + "grad_norm": 0.5624440908432007, + "learning_rate": 0.00015881475674878442, + "loss": 0.9848, + "step": 6751 + }, + { + "epoch": 1.2021011396011396, + "grad_norm": 0.5512633919715881, + "learning_rate": 0.00015880343566617575, + "loss": 1.0308, + "step": 6752 + }, + { + "epoch": 1.2022792022792024, + "grad_norm": 0.5621042251586914, + "learning_rate": 0.0001587921134314359, + "loss": 0.8724, + "step": 6753 + }, + { + "epoch": 1.202457264957265, + "grad_norm": 0.6881251931190491, + "learning_rate": 0.00015878079004478675, + "loss": 0.9771, + "step": 6754 + }, + { + "epoch": 1.2026353276353277, + "grad_norm": 0.729998767375946, + "learning_rate": 0.0001587694655064501, + "loss": 1.002, + "step": 6755 + }, + { + "epoch": 1.2028133903133904, + "grad_norm": 0.5972567200660706, + "learning_rate": 0.00015875813981664787, + "loss": 1.0571, + "step": 6756 + }, + { + "epoch": 1.202991452991453, + "grad_norm": 0.6319229006767273, + "learning_rate": 0.00015874681297560196, + "loss": 0.9294, + "step": 6757 + }, + { + "epoch": 1.2031695156695157, + "grad_norm": 0.6751521825790405, + "learning_rate": 0.00015873548498353428, + "loss": 0.783, + "step": 6758 + }, + { + "epoch": 1.2033475783475784, + "grad_norm": 0.6476554870605469, + "learning_rate": 0.00015872415584066677, + "loss": 0.8939, + "step": 6759 + }, + { + "epoch": 1.203525641025641, + "grad_norm": 0.6530960202217102, + "learning_rate": 0.0001587128255472214, + "loss": 0.9828, + "step": 6760 + }, + { + "epoch": 1.2037037037037037, + "grad_norm": 0.6708502173423767, + "learning_rate": 0.00015870149410342023, + "loss": 0.9285, + "step": 6761 + }, + { + "epoch": 1.2038817663817665, + "grad_norm": 0.7749543190002441, + "learning_rate": 0.0001586901615094852, + "loss": 1.1295, + "step": 6762 + }, + { + "epoch": 1.204059829059829, + "grad_norm": 0.6750495433807373, + "learning_rate": 0.00015867882776563836, + "loss": 1.0562, + "step": 6763 + }, + { + "epoch": 1.2042378917378918, + "grad_norm": 0.6892416477203369, + "learning_rate": 0.00015866749287210178, + "loss": 0.7207, + "step": 6764 + }, + { + "epoch": 1.2044159544159545, + "grad_norm": 0.7066485285758972, + "learning_rate": 0.00015865615682909758, + "loss": 1.0489, + "step": 6765 + }, + { + "epoch": 1.204594017094017, + "grad_norm": 0.5669938325881958, + "learning_rate": 0.00015864481963684783, + "loss": 0.8149, + "step": 6766 + }, + { + "epoch": 1.2047720797720798, + "grad_norm": 0.6467341780662537, + "learning_rate": 0.0001586334812955746, + "loss": 0.9595, + "step": 6767 + }, + { + "epoch": 1.2049501424501425, + "grad_norm": 0.6026045680046082, + "learning_rate": 0.0001586221418055002, + "loss": 0.9832, + "step": 6768 + }, + { + "epoch": 1.205128205128205, + "grad_norm": 0.7655174732208252, + "learning_rate": 0.00015861080116684665, + "loss": 0.9796, + "step": 6769 + }, + { + "epoch": 1.2053062678062678, + "grad_norm": 0.6386621594429016, + "learning_rate": 0.00015859945937983624, + "loss": 0.9368, + "step": 6770 + }, + { + "epoch": 1.2054843304843306, + "grad_norm": 0.7088032364845276, + "learning_rate": 0.0001585881164446911, + "loss": 1.0167, + "step": 6771 + }, + { + "epoch": 1.205662393162393, + "grad_norm": 0.6015275716781616, + "learning_rate": 0.0001585767723616336, + "loss": 0.8551, + "step": 6772 + }, + { + "epoch": 1.2058404558404558, + "grad_norm": 0.7013260722160339, + "learning_rate": 0.00015856542713088583, + "loss": 0.8009, + "step": 6773 + }, + { + "epoch": 1.2060185185185186, + "grad_norm": 0.6931240558624268, + "learning_rate": 0.00015855408075267024, + "loss": 0.9964, + "step": 6774 + }, + { + "epoch": 1.2061965811965811, + "grad_norm": 0.7274388670921326, + "learning_rate": 0.00015854273322720908, + "loss": 1.0991, + "step": 6775 + }, + { + "epoch": 1.2063746438746439, + "grad_norm": 0.6353716254234314, + "learning_rate": 0.00015853138455472466, + "loss": 1.0893, + "step": 6776 + }, + { + "epoch": 1.2065527065527066, + "grad_norm": 0.6958979368209839, + "learning_rate": 0.00015852003473543932, + "loss": 1.0238, + "step": 6777 + }, + { + "epoch": 1.2067307692307692, + "grad_norm": 0.626838743686676, + "learning_rate": 0.00015850868376957551, + "loss": 0.9384, + "step": 6778 + }, + { + "epoch": 1.206908831908832, + "grad_norm": 0.5455024242401123, + "learning_rate": 0.00015849733165735556, + "loss": 0.8068, + "step": 6779 + }, + { + "epoch": 1.2070868945868947, + "grad_norm": 0.6337353587150574, + "learning_rate": 0.0001584859783990019, + "loss": 1.1341, + "step": 6780 + }, + { + "epoch": 1.2072649572649572, + "grad_norm": 0.6318019032478333, + "learning_rate": 0.000158474623994737, + "loss": 1.1095, + "step": 6781 + }, + { + "epoch": 1.20744301994302, + "grad_norm": 0.8183810710906982, + "learning_rate": 0.00015846326844478332, + "loss": 1.1471, + "step": 6782 + }, + { + "epoch": 1.2076210826210827, + "grad_norm": 0.6140483021736145, + "learning_rate": 0.00015845191174936334, + "loss": 0.8538, + "step": 6783 + }, + { + "epoch": 1.2077991452991452, + "grad_norm": 0.7570197582244873, + "learning_rate": 0.0001584405539086996, + "loss": 1.427, + "step": 6784 + }, + { + "epoch": 1.207977207977208, + "grad_norm": 0.7616991996765137, + "learning_rate": 0.00015842919492301455, + "loss": 1.2214, + "step": 6785 + }, + { + "epoch": 1.2081552706552707, + "grad_norm": 0.561996579170227, + "learning_rate": 0.00015841783479253084, + "loss": 0.8916, + "step": 6786 + }, + { + "epoch": 1.2083333333333333, + "grad_norm": 0.6124222874641418, + "learning_rate": 0.000158406473517471, + "loss": 0.9637, + "step": 6787 + }, + { + "epoch": 1.208511396011396, + "grad_norm": 0.6053098440170288, + "learning_rate": 0.00015839511109805762, + "loss": 1.0365, + "step": 6788 + }, + { + "epoch": 1.2086894586894588, + "grad_norm": 0.6451675295829773, + "learning_rate": 0.00015838374753451338, + "loss": 1.0497, + "step": 6789 + }, + { + "epoch": 1.2088675213675213, + "grad_norm": 0.6789399981498718, + "learning_rate": 0.00015837238282706087, + "loss": 0.9286, + "step": 6790 + }, + { + "epoch": 1.209045584045584, + "grad_norm": 0.5742998123168945, + "learning_rate": 0.0001583610169759228, + "loss": 1.082, + "step": 6791 + }, + { + "epoch": 1.2092236467236468, + "grad_norm": 0.6813693642616272, + "learning_rate": 0.0001583496499813218, + "loss": 0.9785, + "step": 6792 + }, + { + "epoch": 1.2094017094017093, + "grad_norm": 0.6150603890419006, + "learning_rate": 0.0001583382818434806, + "loss": 0.9533, + "step": 6793 + }, + { + "epoch": 1.209579772079772, + "grad_norm": 0.6905919909477234, + "learning_rate": 0.000158326912562622, + "loss": 1.0132, + "step": 6794 + }, + { + "epoch": 1.2097578347578348, + "grad_norm": 0.5861411094665527, + "learning_rate": 0.0001583155421389687, + "loss": 0.7071, + "step": 6795 + }, + { + "epoch": 1.2099358974358974, + "grad_norm": 0.6822740435600281, + "learning_rate": 0.0001583041705727435, + "loss": 1.1366, + "step": 6796 + }, + { + "epoch": 1.21011396011396, + "grad_norm": 0.6013675928115845, + "learning_rate": 0.00015829279786416916, + "loss": 0.9232, + "step": 6797 + }, + { + "epoch": 1.2102920227920229, + "grad_norm": 0.650675356388092, + "learning_rate": 0.00015828142401346857, + "loss": 0.887, + "step": 6798 + }, + { + "epoch": 1.2104700854700854, + "grad_norm": 0.6764078736305237, + "learning_rate": 0.00015827004902086456, + "loss": 0.8423, + "step": 6799 + }, + { + "epoch": 1.2106481481481481, + "grad_norm": 0.6460821628570557, + "learning_rate": 0.00015825867288657994, + "loss": 1.0074, + "step": 6800 + }, + { + "epoch": 1.210826210826211, + "grad_norm": 0.692562997341156, + "learning_rate": 0.00015824729561083768, + "loss": 0.7978, + "step": 6801 + }, + { + "epoch": 1.2110042735042734, + "grad_norm": 0.7255034446716309, + "learning_rate": 0.00015823591719386066, + "loss": 1.071, + "step": 6802 + }, + { + "epoch": 1.2111823361823362, + "grad_norm": 0.6598904728889465, + "learning_rate": 0.0001582245376358718, + "loss": 0.9736, + "step": 6803 + }, + { + "epoch": 1.211360398860399, + "grad_norm": 0.6372483968734741, + "learning_rate": 0.0001582131569370941, + "loss": 0.9029, + "step": 6804 + }, + { + "epoch": 1.2115384615384615, + "grad_norm": 0.5907173156738281, + "learning_rate": 0.00015820177509775048, + "loss": 0.918, + "step": 6805 + }, + { + "epoch": 1.2117165242165242, + "grad_norm": 0.6252630949020386, + "learning_rate": 0.00015819039211806404, + "loss": 0.7801, + "step": 6806 + }, + { + "epoch": 1.211894586894587, + "grad_norm": 0.5793096423149109, + "learning_rate": 0.0001581790079982577, + "loss": 0.5769, + "step": 6807 + }, + { + "epoch": 1.2120726495726495, + "grad_norm": 0.7267270684242249, + "learning_rate": 0.00015816762273855454, + "loss": 1.1428, + "step": 6808 + }, + { + "epoch": 1.2122507122507122, + "grad_norm": 0.7481234073638916, + "learning_rate": 0.00015815623633917767, + "loss": 1.0209, + "step": 6809 + }, + { + "epoch": 1.212428774928775, + "grad_norm": 0.6114386916160583, + "learning_rate": 0.00015814484880035017, + "loss": 0.9073, + "step": 6810 + }, + { + "epoch": 1.2126068376068375, + "grad_norm": 0.6871182322502136, + "learning_rate": 0.00015813346012229516, + "loss": 1.151, + "step": 6811 + }, + { + "epoch": 1.2127849002849003, + "grad_norm": 0.6380293965339661, + "learning_rate": 0.0001581220703052357, + "loss": 1.0981, + "step": 6812 + }, + { + "epoch": 1.212962962962963, + "grad_norm": 0.6013718247413635, + "learning_rate": 0.00015811067934939503, + "loss": 0.8832, + "step": 6813 + }, + { + "epoch": 1.2131410256410255, + "grad_norm": 0.5816897749900818, + "learning_rate": 0.00015809928725499632, + "loss": 1.063, + "step": 6814 + }, + { + "epoch": 1.2133190883190883, + "grad_norm": 0.5970914363861084, + "learning_rate": 0.00015808789402226278, + "loss": 1.1177, + "step": 6815 + }, + { + "epoch": 1.213497150997151, + "grad_norm": 0.7624936103820801, + "learning_rate": 0.00015807649965141762, + "loss": 1.048, + "step": 6816 + }, + { + "epoch": 1.2136752136752136, + "grad_norm": 0.636263906955719, + "learning_rate": 0.0001580651041426841, + "loss": 0.9743, + "step": 6817 + }, + { + "epoch": 1.2138532763532763, + "grad_norm": 0.641090452671051, + "learning_rate": 0.00015805370749628547, + "loss": 1.0227, + "step": 6818 + }, + { + "epoch": 1.214031339031339, + "grad_norm": 0.6484021544456482, + "learning_rate": 0.00015804230971244504, + "loss": 0.9615, + "step": 6819 + }, + { + "epoch": 1.2142094017094016, + "grad_norm": 0.6473353505134583, + "learning_rate": 0.00015803091079138613, + "loss": 1.0507, + "step": 6820 + }, + { + "epoch": 1.2143874643874644, + "grad_norm": 0.5477129220962524, + "learning_rate": 0.00015801951073333206, + "loss": 0.7928, + "step": 6821 + }, + { + "epoch": 1.2145655270655271, + "grad_norm": 0.7256210446357727, + "learning_rate": 0.0001580081095385062, + "loss": 1.0172, + "step": 6822 + }, + { + "epoch": 1.2147435897435896, + "grad_norm": 0.5785418748855591, + "learning_rate": 0.00015799670720713195, + "loss": 0.8478, + "step": 6823 + }, + { + "epoch": 1.2149216524216524, + "grad_norm": 0.6782996654510498, + "learning_rate": 0.00015798530373943267, + "loss": 1.1819, + "step": 6824 + }, + { + "epoch": 1.2150997150997151, + "grad_norm": 0.6513699293136597, + "learning_rate": 0.00015797389913563186, + "loss": 0.9626, + "step": 6825 + }, + { + "epoch": 1.2152777777777777, + "grad_norm": 0.6503037214279175, + "learning_rate": 0.0001579624933959529, + "loss": 1.0282, + "step": 6826 + }, + { + "epoch": 1.2154558404558404, + "grad_norm": 0.581501841545105, + "learning_rate": 0.0001579510865206193, + "loss": 0.8976, + "step": 6827 + }, + { + "epoch": 1.2156339031339032, + "grad_norm": 0.6696721911430359, + "learning_rate": 0.00015793967850985454, + "loss": 0.6418, + "step": 6828 + }, + { + "epoch": 1.215811965811966, + "grad_norm": 0.6577274203300476, + "learning_rate": 0.00015792826936388213, + "loss": 1.0615, + "step": 6829 + }, + { + "epoch": 1.2159900284900285, + "grad_norm": 0.66291743516922, + "learning_rate": 0.00015791685908292564, + "loss": 0.8582, + "step": 6830 + }, + { + "epoch": 1.2161680911680912, + "grad_norm": 0.6548362374305725, + "learning_rate": 0.0001579054476672086, + "loss": 1.0343, + "step": 6831 + }, + { + "epoch": 1.2163461538461537, + "grad_norm": 0.6381218433380127, + "learning_rate": 0.00015789403511695457, + "loss": 0.8133, + "step": 6832 + }, + { + "epoch": 1.2165242165242165, + "grad_norm": 0.7217492461204529, + "learning_rate": 0.00015788262143238722, + "loss": 0.9183, + "step": 6833 + }, + { + "epoch": 1.2167022792022792, + "grad_norm": 0.610454797744751, + "learning_rate": 0.00015787120661373013, + "loss": 0.8488, + "step": 6834 + }, + { + "epoch": 1.216880341880342, + "grad_norm": 0.592771053314209, + "learning_rate": 0.00015785979066120696, + "loss": 0.8673, + "step": 6835 + }, + { + "epoch": 1.2170584045584045, + "grad_norm": 0.5787834525108337, + "learning_rate": 0.00015784837357504138, + "loss": 0.7945, + "step": 6836 + }, + { + "epoch": 1.2172364672364673, + "grad_norm": 0.6814196109771729, + "learning_rate": 0.0001578369553554571, + "loss": 0.8906, + "step": 6837 + }, + { + "epoch": 1.2174145299145298, + "grad_norm": 0.6383981108665466, + "learning_rate": 0.00015782553600267787, + "loss": 0.8962, + "step": 6838 + }, + { + "epoch": 1.2175925925925926, + "grad_norm": 0.6733864545822144, + "learning_rate": 0.0001578141155169273, + "loss": 1.2077, + "step": 6839 + }, + { + "epoch": 1.2177706552706553, + "grad_norm": 0.5891284346580505, + "learning_rate": 0.0001578026938984293, + "loss": 0.9477, + "step": 6840 + }, + { + "epoch": 1.217948717948718, + "grad_norm": 0.7220266461372375, + "learning_rate": 0.00015779127114740757, + "loss": 1.0343, + "step": 6841 + }, + { + "epoch": 1.2181267806267806, + "grad_norm": 0.6566546559333801, + "learning_rate": 0.0001577798472640859, + "loss": 0.9576, + "step": 6842 + }, + { + "epoch": 1.2183048433048433, + "grad_norm": 0.6428449153900146, + "learning_rate": 0.0001577684222486882, + "loss": 0.8957, + "step": 6843 + }, + { + "epoch": 1.2184829059829059, + "grad_norm": 0.6542909741401672, + "learning_rate": 0.00015775699610143823, + "loss": 0.9942, + "step": 6844 + }, + { + "epoch": 1.2186609686609686, + "grad_norm": 0.7101675868034363, + "learning_rate": 0.00015774556882255992, + "loss": 1.015, + "step": 6845 + }, + { + "epoch": 1.2188390313390314, + "grad_norm": 0.6606267094612122, + "learning_rate": 0.00015773414041227713, + "loss": 1.1406, + "step": 6846 + }, + { + "epoch": 1.2190170940170941, + "grad_norm": 0.67124342918396, + "learning_rate": 0.00015772271087081383, + "loss": 1.2392, + "step": 6847 + }, + { + "epoch": 1.2191951566951567, + "grad_norm": 0.6615056991577148, + "learning_rate": 0.0001577112801983939, + "loss": 1.1583, + "step": 6848 + }, + { + "epoch": 1.2193732193732194, + "grad_norm": 0.6941317319869995, + "learning_rate": 0.0001576998483952413, + "loss": 1.0255, + "step": 6849 + }, + { + "epoch": 1.219551282051282, + "grad_norm": 0.5740683674812317, + "learning_rate": 0.00015768841546158005, + "loss": 1.0393, + "step": 6850 + }, + { + "epoch": 1.2197293447293447, + "grad_norm": 0.7143667340278625, + "learning_rate": 0.00015767698139763415, + "loss": 0.7564, + "step": 6851 + }, + { + "epoch": 1.2199074074074074, + "grad_norm": 0.6730484366416931, + "learning_rate": 0.00015766554620362758, + "loss": 1.2221, + "step": 6852 + }, + { + "epoch": 1.2200854700854702, + "grad_norm": 0.6883087754249573, + "learning_rate": 0.00015765410987978444, + "loss": 1.0156, + "step": 6853 + }, + { + "epoch": 1.2202635327635327, + "grad_norm": 0.6585961580276489, + "learning_rate": 0.00015764267242632875, + "loss": 1.0888, + "step": 6854 + }, + { + "epoch": 1.2204415954415955, + "grad_norm": 0.6325246691703796, + "learning_rate": 0.00015763123384348465, + "loss": 0.973, + "step": 6855 + }, + { + "epoch": 1.220619658119658, + "grad_norm": 0.5930588245391846, + "learning_rate": 0.00015761979413147627, + "loss": 0.8551, + "step": 6856 + }, + { + "epoch": 1.2207977207977208, + "grad_norm": 0.6440611481666565, + "learning_rate": 0.0001576083532905277, + "loss": 0.8396, + "step": 6857 + }, + { + "epoch": 1.2209757834757835, + "grad_norm": 0.6796659231185913, + "learning_rate": 0.00015759691132086315, + "loss": 1.0662, + "step": 6858 + }, + { + "epoch": 1.2211538461538463, + "grad_norm": 0.6813400983810425, + "learning_rate": 0.00015758546822270674, + "loss": 1.0457, + "step": 6859 + }, + { + "epoch": 1.2213319088319088, + "grad_norm": 0.6871716976165771, + "learning_rate": 0.00015757402399628272, + "loss": 1.1675, + "step": 6860 + }, + { + "epoch": 1.2215099715099715, + "grad_norm": 0.6431481838226318, + "learning_rate": 0.00015756257864181524, + "loss": 0.9366, + "step": 6861 + }, + { + "epoch": 1.221688034188034, + "grad_norm": 0.6061800718307495, + "learning_rate": 0.00015755113215952868, + "loss": 0.9267, + "step": 6862 + }, + { + "epoch": 1.2218660968660968, + "grad_norm": 0.5755770206451416, + "learning_rate": 0.00015753968454964722, + "loss": 0.7342, + "step": 6863 + }, + { + "epoch": 1.2220441595441596, + "grad_norm": 0.571345329284668, + "learning_rate": 0.00015752823581239515, + "loss": 0.8943, + "step": 6864 + }, + { + "epoch": 1.2222222222222223, + "grad_norm": 0.6925615668296814, + "learning_rate": 0.0001575167859479968, + "loss": 0.8801, + "step": 6865 + }, + { + "epoch": 1.2224002849002849, + "grad_norm": 0.6812975406646729, + "learning_rate": 0.00015750533495667655, + "loss": 0.9567, + "step": 6866 + }, + { + "epoch": 1.2225783475783476, + "grad_norm": 0.8216777443885803, + "learning_rate": 0.00015749388283865868, + "loss": 1.0908, + "step": 6867 + }, + { + "epoch": 1.2227564102564104, + "grad_norm": 0.6051010489463806, + "learning_rate": 0.00015748242959416763, + "loss": 0.8851, + "step": 6868 + }, + { + "epoch": 1.2229344729344729, + "grad_norm": 0.7750816345214844, + "learning_rate": 0.00015747097522342775, + "loss": 1.1526, + "step": 6869 + }, + { + "epoch": 1.2231125356125356, + "grad_norm": 0.6240930557250977, + "learning_rate": 0.00015745951972666355, + "loss": 1.0603, + "step": 6870 + }, + { + "epoch": 1.2232905982905984, + "grad_norm": 0.7228875160217285, + "learning_rate": 0.00015744806310409937, + "loss": 1.1028, + "step": 6871 + }, + { + "epoch": 1.223468660968661, + "grad_norm": 0.724075436592102, + "learning_rate": 0.00015743660535595978, + "loss": 0.8983, + "step": 6872 + }, + { + "epoch": 1.2236467236467237, + "grad_norm": 0.6398203372955322, + "learning_rate": 0.00015742514648246916, + "loss": 1.0548, + "step": 6873 + }, + { + "epoch": 1.2238247863247864, + "grad_norm": 0.7024285793304443, + "learning_rate": 0.00015741368648385212, + "loss": 1.0172, + "step": 6874 + }, + { + "epoch": 1.224002849002849, + "grad_norm": 0.6717609763145447, + "learning_rate": 0.00015740222536033316, + "loss": 0.9002, + "step": 6875 + }, + { + "epoch": 1.2241809116809117, + "grad_norm": 0.5886133313179016, + "learning_rate": 0.00015739076311213686, + "loss": 0.8614, + "step": 6876 + }, + { + "epoch": 1.2243589743589745, + "grad_norm": 0.6856684684753418, + "learning_rate": 0.00015737929973948776, + "loss": 1.1633, + "step": 6877 + }, + { + "epoch": 1.224537037037037, + "grad_norm": 0.6771421432495117, + "learning_rate": 0.00015736783524261045, + "loss": 1.0921, + "step": 6878 + }, + { + "epoch": 1.2247150997150997, + "grad_norm": 0.5016412138938904, + "learning_rate": 0.0001573563696217296, + "loss": 0.6732, + "step": 6879 + }, + { + "epoch": 1.2248931623931625, + "grad_norm": 0.7595276236534119, + "learning_rate": 0.00015734490287706984, + "loss": 1.0427, + "step": 6880 + }, + { + "epoch": 1.225071225071225, + "grad_norm": 0.6664281487464905, + "learning_rate": 0.00015733343500885582, + "loss": 1.2836, + "step": 6881 + }, + { + "epoch": 1.2252492877492878, + "grad_norm": 0.6662577390670776, + "learning_rate": 0.00015732196601731224, + "loss": 1.1288, + "step": 6882 + }, + { + "epoch": 1.2254273504273505, + "grad_norm": 0.6238988041877747, + "learning_rate": 0.00015731049590266385, + "loss": 1.0809, + "step": 6883 + }, + { + "epoch": 1.225605413105413, + "grad_norm": 0.6483062505722046, + "learning_rate": 0.00015729902466513532, + "loss": 0.9992, + "step": 6884 + }, + { + "epoch": 1.2257834757834758, + "grad_norm": 0.6890861988067627, + "learning_rate": 0.0001572875523049514, + "loss": 1.1844, + "step": 6885 + }, + { + "epoch": 1.2259615384615385, + "grad_norm": 0.7087607383728027, + "learning_rate": 0.00015727607882233695, + "loss": 1.013, + "step": 6886 + }, + { + "epoch": 1.226139601139601, + "grad_norm": 0.709048867225647, + "learning_rate": 0.00015726460421751668, + "loss": 0.9748, + "step": 6887 + }, + { + "epoch": 1.2263176638176638, + "grad_norm": 0.5918150544166565, + "learning_rate": 0.00015725312849071546, + "loss": 0.9978, + "step": 6888 + }, + { + "epoch": 1.2264957264957266, + "grad_norm": 0.4343377947807312, + "learning_rate": 0.0001572416516421581, + "loss": 0.6233, + "step": 6889 + }, + { + "epoch": 1.226673789173789, + "grad_norm": 0.6360403895378113, + "learning_rate": 0.00015723017367206952, + "loss": 0.9698, + "step": 6890 + }, + { + "epoch": 1.2268518518518519, + "grad_norm": 0.7261984944343567, + "learning_rate": 0.00015721869458067454, + "loss": 1.0426, + "step": 6891 + }, + { + "epoch": 1.2270299145299146, + "grad_norm": 0.6806774139404297, + "learning_rate": 0.0001572072143681981, + "loss": 0.9692, + "step": 6892 + }, + { + "epoch": 1.2272079772079771, + "grad_norm": 0.7140612006187439, + "learning_rate": 0.00015719573303486515, + "loss": 1.0828, + "step": 6893 + }, + { + "epoch": 1.22738603988604, + "grad_norm": 0.5383326411247253, + "learning_rate": 0.0001571842505809006, + "loss": 1.012, + "step": 6894 + }, + { + "epoch": 1.2275641025641026, + "grad_norm": 0.5992259383201599, + "learning_rate": 0.0001571727670065295, + "loss": 0.876, + "step": 6895 + }, + { + "epoch": 1.2277421652421652, + "grad_norm": 0.636696457862854, + "learning_rate": 0.00015716128231197676, + "loss": 1.1001, + "step": 6896 + }, + { + "epoch": 1.227920227920228, + "grad_norm": 0.5980371236801147, + "learning_rate": 0.00015714979649746744, + "loss": 0.937, + "step": 6897 + }, + { + "epoch": 1.2280982905982907, + "grad_norm": 0.7678794860839844, + "learning_rate": 0.00015713830956322656, + "loss": 1.1965, + "step": 6898 + }, + { + "epoch": 1.2282763532763532, + "grad_norm": 0.6918835639953613, + "learning_rate": 0.00015712682150947923, + "loss": 0.8578, + "step": 6899 + }, + { + "epoch": 1.228454415954416, + "grad_norm": 0.6463451385498047, + "learning_rate": 0.00015711533233645048, + "loss": 1.009, + "step": 6900 + }, + { + "epoch": 1.2286324786324787, + "grad_norm": 0.6720646023750305, + "learning_rate": 0.00015710384204436549, + "loss": 1.0031, + "step": 6901 + }, + { + "epoch": 1.2288105413105412, + "grad_norm": 0.6618736982345581, + "learning_rate": 0.00015709235063344926, + "loss": 0.9017, + "step": 6902 + }, + { + "epoch": 1.228988603988604, + "grad_norm": 0.6789427399635315, + "learning_rate": 0.0001570808581039271, + "loss": 1.1289, + "step": 6903 + }, + { + "epoch": 1.2291666666666667, + "grad_norm": 0.6395950317382812, + "learning_rate": 0.00015706936445602403, + "loss": 1.1051, + "step": 6904 + }, + { + "epoch": 1.2293447293447293, + "grad_norm": 0.7023917436599731, + "learning_rate": 0.00015705786968996533, + "loss": 1.2876, + "step": 6905 + }, + { + "epoch": 1.229522792022792, + "grad_norm": 0.7473352551460266, + "learning_rate": 0.00015704637380597623, + "loss": 1.237, + "step": 6906 + }, + { + "epoch": 1.2297008547008548, + "grad_norm": 0.6952672004699707, + "learning_rate": 0.00015703487680428192, + "loss": 1.0674, + "step": 6907 + }, + { + "epoch": 1.2298789173789173, + "grad_norm": 0.5968644022941589, + "learning_rate": 0.0001570233786851077, + "loss": 0.9169, + "step": 6908 + }, + { + "epoch": 1.23005698005698, + "grad_norm": 0.7219798564910889, + "learning_rate": 0.0001570118794486788, + "loss": 1.0556, + "step": 6909 + }, + { + "epoch": 1.2302350427350428, + "grad_norm": 0.6603400707244873, + "learning_rate": 0.0001570003790952206, + "loss": 0.9596, + "step": 6910 + }, + { + "epoch": 1.2304131054131053, + "grad_norm": 0.5972838401794434, + "learning_rate": 0.0001569888776249583, + "loss": 0.9168, + "step": 6911 + }, + { + "epoch": 1.230591168091168, + "grad_norm": 0.792585551738739, + "learning_rate": 0.00015697737503811738, + "loss": 1.1074, + "step": 6912 + }, + { + "epoch": 1.2307692307692308, + "grad_norm": 0.5845609903335571, + "learning_rate": 0.00015696587133492314, + "loss": 0.8413, + "step": 6913 + }, + { + "epoch": 1.2309472934472934, + "grad_norm": 0.6603896021842957, + "learning_rate": 0.000156954366515601, + "loss": 0.9109, + "step": 6914 + }, + { + "epoch": 1.2311253561253561, + "grad_norm": 0.6367142796516418, + "learning_rate": 0.00015694286058037636, + "loss": 1.0119, + "step": 6915 + }, + { + "epoch": 1.2313034188034189, + "grad_norm": 0.693854570388794, + "learning_rate": 0.00015693135352947465, + "loss": 1.0925, + "step": 6916 + }, + { + "epoch": 1.2314814814814814, + "grad_norm": 0.6570404171943665, + "learning_rate": 0.00015691984536312135, + "loss": 0.9731, + "step": 6917 + }, + { + "epoch": 1.2316595441595442, + "grad_norm": 0.6778639554977417, + "learning_rate": 0.0001569083360815419, + "loss": 1.1415, + "step": 6918 + }, + { + "epoch": 1.231837606837607, + "grad_norm": 0.6656233668327332, + "learning_rate": 0.00015689682568496182, + "loss": 0.8603, + "step": 6919 + }, + { + "epoch": 1.2320156695156694, + "grad_norm": 0.6569861173629761, + "learning_rate": 0.00015688531417360665, + "loss": 0.8374, + "step": 6920 + }, + { + "epoch": 1.2321937321937322, + "grad_norm": 0.6746888160705566, + "learning_rate": 0.0001568738015477019, + "loss": 1.1395, + "step": 6921 + }, + { + "epoch": 1.232371794871795, + "grad_norm": 0.6180813908576965, + "learning_rate": 0.00015686228780747316, + "loss": 1.0049, + "step": 6922 + }, + { + "epoch": 1.2325498575498575, + "grad_norm": 0.7326146960258484, + "learning_rate": 0.000156850772953146, + "loss": 1.2389, + "step": 6923 + }, + { + "epoch": 1.2327279202279202, + "grad_norm": 0.5912215709686279, + "learning_rate": 0.00015683925698494608, + "loss": 1.0174, + "step": 6924 + }, + { + "epoch": 1.232905982905983, + "grad_norm": 0.5214745402336121, + "learning_rate": 0.00015682773990309895, + "loss": 0.5778, + "step": 6925 + }, + { + "epoch": 1.2330840455840455, + "grad_norm": 0.6862079501152039, + "learning_rate": 0.00015681622170783034, + "loss": 0.896, + "step": 6926 + }, + { + "epoch": 1.2332621082621082, + "grad_norm": 0.7858926057815552, + "learning_rate": 0.00015680470239936586, + "loss": 1.0714, + "step": 6927 + }, + { + "epoch": 1.233440170940171, + "grad_norm": 0.6706146597862244, + "learning_rate": 0.00015679318197793127, + "loss": 1.0157, + "step": 6928 + }, + { + "epoch": 1.2336182336182335, + "grad_norm": 0.6657105088233948, + "learning_rate": 0.00015678166044375225, + "loss": 0.9674, + "step": 6929 + }, + { + "epoch": 1.2337962962962963, + "grad_norm": 0.6790838837623596, + "learning_rate": 0.0001567701377970545, + "loss": 0.9744, + "step": 6930 + }, + { + "epoch": 1.233974358974359, + "grad_norm": 0.6469771862030029, + "learning_rate": 0.00015675861403806386, + "loss": 1.0205, + "step": 6931 + }, + { + "epoch": 1.2341524216524216, + "grad_norm": 0.4926300346851349, + "learning_rate": 0.0001567470891670061, + "loss": 0.6336, + "step": 6932 + }, + { + "epoch": 1.2343304843304843, + "grad_norm": 0.6762157082557678, + "learning_rate": 0.000156735563184107, + "loss": 1.059, + "step": 6933 + }, + { + "epoch": 1.234508547008547, + "grad_norm": 0.6998521685600281, + "learning_rate": 0.0001567240360895924, + "loss": 1.0586, + "step": 6934 + }, + { + "epoch": 1.2346866096866096, + "grad_norm": 0.5947706699371338, + "learning_rate": 0.00015671250788368814, + "loss": 0.8815, + "step": 6935 + }, + { + "epoch": 1.2348646723646723, + "grad_norm": 0.6966122984886169, + "learning_rate": 0.0001567009785666201, + "loss": 1.0105, + "step": 6936 + }, + { + "epoch": 1.235042735042735, + "grad_norm": 0.6747866272926331, + "learning_rate": 0.0001566894481386142, + "loss": 0.8783, + "step": 6937 + }, + { + "epoch": 1.2352207977207976, + "grad_norm": 0.6348921060562134, + "learning_rate": 0.0001566779165998963, + "loss": 0.7813, + "step": 6938 + }, + { + "epoch": 1.2353988603988604, + "grad_norm": 0.596466600894928, + "learning_rate": 0.00015666638395069236, + "loss": 0.8689, + "step": 6939 + }, + { + "epoch": 1.2355769230769231, + "grad_norm": 0.6926795244216919, + "learning_rate": 0.00015665485019122834, + "loss": 1.0266, + "step": 6940 + }, + { + "epoch": 1.2357549857549857, + "grad_norm": 0.6590100526809692, + "learning_rate": 0.00015664331532173022, + "loss": 1.128, + "step": 6941 + }, + { + "epoch": 1.2359330484330484, + "grad_norm": 0.7422109246253967, + "learning_rate": 0.00015663177934242402, + "loss": 0.8495, + "step": 6942 + }, + { + "epoch": 1.2361111111111112, + "grad_norm": 0.6463228464126587, + "learning_rate": 0.0001566202422535357, + "loss": 1.0941, + "step": 6943 + }, + { + "epoch": 1.236289173789174, + "grad_norm": 0.7278686761856079, + "learning_rate": 0.0001566087040552914, + "loss": 1.2039, + "step": 6944 + }, + { + "epoch": 1.2364672364672364, + "grad_norm": 0.6917086839675903, + "learning_rate": 0.00015659716474791712, + "loss": 1.042, + "step": 6945 + }, + { + "epoch": 1.2366452991452992, + "grad_norm": 0.637205183506012, + "learning_rate": 0.00015658562433163898, + "loss": 1.0379, + "step": 6946 + }, + { + "epoch": 1.2368233618233617, + "grad_norm": 0.6706623435020447, + "learning_rate": 0.00015657408280668307, + "loss": 1.0347, + "step": 6947 + }, + { + "epoch": 1.2370014245014245, + "grad_norm": 0.6435480713844299, + "learning_rate": 0.00015656254017327553, + "loss": 0.7708, + "step": 6948 + }, + { + "epoch": 1.2371794871794872, + "grad_norm": 0.5703113675117493, + "learning_rate": 0.0001565509964316425, + "loss": 0.8786, + "step": 6949 + }, + { + "epoch": 1.23735754985755, + "grad_norm": 0.6438127160072327, + "learning_rate": 0.00015653945158201018, + "loss": 0.9435, + "step": 6950 + }, + { + "epoch": 1.2375356125356125, + "grad_norm": 0.68101966381073, + "learning_rate": 0.00015652790562460474, + "loss": 1.1062, + "step": 6951 + }, + { + "epoch": 1.2377136752136753, + "grad_norm": 0.661230206489563, + "learning_rate": 0.00015651635855965242, + "loss": 1.0113, + "step": 6952 + }, + { + "epoch": 1.2378917378917378, + "grad_norm": 0.6399117708206177, + "learning_rate": 0.0001565048103873795, + "loss": 1.1423, + "step": 6953 + }, + { + "epoch": 1.2380698005698005, + "grad_norm": 0.7614672780036926, + "learning_rate": 0.00015649326110801215, + "loss": 1.0359, + "step": 6954 + }, + { + "epoch": 1.2382478632478633, + "grad_norm": 0.6461986303329468, + "learning_rate": 0.00015648171072177674, + "loss": 1.0145, + "step": 6955 + }, + { + "epoch": 1.238425925925926, + "grad_norm": 0.5902668833732605, + "learning_rate": 0.0001564701592288995, + "loss": 0.9451, + "step": 6956 + }, + { + "epoch": 1.2386039886039886, + "grad_norm": 0.5686020255088806, + "learning_rate": 0.00015645860662960682, + "loss": 0.7512, + "step": 6957 + }, + { + "epoch": 1.2387820512820513, + "grad_norm": 0.6640077829360962, + "learning_rate": 0.00015644705292412503, + "loss": 0.7133, + "step": 6958 + }, + { + "epoch": 1.2389601139601139, + "grad_norm": 0.7402132749557495, + "learning_rate": 0.00015643549811268049, + "loss": 1.0903, + "step": 6959 + }, + { + "epoch": 1.2391381766381766, + "grad_norm": 0.62332683801651, + "learning_rate": 0.00015642394219549962, + "loss": 0.9378, + "step": 6960 + }, + { + "epoch": 1.2393162393162394, + "grad_norm": 0.6374901533126831, + "learning_rate": 0.00015641238517280877, + "loss": 1.0746, + "step": 6961 + }, + { + "epoch": 1.239494301994302, + "grad_norm": 0.5939112901687622, + "learning_rate": 0.00015640082704483443, + "loss": 0.7185, + "step": 6962 + }, + { + "epoch": 1.2396723646723646, + "grad_norm": 0.8378096222877502, + "learning_rate": 0.00015638926781180306, + "loss": 1.1932, + "step": 6963 + }, + { + "epoch": 1.2398504273504274, + "grad_norm": 0.5707982778549194, + "learning_rate": 0.0001563777074739411, + "loss": 0.9834, + "step": 6964 + }, + { + "epoch": 1.24002849002849, + "grad_norm": 0.6339748501777649, + "learning_rate": 0.00015636614603147512, + "loss": 1.0307, + "step": 6965 + }, + { + "epoch": 1.2402065527065527, + "grad_norm": 0.7353155016899109, + "learning_rate": 0.00015635458348463156, + "loss": 1.0311, + "step": 6966 + }, + { + "epoch": 1.2403846153846154, + "grad_norm": 0.8307726979255676, + "learning_rate": 0.00015634301983363704, + "loss": 1.0673, + "step": 6967 + }, + { + "epoch": 1.2405626780626782, + "grad_norm": 0.5299199819564819, + "learning_rate": 0.00015633145507871807, + "loss": 0.6649, + "step": 6968 + }, + { + "epoch": 1.2407407407407407, + "grad_norm": 0.6162533760070801, + "learning_rate": 0.00015631988922010126, + "loss": 0.8096, + "step": 6969 + }, + { + "epoch": 1.2409188034188035, + "grad_norm": 0.6212689876556396, + "learning_rate": 0.0001563083222580132, + "loss": 1.0371, + "step": 6970 + }, + { + "epoch": 1.241096866096866, + "grad_norm": 0.6148123145103455, + "learning_rate": 0.00015629675419268055, + "loss": 1.0439, + "step": 6971 + }, + { + "epoch": 1.2412749287749287, + "grad_norm": 0.6163684129714966, + "learning_rate": 0.00015628518502432994, + "loss": 0.9075, + "step": 6972 + }, + { + "epoch": 1.2414529914529915, + "grad_norm": 0.5127472877502441, + "learning_rate": 0.00015627361475318807, + "loss": 0.6138, + "step": 6973 + }, + { + "epoch": 1.2416310541310542, + "grad_norm": 0.6508103013038635, + "learning_rate": 0.0001562620433794816, + "loss": 0.9608, + "step": 6974 + }, + { + "epoch": 1.2418091168091168, + "grad_norm": 0.6711046695709229, + "learning_rate": 0.0001562504709034373, + "loss": 1.1494, + "step": 6975 + }, + { + "epoch": 1.2419871794871795, + "grad_norm": 0.6831514835357666, + "learning_rate": 0.00015623889732528182, + "loss": 0.9664, + "step": 6976 + }, + { + "epoch": 1.242165242165242, + "grad_norm": 0.693732738494873, + "learning_rate": 0.00015622732264524198, + "loss": 0.9055, + "step": 6977 + }, + { + "epoch": 1.2423433048433048, + "grad_norm": 0.8475173711776733, + "learning_rate": 0.00015621574686354456, + "loss": 1.2014, + "step": 6978 + }, + { + "epoch": 1.2425213675213675, + "grad_norm": 0.6342347264289856, + "learning_rate": 0.0001562041699804164, + "loss": 1.0691, + "step": 6979 + }, + { + "epoch": 1.2426994301994303, + "grad_norm": 0.620517373085022, + "learning_rate": 0.00015619259199608422, + "loss": 0.7318, + "step": 6980 + }, + { + "epoch": 1.2428774928774928, + "grad_norm": 0.589567244052887, + "learning_rate": 0.000156181012910775, + "loss": 1.0656, + "step": 6981 + }, + { + "epoch": 1.2430555555555556, + "grad_norm": 0.7570258975028992, + "learning_rate": 0.00015616943272471546, + "loss": 1.0517, + "step": 6982 + }, + { + "epoch": 1.243233618233618, + "grad_norm": 0.6232032775878906, + "learning_rate": 0.00015615785143813262, + "loss": 0.8867, + "step": 6983 + }, + { + "epoch": 1.2434116809116809, + "grad_norm": 0.630095899105072, + "learning_rate": 0.0001561462690512533, + "loss": 0.9287, + "step": 6984 + }, + { + "epoch": 1.2435897435897436, + "grad_norm": 0.7410848140716553, + "learning_rate": 0.00015613468556430454, + "loss": 1.162, + "step": 6985 + }, + { + "epoch": 1.2437678062678064, + "grad_norm": 0.7574684023857117, + "learning_rate": 0.00015612310097751317, + "loss": 1.2118, + "step": 6986 + }, + { + "epoch": 1.243945868945869, + "grad_norm": 0.580760657787323, + "learning_rate": 0.0001561115152911062, + "loss": 1.0612, + "step": 6987 + }, + { + "epoch": 1.2441239316239316, + "grad_norm": 0.6105104088783264, + "learning_rate": 0.00015609992850531073, + "loss": 0.9262, + "step": 6988 + }, + { + "epoch": 1.2443019943019944, + "grad_norm": 0.669435441493988, + "learning_rate": 0.00015608834062035362, + "loss": 0.9595, + "step": 6989 + }, + { + "epoch": 1.244480056980057, + "grad_norm": 0.6530314683914185, + "learning_rate": 0.00015607675163646206, + "loss": 0.7987, + "step": 6990 + }, + { + "epoch": 1.2446581196581197, + "grad_norm": 0.5801477432250977, + "learning_rate": 0.00015606516155386297, + "loss": 0.7667, + "step": 6991 + }, + { + "epoch": 1.2448361823361824, + "grad_norm": 0.5773885250091553, + "learning_rate": 0.00015605357037278355, + "loss": 0.847, + "step": 6992 + }, + { + "epoch": 1.245014245014245, + "grad_norm": 0.5399810075759888, + "learning_rate": 0.00015604197809345082, + "loss": 0.9284, + "step": 6993 + }, + { + "epoch": 1.2451923076923077, + "grad_norm": 0.5910452604293823, + "learning_rate": 0.000156030384716092, + "loss": 1.0004, + "step": 6994 + }, + { + "epoch": 1.2453703703703705, + "grad_norm": 0.5979224443435669, + "learning_rate": 0.00015601879024093414, + "loss": 0.9027, + "step": 6995 + }, + { + "epoch": 1.245548433048433, + "grad_norm": 0.6092126369476318, + "learning_rate": 0.0001560071946682045, + "loss": 0.9755, + "step": 6996 + }, + { + "epoch": 1.2457264957264957, + "grad_norm": 0.6536708474159241, + "learning_rate": 0.0001559955979981302, + "loss": 1.1828, + "step": 6997 + }, + { + "epoch": 1.2459045584045585, + "grad_norm": 0.6602030992507935, + "learning_rate": 0.00015598400023093847, + "loss": 1.0395, + "step": 6998 + }, + { + "epoch": 1.246082621082621, + "grad_norm": 0.6864825487136841, + "learning_rate": 0.00015597240136685657, + "loss": 1.083, + "step": 6999 + }, + { + "epoch": 1.2462606837606838, + "grad_norm": 0.6194674968719482, + "learning_rate": 0.0001559608014061117, + "loss": 1.0461, + "step": 7000 + }, + { + "epoch": 1.2464387464387465, + "grad_norm": 0.5879074335098267, + "learning_rate": 0.00015594920034893122, + "loss": 1.076, + "step": 7001 + }, + { + "epoch": 1.246616809116809, + "grad_norm": 0.6514387726783752, + "learning_rate": 0.00015593759819554234, + "loss": 1.0396, + "step": 7002 + }, + { + "epoch": 1.2467948717948718, + "grad_norm": 0.5988301634788513, + "learning_rate": 0.00015592599494617247, + "loss": 0.9501, + "step": 7003 + }, + { + "epoch": 1.2469729344729346, + "grad_norm": 0.6282773613929749, + "learning_rate": 0.00015591439060104887, + "loss": 1.1002, + "step": 7004 + }, + { + "epoch": 1.247150997150997, + "grad_norm": 0.6910465955734253, + "learning_rate": 0.00015590278516039896, + "loss": 1.1771, + "step": 7005 + }, + { + "epoch": 1.2473290598290598, + "grad_norm": 0.6097282767295837, + "learning_rate": 0.00015589117862445007, + "loss": 1.0707, + "step": 7006 + }, + { + "epoch": 1.2475071225071226, + "grad_norm": 0.7076875567436218, + "learning_rate": 0.00015587957099342967, + "loss": 1.0078, + "step": 7007 + }, + { + "epoch": 1.2476851851851851, + "grad_norm": 0.6776556372642517, + "learning_rate": 0.00015586796226756518, + "loss": 0.8971, + "step": 7008 + }, + { + "epoch": 1.2478632478632479, + "grad_norm": 0.6506341695785522, + "learning_rate": 0.00015585635244708398, + "loss": 0.9727, + "step": 7009 + }, + { + "epoch": 1.2480413105413106, + "grad_norm": 0.624724805355072, + "learning_rate": 0.00015584474153221357, + "loss": 0.9858, + "step": 7010 + }, + { + "epoch": 1.2482193732193732, + "grad_norm": 0.6070096492767334, + "learning_rate": 0.0001558331295231815, + "loss": 0.9385, + "step": 7011 + }, + { + "epoch": 1.248397435897436, + "grad_norm": 0.6948656439781189, + "learning_rate": 0.00015582151642021524, + "loss": 0.9425, + "step": 7012 + }, + { + "epoch": 1.2485754985754987, + "grad_norm": 0.6559088230133057, + "learning_rate": 0.0001558099022235423, + "loss": 1.0002, + "step": 7013 + }, + { + "epoch": 1.2487535612535612, + "grad_norm": 0.6097117066383362, + "learning_rate": 0.00015579828693339026, + "loss": 1.0234, + "step": 7014 + }, + { + "epoch": 1.248931623931624, + "grad_norm": 0.6612260341644287, + "learning_rate": 0.00015578667054998673, + "loss": 1.1376, + "step": 7015 + }, + { + "epoch": 1.2491096866096867, + "grad_norm": 0.6305607557296753, + "learning_rate": 0.00015577505307355925, + "loss": 0.9127, + "step": 7016 + }, + { + "epoch": 1.2492877492877492, + "grad_norm": 0.6648319959640503, + "learning_rate": 0.00015576343450433549, + "loss": 0.8697, + "step": 7017 + }, + { + "epoch": 1.249465811965812, + "grad_norm": 0.7642946839332581, + "learning_rate": 0.00015575181484254303, + "loss": 1.0998, + "step": 7018 + }, + { + "epoch": 1.2496438746438747, + "grad_norm": 0.6775243282318115, + "learning_rate": 0.00015574019408840962, + "loss": 1.0186, + "step": 7019 + }, + { + "epoch": 1.2498219373219372, + "grad_norm": 0.6075591444969177, + "learning_rate": 0.00015572857224216286, + "loss": 0.9592, + "step": 7020 + }, + { + "epoch": 1.2498219373219372, + "eval_loss": 1.105136752128601, + "eval_runtime": 24.4793, + "eval_samples_per_second": 42.526, + "eval_steps_per_second": 21.283, + "step": 7020 + }, + { + "epoch": 1.25, + "grad_norm": 0.5856962203979492, + "learning_rate": 0.0001557169493040305, + "loss": 0.8336, + "step": 7021 + }, + { + "epoch": 1.2501780626780628, + "grad_norm": 0.6451364159584045, + "learning_rate": 0.00015570532527424028, + "loss": 0.8805, + "step": 7022 + }, + { + "epoch": 1.2503561253561253, + "grad_norm": 0.6266474723815918, + "learning_rate": 0.00015569370015301991, + "loss": 1.0023, + "step": 7023 + }, + { + "epoch": 1.250534188034188, + "grad_norm": 0.5547378063201904, + "learning_rate": 0.00015568207394059722, + "loss": 0.7385, + "step": 7024 + }, + { + "epoch": 1.2507122507122508, + "grad_norm": 0.604169487953186, + "learning_rate": 0.0001556704466371999, + "loss": 0.9194, + "step": 7025 + }, + { + "epoch": 1.2508903133903133, + "grad_norm": 0.7054405212402344, + "learning_rate": 0.00015565881824305586, + "loss": 1.1864, + "step": 7026 + }, + { + "epoch": 1.251068376068376, + "grad_norm": 0.6429929733276367, + "learning_rate": 0.0001556471887583929, + "loss": 1.0129, + "step": 7027 + }, + { + "epoch": 1.2512464387464388, + "grad_norm": 0.695957362651825, + "learning_rate": 0.00015563555818343887, + "loss": 1.2994, + "step": 7028 + }, + { + "epoch": 1.2514245014245013, + "grad_norm": 0.5889938473701477, + "learning_rate": 0.0001556239265184216, + "loss": 1.0109, + "step": 7029 + }, + { + "epoch": 1.251602564102564, + "grad_norm": 0.6424569487571716, + "learning_rate": 0.0001556122937635691, + "loss": 0.8585, + "step": 7030 + }, + { + "epoch": 1.2517806267806268, + "grad_norm": 0.5561244487762451, + "learning_rate": 0.0001556006599191092, + "loss": 0.9994, + "step": 7031 + }, + { + "epoch": 1.2519586894586894, + "grad_norm": 0.6355302333831787, + "learning_rate": 0.00015558902498526988, + "loss": 0.9495, + "step": 7032 + }, + { + "epoch": 1.2521367521367521, + "grad_norm": 0.6272686719894409, + "learning_rate": 0.00015557738896227908, + "loss": 0.7611, + "step": 7033 + }, + { + "epoch": 1.2523148148148149, + "grad_norm": 0.7069199085235596, + "learning_rate": 0.00015556575185036482, + "loss": 1.0612, + "step": 7034 + }, + { + "epoch": 1.2524928774928774, + "grad_norm": 0.6635094285011292, + "learning_rate": 0.00015555411364975505, + "loss": 1.1182, + "step": 7035 + }, + { + "epoch": 1.2526709401709402, + "grad_norm": 0.6112014651298523, + "learning_rate": 0.00015554247436067785, + "loss": 0.8677, + "step": 7036 + }, + { + "epoch": 1.252849002849003, + "grad_norm": 0.678963303565979, + "learning_rate": 0.00015553083398336126, + "loss": 1.1421, + "step": 7037 + }, + { + "epoch": 1.2530270655270654, + "grad_norm": 0.6291939616203308, + "learning_rate": 0.0001555191925180333, + "loss": 0.9157, + "step": 7038 + }, + { + "epoch": 1.2532051282051282, + "grad_norm": 0.6519795656204224, + "learning_rate": 0.0001555075499649221, + "loss": 1.0074, + "step": 7039 + }, + { + "epoch": 1.253383190883191, + "grad_norm": 0.6063529849052429, + "learning_rate": 0.00015549590632425576, + "loss": 1.0205, + "step": 7040 + }, + { + "epoch": 1.2535612535612537, + "grad_norm": 0.7055633664131165, + "learning_rate": 0.00015548426159626242, + "loss": 1.0254, + "step": 7041 + }, + { + "epoch": 1.2537393162393162, + "grad_norm": 0.6783022880554199, + "learning_rate": 0.00015547261578117025, + "loss": 1.1017, + "step": 7042 + }, + { + "epoch": 1.253917378917379, + "grad_norm": 0.7055003643035889, + "learning_rate": 0.0001554609688792074, + "loss": 1.0269, + "step": 7043 + }, + { + "epoch": 1.2540954415954415, + "grad_norm": 0.6465007662773132, + "learning_rate": 0.0001554493208906021, + "loss": 1.0492, + "step": 7044 + }, + { + "epoch": 1.2542735042735043, + "grad_norm": 0.6443775296211243, + "learning_rate": 0.0001554376718155825, + "loss": 0.9778, + "step": 7045 + }, + { + "epoch": 1.254451566951567, + "grad_norm": 0.695214569568634, + "learning_rate": 0.0001554260216543769, + "loss": 0.8792, + "step": 7046 + }, + { + "epoch": 1.2546296296296298, + "grad_norm": 0.6777814626693726, + "learning_rate": 0.00015541437040721354, + "loss": 0.8944, + "step": 7047 + }, + { + "epoch": 1.2548076923076923, + "grad_norm": 0.6269369721412659, + "learning_rate": 0.0001554027180743207, + "loss": 0.8825, + "step": 7048 + }, + { + "epoch": 1.254985754985755, + "grad_norm": 0.6197061538696289, + "learning_rate": 0.0001553910646559267, + "loss": 0.9823, + "step": 7049 + }, + { + "epoch": 1.2551638176638176, + "grad_norm": 0.681347131729126, + "learning_rate": 0.00015537941015225984, + "loss": 0.995, + "step": 7050 + }, + { + "epoch": 1.2553418803418803, + "grad_norm": 0.6224286556243896, + "learning_rate": 0.00015536775456354848, + "loss": 0.7714, + "step": 7051 + }, + { + "epoch": 1.255519943019943, + "grad_norm": 0.6113278269767761, + "learning_rate": 0.00015535609789002098, + "loss": 0.9859, + "step": 7052 + }, + { + "epoch": 1.2556980056980058, + "grad_norm": 0.6985422372817993, + "learning_rate": 0.00015534444013190577, + "loss": 0.8785, + "step": 7053 + }, + { + "epoch": 1.2558760683760684, + "grad_norm": 0.5602933168411255, + "learning_rate": 0.00015533278128943118, + "loss": 0.8341, + "step": 7054 + }, + { + "epoch": 1.256054131054131, + "grad_norm": 0.587684690952301, + "learning_rate": 0.0001553211213628257, + "loss": 0.7933, + "step": 7055 + }, + { + "epoch": 1.2562321937321936, + "grad_norm": 0.692997932434082, + "learning_rate": 0.0001553094603523178, + "loss": 1.0957, + "step": 7056 + }, + { + "epoch": 1.2564102564102564, + "grad_norm": 0.6925587058067322, + "learning_rate": 0.00015529779825813588, + "loss": 0.8602, + "step": 7057 + }, + { + "epoch": 1.2565883190883191, + "grad_norm": 0.6383063197135925, + "learning_rate": 0.0001552861350805085, + "loss": 0.9933, + "step": 7058 + }, + { + "epoch": 1.256766381766382, + "grad_norm": 0.6520544290542603, + "learning_rate": 0.00015527447081966413, + "loss": 0.9498, + "step": 7059 + }, + { + "epoch": 1.2569444444444444, + "grad_norm": 0.7353914380073547, + "learning_rate": 0.00015526280547583133, + "loss": 1.1071, + "step": 7060 + }, + { + "epoch": 1.2571225071225072, + "grad_norm": 0.7141618132591248, + "learning_rate": 0.00015525113904923864, + "loss": 0.8333, + "step": 7061 + }, + { + "epoch": 1.2573005698005697, + "grad_norm": 0.6194499731063843, + "learning_rate": 0.00015523947154011468, + "loss": 0.9421, + "step": 7062 + }, + { + "epoch": 1.2574786324786325, + "grad_norm": 0.7514514327049255, + "learning_rate": 0.00015522780294868803, + "loss": 1.226, + "step": 7063 + }, + { + "epoch": 1.2576566951566952, + "grad_norm": 0.762923538684845, + "learning_rate": 0.0001552161332751873, + "loss": 1.1893, + "step": 7064 + }, + { + "epoch": 1.257834757834758, + "grad_norm": 0.6265730261802673, + "learning_rate": 0.00015520446251984113, + "loss": 0.6604, + "step": 7065 + }, + { + "epoch": 1.2580128205128205, + "grad_norm": 0.6447750329971313, + "learning_rate": 0.0001551927906828782, + "loss": 0.9814, + "step": 7066 + }, + { + "epoch": 1.2581908831908832, + "grad_norm": 0.5791042447090149, + "learning_rate": 0.00015518111776452722, + "loss": 0.8283, + "step": 7067 + }, + { + "epoch": 1.2583689458689458, + "grad_norm": 0.5267777442932129, + "learning_rate": 0.00015516944376501682, + "loss": 0.5748, + "step": 7068 + }, + { + "epoch": 1.2585470085470085, + "grad_norm": 0.7343912720680237, + "learning_rate": 0.0001551577686845758, + "loss": 1.1777, + "step": 7069 + }, + { + "epoch": 1.2587250712250713, + "grad_norm": 0.645746111869812, + "learning_rate": 0.00015514609252343284, + "loss": 0.9356, + "step": 7070 + }, + { + "epoch": 1.258903133903134, + "grad_norm": 0.6993104219436646, + "learning_rate": 0.0001551344152818168, + "loss": 1.06, + "step": 7071 + }, + { + "epoch": 1.2590811965811965, + "grad_norm": 0.6661365628242493, + "learning_rate": 0.0001551227369599564, + "loss": 1.061, + "step": 7072 + }, + { + "epoch": 1.2592592592592593, + "grad_norm": 0.7833736538887024, + "learning_rate": 0.0001551110575580805, + "loss": 0.9674, + "step": 7073 + }, + { + "epoch": 1.2594373219373218, + "grad_norm": 0.5878575444221497, + "learning_rate": 0.00015509937707641787, + "loss": 0.9002, + "step": 7074 + }, + { + "epoch": 1.2596153846153846, + "grad_norm": 0.6402907371520996, + "learning_rate": 0.00015508769551519745, + "loss": 1.0157, + "step": 7075 + }, + { + "epoch": 1.2597934472934473, + "grad_norm": 0.6794611215591431, + "learning_rate": 0.00015507601287464805, + "loss": 1.052, + "step": 7076 + }, + { + "epoch": 1.25997150997151, + "grad_norm": 0.706922173500061, + "learning_rate": 0.0001550643291549986, + "loss": 1.0814, + "step": 7077 + }, + { + "epoch": 1.2601495726495726, + "grad_norm": 0.6722953915596008, + "learning_rate": 0.000155052644356478, + "loss": 1.1402, + "step": 7078 + }, + { + "epoch": 1.2603276353276354, + "grad_norm": 0.6619611978530884, + "learning_rate": 0.00015504095847931518, + "loss": 0.9583, + "step": 7079 + }, + { + "epoch": 1.260505698005698, + "grad_norm": 0.5645583271980286, + "learning_rate": 0.00015502927152373914, + "loss": 0.6746, + "step": 7080 + }, + { + "epoch": 1.2606837606837606, + "grad_norm": 0.6634977459907532, + "learning_rate": 0.00015501758348997882, + "loss": 1.0451, + "step": 7081 + }, + { + "epoch": 1.2608618233618234, + "grad_norm": 0.7167651057243347, + "learning_rate": 0.00015500589437826326, + "loss": 0.931, + "step": 7082 + }, + { + "epoch": 1.2610398860398861, + "grad_norm": 0.6179340481758118, + "learning_rate": 0.00015499420418882146, + "loss": 1.0953, + "step": 7083 + }, + { + "epoch": 1.2612179487179487, + "grad_norm": 0.6948468685150146, + "learning_rate": 0.00015498251292188247, + "loss": 1.0277, + "step": 7084 + }, + { + "epoch": 1.2613960113960114, + "grad_norm": 0.6256045699119568, + "learning_rate": 0.00015497082057767532, + "loss": 1.0154, + "step": 7085 + }, + { + "epoch": 1.261574074074074, + "grad_norm": 0.6457428336143494, + "learning_rate": 0.0001549591271564292, + "loss": 0.9693, + "step": 7086 + }, + { + "epoch": 1.2617521367521367, + "grad_norm": 0.722259521484375, + "learning_rate": 0.0001549474326583731, + "loss": 0.9176, + "step": 7087 + }, + { + "epoch": 1.2619301994301995, + "grad_norm": 0.742477297782898, + "learning_rate": 0.0001549357370837362, + "loss": 0.9813, + "step": 7088 + }, + { + "epoch": 1.2621082621082622, + "grad_norm": 0.5981723666191101, + "learning_rate": 0.0001549240404327477, + "loss": 0.8943, + "step": 7089 + }, + { + "epoch": 1.2622863247863247, + "grad_norm": 0.6266574859619141, + "learning_rate": 0.00015491234270563665, + "loss": 0.8439, + "step": 7090 + }, + { + "epoch": 1.2624643874643875, + "grad_norm": 0.6723998188972473, + "learning_rate": 0.00015490064390263238, + "loss": 1.2278, + "step": 7091 + }, + { + "epoch": 1.26264245014245, + "grad_norm": 0.6628100275993347, + "learning_rate": 0.00015488894402396398, + "loss": 0.9526, + "step": 7092 + }, + { + "epoch": 1.2628205128205128, + "grad_norm": 0.6661350727081299, + "learning_rate": 0.0001548772430698608, + "loss": 0.974, + "step": 7093 + }, + { + "epoch": 1.2629985754985755, + "grad_norm": 0.8210669755935669, + "learning_rate": 0.000154865541040552, + "loss": 1.1142, + "step": 7094 + }, + { + "epoch": 1.2631766381766383, + "grad_norm": 0.6329003572463989, + "learning_rate": 0.0001548538379362669, + "loss": 0.8485, + "step": 7095 + }, + { + "epoch": 1.2633547008547008, + "grad_norm": 0.6288384795188904, + "learning_rate": 0.0001548421337572348, + "loss": 0.816, + "step": 7096 + }, + { + "epoch": 1.2635327635327636, + "grad_norm": 0.631060004234314, + "learning_rate": 0.00015483042850368504, + "loss": 0.8237, + "step": 7097 + }, + { + "epoch": 1.263710826210826, + "grad_norm": 0.7343839406967163, + "learning_rate": 0.0001548187221758469, + "loss": 1.1507, + "step": 7098 + }, + { + "epoch": 1.2638888888888888, + "grad_norm": 0.6313042640686035, + "learning_rate": 0.0001548070147739498, + "loss": 0.7762, + "step": 7099 + }, + { + "epoch": 1.2640669515669516, + "grad_norm": 0.6449850797653198, + "learning_rate": 0.00015479530629822308, + "loss": 0.9225, + "step": 7100 + }, + { + "epoch": 1.2642450142450143, + "grad_norm": 0.6371589303016663, + "learning_rate": 0.00015478359674889617, + "loss": 1.0088, + "step": 7101 + }, + { + "epoch": 1.2644230769230769, + "grad_norm": 0.6483678221702576, + "learning_rate": 0.00015477188612619849, + "loss": 0.6234, + "step": 7102 + }, + { + "epoch": 1.2646011396011396, + "grad_norm": 0.6945441365242004, + "learning_rate": 0.00015476017443035947, + "loss": 1.123, + "step": 7103 + }, + { + "epoch": 1.2647792022792022, + "grad_norm": 0.6356340050697327, + "learning_rate": 0.00015474846166160856, + "loss": 0.9923, + "step": 7104 + }, + { + "epoch": 1.264957264957265, + "grad_norm": 0.6774702668190002, + "learning_rate": 0.00015473674782017532, + "loss": 0.9694, + "step": 7105 + }, + { + "epoch": 1.2651353276353277, + "grad_norm": 0.6332793831825256, + "learning_rate": 0.0001547250329062892, + "loss": 1.0633, + "step": 7106 + }, + { + "epoch": 1.2653133903133904, + "grad_norm": 0.6563684344291687, + "learning_rate": 0.00015471331692017972, + "loss": 1.0893, + "step": 7107 + }, + { + "epoch": 1.265491452991453, + "grad_norm": 0.7318371534347534, + "learning_rate": 0.0001547015998620765, + "loss": 1.1777, + "step": 7108 + }, + { + "epoch": 1.2656695156695157, + "grad_norm": 0.7099173069000244, + "learning_rate": 0.000154689881732209, + "loss": 1.1717, + "step": 7109 + }, + { + "epoch": 1.2658475783475782, + "grad_norm": 0.661078691482544, + "learning_rate": 0.00015467816253080693, + "loss": 1.0448, + "step": 7110 + }, + { + "epoch": 1.266025641025641, + "grad_norm": 0.6206802129745483, + "learning_rate": 0.0001546664422580998, + "loss": 0.9334, + "step": 7111 + }, + { + "epoch": 1.2662037037037037, + "grad_norm": 0.6514355540275574, + "learning_rate": 0.00015465472091431728, + "loss": 0.9533, + "step": 7112 + }, + { + "epoch": 1.2663817663817665, + "grad_norm": 0.6090209484100342, + "learning_rate": 0.0001546429984996891, + "loss": 0.9206, + "step": 7113 + }, + { + "epoch": 1.266559829059829, + "grad_norm": 0.6345987915992737, + "learning_rate": 0.00015463127501444488, + "loss": 1.0537, + "step": 7114 + }, + { + "epoch": 1.2667378917378918, + "grad_norm": 0.6095160245895386, + "learning_rate": 0.0001546195504588143, + "loss": 0.8652, + "step": 7115 + }, + { + "epoch": 1.2669159544159543, + "grad_norm": 0.6751621961593628, + "learning_rate": 0.00015460782483302707, + "loss": 0.9001, + "step": 7116 + }, + { + "epoch": 1.267094017094017, + "grad_norm": 0.6261575222015381, + "learning_rate": 0.00015459609813731295, + "loss": 0.929, + "step": 7117 + }, + { + "epoch": 1.2672720797720798, + "grad_norm": 0.589495837688446, + "learning_rate": 0.0001545843703719017, + "loss": 0.9023, + "step": 7118 + }, + { + "epoch": 1.2674501424501425, + "grad_norm": 0.6364617943763733, + "learning_rate": 0.00015457264153702311, + "loss": 0.8261, + "step": 7119 + }, + { + "epoch": 1.267628205128205, + "grad_norm": 0.6685599684715271, + "learning_rate": 0.00015456091163290698, + "loss": 1.1267, + "step": 7120 + }, + { + "epoch": 1.2678062678062678, + "grad_norm": 0.6440932750701904, + "learning_rate": 0.0001545491806597831, + "loss": 0.9643, + "step": 7121 + }, + { + "epoch": 1.2679843304843303, + "grad_norm": 0.7641597390174866, + "learning_rate": 0.00015453744861788137, + "loss": 1.1577, + "step": 7122 + }, + { + "epoch": 1.268162393162393, + "grad_norm": 0.6965937614440918, + "learning_rate": 0.00015452571550743163, + "loss": 0.7835, + "step": 7123 + }, + { + "epoch": 1.2683404558404558, + "grad_norm": 0.6332844495773315, + "learning_rate": 0.00015451398132866376, + "loss": 0.9794, + "step": 7124 + }, + { + "epoch": 1.2685185185185186, + "grad_norm": 0.6719903349876404, + "learning_rate": 0.00015450224608180765, + "loss": 0.9795, + "step": 7125 + }, + { + "epoch": 1.2686965811965811, + "grad_norm": 0.567414402961731, + "learning_rate": 0.00015449050976709328, + "loss": 0.9737, + "step": 7126 + }, + { + "epoch": 1.2688746438746439, + "grad_norm": 0.6810645461082458, + "learning_rate": 0.0001544787723847505, + "loss": 1.2358, + "step": 7127 + }, + { + "epoch": 1.2690527065527066, + "grad_norm": 0.6693191528320312, + "learning_rate": 0.00015446703393500938, + "loss": 0.9475, + "step": 7128 + }, + { + "epoch": 1.2692307692307692, + "grad_norm": 0.7077522277832031, + "learning_rate": 0.00015445529441809988, + "loss": 1.013, + "step": 7129 + }, + { + "epoch": 1.269408831908832, + "grad_norm": 0.6596258878707886, + "learning_rate": 0.000154443553834252, + "loss": 1.1506, + "step": 7130 + }, + { + "epoch": 1.2695868945868947, + "grad_norm": 0.6721500754356384, + "learning_rate": 0.0001544318121836958, + "loss": 0.8848, + "step": 7131 + }, + { + "epoch": 1.2697649572649572, + "grad_norm": 0.6943998336791992, + "learning_rate": 0.00015442006946666132, + "loss": 1.1118, + "step": 7132 + }, + { + "epoch": 1.26994301994302, + "grad_norm": 0.6132234930992126, + "learning_rate": 0.0001544083256833786, + "loss": 0.9932, + "step": 7133 + }, + { + "epoch": 1.2701210826210827, + "grad_norm": 0.7337939739227295, + "learning_rate": 0.00015439658083407775, + "loss": 1.0973, + "step": 7134 + }, + { + "epoch": 1.2702991452991452, + "grad_norm": 0.6551772356033325, + "learning_rate": 0.00015438483491898893, + "loss": 1.0006, + "step": 7135 + }, + { + "epoch": 1.270477207977208, + "grad_norm": 0.660068929195404, + "learning_rate": 0.00015437308793834223, + "loss": 0.9291, + "step": 7136 + }, + { + "epoch": 1.2706552706552707, + "grad_norm": 0.7622788548469543, + "learning_rate": 0.00015436133989236783, + "loss": 1.0782, + "step": 7137 + }, + { + "epoch": 1.2708333333333333, + "grad_norm": 0.848494291305542, + "learning_rate": 0.00015434959078129587, + "loss": 1.2001, + "step": 7138 + }, + { + "epoch": 1.271011396011396, + "grad_norm": 0.6222602725028992, + "learning_rate": 0.0001543378406053566, + "loss": 1.011, + "step": 7139 + }, + { + "epoch": 1.2711894586894588, + "grad_norm": 0.6164663434028625, + "learning_rate": 0.00015432608936478026, + "loss": 1.0282, + "step": 7140 + }, + { + "epoch": 1.2713675213675213, + "grad_norm": 0.7236546277999878, + "learning_rate": 0.000154314337059797, + "loss": 1.0112, + "step": 7141 + }, + { + "epoch": 1.271545584045584, + "grad_norm": 0.6891111135482788, + "learning_rate": 0.00015430258369063715, + "loss": 1.1191, + "step": 7142 + }, + { + "epoch": 1.2717236467236468, + "grad_norm": 0.6600295901298523, + "learning_rate": 0.00015429082925753099, + "loss": 0.9561, + "step": 7143 + }, + { + "epoch": 1.2719017094017093, + "grad_norm": 0.6819902062416077, + "learning_rate": 0.0001542790737607088, + "loss": 1.0631, + "step": 7144 + }, + { + "epoch": 1.272079772079772, + "grad_norm": 0.6518470644950867, + "learning_rate": 0.0001542673172004009, + "loss": 1.0806, + "step": 7145 + }, + { + "epoch": 1.2722578347578348, + "grad_norm": 0.737501859664917, + "learning_rate": 0.00015425555957683767, + "loss": 1.0144, + "step": 7146 + }, + { + "epoch": 1.2724358974358974, + "grad_norm": 0.6245740652084351, + "learning_rate": 0.00015424380089024944, + "loss": 1.0612, + "step": 7147 + }, + { + "epoch": 1.27261396011396, + "grad_norm": 0.7118125557899475, + "learning_rate": 0.0001542320411408666, + "loss": 1.1458, + "step": 7148 + }, + { + "epoch": 1.2727920227920229, + "grad_norm": 0.6965761780738831, + "learning_rate": 0.00015422028032891958, + "loss": 0.8052, + "step": 7149 + }, + { + "epoch": 1.2729700854700854, + "grad_norm": 0.7661466598510742, + "learning_rate": 0.0001542085184546388, + "loss": 1.1245, + "step": 7150 + }, + { + "epoch": 1.2731481481481481, + "grad_norm": 0.7238876223564148, + "learning_rate": 0.00015419675551825475, + "loss": 0.9346, + "step": 7151 + }, + { + "epoch": 1.273326210826211, + "grad_norm": 0.669562041759491, + "learning_rate": 0.0001541849915199978, + "loss": 0.7816, + "step": 7152 + }, + { + "epoch": 1.2735042735042734, + "grad_norm": 0.6799174547195435, + "learning_rate": 0.00015417322646009855, + "loss": 1.047, + "step": 7153 + }, + { + "epoch": 1.2736823361823362, + "grad_norm": 0.6012796759605408, + "learning_rate": 0.00015416146033878745, + "loss": 1.0101, + "step": 7154 + }, + { + "epoch": 1.273860398860399, + "grad_norm": 0.7008427977561951, + "learning_rate": 0.00015414969315629505, + "loss": 1.1321, + "step": 7155 + }, + { + "epoch": 1.2740384615384617, + "grad_norm": 0.6555556058883667, + "learning_rate": 0.0001541379249128519, + "loss": 0.9926, + "step": 7156 + }, + { + "epoch": 1.2742165242165242, + "grad_norm": 0.6324251294136047, + "learning_rate": 0.00015412615560868854, + "loss": 0.9051, + "step": 7157 + }, + { + "epoch": 1.274394586894587, + "grad_norm": 0.6035568714141846, + "learning_rate": 0.0001541143852440356, + "loss": 0.8248, + "step": 7158 + }, + { + "epoch": 1.2745726495726495, + "grad_norm": 0.6733569502830505, + "learning_rate": 0.0001541026138191237, + "loss": 0.9149, + "step": 7159 + }, + { + "epoch": 1.2747507122507122, + "grad_norm": 0.8306798338890076, + "learning_rate": 0.0001540908413341835, + "loss": 1.0694, + "step": 7160 + }, + { + "epoch": 1.274928774928775, + "grad_norm": 0.6649713516235352, + "learning_rate": 0.00015407906778944563, + "loss": 1.1358, + "step": 7161 + }, + { + "epoch": 1.2751068376068377, + "grad_norm": 0.6889697909355164, + "learning_rate": 0.00015406729318514074, + "loss": 1.0096, + "step": 7162 + }, + { + "epoch": 1.2752849002849003, + "grad_norm": 0.6948645114898682, + "learning_rate": 0.0001540555175214996, + "loss": 1.0649, + "step": 7163 + }, + { + "epoch": 1.275462962962963, + "grad_norm": 0.6844844818115234, + "learning_rate": 0.0001540437407987528, + "loss": 0.884, + "step": 7164 + }, + { + "epoch": 1.2756410256410255, + "grad_norm": 0.7124526500701904, + "learning_rate": 0.00015403196301713124, + "loss": 1.1307, + "step": 7165 + }, + { + "epoch": 1.2758190883190883, + "grad_norm": 0.7328375577926636, + "learning_rate": 0.00015402018417686556, + "loss": 1.0348, + "step": 7166 + }, + { + "epoch": 1.275997150997151, + "grad_norm": 0.5872696042060852, + "learning_rate": 0.00015400840427818663, + "loss": 0.9827, + "step": 7167 + }, + { + "epoch": 1.2761752136752138, + "grad_norm": 0.6370702385902405, + "learning_rate": 0.00015399662332132519, + "loss": 0.9171, + "step": 7168 + }, + { + "epoch": 1.2763532763532763, + "grad_norm": 0.6481866240501404, + "learning_rate": 0.00015398484130651205, + "loss": 0.8704, + "step": 7169 + }, + { + "epoch": 1.276531339031339, + "grad_norm": 0.598739743232727, + "learning_rate": 0.00015397305823397812, + "loss": 0.8097, + "step": 7170 + }, + { + "epoch": 1.2767094017094016, + "grad_norm": 0.5941228270530701, + "learning_rate": 0.00015396127410395423, + "loss": 0.8853, + "step": 7171 + }, + { + "epoch": 1.2768874643874644, + "grad_norm": 0.6485885381698608, + "learning_rate": 0.00015394948891667127, + "loss": 0.702, + "step": 7172 + }, + { + "epoch": 1.2770655270655271, + "grad_norm": 0.5314942598342896, + "learning_rate": 0.00015393770267236017, + "loss": 0.7899, + "step": 7173 + }, + { + "epoch": 1.2772435897435899, + "grad_norm": 0.6113781929016113, + "learning_rate": 0.00015392591537125182, + "loss": 0.9871, + "step": 7174 + }, + { + "epoch": 1.2774216524216524, + "grad_norm": 0.5625866651535034, + "learning_rate": 0.00015391412701357715, + "loss": 0.8246, + "step": 7175 + }, + { + "epoch": 1.2775997150997151, + "grad_norm": 0.6006998419761658, + "learning_rate": 0.00015390233759956718, + "loss": 0.899, + "step": 7176 + }, + { + "epoch": 1.2777777777777777, + "grad_norm": 0.6916918158531189, + "learning_rate": 0.0001538905471294529, + "loss": 1.0443, + "step": 7177 + }, + { + "epoch": 1.2779558404558404, + "grad_norm": 0.6263536810874939, + "learning_rate": 0.00015387875560346525, + "loss": 0.9159, + "step": 7178 + }, + { + "epoch": 1.2781339031339032, + "grad_norm": 0.6563085913658142, + "learning_rate": 0.00015386696302183535, + "loss": 0.994, + "step": 7179 + }, + { + "epoch": 1.278311965811966, + "grad_norm": 0.6312007904052734, + "learning_rate": 0.00015385516938479416, + "loss": 0.9148, + "step": 7180 + }, + { + "epoch": 1.2784900284900285, + "grad_norm": 0.6408209204673767, + "learning_rate": 0.00015384337469257284, + "loss": 1.0508, + "step": 7181 + }, + { + "epoch": 1.2786680911680912, + "grad_norm": 0.656234085559845, + "learning_rate": 0.00015383157894540244, + "loss": 0.9952, + "step": 7182 + }, + { + "epoch": 1.2788461538461537, + "grad_norm": 0.7401639819145203, + "learning_rate": 0.00015381978214351407, + "loss": 1.1615, + "step": 7183 + }, + { + "epoch": 1.2790242165242165, + "grad_norm": 0.5746055841445923, + "learning_rate": 0.00015380798428713885, + "loss": 0.9142, + "step": 7184 + }, + { + "epoch": 1.2792022792022792, + "grad_norm": 0.8061720728874207, + "learning_rate": 0.00015379618537650797, + "loss": 1.13, + "step": 7185 + }, + { + "epoch": 1.279380341880342, + "grad_norm": 0.6336073875427246, + "learning_rate": 0.0001537843854118526, + "loss": 1.0581, + "step": 7186 + }, + { + "epoch": 1.2795584045584045, + "grad_norm": 0.6549856066703796, + "learning_rate": 0.0001537725843934039, + "loss": 1.09, + "step": 7187 + }, + { + "epoch": 1.2797364672364673, + "grad_norm": 0.5759010910987854, + "learning_rate": 0.00015376078232139315, + "loss": 0.8441, + "step": 7188 + }, + { + "epoch": 1.2799145299145298, + "grad_norm": 0.5733884572982788, + "learning_rate": 0.00015374897919605152, + "loss": 0.9086, + "step": 7189 + }, + { + "epoch": 1.2800925925925926, + "grad_norm": 0.6505870819091797, + "learning_rate": 0.0001537371750176103, + "loss": 1.1683, + "step": 7190 + }, + { + "epoch": 1.2802706552706553, + "grad_norm": 0.6744688153266907, + "learning_rate": 0.00015372536978630077, + "loss": 0.9483, + "step": 7191 + }, + { + "epoch": 1.280448717948718, + "grad_norm": 0.598098874092102, + "learning_rate": 0.0001537135635023542, + "loss": 0.7747, + "step": 7192 + }, + { + "epoch": 1.2806267806267806, + "grad_norm": 0.6711761951446533, + "learning_rate": 0.00015370175616600195, + "loss": 1.1897, + "step": 7193 + }, + { + "epoch": 1.2808048433048433, + "grad_norm": 0.6207453608512878, + "learning_rate": 0.00015368994777747536, + "loss": 1.0063, + "step": 7194 + }, + { + "epoch": 1.2809829059829059, + "grad_norm": 0.6701686382293701, + "learning_rate": 0.00015367813833700575, + "loss": 1.0864, + "step": 7195 + }, + { + "epoch": 1.2811609686609686, + "grad_norm": 0.5916469693183899, + "learning_rate": 0.00015366632784482456, + "loss": 0.8786, + "step": 7196 + }, + { + "epoch": 1.2813390313390314, + "grad_norm": 0.6567547917366028, + "learning_rate": 0.00015365451630116312, + "loss": 0.9977, + "step": 7197 + }, + { + "epoch": 1.2815170940170941, + "grad_norm": 0.7287433743476868, + "learning_rate": 0.00015364270370625294, + "loss": 1.1248, + "step": 7198 + }, + { + "epoch": 1.2816951566951567, + "grad_norm": 0.7736039161682129, + "learning_rate": 0.0001536308900603254, + "loss": 0.9832, + "step": 7199 + }, + { + "epoch": 1.2818732193732194, + "grad_norm": 0.6799852252006531, + "learning_rate": 0.00015361907536361194, + "loss": 1.0275, + "step": 7200 + }, + { + "epoch": 1.282051282051282, + "grad_norm": 0.5975812673568726, + "learning_rate": 0.00015360725961634407, + "loss": 1.0516, + "step": 7201 + }, + { + "epoch": 1.2822293447293447, + "grad_norm": 0.616307258605957, + "learning_rate": 0.00015359544281875337, + "loss": 0.8095, + "step": 7202 + }, + { + "epoch": 1.2824074074074074, + "grad_norm": 0.6357580423355103, + "learning_rate": 0.00015358362497107126, + "loss": 0.9186, + "step": 7203 + }, + { + "epoch": 1.2825854700854702, + "grad_norm": 0.679333508014679, + "learning_rate": 0.00015357180607352935, + "loss": 0.9433, + "step": 7204 + }, + { + "epoch": 1.2827635327635327, + "grad_norm": 0.6345439553260803, + "learning_rate": 0.00015355998612635914, + "loss": 0.9186, + "step": 7205 + }, + { + "epoch": 1.2829415954415955, + "grad_norm": 0.6256508827209473, + "learning_rate": 0.00015354816512979231, + "loss": 0.9984, + "step": 7206 + }, + { + "epoch": 1.283119658119658, + "grad_norm": 0.7973852753639221, + "learning_rate": 0.00015353634308406044, + "loss": 1.1145, + "step": 7207 + }, + { + "epoch": 1.2832977207977208, + "grad_norm": 0.711125910282135, + "learning_rate": 0.0001535245199893951, + "loss": 1.1947, + "step": 7208 + }, + { + "epoch": 1.2834757834757835, + "grad_norm": 0.6096055507659912, + "learning_rate": 0.00015351269584602798, + "loss": 1.0078, + "step": 7209 + }, + { + "epoch": 1.2836538461538463, + "grad_norm": 0.7089232802391052, + "learning_rate": 0.00015350087065419077, + "loss": 1.112, + "step": 7210 + }, + { + "epoch": 1.2838319088319088, + "grad_norm": 0.716199517250061, + "learning_rate": 0.00015348904441411508, + "loss": 1.1015, + "step": 7211 + }, + { + "epoch": 1.2840099715099715, + "grad_norm": 0.6374632716178894, + "learning_rate": 0.00015347721712603276, + "loss": 1.0519, + "step": 7212 + }, + { + "epoch": 1.284188034188034, + "grad_norm": 0.6500036716461182, + "learning_rate": 0.0001534653887901754, + "loss": 1.1719, + "step": 7213 + }, + { + "epoch": 1.2843660968660968, + "grad_norm": 0.7249937653541565, + "learning_rate": 0.00015345355940677485, + "loss": 1.0188, + "step": 7214 + }, + { + "epoch": 1.2845441595441596, + "grad_norm": 0.6645919680595398, + "learning_rate": 0.00015344172897606285, + "loss": 0.9788, + "step": 7215 + }, + { + "epoch": 1.2847222222222223, + "grad_norm": 0.7032710313796997, + "learning_rate": 0.00015342989749827113, + "loss": 1.1093, + "step": 7216 + }, + { + "epoch": 1.2849002849002849, + "grad_norm": 0.622767984867096, + "learning_rate": 0.0001534180649736316, + "loss": 0.8978, + "step": 7217 + }, + { + "epoch": 1.2850783475783476, + "grad_norm": 0.7499693036079407, + "learning_rate": 0.00015340623140237605, + "loss": 1.2232, + "step": 7218 + }, + { + "epoch": 1.2852564102564101, + "grad_norm": 0.6308625936508179, + "learning_rate": 0.00015339439678473636, + "loss": 0.8621, + "step": 7219 + }, + { + "epoch": 1.2854344729344729, + "grad_norm": 0.6513667106628418, + "learning_rate": 0.00015338256112094434, + "loss": 1.0541, + "step": 7220 + }, + { + "epoch": 1.2856125356125356, + "grad_norm": 0.6080937385559082, + "learning_rate": 0.00015337072441123193, + "loss": 0.8474, + "step": 7221 + }, + { + "epoch": 1.2857905982905984, + "grad_norm": 0.6742652058601379, + "learning_rate": 0.00015335888665583104, + "loss": 1.0172, + "step": 7222 + }, + { + "epoch": 1.285968660968661, + "grad_norm": 0.620810866355896, + "learning_rate": 0.00015334704785497364, + "loss": 1.049, + "step": 7223 + }, + { + "epoch": 1.2861467236467237, + "grad_norm": 0.5733018517494202, + "learning_rate": 0.00015333520800889165, + "loss": 0.7371, + "step": 7224 + }, + { + "epoch": 1.2863247863247862, + "grad_norm": 0.6447640061378479, + "learning_rate": 0.00015332336711781702, + "loss": 0.9925, + "step": 7225 + }, + { + "epoch": 1.286502849002849, + "grad_norm": 0.6764999628067017, + "learning_rate": 0.00015331152518198183, + "loss": 0.9052, + "step": 7226 + }, + { + "epoch": 1.2866809116809117, + "grad_norm": 0.6492836475372314, + "learning_rate": 0.00015329968220161803, + "loss": 0.9493, + "step": 7227 + }, + { + "epoch": 1.2868589743589745, + "grad_norm": 0.666157603263855, + "learning_rate": 0.00015328783817695766, + "loss": 1.0626, + "step": 7228 + }, + { + "epoch": 1.287037037037037, + "grad_norm": 0.7098026871681213, + "learning_rate": 0.00015327599310823283, + "loss": 1.0461, + "step": 7229 + }, + { + "epoch": 1.2872150997150997, + "grad_norm": 0.637778103351593, + "learning_rate": 0.00015326414699567555, + "loss": 0.9383, + "step": 7230 + }, + { + "epoch": 1.2873931623931623, + "grad_norm": 0.6816399693489075, + "learning_rate": 0.00015325229983951798, + "loss": 1.0647, + "step": 7231 + }, + { + "epoch": 1.287571225071225, + "grad_norm": 0.668689489364624, + "learning_rate": 0.0001532404516399922, + "loss": 1.0479, + "step": 7232 + }, + { + "epoch": 1.2877492877492878, + "grad_norm": 0.6459103226661682, + "learning_rate": 0.0001532286023973304, + "loss": 1.1751, + "step": 7233 + }, + { + "epoch": 1.2879273504273505, + "grad_norm": 0.679999589920044, + "learning_rate": 0.00015321675211176468, + "loss": 0.7541, + "step": 7234 + }, + { + "epoch": 1.288105413105413, + "grad_norm": 0.5415067672729492, + "learning_rate": 0.00015320490078352724, + "loss": 0.822, + "step": 7235 + }, + { + "epoch": 1.2882834757834758, + "grad_norm": 0.6817963719367981, + "learning_rate": 0.00015319304841285032, + "loss": 0.9424, + "step": 7236 + }, + { + "epoch": 1.2884615384615383, + "grad_norm": 0.6187505125999451, + "learning_rate": 0.0001531811949999661, + "loss": 0.8596, + "step": 7237 + }, + { + "epoch": 1.288639601139601, + "grad_norm": 0.6737838387489319, + "learning_rate": 0.00015316934054510685, + "loss": 1.0046, + "step": 7238 + }, + { + "epoch": 1.2888176638176638, + "grad_norm": 0.6445996761322021, + "learning_rate": 0.00015315748504850482, + "loss": 1.01, + "step": 7239 + }, + { + "epoch": 1.2889957264957266, + "grad_norm": 0.7279136180877686, + "learning_rate": 0.0001531456285103923, + "loss": 0.9066, + "step": 7240 + }, + { + "epoch": 1.289173789173789, + "grad_norm": 0.6619178652763367, + "learning_rate": 0.00015313377093100153, + "loss": 0.8977, + "step": 7241 + }, + { + "epoch": 1.2893518518518519, + "grad_norm": 0.7644323110580444, + "learning_rate": 0.000153121912310565, + "loss": 1.3085, + "step": 7242 + }, + { + "epoch": 1.2895299145299146, + "grad_norm": 0.645882248878479, + "learning_rate": 0.00015311005264931487, + "loss": 1.0337, + "step": 7243 + }, + { + "epoch": 1.2897079772079771, + "grad_norm": 0.6868017911911011, + "learning_rate": 0.0001530981919474836, + "loss": 0.9616, + "step": 7244 + }, + { + "epoch": 1.28988603988604, + "grad_norm": 0.7176693677902222, + "learning_rate": 0.00015308633020530362, + "loss": 1.1975, + "step": 7245 + }, + { + "epoch": 1.2900641025641026, + "grad_norm": 0.7358015775680542, + "learning_rate": 0.00015307446742300718, + "loss": 0.9308, + "step": 7246 + }, + { + "epoch": 1.2902421652421652, + "grad_norm": 0.7330248355865479, + "learning_rate": 0.00015306260360082688, + "loss": 0.9518, + "step": 7247 + }, + { + "epoch": 1.290420227920228, + "grad_norm": 0.6571981310844421, + "learning_rate": 0.00015305073873899503, + "loss": 0.9531, + "step": 7248 + }, + { + "epoch": 1.2905982905982907, + "grad_norm": 0.5968486666679382, + "learning_rate": 0.00015303887283774417, + "loss": 0.9245, + "step": 7249 + }, + { + "epoch": 1.2907763532763532, + "grad_norm": 0.6398176550865173, + "learning_rate": 0.0001530270058973068, + "loss": 1.0452, + "step": 7250 + }, + { + "epoch": 1.290954415954416, + "grad_norm": 0.5462267994880676, + "learning_rate": 0.00015301513791791542, + "loss": 0.8451, + "step": 7251 + }, + { + "epoch": 1.2911324786324787, + "grad_norm": 0.7536166906356812, + "learning_rate": 0.00015300326889980252, + "loss": 1.0086, + "step": 7252 + }, + { + "epoch": 1.2913105413105412, + "grad_norm": 0.6208569407463074, + "learning_rate": 0.00015299139884320065, + "loss": 0.7437, + "step": 7253 + }, + { + "epoch": 1.291488603988604, + "grad_norm": 0.7025452852249146, + "learning_rate": 0.00015297952774834242, + "loss": 0.8874, + "step": 7254 + }, + { + "epoch": 1.2916666666666667, + "grad_norm": 0.6758308410644531, + "learning_rate": 0.00015296765561546041, + "loss": 1.0378, + "step": 7255 + }, + { + "epoch": 1.2918447293447293, + "grad_norm": 0.7170431613922119, + "learning_rate": 0.00015295578244478724, + "loss": 1.0111, + "step": 7256 + }, + { + "epoch": 1.292022792022792, + "grad_norm": 0.6263511180877686, + "learning_rate": 0.00015294390823655544, + "loss": 0.7836, + "step": 7257 + }, + { + "epoch": 1.2922008547008548, + "grad_norm": 0.5887803435325623, + "learning_rate": 0.0001529320329909978, + "loss": 1.068, + "step": 7258 + }, + { + "epoch": 1.2923789173789173, + "grad_norm": 0.5955889821052551, + "learning_rate": 0.00015292015670834692, + "loss": 0.8903, + "step": 7259 + }, + { + "epoch": 1.29255698005698, + "grad_norm": 0.630449652671814, + "learning_rate": 0.00015290827938883552, + "loss": 1.1096, + "step": 7260 + }, + { + "epoch": 1.2927350427350428, + "grad_norm": 0.7405480146408081, + "learning_rate": 0.00015289640103269625, + "loss": 1.0648, + "step": 7261 + }, + { + "epoch": 1.2929131054131053, + "grad_norm": 0.6082221865653992, + "learning_rate": 0.00015288452164016191, + "loss": 0.9266, + "step": 7262 + }, + { + "epoch": 1.293091168091168, + "grad_norm": 0.6211720108985901, + "learning_rate": 0.00015287264121146524, + "loss": 0.849, + "step": 7263 + }, + { + "epoch": 1.2932692307692308, + "grad_norm": 0.6481043100357056, + "learning_rate": 0.00015286075974683898, + "loss": 0.7761, + "step": 7264 + }, + { + "epoch": 1.2934472934472934, + "grad_norm": 0.5957167744636536, + "learning_rate": 0.00015284887724651593, + "loss": 0.8942, + "step": 7265 + }, + { + "epoch": 1.2936253561253561, + "grad_norm": 0.7272268533706665, + "learning_rate": 0.00015283699371072894, + "loss": 1.0913, + "step": 7266 + }, + { + "epoch": 1.2938034188034189, + "grad_norm": 0.5902758836746216, + "learning_rate": 0.0001528251091397108, + "loss": 1.1045, + "step": 7267 + }, + { + "epoch": 1.2939814814814814, + "grad_norm": 0.6382482051849365, + "learning_rate": 0.00015281322353369436, + "loss": 0.9265, + "step": 7268 + }, + { + "epoch": 1.2941595441595442, + "grad_norm": 0.6556048393249512, + "learning_rate": 0.00015280133689291256, + "loss": 1.0536, + "step": 7269 + }, + { + "epoch": 1.294337606837607, + "grad_norm": 0.680895209312439, + "learning_rate": 0.00015278944921759822, + "loss": 0.9996, + "step": 7270 + }, + { + "epoch": 1.2945156695156697, + "grad_norm": 0.670317530632019, + "learning_rate": 0.00015277756050798428, + "loss": 1.1402, + "step": 7271 + }, + { + "epoch": 1.2946937321937322, + "grad_norm": 0.6312688589096069, + "learning_rate": 0.0001527656707643037, + "loss": 1.0669, + "step": 7272 + }, + { + "epoch": 1.294871794871795, + "grad_norm": 0.6267009973526001, + "learning_rate": 0.0001527537799867894, + "loss": 0.8985, + "step": 7273 + }, + { + "epoch": 1.2950498575498575, + "grad_norm": 0.7069001197814941, + "learning_rate": 0.00015274188817567436, + "loss": 0.9478, + "step": 7274 + }, + { + "epoch": 1.2952279202279202, + "grad_norm": 0.7229067087173462, + "learning_rate": 0.00015272999533119162, + "loss": 0.9005, + "step": 7275 + }, + { + "epoch": 1.295405982905983, + "grad_norm": 0.6254632472991943, + "learning_rate": 0.00015271810145357412, + "loss": 0.9746, + "step": 7276 + }, + { + "epoch": 1.2955840455840457, + "grad_norm": 0.6772669553756714, + "learning_rate": 0.00015270620654305494, + "loss": 1.1714, + "step": 7277 + }, + { + "epoch": 1.2957621082621082, + "grad_norm": 0.605576753616333, + "learning_rate": 0.00015269431059986713, + "loss": 0.7735, + "step": 7278 + }, + { + "epoch": 1.295940170940171, + "grad_norm": 0.7144771814346313, + "learning_rate": 0.00015268241362424378, + "loss": 0.9757, + "step": 7279 + }, + { + "epoch": 1.2961182336182335, + "grad_norm": 0.5275486707687378, + "learning_rate": 0.00015267051561641798, + "loss": 0.5669, + "step": 7280 + }, + { + "epoch": 1.2962962962962963, + "grad_norm": 0.6619452238082886, + "learning_rate": 0.00015265861657662284, + "loss": 0.9511, + "step": 7281 + }, + { + "epoch": 1.296474358974359, + "grad_norm": 0.6788223385810852, + "learning_rate": 0.00015264671650509147, + "loss": 1.2649, + "step": 7282 + }, + { + "epoch": 1.2966524216524218, + "grad_norm": 0.6198732852935791, + "learning_rate": 0.00015263481540205706, + "loss": 1.0659, + "step": 7283 + }, + { + "epoch": 1.2968304843304843, + "grad_norm": 0.6038815975189209, + "learning_rate": 0.0001526229132677528, + "loss": 1.0655, + "step": 7284 + }, + { + "epoch": 1.297008547008547, + "grad_norm": 0.7616196870803833, + "learning_rate": 0.00015261101010241186, + "loss": 1.131, + "step": 7285 + }, + { + "epoch": 1.2971866096866096, + "grad_norm": 0.7002527713775635, + "learning_rate": 0.00015259910590626746, + "loss": 1.1375, + "step": 7286 + }, + { + "epoch": 1.2973646723646723, + "grad_norm": 0.6067437529563904, + "learning_rate": 0.00015258720067955284, + "loss": 0.9306, + "step": 7287 + }, + { + "epoch": 1.297542735042735, + "grad_norm": 0.653232216835022, + "learning_rate": 0.00015257529442250128, + "loss": 1.107, + "step": 7288 + }, + { + "epoch": 1.2977207977207978, + "grad_norm": 0.6969175934791565, + "learning_rate": 0.00015256338713534603, + "loss": 0.8365, + "step": 7289 + }, + { + "epoch": 1.2978988603988604, + "grad_norm": 0.6176731586456299, + "learning_rate": 0.00015255147881832043, + "loss": 0.9707, + "step": 7290 + }, + { + "epoch": 1.2980769230769231, + "grad_norm": 0.6543741822242737, + "learning_rate": 0.00015253956947165772, + "loss": 0.7714, + "step": 7291 + }, + { + "epoch": 1.2982549857549857, + "grad_norm": 0.5224920511245728, + "learning_rate": 0.00015252765909559135, + "loss": 0.7469, + "step": 7292 + }, + { + "epoch": 1.2984330484330484, + "grad_norm": 0.638708770275116, + "learning_rate": 0.00015251574769035455, + "loss": 1.0965, + "step": 7293 + }, + { + "epoch": 1.2986111111111112, + "grad_norm": 0.6742943525314331, + "learning_rate": 0.0001525038352561808, + "loss": 1.1286, + "step": 7294 + }, + { + "epoch": 1.298789173789174, + "grad_norm": 0.6027839183807373, + "learning_rate": 0.00015249192179330346, + "loss": 0.8824, + "step": 7295 + }, + { + "epoch": 1.2989672364672364, + "grad_norm": 0.7462167143821716, + "learning_rate": 0.00015248000730195597, + "loss": 0.94, + "step": 7296 + }, + { + "epoch": 1.2991452991452992, + "grad_norm": 0.6972534656524658, + "learning_rate": 0.00015246809178237172, + "loss": 1.0664, + "step": 7297 + }, + { + "epoch": 1.2993233618233617, + "grad_norm": 0.569949209690094, + "learning_rate": 0.0001524561752347842, + "loss": 0.691, + "step": 7298 + }, + { + "epoch": 1.2995014245014245, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00015244425765942695, + "loss": 1.083, + "step": 7299 + }, + { + "epoch": 1.2996794871794872, + "grad_norm": 0.6927483677864075, + "learning_rate": 0.00015243233905653337, + "loss": 1.0068, + "step": 7300 + }, + { + "epoch": 1.29985754985755, + "grad_norm": 0.752824604511261, + "learning_rate": 0.00015242041942633704, + "loss": 0.9946, + "step": 7301 + }, + { + "epoch": 1.3000356125356125, + "grad_norm": 0.6532080173492432, + "learning_rate": 0.0001524084987690715, + "loss": 1.2326, + "step": 7302 + }, + { + "epoch": 1.3002136752136753, + "grad_norm": 0.7954180836677551, + "learning_rate": 0.0001523965770849703, + "loss": 1.1105, + "step": 7303 + }, + { + "epoch": 1.3003917378917378, + "grad_norm": 0.5971781015396118, + "learning_rate": 0.000152384654374267, + "loss": 1.0984, + "step": 7304 + }, + { + "epoch": 1.3005698005698005, + "grad_norm": 0.7778682112693787, + "learning_rate": 0.0001523727306371952, + "loss": 1.0795, + "step": 7305 + }, + { + "epoch": 1.3007478632478633, + "grad_norm": 0.6712004542350769, + "learning_rate": 0.00015236080587398856, + "loss": 1.0814, + "step": 7306 + }, + { + "epoch": 1.300925925925926, + "grad_norm": 0.581048846244812, + "learning_rate": 0.00015234888008488066, + "loss": 0.9868, + "step": 7307 + }, + { + "epoch": 1.3011039886039886, + "grad_norm": 0.697695791721344, + "learning_rate": 0.00015233695327010523, + "loss": 1.1045, + "step": 7308 + }, + { + "epoch": 1.3012820512820513, + "grad_norm": 0.6858421564102173, + "learning_rate": 0.00015232502542989593, + "loss": 1.0769, + "step": 7309 + }, + { + "epoch": 1.3014601139601139, + "grad_norm": 0.6312826871871948, + "learning_rate": 0.00015231309656448642, + "loss": 0.9523, + "step": 7310 + }, + { + "epoch": 1.3016381766381766, + "grad_norm": 0.9243300557136536, + "learning_rate": 0.0001523011666741105, + "loss": 0.947, + "step": 7311 + }, + { + "epoch": 1.3018162393162394, + "grad_norm": 0.6808217763900757, + "learning_rate": 0.00015228923575900184, + "loss": 0.8631, + "step": 7312 + }, + { + "epoch": 1.301994301994302, + "grad_norm": 0.6713891625404358, + "learning_rate": 0.00015227730381939424, + "loss": 0.9157, + "step": 7313 + }, + { + "epoch": 1.3021723646723646, + "grad_norm": 0.6802582740783691, + "learning_rate": 0.00015226537085552146, + "loss": 1.041, + "step": 7314 + }, + { + "epoch": 1.3023504273504274, + "grad_norm": 0.6543951034545898, + "learning_rate": 0.0001522534368676173, + "loss": 0.8709, + "step": 7315 + }, + { + "epoch": 1.30252849002849, + "grad_norm": 0.6290678381919861, + "learning_rate": 0.0001522415018559156, + "loss": 1.0568, + "step": 7316 + }, + { + "epoch": 1.3027065527065527, + "grad_norm": 0.6590015292167664, + "learning_rate": 0.0001522295658206502, + "loss": 0.9919, + "step": 7317 + }, + { + "epoch": 1.3028846153846154, + "grad_norm": 0.6374103426933289, + "learning_rate": 0.00015221762876205494, + "loss": 0.878, + "step": 7318 + }, + { + "epoch": 1.3030626780626782, + "grad_norm": 0.7247048616409302, + "learning_rate": 0.00015220569068036372, + "loss": 1.061, + "step": 7319 + }, + { + "epoch": 1.3032407407407407, + "grad_norm": 0.6450991630554199, + "learning_rate": 0.00015219375157581047, + "loss": 0.9389, + "step": 7320 + }, + { + "epoch": 1.3034188034188035, + "grad_norm": 0.8039840459823608, + "learning_rate": 0.00015218181144862903, + "loss": 1.0692, + "step": 7321 + }, + { + "epoch": 1.303596866096866, + "grad_norm": 0.6539456248283386, + "learning_rate": 0.00015216987029905346, + "loss": 1.0478, + "step": 7322 + }, + { + "epoch": 1.3037749287749287, + "grad_norm": 0.60880047082901, + "learning_rate": 0.00015215792812731758, + "loss": 0.8412, + "step": 7323 + }, + { + "epoch": 1.3039529914529915, + "grad_norm": 0.6757258176803589, + "learning_rate": 0.0001521459849336555, + "loss": 0.896, + "step": 7324 + }, + { + "epoch": 1.3041310541310542, + "grad_norm": 0.6735622882843018, + "learning_rate": 0.00015213404071830116, + "loss": 1.1078, + "step": 7325 + }, + { + "epoch": 1.3043091168091168, + "grad_norm": 0.7321233749389648, + "learning_rate": 0.00015212209548148858, + "loss": 1.1021, + "step": 7326 + }, + { + "epoch": 1.3044871794871795, + "grad_norm": 0.6678910851478577, + "learning_rate": 0.00015211014922345182, + "loss": 1.0043, + "step": 7327 + }, + { + "epoch": 1.304665242165242, + "grad_norm": 0.6876940727233887, + "learning_rate": 0.0001520982019444249, + "loss": 1.0376, + "step": 7328 + }, + { + "epoch": 1.3048433048433048, + "grad_norm": 0.6171853542327881, + "learning_rate": 0.00015208625364464195, + "loss": 0.839, + "step": 7329 + }, + { + "epoch": 1.3050213675213675, + "grad_norm": 0.6449569463729858, + "learning_rate": 0.0001520743043243371, + "loss": 1.0908, + "step": 7330 + }, + { + "epoch": 1.3051994301994303, + "grad_norm": 0.6894628405570984, + "learning_rate": 0.00015206235398374443, + "loss": 1.0263, + "step": 7331 + }, + { + "epoch": 1.3053774928774928, + "grad_norm": 0.5853552222251892, + "learning_rate": 0.00015205040262309804, + "loss": 0.8342, + "step": 7332 + }, + { + "epoch": 1.3055555555555556, + "grad_norm": 0.5934799313545227, + "learning_rate": 0.00015203845024263214, + "loss": 0.9464, + "step": 7333 + }, + { + "epoch": 1.305733618233618, + "grad_norm": 0.668927788734436, + "learning_rate": 0.00015202649684258095, + "loss": 0.9018, + "step": 7334 + }, + { + "epoch": 1.3059116809116809, + "grad_norm": 0.676810085773468, + "learning_rate": 0.0001520145424231786, + "loss": 0.9284, + "step": 7335 + }, + { + "epoch": 1.3060897435897436, + "grad_norm": 0.6223878264427185, + "learning_rate": 0.00015200258698465935, + "loss": 1.0779, + "step": 7336 + }, + { + "epoch": 1.3062678062678064, + "grad_norm": 0.6092363595962524, + "learning_rate": 0.00015199063052725745, + "loss": 0.8602, + "step": 7337 + }, + { + "epoch": 1.306445868945869, + "grad_norm": 0.7668731212615967, + "learning_rate": 0.00015197867305120712, + "loss": 1.0756, + "step": 7338 + }, + { + "epoch": 1.3066239316239316, + "grad_norm": 0.6485331654548645, + "learning_rate": 0.00015196671455674268, + "loss": 1.0193, + "step": 7339 + }, + { + "epoch": 1.3068019943019942, + "grad_norm": 0.5661036372184753, + "learning_rate": 0.0001519547550440984, + "loss": 0.8321, + "step": 7340 + }, + { + "epoch": 1.306980056980057, + "grad_norm": 0.6270507574081421, + "learning_rate": 0.00015194279451350866, + "loss": 0.6403, + "step": 7341 + }, + { + "epoch": 1.3071581196581197, + "grad_norm": 0.7283764481544495, + "learning_rate": 0.00015193083296520773, + "loss": 1.0401, + "step": 7342 + }, + { + "epoch": 1.3073361823361824, + "grad_norm": 0.658835232257843, + "learning_rate": 0.00015191887039943, + "loss": 1.0172, + "step": 7343 + }, + { + "epoch": 1.307514245014245, + "grad_norm": 0.6288984417915344, + "learning_rate": 0.00015190690681640988, + "loss": 0.8649, + "step": 7344 + }, + { + "epoch": 1.3076923076923077, + "grad_norm": 0.666442334651947, + "learning_rate": 0.00015189494221638176, + "loss": 1.0757, + "step": 7345 + }, + { + "epoch": 1.3078703703703702, + "grad_norm": 0.6116433143615723, + "learning_rate": 0.00015188297659958003, + "loss": 0.9244, + "step": 7346 + }, + { + "epoch": 1.308048433048433, + "grad_norm": 0.6378964185714722, + "learning_rate": 0.0001518710099662392, + "loss": 0.9629, + "step": 7347 + }, + { + "epoch": 1.3082264957264957, + "grad_norm": 0.6258945465087891, + "learning_rate": 0.00015185904231659357, + "loss": 0.8524, + "step": 7348 + }, + { + "epoch": 1.3084045584045585, + "grad_norm": 0.6498504877090454, + "learning_rate": 0.0001518470736508778, + "loss": 0.9685, + "step": 7349 + }, + { + "epoch": 1.308582621082621, + "grad_norm": 0.6928247809410095, + "learning_rate": 0.00015183510396932635, + "loss": 0.9054, + "step": 7350 + }, + { + "epoch": 1.3087606837606838, + "grad_norm": 0.6350936889648438, + "learning_rate": 0.0001518231332721737, + "loss": 1.0039, + "step": 7351 + }, + { + "epoch": 1.3089387464387463, + "grad_norm": 0.6652286648750305, + "learning_rate": 0.00015181116155965437, + "loss": 0.8946, + "step": 7352 + }, + { + "epoch": 1.309116809116809, + "grad_norm": 0.6554864048957825, + "learning_rate": 0.000151799188832003, + "loss": 0.9518, + "step": 7353 + }, + { + "epoch": 1.3092948717948718, + "grad_norm": 0.7523114085197449, + "learning_rate": 0.0001517872150894541, + "loss": 0.9462, + "step": 7354 + }, + { + "epoch": 1.3094729344729346, + "grad_norm": 0.7113336324691772, + "learning_rate": 0.0001517752403322423, + "loss": 1.2347, + "step": 7355 + }, + { + "epoch": 1.309650997150997, + "grad_norm": 0.6461622714996338, + "learning_rate": 0.00015176326456060223, + "loss": 0.8891, + "step": 7356 + }, + { + "epoch": 1.3098290598290598, + "grad_norm": 0.7429143190383911, + "learning_rate": 0.00015175128777476852, + "loss": 1.1944, + "step": 7357 + }, + { + "epoch": 1.3100071225071226, + "grad_norm": 0.6816306114196777, + "learning_rate": 0.00015173930997497585, + "loss": 1.1445, + "step": 7358 + }, + { + "epoch": 1.3101851851851851, + "grad_norm": 0.6644450426101685, + "learning_rate": 0.00015172733116145884, + "loss": 0.9808, + "step": 7359 + }, + { + "epoch": 1.3103632478632479, + "grad_norm": 0.6921063661575317, + "learning_rate": 0.00015171535133445225, + "loss": 1.0162, + "step": 7360 + }, + { + "epoch": 1.3105413105413106, + "grad_norm": 0.6386187672615051, + "learning_rate": 0.00015170337049419082, + "loss": 0.9951, + "step": 7361 + }, + { + "epoch": 1.3107193732193732, + "grad_norm": 0.6505418419837952, + "learning_rate": 0.0001516913886409092, + "loss": 0.8872, + "step": 7362 + }, + { + "epoch": 1.310897435897436, + "grad_norm": 0.6415576934814453, + "learning_rate": 0.00015167940577484222, + "loss": 1.056, + "step": 7363 + }, + { + "epoch": 1.3110754985754987, + "grad_norm": 0.6691195964813232, + "learning_rate": 0.00015166742189622458, + "loss": 1.0561, + "step": 7364 + }, + { + "epoch": 1.3112535612535612, + "grad_norm": 0.6376257538795471, + "learning_rate": 0.00015165543700529122, + "loss": 0.8499, + "step": 7365 + }, + { + "epoch": 1.311431623931624, + "grad_norm": 0.6270790696144104, + "learning_rate": 0.00015164345110227684, + "loss": 1.0244, + "step": 7366 + }, + { + "epoch": 1.3116096866096867, + "grad_norm": 0.7120122313499451, + "learning_rate": 0.0001516314641874163, + "loss": 1.0476, + "step": 7367 + }, + { + "epoch": 1.3117877492877492, + "grad_norm": 0.6152660250663757, + "learning_rate": 0.0001516194762609445, + "loss": 0.897, + "step": 7368 + }, + { + "epoch": 1.311965811965812, + "grad_norm": 0.7578088045120239, + "learning_rate": 0.00015160748732309626, + "loss": 1.1609, + "step": 7369 + }, + { + "epoch": 1.3121438746438747, + "grad_norm": 0.6594924330711365, + "learning_rate": 0.00015159549737410656, + "loss": 1.1706, + "step": 7370 + }, + { + "epoch": 1.3123219373219372, + "grad_norm": 0.6559173464775085, + "learning_rate": 0.00015158350641421024, + "loss": 0.9452, + "step": 7371 + }, + { + "epoch": 1.3125, + "grad_norm": 0.6667516231536865, + "learning_rate": 0.00015157151444364226, + "loss": 0.8153, + "step": 7372 + }, + { + "epoch": 1.3126780626780628, + "grad_norm": 0.7054803371429443, + "learning_rate": 0.00015155952146263761, + "loss": 0.9887, + "step": 7373 + }, + { + "epoch": 1.3128561253561253, + "grad_norm": 0.7035902142524719, + "learning_rate": 0.00015154752747143123, + "loss": 1.1832, + "step": 7374 + }, + { + "epoch": 1.313034188034188, + "grad_norm": 0.6297488212585449, + "learning_rate": 0.00015153553247025813, + "loss": 0.9602, + "step": 7375 + }, + { + "epoch": 1.3132122507122508, + "grad_norm": 0.6851378083229065, + "learning_rate": 0.00015152353645935335, + "loss": 1.0743, + "step": 7376 + }, + { + "epoch": 1.3133903133903133, + "grad_norm": 0.6215537786483765, + "learning_rate": 0.00015151153943895187, + "loss": 0.9484, + "step": 7377 + }, + { + "epoch": 1.313568376068376, + "grad_norm": 0.6848666071891785, + "learning_rate": 0.0001514995414092888, + "loss": 1.0978, + "step": 7378 + }, + { + "epoch": 1.3137464387464388, + "grad_norm": 0.7527492642402649, + "learning_rate": 0.00015148754237059918, + "loss": 1.083, + "step": 7379 + }, + { + "epoch": 1.3139245014245013, + "grad_norm": 0.6264588236808777, + "learning_rate": 0.00015147554232311814, + "loss": 0.9995, + "step": 7380 + }, + { + "epoch": 1.314102564102564, + "grad_norm": 0.6666619181632996, + "learning_rate": 0.00015146354126708075, + "loss": 1.0156, + "step": 7381 + }, + { + "epoch": 1.3142806267806268, + "grad_norm": 0.6626597046852112, + "learning_rate": 0.00015145153920272222, + "loss": 1.0047, + "step": 7382 + }, + { + "epoch": 1.3144586894586894, + "grad_norm": 0.5975428223609924, + "learning_rate": 0.0001514395361302776, + "loss": 0.806, + "step": 7383 + }, + { + "epoch": 1.3146367521367521, + "grad_norm": 0.6509957909584045, + "learning_rate": 0.00015142753204998218, + "loss": 0.8871, + "step": 7384 + }, + { + "epoch": 1.3148148148148149, + "grad_norm": 0.6672926545143127, + "learning_rate": 0.00015141552696207108, + "loss": 0.9616, + "step": 7385 + }, + { + "epoch": 1.3149928774928774, + "grad_norm": 0.6965435147285461, + "learning_rate": 0.00015140352086677954, + "loss": 1.124, + "step": 7386 + }, + { + "epoch": 1.3151709401709402, + "grad_norm": 0.6559258103370667, + "learning_rate": 0.00015139151376434277, + "loss": 1.0271, + "step": 7387 + }, + { + "epoch": 1.315349002849003, + "grad_norm": 0.7613587379455566, + "learning_rate": 0.00015137950565499608, + "loss": 1.0349, + "step": 7388 + }, + { + "epoch": 1.3155270655270654, + "grad_norm": 0.7001944780349731, + "learning_rate": 0.0001513674965389747, + "loss": 0.8551, + "step": 7389 + }, + { + "epoch": 1.3157051282051282, + "grad_norm": 0.6087043285369873, + "learning_rate": 0.0001513554864165139, + "loss": 0.7118, + "step": 7390 + }, + { + "epoch": 1.315883190883191, + "grad_norm": 0.71526700258255, + "learning_rate": 0.00015134347528784908, + "loss": 1.0478, + "step": 7391 + }, + { + "epoch": 1.3160612535612537, + "grad_norm": 0.6182073950767517, + "learning_rate": 0.00015133146315321548, + "loss": 0.9474, + "step": 7392 + }, + { + "epoch": 1.3162393162393162, + "grad_norm": 0.7771387696266174, + "learning_rate": 0.0001513194500128485, + "loss": 1.0544, + "step": 7393 + }, + { + "epoch": 1.316417378917379, + "grad_norm": 0.7108260989189148, + "learning_rate": 0.00015130743586698353, + "loss": 0.8813, + "step": 7394 + }, + { + "epoch": 1.3165954415954415, + "grad_norm": 0.7057309150695801, + "learning_rate": 0.0001512954207158559, + "loss": 0.899, + "step": 7395 + }, + { + "epoch": 1.3167735042735043, + "grad_norm": 0.6139237880706787, + "learning_rate": 0.00015128340455970106, + "loss": 0.8885, + "step": 7396 + }, + { + "epoch": 1.316951566951567, + "grad_norm": 0.7166598439216614, + "learning_rate": 0.00015127138739875443, + "loss": 0.9792, + "step": 7397 + }, + { + "epoch": 1.3171296296296298, + "grad_norm": 0.6916186809539795, + "learning_rate": 0.00015125936923325153, + "loss": 0.8871, + "step": 7398 + }, + { + "epoch": 1.3173076923076923, + "grad_norm": 0.7189087271690369, + "learning_rate": 0.0001512473500634277, + "loss": 0.8302, + "step": 7399 + }, + { + "epoch": 1.317485754985755, + "grad_norm": 0.5739200115203857, + "learning_rate": 0.00015123532988951853, + "loss": 0.9137, + "step": 7400 + }, + { + "epoch": 1.3176638176638176, + "grad_norm": 0.7661057114601135, + "learning_rate": 0.00015122330871175952, + "loss": 1.1255, + "step": 7401 + }, + { + "epoch": 1.3178418803418803, + "grad_norm": 0.6487592458724976, + "learning_rate": 0.00015121128653038617, + "loss": 1.0519, + "step": 7402 + }, + { + "epoch": 1.318019943019943, + "grad_norm": 0.693134605884552, + "learning_rate": 0.00015119926334563406, + "loss": 0.9585, + "step": 7403 + }, + { + "epoch": 1.3181980056980058, + "grad_norm": 0.5895997285842896, + "learning_rate": 0.0001511872391577387, + "loss": 0.8033, + "step": 7404 + }, + { + "epoch": 1.3183760683760684, + "grad_norm": 0.654876172542572, + "learning_rate": 0.00015117521396693575, + "loss": 1.0082, + "step": 7405 + }, + { + "epoch": 1.318554131054131, + "grad_norm": 0.5877239108085632, + "learning_rate": 0.0001511631877734608, + "loss": 1.0147, + "step": 7406 + }, + { + "epoch": 1.3187321937321936, + "grad_norm": 0.6109837889671326, + "learning_rate": 0.00015115116057754944, + "loss": 0.7498, + "step": 7407 + }, + { + "epoch": 1.3189102564102564, + "grad_norm": 0.643856942653656, + "learning_rate": 0.00015113913237943736, + "loss": 1.0417, + "step": 7408 + }, + { + "epoch": 1.3190883190883191, + "grad_norm": 0.654077410697937, + "learning_rate": 0.00015112710317936022, + "loss": 1.1809, + "step": 7409 + }, + { + "epoch": 1.319266381766382, + "grad_norm": 0.6785375475883484, + "learning_rate": 0.00015111507297755367, + "loss": 0.9447, + "step": 7410 + }, + { + "epoch": 1.3194444444444444, + "grad_norm": 0.6513382196426392, + "learning_rate": 0.00015110304177425347, + "loss": 0.8286, + "step": 7411 + }, + { + "epoch": 1.3196225071225072, + "grad_norm": 0.6536405682563782, + "learning_rate": 0.00015109100956969533, + "loss": 1.1959, + "step": 7412 + }, + { + "epoch": 1.3198005698005697, + "grad_norm": 0.6633172035217285, + "learning_rate": 0.00015107897636411498, + "loss": 0.8839, + "step": 7413 + }, + { + "epoch": 1.3199786324786325, + "grad_norm": 0.5773791670799255, + "learning_rate": 0.00015106694215774821, + "loss": 0.9785, + "step": 7414 + }, + { + "epoch": 1.3201566951566952, + "grad_norm": 0.7005468010902405, + "learning_rate": 0.00015105490695083078, + "loss": 1.0752, + "step": 7415 + }, + { + "epoch": 1.320334757834758, + "grad_norm": 0.6509538888931274, + "learning_rate": 0.0001510428707435985, + "loss": 0.9886, + "step": 7416 + }, + { + "epoch": 1.3205128205128205, + "grad_norm": 0.6607788801193237, + "learning_rate": 0.0001510308335362872, + "loss": 0.9756, + "step": 7417 + }, + { + "epoch": 1.3206908831908832, + "grad_norm": 0.5977858304977417, + "learning_rate": 0.00015101879532913274, + "loss": 1.0574, + "step": 7418 + }, + { + "epoch": 1.3208689458689458, + "grad_norm": 0.6478607058525085, + "learning_rate": 0.00015100675612237096, + "loss": 1.0076, + "step": 7419 + }, + { + "epoch": 1.3210470085470085, + "grad_norm": 0.6386681199073792, + "learning_rate": 0.00015099471591623775, + "loss": 0.9639, + "step": 7420 + }, + { + "epoch": 1.3212250712250713, + "grad_norm": 0.6348143815994263, + "learning_rate": 0.000150982674710969, + "loss": 1.0226, + "step": 7421 + }, + { + "epoch": 1.321403133903134, + "grad_norm": 0.6737388372421265, + "learning_rate": 0.00015097063250680068, + "loss": 0.9985, + "step": 7422 + }, + { + "epoch": 1.3215811965811965, + "grad_norm": 0.7302656769752502, + "learning_rate": 0.00015095858930396866, + "loss": 0.9969, + "step": 7423 + }, + { + "epoch": 1.3217592592592593, + "grad_norm": 0.7062691450119019, + "learning_rate": 0.00015094654510270898, + "loss": 0.9137, + "step": 7424 + }, + { + "epoch": 1.3219373219373218, + "grad_norm": 0.6289888620376587, + "learning_rate": 0.00015093449990325754, + "loss": 0.9231, + "step": 7425 + }, + { + "epoch": 1.3221153846153846, + "grad_norm": 0.643284261226654, + "learning_rate": 0.0001509224537058504, + "loss": 0.8981, + "step": 7426 + }, + { + "epoch": 1.3222934472934473, + "grad_norm": 0.7019244432449341, + "learning_rate": 0.00015091040651072355, + "loss": 0.9994, + "step": 7427 + }, + { + "epoch": 1.32247150997151, + "grad_norm": 0.5982088446617126, + "learning_rate": 0.0001508983583181131, + "loss": 0.9365, + "step": 7428 + }, + { + "epoch": 1.3226495726495726, + "grad_norm": 0.6086063385009766, + "learning_rate": 0.00015088630912825498, + "loss": 0.8621, + "step": 7429 + }, + { + "epoch": 1.3228276353276354, + "grad_norm": 0.6829213500022888, + "learning_rate": 0.00015087425894138535, + "loss": 1.1959, + "step": 7430 + }, + { + "epoch": 1.323005698005698, + "grad_norm": 0.6538017392158508, + "learning_rate": 0.00015086220775774033, + "loss": 0.9412, + "step": 7431 + }, + { + "epoch": 1.3231837606837606, + "grad_norm": 0.6334070563316345, + "learning_rate": 0.00015085015557755597, + "loss": 0.9044, + "step": 7432 + }, + { + "epoch": 1.3233618233618234, + "grad_norm": 0.6514624357223511, + "learning_rate": 0.00015083810240106845, + "loss": 0.8859, + "step": 7433 + }, + { + "epoch": 1.3235398860398861, + "grad_norm": 0.7130434513092041, + "learning_rate": 0.00015082604822851397, + "loss": 1.2845, + "step": 7434 + }, + { + "epoch": 1.3237179487179487, + "grad_norm": 0.609419584274292, + "learning_rate": 0.00015081399306012862, + "loss": 1.0725, + "step": 7435 + }, + { + "epoch": 1.3238960113960114, + "grad_norm": 0.586807370185852, + "learning_rate": 0.0001508019368961486, + "loss": 0.9032, + "step": 7436 + }, + { + "epoch": 1.324074074074074, + "grad_norm": 0.6937291026115417, + "learning_rate": 0.0001507898797368102, + "loss": 0.7975, + "step": 7437 + }, + { + "epoch": 1.3242521367521367, + "grad_norm": 0.6804966330528259, + "learning_rate": 0.00015077782158234962, + "loss": 1.1018, + "step": 7438 + }, + { + "epoch": 1.3244301994301995, + "grad_norm": 0.6110677123069763, + "learning_rate": 0.0001507657624330031, + "loss": 0.7988, + "step": 7439 + }, + { + "epoch": 1.3246082621082622, + "grad_norm": 0.6340961456298828, + "learning_rate": 0.0001507537022890069, + "loss": 0.844, + "step": 7440 + }, + { + "epoch": 1.3247863247863247, + "grad_norm": 0.7291021943092346, + "learning_rate": 0.00015074164115059735, + "loss": 0.9867, + "step": 7441 + }, + { + "epoch": 1.3249643874643875, + "grad_norm": 0.6818505525588989, + "learning_rate": 0.00015072957901801076, + "loss": 1.1541, + "step": 7442 + }, + { + "epoch": 1.32514245014245, + "grad_norm": 0.6174707412719727, + "learning_rate": 0.00015071751589148345, + "loss": 1.1679, + "step": 7443 + }, + { + "epoch": 1.3253205128205128, + "grad_norm": 0.6481367945671082, + "learning_rate": 0.00015070545177125176, + "loss": 1.0955, + "step": 7444 + }, + { + "epoch": 1.3254985754985755, + "grad_norm": 0.6752339005470276, + "learning_rate": 0.00015069338665755203, + "loss": 0.8651, + "step": 7445 + }, + { + "epoch": 1.3256766381766383, + "grad_norm": 0.6608055830001831, + "learning_rate": 0.00015068132055062077, + "loss": 0.9553, + "step": 7446 + }, + { + "epoch": 1.3258547008547008, + "grad_norm": 0.5933246612548828, + "learning_rate": 0.00015066925345069425, + "loss": 0.8584, + "step": 7447 + }, + { + "epoch": 1.3260327635327636, + "grad_norm": 0.6301844716072083, + "learning_rate": 0.000150657185358009, + "loss": 0.8583, + "step": 7448 + }, + { + "epoch": 1.326210826210826, + "grad_norm": 0.7359434962272644, + "learning_rate": 0.00015064511627280145, + "loss": 1.0905, + "step": 7449 + }, + { + "epoch": 1.3263888888888888, + "grad_norm": 0.6334579586982727, + "learning_rate": 0.00015063304619530806, + "loss": 0.9814, + "step": 7450 + }, + { + "epoch": 1.3265669515669516, + "grad_norm": 0.6974197626113892, + "learning_rate": 0.00015062097512576528, + "loss": 0.9302, + "step": 7451 + }, + { + "epoch": 1.3267450142450143, + "grad_norm": 0.6895849704742432, + "learning_rate": 0.00015060890306440965, + "loss": 1.0175, + "step": 7452 + }, + { + "epoch": 1.3269230769230769, + "grad_norm": 0.5938003659248352, + "learning_rate": 0.00015059683001147767, + "loss": 0.8084, + "step": 7453 + }, + { + "epoch": 1.3271011396011396, + "grad_norm": 0.6821470856666565, + "learning_rate": 0.00015058475596720596, + "loss": 0.9897, + "step": 7454 + }, + { + "epoch": 1.3272792022792022, + "grad_norm": 0.5507164001464844, + "learning_rate": 0.00015057268093183104, + "loss": 0.7012, + "step": 7455 + }, + { + "epoch": 1.327457264957265, + "grad_norm": 0.6216199398040771, + "learning_rate": 0.00015056060490558945, + "loss": 1.0281, + "step": 7456 + }, + { + "epoch": 1.3276353276353277, + "grad_norm": 0.6674157977104187, + "learning_rate": 0.00015054852788871787, + "loss": 0.8776, + "step": 7457 + }, + { + "epoch": 1.3278133903133904, + "grad_norm": 0.666963517665863, + "learning_rate": 0.0001505364498814529, + "loss": 1.0742, + "step": 7458 + }, + { + "epoch": 1.327991452991453, + "grad_norm": 0.6205331683158875, + "learning_rate": 0.00015052437088403114, + "loss": 1.1109, + "step": 7459 + }, + { + "epoch": 1.3281695156695157, + "grad_norm": 0.6402750611305237, + "learning_rate": 0.00015051229089668933, + "loss": 1.0648, + "step": 7460 + }, + { + "epoch": 1.3283475783475782, + "grad_norm": 0.7445703744888306, + "learning_rate": 0.00015050020991966406, + "loss": 0.8989, + "step": 7461 + }, + { + "epoch": 1.328525641025641, + "grad_norm": 0.8131299614906311, + "learning_rate": 0.00015048812795319212, + "loss": 0.9552, + "step": 7462 + }, + { + "epoch": 1.3287037037037037, + "grad_norm": 0.7007313966751099, + "learning_rate": 0.00015047604499751017, + "loss": 0.9899, + "step": 7463 + }, + { + "epoch": 1.3288817663817665, + "grad_norm": 0.60536789894104, + "learning_rate": 0.000150463961052855, + "loss": 0.7694, + "step": 7464 + }, + { + "epoch": 1.329059829059829, + "grad_norm": 0.6910434365272522, + "learning_rate": 0.00015045187611946331, + "loss": 0.9575, + "step": 7465 + }, + { + "epoch": 1.3292378917378918, + "grad_norm": 0.7693352103233337, + "learning_rate": 0.00015043979019757194, + "loss": 1.1987, + "step": 7466 + }, + { + "epoch": 1.3294159544159543, + "grad_norm": 0.6675218939781189, + "learning_rate": 0.00015042770328741763, + "loss": 1.0099, + "step": 7467 + }, + { + "epoch": 1.329594017094017, + "grad_norm": 0.8040883541107178, + "learning_rate": 0.00015041561538923722, + "loss": 0.9493, + "step": 7468 + }, + { + "epoch": 1.3297720797720798, + "grad_norm": 0.6765826344490051, + "learning_rate": 0.00015040352650326762, + "loss": 1.1035, + "step": 7469 + }, + { + "epoch": 1.3299501424501425, + "grad_norm": 0.7099924087524414, + "learning_rate": 0.0001503914366297456, + "loss": 0.9198, + "step": 7470 + }, + { + "epoch": 1.330128205128205, + "grad_norm": 0.6673682928085327, + "learning_rate": 0.00015037934576890804, + "loss": 1.0234, + "step": 7471 + }, + { + "epoch": 1.3303062678062678, + "grad_norm": 0.7022300958633423, + "learning_rate": 0.00015036725392099184, + "loss": 1.3875, + "step": 7472 + }, + { + "epoch": 1.3304843304843303, + "grad_norm": 0.6997060179710388, + "learning_rate": 0.00015035516108623394, + "loss": 0.8114, + "step": 7473 + }, + { + "epoch": 1.330662393162393, + "grad_norm": 0.6262350678443909, + "learning_rate": 0.00015034306726487127, + "loss": 1.128, + "step": 7474 + }, + { + "epoch": 1.3308404558404558, + "grad_norm": 0.6330382227897644, + "learning_rate": 0.00015033097245714078, + "loss": 0.9032, + "step": 7475 + }, + { + "epoch": 1.3310185185185186, + "grad_norm": 0.6527551412582397, + "learning_rate": 0.00015031887666327944, + "loss": 0.9311, + "step": 7476 + }, + { + "epoch": 1.3311965811965811, + "grad_norm": 0.6754798889160156, + "learning_rate": 0.00015030677988352422, + "loss": 1.0626, + "step": 7477 + }, + { + "epoch": 1.3313746438746439, + "grad_norm": 0.6397945284843445, + "learning_rate": 0.00015029468211811216, + "loss": 0.9222, + "step": 7478 + }, + { + "epoch": 1.3315527065527066, + "grad_norm": 0.8163481950759888, + "learning_rate": 0.0001502825833672803, + "loss": 1.1827, + "step": 7479 + }, + { + "epoch": 1.3317307692307692, + "grad_norm": 0.6645621657371521, + "learning_rate": 0.00015027048363126566, + "loss": 0.9744, + "step": 7480 + }, + { + "epoch": 1.331908831908832, + "grad_norm": 0.6943182349205017, + "learning_rate": 0.0001502583829103053, + "loss": 1.1597, + "step": 7481 + }, + { + "epoch": 1.3320868945868947, + "grad_norm": 0.6283710598945618, + "learning_rate": 0.00015024628120463636, + "loss": 0.9514, + "step": 7482 + }, + { + "epoch": 1.3322649572649572, + "grad_norm": 0.6159678101539612, + "learning_rate": 0.0001502341785144959, + "loss": 0.9752, + "step": 7483 + }, + { + "epoch": 1.33244301994302, + "grad_norm": 0.6259802579879761, + "learning_rate": 0.00015022207484012107, + "loss": 0.9356, + "step": 7484 + }, + { + "epoch": 1.3326210826210827, + "grad_norm": 0.7322365641593933, + "learning_rate": 0.00015020997018174904, + "loss": 1.2072, + "step": 7485 + }, + { + "epoch": 1.3327991452991452, + "grad_norm": 0.6323443651199341, + "learning_rate": 0.0001501978645396169, + "loss": 1.1661, + "step": 7486 + }, + { + "epoch": 1.332977207977208, + "grad_norm": 0.7811527848243713, + "learning_rate": 0.00015018575791396187, + "loss": 1.0304, + "step": 7487 + }, + { + "epoch": 1.3331552706552707, + "grad_norm": 0.7221232056617737, + "learning_rate": 0.0001501736503050212, + "loss": 0.8838, + "step": 7488 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.6980099081993103, + "learning_rate": 0.00015016154171303207, + "loss": 1.1841, + "step": 7489 + }, + { + "epoch": 1.333511396011396, + "grad_norm": 0.6802879571914673, + "learning_rate": 0.00015014943213823175, + "loss": 0.959, + "step": 7490 + }, + { + "epoch": 1.3336894586894588, + "grad_norm": 0.637698233127594, + "learning_rate": 0.00015013732158085746, + "loss": 1.0517, + "step": 7491 + }, + { + "epoch": 1.3338675213675213, + "grad_norm": 0.6386787295341492, + "learning_rate": 0.0001501252100411465, + "loss": 0.7125, + "step": 7492 + }, + { + "epoch": 1.334045584045584, + "grad_norm": 0.6287358403205872, + "learning_rate": 0.0001501130975193362, + "loss": 0.8913, + "step": 7493 + }, + { + "epoch": 1.3342236467236468, + "grad_norm": 0.6142337322235107, + "learning_rate": 0.00015010098401566386, + "loss": 0.8149, + "step": 7494 + }, + { + "epoch": 1.3344017094017093, + "grad_norm": 0.6369916200637817, + "learning_rate": 0.0001500888695303668, + "loss": 1.0186, + "step": 7495 + }, + { + "epoch": 1.334579772079772, + "grad_norm": 0.7526934146881104, + "learning_rate": 0.0001500767540636824, + "loss": 1.2421, + "step": 7496 + }, + { + "epoch": 1.3347578347578348, + "grad_norm": 0.7278095483779907, + "learning_rate": 0.00015006463761584802, + "loss": 0.9856, + "step": 7497 + }, + { + "epoch": 1.3349358974358974, + "grad_norm": 0.6165127158164978, + "learning_rate": 0.00015005252018710104, + "loss": 1.0041, + "step": 7498 + }, + { + "epoch": 1.33511396011396, + "grad_norm": 0.637856662273407, + "learning_rate": 0.00015004040177767896, + "loss": 0.9134, + "step": 7499 + }, + { + "epoch": 1.3352920227920229, + "grad_norm": 0.661227285861969, + "learning_rate": 0.00015002828238781912, + "loss": 1.0393, + "step": 7500 + }, + { + "epoch": 1.3354700854700854, + "grad_norm": 0.6061869859695435, + "learning_rate": 0.000150016162017759, + "loss": 0.8453, + "step": 7501 + }, + { + "epoch": 1.3356481481481481, + "grad_norm": 0.6938419938087463, + "learning_rate": 0.0001500040406677361, + "loss": 1.0338, + "step": 7502 + }, + { + "epoch": 1.335826210826211, + "grad_norm": 0.6672863960266113, + "learning_rate": 0.0001499919183379879, + "loss": 0.8765, + "step": 7503 + }, + { + "epoch": 1.3360042735042734, + "grad_norm": 0.6200515031814575, + "learning_rate": 0.00014997979502875193, + "loss": 0.8286, + "step": 7504 + }, + { + "epoch": 1.3361823361823362, + "grad_norm": 0.6287549138069153, + "learning_rate": 0.00014996767074026567, + "loss": 0.9761, + "step": 7505 + }, + { + "epoch": 1.336360398860399, + "grad_norm": 0.6036837100982666, + "learning_rate": 0.0001499555454727667, + "loss": 1.0506, + "step": 7506 + }, + { + "epoch": 1.3365384615384617, + "grad_norm": 0.6875260472297668, + "learning_rate": 0.0001499434192264926, + "loss": 1.001, + "step": 7507 + }, + { + "epoch": 1.3367165242165242, + "grad_norm": 0.6558469533920288, + "learning_rate": 0.00014993129200168096, + "loss": 0.6874, + "step": 7508 + }, + { + "epoch": 1.336894586894587, + "grad_norm": 0.604167103767395, + "learning_rate": 0.00014991916379856934, + "loss": 1.0173, + "step": 7509 + }, + { + "epoch": 1.3370726495726495, + "grad_norm": 0.5941442251205444, + "learning_rate": 0.00014990703461739544, + "loss": 0.8569, + "step": 7510 + }, + { + "epoch": 1.3372507122507122, + "grad_norm": 0.7645071148872375, + "learning_rate": 0.00014989490445839687, + "loss": 1.0172, + "step": 7511 + }, + { + "epoch": 1.337428774928775, + "grad_norm": 0.5491678714752197, + "learning_rate": 0.00014988277332181126, + "loss": 0.8018, + "step": 7512 + }, + { + "epoch": 1.3376068376068377, + "grad_norm": 0.583322286605835, + "learning_rate": 0.00014987064120787635, + "loss": 0.8704, + "step": 7513 + }, + { + "epoch": 1.3377849002849003, + "grad_norm": 0.7385724186897278, + "learning_rate": 0.00014985850811682984, + "loss": 1.1121, + "step": 7514 + }, + { + "epoch": 1.337962962962963, + "grad_norm": 0.6842585206031799, + "learning_rate": 0.00014984637404890941, + "loss": 0.914, + "step": 7515 + }, + { + "epoch": 1.3381410256410255, + "grad_norm": 0.6771186590194702, + "learning_rate": 0.00014983423900435285, + "loss": 1.0838, + "step": 7516 + }, + { + "epoch": 1.3383190883190883, + "grad_norm": 0.7562049031257629, + "learning_rate": 0.00014982210298339788, + "loss": 1.123, + "step": 7517 + }, + { + "epoch": 1.338497150997151, + "grad_norm": 0.7617804408073425, + "learning_rate": 0.0001498099659862823, + "loss": 0.9438, + "step": 7518 + }, + { + "epoch": 1.3386752136752138, + "grad_norm": 0.561958909034729, + "learning_rate": 0.00014979782801324392, + "loss": 0.8739, + "step": 7519 + }, + { + "epoch": 1.3388532763532763, + "grad_norm": 0.7726154923439026, + "learning_rate": 0.00014978568906452052, + "loss": 1.1306, + "step": 7520 + }, + { + "epoch": 1.339031339031339, + "grad_norm": 0.6658660173416138, + "learning_rate": 0.00014977354914035002, + "loss": 1.0214, + "step": 7521 + }, + { + "epoch": 1.3392094017094016, + "grad_norm": 0.6385402679443359, + "learning_rate": 0.00014976140824097015, + "loss": 0.8851, + "step": 7522 + }, + { + "epoch": 1.3393874643874644, + "grad_norm": 0.6315767168998718, + "learning_rate": 0.0001497492663666189, + "loss": 0.986, + "step": 7523 + }, + { + "epoch": 1.3395655270655271, + "grad_norm": 0.6379088759422302, + "learning_rate": 0.0001497371235175341, + "loss": 0.9322, + "step": 7524 + }, + { + "epoch": 1.3397435897435899, + "grad_norm": 0.6605859994888306, + "learning_rate": 0.0001497249796939537, + "loss": 1.1112, + "step": 7525 + }, + { + "epoch": 1.3399216524216524, + "grad_norm": 0.7342822551727295, + "learning_rate": 0.0001497128348961156, + "loss": 0.9798, + "step": 7526 + }, + { + "epoch": 1.3400997150997151, + "grad_norm": 0.5667192935943604, + "learning_rate": 0.0001497006891242578, + "loss": 0.7493, + "step": 7527 + }, + { + "epoch": 1.3402777777777777, + "grad_norm": 0.6106827855110168, + "learning_rate": 0.0001496885423786182, + "loss": 1.0924, + "step": 7528 + }, + { + "epoch": 1.3404558404558404, + "grad_norm": 0.6207202076911926, + "learning_rate": 0.00014967639465943486, + "loss": 1.1123, + "step": 7529 + }, + { + "epoch": 1.3406339031339032, + "grad_norm": 0.6272760033607483, + "learning_rate": 0.00014966424596694574, + "loss": 0.9275, + "step": 7530 + }, + { + "epoch": 1.340811965811966, + "grad_norm": 0.6485986113548279, + "learning_rate": 0.0001496520963013889, + "loss": 1.1491, + "step": 7531 + }, + { + "epoch": 1.3409900284900285, + "grad_norm": 0.5743561387062073, + "learning_rate": 0.00014963994566300238, + "loss": 1.1101, + "step": 7532 + }, + { + "epoch": 1.3411680911680912, + "grad_norm": 0.6508657336235046, + "learning_rate": 0.00014962779405202424, + "loss": 1.0368, + "step": 7533 + }, + { + "epoch": 1.3413461538461537, + "grad_norm": 0.6598748564720154, + "learning_rate": 0.00014961564146869259, + "loss": 1.1064, + "step": 7534 + }, + { + "epoch": 1.3415242165242165, + "grad_norm": 0.6722840070724487, + "learning_rate": 0.00014960348791324547, + "loss": 0.9758, + "step": 7535 + }, + { + "epoch": 1.3417022792022792, + "grad_norm": 0.5807220935821533, + "learning_rate": 0.00014959133338592108, + "loss": 0.9936, + "step": 7536 + }, + { + "epoch": 1.341880341880342, + "grad_norm": 0.6318647265434265, + "learning_rate": 0.00014957917788695752, + "loss": 0.907, + "step": 7537 + }, + { + "epoch": 1.3420584045584045, + "grad_norm": 0.6725485324859619, + "learning_rate": 0.00014956702141659295, + "loss": 0.988, + "step": 7538 + }, + { + "epoch": 1.3422364672364673, + "grad_norm": 0.6675217747688293, + "learning_rate": 0.0001495548639750656, + "loss": 1.0194, + "step": 7539 + }, + { + "epoch": 1.3424145299145298, + "grad_norm": 0.6976884603500366, + "learning_rate": 0.0001495427055626136, + "loss": 1.2515, + "step": 7540 + }, + { + "epoch": 1.3425925925925926, + "grad_norm": 0.654941737651825, + "learning_rate": 0.0001495305461794752, + "loss": 1.2072, + "step": 7541 + }, + { + "epoch": 1.3427706552706553, + "grad_norm": 0.7085291743278503, + "learning_rate": 0.00014951838582588864, + "loss": 0.9772, + "step": 7542 + }, + { + "epoch": 1.342948717948718, + "grad_norm": 0.6319566965103149, + "learning_rate": 0.00014950622450209217, + "loss": 1.0162, + "step": 7543 + }, + { + "epoch": 1.3431267806267806, + "grad_norm": 0.6272495985031128, + "learning_rate": 0.00014949406220832407, + "loss": 0.7985, + "step": 7544 + }, + { + "epoch": 1.3433048433048433, + "grad_norm": 0.6352069973945618, + "learning_rate": 0.00014948189894482266, + "loss": 1.0041, + "step": 7545 + }, + { + "epoch": 1.3434829059829059, + "grad_norm": 0.6071867346763611, + "learning_rate": 0.0001494697347118262, + "loss": 0.9486, + "step": 7546 + }, + { + "epoch": 1.3436609686609686, + "grad_norm": 0.6458829641342163, + "learning_rate": 0.00014945756950957308, + "loss": 0.9417, + "step": 7547 + }, + { + "epoch": 1.3438390313390314, + "grad_norm": 0.6472262740135193, + "learning_rate": 0.0001494454033383016, + "loss": 1.056, + "step": 7548 + }, + { + "epoch": 1.3440170940170941, + "grad_norm": 0.6985635161399841, + "learning_rate": 0.00014943323619825017, + "loss": 1.0483, + "step": 7549 + }, + { + "epoch": 1.3441951566951567, + "grad_norm": 0.6379460096359253, + "learning_rate": 0.00014942106808965718, + "loss": 0.9552, + "step": 7550 + }, + { + "epoch": 1.3443732193732194, + "grad_norm": 0.7036557793617249, + "learning_rate": 0.00014940889901276098, + "loss": 0.9647, + "step": 7551 + }, + { + "epoch": 1.344551282051282, + "grad_norm": 0.6697289943695068, + "learning_rate": 0.0001493967289678001, + "loss": 0.9029, + "step": 7552 + }, + { + "epoch": 1.3447293447293447, + "grad_norm": 0.6336250901222229, + "learning_rate": 0.00014938455795501286, + "loss": 0.9458, + "step": 7553 + }, + { + "epoch": 1.3449074074074074, + "grad_norm": 0.7279673218727112, + "learning_rate": 0.00014937238597463785, + "loss": 1.0228, + "step": 7554 + }, + { + "epoch": 1.3450854700854702, + "grad_norm": 0.6514406204223633, + "learning_rate": 0.00014936021302691349, + "loss": 0.8265, + "step": 7555 + }, + { + "epoch": 1.3452635327635327, + "grad_norm": 0.6405338644981384, + "learning_rate": 0.0001493480391120783, + "loss": 0.9516, + "step": 7556 + }, + { + "epoch": 1.3454415954415955, + "grad_norm": 0.6442672610282898, + "learning_rate": 0.00014933586423037076, + "loss": 0.9279, + "step": 7557 + }, + { + "epoch": 1.345619658119658, + "grad_norm": 0.7588633894920349, + "learning_rate": 0.00014932368838202945, + "loss": 1.0976, + "step": 7558 + }, + { + "epoch": 1.3457977207977208, + "grad_norm": 0.5536739230155945, + "learning_rate": 0.00014931151156729296, + "loss": 0.713, + "step": 7559 + }, + { + "epoch": 1.3459757834757835, + "grad_norm": 0.6897570490837097, + "learning_rate": 0.00014929933378639981, + "loss": 0.9521, + "step": 7560 + }, + { + "epoch": 1.3461538461538463, + "grad_norm": 0.6654927134513855, + "learning_rate": 0.00014928715503958863, + "loss": 0.8506, + "step": 7561 + }, + { + "epoch": 1.3463319088319088, + "grad_norm": 0.655806839466095, + "learning_rate": 0.00014927497532709808, + "loss": 0.8636, + "step": 7562 + }, + { + "epoch": 1.3465099715099715, + "grad_norm": 0.6547064185142517, + "learning_rate": 0.00014926279464916667, + "loss": 0.9155, + "step": 7563 + }, + { + "epoch": 1.346688034188034, + "grad_norm": 0.7555415034294128, + "learning_rate": 0.00014925061300603316, + "loss": 0.8791, + "step": 7564 + }, + { + "epoch": 1.3468660968660968, + "grad_norm": 0.7439392805099487, + "learning_rate": 0.0001492384303979362, + "loss": 1.1669, + "step": 7565 + }, + { + "epoch": 1.3470441595441596, + "grad_norm": 0.6016925573348999, + "learning_rate": 0.0001492262468251145, + "loss": 0.9811, + "step": 7566 + }, + { + "epoch": 1.3472222222222223, + "grad_norm": 0.644652783870697, + "learning_rate": 0.00014921406228780675, + "loss": 0.7096, + "step": 7567 + }, + { + "epoch": 1.3474002849002849, + "grad_norm": 0.721814751625061, + "learning_rate": 0.00014920187678625166, + "loss": 0.9933, + "step": 7568 + }, + { + "epoch": 1.3475783475783476, + "grad_norm": 0.6212092638015747, + "learning_rate": 0.000149189690320688, + "loss": 0.8499, + "step": 7569 + }, + { + "epoch": 1.3477564102564101, + "grad_norm": 0.6235958337783813, + "learning_rate": 0.00014917750289135455, + "loss": 0.9189, + "step": 7570 + }, + { + "epoch": 1.3479344729344729, + "grad_norm": 0.6309674978256226, + "learning_rate": 0.0001491653144984901, + "loss": 0.9744, + "step": 7571 + }, + { + "epoch": 1.3481125356125356, + "grad_norm": 0.7606496214866638, + "learning_rate": 0.00014915312514233344, + "loss": 1.0181, + "step": 7572 + }, + { + "epoch": 1.3482905982905984, + "grad_norm": 0.6892654895782471, + "learning_rate": 0.00014914093482312342, + "loss": 0.9517, + "step": 7573 + }, + { + "epoch": 1.348468660968661, + "grad_norm": 0.6746503114700317, + "learning_rate": 0.0001491287435410988, + "loss": 1.056, + "step": 7574 + }, + { + "epoch": 1.3486467236467237, + "grad_norm": 0.5892919301986694, + "learning_rate": 0.00014911655129649858, + "loss": 1.0515, + "step": 7575 + }, + { + "epoch": 1.3488247863247862, + "grad_norm": 0.6278096437454224, + "learning_rate": 0.0001491043580895615, + "loss": 0.864, + "step": 7576 + }, + { + "epoch": 1.349002849002849, + "grad_norm": 0.7017706632614136, + "learning_rate": 0.0001490921639205266, + "loss": 1.0618, + "step": 7577 + }, + { + "epoch": 1.3491809116809117, + "grad_norm": 0.7318746447563171, + "learning_rate": 0.00014907996878963268, + "loss": 0.9905, + "step": 7578 + }, + { + "epoch": 1.3493589743589745, + "grad_norm": 0.6485885977745056, + "learning_rate": 0.00014906777269711873, + "loss": 1.0498, + "step": 7579 + }, + { + "epoch": 1.349537037037037, + "grad_norm": 0.644902229309082, + "learning_rate": 0.00014905557564322372, + "loss": 0.885, + "step": 7580 + }, + { + "epoch": 1.3497150997150997, + "grad_norm": 0.6567610502243042, + "learning_rate": 0.0001490433776281866, + "loss": 0.8938, + "step": 7581 + }, + { + "epoch": 1.3498931623931623, + "grad_norm": 0.6233102679252625, + "learning_rate": 0.0001490311786522464, + "loss": 0.9007, + "step": 7582 + }, + { + "epoch": 1.350071225071225, + "grad_norm": 0.6962146759033203, + "learning_rate": 0.00014901897871564206, + "loss": 0.9257, + "step": 7583 + }, + { + "epoch": 1.3502492877492878, + "grad_norm": 0.6986933350563049, + "learning_rate": 0.00014900677781861266, + "loss": 1.0089, + "step": 7584 + }, + { + "epoch": 1.3504273504273505, + "grad_norm": 0.7527925968170166, + "learning_rate": 0.00014899457596139729, + "loss": 1.0762, + "step": 7585 + }, + { + "epoch": 1.350605413105413, + "grad_norm": 0.69191974401474, + "learning_rate": 0.00014898237314423494, + "loss": 0.9829, + "step": 7586 + }, + { + "epoch": 1.3507834757834758, + "grad_norm": 0.7866443395614624, + "learning_rate": 0.00014897016936736478, + "loss": 1.0911, + "step": 7587 + }, + { + "epoch": 1.3509615384615383, + "grad_norm": 0.7087522745132446, + "learning_rate": 0.00014895796463102587, + "loss": 1.0693, + "step": 7588 + }, + { + "epoch": 1.351139601139601, + "grad_norm": 0.704276442527771, + "learning_rate": 0.00014894575893545736, + "loss": 0.9082, + "step": 7589 + }, + { + "epoch": 1.3513176638176638, + "grad_norm": 0.7074487805366516, + "learning_rate": 0.00014893355228089833, + "loss": 0.8731, + "step": 7590 + }, + { + "epoch": 1.3514957264957266, + "grad_norm": 0.6542425155639648, + "learning_rate": 0.00014892134466758803, + "loss": 0.9325, + "step": 7591 + }, + { + "epoch": 1.351673789173789, + "grad_norm": 0.6577230095863342, + "learning_rate": 0.0001489091360957656, + "loss": 0.8468, + "step": 7592 + }, + { + "epoch": 1.3518518518518519, + "grad_norm": 0.638534426689148, + "learning_rate": 0.00014889692656567025, + "loss": 0.8598, + "step": 7593 + }, + { + "epoch": 1.3520299145299146, + "grad_norm": 0.751133918762207, + "learning_rate": 0.0001488847160775412, + "loss": 1.0006, + "step": 7594 + }, + { + "epoch": 1.3522079772079771, + "grad_norm": 0.6272708773612976, + "learning_rate": 0.00014887250463161767, + "loss": 0.8782, + "step": 7595 + }, + { + "epoch": 1.35238603988604, + "grad_norm": 0.7242439985275269, + "learning_rate": 0.00014886029222813897, + "loss": 1.2443, + "step": 7596 + }, + { + "epoch": 1.3525641025641026, + "grad_norm": 0.6199275851249695, + "learning_rate": 0.0001488480788673443, + "loss": 0.9211, + "step": 7597 + }, + { + "epoch": 1.3527421652421652, + "grad_norm": 0.6401306986808777, + "learning_rate": 0.00014883586454947305, + "loss": 0.8808, + "step": 7598 + }, + { + "epoch": 1.352920227920228, + "grad_norm": 0.6340938806533813, + "learning_rate": 0.00014882364927476443, + "loss": 0.9406, + "step": 7599 + }, + { + "epoch": 1.3530982905982907, + "grad_norm": 0.6388604044914246, + "learning_rate": 0.00014881143304345783, + "loss": 1.0674, + "step": 7600 + }, + { + "epoch": 1.3532763532763532, + "grad_norm": 0.7562061548233032, + "learning_rate": 0.00014879921585579263, + "loss": 1.0959, + "step": 7601 + }, + { + "epoch": 1.353454415954416, + "grad_norm": 0.6303606033325195, + "learning_rate": 0.00014878699771200815, + "loss": 0.9641, + "step": 7602 + }, + { + "epoch": 1.3536324786324787, + "grad_norm": 0.8623232841491699, + "learning_rate": 0.00014877477861234382, + "loss": 1.1529, + "step": 7603 + }, + { + "epoch": 1.3538105413105412, + "grad_norm": 0.6607624888420105, + "learning_rate": 0.00014876255855703896, + "loss": 0.6291, + "step": 7604 + }, + { + "epoch": 1.353988603988604, + "grad_norm": 0.6226931214332581, + "learning_rate": 0.0001487503375463331, + "loss": 0.7485, + "step": 7605 + }, + { + "epoch": 1.3541666666666667, + "grad_norm": 0.7626705169677734, + "learning_rate": 0.00014873811558046565, + "loss": 0.9694, + "step": 7606 + }, + { + "epoch": 1.3543447293447293, + "grad_norm": 0.5436057448387146, + "learning_rate": 0.00014872589265967605, + "loss": 0.6173, + "step": 7607 + }, + { + "epoch": 1.354522792022792, + "grad_norm": 0.7822177410125732, + "learning_rate": 0.00014871366878420382, + "loss": 1.0048, + "step": 7608 + }, + { + "epoch": 1.3547008547008548, + "grad_norm": 0.6955201625823975, + "learning_rate": 0.00014870144395428848, + "loss": 0.9487, + "step": 7609 + }, + { + "epoch": 1.3548789173789173, + "grad_norm": 0.6625505685806274, + "learning_rate": 0.00014868921817016943, + "loss": 0.9389, + "step": 7610 + }, + { + "epoch": 1.35505698005698, + "grad_norm": 0.6625354886054993, + "learning_rate": 0.00014867699143208634, + "loss": 0.9538, + "step": 7611 + }, + { + "epoch": 1.3552350427350428, + "grad_norm": 0.7426592707633972, + "learning_rate": 0.00014866476374027874, + "loss": 1.2566, + "step": 7612 + }, + { + "epoch": 1.3554131054131053, + "grad_norm": 0.6856544017791748, + "learning_rate": 0.00014865253509498616, + "loss": 0.9663, + "step": 7613 + }, + { + "epoch": 1.355591168091168, + "grad_norm": 0.6343915462493896, + "learning_rate": 0.00014864030549644825, + "loss": 0.9416, + "step": 7614 + }, + { + "epoch": 1.3557692307692308, + "grad_norm": 0.6319553256034851, + "learning_rate": 0.00014862807494490454, + "loss": 0.9335, + "step": 7615 + }, + { + "epoch": 1.3559472934472934, + "grad_norm": 0.6919772624969482, + "learning_rate": 0.00014861584344059476, + "loss": 0.8516, + "step": 7616 + }, + { + "epoch": 1.3561253561253561, + "grad_norm": 0.6405790448188782, + "learning_rate": 0.00014860361098375851, + "loss": 1.1278, + "step": 7617 + }, + { + "epoch": 1.3563034188034189, + "grad_norm": 0.7591732144355774, + "learning_rate": 0.00014859137757463548, + "loss": 1.0961, + "step": 7618 + }, + { + "epoch": 1.3564814814814814, + "grad_norm": 0.6166727542877197, + "learning_rate": 0.0001485791432134653, + "loss": 0.9358, + "step": 7619 + }, + { + "epoch": 1.3566595441595442, + "grad_norm": 0.7068707346916199, + "learning_rate": 0.00014856690790048777, + "loss": 0.8325, + "step": 7620 + }, + { + "epoch": 1.356837606837607, + "grad_norm": 0.8465402722358704, + "learning_rate": 0.00014855467163594257, + "loss": 1.0047, + "step": 7621 + }, + { + "epoch": 1.3570156695156697, + "grad_norm": 0.7403460741043091, + "learning_rate": 0.00014854243442006943, + "loss": 1.0907, + "step": 7622 + }, + { + "epoch": 1.3571937321937322, + "grad_norm": 0.6939566135406494, + "learning_rate": 0.00014853019625310813, + "loss": 0.9156, + "step": 7623 + }, + { + "epoch": 1.357371794871795, + "grad_norm": 0.6425924897193909, + "learning_rate": 0.0001485179571352984, + "loss": 0.8156, + "step": 7624 + }, + { + "epoch": 1.3575498575498575, + "grad_norm": 0.7091902494430542, + "learning_rate": 0.00014850571706688013, + "loss": 1.0483, + "step": 7625 + }, + { + "epoch": 1.3577279202279202, + "grad_norm": 0.663342297077179, + "learning_rate": 0.00014849347604809312, + "loss": 1.0405, + "step": 7626 + }, + { + "epoch": 1.357905982905983, + "grad_norm": 0.6727671027183533, + "learning_rate": 0.00014848123407917716, + "loss": 1.0389, + "step": 7627 + }, + { + "epoch": 1.3580840455840457, + "grad_norm": 0.6572692394256592, + "learning_rate": 0.0001484689911603721, + "loss": 1.0489, + "step": 7628 + }, + { + "epoch": 1.3582621082621082, + "grad_norm": 0.7629066109657288, + "learning_rate": 0.0001484567472919179, + "loss": 1.0372, + "step": 7629 + }, + { + "epoch": 1.358440170940171, + "grad_norm": 0.7848913669586182, + "learning_rate": 0.00014844450247405435, + "loss": 0.9437, + "step": 7630 + }, + { + "epoch": 1.3586182336182335, + "grad_norm": 0.715949535369873, + "learning_rate": 0.00014843225670702143, + "loss": 1.1949, + "step": 7631 + }, + { + "epoch": 1.3587962962962963, + "grad_norm": 0.6498245596885681, + "learning_rate": 0.00014842000999105905, + "loss": 0.8845, + "step": 7632 + }, + { + "epoch": 1.358974358974359, + "grad_norm": 0.7251074910163879, + "learning_rate": 0.00014840776232640716, + "loss": 1.093, + "step": 7633 + }, + { + "epoch": 1.3591524216524218, + "grad_norm": 0.6223580837249756, + "learning_rate": 0.0001483955137133057, + "loss": 1.0344, + "step": 7634 + }, + { + "epoch": 1.3593304843304843, + "grad_norm": 0.6504943370819092, + "learning_rate": 0.00014838326415199472, + "loss": 1.109, + "step": 7635 + }, + { + "epoch": 1.359508547008547, + "grad_norm": 0.5912374258041382, + "learning_rate": 0.00014837101364271416, + "loss": 1.0756, + "step": 7636 + }, + { + "epoch": 1.3596866096866096, + "grad_norm": 0.6116467714309692, + "learning_rate": 0.00014835876218570408, + "loss": 0.7871, + "step": 7637 + }, + { + "epoch": 1.3598646723646723, + "grad_norm": 0.7013412117958069, + "learning_rate": 0.0001483465097812045, + "loss": 1.0003, + "step": 7638 + }, + { + "epoch": 1.360042735042735, + "grad_norm": 0.5930750370025635, + "learning_rate": 0.00014833425642945552, + "loss": 0.9926, + "step": 7639 + }, + { + "epoch": 1.3602207977207978, + "grad_norm": 0.732955276966095, + "learning_rate": 0.00014832200213069717, + "loss": 1.2801, + "step": 7640 + }, + { + "epoch": 1.3603988603988604, + "grad_norm": 0.6836149096488953, + "learning_rate": 0.00014830974688516958, + "loss": 0.9292, + "step": 7641 + }, + { + "epoch": 1.3605769230769231, + "grad_norm": 0.6531919836997986, + "learning_rate": 0.00014829749069311283, + "loss": 0.9551, + "step": 7642 + }, + { + "epoch": 1.3607549857549857, + "grad_norm": 0.719093382358551, + "learning_rate": 0.0001482852335547671, + "loss": 0.8588, + "step": 7643 + }, + { + "epoch": 1.3609330484330484, + "grad_norm": 0.6144105792045593, + "learning_rate": 0.00014827297547037252, + "loss": 0.9033, + "step": 7644 + }, + { + "epoch": 1.3611111111111112, + "grad_norm": 0.789241373538971, + "learning_rate": 0.00014826071644016926, + "loss": 1.1916, + "step": 7645 + }, + { + "epoch": 1.361289173789174, + "grad_norm": 0.6137418746948242, + "learning_rate": 0.0001482484564643975, + "loss": 0.9648, + "step": 7646 + }, + { + "epoch": 1.3614672364672364, + "grad_norm": 0.6789261698722839, + "learning_rate": 0.00014823619554329745, + "loss": 0.829, + "step": 7647 + }, + { + "epoch": 1.3616452991452992, + "grad_norm": 0.6508790254592896, + "learning_rate": 0.0001482239336771094, + "loss": 0.942, + "step": 7648 + }, + { + "epoch": 1.3618233618233617, + "grad_norm": 0.6725571751594543, + "learning_rate": 0.00014821167086607353, + "loss": 0.8884, + "step": 7649 + }, + { + "epoch": 1.3620014245014245, + "grad_norm": 0.6252003908157349, + "learning_rate": 0.00014819940711043012, + "loss": 0.9778, + "step": 7650 + }, + { + "epoch": 1.3621794871794872, + "grad_norm": 0.6950626969337463, + "learning_rate": 0.00014818714241041943, + "loss": 1.2104, + "step": 7651 + }, + { + "epoch": 1.36235754985755, + "grad_norm": 0.6527379155158997, + "learning_rate": 0.0001481748767662818, + "loss": 0.7845, + "step": 7652 + }, + { + "epoch": 1.3625356125356125, + "grad_norm": 0.7438235282897949, + "learning_rate": 0.00014816261017825755, + "loss": 0.9513, + "step": 7653 + }, + { + "epoch": 1.3627136752136753, + "grad_norm": 0.6412696838378906, + "learning_rate": 0.000148150342646587, + "loss": 0.8478, + "step": 7654 + }, + { + "epoch": 1.3628917378917378, + "grad_norm": 0.658481240272522, + "learning_rate": 0.00014813807417151046, + "loss": 0.6816, + "step": 7655 + }, + { + "epoch": 1.3630698005698005, + "grad_norm": 0.6170126795768738, + "learning_rate": 0.0001481258047532684, + "loss": 0.8862, + "step": 7656 + }, + { + "epoch": 1.3632478632478633, + "grad_norm": 0.7049173712730408, + "learning_rate": 0.0001481135343921012, + "loss": 1.0027, + "step": 7657 + }, + { + "epoch": 1.363425925925926, + "grad_norm": 0.7780741453170776, + "learning_rate": 0.0001481012630882492, + "loss": 1.0183, + "step": 7658 + }, + { + "epoch": 1.3636039886039886, + "grad_norm": 0.6658362746238708, + "learning_rate": 0.00014808899084195286, + "loss": 0.878, + "step": 7659 + }, + { + "epoch": 1.3637820512820513, + "grad_norm": 0.7192076444625854, + "learning_rate": 0.00014807671765345267, + "loss": 1.2269, + "step": 7660 + }, + { + "epoch": 1.3639601139601139, + "grad_norm": 0.7038660049438477, + "learning_rate": 0.00014806444352298903, + "loss": 0.889, + "step": 7661 + }, + { + "epoch": 1.3641381766381766, + "grad_norm": 0.622803270816803, + "learning_rate": 0.00014805216845080249, + "loss": 0.9623, + "step": 7662 + }, + { + "epoch": 1.3643162393162394, + "grad_norm": 0.9157076478004456, + "learning_rate": 0.00014803989243713353, + "loss": 1.106, + "step": 7663 + }, + { + "epoch": 1.364494301994302, + "grad_norm": 0.6369999647140503, + "learning_rate": 0.00014802761548222268, + "loss": 0.9755, + "step": 7664 + }, + { + "epoch": 1.3646723646723646, + "grad_norm": 0.8318394422531128, + "learning_rate": 0.00014801533758631045, + "loss": 1.1786, + "step": 7665 + }, + { + "epoch": 1.3648504273504274, + "grad_norm": 0.7065796852111816, + "learning_rate": 0.00014800305874963744, + "loss": 1.2066, + "step": 7666 + }, + { + "epoch": 1.36502849002849, + "grad_norm": 0.6570265293121338, + "learning_rate": 0.0001479907789724442, + "loss": 1.0084, + "step": 7667 + }, + { + "epoch": 1.3652065527065527, + "grad_norm": 0.637321949005127, + "learning_rate": 0.00014797849825497135, + "loss": 0.9075, + "step": 7668 + }, + { + "epoch": 1.3653846153846154, + "grad_norm": 0.7656470537185669, + "learning_rate": 0.00014796621659745948, + "loss": 1.1497, + "step": 7669 + }, + { + "epoch": 1.3655626780626782, + "grad_norm": 0.6798120737075806, + "learning_rate": 0.0001479539340001493, + "loss": 0.8154, + "step": 7670 + }, + { + "epoch": 1.3657407407407407, + "grad_norm": 0.7004328966140747, + "learning_rate": 0.0001479416504632813, + "loss": 1.0513, + "step": 7671 + }, + { + "epoch": 1.3659188034188035, + "grad_norm": 0.6551713943481445, + "learning_rate": 0.0001479293659870963, + "loss": 0.8735, + "step": 7672 + }, + { + "epoch": 1.366096866096866, + "grad_norm": 0.7685719132423401, + "learning_rate": 0.00014791708057183494, + "loss": 1.111, + "step": 7673 + }, + { + "epoch": 1.3662749287749287, + "grad_norm": 0.673624575138092, + "learning_rate": 0.0001479047942177379, + "loss": 0.9418, + "step": 7674 + }, + { + "epoch": 1.3664529914529915, + "grad_norm": 0.6281047463417053, + "learning_rate": 0.00014789250692504597, + "loss": 1.0938, + "step": 7675 + }, + { + "epoch": 1.3666310541310542, + "grad_norm": 0.5846312642097473, + "learning_rate": 0.0001478802186939998, + "loss": 0.6352, + "step": 7676 + }, + { + "epoch": 1.3668091168091168, + "grad_norm": 0.7037251591682434, + "learning_rate": 0.00014786792952484025, + "loss": 1.1775, + "step": 7677 + }, + { + "epoch": 1.3669871794871795, + "grad_norm": 0.69822758436203, + "learning_rate": 0.00014785563941780808, + "loss": 1.0877, + "step": 7678 + }, + { + "epoch": 1.367165242165242, + "grad_norm": 0.7229313254356384, + "learning_rate": 0.000147843348373144, + "loss": 1.0305, + "step": 7679 + }, + { + "epoch": 1.3673433048433048, + "grad_norm": 0.665771484375, + "learning_rate": 0.00014783105639108897, + "loss": 0.9056, + "step": 7680 + }, + { + "epoch": 1.3675213675213675, + "grad_norm": 0.6418357491493225, + "learning_rate": 0.00014781876347188367, + "loss": 0.9374, + "step": 7681 + }, + { + "epoch": 1.3676994301994303, + "grad_norm": 0.7255483269691467, + "learning_rate": 0.0001478064696157691, + "loss": 0.8533, + "step": 7682 + }, + { + "epoch": 1.3678774928774928, + "grad_norm": 0.668064534664154, + "learning_rate": 0.00014779417482298603, + "loss": 0.9002, + "step": 7683 + }, + { + "epoch": 1.3680555555555556, + "grad_norm": 0.6797603368759155, + "learning_rate": 0.0001477818790937754, + "loss": 0.9733, + "step": 7684 + }, + { + "epoch": 1.368233618233618, + "grad_norm": 0.6905350685119629, + "learning_rate": 0.0001477695824283781, + "loss": 0.7985, + "step": 7685 + }, + { + "epoch": 1.3684116809116809, + "grad_norm": 0.6846137046813965, + "learning_rate": 0.00014775728482703507, + "loss": 0.9154, + "step": 7686 + }, + { + "epoch": 1.3685897435897436, + "grad_norm": 0.6686832904815674, + "learning_rate": 0.00014774498628998726, + "loss": 0.926, + "step": 7687 + }, + { + "epoch": 1.3687678062678064, + "grad_norm": 0.7050234079360962, + "learning_rate": 0.00014773268681747561, + "loss": 0.9386, + "step": 7688 + }, + { + "epoch": 1.368945868945869, + "grad_norm": 0.7048354744911194, + "learning_rate": 0.00014772038640974112, + "loss": 1.1483, + "step": 7689 + }, + { + "epoch": 1.3691239316239316, + "grad_norm": 0.698192298412323, + "learning_rate": 0.0001477080850670248, + "loss": 1.1452, + "step": 7690 + }, + { + "epoch": 1.3693019943019942, + "grad_norm": 0.6838962435722351, + "learning_rate": 0.00014769578278956766, + "loss": 0.9789, + "step": 7691 + }, + { + "epoch": 1.369480056980057, + "grad_norm": 0.6636955142021179, + "learning_rate": 0.00014768347957761074, + "loss": 0.931, + "step": 7692 + }, + { + "epoch": 1.3696581196581197, + "grad_norm": 0.706030547618866, + "learning_rate": 0.0001476711754313951, + "loss": 1.1096, + "step": 7693 + }, + { + "epoch": 1.3698361823361824, + "grad_norm": 0.6771288514137268, + "learning_rate": 0.00014765887035116178, + "loss": 0.9641, + "step": 7694 + }, + { + "epoch": 1.370014245014245, + "grad_norm": 0.6805008053779602, + "learning_rate": 0.00014764656433715188, + "loss": 0.8724, + "step": 7695 + }, + { + "epoch": 1.3701923076923077, + "grad_norm": 0.6599233746528625, + "learning_rate": 0.00014763425738960657, + "loss": 0.8477, + "step": 7696 + }, + { + "epoch": 1.3703703703703702, + "grad_norm": 0.7036116123199463, + "learning_rate": 0.0001476219495087669, + "loss": 1.0991, + "step": 7697 + }, + { + "epoch": 1.370548433048433, + "grad_norm": 0.6677989363670349, + "learning_rate": 0.0001476096406948741, + "loss": 1.2397, + "step": 7698 + }, + { + "epoch": 1.3707264957264957, + "grad_norm": 0.5652269721031189, + "learning_rate": 0.00014759733094816928, + "loss": 0.9302, + "step": 7699 + }, + { + "epoch": 1.3709045584045585, + "grad_norm": 0.6670156121253967, + "learning_rate": 0.00014758502026889362, + "loss": 0.8362, + "step": 7700 + }, + { + "epoch": 1.371082621082621, + "grad_norm": 0.6705406904220581, + "learning_rate": 0.00014757270865728832, + "loss": 0.876, + "step": 7701 + }, + { + "epoch": 1.3712606837606838, + "grad_norm": 0.6020053625106812, + "learning_rate": 0.00014756039611359465, + "loss": 0.9182, + "step": 7702 + }, + { + "epoch": 1.3714387464387463, + "grad_norm": 0.6370134949684143, + "learning_rate": 0.0001475480826380538, + "loss": 1.1063, + "step": 7703 + }, + { + "epoch": 1.371616809116809, + "grad_norm": 0.6906460523605347, + "learning_rate": 0.00014753576823090705, + "loss": 0.988, + "step": 7704 + }, + { + "epoch": 1.3717948717948718, + "grad_norm": 0.6047569513320923, + "learning_rate": 0.00014752345289239567, + "loss": 1.15, + "step": 7705 + }, + { + "epoch": 1.3719729344729346, + "grad_norm": 0.7019868493080139, + "learning_rate": 0.00014751113662276095, + "loss": 1.1185, + "step": 7706 + }, + { + "epoch": 1.372150997150997, + "grad_norm": 0.6534035801887512, + "learning_rate": 0.00014749881942224417, + "loss": 0.9006, + "step": 7707 + }, + { + "epoch": 1.3723290598290598, + "grad_norm": 0.6111651659011841, + "learning_rate": 0.00014748650129108674, + "loss": 0.935, + "step": 7708 + }, + { + "epoch": 1.3725071225071226, + "grad_norm": 0.6678512096405029, + "learning_rate": 0.00014747418222952995, + "loss": 0.8771, + "step": 7709 + }, + { + "epoch": 1.3726851851851851, + "grad_norm": 0.607829749584198, + "learning_rate": 0.00014746186223781518, + "loss": 1.0509, + "step": 7710 + }, + { + "epoch": 1.3728632478632479, + "grad_norm": 0.7274412512779236, + "learning_rate": 0.00014744954131618382, + "loss": 0.9545, + "step": 7711 + }, + { + "epoch": 1.3730413105413106, + "grad_norm": 0.640333354473114, + "learning_rate": 0.00014743721946487723, + "loss": 1.018, + "step": 7712 + }, + { + "epoch": 1.3732193732193732, + "grad_norm": 0.6772079467773438, + "learning_rate": 0.0001474248966841369, + "loss": 1.0983, + "step": 7713 + }, + { + "epoch": 1.373397435897436, + "grad_norm": 0.49630534648895264, + "learning_rate": 0.00014741257297420422, + "loss": 0.5238, + "step": 7714 + }, + { + "epoch": 1.3735754985754987, + "grad_norm": 0.6316596269607544, + "learning_rate": 0.00014740024833532068, + "loss": 1.1342, + "step": 7715 + }, + { + "epoch": 1.3737535612535612, + "grad_norm": 0.5928404331207275, + "learning_rate": 0.00014738792276772775, + "loss": 0.7987, + "step": 7716 + }, + { + "epoch": 1.373931623931624, + "grad_norm": 0.6773418188095093, + "learning_rate": 0.00014737559627166688, + "loss": 0.934, + "step": 7717 + }, + { + "epoch": 1.3741096866096867, + "grad_norm": 0.7895028591156006, + "learning_rate": 0.00014736326884737963, + "loss": 0.984, + "step": 7718 + }, + { + "epoch": 1.3742877492877492, + "grad_norm": 0.7074753046035767, + "learning_rate": 0.00014735094049510752, + "loss": 1.0093, + "step": 7719 + }, + { + "epoch": 1.374465811965812, + "grad_norm": 0.5389847159385681, + "learning_rate": 0.00014733861121509208, + "loss": 0.8138, + "step": 7720 + }, + { + "epoch": 1.3746438746438747, + "grad_norm": 0.6138495206832886, + "learning_rate": 0.00014732628100757493, + "loss": 0.9282, + "step": 7721 + }, + { + "epoch": 1.3748219373219372, + "grad_norm": 0.7609560489654541, + "learning_rate": 0.00014731394987279757, + "loss": 0.9859, + "step": 7722 + }, + { + "epoch": 1.375, + "grad_norm": 0.6806198954582214, + "learning_rate": 0.00014730161781100165, + "loss": 0.8932, + "step": 7723 + }, + { + "epoch": 1.3751780626780628, + "grad_norm": 0.7229103446006775, + "learning_rate": 0.0001472892848224288, + "loss": 0.956, + "step": 7724 + }, + { + "epoch": 1.3753561253561253, + "grad_norm": 0.6157994866371155, + "learning_rate": 0.00014727695090732066, + "loss": 1.0285, + "step": 7725 + }, + { + "epoch": 1.375534188034188, + "grad_norm": 0.5885980129241943, + "learning_rate": 0.00014726461606591885, + "loss": 0.9174, + "step": 7726 + }, + { + "epoch": 1.3757122507122508, + "grad_norm": 0.6655769944190979, + "learning_rate": 0.0001472522802984651, + "loss": 0.9059, + "step": 7727 + }, + { + "epoch": 1.3758903133903133, + "grad_norm": 0.7075541019439697, + "learning_rate": 0.00014723994360520105, + "loss": 1.0055, + "step": 7728 + }, + { + "epoch": 1.376068376068376, + "grad_norm": 0.6947159171104431, + "learning_rate": 0.00014722760598636847, + "loss": 0.9782, + "step": 7729 + }, + { + "epoch": 1.3762464387464388, + "grad_norm": 0.6629964709281921, + "learning_rate": 0.00014721526744220905, + "loss": 0.9427, + "step": 7730 + }, + { + "epoch": 1.3764245014245013, + "grad_norm": 0.7385284304618835, + "learning_rate": 0.00014720292797296453, + "loss": 0.9953, + "step": 7731 + }, + { + "epoch": 1.376602564102564, + "grad_norm": 0.6123563051223755, + "learning_rate": 0.0001471905875788767, + "loss": 1.0103, + "step": 7732 + }, + { + "epoch": 1.3767806267806268, + "grad_norm": 0.6457047462463379, + "learning_rate": 0.00014717824626018732, + "loss": 0.9779, + "step": 7733 + }, + { + "epoch": 1.3769586894586894, + "grad_norm": 0.6196442246437073, + "learning_rate": 0.00014716590401713824, + "loss": 0.8747, + "step": 7734 + }, + { + "epoch": 1.3771367521367521, + "grad_norm": 0.7932298183441162, + "learning_rate": 0.00014715356084997122, + "loss": 1.1617, + "step": 7735 + }, + { + "epoch": 1.3773148148148149, + "grad_norm": 0.787304699420929, + "learning_rate": 0.00014714121675892815, + "loss": 1.1383, + "step": 7736 + }, + { + "epoch": 1.3774928774928774, + "grad_norm": 0.672795295715332, + "learning_rate": 0.00014712887174425085, + "loss": 1.2563, + "step": 7737 + }, + { + "epoch": 1.3776709401709402, + "grad_norm": 0.6505744457244873, + "learning_rate": 0.00014711652580618123, + "loss": 0.9194, + "step": 7738 + }, + { + "epoch": 1.377849002849003, + "grad_norm": 0.8141193985939026, + "learning_rate": 0.00014710417894496115, + "loss": 1.1428, + "step": 7739 + }, + { + "epoch": 1.3780270655270654, + "grad_norm": 0.6269707679748535, + "learning_rate": 0.00014709183116083253, + "loss": 0.7164, + "step": 7740 + }, + { + "epoch": 1.3782051282051282, + "grad_norm": 0.6737076640129089, + "learning_rate": 0.0001470794824540373, + "loss": 0.9965, + "step": 7741 + }, + { + "epoch": 1.378383190883191, + "grad_norm": 0.6451728343963623, + "learning_rate": 0.0001470671328248174, + "loss": 1.0539, + "step": 7742 + }, + { + "epoch": 1.3785612535612537, + "grad_norm": 0.6480295062065125, + "learning_rate": 0.00014705478227341486, + "loss": 0.9118, + "step": 7743 + }, + { + "epoch": 1.3787393162393162, + "grad_norm": 0.7429090738296509, + "learning_rate": 0.00014704243080007154, + "loss": 1.0031, + "step": 7744 + }, + { + "epoch": 1.378917378917379, + "grad_norm": 0.5601376891136169, + "learning_rate": 0.00014703007840502955, + "loss": 0.849, + "step": 7745 + }, + { + "epoch": 1.3790954415954415, + "grad_norm": 0.7067657113075256, + "learning_rate": 0.00014701772508853088, + "loss": 1.3067, + "step": 7746 + }, + { + "epoch": 1.3792735042735043, + "grad_norm": 0.7016390562057495, + "learning_rate": 0.00014700537085081755, + "loss": 1.0236, + "step": 7747 + }, + { + "epoch": 1.379451566951567, + "grad_norm": 0.6505000591278076, + "learning_rate": 0.0001469930156921316, + "loss": 1.0121, + "step": 7748 + }, + { + "epoch": 1.3796296296296298, + "grad_norm": 0.8515380620956421, + "learning_rate": 0.00014698065961271512, + "loss": 1.0413, + "step": 7749 + }, + { + "epoch": 1.3798076923076923, + "grad_norm": 0.6322008371353149, + "learning_rate": 0.00014696830261281025, + "loss": 0.8306, + "step": 7750 + }, + { + "epoch": 1.379985754985755, + "grad_norm": 0.7090431451797485, + "learning_rate": 0.00014695594469265902, + "loss": 1.1829, + "step": 7751 + }, + { + "epoch": 1.3801638176638176, + "grad_norm": 0.5913167595863342, + "learning_rate": 0.00014694358585250363, + "loss": 0.9769, + "step": 7752 + }, + { + "epoch": 1.3803418803418803, + "grad_norm": 0.7345432639122009, + "learning_rate": 0.00014693122609258616, + "loss": 0.9928, + "step": 7753 + }, + { + "epoch": 1.380519943019943, + "grad_norm": 0.6158214211463928, + "learning_rate": 0.00014691886541314884, + "loss": 1.1166, + "step": 7754 + }, + { + "epoch": 1.3806980056980058, + "grad_norm": 0.6874041557312012, + "learning_rate": 0.0001469065038144338, + "loss": 1.0808, + "step": 7755 + }, + { + "epoch": 1.3808760683760684, + "grad_norm": 0.8135195970535278, + "learning_rate": 0.00014689414129668326, + "loss": 0.9482, + "step": 7756 + }, + { + "epoch": 1.381054131054131, + "grad_norm": 0.6389174461364746, + "learning_rate": 0.00014688177786013944, + "loss": 1.039, + "step": 7757 + }, + { + "epoch": 1.3812321937321936, + "grad_norm": 0.6953016519546509, + "learning_rate": 0.00014686941350504454, + "loss": 0.9426, + "step": 7758 + }, + { + "epoch": 1.3814102564102564, + "grad_norm": 0.8171859383583069, + "learning_rate": 0.00014685704823164087, + "loss": 1.0393, + "step": 7759 + }, + { + "epoch": 1.3815883190883191, + "grad_norm": 0.6968414783477783, + "learning_rate": 0.0001468446820401707, + "loss": 1.1167, + "step": 7760 + }, + { + "epoch": 1.381766381766382, + "grad_norm": 0.6916623711585999, + "learning_rate": 0.00014683231493087628, + "loss": 1.1886, + "step": 7761 + }, + { + "epoch": 1.3819444444444444, + "grad_norm": 0.7351683378219604, + "learning_rate": 0.00014681994690399992, + "loss": 0.9893, + "step": 7762 + }, + { + "epoch": 1.3821225071225072, + "grad_norm": 0.6617491245269775, + "learning_rate": 0.00014680757795978395, + "loss": 1.0505, + "step": 7763 + }, + { + "epoch": 1.3823005698005697, + "grad_norm": 0.6627485156059265, + "learning_rate": 0.00014679520809847074, + "loss": 0.9878, + "step": 7764 + }, + { + "epoch": 1.3824786324786325, + "grad_norm": 0.704636812210083, + "learning_rate": 0.00014678283732030264, + "loss": 0.8332, + "step": 7765 + }, + { + "epoch": 1.3826566951566952, + "grad_norm": 0.698853075504303, + "learning_rate": 0.00014677046562552203, + "loss": 1.0926, + "step": 7766 + }, + { + "epoch": 1.382834757834758, + "grad_norm": 0.6695869565010071, + "learning_rate": 0.0001467580930143713, + "loss": 1.0626, + "step": 7767 + }, + { + "epoch": 1.3830128205128205, + "grad_norm": 0.672173023223877, + "learning_rate": 0.00014674571948709286, + "loss": 0.8842, + "step": 7768 + }, + { + "epoch": 1.3831908831908832, + "grad_norm": 0.6735473871231079, + "learning_rate": 0.00014673334504392916, + "loss": 0.9382, + "step": 7769 + }, + { + "epoch": 1.3833689458689458, + "grad_norm": 0.6864013075828552, + "learning_rate": 0.00014672096968512265, + "loss": 1.1369, + "step": 7770 + }, + { + "epoch": 1.3835470085470085, + "grad_norm": 0.7154954075813293, + "learning_rate": 0.0001467085934109158, + "loss": 1.1447, + "step": 7771 + }, + { + "epoch": 1.3837250712250713, + "grad_norm": 0.5934487581253052, + "learning_rate": 0.0001466962162215511, + "loss": 0.8923, + "step": 7772 + }, + { + "epoch": 1.383903133903134, + "grad_norm": 0.8116832971572876, + "learning_rate": 0.00014668383811727097, + "loss": 1.0997, + "step": 7773 + }, + { + "epoch": 1.3840811965811965, + "grad_norm": 0.8661674857139587, + "learning_rate": 0.00014667145909831808, + "loss": 1.0112, + "step": 7774 + }, + { + "epoch": 1.3842592592592593, + "grad_norm": 0.5173856616020203, + "learning_rate": 0.00014665907916493488, + "loss": 0.6571, + "step": 7775 + }, + { + "epoch": 1.3844373219373218, + "grad_norm": 0.6165067553520203, + "learning_rate": 0.00014664669831736395, + "loss": 1.0992, + "step": 7776 + }, + { + "epoch": 1.3846153846153846, + "grad_norm": 0.6564429998397827, + "learning_rate": 0.00014663431655584787, + "loss": 0.9103, + "step": 7777 + }, + { + "epoch": 1.3847934472934473, + "grad_norm": 0.7162124514579773, + "learning_rate": 0.00014662193388062923, + "loss": 1.0645, + "step": 7778 + }, + { + "epoch": 1.38497150997151, + "grad_norm": 0.6391215920448303, + "learning_rate": 0.00014660955029195064, + "loss": 0.902, + "step": 7779 + }, + { + "epoch": 1.3851495726495726, + "grad_norm": 0.6876635551452637, + "learning_rate": 0.00014659716579005475, + "loss": 1.0924, + "step": 7780 + }, + { + "epoch": 1.3853276353276354, + "grad_norm": 0.7254653573036194, + "learning_rate": 0.00014658478037518418, + "loss": 1.0135, + "step": 7781 + }, + { + "epoch": 1.385505698005698, + "grad_norm": 0.6900535225868225, + "learning_rate": 0.00014657239404758162, + "loss": 0.983, + "step": 7782 + }, + { + "epoch": 1.3856837606837606, + "grad_norm": 0.7477042078971863, + "learning_rate": 0.00014656000680748975, + "loss": 1.0707, + "step": 7783 + }, + { + "epoch": 1.3858618233618234, + "grad_norm": 0.5756927132606506, + "learning_rate": 0.00014654761865515124, + "loss": 0.8881, + "step": 7784 + }, + { + "epoch": 1.3860398860398861, + "grad_norm": 0.6736083626747131, + "learning_rate": 0.00014653522959080884, + "loss": 1.0193, + "step": 7785 + }, + { + "epoch": 1.3862179487179487, + "grad_norm": 0.616179883480072, + "learning_rate": 0.0001465228396147053, + "loss": 0.8676, + "step": 7786 + }, + { + "epoch": 1.3863960113960114, + "grad_norm": 0.7956456542015076, + "learning_rate": 0.00014651044872708338, + "loss": 0.9787, + "step": 7787 + }, + { + "epoch": 1.386574074074074, + "grad_norm": 0.6613463163375854, + "learning_rate": 0.00014649805692818578, + "loss": 1.0032, + "step": 7788 + }, + { + "epoch": 1.3867521367521367, + "grad_norm": 0.6215800642967224, + "learning_rate": 0.0001464856642182554, + "loss": 1.0123, + "step": 7789 + }, + { + "epoch": 1.3869301994301995, + "grad_norm": 0.6701171398162842, + "learning_rate": 0.00014647327059753496, + "loss": 0.9108, + "step": 7790 + }, + { + "epoch": 1.3871082621082622, + "grad_norm": 0.6213465929031372, + "learning_rate": 0.00014646087606626736, + "loss": 0.9313, + "step": 7791 + }, + { + "epoch": 1.3872863247863247, + "grad_norm": 0.7535304427146912, + "learning_rate": 0.00014644848062469535, + "loss": 1.0813, + "step": 7792 + }, + { + "epoch": 1.3874643874643875, + "grad_norm": 0.6778230667114258, + "learning_rate": 0.0001464360842730619, + "loss": 1.0405, + "step": 7793 + }, + { + "epoch": 1.38764245014245, + "grad_norm": 0.7816025614738464, + "learning_rate": 0.0001464236870116098, + "loss": 0.9228, + "step": 7794 + }, + { + "epoch": 1.3878205128205128, + "grad_norm": 0.6815229058265686, + "learning_rate": 0.00014641128884058203, + "loss": 0.9607, + "step": 7795 + }, + { + "epoch": 1.3879985754985755, + "grad_norm": 0.7027714848518372, + "learning_rate": 0.00014639888976022145, + "loss": 0.9379, + "step": 7796 + }, + { + "epoch": 1.3881766381766383, + "grad_norm": 0.7636353373527527, + "learning_rate": 0.00014638648977077104, + "loss": 1.1186, + "step": 7797 + }, + { + "epoch": 1.3883547008547008, + "grad_norm": 0.6732974052429199, + "learning_rate": 0.00014637408887247365, + "loss": 1.1378, + "step": 7798 + }, + { + "epoch": 1.3885327635327636, + "grad_norm": 0.7539397478103638, + "learning_rate": 0.0001463616870655724, + "loss": 0.999, + "step": 7799 + }, + { + "epoch": 1.388710826210826, + "grad_norm": 0.6872972846031189, + "learning_rate": 0.00014634928435031013, + "loss": 0.9564, + "step": 7800 + }, + { + "epoch": 1.3888888888888888, + "grad_norm": 0.6823115348815918, + "learning_rate": 0.00014633688072693, + "loss": 0.9745, + "step": 7801 + }, + { + "epoch": 1.3890669515669516, + "grad_norm": 0.6462571620941162, + "learning_rate": 0.00014632447619567488, + "loss": 0.8314, + "step": 7802 + }, + { + "epoch": 1.3892450142450143, + "grad_norm": 0.7245402932167053, + "learning_rate": 0.0001463120707567879, + "loss": 0.8291, + "step": 7803 + }, + { + "epoch": 1.3894230769230769, + "grad_norm": 0.697179913520813, + "learning_rate": 0.00014629966441051208, + "loss": 1.017, + "step": 7804 + }, + { + "epoch": 1.3896011396011396, + "grad_norm": 0.6304250359535217, + "learning_rate": 0.00014628725715709053, + "loss": 0.9262, + "step": 7805 + }, + { + "epoch": 1.3897792022792022, + "grad_norm": 0.5780240297317505, + "learning_rate": 0.00014627484899676634, + "loss": 0.6596, + "step": 7806 + }, + { + "epoch": 1.389957264957265, + "grad_norm": 0.8030684590339661, + "learning_rate": 0.0001462624399297826, + "loss": 0.9977, + "step": 7807 + }, + { + "epoch": 1.3901353276353277, + "grad_norm": 0.7999774813652039, + "learning_rate": 0.00014625002995638246, + "loss": 1.1036, + "step": 7808 + }, + { + "epoch": 1.3903133903133904, + "grad_norm": 0.7054862976074219, + "learning_rate": 0.00014623761907680904, + "loss": 1.1435, + "step": 7809 + }, + { + "epoch": 1.390491452991453, + "grad_norm": 0.6660647392272949, + "learning_rate": 0.00014622520729130556, + "loss": 0.703, + "step": 7810 + }, + { + "epoch": 1.3906695156695157, + "grad_norm": 0.6339690089225769, + "learning_rate": 0.00014621279460011515, + "loss": 1.0451, + "step": 7811 + }, + { + "epoch": 1.3908475783475782, + "grad_norm": 0.8568736910820007, + "learning_rate": 0.00014620038100348102, + "loss": 1.009, + "step": 7812 + }, + { + "epoch": 1.391025641025641, + "grad_norm": 0.7126797437667847, + "learning_rate": 0.00014618796650164642, + "loss": 0.9592, + "step": 7813 + }, + { + "epoch": 1.3912037037037037, + "grad_norm": 0.6768994331359863, + "learning_rate": 0.00014617555109485453, + "loss": 1.09, + "step": 7814 + }, + { + "epoch": 1.3913817663817665, + "grad_norm": 0.7609471678733826, + "learning_rate": 0.00014616313478334864, + "loss": 0.9781, + "step": 7815 + }, + { + "epoch": 1.391559829059829, + "grad_norm": 0.7107006907463074, + "learning_rate": 0.00014615071756737203, + "loss": 0.9769, + "step": 7816 + }, + { + "epoch": 1.3917378917378918, + "grad_norm": 0.6324763894081116, + "learning_rate": 0.00014613829944716802, + "loss": 1.089, + "step": 7817 + }, + { + "epoch": 1.3919159544159543, + "grad_norm": 0.6617186069488525, + "learning_rate": 0.00014612588042297984, + "loss": 1.0466, + "step": 7818 + }, + { + "epoch": 1.392094017094017, + "grad_norm": 0.7881436944007874, + "learning_rate": 0.00014611346049505083, + "loss": 1.003, + "step": 7819 + }, + { + "epoch": 1.3922720797720798, + "grad_norm": 0.7391049861907959, + "learning_rate": 0.00014610103966362437, + "loss": 1.0531, + "step": 7820 + }, + { + "epoch": 1.3924501424501425, + "grad_norm": 0.6299472451210022, + "learning_rate": 0.00014608861792894383, + "loss": 0.8433, + "step": 7821 + }, + { + "epoch": 1.392628205128205, + "grad_norm": 0.6053452491760254, + "learning_rate": 0.00014607619529125255, + "loss": 0.7945, + "step": 7822 + }, + { + "epoch": 1.3928062678062678, + "grad_norm": 0.7160114645957947, + "learning_rate": 0.0001460637717507939, + "loss": 1.1604, + "step": 7823 + }, + { + "epoch": 1.3929843304843303, + "grad_norm": 0.6308854222297668, + "learning_rate": 0.00014605134730781135, + "loss": 1.0918, + "step": 7824 + }, + { + "epoch": 1.393162393162393, + "grad_norm": 0.7187000513076782, + "learning_rate": 0.00014603892196254833, + "loss": 1.0594, + "step": 7825 + }, + { + "epoch": 1.3933404558404558, + "grad_norm": 0.7516581416130066, + "learning_rate": 0.00014602649571524826, + "loss": 0.9222, + "step": 7826 + }, + { + "epoch": 1.3935185185185186, + "grad_norm": 0.6340481638908386, + "learning_rate": 0.00014601406856615463, + "loss": 0.8131, + "step": 7827 + }, + { + "epoch": 1.3936965811965811, + "grad_norm": 0.8161744475364685, + "learning_rate": 0.0001460016405155109, + "loss": 0.8695, + "step": 7828 + }, + { + "epoch": 1.3938746438746439, + "grad_norm": 0.6926971077919006, + "learning_rate": 0.0001459892115635606, + "loss": 0.9548, + "step": 7829 + }, + { + "epoch": 1.3940527065527066, + "grad_norm": 0.6669796109199524, + "learning_rate": 0.0001459767817105472, + "loss": 0.9255, + "step": 7830 + }, + { + "epoch": 1.3942307692307692, + "grad_norm": 0.6626184582710266, + "learning_rate": 0.00014596435095671432, + "loss": 1.1141, + "step": 7831 + }, + { + "epoch": 1.394408831908832, + "grad_norm": 0.6755738854408264, + "learning_rate": 0.00014595191930230546, + "loss": 0.9596, + "step": 7832 + }, + { + "epoch": 1.3945868945868947, + "grad_norm": 0.6034863591194153, + "learning_rate": 0.00014593948674756417, + "loss": 0.8088, + "step": 7833 + }, + { + "epoch": 1.3947649572649572, + "grad_norm": 0.5638226866722107, + "learning_rate": 0.00014592705329273406, + "loss": 0.5828, + "step": 7834 + }, + { + "epoch": 1.39494301994302, + "grad_norm": 0.6902222633361816, + "learning_rate": 0.0001459146189380588, + "loss": 0.7954, + "step": 7835 + }, + { + "epoch": 1.3951210826210827, + "grad_norm": 0.7579947710037231, + "learning_rate": 0.0001459021836837819, + "loss": 1.1301, + "step": 7836 + }, + { + "epoch": 1.3952991452991452, + "grad_norm": 0.6894911527633667, + "learning_rate": 0.00014588974753014712, + "loss": 1.082, + "step": 7837 + }, + { + "epoch": 1.395477207977208, + "grad_norm": 0.6330230832099915, + "learning_rate": 0.000145877310477398, + "loss": 0.7614, + "step": 7838 + }, + { + "epoch": 1.3956552706552707, + "grad_norm": 0.6164960265159607, + "learning_rate": 0.00014586487252577832, + "loss": 0.8981, + "step": 7839 + }, + { + "epoch": 1.3958333333333333, + "grad_norm": 0.6575061678886414, + "learning_rate": 0.0001458524336755317, + "loss": 0.9735, + "step": 7840 + }, + { + "epoch": 1.396011396011396, + "grad_norm": 0.687921941280365, + "learning_rate": 0.00014583999392690195, + "loss": 0.9207, + "step": 7841 + }, + { + "epoch": 1.3961894586894588, + "grad_norm": 0.6175212860107422, + "learning_rate": 0.00014582755328013274, + "loss": 1.0444, + "step": 7842 + }, + { + "epoch": 1.3963675213675213, + "grad_norm": 0.6351733207702637, + "learning_rate": 0.00014581511173546781, + "loss": 1.0143, + "step": 7843 + }, + { + "epoch": 1.396545584045584, + "grad_norm": 0.7235051989555359, + "learning_rate": 0.00014580266929315093, + "loss": 0.9108, + "step": 7844 + }, + { + "epoch": 1.3967236467236468, + "grad_norm": 0.6432043313980103, + "learning_rate": 0.00014579022595342586, + "loss": 0.8674, + "step": 7845 + }, + { + "epoch": 1.3969017094017093, + "grad_norm": 0.7775412797927856, + "learning_rate": 0.00014577778171653648, + "loss": 1.0637, + "step": 7846 + }, + { + "epoch": 1.397079772079772, + "grad_norm": 0.6748763918876648, + "learning_rate": 0.00014576533658272655, + "loss": 1.0356, + "step": 7847 + }, + { + "epoch": 1.3972578347578348, + "grad_norm": 0.6940401196479797, + "learning_rate": 0.00014575289055223994, + "loss": 0.9937, + "step": 7848 + }, + { + "epoch": 1.3974358974358974, + "grad_norm": 0.6971304416656494, + "learning_rate": 0.00014574044362532045, + "loss": 0.9753, + "step": 7849 + }, + { + "epoch": 1.39761396011396, + "grad_norm": 0.6576017141342163, + "learning_rate": 0.00014572799580221197, + "loss": 1.1233, + "step": 7850 + }, + { + "epoch": 1.3977920227920229, + "grad_norm": 0.6270702481269836, + "learning_rate": 0.00014571554708315843, + "loss": 0.9771, + "step": 7851 + }, + { + "epoch": 1.3979700854700854, + "grad_norm": 0.6898425817489624, + "learning_rate": 0.00014570309746840372, + "loss": 0.9235, + "step": 7852 + }, + { + "epoch": 1.3981481481481481, + "grad_norm": 0.7017102241516113, + "learning_rate": 0.00014569064695819174, + "loss": 1.1056, + "step": 7853 + }, + { + "epoch": 1.398326210826211, + "grad_norm": 0.6298288702964783, + "learning_rate": 0.00014567819555276647, + "loss": 0.8635, + "step": 7854 + }, + { + "epoch": 1.3985042735042734, + "grad_norm": 0.7173134684562683, + "learning_rate": 0.00014566574325237182, + "loss": 1.0893, + "step": 7855 + }, + { + "epoch": 1.3986823361823362, + "grad_norm": 0.7541036605834961, + "learning_rate": 0.0001456532900572518, + "loss": 1.0996, + "step": 7856 + }, + { + "epoch": 1.398860398860399, + "grad_norm": 0.6204771399497986, + "learning_rate": 0.0001456408359676504, + "loss": 0.7601, + "step": 7857 + }, + { + "epoch": 1.3990384615384617, + "grad_norm": 0.629557192325592, + "learning_rate": 0.00014562838098381163, + "loss": 0.9239, + "step": 7858 + }, + { + "epoch": 1.3992165242165242, + "grad_norm": 0.6878390908241272, + "learning_rate": 0.00014561592510597954, + "loss": 0.9641, + "step": 7859 + }, + { + "epoch": 1.399394586894587, + "grad_norm": 0.7490049004554749, + "learning_rate": 0.00014560346833439813, + "loss": 1.0198, + "step": 7860 + }, + { + "epoch": 1.3995726495726495, + "grad_norm": 0.6337960958480835, + "learning_rate": 0.0001455910106693115, + "loss": 0.8709, + "step": 7861 + }, + { + "epoch": 1.3997507122507122, + "grad_norm": 0.6210524439811707, + "learning_rate": 0.0001455785521109637, + "loss": 1.1049, + "step": 7862 + }, + { + "epoch": 1.399928774928775, + "grad_norm": 0.7894936203956604, + "learning_rate": 0.00014556609265959887, + "loss": 0.8933, + "step": 7863 + }, + { + "epoch": 1.4001068376068377, + "grad_norm": 0.6888098120689392, + "learning_rate": 0.00014555363231546112, + "loss": 0.9738, + "step": 7864 + }, + { + "epoch": 1.4002849002849003, + "grad_norm": 0.608799934387207, + "learning_rate": 0.00014554117107879456, + "loss": 0.9103, + "step": 7865 + }, + { + "epoch": 1.400462962962963, + "grad_norm": 0.7390474081039429, + "learning_rate": 0.00014552870894984335, + "loss": 1.2484, + "step": 7866 + }, + { + "epoch": 1.4006410256410255, + "grad_norm": 0.6513381600379944, + "learning_rate": 0.00014551624592885169, + "loss": 0.8523, + "step": 7867 + }, + { + "epoch": 1.4008190883190883, + "grad_norm": 0.6357464790344238, + "learning_rate": 0.00014550378201606373, + "loss": 0.9594, + "step": 7868 + }, + { + "epoch": 1.400997150997151, + "grad_norm": 0.6893286108970642, + "learning_rate": 0.0001454913172117237, + "loss": 0.9798, + "step": 7869 + }, + { + "epoch": 1.4011752136752138, + "grad_norm": 0.6566550731658936, + "learning_rate": 0.0001454788515160758, + "loss": 1.0532, + "step": 7870 + }, + { + "epoch": 1.4013532763532763, + "grad_norm": 0.6442158222198486, + "learning_rate": 0.00014546638492936425, + "loss": 1.0789, + "step": 7871 + }, + { + "epoch": 1.401531339031339, + "grad_norm": 0.7570971846580505, + "learning_rate": 0.0001454539174518334, + "loss": 0.9806, + "step": 7872 + }, + { + "epoch": 1.4017094017094016, + "grad_norm": 0.6180047392845154, + "learning_rate": 0.0001454414490837274, + "loss": 0.857, + "step": 7873 + }, + { + "epoch": 1.4018874643874644, + "grad_norm": 0.7143170237541199, + "learning_rate": 0.0001454289798252906, + "loss": 0.8815, + "step": 7874 + }, + { + "epoch": 1.4020655270655271, + "grad_norm": 0.6388922929763794, + "learning_rate": 0.00014541650967676736, + "loss": 0.95, + "step": 7875 + }, + { + "epoch": 1.4022435897435899, + "grad_norm": 0.7137351632118225, + "learning_rate": 0.00014540403863840193, + "loss": 0.8973, + "step": 7876 + }, + { + "epoch": 1.4024216524216524, + "grad_norm": 0.656315267086029, + "learning_rate": 0.0001453915667104387, + "loss": 1.149, + "step": 7877 + }, + { + "epoch": 1.4025997150997151, + "grad_norm": 0.7234711647033691, + "learning_rate": 0.000145379093893122, + "loss": 0.9798, + "step": 7878 + }, + { + "epoch": 1.4027777777777777, + "grad_norm": 0.6595289707183838, + "learning_rate": 0.00014536662018669623, + "loss": 1.2704, + "step": 7879 + }, + { + "epoch": 1.4029558404558404, + "grad_norm": 0.6760551333427429, + "learning_rate": 0.00014535414559140576, + "loss": 0.8672, + "step": 7880 + }, + { + "epoch": 1.4031339031339032, + "grad_norm": 0.5916706919670105, + "learning_rate": 0.000145341670107495, + "loss": 0.888, + "step": 7881 + }, + { + "epoch": 1.403311965811966, + "grad_norm": 0.7272133231163025, + "learning_rate": 0.00014532919373520846, + "loss": 1.0466, + "step": 7882 + }, + { + "epoch": 1.4034900284900285, + "grad_norm": 0.8512467741966248, + "learning_rate": 0.00014531671647479048, + "loss": 1.2482, + "step": 7883 + }, + { + "epoch": 1.4036680911680912, + "grad_norm": 0.5536492466926575, + "learning_rate": 0.0001453042383264856, + "loss": 0.7823, + "step": 7884 + }, + { + "epoch": 1.4038461538461537, + "grad_norm": 0.7262215614318848, + "learning_rate": 0.0001452917592905383, + "loss": 0.9713, + "step": 7885 + }, + { + "epoch": 1.4040242165242165, + "grad_norm": 0.7146059274673462, + "learning_rate": 0.00014527927936719304, + "loss": 1.1064, + "step": 7886 + }, + { + "epoch": 1.4042022792022792, + "grad_norm": 0.5915318131446838, + "learning_rate": 0.00014526679855669436, + "loss": 0.8567, + "step": 7887 + }, + { + "epoch": 1.404380341880342, + "grad_norm": 0.6548298001289368, + "learning_rate": 0.00014525431685928682, + "loss": 1.1359, + "step": 7888 + }, + { + "epoch": 1.4045584045584045, + "grad_norm": 0.7482563853263855, + "learning_rate": 0.0001452418342752149, + "loss": 0.9095, + "step": 7889 + }, + { + "epoch": 1.4047364672364673, + "grad_norm": 0.6660130023956299, + "learning_rate": 0.0001452293508047233, + "loss": 1.2343, + "step": 7890 + }, + { + "epoch": 1.4049145299145298, + "grad_norm": 0.7457148432731628, + "learning_rate": 0.00014521686644805644, + "loss": 1.2086, + "step": 7891 + }, + { + "epoch": 1.4050925925925926, + "grad_norm": 0.5957929491996765, + "learning_rate": 0.00014520438120545906, + "loss": 0.9724, + "step": 7892 + }, + { + "epoch": 1.4052706552706553, + "grad_norm": 0.6832270622253418, + "learning_rate": 0.00014519189507717573, + "loss": 0.9903, + "step": 7893 + }, + { + "epoch": 1.405448717948718, + "grad_norm": 0.6202489733695984, + "learning_rate": 0.00014517940806345109, + "loss": 0.962, + "step": 7894 + }, + { + "epoch": 1.4056267806267806, + "grad_norm": 0.6419472694396973, + "learning_rate": 0.0001451669201645298, + "loss": 0.8147, + "step": 7895 + }, + { + "epoch": 1.4058048433048433, + "grad_norm": 0.61143958568573, + "learning_rate": 0.00014515443138065652, + "loss": 0.8674, + "step": 7896 + }, + { + "epoch": 1.4059829059829059, + "grad_norm": 0.7527356743812561, + "learning_rate": 0.00014514194171207597, + "loss": 1.0581, + "step": 7897 + }, + { + "epoch": 1.4061609686609686, + "grad_norm": 0.7195194363594055, + "learning_rate": 0.00014512945115903285, + "loss": 1.0268, + "step": 7898 + }, + { + "epoch": 1.4063390313390314, + "grad_norm": 0.7919661998748779, + "learning_rate": 0.00014511695972177187, + "loss": 1.0259, + "step": 7899 + }, + { + "epoch": 1.4065170940170941, + "grad_norm": 0.6774758696556091, + "learning_rate": 0.00014510446740053783, + "loss": 1.1214, + "step": 7900 + }, + { + "epoch": 1.4066951566951567, + "grad_norm": 0.6102406978607178, + "learning_rate": 0.0001450919741955754, + "loss": 1.1846, + "step": 7901 + }, + { + "epoch": 1.4068732193732194, + "grad_norm": 0.7189443707466125, + "learning_rate": 0.00014507948010712942, + "loss": 0.7758, + "step": 7902 + }, + { + "epoch": 1.407051282051282, + "grad_norm": 0.654153048992157, + "learning_rate": 0.00014506698513544467, + "loss": 0.899, + "step": 7903 + }, + { + "epoch": 1.4072293447293447, + "grad_norm": 0.637934684753418, + "learning_rate": 0.00014505448928076598, + "loss": 0.8301, + "step": 7904 + }, + { + "epoch": 1.4074074074074074, + "grad_norm": 0.7504615783691406, + "learning_rate": 0.00014504199254333812, + "loss": 0.9883, + "step": 7905 + }, + { + "epoch": 1.4075854700854702, + "grad_norm": 0.7902522683143616, + "learning_rate": 0.00014502949492340602, + "loss": 0.9615, + "step": 7906 + }, + { + "epoch": 1.4077635327635327, + "grad_norm": 0.5832732319831848, + "learning_rate": 0.0001450169964212145, + "loss": 0.7136, + "step": 7907 + }, + { + "epoch": 1.4079415954415955, + "grad_norm": 0.6025400757789612, + "learning_rate": 0.00014500449703700846, + "loss": 0.8812, + "step": 7908 + }, + { + "epoch": 1.408119658119658, + "grad_norm": 0.6412411332130432, + "learning_rate": 0.0001449919967710328, + "loss": 0.9346, + "step": 7909 + }, + { + "epoch": 1.4082977207977208, + "grad_norm": 0.7546970844268799, + "learning_rate": 0.00014497949562353242, + "loss": 1.0794, + "step": 7910 + }, + { + "epoch": 1.4084757834757835, + "grad_norm": 0.6175593733787537, + "learning_rate": 0.00014496699359475222, + "loss": 0.8939, + "step": 7911 + }, + { + "epoch": 1.4086538461538463, + "grad_norm": 0.6571716666221619, + "learning_rate": 0.00014495449068493722, + "loss": 1.1003, + "step": 7912 + }, + { + "epoch": 1.4088319088319088, + "grad_norm": 0.7038990259170532, + "learning_rate": 0.00014494198689433236, + "loss": 0.8844, + "step": 7913 + }, + { + "epoch": 1.4090099715099715, + "grad_norm": 0.7007337212562561, + "learning_rate": 0.00014492948222318263, + "loss": 1.2038, + "step": 7914 + }, + { + "epoch": 1.409188034188034, + "grad_norm": 0.7318591475486755, + "learning_rate": 0.00014491697667173302, + "loss": 1.0388, + "step": 7915 + }, + { + "epoch": 1.4093660968660968, + "grad_norm": 0.7010329961776733, + "learning_rate": 0.00014490447024022855, + "loss": 1.1485, + "step": 7916 + }, + { + "epoch": 1.4095441595441596, + "grad_norm": 0.7844831347465515, + "learning_rate": 0.0001448919629289143, + "loss": 1.1417, + "step": 7917 + }, + { + "epoch": 1.4097222222222223, + "grad_norm": 0.6953392624855042, + "learning_rate": 0.00014487945473803525, + "loss": 0.9546, + "step": 7918 + }, + { + "epoch": 1.4099002849002849, + "grad_norm": 0.6307587623596191, + "learning_rate": 0.00014486694566783655, + "loss": 0.9912, + "step": 7919 + }, + { + "epoch": 1.4100783475783476, + "grad_norm": 0.6200215816497803, + "learning_rate": 0.00014485443571856326, + "loss": 1.0998, + "step": 7920 + }, + { + "epoch": 1.4102564102564101, + "grad_norm": 0.7096502184867859, + "learning_rate": 0.00014484192489046043, + "loss": 0.9587, + "step": 7921 + }, + { + "epoch": 1.4104344729344729, + "grad_norm": 0.6965526342391968, + "learning_rate": 0.00014482941318377327, + "loss": 0.8791, + "step": 7922 + }, + { + "epoch": 1.4106125356125356, + "grad_norm": 0.7303466200828552, + "learning_rate": 0.00014481690059874687, + "loss": 1.084, + "step": 7923 + }, + { + "epoch": 1.4107905982905984, + "grad_norm": 0.6144066452980042, + "learning_rate": 0.00014480438713562638, + "loss": 0.9646, + "step": 7924 + }, + { + "epoch": 1.410968660968661, + "grad_norm": 0.645222008228302, + "learning_rate": 0.00014479187279465704, + "loss": 0.728, + "step": 7925 + }, + { + "epoch": 1.4111467236467237, + "grad_norm": 0.6069912314414978, + "learning_rate": 0.000144779357576084, + "loss": 0.842, + "step": 7926 + }, + { + "epoch": 1.4113247863247862, + "grad_norm": 0.6212135553359985, + "learning_rate": 0.00014476684148015243, + "loss": 0.9817, + "step": 7927 + }, + { + "epoch": 1.411502849002849, + "grad_norm": 0.6893343329429626, + "learning_rate": 0.00014475432450710763, + "loss": 1.0265, + "step": 7928 + }, + { + "epoch": 1.4116809116809117, + "grad_norm": 0.6842793822288513, + "learning_rate": 0.00014474180665719478, + "loss": 1.0593, + "step": 7929 + }, + { + "epoch": 1.4118589743589745, + "grad_norm": 0.74690842628479, + "learning_rate": 0.0001447292879306592, + "loss": 0.9096, + "step": 7930 + }, + { + "epoch": 1.412037037037037, + "grad_norm": 0.6624761819839478, + "learning_rate": 0.00014471676832774613, + "loss": 1.2244, + "step": 7931 + }, + { + "epoch": 1.4122150997150997, + "grad_norm": 0.6205778121948242, + "learning_rate": 0.00014470424784870088, + "loss": 1.1, + "step": 7932 + }, + { + "epoch": 1.4123931623931623, + "grad_norm": 0.7592337131500244, + "learning_rate": 0.00014469172649376875, + "loss": 0.963, + "step": 7933 + }, + { + "epoch": 1.412571225071225, + "grad_norm": 0.673328697681427, + "learning_rate": 0.00014467920426319508, + "loss": 0.8923, + "step": 7934 + }, + { + "epoch": 1.4127492877492878, + "grad_norm": 0.6064394116401672, + "learning_rate": 0.00014466668115722522, + "loss": 0.9679, + "step": 7935 + }, + { + "epoch": 1.4129273504273505, + "grad_norm": 0.7738677859306335, + "learning_rate": 0.00014465415717610454, + "loss": 1.0678, + "step": 7936 + }, + { + "epoch": 1.413105413105413, + "grad_norm": 0.7013397812843323, + "learning_rate": 0.00014464163232007836, + "loss": 0.9017, + "step": 7937 + }, + { + "epoch": 1.4132834757834758, + "grad_norm": 0.713291347026825, + "learning_rate": 0.0001446291065893922, + "loss": 1.1953, + "step": 7938 + }, + { + "epoch": 1.4134615384615383, + "grad_norm": 0.7538655996322632, + "learning_rate": 0.00014461657998429136, + "loss": 1.0571, + "step": 7939 + }, + { + "epoch": 1.413639601139601, + "grad_norm": 0.6358973383903503, + "learning_rate": 0.00014460405250502133, + "loss": 0.8552, + "step": 7940 + }, + { + "epoch": 1.4138176638176638, + "grad_norm": 0.67508864402771, + "learning_rate": 0.00014459152415182756, + "loss": 1.0293, + "step": 7941 + }, + { + "epoch": 1.4139957264957266, + "grad_norm": 0.7074598670005798, + "learning_rate": 0.00014457899492495546, + "loss": 1.2102, + "step": 7942 + }, + { + "epoch": 1.414173789173789, + "grad_norm": 0.7157037854194641, + "learning_rate": 0.00014456646482465058, + "loss": 1.0566, + "step": 7943 + }, + { + "epoch": 1.4143518518518519, + "grad_norm": 0.7918477058410645, + "learning_rate": 0.00014455393385115844, + "loss": 1.3727, + "step": 7944 + }, + { + "epoch": 1.4145299145299146, + "grad_norm": 0.569144606590271, + "learning_rate": 0.0001445414020047245, + "loss": 0.7251, + "step": 7945 + }, + { + "epoch": 1.4147079772079771, + "grad_norm": 0.7589054107666016, + "learning_rate": 0.0001445288692855943, + "loss": 1.0155, + "step": 7946 + }, + { + "epoch": 1.41488603988604, + "grad_norm": 0.7531685829162598, + "learning_rate": 0.0001445163356940134, + "loss": 0.8404, + "step": 7947 + }, + { + "epoch": 1.4150641025641026, + "grad_norm": 0.5730917453765869, + "learning_rate": 0.0001445038012302274, + "loss": 0.8215, + "step": 7948 + }, + { + "epoch": 1.4152421652421652, + "grad_norm": 0.6960710883140564, + "learning_rate": 0.00014449126589448187, + "loss": 0.7902, + "step": 7949 + }, + { + "epoch": 1.415420227920228, + "grad_norm": 0.8207054138183594, + "learning_rate": 0.0001444787296870224, + "loss": 1.493, + "step": 7950 + }, + { + "epoch": 1.4155982905982907, + "grad_norm": 0.5854668617248535, + "learning_rate": 0.00014446619260809462, + "loss": 0.9262, + "step": 7951 + }, + { + "epoch": 1.4157763532763532, + "grad_norm": 0.5458414554595947, + "learning_rate": 0.00014445365465794413, + "loss": 0.8431, + "step": 7952 + }, + { + "epoch": 1.415954415954416, + "grad_norm": 0.6880569458007812, + "learning_rate": 0.00014444111583681666, + "loss": 1.0184, + "step": 7953 + }, + { + "epoch": 1.4161324786324787, + "grad_norm": 0.6391083598136902, + "learning_rate": 0.00014442857614495783, + "loss": 0.88, + "step": 7954 + }, + { + "epoch": 1.4163105413105412, + "grad_norm": 0.6246135234832764, + "learning_rate": 0.00014441603558261335, + "loss": 0.776, + "step": 7955 + }, + { + "epoch": 1.416488603988604, + "grad_norm": 0.6263493895530701, + "learning_rate": 0.00014440349415002893, + "loss": 0.9069, + "step": 7956 + }, + { + "epoch": 1.4166666666666667, + "grad_norm": 0.7123475670814514, + "learning_rate": 0.00014439095184745024, + "loss": 0.8339, + "step": 7957 + }, + { + "epoch": 1.4168447293447293, + "grad_norm": 0.7171050906181335, + "learning_rate": 0.00014437840867512309, + "loss": 1.0633, + "step": 7958 + }, + { + "epoch": 1.417022792022792, + "grad_norm": 0.7097769975662231, + "learning_rate": 0.00014436586463329322, + "loss": 1.0852, + "step": 7959 + }, + { + "epoch": 1.4172008547008548, + "grad_norm": 0.6889223456382751, + "learning_rate": 0.00014435331972220637, + "loss": 0.916, + "step": 7960 + }, + { + "epoch": 1.4173789173789173, + "grad_norm": 0.6674435138702393, + "learning_rate": 0.0001443407739421084, + "loss": 0.9307, + "step": 7961 + }, + { + "epoch": 1.41755698005698, + "grad_norm": 0.6578894853591919, + "learning_rate": 0.00014432822729324503, + "loss": 0.8767, + "step": 7962 + }, + { + "epoch": 1.4177350427350428, + "grad_norm": 0.7145379781723022, + "learning_rate": 0.00014431567977586212, + "loss": 0.9962, + "step": 7963 + }, + { + "epoch": 1.4179131054131053, + "grad_norm": 0.6916680335998535, + "learning_rate": 0.00014430313139020555, + "loss": 1.0464, + "step": 7964 + }, + { + "epoch": 1.418091168091168, + "grad_norm": 0.6296181678771973, + "learning_rate": 0.00014429058213652116, + "loss": 1.0699, + "step": 7965 + }, + { + "epoch": 1.4182692307692308, + "grad_norm": 0.5640227198600769, + "learning_rate": 0.00014427803201505482, + "loss": 0.7006, + "step": 7966 + }, + { + "epoch": 1.4184472934472934, + "grad_norm": 0.7181212306022644, + "learning_rate": 0.0001442654810260524, + "loss": 1.1648, + "step": 7967 + }, + { + "epoch": 1.4186253561253561, + "grad_norm": 0.6830772757530212, + "learning_rate": 0.00014425292916975984, + "loss": 1.0641, + "step": 7968 + }, + { + "epoch": 1.4188034188034189, + "grad_norm": 0.665716290473938, + "learning_rate": 0.00014424037644642307, + "loss": 0.8769, + "step": 7969 + }, + { + "epoch": 1.4189814814814814, + "grad_norm": 0.8088666796684265, + "learning_rate": 0.00014422782285628802, + "loss": 1.1496, + "step": 7970 + }, + { + "epoch": 1.4191595441595442, + "grad_norm": 0.7186072468757629, + "learning_rate": 0.00014421526839960064, + "loss": 0.7421, + "step": 7971 + }, + { + "epoch": 1.419337606837607, + "grad_norm": 0.6405926942825317, + "learning_rate": 0.00014420271307660694, + "loss": 1.0139, + "step": 7972 + }, + { + "epoch": 1.4195156695156697, + "grad_norm": 0.7097104787826538, + "learning_rate": 0.0001441901568875529, + "loss": 1.1582, + "step": 7973 + }, + { + "epoch": 1.4196937321937322, + "grad_norm": 0.7347947359085083, + "learning_rate": 0.00014417759983268452, + "loss": 0.9751, + "step": 7974 + }, + { + "epoch": 1.419871794871795, + "grad_norm": 0.6999621987342834, + "learning_rate": 0.00014416504191224787, + "loss": 0.9419, + "step": 7975 + }, + { + "epoch": 1.4200498575498575, + "grad_norm": 0.6500616073608398, + "learning_rate": 0.00014415248312648897, + "loss": 0.9407, + "step": 7976 + }, + { + "epoch": 1.4202279202279202, + "grad_norm": 0.6368781328201294, + "learning_rate": 0.00014413992347565383, + "loss": 1.1224, + "step": 7977 + }, + { + "epoch": 1.420405982905983, + "grad_norm": 0.6422648429870605, + "learning_rate": 0.00014412736295998864, + "loss": 0.9573, + "step": 7978 + }, + { + "epoch": 1.4205840455840457, + "grad_norm": 0.744057297706604, + "learning_rate": 0.00014411480157973942, + "loss": 1.1384, + "step": 7979 + }, + { + "epoch": 1.4207621082621082, + "grad_norm": 0.5905839204788208, + "learning_rate": 0.00014410223933515232, + "loss": 0.8212, + "step": 7980 + }, + { + "epoch": 1.420940170940171, + "grad_norm": 0.5905438661575317, + "learning_rate": 0.0001440896762264734, + "loss": 0.8281, + "step": 7981 + }, + { + "epoch": 1.4211182336182335, + "grad_norm": 0.7087140679359436, + "learning_rate": 0.00014407711225394892, + "loss": 1.0165, + "step": 7982 + }, + { + "epoch": 1.4212962962962963, + "grad_norm": 0.6173902153968811, + "learning_rate": 0.00014406454741782495, + "loss": 0.8823, + "step": 7983 + }, + { + "epoch": 1.421474358974359, + "grad_norm": 0.6649761199951172, + "learning_rate": 0.00014405198171834772, + "loss": 0.9489, + "step": 7984 + }, + { + "epoch": 1.4216524216524218, + "grad_norm": 0.619286835193634, + "learning_rate": 0.00014403941515576344, + "loss": 0.8149, + "step": 7985 + }, + { + "epoch": 1.4218304843304843, + "grad_norm": 0.6358469724655151, + "learning_rate": 0.0001440268477303183, + "loss": 1.0558, + "step": 7986 + }, + { + "epoch": 1.422008547008547, + "grad_norm": 0.7239769697189331, + "learning_rate": 0.0001440142794422585, + "loss": 1.0528, + "step": 7987 + }, + { + "epoch": 1.4221866096866096, + "grad_norm": 0.681168794631958, + "learning_rate": 0.00014400171029183036, + "loss": 1.0867, + "step": 7988 + }, + { + "epoch": 1.4223646723646723, + "grad_norm": 0.6741157174110413, + "learning_rate": 0.0001439891402792801, + "loss": 0.9153, + "step": 7989 + }, + { + "epoch": 1.422542735042735, + "grad_norm": 0.5881659984588623, + "learning_rate": 0.00014397656940485403, + "loss": 0.92, + "step": 7990 + }, + { + "epoch": 1.4227207977207978, + "grad_norm": 0.637093722820282, + "learning_rate": 0.00014396399766879842, + "loss": 0.921, + "step": 7991 + }, + { + "epoch": 1.4228988603988604, + "grad_norm": 0.7760605216026306, + "learning_rate": 0.0001439514250713596, + "loss": 1.1451, + "step": 7992 + }, + { + "epoch": 1.4230769230769231, + "grad_norm": 0.6619600653648376, + "learning_rate": 0.00014393885161278393, + "loss": 1.0365, + "step": 7993 + }, + { + "epoch": 1.4232549857549857, + "grad_norm": 0.5354374051094055, + "learning_rate": 0.0001439262772933177, + "loss": 0.8718, + "step": 7994 + }, + { + "epoch": 1.4234330484330484, + "grad_norm": 0.7063560485839844, + "learning_rate": 0.00014391370211320735, + "loss": 0.8258, + "step": 7995 + }, + { + "epoch": 1.4236111111111112, + "grad_norm": 0.6876368522644043, + "learning_rate": 0.00014390112607269923, + "loss": 0.9579, + "step": 7996 + }, + { + "epoch": 1.423789173789174, + "grad_norm": 0.6976612210273743, + "learning_rate": 0.00014388854917203974, + "loss": 1.0376, + "step": 7997 + }, + { + "epoch": 1.4239672364672364, + "grad_norm": 0.6157355308532715, + "learning_rate": 0.00014387597141147525, + "loss": 0.8743, + "step": 7998 + }, + { + "epoch": 1.4241452991452992, + "grad_norm": 0.7273156046867371, + "learning_rate": 0.0001438633927912523, + "loss": 1.101, + "step": 7999 + }, + { + "epoch": 1.4243233618233617, + "grad_norm": 0.918380618095398, + "learning_rate": 0.0001438508133116173, + "loss": 0.9625, + "step": 8000 + }, + { + "epoch": 1.4245014245014245, + "grad_norm": 0.626040518283844, + "learning_rate": 0.00014383823297281666, + "loss": 0.9552, + "step": 8001 + }, + { + "epoch": 1.4246794871794872, + "grad_norm": 0.7320386171340942, + "learning_rate": 0.00014382565177509693, + "loss": 1.0719, + "step": 8002 + }, + { + "epoch": 1.42485754985755, + "grad_norm": 0.7283148169517517, + "learning_rate": 0.0001438130697187046, + "loss": 1.0455, + "step": 8003 + }, + { + "epoch": 1.4250356125356125, + "grad_norm": 0.6614177823066711, + "learning_rate": 0.00014380048680388613, + "loss": 0.9876, + "step": 8004 + }, + { + "epoch": 1.4252136752136753, + "grad_norm": 0.6726453900337219, + "learning_rate": 0.00014378790303088817, + "loss": 0.9861, + "step": 8005 + }, + { + "epoch": 1.4253917378917378, + "grad_norm": 0.7968725562095642, + "learning_rate": 0.00014377531839995718, + "loss": 1.1662, + "step": 8006 + }, + { + "epoch": 1.4255698005698005, + "grad_norm": 0.6510586738586426, + "learning_rate": 0.0001437627329113398, + "loss": 0.9452, + "step": 8007 + }, + { + "epoch": 1.4257478632478633, + "grad_norm": 0.6933155655860901, + "learning_rate": 0.00014375014656528253, + "loss": 1.0149, + "step": 8008 + }, + { + "epoch": 1.425925925925926, + "grad_norm": 0.7141832113265991, + "learning_rate": 0.00014373755936203204, + "loss": 1.0667, + "step": 8009 + }, + { + "epoch": 1.4261039886039886, + "grad_norm": 0.6352181434631348, + "learning_rate": 0.00014372497130183494, + "loss": 0.8652, + "step": 8010 + }, + { + "epoch": 1.4262820512820513, + "grad_norm": 0.7494860291481018, + "learning_rate": 0.00014371238238493786, + "loss": 0.9592, + "step": 8011 + }, + { + "epoch": 1.4264601139601139, + "grad_norm": 0.610556423664093, + "learning_rate": 0.00014369979261158746, + "loss": 0.7015, + "step": 8012 + }, + { + "epoch": 1.4266381766381766, + "grad_norm": 0.7305756211280823, + "learning_rate": 0.00014368720198203037, + "loss": 0.9681, + "step": 8013 + }, + { + "epoch": 1.4268162393162394, + "grad_norm": 0.6964020133018494, + "learning_rate": 0.0001436746104965133, + "loss": 1.1166, + "step": 8014 + }, + { + "epoch": 1.426994301994302, + "grad_norm": 0.7449237108230591, + "learning_rate": 0.00014366201815528302, + "loss": 1.1331, + "step": 8015 + }, + { + "epoch": 1.4271723646723646, + "grad_norm": 0.625834047794342, + "learning_rate": 0.00014364942495858615, + "loss": 0.8796, + "step": 8016 + }, + { + "epoch": 1.4273504273504274, + "grad_norm": 0.664559006690979, + "learning_rate": 0.0001436368309066695, + "loss": 1.0263, + "step": 8017 + }, + { + "epoch": 1.42752849002849, + "grad_norm": Infinity, + "learning_rate": 0.0001436368309066695, + "loss": 1.0731, + "step": 8018 + }, + { + "epoch": 1.4277065527065527, + "grad_norm": 0.6714464426040649, + "learning_rate": 0.00014362423599977977, + "loss": 0.9345, + "step": 8019 + }, + { + "epoch": 1.4278846153846154, + "grad_norm": 0.7595751285552979, + "learning_rate": 0.00014361164023816376, + "loss": 0.9646, + "step": 8020 + }, + { + "epoch": 1.4280626780626782, + "grad_norm": 0.6413954496383667, + "learning_rate": 0.00014359904362206828, + "loss": 1.0471, + "step": 8021 + }, + { + "epoch": 1.4282407407407407, + "grad_norm": 0.7298843264579773, + "learning_rate": 0.00014358644615174008, + "loss": 0.8932, + "step": 8022 + }, + { + "epoch": 1.4284188034188035, + "grad_norm": 0.8022156953811646, + "learning_rate": 0.00014357384782742602, + "loss": 1.0437, + "step": 8023 + }, + { + "epoch": 1.428596866096866, + "grad_norm": 0.7264443635940552, + "learning_rate": 0.00014356124864937296, + "loss": 0.9368, + "step": 8024 + }, + { + "epoch": 1.4287749287749287, + "grad_norm": 0.6819384098052979, + "learning_rate": 0.00014354864861782768, + "loss": 1.0, + "step": 8025 + }, + { + "epoch": 1.4289529914529915, + "grad_norm": 0.5945104956626892, + "learning_rate": 0.0001435360477330371, + "loss": 0.8108, + "step": 8026 + }, + { + "epoch": 1.4291310541310542, + "grad_norm": 0.6497398018836975, + "learning_rate": 0.0001435234459952481, + "loss": 0.8712, + "step": 8027 + }, + { + "epoch": 1.4293091168091168, + "grad_norm": 0.6424077749252319, + "learning_rate": 0.0001435108434047076, + "loss": 0.9172, + "step": 8028 + }, + { + "epoch": 1.4294871794871795, + "grad_norm": 0.6806963086128235, + "learning_rate": 0.00014349823996166253, + "loss": 1.1648, + "step": 8029 + }, + { + "epoch": 1.429665242165242, + "grad_norm": 0.6601083874702454, + "learning_rate": 0.00014348563566635977, + "loss": 0.9453, + "step": 8030 + }, + { + "epoch": 1.4298433048433048, + "grad_norm": 0.7024385929107666, + "learning_rate": 0.00014347303051904636, + "loss": 1.074, + "step": 8031 + }, + { + "epoch": 1.4300213675213675, + "grad_norm": 0.7094005942344666, + "learning_rate": 0.00014346042451996918, + "loss": 0.9976, + "step": 8032 + }, + { + "epoch": 1.4301994301994303, + "grad_norm": 0.6775936484336853, + "learning_rate": 0.0001434478176693753, + "loss": 0.9039, + "step": 8033 + }, + { + "epoch": 1.4303774928774928, + "grad_norm": 0.6920986771583557, + "learning_rate": 0.00014343520996751166, + "loss": 0.9122, + "step": 8034 + }, + { + "epoch": 1.4305555555555556, + "grad_norm": 0.720690906047821, + "learning_rate": 0.00014342260141462528, + "loss": 1.1028, + "step": 8035 + }, + { + "epoch": 1.430733618233618, + "grad_norm": 0.624546229839325, + "learning_rate": 0.00014340999201096328, + "loss": 0.9083, + "step": 8036 + }, + { + "epoch": 1.4309116809116809, + "grad_norm": 0.6560490727424622, + "learning_rate": 0.00014339738175677265, + "loss": 0.8029, + "step": 8037 + }, + { + "epoch": 1.4310897435897436, + "grad_norm": 0.8266100883483887, + "learning_rate": 0.00014338477065230047, + "loss": 0.9655, + "step": 8038 + }, + { + "epoch": 1.4312678062678064, + "grad_norm": 0.6593570113182068, + "learning_rate": 0.00014337215869779385, + "loss": 1.0299, + "step": 8039 + }, + { + "epoch": 1.431445868945869, + "grad_norm": 0.6321794390678406, + "learning_rate": 0.00014335954589349986, + "loss": 0.8755, + "step": 8040 + }, + { + "epoch": 1.4316239316239316, + "grad_norm": 0.7030870318412781, + "learning_rate": 0.00014334693223966562, + "loss": 1.1226, + "step": 8041 + }, + { + "epoch": 1.4318019943019942, + "grad_norm": 0.7794312238693237, + "learning_rate": 0.0001433343177365383, + "loss": 1.1252, + "step": 8042 + }, + { + "epoch": 1.431980056980057, + "grad_norm": 0.6115018129348755, + "learning_rate": 0.00014332170238436507, + "loss": 0.8753, + "step": 8043 + }, + { + "epoch": 1.4321581196581197, + "grad_norm": 0.8525674939155579, + "learning_rate": 0.00014330908618339304, + "loss": 0.9135, + "step": 8044 + }, + { + "epoch": 1.4323361823361824, + "grad_norm": 0.6869912147521973, + "learning_rate": 0.00014329646913386948, + "loss": 0.868, + "step": 8045 + }, + { + "epoch": 1.432514245014245, + "grad_norm": 0.5877542495727539, + "learning_rate": 0.0001432838512360415, + "loss": 0.9051, + "step": 8046 + }, + { + "epoch": 1.4326923076923077, + "grad_norm": 0.6609327793121338, + "learning_rate": 0.0001432712324901564, + "loss": 0.9084, + "step": 8047 + }, + { + "epoch": 1.4328703703703702, + "grad_norm": 0.6318345069885254, + "learning_rate": 0.0001432586128964614, + "loss": 0.8291, + "step": 8048 + }, + { + "epoch": 1.433048433048433, + "grad_norm": 0.6973567008972168, + "learning_rate": 0.0001432459924552037, + "loss": 0.97, + "step": 8049 + }, + { + "epoch": 1.4332264957264957, + "grad_norm": 0.6838201284408569, + "learning_rate": 0.00014323337116663062, + "loss": 1.0957, + "step": 8050 + }, + { + "epoch": 1.4334045584045585, + "grad_norm": 0.7472857236862183, + "learning_rate": 0.00014322074903098944, + "loss": 1.0981, + "step": 8051 + }, + { + "epoch": 1.433582621082621, + "grad_norm": 0.7723061442375183, + "learning_rate": 0.0001432081260485275, + "loss": 1.2231, + "step": 8052 + }, + { + "epoch": 1.4337606837606838, + "grad_norm": 0.681834876537323, + "learning_rate": 0.00014319550221949208, + "loss": 1.073, + "step": 8053 + }, + { + "epoch": 1.4339387464387463, + "grad_norm": 0.6566045880317688, + "learning_rate": 0.00014318287754413051, + "loss": 1.1298, + "step": 8054 + }, + { + "epoch": 1.434116809116809, + "grad_norm": 0.6792440414428711, + "learning_rate": 0.00014317025202269015, + "loss": 1.2224, + "step": 8055 + }, + { + "epoch": 1.4342948717948718, + "grad_norm": 0.7946709394454956, + "learning_rate": 0.00014315762565541838, + "loss": 1.0728, + "step": 8056 + }, + { + "epoch": 1.4344729344729346, + "grad_norm": 0.633466899394989, + "learning_rate": 0.00014314499844256262, + "loss": 0.944, + "step": 8057 + }, + { + "epoch": 1.434650997150997, + "grad_norm": 0.7308502197265625, + "learning_rate": 0.00014313237038437023, + "loss": 1.0684, + "step": 8058 + }, + { + "epoch": 1.4348290598290598, + "grad_norm": 0.6483737230300903, + "learning_rate": 0.00014311974148108862, + "loss": 1.0843, + "step": 8059 + }, + { + "epoch": 1.4350071225071226, + "grad_norm": 0.6301209926605225, + "learning_rate": 0.00014310711173296526, + "loss": 1.0083, + "step": 8060 + }, + { + "epoch": 1.4351851851851851, + "grad_norm": 0.6674302816390991, + "learning_rate": 0.00014309448114024757, + "loss": 0.9877, + "step": 8061 + }, + { + "epoch": 1.4353632478632479, + "grad_norm": 0.6888732314109802, + "learning_rate": 0.00014308184970318307, + "loss": 0.9937, + "step": 8062 + }, + { + "epoch": 1.4355413105413106, + "grad_norm": 0.6922950148582458, + "learning_rate": 0.00014306921742201923, + "loss": 1.0149, + "step": 8063 + }, + { + "epoch": 1.4357193732193732, + "grad_norm": 0.6050686240196228, + "learning_rate": 0.00014305658429700352, + "loss": 0.7882, + "step": 8064 + }, + { + "epoch": 1.435897435897436, + "grad_norm": 0.5080767869949341, + "learning_rate": 0.00014304395032838348, + "loss": 0.7796, + "step": 8065 + }, + { + "epoch": 1.4360754985754987, + "grad_norm": 0.6382707953453064, + "learning_rate": 0.00014303131551640668, + "loss": 0.965, + "step": 8066 + }, + { + "epoch": 1.4362535612535612, + "grad_norm": 0.7153477668762207, + "learning_rate": 0.00014301867986132063, + "loss": 1.1277, + "step": 8067 + }, + { + "epoch": 1.436431623931624, + "grad_norm": 0.6208404898643494, + "learning_rate": 0.00014300604336337292, + "loss": 0.8246, + "step": 8068 + }, + { + "epoch": 1.4366096866096867, + "grad_norm": 0.719695508480072, + "learning_rate": 0.0001429934060228111, + "loss": 0.7681, + "step": 8069 + }, + { + "epoch": 1.4367877492877492, + "grad_norm": 0.6219030618667603, + "learning_rate": 0.0001429807678398828, + "loss": 1.0425, + "step": 8070 + }, + { + "epoch": 1.436965811965812, + "grad_norm": 0.6080238819122314, + "learning_rate": 0.00014296812881483566, + "loss": 0.8762, + "step": 8071 + }, + { + "epoch": 1.4371438746438747, + "grad_norm": 0.6264194846153259, + "learning_rate": 0.00014295548894791729, + "loss": 1.087, + "step": 8072 + }, + { + "epoch": 1.4373219373219372, + "grad_norm": 0.6503600478172302, + "learning_rate": 0.00014294284823937535, + "loss": 1.0583, + "step": 8073 + }, + { + "epoch": 1.4375, + "grad_norm": 0.7623817324638367, + "learning_rate": 0.0001429302066894575, + "loss": 1.2372, + "step": 8074 + }, + { + "epoch": 1.4376780626780628, + "grad_norm": 0.7020344138145447, + "learning_rate": 0.00014291756429841144, + "loss": 1.2163, + "step": 8075 + }, + { + "epoch": 1.4378561253561253, + "grad_norm": 0.7070338129997253, + "learning_rate": 0.00014290492106648484, + "loss": 0.986, + "step": 8076 + }, + { + "epoch": 1.438034188034188, + "grad_norm": 0.6407621502876282, + "learning_rate": 0.00014289227699392545, + "loss": 0.9329, + "step": 8077 + }, + { + "epoch": 1.4382122507122508, + "grad_norm": 0.6836710572242737, + "learning_rate": 0.00014287963208098098, + "loss": 0.9252, + "step": 8078 + }, + { + "epoch": 1.4383903133903133, + "grad_norm": 0.648642897605896, + "learning_rate": 0.00014286698632789922, + "loss": 1.0457, + "step": 8079 + }, + { + "epoch": 1.438568376068376, + "grad_norm": 0.7015881538391113, + "learning_rate": 0.0001428543397349279, + "loss": 1.0516, + "step": 8080 + }, + { + "epoch": 1.4387464387464388, + "grad_norm": 0.6031532883644104, + "learning_rate": 0.0001428416923023148, + "loss": 0.9423, + "step": 8081 + }, + { + "epoch": 1.4389245014245013, + "grad_norm": 0.8235578536987305, + "learning_rate": 0.00014282904403030772, + "loss": 1.3433, + "step": 8082 + }, + { + "epoch": 1.439102564102564, + "grad_norm": 0.7355761528015137, + "learning_rate": 0.00014281639491915452, + "loss": 1.0128, + "step": 8083 + }, + { + "epoch": 1.4392806267806268, + "grad_norm": 0.7429629564285278, + "learning_rate": 0.00014280374496910303, + "loss": 0.8546, + "step": 8084 + }, + { + "epoch": 1.4394586894586894, + "grad_norm": 0.5831776857376099, + "learning_rate": 0.00014279109418040105, + "loss": 0.9021, + "step": 8085 + }, + { + "epoch": 1.4396367521367521, + "grad_norm": 0.6585184931755066, + "learning_rate": 0.00014277844255329645, + "loss": 0.9256, + "step": 8086 + }, + { + "epoch": 1.4398148148148149, + "grad_norm": 0.6412501931190491, + "learning_rate": 0.00014276579008803717, + "loss": 0.9305, + "step": 8087 + }, + { + "epoch": 1.4399928774928774, + "grad_norm": 0.6305423378944397, + "learning_rate": 0.00014275313678487102, + "loss": 0.9471, + "step": 8088 + }, + { + "epoch": 1.4401709401709402, + "grad_norm": 0.7160914540290833, + "learning_rate": 0.00014274048264404602, + "loss": 0.8798, + "step": 8089 + }, + { + "epoch": 1.440349002849003, + "grad_norm": 0.6740858554840088, + "learning_rate": 0.00014272782766581004, + "loss": 0.9022, + "step": 8090 + }, + { + "epoch": 1.4405270655270654, + "grad_norm": 0.7554821968078613, + "learning_rate": 0.000142715171850411, + "loss": 1.0924, + "step": 8091 + }, + { + "epoch": 1.4407051282051282, + "grad_norm": 0.7361162304878235, + "learning_rate": 0.00014270251519809694, + "loss": 0.9907, + "step": 8092 + }, + { + "epoch": 1.440883190883191, + "grad_norm": 0.731813371181488, + "learning_rate": 0.0001426898577091158, + "loss": 1.1765, + "step": 8093 + }, + { + "epoch": 1.4410612535612537, + "grad_norm": 0.6877756714820862, + "learning_rate": 0.00014267719938371558, + "loss": 1.0536, + "step": 8094 + }, + { + "epoch": 1.4412393162393162, + "grad_norm": 0.6724407076835632, + "learning_rate": 0.00014266454022214426, + "loss": 1.1895, + "step": 8095 + }, + { + "epoch": 1.441417378917379, + "grad_norm": 0.6946671605110168, + "learning_rate": 0.0001426518802246499, + "loss": 1.0437, + "step": 8096 + }, + { + "epoch": 1.4415954415954415, + "grad_norm": 0.7032839059829712, + "learning_rate": 0.00014263921939148058, + "loss": 1.1363, + "step": 8097 + }, + { + "epoch": 1.4417735042735043, + "grad_norm": 0.6942192316055298, + "learning_rate": 0.00014262655772288434, + "loss": 1.315, + "step": 8098 + }, + { + "epoch": 1.441951566951567, + "grad_norm": 0.7002301812171936, + "learning_rate": 0.00014261389521910922, + "loss": 1.0546, + "step": 8099 + }, + { + "epoch": 1.4421296296296298, + "grad_norm": 0.7260788083076477, + "learning_rate": 0.00014260123188040335, + "loss": 0.9374, + "step": 8100 + }, + { + "epoch": 1.4423076923076923, + "grad_norm": 0.6629201173782349, + "learning_rate": 0.00014258856770701486, + "loss": 0.8632, + "step": 8101 + }, + { + "epoch": 1.442485754985755, + "grad_norm": 0.6570318937301636, + "learning_rate": 0.0001425759026991918, + "loss": 1.0102, + "step": 8102 + }, + { + "epoch": 1.4426638176638176, + "grad_norm": 0.7696560621261597, + "learning_rate": 0.00014256323685718242, + "loss": 0.9703, + "step": 8103 + }, + { + "epoch": 1.4428418803418803, + "grad_norm": 0.7206611633300781, + "learning_rate": 0.00014255057018123482, + "loss": 1.1728, + "step": 8104 + }, + { + "epoch": 1.443019943019943, + "grad_norm": 0.6871611475944519, + "learning_rate": 0.0001425379026715972, + "loss": 0.9377, + "step": 8105 + }, + { + "epoch": 1.4431980056980058, + "grad_norm": 0.6027442812919617, + "learning_rate": 0.00014252523432851775, + "loss": 0.9212, + "step": 8106 + }, + { + "epoch": 1.4433760683760684, + "grad_norm": 0.7149752378463745, + "learning_rate": 0.00014251256515224463, + "loss": 0.9654, + "step": 8107 + }, + { + "epoch": 1.443554131054131, + "grad_norm": 0.5949522256851196, + "learning_rate": 0.00014249989514302614, + "loss": 1.0646, + "step": 8108 + }, + { + "epoch": 1.4437321937321936, + "grad_norm": 0.7345452904701233, + "learning_rate": 0.0001424872243011105, + "loss": 0.9801, + "step": 8109 + }, + { + "epoch": 1.4439102564102564, + "grad_norm": 0.8045009970664978, + "learning_rate": 0.00014247455262674592, + "loss": 1.3529, + "step": 8110 + }, + { + "epoch": 1.4440883190883191, + "grad_norm": 0.6712123155593872, + "learning_rate": 0.00014246188012018073, + "loss": 1.0416, + "step": 8111 + }, + { + "epoch": 1.444266381766382, + "grad_norm": 0.7811154127120972, + "learning_rate": 0.00014244920678166322, + "loss": 1.2019, + "step": 8112 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 0.6834486126899719, + "learning_rate": 0.00014243653261144167, + "loss": 0.986, + "step": 8113 + }, + { + "epoch": 1.4446225071225072, + "grad_norm": 0.6901041269302368, + "learning_rate": 0.00014242385760976443, + "loss": 1.0988, + "step": 8114 + }, + { + "epoch": 1.4448005698005697, + "grad_norm": 0.6233634948730469, + "learning_rate": 0.00014241118177687982, + "loss": 0.7748, + "step": 8115 + }, + { + "epoch": 1.4449786324786325, + "grad_norm": 0.6899837851524353, + "learning_rate": 0.00014239850511303624, + "loss": 0.9734, + "step": 8116 + }, + { + "epoch": 1.4451566951566952, + "grad_norm": 0.6316244006156921, + "learning_rate": 0.00014238582761848197, + "loss": 0.7888, + "step": 8117 + }, + { + "epoch": 1.445334757834758, + "grad_norm": 0.6074259877204895, + "learning_rate": 0.00014237314929346545, + "loss": 0.8843, + "step": 8118 + }, + { + "epoch": 1.4455128205128205, + "grad_norm": 0.6112192273139954, + "learning_rate": 0.00014236047013823516, + "loss": 0.8529, + "step": 8119 + }, + { + "epoch": 1.4456908831908832, + "grad_norm": 0.6883894801139832, + "learning_rate": 0.0001423477901530394, + "loss": 0.9506, + "step": 8120 + }, + { + "epoch": 1.4458689458689458, + "grad_norm": 0.7248309254646301, + "learning_rate": 0.00014233510933812666, + "loss": 0.9573, + "step": 8121 + }, + { + "epoch": 1.4460470085470085, + "grad_norm": 0.6853367686271667, + "learning_rate": 0.00014232242769374542, + "loss": 0.9903, + "step": 8122 + }, + { + "epoch": 1.4462250712250713, + "grad_norm": 0.7179274559020996, + "learning_rate": 0.0001423097452201441, + "loss": 0.9157, + "step": 8123 + }, + { + "epoch": 1.446403133903134, + "grad_norm": 0.6704817414283752, + "learning_rate": 0.00014229706191757127, + "loss": 1.1361, + "step": 8124 + }, + { + "epoch": 1.4465811965811965, + "grad_norm": 0.6380739212036133, + "learning_rate": 0.00014228437778627533, + "loss": 0.9336, + "step": 8125 + }, + { + "epoch": 1.4467592592592593, + "grad_norm": 0.6275362372398376, + "learning_rate": 0.00014227169282650487, + "loss": 0.9617, + "step": 8126 + }, + { + "epoch": 1.4469373219373218, + "grad_norm": 0.5644828677177429, + "learning_rate": 0.00014225900703850836, + "loss": 0.7384, + "step": 8127 + }, + { + "epoch": 1.4471153846153846, + "grad_norm": 0.6522284150123596, + "learning_rate": 0.00014224632042253443, + "loss": 1.1098, + "step": 8128 + }, + { + "epoch": 1.4472934472934473, + "grad_norm": 0.6228049993515015, + "learning_rate": 0.0001422336329788316, + "loss": 1.1061, + "step": 8129 + }, + { + "epoch": 1.44747150997151, + "grad_norm": 0.6092000603675842, + "learning_rate": 0.00014222094470764848, + "loss": 0.808, + "step": 8130 + }, + { + "epoch": 1.4476495726495726, + "grad_norm": 0.667435348033905, + "learning_rate": 0.00014220825560923363, + "loss": 1.1223, + "step": 8131 + }, + { + "epoch": 1.4478276353276354, + "grad_norm": 0.6080766320228577, + "learning_rate": 0.0001421955656838357, + "loss": 1.0099, + "step": 8132 + }, + { + "epoch": 1.448005698005698, + "grad_norm": 0.7597638368606567, + "learning_rate": 0.00014218287493170332, + "loss": 0.9718, + "step": 8133 + }, + { + "epoch": 1.4481837606837606, + "grad_norm": 0.574130654335022, + "learning_rate": 0.0001421701833530851, + "loss": 0.7745, + "step": 8134 + }, + { + "epoch": 1.4483618233618234, + "grad_norm": 0.6372822523117065, + "learning_rate": 0.0001421574909482298, + "loss": 1.0088, + "step": 8135 + }, + { + "epoch": 1.4485398860398861, + "grad_norm": 0.6759644746780396, + "learning_rate": 0.000142144797717386, + "loss": 0.9684, + "step": 8136 + }, + { + "epoch": 1.4487179487179487, + "grad_norm": 0.706351637840271, + "learning_rate": 0.00014213210366080244, + "loss": 1.021, + "step": 8137 + }, + { + "epoch": 1.4488960113960114, + "grad_norm": 0.6976894736289978, + "learning_rate": 0.0001421194087787278, + "loss": 1.1038, + "step": 8138 + }, + { + "epoch": 1.449074074074074, + "grad_norm": 0.7322551012039185, + "learning_rate": 0.00014210671307141092, + "loss": 1.0213, + "step": 8139 + }, + { + "epoch": 1.4492521367521367, + "grad_norm": 0.5885626077651978, + "learning_rate": 0.0001420940165391004, + "loss": 0.821, + "step": 8140 + }, + { + "epoch": 1.4494301994301995, + "grad_norm": 0.7009791135787964, + "learning_rate": 0.0001420813191820451, + "loss": 0.8647, + "step": 8141 + }, + { + "epoch": 1.4496082621082622, + "grad_norm": 0.5715423822402954, + "learning_rate": 0.00014206862100049375, + "loss": 0.873, + "step": 8142 + }, + { + "epoch": 1.4497863247863247, + "grad_norm": 1.1452178955078125, + "learning_rate": 0.00014205592199469514, + "loss": 1.2523, + "step": 8143 + }, + { + "epoch": 1.4499643874643875, + "grad_norm": 0.8076814413070679, + "learning_rate": 0.00014204322216489814, + "loss": 1.1071, + "step": 8144 + }, + { + "epoch": 1.45014245014245, + "grad_norm": 0.7325751185417175, + "learning_rate": 0.00014203052151135154, + "loss": 0.9846, + "step": 8145 + }, + { + "epoch": 1.4503205128205128, + "grad_norm": 0.7009061574935913, + "learning_rate": 0.00014201782003430417, + "loss": 0.8153, + "step": 8146 + }, + { + "epoch": 1.4504985754985755, + "grad_norm": 0.6502353549003601, + "learning_rate": 0.0001420051177340049, + "loss": 0.8959, + "step": 8147 + }, + { + "epoch": 1.4506766381766383, + "grad_norm": 0.6134430170059204, + "learning_rate": 0.00014199241461070261, + "loss": 0.9683, + "step": 8148 + }, + { + "epoch": 1.4508547008547008, + "grad_norm": 0.720160722732544, + "learning_rate": 0.0001419797106646462, + "loss": 0.9579, + "step": 8149 + }, + { + "epoch": 1.4510327635327636, + "grad_norm": 0.6141422986984253, + "learning_rate": 0.00014196700589608454, + "loss": 0.9427, + "step": 8150 + }, + { + "epoch": 1.451210826210826, + "grad_norm": 0.6835139393806458, + "learning_rate": 0.00014195430030526656, + "loss": 1.0374, + "step": 8151 + }, + { + "epoch": 1.4513888888888888, + "grad_norm": 0.6829691529273987, + "learning_rate": 0.00014194159389244128, + "loss": 0.9418, + "step": 8152 + }, + { + "epoch": 1.4515669515669516, + "grad_norm": 0.7142195701599121, + "learning_rate": 0.00014192888665785755, + "loss": 1.1876, + "step": 8153 + }, + { + "epoch": 1.4517450142450143, + "grad_norm": 0.6719943284988403, + "learning_rate": 0.0001419161786017644, + "loss": 1.1417, + "step": 8154 + }, + { + "epoch": 1.4519230769230769, + "grad_norm": 0.6478939652442932, + "learning_rate": 0.0001419034697244108, + "loss": 0.943, + "step": 8155 + }, + { + "epoch": 1.4521011396011396, + "grad_norm": 0.6308888792991638, + "learning_rate": 0.00014189076002604575, + "loss": 0.9842, + "step": 8156 + }, + { + "epoch": 1.4522792022792022, + "grad_norm": 0.673559844493866, + "learning_rate": 0.00014187804950691827, + "loss": 0.8108, + "step": 8157 + }, + { + "epoch": 1.452457264957265, + "grad_norm": 0.5895359516143799, + "learning_rate": 0.00014186533816727744, + "loss": 0.8187, + "step": 8158 + }, + { + "epoch": 1.4526353276353277, + "grad_norm": 0.6703287363052368, + "learning_rate": 0.00014185262600737225, + "loss": 0.9012, + "step": 8159 + }, + { + "epoch": 1.4528133903133904, + "grad_norm": 0.697728157043457, + "learning_rate": 0.00014183991302745182, + "loss": 1.2572, + "step": 8160 + }, + { + "epoch": 1.452991452991453, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00014182719922776514, + "loss": 1.078, + "step": 8161 + }, + { + "epoch": 1.4531695156695157, + "grad_norm": 0.6774863600730896, + "learning_rate": 0.00014181448460856143, + "loss": 1.0607, + "step": 8162 + }, + { + "epoch": 1.4533475783475782, + "grad_norm": 0.6872009038925171, + "learning_rate": 0.00014180176917008976, + "loss": 1.0713, + "step": 8163 + }, + { + "epoch": 1.453525641025641, + "grad_norm": 0.7949981093406677, + "learning_rate": 0.00014178905291259926, + "loss": 1.0471, + "step": 8164 + }, + { + "epoch": 1.4537037037037037, + "grad_norm": 0.6592127084732056, + "learning_rate": 0.00014177633583633908, + "loss": 0.8409, + "step": 8165 + }, + { + "epoch": 1.4538817663817665, + "grad_norm": 0.6745635867118835, + "learning_rate": 0.00014176361794155837, + "loss": 1.0859, + "step": 8166 + }, + { + "epoch": 1.454059829059829, + "grad_norm": 0.6661605834960938, + "learning_rate": 0.00014175089922850633, + "loss": 1.0587, + "step": 8167 + }, + { + "epoch": 1.4542378917378918, + "grad_norm": 0.6697571873664856, + "learning_rate": 0.00014173817969743212, + "loss": 0.8876, + "step": 8168 + }, + { + "epoch": 1.4544159544159543, + "grad_norm": 0.6162588000297546, + "learning_rate": 0.000141725459348585, + "loss": 0.9575, + "step": 8169 + }, + { + "epoch": 1.454594017094017, + "grad_norm": 0.6235088109970093, + "learning_rate": 0.00014171273818221422, + "loss": 0.9209, + "step": 8170 + }, + { + "epoch": 1.4547720797720798, + "grad_norm": 0.6744212508201599, + "learning_rate": 0.00014170001619856896, + "loss": 0.9704, + "step": 8171 + }, + { + "epoch": 1.4549501424501425, + "grad_norm": 0.6781345009803772, + "learning_rate": 0.0001416872933978985, + "loss": 1.1507, + "step": 8172 + }, + { + "epoch": 1.455128205128205, + "grad_norm": 0.7160060405731201, + "learning_rate": 0.0001416745697804521, + "loss": 1.2529, + "step": 8173 + }, + { + "epoch": 1.4553062678062678, + "grad_norm": 0.6742389798164368, + "learning_rate": 0.00014166184534647913, + "loss": 1.0168, + "step": 8174 + }, + { + "epoch": 1.4554843304843303, + "grad_norm": 0.6685828566551208, + "learning_rate": 0.0001416491200962288, + "loss": 1.0807, + "step": 8175 + }, + { + "epoch": 1.455662393162393, + "grad_norm": 0.6998327374458313, + "learning_rate": 0.0001416363940299505, + "loss": 1.1711, + "step": 8176 + }, + { + "epoch": 1.4558404558404558, + "grad_norm": 0.7132518291473389, + "learning_rate": 0.00014162366714789358, + "loss": 1.1392, + "step": 8177 + }, + { + "epoch": 1.4560185185185186, + "grad_norm": 0.6995887160301208, + "learning_rate": 0.0001416109394503073, + "loss": 1.3335, + "step": 8178 + }, + { + "epoch": 1.4561965811965811, + "grad_norm": 0.7161234021186829, + "learning_rate": 0.00014159821093744115, + "loss": 0.9725, + "step": 8179 + }, + { + "epoch": 1.4563746438746439, + "grad_norm": 0.7678874135017395, + "learning_rate": 0.00014158548160954446, + "loss": 1.1578, + "step": 8180 + }, + { + "epoch": 1.4565527065527066, + "grad_norm": 0.67372065782547, + "learning_rate": 0.00014157275146686662, + "loss": 1.0867, + "step": 8181 + }, + { + "epoch": 1.4567307692307692, + "grad_norm": 0.7757831811904907, + "learning_rate": 0.00014156002050965712, + "loss": 0.9768, + "step": 8182 + }, + { + "epoch": 1.456908831908832, + "grad_norm": 0.7174801230430603, + "learning_rate": 0.00014154728873816533, + "loss": 1.1712, + "step": 8183 + }, + { + "epoch": 1.4570868945868947, + "grad_norm": 0.5972673892974854, + "learning_rate": 0.0001415345561526407, + "loss": 0.9571, + "step": 8184 + }, + { + "epoch": 1.4572649572649572, + "grad_norm": 0.7999650835990906, + "learning_rate": 0.00014152182275333275, + "loss": 1.0583, + "step": 8185 + }, + { + "epoch": 1.45744301994302, + "grad_norm": 0.6737848520278931, + "learning_rate": 0.00014150908854049091, + "loss": 1.0562, + "step": 8186 + }, + { + "epoch": 1.4576210826210827, + "grad_norm": 0.7756418585777283, + "learning_rate": 0.00014149635351436474, + "loss": 1.2301, + "step": 8187 + }, + { + "epoch": 1.4577991452991452, + "grad_norm": 0.5633914470672607, + "learning_rate": 0.00014148361767520374, + "loss": 0.8847, + "step": 8188 + }, + { + "epoch": 1.457977207977208, + "grad_norm": 0.8462759256362915, + "learning_rate": 0.00014147088102325737, + "loss": 0.8046, + "step": 8189 + }, + { + "epoch": 1.4581552706552707, + "grad_norm": 0.7081632614135742, + "learning_rate": 0.00014145814355877526, + "loss": 1.0764, + "step": 8190 + }, + { + "epoch": 1.4583333333333333, + "grad_norm": 0.7357106804847717, + "learning_rate": 0.00014144540528200698, + "loss": 1.0202, + "step": 8191 + }, + { + "epoch": 1.458511396011396, + "grad_norm": 0.603566586971283, + "learning_rate": 0.00014143266619320204, + "loss": 0.8214, + "step": 8192 + }, + { + "epoch": 1.4586894586894588, + "grad_norm": 0.6829110383987427, + "learning_rate": 0.00014141992629261007, + "loss": 0.9479, + "step": 8193 + }, + { + "epoch": 1.4588675213675213, + "grad_norm": 0.6822739839553833, + "learning_rate": 0.00014140718558048072, + "loss": 0.9117, + "step": 8194 + }, + { + "epoch": 1.459045584045584, + "grad_norm": 0.7383607029914856, + "learning_rate": 0.00014139444405706356, + "loss": 0.9819, + "step": 8195 + }, + { + "epoch": 1.4592236467236468, + "grad_norm": 0.6319897770881653, + "learning_rate": 0.00014138170172260826, + "loss": 1.0508, + "step": 8196 + }, + { + "epoch": 1.4594017094017093, + "grad_norm": 0.6804461479187012, + "learning_rate": 0.0001413689585773645, + "loss": 0.992, + "step": 8197 + }, + { + "epoch": 1.459579772079772, + "grad_norm": 0.6198720335960388, + "learning_rate": 0.0001413562146215819, + "loss": 1.0113, + "step": 8198 + }, + { + "epoch": 1.4597578347578348, + "grad_norm": 0.5968540906906128, + "learning_rate": 0.0001413434698555102, + "loss": 0.7562, + "step": 8199 + }, + { + "epoch": 1.4599358974358974, + "grad_norm": 0.5370334982872009, + "learning_rate": 0.00014133072427939913, + "loss": 0.9238, + "step": 8200 + }, + { + "epoch": 1.46011396011396, + "grad_norm": 0.6652548909187317, + "learning_rate": 0.00014131797789349832, + "loss": 0.9464, + "step": 8201 + }, + { + "epoch": 1.4602920227920229, + "grad_norm": 0.637852668762207, + "learning_rate": 0.00014130523069805757, + "loss": 1.0395, + "step": 8202 + }, + { + "epoch": 1.4604700854700854, + "grad_norm": 0.8186550140380859, + "learning_rate": 0.00014129248269332664, + "loss": 1.2116, + "step": 8203 + }, + { + "epoch": 1.4606481481481481, + "grad_norm": 0.5290196537971497, + "learning_rate": 0.00014127973387955528, + "loss": 0.7331, + "step": 8204 + }, + { + "epoch": 1.460826210826211, + "grad_norm": 0.6516342163085938, + "learning_rate": 0.00014126698425699332, + "loss": 0.9275, + "step": 8205 + }, + { + "epoch": 1.4610042735042734, + "grad_norm": 0.767254114151001, + "learning_rate": 0.00014125423382589048, + "loss": 0.9355, + "step": 8206 + }, + { + "epoch": 1.4611823361823362, + "grad_norm": 0.6476777195930481, + "learning_rate": 0.00014124148258649668, + "loss": 0.9263, + "step": 8207 + }, + { + "epoch": 1.461360398860399, + "grad_norm": 0.6737871766090393, + "learning_rate": 0.00014122873053906167, + "loss": 0.9815, + "step": 8208 + }, + { + "epoch": 1.4615384615384617, + "grad_norm": 0.6311159729957581, + "learning_rate": 0.00014121597768383532, + "loss": 0.9607, + "step": 8209 + }, + { + "epoch": 1.4617165242165242, + "grad_norm": 0.6061250567436218, + "learning_rate": 0.00014120322402106752, + "loss": 0.7428, + "step": 8210 + }, + { + "epoch": 1.461894586894587, + "grad_norm": 0.6916252970695496, + "learning_rate": 0.00014119046955100815, + "loss": 0.9664, + "step": 8211 + }, + { + "epoch": 1.4620726495726495, + "grad_norm": 0.6583660840988159, + "learning_rate": 0.00014117771427390706, + "loss": 1.0645, + "step": 8212 + }, + { + "epoch": 1.4622507122507122, + "grad_norm": 0.7034604549407959, + "learning_rate": 0.00014116495819001425, + "loss": 0.9223, + "step": 8213 + }, + { + "epoch": 1.462428774928775, + "grad_norm": 0.6378605961799622, + "learning_rate": 0.00014115220129957954, + "loss": 0.7963, + "step": 8214 + }, + { + "epoch": 1.4626068376068377, + "grad_norm": 0.6251596212387085, + "learning_rate": 0.00014113944360285297, + "loss": 0.9852, + "step": 8215 + }, + { + "epoch": 1.4627849002849003, + "grad_norm": 0.7055560946464539, + "learning_rate": 0.00014112668510008446, + "loss": 0.9342, + "step": 8216 + }, + { + "epoch": 1.462962962962963, + "grad_norm": 0.6250377893447876, + "learning_rate": 0.00014111392579152396, + "loss": 0.9886, + "step": 8217 + }, + { + "epoch": 1.4631410256410255, + "grad_norm": 0.6011185050010681, + "learning_rate": 0.00014110116567742152, + "loss": 0.8465, + "step": 8218 + }, + { + "epoch": 1.4633190883190883, + "grad_norm": 0.6632489562034607, + "learning_rate": 0.0001410884047580271, + "loss": 0.8619, + "step": 8219 + }, + { + "epoch": 1.463497150997151, + "grad_norm": 0.7194828987121582, + "learning_rate": 0.00014107564303359076, + "loss": 1.1231, + "step": 8220 + }, + { + "epoch": 1.4636752136752138, + "grad_norm": 0.7640393376350403, + "learning_rate": 0.0001410628805043625, + "loss": 1.1955, + "step": 8221 + }, + { + "epoch": 1.4638532763532763, + "grad_norm": 0.9118906259536743, + "learning_rate": 0.0001410501171705924, + "loss": 1.0555, + "step": 8222 + }, + { + "epoch": 1.464031339031339, + "grad_norm": 0.7545066475868225, + "learning_rate": 0.00014103735303253053, + "loss": 0.9425, + "step": 8223 + }, + { + "epoch": 1.4642094017094016, + "grad_norm": 0.6848801970481873, + "learning_rate": 0.000141024588090427, + "loss": 1.0418, + "step": 8224 + }, + { + "epoch": 1.4643874643874644, + "grad_norm": 0.6825160384178162, + "learning_rate": 0.00014101182234453185, + "loss": 0.9615, + "step": 8225 + }, + { + "epoch": 1.4645655270655271, + "grad_norm": 0.8258556723594666, + "learning_rate": 0.00014099905579509527, + "loss": 1.1237, + "step": 8226 + }, + { + "epoch": 1.4647435897435899, + "grad_norm": 0.6427522897720337, + "learning_rate": 0.00014098628844236733, + "loss": 1.0853, + "step": 8227 + }, + { + "epoch": 1.4649216524216524, + "grad_norm": 0.6476351022720337, + "learning_rate": 0.00014097352028659825, + "loss": 1.1286, + "step": 8228 + }, + { + "epoch": 1.4650997150997151, + "grad_norm": 0.7621034383773804, + "learning_rate": 0.00014096075132803812, + "loss": 1.1402, + "step": 8229 + }, + { + "epoch": 1.4652777777777777, + "grad_norm": 0.6629892587661743, + "learning_rate": 0.00014094798156693718, + "loss": 0.7108, + "step": 8230 + }, + { + "epoch": 1.4654558404558404, + "grad_norm": 0.6902043223381042, + "learning_rate": 0.00014093521100354557, + "loss": 1.1761, + "step": 8231 + }, + { + "epoch": 1.4656339031339032, + "grad_norm": 0.7422910928726196, + "learning_rate": 0.00014092243963811357, + "loss": 0.867, + "step": 8232 + }, + { + "epoch": 1.465811965811966, + "grad_norm": 0.7424963712692261, + "learning_rate": 0.00014090966747089137, + "loss": 1.015, + "step": 8233 + }, + { + "epoch": 1.4659900284900285, + "grad_norm": 0.6855891942977905, + "learning_rate": 0.0001408968945021292, + "loss": 0.9624, + "step": 8234 + }, + { + "epoch": 1.4661680911680912, + "grad_norm": 0.5968918204307556, + "learning_rate": 0.00014088412073207736, + "loss": 0.9243, + "step": 8235 + }, + { + "epoch": 1.4663461538461537, + "grad_norm": 0.6153344511985779, + "learning_rate": 0.0001408713461609861, + "loss": 1.0305, + "step": 8236 + }, + { + "epoch": 1.4665242165242165, + "grad_norm": 0.6627458333969116, + "learning_rate": 0.0001408585707891057, + "loss": 1.1102, + "step": 8237 + }, + { + "epoch": 1.4667022792022792, + "grad_norm": 0.6475233435630798, + "learning_rate": 0.0001408457946166865, + "loss": 1.0045, + "step": 8238 + }, + { + "epoch": 1.466880341880342, + "grad_norm": 0.6792858839035034, + "learning_rate": 0.00014083301764397876, + "loss": 1.0092, + "step": 8239 + }, + { + "epoch": 1.4670584045584045, + "grad_norm": 0.6916255354881287, + "learning_rate": 0.00014082023987123293, + "loss": 1.0761, + "step": 8240 + }, + { + "epoch": 1.4672364672364673, + "grad_norm": 0.7901251912117004, + "learning_rate": 0.00014080746129869923, + "loss": 0.8002, + "step": 8241 + }, + { + "epoch": 1.4674145299145298, + "grad_norm": 0.8078263401985168, + "learning_rate": 0.00014079468192662812, + "loss": 0.9738, + "step": 8242 + }, + { + "epoch": 1.4675925925925926, + "grad_norm": 0.6370784640312195, + "learning_rate": 0.00014078190175526996, + "loss": 1.0256, + "step": 8243 + }, + { + "epoch": 1.4677706552706553, + "grad_norm": 0.6087532639503479, + "learning_rate": 0.0001407691207848752, + "loss": 0.9747, + "step": 8244 + }, + { + "epoch": 1.467948717948718, + "grad_norm": 0.6333357691764832, + "learning_rate": 0.00014075633901569414, + "loss": 1.0135, + "step": 8245 + }, + { + "epoch": 1.4681267806267806, + "grad_norm": 0.6914255619049072, + "learning_rate": 0.00014074355644797733, + "loss": 1.0261, + "step": 8246 + }, + { + "epoch": 1.4683048433048433, + "grad_norm": 0.6374734044075012, + "learning_rate": 0.00014073077308197513, + "loss": 0.9197, + "step": 8247 + }, + { + "epoch": 1.4684829059829059, + "grad_norm": 0.8023789525032043, + "learning_rate": 0.00014071798891793807, + "loss": 1.1085, + "step": 8248 + }, + { + "epoch": 1.4686609686609686, + "grad_norm": 0.7722933888435364, + "learning_rate": 0.0001407052039561166, + "loss": 1.2018, + "step": 8249 + }, + { + "epoch": 1.4688390313390314, + "grad_norm": 0.6823393106460571, + "learning_rate": 0.0001406924181967612, + "loss": 1.088, + "step": 8250 + }, + { + "epoch": 1.4690170940170941, + "grad_norm": 0.7037357687950134, + "learning_rate": 0.00014067963164012242, + "loss": 1.0324, + "step": 8251 + }, + { + "epoch": 1.4691951566951567, + "grad_norm": 0.6549737453460693, + "learning_rate": 0.00014066684428645074, + "loss": 1.152, + "step": 8252 + }, + { + "epoch": 1.4693732193732194, + "grad_norm": 0.5349790453910828, + "learning_rate": 0.00014065405613599674, + "loss": 0.6996, + "step": 8253 + }, + { + "epoch": 1.469551282051282, + "grad_norm": 0.6760679483413696, + "learning_rate": 0.00014064126718901096, + "loss": 0.9856, + "step": 8254 + }, + { + "epoch": 1.4697293447293447, + "grad_norm": 0.5912436842918396, + "learning_rate": 0.00014062847744574395, + "loss": 1.0076, + "step": 8255 + }, + { + "epoch": 1.4699074074074074, + "grad_norm": 0.75101637840271, + "learning_rate": 0.00014061568690644632, + "loss": 1.0033, + "step": 8256 + }, + { + "epoch": 1.4700854700854702, + "grad_norm": 0.6233504414558411, + "learning_rate": 0.00014060289557136873, + "loss": 0.8525, + "step": 8257 + }, + { + "epoch": 1.4702635327635327, + "grad_norm": 0.659570038318634, + "learning_rate": 0.00014059010344076171, + "loss": 0.855, + "step": 8258 + }, + { + "epoch": 1.4704415954415955, + "grad_norm": 0.8096539974212646, + "learning_rate": 0.00014057731051487593, + "loss": 0.9905, + "step": 8259 + }, + { + "epoch": 1.470619658119658, + "grad_norm": 0.5829728245735168, + "learning_rate": 0.00014056451679396204, + "loss": 0.7974, + "step": 8260 + }, + { + "epoch": 1.4707977207977208, + "grad_norm": 0.6176979541778564, + "learning_rate": 0.0001405517222782707, + "loss": 0.9556, + "step": 8261 + }, + { + "epoch": 1.4709757834757835, + "grad_norm": 0.6322479248046875, + "learning_rate": 0.00014053892696805264, + "loss": 0.8837, + "step": 8262 + }, + { + "epoch": 1.4711538461538463, + "grad_norm": 0.6886917948722839, + "learning_rate": 0.0001405261308635585, + "loss": 0.9242, + "step": 8263 + }, + { + "epoch": 1.4713319088319088, + "grad_norm": 0.7474521994590759, + "learning_rate": 0.00014051333396503901, + "loss": 0.9906, + "step": 8264 + }, + { + "epoch": 1.4715099715099715, + "grad_norm": 0.7120978832244873, + "learning_rate": 0.00014050053627274488, + "loss": 1.1074, + "step": 8265 + }, + { + "epoch": 1.471688034188034, + "grad_norm": 0.6778998374938965, + "learning_rate": 0.0001404877377869269, + "loss": 1.0027, + "step": 8266 + }, + { + "epoch": 1.4718660968660968, + "grad_norm": 0.6832901239395142, + "learning_rate": 0.0001404749385078358, + "loss": 0.9399, + "step": 8267 + }, + { + "epoch": 1.4720441595441596, + "grad_norm": 0.7428423762321472, + "learning_rate": 0.00014046213843572236, + "loss": 1.0591, + "step": 8268 + }, + { + "epoch": 1.4722222222222223, + "grad_norm": 0.7522720098495483, + "learning_rate": 0.00014044933757083737, + "loss": 1.1184, + "step": 8269 + }, + { + "epoch": 1.4724002849002849, + "grad_norm": 0.7714734673500061, + "learning_rate": 0.00014043653591343163, + "loss": 1.0783, + "step": 8270 + }, + { + "epoch": 1.4725783475783476, + "grad_norm": 0.5860890746116638, + "learning_rate": 0.00014042373346375597, + "loss": 0.8394, + "step": 8271 + }, + { + "epoch": 1.4727564102564101, + "grad_norm": 0.6400395035743713, + "learning_rate": 0.0001404109302220612, + "loss": 0.9153, + "step": 8272 + }, + { + "epoch": 1.4729344729344729, + "grad_norm": 0.7441139817237854, + "learning_rate": 0.00014039812618859827, + "loss": 0.9224, + "step": 8273 + }, + { + "epoch": 1.4731125356125356, + "grad_norm": 0.6030932664871216, + "learning_rate": 0.00014038532136361793, + "loss": 1.0783, + "step": 8274 + }, + { + "epoch": 1.4732905982905984, + "grad_norm": 0.7243345975875854, + "learning_rate": 0.0001403725157473711, + "loss": 0.9894, + "step": 8275 + }, + { + "epoch": 1.473468660968661, + "grad_norm": 0.6880641579627991, + "learning_rate": 0.0001403597093401087, + "loss": 0.9459, + "step": 8276 + }, + { + "epoch": 1.4736467236467237, + "grad_norm": 0.6263882517814636, + "learning_rate": 0.00014034690214208165, + "loss": 0.8781, + "step": 8277 + }, + { + "epoch": 1.4738247863247862, + "grad_norm": 0.7159495949745178, + "learning_rate": 0.00014033409415354085, + "loss": 1.0511, + "step": 8278 + }, + { + "epoch": 1.474002849002849, + "grad_norm": 0.7182226181030273, + "learning_rate": 0.00014032128537473727, + "loss": 1.1196, + "step": 8279 + }, + { + "epoch": 1.4741809116809117, + "grad_norm": 0.744478166103363, + "learning_rate": 0.00014030847580592186, + "loss": 1.0747, + "step": 8280 + }, + { + "epoch": 1.4743589743589745, + "grad_norm": 0.6806797385215759, + "learning_rate": 0.00014029566544734558, + "loss": 1.1519, + "step": 8281 + }, + { + "epoch": 1.474537037037037, + "grad_norm": 0.6813502311706543, + "learning_rate": 0.00014028285429925946, + "loss": 0.968, + "step": 8282 + }, + { + "epoch": 1.4747150997150997, + "grad_norm": 0.639784574508667, + "learning_rate": 0.00014027004236191452, + "loss": 1.0685, + "step": 8283 + }, + { + "epoch": 1.4748931623931623, + "grad_norm": 0.6325878500938416, + "learning_rate": 0.00014025722963556173, + "loss": 1.0358, + "step": 8284 + }, + { + "epoch": 1.475071225071225, + "grad_norm": 0.7012955546379089, + "learning_rate": 0.00014024441612045215, + "loss": 1.1059, + "step": 8285 + }, + { + "epoch": 1.4752492877492878, + "grad_norm": 0.690380334854126, + "learning_rate": 0.00014023160181683684, + "loss": 0.9628, + "step": 8286 + }, + { + "epoch": 1.4754273504273505, + "grad_norm": 0.7178516983985901, + "learning_rate": 0.00014021878672496686, + "loss": 0.963, + "step": 8287 + }, + { + "epoch": 1.475605413105413, + "grad_norm": 0.7049064636230469, + "learning_rate": 0.0001402059708450933, + "loss": 0.8996, + "step": 8288 + }, + { + "epoch": 1.4757834757834758, + "grad_norm": 0.6777819395065308, + "learning_rate": 0.00014019315417746728, + "loss": 1.0696, + "step": 8289 + }, + { + "epoch": 1.4759615384615383, + "grad_norm": 0.5948763489723206, + "learning_rate": 0.00014018033672233987, + "loss": 0.928, + "step": 8290 + }, + { + "epoch": 1.476139601139601, + "grad_norm": 0.7183942198753357, + "learning_rate": 0.00014016751847996224, + "loss": 1.1053, + "step": 8291 + }, + { + "epoch": 1.4763176638176638, + "grad_norm": 0.7426177263259888, + "learning_rate": 0.00014015469945058556, + "loss": 0.9504, + "step": 8292 + }, + { + "epoch": 1.4764957264957266, + "grad_norm": 0.6508159041404724, + "learning_rate": 0.0001401418796344609, + "loss": 1.1176, + "step": 8293 + }, + { + "epoch": 1.476673789173789, + "grad_norm": 0.6954567432403564, + "learning_rate": 0.00014012905903183954, + "loss": 0.9238, + "step": 8294 + }, + { + "epoch": 1.4768518518518519, + "grad_norm": 0.7023960947990417, + "learning_rate": 0.0001401162376429726, + "loss": 1.2032, + "step": 8295 + }, + { + "epoch": 1.4770299145299146, + "grad_norm": 0.7174739837646484, + "learning_rate": 0.00014010341546811134, + "loss": 0.9385, + "step": 8296 + }, + { + "epoch": 1.4772079772079771, + "grad_norm": 0.611980140209198, + "learning_rate": 0.00014009059250750695, + "loss": 0.9469, + "step": 8297 + }, + { + "epoch": 1.47738603988604, + "grad_norm": 0.6362917423248291, + "learning_rate": 0.0001400777687614107, + "loss": 1.1406, + "step": 8298 + }, + { + "epoch": 1.4775641025641026, + "grad_norm": 0.6884697675704956, + "learning_rate": 0.00014006494423007381, + "loss": 0.7915, + "step": 8299 + }, + { + "epoch": 1.4777421652421652, + "grad_norm": 0.6266025304794312, + "learning_rate": 0.00014005211891374755, + "loss": 0.94, + "step": 8300 + }, + { + "epoch": 1.477920227920228, + "grad_norm": 0.6130280494689941, + "learning_rate": 0.00014003929281268323, + "loss": 0.9369, + "step": 8301 + }, + { + "epoch": 1.4780982905982907, + "grad_norm": 0.7244207859039307, + "learning_rate": 0.00014002646592713215, + "loss": 1.1449, + "step": 8302 + }, + { + "epoch": 1.4782763532763532, + "grad_norm": 0.6527345776557922, + "learning_rate": 0.0001400136382573456, + "loss": 0.7792, + "step": 8303 + }, + { + "epoch": 1.478454415954416, + "grad_norm": 0.7102689743041992, + "learning_rate": 0.00014000080980357496, + "loss": 0.9577, + "step": 8304 + }, + { + "epoch": 1.4786324786324787, + "grad_norm": 0.6179325580596924, + "learning_rate": 0.00013998798056607154, + "loss": 0.827, + "step": 8305 + }, + { + "epoch": 1.4788105413105412, + "grad_norm": 0.761234700679779, + "learning_rate": 0.00013997515054508668, + "loss": 1.0576, + "step": 8306 + }, + { + "epoch": 1.478988603988604, + "grad_norm": 0.6200914978981018, + "learning_rate": 0.0001399623197408718, + "loss": 1.0514, + "step": 8307 + }, + { + "epoch": 1.4791666666666667, + "grad_norm": 0.5961193442344666, + "learning_rate": 0.0001399494881536783, + "loss": 0.7846, + "step": 8308 + }, + { + "epoch": 1.4793447293447293, + "grad_norm": 0.645984411239624, + "learning_rate": 0.00013993665578375758, + "loss": 0.9927, + "step": 8309 + }, + { + "epoch": 1.479522792022792, + "grad_norm": 0.7258989810943604, + "learning_rate": 0.000139923822631361, + "loss": 0.7567, + "step": 8310 + }, + { + "epoch": 1.4797008547008548, + "grad_norm": 0.708882212638855, + "learning_rate": 0.00013991098869674007, + "loss": 1.1147, + "step": 8311 + }, + { + "epoch": 1.4798789173789173, + "grad_norm": 0.669262707233429, + "learning_rate": 0.00013989815398014624, + "loss": 0.7142, + "step": 8312 + }, + { + "epoch": 1.48005698005698, + "grad_norm": 0.7398767471313477, + "learning_rate": 0.00013988531848183096, + "loss": 1.043, + "step": 8313 + }, + { + "epoch": 1.4802350427350428, + "grad_norm": 0.753197193145752, + "learning_rate": 0.0001398724822020457, + "loss": 1.058, + "step": 8314 + }, + { + "epoch": 1.4804131054131053, + "grad_norm": 0.663526177406311, + "learning_rate": 0.000139859645141042, + "loss": 1.1272, + "step": 8315 + }, + { + "epoch": 1.480591168091168, + "grad_norm": 0.6537514925003052, + "learning_rate": 0.00013984680729907135, + "loss": 1.011, + "step": 8316 + }, + { + "epoch": 1.4807692307692308, + "grad_norm": 0.707554817199707, + "learning_rate": 0.00013983396867638527, + "loss": 1.0593, + "step": 8317 + }, + { + "epoch": 1.4809472934472934, + "grad_norm": 0.6261475086212158, + "learning_rate": 0.00013982112927323533, + "loss": 1.0731, + "step": 8318 + }, + { + "epoch": 1.4811253561253561, + "grad_norm": 0.6694258451461792, + "learning_rate": 0.00013980828908987308, + "loss": 1.0703, + "step": 8319 + }, + { + "epoch": 1.4813034188034189, + "grad_norm": 0.7793164253234863, + "learning_rate": 0.00013979544812655012, + "loss": 1.0447, + "step": 8320 + }, + { + "epoch": 1.4814814814814814, + "grad_norm": 0.6496448516845703, + "learning_rate": 0.00013978260638351802, + "loss": 1.0208, + "step": 8321 + }, + { + "epoch": 1.4816595441595442, + "grad_norm": 0.5992059111595154, + "learning_rate": 0.00013976976386102834, + "loss": 0.9717, + "step": 8322 + }, + { + "epoch": 1.481837606837607, + "grad_norm": 0.7473567128181458, + "learning_rate": 0.0001397569205593328, + "loss": 0.9612, + "step": 8323 + }, + { + "epoch": 1.4820156695156697, + "grad_norm": 0.657558798789978, + "learning_rate": 0.00013974407647868297, + "loss": 1.2137, + "step": 8324 + }, + { + "epoch": 1.4821937321937322, + "grad_norm": 0.7040614485740662, + "learning_rate": 0.00013973123161933055, + "loss": 1.007, + "step": 8325 + }, + { + "epoch": 1.482371794871795, + "grad_norm": 0.6098681092262268, + "learning_rate": 0.00013971838598152717, + "loss": 1.0595, + "step": 8326 + }, + { + "epoch": 1.4825498575498575, + "grad_norm": 0.7194869518280029, + "learning_rate": 0.0001397055395655245, + "loss": 0.9632, + "step": 8327 + }, + { + "epoch": 1.4827279202279202, + "grad_norm": 0.645972728729248, + "learning_rate": 0.00013969269237157426, + "loss": 1.0712, + "step": 8328 + }, + { + "epoch": 1.482905982905983, + "grad_norm": 0.6580560207366943, + "learning_rate": 0.0001396798443999282, + "loss": 1.2117, + "step": 8329 + }, + { + "epoch": 1.4830840455840457, + "grad_norm": 0.6624418497085571, + "learning_rate": 0.00013966699565083802, + "loss": 0.8529, + "step": 8330 + }, + { + "epoch": 1.4832621082621082, + "grad_norm": 0.659896731376648, + "learning_rate": 0.00013965414612455545, + "loss": 0.9359, + "step": 8331 + }, + { + "epoch": 1.483440170940171, + "grad_norm": 0.6690883636474609, + "learning_rate": 0.00013964129582133222, + "loss": 0.971, + "step": 8332 + }, + { + "epoch": 1.4836182336182335, + "grad_norm": 0.6767334938049316, + "learning_rate": 0.00013962844474142022, + "loss": 1.0137, + "step": 8333 + }, + { + "epoch": 1.4837962962962963, + "grad_norm": 0.6412752270698547, + "learning_rate": 0.0001396155928850711, + "loss": 1.2812, + "step": 8334 + }, + { + "epoch": 1.483974358974359, + "grad_norm": 0.6731469035148621, + "learning_rate": 0.0001396027402525368, + "loss": 0.8723, + "step": 8335 + }, + { + "epoch": 1.4841524216524218, + "grad_norm": 0.7327923774719238, + "learning_rate": 0.000139589886844069, + "loss": 0.9606, + "step": 8336 + }, + { + "epoch": 1.4843304843304843, + "grad_norm": 0.6194515824317932, + "learning_rate": 0.00013957703265991963, + "loss": 0.8514, + "step": 8337 + }, + { + "epoch": 1.484508547008547, + "grad_norm": 0.7250012755393982, + "learning_rate": 0.00013956417770034053, + "loss": 0.9755, + "step": 8338 + }, + { + "epoch": 1.4846866096866096, + "grad_norm": 0.7484263181686401, + "learning_rate": 0.00013955132196558358, + "loss": 1.0376, + "step": 8339 + }, + { + "epoch": 1.4848646723646723, + "grad_norm": 0.7593362331390381, + "learning_rate": 0.00013953846545590058, + "loss": 1.3011, + "step": 8340 + }, + { + "epoch": 1.485042735042735, + "grad_norm": 0.6670466065406799, + "learning_rate": 0.00013952560817154352, + "loss": 0.9726, + "step": 8341 + }, + { + "epoch": 1.4852207977207978, + "grad_norm": 0.8001134395599365, + "learning_rate": 0.00013951275011276425, + "loss": 1.1447, + "step": 8342 + }, + { + "epoch": 1.4853988603988604, + "grad_norm": 0.741450309753418, + "learning_rate": 0.00013949989127981475, + "loss": 1.1101, + "step": 8343 + }, + { + "epoch": 1.4855769230769231, + "grad_norm": 0.6594467163085938, + "learning_rate": 0.00013948703167294694, + "loss": 1.0205, + "step": 8344 + }, + { + "epoch": 1.4857549857549857, + "grad_norm": 0.6303030252456665, + "learning_rate": 0.00013947417129241276, + "loss": 0.9179, + "step": 8345 + }, + { + "epoch": 1.4859330484330484, + "grad_norm": 0.6352720856666565, + "learning_rate": 0.00013946131013846418, + "loss": 1.158, + "step": 8346 + }, + { + "epoch": 1.4861111111111112, + "grad_norm": 0.6720923781394958, + "learning_rate": 0.0001394484482113532, + "loss": 0.8805, + "step": 8347 + }, + { + "epoch": 1.486289173789174, + "grad_norm": 0.7186421751976013, + "learning_rate": 0.00013943558551133186, + "loss": 0.8951, + "step": 8348 + }, + { + "epoch": 1.4864672364672364, + "grad_norm": 0.6038698554039001, + "learning_rate": 0.00013942272203865214, + "loss": 1.0079, + "step": 8349 + }, + { + "epoch": 1.4866452991452992, + "grad_norm": 0.665790319442749, + "learning_rate": 0.00013940985779356606, + "loss": 0.8853, + "step": 8350 + }, + { + "epoch": 1.4868233618233617, + "grad_norm": 0.6941595673561096, + "learning_rate": 0.00013939699277632568, + "loss": 1.1404, + "step": 8351 + }, + { + "epoch": 1.4870014245014245, + "grad_norm": 0.7943871021270752, + "learning_rate": 0.00013938412698718305, + "loss": 0.9961, + "step": 8352 + }, + { + "epoch": 1.4871794871794872, + "grad_norm": 0.6363818645477295, + "learning_rate": 0.00013937126042639028, + "loss": 0.8621, + "step": 8353 + }, + { + "epoch": 1.48735754985755, + "grad_norm": 0.7986421585083008, + "learning_rate": 0.00013935839309419943, + "loss": 1.0547, + "step": 8354 + }, + { + "epoch": 1.4875356125356125, + "grad_norm": 0.5890130400657654, + "learning_rate": 0.00013934552499086266, + "loss": 0.9863, + "step": 8355 + }, + { + "epoch": 1.4877136752136753, + "grad_norm": 0.7915370464324951, + "learning_rate": 0.00013933265611663207, + "loss": 1.0385, + "step": 8356 + }, + { + "epoch": 1.4878917378917378, + "grad_norm": 0.7062503695487976, + "learning_rate": 0.00013931978647175973, + "loss": 1.0984, + "step": 8357 + }, + { + "epoch": 1.4880698005698005, + "grad_norm": 0.6496769785881042, + "learning_rate": 0.00013930691605649792, + "loss": 1.0884, + "step": 8358 + }, + { + "epoch": 1.4882478632478633, + "grad_norm": 0.6527266502380371, + "learning_rate": 0.0001392940448710987, + "loss": 1.0366, + "step": 8359 + }, + { + "epoch": 1.488425925925926, + "grad_norm": 0.6269870400428772, + "learning_rate": 0.00013928117291581431, + "loss": 0.9097, + "step": 8360 + }, + { + "epoch": 1.4886039886039886, + "grad_norm": 0.6581160426139832, + "learning_rate": 0.00013926830019089694, + "loss": 0.8694, + "step": 8361 + }, + { + "epoch": 1.4887820512820513, + "grad_norm": 0.6196219325065613, + "learning_rate": 0.0001392554266965988, + "loss": 0.8054, + "step": 8362 + }, + { + "epoch": 1.4889601139601139, + "grad_norm": 0.6246176362037659, + "learning_rate": 0.0001392425524331721, + "loss": 0.9309, + "step": 8363 + }, + { + "epoch": 1.4891381766381766, + "grad_norm": 0.7293874025344849, + "learning_rate": 0.00013922967740086914, + "loss": 1.051, + "step": 8364 + }, + { + "epoch": 1.4893162393162394, + "grad_norm": 0.6581604480743408, + "learning_rate": 0.00013921680159994213, + "loss": 0.8475, + "step": 8365 + }, + { + "epoch": 1.489494301994302, + "grad_norm": 0.6294612288475037, + "learning_rate": 0.00013920392503064335, + "loss": 0.6946, + "step": 8366 + }, + { + "epoch": 1.4896723646723646, + "grad_norm": 0.5725370645523071, + "learning_rate": 0.00013919104769322512, + "loss": 0.7838, + "step": 8367 + }, + { + "epoch": 1.4898504273504274, + "grad_norm": 0.681520402431488, + "learning_rate": 0.00013917816958793967, + "loss": 0.99, + "step": 8368 + }, + { + "epoch": 1.49002849002849, + "grad_norm": 0.6660219430923462, + "learning_rate": 0.00013916529071503943, + "loss": 0.9113, + "step": 8369 + }, + { + "epoch": 1.4902065527065527, + "grad_norm": 0.7567862272262573, + "learning_rate": 0.00013915241107477665, + "loss": 1.2498, + "step": 8370 + }, + { + "epoch": 1.4903846153846154, + "grad_norm": 0.7366036176681519, + "learning_rate": 0.00013913953066740372, + "loss": 1.115, + "step": 8371 + }, + { + "epoch": 1.4905626780626782, + "grad_norm": 0.6201434135437012, + "learning_rate": 0.00013912664949317297, + "loss": 0.8447, + "step": 8372 + }, + { + "epoch": 1.4907407407407407, + "grad_norm": 0.7618655562400818, + "learning_rate": 0.00013911376755233683, + "loss": 0.9696, + "step": 8373 + }, + { + "epoch": 1.4909188034188035, + "grad_norm": 0.6716726422309875, + "learning_rate": 0.00013910088484514764, + "loss": 0.9753, + "step": 8374 + }, + { + "epoch": 1.491096866096866, + "grad_norm": 0.6745659112930298, + "learning_rate": 0.0001390880013718579, + "loss": 1.134, + "step": 8375 + }, + { + "epoch": 1.4912749287749287, + "grad_norm": 0.7524410486221313, + "learning_rate": 0.0001390751171327199, + "loss": 1.0235, + "step": 8376 + }, + { + "epoch": 1.4914529914529915, + "grad_norm": 0.7409411072731018, + "learning_rate": 0.00013906223212798615, + "loss": 0.752, + "step": 8377 + }, + { + "epoch": 1.4916310541310542, + "grad_norm": 0.7016384601593018, + "learning_rate": 0.00013904934635790913, + "loss": 1.1712, + "step": 8378 + }, + { + "epoch": 1.4918091168091168, + "grad_norm": 0.6537824869155884, + "learning_rate": 0.00013903645982274129, + "loss": 1.1162, + "step": 8379 + }, + { + "epoch": 1.4919871794871795, + "grad_norm": 0.6460806727409363, + "learning_rate": 0.0001390235725227351, + "loss": 0.9389, + "step": 8380 + }, + { + "epoch": 1.492165242165242, + "grad_norm": 0.6405501365661621, + "learning_rate": 0.0001390106844581431, + "loss": 1.0508, + "step": 8381 + }, + { + "epoch": 1.4923433048433048, + "grad_norm": 0.6672594547271729, + "learning_rate": 0.00013899779562921775, + "loss": 1.0018, + "step": 8382 + }, + { + "epoch": 1.4925213675213675, + "grad_norm": 0.6303185820579529, + "learning_rate": 0.0001389849060362116, + "loss": 0.9964, + "step": 8383 + }, + { + "epoch": 1.4926994301994303, + "grad_norm": 0.6981508731842041, + "learning_rate": 0.00013897201567937719, + "loss": 1.174, + "step": 8384 + }, + { + "epoch": 1.4928774928774928, + "grad_norm": 0.6195989847183228, + "learning_rate": 0.0001389591245589671, + "loss": 0.9254, + "step": 8385 + }, + { + "epoch": 1.4930555555555556, + "grad_norm": 0.6232163310050964, + "learning_rate": 0.00013894623267523393, + "loss": 0.7151, + "step": 8386 + }, + { + "epoch": 1.493233618233618, + "grad_norm": 0.673067033290863, + "learning_rate": 0.0001389333400284302, + "loss": 1.0156, + "step": 8387 + }, + { + "epoch": 1.4934116809116809, + "grad_norm": 0.706266462802887, + "learning_rate": 0.00013892044661880856, + "loss": 0.9387, + "step": 8388 + }, + { + "epoch": 1.4935897435897436, + "grad_norm": 0.742640495300293, + "learning_rate": 0.00013890755244662161, + "loss": 1.1597, + "step": 8389 + }, + { + "epoch": 1.4937678062678064, + "grad_norm": 0.6856846809387207, + "learning_rate": 0.000138894657512122, + "loss": 0.9998, + "step": 8390 + }, + { + "epoch": 1.493945868945869, + "grad_norm": 0.7214110493659973, + "learning_rate": 0.0001388817618155624, + "loss": 1.1867, + "step": 8391 + }, + { + "epoch": 1.4941239316239316, + "grad_norm": 0.7346787452697754, + "learning_rate": 0.0001388688653571954, + "loss": 0.9071, + "step": 8392 + }, + { + "epoch": 1.4943019943019942, + "grad_norm": 0.7019181847572327, + "learning_rate": 0.00013885596813727373, + "loss": 1.0472, + "step": 8393 + }, + { + "epoch": 1.494480056980057, + "grad_norm": 0.6780814528465271, + "learning_rate": 0.00013884307015605012, + "loss": 1.0031, + "step": 8394 + }, + { + "epoch": 1.4946581196581197, + "grad_norm": 0.6722873449325562, + "learning_rate": 0.0001388301714137772, + "loss": 0.8889, + "step": 8395 + }, + { + "epoch": 1.4948361823361824, + "grad_norm": 0.6736134886741638, + "learning_rate": 0.00013881727191070777, + "loss": 0.8695, + "step": 8396 + }, + { + "epoch": 1.495014245014245, + "grad_norm": 0.632648766040802, + "learning_rate": 0.00013880437164709452, + "loss": 0.9391, + "step": 8397 + }, + { + "epoch": 1.4951923076923077, + "grad_norm": 0.7004299163818359, + "learning_rate": 0.0001387914706231902, + "loss": 1.1423, + "step": 8398 + }, + { + "epoch": 1.4953703703703702, + "grad_norm": 0.5787134766578674, + "learning_rate": 0.0001387785688392476, + "loss": 0.9953, + "step": 8399 + }, + { + "epoch": 1.495548433048433, + "grad_norm": 0.6671785712242126, + "learning_rate": 0.0001387656662955195, + "loss": 0.9356, + "step": 8400 + }, + { + "epoch": 1.4957264957264957, + "grad_norm": 0.7216096520423889, + "learning_rate": 0.0001387527629922587, + "loss": 0.9065, + "step": 8401 + }, + { + "epoch": 1.4959045584045585, + "grad_norm": 0.6469849348068237, + "learning_rate": 0.00013873985892971801, + "loss": 1.0664, + "step": 8402 + }, + { + "epoch": 1.496082621082621, + "grad_norm": 0.5598217248916626, + "learning_rate": 0.00013872695410815027, + "loss": 0.8834, + "step": 8403 + }, + { + "epoch": 1.4962606837606838, + "grad_norm": 0.6860302686691284, + "learning_rate": 0.00013871404852780828, + "loss": 0.9061, + "step": 8404 + }, + { + "epoch": 1.4964387464387463, + "grad_norm": 0.7101688385009766, + "learning_rate": 0.00013870114218894497, + "loss": 1.0236, + "step": 8405 + }, + { + "epoch": 1.496616809116809, + "grad_norm": 0.6494225859642029, + "learning_rate": 0.00013868823509181313, + "loss": 0.9631, + "step": 8406 + }, + { + "epoch": 1.4967948717948718, + "grad_norm": 0.6804189085960388, + "learning_rate": 0.00013867532723666574, + "loss": 0.9341, + "step": 8407 + }, + { + "epoch": 1.4969729344729346, + "grad_norm": 0.8493942022323608, + "learning_rate": 0.00013866241862375562, + "loss": 1.1451, + "step": 8408 + }, + { + "epoch": 1.497150997150997, + "grad_norm": 0.6248497366905212, + "learning_rate": 0.00013864950925333576, + "loss": 0.8584, + "step": 8409 + }, + { + "epoch": 1.4973290598290598, + "grad_norm": 0.6238769292831421, + "learning_rate": 0.00013863659912565903, + "loss": 1.1612, + "step": 8410 + }, + { + "epoch": 1.4975071225071226, + "grad_norm": 0.8538609147071838, + "learning_rate": 0.0001386236882409784, + "loss": 1.0817, + "step": 8411 + }, + { + "epoch": 1.4976851851851851, + "grad_norm": 0.7301406264305115, + "learning_rate": 0.00013861077659954683, + "loss": 0.943, + "step": 8412 + }, + { + "epoch": 1.4978632478632479, + "grad_norm": 0.6573456525802612, + "learning_rate": 0.0001385978642016173, + "loss": 1.0154, + "step": 8413 + }, + { + "epoch": 1.4980413105413106, + "grad_norm": 0.7634185552597046, + "learning_rate": 0.0001385849510474428, + "loss": 1.0432, + "step": 8414 + }, + { + "epoch": 1.4982193732193732, + "grad_norm": 0.6156686544418335, + "learning_rate": 0.00013857203713727633, + "loss": 1.0442, + "step": 8415 + }, + { + "epoch": 1.498397435897436, + "grad_norm": 0.5386871695518494, + "learning_rate": 0.00013855912247137092, + "loss": 0.9055, + "step": 8416 + }, + { + "epoch": 1.4985754985754987, + "grad_norm": 0.7108574509620667, + "learning_rate": 0.00013854620704997962, + "loss": 0.9705, + "step": 8417 + }, + { + "epoch": 1.4987535612535612, + "grad_norm": 0.7313347458839417, + "learning_rate": 0.00013853329087335547, + "loss": 0.7541, + "step": 8418 + }, + { + "epoch": 1.498931623931624, + "grad_norm": 0.8369119167327881, + "learning_rate": 0.0001385203739417515, + "loss": 1.1317, + "step": 8419 + }, + { + "epoch": 1.4991096866096867, + "grad_norm": 0.6763789057731628, + "learning_rate": 0.00013850745625542085, + "loss": 0.7909, + "step": 8420 + }, + { + "epoch": 1.4992877492877492, + "grad_norm": 0.7369635105133057, + "learning_rate": 0.00013849453781461656, + "loss": 1.1454, + "step": 8421 + }, + { + "epoch": 1.499465811965812, + "grad_norm": 0.7165971398353577, + "learning_rate": 0.0001384816186195918, + "loss": 1.1927, + "step": 8422 + }, + { + "epoch": 1.4996438746438747, + "grad_norm": 0.7502337694168091, + "learning_rate": 0.00013846869867059966, + "loss": 1.0592, + "step": 8423 + }, + { + "epoch": 1.4998219373219372, + "grad_norm": 0.7207813858985901, + "learning_rate": 0.00013845577796789326, + "loss": 1.1133, + "step": 8424 + }, + { + "epoch": 1.4998219373219372, + "eval_loss": 1.1057652235031128, + "eval_runtime": 24.7975, + "eval_samples_per_second": 41.98, + "eval_steps_per_second": 21.01, + "step": 8424 + }, + { + "epoch": 1.5, + "grad_norm": 0.6962727308273315, + "learning_rate": 0.00013844285651172576, + "loss": 1.0711, + "step": 8425 + }, + { + "epoch": 1.5001780626780628, + "grad_norm": 0.6585133075714111, + "learning_rate": 0.00013842993430235038, + "loss": 0.9793, + "step": 8426 + }, + { + "epoch": 1.5003561253561255, + "grad_norm": 0.7045056819915771, + "learning_rate": 0.00013841701134002029, + "loss": 1.0046, + "step": 8427 + }, + { + "epoch": 1.500534188034188, + "grad_norm": 0.6788702011108398, + "learning_rate": 0.00013840408762498863, + "loss": 0.9539, + "step": 8428 + }, + { + "epoch": 1.5007122507122506, + "grad_norm": 0.7253114581108093, + "learning_rate": 0.00013839116315750863, + "loss": 0.9446, + "step": 8429 + }, + { + "epoch": 1.5008903133903133, + "grad_norm": 0.6103765368461609, + "learning_rate": 0.0001383782379378336, + "loss": 0.7862, + "step": 8430 + }, + { + "epoch": 1.501068376068376, + "grad_norm": 0.6662353873252869, + "learning_rate": 0.00013836531196621666, + "loss": 1.2178, + "step": 8431 + }, + { + "epoch": 1.5012464387464388, + "grad_norm": 0.6871803998947144, + "learning_rate": 0.00013835238524291117, + "loss": 0.9263, + "step": 8432 + }, + { + "epoch": 1.5014245014245016, + "grad_norm": 0.62713223695755, + "learning_rate": 0.00013833945776817034, + "loss": 0.8879, + "step": 8433 + }, + { + "epoch": 1.501602564102564, + "grad_norm": 0.6698164343833923, + "learning_rate": 0.00013832652954224748, + "loss": 0.9847, + "step": 8434 + }, + { + "epoch": 1.5017806267806266, + "grad_norm": 0.6855883002281189, + "learning_rate": 0.0001383136005653959, + "loss": 0.8614, + "step": 8435 + }, + { + "epoch": 1.5019586894586894, + "grad_norm": 0.7028802037239075, + "learning_rate": 0.0001383006708378689, + "loss": 1.0153, + "step": 8436 + }, + { + "epoch": 1.5021367521367521, + "grad_norm": 0.6710380911827087, + "learning_rate": 0.00013828774035991981, + "loss": 1.0163, + "step": 8437 + }, + { + "epoch": 1.5023148148148149, + "grad_norm": 0.618984580039978, + "learning_rate": 0.000138274809131802, + "loss": 1.0015, + "step": 8438 + }, + { + "epoch": 1.5024928774928776, + "grad_norm": 0.6881645321846008, + "learning_rate": 0.00013826187715376882, + "loss": 0.9776, + "step": 8439 + }, + { + "epoch": 1.5026709401709402, + "grad_norm": 0.6715859770774841, + "learning_rate": 0.00013824894442607358, + "loss": 0.9129, + "step": 8440 + }, + { + "epoch": 1.5028490028490027, + "grad_norm": 0.5940943360328674, + "learning_rate": 0.0001382360109489698, + "loss": 1.0724, + "step": 8441 + }, + { + "epoch": 1.5030270655270654, + "grad_norm": 0.6536458134651184, + "learning_rate": 0.0001382230767227108, + "loss": 1.0162, + "step": 8442 + }, + { + "epoch": 1.5032051282051282, + "grad_norm": 0.6163156628608704, + "learning_rate": 0.00013821014174755, + "loss": 1.0521, + "step": 8443 + }, + { + "epoch": 1.503383190883191, + "grad_norm": 0.7592282891273499, + "learning_rate": 0.00013819720602374082, + "loss": 0.9525, + "step": 8444 + }, + { + "epoch": 1.5035612535612537, + "grad_norm": 0.6672595143318176, + "learning_rate": 0.0001381842695515368, + "loss": 0.9359, + "step": 8445 + }, + { + "epoch": 1.5037393162393162, + "grad_norm": 0.6395034193992615, + "learning_rate": 0.0001381713323311913, + "loss": 1.166, + "step": 8446 + }, + { + "epoch": 1.5039173789173788, + "grad_norm": 0.5958148837089539, + "learning_rate": 0.00013815839436295783, + "loss": 0.9885, + "step": 8447 + }, + { + "epoch": 1.5040954415954415, + "grad_norm": 0.676555871963501, + "learning_rate": 0.0001381454556470899, + "loss": 1.0637, + "step": 8448 + }, + { + "epoch": 1.5042735042735043, + "grad_norm": 0.642428994178772, + "learning_rate": 0.00013813251618384102, + "loss": 0.9288, + "step": 8449 + }, + { + "epoch": 1.504451566951567, + "grad_norm": 0.6730920076370239, + "learning_rate": 0.00013811957597346467, + "loss": 1.1345, + "step": 8450 + }, + { + "epoch": 1.5046296296296298, + "grad_norm": 0.7824259996414185, + "learning_rate": 0.00013810663501621443, + "loss": 0.7532, + "step": 8451 + }, + { + "epoch": 1.5048076923076923, + "grad_norm": 0.8184825778007507, + "learning_rate": 0.00013809369331234386, + "loss": 1.2674, + "step": 8452 + }, + { + "epoch": 1.5049857549857548, + "grad_norm": 0.7369286417961121, + "learning_rate": 0.00013808075086210647, + "loss": 1.0978, + "step": 8453 + }, + { + "epoch": 1.5051638176638176, + "grad_norm": 0.6336679458618164, + "learning_rate": 0.00013806780766575588, + "loss": 1.0922, + "step": 8454 + }, + { + "epoch": 1.5053418803418803, + "grad_norm": 0.700219452381134, + "learning_rate": 0.0001380548637235457, + "loss": 1.0908, + "step": 8455 + }, + { + "epoch": 1.505519943019943, + "grad_norm": 0.6346127986907959, + "learning_rate": 0.0001380419190357295, + "loss": 1.1265, + "step": 8456 + }, + { + "epoch": 1.5056980056980058, + "grad_norm": 0.8653196096420288, + "learning_rate": 0.00013802897360256093, + "loss": 1.0466, + "step": 8457 + }, + { + "epoch": 1.5058760683760684, + "grad_norm": 0.6589069962501526, + "learning_rate": 0.0001380160274242936, + "loss": 1.245, + "step": 8458 + }, + { + "epoch": 1.506054131054131, + "grad_norm": 0.6527602076530457, + "learning_rate": 0.00013800308050118117, + "loss": 1.1539, + "step": 8459 + }, + { + "epoch": 1.5062321937321936, + "grad_norm": 0.6005436182022095, + "learning_rate": 0.00013799013283347734, + "loss": 0.899, + "step": 8460 + }, + { + "epoch": 1.5064102564102564, + "grad_norm": 0.6954274773597717, + "learning_rate": 0.0001379771844214358, + "loss": 1.1245, + "step": 8461 + }, + { + "epoch": 1.5065883190883191, + "grad_norm": 0.658764660358429, + "learning_rate": 0.00013796423526531019, + "loss": 0.9884, + "step": 8462 + }, + { + "epoch": 1.506766381766382, + "grad_norm": 0.652214527130127, + "learning_rate": 0.0001379512853653543, + "loss": 0.9711, + "step": 8463 + }, + { + "epoch": 1.5069444444444444, + "grad_norm": 0.5680044889450073, + "learning_rate": 0.00013793833472182176, + "loss": 0.9055, + "step": 8464 + }, + { + "epoch": 1.5071225071225072, + "grad_norm": 0.7524166703224182, + "learning_rate": 0.0001379253833349664, + "loss": 1.1163, + "step": 8465 + }, + { + "epoch": 1.5073005698005697, + "grad_norm": 0.692936897277832, + "learning_rate": 0.0001379124312050419, + "loss": 0.899, + "step": 8466 + }, + { + "epoch": 1.5074786324786325, + "grad_norm": 0.6871617436408997, + "learning_rate": 0.00013789947833230207, + "loss": 0.9416, + "step": 8467 + }, + { + "epoch": 1.5076566951566952, + "grad_norm": 0.5983462333679199, + "learning_rate": 0.0001378865247170007, + "loss": 0.9776, + "step": 8468 + }, + { + "epoch": 1.507834757834758, + "grad_norm": 0.6486790180206299, + "learning_rate": 0.0001378735703593916, + "loss": 0.9346, + "step": 8469 + }, + { + "epoch": 1.5080128205128205, + "grad_norm": 0.6843809485435486, + "learning_rate": 0.00013786061525972857, + "loss": 1.1276, + "step": 8470 + }, + { + "epoch": 1.5081908831908832, + "grad_norm": 0.5734516382217407, + "learning_rate": 0.00013784765941826538, + "loss": 0.6939, + "step": 8471 + }, + { + "epoch": 1.5083689458689458, + "grad_norm": 0.6126381754875183, + "learning_rate": 0.00013783470283525596, + "loss": 0.8609, + "step": 8472 + }, + { + "epoch": 1.5085470085470085, + "grad_norm": 0.7570928335189819, + "learning_rate": 0.00013782174551095415, + "loss": 0.8809, + "step": 8473 + }, + { + "epoch": 1.5087250712250713, + "grad_norm": 0.6911360025405884, + "learning_rate": 0.00013780878744561377, + "loss": 0.9916, + "step": 8474 + }, + { + "epoch": 1.508903133903134, + "grad_norm": 0.6651954650878906, + "learning_rate": 0.00013779582863948878, + "loss": 1.0012, + "step": 8475 + }, + { + "epoch": 1.5090811965811965, + "grad_norm": 0.845396876335144, + "learning_rate": 0.000137782869092833, + "loss": 0.8455, + "step": 8476 + }, + { + "epoch": 1.5092592592592593, + "grad_norm": 0.6958050727844238, + "learning_rate": 0.00013776990880590042, + "loss": 1.0264, + "step": 8477 + }, + { + "epoch": 1.5094373219373218, + "grad_norm": 0.6950124502182007, + "learning_rate": 0.00013775694777894493, + "loss": 1.0547, + "step": 8478 + }, + { + "epoch": 1.5096153846153846, + "grad_norm": 0.7243088483810425, + "learning_rate": 0.00013774398601222045, + "loss": 1.0999, + "step": 8479 + }, + { + "epoch": 1.5097934472934473, + "grad_norm": 0.6820448040962219, + "learning_rate": 0.00013773102350598097, + "loss": 0.823, + "step": 8480 + }, + { + "epoch": 1.50997150997151, + "grad_norm": 0.689996063709259, + "learning_rate": 0.0001377180602604805, + "loss": 1.049, + "step": 8481 + }, + { + "epoch": 1.5101495726495726, + "grad_norm": 0.6763314604759216, + "learning_rate": 0.000137705096275973, + "loss": 0.9633, + "step": 8482 + }, + { + "epoch": 1.5103276353276354, + "grad_norm": 0.6760517358779907, + "learning_rate": 0.00013769213155271243, + "loss": 1.0326, + "step": 8483 + }, + { + "epoch": 1.510505698005698, + "grad_norm": 0.7181188464164734, + "learning_rate": 0.00013767916609095285, + "loss": 0.9629, + "step": 8484 + }, + { + "epoch": 1.5106837606837606, + "grad_norm": 0.7102212905883789, + "learning_rate": 0.0001376661998909483, + "loss": 1.2714, + "step": 8485 + }, + { + "epoch": 1.5108618233618234, + "grad_norm": 0.6719805598258972, + "learning_rate": 0.00013765323295295278, + "loss": 0.7848, + "step": 8486 + }, + { + "epoch": 1.5110398860398861, + "grad_norm": 0.6592095494270325, + "learning_rate": 0.0001376402652772204, + "loss": 0.882, + "step": 8487 + }, + { + "epoch": 1.5112179487179487, + "grad_norm": 0.6858693361282349, + "learning_rate": 0.00013762729686400522, + "loss": 0.9418, + "step": 8488 + }, + { + "epoch": 1.5113960113960114, + "grad_norm": 0.7183199524879456, + "learning_rate": 0.0001376143277135613, + "loss": 1.0611, + "step": 8489 + }, + { + "epoch": 1.511574074074074, + "grad_norm": 0.6294263005256653, + "learning_rate": 0.00013760135782614277, + "loss": 0.864, + "step": 8490 + }, + { + "epoch": 1.5117521367521367, + "grad_norm": 0.6762619614601135, + "learning_rate": 0.00013758838720200376, + "loss": 1.0295, + "step": 8491 + }, + { + "epoch": 1.5119301994301995, + "grad_norm": 0.6919726133346558, + "learning_rate": 0.00013757541584139834, + "loss": 1.0803, + "step": 8492 + }, + { + "epoch": 1.5121082621082622, + "grad_norm": 0.6801241040229797, + "learning_rate": 0.00013756244374458075, + "loss": 1.1394, + "step": 8493 + }, + { + "epoch": 1.5122863247863247, + "grad_norm": 0.6758754253387451, + "learning_rate": 0.0001375494709118051, + "loss": 1.0053, + "step": 8494 + }, + { + "epoch": 1.5124643874643875, + "grad_norm": 0.6727001070976257, + "learning_rate": 0.00013753649734332555, + "loss": 1.1407, + "step": 8495 + }, + { + "epoch": 1.51264245014245, + "grad_norm": 0.693913459777832, + "learning_rate": 0.00013752352303939632, + "loss": 1.1804, + "step": 8496 + }, + { + "epoch": 1.5128205128205128, + "grad_norm": 0.6122510433197021, + "learning_rate": 0.0001375105480002716, + "loss": 0.917, + "step": 8497 + }, + { + "epoch": 1.5129985754985755, + "grad_norm": 0.6305009722709656, + "learning_rate": 0.00013749757222620562, + "loss": 1.1075, + "step": 8498 + }, + { + "epoch": 1.5131766381766383, + "grad_norm": 0.7249642610549927, + "learning_rate": 0.0001374845957174526, + "loss": 0.9107, + "step": 8499 + }, + { + "epoch": 1.5133547008547008, + "grad_norm": 0.6922136545181274, + "learning_rate": 0.0001374716184742668, + "loss": 0.9974, + "step": 8500 + }, + { + "epoch": 1.5135327635327636, + "grad_norm": 0.6989904046058655, + "learning_rate": 0.00013745864049690245, + "loss": 0.9866, + "step": 8501 + }, + { + "epoch": 1.513710826210826, + "grad_norm": 0.6284058094024658, + "learning_rate": 0.0001374456617856139, + "loss": 0.8658, + "step": 8502 + }, + { + "epoch": 1.5138888888888888, + "grad_norm": 0.615388810634613, + "learning_rate": 0.00013743268234065535, + "loss": 0.7876, + "step": 8503 + }, + { + "epoch": 1.5140669515669516, + "grad_norm": 0.6212600469589233, + "learning_rate": 0.0001374197021622812, + "loss": 0.855, + "step": 8504 + }, + { + "epoch": 1.5142450142450143, + "grad_norm": 0.6312419772148132, + "learning_rate": 0.00013740672125074567, + "loss": 0.9252, + "step": 8505 + }, + { + "epoch": 1.5144230769230769, + "grad_norm": 0.7094576954841614, + "learning_rate": 0.00013739373960630315, + "loss": 0.7655, + "step": 8506 + }, + { + "epoch": 1.5146011396011396, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.000137380757229208, + "loss": 0.7855, + "step": 8507 + }, + { + "epoch": 1.5147792022792022, + "grad_norm": 0.6798399686813354, + "learning_rate": 0.00013736777411971457, + "loss": 0.9935, + "step": 8508 + }, + { + "epoch": 1.514957264957265, + "grad_norm": 0.7835991978645325, + "learning_rate": 0.00013735479027807723, + "loss": 1.1603, + "step": 8509 + }, + { + "epoch": 1.5151353276353277, + "grad_norm": 0.6230790615081787, + "learning_rate": 0.00013734180570455033, + "loss": 1.1463, + "step": 8510 + }, + { + "epoch": 1.5153133903133904, + "grad_norm": 0.646603524684906, + "learning_rate": 0.00013732882039938835, + "loss": 0.9564, + "step": 8511 + }, + { + "epoch": 1.515491452991453, + "grad_norm": 0.6619647145271301, + "learning_rate": 0.0001373158343628457, + "loss": 0.8492, + "step": 8512 + }, + { + "epoch": 1.5156695156695157, + "grad_norm": 0.6458454132080078, + "learning_rate": 0.00013730284759517675, + "loss": 1.0049, + "step": 8513 + }, + { + "epoch": 1.5158475783475782, + "grad_norm": 0.7415743470191956, + "learning_rate": 0.00013728986009663602, + "loss": 0.872, + "step": 8514 + }, + { + "epoch": 1.516025641025641, + "grad_norm": 0.6198840141296387, + "learning_rate": 0.00013727687186747793, + "loss": 0.8645, + "step": 8515 + }, + { + "epoch": 1.5162037037037037, + "grad_norm": 0.7160853147506714, + "learning_rate": 0.00013726388290795697, + "loss": 1.0144, + "step": 8516 + }, + { + "epoch": 1.5163817663817665, + "grad_norm": 0.6604135632514954, + "learning_rate": 0.00013725089321832765, + "loss": 0.9827, + "step": 8517 + }, + { + "epoch": 1.5165598290598292, + "grad_norm": 0.6480790972709656, + "learning_rate": 0.00013723790279884443, + "loss": 1.0357, + "step": 8518 + }, + { + "epoch": 1.5167378917378918, + "grad_norm": 0.6207128167152405, + "learning_rate": 0.00013722491164976187, + "loss": 0.9467, + "step": 8519 + }, + { + "epoch": 1.5169159544159543, + "grad_norm": 0.6024298667907715, + "learning_rate": 0.00013721191977133452, + "loss": 0.8821, + "step": 8520 + }, + { + "epoch": 1.517094017094017, + "grad_norm": 0.684898316860199, + "learning_rate": 0.00013719892716381688, + "loss": 0.9823, + "step": 8521 + }, + { + "epoch": 1.5172720797720798, + "grad_norm": 0.7460635304450989, + "learning_rate": 0.00013718593382746355, + "loss": 1.2573, + "step": 8522 + }, + { + "epoch": 1.5174501424501425, + "grad_norm": 0.7193243503570557, + "learning_rate": 0.00013717293976252907, + "loss": 1.0162, + "step": 8523 + }, + { + "epoch": 1.5176282051282053, + "grad_norm": 0.6328752040863037, + "learning_rate": 0.0001371599449692681, + "loss": 0.8183, + "step": 8524 + }, + { + "epoch": 1.5178062678062678, + "grad_norm": 0.658784806728363, + "learning_rate": 0.00013714694944793517, + "loss": 0.9315, + "step": 8525 + }, + { + "epoch": 1.5179843304843303, + "grad_norm": 0.7875827550888062, + "learning_rate": 0.00013713395319878493, + "loss": 1.0889, + "step": 8526 + }, + { + "epoch": 1.518162393162393, + "grad_norm": 0.6580079793930054, + "learning_rate": 0.00013712095622207203, + "loss": 1.0276, + "step": 8527 + }, + { + "epoch": 1.5183404558404558, + "grad_norm": 0.6214027404785156, + "learning_rate": 0.00013710795851805106, + "loss": 0.9692, + "step": 8528 + }, + { + "epoch": 1.5185185185185186, + "grad_norm": 0.7839403748512268, + "learning_rate": 0.0001370949600869768, + "loss": 0.7378, + "step": 8529 + }, + { + "epoch": 1.5186965811965814, + "grad_norm": 0.6632764339447021, + "learning_rate": 0.0001370819609291038, + "loss": 0.9431, + "step": 8530 + }, + { + "epoch": 1.5188746438746439, + "grad_norm": 0.7071712017059326, + "learning_rate": 0.00013706896104468682, + "loss": 0.7684, + "step": 8531 + }, + { + "epoch": 1.5190527065527064, + "grad_norm": 0.7494829297065735, + "learning_rate": 0.00013705596043398058, + "loss": 0.9709, + "step": 8532 + }, + { + "epoch": 1.5192307692307692, + "grad_norm": 0.6408106088638306, + "learning_rate": 0.00013704295909723973, + "loss": 0.8494, + "step": 8533 + }, + { + "epoch": 1.519408831908832, + "grad_norm": 0.6043150424957275, + "learning_rate": 0.0001370299570347191, + "loss": 0.7485, + "step": 8534 + }, + { + "epoch": 1.5195868945868947, + "grad_norm": 0.6944992542266846, + "learning_rate": 0.00013701695424667336, + "loss": 0.8403, + "step": 8535 + }, + { + "epoch": 1.5197649572649574, + "grad_norm": 0.7730217576026917, + "learning_rate": 0.00013700395073335726, + "loss": 0.9122, + "step": 8536 + }, + { + "epoch": 1.51994301994302, + "grad_norm": 0.6300255060195923, + "learning_rate": 0.00013699094649502564, + "loss": 0.9185, + "step": 8537 + }, + { + "epoch": 1.5201210826210825, + "grad_norm": 0.648676335811615, + "learning_rate": 0.00013697794153193327, + "loss": 0.9897, + "step": 8538 + }, + { + "epoch": 1.5202991452991452, + "grad_norm": 0.7365788817405701, + "learning_rate": 0.00013696493584433494, + "loss": 0.958, + "step": 8539 + }, + { + "epoch": 1.520477207977208, + "grad_norm": 0.6634557247161865, + "learning_rate": 0.00013695192943248552, + "loss": 0.9389, + "step": 8540 + }, + { + "epoch": 1.5206552706552707, + "grad_norm": 0.6110827922821045, + "learning_rate": 0.00013693892229663977, + "loss": 0.9341, + "step": 8541 + }, + { + "epoch": 1.5208333333333335, + "grad_norm": 0.7207275032997131, + "learning_rate": 0.00013692591443705256, + "loss": 0.9526, + "step": 8542 + }, + { + "epoch": 1.521011396011396, + "grad_norm": 0.7071022391319275, + "learning_rate": 0.0001369129058539788, + "loss": 0.9572, + "step": 8543 + }, + { + "epoch": 1.5211894586894585, + "grad_norm": 0.5898227691650391, + "learning_rate": 0.0001368998965476733, + "loss": 0.921, + "step": 8544 + }, + { + "epoch": 1.5213675213675213, + "grad_norm": 0.7542559504508972, + "learning_rate": 0.000136886886518391, + "loss": 0.7799, + "step": 8545 + }, + { + "epoch": 1.521545584045584, + "grad_norm": 0.6904959678649902, + "learning_rate": 0.00013687387576638674, + "loss": 0.9601, + "step": 8546 + }, + { + "epoch": 1.5217236467236468, + "grad_norm": 0.763414204120636, + "learning_rate": 0.00013686086429191553, + "loss": 1.0046, + "step": 8547 + }, + { + "epoch": 1.5219017094017095, + "grad_norm": 0.6879960298538208, + "learning_rate": 0.00013684785209523224, + "loss": 0.9615, + "step": 8548 + }, + { + "epoch": 1.522079772079772, + "grad_norm": 0.7166057229042053, + "learning_rate": 0.00013683483917659186, + "loss": 0.9481, + "step": 8549 + }, + { + "epoch": 1.5222578347578346, + "grad_norm": 0.6384348273277283, + "learning_rate": 0.0001368218255362493, + "loss": 1.1037, + "step": 8550 + }, + { + "epoch": 1.5224358974358974, + "grad_norm": 0.6564528346061707, + "learning_rate": 0.00013680881117445953, + "loss": 0.951, + "step": 8551 + }, + { + "epoch": 1.52261396011396, + "grad_norm": 0.749301016330719, + "learning_rate": 0.00013679579609147762, + "loss": 0.9324, + "step": 8552 + }, + { + "epoch": 1.5227920227920229, + "grad_norm": 0.8130472898483276, + "learning_rate": 0.00013678278028755848, + "loss": 1.0178, + "step": 8553 + }, + { + "epoch": 1.5229700854700856, + "grad_norm": 0.6763297319412231, + "learning_rate": 0.0001367697637629572, + "loss": 0.9224, + "step": 8554 + }, + { + "epoch": 1.5231481481481481, + "grad_norm": 0.6630885601043701, + "learning_rate": 0.00013675674651792878, + "loss": 1.0254, + "step": 8555 + }, + { + "epoch": 1.5233262108262107, + "grad_norm": 0.7377206087112427, + "learning_rate": 0.00013674372855272825, + "loss": 1.0413, + "step": 8556 + }, + { + "epoch": 1.5235042735042734, + "grad_norm": 0.5270320177078247, + "learning_rate": 0.00013673070986761068, + "loss": 0.7124, + "step": 8557 + }, + { + "epoch": 1.5236823361823362, + "grad_norm": 0.5941976308822632, + "learning_rate": 0.00013671769046283116, + "loss": 1.0281, + "step": 8558 + }, + { + "epoch": 1.523860398860399, + "grad_norm": 0.6131376028060913, + "learning_rate": 0.0001367046703386448, + "loss": 0.7593, + "step": 8559 + }, + { + "epoch": 1.5240384615384617, + "grad_norm": 0.7381763458251953, + "learning_rate": 0.00013669164949530664, + "loss": 1.148, + "step": 8560 + }, + { + "epoch": 1.5242165242165242, + "grad_norm": 0.683274507522583, + "learning_rate": 0.00013667862793307185, + "loss": 0.8354, + "step": 8561 + }, + { + "epoch": 1.5243945868945867, + "grad_norm": 0.6912649273872375, + "learning_rate": 0.0001366656056521955, + "loss": 0.9043, + "step": 8562 + }, + { + "epoch": 1.5245726495726495, + "grad_norm": 0.5999594330787659, + "learning_rate": 0.0001366525826529328, + "loss": 0.6138, + "step": 8563 + }, + { + "epoch": 1.5247507122507122, + "grad_norm": 0.7185927629470825, + "learning_rate": 0.00013663955893553892, + "loss": 0.895, + "step": 8564 + }, + { + "epoch": 1.524928774928775, + "grad_norm": 0.5967002511024475, + "learning_rate": 0.00013662653450026893, + "loss": 0.9636, + "step": 8565 + }, + { + "epoch": 1.5251068376068377, + "grad_norm": 0.7122953534126282, + "learning_rate": 0.00013661350934737813, + "loss": 0.9465, + "step": 8566 + }, + { + "epoch": 1.5252849002849003, + "grad_norm": 0.705326497554779, + "learning_rate": 0.00013660048347712163, + "loss": 1.121, + "step": 8567 + }, + { + "epoch": 1.5254629629629628, + "grad_norm": 0.6023733019828796, + "learning_rate": 0.0001365874568897547, + "loss": 0.9881, + "step": 8568 + }, + { + "epoch": 1.5256410256410255, + "grad_norm": 0.6883122324943542, + "learning_rate": 0.0001365744295855326, + "loss": 1.2372, + "step": 8569 + }, + { + "epoch": 1.5258190883190883, + "grad_norm": 0.718126654624939, + "learning_rate": 0.0001365614015647105, + "loss": 1.0888, + "step": 8570 + }, + { + "epoch": 1.525997150997151, + "grad_norm": 0.6649243831634521, + "learning_rate": 0.00013654837282754367, + "loss": 1.0458, + "step": 8571 + }, + { + "epoch": 1.5261752136752138, + "grad_norm": 0.6959797143936157, + "learning_rate": 0.00013653534337428738, + "loss": 0.9282, + "step": 8572 + }, + { + "epoch": 1.5263532763532763, + "grad_norm": 0.6069976687431335, + "learning_rate": 0.00013652231320519697, + "loss": 0.9706, + "step": 8573 + }, + { + "epoch": 1.526531339031339, + "grad_norm": 0.7085374593734741, + "learning_rate": 0.0001365092823205277, + "loss": 1.1241, + "step": 8574 + }, + { + "epoch": 1.5267094017094016, + "grad_norm": 0.575106143951416, + "learning_rate": 0.00013649625072053488, + "loss": 0.9814, + "step": 8575 + }, + { + "epoch": 1.5268874643874644, + "grad_norm": 0.6541273593902588, + "learning_rate": 0.00013648321840547384, + "loss": 1.0155, + "step": 8576 + }, + { + "epoch": 1.5270655270655271, + "grad_norm": 0.6754382848739624, + "learning_rate": 0.0001364701853755999, + "loss": 1.0284, + "step": 8577 + }, + { + "epoch": 1.5272435897435899, + "grad_norm": 0.6219634413719177, + "learning_rate": 0.00013645715163116846, + "loss": 1.1539, + "step": 8578 + }, + { + "epoch": 1.5274216524216524, + "grad_norm": 0.7625157833099365, + "learning_rate": 0.00013644411717243486, + "loss": 1.1157, + "step": 8579 + }, + { + "epoch": 1.5275997150997151, + "grad_norm": 0.6944296956062317, + "learning_rate": 0.0001364310819996545, + "loss": 0.8309, + "step": 8580 + }, + { + "epoch": 1.5277777777777777, + "grad_norm": 0.7198494672775269, + "learning_rate": 0.00013641804611308277, + "loss": 1.0883, + "step": 8581 + }, + { + "epoch": 1.5279558404558404, + "grad_norm": 0.6398822069168091, + "learning_rate": 0.00013640500951297508, + "loss": 1.0173, + "step": 8582 + }, + { + "epoch": 1.5281339031339032, + "grad_norm": 0.7306683659553528, + "learning_rate": 0.00013639197219958682, + "loss": 0.9979, + "step": 8583 + }, + { + "epoch": 1.528311965811966, + "grad_norm": 0.6873512268066406, + "learning_rate": 0.00013637893417317348, + "loss": 0.7883, + "step": 8584 + }, + { + "epoch": 1.5284900284900285, + "grad_norm": 0.6482085585594177, + "learning_rate": 0.00013636589543399052, + "loss": 0.9367, + "step": 8585 + }, + { + "epoch": 1.5286680911680912, + "grad_norm": 0.8161232471466064, + "learning_rate": 0.00013635285598229336, + "loss": 1.0582, + "step": 8586 + }, + { + "epoch": 1.5288461538461537, + "grad_norm": 0.6722155809402466, + "learning_rate": 0.0001363398158183375, + "loss": 0.9805, + "step": 8587 + }, + { + "epoch": 1.5290242165242165, + "grad_norm": 0.7175397872924805, + "learning_rate": 0.00013632677494237845, + "loss": 1.0747, + "step": 8588 + }, + { + "epoch": 1.5292022792022792, + "grad_norm": 0.6665592789649963, + "learning_rate": 0.00013631373335467172, + "loss": 1.006, + "step": 8589 + }, + { + "epoch": 1.529380341880342, + "grad_norm": 0.7002299427986145, + "learning_rate": 0.0001363006910554728, + "loss": 1.0702, + "step": 8590 + }, + { + "epoch": 1.5295584045584045, + "grad_norm": 0.7712168097496033, + "learning_rate": 0.00013628764804503725, + "loss": 1.0628, + "step": 8591 + }, + { + "epoch": 1.5297364672364673, + "grad_norm": 0.6620795130729675, + "learning_rate": 0.0001362746043236206, + "loss": 1.01, + "step": 8592 + }, + { + "epoch": 1.5299145299145298, + "grad_norm": 0.6374393701553345, + "learning_rate": 0.00013626155989147846, + "loss": 0.9106, + "step": 8593 + }, + { + "epoch": 1.5300925925925926, + "grad_norm": 0.6531631946563721, + "learning_rate": 0.00013624851474886636, + "loss": 1.0488, + "step": 8594 + }, + { + "epoch": 1.5302706552706553, + "grad_norm": 0.6843775510787964, + "learning_rate": 0.00013623546889603993, + "loss": 0.8599, + "step": 8595 + }, + { + "epoch": 1.530448717948718, + "grad_norm": 0.7232706546783447, + "learning_rate": 0.00013622242233325476, + "loss": 1.0875, + "step": 8596 + }, + { + "epoch": 1.5306267806267806, + "grad_norm": 0.695691704750061, + "learning_rate": 0.00013620937506076644, + "loss": 0.9835, + "step": 8597 + }, + { + "epoch": 1.5308048433048433, + "grad_norm": 0.6321248412132263, + "learning_rate": 0.00013619632707883065, + "loss": 0.9778, + "step": 8598 + }, + { + "epoch": 1.5309829059829059, + "grad_norm": 0.6469168663024902, + "learning_rate": 0.00013618327838770303, + "loss": 0.9968, + "step": 8599 + }, + { + "epoch": 1.5311609686609686, + "grad_norm": 0.6798683404922485, + "learning_rate": 0.00013617022898763925, + "loss": 0.78, + "step": 8600 + }, + { + "epoch": 1.5313390313390314, + "grad_norm": 0.6932336091995239, + "learning_rate": 0.00013615717887889496, + "loss": 0.9473, + "step": 8601 + }, + { + "epoch": 1.5315170940170941, + "grad_norm": 0.7304185628890991, + "learning_rate": 0.00013614412806172585, + "loss": 1.0478, + "step": 8602 + }, + { + "epoch": 1.5316951566951567, + "grad_norm": 0.6585272550582886, + "learning_rate": 0.00013613107653638763, + "loss": 0.8563, + "step": 8603 + }, + { + "epoch": 1.5318732193732194, + "grad_norm": 0.6804470419883728, + "learning_rate": 0.00013611802430313604, + "loss": 0.9839, + "step": 8604 + }, + { + "epoch": 1.532051282051282, + "grad_norm": 0.7271378040313721, + "learning_rate": 0.0001361049713622268, + "loss": 1.0906, + "step": 8605 + }, + { + "epoch": 1.5322293447293447, + "grad_norm": 0.7731603980064392, + "learning_rate": 0.00013609191771391562, + "loss": 1.1318, + "step": 8606 + }, + { + "epoch": 1.5324074074074074, + "grad_norm": 0.6143709421157837, + "learning_rate": 0.0001360788633584583, + "loss": 0.8726, + "step": 8607 + }, + { + "epoch": 1.5325854700854702, + "grad_norm": 0.6847203373908997, + "learning_rate": 0.00013606580829611056, + "loss": 0.9963, + "step": 8608 + }, + { + "epoch": 1.5327635327635327, + "grad_norm": 0.7561219334602356, + "learning_rate": 0.0001360527525271283, + "loss": 0.8873, + "step": 8609 + }, + { + "epoch": 1.5329415954415955, + "grad_norm": 0.7997925281524658, + "learning_rate": 0.0001360396960517672, + "loss": 0.7675, + "step": 8610 + }, + { + "epoch": 1.533119658119658, + "grad_norm": 0.7206357717514038, + "learning_rate": 0.00013602663887028315, + "loss": 1.0084, + "step": 8611 + }, + { + "epoch": 1.5332977207977208, + "grad_norm": 0.6454238891601562, + "learning_rate": 0.00013601358098293194, + "loss": 0.8194, + "step": 8612 + }, + { + "epoch": 1.5334757834757835, + "grad_norm": 0.5531884431838989, + "learning_rate": 0.0001360005223899694, + "loss": 0.8596, + "step": 8613 + }, + { + "epoch": 1.5336538461538463, + "grad_norm": 0.659161388874054, + "learning_rate": 0.00013598746309165144, + "loss": 1.0363, + "step": 8614 + }, + { + "epoch": 1.5338319088319088, + "grad_norm": 0.6958948373794556, + "learning_rate": 0.00013597440308823385, + "loss": 0.9852, + "step": 8615 + }, + { + "epoch": 1.5340099715099715, + "grad_norm": 0.7147171497344971, + "learning_rate": 0.0001359613423799726, + "loss": 1.0506, + "step": 8616 + }, + { + "epoch": 1.534188034188034, + "grad_norm": 0.604450523853302, + "learning_rate": 0.00013594828096712353, + "loss": 0.9344, + "step": 8617 + }, + { + "epoch": 1.5343660968660968, + "grad_norm": 0.714547336101532, + "learning_rate": 0.00013593521884994257, + "loss": 1.1583, + "step": 8618 + }, + { + "epoch": 1.5345441595441596, + "grad_norm": 0.6864442825317383, + "learning_rate": 0.00013592215602868565, + "loss": 0.991, + "step": 8619 + }, + { + "epoch": 1.5347222222222223, + "grad_norm": 0.6384446620941162, + "learning_rate": 0.00013590909250360873, + "loss": 0.8799, + "step": 8620 + }, + { + "epoch": 1.5349002849002849, + "grad_norm": 0.7307949662208557, + "learning_rate": 0.00013589602827496772, + "loss": 1.0276, + "step": 8621 + }, + { + "epoch": 1.5350783475783476, + "grad_norm": 0.6620129942893982, + "learning_rate": 0.00013588296334301862, + "loss": 0.9378, + "step": 8622 + }, + { + "epoch": 1.5352564102564101, + "grad_norm": 0.7216851711273193, + "learning_rate": 0.00013586989770801735, + "loss": 0.8984, + "step": 8623 + }, + { + "epoch": 1.5354344729344729, + "grad_norm": 0.7319885492324829, + "learning_rate": 0.00013585683137022, + "loss": 1.0357, + "step": 8624 + }, + { + "epoch": 1.5356125356125356, + "grad_norm": 0.7455703616142273, + "learning_rate": 0.00013584376432988247, + "loss": 0.9727, + "step": 8625 + }, + { + "epoch": 1.5357905982905984, + "grad_norm": 0.7285277247428894, + "learning_rate": 0.0001358306965872609, + "loss": 1.1132, + "step": 8626 + }, + { + "epoch": 1.535968660968661, + "grad_norm": 0.6250096559524536, + "learning_rate": 0.00013581762814261124, + "loss": 0.8538, + "step": 8627 + }, + { + "epoch": 1.5361467236467237, + "grad_norm": 0.6252279281616211, + "learning_rate": 0.0001358045589961895, + "loss": 0.822, + "step": 8628 + }, + { + "epoch": 1.5363247863247862, + "grad_norm": 0.7723368406295776, + "learning_rate": 0.0001357914891482519, + "loss": 0.9841, + "step": 8629 + }, + { + "epoch": 1.536502849002849, + "grad_norm": 0.6855236887931824, + "learning_rate": 0.00013577841859905435, + "loss": 0.9512, + "step": 8630 + }, + { + "epoch": 1.5366809116809117, + "grad_norm": 0.8320944309234619, + "learning_rate": 0.00013576534734885303, + "loss": 1.0324, + "step": 8631 + }, + { + "epoch": 1.5368589743589745, + "grad_norm": 0.6970052123069763, + "learning_rate": 0.00013575227539790405, + "loss": 0.9874, + "step": 8632 + }, + { + "epoch": 1.5370370370370372, + "grad_norm": 0.7774853110313416, + "learning_rate": 0.00013573920274646345, + "loss": 0.962, + "step": 8633 + }, + { + "epoch": 1.5372150997150997, + "grad_norm": 0.6479182839393616, + "learning_rate": 0.0001357261293947875, + "loss": 0.9438, + "step": 8634 + }, + { + "epoch": 1.5373931623931623, + "grad_norm": 0.6855679750442505, + "learning_rate": 0.00013571305534313218, + "loss": 1.0898, + "step": 8635 + }, + { + "epoch": 1.537571225071225, + "grad_norm": 0.6527835726737976, + "learning_rate": 0.00013569998059175377, + "loss": 0.954, + "step": 8636 + }, + { + "epoch": 1.5377492877492878, + "grad_norm": 0.6601176857948303, + "learning_rate": 0.00013568690514090837, + "loss": 1.0183, + "step": 8637 + }, + { + "epoch": 1.5379273504273505, + "grad_norm": 0.6628120541572571, + "learning_rate": 0.0001356738289908522, + "loss": 1.0651, + "step": 8638 + }, + { + "epoch": 1.5381054131054133, + "grad_norm": 0.7492203712463379, + "learning_rate": 0.00013566075214184147, + "loss": 1.2438, + "step": 8639 + }, + { + "epoch": 1.5382834757834758, + "grad_norm": 0.6781343817710876, + "learning_rate": 0.00013564767459413237, + "loss": 0.9413, + "step": 8640 + }, + { + "epoch": 1.5384615384615383, + "grad_norm": 0.6890891790390015, + "learning_rate": 0.00013563459634798115, + "loss": 0.9912, + "step": 8641 + }, + { + "epoch": 1.538639601139601, + "grad_norm": 0.722820520401001, + "learning_rate": 0.00013562151740364404, + "loss": 1.1799, + "step": 8642 + }, + { + "epoch": 1.5388176638176638, + "grad_norm": 0.738369882106781, + "learning_rate": 0.0001356084377613773, + "loss": 1.1313, + "step": 8643 + }, + { + "epoch": 1.5389957264957266, + "grad_norm": 0.6232718229293823, + "learning_rate": 0.00013559535742143717, + "loss": 0.9035, + "step": 8644 + }, + { + "epoch": 1.5391737891737893, + "grad_norm": 0.7371624708175659, + "learning_rate": 0.00013558227638407996, + "loss": 1.3377, + "step": 8645 + }, + { + "epoch": 1.5393518518518519, + "grad_norm": 0.658353865146637, + "learning_rate": 0.00013556919464956197, + "loss": 0.9591, + "step": 8646 + }, + { + "epoch": 1.5395299145299144, + "grad_norm": 0.6205827593803406, + "learning_rate": 0.0001355561122181395, + "loss": 0.9217, + "step": 8647 + }, + { + "epoch": 1.5397079772079771, + "grad_norm": 0.5892502069473267, + "learning_rate": 0.00013554302909006888, + "loss": 0.8893, + "step": 8648 + }, + { + "epoch": 1.53988603988604, + "grad_norm": 1.224568486213684, + "learning_rate": 0.0001355299452656064, + "loss": 0.8237, + "step": 8649 + }, + { + "epoch": 1.5400641025641026, + "grad_norm": 0.7732635736465454, + "learning_rate": 0.0001355168607450085, + "loss": 1.1043, + "step": 8650 + }, + { + "epoch": 1.5402421652421654, + "grad_norm": 0.6365402340888977, + "learning_rate": 0.00013550377552853146, + "loss": 1.0345, + "step": 8651 + }, + { + "epoch": 1.540420227920228, + "grad_norm": 0.7046400904655457, + "learning_rate": 0.00013549068961643171, + "loss": 1.0361, + "step": 8652 + }, + { + "epoch": 1.5405982905982905, + "grad_norm": 0.6760256886482239, + "learning_rate": 0.0001354776030089656, + "loss": 0.9437, + "step": 8653 + }, + { + "epoch": 1.5407763532763532, + "grad_norm": 0.6180984973907471, + "learning_rate": 0.00013546451570638958, + "loss": 0.9737, + "step": 8654 + }, + { + "epoch": 1.540954415954416, + "grad_norm": 0.6221960186958313, + "learning_rate": 0.00013545142770896005, + "loss": 0.9313, + "step": 8655 + }, + { + "epoch": 1.5411324786324787, + "grad_norm": 0.6887816786766052, + "learning_rate": 0.0001354383390169334, + "loss": 1.1736, + "step": 8656 + }, + { + "epoch": 1.5413105413105415, + "grad_norm": 0.5840606093406677, + "learning_rate": 0.00013542524963056614, + "loss": 0.9269, + "step": 8657 + }, + { + "epoch": 1.541488603988604, + "grad_norm": 0.7396654486656189, + "learning_rate": 0.00013541215955011472, + "loss": 1.1189, + "step": 8658 + }, + { + "epoch": 1.5416666666666665, + "grad_norm": 0.780616819858551, + "learning_rate": 0.00013539906877583555, + "loss": 1.1251, + "step": 8659 + }, + { + "epoch": 1.5418447293447293, + "grad_norm": 0.6975206732749939, + "learning_rate": 0.0001353859773079852, + "loss": 1.2134, + "step": 8660 + }, + { + "epoch": 1.542022792022792, + "grad_norm": 0.7572869658470154, + "learning_rate": 0.00013537288514682013, + "loss": 0.9396, + "step": 8661 + }, + { + "epoch": 1.5422008547008548, + "grad_norm": 0.6252159476280212, + "learning_rate": 0.00013535979229259686, + "loss": 0.8449, + "step": 8662 + }, + { + "epoch": 1.5423789173789175, + "grad_norm": 0.7321650981903076, + "learning_rate": 0.0001353466987455719, + "loss": 1.3263, + "step": 8663 + }, + { + "epoch": 1.54255698005698, + "grad_norm": 0.7168700695037842, + "learning_rate": 0.00013533360450600177, + "loss": 0.8923, + "step": 8664 + }, + { + "epoch": 1.5427350427350426, + "grad_norm": 0.5931934714317322, + "learning_rate": 0.00013532050957414313, + "loss": 0.8448, + "step": 8665 + }, + { + "epoch": 1.5429131054131053, + "grad_norm": 0.6621279120445251, + "learning_rate": 0.00013530741395025245, + "loss": 1.1023, + "step": 8666 + }, + { + "epoch": 1.543091168091168, + "grad_norm": 0.7133732438087463, + "learning_rate": 0.00013529431763458633, + "loss": 0.9986, + "step": 8667 + }, + { + "epoch": 1.5432692307692308, + "grad_norm": 0.7589015960693359, + "learning_rate": 0.0001352812206274014, + "loss": 1.0111, + "step": 8668 + }, + { + "epoch": 1.5434472934472936, + "grad_norm": 0.6958192586898804, + "learning_rate": 0.0001352681229289542, + "loss": 0.9466, + "step": 8669 + }, + { + "epoch": 1.5436253561253561, + "grad_norm": 0.7539750337600708, + "learning_rate": 0.0001352550245395014, + "loss": 1.0974, + "step": 8670 + }, + { + "epoch": 1.5438034188034186, + "grad_norm": 0.7003816366195679, + "learning_rate": 0.00013524192545929964, + "loss": 1.0354, + "step": 8671 + }, + { + "epoch": 1.5439814814814814, + "grad_norm": 0.6503025889396667, + "learning_rate": 0.00013522882568860558, + "loss": 1.0476, + "step": 8672 + }, + { + "epoch": 1.5441595441595442, + "grad_norm": 0.6757345199584961, + "learning_rate": 0.00013521572522767584, + "loss": 0.864, + "step": 8673 + }, + { + "epoch": 1.544337606837607, + "grad_norm": 0.6857611536979675, + "learning_rate": 0.0001352026240767671, + "loss": 1.1627, + "step": 8674 + }, + { + "epoch": 1.5445156695156697, + "grad_norm": 0.5775430798530579, + "learning_rate": 0.0001351895222361361, + "loss": 0.7444, + "step": 8675 + }, + { + "epoch": 1.5446937321937322, + "grad_norm": 0.7511499524116516, + "learning_rate": 0.00013517641970603952, + "loss": 1.1547, + "step": 8676 + }, + { + "epoch": 1.5448717948717947, + "grad_norm": 0.6727504730224609, + "learning_rate": 0.00013516331648673403, + "loss": 1.0829, + "step": 8677 + }, + { + "epoch": 1.5450498575498575, + "grad_norm": 0.6128812432289124, + "learning_rate": 0.00013515021257847642, + "loss": 0.9318, + "step": 8678 + }, + { + "epoch": 1.5452279202279202, + "grad_norm": 0.7309781312942505, + "learning_rate": 0.00013513710798152343, + "loss": 1.0844, + "step": 8679 + }, + { + "epoch": 1.545405982905983, + "grad_norm": 0.695655882358551, + "learning_rate": 0.00013512400269613176, + "loss": 1.113, + "step": 8680 + }, + { + "epoch": 1.5455840455840457, + "grad_norm": 0.696441650390625, + "learning_rate": 0.00013511089672255824, + "loss": 1.0499, + "step": 8681 + }, + { + "epoch": 1.5457621082621082, + "grad_norm": 0.6309961080551147, + "learning_rate": 0.00013509779006105964, + "loss": 0.8759, + "step": 8682 + }, + { + "epoch": 1.5459401709401708, + "grad_norm": 0.6155984401702881, + "learning_rate": 0.00013508468271189277, + "loss": 0.8967, + "step": 8683 + }, + { + "epoch": 1.5461182336182335, + "grad_norm": 0.6786884665489197, + "learning_rate": 0.00013507157467531442, + "loss": 1.0806, + "step": 8684 + }, + { + "epoch": 1.5462962962962963, + "grad_norm": 0.6494075059890747, + "learning_rate": 0.00013505846595158138, + "loss": 1.0196, + "step": 8685 + }, + { + "epoch": 1.546474358974359, + "grad_norm": 0.7599824070930481, + "learning_rate": 0.00013504535654095055, + "loss": 0.8662, + "step": 8686 + }, + { + "epoch": 1.5466524216524218, + "grad_norm": 0.6017210483551025, + "learning_rate": 0.00013503224644367877, + "loss": 0.872, + "step": 8687 + }, + { + "epoch": 1.5468304843304843, + "grad_norm": 0.7972410321235657, + "learning_rate": 0.00013501913566002288, + "loss": 1.0958, + "step": 8688 + }, + { + "epoch": 1.547008547008547, + "grad_norm": 0.7572960257530212, + "learning_rate": 0.00013500602419023978, + "loss": 1.0219, + "step": 8689 + }, + { + "epoch": 1.5471866096866096, + "grad_norm": 0.6329224109649658, + "learning_rate": 0.00013499291203458635, + "loss": 0.8636, + "step": 8690 + }, + { + "epoch": 1.5473646723646723, + "grad_norm": 0.6777113080024719, + "learning_rate": 0.0001349797991933195, + "loss": 1.0297, + "step": 8691 + }, + { + "epoch": 1.547542735042735, + "grad_norm": 0.6449527144432068, + "learning_rate": 0.00013496668566669617, + "loss": 1.0296, + "step": 8692 + }, + { + "epoch": 1.5477207977207978, + "grad_norm": 0.8236973881721497, + "learning_rate": 0.00013495357145497326, + "loss": 0.8569, + "step": 8693 + }, + { + "epoch": 1.5478988603988604, + "grad_norm": 0.6753743290901184, + "learning_rate": 0.0001349404565584077, + "loss": 1.0733, + "step": 8694 + }, + { + "epoch": 1.5480769230769231, + "grad_norm": 0.6642967462539673, + "learning_rate": 0.0001349273409772565, + "loss": 0.9437, + "step": 8695 + }, + { + "epoch": 1.5482549857549857, + "grad_norm": 0.6470823884010315, + "learning_rate": 0.00013491422471177661, + "loss": 0.999, + "step": 8696 + }, + { + "epoch": 1.5484330484330484, + "grad_norm": 0.7287036776542664, + "learning_rate": 0.000134901107762225, + "loss": 0.9396, + "step": 8697 + }, + { + "epoch": 1.5486111111111112, + "grad_norm": 0.6258324980735779, + "learning_rate": 0.00013488799012885872, + "loss": 1.045, + "step": 8698 + }, + { + "epoch": 1.548789173789174, + "grad_norm": 0.6540539860725403, + "learning_rate": 0.00013487487181193473, + "loss": 0.9939, + "step": 8699 + }, + { + "epoch": 1.5489672364672364, + "grad_norm": 0.7129563093185425, + "learning_rate": 0.00013486175281171003, + "loss": 1.2079, + "step": 8700 + }, + { + "epoch": 1.5491452991452992, + "grad_norm": 0.6383145451545715, + "learning_rate": 0.00013484863312844173, + "loss": 0.9999, + "step": 8701 + }, + { + "epoch": 1.5493233618233617, + "grad_norm": 0.6310200691223145, + "learning_rate": 0.0001348355127623869, + "loss": 1.1193, + "step": 8702 + }, + { + "epoch": 1.5495014245014245, + "grad_norm": 0.6370054483413696, + "learning_rate": 0.0001348223917138025, + "loss": 1.0213, + "step": 8703 + }, + { + "epoch": 1.5496794871794872, + "grad_norm": 0.7052688598632812, + "learning_rate": 0.00013480926998294573, + "loss": 0.8773, + "step": 8704 + }, + { + "epoch": 1.54985754985755, + "grad_norm": 0.6369579434394836, + "learning_rate": 0.00013479614757007355, + "loss": 1.0072, + "step": 8705 + }, + { + "epoch": 1.5500356125356125, + "grad_norm": 0.7152075171470642, + "learning_rate": 0.0001347830244754432, + "loss": 1.0409, + "step": 8706 + }, + { + "epoch": 1.5502136752136753, + "grad_norm": 0.654183566570282, + "learning_rate": 0.00013476990069931173, + "loss": 0.9363, + "step": 8707 + }, + { + "epoch": 1.5503917378917378, + "grad_norm": 0.6700537204742432, + "learning_rate": 0.00013475677624193627, + "loss": 0.985, + "step": 8708 + }, + { + "epoch": 1.5505698005698005, + "grad_norm": 0.7195445895195007, + "learning_rate": 0.00013474365110357402, + "loss": 0.988, + "step": 8709 + }, + { + "epoch": 1.5507478632478633, + "grad_norm": 0.6019890904426575, + "learning_rate": 0.00013473052528448201, + "loss": 0.9915, + "step": 8710 + }, + { + "epoch": 1.550925925925926, + "grad_norm": 0.7787565588951111, + "learning_rate": 0.0001347173987849176, + "loss": 0.9676, + "step": 8711 + }, + { + "epoch": 1.5511039886039886, + "grad_norm": 0.6997103691101074, + "learning_rate": 0.00013470427160513782, + "loss": 1.1158, + "step": 8712 + }, + { + "epoch": 1.5512820512820513, + "grad_norm": 0.6259464025497437, + "learning_rate": 0.00013469114374539998, + "loss": 0.8784, + "step": 8713 + }, + { + "epoch": 1.5514601139601139, + "grad_norm": 0.6159056425094604, + "learning_rate": 0.00013467801520596122, + "loss": 0.9184, + "step": 8714 + }, + { + "epoch": 1.5516381766381766, + "grad_norm": 0.6823606491088867, + "learning_rate": 0.00013466488598707876, + "loss": 0.9542, + "step": 8715 + }, + { + "epoch": 1.5518162393162394, + "grad_norm": 0.6781585812568665, + "learning_rate": 0.0001346517560890099, + "loss": 1.1761, + "step": 8716 + }, + { + "epoch": 1.551994301994302, + "grad_norm": 0.6313831806182861, + "learning_rate": 0.00013463862551201184, + "loss": 0.8935, + "step": 8717 + }, + { + "epoch": 1.5521723646723646, + "grad_norm": 0.7466186881065369, + "learning_rate": 0.0001346254942563419, + "loss": 1.0583, + "step": 8718 + }, + { + "epoch": 1.5523504273504274, + "grad_norm": 0.7073680758476257, + "learning_rate": 0.0001346123623222573, + "loss": 0.9863, + "step": 8719 + }, + { + "epoch": 1.55252849002849, + "grad_norm": 0.6286870241165161, + "learning_rate": 0.00013459922971001536, + "loss": 0.9921, + "step": 8720 + }, + { + "epoch": 1.5527065527065527, + "grad_norm": 0.6047035455703735, + "learning_rate": 0.0001345860964198734, + "loss": 0.9155, + "step": 8721 + }, + { + "epoch": 1.5528846153846154, + "grad_norm": 0.5909964442253113, + "learning_rate": 0.00013457296245208874, + "loss": 0.9593, + "step": 8722 + }, + { + "epoch": 1.5530626780626782, + "grad_norm": 0.7838597893714905, + "learning_rate": 0.00013455982780691869, + "loss": 0.8872, + "step": 8723 + }, + { + "epoch": 1.5532407407407407, + "grad_norm": 0.6914706230163574, + "learning_rate": 0.00013454669248462063, + "loss": 0.9104, + "step": 8724 + }, + { + "epoch": 1.5534188034188035, + "grad_norm": 0.6777952909469604, + "learning_rate": 0.00013453355648545182, + "loss": 0.9839, + "step": 8725 + }, + { + "epoch": 1.553596866096866, + "grad_norm": 0.7482799291610718, + "learning_rate": 0.00013452041980966978, + "loss": 1.1164, + "step": 8726 + }, + { + "epoch": 1.5537749287749287, + "grad_norm": 0.6616327166557312, + "learning_rate": 0.0001345072824575318, + "loss": 0.9574, + "step": 8727 + }, + { + "epoch": 1.5539529914529915, + "grad_norm": 0.7193203568458557, + "learning_rate": 0.00013449414442929532, + "loss": 1.0609, + "step": 8728 + }, + { + "epoch": 1.5541310541310542, + "grad_norm": 0.6599446535110474, + "learning_rate": 0.0001344810057252177, + "loss": 0.9574, + "step": 8729 + }, + { + "epoch": 1.5543091168091168, + "grad_norm": 0.7221707105636597, + "learning_rate": 0.00013446786634555642, + "loss": 0.9819, + "step": 8730 + }, + { + "epoch": 1.5544871794871795, + "grad_norm": 0.6531312465667725, + "learning_rate": 0.0001344547262905689, + "loss": 0.9986, + "step": 8731 + }, + { + "epoch": 1.554665242165242, + "grad_norm": 0.6879804730415344, + "learning_rate": 0.0001344415855605126, + "loss": 1.1078, + "step": 8732 + }, + { + "epoch": 1.5548433048433048, + "grad_norm": 0.708907425403595, + "learning_rate": 0.00013442844415564498, + "loss": 1.0221, + "step": 8733 + }, + { + "epoch": 1.5550213675213675, + "grad_norm": 0.7957375645637512, + "learning_rate": 0.0001344153020762235, + "loss": 1.3101, + "step": 8734 + }, + { + "epoch": 1.5551994301994303, + "grad_norm": 0.7068197727203369, + "learning_rate": 0.00013440215932250567, + "loss": 0.8995, + "step": 8735 + }, + { + "epoch": 1.5553774928774928, + "grad_norm": 0.6455841064453125, + "learning_rate": 0.00013438901589474898, + "loss": 0.7244, + "step": 8736 + }, + { + "epoch": 1.5555555555555556, + "grad_norm": 0.7500516772270203, + "learning_rate": 0.00013437587179321097, + "loss": 1.0161, + "step": 8737 + }, + { + "epoch": 1.555733618233618, + "grad_norm": 0.5983143448829651, + "learning_rate": 0.00013436272701814917, + "loss": 0.9922, + "step": 8738 + }, + { + "epoch": 1.5559116809116809, + "grad_norm": 0.8761729598045349, + "learning_rate": 0.0001343495815698211, + "loss": 1.022, + "step": 8739 + }, + { + "epoch": 1.5560897435897436, + "grad_norm": 0.6901857852935791, + "learning_rate": 0.00013433643544848438, + "loss": 1.0668, + "step": 8740 + }, + { + "epoch": 1.5562678062678064, + "grad_norm": 0.6770836114883423, + "learning_rate": 0.00013432328865439647, + "loss": 0.9516, + "step": 8741 + }, + { + "epoch": 1.556445868945869, + "grad_norm": 0.6138805150985718, + "learning_rate": 0.00013431014118781505, + "loss": 0.8682, + "step": 8742 + }, + { + "epoch": 1.5566239316239316, + "grad_norm": 0.6796693801879883, + "learning_rate": 0.00013429699304899772, + "loss": 1.1132, + "step": 8743 + }, + { + "epoch": 1.5568019943019942, + "grad_norm": 0.6626394987106323, + "learning_rate": 0.000134283844238202, + "loss": 0.9273, + "step": 8744 + }, + { + "epoch": 1.556980056980057, + "grad_norm": 0.7088519334793091, + "learning_rate": 0.00013427069475568563, + "loss": 0.8915, + "step": 8745 + }, + { + "epoch": 1.5571581196581197, + "grad_norm": 0.6244857311248779, + "learning_rate": 0.0001342575446017061, + "loss": 0.9466, + "step": 8746 + }, + { + "epoch": 1.5573361823361824, + "grad_norm": 0.6969038248062134, + "learning_rate": 0.00013424439377652123, + "loss": 1.2307, + "step": 8747 + }, + { + "epoch": 1.5575142450142452, + "grad_norm": 0.6636740565299988, + "learning_rate": 0.0001342312422803886, + "loss": 0.9456, + "step": 8748 + }, + { + "epoch": 1.5576923076923077, + "grad_norm": 0.7863389253616333, + "learning_rate": 0.00013421809011356586, + "loss": 1.1888, + "step": 8749 + }, + { + "epoch": 1.5578703703703702, + "grad_norm": 0.7504058480262756, + "learning_rate": 0.00013420493727631073, + "loss": 1.2602, + "step": 8750 + }, + { + "epoch": 1.558048433048433, + "grad_norm": 0.7173139452934265, + "learning_rate": 0.00013419178376888085, + "loss": 1.0726, + "step": 8751 + }, + { + "epoch": 1.5582264957264957, + "grad_norm": 0.6517474055290222, + "learning_rate": 0.00013417862959153406, + "loss": 1.1299, + "step": 8752 + }, + { + "epoch": 1.5584045584045585, + "grad_norm": 0.8911739587783813, + "learning_rate": 0.00013416547474452803, + "loss": 1.105, + "step": 8753 + }, + { + "epoch": 1.5585826210826212, + "grad_norm": 0.7116649150848389, + "learning_rate": 0.00013415231922812049, + "loss": 0.8037, + "step": 8754 + }, + { + "epoch": 1.5587606837606838, + "grad_norm": 0.6935904026031494, + "learning_rate": 0.00013413916304256916, + "loss": 1.2778, + "step": 8755 + }, + { + "epoch": 1.5589387464387463, + "grad_norm": 0.652763843536377, + "learning_rate": 0.00013412600618813186, + "loss": 0.9188, + "step": 8756 + }, + { + "epoch": 1.559116809116809, + "grad_norm": 0.6545276641845703, + "learning_rate": 0.00013411284866506637, + "loss": 1.0116, + "step": 8757 + }, + { + "epoch": 1.5592948717948718, + "grad_norm": 0.632165253162384, + "learning_rate": 0.0001340996904736305, + "loss": 0.8538, + "step": 8758 + }, + { + "epoch": 1.5594729344729346, + "grad_norm": 0.6719664931297302, + "learning_rate": 0.000134086531614082, + "loss": 1.1877, + "step": 8759 + }, + { + "epoch": 1.5596509971509973, + "grad_norm": 0.6691158413887024, + "learning_rate": 0.00013407337208667873, + "loss": 1.0411, + "step": 8760 + }, + { + "epoch": 1.5598290598290598, + "grad_norm": 0.7711479067802429, + "learning_rate": 0.0001340602118916785, + "loss": 0.9995, + "step": 8761 + }, + { + "epoch": 1.5600071225071224, + "grad_norm": 0.7229881286621094, + "learning_rate": 0.0001340470510293392, + "loss": 1.1751, + "step": 8762 + }, + { + "epoch": 1.5601851851851851, + "grad_norm": 0.7183271646499634, + "learning_rate": 0.00013403388949991864, + "loss": 0.9371, + "step": 8763 + }, + { + "epoch": 1.5603632478632479, + "grad_norm": 0.8142383098602295, + "learning_rate": 0.00013402072730367475, + "loss": 1.0199, + "step": 8764 + }, + { + "epoch": 1.5605413105413106, + "grad_norm": 0.6349362134933472, + "learning_rate": 0.00013400756444086534, + "loss": 0.8453, + "step": 8765 + }, + { + "epoch": 1.5607193732193734, + "grad_norm": 0.651900589466095, + "learning_rate": 0.00013399440091174834, + "loss": 0.8952, + "step": 8766 + }, + { + "epoch": 1.560897435897436, + "grad_norm": 0.6873346567153931, + "learning_rate": 0.00013398123671658172, + "loss": 0.9438, + "step": 8767 + }, + { + "epoch": 1.5610754985754984, + "grad_norm": 0.7404754757881165, + "learning_rate": 0.00013396807185562333, + "loss": 1.123, + "step": 8768 + }, + { + "epoch": 1.5612535612535612, + "grad_norm": 0.7449641227722168, + "learning_rate": 0.00013395490632913111, + "loss": 0.9407, + "step": 8769 + }, + { + "epoch": 1.561431623931624, + "grad_norm": 0.7393384575843811, + "learning_rate": 0.0001339417401373631, + "loss": 1.0209, + "step": 8770 + }, + { + "epoch": 1.5616096866096867, + "grad_norm": 0.6787426471710205, + "learning_rate": 0.00013392857328057713, + "loss": 0.9768, + "step": 8771 + }, + { + "epoch": 1.5617877492877494, + "grad_norm": 0.6295693516731262, + "learning_rate": 0.00013391540575903127, + "loss": 0.9011, + "step": 8772 + }, + { + "epoch": 1.561965811965812, + "grad_norm": 0.7114503979682922, + "learning_rate": 0.00013390223757298354, + "loss": 1.0696, + "step": 8773 + }, + { + "epoch": 1.5621438746438745, + "grad_norm": 0.7540110349655151, + "learning_rate": 0.00013388906872269184, + "loss": 1.0071, + "step": 8774 + }, + { + "epoch": 1.5623219373219372, + "grad_norm": 0.6472305059432983, + "learning_rate": 0.00013387589920841423, + "loss": 1.105, + "step": 8775 + }, + { + "epoch": 1.5625, + "grad_norm": 0.6936793327331543, + "learning_rate": 0.00013386272903040874, + "loss": 0.885, + "step": 8776 + }, + { + "epoch": 1.5626780626780628, + "grad_norm": 0.7487989068031311, + "learning_rate": 0.00013384955818893343, + "loss": 0.7842, + "step": 8777 + }, + { + "epoch": 1.5628561253561255, + "grad_norm": 0.6109505891799927, + "learning_rate": 0.00013383638668424633, + "loss": 0.9461, + "step": 8778 + }, + { + "epoch": 1.563034188034188, + "grad_norm": 0.6650055646896362, + "learning_rate": 0.00013382321451660558, + "loss": 1.0463, + "step": 8779 + }, + { + "epoch": 1.5632122507122506, + "grad_norm": 0.7147329449653625, + "learning_rate": 0.00013381004168626915, + "loss": 0.946, + "step": 8780 + }, + { + "epoch": 1.5633903133903133, + "grad_norm": 0.6919382810592651, + "learning_rate": 0.00013379686819349522, + "loss": 0.8946, + "step": 8781 + }, + { + "epoch": 1.563568376068376, + "grad_norm": 0.7339401245117188, + "learning_rate": 0.00013378369403854184, + "loss": 0.9625, + "step": 8782 + }, + { + "epoch": 1.5637464387464388, + "grad_norm": 0.6337129473686218, + "learning_rate": 0.00013377051922166717, + "loss": 1.0854, + "step": 8783 + }, + { + "epoch": 1.5639245014245016, + "grad_norm": 0.7301266193389893, + "learning_rate": 0.0001337573437431293, + "loss": 1.017, + "step": 8784 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 0.689540684223175, + "learning_rate": 0.00013374416760318644, + "loss": 0.8734, + "step": 8785 + }, + { + "epoch": 1.5642806267806266, + "grad_norm": 0.7121307849884033, + "learning_rate": 0.0001337309908020967, + "loss": 1.0827, + "step": 8786 + }, + { + "epoch": 1.5644586894586894, + "grad_norm": 0.6715386509895325, + "learning_rate": 0.00013371781334011826, + "loss": 0.946, + "step": 8787 + }, + { + "epoch": 1.5646367521367521, + "grad_norm": 0.6895501613616943, + "learning_rate": 0.00013370463521750932, + "loss": 1.1113, + "step": 8788 + }, + { + "epoch": 1.5648148148148149, + "grad_norm": 0.6592531204223633, + "learning_rate": 0.00013369145643452805, + "loss": 0.9952, + "step": 8789 + }, + { + "epoch": 1.5649928774928776, + "grad_norm": 0.7495190501213074, + "learning_rate": 0.0001336782769914327, + "loss": 1.0936, + "step": 8790 + }, + { + "epoch": 1.5651709401709402, + "grad_norm": 0.7273977398872375, + "learning_rate": 0.00013366509688848147, + "loss": 1.1749, + "step": 8791 + }, + { + "epoch": 1.5653490028490027, + "grad_norm": 0.6447354555130005, + "learning_rate": 0.0001336519161259326, + "loss": 0.8638, + "step": 8792 + }, + { + "epoch": 1.5655270655270654, + "grad_norm": 0.6572020053863525, + "learning_rate": 0.00013363873470404432, + "loss": 0.8005, + "step": 8793 + }, + { + "epoch": 1.5657051282051282, + "grad_norm": 0.676418662071228, + "learning_rate": 0.00013362555262307491, + "loss": 0.7651, + "step": 8794 + }, + { + "epoch": 1.565883190883191, + "grad_norm": 0.6886745095252991, + "learning_rate": 0.0001336123698832827, + "loss": 1.0765, + "step": 8795 + }, + { + "epoch": 1.5660612535612537, + "grad_norm": 0.8134182095527649, + "learning_rate": 0.00013359918648492584, + "loss": 1.2228, + "step": 8796 + }, + { + "epoch": 1.5662393162393162, + "grad_norm": 0.7210384011268616, + "learning_rate": 0.00013358600242826277, + "loss": 0.8247, + "step": 8797 + }, + { + "epoch": 1.5664173789173788, + "grad_norm": 0.7086136341094971, + "learning_rate": 0.00013357281771355175, + "loss": 1.0323, + "step": 8798 + }, + { + "epoch": 1.5665954415954415, + "grad_norm": 0.7419785857200623, + "learning_rate": 0.0001335596323410511, + "loss": 1.213, + "step": 8799 + }, + { + "epoch": 1.5667735042735043, + "grad_norm": 0.6390291452407837, + "learning_rate": 0.0001335464463110192, + "loss": 1.0403, + "step": 8800 + }, + { + "epoch": 1.566951566951567, + "grad_norm": 0.6111941337585449, + "learning_rate": 0.00013353325962371434, + "loss": 0.9747, + "step": 8801 + }, + { + "epoch": 1.5671296296296298, + "grad_norm": 0.6792671084403992, + "learning_rate": 0.00013352007227939488, + "loss": 1.1179, + "step": 8802 + }, + { + "epoch": 1.5673076923076923, + "grad_norm": 0.6656535863876343, + "learning_rate": 0.0001335068842783193, + "loss": 0.9214, + "step": 8803 + }, + { + "epoch": 1.5674857549857548, + "grad_norm": 0.6910907626152039, + "learning_rate": 0.0001334936956207459, + "loss": 1.0609, + "step": 8804 + }, + { + "epoch": 1.5676638176638176, + "grad_norm": 0.65049147605896, + "learning_rate": 0.00013348050630693315, + "loss": 0.7189, + "step": 8805 + }, + { + "epoch": 1.5678418803418803, + "grad_norm": 0.6258065104484558, + "learning_rate": 0.0001334673163371394, + "loss": 1.0683, + "step": 8806 + }, + { + "epoch": 1.568019943019943, + "grad_norm": 0.7518934607505798, + "learning_rate": 0.00013345412571162305, + "loss": 1.2415, + "step": 8807 + }, + { + "epoch": 1.5681980056980058, + "grad_norm": 0.7395275235176086, + "learning_rate": 0.00013344093443064267, + "loss": 0.9153, + "step": 8808 + }, + { + "epoch": 1.5683760683760684, + "grad_norm": 0.6789839267730713, + "learning_rate": 0.00013342774249445663, + "loss": 0.8051, + "step": 8809 + }, + { + "epoch": 1.568554131054131, + "grad_norm": 0.786247193813324, + "learning_rate": 0.00013341454990332342, + "loss": 1.203, + "step": 8810 + }, + { + "epoch": 1.5687321937321936, + "grad_norm": 0.6858161687850952, + "learning_rate": 0.00013340135665750153, + "loss": 0.9494, + "step": 8811 + }, + { + "epoch": 1.5689102564102564, + "grad_norm": 0.7245797514915466, + "learning_rate": 0.0001333881627572494, + "loss": 1.0544, + "step": 8812 + }, + { + "epoch": 1.5690883190883191, + "grad_norm": 0.6176164150238037, + "learning_rate": 0.00013337496820282563, + "loss": 0.9084, + "step": 8813 + }, + { + "epoch": 1.569266381766382, + "grad_norm": 0.7342953681945801, + "learning_rate": 0.00013336177299448868, + "loss": 1.0006, + "step": 8814 + }, + { + "epoch": 1.5694444444444444, + "grad_norm": 0.5183523297309875, + "learning_rate": 0.00013334857713249708, + "loss": 0.6295, + "step": 8815 + }, + { + "epoch": 1.5696225071225072, + "grad_norm": 0.6664513349533081, + "learning_rate": 0.00013333538061710936, + "loss": 0.7569, + "step": 8816 + }, + { + "epoch": 1.5698005698005697, + "grad_norm": 0.7051160931587219, + "learning_rate": 0.0001333221834485841, + "loss": 0.9917, + "step": 8817 + }, + { + "epoch": 1.5699786324786325, + "grad_norm": 0.7888057231903076, + "learning_rate": 0.0001333089856271799, + "loss": 1.0337, + "step": 8818 + }, + { + "epoch": 1.5701566951566952, + "grad_norm": 0.6796144247055054, + "learning_rate": 0.00013329578715315534, + "loss": 1.0915, + "step": 8819 + }, + { + "epoch": 1.570334757834758, + "grad_norm": 0.7442883849143982, + "learning_rate": 0.000133282588026769, + "loss": 1.1695, + "step": 8820 + }, + { + "epoch": 1.5705128205128205, + "grad_norm": 0.6164735555648804, + "learning_rate": 0.00013326938824827946, + "loss": 1.0143, + "step": 8821 + }, + { + "epoch": 1.5706908831908832, + "grad_norm": 0.6526502966880798, + "learning_rate": 0.00013325618781794539, + "loss": 0.8402, + "step": 8822 + }, + { + "epoch": 1.5708689458689458, + "grad_norm": 0.6376087069511414, + "learning_rate": 0.00013324298673602535, + "loss": 0.7582, + "step": 8823 + }, + { + "epoch": 1.5710470085470085, + "grad_norm": 0.6888708472251892, + "learning_rate": 0.00013322978500277807, + "loss": 0.997, + "step": 8824 + }, + { + "epoch": 1.5712250712250713, + "grad_norm": 0.553656280040741, + "learning_rate": 0.0001332165826184622, + "loss": 0.6917, + "step": 8825 + }, + { + "epoch": 1.571403133903134, + "grad_norm": 0.643285870552063, + "learning_rate": 0.0001332033795833364, + "loss": 0.8689, + "step": 8826 + }, + { + "epoch": 1.5715811965811965, + "grad_norm": 0.6210280060768127, + "learning_rate": 0.00013319017589765933, + "loss": 0.9047, + "step": 8827 + }, + { + "epoch": 1.5717592592592593, + "grad_norm": 0.7612366676330566, + "learning_rate": 0.0001331769715616897, + "loss": 0.9818, + "step": 8828 + }, + { + "epoch": 1.5719373219373218, + "grad_norm": 0.5970702171325684, + "learning_rate": 0.00013316376657568628, + "loss": 0.82, + "step": 8829 + }, + { + "epoch": 1.5721153846153846, + "grad_norm": 0.7182583808898926, + "learning_rate": 0.0001331505609399077, + "loss": 1.0633, + "step": 8830 + }, + { + "epoch": 1.5722934472934473, + "grad_norm": 0.7230739593505859, + "learning_rate": 0.00013313735465461278, + "loss": 0.977, + "step": 8831 + }, + { + "epoch": 1.57247150997151, + "grad_norm": 0.6752985119819641, + "learning_rate": 0.00013312414772006018, + "loss": 0.9666, + "step": 8832 + }, + { + "epoch": 1.5726495726495726, + "grad_norm": 0.7724275588989258, + "learning_rate": 0.00013311094013650877, + "loss": 1.148, + "step": 8833 + }, + { + "epoch": 1.5728276353276354, + "grad_norm": 0.7216386198997498, + "learning_rate": 0.00013309773190421724, + "loss": 0.9935, + "step": 8834 + }, + { + "epoch": 1.573005698005698, + "grad_norm": 0.6422320008277893, + "learning_rate": 0.0001330845230234444, + "loss": 0.9383, + "step": 8835 + }, + { + "epoch": 1.5731837606837606, + "grad_norm": 0.669538140296936, + "learning_rate": 0.00013307131349444906, + "loss": 1.0866, + "step": 8836 + }, + { + "epoch": 1.5733618233618234, + "grad_norm": 0.6994584798812866, + "learning_rate": 0.00013305810331749003, + "loss": 0.7882, + "step": 8837 + }, + { + "epoch": 1.5735398860398861, + "grad_norm": 0.8094269633293152, + "learning_rate": 0.00013304489249282617, + "loss": 1.2316, + "step": 8838 + }, + { + "epoch": 1.5737179487179487, + "grad_norm": 0.7180120348930359, + "learning_rate": 0.00013303168102071625, + "loss": 0.9795, + "step": 8839 + }, + { + "epoch": 1.5738960113960114, + "grad_norm": 0.6191438436508179, + "learning_rate": 0.00013301846890141918, + "loss": 0.8957, + "step": 8840 + }, + { + "epoch": 1.574074074074074, + "grad_norm": 0.671094536781311, + "learning_rate": 0.00013300525613519382, + "loss": 1.059, + "step": 8841 + }, + { + "epoch": 1.5742521367521367, + "grad_norm": 0.8062624931335449, + "learning_rate": 0.000132992042722299, + "loss": 0.9782, + "step": 8842 + }, + { + "epoch": 1.5744301994301995, + "grad_norm": 0.6674807667732239, + "learning_rate": 0.00013297882866299362, + "loss": 0.7765, + "step": 8843 + }, + { + "epoch": 1.5746082621082622, + "grad_norm": 0.6369131803512573, + "learning_rate": 0.00013296561395753664, + "loss": 0.97, + "step": 8844 + }, + { + "epoch": 1.5747863247863247, + "grad_norm": 0.7913636565208435, + "learning_rate": 0.00013295239860618691, + "loss": 1.0458, + "step": 8845 + }, + { + "epoch": 1.5749643874643875, + "grad_norm": 0.6722261905670166, + "learning_rate": 0.0001329391826092034, + "loss": 1.1118, + "step": 8846 + }, + { + "epoch": 1.57514245014245, + "grad_norm": 0.6936299800872803, + "learning_rate": 0.00013292596596684502, + "loss": 1.009, + "step": 8847 + }, + { + "epoch": 1.5753205128205128, + "grad_norm": 0.7009961009025574, + "learning_rate": 0.00013291274867937073, + "loss": 0.9904, + "step": 8848 + }, + { + "epoch": 1.5754985754985755, + "grad_norm": 0.6900732517242432, + "learning_rate": 0.0001328995307470395, + "loss": 1.0488, + "step": 8849 + }, + { + "epoch": 1.5756766381766383, + "grad_norm": 0.6389018297195435, + "learning_rate": 0.00013288631217011032, + "loss": 0.9444, + "step": 8850 + }, + { + "epoch": 1.5758547008547008, + "grad_norm": 0.6370900869369507, + "learning_rate": 0.00013287309294884216, + "loss": 0.7465, + "step": 8851 + }, + { + "epoch": 1.5760327635327636, + "grad_norm": 0.6463848948478699, + "learning_rate": 0.00013285987308349405, + "loss": 0.896, + "step": 8852 + }, + { + "epoch": 1.576210826210826, + "grad_norm": 0.6022449731826782, + "learning_rate": 0.00013284665257432495, + "loss": 0.8822, + "step": 8853 + }, + { + "epoch": 1.5763888888888888, + "grad_norm": 0.768189013004303, + "learning_rate": 0.00013283343142159396, + "loss": 0.9862, + "step": 8854 + }, + { + "epoch": 1.5765669515669516, + "grad_norm": 0.6642358303070068, + "learning_rate": 0.00013282020962556007, + "loss": 1.0713, + "step": 8855 + }, + { + "epoch": 1.5767450142450143, + "grad_norm": 0.6883034706115723, + "learning_rate": 0.00013280698718648234, + "loss": 1.0351, + "step": 8856 + }, + { + "epoch": 1.5769230769230769, + "grad_norm": 0.602808952331543, + "learning_rate": 0.00013279376410461988, + "loss": 0.7615, + "step": 8857 + }, + { + "epoch": 1.5771011396011396, + "grad_norm": 0.5968614220619202, + "learning_rate": 0.0001327805403802317, + "loss": 0.9443, + "step": 8858 + }, + { + "epoch": 1.5772792022792022, + "grad_norm": 0.7314837574958801, + "learning_rate": 0.00013276731601357696, + "loss": 0.8784, + "step": 8859 + }, + { + "epoch": 1.577457264957265, + "grad_norm": 0.619754433631897, + "learning_rate": 0.0001327540910049147, + "loss": 0.954, + "step": 8860 + }, + { + "epoch": 1.5776353276353277, + "grad_norm": 0.7195139527320862, + "learning_rate": 0.0001327408653545041, + "loss": 1.0227, + "step": 8861 + }, + { + "epoch": 1.5778133903133904, + "grad_norm": 0.6796214580535889, + "learning_rate": 0.0001327276390626042, + "loss": 1.0593, + "step": 8862 + }, + { + "epoch": 1.577991452991453, + "grad_norm": 0.6576255559921265, + "learning_rate": 0.00013271441212947427, + "loss": 0.7921, + "step": 8863 + }, + { + "epoch": 1.5781695156695157, + "grad_norm": 0.7222092151641846, + "learning_rate": 0.00013270118455537336, + "loss": 1.0545, + "step": 8864 + }, + { + "epoch": 1.5783475783475782, + "grad_norm": 0.7159737348556519, + "learning_rate": 0.00013268795634056066, + "loss": 0.9664, + "step": 8865 + }, + { + "epoch": 1.578525641025641, + "grad_norm": 0.7120481133460999, + "learning_rate": 0.00013267472748529536, + "loss": 1.0148, + "step": 8866 + }, + { + "epoch": 1.5787037037037037, + "grad_norm": 0.7353253364562988, + "learning_rate": 0.00013266149798983666, + "loss": 0.9288, + "step": 8867 + }, + { + "epoch": 1.5788817663817665, + "grad_norm": 0.6652441620826721, + "learning_rate": 0.00013264826785444375, + "loss": 0.8246, + "step": 8868 + }, + { + "epoch": 1.5790598290598292, + "grad_norm": 0.7254189252853394, + "learning_rate": 0.00013263503707937584, + "loss": 0.9892, + "step": 8869 + }, + { + "epoch": 1.5792378917378918, + "grad_norm": 0.6305747032165527, + "learning_rate": 0.00013262180566489223, + "loss": 0.8931, + "step": 8870 + }, + { + "epoch": 1.5794159544159543, + "grad_norm": 0.6560617089271545, + "learning_rate": 0.00013260857361125205, + "loss": 0.9245, + "step": 8871 + }, + { + "epoch": 1.579594017094017, + "grad_norm": 0.7304151654243469, + "learning_rate": 0.00013259534091871462, + "loss": 1.009, + "step": 8872 + }, + { + "epoch": 1.5797720797720798, + "grad_norm": 0.782636821269989, + "learning_rate": 0.00013258210758753918, + "loss": 1.1123, + "step": 8873 + }, + { + "epoch": 1.5799501424501425, + "grad_norm": 0.6992011070251465, + "learning_rate": 0.00013256887361798504, + "loss": 1.099, + "step": 8874 + }, + { + "epoch": 1.5801282051282053, + "grad_norm": 0.7159731984138489, + "learning_rate": 0.00013255563901031148, + "loss": 1.0257, + "step": 8875 + }, + { + "epoch": 1.5803062678062678, + "grad_norm": 0.6055454611778259, + "learning_rate": 0.0001325424037647778, + "loss": 0.9199, + "step": 8876 + }, + { + "epoch": 1.5804843304843303, + "grad_norm": 0.6838310360908508, + "learning_rate": 0.00013252916788164334, + "loss": 0.8644, + "step": 8877 + }, + { + "epoch": 1.580662393162393, + "grad_norm": 0.7067445516586304, + "learning_rate": 0.00013251593136116738, + "loss": 1.0285, + "step": 8878 + }, + { + "epoch": 1.5808404558404558, + "grad_norm": 0.7021774649620056, + "learning_rate": 0.00013250269420360928, + "loss": 1.1263, + "step": 8879 + }, + { + "epoch": 1.5810185185185186, + "grad_norm": 0.6586757302284241, + "learning_rate": 0.00013248945640922843, + "loss": 0.906, + "step": 8880 + }, + { + "epoch": 1.5811965811965814, + "grad_norm": 0.6673910021781921, + "learning_rate": 0.00013247621797828418, + "loss": 1.0652, + "step": 8881 + }, + { + "epoch": 1.5813746438746439, + "grad_norm": 0.6763964295387268, + "learning_rate": 0.00013246297891103588, + "loss": 1.0227, + "step": 8882 + }, + { + "epoch": 1.5815527065527064, + "grad_norm": 0.6536892056465149, + "learning_rate": 0.00013244973920774298, + "loss": 0.9026, + "step": 8883 + }, + { + "epoch": 1.5817307692307692, + "grad_norm": 0.8010411858558655, + "learning_rate": 0.0001324364988686648, + "loss": 1.1167, + "step": 8884 + }, + { + "epoch": 1.581908831908832, + "grad_norm": 0.8159251809120178, + "learning_rate": 0.00013242325789406082, + "loss": 1.233, + "step": 8885 + }, + { + "epoch": 1.5820868945868947, + "grad_norm": 0.6487745046615601, + "learning_rate": 0.00013241001628419048, + "loss": 0.9888, + "step": 8886 + }, + { + "epoch": 1.5822649572649574, + "grad_norm": 0.6750285029411316, + "learning_rate": 0.00013239677403931318, + "loss": 0.8874, + "step": 8887 + }, + { + "epoch": 1.58244301994302, + "grad_norm": 0.7164602875709534, + "learning_rate": 0.0001323835311596884, + "loss": 1.2029, + "step": 8888 + }, + { + "epoch": 1.5826210826210825, + "grad_norm": 0.6081351041793823, + "learning_rate": 0.00013237028764557558, + "loss": 0.9593, + "step": 8889 + }, + { + "epoch": 1.5827991452991452, + "grad_norm": 0.7235409021377563, + "learning_rate": 0.00013235704349723424, + "loss": 1.5324, + "step": 8890 + }, + { + "epoch": 1.582977207977208, + "grad_norm": 0.6658480763435364, + "learning_rate": 0.0001323437987149238, + "loss": 0.9756, + "step": 8891 + }, + { + "epoch": 1.5831552706552707, + "grad_norm": 0.7924265265464783, + "learning_rate": 0.00013233055329890387, + "loss": 0.9329, + "step": 8892 + }, + { + "epoch": 1.5833333333333335, + "grad_norm": 0.6262093186378479, + "learning_rate": 0.0001323173072494339, + "loss": 0.8288, + "step": 8893 + }, + { + "epoch": 1.583511396011396, + "grad_norm": 0.6851989030838013, + "learning_rate": 0.0001323040605667734, + "loss": 0.9822, + "step": 8894 + }, + { + "epoch": 1.5836894586894585, + "grad_norm": 0.6963728666305542, + "learning_rate": 0.00013229081325118194, + "loss": 1.0416, + "step": 8895 + }, + { + "epoch": 1.5838675213675213, + "grad_norm": 0.6017457842826843, + "learning_rate": 0.0001322775653029191, + "loss": 0.8123, + "step": 8896 + }, + { + "epoch": 1.584045584045584, + "grad_norm": 0.7396472096443176, + "learning_rate": 0.0001322643167222444, + "loss": 1.0339, + "step": 8897 + }, + { + "epoch": 1.5842236467236468, + "grad_norm": 0.6360299587249756, + "learning_rate": 0.00013225106750941744, + "loss": 0.9463, + "step": 8898 + }, + { + "epoch": 1.5844017094017095, + "grad_norm": 0.6297624111175537, + "learning_rate": 0.00013223781766469783, + "loss": 0.9921, + "step": 8899 + }, + { + "epoch": 1.584579772079772, + "grad_norm": 0.7722037434577942, + "learning_rate": 0.0001322245671883451, + "loss": 0.8394, + "step": 8900 + }, + { + "epoch": 1.5847578347578346, + "grad_norm": 0.677364706993103, + "learning_rate": 0.00013221131608061895, + "loss": 1.0954, + "step": 8901 + }, + { + "epoch": 1.5849358974358974, + "grad_norm": 0.6954908967018127, + "learning_rate": 0.00013219806434177899, + "loss": 1.0637, + "step": 8902 + }, + { + "epoch": 1.58511396011396, + "grad_norm": 0.7079192996025085, + "learning_rate": 0.00013218481197208484, + "loss": 1.039, + "step": 8903 + }, + { + "epoch": 1.5852920227920229, + "grad_norm": 0.7070451378822327, + "learning_rate": 0.00013217155897179611, + "loss": 1.0025, + "step": 8904 + }, + { + "epoch": 1.5854700854700856, + "grad_norm": 0.6940776705741882, + "learning_rate": 0.00013215830534117257, + "loss": 0.8039, + "step": 8905 + }, + { + "epoch": 1.5856481481481481, + "grad_norm": 0.6545892953872681, + "learning_rate": 0.00013214505108047382, + "loss": 0.9347, + "step": 8906 + }, + { + "epoch": 1.5858262108262107, + "grad_norm": 0.6769635081291199, + "learning_rate": 0.00013213179618995957, + "loss": 1.0321, + "step": 8907 + }, + { + "epoch": 1.5860042735042734, + "grad_norm": 0.6505448222160339, + "learning_rate": 0.00013211854066988953, + "loss": 1.0558, + "step": 8908 + }, + { + "epoch": 1.5861823361823362, + "grad_norm": 0.6764090061187744, + "learning_rate": 0.00013210528452052336, + "loss": 0.8407, + "step": 8909 + }, + { + "epoch": 1.586360398860399, + "grad_norm": 0.6454851627349854, + "learning_rate": 0.00013209202774212088, + "loss": 0.7439, + "step": 8910 + }, + { + "epoch": 1.5865384615384617, + "grad_norm": 0.6911695599555969, + "learning_rate": 0.00013207877033494177, + "loss": 0.9625, + "step": 8911 + }, + { + "epoch": 1.5867165242165242, + "grad_norm": 0.7405226826667786, + "learning_rate": 0.0001320655122992458, + "loss": 1.054, + "step": 8912 + }, + { + "epoch": 1.5868945868945867, + "grad_norm": 0.7362869381904602, + "learning_rate": 0.00013205225363529274, + "loss": 1.0516, + "step": 8913 + }, + { + "epoch": 1.5870726495726495, + "grad_norm": 0.6923766136169434, + "learning_rate": 0.0001320389943433423, + "loss": 1.2323, + "step": 8914 + }, + { + "epoch": 1.5872507122507122, + "grad_norm": 0.7980395555496216, + "learning_rate": 0.00013202573442365435, + "loss": 1.0229, + "step": 8915 + }, + { + "epoch": 1.587428774928775, + "grad_norm": 0.7211610078811646, + "learning_rate": 0.00013201247387648868, + "loss": 1.0666, + "step": 8916 + }, + { + "epoch": 1.5876068376068377, + "grad_norm": 0.6728795766830444, + "learning_rate": 0.00013199921270210506, + "loss": 1.0322, + "step": 8917 + }, + { + "epoch": 1.5877849002849003, + "grad_norm": 0.6226436495780945, + "learning_rate": 0.00013198595090076337, + "loss": 1.0517, + "step": 8918 + }, + { + "epoch": 1.5879629629629628, + "grad_norm": 0.6396511197090149, + "learning_rate": 0.0001319726884727234, + "loss": 0.8662, + "step": 8919 + }, + { + "epoch": 1.5881410256410255, + "grad_norm": 0.5664374828338623, + "learning_rate": 0.00013195942541824497, + "loss": 0.6601, + "step": 8920 + }, + { + "epoch": 1.5883190883190883, + "grad_norm": 0.6556946039199829, + "learning_rate": 0.00013194616173758806, + "loss": 0.9662, + "step": 8921 + }, + { + "epoch": 1.588497150997151, + "grad_norm": 0.7332060933113098, + "learning_rate": 0.00013193289743101245, + "loss": 0.7687, + "step": 8922 + }, + { + "epoch": 1.5886752136752138, + "grad_norm": 0.6103306412696838, + "learning_rate": 0.00013191963249877805, + "loss": 0.8329, + "step": 8923 + }, + { + "epoch": 1.5888532763532763, + "grad_norm": 0.63165283203125, + "learning_rate": 0.00013190636694114475, + "loss": 0.8336, + "step": 8924 + }, + { + "epoch": 1.589031339031339, + "grad_norm": 0.6955820322036743, + "learning_rate": 0.00013189310075837246, + "loss": 1.0457, + "step": 8925 + }, + { + "epoch": 1.5892094017094016, + "grad_norm": 0.6911605596542358, + "learning_rate": 0.00013187983395072114, + "loss": 0.9389, + "step": 8926 + }, + { + "epoch": 1.5893874643874644, + "grad_norm": 0.6493414640426636, + "learning_rate": 0.00013186656651845068, + "loss": 0.9821, + "step": 8927 + }, + { + "epoch": 1.5895655270655271, + "grad_norm": 0.6168226599693298, + "learning_rate": 0.00013185329846182107, + "loss": 1.0259, + "step": 8928 + }, + { + "epoch": 1.5897435897435899, + "grad_norm": 0.6460188627243042, + "learning_rate": 0.0001318400297810922, + "loss": 0.9836, + "step": 8929 + }, + { + "epoch": 1.5899216524216524, + "grad_norm": 0.6630695462226868, + "learning_rate": 0.0001318267604765241, + "loss": 0.8936, + "step": 8930 + }, + { + "epoch": 1.5900997150997151, + "grad_norm": 0.6308651566505432, + "learning_rate": 0.00013181349054837676, + "loss": 0.9583, + "step": 8931 + }, + { + "epoch": 1.5902777777777777, + "grad_norm": 0.6508499979972839, + "learning_rate": 0.00013180021999691018, + "loss": 0.7647, + "step": 8932 + }, + { + "epoch": 1.5904558404558404, + "grad_norm": 0.6625795960426331, + "learning_rate": 0.00013178694882238432, + "loss": 1.0329, + "step": 8933 + }, + { + "epoch": 1.5906339031339032, + "grad_norm": 0.6721987128257751, + "learning_rate": 0.00013177367702505924, + "loss": 0.9377, + "step": 8934 + }, + { + "epoch": 1.590811965811966, + "grad_norm": 0.7295519709587097, + "learning_rate": 0.00013176040460519497, + "loss": 0.9396, + "step": 8935 + }, + { + "epoch": 1.5909900284900285, + "grad_norm": 0.6673944592475891, + "learning_rate": 0.0001317471315630515, + "loss": 1.0284, + "step": 8936 + }, + { + "epoch": 1.5911680911680912, + "grad_norm": 0.6858960390090942, + "learning_rate": 0.00013173385789888898, + "loss": 1.2022, + "step": 8937 + }, + { + "epoch": 1.5913461538461537, + "grad_norm": 0.5836796164512634, + "learning_rate": 0.00013172058361296743, + "loss": 1.0078, + "step": 8938 + }, + { + "epoch": 1.5915242165242165, + "grad_norm": 0.7732513546943665, + "learning_rate": 0.00013170730870554694, + "loss": 1.0912, + "step": 8939 + }, + { + "epoch": 1.5917022792022792, + "grad_norm": 0.7095892429351807, + "learning_rate": 0.0001316940331768876, + "loss": 1.0506, + "step": 8940 + }, + { + "epoch": 1.591880341880342, + "grad_norm": 0.757534384727478, + "learning_rate": 0.00013168075702724952, + "loss": 1.036, + "step": 8941 + }, + { + "epoch": 1.5920584045584045, + "grad_norm": 0.6719361543655396, + "learning_rate": 0.00013166748025689282, + "loss": 0.9406, + "step": 8942 + }, + { + "epoch": 1.5922364672364673, + "grad_norm": 0.6955735087394714, + "learning_rate": 0.00013165420286607763, + "loss": 0.9325, + "step": 8943 + }, + { + "epoch": 1.5924145299145298, + "grad_norm": 0.6810322999954224, + "learning_rate": 0.00013164092485506407, + "loss": 1.0402, + "step": 8944 + }, + { + "epoch": 1.5925925925925926, + "grad_norm": 0.6346224546432495, + "learning_rate": 0.00013162764622411233, + "loss": 0.9725, + "step": 8945 + }, + { + "epoch": 1.5927706552706553, + "grad_norm": 0.728705883026123, + "learning_rate": 0.00013161436697348258, + "loss": 0.9665, + "step": 8946 + }, + { + "epoch": 1.592948717948718, + "grad_norm": 0.6838595271110535, + "learning_rate": 0.00013160108710343494, + "loss": 0.9771, + "step": 8947 + }, + { + "epoch": 1.5931267806267806, + "grad_norm": 0.7052602767944336, + "learning_rate": 0.00013158780661422966, + "loss": 0.8819, + "step": 8948 + }, + { + "epoch": 1.5933048433048433, + "grad_norm": 0.7237630486488342, + "learning_rate": 0.00013157452550612697, + "loss": 1.0609, + "step": 8949 + }, + { + "epoch": 1.5934829059829059, + "grad_norm": 0.6554936766624451, + "learning_rate": 0.00013156124377938699, + "loss": 0.8592, + "step": 8950 + }, + { + "epoch": 1.5936609686609686, + "grad_norm": 0.6125665307044983, + "learning_rate": 0.00013154796143427, + "loss": 0.8399, + "step": 8951 + }, + { + "epoch": 1.5938390313390314, + "grad_norm": 0.6930897235870361, + "learning_rate": 0.0001315346784710363, + "loss": 0.9965, + "step": 8952 + }, + { + "epoch": 1.5940170940170941, + "grad_norm": 0.7808064818382263, + "learning_rate": 0.00013152139488994605, + "loss": 1.0527, + "step": 8953 + }, + { + "epoch": 1.5941951566951567, + "grad_norm": 0.6125522255897522, + "learning_rate": 0.0001315081106912595, + "loss": 1.1159, + "step": 8954 + }, + { + "epoch": 1.5943732193732194, + "grad_norm": 0.5863428711891174, + "learning_rate": 0.00013149482587523703, + "loss": 0.84, + "step": 8955 + }, + { + "epoch": 1.594551282051282, + "grad_norm": 0.7170202732086182, + "learning_rate": 0.00013148154044213882, + "loss": 1.0821, + "step": 8956 + }, + { + "epoch": 1.5947293447293447, + "grad_norm": 0.6409463882446289, + "learning_rate": 0.00013146825439222528, + "loss": 1.0097, + "step": 8957 + }, + { + "epoch": 1.5949074074074074, + "grad_norm": 0.7037690281867981, + "learning_rate": 0.00013145496772575666, + "loss": 1.1511, + "step": 8958 + }, + { + "epoch": 1.5950854700854702, + "grad_norm": 0.6400953531265259, + "learning_rate": 0.00013144168044299326, + "loss": 1.0809, + "step": 8959 + }, + { + "epoch": 1.5952635327635327, + "grad_norm": 0.6129940152168274, + "learning_rate": 0.00013142839254419545, + "loss": 0.8481, + "step": 8960 + }, + { + "epoch": 1.5954415954415955, + "grad_norm": 0.7452271580696106, + "learning_rate": 0.00013141510402962358, + "loss": 1.0649, + "step": 8961 + }, + { + "epoch": 1.595619658119658, + "grad_norm": 0.7407623529434204, + "learning_rate": 0.000131401814899538, + "loss": 0.9084, + "step": 8962 + }, + { + "epoch": 1.5957977207977208, + "grad_norm": 0.7103050947189331, + "learning_rate": 0.0001313885251541991, + "loss": 0.946, + "step": 8963 + }, + { + "epoch": 1.5959757834757835, + "grad_norm": 0.5566636323928833, + "learning_rate": 0.00013137523479386727, + "loss": 0.6781, + "step": 8964 + }, + { + "epoch": 1.5961538461538463, + "grad_norm": 0.8137457966804504, + "learning_rate": 0.00013136194381880288, + "loss": 0.9273, + "step": 8965 + }, + { + "epoch": 1.5963319088319088, + "grad_norm": 0.779330849647522, + "learning_rate": 0.0001313486522292663, + "loss": 1.1105, + "step": 8966 + }, + { + "epoch": 1.5965099715099715, + "grad_norm": 0.6807126998901367, + "learning_rate": 0.00013133536002551808, + "loss": 1.0728, + "step": 8967 + }, + { + "epoch": 1.596688034188034, + "grad_norm": 0.7371507287025452, + "learning_rate": 0.00013132206720781853, + "loss": 0.979, + "step": 8968 + }, + { + "epoch": 1.5968660968660968, + "grad_norm": 0.6811465620994568, + "learning_rate": 0.00013130877377642814, + "loss": 0.9821, + "step": 8969 + }, + { + "epoch": 1.5970441595441596, + "grad_norm": 0.6732743978500366, + "learning_rate": 0.00013129547973160738, + "loss": 0.8511, + "step": 8970 + }, + { + "epoch": 1.5972222222222223, + "grad_norm": 0.594901978969574, + "learning_rate": 0.0001312821850736167, + "loss": 0.9674, + "step": 8971 + }, + { + "epoch": 1.5974002849002849, + "grad_norm": 0.6743764281272888, + "learning_rate": 0.00013126888980271657, + "loss": 0.9268, + "step": 8972 + }, + { + "epoch": 1.5975783475783476, + "grad_norm": 0.7532161474227905, + "learning_rate": 0.00013125559391916752, + "loss": 1.0474, + "step": 8973 + }, + { + "epoch": 1.5977564102564101, + "grad_norm": 0.6331499814987183, + "learning_rate": 0.00013124229742323, + "loss": 1.05, + "step": 8974 + }, + { + "epoch": 1.5979344729344729, + "grad_norm": 0.7418690323829651, + "learning_rate": 0.0001312290003151646, + "loss": 0.9475, + "step": 8975 + }, + { + "epoch": 1.5981125356125356, + "grad_norm": 0.6511179804801941, + "learning_rate": 0.0001312157025952318, + "loss": 0.9206, + "step": 8976 + }, + { + "epoch": 1.5982905982905984, + "grad_norm": 0.6380775570869446, + "learning_rate": 0.00013120240426369215, + "loss": 0.9953, + "step": 8977 + }, + { + "epoch": 1.598468660968661, + "grad_norm": 0.8483675122261047, + "learning_rate": 0.00013118910532080623, + "loss": 0.9454, + "step": 8978 + }, + { + "epoch": 1.5986467236467237, + "grad_norm": 0.6700518727302551, + "learning_rate": 0.00013117580576683455, + "loss": 1.0413, + "step": 8979 + }, + { + "epoch": 1.5988247863247862, + "grad_norm": 0.7750083208084106, + "learning_rate": 0.00013116250560203774, + "loss": 1.1868, + "step": 8980 + }, + { + "epoch": 1.599002849002849, + "grad_norm": 0.7474972009658813, + "learning_rate": 0.00013114920482667635, + "loss": 1.0876, + "step": 8981 + }, + { + "epoch": 1.5991809116809117, + "grad_norm": 0.6920070052146912, + "learning_rate": 0.000131135903441011, + "loss": 1.0787, + "step": 8982 + }, + { + "epoch": 1.5993589743589745, + "grad_norm": 0.7572436928749084, + "learning_rate": 0.00013112260144530232, + "loss": 0.9798, + "step": 8983 + }, + { + "epoch": 1.5995370370370372, + "grad_norm": 0.6983019709587097, + "learning_rate": 0.00013110929883981088, + "loss": 1.1115, + "step": 8984 + }, + { + "epoch": 1.5997150997150997, + "grad_norm": 0.6352120041847229, + "learning_rate": 0.0001310959956247974, + "loss": 0.9962, + "step": 8985 + }, + { + "epoch": 1.5998931623931623, + "grad_norm": 0.596858561038971, + "learning_rate": 0.00013108269180052244, + "loss": 0.8686, + "step": 8986 + }, + { + "epoch": 1.600071225071225, + "grad_norm": 0.6237605214118958, + "learning_rate": 0.00013106938736724672, + "loss": 0.9166, + "step": 8987 + }, + { + "epoch": 1.6002492877492878, + "grad_norm": 0.6818585395812988, + "learning_rate": 0.0001310560823252309, + "loss": 0.9993, + "step": 8988 + }, + { + "epoch": 1.6004273504273505, + "grad_norm": 0.6372287273406982, + "learning_rate": 0.00013104277667473564, + "loss": 0.8589, + "step": 8989 + }, + { + "epoch": 1.6006054131054133, + "grad_norm": 0.6057302355766296, + "learning_rate": 0.0001310294704160217, + "loss": 0.9325, + "step": 8990 + }, + { + "epoch": 1.6007834757834758, + "grad_norm": 0.6999384164810181, + "learning_rate": 0.0001310161635493497, + "loss": 0.8691, + "step": 8991 + }, + { + "epoch": 1.6009615384615383, + "grad_norm": 0.6182113289833069, + "learning_rate": 0.00013100285607498045, + "loss": 1.0271, + "step": 8992 + }, + { + "epoch": 1.601139601139601, + "grad_norm": 0.6681149005889893, + "learning_rate": 0.0001309895479931746, + "loss": 0.989, + "step": 8993 + }, + { + "epoch": 1.6013176638176638, + "grad_norm": 0.6187826991081238, + "learning_rate": 0.00013097623930419293, + "loss": 0.8051, + "step": 8994 + }, + { + "epoch": 1.6014957264957266, + "grad_norm": 0.698793888092041, + "learning_rate": 0.00013096293000829621, + "loss": 1.0762, + "step": 8995 + }, + { + "epoch": 1.6016737891737893, + "grad_norm": 0.693149745464325, + "learning_rate": 0.0001309496201057452, + "loss": 1.0894, + "step": 8996 + }, + { + "epoch": 1.6018518518518519, + "grad_norm": 0.6664052605628967, + "learning_rate": 0.00013093630959680068, + "loss": 0.9835, + "step": 8997 + }, + { + "epoch": 1.6020299145299144, + "grad_norm": 0.6919469833374023, + "learning_rate": 0.0001309229984817234, + "loss": 0.9062, + "step": 8998 + }, + { + "epoch": 1.6022079772079771, + "grad_norm": 0.704781174659729, + "learning_rate": 0.00013090968676077427, + "loss": 0.8582, + "step": 8999 + }, + { + "epoch": 1.60238603988604, + "grad_norm": 0.8055264949798584, + "learning_rate": 0.000130896374434214, + "loss": 0.9813, + "step": 9000 + }, + { + "epoch": 1.6025641025641026, + "grad_norm": 0.6301952004432678, + "learning_rate": 0.00013088306150230348, + "loss": 0.7056, + "step": 9001 + }, + { + "epoch": 1.6027421652421654, + "grad_norm": 0.698544442653656, + "learning_rate": 0.00013086974796530347, + "loss": 0.9806, + "step": 9002 + }, + { + "epoch": 1.602920227920228, + "grad_norm": 0.669548511505127, + "learning_rate": 0.00013085643382347491, + "loss": 1.0317, + "step": 9003 + }, + { + "epoch": 1.6030982905982905, + "grad_norm": 0.6404716372489929, + "learning_rate": 0.00013084311907707864, + "loss": 0.8885, + "step": 9004 + }, + { + "epoch": 1.6032763532763532, + "grad_norm": 0.6968616843223572, + "learning_rate": 0.0001308298037263755, + "loss": 1.0665, + "step": 9005 + }, + { + "epoch": 1.603454415954416, + "grad_norm": 0.849311113357544, + "learning_rate": 0.00013081648777162644, + "loss": 1.1404, + "step": 9006 + }, + { + "epoch": 1.6036324786324787, + "grad_norm": 0.6603094935417175, + "learning_rate": 0.00013080317121309223, + "loss": 0.8341, + "step": 9007 + }, + { + "epoch": 1.6038105413105415, + "grad_norm": 0.6777810454368591, + "learning_rate": 0.00013078985405103394, + "loss": 1.044, + "step": 9008 + }, + { + "epoch": 1.603988603988604, + "grad_norm": 0.6783546209335327, + "learning_rate": 0.0001307765362857124, + "loss": 1.042, + "step": 9009 + }, + { + "epoch": 1.6041666666666665, + "grad_norm": 0.7251788377761841, + "learning_rate": 0.00013076321791738858, + "loss": 0.9004, + "step": 9010 + }, + { + "epoch": 1.6043447293447293, + "grad_norm": 0.7885342240333557, + "learning_rate": 0.00013074989894632338, + "loss": 1.1966, + "step": 9011 + }, + { + "epoch": 1.604522792022792, + "grad_norm": 0.7171013355255127, + "learning_rate": 0.0001307365793727778, + "loss": 1.2242, + "step": 9012 + }, + { + "epoch": 1.6047008547008548, + "grad_norm": 0.6027249693870544, + "learning_rate": 0.00013072325919701283, + "loss": 0.917, + "step": 9013 + }, + { + "epoch": 1.6048789173789175, + "grad_norm": 0.5957151055335999, + "learning_rate": 0.00013070993841928936, + "loss": 0.9154, + "step": 9014 + }, + { + "epoch": 1.60505698005698, + "grad_norm": 0.6190659403800964, + "learning_rate": 0.00013069661703986847, + "loss": 0.7071, + "step": 9015 + }, + { + "epoch": 1.6052350427350426, + "grad_norm": 0.6454868316650391, + "learning_rate": 0.00013068329505901117, + "loss": 0.8381, + "step": 9016 + }, + { + "epoch": 1.6054131054131053, + "grad_norm": 0.6255491375923157, + "learning_rate": 0.00013066997247697837, + "loss": 0.7515, + "step": 9017 + }, + { + "epoch": 1.605591168091168, + "grad_norm": 0.6214072108268738, + "learning_rate": 0.0001306566492940312, + "loss": 1.0101, + "step": 9018 + }, + { + "epoch": 1.6057692307692308, + "grad_norm": 0.7244150638580322, + "learning_rate": 0.0001306433255104307, + "loss": 1.2558, + "step": 9019 + }, + { + "epoch": 1.6059472934472936, + "grad_norm": 0.6162270903587341, + "learning_rate": 0.00013063000112643785, + "loss": 1.1009, + "step": 9020 + }, + { + "epoch": 1.6061253561253561, + "grad_norm": 0.7309414744377136, + "learning_rate": 0.0001306166761423138, + "loss": 1.1973, + "step": 9021 + }, + { + "epoch": 1.6063034188034186, + "grad_norm": 0.7150956392288208, + "learning_rate": 0.00013060335055831957, + "loss": 0.9136, + "step": 9022 + }, + { + "epoch": 1.6064814814814814, + "grad_norm": 0.8187742829322815, + "learning_rate": 0.00013059002437471623, + "loss": 1.0524, + "step": 9023 + }, + { + "epoch": 1.6066595441595442, + "grad_norm": 0.7928692698478699, + "learning_rate": 0.00013057669759176493, + "loss": 1.0249, + "step": 9024 + }, + { + "epoch": 1.606837606837607, + "grad_norm": 0.6929279565811157, + "learning_rate": 0.00013056337020972677, + "loss": 1.1804, + "step": 9025 + }, + { + "epoch": 1.6070156695156697, + "grad_norm": 0.6771654486656189, + "learning_rate": 0.00013055004222886285, + "loss": 1.0284, + "step": 9026 + }, + { + "epoch": 1.6071937321937322, + "grad_norm": 0.6689024567604065, + "learning_rate": 0.0001305367136494343, + "loss": 1.0431, + "step": 9027 + }, + { + "epoch": 1.6073717948717947, + "grad_norm": 0.71135413646698, + "learning_rate": 0.0001305233844717023, + "loss": 0.9692, + "step": 9028 + }, + { + "epoch": 1.6075498575498575, + "grad_norm": 0.5459749698638916, + "learning_rate": 0.00013051005469592796, + "loss": 0.5643, + "step": 9029 + }, + { + "epoch": 1.6077279202279202, + "grad_norm": 0.7225865125656128, + "learning_rate": 0.00013049672432237253, + "loss": 1.0954, + "step": 9030 + }, + { + "epoch": 1.607905982905983, + "grad_norm": 0.6878093481063843, + "learning_rate": 0.0001304833933512971, + "loss": 0.894, + "step": 9031 + }, + { + "epoch": 1.6080840455840457, + "grad_norm": 0.6967248320579529, + "learning_rate": 0.00013047006178296288, + "loss": 1.0356, + "step": 9032 + }, + { + "epoch": 1.6082621082621082, + "grad_norm": 0.6404993534088135, + "learning_rate": 0.00013045672961763114, + "loss": 0.8528, + "step": 9033 + }, + { + "epoch": 1.6084401709401708, + "grad_norm": 0.5919156074523926, + "learning_rate": 0.000130443396855563, + "loss": 0.7196, + "step": 9034 + }, + { + "epoch": 1.6086182336182335, + "grad_norm": 0.6792302131652832, + "learning_rate": 0.00013043006349701977, + "loss": 0.9519, + "step": 9035 + }, + { + "epoch": 1.6087962962962963, + "grad_norm": 0.6263542175292969, + "learning_rate": 0.00013041672954226268, + "loss": 1.0483, + "step": 9036 + }, + { + "epoch": 1.608974358974359, + "grad_norm": 0.5865579843521118, + "learning_rate": 0.00013040339499155294, + "loss": 0.8794, + "step": 9037 + }, + { + "epoch": 1.6091524216524218, + "grad_norm": 0.8383142948150635, + "learning_rate": 0.00013039005984515181, + "loss": 0.8929, + "step": 9038 + }, + { + "epoch": 1.6093304843304843, + "grad_norm": 0.6438691020011902, + "learning_rate": 0.00013037672410332063, + "loss": 0.9957, + "step": 9039 + }, + { + "epoch": 1.609508547008547, + "grad_norm": 0.74748694896698, + "learning_rate": 0.0001303633877663206, + "loss": 0.9809, + "step": 9040 + }, + { + "epoch": 1.6096866096866096, + "grad_norm": 0.6697205901145935, + "learning_rate": 0.00013035005083441312, + "loss": 0.9556, + "step": 9041 + }, + { + "epoch": 1.6098646723646723, + "grad_norm": 0.6577828526496887, + "learning_rate": 0.00013033671330785941, + "loss": 0.8956, + "step": 9042 + }, + { + "epoch": 1.610042735042735, + "grad_norm": 0.6423429846763611, + "learning_rate": 0.0001303233751869208, + "loss": 0.8467, + "step": 9043 + }, + { + "epoch": 1.6102207977207978, + "grad_norm": 0.6552175879478455, + "learning_rate": 0.00013031003647185867, + "loss": 0.8656, + "step": 9044 + }, + { + "epoch": 1.6103988603988604, + "grad_norm": 0.6755174398422241, + "learning_rate": 0.00013029669716293433, + "loss": 0.7836, + "step": 9045 + }, + { + "epoch": 1.6105769230769231, + "grad_norm": 0.6832906007766724, + "learning_rate": 0.00013028335726040914, + "loss": 1.1531, + "step": 9046 + }, + { + "epoch": 1.6107549857549857, + "grad_norm": 0.6498637795448303, + "learning_rate": 0.00013027001676454446, + "loss": 0.8637, + "step": 9047 + }, + { + "epoch": 1.6109330484330484, + "grad_norm": 0.6792944073677063, + "learning_rate": 0.0001302566756756017, + "loss": 1.0865, + "step": 9048 + }, + { + "epoch": 1.6111111111111112, + "grad_norm": 0.6801337003707886, + "learning_rate": 0.00013024333399384226, + "loss": 1.0738, + "step": 9049 + }, + { + "epoch": 1.611289173789174, + "grad_norm": 0.675216794013977, + "learning_rate": 0.0001302299917195275, + "loss": 1.1074, + "step": 9050 + }, + { + "epoch": 1.6114672364672364, + "grad_norm": 0.6418983340263367, + "learning_rate": 0.00013021664885291885, + "loss": 1.0025, + "step": 9051 + }, + { + "epoch": 1.6116452991452992, + "grad_norm": 0.7778789401054382, + "learning_rate": 0.0001302033053942777, + "loss": 1.0847, + "step": 9052 + }, + { + "epoch": 1.6118233618233617, + "grad_norm": 0.7672827243804932, + "learning_rate": 0.00013018996134386555, + "loss": 1.0565, + "step": 9053 + }, + { + "epoch": 1.6120014245014245, + "grad_norm": 0.6770617961883545, + "learning_rate": 0.00013017661670194382, + "loss": 0.9069, + "step": 9054 + }, + { + "epoch": 1.6121794871794872, + "grad_norm": 0.7161242961883545, + "learning_rate": 0.00013016327146877393, + "loss": 1.1301, + "step": 9055 + }, + { + "epoch": 1.61235754985755, + "grad_norm": 0.6923251152038574, + "learning_rate": 0.00013014992564461746, + "loss": 0.9546, + "step": 9056 + }, + { + "epoch": 1.6125356125356125, + "grad_norm": 0.622953474521637, + "learning_rate": 0.0001301365792297358, + "loss": 0.8152, + "step": 9057 + }, + { + "epoch": 1.6127136752136753, + "grad_norm": 0.7477008104324341, + "learning_rate": 0.00013012323222439046, + "loss": 0.8428, + "step": 9058 + }, + { + "epoch": 1.6128917378917378, + "grad_norm": 0.6612883806228638, + "learning_rate": 0.000130109884628843, + "loss": 1.0678, + "step": 9059 + }, + { + "epoch": 1.6130698005698005, + "grad_norm": 0.6406781077384949, + "learning_rate": 0.00013009653644335486, + "loss": 0.6792, + "step": 9060 + }, + { + "epoch": 1.6132478632478633, + "grad_norm": 0.6279141902923584, + "learning_rate": 0.00013008318766818763, + "loss": 0.9826, + "step": 9061 + }, + { + "epoch": 1.613425925925926, + "grad_norm": 0.6616412401199341, + "learning_rate": 0.00013006983830360285, + "loss": 1.0691, + "step": 9062 + }, + { + "epoch": 1.6136039886039886, + "grad_norm": 0.6520406603813171, + "learning_rate": 0.000130056488349862, + "loss": 0.9487, + "step": 9063 + }, + { + "epoch": 1.6137820512820513, + "grad_norm": 0.6378647089004517, + "learning_rate": 0.00013004313780722672, + "loss": 0.8557, + "step": 9064 + }, + { + "epoch": 1.6139601139601139, + "grad_norm": 0.6547569036483765, + "learning_rate": 0.00013002978667595857, + "loss": 0.879, + "step": 9065 + }, + { + "epoch": 1.6141381766381766, + "grad_norm": 0.7347842454910278, + "learning_rate": 0.00013001643495631914, + "loss": 1.0757, + "step": 9066 + }, + { + "epoch": 1.6143162393162394, + "grad_norm": 0.5988406538963318, + "learning_rate": 0.00013000308264857002, + "loss": 0.6754, + "step": 9067 + }, + { + "epoch": 1.614494301994302, + "grad_norm": 0.6949366331100464, + "learning_rate": 0.00012998972975297282, + "loss": 1.1236, + "step": 9068 + }, + { + "epoch": 1.6146723646723646, + "grad_norm": 0.7095484137535095, + "learning_rate": 0.00012997637626978913, + "loss": 1.0124, + "step": 9069 + }, + { + "epoch": 1.6148504273504274, + "grad_norm": 0.6634095311164856, + "learning_rate": 0.00012996302219928064, + "loss": 1.2018, + "step": 9070 + }, + { + "epoch": 1.61502849002849, + "grad_norm": 0.6894524693489075, + "learning_rate": 0.000129949667541709, + "loss": 0.9959, + "step": 9071 + }, + { + "epoch": 1.6152065527065527, + "grad_norm": 0.672334611415863, + "learning_rate": 0.00012993631229733582, + "loss": 1.0369, + "step": 9072 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.725759744644165, + "learning_rate": 0.00012992295646642278, + "loss": 1.0079, + "step": 9073 + }, + { + "epoch": 1.6155626780626782, + "grad_norm": 0.7941585779190063, + "learning_rate": 0.00012990960004923154, + "loss": 0.9468, + "step": 9074 + }, + { + "epoch": 1.6157407407407407, + "grad_norm": 0.6556950807571411, + "learning_rate": 0.00012989624304602385, + "loss": 0.9915, + "step": 9075 + }, + { + "epoch": 1.6159188034188035, + "grad_norm": 0.7515892386436462, + "learning_rate": 0.0001298828854570614, + "loss": 1.0924, + "step": 9076 + }, + { + "epoch": 1.616096866096866, + "grad_norm": 0.6944101452827454, + "learning_rate": 0.00012986952728260586, + "loss": 0.9632, + "step": 9077 + }, + { + "epoch": 1.6162749287749287, + "grad_norm": 0.6286170482635498, + "learning_rate": 0.000129856168522919, + "loss": 1.0311, + "step": 9078 + }, + { + "epoch": 1.6164529914529915, + "grad_norm": 0.8362757563591003, + "learning_rate": 0.0001298428091782625, + "loss": 1.1232, + "step": 9079 + }, + { + "epoch": 1.6166310541310542, + "grad_norm": 0.6199851632118225, + "learning_rate": 0.0001298294492488982, + "loss": 0.9454, + "step": 9080 + }, + { + "epoch": 1.6168091168091168, + "grad_norm": 0.7541791796684265, + "learning_rate": 0.0001298160887350878, + "loss": 0.9759, + "step": 9081 + }, + { + "epoch": 1.6169871794871795, + "grad_norm": 0.6940878033638, + "learning_rate": 0.00012980272763709304, + "loss": 0.9258, + "step": 9082 + }, + { + "epoch": 1.617165242165242, + "grad_norm": 0.6934045553207397, + "learning_rate": 0.00012978936595517575, + "loss": 1.0142, + "step": 9083 + }, + { + "epoch": 1.6173433048433048, + "grad_norm": 0.8147503733634949, + "learning_rate": 0.00012977600368959774, + "loss": 0.964, + "step": 9084 + }, + { + "epoch": 1.6175213675213675, + "grad_norm": 0.6583107709884644, + "learning_rate": 0.00012976264084062079, + "loss": 1.0315, + "step": 9085 + }, + { + "epoch": 1.6176994301994303, + "grad_norm": 0.7192013263702393, + "learning_rate": 0.0001297492774085067, + "loss": 0.9528, + "step": 9086 + }, + { + "epoch": 1.6178774928774928, + "grad_norm": 0.665888786315918, + "learning_rate": 0.00012973591339351733, + "loss": 1.0188, + "step": 9087 + }, + { + "epoch": 1.6180555555555556, + "grad_norm": 0.7170987725257874, + "learning_rate": 0.0001297225487959145, + "loss": 0.8969, + "step": 9088 + }, + { + "epoch": 1.618233618233618, + "grad_norm": 0.6768732070922852, + "learning_rate": 0.00012970918361596007, + "loss": 1.1951, + "step": 9089 + }, + { + "epoch": 1.6184116809116809, + "grad_norm": 0.6640290021896362, + "learning_rate": 0.00012969581785391592, + "loss": 0.9649, + "step": 9090 + }, + { + "epoch": 1.6185897435897436, + "grad_norm": 0.6200813055038452, + "learning_rate": 0.00012968245151004392, + "loss": 0.9446, + "step": 9091 + }, + { + "epoch": 1.6187678062678064, + "grad_norm": 0.6815837621688843, + "learning_rate": 0.0001296690845846059, + "loss": 1.0506, + "step": 9092 + }, + { + "epoch": 1.618945868945869, + "grad_norm": 0.7252637147903442, + "learning_rate": 0.0001296557170778638, + "loss": 1.1977, + "step": 9093 + }, + { + "epoch": 1.6191239316239316, + "grad_norm": 0.5609107613563538, + "learning_rate": 0.00012964234899007955, + "loss": 0.8009, + "step": 9094 + }, + { + "epoch": 1.6193019943019942, + "grad_norm": 0.6539437770843506, + "learning_rate": 0.00012962898032151506, + "loss": 0.8482, + "step": 9095 + }, + { + "epoch": 1.619480056980057, + "grad_norm": 0.6993300914764404, + "learning_rate": 0.0001296156110724322, + "loss": 1.0725, + "step": 9096 + }, + { + "epoch": 1.6196581196581197, + "grad_norm": 0.6768273711204529, + "learning_rate": 0.000129602241243093, + "loss": 0.9247, + "step": 9097 + }, + { + "epoch": 1.6198361823361824, + "grad_norm": 0.6896265745162964, + "learning_rate": 0.00012958887083375939, + "loss": 0.9526, + "step": 9098 + }, + { + "epoch": 1.6200142450142452, + "grad_norm": 0.7475146651268005, + "learning_rate": 0.00012957549984469327, + "loss": 0.8302, + "step": 9099 + }, + { + "epoch": 1.6201923076923077, + "grad_norm": 0.6622769236564636, + "learning_rate": 0.00012956212827615674, + "loss": 0.9505, + "step": 9100 + }, + { + "epoch": 1.6203703703703702, + "grad_norm": 0.6938058137893677, + "learning_rate": 0.00012954875612841167, + "loss": 0.9757, + "step": 9101 + }, + { + "epoch": 1.620548433048433, + "grad_norm": 0.7453510761260986, + "learning_rate": 0.0001295353834017201, + "loss": 1.0919, + "step": 9102 + }, + { + "epoch": 1.6207264957264957, + "grad_norm": 0.7868932485580444, + "learning_rate": 0.0001295220100963441, + "loss": 0.9265, + "step": 9103 + }, + { + "epoch": 1.6209045584045585, + "grad_norm": 0.6779825091362, + "learning_rate": 0.00012950863621254558, + "loss": 0.98, + "step": 9104 + }, + { + "epoch": 1.6210826210826212, + "grad_norm": 0.6825897097587585, + "learning_rate": 0.00012949526175058662, + "loss": 0.9218, + "step": 9105 + }, + { + "epoch": 1.6212606837606838, + "grad_norm": 0.6686047911643982, + "learning_rate": 0.00012948188671072934, + "loss": 0.9546, + "step": 9106 + }, + { + "epoch": 1.6214387464387463, + "grad_norm": 0.7456090450286865, + "learning_rate": 0.0001294685110932357, + "loss": 1.0819, + "step": 9107 + }, + { + "epoch": 1.621616809116809, + "grad_norm": 0.7111441493034363, + "learning_rate": 0.0001294551348983678, + "loss": 0.9916, + "step": 9108 + }, + { + "epoch": 1.6217948717948718, + "grad_norm": 0.6534699201583862, + "learning_rate": 0.00012944175812638773, + "loss": 1.0374, + "step": 9109 + }, + { + "epoch": 1.6219729344729346, + "grad_norm": 0.6046397089958191, + "learning_rate": 0.00012942838077755758, + "loss": 0.7922, + "step": 9110 + }, + { + "epoch": 1.6221509971509973, + "grad_norm": 0.7736679911613464, + "learning_rate": 0.00012941500285213942, + "loss": 1.0056, + "step": 9111 + }, + { + "epoch": 1.6223290598290598, + "grad_norm": 0.6850929260253906, + "learning_rate": 0.00012940162435039538, + "loss": 0.9538, + "step": 9112 + }, + { + "epoch": 1.6225071225071224, + "grad_norm": 0.6305751800537109, + "learning_rate": 0.00012938824527258756, + "loss": 0.9341, + "step": 9113 + }, + { + "epoch": 1.6226851851851851, + "grad_norm": 0.6740923523902893, + "learning_rate": 0.0001293748656189782, + "loss": 1.0037, + "step": 9114 + }, + { + "epoch": 1.6228632478632479, + "grad_norm": 0.6579762101173401, + "learning_rate": 0.00012936148538982928, + "loss": 1.0022, + "step": 9115 + }, + { + "epoch": 1.6230413105413106, + "grad_norm": 0.6500434279441833, + "learning_rate": 0.0001293481045854031, + "loss": 0.8589, + "step": 9116 + }, + { + "epoch": 1.6232193732193734, + "grad_norm": 0.7825912237167358, + "learning_rate": 0.00012933472320596177, + "loss": 1.0345, + "step": 9117 + }, + { + "epoch": 1.623397435897436, + "grad_norm": 0.8341414332389832, + "learning_rate": 0.0001293213412517675, + "loss": 1.0314, + "step": 9118 + }, + { + "epoch": 1.6235754985754984, + "grad_norm": 0.63664311170578, + "learning_rate": 0.00012930795872308242, + "loss": 0.819, + "step": 9119 + }, + { + "epoch": 1.6237535612535612, + "grad_norm": 0.6800840497016907, + "learning_rate": 0.00012929457562016878, + "loss": 0.95, + "step": 9120 + }, + { + "epoch": 1.623931623931624, + "grad_norm": 0.754165530204773, + "learning_rate": 0.0001292811919432888, + "loss": 1.1193, + "step": 9121 + }, + { + "epoch": 1.6241096866096867, + "grad_norm": 0.678871750831604, + "learning_rate": 0.00012926780769270465, + "loss": 0.9015, + "step": 9122 + }, + { + "epoch": 1.6242877492877494, + "grad_norm": 0.6642945408821106, + "learning_rate": 0.00012925442286867866, + "loss": 0.9095, + "step": 9123 + }, + { + "epoch": 1.624465811965812, + "grad_norm": 0.6089697480201721, + "learning_rate": 0.000129241037471473, + "loss": 0.8994, + "step": 9124 + }, + { + "epoch": 1.6246438746438745, + "grad_norm": 0.7320881485939026, + "learning_rate": 0.00012922765150134995, + "loss": 1.0518, + "step": 9125 + }, + { + "epoch": 1.6248219373219372, + "grad_norm": 0.7308032512664795, + "learning_rate": 0.0001292142649585718, + "loss": 1.0557, + "step": 9126 + }, + { + "epoch": 1.625, + "grad_norm": 0.6896602511405945, + "learning_rate": 0.0001292008778434008, + "loss": 1.145, + "step": 9127 + }, + { + "epoch": 1.6251780626780628, + "grad_norm": 0.6112532615661621, + "learning_rate": 0.00012918749015609926, + "loss": 0.9611, + "step": 9128 + }, + { + "epoch": 1.6253561253561255, + "grad_norm": 0.6856057643890381, + "learning_rate": 0.00012917410189692947, + "loss": 1.0124, + "step": 9129 + }, + { + "epoch": 1.625534188034188, + "grad_norm": 0.699252188205719, + "learning_rate": 0.00012916071306615378, + "loss": 0.8854, + "step": 9130 + }, + { + "epoch": 1.6257122507122506, + "grad_norm": 0.6306683421134949, + "learning_rate": 0.0001291473236640345, + "loss": 1.0722, + "step": 9131 + }, + { + "epoch": 1.6258903133903133, + "grad_norm": 0.6358118653297424, + "learning_rate": 0.00012913393369083393, + "loss": 0.889, + "step": 9132 + }, + { + "epoch": 1.626068376068376, + "grad_norm": 0.6953601837158203, + "learning_rate": 0.00012912054314681445, + "loss": 1.0168, + "step": 9133 + }, + { + "epoch": 1.6262464387464388, + "grad_norm": 0.6742331385612488, + "learning_rate": 0.00012910715203223844, + "loss": 0.8152, + "step": 9134 + }, + { + "epoch": 1.6264245014245016, + "grad_norm": 0.5872861742973328, + "learning_rate": 0.00012909376034736823, + "loss": 0.8702, + "step": 9135 + }, + { + "epoch": 1.626602564102564, + "grad_norm": 0.7580631971359253, + "learning_rate": 0.00012908036809246623, + "loss": 0.994, + "step": 9136 + }, + { + "epoch": 1.6267806267806266, + "grad_norm": 0.7544930577278137, + "learning_rate": 0.00012906697526779488, + "loss": 0.7475, + "step": 9137 + }, + { + "epoch": 1.6269586894586894, + "grad_norm": 0.6850766539573669, + "learning_rate": 0.00012905358187361647, + "loss": 1.0943, + "step": 9138 + }, + { + "epoch": 1.6271367521367521, + "grad_norm": 0.6821565628051758, + "learning_rate": 0.0001290401879101935, + "loss": 1.2928, + "step": 9139 + }, + { + "epoch": 1.6273148148148149, + "grad_norm": 0.6961034536361694, + "learning_rate": 0.00012902679337778835, + "loss": 0.8694, + "step": 9140 + }, + { + "epoch": 1.6274928774928776, + "grad_norm": 0.7159550786018372, + "learning_rate": 0.00012901339827666353, + "loss": 0.8827, + "step": 9141 + }, + { + "epoch": 1.6276709401709402, + "grad_norm": 0.7491081953048706, + "learning_rate": 0.0001290000026070814, + "loss": 0.8159, + "step": 9142 + }, + { + "epoch": 1.6278490028490027, + "grad_norm": 0.7107849717140198, + "learning_rate": 0.00012898660636930447, + "loss": 1.0625, + "step": 9143 + }, + { + "epoch": 1.6280270655270654, + "grad_norm": 0.7227210998535156, + "learning_rate": 0.0001289732095635952, + "loss": 0.9744, + "step": 9144 + }, + { + "epoch": 1.6282051282051282, + "grad_norm": 0.7141995429992676, + "learning_rate": 0.00012895981219021607, + "loss": 0.9836, + "step": 9145 + }, + { + "epoch": 1.628383190883191, + "grad_norm": 0.6445552706718445, + "learning_rate": 0.00012894641424942958, + "loss": 1.0183, + "step": 9146 + }, + { + "epoch": 1.6285612535612537, + "grad_norm": 0.698783278465271, + "learning_rate": 0.00012893301574149824, + "loss": 0.8392, + "step": 9147 + }, + { + "epoch": 1.6287393162393162, + "grad_norm": 0.6529116034507751, + "learning_rate": 0.00012891961666668458, + "loss": 0.9317, + "step": 9148 + }, + { + "epoch": 1.6289173789173788, + "grad_norm": 0.7780548930168152, + "learning_rate": 0.0001289062170252511, + "loss": 1.2406, + "step": 9149 + }, + { + "epoch": 1.6290954415954415, + "grad_norm": 0.6500990986824036, + "learning_rate": 0.0001288928168174603, + "loss": 1.0381, + "step": 9150 + }, + { + "epoch": 1.6292735042735043, + "grad_norm": 0.7098208665847778, + "learning_rate": 0.00012887941604357482, + "loss": 1.2126, + "step": 9151 + }, + { + "epoch": 1.629451566951567, + "grad_norm": 0.730648398399353, + "learning_rate": 0.0001288660147038572, + "loss": 0.8351, + "step": 9152 + }, + { + "epoch": 1.6296296296296298, + "grad_norm": 0.5520278215408325, + "learning_rate": 0.0001288526127985699, + "loss": 0.5877, + "step": 9153 + }, + { + "epoch": 1.6298076923076923, + "grad_norm": 0.7611770033836365, + "learning_rate": 0.00012883921032797563, + "loss": 1.2227, + "step": 9154 + }, + { + "epoch": 1.6299857549857548, + "grad_norm": 0.636820375919342, + "learning_rate": 0.00012882580729233696, + "loss": 0.8305, + "step": 9155 + }, + { + "epoch": 1.6301638176638176, + "grad_norm": 0.694492518901825, + "learning_rate": 0.00012881240369191644, + "loss": 1.0452, + "step": 9156 + }, + { + "epoch": 1.6303418803418803, + "grad_norm": 0.67826908826828, + "learning_rate": 0.00012879899952697677, + "loss": 0.8345, + "step": 9157 + }, + { + "epoch": 1.630519943019943, + "grad_norm": 0.5891323685646057, + "learning_rate": 0.00012878559479778052, + "loss": 0.8367, + "step": 9158 + }, + { + "epoch": 1.6306980056980058, + "grad_norm": 0.6766192317008972, + "learning_rate": 0.0001287721895045903, + "loss": 0.8319, + "step": 9159 + }, + { + "epoch": 1.6308760683760684, + "grad_norm": 0.5306392908096313, + "learning_rate": 0.0001287587836476688, + "loss": 0.7945, + "step": 9160 + }, + { + "epoch": 1.631054131054131, + "grad_norm": 0.6677970290184021, + "learning_rate": 0.0001287453772272787, + "loss": 1.1228, + "step": 9161 + }, + { + "epoch": 1.6312321937321936, + "grad_norm": 0.810052752494812, + "learning_rate": 0.00012873197024368266, + "loss": 0.8395, + "step": 9162 + }, + { + "epoch": 1.6314102564102564, + "grad_norm": 0.7619220018386841, + "learning_rate": 0.00012871856269714333, + "loss": 1.3713, + "step": 9163 + }, + { + "epoch": 1.6315883190883191, + "grad_norm": 0.6564521193504333, + "learning_rate": 0.00012870515458792342, + "loss": 1.0513, + "step": 9164 + }, + { + "epoch": 1.631766381766382, + "grad_norm": 0.6874445676803589, + "learning_rate": 0.00012869174591628564, + "loss": 1.0255, + "step": 9165 + }, + { + "epoch": 1.6319444444444444, + "grad_norm": 0.6958737373352051, + "learning_rate": 0.0001286783366824927, + "loss": 0.9361, + "step": 9166 + }, + { + "epoch": 1.6321225071225072, + "grad_norm": 0.6909199357032776, + "learning_rate": 0.0001286649268868073, + "loss": 0.9855, + "step": 9167 + }, + { + "epoch": 1.6323005698005697, + "grad_norm": 0.7671375274658203, + "learning_rate": 0.00012865151652949225, + "loss": 1.084, + "step": 9168 + }, + { + "epoch": 1.6324786324786325, + "grad_norm": 0.750200092792511, + "learning_rate": 0.00012863810561081023, + "loss": 0.9341, + "step": 9169 + }, + { + "epoch": 1.6326566951566952, + "grad_norm": 0.6595860123634338, + "learning_rate": 0.00012862469413102402, + "loss": 0.9386, + "step": 9170 + }, + { + "epoch": 1.632834757834758, + "grad_norm": 0.622373640537262, + "learning_rate": 0.0001286112820903964, + "loss": 0.7697, + "step": 9171 + }, + { + "epoch": 1.6330128205128205, + "grad_norm": 0.9628498554229736, + "learning_rate": 0.00012859786948919014, + "loss": 1.2629, + "step": 9172 + }, + { + "epoch": 1.6331908831908832, + "grad_norm": 0.7610561847686768, + "learning_rate": 0.000128584456327668, + "loss": 0.9748, + "step": 9173 + }, + { + "epoch": 1.6333689458689458, + "grad_norm": 0.6585374474525452, + "learning_rate": 0.00012857104260609285, + "loss": 0.9049, + "step": 9174 + }, + { + "epoch": 1.6335470085470085, + "grad_norm": 0.6996221542358398, + "learning_rate": 0.00012855762832472746, + "loss": 0.8893, + "step": 9175 + }, + { + "epoch": 1.6337250712250713, + "grad_norm": 0.6226270198822021, + "learning_rate": 0.00012854421348383466, + "loss": 0.8913, + "step": 9176 + }, + { + "epoch": 1.633903133903134, + "grad_norm": 0.6570866107940674, + "learning_rate": 0.00012853079808367731, + "loss": 0.8632, + "step": 9177 + }, + { + "epoch": 1.6340811965811965, + "grad_norm": 0.6899664402008057, + "learning_rate": 0.00012851738212451826, + "loss": 0.8177, + "step": 9178 + }, + { + "epoch": 1.6342592592592593, + "grad_norm": 0.75257807970047, + "learning_rate": 0.0001285039656066203, + "loss": 0.9096, + "step": 9179 + }, + { + "epoch": 1.6344373219373218, + "grad_norm": 0.6614963412284851, + "learning_rate": 0.00012849054853024638, + "loss": 0.9255, + "step": 9180 + }, + { + "epoch": 1.6346153846153846, + "grad_norm": 0.7245957851409912, + "learning_rate": 0.00012847713089565933, + "loss": 1.0122, + "step": 9181 + }, + { + "epoch": 1.6347934472934473, + "grad_norm": 0.7332839369773865, + "learning_rate": 0.00012846371270312204, + "loss": 0.8484, + "step": 9182 + }, + { + "epoch": 1.63497150997151, + "grad_norm": 0.628089189529419, + "learning_rate": 0.00012845029395289748, + "loss": 1.0171, + "step": 9183 + }, + { + "epoch": 1.6351495726495726, + "grad_norm": 0.7493528723716736, + "learning_rate": 0.00012843687464524848, + "loss": 1.1635, + "step": 9184 + }, + { + "epoch": 1.6353276353276354, + "grad_norm": 0.6328163146972656, + "learning_rate": 0.00012842345478043799, + "loss": 1.1254, + "step": 9185 + }, + { + "epoch": 1.635505698005698, + "grad_norm": 0.6720291376113892, + "learning_rate": 0.00012841003435872894, + "loss": 0.9729, + "step": 9186 + }, + { + "epoch": 1.6356837606837606, + "grad_norm": 0.6657332181930542, + "learning_rate": 0.00012839661338038427, + "loss": 1.1047, + "step": 9187 + }, + { + "epoch": 1.6358618233618234, + "grad_norm": 0.7416180968284607, + "learning_rate": 0.000128383191845667, + "loss": 0.9505, + "step": 9188 + }, + { + "epoch": 1.6360398860398861, + "grad_norm": 0.8737816214561462, + "learning_rate": 0.00012836976975484, + "loss": 1.0518, + "step": 9189 + }, + { + "epoch": 1.6362179487179487, + "grad_norm": 0.7351877093315125, + "learning_rate": 0.0001283563471081663, + "loss": 1.1152, + "step": 9190 + }, + { + "epoch": 1.6363960113960114, + "grad_norm": 0.6442788243293762, + "learning_rate": 0.00012834292390590893, + "loss": 0.9432, + "step": 9191 + }, + { + "epoch": 1.636574074074074, + "grad_norm": 0.6848029494285583, + "learning_rate": 0.0001283295001483308, + "loss": 0.8528, + "step": 9192 + }, + { + "epoch": 1.6367521367521367, + "grad_norm": 0.6627060174942017, + "learning_rate": 0.00012831607583569497, + "loss": 1.0222, + "step": 9193 + }, + { + "epoch": 1.6369301994301995, + "grad_norm": 0.7319555878639221, + "learning_rate": 0.00012830265096826446, + "loss": 0.9392, + "step": 9194 + }, + { + "epoch": 1.6371082621082622, + "grad_norm": 0.6986424326896667, + "learning_rate": 0.0001282892255463023, + "loss": 1.2095, + "step": 9195 + }, + { + "epoch": 1.6372863247863247, + "grad_norm": 0.6649929881095886, + "learning_rate": 0.0001282757995700715, + "loss": 0.9426, + "step": 9196 + }, + { + "epoch": 1.6374643874643875, + "grad_norm": 0.6789031624794006, + "learning_rate": 0.0001282623730398352, + "loss": 0.9705, + "step": 9197 + }, + { + "epoch": 1.63764245014245, + "grad_norm": 0.6388779878616333, + "learning_rate": 0.00012824894595585637, + "loss": 1.0698, + "step": 9198 + }, + { + "epoch": 1.6378205128205128, + "grad_norm": 0.636832594871521, + "learning_rate": 0.00012823551831839814, + "loss": 0.9445, + "step": 9199 + }, + { + "epoch": 1.6379985754985755, + "grad_norm": 0.670190691947937, + "learning_rate": 0.0001282220901277236, + "loss": 0.9847, + "step": 9200 + }, + { + "epoch": 1.6381766381766383, + "grad_norm": 0.6020209193229675, + "learning_rate": 0.0001282086613840958, + "loss": 1.0047, + "step": 9201 + }, + { + "epoch": 1.6383547008547008, + "grad_norm": 0.6648211479187012, + "learning_rate": 0.0001281952320877779, + "loss": 0.8717, + "step": 9202 + }, + { + "epoch": 1.6385327635327636, + "grad_norm": 0.7207710146903992, + "learning_rate": 0.000128181802239033, + "loss": 1.1232, + "step": 9203 + }, + { + "epoch": 1.638710826210826, + "grad_norm": 0.800992488861084, + "learning_rate": 0.0001281683718381242, + "loss": 1.0688, + "step": 9204 + }, + { + "epoch": 1.6388888888888888, + "grad_norm": 0.789398193359375, + "learning_rate": 0.0001281549408853147, + "loss": 1.1772, + "step": 9205 + }, + { + "epoch": 1.6390669515669516, + "grad_norm": 0.6514480710029602, + "learning_rate": 0.0001281415093808676, + "loss": 1.1685, + "step": 9206 + }, + { + "epoch": 1.6392450142450143, + "grad_norm": 0.6914686560630798, + "learning_rate": 0.00012812807732504608, + "loss": 1.1307, + "step": 9207 + }, + { + "epoch": 1.6394230769230769, + "grad_norm": 0.6788144111633301, + "learning_rate": 0.00012811464471811334, + "loss": 1.1735, + "step": 9208 + }, + { + "epoch": 1.6396011396011396, + "grad_norm": 0.7049870491027832, + "learning_rate": 0.00012810121156033252, + "loss": 1.0128, + "step": 9209 + }, + { + "epoch": 1.6397792022792022, + "grad_norm": 0.7156766057014465, + "learning_rate": 0.00012808777785196687, + "loss": 0.9503, + "step": 9210 + }, + { + "epoch": 1.639957264957265, + "grad_norm": 0.651716411113739, + "learning_rate": 0.0001280743435932795, + "loss": 1.1227, + "step": 9211 + }, + { + "epoch": 1.6401353276353277, + "grad_norm": 0.7276262044906616, + "learning_rate": 0.0001280609087845337, + "loss": 1.06, + "step": 9212 + }, + { + "epoch": 1.6403133903133904, + "grad_norm": 0.6591095924377441, + "learning_rate": 0.0001280474734259927, + "loss": 1.0861, + "step": 9213 + }, + { + "epoch": 1.640491452991453, + "grad_norm": 0.6675926446914673, + "learning_rate": 0.00012803403751791975, + "loss": 0.9815, + "step": 9214 + }, + { + "epoch": 1.6406695156695157, + "grad_norm": 0.6391474008560181, + "learning_rate": 0.00012802060106057803, + "loss": 0.8027, + "step": 9215 + }, + { + "epoch": 1.6408475783475782, + "grad_norm": 0.6384556293487549, + "learning_rate": 0.00012800716405423086, + "loss": 0.7877, + "step": 9216 + }, + { + "epoch": 1.641025641025641, + "grad_norm": 0.661191463470459, + "learning_rate": 0.00012799372649914146, + "loss": 0.9725, + "step": 9217 + }, + { + "epoch": 1.6412037037037037, + "grad_norm": 0.7418332695960999, + "learning_rate": 0.0001279802883955732, + "loss": 1.1756, + "step": 9218 + }, + { + "epoch": 1.6413817663817665, + "grad_norm": 0.6588954329490662, + "learning_rate": 0.00012796684974378928, + "loss": 1.0428, + "step": 9219 + }, + { + "epoch": 1.6415598290598292, + "grad_norm": 0.7566093802452087, + "learning_rate": 0.000127953410544053, + "loss": 1.1254, + "step": 9220 + }, + { + "epoch": 1.6417378917378918, + "grad_norm": 0.6801039576530457, + "learning_rate": 0.00012793997079662777, + "loss": 1.0854, + "step": 9221 + }, + { + "epoch": 1.6419159544159543, + "grad_norm": 0.7262716889381409, + "learning_rate": 0.0001279265305017768, + "loss": 0.9343, + "step": 9222 + }, + { + "epoch": 1.642094017094017, + "grad_norm": 0.628625750541687, + "learning_rate": 0.0001279130896597635, + "loss": 0.8942, + "step": 9223 + }, + { + "epoch": 1.6422720797720798, + "grad_norm": 0.6183576583862305, + "learning_rate": 0.0001278996482708512, + "loss": 0.9284, + "step": 9224 + }, + { + "epoch": 1.6424501424501425, + "grad_norm": 0.7912000417709351, + "learning_rate": 0.00012788620633530327, + "loss": 1.3043, + "step": 9225 + }, + { + "epoch": 1.6426282051282053, + "grad_norm": 0.6982026100158691, + "learning_rate": 0.00012787276385338298, + "loss": 1.0224, + "step": 9226 + }, + { + "epoch": 1.6428062678062678, + "grad_norm": 0.6734985709190369, + "learning_rate": 0.00012785932082535386, + "loss": 0.8781, + "step": 9227 + }, + { + "epoch": 1.6429843304843303, + "grad_norm": 0.8799532055854797, + "learning_rate": 0.0001278458772514792, + "loss": 1.1482, + "step": 9228 + }, + { + "epoch": 1.643162393162393, + "grad_norm": 0.590295672416687, + "learning_rate": 0.0001278324331320224, + "loss": 0.9502, + "step": 9229 + }, + { + "epoch": 1.6433404558404558, + "grad_norm": 0.6562125086784363, + "learning_rate": 0.0001278189884672469, + "loss": 0.9834, + "step": 9230 + }, + { + "epoch": 1.6435185185185186, + "grad_norm": 0.6848936676979065, + "learning_rate": 0.00012780554325741612, + "loss": 1.0414, + "step": 9231 + }, + { + "epoch": 1.6436965811965814, + "grad_norm": 0.5985032320022583, + "learning_rate": 0.00012779209750279344, + "loss": 0.9469, + "step": 9232 + }, + { + "epoch": 1.6438746438746439, + "grad_norm": 0.7500917911529541, + "learning_rate": 0.00012777865120364238, + "loss": 0.9626, + "step": 9233 + }, + { + "epoch": 1.6440527065527064, + "grad_norm": 0.6565709114074707, + "learning_rate": 0.00012776520436022634, + "loss": 1.0594, + "step": 9234 + }, + { + "epoch": 1.6442307692307692, + "grad_norm": 0.8005441427230835, + "learning_rate": 0.00012775175697280882, + "loss": 1.2379, + "step": 9235 + }, + { + "epoch": 1.644408831908832, + "grad_norm": 0.6734150648117065, + "learning_rate": 0.00012773830904165326, + "loss": 0.9171, + "step": 9236 + }, + { + "epoch": 1.6445868945868947, + "grad_norm": 0.6950868368148804, + "learning_rate": 0.00012772486056702314, + "loss": 1.1782, + "step": 9237 + }, + { + "epoch": 1.6447649572649574, + "grad_norm": 0.8009599447250366, + "learning_rate": 0.000127711411549182, + "loss": 1.0288, + "step": 9238 + }, + { + "epoch": 1.64494301994302, + "grad_norm": 0.6227970719337463, + "learning_rate": 0.0001276979619883933, + "loss": 0.9327, + "step": 9239 + }, + { + "epoch": 1.6451210826210825, + "grad_norm": 0.6828190088272095, + "learning_rate": 0.00012768451188492058, + "loss": 0.9816, + "step": 9240 + }, + { + "epoch": 1.6452991452991452, + "grad_norm": 0.9689767360687256, + "learning_rate": 0.00012767106123902738, + "loss": 0.9049, + "step": 9241 + }, + { + "epoch": 1.645477207977208, + "grad_norm": 0.677061140537262, + "learning_rate": 0.00012765761005097717, + "loss": 0.9472, + "step": 9242 + }, + { + "epoch": 1.6456552706552707, + "grad_norm": 0.7227110862731934, + "learning_rate": 0.00012764415832103356, + "loss": 1.0384, + "step": 9243 + }, + { + "epoch": 1.6458333333333335, + "grad_norm": 0.6540094614028931, + "learning_rate": 0.0001276307060494601, + "loss": 0.8166, + "step": 9244 + }, + { + "epoch": 1.646011396011396, + "grad_norm": 0.6921904683113098, + "learning_rate": 0.00012761725323652033, + "loss": 0.9746, + "step": 9245 + }, + { + "epoch": 1.6461894586894585, + "grad_norm": 0.6742660999298096, + "learning_rate": 0.0001276037998824779, + "loss": 0.8441, + "step": 9246 + }, + { + "epoch": 1.6463675213675213, + "grad_norm": 0.6611103415489197, + "learning_rate": 0.0001275903459875963, + "loss": 1.087, + "step": 9247 + }, + { + "epoch": 1.646545584045584, + "grad_norm": 0.6805498003959656, + "learning_rate": 0.00012757689155213923, + "loss": 0.923, + "step": 9248 + }, + { + "epoch": 1.6467236467236468, + "grad_norm": 0.6598179340362549, + "learning_rate": 0.00012756343657637024, + "loss": 0.9371, + "step": 9249 + }, + { + "epoch": 1.6469017094017095, + "grad_norm": 0.7147273421287537, + "learning_rate": 0.00012754998106055297, + "loss": 1.053, + "step": 9250 + }, + { + "epoch": 1.647079772079772, + "grad_norm": 0.72414630651474, + "learning_rate": 0.00012753652500495103, + "loss": 1.0547, + "step": 9251 + }, + { + "epoch": 1.6472578347578346, + "grad_norm": 0.7784913182258606, + "learning_rate": 0.00012752306840982811, + "loss": 0.9012, + "step": 9252 + }, + { + "epoch": 1.6474358974358974, + "grad_norm": 0.644026517868042, + "learning_rate": 0.0001275096112754478, + "loss": 1.0911, + "step": 9253 + }, + { + "epoch": 1.64761396011396, + "grad_norm": 0.691124677658081, + "learning_rate": 0.00012749615360207382, + "loss": 0.9918, + "step": 9254 + }, + { + "epoch": 1.6477920227920229, + "grad_norm": 0.6632972359657288, + "learning_rate": 0.00012748269538996986, + "loss": 0.9438, + "step": 9255 + }, + { + "epoch": 1.6479700854700856, + "grad_norm": 0.6548733115196228, + "learning_rate": 0.00012746923663939955, + "loss": 1.1082, + "step": 9256 + }, + { + "epoch": 1.6481481481481481, + "grad_norm": 0.6737542748451233, + "learning_rate": 0.00012745577735062664, + "loss": 0.9255, + "step": 9257 + }, + { + "epoch": 1.6483262108262107, + "grad_norm": 0.686862051486969, + "learning_rate": 0.00012744231752391479, + "loss": 0.9493, + "step": 9258 + }, + { + "epoch": 1.6485042735042734, + "grad_norm": 0.6096474528312683, + "learning_rate": 0.00012742885715952772, + "loss": 0.6849, + "step": 9259 + }, + { + "epoch": 1.6486823361823362, + "grad_norm": 0.702751636505127, + "learning_rate": 0.00012741539625772918, + "loss": 1.0335, + "step": 9260 + }, + { + "epoch": 1.648860398860399, + "grad_norm": 0.7470958232879639, + "learning_rate": 0.0001274019348187829, + "loss": 1.105, + "step": 9261 + }, + { + "epoch": 1.6490384615384617, + "grad_norm": 0.6642739176750183, + "learning_rate": 0.0001273884728429526, + "loss": 1.01, + "step": 9262 + }, + { + "epoch": 1.6492165242165242, + "grad_norm": 0.6470904350280762, + "learning_rate": 0.00012737501033050213, + "loss": 0.9009, + "step": 9263 + }, + { + "epoch": 1.6493945868945867, + "grad_norm": 0.7487246990203857, + "learning_rate": 0.00012736154728169518, + "loss": 0.9832, + "step": 9264 + }, + { + "epoch": 1.6495726495726495, + "grad_norm": 0.7370779514312744, + "learning_rate": 0.00012734808369679553, + "loss": 1.0464, + "step": 9265 + }, + { + "epoch": 1.6497507122507122, + "grad_norm": 0.7942814826965332, + "learning_rate": 0.00012733461957606702, + "loss": 1.102, + "step": 9266 + }, + { + "epoch": 1.649928774928775, + "grad_norm": 0.6535606980323792, + "learning_rate": 0.00012732115491977336, + "loss": 1.0655, + "step": 9267 + }, + { + "epoch": 1.6501068376068377, + "grad_norm": 0.601716935634613, + "learning_rate": 0.00012730768972817847, + "loss": 0.8236, + "step": 9268 + }, + { + "epoch": 1.6502849002849003, + "grad_norm": 0.7375118732452393, + "learning_rate": 0.00012729422400154614, + "loss": 0.9313, + "step": 9269 + }, + { + "epoch": 1.6504629629629628, + "grad_norm": 0.7360411882400513, + "learning_rate": 0.00012728075774014018, + "loss": 0.9254, + "step": 9270 + }, + { + "epoch": 1.6506410256410255, + "grad_norm": 0.8453929424285889, + "learning_rate": 0.00012726729094422444, + "loss": 1.0975, + "step": 9271 + }, + { + "epoch": 1.6508190883190883, + "grad_norm": 0.5615501999855042, + "learning_rate": 0.00012725382361406274, + "loss": 0.8243, + "step": 9272 + }, + { + "epoch": 1.650997150997151, + "grad_norm": 0.6494898796081543, + "learning_rate": 0.000127240355749919, + "loss": 0.9766, + "step": 9273 + }, + { + "epoch": 1.6511752136752138, + "grad_norm": 0.6544778347015381, + "learning_rate": 0.0001272268873520571, + "loss": 0.9969, + "step": 9274 + }, + { + "epoch": 1.6513532763532763, + "grad_norm": 0.6937400698661804, + "learning_rate": 0.00012721341842074092, + "loss": 1.0626, + "step": 9275 + }, + { + "epoch": 1.651531339031339, + "grad_norm": 0.7068421244621277, + "learning_rate": 0.0001271999489562343, + "loss": 1.0068, + "step": 9276 + }, + { + "epoch": 1.6517094017094016, + "grad_norm": 0.6425052285194397, + "learning_rate": 0.0001271864789588012, + "loss": 0.8716, + "step": 9277 + }, + { + "epoch": 1.6518874643874644, + "grad_norm": 0.6895090341567993, + "learning_rate": 0.0001271730084287055, + "loss": 1.081, + "step": 9278 + }, + { + "epoch": 1.6520655270655271, + "grad_norm": 0.6773712038993835, + "learning_rate": 0.00012715953736621116, + "loss": 0.7586, + "step": 9279 + }, + { + "epoch": 1.6522435897435899, + "grad_norm": 0.6085716485977173, + "learning_rate": 0.0001271460657715821, + "loss": 0.8627, + "step": 9280 + }, + { + "epoch": 1.6524216524216524, + "grad_norm": 0.6415461897850037, + "learning_rate": 0.00012713259364508227, + "loss": 0.9751, + "step": 9281 + }, + { + "epoch": 1.6525997150997151, + "grad_norm": 0.6460939645767212, + "learning_rate": 0.00012711912098697565, + "loss": 0.9578, + "step": 9282 + }, + { + "epoch": 1.6527777777777777, + "grad_norm": 0.6076797246932983, + "learning_rate": 0.00012710564779752615, + "loss": 0.9627, + "step": 9283 + }, + { + "epoch": 1.6529558404558404, + "grad_norm": 0.710782527923584, + "learning_rate": 0.00012709217407699783, + "loss": 0.8725, + "step": 9284 + }, + { + "epoch": 1.6531339031339032, + "grad_norm": 0.6793623566627502, + "learning_rate": 0.00012707869982565463, + "loss": 0.908, + "step": 9285 + }, + { + "epoch": 1.653311965811966, + "grad_norm": 0.6841681003570557, + "learning_rate": 0.00012706522504376055, + "loss": 0.8546, + "step": 9286 + }, + { + "epoch": 1.6534900284900285, + "grad_norm": 0.7908675670623779, + "learning_rate": 0.0001270517497315796, + "loss": 0.9409, + "step": 9287 + }, + { + "epoch": 1.6536680911680912, + "grad_norm": 0.6918683648109436, + "learning_rate": 0.0001270382738893758, + "loss": 1.0493, + "step": 9288 + }, + { + "epoch": 1.6538461538461537, + "grad_norm": 0.6891819834709167, + "learning_rate": 0.00012702479751741322, + "loss": 1.0675, + "step": 9289 + }, + { + "epoch": 1.6540242165242165, + "grad_norm": 0.6965166926383972, + "learning_rate": 0.00012701132061595586, + "loss": 0.8563, + "step": 9290 + }, + { + "epoch": 1.6542022792022792, + "grad_norm": 0.7549001574516296, + "learning_rate": 0.00012699784318526779, + "loss": 1.1572, + "step": 9291 + }, + { + "epoch": 1.654380341880342, + "grad_norm": 0.6100513339042664, + "learning_rate": 0.00012698436522561303, + "loss": 0.897, + "step": 9292 + }, + { + "epoch": 1.6545584045584045, + "grad_norm": 0.6477037668228149, + "learning_rate": 0.00012697088673725574, + "loss": 0.7961, + "step": 9293 + }, + { + "epoch": 1.6547364672364673, + "grad_norm": 0.7402619123458862, + "learning_rate": 0.0001269574077204599, + "loss": 1.2001, + "step": 9294 + }, + { + "epoch": 1.6549145299145298, + "grad_norm": 0.7162346243858337, + "learning_rate": 0.0001269439281754897, + "loss": 0.9963, + "step": 9295 + }, + { + "epoch": 1.6550925925925926, + "grad_norm": 0.6757413744926453, + "learning_rate": 0.0001269304481026092, + "loss": 1.0476, + "step": 9296 + }, + { + "epoch": 1.6552706552706553, + "grad_norm": 0.6455655097961426, + "learning_rate": 0.0001269169675020825, + "loss": 0.9716, + "step": 9297 + }, + { + "epoch": 1.655448717948718, + "grad_norm": 0.7705031037330627, + "learning_rate": 0.0001269034863741737, + "loss": 0.9886, + "step": 9298 + }, + { + "epoch": 1.6556267806267806, + "grad_norm": 0.6084272861480713, + "learning_rate": 0.000126890004719147, + "loss": 0.8231, + "step": 9299 + }, + { + "epoch": 1.6558048433048433, + "grad_norm": 0.7051045298576355, + "learning_rate": 0.00012687652253726652, + "loss": 0.8673, + "step": 9300 + }, + { + "epoch": 1.6559829059829059, + "grad_norm": 0.731675386428833, + "learning_rate": 0.0001268630398287964, + "loss": 0.8609, + "step": 9301 + }, + { + "epoch": 1.6561609686609686, + "grad_norm": 0.6796799302101135, + "learning_rate": 0.00012684955659400087, + "loss": 1.0157, + "step": 9302 + }, + { + "epoch": 1.6563390313390314, + "grad_norm": 0.6270264983177185, + "learning_rate": 0.000126836072833144, + "loss": 0.8924, + "step": 9303 + }, + { + "epoch": 1.6565170940170941, + "grad_norm": 0.7235464453697205, + "learning_rate": 0.00012682258854649004, + "loss": 0.8904, + "step": 9304 + }, + { + "epoch": 1.6566951566951567, + "grad_norm": 0.7644724249839783, + "learning_rate": 0.00012680910373430318, + "loss": 0.9119, + "step": 9305 + }, + { + "epoch": 1.6568732193732194, + "grad_norm": 0.661411702632904, + "learning_rate": 0.00012679561839684764, + "loss": 1.0066, + "step": 9306 + }, + { + "epoch": 1.657051282051282, + "grad_norm": 0.6981723308563232, + "learning_rate": 0.0001267821325343876, + "loss": 1.2579, + "step": 9307 + }, + { + "epoch": 1.6572293447293447, + "grad_norm": 0.6469807028770447, + "learning_rate": 0.0001267686461471873, + "loss": 0.8678, + "step": 9308 + }, + { + "epoch": 1.6574074074074074, + "grad_norm": 0.8255495429039001, + "learning_rate": 0.000126755159235511, + "loss": 0.9053, + "step": 9309 + }, + { + "epoch": 1.6575854700854702, + "grad_norm": 0.6882261037826538, + "learning_rate": 0.00012674167179962294, + "loss": 0.8364, + "step": 9310 + }, + { + "epoch": 1.6577635327635327, + "grad_norm": 0.6816701889038086, + "learning_rate": 0.00012672818383978733, + "loss": 0.9627, + "step": 9311 + }, + { + "epoch": 1.6579415954415955, + "grad_norm": 0.6993424892425537, + "learning_rate": 0.00012671469535626852, + "loss": 0.8337, + "step": 9312 + }, + { + "epoch": 1.658119658119658, + "grad_norm": 0.6271458864212036, + "learning_rate": 0.00012670120634933075, + "loss": 0.8322, + "step": 9313 + }, + { + "epoch": 1.6582977207977208, + "grad_norm": 0.7012003660202026, + "learning_rate": 0.00012668771681923827, + "loss": 0.8895, + "step": 9314 + }, + { + "epoch": 1.6584757834757835, + "grad_norm": 0.6704670190811157, + "learning_rate": 0.00012667422676625547, + "loss": 1.0544, + "step": 9315 + }, + { + "epoch": 1.6586538461538463, + "grad_norm": 0.6189491748809814, + "learning_rate": 0.0001266607361906466, + "loss": 0.9623, + "step": 9316 + }, + { + "epoch": 1.6588319088319088, + "grad_norm": 0.7065694332122803, + "learning_rate": 0.000126647245092676, + "loss": 0.8874, + "step": 9317 + }, + { + "epoch": 1.6590099715099715, + "grad_norm": 0.7473452687263489, + "learning_rate": 0.00012663375347260795, + "loss": 1.0576, + "step": 9318 + }, + { + "epoch": 1.659188034188034, + "grad_norm": 0.6839408874511719, + "learning_rate": 0.0001266202613307068, + "loss": 0.9127, + "step": 9319 + }, + { + "epoch": 1.6593660968660968, + "grad_norm": 0.7154020071029663, + "learning_rate": 0.00012660676866723699, + "loss": 1.1174, + "step": 9320 + }, + { + "epoch": 1.6595441595441596, + "grad_norm": 0.7123729586601257, + "learning_rate": 0.0001265932754824628, + "loss": 0.9617, + "step": 9321 + }, + { + "epoch": 1.6597222222222223, + "grad_norm": 0.7537810802459717, + "learning_rate": 0.0001265797817766486, + "loss": 1.0333, + "step": 9322 + }, + { + "epoch": 1.6599002849002849, + "grad_norm": 0.706551730632782, + "learning_rate": 0.00012656628755005884, + "loss": 1.0838, + "step": 9323 + }, + { + "epoch": 1.6600783475783476, + "grad_norm": 0.8104004859924316, + "learning_rate": 0.0001265527928029578, + "loss": 0.9807, + "step": 9324 + }, + { + "epoch": 1.6602564102564101, + "grad_norm": 0.6892881989479065, + "learning_rate": 0.00012653929753560998, + "loss": 0.9941, + "step": 9325 + }, + { + "epoch": 1.6604344729344729, + "grad_norm": 0.5919203758239746, + "learning_rate": 0.00012652580174827974, + "loss": 0.9268, + "step": 9326 + }, + { + "epoch": 1.6606125356125356, + "grad_norm": 0.6715863347053528, + "learning_rate": 0.00012651230544123154, + "loss": 1.0912, + "step": 9327 + }, + { + "epoch": 1.6607905982905984, + "grad_norm": 0.6765137314796448, + "learning_rate": 0.0001264988086147298, + "loss": 1.1576, + "step": 9328 + }, + { + "epoch": 1.660968660968661, + "grad_norm": 0.6781638860702515, + "learning_rate": 0.00012648531126903888, + "loss": 1.1162, + "step": 9329 + }, + { + "epoch": 1.6611467236467237, + "grad_norm": 0.715871274471283, + "learning_rate": 0.00012647181340442337, + "loss": 0.714, + "step": 9330 + }, + { + "epoch": 1.6613247863247862, + "grad_norm": 0.6237258315086365, + "learning_rate": 0.00012645831502114762, + "loss": 0.8512, + "step": 9331 + }, + { + "epoch": 1.661502849002849, + "grad_norm": 0.6668339967727661, + "learning_rate": 0.0001264448161194762, + "loss": 1.0384, + "step": 9332 + }, + { + "epoch": 1.6616809116809117, + "grad_norm": 0.8316730260848999, + "learning_rate": 0.00012643131669967352, + "loss": 0.8931, + "step": 9333 + }, + { + "epoch": 1.6618589743589745, + "grad_norm": 0.7013183832168579, + "learning_rate": 0.00012641781676200406, + "loss": 1.0548, + "step": 9334 + }, + { + "epoch": 1.6620370370370372, + "grad_norm": 0.6980466842651367, + "learning_rate": 0.00012640431630673243, + "loss": 0.8988, + "step": 9335 + }, + { + "epoch": 1.6622150997150997, + "grad_norm": 0.7045995593070984, + "learning_rate": 0.000126390815334123, + "loss": 1.107, + "step": 9336 + }, + { + "epoch": 1.6623931623931623, + "grad_norm": 0.6699773669242859, + "learning_rate": 0.00012637731384444043, + "loss": 1.1757, + "step": 9337 + }, + { + "epoch": 1.662571225071225, + "grad_norm": 0.6489999294281006, + "learning_rate": 0.00012636381183794916, + "loss": 0.9282, + "step": 9338 + }, + { + "epoch": 1.6627492877492878, + "grad_norm": 0.7085952758789062, + "learning_rate": 0.00012635030931491375, + "loss": 1.0221, + "step": 9339 + }, + { + "epoch": 1.6629273504273505, + "grad_norm": 0.6893135905265808, + "learning_rate": 0.00012633680627559878, + "loss": 1.0517, + "step": 9340 + }, + { + "epoch": 1.6631054131054133, + "grad_norm": 0.5659682154655457, + "learning_rate": 0.00012632330272026882, + "loss": 0.6294, + "step": 9341 + }, + { + "epoch": 1.6632834757834758, + "grad_norm": 0.6889018416404724, + "learning_rate": 0.00012630979864918838, + "loss": 1.0735, + "step": 9342 + }, + { + "epoch": 1.6634615384615383, + "grad_norm": 0.7333424687385559, + "learning_rate": 0.00012629629406262212, + "loss": 0.9079, + "step": 9343 + }, + { + "epoch": 1.663639601139601, + "grad_norm": 0.6340580582618713, + "learning_rate": 0.00012628278896083462, + "loss": 0.9738, + "step": 9344 + }, + { + "epoch": 1.6638176638176638, + "grad_norm": 0.7042564749717712, + "learning_rate": 0.00012626928334409044, + "loss": 0.959, + "step": 9345 + }, + { + "epoch": 1.6639957264957266, + "grad_norm": 0.711757242679596, + "learning_rate": 0.00012625577721265424, + "loss": 0.8113, + "step": 9346 + }, + { + "epoch": 1.6641737891737893, + "grad_norm": 0.7723299264907837, + "learning_rate": 0.0001262422705667906, + "loss": 1.1724, + "step": 9347 + }, + { + "epoch": 1.6643518518518519, + "grad_norm": 0.711334228515625, + "learning_rate": 0.00012622876340676422, + "loss": 1.0121, + "step": 9348 + }, + { + "epoch": 1.6645299145299144, + "grad_norm": 0.6954590678215027, + "learning_rate": 0.0001262152557328397, + "loss": 1.2093, + "step": 9349 + }, + { + "epoch": 1.6647079772079771, + "grad_norm": 0.6341620087623596, + "learning_rate": 0.00012620174754528166, + "loss": 1.0535, + "step": 9350 + }, + { + "epoch": 1.66488603988604, + "grad_norm": 0.6434268355369568, + "learning_rate": 0.00012618823884435484, + "loss": 0.8964, + "step": 9351 + }, + { + "epoch": 1.6650641025641026, + "grad_norm": 0.7685084939002991, + "learning_rate": 0.00012617472963032385, + "loss": 1.0639, + "step": 9352 + }, + { + "epoch": 1.6652421652421654, + "grad_norm": 0.6347958445549011, + "learning_rate": 0.00012616121990345345, + "loss": 1.0252, + "step": 9353 + }, + { + "epoch": 1.665420227920228, + "grad_norm": 0.647722601890564, + "learning_rate": 0.0001261477096640083, + "loss": 0.9527, + "step": 9354 + }, + { + "epoch": 1.6655982905982905, + "grad_norm": 0.5942047834396362, + "learning_rate": 0.000126134198912253, + "loss": 1.0062, + "step": 9355 + }, + { + "epoch": 1.6657763532763532, + "grad_norm": 0.683555006980896, + "learning_rate": 0.00012612068764845247, + "loss": 0.8101, + "step": 9356 + }, + { + "epoch": 1.665954415954416, + "grad_norm": 0.6832289099693298, + "learning_rate": 0.00012610717587287128, + "loss": 1.1436, + "step": 9357 + }, + { + "epoch": 1.6661324786324787, + "grad_norm": 0.7035253047943115, + "learning_rate": 0.00012609366358577422, + "loss": 0.9724, + "step": 9358 + }, + { + "epoch": 1.6663105413105415, + "grad_norm": 0.6471409797668457, + "learning_rate": 0.00012608015078742604, + "loss": 0.776, + "step": 9359 + }, + { + "epoch": 1.666488603988604, + "grad_norm": 0.7069687247276306, + "learning_rate": 0.00012606663747809145, + "loss": 0.9667, + "step": 9360 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.6744135618209839, + "learning_rate": 0.00012605312365803525, + "loss": 1.1152, + "step": 9361 + }, + { + "epoch": 1.6668447293447293, + "grad_norm": 0.7212334275245667, + "learning_rate": 0.00012603960932752227, + "loss": 1.1543, + "step": 9362 + }, + { + "epoch": 1.667022792022792, + "grad_norm": 0.6501669883728027, + "learning_rate": 0.0001260260944868172, + "loss": 0.8595, + "step": 9363 + }, + { + "epoch": 1.6672008547008548, + "grad_norm": 0.6970864534378052, + "learning_rate": 0.00012601257913618486, + "loss": 0.9364, + "step": 9364 + }, + { + "epoch": 1.6673789173789175, + "grad_norm": 0.6802223324775696, + "learning_rate": 0.00012599906327589007, + "loss": 0.8429, + "step": 9365 + }, + { + "epoch": 1.66755698005698, + "grad_norm": 0.6842933893203735, + "learning_rate": 0.00012598554690619764, + "loss": 1.1255, + "step": 9366 + }, + { + "epoch": 1.6677350427350426, + "grad_norm": 0.6547088623046875, + "learning_rate": 0.0001259720300273724, + "loss": 0.983, + "step": 9367 + }, + { + "epoch": 1.6679131054131053, + "grad_norm": 0.620424211025238, + "learning_rate": 0.0001259585126396792, + "loss": 0.918, + "step": 9368 + }, + { + "epoch": 1.668091168091168, + "grad_norm": 0.5659816861152649, + "learning_rate": 0.00012594499474338287, + "loss": 0.7788, + "step": 9369 + }, + { + "epoch": 1.6682692307692308, + "grad_norm": 0.5904595255851746, + "learning_rate": 0.00012593147633874826, + "loss": 0.801, + "step": 9370 + }, + { + "epoch": 1.6684472934472936, + "grad_norm": 0.6444024443626404, + "learning_rate": 0.0001259179574260402, + "loss": 1.0997, + "step": 9371 + }, + { + "epoch": 1.6686253561253561, + "grad_norm": 0.6408827304840088, + "learning_rate": 0.00012590443800552365, + "loss": 0.9839, + "step": 9372 + }, + { + "epoch": 1.6688034188034186, + "grad_norm": 0.752391517162323, + "learning_rate": 0.00012589091807746345, + "loss": 1.0249, + "step": 9373 + }, + { + "epoch": 1.6689814814814814, + "grad_norm": 0.8256397247314453, + "learning_rate": 0.00012587739764212448, + "loss": 0.9541, + "step": 9374 + }, + { + "epoch": 1.6691595441595442, + "grad_norm": 0.7878768444061279, + "learning_rate": 0.00012586387669977166, + "loss": 1.0071, + "step": 9375 + }, + { + "epoch": 1.669337606837607, + "grad_norm": 0.6179735660552979, + "learning_rate": 0.0001258503552506699, + "loss": 0.8495, + "step": 9376 + }, + { + "epoch": 1.6695156695156697, + "grad_norm": 0.6699580550193787, + "learning_rate": 0.00012583683329508413, + "loss": 0.8999, + "step": 9377 + }, + { + "epoch": 1.6696937321937322, + "grad_norm": 0.6542006731033325, + "learning_rate": 0.00012582331083327929, + "loss": 1.0357, + "step": 9378 + }, + { + "epoch": 1.6698717948717947, + "grad_norm": 0.7275210618972778, + "learning_rate": 0.0001258097878655203, + "loss": 1.0259, + "step": 9379 + }, + { + "epoch": 1.6700498575498575, + "grad_norm": 0.6836326122283936, + "learning_rate": 0.00012579626439207216, + "loss": 1.0428, + "step": 9380 + }, + { + "epoch": 1.6702279202279202, + "grad_norm": 0.760123610496521, + "learning_rate": 0.00012578274041319978, + "loss": 0.9716, + "step": 9381 + }, + { + "epoch": 1.670405982905983, + "grad_norm": 0.5525194406509399, + "learning_rate": 0.00012576921592916818, + "loss": 0.8253, + "step": 9382 + }, + { + "epoch": 1.6705840455840457, + "grad_norm": 0.6881270408630371, + "learning_rate": 0.00012575569094024232, + "loss": 1.0571, + "step": 9383 + }, + { + "epoch": 1.6707621082621082, + "grad_norm": 0.6776245832443237, + "learning_rate": 0.0001257421654466872, + "loss": 0.9119, + "step": 9384 + }, + { + "epoch": 1.6709401709401708, + "grad_norm": 0.7903014421463013, + "learning_rate": 0.0001257286394487678, + "loss": 1.0626, + "step": 9385 + }, + { + "epoch": 1.6711182336182335, + "grad_norm": 0.61158287525177, + "learning_rate": 0.0001257151129467492, + "loss": 0.9378, + "step": 9386 + }, + { + "epoch": 1.6712962962962963, + "grad_norm": 0.655189573764801, + "learning_rate": 0.00012570158594089637, + "loss": 0.9334, + "step": 9387 + }, + { + "epoch": 1.671474358974359, + "grad_norm": 0.6707320809364319, + "learning_rate": 0.0001256880584314743, + "loss": 1.1802, + "step": 9388 + }, + { + "epoch": 1.6716524216524218, + "grad_norm": 0.847341775894165, + "learning_rate": 0.00012567453041874814, + "loss": 1.1169, + "step": 9389 + }, + { + "epoch": 1.6718304843304843, + "grad_norm": 0.6136410236358643, + "learning_rate": 0.00012566100190298287, + "loss": 0.8959, + "step": 9390 + }, + { + "epoch": 1.672008547008547, + "grad_norm": 0.7203437089920044, + "learning_rate": 0.00012564747288444357, + "loss": 0.9803, + "step": 9391 + }, + { + "epoch": 1.6721866096866096, + "grad_norm": 0.7832576632499695, + "learning_rate": 0.00012563394336339534, + "loss": 0.8696, + "step": 9392 + }, + { + "epoch": 1.6723646723646723, + "grad_norm": 0.6940804719924927, + "learning_rate": 0.00012562041334010323, + "loss": 1.0571, + "step": 9393 + }, + { + "epoch": 1.672542735042735, + "grad_norm": 0.6042298674583435, + "learning_rate": 0.00012560688281483234, + "loss": 0.8835, + "step": 9394 + }, + { + "epoch": 1.6727207977207978, + "grad_norm": 0.7870675921440125, + "learning_rate": 0.00012559335178784776, + "loss": 1.1585, + "step": 9395 + }, + { + "epoch": 1.6728988603988604, + "grad_norm": 0.7448568940162659, + "learning_rate": 0.00012557982025941463, + "loss": 0.9699, + "step": 9396 + }, + { + "epoch": 1.6730769230769231, + "grad_norm": 0.7226544618606567, + "learning_rate": 0.00012556628822979807, + "loss": 0.7817, + "step": 9397 + }, + { + "epoch": 1.6732549857549857, + "grad_norm": 0.5652043223381042, + "learning_rate": 0.0001255527556992632, + "loss": 0.8077, + "step": 9398 + }, + { + "epoch": 1.6734330484330484, + "grad_norm": 0.6459930539131165, + "learning_rate": 0.00012553922266807517, + "loss": 1.22, + "step": 9399 + }, + { + "epoch": 1.6736111111111112, + "grad_norm": 0.7568991780281067, + "learning_rate": 0.00012552568913649912, + "loss": 1.1559, + "step": 9400 + }, + { + "epoch": 1.673789173789174, + "grad_norm": 0.7462680339813232, + "learning_rate": 0.0001255121551048002, + "loss": 1.1438, + "step": 9401 + }, + { + "epoch": 1.6739672364672364, + "grad_norm": 0.6653871536254883, + "learning_rate": 0.0001254986205732436, + "loss": 0.9468, + "step": 9402 + }, + { + "epoch": 1.6741452991452992, + "grad_norm": 0.6261825561523438, + "learning_rate": 0.0001254850855420945, + "loss": 0.8558, + "step": 9403 + }, + { + "epoch": 1.6743233618233617, + "grad_norm": 0.6442354321479797, + "learning_rate": 0.0001254715500116181, + "loss": 0.8605, + "step": 9404 + }, + { + "epoch": 1.6745014245014245, + "grad_norm": 0.7483665943145752, + "learning_rate": 0.00012545801398207958, + "loss": 0.9089, + "step": 9405 + }, + { + "epoch": 1.6746794871794872, + "grad_norm": 0.7319819927215576, + "learning_rate": 0.00012544447745374416, + "loss": 0.9937, + "step": 9406 + }, + { + "epoch": 1.67485754985755, + "grad_norm": 0.703014075756073, + "learning_rate": 0.00012543094042687708, + "loss": 0.9597, + "step": 9407 + }, + { + "epoch": 1.6750356125356125, + "grad_norm": 0.6593887209892273, + "learning_rate": 0.00012541740290174353, + "loss": 0.844, + "step": 9408 + }, + { + "epoch": 1.6752136752136753, + "grad_norm": 0.6567463874816895, + "learning_rate": 0.00012540386487860879, + "loss": 1.0744, + "step": 9409 + }, + { + "epoch": 1.6753917378917378, + "grad_norm": 0.7784611582756042, + "learning_rate": 0.00012539032635773805, + "loss": 0.974, + "step": 9410 + }, + { + "epoch": 1.6755698005698005, + "grad_norm": 0.6760087609291077, + "learning_rate": 0.00012537678733939663, + "loss": 0.8948, + "step": 9411 + }, + { + "epoch": 1.6757478632478633, + "grad_norm": 0.825965940952301, + "learning_rate": 0.0001253632478238498, + "loss": 1.1196, + "step": 9412 + }, + { + "epoch": 1.675925925925926, + "grad_norm": 0.7215564250946045, + "learning_rate": 0.00012534970781136277, + "loss": 1.1774, + "step": 9413 + }, + { + "epoch": 1.6761039886039886, + "grad_norm": 0.6548578143119812, + "learning_rate": 0.00012533616730220094, + "loss": 0.8671, + "step": 9414 + }, + { + "epoch": 1.6762820512820513, + "grad_norm": 0.7257684469223022, + "learning_rate": 0.00012532262629662947, + "loss": 1.105, + "step": 9415 + }, + { + "epoch": 1.6764601139601139, + "grad_norm": 0.6695847511291504, + "learning_rate": 0.00012530908479491378, + "loss": 0.9189, + "step": 9416 + }, + { + "epoch": 1.6766381766381766, + "grad_norm": 0.684695303440094, + "learning_rate": 0.00012529554279731915, + "loss": 1.066, + "step": 9417 + }, + { + "epoch": 1.6768162393162394, + "grad_norm": 0.7107276320457458, + "learning_rate": 0.0001252820003041109, + "loss": 0.9311, + "step": 9418 + }, + { + "epoch": 1.676994301994302, + "grad_norm": 0.6755440831184387, + "learning_rate": 0.0001252684573155544, + "loss": 1.1036, + "step": 9419 + }, + { + "epoch": 1.6771723646723646, + "grad_norm": 0.7571110725402832, + "learning_rate": 0.00012525491383191491, + "loss": 1.0244, + "step": 9420 + }, + { + "epoch": 1.6773504273504274, + "grad_norm": 0.6960614323616028, + "learning_rate": 0.0001252413698534579, + "loss": 0.9077, + "step": 9421 + }, + { + "epoch": 1.67752849002849, + "grad_norm": 0.6675550937652588, + "learning_rate": 0.00012522782538044867, + "loss": 1.0543, + "step": 9422 + }, + { + "epoch": 1.6777065527065527, + "grad_norm": 0.6637391448020935, + "learning_rate": 0.0001252142804131526, + "loss": 0.9471, + "step": 9423 + }, + { + "epoch": 1.6778846153846154, + "grad_norm": 0.6382880210876465, + "learning_rate": 0.00012520073495183508, + "loss": 0.9729, + "step": 9424 + }, + { + "epoch": 1.6780626780626782, + "grad_norm": 0.731922447681427, + "learning_rate": 0.0001251871889967615, + "loss": 1.0385, + "step": 9425 + }, + { + "epoch": 1.6782407407407407, + "grad_norm": 0.5868890285491943, + "learning_rate": 0.00012517364254819728, + "loss": 0.8466, + "step": 9426 + }, + { + "epoch": 1.6784188034188035, + "grad_norm": 0.8535677790641785, + "learning_rate": 0.00012516009560640786, + "loss": 1.1009, + "step": 9427 + }, + { + "epoch": 1.678596866096866, + "grad_norm": 0.7044199705123901, + "learning_rate": 0.0001251465481716586, + "loss": 1.0862, + "step": 9428 + }, + { + "epoch": 1.6787749287749287, + "grad_norm": 0.7207323312759399, + "learning_rate": 0.00012513300024421498, + "loss": 1.064, + "step": 9429 + }, + { + "epoch": 1.6789529914529915, + "grad_norm": 0.7739703059196472, + "learning_rate": 0.0001251194518243424, + "loss": 1.1738, + "step": 9430 + }, + { + "epoch": 1.6791310541310542, + "grad_norm": 0.6829344630241394, + "learning_rate": 0.00012510590291230637, + "loss": 1.0555, + "step": 9431 + }, + { + "epoch": 1.6793091168091168, + "grad_norm": 0.6760238409042358, + "learning_rate": 0.0001250923535083723, + "loss": 1.2177, + "step": 9432 + }, + { + "epoch": 1.6794871794871795, + "grad_norm": 0.6666911840438843, + "learning_rate": 0.0001250788036128057, + "loss": 0.8957, + "step": 9433 + }, + { + "epoch": 1.679665242165242, + "grad_norm": 0.747797429561615, + "learning_rate": 0.00012506525322587207, + "loss": 0.9793, + "step": 9434 + }, + { + "epoch": 1.6798433048433048, + "grad_norm": 0.6261107325553894, + "learning_rate": 0.00012505170234783686, + "loss": 0.7781, + "step": 9435 + }, + { + "epoch": 1.6800213675213675, + "grad_norm": 0.7055163979530334, + "learning_rate": 0.00012503815097896555, + "loss": 1.0617, + "step": 9436 + }, + { + "epoch": 1.6801994301994303, + "grad_norm": 0.5567409992218018, + "learning_rate": 0.00012502459911952371, + "loss": 0.7911, + "step": 9437 + }, + { + "epoch": 1.6803774928774928, + "grad_norm": 0.7410423159599304, + "learning_rate": 0.0001250110467697768, + "loss": 1.1041, + "step": 9438 + }, + { + "epoch": 1.6805555555555556, + "grad_norm": 0.6185283064842224, + "learning_rate": 0.00012499749392999045, + "loss": 0.8101, + "step": 9439 + }, + { + "epoch": 1.680733618233618, + "grad_norm": 0.6988311409950256, + "learning_rate": 0.0001249839406004301, + "loss": 0.8579, + "step": 9440 + }, + { + "epoch": 1.6809116809116809, + "grad_norm": 0.5588746070861816, + "learning_rate": 0.00012497038678136132, + "loss": 0.8035, + "step": 9441 + }, + { + "epoch": 1.6810897435897436, + "grad_norm": 0.6568905711174011, + "learning_rate": 0.0001249568324730497, + "loss": 0.7455, + "step": 9442 + }, + { + "epoch": 1.6812678062678064, + "grad_norm": 0.6924821138381958, + "learning_rate": 0.00012494327767576078, + "loss": 1.134, + "step": 9443 + }, + { + "epoch": 1.681445868945869, + "grad_norm": 0.6940170526504517, + "learning_rate": 0.00012492972238976018, + "loss": 0.9719, + "step": 9444 + }, + { + "epoch": 1.6816239316239316, + "grad_norm": 0.667465090751648, + "learning_rate": 0.00012491616661531343, + "loss": 0.953, + "step": 9445 + }, + { + "epoch": 1.6818019943019942, + "grad_norm": 0.7693275809288025, + "learning_rate": 0.00012490261035268612, + "loss": 1.1342, + "step": 9446 + }, + { + "epoch": 1.681980056980057, + "grad_norm": 0.7243115305900574, + "learning_rate": 0.00012488905360214393, + "loss": 1.1847, + "step": 9447 + }, + { + "epoch": 1.6821581196581197, + "grad_norm": 0.657357931137085, + "learning_rate": 0.00012487549636395245, + "loss": 0.8747, + "step": 9448 + }, + { + "epoch": 1.6823361823361824, + "grad_norm": 0.7471592426300049, + "learning_rate": 0.00012486193863837727, + "loss": 1.0472, + "step": 9449 + }, + { + "epoch": 1.6825142450142452, + "grad_norm": 0.7476530075073242, + "learning_rate": 0.00012484838042568406, + "loss": 1.0708, + "step": 9450 + }, + { + "epoch": 1.6826923076923077, + "grad_norm": 0.6031121611595154, + "learning_rate": 0.00012483482172613846, + "loss": 0.8243, + "step": 9451 + }, + { + "epoch": 1.6828703703703702, + "grad_norm": 0.6733492016792297, + "learning_rate": 0.00012482126254000607, + "loss": 0.7808, + "step": 9452 + }, + { + "epoch": 1.683048433048433, + "grad_norm": 0.5865318179130554, + "learning_rate": 0.00012480770286755265, + "loss": 0.829, + "step": 9453 + }, + { + "epoch": 1.6832264957264957, + "grad_norm": 0.6805713772773743, + "learning_rate": 0.0001247941427090438, + "loss": 0.7206, + "step": 9454 + }, + { + "epoch": 1.6834045584045585, + "grad_norm": 0.6514836549758911, + "learning_rate": 0.0001247805820647453, + "loss": 0.9499, + "step": 9455 + }, + { + "epoch": 1.6835826210826212, + "grad_norm": 0.7432990074157715, + "learning_rate": 0.0001247670209349227, + "loss": 1.1324, + "step": 9456 + }, + { + "epoch": 1.6837606837606838, + "grad_norm": 0.6348414421081543, + "learning_rate": 0.00012475345931984178, + "loss": 0.8246, + "step": 9457 + }, + { + "epoch": 1.6839387464387463, + "grad_norm": 0.7194374203681946, + "learning_rate": 0.00012473989721976825, + "loss": 0.9634, + "step": 9458 + }, + { + "epoch": 1.684116809116809, + "grad_norm": 0.7869647741317749, + "learning_rate": 0.00012472633463496785, + "loss": 1.2115, + "step": 9459 + }, + { + "epoch": 1.6842948717948718, + "grad_norm": 0.6672070026397705, + "learning_rate": 0.00012471277156570623, + "loss": 0.9842, + "step": 9460 + }, + { + "epoch": 1.6844729344729346, + "grad_norm": 0.6611466407775879, + "learning_rate": 0.00012469920801224925, + "loss": 0.9343, + "step": 9461 + }, + { + "epoch": 1.6846509971509973, + "grad_norm": 0.6715068221092224, + "learning_rate": 0.0001246856439748626, + "loss": 0.6852, + "step": 9462 + }, + { + "epoch": 1.6848290598290598, + "grad_norm": 0.641942024230957, + "learning_rate": 0.00012467207945381198, + "loss": 0.8863, + "step": 9463 + }, + { + "epoch": 1.6850071225071224, + "grad_norm": 0.8414762616157532, + "learning_rate": 0.00012465851444936325, + "loss": 1.3404, + "step": 9464 + }, + { + "epoch": 1.6851851851851851, + "grad_norm": 0.715752363204956, + "learning_rate": 0.00012464494896178216, + "loss": 1.123, + "step": 9465 + }, + { + "epoch": 1.6853632478632479, + "grad_norm": 0.6913973093032837, + "learning_rate": 0.00012463138299133447, + "loss": 1.0659, + "step": 9466 + }, + { + "epoch": 1.6855413105413106, + "grad_norm": 0.6998484134674072, + "learning_rate": 0.000124617816538286, + "loss": 1.0555, + "step": 9467 + }, + { + "epoch": 1.6857193732193734, + "grad_norm": 0.7313308119773865, + "learning_rate": 0.00012460424960290256, + "loss": 1.0915, + "step": 9468 + }, + { + "epoch": 1.685897435897436, + "grad_norm": 0.6790569424629211, + "learning_rate": 0.00012459068218544995, + "loss": 1.0214, + "step": 9469 + }, + { + "epoch": 1.6860754985754984, + "grad_norm": 0.6494466662406921, + "learning_rate": 0.00012457711428619402, + "loss": 0.9476, + "step": 9470 + }, + { + "epoch": 1.6862535612535612, + "grad_norm": 0.8048526048660278, + "learning_rate": 0.0001245635459054006, + "loss": 1.1852, + "step": 9471 + }, + { + "epoch": 1.686431623931624, + "grad_norm": 0.6237879395484924, + "learning_rate": 0.0001245499770433355, + "loss": 1.0106, + "step": 9472 + }, + { + "epoch": 1.6866096866096867, + "grad_norm": 0.6282906532287598, + "learning_rate": 0.0001245364077002646, + "loss": 0.9858, + "step": 9473 + }, + { + "epoch": 1.6867877492877494, + "grad_norm": 0.7239370346069336, + "learning_rate": 0.00012452283787645375, + "loss": 0.9586, + "step": 9474 + }, + { + "epoch": 1.686965811965812, + "grad_norm": 0.6438776850700378, + "learning_rate": 0.00012450926757216887, + "loss": 0.9198, + "step": 9475 + }, + { + "epoch": 1.6871438746438745, + "grad_norm": 0.6451360583305359, + "learning_rate": 0.00012449569678767578, + "loss": 1.0183, + "step": 9476 + }, + { + "epoch": 1.6873219373219372, + "grad_norm": 0.6950216293334961, + "learning_rate": 0.0001244821255232404, + "loss": 0.9048, + "step": 9477 + }, + { + "epoch": 1.6875, + "grad_norm": 0.710489809513092, + "learning_rate": 0.00012446855377912865, + "loss": 1.1596, + "step": 9478 + }, + { + "epoch": 1.6876780626780628, + "grad_norm": 0.6819305419921875, + "learning_rate": 0.0001244549815556064, + "loss": 0.8486, + "step": 9479 + }, + { + "epoch": 1.6878561253561255, + "grad_norm": 0.7185879945755005, + "learning_rate": 0.00012444140885293958, + "loss": 0.9539, + "step": 9480 + }, + { + "epoch": 1.688034188034188, + "grad_norm": 0.8181464672088623, + "learning_rate": 0.00012442783567139415, + "loss": 1.0038, + "step": 9481 + }, + { + "epoch": 1.6882122507122506, + "grad_norm": 0.47161349654197693, + "learning_rate": 0.000124414262011236, + "loss": 0.67, + "step": 9482 + }, + { + "epoch": 1.6883903133903133, + "grad_norm": 0.7752482295036316, + "learning_rate": 0.00012440068787273112, + "loss": 0.9944, + "step": 9483 + }, + { + "epoch": 1.688568376068376, + "grad_norm": 0.7119397521018982, + "learning_rate": 0.00012438711325614543, + "loss": 0.9098, + "step": 9484 + }, + { + "epoch": 1.6887464387464388, + "grad_norm": 0.7161153554916382, + "learning_rate": 0.00012437353816174493, + "loss": 1.0003, + "step": 9485 + }, + { + "epoch": 1.6889245014245016, + "grad_norm": 0.5989507436752319, + "learning_rate": 0.0001243599625897956, + "loss": 1.0301, + "step": 9486 + }, + { + "epoch": 1.689102564102564, + "grad_norm": 0.7906841039657593, + "learning_rate": 0.00012434638654056334, + "loss": 1.0388, + "step": 9487 + }, + { + "epoch": 1.6892806267806266, + "grad_norm": 0.6679551601409912, + "learning_rate": 0.00012433281001431428, + "loss": 0.9505, + "step": 9488 + }, + { + "epoch": 1.6894586894586894, + "grad_norm": 0.7090578675270081, + "learning_rate": 0.0001243192330113143, + "loss": 0.8616, + "step": 9489 + }, + { + "epoch": 1.6896367521367521, + "grad_norm": 0.6401308178901672, + "learning_rate": 0.00012430565553182949, + "loss": 0.9099, + "step": 9490 + }, + { + "epoch": 1.6898148148148149, + "grad_norm": 0.7360149621963501, + "learning_rate": 0.00012429207757612586, + "loss": 1.0233, + "step": 9491 + }, + { + "epoch": 1.6899928774928776, + "grad_norm": 0.6736137270927429, + "learning_rate": 0.00012427849914446946, + "loss": 0.9803, + "step": 9492 + }, + { + "epoch": 1.6901709401709402, + "grad_norm": 0.7728668451309204, + "learning_rate": 0.00012426492023712623, + "loss": 1.2316, + "step": 9493 + }, + { + "epoch": 1.6903490028490027, + "grad_norm": 0.789718508720398, + "learning_rate": 0.00012425134085436234, + "loss": 1.1218, + "step": 9494 + }, + { + "epoch": 1.6905270655270654, + "grad_norm": 0.7314121723175049, + "learning_rate": 0.0001242377609964438, + "loss": 1.1294, + "step": 9495 + }, + { + "epoch": 1.6907051282051282, + "grad_norm": 0.7222046256065369, + "learning_rate": 0.0001242241806636367, + "loss": 1.0288, + "step": 9496 + }, + { + "epoch": 1.690883190883191, + "grad_norm": 0.7546363472938538, + "learning_rate": 0.00012421059985620708, + "loss": 0.8781, + "step": 9497 + }, + { + "epoch": 1.6910612535612537, + "grad_norm": 0.7502550482749939, + "learning_rate": 0.00012419701857442104, + "loss": 0.927, + "step": 9498 + }, + { + "epoch": 1.6912393162393162, + "grad_norm": 0.6244059205055237, + "learning_rate": 0.00012418343681854473, + "loss": 0.9689, + "step": 9499 + }, + { + "epoch": 1.6914173789173788, + "grad_norm": 0.7214263677597046, + "learning_rate": 0.00012416985458884417, + "loss": 1.0842, + "step": 9500 + }, + { + "epoch": 1.6915954415954415, + "grad_norm": 0.6960242390632629, + "learning_rate": 0.00012415627188558555, + "loss": 0.9766, + "step": 9501 + }, + { + "epoch": 1.6917735042735043, + "grad_norm": 0.6687830686569214, + "learning_rate": 0.00012414268870903494, + "loss": 1.0222, + "step": 9502 + }, + { + "epoch": 1.691951566951567, + "grad_norm": 0.8611155152320862, + "learning_rate": 0.00012412910505945848, + "loss": 1.1792, + "step": 9503 + }, + { + "epoch": 1.6921296296296298, + "grad_norm": 0.6655587553977966, + "learning_rate": 0.00012411552093712235, + "loss": 0.8763, + "step": 9504 + }, + { + "epoch": 1.6923076923076923, + "grad_norm": 0.7829837799072266, + "learning_rate": 0.00012410193634229268, + "loss": 1.0803, + "step": 9505 + }, + { + "epoch": 1.6924857549857548, + "grad_norm": 0.7951042652130127, + "learning_rate": 0.00012408835127523566, + "loss": 1.0925, + "step": 9506 + }, + { + "epoch": 1.6926638176638176, + "grad_norm": 0.715495228767395, + "learning_rate": 0.0001240747657362174, + "loss": 1.2411, + "step": 9507 + }, + { + "epoch": 1.6928418803418803, + "grad_norm": 0.6779513359069824, + "learning_rate": 0.00012406117972550414, + "loss": 0.8886, + "step": 9508 + }, + { + "epoch": 1.693019943019943, + "grad_norm": 0.647588312625885, + "learning_rate": 0.00012404759324336203, + "loss": 1.107, + "step": 9509 + }, + { + "epoch": 1.6931980056980058, + "grad_norm": 0.7398989796638489, + "learning_rate": 0.00012403400629005726, + "loss": 1.0256, + "step": 9510 + }, + { + "epoch": 1.6933760683760684, + "grad_norm": 0.7572638392448425, + "learning_rate": 0.0001240204188658561, + "loss": 0.9662, + "step": 9511 + }, + { + "epoch": 1.693554131054131, + "grad_norm": 0.7044163346290588, + "learning_rate": 0.00012400683097102473, + "loss": 1.1388, + "step": 9512 + }, + { + "epoch": 1.6937321937321936, + "grad_norm": 0.7889094948768616, + "learning_rate": 0.00012399324260582936, + "loss": 1.0453, + "step": 9513 + }, + { + "epoch": 1.6939102564102564, + "grad_norm": 0.7977854609489441, + "learning_rate": 0.00012397965377053627, + "loss": 1.015, + "step": 9514 + }, + { + "epoch": 1.6940883190883191, + "grad_norm": 0.6223814487457275, + "learning_rate": 0.00012396606446541165, + "loss": 0.7985, + "step": 9515 + }, + { + "epoch": 1.694266381766382, + "grad_norm": 0.8307462334632874, + "learning_rate": 0.0001239524746907218, + "loss": 0.8899, + "step": 9516 + }, + { + "epoch": 1.6944444444444444, + "grad_norm": 0.7780544757843018, + "learning_rate": 0.00012393888444673295, + "loss": 0.9406, + "step": 9517 + }, + { + "epoch": 1.6946225071225072, + "grad_norm": 0.6894499659538269, + "learning_rate": 0.0001239252937337114, + "loss": 0.9412, + "step": 9518 + }, + { + "epoch": 1.6948005698005697, + "grad_norm": 0.7000680565834045, + "learning_rate": 0.00012391170255192342, + "loss": 1.0314, + "step": 9519 + }, + { + "epoch": 1.6949786324786325, + "grad_norm": 0.6772416830062866, + "learning_rate": 0.0001238981109016353, + "loss": 0.9153, + "step": 9520 + }, + { + "epoch": 1.6951566951566952, + "grad_norm": 0.7069609761238098, + "learning_rate": 0.00012388451878311333, + "loss": 1.1777, + "step": 9521 + }, + { + "epoch": 1.695334757834758, + "grad_norm": 0.6138432621955872, + "learning_rate": 0.00012387092619662386, + "loss": 0.8085, + "step": 9522 + }, + { + "epoch": 1.6955128205128205, + "grad_norm": 0.6122859716415405, + "learning_rate": 0.00012385733314243313, + "loss": 0.8534, + "step": 9523 + }, + { + "epoch": 1.6956908831908832, + "grad_norm": 0.7499903440475464, + "learning_rate": 0.00012384373962080755, + "loss": 0.9329, + "step": 9524 + }, + { + "epoch": 1.6958689458689458, + "grad_norm": 0.6413441896438599, + "learning_rate": 0.00012383014563201343, + "loss": 0.9609, + "step": 9525 + }, + { + "epoch": 1.6960470085470085, + "grad_norm": 0.7467969059944153, + "learning_rate": 0.0001238165511763171, + "loss": 0.9142, + "step": 9526 + }, + { + "epoch": 1.6962250712250713, + "grad_norm": 0.6540884375572205, + "learning_rate": 0.00012380295625398494, + "loss": 0.9503, + "step": 9527 + }, + { + "epoch": 1.696403133903134, + "grad_norm": 0.6298567652702332, + "learning_rate": 0.00012378936086528326, + "loss": 0.8853, + "step": 9528 + }, + { + "epoch": 1.6965811965811965, + "grad_norm": 0.8003417253494263, + "learning_rate": 0.00012377576501047845, + "loss": 0.969, + "step": 9529 + }, + { + "epoch": 1.6967592592592593, + "grad_norm": 0.8318493962287903, + "learning_rate": 0.00012376216868983697, + "loss": 1.1413, + "step": 9530 + }, + { + "epoch": 1.6969373219373218, + "grad_norm": 0.8294426202774048, + "learning_rate": 0.00012374857190362515, + "loss": 1.1885, + "step": 9531 + }, + { + "epoch": 1.6971153846153846, + "grad_norm": 0.7502955198287964, + "learning_rate": 0.0001237349746521094, + "loss": 1.233, + "step": 9532 + }, + { + "epoch": 1.6972934472934473, + "grad_norm": 0.6306588649749756, + "learning_rate": 0.00012372137693555612, + "loss": 1.2255, + "step": 9533 + }, + { + "epoch": 1.69747150997151, + "grad_norm": 0.7802746891975403, + "learning_rate": 0.0001237077787542317, + "loss": 1.2054, + "step": 9534 + }, + { + "epoch": 1.6976495726495726, + "grad_norm": 0.685114860534668, + "learning_rate": 0.00012369418010840265, + "loss": 0.9865, + "step": 9535 + }, + { + "epoch": 1.6978276353276354, + "grad_norm": 0.6656857132911682, + "learning_rate": 0.00012368058099833536, + "loss": 1.1579, + "step": 9536 + }, + { + "epoch": 1.698005698005698, + "grad_norm": 0.6596674919128418, + "learning_rate": 0.00012366698142429625, + "loss": 0.9104, + "step": 9537 + }, + { + "epoch": 1.6981837606837606, + "grad_norm": 0.6025584936141968, + "learning_rate": 0.00012365338138655183, + "loss": 1.117, + "step": 9538 + }, + { + "epoch": 1.6983618233618234, + "grad_norm": 0.671585202217102, + "learning_rate": 0.0001236397808853685, + "loss": 1.0271, + "step": 9539 + }, + { + "epoch": 1.6985398860398861, + "grad_norm": 0.7467984557151794, + "learning_rate": 0.0001236261799210128, + "loss": 1.0411, + "step": 9540 + }, + { + "epoch": 1.6987179487179487, + "grad_norm": 0.6251640915870667, + "learning_rate": 0.0001236125784937512, + "loss": 0.7154, + "step": 9541 + }, + { + "epoch": 1.6988960113960114, + "grad_norm": 0.7560956478118896, + "learning_rate": 0.00012359897660385016, + "loss": 1.0048, + "step": 9542 + }, + { + "epoch": 1.699074074074074, + "grad_norm": 0.6144903302192688, + "learning_rate": 0.00012358537425157618, + "loss": 1.1294, + "step": 9543 + }, + { + "epoch": 1.6992521367521367, + "grad_norm": 0.7839425206184387, + "learning_rate": 0.00012357177143719578, + "loss": 1.0725, + "step": 9544 + }, + { + "epoch": 1.6994301994301995, + "grad_norm": 0.6488651037216187, + "learning_rate": 0.00012355816816097553, + "loss": 0.9267, + "step": 9545 + }, + { + "epoch": 1.6996082621082622, + "grad_norm": 0.6848782896995544, + "learning_rate": 0.00012354456442318187, + "loss": 1.0426, + "step": 9546 + }, + { + "epoch": 1.6997863247863247, + "grad_norm": 0.7164611220359802, + "learning_rate": 0.0001235309602240814, + "loss": 0.8208, + "step": 9547 + }, + { + "epoch": 1.6999643874643875, + "grad_norm": 0.6725530624389648, + "learning_rate": 0.0001235173555639406, + "loss": 0.9366, + "step": 9548 + }, + { + "epoch": 1.70014245014245, + "grad_norm": 0.6958004236221313, + "learning_rate": 0.00012350375044302612, + "loss": 1.0185, + "step": 9549 + }, + { + "epoch": 1.7003205128205128, + "grad_norm": 0.8035947680473328, + "learning_rate": 0.00012349014486160445, + "loss": 1.065, + "step": 9550 + }, + { + "epoch": 1.7004985754985755, + "grad_norm": 0.6705633997917175, + "learning_rate": 0.00012347653881994222, + "loss": 0.8381, + "step": 9551 + }, + { + "epoch": 1.7006766381766383, + "grad_norm": 0.6652300357818604, + "learning_rate": 0.00012346293231830596, + "loss": 1.1428, + "step": 9552 + }, + { + "epoch": 1.7008547008547008, + "grad_norm": 0.6719335913658142, + "learning_rate": 0.0001234493253569623, + "loss": 1.0138, + "step": 9553 + }, + { + "epoch": 1.7010327635327636, + "grad_norm": 0.746981680393219, + "learning_rate": 0.0001234357179361778, + "loss": 1.1169, + "step": 9554 + }, + { + "epoch": 1.701210826210826, + "grad_norm": 0.6768170595169067, + "learning_rate": 0.0001234221100562191, + "loss": 0.9065, + "step": 9555 + }, + { + "epoch": 1.7013888888888888, + "grad_norm": 0.7127171754837036, + "learning_rate": 0.00012340850171735278, + "loss": 0.9467, + "step": 9556 + }, + { + "epoch": 1.7015669515669516, + "grad_norm": 0.6802694797515869, + "learning_rate": 0.00012339489291984554, + "loss": 0.8938, + "step": 9557 + }, + { + "epoch": 1.7017450142450143, + "grad_norm": 0.7101455926895142, + "learning_rate": 0.00012338128366396394, + "loss": 1.1939, + "step": 9558 + }, + { + "epoch": 1.7019230769230769, + "grad_norm": 0.621223509311676, + "learning_rate": 0.00012336767394997467, + "loss": 0.7583, + "step": 9559 + }, + { + "epoch": 1.7021011396011396, + "grad_norm": 0.7130763530731201, + "learning_rate": 0.00012335406377814439, + "loss": 0.8684, + "step": 9560 + }, + { + "epoch": 1.7022792022792022, + "grad_norm": 0.6761086583137512, + "learning_rate": 0.00012334045314873972, + "loss": 1.0197, + "step": 9561 + }, + { + "epoch": 1.702457264957265, + "grad_norm": 0.7030459642410278, + "learning_rate": 0.00012332684206202736, + "loss": 0.8627, + "step": 9562 + }, + { + "epoch": 1.7026353276353277, + "grad_norm": 0.6278037428855896, + "learning_rate": 0.000123313230518274, + "loss": 0.8953, + "step": 9563 + }, + { + "epoch": 1.7028133903133904, + "grad_norm": 0.6450623869895935, + "learning_rate": 0.00012329961851774627, + "loss": 0.8826, + "step": 9564 + }, + { + "epoch": 1.702991452991453, + "grad_norm": 0.7324244976043701, + "learning_rate": 0.00012328600606071097, + "loss": 1.0133, + "step": 9565 + }, + { + "epoch": 1.7031695156695157, + "grad_norm": 0.6560033559799194, + "learning_rate": 0.00012327239314743473, + "loss": 0.9601, + "step": 9566 + }, + { + "epoch": 1.7033475783475782, + "grad_norm": 0.6693514585494995, + "learning_rate": 0.0001232587797781843, + "loss": 0.9447, + "step": 9567 + }, + { + "epoch": 1.703525641025641, + "grad_norm": 0.6403199434280396, + "learning_rate": 0.00012324516595322638, + "loss": 0.8554, + "step": 9568 + }, + { + "epoch": 1.7037037037037037, + "grad_norm": 0.8290280103683472, + "learning_rate": 0.00012323155167282774, + "loss": 1.1877, + "step": 9569 + }, + { + "epoch": 1.7038817663817665, + "grad_norm": 0.7207778692245483, + "learning_rate": 0.00012321793693725509, + "loss": 1.0978, + "step": 9570 + }, + { + "epoch": 1.7040598290598292, + "grad_norm": 0.8794265985488892, + "learning_rate": 0.00012320432174677519, + "loss": 0.9387, + "step": 9571 + }, + { + "epoch": 1.7042378917378918, + "grad_norm": 0.6683359146118164, + "learning_rate": 0.00012319070610165484, + "loss": 0.9227, + "step": 9572 + }, + { + "epoch": 1.7044159544159543, + "grad_norm": 0.7342001795768738, + "learning_rate": 0.00012317709000216076, + "loss": 0.9453, + "step": 9573 + }, + { + "epoch": 1.704594017094017, + "grad_norm": 0.6315770149230957, + "learning_rate": 0.00012316347344855973, + "loss": 0.8263, + "step": 9574 + }, + { + "epoch": 1.7047720797720798, + "grad_norm": 0.7697155475616455, + "learning_rate": 0.00012314985644111857, + "loss": 1.0238, + "step": 9575 + }, + { + "epoch": 1.7049501424501425, + "grad_norm": 0.6674068570137024, + "learning_rate": 0.00012313623898010408, + "loss": 1.0823, + "step": 9576 + }, + { + "epoch": 1.7051282051282053, + "grad_norm": 0.6995484232902527, + "learning_rate": 0.00012312262106578304, + "loss": 1.2001, + "step": 9577 + }, + { + "epoch": 1.7053062678062678, + "grad_norm": 0.7639257907867432, + "learning_rate": 0.00012310900269842226, + "loss": 1.3438, + "step": 9578 + }, + { + "epoch": 1.7054843304843303, + "grad_norm": 0.6486390233039856, + "learning_rate": 0.00012309538387828857, + "loss": 0.9924, + "step": 9579 + }, + { + "epoch": 1.705662393162393, + "grad_norm": 0.6737813949584961, + "learning_rate": 0.00012308176460564885, + "loss": 0.8722, + "step": 9580 + }, + { + "epoch": 1.7058404558404558, + "grad_norm": 0.6462090611457825, + "learning_rate": 0.00012306814488076987, + "loss": 1.1013, + "step": 9581 + }, + { + "epoch": 1.7060185185185186, + "grad_norm": 0.7887832522392273, + "learning_rate": 0.00012305452470391852, + "loss": 0.9998, + "step": 9582 + }, + { + "epoch": 1.7061965811965814, + "grad_norm": 0.6345070004463196, + "learning_rate": 0.00012304090407536165, + "loss": 1.0305, + "step": 9583 + }, + { + "epoch": 1.7063746438746439, + "grad_norm": 0.6398460268974304, + "learning_rate": 0.0001230272829953661, + "loss": 1.2243, + "step": 9584 + }, + { + "epoch": 1.7065527065527064, + "grad_norm": 0.6501944065093994, + "learning_rate": 0.00012301366146419879, + "loss": 0.9425, + "step": 9585 + }, + { + "epoch": 1.7067307692307692, + "grad_norm": 0.6406761407852173, + "learning_rate": 0.00012300003948212661, + "loss": 0.948, + "step": 9586 + }, + { + "epoch": 1.706908831908832, + "grad_norm": 0.7114266157150269, + "learning_rate": 0.00012298641704941644, + "loss": 1.1291, + "step": 9587 + }, + { + "epoch": 1.7070868945868947, + "grad_norm": 0.6653099656105042, + "learning_rate": 0.00012297279416633515, + "loss": 1.0156, + "step": 9588 + }, + { + "epoch": 1.7072649572649574, + "grad_norm": 0.5970917344093323, + "learning_rate": 0.0001229591708331497, + "loss": 0.9424, + "step": 9589 + }, + { + "epoch": 1.70744301994302, + "grad_norm": 0.6861461400985718, + "learning_rate": 0.00012294554705012694, + "loss": 0.7581, + "step": 9590 + }, + { + "epoch": 1.7076210826210825, + "grad_norm": 0.6930568218231201, + "learning_rate": 0.00012293192281753393, + "loss": 1.0544, + "step": 9591 + }, + { + "epoch": 1.7077991452991452, + "grad_norm": 0.7420656085014343, + "learning_rate": 0.00012291829813563748, + "loss": 0.7092, + "step": 9592 + }, + { + "epoch": 1.707977207977208, + "grad_norm": 0.6607801914215088, + "learning_rate": 0.0001229046730047046, + "loss": 0.5544, + "step": 9593 + }, + { + "epoch": 1.7081552706552707, + "grad_norm": 0.8419139385223389, + "learning_rate": 0.00012289104742500224, + "loss": 1.0443, + "step": 9594 + }, + { + "epoch": 1.7083333333333335, + "grad_norm": 0.6774617433547974, + "learning_rate": 0.00012287742139679734, + "loss": 1.0098, + "step": 9595 + }, + { + "epoch": 1.708511396011396, + "grad_norm": 0.7517698407173157, + "learning_rate": 0.0001228637949203569, + "loss": 1.1145, + "step": 9596 + }, + { + "epoch": 1.7086894586894585, + "grad_norm": 0.6048635840415955, + "learning_rate": 0.00012285016799594791, + "loss": 0.7398, + "step": 9597 + }, + { + "epoch": 1.7088675213675213, + "grad_norm": 0.8054425716400146, + "learning_rate": 0.00012283654062383734, + "loss": 1.0893, + "step": 9598 + }, + { + "epoch": 1.709045584045584, + "grad_norm": 0.8694897294044495, + "learning_rate": 0.0001228229128042922, + "loss": 1.2366, + "step": 9599 + }, + { + "epoch": 1.7092236467236468, + "grad_norm": 0.7460638880729675, + "learning_rate": 0.00012280928453757946, + "loss": 1.1753, + "step": 9600 + }, + { + "epoch": 1.7094017094017095, + "grad_norm": 0.6714958548545837, + "learning_rate": 0.00012279565582396618, + "loss": 1.0473, + "step": 9601 + }, + { + "epoch": 1.709579772079772, + "grad_norm": 0.6893340945243835, + "learning_rate": 0.00012278202666371937, + "loss": 1.2761, + "step": 9602 + }, + { + "epoch": 1.7097578347578346, + "grad_norm": 0.6816153526306152, + "learning_rate": 0.00012276839705710612, + "loss": 0.991, + "step": 9603 + }, + { + "epoch": 1.7099358974358974, + "grad_norm": 0.6961633563041687, + "learning_rate": 0.0001227547670043934, + "loss": 1.0634, + "step": 9604 + }, + { + "epoch": 1.71011396011396, + "grad_norm": 0.643734872341156, + "learning_rate": 0.0001227411365058483, + "loss": 0.8672, + "step": 9605 + }, + { + "epoch": 1.7102920227920229, + "grad_norm": 0.7313315272331238, + "learning_rate": 0.00012272750556173784, + "loss": 1.1152, + "step": 9606 + }, + { + "epoch": 1.7104700854700856, + "grad_norm": 0.6464954614639282, + "learning_rate": 0.00012271387417232916, + "loss": 0.8798, + "step": 9607 + }, + { + "epoch": 1.7106481481481481, + "grad_norm": 0.8365204334259033, + "learning_rate": 0.00012270024233788929, + "loss": 1.213, + "step": 9608 + }, + { + "epoch": 1.7108262108262107, + "grad_norm": 0.6460705995559692, + "learning_rate": 0.0001226866100586853, + "loss": 0.9232, + "step": 9609 + }, + { + "epoch": 1.7110042735042734, + "grad_norm": 0.6446022987365723, + "learning_rate": 0.00012267297733498434, + "loss": 0.8295, + "step": 9610 + }, + { + "epoch": 1.7111823361823362, + "grad_norm": 0.7692012190818787, + "learning_rate": 0.00012265934416705345, + "loss": 1.0715, + "step": 9611 + }, + { + "epoch": 1.711360398860399, + "grad_norm": 0.671154260635376, + "learning_rate": 0.0001226457105551598, + "loss": 0.9752, + "step": 9612 + }, + { + "epoch": 1.7115384615384617, + "grad_norm": 0.6525935530662537, + "learning_rate": 0.00012263207649957053, + "loss": 1.09, + "step": 9613 + }, + { + "epoch": 1.7117165242165242, + "grad_norm": 0.6984749436378479, + "learning_rate": 0.0001226184420005527, + "loss": 0.9956, + "step": 9614 + }, + { + "epoch": 1.7118945868945867, + "grad_norm": 0.6769809126853943, + "learning_rate": 0.0001226048070583735, + "loss": 1.0151, + "step": 9615 + }, + { + "epoch": 1.7120726495726495, + "grad_norm": 0.6085978746414185, + "learning_rate": 0.00012259117167330005, + "loss": 0.8706, + "step": 9616 + }, + { + "epoch": 1.7122507122507122, + "grad_norm": 0.7335749268531799, + "learning_rate": 0.00012257753584559952, + "loss": 1.0575, + "step": 9617 + }, + { + "epoch": 1.712428774928775, + "grad_norm": 0.7392038106918335, + "learning_rate": 0.0001225638995755391, + "loss": 0.8763, + "step": 9618 + }, + { + "epoch": 1.7126068376068377, + "grad_norm": 0.6708608865737915, + "learning_rate": 0.00012255026286338592, + "loss": 1.131, + "step": 9619 + }, + { + "epoch": 1.7127849002849003, + "grad_norm": 0.726657509803772, + "learning_rate": 0.0001225366257094072, + "loss": 1.0569, + "step": 9620 + }, + { + "epoch": 1.7129629629629628, + "grad_norm": 0.749098002910614, + "learning_rate": 0.0001225229881138701, + "loss": 0.9196, + "step": 9621 + }, + { + "epoch": 1.7131410256410255, + "grad_norm": 0.6550580263137817, + "learning_rate": 0.00012250935007704182, + "loss": 1.0244, + "step": 9622 + }, + { + "epoch": 1.7133190883190883, + "grad_norm": 0.7714282274246216, + "learning_rate": 0.00012249571159918962, + "loss": 1.1025, + "step": 9623 + }, + { + "epoch": 1.713497150997151, + "grad_norm": 0.7869850397109985, + "learning_rate": 0.00012248207268058064, + "loss": 0.9238, + "step": 9624 + }, + { + "epoch": 1.7136752136752138, + "grad_norm": 0.7187856435775757, + "learning_rate": 0.00012246843332148216, + "loss": 1.081, + "step": 9625 + }, + { + "epoch": 1.7138532763532763, + "grad_norm": 0.6634210348129272, + "learning_rate": 0.00012245479352216142, + "loss": 1.1944, + "step": 9626 + }, + { + "epoch": 1.714031339031339, + "grad_norm": 0.6609212160110474, + "learning_rate": 0.00012244115328288567, + "loss": 0.9613, + "step": 9627 + }, + { + "epoch": 1.7142094017094016, + "grad_norm": 0.7906867861747742, + "learning_rate": 0.0001224275126039221, + "loss": 1.2692, + "step": 9628 + }, + { + "epoch": 1.7143874643874644, + "grad_norm": 0.8037096858024597, + "learning_rate": 0.000122413871485538, + "loss": 0.9823, + "step": 9629 + }, + { + "epoch": 1.7145655270655271, + "grad_norm": 0.7740145921707153, + "learning_rate": 0.00012240022992800068, + "loss": 1.1937, + "step": 9630 + }, + { + "epoch": 1.7147435897435899, + "grad_norm": 0.595372200012207, + "learning_rate": 0.00012238658793157738, + "loss": 0.9153, + "step": 9631 + }, + { + "epoch": 1.7149216524216524, + "grad_norm": 0.6671900749206543, + "learning_rate": 0.0001223729454965354, + "loss": 1.0895, + "step": 9632 + }, + { + "epoch": 1.7150997150997151, + "grad_norm": 0.5805774927139282, + "learning_rate": 0.000122359302623142, + "loss": 1.0001, + "step": 9633 + }, + { + "epoch": 1.7152777777777777, + "grad_norm": 0.8851602673530579, + "learning_rate": 0.00012234565931166456, + "loss": 1.2828, + "step": 9634 + }, + { + "epoch": 1.7154558404558404, + "grad_norm": 0.6960011720657349, + "learning_rate": 0.0001223320155623703, + "loss": 1.0622, + "step": 9635 + }, + { + "epoch": 1.7156339031339032, + "grad_norm": 0.5587009191513062, + "learning_rate": 0.0001223183713755266, + "loss": 0.83, + "step": 9636 + }, + { + "epoch": 1.715811965811966, + "grad_norm": 0.6892730593681335, + "learning_rate": 0.00012230472675140076, + "loss": 0.9214, + "step": 9637 + }, + { + "epoch": 1.7159900284900285, + "grad_norm": 0.6545090079307556, + "learning_rate": 0.00012229108169026017, + "loss": 0.829, + "step": 9638 + }, + { + "epoch": 1.7161680911680912, + "grad_norm": 0.6539101600646973, + "learning_rate": 0.00012227743619237213, + "loss": 1.0686, + "step": 9639 + }, + { + "epoch": 1.7163461538461537, + "grad_norm": 0.5887274146080017, + "learning_rate": 0.000122263790258004, + "loss": 0.9285, + "step": 9640 + }, + { + "epoch": 1.7165242165242165, + "grad_norm": 0.6328918933868408, + "learning_rate": 0.00012225014388742313, + "loss": 0.9684, + "step": 9641 + }, + { + "epoch": 1.7167022792022792, + "grad_norm": 0.6377436518669128, + "learning_rate": 0.00012223649708089694, + "loss": 0.9425, + "step": 9642 + }, + { + "epoch": 1.716880341880342, + "grad_norm": 0.6967392563819885, + "learning_rate": 0.00012222284983869275, + "loss": 0.9342, + "step": 9643 + }, + { + "epoch": 1.7170584045584045, + "grad_norm": 0.7051317691802979, + "learning_rate": 0.00012220920216107802, + "loss": 1.1843, + "step": 9644 + }, + { + "epoch": 1.7172364672364673, + "grad_norm": 0.6864503622055054, + "learning_rate": 0.00012219555404832007, + "loss": 1.0371, + "step": 9645 + }, + { + "epoch": 1.7174145299145298, + "grad_norm": 0.583454430103302, + "learning_rate": 0.00012218190550068638, + "loss": 0.6774, + "step": 9646 + }, + { + "epoch": 1.7175925925925926, + "grad_norm": 0.6755677461624146, + "learning_rate": 0.0001221682565184443, + "loss": 0.9517, + "step": 9647 + }, + { + "epoch": 1.7177706552706553, + "grad_norm": 0.7230031490325928, + "learning_rate": 0.0001221546071018613, + "loss": 1.0385, + "step": 9648 + }, + { + "epoch": 1.717948717948718, + "grad_norm": 0.7381200194358826, + "learning_rate": 0.0001221409572512048, + "loss": 0.9893, + "step": 9649 + }, + { + "epoch": 1.7181267806267806, + "grad_norm": 0.7079094648361206, + "learning_rate": 0.0001221273069667422, + "loss": 0.7793, + "step": 9650 + }, + { + "epoch": 1.7183048433048433, + "grad_norm": 0.6666881442070007, + "learning_rate": 0.00012211365624874106, + "loss": 0.9752, + "step": 9651 + }, + { + "epoch": 1.7184829059829059, + "grad_norm": 0.6196922659873962, + "learning_rate": 0.00012210000509746868, + "loss": 0.922, + "step": 9652 + }, + { + "epoch": 1.7186609686609686, + "grad_norm": 0.657879650592804, + "learning_rate": 0.00012208635351319266, + "loss": 1.2583, + "step": 9653 + }, + { + "epoch": 1.7188390313390314, + "grad_norm": 0.7240459322929382, + "learning_rate": 0.00012207270149618043, + "loss": 0.8479, + "step": 9654 + }, + { + "epoch": 1.7190170940170941, + "grad_norm": 0.8293825387954712, + "learning_rate": 0.00012205904904669945, + "loss": 0.9092, + "step": 9655 + }, + { + "epoch": 1.7191951566951567, + "grad_norm": 0.6907553672790527, + "learning_rate": 0.0001220453961650172, + "loss": 1.0543, + "step": 9656 + }, + { + "epoch": 1.7193732193732194, + "grad_norm": 0.7178300023078918, + "learning_rate": 0.00012203174285140124, + "loss": 0.9147, + "step": 9657 + }, + { + "epoch": 1.719551282051282, + "grad_norm": 0.7037166357040405, + "learning_rate": 0.00012201808910611905, + "loss": 0.8685, + "step": 9658 + }, + { + "epoch": 1.7197293447293447, + "grad_norm": 0.5850751996040344, + "learning_rate": 0.00012200443492943813, + "loss": 0.72, + "step": 9659 + }, + { + "epoch": 1.7199074074074074, + "grad_norm": 0.744239330291748, + "learning_rate": 0.00012199078032162603, + "loss": 0.9717, + "step": 9660 + }, + { + "epoch": 1.7200854700854702, + "grad_norm": 0.6509126424789429, + "learning_rate": 0.00012197712528295025, + "loss": 0.9768, + "step": 9661 + }, + { + "epoch": 1.7202635327635327, + "grad_norm": 0.623220682144165, + "learning_rate": 0.00012196346981367837, + "loss": 0.9824, + "step": 9662 + }, + { + "epoch": 1.7204415954415955, + "grad_norm": 0.6376451849937439, + "learning_rate": 0.00012194981391407792, + "loss": 0.8228, + "step": 9663 + }, + { + "epoch": 1.720619658119658, + "grad_norm": 0.794830322265625, + "learning_rate": 0.00012193615758441648, + "loss": 0.9168, + "step": 9664 + }, + { + "epoch": 1.7207977207977208, + "grad_norm": 0.7812975645065308, + "learning_rate": 0.0001219225008249616, + "loss": 0.8625, + "step": 9665 + }, + { + "epoch": 1.7209757834757835, + "grad_norm": 0.6843218207359314, + "learning_rate": 0.0001219088436359808, + "loss": 1.0176, + "step": 9666 + }, + { + "epoch": 1.7211538461538463, + "grad_norm": 0.6924905180931091, + "learning_rate": 0.00012189518601774178, + "loss": 0.855, + "step": 9667 + }, + { + "epoch": 1.7213319088319088, + "grad_norm": 0.6348826289176941, + "learning_rate": 0.00012188152797051202, + "loss": 1.1596, + "step": 9668 + }, + { + "epoch": 1.7215099715099715, + "grad_norm": 0.7170482873916626, + "learning_rate": 0.00012186786949455922, + "loss": 0.9811, + "step": 9669 + }, + { + "epoch": 1.721688034188034, + "grad_norm": 0.7471763491630554, + "learning_rate": 0.00012185421059015094, + "loss": 1.0925, + "step": 9670 + }, + { + "epoch": 1.7218660968660968, + "grad_norm": 0.6771119236946106, + "learning_rate": 0.00012184055125755481, + "loss": 0.9403, + "step": 9671 + }, + { + "epoch": 1.7220441595441596, + "grad_norm": 0.4335343539714813, + "learning_rate": 0.0001218268914970384, + "loss": 0.4925, + "step": 9672 + }, + { + "epoch": 1.7222222222222223, + "grad_norm": 0.6652585864067078, + "learning_rate": 0.00012181323130886943, + "loss": 0.7684, + "step": 9673 + }, + { + "epoch": 1.7224002849002849, + "grad_norm": 0.6465467810630798, + "learning_rate": 0.00012179957069331548, + "loss": 0.9011, + "step": 9674 + }, + { + "epoch": 1.7225783475783476, + "grad_norm": 0.6725688576698303, + "learning_rate": 0.00012178590965064427, + "loss": 0.9563, + "step": 9675 + }, + { + "epoch": 1.7227564102564101, + "grad_norm": 0.6223418712615967, + "learning_rate": 0.00012177224818112341, + "loss": 0.9099, + "step": 9676 + }, + { + "epoch": 1.7229344729344729, + "grad_norm": 0.79325270652771, + "learning_rate": 0.00012175858628502053, + "loss": 1.0318, + "step": 9677 + }, + { + "epoch": 1.7231125356125356, + "grad_norm": 0.6735602617263794, + "learning_rate": 0.0001217449239626034, + "loss": 1.0797, + "step": 9678 + }, + { + "epoch": 1.7232905982905984, + "grad_norm": 0.7082492113113403, + "learning_rate": 0.00012173126121413962, + "loss": 1.1341, + "step": 9679 + }, + { + "epoch": 1.723468660968661, + "grad_norm": 0.6563859581947327, + "learning_rate": 0.00012171759803989696, + "loss": 0.8778, + "step": 9680 + }, + { + "epoch": 1.7236467236467237, + "grad_norm": 0.6867792010307312, + "learning_rate": 0.00012170393444014306, + "loss": 0.8301, + "step": 9681 + }, + { + "epoch": 1.7238247863247862, + "grad_norm": 0.7870511412620544, + "learning_rate": 0.00012169027041514562, + "loss": 0.9165, + "step": 9682 + }, + { + "epoch": 1.724002849002849, + "grad_norm": 0.8006493449211121, + "learning_rate": 0.00012167660596517241, + "loss": 1.0395, + "step": 9683 + }, + { + "epoch": 1.7241809116809117, + "grad_norm": 0.6936125159263611, + "learning_rate": 0.00012166294109049114, + "loss": 1.1037, + "step": 9684 + }, + { + "epoch": 1.7243589743589745, + "grad_norm": 0.8176514506340027, + "learning_rate": 0.00012164927579136956, + "loss": 0.8791, + "step": 9685 + }, + { + "epoch": 1.7245370370370372, + "grad_norm": 0.6948300004005432, + "learning_rate": 0.00012163561006807537, + "loss": 0.9292, + "step": 9686 + }, + { + "epoch": 1.7247150997150997, + "grad_norm": 0.6237453818321228, + "learning_rate": 0.00012162194392087634, + "loss": 0.8553, + "step": 9687 + }, + { + "epoch": 1.7248931623931623, + "grad_norm": 0.6198007464408875, + "learning_rate": 0.00012160827735004021, + "loss": 0.9599, + "step": 9688 + }, + { + "epoch": 1.725071225071225, + "grad_norm": 0.639838695526123, + "learning_rate": 0.00012159461035583482, + "loss": 0.9328, + "step": 9689 + }, + { + "epoch": 1.7252492877492878, + "grad_norm": 0.7264436483383179, + "learning_rate": 0.00012158094293852789, + "loss": 1.0247, + "step": 9690 + }, + { + "epoch": 1.7254273504273505, + "grad_norm": 0.6320534348487854, + "learning_rate": 0.00012156727509838721, + "loss": 1.1222, + "step": 9691 + }, + { + "epoch": 1.7256054131054133, + "grad_norm": 0.6204122304916382, + "learning_rate": 0.00012155360683568056, + "loss": 0.9765, + "step": 9692 + }, + { + "epoch": 1.7257834757834758, + "grad_norm": 0.7026457190513611, + "learning_rate": 0.00012153993815067579, + "loss": 1.0178, + "step": 9693 + }, + { + "epoch": 1.7259615384615383, + "grad_norm": 0.6471006870269775, + "learning_rate": 0.00012152626904364067, + "loss": 1.0035, + "step": 9694 + }, + { + "epoch": 1.726139601139601, + "grad_norm": 0.6875706911087036, + "learning_rate": 0.00012151259951484301, + "loss": 0.7921, + "step": 9695 + }, + { + "epoch": 1.7263176638176638, + "grad_norm": 0.6963251233100891, + "learning_rate": 0.00012149892956455067, + "loss": 0.9677, + "step": 9696 + }, + { + "epoch": 1.7264957264957266, + "grad_norm": 0.9077282547950745, + "learning_rate": 0.00012148525919303142, + "loss": 0.9362, + "step": 9697 + }, + { + "epoch": 1.7266737891737893, + "grad_norm": 0.7347434163093567, + "learning_rate": 0.00012147158840055319, + "loss": 0.8712, + "step": 9698 + }, + { + "epoch": 1.7268518518518519, + "grad_norm": 0.7206630110740662, + "learning_rate": 0.00012145791718738377, + "loss": 1.032, + "step": 9699 + }, + { + "epoch": 1.7270299145299144, + "grad_norm": 0.7174662947654724, + "learning_rate": 0.00012144424555379106, + "loss": 0.954, + "step": 9700 + }, + { + "epoch": 1.7272079772079771, + "grad_norm": 0.7442345023155212, + "learning_rate": 0.0001214305735000429, + "loss": 1.0709, + "step": 9701 + }, + { + "epoch": 1.72738603988604, + "grad_norm": 0.6154376268386841, + "learning_rate": 0.00012141690102640715, + "loss": 0.9365, + "step": 9702 + }, + { + "epoch": 1.7275641025641026, + "grad_norm": 0.6213796734809875, + "learning_rate": 0.00012140322813315172, + "loss": 0.8337, + "step": 9703 + }, + { + "epoch": 1.7277421652421654, + "grad_norm": 0.7682011127471924, + "learning_rate": 0.0001213895548205445, + "loss": 1.1579, + "step": 9704 + }, + { + "epoch": 1.727920227920228, + "grad_norm": 0.6796970963478088, + "learning_rate": 0.0001213758810888534, + "loss": 0.8875, + "step": 9705 + }, + { + "epoch": 1.7280982905982905, + "grad_norm": 0.7203732132911682, + "learning_rate": 0.0001213622069383463, + "loss": 0.7827, + "step": 9706 + }, + { + "epoch": 1.7282763532763532, + "grad_norm": 0.6151877045631409, + "learning_rate": 0.00012134853236929111, + "loss": 1.0282, + "step": 9707 + }, + { + "epoch": 1.728454415954416, + "grad_norm": 0.6665124297142029, + "learning_rate": 0.0001213348573819558, + "loss": 1.0636, + "step": 9708 + }, + { + "epoch": 1.7286324786324787, + "grad_norm": 0.7334614396095276, + "learning_rate": 0.00012132118197660829, + "loss": 1.0889, + "step": 9709 + }, + { + "epoch": 1.7288105413105415, + "grad_norm": 0.7267759442329407, + "learning_rate": 0.00012130750615351649, + "loss": 1.096, + "step": 9710 + }, + { + "epoch": 1.728988603988604, + "grad_norm": 0.6542944312095642, + "learning_rate": 0.00012129382991294837, + "loss": 1.0855, + "step": 9711 + }, + { + "epoch": 1.7291666666666665, + "grad_norm": 0.694523274898529, + "learning_rate": 0.00012128015325517193, + "loss": 0.8482, + "step": 9712 + }, + { + "epoch": 1.7293447293447293, + "grad_norm": 0.7879082560539246, + "learning_rate": 0.00012126647618045504, + "loss": 1.2356, + "step": 9713 + }, + { + "epoch": 1.729522792022792, + "grad_norm": 0.7108420729637146, + "learning_rate": 0.00012125279868906574, + "loss": 1.0185, + "step": 9714 + }, + { + "epoch": 1.7297008547008548, + "grad_norm": 0.6928725838661194, + "learning_rate": 0.000121239120781272, + "loss": 1.1507, + "step": 9715 + }, + { + "epoch": 1.7298789173789175, + "grad_norm": 0.6195241212844849, + "learning_rate": 0.00012122544245734182, + "loss": 0.8656, + "step": 9716 + }, + { + "epoch": 1.73005698005698, + "grad_norm": 0.5962017774581909, + "learning_rate": 0.00012121176371754317, + "loss": 0.918, + "step": 9717 + }, + { + "epoch": 1.7302350427350426, + "grad_norm": 0.7409394979476929, + "learning_rate": 0.00012119808456214407, + "loss": 1.0283, + "step": 9718 + }, + { + "epoch": 1.7304131054131053, + "grad_norm": 0.6571973562240601, + "learning_rate": 0.00012118440499141257, + "loss": 1.1015, + "step": 9719 + }, + { + "epoch": 1.730591168091168, + "grad_norm": 0.681394100189209, + "learning_rate": 0.00012117072500561664, + "loss": 0.8247, + "step": 9720 + }, + { + "epoch": 1.7307692307692308, + "grad_norm": 0.7278251647949219, + "learning_rate": 0.00012115704460502432, + "loss": 1.0693, + "step": 9721 + }, + { + "epoch": 1.7309472934472936, + "grad_norm": 0.6569405794143677, + "learning_rate": 0.0001211433637899037, + "loss": 0.8992, + "step": 9722 + }, + { + "epoch": 1.7311253561253561, + "grad_norm": 0.6305136680603027, + "learning_rate": 0.00012112968256052272, + "loss": 0.8543, + "step": 9723 + }, + { + "epoch": 1.7313034188034186, + "grad_norm": 0.6111339330673218, + "learning_rate": 0.00012111600091714956, + "loss": 0.991, + "step": 9724 + }, + { + "epoch": 1.7314814814814814, + "grad_norm": 0.646973192691803, + "learning_rate": 0.00012110231886005223, + "loss": 0.8855, + "step": 9725 + }, + { + "epoch": 1.7316595441595442, + "grad_norm": 0.7054407000541687, + "learning_rate": 0.00012108863638949879, + "loss": 1.0816, + "step": 9726 + }, + { + "epoch": 1.731837606837607, + "grad_norm": 0.6592162847518921, + "learning_rate": 0.00012107495350575729, + "loss": 1.0961, + "step": 9727 + }, + { + "epoch": 1.7320156695156697, + "grad_norm": 0.6615595817565918, + "learning_rate": 0.00012106127020909587, + "loss": 0.9669, + "step": 9728 + }, + { + "epoch": 1.7321937321937322, + "grad_norm": 0.9030881524085999, + "learning_rate": 0.00012104758649978263, + "loss": 0.9438, + "step": 9729 + }, + { + "epoch": 1.7323717948717947, + "grad_norm": 0.6776516437530518, + "learning_rate": 0.00012103390237808566, + "loss": 0.8967, + "step": 9730 + }, + { + "epoch": 1.7325498575498575, + "grad_norm": 0.6010605096817017, + "learning_rate": 0.00012102021784427306, + "loss": 0.8893, + "step": 9731 + }, + { + "epoch": 1.7327279202279202, + "grad_norm": 0.6540384292602539, + "learning_rate": 0.00012100653289861295, + "loss": 0.9328, + "step": 9732 + }, + { + "epoch": 1.732905982905983, + "grad_norm": 0.6836950182914734, + "learning_rate": 0.00012099284754137345, + "loss": 0.9019, + "step": 9733 + }, + { + "epoch": 1.7330840455840457, + "grad_norm": 0.7597874402999878, + "learning_rate": 0.00012097916177282274, + "loss": 1.0093, + "step": 9734 + }, + { + "epoch": 1.7332621082621082, + "grad_norm": 0.7686513066291809, + "learning_rate": 0.00012096547559322892, + "loss": 0.8685, + "step": 9735 + }, + { + "epoch": 1.7334401709401708, + "grad_norm": 0.613777220249176, + "learning_rate": 0.0001209517890028602, + "loss": 0.8317, + "step": 9736 + }, + { + "epoch": 1.7336182336182335, + "grad_norm": 0.6788455843925476, + "learning_rate": 0.00012093810200198466, + "loss": 0.866, + "step": 9737 + }, + { + "epoch": 1.7337962962962963, + "grad_norm": 0.616801381111145, + "learning_rate": 0.00012092441459087047, + "loss": 0.8299, + "step": 9738 + }, + { + "epoch": 1.733974358974359, + "grad_norm": 0.731987476348877, + "learning_rate": 0.00012091072676978589, + "loss": 1.089, + "step": 9739 + }, + { + "epoch": 1.7341524216524218, + "grad_norm": 0.7042871117591858, + "learning_rate": 0.00012089703853899905, + "loss": 0.8667, + "step": 9740 + }, + { + "epoch": 1.7343304843304843, + "grad_norm": 0.62722247838974, + "learning_rate": 0.00012088334989877817, + "loss": 0.9185, + "step": 9741 + }, + { + "epoch": 1.734508547008547, + "grad_norm": 0.6354684829711914, + "learning_rate": 0.0001208696608493914, + "loss": 0.9951, + "step": 9742 + }, + { + "epoch": 1.7346866096866096, + "grad_norm": 0.658647894859314, + "learning_rate": 0.00012085597139110698, + "loss": 0.9324, + "step": 9743 + }, + { + "epoch": 1.7348646723646723, + "grad_norm": 0.84359210729599, + "learning_rate": 0.00012084228152419312, + "loss": 1.0861, + "step": 9744 + }, + { + "epoch": 1.735042735042735, + "grad_norm": 0.6293938755989075, + "learning_rate": 0.00012082859124891807, + "loss": 0.9676, + "step": 9745 + }, + { + "epoch": 1.7352207977207978, + "grad_norm": 0.6398760676383972, + "learning_rate": 0.00012081490056555004, + "loss": 0.8502, + "step": 9746 + }, + { + "epoch": 1.7353988603988604, + "grad_norm": 0.6918041706085205, + "learning_rate": 0.00012080120947435726, + "loss": 1.0081, + "step": 9747 + }, + { + "epoch": 1.7355769230769231, + "grad_norm": 0.7374079823493958, + "learning_rate": 0.00012078751797560798, + "loss": 0.9485, + "step": 9748 + }, + { + "epoch": 1.7357549857549857, + "grad_norm": 0.7392128705978394, + "learning_rate": 0.00012077382606957049, + "loss": 0.9283, + "step": 9749 + }, + { + "epoch": 1.7359330484330484, + "grad_norm": 0.701320230960846, + "learning_rate": 0.00012076013375651303, + "loss": 1.0339, + "step": 9750 + }, + { + "epoch": 1.7361111111111112, + "grad_norm": 0.6316696405410767, + "learning_rate": 0.00012074644103670387, + "loss": 0.9097, + "step": 9751 + }, + { + "epoch": 1.736289173789174, + "grad_norm": 0.6892024278640747, + "learning_rate": 0.00012073274791041132, + "loss": 1.0863, + "step": 9752 + }, + { + "epoch": 1.7364672364672364, + "grad_norm": 0.6032847762107849, + "learning_rate": 0.00012071905437790361, + "loss": 0.9305, + "step": 9753 + }, + { + "epoch": 1.7366452991452992, + "grad_norm": 0.6659184098243713, + "learning_rate": 0.00012070536043944907, + "loss": 0.9793, + "step": 9754 + }, + { + "epoch": 1.7368233618233617, + "grad_norm": 0.7413665056228638, + "learning_rate": 0.00012069166609531602, + "loss": 1.0523, + "step": 9755 + }, + { + "epoch": 1.7370014245014245, + "grad_norm": 0.7814368009567261, + "learning_rate": 0.00012067797134577275, + "loss": 0.9988, + "step": 9756 + }, + { + "epoch": 1.7371794871794872, + "grad_norm": 0.6174948811531067, + "learning_rate": 0.00012066427619108757, + "loss": 0.9002, + "step": 9757 + }, + { + "epoch": 1.73735754985755, + "grad_norm": 0.6521819233894348, + "learning_rate": 0.00012065058063152885, + "loss": 1.1307, + "step": 9758 + }, + { + "epoch": 1.7375356125356125, + "grad_norm": 0.6797493696212769, + "learning_rate": 0.00012063688466736489, + "loss": 0.84, + "step": 9759 + }, + { + "epoch": 1.7377136752136753, + "grad_norm": 0.6496474146842957, + "learning_rate": 0.00012062318829886404, + "loss": 0.86, + "step": 9760 + }, + { + "epoch": 1.7378917378917378, + "grad_norm": 0.6701306104660034, + "learning_rate": 0.00012060949152629467, + "loss": 0.9422, + "step": 9761 + }, + { + "epoch": 1.7380698005698005, + "grad_norm": 0.7331172823905945, + "learning_rate": 0.00012059579434992512, + "loss": 1.1648, + "step": 9762 + }, + { + "epoch": 1.7382478632478633, + "grad_norm": 0.63930743932724, + "learning_rate": 0.00012058209677002375, + "loss": 1.0617, + "step": 9763 + }, + { + "epoch": 1.738425925925926, + "grad_norm": 0.668851912021637, + "learning_rate": 0.00012056839878685895, + "loss": 0.8219, + "step": 9764 + }, + { + "epoch": 1.7386039886039886, + "grad_norm": 0.7305747270584106, + "learning_rate": 0.00012055470040069912, + "loss": 1.0416, + "step": 9765 + }, + { + "epoch": 1.7387820512820513, + "grad_norm": 0.6931866407394409, + "learning_rate": 0.00012054100161181264, + "loss": 1.0588, + "step": 9766 + }, + { + "epoch": 1.7389601139601139, + "grad_norm": 0.6565485000610352, + "learning_rate": 0.00012052730242046785, + "loss": 0.7885, + "step": 9767 + }, + { + "epoch": 1.7391381766381766, + "grad_norm": 0.739985466003418, + "learning_rate": 0.00012051360282693327, + "loss": 1.0973, + "step": 9768 + }, + { + "epoch": 1.7393162393162394, + "grad_norm": 0.6477079391479492, + "learning_rate": 0.00012049990283147723, + "loss": 0.9841, + "step": 9769 + }, + { + "epoch": 1.739494301994302, + "grad_norm": 0.7018330097198486, + "learning_rate": 0.00012048620243436819, + "loss": 1.0869, + "step": 9770 + }, + { + "epoch": 1.7396723646723646, + "grad_norm": 0.7087421417236328, + "learning_rate": 0.00012047250163587456, + "loss": 0.916, + "step": 9771 + }, + { + "epoch": 1.7398504273504274, + "grad_norm": 0.8747151494026184, + "learning_rate": 0.00012045880043626481, + "loss": 0.8245, + "step": 9772 + }, + { + "epoch": 1.74002849002849, + "grad_norm": 0.777498722076416, + "learning_rate": 0.00012044509883580735, + "loss": 1.071, + "step": 9773 + }, + { + "epoch": 1.7402065527065527, + "grad_norm": 0.6668971180915833, + "learning_rate": 0.00012043139683477062, + "loss": 1.0447, + "step": 9774 + }, + { + "epoch": 1.7403846153846154, + "grad_norm": 0.6702026724815369, + "learning_rate": 0.00012041769443342317, + "loss": 0.8688, + "step": 9775 + }, + { + "epoch": 1.7405626780626782, + "grad_norm": 0.7866267561912537, + "learning_rate": 0.00012040399163203337, + "loss": 1.0842, + "step": 9776 + }, + { + "epoch": 1.7407407407407407, + "grad_norm": 0.7655110955238342, + "learning_rate": 0.00012039028843086977, + "loss": 1.2417, + "step": 9777 + }, + { + "epoch": 1.7409188034188035, + "grad_norm": 0.7084119915962219, + "learning_rate": 0.0001203765848302008, + "loss": 0.9844, + "step": 9778 + }, + { + "epoch": 1.741096866096866, + "grad_norm": 0.7135398983955383, + "learning_rate": 0.00012036288083029497, + "loss": 1.1102, + "step": 9779 + }, + { + "epoch": 1.7412749287749287, + "grad_norm": 0.6784615516662598, + "learning_rate": 0.0001203491764314208, + "loss": 1.0349, + "step": 9780 + }, + { + "epoch": 1.7414529914529915, + "grad_norm": 0.7170301079750061, + "learning_rate": 0.00012033547163384682, + "loss": 1.0899, + "step": 9781 + }, + { + "epoch": 1.7416310541310542, + "grad_norm": 0.6692060828208923, + "learning_rate": 0.0001203217664378415, + "loss": 1.0486, + "step": 9782 + }, + { + "epoch": 1.7418091168091168, + "grad_norm": 0.6730037927627563, + "learning_rate": 0.00012030806084367336, + "loss": 0.9684, + "step": 9783 + }, + { + "epoch": 1.7419871794871795, + "grad_norm": 0.5983504056930542, + "learning_rate": 0.00012029435485161096, + "loss": 0.7106, + "step": 9784 + }, + { + "epoch": 1.742165242165242, + "grad_norm": 0.6834231615066528, + "learning_rate": 0.00012028064846192284, + "loss": 0.803, + "step": 9785 + }, + { + "epoch": 1.7423433048433048, + "grad_norm": 0.621046245098114, + "learning_rate": 0.00012026694167487755, + "loss": 0.9129, + "step": 9786 + }, + { + "epoch": 1.7425213675213675, + "grad_norm": 0.6348989605903625, + "learning_rate": 0.00012025323449074361, + "loss": 1.0076, + "step": 9787 + }, + { + "epoch": 1.7426994301994303, + "grad_norm": 0.6139974594116211, + "learning_rate": 0.00012023952690978966, + "loss": 1.0756, + "step": 9788 + }, + { + "epoch": 1.7428774928774928, + "grad_norm": 0.6473259925842285, + "learning_rate": 0.00012022581893228419, + "loss": 1.0568, + "step": 9789 + }, + { + "epoch": 1.7430555555555556, + "grad_norm": 0.6133778095245361, + "learning_rate": 0.00012021211055849581, + "loss": 0.8722, + "step": 9790 + }, + { + "epoch": 1.743233618233618, + "grad_norm": 0.6934139728546143, + "learning_rate": 0.00012019840178869315, + "loss": 1.0329, + "step": 9791 + }, + { + "epoch": 1.7434116809116809, + "grad_norm": 0.6730150580406189, + "learning_rate": 0.00012018469262314474, + "loss": 0.9326, + "step": 9792 + }, + { + "epoch": 1.7435897435897436, + "grad_norm": 0.6805521249771118, + "learning_rate": 0.0001201709830621192, + "loss": 1.0527, + "step": 9793 + }, + { + "epoch": 1.7437678062678064, + "grad_norm": 0.6972569823265076, + "learning_rate": 0.00012015727310588516, + "loss": 1.0024, + "step": 9794 + }, + { + "epoch": 1.743945868945869, + "grad_norm": 0.7329187989234924, + "learning_rate": 0.00012014356275471122, + "loss": 1.1864, + "step": 9795 + }, + { + "epoch": 1.7441239316239316, + "grad_norm": 0.7220240831375122, + "learning_rate": 0.00012012985200886602, + "loss": 0.8831, + "step": 9796 + }, + { + "epoch": 1.7443019943019942, + "grad_norm": 0.7829749584197998, + "learning_rate": 0.00012011614086861818, + "loss": 1.0365, + "step": 9797 + }, + { + "epoch": 1.744480056980057, + "grad_norm": 0.7148944735527039, + "learning_rate": 0.00012010242933423637, + "loss": 1.0413, + "step": 9798 + }, + { + "epoch": 1.7446581196581197, + "grad_norm": 0.5607262253761292, + "learning_rate": 0.00012008871740598917, + "loss": 0.8154, + "step": 9799 + }, + { + "epoch": 1.7448361823361824, + "grad_norm": 0.754626452922821, + "learning_rate": 0.00012007500508414531, + "loss": 1.0569, + "step": 9800 + }, + { + "epoch": 1.7450142450142452, + "grad_norm": 0.7216293215751648, + "learning_rate": 0.00012006129236897343, + "loss": 1.1641, + "step": 9801 + }, + { + "epoch": 1.7451923076923077, + "grad_norm": 0.6575515270233154, + "learning_rate": 0.0001200475792607422, + "loss": 0.9063, + "step": 9802 + }, + { + "epoch": 1.7453703703703702, + "grad_norm": 0.7411505579948425, + "learning_rate": 0.00012003386575972031, + "loss": 0.9791, + "step": 9803 + }, + { + "epoch": 1.745548433048433, + "grad_norm": 0.6945903301239014, + "learning_rate": 0.0001200201518661764, + "loss": 0.8111, + "step": 9804 + }, + { + "epoch": 1.7457264957264957, + "grad_norm": 0.5760970115661621, + "learning_rate": 0.00012000643758037924, + "loss": 1.1054, + "step": 9805 + }, + { + "epoch": 1.7459045584045585, + "grad_norm": 0.6732224225997925, + "learning_rate": 0.00011999272290259748, + "loss": 0.8992, + "step": 9806 + }, + { + "epoch": 1.7460826210826212, + "grad_norm": 0.673270046710968, + "learning_rate": 0.00011997900783309983, + "loss": 1.0554, + "step": 9807 + }, + { + "epoch": 1.7462606837606838, + "grad_norm": 0.7233314514160156, + "learning_rate": 0.00011996529237215503, + "loss": 1.066, + "step": 9808 + }, + { + "epoch": 1.7464387464387463, + "grad_norm": 0.7016494274139404, + "learning_rate": 0.00011995157652003183, + "loss": 0.891, + "step": 9809 + }, + { + "epoch": 1.746616809116809, + "grad_norm": 0.9377092719078064, + "learning_rate": 0.00011993786027699889, + "loss": 0.8626, + "step": 9810 + }, + { + "epoch": 1.7467948717948718, + "grad_norm": 0.6825845241546631, + "learning_rate": 0.00011992414364332503, + "loss": 0.8996, + "step": 9811 + }, + { + "epoch": 1.7469729344729346, + "grad_norm": 0.6836053729057312, + "learning_rate": 0.00011991042661927896, + "loss": 0.9338, + "step": 9812 + }, + { + "epoch": 1.7471509971509973, + "grad_norm": 0.6462908387184143, + "learning_rate": 0.00011989670920512943, + "loss": 1.1185, + "step": 9813 + }, + { + "epoch": 1.7473290598290598, + "grad_norm": 0.7191921472549438, + "learning_rate": 0.00011988299140114522, + "loss": 0.9084, + "step": 9814 + }, + { + "epoch": 1.7475071225071224, + "grad_norm": 0.6951598525047302, + "learning_rate": 0.00011986927320759508, + "loss": 1.0653, + "step": 9815 + }, + { + "epoch": 1.7476851851851851, + "grad_norm": 0.7512598037719727, + "learning_rate": 0.00011985555462474784, + "loss": 1.0259, + "step": 9816 + }, + { + "epoch": 1.7478632478632479, + "grad_norm": 0.6885492205619812, + "learning_rate": 0.00011984183565287226, + "loss": 0.7148, + "step": 9817 + }, + { + "epoch": 1.7480413105413106, + "grad_norm": 0.6880139708518982, + "learning_rate": 0.00011982811629223709, + "loss": 1.1567, + "step": 9818 + }, + { + "epoch": 1.7482193732193734, + "grad_norm": 0.7381170392036438, + "learning_rate": 0.0001198143965431112, + "loss": 0.8483, + "step": 9819 + }, + { + "epoch": 1.748397435897436, + "grad_norm": 0.6761063933372498, + "learning_rate": 0.00011980067640576333, + "loss": 0.9498, + "step": 9820 + }, + { + "epoch": 1.7485754985754984, + "grad_norm": 0.6454669237136841, + "learning_rate": 0.00011978695588046238, + "loss": 0.7336, + "step": 9821 + }, + { + "epoch": 1.7487535612535612, + "grad_norm": 0.6026871800422668, + "learning_rate": 0.00011977323496747712, + "loss": 0.8618, + "step": 9822 + }, + { + "epoch": 1.748931623931624, + "grad_norm": 0.6877408027648926, + "learning_rate": 0.0001197595136670764, + "loss": 0.9146, + "step": 9823 + }, + { + "epoch": 1.7491096866096867, + "grad_norm": 0.6874892115592957, + "learning_rate": 0.00011974579197952906, + "loss": 1.1628, + "step": 9824 + }, + { + "epoch": 1.7492877492877494, + "grad_norm": 0.7464384436607361, + "learning_rate": 0.00011973206990510393, + "loss": 1.007, + "step": 9825 + }, + { + "epoch": 1.749465811965812, + "grad_norm": 0.7281473278999329, + "learning_rate": 0.00011971834744406986, + "loss": 1.0776, + "step": 9826 + }, + { + "epoch": 1.7496438746438745, + "grad_norm": 0.6112284660339355, + "learning_rate": 0.00011970462459669575, + "loss": 0.7616, + "step": 9827 + }, + { + "epoch": 1.7498219373219372, + "grad_norm": 0.6498035192489624, + "learning_rate": 0.00011969090136325048, + "loss": 0.884, + "step": 9828 + }, + { + "epoch": 1.7498219373219372, + "eval_loss": 1.1018389463424683, + "eval_runtime": 24.5594, + "eval_samples_per_second": 42.387, + "eval_steps_per_second": 21.214, + "step": 9828 + }, + { + "epoch": 1.75, + "grad_norm": 0.6746426224708557, + "learning_rate": 0.00011967717774400289, + "loss": 0.9023, + "step": 9829 + }, + { + "epoch": 1.7501780626780628, + "grad_norm": 0.6513423323631287, + "learning_rate": 0.00011966345373922188, + "loss": 0.9786, + "step": 9830 + }, + { + "epoch": 1.7503561253561255, + "grad_norm": 0.7053804397583008, + "learning_rate": 0.00011964972934917632, + "loss": 1.0667, + "step": 9831 + }, + { + "epoch": 1.750534188034188, + "grad_norm": 0.6769008040428162, + "learning_rate": 0.00011963600457413513, + "loss": 0.8596, + "step": 9832 + }, + { + "epoch": 1.7507122507122506, + "grad_norm": 0.7162246108055115, + "learning_rate": 0.00011962227941436725, + "loss": 1.0746, + "step": 9833 + }, + { + "epoch": 1.7508903133903133, + "grad_norm": 0.7665811777114868, + "learning_rate": 0.00011960855387014156, + "loss": 1.0056, + "step": 9834 + }, + { + "epoch": 1.751068376068376, + "grad_norm": 0.6186950206756592, + "learning_rate": 0.00011959482794172696, + "loss": 0.9016, + "step": 9835 + }, + { + "epoch": 1.7512464387464388, + "grad_norm": 0.8018904328346252, + "learning_rate": 0.00011958110162939245, + "loss": 0.9534, + "step": 9836 + }, + { + "epoch": 1.7514245014245016, + "grad_norm": 0.8239033818244934, + "learning_rate": 0.0001195673749334069, + "loss": 1.214, + "step": 9837 + }, + { + "epoch": 1.751602564102564, + "grad_norm": 0.7886297106742859, + "learning_rate": 0.00011955364785403931, + "loss": 0.9672, + "step": 9838 + }, + { + "epoch": 1.7517806267806266, + "grad_norm": 0.6463177800178528, + "learning_rate": 0.00011953992039155862, + "loss": 0.9184, + "step": 9839 + }, + { + "epoch": 1.7519586894586894, + "grad_norm": 0.7374706864356995, + "learning_rate": 0.00011952619254623374, + "loss": 0.9988, + "step": 9840 + }, + { + "epoch": 1.7521367521367521, + "grad_norm": 0.7456657886505127, + "learning_rate": 0.00011951246431833369, + "loss": 1.2197, + "step": 9841 + }, + { + "epoch": 1.7523148148148149, + "grad_norm": 0.6644248962402344, + "learning_rate": 0.00011949873570812746, + "loss": 0.9449, + "step": 9842 + }, + { + "epoch": 1.7524928774928776, + "grad_norm": 0.707919180393219, + "learning_rate": 0.000119485006715884, + "loss": 0.774, + "step": 9843 + }, + { + "epoch": 1.7526709401709402, + "grad_norm": 0.6273906826972961, + "learning_rate": 0.00011947127734187231, + "loss": 0.8682, + "step": 9844 + }, + { + "epoch": 1.7528490028490027, + "grad_norm": 0.8335350155830383, + "learning_rate": 0.00011945754758636136, + "loss": 1.2282, + "step": 9845 + }, + { + "epoch": 1.7530270655270654, + "grad_norm": 0.6849051117897034, + "learning_rate": 0.00011944381744962022, + "loss": 1.1091, + "step": 9846 + }, + { + "epoch": 1.7532051282051282, + "grad_norm": 0.8571760058403015, + "learning_rate": 0.00011943008693191781, + "loss": 0.9806, + "step": 9847 + }, + { + "epoch": 1.753383190883191, + "grad_norm": 0.7045019268989563, + "learning_rate": 0.00011941635603352328, + "loss": 0.9217, + "step": 9848 + }, + { + "epoch": 1.7535612535612537, + "grad_norm": 0.6820187568664551, + "learning_rate": 0.00011940262475470556, + "loss": 0.9983, + "step": 9849 + }, + { + "epoch": 1.7537393162393162, + "grad_norm": 0.7400697469711304, + "learning_rate": 0.00011938889309573374, + "loss": 0.9521, + "step": 9850 + }, + { + "epoch": 1.7539173789173788, + "grad_norm": 0.7027658820152283, + "learning_rate": 0.00011937516105687678, + "loss": 1.0749, + "step": 9851 + }, + { + "epoch": 1.7540954415954415, + "grad_norm": 0.6778307557106018, + "learning_rate": 0.00011936142863840382, + "loss": 1.0249, + "step": 9852 + }, + { + "epoch": 1.7542735042735043, + "grad_norm": 0.6787961721420288, + "learning_rate": 0.00011934769584058389, + "loss": 1.0014, + "step": 9853 + }, + { + "epoch": 1.754451566951567, + "grad_norm": 0.7515636086463928, + "learning_rate": 0.00011933396266368606, + "loss": 1.0351, + "step": 9854 + }, + { + "epoch": 1.7546296296296298, + "grad_norm": 0.6620134115219116, + "learning_rate": 0.00011932022910797938, + "loss": 1.0294, + "step": 9855 + }, + { + "epoch": 1.7548076923076923, + "grad_norm": 0.8260951638221741, + "learning_rate": 0.00011930649517373294, + "loss": 0.9078, + "step": 9856 + }, + { + "epoch": 1.7549857549857548, + "grad_norm": 0.7680675983428955, + "learning_rate": 0.00011929276086121584, + "loss": 0.92, + "step": 9857 + }, + { + "epoch": 1.7551638176638176, + "grad_norm": 0.7104191184043884, + "learning_rate": 0.00011927902617069717, + "loss": 0.9937, + "step": 9858 + }, + { + "epoch": 1.7553418803418803, + "grad_norm": 0.7185840606689453, + "learning_rate": 0.00011926529110244603, + "loss": 0.9775, + "step": 9859 + }, + { + "epoch": 1.755519943019943, + "grad_norm": 0.7114652991294861, + "learning_rate": 0.00011925155565673151, + "loss": 0.883, + "step": 9860 + }, + { + "epoch": 1.7556980056980058, + "grad_norm": 0.6906639337539673, + "learning_rate": 0.00011923781983382276, + "loss": 0.9789, + "step": 9861 + }, + { + "epoch": 1.7558760683760684, + "grad_norm": 0.706908106803894, + "learning_rate": 0.00011922408363398892, + "loss": 1.1186, + "step": 9862 + }, + { + "epoch": 1.756054131054131, + "grad_norm": 0.7532939910888672, + "learning_rate": 0.00011921034705749908, + "loss": 0.977, + "step": 9863 + }, + { + "epoch": 1.7562321937321936, + "grad_norm": 0.7397763729095459, + "learning_rate": 0.0001191966101046224, + "loss": 1.1121, + "step": 9864 + }, + { + "epoch": 1.7564102564102564, + "grad_norm": 0.6955398321151733, + "learning_rate": 0.00011918287277562801, + "loss": 1.0439, + "step": 9865 + }, + { + "epoch": 1.7565883190883191, + "grad_norm": 0.7485929727554321, + "learning_rate": 0.00011916913507078507, + "loss": 1.1644, + "step": 9866 + }, + { + "epoch": 1.756766381766382, + "grad_norm": 0.6337487101554871, + "learning_rate": 0.00011915539699036274, + "loss": 0.8216, + "step": 9867 + }, + { + "epoch": 1.7569444444444444, + "grad_norm": 0.6628872752189636, + "learning_rate": 0.00011914165853463022, + "loss": 0.9584, + "step": 9868 + }, + { + "epoch": 1.7571225071225072, + "grad_norm": 0.6577547788619995, + "learning_rate": 0.00011912791970385666, + "loss": 0.9484, + "step": 9869 + }, + { + "epoch": 1.7573005698005697, + "grad_norm": 0.6409304738044739, + "learning_rate": 0.00011911418049831127, + "loss": 1.1256, + "step": 9870 + }, + { + "epoch": 1.7574786324786325, + "grad_norm": 0.7499844431877136, + "learning_rate": 0.00011910044091826319, + "loss": 0.7991, + "step": 9871 + }, + { + "epoch": 1.7576566951566952, + "grad_norm": 0.6786715388298035, + "learning_rate": 0.00011908670096398165, + "loss": 1.0368, + "step": 9872 + }, + { + "epoch": 1.757834757834758, + "grad_norm": 0.6432101130485535, + "learning_rate": 0.00011907296063573585, + "loss": 0.9059, + "step": 9873 + }, + { + "epoch": 1.7580128205128205, + "grad_norm": 0.6542613506317139, + "learning_rate": 0.00011905921993379503, + "loss": 0.9866, + "step": 9874 + }, + { + "epoch": 1.7581908831908832, + "grad_norm": 0.6048218011856079, + "learning_rate": 0.00011904547885842838, + "loss": 0.9488, + "step": 9875 + }, + { + "epoch": 1.7583689458689458, + "grad_norm": 0.7694938778877258, + "learning_rate": 0.00011903173740990512, + "loss": 1.1026, + "step": 9876 + }, + { + "epoch": 1.7585470085470085, + "grad_norm": 0.6621627807617188, + "learning_rate": 0.00011901799558849451, + "loss": 1.135, + "step": 9877 + }, + { + "epoch": 1.7587250712250713, + "grad_norm": 0.6561587452888489, + "learning_rate": 0.0001190042533944658, + "loss": 0.9322, + "step": 9878 + }, + { + "epoch": 1.758903133903134, + "grad_norm": 0.7846759557723999, + "learning_rate": 0.00011899051082808821, + "loss": 0.9324, + "step": 9879 + }, + { + "epoch": 1.7590811965811965, + "grad_norm": 0.6004071831703186, + "learning_rate": 0.00011897676788963101, + "loss": 0.9641, + "step": 9880 + }, + { + "epoch": 1.7592592592592593, + "grad_norm": 0.6731070280075073, + "learning_rate": 0.00011896302457936344, + "loss": 1.1437, + "step": 9881 + }, + { + "epoch": 1.7594373219373218, + "grad_norm": 0.6768675446510315, + "learning_rate": 0.00011894928089755481, + "loss": 1.0707, + "step": 9882 + }, + { + "epoch": 1.7596153846153846, + "grad_norm": 0.8368878960609436, + "learning_rate": 0.0001189355368444744, + "loss": 1.0435, + "step": 9883 + }, + { + "epoch": 1.7597934472934473, + "grad_norm": 0.6132324934005737, + "learning_rate": 0.00011892179242039149, + "loss": 0.8889, + "step": 9884 + }, + { + "epoch": 1.75997150997151, + "grad_norm": 0.7598093152046204, + "learning_rate": 0.00011890804762557535, + "loss": 1.151, + "step": 9885 + }, + { + "epoch": 1.7601495726495726, + "grad_norm": 0.7317715883255005, + "learning_rate": 0.00011889430246029527, + "loss": 0.9992, + "step": 9886 + }, + { + "epoch": 1.7603276353276354, + "grad_norm": 0.7664858102798462, + "learning_rate": 0.00011888055692482059, + "loss": 0.8398, + "step": 9887 + }, + { + "epoch": 1.760505698005698, + "grad_norm": 0.6916853189468384, + "learning_rate": 0.00011886681101942063, + "loss": 0.9507, + "step": 9888 + }, + { + "epoch": 1.7606837606837606, + "grad_norm": 0.7103399634361267, + "learning_rate": 0.0001188530647443647, + "loss": 0.915, + "step": 9889 + }, + { + "epoch": 1.7608618233618234, + "grad_norm": 0.6177804470062256, + "learning_rate": 0.00011883931809992215, + "loss": 0.721, + "step": 9890 + }, + { + "epoch": 1.7610398860398861, + "grad_norm": 0.7523959279060364, + "learning_rate": 0.00011882557108636227, + "loss": 0.99, + "step": 9891 + }, + { + "epoch": 1.7612179487179487, + "grad_norm": 0.6211134791374207, + "learning_rate": 0.00011881182370395442, + "loss": 0.8089, + "step": 9892 + }, + { + "epoch": 1.7613960113960114, + "grad_norm": 0.6660307049751282, + "learning_rate": 0.00011879807595296802, + "loss": 1.1062, + "step": 9893 + }, + { + "epoch": 1.761574074074074, + "grad_norm": 0.7039240598678589, + "learning_rate": 0.00011878432783367232, + "loss": 0.9739, + "step": 9894 + }, + { + "epoch": 1.7617521367521367, + "grad_norm": 0.658064603805542, + "learning_rate": 0.00011877057934633675, + "loss": 0.9438, + "step": 9895 + }, + { + "epoch": 1.7619301994301995, + "grad_norm": 0.8227152228355408, + "learning_rate": 0.00011875683049123068, + "loss": 0.8385, + "step": 9896 + }, + { + "epoch": 1.7621082621082622, + "grad_norm": 0.6622483730316162, + "learning_rate": 0.00011874308126862346, + "loss": 0.9432, + "step": 9897 + }, + { + "epoch": 1.7622863247863247, + "grad_norm": 0.7211357951164246, + "learning_rate": 0.00011872933167878453, + "loss": 1.2471, + "step": 9898 + }, + { + "epoch": 1.7624643874643875, + "grad_norm": 0.6177424192428589, + "learning_rate": 0.00011871558172198322, + "loss": 0.8892, + "step": 9899 + }, + { + "epoch": 1.76264245014245, + "grad_norm": 0.6924285888671875, + "learning_rate": 0.00011870183139848898, + "loss": 1.021, + "step": 9900 + }, + { + "epoch": 1.7628205128205128, + "grad_norm": 0.6168648600578308, + "learning_rate": 0.0001186880807085712, + "loss": 0.9013, + "step": 9901 + }, + { + "epoch": 1.7629985754985755, + "grad_norm": 0.6410452723503113, + "learning_rate": 0.00011867432965249929, + "loss": 0.6686, + "step": 9902 + }, + { + "epoch": 1.7631766381766383, + "grad_norm": 0.6959559917449951, + "learning_rate": 0.0001186605782305427, + "loss": 0.9814, + "step": 9903 + }, + { + "epoch": 1.7633547008547008, + "grad_norm": 0.7456178069114685, + "learning_rate": 0.00011864682644297085, + "loss": 1.0151, + "step": 9904 + }, + { + "epoch": 1.7635327635327636, + "grad_norm": 0.6499991416931152, + "learning_rate": 0.00011863307429005317, + "loss": 0.83, + "step": 9905 + }, + { + "epoch": 1.763710826210826, + "grad_norm": 0.643344521522522, + "learning_rate": 0.00011861932177205908, + "loss": 0.8853, + "step": 9906 + }, + { + "epoch": 1.7638888888888888, + "grad_norm": 0.6570441722869873, + "learning_rate": 0.00011860556888925804, + "loss": 0.9179, + "step": 9907 + }, + { + "epoch": 1.7640669515669516, + "grad_norm": 0.6892307996749878, + "learning_rate": 0.00011859181564191957, + "loss": 0.9657, + "step": 9908 + }, + { + "epoch": 1.7642450142450143, + "grad_norm": 0.648158073425293, + "learning_rate": 0.0001185780620303131, + "loss": 0.9179, + "step": 9909 + }, + { + "epoch": 1.7644230769230769, + "grad_norm": 0.5833603143692017, + "learning_rate": 0.00011856430805470808, + "loss": 0.8505, + "step": 9910 + }, + { + "epoch": 1.7646011396011396, + "grad_norm": 0.8302416205406189, + "learning_rate": 0.000118550553715374, + "loss": 0.8948, + "step": 9911 + }, + { + "epoch": 1.7647792022792022, + "grad_norm": 0.7075300216674805, + "learning_rate": 0.00011853679901258035, + "loss": 1.2467, + "step": 9912 + }, + { + "epoch": 1.764957264957265, + "grad_norm": 0.81916344165802, + "learning_rate": 0.00011852304394659666, + "loss": 0.9963, + "step": 9913 + }, + { + "epoch": 1.7651353276353277, + "grad_norm": 0.6492435932159424, + "learning_rate": 0.00011850928851769239, + "loss": 1.0704, + "step": 9914 + }, + { + "epoch": 1.7653133903133904, + "grad_norm": 0.7301090359687805, + "learning_rate": 0.00011849553272613704, + "loss": 1.0477, + "step": 9915 + }, + { + "epoch": 1.765491452991453, + "grad_norm": 0.7280275821685791, + "learning_rate": 0.00011848177657220019, + "loss": 0.9124, + "step": 9916 + }, + { + "epoch": 1.7656695156695157, + "grad_norm": 0.6948845386505127, + "learning_rate": 0.00011846802005615127, + "loss": 1.2275, + "step": 9917 + }, + { + "epoch": 1.7658475783475782, + "grad_norm": 0.6553834676742554, + "learning_rate": 0.0001184542631782599, + "loss": 1.2311, + "step": 9918 + }, + { + "epoch": 1.766025641025641, + "grad_norm": 0.6899739503860474, + "learning_rate": 0.00011844050593879556, + "loss": 0.8936, + "step": 9919 + }, + { + "epoch": 1.7662037037037037, + "grad_norm": 0.6076815128326416, + "learning_rate": 0.00011842674833802782, + "loss": 0.8432, + "step": 9920 + }, + { + "epoch": 1.7663817663817665, + "grad_norm": 0.7650902271270752, + "learning_rate": 0.00011841299037622624, + "loss": 1.0447, + "step": 9921 + }, + { + "epoch": 1.7665598290598292, + "grad_norm": 0.6864938735961914, + "learning_rate": 0.00011839923205366032, + "loss": 0.936, + "step": 9922 + }, + { + "epoch": 1.7667378917378918, + "grad_norm": 0.7176852226257324, + "learning_rate": 0.0001183854733705997, + "loss": 0.9764, + "step": 9923 + }, + { + "epoch": 1.7669159544159543, + "grad_norm": 0.6513439416885376, + "learning_rate": 0.00011837171432731393, + "loss": 1.0095, + "step": 9924 + }, + { + "epoch": 1.767094017094017, + "grad_norm": 0.8031024932861328, + "learning_rate": 0.00011835795492407256, + "loss": 1.1348, + "step": 9925 + }, + { + "epoch": 1.7672720797720798, + "grad_norm": 0.7659830451011658, + "learning_rate": 0.00011834419516114518, + "loss": 0.9058, + "step": 9926 + }, + { + "epoch": 1.7674501424501425, + "grad_norm": 0.8864039778709412, + "learning_rate": 0.00011833043503880145, + "loss": 1.0342, + "step": 9927 + }, + { + "epoch": 1.7676282051282053, + "grad_norm": 0.6870512962341309, + "learning_rate": 0.00011831667455731088, + "loss": 0.9361, + "step": 9928 + }, + { + "epoch": 1.7678062678062678, + "grad_norm": 0.6458830833435059, + "learning_rate": 0.00011830291371694315, + "loss": 0.8215, + "step": 9929 + }, + { + "epoch": 1.7679843304843303, + "grad_norm": 0.7456086874008179, + "learning_rate": 0.00011828915251796787, + "loss": 1.1243, + "step": 9930 + }, + { + "epoch": 1.768162393162393, + "grad_norm": 0.6834850311279297, + "learning_rate": 0.00011827539096065459, + "loss": 0.9536, + "step": 9931 + }, + { + "epoch": 1.7683404558404558, + "grad_norm": 0.643864631652832, + "learning_rate": 0.00011826162904527302, + "loss": 1.1707, + "step": 9932 + }, + { + "epoch": 1.7685185185185186, + "grad_norm": 0.6312864422798157, + "learning_rate": 0.00011824786677209275, + "loss": 0.7937, + "step": 9933 + }, + { + "epoch": 1.7686965811965814, + "grad_norm": 0.6092729568481445, + "learning_rate": 0.00011823410414138343, + "loss": 0.8787, + "step": 9934 + }, + { + "epoch": 1.7688746438746439, + "grad_norm": 0.6859988570213318, + "learning_rate": 0.00011822034115341474, + "loss": 0.9691, + "step": 9935 + }, + { + "epoch": 1.7690527065527064, + "grad_norm": 0.7219935059547424, + "learning_rate": 0.0001182065778084563, + "loss": 1.0606, + "step": 9936 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.6596202850341797, + "learning_rate": 0.00011819281410677778, + "loss": 1.0543, + "step": 9937 + }, + { + "epoch": 1.769408831908832, + "grad_norm": 0.6616338491439819, + "learning_rate": 0.00011817905004864887, + "loss": 0.9757, + "step": 9938 + }, + { + "epoch": 1.7695868945868947, + "grad_norm": 0.6637360453605652, + "learning_rate": 0.00011816528563433924, + "loss": 0.925, + "step": 9939 + }, + { + "epoch": 1.7697649572649574, + "grad_norm": 0.8422333002090454, + "learning_rate": 0.00011815152086411859, + "loss": 1.1343, + "step": 9940 + }, + { + "epoch": 1.76994301994302, + "grad_norm": 0.6638204455375671, + "learning_rate": 0.00011813775573825656, + "loss": 1.2136, + "step": 9941 + }, + { + "epoch": 1.7701210826210825, + "grad_norm": 0.7258831858634949, + "learning_rate": 0.0001181239902570229, + "loss": 0.7308, + "step": 9942 + }, + { + "epoch": 1.7702991452991452, + "grad_norm": 0.730582594871521, + "learning_rate": 0.0001181102244206873, + "loss": 1.1097, + "step": 9943 + }, + { + "epoch": 1.770477207977208, + "grad_norm": 0.7324019074440002, + "learning_rate": 0.00011809645822951946, + "loss": 0.9802, + "step": 9944 + }, + { + "epoch": 1.7706552706552707, + "grad_norm": 0.5565997958183289, + "learning_rate": 0.00011808269168378914, + "loss": 0.7079, + "step": 9945 + }, + { + "epoch": 1.7708333333333335, + "grad_norm": 0.6395503282546997, + "learning_rate": 0.00011806892478376601, + "loss": 1.0048, + "step": 9946 + }, + { + "epoch": 1.771011396011396, + "grad_norm": 0.7670905590057373, + "learning_rate": 0.00011805515752971985, + "loss": 1.2509, + "step": 9947 + }, + { + "epoch": 1.7711894586894585, + "grad_norm": 0.5945813655853271, + "learning_rate": 0.00011804138992192037, + "loss": 0.8856, + "step": 9948 + }, + { + "epoch": 1.7713675213675213, + "grad_norm": 0.7355493307113647, + "learning_rate": 0.00011802762196063737, + "loss": 0.9629, + "step": 9949 + }, + { + "epoch": 1.771545584045584, + "grad_norm": 0.7024806141853333, + "learning_rate": 0.00011801385364614055, + "loss": 1.1351, + "step": 9950 + }, + { + "epoch": 1.7717236467236468, + "grad_norm": 0.6553003191947937, + "learning_rate": 0.00011800008497869968, + "loss": 0.911, + "step": 9951 + }, + { + "epoch": 1.7719017094017095, + "grad_norm": 0.6883971691131592, + "learning_rate": 0.00011798631595858454, + "loss": 1.0099, + "step": 9952 + }, + { + "epoch": 1.772079772079772, + "grad_norm": 0.7106832265853882, + "learning_rate": 0.00011797254658606489, + "loss": 1.0298, + "step": 9953 + }, + { + "epoch": 1.7722578347578346, + "grad_norm": 0.7902877926826477, + "learning_rate": 0.00011795877686141055, + "loss": 1.0572, + "step": 9954 + }, + { + "epoch": 1.7724358974358974, + "grad_norm": 0.7105007171630859, + "learning_rate": 0.00011794500678489126, + "loss": 1.1725, + "step": 9955 + }, + { + "epoch": 1.77261396011396, + "grad_norm": 0.7314959764480591, + "learning_rate": 0.00011793123635677685, + "loss": 1.1074, + "step": 9956 + }, + { + "epoch": 1.7727920227920229, + "grad_norm": 0.6358618140220642, + "learning_rate": 0.00011791746557733712, + "loss": 0.8786, + "step": 9957 + }, + { + "epoch": 1.7729700854700856, + "grad_norm": 0.6441367864608765, + "learning_rate": 0.00011790369444684187, + "loss": 1.1332, + "step": 9958 + }, + { + "epoch": 1.7731481481481481, + "grad_norm": 0.686787486076355, + "learning_rate": 0.0001178899229655609, + "loss": 0.9566, + "step": 9959 + }, + { + "epoch": 1.7733262108262107, + "grad_norm": 0.653840184211731, + "learning_rate": 0.00011787615113376407, + "loss": 0.8763, + "step": 9960 + }, + { + "epoch": 1.7735042735042734, + "grad_norm": 0.7106643915176392, + "learning_rate": 0.00011786237895172119, + "loss": 0.9929, + "step": 9961 + }, + { + "epoch": 1.7736823361823362, + "grad_norm": 0.6634044051170349, + "learning_rate": 0.0001178486064197021, + "loss": 0.7467, + "step": 9962 + }, + { + "epoch": 1.773860398860399, + "grad_norm": 0.7087352871894836, + "learning_rate": 0.00011783483353797663, + "loss": 1.0104, + "step": 9963 + }, + { + "epoch": 1.7740384615384617, + "grad_norm": 0.8088061213493347, + "learning_rate": 0.00011782106030681466, + "loss": 1.0376, + "step": 9964 + }, + { + "epoch": 1.7742165242165242, + "grad_norm": 0.7204688787460327, + "learning_rate": 0.00011780728672648604, + "loss": 0.8556, + "step": 9965 + }, + { + "epoch": 1.7743945868945867, + "grad_norm": 0.7893314957618713, + "learning_rate": 0.0001177935127972606, + "loss": 0.9764, + "step": 9966 + }, + { + "epoch": 1.7745726495726495, + "grad_norm": 0.6098896265029907, + "learning_rate": 0.00011777973851940826, + "loss": 0.9407, + "step": 9967 + }, + { + "epoch": 1.7747507122507122, + "grad_norm": 0.6420868039131165, + "learning_rate": 0.0001177659638931989, + "loss": 1.1328, + "step": 9968 + }, + { + "epoch": 1.774928774928775, + "grad_norm": 0.7732378244400024, + "learning_rate": 0.00011775218891890234, + "loss": 1.1236, + "step": 9969 + }, + { + "epoch": 1.7751068376068377, + "grad_norm": 0.6591582894325256, + "learning_rate": 0.00011773841359678855, + "loss": 1.1523, + "step": 9970 + }, + { + "epoch": 1.7752849002849003, + "grad_norm": 0.6337170004844666, + "learning_rate": 0.00011772463792712738, + "loss": 1.1998, + "step": 9971 + }, + { + "epoch": 1.7754629629629628, + "grad_norm": 0.6400532126426697, + "learning_rate": 0.00011771086191018874, + "loss": 0.9543, + "step": 9972 + }, + { + "epoch": 1.7756410256410255, + "grad_norm": 0.6431527733802795, + "learning_rate": 0.00011769708554624257, + "loss": 0.8164, + "step": 9973 + }, + { + "epoch": 1.7758190883190883, + "grad_norm": 0.7303599119186401, + "learning_rate": 0.00011768330883555876, + "loss": 0.9553, + "step": 9974 + }, + { + "epoch": 1.775997150997151, + "grad_norm": 0.7838605642318726, + "learning_rate": 0.00011766953177840725, + "loss": 0.9759, + "step": 9975 + }, + { + "epoch": 1.7761752136752138, + "grad_norm": 0.6505265831947327, + "learning_rate": 0.00011765575437505796, + "loss": 0.8527, + "step": 9976 + }, + { + "epoch": 1.7763532763532763, + "grad_norm": 0.7336180806159973, + "learning_rate": 0.00011764197662578086, + "loss": 1.1098, + "step": 9977 + }, + { + "epoch": 1.776531339031339, + "grad_norm": 0.7040138244628906, + "learning_rate": 0.00011762819853084586, + "loss": 1.1289, + "step": 9978 + }, + { + "epoch": 1.7767094017094016, + "grad_norm": 0.6414867043495178, + "learning_rate": 0.00011761442009052293, + "loss": 1.0826, + "step": 9979 + }, + { + "epoch": 1.7768874643874644, + "grad_norm": 0.6760666370391846, + "learning_rate": 0.00011760064130508204, + "loss": 1.0188, + "step": 9980 + }, + { + "epoch": 1.7770655270655271, + "grad_norm": 0.7864978909492493, + "learning_rate": 0.00011758686217479316, + "loss": 1.1938, + "step": 9981 + }, + { + "epoch": 1.7772435897435899, + "grad_norm": 0.7964870929718018, + "learning_rate": 0.00011757308269992622, + "loss": 0.9876, + "step": 9982 + }, + { + "epoch": 1.7774216524216524, + "grad_norm": 0.5158692002296448, + "learning_rate": 0.00011755930288075123, + "loss": 0.6508, + "step": 9983 + }, + { + "epoch": 1.7775997150997151, + "grad_norm": 0.7208606600761414, + "learning_rate": 0.00011754552271753819, + "loss": 1.0738, + "step": 9984 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 0.6811334490776062, + "learning_rate": 0.00011753174221055705, + "loss": 1.1216, + "step": 9985 + }, + { + "epoch": 1.7779558404558404, + "grad_norm": 0.6389986276626587, + "learning_rate": 0.00011751796136007787, + "loss": 0.9664, + "step": 9986 + }, + { + "epoch": 1.7781339031339032, + "grad_norm": 0.7081875205039978, + "learning_rate": 0.00011750418016637064, + "loss": 0.9365, + "step": 9987 + }, + { + "epoch": 1.778311965811966, + "grad_norm": 0.7291778326034546, + "learning_rate": 0.00011749039862970535, + "loss": 1.3222, + "step": 9988 + }, + { + "epoch": 1.7784900284900285, + "grad_norm": 0.6790453791618347, + "learning_rate": 0.000117476616750352, + "loss": 0.9537, + "step": 9989 + }, + { + "epoch": 1.7786680911680912, + "grad_norm": 0.6271076202392578, + "learning_rate": 0.00011746283452858069, + "loss": 0.9842, + "step": 9990 + }, + { + "epoch": 1.7788461538461537, + "grad_norm": 0.675628662109375, + "learning_rate": 0.00011744905196466138, + "loss": 0.8675, + "step": 9991 + }, + { + "epoch": 1.7790242165242165, + "grad_norm": 0.7328314185142517, + "learning_rate": 0.00011743526905886417, + "loss": 0.9793, + "step": 9992 + }, + { + "epoch": 1.7792022792022792, + "grad_norm": 0.698764979839325, + "learning_rate": 0.00011742148581145908, + "loss": 0.9527, + "step": 9993 + }, + { + "epoch": 1.779380341880342, + "grad_norm": 0.6911364793777466, + "learning_rate": 0.00011740770222271616, + "loss": 1.1069, + "step": 9994 + }, + { + "epoch": 1.7795584045584045, + "grad_norm": 0.6990836262702942, + "learning_rate": 0.00011739391829290547, + "loss": 0.9132, + "step": 9995 + }, + { + "epoch": 1.7797364672364673, + "grad_norm": 0.7056801319122314, + "learning_rate": 0.0001173801340222971, + "loss": 1.053, + "step": 9996 + }, + { + "epoch": 1.7799145299145298, + "grad_norm": 0.7453791499137878, + "learning_rate": 0.0001173663494111611, + "loss": 0.8806, + "step": 9997 + }, + { + "epoch": 1.7800925925925926, + "grad_norm": 0.7211771011352539, + "learning_rate": 0.00011735256445976757, + "loss": 0.9968, + "step": 9998 + }, + { + "epoch": 1.7802706552706553, + "grad_norm": 0.7259734272956848, + "learning_rate": 0.00011733877916838656, + "loss": 1.167, + "step": 9999 + }, + { + "epoch": 1.780448717948718, + "grad_norm": 0.6931926012039185, + "learning_rate": 0.00011732499353728821, + "loss": 1.0634, + "step": 10000 + }, + { + "epoch": 1.7806267806267806, + "grad_norm": 0.6900074481964111, + "learning_rate": 0.00011731120756674259, + "loss": 0.9718, + "step": 10001 + }, + { + "epoch": 1.7808048433048433, + "grad_norm": 0.6817582845687866, + "learning_rate": 0.00011729742125701984, + "loss": 1.0896, + "step": 10002 + }, + { + "epoch": 1.7809829059829059, + "grad_norm": 0.6901891231536865, + "learning_rate": 0.00011728363460839003, + "loss": 1.0163, + "step": 10003 + }, + { + "epoch": 1.7811609686609686, + "grad_norm": 0.9138323664665222, + "learning_rate": 0.00011726984762112328, + "loss": 1.1713, + "step": 10004 + }, + { + "epoch": 1.7813390313390314, + "grad_norm": 0.6105810403823853, + "learning_rate": 0.00011725606029548977, + "loss": 0.9331, + "step": 10005 + }, + { + "epoch": 1.7815170940170941, + "grad_norm": 0.5605259537696838, + "learning_rate": 0.0001172422726317596, + "loss": 0.7154, + "step": 10006 + }, + { + "epoch": 1.7816951566951567, + "grad_norm": 0.6950963735580444, + "learning_rate": 0.00011722848463020292, + "loss": 1.0093, + "step": 10007 + }, + { + "epoch": 1.7818732193732194, + "grad_norm": 0.6806309819221497, + "learning_rate": 0.00011721469629108988, + "loss": 0.8662, + "step": 10008 + }, + { + "epoch": 1.782051282051282, + "grad_norm": 0.7528520226478577, + "learning_rate": 0.00011720090761469063, + "loss": 0.8567, + "step": 10009 + }, + { + "epoch": 1.7822293447293447, + "grad_norm": 0.6617229580879211, + "learning_rate": 0.00011718711860127529, + "loss": 1.0378, + "step": 10010 + }, + { + "epoch": 1.7824074074074074, + "grad_norm": 0.6468376517295837, + "learning_rate": 0.00011717332925111411, + "loss": 1.0658, + "step": 10011 + }, + { + "epoch": 1.7825854700854702, + "grad_norm": 0.7141897082328796, + "learning_rate": 0.00011715953956447721, + "loss": 1.023, + "step": 10012 + }, + { + "epoch": 1.7827635327635327, + "grad_norm": 0.5777570605278015, + "learning_rate": 0.00011714574954163475, + "loss": 0.9154, + "step": 10013 + }, + { + "epoch": 1.7829415954415955, + "grad_norm": 0.7536137700080872, + "learning_rate": 0.00011713195918285695, + "loss": 0.9651, + "step": 10014 + }, + { + "epoch": 1.783119658119658, + "grad_norm": 0.6977683305740356, + "learning_rate": 0.00011711816848841402, + "loss": 0.7977, + "step": 10015 + }, + { + "epoch": 1.7832977207977208, + "grad_norm": 0.6522472500801086, + "learning_rate": 0.00011710437745857614, + "loss": 0.8834, + "step": 10016 + }, + { + "epoch": 1.7834757834757835, + "grad_norm": 0.6263057589530945, + "learning_rate": 0.0001170905860936135, + "loss": 1.0576, + "step": 10017 + }, + { + "epoch": 1.7836538461538463, + "grad_norm": 0.6470699310302734, + "learning_rate": 0.00011707679439379635, + "loss": 0.9412, + "step": 10018 + }, + { + "epoch": 1.7838319088319088, + "grad_norm": Infinity, + "learning_rate": 0.00011707679439379635, + "loss": 1.1746, + "step": 10019 + }, + { + "epoch": 1.7840099715099715, + "grad_norm": 0.6022017002105713, + "learning_rate": 0.00011706300235939485, + "loss": 0.8945, + "step": 10020 + }, + { + "epoch": 1.784188034188034, + "grad_norm": 0.637208104133606, + "learning_rate": 0.00011704920999067927, + "loss": 1.0215, + "step": 10021 + }, + { + "epoch": 1.7843660968660968, + "grad_norm": 0.7467851042747498, + "learning_rate": 0.00011703541728791987, + "loss": 1.0341, + "step": 10022 + }, + { + "epoch": 1.7845441595441596, + "grad_norm": 0.7562711238861084, + "learning_rate": 0.00011702162425138683, + "loss": 0.9748, + "step": 10023 + }, + { + "epoch": 1.7847222222222223, + "grad_norm": 0.6480089426040649, + "learning_rate": 0.00011700783088135043, + "loss": 1.05, + "step": 10024 + }, + { + "epoch": 1.7849002849002849, + "grad_norm": 0.6293981671333313, + "learning_rate": 0.00011699403717808091, + "loss": 1.0376, + "step": 10025 + }, + { + "epoch": 1.7850783475783476, + "grad_norm": 0.6821253895759583, + "learning_rate": 0.00011698024314184853, + "loss": 1.0542, + "step": 10026 + }, + { + "epoch": 1.7852564102564101, + "grad_norm": 0.6681216359138489, + "learning_rate": 0.00011696644877292356, + "loss": 1.0018, + "step": 10027 + }, + { + "epoch": 1.7854344729344729, + "grad_norm": 0.6788804531097412, + "learning_rate": 0.00011695265407157628, + "loss": 1.1823, + "step": 10028 + }, + { + "epoch": 1.7856125356125356, + "grad_norm": 0.6147881150245667, + "learning_rate": 0.00011693885903807697, + "loss": 0.9246, + "step": 10029 + }, + { + "epoch": 1.7857905982905984, + "grad_norm": 0.7952296137809753, + "learning_rate": 0.00011692506367269588, + "loss": 1.0528, + "step": 10030 + }, + { + "epoch": 1.785968660968661, + "grad_norm": 0.6985954642295837, + "learning_rate": 0.00011691126797570333, + "loss": 0.9173, + "step": 10031 + }, + { + "epoch": 1.7861467236467237, + "grad_norm": 0.6211223602294922, + "learning_rate": 0.00011689747194736961, + "loss": 0.7527, + "step": 10032 + }, + { + "epoch": 1.7863247863247862, + "grad_norm": 0.7531208992004395, + "learning_rate": 0.00011688367558796507, + "loss": 1.1087, + "step": 10033 + }, + { + "epoch": 1.786502849002849, + "grad_norm": 0.7742924690246582, + "learning_rate": 0.00011686987889775996, + "loss": 1.1512, + "step": 10034 + }, + { + "epoch": 1.7866809116809117, + "grad_norm": 0.7046231627464294, + "learning_rate": 0.00011685608187702459, + "loss": 1.0516, + "step": 10035 + }, + { + "epoch": 1.7868589743589745, + "grad_norm": 0.6264076232910156, + "learning_rate": 0.00011684228452602933, + "loss": 0.8938, + "step": 10036 + }, + { + "epoch": 1.7870370370370372, + "grad_norm": 0.6342145800590515, + "learning_rate": 0.00011682848684504448, + "loss": 0.8177, + "step": 10037 + }, + { + "epoch": 1.7872150997150997, + "grad_norm": 0.6609861254692078, + "learning_rate": 0.00011681468883434041, + "loss": 0.9692, + "step": 10038 + }, + { + "epoch": 1.7873931623931623, + "grad_norm": 0.7918622493743896, + "learning_rate": 0.00011680089049418743, + "loss": 0.8246, + "step": 10039 + }, + { + "epoch": 1.787571225071225, + "grad_norm": 0.697712779045105, + "learning_rate": 0.00011678709182485592, + "loss": 0.8981, + "step": 10040 + }, + { + "epoch": 1.7877492877492878, + "grad_norm": 0.6747658252716064, + "learning_rate": 0.00011677329282661617, + "loss": 1.1243, + "step": 10041 + }, + { + "epoch": 1.7879273504273505, + "grad_norm": 0.6525771617889404, + "learning_rate": 0.00011675949349973863, + "loss": 0.852, + "step": 10042 + }, + { + "epoch": 1.7881054131054133, + "grad_norm": 0.7062464952468872, + "learning_rate": 0.00011674569384449363, + "loss": 1.2582, + "step": 10043 + }, + { + "epoch": 1.7882834757834758, + "grad_norm": 0.6453786492347717, + "learning_rate": 0.00011673189386115154, + "loss": 0.868, + "step": 10044 + }, + { + "epoch": 1.7884615384615383, + "grad_norm": 0.7939708232879639, + "learning_rate": 0.00011671809354998273, + "loss": 0.7553, + "step": 10045 + }, + { + "epoch": 1.788639601139601, + "grad_norm": 0.6466066837310791, + "learning_rate": 0.00011670429291125761, + "loss": 0.942, + "step": 10046 + }, + { + "epoch": 1.7888176638176638, + "grad_norm": 0.7380510568618774, + "learning_rate": 0.00011669049194524657, + "loss": 1.044, + "step": 10047 + }, + { + "epoch": 1.7889957264957266, + "grad_norm": 0.6719707250595093, + "learning_rate": 0.00011667669065222002, + "loss": 1.1624, + "step": 10048 + }, + { + "epoch": 1.7891737891737893, + "grad_norm": 0.6996603012084961, + "learning_rate": 0.00011666288903244837, + "loss": 1.001, + "step": 10049 + }, + { + "epoch": 1.7893518518518519, + "grad_norm": 0.696590006351471, + "learning_rate": 0.00011664908708620202, + "loss": 1.17, + "step": 10050 + }, + { + "epoch": 1.7895299145299144, + "grad_norm": 0.7226764559745789, + "learning_rate": 0.00011663528481375137, + "loss": 1.0762, + "step": 10051 + }, + { + "epoch": 1.7897079772079771, + "grad_norm": 0.6117866635322571, + "learning_rate": 0.00011662148221536689, + "loss": 0.9199, + "step": 10052 + }, + { + "epoch": 1.78988603988604, + "grad_norm": 0.6424985527992249, + "learning_rate": 0.000116607679291319, + "loss": 1.1672, + "step": 10053 + }, + { + "epoch": 1.7900641025641026, + "grad_norm": 0.6390290856361389, + "learning_rate": 0.00011659387604187813, + "loss": 1.1895, + "step": 10054 + }, + { + "epoch": 1.7902421652421654, + "grad_norm": 0.6553205251693726, + "learning_rate": 0.00011658007246731473, + "loss": 1.0967, + "step": 10055 + }, + { + "epoch": 1.790420227920228, + "grad_norm": 0.7737570405006409, + "learning_rate": 0.00011656626856789922, + "loss": 0.9637, + "step": 10056 + }, + { + "epoch": 1.7905982905982905, + "grad_norm": 0.644296407699585, + "learning_rate": 0.00011655246434390212, + "loss": 0.9933, + "step": 10057 + }, + { + "epoch": 1.7907763532763532, + "grad_norm": 0.8154410123825073, + "learning_rate": 0.00011653865979559388, + "loss": 0.9623, + "step": 10058 + }, + { + "epoch": 1.790954415954416, + "grad_norm": 0.7181384563446045, + "learning_rate": 0.00011652485492324495, + "loss": 0.9113, + "step": 10059 + }, + { + "epoch": 1.7911324786324787, + "grad_norm": 0.7835097908973694, + "learning_rate": 0.00011651104972712582, + "loss": 1.0804, + "step": 10060 + }, + { + "epoch": 1.7913105413105415, + "grad_norm": 0.6843693852424622, + "learning_rate": 0.00011649724420750691, + "loss": 1.0242, + "step": 10061 + }, + { + "epoch": 1.791488603988604, + "grad_norm": 0.8364703059196472, + "learning_rate": 0.00011648343836465885, + "loss": 0.8445, + "step": 10062 + }, + { + "epoch": 1.7916666666666665, + "grad_norm": 0.7122092843055725, + "learning_rate": 0.00011646963219885201, + "loss": 1.0453, + "step": 10063 + }, + { + "epoch": 1.7918447293447293, + "grad_norm": 0.7018755078315735, + "learning_rate": 0.00011645582571035696, + "loss": 0.9753, + "step": 10064 + }, + { + "epoch": 1.792022792022792, + "grad_norm": 0.6522594094276428, + "learning_rate": 0.00011644201889944419, + "loss": 1.0328, + "step": 10065 + }, + { + "epoch": 1.7922008547008548, + "grad_norm": 0.70301353931427, + "learning_rate": 0.00011642821176638419, + "loss": 0.9143, + "step": 10066 + }, + { + "epoch": 1.7923789173789175, + "grad_norm": 0.6255469918251038, + "learning_rate": 0.0001164144043114475, + "loss": 0.9527, + "step": 10067 + }, + { + "epoch": 1.79255698005698, + "grad_norm": 0.6780602931976318, + "learning_rate": 0.0001164005965349047, + "loss": 0.9192, + "step": 10068 + }, + { + "epoch": 1.7927350427350426, + "grad_norm": 0.6025984287261963, + "learning_rate": 0.00011638678843702626, + "loss": 0.9055, + "step": 10069 + }, + { + "epoch": 1.7929131054131053, + "grad_norm": 0.6430829763412476, + "learning_rate": 0.00011637298001808275, + "loss": 0.9359, + "step": 10070 + }, + { + "epoch": 1.793091168091168, + "grad_norm": 0.6388106942176819, + "learning_rate": 0.0001163591712783447, + "loss": 0.8847, + "step": 10071 + }, + { + "epoch": 1.7932692307692308, + "grad_norm": 0.706347644329071, + "learning_rate": 0.00011634536221808265, + "loss": 0.9055, + "step": 10072 + }, + { + "epoch": 1.7934472934472936, + "grad_norm": 0.661226749420166, + "learning_rate": 0.00011633155283756721, + "loss": 1.118, + "step": 10073 + }, + { + "epoch": 1.7936253561253561, + "grad_norm": 0.543207049369812, + "learning_rate": 0.00011631774313706891, + "loss": 0.8856, + "step": 10074 + }, + { + "epoch": 1.7938034188034186, + "grad_norm": 0.6514154672622681, + "learning_rate": 0.00011630393311685835, + "loss": 0.8967, + "step": 10075 + }, + { + "epoch": 1.7939814814814814, + "grad_norm": 0.8669198155403137, + "learning_rate": 0.00011629012277720607, + "loss": 1.0362, + "step": 10076 + }, + { + "epoch": 1.7941595441595442, + "grad_norm": 0.7256068587303162, + "learning_rate": 0.00011627631211838266, + "loss": 1.1948, + "step": 10077 + }, + { + "epoch": 1.794337606837607, + "grad_norm": 0.6504935622215271, + "learning_rate": 0.00011626250114065875, + "loss": 0.8309, + "step": 10078 + }, + { + "epoch": 1.7945156695156697, + "grad_norm": 0.6964160799980164, + "learning_rate": 0.0001162486898443049, + "loss": 0.9593, + "step": 10079 + }, + { + "epoch": 1.7946937321937322, + "grad_norm": 0.668727695941925, + "learning_rate": 0.00011623487822959174, + "loss": 0.8897, + "step": 10080 + }, + { + "epoch": 1.7948717948717947, + "grad_norm": 0.6907223463058472, + "learning_rate": 0.00011622106629678986, + "loss": 0.897, + "step": 10081 + }, + { + "epoch": 1.7950498575498575, + "grad_norm": 0.6652865409851074, + "learning_rate": 0.00011620725404616985, + "loss": 0.9321, + "step": 10082 + }, + { + "epoch": 1.7952279202279202, + "grad_norm": 0.6523811221122742, + "learning_rate": 0.00011619344147800239, + "loss": 0.8991, + "step": 10083 + }, + { + "epoch": 1.795405982905983, + "grad_norm": 0.6162952184677124, + "learning_rate": 0.0001161796285925581, + "loss": 0.8061, + "step": 10084 + }, + { + "epoch": 1.7955840455840457, + "grad_norm": 0.670606791973114, + "learning_rate": 0.0001161658153901076, + "loss": 0.9341, + "step": 10085 + }, + { + "epoch": 1.7957621082621082, + "grad_norm": 0.6372489333152771, + "learning_rate": 0.00011615200187092148, + "loss": 1.1049, + "step": 10086 + }, + { + "epoch": 1.7959401709401708, + "grad_norm": 0.7311037182807922, + "learning_rate": 0.00011613818803527045, + "loss": 1.0881, + "step": 10087 + }, + { + "epoch": 1.7961182336182335, + "grad_norm": 0.7440751194953918, + "learning_rate": 0.00011612437388342518, + "loss": 0.9487, + "step": 10088 + }, + { + "epoch": 1.7962962962962963, + "grad_norm": 0.6605934500694275, + "learning_rate": 0.00011611055941565629, + "loss": 0.8757, + "step": 10089 + }, + { + "epoch": 1.796474358974359, + "grad_norm": 0.7546001076698303, + "learning_rate": 0.00011609674463223446, + "loss": 0.9368, + "step": 10090 + }, + { + "epoch": 1.7966524216524218, + "grad_norm": 0.7001389861106873, + "learning_rate": 0.00011608292953343036, + "loss": 0.9098, + "step": 10091 + }, + { + "epoch": 1.7968304843304843, + "grad_norm": 0.6898102760314941, + "learning_rate": 0.00011606911411951462, + "loss": 0.8821, + "step": 10092 + }, + { + "epoch": 1.797008547008547, + "grad_norm": 0.7020773887634277, + "learning_rate": 0.00011605529839075801, + "loss": 1.2775, + "step": 10093 + }, + { + "epoch": 1.7971866096866096, + "grad_norm": 0.6061446070671082, + "learning_rate": 0.0001160414823474312, + "loss": 1.0156, + "step": 10094 + }, + { + "epoch": 1.7973646723646723, + "grad_norm": 0.6746069192886353, + "learning_rate": 0.00011602766598980484, + "loss": 0.8223, + "step": 10095 + }, + { + "epoch": 1.797542735042735, + "grad_norm": 0.655829131603241, + "learning_rate": 0.00011601384931814967, + "loss": 0.9482, + "step": 10096 + }, + { + "epoch": 1.7977207977207978, + "grad_norm": 0.6762703061103821, + "learning_rate": 0.00011600003233273636, + "loss": 1.0191, + "step": 10097 + }, + { + "epoch": 1.7978988603988604, + "grad_norm": 0.7610527276992798, + "learning_rate": 0.00011598621503383566, + "loss": 1.0771, + "step": 10098 + }, + { + "epoch": 1.7980769230769231, + "grad_norm": 0.6857240200042725, + "learning_rate": 0.0001159723974217183, + "loss": 0.8325, + "step": 10099 + }, + { + "epoch": 1.7982549857549857, + "grad_norm": 0.6897954940795898, + "learning_rate": 0.00011595857949665501, + "loss": 1.0064, + "step": 10100 + }, + { + "epoch": 1.7984330484330484, + "grad_norm": 0.7023211717605591, + "learning_rate": 0.00011594476125891649, + "loss": 1.1346, + "step": 10101 + }, + { + "epoch": 1.7986111111111112, + "grad_norm": 0.8131003975868225, + "learning_rate": 0.00011593094270877347, + "loss": 1.0384, + "step": 10102 + }, + { + "epoch": 1.798789173789174, + "grad_norm": 0.6504445672035217, + "learning_rate": 0.00011591712384649676, + "loss": 0.8172, + "step": 10103 + }, + { + "epoch": 1.7989672364672364, + "grad_norm": 0.7379748821258545, + "learning_rate": 0.00011590330467235704, + "loss": 1.0118, + "step": 10104 + }, + { + "epoch": 1.7991452991452992, + "grad_norm": 0.8867329955101013, + "learning_rate": 0.0001158894851866251, + "loss": 1.023, + "step": 10105 + }, + { + "epoch": 1.7993233618233617, + "grad_norm": 0.7057412266731262, + "learning_rate": 0.00011587566538957173, + "loss": 0.8415, + "step": 10106 + }, + { + "epoch": 1.7995014245014245, + "grad_norm": 0.7479654550552368, + "learning_rate": 0.00011586184528146769, + "loss": 0.9663, + "step": 10107 + }, + { + "epoch": 1.7996794871794872, + "grad_norm": 0.6280845403671265, + "learning_rate": 0.00011584802486258368, + "loss": 0.973, + "step": 10108 + }, + { + "epoch": 1.79985754985755, + "grad_norm": 0.6735749840736389, + "learning_rate": 0.00011583420413319059, + "loss": 0.8631, + "step": 10109 + }, + { + "epoch": 1.8000356125356125, + "grad_norm": 0.5940406918525696, + "learning_rate": 0.00011582038309355918, + "loss": 0.8533, + "step": 10110 + }, + { + "epoch": 1.8002136752136753, + "grad_norm": 0.6923874020576477, + "learning_rate": 0.00011580656174396021, + "loss": 1.1105, + "step": 10111 + }, + { + "epoch": 1.8003917378917378, + "grad_norm": 0.6996715664863586, + "learning_rate": 0.00011579274008466447, + "loss": 0.9952, + "step": 10112 + }, + { + "epoch": 1.8005698005698005, + "grad_norm": 0.656561553478241, + "learning_rate": 0.00011577891811594281, + "loss": 0.9621, + "step": 10113 + }, + { + "epoch": 1.8007478632478633, + "grad_norm": 0.7121242880821228, + "learning_rate": 0.00011576509583806605, + "loss": 0.8658, + "step": 10114 + }, + { + "epoch": 1.800925925925926, + "grad_norm": 0.7864459753036499, + "learning_rate": 0.00011575127325130498, + "loss": 0.9867, + "step": 10115 + }, + { + "epoch": 1.8011039886039886, + "grad_norm": 0.6086452007293701, + "learning_rate": 0.00011573745035593042, + "loss": 0.8625, + "step": 10116 + }, + { + "epoch": 1.8012820512820513, + "grad_norm": 0.6553642749786377, + "learning_rate": 0.00011572362715221321, + "loss": 0.8475, + "step": 10117 + }, + { + "epoch": 1.8014601139601139, + "grad_norm": 0.6677348017692566, + "learning_rate": 0.00011570980364042419, + "loss": 0.9672, + "step": 10118 + }, + { + "epoch": 1.8016381766381766, + "grad_norm": 0.6275015473365784, + "learning_rate": 0.0001156959798208342, + "loss": 0.8663, + "step": 10119 + }, + { + "epoch": 1.8018162393162394, + "grad_norm": 0.787568211555481, + "learning_rate": 0.0001156821556937141, + "loss": 1.0188, + "step": 10120 + }, + { + "epoch": 1.801994301994302, + "grad_norm": 0.6983163356781006, + "learning_rate": 0.00011566833125933473, + "loss": 1.0767, + "step": 10121 + }, + { + "epoch": 1.8021723646723646, + "grad_norm": 0.7008936405181885, + "learning_rate": 0.00011565450651796695, + "loss": 1.0116, + "step": 10122 + }, + { + "epoch": 1.8023504273504274, + "grad_norm": 0.7694976925849915, + "learning_rate": 0.00011564068146988163, + "loss": 1.0227, + "step": 10123 + }, + { + "epoch": 1.80252849002849, + "grad_norm": 0.9530014991760254, + "learning_rate": 0.00011562685611534967, + "loss": 0.907, + "step": 10124 + }, + { + "epoch": 1.8027065527065527, + "grad_norm": 0.6714984178543091, + "learning_rate": 0.00011561303045464189, + "loss": 0.9501, + "step": 10125 + }, + { + "epoch": 1.8028846153846154, + "grad_norm": 0.7233797311782837, + "learning_rate": 0.00011559920448802925, + "loss": 1.021, + "step": 10126 + }, + { + "epoch": 1.8030626780626782, + "grad_norm": 0.7600540518760681, + "learning_rate": 0.0001155853782157826, + "loss": 1.1056, + "step": 10127 + }, + { + "epoch": 1.8032407407407407, + "grad_norm": 0.7836297750473022, + "learning_rate": 0.00011557155163817281, + "loss": 0.9906, + "step": 10128 + }, + { + "epoch": 1.8034188034188035, + "grad_norm": 0.7161104083061218, + "learning_rate": 0.00011555772475547084, + "loss": 0.9541, + "step": 10129 + }, + { + "epoch": 1.803596866096866, + "grad_norm": 0.6613732576370239, + "learning_rate": 0.00011554389756794757, + "loss": 0.9188, + "step": 10130 + }, + { + "epoch": 1.8037749287749287, + "grad_norm": 0.6415915489196777, + "learning_rate": 0.00011553007007587391, + "loss": 0.9928, + "step": 10131 + }, + { + "epoch": 1.8039529914529915, + "grad_norm": 0.7730516195297241, + "learning_rate": 0.0001155162422795208, + "loss": 1.0654, + "step": 10132 + }, + { + "epoch": 1.8041310541310542, + "grad_norm": 0.6769654750823975, + "learning_rate": 0.00011550241417915913, + "loss": 1.0678, + "step": 10133 + }, + { + "epoch": 1.8043091168091168, + "grad_norm": 0.6542425751686096, + "learning_rate": 0.00011548858577505988, + "loss": 0.9796, + "step": 10134 + }, + { + "epoch": 1.8044871794871795, + "grad_norm": 0.7282404899597168, + "learning_rate": 0.00011547475706749395, + "loss": 1.0314, + "step": 10135 + }, + { + "epoch": 1.804665242165242, + "grad_norm": 0.6450245976448059, + "learning_rate": 0.00011546092805673232, + "loss": 0.9564, + "step": 10136 + }, + { + "epoch": 1.8048433048433048, + "grad_norm": 0.65577632188797, + "learning_rate": 0.0001154470987430459, + "loss": 1.0219, + "step": 10137 + }, + { + "epoch": 1.8050213675213675, + "grad_norm": 0.7151737809181213, + "learning_rate": 0.00011543326912670567, + "loss": 0.9245, + "step": 10138 + }, + { + "epoch": 1.8051994301994303, + "grad_norm": 0.6695905327796936, + "learning_rate": 0.00011541943920798259, + "loss": 0.9535, + "step": 10139 + }, + { + "epoch": 1.8053774928774928, + "grad_norm": 0.7443813681602478, + "learning_rate": 0.00011540560898714767, + "loss": 1.1697, + "step": 10140 + }, + { + "epoch": 1.8055555555555556, + "grad_norm": 0.5701992511749268, + "learning_rate": 0.0001153917784644718, + "loss": 0.7868, + "step": 10141 + }, + { + "epoch": 1.805733618233618, + "grad_norm": 0.6992354989051819, + "learning_rate": 0.00011537794764022605, + "loss": 0.9856, + "step": 10142 + }, + { + "epoch": 1.8059116809116809, + "grad_norm": 0.6354477405548096, + "learning_rate": 0.00011536411651468131, + "loss": 0.8752, + "step": 10143 + }, + { + "epoch": 1.8060897435897436, + "grad_norm": 0.6952932476997375, + "learning_rate": 0.00011535028508810864, + "loss": 0.9446, + "step": 10144 + }, + { + "epoch": 1.8062678062678064, + "grad_norm": 0.5527541637420654, + "learning_rate": 0.00011533645336077901, + "loss": 0.5486, + "step": 10145 + }, + { + "epoch": 1.806445868945869, + "grad_norm": 0.685046374797821, + "learning_rate": 0.00011532262133296345, + "loss": 0.9529, + "step": 10146 + }, + { + "epoch": 1.8066239316239316, + "grad_norm": 0.6927558779716492, + "learning_rate": 0.00011530878900493296, + "loss": 1.1758, + "step": 10147 + }, + { + "epoch": 1.8068019943019942, + "grad_norm": 0.6758309602737427, + "learning_rate": 0.00011529495637695855, + "loss": 1.0076, + "step": 10148 + }, + { + "epoch": 1.806980056980057, + "grad_norm": 0.6739441156387329, + "learning_rate": 0.00011528112344931121, + "loss": 1.1914, + "step": 10149 + }, + { + "epoch": 1.8071581196581197, + "grad_norm": 0.7031944394111633, + "learning_rate": 0.00011526729022226204, + "loss": 0.783, + "step": 10150 + }, + { + "epoch": 1.8073361823361824, + "grad_norm": 0.6476930975914001, + "learning_rate": 0.00011525345669608202, + "loss": 0.9595, + "step": 10151 + }, + { + "epoch": 1.8075142450142452, + "grad_norm": 0.710498571395874, + "learning_rate": 0.00011523962287104222, + "loss": 0.8821, + "step": 10152 + }, + { + "epoch": 1.8076923076923077, + "grad_norm": 0.6664412617683411, + "learning_rate": 0.00011522578874741365, + "loss": 1.0182, + "step": 10153 + }, + { + "epoch": 1.8078703703703702, + "grad_norm": 0.8374263048171997, + "learning_rate": 0.00011521195432546737, + "loss": 0.9394, + "step": 10154 + }, + { + "epoch": 1.808048433048433, + "grad_norm": 0.6770764589309692, + "learning_rate": 0.00011519811960547447, + "loss": 1.0568, + "step": 10155 + }, + { + "epoch": 1.8082264957264957, + "grad_norm": 0.7014045715332031, + "learning_rate": 0.00011518428458770595, + "loss": 1.1705, + "step": 10156 + }, + { + "epoch": 1.8084045584045585, + "grad_norm": 0.6590061187744141, + "learning_rate": 0.00011517044927243295, + "loss": 1.1233, + "step": 10157 + }, + { + "epoch": 1.8085826210826212, + "grad_norm": 0.6093801856040955, + "learning_rate": 0.00011515661365992647, + "loss": 0.953, + "step": 10158 + }, + { + "epoch": 1.8087606837606838, + "grad_norm": 0.6197089552879333, + "learning_rate": 0.00011514277775045768, + "loss": 0.9414, + "step": 10159 + }, + { + "epoch": 1.8089387464387463, + "grad_norm": 0.7530463337898254, + "learning_rate": 0.00011512894154429759, + "loss": 0.9168, + "step": 10160 + }, + { + "epoch": 1.809116809116809, + "grad_norm": 0.6051347851753235, + "learning_rate": 0.00011511510504171735, + "loss": 0.9132, + "step": 10161 + }, + { + "epoch": 1.8092948717948718, + "grad_norm": 0.6388311982154846, + "learning_rate": 0.000115101268242988, + "loss": 0.6551, + "step": 10162 + }, + { + "epoch": 1.8094729344729346, + "grad_norm": 0.7040972709655762, + "learning_rate": 0.00011508743114838063, + "loss": 0.9409, + "step": 10163 + }, + { + "epoch": 1.8096509971509973, + "grad_norm": 0.7669548392295837, + "learning_rate": 0.00011507359375816644, + "loss": 1.0376, + "step": 10164 + }, + { + "epoch": 1.8098290598290598, + "grad_norm": 0.7309662699699402, + "learning_rate": 0.00011505975607261646, + "loss": 0.9071, + "step": 10165 + }, + { + "epoch": 1.8100071225071224, + "grad_norm": 0.6624547839164734, + "learning_rate": 0.00011504591809200187, + "loss": 1.0765, + "step": 10166 + }, + { + "epoch": 1.8101851851851851, + "grad_norm": 0.7719045281410217, + "learning_rate": 0.00011503207981659376, + "loss": 0.9244, + "step": 10167 + }, + { + "epoch": 1.8103632478632479, + "grad_norm": 0.6701484322547913, + "learning_rate": 0.0001150182412466633, + "loss": 0.9475, + "step": 10168 + }, + { + "epoch": 1.8105413105413106, + "grad_norm": 0.5604981184005737, + "learning_rate": 0.00011500440238248154, + "loss": 0.6268, + "step": 10169 + }, + { + "epoch": 1.8107193732193734, + "grad_norm": 0.6736510992050171, + "learning_rate": 0.00011499056322431973, + "loss": 0.9088, + "step": 10170 + }, + { + "epoch": 1.810897435897436, + "grad_norm": 0.7428455948829651, + "learning_rate": 0.00011497672377244897, + "loss": 0.9298, + "step": 10171 + }, + { + "epoch": 1.8110754985754984, + "grad_norm": 0.6543142795562744, + "learning_rate": 0.00011496288402714042, + "loss": 0.8863, + "step": 10172 + }, + { + "epoch": 1.8112535612535612, + "grad_norm": 0.6809250712394714, + "learning_rate": 0.00011494904398866524, + "loss": 0.977, + "step": 10173 + }, + { + "epoch": 1.811431623931624, + "grad_norm": 0.8105120062828064, + "learning_rate": 0.00011493520365729456, + "loss": 1.2115, + "step": 10174 + }, + { + "epoch": 1.8116096866096867, + "grad_norm": 0.6985095143318176, + "learning_rate": 0.00011492136303329964, + "loss": 0.8233, + "step": 10175 + }, + { + "epoch": 1.8117877492877494, + "grad_norm": 0.7198361754417419, + "learning_rate": 0.00011490752211695158, + "loss": 1.0552, + "step": 10176 + }, + { + "epoch": 1.811965811965812, + "grad_norm": 0.7077036499977112, + "learning_rate": 0.0001148936809085216, + "loss": 0.9171, + "step": 10177 + }, + { + "epoch": 1.8121438746438745, + "grad_norm": 0.9362925887107849, + "learning_rate": 0.00011487983940828089, + "loss": 0.9042, + "step": 10178 + }, + { + "epoch": 1.8123219373219372, + "grad_norm": 0.6732819676399231, + "learning_rate": 0.0001148659976165006, + "loss": 1.1033, + "step": 10179 + }, + { + "epoch": 1.8125, + "grad_norm": 0.747702419757843, + "learning_rate": 0.00011485215553345201, + "loss": 1.0692, + "step": 10180 + }, + { + "epoch": 1.8126780626780628, + "grad_norm": 0.7011259198188782, + "learning_rate": 0.00011483831315940627, + "loss": 0.9278, + "step": 10181 + }, + { + "epoch": 1.8128561253561255, + "grad_norm": 0.8542702198028564, + "learning_rate": 0.00011482447049463462, + "loss": 0.9476, + "step": 10182 + }, + { + "epoch": 1.813034188034188, + "grad_norm": 0.6975166201591492, + "learning_rate": 0.00011481062753940825, + "loss": 0.9486, + "step": 10183 + }, + { + "epoch": 1.8132122507122506, + "grad_norm": 0.8239036798477173, + "learning_rate": 0.0001147967842939984, + "loss": 1.0518, + "step": 10184 + }, + { + "epoch": 1.8133903133903133, + "grad_norm": 0.7559717297554016, + "learning_rate": 0.00011478294075867628, + "loss": 1.1877, + "step": 10185 + }, + { + "epoch": 1.813568376068376, + "grad_norm": 0.6755532026290894, + "learning_rate": 0.00011476909693371318, + "loss": 0.9287, + "step": 10186 + }, + { + "epoch": 1.8137464387464388, + "grad_norm": 0.6561332941055298, + "learning_rate": 0.0001147552528193803, + "loss": 0.83, + "step": 10187 + }, + { + "epoch": 1.8139245014245016, + "grad_norm": 0.7223508954048157, + "learning_rate": 0.00011474140841594887, + "loss": 1.1259, + "step": 10188 + }, + { + "epoch": 1.814102564102564, + "grad_norm": 0.7920593023300171, + "learning_rate": 0.0001147275637236902, + "loss": 1.0925, + "step": 10189 + }, + { + "epoch": 1.8142806267806266, + "grad_norm": 0.6896616816520691, + "learning_rate": 0.00011471371874287546, + "loss": 1.0204, + "step": 10190 + }, + { + "epoch": 1.8144586894586894, + "grad_norm": 0.6149865388870239, + "learning_rate": 0.00011469987347377602, + "loss": 1.1249, + "step": 10191 + }, + { + "epoch": 1.8146367521367521, + "grad_norm": 0.6650002598762512, + "learning_rate": 0.00011468602791666307, + "loss": 0.9723, + "step": 10192 + }, + { + "epoch": 1.8148148148148149, + "grad_norm": 0.7298738956451416, + "learning_rate": 0.00011467218207180792, + "loss": 1.0225, + "step": 10193 + }, + { + "epoch": 1.8149928774928776, + "grad_norm": 0.8075628876686096, + "learning_rate": 0.00011465833593948183, + "loss": 1.0429, + "step": 10194 + }, + { + "epoch": 1.8151709401709402, + "grad_norm": 0.8196593523025513, + "learning_rate": 0.0001146444895199561, + "loss": 0.9148, + "step": 10195 + }, + { + "epoch": 1.8153490028490027, + "grad_norm": 0.6394698023796082, + "learning_rate": 0.00011463064281350204, + "loss": 0.9781, + "step": 10196 + }, + { + "epoch": 1.8155270655270654, + "grad_norm": 0.7302836775779724, + "learning_rate": 0.00011461679582039091, + "loss": 1.0394, + "step": 10197 + }, + { + "epoch": 1.8157051282051282, + "grad_norm": 0.7066670060157776, + "learning_rate": 0.00011460294854089404, + "loss": 1.1153, + "step": 10198 + }, + { + "epoch": 1.815883190883191, + "grad_norm": 0.6471068263053894, + "learning_rate": 0.0001145891009752827, + "loss": 1.1533, + "step": 10199 + }, + { + "epoch": 1.8160612535612537, + "grad_norm": 0.6842355132102966, + "learning_rate": 0.00011457525312382826, + "loss": 0.953, + "step": 10200 + }, + { + "epoch": 1.8162393162393162, + "grad_norm": 0.6720319986343384, + "learning_rate": 0.00011456140498680202, + "loss": 1.003, + "step": 10201 + }, + { + "epoch": 1.8164173789173788, + "grad_norm": 0.632017970085144, + "learning_rate": 0.00011454755656447527, + "loss": 0.8148, + "step": 10202 + }, + { + "epoch": 1.8165954415954415, + "grad_norm": 0.7193828225135803, + "learning_rate": 0.00011453370785711939, + "loss": 1.0098, + "step": 10203 + }, + { + "epoch": 1.8167735042735043, + "grad_norm": 0.7098045349121094, + "learning_rate": 0.00011451985886500566, + "loss": 1.1276, + "step": 10204 + }, + { + "epoch": 1.816951566951567, + "grad_norm": 0.7076733708381653, + "learning_rate": 0.00011450600958840547, + "loss": 1.1216, + "step": 10205 + }, + { + "epoch": 1.8171296296296298, + "grad_norm": 0.6864610314369202, + "learning_rate": 0.00011449216002759018, + "loss": 0.9896, + "step": 10206 + }, + { + "epoch": 1.8173076923076923, + "grad_norm": 0.737727701663971, + "learning_rate": 0.0001144783101828311, + "loss": 0.9447, + "step": 10207 + }, + { + "epoch": 1.8174857549857548, + "grad_norm": 0.6562525033950806, + "learning_rate": 0.00011446446005439964, + "loss": 1.1208, + "step": 10208 + }, + { + "epoch": 1.8176638176638176, + "grad_norm": 0.7203826308250427, + "learning_rate": 0.0001144506096425671, + "loss": 1.1339, + "step": 10209 + }, + { + "epoch": 1.8178418803418803, + "grad_norm": 0.6657233834266663, + "learning_rate": 0.00011443675894760489, + "loss": 0.8307, + "step": 10210 + }, + { + "epoch": 1.818019943019943, + "grad_norm": 0.7032586932182312, + "learning_rate": 0.00011442290796978437, + "loss": 0.8546, + "step": 10211 + }, + { + "epoch": 1.8181980056980058, + "grad_norm": 0.6989460587501526, + "learning_rate": 0.00011440905670937696, + "loss": 1.0749, + "step": 10212 + }, + { + "epoch": 1.8183760683760684, + "grad_norm": 0.6461085677146912, + "learning_rate": 0.00011439520516665399, + "loss": 0.984, + "step": 10213 + }, + { + "epoch": 1.818554131054131, + "grad_norm": 0.7077372670173645, + "learning_rate": 0.00011438135334188689, + "loss": 1.0813, + "step": 10214 + }, + { + "epoch": 1.8187321937321936, + "grad_norm": 0.6724075675010681, + "learning_rate": 0.00011436750123534704, + "loss": 0.9975, + "step": 10215 + }, + { + "epoch": 1.8189102564102564, + "grad_norm": 0.6205753684043884, + "learning_rate": 0.00011435364884730583, + "loss": 0.7414, + "step": 10216 + }, + { + "epoch": 1.8190883190883191, + "grad_norm": 0.6416093707084656, + "learning_rate": 0.00011433979617803472, + "loss": 1.0024, + "step": 10217 + }, + { + "epoch": 1.819266381766382, + "grad_norm": 0.7817183136940002, + "learning_rate": 0.00011432594322780508, + "loss": 1.0577, + "step": 10218 + }, + { + "epoch": 1.8194444444444444, + "grad_norm": 0.688220202922821, + "learning_rate": 0.00011431208999688835, + "loss": 1.0301, + "step": 10219 + }, + { + "epoch": 1.8196225071225072, + "grad_norm": 0.6464754343032837, + "learning_rate": 0.0001142982364855559, + "loss": 1.0608, + "step": 10220 + }, + { + "epoch": 1.8198005698005697, + "grad_norm": 0.6607306599617004, + "learning_rate": 0.00011428438269407926, + "loss": 1.1203, + "step": 10221 + }, + { + "epoch": 1.8199786324786325, + "grad_norm": 0.5779942870140076, + "learning_rate": 0.00011427052862272982, + "loss": 0.7895, + "step": 10222 + }, + { + "epoch": 1.8201566951566952, + "grad_norm": 0.7599068880081177, + "learning_rate": 0.000114256674271779, + "loss": 0.883, + "step": 10223 + }, + { + "epoch": 1.820334757834758, + "grad_norm": 0.6578865051269531, + "learning_rate": 0.00011424281964149824, + "loss": 1.101, + "step": 10224 + }, + { + "epoch": 1.8205128205128205, + "grad_norm": 0.7090746760368347, + "learning_rate": 0.00011422896473215905, + "loss": 0.9514, + "step": 10225 + }, + { + "epoch": 1.8206908831908832, + "grad_norm": 0.7537758946418762, + "learning_rate": 0.00011421510954403281, + "loss": 1.2193, + "step": 10226 + }, + { + "epoch": 1.8208689458689458, + "grad_norm": 0.670183002948761, + "learning_rate": 0.00011420125407739106, + "loss": 1.1408, + "step": 10227 + }, + { + "epoch": 1.8210470085470085, + "grad_norm": 0.742520809173584, + "learning_rate": 0.00011418739833250524, + "loss": 0.8826, + "step": 10228 + }, + { + "epoch": 1.8212250712250713, + "grad_norm": 0.6542800664901733, + "learning_rate": 0.00011417354230964683, + "loss": 1.0039, + "step": 10229 + }, + { + "epoch": 1.821403133903134, + "grad_norm": 0.6713709235191345, + "learning_rate": 0.00011415968600908727, + "loss": 0.9351, + "step": 10230 + }, + { + "epoch": 1.8215811965811965, + "grad_norm": 0.6794951558113098, + "learning_rate": 0.0001141458294310981, + "loss": 0.9491, + "step": 10231 + }, + { + "epoch": 1.8217592592592593, + "grad_norm": 0.6921972632408142, + "learning_rate": 0.00011413197257595079, + "loss": 1.1342, + "step": 10232 + }, + { + "epoch": 1.8219373219373218, + "grad_norm": 0.702586829662323, + "learning_rate": 0.00011411811544391682, + "loss": 0.9992, + "step": 10233 + }, + { + "epoch": 1.8221153846153846, + "grad_norm": 0.8147975206375122, + "learning_rate": 0.00011410425803526772, + "loss": 1.0507, + "step": 10234 + }, + { + "epoch": 1.8222934472934473, + "grad_norm": 0.66419517993927, + "learning_rate": 0.00011409040035027496, + "loss": 1.0426, + "step": 10235 + }, + { + "epoch": 1.82247150997151, + "grad_norm": 0.6132485866546631, + "learning_rate": 0.00011407654238921011, + "loss": 0.9859, + "step": 10236 + }, + { + "epoch": 1.8226495726495726, + "grad_norm": 0.7522366046905518, + "learning_rate": 0.00011406268415234462, + "loss": 0.9379, + "step": 10237 + }, + { + "epoch": 1.8228276353276354, + "grad_norm": 0.6335554122924805, + "learning_rate": 0.00011404882563995007, + "loss": 0.9322, + "step": 10238 + }, + { + "epoch": 1.823005698005698, + "grad_norm": 0.7577497363090515, + "learning_rate": 0.00011403496685229797, + "loss": 1.1383, + "step": 10239 + }, + { + "epoch": 1.8231837606837606, + "grad_norm": 0.6796886920928955, + "learning_rate": 0.00011402110778965982, + "loss": 1.0092, + "step": 10240 + }, + { + "epoch": 1.8233618233618234, + "grad_norm": 0.7676617503166199, + "learning_rate": 0.0001140072484523072, + "loss": 1.0137, + "step": 10241 + }, + { + "epoch": 1.8235398860398861, + "grad_norm": 0.7807821035385132, + "learning_rate": 0.00011399338884051165, + "loss": 0.8987, + "step": 10242 + }, + { + "epoch": 1.8237179487179487, + "grad_norm": 0.7169568538665771, + "learning_rate": 0.00011397952895454473, + "loss": 0.8984, + "step": 10243 + }, + { + "epoch": 1.8238960113960114, + "grad_norm": 0.6564654111862183, + "learning_rate": 0.00011396566879467793, + "loss": 1.0255, + "step": 10244 + }, + { + "epoch": 1.824074074074074, + "grad_norm": 0.7290034294128418, + "learning_rate": 0.00011395180836118292, + "loss": 0.9962, + "step": 10245 + }, + { + "epoch": 1.8242521367521367, + "grad_norm": 0.6610758900642395, + "learning_rate": 0.00011393794765433115, + "loss": 1.102, + "step": 10246 + }, + { + "epoch": 1.8244301994301995, + "grad_norm": 0.6875932216644287, + "learning_rate": 0.0001139240866743943, + "loss": 0.9963, + "step": 10247 + }, + { + "epoch": 1.8246082621082622, + "grad_norm": 0.7595645189285278, + "learning_rate": 0.00011391022542164387, + "loss": 1.1285, + "step": 10248 + }, + { + "epoch": 1.8247863247863247, + "grad_norm": 0.6752721667289734, + "learning_rate": 0.0001138963638963515, + "loss": 0.9447, + "step": 10249 + }, + { + "epoch": 1.8249643874643875, + "grad_norm": 0.6697955131530762, + "learning_rate": 0.00011388250209878873, + "loss": 1.0804, + "step": 10250 + }, + { + "epoch": 1.82514245014245, + "grad_norm": 0.6546956896781921, + "learning_rate": 0.00011386864002922713, + "loss": 0.9626, + "step": 10251 + }, + { + "epoch": 1.8253205128205128, + "grad_norm": 0.8002896904945374, + "learning_rate": 0.00011385477768793838, + "loss": 1.1933, + "step": 10252 + }, + { + "epoch": 1.8254985754985755, + "grad_norm": 0.6566781401634216, + "learning_rate": 0.00011384091507519403, + "loss": 0.9802, + "step": 10253 + }, + { + "epoch": 1.8256766381766383, + "grad_norm": 0.617420494556427, + "learning_rate": 0.00011382705219126572, + "loss": 1.1098, + "step": 10254 + }, + { + "epoch": 1.8258547008547008, + "grad_norm": 0.6558036208152771, + "learning_rate": 0.00011381318903642504, + "loss": 1.0291, + "step": 10255 + }, + { + "epoch": 1.8260327635327636, + "grad_norm": 0.6295637488365173, + "learning_rate": 0.00011379932561094358, + "loss": 1.0792, + "step": 10256 + }, + { + "epoch": 1.826210826210826, + "grad_norm": 0.7475154399871826, + "learning_rate": 0.00011378546191509303, + "loss": 1.1362, + "step": 10257 + }, + { + "epoch": 1.8263888888888888, + "grad_norm": 0.6814939379692078, + "learning_rate": 0.00011377159794914498, + "loss": 0.9131, + "step": 10258 + }, + { + "epoch": 1.8265669515669516, + "grad_norm": 0.6726876497268677, + "learning_rate": 0.00011375773371337111, + "loss": 0.9147, + "step": 10259 + }, + { + "epoch": 1.8267450142450143, + "grad_norm": 0.785943865776062, + "learning_rate": 0.00011374386920804298, + "loss": 1.0137, + "step": 10260 + }, + { + "epoch": 1.8269230769230769, + "grad_norm": 0.7614478468894958, + "learning_rate": 0.0001137300044334323, + "loss": 1.2118, + "step": 10261 + }, + { + "epoch": 1.8271011396011396, + "grad_norm": 0.7317564487457275, + "learning_rate": 0.00011371613938981072, + "loss": 1.0602, + "step": 10262 + }, + { + "epoch": 1.8272792022792022, + "grad_norm": 0.6716432571411133, + "learning_rate": 0.00011370227407744987, + "loss": 0.952, + "step": 10263 + }, + { + "epoch": 1.827457264957265, + "grad_norm": 0.6946425437927246, + "learning_rate": 0.00011368840849662139, + "loss": 1.0554, + "step": 10264 + }, + { + "epoch": 1.8276353276353277, + "grad_norm": 0.6692264080047607, + "learning_rate": 0.00011367454264759703, + "loss": 0.8944, + "step": 10265 + }, + { + "epoch": 1.8278133903133904, + "grad_norm": 0.6931505799293518, + "learning_rate": 0.00011366067653064838, + "loss": 0.9045, + "step": 10266 + }, + { + "epoch": 1.827991452991453, + "grad_norm": 0.7233194708824158, + "learning_rate": 0.00011364681014604716, + "loss": 0.9441, + "step": 10267 + }, + { + "epoch": 1.8281695156695157, + "grad_norm": 0.6451242566108704, + "learning_rate": 0.00011363294349406506, + "loss": 0.9948, + "step": 10268 + }, + { + "epoch": 1.8283475783475782, + "grad_norm": 0.6993351578712463, + "learning_rate": 0.00011361907657497375, + "loss": 1.1057, + "step": 10269 + }, + { + "epoch": 1.828525641025641, + "grad_norm": 0.7241137623786926, + "learning_rate": 0.00011360520938904493, + "loss": 0.974, + "step": 10270 + }, + { + "epoch": 1.8287037037037037, + "grad_norm": 0.6349480152130127, + "learning_rate": 0.00011359134193655027, + "loss": 0.9026, + "step": 10271 + }, + { + "epoch": 1.8288817663817665, + "grad_norm": 0.6916826963424683, + "learning_rate": 0.00011357747421776151, + "loss": 0.9153, + "step": 10272 + }, + { + "epoch": 1.8290598290598292, + "grad_norm": 0.879770040512085, + "learning_rate": 0.00011356360623295037, + "loss": 1.0818, + "step": 10273 + }, + { + "epoch": 1.8292378917378918, + "grad_norm": 0.6293807029724121, + "learning_rate": 0.00011354973798238853, + "loss": 1.1164, + "step": 10274 + }, + { + "epoch": 1.8294159544159543, + "grad_norm": 0.7070622444152832, + "learning_rate": 0.0001135358694663477, + "loss": 0.8795, + "step": 10275 + }, + { + "epoch": 1.829594017094017, + "grad_norm": 0.6847673654556274, + "learning_rate": 0.00011352200068509962, + "loss": 0.9173, + "step": 10276 + }, + { + "epoch": 1.8297720797720798, + "grad_norm": 0.6552146077156067, + "learning_rate": 0.00011350813163891605, + "loss": 1.0425, + "step": 10277 + }, + { + "epoch": 1.8299501424501425, + "grad_norm": 0.6432808041572571, + "learning_rate": 0.0001134942623280687, + "loss": 0.9418, + "step": 10278 + }, + { + "epoch": 1.8301282051282053, + "grad_norm": 0.7412393093109131, + "learning_rate": 0.00011348039275282931, + "loss": 1.1212, + "step": 10279 + }, + { + "epoch": 1.8303062678062678, + "grad_norm": 0.6543423533439636, + "learning_rate": 0.00011346652291346965, + "loss": 1.0553, + "step": 10280 + }, + { + "epoch": 1.8304843304843303, + "grad_norm": 0.7159286141395569, + "learning_rate": 0.00011345265281026138, + "loss": 1.0582, + "step": 10281 + }, + { + "epoch": 1.830662393162393, + "grad_norm": 0.6443323493003845, + "learning_rate": 0.00011343878244347639, + "loss": 0.9462, + "step": 10282 + }, + { + "epoch": 1.8308404558404558, + "grad_norm": 0.7592014074325562, + "learning_rate": 0.00011342491181338634, + "loss": 1.2718, + "step": 10283 + }, + { + "epoch": 1.8310185185185186, + "grad_norm": 0.627109944820404, + "learning_rate": 0.00011341104092026302, + "loss": 1.0177, + "step": 10284 + }, + { + "epoch": 1.8311965811965814, + "grad_norm": 0.8061598539352417, + "learning_rate": 0.00011339716976437827, + "loss": 0.9416, + "step": 10285 + }, + { + "epoch": 1.8313746438746439, + "grad_norm": 0.6584261059761047, + "learning_rate": 0.00011338329834600377, + "loss": 0.8297, + "step": 10286 + }, + { + "epoch": 1.8315527065527064, + "grad_norm": 0.6329470276832581, + "learning_rate": 0.00011336942666541133, + "loss": 0.8386, + "step": 10287 + }, + { + "epoch": 1.8317307692307692, + "grad_norm": 0.6833979487419128, + "learning_rate": 0.00011335555472287275, + "loss": 0.9407, + "step": 10288 + }, + { + "epoch": 1.831908831908832, + "grad_norm": 0.7663840651512146, + "learning_rate": 0.00011334168251865985, + "loss": 1.0018, + "step": 10289 + }, + { + "epoch": 1.8320868945868947, + "grad_norm": 0.7751262784004211, + "learning_rate": 0.00011332781005304436, + "loss": 1.0576, + "step": 10290 + }, + { + "epoch": 1.8322649572649574, + "grad_norm": 0.6857370138168335, + "learning_rate": 0.00011331393732629814, + "loss": 0.9888, + "step": 10291 + }, + { + "epoch": 1.83244301994302, + "grad_norm": 0.7534535527229309, + "learning_rate": 0.00011330006433869296, + "loss": 1.0834, + "step": 10292 + }, + { + "epoch": 1.8326210826210825, + "grad_norm": 0.6785250306129456, + "learning_rate": 0.00011328619109050065, + "loss": 1.0471, + "step": 10293 + }, + { + "epoch": 1.8327991452991452, + "grad_norm": 0.7023689150810242, + "learning_rate": 0.00011327231758199303, + "loss": 1.0652, + "step": 10294 + }, + { + "epoch": 1.832977207977208, + "grad_norm": 0.6776610612869263, + "learning_rate": 0.00011325844381344192, + "loss": 0.9504, + "step": 10295 + }, + { + "epoch": 1.8331552706552707, + "grad_norm": 0.7704112529754639, + "learning_rate": 0.00011324456978511917, + "loss": 0.9712, + "step": 10296 + }, + { + "epoch": 1.8333333333333335, + "grad_norm": 0.601502537727356, + "learning_rate": 0.00011323069549729654, + "loss": 1.075, + "step": 10297 + }, + { + "epoch": 1.833511396011396, + "grad_norm": 0.6282439231872559, + "learning_rate": 0.00011321682095024596, + "loss": 0.9238, + "step": 10298 + }, + { + "epoch": 1.8336894586894585, + "grad_norm": 0.6873499155044556, + "learning_rate": 0.00011320294614423921, + "loss": 1.0464, + "step": 10299 + }, + { + "epoch": 1.8338675213675213, + "grad_norm": 0.6063792705535889, + "learning_rate": 0.00011318907107954815, + "loss": 0.9732, + "step": 10300 + }, + { + "epoch": 1.834045584045584, + "grad_norm": 0.5830921530723572, + "learning_rate": 0.00011317519575644464, + "loss": 0.7568, + "step": 10301 + }, + { + "epoch": 1.8342236467236468, + "grad_norm": 0.6394222378730774, + "learning_rate": 0.00011316132017520053, + "loss": 0.9958, + "step": 10302 + }, + { + "epoch": 1.8344017094017095, + "grad_norm": 0.7052412033081055, + "learning_rate": 0.00011314744433608773, + "loss": 0.9129, + "step": 10303 + }, + { + "epoch": 1.834579772079772, + "grad_norm": 0.7287624478340149, + "learning_rate": 0.00011313356823937801, + "loss": 0.8608, + "step": 10304 + }, + { + "epoch": 1.8347578347578346, + "grad_norm": 0.702937662601471, + "learning_rate": 0.00011311969188534334, + "loss": 1.3074, + "step": 10305 + }, + { + "epoch": 1.8349358974358974, + "grad_norm": 0.6693850159645081, + "learning_rate": 0.00011310581527425557, + "loss": 0.928, + "step": 10306 + }, + { + "epoch": 1.83511396011396, + "grad_norm": 0.8153932094573975, + "learning_rate": 0.00011309193840638654, + "loss": 1.1771, + "step": 10307 + }, + { + "epoch": 1.8352920227920229, + "grad_norm": 0.6517418622970581, + "learning_rate": 0.00011307806128200821, + "loss": 0.9634, + "step": 10308 + }, + { + "epoch": 1.8354700854700856, + "grad_norm": 0.6626226305961609, + "learning_rate": 0.00011306418390139245, + "loss": 0.9371, + "step": 10309 + }, + { + "epoch": 1.8356481481481481, + "grad_norm": 0.7397477030754089, + "learning_rate": 0.0001130503062648111, + "loss": 0.9398, + "step": 10310 + }, + { + "epoch": 1.8358262108262107, + "grad_norm": 0.6790265440940857, + "learning_rate": 0.00011303642837253614, + "loss": 0.9728, + "step": 10311 + }, + { + "epoch": 1.8360042735042734, + "grad_norm": 0.6266449093818665, + "learning_rate": 0.00011302255022483941, + "loss": 0.847, + "step": 10312 + }, + { + "epoch": 1.8361823361823362, + "grad_norm": 0.791657030582428, + "learning_rate": 0.00011300867182199288, + "loss": 0.8342, + "step": 10313 + }, + { + "epoch": 1.836360398860399, + "grad_norm": 0.7128583788871765, + "learning_rate": 0.00011299479316426846, + "loss": 0.9591, + "step": 10314 + }, + { + "epoch": 1.8365384615384617, + "grad_norm": 0.659928023815155, + "learning_rate": 0.00011298091425193806, + "loss": 1.0282, + "step": 10315 + }, + { + "epoch": 1.8367165242165242, + "grad_norm": 0.6641396284103394, + "learning_rate": 0.00011296703508527363, + "loss": 1.0161, + "step": 10316 + }, + { + "epoch": 1.8368945868945867, + "grad_norm": 0.7921316027641296, + "learning_rate": 0.00011295315566454702, + "loss": 0.8897, + "step": 10317 + }, + { + "epoch": 1.8370726495726495, + "grad_norm": 0.6900694966316223, + "learning_rate": 0.00011293927599003029, + "loss": 1.0094, + "step": 10318 + }, + { + "epoch": 1.8372507122507122, + "grad_norm": 0.8054366707801819, + "learning_rate": 0.0001129253960619953, + "loss": 0.9489, + "step": 10319 + }, + { + "epoch": 1.837428774928775, + "grad_norm": 0.6623767018318176, + "learning_rate": 0.00011291151588071405, + "loss": 0.92, + "step": 10320 + }, + { + "epoch": 1.8376068376068377, + "grad_norm": 0.6143901348114014, + "learning_rate": 0.00011289763544645846, + "loss": 0.8093, + "step": 10321 + }, + { + "epoch": 1.8377849002849003, + "grad_norm": 0.8207027316093445, + "learning_rate": 0.00011288375475950046, + "loss": 1.2402, + "step": 10322 + }, + { + "epoch": 1.8379629629629628, + "grad_norm": 0.6759985685348511, + "learning_rate": 0.00011286987382011209, + "loss": 0.9179, + "step": 10323 + }, + { + "epoch": 1.8381410256410255, + "grad_norm": 0.745439887046814, + "learning_rate": 0.00011285599262856523, + "loss": 0.8157, + "step": 10324 + }, + { + "epoch": 1.8383190883190883, + "grad_norm": 0.6873317360877991, + "learning_rate": 0.00011284211118513194, + "loss": 0.8681, + "step": 10325 + }, + { + "epoch": 1.838497150997151, + "grad_norm": 0.7060160040855408, + "learning_rate": 0.00011282822949008416, + "loss": 1.0833, + "step": 10326 + }, + { + "epoch": 1.8386752136752138, + "grad_norm": 0.8079642653465271, + "learning_rate": 0.00011281434754369389, + "loss": 0.8639, + "step": 10327 + }, + { + "epoch": 1.8388532763532763, + "grad_norm": 0.6434001922607422, + "learning_rate": 0.00011280046534623303, + "loss": 0.9269, + "step": 10328 + }, + { + "epoch": 1.839031339031339, + "grad_norm": 0.7005292773246765, + "learning_rate": 0.0001127865828979737, + "loss": 1.1475, + "step": 10329 + }, + { + "epoch": 1.8392094017094016, + "grad_norm": 0.7004852890968323, + "learning_rate": 0.00011277270019918784, + "loss": 0.9467, + "step": 10330 + }, + { + "epoch": 1.8393874643874644, + "grad_norm": 0.7542549967765808, + "learning_rate": 0.00011275881725014743, + "loss": 1.0371, + "step": 10331 + }, + { + "epoch": 1.8395655270655271, + "grad_norm": 0.674051821231842, + "learning_rate": 0.00011274493405112452, + "loss": 1.1097, + "step": 10332 + }, + { + "epoch": 1.8397435897435899, + "grad_norm": 0.8136405348777771, + "learning_rate": 0.00011273105060239107, + "loss": 0.9718, + "step": 10333 + }, + { + "epoch": 1.8399216524216524, + "grad_norm": 0.6524073481559753, + "learning_rate": 0.00011271716690421916, + "loss": 0.9953, + "step": 10334 + }, + { + "epoch": 1.8400997150997151, + "grad_norm": 0.7436625957489014, + "learning_rate": 0.00011270328295688077, + "loss": 1.0722, + "step": 10335 + }, + { + "epoch": 1.8402777777777777, + "grad_norm": 0.6815723180770874, + "learning_rate": 0.00011268939876064795, + "loss": 1.0924, + "step": 10336 + }, + { + "epoch": 1.8404558404558404, + "grad_norm": 0.6923388242721558, + "learning_rate": 0.0001126755143157927, + "loss": 0.921, + "step": 10337 + }, + { + "epoch": 1.8406339031339032, + "grad_norm": 0.7464849948883057, + "learning_rate": 0.00011266162962258708, + "loss": 1.0549, + "step": 10338 + }, + { + "epoch": 1.840811965811966, + "grad_norm": 0.6621805429458618, + "learning_rate": 0.00011264774468130315, + "loss": 1.0764, + "step": 10339 + }, + { + "epoch": 1.8409900284900285, + "grad_norm": 0.7370132803916931, + "learning_rate": 0.00011263385949221295, + "loss": 0.7818, + "step": 10340 + }, + { + "epoch": 1.8411680911680912, + "grad_norm": 0.673100471496582, + "learning_rate": 0.00011261997405558848, + "loss": 1.04, + "step": 10341 + }, + { + "epoch": 1.8413461538461537, + "grad_norm": 0.5978201031684875, + "learning_rate": 0.00011260608837170183, + "loss": 0.9644, + "step": 10342 + }, + { + "epoch": 1.8415242165242165, + "grad_norm": 0.6868628263473511, + "learning_rate": 0.00011259220244082507, + "loss": 0.9533, + "step": 10343 + }, + { + "epoch": 1.8417022792022792, + "grad_norm": 0.6580314636230469, + "learning_rate": 0.0001125783162632303, + "loss": 0.9506, + "step": 10344 + }, + { + "epoch": 1.841880341880342, + "grad_norm": 0.7238291501998901, + "learning_rate": 0.00011256442983918951, + "loss": 0.8663, + "step": 10345 + }, + { + "epoch": 1.8420584045584045, + "grad_norm": 0.5838520526885986, + "learning_rate": 0.00011255054316897484, + "loss": 0.9606, + "step": 10346 + }, + { + "epoch": 1.8422364672364673, + "grad_norm": 0.7102842926979065, + "learning_rate": 0.00011253665625285836, + "loss": 0.801, + "step": 10347 + }, + { + "epoch": 1.8424145299145298, + "grad_norm": 0.6449147462844849, + "learning_rate": 0.0001125227690911121, + "loss": 1.0827, + "step": 10348 + }, + { + "epoch": 1.8425925925925926, + "grad_norm": 0.6355304718017578, + "learning_rate": 0.00011250888168400823, + "loss": 1.0369, + "step": 10349 + }, + { + "epoch": 1.8427706552706553, + "grad_norm": 0.678977906703949, + "learning_rate": 0.0001124949940318188, + "loss": 0.9491, + "step": 10350 + }, + { + "epoch": 1.842948717948718, + "grad_norm": 0.6366633772850037, + "learning_rate": 0.00011248110613481592, + "loss": 0.7272, + "step": 10351 + }, + { + "epoch": 1.8431267806267806, + "grad_norm": 0.6639098525047302, + "learning_rate": 0.00011246721799327171, + "loss": 1.0313, + "step": 10352 + }, + { + "epoch": 1.8433048433048433, + "grad_norm": 0.6034720540046692, + "learning_rate": 0.00011245332960745822, + "loss": 0.7141, + "step": 10353 + }, + { + "epoch": 1.8434829059829059, + "grad_norm": 0.8118346333503723, + "learning_rate": 0.00011243944097764763, + "loss": 1.171, + "step": 10354 + }, + { + "epoch": 1.8436609686609686, + "grad_norm": 0.6706618070602417, + "learning_rate": 0.00011242555210411203, + "loss": 0.9578, + "step": 10355 + }, + { + "epoch": 1.8438390313390314, + "grad_norm": 0.619562029838562, + "learning_rate": 0.00011241166298712355, + "loss": 0.9883, + "step": 10356 + }, + { + "epoch": 1.8440170940170941, + "grad_norm": 0.6471936106681824, + "learning_rate": 0.00011239777362695434, + "loss": 0.8897, + "step": 10357 + }, + { + "epoch": 1.8441951566951567, + "grad_norm": 0.7179005742073059, + "learning_rate": 0.00011238388402387645, + "loss": 0.9646, + "step": 10358 + }, + { + "epoch": 1.8443732193732194, + "grad_norm": 0.7726966738700867, + "learning_rate": 0.00011236999417816214, + "loss": 0.8855, + "step": 10359 + }, + { + "epoch": 1.844551282051282, + "grad_norm": 0.6733565330505371, + "learning_rate": 0.00011235610409008346, + "loss": 1.0379, + "step": 10360 + }, + { + "epoch": 1.8447293447293447, + "grad_norm": 0.7317814826965332, + "learning_rate": 0.0001123422137599126, + "loss": 0.8528, + "step": 10361 + }, + { + "epoch": 1.8449074074074074, + "grad_norm": 0.6727005839347839, + "learning_rate": 0.0001123283231879217, + "loss": 0.9612, + "step": 10362 + }, + { + "epoch": 1.8450854700854702, + "grad_norm": 0.6350542306900024, + "learning_rate": 0.00011231443237438289, + "loss": 0.9939, + "step": 10363 + }, + { + "epoch": 1.8452635327635327, + "grad_norm": 0.693148672580719, + "learning_rate": 0.00011230054131956836, + "loss": 1.0149, + "step": 10364 + }, + { + "epoch": 1.8454415954415955, + "grad_norm": 0.7263579368591309, + "learning_rate": 0.0001122866500237503, + "loss": 1.1044, + "step": 10365 + }, + { + "epoch": 1.845619658119658, + "grad_norm": 0.7044230699539185, + "learning_rate": 0.00011227275848720085, + "loss": 1.0677, + "step": 10366 + }, + { + "epoch": 1.8457977207977208, + "grad_norm": 0.6895326972007751, + "learning_rate": 0.00011225886671019219, + "loss": 1.1025, + "step": 10367 + }, + { + "epoch": 1.8459757834757835, + "grad_norm": 0.6045145988464355, + "learning_rate": 0.00011224497469299651, + "loss": 0.8079, + "step": 10368 + }, + { + "epoch": 1.8461538461538463, + "grad_norm": 0.6613210439682007, + "learning_rate": 0.00011223108243588599, + "loss": 1.0345, + "step": 10369 + }, + { + "epoch": 1.8463319088319088, + "grad_norm": 0.6288960576057434, + "learning_rate": 0.0001122171899391328, + "loss": 1.0166, + "step": 10370 + }, + { + "epoch": 1.8465099715099715, + "grad_norm": 0.6158748865127563, + "learning_rate": 0.00011220329720300917, + "loss": 0.895, + "step": 10371 + }, + { + "epoch": 1.846688034188034, + "grad_norm": 0.6583057641983032, + "learning_rate": 0.00011218940422778728, + "loss": 0.8059, + "step": 10372 + }, + { + "epoch": 1.8468660968660968, + "grad_norm": 0.6761550903320312, + "learning_rate": 0.00011217551101373932, + "loss": 0.9253, + "step": 10373 + }, + { + "epoch": 1.8470441595441596, + "grad_norm": 0.5969263315200806, + "learning_rate": 0.0001121616175611375, + "loss": 0.8549, + "step": 10374 + }, + { + "epoch": 1.8472222222222223, + "grad_norm": 0.7994722723960876, + "learning_rate": 0.00011214772387025407, + "loss": 0.9918, + "step": 10375 + }, + { + "epoch": 1.8474002849002849, + "grad_norm": 0.6949167847633362, + "learning_rate": 0.00011213382994136123, + "loss": 1.1853, + "step": 10376 + }, + { + "epoch": 1.8475783475783476, + "grad_norm": 0.7356176376342773, + "learning_rate": 0.00011211993577473121, + "loss": 0.8809, + "step": 10377 + }, + { + "epoch": 1.8477564102564101, + "grad_norm": 0.7110268473625183, + "learning_rate": 0.0001121060413706362, + "loss": 0.9805, + "step": 10378 + }, + { + "epoch": 1.8479344729344729, + "grad_norm": 0.6509962677955627, + "learning_rate": 0.00011209214672934846, + "loss": 0.8899, + "step": 10379 + }, + { + "epoch": 1.8481125356125356, + "grad_norm": 0.6103082299232483, + "learning_rate": 0.00011207825185114025, + "loss": 0.8576, + "step": 10380 + }, + { + "epoch": 1.8482905982905984, + "grad_norm": 0.6261070966720581, + "learning_rate": 0.00011206435673628377, + "loss": 0.8884, + "step": 10381 + }, + { + "epoch": 1.848468660968661, + "grad_norm": 0.7629222273826599, + "learning_rate": 0.00011205046138505126, + "loss": 1.1714, + "step": 10382 + }, + { + "epoch": 1.8486467236467237, + "grad_norm": 0.617957353591919, + "learning_rate": 0.000112036565797715, + "loss": 0.9546, + "step": 10383 + }, + { + "epoch": 1.8488247863247862, + "grad_norm": 0.6926987171173096, + "learning_rate": 0.00011202266997454724, + "loss": 0.8842, + "step": 10384 + }, + { + "epoch": 1.849002849002849, + "grad_norm": 0.602758526802063, + "learning_rate": 0.00011200877391582025, + "loss": 0.9782, + "step": 10385 + }, + { + "epoch": 1.8491809116809117, + "grad_norm": 0.706731915473938, + "learning_rate": 0.00011199487762180627, + "loss": 0.8176, + "step": 10386 + }, + { + "epoch": 1.8493589743589745, + "grad_norm": 0.7135118842124939, + "learning_rate": 0.0001119809810927776, + "loss": 0.9277, + "step": 10387 + }, + { + "epoch": 1.8495370370370372, + "grad_norm": 0.7484592199325562, + "learning_rate": 0.00011196708432900647, + "loss": 1.0733, + "step": 10388 + }, + { + "epoch": 1.8497150997150997, + "grad_norm": 0.7087157964706421, + "learning_rate": 0.00011195318733076519, + "loss": 0.9443, + "step": 10389 + }, + { + "epoch": 1.8498931623931623, + "grad_norm": 0.6511468291282654, + "learning_rate": 0.00011193929009832602, + "loss": 0.955, + "step": 10390 + }, + { + "epoch": 1.850071225071225, + "grad_norm": 0.6386628746986389, + "learning_rate": 0.0001119253926319613, + "loss": 1.0357, + "step": 10391 + }, + { + "epoch": 1.8502492877492878, + "grad_norm": 0.6400021314620972, + "learning_rate": 0.00011191149493194327, + "loss": 0.8094, + "step": 10392 + }, + { + "epoch": 1.8504273504273505, + "grad_norm": 0.7942537069320679, + "learning_rate": 0.00011189759699854423, + "loss": 0.9717, + "step": 10393 + }, + { + "epoch": 1.8506054131054133, + "grad_norm": 0.7230474948883057, + "learning_rate": 0.00011188369883203647, + "loss": 0.9043, + "step": 10394 + }, + { + "epoch": 1.8507834757834758, + "grad_norm": 0.8837162852287292, + "learning_rate": 0.00011186980043269235, + "loss": 1.2821, + "step": 10395 + }, + { + "epoch": 1.8509615384615383, + "grad_norm": 0.7260291576385498, + "learning_rate": 0.00011185590180078413, + "loss": 1.1672, + "step": 10396 + }, + { + "epoch": 1.851139601139601, + "grad_norm": 0.6290066242218018, + "learning_rate": 0.00011184200293658415, + "loss": 0.8942, + "step": 10397 + }, + { + "epoch": 1.8513176638176638, + "grad_norm": 0.6571013331413269, + "learning_rate": 0.00011182810384036475, + "loss": 1.0753, + "step": 10398 + }, + { + "epoch": 1.8514957264957266, + "grad_norm": 0.6494737267494202, + "learning_rate": 0.00011181420451239817, + "loss": 0.8833, + "step": 10399 + }, + { + "epoch": 1.8516737891737893, + "grad_norm": 0.7383694648742676, + "learning_rate": 0.00011180030495295684, + "loss": 1.0094, + "step": 10400 + }, + { + "epoch": 1.8518518518518519, + "grad_norm": 0.6713876724243164, + "learning_rate": 0.00011178640516231302, + "loss": 0.975, + "step": 10401 + }, + { + "epoch": 1.8520299145299144, + "grad_norm": 0.8041042685508728, + "learning_rate": 0.00011177250514073912, + "loss": 1.1419, + "step": 10402 + }, + { + "epoch": 1.8522079772079771, + "grad_norm": 0.7035061120986938, + "learning_rate": 0.00011175860488850738, + "loss": 1.0921, + "step": 10403 + }, + { + "epoch": 1.85238603988604, + "grad_norm": 0.6135673522949219, + "learning_rate": 0.00011174470440589022, + "loss": 0.9611, + "step": 10404 + }, + { + "epoch": 1.8525641025641026, + "grad_norm": 0.7868386507034302, + "learning_rate": 0.00011173080369315999, + "loss": 0.8561, + "step": 10405 + }, + { + "epoch": 1.8527421652421654, + "grad_norm": 0.6575735211372375, + "learning_rate": 0.00011171690275058902, + "loss": 1.0256, + "step": 10406 + }, + { + "epoch": 1.852920227920228, + "grad_norm": 0.7514392137527466, + "learning_rate": 0.00011170300157844969, + "loss": 1.0868, + "step": 10407 + }, + { + "epoch": 1.8530982905982905, + "grad_norm": 0.6915257573127747, + "learning_rate": 0.00011168910017701436, + "loss": 1.1223, + "step": 10408 + }, + { + "epoch": 1.8532763532763532, + "grad_norm": 0.7406772971153259, + "learning_rate": 0.00011167519854655535, + "loss": 1.0922, + "step": 10409 + }, + { + "epoch": 1.853454415954416, + "grad_norm": 0.6632742881774902, + "learning_rate": 0.0001116612966873451, + "loss": 0.9082, + "step": 10410 + }, + { + "epoch": 1.8536324786324787, + "grad_norm": 0.8154461979866028, + "learning_rate": 0.00011164739459965598, + "loss": 1.1126, + "step": 10411 + }, + { + "epoch": 1.8538105413105415, + "grad_norm": 0.895764172077179, + "learning_rate": 0.00011163349228376037, + "loss": 1.0589, + "step": 10412 + }, + { + "epoch": 1.853988603988604, + "grad_norm": 0.6746504902839661, + "learning_rate": 0.00011161958973993063, + "loss": 1.0184, + "step": 10413 + }, + { + "epoch": 1.8541666666666665, + "grad_norm": 0.7271263003349304, + "learning_rate": 0.00011160568696843916, + "loss": 0.9989, + "step": 10414 + }, + { + "epoch": 1.8543447293447293, + "grad_norm": 0.7503132820129395, + "learning_rate": 0.00011159178396955836, + "loss": 1.0783, + "step": 10415 + }, + { + "epoch": 1.854522792022792, + "grad_norm": 0.6768177151679993, + "learning_rate": 0.00011157788074356066, + "loss": 0.9916, + "step": 10416 + }, + { + "epoch": 1.8547008547008548, + "grad_norm": 0.6804978251457214, + "learning_rate": 0.00011156397729071842, + "loss": 0.9534, + "step": 10417 + }, + { + "epoch": 1.8548789173789175, + "grad_norm": 0.7144617438316345, + "learning_rate": 0.00011155007361130408, + "loss": 0.991, + "step": 10418 + }, + { + "epoch": 1.85505698005698, + "grad_norm": 0.6816750168800354, + "learning_rate": 0.00011153616970559, + "loss": 0.9551, + "step": 10419 + }, + { + "epoch": 1.8552350427350426, + "grad_norm": 0.6620030999183655, + "learning_rate": 0.00011152226557384866, + "loss": 0.8854, + "step": 10420 + }, + { + "epoch": 1.8554131054131053, + "grad_norm": 0.8400058746337891, + "learning_rate": 0.00011150836121635249, + "loss": 1.1593, + "step": 10421 + }, + { + "epoch": 1.855591168091168, + "grad_norm": 0.6666815280914307, + "learning_rate": 0.00011149445663337385, + "loss": 1.2112, + "step": 10422 + }, + { + "epoch": 1.8557692307692308, + "grad_norm": 0.7298431396484375, + "learning_rate": 0.00011148055182518522, + "loss": 0.9721, + "step": 10423 + }, + { + "epoch": 1.8559472934472936, + "grad_norm": 0.66816645860672, + "learning_rate": 0.00011146664679205903, + "loss": 1.0945, + "step": 10424 + }, + { + "epoch": 1.8561253561253561, + "grad_norm": 0.5979483127593994, + "learning_rate": 0.00011145274153426771, + "loss": 1.0176, + "step": 10425 + }, + { + "epoch": 1.8563034188034186, + "grad_norm": 0.6579445600509644, + "learning_rate": 0.00011143883605208372, + "loss": 0.9143, + "step": 10426 + }, + { + "epoch": 1.8564814814814814, + "grad_norm": 0.6871697902679443, + "learning_rate": 0.0001114249303457795, + "loss": 1.071, + "step": 10427 + }, + { + "epoch": 1.8566595441595442, + "grad_norm": 0.6683333516120911, + "learning_rate": 0.0001114110244156275, + "loss": 0.7809, + "step": 10428 + }, + { + "epoch": 1.856837606837607, + "grad_norm": 0.6122907996177673, + "learning_rate": 0.0001113971182619002, + "loss": 0.8329, + "step": 10429 + }, + { + "epoch": 1.8570156695156697, + "grad_norm": 0.6510575413703918, + "learning_rate": 0.00011138321188487, + "loss": 1.0068, + "step": 10430 + }, + { + "epoch": 1.8571937321937322, + "grad_norm": 0.6417793035507202, + "learning_rate": 0.00011136930528480945, + "loss": 1.0093, + "step": 10431 + }, + { + "epoch": 1.8573717948717947, + "grad_norm": 0.595824658870697, + "learning_rate": 0.00011135539846199096, + "loss": 0.9856, + "step": 10432 + }, + { + "epoch": 1.8575498575498575, + "grad_norm": 0.7594470381736755, + "learning_rate": 0.00011134149141668704, + "loss": 0.8173, + "step": 10433 + }, + { + "epoch": 1.8577279202279202, + "grad_norm": 0.7078324556350708, + "learning_rate": 0.00011132758414917016, + "loss": 1.0236, + "step": 10434 + }, + { + "epoch": 1.857905982905983, + "grad_norm": 0.6830437779426575, + "learning_rate": 0.00011131367665971275, + "loss": 0.8483, + "step": 10435 + }, + { + "epoch": 1.8580840455840457, + "grad_norm": 0.6856399774551392, + "learning_rate": 0.0001112997689485874, + "loss": 0.8729, + "step": 10436 + }, + { + "epoch": 1.8582621082621082, + "grad_norm": 0.6530426144599915, + "learning_rate": 0.00011128586101606653, + "loss": 0.8616, + "step": 10437 + }, + { + "epoch": 1.8584401709401708, + "grad_norm": 0.6341808438301086, + "learning_rate": 0.00011127195286242267, + "loss": 0.896, + "step": 10438 + }, + { + "epoch": 1.8586182336182335, + "grad_norm": 0.6278257966041565, + "learning_rate": 0.00011125804448792831, + "loss": 0.8309, + "step": 10439 + }, + { + "epoch": 1.8587962962962963, + "grad_norm": 0.708705723285675, + "learning_rate": 0.00011124413589285594, + "loss": 1.1065, + "step": 10440 + }, + { + "epoch": 1.858974358974359, + "grad_norm": 0.6845232248306274, + "learning_rate": 0.00011123022707747808, + "loss": 0.9292, + "step": 10441 + }, + { + "epoch": 1.8591524216524218, + "grad_norm": 0.749204695224762, + "learning_rate": 0.00011121631804206726, + "loss": 1.0487, + "step": 10442 + }, + { + "epoch": 1.8593304843304843, + "grad_norm": 0.7123128771781921, + "learning_rate": 0.00011120240878689599, + "loss": 0.9138, + "step": 10443 + }, + { + "epoch": 1.859508547008547, + "grad_norm": 0.6862115263938904, + "learning_rate": 0.00011118849931223679, + "loss": 1.0675, + "step": 10444 + }, + { + "epoch": 1.8596866096866096, + "grad_norm": 0.7245760560035706, + "learning_rate": 0.00011117458961836215, + "loss": 0.9643, + "step": 10445 + }, + { + "epoch": 1.8598646723646723, + "grad_norm": 0.701574444770813, + "learning_rate": 0.0001111606797055447, + "loss": 1.0022, + "step": 10446 + }, + { + "epoch": 1.860042735042735, + "grad_norm": 0.7292088270187378, + "learning_rate": 0.0001111467695740569, + "loss": 0.9465, + "step": 10447 + }, + { + "epoch": 1.8602207977207978, + "grad_norm": 0.7045044302940369, + "learning_rate": 0.0001111328592241713, + "loss": 1.0942, + "step": 10448 + }, + { + "epoch": 1.8603988603988604, + "grad_norm": 0.7181426286697388, + "learning_rate": 0.00011111894865616046, + "loss": 1.2108, + "step": 10449 + }, + { + "epoch": 1.8605769230769231, + "grad_norm": 0.6083306074142456, + "learning_rate": 0.00011110503787029689, + "loss": 0.929, + "step": 10450 + }, + { + "epoch": 1.8607549857549857, + "grad_norm": 0.6847347617149353, + "learning_rate": 0.00011109112686685319, + "loss": 1.0911, + "step": 10451 + }, + { + "epoch": 1.8609330484330484, + "grad_norm": 0.7131744027137756, + "learning_rate": 0.0001110772156461019, + "loss": 0.9649, + "step": 10452 + }, + { + "epoch": 1.8611111111111112, + "grad_norm": 0.7920312881469727, + "learning_rate": 0.00011106330420831559, + "loss": 0.9965, + "step": 10453 + }, + { + "epoch": 1.861289173789174, + "grad_norm": 0.6640987992286682, + "learning_rate": 0.00011104939255376681, + "loss": 1.2346, + "step": 10454 + }, + { + "epoch": 1.8614672364672364, + "grad_norm": 0.5878208875656128, + "learning_rate": 0.00011103548068272811, + "loss": 0.8565, + "step": 10455 + }, + { + "epoch": 1.8616452991452992, + "grad_norm": 0.6636882424354553, + "learning_rate": 0.0001110215685954721, + "loss": 0.8556, + "step": 10456 + }, + { + "epoch": 1.8618233618233617, + "grad_norm": 0.5985570549964905, + "learning_rate": 0.00011100765629227137, + "loss": 1.0291, + "step": 10457 + }, + { + "epoch": 1.8620014245014245, + "grad_norm": 0.7546643614768982, + "learning_rate": 0.00011099374377339846, + "loss": 1.0199, + "step": 10458 + }, + { + "epoch": 1.8621794871794872, + "grad_norm": 0.6529727578163147, + "learning_rate": 0.00011097983103912602, + "loss": 1.0826, + "step": 10459 + }, + { + "epoch": 1.86235754985755, + "grad_norm": 0.6394338607788086, + "learning_rate": 0.00011096591808972654, + "loss": 0.9896, + "step": 10460 + }, + { + "epoch": 1.8625356125356125, + "grad_norm": 0.6508805751800537, + "learning_rate": 0.00011095200492547271, + "loss": 0.9659, + "step": 10461 + }, + { + "epoch": 1.8627136752136753, + "grad_norm": 0.7085812091827393, + "learning_rate": 0.00011093809154663705, + "loss": 0.9998, + "step": 10462 + }, + { + "epoch": 1.8628917378917378, + "grad_norm": 0.6488457322120667, + "learning_rate": 0.00011092417795349226, + "loss": 0.9757, + "step": 10463 + }, + { + "epoch": 1.8630698005698005, + "grad_norm": 0.6405763626098633, + "learning_rate": 0.0001109102641463109, + "loss": 0.8188, + "step": 10464 + }, + { + "epoch": 1.8632478632478633, + "grad_norm": 0.713361918926239, + "learning_rate": 0.00011089635012536554, + "loss": 0.886, + "step": 10465 + }, + { + "epoch": 1.863425925925926, + "grad_norm": 0.5752255916595459, + "learning_rate": 0.00011088243589092886, + "loss": 1.0223, + "step": 10466 + }, + { + "epoch": 1.8636039886039886, + "grad_norm": 0.6722734570503235, + "learning_rate": 0.00011086852144327344, + "loss": 0.9499, + "step": 10467 + }, + { + "epoch": 1.8637820512820513, + "grad_norm": 0.5516420006752014, + "learning_rate": 0.00011085460678267194, + "loss": 0.7767, + "step": 10468 + }, + { + "epoch": 1.8639601139601139, + "grad_norm": 0.731257438659668, + "learning_rate": 0.00011084069190939697, + "loss": 1.2299, + "step": 10469 + }, + { + "epoch": 1.8641381766381766, + "grad_norm": 0.7977055907249451, + "learning_rate": 0.00011082677682372114, + "loss": 0.9109, + "step": 10470 + }, + { + "epoch": 1.8643162393162394, + "grad_norm": 0.679900586605072, + "learning_rate": 0.0001108128615259171, + "loss": 0.9319, + "step": 10471 + }, + { + "epoch": 1.864494301994302, + "grad_norm": 0.7428545951843262, + "learning_rate": 0.00011079894601625754, + "loss": 0.8585, + "step": 10472 + }, + { + "epoch": 1.8646723646723646, + "grad_norm": 0.6560967564582825, + "learning_rate": 0.00011078503029501504, + "loss": 1.0069, + "step": 10473 + }, + { + "epoch": 1.8648504273504274, + "grad_norm": 0.636202871799469, + "learning_rate": 0.00011077111436246228, + "loss": 1.0329, + "step": 10474 + }, + { + "epoch": 1.86502849002849, + "grad_norm": 0.6666205525398254, + "learning_rate": 0.00011075719821887191, + "loss": 1.0123, + "step": 10475 + }, + { + "epoch": 1.8652065527065527, + "grad_norm": 0.7089471220970154, + "learning_rate": 0.00011074328186451657, + "loss": 0.7851, + "step": 10476 + }, + { + "epoch": 1.8653846153846154, + "grad_norm": 0.6054788827896118, + "learning_rate": 0.00011072936529966895, + "loss": 0.8224, + "step": 10477 + }, + { + "epoch": 1.8655626780626782, + "grad_norm": 0.6009029150009155, + "learning_rate": 0.00011071544852460172, + "loss": 0.865, + "step": 10478 + }, + { + "epoch": 1.8657407407407407, + "grad_norm": 0.6238716244697571, + "learning_rate": 0.00011070153153958753, + "loss": 0.8685, + "step": 10479 + }, + { + "epoch": 1.8659188034188035, + "grad_norm": 0.719985842704773, + "learning_rate": 0.00011068761434489903, + "loss": 1.2204, + "step": 10480 + }, + { + "epoch": 1.866096866096866, + "grad_norm": 0.72972172498703, + "learning_rate": 0.00011067369694080895, + "loss": 1.0454, + "step": 10481 + }, + { + "epoch": 1.8662749287749287, + "grad_norm": 0.6741998791694641, + "learning_rate": 0.00011065977932758995, + "loss": 0.9992, + "step": 10482 + }, + { + "epoch": 1.8664529914529915, + "grad_norm": 0.6150268912315369, + "learning_rate": 0.00011064586150551472, + "loss": 0.8866, + "step": 10483 + }, + { + "epoch": 1.8666310541310542, + "grad_norm": 0.8253782391548157, + "learning_rate": 0.00011063194347485597, + "loss": 1.1173, + "step": 10484 + }, + { + "epoch": 1.8668091168091168, + "grad_norm": 0.7176247835159302, + "learning_rate": 0.00011061802523588636, + "loss": 1.0414, + "step": 10485 + }, + { + "epoch": 1.8669871794871795, + "grad_norm": 0.6372736096382141, + "learning_rate": 0.00011060410678887858, + "loss": 1.0548, + "step": 10486 + }, + { + "epoch": 1.867165242165242, + "grad_norm": 0.7107454538345337, + "learning_rate": 0.00011059018813410538, + "loss": 1.2298, + "step": 10487 + }, + { + "epoch": 1.8673433048433048, + "grad_norm": 0.7113911509513855, + "learning_rate": 0.00011057626927183944, + "loss": 0.9598, + "step": 10488 + }, + { + "epoch": 1.8675213675213675, + "grad_norm": 0.6734410524368286, + "learning_rate": 0.00011056235020235346, + "loss": 0.9475, + "step": 10489 + }, + { + "epoch": 1.8676994301994303, + "grad_norm": 0.6875202655792236, + "learning_rate": 0.0001105484309259202, + "loss": 1.0735, + "step": 10490 + }, + { + "epoch": 1.8678774928774928, + "grad_norm": 0.6908353567123413, + "learning_rate": 0.0001105345114428123, + "loss": 1.0558, + "step": 10491 + }, + { + "epoch": 1.8680555555555556, + "grad_norm": 0.6283324360847473, + "learning_rate": 0.00011052059175330256, + "loss": 0.8872, + "step": 10492 + }, + { + "epoch": 1.868233618233618, + "grad_norm": 0.6422587633132935, + "learning_rate": 0.00011050667185766368, + "loss": 1.1022, + "step": 10493 + }, + { + "epoch": 1.8684116809116809, + "grad_norm": 0.7075859904289246, + "learning_rate": 0.0001104927517561684, + "loss": 1.1389, + "step": 10494 + }, + { + "epoch": 1.8685897435897436, + "grad_norm": 0.5896905064582825, + "learning_rate": 0.00011047883144908944, + "loss": 0.7732, + "step": 10495 + }, + { + "epoch": 1.8687678062678064, + "grad_norm": 0.7647629976272583, + "learning_rate": 0.00011046491093669953, + "loss": 0.9983, + "step": 10496 + }, + { + "epoch": 1.868945868945869, + "grad_norm": 0.5864735841751099, + "learning_rate": 0.00011045099021927144, + "loss": 0.8427, + "step": 10497 + }, + { + "epoch": 1.8691239316239316, + "grad_norm": 0.6766837239265442, + "learning_rate": 0.00011043706929707791, + "loss": 0.9595, + "step": 10498 + }, + { + "epoch": 1.8693019943019942, + "grad_norm": 0.5480074286460876, + "learning_rate": 0.00011042314817039168, + "loss": 0.691, + "step": 10499 + }, + { + "epoch": 1.869480056980057, + "grad_norm": 0.6259615421295166, + "learning_rate": 0.00011040922683948553, + "loss": 0.9991, + "step": 10500 + }, + { + "epoch": 1.8696581196581197, + "grad_norm": 0.5950598120689392, + "learning_rate": 0.00011039530530463218, + "loss": 0.7413, + "step": 10501 + }, + { + "epoch": 1.8698361823361824, + "grad_norm": 0.8099377751350403, + "learning_rate": 0.00011038138356610441, + "loss": 1.1351, + "step": 10502 + }, + { + "epoch": 1.8700142450142452, + "grad_norm": 0.6716185212135315, + "learning_rate": 0.00011036746162417501, + "loss": 1.1057, + "step": 10503 + }, + { + "epoch": 1.8701923076923077, + "grad_norm": 0.7993219494819641, + "learning_rate": 0.00011035353947911675, + "loss": 1.2095, + "step": 10504 + }, + { + "epoch": 1.8703703703703702, + "grad_norm": 0.6381276249885559, + "learning_rate": 0.00011033961713120237, + "loss": 1.0261, + "step": 10505 + }, + { + "epoch": 1.870548433048433, + "grad_norm": 0.6326032280921936, + "learning_rate": 0.00011032569458070469, + "loss": 0.8664, + "step": 10506 + }, + { + "epoch": 1.8707264957264957, + "grad_norm": 0.6864820718765259, + "learning_rate": 0.00011031177182789644, + "loss": 0.9959, + "step": 10507 + }, + { + "epoch": 1.8709045584045585, + "grad_norm": 0.6341838240623474, + "learning_rate": 0.00011029784887305048, + "loss": 0.8029, + "step": 10508 + }, + { + "epoch": 1.8710826210826212, + "grad_norm": 0.6559172868728638, + "learning_rate": 0.00011028392571643957, + "loss": 0.9282, + "step": 10509 + }, + { + "epoch": 1.8712606837606838, + "grad_norm": 0.6976849436759949, + "learning_rate": 0.0001102700023583365, + "loss": 1.0198, + "step": 10510 + }, + { + "epoch": 1.8714387464387463, + "grad_norm": 0.7159395217895508, + "learning_rate": 0.00011025607879901402, + "loss": 1.1585, + "step": 10511 + }, + { + "epoch": 1.871616809116809, + "grad_norm": 0.7168624997138977, + "learning_rate": 0.000110242155038745, + "loss": 1.0558, + "step": 10512 + }, + { + "epoch": 1.8717948717948718, + "grad_norm": 0.5784319043159485, + "learning_rate": 0.00011022823107780224, + "loss": 0.9481, + "step": 10513 + }, + { + "epoch": 1.8719729344729346, + "grad_norm": 0.6602259874343872, + "learning_rate": 0.00011021430691645856, + "loss": 1.0538, + "step": 10514 + }, + { + "epoch": 1.8721509971509973, + "grad_norm": 0.6874588131904602, + "learning_rate": 0.00011020038255498672, + "loss": 1.1396, + "step": 10515 + }, + { + "epoch": 1.8723290598290598, + "grad_norm": 0.7311663031578064, + "learning_rate": 0.00011018645799365956, + "loss": 1.084, + "step": 10516 + }, + { + "epoch": 1.8725071225071224, + "grad_norm": 0.7097118496894836, + "learning_rate": 0.00011017253323274996, + "loss": 0.9872, + "step": 10517 + }, + { + "epoch": 1.8726851851851851, + "grad_norm": 0.6667875051498413, + "learning_rate": 0.00011015860827253068, + "loss": 1.105, + "step": 10518 + }, + { + "epoch": 1.8728632478632479, + "grad_norm": 0.6807677745819092, + "learning_rate": 0.0001101446831132746, + "loss": 0.9093, + "step": 10519 + }, + { + "epoch": 1.8730413105413106, + "grad_norm": 0.6885797381401062, + "learning_rate": 0.0001101307577552545, + "loss": 0.8479, + "step": 10520 + }, + { + "epoch": 1.8732193732193734, + "grad_norm": 0.6269213557243347, + "learning_rate": 0.00011011683219874323, + "loss": 0.9457, + "step": 10521 + }, + { + "epoch": 1.873397435897436, + "grad_norm": 0.7096766829490662, + "learning_rate": 0.00011010290644401364, + "loss": 1.0971, + "step": 10522 + }, + { + "epoch": 1.8735754985754984, + "grad_norm": 0.6909209489822388, + "learning_rate": 0.00011008898049133863, + "loss": 0.9928, + "step": 10523 + }, + { + "epoch": 1.8737535612535612, + "grad_norm": 0.6586211323738098, + "learning_rate": 0.000110075054340991, + "loss": 0.818, + "step": 10524 + }, + { + "epoch": 1.873931623931624, + "grad_norm": 0.5934817790985107, + "learning_rate": 0.0001100611279932436, + "loss": 0.7698, + "step": 10525 + }, + { + "epoch": 1.8741096866096867, + "grad_norm": 0.6361709237098694, + "learning_rate": 0.00011004720144836931, + "loss": 0.9465, + "step": 10526 + }, + { + "epoch": 1.8742877492877494, + "grad_norm": 0.6742212176322937, + "learning_rate": 0.00011003327470664095, + "loss": 1.0998, + "step": 10527 + }, + { + "epoch": 1.874465811965812, + "grad_norm": 0.6634946465492249, + "learning_rate": 0.00011001934776833143, + "loss": 0.8328, + "step": 10528 + }, + { + "epoch": 1.8746438746438745, + "grad_norm": 0.6754063963890076, + "learning_rate": 0.0001100054206337136, + "loss": 1.147, + "step": 10529 + }, + { + "epoch": 1.8748219373219372, + "grad_norm": 0.5951135158538818, + "learning_rate": 0.00010999149330306036, + "loss": 0.8956, + "step": 10530 + }, + { + "epoch": 1.875, + "grad_norm": 0.6140317320823669, + "learning_rate": 0.00010997756577664455, + "loss": 0.9368, + "step": 10531 + }, + { + "epoch": 1.8751780626780628, + "grad_norm": 0.6419258713722229, + "learning_rate": 0.00010996363805473904, + "loss": 0.9817, + "step": 10532 + }, + { + "epoch": 1.8753561253561255, + "grad_norm": 0.7173396348953247, + "learning_rate": 0.00010994971013761677, + "loss": 0.9638, + "step": 10533 + }, + { + "epoch": 1.875534188034188, + "grad_norm": 0.8125925660133362, + "learning_rate": 0.0001099357820255506, + "loss": 1.0996, + "step": 10534 + }, + { + "epoch": 1.8757122507122506, + "grad_norm": 0.6191564798355103, + "learning_rate": 0.00010992185371881341, + "loss": 0.8266, + "step": 10535 + }, + { + "epoch": 1.8758903133903133, + "grad_norm": 0.6632885336875916, + "learning_rate": 0.0001099079252176781, + "loss": 1.1884, + "step": 10536 + }, + { + "epoch": 1.876068376068376, + "grad_norm": 0.7323372960090637, + "learning_rate": 0.00010989399652241759, + "loss": 1.0842, + "step": 10537 + }, + { + "epoch": 1.8762464387464388, + "grad_norm": 0.7553854584693909, + "learning_rate": 0.00010988006763330476, + "loss": 0.9948, + "step": 10538 + }, + { + "epoch": 1.8764245014245016, + "grad_norm": 0.5887658596038818, + "learning_rate": 0.00010986613855061255, + "loss": 0.7653, + "step": 10539 + }, + { + "epoch": 1.876602564102564, + "grad_norm": 0.6849574446678162, + "learning_rate": 0.00010985220927461384, + "loss": 1.152, + "step": 10540 + }, + { + "epoch": 1.8767806267806266, + "grad_norm": 0.6985000371932983, + "learning_rate": 0.00010983827980558155, + "loss": 0.9869, + "step": 10541 + }, + { + "epoch": 1.8769586894586894, + "grad_norm": 0.6885373592376709, + "learning_rate": 0.00010982435014378858, + "loss": 1.1803, + "step": 10542 + }, + { + "epoch": 1.8771367521367521, + "grad_norm": 0.7610142827033997, + "learning_rate": 0.00010981042028950788, + "loss": 0.9219, + "step": 10543 + }, + { + "epoch": 1.8773148148148149, + "grad_norm": 0.6545612215995789, + "learning_rate": 0.00010979649024301242, + "loss": 1.0337, + "step": 10544 + }, + { + "epoch": 1.8774928774928776, + "grad_norm": 0.7307698130607605, + "learning_rate": 0.00010978256000457505, + "loss": 0.9726, + "step": 10545 + }, + { + "epoch": 1.8776709401709402, + "grad_norm": 0.68310546875, + "learning_rate": 0.00010976862957446877, + "loss": 1.161, + "step": 10546 + }, + { + "epoch": 1.8778490028490027, + "grad_norm": 0.6114758253097534, + "learning_rate": 0.00010975469895296646, + "loss": 0.8863, + "step": 10547 + }, + { + "epoch": 1.8780270655270654, + "grad_norm": 0.732390820980072, + "learning_rate": 0.00010974076814034106, + "loss": 1.0339, + "step": 10548 + }, + { + "epoch": 1.8782051282051282, + "grad_norm": 0.6741712689399719, + "learning_rate": 0.0001097268371368656, + "loss": 1.0024, + "step": 10549 + }, + { + "epoch": 1.878383190883191, + "grad_norm": 0.6374897360801697, + "learning_rate": 0.00010971290594281294, + "loss": 0.91, + "step": 10550 + }, + { + "epoch": 1.8785612535612537, + "grad_norm": 0.6434261202812195, + "learning_rate": 0.00010969897455845608, + "loss": 1.0048, + "step": 10551 + }, + { + "epoch": 1.8787393162393162, + "grad_norm": 0.6573047041893005, + "learning_rate": 0.00010968504298406794, + "loss": 1.118, + "step": 10552 + }, + { + "epoch": 1.8789173789173788, + "grad_norm": 0.6686552166938782, + "learning_rate": 0.00010967111121992152, + "loss": 1.089, + "step": 10553 + }, + { + "epoch": 1.8790954415954415, + "grad_norm": 0.7899606823921204, + "learning_rate": 0.00010965717926628976, + "loss": 1.059, + "step": 10554 + }, + { + "epoch": 1.8792735042735043, + "grad_norm": 0.5808879733085632, + "learning_rate": 0.00010964324712344564, + "loss": 0.9369, + "step": 10555 + }, + { + "epoch": 1.879451566951567, + "grad_norm": 0.6322834491729736, + "learning_rate": 0.00010962931479166211, + "loss": 0.8783, + "step": 10556 + }, + { + "epoch": 1.8796296296296298, + "grad_norm": 0.647002637386322, + "learning_rate": 0.00010961538227121218, + "loss": 0.9468, + "step": 10557 + }, + { + "epoch": 1.8798076923076923, + "grad_norm": 0.6581854820251465, + "learning_rate": 0.0001096014495623688, + "loss": 1.0077, + "step": 10558 + }, + { + "epoch": 1.8799857549857548, + "grad_norm": 0.6879259943962097, + "learning_rate": 0.00010958751666540496, + "loss": 0.976, + "step": 10559 + }, + { + "epoch": 1.8801638176638176, + "grad_norm": 0.7055090665817261, + "learning_rate": 0.00010957358358059364, + "loss": 0.8903, + "step": 10560 + }, + { + "epoch": 1.8803418803418803, + "grad_norm": 0.6865016222000122, + "learning_rate": 0.00010955965030820782, + "loss": 0.9872, + "step": 10561 + }, + { + "epoch": 1.880519943019943, + "grad_norm": 0.663436770439148, + "learning_rate": 0.00010954571684852055, + "loss": 1.0485, + "step": 10562 + }, + { + "epoch": 1.8806980056980058, + "grad_norm": 0.6861656904220581, + "learning_rate": 0.00010953178320180475, + "loss": 1.0691, + "step": 10563 + }, + { + "epoch": 1.8808760683760684, + "grad_norm": 0.8045449256896973, + "learning_rate": 0.0001095178493683335, + "loss": 1.1534, + "step": 10564 + }, + { + "epoch": 1.881054131054131, + "grad_norm": 0.6493151187896729, + "learning_rate": 0.00010950391534837973, + "loss": 0.8756, + "step": 10565 + }, + { + "epoch": 1.8812321937321936, + "grad_norm": 0.7057121992111206, + "learning_rate": 0.00010948998114221651, + "loss": 1.1709, + "step": 10566 + }, + { + "epoch": 1.8814102564102564, + "grad_norm": 0.7708197236061096, + "learning_rate": 0.0001094760467501168, + "loss": 1.0037, + "step": 10567 + }, + { + "epoch": 1.8815883190883191, + "grad_norm": 0.7234642505645752, + "learning_rate": 0.00010946211217235364, + "loss": 1.0757, + "step": 10568 + }, + { + "epoch": 1.881766381766382, + "grad_norm": 0.6964395642280579, + "learning_rate": 0.00010944817740920006, + "loss": 1.0769, + "step": 10569 + }, + { + "epoch": 1.8819444444444444, + "grad_norm": 0.7465848922729492, + "learning_rate": 0.00010943424246092906, + "loss": 0.9772, + "step": 10570 + }, + { + "epoch": 1.8821225071225072, + "grad_norm": 0.7145788073539734, + "learning_rate": 0.0001094203073278137, + "loss": 0.9638, + "step": 10571 + }, + { + "epoch": 1.8823005698005697, + "grad_norm": 0.7421764135360718, + "learning_rate": 0.00010940637201012698, + "loss": 1.0324, + "step": 10572 + }, + { + "epoch": 1.8824786324786325, + "grad_norm": 0.7373253107070923, + "learning_rate": 0.0001093924365081419, + "loss": 1.1554, + "step": 10573 + }, + { + "epoch": 1.8826566951566952, + "grad_norm": 0.6861984729766846, + "learning_rate": 0.00010937850082213156, + "loss": 0.9899, + "step": 10574 + }, + { + "epoch": 1.882834757834758, + "grad_norm": 0.6173393130302429, + "learning_rate": 0.000109364564952369, + "loss": 0.8495, + "step": 10575 + }, + { + "epoch": 1.8830128205128205, + "grad_norm": 0.6871610879898071, + "learning_rate": 0.00010935062889912723, + "loss": 1.2164, + "step": 10576 + }, + { + "epoch": 1.8831908831908832, + "grad_norm": 0.7062903642654419, + "learning_rate": 0.00010933669266267931, + "loss": 1.1077, + "step": 10577 + }, + { + "epoch": 1.8833689458689458, + "grad_norm": 0.6574689745903015, + "learning_rate": 0.00010932275624329828, + "loss": 0.9326, + "step": 10578 + }, + { + "epoch": 1.8835470085470085, + "grad_norm": 0.636385440826416, + "learning_rate": 0.00010930881964125723, + "loss": 1.0581, + "step": 10579 + }, + { + "epoch": 1.8837250712250713, + "grad_norm": 0.6178432106971741, + "learning_rate": 0.0001092948828568292, + "loss": 1.1288, + "step": 10580 + }, + { + "epoch": 1.883903133903134, + "grad_norm": 0.6509431600570679, + "learning_rate": 0.00010928094589028721, + "loss": 1.0113, + "step": 10581 + }, + { + "epoch": 1.8840811965811965, + "grad_norm": 0.6543706059455872, + "learning_rate": 0.00010926700874190441, + "loss": 1.0041, + "step": 10582 + }, + { + "epoch": 1.8842592592592593, + "grad_norm": 0.6815463304519653, + "learning_rate": 0.0001092530714119538, + "loss": 1.0892, + "step": 10583 + }, + { + "epoch": 1.8844373219373218, + "grad_norm": 0.6787421107292175, + "learning_rate": 0.00010923913390070846, + "loss": 1.2693, + "step": 10584 + }, + { + "epoch": 1.8846153846153846, + "grad_norm": 0.6953850984573364, + "learning_rate": 0.00010922519620844151, + "loss": 0.9848, + "step": 10585 + }, + { + "epoch": 1.8847934472934473, + "grad_norm": 0.7061360478401184, + "learning_rate": 0.000109211258335426, + "loss": 0.949, + "step": 10586 + }, + { + "epoch": 1.88497150997151, + "grad_norm": 0.6845372915267944, + "learning_rate": 0.00010919732028193504, + "loss": 0.9554, + "step": 10587 + }, + { + "epoch": 1.8851495726495726, + "grad_norm": 0.6524720788002014, + "learning_rate": 0.00010918338204824165, + "loss": 1.1037, + "step": 10588 + }, + { + "epoch": 1.8853276353276354, + "grad_norm": 0.6410523653030396, + "learning_rate": 0.00010916944363461899, + "loss": 0.9085, + "step": 10589 + }, + { + "epoch": 1.885505698005698, + "grad_norm": 0.7109059691429138, + "learning_rate": 0.00010915550504134014, + "loss": 1.0526, + "step": 10590 + }, + { + "epoch": 1.8856837606837606, + "grad_norm": 0.7781991362571716, + "learning_rate": 0.00010914156626867818, + "loss": 0.9737, + "step": 10591 + }, + { + "epoch": 1.8858618233618234, + "grad_norm": 0.7173767685890198, + "learning_rate": 0.00010912762731690623, + "loss": 0.8862, + "step": 10592 + }, + { + "epoch": 1.8860398860398861, + "grad_norm": 0.7650504112243652, + "learning_rate": 0.00010911368818629732, + "loss": 1.2175, + "step": 10593 + }, + { + "epoch": 1.8862179487179487, + "grad_norm": 0.6316116452217102, + "learning_rate": 0.00010909974887712468, + "loss": 0.8332, + "step": 10594 + }, + { + "epoch": 1.8863960113960114, + "grad_norm": 0.6504800319671631, + "learning_rate": 0.00010908580938966138, + "loss": 0.8864, + "step": 10595 + }, + { + "epoch": 1.886574074074074, + "grad_norm": 0.675507128238678, + "learning_rate": 0.00010907186972418049, + "loss": 0.8523, + "step": 10596 + }, + { + "epoch": 1.8867521367521367, + "grad_norm": 0.6535763144493103, + "learning_rate": 0.00010905792988095515, + "loss": 1.0786, + "step": 10597 + }, + { + "epoch": 1.8869301994301995, + "grad_norm": 0.7071853280067444, + "learning_rate": 0.0001090439898602585, + "loss": 0.9319, + "step": 10598 + }, + { + "epoch": 1.8871082621082622, + "grad_norm": 0.699466347694397, + "learning_rate": 0.00010903004966236365, + "loss": 0.9573, + "step": 10599 + }, + { + "epoch": 1.8872863247863247, + "grad_norm": 0.7099201083183289, + "learning_rate": 0.00010901610928754375, + "loss": 0.9447, + "step": 10600 + }, + { + "epoch": 1.8874643874643875, + "grad_norm": 0.6140450835227966, + "learning_rate": 0.00010900216873607189, + "loss": 1.0227, + "step": 10601 + }, + { + "epoch": 1.88764245014245, + "grad_norm": 0.6613629460334778, + "learning_rate": 0.00010898822800822127, + "loss": 1.0152, + "step": 10602 + }, + { + "epoch": 1.8878205128205128, + "grad_norm": 0.7334819436073303, + "learning_rate": 0.00010897428710426498, + "loss": 1.1452, + "step": 10603 + }, + { + "epoch": 1.8879985754985755, + "grad_norm": 0.6819368004798889, + "learning_rate": 0.00010896034602447616, + "loss": 1.0504, + "step": 10604 + }, + { + "epoch": 1.8881766381766383, + "grad_norm": 0.6781361103057861, + "learning_rate": 0.00010894640476912799, + "loss": 0.8719, + "step": 10605 + }, + { + "epoch": 1.8883547008547008, + "grad_norm": 0.621960461139679, + "learning_rate": 0.00010893246333849361, + "loss": 0.9264, + "step": 10606 + }, + { + "epoch": 1.8885327635327636, + "grad_norm": 0.6350592374801636, + "learning_rate": 0.00010891852173284615, + "loss": 1.0042, + "step": 10607 + }, + { + "epoch": 1.888710826210826, + "grad_norm": 0.6650694012641907, + "learning_rate": 0.00010890457995245879, + "loss": 1.1387, + "step": 10608 + }, + { + "epoch": 1.8888888888888888, + "grad_norm": 0.6515723466873169, + "learning_rate": 0.00010889063799760468, + "loss": 0.9508, + "step": 10609 + }, + { + "epoch": 1.8890669515669516, + "grad_norm": 0.6368890404701233, + "learning_rate": 0.000108876695868557, + "loss": 0.8051, + "step": 10610 + }, + { + "epoch": 1.8892450142450143, + "grad_norm": 0.7971013188362122, + "learning_rate": 0.00010886275356558888, + "loss": 0.8629, + "step": 10611 + }, + { + "epoch": 1.8894230769230769, + "grad_norm": 0.6739095449447632, + "learning_rate": 0.00010884881108897353, + "loss": 0.9606, + "step": 10612 + }, + { + "epoch": 1.8896011396011396, + "grad_norm": 0.7754076719284058, + "learning_rate": 0.00010883486843898412, + "loss": 1.0751, + "step": 10613 + }, + { + "epoch": 1.8897792022792022, + "grad_norm": 0.6538285613059998, + "learning_rate": 0.00010882092561589379, + "loss": 0.9288, + "step": 10614 + }, + { + "epoch": 1.889957264957265, + "grad_norm": 0.7373257875442505, + "learning_rate": 0.00010880698261997577, + "loss": 0.9884, + "step": 10615 + }, + { + "epoch": 1.8901353276353277, + "grad_norm": 0.6575660109519958, + "learning_rate": 0.00010879303945150321, + "loss": 1.0307, + "step": 10616 + }, + { + "epoch": 1.8903133903133904, + "grad_norm": 0.7500179409980774, + "learning_rate": 0.00010877909611074932, + "loss": 1.0812, + "step": 10617 + }, + { + "epoch": 1.890491452991453, + "grad_norm": 0.7607308030128479, + "learning_rate": 0.00010876515259798727, + "loss": 0.9746, + "step": 10618 + }, + { + "epoch": 1.8906695156695157, + "grad_norm": 0.7930253744125366, + "learning_rate": 0.00010875120891349024, + "loss": 0.7911, + "step": 10619 + }, + { + "epoch": 1.8908475783475782, + "grad_norm": 0.635254979133606, + "learning_rate": 0.00010873726505753148, + "loss": 1.0468, + "step": 10620 + }, + { + "epoch": 1.891025641025641, + "grad_norm": 0.7579759359359741, + "learning_rate": 0.00010872332103038414, + "loss": 0.9558, + "step": 10621 + }, + { + "epoch": 1.8912037037037037, + "grad_norm": 0.5841903686523438, + "learning_rate": 0.00010870937683232146, + "loss": 0.913, + "step": 10622 + }, + { + "epoch": 1.8913817663817665, + "grad_norm": 0.7088860273361206, + "learning_rate": 0.00010869543246361664, + "loss": 1.0814, + "step": 10623 + }, + { + "epoch": 1.8915598290598292, + "grad_norm": 0.6713772416114807, + "learning_rate": 0.00010868148792454285, + "loss": 0.9972, + "step": 10624 + }, + { + "epoch": 1.8917378917378918, + "grad_norm": 0.6733243465423584, + "learning_rate": 0.00010866754321537338, + "loss": 0.9596, + "step": 10625 + }, + { + "epoch": 1.8919159544159543, + "grad_norm": 0.7747747898101807, + "learning_rate": 0.00010865359833638138, + "loss": 1.0871, + "step": 10626 + }, + { + "epoch": 1.892094017094017, + "grad_norm": 0.677175760269165, + "learning_rate": 0.00010863965328784011, + "loss": 0.9939, + "step": 10627 + }, + { + "epoch": 1.8922720797720798, + "grad_norm": 0.7883930206298828, + "learning_rate": 0.00010862570807002279, + "loss": 1.0708, + "step": 10628 + }, + { + "epoch": 1.8924501424501425, + "grad_norm": 0.7003030180931091, + "learning_rate": 0.00010861176268320261, + "loss": 0.9791, + "step": 10629 + }, + { + "epoch": 1.8926282051282053, + "grad_norm": 0.7450358271598816, + "learning_rate": 0.00010859781712765284, + "loss": 0.9672, + "step": 10630 + }, + { + "epoch": 1.8928062678062678, + "grad_norm": 0.7776696085929871, + "learning_rate": 0.00010858387140364672, + "loss": 1.1037, + "step": 10631 + }, + { + "epoch": 1.8929843304843303, + "grad_norm": 0.6896173357963562, + "learning_rate": 0.00010856992551145745, + "loss": 1.0048, + "step": 10632 + }, + { + "epoch": 1.893162393162393, + "grad_norm": 0.5997697710990906, + "learning_rate": 0.00010855597945135834, + "loss": 0.8025, + "step": 10633 + }, + { + "epoch": 1.8933404558404558, + "grad_norm": 0.8781484365463257, + "learning_rate": 0.00010854203322362251, + "loss": 1.0014, + "step": 10634 + }, + { + "epoch": 1.8935185185185186, + "grad_norm": 0.6348843574523926, + "learning_rate": 0.00010852808682852334, + "loss": 0.9857, + "step": 10635 + }, + { + "epoch": 1.8936965811965814, + "grad_norm": 0.9704267978668213, + "learning_rate": 0.000108514140266334, + "loss": 1.0522, + "step": 10636 + }, + { + "epoch": 1.8938746438746439, + "grad_norm": 0.70372074842453, + "learning_rate": 0.00010850019353732779, + "loss": 1.1044, + "step": 10637 + }, + { + "epoch": 1.8940527065527064, + "grad_norm": 0.6528043150901794, + "learning_rate": 0.00010848624664177793, + "loss": 0.9328, + "step": 10638 + }, + { + "epoch": 1.8942307692307692, + "grad_norm": 0.6299768090248108, + "learning_rate": 0.00010847229957995768, + "loss": 1.0099, + "step": 10639 + }, + { + "epoch": 1.894408831908832, + "grad_norm": 0.6347038149833679, + "learning_rate": 0.00010845835235214034, + "loss": 1.1354, + "step": 10640 + }, + { + "epoch": 1.8945868945868947, + "grad_norm": 0.7087811827659607, + "learning_rate": 0.00010844440495859913, + "loss": 1.0543, + "step": 10641 + }, + { + "epoch": 1.8947649572649574, + "grad_norm": 0.7386305332183838, + "learning_rate": 0.00010843045739960738, + "loss": 0.9192, + "step": 10642 + }, + { + "epoch": 1.89494301994302, + "grad_norm": 0.6047097444534302, + "learning_rate": 0.00010841650967543833, + "loss": 0.8668, + "step": 10643 + }, + { + "epoch": 1.8951210826210825, + "grad_norm": 0.6779503226280212, + "learning_rate": 0.00010840256178636523, + "loss": 0.9263, + "step": 10644 + }, + { + "epoch": 1.8952991452991452, + "grad_norm": 0.7398194670677185, + "learning_rate": 0.00010838861373266138, + "loss": 0.9534, + "step": 10645 + }, + { + "epoch": 1.895477207977208, + "grad_norm": 0.8138558864593506, + "learning_rate": 0.00010837466551460011, + "loss": 0.9835, + "step": 10646 + }, + { + "epoch": 1.8956552706552707, + "grad_norm": 0.8847818374633789, + "learning_rate": 0.00010836071713245466, + "loss": 0.9769, + "step": 10647 + }, + { + "epoch": 1.8958333333333335, + "grad_norm": 0.6824164390563965, + "learning_rate": 0.0001083467685864983, + "loss": 0.9901, + "step": 10648 + }, + { + "epoch": 1.896011396011396, + "grad_norm": 0.6318182945251465, + "learning_rate": 0.00010833281987700436, + "loss": 0.7677, + "step": 10649 + }, + { + "epoch": 1.8961894586894585, + "grad_norm": 0.7372074127197266, + "learning_rate": 0.00010831887100424612, + "loss": 0.9858, + "step": 10650 + }, + { + "epoch": 1.8963675213675213, + "grad_norm": 0.7246516346931458, + "learning_rate": 0.00010830492196849688, + "loss": 0.9644, + "step": 10651 + }, + { + "epoch": 1.896545584045584, + "grad_norm": 0.6517095565795898, + "learning_rate": 0.00010829097277002997, + "loss": 1.1733, + "step": 10652 + }, + { + "epoch": 1.8967236467236468, + "grad_norm": 0.6931695342063904, + "learning_rate": 0.00010827702340911867, + "loss": 0.9923, + "step": 10653 + }, + { + "epoch": 1.8969017094017095, + "grad_norm": 0.6210272312164307, + "learning_rate": 0.00010826307388603628, + "loss": 0.8757, + "step": 10654 + }, + { + "epoch": 1.897079772079772, + "grad_norm": 0.7011165618896484, + "learning_rate": 0.00010824912420105611, + "loss": 1.0011, + "step": 10655 + }, + { + "epoch": 1.8972578347578346, + "grad_norm": 0.7431246638298035, + "learning_rate": 0.0001082351743544515, + "loss": 1.1498, + "step": 10656 + }, + { + "epoch": 1.8974358974358974, + "grad_norm": 0.7099978923797607, + "learning_rate": 0.00010822122434649576, + "loss": 1.0673, + "step": 10657 + }, + { + "epoch": 1.89761396011396, + "grad_norm": 0.7375551462173462, + "learning_rate": 0.00010820727417746219, + "loss": 1.0157, + "step": 10658 + }, + { + "epoch": 1.8977920227920229, + "grad_norm": 0.8155642151832581, + "learning_rate": 0.00010819332384762413, + "loss": 1.229, + "step": 10659 + }, + { + "epoch": 1.8979700854700856, + "grad_norm": 0.6917914748191833, + "learning_rate": 0.00010817937335725493, + "loss": 0.9701, + "step": 10660 + }, + { + "epoch": 1.8981481481481481, + "grad_norm": 0.8498218059539795, + "learning_rate": 0.00010816542270662786, + "loss": 1.0123, + "step": 10661 + }, + { + "epoch": 1.8983262108262107, + "grad_norm": 0.7234359979629517, + "learning_rate": 0.00010815147189601634, + "loss": 1.0755, + "step": 10662 + }, + { + "epoch": 1.8985042735042734, + "grad_norm": 0.6997553110122681, + "learning_rate": 0.00010813752092569365, + "loss": 1.1594, + "step": 10663 + }, + { + "epoch": 1.8986823361823362, + "grad_norm": 0.6519457101821899, + "learning_rate": 0.00010812356979593314, + "loss": 0.9609, + "step": 10664 + }, + { + "epoch": 1.898860398860399, + "grad_norm": 0.7215374708175659, + "learning_rate": 0.00010810961850700813, + "loss": 1.1392, + "step": 10665 + }, + { + "epoch": 1.8990384615384617, + "grad_norm": 0.7766093611717224, + "learning_rate": 0.00010809566705919202, + "loss": 1.0256, + "step": 10666 + }, + { + "epoch": 1.8992165242165242, + "grad_norm": 0.6520358920097351, + "learning_rate": 0.00010808171545275814, + "loss": 1.0434, + "step": 10667 + }, + { + "epoch": 1.8993945868945867, + "grad_norm": 0.7454953193664551, + "learning_rate": 0.00010806776368797982, + "loss": 1.2323, + "step": 10668 + }, + { + "epoch": 1.8995726495726495, + "grad_norm": 0.6891530752182007, + "learning_rate": 0.00010805381176513043, + "loss": 1.1104, + "step": 10669 + }, + { + "epoch": 1.8997507122507122, + "grad_norm": 0.6609626412391663, + "learning_rate": 0.00010803985968448331, + "loss": 0.8565, + "step": 10670 + }, + { + "epoch": 1.899928774928775, + "grad_norm": 0.6650999188423157, + "learning_rate": 0.00010802590744631187, + "loss": 1.1003, + "step": 10671 + }, + { + "epoch": 1.9001068376068377, + "grad_norm": 0.5794292092323303, + "learning_rate": 0.00010801195505088945, + "loss": 0.528, + "step": 10672 + }, + { + "epoch": 1.9002849002849003, + "grad_norm": 1.0802743434906006, + "learning_rate": 0.00010799800249848939, + "loss": 0.8861, + "step": 10673 + }, + { + "epoch": 1.9004629629629628, + "grad_norm": 0.650833249092102, + "learning_rate": 0.00010798404978938513, + "loss": 0.9962, + "step": 10674 + }, + { + "epoch": 1.9006410256410255, + "grad_norm": 0.7290451526641846, + "learning_rate": 0.00010797009692384994, + "loss": 1.0764, + "step": 10675 + }, + { + "epoch": 1.9008190883190883, + "grad_norm": 0.6273928880691528, + "learning_rate": 0.00010795614390215727, + "loss": 0.9478, + "step": 10676 + }, + { + "epoch": 1.900997150997151, + "grad_norm": 0.6939455270767212, + "learning_rate": 0.00010794219072458052, + "loss": 0.8991, + "step": 10677 + }, + { + "epoch": 1.9011752136752138, + "grad_norm": 0.7455828189849854, + "learning_rate": 0.00010792823739139302, + "loss": 0.8902, + "step": 10678 + }, + { + "epoch": 1.9013532763532763, + "grad_norm": 0.6894607543945312, + "learning_rate": 0.00010791428390286817, + "loss": 0.9355, + "step": 10679 + }, + { + "epoch": 1.901531339031339, + "grad_norm": 0.6844658851623535, + "learning_rate": 0.00010790033025927936, + "loss": 0.9835, + "step": 10680 + }, + { + "epoch": 1.9017094017094016, + "grad_norm": 0.6646730899810791, + "learning_rate": 0.00010788637646090001, + "loss": 0.9376, + "step": 10681 + }, + { + "epoch": 1.9018874643874644, + "grad_norm": 0.6494864225387573, + "learning_rate": 0.00010787242250800349, + "loss": 0.8533, + "step": 10682 + }, + { + "epoch": 1.9020655270655271, + "grad_norm": 0.686198353767395, + "learning_rate": 0.0001078584684008632, + "loss": 0.8075, + "step": 10683 + }, + { + "epoch": 1.9022435897435899, + "grad_norm": 0.7014855742454529, + "learning_rate": 0.00010784451413975256, + "loss": 1.0805, + "step": 10684 + }, + { + "epoch": 1.9024216524216524, + "grad_norm": 0.7191864252090454, + "learning_rate": 0.00010783055972494496, + "loss": 0.9375, + "step": 10685 + }, + { + "epoch": 1.9025997150997151, + "grad_norm": 0.8114212155342102, + "learning_rate": 0.00010781660515671379, + "loss": 0.9716, + "step": 10686 + }, + { + "epoch": 1.9027777777777777, + "grad_norm": 0.7423529028892517, + "learning_rate": 0.0001078026504353325, + "loss": 0.9066, + "step": 10687 + }, + { + "epoch": 1.9029558404558404, + "grad_norm": 0.6517882347106934, + "learning_rate": 0.00010778869556107447, + "loss": 0.9908, + "step": 10688 + }, + { + "epoch": 1.9031339031339032, + "grad_norm": 0.6983367800712585, + "learning_rate": 0.00010777474053421315, + "loss": 1.1048, + "step": 10689 + }, + { + "epoch": 1.903311965811966, + "grad_norm": 0.597766101360321, + "learning_rate": 0.00010776078535502193, + "loss": 0.84, + "step": 10690 + }, + { + "epoch": 1.9034900284900285, + "grad_norm": 0.7335455417633057, + "learning_rate": 0.00010774683002377422, + "loss": 1.0387, + "step": 10691 + }, + { + "epoch": 1.9036680911680912, + "grad_norm": 0.6742176413536072, + "learning_rate": 0.0001077328745407435, + "loss": 0.9743, + "step": 10692 + }, + { + "epoch": 1.9038461538461537, + "grad_norm": 0.7954961657524109, + "learning_rate": 0.00010771891890620316, + "loss": 1.1025, + "step": 10693 + }, + { + "epoch": 1.9040242165242165, + "grad_norm": 0.733351469039917, + "learning_rate": 0.00010770496312042664, + "loss": 1.028, + "step": 10694 + }, + { + "epoch": 1.9042022792022792, + "grad_norm": 0.7059772610664368, + "learning_rate": 0.00010769100718368734, + "loss": 1.0103, + "step": 10695 + }, + { + "epoch": 1.904380341880342, + "grad_norm": 0.6234813332557678, + "learning_rate": 0.00010767705109625877, + "loss": 0.6893, + "step": 10696 + }, + { + "epoch": 1.9045584045584045, + "grad_norm": 0.6670311689376831, + "learning_rate": 0.0001076630948584143, + "loss": 1.1386, + "step": 10697 + }, + { + "epoch": 1.9047364672364673, + "grad_norm": 0.7444894909858704, + "learning_rate": 0.00010764913847042744, + "loss": 0.8524, + "step": 10698 + }, + { + "epoch": 1.9049145299145298, + "grad_norm": 0.6252964735031128, + "learning_rate": 0.00010763518193257158, + "loss": 0.9407, + "step": 10699 + }, + { + "epoch": 1.9050925925925926, + "grad_norm": 0.7794382572174072, + "learning_rate": 0.0001076212252451202, + "loss": 1.05, + "step": 10700 + }, + { + "epoch": 1.9052706552706553, + "grad_norm": 0.6313693523406982, + "learning_rate": 0.00010760726840834671, + "loss": 0.8667, + "step": 10701 + }, + { + "epoch": 1.905448717948718, + "grad_norm": 0.6766461730003357, + "learning_rate": 0.00010759331142252462, + "loss": 0.9675, + "step": 10702 + }, + { + "epoch": 1.9056267806267806, + "grad_norm": 0.7457365393638611, + "learning_rate": 0.00010757935428792739, + "loss": 0.9177, + "step": 10703 + }, + { + "epoch": 1.9058048433048433, + "grad_norm": 0.6649872064590454, + "learning_rate": 0.00010756539700482844, + "loss": 0.8703, + "step": 10704 + }, + { + "epoch": 1.9059829059829059, + "grad_norm": 0.8418740034103394, + "learning_rate": 0.00010755143957350127, + "loss": 0.8993, + "step": 10705 + }, + { + "epoch": 1.9061609686609686, + "grad_norm": 0.6767167448997498, + "learning_rate": 0.00010753748199421929, + "loss": 1.0063, + "step": 10706 + }, + { + "epoch": 1.9063390313390314, + "grad_norm": 0.6959242820739746, + "learning_rate": 0.00010752352426725603, + "loss": 1.0516, + "step": 10707 + }, + { + "epoch": 1.9065170940170941, + "grad_norm": 0.7106529474258423, + "learning_rate": 0.00010750956639288493, + "loss": 0.9596, + "step": 10708 + }, + { + "epoch": 1.9066951566951567, + "grad_norm": 0.7611243724822998, + "learning_rate": 0.00010749560837137949, + "loss": 1.0739, + "step": 10709 + }, + { + "epoch": 1.9068732193732194, + "grad_norm": 0.6684338450431824, + "learning_rate": 0.00010748165020301317, + "loss": 1.1437, + "step": 10710 + }, + { + "epoch": 1.907051282051282, + "grad_norm": 0.5957385897636414, + "learning_rate": 0.00010746769188805945, + "loss": 0.8802, + "step": 10711 + }, + { + "epoch": 1.9072293447293447, + "grad_norm": 0.69919353723526, + "learning_rate": 0.00010745373342679184, + "loss": 1.1891, + "step": 10712 + }, + { + "epoch": 1.9074074074074074, + "grad_norm": 0.7562127709388733, + "learning_rate": 0.0001074397748194838, + "loss": 0.8717, + "step": 10713 + }, + { + "epoch": 1.9075854700854702, + "grad_norm": 0.6420038938522339, + "learning_rate": 0.00010742581606640882, + "loss": 1.1196, + "step": 10714 + }, + { + "epoch": 1.9077635327635327, + "grad_norm": 0.7545611262321472, + "learning_rate": 0.00010741185716784039, + "loss": 1.161, + "step": 10715 + }, + { + "epoch": 1.9079415954415955, + "grad_norm": 0.6467727422714233, + "learning_rate": 0.000107397898124052, + "loss": 0.8029, + "step": 10716 + }, + { + "epoch": 1.908119658119658, + "grad_norm": 0.6129235625267029, + "learning_rate": 0.00010738393893531722, + "loss": 0.8802, + "step": 10717 + }, + { + "epoch": 1.9082977207977208, + "grad_norm": 0.6416113376617432, + "learning_rate": 0.00010736997960190946, + "loss": 0.8465, + "step": 10718 + }, + { + "epoch": 1.9084757834757835, + "grad_norm": 0.6609050631523132, + "learning_rate": 0.00010735602012410229, + "loss": 0.9484, + "step": 10719 + }, + { + "epoch": 1.9086538461538463, + "grad_norm": 0.6302639842033386, + "learning_rate": 0.00010734206050216913, + "loss": 0.898, + "step": 10720 + }, + { + "epoch": 1.9088319088319088, + "grad_norm": 0.7291215658187866, + "learning_rate": 0.00010732810073638358, + "loss": 0.9544, + "step": 10721 + }, + { + "epoch": 1.9090099715099715, + "grad_norm": 0.6436966061592102, + "learning_rate": 0.0001073141408270191, + "loss": 0.956, + "step": 10722 + }, + { + "epoch": 1.909188034188034, + "grad_norm": 0.6247875094413757, + "learning_rate": 0.00010730018077434924, + "loss": 0.8704, + "step": 10723 + }, + { + "epoch": 1.9093660968660968, + "grad_norm": 0.7599029541015625, + "learning_rate": 0.00010728622057864753, + "loss": 1.2024, + "step": 10724 + }, + { + "epoch": 1.9095441595441596, + "grad_norm": 0.6894544959068298, + "learning_rate": 0.00010727226024018744, + "loss": 1.1226, + "step": 10725 + }, + { + "epoch": 1.9097222222222223, + "grad_norm": 0.6920733451843262, + "learning_rate": 0.0001072582997592425, + "loss": 0.7682, + "step": 10726 + }, + { + "epoch": 1.9099002849002849, + "grad_norm": 0.6013005375862122, + "learning_rate": 0.00010724433913608627, + "loss": 0.9462, + "step": 10727 + }, + { + "epoch": 1.9100783475783476, + "grad_norm": 0.7466302514076233, + "learning_rate": 0.00010723037837099225, + "loss": 0.9507, + "step": 10728 + }, + { + "epoch": 1.9102564102564101, + "grad_norm": 0.7070091962814331, + "learning_rate": 0.00010721641746423401, + "loss": 1.0704, + "step": 10729 + }, + { + "epoch": 1.9104344729344729, + "grad_norm": 0.6747950315475464, + "learning_rate": 0.00010720245641608506, + "loss": 0.7899, + "step": 10730 + }, + { + "epoch": 1.9106125356125356, + "grad_norm": 0.7338371276855469, + "learning_rate": 0.00010718849522681891, + "loss": 0.9574, + "step": 10731 + }, + { + "epoch": 1.9107905982905984, + "grad_norm": 0.6923216581344604, + "learning_rate": 0.00010717453389670915, + "loss": 1.0725, + "step": 10732 + }, + { + "epoch": 1.910968660968661, + "grad_norm": 0.6050783395767212, + "learning_rate": 0.0001071605724260293, + "loss": 0.9224, + "step": 10733 + }, + { + "epoch": 1.9111467236467237, + "grad_norm": 0.6854597330093384, + "learning_rate": 0.00010714661081505291, + "loss": 0.9749, + "step": 10734 + }, + { + "epoch": 1.9113247863247862, + "grad_norm": 0.7661508321762085, + "learning_rate": 0.00010713264906405351, + "loss": 1.1564, + "step": 10735 + }, + { + "epoch": 1.911502849002849, + "grad_norm": 0.6389622688293457, + "learning_rate": 0.00010711868717330467, + "loss": 0.8148, + "step": 10736 + }, + { + "epoch": 1.9116809116809117, + "grad_norm": 0.6318161487579346, + "learning_rate": 0.00010710472514307996, + "loss": 0.7833, + "step": 10737 + }, + { + "epoch": 1.9118589743589745, + "grad_norm": 0.8646727800369263, + "learning_rate": 0.00010709076297365292, + "loss": 1.2682, + "step": 10738 + }, + { + "epoch": 1.9120370370370372, + "grad_norm": 0.6085501909255981, + "learning_rate": 0.0001070768006652971, + "loss": 0.8706, + "step": 10739 + }, + { + "epoch": 1.9122150997150997, + "grad_norm": 0.8259731531143188, + "learning_rate": 0.00010706283821828607, + "loss": 0.9014, + "step": 10740 + }, + { + "epoch": 1.9123931623931623, + "grad_norm": 0.6509148478507996, + "learning_rate": 0.0001070488756328934, + "loss": 0.8814, + "step": 10741 + }, + { + "epoch": 1.912571225071225, + "grad_norm": 0.7241966128349304, + "learning_rate": 0.00010703491290939264, + "loss": 0.9925, + "step": 10742 + }, + { + "epoch": 1.9127492877492878, + "grad_norm": 0.7736822366714478, + "learning_rate": 0.00010702095004805738, + "loss": 1.0881, + "step": 10743 + }, + { + "epoch": 1.9129273504273505, + "grad_norm": 0.6912824511528015, + "learning_rate": 0.00010700698704916123, + "loss": 1.2334, + "step": 10744 + }, + { + "epoch": 1.9131054131054133, + "grad_norm": 0.825065553188324, + "learning_rate": 0.0001069930239129777, + "loss": 0.9783, + "step": 10745 + }, + { + "epoch": 1.9132834757834758, + "grad_norm": 0.7650560140609741, + "learning_rate": 0.00010697906063978038, + "loss": 0.9788, + "step": 10746 + }, + { + "epoch": 1.9134615384615383, + "grad_norm": 0.7368232607841492, + "learning_rate": 0.00010696509722984287, + "loss": 0.8704, + "step": 10747 + }, + { + "epoch": 1.913639601139601, + "grad_norm": 0.6630628108978271, + "learning_rate": 0.00010695113368343875, + "loss": 1.1993, + "step": 10748 + }, + { + "epoch": 1.9138176638176638, + "grad_norm": 0.6842190027236938, + "learning_rate": 0.0001069371700008416, + "loss": 0.9128, + "step": 10749 + }, + { + "epoch": 1.9139957264957266, + "grad_norm": 0.591655969619751, + "learning_rate": 0.00010692320618232503, + "loss": 1.0607, + "step": 10750 + }, + { + "epoch": 1.9141737891737893, + "grad_norm": 0.74644535779953, + "learning_rate": 0.0001069092422281626, + "loss": 1.0937, + "step": 10751 + }, + { + "epoch": 1.9143518518518519, + "grad_norm": 0.7123813629150391, + "learning_rate": 0.00010689527813862792, + "loss": 0.9043, + "step": 10752 + }, + { + "epoch": 1.9145299145299144, + "grad_norm": 0.6850089430809021, + "learning_rate": 0.0001068813139139946, + "loss": 1.0908, + "step": 10753 + }, + { + "epoch": 1.9147079772079771, + "grad_norm": 0.5882078409194946, + "learning_rate": 0.00010686734955453623, + "loss": 0.829, + "step": 10754 + }, + { + "epoch": 1.91488603988604, + "grad_norm": 0.6741717457771301, + "learning_rate": 0.00010685338506052642, + "loss": 0.9197, + "step": 10755 + }, + { + "epoch": 1.9150641025641026, + "grad_norm": 0.6597354412078857, + "learning_rate": 0.00010683942043223876, + "loss": 0.8778, + "step": 10756 + }, + { + "epoch": 1.9152421652421654, + "grad_norm": 0.6682151556015015, + "learning_rate": 0.00010682545566994684, + "loss": 0.9305, + "step": 10757 + }, + { + "epoch": 1.915420227920228, + "grad_norm": 0.8283176422119141, + "learning_rate": 0.00010681149077392431, + "loss": 1.0164, + "step": 10758 + }, + { + "epoch": 1.9155982905982905, + "grad_norm": 0.648845374584198, + "learning_rate": 0.00010679752574444477, + "loss": 1.0114, + "step": 10759 + }, + { + "epoch": 1.9157763532763532, + "grad_norm": 0.755913496017456, + "learning_rate": 0.00010678356058178182, + "loss": 1.1142, + "step": 10760 + }, + { + "epoch": 1.915954415954416, + "grad_norm": 0.7334780097007751, + "learning_rate": 0.00010676959528620911, + "loss": 0.8758, + "step": 10761 + }, + { + "epoch": 1.9161324786324787, + "grad_norm": 0.9132041335105896, + "learning_rate": 0.00010675562985800025, + "loss": 0.995, + "step": 10762 + }, + { + "epoch": 1.9163105413105415, + "grad_norm": 0.7070860266685486, + "learning_rate": 0.00010674166429742882, + "loss": 0.9856, + "step": 10763 + }, + { + "epoch": 1.916488603988604, + "grad_norm": 0.7143638134002686, + "learning_rate": 0.00010672769860476853, + "loss": 1.0612, + "step": 10764 + }, + { + "epoch": 1.9166666666666665, + "grad_norm": 0.815717339515686, + "learning_rate": 0.00010671373278029293, + "loss": 1.1539, + "step": 10765 + }, + { + "epoch": 1.9168447293447293, + "grad_norm": 0.6379499435424805, + "learning_rate": 0.0001066997668242757, + "loss": 0.8295, + "step": 10766 + }, + { + "epoch": 1.917022792022792, + "grad_norm": 0.6482511758804321, + "learning_rate": 0.00010668580073699044, + "loss": 1.0079, + "step": 10767 + }, + { + "epoch": 1.9172008547008548, + "grad_norm": 0.7382873296737671, + "learning_rate": 0.00010667183451871082, + "loss": 0.8973, + "step": 10768 + }, + { + "epoch": 1.9173789173789175, + "grad_norm": 0.7818579077720642, + "learning_rate": 0.00010665786816971044, + "loss": 1.2131, + "step": 10769 + }, + { + "epoch": 1.91755698005698, + "grad_norm": 0.6960901021957397, + "learning_rate": 0.000106643901690263, + "loss": 1.1466, + "step": 10770 + }, + { + "epoch": 1.9177350427350426, + "grad_norm": 0.696966826915741, + "learning_rate": 0.00010662993508064208, + "loss": 0.854, + "step": 10771 + }, + { + "epoch": 1.9179131054131053, + "grad_norm": 0.6745442152023315, + "learning_rate": 0.00010661596834112133, + "loss": 0.9559, + "step": 10772 + }, + { + "epoch": 1.918091168091168, + "grad_norm": 0.7436230778694153, + "learning_rate": 0.00010660200147197447, + "loss": 1.1367, + "step": 10773 + }, + { + "epoch": 1.9182692307692308, + "grad_norm": 0.6051676869392395, + "learning_rate": 0.00010658803447347509, + "loss": 1.05, + "step": 10774 + }, + { + "epoch": 1.9184472934472936, + "grad_norm": 0.5662530660629272, + "learning_rate": 0.00010657406734589686, + "loss": 0.8697, + "step": 10775 + }, + { + "epoch": 1.9186253561253561, + "grad_norm": 0.6640757322311401, + "learning_rate": 0.00010656010008951344, + "loss": 1.0636, + "step": 10776 + }, + { + "epoch": 1.9188034188034186, + "grad_norm": 0.6994011998176575, + "learning_rate": 0.00010654613270459848, + "loss": 0.9326, + "step": 10777 + }, + { + "epoch": 1.9189814814814814, + "grad_norm": 0.6827420592308044, + "learning_rate": 0.00010653216519142563, + "loss": 0.8667, + "step": 10778 + }, + { + "epoch": 1.9191595441595442, + "grad_norm": 0.6814691424369812, + "learning_rate": 0.00010651819755026862, + "loss": 0.828, + "step": 10779 + }, + { + "epoch": 1.919337606837607, + "grad_norm": 0.7033611536026001, + "learning_rate": 0.00010650422978140103, + "loss": 1.0427, + "step": 10780 + }, + { + "epoch": 1.9195156695156697, + "grad_norm": 0.7098833322525024, + "learning_rate": 0.00010649026188509657, + "loss": 1.1723, + "step": 10781 + }, + { + "epoch": 1.9196937321937322, + "grad_norm": 0.7184767723083496, + "learning_rate": 0.00010647629386162893, + "loss": 0.852, + "step": 10782 + }, + { + "epoch": 1.9198717948717947, + "grad_norm": 0.6682565808296204, + "learning_rate": 0.00010646232571127175, + "loss": 0.8827, + "step": 10783 + }, + { + "epoch": 1.9200498575498575, + "grad_norm": 0.6699280142784119, + "learning_rate": 0.00010644835743429873, + "loss": 0.8346, + "step": 10784 + }, + { + "epoch": 1.9202279202279202, + "grad_norm": 0.8041857481002808, + "learning_rate": 0.00010643438903098355, + "loss": 0.9622, + "step": 10785 + }, + { + "epoch": 1.920405982905983, + "grad_norm": 0.7315110564231873, + "learning_rate": 0.00010642042050159986, + "loss": 1.0443, + "step": 10786 + }, + { + "epoch": 1.9205840455840457, + "grad_norm": 0.5850204229354858, + "learning_rate": 0.0001064064518464214, + "loss": 1.0155, + "step": 10787 + }, + { + "epoch": 1.9207621082621082, + "grad_norm": 0.7320640683174133, + "learning_rate": 0.00010639248306572178, + "loss": 1.1556, + "step": 10788 + }, + { + "epoch": 1.9209401709401708, + "grad_norm": 0.689804196357727, + "learning_rate": 0.00010637851415977478, + "loss": 1.1058, + "step": 10789 + }, + { + "epoch": 1.9211182336182335, + "grad_norm": 0.6433262228965759, + "learning_rate": 0.000106364545128854, + "loss": 1.0916, + "step": 10790 + }, + { + "epoch": 1.9212962962962963, + "grad_norm": 0.6802626252174377, + "learning_rate": 0.00010635057597323323, + "loss": 1.126, + "step": 10791 + }, + { + "epoch": 1.921474358974359, + "grad_norm": 0.7503384351730347, + "learning_rate": 0.00010633660669318608, + "loss": 0.9354, + "step": 10792 + }, + { + "epoch": 1.9216524216524218, + "grad_norm": 0.6370253562927246, + "learning_rate": 0.00010632263728898629, + "loss": 0.9976, + "step": 10793 + }, + { + "epoch": 1.9218304843304843, + "grad_norm": 0.7566042542457581, + "learning_rate": 0.00010630866776090755, + "loss": 1.0311, + "step": 10794 + }, + { + "epoch": 1.922008547008547, + "grad_norm": 0.7011943459510803, + "learning_rate": 0.0001062946981092236, + "loss": 0.8777, + "step": 10795 + }, + { + "epoch": 1.9221866096866096, + "grad_norm": 0.6621114015579224, + "learning_rate": 0.00010628072833420811, + "loss": 0.9615, + "step": 10796 + }, + { + "epoch": 1.9223646723646723, + "grad_norm": 0.6863150000572205, + "learning_rate": 0.00010626675843613478, + "loss": 1.071, + "step": 10797 + }, + { + "epoch": 1.922542735042735, + "grad_norm": 0.597970724105835, + "learning_rate": 0.00010625278841527733, + "loss": 0.8661, + "step": 10798 + }, + { + "epoch": 1.9227207977207978, + "grad_norm": 0.5958755612373352, + "learning_rate": 0.00010623881827190947, + "loss": 0.9075, + "step": 10799 + }, + { + "epoch": 1.9228988603988604, + "grad_norm": 0.7764523029327393, + "learning_rate": 0.00010622484800630494, + "loss": 1.0576, + "step": 10800 + }, + { + "epoch": 1.9230769230769231, + "grad_norm": 0.774156391620636, + "learning_rate": 0.00010621087761873748, + "loss": 0.9273, + "step": 10801 + }, + { + "epoch": 1.9232549857549857, + "grad_norm": 0.6321687698364258, + "learning_rate": 0.00010619690710948074, + "loss": 0.8805, + "step": 10802 + }, + { + "epoch": 1.9234330484330484, + "grad_norm": 0.659538984298706, + "learning_rate": 0.00010618293647880846, + "loss": 0.9845, + "step": 10803 + }, + { + "epoch": 1.9236111111111112, + "grad_norm": 0.6931299567222595, + "learning_rate": 0.00010616896572699442, + "loss": 1.2005, + "step": 10804 + }, + { + "epoch": 1.923789173789174, + "grad_norm": 0.6054762005805969, + "learning_rate": 0.00010615499485431228, + "loss": 0.825, + "step": 10805 + }, + { + "epoch": 1.9239672364672364, + "grad_norm": 0.6631526947021484, + "learning_rate": 0.00010614102386103584, + "loss": 0.9149, + "step": 10806 + }, + { + "epoch": 1.9241452991452992, + "grad_norm": 0.6667893528938293, + "learning_rate": 0.00010612705274743878, + "loss": 1.014, + "step": 10807 + }, + { + "epoch": 1.9243233618233617, + "grad_norm": 0.861302375793457, + "learning_rate": 0.00010611308151379482, + "loss": 1.0809, + "step": 10808 + }, + { + "epoch": 1.9245014245014245, + "grad_norm": 0.6997994780540466, + "learning_rate": 0.00010609911016037777, + "loss": 0.8897, + "step": 10809 + }, + { + "epoch": 1.9246794871794872, + "grad_norm": 0.5689206123352051, + "learning_rate": 0.00010608513868746131, + "loss": 0.7517, + "step": 10810 + }, + { + "epoch": 1.92485754985755, + "grad_norm": 0.5972287654876709, + "learning_rate": 0.00010607116709531918, + "loss": 0.9015, + "step": 10811 + }, + { + "epoch": 1.9250356125356125, + "grad_norm": 0.7115643620491028, + "learning_rate": 0.00010605719538422519, + "loss": 0.6974, + "step": 10812 + }, + { + "epoch": 1.9252136752136753, + "grad_norm": 0.6548098921775818, + "learning_rate": 0.00010604322355445297, + "loss": 0.7075, + "step": 10813 + }, + { + "epoch": 1.9253917378917378, + "grad_norm": 0.6666337847709656, + "learning_rate": 0.00010602925160627639, + "loss": 1.0389, + "step": 10814 + }, + { + "epoch": 1.9255698005698005, + "grad_norm": 0.7754444479942322, + "learning_rate": 0.00010601527953996913, + "loss": 1.0674, + "step": 10815 + }, + { + "epoch": 1.9257478632478633, + "grad_norm": 0.6602712869644165, + "learning_rate": 0.00010600130735580498, + "loss": 1.2622, + "step": 10816 + }, + { + "epoch": 1.925925925925926, + "grad_norm": 0.6974020004272461, + "learning_rate": 0.00010598733505405767, + "loss": 0.9748, + "step": 10817 + }, + { + "epoch": 1.9261039886039886, + "grad_norm": 0.6236271858215332, + "learning_rate": 0.00010597336263500095, + "loss": 0.9463, + "step": 10818 + }, + { + "epoch": 1.9262820512820513, + "grad_norm": 0.6856079697608948, + "learning_rate": 0.00010595939009890859, + "loss": 0.9484, + "step": 10819 + }, + { + "epoch": 1.9264601139601139, + "grad_norm": 0.7300925850868225, + "learning_rate": 0.00010594541744605437, + "loss": 0.9702, + "step": 10820 + }, + { + "epoch": 1.9266381766381766, + "grad_norm": 0.6546478867530823, + "learning_rate": 0.00010593144467671208, + "loss": 0.8235, + "step": 10821 + }, + { + "epoch": 1.9268162393162394, + "grad_norm": 0.7215169072151184, + "learning_rate": 0.00010591747179115543, + "loss": 0.9986, + "step": 10822 + }, + { + "epoch": 1.926994301994302, + "grad_norm": 0.7304712533950806, + "learning_rate": 0.00010590349878965822, + "loss": 1.099, + "step": 10823 + }, + { + "epoch": 1.9271723646723646, + "grad_norm": 0.5883305668830872, + "learning_rate": 0.0001058895256724942, + "loss": 1.0647, + "step": 10824 + }, + { + "epoch": 1.9273504273504274, + "grad_norm": 0.8067272305488586, + "learning_rate": 0.00010587555243993716, + "loss": 1.0295, + "step": 10825 + }, + { + "epoch": 1.92752849002849, + "grad_norm": 0.6607550978660583, + "learning_rate": 0.00010586157909226089, + "loss": 0.8669, + "step": 10826 + }, + { + "epoch": 1.9277065527065527, + "grad_norm": 0.7256106734275818, + "learning_rate": 0.00010584760562973914, + "loss": 1.1674, + "step": 10827 + }, + { + "epoch": 1.9278846153846154, + "grad_norm": 0.6584621071815491, + "learning_rate": 0.00010583363205264574, + "loss": 0.8901, + "step": 10828 + }, + { + "epoch": 1.9280626780626782, + "grad_norm": 0.7200617790222168, + "learning_rate": 0.00010581965836125439, + "loss": 1.0463, + "step": 10829 + }, + { + "epoch": 1.9282407407407407, + "grad_norm": 0.7244223952293396, + "learning_rate": 0.00010580568455583894, + "loss": 1.0973, + "step": 10830 + }, + { + "epoch": 1.9284188034188035, + "grad_norm": 0.7678009867668152, + "learning_rate": 0.00010579171063667317, + "loss": 1.1753, + "step": 10831 + }, + { + "epoch": 1.928596866096866, + "grad_norm": 0.6455881595611572, + "learning_rate": 0.00010577773660403085, + "loss": 0.8988, + "step": 10832 + }, + { + "epoch": 1.9287749287749287, + "grad_norm": 0.6804864406585693, + "learning_rate": 0.0001057637624581858, + "loss": 0.8156, + "step": 10833 + }, + { + "epoch": 1.9289529914529915, + "grad_norm": 0.7874828577041626, + "learning_rate": 0.00010574978819941176, + "loss": 1.1876, + "step": 10834 + }, + { + "epoch": 1.9291310541310542, + "grad_norm": 0.7396490573883057, + "learning_rate": 0.00010573581382798261, + "loss": 0.8709, + "step": 10835 + }, + { + "epoch": 1.9293091168091168, + "grad_norm": 0.6800381541252136, + "learning_rate": 0.00010572183934417209, + "loss": 0.9906, + "step": 10836 + }, + { + "epoch": 1.9294871794871795, + "grad_norm": 0.7077754139900208, + "learning_rate": 0.000105707864748254, + "loss": 0.9785, + "step": 10837 + }, + { + "epoch": 1.929665242165242, + "grad_norm": 0.693249523639679, + "learning_rate": 0.00010569389004050216, + "loss": 0.9515, + "step": 10838 + }, + { + "epoch": 1.9298433048433048, + "grad_norm": 0.706924319267273, + "learning_rate": 0.00010567991522119037, + "loss": 1.074, + "step": 10839 + }, + { + "epoch": 1.9300213675213675, + "grad_norm": 0.6504101157188416, + "learning_rate": 0.00010566594029059244, + "loss": 1.0635, + "step": 10840 + }, + { + "epoch": 1.9301994301994303, + "grad_norm": 0.7620238661766052, + "learning_rate": 0.00010565196524898219, + "loss": 0.944, + "step": 10841 + }, + { + "epoch": 1.9303774928774928, + "grad_norm": 0.6713484525680542, + "learning_rate": 0.00010563799009663344, + "loss": 0.749, + "step": 10842 + }, + { + "epoch": 1.9305555555555556, + "grad_norm": 0.9279242157936096, + "learning_rate": 0.00010562401483381997, + "loss": 0.961, + "step": 10843 + }, + { + "epoch": 1.930733618233618, + "grad_norm": 0.6710723638534546, + "learning_rate": 0.00010561003946081558, + "loss": 1.1288, + "step": 10844 + }, + { + "epoch": 1.9309116809116809, + "grad_norm": 0.7751701474189758, + "learning_rate": 0.00010559606397789416, + "loss": 0.9435, + "step": 10845 + }, + { + "epoch": 1.9310897435897436, + "grad_norm": 0.6741766929626465, + "learning_rate": 0.00010558208838532948, + "loss": 1.0299, + "step": 10846 + }, + { + "epoch": 1.9312678062678064, + "grad_norm": 0.6988041400909424, + "learning_rate": 0.00010556811268339539, + "loss": 1.0236, + "step": 10847 + }, + { + "epoch": 1.931445868945869, + "grad_norm": 0.6353505253791809, + "learning_rate": 0.00010555413687236568, + "loss": 1.0361, + "step": 10848 + }, + { + "epoch": 1.9316239316239316, + "grad_norm": 0.7162703275680542, + "learning_rate": 0.0001055401609525142, + "loss": 1.0931, + "step": 10849 + }, + { + "epoch": 1.9318019943019942, + "grad_norm": 0.61545330286026, + "learning_rate": 0.00010552618492411476, + "loss": 0.8829, + "step": 10850 + }, + { + "epoch": 1.931980056980057, + "grad_norm": 0.6304612159729004, + "learning_rate": 0.00010551220878744124, + "loss": 0.8574, + "step": 10851 + }, + { + "epoch": 1.9321581196581197, + "grad_norm": 0.6372067928314209, + "learning_rate": 0.00010549823254276743, + "loss": 1.0949, + "step": 10852 + }, + { + "epoch": 1.9323361823361824, + "grad_norm": 0.6952856779098511, + "learning_rate": 0.00010548425619036715, + "loss": 0.9232, + "step": 10853 + }, + { + "epoch": 1.9325142450142452, + "grad_norm": 0.6510106325149536, + "learning_rate": 0.00010547027973051427, + "loss": 1.0753, + "step": 10854 + }, + { + "epoch": 1.9326923076923077, + "grad_norm": 0.6377716064453125, + "learning_rate": 0.00010545630316348263, + "loss": 0.8466, + "step": 10855 + }, + { + "epoch": 1.9328703703703702, + "grad_norm": 0.7366968393325806, + "learning_rate": 0.00010544232648954606, + "loss": 0.9351, + "step": 10856 + }, + { + "epoch": 1.933048433048433, + "grad_norm": 0.703652024269104, + "learning_rate": 0.00010542834970897843, + "loss": 1.0032, + "step": 10857 + }, + { + "epoch": 1.9332264957264957, + "grad_norm": 0.6685494780540466, + "learning_rate": 0.00010541437282205355, + "loss": 0.8818, + "step": 10858 + }, + { + "epoch": 1.9334045584045585, + "grad_norm": 0.6594362854957581, + "learning_rate": 0.00010540039582904527, + "loss": 0.9535, + "step": 10859 + }, + { + "epoch": 1.9335826210826212, + "grad_norm": 0.8003259301185608, + "learning_rate": 0.00010538641873022744, + "loss": 0.8852, + "step": 10860 + }, + { + "epoch": 1.9337606837606838, + "grad_norm": 0.6567012071609497, + "learning_rate": 0.00010537244152587393, + "loss": 1.0832, + "step": 10861 + }, + { + "epoch": 1.9339387464387463, + "grad_norm": 0.6714941263198853, + "learning_rate": 0.00010535846421625862, + "loss": 1.1047, + "step": 10862 + }, + { + "epoch": 1.934116809116809, + "grad_norm": 0.6998924612998962, + "learning_rate": 0.00010534448680165531, + "loss": 0.8827, + "step": 10863 + }, + { + "epoch": 1.9342948717948718, + "grad_norm": 0.6065765619277954, + "learning_rate": 0.0001053305092823379, + "loss": 0.5773, + "step": 10864 + }, + { + "epoch": 1.9344729344729346, + "grad_norm": 0.7678273916244507, + "learning_rate": 0.0001053165316585802, + "loss": 0.9199, + "step": 10865 + }, + { + "epoch": 1.9346509971509973, + "grad_norm": 0.7071540951728821, + "learning_rate": 0.00010530255393065613, + "loss": 1.0292, + "step": 10866 + }, + { + "epoch": 1.9348290598290598, + "grad_norm": 0.6329835057258606, + "learning_rate": 0.00010528857609883956, + "loss": 0.9915, + "step": 10867 + }, + { + "epoch": 1.9350071225071224, + "grad_norm": 0.6274038553237915, + "learning_rate": 0.00010527459816340427, + "loss": 0.8499, + "step": 10868 + }, + { + "epoch": 1.9351851851851851, + "grad_norm": 0.6564371585845947, + "learning_rate": 0.00010526062012462424, + "loss": 1.1707, + "step": 10869 + }, + { + "epoch": 1.9353632478632479, + "grad_norm": 0.8561269044876099, + "learning_rate": 0.00010524664198277326, + "loss": 1.148, + "step": 10870 + }, + { + "epoch": 1.9355413105413106, + "grad_norm": 0.6322671175003052, + "learning_rate": 0.00010523266373812521, + "loss": 0.9165, + "step": 10871 + }, + { + "epoch": 1.9357193732193734, + "grad_norm": 0.7602947354316711, + "learning_rate": 0.00010521868539095403, + "loss": 0.9647, + "step": 10872 + }, + { + "epoch": 1.935897435897436, + "grad_norm": 0.5962168574333191, + "learning_rate": 0.00010520470694153353, + "loss": 0.8585, + "step": 10873 + }, + { + "epoch": 1.9360754985754984, + "grad_norm": 0.7498637437820435, + "learning_rate": 0.00010519072839013757, + "loss": 0.9828, + "step": 10874 + }, + { + "epoch": 1.9362535612535612, + "grad_norm": 0.6841256022453308, + "learning_rate": 0.00010517674973704012, + "loss": 0.9991, + "step": 10875 + }, + { + "epoch": 1.936431623931624, + "grad_norm": 0.8281826972961426, + "learning_rate": 0.00010516277098251499, + "loss": 1.028, + "step": 10876 + }, + { + "epoch": 1.9366096866096867, + "grad_norm": 0.6673563718795776, + "learning_rate": 0.0001051487921268361, + "loss": 1.1594, + "step": 10877 + }, + { + "epoch": 1.9367877492877494, + "grad_norm": 0.7833667993545532, + "learning_rate": 0.00010513481317027733, + "loss": 0.7675, + "step": 10878 + }, + { + "epoch": 1.936965811965812, + "grad_norm": 0.6087225675582886, + "learning_rate": 0.00010512083411311253, + "loss": 0.7803, + "step": 10879 + }, + { + "epoch": 1.9371438746438745, + "grad_norm": 0.6758120656013489, + "learning_rate": 0.00010510685495561563, + "loss": 1.0621, + "step": 10880 + }, + { + "epoch": 1.9373219373219372, + "grad_norm": 0.6720096468925476, + "learning_rate": 0.00010509287569806055, + "loss": 0.8502, + "step": 10881 + }, + { + "epoch": 1.9375, + "grad_norm": 0.6233887672424316, + "learning_rate": 0.00010507889634072113, + "loss": 1.0127, + "step": 10882 + }, + { + "epoch": 1.9376780626780628, + "grad_norm": 0.667742908000946, + "learning_rate": 0.00010506491688387127, + "loss": 0.9086, + "step": 10883 + }, + { + "epoch": 1.9378561253561255, + "grad_norm": 0.6533677577972412, + "learning_rate": 0.00010505093732778492, + "loss": 0.9724, + "step": 10884 + }, + { + "epoch": 1.938034188034188, + "grad_norm": 0.7171359062194824, + "learning_rate": 0.00010503695767273591, + "loss": 0.9915, + "step": 10885 + }, + { + "epoch": 1.9382122507122506, + "grad_norm": 0.723655641078949, + "learning_rate": 0.0001050229779189982, + "loss": 0.8981, + "step": 10886 + }, + { + "epoch": 1.9383903133903133, + "grad_norm": 0.6863494515419006, + "learning_rate": 0.00010500899806684568, + "loss": 1.2577, + "step": 10887 + }, + { + "epoch": 1.938568376068376, + "grad_norm": 0.8174706697463989, + "learning_rate": 0.00010499501811655224, + "loss": 0.9848, + "step": 10888 + }, + { + "epoch": 1.9387464387464388, + "grad_norm": 0.6378024220466614, + "learning_rate": 0.00010498103806839179, + "loss": 0.8499, + "step": 10889 + }, + { + "epoch": 1.9389245014245016, + "grad_norm": 0.6734544634819031, + "learning_rate": 0.00010496705792263823, + "loss": 0.8446, + "step": 10890 + }, + { + "epoch": 1.939102564102564, + "grad_norm": 0.6802361607551575, + "learning_rate": 0.00010495307767956551, + "loss": 0.9285, + "step": 10891 + }, + { + "epoch": 1.9392806267806266, + "grad_norm": 0.7821299433708191, + "learning_rate": 0.00010493909733944752, + "loss": 1.08, + "step": 10892 + }, + { + "epoch": 1.9394586894586894, + "grad_norm": 0.6204990148544312, + "learning_rate": 0.00010492511690255818, + "loss": 0.7861, + "step": 10893 + }, + { + "epoch": 1.9396367521367521, + "grad_norm": 0.6386391520500183, + "learning_rate": 0.0001049111363691714, + "loss": 0.9162, + "step": 10894 + }, + { + "epoch": 1.9398148148148149, + "grad_norm": 0.6885092854499817, + "learning_rate": 0.0001048971557395611, + "loss": 1.0026, + "step": 10895 + }, + { + "epoch": 1.9399928774928776, + "grad_norm": 0.6962558627128601, + "learning_rate": 0.00010488317501400122, + "loss": 1.146, + "step": 10896 + }, + { + "epoch": 1.9401709401709402, + "grad_norm": 0.6283716559410095, + "learning_rate": 0.00010486919419276566, + "loss": 1.0268, + "step": 10897 + }, + { + "epoch": 1.9403490028490027, + "grad_norm": 0.7183622717857361, + "learning_rate": 0.00010485521327612835, + "loss": 1.0123, + "step": 10898 + }, + { + "epoch": 1.9405270655270654, + "grad_norm": 0.6354197263717651, + "learning_rate": 0.00010484123226436321, + "loss": 0.871, + "step": 10899 + }, + { + "epoch": 1.9407051282051282, + "grad_norm": 0.804358184337616, + "learning_rate": 0.00010482725115774421, + "loss": 1.1001, + "step": 10900 + }, + { + "epoch": 1.940883190883191, + "grad_norm": 0.6896754503250122, + "learning_rate": 0.00010481326995654524, + "loss": 1.0976, + "step": 10901 + }, + { + "epoch": 1.9410612535612537, + "grad_norm": 0.9108015894889832, + "learning_rate": 0.00010479928866104023, + "loss": 0.8785, + "step": 10902 + }, + { + "epoch": 1.9412393162393162, + "grad_norm": 0.6963121294975281, + "learning_rate": 0.00010478530727150316, + "loss": 1.0458, + "step": 10903 + }, + { + "epoch": 1.9414173789173788, + "grad_norm": 0.6657114624977112, + "learning_rate": 0.00010477132578820792, + "loss": 0.8188, + "step": 10904 + }, + { + "epoch": 1.9415954415954415, + "grad_norm": 0.671716034412384, + "learning_rate": 0.00010475734421142847, + "loss": 1.0915, + "step": 10905 + }, + { + "epoch": 1.9417735042735043, + "grad_norm": 0.6790717244148254, + "learning_rate": 0.0001047433625414387, + "loss": 0.9688, + "step": 10906 + }, + { + "epoch": 1.941951566951567, + "grad_norm": 0.6411764621734619, + "learning_rate": 0.00010472938077851264, + "loss": 1.0387, + "step": 10907 + }, + { + "epoch": 1.9421296296296298, + "grad_norm": 0.8579615950584412, + "learning_rate": 0.00010471539892292417, + "loss": 1.1635, + "step": 10908 + }, + { + "epoch": 1.9423076923076923, + "grad_norm": 0.7031029462814331, + "learning_rate": 0.00010470141697494726, + "loss": 0.9813, + "step": 10909 + }, + { + "epoch": 1.9424857549857548, + "grad_norm": 0.6657388806343079, + "learning_rate": 0.00010468743493485584, + "loss": 0.7947, + "step": 10910 + }, + { + "epoch": 1.9426638176638176, + "grad_norm": 0.6364194750785828, + "learning_rate": 0.00010467345280292389, + "loss": 0.8554, + "step": 10911 + }, + { + "epoch": 1.9428418803418803, + "grad_norm": 0.7394127249717712, + "learning_rate": 0.00010465947057942534, + "loss": 0.822, + "step": 10912 + }, + { + "epoch": 1.943019943019943, + "grad_norm": 0.6557473540306091, + "learning_rate": 0.00010464548826463411, + "loss": 1.0025, + "step": 10913 + }, + { + "epoch": 1.9431980056980058, + "grad_norm": 0.6530601382255554, + "learning_rate": 0.00010463150585882422, + "loss": 1.0828, + "step": 10914 + }, + { + "epoch": 1.9433760683760684, + "grad_norm": 0.7376404404640198, + "learning_rate": 0.00010461752336226957, + "loss": 0.9413, + "step": 10915 + }, + { + "epoch": 1.943554131054131, + "grad_norm": 0.7110656499862671, + "learning_rate": 0.00010460354077524417, + "loss": 0.9162, + "step": 10916 + }, + { + "epoch": 1.9437321937321936, + "grad_norm": 0.6515666246414185, + "learning_rate": 0.00010458955809802194, + "loss": 0.9211, + "step": 10917 + }, + { + "epoch": 1.9439102564102564, + "grad_norm": 0.6888720989227295, + "learning_rate": 0.00010457557533087683, + "loss": 1.0632, + "step": 10918 + }, + { + "epoch": 1.9440883190883191, + "grad_norm": 0.7246627807617188, + "learning_rate": 0.00010456159247408286, + "loss": 0.9807, + "step": 10919 + }, + { + "epoch": 1.944266381766382, + "grad_norm": 0.727834165096283, + "learning_rate": 0.00010454760952791394, + "loss": 1.0793, + "step": 10920 + }, + { + "epoch": 1.9444444444444444, + "grad_norm": 0.6365306377410889, + "learning_rate": 0.00010453362649264407, + "loss": 1.0415, + "step": 10921 + }, + { + "epoch": 1.9446225071225072, + "grad_norm": 0.7187839150428772, + "learning_rate": 0.0001045196433685472, + "loss": 1.007, + "step": 10922 + }, + { + "epoch": 1.9448005698005697, + "grad_norm": 0.5905138254165649, + "learning_rate": 0.00010450566015589732, + "loss": 0.9818, + "step": 10923 + }, + { + "epoch": 1.9449786324786325, + "grad_norm": 0.7008894085884094, + "learning_rate": 0.00010449167685496837, + "loss": 0.8444, + "step": 10924 + }, + { + "epoch": 1.9451566951566952, + "grad_norm": 0.6126312017440796, + "learning_rate": 0.00010447769346603435, + "loss": 0.7207, + "step": 10925 + }, + { + "epoch": 1.945334757834758, + "grad_norm": 0.7513176202774048, + "learning_rate": 0.00010446370998936922, + "loss": 0.8693, + "step": 10926 + }, + { + "epoch": 1.9455128205128205, + "grad_norm": 0.6382531523704529, + "learning_rate": 0.00010444972642524697, + "loss": 0.8379, + "step": 10927 + }, + { + "epoch": 1.9456908831908832, + "grad_norm": 0.7062170505523682, + "learning_rate": 0.0001044357427739416, + "loss": 1.0525, + "step": 10928 + }, + { + "epoch": 1.9458689458689458, + "grad_norm": 0.6954067349433899, + "learning_rate": 0.00010442175903572703, + "loss": 1.0238, + "step": 10929 + }, + { + "epoch": 1.9460470085470085, + "grad_norm": 0.7257117033004761, + "learning_rate": 0.00010440777521087731, + "loss": 1.1413, + "step": 10930 + }, + { + "epoch": 1.9462250712250713, + "grad_norm": 0.6617701053619385, + "learning_rate": 0.00010439379129966635, + "loss": 1.0089, + "step": 10931 + }, + { + "epoch": 1.946403133903134, + "grad_norm": 0.6860800385475159, + "learning_rate": 0.00010437980730236821, + "loss": 1.1778, + "step": 10932 + }, + { + "epoch": 1.9465811965811965, + "grad_norm": 0.846235454082489, + "learning_rate": 0.00010436582321925684, + "loss": 0.9851, + "step": 10933 + }, + { + "epoch": 1.9467592592592593, + "grad_norm": 0.6385617852210999, + "learning_rate": 0.00010435183905060623, + "loss": 0.9542, + "step": 10934 + }, + { + "epoch": 1.9469373219373218, + "grad_norm": 0.7137401700019836, + "learning_rate": 0.00010433785479669038, + "loss": 1.0499, + "step": 10935 + }, + { + "epoch": 1.9471153846153846, + "grad_norm": 0.6269308924674988, + "learning_rate": 0.00010432387045778324, + "loss": 0.8929, + "step": 10936 + }, + { + "epoch": 1.9472934472934473, + "grad_norm": 0.7903163433074951, + "learning_rate": 0.00010430988603415888, + "loss": 0.9812, + "step": 10937 + }, + { + "epoch": 1.94747150997151, + "grad_norm": 0.6006736159324646, + "learning_rate": 0.00010429590152609121, + "loss": 0.7959, + "step": 10938 + }, + { + "epoch": 1.9476495726495726, + "grad_norm": 0.6061521768569946, + "learning_rate": 0.00010428191693385431, + "loss": 0.8748, + "step": 10939 + }, + { + "epoch": 1.9478276353276354, + "grad_norm": 0.6637623906135559, + "learning_rate": 0.00010426793225772216, + "loss": 0.7047, + "step": 10940 + }, + { + "epoch": 1.948005698005698, + "grad_norm": 0.7650586366653442, + "learning_rate": 0.00010425394749796874, + "loss": 1.0018, + "step": 10941 + }, + { + "epoch": 1.9481837606837606, + "grad_norm": 0.6575125455856323, + "learning_rate": 0.000104239962654868, + "loss": 0.8915, + "step": 10942 + }, + { + "epoch": 1.9483618233618234, + "grad_norm": 0.6315393447875977, + "learning_rate": 0.00010422597772869404, + "loss": 1.1884, + "step": 10943 + }, + { + "epoch": 1.9485398860398861, + "grad_norm": 0.7607148885726929, + "learning_rate": 0.00010421199271972083, + "loss": 0.9341, + "step": 10944 + }, + { + "epoch": 1.9487179487179487, + "grad_norm": 0.6491827964782715, + "learning_rate": 0.00010419800762822239, + "loss": 0.9991, + "step": 10945 + }, + { + "epoch": 1.9488960113960114, + "grad_norm": 0.6294243335723877, + "learning_rate": 0.00010418402245447265, + "loss": 0.9253, + "step": 10946 + }, + { + "epoch": 1.949074074074074, + "grad_norm": 0.6472215056419373, + "learning_rate": 0.00010417003719874571, + "loss": 1.0402, + "step": 10947 + }, + { + "epoch": 1.9492521367521367, + "grad_norm": 0.7377899885177612, + "learning_rate": 0.00010415605186131559, + "loss": 1.046, + "step": 10948 + }, + { + "epoch": 1.9494301994301995, + "grad_norm": 0.6391907334327698, + "learning_rate": 0.00010414206644245623, + "loss": 0.8529, + "step": 10949 + }, + { + "epoch": 1.9496082621082622, + "grad_norm": 0.7101355195045471, + "learning_rate": 0.0001041280809424417, + "loss": 0.925, + "step": 10950 + }, + { + "epoch": 1.9497863247863247, + "grad_norm": 0.7891978025436401, + "learning_rate": 0.00010411409536154597, + "loss": 1.0691, + "step": 10951 + }, + { + "epoch": 1.9499643874643875, + "grad_norm": 0.7225242853164673, + "learning_rate": 0.00010410010970004311, + "loss": 1.158, + "step": 10952 + }, + { + "epoch": 1.95014245014245, + "grad_norm": 0.6073256731033325, + "learning_rate": 0.00010408612395820714, + "loss": 0.9977, + "step": 10953 + }, + { + "epoch": 1.9503205128205128, + "grad_norm": 0.6373769044876099, + "learning_rate": 0.00010407213813631203, + "loss": 1.019, + "step": 10954 + }, + { + "epoch": 1.9504985754985755, + "grad_norm": 0.7451884746551514, + "learning_rate": 0.00010405815223463184, + "loss": 0.9497, + "step": 10955 + }, + { + "epoch": 1.9506766381766383, + "grad_norm": 0.7760418057441711, + "learning_rate": 0.00010404416625344058, + "loss": 1.0378, + "step": 10956 + }, + { + "epoch": 1.9508547008547008, + "grad_norm": 0.7057808041572571, + "learning_rate": 0.00010403018019301228, + "loss": 0.8953, + "step": 10957 + }, + { + "epoch": 1.9510327635327636, + "grad_norm": 0.6599584817886353, + "learning_rate": 0.00010401619405362095, + "loss": 0.8859, + "step": 10958 + }, + { + "epoch": 1.951210826210826, + "grad_norm": 0.6977253556251526, + "learning_rate": 0.00010400220783554069, + "loss": 0.9038, + "step": 10959 + }, + { + "epoch": 1.9513888888888888, + "grad_norm": 0.6930267810821533, + "learning_rate": 0.00010398822153904546, + "loss": 1.1547, + "step": 10960 + }, + { + "epoch": 1.9515669515669516, + "grad_norm": 0.6301694512367249, + "learning_rate": 0.00010397423516440931, + "loss": 0.8875, + "step": 10961 + }, + { + "epoch": 1.9517450142450143, + "grad_norm": 0.7447484135627747, + "learning_rate": 0.00010396024871190628, + "loss": 1.0454, + "step": 10962 + }, + { + "epoch": 1.9519230769230769, + "grad_norm": 0.8666765093803406, + "learning_rate": 0.00010394626218181041, + "loss": 1.2211, + "step": 10963 + }, + { + "epoch": 1.9521011396011396, + "grad_norm": 0.599354088306427, + "learning_rate": 0.00010393227557439573, + "loss": 1.0419, + "step": 10964 + }, + { + "epoch": 1.9522792022792022, + "grad_norm": 0.6991702914237976, + "learning_rate": 0.00010391828888993627, + "loss": 0.8217, + "step": 10965 + }, + { + "epoch": 1.952457264957265, + "grad_norm": 0.7467028498649597, + "learning_rate": 0.0001039043021287061, + "loss": 0.8708, + "step": 10966 + }, + { + "epoch": 1.9526353276353277, + "grad_norm": 0.6806215047836304, + "learning_rate": 0.0001038903152909792, + "loss": 1.218, + "step": 10967 + }, + { + "epoch": 1.9528133903133904, + "grad_norm": 0.6704212427139282, + "learning_rate": 0.00010387632837702968, + "loss": 0.8428, + "step": 10968 + }, + { + "epoch": 1.952991452991453, + "grad_norm": 0.6843154430389404, + "learning_rate": 0.00010386234138713155, + "loss": 0.9729, + "step": 10969 + }, + { + "epoch": 1.9531695156695157, + "grad_norm": 0.6619821190834045, + "learning_rate": 0.00010384835432155888, + "loss": 1.021, + "step": 10970 + }, + { + "epoch": 1.9533475783475782, + "grad_norm": 0.6249803900718689, + "learning_rate": 0.0001038343671805857, + "loss": 0.9321, + "step": 10971 + }, + { + "epoch": 1.953525641025641, + "grad_norm": 0.7361689805984497, + "learning_rate": 0.00010382037996448604, + "loss": 0.9451, + "step": 10972 + }, + { + "epoch": 1.9537037037037037, + "grad_norm": 0.6464847922325134, + "learning_rate": 0.00010380639267353398, + "loss": 1.0188, + "step": 10973 + }, + { + "epoch": 1.9538817663817665, + "grad_norm": 0.5975635647773743, + "learning_rate": 0.00010379240530800356, + "loss": 0.9025, + "step": 10974 + }, + { + "epoch": 1.9540598290598292, + "grad_norm": 0.6734475493431091, + "learning_rate": 0.00010377841786816884, + "loss": 1.0742, + "step": 10975 + }, + { + "epoch": 1.9542378917378918, + "grad_norm": 0.7318592667579651, + "learning_rate": 0.00010376443035430386, + "loss": 1.1082, + "step": 10976 + }, + { + "epoch": 1.9544159544159543, + "grad_norm": 0.7696142792701721, + "learning_rate": 0.00010375044276668271, + "loss": 0.8421, + "step": 10977 + }, + { + "epoch": 1.954594017094017, + "grad_norm": 0.68442302942276, + "learning_rate": 0.00010373645510557939, + "loss": 1.0794, + "step": 10978 + }, + { + "epoch": 1.9547720797720798, + "grad_norm": 0.7582547068595886, + "learning_rate": 0.00010372246737126801, + "loss": 1.0332, + "step": 10979 + }, + { + "epoch": 1.9549501424501425, + "grad_norm": 0.6529998183250427, + "learning_rate": 0.00010370847956402262, + "loss": 1.1833, + "step": 10980 + }, + { + "epoch": 1.9551282051282053, + "grad_norm": 0.7565605044364929, + "learning_rate": 0.00010369449168411729, + "loss": 1.0494, + "step": 10981 + }, + { + "epoch": 1.9553062678062678, + "grad_norm": 0.6346915364265442, + "learning_rate": 0.00010368050373182605, + "loss": 1.0052, + "step": 10982 + }, + { + "epoch": 1.9554843304843303, + "grad_norm": 0.7021830081939697, + "learning_rate": 0.00010366651570742298, + "loss": 0.9716, + "step": 10983 + }, + { + "epoch": 1.955662393162393, + "grad_norm": 0.6464530825614929, + "learning_rate": 0.00010365252761118218, + "loss": 0.9802, + "step": 10984 + }, + { + "epoch": 1.9558404558404558, + "grad_norm": 0.6845090985298157, + "learning_rate": 0.00010363853944337768, + "loss": 0.9529, + "step": 10985 + }, + { + "epoch": 1.9560185185185186, + "grad_norm": 0.7178115248680115, + "learning_rate": 0.00010362455120428356, + "loss": 0.9968, + "step": 10986 + }, + { + "epoch": 1.9561965811965814, + "grad_norm": 0.6131038069725037, + "learning_rate": 0.00010361056289417385, + "loss": 1.0559, + "step": 10987 + }, + { + "epoch": 1.9563746438746439, + "grad_norm": 0.6946909427642822, + "learning_rate": 0.0001035965745133227, + "loss": 1.0457, + "step": 10988 + }, + { + "epoch": 1.9565527065527064, + "grad_norm": 0.7376706600189209, + "learning_rate": 0.00010358258606200413, + "loss": 0.7775, + "step": 10989 + }, + { + "epoch": 1.9567307692307692, + "grad_norm": 0.6864920854568481, + "learning_rate": 0.00010356859754049225, + "loss": 0.8798, + "step": 10990 + }, + { + "epoch": 1.956908831908832, + "grad_norm": 0.6301153302192688, + "learning_rate": 0.0001035546089490611, + "loss": 0.8757, + "step": 10991 + }, + { + "epoch": 1.9570868945868947, + "grad_norm": 0.7184807062149048, + "learning_rate": 0.00010354062028798474, + "loss": 1.0783, + "step": 10992 + }, + { + "epoch": 1.9572649572649574, + "grad_norm": 0.7138563394546509, + "learning_rate": 0.00010352663155753732, + "loss": 1.0328, + "step": 10993 + }, + { + "epoch": 1.95744301994302, + "grad_norm": 0.6565547585487366, + "learning_rate": 0.00010351264275799286, + "loss": 1.1312, + "step": 10994 + }, + { + "epoch": 1.9576210826210825, + "grad_norm": 0.7055862545967102, + "learning_rate": 0.00010349865388962547, + "loss": 1.0787, + "step": 10995 + }, + { + "epoch": 1.9577991452991452, + "grad_norm": 0.6184022426605225, + "learning_rate": 0.00010348466495270926, + "loss": 0.9635, + "step": 10996 + }, + { + "epoch": 1.957977207977208, + "grad_norm": 0.6563652753829956, + "learning_rate": 0.0001034706759475182, + "loss": 0.772, + "step": 10997 + }, + { + "epoch": 1.9581552706552707, + "grad_norm": 0.6103591322898865, + "learning_rate": 0.00010345668687432651, + "loss": 0.8113, + "step": 10998 + }, + { + "epoch": 1.9583333333333335, + "grad_norm": 0.6715512275695801, + "learning_rate": 0.0001034426977334082, + "loss": 1.1841, + "step": 10999 + }, + { + "epoch": 1.958511396011396, + "grad_norm": 0.680092453956604, + "learning_rate": 0.00010342870852503739, + "loss": 0.9992, + "step": 11000 + }, + { + "epoch": 1.9586894586894585, + "grad_norm": 0.828472375869751, + "learning_rate": 0.00010341471924948816, + "loss": 1.0975, + "step": 11001 + }, + { + "epoch": 1.9588675213675213, + "grad_norm": 0.758441686630249, + "learning_rate": 0.00010340072990703463, + "loss": 1.0632, + "step": 11002 + }, + { + "epoch": 1.959045584045584, + "grad_norm": 0.6847560405731201, + "learning_rate": 0.00010338674049795079, + "loss": 1.0054, + "step": 11003 + }, + { + "epoch": 1.9592236467236468, + "grad_norm": 0.707626223564148, + "learning_rate": 0.00010337275102251085, + "loss": 0.9427, + "step": 11004 + }, + { + "epoch": 1.9594017094017095, + "grad_norm": 0.769036591053009, + "learning_rate": 0.00010335876148098887, + "loss": 1.0424, + "step": 11005 + }, + { + "epoch": 1.959579772079772, + "grad_norm": 0.822695791721344, + "learning_rate": 0.00010334477187365892, + "loss": 1.1573, + "step": 11006 + }, + { + "epoch": 1.9597578347578346, + "grad_norm": 0.6290286183357239, + "learning_rate": 0.00010333078220079513, + "loss": 0.936, + "step": 11007 + }, + { + "epoch": 1.9599358974358974, + "grad_norm": 0.6802252531051636, + "learning_rate": 0.00010331679246267155, + "loss": 0.8049, + "step": 11008 + }, + { + "epoch": 1.96011396011396, + "grad_norm": 0.6652607321739197, + "learning_rate": 0.00010330280265956232, + "loss": 0.926, + "step": 11009 + }, + { + "epoch": 1.9602920227920229, + "grad_norm": 0.7057216763496399, + "learning_rate": 0.00010328881279174154, + "loss": 0.9464, + "step": 11010 + }, + { + "epoch": 1.9604700854700856, + "grad_norm": 0.6951601505279541, + "learning_rate": 0.00010327482285948331, + "loss": 0.9882, + "step": 11011 + }, + { + "epoch": 1.9606481481481481, + "grad_norm": 0.6537632942199707, + "learning_rate": 0.00010326083286306174, + "loss": 0.8663, + "step": 11012 + }, + { + "epoch": 1.9608262108262107, + "grad_norm": 0.7252047657966614, + "learning_rate": 0.0001032468428027509, + "loss": 1.1377, + "step": 11013 + }, + { + "epoch": 1.9610042735042734, + "grad_norm": 0.6494104266166687, + "learning_rate": 0.00010323285267882492, + "loss": 0.8072, + "step": 11014 + }, + { + "epoch": 1.9611823361823362, + "grad_norm": 0.8463460206985474, + "learning_rate": 0.00010321886249155792, + "loss": 1.22, + "step": 11015 + }, + { + "epoch": 1.961360398860399, + "grad_norm": 0.6071396470069885, + "learning_rate": 0.00010320487224122401, + "loss": 0.7975, + "step": 11016 + }, + { + "epoch": 1.9615384615384617, + "grad_norm": 0.6546960473060608, + "learning_rate": 0.00010319088192809725, + "loss": 1.1729, + "step": 11017 + }, + { + "epoch": 1.9617165242165242, + "grad_norm": 0.7399442791938782, + "learning_rate": 0.00010317689155245178, + "loss": 1.092, + "step": 11018 + }, + { + "epoch": 1.9618945868945867, + "grad_norm": 0.7103837728500366, + "learning_rate": 0.00010316290111456175, + "loss": 0.8436, + "step": 11019 + }, + { + "epoch": 1.9620726495726495, + "grad_norm": 0.6990065574645996, + "learning_rate": 0.00010314891061470125, + "loss": 0.9003, + "step": 11020 + }, + { + "epoch": 1.9622507122507122, + "grad_norm": 0.7945666313171387, + "learning_rate": 0.00010313492005314438, + "loss": 0.8812, + "step": 11021 + }, + { + "epoch": 1.962428774928775, + "grad_norm": 0.6177538633346558, + "learning_rate": 0.00010312092943016527, + "loss": 1.0091, + "step": 11022 + }, + { + "epoch": 1.9626068376068377, + "grad_norm": 0.7260771989822388, + "learning_rate": 0.000103106938746038, + "loss": 0.9376, + "step": 11023 + }, + { + "epoch": 1.9627849002849003, + "grad_norm": 0.6726518273353577, + "learning_rate": 0.00010309294800103674, + "loss": 0.8048, + "step": 11024 + }, + { + "epoch": 1.9629629629629628, + "grad_norm": 0.8759992122650146, + "learning_rate": 0.00010307895719543562, + "loss": 1.0248, + "step": 11025 + }, + { + "epoch": 1.9631410256410255, + "grad_norm": 0.683437168598175, + "learning_rate": 0.00010306496632950868, + "loss": 1.0314, + "step": 11026 + }, + { + "epoch": 1.9633190883190883, + "grad_norm": 0.7255756258964539, + "learning_rate": 0.00010305097540353012, + "loss": 0.9828, + "step": 11027 + }, + { + "epoch": 1.963497150997151, + "grad_norm": 0.6904804706573486, + "learning_rate": 0.000103036984417774, + "loss": 0.9054, + "step": 11028 + }, + { + "epoch": 1.9636752136752138, + "grad_norm": 0.6906846761703491, + "learning_rate": 0.00010302299337251451, + "loss": 1.0287, + "step": 11029 + }, + { + "epoch": 1.9638532763532763, + "grad_norm": 0.6677078008651733, + "learning_rate": 0.00010300900226802575, + "loss": 0.8742, + "step": 11030 + }, + { + "epoch": 1.964031339031339, + "grad_norm": 0.6144888997077942, + "learning_rate": 0.00010299501110458183, + "loss": 0.6942, + "step": 11031 + }, + { + "epoch": 1.9642094017094016, + "grad_norm": 0.753010094165802, + "learning_rate": 0.0001029810198824569, + "loss": 0.9018, + "step": 11032 + }, + { + "epoch": 1.9643874643874644, + "grad_norm": 0.6872276663780212, + "learning_rate": 0.00010296702860192505, + "loss": 1.1647, + "step": 11033 + }, + { + "epoch": 1.9645655270655271, + "grad_norm": 0.709000289440155, + "learning_rate": 0.00010295303726326047, + "loss": 0.9143, + "step": 11034 + }, + { + "epoch": 1.9647435897435899, + "grad_norm": 0.6507021188735962, + "learning_rate": 0.00010293904586673723, + "loss": 1.006, + "step": 11035 + }, + { + "epoch": 1.9649216524216524, + "grad_norm": 0.6789946556091309, + "learning_rate": 0.00010292505441262952, + "loss": 0.9049, + "step": 11036 + }, + { + "epoch": 1.9650997150997151, + "grad_norm": 0.7156081795692444, + "learning_rate": 0.00010291106290121143, + "loss": 0.9195, + "step": 11037 + }, + { + "epoch": 1.9652777777777777, + "grad_norm": 0.6770932078361511, + "learning_rate": 0.0001028970713327571, + "loss": 0.9524, + "step": 11038 + }, + { + "epoch": 1.9654558404558404, + "grad_norm": 0.7304288148880005, + "learning_rate": 0.00010288307970754067, + "loss": 0.9276, + "step": 11039 + }, + { + "epoch": 1.9656339031339032, + "grad_norm": 0.7603645324707031, + "learning_rate": 0.0001028690880258363, + "loss": 1.2157, + "step": 11040 + }, + { + "epoch": 1.965811965811966, + "grad_norm": 0.6875246167182922, + "learning_rate": 0.00010285509628791811, + "loss": 1.0269, + "step": 11041 + }, + { + "epoch": 1.9659900284900285, + "grad_norm": 0.7234818935394287, + "learning_rate": 0.00010284110449406026, + "loss": 0.9695, + "step": 11042 + }, + { + "epoch": 1.9661680911680912, + "grad_norm": 0.7322804927825928, + "learning_rate": 0.00010282711264453684, + "loss": 0.9752, + "step": 11043 + }, + { + "epoch": 1.9663461538461537, + "grad_norm": 0.7524822950363159, + "learning_rate": 0.00010281312073962202, + "loss": 1.2144, + "step": 11044 + }, + { + "epoch": 1.9665242165242165, + "grad_norm": 0.6623101234436035, + "learning_rate": 0.00010279912877958995, + "loss": 1.1334, + "step": 11045 + }, + { + "epoch": 1.9667022792022792, + "grad_norm": 0.7814893126487732, + "learning_rate": 0.00010278513676471477, + "loss": 1.266, + "step": 11046 + }, + { + "epoch": 1.966880341880342, + "grad_norm": 0.7129884362220764, + "learning_rate": 0.00010277114469527063, + "loss": 1.0918, + "step": 11047 + }, + { + "epoch": 1.9670584045584045, + "grad_norm": 0.6996828317642212, + "learning_rate": 0.00010275715257153164, + "loss": 0.9269, + "step": 11048 + }, + { + "epoch": 1.9672364672364673, + "grad_norm": 0.6439059972763062, + "learning_rate": 0.00010274316039377198, + "loss": 1.1998, + "step": 11049 + }, + { + "epoch": 1.9674145299145298, + "grad_norm": 0.6837672591209412, + "learning_rate": 0.00010272916816226581, + "loss": 0.8899, + "step": 11050 + }, + { + "epoch": 1.9675925925925926, + "grad_norm": 0.702583909034729, + "learning_rate": 0.00010271517587728726, + "loss": 1.1862, + "step": 11051 + }, + { + "epoch": 1.9677706552706553, + "grad_norm": 0.6627798676490784, + "learning_rate": 0.00010270118353911047, + "loss": 0.898, + "step": 11052 + }, + { + "epoch": 1.967948717948718, + "grad_norm": 0.7628579139709473, + "learning_rate": 0.00010268719114800957, + "loss": 1.006, + "step": 11053 + }, + { + "epoch": 1.9681267806267806, + "grad_norm": 0.6425395607948303, + "learning_rate": 0.00010267319870425877, + "loss": 0.962, + "step": 11054 + }, + { + "epoch": 1.9683048433048433, + "grad_norm": 0.7462666630744934, + "learning_rate": 0.00010265920620813219, + "loss": 1.0703, + "step": 11055 + }, + { + "epoch": 1.9684829059829059, + "grad_norm": 0.67641681432724, + "learning_rate": 0.00010264521365990401, + "loss": 1.1077, + "step": 11056 + }, + { + "epoch": 1.9686609686609686, + "grad_norm": 0.6716381311416626, + "learning_rate": 0.0001026312210598483, + "loss": 1.1048, + "step": 11057 + }, + { + "epoch": 1.9688390313390314, + "grad_norm": 0.7207448482513428, + "learning_rate": 0.00010261722840823935, + "loss": 0.9236, + "step": 11058 + }, + { + "epoch": 1.9690170940170941, + "grad_norm": 0.7208544015884399, + "learning_rate": 0.0001026032357053512, + "loss": 1.0814, + "step": 11059 + }, + { + "epoch": 1.9691951566951567, + "grad_norm": 0.6076363325119019, + "learning_rate": 0.00010258924295145807, + "loss": 0.9388, + "step": 11060 + }, + { + "epoch": 1.9693732193732194, + "grad_norm": 0.6460439562797546, + "learning_rate": 0.00010257525014683411, + "loss": 0.9506, + "step": 11061 + }, + { + "epoch": 1.969551282051282, + "grad_norm": 0.7449939250946045, + "learning_rate": 0.00010256125729175348, + "loss": 1.0209, + "step": 11062 + }, + { + "epoch": 1.9697293447293447, + "grad_norm": 0.640885055065155, + "learning_rate": 0.00010254726438649031, + "loss": 1.0235, + "step": 11063 + }, + { + "epoch": 1.9699074074074074, + "grad_norm": 0.6872261166572571, + "learning_rate": 0.00010253327143131879, + "loss": 0.9217, + "step": 11064 + }, + { + "epoch": 1.9700854700854702, + "grad_norm": 0.6213285326957703, + "learning_rate": 0.0001025192784265131, + "loss": 0.8204, + "step": 11065 + }, + { + "epoch": 1.9702635327635327, + "grad_norm": 0.6594449281692505, + "learning_rate": 0.00010250528537234736, + "loss": 0.9789, + "step": 11066 + }, + { + "epoch": 1.9704415954415955, + "grad_norm": 0.7098729610443115, + "learning_rate": 0.00010249129226909577, + "loss": 1.2551, + "step": 11067 + }, + { + "epoch": 1.970619658119658, + "grad_norm": 0.7455953359603882, + "learning_rate": 0.0001024772991170325, + "loss": 1.0281, + "step": 11068 + }, + { + "epoch": 1.9707977207977208, + "grad_norm": 0.6657416224479675, + "learning_rate": 0.00010246330591643166, + "loss": 0.9421, + "step": 11069 + }, + { + "epoch": 1.9709757834757835, + "grad_norm": 0.6480659246444702, + "learning_rate": 0.00010244931266756748, + "loss": 0.9424, + "step": 11070 + }, + { + "epoch": 1.9711538461538463, + "grad_norm": 0.6440510749816895, + "learning_rate": 0.00010243531937071411, + "loss": 0.9651, + "step": 11071 + }, + { + "epoch": 1.9713319088319088, + "grad_norm": 0.6329794526100159, + "learning_rate": 0.00010242132602614571, + "loss": 0.9233, + "step": 11072 + }, + { + "epoch": 1.9715099715099715, + "grad_norm": 0.6694819927215576, + "learning_rate": 0.00010240733263413646, + "loss": 0.884, + "step": 11073 + }, + { + "epoch": 1.971688034188034, + "grad_norm": 0.7702556848526001, + "learning_rate": 0.0001023933391949605, + "loss": 1.216, + "step": 11074 + }, + { + "epoch": 1.9718660968660968, + "grad_norm": 0.6587536931037903, + "learning_rate": 0.00010237934570889207, + "loss": 0.9324, + "step": 11075 + }, + { + "epoch": 1.9720441595441596, + "grad_norm": 0.7919837832450867, + "learning_rate": 0.00010236535217620529, + "loss": 1.0011, + "step": 11076 + }, + { + "epoch": 1.9722222222222223, + "grad_norm": 0.6604606509208679, + "learning_rate": 0.00010235135859717433, + "loss": 0.929, + "step": 11077 + }, + { + "epoch": 1.9724002849002849, + "grad_norm": 0.7158446907997131, + "learning_rate": 0.0001023373649720734, + "loss": 0.8912, + "step": 11078 + }, + { + "epoch": 1.9725783475783476, + "grad_norm": 0.7450904846191406, + "learning_rate": 0.00010232337130117666, + "loss": 1.0782, + "step": 11079 + }, + { + "epoch": 1.9727564102564101, + "grad_norm": 0.6687077283859253, + "learning_rate": 0.00010230937758475827, + "loss": 1.0662, + "step": 11080 + }, + { + "epoch": 1.9729344729344729, + "grad_norm": 0.7188364267349243, + "learning_rate": 0.00010229538382309245, + "loss": 1.024, + "step": 11081 + }, + { + "epoch": 1.9731125356125356, + "grad_norm": 0.6787814497947693, + "learning_rate": 0.00010228139001645334, + "loss": 0.9559, + "step": 11082 + }, + { + "epoch": 1.9732905982905984, + "grad_norm": 0.6834072470664978, + "learning_rate": 0.00010226739616511513, + "loss": 0.8143, + "step": 11083 + }, + { + "epoch": 1.973468660968661, + "grad_norm": 0.6651090979576111, + "learning_rate": 0.00010225340226935201, + "loss": 1.05, + "step": 11084 + }, + { + "epoch": 1.9736467236467237, + "grad_norm": 0.7125018835067749, + "learning_rate": 0.00010223940832943813, + "loss": 1.0275, + "step": 11085 + }, + { + "epoch": 1.9738247863247862, + "grad_norm": 0.6886870861053467, + "learning_rate": 0.00010222541434564772, + "loss": 1.0972, + "step": 11086 + }, + { + "epoch": 1.974002849002849, + "grad_norm": 0.7068913578987122, + "learning_rate": 0.00010221142031825492, + "loss": 0.9248, + "step": 11087 + }, + { + "epoch": 1.9741809116809117, + "grad_norm": 0.7752319574356079, + "learning_rate": 0.00010219742624753397, + "loss": 0.9754, + "step": 11088 + }, + { + "epoch": 1.9743589743589745, + "grad_norm": 0.7915459871292114, + "learning_rate": 0.00010218343213375896, + "loss": 1.2589, + "step": 11089 + }, + { + "epoch": 1.9745370370370372, + "grad_norm": 0.6597068309783936, + "learning_rate": 0.00010216943797720418, + "loss": 1.0004, + "step": 11090 + }, + { + "epoch": 1.9747150997150997, + "grad_norm": 0.7060620188713074, + "learning_rate": 0.00010215544377814375, + "loss": 0.9968, + "step": 11091 + }, + { + "epoch": 1.9748931623931623, + "grad_norm": 0.6815677881240845, + "learning_rate": 0.0001021414495368519, + "loss": 0.8889, + "step": 11092 + }, + { + "epoch": 1.975071225071225, + "grad_norm": 0.6872935891151428, + "learning_rate": 0.00010212745525360277, + "loss": 1.1582, + "step": 11093 + }, + { + "epoch": 1.9752492877492878, + "grad_norm": 0.6781140565872192, + "learning_rate": 0.00010211346092867056, + "loss": 0.9988, + "step": 11094 + }, + { + "epoch": 1.9754273504273505, + "grad_norm": 0.6959224343299866, + "learning_rate": 0.00010209946656232949, + "loss": 1.1097, + "step": 11095 + }, + { + "epoch": 1.9756054131054133, + "grad_norm": 0.7205058336257935, + "learning_rate": 0.00010208547215485376, + "loss": 0.9951, + "step": 11096 + }, + { + "epoch": 1.9757834757834758, + "grad_norm": 0.6968751549720764, + "learning_rate": 0.00010207147770651748, + "loss": 0.9313, + "step": 11097 + }, + { + "epoch": 1.9759615384615383, + "grad_norm": 0.6688823103904724, + "learning_rate": 0.00010205748321759494, + "loss": 0.9439, + "step": 11098 + }, + { + "epoch": 1.976139601139601, + "grad_norm": 0.6169568300247192, + "learning_rate": 0.00010204348868836028, + "loss": 1.123, + "step": 11099 + }, + { + "epoch": 1.9763176638176638, + "grad_norm": 0.6995537281036377, + "learning_rate": 0.00010202949411908768, + "loss": 1.1928, + "step": 11100 + }, + { + "epoch": 1.9764957264957266, + "grad_norm": 0.7102637887001038, + "learning_rate": 0.00010201549951005138, + "loss": 1.0265, + "step": 11101 + }, + { + "epoch": 1.9766737891737893, + "grad_norm": 0.6820045113563538, + "learning_rate": 0.00010200150486152558, + "loss": 0.9309, + "step": 11102 + }, + { + "epoch": 1.9768518518518519, + "grad_norm": 0.7050938010215759, + "learning_rate": 0.00010198751017378443, + "loss": 1.0047, + "step": 11103 + }, + { + "epoch": 1.9770299145299144, + "grad_norm": 0.6418201923370361, + "learning_rate": 0.00010197351544710214, + "loss": 1.1172, + "step": 11104 + }, + { + "epoch": 1.9772079772079771, + "grad_norm": 0.6681215763092041, + "learning_rate": 0.0001019595206817529, + "loss": 1.0621, + "step": 11105 + }, + { + "epoch": 1.97738603988604, + "grad_norm": 0.7725709676742554, + "learning_rate": 0.00010194552587801094, + "loss": 1.0044, + "step": 11106 + }, + { + "epoch": 1.9775641025641026, + "grad_norm": 0.6870455741882324, + "learning_rate": 0.00010193153103615045, + "loss": 1.2652, + "step": 11107 + }, + { + "epoch": 1.9777421652421654, + "grad_norm": 0.6352108120918274, + "learning_rate": 0.00010191753615644561, + "loss": 1.1081, + "step": 11108 + }, + { + "epoch": 1.977920227920228, + "grad_norm": 0.7322626113891602, + "learning_rate": 0.00010190354123917066, + "loss": 1.0003, + "step": 11109 + }, + { + "epoch": 1.9780982905982905, + "grad_norm": 0.6240935921669006, + "learning_rate": 0.00010188954628459972, + "loss": 0.8925, + "step": 11110 + }, + { + "epoch": 1.9782763532763532, + "grad_norm": 0.6648945212364197, + "learning_rate": 0.00010187555129300708, + "loss": 1.0882, + "step": 11111 + }, + { + "epoch": 1.978454415954416, + "grad_norm": 0.6704208850860596, + "learning_rate": 0.00010186155626466692, + "loss": 0.8873, + "step": 11112 + }, + { + "epoch": 1.9786324786324787, + "grad_norm": 0.6716459393501282, + "learning_rate": 0.00010184756119985341, + "loss": 1.0045, + "step": 11113 + }, + { + "epoch": 1.9788105413105415, + "grad_norm": 0.81277996301651, + "learning_rate": 0.0001018335660988408, + "loss": 0.8867, + "step": 11114 + }, + { + "epoch": 1.978988603988604, + "grad_norm": 0.7008311748504639, + "learning_rate": 0.00010181957096190323, + "loss": 0.9391, + "step": 11115 + }, + { + "epoch": 1.9791666666666665, + "grad_norm": 0.727676272392273, + "learning_rate": 0.00010180557578931498, + "loss": 1.0157, + "step": 11116 + }, + { + "epoch": 1.9793447293447293, + "grad_norm": 0.7058015465736389, + "learning_rate": 0.00010179158058135018, + "loss": 1.0, + "step": 11117 + }, + { + "epoch": 1.979522792022792, + "grad_norm": 0.7770412564277649, + "learning_rate": 0.00010177758533828312, + "loss": 1.0428, + "step": 11118 + }, + { + "epoch": 1.9797008547008548, + "grad_norm": 0.6557414531707764, + "learning_rate": 0.00010176359006038798, + "loss": 0.8557, + "step": 11119 + }, + { + "epoch": 1.9798789173789175, + "grad_norm": 0.7681090235710144, + "learning_rate": 0.00010174959474793894, + "loss": 0.867, + "step": 11120 + }, + { + "epoch": 1.98005698005698, + "grad_norm": 0.7915860414505005, + "learning_rate": 0.0001017355994012102, + "loss": 0.9961, + "step": 11121 + }, + { + "epoch": 1.9802350427350426, + "grad_norm": 0.8039166927337646, + "learning_rate": 0.00010172160402047604, + "loss": 1.1378, + "step": 11122 + }, + { + "epoch": 1.9804131054131053, + "grad_norm": 0.6641189455986023, + "learning_rate": 0.0001017076086060106, + "loss": 0.8914, + "step": 11123 + }, + { + "epoch": 1.980591168091168, + "grad_norm": 0.7673811316490173, + "learning_rate": 0.00010169361315808812, + "loss": 1.018, + "step": 11124 + }, + { + "epoch": 1.9807692307692308, + "grad_norm": 0.7320558428764343, + "learning_rate": 0.00010167961767698279, + "loss": 1.0515, + "step": 11125 + }, + { + "epoch": 1.9809472934472936, + "grad_norm": 0.5717357993125916, + "learning_rate": 0.00010166562216296886, + "loss": 0.7619, + "step": 11126 + }, + { + "epoch": 1.9811253561253561, + "grad_norm": 0.6638465523719788, + "learning_rate": 0.00010165162661632052, + "loss": 1.0161, + "step": 11127 + }, + { + "epoch": 1.9813034188034186, + "grad_norm": 0.7293243408203125, + "learning_rate": 0.00010163763103731201, + "loss": 1.063, + "step": 11128 + }, + { + "epoch": 1.9814814814814814, + "grad_norm": 0.634694516658783, + "learning_rate": 0.00010162363542621752, + "loss": 0.8945, + "step": 11129 + }, + { + "epoch": 1.9816595441595442, + "grad_norm": 0.7086902856826782, + "learning_rate": 0.00010160963978331122, + "loss": 1.0542, + "step": 11130 + }, + { + "epoch": 1.981837606837607, + "grad_norm": 0.5939825773239136, + "learning_rate": 0.00010159564410886742, + "loss": 0.7822, + "step": 11131 + }, + { + "epoch": 1.9820156695156697, + "grad_norm": 0.722183346748352, + "learning_rate": 0.00010158164840316027, + "loss": 1.0252, + "step": 11132 + }, + { + "epoch": 1.9821937321937322, + "grad_norm": 0.7300103306770325, + "learning_rate": 0.000101567652666464, + "loss": 0.9099, + "step": 11133 + }, + { + "epoch": 1.9823717948717947, + "grad_norm": 0.7148736119270325, + "learning_rate": 0.00010155365689905285, + "loss": 1.0149, + "step": 11134 + }, + { + "epoch": 1.9825498575498575, + "grad_norm": 0.8214462995529175, + "learning_rate": 0.000101539661101201, + "loss": 1.0127, + "step": 11135 + }, + { + "epoch": 1.9827279202279202, + "grad_norm": 0.7111126780509949, + "learning_rate": 0.00010152566527318265, + "loss": 1.045, + "step": 11136 + }, + { + "epoch": 1.982905982905983, + "grad_norm": 0.6640021800994873, + "learning_rate": 0.00010151166941527213, + "loss": 0.9618, + "step": 11137 + }, + { + "epoch": 1.9830840455840457, + "grad_norm": 0.7177722454071045, + "learning_rate": 0.00010149767352774358, + "loss": 1.0373, + "step": 11138 + }, + { + "epoch": 1.9832621082621082, + "grad_norm": 0.6728883981704712, + "learning_rate": 0.00010148367761087121, + "loss": 0.9886, + "step": 11139 + }, + { + "epoch": 1.9834401709401708, + "grad_norm": 0.7060428857803345, + "learning_rate": 0.00010146968166492926, + "loss": 1.042, + "step": 11140 + }, + { + "epoch": 1.9836182336182335, + "grad_norm": 0.706253707408905, + "learning_rate": 0.00010145568569019192, + "loss": 1.2249, + "step": 11141 + }, + { + "epoch": 1.9837962962962963, + "grad_norm": 0.618221640586853, + "learning_rate": 0.00010144168968693348, + "loss": 0.9223, + "step": 11142 + }, + { + "epoch": 1.983974358974359, + "grad_norm": 0.7005748748779297, + "learning_rate": 0.00010142769365542814, + "loss": 1.2735, + "step": 11143 + }, + { + "epoch": 1.9841524216524218, + "grad_norm": 0.6059799194335938, + "learning_rate": 0.0001014136975959501, + "loss": 0.7216, + "step": 11144 + }, + { + "epoch": 1.9843304843304843, + "grad_norm": 0.7169116735458374, + "learning_rate": 0.00010139970150877358, + "loss": 0.9541, + "step": 11145 + }, + { + "epoch": 1.984508547008547, + "grad_norm": 0.7402058839797974, + "learning_rate": 0.00010138570539417281, + "loss": 1.1268, + "step": 11146 + }, + { + "epoch": 1.9846866096866096, + "grad_norm": 0.7204117178916931, + "learning_rate": 0.00010137170925242201, + "loss": 1.1557, + "step": 11147 + }, + { + "epoch": 1.9848646723646723, + "grad_norm": 0.589163064956665, + "learning_rate": 0.00010135771308379545, + "loss": 0.9863, + "step": 11148 + }, + { + "epoch": 1.985042735042735, + "grad_norm": 0.6342785358428955, + "learning_rate": 0.00010134371688856732, + "loss": 0.9294, + "step": 11149 + }, + { + "epoch": 1.9852207977207978, + "grad_norm": 0.7144256234169006, + "learning_rate": 0.00010132972066701183, + "loss": 0.9428, + "step": 11150 + }, + { + "epoch": 1.9853988603988604, + "grad_norm": 0.658032238483429, + "learning_rate": 0.00010131572441940322, + "loss": 0.9749, + "step": 11151 + }, + { + "epoch": 1.9855769230769231, + "grad_norm": 0.7609163522720337, + "learning_rate": 0.00010130172814601576, + "loss": 1.1771, + "step": 11152 + }, + { + "epoch": 1.9857549857549857, + "grad_norm": 0.6531760692596436, + "learning_rate": 0.00010128773184712361, + "loss": 0.8529, + "step": 11153 + }, + { + "epoch": 1.9859330484330484, + "grad_norm": 0.6983599066734314, + "learning_rate": 0.00010127373552300103, + "loss": 1.0307, + "step": 11154 + }, + { + "epoch": 1.9861111111111112, + "grad_norm": 0.7121559381484985, + "learning_rate": 0.00010125973917392224, + "loss": 0.9426, + "step": 11155 + }, + { + "epoch": 1.986289173789174, + "grad_norm": 0.6282170414924622, + "learning_rate": 0.0001012457428001615, + "loss": 0.8983, + "step": 11156 + }, + { + "epoch": 1.9864672364672364, + "grad_norm": 0.6960387825965881, + "learning_rate": 0.000101231746401993, + "loss": 0.9001, + "step": 11157 + }, + { + "epoch": 1.9866452991452992, + "grad_norm": 0.7523152232170105, + "learning_rate": 0.000101217749979691, + "loss": 1.3462, + "step": 11158 + }, + { + "epoch": 1.9868233618233617, + "grad_norm": 0.71713787317276, + "learning_rate": 0.00010120375353352971, + "loss": 1.0147, + "step": 11159 + }, + { + "epoch": 1.9870014245014245, + "grad_norm": 0.7304390072822571, + "learning_rate": 0.00010118975706378339, + "loss": 0.8436, + "step": 11160 + }, + { + "epoch": 1.9871794871794872, + "grad_norm": 0.789968729019165, + "learning_rate": 0.00010117576057072622, + "loss": 1.1162, + "step": 11161 + }, + { + "epoch": 1.98735754985755, + "grad_norm": 0.6752170920372009, + "learning_rate": 0.00010116176405463249, + "loss": 1.0619, + "step": 11162 + }, + { + "epoch": 1.9875356125356125, + "grad_norm": 0.681398868560791, + "learning_rate": 0.0001011477675157764, + "loss": 0.8981, + "step": 11163 + }, + { + "epoch": 1.9877136752136753, + "grad_norm": 0.61469566822052, + "learning_rate": 0.0001011337709544322, + "loss": 1.0139, + "step": 11164 + }, + { + "epoch": 1.9878917378917378, + "grad_norm": 0.7524265050888062, + "learning_rate": 0.0001011197743708741, + "loss": 1.1571, + "step": 11165 + }, + { + "epoch": 1.9880698005698005, + "grad_norm": 0.6289594173431396, + "learning_rate": 0.00010110577776537633, + "loss": 0.93, + "step": 11166 + }, + { + "epoch": 1.9882478632478633, + "grad_norm": 0.6991903781890869, + "learning_rate": 0.00010109178113821318, + "loss": 1.1176, + "step": 11167 + }, + { + "epoch": 1.988425925925926, + "grad_norm": 0.7604053020477295, + "learning_rate": 0.00010107778448965883, + "loss": 1.0497, + "step": 11168 + }, + { + "epoch": 1.9886039886039886, + "grad_norm": 0.7166453003883362, + "learning_rate": 0.00010106378781998753, + "loss": 1.1237, + "step": 11169 + }, + { + "epoch": 1.9887820512820513, + "grad_norm": 0.6071686744689941, + "learning_rate": 0.00010104979112947352, + "loss": 0.8934, + "step": 11170 + }, + { + "epoch": 1.9889601139601139, + "grad_norm": 0.6618169546127319, + "learning_rate": 0.00010103579441839101, + "loss": 1.0596, + "step": 11171 + }, + { + "epoch": 1.9891381766381766, + "grad_norm": 0.6838458776473999, + "learning_rate": 0.0001010217976870143, + "loss": 1.0167, + "step": 11172 + }, + { + "epoch": 1.9893162393162394, + "grad_norm": 0.6369979381561279, + "learning_rate": 0.00010100780093561757, + "loss": 0.9001, + "step": 11173 + }, + { + "epoch": 1.989494301994302, + "grad_norm": 0.661313533782959, + "learning_rate": 0.00010099380416447508, + "loss": 0.8952, + "step": 11174 + }, + { + "epoch": 1.9896723646723646, + "grad_norm": 0.6991600394248962, + "learning_rate": 0.00010097980737386106, + "loss": 1.0083, + "step": 11175 + }, + { + "epoch": 1.9898504273504274, + "grad_norm": 0.618748664855957, + "learning_rate": 0.00010096581056404972, + "loss": 0.8797, + "step": 11176 + }, + { + "epoch": 1.99002849002849, + "grad_norm": 0.7039223909378052, + "learning_rate": 0.00010095181373531535, + "loss": 1.0385, + "step": 11177 + }, + { + "epoch": 1.9902065527065527, + "grad_norm": 0.7598999738693237, + "learning_rate": 0.00010093781688793216, + "loss": 0.9205, + "step": 11178 + }, + { + "epoch": 1.9903846153846154, + "grad_norm": 0.6355955600738525, + "learning_rate": 0.00010092382002217441, + "loss": 0.8646, + "step": 11179 + }, + { + "epoch": 1.9905626780626782, + "grad_norm": 0.8024569153785706, + "learning_rate": 0.00010090982313831634, + "loss": 1.1678, + "step": 11180 + }, + { + "epoch": 1.9907407407407407, + "grad_norm": 0.5960529446601868, + "learning_rate": 0.00010089582623663216, + "loss": 0.8277, + "step": 11181 + }, + { + "epoch": 1.9909188034188035, + "grad_norm": 0.6323728561401367, + "learning_rate": 0.00010088182931739609, + "loss": 0.948, + "step": 11182 + }, + { + "epoch": 1.991096866096866, + "grad_norm": 0.7532381415367126, + "learning_rate": 0.00010086783238088244, + "loss": 1.2948, + "step": 11183 + }, + { + "epoch": 1.9912749287749287, + "grad_norm": 0.5740166306495667, + "learning_rate": 0.00010085383542736543, + "loss": 0.7019, + "step": 11184 + }, + { + "epoch": 1.9914529914529915, + "grad_norm": 0.616985559463501, + "learning_rate": 0.00010083983845711929, + "loss": 1.0802, + "step": 11185 + }, + { + "epoch": 1.9916310541310542, + "grad_norm": 0.7505929470062256, + "learning_rate": 0.00010082584147041824, + "loss": 1.0523, + "step": 11186 + }, + { + "epoch": 1.9918091168091168, + "grad_norm": 0.7147656679153442, + "learning_rate": 0.00010081184446753653, + "loss": 1.0019, + "step": 11187 + }, + { + "epoch": 1.9919871794871795, + "grad_norm": 0.7301992774009705, + "learning_rate": 0.00010079784744874845, + "loss": 1.0329, + "step": 11188 + }, + { + "epoch": 1.992165242165242, + "grad_norm": 0.6847206354141235, + "learning_rate": 0.00010078385041432819, + "loss": 1.0367, + "step": 11189 + }, + { + "epoch": 1.9923433048433048, + "grad_norm": 0.7310990691184998, + "learning_rate": 0.00010076985336455, + "loss": 1.1675, + "step": 11190 + }, + { + "epoch": 1.9925213675213675, + "grad_norm": 0.6916858553886414, + "learning_rate": 0.00010075585629968813, + "loss": 0.8615, + "step": 11191 + }, + { + "epoch": 1.9926994301994303, + "grad_norm": 0.6519390344619751, + "learning_rate": 0.00010074185922001685, + "loss": 0.8105, + "step": 11192 + }, + { + "epoch": 1.9928774928774928, + "grad_norm": 0.7437400817871094, + "learning_rate": 0.00010072786212581036, + "loss": 0.9993, + "step": 11193 + }, + { + "epoch": 1.9930555555555556, + "grad_norm": 0.5048928260803223, + "learning_rate": 0.00010071386501734292, + "loss": 0.7912, + "step": 11194 + }, + { + "epoch": 1.993233618233618, + "grad_norm": 0.8042343258857727, + "learning_rate": 0.00010069986789488882, + "loss": 0.9156, + "step": 11195 + }, + { + "epoch": 1.9934116809116809, + "grad_norm": 0.7188669443130493, + "learning_rate": 0.0001006858707587222, + "loss": 1.0474, + "step": 11196 + }, + { + "epoch": 1.9935897435897436, + "grad_norm": 0.7377660870552063, + "learning_rate": 0.00010067187360911738, + "loss": 0.7013, + "step": 11197 + }, + { + "epoch": 1.9937678062678064, + "grad_norm": 0.6684696078300476, + "learning_rate": 0.00010065787644634861, + "loss": 0.9199, + "step": 11198 + }, + { + "epoch": 1.993945868945869, + "grad_norm": 0.7341524958610535, + "learning_rate": 0.00010064387927069012, + "loss": 1.0925, + "step": 11199 + }, + { + "epoch": 1.9941239316239316, + "grad_norm": 0.685745120048523, + "learning_rate": 0.00010062988208241614, + "loss": 1.083, + "step": 11200 + }, + { + "epoch": 1.9943019943019942, + "grad_norm": 0.6923556327819824, + "learning_rate": 0.00010061588488180096, + "loss": 1.2728, + "step": 11201 + }, + { + "epoch": 1.994480056980057, + "grad_norm": 0.6663293242454529, + "learning_rate": 0.00010060188766911876, + "loss": 1.0937, + "step": 11202 + }, + { + "epoch": 1.9946581196581197, + "grad_norm": 0.7963639497756958, + "learning_rate": 0.00010058789044464383, + "loss": 1.0592, + "step": 11203 + }, + { + "epoch": 1.9948361823361824, + "grad_norm": 0.6362990140914917, + "learning_rate": 0.00010057389320865042, + "loss": 0.8872, + "step": 11204 + }, + { + "epoch": 1.9950142450142452, + "grad_norm": 0.7752974033355713, + "learning_rate": 0.00010055989596141278, + "loss": 1.043, + "step": 11205 + }, + { + "epoch": 1.9951923076923077, + "grad_norm": 0.7125133275985718, + "learning_rate": 0.00010054589870320512, + "loss": 1.0015, + "step": 11206 + }, + { + "epoch": 1.9953703703703702, + "grad_norm": 0.7102736830711365, + "learning_rate": 0.00010053190143430169, + "loss": 1.0052, + "step": 11207 + }, + { + "epoch": 1.995548433048433, + "grad_norm": 0.8628628849983215, + "learning_rate": 0.00010051790415497677, + "loss": 1.2351, + "step": 11208 + }, + { + "epoch": 1.9957264957264957, + "grad_norm": 0.7233129739761353, + "learning_rate": 0.00010050390686550462, + "loss": 1.0848, + "step": 11209 + }, + { + "epoch": 1.9959045584045585, + "grad_norm": 0.5936228036880493, + "learning_rate": 0.00010048990956615944, + "loss": 0.7998, + "step": 11210 + }, + { + "epoch": 1.9960826210826212, + "grad_norm": 0.7345388531684875, + "learning_rate": 0.0001004759122572155, + "loss": 1.0329, + "step": 11211 + }, + { + "epoch": 1.9962606837606838, + "grad_norm": 0.7344130873680115, + "learning_rate": 0.00010046191493894703, + "loss": 1.1563, + "step": 11212 + }, + { + "epoch": 1.9964387464387463, + "grad_norm": 0.6979942321777344, + "learning_rate": 0.00010044791761162833, + "loss": 0.9269, + "step": 11213 + }, + { + "epoch": 1.996616809116809, + "grad_norm": 0.67514967918396, + "learning_rate": 0.0001004339202755336, + "loss": 0.9028, + "step": 11214 + }, + { + "epoch": 1.9967948717948718, + "grad_norm": 0.6379111409187317, + "learning_rate": 0.00010041992293093712, + "loss": 0.7816, + "step": 11215 + }, + { + "epoch": 1.9969729344729346, + "grad_norm": 0.693976104259491, + "learning_rate": 0.00010040592557811308, + "loss": 0.8411, + "step": 11216 + }, + { + "epoch": 1.9971509971509973, + "grad_norm": 0.5952646732330322, + "learning_rate": 0.0001003919282173358, + "loss": 0.8681, + "step": 11217 + }, + { + "epoch": 1.9973290598290598, + "grad_norm": 0.7452160716056824, + "learning_rate": 0.00010037793084887948, + "loss": 1.0198, + "step": 11218 + }, + { + "epoch": 1.9975071225071224, + "grad_norm": 0.6683938503265381, + "learning_rate": 0.00010036393347301841, + "loss": 0.8162, + "step": 11219 + }, + { + "epoch": 1.9976851851851851, + "grad_norm": 0.6849120855331421, + "learning_rate": 0.00010034993609002683, + "loss": 1.0668, + "step": 11220 + }, + { + "epoch": 1.9978632478632479, + "grad_norm": 0.8782517910003662, + "learning_rate": 0.00010033593870017897, + "loss": 1.222, + "step": 11221 + }, + { + "epoch": 1.9980413105413106, + "grad_norm": 0.6482772827148438, + "learning_rate": 0.00010032194130374908, + "loss": 0.7722, + "step": 11222 + }, + { + "epoch": 1.9982193732193734, + "grad_norm": 0.8595399260520935, + "learning_rate": 0.00010030794390101142, + "loss": 1.3004, + "step": 11223 + }, + { + "epoch": 1.998397435897436, + "grad_norm": 0.7258931994438171, + "learning_rate": 0.00010029394649224024, + "loss": 0.8825, + "step": 11224 + }, + { + "epoch": 1.9985754985754984, + "grad_norm": 0.6291348934173584, + "learning_rate": 0.00010027994907770981, + "loss": 0.8681, + "step": 11225 + }, + { + "epoch": 1.9987535612535612, + "grad_norm": 0.7528844475746155, + "learning_rate": 0.00010026595165769434, + "loss": 1.1443, + "step": 11226 + }, + { + "epoch": 1.998931623931624, + "grad_norm": 0.654017984867096, + "learning_rate": 0.0001002519542324681, + "loss": 0.8585, + "step": 11227 + }, + { + "epoch": 1.9991096866096867, + "grad_norm": 0.6812533736228943, + "learning_rate": 0.00010023795680230532, + "loss": 0.8757, + "step": 11228 + }, + { + "epoch": 1.9992877492877494, + "grad_norm": 0.7120179533958435, + "learning_rate": 0.0001002239593674803, + "loss": 1.0159, + "step": 11229 + }, + { + "epoch": 1.999465811965812, + "grad_norm": 0.6943802237510681, + "learning_rate": 0.00010020996192826725, + "loss": 1.0193, + "step": 11230 + }, + { + "epoch": 1.9996438746438745, + "grad_norm": 0.7227906584739685, + "learning_rate": 0.00010019596448494047, + "loss": 1.1536, + "step": 11231 + }, + { + "epoch": 1.9998219373219372, + "grad_norm": 0.6233312487602234, + "learning_rate": 0.00010018196703777411, + "loss": 0.9117, + "step": 11232 + }, + { + "epoch": 1.9998219373219372, + "eval_loss": 1.0963108539581299, + "eval_runtime": 24.4478, + "eval_samples_per_second": 42.58, + "eval_steps_per_second": 21.311, + "step": 11232 + }, + { + "epoch": 2.0, + "grad_norm": 0.67911696434021, + "learning_rate": 0.00010016796958704254, + "loss": 0.9516, + "step": 11233 + }, + { + "epoch": 2.0001780626780628, + "grad_norm": 0.7372198700904846, + "learning_rate": 0.00010015397213301992, + "loss": 1.3066, + "step": 11234 + }, + { + "epoch": 2.0001780626780628, + "grad_norm": 0.7573498487472534, + "learning_rate": 0.00010013997467598055, + "loss": 1.0204, + "step": 11235 + }, + { + "epoch": 2.0003561253561255, + "grad_norm": 0.6862889528274536, + "learning_rate": 0.00010012597721619863, + "loss": 0.9447, + "step": 11236 + }, + { + "epoch": 2.0005341880341883, + "grad_norm": 0.5977628231048584, + "learning_rate": 0.00010011197975394851, + "loss": 0.9849, + "step": 11237 + }, + { + "epoch": 2.0007122507122506, + "grad_norm": 0.620206892490387, + "learning_rate": 0.00010009798228950431, + "loss": 0.7498, + "step": 11238 + }, + { + "epoch": 2.0008903133903133, + "grad_norm": 0.5694536566734314, + "learning_rate": 0.0001000839848231404, + "loss": 0.7092, + "step": 11239 + }, + { + "epoch": 2.001068376068376, + "grad_norm": 0.5880212783813477, + "learning_rate": 0.00010006998735513098, + "loss": 0.9057, + "step": 11240 + }, + { + "epoch": 2.001246438746439, + "grad_norm": 0.6152323484420776, + "learning_rate": 0.00010005598988575029, + "loss": 0.8356, + "step": 11241 + }, + { + "epoch": 2.0014245014245016, + "grad_norm": 0.6827659010887146, + "learning_rate": 0.00010004199241527261, + "loss": 0.8302, + "step": 11242 + }, + { + "epoch": 2.0016025641025643, + "grad_norm": 0.5883491635322571, + "learning_rate": 0.00010002799494397215, + "loss": 0.8616, + "step": 11243 + }, + { + "epoch": 2.0017806267806266, + "grad_norm": 0.7649462819099426, + "learning_rate": 0.00010001399747212322, + "loss": 1.1643, + "step": 11244 + }, + { + "epoch": 2.0019586894586894, + "grad_norm": 0.6435316205024719, + "learning_rate": 0.0001, + "loss": 0.9239, + "step": 11245 + }, + { + "epoch": 2.002136752136752, + "grad_norm": 0.5662951469421387, + "learning_rate": 9.99860025278768e-05, + "loss": 0.9111, + "step": 11246 + }, + { + "epoch": 2.002314814814815, + "grad_norm": 0.6234064102172852, + "learning_rate": 9.997200505602787e-05, + "loss": 0.4852, + "step": 11247 + }, + { + "epoch": 2.0024928774928776, + "grad_norm": 0.6322146058082581, + "learning_rate": 9.995800758472741e-05, + "loss": 0.8599, + "step": 11248 + }, + { + "epoch": 2.0026709401709404, + "grad_norm": 0.6131469011306763, + "learning_rate": 9.994401011424972e-05, + "loss": 0.8504, + "step": 11249 + }, + { + "epoch": 2.0028490028490027, + "grad_norm": 0.6809168457984924, + "learning_rate": 9.993001264486903e-05, + "loss": 0.761, + "step": 11250 + }, + { + "epoch": 2.0030270655270654, + "grad_norm": 0.6721677184104919, + "learning_rate": 9.991601517685962e-05, + "loss": 0.9146, + "step": 11251 + }, + { + "epoch": 2.003205128205128, + "grad_norm": 0.6395483016967773, + "learning_rate": 9.990201771049569e-05, + "loss": 0.8583, + "step": 11252 + }, + { + "epoch": 2.003383190883191, + "grad_norm": 0.8524805903434753, + "learning_rate": 9.988802024605153e-05, + "loss": 0.859, + "step": 11253 + }, + { + "epoch": 2.0035612535612537, + "grad_norm": 0.6186681389808655, + "learning_rate": 9.987402278380136e-05, + "loss": 0.6695, + "step": 11254 + }, + { + "epoch": 2.0037393162393164, + "grad_norm": 0.593245267868042, + "learning_rate": 9.98600253240195e-05, + "loss": 0.7104, + "step": 11255 + }, + { + "epoch": 2.0039173789173788, + "grad_norm": 0.6806482672691345, + "learning_rate": 9.98460278669801e-05, + "loss": 0.6208, + "step": 11256 + }, + { + "epoch": 2.0040954415954415, + "grad_norm": 0.7329097390174866, + "learning_rate": 9.983203041295753e-05, + "loss": 0.8264, + "step": 11257 + }, + { + "epoch": 2.0042735042735043, + "grad_norm": 0.7579078078269958, + "learning_rate": 9.981803296222591e-05, + "loss": 0.7572, + "step": 11258 + }, + { + "epoch": 2.004451566951567, + "grad_norm": 0.7945193648338318, + "learning_rate": 9.980403551505958e-05, + "loss": 0.7916, + "step": 11259 + }, + { + "epoch": 2.0046296296296298, + "grad_norm": 0.5528121590614319, + "learning_rate": 9.979003807173276e-05, + "loss": 0.5609, + "step": 11260 + }, + { + "epoch": 2.0048076923076925, + "grad_norm": 0.7321668863296509, + "learning_rate": 9.977604063251973e-05, + "loss": 0.9041, + "step": 11261 + }, + { + "epoch": 2.004985754985755, + "grad_norm": 0.6553691029548645, + "learning_rate": 9.976204319769469e-05, + "loss": 0.853, + "step": 11262 + }, + { + "epoch": 2.0051638176638176, + "grad_norm": 0.789152979850769, + "learning_rate": 9.974804576753194e-05, + "loss": 0.9909, + "step": 11263 + }, + { + "epoch": 2.0053418803418803, + "grad_norm": 0.6342231631278992, + "learning_rate": 9.973404834230568e-05, + "loss": 0.8841, + "step": 11264 + }, + { + "epoch": 2.005519943019943, + "grad_norm": 0.671882688999176, + "learning_rate": 9.97200509222902e-05, + "loss": 0.8147, + "step": 11265 + }, + { + "epoch": 2.005698005698006, + "grad_norm": 0.8409315347671509, + "learning_rate": 9.970605350775978e-05, + "loss": 1.0466, + "step": 11266 + }, + { + "epoch": 2.0058760683760686, + "grad_norm": 0.6155081987380981, + "learning_rate": 9.969205609898858e-05, + "loss": 0.9461, + "step": 11267 + }, + { + "epoch": 2.006054131054131, + "grad_norm": 0.656370997428894, + "learning_rate": 9.967805869625093e-05, + "loss": 0.6152, + "step": 11268 + }, + { + "epoch": 2.0062321937321936, + "grad_norm": 0.6441524624824524, + "learning_rate": 9.966406129982103e-05, + "loss": 0.9023, + "step": 11269 + }, + { + "epoch": 2.0064102564102564, + "grad_norm": 0.7976031303405762, + "learning_rate": 9.96500639099732e-05, + "loss": 0.8886, + "step": 11270 + }, + { + "epoch": 2.006588319088319, + "grad_norm": 0.6888235807418823, + "learning_rate": 9.963606652698159e-05, + "loss": 0.7216, + "step": 11271 + }, + { + "epoch": 2.006766381766382, + "grad_norm": 0.8439735770225525, + "learning_rate": 9.962206915112054e-05, + "loss": 1.0161, + "step": 11272 + }, + { + "epoch": 2.0069444444444446, + "grad_norm": 0.6425265669822693, + "learning_rate": 9.960807178266423e-05, + "loss": 0.8718, + "step": 11273 + }, + { + "epoch": 2.007122507122507, + "grad_norm": 0.7393937110900879, + "learning_rate": 9.959407442188696e-05, + "loss": 0.7615, + "step": 11274 + }, + { + "epoch": 2.0073005698005697, + "grad_norm": 0.5919229984283447, + "learning_rate": 9.958007706906292e-05, + "loss": 0.6761, + "step": 11275 + }, + { + "epoch": 2.0074786324786325, + "grad_norm": 0.765935480594635, + "learning_rate": 9.956607972446644e-05, + "loss": 0.9057, + "step": 11276 + }, + { + "epoch": 2.007656695156695, + "grad_norm": 0.7091122269630432, + "learning_rate": 9.955208238837169e-05, + "loss": 0.8322, + "step": 11277 + }, + { + "epoch": 2.007834757834758, + "grad_norm": 0.850652813911438, + "learning_rate": 9.953808506105299e-05, + "loss": 0.9942, + "step": 11278 + }, + { + "epoch": 2.0080128205128207, + "grad_norm": 0.7341200113296509, + "learning_rate": 9.952408774278452e-05, + "loss": 0.7826, + "step": 11279 + }, + { + "epoch": 2.008190883190883, + "grad_norm": 0.6891999840736389, + "learning_rate": 9.95100904338406e-05, + "loss": 0.8939, + "step": 11280 + }, + { + "epoch": 2.0083689458689458, + "grad_norm": 0.800881028175354, + "learning_rate": 9.94960931344954e-05, + "loss": 0.8036, + "step": 11281 + }, + { + "epoch": 2.0085470085470085, + "grad_norm": 0.7483115792274475, + "learning_rate": 9.948209584502328e-05, + "loss": 0.7203, + "step": 11282 + }, + { + "epoch": 2.0087250712250713, + "grad_norm": 0.7314630150794983, + "learning_rate": 9.946809856569833e-05, + "loss": 0.8907, + "step": 11283 + }, + { + "epoch": 2.008903133903134, + "grad_norm": 0.7317429184913635, + "learning_rate": 9.945410129679493e-05, + "loss": 0.8971, + "step": 11284 + }, + { + "epoch": 2.0090811965811968, + "grad_norm": 0.6968898177146912, + "learning_rate": 9.944010403858726e-05, + "loss": 0.8638, + "step": 11285 + }, + { + "epoch": 2.009259259259259, + "grad_norm": 0.6680058240890503, + "learning_rate": 9.942610679134957e-05, + "loss": 0.7524, + "step": 11286 + }, + { + "epoch": 2.009437321937322, + "grad_norm": 0.6863839030265808, + "learning_rate": 9.941210955535618e-05, + "loss": 0.9647, + "step": 11287 + }, + { + "epoch": 2.0096153846153846, + "grad_norm": 0.7137607336044312, + "learning_rate": 9.939811233088125e-05, + "loss": 0.7089, + "step": 11288 + }, + { + "epoch": 2.0097934472934473, + "grad_norm": 0.8341759443283081, + "learning_rate": 9.938411511819907e-05, + "loss": 0.9461, + "step": 11289 + }, + { + "epoch": 2.00997150997151, + "grad_norm": 0.7326228022575378, + "learning_rate": 9.937011791758384e-05, + "loss": 0.8795, + "step": 11290 + }, + { + "epoch": 2.010149572649573, + "grad_norm": 0.6795905232429504, + "learning_rate": 9.935612072930989e-05, + "loss": 0.8298, + "step": 11291 + }, + { + "epoch": 2.010327635327635, + "grad_norm": 0.7060360312461853, + "learning_rate": 9.934212355365139e-05, + "loss": 0.8483, + "step": 11292 + }, + { + "epoch": 2.010505698005698, + "grad_norm": 0.7532246112823486, + "learning_rate": 9.932812639088265e-05, + "loss": 0.9061, + "step": 11293 + }, + { + "epoch": 2.0106837606837606, + "grad_norm": 0.6563972234725952, + "learning_rate": 9.931412924127781e-05, + "loss": 0.8511, + "step": 11294 + }, + { + "epoch": 2.0108618233618234, + "grad_norm": 0.6672948002815247, + "learning_rate": 9.930013210511125e-05, + "loss": 0.7875, + "step": 11295 + }, + { + "epoch": 2.011039886039886, + "grad_norm": 0.7173593640327454, + "learning_rate": 9.928613498265709e-05, + "loss": 0.8602, + "step": 11296 + }, + { + "epoch": 2.011217948717949, + "grad_norm": 0.7399459481239319, + "learning_rate": 9.927213787418968e-05, + "loss": 0.8711, + "step": 11297 + }, + { + "epoch": 2.011396011396011, + "grad_norm": 0.7693262696266174, + "learning_rate": 9.925814077998317e-05, + "loss": 0.9927, + "step": 11298 + }, + { + "epoch": 2.011574074074074, + "grad_norm": 0.7998616695404053, + "learning_rate": 9.92441437003119e-05, + "loss": 1.1585, + "step": 11299 + }, + { + "epoch": 2.0117521367521367, + "grad_norm": 0.7239874005317688, + "learning_rate": 9.923014663545002e-05, + "loss": 0.8736, + "step": 11300 + }, + { + "epoch": 2.0119301994301995, + "grad_norm": 0.8565806150436401, + "learning_rate": 9.921614958567186e-05, + "loss": 0.9768, + "step": 11301 + }, + { + "epoch": 2.012108262108262, + "grad_norm": 0.6341429948806763, + "learning_rate": 9.920215255125158e-05, + "loss": 0.6553, + "step": 11302 + }, + { + "epoch": 2.012286324786325, + "grad_norm": 0.824182391166687, + "learning_rate": 9.91881555324635e-05, + "loss": 1.0138, + "step": 11303 + }, + { + "epoch": 2.0124643874643873, + "grad_norm": 0.6309344172477722, + "learning_rate": 9.917415852958178e-05, + "loss": 0.619, + "step": 11304 + }, + { + "epoch": 2.01264245014245, + "grad_norm": 0.7469239830970764, + "learning_rate": 9.916016154288071e-05, + "loss": 0.8537, + "step": 11305 + }, + { + "epoch": 2.0128205128205128, + "grad_norm": 0.7433663606643677, + "learning_rate": 9.914616457263459e-05, + "loss": 0.8518, + "step": 11306 + }, + { + "epoch": 2.0129985754985755, + "grad_norm": 0.6550318002700806, + "learning_rate": 9.913216761911755e-05, + "loss": 0.8021, + "step": 11307 + }, + { + "epoch": 2.0131766381766383, + "grad_norm": 0.7360837459564209, + "learning_rate": 9.911817068260392e-05, + "loss": 0.7002, + "step": 11308 + }, + { + "epoch": 2.013354700854701, + "grad_norm": 0.7208407521247864, + "learning_rate": 9.910417376336786e-05, + "loss": 0.8633, + "step": 11309 + }, + { + "epoch": 2.0135327635327633, + "grad_norm": 0.7758026719093323, + "learning_rate": 9.909017686168369e-05, + "loss": 0.764, + "step": 11310 + }, + { + "epoch": 2.013710826210826, + "grad_norm": 0.8215547204017639, + "learning_rate": 9.90761799778256e-05, + "loss": 0.7062, + "step": 11311 + }, + { + "epoch": 2.013888888888889, + "grad_norm": 0.6731052994728088, + "learning_rate": 9.906218311206786e-05, + "loss": 0.902, + "step": 11312 + }, + { + "epoch": 2.0140669515669516, + "grad_norm": 0.74113929271698, + "learning_rate": 9.904818626468466e-05, + "loss": 0.7229, + "step": 11313 + }, + { + "epoch": 2.0142450142450143, + "grad_norm": 0.6673575639724731, + "learning_rate": 9.90341894359503e-05, + "loss": 0.7299, + "step": 11314 + }, + { + "epoch": 2.014423076923077, + "grad_norm": 0.7665545344352722, + "learning_rate": 9.902019262613897e-05, + "loss": 0.6993, + "step": 11315 + }, + { + "epoch": 2.0146011396011394, + "grad_norm": 0.6423895359039307, + "learning_rate": 9.900619583552497e-05, + "loss": 0.7344, + "step": 11316 + }, + { + "epoch": 2.014779202279202, + "grad_norm": 0.7071038484573364, + "learning_rate": 9.899219906438245e-05, + "loss": 0.6951, + "step": 11317 + }, + { + "epoch": 2.014957264957265, + "grad_norm": 0.689984142780304, + "learning_rate": 9.897820231298574e-05, + "loss": 0.8496, + "step": 11318 + }, + { + "epoch": 2.0151353276353277, + "grad_norm": 0.8747256398200989, + "learning_rate": 9.896420558160901e-05, + "loss": 0.9752, + "step": 11319 + }, + { + "epoch": 2.0153133903133904, + "grad_norm": 0.6828433275222778, + "learning_rate": 9.895020887052651e-05, + "loss": 0.8369, + "step": 11320 + }, + { + "epoch": 2.015491452991453, + "grad_norm": 0.7334261536598206, + "learning_rate": 9.89362121800125e-05, + "loss": 0.7744, + "step": 11321 + }, + { + "epoch": 2.0156695156695155, + "grad_norm": 0.7896139621734619, + "learning_rate": 9.892221551034122e-05, + "loss": 0.8353, + "step": 11322 + }, + { + "epoch": 2.015847578347578, + "grad_norm": 0.6673476099967957, + "learning_rate": 9.890821886178684e-05, + "loss": 0.8644, + "step": 11323 + }, + { + "epoch": 2.016025641025641, + "grad_norm": 0.7475691437721252, + "learning_rate": 9.889422223462368e-05, + "loss": 0.8034, + "step": 11324 + }, + { + "epoch": 2.0162037037037037, + "grad_norm": 0.9086315631866455, + "learning_rate": 9.888022562912593e-05, + "loss": 1.1878, + "step": 11325 + }, + { + "epoch": 2.0163817663817665, + "grad_norm": 0.6634678244590759, + "learning_rate": 9.88662290455678e-05, + "loss": 0.9655, + "step": 11326 + }, + { + "epoch": 2.0165598290598292, + "grad_norm": 0.7184932827949524, + "learning_rate": 9.885223248422361e-05, + "loss": 0.5964, + "step": 11327 + }, + { + "epoch": 2.0167378917378915, + "grad_norm": 0.6319148540496826, + "learning_rate": 9.883823594536751e-05, + "loss": 0.5692, + "step": 11328 + }, + { + "epoch": 2.0169159544159543, + "grad_norm": 0.6232550144195557, + "learning_rate": 9.88242394292738e-05, + "loss": 0.6492, + "step": 11329 + }, + { + "epoch": 2.017094017094017, + "grad_norm": 0.7149667143821716, + "learning_rate": 9.881024293621663e-05, + "loss": 0.7023, + "step": 11330 + }, + { + "epoch": 2.01727207977208, + "grad_norm": 0.8871679902076721, + "learning_rate": 9.879624646647031e-05, + "loss": 0.954, + "step": 11331 + }, + { + "epoch": 2.0174501424501425, + "grad_norm": 0.6905941367149353, + "learning_rate": 9.878225002030901e-05, + "loss": 0.8534, + "step": 11332 + }, + { + "epoch": 2.0176282051282053, + "grad_norm": 0.8891478776931763, + "learning_rate": 9.876825359800703e-05, + "loss": 0.8324, + "step": 11333 + }, + { + "epoch": 2.0178062678062676, + "grad_norm": 0.8125092387199402, + "learning_rate": 9.875425719983852e-05, + "loss": 0.9604, + "step": 11334 + }, + { + "epoch": 2.0179843304843303, + "grad_norm": 0.7362027764320374, + "learning_rate": 9.874026082607778e-05, + "loss": 0.7879, + "step": 11335 + }, + { + "epoch": 2.018162393162393, + "grad_norm": 0.6763492226600647, + "learning_rate": 9.872626447699899e-05, + "loss": 0.8839, + "step": 11336 + }, + { + "epoch": 2.018340455840456, + "grad_norm": 0.7350467443466187, + "learning_rate": 9.871226815287644e-05, + "loss": 0.834, + "step": 11337 + }, + { + "epoch": 2.0185185185185186, + "grad_norm": 0.7768327593803406, + "learning_rate": 9.869827185398428e-05, + "loss": 1.1123, + "step": 11338 + }, + { + "epoch": 2.0186965811965814, + "grad_norm": 0.9218043088912964, + "learning_rate": 9.868427558059681e-05, + "loss": 0.9439, + "step": 11339 + }, + { + "epoch": 2.0188746438746437, + "grad_norm": 0.6613419651985168, + "learning_rate": 9.867027933298819e-05, + "loss": 0.836, + "step": 11340 + }, + { + "epoch": 2.0190527065527064, + "grad_norm": 0.7251055240631104, + "learning_rate": 9.865628311143273e-05, + "loss": 0.973, + "step": 11341 + }, + { + "epoch": 2.019230769230769, + "grad_norm": 0.6571859121322632, + "learning_rate": 9.864228691620458e-05, + "loss": 0.8811, + "step": 11342 + }, + { + "epoch": 2.019408831908832, + "grad_norm": 0.7552264928817749, + "learning_rate": 9.862829074757802e-05, + "loss": 0.9128, + "step": 11343 + }, + { + "epoch": 2.0195868945868947, + "grad_norm": 0.6724083423614502, + "learning_rate": 9.861429460582723e-05, + "loss": 0.8894, + "step": 11344 + }, + { + "epoch": 2.0197649572649574, + "grad_norm": 0.8309593200683594, + "learning_rate": 9.860029849122644e-05, + "loss": 0.9374, + "step": 11345 + }, + { + "epoch": 2.0199430199430197, + "grad_norm": 0.7709865570068359, + "learning_rate": 9.858630240404993e-05, + "loss": 0.8195, + "step": 11346 + }, + { + "epoch": 2.0201210826210825, + "grad_norm": 0.8163080811500549, + "learning_rate": 9.857230634457187e-05, + "loss": 0.9329, + "step": 11347 + }, + { + "epoch": 2.0202991452991452, + "grad_norm": 0.8424021005630493, + "learning_rate": 9.855831031306653e-05, + "loss": 0.8732, + "step": 11348 + }, + { + "epoch": 2.020477207977208, + "grad_norm": 0.7816365361213684, + "learning_rate": 9.854431430980808e-05, + "loss": 0.8858, + "step": 11349 + }, + { + "epoch": 2.0206552706552707, + "grad_norm": 0.7559000253677368, + "learning_rate": 9.853031833507075e-05, + "loss": 0.7146, + "step": 11350 + }, + { + "epoch": 2.0208333333333335, + "grad_norm": 0.6723140478134155, + "learning_rate": 9.85163223891288e-05, + "loss": 0.813, + "step": 11351 + }, + { + "epoch": 2.021011396011396, + "grad_norm": 0.757641077041626, + "learning_rate": 9.850232647225646e-05, + "loss": 0.794, + "step": 11352 + }, + { + "epoch": 2.0211894586894585, + "grad_norm": 0.8217115998268127, + "learning_rate": 9.848833058472787e-05, + "loss": 1.0407, + "step": 11353 + }, + { + "epoch": 2.0213675213675213, + "grad_norm": 0.8016467690467834, + "learning_rate": 9.847433472681736e-05, + "loss": 0.8967, + "step": 11354 + }, + { + "epoch": 2.021545584045584, + "grad_norm": 0.7703533172607422, + "learning_rate": 9.846033889879903e-05, + "loss": 0.9669, + "step": 11355 + }, + { + "epoch": 2.021723646723647, + "grad_norm": 0.7372044920921326, + "learning_rate": 9.84463431009472e-05, + "loss": 0.8581, + "step": 11356 + }, + { + "epoch": 2.0219017094017095, + "grad_norm": 0.7676188945770264, + "learning_rate": 9.8432347333536e-05, + "loss": 0.8498, + "step": 11357 + }, + { + "epoch": 2.0220797720797723, + "grad_norm": 0.7485190629959106, + "learning_rate": 9.841835159683977e-05, + "loss": 0.8492, + "step": 11358 + }, + { + "epoch": 2.0222578347578346, + "grad_norm": 0.7287883758544922, + "learning_rate": 9.840435589113262e-05, + "loss": 0.9072, + "step": 11359 + }, + { + "epoch": 2.0224358974358974, + "grad_norm": 0.7719354033470154, + "learning_rate": 9.83903602166888e-05, + "loss": 0.7657, + "step": 11360 + }, + { + "epoch": 2.02261396011396, + "grad_norm": 0.7679458260536194, + "learning_rate": 9.837636457378251e-05, + "loss": 0.7098, + "step": 11361 + }, + { + "epoch": 2.022792022792023, + "grad_norm": 0.7496665120124817, + "learning_rate": 9.836236896268803e-05, + "loss": 0.8459, + "step": 11362 + }, + { + "epoch": 2.0229700854700856, + "grad_norm": 0.8511863350868225, + "learning_rate": 9.834837338367949e-05, + "loss": 0.9782, + "step": 11363 + }, + { + "epoch": 2.0231481481481484, + "grad_norm": 0.5752342343330383, + "learning_rate": 9.833437783703114e-05, + "loss": 0.4539, + "step": 11364 + }, + { + "epoch": 2.0233262108262107, + "grad_norm": 0.6654593348503113, + "learning_rate": 9.832038232301722e-05, + "loss": 0.8009, + "step": 11365 + }, + { + "epoch": 2.0235042735042734, + "grad_norm": 0.7296777963638306, + "learning_rate": 9.83063868419119e-05, + "loss": 0.7841, + "step": 11366 + }, + { + "epoch": 2.023682336182336, + "grad_norm": 0.8404465913772583, + "learning_rate": 9.829239139398943e-05, + "loss": 0.9152, + "step": 11367 + }, + { + "epoch": 2.023860398860399, + "grad_norm": 0.6407002806663513, + "learning_rate": 9.827839597952397e-05, + "loss": 0.6953, + "step": 11368 + }, + { + "epoch": 2.0240384615384617, + "grad_norm": 0.8107042908668518, + "learning_rate": 9.826440059878982e-05, + "loss": 0.8726, + "step": 11369 + }, + { + "epoch": 2.0242165242165244, + "grad_norm": 0.803804874420166, + "learning_rate": 9.825040525206108e-05, + "loss": 0.8906, + "step": 11370 + }, + { + "epoch": 2.0243945868945867, + "grad_norm": 0.7625358700752258, + "learning_rate": 9.823640993961205e-05, + "loss": 0.8938, + "step": 11371 + }, + { + "epoch": 2.0245726495726495, + "grad_norm": 0.690793514251709, + "learning_rate": 9.822241466171686e-05, + "loss": 0.7926, + "step": 11372 + }, + { + "epoch": 2.0247507122507122, + "grad_norm": 0.7006554007530212, + "learning_rate": 9.820841941864983e-05, + "loss": 0.793, + "step": 11373 + }, + { + "epoch": 2.024928774928775, + "grad_norm": 0.8029078841209412, + "learning_rate": 9.819442421068504e-05, + "loss": 0.867, + "step": 11374 + }, + { + "epoch": 2.0251068376068377, + "grad_norm": 0.6999112367630005, + "learning_rate": 9.818042903809678e-05, + "loss": 0.688, + "step": 11375 + }, + { + "epoch": 2.0252849002849005, + "grad_norm": 0.6848462224006653, + "learning_rate": 9.816643390115923e-05, + "loss": 0.7337, + "step": 11376 + }, + { + "epoch": 2.025462962962963, + "grad_norm": 0.7698155641555786, + "learning_rate": 9.815243880014663e-05, + "loss": 0.9712, + "step": 11377 + }, + { + "epoch": 2.0256410256410255, + "grad_norm": 0.8449836373329163, + "learning_rate": 9.81384437353331e-05, + "loss": 0.9144, + "step": 11378 + }, + { + "epoch": 2.0258190883190883, + "grad_norm": 0.6340110301971436, + "learning_rate": 9.812444870699296e-05, + "loss": 0.6365, + "step": 11379 + }, + { + "epoch": 2.025997150997151, + "grad_norm": 0.7104073762893677, + "learning_rate": 9.81104537154003e-05, + "loss": 0.7781, + "step": 11380 + }, + { + "epoch": 2.026175213675214, + "grad_norm": 0.7287606000900269, + "learning_rate": 9.809645876082939e-05, + "loss": 0.9351, + "step": 11381 + }, + { + "epoch": 2.0263532763532766, + "grad_norm": 0.9640787243843079, + "learning_rate": 9.80824638435544e-05, + "loss": 0.8745, + "step": 11382 + }, + { + "epoch": 2.026531339031339, + "grad_norm": 0.5718010067939758, + "learning_rate": 9.806846896384959e-05, + "loss": 0.4711, + "step": 11383 + }, + { + "epoch": 2.0267094017094016, + "grad_norm": 0.7903527021408081, + "learning_rate": 9.805447412198907e-05, + "loss": 0.8241, + "step": 11384 + }, + { + "epoch": 2.0268874643874644, + "grad_norm": 0.8579357862472534, + "learning_rate": 9.80404793182471e-05, + "loss": 0.8621, + "step": 11385 + }, + { + "epoch": 2.027065527065527, + "grad_norm": 0.8466464877128601, + "learning_rate": 9.802648455289787e-05, + "loss": 0.8772, + "step": 11386 + }, + { + "epoch": 2.02724358974359, + "grad_norm": 0.7888286709785461, + "learning_rate": 9.801248982621557e-05, + "loss": 0.8352, + "step": 11387 + }, + { + "epoch": 2.0274216524216526, + "grad_norm": 0.6967005133628845, + "learning_rate": 9.799849513847444e-05, + "loss": 0.7936, + "step": 11388 + }, + { + "epoch": 2.027599715099715, + "grad_norm": 0.6987027525901794, + "learning_rate": 9.79845004899486e-05, + "loss": 0.7778, + "step": 11389 + }, + { + "epoch": 2.0277777777777777, + "grad_norm": 0.7414312362670898, + "learning_rate": 9.797050588091233e-05, + "loss": 0.9017, + "step": 11390 + }, + { + "epoch": 2.0279558404558404, + "grad_norm": 0.7932028770446777, + "learning_rate": 9.795651131163974e-05, + "loss": 0.8662, + "step": 11391 + }, + { + "epoch": 2.028133903133903, + "grad_norm": 0.8166332244873047, + "learning_rate": 9.79425167824051e-05, + "loss": 1.0489, + "step": 11392 + }, + { + "epoch": 2.028311965811966, + "grad_norm": 0.7265253663063049, + "learning_rate": 9.792852229348251e-05, + "loss": 0.9458, + "step": 11393 + }, + { + "epoch": 2.0284900284900287, + "grad_norm": 0.7374703288078308, + "learning_rate": 9.791452784514629e-05, + "loss": 0.9203, + "step": 11394 + }, + { + "epoch": 2.028668091168091, + "grad_norm": 0.6912441253662109, + "learning_rate": 9.790053343767052e-05, + "loss": 0.8986, + "step": 11395 + }, + { + "epoch": 2.0288461538461537, + "grad_norm": 0.871231734752655, + "learning_rate": 9.788653907132946e-05, + "loss": 0.6811, + "step": 11396 + }, + { + "epoch": 2.0290242165242165, + "grad_norm": 0.7361812591552734, + "learning_rate": 9.787254474639726e-05, + "loss": 0.7868, + "step": 11397 + }, + { + "epoch": 2.0292022792022792, + "grad_norm": 0.6828895211219788, + "learning_rate": 9.785855046314815e-05, + "loss": 0.7739, + "step": 11398 + }, + { + "epoch": 2.029380341880342, + "grad_norm": 0.7203328609466553, + "learning_rate": 9.784455622185626e-05, + "loss": 0.6474, + "step": 11399 + }, + { + "epoch": 2.0295584045584047, + "grad_norm": 0.774886429309845, + "learning_rate": 9.783056202279587e-05, + "loss": 0.8073, + "step": 11400 + }, + { + "epoch": 2.029736467236467, + "grad_norm": 0.6479005813598633, + "learning_rate": 9.781656786624106e-05, + "loss": 0.7237, + "step": 11401 + }, + { + "epoch": 2.02991452991453, + "grad_norm": 0.7269866466522217, + "learning_rate": 9.78025737524661e-05, + "loss": 0.9089, + "step": 11402 + }, + { + "epoch": 2.0300925925925926, + "grad_norm": 0.7265415191650391, + "learning_rate": 9.778857968174509e-05, + "loss": 0.827, + "step": 11403 + }, + { + "epoch": 2.0302706552706553, + "grad_norm": 0.8174277544021606, + "learning_rate": 9.777458565435227e-05, + "loss": 0.6752, + "step": 11404 + }, + { + "epoch": 2.030448717948718, + "grad_norm": 0.9333333969116211, + "learning_rate": 9.77605916705619e-05, + "loss": 0.9542, + "step": 11405 + }, + { + "epoch": 2.030626780626781, + "grad_norm": 0.6854027509689331, + "learning_rate": 9.774659773064801e-05, + "loss": 0.8526, + "step": 11406 + }, + { + "epoch": 2.030804843304843, + "grad_norm": 0.7711043357849121, + "learning_rate": 9.773260383488489e-05, + "loss": 0.9009, + "step": 11407 + }, + { + "epoch": 2.030982905982906, + "grad_norm": 0.6915287971496582, + "learning_rate": 9.771860998354667e-05, + "loss": 0.9635, + "step": 11408 + }, + { + "epoch": 2.0311609686609686, + "grad_norm": 0.7978841066360474, + "learning_rate": 9.770461617690758e-05, + "loss": 0.7563, + "step": 11409 + }, + { + "epoch": 2.0313390313390314, + "grad_norm": 0.6686414480209351, + "learning_rate": 9.769062241524172e-05, + "loss": 0.8282, + "step": 11410 + }, + { + "epoch": 2.031517094017094, + "grad_norm": 0.7024029493331909, + "learning_rate": 9.767662869882335e-05, + "loss": 0.9176, + "step": 11411 + }, + { + "epoch": 2.031695156695157, + "grad_norm": 0.6945844292640686, + "learning_rate": 9.766263502792659e-05, + "loss": 0.86, + "step": 11412 + }, + { + "epoch": 2.031873219373219, + "grad_norm": 0.7351676821708679, + "learning_rate": 9.764864140282569e-05, + "loss": 0.865, + "step": 11413 + }, + { + "epoch": 2.032051282051282, + "grad_norm": 0.7663825750350952, + "learning_rate": 9.763464782379472e-05, + "loss": 0.9309, + "step": 11414 + }, + { + "epoch": 2.0322293447293447, + "grad_norm": 0.7552894949913025, + "learning_rate": 9.762065429110798e-05, + "loss": 0.8366, + "step": 11415 + }, + { + "epoch": 2.0324074074074074, + "grad_norm": 0.6852208971977234, + "learning_rate": 9.760666080503951e-05, + "loss": 0.9095, + "step": 11416 + }, + { + "epoch": 2.03258547008547, + "grad_norm": 0.7759820222854614, + "learning_rate": 9.759266736586358e-05, + "loss": 0.7461, + "step": 11417 + }, + { + "epoch": 2.032763532763533, + "grad_norm": 0.6514183878898621, + "learning_rate": 9.757867397385431e-05, + "loss": 0.5479, + "step": 11418 + }, + { + "epoch": 2.0329415954415953, + "grad_norm": 0.7703103423118591, + "learning_rate": 9.756468062928593e-05, + "loss": 0.9588, + "step": 11419 + }, + { + "epoch": 2.033119658119658, + "grad_norm": 0.6937198638916016, + "learning_rate": 9.755068733243255e-05, + "loss": 0.8661, + "step": 11420 + }, + { + "epoch": 2.0332977207977208, + "grad_norm": 0.6675645112991333, + "learning_rate": 9.753669408356835e-05, + "loss": 0.7484, + "step": 11421 + }, + { + "epoch": 2.0334757834757835, + "grad_norm": 0.6653266549110413, + "learning_rate": 9.752270088296753e-05, + "loss": 0.7217, + "step": 11422 + }, + { + "epoch": 2.0336538461538463, + "grad_norm": 0.7893908023834229, + "learning_rate": 9.750870773090425e-05, + "loss": 0.6346, + "step": 11423 + }, + { + "epoch": 2.033831908831909, + "grad_norm": 0.7442745566368103, + "learning_rate": 9.749471462765265e-05, + "loss": 0.9452, + "step": 11424 + }, + { + "epoch": 2.0340099715099713, + "grad_norm": 0.8270035982131958, + "learning_rate": 9.748072157348691e-05, + "loss": 0.9906, + "step": 11425 + }, + { + "epoch": 2.034188034188034, + "grad_norm": 0.7195143699645996, + "learning_rate": 9.746672856868123e-05, + "loss": 0.9564, + "step": 11426 + }, + { + "epoch": 2.034366096866097, + "grad_norm": 0.75486820936203, + "learning_rate": 9.745273561350971e-05, + "loss": 0.9052, + "step": 11427 + }, + { + "epoch": 2.0345441595441596, + "grad_norm": 0.6710293889045715, + "learning_rate": 9.743874270824655e-05, + "loss": 0.6801, + "step": 11428 + }, + { + "epoch": 2.0347222222222223, + "grad_norm": 0.704175591468811, + "learning_rate": 9.742474985316588e-05, + "loss": 0.8619, + "step": 11429 + }, + { + "epoch": 2.034900284900285, + "grad_norm": 0.7941717505455017, + "learning_rate": 9.741075704854196e-05, + "loss": 0.8318, + "step": 11430 + }, + { + "epoch": 2.0350783475783474, + "grad_norm": 0.8592050671577454, + "learning_rate": 9.739676429464881e-05, + "loss": 0.8203, + "step": 11431 + }, + { + "epoch": 2.03525641025641, + "grad_norm": 0.9149407148361206, + "learning_rate": 9.738277159176068e-05, + "loss": 0.87, + "step": 11432 + }, + { + "epoch": 2.035434472934473, + "grad_norm": 0.780890941619873, + "learning_rate": 9.736877894015169e-05, + "loss": 0.6971, + "step": 11433 + }, + { + "epoch": 2.0356125356125356, + "grad_norm": 0.7540209293365479, + "learning_rate": 9.735478634009605e-05, + "loss": 0.8927, + "step": 11434 + }, + { + "epoch": 2.0357905982905984, + "grad_norm": 0.8556281924247742, + "learning_rate": 9.734079379186782e-05, + "loss": 0.7498, + "step": 11435 + }, + { + "epoch": 2.035968660968661, + "grad_norm": 0.8710931539535522, + "learning_rate": 9.732680129574128e-05, + "loss": 0.6009, + "step": 11436 + }, + { + "epoch": 2.0361467236467234, + "grad_norm": 0.6873082518577576, + "learning_rate": 9.731280885199045e-05, + "loss": 0.8441, + "step": 11437 + }, + { + "epoch": 2.036324786324786, + "grad_norm": 0.8333037495613098, + "learning_rate": 9.729881646088958e-05, + "loss": 0.888, + "step": 11438 + }, + { + "epoch": 2.036502849002849, + "grad_norm": 0.859365701675415, + "learning_rate": 9.728482412271277e-05, + "loss": 1.0272, + "step": 11439 + }, + { + "epoch": 2.0366809116809117, + "grad_norm": 0.7239334583282471, + "learning_rate": 9.727083183773423e-05, + "loss": 0.9428, + "step": 11440 + }, + { + "epoch": 2.0368589743589745, + "grad_norm": 0.8341524004936218, + "learning_rate": 9.725683960622804e-05, + "loss": 0.9275, + "step": 11441 + }, + { + "epoch": 2.037037037037037, + "grad_norm": 0.6992602348327637, + "learning_rate": 9.724284742846838e-05, + "loss": 0.8492, + "step": 11442 + }, + { + "epoch": 2.0372150997150995, + "grad_norm": 0.7429133057594299, + "learning_rate": 9.72288553047294e-05, + "loss": 0.8246, + "step": 11443 + }, + { + "epoch": 2.0373931623931623, + "grad_norm": 0.7765250205993652, + "learning_rate": 9.721486323528522e-05, + "loss": 0.8624, + "step": 11444 + }, + { + "epoch": 2.037571225071225, + "grad_norm": 0.9104889631271362, + "learning_rate": 9.720087122041007e-05, + "loss": 0.8369, + "step": 11445 + }, + { + "epoch": 2.0377492877492878, + "grad_norm": 0.6483191251754761, + "learning_rate": 9.718687926037798e-05, + "loss": 0.7347, + "step": 11446 + }, + { + "epoch": 2.0379273504273505, + "grad_norm": 0.7816178202629089, + "learning_rate": 9.717288735546317e-05, + "loss": 0.8607, + "step": 11447 + }, + { + "epoch": 2.0381054131054133, + "grad_norm": 0.6909009218215942, + "learning_rate": 9.715889550593975e-05, + "loss": 0.8764, + "step": 11448 + }, + { + "epoch": 2.0382834757834756, + "grad_norm": 0.8101255297660828, + "learning_rate": 9.71449037120819e-05, + "loss": 0.8858, + "step": 11449 + }, + { + "epoch": 2.0384615384615383, + "grad_norm": 0.7476511001586914, + "learning_rate": 9.71309119741637e-05, + "loss": 0.8765, + "step": 11450 + }, + { + "epoch": 2.038639601139601, + "grad_norm": 0.7514875531196594, + "learning_rate": 9.711692029245934e-05, + "loss": 0.942, + "step": 11451 + }, + { + "epoch": 2.038817663817664, + "grad_norm": 0.7400087118148804, + "learning_rate": 9.710292866724292e-05, + "loss": 0.8327, + "step": 11452 + }, + { + "epoch": 2.0389957264957266, + "grad_norm": 0.832979142665863, + "learning_rate": 9.70889370987886e-05, + "loss": 0.9714, + "step": 11453 + }, + { + "epoch": 2.0391737891737893, + "grad_norm": 0.6918326616287231, + "learning_rate": 9.70749455873705e-05, + "loss": 0.7765, + "step": 11454 + }, + { + "epoch": 2.0393518518518516, + "grad_norm": 0.8286036849021912, + "learning_rate": 9.70609541332628e-05, + "loss": 0.9138, + "step": 11455 + }, + { + "epoch": 2.0395299145299144, + "grad_norm": 0.6436729431152344, + "learning_rate": 9.704696273673955e-05, + "loss": 0.738, + "step": 11456 + }, + { + "epoch": 2.039707977207977, + "grad_norm": 0.7057681679725647, + "learning_rate": 9.703297139807496e-05, + "loss": 0.8107, + "step": 11457 + }, + { + "epoch": 2.03988603988604, + "grad_norm": 0.7444550395011902, + "learning_rate": 9.701898011754313e-05, + "loss": 0.8188, + "step": 11458 + }, + { + "epoch": 2.0400641025641026, + "grad_norm": 0.7622130513191223, + "learning_rate": 9.70049888954182e-05, + "loss": 0.8451, + "step": 11459 + }, + { + "epoch": 2.0402421652421654, + "grad_norm": 0.8166092038154602, + "learning_rate": 9.699099773197426e-05, + "loss": 0.9399, + "step": 11460 + }, + { + "epoch": 2.0404202279202277, + "grad_norm": 0.7235924601554871, + "learning_rate": 9.697700662748552e-05, + "loss": 0.7863, + "step": 11461 + }, + { + "epoch": 2.0405982905982905, + "grad_norm": 0.7150312662124634, + "learning_rate": 9.696301558222601e-05, + "loss": 0.8288, + "step": 11462 + }, + { + "epoch": 2.040776353276353, + "grad_norm": 0.8007016777992249, + "learning_rate": 9.694902459646993e-05, + "loss": 0.9203, + "step": 11463 + }, + { + "epoch": 2.040954415954416, + "grad_norm": 0.7665491700172424, + "learning_rate": 9.693503367049134e-05, + "loss": 0.7956, + "step": 11464 + }, + { + "epoch": 2.0411324786324787, + "grad_norm": 0.7499460577964783, + "learning_rate": 9.692104280456439e-05, + "loss": 0.9973, + "step": 11465 + }, + { + "epoch": 2.0413105413105415, + "grad_norm": 0.7598159909248352, + "learning_rate": 9.690705199896327e-05, + "loss": 0.95, + "step": 11466 + }, + { + "epoch": 2.041488603988604, + "grad_norm": 0.7699945569038391, + "learning_rate": 9.689306125396201e-05, + "loss": 0.8731, + "step": 11467 + }, + { + "epoch": 2.0416666666666665, + "grad_norm": 0.6724731922149658, + "learning_rate": 9.687907056983476e-05, + "loss": 0.906, + "step": 11468 + }, + { + "epoch": 2.0418447293447293, + "grad_norm": 0.9238275289535522, + "learning_rate": 9.686507994685562e-05, + "loss": 0.8397, + "step": 11469 + }, + { + "epoch": 2.042022792022792, + "grad_norm": 0.744969367980957, + "learning_rate": 9.685108938529876e-05, + "loss": 0.8436, + "step": 11470 + }, + { + "epoch": 2.0422008547008548, + "grad_norm": 0.6983298063278198, + "learning_rate": 9.683709888543824e-05, + "loss": 0.8235, + "step": 11471 + }, + { + "epoch": 2.0423789173789175, + "grad_norm": 0.7098708748817444, + "learning_rate": 9.682310844754824e-05, + "loss": 0.8235, + "step": 11472 + }, + { + "epoch": 2.04255698005698, + "grad_norm": 0.7492793798446655, + "learning_rate": 9.680911807190277e-05, + "loss": 0.7988, + "step": 11473 + }, + { + "epoch": 2.0427350427350426, + "grad_norm": 0.6952250003814697, + "learning_rate": 9.679512775877604e-05, + "loss": 0.7928, + "step": 11474 + }, + { + "epoch": 2.0429131054131053, + "grad_norm": 0.6442983150482178, + "learning_rate": 9.678113750844209e-05, + "loss": 0.8206, + "step": 11475 + }, + { + "epoch": 2.043091168091168, + "grad_norm": 0.7408245205879211, + "learning_rate": 9.67671473211751e-05, + "loss": 0.6941, + "step": 11476 + }, + { + "epoch": 2.043269230769231, + "grad_norm": 0.8277738094329834, + "learning_rate": 9.675315719724913e-05, + "loss": 1.3153, + "step": 11477 + }, + { + "epoch": 2.0434472934472936, + "grad_norm": 0.7535714507102966, + "learning_rate": 9.67391671369383e-05, + "loss": 0.9238, + "step": 11478 + }, + { + "epoch": 2.0436253561253563, + "grad_norm": 0.8341996073722839, + "learning_rate": 9.67251771405167e-05, + "loss": 0.8149, + "step": 11479 + }, + { + "epoch": 2.0438034188034186, + "grad_norm": 0.7365956902503967, + "learning_rate": 9.671118720825849e-05, + "loss": 0.799, + "step": 11480 + }, + { + "epoch": 2.0439814814814814, + "grad_norm": 0.7630738615989685, + "learning_rate": 9.669719734043769e-05, + "loss": 0.9284, + "step": 11481 + }, + { + "epoch": 2.044159544159544, + "grad_norm": 0.659172773361206, + "learning_rate": 9.668320753732848e-05, + "loss": 0.7594, + "step": 11482 + }, + { + "epoch": 2.044337606837607, + "grad_norm": 0.7724705934524536, + "learning_rate": 9.66692177992049e-05, + "loss": 0.8623, + "step": 11483 + }, + { + "epoch": 2.0445156695156697, + "grad_norm": 0.7140040993690491, + "learning_rate": 9.665522812634108e-05, + "loss": 0.851, + "step": 11484 + }, + { + "epoch": 2.0446937321937324, + "grad_norm": 0.9072890877723694, + "learning_rate": 9.664123851901115e-05, + "loss": 0.9459, + "step": 11485 + }, + { + "epoch": 2.0448717948717947, + "grad_norm": 0.8145443201065063, + "learning_rate": 9.662724897748915e-05, + "loss": 0.9067, + "step": 11486 + }, + { + "epoch": 2.0450498575498575, + "grad_norm": 0.8471246957778931, + "learning_rate": 9.661325950204922e-05, + "loss": 0.7194, + "step": 11487 + }, + { + "epoch": 2.04522792022792, + "grad_norm": 0.8465375304222107, + "learning_rate": 9.659927009296541e-05, + "loss": 0.9495, + "step": 11488 + }, + { + "epoch": 2.045405982905983, + "grad_norm": 0.7597832083702087, + "learning_rate": 9.658528075051185e-05, + "loss": 0.7526, + "step": 11489 + }, + { + "epoch": 2.0455840455840457, + "grad_norm": 0.8013564944267273, + "learning_rate": 9.657129147496261e-05, + "loss": 1.0514, + "step": 11490 + }, + { + "epoch": 2.0457621082621085, + "grad_norm": 0.8695764541625977, + "learning_rate": 9.655730226659182e-05, + "loss": 0.9925, + "step": 11491 + }, + { + "epoch": 2.0459401709401708, + "grad_norm": 0.7295607328414917, + "learning_rate": 9.65433131256735e-05, + "loss": 0.8652, + "step": 11492 + }, + { + "epoch": 2.0461182336182335, + "grad_norm": 0.7819971442222595, + "learning_rate": 9.652932405248181e-05, + "loss": 0.8601, + "step": 11493 + }, + { + "epoch": 2.0462962962962963, + "grad_norm": 0.7244205474853516, + "learning_rate": 9.651533504729078e-05, + "loss": 0.752, + "step": 11494 + }, + { + "epoch": 2.046474358974359, + "grad_norm": 0.7774363160133362, + "learning_rate": 9.650134611037456e-05, + "loss": 0.8638, + "step": 11495 + }, + { + "epoch": 2.046652421652422, + "grad_norm": 0.7955372929573059, + "learning_rate": 9.648735724200715e-05, + "loss": 0.9662, + "step": 11496 + }, + { + "epoch": 2.0468304843304845, + "grad_norm": 0.7114127278327942, + "learning_rate": 9.647336844246273e-05, + "loss": 0.9523, + "step": 11497 + }, + { + "epoch": 2.047008547008547, + "grad_norm": 0.7449100017547607, + "learning_rate": 9.645937971201527e-05, + "loss": 0.7898, + "step": 11498 + }, + { + "epoch": 2.0471866096866096, + "grad_norm": 0.7541512846946716, + "learning_rate": 9.644539105093895e-05, + "loss": 0.9286, + "step": 11499 + }, + { + "epoch": 2.0473646723646723, + "grad_norm": 0.6816682815551758, + "learning_rate": 9.643140245950778e-05, + "loss": 0.7757, + "step": 11500 + }, + { + "epoch": 2.047542735042735, + "grad_norm": 0.7222850918769836, + "learning_rate": 9.641741393799591e-05, + "loss": 0.8415, + "step": 11501 + }, + { + "epoch": 2.047720797720798, + "grad_norm": 0.7605552077293396, + "learning_rate": 9.640342548667732e-05, + "loss": 0.8875, + "step": 11502 + }, + { + "epoch": 2.0478988603988606, + "grad_norm": 0.7442240118980408, + "learning_rate": 9.638943710582615e-05, + "loss": 0.8755, + "step": 11503 + }, + { + "epoch": 2.048076923076923, + "grad_norm": 0.7065736651420593, + "learning_rate": 9.637544879571648e-05, + "loss": 0.6885, + "step": 11504 + }, + { + "epoch": 2.0482549857549857, + "grad_norm": 0.6400303244590759, + "learning_rate": 9.636146055662232e-05, + "loss": 0.5775, + "step": 11505 + }, + { + "epoch": 2.0484330484330484, + "grad_norm": 0.7955389022827148, + "learning_rate": 9.634747238881783e-05, + "loss": 1.0182, + "step": 11506 + }, + { + "epoch": 2.048611111111111, + "grad_norm": 0.8283255696296692, + "learning_rate": 9.6333484292577e-05, + "loss": 0.9247, + "step": 11507 + }, + { + "epoch": 2.048789173789174, + "grad_norm": 0.7619521617889404, + "learning_rate": 9.631949626817399e-05, + "loss": 0.8355, + "step": 11508 + }, + { + "epoch": 2.0489672364672367, + "grad_norm": 0.7204191088676453, + "learning_rate": 9.630550831588273e-05, + "loss": 0.8571, + "step": 11509 + }, + { + "epoch": 2.049145299145299, + "grad_norm": 0.7568399310112, + "learning_rate": 9.629152043597738e-05, + "loss": 0.7349, + "step": 11510 + }, + { + "epoch": 2.0493233618233617, + "grad_norm": 0.8594959378242493, + "learning_rate": 9.627753262873199e-05, + "loss": 0.6918, + "step": 11511 + }, + { + "epoch": 2.0495014245014245, + "grad_norm": 0.6345391869544983, + "learning_rate": 9.626354489442064e-05, + "loss": 0.871, + "step": 11512 + }, + { + "epoch": 2.0496794871794872, + "grad_norm": 0.7671827673912048, + "learning_rate": 9.624955723331732e-05, + "loss": 0.8518, + "step": 11513 + }, + { + "epoch": 2.04985754985755, + "grad_norm": 0.7182049751281738, + "learning_rate": 9.623556964569616e-05, + "loss": 0.889, + "step": 11514 + }, + { + "epoch": 2.0500356125356127, + "grad_norm": 0.8342016339302063, + "learning_rate": 9.622158213183118e-05, + "loss": 0.9712, + "step": 11515 + }, + { + "epoch": 2.050213675213675, + "grad_norm": 0.7007761001586914, + "learning_rate": 9.620759469199649e-05, + "loss": 0.8113, + "step": 11516 + }, + { + "epoch": 2.050391737891738, + "grad_norm": 0.7129531502723694, + "learning_rate": 9.619360732646605e-05, + "loss": 0.8666, + "step": 11517 + }, + { + "epoch": 2.0505698005698005, + "grad_norm": 0.7505812048912048, + "learning_rate": 9.6179620035514e-05, + "loss": 0.7321, + "step": 11518 + }, + { + "epoch": 2.0507478632478633, + "grad_norm": 0.7407607436180115, + "learning_rate": 9.616563281941433e-05, + "loss": 0.9275, + "step": 11519 + }, + { + "epoch": 2.050925925925926, + "grad_norm": 0.769345223903656, + "learning_rate": 9.615164567844116e-05, + "loss": 0.9731, + "step": 11520 + }, + { + "epoch": 2.051103988603989, + "grad_norm": 0.7782812118530273, + "learning_rate": 9.613765861286846e-05, + "loss": 0.9702, + "step": 11521 + }, + { + "epoch": 2.051282051282051, + "grad_norm": 0.7071413993835449, + "learning_rate": 9.612367162297037e-05, + "loss": 0.8451, + "step": 11522 + }, + { + "epoch": 2.051460113960114, + "grad_norm": 0.7598503232002258, + "learning_rate": 9.610968470902082e-05, + "loss": 0.8641, + "step": 11523 + }, + { + "epoch": 2.0516381766381766, + "grad_norm": 0.7951003313064575, + "learning_rate": 9.609569787129394e-05, + "loss": 0.9131, + "step": 11524 + }, + { + "epoch": 2.0518162393162394, + "grad_norm": 0.8029175996780396, + "learning_rate": 9.608171111006374e-05, + "loss": 0.8618, + "step": 11525 + }, + { + "epoch": 2.051994301994302, + "grad_norm": 0.6993120908737183, + "learning_rate": 9.606772442560428e-05, + "loss": 0.8487, + "step": 11526 + }, + { + "epoch": 2.052172364672365, + "grad_norm": 0.8039231896400452, + "learning_rate": 9.605373781818961e-05, + "loss": 1.0102, + "step": 11527 + }, + { + "epoch": 2.052350427350427, + "grad_norm": 0.714849054813385, + "learning_rate": 9.603975128809373e-05, + "loss": 0.8977, + "step": 11528 + }, + { + "epoch": 2.05252849002849, + "grad_norm": 0.8728037476539612, + "learning_rate": 9.60257648355907e-05, + "loss": 0.8004, + "step": 11529 + }, + { + "epoch": 2.0527065527065527, + "grad_norm": 0.764776885509491, + "learning_rate": 9.601177846095454e-05, + "loss": 0.9205, + "step": 11530 + }, + { + "epoch": 2.0528846153846154, + "grad_norm": 0.6948725581169128, + "learning_rate": 9.599779216445934e-05, + "loss": 0.7864, + "step": 11531 + }, + { + "epoch": 2.053062678062678, + "grad_norm": 0.7663996815681458, + "learning_rate": 9.598380594637903e-05, + "loss": 0.8877, + "step": 11532 + }, + { + "epoch": 2.053240740740741, + "grad_norm": 0.7584146857261658, + "learning_rate": 9.596981980698776e-05, + "loss": 1.1328, + "step": 11533 + }, + { + "epoch": 2.0534188034188032, + "grad_norm": 0.7701094150543213, + "learning_rate": 9.595583374655945e-05, + "loss": 0.7551, + "step": 11534 + }, + { + "epoch": 2.053596866096866, + "grad_norm": 0.7745714783668518, + "learning_rate": 9.594184776536821e-05, + "loss": 0.8862, + "step": 11535 + }, + { + "epoch": 2.0537749287749287, + "grad_norm": 0.7832430005073547, + "learning_rate": 9.5927861863688e-05, + "loss": 0.8736, + "step": 11536 + }, + { + "epoch": 2.0539529914529915, + "grad_norm": 0.7354840040206909, + "learning_rate": 9.591387604179291e-05, + "loss": 0.8183, + "step": 11537 + }, + { + "epoch": 2.0541310541310542, + "grad_norm": 0.7516480684280396, + "learning_rate": 9.589989029995691e-05, + "loss": 0.924, + "step": 11538 + }, + { + "epoch": 2.054309116809117, + "grad_norm": 0.7942310571670532, + "learning_rate": 9.588590463845405e-05, + "loss": 1.0283, + "step": 11539 + }, + { + "epoch": 2.0544871794871793, + "grad_norm": 0.7716572880744934, + "learning_rate": 9.587191905755832e-05, + "loss": 0.8686, + "step": 11540 + }, + { + "epoch": 2.054665242165242, + "grad_norm": 0.8075140118598938, + "learning_rate": 9.585793355754381e-05, + "loss": 0.8731, + "step": 11541 + }, + { + "epoch": 2.054843304843305, + "grad_norm": 0.8119283318519592, + "learning_rate": 9.584394813868444e-05, + "loss": 0.9543, + "step": 11542 + }, + { + "epoch": 2.0550213675213675, + "grad_norm": 0.6476314067840576, + "learning_rate": 9.582996280125427e-05, + "loss": 0.6943, + "step": 11543 + }, + { + "epoch": 2.0551994301994303, + "grad_norm": 0.7617185711860657, + "learning_rate": 9.581597754552737e-05, + "loss": 0.6942, + "step": 11544 + }, + { + "epoch": 2.055377492877493, + "grad_norm": 0.879355788230896, + "learning_rate": 9.580199237177765e-05, + "loss": 0.825, + "step": 11545 + }, + { + "epoch": 2.0555555555555554, + "grad_norm": 0.8229055404663086, + "learning_rate": 9.578800728027919e-05, + "loss": 0.9973, + "step": 11546 + }, + { + "epoch": 2.055733618233618, + "grad_norm": 0.7808930277824402, + "learning_rate": 9.577402227130596e-05, + "loss": 0.9525, + "step": 11547 + }, + { + "epoch": 2.055911680911681, + "grad_norm": 0.870499849319458, + "learning_rate": 9.576003734513201e-05, + "loss": 0.8874, + "step": 11548 + }, + { + "epoch": 2.0560897435897436, + "grad_norm": 0.8254318833351135, + "learning_rate": 9.57460525020313e-05, + "loss": 0.6656, + "step": 11549 + }, + { + "epoch": 2.0562678062678064, + "grad_norm": 0.8358132243156433, + "learning_rate": 9.573206774227786e-05, + "loss": 0.7946, + "step": 11550 + }, + { + "epoch": 2.056445868945869, + "grad_norm": 0.636366605758667, + "learning_rate": 9.571808306614568e-05, + "loss": 0.6757, + "step": 11551 + }, + { + "epoch": 2.0566239316239314, + "grad_norm": 0.8884546160697937, + "learning_rate": 9.57040984739088e-05, + "loss": 0.6775, + "step": 11552 + }, + { + "epoch": 2.056801994301994, + "grad_norm": 0.7240797877311707, + "learning_rate": 9.569011396584115e-05, + "loss": 0.8033, + "step": 11553 + }, + { + "epoch": 2.056980056980057, + "grad_norm": 0.8730767965316772, + "learning_rate": 9.567612954221678e-05, + "loss": 0.9577, + "step": 11554 + }, + { + "epoch": 2.0571581196581197, + "grad_norm": 0.6785064339637756, + "learning_rate": 9.566214520330966e-05, + "loss": 0.6241, + "step": 11555 + }, + { + "epoch": 2.0573361823361824, + "grad_norm": 0.7757805585861206, + "learning_rate": 9.564816094939382e-05, + "loss": 0.7926, + "step": 11556 + }, + { + "epoch": 2.057514245014245, + "grad_norm": 0.7630164623260498, + "learning_rate": 9.563417678074319e-05, + "loss": 0.8547, + "step": 11557 + }, + { + "epoch": 2.0576923076923075, + "grad_norm": 0.7690725922584534, + "learning_rate": 9.562019269763184e-05, + "loss": 0.9172, + "step": 11558 + }, + { + "epoch": 2.0578703703703702, + "grad_norm": 0.81644207239151, + "learning_rate": 9.560620870033367e-05, + "loss": 0.811, + "step": 11559 + }, + { + "epoch": 2.058048433048433, + "grad_norm": 0.8240723013877869, + "learning_rate": 9.559222478912273e-05, + "loss": 0.9094, + "step": 11560 + }, + { + "epoch": 2.0582264957264957, + "grad_norm": 0.7168204188346863, + "learning_rate": 9.557824096427297e-05, + "loss": 1.0617, + "step": 11561 + }, + { + "epoch": 2.0584045584045585, + "grad_norm": 0.6648391485214233, + "learning_rate": 9.556425722605846e-05, + "loss": 0.6556, + "step": 11562 + }, + { + "epoch": 2.0585826210826212, + "grad_norm": 0.7291145324707031, + "learning_rate": 9.555027357475305e-05, + "loss": 0.784, + "step": 11563 + }, + { + "epoch": 2.0587606837606836, + "grad_norm": 0.6910824775695801, + "learning_rate": 9.553629001063079e-05, + "loss": 0.9332, + "step": 11564 + }, + { + "epoch": 2.0589387464387463, + "grad_norm": 0.757247805595398, + "learning_rate": 9.552230653396566e-05, + "loss": 0.6598, + "step": 11565 + }, + { + "epoch": 2.059116809116809, + "grad_norm": 0.7778435349464417, + "learning_rate": 9.550832314503163e-05, + "loss": 0.8899, + "step": 11566 + }, + { + "epoch": 2.059294871794872, + "grad_norm": 0.7827669978141785, + "learning_rate": 9.54943398441027e-05, + "loss": 0.8036, + "step": 11567 + }, + { + "epoch": 2.0594729344729346, + "grad_norm": 0.7462462186813354, + "learning_rate": 9.54803566314528e-05, + "loss": 0.9306, + "step": 11568 + }, + { + "epoch": 2.0596509971509973, + "grad_norm": 0.8088639974594116, + "learning_rate": 9.546637350735597e-05, + "loss": 0.8766, + "step": 11569 + }, + { + "epoch": 2.0598290598290596, + "grad_norm": 0.6477743983268738, + "learning_rate": 9.545239047208607e-05, + "loss": 0.7239, + "step": 11570 + }, + { + "epoch": 2.0600071225071224, + "grad_norm": 0.7535004615783691, + "learning_rate": 9.543840752591718e-05, + "loss": 0.8891, + "step": 11571 + }, + { + "epoch": 2.060185185185185, + "grad_norm": 0.7085242867469788, + "learning_rate": 9.542442466912316e-05, + "loss": 0.8105, + "step": 11572 + }, + { + "epoch": 2.060363247863248, + "grad_norm": 0.8129137754440308, + "learning_rate": 9.541044190197811e-05, + "loss": 0.6955, + "step": 11573 + }, + { + "epoch": 2.0605413105413106, + "grad_norm": 0.7160677909851074, + "learning_rate": 9.539645922475586e-05, + "loss": 0.833, + "step": 11574 + }, + { + "epoch": 2.0607193732193734, + "grad_norm": 0.6983035206794739, + "learning_rate": 9.538247663773044e-05, + "loss": 0.6439, + "step": 11575 + }, + { + "epoch": 2.0608974358974357, + "grad_norm": 0.8732622861862183, + "learning_rate": 9.536849414117578e-05, + "loss": 0.7763, + "step": 11576 + }, + { + "epoch": 2.0610754985754984, + "grad_norm": 0.7745480537414551, + "learning_rate": 9.535451173536591e-05, + "loss": 0.8272, + "step": 11577 + }, + { + "epoch": 2.061253561253561, + "grad_norm": 0.8210037350654602, + "learning_rate": 9.53405294205747e-05, + "loss": 0.9539, + "step": 11578 + }, + { + "epoch": 2.061431623931624, + "grad_norm": 0.6742323637008667, + "learning_rate": 9.532654719707617e-05, + "loss": 0.6525, + "step": 11579 + }, + { + "epoch": 2.0616096866096867, + "grad_norm": 0.8312603831291199, + "learning_rate": 9.531256506514418e-05, + "loss": 0.7776, + "step": 11580 + }, + { + "epoch": 2.0617877492877494, + "grad_norm": 0.7817347049713135, + "learning_rate": 9.529858302505278e-05, + "loss": 1.0148, + "step": 11581 + }, + { + "epoch": 2.0619658119658117, + "grad_norm": 0.751153290271759, + "learning_rate": 9.528460107707584e-05, + "loss": 0.8064, + "step": 11582 + }, + { + "epoch": 2.0621438746438745, + "grad_norm": 0.7483627200126648, + "learning_rate": 9.527061922148737e-05, + "loss": 0.8706, + "step": 11583 + }, + { + "epoch": 2.0623219373219372, + "grad_norm": 0.7044979929924011, + "learning_rate": 9.525663745856132e-05, + "loss": 0.8008, + "step": 11584 + }, + { + "epoch": 2.0625, + "grad_norm": 0.8249054551124573, + "learning_rate": 9.524265578857157e-05, + "loss": 0.9339, + "step": 11585 + }, + { + "epoch": 2.0626780626780628, + "grad_norm": 0.7184668183326721, + "learning_rate": 9.522867421179211e-05, + "loss": 0.9191, + "step": 11586 + }, + { + "epoch": 2.0628561253561255, + "grad_norm": 0.8135001063346863, + "learning_rate": 9.521469272849685e-05, + "loss": 0.966, + "step": 11587 + }, + { + "epoch": 2.0630341880341883, + "grad_norm": 0.8151242733001709, + "learning_rate": 9.520071133895978e-05, + "loss": 0.7947, + "step": 11588 + }, + { + "epoch": 2.0632122507122506, + "grad_norm": 0.8044771552085876, + "learning_rate": 9.518673004345477e-05, + "loss": 0.893, + "step": 11589 + }, + { + "epoch": 2.0633903133903133, + "grad_norm": 0.730505645275116, + "learning_rate": 9.517274884225581e-05, + "loss": 0.8691, + "step": 11590 + }, + { + "epoch": 2.063568376068376, + "grad_norm": 0.7419933676719666, + "learning_rate": 9.515876773563678e-05, + "loss": 0.809, + "step": 11591 + }, + { + "epoch": 2.063746438746439, + "grad_norm": 0.7809683084487915, + "learning_rate": 9.514478672387169e-05, + "loss": 0.6926, + "step": 11592 + }, + { + "epoch": 2.0639245014245016, + "grad_norm": 1.0065315961837769, + "learning_rate": 9.513080580723435e-05, + "loss": 0.7506, + "step": 11593 + }, + { + "epoch": 2.064102564102564, + "grad_norm": 0.7424543499946594, + "learning_rate": 9.511682498599883e-05, + "loss": 0.9158, + "step": 11594 + }, + { + "epoch": 2.0642806267806266, + "grad_norm": 0.6907097697257996, + "learning_rate": 9.510284426043893e-05, + "loss": 0.8795, + "step": 11595 + }, + { + "epoch": 2.0644586894586894, + "grad_norm": 0.7849169969558716, + "learning_rate": 9.508886363082864e-05, + "loss": 1.0065, + "step": 11596 + }, + { + "epoch": 2.064636752136752, + "grad_norm": 0.7421438694000244, + "learning_rate": 9.507488309744183e-05, + "loss": 0.7574, + "step": 11597 + }, + { + "epoch": 2.064814814814815, + "grad_norm": 0.6636283993721008, + "learning_rate": 9.506090266055252e-05, + "loss": 0.7489, + "step": 11598 + }, + { + "epoch": 2.0649928774928776, + "grad_norm": 0.7133244872093201, + "learning_rate": 9.504692232043452e-05, + "loss": 0.7652, + "step": 11599 + }, + { + "epoch": 2.0651709401709404, + "grad_norm": 0.7891597151756287, + "learning_rate": 9.50329420773618e-05, + "loss": 0.7268, + "step": 11600 + }, + { + "epoch": 2.0653490028490027, + "grad_norm": 0.8578699827194214, + "learning_rate": 9.501896193160822e-05, + "loss": 0.9872, + "step": 11601 + }, + { + "epoch": 2.0655270655270654, + "grad_norm": 0.7071980834007263, + "learning_rate": 9.500498188344777e-05, + "loss": 0.7278, + "step": 11602 + }, + { + "epoch": 2.065705128205128, + "grad_norm": 0.8434318900108337, + "learning_rate": 9.499100193315436e-05, + "loss": 0.8811, + "step": 11603 + }, + { + "epoch": 2.065883190883191, + "grad_norm": 0.7429414391517639, + "learning_rate": 9.49770220810018e-05, + "loss": 0.8256, + "step": 11604 + }, + { + "epoch": 2.0660612535612537, + "grad_norm": 0.7059712409973145, + "learning_rate": 9.496304232726412e-05, + "loss": 0.8012, + "step": 11605 + }, + { + "epoch": 2.0662393162393164, + "grad_norm": 0.7095850706100464, + "learning_rate": 9.49490626722151e-05, + "loss": 0.8767, + "step": 11606 + }, + { + "epoch": 2.0664173789173788, + "grad_norm": 0.8135038018226624, + "learning_rate": 9.493508311612874e-05, + "loss": 1.1402, + "step": 11607 + }, + { + "epoch": 2.0665954415954415, + "grad_norm": 0.9023036360740662, + "learning_rate": 9.492110365927888e-05, + "loss": 0.7307, + "step": 11608 + }, + { + "epoch": 2.0667735042735043, + "grad_norm": 0.6990833282470703, + "learning_rate": 9.490712430193949e-05, + "loss": 0.7996, + "step": 11609 + }, + { + "epoch": 2.066951566951567, + "grad_norm": 0.7765957713127136, + "learning_rate": 9.489314504438437e-05, + "loss": 0.9721, + "step": 11610 + }, + { + "epoch": 2.0671296296296298, + "grad_norm": 0.883575975894928, + "learning_rate": 9.487916588688749e-05, + "loss": 0.8778, + "step": 11611 + }, + { + "epoch": 2.0673076923076925, + "grad_norm": 0.7226536870002747, + "learning_rate": 9.48651868297227e-05, + "loss": 1.0438, + "step": 11612 + }, + { + "epoch": 2.067485754985755, + "grad_norm": 0.7399018406867981, + "learning_rate": 9.485120787316394e-05, + "loss": 0.8154, + "step": 11613 + }, + { + "epoch": 2.0676638176638176, + "grad_norm": 1.0130186080932617, + "learning_rate": 9.483722901748502e-05, + "loss": 0.9832, + "step": 11614 + }, + { + "epoch": 2.0678418803418803, + "grad_norm": 0.8163331151008606, + "learning_rate": 9.482325026295993e-05, + "loss": 0.8854, + "step": 11615 + }, + { + "epoch": 2.068019943019943, + "grad_norm": 0.7165096998214722, + "learning_rate": 9.480927160986244e-05, + "loss": 0.7636, + "step": 11616 + }, + { + "epoch": 2.068198005698006, + "grad_norm": 0.8579450845718384, + "learning_rate": 9.479529305846652e-05, + "loss": 0.9105, + "step": 11617 + }, + { + "epoch": 2.0683760683760686, + "grad_norm": 0.7062679529190063, + "learning_rate": 9.4781314609046e-05, + "loss": 0.7601, + "step": 11618 + }, + { + "epoch": 2.068554131054131, + "grad_norm": 0.8178739547729492, + "learning_rate": 9.476733626187483e-05, + "loss": 0.8393, + "step": 11619 + }, + { + "epoch": 2.0687321937321936, + "grad_norm": 0.6667241454124451, + "learning_rate": 9.475335801722678e-05, + "loss": 0.8791, + "step": 11620 + }, + { + "epoch": 2.0689102564102564, + "grad_norm": 0.6603145599365234, + "learning_rate": 9.47393798753758e-05, + "loss": 0.7073, + "step": 11621 + }, + { + "epoch": 2.069088319088319, + "grad_norm": 0.7719821333885193, + "learning_rate": 9.472540183659573e-05, + "loss": 0.9195, + "step": 11622 + }, + { + "epoch": 2.069266381766382, + "grad_norm": 0.8059320449829102, + "learning_rate": 9.471142390116045e-05, + "loss": 0.927, + "step": 11623 + }, + { + "epoch": 2.0694444444444446, + "grad_norm": 0.6513992547988892, + "learning_rate": 9.469744606934388e-05, + "loss": 0.857, + "step": 11624 + }, + { + "epoch": 2.069622507122507, + "grad_norm": 0.6948497295379639, + "learning_rate": 9.468346834141979e-05, + "loss": 0.65, + "step": 11625 + }, + { + "epoch": 2.0698005698005697, + "grad_norm": 0.8086618781089783, + "learning_rate": 9.466949071766213e-05, + "loss": 0.8328, + "step": 11626 + }, + { + "epoch": 2.0699786324786325, + "grad_norm": 0.794731616973877, + "learning_rate": 9.465551319834468e-05, + "loss": 0.8641, + "step": 11627 + }, + { + "epoch": 2.070156695156695, + "grad_norm": 0.7312739491462708, + "learning_rate": 9.46415357837414e-05, + "loss": 0.7878, + "step": 11628 + }, + { + "epoch": 2.070334757834758, + "grad_norm": 0.8025211691856384, + "learning_rate": 9.462755847412606e-05, + "loss": 0.8624, + "step": 11629 + }, + { + "epoch": 2.0705128205128207, + "grad_norm": 0.7296801209449768, + "learning_rate": 9.461358126977259e-05, + "loss": 0.7299, + "step": 11630 + }, + { + "epoch": 2.070690883190883, + "grad_norm": 0.7176340222358704, + "learning_rate": 9.459960417095477e-05, + "loss": 0.7374, + "step": 11631 + }, + { + "epoch": 2.0708689458689458, + "grad_norm": 0.7656565308570862, + "learning_rate": 9.45856271779465e-05, + "loss": 0.7791, + "step": 11632 + }, + { + "epoch": 2.0710470085470085, + "grad_norm": 0.7232711315155029, + "learning_rate": 9.457165029102159e-05, + "loss": 0.85, + "step": 11633 + }, + { + "epoch": 2.0712250712250713, + "grad_norm": 0.7342440485954285, + "learning_rate": 9.455767351045397e-05, + "loss": 0.8423, + "step": 11634 + }, + { + "epoch": 2.071403133903134, + "grad_norm": 0.7844834923744202, + "learning_rate": 9.45436968365174e-05, + "loss": 0.9179, + "step": 11635 + }, + { + "epoch": 2.0715811965811968, + "grad_norm": 0.8880203366279602, + "learning_rate": 9.452972026948575e-05, + "loss": 0.7139, + "step": 11636 + }, + { + "epoch": 2.071759259259259, + "grad_norm": 0.7611206769943237, + "learning_rate": 9.451574380963286e-05, + "loss": 0.8915, + "step": 11637 + }, + { + "epoch": 2.071937321937322, + "grad_norm": 0.8123503923416138, + "learning_rate": 9.450176745723262e-05, + "loss": 0.7011, + "step": 11638 + }, + { + "epoch": 2.0721153846153846, + "grad_norm": 0.7703253030776978, + "learning_rate": 9.448779121255879e-05, + "loss": 0.8493, + "step": 11639 + }, + { + "epoch": 2.0722934472934473, + "grad_norm": 0.755836009979248, + "learning_rate": 9.447381507588527e-05, + "loss": 0.9145, + "step": 11640 + }, + { + "epoch": 2.07247150997151, + "grad_norm": 0.7879568338394165, + "learning_rate": 9.445983904748583e-05, + "loss": 0.7761, + "step": 11641 + }, + { + "epoch": 2.072649572649573, + "grad_norm": 0.6695574522018433, + "learning_rate": 9.444586312763434e-05, + "loss": 0.8594, + "step": 11642 + }, + { + "epoch": 2.072827635327635, + "grad_norm": 0.6734640002250671, + "learning_rate": 9.443188731660462e-05, + "loss": 0.7324, + "step": 11643 + }, + { + "epoch": 2.073005698005698, + "grad_norm": 0.7823841571807861, + "learning_rate": 9.441791161467051e-05, + "loss": 1.0223, + "step": 11644 + }, + { + "epoch": 2.0731837606837606, + "grad_norm": 0.8152045011520386, + "learning_rate": 9.440393602210585e-05, + "loss": 1.0364, + "step": 11645 + }, + { + "epoch": 2.0733618233618234, + "grad_norm": 0.8664864897727966, + "learning_rate": 9.438996053918441e-05, + "loss": 0.7607, + "step": 11646 + }, + { + "epoch": 2.073539886039886, + "grad_norm": 0.7949544787406921, + "learning_rate": 9.437598516618006e-05, + "loss": 0.7644, + "step": 11647 + }, + { + "epoch": 2.073717948717949, + "grad_norm": 0.767045259475708, + "learning_rate": 9.436200990336657e-05, + "loss": 0.7563, + "step": 11648 + }, + { + "epoch": 2.073896011396011, + "grad_norm": 0.669129490852356, + "learning_rate": 9.434803475101782e-05, + "loss": 0.7644, + "step": 11649 + }, + { + "epoch": 2.074074074074074, + "grad_norm": 0.7969587445259094, + "learning_rate": 9.433405970940755e-05, + "loss": 1.0249, + "step": 11650 + }, + { + "epoch": 2.0742521367521367, + "grad_norm": 0.6744855642318726, + "learning_rate": 9.432008477880966e-05, + "loss": 0.7478, + "step": 11651 + }, + { + "epoch": 2.0744301994301995, + "grad_norm": 0.6236920356750488, + "learning_rate": 9.430610995949786e-05, + "loss": 0.7309, + "step": 11652 + }, + { + "epoch": 2.074608262108262, + "grad_norm": 0.7952008843421936, + "learning_rate": 9.429213525174603e-05, + "loss": 0.927, + "step": 11653 + }, + { + "epoch": 2.074786324786325, + "grad_norm": 0.7075965404510498, + "learning_rate": 9.427816065582792e-05, + "loss": 0.8494, + "step": 11654 + }, + { + "epoch": 2.0749643874643873, + "grad_norm": 0.8018102049827576, + "learning_rate": 9.426418617201744e-05, + "loss": 0.9261, + "step": 11655 + }, + { + "epoch": 2.07514245014245, + "grad_norm": 0.7155446410179138, + "learning_rate": 9.425021180058824e-05, + "loss": 0.8296, + "step": 11656 + }, + { + "epoch": 2.0753205128205128, + "grad_norm": 0.6611294150352478, + "learning_rate": 9.423623754181425e-05, + "loss": 0.687, + "step": 11657 + }, + { + "epoch": 2.0754985754985755, + "grad_norm": 0.706280529499054, + "learning_rate": 9.422226339596917e-05, + "loss": 0.8664, + "step": 11658 + }, + { + "epoch": 2.0756766381766383, + "grad_norm": 0.7512072324752808, + "learning_rate": 9.420828936332687e-05, + "loss": 0.9074, + "step": 11659 + }, + { + "epoch": 2.075854700854701, + "grad_norm": 0.8833743333816528, + "learning_rate": 9.419431544416108e-05, + "loss": 1.0541, + "step": 11660 + }, + { + "epoch": 2.0760327635327633, + "grad_norm": 0.6991413235664368, + "learning_rate": 9.418034163874564e-05, + "loss": 0.7543, + "step": 11661 + }, + { + "epoch": 2.076210826210826, + "grad_norm": 0.784294605255127, + "learning_rate": 9.41663679473543e-05, + "loss": 0.8156, + "step": 11662 + }, + { + "epoch": 2.076388888888889, + "grad_norm": 0.7716241478919983, + "learning_rate": 9.415239437026086e-05, + "loss": 0.9613, + "step": 11663 + }, + { + "epoch": 2.0765669515669516, + "grad_norm": 0.8247698545455933, + "learning_rate": 9.413842090773914e-05, + "loss": 0.8811, + "step": 11664 + }, + { + "epoch": 2.0767450142450143, + "grad_norm": 0.7988204956054688, + "learning_rate": 9.412444756006283e-05, + "loss": 1.017, + "step": 11665 + }, + { + "epoch": 2.076923076923077, + "grad_norm": 0.8069472908973694, + "learning_rate": 9.411047432750583e-05, + "loss": 0.8292, + "step": 11666 + }, + { + "epoch": 2.0771011396011394, + "grad_norm": 0.8177345991134644, + "learning_rate": 9.40965012103418e-05, + "loss": 0.8004, + "step": 11667 + }, + { + "epoch": 2.077279202279202, + "grad_norm": 0.8589172959327698, + "learning_rate": 9.40825282088446e-05, + "loss": 0.8274, + "step": 11668 + }, + { + "epoch": 2.077457264957265, + "grad_norm": 0.677379846572876, + "learning_rate": 9.406855532328792e-05, + "loss": 0.7044, + "step": 11669 + }, + { + "epoch": 2.0776353276353277, + "grad_norm": 0.9417888522148132, + "learning_rate": 9.405458255394564e-05, + "loss": 0.8418, + "step": 11670 + }, + { + "epoch": 2.0778133903133904, + "grad_norm": 0.7226679921150208, + "learning_rate": 9.404060990109141e-05, + "loss": 0.8496, + "step": 11671 + }, + { + "epoch": 2.077991452991453, + "grad_norm": 0.7451614737510681, + "learning_rate": 9.402663736499909e-05, + "loss": 0.7569, + "step": 11672 + }, + { + "epoch": 2.0781695156695155, + "grad_norm": 0.6516944169998169, + "learning_rate": 9.401266494594235e-05, + "loss": 0.5591, + "step": 11673 + }, + { + "epoch": 2.078347578347578, + "grad_norm": 0.7473219633102417, + "learning_rate": 9.399869264419507e-05, + "loss": 1.0098, + "step": 11674 + }, + { + "epoch": 2.078525641025641, + "grad_norm": 0.8346691131591797, + "learning_rate": 9.398472046003088e-05, + "loss": 1.2654, + "step": 11675 + }, + { + "epoch": 2.0787037037037037, + "grad_norm": 0.8611979484558105, + "learning_rate": 9.397074839372366e-05, + "loss": 0.571, + "step": 11676 + }, + { + "epoch": 2.0788817663817665, + "grad_norm": 0.8093259334564209, + "learning_rate": 9.395677644554705e-05, + "loss": 1.0383, + "step": 11677 + }, + { + "epoch": 2.0790598290598292, + "grad_norm": 0.7954222559928894, + "learning_rate": 9.394280461577488e-05, + "loss": 0.8078, + "step": 11678 + }, + { + "epoch": 2.0792378917378915, + "grad_norm": 0.8380635380744934, + "learning_rate": 9.392883290468083e-05, + "loss": 1.0368, + "step": 11679 + }, + { + "epoch": 2.0794159544159543, + "grad_norm": 0.8427146077156067, + "learning_rate": 9.391486131253874e-05, + "loss": 0.8638, + "step": 11680 + }, + { + "epoch": 2.079594017094017, + "grad_norm": 0.7211564779281616, + "learning_rate": 9.390088983962227e-05, + "loss": 0.8211, + "step": 11681 + }, + { + "epoch": 2.07977207977208, + "grad_norm": 0.7480773329734802, + "learning_rate": 9.388691848620517e-05, + "loss": 0.9313, + "step": 11682 + }, + { + "epoch": 2.0799501424501425, + "grad_norm": 0.8421902060508728, + "learning_rate": 9.387294725256123e-05, + "loss": 0.8808, + "step": 11683 + }, + { + "epoch": 2.0801282051282053, + "grad_norm": 0.7753815650939941, + "learning_rate": 9.385897613896416e-05, + "loss": 0.7738, + "step": 11684 + }, + { + "epoch": 2.0803062678062676, + "grad_norm": 0.8053030967712402, + "learning_rate": 9.384500514568773e-05, + "loss": 0.9194, + "step": 11685 + }, + { + "epoch": 2.0804843304843303, + "grad_norm": 0.7628602981567383, + "learning_rate": 9.383103427300559e-05, + "loss": 0.7866, + "step": 11686 + }, + { + "epoch": 2.080662393162393, + "grad_norm": 0.7087932825088501, + "learning_rate": 9.381706352119156e-05, + "loss": 0.8817, + "step": 11687 + }, + { + "epoch": 2.080840455840456, + "grad_norm": 0.8687152862548828, + "learning_rate": 9.380309289051929e-05, + "loss": 0.8902, + "step": 11688 + }, + { + "epoch": 2.0810185185185186, + "grad_norm": 0.8181152939796448, + "learning_rate": 9.378912238126256e-05, + "loss": 0.9244, + "step": 11689 + }, + { + "epoch": 2.0811965811965814, + "grad_norm": 0.7961983680725098, + "learning_rate": 9.377515199369506e-05, + "loss": 0.714, + "step": 11690 + }, + { + "epoch": 2.0813746438746437, + "grad_norm": 0.8307793736457825, + "learning_rate": 9.376118172809056e-05, + "loss": 0.9573, + "step": 11691 + }, + { + "epoch": 2.0815527065527064, + "grad_norm": 0.7349256277084351, + "learning_rate": 9.374721158472269e-05, + "loss": 0.7533, + "step": 11692 + }, + { + "epoch": 2.081730769230769, + "grad_norm": 0.7625117897987366, + "learning_rate": 9.373324156386526e-05, + "loss": 0.8387, + "step": 11693 + }, + { + "epoch": 2.081908831908832, + "grad_norm": 0.9537683129310608, + "learning_rate": 9.371927166579191e-05, + "loss": 0.9444, + "step": 11694 + }, + { + "epoch": 2.0820868945868947, + "grad_norm": 0.7170497179031372, + "learning_rate": 9.370530189077644e-05, + "loss": 0.9132, + "step": 11695 + }, + { + "epoch": 2.0822649572649574, + "grad_norm": 0.7750041484832764, + "learning_rate": 9.369133223909246e-05, + "loss": 0.6635, + "step": 11696 + }, + { + "epoch": 2.08244301994302, + "grad_norm": 0.8990386128425598, + "learning_rate": 9.367736271101373e-05, + "loss": 0.8692, + "step": 11697 + }, + { + "epoch": 2.0826210826210825, + "grad_norm": 0.5909343361854553, + "learning_rate": 9.366339330681393e-05, + "loss": 0.6811, + "step": 11698 + }, + { + "epoch": 2.0827991452991452, + "grad_norm": 0.7783302068710327, + "learning_rate": 9.364942402676682e-05, + "loss": 1.1024, + "step": 11699 + }, + { + "epoch": 2.082977207977208, + "grad_norm": 0.8926466703414917, + "learning_rate": 9.3635454871146e-05, + "loss": 0.971, + "step": 11700 + }, + { + "epoch": 2.0831552706552707, + "grad_norm": 0.7374816536903381, + "learning_rate": 9.362148584022527e-05, + "loss": 0.7151, + "step": 11701 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 0.7491161227226257, + "learning_rate": 9.360751693427823e-05, + "loss": 0.9213, + "step": 11702 + }, + { + "epoch": 2.083511396011396, + "grad_norm": 0.726859986782074, + "learning_rate": 9.359354815357862e-05, + "loss": 0.8412, + "step": 11703 + }, + { + "epoch": 2.0836894586894585, + "grad_norm": 0.756703794002533, + "learning_rate": 9.357957949840015e-05, + "loss": 0.8074, + "step": 11704 + }, + { + "epoch": 2.0838675213675213, + "grad_norm": 0.8475984334945679, + "learning_rate": 9.356561096901646e-05, + "loss": 0.8926, + "step": 11705 + }, + { + "epoch": 2.084045584045584, + "grad_norm": 0.9776971936225891, + "learning_rate": 9.355164256570129e-05, + "loss": 0.8543, + "step": 11706 + }, + { + "epoch": 2.084223646723647, + "grad_norm": 0.7185834646224976, + "learning_rate": 9.353767428872826e-05, + "loss": 0.6946, + "step": 11707 + }, + { + "epoch": 2.0844017094017095, + "grad_norm": 0.7075535655021667, + "learning_rate": 9.352370613837109e-05, + "loss": 0.7171, + "step": 11708 + }, + { + "epoch": 2.0845797720797723, + "grad_norm": 0.8549726009368896, + "learning_rate": 9.350973811490343e-05, + "loss": 0.9028, + "step": 11709 + }, + { + "epoch": 2.0847578347578346, + "grad_norm": 0.731235682964325, + "learning_rate": 9.3495770218599e-05, + "loss": 0.7703, + "step": 11710 + }, + { + "epoch": 2.0849358974358974, + "grad_norm": 0.8660612106323242, + "learning_rate": 9.34818024497314e-05, + "loss": 0.8464, + "step": 11711 + }, + { + "epoch": 2.08511396011396, + "grad_norm": 0.7687711715698242, + "learning_rate": 9.346783480857439e-05, + "loss": 0.8199, + "step": 11712 + }, + { + "epoch": 2.085292022792023, + "grad_norm": 0.6802884936332703, + "learning_rate": 9.345386729540155e-05, + "loss": 0.7537, + "step": 11713 + }, + { + "epoch": 2.0854700854700856, + "grad_norm": 0.7688863277435303, + "learning_rate": 9.34398999104866e-05, + "loss": 0.8374, + "step": 11714 + }, + { + "epoch": 2.0856481481481484, + "grad_norm": 0.7872602939605713, + "learning_rate": 9.342593265410315e-05, + "loss": 0.8786, + "step": 11715 + }, + { + "epoch": 2.0858262108262107, + "grad_norm": 0.9752106666564941, + "learning_rate": 9.341196552652496e-05, + "loss": 0.9572, + "step": 11716 + }, + { + "epoch": 2.0860042735042734, + "grad_norm": 0.7023422718048096, + "learning_rate": 9.339799852802555e-05, + "loss": 0.7613, + "step": 11717 + }, + { + "epoch": 2.086182336182336, + "grad_norm": 0.8366875052452087, + "learning_rate": 9.338403165887868e-05, + "loss": 0.8206, + "step": 11718 + }, + { + "epoch": 2.086360398860399, + "grad_norm": 0.8534985184669495, + "learning_rate": 9.337006491935794e-05, + "loss": 0.8549, + "step": 11719 + }, + { + "epoch": 2.0865384615384617, + "grad_norm": 0.7902935743331909, + "learning_rate": 9.335609830973707e-05, + "loss": 0.8399, + "step": 11720 + }, + { + "epoch": 2.0867165242165244, + "grad_norm": 0.8064647316932678, + "learning_rate": 9.334213183028958e-05, + "loss": 0.7978, + "step": 11721 + }, + { + "epoch": 2.0868945868945867, + "grad_norm": 0.816412627696991, + "learning_rate": 9.332816548128919e-05, + "loss": 0.8814, + "step": 11722 + }, + { + "epoch": 2.0870726495726495, + "grad_norm": 0.7778908610343933, + "learning_rate": 9.33141992630096e-05, + "loss": 0.9916, + "step": 11723 + }, + { + "epoch": 2.0872507122507122, + "grad_norm": 0.7899400591850281, + "learning_rate": 9.330023317572433e-05, + "loss": 0.5682, + "step": 11724 + }, + { + "epoch": 2.087428774928775, + "grad_norm": 0.6770033836364746, + "learning_rate": 9.32862672197071e-05, + "loss": 0.7327, + "step": 11725 + }, + { + "epoch": 2.0876068376068377, + "grad_norm": 0.8385946750640869, + "learning_rate": 9.327230139523148e-05, + "loss": 0.7793, + "step": 11726 + }, + { + "epoch": 2.0877849002849005, + "grad_norm": 0.708091139793396, + "learning_rate": 9.32583357025712e-05, + "loss": 0.6199, + "step": 11727 + }, + { + "epoch": 2.087962962962963, + "grad_norm": 0.8172122836112976, + "learning_rate": 9.324437014199978e-05, + "loss": 0.749, + "step": 11728 + }, + { + "epoch": 2.0881410256410255, + "grad_norm": 0.818324625492096, + "learning_rate": 9.323040471379091e-05, + "loss": 0.6605, + "step": 11729 + }, + { + "epoch": 2.0883190883190883, + "grad_norm": 0.9074803590774536, + "learning_rate": 9.321643941821819e-05, + "loss": 0.9231, + "step": 11730 + }, + { + "epoch": 2.088497150997151, + "grad_norm": 0.7559560537338257, + "learning_rate": 9.320247425555527e-05, + "loss": 0.7291, + "step": 11731 + }, + { + "epoch": 2.088675213675214, + "grad_norm": 0.8001563549041748, + "learning_rate": 9.318850922607571e-05, + "loss": 0.7781, + "step": 11732 + }, + { + "epoch": 2.0888532763532766, + "grad_norm": 0.7365888953208923, + "learning_rate": 9.31745443300532e-05, + "loss": 0.8243, + "step": 11733 + }, + { + "epoch": 2.089031339031339, + "grad_norm": 0.7861692309379578, + "learning_rate": 9.316057956776126e-05, + "loss": 0.7568, + "step": 11734 + }, + { + "epoch": 2.0892094017094016, + "grad_norm": 0.8399034738540649, + "learning_rate": 9.314661493947363e-05, + "loss": 0.8019, + "step": 11735 + }, + { + "epoch": 2.0893874643874644, + "grad_norm": 0.7718507051467896, + "learning_rate": 9.313265044546378e-05, + "loss": 0.909, + "step": 11736 + }, + { + "epoch": 2.089565527065527, + "grad_norm": 0.8940733671188354, + "learning_rate": 9.311868608600543e-05, + "loss": 0.7154, + "step": 11737 + }, + { + "epoch": 2.08974358974359, + "grad_norm": 0.8506718873977661, + "learning_rate": 9.31047218613721e-05, + "loss": 0.8367, + "step": 11738 + }, + { + "epoch": 2.0899216524216526, + "grad_norm": 0.8431367874145508, + "learning_rate": 9.309075777183743e-05, + "loss": 0.9532, + "step": 11739 + }, + { + "epoch": 2.090099715099715, + "grad_norm": 0.7683414220809937, + "learning_rate": 9.307679381767499e-05, + "loss": 0.9301, + "step": 11740 + }, + { + "epoch": 2.0902777777777777, + "grad_norm": 0.7601380348205566, + "learning_rate": 9.306282999915839e-05, + "loss": 0.8462, + "step": 11741 + }, + { + "epoch": 2.0904558404558404, + "grad_norm": 0.7531782388687134, + "learning_rate": 9.304886631656127e-05, + "loss": 0.8012, + "step": 11742 + }, + { + "epoch": 2.090633903133903, + "grad_norm": 0.7869617938995361, + "learning_rate": 9.303490277015714e-05, + "loss": 0.6645, + "step": 11743 + }, + { + "epoch": 2.090811965811966, + "grad_norm": 0.8042751550674438, + "learning_rate": 9.302093936021964e-05, + "loss": 1.1078, + "step": 11744 + }, + { + "epoch": 2.0909900284900287, + "grad_norm": 0.750350296497345, + "learning_rate": 9.300697608702231e-05, + "loss": 0.8552, + "step": 11745 + }, + { + "epoch": 2.091168091168091, + "grad_norm": 0.7624406814575195, + "learning_rate": 9.29930129508388e-05, + "loss": 0.861, + "step": 11746 + }, + { + "epoch": 2.0913461538461537, + "grad_norm": 0.7634474635124207, + "learning_rate": 9.29790499519426e-05, + "loss": 0.9483, + "step": 11747 + }, + { + "epoch": 2.0915242165242165, + "grad_norm": 0.7312899231910706, + "learning_rate": 9.296508709060738e-05, + "loss": 0.655, + "step": 11748 + }, + { + "epoch": 2.0917022792022792, + "grad_norm": 0.8181857466697693, + "learning_rate": 9.295112436710662e-05, + "loss": 0.7912, + "step": 11749 + }, + { + "epoch": 2.091880341880342, + "grad_norm": 0.6349542737007141, + "learning_rate": 9.293716178171396e-05, + "loss": 0.6268, + "step": 11750 + }, + { + "epoch": 2.0920584045584047, + "grad_norm": 0.8832548260688782, + "learning_rate": 9.292319933470291e-05, + "loss": 0.7805, + "step": 11751 + }, + { + "epoch": 2.092236467236467, + "grad_norm": 0.7251408100128174, + "learning_rate": 9.290923702634712e-05, + "loss": 0.7553, + "step": 11752 + }, + { + "epoch": 2.09241452991453, + "grad_norm": 0.8794457912445068, + "learning_rate": 9.289527485692006e-05, + "loss": 0.9187, + "step": 11753 + }, + { + "epoch": 2.0925925925925926, + "grad_norm": 0.7768839597702026, + "learning_rate": 9.288131282669534e-05, + "loss": 0.9267, + "step": 11754 + }, + { + "epoch": 2.0927706552706553, + "grad_norm": 0.744144856929779, + "learning_rate": 9.28673509359465e-05, + "loss": 0.714, + "step": 11755 + }, + { + "epoch": 2.092948717948718, + "grad_norm": 0.9117433428764343, + "learning_rate": 9.285338918494714e-05, + "loss": 0.9965, + "step": 11756 + }, + { + "epoch": 2.093126780626781, + "grad_norm": 0.8105267286300659, + "learning_rate": 9.283942757397073e-05, + "loss": 0.7517, + "step": 11757 + }, + { + "epoch": 2.093304843304843, + "grad_norm": 0.7348153591156006, + "learning_rate": 9.28254661032909e-05, + "loss": 0.7101, + "step": 11758 + }, + { + "epoch": 2.093482905982906, + "grad_norm": 0.7625702023506165, + "learning_rate": 9.281150477318113e-05, + "loss": 0.6863, + "step": 11759 + }, + { + "epoch": 2.0936609686609686, + "grad_norm": 0.7987569570541382, + "learning_rate": 9.2797543583915e-05, + "loss": 0.8848, + "step": 11760 + }, + { + "epoch": 2.0938390313390314, + "grad_norm": 0.706235408782959, + "learning_rate": 9.278358253576601e-05, + "loss": 0.7375, + "step": 11761 + }, + { + "epoch": 2.094017094017094, + "grad_norm": 0.9716742038726807, + "learning_rate": 9.276962162900774e-05, + "loss": 0.8602, + "step": 11762 + }, + { + "epoch": 2.094195156695157, + "grad_norm": 0.7711777687072754, + "learning_rate": 9.275566086391377e-05, + "loss": 0.8553, + "step": 11763 + }, + { + "epoch": 2.094373219373219, + "grad_norm": 0.8542511463165283, + "learning_rate": 9.274170024075751e-05, + "loss": 0.8412, + "step": 11764 + }, + { + "epoch": 2.094551282051282, + "grad_norm": 0.8255360126495361, + "learning_rate": 9.272773975981259e-05, + "loss": 1.0245, + "step": 11765 + }, + { + "epoch": 2.0947293447293447, + "grad_norm": 0.7416045665740967, + "learning_rate": 9.271377942135248e-05, + "loss": 0.57, + "step": 11766 + }, + { + "epoch": 2.0949074074074074, + "grad_norm": 0.8805620670318604, + "learning_rate": 9.269981922565078e-05, + "loss": 1.0262, + "step": 11767 + }, + { + "epoch": 2.09508547008547, + "grad_norm": 0.7293491363525391, + "learning_rate": 9.26858591729809e-05, + "loss": 0.7945, + "step": 11768 + }, + { + "epoch": 2.095263532763533, + "grad_norm": 0.7949206233024597, + "learning_rate": 9.267189926361643e-05, + "loss": 0.7071, + "step": 11769 + }, + { + "epoch": 2.0954415954415953, + "grad_norm": 0.771806538105011, + "learning_rate": 9.265793949783087e-05, + "loss": 0.8125, + "step": 11770 + }, + { + "epoch": 2.095619658119658, + "grad_norm": 0.7256866693496704, + "learning_rate": 9.264397987589776e-05, + "loss": 0.7607, + "step": 11771 + }, + { + "epoch": 2.0957977207977208, + "grad_norm": 0.8175343871116638, + "learning_rate": 9.263002039809055e-05, + "loss": 0.8486, + "step": 11772 + }, + { + "epoch": 2.0959757834757835, + "grad_norm": 0.7618881464004517, + "learning_rate": 9.261606106468282e-05, + "loss": 0.8182, + "step": 11773 + }, + { + "epoch": 2.0961538461538463, + "grad_norm": 0.7574927806854248, + "learning_rate": 9.2602101875948e-05, + "loss": 0.8703, + "step": 11774 + }, + { + "epoch": 2.096331908831909, + "grad_norm": 0.8639108538627625, + "learning_rate": 9.258814283215964e-05, + "loss": 0.9044, + "step": 11775 + }, + { + "epoch": 2.0965099715099713, + "grad_norm": 0.7221997380256653, + "learning_rate": 9.25741839335912e-05, + "loss": 0.7599, + "step": 11776 + }, + { + "epoch": 2.096688034188034, + "grad_norm": 0.9379764795303345, + "learning_rate": 9.256022518051626e-05, + "loss": 1.0002, + "step": 11777 + }, + { + "epoch": 2.096866096866097, + "grad_norm": 0.8430935740470886, + "learning_rate": 9.25462665732082e-05, + "loss": 0.7711, + "step": 11778 + }, + { + "epoch": 2.0970441595441596, + "grad_norm": 0.8371061086654663, + "learning_rate": 9.253230811194057e-05, + "loss": 0.9028, + "step": 11779 + }, + { + "epoch": 2.0972222222222223, + "grad_norm": 0.6960258483886719, + "learning_rate": 9.251834979698684e-05, + "loss": 0.7491, + "step": 11780 + }, + { + "epoch": 2.097400284900285, + "grad_norm": 0.7736398577690125, + "learning_rate": 9.25043916286205e-05, + "loss": 0.8985, + "step": 11781 + }, + { + "epoch": 2.0975783475783474, + "grad_norm": 0.6901512145996094, + "learning_rate": 9.249043360711509e-05, + "loss": 0.5881, + "step": 11782 + }, + { + "epoch": 2.09775641025641, + "grad_norm": 0.6741603016853333, + "learning_rate": 9.247647573274397e-05, + "loss": 0.7641, + "step": 11783 + }, + { + "epoch": 2.097934472934473, + "grad_norm": 0.736657440662384, + "learning_rate": 9.246251800578074e-05, + "loss": 0.8286, + "step": 11784 + }, + { + "epoch": 2.0981125356125356, + "grad_norm": 0.8235752582550049, + "learning_rate": 9.244856042649877e-05, + "loss": 0.8835, + "step": 11785 + }, + { + "epoch": 2.0982905982905984, + "grad_norm": 0.8083409667015076, + "learning_rate": 9.243460299517158e-05, + "loss": 0.9032, + "step": 11786 + }, + { + "epoch": 2.098468660968661, + "grad_norm": 0.7650952339172363, + "learning_rate": 9.242064571207262e-05, + "loss": 0.775, + "step": 11787 + }, + { + "epoch": 2.0986467236467234, + "grad_norm": 0.7961280345916748, + "learning_rate": 9.24066885774754e-05, + "loss": 0.6308, + "step": 11788 + }, + { + "epoch": 2.098824786324786, + "grad_norm": 0.8032481670379639, + "learning_rate": 9.23927315916533e-05, + "loss": 0.7544, + "step": 11789 + }, + { + "epoch": 2.099002849002849, + "grad_norm": 0.7452995777130127, + "learning_rate": 9.237877475487984e-05, + "loss": 0.8573, + "step": 11790 + }, + { + "epoch": 2.0991809116809117, + "grad_norm": 0.8141751289367676, + "learning_rate": 9.236481806742844e-05, + "loss": 0.9055, + "step": 11791 + }, + { + "epoch": 2.0993589743589745, + "grad_norm": 0.7862252593040466, + "learning_rate": 9.235086152957261e-05, + "loss": 0.6967, + "step": 11792 + }, + { + "epoch": 2.099537037037037, + "grad_norm": 0.771587073802948, + "learning_rate": 9.233690514158571e-05, + "loss": 0.7544, + "step": 11793 + }, + { + "epoch": 2.0997150997150995, + "grad_norm": 0.851445198059082, + "learning_rate": 9.23229489037413e-05, + "loss": 0.9249, + "step": 11794 + }, + { + "epoch": 2.0998931623931623, + "grad_norm": 0.7483612895011902, + "learning_rate": 9.23089928163127e-05, + "loss": 0.747, + "step": 11795 + }, + { + "epoch": 2.100071225071225, + "grad_norm": 0.8493219017982483, + "learning_rate": 9.229503687957342e-05, + "loss": 0.8898, + "step": 11796 + }, + { + "epoch": 2.1002492877492878, + "grad_norm": 0.8331718444824219, + "learning_rate": 9.228108109379687e-05, + "loss": 0.8943, + "step": 11797 + }, + { + "epoch": 2.1004273504273505, + "grad_norm": 0.7756054997444153, + "learning_rate": 9.226712545925655e-05, + "loss": 0.8586, + "step": 11798 + }, + { + "epoch": 2.1006054131054133, + "grad_norm": 0.7292607426643372, + "learning_rate": 9.225316997622579e-05, + "loss": 0.7591, + "step": 11799 + }, + { + "epoch": 2.1007834757834756, + "grad_norm": 0.8575723767280579, + "learning_rate": 9.223921464497811e-05, + "loss": 1.0147, + "step": 11800 + }, + { + "epoch": 2.1009615384615383, + "grad_norm": 0.7882707118988037, + "learning_rate": 9.222525946578687e-05, + "loss": 0.8297, + "step": 11801 + }, + { + "epoch": 2.101139601139601, + "grad_norm": 0.7982630729675293, + "learning_rate": 9.221130443892551e-05, + "loss": 0.9308, + "step": 11802 + }, + { + "epoch": 2.101317663817664, + "grad_norm": 0.7577962279319763, + "learning_rate": 9.219734956466752e-05, + "loss": 0.8474, + "step": 11803 + }, + { + "epoch": 2.1014957264957266, + "grad_norm": 0.7103776335716248, + "learning_rate": 9.218339484328621e-05, + "loss": 0.7863, + "step": 11804 + }, + { + "epoch": 2.1016737891737893, + "grad_norm": 0.8307296633720398, + "learning_rate": 9.216944027505505e-05, + "loss": 0.7633, + "step": 11805 + }, + { + "epoch": 2.1018518518518516, + "grad_norm": 0.8197653293609619, + "learning_rate": 9.215548586024743e-05, + "loss": 0.8987, + "step": 11806 + }, + { + "epoch": 2.1020299145299144, + "grad_norm": 0.9192719459533691, + "learning_rate": 9.21415315991368e-05, + "loss": 0.7829, + "step": 11807 + }, + { + "epoch": 2.102207977207977, + "grad_norm": 0.7249892354011536, + "learning_rate": 9.21275774919965e-05, + "loss": 0.9143, + "step": 11808 + }, + { + "epoch": 2.10238603988604, + "grad_norm": 0.7942582368850708, + "learning_rate": 9.211362353910002e-05, + "loss": 0.8634, + "step": 11809 + }, + { + "epoch": 2.1025641025641026, + "grad_norm": 0.7773341536521912, + "learning_rate": 9.209966974072065e-05, + "loss": 0.7865, + "step": 11810 + }, + { + "epoch": 2.1027421652421654, + "grad_norm": 0.802175760269165, + "learning_rate": 9.208571609713185e-05, + "loss": 0.7473, + "step": 11811 + }, + { + "epoch": 2.1029202279202277, + "grad_norm": 1.0248547792434692, + "learning_rate": 9.207176260860701e-05, + "loss": 1.0097, + "step": 11812 + }, + { + "epoch": 2.1030982905982905, + "grad_norm": 0.5781275629997253, + "learning_rate": 9.205780927541954e-05, + "loss": 0.5813, + "step": 11813 + }, + { + "epoch": 2.103276353276353, + "grad_norm": 0.7252389192581177, + "learning_rate": 9.204385609784274e-05, + "loss": 0.7978, + "step": 11814 + }, + { + "epoch": 2.103454415954416, + "grad_norm": 0.8497771620750427, + "learning_rate": 9.20299030761501e-05, + "loss": 0.95, + "step": 11815 + }, + { + "epoch": 2.1036324786324787, + "grad_norm": 0.8420650362968445, + "learning_rate": 9.201595021061491e-05, + "loss": 0.9693, + "step": 11816 + }, + { + "epoch": 2.1038105413105415, + "grad_norm": 0.8286302089691162, + "learning_rate": 9.200199750151063e-05, + "loss": 0.9457, + "step": 11817 + }, + { + "epoch": 2.103988603988604, + "grad_norm": 0.877740204334259, + "learning_rate": 9.198804494911057e-05, + "loss": 0.9082, + "step": 11818 + }, + { + "epoch": 2.1041666666666665, + "grad_norm": 0.7579863667488098, + "learning_rate": 9.197409255368817e-05, + "loss": 0.7681, + "step": 11819 + }, + { + "epoch": 2.1043447293447293, + "grad_norm": 0.7141458988189697, + "learning_rate": 9.19601403155167e-05, + "loss": 0.659, + "step": 11820 + }, + { + "epoch": 2.104522792022792, + "grad_norm": 0.8493850827217102, + "learning_rate": 9.194618823486958e-05, + "loss": 0.8197, + "step": 11821 + }, + { + "epoch": 2.1047008547008548, + "grad_norm": 0.8319337368011475, + "learning_rate": 9.193223631202019e-05, + "loss": 0.8955, + "step": 11822 + }, + { + "epoch": 2.1048789173789175, + "grad_norm": 0.7180153727531433, + "learning_rate": 9.191828454724186e-05, + "loss": 0.8068, + "step": 11823 + }, + { + "epoch": 2.10505698005698, + "grad_norm": 0.6748450398445129, + "learning_rate": 9.190433294080799e-05, + "loss": 0.7469, + "step": 11824 + }, + { + "epoch": 2.1052350427350426, + "grad_norm": 0.7750198841094971, + "learning_rate": 9.189038149299186e-05, + "loss": 0.932, + "step": 11825 + }, + { + "epoch": 2.1054131054131053, + "grad_norm": 0.7763389945030212, + "learning_rate": 9.187643020406688e-05, + "loss": 0.8027, + "step": 11826 + }, + { + "epoch": 2.105591168091168, + "grad_norm": 0.8382455110549927, + "learning_rate": 9.186247907430636e-05, + "loss": 0.8288, + "step": 11827 + }, + { + "epoch": 2.105769230769231, + "grad_norm": 0.6744221448898315, + "learning_rate": 9.184852810398367e-05, + "loss": 0.6807, + "step": 11828 + }, + { + "epoch": 2.1059472934472936, + "grad_norm": 0.7798452377319336, + "learning_rate": 9.183457729337212e-05, + "loss": 0.9853, + "step": 11829 + }, + { + "epoch": 2.1061253561253563, + "grad_norm": 0.7377058863639832, + "learning_rate": 9.182062664274513e-05, + "loss": 0.9043, + "step": 11830 + }, + { + "epoch": 2.1063034188034186, + "grad_norm": 0.8190791010856628, + "learning_rate": 9.180667615237589e-05, + "loss": 0.9786, + "step": 11831 + }, + { + "epoch": 2.1064814814814814, + "grad_norm": 0.7629963755607605, + "learning_rate": 9.179272582253785e-05, + "loss": 0.9168, + "step": 11832 + }, + { + "epoch": 2.106659544159544, + "grad_norm": 0.7753663063049316, + "learning_rate": 9.177877565350426e-05, + "loss": 0.963, + "step": 11833 + }, + { + "epoch": 2.106837606837607, + "grad_norm": 0.7842921614646912, + "learning_rate": 9.176482564554855e-05, + "loss": 0.8194, + "step": 11834 + }, + { + "epoch": 2.1070156695156697, + "grad_norm": 0.6640288233757019, + "learning_rate": 9.175087579894393e-05, + "loss": 0.6227, + "step": 11835 + }, + { + "epoch": 2.1071937321937324, + "grad_norm": 0.8474540710449219, + "learning_rate": 9.173692611396376e-05, + "loss": 0.7817, + "step": 11836 + }, + { + "epoch": 2.1073717948717947, + "grad_norm": 0.7123007774353027, + "learning_rate": 9.172297659088135e-05, + "loss": 0.9508, + "step": 11837 + }, + { + "epoch": 2.1075498575498575, + "grad_norm": 0.7418060898780823, + "learning_rate": 9.170902722997007e-05, + "loss": 0.8832, + "step": 11838 + }, + { + "epoch": 2.10772792022792, + "grad_norm": 0.7899464964866638, + "learning_rate": 9.169507803150313e-05, + "loss": 0.8474, + "step": 11839 + }, + { + "epoch": 2.107905982905983, + "grad_norm": 0.7543701529502869, + "learning_rate": 9.168112899575388e-05, + "loss": 0.8113, + "step": 11840 + }, + { + "epoch": 2.1080840455840457, + "grad_norm": 0.8057922720909119, + "learning_rate": 9.166718012299565e-05, + "loss": 0.998, + "step": 11841 + }, + { + "epoch": 2.1082621082621085, + "grad_norm": 0.7879176139831543, + "learning_rate": 9.16532314135017e-05, + "loss": 1.0509, + "step": 11842 + }, + { + "epoch": 2.1084401709401708, + "grad_norm": 0.8796642422676086, + "learning_rate": 9.163928286754537e-05, + "loss": 1.0481, + "step": 11843 + }, + { + "epoch": 2.1086182336182335, + "grad_norm": 0.7158889174461365, + "learning_rate": 9.16253344853999e-05, + "loss": 0.796, + "step": 11844 + }, + { + "epoch": 2.1087962962962963, + "grad_norm": 0.8020899295806885, + "learning_rate": 9.161138626733863e-05, + "loss": 0.822, + "step": 11845 + }, + { + "epoch": 2.108974358974359, + "grad_norm": 0.7217469215393066, + "learning_rate": 9.159743821363478e-05, + "loss": 1.0037, + "step": 11846 + }, + { + "epoch": 2.109152421652422, + "grad_norm": 0.762450098991394, + "learning_rate": 9.158349032456171e-05, + "loss": 1.0047, + "step": 11847 + }, + { + "epoch": 2.1093304843304845, + "grad_norm": 0.7227019667625427, + "learning_rate": 9.156954260039263e-05, + "loss": 0.8034, + "step": 11848 + }, + { + "epoch": 2.109508547008547, + "grad_norm": 0.7358957529067993, + "learning_rate": 9.155559504140089e-05, + "loss": 0.9483, + "step": 11849 + }, + { + "epoch": 2.1096866096866096, + "grad_norm": 0.7039931416511536, + "learning_rate": 9.154164764785968e-05, + "loss": 0.9255, + "step": 11850 + }, + { + "epoch": 2.1098646723646723, + "grad_norm": 0.8479618430137634, + "learning_rate": 9.152770042004234e-05, + "loss": 0.7379, + "step": 11851 + }, + { + "epoch": 2.110042735042735, + "grad_norm": 0.8320785164833069, + "learning_rate": 9.151375335822208e-05, + "loss": 0.944, + "step": 11852 + }, + { + "epoch": 2.110220797720798, + "grad_norm": 0.8186322450637817, + "learning_rate": 9.149980646267225e-05, + "loss": 0.7757, + "step": 11853 + }, + { + "epoch": 2.1103988603988606, + "grad_norm": 0.7816671133041382, + "learning_rate": 9.148585973366601e-05, + "loss": 0.8592, + "step": 11854 + }, + { + "epoch": 2.110576923076923, + "grad_norm": 0.8747152090072632, + "learning_rate": 9.147191317147671e-05, + "loss": 1.0852, + "step": 11855 + }, + { + "epoch": 2.1107549857549857, + "grad_norm": 0.7762712240219116, + "learning_rate": 9.14579667763775e-05, + "loss": 0.8466, + "step": 11856 + }, + { + "epoch": 2.1109330484330484, + "grad_norm": 0.8426344394683838, + "learning_rate": 9.144402054864171e-05, + "loss": 0.9949, + "step": 11857 + }, + { + "epoch": 2.111111111111111, + "grad_norm": 0.7581121921539307, + "learning_rate": 9.143007448854256e-05, + "loss": 0.748, + "step": 11858 + }, + { + "epoch": 2.111289173789174, + "grad_norm": 0.837939977645874, + "learning_rate": 9.141612859635333e-05, + "loss": 0.9479, + "step": 11859 + }, + { + "epoch": 2.1114672364672367, + "grad_norm": 0.7402070760726929, + "learning_rate": 9.140218287234718e-05, + "loss": 0.7829, + "step": 11860 + }, + { + "epoch": 2.111645299145299, + "grad_norm": 0.7125605344772339, + "learning_rate": 9.13882373167974e-05, + "loss": 1.0175, + "step": 11861 + }, + { + "epoch": 2.1118233618233617, + "grad_norm": 0.8021374344825745, + "learning_rate": 9.137429192997723e-05, + "loss": 0.9258, + "step": 11862 + }, + { + "epoch": 2.1120014245014245, + "grad_norm": 0.7860891222953796, + "learning_rate": 9.136034671215988e-05, + "loss": 0.7351, + "step": 11863 + }, + { + "epoch": 2.1121794871794872, + "grad_norm": 0.8324207067489624, + "learning_rate": 9.134640166361864e-05, + "loss": 0.8933, + "step": 11864 + }, + { + "epoch": 2.11235754985755, + "grad_norm": 0.8209179639816284, + "learning_rate": 9.133245678462663e-05, + "loss": 0.6983, + "step": 11865 + }, + { + "epoch": 2.1125356125356127, + "grad_norm": 0.7071694731712341, + "learning_rate": 9.131851207545716e-05, + "loss": 0.7796, + "step": 11866 + }, + { + "epoch": 2.112713675213675, + "grad_norm": 0.8126310110092163, + "learning_rate": 9.130456753638339e-05, + "loss": 0.8887, + "step": 11867 + }, + { + "epoch": 2.112891737891738, + "grad_norm": 0.7713829874992371, + "learning_rate": 9.129062316767855e-05, + "loss": 0.8169, + "step": 11868 + }, + { + "epoch": 2.1130698005698005, + "grad_norm": 0.8065944314002991, + "learning_rate": 9.127667896961585e-05, + "loss": 0.9295, + "step": 11869 + }, + { + "epoch": 2.1132478632478633, + "grad_norm": 0.7433435320854187, + "learning_rate": 9.126273494246856e-05, + "loss": 1.089, + "step": 11870 + }, + { + "epoch": 2.113425925925926, + "grad_norm": 0.8168141841888428, + "learning_rate": 9.124879108650978e-05, + "loss": 0.7914, + "step": 11871 + }, + { + "epoch": 2.113603988603989, + "grad_norm": 0.7703335285186768, + "learning_rate": 9.123484740201276e-05, + "loss": 1.0599, + "step": 11872 + }, + { + "epoch": 2.113782051282051, + "grad_norm": 0.810584545135498, + "learning_rate": 9.12209038892507e-05, + "loss": 0.886, + "step": 11873 + }, + { + "epoch": 2.113960113960114, + "grad_norm": 0.8441819548606873, + "learning_rate": 9.120696054849683e-05, + "loss": 0.9069, + "step": 11874 + }, + { + "epoch": 2.1141381766381766, + "grad_norm": 0.816067636013031, + "learning_rate": 9.119301738002425e-05, + "loss": 0.9084, + "step": 11875 + }, + { + "epoch": 2.1143162393162394, + "grad_norm": 0.8595525622367859, + "learning_rate": 9.117907438410622e-05, + "loss": 0.916, + "step": 11876 + }, + { + "epoch": 2.114494301994302, + "grad_norm": 0.8604792356491089, + "learning_rate": 9.116513156101589e-05, + "loss": 1.1207, + "step": 11877 + }, + { + "epoch": 2.114672364672365, + "grad_norm": 0.673664927482605, + "learning_rate": 9.115118891102649e-05, + "loss": 0.9767, + "step": 11878 + }, + { + "epoch": 2.114850427350427, + "grad_norm": 0.7064382433891296, + "learning_rate": 9.113724643441113e-05, + "loss": 0.91, + "step": 11879 + }, + { + "epoch": 2.11502849002849, + "grad_norm": 0.7256918549537659, + "learning_rate": 9.112330413144301e-05, + "loss": 0.9061, + "step": 11880 + }, + { + "epoch": 2.1152065527065527, + "grad_norm": 0.7914155721664429, + "learning_rate": 9.110936200239534e-05, + "loss": 0.6652, + "step": 11881 + }, + { + "epoch": 2.1153846153846154, + "grad_norm": 0.7484595775604248, + "learning_rate": 9.109542004754122e-05, + "loss": 0.8049, + "step": 11882 + }, + { + "epoch": 2.115562678062678, + "grad_norm": 0.8062677979469299, + "learning_rate": 9.108147826715387e-05, + "loss": 0.8671, + "step": 11883 + }, + { + "epoch": 2.115740740740741, + "grad_norm": 0.9595313668251038, + "learning_rate": 9.10675366615064e-05, + "loss": 0.995, + "step": 11884 + }, + { + "epoch": 2.1159188034188032, + "grad_norm": 0.7263179421424866, + "learning_rate": 9.105359523087203e-05, + "loss": 0.9177, + "step": 11885 + }, + { + "epoch": 2.116096866096866, + "grad_norm": 0.900650680065155, + "learning_rate": 9.103965397552385e-05, + "loss": 0.8599, + "step": 11886 + }, + { + "epoch": 2.1162749287749287, + "grad_norm": 0.7682752013206482, + "learning_rate": 9.102571289573506e-05, + "loss": 0.8942, + "step": 11887 + }, + { + "epoch": 2.1164529914529915, + "grad_norm": 0.7076446413993835, + "learning_rate": 9.101177199177874e-05, + "loss": 0.7498, + "step": 11888 + }, + { + "epoch": 2.1166310541310542, + "grad_norm": 0.711475133895874, + "learning_rate": 9.099783126392813e-05, + "loss": 0.7035, + "step": 11889 + }, + { + "epoch": 2.116809116809117, + "grad_norm": 0.6720870137214661, + "learning_rate": 9.098389071245627e-05, + "loss": 0.7315, + "step": 11890 + }, + { + "epoch": 2.1169871794871793, + "grad_norm": 0.8207699656486511, + "learning_rate": 9.096995033763639e-05, + "loss": 0.7465, + "step": 11891 + }, + { + "epoch": 2.117165242165242, + "grad_norm": 0.9032317996025085, + "learning_rate": 9.095601013974153e-05, + "loss": 0.9209, + "step": 11892 + }, + { + "epoch": 2.117343304843305, + "grad_norm": 0.886545717716217, + "learning_rate": 9.094207011904489e-05, + "loss": 0.9411, + "step": 11893 + }, + { + "epoch": 2.1175213675213675, + "grad_norm": 0.8235130906105042, + "learning_rate": 9.092813027581953e-05, + "loss": 0.9264, + "step": 11894 + }, + { + "epoch": 2.1176994301994303, + "grad_norm": 0.7530205845832825, + "learning_rate": 9.091419061033867e-05, + "loss": 0.8926, + "step": 11895 + }, + { + "epoch": 2.117877492877493, + "grad_norm": 0.8329548835754395, + "learning_rate": 9.090025112287533e-05, + "loss": 0.9615, + "step": 11896 + }, + { + "epoch": 2.1180555555555554, + "grad_norm": 0.8184738755226135, + "learning_rate": 9.088631181370269e-05, + "loss": 0.9069, + "step": 11897 + }, + { + "epoch": 2.118233618233618, + "grad_norm": 0.8071370720863342, + "learning_rate": 9.087237268309381e-05, + "loss": 0.8721, + "step": 11898 + }, + { + "epoch": 2.118411680911681, + "grad_norm": 0.8995245695114136, + "learning_rate": 9.085843373132187e-05, + "loss": 0.8815, + "step": 11899 + }, + { + "epoch": 2.1185897435897436, + "grad_norm": 0.7601714730262756, + "learning_rate": 9.084449495865989e-05, + "loss": 0.6824, + "step": 11900 + }, + { + "epoch": 2.1187678062678064, + "grad_norm": 0.8499618172645569, + "learning_rate": 9.083055636538101e-05, + "loss": 0.9868, + "step": 11901 + }, + { + "epoch": 2.118945868945869, + "grad_norm": 0.8190310001373291, + "learning_rate": 9.081661795175837e-05, + "loss": 0.8156, + "step": 11902 + }, + { + "epoch": 2.1191239316239314, + "grad_norm": 0.8340418934822083, + "learning_rate": 9.080267971806498e-05, + "loss": 1.0153, + "step": 11903 + }, + { + "epoch": 2.119301994301994, + "grad_norm": 0.8460756540298462, + "learning_rate": 9.0788741664574e-05, + "loss": 0.8752, + "step": 11904 + }, + { + "epoch": 2.119480056980057, + "grad_norm": 0.7457373738288879, + "learning_rate": 9.077480379155848e-05, + "loss": 0.9105, + "step": 11905 + }, + { + "epoch": 2.1196581196581197, + "grad_norm": 0.7883822917938232, + "learning_rate": 9.076086609929155e-05, + "loss": 0.8782, + "step": 11906 + }, + { + "epoch": 2.1198361823361824, + "grad_norm": 0.912143886089325, + "learning_rate": 9.074692858804622e-05, + "loss": 0.9898, + "step": 11907 + }, + { + "epoch": 2.120014245014245, + "grad_norm": 0.7801905274391174, + "learning_rate": 9.073299125809562e-05, + "loss": 1.091, + "step": 11908 + }, + { + "epoch": 2.1201923076923075, + "grad_norm": 0.6836256384849548, + "learning_rate": 9.071905410971279e-05, + "loss": 0.7967, + "step": 11909 + }, + { + "epoch": 2.1203703703703702, + "grad_norm": 0.7656795382499695, + "learning_rate": 9.070511714317085e-05, + "loss": 0.9696, + "step": 11910 + }, + { + "epoch": 2.120548433048433, + "grad_norm": 0.7010015249252319, + "learning_rate": 9.06911803587428e-05, + "loss": 0.6501, + "step": 11911 + }, + { + "epoch": 2.1207264957264957, + "grad_norm": 0.6673064827919006, + "learning_rate": 9.067724375670174e-05, + "loss": 0.5663, + "step": 11912 + }, + { + "epoch": 2.1209045584045585, + "grad_norm": 0.8683220148086548, + "learning_rate": 9.06633073373207e-05, + "loss": 0.9722, + "step": 11913 + }, + { + "epoch": 2.1210826210826212, + "grad_norm": 0.7793976068496704, + "learning_rate": 9.06493711008728e-05, + "loss": 0.7595, + "step": 11914 + }, + { + "epoch": 2.1212606837606836, + "grad_norm": 0.7803528308868408, + "learning_rate": 9.0635435047631e-05, + "loss": 0.9262, + "step": 11915 + }, + { + "epoch": 2.1214387464387463, + "grad_norm": 0.8067244291305542, + "learning_rate": 9.062149917786846e-05, + "loss": 0.9376, + "step": 11916 + }, + { + "epoch": 2.121616809116809, + "grad_norm": 0.7389153838157654, + "learning_rate": 9.060756349185812e-05, + "loss": 0.7414, + "step": 11917 + }, + { + "epoch": 2.121794871794872, + "grad_norm": 0.7717151045799255, + "learning_rate": 9.059362798987308e-05, + "loss": 0.7261, + "step": 11918 + }, + { + "epoch": 2.1219729344729346, + "grad_norm": 0.7668650150299072, + "learning_rate": 9.057969267218632e-05, + "loss": 0.8145, + "step": 11919 + }, + { + "epoch": 2.1221509971509973, + "grad_norm": 1.0015910863876343, + "learning_rate": 9.056575753907093e-05, + "loss": 0.8997, + "step": 11920 + }, + { + "epoch": 2.1223290598290596, + "grad_norm": 0.8731024861335754, + "learning_rate": 9.055182259079997e-05, + "loss": 1.0101, + "step": 11921 + }, + { + "epoch": 2.1225071225071224, + "grad_norm": 0.7662718892097473, + "learning_rate": 9.053788782764637e-05, + "loss": 0.697, + "step": 11922 + }, + { + "epoch": 2.122685185185185, + "grad_norm": 0.7783135771751404, + "learning_rate": 9.05239532498832e-05, + "loss": 0.8506, + "step": 11923 + }, + { + "epoch": 2.122863247863248, + "grad_norm": 0.8667652606964111, + "learning_rate": 9.05100188577835e-05, + "loss": 0.9851, + "step": 11924 + }, + { + "epoch": 2.1230413105413106, + "grad_norm": 0.7785412073135376, + "learning_rate": 9.049608465162028e-05, + "loss": 0.5924, + "step": 11925 + }, + { + "epoch": 2.1232193732193734, + "grad_norm": 0.7968559861183167, + "learning_rate": 9.04821506316665e-05, + "loss": 0.8114, + "step": 11926 + }, + { + "epoch": 2.123397435897436, + "grad_norm": 0.8065921068191528, + "learning_rate": 9.046821679819527e-05, + "loss": 0.9045, + "step": 11927 + }, + { + "epoch": 2.1235754985754984, + "grad_norm": 0.7509779930114746, + "learning_rate": 9.045428315147948e-05, + "loss": 0.7337, + "step": 11928 + }, + { + "epoch": 2.123753561253561, + "grad_norm": 0.8174976110458374, + "learning_rate": 9.044034969179219e-05, + "loss": 1.0113, + "step": 11929 + }, + { + "epoch": 2.123931623931624, + "grad_norm": 0.8723294734954834, + "learning_rate": 9.042641641940638e-05, + "loss": 0.9657, + "step": 11930 + }, + { + "epoch": 2.1241096866096867, + "grad_norm": 0.7412081360816956, + "learning_rate": 9.041248333459509e-05, + "loss": 0.9311, + "step": 11931 + }, + { + "epoch": 2.1242877492877494, + "grad_norm": 0.7376424670219421, + "learning_rate": 9.039855043763124e-05, + "loss": 0.7039, + "step": 11932 + }, + { + "epoch": 2.1244658119658117, + "grad_norm": 0.8002118468284607, + "learning_rate": 9.038461772878786e-05, + "loss": 0.9555, + "step": 11933 + }, + { + "epoch": 2.1246438746438745, + "grad_norm": 0.7221434712409973, + "learning_rate": 9.03706852083379e-05, + "loss": 0.8462, + "step": 11934 + }, + { + "epoch": 2.1248219373219372, + "grad_norm": 0.8506385684013367, + "learning_rate": 9.035675287655441e-05, + "loss": 0.7977, + "step": 11935 + }, + { + "epoch": 2.125, + "grad_norm": 0.8088411688804626, + "learning_rate": 9.034282073371025e-05, + "loss": 1.0146, + "step": 11936 + }, + { + "epoch": 2.1251780626780628, + "grad_norm": 0.9231638312339783, + "learning_rate": 9.032888878007853e-05, + "loss": 0.7017, + "step": 11937 + }, + { + "epoch": 2.1253561253561255, + "grad_norm": 0.721066951751709, + "learning_rate": 9.03149570159321e-05, + "loss": 0.7662, + "step": 11938 + }, + { + "epoch": 2.1255341880341883, + "grad_norm": 0.7804762721061707, + "learning_rate": 9.030102544154395e-05, + "loss": 0.6835, + "step": 11939 + }, + { + "epoch": 2.1257122507122506, + "grad_norm": 0.9728445410728455, + "learning_rate": 9.028709405718707e-05, + "loss": 0.9161, + "step": 11940 + }, + { + "epoch": 2.1258903133903133, + "grad_norm": 0.8209855556488037, + "learning_rate": 9.02731628631344e-05, + "loss": 0.7492, + "step": 11941 + }, + { + "epoch": 2.126068376068376, + "grad_norm": 0.7054622769355774, + "learning_rate": 9.025923185965896e-05, + "loss": 0.7908, + "step": 11942 + }, + { + "epoch": 2.126246438746439, + "grad_norm": 0.73018878698349, + "learning_rate": 9.024530104703358e-05, + "loss": 0.7902, + "step": 11943 + }, + { + "epoch": 2.1264245014245016, + "grad_norm": 0.73788982629776, + "learning_rate": 9.023137042553127e-05, + "loss": 0.8473, + "step": 11944 + }, + { + "epoch": 2.126602564102564, + "grad_norm": 0.7733396291732788, + "learning_rate": 9.021743999542495e-05, + "loss": 0.9595, + "step": 11945 + }, + { + "epoch": 2.1267806267806266, + "grad_norm": 0.9066760540008545, + "learning_rate": 9.020350975698761e-05, + "loss": 0.8517, + "step": 11946 + }, + { + "epoch": 2.1269586894586894, + "grad_norm": 0.7552717328071594, + "learning_rate": 9.018957971049211e-05, + "loss": 0.6802, + "step": 11947 + }, + { + "epoch": 2.127136752136752, + "grad_norm": 0.7437541484832764, + "learning_rate": 9.017564985621144e-05, + "loss": 0.9365, + "step": 11948 + }, + { + "epoch": 2.127314814814815, + "grad_norm": 0.8216256499290466, + "learning_rate": 9.016172019441847e-05, + "loss": 0.9019, + "step": 11949 + }, + { + "epoch": 2.1274928774928776, + "grad_norm": 0.752247154712677, + "learning_rate": 9.014779072538621e-05, + "loss": 0.7771, + "step": 11950 + }, + { + "epoch": 2.1276709401709404, + "grad_norm": 0.7714348435401917, + "learning_rate": 9.013386144938748e-05, + "loss": 0.8495, + "step": 11951 + }, + { + "epoch": 2.1278490028490027, + "grad_norm": 0.8347537517547607, + "learning_rate": 9.011993236669529e-05, + "loss": 0.861, + "step": 11952 + }, + { + "epoch": 2.1280270655270654, + "grad_norm": 0.8180193901062012, + "learning_rate": 9.010600347758245e-05, + "loss": 0.9059, + "step": 11953 + }, + { + "epoch": 2.128205128205128, + "grad_norm": 0.7328528761863708, + "learning_rate": 9.009207478232193e-05, + "loss": 0.9144, + "step": 11954 + }, + { + "epoch": 2.128383190883191, + "grad_norm": 0.7590839862823486, + "learning_rate": 9.007814628118661e-05, + "loss": 0.8642, + "step": 11955 + }, + { + "epoch": 2.1285612535612537, + "grad_norm": 0.7962782382965088, + "learning_rate": 9.006421797444945e-05, + "loss": 0.8958, + "step": 11956 + }, + { + "epoch": 2.128739316239316, + "grad_norm": 0.7302426695823669, + "learning_rate": 9.005028986238325e-05, + "loss": 0.9419, + "step": 11957 + }, + { + "epoch": 2.1289173789173788, + "grad_norm": 0.9223780632019043, + "learning_rate": 9.003636194526098e-05, + "loss": 0.7631, + "step": 11958 + }, + { + "epoch": 2.1290954415954415, + "grad_norm": 0.728225588798523, + "learning_rate": 9.002243422335547e-05, + "loss": 0.7705, + "step": 11959 + }, + { + "epoch": 2.1292735042735043, + "grad_norm": 0.8519338369369507, + "learning_rate": 9.000850669693964e-05, + "loss": 0.8962, + "step": 11960 + }, + { + "epoch": 2.129451566951567, + "grad_norm": 0.8920532464981079, + "learning_rate": 8.999457936628641e-05, + "loss": 0.618, + "step": 11961 + }, + { + "epoch": 2.1296296296296298, + "grad_norm": 0.9719427824020386, + "learning_rate": 8.998065223166857e-05, + "loss": 0.9142, + "step": 11962 + }, + { + "epoch": 2.1298076923076925, + "grad_norm": 0.8130887150764465, + "learning_rate": 8.996672529335908e-05, + "loss": 1.0246, + "step": 11963 + }, + { + "epoch": 2.129985754985755, + "grad_norm": 0.7682677507400513, + "learning_rate": 8.995279855163073e-05, + "loss": 0.7964, + "step": 11964 + }, + { + "epoch": 2.1301638176638176, + "grad_norm": 0.8507778644561768, + "learning_rate": 8.993887200675641e-05, + "loss": 0.756, + "step": 11965 + }, + { + "epoch": 2.1303418803418803, + "grad_norm": 0.815487802028656, + "learning_rate": 8.992494565900901e-05, + "loss": 0.7596, + "step": 11966 + }, + { + "epoch": 2.130519943019943, + "grad_norm": 0.8560892939567566, + "learning_rate": 8.991101950866138e-05, + "loss": 0.9939, + "step": 11967 + }, + { + "epoch": 2.130698005698006, + "grad_norm": 0.8737899661064148, + "learning_rate": 8.989709355598635e-05, + "loss": 0.9235, + "step": 11968 + }, + { + "epoch": 2.1308760683760686, + "grad_norm": 0.8434267640113831, + "learning_rate": 8.98831678012568e-05, + "loss": 0.7832, + "step": 11969 + }, + { + "epoch": 2.131054131054131, + "grad_norm": 0.8286582827568054, + "learning_rate": 8.986924224474553e-05, + "loss": 1.0591, + "step": 11970 + }, + { + "epoch": 2.1312321937321936, + "grad_norm": 0.8023663759231567, + "learning_rate": 8.985531688672546e-05, + "loss": 0.935, + "step": 11971 + }, + { + "epoch": 2.1314102564102564, + "grad_norm": 0.6504420042037964, + "learning_rate": 8.984139172746933e-05, + "loss": 0.79, + "step": 11972 + }, + { + "epoch": 2.131588319088319, + "grad_norm": 0.8969349265098572, + "learning_rate": 8.982746676725009e-05, + "loss": 1.0531, + "step": 11973 + }, + { + "epoch": 2.131766381766382, + "grad_norm": 0.802094042301178, + "learning_rate": 8.981354200634046e-05, + "loss": 0.8873, + "step": 11974 + }, + { + "epoch": 2.1319444444444446, + "grad_norm": 0.7630797624588013, + "learning_rate": 8.979961744501332e-05, + "loss": 0.9299, + "step": 11975 + }, + { + "epoch": 2.132122507122507, + "grad_norm": 0.8395546674728394, + "learning_rate": 8.978569308354148e-05, + "loss": 0.922, + "step": 11976 + }, + { + "epoch": 2.1323005698005697, + "grad_norm": 0.9325534701347351, + "learning_rate": 8.97717689221978e-05, + "loss": 0.9156, + "step": 11977 + }, + { + "epoch": 2.1324786324786325, + "grad_norm": 0.8139503002166748, + "learning_rate": 8.975784496125502e-05, + "loss": 0.8882, + "step": 11978 + }, + { + "epoch": 2.132656695156695, + "grad_norm": 1.0311007499694824, + "learning_rate": 8.974392120098599e-05, + "loss": 1.0068, + "step": 11979 + }, + { + "epoch": 2.132834757834758, + "grad_norm": 0.9328663945198059, + "learning_rate": 8.972999764166354e-05, + "loss": 0.8313, + "step": 11980 + }, + { + "epoch": 2.1330128205128207, + "grad_norm": 0.747276782989502, + "learning_rate": 8.971607428356044e-05, + "loss": 0.9302, + "step": 11981 + }, + { + "epoch": 2.133190883190883, + "grad_norm": 0.7572789788246155, + "learning_rate": 8.970215112694953e-05, + "loss": 0.8016, + "step": 11982 + }, + { + "epoch": 2.1333689458689458, + "grad_norm": 0.8988085389137268, + "learning_rate": 8.968822817210354e-05, + "loss": 0.9307, + "step": 11983 + }, + { + "epoch": 2.1335470085470085, + "grad_norm": 0.7537818551063538, + "learning_rate": 8.967430541929532e-05, + "loss": 0.6423, + "step": 11984 + }, + { + "epoch": 2.1337250712250713, + "grad_norm": 0.7470884323120117, + "learning_rate": 8.966038286879763e-05, + "loss": 0.7753, + "step": 11985 + }, + { + "epoch": 2.133903133903134, + "grad_norm": 0.8670676946640015, + "learning_rate": 8.964646052088328e-05, + "loss": 1.0407, + "step": 11986 + }, + { + "epoch": 2.1340811965811968, + "grad_norm": 0.8322215676307678, + "learning_rate": 8.9632538375825e-05, + "loss": 0.6498, + "step": 11987 + }, + { + "epoch": 2.134259259259259, + "grad_norm": 0.7089048027992249, + "learning_rate": 8.961861643389562e-05, + "loss": 0.8778, + "step": 11988 + }, + { + "epoch": 2.134437321937322, + "grad_norm": 0.7980125546455383, + "learning_rate": 8.960469469536786e-05, + "loss": 0.7797, + "step": 11989 + }, + { + "epoch": 2.1346153846153846, + "grad_norm": 0.9979715943336487, + "learning_rate": 8.959077316051452e-05, + "loss": 0.7388, + "step": 11990 + }, + { + "epoch": 2.1347934472934473, + "grad_norm": 1.0040662288665771, + "learning_rate": 8.957685182960833e-05, + "loss": 0.954, + "step": 11991 + }, + { + "epoch": 2.13497150997151, + "grad_norm": 0.7885099053382874, + "learning_rate": 8.956293070292214e-05, + "loss": 0.9232, + "step": 11992 + }, + { + "epoch": 2.135149572649573, + "grad_norm": 0.7242771983146667, + "learning_rate": 8.954900978072859e-05, + "loss": 0.8614, + "step": 11993 + }, + { + "epoch": 2.135327635327635, + "grad_norm": 0.7970352172851562, + "learning_rate": 8.95350890633005e-05, + "loss": 0.8959, + "step": 11994 + }, + { + "epoch": 2.135505698005698, + "grad_norm": 0.8587128520011902, + "learning_rate": 8.952116855091059e-05, + "loss": 0.9981, + "step": 11995 + }, + { + "epoch": 2.1356837606837606, + "grad_norm": 0.8206220269203186, + "learning_rate": 8.950724824383164e-05, + "loss": 1.0271, + "step": 11996 + }, + { + "epoch": 2.1358618233618234, + "grad_norm": 0.8085001707077026, + "learning_rate": 8.949332814233635e-05, + "loss": 0.925, + "step": 11997 + }, + { + "epoch": 2.136039886039886, + "grad_norm": 0.8361417651176453, + "learning_rate": 8.947940824669748e-05, + "loss": 0.8744, + "step": 11998 + }, + { + "epoch": 2.136217948717949, + "grad_norm": 0.7548407316207886, + "learning_rate": 8.946548855718773e-05, + "loss": 0.7365, + "step": 11999 + }, + { + "epoch": 2.136396011396011, + "grad_norm": 0.8671223521232605, + "learning_rate": 8.945156907407983e-05, + "loss": 0.8958, + "step": 12000 + }, + { + "epoch": 2.136574074074074, + "grad_norm": 0.8007429838180542, + "learning_rate": 8.943764979764656e-05, + "loss": 0.955, + "step": 12001 + }, + { + "epoch": 2.1367521367521367, + "grad_norm": 0.7834315299987793, + "learning_rate": 8.942373072816057e-05, + "loss": 0.9226, + "step": 12002 + }, + { + "epoch": 2.1369301994301995, + "grad_norm": 0.896920919418335, + "learning_rate": 8.940981186589466e-05, + "loss": 0.8779, + "step": 12003 + }, + { + "epoch": 2.137108262108262, + "grad_norm": 0.7473411560058594, + "learning_rate": 8.939589321112143e-05, + "loss": 0.8993, + "step": 12004 + }, + { + "epoch": 2.137286324786325, + "grad_norm": 0.8071674704551697, + "learning_rate": 8.938197476411367e-05, + "loss": 0.9998, + "step": 12005 + }, + { + "epoch": 2.1374643874643873, + "grad_norm": 0.839290976524353, + "learning_rate": 8.936805652514404e-05, + "loss": 0.8311, + "step": 12006 + }, + { + "epoch": 2.13764245014245, + "grad_norm": 0.7217035293579102, + "learning_rate": 8.93541384944853e-05, + "loss": 0.8009, + "step": 12007 + }, + { + "epoch": 2.1378205128205128, + "grad_norm": 0.7392259836196899, + "learning_rate": 8.934022067241004e-05, + "loss": 0.9854, + "step": 12008 + }, + { + "epoch": 2.1379985754985755, + "grad_norm": 0.7470507621765137, + "learning_rate": 8.932630305919107e-05, + "loss": 0.8111, + "step": 12009 + }, + { + "epoch": 2.1381766381766383, + "grad_norm": 0.7988318204879761, + "learning_rate": 8.931238565510098e-05, + "loss": 0.8492, + "step": 12010 + }, + { + "epoch": 2.138354700854701, + "grad_norm": 0.9267526268959045, + "learning_rate": 8.929846846041251e-05, + "loss": 1.2238, + "step": 12011 + }, + { + "epoch": 2.1385327635327633, + "grad_norm": 0.8036465644836426, + "learning_rate": 8.92845514753983e-05, + "loss": 0.8837, + "step": 12012 + }, + { + "epoch": 2.138710826210826, + "grad_norm": 0.809256911277771, + "learning_rate": 8.927063470033109e-05, + "loss": 0.8836, + "step": 12013 + }, + { + "epoch": 2.138888888888889, + "grad_norm": 0.754692792892456, + "learning_rate": 8.925671813548345e-05, + "loss": 0.9469, + "step": 12014 + }, + { + "epoch": 2.1390669515669516, + "grad_norm": 0.9183036088943481, + "learning_rate": 8.924280178112814e-05, + "loss": 0.7654, + "step": 12015 + }, + { + "epoch": 2.1392450142450143, + "grad_norm": 0.82411128282547, + "learning_rate": 8.922888563753775e-05, + "loss": 0.9132, + "step": 12016 + }, + { + "epoch": 2.139423076923077, + "grad_norm": 0.8455918431282043, + "learning_rate": 8.9214969704985e-05, + "loss": 1.0041, + "step": 12017 + }, + { + "epoch": 2.1396011396011394, + "grad_norm": 0.9235896468162537, + "learning_rate": 8.92010539837425e-05, + "loss": 0.7842, + "step": 12018 + }, + { + "epoch": 2.139779202279202, + "grad_norm": 0.8965059518814087, + "learning_rate": 8.918713847408289e-05, + "loss": 1.0665, + "step": 12019 + }, + { + "epoch": 2.139957264957265, + "grad_norm": Infinity, + "learning_rate": 8.918713847408289e-05, + "loss": 1.0773, + "step": 12020 + }, + { + "epoch": 2.1401353276353277, + "grad_norm": 0.8859738707542419, + "learning_rate": 8.917322317627887e-05, + "loss": 0.9175, + "step": 12021 + }, + { + "epoch": 2.1403133903133904, + "grad_norm": 0.7828214764595032, + "learning_rate": 8.915930809060304e-05, + "loss": 0.8433, + "step": 12022 + }, + { + "epoch": 2.140491452991453, + "grad_norm": 0.7705734372138977, + "learning_rate": 8.914539321732808e-05, + "loss": 0.8696, + "step": 12023 + }, + { + "epoch": 2.1406695156695155, + "grad_norm": 0.7999989986419678, + "learning_rate": 8.913147855672655e-05, + "loss": 1.0531, + "step": 12024 + }, + { + "epoch": 2.140847578347578, + "grad_norm": 0.7210655212402344, + "learning_rate": 8.911756410907118e-05, + "loss": 0.6703, + "step": 12025 + }, + { + "epoch": 2.141025641025641, + "grad_norm": 0.7153459191322327, + "learning_rate": 8.910364987463447e-05, + "loss": 0.7166, + "step": 12026 + }, + { + "epoch": 2.1412037037037037, + "grad_norm": 0.771530032157898, + "learning_rate": 8.908973585368913e-05, + "loss": 0.6881, + "step": 12027 + }, + { + "epoch": 2.1413817663817665, + "grad_norm": 0.9988116025924683, + "learning_rate": 8.907582204650774e-05, + "loss": 0.8329, + "step": 12028 + }, + { + "epoch": 2.1415598290598292, + "grad_norm": 0.6992440819740295, + "learning_rate": 8.906190845336296e-05, + "loss": 0.6262, + "step": 12029 + }, + { + "epoch": 2.1417378917378915, + "grad_norm": 0.8061181902885437, + "learning_rate": 8.904799507452731e-05, + "loss": 0.8325, + "step": 12030 + }, + { + "epoch": 2.1419159544159543, + "grad_norm": 0.8372871279716492, + "learning_rate": 8.903408191027349e-05, + "loss": 0.8894, + "step": 12031 + }, + { + "epoch": 2.142094017094017, + "grad_norm": 0.803719162940979, + "learning_rate": 8.902016896087402e-05, + "loss": 0.9031, + "step": 12032 + }, + { + "epoch": 2.14227207977208, + "grad_norm": 0.8168890476226807, + "learning_rate": 8.900625622660158e-05, + "loss": 0.8174, + "step": 12033 + }, + { + "epoch": 2.1424501424501425, + "grad_norm": 0.8011388182640076, + "learning_rate": 8.899234370772865e-05, + "loss": 0.8267, + "step": 12034 + }, + { + "epoch": 2.1426282051282053, + "grad_norm": 0.8209220767021179, + "learning_rate": 8.897843140452795e-05, + "loss": 0.9303, + "step": 12035 + }, + { + "epoch": 2.142806267806268, + "grad_norm": 0.773525059223175, + "learning_rate": 8.896451931727192e-05, + "loss": 0.7037, + "step": 12036 + }, + { + "epoch": 2.1429843304843303, + "grad_norm": 0.7568892240524292, + "learning_rate": 8.895060744623324e-05, + "loss": 0.8568, + "step": 12037 + }, + { + "epoch": 2.143162393162393, + "grad_norm": 0.713636040687561, + "learning_rate": 8.893669579168444e-05, + "loss": 0.7838, + "step": 12038 + }, + { + "epoch": 2.143340455840456, + "grad_norm": 0.7462167739868164, + "learning_rate": 8.892278435389814e-05, + "loss": 0.6311, + "step": 12039 + }, + { + "epoch": 2.1435185185185186, + "grad_norm": 0.7164530158042908, + "learning_rate": 8.890887313314685e-05, + "loss": 1.0228, + "step": 12040 + }, + { + "epoch": 2.1436965811965814, + "grad_norm": 0.7540927529335022, + "learning_rate": 8.889496212970312e-05, + "loss": 0.8958, + "step": 12041 + }, + { + "epoch": 2.1438746438746437, + "grad_norm": 0.8119065761566162, + "learning_rate": 8.888105134383957e-05, + "loss": 0.8925, + "step": 12042 + }, + { + "epoch": 2.1440527065527064, + "grad_norm": 0.7905679941177368, + "learning_rate": 8.88671407758287e-05, + "loss": 0.7579, + "step": 12043 + }, + { + "epoch": 2.144230769230769, + "grad_norm": 0.8901177048683167, + "learning_rate": 8.885323042594312e-05, + "loss": 0.8849, + "step": 12044 + }, + { + "epoch": 2.144408831908832, + "grad_norm": 0.6958974599838257, + "learning_rate": 8.88393202944553e-05, + "loss": 0.8072, + "step": 12045 + }, + { + "epoch": 2.1445868945868947, + "grad_norm": 0.790036141872406, + "learning_rate": 8.882541038163786e-05, + "loss": 0.796, + "step": 12046 + }, + { + "epoch": 2.1447649572649574, + "grad_norm": 0.757655680179596, + "learning_rate": 8.881150068776324e-05, + "loss": 0.8094, + "step": 12047 + }, + { + "epoch": 2.14494301994302, + "grad_norm": 0.7525215148925781, + "learning_rate": 8.879759121310404e-05, + "loss": 0.6746, + "step": 12048 + }, + { + "epoch": 2.1451210826210825, + "grad_norm": 0.740566074848175, + "learning_rate": 8.878368195793276e-05, + "loss": 0.688, + "step": 12049 + }, + { + "epoch": 2.1452991452991452, + "grad_norm": 0.7771985530853271, + "learning_rate": 8.876977292252196e-05, + "loss": 1.0013, + "step": 12050 + }, + { + "epoch": 2.145477207977208, + "grad_norm": 0.8582369685173035, + "learning_rate": 8.875586410714409e-05, + "loss": 0.9185, + "step": 12051 + }, + { + "epoch": 2.1456552706552707, + "grad_norm": 0.7992526292800903, + "learning_rate": 8.874195551207174e-05, + "loss": 0.8388, + "step": 12052 + }, + { + "epoch": 2.1458333333333335, + "grad_norm": 0.795129120349884, + "learning_rate": 8.872804713757735e-05, + "loss": 0.88, + "step": 12053 + }, + { + "epoch": 2.146011396011396, + "grad_norm": 0.7467540502548218, + "learning_rate": 8.871413898393351e-05, + "loss": 0.8092, + "step": 12054 + }, + { + "epoch": 2.1461894586894585, + "grad_norm": 0.9468266367912292, + "learning_rate": 8.870023105141264e-05, + "loss": 0.7759, + "step": 12055 + }, + { + "epoch": 2.1463675213675213, + "grad_norm": 0.7893772721290588, + "learning_rate": 8.868632334028727e-05, + "loss": 0.8508, + "step": 12056 + }, + { + "epoch": 2.146545584045584, + "grad_norm": 0.6931375861167908, + "learning_rate": 8.867241585082988e-05, + "loss": 0.5013, + "step": 12057 + }, + { + "epoch": 2.146723646723647, + "grad_norm": 0.8978447318077087, + "learning_rate": 8.865850858331301e-05, + "loss": 0.9518, + "step": 12058 + }, + { + "epoch": 2.1469017094017095, + "grad_norm": 0.7293453812599182, + "learning_rate": 8.864460153800906e-05, + "loss": 0.731, + "step": 12059 + }, + { + "epoch": 2.1470797720797723, + "grad_norm": 0.8537824749946594, + "learning_rate": 8.863069471519056e-05, + "loss": 0.7935, + "step": 12060 + }, + { + "epoch": 2.1472578347578346, + "grad_norm": 0.6527614593505859, + "learning_rate": 8.861678811513002e-05, + "loss": 0.6579, + "step": 12061 + }, + { + "epoch": 2.1474358974358974, + "grad_norm": 0.9407904148101807, + "learning_rate": 8.860288173809983e-05, + "loss": 0.9057, + "step": 12062 + }, + { + "epoch": 2.14761396011396, + "grad_norm": 0.9314194321632385, + "learning_rate": 8.858897558437251e-05, + "loss": 1.0826, + "step": 12063 + }, + { + "epoch": 2.147792022792023, + "grad_norm": 0.7872337102890015, + "learning_rate": 8.85750696542205e-05, + "loss": 0.8919, + "step": 12064 + }, + { + "epoch": 2.1479700854700856, + "grad_norm": 0.8379341959953308, + "learning_rate": 8.85611639479163e-05, + "loss": 0.8193, + "step": 12065 + }, + { + "epoch": 2.148148148148148, + "grad_norm": 0.801295280456543, + "learning_rate": 8.85472584657323e-05, + "loss": 0.9305, + "step": 12066 + }, + { + "epoch": 2.1483262108262107, + "grad_norm": 0.7625086903572083, + "learning_rate": 8.853335320794098e-05, + "loss": 0.8379, + "step": 12067 + }, + { + "epoch": 2.1485042735042734, + "grad_norm": 0.8256231546401978, + "learning_rate": 8.851944817481478e-05, + "loss": 0.8901, + "step": 12068 + }, + { + "epoch": 2.148682336182336, + "grad_norm": 0.6940581202507019, + "learning_rate": 8.850554336662618e-05, + "loss": 0.6706, + "step": 12069 + }, + { + "epoch": 2.148860398860399, + "grad_norm": 0.910836398601532, + "learning_rate": 8.849163878364755e-05, + "loss": 0.9326, + "step": 12070 + }, + { + "epoch": 2.1490384615384617, + "grad_norm": 0.8550460934638977, + "learning_rate": 8.847773442615138e-05, + "loss": 0.8474, + "step": 12071 + }, + { + "epoch": 2.1492165242165244, + "grad_norm": 0.8178627490997314, + "learning_rate": 8.846383029441002e-05, + "loss": 0.8331, + "step": 12072 + }, + { + "epoch": 2.1493945868945867, + "grad_norm": 0.7606281638145447, + "learning_rate": 8.844992638869599e-05, + "loss": 0.6571, + "step": 12073 + }, + { + "epoch": 2.1495726495726495, + "grad_norm": 0.7166888117790222, + "learning_rate": 8.84360227092816e-05, + "loss": 0.8592, + "step": 12074 + }, + { + "epoch": 2.1497507122507122, + "grad_norm": 0.7688186764717102, + "learning_rate": 8.84221192564394e-05, + "loss": 0.691, + "step": 12075 + }, + { + "epoch": 2.149928774928775, + "grad_norm": 0.876740038394928, + "learning_rate": 8.840821603044166e-05, + "loss": 0.9962, + "step": 12076 + }, + { + "epoch": 2.1501068376068377, + "grad_norm": 0.7910363078117371, + "learning_rate": 8.839431303156087e-05, + "loss": 0.8061, + "step": 12077 + }, + { + "epoch": 2.1502849002849005, + "grad_norm": 0.6880493760108948, + "learning_rate": 8.83804102600694e-05, + "loss": 0.8078, + "step": 12078 + }, + { + "epoch": 2.150462962962963, + "grad_norm": 0.7795937061309814, + "learning_rate": 8.836650771623963e-05, + "loss": 0.8504, + "step": 12079 + }, + { + "epoch": 2.1506410256410255, + "grad_norm": 0.7761844992637634, + "learning_rate": 8.835260540034403e-05, + "loss": 0.7253, + "step": 12080 + }, + { + "epoch": 2.1508190883190883, + "grad_norm": 0.7070515751838684, + "learning_rate": 8.83387033126549e-05, + "loss": 0.7156, + "step": 12081 + }, + { + "epoch": 2.150997150997151, + "grad_norm": 0.7666274309158325, + "learning_rate": 8.832480145344467e-05, + "loss": 0.813, + "step": 12082 + }, + { + "epoch": 2.151175213675214, + "grad_norm": 0.9145975708961487, + "learning_rate": 8.831089982298568e-05, + "loss": 0.8889, + "step": 12083 + }, + { + "epoch": 2.1513532763532766, + "grad_norm": 0.7735843062400818, + "learning_rate": 8.829699842155035e-05, + "loss": 0.7152, + "step": 12084 + }, + { + "epoch": 2.151531339031339, + "grad_norm": 0.7625414729118347, + "learning_rate": 8.828309724941099e-05, + "loss": 0.8752, + "step": 12085 + }, + { + "epoch": 2.1517094017094016, + "grad_norm": 0.8874264359474182, + "learning_rate": 8.826919630684005e-05, + "loss": 0.8175, + "step": 12086 + }, + { + "epoch": 2.1518874643874644, + "grad_norm": 0.7425693273544312, + "learning_rate": 8.82552955941098e-05, + "loss": 0.7505, + "step": 12087 + }, + { + "epoch": 2.152065527065527, + "grad_norm": 0.7098270058631897, + "learning_rate": 8.824139511149265e-05, + "loss": 0.7129, + "step": 12088 + }, + { + "epoch": 2.15224358974359, + "grad_norm": 0.8470510840415955, + "learning_rate": 8.822749485926092e-05, + "loss": 0.9656, + "step": 12089 + }, + { + "epoch": 2.1524216524216526, + "grad_norm": 0.7690402865409851, + "learning_rate": 8.8213594837687e-05, + "loss": 0.9436, + "step": 12090 + }, + { + "epoch": 2.152599715099715, + "grad_norm": 0.77431321144104, + "learning_rate": 8.819969504704318e-05, + "loss": 0.9912, + "step": 12091 + }, + { + "epoch": 2.1527777777777777, + "grad_norm": 0.7590892314910889, + "learning_rate": 8.818579548760184e-05, + "loss": 0.7412, + "step": 12092 + }, + { + "epoch": 2.1529558404558404, + "grad_norm": 0.870966374874115, + "learning_rate": 8.817189615963528e-05, + "loss": 1.0248, + "step": 12093 + }, + { + "epoch": 2.153133903133903, + "grad_norm": 0.7989356517791748, + "learning_rate": 8.815799706341587e-05, + "loss": 0.7104, + "step": 12094 + }, + { + "epoch": 2.153311965811966, + "grad_norm": 0.9615582227706909, + "learning_rate": 8.814409819921589e-05, + "loss": 1.0191, + "step": 12095 + }, + { + "epoch": 2.1534900284900287, + "grad_norm": 0.7063159346580505, + "learning_rate": 8.81301995673077e-05, + "loss": 0.8209, + "step": 12096 + }, + { + "epoch": 2.153668091168091, + "grad_norm": 0.8179874420166016, + "learning_rate": 8.811630116796356e-05, + "loss": 0.9457, + "step": 12097 + }, + { + "epoch": 2.1538461538461537, + "grad_norm": 0.7227353453636169, + "learning_rate": 8.810240300145582e-05, + "loss": 0.8112, + "step": 12098 + }, + { + "epoch": 2.1540242165242165, + "grad_norm": 0.7480359077453613, + "learning_rate": 8.808850506805677e-05, + "loss": 0.6293, + "step": 12099 + }, + { + "epoch": 2.1542022792022792, + "grad_norm": 0.7610893845558167, + "learning_rate": 8.807460736803871e-05, + "loss": 0.911, + "step": 12100 + }, + { + "epoch": 2.154380341880342, + "grad_norm": 0.774640679359436, + "learning_rate": 8.806070990167399e-05, + "loss": 0.8144, + "step": 12101 + }, + { + "epoch": 2.1545584045584047, + "grad_norm": 0.7785552144050598, + "learning_rate": 8.804681266923482e-05, + "loss": 0.8841, + "step": 12102 + }, + { + "epoch": 2.154736467236467, + "grad_norm": 0.843715488910675, + "learning_rate": 8.803291567099354e-05, + "loss": 0.9056, + "step": 12103 + }, + { + "epoch": 2.15491452991453, + "grad_norm": 0.7996346354484558, + "learning_rate": 8.801901890722241e-05, + "loss": 0.8916, + "step": 12104 + }, + { + "epoch": 2.1550925925925926, + "grad_norm": 0.9159125685691833, + "learning_rate": 8.800512237819376e-05, + "loss": 0.8225, + "step": 12105 + }, + { + "epoch": 2.1552706552706553, + "grad_norm": 0.8341643810272217, + "learning_rate": 8.799122608417976e-05, + "loss": 0.8702, + "step": 12106 + }, + { + "epoch": 2.155448717948718, + "grad_norm": 0.8075932264328003, + "learning_rate": 8.797733002545278e-05, + "loss": 0.9167, + "step": 12107 + }, + { + "epoch": 2.155626780626781, + "grad_norm": 0.8370183706283569, + "learning_rate": 8.7963434202285e-05, + "loss": 0.9213, + "step": 12108 + }, + { + "epoch": 2.155804843304843, + "grad_norm": 0.7500374913215637, + "learning_rate": 8.794953861494877e-05, + "loss": 0.7702, + "step": 12109 + }, + { + "epoch": 2.155982905982906, + "grad_norm": 0.7347766160964966, + "learning_rate": 8.793564326371626e-05, + "loss": 0.7057, + "step": 12110 + }, + { + "epoch": 2.1561609686609686, + "grad_norm": 0.754917562007904, + "learning_rate": 8.79217481488598e-05, + "loss": 0.8725, + "step": 12111 + }, + { + "epoch": 2.1563390313390314, + "grad_norm": 0.6942774057388306, + "learning_rate": 8.790785327065155e-05, + "loss": 0.85, + "step": 12112 + }, + { + "epoch": 2.156517094017094, + "grad_norm": 0.8082157969474792, + "learning_rate": 8.789395862936383e-05, + "loss": 1.1462, + "step": 12113 + }, + { + "epoch": 2.156695156695157, + "grad_norm": 0.898435652256012, + "learning_rate": 8.788006422526881e-05, + "loss": 0.8044, + "step": 12114 + }, + { + "epoch": 2.156873219373219, + "grad_norm": 0.9474737048149109, + "learning_rate": 8.786617005863879e-05, + "loss": 0.9089, + "step": 12115 + }, + { + "epoch": 2.157051282051282, + "grad_norm": 0.7898718118667603, + "learning_rate": 8.785227612974594e-05, + "loss": 0.9758, + "step": 12116 + }, + { + "epoch": 2.1572293447293447, + "grad_norm": 0.6734052300453186, + "learning_rate": 8.783838243886253e-05, + "loss": 0.7835, + "step": 12117 + }, + { + "epoch": 2.1574074074074074, + "grad_norm": 0.9381069540977478, + "learning_rate": 8.782448898626072e-05, + "loss": 0.7666, + "step": 12118 + }, + { + "epoch": 2.15758547008547, + "grad_norm": 0.8677506446838379, + "learning_rate": 8.781059577221276e-05, + "loss": 0.7442, + "step": 12119 + }, + { + "epoch": 2.157763532763533, + "grad_norm": 0.8244445323944092, + "learning_rate": 8.779670279699086e-05, + "loss": 0.8104, + "step": 12120 + }, + { + "epoch": 2.1579415954415953, + "grad_norm": 0.7984805703163147, + "learning_rate": 8.77828100608672e-05, + "loss": 1.029, + "step": 12121 + }, + { + "epoch": 2.158119658119658, + "grad_norm": 0.7817366123199463, + "learning_rate": 8.776891756411405e-05, + "loss": 0.6797, + "step": 12122 + }, + { + "epoch": 2.1582977207977208, + "grad_norm": 0.7084082365036011, + "learning_rate": 8.77550253070035e-05, + "loss": 0.688, + "step": 12123 + }, + { + "epoch": 2.1584757834757835, + "grad_norm": 0.7659782767295837, + "learning_rate": 8.774113328980782e-05, + "loss": 0.8691, + "step": 12124 + }, + { + "epoch": 2.1586538461538463, + "grad_norm": 0.7010130286216736, + "learning_rate": 8.772724151279913e-05, + "loss": 0.7587, + "step": 12125 + }, + { + "epoch": 2.158831908831909, + "grad_norm": 0.8183525800704956, + "learning_rate": 8.771334997624973e-05, + "loss": 0.8696, + "step": 12126 + }, + { + "epoch": 2.1590099715099713, + "grad_norm": 0.7944908142089844, + "learning_rate": 8.769945868043164e-05, + "loss": 0.8625, + "step": 12127 + }, + { + "epoch": 2.159188034188034, + "grad_norm": 0.7710323333740234, + "learning_rate": 8.768556762561713e-05, + "loss": 0.7765, + "step": 12128 + }, + { + "epoch": 2.159366096866097, + "grad_norm": 0.7416872382164001, + "learning_rate": 8.767167681207833e-05, + "loss": 0.9151, + "step": 12129 + }, + { + "epoch": 2.1595441595441596, + "grad_norm": 0.9230012893676758, + "learning_rate": 8.765778624008744e-05, + "loss": 0.914, + "step": 12130 + }, + { + "epoch": 2.1597222222222223, + "grad_norm": 0.7468557357788086, + "learning_rate": 8.764389590991657e-05, + "loss": 0.8624, + "step": 12131 + }, + { + "epoch": 2.159900284900285, + "grad_norm": 0.7746220827102661, + "learning_rate": 8.763000582183791e-05, + "loss": 0.9683, + "step": 12132 + }, + { + "epoch": 2.1600783475783474, + "grad_norm": 0.8429577350616455, + "learning_rate": 8.761611597612356e-05, + "loss": 0.8808, + "step": 12133 + }, + { + "epoch": 2.16025641025641, + "grad_norm": 0.8117298483848572, + "learning_rate": 8.760222637304572e-05, + "loss": 0.7067, + "step": 12134 + }, + { + "epoch": 2.160434472934473, + "grad_norm": 0.7717329859733582, + "learning_rate": 8.758833701287647e-05, + "loss": 1.0001, + "step": 12135 + }, + { + "epoch": 2.1606125356125356, + "grad_norm": 0.8493856191635132, + "learning_rate": 8.7574447895888e-05, + "loss": 0.9883, + "step": 12136 + }, + { + "epoch": 2.1607905982905984, + "grad_norm": 0.8592587113380432, + "learning_rate": 8.75605590223524e-05, + "loss": 0.7437, + "step": 12137 + }, + { + "epoch": 2.160968660968661, + "grad_norm": 0.6487032771110535, + "learning_rate": 8.75466703925418e-05, + "loss": 0.6841, + "step": 12138 + }, + { + "epoch": 2.1611467236467234, + "grad_norm": 0.8449310660362244, + "learning_rate": 8.753278200672832e-05, + "loss": 0.9221, + "step": 12139 + }, + { + "epoch": 2.161324786324786, + "grad_norm": 0.9603136777877808, + "learning_rate": 8.751889386518407e-05, + "loss": 0.8664, + "step": 12140 + }, + { + "epoch": 2.161502849002849, + "grad_norm": 0.7288493514060974, + "learning_rate": 8.750500596818121e-05, + "loss": 0.5745, + "step": 12141 + }, + { + "epoch": 2.1616809116809117, + "grad_norm": 0.8626441955566406, + "learning_rate": 8.749111831599178e-05, + "loss": 0.8346, + "step": 12142 + }, + { + "epoch": 2.1618589743589745, + "grad_norm": 0.7634188532829285, + "learning_rate": 8.74772309088879e-05, + "loss": 0.7503, + "step": 12143 + }, + { + "epoch": 2.162037037037037, + "grad_norm": 0.8641456365585327, + "learning_rate": 8.746334374714167e-05, + "loss": 0.9033, + "step": 12144 + }, + { + "epoch": 2.1622150997150995, + "grad_norm": 0.8103315234184265, + "learning_rate": 8.744945683102517e-05, + "loss": 0.8181, + "step": 12145 + }, + { + "epoch": 2.1623931623931623, + "grad_norm": 1.2493078708648682, + "learning_rate": 8.743557016081047e-05, + "loss": 1.0308, + "step": 12146 + }, + { + "epoch": 2.162571225071225, + "grad_norm": 0.6447771191596985, + "learning_rate": 8.742168373676973e-05, + "loss": 0.6886, + "step": 12147 + }, + { + "epoch": 2.1627492877492878, + "grad_norm": 0.90229332447052, + "learning_rate": 8.740779755917492e-05, + "loss": 1.0361, + "step": 12148 + }, + { + "epoch": 2.1629273504273505, + "grad_norm": 0.7414017915725708, + "learning_rate": 8.739391162829818e-05, + "loss": 0.781, + "step": 12149 + }, + { + "epoch": 2.1631054131054133, + "grad_norm": 0.8897294998168945, + "learning_rate": 8.738002594441154e-05, + "loss": 0.7712, + "step": 12150 + }, + { + "epoch": 2.1632834757834756, + "grad_norm": 0.8515656590461731, + "learning_rate": 8.73661405077871e-05, + "loss": 0.8843, + "step": 12151 + }, + { + "epoch": 2.1634615384615383, + "grad_norm": 0.7901699542999268, + "learning_rate": 8.735225531869686e-05, + "loss": 0.8588, + "step": 12152 + }, + { + "epoch": 2.163639601139601, + "grad_norm": 0.7262305021286011, + "learning_rate": 8.733837037741295e-05, + "loss": 0.9257, + "step": 12153 + }, + { + "epoch": 2.163817663817664, + "grad_norm": 1.1076871156692505, + "learning_rate": 8.732448568420732e-05, + "loss": 0.9817, + "step": 12154 + }, + { + "epoch": 2.1639957264957266, + "grad_norm": 0.8384785652160645, + "learning_rate": 8.731060123935209e-05, + "loss": 0.8024, + "step": 12155 + }, + { + "epoch": 2.1641737891737893, + "grad_norm": 0.8376259803771973, + "learning_rate": 8.729671704311924e-05, + "loss": 1.0299, + "step": 12156 + }, + { + "epoch": 2.164351851851852, + "grad_norm": 0.8248558044433594, + "learning_rate": 8.728283309578089e-05, + "loss": 0.9557, + "step": 12157 + }, + { + "epoch": 2.1645299145299144, + "grad_norm": 0.7452875375747681, + "learning_rate": 8.726894939760894e-05, + "loss": 0.8267, + "step": 12158 + }, + { + "epoch": 2.164707977207977, + "grad_norm": 0.8329267501831055, + "learning_rate": 8.72550659488755e-05, + "loss": 0.9288, + "step": 12159 + }, + { + "epoch": 2.16488603988604, + "grad_norm": 0.8748268485069275, + "learning_rate": 8.724118274985259e-05, + "loss": 0.8663, + "step": 12160 + }, + { + "epoch": 2.1650641025641026, + "grad_norm": 0.6839116811752319, + "learning_rate": 8.722729980081217e-05, + "loss": 0.6067, + "step": 12161 + }, + { + "epoch": 2.1652421652421654, + "grad_norm": 0.8343674540519714, + "learning_rate": 8.721341710202632e-05, + "loss": 0.9611, + "step": 12162 + }, + { + "epoch": 2.1654202279202277, + "grad_norm": 0.7783843874931335, + "learning_rate": 8.719953465376695e-05, + "loss": 0.8921, + "step": 12163 + }, + { + "epoch": 2.1655982905982905, + "grad_norm": 0.8357030749320984, + "learning_rate": 8.718565245630615e-05, + "loss": 0.9189, + "step": 12164 + }, + { + "epoch": 2.165776353276353, + "grad_norm": 0.8150131702423096, + "learning_rate": 8.717177050991582e-05, + "loss": 0.7486, + "step": 12165 + }, + { + "epoch": 2.165954415954416, + "grad_norm": 0.7282506823539734, + "learning_rate": 8.715788881486807e-05, + "loss": 0.8894, + "step": 12166 + }, + { + "epoch": 2.1661324786324787, + "grad_norm": 0.9958226680755615, + "learning_rate": 8.714400737143475e-05, + "loss": 1.0359, + "step": 12167 + }, + { + "epoch": 2.1663105413105415, + "grad_norm": 0.7162553071975708, + "learning_rate": 8.713012617988796e-05, + "loss": 0.7728, + "step": 12168 + }, + { + "epoch": 2.166488603988604, + "grad_norm": 0.6364821791648865, + "learning_rate": 8.711624524049955e-05, + "loss": 0.5881, + "step": 12169 + }, + { + "epoch": 2.1666666666666665, + "grad_norm": 0.9431148767471313, + "learning_rate": 8.710236455354159e-05, + "loss": 0.8804, + "step": 12170 + }, + { + "epoch": 2.1668447293447293, + "grad_norm": 0.7328855395317078, + "learning_rate": 8.708848411928598e-05, + "loss": 0.7762, + "step": 12171 + }, + { + "epoch": 2.167022792022792, + "grad_norm": 0.7855633497238159, + "learning_rate": 8.707460393800472e-05, + "loss": 0.7687, + "step": 12172 + }, + { + "epoch": 2.1672008547008548, + "grad_norm": 0.8694273233413696, + "learning_rate": 8.706072400996973e-05, + "loss": 0.7153, + "step": 12173 + }, + { + "epoch": 2.1673789173789175, + "grad_norm": 0.7371255159378052, + "learning_rate": 8.704684433545299e-05, + "loss": 0.7901, + "step": 12174 + }, + { + "epoch": 2.16755698005698, + "grad_norm": 0.7719849944114685, + "learning_rate": 8.70329649147264e-05, + "loss": 0.7569, + "step": 12175 + }, + { + "epoch": 2.1677350427350426, + "grad_norm": 0.883618175983429, + "learning_rate": 8.701908574806197e-05, + "loss": 0.8941, + "step": 12176 + }, + { + "epoch": 2.1679131054131053, + "grad_norm": 0.9455791711807251, + "learning_rate": 8.700520683573155e-05, + "loss": 0.8596, + "step": 12177 + }, + { + "epoch": 2.168091168091168, + "grad_norm": 0.7487229108810425, + "learning_rate": 8.69913281780071e-05, + "loss": 0.7353, + "step": 12178 + }, + { + "epoch": 2.168269230769231, + "grad_norm": 0.8050364255905151, + "learning_rate": 8.697744977516062e-05, + "loss": 0.8564, + "step": 12179 + }, + { + "epoch": 2.1684472934472936, + "grad_norm": 0.759355902671814, + "learning_rate": 8.69635716274639e-05, + "loss": 0.7128, + "step": 12180 + }, + { + "epoch": 2.1686253561253563, + "grad_norm": 0.8730760216712952, + "learning_rate": 8.694969373518892e-05, + "loss": 0.9944, + "step": 12181 + }, + { + "epoch": 2.1688034188034186, + "grad_norm": 0.7761130332946777, + "learning_rate": 8.693581609860756e-05, + "loss": 0.6845, + "step": 12182 + }, + { + "epoch": 2.1689814814814814, + "grad_norm": 0.8118788003921509, + "learning_rate": 8.692193871799181e-05, + "loss": 0.798, + "step": 12183 + }, + { + "epoch": 2.169159544159544, + "grad_norm": 0.8340219855308533, + "learning_rate": 8.690806159361344e-05, + "loss": 0.9754, + "step": 12184 + }, + { + "epoch": 2.169337606837607, + "grad_norm": 0.7515831589698792, + "learning_rate": 8.689418472574444e-05, + "loss": 0.998, + "step": 12185 + }, + { + "epoch": 2.1695156695156697, + "grad_norm": 0.7781083583831787, + "learning_rate": 8.688030811465665e-05, + "loss": 1.0152, + "step": 12186 + }, + { + "epoch": 2.169693732193732, + "grad_norm": 0.775097131729126, + "learning_rate": 8.6866431760622e-05, + "loss": 0.808, + "step": 12187 + }, + { + "epoch": 2.1698717948717947, + "grad_norm": 0.8288158178329468, + "learning_rate": 8.68525556639123e-05, + "loss": 0.9172, + "step": 12188 + }, + { + "epoch": 2.1700498575498575, + "grad_norm": 0.7754917740821838, + "learning_rate": 8.68386798247995e-05, + "loss": 0.719, + "step": 12189 + }, + { + "epoch": 2.17022792022792, + "grad_norm": 0.786685585975647, + "learning_rate": 8.682480424355539e-05, + "loss": 0.8883, + "step": 12190 + }, + { + "epoch": 2.170405982905983, + "grad_norm": 0.9970952272415161, + "learning_rate": 8.681092892045189e-05, + "loss": 0.9258, + "step": 12191 + }, + { + "epoch": 2.1705840455840457, + "grad_norm": 0.9732664227485657, + "learning_rate": 8.679705385576082e-05, + "loss": 0.8916, + "step": 12192 + }, + { + "epoch": 2.1707621082621085, + "grad_norm": 0.8557142615318298, + "learning_rate": 8.67831790497541e-05, + "loss": 0.8908, + "step": 12193 + }, + { + "epoch": 2.1709401709401708, + "grad_norm": 0.8564930558204651, + "learning_rate": 8.676930450270347e-05, + "loss": 1.054, + "step": 12194 + }, + { + "epoch": 2.1711182336182335, + "grad_norm": 0.785732090473175, + "learning_rate": 8.675543021488087e-05, + "loss": 0.7459, + "step": 12195 + }, + { + "epoch": 2.1712962962962963, + "grad_norm": 0.8739910125732422, + "learning_rate": 8.674155618655809e-05, + "loss": 0.8464, + "step": 12196 + }, + { + "epoch": 2.171474358974359, + "grad_norm": 0.8624834418296814, + "learning_rate": 8.672768241800699e-05, + "loss": 0.9405, + "step": 12197 + }, + { + "epoch": 2.171652421652422, + "grad_norm": 0.7948583364486694, + "learning_rate": 8.671380890949936e-05, + "loss": 1.0271, + "step": 12198 + }, + { + "epoch": 2.1718304843304845, + "grad_norm": 0.8078029155731201, + "learning_rate": 8.669993566130704e-05, + "loss": 0.6845, + "step": 12199 + }, + { + "epoch": 2.172008547008547, + "grad_norm": 0.7599586844444275, + "learning_rate": 8.668606267370187e-05, + "loss": 0.8438, + "step": 12200 + }, + { + "epoch": 2.1721866096866096, + "grad_norm": 0.8085161447525024, + "learning_rate": 8.667218994695562e-05, + "loss": 0.8398, + "step": 12201 + }, + { + "epoch": 2.1723646723646723, + "grad_norm": 0.9033090472221375, + "learning_rate": 8.665831748134019e-05, + "loss": 0.8591, + "step": 12202 + }, + { + "epoch": 2.172542735042735, + "grad_norm": 0.8638277649879456, + "learning_rate": 8.664444527712726e-05, + "loss": 0.8276, + "step": 12203 + }, + { + "epoch": 2.172720797720798, + "grad_norm": 0.8230745792388916, + "learning_rate": 8.663057333458871e-05, + "loss": 0.8663, + "step": 12204 + }, + { + "epoch": 2.1728988603988606, + "grad_norm": 0.8588439226150513, + "learning_rate": 8.661670165399626e-05, + "loss": 0.8543, + "step": 12205 + }, + { + "epoch": 2.173076923076923, + "grad_norm": 0.8461976051330566, + "learning_rate": 8.660283023562177e-05, + "loss": 0.8973, + "step": 12206 + }, + { + "epoch": 2.1732549857549857, + "grad_norm": 0.7768828868865967, + "learning_rate": 8.658895907973697e-05, + "loss": 0.7285, + "step": 12207 + }, + { + "epoch": 2.1734330484330484, + "grad_norm": 0.7384130954742432, + "learning_rate": 8.65750881866137e-05, + "loss": 0.7654, + "step": 12208 + }, + { + "epoch": 2.173611111111111, + "grad_norm": 0.8700957298278809, + "learning_rate": 8.656121755652365e-05, + "loss": 1.067, + "step": 12209 + }, + { + "epoch": 2.173789173789174, + "grad_norm": 0.8067826628684998, + "learning_rate": 8.654734718973863e-05, + "loss": 0.9863, + "step": 12210 + }, + { + "epoch": 2.1739672364672367, + "grad_norm": 0.7515989542007446, + "learning_rate": 8.653347708653039e-05, + "loss": 0.8434, + "step": 12211 + }, + { + "epoch": 2.174145299145299, + "grad_norm": 0.8280966877937317, + "learning_rate": 8.651960724717072e-05, + "loss": 1.0065, + "step": 12212 + }, + { + "epoch": 2.1743233618233617, + "grad_norm": 0.7988734841346741, + "learning_rate": 8.650573767193132e-05, + "loss": 0.7892, + "step": 12213 + }, + { + "epoch": 2.1745014245014245, + "grad_norm": 0.785323977470398, + "learning_rate": 8.649186836108399e-05, + "loss": 0.7563, + "step": 12214 + }, + { + "epoch": 2.1746794871794872, + "grad_norm": 0.7884892821311951, + "learning_rate": 8.64779993149004e-05, + "loss": 0.7225, + "step": 12215 + }, + { + "epoch": 2.17485754985755, + "grad_norm": 0.9376154541969299, + "learning_rate": 8.646413053365235e-05, + "loss": 0.8103, + "step": 12216 + }, + { + "epoch": 2.1750356125356127, + "grad_norm": 0.9680297374725342, + "learning_rate": 8.64502620176115e-05, + "loss": 0.8924, + "step": 12217 + }, + { + "epoch": 2.175213675213675, + "grad_norm": 1.016848087310791, + "learning_rate": 8.643639376704964e-05, + "loss": 0.9017, + "step": 12218 + }, + { + "epoch": 2.175391737891738, + "grad_norm": 0.790868878364563, + "learning_rate": 8.64225257822385e-05, + "loss": 0.7597, + "step": 12219 + }, + { + "epoch": 2.1755698005698005, + "grad_norm": 0.7539415955543518, + "learning_rate": 8.640865806344974e-05, + "loss": 0.7801, + "step": 12220 + }, + { + "epoch": 2.1757478632478633, + "grad_norm": 0.7288404703140259, + "learning_rate": 8.63947906109551e-05, + "loss": 0.6753, + "step": 12221 + }, + { + "epoch": 2.175925925925926, + "grad_norm": 0.8449869155883789, + "learning_rate": 8.638092342502623e-05, + "loss": 0.7165, + "step": 12222 + }, + { + "epoch": 2.176103988603989, + "grad_norm": 0.8210735321044922, + "learning_rate": 8.636705650593495e-05, + "loss": 0.8677, + "step": 12223 + }, + { + "epoch": 2.176282051282051, + "grad_norm": 0.7431774735450745, + "learning_rate": 8.635318985395284e-05, + "loss": 0.7914, + "step": 12224 + }, + { + "epoch": 2.176460113960114, + "grad_norm": 0.8223997354507446, + "learning_rate": 8.633932346935165e-05, + "loss": 0.7243, + "step": 12225 + }, + { + "epoch": 2.1766381766381766, + "grad_norm": 1.0101778507232666, + "learning_rate": 8.632545735240299e-05, + "loss": 0.8608, + "step": 12226 + }, + { + "epoch": 2.1768162393162394, + "grad_norm": 0.7270255088806152, + "learning_rate": 8.631159150337862e-05, + "loss": 0.8699, + "step": 12227 + }, + { + "epoch": 2.176994301994302, + "grad_norm": 0.8687323331832886, + "learning_rate": 8.629772592255016e-05, + "loss": 0.9168, + "step": 12228 + }, + { + "epoch": 2.177172364672365, + "grad_norm": 0.7623698115348816, + "learning_rate": 8.628386061018934e-05, + "loss": 0.9012, + "step": 12229 + }, + { + "epoch": 2.177350427350427, + "grad_norm": 0.7458708882331848, + "learning_rate": 8.626999556656771e-05, + "loss": 0.9068, + "step": 12230 + }, + { + "epoch": 2.17752849002849, + "grad_norm": 0.8262876868247986, + "learning_rate": 8.625613079195704e-05, + "loss": 0.9425, + "step": 12231 + }, + { + "epoch": 2.1777065527065527, + "grad_norm": 0.8737035989761353, + "learning_rate": 8.624226628662893e-05, + "loss": 0.9943, + "step": 12232 + }, + { + "epoch": 2.1778846153846154, + "grad_norm": 0.8250965476036072, + "learning_rate": 8.622840205085505e-05, + "loss": 0.9237, + "step": 12233 + }, + { + "epoch": 2.178062678062678, + "grad_norm": 0.8689019680023193, + "learning_rate": 8.621453808490699e-05, + "loss": 0.8544, + "step": 12234 + }, + { + "epoch": 2.178240740740741, + "grad_norm": 0.8672708868980408, + "learning_rate": 8.620067438905643e-05, + "loss": 0.7623, + "step": 12235 + }, + { + "epoch": 2.1784188034188032, + "grad_norm": 0.7077436447143555, + "learning_rate": 8.6186810963575e-05, + "loss": 0.7468, + "step": 12236 + }, + { + "epoch": 2.178596866096866, + "grad_norm": 0.838474452495575, + "learning_rate": 8.617294780873433e-05, + "loss": 0.9207, + "step": 12237 + }, + { + "epoch": 2.1787749287749287, + "grad_norm": 0.7240039110183716, + "learning_rate": 8.615908492480598e-05, + "loss": 0.8981, + "step": 12238 + }, + { + "epoch": 2.1789529914529915, + "grad_norm": 0.6995998620986938, + "learning_rate": 8.614522231206162e-05, + "loss": 0.7131, + "step": 12239 + }, + { + "epoch": 2.1791310541310542, + "grad_norm": 0.7011054158210754, + "learning_rate": 8.613135997077288e-05, + "loss": 0.8138, + "step": 12240 + }, + { + "epoch": 2.179309116809117, + "grad_norm": 0.9815019369125366, + "learning_rate": 8.611749790121131e-05, + "loss": 0.9637, + "step": 12241 + }, + { + "epoch": 2.1794871794871793, + "grad_norm": 0.7523870468139648, + "learning_rate": 8.610363610364853e-05, + "loss": 0.7555, + "step": 12242 + }, + { + "epoch": 2.179665242165242, + "grad_norm": 0.8286668658256531, + "learning_rate": 8.608977457835612e-05, + "loss": 0.7911, + "step": 12243 + }, + { + "epoch": 2.179843304843305, + "grad_norm": 0.8183441758155823, + "learning_rate": 8.607591332560573e-05, + "loss": 0.793, + "step": 12244 + }, + { + "epoch": 2.1800213675213675, + "grad_norm": 0.7104299068450928, + "learning_rate": 8.606205234566885e-05, + "loss": 0.6856, + "step": 12245 + }, + { + "epoch": 2.1801994301994303, + "grad_norm": 0.871588945388794, + "learning_rate": 8.60481916388171e-05, + "loss": 0.8342, + "step": 12246 + }, + { + "epoch": 2.180377492877493, + "grad_norm": 0.8002356290817261, + "learning_rate": 8.603433120532206e-05, + "loss": 0.9451, + "step": 12247 + }, + { + "epoch": 2.1805555555555554, + "grad_norm": 0.8223865032196045, + "learning_rate": 8.602047104545532e-05, + "loss": 0.8446, + "step": 12248 + }, + { + "epoch": 2.180733618233618, + "grad_norm": 0.7381762266159058, + "learning_rate": 8.600661115948836e-05, + "loss": 0.7575, + "step": 12249 + }, + { + "epoch": 2.180911680911681, + "grad_norm": 0.8717563152313232, + "learning_rate": 8.599275154769284e-05, + "loss": 0.9615, + "step": 12250 + }, + { + "epoch": 2.1810897435897436, + "grad_norm": 0.7935179471969604, + "learning_rate": 8.597889221034022e-05, + "loss": 0.8603, + "step": 12251 + }, + { + "epoch": 2.1812678062678064, + "grad_norm": 0.6350329518318176, + "learning_rate": 8.596503314770208e-05, + "loss": 0.4981, + "step": 12252 + }, + { + "epoch": 2.181445868945869, + "grad_norm": 0.8739648461341858, + "learning_rate": 8.595117436004995e-05, + "loss": 0.7872, + "step": 12253 + }, + { + "epoch": 2.1816239316239314, + "grad_norm": 0.8199412822723389, + "learning_rate": 8.593731584765542e-05, + "loss": 0.7905, + "step": 12254 + }, + { + "epoch": 2.181801994301994, + "grad_norm": 0.7289649844169617, + "learning_rate": 8.592345761078993e-05, + "loss": 0.6981, + "step": 12255 + }, + { + "epoch": 2.181980056980057, + "grad_norm": 0.8234626650810242, + "learning_rate": 8.590959964972506e-05, + "loss": 1.0442, + "step": 12256 + }, + { + "epoch": 2.1821581196581197, + "grad_norm": 0.7804498076438904, + "learning_rate": 8.589574196473229e-05, + "loss": 0.9386, + "step": 12257 + }, + { + "epoch": 2.1823361823361824, + "grad_norm": 0.9459218382835388, + "learning_rate": 8.588188455608317e-05, + "loss": 0.8075, + "step": 12258 + }, + { + "epoch": 2.182514245014245, + "grad_norm": 0.8133191466331482, + "learning_rate": 8.586802742404924e-05, + "loss": 1.0275, + "step": 12259 + }, + { + "epoch": 2.1826923076923075, + "grad_norm": 0.8302663564682007, + "learning_rate": 8.58541705689019e-05, + "loss": 0.9887, + "step": 12260 + }, + { + "epoch": 2.1828703703703702, + "grad_norm": 0.7839202284812927, + "learning_rate": 8.584031399091274e-05, + "loss": 1.0256, + "step": 12261 + }, + { + "epoch": 2.183048433048433, + "grad_norm": 0.8050578236579895, + "learning_rate": 8.582645769035319e-05, + "loss": 0.843, + "step": 12262 + }, + { + "epoch": 2.1832264957264957, + "grad_norm": 0.749110221862793, + "learning_rate": 8.581260166749477e-05, + "loss": 0.7683, + "step": 12263 + }, + { + "epoch": 2.1834045584045585, + "grad_norm": 0.7982701659202576, + "learning_rate": 8.579874592260894e-05, + "loss": 0.942, + "step": 12264 + }, + { + "epoch": 2.1835826210826212, + "grad_norm": 0.7571866512298584, + "learning_rate": 8.57848904559672e-05, + "loss": 0.8828, + "step": 12265 + }, + { + "epoch": 2.183760683760684, + "grad_norm": 0.7445113658905029, + "learning_rate": 8.577103526784098e-05, + "loss": 0.8869, + "step": 12266 + }, + { + "epoch": 2.1839387464387463, + "grad_norm": 0.7999380230903625, + "learning_rate": 8.575718035850177e-05, + "loss": 0.9476, + "step": 12267 + }, + { + "epoch": 2.184116809116809, + "grad_norm": 0.7188777923583984, + "learning_rate": 8.574332572822103e-05, + "loss": 0.7961, + "step": 12268 + }, + { + "epoch": 2.184294871794872, + "grad_norm": 0.9545742869377136, + "learning_rate": 8.572947137727023e-05, + "loss": 0.8629, + "step": 12269 + }, + { + "epoch": 2.1844729344729346, + "grad_norm": 0.8066838979721069, + "learning_rate": 8.571561730592075e-05, + "loss": 0.7728, + "step": 12270 + }, + { + "epoch": 2.1846509971509973, + "grad_norm": 0.7819525003433228, + "learning_rate": 8.57017635144441e-05, + "loss": 0.9897, + "step": 12271 + }, + { + "epoch": 2.1848290598290596, + "grad_norm": 0.9249349236488342, + "learning_rate": 8.568791000311166e-05, + "loss": 0.8562, + "step": 12272 + }, + { + "epoch": 2.1850071225071224, + "grad_norm": 0.8118993043899536, + "learning_rate": 8.567405677219497e-05, + "loss": 0.819, + "step": 12273 + }, + { + "epoch": 2.185185185185185, + "grad_norm": 0.7858524322509766, + "learning_rate": 8.566020382196532e-05, + "loss": 1.0708, + "step": 12274 + }, + { + "epoch": 2.185363247863248, + "grad_norm": 1.0223300457000732, + "learning_rate": 8.564635115269422e-05, + "loss": 0.9929, + "step": 12275 + }, + { + "epoch": 2.1855413105413106, + "grad_norm": 0.7749526500701904, + "learning_rate": 8.5632498764653e-05, + "loss": 0.7555, + "step": 12276 + }, + { + "epoch": 2.1857193732193734, + "grad_norm": 0.8443665504455566, + "learning_rate": 8.561864665811313e-05, + "loss": 0.8488, + "step": 12277 + }, + { + "epoch": 2.185897435897436, + "grad_norm": 0.7482786178588867, + "learning_rate": 8.560479483334603e-05, + "loss": 0.6535, + "step": 12278 + }, + { + "epoch": 2.1860754985754984, + "grad_norm": 0.7981070876121521, + "learning_rate": 8.559094329062305e-05, + "loss": 0.7698, + "step": 12279 + }, + { + "epoch": 2.186253561253561, + "grad_norm": 0.7612428665161133, + "learning_rate": 8.557709203021564e-05, + "loss": 0.9086, + "step": 12280 + }, + { + "epoch": 2.186431623931624, + "grad_norm": 0.8246445059776306, + "learning_rate": 8.556324105239512e-05, + "loss": 0.9927, + "step": 12281 + }, + { + "epoch": 2.1866096866096867, + "grad_norm": 0.8902820348739624, + "learning_rate": 8.554939035743292e-05, + "loss": 0.8474, + "step": 12282 + }, + { + "epoch": 2.1867877492877494, + "grad_norm": 0.9992623329162598, + "learning_rate": 8.553553994560037e-05, + "loss": 1.0898, + "step": 12283 + }, + { + "epoch": 2.1869658119658117, + "grad_norm": 0.9124125838279724, + "learning_rate": 8.552168981716892e-05, + "loss": 0.9542, + "step": 12284 + }, + { + "epoch": 2.1871438746438745, + "grad_norm": 0.6818730235099792, + "learning_rate": 8.550783997240983e-05, + "loss": 0.678, + "step": 12285 + }, + { + "epoch": 2.1873219373219372, + "grad_norm": 0.8302112817764282, + "learning_rate": 8.549399041159455e-05, + "loss": 0.9955, + "step": 12286 + }, + { + "epoch": 2.1875, + "grad_norm": 0.840419352054596, + "learning_rate": 8.548014113499436e-05, + "loss": 0.9299, + "step": 12287 + }, + { + "epoch": 2.1876780626780628, + "grad_norm": 0.8317474722862244, + "learning_rate": 8.546629214288067e-05, + "loss": 0.754, + "step": 12288 + }, + { + "epoch": 2.1878561253561255, + "grad_norm": 0.879148542881012, + "learning_rate": 8.545244343552476e-05, + "loss": 0.9238, + "step": 12289 + }, + { + "epoch": 2.1880341880341883, + "grad_norm": 0.7899607419967651, + "learning_rate": 8.543859501319805e-05, + "loss": 0.8277, + "step": 12290 + }, + { + "epoch": 2.1882122507122506, + "grad_norm": 0.837785542011261, + "learning_rate": 8.542474687617176e-05, + "loss": 0.8801, + "step": 12291 + }, + { + "epoch": 2.1883903133903133, + "grad_norm": 0.856213390827179, + "learning_rate": 8.541089902471733e-05, + "loss": 0.7589, + "step": 12292 + }, + { + "epoch": 2.188568376068376, + "grad_norm": 0.7915818095207214, + "learning_rate": 8.539705145910599e-05, + "loss": 0.8071, + "step": 12293 + }, + { + "epoch": 2.188746438746439, + "grad_norm": 0.8266519904136658, + "learning_rate": 8.538320417960914e-05, + "loss": 0.9482, + "step": 12294 + }, + { + "epoch": 2.1889245014245016, + "grad_norm": 0.8505687713623047, + "learning_rate": 8.536935718649799e-05, + "loss": 0.9272, + "step": 12295 + }, + { + "epoch": 2.189102564102564, + "grad_norm": 0.7530698776245117, + "learning_rate": 8.535551048004394e-05, + "loss": 0.7908, + "step": 12296 + }, + { + "epoch": 2.1892806267806266, + "grad_norm": 0.7904362678527832, + "learning_rate": 8.534166406051818e-05, + "loss": 0.9771, + "step": 12297 + }, + { + "epoch": 2.1894586894586894, + "grad_norm": 0.7860299944877625, + "learning_rate": 8.532781792819209e-05, + "loss": 0.605, + "step": 12298 + }, + { + "epoch": 2.189636752136752, + "grad_norm": 0.7718655467033386, + "learning_rate": 8.531397208333695e-05, + "loss": 0.8844, + "step": 12299 + }, + { + "epoch": 2.189814814814815, + "grad_norm": 0.8069637417793274, + "learning_rate": 8.530012652622397e-05, + "loss": 0.8571, + "step": 12300 + }, + { + "epoch": 2.1899928774928776, + "grad_norm": 0.8557140231132507, + "learning_rate": 8.528628125712455e-05, + "loss": 0.7396, + "step": 12301 + }, + { + "epoch": 2.1901709401709404, + "grad_norm": 0.8547600507736206, + "learning_rate": 8.527243627630983e-05, + "loss": 0.8073, + "step": 12302 + }, + { + "epoch": 2.1903490028490027, + "grad_norm": 0.8217329382896423, + "learning_rate": 8.525859158405114e-05, + "loss": 0.9723, + "step": 12303 + }, + { + "epoch": 2.1905270655270654, + "grad_norm": 0.896946132183075, + "learning_rate": 8.524474718061972e-05, + "loss": 0.7896, + "step": 12304 + }, + { + "epoch": 2.190705128205128, + "grad_norm": 1.0846823453903198, + "learning_rate": 8.523090306628685e-05, + "loss": 0.7689, + "step": 12305 + }, + { + "epoch": 2.190883190883191, + "grad_norm": 0.7265166640281677, + "learning_rate": 8.521705924132373e-05, + "loss": 0.8451, + "step": 12306 + }, + { + "epoch": 2.1910612535612537, + "grad_norm": 0.8806917071342468, + "learning_rate": 8.520321570600162e-05, + "loss": 0.8769, + "step": 12307 + }, + { + "epoch": 2.191239316239316, + "grad_norm": 0.7528414130210876, + "learning_rate": 8.518937246059176e-05, + "loss": 0.7137, + "step": 12308 + }, + { + "epoch": 2.1914173789173788, + "grad_norm": 0.9313900470733643, + "learning_rate": 8.517552950536543e-05, + "loss": 0.933, + "step": 12309 + }, + { + "epoch": 2.1915954415954415, + "grad_norm": 0.8363727331161499, + "learning_rate": 8.516168684059375e-05, + "loss": 0.899, + "step": 12310 + }, + { + "epoch": 2.1917735042735043, + "grad_norm": 0.7939122915267944, + "learning_rate": 8.514784446654803e-05, + "loss": 1.0323, + "step": 12311 + }, + { + "epoch": 2.191951566951567, + "grad_norm": 0.8744710087776184, + "learning_rate": 8.51340023834994e-05, + "loss": 0.8738, + "step": 12312 + }, + { + "epoch": 2.1921296296296298, + "grad_norm": 0.779353678226471, + "learning_rate": 8.512016059171916e-05, + "loss": 0.7692, + "step": 12313 + }, + { + "epoch": 2.1923076923076925, + "grad_norm": 0.8578362464904785, + "learning_rate": 8.510631909147841e-05, + "loss": 1.0636, + "step": 12314 + }, + { + "epoch": 2.192485754985755, + "grad_norm": 0.7210206985473633, + "learning_rate": 8.509247788304846e-05, + "loss": 0.6342, + "step": 12315 + }, + { + "epoch": 2.1926638176638176, + "grad_norm": 0.7221980690956116, + "learning_rate": 8.50786369667004e-05, + "loss": 0.7022, + "step": 12316 + }, + { + "epoch": 2.1928418803418803, + "grad_norm": 0.7871465086936951, + "learning_rate": 8.506479634270544e-05, + "loss": 0.9349, + "step": 12317 + }, + { + "epoch": 2.193019943019943, + "grad_norm": 0.7396262884140015, + "learning_rate": 8.505095601133479e-05, + "loss": 0.8644, + "step": 12318 + }, + { + "epoch": 2.193198005698006, + "grad_norm": 0.7513349652290344, + "learning_rate": 8.503711597285959e-05, + "loss": 0.7881, + "step": 12319 + }, + { + "epoch": 2.1933760683760686, + "grad_norm": 0.7280148863792419, + "learning_rate": 8.502327622755106e-05, + "loss": 0.7115, + "step": 12320 + }, + { + "epoch": 2.193554131054131, + "grad_norm": 0.792238712310791, + "learning_rate": 8.500943677568028e-05, + "loss": 0.8193, + "step": 12321 + }, + { + "epoch": 2.1937321937321936, + "grad_norm": 0.8709526062011719, + "learning_rate": 8.499559761751847e-05, + "loss": 0.8957, + "step": 12322 + }, + { + "epoch": 2.1939102564102564, + "grad_norm": 0.6865217685699463, + "learning_rate": 8.498175875333674e-05, + "loss": 0.6853, + "step": 12323 + }, + { + "epoch": 2.194088319088319, + "grad_norm": 0.7797526121139526, + "learning_rate": 8.496792018340625e-05, + "loss": 0.8885, + "step": 12324 + }, + { + "epoch": 2.194266381766382, + "grad_norm": 0.8806295394897461, + "learning_rate": 8.495408190799814e-05, + "loss": 0.9322, + "step": 12325 + }, + { + "epoch": 2.1944444444444446, + "grad_norm": 0.8566734790802002, + "learning_rate": 8.494024392738355e-05, + "loss": 0.9363, + "step": 12326 + }, + { + "epoch": 2.194622507122507, + "grad_norm": 0.8058465123176575, + "learning_rate": 8.49264062418336e-05, + "loss": 0.9007, + "step": 12327 + }, + { + "epoch": 2.1948005698005697, + "grad_norm": 0.7895804643630981, + "learning_rate": 8.491256885161938e-05, + "loss": 0.8486, + "step": 12328 + }, + { + "epoch": 2.1949786324786325, + "grad_norm": 0.7626506686210632, + "learning_rate": 8.489873175701204e-05, + "loss": 0.8208, + "step": 12329 + }, + { + "epoch": 2.195156695156695, + "grad_norm": 0.8917649388313293, + "learning_rate": 8.488489495828272e-05, + "loss": 1.1036, + "step": 12330 + }, + { + "epoch": 2.195334757834758, + "grad_norm": 0.7614438533782959, + "learning_rate": 8.487105845570242e-05, + "loss": 0.7124, + "step": 12331 + }, + { + "epoch": 2.1955128205128207, + "grad_norm": 0.7697421312332153, + "learning_rate": 8.485722224954237e-05, + "loss": 0.8831, + "step": 12332 + }, + { + "epoch": 2.195690883190883, + "grad_norm": 0.7449761629104614, + "learning_rate": 8.484338634007354e-05, + "loss": 0.9115, + "step": 12333 + }, + { + "epoch": 2.1958689458689458, + "grad_norm": 0.7099741101264954, + "learning_rate": 8.482955072756709e-05, + "loss": 0.6907, + "step": 12334 + }, + { + "epoch": 2.1960470085470085, + "grad_norm": 0.7856435775756836, + "learning_rate": 8.481571541229406e-05, + "loss": 1.0565, + "step": 12335 + }, + { + "epoch": 2.1962250712250713, + "grad_norm": 0.8374622464179993, + "learning_rate": 8.48018803945256e-05, + "loss": 0.9191, + "step": 12336 + }, + { + "epoch": 2.196403133903134, + "grad_norm": 0.7530848383903503, + "learning_rate": 8.478804567453265e-05, + "loss": 0.6576, + "step": 12337 + }, + { + "epoch": 2.1965811965811968, + "grad_norm": 0.774861216545105, + "learning_rate": 8.477421125258637e-05, + "loss": 1.0258, + "step": 12338 + }, + { + "epoch": 2.196759259259259, + "grad_norm": 0.9623909592628479, + "learning_rate": 8.47603771289578e-05, + "loss": 1.0192, + "step": 12339 + }, + { + "epoch": 2.196937321937322, + "grad_norm": 0.8253501653671265, + "learning_rate": 8.474654330391797e-05, + "loss": 0.7823, + "step": 12340 + }, + { + "epoch": 2.1971153846153846, + "grad_norm": 0.8683596849441528, + "learning_rate": 8.473270977773797e-05, + "loss": 0.8002, + "step": 12341 + }, + { + "epoch": 2.1972934472934473, + "grad_norm": 0.9093332886695862, + "learning_rate": 8.471887655068877e-05, + "loss": 1.0315, + "step": 12342 + }, + { + "epoch": 2.19747150997151, + "grad_norm": 0.7313206791877747, + "learning_rate": 8.470504362304147e-05, + "loss": 0.8238, + "step": 12343 + }, + { + "epoch": 2.197649572649573, + "grad_norm": 0.8464672565460205, + "learning_rate": 8.469121099506703e-05, + "loss": 0.8104, + "step": 12344 + }, + { + "epoch": 2.197827635327635, + "grad_norm": 0.9213936924934387, + "learning_rate": 8.467737866703657e-05, + "loss": 0.9963, + "step": 12345 + }, + { + "epoch": 2.198005698005698, + "grad_norm": 0.8033352494239807, + "learning_rate": 8.466354663922099e-05, + "loss": 0.9788, + "step": 12346 + }, + { + "epoch": 2.1981837606837606, + "grad_norm": 0.7210986018180847, + "learning_rate": 8.464971491189141e-05, + "loss": 0.7597, + "step": 12347 + }, + { + "epoch": 2.1983618233618234, + "grad_norm": 0.8128374814987183, + "learning_rate": 8.463588348531872e-05, + "loss": 0.9575, + "step": 12348 + }, + { + "epoch": 2.198539886039886, + "grad_norm": 0.7276061773300171, + "learning_rate": 8.4622052359774e-05, + "loss": 0.815, + "step": 12349 + }, + { + "epoch": 2.198717948717949, + "grad_norm": 0.7463665008544922, + "learning_rate": 8.46082215355282e-05, + "loss": 0.9782, + "step": 12350 + }, + { + "epoch": 2.198896011396011, + "grad_norm": 0.8288317918777466, + "learning_rate": 8.459439101285238e-05, + "loss": 0.8206, + "step": 12351 + }, + { + "epoch": 2.199074074074074, + "grad_norm": 0.8286055326461792, + "learning_rate": 8.458056079201742e-05, + "loss": 0.9819, + "step": 12352 + }, + { + "epoch": 2.1992521367521367, + "grad_norm": 0.8138381242752075, + "learning_rate": 8.456673087329436e-05, + "loss": 0.8565, + "step": 12353 + }, + { + "epoch": 2.1994301994301995, + "grad_norm": 0.9059311747550964, + "learning_rate": 8.455290125695412e-05, + "loss": 0.8727, + "step": 12354 + }, + { + "epoch": 2.199608262108262, + "grad_norm": 0.6138933300971985, + "learning_rate": 8.453907194326773e-05, + "loss": 0.5635, + "step": 12355 + }, + { + "epoch": 2.199786324786325, + "grad_norm": 0.870585560798645, + "learning_rate": 8.452524293250608e-05, + "loss": 0.7401, + "step": 12356 + }, + { + "epoch": 2.1999643874643873, + "grad_norm": 0.8393024802207947, + "learning_rate": 8.451141422494013e-05, + "loss": 1.0083, + "step": 12357 + }, + { + "epoch": 2.20014245014245, + "grad_norm": 0.7667146325111389, + "learning_rate": 8.449758582084091e-05, + "loss": 0.8915, + "step": 12358 + }, + { + "epoch": 2.2003205128205128, + "grad_norm": 1.0229144096374512, + "learning_rate": 8.448375772047923e-05, + "loss": 0.8879, + "step": 12359 + }, + { + "epoch": 2.2004985754985755, + "grad_norm": 0.7670294046401978, + "learning_rate": 8.446992992412611e-05, + "loss": 0.8233, + "step": 12360 + }, + { + "epoch": 2.2006766381766383, + "grad_norm": 0.7110083103179932, + "learning_rate": 8.445610243205244e-05, + "loss": 0.6315, + "step": 12361 + }, + { + "epoch": 2.200854700854701, + "grad_norm": 0.7801400423049927, + "learning_rate": 8.444227524452918e-05, + "loss": 0.7758, + "step": 12362 + }, + { + "epoch": 2.2010327635327633, + "grad_norm": 0.8762022852897644, + "learning_rate": 8.44284483618272e-05, + "loss": 0.9308, + "step": 12363 + }, + { + "epoch": 2.201210826210826, + "grad_norm": 0.811890184879303, + "learning_rate": 8.441462178421742e-05, + "loss": 1.0322, + "step": 12364 + }, + { + "epoch": 2.201388888888889, + "grad_norm": 0.8128690719604492, + "learning_rate": 8.440079551197076e-05, + "loss": 1.0669, + "step": 12365 + }, + { + "epoch": 2.2015669515669516, + "grad_norm": 0.8925766348838806, + "learning_rate": 8.438696954535812e-05, + "loss": 0.8848, + "step": 12366 + }, + { + "epoch": 2.2017450142450143, + "grad_norm": 0.9104064106941223, + "learning_rate": 8.437314388465036e-05, + "loss": 0.8227, + "step": 12367 + }, + { + "epoch": 2.201923076923077, + "grad_norm": 0.7956777215003967, + "learning_rate": 8.43593185301184e-05, + "loss": 0.7616, + "step": 12368 + }, + { + "epoch": 2.2021011396011394, + "grad_norm": 0.7658423185348511, + "learning_rate": 8.434549348203309e-05, + "loss": 0.9406, + "step": 12369 + }, + { + "epoch": 2.202279202279202, + "grad_norm": 0.7650682926177979, + "learning_rate": 8.433166874066532e-05, + "loss": 0.9031, + "step": 12370 + }, + { + "epoch": 2.202457264957265, + "grad_norm": 0.8613301515579224, + "learning_rate": 8.431784430628594e-05, + "loss": 0.9184, + "step": 12371 + }, + { + "epoch": 2.2026353276353277, + "grad_norm": 0.8446599245071411, + "learning_rate": 8.430402017916586e-05, + "loss": 0.8639, + "step": 12372 + }, + { + "epoch": 2.2028133903133904, + "grad_norm": 0.8082340955734253, + "learning_rate": 8.429019635957585e-05, + "loss": 0.7365, + "step": 12373 + }, + { + "epoch": 2.202991452991453, + "grad_norm": 0.8843092918395996, + "learning_rate": 8.427637284778683e-05, + "loss": 0.8679, + "step": 12374 + }, + { + "epoch": 2.2031695156695155, + "grad_norm": 0.8475705981254578, + "learning_rate": 8.426254964406961e-05, + "loss": 0.6614, + "step": 12375 + }, + { + "epoch": 2.203347578347578, + "grad_norm": 0.9980667233467102, + "learning_rate": 8.424872674869507e-05, + "loss": 0.9103, + "step": 12376 + }, + { + "epoch": 2.203525641025641, + "grad_norm": 0.8033170104026794, + "learning_rate": 8.423490416193398e-05, + "loss": 0.7668, + "step": 12377 + }, + { + "epoch": 2.2037037037037037, + "grad_norm": 0.8275265097618103, + "learning_rate": 8.422108188405718e-05, + "loss": 0.7448, + "step": 12378 + }, + { + "epoch": 2.2038817663817665, + "grad_norm": 0.7622979283332825, + "learning_rate": 8.420725991533554e-05, + "loss": 0.8121, + "step": 12379 + }, + { + "epoch": 2.2040598290598292, + "grad_norm": 0.8580977320671082, + "learning_rate": 8.41934382560398e-05, + "loss": 0.8437, + "step": 12380 + }, + { + "epoch": 2.2042378917378915, + "grad_norm": 0.8443751931190491, + "learning_rate": 8.417961690644086e-05, + "loss": 0.971, + "step": 12381 + }, + { + "epoch": 2.2044159544159543, + "grad_norm": 0.782430112361908, + "learning_rate": 8.416579586680939e-05, + "loss": 0.8367, + "step": 12382 + }, + { + "epoch": 2.204594017094017, + "grad_norm": 0.8664544820785522, + "learning_rate": 8.415197513741633e-05, + "loss": 0.8288, + "step": 12383 + }, + { + "epoch": 2.20477207977208, + "grad_norm": 0.7207586169242859, + "learning_rate": 8.413815471853235e-05, + "loss": 0.8038, + "step": 12384 + }, + { + "epoch": 2.2049501424501425, + "grad_norm": 0.743195652961731, + "learning_rate": 8.412433461042828e-05, + "loss": 0.705, + "step": 12385 + }, + { + "epoch": 2.2051282051282053, + "grad_norm": 0.7891412377357483, + "learning_rate": 8.411051481337488e-05, + "loss": 0.9729, + "step": 12386 + }, + { + "epoch": 2.205306267806268, + "grad_norm": 0.838847815990448, + "learning_rate": 8.4096695327643e-05, + "loss": 0.9053, + "step": 12387 + }, + { + "epoch": 2.2054843304843303, + "grad_norm": 0.7717056274414062, + "learning_rate": 8.408287615350328e-05, + "loss": 0.7388, + "step": 12388 + }, + { + "epoch": 2.205662393162393, + "grad_norm": 0.7209389209747314, + "learning_rate": 8.406905729122654e-05, + "loss": 0.6411, + "step": 12389 + }, + { + "epoch": 2.205840455840456, + "grad_norm": 0.822475790977478, + "learning_rate": 8.405523874108354e-05, + "loss": 0.9574, + "step": 12390 + }, + { + "epoch": 2.2060185185185186, + "grad_norm": 0.9401286840438843, + "learning_rate": 8.404142050334504e-05, + "loss": 0.8915, + "step": 12391 + }, + { + "epoch": 2.2061965811965814, + "grad_norm": 0.8247103691101074, + "learning_rate": 8.40276025782817e-05, + "loss": 0.8369, + "step": 12392 + }, + { + "epoch": 2.2063746438746437, + "grad_norm": 0.8082301020622253, + "learning_rate": 8.401378496616437e-05, + "loss": 0.9321, + "step": 12393 + }, + { + "epoch": 2.2065527065527064, + "grad_norm": 0.8156028389930725, + "learning_rate": 8.399996766726367e-05, + "loss": 0.7599, + "step": 12394 + }, + { + "epoch": 2.206730769230769, + "grad_norm": 0.7941898107528687, + "learning_rate": 8.398615068185038e-05, + "loss": 0.812, + "step": 12395 + }, + { + "epoch": 2.206908831908832, + "grad_norm": 0.7013470530509949, + "learning_rate": 8.397233401019518e-05, + "loss": 0.7914, + "step": 12396 + }, + { + "epoch": 2.2070868945868947, + "grad_norm": 0.6028649210929871, + "learning_rate": 8.395851765256881e-05, + "loss": 0.5787, + "step": 12397 + }, + { + "epoch": 2.2072649572649574, + "grad_norm": 0.9031504392623901, + "learning_rate": 8.3944701609242e-05, + "loss": 0.8677, + "step": 12398 + }, + { + "epoch": 2.20744301994302, + "grad_norm": 0.7370864748954773, + "learning_rate": 8.393088588048536e-05, + "loss": 0.9025, + "step": 12399 + }, + { + "epoch": 2.2076210826210825, + "grad_norm": 0.7764220237731934, + "learning_rate": 8.391707046656968e-05, + "loss": 0.8805, + "step": 12400 + }, + { + "epoch": 2.2077991452991452, + "grad_norm": 0.7456721663475037, + "learning_rate": 8.390325536776553e-05, + "loss": 0.7739, + "step": 12401 + }, + { + "epoch": 2.207977207977208, + "grad_norm": 0.8032360076904297, + "learning_rate": 8.388944058434373e-05, + "loss": 0.9765, + "step": 12402 + }, + { + "epoch": 2.2081552706552707, + "grad_norm": 0.8502830266952515, + "learning_rate": 8.387562611657483e-05, + "loss": 0.9356, + "step": 12403 + }, + { + "epoch": 2.2083333333333335, + "grad_norm": 0.812216579914093, + "learning_rate": 8.386181196472956e-05, + "loss": 0.8846, + "step": 12404 + }, + { + "epoch": 2.208511396011396, + "grad_norm": 0.6996115446090698, + "learning_rate": 8.384799812907853e-05, + "loss": 0.7035, + "step": 12405 + }, + { + "epoch": 2.2086894586894585, + "grad_norm": 0.7909261584281921, + "learning_rate": 8.383418460989245e-05, + "loss": 0.8025, + "step": 12406 + }, + { + "epoch": 2.2088675213675213, + "grad_norm": 0.8278310894966125, + "learning_rate": 8.382037140744192e-05, + "loss": 0.7982, + "step": 12407 + }, + { + "epoch": 2.209045584045584, + "grad_norm": 0.7558199167251587, + "learning_rate": 8.380655852199763e-05, + "loss": 0.854, + "step": 12408 + }, + { + "epoch": 2.209223646723647, + "grad_norm": 0.8516034483909607, + "learning_rate": 8.379274595383016e-05, + "loss": 0.7497, + "step": 12409 + }, + { + "epoch": 2.2094017094017095, + "grad_norm": 0.777004599571228, + "learning_rate": 8.377893370321018e-05, + "loss": 0.797, + "step": 12410 + }, + { + "epoch": 2.2095797720797723, + "grad_norm": 0.8820251822471619, + "learning_rate": 8.376512177040829e-05, + "loss": 0.9229, + "step": 12411 + }, + { + "epoch": 2.2097578347578346, + "grad_norm": 0.8623200058937073, + "learning_rate": 8.375131015569514e-05, + "loss": 1.011, + "step": 12412 + }, + { + "epoch": 2.2099358974358974, + "grad_norm": 0.9192054271697998, + "learning_rate": 8.373749885934127e-05, + "loss": 0.8711, + "step": 12413 + }, + { + "epoch": 2.21011396011396, + "grad_norm": 0.7627860903739929, + "learning_rate": 8.372368788161736e-05, + "loss": 0.5937, + "step": 12414 + }, + { + "epoch": 2.210292022792023, + "grad_norm": 0.74603670835495, + "learning_rate": 8.370987722279395e-05, + "loss": 0.8238, + "step": 12415 + }, + { + "epoch": 2.2104700854700856, + "grad_norm": 0.884469211101532, + "learning_rate": 8.369606688314165e-05, + "loss": 1.1957, + "step": 12416 + }, + { + "epoch": 2.210648148148148, + "grad_norm": 0.8145224452018738, + "learning_rate": 8.36822568629311e-05, + "loss": 0.8517, + "step": 12417 + }, + { + "epoch": 2.2108262108262107, + "grad_norm": 0.8167604207992554, + "learning_rate": 8.366844716243279e-05, + "loss": 0.9701, + "step": 12418 + }, + { + "epoch": 2.2110042735042734, + "grad_norm": 0.7668562531471252, + "learning_rate": 8.365463778191736e-05, + "loss": 1.0281, + "step": 12419 + }, + { + "epoch": 2.211182336182336, + "grad_norm": 0.8455148339271545, + "learning_rate": 8.364082872165532e-05, + "loss": 0.7812, + "step": 12420 + }, + { + "epoch": 2.211360398860399, + "grad_norm": 0.8756504654884338, + "learning_rate": 8.362701998191728e-05, + "loss": 0.779, + "step": 12421 + }, + { + "epoch": 2.2115384615384617, + "grad_norm": 0.8239594101905823, + "learning_rate": 8.361321156297374e-05, + "loss": 0.8581, + "step": 12422 + }, + { + "epoch": 2.2117165242165244, + "grad_norm": 0.7719405889511108, + "learning_rate": 8.359940346509533e-05, + "loss": 0.7593, + "step": 12423 + }, + { + "epoch": 2.2118945868945867, + "grad_norm": 0.8607308268547058, + "learning_rate": 8.358559568855249e-05, + "loss": 1.0618, + "step": 12424 + }, + { + "epoch": 2.2120726495726495, + "grad_norm": 0.750431478023529, + "learning_rate": 8.357178823361582e-05, + "loss": 0.7779, + "step": 12425 + }, + { + "epoch": 2.2122507122507122, + "grad_norm": 0.7770674824714661, + "learning_rate": 8.355798110055583e-05, + "loss": 0.6837, + "step": 12426 + }, + { + "epoch": 2.212428774928775, + "grad_norm": 0.7924200296401978, + "learning_rate": 8.354417428964307e-05, + "loss": 0.8092, + "step": 12427 + }, + { + "epoch": 2.2126068376068377, + "grad_norm": 0.7784677743911743, + "learning_rate": 8.3530367801148e-05, + "loss": 0.7168, + "step": 12428 + }, + { + "epoch": 2.2127849002849005, + "grad_norm": 1.0548151731491089, + "learning_rate": 8.351656163534121e-05, + "loss": 0.9286, + "step": 12429 + }, + { + "epoch": 2.212962962962963, + "grad_norm": 0.8983006477355957, + "learning_rate": 8.35027557924931e-05, + "loss": 0.878, + "step": 12430 + }, + { + "epoch": 2.2131410256410255, + "grad_norm": 0.8136780261993408, + "learning_rate": 8.348895027287424e-05, + "loss": 0.7901, + "step": 12431 + }, + { + "epoch": 2.2133190883190883, + "grad_norm": 0.8186678290367126, + "learning_rate": 8.347514507675508e-05, + "loss": 0.8994, + "step": 12432 + }, + { + "epoch": 2.213497150997151, + "grad_norm": 0.880790650844574, + "learning_rate": 8.346134020440617e-05, + "loss": 1.0681, + "step": 12433 + }, + { + "epoch": 2.213675213675214, + "grad_norm": 0.8061994910240173, + "learning_rate": 8.344753565609789e-05, + "loss": 0.8466, + "step": 12434 + }, + { + "epoch": 2.2138532763532766, + "grad_norm": 0.8041423559188843, + "learning_rate": 8.34337314321008e-05, + "loss": 0.897, + "step": 12435 + }, + { + "epoch": 2.214031339031339, + "grad_norm": 0.5797891616821289, + "learning_rate": 8.34199275326853e-05, + "loss": 0.4827, + "step": 12436 + }, + { + "epoch": 2.2142094017094016, + "grad_norm": 0.7373392581939697, + "learning_rate": 8.340612395812188e-05, + "loss": 0.779, + "step": 12437 + }, + { + "epoch": 2.2143874643874644, + "grad_norm": 0.7852202653884888, + "learning_rate": 8.339232070868102e-05, + "loss": 0.8001, + "step": 12438 + }, + { + "epoch": 2.214565527065527, + "grad_norm": 0.8209689259529114, + "learning_rate": 8.337851778463311e-05, + "loss": 0.7492, + "step": 12439 + }, + { + "epoch": 2.21474358974359, + "grad_norm": 0.9393492937088013, + "learning_rate": 8.336471518624867e-05, + "loss": 0.884, + "step": 12440 + }, + { + "epoch": 2.2149216524216526, + "grad_norm": 0.6966122984886169, + "learning_rate": 8.3350912913798e-05, + "loss": 0.7364, + "step": 12441 + }, + { + "epoch": 2.215099715099715, + "grad_norm": 0.7379066944122314, + "learning_rate": 8.333711096755165e-05, + "loss": 0.7345, + "step": 12442 + }, + { + "epoch": 2.2152777777777777, + "grad_norm": 0.9011021256446838, + "learning_rate": 8.332330934777999e-05, + "loss": 0.8392, + "step": 12443 + }, + { + "epoch": 2.2154558404558404, + "grad_norm": 0.7718381285667419, + "learning_rate": 8.330950805475346e-05, + "loss": 0.9062, + "step": 12444 + }, + { + "epoch": 2.215633903133903, + "grad_norm": 0.8584564328193665, + "learning_rate": 8.329570708874241e-05, + "loss": 0.9612, + "step": 12445 + }, + { + "epoch": 2.215811965811966, + "grad_norm": 0.7711616158485413, + "learning_rate": 8.32819064500173e-05, + "loss": 0.731, + "step": 12446 + }, + { + "epoch": 2.2159900284900287, + "grad_norm": 0.8014609217643738, + "learning_rate": 8.326810613884849e-05, + "loss": 1.0128, + "step": 12447 + }, + { + "epoch": 2.216168091168091, + "grad_norm": 0.7837486863136292, + "learning_rate": 8.325430615550642e-05, + "loss": 0.8271, + "step": 12448 + }, + { + "epoch": 2.2163461538461537, + "grad_norm": 0.9399738907814026, + "learning_rate": 8.324050650026139e-05, + "loss": 1.0433, + "step": 12449 + }, + { + "epoch": 2.2165242165242165, + "grad_norm": 0.8302193284034729, + "learning_rate": 8.322670717338385e-05, + "loss": 1.0259, + "step": 12450 + }, + { + "epoch": 2.2167022792022792, + "grad_norm": 0.7707721590995789, + "learning_rate": 8.321290817514411e-05, + "loss": 0.6972, + "step": 12451 + }, + { + "epoch": 2.216880341880342, + "grad_norm": 0.5814536809921265, + "learning_rate": 8.319910950581261e-05, + "loss": 0.5846, + "step": 12452 + }, + { + "epoch": 2.2170584045584047, + "grad_norm": 0.8249124884605408, + "learning_rate": 8.318531116565962e-05, + "loss": 0.7417, + "step": 12453 + }, + { + "epoch": 2.217236467236467, + "grad_norm": 0.7116015553474426, + "learning_rate": 8.317151315495556e-05, + "loss": 0.8698, + "step": 12454 + }, + { + "epoch": 2.21741452991453, + "grad_norm": 0.8025332689285278, + "learning_rate": 8.31577154739707e-05, + "loss": 0.825, + "step": 12455 + }, + { + "epoch": 2.2175925925925926, + "grad_norm": 0.8962773680686951, + "learning_rate": 8.314391812297542e-05, + "loss": 0.9987, + "step": 12456 + }, + { + "epoch": 2.2177706552706553, + "grad_norm": 0.8446899652481079, + "learning_rate": 8.313012110224008e-05, + "loss": 0.8554, + "step": 12457 + }, + { + "epoch": 2.217948717948718, + "grad_norm": 0.7759326696395874, + "learning_rate": 8.311632441203494e-05, + "loss": 0.8206, + "step": 12458 + }, + { + "epoch": 2.218126780626781, + "grad_norm": 0.9782015085220337, + "learning_rate": 8.31025280526304e-05, + "loss": 0.8183, + "step": 12459 + }, + { + "epoch": 2.218304843304843, + "grad_norm": 0.7445226907730103, + "learning_rate": 8.308873202429666e-05, + "loss": 0.6819, + "step": 12460 + }, + { + "epoch": 2.218482905982906, + "grad_norm": 0.7613980770111084, + "learning_rate": 8.307493632730413e-05, + "loss": 0.6283, + "step": 12461 + }, + { + "epoch": 2.2186609686609686, + "grad_norm": 0.7437549829483032, + "learning_rate": 8.306114096192304e-05, + "loss": 0.7511, + "step": 12462 + }, + { + "epoch": 2.2188390313390314, + "grad_norm": 0.7600140571594238, + "learning_rate": 8.304734592842373e-05, + "loss": 0.8784, + "step": 12463 + }, + { + "epoch": 2.219017094017094, + "grad_norm": 0.9086898565292358, + "learning_rate": 8.303355122707644e-05, + "loss": 1.0818, + "step": 12464 + }, + { + "epoch": 2.219195156695157, + "grad_norm": 0.8674180507659912, + "learning_rate": 8.30197568581515e-05, + "loss": 0.8925, + "step": 12465 + }, + { + "epoch": 2.219373219373219, + "grad_norm": 0.893606960773468, + "learning_rate": 8.300596282191911e-05, + "loss": 0.9382, + "step": 12466 + }, + { + "epoch": 2.219551282051282, + "grad_norm": 0.7664543390274048, + "learning_rate": 8.29921691186496e-05, + "loss": 0.7893, + "step": 12467 + }, + { + "epoch": 2.2197293447293447, + "grad_norm": 0.8730209469795227, + "learning_rate": 8.297837574861318e-05, + "loss": 1.0509, + "step": 12468 + }, + { + "epoch": 2.2199074074074074, + "grad_norm": 0.8138112425804138, + "learning_rate": 8.296458271208018e-05, + "loss": 0.784, + "step": 12469 + }, + { + "epoch": 2.22008547008547, + "grad_norm": 0.8362413644790649, + "learning_rate": 8.295079000932073e-05, + "loss": 1.0236, + "step": 12470 + }, + { + "epoch": 2.220263532763533, + "grad_norm": 0.8422487378120422, + "learning_rate": 8.293699764060518e-05, + "loss": 0.9677, + "step": 12471 + }, + { + "epoch": 2.2204415954415953, + "grad_norm": 0.7290427088737488, + "learning_rate": 8.292320560620369e-05, + "loss": 0.7514, + "step": 12472 + }, + { + "epoch": 2.220619658119658, + "grad_norm": 0.8083370923995972, + "learning_rate": 8.290941390638653e-05, + "loss": 0.8136, + "step": 12473 + }, + { + "epoch": 2.2207977207977208, + "grad_norm": 0.8045510053634644, + "learning_rate": 8.289562254142389e-05, + "loss": 0.6753, + "step": 12474 + }, + { + "epoch": 2.2209757834757835, + "grad_norm": 0.8019934892654419, + "learning_rate": 8.288183151158602e-05, + "loss": 0.8147, + "step": 12475 + }, + { + "epoch": 2.2211538461538463, + "grad_norm": 0.8129584193229675, + "learning_rate": 8.286804081714306e-05, + "loss": 0.9137, + "step": 12476 + }, + { + "epoch": 2.221331908831909, + "grad_norm": 0.9729450345039368, + "learning_rate": 8.285425045836526e-05, + "loss": 0.7884, + "step": 12477 + }, + { + "epoch": 2.2215099715099713, + "grad_norm": 0.755081295967102, + "learning_rate": 8.284046043552282e-05, + "loss": 0.8496, + "step": 12478 + }, + { + "epoch": 2.221688034188034, + "grad_norm": 0.725267767906189, + "learning_rate": 8.282667074888589e-05, + "loss": 0.7054, + "step": 12479 + }, + { + "epoch": 2.221866096866097, + "grad_norm": 0.832098662853241, + "learning_rate": 8.281288139872472e-05, + "loss": 0.8729, + "step": 12480 + }, + { + "epoch": 2.2220441595441596, + "grad_norm": 0.9908086657524109, + "learning_rate": 8.27990923853094e-05, + "loss": 0.9106, + "step": 12481 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 0.8001172542572021, + "learning_rate": 8.278530370891013e-05, + "loss": 0.906, + "step": 12482 + }, + { + "epoch": 2.222400284900285, + "grad_norm": 0.7607424259185791, + "learning_rate": 8.277151536979709e-05, + "loss": 0.8125, + "step": 12483 + }, + { + "epoch": 2.2225783475783474, + "grad_norm": 0.7850996255874634, + "learning_rate": 8.275772736824042e-05, + "loss": 0.7017, + "step": 12484 + }, + { + "epoch": 2.22275641025641, + "grad_norm": 0.8376613855361938, + "learning_rate": 8.274393970451024e-05, + "loss": 1.0453, + "step": 12485 + }, + { + "epoch": 2.222934472934473, + "grad_norm": 0.7973353266716003, + "learning_rate": 8.273015237887673e-05, + "loss": 0.8337, + "step": 12486 + }, + { + "epoch": 2.2231125356125356, + "grad_norm": 0.7622607350349426, + "learning_rate": 8.271636539161e-05, + "loss": 0.8574, + "step": 12487 + }, + { + "epoch": 2.2232905982905984, + "grad_norm": 0.7839400768280029, + "learning_rate": 8.270257874298022e-05, + "loss": 0.8857, + "step": 12488 + }, + { + "epoch": 2.223468660968661, + "grad_norm": 0.7730473875999451, + "learning_rate": 8.268879243325743e-05, + "loss": 0.9578, + "step": 12489 + }, + { + "epoch": 2.2236467236467234, + "grad_norm": 0.7811899185180664, + "learning_rate": 8.267500646271184e-05, + "loss": 0.9469, + "step": 12490 + }, + { + "epoch": 2.223824786324786, + "grad_norm": 0.8570041060447693, + "learning_rate": 8.266122083161347e-05, + "loss": 0.8853, + "step": 12491 + }, + { + "epoch": 2.224002849002849, + "grad_norm": 0.7989770174026489, + "learning_rate": 8.264743554023248e-05, + "loss": 0.7467, + "step": 12492 + }, + { + "epoch": 2.2241809116809117, + "grad_norm": 0.8287475109100342, + "learning_rate": 8.263365058883891e-05, + "loss": 0.9987, + "step": 12493 + }, + { + "epoch": 2.2243589743589745, + "grad_norm": 0.8879026174545288, + "learning_rate": 8.261986597770295e-05, + "loss": 0.9503, + "step": 12494 + }, + { + "epoch": 2.224537037037037, + "grad_norm": 0.8153596520423889, + "learning_rate": 8.260608170709456e-05, + "loss": 0.9715, + "step": 12495 + }, + { + "epoch": 2.2247150997150995, + "grad_norm": 0.8294584155082703, + "learning_rate": 8.259229777728384e-05, + "loss": 0.958, + "step": 12496 + }, + { + "epoch": 2.2248931623931623, + "grad_norm": 0.76850426197052, + "learning_rate": 8.257851418854093e-05, + "loss": 0.7666, + "step": 12497 + }, + { + "epoch": 2.225071225071225, + "grad_norm": 0.743966817855835, + "learning_rate": 8.256473094113582e-05, + "loss": 0.8893, + "step": 12498 + }, + { + "epoch": 2.2252492877492878, + "grad_norm": 0.7339308857917786, + "learning_rate": 8.255094803533863e-05, + "loss": 0.7317, + "step": 12499 + }, + { + "epoch": 2.2254273504273505, + "grad_norm": 1.0800104141235352, + "learning_rate": 8.253716547141932e-05, + "loss": 1.0147, + "step": 12500 + }, + { + "epoch": 2.2256054131054133, + "grad_norm": 0.8518815636634827, + "learning_rate": 8.252338324964802e-05, + "loss": 0.9695, + "step": 12501 + }, + { + "epoch": 2.2257834757834756, + "grad_norm": 0.8706745505332947, + "learning_rate": 8.250960137029469e-05, + "loss": 0.7735, + "step": 12502 + }, + { + "epoch": 2.2259615384615383, + "grad_norm": 1.0482546091079712, + "learning_rate": 8.24958198336294e-05, + "loss": 1.0882, + "step": 12503 + }, + { + "epoch": 2.226139601139601, + "grad_norm": 0.8025278449058533, + "learning_rate": 8.248203863992213e-05, + "loss": 0.8573, + "step": 12504 + }, + { + "epoch": 2.226317663817664, + "grad_norm": 0.8267400860786438, + "learning_rate": 8.246825778944297e-05, + "loss": 0.8609, + "step": 12505 + }, + { + "epoch": 2.2264957264957266, + "grad_norm": 0.703681230545044, + "learning_rate": 8.245447728246184e-05, + "loss": 0.6934, + "step": 12506 + }, + { + "epoch": 2.2266737891737893, + "grad_norm": 0.807736873626709, + "learning_rate": 8.24406971192488e-05, + "loss": 0.7258, + "step": 12507 + }, + { + "epoch": 2.226851851851852, + "grad_norm": 0.7663748860359192, + "learning_rate": 8.24269173000738e-05, + "loss": 0.7825, + "step": 12508 + }, + { + "epoch": 2.2270299145299144, + "grad_norm": 0.7799240946769714, + "learning_rate": 8.24131378252069e-05, + "loss": 0.7868, + "step": 12509 + }, + { + "epoch": 2.227207977207977, + "grad_norm": 0.8309668302536011, + "learning_rate": 8.239935869491799e-05, + "loss": 0.7697, + "step": 12510 + }, + { + "epoch": 2.22738603988604, + "grad_norm": 0.7257094979286194, + "learning_rate": 8.23855799094771e-05, + "loss": 0.8168, + "step": 12511 + }, + { + "epoch": 2.2275641025641026, + "grad_norm": 0.8902100920677185, + "learning_rate": 8.237180146915416e-05, + "loss": 0.8606, + "step": 12512 + }, + { + "epoch": 2.2277421652421654, + "grad_norm": 0.8100315928459167, + "learning_rate": 8.235802337421919e-05, + "loss": 0.9225, + "step": 12513 + }, + { + "epoch": 2.2279202279202277, + "grad_norm": 0.6804848909378052, + "learning_rate": 8.234424562494205e-05, + "loss": 0.7047, + "step": 12514 + }, + { + "epoch": 2.2280982905982905, + "grad_norm": 0.8664964437484741, + "learning_rate": 8.233046822159276e-05, + "loss": 1.0255, + "step": 12515 + }, + { + "epoch": 2.228276353276353, + "grad_norm": 0.836857795715332, + "learning_rate": 8.231669116444128e-05, + "loss": 0.9818, + "step": 12516 + }, + { + "epoch": 2.228454415954416, + "grad_norm": 0.6999024748802185, + "learning_rate": 8.230291445375744e-05, + "loss": 0.7298, + "step": 12517 + }, + { + "epoch": 2.2286324786324787, + "grad_norm": 0.8676811456680298, + "learning_rate": 8.228913808981127e-05, + "loss": 0.9592, + "step": 12518 + }, + { + "epoch": 2.2288105413105415, + "grad_norm": 0.8088808655738831, + "learning_rate": 8.227536207287263e-05, + "loss": 1.0021, + "step": 12519 + }, + { + "epoch": 2.228988603988604, + "grad_norm": 0.7802120447158813, + "learning_rate": 8.226158640321149e-05, + "loss": 0.8519, + "step": 12520 + }, + { + "epoch": 2.2291666666666665, + "grad_norm": 0.7560334801673889, + "learning_rate": 8.224781108109766e-05, + "loss": 0.7676, + "step": 12521 + }, + { + "epoch": 2.2293447293447293, + "grad_norm": 0.7806954383850098, + "learning_rate": 8.223403610680113e-05, + "loss": 0.9151, + "step": 12522 + }, + { + "epoch": 2.229522792022792, + "grad_norm": 0.7972870469093323, + "learning_rate": 8.222026148059173e-05, + "loss": 0.8785, + "step": 12523 + }, + { + "epoch": 2.2297008547008548, + "grad_norm": 0.7868863344192505, + "learning_rate": 8.220648720273941e-05, + "loss": 0.8981, + "step": 12524 + }, + { + "epoch": 2.2298789173789175, + "grad_norm": 0.7388648390769958, + "learning_rate": 8.219271327351397e-05, + "loss": 0.7361, + "step": 12525 + }, + { + "epoch": 2.23005698005698, + "grad_norm": 0.7367138862609863, + "learning_rate": 8.217893969318538e-05, + "loss": 0.7357, + "step": 12526 + }, + { + "epoch": 2.2302350427350426, + "grad_norm": 0.8345077037811279, + "learning_rate": 8.216516646202339e-05, + "loss": 0.9671, + "step": 12527 + }, + { + "epoch": 2.2304131054131053, + "grad_norm": 0.7875744104385376, + "learning_rate": 8.215139358029793e-05, + "loss": 0.7991, + "step": 12528 + }, + { + "epoch": 2.230591168091168, + "grad_norm": 0.7444638609886169, + "learning_rate": 8.213762104827882e-05, + "loss": 0.6524, + "step": 12529 + }, + { + "epoch": 2.230769230769231, + "grad_norm": 0.6670697927474976, + "learning_rate": 8.212384886623597e-05, + "loss": 0.639, + "step": 12530 + }, + { + "epoch": 2.2309472934472936, + "grad_norm": 0.8348705172538757, + "learning_rate": 8.211007703443913e-05, + "loss": 0.8904, + "step": 12531 + }, + { + "epoch": 2.2311253561253563, + "grad_norm": 0.8458212614059448, + "learning_rate": 8.209630555315817e-05, + "loss": 0.8398, + "step": 12532 + }, + { + "epoch": 2.2313034188034186, + "grad_norm": 0.9043961763381958, + "learning_rate": 8.20825344226629e-05, + "loss": 0.904, + "step": 12533 + }, + { + "epoch": 2.2314814814814814, + "grad_norm": 0.8207734227180481, + "learning_rate": 8.206876364322319e-05, + "loss": 0.853, + "step": 12534 + }, + { + "epoch": 2.231659544159544, + "grad_norm": 0.9311240911483765, + "learning_rate": 8.205499321510876e-05, + "loss": 0.9807, + "step": 12535 + }, + { + "epoch": 2.231837606837607, + "grad_norm": 0.8379791378974915, + "learning_rate": 8.204122313858946e-05, + "loss": 0.8318, + "step": 12536 + }, + { + "epoch": 2.2320156695156697, + "grad_norm": 0.8078454732894897, + "learning_rate": 8.202745341393515e-05, + "loss": 0.8692, + "step": 12537 + }, + { + "epoch": 2.232193732193732, + "grad_norm": 0.7555927038192749, + "learning_rate": 8.201368404141547e-05, + "loss": 0.8514, + "step": 12538 + }, + { + "epoch": 2.2323717948717947, + "grad_norm": 0.7724241018295288, + "learning_rate": 8.199991502130035e-05, + "loss": 0.5758, + "step": 12539 + }, + { + "epoch": 2.2325498575498575, + "grad_norm": 0.7388870120048523, + "learning_rate": 8.198614635385946e-05, + "loss": 0.6265, + "step": 12540 + }, + { + "epoch": 2.23272792022792, + "grad_norm": 0.9006723761558533, + "learning_rate": 8.197237803936267e-05, + "loss": 0.8238, + "step": 12541 + }, + { + "epoch": 2.232905982905983, + "grad_norm": 0.917884349822998, + "learning_rate": 8.195861007807962e-05, + "loss": 0.9447, + "step": 12542 + }, + { + "epoch": 2.2330840455840457, + "grad_norm": 0.81849205493927, + "learning_rate": 8.194484247028016e-05, + "loss": 0.9071, + "step": 12543 + }, + { + "epoch": 2.2332621082621085, + "grad_norm": 0.8572089076042175, + "learning_rate": 8.193107521623398e-05, + "loss": 0.9068, + "step": 12544 + }, + { + "epoch": 2.2334401709401708, + "grad_norm": 0.7870976328849792, + "learning_rate": 8.19173083162109e-05, + "loss": 0.7595, + "step": 12545 + }, + { + "epoch": 2.2336182336182335, + "grad_norm": 0.8728759288787842, + "learning_rate": 8.190354177048055e-05, + "loss": 1.0974, + "step": 12546 + }, + { + "epoch": 2.2337962962962963, + "grad_norm": 0.7679606080055237, + "learning_rate": 8.188977557931274e-05, + "loss": 0.7068, + "step": 12547 + }, + { + "epoch": 2.233974358974359, + "grad_norm": 0.7753520011901855, + "learning_rate": 8.187600974297714e-05, + "loss": 0.8008, + "step": 12548 + }, + { + "epoch": 2.234152421652422, + "grad_norm": 0.7785305976867676, + "learning_rate": 8.186224426174348e-05, + "loss": 0.8528, + "step": 12549 + }, + { + "epoch": 2.2343304843304845, + "grad_norm": 0.7762976288795471, + "learning_rate": 8.184847913588145e-05, + "loss": 0.9264, + "step": 12550 + }, + { + "epoch": 2.234508547008547, + "grad_norm": 1.0543726682662964, + "learning_rate": 8.18347143656608e-05, + "loss": 0.9201, + "step": 12551 + }, + { + "epoch": 2.2346866096866096, + "grad_norm": 0.815389096736908, + "learning_rate": 8.182094995135116e-05, + "loss": 0.8834, + "step": 12552 + }, + { + "epoch": 2.2348646723646723, + "grad_norm": 0.774773895740509, + "learning_rate": 8.180718589322225e-05, + "loss": 0.8864, + "step": 12553 + }, + { + "epoch": 2.235042735042735, + "grad_norm": 0.8139658570289612, + "learning_rate": 8.179342219154372e-05, + "loss": 0.8696, + "step": 12554 + }, + { + "epoch": 2.235220797720798, + "grad_norm": 0.7804924249649048, + "learning_rate": 8.177965884658527e-05, + "loss": 0.8854, + "step": 12555 + }, + { + "epoch": 2.2353988603988606, + "grad_norm": 0.8601226210594177, + "learning_rate": 8.176589585861659e-05, + "loss": 0.9115, + "step": 12556 + }, + { + "epoch": 2.235576923076923, + "grad_norm": 0.7518162727355957, + "learning_rate": 8.175213322790726e-05, + "loss": 0.7871, + "step": 12557 + }, + { + "epoch": 2.2357549857549857, + "grad_norm": 0.7595868110656738, + "learning_rate": 8.1738370954727e-05, + "loss": 0.7597, + "step": 12558 + }, + { + "epoch": 2.2359330484330484, + "grad_norm": 0.8191643357276917, + "learning_rate": 8.17246090393454e-05, + "loss": 0.9443, + "step": 12559 + }, + { + "epoch": 2.236111111111111, + "grad_norm": 0.7854904532432556, + "learning_rate": 8.171084748203217e-05, + "loss": 0.8547, + "step": 12560 + }, + { + "epoch": 2.236289173789174, + "grad_norm": 0.8610023260116577, + "learning_rate": 8.169708628305684e-05, + "loss": 0.7846, + "step": 12561 + }, + { + "epoch": 2.2364672364672367, + "grad_norm": 0.8254715204238892, + "learning_rate": 8.168332544268914e-05, + "loss": 0.7493, + "step": 12562 + }, + { + "epoch": 2.236645299145299, + "grad_norm": 0.8390897512435913, + "learning_rate": 8.166956496119857e-05, + "loss": 0.9867, + "step": 12563 + }, + { + "epoch": 2.2368233618233617, + "grad_norm": 0.8179677128791809, + "learning_rate": 8.165580483885483e-05, + "loss": 0.8039, + "step": 12564 + }, + { + "epoch": 2.2370014245014245, + "grad_norm": 0.6722155809402466, + "learning_rate": 8.164204507592745e-05, + "loss": 0.695, + "step": 12565 + }, + { + "epoch": 2.2371794871794872, + "grad_norm": 0.8228170871734619, + "learning_rate": 8.162828567268612e-05, + "loss": 1.0414, + "step": 12566 + }, + { + "epoch": 2.23735754985755, + "grad_norm": 0.8676900267601013, + "learning_rate": 8.161452662940032e-05, + "loss": 1.0157, + "step": 12567 + }, + { + "epoch": 2.2375356125356127, + "grad_norm": 0.8174694180488586, + "learning_rate": 8.16007679463397e-05, + "loss": 0.691, + "step": 12568 + }, + { + "epoch": 2.237713675213675, + "grad_norm": 0.8137148022651672, + "learning_rate": 8.158700962377379e-05, + "loss": 1.0022, + "step": 12569 + }, + { + "epoch": 2.237891737891738, + "grad_norm": 0.970250129699707, + "learning_rate": 8.157325166197221e-05, + "loss": 0.7946, + "step": 12570 + }, + { + "epoch": 2.2380698005698005, + "grad_norm": 0.7366915941238403, + "learning_rate": 8.155949406120446e-05, + "loss": 0.9039, + "step": 12571 + }, + { + "epoch": 2.2382478632478633, + "grad_norm": 0.878358781337738, + "learning_rate": 8.154573682174014e-05, + "loss": 0.8172, + "step": 12572 + }, + { + "epoch": 2.238425925925926, + "grad_norm": 0.7552989721298218, + "learning_rate": 8.153197994384875e-05, + "loss": 0.9955, + "step": 12573 + }, + { + "epoch": 2.238603988603989, + "grad_norm": 0.8198257684707642, + "learning_rate": 8.151822342779985e-05, + "loss": 0.8677, + "step": 12574 + }, + { + "epoch": 2.238782051282051, + "grad_norm": 0.9128977656364441, + "learning_rate": 8.150446727386297e-05, + "loss": 0.9531, + "step": 12575 + }, + { + "epoch": 2.238960113960114, + "grad_norm": 0.867671549320221, + "learning_rate": 8.149071148230762e-05, + "loss": 1.0226, + "step": 12576 + }, + { + "epoch": 2.2391381766381766, + "grad_norm": 0.8640758395195007, + "learning_rate": 8.147695605340337e-05, + "loss": 1.1284, + "step": 12577 + }, + { + "epoch": 2.2393162393162394, + "grad_norm": 0.7453210353851318, + "learning_rate": 8.146320098741964e-05, + "loss": 0.7812, + "step": 12578 + }, + { + "epoch": 2.239494301994302, + "grad_norm": 0.9207521080970764, + "learning_rate": 8.144944628462602e-05, + "loss": 0.9955, + "step": 12579 + }, + { + "epoch": 2.239672364672365, + "grad_norm": 0.751732349395752, + "learning_rate": 8.143569194529193e-05, + "loss": 0.7858, + "step": 12580 + }, + { + "epoch": 2.239850427350427, + "grad_norm": 0.7955539226531982, + "learning_rate": 8.142193796968694e-05, + "loss": 0.8482, + "step": 12581 + }, + { + "epoch": 2.24002849002849, + "grad_norm": 0.8020164370536804, + "learning_rate": 8.140818435808043e-05, + "loss": 0.8069, + "step": 12582 + }, + { + "epoch": 2.2402065527065527, + "grad_norm": 0.7460235357284546, + "learning_rate": 8.139443111074198e-05, + "loss": 0.6478, + "step": 12583 + }, + { + "epoch": 2.2403846153846154, + "grad_norm": 0.7504379153251648, + "learning_rate": 8.138067822794096e-05, + "loss": 0.726, + "step": 12584 + }, + { + "epoch": 2.240562678062678, + "grad_norm": 0.8214267492294312, + "learning_rate": 8.136692570994688e-05, + "loss": 1.0114, + "step": 12585 + }, + { + "epoch": 2.240740740740741, + "grad_norm": 0.9436941742897034, + "learning_rate": 8.135317355702917e-05, + "loss": 0.873, + "step": 12586 + }, + { + "epoch": 2.2409188034188032, + "grad_norm": 0.7541804909706116, + "learning_rate": 8.133942176945733e-05, + "loss": 0.8013, + "step": 12587 + }, + { + "epoch": 2.241096866096866, + "grad_norm": 0.8725557327270508, + "learning_rate": 8.132567034750073e-05, + "loss": 0.8506, + "step": 12588 + }, + { + "epoch": 2.2412749287749287, + "grad_norm": 0.7766169905662537, + "learning_rate": 8.131191929142882e-05, + "loss": 0.9076, + "step": 12589 + }, + { + "epoch": 2.2414529914529915, + "grad_norm": 0.8852736353874207, + "learning_rate": 8.129816860151104e-05, + "loss": 0.9278, + "step": 12590 + }, + { + "epoch": 2.2416310541310542, + "grad_norm": 0.6939527988433838, + "learning_rate": 8.128441827801681e-05, + "loss": 0.753, + "step": 12591 + }, + { + "epoch": 2.241809116809117, + "grad_norm": 0.8932832479476929, + "learning_rate": 8.127066832121551e-05, + "loss": 0.8089, + "step": 12592 + }, + { + "epoch": 2.2419871794871793, + "grad_norm": 0.7399743795394897, + "learning_rate": 8.125691873137656e-05, + "loss": 0.6905, + "step": 12593 + }, + { + "epoch": 2.242165242165242, + "grad_norm": 0.7664098143577576, + "learning_rate": 8.124316950876933e-05, + "loss": 0.8698, + "step": 12594 + }, + { + "epoch": 2.242343304843305, + "grad_norm": 0.8222574591636658, + "learning_rate": 8.122942065366323e-05, + "loss": 0.8922, + "step": 12595 + }, + { + "epoch": 2.2425213675213675, + "grad_norm": 0.8072433471679688, + "learning_rate": 8.121567216632771e-05, + "loss": 0.7613, + "step": 12596 + }, + { + "epoch": 2.2426994301994303, + "grad_norm": 0.7647300362586975, + "learning_rate": 8.120192404703199e-05, + "loss": 0.8736, + "step": 12597 + }, + { + "epoch": 2.242877492877493, + "grad_norm": 0.7536396980285645, + "learning_rate": 8.118817629604559e-05, + "loss": 0.7697, + "step": 12598 + }, + { + "epoch": 2.2430555555555554, + "grad_norm": 0.7295291423797607, + "learning_rate": 8.117442891363774e-05, + "loss": 0.8477, + "step": 12599 + }, + { + "epoch": 2.243233618233618, + "grad_norm": 0.7677894830703735, + "learning_rate": 8.116068190007787e-05, + "loss": 0.8113, + "step": 12600 + }, + { + "epoch": 2.243411680911681, + "grad_norm": 0.825614869594574, + "learning_rate": 8.114693525563529e-05, + "loss": 0.915, + "step": 12601 + }, + { + "epoch": 2.2435897435897436, + "grad_norm": 0.7841798663139343, + "learning_rate": 8.113318898057939e-05, + "loss": 0.7028, + "step": 12602 + }, + { + "epoch": 2.2437678062678064, + "grad_norm": 1.085337519645691, + "learning_rate": 8.111944307517942e-05, + "loss": 0.8354, + "step": 12603 + }, + { + "epoch": 2.243945868945869, + "grad_norm": 0.7831527590751648, + "learning_rate": 8.110569753970475e-05, + "loss": 1.0275, + "step": 12604 + }, + { + "epoch": 2.2441239316239314, + "grad_norm": 0.800504744052887, + "learning_rate": 8.109195237442467e-05, + "loss": 0.7006, + "step": 12605 + }, + { + "epoch": 2.244301994301994, + "grad_norm": 0.8189738392829895, + "learning_rate": 8.107820757960856e-05, + "loss": 0.8036, + "step": 12606 + }, + { + "epoch": 2.244480056980057, + "grad_norm": 0.8892425298690796, + "learning_rate": 8.106446315552562e-05, + "loss": 0.8274, + "step": 12607 + }, + { + "epoch": 2.2446581196581197, + "grad_norm": 0.8144643306732178, + "learning_rate": 8.105071910244521e-05, + "loss": 1.0648, + "step": 12608 + }, + { + "epoch": 2.2448361823361824, + "grad_norm": 0.914513111114502, + "learning_rate": 8.103697542063657e-05, + "loss": 0.8999, + "step": 12609 + }, + { + "epoch": 2.245014245014245, + "grad_norm": 0.8273763656616211, + "learning_rate": 8.102323211036904e-05, + "loss": 0.8554, + "step": 12610 + }, + { + "epoch": 2.2451923076923075, + "grad_norm": 0.9459149837493896, + "learning_rate": 8.100948917191181e-05, + "loss": 1.2345, + "step": 12611 + }, + { + "epoch": 2.2453703703703702, + "grad_norm": 0.8377025723457336, + "learning_rate": 8.099574660553425e-05, + "loss": 0.8096, + "step": 12612 + }, + { + "epoch": 2.245548433048433, + "grad_norm": 0.8639607429504395, + "learning_rate": 8.098200441150551e-05, + "loss": 0.8238, + "step": 12613 + }, + { + "epoch": 2.2457264957264957, + "grad_norm": 1.0107637643814087, + "learning_rate": 8.09682625900949e-05, + "loss": 0.8747, + "step": 12614 + }, + { + "epoch": 2.2459045584045585, + "grad_norm": 0.8153043985366821, + "learning_rate": 8.095452114157164e-05, + "loss": 0.9357, + "step": 12615 + }, + { + "epoch": 2.2460826210826212, + "grad_norm": 0.8948562741279602, + "learning_rate": 8.094078006620497e-05, + "loss": 0.8245, + "step": 12616 + }, + { + "epoch": 2.246260683760684, + "grad_norm": 0.7983259558677673, + "learning_rate": 8.092703936426416e-05, + "loss": 0.8936, + "step": 12617 + }, + { + "epoch": 2.2464387464387463, + "grad_norm": 0.9016979336738586, + "learning_rate": 8.091329903601835e-05, + "loss": 1.0685, + "step": 12618 + }, + { + "epoch": 2.246616809116809, + "grad_norm": 0.7192493677139282, + "learning_rate": 8.089955908173685e-05, + "loss": 0.8622, + "step": 12619 + }, + { + "epoch": 2.246794871794872, + "grad_norm": 0.78288334608078, + "learning_rate": 8.088581950168877e-05, + "loss": 0.7874, + "step": 12620 + }, + { + "epoch": 2.2469729344729346, + "grad_norm": 0.8438683152198792, + "learning_rate": 8.087208029614336e-05, + "loss": 0.9262, + "step": 12621 + }, + { + "epoch": 2.2471509971509973, + "grad_norm": 0.8384907245635986, + "learning_rate": 8.085834146536978e-05, + "loss": 0.9069, + "step": 12622 + }, + { + "epoch": 2.2473290598290596, + "grad_norm": 0.8209545016288757, + "learning_rate": 8.084460300963729e-05, + "loss": 0.9457, + "step": 12623 + }, + { + "epoch": 2.2475071225071224, + "grad_norm": 0.8220782279968262, + "learning_rate": 8.083086492921496e-05, + "loss": 0.9224, + "step": 12624 + }, + { + "epoch": 2.247685185185185, + "grad_norm": 0.8927256464958191, + "learning_rate": 8.081712722437204e-05, + "loss": 0.7091, + "step": 12625 + }, + { + "epoch": 2.247863247863248, + "grad_norm": 0.8878564238548279, + "learning_rate": 8.080338989537764e-05, + "loss": 0.8879, + "step": 12626 + }, + { + "epoch": 2.2480413105413106, + "grad_norm": 0.8380948305130005, + "learning_rate": 8.078965294250097e-05, + "loss": 0.8504, + "step": 12627 + }, + { + "epoch": 2.2482193732193734, + "grad_norm": 0.8005350828170776, + "learning_rate": 8.07759163660111e-05, + "loss": 1.2119, + "step": 12628 + }, + { + "epoch": 2.248397435897436, + "grad_norm": 0.7990152835845947, + "learning_rate": 8.076218016617726e-05, + "loss": 0.72, + "step": 12629 + }, + { + "epoch": 2.2485754985754984, + "grad_norm": 0.9264963269233704, + "learning_rate": 8.07484443432685e-05, + "loss": 0.8398, + "step": 12630 + }, + { + "epoch": 2.248753561253561, + "grad_norm": 0.9103235602378845, + "learning_rate": 8.073470889755402e-05, + "loss": 0.9122, + "step": 12631 + }, + { + "epoch": 2.248931623931624, + "grad_norm": 0.8042106032371521, + "learning_rate": 8.072097382930285e-05, + "loss": 0.8065, + "step": 12632 + }, + { + "epoch": 2.2491096866096867, + "grad_norm": 0.8464857935905457, + "learning_rate": 8.070723913878421e-05, + "loss": 0.9117, + "step": 12633 + }, + { + "epoch": 2.2492877492877494, + "grad_norm": 0.7476474642753601, + "learning_rate": 8.06935048262671e-05, + "loss": 1.0252, + "step": 12634 + }, + { + "epoch": 2.2494658119658117, + "grad_norm": 0.8098256587982178, + "learning_rate": 8.067977089202065e-05, + "loss": 0.911, + "step": 12635 + }, + { + "epoch": 2.2496438746438745, + "grad_norm": 0.9311509728431702, + "learning_rate": 8.066603733631398e-05, + "loss": 0.9594, + "step": 12636 + }, + { + "epoch": 2.2496438746438745, + "eval_loss": 1.1335573196411133, + "eval_runtime": 24.2688, + "eval_samples_per_second": 42.895, + "eval_steps_per_second": 21.468, + "step": 12636 + }, + { + "epoch": 2.2498219373219372, + "grad_norm": 0.7744980454444885, + "learning_rate": 8.065230415941612e-05, + "loss": 0.8983, + "step": 12637 + }, + { + "epoch": 2.25, + "grad_norm": 0.9464056491851807, + "learning_rate": 8.06385713615962e-05, + "loss": 0.8646, + "step": 12638 + }, + { + "epoch": 2.2501780626780628, + "grad_norm": 0.8263896107673645, + "learning_rate": 8.062483894312323e-05, + "loss": 0.8557, + "step": 12639 + }, + { + "epoch": 2.2503561253561255, + "grad_norm": 0.8827885389328003, + "learning_rate": 8.06111069042663e-05, + "loss": 0.7632, + "step": 12640 + }, + { + "epoch": 2.2505341880341883, + "grad_norm": 0.8537881374359131, + "learning_rate": 8.059737524529443e-05, + "loss": 0.8004, + "step": 12641 + }, + { + "epoch": 2.2507122507122506, + "grad_norm": 0.8397842049598694, + "learning_rate": 8.058364396647674e-05, + "loss": 0.9487, + "step": 12642 + }, + { + "epoch": 2.2508903133903133, + "grad_norm": 1.071976661682129, + "learning_rate": 8.056991306808217e-05, + "loss": 1.0699, + "step": 12643 + }, + { + "epoch": 2.251068376068376, + "grad_norm": 0.8712023496627808, + "learning_rate": 8.055618255037983e-05, + "loss": 0.6518, + "step": 12644 + }, + { + "epoch": 2.251246438746439, + "grad_norm": 0.7885438799858093, + "learning_rate": 8.054245241363866e-05, + "loss": 0.8458, + "step": 12645 + }, + { + "epoch": 2.2514245014245016, + "grad_norm": 0.947169840335846, + "learning_rate": 8.052872265812774e-05, + "loss": 0.6631, + "step": 12646 + }, + { + "epoch": 2.251602564102564, + "grad_norm": 0.8554182052612305, + "learning_rate": 8.051499328411603e-05, + "loss": 0.8945, + "step": 12647 + }, + { + "epoch": 2.2517806267806266, + "grad_norm": 0.8081278800964355, + "learning_rate": 8.050126429187259e-05, + "loss": 0.8969, + "step": 12648 + }, + { + "epoch": 2.2519586894586894, + "grad_norm": 0.7826179265975952, + "learning_rate": 8.048753568166633e-05, + "loss": 0.6965, + "step": 12649 + }, + { + "epoch": 2.252136752136752, + "grad_norm": 0.9688517451286316, + "learning_rate": 8.04738074537663e-05, + "loss": 1.0044, + "step": 12650 + }, + { + "epoch": 2.252314814814815, + "grad_norm": 0.7780970931053162, + "learning_rate": 8.04600796084414e-05, + "loss": 0.8712, + "step": 12651 + }, + { + "epoch": 2.2524928774928776, + "grad_norm": 0.8360016942024231, + "learning_rate": 8.044635214596073e-05, + "loss": 0.9522, + "step": 12652 + }, + { + "epoch": 2.2526709401709404, + "grad_norm": 0.8137710094451904, + "learning_rate": 8.043262506659311e-05, + "loss": 0.7953, + "step": 12653 + }, + { + "epoch": 2.2528490028490027, + "grad_norm": 0.8394312858581543, + "learning_rate": 8.041889837060755e-05, + "loss": 0.77, + "step": 12654 + }, + { + "epoch": 2.2530270655270654, + "grad_norm": 0.7245169878005981, + "learning_rate": 8.040517205827307e-05, + "loss": 0.7657, + "step": 12655 + }, + { + "epoch": 2.253205128205128, + "grad_norm": 0.8018792271614075, + "learning_rate": 8.039144612985846e-05, + "loss": 0.8974, + "step": 12656 + }, + { + "epoch": 2.253383190883191, + "grad_norm": 0.8204617500305176, + "learning_rate": 8.037772058563278e-05, + "loss": 0.8635, + "step": 12657 + }, + { + "epoch": 2.2535612535612537, + "grad_norm": 0.906288743019104, + "learning_rate": 8.036399542586485e-05, + "loss": 1.0498, + "step": 12658 + }, + { + "epoch": 2.253739316239316, + "grad_norm": 0.8674196600914001, + "learning_rate": 8.035027065082371e-05, + "loss": 0.8621, + "step": 12659 + }, + { + "epoch": 2.2539173789173788, + "grad_norm": 0.8112890124320984, + "learning_rate": 8.033654626077816e-05, + "loss": 0.9937, + "step": 12660 + }, + { + "epoch": 2.2540954415954415, + "grad_norm": 0.8072839975357056, + "learning_rate": 8.032282225599714e-05, + "loss": 0.8555, + "step": 12661 + }, + { + "epoch": 2.2542735042735043, + "grad_norm": 0.7853979468345642, + "learning_rate": 8.030909863674952e-05, + "loss": 0.8698, + "step": 12662 + }, + { + "epoch": 2.254451566951567, + "grad_norm": 0.7456761598587036, + "learning_rate": 8.029537540330426e-05, + "loss": 0.6214, + "step": 12663 + }, + { + "epoch": 2.2546296296296298, + "grad_norm": 0.7207663059234619, + "learning_rate": 8.028165255593015e-05, + "loss": 0.7641, + "step": 12664 + }, + { + "epoch": 2.2548076923076925, + "grad_norm": 0.6541373133659363, + "learning_rate": 8.02679300948961e-05, + "loss": 0.7438, + "step": 12665 + }, + { + "epoch": 2.254985754985755, + "grad_norm": 0.7535310983657837, + "learning_rate": 8.025420802047096e-05, + "loss": 0.9417, + "step": 12666 + }, + { + "epoch": 2.2551638176638176, + "grad_norm": 0.88471919298172, + "learning_rate": 8.024048633292364e-05, + "loss": 0.9122, + "step": 12667 + }, + { + "epoch": 2.2553418803418803, + "grad_norm": 0.8621570467948914, + "learning_rate": 8.02267650325229e-05, + "loss": 0.7026, + "step": 12668 + }, + { + "epoch": 2.255519943019943, + "grad_norm": 0.8574202060699463, + "learning_rate": 8.021304411953767e-05, + "loss": 0.8997, + "step": 12669 + }, + { + "epoch": 2.255698005698006, + "grad_norm": 0.8038806915283203, + "learning_rate": 8.019932359423667e-05, + "loss": 0.9386, + "step": 12670 + }, + { + "epoch": 2.255876068376068, + "grad_norm": 0.7760711908340454, + "learning_rate": 8.018560345688883e-05, + "loss": 0.7777, + "step": 12671 + }, + { + "epoch": 2.256054131054131, + "grad_norm": 0.7433146834373474, + "learning_rate": 8.017188370776292e-05, + "loss": 0.7245, + "step": 12672 + }, + { + "epoch": 2.2562321937321936, + "grad_norm": 0.8710882067680359, + "learning_rate": 8.01581643471278e-05, + "loss": 0.9478, + "step": 12673 + }, + { + "epoch": 2.2564102564102564, + "grad_norm": 0.7726423740386963, + "learning_rate": 8.014444537525218e-05, + "loss": 0.8388, + "step": 12674 + }, + { + "epoch": 2.256588319088319, + "grad_norm": 0.7967063188552856, + "learning_rate": 8.01307267924049e-05, + "loss": 0.8226, + "step": 12675 + }, + { + "epoch": 2.256766381766382, + "grad_norm": 0.7524598836898804, + "learning_rate": 8.011700859885479e-05, + "loss": 0.7285, + "step": 12676 + }, + { + "epoch": 2.2569444444444446, + "grad_norm": 0.808729887008667, + "learning_rate": 8.010329079487055e-05, + "loss": 0.7498, + "step": 12677 + }, + { + "epoch": 2.257122507122507, + "grad_norm": 0.7842788100242615, + "learning_rate": 8.008957338072106e-05, + "loss": 0.9216, + "step": 12678 + }, + { + "epoch": 2.2573005698005697, + "grad_norm": 0.8905709981918335, + "learning_rate": 8.007585635667497e-05, + "loss": 0.9254, + "step": 12679 + }, + { + "epoch": 2.2574786324786325, + "grad_norm": 0.7495295405387878, + "learning_rate": 8.006213972300112e-05, + "loss": 0.8407, + "step": 12680 + }, + { + "epoch": 2.257656695156695, + "grad_norm": 0.7425774335861206, + "learning_rate": 8.004842347996819e-05, + "loss": 0.7893, + "step": 12681 + }, + { + "epoch": 2.257834757834758, + "grad_norm": 0.8028583526611328, + "learning_rate": 8.003470762784498e-05, + "loss": 0.8106, + "step": 12682 + }, + { + "epoch": 2.2580128205128207, + "grad_norm": 0.8874917030334473, + "learning_rate": 8.002099216690017e-05, + "loss": 0.97, + "step": 12683 + }, + { + "epoch": 2.258190883190883, + "grad_norm": 0.8830558061599731, + "learning_rate": 8.000727709740257e-05, + "loss": 1.028, + "step": 12684 + }, + { + "epoch": 2.2583689458689458, + "grad_norm": 0.8720272779464722, + "learning_rate": 7.99935624196208e-05, + "loss": 0.9401, + "step": 12685 + }, + { + "epoch": 2.2585470085470085, + "grad_norm": 0.736709713935852, + "learning_rate": 7.997984813382362e-05, + "loss": 0.8479, + "step": 12686 + }, + { + "epoch": 2.2587250712250713, + "grad_norm": 0.8028469085693359, + "learning_rate": 7.996613424027973e-05, + "loss": 0.9291, + "step": 12687 + }, + { + "epoch": 2.258903133903134, + "grad_norm": 0.777618944644928, + "learning_rate": 7.995242073925784e-05, + "loss": 0.7021, + "step": 12688 + }, + { + "epoch": 2.2590811965811968, + "grad_norm": 0.8371155261993408, + "learning_rate": 7.993870763102659e-05, + "loss": 0.8309, + "step": 12689 + }, + { + "epoch": 2.259259259259259, + "grad_norm": 0.8853654861450195, + "learning_rate": 7.992499491585473e-05, + "loss": 0.762, + "step": 12690 + }, + { + "epoch": 2.259437321937322, + "grad_norm": 0.742594301700592, + "learning_rate": 7.991128259401086e-05, + "loss": 0.8025, + "step": 12691 + }, + { + "epoch": 2.2596153846153846, + "grad_norm": 1.0678842067718506, + "learning_rate": 7.989757066576369e-05, + "loss": 0.9127, + "step": 12692 + }, + { + "epoch": 2.2597934472934473, + "grad_norm": 0.7917565703392029, + "learning_rate": 7.988385913138183e-05, + "loss": 0.8078, + "step": 12693 + }, + { + "epoch": 2.25997150997151, + "grad_norm": 0.6907288432121277, + "learning_rate": 7.987014799113397e-05, + "loss": 0.6313, + "step": 12694 + }, + { + "epoch": 2.260149572649573, + "grad_norm": 0.9007455706596375, + "learning_rate": 7.98564372452888e-05, + "loss": 1.0734, + "step": 12695 + }, + { + "epoch": 2.260327635327635, + "grad_norm": 0.7732774615287781, + "learning_rate": 7.984272689411484e-05, + "loss": 0.9925, + "step": 12696 + }, + { + "epoch": 2.260505698005698, + "grad_norm": 0.7470823526382446, + "learning_rate": 7.982901693788082e-05, + "loss": 0.8518, + "step": 12697 + }, + { + "epoch": 2.2606837606837606, + "grad_norm": 0.8018864989280701, + "learning_rate": 7.981530737685526e-05, + "loss": 0.8668, + "step": 12698 + }, + { + "epoch": 2.2608618233618234, + "grad_norm": 0.8459745049476624, + "learning_rate": 7.980159821130688e-05, + "loss": 0.8972, + "step": 12699 + }, + { + "epoch": 2.261039886039886, + "grad_norm": 0.8255595564842224, + "learning_rate": 7.978788944150419e-05, + "loss": 0.9562, + "step": 12700 + }, + { + "epoch": 2.261217948717949, + "grad_norm": 0.8243128061294556, + "learning_rate": 7.977418106771582e-05, + "loss": 0.6634, + "step": 12701 + }, + { + "epoch": 2.261396011396011, + "grad_norm": 0.802949845790863, + "learning_rate": 7.976047309021034e-05, + "loss": 0.8155, + "step": 12702 + }, + { + "epoch": 2.261574074074074, + "grad_norm": 0.8480857014656067, + "learning_rate": 7.97467655092564e-05, + "loss": 0.8568, + "step": 12703 + }, + { + "epoch": 2.2617521367521367, + "grad_norm": 0.8777545690536499, + "learning_rate": 7.973305832512247e-05, + "loss": 0.8688, + "step": 12704 + }, + { + "epoch": 2.2619301994301995, + "grad_norm": 0.8334060907363892, + "learning_rate": 7.971935153807719e-05, + "loss": 0.932, + "step": 12705 + }, + { + "epoch": 2.262108262108262, + "grad_norm": 0.836976170539856, + "learning_rate": 7.970564514838907e-05, + "loss": 0.8205, + "step": 12706 + }, + { + "epoch": 2.262286324786325, + "grad_norm": 0.782866895198822, + "learning_rate": 7.969193915632667e-05, + "loss": 0.8362, + "step": 12707 + }, + { + "epoch": 2.2624643874643873, + "grad_norm": 0.9018504619598389, + "learning_rate": 7.967823356215854e-05, + "loss": 0.8354, + "step": 12708 + }, + { + "epoch": 2.26264245014245, + "grad_norm": 0.7974916696548462, + "learning_rate": 7.966452836615324e-05, + "loss": 0.8035, + "step": 12709 + }, + { + "epoch": 2.2628205128205128, + "grad_norm": 0.8745712637901306, + "learning_rate": 7.965082356857922e-05, + "loss": 0.8803, + "step": 12710 + }, + { + "epoch": 2.2629985754985755, + "grad_norm": 0.8667176365852356, + "learning_rate": 7.963711916970505e-05, + "loss": 0.8005, + "step": 12711 + }, + { + "epoch": 2.2631766381766383, + "grad_norm": 0.849998950958252, + "learning_rate": 7.962341516979922e-05, + "loss": 0.8208, + "step": 12712 + }, + { + "epoch": 2.263354700854701, + "grad_norm": 0.803727388381958, + "learning_rate": 7.960971156913028e-05, + "loss": 0.7232, + "step": 12713 + }, + { + "epoch": 2.263532763532764, + "grad_norm": 0.842913031578064, + "learning_rate": 7.959600836796664e-05, + "loss": 0.8182, + "step": 12714 + }, + { + "epoch": 2.263710826210826, + "grad_norm": 0.8191903829574585, + "learning_rate": 7.958230556657684e-05, + "loss": 0.8353, + "step": 12715 + }, + { + "epoch": 2.263888888888889, + "grad_norm": 0.8525017499923706, + "learning_rate": 7.95686031652294e-05, + "loss": 0.9824, + "step": 12716 + }, + { + "epoch": 2.2640669515669516, + "grad_norm": 0.7176641225814819, + "learning_rate": 7.955490116419267e-05, + "loss": 0.7722, + "step": 12717 + }, + { + "epoch": 2.2642450142450143, + "grad_norm": 0.8740555047988892, + "learning_rate": 7.954119956373521e-05, + "loss": 0.8286, + "step": 12718 + }, + { + "epoch": 2.264423076923077, + "grad_norm": 0.7928949594497681, + "learning_rate": 7.952749836412543e-05, + "loss": 0.9183, + "step": 12719 + }, + { + "epoch": 2.2646011396011394, + "grad_norm": 0.787661612033844, + "learning_rate": 7.951379756563185e-05, + "loss": 0.7741, + "step": 12720 + }, + { + "epoch": 2.264779202279202, + "grad_norm": 0.8369856476783752, + "learning_rate": 7.950009716852277e-05, + "loss": 0.911, + "step": 12721 + }, + { + "epoch": 2.264957264957265, + "grad_norm": 0.7838568687438965, + "learning_rate": 7.948639717306675e-05, + "loss": 0.8532, + "step": 12722 + }, + { + "epoch": 2.2651353276353277, + "grad_norm": 0.8287179470062256, + "learning_rate": 7.947269757953213e-05, + "loss": 0.893, + "step": 12723 + }, + { + "epoch": 2.2653133903133904, + "grad_norm": 0.7754728198051453, + "learning_rate": 7.945899838818741e-05, + "loss": 0.9516, + "step": 12724 + }, + { + "epoch": 2.265491452991453, + "grad_norm": 0.7088906764984131, + "learning_rate": 7.94452995993009e-05, + "loss": 0.6797, + "step": 12725 + }, + { + "epoch": 2.265669515669516, + "grad_norm": 0.8004380464553833, + "learning_rate": 7.94316012131411e-05, + "loss": 0.8583, + "step": 12726 + }, + { + "epoch": 2.265847578347578, + "grad_norm": 0.8221408128738403, + "learning_rate": 7.941790322997629e-05, + "loss": 0.9575, + "step": 12727 + }, + { + "epoch": 2.266025641025641, + "grad_norm": 0.8640061020851135, + "learning_rate": 7.940420565007492e-05, + "loss": 0.9471, + "step": 12728 + }, + { + "epoch": 2.2662037037037037, + "grad_norm": 0.8151915669441223, + "learning_rate": 7.939050847370536e-05, + "loss": 0.7841, + "step": 12729 + }, + { + "epoch": 2.2663817663817665, + "grad_norm": 0.7910612225532532, + "learning_rate": 7.9376811701136e-05, + "loss": 0.8826, + "step": 12730 + }, + { + "epoch": 2.2665598290598292, + "grad_norm": 0.7158875465393066, + "learning_rate": 7.936311533263514e-05, + "loss": 0.8598, + "step": 12731 + }, + { + "epoch": 2.2667378917378915, + "grad_norm": 0.6968050003051758, + "learning_rate": 7.934941936847119e-05, + "loss": 0.742, + "step": 12732 + }, + { + "epoch": 2.2669159544159543, + "grad_norm": 0.8630516529083252, + "learning_rate": 7.933572380891245e-05, + "loss": 0.789, + "step": 12733 + }, + { + "epoch": 2.267094017094017, + "grad_norm": 0.8060851097106934, + "learning_rate": 7.932202865422726e-05, + "loss": 0.8447, + "step": 12734 + }, + { + "epoch": 2.26727207977208, + "grad_norm": 0.9570813775062561, + "learning_rate": 7.930833390468402e-05, + "loss": 0.8948, + "step": 12735 + }, + { + "epoch": 2.2674501424501425, + "grad_norm": 0.7649935483932495, + "learning_rate": 7.929463956055094e-05, + "loss": 0.905, + "step": 12736 + }, + { + "epoch": 2.2676282051282053, + "grad_norm": 0.7498226165771484, + "learning_rate": 7.928094562209641e-05, + "loss": 0.9168, + "step": 12737 + }, + { + "epoch": 2.267806267806268, + "grad_norm": 0.7915979027748108, + "learning_rate": 7.926725208958869e-05, + "loss": 0.8628, + "step": 12738 + }, + { + "epoch": 2.2679843304843303, + "grad_norm": 0.7620252966880798, + "learning_rate": 7.925355896329615e-05, + "loss": 0.8768, + "step": 12739 + }, + { + "epoch": 2.268162393162393, + "grad_norm": 0.9785344004631042, + "learning_rate": 7.923986624348697e-05, + "loss": 0.9579, + "step": 12740 + }, + { + "epoch": 2.268340455840456, + "grad_norm": 0.9146337509155273, + "learning_rate": 7.922617393042954e-05, + "loss": 1.2241, + "step": 12741 + }, + { + "epoch": 2.2685185185185186, + "grad_norm": 0.7815660238265991, + "learning_rate": 7.921248202439203e-05, + "loss": 0.7367, + "step": 12742 + }, + { + "epoch": 2.2686965811965814, + "grad_norm": 0.8466008305549622, + "learning_rate": 7.919879052564276e-05, + "loss": 0.923, + "step": 12743 + }, + { + "epoch": 2.2688746438746437, + "grad_norm": 0.742203950881958, + "learning_rate": 7.918509943444998e-05, + "loss": 0.8282, + "step": 12744 + }, + { + "epoch": 2.2690527065527064, + "grad_norm": 0.785446286201477, + "learning_rate": 7.917140875108196e-05, + "loss": 0.7689, + "step": 12745 + }, + { + "epoch": 2.269230769230769, + "grad_norm": 0.912765383720398, + "learning_rate": 7.915771847580689e-05, + "loss": 0.8259, + "step": 12746 + }, + { + "epoch": 2.269408831908832, + "grad_norm": 0.7319221496582031, + "learning_rate": 7.914402860889305e-05, + "loss": 0.8861, + "step": 12747 + }, + { + "epoch": 2.2695868945868947, + "grad_norm": 1.0215193033218384, + "learning_rate": 7.913033915060861e-05, + "loss": 0.8556, + "step": 12748 + }, + { + "epoch": 2.2697649572649574, + "grad_norm": 0.9348630905151367, + "learning_rate": 7.911665010122188e-05, + "loss": 1.0269, + "step": 12749 + }, + { + "epoch": 2.26994301994302, + "grad_norm": 0.7521753311157227, + "learning_rate": 7.910296146100096e-05, + "loss": 0.8492, + "step": 12750 + }, + { + "epoch": 2.2701210826210825, + "grad_norm": 0.7274978756904602, + "learning_rate": 7.908927323021414e-05, + "loss": 0.7029, + "step": 12751 + }, + { + "epoch": 2.2702991452991452, + "grad_norm": 0.8103266954421997, + "learning_rate": 7.907558540912954e-05, + "loss": 0.5268, + "step": 12752 + }, + { + "epoch": 2.270477207977208, + "grad_norm": 0.8645551800727844, + "learning_rate": 7.906189799801538e-05, + "loss": 0.8172, + "step": 12753 + }, + { + "epoch": 2.2706552706552707, + "grad_norm": 0.8652981519699097, + "learning_rate": 7.904821099713984e-05, + "loss": 0.8711, + "step": 12754 + }, + { + "epoch": 2.2708333333333335, + "grad_norm": 0.7020241618156433, + "learning_rate": 7.903452440677106e-05, + "loss": 0.7202, + "step": 12755 + }, + { + "epoch": 2.271011396011396, + "grad_norm": 0.9812583923339844, + "learning_rate": 7.902083822717727e-05, + "loss": 0.8274, + "step": 12756 + }, + { + "epoch": 2.2711894586894585, + "grad_norm": 0.9119269847869873, + "learning_rate": 7.900715245862655e-05, + "loss": 0.8695, + "step": 12757 + }, + { + "epoch": 2.2713675213675213, + "grad_norm": 0.7336047291755676, + "learning_rate": 7.899346710138706e-05, + "loss": 0.8138, + "step": 12758 + }, + { + "epoch": 2.271545584045584, + "grad_norm": 0.900337278842926, + "learning_rate": 7.897978215572695e-05, + "loss": 0.8346, + "step": 12759 + }, + { + "epoch": 2.271723646723647, + "grad_norm": 0.736018717288971, + "learning_rate": 7.896609762191437e-05, + "loss": 0.7045, + "step": 12760 + }, + { + "epoch": 2.2719017094017095, + "grad_norm": 0.8484935760498047, + "learning_rate": 7.895241350021737e-05, + "loss": 1.05, + "step": 12761 + }, + { + "epoch": 2.2720797720797723, + "grad_norm": 0.8032337427139282, + "learning_rate": 7.893872979090415e-05, + "loss": 0.8024, + "step": 12762 + }, + { + "epoch": 2.2722578347578346, + "grad_norm": 0.8957629203796387, + "learning_rate": 7.892504649424272e-05, + "loss": 0.9593, + "step": 12763 + }, + { + "epoch": 2.2724358974358974, + "grad_norm": 0.9227191805839539, + "learning_rate": 7.891136361050126e-05, + "loss": 0.9978, + "step": 12764 + }, + { + "epoch": 2.27261396011396, + "grad_norm": 0.8649391531944275, + "learning_rate": 7.88976811399478e-05, + "loss": 0.8525, + "step": 12765 + }, + { + "epoch": 2.272792022792023, + "grad_norm": 0.8762859106063843, + "learning_rate": 7.888399908285046e-05, + "loss": 0.9526, + "step": 12766 + }, + { + "epoch": 2.2729700854700856, + "grad_norm": 0.8566350340843201, + "learning_rate": 7.887031743947729e-05, + "loss": 0.7886, + "step": 12767 + }, + { + "epoch": 2.273148148148148, + "grad_norm": 0.9285386800765991, + "learning_rate": 7.885663621009634e-05, + "loss": 1.013, + "step": 12768 + }, + { + "epoch": 2.2733262108262107, + "grad_norm": 0.9326284527778625, + "learning_rate": 7.884295539497569e-05, + "loss": 0.9908, + "step": 12769 + }, + { + "epoch": 2.2735042735042734, + "grad_norm": 0.8035810589790344, + "learning_rate": 7.882927499438341e-05, + "loss": 0.7452, + "step": 12770 + }, + { + "epoch": 2.273682336182336, + "grad_norm": 0.831741988658905, + "learning_rate": 7.881559500858747e-05, + "loss": 0.8782, + "step": 12771 + }, + { + "epoch": 2.273860398860399, + "grad_norm": 0.7790034413337708, + "learning_rate": 7.880191543785594e-05, + "loss": 0.9494, + "step": 12772 + }, + { + "epoch": 2.2740384615384617, + "grad_norm": 0.7070899605751038, + "learning_rate": 7.878823628245684e-05, + "loss": 0.7007, + "step": 12773 + }, + { + "epoch": 2.2742165242165244, + "grad_norm": 0.739573061466217, + "learning_rate": 7.877455754265818e-05, + "loss": 0.758, + "step": 12774 + }, + { + "epoch": 2.2743945868945867, + "grad_norm": 1.091391921043396, + "learning_rate": 7.876087921872803e-05, + "loss": 1.1333, + "step": 12775 + }, + { + "epoch": 2.2745726495726495, + "grad_norm": 0.623710036277771, + "learning_rate": 7.874720131093427e-05, + "loss": 0.7068, + "step": 12776 + }, + { + "epoch": 2.2747507122507122, + "grad_norm": 0.7989393472671509, + "learning_rate": 7.8733523819545e-05, + "loss": 0.77, + "step": 12777 + }, + { + "epoch": 2.274928774928775, + "grad_norm": 0.8401352167129517, + "learning_rate": 7.87198467448281e-05, + "loss": 0.8192, + "step": 12778 + }, + { + "epoch": 2.2751068376068377, + "grad_norm": 0.7962843179702759, + "learning_rate": 7.870617008705164e-05, + "loss": 0.8071, + "step": 12779 + }, + { + "epoch": 2.2752849002849, + "grad_norm": 0.9518889784812927, + "learning_rate": 7.869249384648351e-05, + "loss": 0.8956, + "step": 12780 + }, + { + "epoch": 2.275462962962963, + "grad_norm": 0.7469878792762756, + "learning_rate": 7.867881802339175e-05, + "loss": 0.6816, + "step": 12781 + }, + { + "epoch": 2.2756410256410255, + "grad_norm": 0.8888431191444397, + "learning_rate": 7.866514261804421e-05, + "loss": 0.7906, + "step": 12782 + }, + { + "epoch": 2.2758190883190883, + "grad_norm": 0.9856036305427551, + "learning_rate": 7.86514676307089e-05, + "loss": 0.9371, + "step": 12783 + }, + { + "epoch": 2.275997150997151, + "grad_norm": 0.9144912958145142, + "learning_rate": 7.863779306165371e-05, + "loss": 0.9613, + "step": 12784 + }, + { + "epoch": 2.276175213675214, + "grad_norm": 0.7898108959197998, + "learning_rate": 7.862411891114665e-05, + "loss": 0.8631, + "step": 12785 + }, + { + "epoch": 2.2763532763532766, + "grad_norm": 0.8524056077003479, + "learning_rate": 7.861044517945552e-05, + "loss": 0.7011, + "step": 12786 + }, + { + "epoch": 2.276531339031339, + "grad_norm": 0.8811307549476624, + "learning_rate": 7.859677186684831e-05, + "loss": 0.8138, + "step": 12787 + }, + { + "epoch": 2.2767094017094016, + "grad_norm": 0.9247646331787109, + "learning_rate": 7.858309897359289e-05, + "loss": 0.8971, + "step": 12788 + }, + { + "epoch": 2.2768874643874644, + "grad_norm": 0.8655884861946106, + "learning_rate": 7.856942649995715e-05, + "loss": 0.885, + "step": 12789 + }, + { + "epoch": 2.277065527065527, + "grad_norm": 0.9330910444259644, + "learning_rate": 7.855575444620897e-05, + "loss": 0.8493, + "step": 12790 + }, + { + "epoch": 2.27724358974359, + "grad_norm": 0.746694028377533, + "learning_rate": 7.854208281261626e-05, + "loss": 0.7169, + "step": 12791 + }, + { + "epoch": 2.277421652421652, + "grad_norm": 0.9785143136978149, + "learning_rate": 7.852841159944685e-05, + "loss": 1.023, + "step": 12792 + }, + { + "epoch": 2.277599715099715, + "grad_norm": 0.6107021570205688, + "learning_rate": 7.851474080696859e-05, + "loss": 0.652, + "step": 12793 + }, + { + "epoch": 2.2777777777777777, + "grad_norm": 0.9269224405288696, + "learning_rate": 7.850107043544937e-05, + "loss": 0.8024, + "step": 12794 + }, + { + "epoch": 2.2779558404558404, + "grad_norm": 0.8488328456878662, + "learning_rate": 7.8487400485157e-05, + "loss": 0.8626, + "step": 12795 + }, + { + "epoch": 2.278133903133903, + "grad_norm": 0.7187852263450623, + "learning_rate": 7.847373095635937e-05, + "loss": 0.7416, + "step": 12796 + }, + { + "epoch": 2.278311965811966, + "grad_norm": 1.00519859790802, + "learning_rate": 7.846006184932422e-05, + "loss": 1.0577, + "step": 12797 + }, + { + "epoch": 2.2784900284900287, + "grad_norm": 0.8175216913223267, + "learning_rate": 7.844639316431945e-05, + "loss": 0.9685, + "step": 12798 + }, + { + "epoch": 2.278668091168091, + "grad_norm": 0.8239067792892456, + "learning_rate": 7.843272490161281e-05, + "loss": 0.8714, + "step": 12799 + }, + { + "epoch": 2.2788461538461537, + "grad_norm": 0.8089447617530823, + "learning_rate": 7.841905706147212e-05, + "loss": 0.8397, + "step": 12800 + }, + { + "epoch": 2.2790242165242165, + "grad_norm": 0.8505867719650269, + "learning_rate": 7.840538964416518e-05, + "loss": 0.6872, + "step": 12801 + }, + { + "epoch": 2.2792022792022792, + "grad_norm": 0.8512473702430725, + "learning_rate": 7.83917226499598e-05, + "loss": 1.0422, + "step": 12802 + }, + { + "epoch": 2.279380341880342, + "grad_norm": 0.8935198187828064, + "learning_rate": 7.837805607912369e-05, + "loss": 0.9874, + "step": 12803 + }, + { + "epoch": 2.2795584045584047, + "grad_norm": 0.6903477907180786, + "learning_rate": 7.836438993192466e-05, + "loss": 0.7301, + "step": 12804 + }, + { + "epoch": 2.279736467236467, + "grad_norm": 0.7140037417411804, + "learning_rate": 7.835072420863046e-05, + "loss": 0.8323, + "step": 12805 + }, + { + "epoch": 2.27991452991453, + "grad_norm": 0.7974498867988586, + "learning_rate": 7.833705890950888e-05, + "loss": 0.7784, + "step": 12806 + }, + { + "epoch": 2.2800925925925926, + "grad_norm": 0.8191199898719788, + "learning_rate": 7.83233940348276e-05, + "loss": 0.8584, + "step": 12807 + }, + { + "epoch": 2.2802706552706553, + "grad_norm": 0.843112587928772, + "learning_rate": 7.83097295848544e-05, + "loss": 1.043, + "step": 12808 + }, + { + "epoch": 2.280448717948718, + "grad_norm": 0.8029288053512573, + "learning_rate": 7.829606555985698e-05, + "loss": 0.6806, + "step": 12809 + }, + { + "epoch": 2.280626780626781, + "grad_norm": 0.712228536605835, + "learning_rate": 7.828240196010311e-05, + "loss": 0.7748, + "step": 12810 + }, + { + "epoch": 2.280804843304843, + "grad_norm": 0.801659882068634, + "learning_rate": 7.82687387858604e-05, + "loss": 0.9374, + "step": 12811 + }, + { + "epoch": 2.280982905982906, + "grad_norm": 0.8457205891609192, + "learning_rate": 7.825507603739666e-05, + "loss": 0.9453, + "step": 12812 + }, + { + "epoch": 2.2811609686609686, + "grad_norm": 0.9129060506820679, + "learning_rate": 7.824141371497948e-05, + "loss": 0.9324, + "step": 12813 + }, + { + "epoch": 2.2813390313390314, + "grad_norm": 0.947914183139801, + "learning_rate": 7.822775181887663e-05, + "loss": 0.8275, + "step": 12814 + }, + { + "epoch": 2.281517094017094, + "grad_norm": 0.7204358577728271, + "learning_rate": 7.821409034935576e-05, + "loss": 0.6538, + "step": 12815 + }, + { + "epoch": 2.281695156695157, + "grad_norm": 0.8021003603935242, + "learning_rate": 7.82004293066845e-05, + "loss": 1.0464, + "step": 12816 + }, + { + "epoch": 2.281873219373219, + "grad_norm": 0.9530314803123474, + "learning_rate": 7.818676869113059e-05, + "loss": 0.8854, + "step": 12817 + }, + { + "epoch": 2.282051282051282, + "grad_norm": 0.7932098507881165, + "learning_rate": 7.81731085029616e-05, + "loss": 0.8219, + "step": 12818 + }, + { + "epoch": 2.2822293447293447, + "grad_norm": 0.7955925464630127, + "learning_rate": 7.815944874244523e-05, + "loss": 0.801, + "step": 12819 + }, + { + "epoch": 2.2824074074074074, + "grad_norm": 0.8490158915519714, + "learning_rate": 7.814578940984907e-05, + "loss": 0.8666, + "step": 12820 + }, + { + "epoch": 2.28258547008547, + "grad_norm": 0.7325232028961182, + "learning_rate": 7.813213050544081e-05, + "loss": 0.9579, + "step": 12821 + }, + { + "epoch": 2.282763532763533, + "grad_norm": 0.9203488230705261, + "learning_rate": 7.811847202948798e-05, + "loss": 1.0581, + "step": 12822 + }, + { + "epoch": 2.2829415954415953, + "grad_norm": 0.8207429647445679, + "learning_rate": 7.810481398225827e-05, + "loss": 0.8613, + "step": 12823 + }, + { + "epoch": 2.283119658119658, + "grad_norm": 0.872207522392273, + "learning_rate": 7.809115636401921e-05, + "loss": 0.9155, + "step": 12824 + }, + { + "epoch": 2.2832977207977208, + "grad_norm": 0.8032099604606628, + "learning_rate": 7.807749917503845e-05, + "loss": 0.8294, + "step": 12825 + }, + { + "epoch": 2.2834757834757835, + "grad_norm": 0.8824980854988098, + "learning_rate": 7.806384241558354e-05, + "loss": 0.8618, + "step": 12826 + }, + { + "epoch": 2.2836538461538463, + "grad_norm": 0.9057566523551941, + "learning_rate": 7.805018608592212e-05, + "loss": 0.826, + "step": 12827 + }, + { + "epoch": 2.283831908831909, + "grad_norm": 0.8092000484466553, + "learning_rate": 7.803653018632164e-05, + "loss": 0.8091, + "step": 12828 + }, + { + "epoch": 2.2840099715099713, + "grad_norm": 0.8372754454612732, + "learning_rate": 7.802287471704976e-05, + "loss": 1.108, + "step": 12829 + }, + { + "epoch": 2.284188034188034, + "grad_norm": 0.8702181577682495, + "learning_rate": 7.800921967837398e-05, + "loss": 0.9654, + "step": 12830 + }, + { + "epoch": 2.284366096866097, + "grad_norm": 0.9543859958648682, + "learning_rate": 7.79955650705619e-05, + "loss": 0.9268, + "step": 12831 + }, + { + "epoch": 2.2845441595441596, + "grad_norm": 0.7992038726806641, + "learning_rate": 7.798191089388096e-05, + "loss": 0.8299, + "step": 12832 + }, + { + "epoch": 2.2847222222222223, + "grad_norm": 0.8655165433883667, + "learning_rate": 7.796825714859874e-05, + "loss": 0.9656, + "step": 12833 + }, + { + "epoch": 2.284900284900285, + "grad_norm": 0.9013311862945557, + "learning_rate": 7.795460383498281e-05, + "loss": 0.9373, + "step": 12834 + }, + { + "epoch": 2.285078347578348, + "grad_norm": 0.8453806638717651, + "learning_rate": 7.794095095330058e-05, + "loss": 0.7711, + "step": 12835 + }, + { + "epoch": 2.28525641025641, + "grad_norm": 0.8016965985298157, + "learning_rate": 7.792729850381959e-05, + "loss": 0.7492, + "step": 12836 + }, + { + "epoch": 2.285434472934473, + "grad_norm": 0.7191343307495117, + "learning_rate": 7.791364648680734e-05, + "loss": 0.7541, + "step": 12837 + }, + { + "epoch": 2.2856125356125356, + "grad_norm": 0.8210958242416382, + "learning_rate": 7.789999490253133e-05, + "loss": 0.7448, + "step": 12838 + }, + { + "epoch": 2.2857905982905984, + "grad_norm": 0.904022216796875, + "learning_rate": 7.788634375125898e-05, + "loss": 1.0329, + "step": 12839 + }, + { + "epoch": 2.285968660968661, + "grad_norm": 0.8934714794158936, + "learning_rate": 7.787269303325779e-05, + "loss": 0.8982, + "step": 12840 + }, + { + "epoch": 2.2861467236467234, + "grad_norm": 0.9424307942390442, + "learning_rate": 7.785904274879521e-05, + "loss": 1.0298, + "step": 12841 + }, + { + "epoch": 2.286324786324786, + "grad_norm": 0.7753969430923462, + "learning_rate": 7.784539289813873e-05, + "loss": 0.7811, + "step": 12842 + }, + { + "epoch": 2.286502849002849, + "grad_norm": 0.7851901054382324, + "learning_rate": 7.78317434815557e-05, + "loss": 0.8395, + "step": 12843 + }, + { + "epoch": 2.2866809116809117, + "grad_norm": 0.7734000086784363, + "learning_rate": 7.781809449931365e-05, + "loss": 0.6572, + "step": 12844 + }, + { + "epoch": 2.2868589743589745, + "grad_norm": 0.8322952389717102, + "learning_rate": 7.780444595167992e-05, + "loss": 0.9707, + "step": 12845 + }, + { + "epoch": 2.287037037037037, + "grad_norm": 0.8243176341056824, + "learning_rate": 7.779079783892203e-05, + "loss": 0.8413, + "step": 12846 + }, + { + "epoch": 2.2872150997151, + "grad_norm": 0.8600375056266785, + "learning_rate": 7.777715016130727e-05, + "loss": 0.8471, + "step": 12847 + }, + { + "epoch": 2.2873931623931623, + "grad_norm": 0.9846388101577759, + "learning_rate": 7.776350291910311e-05, + "loss": 1.0187, + "step": 12848 + }, + { + "epoch": 2.287571225071225, + "grad_norm": 0.8445034623146057, + "learning_rate": 7.774985611257688e-05, + "loss": 0.9113, + "step": 12849 + }, + { + "epoch": 2.2877492877492878, + "grad_norm": 0.804595947265625, + "learning_rate": 7.773620974199604e-05, + "loss": 0.8331, + "step": 12850 + }, + { + "epoch": 2.2879273504273505, + "grad_norm": 0.7600802779197693, + "learning_rate": 7.772256380762789e-05, + "loss": 0.8448, + "step": 12851 + }, + { + "epoch": 2.2881054131054133, + "grad_norm": 0.7406377792358398, + "learning_rate": 7.770891830973984e-05, + "loss": 0.7904, + "step": 12852 + }, + { + "epoch": 2.2882834757834756, + "grad_norm": 0.7294487357139587, + "learning_rate": 7.769527324859924e-05, + "loss": 0.8799, + "step": 12853 + }, + { + "epoch": 2.2884615384615383, + "grad_norm": 0.8864750266075134, + "learning_rate": 7.768162862447342e-05, + "loss": 0.9038, + "step": 12854 + }, + { + "epoch": 2.288639601139601, + "grad_norm": 0.8933553099632263, + "learning_rate": 7.766798443762972e-05, + "loss": 0.929, + "step": 12855 + }, + { + "epoch": 2.288817663817664, + "grad_norm": 0.8065192103385925, + "learning_rate": 7.765434068833545e-05, + "loss": 0.9335, + "step": 12856 + }, + { + "epoch": 2.2889957264957266, + "grad_norm": 0.8644578456878662, + "learning_rate": 7.764069737685802e-05, + "loss": 0.7717, + "step": 12857 + }, + { + "epoch": 2.2891737891737893, + "grad_norm": 0.8957899212837219, + "learning_rate": 7.762705450346462e-05, + "loss": 0.8625, + "step": 12858 + }, + { + "epoch": 2.289351851851852, + "grad_norm": 0.7164827585220337, + "learning_rate": 7.761341206842265e-05, + "loss": 0.8018, + "step": 12859 + }, + { + "epoch": 2.2895299145299144, + "grad_norm": 0.8752971291542053, + "learning_rate": 7.759977007199933e-05, + "loss": 0.8517, + "step": 12860 + }, + { + "epoch": 2.289707977207977, + "grad_norm": 0.8448139429092407, + "learning_rate": 7.758612851446201e-05, + "loss": 1.0449, + "step": 12861 + }, + { + "epoch": 2.28988603988604, + "grad_norm": 0.81675785779953, + "learning_rate": 7.75724873960779e-05, + "loss": 1.0952, + "step": 12862 + }, + { + "epoch": 2.2900641025641026, + "grad_norm": 0.8215656876564026, + "learning_rate": 7.755884671711437e-05, + "loss": 0.8419, + "step": 12863 + }, + { + "epoch": 2.2902421652421654, + "grad_norm": 0.8270167708396912, + "learning_rate": 7.754520647783859e-05, + "loss": 0.9065, + "step": 12864 + }, + { + "epoch": 2.2904202279202277, + "grad_norm": 0.8222723603248596, + "learning_rate": 7.753156667851784e-05, + "loss": 0.8536, + "step": 12865 + }, + { + "epoch": 2.2905982905982905, + "grad_norm": 0.8383764028549194, + "learning_rate": 7.751792731941936e-05, + "loss": 0.8829, + "step": 12866 + }, + { + "epoch": 2.290776353276353, + "grad_norm": 0.8115772008895874, + "learning_rate": 7.750428840081043e-05, + "loss": 0.8969, + "step": 12867 + }, + { + "epoch": 2.290954415954416, + "grad_norm": 0.8721897602081299, + "learning_rate": 7.74906499229582e-05, + "loss": 1.031, + "step": 12868 + }, + { + "epoch": 2.2911324786324787, + "grad_norm": 0.6958467364311218, + "learning_rate": 7.747701188612996e-05, + "loss": 0.7528, + "step": 12869 + }, + { + "epoch": 2.2913105413105415, + "grad_norm": 0.8352338671684265, + "learning_rate": 7.746337429059285e-05, + "loss": 0.9297, + "step": 12870 + }, + { + "epoch": 2.291488603988604, + "grad_norm": 0.8407408595085144, + "learning_rate": 7.744973713661411e-05, + "loss": 0.8209, + "step": 12871 + }, + { + "epoch": 2.2916666666666665, + "grad_norm": 0.9509777426719666, + "learning_rate": 7.743610042446092e-05, + "loss": 0.9408, + "step": 12872 + }, + { + "epoch": 2.2918447293447293, + "grad_norm": 0.7913112640380859, + "learning_rate": 7.742246415440048e-05, + "loss": 0.9063, + "step": 12873 + }, + { + "epoch": 2.292022792022792, + "grad_norm": 0.90866619348526, + "learning_rate": 7.740882832669998e-05, + "loss": 1.0178, + "step": 12874 + }, + { + "epoch": 2.2922008547008548, + "grad_norm": 0.5832980871200562, + "learning_rate": 7.739519294162652e-05, + "loss": 0.4138, + "step": 12875 + }, + { + "epoch": 2.2923789173789175, + "grad_norm": 0.717993974685669, + "learning_rate": 7.738155799944732e-05, + "loss": 0.7303, + "step": 12876 + }, + { + "epoch": 2.29255698005698, + "grad_norm": 0.7821396589279175, + "learning_rate": 7.736792350042948e-05, + "loss": 0.829, + "step": 12877 + }, + { + "epoch": 2.2927350427350426, + "grad_norm": 0.8877809047698975, + "learning_rate": 7.735428944484021e-05, + "loss": 0.8883, + "step": 12878 + }, + { + "epoch": 2.2929131054131053, + "grad_norm": 0.7754776477813721, + "learning_rate": 7.734065583294656e-05, + "loss": 0.807, + "step": 12879 + }, + { + "epoch": 2.293091168091168, + "grad_norm": 0.851157009601593, + "learning_rate": 7.73270226650157e-05, + "loss": 0.8859, + "step": 12880 + }, + { + "epoch": 2.293269230769231, + "grad_norm": 0.7635365128517151, + "learning_rate": 7.731338994131472e-05, + "loss": 0.9796, + "step": 12881 + }, + { + "epoch": 2.2934472934472936, + "grad_norm": 0.8386050462722778, + "learning_rate": 7.729975766211078e-05, + "loss": 0.788, + "step": 12882 + }, + { + "epoch": 2.2936253561253563, + "grad_norm": 0.7092825174331665, + "learning_rate": 7.728612582767088e-05, + "loss": 0.6855, + "step": 12883 + }, + { + "epoch": 2.2938034188034186, + "grad_norm": 0.8651222586631775, + "learning_rate": 7.72724944382622e-05, + "loss": 0.8875, + "step": 12884 + }, + { + "epoch": 2.2939814814814814, + "grad_norm": 0.89743572473526, + "learning_rate": 7.725886349415175e-05, + "loss": 0.9256, + "step": 12885 + }, + { + "epoch": 2.294159544159544, + "grad_norm": 0.8257600665092468, + "learning_rate": 7.724523299560664e-05, + "loss": 0.6703, + "step": 12886 + }, + { + "epoch": 2.294337606837607, + "grad_norm": 0.8133751153945923, + "learning_rate": 7.72316029428939e-05, + "loss": 0.8991, + "step": 12887 + }, + { + "epoch": 2.2945156695156697, + "grad_norm": 0.7874962687492371, + "learning_rate": 7.721797333628065e-05, + "loss": 0.8679, + "step": 12888 + }, + { + "epoch": 2.294693732193732, + "grad_norm": 0.8284404277801514, + "learning_rate": 7.720434417603384e-05, + "loss": 0.873, + "step": 12889 + }, + { + "epoch": 2.2948717948717947, + "grad_norm": 0.8751698136329651, + "learning_rate": 7.719071546242058e-05, + "loss": 1.0671, + "step": 12890 + }, + { + "epoch": 2.2950498575498575, + "grad_norm": 0.9355120062828064, + "learning_rate": 7.717708719570784e-05, + "loss": 0.93, + "step": 12891 + }, + { + "epoch": 2.29522792022792, + "grad_norm": 0.8643141984939575, + "learning_rate": 7.716345937616267e-05, + "loss": 0.7635, + "step": 12892 + }, + { + "epoch": 2.295405982905983, + "grad_norm": 0.9343852996826172, + "learning_rate": 7.714983200405212e-05, + "loss": 1.0624, + "step": 12893 + }, + { + "epoch": 2.2955840455840457, + "grad_norm": 0.893825352191925, + "learning_rate": 7.71362050796431e-05, + "loss": 1.0843, + "step": 12894 + }, + { + "epoch": 2.2957621082621085, + "grad_norm": 0.920723021030426, + "learning_rate": 7.712257860320269e-05, + "loss": 0.9681, + "step": 12895 + }, + { + "epoch": 2.2959401709401708, + "grad_norm": 0.9275181293487549, + "learning_rate": 7.710895257499778e-05, + "loss": 0.8904, + "step": 12896 + }, + { + "epoch": 2.2961182336182335, + "grad_norm": 0.9343428611755371, + "learning_rate": 7.709532699529543e-05, + "loss": 0.9338, + "step": 12897 + }, + { + "epoch": 2.2962962962962963, + "grad_norm": 0.7457774877548218, + "learning_rate": 7.708170186436252e-05, + "loss": 0.6521, + "step": 12898 + }, + { + "epoch": 2.296474358974359, + "grad_norm": 0.7977834343910217, + "learning_rate": 7.706807718246611e-05, + "loss": 0.887, + "step": 12899 + }, + { + "epoch": 2.296652421652422, + "grad_norm": 0.774459719657898, + "learning_rate": 7.705445294987304e-05, + "loss": 0.914, + "step": 12900 + }, + { + "epoch": 2.296830484330484, + "grad_norm": 0.8464851379394531, + "learning_rate": 7.704082916685034e-05, + "loss": 1.0116, + "step": 12901 + }, + { + "epoch": 2.297008547008547, + "grad_norm": 0.8497290015220642, + "learning_rate": 7.702720583366486e-05, + "loss": 0.9242, + "step": 12902 + }, + { + "epoch": 2.2971866096866096, + "grad_norm": 0.8673670291900635, + "learning_rate": 7.70135829505836e-05, + "loss": 0.8172, + "step": 12903 + }, + { + "epoch": 2.2973646723646723, + "grad_norm": 0.786389172077179, + "learning_rate": 7.699996051787341e-05, + "loss": 0.6713, + "step": 12904 + }, + { + "epoch": 2.297542735042735, + "grad_norm": 0.8441919088363647, + "learning_rate": 7.698633853580124e-05, + "loss": 0.7835, + "step": 12905 + }, + { + "epoch": 2.297720797720798, + "grad_norm": 0.8806493878364563, + "learning_rate": 7.697271700463392e-05, + "loss": 0.9103, + "step": 12906 + }, + { + "epoch": 2.2978988603988606, + "grad_norm": 0.7418580651283264, + "learning_rate": 7.69590959246384e-05, + "loss": 0.9052, + "step": 12907 + }, + { + "epoch": 2.298076923076923, + "grad_norm": 0.7883853316307068, + "learning_rate": 7.694547529608152e-05, + "loss": 0.7689, + "step": 12908 + }, + { + "epoch": 2.2982549857549857, + "grad_norm": 0.7842690944671631, + "learning_rate": 7.693185511923017e-05, + "loss": 0.9587, + "step": 12909 + }, + { + "epoch": 2.2984330484330484, + "grad_norm": 0.884484052658081, + "learning_rate": 7.691823539435119e-05, + "loss": 0.9562, + "step": 12910 + }, + { + "epoch": 2.298611111111111, + "grad_norm": 0.8152852058410645, + "learning_rate": 7.690461612171145e-05, + "loss": 0.9857, + "step": 12911 + }, + { + "epoch": 2.298789173789174, + "grad_norm": 0.8502064943313599, + "learning_rate": 7.689099730157776e-05, + "loss": 0.7806, + "step": 12912 + }, + { + "epoch": 2.298967236467236, + "grad_norm": 0.9655177593231201, + "learning_rate": 7.687737893421697e-05, + "loss": 0.9693, + "step": 12913 + }, + { + "epoch": 2.299145299145299, + "grad_norm": 0.7759003639221191, + "learning_rate": 7.686376101989596e-05, + "loss": 0.9137, + "step": 12914 + }, + { + "epoch": 2.2993233618233617, + "grad_norm": 0.6987054944038391, + "learning_rate": 7.685014355888143e-05, + "loss": 0.9026, + "step": 12915 + }, + { + "epoch": 2.2995014245014245, + "grad_norm": 0.762819230556488, + "learning_rate": 7.683652655144027e-05, + "loss": 0.8358, + "step": 12916 + }, + { + "epoch": 2.2996794871794872, + "grad_norm": 0.8233383893966675, + "learning_rate": 7.682290999783924e-05, + "loss": 0.8468, + "step": 12917 + }, + { + "epoch": 2.29985754985755, + "grad_norm": 0.8558689951896667, + "learning_rate": 7.68092938983452e-05, + "loss": 0.9018, + "step": 12918 + }, + { + "epoch": 2.3000356125356127, + "grad_norm": 0.741760790348053, + "learning_rate": 7.67956782532248e-05, + "loss": 0.7968, + "step": 12919 + }, + { + "epoch": 2.300213675213675, + "grad_norm": 0.9132583737373352, + "learning_rate": 7.678206306274495e-05, + "loss": 0.9952, + "step": 12920 + }, + { + "epoch": 2.300391737891738, + "grad_norm": 0.7656551003456116, + "learning_rate": 7.67684483271723e-05, + "loss": 0.8772, + "step": 12921 + }, + { + "epoch": 2.3005698005698005, + "grad_norm": 0.7407111525535583, + "learning_rate": 7.675483404677364e-05, + "loss": 0.8199, + "step": 12922 + }, + { + "epoch": 2.3007478632478633, + "grad_norm": 0.9602083563804626, + "learning_rate": 7.674122022181571e-05, + "loss": 1.0837, + "step": 12923 + }, + { + "epoch": 2.300925925925926, + "grad_norm": 0.7562392354011536, + "learning_rate": 7.672760685256531e-05, + "loss": 0.8148, + "step": 12924 + }, + { + "epoch": 2.301103988603989, + "grad_norm": 0.9260091185569763, + "learning_rate": 7.671399393928906e-05, + "loss": 0.9508, + "step": 12925 + }, + { + "epoch": 2.301282051282051, + "grad_norm": 0.8745924234390259, + "learning_rate": 7.670038148225374e-05, + "loss": 0.8688, + "step": 12926 + }, + { + "epoch": 2.301460113960114, + "grad_norm": 0.7802116274833679, + "learning_rate": 7.668676948172602e-05, + "loss": 0.7698, + "step": 12927 + }, + { + "epoch": 2.3016381766381766, + "grad_norm": 0.7701709866523743, + "learning_rate": 7.667315793797268e-05, + "loss": 0.7633, + "step": 12928 + }, + { + "epoch": 2.3018162393162394, + "grad_norm": 0.8084021806716919, + "learning_rate": 7.66595468512603e-05, + "loss": 0.8502, + "step": 12929 + }, + { + "epoch": 2.301994301994302, + "grad_norm": 1.0485330820083618, + "learning_rate": 7.664593622185568e-05, + "loss": 0.8049, + "step": 12930 + }, + { + "epoch": 2.302172364672365, + "grad_norm": 0.7852743864059448, + "learning_rate": 7.663232605002535e-05, + "loss": 0.882, + "step": 12931 + }, + { + "epoch": 2.302350427350427, + "grad_norm": 0.7795702815055847, + "learning_rate": 7.661871633603607e-05, + "loss": 0.7841, + "step": 12932 + }, + { + "epoch": 2.30252849002849, + "grad_norm": 0.8882975578308105, + "learning_rate": 7.660510708015448e-05, + "loss": 1.117, + "step": 12933 + }, + { + "epoch": 2.3027065527065527, + "grad_norm": 0.7987662553787231, + "learning_rate": 7.65914982826472e-05, + "loss": 0.8552, + "step": 12934 + }, + { + "epoch": 2.3028846153846154, + "grad_norm": 0.8141679167747498, + "learning_rate": 7.657788994378095e-05, + "loss": 0.8288, + "step": 12935 + }, + { + "epoch": 2.303062678062678, + "grad_norm": 0.8506320118904114, + "learning_rate": 7.656428206382222e-05, + "loss": 0.7521, + "step": 12936 + }, + { + "epoch": 2.303240740740741, + "grad_norm": 0.7666227221488953, + "learning_rate": 7.655067464303773e-05, + "loss": 0.8394, + "step": 12937 + }, + { + "epoch": 2.3034188034188032, + "grad_norm": 0.8018062710762024, + "learning_rate": 7.653706768169405e-05, + "loss": 0.7566, + "step": 12938 + }, + { + "epoch": 2.303596866096866, + "grad_norm": 0.8054059743881226, + "learning_rate": 7.652346118005779e-05, + "loss": 0.8749, + "step": 12939 + }, + { + "epoch": 2.3037749287749287, + "grad_norm": 0.8663263320922852, + "learning_rate": 7.650985513839554e-05, + "loss": 0.7799, + "step": 12940 + }, + { + "epoch": 2.3039529914529915, + "grad_norm": 0.7591161727905273, + "learning_rate": 7.64962495569739e-05, + "loss": 0.7378, + "step": 12941 + }, + { + "epoch": 2.3041310541310542, + "grad_norm": 0.8118969202041626, + "learning_rate": 7.64826444360594e-05, + "loss": 0.7948, + "step": 12942 + }, + { + "epoch": 2.304309116809117, + "grad_norm": 0.6880847811698914, + "learning_rate": 7.646903977591865e-05, + "loss": 0.9164, + "step": 12943 + }, + { + "epoch": 2.3044871794871793, + "grad_norm": 0.814386248588562, + "learning_rate": 7.645543557681816e-05, + "loss": 0.7998, + "step": 12944 + }, + { + "epoch": 2.304665242165242, + "grad_norm": 0.8295530676841736, + "learning_rate": 7.644183183902454e-05, + "loss": 0.812, + "step": 12945 + }, + { + "epoch": 2.304843304843305, + "grad_norm": 0.7872505187988281, + "learning_rate": 7.642822856280424e-05, + "loss": 0.9073, + "step": 12946 + }, + { + "epoch": 2.3050213675213675, + "grad_norm": 0.9217497110366821, + "learning_rate": 7.641462574842387e-05, + "loss": 0.7762, + "step": 12947 + }, + { + "epoch": 2.3051994301994303, + "grad_norm": 0.7502169609069824, + "learning_rate": 7.640102339614987e-05, + "loss": 0.9374, + "step": 12948 + }, + { + "epoch": 2.305377492877493, + "grad_norm": 0.8262767195701599, + "learning_rate": 7.638742150624886e-05, + "loss": 0.5363, + "step": 12949 + }, + { + "epoch": 2.3055555555555554, + "grad_norm": 0.7571384310722351, + "learning_rate": 7.637382007898722e-05, + "loss": 0.9548, + "step": 12950 + }, + { + "epoch": 2.305733618233618, + "grad_norm": 0.7899317145347595, + "learning_rate": 7.636021911463152e-05, + "loss": 0.7718, + "step": 12951 + }, + { + "epoch": 2.305911680911681, + "grad_norm": 0.7772458791732788, + "learning_rate": 7.634661861344819e-05, + "loss": 0.7158, + "step": 12952 + }, + { + "epoch": 2.3060897435897436, + "grad_norm": 0.8279168009757996, + "learning_rate": 7.633301857570374e-05, + "loss": 0.7835, + "step": 12953 + }, + { + "epoch": 2.3062678062678064, + "grad_norm": 0.751268208026886, + "learning_rate": 7.631941900166468e-05, + "loss": 0.8609, + "step": 12954 + }, + { + "epoch": 2.306445868945869, + "grad_norm": 0.8101294636726379, + "learning_rate": 7.630581989159736e-05, + "loss": 1.0242, + "step": 12955 + }, + { + "epoch": 2.306623931623932, + "grad_norm": 0.7707645297050476, + "learning_rate": 7.629222124576831e-05, + "loss": 0.7969, + "step": 12956 + }, + { + "epoch": 2.306801994301994, + "grad_norm": 0.6519944667816162, + "learning_rate": 7.627862306444391e-05, + "loss": 0.5459, + "step": 12957 + }, + { + "epoch": 2.306980056980057, + "grad_norm": 0.7738897800445557, + "learning_rate": 7.626502534789063e-05, + "loss": 0.957, + "step": 12958 + }, + { + "epoch": 2.3071581196581197, + "grad_norm": 0.7059842944145203, + "learning_rate": 7.625142809637485e-05, + "loss": 0.6316, + "step": 12959 + }, + { + "epoch": 2.3073361823361824, + "grad_norm": 0.8380797505378723, + "learning_rate": 7.623783131016305e-05, + "loss": 0.9685, + "step": 12960 + }, + { + "epoch": 2.307514245014245, + "grad_norm": 0.8272121548652649, + "learning_rate": 7.622423498952154e-05, + "loss": 0.9425, + "step": 12961 + }, + { + "epoch": 2.3076923076923075, + "grad_norm": 0.763522744178772, + "learning_rate": 7.621063913471678e-05, + "loss": 0.7778, + "step": 12962 + }, + { + "epoch": 2.3078703703703702, + "grad_norm": 0.8345584273338318, + "learning_rate": 7.61970437460151e-05, + "loss": 0.9652, + "step": 12963 + }, + { + "epoch": 2.308048433048433, + "grad_norm": 0.943286657333374, + "learning_rate": 7.618344882368294e-05, + "loss": 0.9088, + "step": 12964 + }, + { + "epoch": 2.3082264957264957, + "grad_norm": 0.8568450212478638, + "learning_rate": 7.616985436798659e-05, + "loss": 0.7535, + "step": 12965 + }, + { + "epoch": 2.3084045584045585, + "grad_norm": 0.8722548484802246, + "learning_rate": 7.615626037919248e-05, + "loss": 0.9802, + "step": 12966 + }, + { + "epoch": 2.3085826210826212, + "grad_norm": 1.0332363843917847, + "learning_rate": 7.614266685756688e-05, + "loss": 0.9105, + "step": 12967 + }, + { + "epoch": 2.308760683760684, + "grad_norm": 0.7503480315208435, + "learning_rate": 7.612907380337619e-05, + "loss": 0.7345, + "step": 12968 + }, + { + "epoch": 2.3089387464387463, + "grad_norm": 0.7406014204025269, + "learning_rate": 7.611548121688668e-05, + "loss": 0.9222, + "step": 12969 + }, + { + "epoch": 2.309116809116809, + "grad_norm": 0.7574487328529358, + "learning_rate": 7.610188909836474e-05, + "loss": 0.7709, + "step": 12970 + }, + { + "epoch": 2.309294871794872, + "grad_norm": 0.8669037818908691, + "learning_rate": 7.608829744807661e-05, + "loss": 0.838, + "step": 12971 + }, + { + "epoch": 2.3094729344729346, + "grad_norm": 0.7544569373130798, + "learning_rate": 7.607470626628861e-05, + "loss": 0.6966, + "step": 12972 + }, + { + "epoch": 2.3096509971509973, + "grad_norm": 0.8201249241828918, + "learning_rate": 7.606111555326706e-05, + "loss": 0.9322, + "step": 12973 + }, + { + "epoch": 2.3098290598290596, + "grad_norm": 0.7935477495193481, + "learning_rate": 7.60475253092782e-05, + "loss": 0.7981, + "step": 12974 + }, + { + "epoch": 2.3100071225071224, + "grad_norm": 0.8775026798248291, + "learning_rate": 7.603393553458838e-05, + "loss": 0.8352, + "step": 12975 + }, + { + "epoch": 2.310185185185185, + "grad_norm": 0.8422450423240662, + "learning_rate": 7.602034622946374e-05, + "loss": 0.577, + "step": 12976 + }, + { + "epoch": 2.310363247863248, + "grad_norm": 0.8584204316139221, + "learning_rate": 7.600675739417067e-05, + "loss": 0.9767, + "step": 12977 + }, + { + "epoch": 2.3105413105413106, + "grad_norm": 0.7818547487258911, + "learning_rate": 7.599316902897528e-05, + "loss": 0.7675, + "step": 12978 + }, + { + "epoch": 2.3107193732193734, + "grad_norm": 0.93815016746521, + "learning_rate": 7.597958113414391e-05, + "loss": 0.8517, + "step": 12979 + }, + { + "epoch": 2.310897435897436, + "grad_norm": 0.8092408776283264, + "learning_rate": 7.596599370994272e-05, + "loss": 0.7266, + "step": 12980 + }, + { + "epoch": 2.3110754985754984, + "grad_norm": 0.8577243089675903, + "learning_rate": 7.595240675663802e-05, + "loss": 0.9138, + "step": 12981 + }, + { + "epoch": 2.311253561253561, + "grad_norm": 0.8600401878356934, + "learning_rate": 7.59388202744959e-05, + "loss": 0.8348, + "step": 12982 + }, + { + "epoch": 2.311431623931624, + "grad_norm": 0.6399564743041992, + "learning_rate": 7.592523426378264e-05, + "loss": 0.6649, + "step": 12983 + }, + { + "epoch": 2.3116096866096867, + "grad_norm": 0.7916820049285889, + "learning_rate": 7.591164872476438e-05, + "loss": 0.8048, + "step": 12984 + }, + { + "epoch": 2.3117877492877494, + "grad_norm": 0.7748355269432068, + "learning_rate": 7.589806365770736e-05, + "loss": 1.0101, + "step": 12985 + }, + { + "epoch": 2.3119658119658117, + "grad_norm": 0.8463436365127563, + "learning_rate": 7.588447906287767e-05, + "loss": 0.7547, + "step": 12986 + }, + { + "epoch": 2.3121438746438745, + "grad_norm": 0.8257808685302734, + "learning_rate": 7.587089494054155e-05, + "loss": 0.8093, + "step": 12987 + }, + { + "epoch": 2.3123219373219372, + "grad_norm": 0.843781054019928, + "learning_rate": 7.58573112909651e-05, + "loss": 0.8379, + "step": 12988 + }, + { + "epoch": 2.3125, + "grad_norm": 0.8782341480255127, + "learning_rate": 7.584372811441452e-05, + "loss": 0.9372, + "step": 12989 + }, + { + "epoch": 2.3126780626780628, + "grad_norm": 0.8465158343315125, + "learning_rate": 7.583014541115585e-05, + "loss": 0.8427, + "step": 12990 + }, + { + "epoch": 2.3128561253561255, + "grad_norm": 0.7140238285064697, + "learning_rate": 7.58165631814553e-05, + "loss": 0.7896, + "step": 12991 + }, + { + "epoch": 2.3130341880341883, + "grad_norm": 0.9414699077606201, + "learning_rate": 7.580298142557898e-05, + "loss": 1.0464, + "step": 12992 + }, + { + "epoch": 2.3132122507122506, + "grad_norm": 0.7970326542854309, + "learning_rate": 7.578940014379293e-05, + "loss": 0.696, + "step": 12993 + }, + { + "epoch": 2.3133903133903133, + "grad_norm": 0.7377375960350037, + "learning_rate": 7.577581933636332e-05, + "loss": 0.7205, + "step": 12994 + }, + { + "epoch": 2.313568376068376, + "grad_norm": 0.7740316987037659, + "learning_rate": 7.576223900355619e-05, + "loss": 0.7448, + "step": 12995 + }, + { + "epoch": 2.313746438746439, + "grad_norm": 0.7820385098457336, + "learning_rate": 7.574865914563767e-05, + "loss": 0.7289, + "step": 12996 + }, + { + "epoch": 2.3139245014245016, + "grad_norm": 0.8180822134017944, + "learning_rate": 7.573507976287376e-05, + "loss": 0.8709, + "step": 12997 + }, + { + "epoch": 2.314102564102564, + "grad_norm": 0.9008440971374512, + "learning_rate": 7.572150085553058e-05, + "loss": 0.7938, + "step": 12998 + }, + { + "epoch": 2.3142806267806266, + "grad_norm": 0.786400318145752, + "learning_rate": 7.570792242387414e-05, + "loss": 0.9866, + "step": 12999 + }, + { + "epoch": 2.3144586894586894, + "grad_norm": 0.872160792350769, + "learning_rate": 7.569434446817052e-05, + "loss": 0.7319, + "step": 13000 + }, + { + "epoch": 2.314636752136752, + "grad_norm": 0.7858988642692566, + "learning_rate": 7.56807669886857e-05, + "loss": 0.8786, + "step": 13001 + }, + { + "epoch": 2.314814814814815, + "grad_norm": 0.7090579271316528, + "learning_rate": 7.566718998568579e-05, + "loss": 0.7092, + "step": 13002 + }, + { + "epoch": 2.3149928774928776, + "grad_norm": 0.7881498336791992, + "learning_rate": 7.565361345943668e-05, + "loss": 0.876, + "step": 13003 + }, + { + "epoch": 2.3151709401709404, + "grad_norm": 0.8768819570541382, + "learning_rate": 7.564003741020447e-05, + "loss": 1.0374, + "step": 13004 + }, + { + "epoch": 2.3153490028490027, + "grad_norm": 0.7608295679092407, + "learning_rate": 7.56264618382551e-05, + "loss": 0.7551, + "step": 13005 + }, + { + "epoch": 2.3155270655270654, + "grad_norm": 0.6947942972183228, + "learning_rate": 7.561288674385462e-05, + "loss": 0.8132, + "step": 13006 + }, + { + "epoch": 2.315705128205128, + "grad_norm": 0.8722706437110901, + "learning_rate": 7.559931212726892e-05, + "loss": 0.7486, + "step": 13007 + }, + { + "epoch": 2.315883190883191, + "grad_norm": 0.9804681539535522, + "learning_rate": 7.558573798876404e-05, + "loss": 0.899, + "step": 13008 + }, + { + "epoch": 2.3160612535612537, + "grad_norm": 0.9246195554733276, + "learning_rate": 7.557216432860587e-05, + "loss": 0.7742, + "step": 13009 + }, + { + "epoch": 2.316239316239316, + "grad_norm": 0.8792895078659058, + "learning_rate": 7.555859114706046e-05, + "loss": 0.8299, + "step": 13010 + }, + { + "epoch": 2.3164173789173788, + "grad_norm": 0.8280500769615173, + "learning_rate": 7.554501844439362e-05, + "loss": 0.8708, + "step": 13011 + }, + { + "epoch": 2.3165954415954415, + "grad_norm": 0.8560570478439331, + "learning_rate": 7.553144622087136e-05, + "loss": 0.9571, + "step": 13012 + }, + { + "epoch": 2.3167735042735043, + "grad_norm": 0.8504697680473328, + "learning_rate": 7.551787447675962e-05, + "loss": 0.7609, + "step": 13013 + }, + { + "epoch": 2.316951566951567, + "grad_norm": 0.8199480772018433, + "learning_rate": 7.550430321232423e-05, + "loss": 0.7077, + "step": 13014 + }, + { + "epoch": 2.3171296296296298, + "grad_norm": 0.854341447353363, + "learning_rate": 7.549073242783115e-05, + "loss": 0.9602, + "step": 13015 + }, + { + "epoch": 2.3173076923076925, + "grad_norm": 0.7619566917419434, + "learning_rate": 7.547716212354623e-05, + "loss": 0.9967, + "step": 13016 + }, + { + "epoch": 2.317485754985755, + "grad_norm": 0.7371547222137451, + "learning_rate": 7.546359229973543e-05, + "loss": 0.7239, + "step": 13017 + }, + { + "epoch": 2.3176638176638176, + "grad_norm": 0.7797415852546692, + "learning_rate": 7.545002295666453e-05, + "loss": 0.7472, + "step": 13018 + }, + { + "epoch": 2.3178418803418803, + "grad_norm": 0.8549608588218689, + "learning_rate": 7.543645409459943e-05, + "loss": 0.8968, + "step": 13019 + }, + { + "epoch": 2.318019943019943, + "grad_norm": 0.7931239008903503, + "learning_rate": 7.542288571380598e-05, + "loss": 0.9853, + "step": 13020 + }, + { + "epoch": 2.318198005698006, + "grad_norm": 0.797726035118103, + "learning_rate": 7.540931781455008e-05, + "loss": 0.9366, + "step": 13021 + }, + { + "epoch": 2.318376068376068, + "grad_norm": 0.7382092475891113, + "learning_rate": 7.539575039709747e-05, + "loss": 0.6484, + "step": 13022 + }, + { + "epoch": 2.318554131054131, + "grad_norm": 0.83231121301651, + "learning_rate": 7.538218346171403e-05, + "loss": 1.0184, + "step": 13023 + }, + { + "epoch": 2.3187321937321936, + "grad_norm": 0.8613845109939575, + "learning_rate": 7.536861700866554e-05, + "loss": 0.8019, + "step": 13024 + }, + { + "epoch": 2.3189102564102564, + "grad_norm": 0.7736538648605347, + "learning_rate": 7.53550510382179e-05, + "loss": 0.8228, + "step": 13025 + }, + { + "epoch": 2.319088319088319, + "grad_norm": 0.7894346714019775, + "learning_rate": 7.534148555063678e-05, + "loss": 0.8189, + "step": 13026 + }, + { + "epoch": 2.319266381766382, + "grad_norm": 0.7333146333694458, + "learning_rate": 7.532792054618807e-05, + "loss": 0.8456, + "step": 13027 + }, + { + "epoch": 2.3194444444444446, + "grad_norm": 1.0321780443191528, + "learning_rate": 7.531435602513745e-05, + "loss": 0.8594, + "step": 13028 + }, + { + "epoch": 2.319622507122507, + "grad_norm": 0.8658601641654968, + "learning_rate": 7.530079198775079e-05, + "loss": 0.9638, + "step": 13029 + }, + { + "epoch": 2.3198005698005697, + "grad_norm": 0.7287920713424683, + "learning_rate": 7.528722843429376e-05, + "loss": 0.7001, + "step": 13030 + }, + { + "epoch": 2.3199786324786325, + "grad_norm": 0.7398431301116943, + "learning_rate": 7.527366536503218e-05, + "loss": 0.8595, + "step": 13031 + }, + { + "epoch": 2.320156695156695, + "grad_norm": 0.8127739429473877, + "learning_rate": 7.526010278023178e-05, + "loss": 0.7641, + "step": 13032 + }, + { + "epoch": 2.320334757834758, + "grad_norm": 0.776497483253479, + "learning_rate": 7.524654068015824e-05, + "loss": 0.8666, + "step": 13033 + }, + { + "epoch": 2.3205128205128207, + "grad_norm": 0.8524185419082642, + "learning_rate": 7.523297906507733e-05, + "loss": 0.9815, + "step": 13034 + }, + { + "epoch": 2.320690883190883, + "grad_norm": 0.7745016813278198, + "learning_rate": 7.521941793525474e-05, + "loss": 0.8527, + "step": 13035 + }, + { + "epoch": 2.3208689458689458, + "grad_norm": 0.8695911169052124, + "learning_rate": 7.52058572909562e-05, + "loss": 0.8881, + "step": 13036 + }, + { + "epoch": 2.3210470085470085, + "grad_norm": 0.788969099521637, + "learning_rate": 7.519229713244736e-05, + "loss": 0.7886, + "step": 13037 + }, + { + "epoch": 2.3212250712250713, + "grad_norm": 0.776520311832428, + "learning_rate": 7.517873745999394e-05, + "loss": 0.5986, + "step": 13038 + }, + { + "epoch": 2.321403133903134, + "grad_norm": 0.8118561506271362, + "learning_rate": 7.516517827386158e-05, + "loss": 0.8805, + "step": 13039 + }, + { + "epoch": 2.3215811965811968, + "grad_norm": 0.8859134912490845, + "learning_rate": 7.515161957431596e-05, + "loss": 0.8861, + "step": 13040 + }, + { + "epoch": 2.321759259259259, + "grad_norm": 0.8181297779083252, + "learning_rate": 7.513806136162273e-05, + "loss": 0.9015, + "step": 13041 + }, + { + "epoch": 2.321937321937322, + "grad_norm": 0.8488339185714722, + "learning_rate": 7.512450363604759e-05, + "loss": 1.0423, + "step": 13042 + }, + { + "epoch": 2.3221153846153846, + "grad_norm": 0.7755734920501709, + "learning_rate": 7.511094639785607e-05, + "loss": 0.7595, + "step": 13043 + }, + { + "epoch": 2.3222934472934473, + "grad_norm": 0.8437283635139465, + "learning_rate": 7.509738964731389e-05, + "loss": 0.9011, + "step": 13044 + }, + { + "epoch": 2.32247150997151, + "grad_norm": 0.7508310675621033, + "learning_rate": 7.508383338468659e-05, + "loss": 0.8335, + "step": 13045 + }, + { + "epoch": 2.322649572649573, + "grad_norm": 0.8001464605331421, + "learning_rate": 7.507027761023987e-05, + "loss": 0.9785, + "step": 13046 + }, + { + "epoch": 2.322827635327635, + "grad_norm": 0.8142531514167786, + "learning_rate": 7.505672232423923e-05, + "loss": 0.8577, + "step": 13047 + }, + { + "epoch": 2.323005698005698, + "grad_norm": 0.7852125763893127, + "learning_rate": 7.504316752695035e-05, + "loss": 0.798, + "step": 13048 + }, + { + "epoch": 2.3231837606837606, + "grad_norm": 0.8998631238937378, + "learning_rate": 7.502961321863871e-05, + "loss": 0.9291, + "step": 13049 + }, + { + "epoch": 2.3233618233618234, + "grad_norm": 0.8850175738334656, + "learning_rate": 7.501605939956995e-05, + "loss": 0.9536, + "step": 13050 + }, + { + "epoch": 2.323539886039886, + "grad_norm": 0.8305020928382874, + "learning_rate": 7.500250607000959e-05, + "loss": 0.8695, + "step": 13051 + }, + { + "epoch": 2.323717948717949, + "grad_norm": 0.8073359727859497, + "learning_rate": 7.498895323022317e-05, + "loss": 0.6831, + "step": 13052 + }, + { + "epoch": 2.323896011396011, + "grad_norm": 0.8435724973678589, + "learning_rate": 7.497540088047632e-05, + "loss": 0.9419, + "step": 13053 + }, + { + "epoch": 2.324074074074074, + "grad_norm": 0.927147388458252, + "learning_rate": 7.496184902103446e-05, + "loss": 0.957, + "step": 13054 + }, + { + "epoch": 2.3242521367521367, + "grad_norm": 0.7923009395599365, + "learning_rate": 7.494829765216319e-05, + "loss": 0.839, + "step": 13055 + }, + { + "epoch": 2.3244301994301995, + "grad_norm": 0.7830277681350708, + "learning_rate": 7.493474677412794e-05, + "loss": 1.0236, + "step": 13056 + }, + { + "epoch": 2.324608262108262, + "grad_norm": 0.8470967411994934, + "learning_rate": 7.492119638719432e-05, + "loss": 0.9144, + "step": 13057 + }, + { + "epoch": 2.324786324786325, + "grad_norm": 0.7469272613525391, + "learning_rate": 7.490764649162771e-05, + "loss": 0.7101, + "step": 13058 + }, + { + "epoch": 2.3249643874643873, + "grad_norm": 0.9236082434654236, + "learning_rate": 7.489409708769366e-05, + "loss": 0.6658, + "step": 13059 + }, + { + "epoch": 2.32514245014245, + "grad_norm": 0.8271692395210266, + "learning_rate": 7.48805481756576e-05, + "loss": 0.775, + "step": 13060 + }, + { + "epoch": 2.3253205128205128, + "grad_norm": 0.9878279566764832, + "learning_rate": 7.486699975578507e-05, + "loss": 0.6881, + "step": 13061 + }, + { + "epoch": 2.3254985754985755, + "grad_norm": 0.7332003712654114, + "learning_rate": 7.485345182834142e-05, + "loss": 0.8384, + "step": 13062 + }, + { + "epoch": 2.3256766381766383, + "grad_norm": 0.9525214433670044, + "learning_rate": 7.483990439359221e-05, + "loss": 0.9892, + "step": 13063 + }, + { + "epoch": 2.325854700854701, + "grad_norm": 0.7413233518600464, + "learning_rate": 7.482635745180273e-05, + "loss": 0.8749, + "step": 13064 + }, + { + "epoch": 2.326032763532764, + "grad_norm": 0.8286891579627991, + "learning_rate": 7.481281100323854e-05, + "loss": 1.0313, + "step": 13065 + }, + { + "epoch": 2.326210826210826, + "grad_norm": 0.868653416633606, + "learning_rate": 7.479926504816495e-05, + "loss": 0.7407, + "step": 13066 + }, + { + "epoch": 2.326388888888889, + "grad_norm": 0.790052056312561, + "learning_rate": 7.478571958684746e-05, + "loss": 0.7156, + "step": 13067 + }, + { + "epoch": 2.3265669515669516, + "grad_norm": 0.8799049854278564, + "learning_rate": 7.477217461955137e-05, + "loss": 0.7706, + "step": 13068 + }, + { + "epoch": 2.3267450142450143, + "grad_norm": 0.8246361017227173, + "learning_rate": 7.475863014654214e-05, + "loss": 0.9402, + "step": 13069 + }, + { + "epoch": 2.326923076923077, + "grad_norm": 0.870917022228241, + "learning_rate": 7.474508616808508e-05, + "loss": 0.962, + "step": 13070 + }, + { + "epoch": 2.3271011396011394, + "grad_norm": 0.8706079125404358, + "learning_rate": 7.473154268444563e-05, + "loss": 0.9094, + "step": 13071 + }, + { + "epoch": 2.327279202279202, + "grad_norm": 0.9031453728675842, + "learning_rate": 7.471799969588912e-05, + "loss": 0.8447, + "step": 13072 + }, + { + "epoch": 2.327457264957265, + "grad_norm": 0.9153435230255127, + "learning_rate": 7.470445720268086e-05, + "loss": 0.9935, + "step": 13073 + }, + { + "epoch": 2.3276353276353277, + "grad_norm": 0.8236302733421326, + "learning_rate": 7.469091520508624e-05, + "loss": 0.7911, + "step": 13074 + }, + { + "epoch": 2.3278133903133904, + "grad_norm": 0.7344710826873779, + "learning_rate": 7.467737370337054e-05, + "loss": 0.6565, + "step": 13075 + }, + { + "epoch": 2.327991452991453, + "grad_norm": 0.8711966276168823, + "learning_rate": 7.466383269779911e-05, + "loss": 0.8332, + "step": 13076 + }, + { + "epoch": 2.328169515669516, + "grad_norm": 0.836825966835022, + "learning_rate": 7.465029218863723e-05, + "loss": 0.9359, + "step": 13077 + }, + { + "epoch": 2.328347578347578, + "grad_norm": 0.9791260361671448, + "learning_rate": 7.463675217615024e-05, + "loss": 0.8938, + "step": 13078 + }, + { + "epoch": 2.328525641025641, + "grad_norm": 0.7260454893112183, + "learning_rate": 7.46232126606034e-05, + "loss": 0.6871, + "step": 13079 + }, + { + "epoch": 2.3287037037037037, + "grad_norm": 0.7887428998947144, + "learning_rate": 7.460967364226197e-05, + "loss": 0.8098, + "step": 13080 + }, + { + "epoch": 2.3288817663817665, + "grad_norm": 0.8303743004798889, + "learning_rate": 7.459613512139124e-05, + "loss": 0.8897, + "step": 13081 + }, + { + "epoch": 2.3290598290598292, + "grad_norm": 0.7933324575424194, + "learning_rate": 7.458259709825652e-05, + "loss": 0.754, + "step": 13082 + }, + { + "epoch": 2.3292378917378915, + "grad_norm": 0.8998779058456421, + "learning_rate": 7.456905957312296e-05, + "loss": 0.866, + "step": 13083 + }, + { + "epoch": 2.3294159544159543, + "grad_norm": 0.8205044269561768, + "learning_rate": 7.455552254625588e-05, + "loss": 0.7455, + "step": 13084 + }, + { + "epoch": 2.329594017094017, + "grad_norm": 0.8731769323348999, + "learning_rate": 7.454198601792046e-05, + "loss": 0.8876, + "step": 13085 + }, + { + "epoch": 2.32977207977208, + "grad_norm": 0.9183599352836609, + "learning_rate": 7.452844998838194e-05, + "loss": 1.0991, + "step": 13086 + }, + { + "epoch": 2.3299501424501425, + "grad_norm": 0.8820931315422058, + "learning_rate": 7.451491445790553e-05, + "loss": 0.7591, + "step": 13087 + }, + { + "epoch": 2.3301282051282053, + "grad_norm": 0.7837240099906921, + "learning_rate": 7.450137942675646e-05, + "loss": 0.792, + "step": 13088 + }, + { + "epoch": 2.330306267806268, + "grad_norm": 0.8960266709327698, + "learning_rate": 7.448784489519984e-05, + "loss": 0.9725, + "step": 13089 + }, + { + "epoch": 2.3304843304843303, + "grad_norm": 0.8010196685791016, + "learning_rate": 7.447431086350092e-05, + "loss": 0.8418, + "step": 13090 + }, + { + "epoch": 2.330662393162393, + "grad_norm": 0.8652680516242981, + "learning_rate": 7.446077733192486e-05, + "loss": 0.93, + "step": 13091 + }, + { + "epoch": 2.330840455840456, + "grad_norm": 0.9385902285575867, + "learning_rate": 7.44472443007368e-05, + "loss": 0.968, + "step": 13092 + }, + { + "epoch": 2.3310185185185186, + "grad_norm": 0.8097951412200928, + "learning_rate": 7.443371177020195e-05, + "loss": 0.8715, + "step": 13093 + }, + { + "epoch": 2.3311965811965814, + "grad_norm": 0.7931473255157471, + "learning_rate": 7.442017974058537e-05, + "loss": 0.865, + "step": 13094 + }, + { + "epoch": 2.3313746438746437, + "grad_norm": 0.7680486440658569, + "learning_rate": 7.440664821215224e-05, + "loss": 0.9155, + "step": 13095 + }, + { + "epoch": 2.3315527065527064, + "grad_norm": 0.8128345012664795, + "learning_rate": 7.439311718516766e-05, + "loss": 0.8707, + "step": 13096 + }, + { + "epoch": 2.331730769230769, + "grad_norm": 0.9534463286399841, + "learning_rate": 7.43795866598968e-05, + "loss": 1.1102, + "step": 13097 + }, + { + "epoch": 2.331908831908832, + "grad_norm": 0.9140331745147705, + "learning_rate": 7.436605663660468e-05, + "loss": 0.7744, + "step": 13098 + }, + { + "epoch": 2.3320868945868947, + "grad_norm": 0.8316463828086853, + "learning_rate": 7.435252711555645e-05, + "loss": 0.7201, + "step": 13099 + }, + { + "epoch": 2.3322649572649574, + "grad_norm": 0.7714298963546753, + "learning_rate": 7.433899809701714e-05, + "loss": 0.9111, + "step": 13100 + }, + { + "epoch": 2.33244301994302, + "grad_norm": 0.999081552028656, + "learning_rate": 7.432546958125188e-05, + "loss": 0.8839, + "step": 13101 + }, + { + "epoch": 2.3326210826210825, + "grad_norm": 0.8515602350234985, + "learning_rate": 7.43119415685257e-05, + "loss": 0.7245, + "step": 13102 + }, + { + "epoch": 2.3327991452991452, + "grad_norm": 0.9441094398498535, + "learning_rate": 7.42984140591037e-05, + "loss": 0.8678, + "step": 13103 + }, + { + "epoch": 2.332977207977208, + "grad_norm": 0.8220996856689453, + "learning_rate": 7.428488705325084e-05, + "loss": 0.9737, + "step": 13104 + }, + { + "epoch": 2.3331552706552707, + "grad_norm": 0.8164090514183044, + "learning_rate": 7.427136055123222e-05, + "loss": 0.9138, + "step": 13105 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.9672707319259644, + "learning_rate": 7.425783455331281e-05, + "loss": 0.7723, + "step": 13106 + }, + { + "epoch": 2.333511396011396, + "grad_norm": 0.7953858971595764, + "learning_rate": 7.424430905975773e-05, + "loss": 0.8876, + "step": 13107 + }, + { + "epoch": 2.3336894586894585, + "grad_norm": 0.7809541821479797, + "learning_rate": 7.423078407083183e-05, + "loss": 1.0268, + "step": 13108 + }, + { + "epoch": 2.3338675213675213, + "grad_norm": 0.805270254611969, + "learning_rate": 7.421725958680025e-05, + "loss": 0.7515, + "step": 13109 + }, + { + "epoch": 2.334045584045584, + "grad_norm": 0.8066652417182922, + "learning_rate": 7.420373560792788e-05, + "loss": 1.004, + "step": 13110 + }, + { + "epoch": 2.334223646723647, + "grad_norm": 0.9382686018943787, + "learning_rate": 7.41902121344797e-05, + "loss": 0.8769, + "step": 13111 + }, + { + "epoch": 2.3344017094017095, + "grad_norm": 0.7908356785774231, + "learning_rate": 7.417668916672074e-05, + "loss": 0.8491, + "step": 13112 + }, + { + "epoch": 2.3345797720797723, + "grad_norm": 0.7188867330551147, + "learning_rate": 7.416316670491588e-05, + "loss": 0.7138, + "step": 13113 + }, + { + "epoch": 2.3347578347578346, + "grad_norm": 0.8477714657783508, + "learning_rate": 7.414964474933012e-05, + "loss": 0.9715, + "step": 13114 + }, + { + "epoch": 2.3349358974358974, + "grad_norm": 0.8769845366477966, + "learning_rate": 7.413612330022835e-05, + "loss": 0.9029, + "step": 13115 + }, + { + "epoch": 2.33511396011396, + "grad_norm": 0.9011028409004211, + "learning_rate": 7.412260235787554e-05, + "loss": 1.026, + "step": 13116 + }, + { + "epoch": 2.335292022792023, + "grad_norm": 0.7775689363479614, + "learning_rate": 7.410908192253656e-05, + "loss": 0.8492, + "step": 13117 + }, + { + "epoch": 2.3354700854700856, + "grad_norm": 0.9587660431861877, + "learning_rate": 7.409556199447637e-05, + "loss": 0.8731, + "step": 13118 + }, + { + "epoch": 2.335648148148148, + "grad_norm": 0.8117266297340393, + "learning_rate": 7.408204257395979e-05, + "loss": 0.8827, + "step": 13119 + }, + { + "epoch": 2.3358262108262107, + "grad_norm": 0.7382497787475586, + "learning_rate": 7.40685236612518e-05, + "loss": 0.6617, + "step": 13120 + }, + { + "epoch": 2.3360042735042734, + "grad_norm": 0.8630974888801575, + "learning_rate": 7.405500525661717e-05, + "loss": 0.843, + "step": 13121 + }, + { + "epoch": 2.336182336182336, + "grad_norm": 0.7496539950370789, + "learning_rate": 7.404148736032083e-05, + "loss": 1.0354, + "step": 13122 + }, + { + "epoch": 2.336360398860399, + "grad_norm": 0.8409397602081299, + "learning_rate": 7.402796997262761e-05, + "loss": 0.9848, + "step": 13123 + }, + { + "epoch": 2.3365384615384617, + "grad_norm": 0.8018865585327148, + "learning_rate": 7.40144530938024e-05, + "loss": 0.8929, + "step": 13124 + }, + { + "epoch": 2.3367165242165244, + "grad_norm": 0.7378625869750977, + "learning_rate": 7.400093672410996e-05, + "loss": 0.7749, + "step": 13125 + }, + { + "epoch": 2.3368945868945867, + "grad_norm": 0.8906251192092896, + "learning_rate": 7.398742086381519e-05, + "loss": 1.007, + "step": 13126 + }, + { + "epoch": 2.3370726495726495, + "grad_norm": 0.8324725031852722, + "learning_rate": 7.397390551318283e-05, + "loss": 0.8493, + "step": 13127 + }, + { + "epoch": 2.3372507122507122, + "grad_norm": 0.781080961227417, + "learning_rate": 7.39603906724778e-05, + "loss": 0.696, + "step": 13128 + }, + { + "epoch": 2.337428774928775, + "grad_norm": 0.8068976402282715, + "learning_rate": 7.394687634196476e-05, + "loss": 0.7196, + "step": 13129 + }, + { + "epoch": 2.3376068376068377, + "grad_norm": 0.7588358521461487, + "learning_rate": 7.393336252190854e-05, + "loss": 0.9179, + "step": 13130 + }, + { + "epoch": 2.3377849002849, + "grad_norm": 0.8334088325500488, + "learning_rate": 7.391984921257398e-05, + "loss": 0.94, + "step": 13131 + }, + { + "epoch": 2.337962962962963, + "grad_norm": 0.9485353231430054, + "learning_rate": 7.390633641422578e-05, + "loss": 0.9253, + "step": 13132 + }, + { + "epoch": 2.3381410256410255, + "grad_norm": 0.9447978734970093, + "learning_rate": 7.389282412712874e-05, + "loss": 0.9112, + "step": 13133 + }, + { + "epoch": 2.3383190883190883, + "grad_norm": 0.7348376512527466, + "learning_rate": 7.387931235154754e-05, + "loss": 0.7817, + "step": 13134 + }, + { + "epoch": 2.338497150997151, + "grad_norm": 0.8610092401504517, + "learning_rate": 7.386580108774699e-05, + "loss": 0.8231, + "step": 13135 + }, + { + "epoch": 2.338675213675214, + "grad_norm": 0.8314286470413208, + "learning_rate": 7.385229033599175e-05, + "loss": 0.7323, + "step": 13136 + }, + { + "epoch": 2.3388532763532766, + "grad_norm": 0.7775855660438538, + "learning_rate": 7.383878009654657e-05, + "loss": 0.9897, + "step": 13137 + }, + { + "epoch": 2.339031339031339, + "grad_norm": 0.8140097260475159, + "learning_rate": 7.382527036967614e-05, + "loss": 0.7815, + "step": 13138 + }, + { + "epoch": 2.3392094017094016, + "grad_norm": 0.8154003620147705, + "learning_rate": 7.38117611556452e-05, + "loss": 0.7646, + "step": 13139 + }, + { + "epoch": 2.3393874643874644, + "grad_norm": 0.7705643177032471, + "learning_rate": 7.379825245471836e-05, + "loss": 0.7633, + "step": 13140 + }, + { + "epoch": 2.339565527065527, + "grad_norm": 0.7856985330581665, + "learning_rate": 7.378474426716035e-05, + "loss": 0.6803, + "step": 13141 + }, + { + "epoch": 2.33974358974359, + "grad_norm": 0.8384547233581543, + "learning_rate": 7.377123659323579e-05, + "loss": 0.9092, + "step": 13142 + }, + { + "epoch": 2.339921652421652, + "grad_norm": 0.7456032633781433, + "learning_rate": 7.375772943320942e-05, + "loss": 0.8393, + "step": 13143 + }, + { + "epoch": 2.340099715099715, + "grad_norm": 0.9527342319488525, + "learning_rate": 7.374422278734579e-05, + "loss": 1.0272, + "step": 13144 + }, + { + "epoch": 2.3402777777777777, + "grad_norm": 0.8976300954818726, + "learning_rate": 7.37307166559096e-05, + "loss": 0.7184, + "step": 13145 + }, + { + "epoch": 2.3404558404558404, + "grad_norm": 0.7698291540145874, + "learning_rate": 7.371721103916542e-05, + "loss": 0.6783, + "step": 13146 + }, + { + "epoch": 2.340633903133903, + "grad_norm": 0.8646810054779053, + "learning_rate": 7.37037059373779e-05, + "loss": 0.8559, + "step": 13147 + }, + { + "epoch": 2.340811965811966, + "grad_norm": 0.7534750699996948, + "learning_rate": 7.369020135081161e-05, + "loss": 0.8087, + "step": 13148 + }, + { + "epoch": 2.3409900284900287, + "grad_norm": 0.7408546209335327, + "learning_rate": 7.367669727973123e-05, + "loss": 0.675, + "step": 13149 + }, + { + "epoch": 2.341168091168091, + "grad_norm": 0.8753145933151245, + "learning_rate": 7.366319372440124e-05, + "loss": 0.8163, + "step": 13150 + }, + { + "epoch": 2.3413461538461537, + "grad_norm": 0.7065265774726868, + "learning_rate": 7.364969068508624e-05, + "loss": 0.5786, + "step": 13151 + }, + { + "epoch": 2.3415242165242165, + "grad_norm": 0.7976117730140686, + "learning_rate": 7.363618816205087e-05, + "loss": 0.9053, + "step": 13152 + }, + { + "epoch": 2.3417022792022792, + "grad_norm": 0.7261707782745361, + "learning_rate": 7.362268615555958e-05, + "loss": 0.7677, + "step": 13153 + }, + { + "epoch": 2.341880341880342, + "grad_norm": 0.7868889570236206, + "learning_rate": 7.360918466587701e-05, + "loss": 0.8648, + "step": 13154 + }, + { + "epoch": 2.3420584045584047, + "grad_norm": 0.8473666310310364, + "learning_rate": 7.35956836932676e-05, + "loss": 0.8533, + "step": 13155 + }, + { + "epoch": 2.342236467236467, + "grad_norm": 0.7456569671630859, + "learning_rate": 7.358218323799594e-05, + "loss": 0.7617, + "step": 13156 + }, + { + "epoch": 2.34241452991453, + "grad_norm": 0.8130928874015808, + "learning_rate": 7.356868330032652e-05, + "loss": 0.8667, + "step": 13157 + }, + { + "epoch": 2.3425925925925926, + "grad_norm": 0.8743309378623962, + "learning_rate": 7.355518388052384e-05, + "loss": 0.9196, + "step": 13158 + }, + { + "epoch": 2.3427706552706553, + "grad_norm": 0.8228809237480164, + "learning_rate": 7.354168497885237e-05, + "loss": 0.6509, + "step": 13159 + }, + { + "epoch": 2.342948717948718, + "grad_norm": 0.6998807191848755, + "learning_rate": 7.352818659557668e-05, + "loss": 0.5762, + "step": 13160 + }, + { + "epoch": 2.343126780626781, + "grad_norm": 0.8757675290107727, + "learning_rate": 7.351468873096114e-05, + "loss": 0.9094, + "step": 13161 + }, + { + "epoch": 2.343304843304843, + "grad_norm": 0.7495744824409485, + "learning_rate": 7.350119138527026e-05, + "loss": 0.653, + "step": 13162 + }, + { + "epoch": 2.343482905982906, + "grad_norm": 0.8229764103889465, + "learning_rate": 7.348769455876849e-05, + "loss": 0.8146, + "step": 13163 + }, + { + "epoch": 2.3436609686609686, + "grad_norm": 0.8317791819572449, + "learning_rate": 7.347419825172029e-05, + "loss": 0.7754, + "step": 13164 + }, + { + "epoch": 2.3438390313390314, + "grad_norm": 0.8210344910621643, + "learning_rate": 7.346070246439005e-05, + "loss": 0.876, + "step": 13165 + }, + { + "epoch": 2.344017094017094, + "grad_norm": 0.7711526155471802, + "learning_rate": 7.344720719704223e-05, + "loss": 0.7426, + "step": 13166 + }, + { + "epoch": 2.344195156695157, + "grad_norm": 0.8231741189956665, + "learning_rate": 7.343371244994119e-05, + "loss": 0.8992, + "step": 13167 + }, + { + "epoch": 2.344373219373219, + "grad_norm": 0.7145521640777588, + "learning_rate": 7.342021822335143e-05, + "loss": 0.8787, + "step": 13168 + }, + { + "epoch": 2.344551282051282, + "grad_norm": 0.8323171734809875, + "learning_rate": 7.340672451753723e-05, + "loss": 0.7035, + "step": 13169 + }, + { + "epoch": 2.3447293447293447, + "grad_norm": 0.7061881422996521, + "learning_rate": 7.339323133276301e-05, + "loss": 0.8077, + "step": 13170 + }, + { + "epoch": 2.3449074074074074, + "grad_norm": 0.8705938458442688, + "learning_rate": 7.33797386692932e-05, + "loss": 0.8616, + "step": 13171 + }, + { + "epoch": 2.34508547008547, + "grad_norm": 0.8777729868888855, + "learning_rate": 7.336624652739208e-05, + "loss": 0.9524, + "step": 13172 + }, + { + "epoch": 2.345263532763533, + "grad_norm": 0.9099276065826416, + "learning_rate": 7.335275490732406e-05, + "loss": 0.8248, + "step": 13173 + }, + { + "epoch": 2.3454415954415953, + "grad_norm": 0.7963444590568542, + "learning_rate": 7.333926380935341e-05, + "loss": 0.78, + "step": 13174 + }, + { + "epoch": 2.345619658119658, + "grad_norm": 0.9400636553764343, + "learning_rate": 7.332577323374454e-05, + "loss": 1.0062, + "step": 13175 + }, + { + "epoch": 2.3457977207977208, + "grad_norm": 0.7794054746627808, + "learning_rate": 7.331228318076171e-05, + "loss": 0.8564, + "step": 13176 + }, + { + "epoch": 2.3459757834757835, + "grad_norm": 0.7767263054847717, + "learning_rate": 7.329879365066927e-05, + "loss": 0.8191, + "step": 13177 + }, + { + "epoch": 2.3461538461538463, + "grad_norm": 0.7170942425727844, + "learning_rate": 7.328530464373148e-05, + "loss": 0.7018, + "step": 13178 + }, + { + "epoch": 2.346331908831909, + "grad_norm": 0.8246886134147644, + "learning_rate": 7.327181616021268e-05, + "loss": 0.8498, + "step": 13179 + }, + { + "epoch": 2.3465099715099713, + "grad_norm": 0.9531362652778625, + "learning_rate": 7.325832820037711e-05, + "loss": 0.7031, + "step": 13180 + }, + { + "epoch": 2.346688034188034, + "grad_norm": 0.8561878204345703, + "learning_rate": 7.324484076448905e-05, + "loss": 0.8627, + "step": 13181 + }, + { + "epoch": 2.346866096866097, + "grad_norm": 0.7890949845314026, + "learning_rate": 7.323135385281274e-05, + "loss": 0.7675, + "step": 13182 + }, + { + "epoch": 2.3470441595441596, + "grad_norm": 0.72523033618927, + "learning_rate": 7.321786746561246e-05, + "loss": 0.8847, + "step": 13183 + }, + { + "epoch": 2.3472222222222223, + "grad_norm": 0.7866469025611877, + "learning_rate": 7.32043816031524e-05, + "loss": 0.9657, + "step": 13184 + }, + { + "epoch": 2.347400284900285, + "grad_norm": 0.8669828176498413, + "learning_rate": 7.319089626569687e-05, + "loss": 0.9098, + "step": 13185 + }, + { + "epoch": 2.347578347578348, + "grad_norm": 0.7874458432197571, + "learning_rate": 7.317741145351e-05, + "loss": 0.9545, + "step": 13186 + }, + { + "epoch": 2.34775641025641, + "grad_norm": 0.7924689054489136, + "learning_rate": 7.316392716685604e-05, + "loss": 0.8577, + "step": 13187 + }, + { + "epoch": 2.347934472934473, + "grad_norm": 0.731119692325592, + "learning_rate": 7.315044340599918e-05, + "loss": 0.9251, + "step": 13188 + }, + { + "epoch": 2.3481125356125356, + "grad_norm": 0.914900004863739, + "learning_rate": 7.313696017120361e-05, + "loss": 0.9224, + "step": 13189 + }, + { + "epoch": 2.3482905982905984, + "grad_norm": 0.7616490125656128, + "learning_rate": 7.312347746273349e-05, + "loss": 0.7263, + "step": 13190 + }, + { + "epoch": 2.348468660968661, + "grad_norm": 0.8357210159301758, + "learning_rate": 7.310999528085301e-05, + "loss": 0.8572, + "step": 13191 + }, + { + "epoch": 2.3486467236467234, + "grad_norm": 0.8404232263565063, + "learning_rate": 7.309651362582633e-05, + "loss": 0.6822, + "step": 13192 + }, + { + "epoch": 2.348824786324786, + "grad_norm": 0.8992070555686951, + "learning_rate": 7.308303249791754e-05, + "loss": 0.91, + "step": 13193 + }, + { + "epoch": 2.349002849002849, + "grad_norm": 0.8150524497032166, + "learning_rate": 7.306955189739084e-05, + "loss": 0.984, + "step": 13194 + }, + { + "epoch": 2.3491809116809117, + "grad_norm": 0.9042861461639404, + "learning_rate": 7.305607182451031e-05, + "loss": 1.0111, + "step": 13195 + }, + { + "epoch": 2.3493589743589745, + "grad_norm": 0.8402968049049377, + "learning_rate": 7.30425922795401e-05, + "loss": 0.801, + "step": 13196 + }, + { + "epoch": 2.349537037037037, + "grad_norm": 0.7742997407913208, + "learning_rate": 7.302911326274428e-05, + "loss": 0.659, + "step": 13197 + }, + { + "epoch": 2.3497150997151, + "grad_norm": 0.8005271553993225, + "learning_rate": 7.301563477438698e-05, + "loss": 0.8549, + "step": 13198 + }, + { + "epoch": 2.3498931623931623, + "grad_norm": 0.8253805637359619, + "learning_rate": 7.300215681473224e-05, + "loss": 0.9049, + "step": 13199 + }, + { + "epoch": 2.350071225071225, + "grad_norm": 0.8539033532142639, + "learning_rate": 7.29886793840442e-05, + "loss": 0.9359, + "step": 13200 + }, + { + "epoch": 2.3502492877492878, + "grad_norm": 0.827608048915863, + "learning_rate": 7.297520248258681e-05, + "loss": 1.0105, + "step": 13201 + }, + { + "epoch": 2.3504273504273505, + "grad_norm": 0.8418487310409546, + "learning_rate": 7.296172611062422e-05, + "loss": 0.8138, + "step": 13202 + }, + { + "epoch": 2.3506054131054133, + "grad_norm": 0.7853255867958069, + "learning_rate": 7.294825026842042e-05, + "loss": 0.9279, + "step": 13203 + }, + { + "epoch": 2.3507834757834756, + "grad_norm": 0.8454880714416504, + "learning_rate": 7.293477495623951e-05, + "loss": 0.7687, + "step": 13204 + }, + { + "epoch": 2.3509615384615383, + "grad_norm": 0.7620453238487244, + "learning_rate": 7.29213001743454e-05, + "loss": 0.7567, + "step": 13205 + }, + { + "epoch": 2.351139601139601, + "grad_norm": 0.8993792533874512, + "learning_rate": 7.290782592300223e-05, + "loss": 0.9716, + "step": 13206 + }, + { + "epoch": 2.351317663817664, + "grad_norm": 1.1063668727874756, + "learning_rate": 7.289435220247387e-05, + "loss": 0.9763, + "step": 13207 + }, + { + "epoch": 2.3514957264957266, + "grad_norm": 0.8205364346504211, + "learning_rate": 7.288087901302439e-05, + "loss": 0.9395, + "step": 13208 + }, + { + "epoch": 2.3516737891737893, + "grad_norm": 0.680487871170044, + "learning_rate": 7.286740635491774e-05, + "loss": 0.6252, + "step": 13209 + }, + { + "epoch": 2.351851851851852, + "grad_norm": 0.8450767397880554, + "learning_rate": 7.285393422841791e-05, + "loss": 0.8707, + "step": 13210 + }, + { + "epoch": 2.3520299145299144, + "grad_norm": 0.6871187686920166, + "learning_rate": 7.284046263378888e-05, + "loss": 0.5695, + "step": 13211 + }, + { + "epoch": 2.352207977207977, + "grad_norm": 0.7968555688858032, + "learning_rate": 7.282699157129451e-05, + "loss": 0.7014, + "step": 13212 + }, + { + "epoch": 2.35238603988604, + "grad_norm": 0.863798201084137, + "learning_rate": 7.281352104119883e-05, + "loss": 0.9241, + "step": 13213 + }, + { + "epoch": 2.3525641025641026, + "grad_norm": 0.8848825693130493, + "learning_rate": 7.28000510437657e-05, + "loss": 0.8252, + "step": 13214 + }, + { + "epoch": 2.3527421652421654, + "grad_norm": 0.7528855800628662, + "learning_rate": 7.278658157925912e-05, + "loss": 0.7428, + "step": 13215 + }, + { + "epoch": 2.3529202279202277, + "grad_norm": 0.7636159062385559, + "learning_rate": 7.277311264794288e-05, + "loss": 0.8952, + "step": 13216 + }, + { + "epoch": 2.3530982905982905, + "grad_norm": 1.0585514307022095, + "learning_rate": 7.2759644250081e-05, + "loss": 0.9153, + "step": 13217 + }, + { + "epoch": 2.353276353276353, + "grad_norm": 0.7691277265548706, + "learning_rate": 7.274617638593725e-05, + "loss": 1.057, + "step": 13218 + }, + { + "epoch": 2.353454415954416, + "grad_norm": 0.8324813842773438, + "learning_rate": 7.273270905577561e-05, + "loss": 0.9253, + "step": 13219 + }, + { + "epoch": 2.3536324786324787, + "grad_norm": 0.835491418838501, + "learning_rate": 7.271924225985984e-05, + "loss": 1.0103, + "step": 13220 + }, + { + "epoch": 2.3538105413105415, + "grad_norm": 0.8318347930908203, + "learning_rate": 7.270577599845389e-05, + "loss": 0.8896, + "step": 13221 + }, + { + "epoch": 2.353988603988604, + "grad_norm": 0.7801460027694702, + "learning_rate": 7.269231027182153e-05, + "loss": 0.9274, + "step": 13222 + }, + { + "epoch": 2.3541666666666665, + "grad_norm": 0.8195397257804871, + "learning_rate": 7.267884508022665e-05, + "loss": 0.8126, + "step": 13223 + }, + { + "epoch": 2.3543447293447293, + "grad_norm": 0.7978246212005615, + "learning_rate": 7.2665380423933e-05, + "loss": 0.8426, + "step": 13224 + }, + { + "epoch": 2.354522792022792, + "grad_norm": 0.7614684104919434, + "learning_rate": 7.265191630320452e-05, + "loss": 0.7277, + "step": 13225 + }, + { + "epoch": 2.3547008547008548, + "grad_norm": 0.8684967756271362, + "learning_rate": 7.263845271830485e-05, + "loss": 0.7054, + "step": 13226 + }, + { + "epoch": 2.3548789173789175, + "grad_norm": 0.878842830657959, + "learning_rate": 7.262498966949791e-05, + "loss": 1.0478, + "step": 13227 + }, + { + "epoch": 2.35505698005698, + "grad_norm": 0.8321235179901123, + "learning_rate": 7.26115271570474e-05, + "loss": 0.8119, + "step": 13228 + }, + { + "epoch": 2.3552350427350426, + "grad_norm": 0.9144030213356018, + "learning_rate": 7.259806518121713e-05, + "loss": 0.8626, + "step": 13229 + }, + { + "epoch": 2.3554131054131053, + "grad_norm": 0.8437082767486572, + "learning_rate": 7.258460374227085e-05, + "loss": 0.8439, + "step": 13230 + }, + { + "epoch": 2.355591168091168, + "grad_norm": 0.8405697345733643, + "learning_rate": 7.257114284047229e-05, + "loss": 0.863, + "step": 13231 + }, + { + "epoch": 2.355769230769231, + "grad_norm": 0.8674731850624084, + "learning_rate": 7.255768247608525e-05, + "loss": 0.9823, + "step": 13232 + }, + { + "epoch": 2.3559472934472936, + "grad_norm": 0.8844531178474426, + "learning_rate": 7.254422264937337e-05, + "loss": 0.9018, + "step": 13233 + }, + { + "epoch": 2.3561253561253563, + "grad_norm": 0.8394746780395508, + "learning_rate": 7.253076336060045e-05, + "loss": 0.8407, + "step": 13234 + }, + { + "epoch": 2.3563034188034186, + "grad_norm": 0.8759872317314148, + "learning_rate": 7.251730461003012e-05, + "loss": 0.841, + "step": 13235 + }, + { + "epoch": 2.3564814814814814, + "grad_norm": 0.7240089774131775, + "learning_rate": 7.250384639792617e-05, + "loss": 0.7918, + "step": 13236 + }, + { + "epoch": 2.356659544159544, + "grad_norm": 0.8619599342346191, + "learning_rate": 7.24903887245522e-05, + "loss": 0.761, + "step": 13237 + }, + { + "epoch": 2.356837606837607, + "grad_norm": 0.7291443943977356, + "learning_rate": 7.247693159017192e-05, + "loss": 0.8189, + "step": 13238 + }, + { + "epoch": 2.3570156695156697, + "grad_norm": 0.8006066083908081, + "learning_rate": 7.246347499504898e-05, + "loss": 0.8924, + "step": 13239 + }, + { + "epoch": 2.357193732193732, + "grad_norm": 0.7774627208709717, + "learning_rate": 7.245001893944707e-05, + "loss": 0.946, + "step": 13240 + }, + { + "epoch": 2.3573717948717947, + "grad_norm": 0.7643784284591675, + "learning_rate": 7.243656342362978e-05, + "loss": 0.8717, + "step": 13241 + }, + { + "epoch": 2.3575498575498575, + "grad_norm": 0.7197792530059814, + "learning_rate": 7.242310844786082e-05, + "loss": 0.6792, + "step": 13242 + }, + { + "epoch": 2.35772792022792, + "grad_norm": 0.9124938249588013, + "learning_rate": 7.240965401240371e-05, + "loss": 0.841, + "step": 13243 + }, + { + "epoch": 2.357905982905983, + "grad_norm": 0.7350388765335083, + "learning_rate": 7.239620011752215e-05, + "loss": 0.8294, + "step": 13244 + }, + { + "epoch": 2.3580840455840457, + "grad_norm": 0.8814936280250549, + "learning_rate": 7.238274676347967e-05, + "loss": 0.9732, + "step": 13245 + }, + { + "epoch": 2.3582621082621085, + "grad_norm": 0.8379302024841309, + "learning_rate": 7.236929395053995e-05, + "loss": 0.8896, + "step": 13246 + }, + { + "epoch": 2.3584401709401708, + "grad_norm": 0.8200546503067017, + "learning_rate": 7.235584167896648e-05, + "loss": 0.7991, + "step": 13247 + }, + { + "epoch": 2.3586182336182335, + "grad_norm": 0.7842608690261841, + "learning_rate": 7.234238994902287e-05, + "loss": 0.695, + "step": 13248 + }, + { + "epoch": 2.3587962962962963, + "grad_norm": 0.8872218132019043, + "learning_rate": 7.232893876097266e-05, + "loss": 0.8611, + "step": 13249 + }, + { + "epoch": 2.358974358974359, + "grad_norm": 0.8358500599861145, + "learning_rate": 7.231548811507942e-05, + "loss": 0.7829, + "step": 13250 + }, + { + "epoch": 2.359152421652422, + "grad_norm": 0.8269400000572205, + "learning_rate": 7.23020380116067e-05, + "loss": 0.6904, + "step": 13251 + }, + { + "epoch": 2.359330484330484, + "grad_norm": 0.8693541288375854, + "learning_rate": 7.2288588450818e-05, + "loss": 0.8659, + "step": 13252 + }, + { + "epoch": 2.359508547008547, + "grad_norm": 0.858076810836792, + "learning_rate": 7.227513943297688e-05, + "loss": 0.8824, + "step": 13253 + }, + { + "epoch": 2.3596866096866096, + "grad_norm": 0.796541154384613, + "learning_rate": 7.226169095834675e-05, + "loss": 0.8999, + "step": 13254 + }, + { + "epoch": 2.3598646723646723, + "grad_norm": 0.7692779898643494, + "learning_rate": 7.22482430271912e-05, + "loss": 0.9492, + "step": 13255 + }, + { + "epoch": 2.360042735042735, + "grad_norm": 0.9259434342384338, + "learning_rate": 7.223479563977364e-05, + "loss": 0.9115, + "step": 13256 + }, + { + "epoch": 2.360220797720798, + "grad_norm": 0.9048989415168762, + "learning_rate": 7.222134879635764e-05, + "loss": 0.8057, + "step": 13257 + }, + { + "epoch": 2.3603988603988606, + "grad_norm": 0.9342616200447083, + "learning_rate": 7.220790249720656e-05, + "loss": 0.8554, + "step": 13258 + }, + { + "epoch": 2.360576923076923, + "grad_norm": 0.7747787237167358, + "learning_rate": 7.219445674258392e-05, + "loss": 0.7555, + "step": 13259 + }, + { + "epoch": 2.3607549857549857, + "grad_norm": 0.805437445640564, + "learning_rate": 7.218101153275311e-05, + "loss": 0.7442, + "step": 13260 + }, + { + "epoch": 2.3609330484330484, + "grad_norm": 0.9797805547714233, + "learning_rate": 7.216756686797764e-05, + "loss": 1.0975, + "step": 13261 + }, + { + "epoch": 2.361111111111111, + "grad_norm": 0.7361458539962769, + "learning_rate": 7.215412274852083e-05, + "loss": 0.6597, + "step": 13262 + }, + { + "epoch": 2.361289173789174, + "grad_norm": 0.8041569590568542, + "learning_rate": 7.21406791746462e-05, + "loss": 0.8343, + "step": 13263 + }, + { + "epoch": 2.361467236467236, + "grad_norm": 0.8364384770393372, + "learning_rate": 7.212723614661703e-05, + "loss": 0.9486, + "step": 13264 + }, + { + "epoch": 2.361645299145299, + "grad_norm": 0.714241623878479, + "learning_rate": 7.21137936646968e-05, + "loss": 0.5978, + "step": 13265 + }, + { + "epoch": 2.3618233618233617, + "grad_norm": 0.8830710053443909, + "learning_rate": 7.210035172914882e-05, + "loss": 0.9584, + "step": 13266 + }, + { + "epoch": 2.3620014245014245, + "grad_norm": 0.714112401008606, + "learning_rate": 7.208691034023653e-05, + "loss": 0.8878, + "step": 13267 + }, + { + "epoch": 2.3621794871794872, + "grad_norm": 0.7654083371162415, + "learning_rate": 7.207346949822322e-05, + "loss": 0.822, + "step": 13268 + }, + { + "epoch": 2.36235754985755, + "grad_norm": 0.772693395614624, + "learning_rate": 7.206002920337225e-05, + "loss": 0.7993, + "step": 13269 + }, + { + "epoch": 2.3625356125356127, + "grad_norm": 0.9678596258163452, + "learning_rate": 7.2046589455947e-05, + "loss": 0.948, + "step": 13270 + }, + { + "epoch": 2.362713675213675, + "grad_norm": 0.8254278302192688, + "learning_rate": 7.203315025621073e-05, + "loss": 0.8654, + "step": 13271 + }, + { + "epoch": 2.362891737891738, + "grad_norm": 0.7527315020561218, + "learning_rate": 7.201971160442685e-05, + "loss": 0.6881, + "step": 13272 + }, + { + "epoch": 2.3630698005698005, + "grad_norm": 0.7658267021179199, + "learning_rate": 7.200627350085853e-05, + "loss": 0.7332, + "step": 13273 + }, + { + "epoch": 2.3632478632478633, + "grad_norm": 0.8590806126594543, + "learning_rate": 7.199283594576916e-05, + "loss": 0.879, + "step": 13274 + }, + { + "epoch": 2.363425925925926, + "grad_norm": 0.7533347606658936, + "learning_rate": 7.197939893942197e-05, + "loss": 0.8738, + "step": 13275 + }, + { + "epoch": 2.363603988603989, + "grad_norm": Infinity, + "learning_rate": 7.197939893942197e-05, + "loss": 0.7641, + "step": 13276 + }, + { + "epoch": 2.363782051282051, + "grad_norm": 0.6873685121536255, + "learning_rate": 7.196596248208029e-05, + "loss": 0.4708, + "step": 13277 + }, + { + "epoch": 2.363960113960114, + "grad_norm": 0.7659112215042114, + "learning_rate": 7.195252657400729e-05, + "loss": 0.839, + "step": 13278 + }, + { + "epoch": 2.3641381766381766, + "grad_norm": 0.8355028629302979, + "learning_rate": 7.193909121546631e-05, + "loss": 0.9792, + "step": 13279 + }, + { + "epoch": 2.3643162393162394, + "grad_norm": 0.9633997678756714, + "learning_rate": 7.192565640672052e-05, + "loss": 0.9891, + "step": 13280 + }, + { + "epoch": 2.364494301994302, + "grad_norm": 0.7984298467636108, + "learning_rate": 7.191222214803318e-05, + "loss": 0.8343, + "step": 13281 + }, + { + "epoch": 2.364672364672365, + "grad_norm": 0.8239994645118713, + "learning_rate": 7.189878843966749e-05, + "loss": 0.8586, + "step": 13282 + }, + { + "epoch": 2.364850427350427, + "grad_norm": 0.8695420026779175, + "learning_rate": 7.188535528188671e-05, + "loss": 0.9161, + "step": 13283 + }, + { + "epoch": 2.36502849002849, + "grad_norm": 0.8272924423217773, + "learning_rate": 7.187192267495393e-05, + "loss": 0.8158, + "step": 13284 + }, + { + "epoch": 2.3652065527065527, + "grad_norm": 0.8217222690582275, + "learning_rate": 7.185849061913243e-05, + "loss": 0.892, + "step": 13285 + }, + { + "epoch": 2.3653846153846154, + "grad_norm": 0.9041243195533752, + "learning_rate": 7.184505911468532e-05, + "loss": 0.9093, + "step": 13286 + }, + { + "epoch": 2.365562678062678, + "grad_norm": 0.8325521349906921, + "learning_rate": 7.183162816187582e-05, + "loss": 0.7546, + "step": 13287 + }, + { + "epoch": 2.365740740740741, + "grad_norm": 0.9160267114639282, + "learning_rate": 7.181819776096704e-05, + "loss": 0.9662, + "step": 13288 + }, + { + "epoch": 2.3659188034188032, + "grad_norm": 0.8771381974220276, + "learning_rate": 7.180476791222215e-05, + "loss": 1.0083, + "step": 13289 + }, + { + "epoch": 2.366096866096866, + "grad_norm": 0.8251327872276306, + "learning_rate": 7.179133861590421e-05, + "loss": 0.8209, + "step": 13290 + }, + { + "epoch": 2.3662749287749287, + "grad_norm": 0.8760706186294556, + "learning_rate": 7.177790987227641e-05, + "loss": 0.7479, + "step": 13291 + }, + { + "epoch": 2.3664529914529915, + "grad_norm": 0.7857288122177124, + "learning_rate": 7.176448168160187e-05, + "loss": 0.6511, + "step": 13292 + }, + { + "epoch": 2.3666310541310542, + "grad_norm": 0.9548102021217346, + "learning_rate": 7.175105404414362e-05, + "loss": 0.731, + "step": 13293 + }, + { + "epoch": 2.366809116809117, + "grad_norm": 0.7604304552078247, + "learning_rate": 7.173762696016484e-05, + "loss": 0.8212, + "step": 13294 + }, + { + "epoch": 2.3669871794871793, + "grad_norm": 0.9121061563491821, + "learning_rate": 7.172420042992849e-05, + "loss": 0.939, + "step": 13295 + }, + { + "epoch": 2.367165242165242, + "grad_norm": 0.8128613233566284, + "learning_rate": 7.171077445369772e-05, + "loss": 0.8908, + "step": 13296 + }, + { + "epoch": 2.367343304843305, + "grad_norm": 0.9184401035308838, + "learning_rate": 7.169734903173555e-05, + "loss": 0.957, + "step": 13297 + }, + { + "epoch": 2.3675213675213675, + "grad_norm": 0.9234427809715271, + "learning_rate": 7.168392416430507e-05, + "loss": 0.8403, + "step": 13298 + }, + { + "epoch": 2.3676994301994303, + "grad_norm": 0.8810806274414062, + "learning_rate": 7.167049985166922e-05, + "loss": 0.9754, + "step": 13299 + }, + { + "epoch": 2.367877492877493, + "grad_norm": 0.8208937048912048, + "learning_rate": 7.165707609409113e-05, + "loss": 0.9418, + "step": 13300 + }, + { + "epoch": 2.3680555555555554, + "grad_norm": 0.8666219711303711, + "learning_rate": 7.164365289183371e-05, + "loss": 0.8936, + "step": 13301 + }, + { + "epoch": 2.368233618233618, + "grad_norm": 0.9385154843330383, + "learning_rate": 7.163023024516002e-05, + "loss": 0.8158, + "step": 13302 + }, + { + "epoch": 2.368411680911681, + "grad_norm": 1.0415911674499512, + "learning_rate": 7.161680815433303e-05, + "loss": 1.0445, + "step": 13303 + }, + { + "epoch": 2.3685897435897436, + "grad_norm": 0.6882192492485046, + "learning_rate": 7.160338661961577e-05, + "loss": 0.4929, + "step": 13304 + }, + { + "epoch": 2.3687678062678064, + "grad_norm": 0.8695144653320312, + "learning_rate": 7.15899656412711e-05, + "loss": 0.8991, + "step": 13305 + }, + { + "epoch": 2.368945868945869, + "grad_norm": 0.8973569273948669, + "learning_rate": 7.157654521956206e-05, + "loss": 0.8423, + "step": 13306 + }, + { + "epoch": 2.369123931623932, + "grad_norm": 0.7656881213188171, + "learning_rate": 7.156312535475155e-05, + "loss": 0.7351, + "step": 13307 + }, + { + "epoch": 2.369301994301994, + "grad_norm": 0.8023402690887451, + "learning_rate": 7.154970604710258e-05, + "loss": 0.9943, + "step": 13308 + }, + { + "epoch": 2.369480056980057, + "grad_norm": 0.916946530342102, + "learning_rate": 7.153628729687797e-05, + "loss": 0.8649, + "step": 13309 + }, + { + "epoch": 2.3696581196581197, + "grad_norm": 0.8764750361442566, + "learning_rate": 7.152286910434068e-05, + "loss": 0.9799, + "step": 13310 + }, + { + "epoch": 2.3698361823361824, + "grad_norm": 0.8732671737670898, + "learning_rate": 7.150945146975364e-05, + "loss": 1.0431, + "step": 13311 + }, + { + "epoch": 2.370014245014245, + "grad_norm": 0.8447144031524658, + "learning_rate": 7.149603439337969e-05, + "loss": 0.7805, + "step": 13312 + }, + { + "epoch": 2.3701923076923075, + "grad_norm": 0.9017399549484253, + "learning_rate": 7.148261787548178e-05, + "loss": 0.8102, + "step": 13313 + }, + { + "epoch": 2.3703703703703702, + "grad_norm": 0.7187124490737915, + "learning_rate": 7.14692019163227e-05, + "loss": 0.7327, + "step": 13314 + }, + { + "epoch": 2.370548433048433, + "grad_norm": 0.8579949736595154, + "learning_rate": 7.145578651616536e-05, + "loss": 0.8685, + "step": 13315 + }, + { + "epoch": 2.3707264957264957, + "grad_norm": 0.6088887453079224, + "learning_rate": 7.144237167527256e-05, + "loss": 0.7004, + "step": 13316 + }, + { + "epoch": 2.3709045584045585, + "grad_norm": 0.6400231719017029, + "learning_rate": 7.142895739390718e-05, + "loss": 0.7273, + "step": 13317 + }, + { + "epoch": 2.3710826210826212, + "grad_norm": 0.8680049180984497, + "learning_rate": 7.141554367233201e-05, + "loss": 0.7886, + "step": 13318 + }, + { + "epoch": 2.371260683760684, + "grad_norm": 0.8894832134246826, + "learning_rate": 7.140213051080991e-05, + "loss": 1.0597, + "step": 13319 + }, + { + "epoch": 2.3714387464387463, + "grad_norm": 0.7371698021888733, + "learning_rate": 7.138871790960365e-05, + "loss": 0.8344, + "step": 13320 + }, + { + "epoch": 2.371616809116809, + "grad_norm": 0.7396906018257141, + "learning_rate": 7.137530586897601e-05, + "loss": 0.7185, + "step": 13321 + }, + { + "epoch": 2.371794871794872, + "grad_norm": 0.7884365320205688, + "learning_rate": 7.136189438918978e-05, + "loss": 0.8311, + "step": 13322 + }, + { + "epoch": 2.3719729344729346, + "grad_norm": 0.8064826130867004, + "learning_rate": 7.13484834705078e-05, + "loss": 0.6933, + "step": 13323 + }, + { + "epoch": 2.3721509971509973, + "grad_norm": 0.8865584135055542, + "learning_rate": 7.13350731131927e-05, + "loss": 0.979, + "step": 13324 + }, + { + "epoch": 2.3723290598290596, + "grad_norm": 0.7782325148582458, + "learning_rate": 7.132166331750736e-05, + "loss": 0.8147, + "step": 13325 + }, + { + "epoch": 2.3725071225071224, + "grad_norm": 0.8515480160713196, + "learning_rate": 7.13082540837144e-05, + "loss": 0.7571, + "step": 13326 + }, + { + "epoch": 2.372685185185185, + "grad_norm": 0.8665108680725098, + "learning_rate": 7.129484541207662e-05, + "loss": 0.8171, + "step": 13327 + }, + { + "epoch": 2.372863247863248, + "grad_norm": 0.7640653252601624, + "learning_rate": 7.128143730285668e-05, + "loss": 0.7118, + "step": 13328 + }, + { + "epoch": 2.3730413105413106, + "grad_norm": 0.844083309173584, + "learning_rate": 7.126802975631735e-05, + "loss": 0.8394, + "step": 13329 + }, + { + "epoch": 2.3732193732193734, + "grad_norm": 0.8718371391296387, + "learning_rate": 7.12546227727213e-05, + "loss": 0.8729, + "step": 13330 + }, + { + "epoch": 2.373397435897436, + "grad_norm": 0.7254782319068909, + "learning_rate": 7.124121635233118e-05, + "loss": 0.8178, + "step": 13331 + }, + { + "epoch": 2.3735754985754984, + "grad_norm": 0.7211804389953613, + "learning_rate": 7.12278104954097e-05, + "loss": 0.9415, + "step": 13332 + }, + { + "epoch": 2.373753561253561, + "grad_norm": 0.8538317680358887, + "learning_rate": 7.121440520221949e-05, + "loss": 0.8614, + "step": 13333 + }, + { + "epoch": 2.373931623931624, + "grad_norm": 0.8942680358886719, + "learning_rate": 7.120100047302324e-05, + "loss": 0.985, + "step": 13334 + }, + { + "epoch": 2.3741096866096867, + "grad_norm": 0.8282434344291687, + "learning_rate": 7.118759630808354e-05, + "loss": 0.94, + "step": 13335 + }, + { + "epoch": 2.3742877492877494, + "grad_norm": 0.8036409616470337, + "learning_rate": 7.117419270766308e-05, + "loss": 0.7145, + "step": 13336 + }, + { + "epoch": 2.3744658119658117, + "grad_norm": 0.9169675707817078, + "learning_rate": 7.116078967202437e-05, + "loss": 1.1078, + "step": 13337 + }, + { + "epoch": 2.3746438746438745, + "grad_norm": 0.7805418372154236, + "learning_rate": 7.114738720143011e-05, + "loss": 0.8216, + "step": 13338 + }, + { + "epoch": 2.3748219373219372, + "grad_norm": 1.0444506406784058, + "learning_rate": 7.113398529614285e-05, + "loss": 0.8153, + "step": 13339 + }, + { + "epoch": 2.375, + "grad_norm": 0.8254665732383728, + "learning_rate": 7.112058395642522e-05, + "loss": 0.8127, + "step": 13340 + }, + { + "epoch": 2.3751780626780628, + "grad_norm": 0.8327687382698059, + "learning_rate": 7.11071831825397e-05, + "loss": 0.7014, + "step": 13341 + }, + { + "epoch": 2.3753561253561255, + "grad_norm": 0.7473437786102295, + "learning_rate": 7.109378297474894e-05, + "loss": 0.7621, + "step": 13342 + }, + { + "epoch": 2.3755341880341883, + "grad_norm": 0.8537931442260742, + "learning_rate": 7.108038333331544e-05, + "loss": 0.9302, + "step": 13343 + }, + { + "epoch": 2.3757122507122506, + "grad_norm": 0.81959468126297, + "learning_rate": 7.106698425850178e-05, + "loss": 0.9157, + "step": 13344 + }, + { + "epoch": 2.3758903133903133, + "grad_norm": 0.769257128238678, + "learning_rate": 7.105358575057043e-05, + "loss": 0.8739, + "step": 13345 + }, + { + "epoch": 2.376068376068376, + "grad_norm": 0.7428072690963745, + "learning_rate": 7.104018780978394e-05, + "loss": 0.7001, + "step": 13346 + }, + { + "epoch": 2.376246438746439, + "grad_norm": 0.8152543306350708, + "learning_rate": 7.102679043640481e-05, + "loss": 0.9866, + "step": 13347 + }, + { + "epoch": 2.3764245014245016, + "grad_norm": 0.8732424974441528, + "learning_rate": 7.101339363069556e-05, + "loss": 1.0207, + "step": 13348 + }, + { + "epoch": 2.376602564102564, + "grad_norm": 0.759279191493988, + "learning_rate": 7.099999739291862e-05, + "loss": 0.8703, + "step": 13349 + }, + { + "epoch": 2.3767806267806266, + "grad_norm": 0.8751664161682129, + "learning_rate": 7.098660172333648e-05, + "loss": 0.9805, + "step": 13350 + }, + { + "epoch": 2.3769586894586894, + "grad_norm": 0.9646390080451965, + "learning_rate": 7.097320662221168e-05, + "loss": 0.8623, + "step": 13351 + }, + { + "epoch": 2.377136752136752, + "grad_norm": 0.8626869320869446, + "learning_rate": 7.095981208980652e-05, + "loss": 0.7175, + "step": 13352 + }, + { + "epoch": 2.377314814814815, + "grad_norm": 0.8075738549232483, + "learning_rate": 7.094641812638354e-05, + "loss": 0.7741, + "step": 13353 + }, + { + "epoch": 2.3774928774928776, + "grad_norm": 0.7733559608459473, + "learning_rate": 7.093302473220513e-05, + "loss": 0.8553, + "step": 13354 + }, + { + "epoch": 2.3776709401709404, + "grad_norm": 0.7372797727584839, + "learning_rate": 7.091963190753376e-05, + "loss": 0.8554, + "step": 13355 + }, + { + "epoch": 2.3778490028490027, + "grad_norm": 0.804649293422699, + "learning_rate": 7.090623965263177e-05, + "loss": 0.8704, + "step": 13356 + }, + { + "epoch": 2.3780270655270654, + "grad_norm": 0.8370727300643921, + "learning_rate": 7.089284796776157e-05, + "loss": 0.9786, + "step": 13357 + }, + { + "epoch": 2.378205128205128, + "grad_norm": 0.7565299272537231, + "learning_rate": 7.087945685318554e-05, + "loss": 0.8096, + "step": 13358 + }, + { + "epoch": 2.378383190883191, + "grad_norm": 0.9046086072921753, + "learning_rate": 7.086606630916611e-05, + "loss": 0.8108, + "step": 13359 + }, + { + "epoch": 2.3785612535612537, + "grad_norm": 0.8453067541122437, + "learning_rate": 7.085267633596552e-05, + "loss": 0.8226, + "step": 13360 + }, + { + "epoch": 2.378739316239316, + "grad_norm": 0.8499273061752319, + "learning_rate": 7.083928693384628e-05, + "loss": 1.001, + "step": 13361 + }, + { + "epoch": 2.3789173789173788, + "grad_norm": 0.8358726501464844, + "learning_rate": 7.082589810307055e-05, + "loss": 0.7891, + "step": 13362 + }, + { + "epoch": 2.3790954415954415, + "grad_norm": 0.9156573414802551, + "learning_rate": 7.081250984390078e-05, + "loss": 0.9381, + "step": 13363 + }, + { + "epoch": 2.3792735042735043, + "grad_norm": 0.8704338669776917, + "learning_rate": 7.079912215659923e-05, + "loss": 0.9004, + "step": 13364 + }, + { + "epoch": 2.379451566951567, + "grad_norm": 0.8201949000358582, + "learning_rate": 7.078573504142824e-05, + "loss": 0.7501, + "step": 13365 + }, + { + "epoch": 2.3796296296296298, + "grad_norm": 0.9453420639038086, + "learning_rate": 7.077234849865008e-05, + "loss": 0.9658, + "step": 13366 + }, + { + "epoch": 2.3798076923076925, + "grad_norm": 0.8556796908378601, + "learning_rate": 7.075896252852703e-05, + "loss": 0.8054, + "step": 13367 + }, + { + "epoch": 2.379985754985755, + "grad_norm": 0.7961027026176453, + "learning_rate": 7.074557713132136e-05, + "loss": 0.8065, + "step": 13368 + }, + { + "epoch": 2.3801638176638176, + "grad_norm": 0.8777903318405151, + "learning_rate": 7.073219230729533e-05, + "loss": 0.9399, + "step": 13369 + }, + { + "epoch": 2.3803418803418803, + "grad_norm": 0.8569813370704651, + "learning_rate": 7.071880805671123e-05, + "loss": 0.9424, + "step": 13370 + }, + { + "epoch": 2.380519943019943, + "grad_norm": 0.8810455203056335, + "learning_rate": 7.070542437983123e-05, + "loss": 1.1313, + "step": 13371 + }, + { + "epoch": 2.380698005698006, + "grad_norm": 0.8691363334655762, + "learning_rate": 7.069204127691761e-05, + "loss": 0.9114, + "step": 13372 + }, + { + "epoch": 2.380876068376068, + "grad_norm": 0.7922945618629456, + "learning_rate": 7.067865874823253e-05, + "loss": 0.9158, + "step": 13373 + }, + { + "epoch": 2.381054131054131, + "grad_norm": 0.7465389370918274, + "learning_rate": 7.066527679403825e-05, + "loss": 0.6597, + "step": 13374 + }, + { + "epoch": 2.3812321937321936, + "grad_norm": 0.8386009931564331, + "learning_rate": 7.065189541459689e-05, + "loss": 0.7194, + "step": 13375 + }, + { + "epoch": 2.3814102564102564, + "grad_norm": 0.8633689880371094, + "learning_rate": 7.063851461017073e-05, + "loss": 0.8877, + "step": 13376 + }, + { + "epoch": 2.381588319088319, + "grad_norm": 0.8689528107643127, + "learning_rate": 7.062513438102184e-05, + "loss": 0.8384, + "step": 13377 + }, + { + "epoch": 2.381766381766382, + "grad_norm": 0.7648544311523438, + "learning_rate": 7.061175472741243e-05, + "loss": 0.7669, + "step": 13378 + }, + { + "epoch": 2.3819444444444446, + "grad_norm": 0.8502510786056519, + "learning_rate": 7.059837564960465e-05, + "loss": 0.9379, + "step": 13379 + }, + { + "epoch": 2.382122507122507, + "grad_norm": 0.8277843594551086, + "learning_rate": 7.058499714786063e-05, + "loss": 0.7372, + "step": 13380 + }, + { + "epoch": 2.3823005698005697, + "grad_norm": 0.7394976615905762, + "learning_rate": 7.057161922244246e-05, + "loss": 0.7628, + "step": 13381 + }, + { + "epoch": 2.3824786324786325, + "grad_norm": 0.7906123399734497, + "learning_rate": 7.05582418736123e-05, + "loss": 0.7645, + "step": 13382 + }, + { + "epoch": 2.382656695156695, + "grad_norm": 0.7889885902404785, + "learning_rate": 7.054486510163221e-05, + "loss": 0.8316, + "step": 13383 + }, + { + "epoch": 2.382834757834758, + "grad_norm": 0.7983359098434448, + "learning_rate": 7.053148890676434e-05, + "loss": 0.7925, + "step": 13384 + }, + { + "epoch": 2.3830128205128207, + "grad_norm": 0.9067932963371277, + "learning_rate": 7.051811328927067e-05, + "loss": 0.9385, + "step": 13385 + }, + { + "epoch": 2.383190883190883, + "grad_norm": 0.7210679650306702, + "learning_rate": 7.05047382494134e-05, + "loss": 0.591, + "step": 13386 + }, + { + "epoch": 2.3833689458689458, + "grad_norm": 0.9977821707725525, + "learning_rate": 7.049136378745445e-05, + "loss": 0.8362, + "step": 13387 + }, + { + "epoch": 2.3835470085470085, + "grad_norm": 0.9260198473930359, + "learning_rate": 7.047798990365595e-05, + "loss": 1.0051, + "step": 13388 + }, + { + "epoch": 2.3837250712250713, + "grad_norm": 0.8903454542160034, + "learning_rate": 7.04646165982799e-05, + "loss": 0.7055, + "step": 13389 + }, + { + "epoch": 2.383903133903134, + "grad_norm": 0.9634504914283752, + "learning_rate": 7.045124387158832e-05, + "loss": 0.7681, + "step": 13390 + }, + { + "epoch": 2.3840811965811968, + "grad_norm": 0.8645864129066467, + "learning_rate": 7.043787172384329e-05, + "loss": 0.9271, + "step": 13391 + }, + { + "epoch": 2.384259259259259, + "grad_norm": 0.8738446235656738, + "learning_rate": 7.04245001553067e-05, + "loss": 0.9, + "step": 13392 + }, + { + "epoch": 2.384437321937322, + "grad_norm": 0.7869822382926941, + "learning_rate": 7.041112916624062e-05, + "loss": 0.8639, + "step": 13393 + }, + { + "epoch": 2.3846153846153846, + "grad_norm": 0.8728111386299133, + "learning_rate": 7.039775875690698e-05, + "loss": 1.0367, + "step": 13394 + }, + { + "epoch": 2.3847934472934473, + "grad_norm": 0.7883852124214172, + "learning_rate": 7.03843889275678e-05, + "loss": 0.8338, + "step": 13395 + }, + { + "epoch": 2.38497150997151, + "grad_norm": 0.9267113208770752, + "learning_rate": 7.037101967848496e-05, + "loss": 0.8931, + "step": 13396 + }, + { + "epoch": 2.385149572649573, + "grad_norm": 0.8940320611000061, + "learning_rate": 7.035765100992048e-05, + "loss": 0.8071, + "step": 13397 + }, + { + "epoch": 2.385327635327635, + "grad_norm": 0.8109263777732849, + "learning_rate": 7.03442829221362e-05, + "loss": 0.8083, + "step": 13398 + }, + { + "epoch": 2.385505698005698, + "grad_norm": 0.8223438262939453, + "learning_rate": 7.033091541539413e-05, + "loss": 0.9296, + "step": 13399 + }, + { + "epoch": 2.3856837606837606, + "grad_norm": 0.817894697189331, + "learning_rate": 7.031754848995612e-05, + "loss": 0.9168, + "step": 13400 + }, + { + "epoch": 2.3858618233618234, + "grad_norm": 0.831462562084198, + "learning_rate": 7.030418214608411e-05, + "loss": 0.8613, + "step": 13401 + }, + { + "epoch": 2.386039886039886, + "grad_norm": 0.8388770818710327, + "learning_rate": 7.029081638403994e-05, + "loss": 0.7477, + "step": 13402 + }, + { + "epoch": 2.386217948717949, + "grad_norm": 0.9557843804359436, + "learning_rate": 7.02774512040855e-05, + "loss": 0.8932, + "step": 13403 + }, + { + "epoch": 2.386396011396011, + "grad_norm": 0.8249707221984863, + "learning_rate": 7.026408660648268e-05, + "loss": 1.0301, + "step": 13404 + }, + { + "epoch": 2.386574074074074, + "grad_norm": 0.8355069160461426, + "learning_rate": 7.025072259149333e-05, + "loss": 0.8081, + "step": 13405 + }, + { + "epoch": 2.3867521367521367, + "grad_norm": 0.8373300433158875, + "learning_rate": 7.023735915937924e-05, + "loss": 0.9911, + "step": 13406 + }, + { + "epoch": 2.3869301994301995, + "grad_norm": 0.7177539467811584, + "learning_rate": 7.022399631040228e-05, + "loss": 0.6397, + "step": 13407 + }, + { + "epoch": 2.387108262108262, + "grad_norm": 0.7371904253959656, + "learning_rate": 7.021063404482426e-05, + "loss": 0.8634, + "step": 13408 + }, + { + "epoch": 2.387286324786325, + "grad_norm": 0.8919385671615601, + "learning_rate": 7.019727236290696e-05, + "loss": 0.9514, + "step": 13409 + }, + { + "epoch": 2.3874643874643873, + "grad_norm": 0.7673050761222839, + "learning_rate": 7.018391126491225e-05, + "loss": 0.8957, + "step": 13410 + }, + { + "epoch": 2.38764245014245, + "grad_norm": 0.8401889801025391, + "learning_rate": 7.01705507511018e-05, + "loss": 0.908, + "step": 13411 + }, + { + "epoch": 2.3878205128205128, + "grad_norm": 0.822903037071228, + "learning_rate": 7.01571908217375e-05, + "loss": 0.911, + "step": 13412 + }, + { + "epoch": 2.3879985754985755, + "grad_norm": 0.9824740290641785, + "learning_rate": 7.014383147708102e-05, + "loss": 0.8314, + "step": 13413 + }, + { + "epoch": 2.3881766381766383, + "grad_norm": 0.9485064148902893, + "learning_rate": 7.013047271739414e-05, + "loss": 0.9819, + "step": 13414 + }, + { + "epoch": 2.388354700854701, + "grad_norm": 0.7565387487411499, + "learning_rate": 7.01171145429386e-05, + "loss": 0.9702, + "step": 13415 + }, + { + "epoch": 2.388532763532764, + "grad_norm": 0.8159620761871338, + "learning_rate": 7.010375695397615e-05, + "loss": 0.7302, + "step": 13416 + }, + { + "epoch": 2.388710826210826, + "grad_norm": 0.7818536162376404, + "learning_rate": 7.009039995076844e-05, + "loss": 0.6821, + "step": 13417 + }, + { + "epoch": 2.388888888888889, + "grad_norm": 0.7958348989486694, + "learning_rate": 7.007704353357724e-05, + "loss": 0.7996, + "step": 13418 + }, + { + "epoch": 2.3890669515669516, + "grad_norm": 0.8097305297851562, + "learning_rate": 7.006368770266421e-05, + "loss": 0.8177, + "step": 13419 + }, + { + "epoch": 2.3892450142450143, + "grad_norm": 0.9326507449150085, + "learning_rate": 7.005033245829105e-05, + "loss": 0.9307, + "step": 13420 + }, + { + "epoch": 2.389423076923077, + "grad_norm": 0.8954049944877625, + "learning_rate": 7.003697780071936e-05, + "loss": 0.8527, + "step": 13421 + }, + { + "epoch": 2.3896011396011394, + "grad_norm": 0.890548586845398, + "learning_rate": 7.00236237302109e-05, + "loss": 0.8203, + "step": 13422 + }, + { + "epoch": 2.389779202279202, + "grad_norm": 0.7508596181869507, + "learning_rate": 7.001027024702722e-05, + "loss": 0.7056, + "step": 13423 + }, + { + "epoch": 2.389957264957265, + "grad_norm": 0.9403550624847412, + "learning_rate": 6.999691735143002e-05, + "loss": 0.7336, + "step": 13424 + }, + { + "epoch": 2.3901353276353277, + "grad_norm": 0.8187662959098816, + "learning_rate": 6.998356504368087e-05, + "loss": 0.6897, + "step": 13425 + }, + { + "epoch": 2.3903133903133904, + "grad_norm": 0.8584417104721069, + "learning_rate": 6.997021332404145e-05, + "loss": 0.9143, + "step": 13426 + }, + { + "epoch": 2.390491452991453, + "grad_norm": 0.8739892840385437, + "learning_rate": 6.995686219277329e-05, + "loss": 0.8028, + "step": 13427 + }, + { + "epoch": 2.390669515669516, + "grad_norm": 0.9291013479232788, + "learning_rate": 6.994351165013799e-05, + "loss": 1.0305, + "step": 13428 + }, + { + "epoch": 2.390847578347578, + "grad_norm": 0.7937391400337219, + "learning_rate": 6.993016169639719e-05, + "loss": 0.8326, + "step": 13429 + }, + { + "epoch": 2.391025641025641, + "grad_norm": 0.655261754989624, + "learning_rate": 6.991681233181236e-05, + "loss": 0.7939, + "step": 13430 + }, + { + "epoch": 2.3912037037037037, + "grad_norm": 0.9606142640113831, + "learning_rate": 6.990346355664515e-05, + "loss": 1.1344, + "step": 13431 + }, + { + "epoch": 2.3913817663817665, + "grad_norm": 0.8111617565155029, + "learning_rate": 6.9890115371157e-05, + "loss": 0.8398, + "step": 13432 + }, + { + "epoch": 2.3915598290598292, + "grad_norm": 0.8111898899078369, + "learning_rate": 6.987676777560955e-05, + "loss": 0.9189, + "step": 13433 + }, + { + "epoch": 2.3917378917378915, + "grad_norm": 0.7850473523139954, + "learning_rate": 6.98634207702642e-05, + "loss": 0.9563, + "step": 13434 + }, + { + "epoch": 2.3919159544159543, + "grad_norm": 0.7740257978439331, + "learning_rate": 6.985007435538256e-05, + "loss": 0.7446, + "step": 13435 + }, + { + "epoch": 2.392094017094017, + "grad_norm": 0.9354606866836548, + "learning_rate": 6.983672853122604e-05, + "loss": 0.879, + "step": 13436 + }, + { + "epoch": 2.39227207977208, + "grad_norm": 0.8909385800361633, + "learning_rate": 6.982338329805622e-05, + "loss": 0.9381, + "step": 13437 + }, + { + "epoch": 2.3924501424501425, + "grad_norm": 0.7748416066169739, + "learning_rate": 6.981003865613448e-05, + "loss": 0.7169, + "step": 13438 + }, + { + "epoch": 2.3926282051282053, + "grad_norm": 0.7357833981513977, + "learning_rate": 6.979669460572234e-05, + "loss": 0.669, + "step": 13439 + }, + { + "epoch": 2.392806267806268, + "grad_norm": 0.8370460271835327, + "learning_rate": 6.978335114708119e-05, + "loss": 0.6215, + "step": 13440 + }, + { + "epoch": 2.3929843304843303, + "grad_norm": 0.7578476071357727, + "learning_rate": 6.977000828047256e-05, + "loss": 0.871, + "step": 13441 + }, + { + "epoch": 2.393162393162393, + "grad_norm": 0.8111903071403503, + "learning_rate": 6.975666600615776e-05, + "loss": 0.7888, + "step": 13442 + }, + { + "epoch": 2.393340455840456, + "grad_norm": 0.9584433436393738, + "learning_rate": 6.974332432439831e-05, + "loss": 1.0011, + "step": 13443 + }, + { + "epoch": 2.3935185185185186, + "grad_norm": 0.9105294942855835, + "learning_rate": 6.972998323545555e-05, + "loss": 1.0832, + "step": 13444 + }, + { + "epoch": 2.3936965811965814, + "grad_norm": 0.7990328669548035, + "learning_rate": 6.971664273959089e-05, + "loss": 0.9561, + "step": 13445 + }, + { + "epoch": 2.3938746438746437, + "grad_norm": 0.8575631976127625, + "learning_rate": 6.970330283706569e-05, + "loss": 0.7965, + "step": 13446 + }, + { + "epoch": 2.3940527065527064, + "grad_norm": 0.8147784471511841, + "learning_rate": 6.968996352814139e-05, + "loss": 0.806, + "step": 13447 + }, + { + "epoch": 2.394230769230769, + "grad_norm": 0.8284323215484619, + "learning_rate": 6.967662481307923e-05, + "loss": 0.942, + "step": 13448 + }, + { + "epoch": 2.394408831908832, + "grad_norm": 0.8238104581832886, + "learning_rate": 6.966328669214062e-05, + "loss": 0.9163, + "step": 13449 + }, + { + "epoch": 2.3945868945868947, + "grad_norm": 0.8855763673782349, + "learning_rate": 6.964994916558692e-05, + "loss": 0.7683, + "step": 13450 + }, + { + "epoch": 2.3947649572649574, + "grad_norm": 1.02780020236969, + "learning_rate": 6.963661223367937e-05, + "loss": 0.904, + "step": 13451 + }, + { + "epoch": 2.39494301994302, + "grad_norm": 0.8001773953437805, + "learning_rate": 6.96232758966794e-05, + "loss": 0.8459, + "step": 13452 + }, + { + "epoch": 2.3951210826210825, + "grad_norm": 0.755388617515564, + "learning_rate": 6.960994015484818e-05, + "loss": 0.7759, + "step": 13453 + }, + { + "epoch": 2.3952991452991452, + "grad_norm": 0.7774340510368347, + "learning_rate": 6.959660500844708e-05, + "loss": 0.7353, + "step": 13454 + }, + { + "epoch": 2.395477207977208, + "grad_norm": 0.8696026802062988, + "learning_rate": 6.958327045773733e-05, + "loss": 0.635, + "step": 13455 + }, + { + "epoch": 2.3956552706552707, + "grad_norm": 0.8419780731201172, + "learning_rate": 6.956993650298025e-05, + "loss": 0.8515, + "step": 13456 + }, + { + "epoch": 2.3958333333333335, + "grad_norm": 0.9125590324401855, + "learning_rate": 6.955660314443699e-05, + "loss": 0.9099, + "step": 13457 + }, + { + "epoch": 2.396011396011396, + "grad_norm": 0.6847489476203918, + "learning_rate": 6.954327038236891e-05, + "loss": 0.6652, + "step": 13458 + }, + { + "epoch": 2.3961894586894585, + "grad_norm": 0.8674905896186829, + "learning_rate": 6.952993821703713e-05, + "loss": 0.7049, + "step": 13459 + }, + { + "epoch": 2.3963675213675213, + "grad_norm": 0.7777035236358643, + "learning_rate": 6.951660664870296e-05, + "loss": 0.818, + "step": 13460 + }, + { + "epoch": 2.396545584045584, + "grad_norm": 0.8349783420562744, + "learning_rate": 6.950327567762751e-05, + "loss": 0.9203, + "step": 13461 + }, + { + "epoch": 2.396723646723647, + "grad_norm": 0.7589834332466125, + "learning_rate": 6.948994530407206e-05, + "loss": 1.015, + "step": 13462 + }, + { + "epoch": 2.3969017094017095, + "grad_norm": 0.9340610504150391, + "learning_rate": 6.947661552829773e-05, + "loss": 1.0575, + "step": 13463 + }, + { + "epoch": 2.3970797720797723, + "grad_norm": 0.9100959300994873, + "learning_rate": 6.946328635056573e-05, + "loss": 0.8824, + "step": 13464 + }, + { + "epoch": 2.3972578347578346, + "grad_norm": 0.8255945444107056, + "learning_rate": 6.944995777113717e-05, + "loss": 0.7701, + "step": 13465 + }, + { + "epoch": 2.3974358974358974, + "grad_norm": 0.8572675585746765, + "learning_rate": 6.943662979027328e-05, + "loss": 0.9425, + "step": 13466 + }, + { + "epoch": 2.39761396011396, + "grad_norm": 0.8219536542892456, + "learning_rate": 6.94233024082351e-05, + "loss": 0.8184, + "step": 13467 + }, + { + "epoch": 2.397792022792023, + "grad_norm": 0.8260995149612427, + "learning_rate": 6.940997562528377e-05, + "loss": 0.8324, + "step": 13468 + }, + { + "epoch": 2.3979700854700856, + "grad_norm": 0.9707075357437134, + "learning_rate": 6.939664944168047e-05, + "loss": 0.9865, + "step": 13469 + }, + { + "epoch": 2.398148148148148, + "grad_norm": 0.9030438661575317, + "learning_rate": 6.938332385768622e-05, + "loss": 1.0244, + "step": 13470 + }, + { + "epoch": 2.3983262108262107, + "grad_norm": 0.8425108194351196, + "learning_rate": 6.936999887356214e-05, + "loss": 0.7053, + "step": 13471 + }, + { + "epoch": 2.3985042735042734, + "grad_norm": 1.0073270797729492, + "learning_rate": 6.93566744895693e-05, + "loss": 0.9324, + "step": 13472 + }, + { + "epoch": 2.398682336182336, + "grad_norm": 0.7647563219070435, + "learning_rate": 6.93433507059688e-05, + "loss": 0.7233, + "step": 13473 + }, + { + "epoch": 2.398860398860399, + "grad_norm": 0.7632454633712769, + "learning_rate": 6.933002752302162e-05, + "loss": 0.8678, + "step": 13474 + }, + { + "epoch": 2.3990384615384617, + "grad_norm": 0.7943702936172485, + "learning_rate": 6.931670494098887e-05, + "loss": 0.8805, + "step": 13475 + }, + { + "epoch": 2.3992165242165244, + "grad_norm": 0.9440419673919678, + "learning_rate": 6.930338296013153e-05, + "loss": 1.0103, + "step": 13476 + }, + { + "epoch": 2.3993945868945867, + "grad_norm": 0.9119253754615784, + "learning_rate": 6.929006158071065e-05, + "loss": 1.0235, + "step": 13477 + }, + { + "epoch": 2.3995726495726495, + "grad_norm": 0.7750248908996582, + "learning_rate": 6.927674080298721e-05, + "loss": 0.957, + "step": 13478 + }, + { + "epoch": 2.3997507122507122, + "grad_norm": 0.8847192525863647, + "learning_rate": 6.926342062722223e-05, + "loss": 0.9066, + "step": 13479 + }, + { + "epoch": 2.399928774928775, + "grad_norm": 0.814396321773529, + "learning_rate": 6.925010105367665e-05, + "loss": 1.0001, + "step": 13480 + }, + { + "epoch": 2.4001068376068377, + "grad_norm": 0.8323664665222168, + "learning_rate": 6.923678208261147e-05, + "loss": 1.0027, + "step": 13481 + }, + { + "epoch": 2.4002849002849, + "grad_norm": 0.8351104259490967, + "learning_rate": 6.92234637142876e-05, + "loss": 0.8629, + "step": 13482 + }, + { + "epoch": 2.400462962962963, + "grad_norm": 0.9298360347747803, + "learning_rate": 6.92101459489661e-05, + "loss": 0.9161, + "step": 13483 + }, + { + "epoch": 2.4006410256410255, + "grad_norm": 0.9423344135284424, + "learning_rate": 6.919682878690777e-05, + "loss": 1.4416, + "step": 13484 + }, + { + "epoch": 2.4008190883190883, + "grad_norm": 0.8340599536895752, + "learning_rate": 6.918351222837363e-05, + "loss": 0.8696, + "step": 13485 + }, + { + "epoch": 2.400997150997151, + "grad_norm": 0.8533751368522644, + "learning_rate": 6.917019627362451e-05, + "loss": 1.1383, + "step": 13486 + }, + { + "epoch": 2.401175213675214, + "grad_norm": 0.8060563206672668, + "learning_rate": 6.91568809229214e-05, + "loss": 0.8544, + "step": 13487 + }, + { + "epoch": 2.4013532763532766, + "grad_norm": 0.865485668182373, + "learning_rate": 6.914356617652511e-05, + "loss": 0.9286, + "step": 13488 + }, + { + "epoch": 2.401531339031339, + "grad_norm": 0.8785045742988586, + "learning_rate": 6.913025203469652e-05, + "loss": 0.7339, + "step": 13489 + }, + { + "epoch": 2.4017094017094016, + "grad_norm": 0.7718466520309448, + "learning_rate": 6.911693849769654e-05, + "loss": 0.8821, + "step": 13490 + }, + { + "epoch": 2.4018874643874644, + "grad_norm": 0.7274343371391296, + "learning_rate": 6.910362556578599e-05, + "loss": 0.6179, + "step": 13491 + }, + { + "epoch": 2.402065527065527, + "grad_norm": 0.8848530054092407, + "learning_rate": 6.909031323922574e-05, + "loss": 0.7848, + "step": 13492 + }, + { + "epoch": 2.40224358974359, + "grad_norm": 0.7384527325630188, + "learning_rate": 6.907700151827657e-05, + "loss": 0.5, + "step": 13493 + }, + { + "epoch": 2.402421652421652, + "grad_norm": 0.865505576133728, + "learning_rate": 6.906369040319936e-05, + "loss": 0.7127, + "step": 13494 + }, + { + "epoch": 2.402599715099715, + "grad_norm": 0.8588849902153015, + "learning_rate": 6.90503798942548e-05, + "loss": 0.8833, + "step": 13495 + }, + { + "epoch": 2.4027777777777777, + "grad_norm": 0.8570847511291504, + "learning_rate": 6.903706999170381e-05, + "loss": 0.9765, + "step": 13496 + }, + { + "epoch": 2.4029558404558404, + "grad_norm": 0.9193849563598633, + "learning_rate": 6.902376069580706e-05, + "loss": 0.8654, + "step": 13497 + }, + { + "epoch": 2.403133903133903, + "grad_norm": 0.8181582093238831, + "learning_rate": 6.901045200682545e-05, + "loss": 0.8815, + "step": 13498 + }, + { + "epoch": 2.403311965811966, + "grad_norm": 0.783163845539093, + "learning_rate": 6.89971439250196e-05, + "loss": 0.8383, + "step": 13499 + }, + { + "epoch": 2.4034900284900287, + "grad_norm": 1.0679216384887695, + "learning_rate": 6.898383645065032e-05, + "loss": 1.0525, + "step": 13500 + }, + { + "epoch": 2.403668091168091, + "grad_norm": 0.7945899367332458, + "learning_rate": 6.897052958397831e-05, + "loss": 1.0091, + "step": 13501 + }, + { + "epoch": 2.4038461538461537, + "grad_norm": 0.8310369253158569, + "learning_rate": 6.895722332526438e-05, + "loss": 0.8909, + "step": 13502 + }, + { + "epoch": 2.4040242165242165, + "grad_norm": 0.8811371922492981, + "learning_rate": 6.894391767476911e-05, + "loss": 0.8354, + "step": 13503 + }, + { + "epoch": 2.4042022792022792, + "grad_norm": 1.011495590209961, + "learning_rate": 6.893061263275332e-05, + "loss": 0.8846, + "step": 13504 + }, + { + "epoch": 2.404380341880342, + "grad_norm": 0.7587227821350098, + "learning_rate": 6.891730819947758e-05, + "loss": 0.8886, + "step": 13505 + }, + { + "epoch": 2.4045584045584047, + "grad_norm": 0.8367353677749634, + "learning_rate": 6.890400437520265e-05, + "loss": 1.008, + "step": 13506 + }, + { + "epoch": 2.404736467236467, + "grad_norm": 0.7200010418891907, + "learning_rate": 6.889070116018911e-05, + "loss": 0.8405, + "step": 13507 + }, + { + "epoch": 2.40491452991453, + "grad_norm": 0.9391907453536987, + "learning_rate": 6.887739855469769e-05, + "loss": 0.8904, + "step": 13508 + }, + { + "epoch": 2.4050925925925926, + "grad_norm": 0.8687568306922913, + "learning_rate": 6.886409655898902e-05, + "loss": 0.7145, + "step": 13509 + }, + { + "epoch": 2.4052706552706553, + "grad_norm": 0.7382767796516418, + "learning_rate": 6.885079517332366e-05, + "loss": 0.7639, + "step": 13510 + }, + { + "epoch": 2.405448717948718, + "grad_norm": 0.8322962522506714, + "learning_rate": 6.883749439796227e-05, + "loss": 1.0002, + "step": 13511 + }, + { + "epoch": 2.405626780626781, + "grad_norm": 0.815183162689209, + "learning_rate": 6.882419423316544e-05, + "loss": 0.8628, + "step": 13512 + }, + { + "epoch": 2.405804843304843, + "grad_norm": 0.9304860234260559, + "learning_rate": 6.881089467919381e-05, + "loss": 0.9489, + "step": 13513 + }, + { + "epoch": 2.405982905982906, + "grad_norm": 0.9071274995803833, + "learning_rate": 6.879759573630784e-05, + "loss": 0.8117, + "step": 13514 + }, + { + "epoch": 2.4061609686609686, + "grad_norm": 0.9378795027732849, + "learning_rate": 6.878429740476822e-05, + "loss": 1.22, + "step": 13515 + }, + { + "epoch": 2.4063390313390314, + "grad_norm": 0.7354511618614197, + "learning_rate": 6.877099968483541e-05, + "loss": 0.6696, + "step": 13516 + }, + { + "epoch": 2.406517094017094, + "grad_norm": 0.8701893091201782, + "learning_rate": 6.875770257677002e-05, + "loss": 0.8691, + "step": 13517 + }, + { + "epoch": 2.406695156695157, + "grad_norm": 0.8819001913070679, + "learning_rate": 6.87444060808325e-05, + "loss": 0.7428, + "step": 13518 + }, + { + "epoch": 2.406873219373219, + "grad_norm": 0.7339609265327454, + "learning_rate": 6.873111019728347e-05, + "loss": 0.7959, + "step": 13519 + }, + { + "epoch": 2.407051282051282, + "grad_norm": 0.8365123867988586, + "learning_rate": 6.871781492638335e-05, + "loss": 0.8199, + "step": 13520 + }, + { + "epoch": 2.4072293447293447, + "grad_norm": 0.9667043685913086, + "learning_rate": 6.870452026839266e-05, + "loss": 0.8261, + "step": 13521 + }, + { + "epoch": 2.4074074074074074, + "grad_norm": 0.6979679465293884, + "learning_rate": 6.869122622357187e-05, + "loss": 0.5909, + "step": 13522 + }, + { + "epoch": 2.40758547008547, + "grad_norm": 0.7326778769493103, + "learning_rate": 6.867793279218152e-05, + "loss": 0.9297, + "step": 13523 + }, + { + "epoch": 2.407763532763533, + "grad_norm": 0.8808563351631165, + "learning_rate": 6.866463997448196e-05, + "loss": 0.7481, + "step": 13524 + }, + { + "epoch": 2.4079415954415953, + "grad_norm": 0.7830268740653992, + "learning_rate": 6.86513477707337e-05, + "loss": 0.7902, + "step": 13525 + }, + { + "epoch": 2.408119658119658, + "grad_norm": 0.9482602477073669, + "learning_rate": 6.863805618119713e-05, + "loss": 1.1541, + "step": 13526 + }, + { + "epoch": 2.4082977207977208, + "grad_norm": 0.8369114995002747, + "learning_rate": 6.862476520613276e-05, + "loss": 0.874, + "step": 13527 + }, + { + "epoch": 2.4084757834757835, + "grad_norm": 0.9107078909873962, + "learning_rate": 6.86114748458009e-05, + "loss": 0.9412, + "step": 13528 + }, + { + "epoch": 2.4086538461538463, + "grad_norm": 0.8086137771606445, + "learning_rate": 6.859818510046199e-05, + "loss": 0.8495, + "step": 13529 + }, + { + "epoch": 2.408831908831909, + "grad_norm": 0.8824704885482788, + "learning_rate": 6.858489597037646e-05, + "loss": 0.8967, + "step": 13530 + }, + { + "epoch": 2.4090099715099713, + "grad_norm": 0.8514662384986877, + "learning_rate": 6.857160745580455e-05, + "loss": 0.9171, + "step": 13531 + }, + { + "epoch": 2.409188034188034, + "grad_norm": 0.7788167595863342, + "learning_rate": 6.855831955700675e-05, + "loss": 0.904, + "step": 13532 + }, + { + "epoch": 2.409366096866097, + "grad_norm": 0.913113534450531, + "learning_rate": 6.854503227424337e-05, + "loss": 0.8696, + "step": 13533 + }, + { + "epoch": 2.4095441595441596, + "grad_norm": 0.8424487113952637, + "learning_rate": 6.853174560777475e-05, + "loss": 0.8388, + "step": 13534 + }, + { + "epoch": 2.4097222222222223, + "grad_norm": 0.8609711527824402, + "learning_rate": 6.851845955786116e-05, + "loss": 0.7142, + "step": 13535 + }, + { + "epoch": 2.409900284900285, + "grad_norm": 0.8141375184059143, + "learning_rate": 6.850517412476301e-05, + "loss": 0.7198, + "step": 13536 + }, + { + "epoch": 2.410078347578348, + "grad_norm": 0.8615440130233765, + "learning_rate": 6.84918893087405e-05, + "loss": 0.958, + "step": 13537 + }, + { + "epoch": 2.41025641025641, + "grad_norm": 0.7733060717582703, + "learning_rate": 6.847860511005401e-05, + "loss": 0.7639, + "step": 13538 + }, + { + "epoch": 2.410434472934473, + "grad_norm": 0.9519185423851013, + "learning_rate": 6.846532152896375e-05, + "loss": 0.8239, + "step": 13539 + }, + { + "epoch": 2.4106125356125356, + "grad_norm": 0.774053692817688, + "learning_rate": 6.845203856573002e-05, + "loss": 0.891, + "step": 13540 + }, + { + "epoch": 2.4107905982905984, + "grad_norm": 0.8791571259498596, + "learning_rate": 6.843875622061304e-05, + "loss": 1.0107, + "step": 13541 + }, + { + "epoch": 2.410968660968661, + "grad_norm": 0.9431949853897095, + "learning_rate": 6.842547449387309e-05, + "loss": 0.8575, + "step": 13542 + }, + { + "epoch": 2.4111467236467234, + "grad_norm": 1.0521612167358398, + "learning_rate": 6.841219338577034e-05, + "loss": 0.9446, + "step": 13543 + }, + { + "epoch": 2.411324786324786, + "grad_norm": 0.7592857480049133, + "learning_rate": 6.83989128965651e-05, + "loss": 1.0595, + "step": 13544 + }, + { + "epoch": 2.411502849002849, + "grad_norm": 0.9002043604850769, + "learning_rate": 6.838563302651747e-05, + "loss": 0.9067, + "step": 13545 + }, + { + "epoch": 2.4116809116809117, + "grad_norm": 0.7144047021865845, + "learning_rate": 6.83723537758877e-05, + "loss": 0.6699, + "step": 13546 + }, + { + "epoch": 2.4118589743589745, + "grad_norm": 0.8226693868637085, + "learning_rate": 6.835907514493594e-05, + "loss": 0.987, + "step": 13547 + }, + { + "epoch": 2.412037037037037, + "grad_norm": 0.8507830500602722, + "learning_rate": 6.834579713392237e-05, + "loss": 0.7803, + "step": 13548 + }, + { + "epoch": 2.4122150997151, + "grad_norm": 0.727870762348175, + "learning_rate": 6.83325197431072e-05, + "loss": 0.8071, + "step": 13549 + }, + { + "epoch": 2.4123931623931623, + "grad_norm": 0.7601624727249146, + "learning_rate": 6.831924297275049e-05, + "loss": 0.7627, + "step": 13550 + }, + { + "epoch": 2.412571225071225, + "grad_norm": 0.8519877791404724, + "learning_rate": 6.830596682311243e-05, + "loss": 0.9271, + "step": 13551 + }, + { + "epoch": 2.4127492877492878, + "grad_norm": 1.0122307538986206, + "learning_rate": 6.829269129445307e-05, + "loss": 0.6424, + "step": 13552 + }, + { + "epoch": 2.4129273504273505, + "grad_norm": 0.8992687463760376, + "learning_rate": 6.827941638703258e-05, + "loss": 0.8034, + "step": 13553 + }, + { + "epoch": 2.4131054131054133, + "grad_norm": 0.7740746140480042, + "learning_rate": 6.826614210111102e-05, + "loss": 0.955, + "step": 13554 + }, + { + "epoch": 2.4132834757834756, + "grad_norm": 0.8176493048667908, + "learning_rate": 6.825286843694852e-05, + "loss": 0.7844, + "step": 13555 + }, + { + "epoch": 2.4134615384615383, + "grad_norm": 0.8112488985061646, + "learning_rate": 6.823959539480507e-05, + "loss": 0.8495, + "step": 13556 + }, + { + "epoch": 2.413639601139601, + "grad_norm": 0.8186960220336914, + "learning_rate": 6.822632297494078e-05, + "loss": 0.8922, + "step": 13557 + }, + { + "epoch": 2.413817663817664, + "grad_norm": 0.9498438835144043, + "learning_rate": 6.821305117761569e-05, + "loss": 0.8862, + "step": 13558 + }, + { + "epoch": 2.4139957264957266, + "grad_norm": 0.8591099381446838, + "learning_rate": 6.819978000308987e-05, + "loss": 0.837, + "step": 13559 + }, + { + "epoch": 2.4141737891737893, + "grad_norm": 0.8130860328674316, + "learning_rate": 6.818650945162324e-05, + "loss": 1.0723, + "step": 13560 + }, + { + "epoch": 2.414351851851852, + "grad_norm": 0.8800109624862671, + "learning_rate": 6.81732395234759e-05, + "loss": 0.8067, + "step": 13561 + }, + { + "epoch": 2.4145299145299144, + "grad_norm": 0.7786064147949219, + "learning_rate": 6.81599702189078e-05, + "loss": 0.9, + "step": 13562 + }, + { + "epoch": 2.414707977207977, + "grad_norm": 0.8343027234077454, + "learning_rate": 6.814670153817898e-05, + "loss": 0.7487, + "step": 13563 + }, + { + "epoch": 2.41488603988604, + "grad_norm": 0.7904187440872192, + "learning_rate": 6.813343348154934e-05, + "loss": 0.7904, + "step": 13564 + }, + { + "epoch": 2.4150641025641026, + "grad_norm": 0.7609010934829712, + "learning_rate": 6.81201660492789e-05, + "loss": 0.6734, + "step": 13565 + }, + { + "epoch": 2.4152421652421654, + "grad_norm": 0.8402243256568909, + "learning_rate": 6.810689924162756e-05, + "loss": 0.9581, + "step": 13566 + }, + { + "epoch": 2.4154202279202277, + "grad_norm": 0.8557454943656921, + "learning_rate": 6.809363305885527e-05, + "loss": 0.8387, + "step": 13567 + }, + { + "epoch": 2.4155982905982905, + "grad_norm": 0.8983132243156433, + "learning_rate": 6.808036750122197e-05, + "loss": 0.832, + "step": 13568 + }, + { + "epoch": 2.415776353276353, + "grad_norm": 0.8552190065383911, + "learning_rate": 6.806710256898755e-05, + "loss": 0.9257, + "step": 13569 + }, + { + "epoch": 2.415954415954416, + "grad_norm": 1.0639078617095947, + "learning_rate": 6.805383826241197e-05, + "loss": 0.9743, + "step": 13570 + }, + { + "epoch": 2.4161324786324787, + "grad_norm": 0.7951667904853821, + "learning_rate": 6.804057458175501e-05, + "loss": 0.9326, + "step": 13571 + }, + { + "epoch": 2.4163105413105415, + "grad_norm": 0.8652639985084534, + "learning_rate": 6.802731152727664e-05, + "loss": 1.0114, + "step": 13572 + }, + { + "epoch": 2.416488603988604, + "grad_norm": 0.8777487874031067, + "learning_rate": 6.801404909923664e-05, + "loss": 0.759, + "step": 13573 + }, + { + "epoch": 2.4166666666666665, + "grad_norm": 0.7922869920730591, + "learning_rate": 6.800078729789497e-05, + "loss": 0.8392, + "step": 13574 + }, + { + "epoch": 2.4168447293447293, + "grad_norm": 0.9189477562904358, + "learning_rate": 6.798752612351133e-05, + "loss": 0.9485, + "step": 13575 + }, + { + "epoch": 2.417022792022792, + "grad_norm": 0.8752175569534302, + "learning_rate": 6.797426557634567e-05, + "loss": 0.725, + "step": 13576 + }, + { + "epoch": 2.4172008547008548, + "grad_norm": 0.8646897077560425, + "learning_rate": 6.79610056566577e-05, + "loss": 0.9791, + "step": 13577 + }, + { + "epoch": 2.4173789173789175, + "grad_norm": 0.8749415278434753, + "learning_rate": 6.794774636470731e-05, + "loss": 0.9059, + "step": 13578 + }, + { + "epoch": 2.41755698005698, + "grad_norm": 0.9642252922058105, + "learning_rate": 6.793448770075422e-05, + "loss": 0.9972, + "step": 13579 + }, + { + "epoch": 2.4177350427350426, + "grad_norm": 0.8430541157722473, + "learning_rate": 6.792122966505827e-05, + "loss": 0.7126, + "step": 13580 + }, + { + "epoch": 2.4179131054131053, + "grad_norm": 0.8478374481201172, + "learning_rate": 6.790797225787913e-05, + "loss": 0.8995, + "step": 13581 + }, + { + "epoch": 2.418091168091168, + "grad_norm": 0.7008727788925171, + "learning_rate": 6.789471547947665e-05, + "loss": 0.6694, + "step": 13582 + }, + { + "epoch": 2.418269230769231, + "grad_norm": 0.873543918132782, + "learning_rate": 6.78814593301105e-05, + "loss": 1.0418, + "step": 13583 + }, + { + "epoch": 2.4184472934472936, + "grad_norm": 0.7208766341209412, + "learning_rate": 6.786820381004047e-05, + "loss": 0.8095, + "step": 13584 + }, + { + "epoch": 2.4186253561253563, + "grad_norm": 0.7272628545761108, + "learning_rate": 6.78549489195262e-05, + "loss": 0.5801, + "step": 13585 + }, + { + "epoch": 2.4188034188034186, + "grad_norm": 0.7155343294143677, + "learning_rate": 6.784169465882747e-05, + "loss": 0.748, + "step": 13586 + }, + { + "epoch": 2.4189814814814814, + "grad_norm": 0.928404688835144, + "learning_rate": 6.78284410282039e-05, + "loss": 0.9175, + "step": 13587 + }, + { + "epoch": 2.419159544159544, + "grad_norm": 0.7239044308662415, + "learning_rate": 6.781518802791519e-05, + "loss": 0.5855, + "step": 13588 + }, + { + "epoch": 2.419337606837607, + "grad_norm": 0.8126311302185059, + "learning_rate": 6.780193565822104e-05, + "loss": 0.832, + "step": 13589 + }, + { + "epoch": 2.4195156695156697, + "grad_norm": 0.7470774054527283, + "learning_rate": 6.778868391938103e-05, + "loss": 0.6202, + "step": 13590 + }, + { + "epoch": 2.419693732193732, + "grad_norm": 0.9161462187767029, + "learning_rate": 6.77754328116549e-05, + "loss": 0.8674, + "step": 13591 + }, + { + "epoch": 2.4198717948717947, + "grad_norm": 0.7225745320320129, + "learning_rate": 6.77621823353022e-05, + "loss": 0.8745, + "step": 13592 + }, + { + "epoch": 2.4200498575498575, + "grad_norm": 0.8380082845687866, + "learning_rate": 6.774893249058257e-05, + "loss": 0.5501, + "step": 13593 + }, + { + "epoch": 2.42022792022792, + "grad_norm": 0.8031942844390869, + "learning_rate": 6.77356832777556e-05, + "loss": 0.9925, + "step": 13594 + }, + { + "epoch": 2.420405982905983, + "grad_norm": 0.8278502821922302, + "learning_rate": 6.772243469708093e-05, + "loss": 0.6411, + "step": 13595 + }, + { + "epoch": 2.4205840455840457, + "grad_norm": 0.7655481100082397, + "learning_rate": 6.770918674881805e-05, + "loss": 0.7896, + "step": 13596 + }, + { + "epoch": 2.4207621082621085, + "grad_norm": 0.8260186314582825, + "learning_rate": 6.769593943322661e-05, + "loss": 0.8531, + "step": 13597 + }, + { + "epoch": 2.4209401709401708, + "grad_norm": 0.8293251395225525, + "learning_rate": 6.76826927505661e-05, + "loss": 0.8193, + "step": 13598 + }, + { + "epoch": 2.4211182336182335, + "grad_norm": 0.8868293762207031, + "learning_rate": 6.766944670109616e-05, + "loss": 0.8453, + "step": 13599 + }, + { + "epoch": 2.4212962962962963, + "grad_norm": 0.769124448299408, + "learning_rate": 6.765620128507619e-05, + "loss": 0.7412, + "step": 13600 + }, + { + "epoch": 2.421474358974359, + "grad_norm": 0.7727167010307312, + "learning_rate": 6.764295650276581e-05, + "loss": 0.8721, + "step": 13601 + }, + { + "epoch": 2.421652421652422, + "grad_norm": 0.9975818395614624, + "learning_rate": 6.762971235442444e-05, + "loss": 0.8128, + "step": 13602 + }, + { + "epoch": 2.421830484330484, + "grad_norm": 0.8000788688659668, + "learning_rate": 6.761646884031164e-05, + "loss": 0.8328, + "step": 13603 + }, + { + "epoch": 2.422008547008547, + "grad_norm": 0.7196731567382812, + "learning_rate": 6.760322596068684e-05, + "loss": 0.8912, + "step": 13604 + }, + { + "epoch": 2.4221866096866096, + "grad_norm": 0.8092321753501892, + "learning_rate": 6.758998371580955e-05, + "loss": 0.9066, + "step": 13605 + }, + { + "epoch": 2.4223646723646723, + "grad_norm": 0.7664031982421875, + "learning_rate": 6.757674210593918e-05, + "loss": 0.9854, + "step": 13606 + }, + { + "epoch": 2.422542735042735, + "grad_norm": 0.794507622718811, + "learning_rate": 6.75635011313352e-05, + "loss": 0.9296, + "step": 13607 + }, + { + "epoch": 2.422720797720798, + "grad_norm": 0.9127107858657837, + "learning_rate": 6.755026079225705e-05, + "loss": 0.9516, + "step": 13608 + }, + { + "epoch": 2.4228988603988606, + "grad_norm": 0.8025720715522766, + "learning_rate": 6.753702108896411e-05, + "loss": 1.0664, + "step": 13609 + }, + { + "epoch": 2.423076923076923, + "grad_norm": 0.7304871678352356, + "learning_rate": 6.752378202171585e-05, + "loss": 0.9071, + "step": 13610 + }, + { + "epoch": 2.4232549857549857, + "grad_norm": 0.9048241972923279, + "learning_rate": 6.751054359077157e-05, + "loss": 0.8649, + "step": 13611 + }, + { + "epoch": 2.4234330484330484, + "grad_norm": 0.8589995503425598, + "learning_rate": 6.749730579639074e-05, + "loss": 0.8895, + "step": 13612 + }, + { + "epoch": 2.423611111111111, + "grad_norm": 0.8098960518836975, + "learning_rate": 6.748406863883265e-05, + "loss": 0.6651, + "step": 13613 + }, + { + "epoch": 2.423789173789174, + "grad_norm": 0.9875120520591736, + "learning_rate": 6.74708321183567e-05, + "loss": 0.9605, + "step": 13614 + }, + { + "epoch": 2.423967236467236, + "grad_norm": 0.8211431503295898, + "learning_rate": 6.74575962352222e-05, + "loss": 0.8721, + "step": 13615 + }, + { + "epoch": 2.424145299145299, + "grad_norm": 0.9732884168624878, + "learning_rate": 6.744436098968855e-05, + "loss": 0.7501, + "step": 13616 + }, + { + "epoch": 2.4243233618233617, + "grad_norm": 0.9813733696937561, + "learning_rate": 6.743112638201496e-05, + "loss": 0.9823, + "step": 13617 + }, + { + "epoch": 2.4245014245014245, + "grad_norm": 0.8075012564659119, + "learning_rate": 6.741789241246083e-05, + "loss": 0.7018, + "step": 13618 + }, + { + "epoch": 2.4246794871794872, + "grad_norm": 0.7845864295959473, + "learning_rate": 6.740465908128539e-05, + "loss": 0.7423, + "step": 13619 + }, + { + "epoch": 2.42485754985755, + "grad_norm": 0.7754862308502197, + "learning_rate": 6.739142638874799e-05, + "loss": 0.8735, + "step": 13620 + }, + { + "epoch": 2.4250356125356127, + "grad_norm": 0.7971537709236145, + "learning_rate": 6.737819433510781e-05, + "loss": 0.7663, + "step": 13621 + }, + { + "epoch": 2.425213675213675, + "grad_norm": 0.8043563365936279, + "learning_rate": 6.736496292062416e-05, + "loss": 0.9311, + "step": 13622 + }, + { + "epoch": 2.425391737891738, + "grad_norm": 0.8150136470794678, + "learning_rate": 6.735173214555628e-05, + "loss": 0.9164, + "step": 13623 + }, + { + "epoch": 2.4255698005698005, + "grad_norm": 0.9853758811950684, + "learning_rate": 6.733850201016338e-05, + "loss": 0.6253, + "step": 13624 + }, + { + "epoch": 2.4257478632478633, + "grad_norm": 1.2138506174087524, + "learning_rate": 6.732527251470465e-05, + "loss": 1.0536, + "step": 13625 + }, + { + "epoch": 2.425925925925926, + "grad_norm": 0.9306546449661255, + "learning_rate": 6.73120436594394e-05, + "loss": 0.8607, + "step": 13626 + }, + { + "epoch": 2.426103988603989, + "grad_norm": 0.8536837697029114, + "learning_rate": 6.729881544462668e-05, + "loss": 0.9418, + "step": 13627 + }, + { + "epoch": 2.426282051282051, + "grad_norm": 0.8561417460441589, + "learning_rate": 6.728558787052574e-05, + "loss": 0.9556, + "step": 13628 + }, + { + "epoch": 2.426460113960114, + "grad_norm": 0.7499847412109375, + "learning_rate": 6.727236093739579e-05, + "loss": 0.7795, + "step": 13629 + }, + { + "epoch": 2.4266381766381766, + "grad_norm": 0.8541018962860107, + "learning_rate": 6.725913464549591e-05, + "loss": 1.0322, + "step": 13630 + }, + { + "epoch": 2.4268162393162394, + "grad_norm": 0.9659489989280701, + "learning_rate": 6.724590899508532e-05, + "loss": 1.1907, + "step": 13631 + }, + { + "epoch": 2.426994301994302, + "grad_norm": 0.9548102617263794, + "learning_rate": 6.723268398642307e-05, + "loss": 1.0545, + "step": 13632 + }, + { + "epoch": 2.427172364672365, + "grad_norm": 0.8543868660926819, + "learning_rate": 6.72194596197683e-05, + "loss": 0.8041, + "step": 13633 + }, + { + "epoch": 2.427350427350427, + "grad_norm": 0.838178277015686, + "learning_rate": 6.720623589538013e-05, + "loss": 0.7772, + "step": 13634 + }, + { + "epoch": 2.42752849002849, + "grad_norm": 0.8207933306694031, + "learning_rate": 6.719301281351768e-05, + "loss": 0.9222, + "step": 13635 + }, + { + "epoch": 2.4277065527065527, + "grad_norm": 0.7705093026161194, + "learning_rate": 6.717979037443996e-05, + "loss": 0.8989, + "step": 13636 + }, + { + "epoch": 2.4278846153846154, + "grad_norm": 0.8627061247825623, + "learning_rate": 6.716656857840609e-05, + "loss": 0.8834, + "step": 13637 + }, + { + "epoch": 2.428062678062678, + "grad_norm": 0.8404269218444824, + "learning_rate": 6.715334742567507e-05, + "loss": 0.8087, + "step": 13638 + }, + { + "epoch": 2.428240740740741, + "grad_norm": 0.9098958373069763, + "learning_rate": 6.7140126916506e-05, + "loss": 0.803, + "step": 13639 + }, + { + "epoch": 2.4284188034188032, + "grad_norm": 0.7482922673225403, + "learning_rate": 6.712690705115785e-05, + "loss": 0.8254, + "step": 13640 + }, + { + "epoch": 2.428596866096866, + "grad_norm": 0.8636375665664673, + "learning_rate": 6.711368782988972e-05, + "loss": 0.8788, + "step": 13641 + }, + { + "epoch": 2.4287749287749287, + "grad_norm": 0.8261808753013611, + "learning_rate": 6.710046925296052e-05, + "loss": 0.9135, + "step": 13642 + }, + { + "epoch": 2.4289529914529915, + "grad_norm": 1.060263752937317, + "learning_rate": 6.70872513206293e-05, + "loss": 0.9589, + "step": 13643 + }, + { + "epoch": 2.4291310541310542, + "grad_norm": 0.8128657341003418, + "learning_rate": 6.7074034033155e-05, + "loss": 0.7735, + "step": 13644 + }, + { + "epoch": 2.429309116809117, + "grad_norm": 0.9948938488960266, + "learning_rate": 6.706081739079663e-05, + "loss": 0.7242, + "step": 13645 + }, + { + "epoch": 2.4294871794871793, + "grad_norm": 0.8850025534629822, + "learning_rate": 6.704760139381311e-05, + "loss": 0.9393, + "step": 13646 + }, + { + "epoch": 2.429665242165242, + "grad_norm": 0.833534300327301, + "learning_rate": 6.703438604246337e-05, + "loss": 0.7824, + "step": 13647 + }, + { + "epoch": 2.429843304843305, + "grad_norm": 0.7362738251686096, + "learning_rate": 6.70211713370064e-05, + "loss": 0.7594, + "step": 13648 + }, + { + "epoch": 2.4300213675213675, + "grad_norm": 0.97635817527771, + "learning_rate": 6.700795727770101e-05, + "loss": 1.0097, + "step": 13649 + }, + { + "epoch": 2.4301994301994303, + "grad_norm": 0.8434939980506897, + "learning_rate": 6.699474386480622e-05, + "loss": 0.6639, + "step": 13650 + }, + { + "epoch": 2.430377492877493, + "grad_norm": 0.7960709929466248, + "learning_rate": 6.69815310985808e-05, + "loss": 0.9557, + "step": 13651 + }, + { + "epoch": 2.4305555555555554, + "grad_norm": 0.8336359262466431, + "learning_rate": 6.696831897928376e-05, + "loss": 1.0112, + "step": 13652 + }, + { + "epoch": 2.430733618233618, + "grad_norm": 0.8353996872901917, + "learning_rate": 6.695510750717384e-05, + "loss": 0.8578, + "step": 13653 + }, + { + "epoch": 2.430911680911681, + "grad_norm": 0.8968163132667542, + "learning_rate": 6.694189668250996e-05, + "loss": 1.0412, + "step": 13654 + }, + { + "epoch": 2.4310897435897436, + "grad_norm": 0.8091850876808167, + "learning_rate": 6.692868650555093e-05, + "loss": 0.761, + "step": 13655 + }, + { + "epoch": 2.4312678062678064, + "grad_norm": 0.7735705375671387, + "learning_rate": 6.691547697655563e-05, + "loss": 1.0507, + "step": 13656 + }, + { + "epoch": 2.431445868945869, + "grad_norm": 0.7707101702690125, + "learning_rate": 6.690226809578279e-05, + "loss": 0.8883, + "step": 13657 + }, + { + "epoch": 2.431623931623932, + "grad_norm": 0.8384747505187988, + "learning_rate": 6.688905986349127e-05, + "loss": 0.7772, + "step": 13658 + }, + { + "epoch": 2.431801994301994, + "grad_norm": 0.7676185369491577, + "learning_rate": 6.687585227993985e-05, + "loss": 0.844, + "step": 13659 + }, + { + "epoch": 2.431980056980057, + "grad_norm": 0.8745819926261902, + "learning_rate": 6.686264534538726e-05, + "loss": 0.9996, + "step": 13660 + }, + { + "epoch": 2.4321581196581197, + "grad_norm": 0.7455142736434937, + "learning_rate": 6.684943906009232e-05, + "loss": 0.7133, + "step": 13661 + }, + { + "epoch": 2.4323361823361824, + "grad_norm": 0.8742238879203796, + "learning_rate": 6.683623342431378e-05, + "loss": 0.8155, + "step": 13662 + }, + { + "epoch": 2.432514245014245, + "grad_norm": 0.7863791584968567, + "learning_rate": 6.68230284383103e-05, + "loss": 1.0017, + "step": 13663 + }, + { + "epoch": 2.4326923076923075, + "grad_norm": 0.9469232559204102, + "learning_rate": 6.68098241023407e-05, + "loss": 0.8172, + "step": 13664 + }, + { + "epoch": 2.4328703703703702, + "grad_norm": 0.808024525642395, + "learning_rate": 6.679662041666362e-05, + "loss": 0.8813, + "step": 13665 + }, + { + "epoch": 2.433048433048433, + "grad_norm": 0.834863543510437, + "learning_rate": 6.67834173815378e-05, + "loss": 0.9765, + "step": 13666 + }, + { + "epoch": 2.4332264957264957, + "grad_norm": 0.8903583288192749, + "learning_rate": 6.677021499722193e-05, + "loss": 0.9393, + "step": 13667 + }, + { + "epoch": 2.4334045584045585, + "grad_norm": 0.8341929912567139, + "learning_rate": 6.675701326397466e-05, + "loss": 0.8681, + "step": 13668 + }, + { + "epoch": 2.4335826210826212, + "grad_norm": 0.9348049163818359, + "learning_rate": 6.674381218205465e-05, + "loss": 0.7396, + "step": 13669 + }, + { + "epoch": 2.433760683760684, + "grad_norm": 0.8898159861564636, + "learning_rate": 6.673061175172055e-05, + "loss": 0.8638, + "step": 13670 + }, + { + "epoch": 2.4339387464387463, + "grad_norm": 0.8101391792297363, + "learning_rate": 6.671741197323105e-05, + "loss": 0.8064, + "step": 13671 + }, + { + "epoch": 2.434116809116809, + "grad_norm": 0.8756688237190247, + "learning_rate": 6.670421284684467e-05, + "loss": 0.7718, + "step": 13672 + }, + { + "epoch": 2.434294871794872, + "grad_norm": 0.8060923218727112, + "learning_rate": 6.669101437282012e-05, + "loss": 0.8137, + "step": 13673 + }, + { + "epoch": 2.4344729344729346, + "grad_norm": 0.792891800403595, + "learning_rate": 6.667781655141589e-05, + "loss": 0.9104, + "step": 13674 + }, + { + "epoch": 2.4346509971509973, + "grad_norm": 0.8590527772903442, + "learning_rate": 6.666461938289068e-05, + "loss": 0.9578, + "step": 13675 + }, + { + "epoch": 2.4348290598290596, + "grad_norm": 0.8593253493309021, + "learning_rate": 6.665142286750297e-05, + "loss": 0.8083, + "step": 13676 + }, + { + "epoch": 2.4350071225071224, + "grad_norm": 0.8237900733947754, + "learning_rate": 6.663822700551137e-05, + "loss": 0.8096, + "step": 13677 + }, + { + "epoch": 2.435185185185185, + "grad_norm": 0.9017227292060852, + "learning_rate": 6.66250317971744e-05, + "loss": 0.9599, + "step": 13678 + }, + { + "epoch": 2.435363247863248, + "grad_norm": 0.7811765670776367, + "learning_rate": 6.661183724275061e-05, + "loss": 0.8392, + "step": 13679 + }, + { + "epoch": 2.4355413105413106, + "grad_norm": 0.8113176822662354, + "learning_rate": 6.659864334249848e-05, + "loss": 0.6788, + "step": 13680 + }, + { + "epoch": 2.4357193732193734, + "grad_norm": 0.795261561870575, + "learning_rate": 6.65854500966766e-05, + "loss": 0.783, + "step": 13681 + }, + { + "epoch": 2.435897435897436, + "grad_norm": 0.7738518714904785, + "learning_rate": 6.657225750554338e-05, + "loss": 0.676, + "step": 13682 + }, + { + "epoch": 2.4360754985754984, + "grad_norm": 0.8513518571853638, + "learning_rate": 6.655906556935737e-05, + "loss": 0.8743, + "step": 13683 + }, + { + "epoch": 2.436253561253561, + "grad_norm": 0.9595896005630493, + "learning_rate": 6.654587428837696e-05, + "loss": 0.9333, + "step": 13684 + }, + { + "epoch": 2.436431623931624, + "grad_norm": 0.7720373272895813, + "learning_rate": 6.653268366286066e-05, + "loss": 0.77, + "step": 13685 + }, + { + "epoch": 2.4366096866096867, + "grad_norm": 0.9022032022476196, + "learning_rate": 6.651949369306689e-05, + "loss": 0.8786, + "step": 13686 + }, + { + "epoch": 2.4367877492877494, + "grad_norm": 0.795092761516571, + "learning_rate": 6.650630437925409e-05, + "loss": 0.8557, + "step": 13687 + }, + { + "epoch": 2.4369658119658117, + "grad_norm": 0.8517789244651794, + "learning_rate": 6.649311572168072e-05, + "loss": 0.9781, + "step": 13688 + }, + { + "epoch": 2.4371438746438745, + "grad_norm": 1.0523463487625122, + "learning_rate": 6.64799277206051e-05, + "loss": 0.9046, + "step": 13689 + }, + { + "epoch": 2.4373219373219372, + "grad_norm": 0.9401832818984985, + "learning_rate": 6.646674037628568e-05, + "loss": 1.0966, + "step": 13690 + }, + { + "epoch": 2.4375, + "grad_norm": 0.9895738959312439, + "learning_rate": 6.645355368898082e-05, + "loss": 0.714, + "step": 13691 + }, + { + "epoch": 2.4376780626780628, + "grad_norm": 0.8458610773086548, + "learning_rate": 6.644036765894892e-05, + "loss": 0.6265, + "step": 13692 + }, + { + "epoch": 2.4378561253561255, + "grad_norm": 0.7908345460891724, + "learning_rate": 6.642718228644826e-05, + "loss": 0.7943, + "step": 13693 + }, + { + "epoch": 2.4380341880341883, + "grad_norm": 0.8119938373565674, + "learning_rate": 6.641399757173725e-05, + "loss": 0.5948, + "step": 13694 + }, + { + "epoch": 2.4382122507122506, + "grad_norm": 0.8175633549690247, + "learning_rate": 6.640081351507417e-05, + "loss": 0.9098, + "step": 13695 + }, + { + "epoch": 2.4383903133903133, + "grad_norm": 0.8546686768531799, + "learning_rate": 6.638763011671736e-05, + "loss": 1.0347, + "step": 13696 + }, + { + "epoch": 2.438568376068376, + "grad_norm": 0.812406599521637, + "learning_rate": 6.637444737692508e-05, + "loss": 0.8469, + "step": 13697 + }, + { + "epoch": 2.438746438746439, + "grad_norm": 0.7802549004554749, + "learning_rate": 6.636126529595572e-05, + "loss": 0.7024, + "step": 13698 + }, + { + "epoch": 2.4389245014245016, + "grad_norm": 0.8046648502349854, + "learning_rate": 6.634808387406744e-05, + "loss": 0.8292, + "step": 13699 + }, + { + "epoch": 2.439102564102564, + "grad_norm": 0.8544600009918213, + "learning_rate": 6.633490311151857e-05, + "loss": 0.8033, + "step": 13700 + }, + { + "epoch": 2.4392806267806266, + "grad_norm": 0.8327271938323975, + "learning_rate": 6.632172300856731e-05, + "loss": 0.8641, + "step": 13701 + }, + { + "epoch": 2.4394586894586894, + "grad_norm": 0.9563352465629578, + "learning_rate": 6.630854356547199e-05, + "loss": 0.8144, + "step": 13702 + }, + { + "epoch": 2.439636752136752, + "grad_norm": 0.8993256092071533, + "learning_rate": 6.629536478249071e-05, + "loss": 0.8688, + "step": 13703 + }, + { + "epoch": 2.439814814814815, + "grad_norm": 0.8741861581802368, + "learning_rate": 6.628218665988178e-05, + "loss": 0.9868, + "step": 13704 + }, + { + "epoch": 2.4399928774928776, + "grad_norm": 0.7898648381233215, + "learning_rate": 6.626900919790332e-05, + "loss": 0.7802, + "step": 13705 + }, + { + "epoch": 2.4401709401709404, + "grad_norm": 0.7651925086975098, + "learning_rate": 6.625583239681357e-05, + "loss": 0.8131, + "step": 13706 + }, + { + "epoch": 2.4403490028490027, + "grad_norm": 0.7917741537094116, + "learning_rate": 6.624265625687071e-05, + "loss": 0.8581, + "step": 13707 + }, + { + "epoch": 2.4405270655270654, + "grad_norm": 0.7631075978279114, + "learning_rate": 6.622948077833284e-05, + "loss": 0.6069, + "step": 13708 + }, + { + "epoch": 2.440705128205128, + "grad_norm": 0.920765221118927, + "learning_rate": 6.621630596145819e-05, + "loss": 0.6846, + "step": 13709 + }, + { + "epoch": 2.440883190883191, + "grad_norm": 0.822335422039032, + "learning_rate": 6.62031318065048e-05, + "loss": 1.0309, + "step": 13710 + }, + { + "epoch": 2.4410612535612537, + "grad_norm": 0.7978029251098633, + "learning_rate": 6.618995831373086e-05, + "loss": 0.9593, + "step": 13711 + }, + { + "epoch": 2.441239316239316, + "grad_norm": 0.8908950686454773, + "learning_rate": 6.617678548339443e-05, + "loss": 0.7147, + "step": 13712 + }, + { + "epoch": 2.4414173789173788, + "grad_norm": 0.7772884368896484, + "learning_rate": 6.616361331575368e-05, + "loss": 0.8839, + "step": 13713 + }, + { + "epoch": 2.4415954415954415, + "grad_norm": 0.8437771797180176, + "learning_rate": 6.615044181106658e-05, + "loss": 0.8388, + "step": 13714 + }, + { + "epoch": 2.4417735042735043, + "grad_norm": 0.8549850583076477, + "learning_rate": 6.613727096959128e-05, + "loss": 0.9251, + "step": 13715 + }, + { + "epoch": 2.441951566951567, + "grad_norm": 0.8402581810951233, + "learning_rate": 6.612410079158579e-05, + "loss": 0.8607, + "step": 13716 + }, + { + "epoch": 2.4421296296296298, + "grad_norm": 0.8181160688400269, + "learning_rate": 6.611093127730821e-05, + "loss": 0.6082, + "step": 13717 + }, + { + "epoch": 2.4423076923076925, + "grad_norm": 0.9006236791610718, + "learning_rate": 6.609776242701651e-05, + "loss": 0.9091, + "step": 13718 + }, + { + "epoch": 2.442485754985755, + "grad_norm": 0.7759920358657837, + "learning_rate": 6.608459424096876e-05, + "loss": 0.842, + "step": 13719 + }, + { + "epoch": 2.4426638176638176, + "grad_norm": 0.825701117515564, + "learning_rate": 6.60714267194229e-05, + "loss": 0.9325, + "step": 13720 + }, + { + "epoch": 2.4428418803418803, + "grad_norm": 0.7646961212158203, + "learning_rate": 6.605825986263697e-05, + "loss": 0.8124, + "step": 13721 + }, + { + "epoch": 2.443019943019943, + "grad_norm": 0.896112322807312, + "learning_rate": 6.604509367086888e-05, + "loss": 1.0962, + "step": 13722 + }, + { + "epoch": 2.443198005698006, + "grad_norm": 0.8079821467399597, + "learning_rate": 6.603192814437672e-05, + "loss": 0.8195, + "step": 13723 + }, + { + "epoch": 2.443376068376068, + "grad_norm": 0.8901529908180237, + "learning_rate": 6.601876328341831e-05, + "loss": 0.7886, + "step": 13724 + }, + { + "epoch": 2.443554131054131, + "grad_norm": 1.0454550981521606, + "learning_rate": 6.600559908825168e-05, + "loss": 0.9642, + "step": 13725 + }, + { + "epoch": 2.4437321937321936, + "grad_norm": 0.7995026707649231, + "learning_rate": 6.599243555913469e-05, + "loss": 0.6927, + "step": 13726 + }, + { + "epoch": 2.4439102564102564, + "grad_norm": 0.9235756397247314, + "learning_rate": 6.597927269632526e-05, + "loss": 0.8986, + "step": 13727 + }, + { + "epoch": 2.444088319088319, + "grad_norm": 0.7869365215301514, + "learning_rate": 6.596611050008137e-05, + "loss": 0.7592, + "step": 13728 + }, + { + "epoch": 2.444266381766382, + "grad_norm": 0.8172873258590698, + "learning_rate": 6.595294897066081e-05, + "loss": 0.8048, + "step": 13729 + }, + { + "epoch": 2.4444444444444446, + "grad_norm": 0.8021790981292725, + "learning_rate": 6.593978810832152e-05, + "loss": 0.767, + "step": 13730 + }, + { + "epoch": 2.444622507122507, + "grad_norm": 0.7781784534454346, + "learning_rate": 6.592662791332129e-05, + "loss": 0.8364, + "step": 13731 + }, + { + "epoch": 2.4448005698005697, + "grad_norm": 0.8227871656417847, + "learning_rate": 6.591346838591803e-05, + "loss": 0.8967, + "step": 13732 + }, + { + "epoch": 2.4449786324786325, + "grad_norm": 0.8349295854568481, + "learning_rate": 6.590030952636952e-05, + "loss": 0.8593, + "step": 13733 + }, + { + "epoch": 2.445156695156695, + "grad_norm": 1.0261762142181396, + "learning_rate": 6.588715133493365e-05, + "loss": 0.701, + "step": 13734 + }, + { + "epoch": 2.445334757834758, + "grad_norm": 0.8612635731697083, + "learning_rate": 6.587399381186814e-05, + "loss": 0.9803, + "step": 13735 + }, + { + "epoch": 2.4455128205128207, + "grad_norm": 0.7890039682388306, + "learning_rate": 6.586083695743086e-05, + "loss": 0.8001, + "step": 13736 + }, + { + "epoch": 2.445690883190883, + "grad_norm": 0.9934018850326538, + "learning_rate": 6.584768077187955e-05, + "loss": 1.0089, + "step": 13737 + }, + { + "epoch": 2.4458689458689458, + "grad_norm": 0.8232909440994263, + "learning_rate": 6.583452525547202e-05, + "loss": 0.8088, + "step": 13738 + }, + { + "epoch": 2.4460470085470085, + "grad_norm": 0.8635872006416321, + "learning_rate": 6.582137040846595e-05, + "loss": 1.0007, + "step": 13739 + }, + { + "epoch": 2.4462250712250713, + "grad_norm": 0.905575692653656, + "learning_rate": 6.580821623111914e-05, + "loss": 0.7577, + "step": 13740 + }, + { + "epoch": 2.446403133903134, + "grad_norm": 0.8264324069023132, + "learning_rate": 6.579506272368931e-05, + "loss": 0.9337, + "step": 13741 + }, + { + "epoch": 2.4465811965811968, + "grad_norm": 0.9100900292396545, + "learning_rate": 6.57819098864342e-05, + "loss": 0.8115, + "step": 13742 + }, + { + "epoch": 2.446759259259259, + "grad_norm": 0.8536351919174194, + "learning_rate": 6.576875771961145e-05, + "loss": 0.8612, + "step": 13743 + }, + { + "epoch": 2.446937321937322, + "grad_norm": 0.8968019485473633, + "learning_rate": 6.57556062234788e-05, + "loss": 0.8798, + "step": 13744 + }, + { + "epoch": 2.4471153846153846, + "grad_norm": 0.8745046854019165, + "learning_rate": 6.574245539829389e-05, + "loss": 0.8992, + "step": 13745 + }, + { + "epoch": 2.4472934472934473, + "grad_norm": 0.8336703777313232, + "learning_rate": 6.57293052443144e-05, + "loss": 0.7947, + "step": 13746 + }, + { + "epoch": 2.44747150997151, + "grad_norm": 0.8544902801513672, + "learning_rate": 6.571615576179801e-05, + "loss": 1.0315, + "step": 13747 + }, + { + "epoch": 2.447649572649573, + "grad_norm": 0.848242461681366, + "learning_rate": 6.570300695100229e-05, + "loss": 0.853, + "step": 13748 + }, + { + "epoch": 2.447827635327635, + "grad_norm": 0.7753778100013733, + "learning_rate": 6.568985881218496e-05, + "loss": 0.934, + "step": 13749 + }, + { + "epoch": 2.448005698005698, + "grad_norm": 0.8294853568077087, + "learning_rate": 6.567671134560351e-05, + "loss": 0.9328, + "step": 13750 + }, + { + "epoch": 2.4481837606837606, + "grad_norm": 0.8720992803573608, + "learning_rate": 6.566356455151565e-05, + "loss": 0.875, + "step": 13751 + }, + { + "epoch": 2.4483618233618234, + "grad_norm": 0.8204464316368103, + "learning_rate": 6.565041843017888e-05, + "loss": 0.923, + "step": 13752 + }, + { + "epoch": 2.448539886039886, + "grad_norm": 0.940037190914154, + "learning_rate": 6.563727298185085e-05, + "loss": 1.1596, + "step": 13753 + }, + { + "epoch": 2.448717948717949, + "grad_norm": 0.8390263915061951, + "learning_rate": 6.562412820678902e-05, + "loss": 0.8256, + "step": 13754 + }, + { + "epoch": 2.448896011396011, + "grad_norm": 0.8572748303413391, + "learning_rate": 6.561098410525106e-05, + "loss": 0.7833, + "step": 13755 + }, + { + "epoch": 2.449074074074074, + "grad_norm": 0.7981020212173462, + "learning_rate": 6.559784067749436e-05, + "loss": 0.7609, + "step": 13756 + }, + { + "epoch": 2.4492521367521367, + "grad_norm": 0.8356930613517761, + "learning_rate": 6.558469792377653e-05, + "loss": 0.7542, + "step": 13757 + }, + { + "epoch": 2.4494301994301995, + "grad_norm": 0.9340906739234924, + "learning_rate": 6.557155584435504e-05, + "loss": 0.9898, + "step": 13758 + }, + { + "epoch": 2.449608262108262, + "grad_norm": 1.0551100969314575, + "learning_rate": 6.555841443948743e-05, + "loss": 0.7189, + "step": 13759 + }, + { + "epoch": 2.449786324786325, + "grad_norm": 0.9572125673294067, + "learning_rate": 6.554527370943111e-05, + "loss": 0.8878, + "step": 13760 + }, + { + "epoch": 2.4499643874643873, + "grad_norm": 0.8760324716567993, + "learning_rate": 6.55321336544436e-05, + "loss": 0.75, + "step": 13761 + }, + { + "epoch": 2.45014245014245, + "grad_norm": 0.7599226236343384, + "learning_rate": 6.55189942747823e-05, + "loss": 0.9222, + "step": 13762 + }, + { + "epoch": 2.4503205128205128, + "grad_norm": 0.7307319045066833, + "learning_rate": 6.550585557070473e-05, + "loss": 0.833, + "step": 13763 + }, + { + "epoch": 2.4504985754985755, + "grad_norm": 0.8022613525390625, + "learning_rate": 6.549271754246822e-05, + "loss": 0.9439, + "step": 13764 + }, + { + "epoch": 2.4506766381766383, + "grad_norm": 0.7447740435600281, + "learning_rate": 6.547958019033024e-05, + "loss": 0.7803, + "step": 13765 + }, + { + "epoch": 2.450854700854701, + "grad_norm": 0.9021183252334595, + "learning_rate": 6.546644351454818e-05, + "loss": 0.8373, + "step": 13766 + }, + { + "epoch": 2.451032763532764, + "grad_norm": 0.8230152726173401, + "learning_rate": 6.545330751537941e-05, + "loss": 0.7023, + "step": 13767 + }, + { + "epoch": 2.451210826210826, + "grad_norm": 0.9581316113471985, + "learning_rate": 6.544017219308132e-05, + "loss": 1.0024, + "step": 13768 + }, + { + "epoch": 2.451388888888889, + "grad_norm": 0.7969945073127747, + "learning_rate": 6.542703754791127e-05, + "loss": 0.8996, + "step": 13769 + }, + { + "epoch": 2.4515669515669516, + "grad_norm": 0.864604115486145, + "learning_rate": 6.54139035801266e-05, + "loss": 0.9325, + "step": 13770 + }, + { + "epoch": 2.4517450142450143, + "grad_norm": 0.8156671524047852, + "learning_rate": 6.540077028998463e-05, + "loss": 0.7926, + "step": 13771 + }, + { + "epoch": 2.451923076923077, + "grad_norm": 0.8704202175140381, + "learning_rate": 6.538763767774272e-05, + "loss": 0.8855, + "step": 13772 + }, + { + "epoch": 2.4521011396011394, + "grad_norm": 0.7533015012741089, + "learning_rate": 6.537450574365811e-05, + "loss": 0.9322, + "step": 13773 + }, + { + "epoch": 2.452279202279202, + "grad_norm": 0.8272553086280823, + "learning_rate": 6.536137448798819e-05, + "loss": 0.7474, + "step": 13774 + }, + { + "epoch": 2.452457264957265, + "grad_norm": 0.7788257598876953, + "learning_rate": 6.534824391099013e-05, + "loss": 0.7163, + "step": 13775 + }, + { + "epoch": 2.4526353276353277, + "grad_norm": 0.8309275507926941, + "learning_rate": 6.533511401292125e-05, + "loss": 1.1595, + "step": 13776 + }, + { + "epoch": 2.4528133903133904, + "grad_norm": 0.8369085788726807, + "learning_rate": 6.53219847940388e-05, + "loss": 0.7211, + "step": 13777 + }, + { + "epoch": 2.452991452991453, + "grad_norm": 0.8571248054504395, + "learning_rate": 6.530885625460007e-05, + "loss": 0.729, + "step": 13778 + }, + { + "epoch": 2.453169515669516, + "grad_norm": 0.7579928040504456, + "learning_rate": 6.529572839486217e-05, + "loss": 0.8799, + "step": 13779 + }, + { + "epoch": 2.453347578347578, + "grad_norm": 0.822463870048523, + "learning_rate": 6.528260121508245e-05, + "loss": 0.7948, + "step": 13780 + }, + { + "epoch": 2.453525641025641, + "grad_norm": 0.7910317778587341, + "learning_rate": 6.526947471551798e-05, + "loss": 0.7727, + "step": 13781 + }, + { + "epoch": 2.4537037037037037, + "grad_norm": 0.9321692585945129, + "learning_rate": 6.525634889642605e-05, + "loss": 0.8754, + "step": 13782 + }, + { + "epoch": 2.4538817663817665, + "grad_norm": 1.0130813121795654, + "learning_rate": 6.524322375806374e-05, + "loss": 0.7845, + "step": 13783 + }, + { + "epoch": 2.4540598290598292, + "grad_norm": 0.7254214882850647, + "learning_rate": 6.52300993006883e-05, + "loss": 0.4665, + "step": 13784 + }, + { + "epoch": 2.4542378917378915, + "grad_norm": 0.7874964475631714, + "learning_rate": 6.521697552455683e-05, + "loss": 0.8535, + "step": 13785 + }, + { + "epoch": 2.4544159544159543, + "grad_norm": 0.8275010585784912, + "learning_rate": 6.520385242992644e-05, + "loss": 0.7744, + "step": 13786 + }, + { + "epoch": 2.454594017094017, + "grad_norm": 0.7972453236579895, + "learning_rate": 6.519073001705431e-05, + "loss": 0.9494, + "step": 13787 + }, + { + "epoch": 2.45477207977208, + "grad_norm": 0.8763988018035889, + "learning_rate": 6.517760828619748e-05, + "loss": 0.8043, + "step": 13788 + }, + { + "epoch": 2.4549501424501425, + "grad_norm": 0.7948910593986511, + "learning_rate": 6.516448723761315e-05, + "loss": 0.7218, + "step": 13789 + }, + { + "epoch": 2.4551282051282053, + "grad_norm": 0.9416671395301819, + "learning_rate": 6.515136687155825e-05, + "loss": 0.7866, + "step": 13790 + }, + { + "epoch": 2.455306267806268, + "grad_norm": 0.8702704906463623, + "learning_rate": 6.513824718828999e-05, + "loss": 1.1579, + "step": 13791 + }, + { + "epoch": 2.4554843304843303, + "grad_norm": 0.8148752450942993, + "learning_rate": 6.51251281880653e-05, + "loss": 0.897, + "step": 13792 + }, + { + "epoch": 2.455662393162393, + "grad_norm": 0.8088299036026001, + "learning_rate": 6.511200987114132e-05, + "loss": 0.939, + "step": 13793 + }, + { + "epoch": 2.455840455840456, + "grad_norm": 0.9836809635162354, + "learning_rate": 6.509889223777499e-05, + "loss": 0.8841, + "step": 13794 + }, + { + "epoch": 2.4560185185185186, + "grad_norm": 0.7677251696586609, + "learning_rate": 6.508577528822342e-05, + "loss": 0.7816, + "step": 13795 + }, + { + "epoch": 2.4561965811965814, + "grad_norm": 0.835421085357666, + "learning_rate": 6.507265902274351e-05, + "loss": 0.9302, + "step": 13796 + }, + { + "epoch": 2.4563746438746437, + "grad_norm": 0.8892473578453064, + "learning_rate": 6.50595434415923e-05, + "loss": 0.8281, + "step": 13797 + }, + { + "epoch": 2.4565527065527064, + "grad_norm": 0.810459315776825, + "learning_rate": 6.504642854502676e-05, + "loss": 0.81, + "step": 13798 + }, + { + "epoch": 2.456730769230769, + "grad_norm": 0.9277065992355347, + "learning_rate": 6.503331433330386e-05, + "loss": 0.7956, + "step": 13799 + }, + { + "epoch": 2.456908831908832, + "grad_norm": 0.861725389957428, + "learning_rate": 6.502020080668051e-05, + "loss": 0.7788, + "step": 13800 + }, + { + "epoch": 2.4570868945868947, + "grad_norm": 1.0000818967819214, + "learning_rate": 6.500708796541366e-05, + "loss": 0.9197, + "step": 13801 + }, + { + "epoch": 2.4572649572649574, + "grad_norm": 0.920998215675354, + "learning_rate": 6.499397580976024e-05, + "loss": 0.7816, + "step": 13802 + }, + { + "epoch": 2.45744301994302, + "grad_norm": 0.7574821710586548, + "learning_rate": 6.498086433997715e-05, + "loss": 0.8982, + "step": 13803 + }, + { + "epoch": 2.4576210826210825, + "grad_norm": 1.026700496673584, + "learning_rate": 6.496775355632125e-05, + "loss": 1.131, + "step": 13804 + }, + { + "epoch": 2.4577991452991452, + "grad_norm": 0.7532633543014526, + "learning_rate": 6.495464345904945e-05, + "loss": 0.7998, + "step": 13805 + }, + { + "epoch": 2.457977207977208, + "grad_norm": 0.7380105257034302, + "learning_rate": 6.494153404841865e-05, + "loss": 0.7656, + "step": 13806 + }, + { + "epoch": 2.4581552706552707, + "grad_norm": 0.7933080792427063, + "learning_rate": 6.492842532468561e-05, + "loss": 0.7419, + "step": 13807 + }, + { + "epoch": 2.4583333333333335, + "grad_norm": 0.7731907963752747, + "learning_rate": 6.491531728810724e-05, + "loss": 0.8334, + "step": 13808 + }, + { + "epoch": 2.458511396011396, + "grad_norm": 0.7368177771568298, + "learning_rate": 6.490220993894035e-05, + "loss": 0.6184, + "step": 13809 + }, + { + "epoch": 2.4586894586894585, + "grad_norm": 0.8381120562553406, + "learning_rate": 6.488910327744178e-05, + "loss": 0.7875, + "step": 13810 + }, + { + "epoch": 2.4588675213675213, + "grad_norm": 0.910142183303833, + "learning_rate": 6.487599730386824e-05, + "loss": 0.8216, + "step": 13811 + }, + { + "epoch": 2.459045584045584, + "grad_norm": 0.9005017876625061, + "learning_rate": 6.48628920184766e-05, + "loss": 0.8928, + "step": 13812 + }, + { + "epoch": 2.459223646723647, + "grad_norm": 0.8437321782112122, + "learning_rate": 6.484978742152358e-05, + "loss": 0.9243, + "step": 13813 + }, + { + "epoch": 2.4594017094017095, + "grad_norm": 0.9145610928535461, + "learning_rate": 6.483668351326599e-05, + "loss": 0.9759, + "step": 13814 + }, + { + "epoch": 2.4595797720797723, + "grad_norm": 0.8391930460929871, + "learning_rate": 6.48235802939605e-05, + "loss": 0.8021, + "step": 13815 + }, + { + "epoch": 2.4597578347578346, + "grad_norm": 0.8035653233528137, + "learning_rate": 6.481047776386394e-05, + "loss": 0.8622, + "step": 13816 + }, + { + "epoch": 2.4599358974358974, + "grad_norm": 0.7238573431968689, + "learning_rate": 6.479737592323291e-05, + "loss": 0.6864, + "step": 13817 + }, + { + "epoch": 2.46011396011396, + "grad_norm": 0.8629193902015686, + "learning_rate": 6.47842747723242e-05, + "loss": 0.6107, + "step": 13818 + }, + { + "epoch": 2.460292022792023, + "grad_norm": 0.9015333652496338, + "learning_rate": 6.477117431139444e-05, + "loss": 0.845, + "step": 13819 + }, + { + "epoch": 2.4604700854700856, + "grad_norm": 0.8544989824295044, + "learning_rate": 6.47580745407004e-05, + "loss": 0.7103, + "step": 13820 + }, + { + "epoch": 2.460648148148148, + "grad_norm": 0.7665401101112366, + "learning_rate": 6.474497546049862e-05, + "loss": 0.6864, + "step": 13821 + }, + { + "epoch": 2.4608262108262107, + "grad_norm": 0.8640002012252808, + "learning_rate": 6.473187707104584e-05, + "loss": 0.953, + "step": 13822 + }, + { + "epoch": 2.4610042735042734, + "grad_norm": 0.8563477396965027, + "learning_rate": 6.471877937259864e-05, + "loss": 0.7776, + "step": 13823 + }, + { + "epoch": 2.461182336182336, + "grad_norm": 0.8089157938957214, + "learning_rate": 6.470568236541371e-05, + "loss": 0.8273, + "step": 13824 + }, + { + "epoch": 2.461360398860399, + "grad_norm": 0.8710005283355713, + "learning_rate": 6.469258604974757e-05, + "loss": 0.8827, + "step": 13825 + }, + { + "epoch": 2.4615384615384617, + "grad_norm": 0.7780489325523376, + "learning_rate": 6.467949042585688e-05, + "loss": 0.8035, + "step": 13826 + }, + { + "epoch": 2.4617165242165244, + "grad_norm": 1.010976791381836, + "learning_rate": 6.466639549399822e-05, + "loss": 0.7442, + "step": 13827 + }, + { + "epoch": 2.4618945868945867, + "grad_norm": 0.714077889919281, + "learning_rate": 6.465330125442812e-05, + "loss": 0.694, + "step": 13828 + }, + { + "epoch": 2.4620726495726495, + "grad_norm": 0.7408512234687805, + "learning_rate": 6.464020770740316e-05, + "loss": 0.6709, + "step": 13829 + }, + { + "epoch": 2.4622507122507122, + "grad_norm": 0.8433945178985596, + "learning_rate": 6.462711485317987e-05, + "loss": 0.9127, + "step": 13830 + }, + { + "epoch": 2.462428774928775, + "grad_norm": 1.031745195388794, + "learning_rate": 6.461402269201481e-05, + "loss": 0.9105, + "step": 13831 + }, + { + "epoch": 2.4626068376068377, + "grad_norm": 0.8884360790252686, + "learning_rate": 6.460093122416444e-05, + "loss": 0.9354, + "step": 13832 + }, + { + "epoch": 2.4627849002849, + "grad_norm": 0.8466372489929199, + "learning_rate": 6.45878404498853e-05, + "loss": 0.9554, + "step": 13833 + }, + { + "epoch": 2.462962962962963, + "grad_norm": 0.9026118516921997, + "learning_rate": 6.457475036943386e-05, + "loss": 0.928, + "step": 13834 + }, + { + "epoch": 2.4631410256410255, + "grad_norm": 0.9034590721130371, + "learning_rate": 6.456166098306661e-05, + "loss": 0.7694, + "step": 13835 + }, + { + "epoch": 2.4633190883190883, + "grad_norm": 0.8369483947753906, + "learning_rate": 6.454857229103998e-05, + "loss": 0.928, + "step": 13836 + }, + { + "epoch": 2.463497150997151, + "grad_norm": 0.8670645356178284, + "learning_rate": 6.453548429361045e-05, + "loss": 0.971, + "step": 13837 + }, + { + "epoch": 2.463675213675214, + "grad_norm": 0.8415539860725403, + "learning_rate": 6.452239699103442e-05, + "loss": 0.8461, + "step": 13838 + }, + { + "epoch": 2.4638532763532766, + "grad_norm": 0.7434490323066711, + "learning_rate": 6.450931038356834e-05, + "loss": 0.6677, + "step": 13839 + }, + { + "epoch": 2.464031339031339, + "grad_norm": 0.8113850355148315, + "learning_rate": 6.449622447146855e-05, + "loss": 0.8644, + "step": 13840 + }, + { + "epoch": 2.4642094017094016, + "grad_norm": 0.7424083352088928, + "learning_rate": 6.448313925499154e-05, + "loss": 0.7469, + "step": 13841 + }, + { + "epoch": 2.4643874643874644, + "grad_norm": 1.006949782371521, + "learning_rate": 6.44700547343936e-05, + "loss": 0.8426, + "step": 13842 + }, + { + "epoch": 2.464565527065527, + "grad_norm": 1.0643857717514038, + "learning_rate": 6.445697090993117e-05, + "loss": 0.7793, + "step": 13843 + }, + { + "epoch": 2.46474358974359, + "grad_norm": 0.8716835975646973, + "learning_rate": 6.444388778186051e-05, + "loss": 0.9684, + "step": 13844 + }, + { + "epoch": 2.464921652421652, + "grad_norm": 0.8677120804786682, + "learning_rate": 6.443080535043802e-05, + "loss": 0.9772, + "step": 13845 + }, + { + "epoch": 2.465099715099715, + "grad_norm": 0.8955141305923462, + "learning_rate": 6.441772361592005e-05, + "loss": 0.881, + "step": 13846 + }, + { + "epoch": 2.4652777777777777, + "grad_norm": 0.806794285774231, + "learning_rate": 6.440464257856283e-05, + "loss": 0.6393, + "step": 13847 + }, + { + "epoch": 2.4654558404558404, + "grad_norm": 0.8438352346420288, + "learning_rate": 6.439156223862272e-05, + "loss": 0.8305, + "step": 13848 + }, + { + "epoch": 2.465633903133903, + "grad_norm": 0.828960120677948, + "learning_rate": 6.437848259635594e-05, + "loss": 0.8564, + "step": 13849 + }, + { + "epoch": 2.465811965811966, + "grad_norm": 0.790199339389801, + "learning_rate": 6.436540365201886e-05, + "loss": 0.6573, + "step": 13850 + }, + { + "epoch": 2.4659900284900287, + "grad_norm": 0.8476296663284302, + "learning_rate": 6.435232540586763e-05, + "loss": 0.6979, + "step": 13851 + }, + { + "epoch": 2.466168091168091, + "grad_norm": 0.6880464553833008, + "learning_rate": 6.433924785815857e-05, + "loss": 0.6925, + "step": 13852 + }, + { + "epoch": 2.4663461538461537, + "grad_norm": 0.9369434118270874, + "learning_rate": 6.432617100914782e-05, + "loss": 0.6891, + "step": 13853 + }, + { + "epoch": 2.4665242165242165, + "grad_norm": 0.8522159457206726, + "learning_rate": 6.431309485909166e-05, + "loss": 1.0786, + "step": 13854 + }, + { + "epoch": 2.4667022792022792, + "grad_norm": 0.8479002714157104, + "learning_rate": 6.430001940824625e-05, + "loss": 0.9293, + "step": 13855 + }, + { + "epoch": 2.466880341880342, + "grad_norm": 0.8382098078727722, + "learning_rate": 6.428694465686787e-05, + "loss": 1.1279, + "step": 13856 + }, + { + "epoch": 2.4670584045584047, + "grad_norm": 0.9641128182411194, + "learning_rate": 6.427387060521255e-05, + "loss": 0.7938, + "step": 13857 + }, + { + "epoch": 2.467236467236467, + "grad_norm": 0.8458924293518066, + "learning_rate": 6.426079725353656e-05, + "loss": 0.7804, + "step": 13858 + }, + { + "epoch": 2.46741452991453, + "grad_norm": 0.8023849725723267, + "learning_rate": 6.424772460209597e-05, + "loss": 0.718, + "step": 13859 + }, + { + "epoch": 2.4675925925925926, + "grad_norm": 0.800864040851593, + "learning_rate": 6.423465265114699e-05, + "loss": 0.7627, + "step": 13860 + }, + { + "epoch": 2.4677706552706553, + "grad_norm": 0.9159586429595947, + "learning_rate": 6.422158140094566e-05, + "loss": 0.8348, + "step": 13861 + }, + { + "epoch": 2.467948717948718, + "grad_norm": 0.7982872724533081, + "learning_rate": 6.420851085174817e-05, + "loss": 0.8092, + "step": 13862 + }, + { + "epoch": 2.468126780626781, + "grad_norm": 0.8847397565841675, + "learning_rate": 6.41954410038105e-05, + "loss": 0.9165, + "step": 13863 + }, + { + "epoch": 2.468304843304843, + "grad_norm": 0.7885190844535828, + "learning_rate": 6.418237185738882e-05, + "loss": 0.733, + "step": 13864 + }, + { + "epoch": 2.468482905982906, + "grad_norm": 0.902428150177002, + "learning_rate": 6.416930341273914e-05, + "loss": 0.8345, + "step": 13865 + }, + { + "epoch": 2.4686609686609686, + "grad_norm": 0.9344130158424377, + "learning_rate": 6.415623567011751e-05, + "loss": 0.8651, + "step": 13866 + }, + { + "epoch": 2.4688390313390314, + "grad_norm": 0.840679407119751, + "learning_rate": 6.414316862978003e-05, + "loss": 0.7534, + "step": 13867 + }, + { + "epoch": 2.469017094017094, + "grad_norm": 0.8799613118171692, + "learning_rate": 6.413010229198263e-05, + "loss": 0.9335, + "step": 13868 + }, + { + "epoch": 2.469195156695157, + "grad_norm": 1.1401816606521606, + "learning_rate": 6.411703665698142e-05, + "loss": 0.8902, + "step": 13869 + }, + { + "epoch": 2.469373219373219, + "grad_norm": 0.9733933210372925, + "learning_rate": 6.410397172503227e-05, + "loss": 0.9288, + "step": 13870 + }, + { + "epoch": 2.469551282051282, + "grad_norm": 0.9260223507881165, + "learning_rate": 6.409090749639128e-05, + "loss": 0.9664, + "step": 13871 + }, + { + "epoch": 2.4697293447293447, + "grad_norm": 1.0200423002243042, + "learning_rate": 6.407784397131433e-05, + "loss": 0.9095, + "step": 13872 + }, + { + "epoch": 2.4699074074074074, + "grad_norm": 0.8015561103820801, + "learning_rate": 6.406478115005743e-05, + "loss": 0.8015, + "step": 13873 + }, + { + "epoch": 2.47008547008547, + "grad_norm": 0.8035915493965149, + "learning_rate": 6.40517190328765e-05, + "loss": 0.9275, + "step": 13874 + }, + { + "epoch": 2.470263532763533, + "grad_norm": 0.8248090147972107, + "learning_rate": 6.403865762002743e-05, + "loss": 0.8736, + "step": 13875 + }, + { + "epoch": 2.4704415954415953, + "grad_norm": 0.9310920834541321, + "learning_rate": 6.402559691176616e-05, + "loss": 0.794, + "step": 13876 + }, + { + "epoch": 2.470619658119658, + "grad_norm": 0.7796428799629211, + "learning_rate": 6.401253690834863e-05, + "loss": 0.832, + "step": 13877 + }, + { + "epoch": 2.4707977207977208, + "grad_norm": 0.9046199321746826, + "learning_rate": 6.399947761003063e-05, + "loss": 0.9988, + "step": 13878 + }, + { + "epoch": 2.4709757834757835, + "grad_norm": 0.8970019221305847, + "learning_rate": 6.398641901706812e-05, + "loss": 0.6805, + "step": 13879 + }, + { + "epoch": 2.4711538461538463, + "grad_norm": 0.935786247253418, + "learning_rate": 6.397336112971688e-05, + "loss": 0.9826, + "step": 13880 + }, + { + "epoch": 2.471331908831909, + "grad_norm": 0.8497617244720459, + "learning_rate": 6.396030394823285e-05, + "loss": 0.8842, + "step": 13881 + }, + { + "epoch": 2.4715099715099713, + "grad_norm": 0.9159898161888123, + "learning_rate": 6.394724747287173e-05, + "loss": 0.8592, + "step": 13882 + }, + { + "epoch": 2.471688034188034, + "grad_norm": 0.891951858997345, + "learning_rate": 6.393419170388943e-05, + "loss": 0.8197, + "step": 13883 + }, + { + "epoch": 2.471866096866097, + "grad_norm": 0.9038097858428955, + "learning_rate": 6.392113664154172e-05, + "loss": 0.7535, + "step": 13884 + }, + { + "epoch": 2.4720441595441596, + "grad_norm": 0.8502489924430847, + "learning_rate": 6.390808228608438e-05, + "loss": 0.9183, + "step": 13885 + }, + { + "epoch": 2.4722222222222223, + "grad_norm": 0.7442654371261597, + "learning_rate": 6.389502863777323e-05, + "loss": 0.7741, + "step": 13886 + }, + { + "epoch": 2.472400284900285, + "grad_norm": 0.8811324238777161, + "learning_rate": 6.388197569686395e-05, + "loss": 0.7261, + "step": 13887 + }, + { + "epoch": 2.472578347578348, + "grad_norm": 0.9487552046775818, + "learning_rate": 6.386892346361239e-05, + "loss": 0.779, + "step": 13888 + }, + { + "epoch": 2.47275641025641, + "grad_norm": 0.8138917088508606, + "learning_rate": 6.385587193827416e-05, + "loss": 0.915, + "step": 13889 + }, + { + "epoch": 2.472934472934473, + "grad_norm": 0.7842695713043213, + "learning_rate": 6.384282112110506e-05, + "loss": 0.9458, + "step": 13890 + }, + { + "epoch": 2.4731125356125356, + "grad_norm": 0.7852116823196411, + "learning_rate": 6.382977101236074e-05, + "loss": 0.9515, + "step": 13891 + }, + { + "epoch": 2.4732905982905984, + "grad_norm": 0.8429296016693115, + "learning_rate": 6.381672161229698e-05, + "loss": 0.9466, + "step": 13892 + }, + { + "epoch": 2.473468660968661, + "grad_norm": 0.8713327050209045, + "learning_rate": 6.380367292116933e-05, + "loss": 0.7552, + "step": 13893 + }, + { + "epoch": 2.4736467236467234, + "grad_norm": 0.8153441548347473, + "learning_rate": 6.379062493923355e-05, + "loss": 0.7833, + "step": 13894 + }, + { + "epoch": 2.473824786324786, + "grad_norm": 0.8283601999282837, + "learning_rate": 6.377757766674526e-05, + "loss": 0.8986, + "step": 13895 + }, + { + "epoch": 2.474002849002849, + "grad_norm": 0.8116408586502075, + "learning_rate": 6.37645311039601e-05, + "loss": 0.8549, + "step": 13896 + }, + { + "epoch": 2.4741809116809117, + "grad_norm": 0.7999116778373718, + "learning_rate": 6.375148525113365e-05, + "loss": 0.8082, + "step": 13897 + }, + { + "epoch": 2.4743589743589745, + "grad_norm": 0.7307565808296204, + "learning_rate": 6.373844010852159e-05, + "loss": 0.7946, + "step": 13898 + }, + { + "epoch": 2.474537037037037, + "grad_norm": 0.7519806027412415, + "learning_rate": 6.372539567637941e-05, + "loss": 0.6511, + "step": 13899 + }, + { + "epoch": 2.4747150997151, + "grad_norm": 0.8571820259094238, + "learning_rate": 6.371235195496279e-05, + "loss": 0.8266, + "step": 13900 + }, + { + "epoch": 2.4748931623931623, + "grad_norm": 0.8118062019348145, + "learning_rate": 6.369930894452723e-05, + "loss": 0.8573, + "step": 13901 + }, + { + "epoch": 2.475071225071225, + "grad_norm": 0.8729892373085022, + "learning_rate": 6.368626664532833e-05, + "loss": 0.812, + "step": 13902 + }, + { + "epoch": 2.4752492877492878, + "grad_norm": 0.7663209438323975, + "learning_rate": 6.367322505762157e-05, + "loss": 0.6648, + "step": 13903 + }, + { + "epoch": 2.4754273504273505, + "grad_norm": 0.7913058996200562, + "learning_rate": 6.366018418166251e-05, + "loss": 0.7486, + "step": 13904 + }, + { + "epoch": 2.4756054131054133, + "grad_norm": 0.7714928984642029, + "learning_rate": 6.364714401770666e-05, + "loss": 0.9134, + "step": 13905 + }, + { + "epoch": 2.4757834757834756, + "grad_norm": 0.8226378560066223, + "learning_rate": 6.363410456600949e-05, + "loss": 0.903, + "step": 13906 + }, + { + "epoch": 2.4759615384615383, + "grad_norm": 0.8643919825553894, + "learning_rate": 6.362106582682653e-05, + "loss": 0.9068, + "step": 13907 + }, + { + "epoch": 2.476139601139601, + "grad_norm": 0.8390868306159973, + "learning_rate": 6.360802780041317e-05, + "loss": 0.8938, + "step": 13908 + }, + { + "epoch": 2.476317663817664, + "grad_norm": 0.9549261927604675, + "learning_rate": 6.359499048702495e-05, + "loss": 0.8431, + "step": 13909 + }, + { + "epoch": 2.4764957264957266, + "grad_norm": 0.7509152889251709, + "learning_rate": 6.358195388691726e-05, + "loss": 0.8661, + "step": 13910 + }, + { + "epoch": 2.4766737891737893, + "grad_norm": 0.9447416067123413, + "learning_rate": 6.356891800034552e-05, + "loss": 0.8766, + "step": 13911 + }, + { + "epoch": 2.476851851851852, + "grad_norm": 0.8972395062446594, + "learning_rate": 6.355588282756515e-05, + "loss": 0.8517, + "step": 13912 + }, + { + "epoch": 2.4770299145299144, + "grad_norm": 0.8645047545433044, + "learning_rate": 6.354284836883156e-05, + "loss": 0.8188, + "step": 13913 + }, + { + "epoch": 2.477207977207977, + "grad_norm": 0.7939230799674988, + "learning_rate": 6.35298146244001e-05, + "loss": 0.778, + "step": 13914 + }, + { + "epoch": 2.47738603988604, + "grad_norm": 0.7714613676071167, + "learning_rate": 6.351678159452618e-05, + "loss": 0.7964, + "step": 13915 + }, + { + "epoch": 2.4775641025641026, + "grad_norm": 0.9293754696846008, + "learning_rate": 6.350374927946512e-05, + "loss": 0.7965, + "step": 13916 + }, + { + "epoch": 2.4777421652421654, + "grad_norm": 0.8025050759315491, + "learning_rate": 6.349071767947233e-05, + "loss": 0.8222, + "step": 13917 + }, + { + "epoch": 2.4779202279202277, + "grad_norm": 0.7790399193763733, + "learning_rate": 6.347768679480304e-05, + "loss": 1.0294, + "step": 13918 + }, + { + "epoch": 2.4780982905982905, + "grad_norm": 1.0063512325286865, + "learning_rate": 6.346465662571261e-05, + "loss": 0.953, + "step": 13919 + }, + { + "epoch": 2.478276353276353, + "grad_norm": 0.8742708563804626, + "learning_rate": 6.345162717245634e-05, + "loss": 0.7119, + "step": 13920 + }, + { + "epoch": 2.478454415954416, + "grad_norm": 0.760497510433197, + "learning_rate": 6.343859843528955e-05, + "loss": 0.7446, + "step": 13921 + }, + { + "epoch": 2.4786324786324787, + "grad_norm": 0.7635362148284912, + "learning_rate": 6.342557041446743e-05, + "loss": 0.8286, + "step": 13922 + }, + { + "epoch": 2.4788105413105415, + "grad_norm": 0.7876720428466797, + "learning_rate": 6.341254311024532e-05, + "loss": 0.8007, + "step": 13923 + }, + { + "epoch": 2.478988603988604, + "grad_norm": 0.8144401907920837, + "learning_rate": 6.339951652287839e-05, + "loss": 0.9177, + "step": 13924 + }, + { + "epoch": 2.4791666666666665, + "grad_norm": 0.7400189638137817, + "learning_rate": 6.338649065262189e-05, + "loss": 0.8935, + "step": 13925 + }, + { + "epoch": 2.4793447293447293, + "grad_norm": 0.8412175178527832, + "learning_rate": 6.337346549973106e-05, + "loss": 0.6929, + "step": 13926 + }, + { + "epoch": 2.479522792022792, + "grad_norm": 1.0156967639923096, + "learning_rate": 6.336044106446108e-05, + "loss": 0.9843, + "step": 13927 + }, + { + "epoch": 2.4797008547008548, + "grad_norm": 0.8556809425354004, + "learning_rate": 6.33474173470672e-05, + "loss": 0.81, + "step": 13928 + }, + { + "epoch": 2.4798789173789175, + "grad_norm": 0.8103616237640381, + "learning_rate": 6.333439434780448e-05, + "loss": 0.925, + "step": 13929 + }, + { + "epoch": 2.48005698005698, + "grad_norm": 0.9460168480873108, + "learning_rate": 6.332137206692817e-05, + "loss": 0.7059, + "step": 13930 + }, + { + "epoch": 2.4802350427350426, + "grad_norm": 0.847226619720459, + "learning_rate": 6.330835050469334e-05, + "loss": 1.0139, + "step": 13931 + }, + { + "epoch": 2.4804131054131053, + "grad_norm": 0.7639240622520447, + "learning_rate": 6.329532966135523e-05, + "loss": 0.8141, + "step": 13932 + }, + { + "epoch": 2.480591168091168, + "grad_norm": 0.9273494482040405, + "learning_rate": 6.328230953716883e-05, + "loss": 0.9864, + "step": 13933 + }, + { + "epoch": 2.480769230769231, + "grad_norm": 0.7709840536117554, + "learning_rate": 6.326929013238934e-05, + "loss": 0.7275, + "step": 13934 + }, + { + "epoch": 2.4809472934472936, + "grad_norm": 0.755933940410614, + "learning_rate": 6.325627144727177e-05, + "loss": 0.7785, + "step": 13935 + }, + { + "epoch": 2.4811253561253563, + "grad_norm": 0.9058536291122437, + "learning_rate": 6.324325348207125e-05, + "loss": 0.9694, + "step": 13936 + }, + { + "epoch": 2.4813034188034186, + "grad_norm": 0.8490056395530701, + "learning_rate": 6.323023623704282e-05, + "loss": 0.8882, + "step": 13937 + }, + { + "epoch": 2.4814814814814814, + "grad_norm": 0.9559429883956909, + "learning_rate": 6.321721971244155e-05, + "loss": 0.9243, + "step": 13938 + }, + { + "epoch": 2.481659544159544, + "grad_norm": 0.8607096076011658, + "learning_rate": 6.320420390852242e-05, + "loss": 0.9281, + "step": 13939 + }, + { + "epoch": 2.481837606837607, + "grad_norm": 1.1263439655303955, + "learning_rate": 6.319118882554049e-05, + "loss": 0.8772, + "step": 13940 + }, + { + "epoch": 2.4820156695156697, + "grad_norm": 0.9691354632377625, + "learning_rate": 6.317817446375074e-05, + "loss": 0.9349, + "step": 13941 + }, + { + "epoch": 2.482193732193732, + "grad_norm": 0.8636828064918518, + "learning_rate": 6.31651608234082e-05, + "loss": 0.9021, + "step": 13942 + }, + { + "epoch": 2.4823717948717947, + "grad_norm": 0.8405864238739014, + "learning_rate": 6.315214790476777e-05, + "loss": 0.8917, + "step": 13943 + }, + { + "epoch": 2.4825498575498575, + "grad_norm": 0.8082821369171143, + "learning_rate": 6.313913570808448e-05, + "loss": 0.914, + "step": 13944 + }, + { + "epoch": 2.48272792022792, + "grad_norm": 0.8734335899353027, + "learning_rate": 6.312612423361328e-05, + "loss": 0.6885, + "step": 13945 + }, + { + "epoch": 2.482905982905983, + "grad_norm": 0.943190336227417, + "learning_rate": 6.311311348160904e-05, + "loss": 0.894, + "step": 13946 + }, + { + "epoch": 2.4830840455840457, + "grad_norm": 0.9425446391105652, + "learning_rate": 6.310010345232673e-05, + "loss": 0.7893, + "step": 13947 + }, + { + "epoch": 2.4832621082621085, + "grad_norm": 0.9668664932250977, + "learning_rate": 6.308709414602123e-05, + "loss": 0.9715, + "step": 13948 + }, + { + "epoch": 2.4834401709401708, + "grad_norm": 0.7771579027175903, + "learning_rate": 6.307408556294747e-05, + "loss": 0.872, + "step": 13949 + }, + { + "epoch": 2.4836182336182335, + "grad_norm": 0.8183084726333618, + "learning_rate": 6.306107770336025e-05, + "loss": 1.0008, + "step": 13950 + }, + { + "epoch": 2.4837962962962963, + "grad_norm": 0.8284399509429932, + "learning_rate": 6.304807056751452e-05, + "loss": 0.687, + "step": 13951 + }, + { + "epoch": 2.483974358974359, + "grad_norm": 0.8663082718849182, + "learning_rate": 6.303506415566504e-05, + "loss": 0.8792, + "step": 13952 + }, + { + "epoch": 2.484152421652422, + "grad_norm": 0.8272001147270203, + "learning_rate": 6.302205846806675e-05, + "loss": 0.9255, + "step": 13953 + }, + { + "epoch": 2.484330484330484, + "grad_norm": 0.9398671388626099, + "learning_rate": 6.300905350497437e-05, + "loss": 0.9605, + "step": 13954 + }, + { + "epoch": 2.484508547008547, + "grad_norm": 0.7368931770324707, + "learning_rate": 6.299604926664276e-05, + "loss": 0.6734, + "step": 13955 + }, + { + "epoch": 2.4846866096866096, + "grad_norm": 0.8315541744232178, + "learning_rate": 6.298304575332668e-05, + "loss": 0.9578, + "step": 13956 + }, + { + "epoch": 2.4848646723646723, + "grad_norm": 0.8656954169273376, + "learning_rate": 6.297004296528095e-05, + "loss": 0.8897, + "step": 13957 + }, + { + "epoch": 2.485042735042735, + "grad_norm": 0.9047118425369263, + "learning_rate": 6.295704090276026e-05, + "loss": 1.0308, + "step": 13958 + }, + { + "epoch": 2.485220797720798, + "grad_norm": 0.8771422505378723, + "learning_rate": 6.294403956601946e-05, + "loss": 0.9505, + "step": 13959 + }, + { + "epoch": 2.4853988603988606, + "grad_norm": 0.8935427069664001, + "learning_rate": 6.293103895531319e-05, + "loss": 1.0485, + "step": 13960 + }, + { + "epoch": 2.485576923076923, + "grad_norm": 0.7345624566078186, + "learning_rate": 6.291803907089621e-05, + "loss": 0.7485, + "step": 13961 + }, + { + "epoch": 2.4857549857549857, + "grad_norm": 0.6415224075317383, + "learning_rate": 6.290503991302324e-05, + "loss": 0.6237, + "step": 13962 + }, + { + "epoch": 2.4859330484330484, + "grad_norm": 0.8547754883766174, + "learning_rate": 6.289204148194896e-05, + "loss": 0.9152, + "step": 13963 + }, + { + "epoch": 2.486111111111111, + "grad_norm": 0.7434722185134888, + "learning_rate": 6.2879043777928e-05, + "loss": 0.8499, + "step": 13964 + }, + { + "epoch": 2.486289173789174, + "grad_norm": 0.8609980940818787, + "learning_rate": 6.286604680121509e-05, + "loss": 0.8101, + "step": 13965 + }, + { + "epoch": 2.486467236467236, + "grad_norm": 0.8709290027618408, + "learning_rate": 6.285305055206486e-05, + "loss": 0.8335, + "step": 13966 + }, + { + "epoch": 2.486645299145299, + "grad_norm": 0.7758293747901917, + "learning_rate": 6.284005503073191e-05, + "loss": 0.8408, + "step": 13967 + }, + { + "epoch": 2.4868233618233617, + "grad_norm": 0.9778353571891785, + "learning_rate": 6.282706023747094e-05, + "loss": 0.8595, + "step": 13968 + }, + { + "epoch": 2.4870014245014245, + "grad_norm": 0.8438369631767273, + "learning_rate": 6.281406617253646e-05, + "loss": 0.8843, + "step": 13969 + }, + { + "epoch": 2.4871794871794872, + "grad_norm": 0.9704681634902954, + "learning_rate": 6.280107283618315e-05, + "loss": 0.62, + "step": 13970 + }, + { + "epoch": 2.48735754985755, + "grad_norm": 0.7774441242218018, + "learning_rate": 6.278808022866549e-05, + "loss": 0.9087, + "step": 13971 + }, + { + "epoch": 2.4875356125356127, + "grad_norm": 0.8387142419815063, + "learning_rate": 6.277508835023813e-05, + "loss": 0.854, + "step": 13972 + }, + { + "epoch": 2.487713675213675, + "grad_norm": 0.8483029007911682, + "learning_rate": 6.276209720115556e-05, + "loss": 0.8665, + "step": 13973 + }, + { + "epoch": 2.487891737891738, + "grad_norm": 0.8251432180404663, + "learning_rate": 6.274910678167239e-05, + "loss": 0.7816, + "step": 13974 + }, + { + "epoch": 2.4880698005698005, + "grad_norm": 0.8503836989402771, + "learning_rate": 6.273611709204304e-05, + "loss": 0.939, + "step": 13975 + }, + { + "epoch": 2.4882478632478633, + "grad_norm": 0.6545158624649048, + "learning_rate": 6.27231281325221e-05, + "loss": 0.564, + "step": 13976 + }, + { + "epoch": 2.488425925925926, + "grad_norm": 0.7353499531745911, + "learning_rate": 6.2710139903364e-05, + "loss": 0.7103, + "step": 13977 + }, + { + "epoch": 2.488603988603989, + "grad_norm": 0.9032405614852905, + "learning_rate": 6.269715240482327e-05, + "loss": 0.8802, + "step": 13978 + }, + { + "epoch": 2.488782051282051, + "grad_norm": 0.8141019940376282, + "learning_rate": 6.268416563715434e-05, + "loss": 0.9836, + "step": 13979 + }, + { + "epoch": 2.488960113960114, + "grad_norm": 0.9087637066841125, + "learning_rate": 6.267117960061167e-05, + "loss": 0.8488, + "step": 13980 + }, + { + "epoch": 2.4891381766381766, + "grad_norm": 0.6649556756019592, + "learning_rate": 6.265819429544969e-05, + "loss": 0.604, + "step": 13981 + }, + { + "epoch": 2.4893162393162394, + "grad_norm": 0.7872918248176575, + "learning_rate": 6.264520972192283e-05, + "loss": 0.8537, + "step": 13982 + }, + { + "epoch": 2.489494301994302, + "grad_norm": 0.8821072578430176, + "learning_rate": 6.263222588028546e-05, + "loss": 0.7485, + "step": 13983 + }, + { + "epoch": 2.489672364672365, + "grad_norm": 0.8077933192253113, + "learning_rate": 6.2619242770792e-05, + "loss": 0.9174, + "step": 13984 + }, + { + "epoch": 2.489850427350427, + "grad_norm": 0.8087183237075806, + "learning_rate": 6.260626039369686e-05, + "loss": 0.9184, + "step": 13985 + }, + { + "epoch": 2.49002849002849, + "grad_norm": 0.8849205374717712, + "learning_rate": 6.259327874925434e-05, + "loss": 0.9095, + "step": 13986 + }, + { + "epoch": 2.4902065527065527, + "grad_norm": 0.7899976968765259, + "learning_rate": 6.258029783771884e-05, + "loss": 0.7286, + "step": 13987 + }, + { + "epoch": 2.4903846153846154, + "grad_norm": 0.7907543182373047, + "learning_rate": 6.256731765934464e-05, + "loss": 0.8433, + "step": 13988 + }, + { + "epoch": 2.490562678062678, + "grad_norm": 0.7794694304466248, + "learning_rate": 6.255433821438614e-05, + "loss": 0.7197, + "step": 13989 + }, + { + "epoch": 2.490740740740741, + "grad_norm": 0.8443161249160767, + "learning_rate": 6.254135950309753e-05, + "loss": 0.7982, + "step": 13990 + }, + { + "epoch": 2.4909188034188032, + "grad_norm": 0.972024142742157, + "learning_rate": 6.252838152573323e-05, + "loss": 0.9212, + "step": 13991 + }, + { + "epoch": 2.491096866096866, + "grad_norm": 0.8017764687538147, + "learning_rate": 6.25154042825474e-05, + "loss": 0.9544, + "step": 13992 + }, + { + "epoch": 2.4912749287749287, + "grad_norm": 0.8622884154319763, + "learning_rate": 6.250242777379442e-05, + "loss": 0.8411, + "step": 13993 + }, + { + "epoch": 2.4914529914529915, + "grad_norm": 0.7384446263313293, + "learning_rate": 6.248945199972842e-05, + "loss": 0.8357, + "step": 13994 + }, + { + "epoch": 2.4916310541310542, + "grad_norm": 0.7748960256576538, + "learning_rate": 6.247647696060372e-05, + "loss": 0.7739, + "step": 13995 + }, + { + "epoch": 2.491809116809117, + "grad_norm": 0.8295742273330688, + "learning_rate": 6.246350265667448e-05, + "loss": 0.8032, + "step": 13996 + }, + { + "epoch": 2.4919871794871793, + "grad_norm": 0.8604934811592102, + "learning_rate": 6.245052908819494e-05, + "loss": 0.8738, + "step": 13997 + }, + { + "epoch": 2.492165242165242, + "grad_norm": 0.8381406664848328, + "learning_rate": 6.243755625541926e-05, + "loss": 0.8351, + "step": 13998 + }, + { + "epoch": 2.492343304843305, + "grad_norm": 0.9238134026527405, + "learning_rate": 6.242458415860168e-05, + "loss": 0.9529, + "step": 13999 + }, + { + "epoch": 2.4925213675213675, + "grad_norm": 0.9234444499015808, + "learning_rate": 6.241161279799628e-05, + "loss": 1.1086, + "step": 14000 + }, + { + "epoch": 2.4926994301994303, + "grad_norm": 0.8056737780570984, + "learning_rate": 6.239864217385727e-05, + "loss": 0.7957, + "step": 14001 + }, + { + "epoch": 2.492877492877493, + "grad_norm": 0.7877696752548218, + "learning_rate": 6.238567228643872e-05, + "loss": 0.9577, + "step": 14002 + }, + { + "epoch": 2.4930555555555554, + "grad_norm": 0.7437340021133423, + "learning_rate": 6.237270313599479e-05, + "loss": 0.6171, + "step": 14003 + }, + { + "epoch": 2.493233618233618, + "grad_norm": 0.8503403067588806, + "learning_rate": 6.235973472277962e-05, + "loss": 0.7608, + "step": 14004 + }, + { + "epoch": 2.493411680911681, + "grad_norm": 0.8557562232017517, + "learning_rate": 6.234676704704722e-05, + "loss": 0.8414, + "step": 14005 + }, + { + "epoch": 2.4935897435897436, + "grad_norm": 0.9188289046287537, + "learning_rate": 6.233380010905174e-05, + "loss": 0.89, + "step": 14006 + }, + { + "epoch": 2.4937678062678064, + "grad_norm": 0.9433556199073792, + "learning_rate": 6.232083390904716e-05, + "loss": 0.9316, + "step": 14007 + }, + { + "epoch": 2.493945868945869, + "grad_norm": 0.9278882145881653, + "learning_rate": 6.230786844728759e-05, + "loss": 0.9211, + "step": 14008 + }, + { + "epoch": 2.494123931623932, + "grad_norm": 0.8365640640258789, + "learning_rate": 6.229490372402702e-05, + "loss": 0.9578, + "step": 14009 + }, + { + "epoch": 2.494301994301994, + "grad_norm": 0.7987647652626038, + "learning_rate": 6.228193973951953e-05, + "loss": 0.9279, + "step": 14010 + }, + { + "epoch": 2.494480056980057, + "grad_norm": 0.7707502841949463, + "learning_rate": 6.226897649401902e-05, + "loss": 0.8879, + "step": 14011 + }, + { + "epoch": 2.4946581196581197, + "grad_norm": 0.8623191118240356, + "learning_rate": 6.225601398777957e-05, + "loss": 0.7427, + "step": 14012 + }, + { + "epoch": 2.4948361823361824, + "grad_norm": 0.8470782041549683, + "learning_rate": 6.22430522210551e-05, + "loss": 0.9425, + "step": 14013 + }, + { + "epoch": 2.495014245014245, + "grad_norm": 0.9169524908065796, + "learning_rate": 6.223009119409963e-05, + "loss": 0.9595, + "step": 14014 + }, + { + "epoch": 2.4951923076923075, + "grad_norm": 0.8541738986968994, + "learning_rate": 6.221713090716701e-05, + "loss": 1.0726, + "step": 14015 + }, + { + "epoch": 2.4953703703703702, + "grad_norm": 0.8801444172859192, + "learning_rate": 6.220417136051126e-05, + "loss": 0.8323, + "step": 14016 + }, + { + "epoch": 2.495548433048433, + "grad_norm": 0.845448911190033, + "learning_rate": 6.219121255438624e-05, + "loss": 0.835, + "step": 14017 + }, + { + "epoch": 2.4957264957264957, + "grad_norm": 0.7653858661651611, + "learning_rate": 6.217825448904588e-05, + "loss": 0.7027, + "step": 14018 + }, + { + "epoch": 2.4959045584045585, + "grad_norm": 0.7779282927513123, + "learning_rate": 6.216529716474404e-05, + "loss": 0.7881, + "step": 14019 + }, + { + "epoch": 2.4960826210826212, + "grad_norm": 0.8739959597587585, + "learning_rate": 6.215234058173465e-05, + "loss": 0.9738, + "step": 14020 + }, + { + "epoch": 2.496260683760684, + "grad_norm": 0.8388087749481201, + "learning_rate": 6.213938474027148e-05, + "loss": 1.0128, + "step": 14021 + }, + { + "epoch": 2.4964387464387463, + "grad_norm": 0.8963341116905212, + "learning_rate": 6.212642964060843e-05, + "loss": 0.9669, + "step": 14022 + }, + { + "epoch": 2.496616809116809, + "grad_norm": 0.8959031701087952, + "learning_rate": 6.211347528299928e-05, + "loss": 0.9558, + "step": 14023 + }, + { + "epoch": 2.496794871794872, + "grad_norm": 0.8463472127914429, + "learning_rate": 6.210052166769791e-05, + "loss": 0.9835, + "step": 14024 + }, + { + "epoch": 2.4969729344729346, + "grad_norm": 0.7827564477920532, + "learning_rate": 6.208756879495812e-05, + "loss": 0.8411, + "step": 14025 + }, + { + "epoch": 2.4971509971509973, + "grad_norm": 0.8851028084754944, + "learning_rate": 6.207461666503363e-05, + "loss": 1.0409, + "step": 14026 + }, + { + "epoch": 2.4973290598290596, + "grad_norm": 1.0151652097702026, + "learning_rate": 6.206166527817825e-05, + "loss": 0.6671, + "step": 14027 + }, + { + "epoch": 2.4975071225071224, + "grad_norm": 0.7924346923828125, + "learning_rate": 6.204871463464572e-05, + "loss": 0.9971, + "step": 14028 + }, + { + "epoch": 2.497685185185185, + "grad_norm": 0.8524144887924194, + "learning_rate": 6.203576473468981e-05, + "loss": 0.9228, + "step": 14029 + }, + { + "epoch": 2.497863247863248, + "grad_norm": 0.7936401963233948, + "learning_rate": 6.20228155785642e-05, + "loss": 0.8128, + "step": 14030 + }, + { + "epoch": 2.4980413105413106, + "grad_norm": 1.0074050426483154, + "learning_rate": 6.200986716652267e-05, + "loss": 0.7846, + "step": 14031 + }, + { + "epoch": 2.4982193732193734, + "grad_norm": 0.7972239851951599, + "learning_rate": 6.199691949881882e-05, + "loss": 0.8689, + "step": 14032 + }, + { + "epoch": 2.498397435897436, + "grad_norm": 0.8810364007949829, + "learning_rate": 6.198397257570643e-05, + "loss": 0.7775, + "step": 14033 + }, + { + "epoch": 2.4985754985754984, + "grad_norm": 0.8819566965103149, + "learning_rate": 6.19710263974391e-05, + "loss": 0.8852, + "step": 14034 + }, + { + "epoch": 2.498753561253561, + "grad_norm": 0.8020595908164978, + "learning_rate": 6.195808096427054e-05, + "loss": 0.9691, + "step": 14035 + }, + { + "epoch": 2.498931623931624, + "grad_norm": 0.83958899974823, + "learning_rate": 6.194513627645433e-05, + "loss": 0.8072, + "step": 14036 + }, + { + "epoch": 2.4991096866096867, + "grad_norm": 0.7525333166122437, + "learning_rate": 6.193219233424414e-05, + "loss": 0.7839, + "step": 14037 + }, + { + "epoch": 2.4992877492877494, + "grad_norm": 0.8687964677810669, + "learning_rate": 6.191924913789353e-05, + "loss": 0.9512, + "step": 14038 + }, + { + "epoch": 2.4994658119658117, + "grad_norm": 0.9080697298049927, + "learning_rate": 6.190630668765617e-05, + "loss": 0.8635, + "step": 14039 + }, + { + "epoch": 2.4996438746438745, + "grad_norm": 0.8174137473106384, + "learning_rate": 6.189336498378557e-05, + "loss": 0.9034, + "step": 14040 + }, + { + "epoch": 2.4996438746438745, + "eval_loss": 1.1338438987731934, + "eval_runtime": 24.4013, + "eval_samples_per_second": 42.662, + "eval_steps_per_second": 21.351, + "step": 14040 + }, + { + "epoch": 2.4998219373219372, + "grad_norm": 0.9711320996284485, + "learning_rate": 6.188042402653536e-05, + "loss": 0.9892, + "step": 14041 + }, + { + "epoch": 2.5, + "grad_norm": 0.8726856112480164, + "learning_rate": 6.1867483816159e-05, + "loss": 0.7482, + "step": 14042 + }, + { + "epoch": 2.5001780626780628, + "grad_norm": 0.875801682472229, + "learning_rate": 6.18545443529101e-05, + "loss": 0.7599, + "step": 14043 + }, + { + "epoch": 2.5003561253561255, + "grad_norm": 0.8867987990379333, + "learning_rate": 6.184160563704218e-05, + "loss": 0.805, + "step": 14044 + }, + { + "epoch": 2.5005341880341883, + "grad_norm": 0.8766322135925293, + "learning_rate": 6.18286676688087e-05, + "loss": 0.7343, + "step": 14045 + }, + { + "epoch": 2.5007122507122506, + "grad_norm": 0.8096646070480347, + "learning_rate": 6.181573044846323e-05, + "loss": 0.7957, + "step": 14046 + }, + { + "epoch": 2.5008903133903133, + "grad_norm": 1.0121821165084839, + "learning_rate": 6.180279397625917e-05, + "loss": 0.8775, + "step": 14047 + }, + { + "epoch": 2.501068376068376, + "grad_norm": 0.79291170835495, + "learning_rate": 6.178985825245003e-05, + "loss": 0.91, + "step": 14048 + }, + { + "epoch": 2.501246438746439, + "grad_norm": 0.83204185962677, + "learning_rate": 6.177692327728922e-05, + "loss": 0.799, + "step": 14049 + }, + { + "epoch": 2.5014245014245016, + "grad_norm": 0.8746328949928284, + "learning_rate": 6.176398905103023e-05, + "loss": 0.9595, + "step": 14050 + }, + { + "epoch": 2.501602564102564, + "grad_norm": 0.7665601968765259, + "learning_rate": 6.17510555739264e-05, + "loss": 0.7935, + "step": 14051 + }, + { + "epoch": 2.5017806267806266, + "grad_norm": 0.8761195540428162, + "learning_rate": 6.173812284623122e-05, + "loss": 1.017, + "step": 14052 + }, + { + "epoch": 2.5019586894586894, + "grad_norm": 0.8847656846046448, + "learning_rate": 6.172519086819802e-05, + "loss": 0.7684, + "step": 14053 + }, + { + "epoch": 2.502136752136752, + "grad_norm": 0.8320107460021973, + "learning_rate": 6.171225964008021e-05, + "loss": 0.8828, + "step": 14054 + }, + { + "epoch": 2.502314814814815, + "grad_norm": 1.0184354782104492, + "learning_rate": 6.169932916213111e-05, + "loss": 0.887, + "step": 14055 + }, + { + "epoch": 2.5024928774928776, + "grad_norm": 0.7870062589645386, + "learning_rate": 6.168639943460415e-05, + "loss": 0.8595, + "step": 14056 + }, + { + "epoch": 2.5026709401709404, + "grad_norm": 0.8314430117607117, + "learning_rate": 6.167347045775254e-05, + "loss": 0.7822, + "step": 14057 + }, + { + "epoch": 2.5028490028490027, + "grad_norm": 0.7521854043006897, + "learning_rate": 6.166054223182968e-05, + "loss": 0.916, + "step": 14058 + }, + { + "epoch": 2.5030270655270654, + "grad_norm": 0.8240202069282532, + "learning_rate": 6.164761475708885e-05, + "loss": 0.9404, + "step": 14059 + }, + { + "epoch": 2.503205128205128, + "grad_norm": 0.8467113375663757, + "learning_rate": 6.163468803378338e-05, + "loss": 0.7554, + "step": 14060 + }, + { + "epoch": 2.503383190883191, + "grad_norm": 0.7639012336730957, + "learning_rate": 6.162176206216645e-05, + "loss": 0.8141, + "step": 14061 + }, + { + "epoch": 2.5035612535612537, + "grad_norm": 0.8578195571899414, + "learning_rate": 6.160883684249138e-05, + "loss": 0.821, + "step": 14062 + }, + { + "epoch": 2.503739316239316, + "grad_norm": 0.7478210926055908, + "learning_rate": 6.159591237501139e-05, + "loss": 0.5878, + "step": 14063 + }, + { + "epoch": 2.5039173789173788, + "grad_norm": 0.7936450242996216, + "learning_rate": 6.158298865997972e-05, + "loss": 0.9616, + "step": 14064 + }, + { + "epoch": 2.5040954415954415, + "grad_norm": 0.9196288585662842, + "learning_rate": 6.157006569764963e-05, + "loss": 0.8147, + "step": 14065 + }, + { + "epoch": 2.5042735042735043, + "grad_norm": 1.0488382577896118, + "learning_rate": 6.155714348827422e-05, + "loss": 0.7941, + "step": 14066 + }, + { + "epoch": 2.504451566951567, + "grad_norm": 0.9195658564567566, + "learning_rate": 6.154422203210676e-05, + "loss": 1.0186, + "step": 14067 + }, + { + "epoch": 2.5046296296296298, + "grad_norm": 0.9088640213012695, + "learning_rate": 6.153130132940037e-05, + "loss": 0.7611, + "step": 14068 + }, + { + "epoch": 2.5048076923076925, + "grad_norm": 0.8168773651123047, + "learning_rate": 6.151838138040821e-05, + "loss": 0.8466, + "step": 14069 + }, + { + "epoch": 2.504985754985755, + "grad_norm": 0.9976982474327087, + "learning_rate": 6.150546218538342e-05, + "loss": 0.9438, + "step": 14070 + }, + { + "epoch": 2.5051638176638176, + "grad_norm": 0.9469537138938904, + "learning_rate": 6.149254374457917e-05, + "loss": 0.9485, + "step": 14071 + }, + { + "epoch": 2.5053418803418803, + "grad_norm": 0.8861194849014282, + "learning_rate": 6.147962605824851e-05, + "loss": 0.9676, + "step": 14072 + }, + { + "epoch": 2.505519943019943, + "grad_norm": 0.9008424878120422, + "learning_rate": 6.146670912664457e-05, + "loss": 0.7343, + "step": 14073 + }, + { + "epoch": 2.505698005698006, + "grad_norm": 0.8957796096801758, + "learning_rate": 6.145379295002038e-05, + "loss": 0.8538, + "step": 14074 + }, + { + "epoch": 2.505876068376068, + "grad_norm": 0.8739160895347595, + "learning_rate": 6.14408775286291e-05, + "loss": 0.9029, + "step": 14075 + }, + { + "epoch": 2.506054131054131, + "grad_norm": 0.7713274955749512, + "learning_rate": 6.142796286272368e-05, + "loss": 0.6962, + "step": 14076 + }, + { + "epoch": 2.5062321937321936, + "grad_norm": 0.8545170426368713, + "learning_rate": 6.141504895255725e-05, + "loss": 0.9208, + "step": 14077 + }, + { + "epoch": 2.5064102564102564, + "grad_norm": 0.8102772235870361, + "learning_rate": 6.140213579838274e-05, + "loss": 0.7785, + "step": 14078 + }, + { + "epoch": 2.506588319088319, + "grad_norm": 0.8055099844932556, + "learning_rate": 6.138922340045321e-05, + "loss": 0.8502, + "step": 14079 + }, + { + "epoch": 2.506766381766382, + "grad_norm": 0.8132893443107605, + "learning_rate": 6.137631175902164e-05, + "loss": 0.7559, + "step": 14080 + }, + { + "epoch": 2.5069444444444446, + "grad_norm": 0.8608863949775696, + "learning_rate": 6.136340087434102e-05, + "loss": 0.838, + "step": 14081 + }, + { + "epoch": 2.5071225071225074, + "grad_norm": 0.8480643630027771, + "learning_rate": 6.135049074666428e-05, + "loss": 0.9062, + "step": 14082 + }, + { + "epoch": 2.5073005698005697, + "grad_norm": 0.7107672691345215, + "learning_rate": 6.133758137624437e-05, + "loss": 0.7494, + "step": 14083 + }, + { + "epoch": 2.5074786324786325, + "grad_norm": 0.812416672706604, + "learning_rate": 6.132467276333427e-05, + "loss": 0.6428, + "step": 14084 + }, + { + "epoch": 2.507656695156695, + "grad_norm": 0.8304431438446045, + "learning_rate": 6.131176490818684e-05, + "loss": 0.9931, + "step": 14085 + }, + { + "epoch": 2.507834757834758, + "grad_norm": 0.8344886302947998, + "learning_rate": 6.129885781105507e-05, + "loss": 0.921, + "step": 14086 + }, + { + "epoch": 2.5080128205128203, + "grad_norm": 0.8137457966804504, + "learning_rate": 6.128595147219172e-05, + "loss": 0.9113, + "step": 14087 + }, + { + "epoch": 2.508190883190883, + "grad_norm": 0.7404686212539673, + "learning_rate": 6.127304589184976e-05, + "loss": 0.7625, + "step": 14088 + }, + { + "epoch": 2.5083689458689458, + "grad_norm": 0.8179733157157898, + "learning_rate": 6.126014107028202e-05, + "loss": 0.9049, + "step": 14089 + }, + { + "epoch": 2.5085470085470085, + "grad_norm": 0.7788520455360413, + "learning_rate": 6.124723700774133e-05, + "loss": 0.7391, + "step": 14090 + }, + { + "epoch": 2.5087250712250713, + "grad_norm": 0.8127198219299316, + "learning_rate": 6.123433370448052e-05, + "loss": 0.8551, + "step": 14091 + }, + { + "epoch": 2.508903133903134, + "grad_norm": 0.8134245276451111, + "learning_rate": 6.122143116075245e-05, + "loss": 0.7422, + "step": 14092 + }, + { + "epoch": 2.5090811965811968, + "grad_norm": 0.9117823243141174, + "learning_rate": 6.120852937680983e-05, + "loss": 0.8649, + "step": 14093 + }, + { + "epoch": 2.5092592592592595, + "grad_norm": 0.8417702913284302, + "learning_rate": 6.119562835290553e-05, + "loss": 0.6902, + "step": 14094 + }, + { + "epoch": 2.509437321937322, + "grad_norm": 0.8655431866645813, + "learning_rate": 6.118272808929225e-05, + "loss": 1.0778, + "step": 14095 + }, + { + "epoch": 2.5096153846153846, + "grad_norm": 0.9228867888450623, + "learning_rate": 6.116982858622282e-05, + "loss": 0.9991, + "step": 14096 + }, + { + "epoch": 2.5097934472934473, + "grad_norm": 0.819505512714386, + "learning_rate": 6.115692984394992e-05, + "loss": 1.0351, + "step": 14097 + }, + { + "epoch": 2.50997150997151, + "grad_norm": 0.8419737219810486, + "learning_rate": 6.114403186272628e-05, + "loss": 0.92, + "step": 14098 + }, + { + "epoch": 2.5101495726495724, + "grad_norm": 0.8294256329536438, + "learning_rate": 6.11311346428046e-05, + "loss": 0.792, + "step": 14099 + }, + { + "epoch": 2.510327635327635, + "grad_norm": 0.8549113273620605, + "learning_rate": 6.111823818443765e-05, + "loss": 0.9453, + "step": 14100 + }, + { + "epoch": 2.510505698005698, + "grad_norm": 1.0194092988967896, + "learning_rate": 6.1105342487878e-05, + "loss": 0.9747, + "step": 14101 + }, + { + "epoch": 2.5106837606837606, + "grad_norm": 0.7633654475212097, + "learning_rate": 6.109244755337842e-05, + "loss": 0.8227, + "step": 14102 + }, + { + "epoch": 2.5108618233618234, + "grad_norm": 0.9679104685783386, + "learning_rate": 6.107955338119147e-05, + "loss": 1.0407, + "step": 14103 + }, + { + "epoch": 2.511039886039886, + "grad_norm": 0.8342793583869934, + "learning_rate": 6.10666599715698e-05, + "loss": 0.7868, + "step": 14104 + }, + { + "epoch": 2.511217948717949, + "grad_norm": 0.9264410734176636, + "learning_rate": 6.105376732476609e-05, + "loss": 0.802, + "step": 14105 + }, + { + "epoch": 2.5113960113960117, + "grad_norm": 0.7511885762214661, + "learning_rate": 6.104087544103287e-05, + "loss": 0.7561, + "step": 14106 + }, + { + "epoch": 2.511574074074074, + "grad_norm": 0.8330591320991516, + "learning_rate": 6.102798432062282e-05, + "loss": 0.8511, + "step": 14107 + }, + { + "epoch": 2.5117521367521367, + "grad_norm": 0.8971241116523743, + "learning_rate": 6.1015093963788415e-05, + "loss": 0.8854, + "step": 14108 + }, + { + "epoch": 2.5119301994301995, + "grad_norm": 0.7926762700080872, + "learning_rate": 6.100220437078228e-05, + "loss": 0.6803, + "step": 14109 + }, + { + "epoch": 2.512108262108262, + "grad_norm": 0.7384431958198547, + "learning_rate": 6.098931554185692e-05, + "loss": 0.6162, + "step": 14110 + }, + { + "epoch": 2.5122863247863245, + "grad_norm": 0.9201281070709229, + "learning_rate": 6.097642747726491e-05, + "loss": 1.1095, + "step": 14111 + }, + { + "epoch": 2.5124643874643873, + "grad_norm": 0.8822020888328552, + "learning_rate": 6.0963540177258716e-05, + "loss": 0.87, + "step": 14112 + }, + { + "epoch": 2.51264245014245, + "grad_norm": 0.8243268728256226, + "learning_rate": 6.09506536420909e-05, + "loss": 1.0899, + "step": 14113 + }, + { + "epoch": 2.5128205128205128, + "grad_norm": 0.8657538294792175, + "learning_rate": 6.093776787201386e-05, + "loss": 0.8218, + "step": 14114 + }, + { + "epoch": 2.5129985754985755, + "grad_norm": 0.8651030659675598, + "learning_rate": 6.092488286728013e-05, + "loss": 0.7903, + "step": 14115 + }, + { + "epoch": 2.5131766381766383, + "grad_norm": 0.8341799378395081, + "learning_rate": 6.091199862814214e-05, + "loss": 0.8612, + "step": 14116 + }, + { + "epoch": 2.513354700854701, + "grad_norm": 0.7693229913711548, + "learning_rate": 6.0899115154852384e-05, + "loss": 0.819, + "step": 14117 + }, + { + "epoch": 2.513532763532764, + "grad_norm": 0.8883055448532104, + "learning_rate": 6.088623244766318e-05, + "loss": 0.9026, + "step": 14118 + }, + { + "epoch": 2.513710826210826, + "grad_norm": 0.7761621475219727, + "learning_rate": 6.087335050682703e-05, + "loss": 0.7505, + "step": 14119 + }, + { + "epoch": 2.513888888888889, + "grad_norm": 0.8152571320533752, + "learning_rate": 6.086046933259628e-05, + "loss": 0.7637, + "step": 14120 + }, + { + "epoch": 2.5140669515669516, + "grad_norm": 0.7990148663520813, + "learning_rate": 6.0847588925223376e-05, + "loss": 0.8615, + "step": 14121 + }, + { + "epoch": 2.5142450142450143, + "grad_norm": 0.844756007194519, + "learning_rate": 6.083470928496058e-05, + "loss": 0.8696, + "step": 14122 + }, + { + "epoch": 2.5144230769230766, + "grad_norm": 0.7533631324768066, + "learning_rate": 6.082183041206031e-05, + "loss": 0.7674, + "step": 14123 + }, + { + "epoch": 2.5146011396011394, + "grad_norm": 0.7914009690284729, + "learning_rate": 6.0808952306774905e-05, + "loss": 0.7577, + "step": 14124 + }, + { + "epoch": 2.514779202279202, + "grad_norm": 0.8341572284698486, + "learning_rate": 6.079607496935666e-05, + "loss": 0.8899, + "step": 14125 + }, + { + "epoch": 2.514957264957265, + "grad_norm": 0.9185548424720764, + "learning_rate": 6.078319840005788e-05, + "loss": 0.9486, + "step": 14126 + }, + { + "epoch": 2.5151353276353277, + "grad_norm": 0.8611742854118347, + "learning_rate": 6.0770322599130856e-05, + "loss": 0.8267, + "step": 14127 + }, + { + "epoch": 2.5153133903133904, + "grad_norm": 0.899135410785675, + "learning_rate": 6.0757447566827906e-05, + "loss": 1.0829, + "step": 14128 + }, + { + "epoch": 2.515491452991453, + "grad_norm": 0.8016429543495178, + "learning_rate": 6.074457330340122e-05, + "loss": 0.8582, + "step": 14129 + }, + { + "epoch": 2.515669515669516, + "grad_norm": 0.7781331539154053, + "learning_rate": 6.073169980910307e-05, + "loss": 0.8435, + "step": 14130 + }, + { + "epoch": 2.515847578347578, + "grad_norm": 0.7605105042457581, + "learning_rate": 6.071882708418568e-05, + "loss": 0.6961, + "step": 14131 + }, + { + "epoch": 2.516025641025641, + "grad_norm": 0.9337655901908875, + "learning_rate": 6.0705955128901326e-05, + "loss": 0.8673, + "step": 14132 + }, + { + "epoch": 2.5162037037037037, + "grad_norm": 0.7868272662162781, + "learning_rate": 6.06930839435021e-05, + "loss": 0.7526, + "step": 14133 + }, + { + "epoch": 2.5163817663817665, + "grad_norm": 0.8722387552261353, + "learning_rate": 6.068021352824027e-05, + "loss": 0.9541, + "step": 14134 + }, + { + "epoch": 2.5165598290598292, + "grad_norm": 0.7682648301124573, + "learning_rate": 6.066734388336794e-05, + "loss": 0.7191, + "step": 14135 + }, + { + "epoch": 2.5167378917378915, + "grad_norm": 0.9540650844573975, + "learning_rate": 6.065447500913737e-05, + "loss": 1.0638, + "step": 14136 + }, + { + "epoch": 2.5169159544159543, + "grad_norm": 0.8276218175888062, + "learning_rate": 6.064160690580056e-05, + "loss": 0.7967, + "step": 14137 + }, + { + "epoch": 2.517094017094017, + "grad_norm": 0.7966098785400391, + "learning_rate": 6.062873957360976e-05, + "loss": 0.8913, + "step": 14138 + }, + { + "epoch": 2.51727207977208, + "grad_norm": 0.9670028686523438, + "learning_rate": 6.0615873012816974e-05, + "loss": 0.8846, + "step": 14139 + }, + { + "epoch": 2.5174501424501425, + "grad_norm": 0.819952666759491, + "learning_rate": 6.0603007223674366e-05, + "loss": 0.8409, + "step": 14140 + }, + { + "epoch": 2.5176282051282053, + "grad_norm": 0.7746681571006775, + "learning_rate": 6.0590142206433973e-05, + "loss": 0.7382, + "step": 14141 + }, + { + "epoch": 2.517806267806268, + "grad_norm": 0.9452744722366333, + "learning_rate": 6.057727796134787e-05, + "loss": 0.8878, + "step": 14142 + }, + { + "epoch": 2.5179843304843303, + "grad_norm": 0.7940170168876648, + "learning_rate": 6.0564414488668165e-05, + "loss": 0.8289, + "step": 14143 + }, + { + "epoch": 2.518162393162393, + "grad_norm": 0.9046176671981812, + "learning_rate": 6.0551551788646774e-05, + "loss": 0.8596, + "step": 14144 + }, + { + "epoch": 2.518340455840456, + "grad_norm": 0.8460658192634583, + "learning_rate": 6.053868986153581e-05, + "loss": 0.7678, + "step": 14145 + }, + { + "epoch": 2.5185185185185186, + "grad_norm": 0.9131760597229004, + "learning_rate": 6.052582870758723e-05, + "loss": 0.8845, + "step": 14146 + }, + { + "epoch": 2.5186965811965814, + "grad_norm": 0.8375167846679688, + "learning_rate": 6.0512968327053076e-05, + "loss": 1.0082, + "step": 14147 + }, + { + "epoch": 2.5188746438746437, + "grad_norm": 0.8587140440940857, + "learning_rate": 6.050010872018523e-05, + "loss": 0.8745, + "step": 14148 + }, + { + "epoch": 2.5190527065527064, + "grad_norm": 0.8347265124320984, + "learning_rate": 6.048724988723575e-05, + "loss": 0.8345, + "step": 14149 + }, + { + "epoch": 2.519230769230769, + "grad_norm": 1.0271183252334595, + "learning_rate": 6.047439182845649e-05, + "loss": 0.9863, + "step": 14150 + }, + { + "epoch": 2.519408831908832, + "grad_norm": 0.6951111555099487, + "learning_rate": 6.046153454409943e-05, + "loss": 0.6162, + "step": 14151 + }, + { + "epoch": 2.5195868945868947, + "grad_norm": 0.7702959179878235, + "learning_rate": 6.044867803441645e-05, + "loss": 0.8127, + "step": 14152 + }, + { + "epoch": 2.5197649572649574, + "grad_norm": 0.7997276186943054, + "learning_rate": 6.0435822299659496e-05, + "loss": 0.6777, + "step": 14153 + }, + { + "epoch": 2.51994301994302, + "grad_norm": 0.8006166815757751, + "learning_rate": 6.0422967340080385e-05, + "loss": 0.9122, + "step": 14154 + }, + { + "epoch": 2.5201210826210825, + "grad_norm": 0.888225793838501, + "learning_rate": 6.041011315593102e-05, + "loss": 0.7621, + "step": 14155 + }, + { + "epoch": 2.5202991452991452, + "grad_norm": 0.928814172744751, + "learning_rate": 6.039725974746324e-05, + "loss": 1.0245, + "step": 14156 + }, + { + "epoch": 2.520477207977208, + "grad_norm": 0.7914403676986694, + "learning_rate": 6.038440711492892e-05, + "loss": 0.6585, + "step": 14157 + }, + { + "epoch": 2.5206552706552707, + "grad_norm": 0.82389897108078, + "learning_rate": 6.0371555258579826e-05, + "loss": 0.7862, + "step": 14158 + }, + { + "epoch": 2.5208333333333335, + "grad_norm": 0.952135443687439, + "learning_rate": 6.035870417866778e-05, + "loss": 0.8952, + "step": 14159 + }, + { + "epoch": 2.521011396011396, + "grad_norm": 0.8626661896705627, + "learning_rate": 6.034585387544458e-05, + "loss": 0.9166, + "step": 14160 + }, + { + "epoch": 2.5211894586894585, + "grad_norm": 0.9641584157943726, + "learning_rate": 6.033300434916203e-05, + "loss": 0.8481, + "step": 14161 + }, + { + "epoch": 2.5213675213675213, + "grad_norm": 0.949110209941864, + "learning_rate": 6.0320155600071814e-05, + "loss": 0.9628, + "step": 14162 + }, + { + "epoch": 2.521545584045584, + "grad_norm": 0.8198522329330444, + "learning_rate": 6.030730762842573e-05, + "loss": 0.817, + "step": 14163 + }, + { + "epoch": 2.521723646723647, + "grad_norm": 0.9209866523742676, + "learning_rate": 6.029446043447553e-05, + "loss": 0.925, + "step": 14164 + }, + { + "epoch": 2.5219017094017095, + "grad_norm": 0.8604369163513184, + "learning_rate": 6.0281614018472854e-05, + "loss": 0.7846, + "step": 14165 + }, + { + "epoch": 2.5220797720797723, + "grad_norm": 0.882255494594574, + "learning_rate": 6.026876838066948e-05, + "loss": 0.8715, + "step": 14166 + }, + { + "epoch": 2.5222578347578346, + "grad_norm": 0.8609021306037903, + "learning_rate": 6.0255923521317015e-05, + "loss": 0.8627, + "step": 14167 + }, + { + "epoch": 2.5224358974358974, + "grad_norm": 0.9782202243804932, + "learning_rate": 6.0243079440667226e-05, + "loss": 0.8499, + "step": 14168 + }, + { + "epoch": 2.52261396011396, + "grad_norm": 0.7932701706886292, + "learning_rate": 6.023023613897165e-05, + "loss": 0.9174, + "step": 14169 + }, + { + "epoch": 2.522792022792023, + "grad_norm": 0.8827422261238098, + "learning_rate": 6.021739361648202e-05, + "loss": 0.9384, + "step": 14170 + }, + { + "epoch": 2.5229700854700856, + "grad_norm": 0.9764171838760376, + "learning_rate": 6.020455187344989e-05, + "loss": 0.8806, + "step": 14171 + }, + { + "epoch": 2.523148148148148, + "grad_norm": 0.7635362148284912, + "learning_rate": 6.019171091012694e-05, + "loss": 0.7519, + "step": 14172 + }, + { + "epoch": 2.5233262108262107, + "grad_norm": 0.9925556182861328, + "learning_rate": 6.017887072676468e-05, + "loss": 0.8467, + "step": 14173 + }, + { + "epoch": 2.5235042735042734, + "grad_norm": 0.9624950289726257, + "learning_rate": 6.016603132361477e-05, + "loss": 0.9492, + "step": 14174 + }, + { + "epoch": 2.523682336182336, + "grad_norm": 0.7960891127586365, + "learning_rate": 6.0153192700928685e-05, + "loss": 1.0111, + "step": 14175 + }, + { + "epoch": 2.523860398860399, + "grad_norm": 0.8387307524681091, + "learning_rate": 6.014035485895804e-05, + "loss": 0.8013, + "step": 14176 + }, + { + "epoch": 2.5240384615384617, + "grad_norm": 0.8488287925720215, + "learning_rate": 6.0127517797954316e-05, + "loss": 0.8508, + "step": 14177 + }, + { + "epoch": 2.5242165242165244, + "grad_norm": 0.7339358329772949, + "learning_rate": 6.011468151816908e-05, + "loss": 0.7225, + "step": 14178 + }, + { + "epoch": 2.5243945868945867, + "grad_norm": 0.9265308976173401, + "learning_rate": 6.010184601985378e-05, + "loss": 0.7993, + "step": 14179 + }, + { + "epoch": 2.5245726495726495, + "grad_norm": 0.7752045392990112, + "learning_rate": 6.0089011303259944e-05, + "loss": 0.8315, + "step": 14180 + }, + { + "epoch": 2.5247507122507122, + "grad_norm": 0.7794929146766663, + "learning_rate": 6.007617736863901e-05, + "loss": 0.9174, + "step": 14181 + }, + { + "epoch": 2.524928774928775, + "grad_norm": 0.9099361896514893, + "learning_rate": 6.0063344216242434e-05, + "loss": 0.8948, + "step": 14182 + }, + { + "epoch": 2.5251068376068377, + "grad_norm": 0.8161521553993225, + "learning_rate": 6.005051184632171e-05, + "loss": 0.8018, + "step": 14183 + }, + { + "epoch": 2.5252849002849, + "grad_norm": 0.9279208183288574, + "learning_rate": 6.003768025912819e-05, + "loss": 0.9032, + "step": 14184 + }, + { + "epoch": 2.525462962962963, + "grad_norm": 0.9689664840698242, + "learning_rate": 6.002484945491333e-05, + "loss": 0.9463, + "step": 14185 + }, + { + "epoch": 2.5256410256410255, + "grad_norm": 0.8367486596107483, + "learning_rate": 6.001201943392848e-05, + "loss": 0.7866, + "step": 14186 + }, + { + "epoch": 2.5258190883190883, + "grad_norm": 0.8383589386940002, + "learning_rate": 5.9999190196425056e-05, + "loss": 0.7642, + "step": 14187 + }, + { + "epoch": 2.525997150997151, + "grad_norm": 0.9113569855690002, + "learning_rate": 5.99863617426544e-05, + "loss": 1.0451, + "step": 14188 + }, + { + "epoch": 2.526175213675214, + "grad_norm": 0.896575391292572, + "learning_rate": 5.997353407286788e-05, + "loss": 0.8559, + "step": 14189 + }, + { + "epoch": 2.5263532763532766, + "grad_norm": 0.899214506149292, + "learning_rate": 5.996070718731679e-05, + "loss": 1.0006, + "step": 14190 + }, + { + "epoch": 2.5265313390313393, + "grad_norm": 0.9739418625831604, + "learning_rate": 5.994788108625247e-05, + "loss": 0.9412, + "step": 14191 + }, + { + "epoch": 2.5267094017094016, + "grad_norm": 0.7940781712532043, + "learning_rate": 5.9935055769926215e-05, + "loss": 0.8802, + "step": 14192 + }, + { + "epoch": 2.5268874643874644, + "grad_norm": 0.802066445350647, + "learning_rate": 5.9922231238589346e-05, + "loss": 0.8459, + "step": 14193 + }, + { + "epoch": 2.527065527065527, + "grad_norm": 0.827560544013977, + "learning_rate": 5.990940749249306e-05, + "loss": 0.8732, + "step": 14194 + }, + { + "epoch": 2.52724358974359, + "grad_norm": 0.7782348394393921, + "learning_rate": 5.989658453188869e-05, + "loss": 0.6742, + "step": 14195 + }, + { + "epoch": 2.527421652421652, + "grad_norm": 0.8418310284614563, + "learning_rate": 5.9883762357027416e-05, + "loss": 0.8699, + "step": 14196 + }, + { + "epoch": 2.527599715099715, + "grad_norm": 0.7925812602043152, + "learning_rate": 5.987094096816051e-05, + "loss": 0.6807, + "step": 14197 + }, + { + "epoch": 2.5277777777777777, + "grad_norm": 0.8200794458389282, + "learning_rate": 5.9858120365539105e-05, + "loss": 0.8249, + "step": 14198 + }, + { + "epoch": 2.5279558404558404, + "grad_norm": 0.7137587070465088, + "learning_rate": 5.9845300549414505e-05, + "loss": 0.7882, + "step": 14199 + }, + { + "epoch": 2.528133903133903, + "grad_norm": 0.8084787726402283, + "learning_rate": 5.983248152003778e-05, + "loss": 1.0161, + "step": 14200 + }, + { + "epoch": 2.528311965811966, + "grad_norm": 0.7717064023017883, + "learning_rate": 5.9819663277660156e-05, + "loss": 0.798, + "step": 14201 + }, + { + "epoch": 2.5284900284900287, + "grad_norm": 0.7722328305244446, + "learning_rate": 5.980684582253275e-05, + "loss": 0.8324, + "step": 14202 + }, + { + "epoch": 2.5286680911680914, + "grad_norm": 0.8357635140419006, + "learning_rate": 5.9794029154906696e-05, + "loss": 0.9224, + "step": 14203 + }, + { + "epoch": 2.5288461538461537, + "grad_norm": 0.8159863352775574, + "learning_rate": 5.978121327503317e-05, + "loss": 0.7529, + "step": 14204 + }, + { + "epoch": 2.5290242165242165, + "grad_norm": 0.8255389332771301, + "learning_rate": 5.976839818316317e-05, + "loss": 0.9674, + "step": 14205 + }, + { + "epoch": 2.5292022792022792, + "grad_norm": 0.8204228281974792, + "learning_rate": 5.975558387954787e-05, + "loss": 0.9138, + "step": 14206 + }, + { + "epoch": 2.529380341880342, + "grad_norm": 0.8232463598251343, + "learning_rate": 5.9742770364438275e-05, + "loss": 0.7949, + "step": 14207 + }, + { + "epoch": 2.5295584045584043, + "grad_norm": 0.8164107203483582, + "learning_rate": 5.972995763808551e-05, + "loss": 0.7087, + "step": 14208 + }, + { + "epoch": 2.529736467236467, + "grad_norm": 0.8100822567939758, + "learning_rate": 5.971714570074052e-05, + "loss": 0.9187, + "step": 14209 + }, + { + "epoch": 2.52991452991453, + "grad_norm": 0.7401103377342224, + "learning_rate": 5.970433455265443e-05, + "loss": 0.798, + "step": 14210 + }, + { + "epoch": 2.5300925925925926, + "grad_norm": 0.798327624797821, + "learning_rate": 5.9691524194078154e-05, + "loss": 0.7312, + "step": 14211 + }, + { + "epoch": 2.5302706552706553, + "grad_norm": 0.8566045165061951, + "learning_rate": 5.9678714625262754e-05, + "loss": 0.8555, + "step": 14212 + }, + { + "epoch": 2.530448717948718, + "grad_norm": 0.8005902767181396, + "learning_rate": 5.9665905846459155e-05, + "loss": 0.7979, + "step": 14213 + }, + { + "epoch": 2.530626780626781, + "grad_norm": 0.815990686416626, + "learning_rate": 5.9653097857918396e-05, + "loss": 0.8739, + "step": 14214 + }, + { + "epoch": 2.5308048433048436, + "grad_norm": 0.7694230079650879, + "learning_rate": 5.9640290659891316e-05, + "loss": 0.7249, + "step": 14215 + }, + { + "epoch": 2.530982905982906, + "grad_norm": 0.8469253182411194, + "learning_rate": 5.962748425262892e-05, + "loss": 0.8505, + "step": 14216 + }, + { + "epoch": 2.5311609686609686, + "grad_norm": 0.8061797022819519, + "learning_rate": 5.961467863638209e-05, + "loss": 0.8979, + "step": 14217 + }, + { + "epoch": 2.5313390313390314, + "grad_norm": 1.0380569696426392, + "learning_rate": 5.960187381140179e-05, + "loss": 0.8664, + "step": 14218 + }, + { + "epoch": 2.531517094017094, + "grad_norm": 0.9435166716575623, + "learning_rate": 5.9589069777938786e-05, + "loss": 0.7566, + "step": 14219 + }, + { + "epoch": 2.5316951566951564, + "grad_norm": 0.8882613182067871, + "learning_rate": 5.957626653624407e-05, + "loss": 0.6999, + "step": 14220 + }, + { + "epoch": 2.531873219373219, + "grad_norm": 0.8544003963470459, + "learning_rate": 5.95634640865684e-05, + "loss": 0.8028, + "step": 14221 + }, + { + "epoch": 2.532051282051282, + "grad_norm": 0.8407679796218872, + "learning_rate": 5.9550662429162655e-05, + "loss": 0.6868, + "step": 14222 + }, + { + "epoch": 2.5322293447293447, + "grad_norm": 0.9049725532531738, + "learning_rate": 5.9537861564277654e-05, + "loss": 0.8177, + "step": 14223 + }, + { + "epoch": 2.5324074074074074, + "grad_norm": 0.938050389289856, + "learning_rate": 5.952506149216419e-05, + "loss": 0.9187, + "step": 14224 + }, + { + "epoch": 2.53258547008547, + "grad_norm": 0.9515482783317566, + "learning_rate": 5.951226221307312e-05, + "loss": 0.6735, + "step": 14225 + }, + { + "epoch": 2.532763532763533, + "grad_norm": 0.8545815348625183, + "learning_rate": 5.949946372725512e-05, + "loss": 0.7643, + "step": 14226 + }, + { + "epoch": 2.5329415954415957, + "grad_norm": 0.8388620615005493, + "learning_rate": 5.9486666034961e-05, + "loss": 0.9437, + "step": 14227 + }, + { + "epoch": 2.533119658119658, + "grad_norm": 0.7202512621879578, + "learning_rate": 5.9473869136441506e-05, + "loss": 0.718, + "step": 14228 + }, + { + "epoch": 2.5332977207977208, + "grad_norm": 0.8375558853149414, + "learning_rate": 5.946107303194739e-05, + "loss": 0.852, + "step": 14229 + }, + { + "epoch": 2.5334757834757835, + "grad_norm": 0.8980572819709778, + "learning_rate": 5.94482777217293e-05, + "loss": 0.9684, + "step": 14230 + }, + { + "epoch": 2.5336538461538463, + "grad_norm": 0.7374732494354248, + "learning_rate": 5.9435483206037977e-05, + "loss": 0.6498, + "step": 14231 + }, + { + "epoch": 2.5338319088319086, + "grad_norm": 1.073758602142334, + "learning_rate": 5.942268948512409e-05, + "loss": 1.0315, + "step": 14232 + }, + { + "epoch": 2.5340099715099713, + "grad_norm": 0.9503611326217651, + "learning_rate": 5.940989655923832e-05, + "loss": 1.0644, + "step": 14233 + }, + { + "epoch": 2.534188034188034, + "grad_norm": 0.870490550994873, + "learning_rate": 5.939710442863129e-05, + "loss": 1.1, + "step": 14234 + }, + { + "epoch": 2.534366096866097, + "grad_norm": 0.8019965887069702, + "learning_rate": 5.93843130935537e-05, + "loss": 0.9169, + "step": 14235 + }, + { + "epoch": 2.5345441595441596, + "grad_norm": 0.8333065509796143, + "learning_rate": 5.9371522554256076e-05, + "loss": 0.7848, + "step": 14236 + }, + { + "epoch": 2.5347222222222223, + "grad_norm": 0.8606435656547546, + "learning_rate": 5.935873281098909e-05, + "loss": 0.9941, + "step": 14237 + }, + { + "epoch": 2.534900284900285, + "grad_norm": 0.7711295485496521, + "learning_rate": 5.934594386400328e-05, + "loss": 0.8495, + "step": 14238 + }, + { + "epoch": 2.535078347578348, + "grad_norm": 0.871533215045929, + "learning_rate": 5.93331557135493e-05, + "loss": 0.9071, + "step": 14239 + }, + { + "epoch": 2.53525641025641, + "grad_norm": 0.9828163981437683, + "learning_rate": 5.932036835987762e-05, + "loss": 0.9561, + "step": 14240 + }, + { + "epoch": 2.535434472934473, + "grad_norm": 0.8485092520713806, + "learning_rate": 5.930758180323881e-05, + "loss": 0.7278, + "step": 14241 + }, + { + "epoch": 2.5356125356125356, + "grad_norm": 0.7608986496925354, + "learning_rate": 5.929479604388342e-05, + "loss": 1.0449, + "step": 14242 + }, + { + "epoch": 2.5357905982905984, + "grad_norm": 0.7852896451950073, + "learning_rate": 5.928201108206193e-05, + "loss": 0.8844, + "step": 14243 + }, + { + "epoch": 2.5359686609686607, + "grad_norm": 0.7636764645576477, + "learning_rate": 5.9269226918024875e-05, + "loss": 0.8259, + "step": 14244 + }, + { + "epoch": 2.5361467236467234, + "grad_norm": 0.9067455530166626, + "learning_rate": 5.925644355202269e-05, + "loss": 0.8742, + "step": 14245 + }, + { + "epoch": 2.536324786324786, + "grad_norm": 0.7911350727081299, + "learning_rate": 5.924366098430588e-05, + "loss": 0.8586, + "step": 14246 + }, + { + "epoch": 2.536502849002849, + "grad_norm": 0.8010593056678772, + "learning_rate": 5.923087921512483e-05, + "loss": 0.8524, + "step": 14247 + }, + { + "epoch": 2.5366809116809117, + "grad_norm": 1.232219934463501, + "learning_rate": 5.9218098244730034e-05, + "loss": 0.8302, + "step": 14248 + }, + { + "epoch": 2.5368589743589745, + "grad_norm": 0.8717244267463684, + "learning_rate": 5.9205318073371874e-05, + "loss": 0.8692, + "step": 14249 + }, + { + "epoch": 2.537037037037037, + "grad_norm": 0.9757453799247742, + "learning_rate": 5.919253870130079e-05, + "loss": 0.7986, + "step": 14250 + }, + { + "epoch": 2.5372150997151, + "grad_norm": 0.8183274865150452, + "learning_rate": 5.917976012876712e-05, + "loss": 0.8277, + "step": 14251 + }, + { + "epoch": 2.5373931623931623, + "grad_norm": 0.823930025100708, + "learning_rate": 5.916698235602125e-05, + "loss": 0.7972, + "step": 14252 + }, + { + "epoch": 2.537571225071225, + "grad_norm": 0.8480231761932373, + "learning_rate": 5.915420538331353e-05, + "loss": 0.8234, + "step": 14253 + }, + { + "epoch": 2.5377492877492878, + "grad_norm": 0.6718716621398926, + "learning_rate": 5.914142921089434e-05, + "loss": 0.5984, + "step": 14254 + }, + { + "epoch": 2.5379273504273505, + "grad_norm": 0.8506333827972412, + "learning_rate": 5.912865383901394e-05, + "loss": 0.8004, + "step": 14255 + }, + { + "epoch": 2.5381054131054133, + "grad_norm": 0.755740225315094, + "learning_rate": 5.911587926792269e-05, + "loss": 0.7465, + "step": 14256 + }, + { + "epoch": 2.5382834757834756, + "grad_norm": 0.8908467888832092, + "learning_rate": 5.9103105497870815e-05, + "loss": 0.7822, + "step": 14257 + }, + { + "epoch": 2.5384615384615383, + "grad_norm": 0.9572851061820984, + "learning_rate": 5.909033252910867e-05, + "loss": 0.7891, + "step": 14258 + }, + { + "epoch": 2.538639601139601, + "grad_norm": 0.8606489896774292, + "learning_rate": 5.907756036188644e-05, + "loss": 0.9071, + "step": 14259 + }, + { + "epoch": 2.538817663817664, + "grad_norm": 0.837658166885376, + "learning_rate": 5.906478899645444e-05, + "loss": 0.829, + "step": 14260 + }, + { + "epoch": 2.5389957264957266, + "grad_norm": 0.8964337706565857, + "learning_rate": 5.905201843306285e-05, + "loss": 1.0385, + "step": 14261 + }, + { + "epoch": 2.5391737891737893, + "grad_norm": 0.7854750156402588, + "learning_rate": 5.903924867196189e-05, + "loss": 0.8905, + "step": 14262 + }, + { + "epoch": 2.539351851851852, + "grad_norm": 0.8828065991401672, + "learning_rate": 5.902647971340176e-05, + "loss": 0.8541, + "step": 14263 + }, + { + "epoch": 2.5395299145299144, + "grad_norm": 0.804121196269989, + "learning_rate": 5.9013711557632645e-05, + "loss": 0.8333, + "step": 14264 + }, + { + "epoch": 2.539707977207977, + "grad_norm": 0.8868918418884277, + "learning_rate": 5.900094420490475e-05, + "loss": 0.7959, + "step": 14265 + }, + { + "epoch": 2.53988603988604, + "grad_norm": 0.9231327176094055, + "learning_rate": 5.8988177655468134e-05, + "loss": 0.8007, + "step": 14266 + }, + { + "epoch": 2.5400641025641026, + "grad_norm": 0.9918177127838135, + "learning_rate": 5.897541190957301e-05, + "loss": 0.7766, + "step": 14267 + }, + { + "epoch": 2.5402421652421654, + "grad_norm": 0.8468625545501709, + "learning_rate": 5.896264696746947e-05, + "loss": 0.9209, + "step": 14268 + }, + { + "epoch": 2.5404202279202277, + "grad_norm": 0.9160833358764648, + "learning_rate": 5.894988282940761e-05, + "loss": 0.8994, + "step": 14269 + }, + { + "epoch": 2.5405982905982905, + "grad_norm": 0.8029152154922485, + "learning_rate": 5.8937119495637515e-05, + "loss": 0.7936, + "step": 14270 + }, + { + "epoch": 2.540776353276353, + "grad_norm": 0.8829928636550903, + "learning_rate": 5.8924356966409286e-05, + "loss": 0.7368, + "step": 14271 + }, + { + "epoch": 2.540954415954416, + "grad_norm": 0.9698056578636169, + "learning_rate": 5.8911595241972925e-05, + "loss": 0.789, + "step": 14272 + }, + { + "epoch": 2.5411324786324787, + "grad_norm": 0.7949244379997253, + "learning_rate": 5.8898834322578524e-05, + "loss": 0.8885, + "step": 14273 + }, + { + "epoch": 2.5413105413105415, + "grad_norm": 1.2430917024612427, + "learning_rate": 5.888607420847605e-05, + "loss": 0.861, + "step": 14274 + }, + { + "epoch": 2.541488603988604, + "grad_norm": 0.7476705312728882, + "learning_rate": 5.887331489991559e-05, + "loss": 0.7942, + "step": 14275 + }, + { + "epoch": 2.5416666666666665, + "grad_norm": 0.9204338192939758, + "learning_rate": 5.886055639714706e-05, + "loss": 0.8633, + "step": 14276 + }, + { + "epoch": 2.5418447293447293, + "grad_norm": 0.8812162280082703, + "learning_rate": 5.884779870042047e-05, + "loss": 0.7162, + "step": 14277 + }, + { + "epoch": 2.542022792022792, + "grad_norm": 0.7859770655632019, + "learning_rate": 5.883504180998578e-05, + "loss": 0.7965, + "step": 14278 + }, + { + "epoch": 2.5422008547008548, + "grad_norm": 0.7732986211776733, + "learning_rate": 5.882228572609296e-05, + "loss": 0.9671, + "step": 14279 + }, + { + "epoch": 2.5423789173789175, + "grad_norm": 0.8555598855018616, + "learning_rate": 5.880953044899189e-05, + "loss": 0.8993, + "step": 14280 + }, + { + "epoch": 2.54255698005698, + "grad_norm": 0.7980908155441284, + "learning_rate": 5.879677597893248e-05, + "loss": 0.873, + "step": 14281 + }, + { + "epoch": 2.5427350427350426, + "grad_norm": 0.9244991540908813, + "learning_rate": 5.878402231616471e-05, + "loss": 0.934, + "step": 14282 + }, + { + "epoch": 2.5429131054131053, + "grad_norm": 1.0128331184387207, + "learning_rate": 5.877126946093835e-05, + "loss": 0.9607, + "step": 14283 + }, + { + "epoch": 2.543091168091168, + "grad_norm": 0.7916569709777832, + "learning_rate": 5.875851741350334e-05, + "loss": 0.831, + "step": 14284 + }, + { + "epoch": 2.543269230769231, + "grad_norm": 0.705007791519165, + "learning_rate": 5.8745766174109495e-05, + "loss": 0.6399, + "step": 14285 + }, + { + "epoch": 2.5434472934472936, + "grad_norm": 0.8785403966903687, + "learning_rate": 5.873301574300671e-05, + "loss": 0.9336, + "step": 14286 + }, + { + "epoch": 2.5436253561253563, + "grad_norm": 0.8225776553153992, + "learning_rate": 5.872026612044471e-05, + "loss": 0.8252, + "step": 14287 + }, + { + "epoch": 2.5438034188034186, + "grad_norm": 0.9629518985748291, + "learning_rate": 5.870751730667337e-05, + "loss": 1.0213, + "step": 14288 + }, + { + "epoch": 2.5439814814814814, + "grad_norm": 0.8242672681808472, + "learning_rate": 5.869476930194242e-05, + "loss": 0.9642, + "step": 14289 + }, + { + "epoch": 2.544159544159544, + "grad_norm": 0.5798216462135315, + "learning_rate": 5.868202210650171e-05, + "loss": 0.4366, + "step": 14290 + }, + { + "epoch": 2.544337606837607, + "grad_norm": 0.7945725917816162, + "learning_rate": 5.86692757206009e-05, + "loss": 0.9252, + "step": 14291 + }, + { + "epoch": 2.5445156695156697, + "grad_norm": 0.9078665375709534, + "learning_rate": 5.865653014448982e-05, + "loss": 1.0551, + "step": 14292 + }, + { + "epoch": 2.544693732193732, + "grad_norm": 0.8044732809066772, + "learning_rate": 5.86437853784181e-05, + "loss": 0.7778, + "step": 14293 + }, + { + "epoch": 2.5448717948717947, + "grad_norm": 0.8317133784294128, + "learning_rate": 5.863104142263553e-05, + "loss": 1.0047, + "step": 14294 + }, + { + "epoch": 2.5450498575498575, + "grad_norm": 0.8330327272415161, + "learning_rate": 5.861829827739174e-05, + "loss": 0.8074, + "step": 14295 + }, + { + "epoch": 2.54522792022792, + "grad_norm": 0.8731801509857178, + "learning_rate": 5.8605555942936474e-05, + "loss": 0.9311, + "step": 14296 + }, + { + "epoch": 2.545405982905983, + "grad_norm": 0.8906812071800232, + "learning_rate": 5.85928144195193e-05, + "loss": 0.9084, + "step": 14297 + }, + { + "epoch": 2.5455840455840457, + "grad_norm": 0.948535144329071, + "learning_rate": 5.8580073707389935e-05, + "loss": 0.923, + "step": 14298 + }, + { + "epoch": 2.5457621082621085, + "grad_norm": 1.0418797731399536, + "learning_rate": 5.8567333806797975e-05, + "loss": 0.9786, + "step": 14299 + }, + { + "epoch": 2.5459401709401708, + "grad_norm": 0.8591430187225342, + "learning_rate": 5.8554594717993075e-05, + "loss": 0.8706, + "step": 14300 + }, + { + "epoch": 2.5461182336182335, + "grad_norm": 1.1056550741195679, + "learning_rate": 5.854185644122475e-05, + "loss": 0.891, + "step": 14301 + }, + { + "epoch": 2.5462962962962963, + "grad_norm": 0.8945133090019226, + "learning_rate": 5.8529118976742624e-05, + "loss": 0.9584, + "step": 14302 + }, + { + "epoch": 2.546474358974359, + "grad_norm": 0.8568279147148132, + "learning_rate": 5.851638232479629e-05, + "loss": 0.7462, + "step": 14303 + }, + { + "epoch": 2.546652421652422, + "grad_norm": 0.934648871421814, + "learning_rate": 5.850364648563527e-05, + "loss": 0.8977, + "step": 14304 + }, + { + "epoch": 2.546830484330484, + "grad_norm": 0.8074216842651367, + "learning_rate": 5.849091145950909e-05, + "loss": 0.8779, + "step": 14305 + }, + { + "epoch": 2.547008547008547, + "grad_norm": 0.8781399726867676, + "learning_rate": 5.8478177246667266e-05, + "loss": 0.8715, + "step": 14306 + }, + { + "epoch": 2.5471866096866096, + "grad_norm": 0.8237441182136536, + "learning_rate": 5.846544384735933e-05, + "loss": 0.8806, + "step": 14307 + }, + { + "epoch": 2.5473646723646723, + "grad_norm": 0.933709442615509, + "learning_rate": 5.8452711261834717e-05, + "loss": 0.8555, + "step": 14308 + }, + { + "epoch": 2.547542735042735, + "grad_norm": 0.9045436978340149, + "learning_rate": 5.843997949034292e-05, + "loss": 1.0105, + "step": 14309 + }, + { + "epoch": 2.547720797720798, + "grad_norm": 0.8088112473487854, + "learning_rate": 5.842724853313337e-05, + "loss": 0.8667, + "step": 14310 + }, + { + "epoch": 2.5478988603988606, + "grad_norm": 1.1022162437438965, + "learning_rate": 5.841451839045559e-05, + "loss": 0.9818, + "step": 14311 + }, + { + "epoch": 2.5480769230769234, + "grad_norm": 0.8974189162254333, + "learning_rate": 5.8401789062558876e-05, + "loss": 0.843, + "step": 14312 + }, + { + "epoch": 2.5482549857549857, + "grad_norm": 0.7816309928894043, + "learning_rate": 5.838906054969272e-05, + "loss": 0.8665, + "step": 14313 + }, + { + "epoch": 2.5484330484330484, + "grad_norm": 0.8243623971939087, + "learning_rate": 5.8376332852106485e-05, + "loss": 0.7291, + "step": 14314 + }, + { + "epoch": 2.548611111111111, + "grad_norm": 0.8475931286811829, + "learning_rate": 5.8363605970049526e-05, + "loss": 0.7551, + "step": 14315 + }, + { + "epoch": 2.548789173789174, + "grad_norm": 0.8949251770973206, + "learning_rate": 5.835087990377124e-05, + "loss": 0.9138, + "step": 14316 + }, + { + "epoch": 2.548967236467236, + "grad_norm": 1.0444703102111816, + "learning_rate": 5.833815465352093e-05, + "loss": 0.8663, + "step": 14317 + }, + { + "epoch": 2.549145299145299, + "grad_norm": 0.8611619472503662, + "learning_rate": 5.8325430219547895e-05, + "loss": 0.795, + "step": 14318 + }, + { + "epoch": 2.5493233618233617, + "grad_norm": 0.7808047533035278, + "learning_rate": 5.8312706602101564e-05, + "loss": 0.905, + "step": 14319 + }, + { + "epoch": 2.5495014245014245, + "grad_norm": 0.9137473106384277, + "learning_rate": 5.8299983801431066e-05, + "loss": 0.8763, + "step": 14320 + }, + { + "epoch": 2.5496794871794872, + "grad_norm": 0.9511715769767761, + "learning_rate": 5.828726181778581e-05, + "loss": 0.8385, + "step": 14321 + }, + { + "epoch": 2.54985754985755, + "grad_norm": 0.9250940084457397, + "learning_rate": 5.8274540651415e-05, + "loss": 0.7026, + "step": 14322 + }, + { + "epoch": 2.5500356125356127, + "grad_norm": 1.001017689704895, + "learning_rate": 5.826182030256786e-05, + "loss": 0.7952, + "step": 14323 + }, + { + "epoch": 2.5502136752136755, + "grad_norm": 0.7638011574745178, + "learning_rate": 5.824910077149371e-05, + "loss": 0.7178, + "step": 14324 + }, + { + "epoch": 2.550391737891738, + "grad_norm": 0.9289371967315674, + "learning_rate": 5.823638205844164e-05, + "loss": 0.8492, + "step": 14325 + }, + { + "epoch": 2.5505698005698005, + "grad_norm": 0.8494341969490051, + "learning_rate": 5.822366416366093e-05, + "loss": 0.9095, + "step": 14326 + }, + { + "epoch": 2.5507478632478633, + "grad_norm": 0.8686699867248535, + "learning_rate": 5.8210947087400746e-05, + "loss": 0.9548, + "step": 14327 + }, + { + "epoch": 2.550925925925926, + "grad_norm": 1.1318142414093018, + "learning_rate": 5.819823082991025e-05, + "loss": 1.0554, + "step": 14328 + }, + { + "epoch": 2.5511039886039883, + "grad_norm": 0.8405448198318481, + "learning_rate": 5.818551539143857e-05, + "loss": 0.704, + "step": 14329 + }, + { + "epoch": 2.551282051282051, + "grad_norm": 0.9133256673812866, + "learning_rate": 5.8172800772234856e-05, + "loss": 0.922, + "step": 14330 + }, + { + "epoch": 2.551460113960114, + "grad_norm": 0.8526531457901001, + "learning_rate": 5.816008697254824e-05, + "loss": 0.6452, + "step": 14331 + }, + { + "epoch": 2.5516381766381766, + "grad_norm": 0.7987905144691467, + "learning_rate": 5.81473739926278e-05, + "loss": 0.8159, + "step": 14332 + }, + { + "epoch": 2.5518162393162394, + "grad_norm": 0.8217538595199585, + "learning_rate": 5.813466183272257e-05, + "loss": 0.6703, + "step": 14333 + }, + { + "epoch": 2.551994301994302, + "grad_norm": 0.7654905915260315, + "learning_rate": 5.8121950493081765e-05, + "loss": 0.8711, + "step": 14334 + }, + { + "epoch": 2.552172364672365, + "grad_norm": 0.872327983379364, + "learning_rate": 5.8109239973954264e-05, + "loss": 1.0079, + "step": 14335 + }, + { + "epoch": 2.5523504273504276, + "grad_norm": 0.7675468325614929, + "learning_rate": 5.809653027558922e-05, + "loss": 0.7541, + "step": 14336 + }, + { + "epoch": 2.55252849002849, + "grad_norm": 0.8367551565170288, + "learning_rate": 5.808382139823563e-05, + "loss": 0.9325, + "step": 14337 + }, + { + "epoch": 2.5527065527065527, + "grad_norm": 0.7946585416793823, + "learning_rate": 5.807111334214248e-05, + "loss": 0.6703, + "step": 14338 + }, + { + "epoch": 2.5528846153846154, + "grad_norm": 0.8752394318580627, + "learning_rate": 5.805840610755876e-05, + "loss": 0.82, + "step": 14339 + }, + { + "epoch": 2.553062678062678, + "grad_norm": 0.9394813776016235, + "learning_rate": 5.804569969473341e-05, + "loss": 1.0094, + "step": 14340 + }, + { + "epoch": 2.5532407407407405, + "grad_norm": 0.7763680219650269, + "learning_rate": 5.803299410391551e-05, + "loss": 0.8424, + "step": 14341 + }, + { + "epoch": 2.5534188034188032, + "grad_norm": 0.9148688912391663, + "learning_rate": 5.8020289335353816e-05, + "loss": 0.9344, + "step": 14342 + }, + { + "epoch": 2.553596866096866, + "grad_norm": 0.8666844367980957, + "learning_rate": 5.80075853892974e-05, + "loss": 0.8651, + "step": 14343 + }, + { + "epoch": 2.5537749287749287, + "grad_norm": 0.7240473628044128, + "learning_rate": 5.799488226599511e-05, + "loss": 0.6913, + "step": 14344 + }, + { + "epoch": 2.5539529914529915, + "grad_norm": 0.8949013352394104, + "learning_rate": 5.798217996569585e-05, + "loss": 0.7419, + "step": 14345 + }, + { + "epoch": 2.5541310541310542, + "grad_norm": 0.7760846614837646, + "learning_rate": 5.796947848864849e-05, + "loss": 0.8292, + "step": 14346 + }, + { + "epoch": 2.554309116809117, + "grad_norm": 0.8448507785797119, + "learning_rate": 5.795677783510187e-05, + "loss": 0.9953, + "step": 14347 + }, + { + "epoch": 2.5544871794871797, + "grad_norm": 0.834007203578949, + "learning_rate": 5.794407800530484e-05, + "loss": 0.8135, + "step": 14348 + }, + { + "epoch": 2.554665242165242, + "grad_norm": 0.8247915506362915, + "learning_rate": 5.793137899950629e-05, + "loss": 0.8607, + "step": 14349 + }, + { + "epoch": 2.554843304843305, + "grad_norm": 0.8796288967132568, + "learning_rate": 5.7918680817954906e-05, + "loss": 1.0479, + "step": 14350 + }, + { + "epoch": 2.5550213675213675, + "grad_norm": 0.8384763598442078, + "learning_rate": 5.790598346089964e-05, + "loss": 0.98, + "step": 14351 + }, + { + "epoch": 2.5551994301994303, + "grad_norm": 0.9394076466560364, + "learning_rate": 5.7893286928589107e-05, + "loss": 0.922, + "step": 14352 + }, + { + "epoch": 2.5553774928774926, + "grad_norm": 0.9548128843307495, + "learning_rate": 5.7880591221272184e-05, + "loss": 0.9877, + "step": 14353 + }, + { + "epoch": 2.5555555555555554, + "grad_norm": 0.7609717845916748, + "learning_rate": 5.786789633919758e-05, + "loss": 0.8115, + "step": 14354 + }, + { + "epoch": 2.555733618233618, + "grad_norm": 0.7415568232536316, + "learning_rate": 5.785520228261403e-05, + "loss": 0.6336, + "step": 14355 + }, + { + "epoch": 2.555911680911681, + "grad_norm": 0.8595952391624451, + "learning_rate": 5.7842509051770246e-05, + "loss": 0.8065, + "step": 14356 + }, + { + "epoch": 2.5560897435897436, + "grad_norm": 1.0075218677520752, + "learning_rate": 5.782981664691491e-05, + "loss": 0.9967, + "step": 14357 + }, + { + "epoch": 2.5562678062678064, + "grad_norm": 0.8405288457870483, + "learning_rate": 5.781712506829669e-05, + "loss": 0.7953, + "step": 14358 + }, + { + "epoch": 2.556445868945869, + "grad_norm": 0.8259321451187134, + "learning_rate": 5.780443431616435e-05, + "loss": 0.925, + "step": 14359 + }, + { + "epoch": 2.556623931623932, + "grad_norm": 0.8155162334442139, + "learning_rate": 5.7791744390766376e-05, + "loss": 0.9658, + "step": 14360 + }, + { + "epoch": 2.556801994301994, + "grad_norm": 0.8670404553413391, + "learning_rate": 5.7779055292351545e-05, + "loss": 1.0029, + "step": 14361 + }, + { + "epoch": 2.556980056980057, + "grad_norm": 0.8574714660644531, + "learning_rate": 5.7766367021168423e-05, + "loss": 0.9208, + "step": 14362 + }, + { + "epoch": 2.5571581196581197, + "grad_norm": 1.0231248140335083, + "learning_rate": 5.775367957746556e-05, + "loss": 1.0422, + "step": 14363 + }, + { + "epoch": 2.5573361823361824, + "grad_norm": 0.8403676152229309, + "learning_rate": 5.7740992961491655e-05, + "loss": 0.8068, + "step": 14364 + }, + { + "epoch": 2.557514245014245, + "grad_norm": 0.8792767524719238, + "learning_rate": 5.7728307173495136e-05, + "loss": 1.0405, + "step": 14365 + }, + { + "epoch": 2.5576923076923075, + "grad_norm": 0.8546510934829712, + "learning_rate": 5.771562221372471e-05, + "loss": 0.8246, + "step": 14366 + }, + { + "epoch": 2.5578703703703702, + "grad_norm": 0.7620588541030884, + "learning_rate": 5.770293808242875e-05, + "loss": 0.7588, + "step": 14367 + }, + { + "epoch": 2.558048433048433, + "grad_norm": 0.8154500722885132, + "learning_rate": 5.769025477985588e-05, + "loss": 0.8217, + "step": 14368 + }, + { + "epoch": 2.5582264957264957, + "grad_norm": 0.8630158305168152, + "learning_rate": 5.767757230625459e-05, + "loss": 0.8486, + "step": 14369 + }, + { + "epoch": 2.5584045584045585, + "grad_norm": 0.8991047143936157, + "learning_rate": 5.766489066187335e-05, + "loss": 0.9012, + "step": 14370 + }, + { + "epoch": 2.5585826210826212, + "grad_norm": 1.056725263595581, + "learning_rate": 5.7652209846960626e-05, + "loss": 0.8764, + "step": 14371 + }, + { + "epoch": 2.558760683760684, + "grad_norm": 0.7467330694198608, + "learning_rate": 5.7639529861764885e-05, + "loss": 0.6614, + "step": 14372 + }, + { + "epoch": 2.5589387464387463, + "grad_norm": 0.7930710315704346, + "learning_rate": 5.762685070653453e-05, + "loss": 0.6866, + "step": 14373 + }, + { + "epoch": 2.559116809116809, + "grad_norm": 0.9234277606010437, + "learning_rate": 5.7614172381518085e-05, + "loss": 0.9158, + "step": 14374 + }, + { + "epoch": 2.559294871794872, + "grad_norm": 1.0100786685943604, + "learning_rate": 5.7601494886963806e-05, + "loss": 0.9061, + "step": 14375 + }, + { + "epoch": 2.5594729344729346, + "grad_norm": 0.9864867925643921, + "learning_rate": 5.758881822312023e-05, + "loss": 0.9955, + "step": 14376 + }, + { + "epoch": 2.5596509971509973, + "grad_norm": 0.7328418493270874, + "learning_rate": 5.757614239023559e-05, + "loss": 0.874, + "step": 14377 + }, + { + "epoch": 2.5598290598290596, + "grad_norm": 0.8538700938224792, + "learning_rate": 5.7563467388558355e-05, + "loss": 0.8251, + "step": 14378 + }, + { + "epoch": 2.5600071225071224, + "grad_norm": 0.7603667378425598, + "learning_rate": 5.755079321833681e-05, + "loss": 0.8466, + "step": 14379 + }, + { + "epoch": 2.560185185185185, + "grad_norm": 0.8983954787254333, + "learning_rate": 5.753811987981925e-05, + "loss": 0.754, + "step": 14380 + }, + { + "epoch": 2.560363247863248, + "grad_norm": 0.8304823040962219, + "learning_rate": 5.752544737325411e-05, + "loss": 0.7057, + "step": 14381 + }, + { + "epoch": 2.5605413105413106, + "grad_norm": 0.8694877028465271, + "learning_rate": 5.751277569888952e-05, + "loss": 0.843, + "step": 14382 + }, + { + "epoch": 2.5607193732193734, + "grad_norm": 0.7965344786643982, + "learning_rate": 5.750010485697387e-05, + "loss": 0.7679, + "step": 14383 + }, + { + "epoch": 2.560897435897436, + "grad_norm": 0.8181809782981873, + "learning_rate": 5.7487434847755386e-05, + "loss": 0.8408, + "step": 14384 + }, + { + "epoch": 2.5610754985754984, + "grad_norm": 0.8492250442504883, + "learning_rate": 5.747476567148229e-05, + "loss": 0.6768, + "step": 14385 + }, + { + "epoch": 2.561253561253561, + "grad_norm": 0.9129379987716675, + "learning_rate": 5.746209732840282e-05, + "loss": 0.8804, + "step": 14386 + }, + { + "epoch": 2.561431623931624, + "grad_norm": 0.8701111078262329, + "learning_rate": 5.74494298187652e-05, + "loss": 1.0908, + "step": 14387 + }, + { + "epoch": 2.5616096866096867, + "grad_norm": 0.9152243733406067, + "learning_rate": 5.7436763142817606e-05, + "loss": 0.8197, + "step": 14388 + }, + { + "epoch": 2.5617877492877494, + "grad_norm": 0.8663429617881775, + "learning_rate": 5.742409730080822e-05, + "loss": 0.8065, + "step": 14389 + }, + { + "epoch": 2.5619658119658117, + "grad_norm": 0.8722090721130371, + "learning_rate": 5.741143229298516e-05, + "loss": 0.8634, + "step": 14390 + }, + { + "epoch": 2.5621438746438745, + "grad_norm": 0.8126732707023621, + "learning_rate": 5.7398768119596704e-05, + "loss": 0.831, + "step": 14391 + }, + { + "epoch": 2.5623219373219372, + "grad_norm": 0.9060684442520142, + "learning_rate": 5.7386104780890794e-05, + "loss": 0.8757, + "step": 14392 + }, + { + "epoch": 2.5625, + "grad_norm": 0.947692334651947, + "learning_rate": 5.7373442277115696e-05, + "loss": 0.8606, + "step": 14393 + }, + { + "epoch": 2.5626780626780628, + "grad_norm": 0.8826618790626526, + "learning_rate": 5.736078060851944e-05, + "loss": 0.9942, + "step": 14394 + }, + { + "epoch": 2.5628561253561255, + "grad_norm": 0.915372908115387, + "learning_rate": 5.734811977535011e-05, + "loss": 0.7871, + "step": 14395 + }, + { + "epoch": 2.5630341880341883, + "grad_norm": 0.8202184438705444, + "learning_rate": 5.733545977785577e-05, + "loss": 0.8889, + "step": 14396 + }, + { + "epoch": 2.5632122507122506, + "grad_norm": 1.0160186290740967, + "learning_rate": 5.7322800616284475e-05, + "loss": 0.8973, + "step": 14397 + }, + { + "epoch": 2.5633903133903133, + "grad_norm": 0.848753809928894, + "learning_rate": 5.7310142290884206e-05, + "loss": 0.9517, + "step": 14398 + }, + { + "epoch": 2.563568376068376, + "grad_norm": 0.7473777532577515, + "learning_rate": 5.72974848019031e-05, + "loss": 0.6549, + "step": 14399 + }, + { + "epoch": 2.563746438746439, + "grad_norm": 0.7396529316902161, + "learning_rate": 5.728482814958899e-05, + "loss": 0.7814, + "step": 14400 + }, + { + "epoch": 2.5639245014245016, + "grad_norm": 0.8617672920227051, + "learning_rate": 5.727217233418998e-05, + "loss": 0.885, + "step": 14401 + }, + { + "epoch": 2.564102564102564, + "grad_norm": 0.920477569103241, + "learning_rate": 5.7259517355953984e-05, + "loss": 1.0269, + "step": 14402 + }, + { + "epoch": 2.5642806267806266, + "grad_norm": 0.8460386991500854, + "learning_rate": 5.7246863215128975e-05, + "loss": 0.7418, + "step": 14403 + }, + { + "epoch": 2.5644586894586894, + "grad_norm": 0.8857694268226624, + "learning_rate": 5.723420991196287e-05, + "loss": 0.7771, + "step": 14404 + }, + { + "epoch": 2.564636752136752, + "grad_norm": 0.9912863969802856, + "learning_rate": 5.722155744670352e-05, + "loss": 1.101, + "step": 14405 + }, + { + "epoch": 2.564814814814815, + "grad_norm": 0.8062789440155029, + "learning_rate": 5.720890581959899e-05, + "loss": 0.8602, + "step": 14406 + }, + { + "epoch": 2.5649928774928776, + "grad_norm": 0.8620314598083496, + "learning_rate": 5.719625503089698e-05, + "loss": 0.9433, + "step": 14407 + }, + { + "epoch": 2.5651709401709404, + "grad_norm": 0.8119623064994812, + "learning_rate": 5.718360508084546e-05, + "loss": 0.884, + "step": 14408 + }, + { + "epoch": 2.5653490028490027, + "grad_norm": 0.7872169613838196, + "learning_rate": 5.7170955969692265e-05, + "loss": 0.8247, + "step": 14409 + }, + { + "epoch": 2.5655270655270654, + "grad_norm": 0.8314040303230286, + "learning_rate": 5.715830769768522e-05, + "loss": 0.8643, + "step": 14410 + }, + { + "epoch": 2.565705128205128, + "grad_norm": 0.9003102779388428, + "learning_rate": 5.7145660265072145e-05, + "loss": 0.9426, + "step": 14411 + }, + { + "epoch": 2.565883190883191, + "grad_norm": 0.6572127938270569, + "learning_rate": 5.713301367210082e-05, + "loss": 0.4852, + "step": 14412 + }, + { + "epoch": 2.5660612535612537, + "grad_norm": 0.9557960629463196, + "learning_rate": 5.7120367919019044e-05, + "loss": 0.9281, + "step": 14413 + }, + { + "epoch": 2.566239316239316, + "grad_norm": 0.9009736180305481, + "learning_rate": 5.71077230060746e-05, + "loss": 1.0154, + "step": 14414 + }, + { + "epoch": 2.5664173789173788, + "grad_norm": 0.8672121167182922, + "learning_rate": 5.7095078933515175e-05, + "loss": 0.805, + "step": 14415 + }, + { + "epoch": 2.5665954415954415, + "grad_norm": 0.9077832698822021, + "learning_rate": 5.708243570158862e-05, + "loss": 0.7446, + "step": 14416 + }, + { + "epoch": 2.5667735042735043, + "grad_norm": 0.850246787071228, + "learning_rate": 5.706979331054252e-05, + "loss": 0.7773, + "step": 14417 + }, + { + "epoch": 2.566951566951567, + "grad_norm": 0.803983211517334, + "learning_rate": 5.705715176062467e-05, + "loss": 0.9361, + "step": 14418 + }, + { + "epoch": 2.5671296296296298, + "grad_norm": 0.8956922888755798, + "learning_rate": 5.704451105208273e-05, + "loss": 0.8962, + "step": 14419 + }, + { + "epoch": 2.5673076923076925, + "grad_norm": 0.8994067907333374, + "learning_rate": 5.703187118516433e-05, + "loss": 0.9902, + "step": 14420 + }, + { + "epoch": 2.567485754985755, + "grad_norm": 0.7383418679237366, + "learning_rate": 5.701923216011722e-05, + "loss": 0.8188, + "step": 14421 + }, + { + "epoch": 2.5676638176638176, + "grad_norm": 0.8397318720817566, + "learning_rate": 5.70065939771889e-05, + "loss": 0.8557, + "step": 14422 + }, + { + "epoch": 2.5678418803418803, + "grad_norm": 0.8804301023483276, + "learning_rate": 5.699395663662714e-05, + "loss": 0.7248, + "step": 14423 + }, + { + "epoch": 2.568019943019943, + "grad_norm": 0.8391412496566772, + "learning_rate": 5.698132013867938e-05, + "loss": 0.6986, + "step": 14424 + }, + { + "epoch": 2.568198005698006, + "grad_norm": 0.7337331771850586, + "learning_rate": 5.6968684483593334e-05, + "loss": 0.7911, + "step": 14425 + }, + { + "epoch": 2.568376068376068, + "grad_norm": 1.006412386894226, + "learning_rate": 5.695604967161652e-05, + "loss": 1.0131, + "step": 14426 + }, + { + "epoch": 2.568554131054131, + "grad_norm": 0.7777771353721619, + "learning_rate": 5.6943415702996494e-05, + "loss": 0.67, + "step": 14427 + }, + { + "epoch": 2.5687321937321936, + "grad_norm": 0.8864775896072388, + "learning_rate": 5.6930782577980803e-05, + "loss": 0.9513, + "step": 14428 + }, + { + "epoch": 2.5689102564102564, + "grad_norm": 0.8505052328109741, + "learning_rate": 5.691815029681695e-05, + "loss": 0.7213, + "step": 14429 + }, + { + "epoch": 2.569088319088319, + "grad_norm": 0.705781877040863, + "learning_rate": 5.6905518859752416e-05, + "loss": 0.8273, + "step": 14430 + }, + { + "epoch": 2.569266381766382, + "grad_norm": 0.7157384753227234, + "learning_rate": 5.689288826703479e-05, + "loss": 0.6854, + "step": 14431 + }, + { + "epoch": 2.5694444444444446, + "grad_norm": 0.871244490146637, + "learning_rate": 5.68802585189114e-05, + "loss": 0.8786, + "step": 14432 + }, + { + "epoch": 2.5696225071225074, + "grad_norm": 0.8742622137069702, + "learning_rate": 5.686762961562981e-05, + "loss": 0.7253, + "step": 14433 + }, + { + "epoch": 2.5698005698005697, + "grad_norm": 0.8194206357002258, + "learning_rate": 5.685500155743742e-05, + "loss": 0.9028, + "step": 14434 + }, + { + "epoch": 2.5699786324786325, + "grad_norm": 0.7505850195884705, + "learning_rate": 5.684237434458164e-05, + "loss": 0.7711, + "step": 14435 + }, + { + "epoch": 2.570156695156695, + "grad_norm": 0.9128859639167786, + "learning_rate": 5.6829747977309885e-05, + "loss": 0.8495, + "step": 14436 + }, + { + "epoch": 2.570334757834758, + "grad_norm": 0.6996384263038635, + "learning_rate": 5.681712245586954e-05, + "loss": 0.6938, + "step": 14437 + }, + { + "epoch": 2.5705128205128203, + "grad_norm": 0.8720461130142212, + "learning_rate": 5.680449778050798e-05, + "loss": 1.0547, + "step": 14438 + }, + { + "epoch": 2.570690883190883, + "grad_norm": 0.7767693996429443, + "learning_rate": 5.6791873951472544e-05, + "loss": 0.8718, + "step": 14439 + }, + { + "epoch": 2.5708689458689458, + "grad_norm": 0.8596739768981934, + "learning_rate": 5.6779250969010554e-05, + "loss": 0.792, + "step": 14440 + }, + { + "epoch": 2.5710470085470085, + "grad_norm": 1.0065197944641113, + "learning_rate": 5.676662883336939e-05, + "loss": 0.9199, + "step": 14441 + }, + { + "epoch": 2.5712250712250713, + "grad_norm": 0.8707680702209473, + "learning_rate": 5.6754007544796316e-05, + "loss": 0.9696, + "step": 14442 + }, + { + "epoch": 2.571403133903134, + "grad_norm": 0.8060235977172852, + "learning_rate": 5.674138710353865e-05, + "loss": 0.6626, + "step": 14443 + }, + { + "epoch": 2.5715811965811968, + "grad_norm": 0.7640239596366882, + "learning_rate": 5.6728767509843627e-05, + "loss": 0.8177, + "step": 14444 + }, + { + "epoch": 2.5717592592592595, + "grad_norm": 1.0821335315704346, + "learning_rate": 5.671614876395848e-05, + "loss": 1.0084, + "step": 14445 + }, + { + "epoch": 2.571937321937322, + "grad_norm": 0.874721884727478, + "learning_rate": 5.670353086613056e-05, + "loss": 0.9508, + "step": 14446 + }, + { + "epoch": 2.5721153846153846, + "grad_norm": 0.7837753891944885, + "learning_rate": 5.669091381660694e-05, + "loss": 0.6546, + "step": 14447 + }, + { + "epoch": 2.5722934472934473, + "grad_norm": 0.832924485206604, + "learning_rate": 5.6678297615634965e-05, + "loss": 0.9055, + "step": 14448 + }, + { + "epoch": 2.57247150997151, + "grad_norm": 0.8463562726974487, + "learning_rate": 5.6665682263461696e-05, + "loss": 0.8234, + "step": 14449 + }, + { + "epoch": 2.5726495726495724, + "grad_norm": 0.8785214424133301, + "learning_rate": 5.6653067760334386e-05, + "loss": 0.8478, + "step": 14450 + }, + { + "epoch": 2.572827635327635, + "grad_norm": 0.7375151515007019, + "learning_rate": 5.664045410650017e-05, + "loss": 0.8629, + "step": 14451 + }, + { + "epoch": 2.573005698005698, + "grad_norm": 0.7428547143936157, + "learning_rate": 5.6627841302206196e-05, + "loss": 0.9198, + "step": 14452 + }, + { + "epoch": 2.5731837606837606, + "grad_norm": 0.7373468279838562, + "learning_rate": 5.661522934769956e-05, + "loss": 0.6931, + "step": 14453 + }, + { + "epoch": 2.5733618233618234, + "grad_norm": 0.9162034392356873, + "learning_rate": 5.660261824322739e-05, + "loss": 0.9971, + "step": 14454 + }, + { + "epoch": 2.573539886039886, + "grad_norm": 0.7816632390022278, + "learning_rate": 5.659000798903672e-05, + "loss": 1.0481, + "step": 14455 + }, + { + "epoch": 2.573717948717949, + "grad_norm": 0.8594158291816711, + "learning_rate": 5.657739858537474e-05, + "loss": 1.1846, + "step": 14456 + }, + { + "epoch": 2.5738960113960117, + "grad_norm": 0.8171747922897339, + "learning_rate": 5.656479003248836e-05, + "loss": 0.8435, + "step": 14457 + }, + { + "epoch": 2.574074074074074, + "grad_norm": 0.8568267822265625, + "learning_rate": 5.6552182330624784e-05, + "loss": 1.031, + "step": 14458 + }, + { + "epoch": 2.5742521367521367, + "grad_norm": 0.8238523602485657, + "learning_rate": 5.653957548003084e-05, + "loss": 0.8917, + "step": 14459 + }, + { + "epoch": 2.5744301994301995, + "grad_norm": 0.7226746082305908, + "learning_rate": 5.652696948095369e-05, + "loss": 0.832, + "step": 14460 + }, + { + "epoch": 2.574608262108262, + "grad_norm": 0.9448554515838623, + "learning_rate": 5.651436433364024e-05, + "loss": 0.9696, + "step": 14461 + }, + { + "epoch": 2.5747863247863245, + "grad_norm": 0.9404924511909485, + "learning_rate": 5.650176003833747e-05, + "loss": 0.9813, + "step": 14462 + }, + { + "epoch": 2.5749643874643873, + "grad_norm": 0.9445366859436035, + "learning_rate": 5.648915659529241e-05, + "loss": 0.7205, + "step": 14463 + }, + { + "epoch": 2.57514245014245, + "grad_norm": 0.9205772876739502, + "learning_rate": 5.647655400475189e-05, + "loss": 0.958, + "step": 14464 + }, + { + "epoch": 2.5753205128205128, + "grad_norm": 0.9025790691375732, + "learning_rate": 5.646395226696291e-05, + "loss": 0.9107, + "step": 14465 + }, + { + "epoch": 2.5754985754985755, + "grad_norm": 0.9562451839447021, + "learning_rate": 5.645135138217235e-05, + "loss": 0.7618, + "step": 14466 + }, + { + "epoch": 2.5756766381766383, + "grad_norm": 0.8896244764328003, + "learning_rate": 5.6438751350627085e-05, + "loss": 0.9696, + "step": 14467 + }, + { + "epoch": 2.575854700854701, + "grad_norm": 0.9051744937896729, + "learning_rate": 5.6426152172574e-05, + "loss": 0.9537, + "step": 14468 + }, + { + "epoch": 2.576032763532764, + "grad_norm": 0.844556450843811, + "learning_rate": 5.641355384825995e-05, + "loss": 0.8686, + "step": 14469 + }, + { + "epoch": 2.576210826210826, + "grad_norm": 0.7751742601394653, + "learning_rate": 5.6400956377931726e-05, + "loss": 0.8373, + "step": 14470 + }, + { + "epoch": 2.576388888888889, + "grad_norm": 0.8988052010536194, + "learning_rate": 5.638835976183627e-05, + "loss": 0.8661, + "step": 14471 + }, + { + "epoch": 2.5765669515669516, + "grad_norm": 0.9114456176757812, + "learning_rate": 5.637576400022023e-05, + "loss": 1.0583, + "step": 14472 + }, + { + "epoch": 2.5767450142450143, + "grad_norm": 0.8742861151695251, + "learning_rate": 5.636316909333056e-05, + "loss": 0.8392, + "step": 14473 + }, + { + "epoch": 2.5769230769230766, + "grad_norm": 0.8418447375297546, + "learning_rate": 5.6350575041413854e-05, + "loss": 0.7494, + "step": 14474 + }, + { + "epoch": 2.5771011396011394, + "grad_norm": 0.9942673444747925, + "learning_rate": 5.633798184471701e-05, + "loss": 1.0183, + "step": 14475 + }, + { + "epoch": 2.577279202279202, + "grad_norm": 0.7663289308547974, + "learning_rate": 5.63253895034867e-05, + "loss": 0.7551, + "step": 14476 + }, + { + "epoch": 2.577457264957265, + "grad_norm": 0.8866778016090393, + "learning_rate": 5.631279801796966e-05, + "loss": 0.8623, + "step": 14477 + }, + { + "epoch": 2.5776353276353277, + "grad_norm": 0.9198449850082397, + "learning_rate": 5.6300207388412595e-05, + "loss": 1.0388, + "step": 14478 + }, + { + "epoch": 2.5778133903133904, + "grad_norm": 0.8202611804008484, + "learning_rate": 5.628761761506214e-05, + "loss": 0.7556, + "step": 14479 + }, + { + "epoch": 2.577991452991453, + "grad_norm": 0.751899003982544, + "learning_rate": 5.627502869816505e-05, + "loss": 0.8231, + "step": 14480 + }, + { + "epoch": 2.578169515669516, + "grad_norm": 1.0094623565673828, + "learning_rate": 5.626244063796795e-05, + "loss": 0.9778, + "step": 14481 + }, + { + "epoch": 2.578347578347578, + "grad_norm": 0.8163259625434875, + "learning_rate": 5.624985343471747e-05, + "loss": 0.8355, + "step": 14482 + }, + { + "epoch": 2.578525641025641, + "grad_norm": 0.8190516829490662, + "learning_rate": 5.623726708866023e-05, + "loss": 0.7736, + "step": 14483 + }, + { + "epoch": 2.5787037037037037, + "grad_norm": 0.884303629398346, + "learning_rate": 5.622468160004283e-05, + "loss": 0.8618, + "step": 14484 + }, + { + "epoch": 2.5788817663817665, + "grad_norm": 0.8564121723175049, + "learning_rate": 5.621209696911185e-05, + "loss": 0.9691, + "step": 14485 + }, + { + "epoch": 2.5790598290598292, + "grad_norm": 0.8122418522834778, + "learning_rate": 5.619951319611388e-05, + "loss": 0.7539, + "step": 14486 + }, + { + "epoch": 2.5792378917378915, + "grad_norm": 0.764470636844635, + "learning_rate": 5.6186930281295425e-05, + "loss": 0.7065, + "step": 14487 + }, + { + "epoch": 2.5794159544159543, + "grad_norm": 0.7477477192878723, + "learning_rate": 5.617434822490313e-05, + "loss": 0.7456, + "step": 14488 + }, + { + "epoch": 2.579594017094017, + "grad_norm": 0.9460917711257935, + "learning_rate": 5.616176702718335e-05, + "loss": 0.8427, + "step": 14489 + }, + { + "epoch": 2.57977207977208, + "grad_norm": 0.858561098575592, + "learning_rate": 5.614918668838274e-05, + "loss": 0.8913, + "step": 14490 + }, + { + "epoch": 2.5799501424501425, + "grad_norm": 0.8664894104003906, + "learning_rate": 5.613660720874772e-05, + "loss": 0.9211, + "step": 14491 + }, + { + "epoch": 2.5801282051282053, + "grad_norm": 0.8475569486618042, + "learning_rate": 5.612402858852475e-05, + "loss": 0.8149, + "step": 14492 + }, + { + "epoch": 2.580306267806268, + "grad_norm": 0.9543033838272095, + "learning_rate": 5.6111450827960296e-05, + "loss": 0.679, + "step": 14493 + }, + { + "epoch": 2.5804843304843303, + "grad_norm": 0.8219496011734009, + "learning_rate": 5.60988739273008e-05, + "loss": 0.9586, + "step": 14494 + }, + { + "epoch": 2.580662393162393, + "grad_norm": 0.8432445526123047, + "learning_rate": 5.6086297886792684e-05, + "loss": 0.8939, + "step": 14495 + }, + { + "epoch": 2.580840455840456, + "grad_norm": 0.9026654362678528, + "learning_rate": 5.607372270668232e-05, + "loss": 0.7422, + "step": 14496 + }, + { + "epoch": 2.5810185185185186, + "grad_norm": 1.0681802034378052, + "learning_rate": 5.606114838721608e-05, + "loss": 0.8208, + "step": 14497 + }, + { + "epoch": 2.5811965811965814, + "grad_norm": 0.8807427883148193, + "learning_rate": 5.604857492864044e-05, + "loss": 0.8463, + "step": 14498 + }, + { + "epoch": 2.5813746438746437, + "grad_norm": 0.7520862221717834, + "learning_rate": 5.603600233120159e-05, + "loss": 0.6691, + "step": 14499 + }, + { + "epoch": 2.5815527065527064, + "grad_norm": 0.8214079737663269, + "learning_rate": 5.602343059514599e-05, + "loss": 0.8416, + "step": 14500 + }, + { + "epoch": 2.581730769230769, + "grad_norm": 0.9263389110565186, + "learning_rate": 5.601085972071991e-05, + "loss": 1.1466, + "step": 14501 + }, + { + "epoch": 2.581908831908832, + "grad_norm": 0.8501101136207581, + "learning_rate": 5.5998289708169626e-05, + "loss": 0.855, + "step": 14502 + }, + { + "epoch": 2.5820868945868947, + "grad_norm": 0.8312939405441284, + "learning_rate": 5.598572055774152e-05, + "loss": 0.9843, + "step": 14503 + }, + { + "epoch": 2.5822649572649574, + "grad_norm": 0.7309035658836365, + "learning_rate": 5.5973152269681714e-05, + "loss": 0.813, + "step": 14504 + }, + { + "epoch": 2.58244301994302, + "grad_norm": 0.8962578177452087, + "learning_rate": 5.596058484423656e-05, + "loss": 0.7619, + "step": 14505 + }, + { + "epoch": 2.5826210826210825, + "grad_norm": 0.7805112600326538, + "learning_rate": 5.594801828165228e-05, + "loss": 1.1011, + "step": 14506 + }, + { + "epoch": 2.5827991452991452, + "grad_norm": 1.224509358406067, + "learning_rate": 5.593545258217505e-05, + "loss": 0.9764, + "step": 14507 + }, + { + "epoch": 2.582977207977208, + "grad_norm": 0.8085877895355225, + "learning_rate": 5.59228877460511e-05, + "loss": 0.9324, + "step": 14508 + }, + { + "epoch": 2.5831552706552707, + "grad_norm": 0.7962629199028015, + "learning_rate": 5.591032377352661e-05, + "loss": 0.6294, + "step": 14509 + }, + { + "epoch": 2.5833333333333335, + "grad_norm": 0.8638611435890198, + "learning_rate": 5.589776066484773e-05, + "loss": 0.7355, + "step": 14510 + }, + { + "epoch": 2.583511396011396, + "grad_norm": 0.8975821733474731, + "learning_rate": 5.588519842026061e-05, + "loss": 1.0264, + "step": 14511 + }, + { + "epoch": 2.5836894586894585, + "grad_norm": 0.8327218890190125, + "learning_rate": 5.5872637040011355e-05, + "loss": 0.8864, + "step": 14512 + }, + { + "epoch": 2.5838675213675213, + "grad_norm": 0.8141334652900696, + "learning_rate": 5.5860076524346197e-05, + "loss": 1.0277, + "step": 14513 + }, + { + "epoch": 2.584045584045584, + "grad_norm": 0.8557519316673279, + "learning_rate": 5.584751687351105e-05, + "loss": 0.9215, + "step": 14514 + }, + { + "epoch": 2.584223646723647, + "grad_norm": 0.902601957321167, + "learning_rate": 5.583495808775214e-05, + "loss": 0.8527, + "step": 14515 + }, + { + "epoch": 2.5844017094017095, + "grad_norm": 0.826359212398529, + "learning_rate": 5.582240016731548e-05, + "loss": 0.8524, + "step": 14516 + }, + { + "epoch": 2.5845797720797723, + "grad_norm": 0.7099179029464722, + "learning_rate": 5.580984311244713e-05, + "loss": 0.6923, + "step": 14517 + }, + { + "epoch": 2.5847578347578346, + "grad_norm": 0.829795777797699, + "learning_rate": 5.5797286923393086e-05, + "loss": 0.7211, + "step": 14518 + }, + { + "epoch": 2.5849358974358974, + "grad_norm": 0.8006768226623535, + "learning_rate": 5.5784731600399355e-05, + "loss": 0.7237, + "step": 14519 + }, + { + "epoch": 2.58511396011396, + "grad_norm": 0.7596119046211243, + "learning_rate": 5.577217714371203e-05, + "loss": 0.7651, + "step": 14520 + }, + { + "epoch": 2.585292022792023, + "grad_norm": 0.7901585102081299, + "learning_rate": 5.575962355357694e-05, + "loss": 0.7672, + "step": 14521 + }, + { + "epoch": 2.5854700854700856, + "grad_norm": 0.8586403131484985, + "learning_rate": 5.574707083024018e-05, + "loss": 1.084, + "step": 14522 + }, + { + "epoch": 2.585648148148148, + "grad_norm": 0.7670607566833496, + "learning_rate": 5.5734518973947616e-05, + "loss": 0.7929, + "step": 14523 + }, + { + "epoch": 2.5858262108262107, + "grad_norm": 0.8114384412765503, + "learning_rate": 5.572196798494522e-05, + "loss": 0.8154, + "step": 14524 + }, + { + "epoch": 2.5860042735042734, + "grad_norm": 0.8050188422203064, + "learning_rate": 5.570941786347888e-05, + "loss": 0.7969, + "step": 14525 + }, + { + "epoch": 2.586182336182336, + "grad_norm": 0.8641461133956909, + "learning_rate": 5.569686860979447e-05, + "loss": 0.8469, + "step": 14526 + }, + { + "epoch": 2.586360398860399, + "grad_norm": 0.7644940614700317, + "learning_rate": 5.568432022413787e-05, + "loss": 0.563, + "step": 14527 + }, + { + "epoch": 2.5865384615384617, + "grad_norm": 0.7620565891265869, + "learning_rate": 5.567177270675503e-05, + "loss": 0.657, + "step": 14528 + }, + { + "epoch": 2.5867165242165244, + "grad_norm": 0.8371306657791138, + "learning_rate": 5.5659226057891634e-05, + "loss": 0.8862, + "step": 14529 + }, + { + "epoch": 2.5868945868945867, + "grad_norm": 0.8996389508247375, + "learning_rate": 5.564668027779367e-05, + "loss": 0.6031, + "step": 14530 + }, + { + "epoch": 2.5870726495726495, + "grad_norm": 0.8691734671592712, + "learning_rate": 5.5634135366706806e-05, + "loss": 0.9198, + "step": 14531 + }, + { + "epoch": 2.5872507122507122, + "grad_norm": 0.8926620483398438, + "learning_rate": 5.562159132487693e-05, + "loss": 0.8691, + "step": 14532 + }, + { + "epoch": 2.587428774928775, + "grad_norm": 1.0852068662643433, + "learning_rate": 5.5609048152549794e-05, + "loss": 1.2338, + "step": 14533 + }, + { + "epoch": 2.5876068376068377, + "grad_norm": 0.7894790172576904, + "learning_rate": 5.5596505849971124e-05, + "loss": 0.907, + "step": 14534 + }, + { + "epoch": 2.5877849002849, + "grad_norm": 0.8084964156150818, + "learning_rate": 5.558396441738669e-05, + "loss": 0.9082, + "step": 14535 + }, + { + "epoch": 2.587962962962963, + "grad_norm": 1.0563920736312866, + "learning_rate": 5.557142385504222e-05, + "loss": 1.0364, + "step": 14536 + }, + { + "epoch": 2.5881410256410255, + "grad_norm": 0.7996996641159058, + "learning_rate": 5.5558884163183354e-05, + "loss": 0.925, + "step": 14537 + }, + { + "epoch": 2.5883190883190883, + "grad_norm": 0.7493244409561157, + "learning_rate": 5.5546345342055916e-05, + "loss": 0.9516, + "step": 14538 + }, + { + "epoch": 2.588497150997151, + "grad_norm": 0.8916776776313782, + "learning_rate": 5.553380739190541e-05, + "loss": 0.8164, + "step": 14539 + }, + { + "epoch": 2.588675213675214, + "grad_norm": 0.8178156614303589, + "learning_rate": 5.552127031297762e-05, + "loss": 0.905, + "step": 14540 + }, + { + "epoch": 2.5888532763532766, + "grad_norm": 0.8305806517601013, + "learning_rate": 5.550873410551816e-05, + "loss": 0.789, + "step": 14541 + }, + { + "epoch": 2.5890313390313393, + "grad_norm": 0.9307064414024353, + "learning_rate": 5.549619876977258e-05, + "loss": 0.8529, + "step": 14542 + }, + { + "epoch": 2.5892094017094016, + "grad_norm": 0.8526419401168823, + "learning_rate": 5.5483664305986614e-05, + "loss": 0.8314, + "step": 14543 + }, + { + "epoch": 2.5893874643874644, + "grad_norm": 0.884918212890625, + "learning_rate": 5.547113071440568e-05, + "loss": 0.7957, + "step": 14544 + }, + { + "epoch": 2.589565527065527, + "grad_norm": 0.7517948746681213, + "learning_rate": 5.5458597995275554e-05, + "loss": 0.7012, + "step": 14545 + }, + { + "epoch": 2.58974358974359, + "grad_norm": 0.8321232199668884, + "learning_rate": 5.5446066148841556e-05, + "loss": 1.0017, + "step": 14546 + }, + { + "epoch": 2.589921652421652, + "grad_norm": 0.8279885053634644, + "learning_rate": 5.543353517534939e-05, + "loss": 0.987, + "step": 14547 + }, + { + "epoch": 2.590099715099715, + "grad_norm": 0.8651175498962402, + "learning_rate": 5.542100507504454e-05, + "loss": 0.8929, + "step": 14548 + }, + { + "epoch": 2.5902777777777777, + "grad_norm": 0.9273492097854614, + "learning_rate": 5.540847584817248e-05, + "loss": 0.9503, + "step": 14549 + }, + { + "epoch": 2.5904558404558404, + "grad_norm": 0.8779071569442749, + "learning_rate": 5.5395947494978696e-05, + "loss": 0.9099, + "step": 14550 + }, + { + "epoch": 2.590633903133903, + "grad_norm": 0.8860164880752563, + "learning_rate": 5.538342001570868e-05, + "loss": 0.9559, + "step": 14551 + }, + { + "epoch": 2.590811965811966, + "grad_norm": 0.9232339859008789, + "learning_rate": 5.5370893410607816e-05, + "loss": 0.9495, + "step": 14552 + }, + { + "epoch": 2.5909900284900287, + "grad_norm": 0.8176831007003784, + "learning_rate": 5.5358367679921666e-05, + "loss": 0.8897, + "step": 14553 + }, + { + "epoch": 2.5911680911680914, + "grad_norm": 0.7926605939865112, + "learning_rate": 5.5345842823895486e-05, + "loss": 0.7609, + "step": 14554 + }, + { + "epoch": 2.5913461538461537, + "grad_norm": 0.9837173819541931, + "learning_rate": 5.533331884277484e-05, + "loss": 0.7842, + "step": 14555 + }, + { + "epoch": 2.5915242165242165, + "grad_norm": 0.7303726673126221, + "learning_rate": 5.5320795736804945e-05, + "loss": 0.824, + "step": 14556 + }, + { + "epoch": 2.5917022792022792, + "grad_norm": 0.8379296660423279, + "learning_rate": 5.530827350623128e-05, + "loss": 0.8005, + "step": 14557 + }, + { + "epoch": 2.591880341880342, + "grad_norm": 0.8562047481536865, + "learning_rate": 5.529575215129916e-05, + "loss": 1.0048, + "step": 14558 + }, + { + "epoch": 2.5920584045584043, + "grad_norm": 0.7543022632598877, + "learning_rate": 5.528323167225386e-05, + "loss": 0.7543, + "step": 14559 + }, + { + "epoch": 2.592236467236467, + "grad_norm": 0.8205977082252502, + "learning_rate": 5.5270712069340847e-05, + "loss": 0.997, + "step": 14560 + }, + { + "epoch": 2.59241452991453, + "grad_norm": 0.8566918969154358, + "learning_rate": 5.525819334280522e-05, + "loss": 0.9222, + "step": 14561 + }, + { + "epoch": 2.5925925925925926, + "grad_norm": 0.8513971567153931, + "learning_rate": 5.524567549289239e-05, + "loss": 0.7007, + "step": 14562 + }, + { + "epoch": 2.5927706552706553, + "grad_norm": 0.8939194679260254, + "learning_rate": 5.523315851984758e-05, + "loss": 0.8597, + "step": 14563 + }, + { + "epoch": 2.592948717948718, + "grad_norm": 0.7597625851631165, + "learning_rate": 5.5220642423916035e-05, + "loss": 0.7122, + "step": 14564 + }, + { + "epoch": 2.593126780626781, + "grad_norm": 0.9511955976486206, + "learning_rate": 5.5208127205342983e-05, + "loss": 1.0905, + "step": 14565 + }, + { + "epoch": 2.5933048433048436, + "grad_norm": 0.8359304070472717, + "learning_rate": 5.5195612864373626e-05, + "loss": 0.7132, + "step": 14566 + }, + { + "epoch": 2.593482905982906, + "grad_norm": 0.8302733302116394, + "learning_rate": 5.518309940125317e-05, + "loss": 0.9123, + "step": 14567 + }, + { + "epoch": 2.5936609686609686, + "grad_norm": 0.7923629283905029, + "learning_rate": 5.517058681622678e-05, + "loss": 0.8384, + "step": 14568 + }, + { + "epoch": 2.5938390313390314, + "grad_norm": 1.0625137090682983, + "learning_rate": 5.515807510953956e-05, + "loss": 1.0262, + "step": 14569 + }, + { + "epoch": 2.594017094017094, + "grad_norm": 1.0595879554748535, + "learning_rate": 5.5145564281436804e-05, + "loss": 0.9112, + "step": 14570 + }, + { + "epoch": 2.5941951566951564, + "grad_norm": 0.7307499647140503, + "learning_rate": 5.513305433216346e-05, + "loss": 0.8273, + "step": 14571 + }, + { + "epoch": 2.594373219373219, + "grad_norm": 0.9221912026405334, + "learning_rate": 5.512054526196475e-05, + "loss": 1.0679, + "step": 14572 + }, + { + "epoch": 2.594551282051282, + "grad_norm": 0.8098722100257874, + "learning_rate": 5.5108037071085725e-05, + "loss": 0.922, + "step": 14573 + }, + { + "epoch": 2.5947293447293447, + "grad_norm": 0.984785258769989, + "learning_rate": 5.509552975977146e-05, + "loss": 0.7525, + "step": 14574 + }, + { + "epoch": 2.5949074074074074, + "grad_norm": 0.8076850771903992, + "learning_rate": 5.5083023328267006e-05, + "loss": 1.008, + "step": 14575 + }, + { + "epoch": 2.59508547008547, + "grad_norm": 0.8375436067581177, + "learning_rate": 5.507051777681741e-05, + "loss": 0.8822, + "step": 14576 + }, + { + "epoch": 2.595263532763533, + "grad_norm": 0.779228687286377, + "learning_rate": 5.505801310566764e-05, + "loss": 0.8072, + "step": 14577 + }, + { + "epoch": 2.5954415954415957, + "grad_norm": 0.7347875833511353, + "learning_rate": 5.504550931506278e-05, + "loss": 0.7796, + "step": 14578 + }, + { + "epoch": 2.595619658119658, + "grad_norm": 0.8229580521583557, + "learning_rate": 5.503300640524779e-05, + "loss": 0.9337, + "step": 14579 + }, + { + "epoch": 2.5957977207977208, + "grad_norm": 0.8643096089363098, + "learning_rate": 5.502050437646762e-05, + "loss": 0.9101, + "step": 14580 + }, + { + "epoch": 2.5959757834757835, + "grad_norm": 0.769158661365509, + "learning_rate": 5.500800322896723e-05, + "loss": 0.8417, + "step": 14581 + }, + { + "epoch": 2.5961538461538463, + "grad_norm": 0.7792086005210876, + "learning_rate": 5.4995502962991566e-05, + "loss": 0.6965, + "step": 14582 + }, + { + "epoch": 2.5963319088319086, + "grad_norm": 0.7833219170570374, + "learning_rate": 5.498300357878552e-05, + "loss": 0.641, + "step": 14583 + }, + { + "epoch": 2.5965099715099713, + "grad_norm": 0.9491978287696838, + "learning_rate": 5.4970505076593956e-05, + "loss": 0.9229, + "step": 14584 + }, + { + "epoch": 2.596688034188034, + "grad_norm": 0.9128090739250183, + "learning_rate": 5.495800745666191e-05, + "loss": 0.8047, + "step": 14585 + }, + { + "epoch": 2.596866096866097, + "grad_norm": 0.9235281944274902, + "learning_rate": 5.494551071923404e-05, + "loss": 0.961, + "step": 14586 + }, + { + "epoch": 2.5970441595441596, + "grad_norm": 0.8582631349563599, + "learning_rate": 5.493301486455536e-05, + "loss": 0.7203, + "step": 14587 + }, + { + "epoch": 2.5972222222222223, + "grad_norm": 0.9605505466461182, + "learning_rate": 5.4920519892870605e-05, + "loss": 0.8315, + "step": 14588 + }, + { + "epoch": 2.597400284900285, + "grad_norm": 0.9344304203987122, + "learning_rate": 5.490802580442462e-05, + "loss": 0.9031, + "step": 14589 + }, + { + "epoch": 2.597578347578348, + "grad_norm": 1.0027791261672974, + "learning_rate": 5.4895532599462216e-05, + "loss": 1.0361, + "step": 14590 + }, + { + "epoch": 2.59775641025641, + "grad_norm": 0.8774647116661072, + "learning_rate": 5.488304027822815e-05, + "loss": 0.9533, + "step": 14591 + }, + { + "epoch": 2.597934472934473, + "grad_norm": 0.886246919631958, + "learning_rate": 5.487054884096718e-05, + "loss": 0.8588, + "step": 14592 + }, + { + "epoch": 2.5981125356125356, + "grad_norm": 0.8963425755500793, + "learning_rate": 5.485805828792408e-05, + "loss": 0.8685, + "step": 14593 + }, + { + "epoch": 2.5982905982905984, + "grad_norm": 0.7650768756866455, + "learning_rate": 5.484556861934349e-05, + "loss": 0.7441, + "step": 14594 + }, + { + "epoch": 2.5984686609686607, + "grad_norm": 0.8266916871070862, + "learning_rate": 5.483307983547026e-05, + "loss": 0.9625, + "step": 14595 + }, + { + "epoch": 2.5986467236467234, + "grad_norm": 0.8243923783302307, + "learning_rate": 5.482059193654894e-05, + "loss": 0.8553, + "step": 14596 + }, + { + "epoch": 2.598824786324786, + "grad_norm": 0.8200470209121704, + "learning_rate": 5.48081049228243e-05, + "loss": 0.6682, + "step": 14597 + }, + { + "epoch": 2.599002849002849, + "grad_norm": 0.8360442519187927, + "learning_rate": 5.479561879454097e-05, + "loss": 0.8996, + "step": 14598 + }, + { + "epoch": 2.5991809116809117, + "grad_norm": 0.8326625227928162, + "learning_rate": 5.4783133551943546e-05, + "loss": 0.6532, + "step": 14599 + }, + { + "epoch": 2.5993589743589745, + "grad_norm": 0.8162251114845276, + "learning_rate": 5.4770649195276766e-05, + "loss": 1.0514, + "step": 14600 + }, + { + "epoch": 2.599537037037037, + "grad_norm": 1.0407251119613647, + "learning_rate": 5.4758165724785084e-05, + "loss": 0.7991, + "step": 14601 + }, + { + "epoch": 2.5997150997151, + "grad_norm": 0.9161550998687744, + "learning_rate": 5.474568314071323e-05, + "loss": 0.8623, + "step": 14602 + }, + { + "epoch": 2.5998931623931623, + "grad_norm": 0.8405734896659851, + "learning_rate": 5.4733201443305646e-05, + "loss": 0.8406, + "step": 14603 + }, + { + "epoch": 2.600071225071225, + "grad_norm": 0.937198281288147, + "learning_rate": 5.472072063280698e-05, + "loss": 1.0887, + "step": 14604 + }, + { + "epoch": 2.6002492877492878, + "grad_norm": 0.8800520896911621, + "learning_rate": 5.470824070946172e-05, + "loss": 0.8738, + "step": 14605 + }, + { + "epoch": 2.6004273504273505, + "grad_norm": 0.9473027586936951, + "learning_rate": 5.4695761673514425e-05, + "loss": 0.8188, + "step": 14606 + }, + { + "epoch": 2.6006054131054133, + "grad_norm": 0.8547683954238892, + "learning_rate": 5.468328352520955e-05, + "loss": 0.7619, + "step": 14607 + }, + { + "epoch": 2.6007834757834756, + "grad_norm": 1.0138040781021118, + "learning_rate": 5.4670806264791595e-05, + "loss": 1.0805, + "step": 14608 + }, + { + "epoch": 2.6009615384615383, + "grad_norm": 0.8458215594291687, + "learning_rate": 5.465832989250499e-05, + "loss": 0.8386, + "step": 14609 + }, + { + "epoch": 2.601139601139601, + "grad_norm": 0.811152458190918, + "learning_rate": 5.464585440859431e-05, + "loss": 0.8158, + "step": 14610 + }, + { + "epoch": 2.601317663817664, + "grad_norm": 0.9584031701087952, + "learning_rate": 5.463337981330381e-05, + "loss": 0.8537, + "step": 14611 + }, + { + "epoch": 2.6014957264957266, + "grad_norm": 0.8734773397445679, + "learning_rate": 5.462090610687802e-05, + "loss": 1.0246, + "step": 14612 + }, + { + "epoch": 2.6016737891737893, + "grad_norm": 0.8463562726974487, + "learning_rate": 5.460843328956133e-05, + "loss": 0.8763, + "step": 14613 + }, + { + "epoch": 2.601851851851852, + "grad_norm": 0.8010903000831604, + "learning_rate": 5.459596136159808e-05, + "loss": 0.8438, + "step": 14614 + }, + { + "epoch": 2.6020299145299144, + "grad_norm": 0.7927500009536743, + "learning_rate": 5.458349032323267e-05, + "loss": 0.7388, + "step": 14615 + }, + { + "epoch": 2.602207977207977, + "grad_norm": 0.784017026424408, + "learning_rate": 5.4571020174709407e-05, + "loss": 0.6981, + "step": 14616 + }, + { + "epoch": 2.60238603988604, + "grad_norm": 0.8732004761695862, + "learning_rate": 5.455855091627263e-05, + "loss": 1.043, + "step": 14617 + }, + { + "epoch": 2.6025641025641026, + "grad_norm": 0.7947654128074646, + "learning_rate": 5.454608254816662e-05, + "loss": 0.9487, + "step": 14618 + }, + { + "epoch": 2.6027421652421654, + "grad_norm": 0.8809077739715576, + "learning_rate": 5.4533615070635734e-05, + "loss": 0.8499, + "step": 14619 + }, + { + "epoch": 2.6029202279202277, + "grad_norm": 0.9094803333282471, + "learning_rate": 5.452114848392422e-05, + "loss": 0.9522, + "step": 14620 + }, + { + "epoch": 2.6030982905982905, + "grad_norm": 0.8943446278572083, + "learning_rate": 5.4508682788276324e-05, + "loss": 0.7328, + "step": 14621 + }, + { + "epoch": 2.603276353276353, + "grad_norm": 0.856849730014801, + "learning_rate": 5.449621798393628e-05, + "loss": 0.7536, + "step": 14622 + }, + { + "epoch": 2.603454415954416, + "grad_norm": 0.8199608325958252, + "learning_rate": 5.448375407114833e-05, + "loss": 0.6377, + "step": 14623 + }, + { + "epoch": 2.6036324786324787, + "grad_norm": 0.8981915712356567, + "learning_rate": 5.4471291050156626e-05, + "loss": 1.0372, + "step": 14624 + }, + { + "epoch": 2.6038105413105415, + "grad_norm": 0.8449446558952332, + "learning_rate": 5.4458828921205465e-05, + "loss": 0.9948, + "step": 14625 + }, + { + "epoch": 2.603988603988604, + "grad_norm": 0.8807474970817566, + "learning_rate": 5.444636768453888e-05, + "loss": 0.9752, + "step": 14626 + }, + { + "epoch": 2.6041666666666665, + "grad_norm": 0.8212316036224365, + "learning_rate": 5.443390734040117e-05, + "loss": 0.9221, + "step": 14627 + }, + { + "epoch": 2.6043447293447293, + "grad_norm": 0.8049453496932983, + "learning_rate": 5.4421447889036304e-05, + "loss": 0.7726, + "step": 14628 + }, + { + "epoch": 2.604522792022792, + "grad_norm": 0.8091840744018555, + "learning_rate": 5.440898933068853e-05, + "loss": 0.9152, + "step": 14629 + }, + { + "epoch": 2.6047008547008548, + "grad_norm": 0.8409022688865662, + "learning_rate": 5.43965316656019e-05, + "loss": 0.8672, + "step": 14630 + }, + { + "epoch": 2.6048789173789175, + "grad_norm": 0.7622308731079102, + "learning_rate": 5.4384074894020496e-05, + "loss": 0.9021, + "step": 14631 + }, + { + "epoch": 2.60505698005698, + "grad_norm": 0.8272425532341003, + "learning_rate": 5.437161901618839e-05, + "loss": 0.7729, + "step": 14632 + }, + { + "epoch": 2.6052350427350426, + "grad_norm": 0.8699020743370056, + "learning_rate": 5.435916403234963e-05, + "loss": 0.8211, + "step": 14633 + }, + { + "epoch": 2.6054131054131053, + "grad_norm": 0.8145751357078552, + "learning_rate": 5.4346709942748196e-05, + "loss": 0.8996, + "step": 14634 + }, + { + "epoch": 2.605591168091168, + "grad_norm": 0.9398832321166992, + "learning_rate": 5.433425674762822e-05, + "loss": 0.8116, + "step": 14635 + }, + { + "epoch": 2.605769230769231, + "grad_norm": 0.9191767573356628, + "learning_rate": 5.4321804447233535e-05, + "loss": 0.8933, + "step": 14636 + }, + { + "epoch": 2.6059472934472936, + "grad_norm": 0.7511529326438904, + "learning_rate": 5.430935304180831e-05, + "loss": 0.7595, + "step": 14637 + }, + { + "epoch": 2.6061253561253563, + "grad_norm": 0.9087170362472534, + "learning_rate": 5.4296902531596296e-05, + "loss": 0.9781, + "step": 14638 + }, + { + "epoch": 2.6063034188034186, + "grad_norm": 0.8496448397636414, + "learning_rate": 5.4284452916841575e-05, + "loss": 0.9852, + "step": 14639 + }, + { + "epoch": 2.6064814814814814, + "grad_norm": 0.868609607219696, + "learning_rate": 5.427200419778804e-05, + "loss": 0.781, + "step": 14640 + }, + { + "epoch": 2.606659544159544, + "grad_norm": 0.7752132415771484, + "learning_rate": 5.4259556374679553e-05, + "loss": 0.7319, + "step": 14641 + }, + { + "epoch": 2.606837606837607, + "grad_norm": 0.8950543999671936, + "learning_rate": 5.4247109447760124e-05, + "loss": 0.7637, + "step": 14642 + }, + { + "epoch": 2.6070156695156697, + "grad_norm": 0.892699658870697, + "learning_rate": 5.423466341727346e-05, + "loss": 0.8274, + "step": 14643 + }, + { + "epoch": 2.607193732193732, + "grad_norm": 0.9283786416053772, + "learning_rate": 5.422221828346352e-05, + "loss": 1.1009, + "step": 14644 + }, + { + "epoch": 2.6073717948717947, + "grad_norm": 0.7551446557044983, + "learning_rate": 5.420977404657413e-05, + "loss": 0.8105, + "step": 14645 + }, + { + "epoch": 2.6075498575498575, + "grad_norm": 0.8014101386070251, + "learning_rate": 5.41973307068491e-05, + "loss": 0.838, + "step": 14646 + }, + { + "epoch": 2.60772792022792, + "grad_norm": 0.8941731452941895, + "learning_rate": 5.418488826453223e-05, + "loss": 0.9557, + "step": 14647 + }, + { + "epoch": 2.607905982905983, + "grad_norm": 0.7990903258323669, + "learning_rate": 5.41724467198673e-05, + "loss": 0.9634, + "step": 14648 + }, + { + "epoch": 2.6080840455840457, + "grad_norm": 1.0688040256500244, + "learning_rate": 5.4160006073098035e-05, + "loss": 0.8976, + "step": 14649 + }, + { + "epoch": 2.6082621082621085, + "grad_norm": 0.8451266884803772, + "learning_rate": 5.4147566324468313e-05, + "loss": 0.8703, + "step": 14650 + }, + { + "epoch": 2.6084401709401708, + "grad_norm": 0.8196333050727844, + "learning_rate": 5.413512747422169e-05, + "loss": 0.9423, + "step": 14651 + }, + { + "epoch": 2.6086182336182335, + "grad_norm": 0.7639298439025879, + "learning_rate": 5.412268952260204e-05, + "loss": 0.9092, + "step": 14652 + }, + { + "epoch": 2.6087962962962963, + "grad_norm": 0.88963782787323, + "learning_rate": 5.411025246985293e-05, + "loss": 1.2503, + "step": 14653 + }, + { + "epoch": 2.608974358974359, + "grad_norm": 0.831516683101654, + "learning_rate": 5.409781631621812e-05, + "loss": 0.8643, + "step": 14654 + }, + { + "epoch": 2.609152421652422, + "grad_norm": 0.7729721069335938, + "learning_rate": 5.408538106194125e-05, + "loss": 0.8289, + "step": 14655 + }, + { + "epoch": 2.609330484330484, + "grad_norm": 0.8360101580619812, + "learning_rate": 5.407294670726596e-05, + "loss": 0.8619, + "step": 14656 + }, + { + "epoch": 2.609508547008547, + "grad_norm": 0.7525733709335327, + "learning_rate": 5.406051325243586e-05, + "loss": 0.8353, + "step": 14657 + }, + { + "epoch": 2.6096866096866096, + "grad_norm": 0.8943357467651367, + "learning_rate": 5.404808069769456e-05, + "loss": 0.9291, + "step": 14658 + }, + { + "epoch": 2.6098646723646723, + "grad_norm": 1.024953007698059, + "learning_rate": 5.403564904328568e-05, + "loss": 1.0414, + "step": 14659 + }, + { + "epoch": 2.610042735042735, + "grad_norm": 0.8671780228614807, + "learning_rate": 5.402321828945278e-05, + "loss": 0.9309, + "step": 14660 + }, + { + "epoch": 2.610220797720798, + "grad_norm": 0.9765334725379944, + "learning_rate": 5.4010788436439406e-05, + "loss": 0.9399, + "step": 14661 + }, + { + "epoch": 2.6103988603988606, + "grad_norm": 0.8996732234954834, + "learning_rate": 5.3998359484489106e-05, + "loss": 0.9868, + "step": 14662 + }, + { + "epoch": 2.6105769230769234, + "grad_norm": 0.8597404956817627, + "learning_rate": 5.398593143384538e-05, + "loss": 1.0328, + "step": 14663 + }, + { + "epoch": 2.6107549857549857, + "grad_norm": 0.8909318447113037, + "learning_rate": 5.397350428475176e-05, + "loss": 0.9362, + "step": 14664 + }, + { + "epoch": 2.6109330484330484, + "grad_norm": 0.8874006867408752, + "learning_rate": 5.39610780374517e-05, + "loss": 0.9254, + "step": 14665 + }, + { + "epoch": 2.611111111111111, + "grad_norm": 0.8325822949409485, + "learning_rate": 5.3948652692188626e-05, + "loss": 0.8495, + "step": 14666 + }, + { + "epoch": 2.611289173789174, + "grad_norm": 0.847998857498169, + "learning_rate": 5.393622824920614e-05, + "loss": 0.8372, + "step": 14667 + }, + { + "epoch": 2.611467236467236, + "grad_norm": 0.8439756631851196, + "learning_rate": 5.392380470874749e-05, + "loss": 0.8934, + "step": 14668 + }, + { + "epoch": 2.611645299145299, + "grad_norm": 0.9563834071159363, + "learning_rate": 5.39113820710562e-05, + "loss": 1.1213, + "step": 14669 + }, + { + "epoch": 2.6118233618233617, + "grad_norm": 0.7761119604110718, + "learning_rate": 5.3898960336375646e-05, + "loss": 0.6104, + "step": 14670 + }, + { + "epoch": 2.6120014245014245, + "grad_norm": 0.8661524653434753, + "learning_rate": 5.38865395049492e-05, + "loss": 0.7562, + "step": 14671 + }, + { + "epoch": 2.6121794871794872, + "grad_norm": 0.854347825050354, + "learning_rate": 5.387411957702021e-05, + "loss": 0.8613, + "step": 14672 + }, + { + "epoch": 2.61235754985755, + "grad_norm": 0.7728402614593506, + "learning_rate": 5.386170055283204e-05, + "loss": 0.7879, + "step": 14673 + }, + { + "epoch": 2.6125356125356127, + "grad_norm": 0.8647109270095825, + "learning_rate": 5.384928243262799e-05, + "loss": 0.835, + "step": 14674 + }, + { + "epoch": 2.6127136752136755, + "grad_norm": 0.6764749884605408, + "learning_rate": 5.383686521665139e-05, + "loss": 0.7233, + "step": 14675 + }, + { + "epoch": 2.612891737891738, + "grad_norm": 0.8431640863418579, + "learning_rate": 5.382444890514548e-05, + "loss": 1.1699, + "step": 14676 + }, + { + "epoch": 2.6130698005698005, + "grad_norm": 0.9196193814277649, + "learning_rate": 5.381203349835364e-05, + "loss": 0.8668, + "step": 14677 + }, + { + "epoch": 2.6132478632478633, + "grad_norm": 0.9449048638343811, + "learning_rate": 5.3799618996519e-05, + "loss": 0.8353, + "step": 14678 + }, + { + "epoch": 2.613425925925926, + "grad_norm": 0.9835928678512573, + "learning_rate": 5.378720539988488e-05, + "loss": 1.1129, + "step": 14679 + }, + { + "epoch": 2.6136039886039883, + "grad_norm": 0.763592004776001, + "learning_rate": 5.377479270869448e-05, + "loss": 0.7929, + "step": 14680 + }, + { + "epoch": 2.613782051282051, + "grad_norm": 0.8119748830795288, + "learning_rate": 5.376238092319094e-05, + "loss": 1.0257, + "step": 14681 + }, + { + "epoch": 2.613960113960114, + "grad_norm": 0.7605236172676086, + "learning_rate": 5.374997004361757e-05, + "loss": 0.7005, + "step": 14682 + }, + { + "epoch": 2.6141381766381766, + "grad_norm": 0.9077369570732117, + "learning_rate": 5.3737560070217394e-05, + "loss": 0.9208, + "step": 14683 + }, + { + "epoch": 2.6143162393162394, + "grad_norm": 0.9089310765266418, + "learning_rate": 5.3725151003233665e-05, + "loss": 0.6855, + "step": 14684 + }, + { + "epoch": 2.614494301994302, + "grad_norm": 0.8387685418128967, + "learning_rate": 5.371274284290947e-05, + "loss": 0.8682, + "step": 14685 + }, + { + "epoch": 2.614672364672365, + "grad_norm": 0.7626301050186157, + "learning_rate": 5.3700335589487925e-05, + "loss": 0.6928, + "step": 14686 + }, + { + "epoch": 2.6148504273504276, + "grad_norm": 1.2667319774627686, + "learning_rate": 5.368792924321213e-05, + "loss": 0.9288, + "step": 14687 + }, + { + "epoch": 2.61502849002849, + "grad_norm": 0.8570333123207092, + "learning_rate": 5.3675523804325154e-05, + "loss": 0.9916, + "step": 14688 + }, + { + "epoch": 2.6152065527065527, + "grad_norm": 0.9050240516662598, + "learning_rate": 5.366311927307006e-05, + "loss": 0.7734, + "step": 14689 + }, + { + "epoch": 2.6153846153846154, + "grad_norm": 1.000036358833313, + "learning_rate": 5.365071564968989e-05, + "loss": 0.7932, + "step": 14690 + }, + { + "epoch": 2.615562678062678, + "grad_norm": 0.8147441744804382, + "learning_rate": 5.363831293442763e-05, + "loss": 0.8867, + "step": 14691 + }, + { + "epoch": 2.6157407407407405, + "grad_norm": 0.8662015795707703, + "learning_rate": 5.3625911127526375e-05, + "loss": 0.6742, + "step": 14692 + }, + { + "epoch": 2.6159188034188032, + "grad_norm": 0.8576271533966064, + "learning_rate": 5.3613510229229e-05, + "loss": 0.8161, + "step": 14693 + }, + { + "epoch": 2.616096866096866, + "grad_norm": 0.8862481713294983, + "learning_rate": 5.360111023977856e-05, + "loss": 0.8774, + "step": 14694 + }, + { + "epoch": 2.6162749287749287, + "grad_norm": 0.8384450674057007, + "learning_rate": 5.358871115941799e-05, + "loss": 0.9149, + "step": 14695 + }, + { + "epoch": 2.6164529914529915, + "grad_norm": 0.9055412411689758, + "learning_rate": 5.357631298839021e-05, + "loss": 0.8197, + "step": 14696 + }, + { + "epoch": 2.6166310541310542, + "grad_norm": 0.937764585018158, + "learning_rate": 5.356391572693813e-05, + "loss": 1.0392, + "step": 14697 + }, + { + "epoch": 2.616809116809117, + "grad_norm": 0.8917306661605835, + "learning_rate": 5.355151937530463e-05, + "loss": 0.868, + "step": 14698 + }, + { + "epoch": 2.6169871794871797, + "grad_norm": 0.7353024482727051, + "learning_rate": 5.3539123933732705e-05, + "loss": 0.7788, + "step": 14699 + }, + { + "epoch": 2.617165242165242, + "grad_norm": 0.8607454299926758, + "learning_rate": 5.352672940246504e-05, + "loss": 0.7746, + "step": 14700 + }, + { + "epoch": 2.617343304843305, + "grad_norm": 0.9775658249855042, + "learning_rate": 5.3514335781744616e-05, + "loss": 0.9438, + "step": 14701 + }, + { + "epoch": 2.6175213675213675, + "grad_norm": 0.9416237473487854, + "learning_rate": 5.350194307181422e-05, + "loss": 0.9581, + "step": 14702 + }, + { + "epoch": 2.6176994301994303, + "grad_norm": 0.8378105163574219, + "learning_rate": 5.348955127291666e-05, + "loss": 1.0038, + "step": 14703 + }, + { + "epoch": 2.6178774928774926, + "grad_norm": 0.8199161887168884, + "learning_rate": 5.347716038529471e-05, + "loss": 0.9492, + "step": 14704 + }, + { + "epoch": 2.6180555555555554, + "grad_norm": 0.9511042833328247, + "learning_rate": 5.3464770409191176e-05, + "loss": 1.2101, + "step": 14705 + }, + { + "epoch": 2.618233618233618, + "grad_norm": 0.8017105460166931, + "learning_rate": 5.3452381344848754e-05, + "loss": 0.9524, + "step": 14706 + }, + { + "epoch": 2.618411680911681, + "grad_norm": 0.8174898624420166, + "learning_rate": 5.34399931925103e-05, + "loss": 0.911, + "step": 14707 + }, + { + "epoch": 2.6185897435897436, + "grad_norm": 0.8134239315986633, + "learning_rate": 5.342760595241838e-05, + "loss": 0.8971, + "step": 14708 + }, + { + "epoch": 2.6187678062678064, + "grad_norm": 0.817252516746521, + "learning_rate": 5.341521962481586e-05, + "loss": 0.8472, + "step": 14709 + }, + { + "epoch": 2.618945868945869, + "grad_norm": 0.8675270080566406, + "learning_rate": 5.3402834209945264e-05, + "loss": 0.9607, + "step": 14710 + }, + { + "epoch": 2.619123931623932, + "grad_norm": 1.0281410217285156, + "learning_rate": 5.339044970804936e-05, + "loss": 1.0487, + "step": 14711 + }, + { + "epoch": 2.619301994301994, + "grad_norm": 0.9276307225227356, + "learning_rate": 5.33780661193708e-05, + "loss": 0.8915, + "step": 14712 + }, + { + "epoch": 2.619480056980057, + "grad_norm": 0.8479217290878296, + "learning_rate": 5.336568344415216e-05, + "loss": 0.929, + "step": 14713 + }, + { + "epoch": 2.6196581196581197, + "grad_norm": 0.8695724010467529, + "learning_rate": 5.335330168263608e-05, + "loss": 0.8651, + "step": 14714 + }, + { + "epoch": 2.6198361823361824, + "grad_norm": 0.7740936875343323, + "learning_rate": 5.3340920835065155e-05, + "loss": 0.8572, + "step": 14715 + }, + { + "epoch": 2.620014245014245, + "grad_norm": 0.8619815111160278, + "learning_rate": 5.332854090168192e-05, + "loss": 0.6934, + "step": 14716 + }, + { + "epoch": 2.6201923076923075, + "grad_norm": 0.8866271376609802, + "learning_rate": 5.331616188272902e-05, + "loss": 1.038, + "step": 14717 + }, + { + "epoch": 2.6203703703703702, + "grad_norm": 0.7526047825813293, + "learning_rate": 5.330378377844896e-05, + "loss": 0.8534, + "step": 14718 + }, + { + "epoch": 2.620548433048433, + "grad_norm": 0.6914070248603821, + "learning_rate": 5.329140658908423e-05, + "loss": 0.5355, + "step": 14719 + }, + { + "epoch": 2.6207264957264957, + "grad_norm": 0.886074423789978, + "learning_rate": 5.3279030314877374e-05, + "loss": 0.8277, + "step": 14720 + }, + { + "epoch": 2.6209045584045585, + "grad_norm": 0.9101460576057434, + "learning_rate": 5.326665495607082e-05, + "loss": 0.8711, + "step": 14721 + }, + { + "epoch": 2.6210826210826212, + "grad_norm": 0.9744461178779602, + "learning_rate": 5.3254280512907175e-05, + "loss": 1.2376, + "step": 14722 + }, + { + "epoch": 2.621260683760684, + "grad_norm": 1.013480544090271, + "learning_rate": 5.32419069856287e-05, + "loss": 0.8946, + "step": 14723 + }, + { + "epoch": 2.6214387464387463, + "grad_norm": 0.82442706823349, + "learning_rate": 5.3229534374478005e-05, + "loss": 0.732, + "step": 14724 + }, + { + "epoch": 2.621616809116809, + "grad_norm": 0.7960239052772522, + "learning_rate": 5.3217162679697366e-05, + "loss": 0.7633, + "step": 14725 + }, + { + "epoch": 2.621794871794872, + "grad_norm": 0.819844126701355, + "learning_rate": 5.320479190152926e-05, + "loss": 0.974, + "step": 14726 + }, + { + "epoch": 2.6219729344729346, + "grad_norm": 0.8245221376419067, + "learning_rate": 5.319242204021606e-05, + "loss": 0.9122, + "step": 14727 + }, + { + "epoch": 2.6221509971509973, + "grad_norm": 0.7574561834335327, + "learning_rate": 5.318005309600011e-05, + "loss": 0.8427, + "step": 14728 + }, + { + "epoch": 2.6223290598290596, + "grad_norm": 1.0385704040527344, + "learning_rate": 5.316768506912377e-05, + "loss": 0.8214, + "step": 14729 + }, + { + "epoch": 2.6225071225071224, + "grad_norm": 0.8616722822189331, + "learning_rate": 5.3155317959829346e-05, + "loss": 0.8469, + "step": 14730 + }, + { + "epoch": 2.622685185185185, + "grad_norm": 0.909667432308197, + "learning_rate": 5.314295176835912e-05, + "loss": 0.9156, + "step": 14731 + }, + { + "epoch": 2.622863247863248, + "grad_norm": 0.9016293883323669, + "learning_rate": 5.3130586494955494e-05, + "loss": 0.9183, + "step": 14732 + }, + { + "epoch": 2.6230413105413106, + "grad_norm": 0.8828284740447998, + "learning_rate": 5.311822213986057e-05, + "loss": 0.8338, + "step": 14733 + }, + { + "epoch": 2.6232193732193734, + "grad_norm": 0.8159047365188599, + "learning_rate": 5.3105858703316794e-05, + "loss": 0.7055, + "step": 14734 + }, + { + "epoch": 2.623397435897436, + "grad_norm": 0.9240905046463013, + "learning_rate": 5.309349618556623e-05, + "loss": 0.9078, + "step": 14735 + }, + { + "epoch": 2.6235754985754984, + "grad_norm": 0.8881595134735107, + "learning_rate": 5.308113458685118e-05, + "loss": 0.9946, + "step": 14736 + }, + { + "epoch": 2.623753561253561, + "grad_norm": 0.8781841397285461, + "learning_rate": 5.306877390741385e-05, + "loss": 0.8252, + "step": 14737 + }, + { + "epoch": 2.623931623931624, + "grad_norm": 0.8348106741905212, + "learning_rate": 5.3056414147496355e-05, + "loss": 0.8653, + "step": 14738 + }, + { + "epoch": 2.6241096866096867, + "grad_norm": 0.9692304134368896, + "learning_rate": 5.3044055307341e-05, + "loss": 0.7814, + "step": 14739 + }, + { + "epoch": 2.6242877492877494, + "grad_norm": 0.866179347038269, + "learning_rate": 5.303169738718976e-05, + "loss": 0.9255, + "step": 14740 + }, + { + "epoch": 2.6244658119658117, + "grad_norm": 0.9306690692901611, + "learning_rate": 5.301934038728487e-05, + "loss": 0.9123, + "step": 14741 + }, + { + "epoch": 2.6246438746438745, + "grad_norm": 0.949357271194458, + "learning_rate": 5.3006984307868415e-05, + "loss": 0.8452, + "step": 14742 + }, + { + "epoch": 2.6248219373219372, + "grad_norm": 0.8638128042221069, + "learning_rate": 5.299462914918249e-05, + "loss": 0.8026, + "step": 14743 + }, + { + "epoch": 2.625, + "grad_norm": 0.9075117707252502, + "learning_rate": 5.2982274911469154e-05, + "loss": 1.0644, + "step": 14744 + }, + { + "epoch": 2.6251780626780628, + "grad_norm": 0.8146225810050964, + "learning_rate": 5.296992159497047e-05, + "loss": 0.8494, + "step": 14745 + }, + { + "epoch": 2.6253561253561255, + "grad_norm": 0.8887025713920593, + "learning_rate": 5.295756919992847e-05, + "loss": 0.8143, + "step": 14746 + }, + { + "epoch": 2.6255341880341883, + "grad_norm": 0.8262654542922974, + "learning_rate": 5.29452177265852e-05, + "loss": 0.7559, + "step": 14747 + }, + { + "epoch": 2.6257122507122506, + "grad_norm": 0.8126912117004395, + "learning_rate": 5.2932867175182574e-05, + "loss": 0.8528, + "step": 14748 + }, + { + "epoch": 2.6258903133903133, + "grad_norm": 0.8970595598220825, + "learning_rate": 5.2920517545962746e-05, + "loss": 0.8584, + "step": 14749 + }, + { + "epoch": 2.626068376068376, + "grad_norm": 0.8678651452064514, + "learning_rate": 5.290816883916748e-05, + "loss": 0.8686, + "step": 14750 + }, + { + "epoch": 2.626246438746439, + "grad_norm": 0.8069576621055603, + "learning_rate": 5.289582105503887e-05, + "loss": 0.868, + "step": 14751 + }, + { + "epoch": 2.6264245014245016, + "grad_norm": 1.0322144031524658, + "learning_rate": 5.28834741938188e-05, + "loss": 1.1537, + "step": 14752 + }, + { + "epoch": 2.626602564102564, + "grad_norm": 0.8274349570274353, + "learning_rate": 5.287112825574917e-05, + "loss": 1.0126, + "step": 14753 + }, + { + "epoch": 2.6267806267806266, + "grad_norm": 0.8820709586143494, + "learning_rate": 5.2858783241071875e-05, + "loss": 0.893, + "step": 14754 + }, + { + "epoch": 2.6269586894586894, + "grad_norm": 1.0102146863937378, + "learning_rate": 5.28464391500288e-05, + "loss": 0.8524, + "step": 14755 + }, + { + "epoch": 2.627136752136752, + "grad_norm": 0.875468373298645, + "learning_rate": 5.2834095982861764e-05, + "loss": 1.0991, + "step": 14756 + }, + { + "epoch": 2.627314814814815, + "grad_norm": 0.8155242800712585, + "learning_rate": 5.282175373981267e-05, + "loss": 0.666, + "step": 14757 + }, + { + "epoch": 2.6274928774928776, + "grad_norm": 0.8777057528495789, + "learning_rate": 5.280941242112332e-05, + "loss": 0.892, + "step": 14758 + }, + { + "epoch": 2.6276709401709404, + "grad_norm": 0.8357667922973633, + "learning_rate": 5.279707202703549e-05, + "loss": 0.8118, + "step": 14759 + }, + { + "epoch": 2.6278490028490027, + "grad_norm": 0.7862337827682495, + "learning_rate": 5.278473255779097e-05, + "loss": 0.7287, + "step": 14760 + }, + { + "epoch": 2.6280270655270654, + "grad_norm": 0.8340336084365845, + "learning_rate": 5.277239401363155e-05, + "loss": 0.7697, + "step": 14761 + }, + { + "epoch": 2.628205128205128, + "grad_norm": 0.7986457943916321, + "learning_rate": 5.276005639479896e-05, + "loss": 0.9358, + "step": 14762 + }, + { + "epoch": 2.628383190883191, + "grad_norm": 0.7377769947052002, + "learning_rate": 5.2747719701534895e-05, + "loss": 0.8091, + "step": 14763 + }, + { + "epoch": 2.6285612535612537, + "grad_norm": 0.9749723672866821, + "learning_rate": 5.273538393408117e-05, + "loss": 0.8163, + "step": 14764 + }, + { + "epoch": 2.628739316239316, + "grad_norm": 0.8718321323394775, + "learning_rate": 5.2723049092679354e-05, + "loss": 1.1587, + "step": 14765 + }, + { + "epoch": 2.6289173789173788, + "grad_norm": 0.9394767880439758, + "learning_rate": 5.27107151775712e-05, + "loss": 0.9409, + "step": 14766 + }, + { + "epoch": 2.6290954415954415, + "grad_norm": 0.9763813614845276, + "learning_rate": 5.269838218899836e-05, + "loss": 1.0171, + "step": 14767 + }, + { + "epoch": 2.6292735042735043, + "grad_norm": 0.878968358039856, + "learning_rate": 5.268605012720247e-05, + "loss": 0.9117, + "step": 14768 + }, + { + "epoch": 2.629451566951567, + "grad_norm": 0.8240547776222229, + "learning_rate": 5.267371899242512e-05, + "loss": 0.9351, + "step": 14769 + }, + { + "epoch": 2.6296296296296298, + "grad_norm": 0.8048275709152222, + "learning_rate": 5.266138878490795e-05, + "loss": 0.9331, + "step": 14770 + }, + { + "epoch": 2.6298076923076925, + "grad_norm": 0.7176041007041931, + "learning_rate": 5.264905950489252e-05, + "loss": 0.6424, + "step": 14771 + }, + { + "epoch": 2.629985754985755, + "grad_norm": 0.973258912563324, + "learning_rate": 5.263673115262041e-05, + "loss": 0.9295, + "step": 14772 + }, + { + "epoch": 2.6301638176638176, + "grad_norm": 0.8955824971199036, + "learning_rate": 5.262440372833313e-05, + "loss": 0.9306, + "step": 14773 + }, + { + "epoch": 2.6303418803418803, + "grad_norm": 0.8430632948875427, + "learning_rate": 5.2612077232272305e-05, + "loss": 0.9343, + "step": 14774 + }, + { + "epoch": 2.630519943019943, + "grad_norm": 1.0231794118881226, + "learning_rate": 5.2599751664679334e-05, + "loss": 0.941, + "step": 14775 + }, + { + "epoch": 2.630698005698006, + "grad_norm": 0.9726024866104126, + "learning_rate": 5.258742702579579e-05, + "loss": 1.1726, + "step": 14776 + }, + { + "epoch": 2.630876068376068, + "grad_norm": 0.8575723171234131, + "learning_rate": 5.257510331586312e-05, + "loss": 0.5644, + "step": 14777 + }, + { + "epoch": 2.631054131054131, + "grad_norm": 0.853165864944458, + "learning_rate": 5.2562780535122744e-05, + "loss": 0.8555, + "step": 14778 + }, + { + "epoch": 2.6312321937321936, + "grad_norm": 0.861574649810791, + "learning_rate": 5.255045868381623e-05, + "loss": 0.8298, + "step": 14779 + }, + { + "epoch": 2.6314102564102564, + "grad_norm": 0.8744526505470276, + "learning_rate": 5.2538137762184816e-05, + "loss": 0.9889, + "step": 14780 + }, + { + "epoch": 2.631588319088319, + "grad_norm": 0.7891412973403931, + "learning_rate": 5.2525817770470084e-05, + "loss": 0.9765, + "step": 14781 + }, + { + "epoch": 2.631766381766382, + "grad_norm": 0.9155156016349792, + "learning_rate": 5.251349870891327e-05, + "loss": 0.8927, + "step": 14782 + }, + { + "epoch": 2.6319444444444446, + "grad_norm": 0.8547508120536804, + "learning_rate": 5.250118057775582e-05, + "loss": 0.8479, + "step": 14783 + }, + { + "epoch": 2.6321225071225074, + "grad_norm": 0.7606263756752014, + "learning_rate": 5.248886337723908e-05, + "loss": 0.7557, + "step": 14784 + }, + { + "epoch": 2.6323005698005697, + "grad_norm": 0.855315625667572, + "learning_rate": 5.247654710760437e-05, + "loss": 0.8527, + "step": 14785 + }, + { + "epoch": 2.6324786324786325, + "grad_norm": 0.7656288743019104, + "learning_rate": 5.246423176909298e-05, + "loss": 0.8881, + "step": 14786 + }, + { + "epoch": 2.632656695156695, + "grad_norm": 0.817034125328064, + "learning_rate": 5.2451917361946236e-05, + "loss": 1.042, + "step": 14787 + }, + { + "epoch": 2.632834757834758, + "grad_norm": 0.8473303318023682, + "learning_rate": 5.2439603886405356e-05, + "loss": 0.8804, + "step": 14788 + }, + { + "epoch": 2.6330128205128203, + "grad_norm": 0.9563126564025879, + "learning_rate": 5.242729134271171e-05, + "loss": 0.8463, + "step": 14789 + }, + { + "epoch": 2.633190883190883, + "grad_norm": 0.8297066688537598, + "learning_rate": 5.241497973110641e-05, + "loss": 0.7776, + "step": 14790 + }, + { + "epoch": 2.6333689458689458, + "grad_norm": 0.8433563709259033, + "learning_rate": 5.240266905183075e-05, + "loss": 0.8712, + "step": 14791 + }, + { + "epoch": 2.6335470085470085, + "grad_norm": 0.814725935459137, + "learning_rate": 5.239035930512593e-05, + "loss": 0.9819, + "step": 14792 + }, + { + "epoch": 2.6337250712250713, + "grad_norm": 0.844292163848877, + "learning_rate": 5.23780504912331e-05, + "loss": 0.8693, + "step": 14793 + }, + { + "epoch": 2.633903133903134, + "grad_norm": 0.8194862008094788, + "learning_rate": 5.2365742610393464e-05, + "loss": 0.7878, + "step": 14794 + }, + { + "epoch": 2.6340811965811968, + "grad_norm": 0.8570502400398254, + "learning_rate": 5.2353435662848135e-05, + "loss": 0.815, + "step": 14795 + }, + { + "epoch": 2.6342592592592595, + "grad_norm": 0.9301772713661194, + "learning_rate": 5.2341129648838275e-05, + "loss": 0.9092, + "step": 14796 + }, + { + "epoch": 2.634437321937322, + "grad_norm": 0.7605858445167542, + "learning_rate": 5.232882456860493e-05, + "loss": 0.8753, + "step": 14797 + }, + { + "epoch": 2.6346153846153846, + "grad_norm": 0.8265452980995178, + "learning_rate": 5.231652042238927e-05, + "loss": 0.9134, + "step": 14798 + }, + { + "epoch": 2.6347934472934473, + "grad_norm": 0.7440468072891235, + "learning_rate": 5.230421721043235e-05, + "loss": 0.7471, + "step": 14799 + }, + { + "epoch": 2.63497150997151, + "grad_norm": 0.9172230958938599, + "learning_rate": 5.2291914932975205e-05, + "loss": 1.0155, + "step": 14800 + }, + { + "epoch": 2.6351495726495724, + "grad_norm": 0.8364499807357788, + "learning_rate": 5.227961359025888e-05, + "loss": 0.9561, + "step": 14801 + }, + { + "epoch": 2.635327635327635, + "grad_norm": 0.7756382822990417, + "learning_rate": 5.22673131825244e-05, + "loss": 0.6893, + "step": 14802 + }, + { + "epoch": 2.635505698005698, + "grad_norm": 0.9042136669158936, + "learning_rate": 5.225501371001273e-05, + "loss": 0.7613, + "step": 14803 + }, + { + "epoch": 2.6356837606837606, + "grad_norm": 0.8989379405975342, + "learning_rate": 5.224271517296495e-05, + "loss": 0.8092, + "step": 14804 + }, + { + "epoch": 2.6358618233618234, + "grad_norm": 0.7999827265739441, + "learning_rate": 5.2230417571621906e-05, + "loss": 0.8115, + "step": 14805 + }, + { + "epoch": 2.636039886039886, + "grad_norm": 0.9071131348609924, + "learning_rate": 5.221812090622464e-05, + "loss": 0.9072, + "step": 14806 + }, + { + "epoch": 2.636217948717949, + "grad_norm": 0.7227704524993896, + "learning_rate": 5.220582517701398e-05, + "loss": 0.7598, + "step": 14807 + }, + { + "epoch": 2.6363960113960117, + "grad_norm": 0.8520537614822388, + "learning_rate": 5.219353038423094e-05, + "loss": 1.1072, + "step": 14808 + }, + { + "epoch": 2.636574074074074, + "grad_norm": 0.8690574765205383, + "learning_rate": 5.218123652811634e-05, + "loss": 0.773, + "step": 14809 + }, + { + "epoch": 2.6367521367521367, + "grad_norm": 0.7897602319717407, + "learning_rate": 5.216894360891109e-05, + "loss": 0.792, + "step": 14810 + }, + { + "epoch": 2.6369301994301995, + "grad_norm": 0.8746532201766968, + "learning_rate": 5.215665162685601e-05, + "loss": 0.8853, + "step": 14811 + }, + { + "epoch": 2.637108262108262, + "grad_norm": 0.8525128364562988, + "learning_rate": 5.214436058219199e-05, + "loss": 0.7293, + "step": 14812 + }, + { + "epoch": 2.6372863247863245, + "grad_norm": 0.979969322681427, + "learning_rate": 5.213207047515975e-05, + "loss": 0.8485, + "step": 14813 + }, + { + "epoch": 2.6374643874643873, + "grad_norm": 0.8439529538154602, + "learning_rate": 5.211978130600024e-05, + "loss": 0.7492, + "step": 14814 + }, + { + "epoch": 2.63764245014245, + "grad_norm": 0.8356610536575317, + "learning_rate": 5.2107493074954064e-05, + "loss": 0.8255, + "step": 14815 + }, + { + "epoch": 2.6378205128205128, + "grad_norm": 0.7857736349105835, + "learning_rate": 5.2095205782262116e-05, + "loss": 0.766, + "step": 14816 + }, + { + "epoch": 2.6379985754985755, + "grad_norm": 0.919058084487915, + "learning_rate": 5.20829194281651e-05, + "loss": 1.0661, + "step": 14817 + }, + { + "epoch": 2.6381766381766383, + "grad_norm": 0.8793047070503235, + "learning_rate": 5.207063401290373e-05, + "loss": 0.8297, + "step": 14818 + }, + { + "epoch": 2.638354700854701, + "grad_norm": 0.7848390340805054, + "learning_rate": 5.205834953671873e-05, + "loss": 0.8051, + "step": 14819 + }, + { + "epoch": 2.638532763532764, + "grad_norm": 0.8391907215118408, + "learning_rate": 5.2046065999850736e-05, + "loss": 0.8444, + "step": 14820 + }, + { + "epoch": 2.638710826210826, + "grad_norm": 0.8137226700782776, + "learning_rate": 5.2033783402540546e-05, + "loss": 0.7908, + "step": 14821 + }, + { + "epoch": 2.638888888888889, + "grad_norm": 0.8440108299255371, + "learning_rate": 5.2021501745028645e-05, + "loss": 0.7985, + "step": 14822 + }, + { + "epoch": 2.6390669515669516, + "grad_norm": 0.7432600855827332, + "learning_rate": 5.200922102755581e-05, + "loss": 0.7816, + "step": 14823 + }, + { + "epoch": 2.6392450142450143, + "grad_norm": 0.9003379344940186, + "learning_rate": 5.199694125036257e-05, + "loss": 0.9171, + "step": 14824 + }, + { + "epoch": 2.6394230769230766, + "grad_norm": 0.8994988203048706, + "learning_rate": 5.198466241368957e-05, + "loss": 0.8333, + "step": 14825 + }, + { + "epoch": 2.6396011396011394, + "grad_norm": 0.9042859077453613, + "learning_rate": 5.197238451777735e-05, + "loss": 0.7491, + "step": 14826 + }, + { + "epoch": 2.639779202279202, + "grad_norm": 0.8024145364761353, + "learning_rate": 5.196010756286649e-05, + "loss": 0.882, + "step": 14827 + }, + { + "epoch": 2.639957264957265, + "grad_norm": 0.73011714220047, + "learning_rate": 5.1947831549197504e-05, + "loss": 0.865, + "step": 14828 + }, + { + "epoch": 2.6401353276353277, + "grad_norm": 0.845160186290741, + "learning_rate": 5.1935556477011006e-05, + "loss": 0.8912, + "step": 14829 + }, + { + "epoch": 2.6403133903133904, + "grad_norm": 0.8264908194541931, + "learning_rate": 5.192328234654735e-05, + "loss": 0.8597, + "step": 14830 + }, + { + "epoch": 2.640491452991453, + "grad_norm": 0.8400609493255615, + "learning_rate": 5.191100915804718e-05, + "loss": 0.5906, + "step": 14831 + }, + { + "epoch": 2.640669515669516, + "grad_norm": 0.8633815050125122, + "learning_rate": 5.189873691175082e-05, + "loss": 1.031, + "step": 14832 + }, + { + "epoch": 2.640847578347578, + "grad_norm": 0.9047896862030029, + "learning_rate": 5.188646560789884e-05, + "loss": 0.6929, + "step": 14833 + }, + { + "epoch": 2.641025641025641, + "grad_norm": 0.8293144106864929, + "learning_rate": 5.18741952467316e-05, + "loss": 0.9608, + "step": 14834 + }, + { + "epoch": 2.6412037037037037, + "grad_norm": 0.7980968356132507, + "learning_rate": 5.186192582848955e-05, + "loss": 0.6021, + "step": 14835 + }, + { + "epoch": 2.6413817663817665, + "grad_norm": 0.7945372462272644, + "learning_rate": 5.184965735341305e-05, + "loss": 0.7069, + "step": 14836 + }, + { + "epoch": 2.6415598290598292, + "grad_norm": 0.8388827443122864, + "learning_rate": 5.183738982174246e-05, + "loss": 1.2404, + "step": 14837 + }, + { + "epoch": 2.6417378917378915, + "grad_norm": 0.8332177400588989, + "learning_rate": 5.18251232337182e-05, + "loss": 0.9353, + "step": 14838 + }, + { + "epoch": 2.6419159544159543, + "grad_norm": 0.9658130407333374, + "learning_rate": 5.1812857589580565e-05, + "loss": 0.8, + "step": 14839 + }, + { + "epoch": 2.642094017094017, + "grad_norm": 0.9074252247810364, + "learning_rate": 5.180059288956991e-05, + "loss": 0.7567, + "step": 14840 + }, + { + "epoch": 2.64227207977208, + "grad_norm": 0.8543582558631897, + "learning_rate": 5.178832913392649e-05, + "loss": 0.9754, + "step": 14841 + }, + { + "epoch": 2.6424501424501425, + "grad_norm": 0.8235877156257629, + "learning_rate": 5.177606632289063e-05, + "loss": 0.825, + "step": 14842 + }, + { + "epoch": 2.6426282051282053, + "grad_norm": 0.8550012111663818, + "learning_rate": 5.1763804456702545e-05, + "loss": 1.0286, + "step": 14843 + }, + { + "epoch": 2.642806267806268, + "grad_norm": 0.8879600763320923, + "learning_rate": 5.175154353560254e-05, + "loss": 0.8935, + "step": 14844 + }, + { + "epoch": 2.6429843304843303, + "grad_norm": 0.8822683095932007, + "learning_rate": 5.1739283559830754e-05, + "loss": 0.8659, + "step": 14845 + }, + { + "epoch": 2.643162393162393, + "grad_norm": 1.0260087251663208, + "learning_rate": 5.1727024529627544e-05, + "loss": 0.8952, + "step": 14846 + }, + { + "epoch": 2.643340455840456, + "grad_norm": 0.8105470538139343, + "learning_rate": 5.171476644523292e-05, + "loss": 0.7987, + "step": 14847 + }, + { + "epoch": 2.6435185185185186, + "grad_norm": 0.8861166834831238, + "learning_rate": 5.170250930688719e-05, + "loss": 0.8476, + "step": 14848 + }, + { + "epoch": 2.6436965811965814, + "grad_norm": 0.8035899996757507, + "learning_rate": 5.169025311483047e-05, + "loss": 0.7366, + "step": 14849 + }, + { + "epoch": 2.6438746438746437, + "grad_norm": 0.8359752297401428, + "learning_rate": 5.1677997869302874e-05, + "loss": 0.8931, + "step": 14850 + }, + { + "epoch": 2.6440527065527064, + "grad_norm": 0.8483668565750122, + "learning_rate": 5.166574357054452e-05, + "loss": 0.7662, + "step": 14851 + }, + { + "epoch": 2.644230769230769, + "grad_norm": 0.9865937829017639, + "learning_rate": 5.165349021879553e-05, + "loss": 1.101, + "step": 14852 + }, + { + "epoch": 2.644408831908832, + "grad_norm": 0.8491073250770569, + "learning_rate": 5.164123781429596e-05, + "loss": 0.9576, + "step": 14853 + }, + { + "epoch": 2.6445868945868947, + "grad_norm": 0.8185597062110901, + "learning_rate": 5.162898635728588e-05, + "loss": 0.6353, + "step": 14854 + }, + { + "epoch": 2.6447649572649574, + "grad_norm": 0.8583887815475464, + "learning_rate": 5.1616735848005306e-05, + "loss": 0.8715, + "step": 14855 + }, + { + "epoch": 2.64494301994302, + "grad_norm": 0.8107531666755676, + "learning_rate": 5.16044862866943e-05, + "loss": 0.7111, + "step": 14856 + }, + { + "epoch": 2.6451210826210825, + "grad_norm": 0.7675925493240356, + "learning_rate": 5.1592237673592867e-05, + "loss": 0.8145, + "step": 14857 + }, + { + "epoch": 2.6452991452991452, + "grad_norm": 0.9418326020240784, + "learning_rate": 5.157999000894098e-05, + "loss": 0.8454, + "step": 14858 + }, + { + "epoch": 2.645477207977208, + "grad_norm": 0.8420053720474243, + "learning_rate": 5.15677432929786e-05, + "loss": 0.7343, + "step": 14859 + }, + { + "epoch": 2.6456552706552707, + "grad_norm": 0.9815202951431274, + "learning_rate": 5.155549752594564e-05, + "loss": 0.9252, + "step": 14860 + }, + { + "epoch": 2.6458333333333335, + "grad_norm": 0.8282185792922974, + "learning_rate": 5.1543252708082146e-05, + "loss": 0.9935, + "step": 14861 + }, + { + "epoch": 2.646011396011396, + "grad_norm": 0.7398781180381775, + "learning_rate": 5.153100883962788e-05, + "loss": 0.5024, + "step": 14862 + }, + { + "epoch": 2.6461894586894585, + "grad_norm": 1.0273998975753784, + "learning_rate": 5.1518765920822856e-05, + "loss": 0.9023, + "step": 14863 + }, + { + "epoch": 2.6463675213675213, + "grad_norm": 0.8017948269844055, + "learning_rate": 5.150652395190689e-05, + "loss": 0.6755, + "step": 14864 + }, + { + "epoch": 2.646545584045584, + "grad_norm": 0.7470258474349976, + "learning_rate": 5.1494282933119864e-05, + "loss": 0.5408, + "step": 14865 + }, + { + "epoch": 2.646723646723647, + "grad_norm": 0.8118627071380615, + "learning_rate": 5.1482042864701595e-05, + "loss": 0.8032, + "step": 14866 + }, + { + "epoch": 2.6469017094017095, + "grad_norm": 0.8302956223487854, + "learning_rate": 5.146980374689192e-05, + "loss": 0.7428, + "step": 14867 + }, + { + "epoch": 2.6470797720797723, + "grad_norm": 0.8660209774971008, + "learning_rate": 5.145756557993061e-05, + "loss": 0.8284, + "step": 14868 + }, + { + "epoch": 2.6472578347578346, + "grad_norm": 1.0153858661651611, + "learning_rate": 5.1445328364057475e-05, + "loss": 0.9766, + "step": 14869 + }, + { + "epoch": 2.6474358974358974, + "grad_norm": 0.9047706127166748, + "learning_rate": 5.143309209951223e-05, + "loss": 1.0099, + "step": 14870 + }, + { + "epoch": 2.64761396011396, + "grad_norm": 0.7924295663833618, + "learning_rate": 5.1420856786534724e-05, + "loss": 0.8385, + "step": 14871 + }, + { + "epoch": 2.647792022792023, + "grad_norm": 0.8885742425918579, + "learning_rate": 5.140862242536455e-05, + "loss": 1.0259, + "step": 14872 + }, + { + "epoch": 2.6479700854700856, + "grad_norm": 0.8826889991760254, + "learning_rate": 5.139638901624151e-05, + "loss": 1.0755, + "step": 14873 + }, + { + "epoch": 2.648148148148148, + "grad_norm": 0.7793754935264587, + "learning_rate": 5.138415655940526e-05, + "loss": 0.8289, + "step": 14874 + }, + { + "epoch": 2.6483262108262107, + "grad_norm": 0.8587870597839355, + "learning_rate": 5.137192505509547e-05, + "loss": 0.9185, + "step": 14875 + }, + { + "epoch": 2.6485042735042734, + "grad_norm": 0.8799259066581726, + "learning_rate": 5.13596945035518e-05, + "loss": 0.9493, + "step": 14876 + }, + { + "epoch": 2.648682336182336, + "grad_norm": 0.8108882308006287, + "learning_rate": 5.1347464905013834e-05, + "loss": 0.8892, + "step": 14877 + }, + { + "epoch": 2.648860398860399, + "grad_norm": 0.8387644290924072, + "learning_rate": 5.1335236259721296e-05, + "loss": 0.8723, + "step": 14878 + }, + { + "epoch": 2.6490384615384617, + "grad_norm": 0.8750926852226257, + "learning_rate": 5.1323008567913655e-05, + "loss": 0.6978, + "step": 14879 + }, + { + "epoch": 2.6492165242165244, + "grad_norm": 0.7837518453598022, + "learning_rate": 5.131078182983055e-05, + "loss": 0.8747, + "step": 14880 + }, + { + "epoch": 2.6493945868945867, + "grad_norm": 0.8998439311981201, + "learning_rate": 5.1298556045711566e-05, + "loss": 0.9903, + "step": 14881 + }, + { + "epoch": 2.6495726495726495, + "grad_norm": 0.8012915253639221, + "learning_rate": 5.128633121579619e-05, + "loss": 0.9767, + "step": 14882 + }, + { + "epoch": 2.6497507122507122, + "grad_norm": 0.9051218032836914, + "learning_rate": 5.1274107340323964e-05, + "loss": 0.7454, + "step": 14883 + }, + { + "epoch": 2.649928774928775, + "grad_norm": 0.8373401761054993, + "learning_rate": 5.1261884419534376e-05, + "loss": 0.821, + "step": 14884 + }, + { + "epoch": 2.6501068376068377, + "grad_norm": 0.7482876181602478, + "learning_rate": 5.124966245366689e-05, + "loss": 0.7051, + "step": 14885 + }, + { + "epoch": 2.6502849002849, + "grad_norm": 0.8445764183998108, + "learning_rate": 5.1237441442961074e-05, + "loss": 0.7416, + "step": 14886 + }, + { + "epoch": 2.650462962962963, + "grad_norm": 0.887598991394043, + "learning_rate": 5.122522138765622e-05, + "loss": 0.9027, + "step": 14887 + }, + { + "epoch": 2.6506410256410255, + "grad_norm": 0.8089238405227661, + "learning_rate": 5.1213002287991905e-05, + "loss": 0.9294, + "step": 14888 + }, + { + "epoch": 2.6508190883190883, + "grad_norm": 0.8614209890365601, + "learning_rate": 5.120078414420739e-05, + "loss": 0.7716, + "step": 14889 + }, + { + "epoch": 2.650997150997151, + "grad_norm": 0.6805269718170166, + "learning_rate": 5.118856695654217e-05, + "loss": 0.6183, + "step": 14890 + }, + { + "epoch": 2.651175213675214, + "grad_norm": 0.9024596214294434, + "learning_rate": 5.117635072523559e-05, + "loss": 0.9516, + "step": 14891 + }, + { + "epoch": 2.6513532763532766, + "grad_norm": 0.906373143196106, + "learning_rate": 5.116413545052701e-05, + "loss": 0.7522, + "step": 14892 + }, + { + "epoch": 2.6515313390313393, + "grad_norm": 0.827235996723175, + "learning_rate": 5.1151921132655725e-05, + "loss": 0.6776, + "step": 14893 + }, + { + "epoch": 2.6517094017094016, + "grad_norm": 0.7769291996955872, + "learning_rate": 5.113970777186108e-05, + "loss": 0.6682, + "step": 14894 + }, + { + "epoch": 2.6518874643874644, + "grad_norm": 0.8420324921607971, + "learning_rate": 5.112749536838233e-05, + "loss": 0.8303, + "step": 14895 + }, + { + "epoch": 2.652065527065527, + "grad_norm": 0.789368748664856, + "learning_rate": 5.1115283922458814e-05, + "loss": 0.773, + "step": 14896 + }, + { + "epoch": 2.65224358974359, + "grad_norm": 0.9156190752983093, + "learning_rate": 5.1103073434329766e-05, + "loss": 1.0318, + "step": 14897 + }, + { + "epoch": 2.652421652421652, + "grad_norm": 1.0411027669906616, + "learning_rate": 5.109086390423441e-05, + "loss": 0.81, + "step": 14898 + }, + { + "epoch": 2.652599715099715, + "grad_norm": 0.9908538460731506, + "learning_rate": 5.107865533241198e-05, + "loss": 0.9386, + "step": 14899 + }, + { + "epoch": 2.6527777777777777, + "grad_norm": 0.7364035844802856, + "learning_rate": 5.106644771910165e-05, + "loss": 0.675, + "step": 14900 + }, + { + "epoch": 2.6529558404558404, + "grad_norm": 0.8409245014190674, + "learning_rate": 5.1054241064542686e-05, + "loss": 0.9446, + "step": 14901 + }, + { + "epoch": 2.653133903133903, + "grad_norm": 0.7731066942214966, + "learning_rate": 5.104203536897412e-05, + "loss": 0.4684, + "step": 14902 + }, + { + "epoch": 2.653311965811966, + "grad_norm": 0.9114529490470886, + "learning_rate": 5.102983063263525e-05, + "loss": 0.9551, + "step": 14903 + }, + { + "epoch": 2.6534900284900287, + "grad_norm": 0.7949321269989014, + "learning_rate": 5.101762685576503e-05, + "loss": 0.989, + "step": 14904 + }, + { + "epoch": 2.6536680911680914, + "grad_norm": 0.940191924571991, + "learning_rate": 5.1005424038602724e-05, + "loss": 1.0377, + "step": 14905 + }, + { + "epoch": 2.6538461538461537, + "grad_norm": 0.7629654407501221, + "learning_rate": 5.0993222181387334e-05, + "loss": 0.7908, + "step": 14906 + }, + { + "epoch": 2.6540242165242165, + "grad_norm": 0.9712302684783936, + "learning_rate": 5.098102128435797e-05, + "loss": 1.1486, + "step": 14907 + }, + { + "epoch": 2.6542022792022792, + "grad_norm": 0.9054526686668396, + "learning_rate": 5.096882134775365e-05, + "loss": 0.8078, + "step": 14908 + }, + { + "epoch": 2.654380341880342, + "grad_norm": 0.824647068977356, + "learning_rate": 5.095662237181343e-05, + "loss": 0.9095, + "step": 14909 + }, + { + "epoch": 2.6545584045584043, + "grad_norm": 0.8760488033294678, + "learning_rate": 5.0944424356776287e-05, + "loss": 0.8538, + "step": 14910 + }, + { + "epoch": 2.654736467236467, + "grad_norm": 0.8012890219688416, + "learning_rate": 5.093222730288131e-05, + "loss": 0.7972, + "step": 14911 + }, + { + "epoch": 2.65491452991453, + "grad_norm": 0.9025147557258606, + "learning_rate": 5.0920031210367326e-05, + "loss": 0.8485, + "step": 14912 + }, + { + "epoch": 2.6550925925925926, + "grad_norm": 0.8621100783348083, + "learning_rate": 5.090783607947347e-05, + "loss": 1.1856, + "step": 14913 + }, + { + "epoch": 2.6552706552706553, + "grad_norm": 0.7914317846298218, + "learning_rate": 5.08956419104385e-05, + "loss": 0.78, + "step": 14914 + }, + { + "epoch": 2.655448717948718, + "grad_norm": 0.8691070675849915, + "learning_rate": 5.088344870350146e-05, + "loss": 0.8406, + "step": 14915 + }, + { + "epoch": 2.655626780626781, + "grad_norm": 0.8521141409873962, + "learning_rate": 5.087125645890121e-05, + "loss": 1.0077, + "step": 14916 + }, + { + "epoch": 2.6558048433048436, + "grad_norm": 0.7918437123298645, + "learning_rate": 5.08590651768766e-05, + "loss": 0.8367, + "step": 14917 + }, + { + "epoch": 2.655982905982906, + "grad_norm": 0.8580697178840637, + "learning_rate": 5.084687485766659e-05, + "loss": 0.921, + "step": 14918 + }, + { + "epoch": 2.6561609686609686, + "grad_norm": 0.7943900227546692, + "learning_rate": 5.0834685501509894e-05, + "loss": 0.7934, + "step": 14919 + }, + { + "epoch": 2.6563390313390314, + "grad_norm": 0.7467655539512634, + "learning_rate": 5.082249710864544e-05, + "loss": 0.8625, + "step": 14920 + }, + { + "epoch": 2.656517094017094, + "grad_norm": 0.7654036283493042, + "learning_rate": 5.0810309679311996e-05, + "loss": 0.888, + "step": 14921 + }, + { + "epoch": 2.6566951566951564, + "grad_norm": 0.8428319692611694, + "learning_rate": 5.079812321374836e-05, + "loss": 0.858, + "step": 14922 + }, + { + "epoch": 2.656873219373219, + "grad_norm": 0.8273693323135376, + "learning_rate": 5.078593771219329e-05, + "loss": 0.8982, + "step": 14923 + }, + { + "epoch": 2.657051282051282, + "grad_norm": 0.9037185311317444, + "learning_rate": 5.077375317488553e-05, + "loss": 0.7022, + "step": 14924 + }, + { + "epoch": 2.6572293447293447, + "grad_norm": 0.916585385799408, + "learning_rate": 5.0761569602063816e-05, + "loss": 0.8058, + "step": 14925 + }, + { + "epoch": 2.6574074074074074, + "grad_norm": 0.8697561621665955, + "learning_rate": 5.074938699396687e-05, + "loss": 0.8142, + "step": 14926 + }, + { + "epoch": 2.65758547008547, + "grad_norm": 1.024512529373169, + "learning_rate": 5.073720535083334e-05, + "loss": 0.7462, + "step": 14927 + }, + { + "epoch": 2.657763532763533, + "grad_norm": 0.8258776664733887, + "learning_rate": 5.072502467290201e-05, + "loss": 0.7467, + "step": 14928 + }, + { + "epoch": 2.6579415954415957, + "grad_norm": 0.8279047012329102, + "learning_rate": 5.071284496041138e-05, + "loss": 0.9148, + "step": 14929 + }, + { + "epoch": 2.658119658119658, + "grad_norm": 0.8176717758178711, + "learning_rate": 5.070066621360021e-05, + "loss": 1.0971, + "step": 14930 + }, + { + "epoch": 2.6582977207977208, + "grad_norm": 0.7482925057411194, + "learning_rate": 5.0688488432707074e-05, + "loss": 0.8666, + "step": 14931 + }, + { + "epoch": 2.6584757834757835, + "grad_norm": 0.9302734136581421, + "learning_rate": 5.067631161797057e-05, + "loss": 0.9994, + "step": 14932 + }, + { + "epoch": 2.6586538461538463, + "grad_norm": 0.7811494469642639, + "learning_rate": 5.066413576962927e-05, + "loss": 0.5959, + "step": 14933 + }, + { + "epoch": 2.6588319088319086, + "grad_norm": 0.8109773993492126, + "learning_rate": 5.065196088792177e-05, + "loss": 0.7342, + "step": 14934 + }, + { + "epoch": 2.6590099715099713, + "grad_norm": 0.8351961374282837, + "learning_rate": 5.0639786973086525e-05, + "loss": 0.775, + "step": 14935 + }, + { + "epoch": 2.659188034188034, + "grad_norm": 0.8558792471885681, + "learning_rate": 5.062761402536216e-05, + "loss": 0.8819, + "step": 14936 + }, + { + "epoch": 2.659366096866097, + "grad_norm": 0.7928652167320251, + "learning_rate": 5.061544204498714e-05, + "loss": 0.8313, + "step": 14937 + }, + { + "epoch": 2.6595441595441596, + "grad_norm": 0.8388734459877014, + "learning_rate": 5.060327103219993e-05, + "loss": 0.7208, + "step": 14938 + }, + { + "epoch": 2.6597222222222223, + "grad_norm": 0.8921391367912292, + "learning_rate": 5.059110098723903e-05, + "loss": 0.8974, + "step": 14939 + }, + { + "epoch": 2.659900284900285, + "grad_norm": 0.8111342787742615, + "learning_rate": 5.057893191034286e-05, + "loss": 0.6879, + "step": 14940 + }, + { + "epoch": 2.660078347578348, + "grad_norm": 0.8677322864532471, + "learning_rate": 5.056676380174985e-05, + "loss": 0.8643, + "step": 14941 + }, + { + "epoch": 2.66025641025641, + "grad_norm": 0.7969355583190918, + "learning_rate": 5.055459666169839e-05, + "loss": 0.8462, + "step": 14942 + }, + { + "epoch": 2.660434472934473, + "grad_norm": 0.9927026629447937, + "learning_rate": 5.0542430490426975e-05, + "loss": 0.7954, + "step": 14943 + }, + { + "epoch": 2.6606125356125356, + "grad_norm": 1.0181084871292114, + "learning_rate": 5.053026528817379e-05, + "loss": 0.9597, + "step": 14944 + }, + { + "epoch": 2.6607905982905984, + "grad_norm": 1.0274122953414917, + "learning_rate": 5.0518101055177355e-05, + "loss": 0.7321, + "step": 14945 + }, + { + "epoch": 2.6609686609686607, + "grad_norm": 1.056132197380066, + "learning_rate": 5.050593779167594e-05, + "loss": 0.8405, + "step": 14946 + }, + { + "epoch": 2.6611467236467234, + "grad_norm": 0.8586339950561523, + "learning_rate": 5.0493775497907846e-05, + "loss": 1.0238, + "step": 14947 + }, + { + "epoch": 2.661324786324786, + "grad_norm": 0.8103144764900208, + "learning_rate": 5.048161417411139e-05, + "loss": 0.5885, + "step": 14948 + }, + { + "epoch": 2.661502849002849, + "grad_norm": 0.7321345210075378, + "learning_rate": 5.0469453820524834e-05, + "loss": 0.7987, + "step": 14949 + }, + { + "epoch": 2.6616809116809117, + "grad_norm": 0.8244233727455139, + "learning_rate": 5.045729443738645e-05, + "loss": 0.8855, + "step": 14950 + }, + { + "epoch": 2.6618589743589745, + "grad_norm": 0.7888374924659729, + "learning_rate": 5.0445136024934456e-05, + "loss": 0.9192, + "step": 14951 + }, + { + "epoch": 2.662037037037037, + "grad_norm": 0.8414669036865234, + "learning_rate": 5.0432978583407044e-05, + "loss": 0.8152, + "step": 14952 + }, + { + "epoch": 2.6622150997151, + "grad_norm": 0.9176363348960876, + "learning_rate": 5.042082211304252e-05, + "loss": 0.8836, + "step": 14953 + }, + { + "epoch": 2.6623931623931623, + "grad_norm": 0.9827163219451904, + "learning_rate": 5.040866661407893e-05, + "loss": 0.963, + "step": 14954 + }, + { + "epoch": 2.662571225071225, + "grad_norm": 0.8765084743499756, + "learning_rate": 5.0396512086754535e-05, + "loss": 1.022, + "step": 14955 + }, + { + "epoch": 2.6627492877492878, + "grad_norm": 0.9236209392547607, + "learning_rate": 5.038435853130743e-05, + "loss": 0.9152, + "step": 14956 + }, + { + "epoch": 2.6629273504273505, + "grad_norm": 0.8300418853759766, + "learning_rate": 5.037220594797574e-05, + "loss": 0.8063, + "step": 14957 + }, + { + "epoch": 2.6631054131054133, + "grad_norm": 0.9248050451278687, + "learning_rate": 5.036005433699764e-05, + "loss": 0.8799, + "step": 14958 + }, + { + "epoch": 2.6632834757834756, + "grad_norm": 0.9670597910881042, + "learning_rate": 5.0347903698611085e-05, + "loss": 0.9068, + "step": 14959 + }, + { + "epoch": 2.6634615384615383, + "grad_norm": 0.851403534412384, + "learning_rate": 5.033575403305428e-05, + "loss": 0.8058, + "step": 14960 + }, + { + "epoch": 2.663639601139601, + "grad_norm": 0.9643952250480652, + "learning_rate": 5.032360534056515e-05, + "loss": 1.076, + "step": 14961 + }, + { + "epoch": 2.663817663817664, + "grad_norm": 0.8473731279373169, + "learning_rate": 5.031145762138181e-05, + "loss": 0.9585, + "step": 14962 + }, + { + "epoch": 2.6639957264957266, + "grad_norm": 0.8265015482902527, + "learning_rate": 5.029931087574222e-05, + "loss": 0.8602, + "step": 14963 + }, + { + "epoch": 2.6641737891737893, + "grad_norm": 0.8004183173179626, + "learning_rate": 5.0287165103884416e-05, + "loss": 0.7293, + "step": 14964 + }, + { + "epoch": 2.664351851851852, + "grad_norm": 0.8410465121269226, + "learning_rate": 5.027502030604633e-05, + "loss": 0.9479, + "step": 14965 + }, + { + "epoch": 2.6645299145299144, + "grad_norm": 0.8365132808685303, + "learning_rate": 5.0262876482465925e-05, + "loss": 0.7373, + "step": 14966 + }, + { + "epoch": 2.664707977207977, + "grad_norm": 0.9017055630683899, + "learning_rate": 5.025073363338111e-05, + "loss": 0.9463, + "step": 14967 + }, + { + "epoch": 2.66488603988604, + "grad_norm": 0.7985300421714783, + "learning_rate": 5.023859175902988e-05, + "loss": 0.7074, + "step": 14968 + }, + { + "epoch": 2.6650641025641026, + "grad_norm": 0.8032601475715637, + "learning_rate": 5.022645085965001e-05, + "loss": 0.6796, + "step": 14969 + }, + { + "epoch": 2.6652421652421654, + "grad_norm": 0.7785899639129639, + "learning_rate": 5.021431093547948e-05, + "loss": 0.7256, + "step": 14970 + }, + { + "epoch": 2.6654202279202277, + "grad_norm": 0.8083044290542603, + "learning_rate": 5.02021719867561e-05, + "loss": 0.9254, + "step": 14971 + }, + { + "epoch": 2.6655982905982905, + "grad_norm": 0.8896783590316772, + "learning_rate": 5.019003401371771e-05, + "loss": 0.9231, + "step": 14972 + }, + { + "epoch": 2.665776353276353, + "grad_norm": 0.9304720163345337, + "learning_rate": 5.017789701660215e-05, + "loss": 0.8915, + "step": 14973 + }, + { + "epoch": 2.665954415954416, + "grad_norm": 0.8683121204376221, + "learning_rate": 5.016576099564718e-05, + "loss": 0.8654, + "step": 14974 + }, + { + "epoch": 2.6661324786324787, + "grad_norm": 1.1082890033721924, + "learning_rate": 5.015362595109062e-05, + "loss": 1.0669, + "step": 14975 + }, + { + "epoch": 2.6663105413105415, + "grad_norm": 1.1696041822433472, + "learning_rate": 5.014149188317017e-05, + "loss": 0.9273, + "step": 14976 + }, + { + "epoch": 2.666488603988604, + "grad_norm": 0.8726202845573425, + "learning_rate": 5.0129358792123637e-05, + "loss": 0.6615, + "step": 14977 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.8246448636054993, + "learning_rate": 5.011722667818875e-05, + "loss": 0.8263, + "step": 14978 + }, + { + "epoch": 2.6668447293447293, + "grad_norm": 0.7201130390167236, + "learning_rate": 5.010509554160316e-05, + "loss": 0.7122, + "step": 14979 + }, + { + "epoch": 2.667022792022792, + "grad_norm": 0.8296586275100708, + "learning_rate": 5.009296538260457e-05, + "loss": 0.8816, + "step": 14980 + }, + { + "epoch": 2.6672008547008548, + "grad_norm": 0.8647085428237915, + "learning_rate": 5.008083620143067e-05, + "loss": 1.0, + "step": 14981 + }, + { + "epoch": 2.6673789173789175, + "grad_norm": 0.8175796270370483, + "learning_rate": 5.0068707998319045e-05, + "loss": 0.6727, + "step": 14982 + }, + { + "epoch": 2.66755698005698, + "grad_norm": 0.8537090420722961, + "learning_rate": 5.0056580773507434e-05, + "loss": 0.8034, + "step": 14983 + }, + { + "epoch": 2.6677350427350426, + "grad_norm": 0.7980232238769531, + "learning_rate": 5.00444545272333e-05, + "loss": 0.83, + "step": 14984 + }, + { + "epoch": 2.6679131054131053, + "grad_norm": 0.8231784701347351, + "learning_rate": 5.003232925973438e-05, + "loss": 0.6292, + "step": 14985 + }, + { + "epoch": 2.668091168091168, + "grad_norm": 0.9140519499778748, + "learning_rate": 5.0020204971248096e-05, + "loss": 0.893, + "step": 14986 + }, + { + "epoch": 2.668269230769231, + "grad_norm": 0.7462875247001648, + "learning_rate": 5.000808166201212e-05, + "loss": 0.7335, + "step": 14987 + }, + { + "epoch": 2.6684472934472936, + "grad_norm": 0.8201214671134949, + "learning_rate": 4.999595933226392e-05, + "loss": 0.7888, + "step": 14988 + }, + { + "epoch": 2.6686253561253563, + "grad_norm": 0.9165699481964111, + "learning_rate": 4.9983837982241024e-05, + "loss": 0.8808, + "step": 14989 + }, + { + "epoch": 2.6688034188034186, + "grad_norm": 0.9286229610443115, + "learning_rate": 4.997171761218092e-05, + "loss": 0.969, + "step": 14990 + }, + { + "epoch": 2.6689814814814814, + "grad_norm": 0.6710283160209656, + "learning_rate": 4.995959822232109e-05, + "loss": 0.6046, + "step": 14991 + }, + { + "epoch": 2.669159544159544, + "grad_norm": 0.9091618061065674, + "learning_rate": 4.994747981289895e-05, + "loss": 0.9747, + "step": 14992 + }, + { + "epoch": 2.669337606837607, + "grad_norm": 0.7992748618125916, + "learning_rate": 4.993536238415204e-05, + "loss": 0.8441, + "step": 14993 + }, + { + "epoch": 2.6695156695156697, + "grad_norm": 0.926811695098877, + "learning_rate": 4.992324593631762e-05, + "loss": 1.0308, + "step": 14994 + }, + { + "epoch": 2.669693732193732, + "grad_norm": 0.8966291546821594, + "learning_rate": 4.9911130469633216e-05, + "loss": 1.0689, + "step": 14995 + }, + { + "epoch": 2.6698717948717947, + "grad_norm": 0.8300046324729919, + "learning_rate": 4.989901598433616e-05, + "loss": 0.8539, + "step": 14996 + }, + { + "epoch": 2.6700498575498575, + "grad_norm": 0.9567606449127197, + "learning_rate": 4.988690248066381e-05, + "loss": 0.7707, + "step": 14997 + }, + { + "epoch": 2.67022792022792, + "grad_norm": 0.7993598580360413, + "learning_rate": 4.987478995885351e-05, + "loss": 0.9241, + "step": 14998 + }, + { + "epoch": 2.670405982905983, + "grad_norm": 0.9573900103569031, + "learning_rate": 4.986267841914253e-05, + "loss": 0.8051, + "step": 14999 + }, + { + "epoch": 2.6705840455840457, + "grad_norm": 0.8562188148498535, + "learning_rate": 4.985056786176828e-05, + "loss": 0.8818, + "step": 15000 + }, + { + "epoch": 2.6707621082621085, + "grad_norm": 0.7997880578041077, + "learning_rate": 4.983845828696792e-05, + "loss": 0.8749, + "step": 15001 + }, + { + "epoch": 2.6709401709401708, + "grad_norm": 0.8442137837409973, + "learning_rate": 4.982634969497879e-05, + "loss": 1.0239, + "step": 15002 + }, + { + "epoch": 2.6711182336182335, + "grad_norm": 0.775762140750885, + "learning_rate": 4.981424208603812e-05, + "loss": 0.728, + "step": 15003 + }, + { + "epoch": 2.6712962962962963, + "grad_norm": 0.7570006251335144, + "learning_rate": 4.9802135460383126e-05, + "loss": 0.6964, + "step": 15004 + }, + { + "epoch": 2.671474358974359, + "grad_norm": 0.8406931161880493, + "learning_rate": 4.979002981825101e-05, + "loss": 0.783, + "step": 15005 + }, + { + "epoch": 2.671652421652422, + "grad_norm": 0.827357828617096, + "learning_rate": 4.977792515987896e-05, + "loss": 0.9294, + "step": 15006 + }, + { + "epoch": 2.671830484330484, + "grad_norm": 0.9244057536125183, + "learning_rate": 4.9765821485504094e-05, + "loss": 0.8993, + "step": 15007 + }, + { + "epoch": 2.672008547008547, + "grad_norm": 0.7569696307182312, + "learning_rate": 4.975371879536368e-05, + "loss": 0.8305, + "step": 15008 + }, + { + "epoch": 2.6721866096866096, + "grad_norm": 0.8337959051132202, + "learning_rate": 4.9741617089694695e-05, + "loss": 0.8793, + "step": 15009 + }, + { + "epoch": 2.6723646723646723, + "grad_norm": 0.7254770994186401, + "learning_rate": 4.97295163687344e-05, + "loss": 0.9325, + "step": 15010 + }, + { + "epoch": 2.672542735042735, + "grad_norm": 0.7988013029098511, + "learning_rate": 4.971741663271972e-05, + "loss": 0.9787, + "step": 15011 + }, + { + "epoch": 2.672720797720798, + "grad_norm": 0.8326970338821411, + "learning_rate": 4.9705317881887845e-05, + "loss": 0.9164, + "step": 15012 + }, + { + "epoch": 2.6728988603988606, + "grad_norm": 0.7416687607765198, + "learning_rate": 4.96932201164758e-05, + "loss": 0.9041, + "step": 15013 + }, + { + "epoch": 2.6730769230769234, + "grad_norm": 0.868765652179718, + "learning_rate": 4.968112333672059e-05, + "loss": 0.646, + "step": 15014 + }, + { + "epoch": 2.6732549857549857, + "grad_norm": 0.7440044283866882, + "learning_rate": 4.966902754285925e-05, + "loss": 0.9147, + "step": 15015 + }, + { + "epoch": 2.6734330484330484, + "grad_norm": 0.8410077691078186, + "learning_rate": 4.9656932735128724e-05, + "loss": 0.8044, + "step": 15016 + }, + { + "epoch": 2.673611111111111, + "grad_norm": 0.8185286521911621, + "learning_rate": 4.964483891376606e-05, + "loss": 0.8057, + "step": 15017 + }, + { + "epoch": 2.673789173789174, + "grad_norm": 0.8550063967704773, + "learning_rate": 4.9632746079008166e-05, + "loss": 0.7841, + "step": 15018 + }, + { + "epoch": 2.673967236467236, + "grad_norm": 0.9171682000160217, + "learning_rate": 4.962065423109199e-05, + "loss": 0.8731, + "step": 15019 + }, + { + "epoch": 2.674145299145299, + "grad_norm": 0.8567686676979065, + "learning_rate": 4.9608563370254436e-05, + "loss": 0.9284, + "step": 15020 + }, + { + "epoch": 2.6743233618233617, + "grad_norm": 0.8641629219055176, + "learning_rate": 4.959647349673241e-05, + "loss": 1.0165, + "step": 15021 + }, + { + "epoch": 2.6745014245014245, + "grad_norm": 0.8058172464370728, + "learning_rate": 4.958438461076277e-05, + "loss": 0.9737, + "step": 15022 + }, + { + "epoch": 2.6746794871794872, + "grad_norm": 0.8329246640205383, + "learning_rate": 4.95722967125824e-05, + "loss": 0.7943, + "step": 15023 + }, + { + "epoch": 2.67485754985755, + "grad_norm": 0.9603211879730225, + "learning_rate": 4.956020980242807e-05, + "loss": 0.9453, + "step": 15024 + }, + { + "epoch": 2.6750356125356127, + "grad_norm": 0.8635705709457397, + "learning_rate": 4.9548123880536736e-05, + "loss": 0.9028, + "step": 15025 + }, + { + "epoch": 2.6752136752136755, + "grad_norm": 0.8909839987754822, + "learning_rate": 4.9536038947145024e-05, + "loss": 1.0376, + "step": 15026 + }, + { + "epoch": 2.675391737891738, + "grad_norm": 0.7507481575012207, + "learning_rate": 4.952395500248984e-05, + "loss": 0.7151, + "step": 15027 + }, + { + "epoch": 2.6755698005698005, + "grad_norm": 0.9425675272941589, + "learning_rate": 4.951187204680791e-05, + "loss": 0.839, + "step": 15028 + }, + { + "epoch": 2.6757478632478633, + "grad_norm": 0.8826829195022583, + "learning_rate": 4.949979008033596e-05, + "loss": 1.0107, + "step": 15029 + }, + { + "epoch": 2.675925925925926, + "grad_norm": 0.9209766387939453, + "learning_rate": 4.948770910331072e-05, + "loss": 0.8685, + "step": 15030 + }, + { + "epoch": 2.6761039886039883, + "grad_norm": 0.8018497824668884, + "learning_rate": 4.947562911596889e-05, + "loss": 0.7417, + "step": 15031 + }, + { + "epoch": 2.676282051282051, + "grad_norm": 0.7865417003631592, + "learning_rate": 4.9463550118547155e-05, + "loss": 0.9332, + "step": 15032 + }, + { + "epoch": 2.676460113960114, + "grad_norm": 0.8146806955337524, + "learning_rate": 4.945147211128216e-05, + "loss": 0.8658, + "step": 15033 + }, + { + "epoch": 2.6766381766381766, + "grad_norm": 0.8176286816596985, + "learning_rate": 4.943939509441054e-05, + "loss": 1.0603, + "step": 15034 + }, + { + "epoch": 2.6768162393162394, + "grad_norm": 0.8441028594970703, + "learning_rate": 4.942731906816897e-05, + "loss": 0.8699, + "step": 15035 + }, + { + "epoch": 2.676994301994302, + "grad_norm": 1.0035977363586426, + "learning_rate": 4.941524403279405e-05, + "loss": 0.8149, + "step": 15036 + }, + { + "epoch": 2.677172364672365, + "grad_norm": 0.8316586017608643, + "learning_rate": 4.9403169988522324e-05, + "loss": 0.9674, + "step": 15037 + }, + { + "epoch": 2.6773504273504276, + "grad_norm": 0.7379693388938904, + "learning_rate": 4.9391096935590375e-05, + "loss": 0.7097, + "step": 15038 + }, + { + "epoch": 2.67752849002849, + "grad_norm": 0.8861358165740967, + "learning_rate": 4.937902487423473e-05, + "loss": 0.9145, + "step": 15039 + }, + { + "epoch": 2.6777065527065527, + "grad_norm": 0.8769996166229248, + "learning_rate": 4.9366953804691994e-05, + "loss": 0.92, + "step": 15040 + }, + { + "epoch": 2.6778846153846154, + "grad_norm": 0.891703724861145, + "learning_rate": 4.9354883727198545e-05, + "loss": 0.8898, + "step": 15041 + }, + { + "epoch": 2.678062678062678, + "grad_norm": 0.8371208310127258, + "learning_rate": 4.934281464199099e-05, + "loss": 0.8868, + "step": 15042 + }, + { + "epoch": 2.6782407407407405, + "grad_norm": 0.8618297576904297, + "learning_rate": 4.933074654930574e-05, + "loss": 0.8577, + "step": 15043 + }, + { + "epoch": 2.6784188034188032, + "grad_norm": 0.7748361229896545, + "learning_rate": 4.931867944937926e-05, + "loss": 0.7273, + "step": 15044 + }, + { + "epoch": 2.678596866096866, + "grad_norm": 0.8320143222808838, + "learning_rate": 4.930661334244797e-05, + "loss": 0.8654, + "step": 15045 + }, + { + "epoch": 2.6787749287749287, + "grad_norm": 0.8370615243911743, + "learning_rate": 4.929454822874829e-05, + "loss": 0.751, + "step": 15046 + }, + { + "epoch": 2.6789529914529915, + "grad_norm": 0.9115342497825623, + "learning_rate": 4.9282484108516614e-05, + "loss": 0.823, + "step": 15047 + }, + { + "epoch": 2.6791310541310542, + "grad_norm": 0.9542914032936096, + "learning_rate": 4.9270420981989294e-05, + "loss": 0.9271, + "step": 15048 + }, + { + "epoch": 2.679309116809117, + "grad_norm": 0.765336275100708, + "learning_rate": 4.9258358849402655e-05, + "loss": 0.5523, + "step": 15049 + }, + { + "epoch": 2.6794871794871797, + "grad_norm": 0.8169335722923279, + "learning_rate": 4.924629771099315e-05, + "loss": 0.7437, + "step": 15050 + }, + { + "epoch": 2.679665242165242, + "grad_norm": 0.8192304968833923, + "learning_rate": 4.9234237566996935e-05, + "loss": 0.8888, + "step": 15051 + }, + { + "epoch": 2.679843304843305, + "grad_norm": 0.8657594919204712, + "learning_rate": 4.922217841765041e-05, + "loss": 0.9858, + "step": 15052 + }, + { + "epoch": 2.6800213675213675, + "grad_norm": 0.9291723370552063, + "learning_rate": 4.921012026318982e-05, + "loss": 0.9731, + "step": 15053 + }, + { + "epoch": 2.6801994301994303, + "grad_norm": 0.7988953590393066, + "learning_rate": 4.919806310385138e-05, + "loss": 0.8467, + "step": 15054 + }, + { + "epoch": 2.6803774928774926, + "grad_norm": 0.8022913336753845, + "learning_rate": 4.9186006939871434e-05, + "loss": 0.9009, + "step": 15055 + }, + { + "epoch": 2.6805555555555554, + "grad_norm": 0.8444825410842896, + "learning_rate": 4.917395177148605e-05, + "loss": 0.8851, + "step": 15056 + }, + { + "epoch": 2.680733618233618, + "grad_norm": 0.8054760694503784, + "learning_rate": 4.9161897598931575e-05, + "loss": 0.8679, + "step": 15057 + }, + { + "epoch": 2.680911680911681, + "grad_norm": 0.8291507959365845, + "learning_rate": 4.9149844422444023e-05, + "loss": 0.7229, + "step": 15058 + }, + { + "epoch": 2.6810897435897436, + "grad_norm": 0.9225491285324097, + "learning_rate": 4.91377922422597e-05, + "loss": 0.7584, + "step": 15059 + }, + { + "epoch": 2.6812678062678064, + "grad_norm": 0.9598490595817566, + "learning_rate": 4.912574105861466e-05, + "loss": 1.0548, + "step": 15060 + }, + { + "epoch": 2.681445868945869, + "grad_norm": 0.7480899691581726, + "learning_rate": 4.911369087174504e-05, + "loss": 0.8389, + "step": 15061 + }, + { + "epoch": 2.681623931623932, + "grad_norm": 1.0396811962127686, + "learning_rate": 4.910164168188696e-05, + "loss": 0.8776, + "step": 15062 + }, + { + "epoch": 2.681801994301994, + "grad_norm": 0.8191503882408142, + "learning_rate": 4.9089593489276465e-05, + "loss": 0.7601, + "step": 15063 + }, + { + "epoch": 2.681980056980057, + "grad_norm": 0.8405289053916931, + "learning_rate": 4.907754629414959e-05, + "loss": 1.0859, + "step": 15064 + }, + { + "epoch": 2.6821581196581197, + "grad_norm": 0.8369600176811218, + "learning_rate": 4.90655000967425e-05, + "loss": 0.9159, + "step": 15065 + }, + { + "epoch": 2.6823361823361824, + "grad_norm": 0.8304924368858337, + "learning_rate": 4.905345489729104e-05, + "loss": 0.743, + "step": 15066 + }, + { + "epoch": 2.682514245014245, + "grad_norm": 0.7378702163696289, + "learning_rate": 4.904141069603139e-05, + "loss": 0.9386, + "step": 15067 + }, + { + "epoch": 2.6826923076923075, + "grad_norm": 0.9135075807571411, + "learning_rate": 4.902936749319935e-05, + "loss": 0.7341, + "step": 15068 + }, + { + "epoch": 2.6828703703703702, + "grad_norm": 0.77586430311203, + "learning_rate": 4.901732528903101e-05, + "loss": 0.5586, + "step": 15069 + }, + { + "epoch": 2.683048433048433, + "grad_norm": 0.8733307719230652, + "learning_rate": 4.900528408376228e-05, + "loss": 0.8173, + "step": 15070 + }, + { + "epoch": 2.6832264957264957, + "grad_norm": 0.7499578595161438, + "learning_rate": 4.8993243877629066e-05, + "loss": 0.7355, + "step": 15071 + }, + { + "epoch": 2.6834045584045585, + "grad_norm": 0.8372282385826111, + "learning_rate": 4.8981204670867295e-05, + "loss": 0.8169, + "step": 15072 + }, + { + "epoch": 2.6835826210826212, + "grad_norm": 0.7705212235450745, + "learning_rate": 4.8969166463712834e-05, + "loss": 0.7382, + "step": 15073 + }, + { + "epoch": 2.683760683760684, + "grad_norm": 0.8367058038711548, + "learning_rate": 4.89571292564015e-05, + "loss": 0.7268, + "step": 15074 + }, + { + "epoch": 2.6839387464387463, + "grad_norm": 0.8421934843063354, + "learning_rate": 4.8945093049169233e-05, + "loss": 0.8319, + "step": 15075 + }, + { + "epoch": 2.684116809116809, + "grad_norm": 0.8927276730537415, + "learning_rate": 4.893305784225181e-05, + "loss": 0.8669, + "step": 15076 + }, + { + "epoch": 2.684294871794872, + "grad_norm": 0.8147335052490234, + "learning_rate": 4.892102363588503e-05, + "loss": 0.7722, + "step": 15077 + }, + { + "epoch": 2.6844729344729346, + "grad_norm": 0.9491320848464966, + "learning_rate": 4.890899043030469e-05, + "loss": 1.0213, + "step": 15078 + }, + { + "epoch": 2.6846509971509973, + "grad_norm": 0.8635398745536804, + "learning_rate": 4.889695822574651e-05, + "loss": 0.797, + "step": 15079 + }, + { + "epoch": 2.6848290598290596, + "grad_norm": 0.7290985584259033, + "learning_rate": 4.888492702244636e-05, + "loss": 1.0142, + "step": 15080 + }, + { + "epoch": 2.6850071225071224, + "grad_norm": 0.7667058110237122, + "learning_rate": 4.8872896820639794e-05, + "loss": 0.7547, + "step": 15081 + }, + { + "epoch": 2.685185185185185, + "grad_norm": 0.9096128344535828, + "learning_rate": 4.886086762056269e-05, + "loss": 0.7972, + "step": 15082 + }, + { + "epoch": 2.685363247863248, + "grad_norm": 0.7461803555488586, + "learning_rate": 4.884883942245057e-05, + "loss": 0.8994, + "step": 15083 + }, + { + "epoch": 2.6855413105413106, + "grad_norm": 0.7640016674995422, + "learning_rate": 4.883681222653923e-05, + "loss": 0.7607, + "step": 15084 + }, + { + "epoch": 2.6857193732193734, + "grad_norm": 0.7481253743171692, + "learning_rate": 4.882478603306427e-05, + "loss": 0.7089, + "step": 15085 + }, + { + "epoch": 2.685897435897436, + "grad_norm": 0.825998842716217, + "learning_rate": 4.881276084226132e-05, + "loss": 0.6617, + "step": 15086 + }, + { + "epoch": 2.6860754985754984, + "grad_norm": 0.9775291085243225, + "learning_rate": 4.8800736654365986e-05, + "loss": 0.9345, + "step": 15087 + }, + { + "epoch": 2.686253561253561, + "grad_norm": 0.8158339262008667, + "learning_rate": 4.878871346961387e-05, + "loss": 0.8198, + "step": 15088 + }, + { + "epoch": 2.686431623931624, + "grad_norm": 0.8778133988380432, + "learning_rate": 4.8776691288240486e-05, + "loss": 0.8323, + "step": 15089 + }, + { + "epoch": 2.6866096866096867, + "grad_norm": 0.9657309055328369, + "learning_rate": 4.8764670110481505e-05, + "loss": 0.907, + "step": 15090 + }, + { + "epoch": 2.6867877492877494, + "grad_norm": 1.0467438697814941, + "learning_rate": 4.8752649936572304e-05, + "loss": 1.0128, + "step": 15091 + }, + { + "epoch": 2.6869658119658117, + "grad_norm": 0.7682142853736877, + "learning_rate": 4.874063076674854e-05, + "loss": 1.0164, + "step": 15092 + }, + { + "epoch": 2.6871438746438745, + "grad_norm": 0.8184331059455872, + "learning_rate": 4.8728612601245574e-05, + "loss": 0.6614, + "step": 15093 + }, + { + "epoch": 2.6873219373219372, + "grad_norm": 0.8372936248779297, + "learning_rate": 4.871659544029896e-05, + "loss": 0.9011, + "step": 15094 + }, + { + "epoch": 2.6875, + "grad_norm": 0.7872710824012756, + "learning_rate": 4.870457928414414e-05, + "loss": 0.6986, + "step": 15095 + }, + { + "epoch": 2.6876780626780628, + "grad_norm": 0.7297250628471375, + "learning_rate": 4.8692564133016485e-05, + "loss": 0.5399, + "step": 15096 + }, + { + "epoch": 2.6878561253561255, + "grad_norm": 0.8855645060539246, + "learning_rate": 4.868054998715153e-05, + "loss": 0.8992, + "step": 15097 + }, + { + "epoch": 2.6880341880341883, + "grad_norm": 0.9055765271186829, + "learning_rate": 4.866853684678452e-05, + "loss": 0.888, + "step": 15098 + }, + { + "epoch": 2.6882122507122506, + "grad_norm": 1.0414996147155762, + "learning_rate": 4.865652471215093e-05, + "loss": 1.0375, + "step": 15099 + }, + { + "epoch": 2.6883903133903133, + "grad_norm": 0.8606446385383606, + "learning_rate": 4.8644513583486086e-05, + "loss": 0.8906, + "step": 15100 + }, + { + "epoch": 2.688568376068376, + "grad_norm": 0.9065528512001038, + "learning_rate": 4.8632503461025316e-05, + "loss": 0.848, + "step": 15101 + }, + { + "epoch": 2.688746438746439, + "grad_norm": 0.7832834720611572, + "learning_rate": 4.862049434500393e-05, + "loss": 0.7028, + "step": 15102 + }, + { + "epoch": 2.6889245014245016, + "grad_norm": 0.7107385396957397, + "learning_rate": 4.860848623565723e-05, + "loss": 0.7249, + "step": 15103 + }, + { + "epoch": 2.689102564102564, + "grad_norm": 0.8936449289321899, + "learning_rate": 4.8596479133220485e-05, + "loss": 0.9651, + "step": 15104 + }, + { + "epoch": 2.6892806267806266, + "grad_norm": 0.9019163846969604, + "learning_rate": 4.8584473037928944e-05, + "loss": 0.7165, + "step": 15105 + }, + { + "epoch": 2.6894586894586894, + "grad_norm": 0.8838223218917847, + "learning_rate": 4.857246795001782e-05, + "loss": 0.8148, + "step": 15106 + }, + { + "epoch": 2.689636752136752, + "grad_norm": 0.8004612922668457, + "learning_rate": 4.856046386972243e-05, + "loss": 0.9109, + "step": 15107 + }, + { + "epoch": 2.689814814814815, + "grad_norm": 0.9337486028671265, + "learning_rate": 4.854846079727781e-05, + "loss": 1.0952, + "step": 15108 + }, + { + "epoch": 2.6899928774928776, + "grad_norm": 0.6513102650642395, + "learning_rate": 4.853645873291926e-05, + "loss": 0.5435, + "step": 15109 + }, + { + "epoch": 2.6901709401709404, + "grad_norm": 0.8750485181808472, + "learning_rate": 4.85244576768819e-05, + "loss": 0.8783, + "step": 15110 + }, + { + "epoch": 2.6903490028490027, + "grad_norm": 0.9513342976570129, + "learning_rate": 4.851245762940085e-05, + "loss": 0.8822, + "step": 15111 + }, + { + "epoch": 2.6905270655270654, + "grad_norm": 0.8832191824913025, + "learning_rate": 4.850045859071125e-05, + "loss": 0.9216, + "step": 15112 + }, + { + "epoch": 2.690705128205128, + "grad_norm": 0.875396728515625, + "learning_rate": 4.8488460561048175e-05, + "loss": 0.998, + "step": 15113 + }, + { + "epoch": 2.690883190883191, + "grad_norm": 0.8847890496253967, + "learning_rate": 4.847646354064668e-05, + "loss": 1.0916, + "step": 15114 + }, + { + "epoch": 2.6910612535612537, + "grad_norm": 0.8235226273536682, + "learning_rate": 4.846446752974187e-05, + "loss": 0.8154, + "step": 15115 + }, + { + "epoch": 2.691239316239316, + "grad_norm": 0.8099366426467896, + "learning_rate": 4.845247252856878e-05, + "loss": 0.9392, + "step": 15116 + }, + { + "epoch": 2.6914173789173788, + "grad_norm": 0.8525599837303162, + "learning_rate": 4.84404785373624e-05, + "loss": 0.6619, + "step": 15117 + }, + { + "epoch": 2.6915954415954415, + "grad_norm": 1.0223274230957031, + "learning_rate": 4.842848555635775e-05, + "loss": 0.9479, + "step": 15118 + }, + { + "epoch": 2.6917735042735043, + "grad_norm": 0.7834655046463013, + "learning_rate": 4.841649358578978e-05, + "loss": 0.6962, + "step": 15119 + }, + { + "epoch": 2.691951566951567, + "grad_norm": 0.787391185760498, + "learning_rate": 4.8404502625893474e-05, + "loss": 0.8598, + "step": 15120 + }, + { + "epoch": 2.6921296296296298, + "grad_norm": 0.907228410243988, + "learning_rate": 4.839251267690371e-05, + "loss": 0.9913, + "step": 15121 + }, + { + "epoch": 2.6923076923076925, + "grad_norm": 0.8313533663749695, + "learning_rate": 4.838052373905554e-05, + "loss": 0.9542, + "step": 15122 + }, + { + "epoch": 2.692485754985755, + "grad_norm": 0.8444675207138062, + "learning_rate": 4.83685358125837e-05, + "loss": 0.7437, + "step": 15123 + }, + { + "epoch": 2.6926638176638176, + "grad_norm": 0.8656189441680908, + "learning_rate": 4.835654889772319e-05, + "loss": 1.104, + "step": 15124 + }, + { + "epoch": 2.6928418803418803, + "grad_norm": 0.9181584715843201, + "learning_rate": 4.8344562994708805e-05, + "loss": 0.8533, + "step": 15125 + }, + { + "epoch": 2.693019943019943, + "grad_norm": 0.5977702140808105, + "learning_rate": 4.833257810377542e-05, + "loss": 0.495, + "step": 15126 + }, + { + "epoch": 2.693198005698006, + "grad_norm": 0.8839932084083557, + "learning_rate": 4.8320594225157834e-05, + "loss": 0.8026, + "step": 15127 + }, + { + "epoch": 2.693376068376068, + "grad_norm": 0.876559853553772, + "learning_rate": 4.8308611359090846e-05, + "loss": 0.893, + "step": 15128 + }, + { + "epoch": 2.693554131054131, + "grad_norm": 0.7847880721092224, + "learning_rate": 4.829662950580924e-05, + "loss": 0.7794, + "step": 15129 + }, + { + "epoch": 2.6937321937321936, + "grad_norm": 0.8713442087173462, + "learning_rate": 4.828464866554778e-05, + "loss": 1.0394, + "step": 15130 + }, + { + "epoch": 2.6939102564102564, + "grad_norm": 0.9720988869667053, + "learning_rate": 4.827266883854116e-05, + "loss": 0.7844, + "step": 15131 + }, + { + "epoch": 2.694088319088319, + "grad_norm": 0.8163195252418518, + "learning_rate": 4.82606900250242e-05, + "loss": 0.711, + "step": 15132 + }, + { + "epoch": 2.694266381766382, + "grad_norm": 0.7119855880737305, + "learning_rate": 4.8248712225231486e-05, + "loss": 0.6224, + "step": 15133 + }, + { + "epoch": 2.6944444444444446, + "grad_norm": 0.8176950812339783, + "learning_rate": 4.823673543939777e-05, + "loss": 0.8695, + "step": 15134 + }, + { + "epoch": 2.6946225071225074, + "grad_norm": 0.8138632774353027, + "learning_rate": 4.822475966775771e-05, + "loss": 0.7331, + "step": 15135 + }, + { + "epoch": 2.6948005698005697, + "grad_norm": 0.9323116540908813, + "learning_rate": 4.821278491054589e-05, + "loss": 0.8275, + "step": 15136 + }, + { + "epoch": 2.6949786324786325, + "grad_norm": 0.7593950033187866, + "learning_rate": 4.820081116799704e-05, + "loss": 0.9571, + "step": 15137 + }, + { + "epoch": 2.695156695156695, + "grad_norm": 0.9058876037597656, + "learning_rate": 4.818883844034563e-05, + "loss": 0.7676, + "step": 15138 + }, + { + "epoch": 2.695334757834758, + "grad_norm": 1.0943962335586548, + "learning_rate": 4.8176866727826365e-05, + "loss": 0.7542, + "step": 15139 + }, + { + "epoch": 2.6955128205128203, + "grad_norm": 0.9133912324905396, + "learning_rate": 4.8164896030673664e-05, + "loss": 0.8419, + "step": 15140 + }, + { + "epoch": 2.695690883190883, + "grad_norm": 0.8556821942329407, + "learning_rate": 4.8152926349122195e-05, + "loss": 0.8234, + "step": 15141 + }, + { + "epoch": 2.6958689458689458, + "grad_norm": 1.0329471826553345, + "learning_rate": 4.814095768340643e-05, + "loss": 0.8181, + "step": 15142 + }, + { + "epoch": 2.6960470085470085, + "grad_norm": 0.89934903383255, + "learning_rate": 4.812899003376087e-05, + "loss": 0.8392, + "step": 15143 + }, + { + "epoch": 2.6962250712250713, + "grad_norm": 0.7836576104164124, + "learning_rate": 4.811702340042e-05, + "loss": 0.9491, + "step": 15144 + }, + { + "epoch": 2.696403133903134, + "grad_norm": 0.9841184020042419, + "learning_rate": 4.810505778361828e-05, + "loss": 1.0763, + "step": 15145 + }, + { + "epoch": 2.6965811965811968, + "grad_norm": 1.0479893684387207, + "learning_rate": 4.80930931835901e-05, + "loss": 1.054, + "step": 15146 + }, + { + "epoch": 2.6967592592592595, + "grad_norm": 0.895803689956665, + "learning_rate": 4.808112960057002e-05, + "loss": 0.8769, + "step": 15147 + }, + { + "epoch": 2.696937321937322, + "grad_norm": 0.8467312455177307, + "learning_rate": 4.806916703479227e-05, + "loss": 0.8036, + "step": 15148 + }, + { + "epoch": 2.6971153846153846, + "grad_norm": 0.7371073365211487, + "learning_rate": 4.8057205486491366e-05, + "loss": 0.72, + "step": 15149 + }, + { + "epoch": 2.6972934472934473, + "grad_norm": 0.9631866812705994, + "learning_rate": 4.80452449559016e-05, + "loss": 0.8661, + "step": 15150 + }, + { + "epoch": 2.69747150997151, + "grad_norm": 0.8467531204223633, + "learning_rate": 4.803328544325735e-05, + "loss": 0.9359, + "step": 15151 + }, + { + "epoch": 2.6976495726495724, + "grad_norm": 0.8170605897903442, + "learning_rate": 4.802132694879291e-05, + "loss": 0.9086, + "step": 15152 + }, + { + "epoch": 2.697827635327635, + "grad_norm": 0.8378857970237732, + "learning_rate": 4.800936947274255e-05, + "loss": 0.6255, + "step": 15153 + }, + { + "epoch": 2.698005698005698, + "grad_norm": 0.8074176907539368, + "learning_rate": 4.799741301534067e-05, + "loss": 0.9129, + "step": 15154 + }, + { + "epoch": 2.6981837606837606, + "grad_norm": 0.862147331237793, + "learning_rate": 4.798545757682139e-05, + "loss": 0.8298, + "step": 15155 + }, + { + "epoch": 2.6983618233618234, + "grad_norm": 0.8020915985107422, + "learning_rate": 4.797350315741905e-05, + "loss": 0.8364, + "step": 15156 + }, + { + "epoch": 2.698539886039886, + "grad_norm": 0.7929054498672485, + "learning_rate": 4.7961549757367854e-05, + "loss": 1.0302, + "step": 15157 + }, + { + "epoch": 2.698717948717949, + "grad_norm": 0.8528931140899658, + "learning_rate": 4.7949597376901964e-05, + "loss": 0.7891, + "step": 15158 + }, + { + "epoch": 2.6988960113960117, + "grad_norm": 0.8090588450431824, + "learning_rate": 4.793764601625561e-05, + "loss": 0.7905, + "step": 15159 + }, + { + "epoch": 2.699074074074074, + "grad_norm": 0.8221202492713928, + "learning_rate": 4.7925695675662916e-05, + "loss": 0.8156, + "step": 15160 + }, + { + "epoch": 2.6992521367521367, + "grad_norm": 0.8121498823165894, + "learning_rate": 4.791374635535802e-05, + "loss": 0.865, + "step": 15161 + }, + { + "epoch": 2.6994301994301995, + "grad_norm": 0.7626228928565979, + "learning_rate": 4.790179805557513e-05, + "loss": 0.8033, + "step": 15162 + }, + { + "epoch": 2.699608262108262, + "grad_norm": 0.8483169078826904, + "learning_rate": 4.7889850776548205e-05, + "loss": 0.9239, + "step": 15163 + }, + { + "epoch": 2.6997863247863245, + "grad_norm": 0.8302589058876038, + "learning_rate": 4.7877904518511485e-05, + "loss": 0.8445, + "step": 15164 + }, + { + "epoch": 2.6999643874643873, + "grad_norm": 0.9140453338623047, + "learning_rate": 4.786595928169887e-05, + "loss": 1.0492, + "step": 15165 + }, + { + "epoch": 2.70014245014245, + "grad_norm": 0.8046873807907104, + "learning_rate": 4.785401506634453e-05, + "loss": 1.0009, + "step": 15166 + }, + { + "epoch": 2.7003205128205128, + "grad_norm": 0.8879752159118652, + "learning_rate": 4.7842071872682434e-05, + "loss": 0.7788, + "step": 15167 + }, + { + "epoch": 2.7004985754985755, + "grad_norm": 0.8190163969993591, + "learning_rate": 4.783012970094659e-05, + "loss": 0.9063, + "step": 15168 + }, + { + "epoch": 2.7006766381766383, + "grad_norm": 0.9363130331039429, + "learning_rate": 4.781818855137099e-05, + "loss": 0.9723, + "step": 15169 + }, + { + "epoch": 2.700854700854701, + "grad_norm": 0.8428171873092651, + "learning_rate": 4.780624842418958e-05, + "loss": 0.9173, + "step": 15170 + }, + { + "epoch": 2.701032763532764, + "grad_norm": 0.8089821934700012, + "learning_rate": 4.779430931963627e-05, + "loss": 0.8996, + "step": 15171 + }, + { + "epoch": 2.701210826210826, + "grad_norm": 0.8893290758132935, + "learning_rate": 4.77823712379451e-05, + "loss": 0.9483, + "step": 15172 + }, + { + "epoch": 2.701388888888889, + "grad_norm": 0.8589824438095093, + "learning_rate": 4.777043417934981e-05, + "loss": 0.8765, + "step": 15173 + }, + { + "epoch": 2.7015669515669516, + "grad_norm": 0.8665438294410706, + "learning_rate": 4.7758498144084405e-05, + "loss": 0.8546, + "step": 15174 + }, + { + "epoch": 2.7017450142450143, + "grad_norm": 0.743841826915741, + "learning_rate": 4.774656313238272e-05, + "loss": 0.6866, + "step": 15175 + }, + { + "epoch": 2.7019230769230766, + "grad_norm": 0.9317346811294556, + "learning_rate": 4.7734629144478574e-05, + "loss": 0.8004, + "step": 15176 + }, + { + "epoch": 2.7021011396011394, + "grad_norm": 0.8244655132293701, + "learning_rate": 4.77226961806058e-05, + "loss": 0.9302, + "step": 15177 + }, + { + "epoch": 2.702279202279202, + "grad_norm": 1.0759600400924683, + "learning_rate": 4.771076424099815e-05, + "loss": 0.9073, + "step": 15178 + }, + { + "epoch": 2.702457264957265, + "grad_norm": 0.8852303624153137, + "learning_rate": 4.769883332588954e-05, + "loss": 0.8084, + "step": 15179 + }, + { + "epoch": 2.7026353276353277, + "grad_norm": 0.8642051815986633, + "learning_rate": 4.7686903435513564e-05, + "loss": 1.0018, + "step": 15180 + }, + { + "epoch": 2.7028133903133904, + "grad_norm": 0.9442928433418274, + "learning_rate": 4.767497457010408e-05, + "loss": 0.8099, + "step": 15181 + }, + { + "epoch": 2.702991452991453, + "grad_norm": 0.8357751965522766, + "learning_rate": 4.7663046729894776e-05, + "loss": 0.8594, + "step": 15182 + }, + { + "epoch": 2.703169515669516, + "grad_norm": 1.0791765451431274, + "learning_rate": 4.765111991511936e-05, + "loss": 1.1203, + "step": 15183 + }, + { + "epoch": 2.703347578347578, + "grad_norm": 0.7855654954910278, + "learning_rate": 4.7639194126011485e-05, + "loss": 0.7218, + "step": 15184 + }, + { + "epoch": 2.703525641025641, + "grad_norm": 0.8058420419692993, + "learning_rate": 4.762726936280485e-05, + "loss": 0.7885, + "step": 15185 + }, + { + "epoch": 2.7037037037037037, + "grad_norm": 0.7701787352561951, + "learning_rate": 4.761534562573302e-05, + "loss": 0.6378, + "step": 15186 + }, + { + "epoch": 2.7038817663817665, + "grad_norm": 0.9011744856834412, + "learning_rate": 4.760342291502976e-05, + "loss": 0.9106, + "step": 15187 + }, + { + "epoch": 2.7040598290598292, + "grad_norm": 0.7268012762069702, + "learning_rate": 4.759150123092851e-05, + "loss": 0.6303, + "step": 15188 + }, + { + "epoch": 2.7042378917378915, + "grad_norm": 0.8369283676147461, + "learning_rate": 4.7579580573663e-05, + "loss": 0.8013, + "step": 15189 + }, + { + "epoch": 2.7044159544159543, + "grad_norm": 0.9511098861694336, + "learning_rate": 4.756766094346663e-05, + "loss": 1.0211, + "step": 15190 + }, + { + "epoch": 2.704594017094017, + "grad_norm": 0.8408896327018738, + "learning_rate": 4.7555742340573074e-05, + "loss": 1.1018, + "step": 15191 + }, + { + "epoch": 2.70477207977208, + "grad_norm": 0.9166504740715027, + "learning_rate": 4.7543824765215795e-05, + "loss": 0.9222, + "step": 15192 + }, + { + "epoch": 2.7049501424501425, + "grad_norm": 0.8373738527297974, + "learning_rate": 4.753190821762826e-05, + "loss": 0.9735, + "step": 15193 + }, + { + "epoch": 2.7051282051282053, + "grad_norm": 0.8610605597496033, + "learning_rate": 4.751999269804408e-05, + "loss": 0.7942, + "step": 15194 + }, + { + "epoch": 2.705306267806268, + "grad_norm": 0.8778019547462463, + "learning_rate": 4.750807820669654e-05, + "loss": 0.8055, + "step": 15195 + }, + { + "epoch": 2.7054843304843303, + "grad_norm": 0.9997664093971252, + "learning_rate": 4.749616474381921e-05, + "loss": 0.8461, + "step": 15196 + }, + { + "epoch": 2.705662393162393, + "grad_norm": 0.8362101316452026, + "learning_rate": 4.748425230964545e-05, + "loss": 1.0008, + "step": 15197 + }, + { + "epoch": 2.705840455840456, + "grad_norm": 0.870482861995697, + "learning_rate": 4.747234090440869e-05, + "loss": 0.9547, + "step": 15198 + }, + { + "epoch": 2.7060185185185186, + "grad_norm": 0.867431104183197, + "learning_rate": 4.746043052834228e-05, + "loss": 0.8533, + "step": 15199 + }, + { + "epoch": 2.7061965811965814, + "grad_norm": 0.842071533203125, + "learning_rate": 4.7448521181679604e-05, + "loss": 0.8919, + "step": 15200 + }, + { + "epoch": 2.7063746438746437, + "grad_norm": 0.9487791657447815, + "learning_rate": 4.743661286465398e-05, + "loss": 0.8072, + "step": 15201 + }, + { + "epoch": 2.7065527065527064, + "grad_norm": 0.8469042181968689, + "learning_rate": 4.742470557749874e-05, + "loss": 0.8792, + "step": 15202 + }, + { + "epoch": 2.706730769230769, + "grad_norm": 0.86415696144104, + "learning_rate": 4.7412799320447145e-05, + "loss": 0.9725, + "step": 15203 + }, + { + "epoch": 2.706908831908832, + "grad_norm": 0.9035004377365112, + "learning_rate": 4.740089409373257e-05, + "loss": 0.9915, + "step": 15204 + }, + { + "epoch": 2.7070868945868947, + "grad_norm": 0.8122807741165161, + "learning_rate": 4.7388989897588156e-05, + "loss": 0.946, + "step": 15205 + }, + { + "epoch": 2.7072649572649574, + "grad_norm": 0.9801422357559204, + "learning_rate": 4.737708673224721e-05, + "loss": 0.9357, + "step": 15206 + }, + { + "epoch": 2.70744301994302, + "grad_norm": 1.0265265703201294, + "learning_rate": 4.736518459794295e-05, + "loss": 0.7982, + "step": 15207 + }, + { + "epoch": 2.7076210826210825, + "grad_norm": 0.828814685344696, + "learning_rate": 4.735328349490855e-05, + "loss": 0.6864, + "step": 15208 + }, + { + "epoch": 2.7077991452991452, + "grad_norm": 0.7948212623596191, + "learning_rate": 4.7341383423377195e-05, + "loss": 0.8661, + "step": 15209 + }, + { + "epoch": 2.707977207977208, + "grad_norm": 0.8372616767883301, + "learning_rate": 4.7329484383582046e-05, + "loss": 0.8818, + "step": 15210 + }, + { + "epoch": 2.7081552706552707, + "grad_norm": 0.8000285029411316, + "learning_rate": 4.731758637575624e-05, + "loss": 0.8006, + "step": 15211 + }, + { + "epoch": 2.7083333333333335, + "grad_norm": 0.7860875725746155, + "learning_rate": 4.730568940013289e-05, + "loss": 0.926, + "step": 15212 + }, + { + "epoch": 2.708511396011396, + "grad_norm": 0.9157412052154541, + "learning_rate": 4.7293793456945054e-05, + "loss": 0.7042, + "step": 15213 + }, + { + "epoch": 2.7086894586894585, + "grad_norm": 0.8802906274795532, + "learning_rate": 4.728189854642589e-05, + "loss": 0.8639, + "step": 15214 + }, + { + "epoch": 2.7088675213675213, + "grad_norm": 0.8047248721122742, + "learning_rate": 4.7270004668808397e-05, + "loss": 0.7603, + "step": 15215 + }, + { + "epoch": 2.709045584045584, + "grad_norm": 0.9848080277442932, + "learning_rate": 4.725811182432564e-05, + "loss": 0.8213, + "step": 15216 + }, + { + "epoch": 2.709223646723647, + "grad_norm": 0.8568090200424194, + "learning_rate": 4.724622001321062e-05, + "loss": 0.7663, + "step": 15217 + }, + { + "epoch": 2.7094017094017095, + "grad_norm": 0.7926214337348938, + "learning_rate": 4.7234329235696284e-05, + "loss": 0.874, + "step": 15218 + }, + { + "epoch": 2.7095797720797723, + "grad_norm": 0.8389978408813477, + "learning_rate": 4.7222439492015734e-05, + "loss": 0.623, + "step": 15219 + }, + { + "epoch": 2.7097578347578346, + "grad_norm": 0.8635174036026001, + "learning_rate": 4.7210550782401773e-05, + "loss": 0.822, + "step": 15220 + }, + { + "epoch": 2.7099358974358974, + "grad_norm": 0.8381666541099548, + "learning_rate": 4.7198663107087446e-05, + "loss": 1.0864, + "step": 15221 + }, + { + "epoch": 2.71011396011396, + "grad_norm": 1.0722376108169556, + "learning_rate": 4.718677646630564e-05, + "loss": 0.8527, + "step": 15222 + }, + { + "epoch": 2.710292022792023, + "grad_norm": 0.9505516290664673, + "learning_rate": 4.7174890860289224e-05, + "loss": 1.0645, + "step": 15223 + }, + { + "epoch": 2.7104700854700856, + "grad_norm": 0.7757406234741211, + "learning_rate": 4.7163006289271095e-05, + "loss": 0.5924, + "step": 15224 + }, + { + "epoch": 2.710648148148148, + "grad_norm": 0.816387414932251, + "learning_rate": 4.71511227534841e-05, + "loss": 0.8337, + "step": 15225 + }, + { + "epoch": 2.7108262108262107, + "grad_norm": 0.7817156910896301, + "learning_rate": 4.7139240253161065e-05, + "loss": 0.8315, + "step": 15226 + }, + { + "epoch": 2.7110042735042734, + "grad_norm": 0.9753041863441467, + "learning_rate": 4.7127358788534816e-05, + "loss": 0.851, + "step": 15227 + }, + { + "epoch": 2.711182336182336, + "grad_norm": 0.7564638257026672, + "learning_rate": 4.7115478359838095e-05, + "loss": 0.8132, + "step": 15228 + }, + { + "epoch": 2.711360398860399, + "grad_norm": 0.8709259629249573, + "learning_rate": 4.710359896730379e-05, + "loss": 1.0277, + "step": 15229 + }, + { + "epoch": 2.7115384615384617, + "grad_norm": 0.9849836230278015, + "learning_rate": 4.7091720611164504e-05, + "loss": 0.9778, + "step": 15230 + }, + { + "epoch": 2.7117165242165244, + "grad_norm": 0.8330100178718567, + "learning_rate": 4.707984329165309e-05, + "loss": 0.7138, + "step": 15231 + }, + { + "epoch": 2.7118945868945867, + "grad_norm": 1.005644679069519, + "learning_rate": 4.706796700900221e-05, + "loss": 1.0089, + "step": 15232 + }, + { + "epoch": 2.7120726495726495, + "grad_norm": 0.8292263746261597, + "learning_rate": 4.705609176344452e-05, + "loss": 0.8323, + "step": 15233 + }, + { + "epoch": 2.7122507122507122, + "grad_norm": 0.860713005065918, + "learning_rate": 4.704421755521281e-05, + "loss": 0.821, + "step": 15234 + }, + { + "epoch": 2.712428774928775, + "grad_norm": 0.8316803574562073, + "learning_rate": 4.703234438453958e-05, + "loss": 0.9181, + "step": 15235 + }, + { + "epoch": 2.7126068376068377, + "grad_norm": 0.7368014454841614, + "learning_rate": 4.70204722516576e-05, + "loss": 0.8206, + "step": 15236 + }, + { + "epoch": 2.7127849002849, + "grad_norm": 1.0202926397323608, + "learning_rate": 4.7008601156799336e-05, + "loss": 0.8101, + "step": 15237 + }, + { + "epoch": 2.712962962962963, + "grad_norm": 0.8069320917129517, + "learning_rate": 4.69967311001975e-05, + "loss": 0.9042, + "step": 15238 + }, + { + "epoch": 2.7131410256410255, + "grad_norm": 0.8426684737205505, + "learning_rate": 4.69848620820846e-05, + "loss": 0.7318, + "step": 15239 + }, + { + "epoch": 2.7133190883190883, + "grad_norm": 0.8863842487335205, + "learning_rate": 4.69729941026932e-05, + "loss": 1.0172, + "step": 15240 + }, + { + "epoch": 2.713497150997151, + "grad_norm": 0.7984182834625244, + "learning_rate": 4.696112716225582e-05, + "loss": 0.8298, + "step": 15241 + }, + { + "epoch": 2.713675213675214, + "grad_norm": 0.8328375220298767, + "learning_rate": 4.6949261261005e-05, + "loss": 0.7663, + "step": 15242 + }, + { + "epoch": 2.7138532763532766, + "grad_norm": 0.9197641015052795, + "learning_rate": 4.693739639917314e-05, + "loss": 0.8951, + "step": 15243 + }, + { + "epoch": 2.7140313390313393, + "grad_norm": 0.7421545386314392, + "learning_rate": 4.692553257699286e-05, + "loss": 0.7235, + "step": 15244 + }, + { + "epoch": 2.7142094017094016, + "grad_norm": 0.8033188581466675, + "learning_rate": 4.691366979469642e-05, + "loss": 0.9693, + "step": 15245 + }, + { + "epoch": 2.7143874643874644, + "grad_norm": 0.8765473365783691, + "learning_rate": 4.6901808052516436e-05, + "loss": 0.8851, + "step": 15246 + }, + { + "epoch": 2.714565527065527, + "grad_norm": 0.8351873755455017, + "learning_rate": 4.688994735068515e-05, + "loss": 1.0156, + "step": 15247 + }, + { + "epoch": 2.71474358974359, + "grad_norm": 0.8569470643997192, + "learning_rate": 4.6878087689435046e-05, + "loss": 0.7149, + "step": 15248 + }, + { + "epoch": 2.714921652421652, + "grad_norm": 0.8334367871284485, + "learning_rate": 4.686622906899847e-05, + "loss": 0.9218, + "step": 15249 + }, + { + "epoch": 2.715099715099715, + "grad_norm": 0.8889651298522949, + "learning_rate": 4.685437148960775e-05, + "loss": 0.8987, + "step": 15250 + }, + { + "epoch": 2.7152777777777777, + "grad_norm": 0.9381657838821411, + "learning_rate": 4.684251495149522e-05, + "loss": 0.7798, + "step": 15251 + }, + { + "epoch": 2.7154558404558404, + "grad_norm": 0.7698730826377869, + "learning_rate": 4.68306594548932e-05, + "loss": 0.8248, + "step": 15252 + }, + { + "epoch": 2.715633903133903, + "grad_norm": 0.8980026245117188, + "learning_rate": 4.681880500003391e-05, + "loss": 1.0156, + "step": 15253 + }, + { + "epoch": 2.715811965811966, + "grad_norm": 0.7872338891029358, + "learning_rate": 4.6806951587149694e-05, + "loss": 0.6389, + "step": 15254 + }, + { + "epoch": 2.7159900284900287, + "grad_norm": 0.8155974745750427, + "learning_rate": 4.6795099216472774e-05, + "loss": 0.9081, + "step": 15255 + }, + { + "epoch": 2.7161680911680914, + "grad_norm": 0.7678217887878418, + "learning_rate": 4.678324788823535e-05, + "loss": 0.6193, + "step": 15256 + }, + { + "epoch": 2.7163461538461537, + "grad_norm": 0.75429767370224, + "learning_rate": 4.6771397602669643e-05, + "loss": 0.9384, + "step": 15257 + }, + { + "epoch": 2.7165242165242165, + "grad_norm": 0.8755250573158264, + "learning_rate": 4.675954836000779e-05, + "loss": 0.8563, + "step": 15258 + }, + { + "epoch": 2.7167022792022792, + "grad_norm": 0.8393009305000305, + "learning_rate": 4.6747700160482053e-05, + "loss": 0.9407, + "step": 15259 + }, + { + "epoch": 2.716880341880342, + "grad_norm": 0.8478221297264099, + "learning_rate": 4.673585300432445e-05, + "loss": 0.7562, + "step": 15260 + }, + { + "epoch": 2.7170584045584043, + "grad_norm": 0.7497259974479675, + "learning_rate": 4.672400689176722e-05, + "loss": 0.8406, + "step": 15261 + }, + { + "epoch": 2.717236467236467, + "grad_norm": 0.9695250391960144, + "learning_rate": 4.671216182304234e-05, + "loss": 0.9505, + "step": 15262 + }, + { + "epoch": 2.71741452991453, + "grad_norm": 0.9375512599945068, + "learning_rate": 4.6700317798382e-05, + "loss": 0.9024, + "step": 15263 + }, + { + "epoch": 2.7175925925925926, + "grad_norm": 0.7930737137794495, + "learning_rate": 4.6688474818018194e-05, + "loss": 0.8416, + "step": 15264 + }, + { + "epoch": 2.7177706552706553, + "grad_norm": 0.9707022309303284, + "learning_rate": 4.667663288218298e-05, + "loss": 1.1172, + "step": 15265 + }, + { + "epoch": 2.717948717948718, + "grad_norm": 0.7616816759109497, + "learning_rate": 4.666479199110838e-05, + "loss": 0.8557, + "step": 15266 + }, + { + "epoch": 2.718126780626781, + "grad_norm": 0.7836055159568787, + "learning_rate": 4.66529521450264e-05, + "loss": 0.7299, + "step": 15267 + }, + { + "epoch": 2.7183048433048436, + "grad_norm": 0.8313519954681396, + "learning_rate": 4.664111334416894e-05, + "loss": 0.8268, + "step": 15268 + }, + { + "epoch": 2.718482905982906, + "grad_norm": 0.9130576252937317, + "learning_rate": 4.662927558876812e-05, + "loss": 0.8913, + "step": 15269 + }, + { + "epoch": 2.7186609686609686, + "grad_norm": 0.8552213907241821, + "learning_rate": 4.661743887905569e-05, + "loss": 0.9396, + "step": 15270 + }, + { + "epoch": 2.7188390313390314, + "grad_norm": 0.7953839898109436, + "learning_rate": 4.660560321526373e-05, + "loss": 0.74, + "step": 15271 + }, + { + "epoch": 2.719017094017094, + "grad_norm": 0.9148657321929932, + "learning_rate": 4.6593768597623974e-05, + "loss": 0.7821, + "step": 15272 + }, + { + "epoch": 2.7191951566951564, + "grad_norm": 0.8587655425071716, + "learning_rate": 4.658193502636843e-05, + "loss": 0.9495, + "step": 15273 + }, + { + "epoch": 2.719373219373219, + "grad_norm": 0.8915669322013855, + "learning_rate": 4.6570102501728896e-05, + "loss": 0.8612, + "step": 15274 + }, + { + "epoch": 2.719551282051282, + "grad_norm": 0.957039475440979, + "learning_rate": 4.655827102393717e-05, + "loss": 0.8506, + "step": 15275 + }, + { + "epoch": 2.7197293447293447, + "grad_norm": 0.7784267067909241, + "learning_rate": 4.654644059322519e-05, + "loss": 0.6864, + "step": 15276 + }, + { + "epoch": 2.7199074074074074, + "grad_norm": 0.9508241415023804, + "learning_rate": 4.65346112098246e-05, + "loss": 1.0097, + "step": 15277 + }, + { + "epoch": 2.72008547008547, + "grad_norm": 0.8316742777824402, + "learning_rate": 4.6522782873967265e-05, + "loss": 0.7444, + "step": 15278 + }, + { + "epoch": 2.720263532763533, + "grad_norm": 0.8781944513320923, + "learning_rate": 4.651095558588491e-05, + "loss": 0.8725, + "step": 15279 + }, + { + "epoch": 2.7204415954415957, + "grad_norm": 0.9407825469970703, + "learning_rate": 4.649912934580927e-05, + "loss": 0.9788, + "step": 15280 + }, + { + "epoch": 2.720619658119658, + "grad_norm": 0.9863289594650269, + "learning_rate": 4.6487304153972045e-05, + "loss": 0.7777, + "step": 15281 + }, + { + "epoch": 2.7207977207977208, + "grad_norm": 0.7580869793891907, + "learning_rate": 4.6475480010604945e-05, + "loss": 0.5835, + "step": 15282 + }, + { + "epoch": 2.7209757834757835, + "grad_norm": 0.7973836660385132, + "learning_rate": 4.646365691593961e-05, + "loss": 0.7633, + "step": 15283 + }, + { + "epoch": 2.7211538461538463, + "grad_norm": 0.8107978701591492, + "learning_rate": 4.645183487020772e-05, + "loss": 0.7149, + "step": 15284 + }, + { + "epoch": 2.7213319088319086, + "grad_norm": 0.8944578170776367, + "learning_rate": 4.644001387364084e-05, + "loss": 0.9227, + "step": 15285 + }, + { + "epoch": 2.7215099715099713, + "grad_norm": 0.7592978477478027, + "learning_rate": 4.642819392647071e-05, + "loss": 0.464, + "step": 15286 + }, + { + "epoch": 2.721688034188034, + "grad_norm": 0.8484344482421875, + "learning_rate": 4.641637502892876e-05, + "loss": 1.0439, + "step": 15287 + }, + { + "epoch": 2.721866096866097, + "grad_norm": 0.8766823410987854, + "learning_rate": 4.640455718124667e-05, + "loss": 0.7561, + "step": 15288 + }, + { + "epoch": 2.7220441595441596, + "grad_norm": 0.8039024472236633, + "learning_rate": 4.639274038365594e-05, + "loss": 0.6774, + "step": 15289 + }, + { + "epoch": 2.7222222222222223, + "grad_norm": 0.8199611902236938, + "learning_rate": 4.63809246363881e-05, + "loss": 0.7721, + "step": 15290 + }, + { + "epoch": 2.722400284900285, + "grad_norm": 0.8209745287895203, + "learning_rate": 4.636910993967467e-05, + "loss": 0.7017, + "step": 15291 + }, + { + "epoch": 2.722578347578348, + "grad_norm": 0.8822476267814636, + "learning_rate": 4.6357296293747075e-05, + "loss": 0.8742, + "step": 15292 + }, + { + "epoch": 2.72275641025641, + "grad_norm": 0.8172603249549866, + "learning_rate": 4.634548369883687e-05, + "loss": 0.8165, + "step": 15293 + }, + { + "epoch": 2.722934472934473, + "grad_norm": 0.8601866960525513, + "learning_rate": 4.633367215517546e-05, + "loss": 0.7961, + "step": 15294 + }, + { + "epoch": 2.7231125356125356, + "grad_norm": 0.9346174001693726, + "learning_rate": 4.632186166299425e-05, + "loss": 0.9229, + "step": 15295 + }, + { + "epoch": 2.7232905982905984, + "grad_norm": 0.8956635594367981, + "learning_rate": 4.631005222252465e-05, + "loss": 0.7886, + "step": 15296 + }, + { + "epoch": 2.7234686609686607, + "grad_norm": 0.8453384637832642, + "learning_rate": 4.629824383399805e-05, + "loss": 0.8513, + "step": 15297 + }, + { + "epoch": 2.7236467236467234, + "grad_norm": 0.8931429982185364, + "learning_rate": 4.628643649764581e-05, + "loss": 1.0195, + "step": 15298 + }, + { + "epoch": 2.723824786324786, + "grad_norm": 0.7326723337173462, + "learning_rate": 4.6274630213699265e-05, + "loss": 0.7616, + "step": 15299 + }, + { + "epoch": 2.724002849002849, + "grad_norm": 0.8572023510932922, + "learning_rate": 4.6262824982389706e-05, + "loss": 0.8266, + "step": 15300 + }, + { + "epoch": 2.7241809116809117, + "grad_norm": 0.7753783464431763, + "learning_rate": 4.625102080394853e-05, + "loss": 0.6907, + "step": 15301 + }, + { + "epoch": 2.7243589743589745, + "grad_norm": 0.8758052587509155, + "learning_rate": 4.623921767860687e-05, + "loss": 0.6369, + "step": 15302 + }, + { + "epoch": 2.724537037037037, + "grad_norm": 0.8508220314979553, + "learning_rate": 4.6227415606596104e-05, + "loss": 0.933, + "step": 15303 + }, + { + "epoch": 2.7247150997151, + "grad_norm": 0.7440072298049927, + "learning_rate": 4.621561458814743e-05, + "loss": 0.7172, + "step": 15304 + }, + { + "epoch": 2.7248931623931623, + "grad_norm": 0.9081870317459106, + "learning_rate": 4.6203814623492046e-05, + "loss": 0.8964, + "step": 15305 + }, + { + "epoch": 2.725071225071225, + "grad_norm": 0.9127907156944275, + "learning_rate": 4.619201571286117e-05, + "loss": 1.0081, + "step": 15306 + }, + { + "epoch": 2.7252492877492878, + "grad_norm": 0.9508554935455322, + "learning_rate": 4.618021785648597e-05, + "loss": 0.94, + "step": 15307 + }, + { + "epoch": 2.7254273504273505, + "grad_norm": 0.8726735711097717, + "learning_rate": 4.616842105459761e-05, + "loss": 0.9284, + "step": 15308 + }, + { + "epoch": 2.7256054131054133, + "grad_norm": 0.9266753792762756, + "learning_rate": 4.6156625307427206e-05, + "loss": 0.9069, + "step": 15309 + }, + { + "epoch": 2.7257834757834756, + "grad_norm": 0.952553391456604, + "learning_rate": 4.614483061520584e-05, + "loss": 0.8604, + "step": 15310 + }, + { + "epoch": 2.7259615384615383, + "grad_norm": 0.7702621817588806, + "learning_rate": 4.613303697816471e-05, + "loss": 0.808, + "step": 15311 + }, + { + "epoch": 2.726139601139601, + "grad_norm": 0.8052653670310974, + "learning_rate": 4.612124439653477e-05, + "loss": 0.8696, + "step": 15312 + }, + { + "epoch": 2.726317663817664, + "grad_norm": 0.8808547854423523, + "learning_rate": 4.610945287054714e-05, + "loss": 0.9595, + "step": 15313 + }, + { + "epoch": 2.7264957264957266, + "grad_norm": 1.0233266353607178, + "learning_rate": 4.609766240043284e-05, + "loss": 1.0691, + "step": 15314 + }, + { + "epoch": 2.7266737891737893, + "grad_norm": 0.8129898309707642, + "learning_rate": 4.6085872986422826e-05, + "loss": 0.9269, + "step": 15315 + }, + { + "epoch": 2.726851851851852, + "grad_norm": 1.2745141983032227, + "learning_rate": 4.607408462874823e-05, + "loss": 0.937, + "step": 15316 + }, + { + "epoch": 2.7270299145299144, + "grad_norm": 0.808274507522583, + "learning_rate": 4.606229732763984e-05, + "loss": 0.789, + "step": 15317 + }, + { + "epoch": 2.727207977207977, + "grad_norm": 0.8849375247955322, + "learning_rate": 4.605051108332875e-05, + "loss": 0.8993, + "step": 15318 + }, + { + "epoch": 2.72738603988604, + "grad_norm": 0.8251593112945557, + "learning_rate": 4.603872589604576e-05, + "loss": 0.9057, + "step": 15319 + }, + { + "epoch": 2.7275641025641026, + "grad_norm": 0.8271582126617432, + "learning_rate": 4.602694176602188e-05, + "loss": 0.8378, + "step": 15320 + }, + { + "epoch": 2.7277421652421654, + "grad_norm": 0.8139070868492126, + "learning_rate": 4.6015158693487956e-05, + "loss": 0.8014, + "step": 15321 + }, + { + "epoch": 2.7279202279202277, + "grad_norm": 0.8873880505561829, + "learning_rate": 4.600337667867486e-05, + "loss": 0.8707, + "step": 15322 + }, + { + "epoch": 2.7280982905982905, + "grad_norm": 0.8616414666175842, + "learning_rate": 4.599159572181342e-05, + "loss": 0.8538, + "step": 15323 + }, + { + "epoch": 2.728276353276353, + "grad_norm": 0.8280995488166809, + "learning_rate": 4.5979815823134466e-05, + "loss": 0.8444, + "step": 15324 + }, + { + "epoch": 2.728454415954416, + "grad_norm": 0.8684375882148743, + "learning_rate": 4.596803698286878e-05, + "loss": 0.7562, + "step": 15325 + }, + { + "epoch": 2.7286324786324787, + "grad_norm": 0.8113002181053162, + "learning_rate": 4.595625920124723e-05, + "loss": 0.8331, + "step": 15326 + }, + { + "epoch": 2.7288105413105415, + "grad_norm": 0.8675588965415955, + "learning_rate": 4.5944482478500436e-05, + "loss": 1.1016, + "step": 15327 + }, + { + "epoch": 2.728988603988604, + "grad_norm": 0.9015034437179565, + "learning_rate": 4.593270681485927e-05, + "loss": 0.9002, + "step": 15328 + }, + { + "epoch": 2.7291666666666665, + "grad_norm": 0.9215324521064758, + "learning_rate": 4.592093221055439e-05, + "loss": 0.9491, + "step": 15329 + }, + { + "epoch": 2.7293447293447293, + "grad_norm": 0.8969921469688416, + "learning_rate": 4.590915866581651e-05, + "loss": 0.8791, + "step": 15330 + }, + { + "epoch": 2.729522792022792, + "grad_norm": 0.9012344479560852, + "learning_rate": 4.5897386180876304e-05, + "loss": 0.9114, + "step": 15331 + }, + { + "epoch": 2.7297008547008548, + "grad_norm": 1.0024429559707642, + "learning_rate": 4.588561475596438e-05, + "loss": 0.8782, + "step": 15332 + }, + { + "epoch": 2.7298789173789175, + "grad_norm": 0.9079484343528748, + "learning_rate": 4.5873844391311496e-05, + "loss": 1.0012, + "step": 15333 + }, + { + "epoch": 2.73005698005698, + "grad_norm": 0.709800660610199, + "learning_rate": 4.5862075087148124e-05, + "loss": 0.7473, + "step": 15334 + }, + { + "epoch": 2.7302350427350426, + "grad_norm": 0.9776272773742676, + "learning_rate": 4.585030684370497e-05, + "loss": 1.1927, + "step": 15335 + }, + { + "epoch": 2.7304131054131053, + "grad_norm": 0.8624512553215027, + "learning_rate": 4.5838539661212565e-05, + "loss": 0.6661, + "step": 15336 + }, + { + "epoch": 2.730591168091168, + "grad_norm": 0.7901379466056824, + "learning_rate": 4.5826773539901456e-05, + "loss": 0.8244, + "step": 15337 + }, + { + "epoch": 2.730769230769231, + "grad_norm": 0.8546316027641296, + "learning_rate": 4.58150084800022e-05, + "loss": 0.8232, + "step": 15338 + }, + { + "epoch": 2.7309472934472936, + "grad_norm": 1.0038648843765259, + "learning_rate": 4.5803244481745275e-05, + "loss": 0.8363, + "step": 15339 + }, + { + "epoch": 2.7311253561253563, + "grad_norm": 0.7757763266563416, + "learning_rate": 4.579148154536117e-05, + "loss": 0.6935, + "step": 15340 + }, + { + "epoch": 2.7313034188034186, + "grad_norm": 0.8671833276748657, + "learning_rate": 4.5779719671080436e-05, + "loss": 0.8453, + "step": 15341 + }, + { + "epoch": 2.7314814814814814, + "grad_norm": 0.8507152795791626, + "learning_rate": 4.57679588591334e-05, + "loss": 0.8142, + "step": 15342 + }, + { + "epoch": 2.731659544159544, + "grad_norm": 0.8205499053001404, + "learning_rate": 4.575619910975062e-05, + "loss": 0.8442, + "step": 15343 + }, + { + "epoch": 2.731837606837607, + "grad_norm": 0.8809645771980286, + "learning_rate": 4.574444042316236e-05, + "loss": 1.024, + "step": 15344 + }, + { + "epoch": 2.7320156695156697, + "grad_norm": 0.825038492679596, + "learning_rate": 4.573268279959912e-05, + "loss": 0.9089, + "step": 15345 + }, + { + "epoch": 2.732193732193732, + "grad_norm": 0.7646815776824951, + "learning_rate": 4.572092623929124e-05, + "loss": 0.9239, + "step": 15346 + }, + { + "epoch": 2.7323717948717947, + "grad_norm": 0.8372252583503723, + "learning_rate": 4.570917074246905e-05, + "loss": 0.8687, + "step": 15347 + }, + { + "epoch": 2.7325498575498575, + "grad_norm": 0.9108861088752747, + "learning_rate": 4.5697416309362885e-05, + "loss": 0.9451, + "step": 15348 + }, + { + "epoch": 2.73272792022792, + "grad_norm": 0.8537670969963074, + "learning_rate": 4.568566294020303e-05, + "loss": 0.7437, + "step": 15349 + }, + { + "epoch": 2.732905982905983, + "grad_norm": 0.8660921454429626, + "learning_rate": 4.5673910635219766e-05, + "loss": 1.0617, + "step": 15350 + }, + { + "epoch": 2.7330840455840457, + "grad_norm": 0.8499423265457153, + "learning_rate": 4.5662159394643424e-05, + "loss": 0.9355, + "step": 15351 + }, + { + "epoch": 2.7332621082621085, + "grad_norm": 0.7776598930358887, + "learning_rate": 4.565040921870413e-05, + "loss": 0.8132, + "step": 15352 + }, + { + "epoch": 2.7334401709401708, + "grad_norm": 0.9197307229042053, + "learning_rate": 4.563866010763219e-05, + "loss": 0.8648, + "step": 15353 + }, + { + "epoch": 2.7336182336182335, + "grad_norm": 0.8543015122413635, + "learning_rate": 4.5626912061657786e-05, + "loss": 0.8453, + "step": 15354 + }, + { + "epoch": 2.7337962962962963, + "grad_norm": 0.9448479413986206, + "learning_rate": 4.561516508101109e-05, + "loss": 0.9515, + "step": 15355 + }, + { + "epoch": 2.733974358974359, + "grad_norm": 0.7448729276657104, + "learning_rate": 4.5603419165922265e-05, + "loss": 0.64, + "step": 15356 + }, + { + "epoch": 2.734152421652422, + "grad_norm": 0.8229237198829651, + "learning_rate": 4.5591674316621405e-05, + "loss": 0.7936, + "step": 15357 + }, + { + "epoch": 2.734330484330484, + "grad_norm": 0.8518769145011902, + "learning_rate": 4.557993053333873e-05, + "loss": 1.1976, + "step": 15358 + }, + { + "epoch": 2.734508547008547, + "grad_norm": 0.8680224418640137, + "learning_rate": 4.55681878163042e-05, + "loss": 0.8223, + "step": 15359 + }, + { + "epoch": 2.7346866096866096, + "grad_norm": 0.8199124336242676, + "learning_rate": 4.555644616574799e-05, + "loss": 0.634, + "step": 15360 + }, + { + "epoch": 2.7348646723646723, + "grad_norm": 0.8262977004051208, + "learning_rate": 4.554470558190013e-05, + "loss": 0.6373, + "step": 15361 + }, + { + "epoch": 2.735042735042735, + "grad_norm": 0.8114070296287537, + "learning_rate": 4.553296606499062e-05, + "loss": 0.6624, + "step": 15362 + }, + { + "epoch": 2.735220797720798, + "grad_norm": 0.9944671392440796, + "learning_rate": 4.552122761524952e-05, + "loss": 0.8246, + "step": 15363 + }, + { + "epoch": 2.7353988603988606, + "grad_norm": 0.8174465298652649, + "learning_rate": 4.550949023290678e-05, + "loss": 0.8431, + "step": 15364 + }, + { + "epoch": 2.7355769230769234, + "grad_norm": 0.8303970694541931, + "learning_rate": 4.5497753918192356e-05, + "loss": 0.417, + "step": 15365 + }, + { + "epoch": 2.7357549857549857, + "grad_norm": 0.8428391218185425, + "learning_rate": 4.548601867133629e-05, + "loss": 0.751, + "step": 15366 + }, + { + "epoch": 2.7359330484330484, + "grad_norm": 0.8769099116325378, + "learning_rate": 4.5474284492568384e-05, + "loss": 0.8984, + "step": 15367 + }, + { + "epoch": 2.736111111111111, + "grad_norm": 0.8389245867729187, + "learning_rate": 4.546255138211867e-05, + "loss": 0.8503, + "step": 15368 + }, + { + "epoch": 2.736289173789174, + "grad_norm": 0.8404824137687683, + "learning_rate": 4.5450819340216896e-05, + "loss": 0.836, + "step": 15369 + }, + { + "epoch": 2.736467236467236, + "grad_norm": 1.0007327795028687, + "learning_rate": 4.543908836709304e-05, + "loss": 0.9746, + "step": 15370 + }, + { + "epoch": 2.736645299145299, + "grad_norm": 0.7373863458633423, + "learning_rate": 4.542735846297691e-05, + "loss": 0.6752, + "step": 15371 + }, + { + "epoch": 2.7368233618233617, + "grad_norm": 0.8973239660263062, + "learning_rate": 4.541562962809829e-05, + "loss": 0.8988, + "step": 15372 + }, + { + "epoch": 2.7370014245014245, + "grad_norm": 0.8576705455780029, + "learning_rate": 4.5403901862687095e-05, + "loss": 0.9279, + "step": 15373 + }, + { + "epoch": 2.7371794871794872, + "grad_norm": 0.7462539076805115, + "learning_rate": 4.539217516697295e-05, + "loss": 0.7228, + "step": 15374 + }, + { + "epoch": 2.73735754985755, + "grad_norm": 0.8082219362258911, + "learning_rate": 4.538044954118573e-05, + "loss": 0.9562, + "step": 15375 + }, + { + "epoch": 2.7375356125356127, + "grad_norm": 0.7067760825157166, + "learning_rate": 4.5368724985555134e-05, + "loss": 0.5306, + "step": 15376 + }, + { + "epoch": 2.7377136752136755, + "grad_norm": 0.9178285598754883, + "learning_rate": 4.535700150031089e-05, + "loss": 0.8087, + "step": 15377 + }, + { + "epoch": 2.737891737891738, + "grad_norm": 0.803240180015564, + "learning_rate": 4.53452790856827e-05, + "loss": 0.7315, + "step": 15378 + }, + { + "epoch": 2.7380698005698005, + "grad_norm": 0.7200242877006531, + "learning_rate": 4.5333557741900226e-05, + "loss": 0.6329, + "step": 15379 + }, + { + "epoch": 2.7382478632478633, + "grad_norm": 0.8744874596595764, + "learning_rate": 4.5321837469193117e-05, + "loss": 0.8279, + "step": 15380 + }, + { + "epoch": 2.738425925925926, + "grad_norm": 0.7736984491348267, + "learning_rate": 4.531011826779103e-05, + "loss": 0.7267, + "step": 15381 + }, + { + "epoch": 2.7386039886039883, + "grad_norm": 0.893189549446106, + "learning_rate": 4.5298400137923527e-05, + "loss": 0.754, + "step": 15382 + }, + { + "epoch": 2.738782051282051, + "grad_norm": 0.7637171149253845, + "learning_rate": 4.5286683079820314e-05, + "loss": 0.6201, + "step": 15383 + }, + { + "epoch": 2.738960113960114, + "grad_norm": 0.8324360847473145, + "learning_rate": 4.527496709371082e-05, + "loss": 0.7617, + "step": 15384 + }, + { + "epoch": 2.7391381766381766, + "grad_norm": 0.7570679783821106, + "learning_rate": 4.52632521798247e-05, + "loss": 0.7275, + "step": 15385 + }, + { + "epoch": 2.7393162393162394, + "grad_norm": 0.7802938222885132, + "learning_rate": 4.525153833839144e-05, + "loss": 0.7423, + "step": 15386 + }, + { + "epoch": 2.739494301994302, + "grad_norm": 0.8645743727684021, + "learning_rate": 4.523982556964056e-05, + "loss": 0.8066, + "step": 15387 + }, + { + "epoch": 2.739672364672365, + "grad_norm": 0.8080064654350281, + "learning_rate": 4.522811387380155e-05, + "loss": 0.7418, + "step": 15388 + }, + { + "epoch": 2.7398504273504276, + "grad_norm": 0.925401508808136, + "learning_rate": 4.521640325110387e-05, + "loss": 0.9622, + "step": 15389 + }, + { + "epoch": 2.74002849002849, + "grad_norm": 0.8898165822029114, + "learning_rate": 4.520469370177696e-05, + "loss": 0.8543, + "step": 15390 + }, + { + "epoch": 2.7402065527065527, + "grad_norm": 0.8610122799873352, + "learning_rate": 4.519298522605021e-05, + "loss": 0.9502, + "step": 15391 + }, + { + "epoch": 2.7403846153846154, + "grad_norm": 0.8111294507980347, + "learning_rate": 4.51812778241531e-05, + "loss": 0.7526, + "step": 15392 + }, + { + "epoch": 2.740562678062678, + "grad_norm": 0.8465895056724548, + "learning_rate": 4.516957149631498e-05, + "loss": 0.9076, + "step": 15393 + }, + { + "epoch": 2.7407407407407405, + "grad_norm": 0.8541668057441711, + "learning_rate": 4.51578662427652e-05, + "loss": 0.8996, + "step": 15394 + }, + { + "epoch": 2.7409188034188032, + "grad_norm": 0.9113210439682007, + "learning_rate": 4.514616206373311e-05, + "loss": 0.9129, + "step": 15395 + }, + { + "epoch": 2.741096866096866, + "grad_norm": 0.7553523182868958, + "learning_rate": 4.513445895944802e-05, + "loss": 0.7265, + "step": 15396 + }, + { + "epoch": 2.7412749287749287, + "grad_norm": 0.8949921131134033, + "learning_rate": 4.5122756930139206e-05, + "loss": 0.9176, + "step": 15397 + }, + { + "epoch": 2.7414529914529915, + "grad_norm": 0.7957020401954651, + "learning_rate": 4.5111055976036044e-05, + "loss": 0.9384, + "step": 15398 + }, + { + "epoch": 2.7416310541310542, + "grad_norm": 0.759608805179596, + "learning_rate": 4.509935609736764e-05, + "loss": 0.7791, + "step": 15399 + }, + { + "epoch": 2.741809116809117, + "grad_norm": 0.927768886089325, + "learning_rate": 4.508765729436335e-05, + "loss": 0.9113, + "step": 15400 + }, + { + "epoch": 2.7419871794871797, + "grad_norm": 0.910513162612915, + "learning_rate": 4.5075959567252335e-05, + "loss": 0.9334, + "step": 15401 + }, + { + "epoch": 2.742165242165242, + "grad_norm": 0.9029644727706909, + "learning_rate": 4.5064262916263814e-05, + "loss": 0.9487, + "step": 15402 + }, + { + "epoch": 2.742343304843305, + "grad_norm": 0.8001708984375, + "learning_rate": 4.505256734162693e-05, + "loss": 0.8447, + "step": 15403 + }, + { + "epoch": 2.7425213675213675, + "grad_norm": 0.8000209927558899, + "learning_rate": 4.504087284357085e-05, + "loss": 0.6764, + "step": 15404 + }, + { + "epoch": 2.7426994301994303, + "grad_norm": 0.7368536591529846, + "learning_rate": 4.5029179422324686e-05, + "loss": 0.563, + "step": 15405 + }, + { + "epoch": 2.7428774928774926, + "grad_norm": 0.9035481214523315, + "learning_rate": 4.501748707811757e-05, + "loss": 0.8165, + "step": 15406 + }, + { + "epoch": 2.7430555555555554, + "grad_norm": 0.7985709309577942, + "learning_rate": 4.500579581117854e-05, + "loss": 1.0773, + "step": 15407 + }, + { + "epoch": 2.743233618233618, + "grad_norm": 0.7867546677589417, + "learning_rate": 4.499410562173678e-05, + "loss": 0.7416, + "step": 15408 + }, + { + "epoch": 2.743411680911681, + "grad_norm": 0.8863609433174133, + "learning_rate": 4.498241651002117e-05, + "loss": 0.8609, + "step": 15409 + }, + { + "epoch": 2.7435897435897436, + "grad_norm": 0.8197270631790161, + "learning_rate": 4.497072847626087e-05, + "loss": 1.0664, + "step": 15410 + }, + { + "epoch": 2.7437678062678064, + "grad_norm": 0.843718409538269, + "learning_rate": 4.495904152068483e-05, + "loss": 0.9831, + "step": 15411 + }, + { + "epoch": 2.743945868945869, + "grad_norm": 0.8311102986335754, + "learning_rate": 4.4947355643521985e-05, + "loss": 0.8035, + "step": 15412 + }, + { + "epoch": 2.744123931623932, + "grad_norm": 0.8396357297897339, + "learning_rate": 4.493567084500143e-05, + "loss": 1.0015, + "step": 15413 + }, + { + "epoch": 2.744301994301994, + "grad_norm": 0.7959007620811462, + "learning_rate": 4.492398712535194e-05, + "loss": 0.8414, + "step": 15414 + }, + { + "epoch": 2.744480056980057, + "grad_norm": 0.7720336318016052, + "learning_rate": 4.491230448480258e-05, + "loss": 0.8185, + "step": 15415 + }, + { + "epoch": 2.7446581196581197, + "grad_norm": 0.7999769449234009, + "learning_rate": 4.4900622923582115e-05, + "loss": 0.6807, + "step": 15416 + }, + { + "epoch": 2.7448361823361824, + "grad_norm": 0.9882165789604187, + "learning_rate": 4.488894244191951e-05, + "loss": 0.975, + "step": 15417 + }, + { + "epoch": 2.745014245014245, + "grad_norm": 0.8275474309921265, + "learning_rate": 4.48772630400436e-05, + "loss": 0.8459, + "step": 15418 + }, + { + "epoch": 2.7451923076923075, + "grad_norm": 0.8468943238258362, + "learning_rate": 4.486558471818322e-05, + "loss": 0.8217, + "step": 15419 + }, + { + "epoch": 2.7453703703703702, + "grad_norm": 0.8845008015632629, + "learning_rate": 4.485390747656717e-05, + "loss": 0.9811, + "step": 15420 + }, + { + "epoch": 2.745548433048433, + "grad_norm": 1.0010331869125366, + "learning_rate": 4.4842231315424255e-05, + "loss": 0.9437, + "step": 15421 + }, + { + "epoch": 2.7457264957264957, + "grad_norm": 0.7468565106391907, + "learning_rate": 4.483055623498319e-05, + "loss": 0.7205, + "step": 15422 + }, + { + "epoch": 2.7459045584045585, + "grad_norm": 0.9002050757408142, + "learning_rate": 4.4818882235472845e-05, + "loss": 0.8812, + "step": 15423 + }, + { + "epoch": 2.7460826210826212, + "grad_norm": 0.8684462904930115, + "learning_rate": 4.48072093171218e-05, + "loss": 0.7929, + "step": 15424 + }, + { + "epoch": 2.746260683760684, + "grad_norm": 0.8685877323150635, + "learning_rate": 4.479553748015891e-05, + "loss": 0.805, + "step": 15425 + }, + { + "epoch": 2.7464387464387463, + "grad_norm": 0.8292124271392822, + "learning_rate": 4.478386672481272e-05, + "loss": 0.9622, + "step": 15426 + }, + { + "epoch": 2.746616809116809, + "grad_norm": 0.8269517421722412, + "learning_rate": 4.477219705131199e-05, + "loss": 0.8011, + "step": 15427 + }, + { + "epoch": 2.746794871794872, + "grad_norm": 0.8913753628730774, + "learning_rate": 4.4760528459885334e-05, + "loss": 0.9794, + "step": 15428 + }, + { + "epoch": 2.7469729344729346, + "grad_norm": 0.8017858266830444, + "learning_rate": 4.474886095076137e-05, + "loss": 0.8593, + "step": 15429 + }, + { + "epoch": 2.7471509971509973, + "grad_norm": 1.0657325983047485, + "learning_rate": 4.47371945241687e-05, + "loss": 0.8043, + "step": 15430 + }, + { + "epoch": 2.7473290598290596, + "grad_norm": 0.8358477354049683, + "learning_rate": 4.472552918033588e-05, + "loss": 0.818, + "step": 15431 + }, + { + "epoch": 2.7475071225071224, + "grad_norm": 1.0436886548995972, + "learning_rate": 4.4713864919491514e-05, + "loss": 0.9246, + "step": 15432 + }, + { + "epoch": 2.747685185185185, + "grad_norm": 0.9838647246360779, + "learning_rate": 4.470220174186413e-05, + "loss": 1.0345, + "step": 15433 + }, + { + "epoch": 2.747863247863248, + "grad_norm": 0.8583347201347351, + "learning_rate": 4.469053964768222e-05, + "loss": 0.6242, + "step": 15434 + }, + { + "epoch": 2.7480413105413106, + "grad_norm": 0.832467794418335, + "learning_rate": 4.4678878637174304e-05, + "loss": 0.7988, + "step": 15435 + }, + { + "epoch": 2.7482193732193734, + "grad_norm": 0.7854242324829102, + "learning_rate": 4.4667218710568825e-05, + "loss": 0.9386, + "step": 15436 + }, + { + "epoch": 2.748397435897436, + "grad_norm": 0.7748091816902161, + "learning_rate": 4.465555986809423e-05, + "loss": 0.7322, + "step": 15437 + }, + { + "epoch": 2.7485754985754984, + "grad_norm": 0.818305492401123, + "learning_rate": 4.464390210997904e-05, + "loss": 0.822, + "step": 15438 + }, + { + "epoch": 2.748753561253561, + "grad_norm": 0.8253993391990662, + "learning_rate": 4.463224543645151e-05, + "loss": 0.7631, + "step": 15439 + }, + { + "epoch": 2.748931623931624, + "grad_norm": 0.804768443107605, + "learning_rate": 4.46205898477402e-05, + "loss": 0.8783, + "step": 15440 + }, + { + "epoch": 2.7491096866096867, + "grad_norm": 0.8612813949584961, + "learning_rate": 4.460893534407332e-05, + "loss": 0.944, + "step": 15441 + }, + { + "epoch": 2.7492877492877494, + "grad_norm": 0.8149600625038147, + "learning_rate": 4.459728192567932e-05, + "loss": 0.8592, + "step": 15442 + }, + { + "epoch": 2.7494658119658117, + "grad_norm": 0.996081531047821, + "learning_rate": 4.4585629592786496e-05, + "loss": 0.8648, + "step": 15443 + }, + { + "epoch": 2.7496438746438745, + "grad_norm": 0.8563137650489807, + "learning_rate": 4.457397834562314e-05, + "loss": 0.6645, + "step": 15444 + }, + { + "epoch": 2.7496438746438745, + "eval_loss": 1.1326396465301514, + "eval_runtime": 24.3757, + "eval_samples_per_second": 42.707, + "eval_steps_per_second": 21.374, + "step": 15444 + }, + { + "epoch": 2.7498219373219372, + "grad_norm": 0.7415599226951599, + "learning_rate": 4.4562328184417547e-05, + "loss": 0.7481, + "step": 15445 + }, + { + "epoch": 2.75, + "grad_norm": 0.8192741274833679, + "learning_rate": 4.455067910939796e-05, + "loss": 0.8367, + "step": 15446 + }, + { + "epoch": 2.7501780626780628, + "grad_norm": 0.8514624834060669, + "learning_rate": 4.4539031120792604e-05, + "loss": 0.7107, + "step": 15447 + }, + { + "epoch": 2.7503561253561255, + "grad_norm": 0.8594211339950562, + "learning_rate": 4.4527384218829796e-05, + "loss": 0.9332, + "step": 15448 + }, + { + "epoch": 2.7505341880341883, + "grad_norm": 0.7828420996665955, + "learning_rate": 4.4515738403737585e-05, + "loss": 0.719, + "step": 15449 + }, + { + "epoch": 2.7507122507122506, + "grad_norm": 0.9195737242698669, + "learning_rate": 4.4504093675744285e-05, + "loss": 0.8581, + "step": 15450 + }, + { + "epoch": 2.7508903133903133, + "grad_norm": 0.8090249300003052, + "learning_rate": 4.449245003507793e-05, + "loss": 0.8012, + "step": 15451 + }, + { + "epoch": 2.751068376068376, + "grad_norm": 0.9179023504257202, + "learning_rate": 4.4480807481966736e-05, + "loss": 0.8968, + "step": 15452 + }, + { + "epoch": 2.751246438746439, + "grad_norm": 0.8140867352485657, + "learning_rate": 4.446916601663879e-05, + "loss": 0.791, + "step": 15453 + }, + { + "epoch": 2.7514245014245016, + "grad_norm": 0.8313645720481873, + "learning_rate": 4.445752563932214e-05, + "loss": 0.8658, + "step": 15454 + }, + { + "epoch": 2.751602564102564, + "grad_norm": 1.0235247611999512, + "learning_rate": 4.444588635024497e-05, + "loss": 0.9383, + "step": 15455 + }, + { + "epoch": 2.7517806267806266, + "grad_norm": 0.8270257711410522, + "learning_rate": 4.443424814963518e-05, + "loss": 0.7666, + "step": 15456 + }, + { + "epoch": 2.7519586894586894, + "grad_norm": 0.8711290955543518, + "learning_rate": 4.442261103772092e-05, + "loss": 1.0819, + "step": 15457 + }, + { + "epoch": 2.752136752136752, + "grad_norm": 0.7851848602294922, + "learning_rate": 4.441097501473013e-05, + "loss": 0.8788, + "step": 15458 + }, + { + "epoch": 2.752314814814815, + "grad_norm": 0.9703593850135803, + "learning_rate": 4.4399340080890816e-05, + "loss": 0.8661, + "step": 15459 + }, + { + "epoch": 2.7524928774928776, + "grad_norm": 0.8575040102005005, + "learning_rate": 4.438770623643093e-05, + "loss": 0.8318, + "step": 15460 + }, + { + "epoch": 2.7526709401709404, + "grad_norm": 0.9393935799598694, + "learning_rate": 4.43760734815784e-05, + "loss": 0.8899, + "step": 15461 + }, + { + "epoch": 2.7528490028490027, + "grad_norm": 0.9310712814331055, + "learning_rate": 4.4364441816561185e-05, + "loss": 0.9519, + "step": 15462 + }, + { + "epoch": 2.7530270655270654, + "grad_norm": 0.8066901564598083, + "learning_rate": 4.435281124160715e-05, + "loss": 0.8102, + "step": 15463 + }, + { + "epoch": 2.753205128205128, + "grad_norm": 0.8681934475898743, + "learning_rate": 4.434118175694415e-05, + "loss": 0.7745, + "step": 15464 + }, + { + "epoch": 2.753383190883191, + "grad_norm": 0.7921330332756042, + "learning_rate": 4.432955336280014e-05, + "loss": 0.8396, + "step": 15465 + }, + { + "epoch": 2.7535612535612537, + "grad_norm": 0.8818981051445007, + "learning_rate": 4.4317926059402816e-05, + "loss": 0.9268, + "step": 15466 + }, + { + "epoch": 2.753739316239316, + "grad_norm": 0.8018338084220886, + "learning_rate": 4.4306299846980096e-05, + "loss": 0.9018, + "step": 15467 + }, + { + "epoch": 2.7539173789173788, + "grad_norm": 0.8704143762588501, + "learning_rate": 4.4294674725759734e-05, + "loss": 0.9517, + "step": 15468 + }, + { + "epoch": 2.7540954415954415, + "grad_norm": 1.1460380554199219, + "learning_rate": 4.4283050695969506e-05, + "loss": 0.7642, + "step": 15469 + }, + { + "epoch": 2.7542735042735043, + "grad_norm": 0.8134510517120361, + "learning_rate": 4.427142775783716e-05, + "loss": 1.0405, + "step": 15470 + }, + { + "epoch": 2.754451566951567, + "grad_norm": 0.8054876327514648, + "learning_rate": 4.425980591159038e-05, + "loss": 0.8929, + "step": 15471 + }, + { + "epoch": 2.7546296296296298, + "grad_norm": 0.8607433438301086, + "learning_rate": 4.4248185157456953e-05, + "loss": 0.9187, + "step": 15472 + }, + { + "epoch": 2.7548076923076925, + "grad_norm": 0.7448357939720154, + "learning_rate": 4.423656549566453e-05, + "loss": 0.7298, + "step": 15473 + }, + { + "epoch": 2.754985754985755, + "grad_norm": 0.9228075742721558, + "learning_rate": 4.422494692644076e-05, + "loss": 0.8704, + "step": 15474 + }, + { + "epoch": 2.7551638176638176, + "grad_norm": 0.8781694173812866, + "learning_rate": 4.421332945001329e-05, + "loss": 0.9555, + "step": 15475 + }, + { + "epoch": 2.7553418803418803, + "grad_norm": 0.8632338643074036, + "learning_rate": 4.420171306660975e-05, + "loss": 0.7322, + "step": 15476 + }, + { + "epoch": 2.755519943019943, + "grad_norm": 0.9808199405670166, + "learning_rate": 4.4190097776457716e-05, + "loss": 0.619, + "step": 15477 + }, + { + "epoch": 2.755698005698006, + "grad_norm": 1.025109052658081, + "learning_rate": 4.41784835797848e-05, + "loss": 1.0617, + "step": 15478 + }, + { + "epoch": 2.755876068376068, + "grad_norm": 0.8132767677307129, + "learning_rate": 4.416687047681849e-05, + "loss": 1.0045, + "step": 15479 + }, + { + "epoch": 2.756054131054131, + "grad_norm": 0.9630453586578369, + "learning_rate": 4.415525846778645e-05, + "loss": 0.9576, + "step": 15480 + }, + { + "epoch": 2.7562321937321936, + "grad_norm": 0.9891922473907471, + "learning_rate": 4.4143647552916034e-05, + "loss": 0.9333, + "step": 15481 + }, + { + "epoch": 2.7564102564102564, + "grad_norm": 0.7974509000778198, + "learning_rate": 4.413203773243486e-05, + "loss": 0.7809, + "step": 15482 + }, + { + "epoch": 2.756588319088319, + "grad_norm": 0.8173473477363586, + "learning_rate": 4.412042900657034e-05, + "loss": 1.1023, + "step": 15483 + }, + { + "epoch": 2.756766381766382, + "grad_norm": 0.8502877950668335, + "learning_rate": 4.410882137554994e-05, + "loss": 0.8705, + "step": 15484 + }, + { + "epoch": 2.7569444444444446, + "grad_norm": 0.8519158959388733, + "learning_rate": 4.4097214839601074e-05, + "loss": 0.8901, + "step": 15485 + }, + { + "epoch": 2.7571225071225074, + "grad_norm": 0.7851125001907349, + "learning_rate": 4.4085609398951164e-05, + "loss": 1.0612, + "step": 15486 + }, + { + "epoch": 2.7573005698005697, + "grad_norm": 0.9585029482841492, + "learning_rate": 4.407400505382758e-05, + "loss": 0.9229, + "step": 15487 + }, + { + "epoch": 2.7574786324786325, + "grad_norm": 0.775071918964386, + "learning_rate": 4.4062401804457686e-05, + "loss": 0.8246, + "step": 15488 + }, + { + "epoch": 2.757656695156695, + "grad_norm": 0.8049488067626953, + "learning_rate": 4.405079965106881e-05, + "loss": 0.9681, + "step": 15489 + }, + { + "epoch": 2.757834757834758, + "grad_norm": 0.9452522993087769, + "learning_rate": 4.4039198593888306e-05, + "loss": 0.7288, + "step": 15490 + }, + { + "epoch": 2.7580128205128203, + "grad_norm": 0.8296085596084595, + "learning_rate": 4.402759863314346e-05, + "loss": 0.8053, + "step": 15491 + }, + { + "epoch": 2.758190883190883, + "grad_norm": 0.8086248636245728, + "learning_rate": 4.4015999769061556e-05, + "loss": 0.8692, + "step": 15492 + }, + { + "epoch": 2.7583689458689458, + "grad_norm": 0.8784860372543335, + "learning_rate": 4.4004402001869836e-05, + "loss": 1.0503, + "step": 15493 + }, + { + "epoch": 2.7585470085470085, + "grad_norm": 0.82901930809021, + "learning_rate": 4.399280533179551e-05, + "loss": 0.8479, + "step": 15494 + }, + { + "epoch": 2.7587250712250713, + "grad_norm": 0.7654509544372559, + "learning_rate": 4.3981209759065875e-05, + "loss": 0.743, + "step": 15495 + }, + { + "epoch": 2.758903133903134, + "grad_norm": 0.8240879774093628, + "learning_rate": 4.3969615283908e-05, + "loss": 0.8303, + "step": 15496 + }, + { + "epoch": 2.7590811965811968, + "grad_norm": 0.9411282539367676, + "learning_rate": 4.3958021906549195e-05, + "loss": 0.8217, + "step": 15497 + }, + { + "epoch": 2.7592592592592595, + "grad_norm": 0.8222329616546631, + "learning_rate": 4.394642962721647e-05, + "loss": 0.9596, + "step": 15498 + }, + { + "epoch": 2.759437321937322, + "grad_norm": 0.8462044596672058, + "learning_rate": 4.393483844613704e-05, + "loss": 0.8029, + "step": 15499 + }, + { + "epoch": 2.7596153846153846, + "grad_norm": 1.0385619401931763, + "learning_rate": 4.392324836353798e-05, + "loss": 1.0352, + "step": 15500 + }, + { + "epoch": 2.7597934472934473, + "grad_norm": 0.9049911499023438, + "learning_rate": 4.3911659379646384e-05, + "loss": 1.0761, + "step": 15501 + }, + { + "epoch": 2.75997150997151, + "grad_norm": 0.8253830671310425, + "learning_rate": 4.390007149468932e-05, + "loss": 0.7693, + "step": 15502 + }, + { + "epoch": 2.7601495726495724, + "grad_norm": 0.7939008474349976, + "learning_rate": 4.388848470889381e-05, + "loss": 0.8847, + "step": 15503 + }, + { + "epoch": 2.760327635327635, + "grad_norm": 1.048941969871521, + "learning_rate": 4.387689902248684e-05, + "loss": 0.7012, + "step": 15504 + }, + { + "epoch": 2.760505698005698, + "grad_norm": 0.8834842443466187, + "learning_rate": 4.386531443569553e-05, + "loss": 0.9561, + "step": 15505 + }, + { + "epoch": 2.7606837606837606, + "grad_norm": 0.9147583842277527, + "learning_rate": 4.385373094874669e-05, + "loss": 0.9736, + "step": 15506 + }, + { + "epoch": 2.7608618233618234, + "grad_norm": 0.6820386648178101, + "learning_rate": 4.38421485618674e-05, + "loss": 0.4449, + "step": 15507 + }, + { + "epoch": 2.761039886039886, + "grad_norm": 0.9519942402839661, + "learning_rate": 4.383056727528455e-05, + "loss": 1.0385, + "step": 15508 + }, + { + "epoch": 2.761217948717949, + "grad_norm": 0.8701474070549011, + "learning_rate": 4.381898708922505e-05, + "loss": 0.9896, + "step": 15509 + }, + { + "epoch": 2.7613960113960117, + "grad_norm": 0.8756018877029419, + "learning_rate": 4.38074080039158e-05, + "loss": 0.9378, + "step": 15510 + }, + { + "epoch": 2.761574074074074, + "grad_norm": 0.8670514225959778, + "learning_rate": 4.379583001958362e-05, + "loss": 0.9175, + "step": 15511 + }, + { + "epoch": 2.7617521367521367, + "grad_norm": 0.8227131366729736, + "learning_rate": 4.378425313645547e-05, + "loss": 0.7864, + "step": 15512 + }, + { + "epoch": 2.7619301994301995, + "grad_norm": 0.9121497273445129, + "learning_rate": 4.377267735475802e-05, + "loss": 0.877, + "step": 15513 + }, + { + "epoch": 2.762108262108262, + "grad_norm": 0.8347102999687195, + "learning_rate": 4.3761102674718205e-05, + "loss": 0.8223, + "step": 15514 + }, + { + "epoch": 2.7622863247863245, + "grad_norm": 0.8657951951026917, + "learning_rate": 4.374952909656275e-05, + "loss": 0.7117, + "step": 15515 + }, + { + "epoch": 2.7624643874643873, + "grad_norm": 0.8934728503227234, + "learning_rate": 4.3737956620518414e-05, + "loss": 0.84, + "step": 15516 + }, + { + "epoch": 2.76264245014245, + "grad_norm": 0.7592045068740845, + "learning_rate": 4.3726385246811964e-05, + "loss": 0.7856, + "step": 15517 + }, + { + "epoch": 2.7628205128205128, + "grad_norm": 0.8480674028396606, + "learning_rate": 4.371481497567008e-05, + "loss": 0.7805, + "step": 15518 + }, + { + "epoch": 2.7629985754985755, + "grad_norm": 1.0231767892837524, + "learning_rate": 4.3703245807319437e-05, + "loss": 1.1517, + "step": 15519 + }, + { + "epoch": 2.7631766381766383, + "grad_norm": 0.9852092862129211, + "learning_rate": 4.369167774198684e-05, + "loss": 0.8735, + "step": 15520 + }, + { + "epoch": 2.763354700854701, + "grad_norm": 0.8751610517501831, + "learning_rate": 4.368011077989875e-05, + "loss": 0.9975, + "step": 15521 + }, + { + "epoch": 2.763532763532764, + "grad_norm": 0.8397828340530396, + "learning_rate": 4.3668544921281976e-05, + "loss": 1.0675, + "step": 15522 + }, + { + "epoch": 2.763710826210826, + "grad_norm": 0.7040372490882874, + "learning_rate": 4.3656980166362974e-05, + "loss": 0.7123, + "step": 15523 + }, + { + "epoch": 2.763888888888889, + "grad_norm": 1.0610599517822266, + "learning_rate": 4.364541651536844e-05, + "loss": 1.0854, + "step": 15524 + }, + { + "epoch": 2.7640669515669516, + "grad_norm": 0.78865647315979, + "learning_rate": 4.363385396852491e-05, + "loss": 0.924, + "step": 15525 + }, + { + "epoch": 2.7642450142450143, + "grad_norm": 0.87164705991745, + "learning_rate": 4.362229252605891e-05, + "loss": 0.9739, + "step": 15526 + }, + { + "epoch": 2.7644230769230766, + "grad_norm": 0.9362281560897827, + "learning_rate": 4.361073218819698e-05, + "loss": 0.7751, + "step": 15527 + }, + { + "epoch": 2.7646011396011394, + "grad_norm": 0.7944566011428833, + "learning_rate": 4.3599172955165605e-05, + "loss": 0.8913, + "step": 15528 + }, + { + "epoch": 2.764779202279202, + "grad_norm": 0.9346068501472473, + "learning_rate": 4.358761482719125e-05, + "loss": 0.8286, + "step": 15529 + }, + { + "epoch": 2.764957264957265, + "grad_norm": 0.8570913076400757, + "learning_rate": 4.3576057804500414e-05, + "loss": 1.0334, + "step": 15530 + }, + { + "epoch": 2.7651353276353277, + "grad_norm": 0.801908552646637, + "learning_rate": 4.356450188731953e-05, + "loss": 0.9021, + "step": 15531 + }, + { + "epoch": 2.7653133903133904, + "grad_norm": 0.848849892616272, + "learning_rate": 4.355294707587499e-05, + "loss": 0.9132, + "step": 15532 + }, + { + "epoch": 2.765491452991453, + "grad_norm": 0.7961751818656921, + "learning_rate": 4.35413933703932e-05, + "loss": 0.841, + "step": 15533 + }, + { + "epoch": 2.765669515669516, + "grad_norm": 0.8609708547592163, + "learning_rate": 4.352984077110052e-05, + "loss": 0.8176, + "step": 15534 + }, + { + "epoch": 2.765847578347578, + "grad_norm": 0.8779369592666626, + "learning_rate": 4.35182892782233e-05, + "loss": 0.8252, + "step": 15535 + }, + { + "epoch": 2.766025641025641, + "grad_norm": 0.7878577709197998, + "learning_rate": 4.3506738891987844e-05, + "loss": 0.7498, + "step": 15536 + }, + { + "epoch": 2.7662037037037037, + "grad_norm": 0.9531580805778503, + "learning_rate": 4.3495189612620557e-05, + "loss": 0.8438, + "step": 15537 + }, + { + "epoch": 2.7663817663817665, + "grad_norm": 0.7791294455528259, + "learning_rate": 4.3483641440347564e-05, + "loss": 0.9188, + "step": 15538 + }, + { + "epoch": 2.7665598290598292, + "grad_norm": 0.8683488965034485, + "learning_rate": 4.347209437539527e-05, + "loss": 1.0422, + "step": 15539 + }, + { + "epoch": 2.7667378917378915, + "grad_norm": 0.8904309272766113, + "learning_rate": 4.346054841798984e-05, + "loss": 0.8072, + "step": 15540 + }, + { + "epoch": 2.7669159544159543, + "grad_norm": 0.7409844398498535, + "learning_rate": 4.344900356835753e-05, + "loss": 0.7179, + "step": 15541 + }, + { + "epoch": 2.767094017094017, + "grad_norm": 0.9663724899291992, + "learning_rate": 4.343745982672451e-05, + "loss": 1.0568, + "step": 15542 + }, + { + "epoch": 2.76727207977208, + "grad_norm": 0.8481591939926147, + "learning_rate": 4.342591719331698e-05, + "loss": 0.8678, + "step": 15543 + }, + { + "epoch": 2.7674501424501425, + "grad_norm": 0.7301938533782959, + "learning_rate": 4.341437566836103e-05, + "loss": 0.6357, + "step": 15544 + }, + { + "epoch": 2.7676282051282053, + "grad_norm": 0.8628479242324829, + "learning_rate": 4.340283525208292e-05, + "loss": 0.9622, + "step": 15545 + }, + { + "epoch": 2.767806267806268, + "grad_norm": 0.953744113445282, + "learning_rate": 4.339129594470861e-05, + "loss": 0.683, + "step": 15546 + }, + { + "epoch": 2.7679843304843303, + "grad_norm": 0.7589353322982788, + "learning_rate": 4.3379757746464336e-05, + "loss": 0.8343, + "step": 15547 + }, + { + "epoch": 2.768162393162393, + "grad_norm": 0.8304651379585266, + "learning_rate": 4.336822065757601e-05, + "loss": 0.9084, + "step": 15548 + }, + { + "epoch": 2.768340455840456, + "grad_norm": 0.8092817068099976, + "learning_rate": 4.33566846782698e-05, + "loss": 0.8371, + "step": 15549 + }, + { + "epoch": 2.7685185185185186, + "grad_norm": 0.8983978033065796, + "learning_rate": 4.334514980877169e-05, + "loss": 0.9631, + "step": 15550 + }, + { + "epoch": 2.7686965811965814, + "grad_norm": 0.766621470451355, + "learning_rate": 4.3333616049307636e-05, + "loss": 0.768, + "step": 15551 + }, + { + "epoch": 2.7688746438746437, + "grad_norm": 0.8753345012664795, + "learning_rate": 4.332208340010374e-05, + "loss": 0.8854, + "step": 15552 + }, + { + "epoch": 2.7690527065527064, + "grad_norm": 0.8396589756011963, + "learning_rate": 4.331055186138581e-05, + "loss": 0.8322, + "step": 15553 + }, + { + "epoch": 2.769230769230769, + "grad_norm": 0.8134872317314148, + "learning_rate": 4.3299021433379885e-05, + "loss": 0.945, + "step": 15554 + }, + { + "epoch": 2.769408831908832, + "grad_norm": 0.8712667226791382, + "learning_rate": 4.3287492116311854e-05, + "loss": 0.8487, + "step": 15555 + }, + { + "epoch": 2.7695868945868947, + "grad_norm": 0.8938018083572388, + "learning_rate": 4.32759639104076e-05, + "loss": 0.854, + "step": 15556 + }, + { + "epoch": 2.7697649572649574, + "grad_norm": 0.8213987946510315, + "learning_rate": 4.3264436815893005e-05, + "loss": 0.9055, + "step": 15557 + }, + { + "epoch": 2.76994301994302, + "grad_norm": 0.9587214589118958, + "learning_rate": 4.3252910832993906e-05, + "loss": 0.9664, + "step": 15558 + }, + { + "epoch": 2.7701210826210825, + "grad_norm": 0.8746159076690674, + "learning_rate": 4.3241385961936146e-05, + "loss": 0.8394, + "step": 15559 + }, + { + "epoch": 2.7702991452991452, + "grad_norm": 0.8395819067955017, + "learning_rate": 4.3229862202945517e-05, + "loss": 0.7991, + "step": 15560 + }, + { + "epoch": 2.770477207977208, + "grad_norm": 0.856765627861023, + "learning_rate": 4.321833955624777e-05, + "loss": 0.8106, + "step": 15561 + }, + { + "epoch": 2.7706552706552707, + "grad_norm": 0.8630124926567078, + "learning_rate": 4.3206818022068776e-05, + "loss": 0.7293, + "step": 15562 + }, + { + "epoch": 2.7708333333333335, + "grad_norm": 0.8196776509284973, + "learning_rate": 4.319529760063414e-05, + "loss": 0.8831, + "step": 15563 + }, + { + "epoch": 2.771011396011396, + "grad_norm": 0.8283860683441162, + "learning_rate": 4.3183778292169674e-05, + "loss": 0.9249, + "step": 15564 + }, + { + "epoch": 2.7711894586894585, + "grad_norm": 0.8983619809150696, + "learning_rate": 4.3172260096901054e-05, + "loss": 1.2334, + "step": 15565 + }, + { + "epoch": 2.7713675213675213, + "grad_norm": 0.8437079191207886, + "learning_rate": 4.316074301505395e-05, + "loss": 0.771, + "step": 15566 + }, + { + "epoch": 2.771545584045584, + "grad_norm": 0.9565808773040771, + "learning_rate": 4.314922704685401e-05, + "loss": 0.8927, + "step": 15567 + }, + { + "epoch": 2.771723646723647, + "grad_norm": 0.7943497896194458, + "learning_rate": 4.313771219252687e-05, + "loss": 0.656, + "step": 15568 + }, + { + "epoch": 2.7719017094017095, + "grad_norm": 0.862404465675354, + "learning_rate": 4.3126198452298126e-05, + "loss": 0.8783, + "step": 15569 + }, + { + "epoch": 2.7720797720797723, + "grad_norm": 0.7928122878074646, + "learning_rate": 4.3114685826393365e-05, + "loss": 0.9799, + "step": 15570 + }, + { + "epoch": 2.7722578347578346, + "grad_norm": 0.8270733952522278, + "learning_rate": 4.3103174315038184e-05, + "loss": 0.7878, + "step": 15571 + }, + { + "epoch": 2.7724358974358974, + "grad_norm": 0.8223987817764282, + "learning_rate": 4.309166391845811e-05, + "loss": 0.7222, + "step": 15572 + }, + { + "epoch": 2.77261396011396, + "grad_norm": 0.8159852027893066, + "learning_rate": 4.3080154636878675e-05, + "loss": 0.8082, + "step": 15573 + }, + { + "epoch": 2.772792022792023, + "grad_norm": 0.882792055606842, + "learning_rate": 4.306864647052537e-05, + "loss": 1.0659, + "step": 15574 + }, + { + "epoch": 2.7729700854700856, + "grad_norm": 0.8734562993049622, + "learning_rate": 4.305713941962366e-05, + "loss": 0.9301, + "step": 15575 + }, + { + "epoch": 2.773148148148148, + "grad_norm": 0.8544983267784119, + "learning_rate": 4.304563348439898e-05, + "loss": 0.7442, + "step": 15576 + }, + { + "epoch": 2.7733262108262107, + "grad_norm": 0.9045799374580383, + "learning_rate": 4.303412866507689e-05, + "loss": 1.1023, + "step": 15577 + }, + { + "epoch": 2.7735042735042734, + "grad_norm": 0.8132993578910828, + "learning_rate": 4.3022624961882615e-05, + "loss": 0.7032, + "step": 15578 + }, + { + "epoch": 2.773682336182336, + "grad_norm": 0.7072446942329407, + "learning_rate": 4.30111223750417e-05, + "loss": 0.7176, + "step": 15579 + }, + { + "epoch": 2.773860398860399, + "grad_norm": 0.8212466239929199, + "learning_rate": 4.299962090477945e-05, + "loss": 0.7664, + "step": 15580 + }, + { + "epoch": 2.7740384615384617, + "grad_norm": 0.7781338095664978, + "learning_rate": 4.298812055132122e-05, + "loss": 0.7439, + "step": 15581 + }, + { + "epoch": 2.7742165242165244, + "grad_norm": 0.9289973378181458, + "learning_rate": 4.297662131489234e-05, + "loss": 0.9504, + "step": 15582 + }, + { + "epoch": 2.7743945868945867, + "grad_norm": 0.8571373224258423, + "learning_rate": 4.2965123195718105e-05, + "loss": 0.9959, + "step": 15583 + }, + { + "epoch": 2.7745726495726495, + "grad_norm": 0.8670883774757385, + "learning_rate": 4.29536261940238e-05, + "loss": 0.8207, + "step": 15584 + }, + { + "epoch": 2.7747507122507122, + "grad_norm": 0.8684807419776917, + "learning_rate": 4.294213031003469e-05, + "loss": 0.7508, + "step": 15585 + }, + { + "epoch": 2.774928774928775, + "grad_norm": 0.7746252417564392, + "learning_rate": 4.293063554397597e-05, + "loss": 0.909, + "step": 15586 + }, + { + "epoch": 2.7751068376068377, + "grad_norm": 0.8363521099090576, + "learning_rate": 4.291914189607297e-05, + "loss": 0.8564, + "step": 15587 + }, + { + "epoch": 2.7752849002849, + "grad_norm": 0.8843217492103577, + "learning_rate": 4.2907649366550726e-05, + "loss": 0.8187, + "step": 15588 + }, + { + "epoch": 2.775462962962963, + "grad_norm": 0.9330897331237793, + "learning_rate": 4.2896157955634545e-05, + "loss": 0.8179, + "step": 15589 + }, + { + "epoch": 2.7756410256410255, + "grad_norm": 0.7950356602668762, + "learning_rate": 4.288466766354953e-05, + "loss": 0.7091, + "step": 15590 + }, + { + "epoch": 2.7758190883190883, + "grad_norm": 0.9085933566093445, + "learning_rate": 4.287317849052075e-05, + "loss": 0.8015, + "step": 15591 + }, + { + "epoch": 2.775997150997151, + "grad_norm": 0.9285191893577576, + "learning_rate": 4.286169043677345e-05, + "loss": 0.9967, + "step": 15592 + }, + { + "epoch": 2.776175213675214, + "grad_norm": 0.8202041387557983, + "learning_rate": 4.285020350253256e-05, + "loss": 0.9286, + "step": 15593 + }, + { + "epoch": 2.7763532763532766, + "grad_norm": 1.0619434118270874, + "learning_rate": 4.283871768802328e-05, + "loss": 0.7863, + "step": 15594 + }, + { + "epoch": 2.7765313390313393, + "grad_norm": 0.8250051140785217, + "learning_rate": 4.282723299347052e-05, + "loss": 0.8531, + "step": 15595 + }, + { + "epoch": 2.7767094017094016, + "grad_norm": 0.8794218897819519, + "learning_rate": 4.281574941909939e-05, + "loss": 0.906, + "step": 15596 + }, + { + "epoch": 2.7768874643874644, + "grad_norm": 0.7725922465324402, + "learning_rate": 4.2804266965134866e-05, + "loss": 0.9084, + "step": 15597 + }, + { + "epoch": 2.777065527065527, + "grad_norm": 0.7845144867897034, + "learning_rate": 4.279278563180192e-05, + "loss": 0.9768, + "step": 15598 + }, + { + "epoch": 2.77724358974359, + "grad_norm": 0.9395498633384705, + "learning_rate": 4.27813054193255e-05, + "loss": 0.9055, + "step": 15599 + }, + { + "epoch": 2.777421652421652, + "grad_norm": 0.8043427467346191, + "learning_rate": 4.276982632793054e-05, + "loss": 0.8244, + "step": 15600 + }, + { + "epoch": 2.777599715099715, + "grad_norm": 0.7874096632003784, + "learning_rate": 4.27583483578419e-05, + "loss": 0.8861, + "step": 15601 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 0.8874611258506775, + "learning_rate": 4.27468715092846e-05, + "loss": 1.0457, + "step": 15602 + }, + { + "epoch": 2.7779558404558404, + "grad_norm": 1.0025757551193237, + "learning_rate": 4.273539578248334e-05, + "loss": 1.1114, + "step": 15603 + }, + { + "epoch": 2.778133903133903, + "grad_norm": 0.9982876777648926, + "learning_rate": 4.272392117766313e-05, + "loss": 0.9142, + "step": 15604 + }, + { + "epoch": 2.778311965811966, + "grad_norm": 0.8762221932411194, + "learning_rate": 4.2712447695048616e-05, + "loss": 1.0114, + "step": 15605 + }, + { + "epoch": 2.7784900284900287, + "grad_norm": 0.9136927723884583, + "learning_rate": 4.2700975334864726e-05, + "loss": 0.8224, + "step": 15606 + }, + { + "epoch": 2.7786680911680914, + "grad_norm": 0.8845604062080383, + "learning_rate": 4.2689504097336184e-05, + "loss": 0.8135, + "step": 15607 + }, + { + "epoch": 2.7788461538461537, + "grad_norm": 0.8584510087966919, + "learning_rate": 4.267803398268777e-05, + "loss": 0.8928, + "step": 15608 + }, + { + "epoch": 2.7790242165242165, + "grad_norm": 0.711402177810669, + "learning_rate": 4.266656499114421e-05, + "loss": 0.792, + "step": 15609 + }, + { + "epoch": 2.7792022792022792, + "grad_norm": 0.9480760097503662, + "learning_rate": 4.2655097122930165e-05, + "loss": 0.8707, + "step": 15610 + }, + { + "epoch": 2.779380341880342, + "grad_norm": 0.863855242729187, + "learning_rate": 4.264363037827041e-05, + "loss": 1.0114, + "step": 15611 + }, + { + "epoch": 2.7795584045584043, + "grad_norm": 0.8010865449905396, + "learning_rate": 4.2632164757389556e-05, + "loss": 0.9221, + "step": 15612 + }, + { + "epoch": 2.779736467236467, + "grad_norm": 0.7950930595397949, + "learning_rate": 4.262070026051227e-05, + "loss": 0.6951, + "step": 15613 + }, + { + "epoch": 2.77991452991453, + "grad_norm": 0.8252870440483093, + "learning_rate": 4.260923688786317e-05, + "loss": 0.6976, + "step": 15614 + }, + { + "epoch": 2.7800925925925926, + "grad_norm": 0.7855920791625977, + "learning_rate": 4.259777463966686e-05, + "loss": 0.8277, + "step": 15615 + }, + { + "epoch": 2.7802706552706553, + "grad_norm": 0.8783130645751953, + "learning_rate": 4.258631351614786e-05, + "loss": 0.6995, + "step": 15616 + }, + { + "epoch": 2.780448717948718, + "grad_norm": 0.8904485106468201, + "learning_rate": 4.257485351753085e-05, + "loss": 0.8226, + "step": 15617 + }, + { + "epoch": 2.780626780626781, + "grad_norm": 0.8761011958122253, + "learning_rate": 4.2563394644040244e-05, + "loss": 0.9187, + "step": 15618 + }, + { + "epoch": 2.7808048433048436, + "grad_norm": 0.897404670715332, + "learning_rate": 4.255193689590067e-05, + "loss": 1.0234, + "step": 15619 + }, + { + "epoch": 2.780982905982906, + "grad_norm": 0.8966960906982422, + "learning_rate": 4.254048027333648e-05, + "loss": 0.9, + "step": 15620 + }, + { + "epoch": 2.7811609686609686, + "grad_norm": 0.7506237030029297, + "learning_rate": 4.2529024776572245e-05, + "loss": 0.8939, + "step": 15621 + }, + { + "epoch": 2.7813390313390314, + "grad_norm": 0.8073886036872864, + "learning_rate": 4.2517570405832396e-05, + "loss": 0.7779, + "step": 15622 + }, + { + "epoch": 2.781517094017094, + "grad_norm": 0.7928911447525024, + "learning_rate": 4.250611716134134e-05, + "loss": 0.8278, + "step": 15623 + }, + { + "epoch": 2.7816951566951564, + "grad_norm": 0.7301982045173645, + "learning_rate": 4.249466504332349e-05, + "loss": 0.7515, + "step": 15624 + }, + { + "epoch": 2.781873219373219, + "grad_norm": 0.8215289115905762, + "learning_rate": 4.248321405200322e-05, + "loss": 0.9752, + "step": 15625 + }, + { + "epoch": 2.782051282051282, + "grad_norm": 0.8281431198120117, + "learning_rate": 4.247176418760486e-05, + "loss": 0.9625, + "step": 15626 + }, + { + "epoch": 2.7822293447293447, + "grad_norm": 0.9202759265899658, + "learning_rate": 4.246031545035283e-05, + "loss": 0.8757, + "step": 15627 + }, + { + "epoch": 2.7824074074074074, + "grad_norm": 0.8628471493721008, + "learning_rate": 4.244886784047133e-05, + "loss": 0.7626, + "step": 15628 + }, + { + "epoch": 2.78258547008547, + "grad_norm": 0.9345491528511047, + "learning_rate": 4.2437421358184747e-05, + "loss": 0.8714, + "step": 15629 + }, + { + "epoch": 2.782763532763533, + "grad_norm": 0.893713116645813, + "learning_rate": 4.2425976003717314e-05, + "loss": 0.9953, + "step": 15630 + }, + { + "epoch": 2.7829415954415957, + "grad_norm": 0.8794371485710144, + "learning_rate": 4.2414531777293286e-05, + "loss": 0.7899, + "step": 15631 + }, + { + "epoch": 2.783119658119658, + "grad_norm": 0.9003345370292664, + "learning_rate": 4.240308867913688e-05, + "loss": 0.9607, + "step": 15632 + }, + { + "epoch": 2.7832977207977208, + "grad_norm": 0.8352270126342773, + "learning_rate": 4.239164670947228e-05, + "loss": 1.0315, + "step": 15633 + }, + { + "epoch": 2.7834757834757835, + "grad_norm": 0.825252890586853, + "learning_rate": 4.238020586852375e-05, + "loss": 1.0493, + "step": 15634 + }, + { + "epoch": 2.7836538461538463, + "grad_norm": 1.1185758113861084, + "learning_rate": 4.2368766156515324e-05, + "loss": 0.8962, + "step": 15635 + }, + { + "epoch": 2.7838319088319086, + "grad_norm": 0.840336799621582, + "learning_rate": 4.235732757367125e-05, + "loss": 0.8289, + "step": 15636 + }, + { + "epoch": 2.7840099715099713, + "grad_norm": 0.9251887202262878, + "learning_rate": 4.2345890120215595e-05, + "loss": 0.9306, + "step": 15637 + }, + { + "epoch": 2.784188034188034, + "grad_norm": 0.9645969867706299, + "learning_rate": 4.233445379637244e-05, + "loss": 0.8453, + "step": 15638 + }, + { + "epoch": 2.784366096866097, + "grad_norm": 0.9010009765625, + "learning_rate": 4.232301860236589e-05, + "loss": 0.9796, + "step": 15639 + }, + { + "epoch": 2.7845441595441596, + "grad_norm": 0.920427143573761, + "learning_rate": 4.231158453841998e-05, + "loss": 0.7905, + "step": 15640 + }, + { + "epoch": 2.7847222222222223, + "grad_norm": 0.8292316794395447, + "learning_rate": 4.2300151604758734e-05, + "loss": 0.938, + "step": 15641 + }, + { + "epoch": 2.784900284900285, + "grad_norm": 0.8550885319709778, + "learning_rate": 4.228871980160615e-05, + "loss": 0.7728, + "step": 15642 + }, + { + "epoch": 2.785078347578348, + "grad_norm": 0.8785567283630371, + "learning_rate": 4.227728912918617e-05, + "loss": 1.0367, + "step": 15643 + }, + { + "epoch": 2.78525641025641, + "grad_norm": 0.8732814788818359, + "learning_rate": 4.226585958772289e-05, + "loss": 0.9914, + "step": 15644 + }, + { + "epoch": 2.785434472934473, + "grad_norm": 0.9473167061805725, + "learning_rate": 4.225443117744008e-05, + "loss": 1.1311, + "step": 15645 + }, + { + "epoch": 2.7856125356125356, + "grad_norm": 0.8819913864135742, + "learning_rate": 4.224300389856177e-05, + "loss": 0.8798, + "step": 15646 + }, + { + "epoch": 2.7857905982905984, + "grad_norm": 0.835367739200592, + "learning_rate": 4.223157775131182e-05, + "loss": 0.6977, + "step": 15647 + }, + { + "epoch": 2.7859686609686607, + "grad_norm": 0.8122659921646118, + "learning_rate": 4.222015273591411e-05, + "loss": 0.9656, + "step": 15648 + }, + { + "epoch": 2.7861467236467234, + "grad_norm": 0.8085313439369202, + "learning_rate": 4.220872885259247e-05, + "loss": 0.7456, + "step": 15649 + }, + { + "epoch": 2.786324786324786, + "grad_norm": 0.681515097618103, + "learning_rate": 4.21973061015707e-05, + "loss": 0.5008, + "step": 15650 + }, + { + "epoch": 2.786502849002849, + "grad_norm": 0.8021831512451172, + "learning_rate": 4.2185884483072676e-05, + "loss": 0.8954, + "step": 15651 + }, + { + "epoch": 2.7866809116809117, + "grad_norm": 0.9254723787307739, + "learning_rate": 4.217446399732216e-05, + "loss": 0.7855, + "step": 15652 + }, + { + "epoch": 2.7868589743589745, + "grad_norm": 0.8415037989616394, + "learning_rate": 4.2163044644542894e-05, + "loss": 0.8835, + "step": 15653 + }, + { + "epoch": 2.787037037037037, + "grad_norm": 0.9031959772109985, + "learning_rate": 4.2151626424958614e-05, + "loss": 1.0048, + "step": 15654 + }, + { + "epoch": 2.7872150997151, + "grad_norm": Infinity, + "learning_rate": 4.2151626424958614e-05, + "loss": 0.9344, + "step": 15655 + }, + { + "epoch": 2.7873931623931623, + "grad_norm": 0.815680742263794, + "learning_rate": 4.214020933879306e-05, + "loss": 0.6897, + "step": 15656 + }, + { + "epoch": 2.787571225071225, + "grad_norm": 0.9080044627189636, + "learning_rate": 4.212879338626989e-05, + "loss": 1.0366, + "step": 15657 + }, + { + "epoch": 2.7877492877492878, + "grad_norm": 0.8387414813041687, + "learning_rate": 4.211737856761281e-05, + "loss": 0.9255, + "step": 15658 + }, + { + "epoch": 2.7879273504273505, + "grad_norm": 0.9269571304321289, + "learning_rate": 4.210596488304542e-05, + "loss": 0.8971, + "step": 15659 + }, + { + "epoch": 2.7881054131054133, + "grad_norm": 0.7987017035484314, + "learning_rate": 4.2094552332791456e-05, + "loss": 0.8293, + "step": 15660 + }, + { + "epoch": 2.7882834757834756, + "grad_norm": 0.8481683731079102, + "learning_rate": 4.208314091707437e-05, + "loss": 0.9159, + "step": 15661 + }, + { + "epoch": 2.7884615384615383, + "grad_norm": 0.944736897945404, + "learning_rate": 4.207173063611788e-05, + "loss": 0.9398, + "step": 15662 + }, + { + "epoch": 2.788639601139601, + "grad_norm": 0.8471882343292236, + "learning_rate": 4.206032149014547e-05, + "loss": 0.7534, + "step": 15663 + }, + { + "epoch": 2.788817663817664, + "grad_norm": 0.8500807881355286, + "learning_rate": 4.2048913479380714e-05, + "loss": 0.8874, + "step": 15664 + }, + { + "epoch": 2.7889957264957266, + "grad_norm": 0.7949451804161072, + "learning_rate": 4.2037506604047115e-05, + "loss": 0.8691, + "step": 15665 + }, + { + "epoch": 2.7891737891737893, + "grad_norm": 0.8587945103645325, + "learning_rate": 4.202610086436817e-05, + "loss": 0.8288, + "step": 15666 + }, + { + "epoch": 2.789351851851852, + "grad_norm": 0.9155020117759705, + "learning_rate": 4.201469626056734e-05, + "loss": 0.8384, + "step": 15667 + }, + { + "epoch": 2.7895299145299144, + "grad_norm": 0.9402222037315369, + "learning_rate": 4.200329279286809e-05, + "loss": 0.8239, + "step": 15668 + }, + { + "epoch": 2.789707977207977, + "grad_norm": 0.9111437201499939, + "learning_rate": 4.19918904614938e-05, + "loss": 0.9251, + "step": 15669 + }, + { + "epoch": 2.78988603988604, + "grad_norm": 0.9434856176376343, + "learning_rate": 4.198048926666795e-05, + "loss": 1.0517, + "step": 15670 + }, + { + "epoch": 2.7900641025641026, + "grad_norm": 0.9518313407897949, + "learning_rate": 4.1969089208613896e-05, + "loss": 0.8893, + "step": 15671 + }, + { + "epoch": 2.7902421652421654, + "grad_norm": 0.8107752799987793, + "learning_rate": 4.1957690287554986e-05, + "loss": 0.6548, + "step": 15672 + }, + { + "epoch": 2.7904202279202277, + "grad_norm": 0.8361678719520569, + "learning_rate": 4.1946292503714556e-05, + "loss": 0.9224, + "step": 15673 + }, + { + "epoch": 2.7905982905982905, + "grad_norm": 0.7812657952308655, + "learning_rate": 4.1934895857315904e-05, + "loss": 0.7126, + "step": 15674 + }, + { + "epoch": 2.790776353276353, + "grad_norm": 0.9054265022277832, + "learning_rate": 4.192350034858241e-05, + "loss": 0.6891, + "step": 15675 + }, + { + "epoch": 2.790954415954416, + "grad_norm": 0.9675585627555847, + "learning_rate": 4.1912105977737214e-05, + "loss": 0.8429, + "step": 15676 + }, + { + "epoch": 2.7911324786324787, + "grad_norm": 0.9077114462852478, + "learning_rate": 4.19007127450037e-05, + "loss": 0.8864, + "step": 15677 + }, + { + "epoch": 2.7913105413105415, + "grad_norm": 0.9230541586875916, + "learning_rate": 4.188932065060497e-05, + "loss": 1.0065, + "step": 15678 + }, + { + "epoch": 2.791488603988604, + "grad_norm": 0.8667981028556824, + "learning_rate": 4.1877929694764315e-05, + "loss": 0.7584, + "step": 15679 + }, + { + "epoch": 2.7916666666666665, + "grad_norm": 0.8986212015151978, + "learning_rate": 4.1866539877704894e-05, + "loss": 0.9205, + "step": 15680 + }, + { + "epoch": 2.7918447293447293, + "grad_norm": 0.8524685502052307, + "learning_rate": 4.185515119964986e-05, + "loss": 0.8516, + "step": 15681 + }, + { + "epoch": 2.792022792022792, + "grad_norm": 0.8247089385986328, + "learning_rate": 4.184376366082234e-05, + "loss": 0.8733, + "step": 15682 + }, + { + "epoch": 2.7922008547008548, + "grad_norm": 0.8236528635025024, + "learning_rate": 4.183237726144549e-05, + "loss": 0.8715, + "step": 15683 + }, + { + "epoch": 2.7923789173789175, + "grad_norm": 0.8853272199630737, + "learning_rate": 4.182099200174232e-05, + "loss": 0.7741, + "step": 15684 + }, + { + "epoch": 2.79255698005698, + "grad_norm": 0.8243789672851562, + "learning_rate": 4.180960788193603e-05, + "loss": 0.9196, + "step": 15685 + }, + { + "epoch": 2.7927350427350426, + "grad_norm": 0.9670386910438538, + "learning_rate": 4.1798224902249515e-05, + "loss": 0.828, + "step": 15686 + }, + { + "epoch": 2.7929131054131053, + "grad_norm": 0.7831283211708069, + "learning_rate": 4.178684306290592e-05, + "loss": 0.8389, + "step": 15687 + }, + { + "epoch": 2.793091168091168, + "grad_norm": 0.9372588396072388, + "learning_rate": 4.177546236412822e-05, + "loss": 1.2267, + "step": 15688 + }, + { + "epoch": 2.793269230769231, + "grad_norm": 0.9065600633621216, + "learning_rate": 4.176408280613937e-05, + "loss": 0.9674, + "step": 15689 + }, + { + "epoch": 2.7934472934472936, + "grad_norm": 0.8220530152320862, + "learning_rate": 4.1752704389162344e-05, + "loss": 0.8717, + "step": 15690 + }, + { + "epoch": 2.7936253561253563, + "grad_norm": 0.8952174782752991, + "learning_rate": 4.174132711342005e-05, + "loss": 0.8904, + "step": 15691 + }, + { + "epoch": 2.7938034188034186, + "grad_norm": 0.8454076647758484, + "learning_rate": 4.172995097913549e-05, + "loss": 0.9784, + "step": 15692 + }, + { + "epoch": 2.7939814814814814, + "grad_norm": 0.8697866797447205, + "learning_rate": 4.171857598653143e-05, + "loss": 1.0042, + "step": 15693 + }, + { + "epoch": 2.794159544159544, + "grad_norm": 0.8736211657524109, + "learning_rate": 4.170720213583084e-05, + "loss": 0.8787, + "step": 15694 + }, + { + "epoch": 2.794337606837607, + "grad_norm": 1.0082578659057617, + "learning_rate": 4.1695829427256525e-05, + "loss": 1.2508, + "step": 15695 + }, + { + "epoch": 2.7945156695156697, + "grad_norm": 0.8092042207717896, + "learning_rate": 4.1684457861031325e-05, + "loss": 0.8968, + "step": 15696 + }, + { + "epoch": 2.794693732193732, + "grad_norm": 0.847034752368927, + "learning_rate": 4.167308743737802e-05, + "loss": 0.8019, + "step": 15697 + }, + { + "epoch": 2.7948717948717947, + "grad_norm": 0.9059078097343445, + "learning_rate": 4.1661718156519414e-05, + "loss": 1.0393, + "step": 15698 + }, + { + "epoch": 2.7950498575498575, + "grad_norm": 0.8907228112220764, + "learning_rate": 4.165035001867822e-05, + "loss": 0.7388, + "step": 15699 + }, + { + "epoch": 2.79522792022792, + "grad_norm": 0.8089052438735962, + "learning_rate": 4.163898302407727e-05, + "loss": 0.8052, + "step": 15700 + }, + { + "epoch": 2.795405982905983, + "grad_norm": 1.1158883571624756, + "learning_rate": 4.162761717293915e-05, + "loss": 0.6923, + "step": 15701 + }, + { + "epoch": 2.7955840455840457, + "grad_norm": 0.8157755732536316, + "learning_rate": 4.1616252465486684e-05, + "loss": 0.606, + "step": 15702 + }, + { + "epoch": 2.7957621082621085, + "grad_norm": 0.8905386328697205, + "learning_rate": 4.1604888901942386e-05, + "loss": 0.7798, + "step": 15703 + }, + { + "epoch": 2.7959401709401708, + "grad_norm": 0.7655990719795227, + "learning_rate": 4.1593526482529034e-05, + "loss": 0.7435, + "step": 15704 + }, + { + "epoch": 2.7961182336182335, + "grad_norm": 0.900643527507782, + "learning_rate": 4.1582165207469195e-05, + "loss": 0.8314, + "step": 15705 + }, + { + "epoch": 2.7962962962962963, + "grad_norm": 0.7737550735473633, + "learning_rate": 4.1570805076985475e-05, + "loss": 0.8281, + "step": 15706 + }, + { + "epoch": 2.796474358974359, + "grad_norm": 0.8385021090507507, + "learning_rate": 4.1559446091300455e-05, + "loss": 0.896, + "step": 15707 + }, + { + "epoch": 2.796652421652422, + "grad_norm": 0.8830214142799377, + "learning_rate": 4.1548088250636687e-05, + "loss": 0.8856, + "step": 15708 + }, + { + "epoch": 2.796830484330484, + "grad_norm": 0.9748533368110657, + "learning_rate": 4.1536731555216676e-05, + "loss": 0.8768, + "step": 15709 + }, + { + "epoch": 2.797008547008547, + "grad_norm": 0.8918380737304688, + "learning_rate": 4.1525376005263e-05, + "loss": 0.8332, + "step": 15710 + }, + { + "epoch": 2.7971866096866096, + "grad_norm": 0.9205654263496399, + "learning_rate": 4.15140216009981e-05, + "loss": 0.7698, + "step": 15711 + }, + { + "epoch": 2.7973646723646723, + "grad_norm": 0.9631472229957581, + "learning_rate": 4.1502668342644455e-05, + "loss": 0.9604, + "step": 15712 + }, + { + "epoch": 2.797542735042735, + "grad_norm": 0.8770546913146973, + "learning_rate": 4.1491316230424516e-05, + "loss": 0.7661, + "step": 15713 + }, + { + "epoch": 2.797720797720798, + "grad_norm": 0.8872628808021545, + "learning_rate": 4.147996526456069e-05, + "loss": 1.0847, + "step": 15714 + }, + { + "epoch": 2.7978988603988606, + "grad_norm": 0.8924010396003723, + "learning_rate": 4.146861544527538e-05, + "loss": 0.8159, + "step": 15715 + }, + { + "epoch": 2.7980769230769234, + "grad_norm": 0.9251703023910522, + "learning_rate": 4.1457266772790923e-05, + "loss": 0.7204, + "step": 15716 + }, + { + "epoch": 2.7982549857549857, + "grad_norm": 0.8891414403915405, + "learning_rate": 4.144591924732979e-05, + "loss": 0.8576, + "step": 15717 + }, + { + "epoch": 2.7984330484330484, + "grad_norm": 0.9676079154014587, + "learning_rate": 4.143457286911415e-05, + "loss": 1.1912, + "step": 15718 + }, + { + "epoch": 2.798611111111111, + "grad_norm": 0.8125061392784119, + "learning_rate": 4.142322763836645e-05, + "loss": 0.8111, + "step": 15719 + }, + { + "epoch": 2.798789173789174, + "grad_norm": 0.8612900972366333, + "learning_rate": 4.141188355530891e-05, + "loss": 0.7537, + "step": 15720 + }, + { + "epoch": 2.798967236467236, + "grad_norm": 1.0774086713790894, + "learning_rate": 4.14005406201638e-05, + "loss": 0.9364, + "step": 15721 + }, + { + "epoch": 2.799145299145299, + "grad_norm": 0.8296873569488525, + "learning_rate": 4.138919883315338e-05, + "loss": 0.9329, + "step": 15722 + }, + { + "epoch": 2.7993233618233617, + "grad_norm": 0.757978618144989, + "learning_rate": 4.137785819449984e-05, + "loss": 1.0353, + "step": 15723 + }, + { + "epoch": 2.7995014245014245, + "grad_norm": 0.8584328293800354, + "learning_rate": 4.136651870442536e-05, + "loss": 1.0577, + "step": 15724 + }, + { + "epoch": 2.7996794871794872, + "grad_norm": 0.7919153571128845, + "learning_rate": 4.135518036315222e-05, + "loss": 0.8013, + "step": 15725 + }, + { + "epoch": 2.79985754985755, + "grad_norm": 0.968519926071167, + "learning_rate": 4.134384317090243e-05, + "loss": 0.7684, + "step": 15726 + }, + { + "epoch": 2.8000356125356127, + "grad_norm": 0.8565614819526672, + "learning_rate": 4.133250712789826e-05, + "loss": 0.9177, + "step": 15727 + }, + { + "epoch": 2.8002136752136755, + "grad_norm": 0.8614934086799622, + "learning_rate": 4.1321172234361647e-05, + "loss": 0.9613, + "step": 15728 + }, + { + "epoch": 2.800391737891738, + "grad_norm": 0.8621053099632263, + "learning_rate": 4.130983849051483e-05, + "loss": 0.8254, + "step": 15729 + }, + { + "epoch": 2.8005698005698005, + "grad_norm": 0.8108318448066711, + "learning_rate": 4.12985058965798e-05, + "loss": 0.7577, + "step": 15730 + }, + { + "epoch": 2.8007478632478633, + "grad_norm": 0.9211961627006531, + "learning_rate": 4.1287174452778564e-05, + "loss": 0.8204, + "step": 15731 + }, + { + "epoch": 2.800925925925926, + "grad_norm": 0.8582359552383423, + "learning_rate": 4.127584415933326e-05, + "loss": 0.8915, + "step": 15732 + }, + { + "epoch": 2.8011039886039883, + "grad_norm": 0.9122742414474487, + "learning_rate": 4.126451501646573e-05, + "loss": 1.0896, + "step": 15733 + }, + { + "epoch": 2.801282051282051, + "grad_norm": 0.7392016649246216, + "learning_rate": 4.125318702439804e-05, + "loss": 0.6354, + "step": 15734 + }, + { + "epoch": 2.801460113960114, + "grad_norm": 0.8227471113204956, + "learning_rate": 4.124186018335213e-05, + "loss": 0.7903, + "step": 15735 + }, + { + "epoch": 2.8016381766381766, + "grad_norm": 0.8843638300895691, + "learning_rate": 4.12305344935499e-05, + "loss": 0.7545, + "step": 15736 + }, + { + "epoch": 2.8018162393162394, + "grad_norm": 0.9978471994400024, + "learning_rate": 4.121920995521327e-05, + "loss": 1.0127, + "step": 15737 + }, + { + "epoch": 2.801994301994302, + "grad_norm": 0.9969626069068909, + "learning_rate": 4.12078865685641e-05, + "loss": 0.8137, + "step": 15738 + }, + { + "epoch": 2.802172364672365, + "grad_norm": 1.0768957138061523, + "learning_rate": 4.119656433382428e-05, + "loss": 0.6866, + "step": 15739 + }, + { + "epoch": 2.8023504273504276, + "grad_norm": 0.7401831746101379, + "learning_rate": 4.1185243251215624e-05, + "loss": 0.9103, + "step": 15740 + }, + { + "epoch": 2.80252849002849, + "grad_norm": 0.9753470420837402, + "learning_rate": 4.1173923320959905e-05, + "loss": 0.9499, + "step": 15741 + }, + { + "epoch": 2.8027065527065527, + "grad_norm": 0.9174960851669312, + "learning_rate": 4.116260454327904e-05, + "loss": 1.0355, + "step": 15742 + }, + { + "epoch": 2.8028846153846154, + "grad_norm": 0.8292258381843567, + "learning_rate": 4.115128691839464e-05, + "loss": 0.8806, + "step": 15743 + }, + { + "epoch": 2.803062678062678, + "grad_norm": 0.9542452096939087, + "learning_rate": 4.1139970446528564e-05, + "loss": 0.8378, + "step": 15744 + }, + { + "epoch": 2.8032407407407405, + "grad_norm": 0.848686933517456, + "learning_rate": 4.1128655127902485e-05, + "loss": 0.7939, + "step": 15745 + }, + { + "epoch": 2.8034188034188032, + "grad_norm": 0.8277645111083984, + "learning_rate": 4.1117340962738125e-05, + "loss": 0.9277, + "step": 15746 + }, + { + "epoch": 2.803596866096866, + "grad_norm": 0.8613318204879761, + "learning_rate": 4.110602795125714e-05, + "loss": 0.7622, + "step": 15747 + }, + { + "epoch": 2.8037749287749287, + "grad_norm": 0.7106199860572815, + "learning_rate": 4.109471609368121e-05, + "loss": 0.787, + "step": 15748 + }, + { + "epoch": 2.8039529914529915, + "grad_norm": 0.8933543562889099, + "learning_rate": 4.108340539023194e-05, + "loss": 0.857, + "step": 15749 + }, + { + "epoch": 2.8041310541310542, + "grad_norm": 0.8682022094726562, + "learning_rate": 4.107209584113092e-05, + "loss": 0.7931, + "step": 15750 + }, + { + "epoch": 2.804309116809117, + "grad_norm": 0.828279435634613, + "learning_rate": 4.106078744659981e-05, + "loss": 0.8404, + "step": 15751 + }, + { + "epoch": 2.8044871794871797, + "grad_norm": 0.9503956437110901, + "learning_rate": 4.1049480206860136e-05, + "loss": 0.8588, + "step": 15752 + }, + { + "epoch": 2.804665242165242, + "grad_norm": 0.810714602470398, + "learning_rate": 4.1038174122133435e-05, + "loss": 0.8592, + "step": 15753 + }, + { + "epoch": 2.804843304843305, + "grad_norm": 0.9458156824111938, + "learning_rate": 4.1026869192641225e-05, + "loss": 0.7441, + "step": 15754 + }, + { + "epoch": 2.8050213675213675, + "grad_norm": 0.8267046809196472, + "learning_rate": 4.1015565418605016e-05, + "loss": 0.8574, + "step": 15755 + }, + { + "epoch": 2.8051994301994303, + "grad_norm": 0.8413352370262146, + "learning_rate": 4.100426280024623e-05, + "loss": 0.7598, + "step": 15756 + }, + { + "epoch": 2.8053774928774926, + "grad_norm": 0.9205049872398376, + "learning_rate": 4.099296133778644e-05, + "loss": 0.8346, + "step": 15757 + }, + { + "epoch": 2.8055555555555554, + "grad_norm": 0.7986966967582703, + "learning_rate": 4.098166103144692e-05, + "loss": 0.7607, + "step": 15758 + }, + { + "epoch": 2.805733618233618, + "grad_norm": 0.9400181174278259, + "learning_rate": 4.097036188144918e-05, + "loss": 0.8947, + "step": 15759 + }, + { + "epoch": 2.805911680911681, + "grad_norm": 0.9014864563941956, + "learning_rate": 4.0959063888014594e-05, + "loss": 0.7781, + "step": 15760 + }, + { + "epoch": 2.8060897435897436, + "grad_norm": 0.8163666725158691, + "learning_rate": 4.094776705136448e-05, + "loss": 0.9042, + "step": 15761 + }, + { + "epoch": 2.8062678062678064, + "grad_norm": 0.8435617685317993, + "learning_rate": 4.0936471371720217e-05, + "loss": 0.9229, + "step": 15762 + }, + { + "epoch": 2.806445868945869, + "grad_norm": 0.7995414137840271, + "learning_rate": 4.0925176849303104e-05, + "loss": 0.7294, + "step": 15763 + }, + { + "epoch": 2.806623931623932, + "grad_norm": 0.9147883057594299, + "learning_rate": 4.091388348433442e-05, + "loss": 0.955, + "step": 15764 + }, + { + "epoch": 2.806801994301994, + "grad_norm": 0.7326688170433044, + "learning_rate": 4.0902591277035474e-05, + "loss": 0.5885, + "step": 15765 + }, + { + "epoch": 2.806980056980057, + "grad_norm": 0.8756957054138184, + "learning_rate": 4.0891300227627425e-05, + "loss": 0.9404, + "step": 15766 + }, + { + "epoch": 2.8071581196581197, + "grad_norm": 0.8897661566734314, + "learning_rate": 4.088001033633165e-05, + "loss": 1.0214, + "step": 15767 + }, + { + "epoch": 2.8073361823361824, + "grad_norm": 0.8007500171661377, + "learning_rate": 4.086872160336919e-05, + "loss": 0.6553, + "step": 15768 + }, + { + "epoch": 2.807514245014245, + "grad_norm": 0.8175814747810364, + "learning_rate": 4.0857434028961316e-05, + "loss": 0.8572, + "step": 15769 + }, + { + "epoch": 2.8076923076923075, + "grad_norm": 0.8290224671363831, + "learning_rate": 4.084614761332917e-05, + "loss": 0.9196, + "step": 15770 + }, + { + "epoch": 2.8078703703703702, + "grad_norm": 0.9355306625366211, + "learning_rate": 4.083486235669385e-05, + "loss": 0.8198, + "step": 15771 + }, + { + "epoch": 2.808048433048433, + "grad_norm": 0.942641019821167, + "learning_rate": 4.082357825927656e-05, + "loss": 0.7817, + "step": 15772 + }, + { + "epoch": 2.8082264957264957, + "grad_norm": 0.9115625023841858, + "learning_rate": 4.081229532129827e-05, + "loss": 1.046, + "step": 15773 + }, + { + "epoch": 2.8084045584045585, + "grad_norm": 0.8559226393699646, + "learning_rate": 4.080101354298016e-05, + "loss": 0.7085, + "step": 15774 + }, + { + "epoch": 2.8085826210826212, + "grad_norm": 0.8223599195480347, + "learning_rate": 4.0789732924543144e-05, + "loss": 0.9019, + "step": 15775 + }, + { + "epoch": 2.808760683760684, + "grad_norm": 0.8511637449264526, + "learning_rate": 4.0778453466208344e-05, + "loss": 0.6845, + "step": 15776 + }, + { + "epoch": 2.8089387464387463, + "grad_norm": 0.9633409976959229, + "learning_rate": 4.076717516819674e-05, + "loss": 1.0478, + "step": 15777 + }, + { + "epoch": 2.809116809116809, + "grad_norm": 0.8657141327857971, + "learning_rate": 4.075589803072928e-05, + "loss": 0.8694, + "step": 15778 + }, + { + "epoch": 2.809294871794872, + "grad_norm": 0.8126863241195679, + "learning_rate": 4.0744622054026936e-05, + "loss": 0.8529, + "step": 15779 + }, + { + "epoch": 2.8094729344729346, + "grad_norm": 0.8671838045120239, + "learning_rate": 4.0733347238310626e-05, + "loss": 0.81, + "step": 15780 + }, + { + "epoch": 2.8096509971509973, + "grad_norm": 0.8336054086685181, + "learning_rate": 4.0722073583801223e-05, + "loss": 0.7103, + "step": 15781 + }, + { + "epoch": 2.8098290598290596, + "grad_norm": 0.7833893299102783, + "learning_rate": 4.071080109071973e-05, + "loss": 0.875, + "step": 15782 + }, + { + "epoch": 2.8100071225071224, + "grad_norm": 0.9142106771469116, + "learning_rate": 4.0699529759286844e-05, + "loss": 0.9063, + "step": 15783 + }, + { + "epoch": 2.810185185185185, + "grad_norm": 0.7767373919487, + "learning_rate": 4.0688259589723565e-05, + "loss": 0.841, + "step": 15784 + }, + { + "epoch": 2.810363247863248, + "grad_norm": 0.8679327368736267, + "learning_rate": 4.067699058225056e-05, + "loss": 0.8581, + "step": 15785 + }, + { + "epoch": 2.8105413105413106, + "grad_norm": 0.9434911608695984, + "learning_rate": 4.066572273708873e-05, + "loss": 1.0166, + "step": 15786 + }, + { + "epoch": 2.8107193732193734, + "grad_norm": 0.8053399920463562, + "learning_rate": 4.06544560544588e-05, + "loss": 0.874, + "step": 15787 + }, + { + "epoch": 2.810897435897436, + "grad_norm": 1.0207599401474, + "learning_rate": 4.0643190534581524e-05, + "loss": 0.8296, + "step": 15788 + }, + { + "epoch": 2.8110754985754984, + "grad_norm": 0.8879590034484863, + "learning_rate": 4.0631926177677626e-05, + "loss": 0.8787, + "step": 15789 + }, + { + "epoch": 2.811253561253561, + "grad_norm": 1.0816758871078491, + "learning_rate": 4.062066298396778e-05, + "loss": 0.8129, + "step": 15790 + }, + { + "epoch": 2.811431623931624, + "grad_norm": 0.8332961797714233, + "learning_rate": 4.060940095367272e-05, + "loss": 0.8855, + "step": 15791 + }, + { + "epoch": 2.8116096866096867, + "grad_norm": 0.98028963804245, + "learning_rate": 4.059814008701308e-05, + "loss": 1.0065, + "step": 15792 + }, + { + "epoch": 2.8117877492877494, + "grad_norm": 0.8983020782470703, + "learning_rate": 4.058688038420949e-05, + "loss": 0.8259, + "step": 15793 + }, + { + "epoch": 2.8119658119658117, + "grad_norm": 0.8029065728187561, + "learning_rate": 4.057562184548255e-05, + "loss": 0.7639, + "step": 15794 + }, + { + "epoch": 2.8121438746438745, + "grad_norm": 0.8188722729682922, + "learning_rate": 4.056436447105286e-05, + "loss": 0.7179, + "step": 15795 + }, + { + "epoch": 2.8123219373219372, + "grad_norm": 0.8493495583534241, + "learning_rate": 4.055310826114095e-05, + "loss": 0.9479, + "step": 15796 + }, + { + "epoch": 2.8125, + "grad_norm": 0.7770833373069763, + "learning_rate": 4.0541853215967466e-05, + "loss": 0.6915, + "step": 15797 + }, + { + "epoch": 2.8126780626780628, + "grad_norm": 0.8238499760627747, + "learning_rate": 4.053059933575277e-05, + "loss": 1.0257, + "step": 15798 + }, + { + "epoch": 2.8128561253561255, + "grad_norm": 0.8537020683288574, + "learning_rate": 4.051934662071754e-05, + "loss": 0.7037, + "step": 15799 + }, + { + "epoch": 2.8130341880341883, + "grad_norm": 0.8120609521865845, + "learning_rate": 4.0508095071082055e-05, + "loss": 0.9533, + "step": 15800 + }, + { + "epoch": 2.8132122507122506, + "grad_norm": 0.8701691627502441, + "learning_rate": 4.0496844687066915e-05, + "loss": 0.7129, + "step": 15801 + }, + { + "epoch": 2.8133903133903133, + "grad_norm": 0.9007154107093811, + "learning_rate": 4.048559546889249e-05, + "loss": 0.8879, + "step": 15802 + }, + { + "epoch": 2.813568376068376, + "grad_norm": 0.8858364820480347, + "learning_rate": 4.047434741677919e-05, + "loss": 0.9391, + "step": 15803 + }, + { + "epoch": 2.813746438746439, + "grad_norm": 0.8597678542137146, + "learning_rate": 4.046310053094739e-05, + "loss": 0.7897, + "step": 15804 + }, + { + "epoch": 2.8139245014245016, + "grad_norm": 0.8493016362190247, + "learning_rate": 4.0451854811617475e-05, + "loss": 0.914, + "step": 15805 + }, + { + "epoch": 2.814102564102564, + "grad_norm": 0.8347373008728027, + "learning_rate": 4.044061025900973e-05, + "loss": 0.671, + "step": 15806 + }, + { + "epoch": 2.8142806267806266, + "grad_norm": 0.7406242489814758, + "learning_rate": 4.042936687334455e-05, + "loss": 0.6592, + "step": 15807 + }, + { + "epoch": 2.8144586894586894, + "grad_norm": 0.93736732006073, + "learning_rate": 4.041812465484214e-05, + "loss": 0.8301, + "step": 15808 + }, + { + "epoch": 2.814636752136752, + "grad_norm": 0.8744384050369263, + "learning_rate": 4.040688360372282e-05, + "loss": 0.9104, + "step": 15809 + }, + { + "epoch": 2.814814814814815, + "grad_norm": 0.7417266964912415, + "learning_rate": 4.0395643720206834e-05, + "loss": 0.7698, + "step": 15810 + }, + { + "epoch": 2.8149928774928776, + "grad_norm": 0.8601716160774231, + "learning_rate": 4.038440500451438e-05, + "loss": 0.8459, + "step": 15811 + }, + { + "epoch": 2.8151709401709404, + "grad_norm": 0.9801309108734131, + "learning_rate": 4.037316745686568e-05, + "loss": 0.9555, + "step": 15812 + }, + { + "epoch": 2.8153490028490027, + "grad_norm": 0.8559770584106445, + "learning_rate": 4.036193107748085e-05, + "loss": 0.9912, + "step": 15813 + }, + { + "epoch": 2.8155270655270654, + "grad_norm": 0.8155802488327026, + "learning_rate": 4.035069586658018e-05, + "loss": 0.8471, + "step": 15814 + }, + { + "epoch": 2.815705128205128, + "grad_norm": 0.96893310546875, + "learning_rate": 4.033946182438364e-05, + "loss": 0.9329, + "step": 15815 + }, + { + "epoch": 2.815883190883191, + "grad_norm": 0.9981120228767395, + "learning_rate": 4.032822895111144e-05, + "loss": 1.0333, + "step": 15816 + }, + { + "epoch": 2.8160612535612537, + "grad_norm": 0.9491816163063049, + "learning_rate": 4.031699724698363e-05, + "loss": 0.8838, + "step": 15817 + }, + { + "epoch": 2.816239316239316, + "grad_norm": 0.7149206399917603, + "learning_rate": 4.030576671222029e-05, + "loss": 0.8248, + "step": 15818 + }, + { + "epoch": 2.8164173789173788, + "grad_norm": 0.8366861343383789, + "learning_rate": 4.029453734704145e-05, + "loss": 0.8041, + "step": 15819 + }, + { + "epoch": 2.8165954415954415, + "grad_norm": 0.8147984147071838, + "learning_rate": 4.0283309151667116e-05, + "loss": 0.8702, + "step": 15820 + }, + { + "epoch": 2.8167735042735043, + "grad_norm": 0.8462722897529602, + "learning_rate": 4.027208212631729e-05, + "loss": 0.9899, + "step": 15821 + }, + { + "epoch": 2.816951566951567, + "grad_norm": 0.9278651475906372, + "learning_rate": 4.0260856271211946e-05, + "loss": 0.8684, + "step": 15822 + }, + { + "epoch": 2.8171296296296298, + "grad_norm": 0.8845569491386414, + "learning_rate": 4.0249631586570993e-05, + "loss": 0.6956, + "step": 15823 + }, + { + "epoch": 2.8173076923076925, + "grad_norm": 0.7803958654403687, + "learning_rate": 4.0238408072614453e-05, + "loss": 0.758, + "step": 15824 + }, + { + "epoch": 2.817485754985755, + "grad_norm": 0.8920331597328186, + "learning_rate": 4.022718572956209e-05, + "loss": 0.9131, + "step": 15825 + }, + { + "epoch": 2.8176638176638176, + "grad_norm": 0.9914098978042603, + "learning_rate": 4.021596455763389e-05, + "loss": 0.973, + "step": 15826 + }, + { + "epoch": 2.8178418803418803, + "grad_norm": 0.8329834938049316, + "learning_rate": 4.020474455704968e-05, + "loss": 0.8746, + "step": 15827 + }, + { + "epoch": 2.818019943019943, + "grad_norm": 0.7789189219474792, + "learning_rate": 4.019352572802928e-05, + "loss": 0.7359, + "step": 15828 + }, + { + "epoch": 2.818198005698006, + "grad_norm": 0.8405792713165283, + "learning_rate": 4.0182308070792505e-05, + "loss": 0.7897, + "step": 15829 + }, + { + "epoch": 2.818376068376068, + "grad_norm": 0.9215472936630249, + "learning_rate": 4.0171091585559116e-05, + "loss": 0.9017, + "step": 15830 + }, + { + "epoch": 2.818554131054131, + "grad_norm": 0.8310655355453491, + "learning_rate": 4.0159876272548933e-05, + "loss": 0.7499, + "step": 15831 + }, + { + "epoch": 2.8187321937321936, + "grad_norm": 0.8645792603492737, + "learning_rate": 4.014866213198167e-05, + "loss": 0.9009, + "step": 15832 + }, + { + "epoch": 2.8189102564102564, + "grad_norm": 0.8059788942337036, + "learning_rate": 4.013744916407703e-05, + "loss": 0.7367, + "step": 15833 + }, + { + "epoch": 2.819088319088319, + "grad_norm": 0.7990314960479736, + "learning_rate": 4.0126237369054745e-05, + "loss": 1.0172, + "step": 15834 + }, + { + "epoch": 2.819266381766382, + "grad_norm": 0.9321257472038269, + "learning_rate": 4.0115026747134446e-05, + "loss": 1.1224, + "step": 15835 + }, + { + "epoch": 2.8194444444444446, + "grad_norm": 0.9674378037452698, + "learning_rate": 4.0103817298535794e-05, + "loss": 0.9881, + "step": 15836 + }, + { + "epoch": 2.8196225071225074, + "grad_norm": 0.8573030829429626, + "learning_rate": 4.009260902347842e-05, + "loss": 0.7515, + "step": 15837 + }, + { + "epoch": 2.8198005698005697, + "grad_norm": 0.9248984456062317, + "learning_rate": 4.00814019221819e-05, + "loss": 0.9265, + "step": 15838 + }, + { + "epoch": 2.8199786324786325, + "grad_norm": 0.8749010562896729, + "learning_rate": 4.00701959948659e-05, + "loss": 0.8659, + "step": 15839 + }, + { + "epoch": 2.820156695156695, + "grad_norm": 0.8070803880691528, + "learning_rate": 4.005899124174986e-05, + "loss": 0.8813, + "step": 15840 + }, + { + "epoch": 2.820334757834758, + "grad_norm": 1.0711981058120728, + "learning_rate": 4.004778766305339e-05, + "loss": 1.019, + "step": 15841 + }, + { + "epoch": 2.8205128205128203, + "grad_norm": 0.7646795511245728, + "learning_rate": 4.0036585258995985e-05, + "loss": 0.8326, + "step": 15842 + }, + { + "epoch": 2.820690883190883, + "grad_norm": 0.6720184683799744, + "learning_rate": 4.002538402979713e-05, + "loss": 0.6642, + "step": 15843 + }, + { + "epoch": 2.8208689458689458, + "grad_norm": 0.8062998056411743, + "learning_rate": 4.001418397567629e-05, + "loss": 0.9585, + "step": 15844 + }, + { + "epoch": 2.8210470085470085, + "grad_norm": 0.835515558719635, + "learning_rate": 4.0002985096852893e-05, + "loss": 0.7992, + "step": 15845 + }, + { + "epoch": 2.8212250712250713, + "grad_norm": 0.8308731913566589, + "learning_rate": 3.9991787393546385e-05, + "loss": 0.9982, + "step": 15846 + }, + { + "epoch": 2.821403133903134, + "grad_norm": 0.9056837558746338, + "learning_rate": 3.998059086597614e-05, + "loss": 0.8283, + "step": 15847 + }, + { + "epoch": 2.8215811965811968, + "grad_norm": 0.853070080280304, + "learning_rate": 3.9969395514361506e-05, + "loss": 1.033, + "step": 15848 + }, + { + "epoch": 2.8217592592592595, + "grad_norm": 0.9179061055183411, + "learning_rate": 3.99582013389219e-05, + "loss": 0.929, + "step": 15849 + }, + { + "epoch": 2.821937321937322, + "grad_norm": 0.8702627420425415, + "learning_rate": 3.9947008339876616e-05, + "loss": 0.9994, + "step": 15850 + }, + { + "epoch": 2.8221153846153846, + "grad_norm": 0.9594024419784546, + "learning_rate": 3.9935816517444935e-05, + "loss": 0.88, + "step": 15851 + }, + { + "epoch": 2.8222934472934473, + "grad_norm": 0.8474575877189636, + "learning_rate": 3.992462587184618e-05, + "loss": 0.7817, + "step": 15852 + }, + { + "epoch": 2.82247150997151, + "grad_norm": 0.7588878870010376, + "learning_rate": 3.9913436403299533e-05, + "loss": 0.7911, + "step": 15853 + }, + { + "epoch": 2.8226495726495724, + "grad_norm": 0.8467457890510559, + "learning_rate": 3.9902248112024366e-05, + "loss": 0.7433, + "step": 15854 + }, + { + "epoch": 2.822827635327635, + "grad_norm": 0.839137077331543, + "learning_rate": 3.989106099823972e-05, + "loss": 0.7303, + "step": 15855 + }, + { + "epoch": 2.823005698005698, + "grad_norm": 0.8945586085319519, + "learning_rate": 3.987987506216495e-05, + "loss": 0.6697, + "step": 15856 + }, + { + "epoch": 2.8231837606837606, + "grad_norm": 0.749971330165863, + "learning_rate": 3.9868690304019064e-05, + "loss": 0.65, + "step": 15857 + }, + { + "epoch": 2.8233618233618234, + "grad_norm": 0.9841105341911316, + "learning_rate": 3.985750672402131e-05, + "loss": 0.8382, + "step": 15858 + }, + { + "epoch": 2.823539886039886, + "grad_norm": 0.8696077466011047, + "learning_rate": 3.984632432239078e-05, + "loss": 0.7895, + "step": 15859 + }, + { + "epoch": 2.823717948717949, + "grad_norm": 0.7845979928970337, + "learning_rate": 3.9835143099346575e-05, + "loss": 0.8673, + "step": 15860 + }, + { + "epoch": 2.8238960113960117, + "grad_norm": 0.8145211338996887, + "learning_rate": 3.982396305510775e-05, + "loss": 0.905, + "step": 15861 + }, + { + "epoch": 2.824074074074074, + "grad_norm": 0.8533337116241455, + "learning_rate": 3.981278418989336e-05, + "loss": 0.9597, + "step": 15862 + }, + { + "epoch": 2.8242521367521367, + "grad_norm": 0.9430350065231323, + "learning_rate": 3.980160650392241e-05, + "loss": 0.9043, + "step": 15863 + }, + { + "epoch": 2.8244301994301995, + "grad_norm": 0.7469115257263184, + "learning_rate": 3.9790429997414e-05, + "loss": 0.9074, + "step": 15864 + }, + { + "epoch": 2.824608262108262, + "grad_norm": 0.8558746576309204, + "learning_rate": 3.977925467058696e-05, + "loss": 0.8958, + "step": 15865 + }, + { + "epoch": 2.8247863247863245, + "grad_norm": 1.1557669639587402, + "learning_rate": 3.976808052366037e-05, + "loss": 0.7884, + "step": 15866 + }, + { + "epoch": 2.8249643874643873, + "grad_norm": 0.8448477983474731, + "learning_rate": 3.975690755685312e-05, + "loss": 0.9542, + "step": 15867 + }, + { + "epoch": 2.82514245014245, + "grad_norm": 0.833256721496582, + "learning_rate": 3.9745735770384086e-05, + "loss": 0.9196, + "step": 15868 + }, + { + "epoch": 2.8253205128205128, + "grad_norm": 0.9318852424621582, + "learning_rate": 3.973456516447226e-05, + "loss": 0.7792, + "step": 15869 + }, + { + "epoch": 2.8254985754985755, + "grad_norm": 0.832655131816864, + "learning_rate": 3.972339573933638e-05, + "loss": 0.9526, + "step": 15870 + }, + { + "epoch": 2.8256766381766383, + "grad_norm": 0.7546842694282532, + "learning_rate": 3.9712227495195406e-05, + "loss": 0.5969, + "step": 15871 + }, + { + "epoch": 2.825854700854701, + "grad_norm": 0.8538267016410828, + "learning_rate": 3.970106043226802e-05, + "loss": 0.8369, + "step": 15872 + }, + { + "epoch": 2.826032763532764, + "grad_norm": 0.8023465871810913, + "learning_rate": 3.968989455077314e-05, + "loss": 0.7257, + "step": 15873 + }, + { + "epoch": 2.826210826210826, + "grad_norm": 0.7905409932136536, + "learning_rate": 3.96787298509295e-05, + "loss": 0.7965, + "step": 15874 + }, + { + "epoch": 2.826388888888889, + "grad_norm": 0.8316642642021179, + "learning_rate": 3.966756633295583e-05, + "loss": 0.808, + "step": 15875 + }, + { + "epoch": 2.8265669515669516, + "grad_norm": 0.9130233526229858, + "learning_rate": 3.965640399707088e-05, + "loss": 0.882, + "step": 15876 + }, + { + "epoch": 2.8267450142450143, + "grad_norm": 0.873147189617157, + "learning_rate": 3.9645242843493325e-05, + "loss": 0.8347, + "step": 15877 + }, + { + "epoch": 2.8269230769230766, + "grad_norm": 0.8806825280189514, + "learning_rate": 3.963408287244183e-05, + "loss": 1.0496, + "step": 15878 + }, + { + "epoch": 2.8271011396011394, + "grad_norm": 0.8899962306022644, + "learning_rate": 3.962292408413516e-05, + "loss": 1.0014, + "step": 15879 + }, + { + "epoch": 2.827279202279202, + "grad_norm": 0.900303065776825, + "learning_rate": 3.961176647879179e-05, + "loss": 0.7345, + "step": 15880 + }, + { + "epoch": 2.827457264957265, + "grad_norm": 1.1055848598480225, + "learning_rate": 3.960061005663049e-05, + "loss": 0.9697, + "step": 15881 + }, + { + "epoch": 2.8276353276353277, + "grad_norm": 0.891404390335083, + "learning_rate": 3.958945481786969e-05, + "loss": 0.8878, + "step": 15882 + }, + { + "epoch": 2.8278133903133904, + "grad_norm": 0.9403249025344849, + "learning_rate": 3.957830076272807e-05, + "loss": 0.9536, + "step": 15883 + }, + { + "epoch": 2.827991452991453, + "grad_norm": 0.8735896944999695, + "learning_rate": 3.9567147891424126e-05, + "loss": 0.9113, + "step": 15884 + }, + { + "epoch": 2.828169515669516, + "grad_norm": 0.7758442759513855, + "learning_rate": 3.9555996204176385e-05, + "loss": 0.7336, + "step": 15885 + }, + { + "epoch": 2.828347578347578, + "grad_norm": 0.8632972836494446, + "learning_rate": 3.9544845701203335e-05, + "loss": 0.6883, + "step": 15886 + }, + { + "epoch": 2.828525641025641, + "grad_norm": 0.8639035224914551, + "learning_rate": 3.953369638272345e-05, + "loss": 0.8836, + "step": 15887 + }, + { + "epoch": 2.8287037037037037, + "grad_norm": 0.8131114840507507, + "learning_rate": 3.952254824895514e-05, + "loss": 0.8175, + "step": 15888 + }, + { + "epoch": 2.8288817663817665, + "grad_norm": 0.7421914935112, + "learning_rate": 3.9511401300116904e-05, + "loss": 0.7721, + "step": 15889 + }, + { + "epoch": 2.8290598290598292, + "grad_norm": 0.8358926177024841, + "learning_rate": 3.95002555364271e-05, + "loss": 0.7327, + "step": 15890 + }, + { + "epoch": 2.8292378917378915, + "grad_norm": 0.8913134932518005, + "learning_rate": 3.9489110958104115e-05, + "loss": 1.0415, + "step": 15891 + }, + { + "epoch": 2.8294159544159543, + "grad_norm": 0.9675887227058411, + "learning_rate": 3.94779675653663e-05, + "loss": 0.8295, + "step": 15892 + }, + { + "epoch": 2.829594017094017, + "grad_norm": 0.8618438839912415, + "learning_rate": 3.946682535843199e-05, + "loss": 1.0754, + "step": 15893 + }, + { + "epoch": 2.82977207977208, + "grad_norm": 0.820209801197052, + "learning_rate": 3.945568433751948e-05, + "loss": 0.9484, + "step": 15894 + }, + { + "epoch": 2.8299501424501425, + "grad_norm": 0.8641984462738037, + "learning_rate": 3.944454450284705e-05, + "loss": 0.964, + "step": 15895 + }, + { + "epoch": 2.8301282051282053, + "grad_norm": 0.8229194283485413, + "learning_rate": 3.943340585463303e-05, + "loss": 0.7069, + "step": 15896 + }, + { + "epoch": 2.830306267806268, + "grad_norm": 0.7874621748924255, + "learning_rate": 3.942226839309554e-05, + "loss": 0.7815, + "step": 15897 + }, + { + "epoch": 2.8304843304843303, + "grad_norm": 0.8581945896148682, + "learning_rate": 3.9411132118452896e-05, + "loss": 0.9119, + "step": 15898 + }, + { + "epoch": 2.830662393162393, + "grad_norm": 0.9327018857002258, + "learning_rate": 3.939999703092326e-05, + "loss": 1.1002, + "step": 15899 + }, + { + "epoch": 2.830840455840456, + "grad_norm": 0.7793048024177551, + "learning_rate": 3.9388863130724794e-05, + "loss": 0.6529, + "step": 15900 + }, + { + "epoch": 2.8310185185185186, + "grad_norm": 0.9133790135383606, + "learning_rate": 3.9377730418075645e-05, + "loss": 0.7354, + "step": 15901 + }, + { + "epoch": 2.8311965811965814, + "grad_norm": 0.7800240516662598, + "learning_rate": 3.936659889319394e-05, + "loss": 0.8541, + "step": 15902 + }, + { + "epoch": 2.8313746438746437, + "grad_norm": 0.782433271408081, + "learning_rate": 3.9355468556297737e-05, + "loss": 0.9084, + "step": 15903 + }, + { + "epoch": 2.8315527065527064, + "grad_norm": 0.8926814198493958, + "learning_rate": 3.9344339407605226e-05, + "loss": 1.0252, + "step": 15904 + }, + { + "epoch": 2.831730769230769, + "grad_norm": 0.92144376039505, + "learning_rate": 3.93332114473343e-05, + "loss": 0.7859, + "step": 15905 + }, + { + "epoch": 2.831908831908832, + "grad_norm": 0.7403308749198914, + "learning_rate": 3.932208467570315e-05, + "loss": 0.796, + "step": 15906 + }, + { + "epoch": 2.8320868945868947, + "grad_norm": 0.939708411693573, + "learning_rate": 3.9310959092929636e-05, + "loss": 0.9622, + "step": 15907 + }, + { + "epoch": 2.8322649572649574, + "grad_norm": 0.7546647787094116, + "learning_rate": 3.929983469923184e-05, + "loss": 0.7551, + "step": 15908 + }, + { + "epoch": 2.83244301994302, + "grad_norm": 0.8301447033882141, + "learning_rate": 3.928871149482768e-05, + "loss": 0.9377, + "step": 15909 + }, + { + "epoch": 2.8326210826210825, + "grad_norm": 0.8961447477340698, + "learning_rate": 3.927758947993508e-05, + "loss": 0.8245, + "step": 15910 + }, + { + "epoch": 2.8327991452991452, + "grad_norm": 0.7845988869667053, + "learning_rate": 3.926646865477204e-05, + "loss": 0.8662, + "step": 15911 + }, + { + "epoch": 2.832977207977208, + "grad_norm": 0.9339789152145386, + "learning_rate": 3.925534901955631e-05, + "loss": 0.8533, + "step": 15912 + }, + { + "epoch": 2.8331552706552707, + "grad_norm": 0.936855137348175, + "learning_rate": 3.924423057450587e-05, + "loss": 0.8809, + "step": 15913 + }, + { + "epoch": 2.8333333333333335, + "grad_norm": 0.972535252571106, + "learning_rate": 3.923311331983852e-05, + "loss": 1.0103, + "step": 15914 + }, + { + "epoch": 2.833511396011396, + "grad_norm": 0.9115430116653442, + "learning_rate": 3.922199725577208e-05, + "loss": 1.0041, + "step": 15915 + }, + { + "epoch": 2.8336894586894585, + "grad_norm": 0.8378027081489563, + "learning_rate": 3.921088238252435e-05, + "loss": 0.8475, + "step": 15916 + }, + { + "epoch": 2.8338675213675213, + "grad_norm": 0.9383054375648499, + "learning_rate": 3.91997687003131e-05, + "loss": 0.7305, + "step": 15917 + }, + { + "epoch": 2.834045584045584, + "grad_norm": 0.7996332049369812, + "learning_rate": 3.918865620935609e-05, + "loss": 0.8749, + "step": 15918 + }, + { + "epoch": 2.834223646723647, + "grad_norm": 0.8876177668571472, + "learning_rate": 3.917754490987103e-05, + "loss": 0.8836, + "step": 15919 + }, + { + "epoch": 2.8344017094017095, + "grad_norm": 0.8004130125045776, + "learning_rate": 3.9166434802075594e-05, + "loss": 1.0383, + "step": 15920 + }, + { + "epoch": 2.8345797720797723, + "grad_norm": 0.76146000623703, + "learning_rate": 3.915532588618756e-05, + "loss": 0.9616, + "step": 15921 + }, + { + "epoch": 2.8347578347578346, + "grad_norm": 0.9333193898200989, + "learning_rate": 3.914421816242446e-05, + "loss": 0.7386, + "step": 15922 + }, + { + "epoch": 2.8349358974358974, + "grad_norm": 0.9340601563453674, + "learning_rate": 3.913311163100403e-05, + "loss": 0.8894, + "step": 15923 + }, + { + "epoch": 2.83511396011396, + "grad_norm": 0.8401036858558655, + "learning_rate": 3.912200629214383e-05, + "loss": 0.8366, + "step": 15924 + }, + { + "epoch": 2.835292022792023, + "grad_norm": 0.9298731088638306, + "learning_rate": 3.911090214606146e-05, + "loss": 0.899, + "step": 15925 + }, + { + "epoch": 2.8354700854700856, + "grad_norm": 0.8085874915122986, + "learning_rate": 3.909979919297446e-05, + "loss": 0.7597, + "step": 15926 + }, + { + "epoch": 2.835648148148148, + "grad_norm": 0.7841027975082397, + "learning_rate": 3.9088697433100396e-05, + "loss": 0.897, + "step": 15927 + }, + { + "epoch": 2.8358262108262107, + "grad_norm": 1.0678621530532837, + "learning_rate": 3.907759686665677e-05, + "loss": 0.976, + "step": 15928 + }, + { + "epoch": 2.8360042735042734, + "grad_norm": 0.7748154997825623, + "learning_rate": 3.906649749386106e-05, + "loss": 0.6834, + "step": 15929 + }, + { + "epoch": 2.836182336182336, + "grad_norm": 0.8119567632675171, + "learning_rate": 3.905539931493076e-05, + "loss": 0.8076, + "step": 15930 + }, + { + "epoch": 2.836360398860399, + "grad_norm": 0.8723282814025879, + "learning_rate": 3.9044302330083326e-05, + "loss": 0.9057, + "step": 15931 + }, + { + "epoch": 2.8365384615384617, + "grad_norm": 0.7785065174102783, + "learning_rate": 3.903320653953616e-05, + "loss": 0.8899, + "step": 15932 + }, + { + "epoch": 2.8367165242165244, + "grad_norm": 0.9053105115890503, + "learning_rate": 3.902211194350667e-05, + "loss": 0.9038, + "step": 15933 + }, + { + "epoch": 2.8368945868945867, + "grad_norm": 0.8937689065933228, + "learning_rate": 3.9011018542212216e-05, + "loss": 0.736, + "step": 15934 + }, + { + "epoch": 2.8370726495726495, + "grad_norm": 0.7169269323348999, + "learning_rate": 3.899992633587014e-05, + "loss": 0.6632, + "step": 15935 + }, + { + "epoch": 2.8372507122507122, + "grad_norm": 0.8168412446975708, + "learning_rate": 3.898883532469785e-05, + "loss": 0.8482, + "step": 15936 + }, + { + "epoch": 2.837428774928775, + "grad_norm": 0.7374065518379211, + "learning_rate": 3.897774550891252e-05, + "loss": 0.8141, + "step": 15937 + }, + { + "epoch": 2.8376068376068377, + "grad_norm": 0.8844853043556213, + "learning_rate": 3.8966656888731546e-05, + "loss": 0.8204, + "step": 15938 + }, + { + "epoch": 2.8377849002849, + "grad_norm": 0.9031739234924316, + "learning_rate": 3.895556946437213e-05, + "loss": 0.862, + "step": 15939 + }, + { + "epoch": 2.837962962962963, + "grad_norm": 0.8141549229621887, + "learning_rate": 3.894448323605154e-05, + "loss": 0.6687, + "step": 15940 + }, + { + "epoch": 2.8381410256410255, + "grad_norm": 0.796144962310791, + "learning_rate": 3.893339820398696e-05, + "loss": 0.9021, + "step": 15941 + }, + { + "epoch": 2.8383190883190883, + "grad_norm": 0.8840420246124268, + "learning_rate": 3.8922314368395584e-05, + "loss": 0.9608, + "step": 15942 + }, + { + "epoch": 2.838497150997151, + "grad_norm": 0.8297450542449951, + "learning_rate": 3.891123172949459e-05, + "loss": 0.9442, + "step": 15943 + }, + { + "epoch": 2.838675213675214, + "grad_norm": 0.7875503301620483, + "learning_rate": 3.89001502875011e-05, + "loss": 0.9269, + "step": 15944 + }, + { + "epoch": 2.8388532763532766, + "grad_norm": 0.9460122585296631, + "learning_rate": 3.8889070042632217e-05, + "loss": 0.9459, + "step": 15945 + }, + { + "epoch": 2.8390313390313393, + "grad_norm": 0.8144980669021606, + "learning_rate": 3.887799099510512e-05, + "loss": 0.8409, + "step": 15946 + }, + { + "epoch": 2.8392094017094016, + "grad_norm": 0.8182117342948914, + "learning_rate": 3.886691314513675e-05, + "loss": 0.8093, + "step": 15947 + }, + { + "epoch": 2.8393874643874644, + "grad_norm": 0.8287648558616638, + "learning_rate": 3.885583649294426e-05, + "loss": 0.7792, + "step": 15948 + }, + { + "epoch": 2.839565527065527, + "grad_norm": 0.9165690541267395, + "learning_rate": 3.884476103874464e-05, + "loss": 0.9865, + "step": 15949 + }, + { + "epoch": 2.83974358974359, + "grad_norm": 0.7819885015487671, + "learning_rate": 3.883368678275485e-05, + "loss": 0.7245, + "step": 15950 + }, + { + "epoch": 2.839921652421652, + "grad_norm": 0.8354606628417969, + "learning_rate": 3.882261372519198e-05, + "loss": 1.0513, + "step": 15951 + }, + { + "epoch": 2.840099715099715, + "grad_norm": 0.7606815099716187, + "learning_rate": 3.881154186627284e-05, + "loss": 0.9357, + "step": 15952 + }, + { + "epoch": 2.8402777777777777, + "grad_norm": 0.9649691581726074, + "learning_rate": 3.88004712062145e-05, + "loss": 0.8756, + "step": 15953 + }, + { + "epoch": 2.8404558404558404, + "grad_norm": 0.8770344853401184, + "learning_rate": 3.878940174523371e-05, + "loss": 0.868, + "step": 15954 + }, + { + "epoch": 2.840633903133903, + "grad_norm": 0.898287832736969, + "learning_rate": 3.877833348354749e-05, + "loss": 0.8734, + "step": 15955 + }, + { + "epoch": 2.840811965811966, + "grad_norm": 0.84062260389328, + "learning_rate": 3.876726642137264e-05, + "loss": 0.9362, + "step": 15956 + }, + { + "epoch": 2.8409900284900287, + "grad_norm": 0.7898240685462952, + "learning_rate": 3.8756200558926013e-05, + "loss": 0.7788, + "step": 15957 + }, + { + "epoch": 2.8411680911680914, + "grad_norm": 0.7237298488616943, + "learning_rate": 3.874513589642441e-05, + "loss": 0.8426, + "step": 15958 + }, + { + "epoch": 2.8413461538461537, + "grad_norm": 0.9025090932846069, + "learning_rate": 3.873407243408462e-05, + "loss": 0.7135, + "step": 15959 + }, + { + "epoch": 2.8415242165242165, + "grad_norm": 0.807295560836792, + "learning_rate": 3.872301017212337e-05, + "loss": 0.6889, + "step": 15960 + }, + { + "epoch": 2.8417022792022792, + "grad_norm": 0.8537244200706482, + "learning_rate": 3.8711949110757525e-05, + "loss": 0.798, + "step": 15961 + }, + { + "epoch": 2.841880341880342, + "grad_norm": 0.8148910999298096, + "learning_rate": 3.870088925020366e-05, + "loss": 0.8783, + "step": 15962 + }, + { + "epoch": 2.8420584045584043, + "grad_norm": 0.8254446983337402, + "learning_rate": 3.868983059067859e-05, + "loss": 0.7043, + "step": 15963 + }, + { + "epoch": 2.842236467236467, + "grad_norm": 0.8392706513404846, + "learning_rate": 3.867877313239886e-05, + "loss": 0.9942, + "step": 15964 + }, + { + "epoch": 2.84241452991453, + "grad_norm": 0.8974948525428772, + "learning_rate": 3.8667716875581217e-05, + "loss": 0.7646, + "step": 15965 + }, + { + "epoch": 2.8425925925925926, + "grad_norm": 0.9764110445976257, + "learning_rate": 3.8656661820442264e-05, + "loss": 0.8803, + "step": 15966 + }, + { + "epoch": 2.8427706552706553, + "grad_norm": 0.9663669466972351, + "learning_rate": 3.864560796719855e-05, + "loss": 0.8764, + "step": 15967 + }, + { + "epoch": 2.842948717948718, + "grad_norm": 0.837733268737793, + "learning_rate": 3.863455531606677e-05, + "loss": 0.8992, + "step": 15968 + }, + { + "epoch": 2.843126780626781, + "grad_norm": 0.8458481431007385, + "learning_rate": 3.8623503867263335e-05, + "loss": 0.9025, + "step": 15969 + }, + { + "epoch": 2.8433048433048436, + "grad_norm": 0.901089072227478, + "learning_rate": 3.861245362100488e-05, + "loss": 0.8025, + "step": 15970 + }, + { + "epoch": 2.843482905982906, + "grad_norm": 0.9032089114189148, + "learning_rate": 3.860140457750786e-05, + "loss": 0.7217, + "step": 15971 + }, + { + "epoch": 2.8436609686609686, + "grad_norm": 0.7998839616775513, + "learning_rate": 3.859035673698879e-05, + "loss": 0.9127, + "step": 15972 + }, + { + "epoch": 2.8438390313390314, + "grad_norm": 0.8568583726882935, + "learning_rate": 3.85793100996641e-05, + "loss": 0.8847, + "step": 15973 + }, + { + "epoch": 2.844017094017094, + "grad_norm": 0.8720089793205261, + "learning_rate": 3.856826466575024e-05, + "loss": 0.7822, + "step": 15974 + }, + { + "epoch": 2.8441951566951564, + "grad_norm": 0.8872382640838623, + "learning_rate": 3.8557220435463594e-05, + "loss": 0.9601, + "step": 15975 + }, + { + "epoch": 2.844373219373219, + "grad_norm": 1.1950596570968628, + "learning_rate": 3.8546177409020634e-05, + "loss": 1.075, + "step": 15976 + }, + { + "epoch": 2.844551282051282, + "grad_norm": 0.9111549854278564, + "learning_rate": 3.85351355866376e-05, + "loss": 1.0103, + "step": 15977 + }, + { + "epoch": 2.8447293447293447, + "grad_norm": 0.9310214519500732, + "learning_rate": 3.852409496853099e-05, + "loss": 1.0163, + "step": 15978 + }, + { + "epoch": 2.8449074074074074, + "grad_norm": 0.8177474737167358, + "learning_rate": 3.851305555491695e-05, + "loss": 0.7488, + "step": 15979 + }, + { + "epoch": 2.84508547008547, + "grad_norm": 0.9321249127388, + "learning_rate": 3.85020173460119e-05, + "loss": 0.9914, + "step": 15980 + }, + { + "epoch": 2.845263532763533, + "grad_norm": 0.7649266719818115, + "learning_rate": 3.849098034203206e-05, + "loss": 0.692, + "step": 15981 + }, + { + "epoch": 2.8454415954415957, + "grad_norm": 0.7714266777038574, + "learning_rate": 3.847994454319369e-05, + "loss": 0.8859, + "step": 15982 + }, + { + "epoch": 2.845619658119658, + "grad_norm": 0.9535303711891174, + "learning_rate": 3.846890994971302e-05, + "loss": 0.8992, + "step": 15983 + }, + { + "epoch": 2.8457977207977208, + "grad_norm": 0.8171879649162292, + "learning_rate": 3.845787656180623e-05, + "loss": 0.7125, + "step": 15984 + }, + { + "epoch": 2.8459757834757835, + "grad_norm": 0.8546884655952454, + "learning_rate": 3.8446844379689464e-05, + "loss": 0.9895, + "step": 15985 + }, + { + "epoch": 2.8461538461538463, + "grad_norm": 0.9110364317893982, + "learning_rate": 3.843581340357899e-05, + "loss": 1.0702, + "step": 15986 + }, + { + "epoch": 2.8463319088319086, + "grad_norm": 0.8862065672874451, + "learning_rate": 3.84247836336908e-05, + "loss": 0.9138, + "step": 15987 + }, + { + "epoch": 2.8465099715099713, + "grad_norm": 0.8485249876976013, + "learning_rate": 3.84137550702411e-05, + "loss": 1.2831, + "step": 15988 + }, + { + "epoch": 2.846688034188034, + "grad_norm": 0.8271495699882507, + "learning_rate": 3.840272771344593e-05, + "loss": 0.9497, + "step": 15989 + }, + { + "epoch": 2.846866096866097, + "grad_norm": 0.7829293608665466, + "learning_rate": 3.839170156352135e-05, + "loss": 0.6503, + "step": 15990 + }, + { + "epoch": 2.8470441595441596, + "grad_norm": 0.9366582036018372, + "learning_rate": 3.838067662068341e-05, + "loss": 0.7805, + "step": 15991 + }, + { + "epoch": 2.8472222222222223, + "grad_norm": 0.8666117787361145, + "learning_rate": 3.836965288514807e-05, + "loss": 0.7721, + "step": 15992 + }, + { + "epoch": 2.847400284900285, + "grad_norm": 0.7855546474456787, + "learning_rate": 3.835863035713142e-05, + "loss": 0.7457, + "step": 15993 + }, + { + "epoch": 2.847578347578348, + "grad_norm": 0.8234511017799377, + "learning_rate": 3.8347609036849284e-05, + "loss": 0.8937, + "step": 15994 + }, + { + "epoch": 2.84775641025641, + "grad_norm": 0.8896345496177673, + "learning_rate": 3.833658892451773e-05, + "loss": 0.9146, + "step": 15995 + }, + { + "epoch": 2.847934472934473, + "grad_norm": 0.8099349737167358, + "learning_rate": 3.83255700203526e-05, + "loss": 0.8353, + "step": 15996 + }, + { + "epoch": 2.8481125356125356, + "grad_norm": 0.874100387096405, + "learning_rate": 3.831455232456982e-05, + "loss": 0.6829, + "step": 15997 + }, + { + "epoch": 2.8482905982905984, + "grad_norm": 0.9338345527648926, + "learning_rate": 3.830353583738524e-05, + "loss": 1.0345, + "step": 15998 + }, + { + "epoch": 2.8484686609686607, + "grad_norm": 0.7876978516578674, + "learning_rate": 3.829252055901472e-05, + "loss": 0.6703, + "step": 15999 + }, + { + "epoch": 2.8486467236467234, + "grad_norm": 0.8565872311592102, + "learning_rate": 3.828150648967408e-05, + "loss": 0.8227, + "step": 16000 + }, + { + "epoch": 2.848824786324786, + "grad_norm": 1.0180596113204956, + "learning_rate": 3.82704936295791e-05, + "loss": 0.8933, + "step": 16001 + }, + { + "epoch": 2.849002849002849, + "grad_norm": 0.8450096845626831, + "learning_rate": 3.825948197894553e-05, + "loss": 0.851, + "step": 16002 + }, + { + "epoch": 2.8491809116809117, + "grad_norm": 0.7936033010482788, + "learning_rate": 3.824847153798923e-05, + "loss": 0.8573, + "step": 16003 + }, + { + "epoch": 2.8493589743589745, + "grad_norm": 0.9499372839927673, + "learning_rate": 3.8237462306925774e-05, + "loss": 0.8269, + "step": 16004 + }, + { + "epoch": 2.849537037037037, + "grad_norm": 0.874855101108551, + "learning_rate": 3.822645428597099e-05, + "loss": 0.9657, + "step": 16005 + }, + { + "epoch": 2.8497150997151, + "grad_norm": 0.8966119885444641, + "learning_rate": 3.8215447475340506e-05, + "loss": 0.9239, + "step": 16006 + }, + { + "epoch": 2.8498931623931623, + "grad_norm": 0.8341490030288696, + "learning_rate": 3.820444187524994e-05, + "loss": 1.0051, + "step": 16007 + }, + { + "epoch": 2.850071225071225, + "grad_norm": 0.7965613007545471, + "learning_rate": 3.8193437485915054e-05, + "loss": 0.8591, + "step": 16008 + }, + { + "epoch": 2.8502492877492878, + "grad_norm": 0.7846593856811523, + "learning_rate": 3.818243430755128e-05, + "loss": 0.8095, + "step": 16009 + }, + { + "epoch": 2.8504273504273505, + "grad_norm": 0.9422695636749268, + "learning_rate": 3.8171432340374334e-05, + "loss": 0.8902, + "step": 16010 + }, + { + "epoch": 2.8506054131054133, + "grad_norm": 0.8810960650444031, + "learning_rate": 3.8160431584599744e-05, + "loss": 0.8483, + "step": 16011 + }, + { + "epoch": 2.8507834757834756, + "grad_norm": 0.8513348698616028, + "learning_rate": 3.814943204044302e-05, + "loss": 0.832, + "step": 16012 + }, + { + "epoch": 2.8509615384615383, + "grad_norm": 0.8906846046447754, + "learning_rate": 3.8138433708119704e-05, + "loss": 0.9702, + "step": 16013 + }, + { + "epoch": 2.851139601139601, + "grad_norm": 0.9517511129379272, + "learning_rate": 3.812743658784526e-05, + "loss": 0.8138, + "step": 16014 + }, + { + "epoch": 2.851317663817664, + "grad_norm": 0.7989702820777893, + "learning_rate": 3.811644067983517e-05, + "loss": 0.8653, + "step": 16015 + }, + { + "epoch": 2.8514957264957266, + "grad_norm": 0.8255589008331299, + "learning_rate": 3.8105445984304874e-05, + "loss": 0.9456, + "step": 16016 + }, + { + "epoch": 2.8516737891737893, + "grad_norm": 0.7919938564300537, + "learning_rate": 3.809445250146977e-05, + "loss": 0.6261, + "step": 16017 + }, + { + "epoch": 2.851851851851852, + "grad_norm": 0.866316020488739, + "learning_rate": 3.808346023154532e-05, + "loss": 0.8171, + "step": 16018 + }, + { + "epoch": 2.8520299145299144, + "grad_norm": 1.0050057172775269, + "learning_rate": 3.8072469174746794e-05, + "loss": 0.9094, + "step": 16019 + }, + { + "epoch": 2.852207977207977, + "grad_norm": 0.8405657410621643, + "learning_rate": 3.806147933128962e-05, + "loss": 0.7737, + "step": 16020 + }, + { + "epoch": 2.85238603988604, + "grad_norm": 0.8127378821372986, + "learning_rate": 3.8050490701389085e-05, + "loss": 0.9102, + "step": 16021 + }, + { + "epoch": 2.8525641025641026, + "grad_norm": 0.8622255921363831, + "learning_rate": 3.8039503285260506e-05, + "loss": 0.8815, + "step": 16022 + }, + { + "epoch": 2.8527421652421654, + "grad_norm": 0.8802367448806763, + "learning_rate": 3.802851708311913e-05, + "loss": 1.0123, + "step": 16023 + }, + { + "epoch": 2.8529202279202277, + "grad_norm": 0.908149778842926, + "learning_rate": 3.801753209518024e-05, + "loss": 0.6808, + "step": 16024 + }, + { + "epoch": 2.8530982905982905, + "grad_norm": 0.9346339702606201, + "learning_rate": 3.8006548321659055e-05, + "loss": 1.1107, + "step": 16025 + }, + { + "epoch": 2.853276353276353, + "grad_norm": 0.945125937461853, + "learning_rate": 3.799556576277077e-05, + "loss": 0.6578, + "step": 16026 + }, + { + "epoch": 2.853454415954416, + "grad_norm": 0.8294890522956848, + "learning_rate": 3.798458441873054e-05, + "loss": 0.869, + "step": 16027 + }, + { + "epoch": 2.8536324786324787, + "grad_norm": 0.7922961115837097, + "learning_rate": 3.797360428975358e-05, + "loss": 0.718, + "step": 16028 + }, + { + "epoch": 2.8538105413105415, + "grad_norm": 1.0540844202041626, + "learning_rate": 3.7962625376055005e-05, + "loss": 0.8287, + "step": 16029 + }, + { + "epoch": 2.853988603988604, + "grad_norm": 0.9409742951393127, + "learning_rate": 3.795164767784991e-05, + "loss": 1.0484, + "step": 16030 + }, + { + "epoch": 2.8541666666666665, + "grad_norm": 0.7328341603279114, + "learning_rate": 3.7940671195353385e-05, + "loss": 0.7603, + "step": 16031 + }, + { + "epoch": 2.8543447293447293, + "grad_norm": 0.9151208996772766, + "learning_rate": 3.792969592878045e-05, + "loss": 0.7523, + "step": 16032 + }, + { + "epoch": 2.854522792022792, + "grad_norm": 0.7935783267021179, + "learning_rate": 3.791872187834626e-05, + "loss": 0.7559, + "step": 16033 + }, + { + "epoch": 2.8547008547008548, + "grad_norm": 0.8030906915664673, + "learning_rate": 3.790774904426568e-05, + "loss": 0.7905, + "step": 16034 + }, + { + "epoch": 2.8548789173789175, + "grad_norm": 0.8756175637245178, + "learning_rate": 3.789677742675384e-05, + "loss": 0.9067, + "step": 16035 + }, + { + "epoch": 2.85505698005698, + "grad_norm": 0.7602807283401489, + "learning_rate": 3.788580702602558e-05, + "loss": 0.9192, + "step": 16036 + }, + { + "epoch": 2.8552350427350426, + "grad_norm": 0.9411010146141052, + "learning_rate": 3.787483784229592e-05, + "loss": 0.8192, + "step": 16037 + }, + { + "epoch": 2.8554131054131053, + "grad_norm": 0.9473391175270081, + "learning_rate": 3.786386987577976e-05, + "loss": 0.7845, + "step": 16038 + }, + { + "epoch": 2.855591168091168, + "grad_norm": 0.9226218461990356, + "learning_rate": 3.7852903126692e-05, + "loss": 0.8712, + "step": 16039 + }, + { + "epoch": 2.855769230769231, + "grad_norm": 0.9519350528717041, + "learning_rate": 3.78419375952475e-05, + "loss": 0.9584, + "step": 16040 + }, + { + "epoch": 2.8559472934472936, + "grad_norm": 0.825547456741333, + "learning_rate": 3.783097328166111e-05, + "loss": 0.9279, + "step": 16041 + }, + { + "epoch": 2.8561253561253563, + "grad_norm": 0.8645279407501221, + "learning_rate": 3.782001018614763e-05, + "loss": 0.83, + "step": 16042 + }, + { + "epoch": 2.8563034188034186, + "grad_norm": 0.827126145362854, + "learning_rate": 3.7809048308921936e-05, + "loss": 0.7661, + "step": 16043 + }, + { + "epoch": 2.8564814814814814, + "grad_norm": 0.9441137909889221, + "learning_rate": 3.779808765019869e-05, + "loss": 0.8745, + "step": 16044 + }, + { + "epoch": 2.856659544159544, + "grad_norm": 0.8505343794822693, + "learning_rate": 3.7787128210192736e-05, + "loss": 0.8176, + "step": 16045 + }, + { + "epoch": 2.856837606837607, + "grad_norm": 0.8797150254249573, + "learning_rate": 3.777616998911876e-05, + "loss": 0.7018, + "step": 16046 + }, + { + "epoch": 2.8570156695156697, + "grad_norm": 0.8386834263801575, + "learning_rate": 3.776521298719144e-05, + "loss": 0.9805, + "step": 16047 + }, + { + "epoch": 2.857193732193732, + "grad_norm": 0.818373441696167, + "learning_rate": 3.775425720462558e-05, + "loss": 0.7782, + "step": 16048 + }, + { + "epoch": 2.8573717948717947, + "grad_norm": 0.760405957698822, + "learning_rate": 3.774330264163566e-05, + "loss": 0.7283, + "step": 16049 + }, + { + "epoch": 2.8575498575498575, + "grad_norm": 0.9082552194595337, + "learning_rate": 3.7732349298436465e-05, + "loss": 0.7594, + "step": 16050 + }, + { + "epoch": 2.85772792022792, + "grad_norm": 0.859868586063385, + "learning_rate": 3.7721397175242477e-05, + "loss": 0.7841, + "step": 16051 + }, + { + "epoch": 2.857905982905983, + "grad_norm": 1.031545877456665, + "learning_rate": 3.771044627226836e-05, + "loss": 1.2157, + "step": 16052 + }, + { + "epoch": 2.8580840455840457, + "grad_norm": 0.859491765499115, + "learning_rate": 3.769949658972867e-05, + "loss": 0.7942, + "step": 16053 + }, + { + "epoch": 2.8582621082621085, + "grad_norm": 0.775382936000824, + "learning_rate": 3.768854812783791e-05, + "loss": 0.7321, + "step": 16054 + }, + { + "epoch": 2.8584401709401708, + "grad_norm": 0.9268868565559387, + "learning_rate": 3.767760088681062e-05, + "loss": 1.0104, + "step": 16055 + }, + { + "epoch": 2.8586182336182335, + "grad_norm": 0.8408828973770142, + "learning_rate": 3.7666654866861274e-05, + "loss": 0.5298, + "step": 16056 + }, + { + "epoch": 2.8587962962962963, + "grad_norm": 0.8417157530784607, + "learning_rate": 3.76557100682043e-05, + "loss": 0.9123, + "step": 16057 + }, + { + "epoch": 2.858974358974359, + "grad_norm": 0.8079593777656555, + "learning_rate": 3.764476649105425e-05, + "loss": 0.794, + "step": 16058 + }, + { + "epoch": 2.859152421652422, + "grad_norm": 0.9449031352996826, + "learning_rate": 3.763382413562541e-05, + "loss": 1.1281, + "step": 16059 + }, + { + "epoch": 2.859330484330484, + "grad_norm": 0.8985004425048828, + "learning_rate": 3.762288300213228e-05, + "loss": 0.929, + "step": 16060 + }, + { + "epoch": 2.859508547008547, + "grad_norm": 0.9850391149520874, + "learning_rate": 3.761194309078913e-05, + "loss": 0.9796, + "step": 16061 + }, + { + "epoch": 2.8596866096866096, + "grad_norm": 0.9231089949607849, + "learning_rate": 3.760100440181038e-05, + "loss": 0.6981, + "step": 16062 + }, + { + "epoch": 2.8598646723646723, + "grad_norm": 0.8458681702613831, + "learning_rate": 3.759006693541033e-05, + "loss": 0.9502, + "step": 16063 + }, + { + "epoch": 2.860042735042735, + "grad_norm": 0.8494541645050049, + "learning_rate": 3.7579130691803266e-05, + "loss": 0.7477, + "step": 16064 + }, + { + "epoch": 2.860220797720798, + "grad_norm": 0.879878580570221, + "learning_rate": 3.756819567120348e-05, + "loss": 0.8426, + "step": 16065 + }, + { + "epoch": 2.8603988603988606, + "grad_norm": 0.8161541223526001, + "learning_rate": 3.7557261873825155e-05, + "loss": 0.7411, + "step": 16066 + }, + { + "epoch": 2.8605769230769234, + "grad_norm": 0.9438506364822388, + "learning_rate": 3.754632929988262e-05, + "loss": 0.8494, + "step": 16067 + }, + { + "epoch": 2.8607549857549857, + "grad_norm": 0.8552418351173401, + "learning_rate": 3.753539794959002e-05, + "loss": 0.766, + "step": 16068 + }, + { + "epoch": 2.8609330484330484, + "grad_norm": 0.8670600056648254, + "learning_rate": 3.7524467823161546e-05, + "loss": 0.7462, + "step": 16069 + }, + { + "epoch": 2.861111111111111, + "grad_norm": 0.7906678318977356, + "learning_rate": 3.751353892081134e-05, + "loss": 0.791, + "step": 16070 + }, + { + "epoch": 2.861289173789174, + "grad_norm": 0.8461915254592896, + "learning_rate": 3.7502611242753536e-05, + "loss": 0.6979, + "step": 16071 + }, + { + "epoch": 2.861467236467236, + "grad_norm": 0.8197309970855713, + "learning_rate": 3.749168478920223e-05, + "loss": 0.8095, + "step": 16072 + }, + { + "epoch": 2.861645299145299, + "grad_norm": 0.9489047527313232, + "learning_rate": 3.7480759560371516e-05, + "loss": 0.9462, + "step": 16073 + }, + { + "epoch": 2.8618233618233617, + "grad_norm": 0.8539329767227173, + "learning_rate": 3.7469835556475405e-05, + "loss": 0.853, + "step": 16074 + }, + { + "epoch": 2.8620014245014245, + "grad_norm": 0.9104743003845215, + "learning_rate": 3.745891277772805e-05, + "loss": 0.757, + "step": 16075 + }, + { + "epoch": 2.8621794871794872, + "grad_norm": 0.8277523517608643, + "learning_rate": 3.744799122434332e-05, + "loss": 0.8073, + "step": 16076 + }, + { + "epoch": 2.86235754985755, + "grad_norm": 0.866422176361084, + "learning_rate": 3.743707089653527e-05, + "loss": 0.7201, + "step": 16077 + }, + { + "epoch": 2.8625356125356127, + "grad_norm": 0.8722748160362244, + "learning_rate": 3.742615179451787e-05, + "loss": 0.7901, + "step": 16078 + }, + { + "epoch": 2.8627136752136755, + "grad_norm": 0.74676513671875, + "learning_rate": 3.741523391850504e-05, + "loss": 0.7558, + "step": 16079 + }, + { + "epoch": 2.862891737891738, + "grad_norm": 0.7674166560173035, + "learning_rate": 3.740431726871069e-05, + "loss": 0.8699, + "step": 16080 + }, + { + "epoch": 2.8630698005698005, + "grad_norm": 0.9028998613357544, + "learning_rate": 3.739340184534871e-05, + "loss": 1.136, + "step": 16081 + }, + { + "epoch": 2.8632478632478633, + "grad_norm": 0.8240773677825928, + "learning_rate": 3.7382487648632936e-05, + "loss": 0.9357, + "step": 16082 + }, + { + "epoch": 2.863425925925926, + "grad_norm": 0.8877659440040588, + "learning_rate": 3.737157467877731e-05, + "loss": 0.9167, + "step": 16083 + }, + { + "epoch": 2.8636039886039883, + "grad_norm": 0.9677366614341736, + "learning_rate": 3.73606629359955e-05, + "loss": 0.8342, + "step": 16084 + }, + { + "epoch": 2.863782051282051, + "grad_norm": 0.8721164464950562, + "learning_rate": 3.734975242050146e-05, + "loss": 0.9195, + "step": 16085 + }, + { + "epoch": 2.863960113960114, + "grad_norm": 0.9791151881217957, + "learning_rate": 3.733884313250879e-05, + "loss": 1.0011, + "step": 16086 + }, + { + "epoch": 2.8641381766381766, + "grad_norm": 0.7869369983673096, + "learning_rate": 3.7327935072231366e-05, + "loss": 0.7998, + "step": 16087 + }, + { + "epoch": 2.8643162393162394, + "grad_norm": 0.891656756401062, + "learning_rate": 3.731702823988287e-05, + "loss": 0.7287, + "step": 16088 + }, + { + "epoch": 2.864494301994302, + "grad_norm": 0.8720460534095764, + "learning_rate": 3.7306122635676955e-05, + "loss": 0.9492, + "step": 16089 + }, + { + "epoch": 2.864672364672365, + "grad_norm": 0.7878959774971008, + "learning_rate": 3.72952182598274e-05, + "loss": 0.7652, + "step": 16090 + }, + { + "epoch": 2.8648504273504276, + "grad_norm": 0.9350453019142151, + "learning_rate": 3.728431511254772e-05, + "loss": 0.8661, + "step": 16091 + }, + { + "epoch": 2.86502849002849, + "grad_norm": 0.7575289011001587, + "learning_rate": 3.727341319405163e-05, + "loss": 0.7943, + "step": 16092 + }, + { + "epoch": 2.8652065527065527, + "grad_norm": 0.8256776928901672, + "learning_rate": 3.7262512504552716e-05, + "loss": 1.0025, + "step": 16093 + }, + { + "epoch": 2.8653846153846154, + "grad_norm": 0.7274962067604065, + "learning_rate": 3.7251613044264536e-05, + "loss": 0.7628, + "step": 16094 + }, + { + "epoch": 2.865562678062678, + "grad_norm": 0.867734968662262, + "learning_rate": 3.7240714813400646e-05, + "loss": 0.8931, + "step": 16095 + }, + { + "epoch": 2.8657407407407405, + "grad_norm": 0.8225845098495483, + "learning_rate": 3.722981781217458e-05, + "loss": 0.7934, + "step": 16096 + }, + { + "epoch": 2.8659188034188032, + "grad_norm": 0.8180573582649231, + "learning_rate": 3.721892204079985e-05, + "loss": 0.8119, + "step": 16097 + }, + { + "epoch": 2.866096866096866, + "grad_norm": 1.0235565900802612, + "learning_rate": 3.720802749948993e-05, + "loss": 0.8281, + "step": 16098 + }, + { + "epoch": 2.8662749287749287, + "grad_norm": 0.7290656566619873, + "learning_rate": 3.719713418845823e-05, + "loss": 0.7099, + "step": 16099 + }, + { + "epoch": 2.8664529914529915, + "grad_norm": 0.8408772349357605, + "learning_rate": 3.718624210791828e-05, + "loss": 0.8631, + "step": 16100 + }, + { + "epoch": 2.8666310541310542, + "grad_norm": 0.8182529807090759, + "learning_rate": 3.717535125808338e-05, + "loss": 0.7584, + "step": 16101 + }, + { + "epoch": 2.866809116809117, + "grad_norm": 0.8381599187850952, + "learning_rate": 3.716446163916699e-05, + "loss": 0.8735, + "step": 16102 + }, + { + "epoch": 2.8669871794871797, + "grad_norm": 0.8975555300712585, + "learning_rate": 3.715357325138245e-05, + "loss": 0.7564, + "step": 16103 + }, + { + "epoch": 2.867165242165242, + "grad_norm": 0.9531118869781494, + "learning_rate": 3.714268609494309e-05, + "loss": 0.7627, + "step": 16104 + }, + { + "epoch": 2.867343304843305, + "grad_norm": 0.853065550327301, + "learning_rate": 3.7131800170062216e-05, + "loss": 0.8001, + "step": 16105 + }, + { + "epoch": 2.8675213675213675, + "grad_norm": 0.788351833820343, + "learning_rate": 3.7120915476953085e-05, + "loss": 0.7935, + "step": 16106 + }, + { + "epoch": 2.8676994301994303, + "grad_norm": 0.9121149778366089, + "learning_rate": 3.711003201582908e-05, + "loss": 0.9212, + "step": 16107 + }, + { + "epoch": 2.8678774928774926, + "grad_norm": 0.8156226277351379, + "learning_rate": 3.7099149786903263e-05, + "loss": 0.86, + "step": 16108 + }, + { + "epoch": 2.8680555555555554, + "grad_norm": 0.8555662035942078, + "learning_rate": 3.708826879038899e-05, + "loss": 0.8872, + "step": 16109 + }, + { + "epoch": 2.868233618233618, + "grad_norm": 1.0395163297653198, + "learning_rate": 3.70773890264994e-05, + "loss": 0.9794, + "step": 16110 + }, + { + "epoch": 2.868411680911681, + "grad_norm": 0.7535551190376282, + "learning_rate": 3.706651049544766e-05, + "loss": 0.91, + "step": 16111 + }, + { + "epoch": 2.8685897435897436, + "grad_norm": 1.0145034790039062, + "learning_rate": 3.705563319744691e-05, + "loss": 0.9386, + "step": 16112 + }, + { + "epoch": 2.8687678062678064, + "grad_norm": 0.8577025532722473, + "learning_rate": 3.704475713271029e-05, + "loss": 1.0491, + "step": 16113 + }, + { + "epoch": 2.868945868945869, + "grad_norm": 0.8291150331497192, + "learning_rate": 3.7033882301450815e-05, + "loss": 0.8418, + "step": 16114 + }, + { + "epoch": 2.869123931623932, + "grad_norm": 0.7628613710403442, + "learning_rate": 3.70230087038817e-05, + "loss": 0.681, + "step": 16115 + }, + { + "epoch": 2.869301994301994, + "grad_norm": 0.8664639592170715, + "learning_rate": 3.701213634021583e-05, + "loss": 1.0247, + "step": 16116 + }, + { + "epoch": 2.869480056980057, + "grad_norm": 0.9613258838653564, + "learning_rate": 3.700126521066635e-05, + "loss": 1.1058, + "step": 16117 + }, + { + "epoch": 2.8696581196581197, + "grad_norm": 0.8279051780700684, + "learning_rate": 3.699039531544619e-05, + "loss": 0.859, + "step": 16118 + }, + { + "epoch": 2.8698361823361824, + "grad_norm": 0.8285593390464783, + "learning_rate": 3.697952665476836e-05, + "loss": 0.9008, + "step": 16119 + }, + { + "epoch": 2.870014245014245, + "grad_norm": 0.9056670069694519, + "learning_rate": 3.696865922884578e-05, + "loss": 0.965, + "step": 16120 + }, + { + "epoch": 2.8701923076923075, + "grad_norm": 0.7092664837837219, + "learning_rate": 3.69577930378914e-05, + "loss": 0.7417, + "step": 16121 + }, + { + "epoch": 2.8703703703703702, + "grad_norm": 0.8726393580436707, + "learning_rate": 3.6946928082118096e-05, + "loss": 0.8582, + "step": 16122 + }, + { + "epoch": 2.870548433048433, + "grad_norm": 1.0046098232269287, + "learning_rate": 3.693606436173875e-05, + "loss": 0.9616, + "step": 16123 + }, + { + "epoch": 2.8707264957264957, + "grad_norm": 0.7739760875701904, + "learning_rate": 3.69252018769662e-05, + "loss": 0.8775, + "step": 16124 + }, + { + "epoch": 2.8709045584045585, + "grad_norm": 0.9054580926895142, + "learning_rate": 3.6914340628013344e-05, + "loss": 0.9205, + "step": 16125 + }, + { + "epoch": 2.8710826210826212, + "grad_norm": 0.8324142694473267, + "learning_rate": 3.690348061509288e-05, + "loss": 0.8712, + "step": 16126 + }, + { + "epoch": 2.871260683760684, + "grad_norm": 0.9162326455116272, + "learning_rate": 3.6892621838417664e-05, + "loss": 0.9102, + "step": 16127 + }, + { + "epoch": 2.8714387464387463, + "grad_norm": 0.8579963445663452, + "learning_rate": 3.688176429820044e-05, + "loss": 0.8337, + "step": 16128 + }, + { + "epoch": 2.871616809116809, + "grad_norm": 0.7649274468421936, + "learning_rate": 3.687090799465388e-05, + "loss": 0.6982, + "step": 16129 + }, + { + "epoch": 2.871794871794872, + "grad_norm": 0.9612696766853333, + "learning_rate": 3.6860052927990816e-05, + "loss": 0.8779, + "step": 16130 + }, + { + "epoch": 2.8719729344729346, + "grad_norm": 0.982455313205719, + "learning_rate": 3.6849199098423795e-05, + "loss": 0.8145, + "step": 16131 + }, + { + "epoch": 2.8721509971509973, + "grad_norm": 0.7292434573173523, + "learning_rate": 3.6838346506165587e-05, + "loss": 0.6711, + "step": 16132 + }, + { + "epoch": 2.8723290598290596, + "grad_norm": 0.877310574054718, + "learning_rate": 3.68274951514287e-05, + "loss": 0.7682, + "step": 16133 + }, + { + "epoch": 2.8725071225071224, + "grad_norm": 0.9633384943008423, + "learning_rate": 3.681664503442586e-05, + "loss": 1.0046, + "step": 16134 + }, + { + "epoch": 2.872685185185185, + "grad_norm": 0.940661609172821, + "learning_rate": 3.680579615536961e-05, + "loss": 0.897, + "step": 16135 + }, + { + "epoch": 2.872863247863248, + "grad_norm": 1.0335214138031006, + "learning_rate": 3.6794948514472505e-05, + "loss": 0.8836, + "step": 16136 + }, + { + "epoch": 2.8730413105413106, + "grad_norm": 0.8682044148445129, + "learning_rate": 3.6784102111947084e-05, + "loss": 0.7733, + "step": 16137 + }, + { + "epoch": 2.8732193732193734, + "grad_norm": 0.8767847418785095, + "learning_rate": 3.677325694800586e-05, + "loss": 0.8828, + "step": 16138 + }, + { + "epoch": 2.873397435897436, + "grad_norm": 0.954585075378418, + "learning_rate": 3.6762413022861305e-05, + "loss": 0.9294, + "step": 16139 + }, + { + "epoch": 2.8735754985754984, + "grad_norm": 0.8497310876846313, + "learning_rate": 3.675157033672596e-05, + "loss": 0.9396, + "step": 16140 + }, + { + "epoch": 2.873753561253561, + "grad_norm": 0.7619023323059082, + "learning_rate": 3.674072888981214e-05, + "loss": 0.7467, + "step": 16141 + }, + { + "epoch": 2.873931623931624, + "grad_norm": 0.9939205646514893, + "learning_rate": 3.6729888682332394e-05, + "loss": 0.9122, + "step": 16142 + }, + { + "epoch": 2.8741096866096867, + "grad_norm": 0.943192183971405, + "learning_rate": 3.671904971449899e-05, + "loss": 0.8878, + "step": 16143 + }, + { + "epoch": 2.8742877492877494, + "grad_norm": 0.8002169728279114, + "learning_rate": 3.6708211986524365e-05, + "loss": 0.8337, + "step": 16144 + }, + { + "epoch": 2.8744658119658117, + "grad_norm": 0.7558008432388306, + "learning_rate": 3.669737549862087e-05, + "loss": 0.6592, + "step": 16145 + }, + { + "epoch": 2.8746438746438745, + "grad_norm": 0.8732983469963074, + "learning_rate": 3.6686540251000756e-05, + "loss": 0.9166, + "step": 16146 + }, + { + "epoch": 2.8748219373219372, + "grad_norm": 0.8272808194160461, + "learning_rate": 3.667570624387643e-05, + "loss": 0.6342, + "step": 16147 + }, + { + "epoch": 2.875, + "grad_norm": 0.8012139797210693, + "learning_rate": 3.666487347746004e-05, + "loss": 0.8436, + "step": 16148 + }, + { + "epoch": 2.8751780626780628, + "grad_norm": 0.8791360259056091, + "learning_rate": 3.66540419519639e-05, + "loss": 0.8634, + "step": 16149 + }, + { + "epoch": 2.8753561253561255, + "grad_norm": 0.8274601101875305, + "learning_rate": 3.6643211667600244e-05, + "loss": 0.7488, + "step": 16150 + }, + { + "epoch": 2.8755341880341883, + "grad_norm": 0.8390287756919861, + "learning_rate": 3.663238262458122e-05, + "loss": 0.7084, + "step": 16151 + }, + { + "epoch": 2.8757122507122506, + "grad_norm": 0.921089768409729, + "learning_rate": 3.662155482311903e-05, + "loss": 0.8909, + "step": 16152 + }, + { + "epoch": 2.8758903133903133, + "grad_norm": 0.8790102601051331, + "learning_rate": 3.661072826342583e-05, + "loss": 0.8235, + "step": 16153 + }, + { + "epoch": 2.876068376068376, + "grad_norm": 0.8030495643615723, + "learning_rate": 3.659990294571368e-05, + "loss": 0.874, + "step": 16154 + }, + { + "epoch": 2.876246438746439, + "grad_norm": 0.9690510034561157, + "learning_rate": 3.6589078870194804e-05, + "loss": 1.0926, + "step": 16155 + }, + { + "epoch": 2.8764245014245016, + "grad_norm": 0.8150941133499146, + "learning_rate": 3.657825603708114e-05, + "loss": 0.861, + "step": 16156 + }, + { + "epoch": 2.876602564102564, + "grad_norm": 0.8865286111831665, + "learning_rate": 3.656743444658486e-05, + "loss": 0.8219, + "step": 16157 + }, + { + "epoch": 2.8767806267806266, + "grad_norm": 0.8591124415397644, + "learning_rate": 3.655661409891786e-05, + "loss": 0.889, + "step": 16158 + }, + { + "epoch": 2.8769586894586894, + "grad_norm": 0.8625560402870178, + "learning_rate": 3.6545794994292256e-05, + "loss": 0.9581, + "step": 16159 + }, + { + "epoch": 2.877136752136752, + "grad_norm": 0.8699239492416382, + "learning_rate": 3.653497713291999e-05, + "loss": 0.7275, + "step": 16160 + }, + { + "epoch": 2.877314814814815, + "grad_norm": 0.9005762338638306, + "learning_rate": 3.652416051501301e-05, + "loss": 0.8894, + "step": 16161 + }, + { + "epoch": 2.8774928774928776, + "grad_norm": 0.7541293501853943, + "learning_rate": 3.651334514078323e-05, + "loss": 0.7265, + "step": 16162 + }, + { + "epoch": 2.8776709401709404, + "grad_norm": 0.9799004793167114, + "learning_rate": 3.650253101044258e-05, + "loss": 0.8817, + "step": 16163 + }, + { + "epoch": 2.8778490028490027, + "grad_norm": 0.7796139717102051, + "learning_rate": 3.64917181242029e-05, + "loss": 0.7105, + "step": 16164 + }, + { + "epoch": 2.8780270655270654, + "grad_norm": 0.8818691968917847, + "learning_rate": 3.648090648227613e-05, + "loss": 0.8868, + "step": 16165 + }, + { + "epoch": 2.878205128205128, + "grad_norm": 0.7982428073883057, + "learning_rate": 3.647009608487399e-05, + "loss": 0.9269, + "step": 16166 + }, + { + "epoch": 2.878383190883191, + "grad_norm": 0.9602195620536804, + "learning_rate": 3.645928693220838e-05, + "loss": 0.8264, + "step": 16167 + }, + { + "epoch": 2.8785612535612537, + "grad_norm": 0.8941618800163269, + "learning_rate": 3.6448479024491054e-05, + "loss": 0.8009, + "step": 16168 + }, + { + "epoch": 2.878739316239316, + "grad_norm": 0.7777221202850342, + "learning_rate": 3.643767236193375e-05, + "loss": 0.7734, + "step": 16169 + }, + { + "epoch": 2.8789173789173788, + "grad_norm": 0.8050239086151123, + "learning_rate": 3.642686694474823e-05, + "loss": 0.9732, + "step": 16170 + }, + { + "epoch": 2.8790954415954415, + "grad_norm": 0.8437817096710205, + "learning_rate": 3.6416062773146156e-05, + "loss": 0.8161, + "step": 16171 + }, + { + "epoch": 2.8792735042735043, + "grad_norm": 0.7887414693832397, + "learning_rate": 3.64052598473393e-05, + "loss": 0.9165, + "step": 16172 + }, + { + "epoch": 2.879451566951567, + "grad_norm": 0.8671287894248962, + "learning_rate": 3.639445816753921e-05, + "loss": 0.9519, + "step": 16173 + }, + { + "epoch": 2.8796296296296298, + "grad_norm": 0.8444932103157043, + "learning_rate": 3.638365773395763e-05, + "loss": 0.784, + "step": 16174 + }, + { + "epoch": 2.8798076923076925, + "grad_norm": 0.8580447435379028, + "learning_rate": 3.637285854680612e-05, + "loss": 0.797, + "step": 16175 + }, + { + "epoch": 2.879985754985755, + "grad_norm": 0.7977848052978516, + "learning_rate": 3.636206060629627e-05, + "loss": 0.8552, + "step": 16176 + }, + { + "epoch": 2.8801638176638176, + "grad_norm": 0.8350155353546143, + "learning_rate": 3.6351263912639644e-05, + "loss": 0.9481, + "step": 16177 + }, + { + "epoch": 2.8803418803418803, + "grad_norm": 0.8759897351264954, + "learning_rate": 3.634046846604778e-05, + "loss": 0.9972, + "step": 16178 + }, + { + "epoch": 2.880519943019943, + "grad_norm": 0.8259425163269043, + "learning_rate": 3.6329674266732194e-05, + "loss": 0.8947, + "step": 16179 + }, + { + "epoch": 2.880698005698006, + "grad_norm": 0.7992371320724487, + "learning_rate": 3.631888131490438e-05, + "loss": 0.8346, + "step": 16180 + }, + { + "epoch": 2.880876068376068, + "grad_norm": 0.9702637195587158, + "learning_rate": 3.6308089610775775e-05, + "loss": 0.9543, + "step": 16181 + }, + { + "epoch": 2.881054131054131, + "grad_norm": 0.8494347929954529, + "learning_rate": 3.6297299154557905e-05, + "loss": 0.8672, + "step": 16182 + }, + { + "epoch": 2.8812321937321936, + "grad_norm": 0.8098355531692505, + "learning_rate": 3.628650994646207e-05, + "loss": 0.8356, + "step": 16183 + }, + { + "epoch": 2.8814102564102564, + "grad_norm": 0.7736916542053223, + "learning_rate": 3.627572198669974e-05, + "loss": 0.849, + "step": 16184 + }, + { + "epoch": 2.881588319088319, + "grad_norm": 0.806710422039032, + "learning_rate": 3.626493527548226e-05, + "loss": 0.9479, + "step": 16185 + }, + { + "epoch": 2.881766381766382, + "grad_norm": 1.0013649463653564, + "learning_rate": 3.625414981302095e-05, + "loss": 0.967, + "step": 16186 + }, + { + "epoch": 2.8819444444444446, + "grad_norm": 0.8116905093193054, + "learning_rate": 3.624336559952723e-05, + "loss": 0.8608, + "step": 16187 + }, + { + "epoch": 2.8821225071225074, + "grad_norm": 0.7848439812660217, + "learning_rate": 3.6232582635212233e-05, + "loss": 0.9224, + "step": 16188 + }, + { + "epoch": 2.8823005698005697, + "grad_norm": 0.9881102442741394, + "learning_rate": 3.622180092028735e-05, + "loss": 0.8089, + "step": 16189 + }, + { + "epoch": 2.8824786324786325, + "grad_norm": 0.790452778339386, + "learning_rate": 3.6211020454963796e-05, + "loss": 0.8312, + "step": 16190 + }, + { + "epoch": 2.882656695156695, + "grad_norm": 0.8228929042816162, + "learning_rate": 3.6200241239452784e-05, + "loss": 0.8277, + "step": 16191 + }, + { + "epoch": 2.882834757834758, + "grad_norm": 0.871133029460907, + "learning_rate": 3.618946327396552e-05, + "loss": 0.8861, + "step": 16192 + }, + { + "epoch": 2.8830128205128203, + "grad_norm": 0.8964230418205261, + "learning_rate": 3.617868655871316e-05, + "loss": 0.9987, + "step": 16193 + }, + { + "epoch": 2.883190883190883, + "grad_norm": 0.8292636275291443, + "learning_rate": 3.6167911093906856e-05, + "loss": 0.9823, + "step": 16194 + }, + { + "epoch": 2.8833689458689458, + "grad_norm": 0.9594070315361023, + "learning_rate": 3.615713687975774e-05, + "loss": 0.7963, + "step": 16195 + }, + { + "epoch": 2.8835470085470085, + "grad_norm": 1.014891266822815, + "learning_rate": 3.6146363916476864e-05, + "loss": 0.9203, + "step": 16196 + }, + { + "epoch": 2.8837250712250713, + "grad_norm": 0.8205485343933105, + "learning_rate": 3.6135592204275424e-05, + "loss": 0.894, + "step": 16197 + }, + { + "epoch": 2.883903133903134, + "grad_norm": 0.9057072997093201, + "learning_rate": 3.6124821743364315e-05, + "loss": 0.9131, + "step": 16198 + }, + { + "epoch": 2.8840811965811968, + "grad_norm": 0.8080529570579529, + "learning_rate": 3.6114052533954665e-05, + "loss": 0.9995, + "step": 16199 + }, + { + "epoch": 2.8842592592592595, + "grad_norm": 0.762264609336853, + "learning_rate": 3.6103284576257446e-05, + "loss": 0.915, + "step": 16200 + }, + { + "epoch": 2.884437321937322, + "grad_norm": 0.8386275172233582, + "learning_rate": 3.609251787048363e-05, + "loss": 0.9016, + "step": 16201 + }, + { + "epoch": 2.8846153846153846, + "grad_norm": 0.9676657319068909, + "learning_rate": 3.608175241684417e-05, + "loss": 1.1004, + "step": 16202 + }, + { + "epoch": 2.8847934472934473, + "grad_norm": 0.8155630826950073, + "learning_rate": 3.607098821554999e-05, + "loss": 0.8579, + "step": 16203 + }, + { + "epoch": 2.88497150997151, + "grad_norm": 0.8426685333251953, + "learning_rate": 3.606022526681201e-05, + "loss": 0.8722, + "step": 16204 + }, + { + "epoch": 2.8851495726495724, + "grad_norm": 0.7646408081054688, + "learning_rate": 3.604946357084105e-05, + "loss": 0.7806, + "step": 16205 + }, + { + "epoch": 2.885327635327635, + "grad_norm": 0.808560848236084, + "learning_rate": 3.603870312784803e-05, + "loss": 0.8419, + "step": 16206 + }, + { + "epoch": 2.885505698005698, + "grad_norm": 0.7197920083999634, + "learning_rate": 3.602794393804376e-05, + "loss": 0.7356, + "step": 16207 + }, + { + "epoch": 2.8856837606837606, + "grad_norm": 0.9405228495597839, + "learning_rate": 3.6017186001639036e-05, + "loss": 0.7279, + "step": 16208 + }, + { + "epoch": 2.8858618233618234, + "grad_norm": 0.8910958170890808, + "learning_rate": 3.600642931884465e-05, + "loss": 0.7653, + "step": 16209 + }, + { + "epoch": 2.886039886039886, + "grad_norm": 0.7354677319526672, + "learning_rate": 3.599567388987134e-05, + "loss": 0.6884, + "step": 16210 + }, + { + "epoch": 2.886217948717949, + "grad_norm": 0.75583815574646, + "learning_rate": 3.598491971492981e-05, + "loss": 0.7555, + "step": 16211 + }, + { + "epoch": 2.8863960113960117, + "grad_norm": 0.839042603969574, + "learning_rate": 3.597416679423086e-05, + "loss": 1.0522, + "step": 16212 + }, + { + "epoch": 2.886574074074074, + "grad_norm": 0.7965270280838013, + "learning_rate": 3.596341512798505e-05, + "loss": 0.9101, + "step": 16213 + }, + { + "epoch": 2.8867521367521367, + "grad_norm": 0.9170811772346497, + "learning_rate": 3.5952664716403154e-05, + "loss": 0.9078, + "step": 16214 + }, + { + "epoch": 2.8869301994301995, + "grad_norm": 0.8612267374992371, + "learning_rate": 3.5941915559695685e-05, + "loss": 0.8615, + "step": 16215 + }, + { + "epoch": 2.887108262108262, + "grad_norm": 0.9182412028312683, + "learning_rate": 3.593116765807335e-05, + "loss": 0.8918, + "step": 16216 + }, + { + "epoch": 2.8872863247863245, + "grad_norm": 0.9452505707740784, + "learning_rate": 3.592042101174669e-05, + "loss": 1.0386, + "step": 16217 + }, + { + "epoch": 2.8874643874643873, + "grad_norm": 0.8544769287109375, + "learning_rate": 3.5909675620926255e-05, + "loss": 1.0305, + "step": 16218 + }, + { + "epoch": 2.88764245014245, + "grad_norm": 0.8184592127799988, + "learning_rate": 3.5898931485822605e-05, + "loss": 0.6815, + "step": 16219 + }, + { + "epoch": 2.8878205128205128, + "grad_norm": 0.8263654708862305, + "learning_rate": 3.5888188606646224e-05, + "loss": 1.0097, + "step": 16220 + }, + { + "epoch": 2.8879985754985755, + "grad_norm": 0.8290582299232483, + "learning_rate": 3.587744698360758e-05, + "loss": 0.6539, + "step": 16221 + }, + { + "epoch": 2.8881766381766383, + "grad_norm": 0.7936849594116211, + "learning_rate": 3.5866706616917226e-05, + "loss": 0.9524, + "step": 16222 + }, + { + "epoch": 2.888354700854701, + "grad_norm": 0.9449033737182617, + "learning_rate": 3.585596750678546e-05, + "loss": 0.9243, + "step": 16223 + }, + { + "epoch": 2.888532763532764, + "grad_norm": 0.7599559426307678, + "learning_rate": 3.58452296534228e-05, + "loss": 0.8597, + "step": 16224 + }, + { + "epoch": 2.888710826210826, + "grad_norm": 0.8485760688781738, + "learning_rate": 3.583449305703959e-05, + "loss": 0.7236, + "step": 16225 + }, + { + "epoch": 2.888888888888889, + "grad_norm": 0.8510624170303345, + "learning_rate": 3.582375771784616e-05, + "loss": 0.9081, + "step": 16226 + }, + { + "epoch": 2.8890669515669516, + "grad_norm": 0.815827488899231, + "learning_rate": 3.581302363605296e-05, + "loss": 0.8259, + "step": 16227 + }, + { + "epoch": 2.8892450142450143, + "grad_norm": 0.7588803768157959, + "learning_rate": 3.580229081187016e-05, + "loss": 0.8722, + "step": 16228 + }, + { + "epoch": 2.8894230769230766, + "grad_norm": 1.0699365139007568, + "learning_rate": 3.579155924550817e-05, + "loss": 1.0261, + "step": 16229 + }, + { + "epoch": 2.8896011396011394, + "grad_norm": 0.9127700924873352, + "learning_rate": 3.5780828937177126e-05, + "loss": 0.8993, + "step": 16230 + }, + { + "epoch": 2.889779202279202, + "grad_norm": 0.8101344108581543, + "learning_rate": 3.577009988708737e-05, + "loss": 1.0002, + "step": 16231 + }, + { + "epoch": 2.889957264957265, + "grad_norm": 0.8259516358375549, + "learning_rate": 3.5759372095449085e-05, + "loss": 0.9223, + "step": 16232 + }, + { + "epoch": 2.8901353276353277, + "grad_norm": 0.8366032242774963, + "learning_rate": 3.574864556247246e-05, + "loss": 0.9635, + "step": 16233 + }, + { + "epoch": 2.8903133903133904, + "grad_norm": 0.8667864203453064, + "learning_rate": 3.573792028836764e-05, + "loss": 0.9923, + "step": 16234 + }, + { + "epoch": 2.890491452991453, + "grad_norm": 0.8982881903648376, + "learning_rate": 3.5727196273344784e-05, + "loss": 0.8735, + "step": 16235 + }, + { + "epoch": 2.890669515669516, + "grad_norm": 0.8208550214767456, + "learning_rate": 3.571647351761398e-05, + "loss": 0.809, + "step": 16236 + }, + { + "epoch": 2.890847578347578, + "grad_norm": 1.0483866930007935, + "learning_rate": 3.5705752021385395e-05, + "loss": 0.9106, + "step": 16237 + }, + { + "epoch": 2.891025641025641, + "grad_norm": 0.8783283829689026, + "learning_rate": 3.5695031784868984e-05, + "loss": 0.8061, + "step": 16238 + }, + { + "epoch": 2.8912037037037037, + "grad_norm": 0.7530997395515442, + "learning_rate": 3.56843128082749e-05, + "loss": 0.58, + "step": 16239 + }, + { + "epoch": 2.8913817663817665, + "grad_norm": 1.0263420343399048, + "learning_rate": 3.567359509181304e-05, + "loss": 0.9946, + "step": 16240 + }, + { + "epoch": 2.8915598290598292, + "grad_norm": 0.7237655520439148, + "learning_rate": 3.5662878635693484e-05, + "loss": 0.6941, + "step": 16241 + }, + { + "epoch": 2.8917378917378915, + "grad_norm": 0.8415992856025696, + "learning_rate": 3.565216344012618e-05, + "loss": 0.7908, + "step": 16242 + }, + { + "epoch": 2.8919159544159543, + "grad_norm": 0.8285344243049622, + "learning_rate": 3.564144950532107e-05, + "loss": 0.8044, + "step": 16243 + }, + { + "epoch": 2.892094017094017, + "grad_norm": 0.9715448617935181, + "learning_rate": 3.5630736831488046e-05, + "loss": 0.8236, + "step": 16244 + }, + { + "epoch": 2.89227207977208, + "grad_norm": 0.8461307287216187, + "learning_rate": 3.5620025418836985e-05, + "loss": 0.9222, + "step": 16245 + }, + { + "epoch": 2.8924501424501425, + "grad_norm": 0.7191013097763062, + "learning_rate": 3.5609315267577836e-05, + "loss": 0.6089, + "step": 16246 + }, + { + "epoch": 2.8926282051282053, + "grad_norm": 0.8537501692771912, + "learning_rate": 3.559860637792038e-05, + "loss": 0.8771, + "step": 16247 + }, + { + "epoch": 2.892806267806268, + "grad_norm": 0.8684942126274109, + "learning_rate": 3.558789875007447e-05, + "loss": 0.9009, + "step": 16248 + }, + { + "epoch": 2.8929843304843303, + "grad_norm": 1.0619043111801147, + "learning_rate": 3.5577192384249856e-05, + "loss": 0.7283, + "step": 16249 + }, + { + "epoch": 2.893162393162393, + "grad_norm": 0.8889201879501343, + "learning_rate": 3.556648728065635e-05, + "loss": 0.7538, + "step": 16250 + }, + { + "epoch": 2.893340455840456, + "grad_norm": 0.8162542581558228, + "learning_rate": 3.555578343950367e-05, + "loss": 0.8338, + "step": 16251 + }, + { + "epoch": 2.8935185185185186, + "grad_norm": 0.8750036954879761, + "learning_rate": 3.5545080861001535e-05, + "loss": 0.6956, + "step": 16252 + }, + { + "epoch": 2.8936965811965814, + "grad_norm": 0.911232054233551, + "learning_rate": 3.553437954535962e-05, + "loss": 0.7559, + "step": 16253 + }, + { + "epoch": 2.8938746438746437, + "grad_norm": 0.889566957950592, + "learning_rate": 3.5523679492787685e-05, + "loss": 0.969, + "step": 16254 + }, + { + "epoch": 2.8940527065527064, + "grad_norm": 0.933595597743988, + "learning_rate": 3.551298070349525e-05, + "loss": 0.9767, + "step": 16255 + }, + { + "epoch": 2.894230769230769, + "grad_norm": 0.8633596897125244, + "learning_rate": 3.550228317769203e-05, + "loss": 0.7823, + "step": 16256 + }, + { + "epoch": 2.894408831908832, + "grad_norm": 0.8595561981201172, + "learning_rate": 3.5491586915587585e-05, + "loss": 0.9583, + "step": 16257 + }, + { + "epoch": 2.8945868945868947, + "grad_norm": 0.7595796585083008, + "learning_rate": 3.548089191739149e-05, + "loss": 0.7167, + "step": 16258 + }, + { + "epoch": 2.8947649572649574, + "grad_norm": 0.8662711381912231, + "learning_rate": 3.54701981833133e-05, + "loss": 0.9439, + "step": 16259 + }, + { + "epoch": 2.89494301994302, + "grad_norm": 0.8028330206871033, + "learning_rate": 3.5459505713562525e-05, + "loss": 0.7442, + "step": 16260 + }, + { + "epoch": 2.8951210826210825, + "grad_norm": 0.9413794279098511, + "learning_rate": 3.5448814508348616e-05, + "loss": 0.97, + "step": 16261 + }, + { + "epoch": 2.8952991452991452, + "grad_norm": 0.9300761222839355, + "learning_rate": 3.543812456788118e-05, + "loss": 0.9294, + "step": 16262 + }, + { + "epoch": 2.895477207977208, + "grad_norm": 0.7364256381988525, + "learning_rate": 3.54274358923695e-05, + "loss": 0.5352, + "step": 16263 + }, + { + "epoch": 2.8956552706552707, + "grad_norm": 0.8452964425086975, + "learning_rate": 3.541674848202314e-05, + "loss": 0.7362, + "step": 16264 + }, + { + "epoch": 2.8958333333333335, + "grad_norm": 0.8745927214622498, + "learning_rate": 3.540606233705137e-05, + "loss": 1.0222, + "step": 16265 + }, + { + "epoch": 2.896011396011396, + "grad_norm": 0.9202282428741455, + "learning_rate": 3.539537745766367e-05, + "loss": 1.0853, + "step": 16266 + }, + { + "epoch": 2.8961894586894585, + "grad_norm": 0.8450053930282593, + "learning_rate": 3.538469384406933e-05, + "loss": 0.9295, + "step": 16267 + }, + { + "epoch": 2.8963675213675213, + "grad_norm": 0.8761671781539917, + "learning_rate": 3.5374011496477656e-05, + "loss": 0.9273, + "step": 16268 + }, + { + "epoch": 2.896545584045584, + "grad_norm": 0.8562198281288147, + "learning_rate": 3.536333041509805e-05, + "loss": 0.7654, + "step": 16269 + }, + { + "epoch": 2.896723646723647, + "grad_norm": 0.9013510942459106, + "learning_rate": 3.535265060013965e-05, + "loss": 0.7432, + "step": 16270 + }, + { + "epoch": 2.8969017094017095, + "grad_norm": 1.1163274049758911, + "learning_rate": 3.534197205181179e-05, + "loss": 1.1223, + "step": 16271 + }, + { + "epoch": 2.8970797720797723, + "grad_norm": 0.869686484336853, + "learning_rate": 3.5331294770323674e-05, + "loss": 0.9255, + "step": 16272 + }, + { + "epoch": 2.8972578347578346, + "grad_norm": 0.8521125316619873, + "learning_rate": 3.53206187558845e-05, + "loss": 1.0221, + "step": 16273 + }, + { + "epoch": 2.8974358974358974, + "grad_norm": 0.9807026982307434, + "learning_rate": 3.530994400870345e-05, + "loss": 1.0133, + "step": 16274 + }, + { + "epoch": 2.89761396011396, + "grad_norm": 0.9236428141593933, + "learning_rate": 3.529927052898967e-05, + "loss": 0.8711, + "step": 16275 + }, + { + "epoch": 2.897792022792023, + "grad_norm": 0.8108885884284973, + "learning_rate": 3.528859831695227e-05, + "loss": 0.917, + "step": 16276 + }, + { + "epoch": 2.8979700854700856, + "grad_norm": 0.7522720098495483, + "learning_rate": 3.527792737280036e-05, + "loss": 0.6776, + "step": 16277 + }, + { + "epoch": 2.898148148148148, + "grad_norm": 0.9261712431907654, + "learning_rate": 3.526725769674297e-05, + "loss": 1.0728, + "step": 16278 + }, + { + "epoch": 2.8983262108262107, + "grad_norm": 0.849559485912323, + "learning_rate": 3.5256589288989285e-05, + "loss": 0.9886, + "step": 16279 + }, + { + "epoch": 2.8985042735042734, + "grad_norm": 0.9024273157119751, + "learning_rate": 3.5245922149748155e-05, + "loss": 0.9296, + "step": 16280 + }, + { + "epoch": 2.898682336182336, + "grad_norm": 0.8285173773765564, + "learning_rate": 3.52352562792287e-05, + "loss": 0.8776, + "step": 16281 + }, + { + "epoch": 2.898860398860399, + "grad_norm": 0.895517885684967, + "learning_rate": 3.522459167763987e-05, + "loss": 0.8975, + "step": 16282 + }, + { + "epoch": 2.8990384615384617, + "grad_norm": 0.8449265956878662, + "learning_rate": 3.521392834519061e-05, + "loss": 0.6736, + "step": 16283 + }, + { + "epoch": 2.8992165242165244, + "grad_norm": 0.8652997612953186, + "learning_rate": 3.520326628208983e-05, + "loss": 0.7834, + "step": 16284 + }, + { + "epoch": 2.8993945868945867, + "grad_norm": 0.9012393951416016, + "learning_rate": 3.519260548854642e-05, + "loss": 0.8082, + "step": 16285 + }, + { + "epoch": 2.8995726495726495, + "grad_norm": 0.9048463106155396, + "learning_rate": 3.5181945964769333e-05, + "loss": 0.8078, + "step": 16286 + }, + { + "epoch": 2.8997507122507122, + "grad_norm": 0.8788473010063171, + "learning_rate": 3.5171287710967314e-05, + "loss": 0.8022, + "step": 16287 + }, + { + "epoch": 2.899928774928775, + "grad_norm": 0.8322813510894775, + "learning_rate": 3.516063072734928e-05, + "loss": 0.8855, + "step": 16288 + }, + { + "epoch": 2.9001068376068377, + "grad_norm": 0.8762373328208923, + "learning_rate": 3.514997501412398e-05, + "loss": 0.7858, + "step": 16289 + }, + { + "epoch": 2.9002849002849, + "grad_norm": 0.7718746066093445, + "learning_rate": 3.513932057150021e-05, + "loss": 0.6881, + "step": 16290 + }, + { + "epoch": 2.900462962962963, + "grad_norm": 1.0138204097747803, + "learning_rate": 3.5128667399686724e-05, + "loss": 0.9378, + "step": 16291 + }, + { + "epoch": 2.9006410256410255, + "grad_norm": 0.6968120336532593, + "learning_rate": 3.5118015498892234e-05, + "loss": 0.7135, + "step": 16292 + }, + { + "epoch": 2.9008190883190883, + "grad_norm": 0.7925532460212708, + "learning_rate": 3.510736486932542e-05, + "loss": 0.7414, + "step": 16293 + }, + { + "epoch": 2.900997150997151, + "grad_norm": 1.0432425737380981, + "learning_rate": 3.5096715511195056e-05, + "loss": 0.8957, + "step": 16294 + }, + { + "epoch": 2.901175213675214, + "grad_norm": 0.8664390444755554, + "learning_rate": 3.508606742470966e-05, + "loss": 0.7832, + "step": 16295 + }, + { + "epoch": 2.9013532763532766, + "grad_norm": 0.8470353484153748, + "learning_rate": 3.507542061007795e-05, + "loss": 0.8133, + "step": 16296 + }, + { + "epoch": 2.9015313390313393, + "grad_norm": 0.8339848518371582, + "learning_rate": 3.5064775067508514e-05, + "loss": 0.9275, + "step": 16297 + }, + { + "epoch": 2.9017094017094016, + "grad_norm": 0.8686776757240295, + "learning_rate": 3.5054130797209916e-05, + "loss": 0.9238, + "step": 16298 + }, + { + "epoch": 2.9018874643874644, + "grad_norm": 0.8178901076316833, + "learning_rate": 3.504348779939071e-05, + "loss": 1.0776, + "step": 16299 + }, + { + "epoch": 2.902065527065527, + "grad_norm": 0.8446379899978638, + "learning_rate": 3.5032846074259426e-05, + "loss": 0.715, + "step": 16300 + }, + { + "epoch": 2.90224358974359, + "grad_norm": 0.8320762515068054, + "learning_rate": 3.502220562202457e-05, + "loss": 0.8543, + "step": 16301 + }, + { + "epoch": 2.902421652421652, + "grad_norm": 0.9112939238548279, + "learning_rate": 3.501156644289462e-05, + "loss": 0.8239, + "step": 16302 + }, + { + "epoch": 2.902599715099715, + "grad_norm": 0.8749213218688965, + "learning_rate": 3.500092853707797e-05, + "loss": 0.8057, + "step": 16303 + }, + { + "epoch": 2.9027777777777777, + "grad_norm": 0.8245106339454651, + "learning_rate": 3.4990291904783143e-05, + "loss": 0.8955, + "step": 16304 + }, + { + "epoch": 2.9029558404558404, + "grad_norm": 0.7982145547866821, + "learning_rate": 3.4979656546218506e-05, + "loss": 0.6641, + "step": 16305 + }, + { + "epoch": 2.903133903133903, + "grad_norm": 0.8777986168861389, + "learning_rate": 3.496902246159244e-05, + "loss": 0.7365, + "step": 16306 + }, + { + "epoch": 2.903311965811966, + "grad_norm": 0.8463431596755981, + "learning_rate": 3.4958389651113275e-05, + "loss": 1.0434, + "step": 16307 + }, + { + "epoch": 2.9034900284900287, + "grad_norm": 0.849039614200592, + "learning_rate": 3.494775811498931e-05, + "loss": 0.8813, + "step": 16308 + }, + { + "epoch": 2.9036680911680914, + "grad_norm": 0.7352656126022339, + "learning_rate": 3.4937127853428976e-05, + "loss": 0.7936, + "step": 16309 + }, + { + "epoch": 2.9038461538461537, + "grad_norm": 0.764543354511261, + "learning_rate": 3.49264988666404e-05, + "loss": 0.7253, + "step": 16310 + }, + { + "epoch": 2.9040242165242165, + "grad_norm": 0.7932603359222412, + "learning_rate": 3.491587115483196e-05, + "loss": 0.9217, + "step": 16311 + }, + { + "epoch": 2.9042022792022792, + "grad_norm": 0.9001819491386414, + "learning_rate": 3.490524471821175e-05, + "loss": 0.8628, + "step": 16312 + }, + { + "epoch": 2.904380341880342, + "grad_norm": 0.7983253002166748, + "learning_rate": 3.4894619556988085e-05, + "loss": 0.9858, + "step": 16313 + }, + { + "epoch": 2.9045584045584043, + "grad_norm": 0.747611403465271, + "learning_rate": 3.488399567136911e-05, + "loss": 0.8656, + "step": 16314 + }, + { + "epoch": 2.904736467236467, + "grad_norm": 0.8770463466644287, + "learning_rate": 3.487337306156296e-05, + "loss": 1.0154, + "step": 16315 + }, + { + "epoch": 2.90491452991453, + "grad_norm": 0.8757193088531494, + "learning_rate": 3.4862751727777797e-05, + "loss": 0.7931, + "step": 16316 + }, + { + "epoch": 2.9050925925925926, + "grad_norm": 0.8232926726341248, + "learning_rate": 3.485213167022169e-05, + "loss": 0.8477, + "step": 16317 + }, + { + "epoch": 2.9052706552706553, + "grad_norm": 0.8445250391960144, + "learning_rate": 3.48415128891027e-05, + "loss": 1.0618, + "step": 16318 + }, + { + "epoch": 2.905448717948718, + "grad_norm": 0.8172810673713684, + "learning_rate": 3.483089538462897e-05, + "loss": 0.7882, + "step": 16319 + }, + { + "epoch": 2.905626780626781, + "grad_norm": 0.910757303237915, + "learning_rate": 3.4820279157008404e-05, + "loss": 0.9383, + "step": 16320 + }, + { + "epoch": 2.9058048433048436, + "grad_norm": 0.8837474584579468, + "learning_rate": 3.480966420644911e-05, + "loss": 1.1006, + "step": 16321 + }, + { + "epoch": 2.905982905982906, + "grad_norm": 0.7739782333374023, + "learning_rate": 3.4799050533159014e-05, + "loss": 0.6885, + "step": 16322 + }, + { + "epoch": 2.9061609686609686, + "grad_norm": 0.911738932132721, + "learning_rate": 3.478843813734609e-05, + "loss": 0.993, + "step": 16323 + }, + { + "epoch": 2.9063390313390314, + "grad_norm": 0.8834345936775208, + "learning_rate": 3.477782701921825e-05, + "loss": 0.8724, + "step": 16324 + }, + { + "epoch": 2.906517094017094, + "grad_norm": 0.8103434443473816, + "learning_rate": 3.476721717898337e-05, + "loss": 1.0517, + "step": 16325 + }, + { + "epoch": 2.9066951566951564, + "grad_norm": 0.8648924827575684, + "learning_rate": 3.475660861684943e-05, + "loss": 0.7223, + "step": 16326 + }, + { + "epoch": 2.906873219373219, + "grad_norm": 0.8762979507446289, + "learning_rate": 3.4746001333024134e-05, + "loss": 0.8421, + "step": 16327 + }, + { + "epoch": 2.907051282051282, + "grad_norm": 0.9596083164215088, + "learning_rate": 3.4735395327715434e-05, + "loss": 1.0206, + "step": 16328 + }, + { + "epoch": 2.9072293447293447, + "grad_norm": 0.8210311532020569, + "learning_rate": 3.472479060113107e-05, + "loss": 0.836, + "step": 16329 + }, + { + "epoch": 2.9074074074074074, + "grad_norm": 0.8436611890792847, + "learning_rate": 3.471418715347886e-05, + "loss": 0.8459, + "step": 16330 + }, + { + "epoch": 2.90758547008547, + "grad_norm": 0.9176212549209595, + "learning_rate": 3.470358498496652e-05, + "loss": 0.8227, + "step": 16331 + }, + { + "epoch": 2.907763532763533, + "grad_norm": 0.7709631323814392, + "learning_rate": 3.4692984095801796e-05, + "loss": 0.8096, + "step": 16332 + }, + { + "epoch": 2.9079415954415957, + "grad_norm": 0.7727495431900024, + "learning_rate": 3.4682384486192346e-05, + "loss": 0.6843, + "step": 16333 + }, + { + "epoch": 2.908119658119658, + "grad_norm": 0.9743461608886719, + "learning_rate": 3.4671786156345955e-05, + "loss": 0.7542, + "step": 16334 + }, + { + "epoch": 2.9082977207977208, + "grad_norm": 0.9035171270370483, + "learning_rate": 3.466118910647014e-05, + "loss": 0.8827, + "step": 16335 + }, + { + "epoch": 2.9084757834757835, + "grad_norm": 1.1434134244918823, + "learning_rate": 3.465059333677266e-05, + "loss": 0.8026, + "step": 16336 + }, + { + "epoch": 2.9086538461538463, + "grad_norm": 0.8229905962944031, + "learning_rate": 3.4639998847461e-05, + "loss": 0.7415, + "step": 16337 + }, + { + "epoch": 2.9088319088319086, + "grad_norm": 0.9193732738494873, + "learning_rate": 3.462940563874281e-05, + "loss": 0.9773, + "step": 16338 + }, + { + "epoch": 2.9090099715099713, + "grad_norm": 0.8461189270019531, + "learning_rate": 3.4618813710825614e-05, + "loss": 0.9169, + "step": 16339 + }, + { + "epoch": 2.909188034188034, + "grad_norm": 0.9471017718315125, + "learning_rate": 3.460822306391696e-05, + "loss": 0.9941, + "step": 16340 + }, + { + "epoch": 2.909366096866097, + "grad_norm": 0.8515542149543762, + "learning_rate": 3.459763369822432e-05, + "loss": 0.8209, + "step": 16341 + }, + { + "epoch": 2.9095441595441596, + "grad_norm": 0.8520704507827759, + "learning_rate": 3.458704561395519e-05, + "loss": 0.8443, + "step": 16342 + }, + { + "epoch": 2.9097222222222223, + "grad_norm": 0.9236885905265808, + "learning_rate": 3.457645881131699e-05, + "loss": 0.8407, + "step": 16343 + }, + { + "epoch": 2.909900284900285, + "grad_norm": 0.9255889654159546, + "learning_rate": 3.4565873290517203e-05, + "loss": 0.9754, + "step": 16344 + }, + { + "epoch": 2.910078347578348, + "grad_norm": 0.7904002666473389, + "learning_rate": 3.455528905176321e-05, + "loss": 0.7123, + "step": 16345 + }, + { + "epoch": 2.91025641025641, + "grad_norm": 0.821877658367157, + "learning_rate": 3.454470609526237e-05, + "loss": 0.7024, + "step": 16346 + }, + { + "epoch": 2.910434472934473, + "grad_norm": 0.84690260887146, + "learning_rate": 3.453412442122205e-05, + "loss": 0.7551, + "step": 16347 + }, + { + "epoch": 2.9106125356125356, + "grad_norm": 1.0308923721313477, + "learning_rate": 3.452354402984955e-05, + "loss": 1.0316, + "step": 16348 + }, + { + "epoch": 2.9107905982905984, + "grad_norm": 0.7016192674636841, + "learning_rate": 3.451296492135221e-05, + "loss": 0.6816, + "step": 16349 + }, + { + "epoch": 2.9109686609686607, + "grad_norm": 0.942915141582489, + "learning_rate": 3.4502387095937237e-05, + "loss": 0.9029, + "step": 16350 + }, + { + "epoch": 2.9111467236467234, + "grad_norm": 0.7511810660362244, + "learning_rate": 3.449181055381201e-05, + "loss": 0.641, + "step": 16351 + }, + { + "epoch": 2.911324786324786, + "grad_norm": 0.8904419541358948, + "learning_rate": 3.44812352951836e-05, + "loss": 0.8176, + "step": 16352 + }, + { + "epoch": 2.911502849002849, + "grad_norm": 0.7585623264312744, + "learning_rate": 3.447066132025931e-05, + "loss": 0.7815, + "step": 16353 + }, + { + "epoch": 2.9116809116809117, + "grad_norm": 0.8587543964385986, + "learning_rate": 3.446008862924629e-05, + "loss": 0.951, + "step": 16354 + }, + { + "epoch": 2.9118589743589745, + "grad_norm": 0.7638232707977295, + "learning_rate": 3.444951722235169e-05, + "loss": 0.6784, + "step": 16355 + }, + { + "epoch": 2.912037037037037, + "grad_norm": 0.7942266464233398, + "learning_rate": 3.4438947099782624e-05, + "loss": 0.7211, + "step": 16356 + }, + { + "epoch": 2.9122150997151, + "grad_norm": 0.7207983732223511, + "learning_rate": 3.4428378261746195e-05, + "loss": 0.7515, + "step": 16357 + }, + { + "epoch": 2.9123931623931623, + "grad_norm": 0.8683337569236755, + "learning_rate": 3.4417810708449495e-05, + "loss": 0.8724, + "step": 16358 + }, + { + "epoch": 2.912571225071225, + "grad_norm": 0.8052859902381897, + "learning_rate": 3.440724444009955e-05, + "loss": 0.8331, + "step": 16359 + }, + { + "epoch": 2.9127492877492878, + "grad_norm": 0.7425459027290344, + "learning_rate": 3.439667945690336e-05, + "loss": 0.5711, + "step": 16360 + }, + { + "epoch": 2.9129273504273505, + "grad_norm": 0.8537404537200928, + "learning_rate": 3.438611575906803e-05, + "loss": 0.8283, + "step": 16361 + }, + { + "epoch": 2.9131054131054133, + "grad_norm": 0.794684648513794, + "learning_rate": 3.437555334680038e-05, + "loss": 0.679, + "step": 16362 + }, + { + "epoch": 2.9132834757834756, + "grad_norm": 0.8330501914024353, + "learning_rate": 3.436499222030748e-05, + "loss": 0.6892, + "step": 16363 + }, + { + "epoch": 2.9134615384615383, + "grad_norm": 0.9036495089530945, + "learning_rate": 3.435443237979621e-05, + "loss": 0.8349, + "step": 16364 + }, + { + "epoch": 2.913639601139601, + "grad_norm": 0.776745080947876, + "learning_rate": 3.434387382547344e-05, + "loss": 0.6691, + "step": 16365 + }, + { + "epoch": 2.913817663817664, + "grad_norm": 0.8921083807945251, + "learning_rate": 3.4333316557546145e-05, + "loss": 0.9497, + "step": 16366 + }, + { + "epoch": 2.9139957264957266, + "grad_norm": 0.8435728549957275, + "learning_rate": 3.4322760576221023e-05, + "loss": 0.787, + "step": 16367 + }, + { + "epoch": 2.9141737891737893, + "grad_norm": 0.7281554341316223, + "learning_rate": 3.4312205881705015e-05, + "loss": 0.7604, + "step": 16368 + }, + { + "epoch": 2.914351851851852, + "grad_norm": 0.8474677801132202, + "learning_rate": 3.430165247420488e-05, + "loss": 0.9052, + "step": 16369 + }, + { + "epoch": 2.9145299145299144, + "grad_norm": 0.9927265048027039, + "learning_rate": 3.42911003539274e-05, + "loss": 0.7486, + "step": 16370 + }, + { + "epoch": 2.914707977207977, + "grad_norm": 0.7754966020584106, + "learning_rate": 3.4280549521079286e-05, + "loss": 0.8368, + "step": 16371 + }, + { + "epoch": 2.91488603988604, + "grad_norm": 0.7774887084960938, + "learning_rate": 3.4269999975867295e-05, + "loss": 0.7014, + "step": 16372 + }, + { + "epoch": 2.9150641025641026, + "grad_norm": 0.9286267161369324, + "learning_rate": 3.4259451718498115e-05, + "loss": 1.0414, + "step": 16373 + }, + { + "epoch": 2.9152421652421654, + "grad_norm": 0.8269815444946289, + "learning_rate": 3.424890474917841e-05, + "loss": 0.9731, + "step": 16374 + }, + { + "epoch": 2.9154202279202277, + "grad_norm": 0.8319926857948303, + "learning_rate": 3.42383590681148e-05, + "loss": 0.9047, + "step": 16375 + }, + { + "epoch": 2.9155982905982905, + "grad_norm": 0.8976882696151733, + "learning_rate": 3.4227814675514e-05, + "loss": 0.8608, + "step": 16376 + }, + { + "epoch": 2.915776353276353, + "grad_norm": 0.80831378698349, + "learning_rate": 3.421727157158248e-05, + "loss": 0.9213, + "step": 16377 + }, + { + "epoch": 2.915954415954416, + "grad_norm": 0.881201982498169, + "learning_rate": 3.42067297565269e-05, + "loss": 0.9015, + "step": 16378 + }, + { + "epoch": 2.9161324786324787, + "grad_norm": 0.7797132134437561, + "learning_rate": 3.4196189230553775e-05, + "loss": 0.9002, + "step": 16379 + }, + { + "epoch": 2.9163105413105415, + "grad_norm": 1.0163915157318115, + "learning_rate": 3.4185649993869626e-05, + "loss": 1.0185, + "step": 16380 + }, + { + "epoch": 2.916488603988604, + "grad_norm": 0.8868339657783508, + "learning_rate": 3.417511204668096e-05, + "loss": 0.7316, + "step": 16381 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 0.8215135931968689, + "learning_rate": 3.416457538919422e-05, + "loss": 0.8693, + "step": 16382 + }, + { + "epoch": 2.9168447293447293, + "grad_norm": 0.8070623278617859, + "learning_rate": 3.4154040021615876e-05, + "loss": 0.8532, + "step": 16383 + }, + { + "epoch": 2.917022792022792, + "grad_norm": 0.8857349157333374, + "learning_rate": 3.4143505944152284e-05, + "loss": 0.961, + "step": 16384 + }, + { + "epoch": 2.9172008547008548, + "grad_norm": 0.8146688342094421, + "learning_rate": 3.4132973157009936e-05, + "loss": 0.8597, + "step": 16385 + }, + { + "epoch": 2.9173789173789175, + "grad_norm": 0.8757439851760864, + "learning_rate": 3.4122441660395156e-05, + "loss": 0.7683, + "step": 16386 + }, + { + "epoch": 2.91755698005698, + "grad_norm": 0.8319665789604187, + "learning_rate": 3.4111911454514266e-05, + "loss": 0.8034, + "step": 16387 + }, + { + "epoch": 2.9177350427350426, + "grad_norm": 0.8103782534599304, + "learning_rate": 3.410138253957361e-05, + "loss": 0.8469, + "step": 16388 + }, + { + "epoch": 2.9179131054131053, + "grad_norm": 0.8783053755760193, + "learning_rate": 3.4090854915779466e-05, + "loss": 0.7997, + "step": 16389 + }, + { + "epoch": 2.918091168091168, + "grad_norm": 0.9793184995651245, + "learning_rate": 3.408032858333808e-05, + "loss": 0.8934, + "step": 16390 + }, + { + "epoch": 2.918269230769231, + "grad_norm": 0.8603301048278809, + "learning_rate": 3.406980354245578e-05, + "loss": 0.847, + "step": 16391 + }, + { + "epoch": 2.9184472934472936, + "grad_norm": 0.9763472080230713, + "learning_rate": 3.405927979333866e-05, + "loss": 0.9491, + "step": 16392 + }, + { + "epoch": 2.9186253561253563, + "grad_norm": 0.9340085983276367, + "learning_rate": 3.4048757336193036e-05, + "loss": 0.8913, + "step": 16393 + }, + { + "epoch": 2.9188034188034186, + "grad_norm": 1.051066517829895, + "learning_rate": 3.4038236171224946e-05, + "loss": 0.9903, + "step": 16394 + }, + { + "epoch": 2.9189814814814814, + "grad_norm": 0.8462432026863098, + "learning_rate": 3.402771629864062e-05, + "loss": 0.868, + "step": 16395 + }, + { + "epoch": 2.919159544159544, + "grad_norm": 0.9149147868156433, + "learning_rate": 3.401719771864615e-05, + "loss": 0.9215, + "step": 16396 + }, + { + "epoch": 2.919337606837607, + "grad_norm": 0.7974846959114075, + "learning_rate": 3.4006680431447636e-05, + "loss": 0.9577, + "step": 16397 + }, + { + "epoch": 2.9195156695156697, + "grad_norm": 0.9298149943351746, + "learning_rate": 3.399616443725112e-05, + "loss": 0.8977, + "step": 16398 + }, + { + "epoch": 2.919693732193732, + "grad_norm": 0.8799013495445251, + "learning_rate": 3.398564973626265e-05, + "loss": 1.0161, + "step": 16399 + }, + { + "epoch": 2.9198717948717947, + "grad_norm": 0.8730150461196899, + "learning_rate": 3.3975136328688206e-05, + "loss": 0.7953, + "step": 16400 + }, + { + "epoch": 2.9200498575498575, + "grad_norm": 1.0036572217941284, + "learning_rate": 3.396462421473387e-05, + "loss": 0.5941, + "step": 16401 + }, + { + "epoch": 2.92022792022792, + "grad_norm": 0.8903291821479797, + "learning_rate": 3.39541133946055e-05, + "loss": 0.7715, + "step": 16402 + }, + { + "epoch": 2.920405982905983, + "grad_norm": 0.8254680633544922, + "learning_rate": 3.39436038685091e-05, + "loss": 0.7268, + "step": 16403 + }, + { + "epoch": 2.9205840455840457, + "grad_norm": 0.8490110039710999, + "learning_rate": 3.393309563665057e-05, + "loss": 0.8684, + "step": 16404 + }, + { + "epoch": 2.9207621082621085, + "grad_norm": 0.8137654066085815, + "learning_rate": 3.392258869923575e-05, + "loss": 0.7068, + "step": 16405 + }, + { + "epoch": 2.9209401709401708, + "grad_norm": 0.8084680438041687, + "learning_rate": 3.39120830564706e-05, + "loss": 0.8214, + "step": 16406 + }, + { + "epoch": 2.9211182336182335, + "grad_norm": 0.9359310269355774, + "learning_rate": 3.3901578708560835e-05, + "loss": 1.0247, + "step": 16407 + }, + { + "epoch": 2.9212962962962963, + "grad_norm": 0.826842725276947, + "learning_rate": 3.38910756557124e-05, + "loss": 0.8018, + "step": 16408 + }, + { + "epoch": 2.921474358974359, + "grad_norm": 0.9049538373947144, + "learning_rate": 3.388057389813093e-05, + "loss": 0.8825, + "step": 16409 + }, + { + "epoch": 2.921652421652422, + "grad_norm": 0.7966105937957764, + "learning_rate": 3.387007343602231e-05, + "loss": 0.8716, + "step": 16410 + }, + { + "epoch": 2.921830484330484, + "grad_norm": 0.8647517561912537, + "learning_rate": 3.385957426959222e-05, + "loss": 0.9947, + "step": 16411 + }, + { + "epoch": 2.922008547008547, + "grad_norm": 0.8983362317085266, + "learning_rate": 3.384907639904638e-05, + "loss": 0.8701, + "step": 16412 + }, + { + "epoch": 2.9221866096866096, + "grad_norm": 0.7970262765884399, + "learning_rate": 3.3838579824590465e-05, + "loss": 0.8921, + "step": 16413 + }, + { + "epoch": 2.9223646723646723, + "grad_norm": 0.8115224242210388, + "learning_rate": 3.382808454643015e-05, + "loss": 0.5948, + "step": 16414 + }, + { + "epoch": 2.922542735042735, + "grad_norm": 0.832318902015686, + "learning_rate": 3.381759056477102e-05, + "loss": 0.8399, + "step": 16415 + }, + { + "epoch": 2.922720797720798, + "grad_norm": 0.9396497011184692, + "learning_rate": 3.380709787981878e-05, + "loss": 0.7322, + "step": 16416 + }, + { + "epoch": 2.9228988603988606, + "grad_norm": 0.7956545352935791, + "learning_rate": 3.3796606491778904e-05, + "loss": 0.8566, + "step": 16417 + }, + { + "epoch": 2.9230769230769234, + "grad_norm": 0.8257092833518982, + "learning_rate": 3.378611640085705e-05, + "loss": 0.8682, + "step": 16418 + }, + { + "epoch": 2.9232549857549857, + "grad_norm": 0.7565430402755737, + "learning_rate": 3.377562760725863e-05, + "loss": 0.8513, + "step": 16419 + }, + { + "epoch": 2.9234330484330484, + "grad_norm": 0.769944965839386, + "learning_rate": 3.3765140111189265e-05, + "loss": 0.8869, + "step": 16420 + }, + { + "epoch": 2.923611111111111, + "grad_norm": 0.8117483854293823, + "learning_rate": 3.375465391285438e-05, + "loss": 0.8718, + "step": 16421 + }, + { + "epoch": 2.923789173789174, + "grad_norm": 0.8011773824691772, + "learning_rate": 3.374416901245944e-05, + "loss": 0.9326, + "step": 16422 + }, + { + "epoch": 2.923967236467236, + "grad_norm": 0.8096779584884644, + "learning_rate": 3.373368541020987e-05, + "loss": 0.9317, + "step": 16423 + }, + { + "epoch": 2.924145299145299, + "grad_norm": 0.8383152484893799, + "learning_rate": 3.3723203106311055e-05, + "loss": 0.9399, + "step": 16424 + }, + { + "epoch": 2.9243233618233617, + "grad_norm": 0.9268670678138733, + "learning_rate": 3.371272210096842e-05, + "loss": 0.6867, + "step": 16425 + }, + { + "epoch": 2.9245014245014245, + "grad_norm": 0.941338837146759, + "learning_rate": 3.3702242394387294e-05, + "loss": 0.9164, + "step": 16426 + }, + { + "epoch": 2.9246794871794872, + "grad_norm": 0.7108882665634155, + "learning_rate": 3.3691763986773014e-05, + "loss": 0.6956, + "step": 16427 + }, + { + "epoch": 2.92485754985755, + "grad_norm": 0.9144331812858582, + "learning_rate": 3.3681286878330876e-05, + "loss": 0.8148, + "step": 16428 + }, + { + "epoch": 2.9250356125356127, + "grad_norm": 0.882851243019104, + "learning_rate": 3.367081106926615e-05, + "loss": 0.846, + "step": 16429 + }, + { + "epoch": 2.9252136752136755, + "grad_norm": 0.9508523941040039, + "learning_rate": 3.366033655978409e-05, + "loss": 0.9637, + "step": 16430 + }, + { + "epoch": 2.925391737891738, + "grad_norm": 0.7770804166793823, + "learning_rate": 3.3649863350089935e-05, + "loss": 0.6673, + "step": 16431 + }, + { + "epoch": 2.9255698005698005, + "grad_norm": 0.8342770934104919, + "learning_rate": 3.3639391440388845e-05, + "loss": 1.0366, + "step": 16432 + }, + { + "epoch": 2.9257478632478633, + "grad_norm": 0.8854461908340454, + "learning_rate": 3.362892083088609e-05, + "loss": 0.7885, + "step": 16433 + }, + { + "epoch": 2.925925925925926, + "grad_norm": 0.8054807186126709, + "learning_rate": 3.36184515217867e-05, + "loss": 0.7705, + "step": 16434 + }, + { + "epoch": 2.9261039886039883, + "grad_norm": 0.9037294983863831, + "learning_rate": 3.360798351329587e-05, + "loss": 1.077, + "step": 16435 + }, + { + "epoch": 2.926282051282051, + "grad_norm": 0.8743478059768677, + "learning_rate": 3.35975168056187e-05, + "loss": 0.8633, + "step": 16436 + }, + { + "epoch": 2.926460113960114, + "grad_norm": 0.8879369497299194, + "learning_rate": 3.3587051398960245e-05, + "loss": 0.8059, + "step": 16437 + }, + { + "epoch": 2.9266381766381766, + "grad_norm": 0.9445768594741821, + "learning_rate": 3.3576587293525564e-05, + "loss": 0.861, + "step": 16438 + }, + { + "epoch": 2.9268162393162394, + "grad_norm": 0.8131009340286255, + "learning_rate": 3.356612448951967e-05, + "loss": 0.8299, + "step": 16439 + }, + { + "epoch": 2.926994301994302, + "grad_norm": 0.8781847953796387, + "learning_rate": 3.3555662987147515e-05, + "loss": 0.8447, + "step": 16440 + }, + { + "epoch": 2.927172364672365, + "grad_norm": 0.9173591732978821, + "learning_rate": 3.3545202786614206e-05, + "loss": 0.9297, + "step": 16441 + }, + { + "epoch": 2.9273504273504276, + "grad_norm": 0.9596586227416992, + "learning_rate": 3.353474388812452e-05, + "loss": 0.9109, + "step": 16442 + }, + { + "epoch": 2.92752849002849, + "grad_norm": 0.8325486183166504, + "learning_rate": 3.352428629188349e-05, + "loss": 0.6822, + "step": 16443 + }, + { + "epoch": 2.9277065527065527, + "grad_norm": 0.8758354187011719, + "learning_rate": 3.351382999809599e-05, + "loss": 0.9522, + "step": 16444 + }, + { + "epoch": 2.9278846153846154, + "grad_norm": 0.909718930721283, + "learning_rate": 3.3503375006966866e-05, + "loss": 0.8226, + "step": 16445 + }, + { + "epoch": 2.928062678062678, + "grad_norm": 0.9259094595909119, + "learning_rate": 3.3492921318700974e-05, + "loss": 0.9895, + "step": 16446 + }, + { + "epoch": 2.9282407407407405, + "grad_norm": 0.7815300226211548, + "learning_rate": 3.348246893350311e-05, + "loss": 0.8854, + "step": 16447 + }, + { + "epoch": 2.9284188034188032, + "grad_norm": 0.8576910495758057, + "learning_rate": 3.3472017851578154e-05, + "loss": 0.9379, + "step": 16448 + }, + { + "epoch": 2.928596866096866, + "grad_norm": 0.9139176607131958, + "learning_rate": 3.3461568073130735e-05, + "loss": 0.9372, + "step": 16449 + }, + { + "epoch": 2.9287749287749287, + "grad_norm": 0.8304639458656311, + "learning_rate": 3.34511195983657e-05, + "loss": 0.7105, + "step": 16450 + }, + { + "epoch": 2.9289529914529915, + "grad_norm": 0.8689056634902954, + "learning_rate": 3.344067242748774e-05, + "loss": 0.9518, + "step": 16451 + }, + { + "epoch": 2.9291310541310542, + "grad_norm": 0.9180546998977661, + "learning_rate": 3.343022656070154e-05, + "loss": 0.982, + "step": 16452 + }, + { + "epoch": 2.929309116809117, + "grad_norm": 0.9291700124740601, + "learning_rate": 3.341978199821175e-05, + "loss": 1.0003, + "step": 16453 + }, + { + "epoch": 2.9294871794871797, + "grad_norm": 0.970675528049469, + "learning_rate": 3.340933874022304e-05, + "loss": 0.8941, + "step": 16454 + }, + { + "epoch": 2.929665242165242, + "grad_norm": 0.8425672650337219, + "learning_rate": 3.339889678693999e-05, + "loss": 0.7173, + "step": 16455 + }, + { + "epoch": 2.929843304843305, + "grad_norm": 0.9666314721107483, + "learning_rate": 3.3388456138567225e-05, + "loss": 1.0403, + "step": 16456 + }, + { + "epoch": 2.9300213675213675, + "grad_norm": 0.9586226344108582, + "learning_rate": 3.337801679530924e-05, + "loss": 0.7691, + "step": 16457 + }, + { + "epoch": 2.9301994301994303, + "grad_norm": 0.888327419757843, + "learning_rate": 3.33675787573707e-05, + "loss": 0.7019, + "step": 16458 + }, + { + "epoch": 2.9303774928774926, + "grad_norm": 0.7410684823989868, + "learning_rate": 3.335714202495596e-05, + "loss": 0.9935, + "step": 16459 + }, + { + "epoch": 2.9305555555555554, + "grad_norm": 1.0408822298049927, + "learning_rate": 3.3346706598269617e-05, + "loss": 0.9454, + "step": 16460 + }, + { + "epoch": 2.930733618233618, + "grad_norm": 0.9065089821815491, + "learning_rate": 3.333627247751611e-05, + "loss": 0.9665, + "step": 16461 + }, + { + "epoch": 2.930911680911681, + "grad_norm": 0.8904961943626404, + "learning_rate": 3.332583966289985e-05, + "loss": 0.9402, + "step": 16462 + }, + { + "epoch": 2.9310897435897436, + "grad_norm": 0.920364260673523, + "learning_rate": 3.331540815462526e-05, + "loss": 0.8524, + "step": 16463 + }, + { + "epoch": 2.9312678062678064, + "grad_norm": 0.9185073375701904, + "learning_rate": 3.330497795289669e-05, + "loss": 0.8982, + "step": 16464 + }, + { + "epoch": 2.931445868945869, + "grad_norm": 0.9365581274032593, + "learning_rate": 3.32945490579186e-05, + "loss": 0.9272, + "step": 16465 + }, + { + "epoch": 2.931623931623932, + "grad_norm": 0.9139016270637512, + "learning_rate": 3.328412146989518e-05, + "loss": 1.0019, + "step": 16466 + }, + { + "epoch": 2.931801994301994, + "grad_norm": 0.9021140336990356, + "learning_rate": 3.327369518903085e-05, + "loss": 0.866, + "step": 16467 + }, + { + "epoch": 2.931980056980057, + "grad_norm": 0.8053449988365173, + "learning_rate": 3.326327021552984e-05, + "loss": 0.8915, + "step": 16468 + }, + { + "epoch": 2.9321581196581197, + "grad_norm": 1.0013985633850098, + "learning_rate": 3.325284654959643e-05, + "loss": 0.9063, + "step": 16469 + }, + { + "epoch": 2.9323361823361824, + "grad_norm": 0.8236168622970581, + "learning_rate": 3.324242419143483e-05, + "loss": 0.7113, + "step": 16470 + }, + { + "epoch": 2.932514245014245, + "grad_norm": 0.7256088852882385, + "learning_rate": 3.323200314124925e-05, + "loss": 0.6009, + "step": 16471 + }, + { + "epoch": 2.9326923076923075, + "grad_norm": 0.9991055727005005, + "learning_rate": 3.322158339924384e-05, + "loss": 1.039, + "step": 16472 + }, + { + "epoch": 2.9328703703703702, + "grad_norm": 0.8769686818122864, + "learning_rate": 3.3211164965622845e-05, + "loss": 0.989, + "step": 16473 + }, + { + "epoch": 2.933048433048433, + "grad_norm": 0.8226488828659058, + "learning_rate": 3.320074784059026e-05, + "loss": 0.9269, + "step": 16474 + }, + { + "epoch": 2.9332264957264957, + "grad_norm": 0.7763178944587708, + "learning_rate": 3.31903320243503e-05, + "loss": 0.9245, + "step": 16475 + }, + { + "epoch": 2.9334045584045585, + "grad_norm": 0.9204390645027161, + "learning_rate": 3.3179917517107e-05, + "loss": 1.2374, + "step": 16476 + }, + { + "epoch": 2.9335826210826212, + "grad_norm": 0.9970325827598572, + "learning_rate": 3.31695043190644e-05, + "loss": 0.7862, + "step": 16477 + }, + { + "epoch": 2.933760683760684, + "grad_norm": 0.8539462685585022, + "learning_rate": 3.315909243042654e-05, + "loss": 0.8015, + "step": 16478 + }, + { + "epoch": 2.9339387464387463, + "grad_norm": 0.8208832144737244, + "learning_rate": 3.314868185139742e-05, + "loss": 0.8571, + "step": 16479 + }, + { + "epoch": 2.934116809116809, + "grad_norm": 0.8628700375556946, + "learning_rate": 3.313827258218101e-05, + "loss": 0.8274, + "step": 16480 + }, + { + "epoch": 2.934294871794872, + "grad_norm": 1.0472661256790161, + "learning_rate": 3.312786462298124e-05, + "loss": 0.9252, + "step": 16481 + }, + { + "epoch": 2.9344729344729346, + "grad_norm": 0.8292158842086792, + "learning_rate": 3.311745797400202e-05, + "loss": 0.8548, + "step": 16482 + }, + { + "epoch": 2.9346509971509973, + "grad_norm": 0.85657799243927, + "learning_rate": 3.310705263544731e-05, + "loss": 0.7657, + "step": 16483 + }, + { + "epoch": 2.9348290598290596, + "grad_norm": 0.7560283541679382, + "learning_rate": 3.309664860752095e-05, + "loss": 0.8063, + "step": 16484 + }, + { + "epoch": 2.9350071225071224, + "grad_norm": 0.7933059334754944, + "learning_rate": 3.308624589042677e-05, + "loss": 0.7102, + "step": 16485 + }, + { + "epoch": 2.935185185185185, + "grad_norm": 0.8430653810501099, + "learning_rate": 3.3075844484368615e-05, + "loss": 0.8815, + "step": 16486 + }, + { + "epoch": 2.935363247863248, + "grad_norm": 0.8113032579421997, + "learning_rate": 3.306544438955021e-05, + "loss": 0.8144, + "step": 16487 + }, + { + "epoch": 2.9355413105413106, + "grad_norm": 0.8344797492027283, + "learning_rate": 3.3055045606175474e-05, + "loss": 0.7164, + "step": 16488 + }, + { + "epoch": 2.9357193732193734, + "grad_norm": 1.0109050273895264, + "learning_rate": 3.3044648134447964e-05, + "loss": 0.6619, + "step": 16489 + }, + { + "epoch": 2.935897435897436, + "grad_norm": 0.8746094703674316, + "learning_rate": 3.303425197457156e-05, + "loss": 0.73, + "step": 16490 + }, + { + "epoch": 2.9360754985754984, + "grad_norm": 0.919924795627594, + "learning_rate": 3.302385712674981e-05, + "loss": 1.1008, + "step": 16491 + }, + { + "epoch": 2.936253561253561, + "grad_norm": 0.8407595753669739, + "learning_rate": 3.301346359118648e-05, + "loss": 1.0963, + "step": 16492 + }, + { + "epoch": 2.936431623931624, + "grad_norm": 0.8145756721496582, + "learning_rate": 3.3003071368085184e-05, + "loss": 0.805, + "step": 16493 + }, + { + "epoch": 2.9366096866096867, + "grad_norm": 0.8298826813697815, + "learning_rate": 3.299268045764953e-05, + "loss": 0.7075, + "step": 16494 + }, + { + "epoch": 2.9367877492877494, + "grad_norm": 1.1027754545211792, + "learning_rate": 3.2982290860083106e-05, + "loss": 1.0381, + "step": 16495 + }, + { + "epoch": 2.9369658119658117, + "grad_norm": 0.8431075215339661, + "learning_rate": 3.2971902575589476e-05, + "loss": 0.9013, + "step": 16496 + }, + { + "epoch": 2.9371438746438745, + "grad_norm": 1.0045086145401, + "learning_rate": 3.296151560437214e-05, + "loss": 0.9054, + "step": 16497 + }, + { + "epoch": 2.9373219373219372, + "grad_norm": 0.8290889263153076, + "learning_rate": 3.295112994663471e-05, + "loss": 0.8054, + "step": 16498 + }, + { + "epoch": 2.9375, + "grad_norm": 0.7854097485542297, + "learning_rate": 3.2940745602580544e-05, + "loss": 0.7585, + "step": 16499 + }, + { + "epoch": 2.9376780626780628, + "grad_norm": 0.7470735907554626, + "learning_rate": 3.29303625724132e-05, + "loss": 0.7122, + "step": 16500 + }, + { + "epoch": 2.9378561253561255, + "grad_norm": 0.847463071346283, + "learning_rate": 3.2919980856336075e-05, + "loss": 0.7949, + "step": 16501 + }, + { + "epoch": 2.9380341880341883, + "grad_norm": 0.9595068097114563, + "learning_rate": 3.290960045455257e-05, + "loss": 0.8335, + "step": 16502 + }, + { + "epoch": 2.9382122507122506, + "grad_norm": 0.9636897444725037, + "learning_rate": 3.289922136726609e-05, + "loss": 0.7454, + "step": 16503 + }, + { + "epoch": 2.9383903133903133, + "grad_norm": 0.9030486941337585, + "learning_rate": 3.288884359467993e-05, + "loss": 0.8346, + "step": 16504 + }, + { + "epoch": 2.938568376068376, + "grad_norm": 0.8722931146621704, + "learning_rate": 3.287846713699755e-05, + "loss": 1.0182, + "step": 16505 + }, + { + "epoch": 2.938746438746439, + "grad_norm": 0.7612178325653076, + "learning_rate": 3.286809199442209e-05, + "loss": 0.8759, + "step": 16506 + }, + { + "epoch": 2.9389245014245016, + "grad_norm": 0.951334536075592, + "learning_rate": 3.2857718167156956e-05, + "loss": 0.8035, + "step": 16507 + }, + { + "epoch": 2.939102564102564, + "grad_norm": 0.8043029308319092, + "learning_rate": 3.284734565540536e-05, + "loss": 0.7652, + "step": 16508 + }, + { + "epoch": 2.9392806267806266, + "grad_norm": 0.8762648701667786, + "learning_rate": 3.283697445937053e-05, + "loss": 0.9362, + "step": 16509 + }, + { + "epoch": 2.9394586894586894, + "grad_norm": 0.8046880960464478, + "learning_rate": 3.282660457925566e-05, + "loss": 0.7757, + "step": 16510 + }, + { + "epoch": 2.939636752136752, + "grad_norm": 0.7703250050544739, + "learning_rate": 3.281623601526394e-05, + "loss": 0.6699, + "step": 16511 + }, + { + "epoch": 2.939814814814815, + "grad_norm": 0.9165888428688049, + "learning_rate": 3.280586876759847e-05, + "loss": 0.8321, + "step": 16512 + }, + { + "epoch": 2.9399928774928776, + "grad_norm": 0.7389699816703796, + "learning_rate": 3.279550283646249e-05, + "loss": 0.7621, + "step": 16513 + }, + { + "epoch": 2.9401709401709404, + "grad_norm": 0.9256302714347839, + "learning_rate": 3.278513822205897e-05, + "loss": 0.8634, + "step": 16514 + }, + { + "epoch": 2.9403490028490027, + "grad_norm": 0.8384902477264404, + "learning_rate": 3.277477492459109e-05, + "loss": 0.7645, + "step": 16515 + }, + { + "epoch": 2.9405270655270654, + "grad_norm": 0.9446337819099426, + "learning_rate": 3.276441294426178e-05, + "loss": 0.9729, + "step": 16516 + }, + { + "epoch": 2.940705128205128, + "grad_norm": 0.920237123966217, + "learning_rate": 3.275405228127417e-05, + "loss": 0.7834, + "step": 16517 + }, + { + "epoch": 2.940883190883191, + "grad_norm": 0.8432943224906921, + "learning_rate": 3.274369293583121e-05, + "loss": 0.8246, + "step": 16518 + }, + { + "epoch": 2.9410612535612537, + "grad_norm": 0.8046762943267822, + "learning_rate": 3.2733334908135885e-05, + "loss": 0.9363, + "step": 16519 + }, + { + "epoch": 2.941239316239316, + "grad_norm": 0.9555963277816772, + "learning_rate": 3.2722978198391106e-05, + "loss": 0.8699, + "step": 16520 + }, + { + "epoch": 2.9414173789173788, + "grad_norm": 0.8619177937507629, + "learning_rate": 3.2712622806799834e-05, + "loss": 0.8482, + "step": 16521 + }, + { + "epoch": 2.9415954415954415, + "grad_norm": 0.8801655769348145, + "learning_rate": 3.27022687335649e-05, + "loss": 0.9601, + "step": 16522 + }, + { + "epoch": 2.9417735042735043, + "grad_norm": 0.9054547548294067, + "learning_rate": 3.2691915978889244e-05, + "loss": 0.8752, + "step": 16523 + }, + { + "epoch": 2.941951566951567, + "grad_norm": 0.9078481197357178, + "learning_rate": 3.2681564542975675e-05, + "loss": 0.8225, + "step": 16524 + }, + { + "epoch": 2.9421296296296298, + "grad_norm": 0.9574032425880432, + "learning_rate": 3.267121442602701e-05, + "loss": 0.9133, + "step": 16525 + }, + { + "epoch": 2.9423076923076925, + "grad_norm": 0.8808075189590454, + "learning_rate": 3.2660865628246026e-05, + "loss": 0.8544, + "step": 16526 + }, + { + "epoch": 2.942485754985755, + "grad_norm": 0.8007816076278687, + "learning_rate": 3.26505181498355e-05, + "loss": 0.7846, + "step": 16527 + }, + { + "epoch": 2.9426638176638176, + "grad_norm": 0.8623418211936951, + "learning_rate": 3.264017199099816e-05, + "loss": 0.9056, + "step": 16528 + }, + { + "epoch": 2.9428418803418803, + "grad_norm": 0.870961606502533, + "learning_rate": 3.2629827151936695e-05, + "loss": 0.9883, + "step": 16529 + }, + { + "epoch": 2.943019943019943, + "grad_norm": 0.9122142791748047, + "learning_rate": 3.2619483632853885e-05, + "loss": 0.8012, + "step": 16530 + }, + { + "epoch": 2.943198005698006, + "grad_norm": 0.9072979688644409, + "learning_rate": 3.260914143395225e-05, + "loss": 0.9999, + "step": 16531 + }, + { + "epoch": 2.943376068376068, + "grad_norm": 0.8746095299720764, + "learning_rate": 3.259880055543454e-05, + "loss": 0.8022, + "step": 16532 + }, + { + "epoch": 2.943554131054131, + "grad_norm": 1.0012668371200562, + "learning_rate": 3.2588460997503314e-05, + "loss": 0.8909, + "step": 16533 + }, + { + "epoch": 2.9437321937321936, + "grad_norm": 0.913070023059845, + "learning_rate": 3.2578122760361154e-05, + "loss": 0.7729, + "step": 16534 + }, + { + "epoch": 2.9439102564102564, + "grad_norm": 0.791753888130188, + "learning_rate": 3.2567785844210616e-05, + "loss": 0.9731, + "step": 16535 + }, + { + "epoch": 2.944088319088319, + "grad_norm": 0.9673776030540466, + "learning_rate": 3.255745024925425e-05, + "loss": 0.8008, + "step": 16536 + }, + { + "epoch": 2.944266381766382, + "grad_norm": 0.9382752180099487, + "learning_rate": 3.254711597569454e-05, + "loss": 0.9611, + "step": 16537 + }, + { + "epoch": 2.9444444444444446, + "grad_norm": 0.8816630840301514, + "learning_rate": 3.2536783023733975e-05, + "loss": 0.7565, + "step": 16538 + }, + { + "epoch": 2.9446225071225074, + "grad_norm": 0.8474457859992981, + "learning_rate": 3.2526451393574964e-05, + "loss": 0.7766, + "step": 16539 + }, + { + "epoch": 2.9448005698005697, + "grad_norm": 0.9510074257850647, + "learning_rate": 3.251612108542005e-05, + "loss": 0.9899, + "step": 16540 + }, + { + "epoch": 2.9449786324786325, + "grad_norm": 0.776924192905426, + "learning_rate": 3.250579209947149e-05, + "loss": 0.7845, + "step": 16541 + }, + { + "epoch": 2.945156695156695, + "grad_norm": 0.9543585181236267, + "learning_rate": 3.2495464435931756e-05, + "loss": 1.1313, + "step": 16542 + }, + { + "epoch": 2.945334757834758, + "grad_norm": 0.9087918400764465, + "learning_rate": 3.2485138095003164e-05, + "loss": 0.774, + "step": 16543 + }, + { + "epoch": 2.9455128205128203, + "grad_norm": 0.7719675302505493, + "learning_rate": 3.247481307688801e-05, + "loss": 0.7755, + "step": 16544 + }, + { + "epoch": 2.945690883190883, + "grad_norm": 0.8550716638565063, + "learning_rate": 3.246448938178869e-05, + "loss": 0.8455, + "step": 16545 + }, + { + "epoch": 2.9458689458689458, + "grad_norm": 0.8585572838783264, + "learning_rate": 3.2454167009907346e-05, + "loss": 0.7048, + "step": 16546 + }, + { + "epoch": 2.9460470085470085, + "grad_norm": 0.819939136505127, + "learning_rate": 3.2443845961446315e-05, + "loss": 0.7671, + "step": 16547 + }, + { + "epoch": 2.9462250712250713, + "grad_norm": 0.8811594247817993, + "learning_rate": 3.243352623660778e-05, + "loss": 0.9244, + "step": 16548 + }, + { + "epoch": 2.946403133903134, + "grad_norm": 0.8128607869148254, + "learning_rate": 3.242320783559395e-05, + "loss": 0.8103, + "step": 16549 + }, + { + "epoch": 2.9465811965811968, + "grad_norm": 0.778759241104126, + "learning_rate": 3.2412890758606985e-05, + "loss": 0.6261, + "step": 16550 + }, + { + "epoch": 2.9467592592592595, + "grad_norm": 1.0277715921401978, + "learning_rate": 3.240257500584901e-05, + "loss": 0.8728, + "step": 16551 + }, + { + "epoch": 2.946937321937322, + "grad_norm": 0.7647507786750793, + "learning_rate": 3.239226057752217e-05, + "loss": 0.7142, + "step": 16552 + }, + { + "epoch": 2.9471153846153846, + "grad_norm": 0.8381546139717102, + "learning_rate": 3.238194747382855e-05, + "loss": 0.737, + "step": 16553 + }, + { + "epoch": 2.9472934472934473, + "grad_norm": 0.7928317189216614, + "learning_rate": 3.237163569497016e-05, + "loss": 0.8926, + "step": 16554 + }, + { + "epoch": 2.94747150997151, + "grad_norm": 0.7086058259010315, + "learning_rate": 3.236132524114914e-05, + "loss": 0.6845, + "step": 16555 + }, + { + "epoch": 2.9476495726495724, + "grad_norm": 0.83002769947052, + "learning_rate": 3.235101611256739e-05, + "loss": 0.7365, + "step": 16556 + }, + { + "epoch": 2.947827635327635, + "grad_norm": 0.9012778401374817, + "learning_rate": 3.234070830942698e-05, + "loss": 0.9999, + "step": 16557 + }, + { + "epoch": 2.948005698005698, + "grad_norm": 0.8554810881614685, + "learning_rate": 3.233040183192985e-05, + "loss": 0.6699, + "step": 16558 + }, + { + "epoch": 2.9481837606837606, + "grad_norm": 0.9322055578231812, + "learning_rate": 3.2320096680277915e-05, + "loss": 0.8969, + "step": 16559 + }, + { + "epoch": 2.9483618233618234, + "grad_norm": 0.8755966424942017, + "learning_rate": 3.2309792854673095e-05, + "loss": 0.8398, + "step": 16560 + }, + { + "epoch": 2.948539886039886, + "grad_norm": 0.8203766345977783, + "learning_rate": 3.229949035531726e-05, + "loss": 0.8354, + "step": 16561 + }, + { + "epoch": 2.948717948717949, + "grad_norm": 0.8970799446105957, + "learning_rate": 3.228918918241229e-05, + "loss": 0.9074, + "step": 16562 + }, + { + "epoch": 2.9488960113960117, + "grad_norm": 0.8263736963272095, + "learning_rate": 3.227888933615997e-05, + "loss": 0.8896, + "step": 16563 + }, + { + "epoch": 2.949074074074074, + "grad_norm": 1.0277043581008911, + "learning_rate": 3.2268590816762155e-05, + "loss": 0.8706, + "step": 16564 + }, + { + "epoch": 2.9492521367521367, + "grad_norm": 0.8965407013893127, + "learning_rate": 3.225829362442061e-05, + "loss": 0.6796, + "step": 16565 + }, + { + "epoch": 2.9494301994301995, + "grad_norm": 0.8175839185714722, + "learning_rate": 3.224799775933708e-05, + "loss": 0.9243, + "step": 16566 + }, + { + "epoch": 2.949608262108262, + "grad_norm": 0.7979576587677002, + "learning_rate": 3.2237703221713286e-05, + "loss": 0.8008, + "step": 16567 + }, + { + "epoch": 2.9497863247863245, + "grad_norm": 1.054843783378601, + "learning_rate": 3.2227410011750945e-05, + "loss": 1.0279, + "step": 16568 + }, + { + "epoch": 2.9499643874643873, + "grad_norm": 0.7947831749916077, + "learning_rate": 3.221711812965168e-05, + "loss": 0.6767, + "step": 16569 + }, + { + "epoch": 2.95014245014245, + "grad_norm": 0.8702623844146729, + "learning_rate": 3.220682757561725e-05, + "loss": 0.9844, + "step": 16570 + }, + { + "epoch": 2.9503205128205128, + "grad_norm": 0.8653056621551514, + "learning_rate": 3.2196538349849123e-05, + "loss": 1.0747, + "step": 16571 + }, + { + "epoch": 2.9504985754985755, + "grad_norm": 0.9718163013458252, + "learning_rate": 3.2186250452549026e-05, + "loss": 0.7793, + "step": 16572 + }, + { + "epoch": 2.9506766381766383, + "grad_norm": 0.8626788854598999, + "learning_rate": 3.217596388391848e-05, + "loss": 0.9145, + "step": 16573 + }, + { + "epoch": 2.950854700854701, + "grad_norm": 0.9753466844558716, + "learning_rate": 3.2165678644159025e-05, + "loss": 0.9784, + "step": 16574 + }, + { + "epoch": 2.951032763532764, + "grad_norm": 0.795011043548584, + "learning_rate": 3.2155394733472186e-05, + "loss": 0.7457, + "step": 16575 + }, + { + "epoch": 2.951210826210826, + "grad_norm": 0.842991828918457, + "learning_rate": 3.2145112152059454e-05, + "loss": 0.9126, + "step": 16576 + }, + { + "epoch": 2.951388888888889, + "grad_norm": 0.8642476797103882, + "learning_rate": 3.213483090012228e-05, + "loss": 0.7213, + "step": 16577 + }, + { + "epoch": 2.9515669515669516, + "grad_norm": 0.9114034175872803, + "learning_rate": 3.212455097786214e-05, + "loss": 0.651, + "step": 16578 + }, + { + "epoch": 2.9517450142450143, + "grad_norm": 1.0575958490371704, + "learning_rate": 3.211427238548037e-05, + "loss": 0.9727, + "step": 16579 + }, + { + "epoch": 2.9519230769230766, + "grad_norm": 0.8618924617767334, + "learning_rate": 3.210399512317849e-05, + "loss": 0.8593, + "step": 16580 + }, + { + "epoch": 2.9521011396011394, + "grad_norm": 0.768313467502594, + "learning_rate": 3.209371919115771e-05, + "loss": 0.8305, + "step": 16581 + }, + { + "epoch": 2.952279202279202, + "grad_norm": 0.8160355091094971, + "learning_rate": 3.208344458961947e-05, + "loss": 0.7266, + "step": 16582 + }, + { + "epoch": 2.952457264957265, + "grad_norm": 0.908545196056366, + "learning_rate": 3.207317131876506e-05, + "loss": 0.8022, + "step": 16583 + }, + { + "epoch": 2.9526353276353277, + "grad_norm": 1.2593516111373901, + "learning_rate": 3.206289937879571e-05, + "loss": 1.0274, + "step": 16584 + }, + { + "epoch": 2.9528133903133904, + "grad_norm": 0.7789214849472046, + "learning_rate": 3.2052628769912795e-05, + "loss": 0.8993, + "step": 16585 + }, + { + "epoch": 2.952991452991453, + "grad_norm": 0.8475270867347717, + "learning_rate": 3.20423594923174e-05, + "loss": 1.0145, + "step": 16586 + }, + { + "epoch": 2.953169515669516, + "grad_norm": 1.0083874464035034, + "learning_rate": 3.203209154621086e-05, + "loss": 1.02, + "step": 16587 + }, + { + "epoch": 2.953347578347578, + "grad_norm": 0.7013131380081177, + "learning_rate": 3.2021824931794245e-05, + "loss": 0.5449, + "step": 16588 + }, + { + "epoch": 2.953525641025641, + "grad_norm": 0.8298764824867249, + "learning_rate": 3.201155964926878e-05, + "loss": 0.9914, + "step": 16589 + }, + { + "epoch": 2.9537037037037037, + "grad_norm": 0.9371963143348694, + "learning_rate": 3.200129569883556e-05, + "loss": 0.7349, + "step": 16590 + }, + { + "epoch": 2.9538817663817665, + "grad_norm": 0.7932438850402832, + "learning_rate": 3.199103308069571e-05, + "loss": 0.7911, + "step": 16591 + }, + { + "epoch": 2.9540598290598292, + "grad_norm": 0.8415567278862, + "learning_rate": 3.198077179505029e-05, + "loss": 0.766, + "step": 16592 + }, + { + "epoch": 2.9542378917378915, + "grad_norm": 0.9155336618423462, + "learning_rate": 3.197051184210035e-05, + "loss": 0.9622, + "step": 16593 + }, + { + "epoch": 2.9544159544159543, + "grad_norm": 0.8972458839416504, + "learning_rate": 3.196025322204688e-05, + "loss": 1.0202, + "step": 16594 + }, + { + "epoch": 2.954594017094017, + "grad_norm": 0.9620199799537659, + "learning_rate": 3.194999593509096e-05, + "loss": 0.8114, + "step": 16595 + }, + { + "epoch": 2.95477207977208, + "grad_norm": 0.819244921207428, + "learning_rate": 3.1939739981433456e-05, + "loss": 0.7858, + "step": 16596 + }, + { + "epoch": 2.9549501424501425, + "grad_norm": 0.8560270667076111, + "learning_rate": 3.192948536127542e-05, + "loss": 0.9163, + "step": 16597 + }, + { + "epoch": 2.9551282051282053, + "grad_norm": 0.9105932116508484, + "learning_rate": 3.191923207481765e-05, + "loss": 0.7372, + "step": 16598 + }, + { + "epoch": 2.955306267806268, + "grad_norm": 0.888846218585968, + "learning_rate": 3.190898012226114e-05, + "loss": 0.8624, + "step": 16599 + }, + { + "epoch": 2.9554843304843303, + "grad_norm": 0.8116851449012756, + "learning_rate": 3.1898729503806726e-05, + "loss": 0.8549, + "step": 16600 + }, + { + "epoch": 2.955662393162393, + "grad_norm": 0.8970577120780945, + "learning_rate": 3.188848021965522e-05, + "loss": 0.8228, + "step": 16601 + }, + { + "epoch": 2.955840455840456, + "grad_norm": 0.8820711374282837, + "learning_rate": 3.187823227000747e-05, + "loss": 0.7976, + "step": 16602 + }, + { + "epoch": 2.9560185185185186, + "grad_norm": 0.9349139928817749, + "learning_rate": 3.1867985655064205e-05, + "loss": 0.8965, + "step": 16603 + }, + { + "epoch": 2.9561965811965814, + "grad_norm": 0.939132571220398, + "learning_rate": 3.185774037502627e-05, + "loss": 0.9152, + "step": 16604 + }, + { + "epoch": 2.9563746438746437, + "grad_norm": 1.029159665107727, + "learning_rate": 3.184749643009435e-05, + "loss": 1.0904, + "step": 16605 + }, + { + "epoch": 2.9565527065527064, + "grad_norm": 0.966548502445221, + "learning_rate": 3.183725382046917e-05, + "loss": 0.9508, + "step": 16606 + }, + { + "epoch": 2.956730769230769, + "grad_norm": 0.8782771229743958, + "learning_rate": 3.18270125463514e-05, + "loss": 0.9929, + "step": 16607 + }, + { + "epoch": 2.956908831908832, + "grad_norm": 0.9193231463432312, + "learning_rate": 3.18167726079417e-05, + "loss": 0.9942, + "step": 16608 + }, + { + "epoch": 2.9570868945868947, + "grad_norm": 0.8940062522888184, + "learning_rate": 3.180653400544071e-05, + "loss": 1.0594, + "step": 16609 + }, + { + "epoch": 2.9572649572649574, + "grad_norm": 0.8237268328666687, + "learning_rate": 3.179629673904903e-05, + "loss": 0.8211, + "step": 16610 + }, + { + "epoch": 2.95744301994302, + "grad_norm": 0.92745041847229, + "learning_rate": 3.17860608089672e-05, + "loss": 0.8287, + "step": 16611 + }, + { + "epoch": 2.9576210826210825, + "grad_norm": 0.8203856348991394, + "learning_rate": 3.177582621539586e-05, + "loss": 0.9457, + "step": 16612 + }, + { + "epoch": 2.9577991452991452, + "grad_norm": 0.8719314932823181, + "learning_rate": 3.176559295853543e-05, + "loss": 0.8022, + "step": 16613 + }, + { + "epoch": 2.957977207977208, + "grad_norm": 0.8742199540138245, + "learning_rate": 3.175536103858648e-05, + "loss": 0.976, + "step": 16614 + }, + { + "epoch": 2.9581552706552707, + "grad_norm": 0.8677577972412109, + "learning_rate": 3.174513045574947e-05, + "loss": 0.9513, + "step": 16615 + }, + { + "epoch": 2.9583333333333335, + "grad_norm": 0.809138298034668, + "learning_rate": 3.173490121022485e-05, + "loss": 0.7001, + "step": 16616 + }, + { + "epoch": 2.958511396011396, + "grad_norm": 0.9672527313232422, + "learning_rate": 3.1724673302213025e-05, + "loss": 0.9932, + "step": 16617 + }, + { + "epoch": 2.9586894586894585, + "grad_norm": 0.8101853728294373, + "learning_rate": 3.17144467319144e-05, + "loss": 0.8236, + "step": 16618 + }, + { + "epoch": 2.9588675213675213, + "grad_norm": 0.8965981006622314, + "learning_rate": 3.170422149952931e-05, + "loss": 0.9353, + "step": 16619 + }, + { + "epoch": 2.959045584045584, + "grad_norm": 0.8267533183097839, + "learning_rate": 3.1693997605258184e-05, + "loss": 0.6445, + "step": 16620 + }, + { + "epoch": 2.959223646723647, + "grad_norm": 0.9084979891777039, + "learning_rate": 3.168377504930122e-05, + "loss": 0.9311, + "step": 16621 + }, + { + "epoch": 2.9594017094017095, + "grad_norm": 0.8414687514305115, + "learning_rate": 3.1673553831858805e-05, + "loss": 0.7663, + "step": 16622 + }, + { + "epoch": 2.9595797720797723, + "grad_norm": 0.8619266748428345, + "learning_rate": 3.166333395313116e-05, + "loss": 0.8524, + "step": 16623 + }, + { + "epoch": 2.9597578347578346, + "grad_norm": 0.9963071346282959, + "learning_rate": 3.1653115413318534e-05, + "loss": 0.813, + "step": 16624 + }, + { + "epoch": 2.9599358974358974, + "grad_norm": 0.8264978528022766, + "learning_rate": 3.164289821262113e-05, + "loss": 0.9716, + "step": 16625 + }, + { + "epoch": 2.96011396011396, + "grad_norm": 0.8166584372520447, + "learning_rate": 3.163268235123911e-05, + "loss": 0.7492, + "step": 16626 + }, + { + "epoch": 2.960292022792023, + "grad_norm": 0.809241533279419, + "learning_rate": 3.1622467829372724e-05, + "loss": 0.9033, + "step": 16627 + }, + { + "epoch": 2.9604700854700856, + "grad_norm": 1.0441617965698242, + "learning_rate": 3.161225464722197e-05, + "loss": 0.7305, + "step": 16628 + }, + { + "epoch": 2.960648148148148, + "grad_norm": 0.9226490259170532, + "learning_rate": 3.160204280498705e-05, + "loss": 0.6896, + "step": 16629 + }, + { + "epoch": 2.9608262108262107, + "grad_norm": 0.7890266180038452, + "learning_rate": 3.159183230286803e-05, + "loss": 0.873, + "step": 16630 + }, + { + "epoch": 2.9610042735042734, + "grad_norm": 0.8491777181625366, + "learning_rate": 3.1581623141064934e-05, + "loss": 0.8828, + "step": 16631 + }, + { + "epoch": 2.961182336182336, + "grad_norm": 0.7522078156471252, + "learning_rate": 3.157141531977782e-05, + "loss": 0.7717, + "step": 16632 + }, + { + "epoch": 2.961360398860399, + "grad_norm": 0.9260183572769165, + "learning_rate": 3.156120883920667e-05, + "loss": 0.9715, + "step": 16633 + }, + { + "epoch": 2.9615384615384617, + "grad_norm": 0.6620128154754639, + "learning_rate": 3.1551003699551465e-05, + "loss": 0.5624, + "step": 16634 + }, + { + "epoch": 2.9617165242165244, + "grad_norm": 0.7571594715118408, + "learning_rate": 3.154079990101214e-05, + "loss": 0.6911, + "step": 16635 + }, + { + "epoch": 2.9618945868945867, + "grad_norm": 0.8571279644966125, + "learning_rate": 3.153059744378861e-05, + "loss": 0.9057, + "step": 16636 + }, + { + "epoch": 2.9620726495726495, + "grad_norm": 0.8895478248596191, + "learning_rate": 3.152039632808085e-05, + "loss": 0.7097, + "step": 16637 + }, + { + "epoch": 2.9622507122507122, + "grad_norm": 0.8340024352073669, + "learning_rate": 3.15101965540886e-05, + "loss": 0.7719, + "step": 16638 + }, + { + "epoch": 2.962428774928775, + "grad_norm": 0.8263829946517944, + "learning_rate": 3.149999812201182e-05, + "loss": 0.8561, + "step": 16639 + }, + { + "epoch": 2.9626068376068377, + "grad_norm": 0.9083819389343262, + "learning_rate": 3.148980103205027e-05, + "loss": 0.9319, + "step": 16640 + }, + { + "epoch": 2.9627849002849, + "grad_norm": 0.9346078038215637, + "learning_rate": 3.147960528440372e-05, + "loss": 0.8036, + "step": 16641 + }, + { + "epoch": 2.962962962962963, + "grad_norm": 0.907319188117981, + "learning_rate": 3.146941087927203e-05, + "loss": 0.9228, + "step": 16642 + }, + { + "epoch": 2.9631410256410255, + "grad_norm": 0.7912126183509827, + "learning_rate": 3.1459217816854815e-05, + "loss": 0.7996, + "step": 16643 + }, + { + "epoch": 2.9633190883190883, + "grad_norm": 0.8844919800758362, + "learning_rate": 3.1449026097351896e-05, + "loss": 1.0235, + "step": 16644 + }, + { + "epoch": 2.963497150997151, + "grad_norm": 0.7468230128288269, + "learning_rate": 3.143883572096286e-05, + "loss": 0.7292, + "step": 16645 + }, + { + "epoch": 2.963675213675214, + "grad_norm": 0.8521941900253296, + "learning_rate": 3.142864668788744e-05, + "loss": 0.9012, + "step": 16646 + }, + { + "epoch": 2.9638532763532766, + "grad_norm": 0.9340695738792419, + "learning_rate": 3.141845899832524e-05, + "loss": 0.8736, + "step": 16647 + }, + { + "epoch": 2.9640313390313393, + "grad_norm": 0.859395444393158, + "learning_rate": 3.140827265247588e-05, + "loss": 0.796, + "step": 16648 + }, + { + "epoch": 2.9642094017094016, + "grad_norm": 0.8320850729942322, + "learning_rate": 3.139808765053892e-05, + "loss": 0.86, + "step": 16649 + }, + { + "epoch": 2.9643874643874644, + "grad_norm": 0.8911257386207581, + "learning_rate": 3.138790399271393e-05, + "loss": 0.966, + "step": 16650 + }, + { + "epoch": 2.964565527065527, + "grad_norm": 0.8817025423049927, + "learning_rate": 3.13777216792004e-05, + "loss": 0.983, + "step": 16651 + }, + { + "epoch": 2.96474358974359, + "grad_norm": 0.7765538692474365, + "learning_rate": 3.136754071019793e-05, + "loss": 0.643, + "step": 16652 + }, + { + "epoch": 2.964921652421652, + "grad_norm": 0.7961843609809875, + "learning_rate": 3.135736108590586e-05, + "loss": 0.733, + "step": 16653 + }, + { + "epoch": 2.965099715099715, + "grad_norm": 0.7910877466201782, + "learning_rate": 3.134718280652373e-05, + "loss": 0.7291, + "step": 16654 + }, + { + "epoch": 2.9652777777777777, + "grad_norm": 0.9326547384262085, + "learning_rate": 3.1337005872250956e-05, + "loss": 0.8834, + "step": 16655 + }, + { + "epoch": 2.9654558404558404, + "grad_norm": 0.9362995624542236, + "learning_rate": 3.132683028328691e-05, + "loss": 0.9693, + "step": 16656 + }, + { + "epoch": 2.965633903133903, + "grad_norm": 0.8322434425354004, + "learning_rate": 3.131665603983096e-05, + "loss": 0.9782, + "step": 16657 + }, + { + "epoch": 2.965811965811966, + "grad_norm": 0.8336097598075867, + "learning_rate": 3.130648314208247e-05, + "loss": 0.9995, + "step": 16658 + }, + { + "epoch": 2.9659900284900287, + "grad_norm": 0.8637044429779053, + "learning_rate": 3.129631159024074e-05, + "loss": 0.8291, + "step": 16659 + }, + { + "epoch": 2.9661680911680914, + "grad_norm": 0.7853943109512329, + "learning_rate": 3.128614138450506e-05, + "loss": 0.9016, + "step": 16660 + }, + { + "epoch": 2.9663461538461537, + "grad_norm": 0.9506492614746094, + "learning_rate": 3.1275972525074674e-05, + "loss": 0.9762, + "step": 16661 + }, + { + "epoch": 2.9665242165242165, + "grad_norm": 0.9375113248825073, + "learning_rate": 3.126580501214887e-05, + "loss": 0.8981, + "step": 16662 + }, + { + "epoch": 2.9667022792022792, + "grad_norm": 0.9404717683792114, + "learning_rate": 3.125563884592684e-05, + "loss": 0.7506, + "step": 16663 + }, + { + "epoch": 2.966880341880342, + "grad_norm": 0.9678782820701599, + "learning_rate": 3.124547402660776e-05, + "loss": 0.896, + "step": 16664 + }, + { + "epoch": 2.9670584045584043, + "grad_norm": 0.8530639410018921, + "learning_rate": 3.12353105543908e-05, + "loss": 0.8792, + "step": 16665 + }, + { + "epoch": 2.967236467236467, + "grad_norm": 0.8015365600585938, + "learning_rate": 3.122514842947504e-05, + "loss": 0.7052, + "step": 16666 + }, + { + "epoch": 2.96741452991453, + "grad_norm": 0.8753054141998291, + "learning_rate": 3.121498765205969e-05, + "loss": 0.9032, + "step": 16667 + }, + { + "epoch": 2.9675925925925926, + "grad_norm": 1.0053389072418213, + "learning_rate": 3.1204828222343716e-05, + "loss": 1.0236, + "step": 16668 + }, + { + "epoch": 2.9677706552706553, + "grad_norm": 0.8506449460983276, + "learning_rate": 3.119467014052628e-05, + "loss": 0.6762, + "step": 16669 + }, + { + "epoch": 2.967948717948718, + "grad_norm": 0.9376404881477356, + "learning_rate": 3.118451340680629e-05, + "loss": 0.9391, + "step": 16670 + }, + { + "epoch": 2.968126780626781, + "grad_norm": 0.8381097912788391, + "learning_rate": 3.117435802138284e-05, + "loss": 0.7343, + "step": 16671 + }, + { + "epoch": 2.9683048433048436, + "grad_norm": 0.8525682091712952, + "learning_rate": 3.116420398445488e-05, + "loss": 0.7928, + "step": 16672 + }, + { + "epoch": 2.968482905982906, + "grad_norm": 0.9399489760398865, + "learning_rate": 3.115405129622133e-05, + "loss": 0.975, + "step": 16673 + }, + { + "epoch": 2.9686609686609686, + "grad_norm": 0.8394346237182617, + "learning_rate": 3.114389995688114e-05, + "loss": 0.7591, + "step": 16674 + }, + { + "epoch": 2.9688390313390314, + "grad_norm": 0.8935068845748901, + "learning_rate": 3.11337499666332e-05, + "loss": 0.8899, + "step": 16675 + }, + { + "epoch": 2.969017094017094, + "grad_norm": 0.8111040592193604, + "learning_rate": 3.112360132567633e-05, + "loss": 0.795, + "step": 16676 + }, + { + "epoch": 2.9691951566951564, + "grad_norm": 0.8854177594184875, + "learning_rate": 3.1113454034209486e-05, + "loss": 1.0677, + "step": 16677 + }, + { + "epoch": 2.969373219373219, + "grad_norm": 0.9821479916572571, + "learning_rate": 3.110330809243134e-05, + "loss": 0.7859, + "step": 16678 + }, + { + "epoch": 2.969551282051282, + "grad_norm": 0.9066275954246521, + "learning_rate": 3.109316350054079e-05, + "loss": 1.1727, + "step": 16679 + }, + { + "epoch": 2.9697293447293447, + "grad_norm": 0.981238603591919, + "learning_rate": 3.108302025873656e-05, + "loss": 1.0036, + "step": 16680 + }, + { + "epoch": 2.9699074074074074, + "grad_norm": 0.8290690779685974, + "learning_rate": 3.107287836721737e-05, + "loss": 0.911, + "step": 16681 + }, + { + "epoch": 2.97008547008547, + "grad_norm": 0.8419190049171448, + "learning_rate": 3.106273782618196e-05, + "loss": 0.688, + "step": 16682 + }, + { + "epoch": 2.970263532763533, + "grad_norm": 0.9250679612159729, + "learning_rate": 3.1052598635828964e-05, + "loss": 0.8506, + "step": 16683 + }, + { + "epoch": 2.9704415954415957, + "grad_norm": 0.9414278864860535, + "learning_rate": 3.104246079635713e-05, + "loss": 0.8501, + "step": 16684 + }, + { + "epoch": 2.970619658119658, + "grad_norm": 0.9107208847999573, + "learning_rate": 3.1032324307964974e-05, + "loss": 0.8234, + "step": 16685 + }, + { + "epoch": 2.9707977207977208, + "grad_norm": 0.8691245317459106, + "learning_rate": 3.102218917085119e-05, + "loss": 0.9341, + "step": 16686 + }, + { + "epoch": 2.9709757834757835, + "grad_norm": 0.8816282153129578, + "learning_rate": 3.101205538521431e-05, + "loss": 0.9412, + "step": 16687 + }, + { + "epoch": 2.9711538461538463, + "grad_norm": 0.8969736695289612, + "learning_rate": 3.100192295125289e-05, + "loss": 0.9468, + "step": 16688 + }, + { + "epoch": 2.9713319088319086, + "grad_norm": 1.0070735216140747, + "learning_rate": 3.099179186916548e-05, + "loss": 0.8587, + "step": 16689 + }, + { + "epoch": 2.9715099715099713, + "grad_norm": 0.9801154136657715, + "learning_rate": 3.0981662139150537e-05, + "loss": 0.8999, + "step": 16690 + }, + { + "epoch": 2.971688034188034, + "grad_norm": 0.9949473142623901, + "learning_rate": 3.0971533761406526e-05, + "loss": 1.0552, + "step": 16691 + }, + { + "epoch": 2.971866096866097, + "grad_norm": 0.9654440879821777, + "learning_rate": 3.096140673613198e-05, + "loss": 0.8618, + "step": 16692 + }, + { + "epoch": 2.9720441595441596, + "grad_norm": 0.9247317910194397, + "learning_rate": 3.0951281063525185e-05, + "loss": 0.8915, + "step": 16693 + }, + { + "epoch": 2.9722222222222223, + "grad_norm": 1.0254271030426025, + "learning_rate": 3.094115674378467e-05, + "loss": 0.9098, + "step": 16694 + }, + { + "epoch": 2.972400284900285, + "grad_norm": 0.7822396159172058, + "learning_rate": 3.0931033777108666e-05, + "loss": 0.8497, + "step": 16695 + }, + { + "epoch": 2.972578347578348, + "grad_norm": 0.9368909001350403, + "learning_rate": 3.092091216369561e-05, + "loss": 0.8978, + "step": 16696 + }, + { + "epoch": 2.97275641025641, + "grad_norm": 1.0181694030761719, + "learning_rate": 3.0910791903743786e-05, + "loss": 1.0605, + "step": 16697 + }, + { + "epoch": 2.972934472934473, + "grad_norm": 0.8718426823616028, + "learning_rate": 3.090067299745146e-05, + "loss": 0.9844, + "step": 16698 + }, + { + "epoch": 2.9731125356125356, + "grad_norm": 0.8327153325080872, + "learning_rate": 3.089055544501692e-05, + "loss": 0.8795, + "step": 16699 + }, + { + "epoch": 2.9732905982905984, + "grad_norm": 0.7970293164253235, + "learning_rate": 3.088043924663838e-05, + "loss": 0.794, + "step": 16700 + }, + { + "epoch": 2.9734686609686607, + "grad_norm": 0.8885688185691833, + "learning_rate": 3.0870324402514014e-05, + "loss": 0.984, + "step": 16701 + }, + { + "epoch": 2.9736467236467234, + "grad_norm": 0.9126071333885193, + "learning_rate": 3.086021091284207e-05, + "loss": 0.8962, + "step": 16702 + }, + { + "epoch": 2.973824786324786, + "grad_norm": 0.9220738410949707, + "learning_rate": 3.085009877782067e-05, + "loss": 0.9129, + "step": 16703 + }, + { + "epoch": 2.974002849002849, + "grad_norm": 0.8738197684288025, + "learning_rate": 3.0839987997647935e-05, + "loss": 0.871, + "step": 16704 + }, + { + "epoch": 2.9741809116809117, + "grad_norm": 0.7505079507827759, + "learning_rate": 3.0829878572521964e-05, + "loss": 0.7934, + "step": 16705 + }, + { + "epoch": 2.9743589743589745, + "grad_norm": 0.8328925371170044, + "learning_rate": 3.081977050264084e-05, + "loss": 0.8431, + "step": 16706 + }, + { + "epoch": 2.974537037037037, + "grad_norm": 0.8248029947280884, + "learning_rate": 3.08096637882026e-05, + "loss": 0.7877, + "step": 16707 + }, + { + "epoch": 2.9747150997151, + "grad_norm": 1.1086161136627197, + "learning_rate": 3.079955842940524e-05, + "loss": 1.1196, + "step": 16708 + }, + { + "epoch": 2.9748931623931623, + "grad_norm": 0.8271856307983398, + "learning_rate": 3.078945442644684e-05, + "loss": 0.946, + "step": 16709 + }, + { + "epoch": 2.975071225071225, + "grad_norm": 0.9062432050704956, + "learning_rate": 3.0779351779525246e-05, + "loss": 0.9369, + "step": 16710 + }, + { + "epoch": 2.9752492877492878, + "grad_norm": 0.9849454760551453, + "learning_rate": 3.0769250488838496e-05, + "loss": 0.7748, + "step": 16711 + }, + { + "epoch": 2.9754273504273505, + "grad_norm": 0.8420920372009277, + "learning_rate": 3.075915055458447e-05, + "loss": 0.9954, + "step": 16712 + }, + { + "epoch": 2.9756054131054133, + "grad_norm": 1.027130126953125, + "learning_rate": 3.074905197696104e-05, + "loss": 1.1053, + "step": 16713 + }, + { + "epoch": 2.9757834757834756, + "grad_norm": 0.8968556523323059, + "learning_rate": 3.073895475616609e-05, + "loss": 0.7487, + "step": 16714 + }, + { + "epoch": 2.9759615384615383, + "grad_norm": 0.9237984418869019, + "learning_rate": 3.072885889239745e-05, + "loss": 1.2853, + "step": 16715 + }, + { + "epoch": 2.976139601139601, + "grad_norm": 0.7972691059112549, + "learning_rate": 3.071876438585292e-05, + "loss": 0.6724, + "step": 16716 + }, + { + "epoch": 2.976317663817664, + "grad_norm": 0.8158389925956726, + "learning_rate": 3.0708671236730266e-05, + "loss": 0.8302, + "step": 16717 + }, + { + "epoch": 2.9764957264957266, + "grad_norm": 0.7947682738304138, + "learning_rate": 3.0698579445227236e-05, + "loss": 0.7817, + "step": 16718 + }, + { + "epoch": 2.9766737891737893, + "grad_norm": 0.847747802734375, + "learning_rate": 3.068848901154166e-05, + "loss": 0.8391, + "step": 16719 + }, + { + "epoch": 2.976851851851852, + "grad_norm": 0.7906867265701294, + "learning_rate": 3.067839993587107e-05, + "loss": 0.8935, + "step": 16720 + }, + { + "epoch": 2.9770299145299144, + "grad_norm": 0.9028356075286865, + "learning_rate": 3.066831221841328e-05, + "loss": 0.821, + "step": 16721 + }, + { + "epoch": 2.977207977207977, + "grad_norm": 0.8177092671394348, + "learning_rate": 3.065822585936589e-05, + "loss": 0.7956, + "step": 16722 + }, + { + "epoch": 2.97738603988604, + "grad_norm": 0.771271824836731, + "learning_rate": 3.064814085892647e-05, + "loss": 0.8675, + "step": 16723 + }, + { + "epoch": 2.9775641025641026, + "grad_norm": 0.8575150966644287, + "learning_rate": 3.063805721729274e-05, + "loss": 0.6999, + "step": 16724 + }, + { + "epoch": 2.9777421652421654, + "grad_norm": 0.8909936547279358, + "learning_rate": 3.062797493466212e-05, + "loss": 0.9325, + "step": 16725 + }, + { + "epoch": 2.9779202279202277, + "grad_norm": 0.836875319480896, + "learning_rate": 3.061789401123226e-05, + "loss": 0.9276, + "step": 16726 + }, + { + "epoch": 2.9780982905982905, + "grad_norm": 0.8965426683425903, + "learning_rate": 3.060781444720065e-05, + "loss": 0.7728, + "step": 16727 + }, + { + "epoch": 2.978276353276353, + "grad_norm": 0.8298100829124451, + "learning_rate": 3.059773624276475e-05, + "loss": 0.926, + "step": 16728 + }, + { + "epoch": 2.978454415954416, + "grad_norm": 0.9369875192642212, + "learning_rate": 3.058765939812204e-05, + "loss": 0.8598, + "step": 16729 + }, + { + "epoch": 2.9786324786324787, + "grad_norm": 1.019013524055481, + "learning_rate": 3.057758391346995e-05, + "loss": 0.9075, + "step": 16730 + }, + { + "epoch": 2.9788105413105415, + "grad_norm": 0.8597549200057983, + "learning_rate": 3.05675097890059e-05, + "loss": 0.8279, + "step": 16731 + }, + { + "epoch": 2.978988603988604, + "grad_norm": 0.8331323862075806, + "learning_rate": 3.055743702492726e-05, + "loss": 0.8366, + "step": 16732 + }, + { + "epoch": 2.9791666666666665, + "grad_norm": 0.7987640500068665, + "learning_rate": 3.054736562143135e-05, + "loss": 0.7618, + "step": 16733 + }, + { + "epoch": 2.9793447293447293, + "grad_norm": 0.84430330991745, + "learning_rate": 3.0537295578715606e-05, + "loss": 0.9791, + "step": 16734 + }, + { + "epoch": 2.979522792022792, + "grad_norm": 0.7811259627342224, + "learning_rate": 3.052722689697719e-05, + "loss": 0.7342, + "step": 16735 + }, + { + "epoch": 2.9797008547008548, + "grad_norm": 0.8311488032341003, + "learning_rate": 3.0517159576413477e-05, + "loss": 0.64, + "step": 16736 + }, + { + "epoch": 2.9798789173789175, + "grad_norm": 0.8865106105804443, + "learning_rate": 3.0507093617221683e-05, + "loss": 0.9196, + "step": 16737 + }, + { + "epoch": 2.98005698005698, + "grad_norm": 0.7360345125198364, + "learning_rate": 3.0497029019599033e-05, + "loss": 0.8709, + "step": 16738 + }, + { + "epoch": 2.9802350427350426, + "grad_norm": 0.8705546855926514, + "learning_rate": 3.0486965783742726e-05, + "loss": 0.7567, + "step": 16739 + }, + { + "epoch": 2.9804131054131053, + "grad_norm": 0.8622585535049438, + "learning_rate": 3.0476903909849908e-05, + "loss": 0.7909, + "step": 16740 + }, + { + "epoch": 2.980591168091168, + "grad_norm": 0.7877208590507507, + "learning_rate": 3.046684339811775e-05, + "loss": 0.7046, + "step": 16741 + }, + { + "epoch": 2.980769230769231, + "grad_norm": 0.915133535861969, + "learning_rate": 3.045678424874332e-05, + "loss": 0.9142, + "step": 16742 + }, + { + "epoch": 2.9809472934472936, + "grad_norm": 0.8539748191833496, + "learning_rate": 3.044672646192377e-05, + "loss": 0.9333, + "step": 16743 + }, + { + "epoch": 2.9811253561253563, + "grad_norm": 1.1402504444122314, + "learning_rate": 3.043667003785612e-05, + "loss": 0.8041, + "step": 16744 + }, + { + "epoch": 2.9813034188034186, + "grad_norm": 0.7355749607086182, + "learning_rate": 3.042661497673742e-05, + "loss": 0.8057, + "step": 16745 + }, + { + "epoch": 2.9814814814814814, + "grad_norm": 0.8524863719940186, + "learning_rate": 3.041656127876469e-05, + "loss": 0.7118, + "step": 16746 + }, + { + "epoch": 2.981659544159544, + "grad_norm": 0.9434519410133362, + "learning_rate": 3.040650894413487e-05, + "loss": 0.8104, + "step": 16747 + }, + { + "epoch": 2.981837606837607, + "grad_norm": 0.8716524839401245, + "learning_rate": 3.0396457973044923e-05, + "loss": 0.7502, + "step": 16748 + }, + { + "epoch": 2.9820156695156697, + "grad_norm": 0.942481279373169, + "learning_rate": 3.0386408365691855e-05, + "loss": 0.8506, + "step": 16749 + }, + { + "epoch": 2.982193732193732, + "grad_norm": 0.7921879291534424, + "learning_rate": 3.037636012227244e-05, + "loss": 0.7005, + "step": 16750 + }, + { + "epoch": 2.9823717948717947, + "grad_norm": 0.8415821194648743, + "learning_rate": 3.0366313242983645e-05, + "loss": 0.987, + "step": 16751 + }, + { + "epoch": 2.9825498575498575, + "grad_norm": 0.9556678533554077, + "learning_rate": 3.0356267728022293e-05, + "loss": 0.9198, + "step": 16752 + }, + { + "epoch": 2.98272792022792, + "grad_norm": 0.856810450553894, + "learning_rate": 3.0346223577585208e-05, + "loss": 0.9634, + "step": 16753 + }, + { + "epoch": 2.982905982905983, + "grad_norm": 0.8425500392913818, + "learning_rate": 3.0336180791869172e-05, + "loss": 0.8725, + "step": 16754 + }, + { + "epoch": 2.9830840455840457, + "grad_norm": 0.773210883140564, + "learning_rate": 3.032613937107096e-05, + "loss": 0.8363, + "step": 16755 + }, + { + "epoch": 2.9832621082621085, + "grad_norm": 0.7771822214126587, + "learning_rate": 3.0316099315387324e-05, + "loss": 0.8586, + "step": 16756 + }, + { + "epoch": 2.9834401709401708, + "grad_norm": 0.8261004686355591, + "learning_rate": 3.0306060625014954e-05, + "loss": 0.8185, + "step": 16757 + }, + { + "epoch": 2.9836182336182335, + "grad_norm": 1.074456810951233, + "learning_rate": 3.029602330015052e-05, + "loss": 0.9008, + "step": 16758 + }, + { + "epoch": 2.9837962962962963, + "grad_norm": 0.8160609006881714, + "learning_rate": 3.0285987340990774e-05, + "loss": 0.8494, + "step": 16759 + }, + { + "epoch": 2.983974358974359, + "grad_norm": 0.8296660780906677, + "learning_rate": 3.0275952747732227e-05, + "loss": 0.9089, + "step": 16760 + }, + { + "epoch": 2.984152421652422, + "grad_norm": 0.815726637840271, + "learning_rate": 3.0265919520571572e-05, + "loss": 0.8605, + "step": 16761 + }, + { + "epoch": 2.984330484330484, + "grad_norm": 0.7600420117378235, + "learning_rate": 3.0255887659705374e-05, + "loss": 0.6833, + "step": 16762 + }, + { + "epoch": 2.984508547008547, + "grad_norm": 0.8039379119873047, + "learning_rate": 3.024585716533014e-05, + "loss": 0.8156, + "step": 16763 + }, + { + "epoch": 2.9846866096866096, + "grad_norm": 0.8797104358673096, + "learning_rate": 3.02358280376425e-05, + "loss": 0.9965, + "step": 16764 + }, + { + "epoch": 2.9848646723646723, + "grad_norm": 0.8681818246841431, + "learning_rate": 3.0225800276838822e-05, + "loss": 0.8967, + "step": 16765 + }, + { + "epoch": 2.985042735042735, + "grad_norm": 0.804710328578949, + "learning_rate": 3.0215773883115706e-05, + "loss": 0.7424, + "step": 16766 + }, + { + "epoch": 2.985220797720798, + "grad_norm": 0.7346279621124268, + "learning_rate": 3.0205748856669467e-05, + "loss": 0.6986, + "step": 16767 + }, + { + "epoch": 2.9853988603988606, + "grad_norm": 0.8896076083183289, + "learning_rate": 3.0195725197696634e-05, + "loss": 0.8472, + "step": 16768 + }, + { + "epoch": 2.9855769230769234, + "grad_norm": 0.8281870484352112, + "learning_rate": 3.0185702906393555e-05, + "loss": 0.9873, + "step": 16769 + }, + { + "epoch": 2.9857549857549857, + "grad_norm": 0.7376837134361267, + "learning_rate": 3.0175681982956607e-05, + "loss": 1.0171, + "step": 16770 + }, + { + "epoch": 2.9859330484330484, + "grad_norm": 0.9346923828125, + "learning_rate": 3.016566242758212e-05, + "loss": 0.8742, + "step": 16771 + }, + { + "epoch": 2.986111111111111, + "grad_norm": 0.918557345867157, + "learning_rate": 3.0155644240466406e-05, + "loss": 1.0504, + "step": 16772 + }, + { + "epoch": 2.986289173789174, + "grad_norm": 0.8885583281517029, + "learning_rate": 3.0145627421805733e-05, + "loss": 0.7462, + "step": 16773 + }, + { + "epoch": 2.986467236467236, + "grad_norm": 0.8475548028945923, + "learning_rate": 3.013561197179644e-05, + "loss": 0.7353, + "step": 16774 + }, + { + "epoch": 2.986645299145299, + "grad_norm": 0.8925435543060303, + "learning_rate": 3.0125597890634626e-05, + "loss": 0.8462, + "step": 16775 + }, + { + "epoch": 2.9868233618233617, + "grad_norm": 0.923281729221344, + "learning_rate": 3.0115585178516648e-05, + "loss": 0.9725, + "step": 16776 + }, + { + "epoch": 2.9870014245014245, + "grad_norm": 0.9134986996650696, + "learning_rate": 3.0105573835638533e-05, + "loss": 0.9428, + "step": 16777 + }, + { + "epoch": 2.9871794871794872, + "grad_norm": 0.8284963369369507, + "learning_rate": 3.009556386219654e-05, + "loss": 0.9134, + "step": 16778 + }, + { + "epoch": 2.98735754985755, + "grad_norm": 0.9571327567100525, + "learning_rate": 3.0085555258386756e-05, + "loss": 0.8407, + "step": 16779 + }, + { + "epoch": 2.9875356125356127, + "grad_norm": 0.8699741363525391, + "learning_rate": 3.0075548024405254e-05, + "loss": 0.862, + "step": 16780 + }, + { + "epoch": 2.9877136752136755, + "grad_norm": 0.9462196826934814, + "learning_rate": 3.006554216044819e-05, + "loss": 0.8936, + "step": 16781 + }, + { + "epoch": 2.987891737891738, + "grad_norm": 0.9415904879570007, + "learning_rate": 3.0055537666711496e-05, + "loss": 1.0089, + "step": 16782 + }, + { + "epoch": 2.9880698005698005, + "grad_norm": 0.8529258370399475, + "learning_rate": 3.0045534543391275e-05, + "loss": 0.7896, + "step": 16783 + }, + { + "epoch": 2.9882478632478633, + "grad_norm": 0.7967036962509155, + "learning_rate": 3.0035532790683486e-05, + "loss": 0.7661, + "step": 16784 + }, + { + "epoch": 2.988425925925926, + "grad_norm": 1.087546944618225, + "learning_rate": 3.0025532408784097e-05, + "loss": 0.7232, + "step": 16785 + }, + { + "epoch": 2.9886039886039883, + "grad_norm": 1.0032312870025635, + "learning_rate": 3.001553339788903e-05, + "loss": 0.8462, + "step": 16786 + }, + { + "epoch": 2.988782051282051, + "grad_norm": 0.7726916074752808, + "learning_rate": 3.0005535758194216e-05, + "loss": 0.8318, + "step": 16787 + }, + { + "epoch": 2.988960113960114, + "grad_norm": 0.8814677000045776, + "learning_rate": 2.999553948989552e-05, + "loss": 0.8058, + "step": 16788 + }, + { + "epoch": 2.9891381766381766, + "grad_norm": 0.8866276741027832, + "learning_rate": 2.9985544593188818e-05, + "loss": 0.8576, + "step": 16789 + }, + { + "epoch": 2.9893162393162394, + "grad_norm": 0.9877329468727112, + "learning_rate": 2.997555106826988e-05, + "loss": 0.8675, + "step": 16790 + }, + { + "epoch": 2.989494301994302, + "grad_norm": 0.842682421207428, + "learning_rate": 2.9965558915334636e-05, + "loss": 1.094, + "step": 16791 + }, + { + "epoch": 2.989672364672365, + "grad_norm": 0.8740942478179932, + "learning_rate": 2.9955568134578703e-05, + "loss": 0.904, + "step": 16792 + }, + { + "epoch": 2.9898504273504276, + "grad_norm": 0.9166956543922424, + "learning_rate": 2.9945578726197944e-05, + "loss": 1.0695, + "step": 16793 + }, + { + "epoch": 2.99002849002849, + "grad_norm": 0.7124375104904175, + "learning_rate": 2.993559069038805e-05, + "loss": 0.6606, + "step": 16794 + }, + { + "epoch": 2.9902065527065527, + "grad_norm": 0.8711931109428406, + "learning_rate": 2.99256040273447e-05, + "loss": 0.9049, + "step": 16795 + }, + { + "epoch": 2.9903846153846154, + "grad_norm": 0.8475961089134216, + "learning_rate": 2.9915618737263584e-05, + "loss": 0.8487, + "step": 16796 + }, + { + "epoch": 2.990562678062678, + "grad_norm": 0.8962427377700806, + "learning_rate": 2.9905634820340324e-05, + "loss": 0.9318, + "step": 16797 + }, + { + "epoch": 2.9907407407407405, + "grad_norm": 0.9120275378227234, + "learning_rate": 2.9895652276770514e-05, + "loss": 0.8787, + "step": 16798 + }, + { + "epoch": 2.9909188034188032, + "grad_norm": 0.8273578882217407, + "learning_rate": 2.9885671106749822e-05, + "loss": 0.8566, + "step": 16799 + }, + { + "epoch": 2.991096866096866, + "grad_norm": 0.8050937056541443, + "learning_rate": 2.9875691310473697e-05, + "loss": 1.1276, + "step": 16800 + }, + { + "epoch": 2.9912749287749287, + "grad_norm": 0.8434747457504272, + "learning_rate": 2.9865712888137766e-05, + "loss": 0.6286, + "step": 16801 + }, + { + "epoch": 2.9914529914529915, + "grad_norm": 0.7851191759109497, + "learning_rate": 2.9855735839937493e-05, + "loss": 0.8116, + "step": 16802 + }, + { + "epoch": 2.9916310541310542, + "grad_norm": 0.8858240842819214, + "learning_rate": 2.984576016606837e-05, + "loss": 0.8663, + "step": 16803 + }, + { + "epoch": 2.991809116809117, + "grad_norm": 0.8688439726829529, + "learning_rate": 2.9835785866725842e-05, + "loss": 0.9229, + "step": 16804 + }, + { + "epoch": 2.9919871794871797, + "grad_norm": 0.7873746752738953, + "learning_rate": 2.9825812942105302e-05, + "loss": 0.779, + "step": 16805 + }, + { + "epoch": 2.992165242165242, + "grad_norm": 0.9577028751373291, + "learning_rate": 2.9815841392402255e-05, + "loss": 0.7068, + "step": 16806 + }, + { + "epoch": 2.992343304843305, + "grad_norm": 0.8219005465507507, + "learning_rate": 2.980587121781193e-05, + "loss": 0.7712, + "step": 16807 + }, + { + "epoch": 2.9925213675213675, + "grad_norm": 0.9321123957633972, + "learning_rate": 2.9795902418529776e-05, + "loss": 0.9309, + "step": 16808 + }, + { + "epoch": 2.9926994301994303, + "grad_norm": 0.8010317087173462, + "learning_rate": 2.9785934994751074e-05, + "loss": 0.8973, + "step": 16809 + }, + { + "epoch": 2.9928774928774926, + "grad_norm": 0.8819838762283325, + "learning_rate": 2.9775968946671117e-05, + "loss": 0.6899, + "step": 16810 + }, + { + "epoch": 2.9930555555555554, + "grad_norm": 0.8337511420249939, + "learning_rate": 2.976600427448518e-05, + "loss": 0.8005, + "step": 16811 + }, + { + "epoch": 2.993233618233618, + "grad_norm": 1.0560593605041504, + "learning_rate": 2.975604097838849e-05, + "loss": 0.8804, + "step": 16812 + }, + { + "epoch": 2.993411680911681, + "grad_norm": 0.8482444882392883, + "learning_rate": 2.9746079058576247e-05, + "loss": 0.8606, + "step": 16813 + }, + { + "epoch": 2.9935897435897436, + "grad_norm": 0.8289902210235596, + "learning_rate": 2.973611851524366e-05, + "loss": 0.8163, + "step": 16814 + }, + { + "epoch": 2.9937678062678064, + "grad_norm": 0.780939519405365, + "learning_rate": 2.9726159348585826e-05, + "loss": 0.7041, + "step": 16815 + }, + { + "epoch": 2.993945868945869, + "grad_norm": 0.8907291889190674, + "learning_rate": 2.9716201558797984e-05, + "loss": 0.6484, + "step": 16816 + }, + { + "epoch": 2.994123931623932, + "grad_norm": 0.9351929426193237, + "learning_rate": 2.9706245146075118e-05, + "loss": 0.7232, + "step": 16817 + }, + { + "epoch": 2.994301994301994, + "grad_norm": 0.934117317199707, + "learning_rate": 2.969629011061238e-05, + "loss": 0.9189, + "step": 16818 + }, + { + "epoch": 2.994480056980057, + "grad_norm": 0.8751780390739441, + "learning_rate": 2.968633645260479e-05, + "loss": 0.9695, + "step": 16819 + }, + { + "epoch": 2.9946581196581197, + "grad_norm": 0.921024739742279, + "learning_rate": 2.967638417224734e-05, + "loss": 0.9011, + "step": 16820 + }, + { + "epoch": 2.9948361823361824, + "grad_norm": 0.7881745100021362, + "learning_rate": 2.9666433269735126e-05, + "loss": 0.8001, + "step": 16821 + }, + { + "epoch": 2.995014245014245, + "grad_norm": 0.8107823729515076, + "learning_rate": 2.9656483745262985e-05, + "loss": 0.9239, + "step": 16822 + }, + { + "epoch": 2.9951923076923075, + "grad_norm": 0.8491915464401245, + "learning_rate": 2.964653559902595e-05, + "loss": 0.8193, + "step": 16823 + }, + { + "epoch": 2.9953703703703702, + "grad_norm": 0.9521864056587219, + "learning_rate": 2.9636588831218904e-05, + "loss": 1.0745, + "step": 16824 + }, + { + "epoch": 2.995548433048433, + "grad_norm": 0.9784126281738281, + "learning_rate": 2.9626643442036727e-05, + "loss": 0.9899, + "step": 16825 + }, + { + "epoch": 2.9957264957264957, + "grad_norm": 0.8035039305686951, + "learning_rate": 2.961669943167429e-05, + "loss": 0.9172, + "step": 16826 + }, + { + "epoch": 2.9959045584045585, + "grad_norm": 0.9292920827865601, + "learning_rate": 2.9606756800326408e-05, + "loss": 0.9228, + "step": 16827 + }, + { + "epoch": 2.9960826210826212, + "grad_norm": 0.9243139028549194, + "learning_rate": 2.9596815548187908e-05, + "loss": 0.8698, + "step": 16828 + }, + { + "epoch": 2.996260683760684, + "grad_norm": 0.9847014546394348, + "learning_rate": 2.958687567545355e-05, + "loss": 0.8885, + "step": 16829 + }, + { + "epoch": 2.9964387464387463, + "grad_norm": 0.9667131900787354, + "learning_rate": 2.9576937182318078e-05, + "loss": 0.7951, + "step": 16830 + }, + { + "epoch": 2.996616809116809, + "grad_norm": 0.8431822061538696, + "learning_rate": 2.956700006897628e-05, + "loss": 1.0645, + "step": 16831 + }, + { + "epoch": 2.996794871794872, + "grad_norm": 0.7381225228309631, + "learning_rate": 2.955706433562274e-05, + "loss": 0.5631, + "step": 16832 + }, + { + "epoch": 2.9969729344729346, + "grad_norm": 0.7975440621376038, + "learning_rate": 2.9547129982452228e-05, + "loss": 0.8324, + "step": 16833 + }, + { + "epoch": 2.9971509971509973, + "grad_norm": 0.9004024863243103, + "learning_rate": 2.9537197009659344e-05, + "loss": 1.045, + "step": 16834 + }, + { + "epoch": 2.9973290598290596, + "grad_norm": 0.786651074886322, + "learning_rate": 2.952726541743871e-05, + "loss": 0.7569, + "step": 16835 + }, + { + "epoch": 2.9975071225071224, + "grad_norm": 0.8053869605064392, + "learning_rate": 2.951733520598492e-05, + "loss": 0.8327, + "step": 16836 + }, + { + "epoch": 2.997685185185185, + "grad_norm": 0.8725607991218567, + "learning_rate": 2.9507406375492543e-05, + "loss": 0.7702, + "step": 16837 + }, + { + "epoch": 2.997863247863248, + "grad_norm": 0.9187145233154297, + "learning_rate": 2.9497478926156087e-05, + "loss": 0.8946, + "step": 16838 + }, + { + "epoch": 2.9980413105413106, + "grad_norm": 0.9324785470962524, + "learning_rate": 2.9487552858170076e-05, + "loss": 0.8669, + "step": 16839 + }, + { + "epoch": 2.9982193732193734, + "grad_norm": 0.8439409136772156, + "learning_rate": 2.9477628171728964e-05, + "loss": 0.7187, + "step": 16840 + }, + { + "epoch": 2.998397435897436, + "grad_norm": 0.7890669703483582, + "learning_rate": 2.9467704867027258e-05, + "loss": 0.7856, + "step": 16841 + }, + { + "epoch": 2.9985754985754984, + "grad_norm": 0.7931693196296692, + "learning_rate": 2.9457782944259362e-05, + "loss": 0.7212, + "step": 16842 + }, + { + "epoch": 2.998753561253561, + "grad_norm": 0.8256354331970215, + "learning_rate": 2.9447862403619665e-05, + "loss": 0.8255, + "step": 16843 + }, + { + "epoch": 2.998931623931624, + "grad_norm": 0.7902492880821228, + "learning_rate": 2.9437943245302547e-05, + "loss": 0.8298, + "step": 16844 + }, + { + "epoch": 2.9991096866096867, + "grad_norm": 1.0743845701217651, + "learning_rate": 2.9428025469502318e-05, + "loss": 0.92, + "step": 16845 + }, + { + "epoch": 2.9992877492877494, + "grad_norm": 0.9262487292289734, + "learning_rate": 2.9418109076413404e-05, + "loss": 0.9108, + "step": 16846 + }, + { + "epoch": 2.9994658119658117, + "grad_norm": 0.855722963809967, + "learning_rate": 2.9408194066229944e-05, + "loss": 0.8596, + "step": 16847 + }, + { + "epoch": 2.9996438746438745, + "grad_norm": 0.8279969692230225, + "learning_rate": 2.939828043914634e-05, + "loss": 0.8913, + "step": 16848 + }, + { + "epoch": 2.9996438746438745, + "eval_loss": 1.1308976411819458, + "eval_runtime": 24.7916, + "eval_samples_per_second": 41.99, + "eval_steps_per_second": 21.015, + "step": 16848 + }, + { + "epoch": 2.9998219373219372, + "grad_norm": 0.8992627263069153, + "learning_rate": 2.9388368195356718e-05, + "loss": 1.0347, + "step": 16849 + }, + { + "epoch": 3.0, + "grad_norm": 0.8295336961746216, + "learning_rate": 2.9378457335055364e-05, + "loss": 0.797, + "step": 16850 + }, + { + "epoch": 3.0001780626780628, + "grad_norm": 0.9085283875465393, + "learning_rate": 2.9368547858436434e-05, + "loss": 0.7888, + "step": 16851 + }, + { + "epoch": 3.0001780626780628, + "grad_norm": 0.7804644107818604, + "learning_rate": 2.9358639765694085e-05, + "loss": 0.6619, + "step": 16852 + }, + { + "epoch": 3.0003561253561255, + "grad_norm": 0.7937109470367432, + "learning_rate": 2.9348733057022447e-05, + "loss": 0.6769, + "step": 16853 + }, + { + "epoch": 3.0005341880341883, + "grad_norm": 0.8836156725883484, + "learning_rate": 2.9338827732615627e-05, + "loss": 0.8822, + "step": 16854 + }, + { + "epoch": 3.0007122507122506, + "grad_norm": 0.7436475157737732, + "learning_rate": 2.932892379266765e-05, + "loss": 0.8601, + "step": 16855 + }, + { + "epoch": 3.0008903133903133, + "grad_norm": 0.825633704662323, + "learning_rate": 2.9319021237372668e-05, + "loss": 0.8905, + "step": 16856 + }, + { + "epoch": 3.001068376068376, + "grad_norm": 0.7219718098640442, + "learning_rate": 2.930912006692458e-05, + "loss": 0.8757, + "step": 16857 + }, + { + "epoch": 3.001246438746439, + "grad_norm": 0.8130925893783569, + "learning_rate": 2.929922028151746e-05, + "loss": 0.4952, + "step": 16858 + }, + { + "epoch": 3.0014245014245016, + "grad_norm": 0.6775509119033813, + "learning_rate": 2.9289321881345254e-05, + "loss": 0.5514, + "step": 16859 + }, + { + "epoch": 3.0016025641025643, + "grad_norm": 0.7436477541923523, + "learning_rate": 2.92794248666019e-05, + "loss": 0.9079, + "step": 16860 + }, + { + "epoch": 3.0017806267806266, + "grad_norm": 0.7549178600311279, + "learning_rate": 2.9269529237481307e-05, + "loss": 0.7723, + "step": 16861 + }, + { + "epoch": 3.0019586894586894, + "grad_norm": 0.8569880127906799, + "learning_rate": 2.9259634994177322e-05, + "loss": 0.8593, + "step": 16862 + }, + { + "epoch": 3.002136752136752, + "grad_norm": 0.7354894280433655, + "learning_rate": 2.9249742136883906e-05, + "loss": 0.7426, + "step": 16863 + }, + { + "epoch": 3.002314814814815, + "grad_norm": 0.8491479158401489, + "learning_rate": 2.9239850665794755e-05, + "loss": 0.8501, + "step": 16864 + }, + { + "epoch": 3.0024928774928776, + "grad_norm": 0.7588241696357727, + "learning_rate": 2.9229960581103776e-05, + "loss": 0.8064, + "step": 16865 + }, + { + "epoch": 3.0026709401709404, + "grad_norm": 0.7409141659736633, + "learning_rate": 2.9220071883004708e-05, + "loss": 0.7756, + "step": 16866 + }, + { + "epoch": 3.0028490028490027, + "grad_norm": 0.8201326131820679, + "learning_rate": 2.921018457169129e-05, + "loss": 0.9796, + "step": 16867 + }, + { + "epoch": 3.0030270655270654, + "grad_norm": 0.7816381454467773, + "learning_rate": 2.920029864735726e-05, + "loss": 0.8362, + "step": 16868 + }, + { + "epoch": 3.003205128205128, + "grad_norm": 0.9872122406959534, + "learning_rate": 2.9190414110196295e-05, + "loss": 0.9468, + "step": 16869 + }, + { + "epoch": 3.003383190883191, + "grad_norm": 0.9007967114448547, + "learning_rate": 2.918053096040205e-05, + "loss": 0.7311, + "step": 16870 + }, + { + "epoch": 3.0035612535612537, + "grad_norm": 0.8848899006843567, + "learning_rate": 2.9170649198168243e-05, + "loss": 0.6342, + "step": 16871 + }, + { + "epoch": 3.0037393162393164, + "grad_norm": 0.6995126605033875, + "learning_rate": 2.9160768823688377e-05, + "loss": 0.6895, + "step": 16872 + }, + { + "epoch": 3.0039173789173788, + "grad_norm": 0.7542599439620972, + "learning_rate": 2.9150889837156135e-05, + "loss": 0.7411, + "step": 16873 + }, + { + "epoch": 3.0040954415954415, + "grad_norm": 0.8360390663146973, + "learning_rate": 2.9141012238764985e-05, + "loss": 0.7273, + "step": 16874 + }, + { + "epoch": 3.0042735042735043, + "grad_norm": 0.8042497038841248, + "learning_rate": 2.9131136028708526e-05, + "loss": 0.6749, + "step": 16875 + }, + { + "epoch": 3.004451566951567, + "grad_norm": 0.7112155556678772, + "learning_rate": 2.912126120718024e-05, + "loss": 0.7799, + "step": 16876 + }, + { + "epoch": 3.0046296296296298, + "grad_norm": 0.7784724831581116, + "learning_rate": 2.9111387774373612e-05, + "loss": 0.7181, + "step": 16877 + }, + { + "epoch": 3.0048076923076925, + "grad_norm": 0.8795644044876099, + "learning_rate": 2.9101515730482064e-05, + "loss": 0.8351, + "step": 16878 + }, + { + "epoch": 3.004985754985755, + "grad_norm": 0.7584426999092102, + "learning_rate": 2.9091645075699014e-05, + "loss": 0.7506, + "step": 16879 + }, + { + "epoch": 3.0051638176638176, + "grad_norm": 0.881862461566925, + "learning_rate": 2.9081775810217915e-05, + "loss": 0.7594, + "step": 16880 + }, + { + "epoch": 3.0053418803418803, + "grad_norm": 0.8785306215286255, + "learning_rate": 2.90719079342321e-05, + "loss": 0.701, + "step": 16881 + }, + { + "epoch": 3.005519943019943, + "grad_norm": 0.749655544757843, + "learning_rate": 2.9062041447934908e-05, + "loss": 0.7694, + "step": 16882 + }, + { + "epoch": 3.005698005698006, + "grad_norm": 0.8743281364440918, + "learning_rate": 2.9052176351519646e-05, + "loss": 0.9534, + "step": 16883 + }, + { + "epoch": 3.0058760683760686, + "grad_norm": 0.7728061079978943, + "learning_rate": 2.9042312645179613e-05, + "loss": 0.6966, + "step": 16884 + }, + { + "epoch": 3.006054131054131, + "grad_norm": 0.8588292002677917, + "learning_rate": 2.9032450329108052e-05, + "loss": 0.6174, + "step": 16885 + }, + { + "epoch": 3.0062321937321936, + "grad_norm": 0.8733890056610107, + "learning_rate": 2.9022589403498214e-05, + "loss": 0.5899, + "step": 16886 + }, + { + "epoch": 3.0064102564102564, + "grad_norm": 0.8324486613273621, + "learning_rate": 2.9012729868543264e-05, + "loss": 0.7532, + "step": 16887 + }, + { + "epoch": 3.006588319088319, + "grad_norm": 0.7193008661270142, + "learning_rate": 2.9002871724436465e-05, + "loss": 0.6136, + "step": 16888 + }, + { + "epoch": 3.006766381766382, + "grad_norm": 0.7795392870903015, + "learning_rate": 2.899301497137086e-05, + "loss": 0.8326, + "step": 16889 + }, + { + "epoch": 3.0069444444444446, + "grad_norm": 0.7805325984954834, + "learning_rate": 2.8983159609539635e-05, + "loss": 0.6532, + "step": 16890 + }, + { + "epoch": 3.007122507122507, + "grad_norm": 0.78617262840271, + "learning_rate": 2.8973305639135883e-05, + "loss": 0.5964, + "step": 16891 + }, + { + "epoch": 3.0073005698005697, + "grad_norm": 0.8331096172332764, + "learning_rate": 2.8963453060352662e-05, + "loss": 0.7982, + "step": 16892 + }, + { + "epoch": 3.0074786324786325, + "grad_norm": 0.8765615820884705, + "learning_rate": 2.8953601873383017e-05, + "loss": 0.8202, + "step": 16893 + }, + { + "epoch": 3.007656695156695, + "grad_norm": 0.8578206896781921, + "learning_rate": 2.894375207841995e-05, + "loss": 0.5168, + "step": 16894 + }, + { + "epoch": 3.007834757834758, + "grad_norm": 0.8777512311935425, + "learning_rate": 2.893390367565646e-05, + "loss": 0.7154, + "step": 16895 + }, + { + "epoch": 3.0080128205128207, + "grad_norm": 0.8647478222846985, + "learning_rate": 2.8924056665285494e-05, + "loss": 0.6488, + "step": 16896 + }, + { + "epoch": 3.008190883190883, + "grad_norm": 0.7841095924377441, + "learning_rate": 2.8914211047499963e-05, + "loss": 0.8219, + "step": 16897 + }, + { + "epoch": 3.0083689458689458, + "grad_norm": 0.9699451923370361, + "learning_rate": 2.8904366822492855e-05, + "loss": 0.6815, + "step": 16898 + }, + { + "epoch": 3.0085470085470085, + "grad_norm": 0.8264166116714478, + "learning_rate": 2.8894523990456946e-05, + "loss": 0.8515, + "step": 16899 + }, + { + "epoch": 3.0087250712250713, + "grad_norm": 0.8300934433937073, + "learning_rate": 2.888468255158514e-05, + "loss": 0.6219, + "step": 16900 + }, + { + "epoch": 3.008903133903134, + "grad_norm": 0.8727999925613403, + "learning_rate": 2.8874842506070265e-05, + "loss": 0.8451, + "step": 16901 + }, + { + "epoch": 3.0090811965811968, + "grad_norm": 0.8788085579872131, + "learning_rate": 2.8865003854105066e-05, + "loss": 0.8576, + "step": 16902 + }, + { + "epoch": 3.009259259259259, + "grad_norm": 0.8885243535041809, + "learning_rate": 2.885516659588241e-05, + "loss": 0.7011, + "step": 16903 + }, + { + "epoch": 3.009437321937322, + "grad_norm": 0.8000853061676025, + "learning_rate": 2.8845330731594898e-05, + "loss": 0.7582, + "step": 16904 + }, + { + "epoch": 3.0096153846153846, + "grad_norm": 0.9916318655014038, + "learning_rate": 2.883549626143537e-05, + "loss": 0.8114, + "step": 16905 + }, + { + "epoch": 3.0097934472934473, + "grad_norm": 0.9104628562927246, + "learning_rate": 2.882566318559645e-05, + "loss": 1.059, + "step": 16906 + }, + { + "epoch": 3.00997150997151, + "grad_norm": 0.8790180683135986, + "learning_rate": 2.8815831504270807e-05, + "loss": 0.8905, + "step": 16907 + }, + { + "epoch": 3.010149572649573, + "grad_norm": 0.954405665397644, + "learning_rate": 2.880600121765108e-05, + "loss": 0.9691, + "step": 16908 + }, + { + "epoch": 3.010327635327635, + "grad_norm": 0.8337442874908447, + "learning_rate": 2.879617232592986e-05, + "loss": 0.6808, + "step": 16909 + }, + { + "epoch": 3.010505698005698, + "grad_norm": 0.9617137908935547, + "learning_rate": 2.878634482929974e-05, + "loss": 0.7545, + "step": 16910 + }, + { + "epoch": 3.0106837606837606, + "grad_norm": 0.8164490461349487, + "learning_rate": 2.877651872795325e-05, + "loss": 0.7219, + "step": 16911 + }, + { + "epoch": 3.0108618233618234, + "grad_norm": 0.7971426248550415, + "learning_rate": 2.8766694022082895e-05, + "loss": 0.7519, + "step": 16912 + }, + { + "epoch": 3.011039886039886, + "grad_norm": 0.8035764694213867, + "learning_rate": 2.8756870711881255e-05, + "loss": 0.5414, + "step": 16913 + }, + { + "epoch": 3.011217948717949, + "grad_norm": 0.851365864276886, + "learning_rate": 2.8747048797540676e-05, + "loss": 0.8258, + "step": 16914 + }, + { + "epoch": 3.011396011396011, + "grad_norm": 0.9337644577026367, + "learning_rate": 2.87372282792537e-05, + "loss": 0.7766, + "step": 16915 + }, + { + "epoch": 3.011574074074074, + "grad_norm": 0.805377721786499, + "learning_rate": 2.8727409157212693e-05, + "loss": 0.5635, + "step": 16916 + }, + { + "epoch": 3.0117521367521367, + "grad_norm": 0.834541916847229, + "learning_rate": 2.8717591431610046e-05, + "loss": 0.6602, + "step": 16917 + }, + { + "epoch": 3.0119301994301995, + "grad_norm": 0.7747098207473755, + "learning_rate": 2.8707775102638124e-05, + "loss": 0.6742, + "step": 16918 + }, + { + "epoch": 3.012108262108262, + "grad_norm": 0.8491533398628235, + "learning_rate": 2.8697960170489226e-05, + "loss": 0.7951, + "step": 16919 + }, + { + "epoch": 3.012286324786325, + "grad_norm": 0.8623409867286682, + "learning_rate": 2.8688146635355727e-05, + "loss": 0.7455, + "step": 16920 + }, + { + "epoch": 3.0124643874643873, + "grad_norm": 0.8398233652114868, + "learning_rate": 2.8678334497429804e-05, + "loss": 0.8699, + "step": 16921 + }, + { + "epoch": 3.01264245014245, + "grad_norm": 0.8916721940040588, + "learning_rate": 2.8668523756903798e-05, + "loss": 0.7295, + "step": 16922 + }, + { + "epoch": 3.0128205128205128, + "grad_norm": 0.8889368772506714, + "learning_rate": 2.865871441396988e-05, + "loss": 0.7543, + "step": 16923 + }, + { + "epoch": 3.0129985754985755, + "grad_norm": 0.8547541499137878, + "learning_rate": 2.8648906468820257e-05, + "loss": 0.7837, + "step": 16924 + }, + { + "epoch": 3.0131766381766383, + "grad_norm": 0.7525152564048767, + "learning_rate": 2.8639099921647096e-05, + "loss": 0.8279, + "step": 16925 + }, + { + "epoch": 3.013354700854701, + "grad_norm": 0.8209140300750732, + "learning_rate": 2.8629294772642523e-05, + "loss": 0.8206, + "step": 16926 + }, + { + "epoch": 3.0135327635327633, + "grad_norm": 0.8411595821380615, + "learning_rate": 2.8619491021998633e-05, + "loss": 0.7606, + "step": 16927 + }, + { + "epoch": 3.013710826210826, + "grad_norm": 0.8703480362892151, + "learning_rate": 2.860968866990761e-05, + "loss": 0.8538, + "step": 16928 + }, + { + "epoch": 3.013888888888889, + "grad_norm": 0.9207778573036194, + "learning_rate": 2.8599887716561357e-05, + "loss": 0.7485, + "step": 16929 + }, + { + "epoch": 3.0140669515669516, + "grad_norm": 0.8924398422241211, + "learning_rate": 2.8590088162152016e-05, + "loss": 0.9015, + "step": 16930 + }, + { + "epoch": 3.0142450142450143, + "grad_norm": 0.8561375141143799, + "learning_rate": 2.858029000687157e-05, + "loss": 0.7663, + "step": 16931 + }, + { + "epoch": 3.014423076923077, + "grad_norm": 0.9084187746047974, + "learning_rate": 2.8570493250911967e-05, + "loss": 0.8277, + "step": 16932 + }, + { + "epoch": 3.0146011396011394, + "grad_norm": 0.9076225757598877, + "learning_rate": 2.856069789446517e-05, + "loss": 0.7972, + "step": 16933 + }, + { + "epoch": 3.014779202279202, + "grad_norm": 0.9396339058876038, + "learning_rate": 2.8550903937723104e-05, + "loss": 0.6968, + "step": 16934 + }, + { + "epoch": 3.014957264957265, + "grad_norm": 0.8107509613037109, + "learning_rate": 2.854111138087765e-05, + "loss": 0.6931, + "step": 16935 + }, + { + "epoch": 3.0151353276353277, + "grad_norm": 0.908308207988739, + "learning_rate": 2.8531320224120663e-05, + "loss": 0.8363, + "step": 16936 + }, + { + "epoch": 3.0153133903133904, + "grad_norm": 0.8976947665214539, + "learning_rate": 2.8521530467643976e-05, + "loss": 0.7907, + "step": 16937 + }, + { + "epoch": 3.015491452991453, + "grad_norm": 0.8451889753341675, + "learning_rate": 2.8511742111639473e-05, + "loss": 0.8333, + "step": 16938 + }, + { + "epoch": 3.0156695156695155, + "grad_norm": 0.8719599843025208, + "learning_rate": 2.8501955156298808e-05, + "loss": 0.8047, + "step": 16939 + }, + { + "epoch": 3.015847578347578, + "grad_norm": 0.7666802406311035, + "learning_rate": 2.8492169601813856e-05, + "loss": 0.6434, + "step": 16940 + }, + { + "epoch": 3.016025641025641, + "grad_norm": 0.9386308789253235, + "learning_rate": 2.848238544837628e-05, + "loss": 0.7351, + "step": 16941 + }, + { + "epoch": 3.0162037037037037, + "grad_norm": 0.9501772522926331, + "learning_rate": 2.8472602696177773e-05, + "loss": 0.7057, + "step": 16942 + }, + { + "epoch": 3.0163817663817665, + "grad_norm": 1.0069975852966309, + "learning_rate": 2.8462821345410097e-05, + "loss": 0.8611, + "step": 16943 + }, + { + "epoch": 3.0165598290598292, + "grad_norm": 0.9124653935432434, + "learning_rate": 2.8453041396264756e-05, + "loss": 0.6097, + "step": 16944 + }, + { + "epoch": 3.0167378917378915, + "grad_norm": 0.8535616993904114, + "learning_rate": 2.8443262848933515e-05, + "loss": 0.7971, + "step": 16945 + }, + { + "epoch": 3.0169159544159543, + "grad_norm": 0.9309000968933105, + "learning_rate": 2.843348570360783e-05, + "loss": 0.8864, + "step": 16946 + }, + { + "epoch": 3.017094017094017, + "grad_norm": 0.8818594813346863, + "learning_rate": 2.842370996047935e-05, + "loss": 0.7748, + "step": 16947 + }, + { + "epoch": 3.01727207977208, + "grad_norm": 1.034568428993225, + "learning_rate": 2.8413935619739596e-05, + "loss": 0.6979, + "step": 16948 + }, + { + "epoch": 3.0174501424501425, + "grad_norm": 0.9504423141479492, + "learning_rate": 2.840416268158006e-05, + "loss": 0.7447, + "step": 16949 + }, + { + "epoch": 3.0176282051282053, + "grad_norm": 0.9099027514457703, + "learning_rate": 2.8394391146192234e-05, + "loss": 0.8136, + "step": 16950 + }, + { + "epoch": 3.0178062678062676, + "grad_norm": 0.8165521025657654, + "learning_rate": 2.8384621013767565e-05, + "loss": 0.7306, + "step": 16951 + }, + { + "epoch": 3.0179843304843303, + "grad_norm": 0.832309901714325, + "learning_rate": 2.8374852284497446e-05, + "loss": 0.6333, + "step": 16952 + }, + { + "epoch": 3.018162393162393, + "grad_norm": 0.8041273951530457, + "learning_rate": 2.8365084958573363e-05, + "loss": 0.6899, + "step": 16953 + }, + { + "epoch": 3.018340455840456, + "grad_norm": 0.9186970591545105, + "learning_rate": 2.8355319036186577e-05, + "loss": 0.7873, + "step": 16954 + }, + { + "epoch": 3.0185185185185186, + "grad_norm": 0.8950297832489014, + "learning_rate": 2.834555451752855e-05, + "loss": 0.945, + "step": 16955 + }, + { + "epoch": 3.0186965811965814, + "grad_norm": 1.0703802108764648, + "learning_rate": 2.8335791402790468e-05, + "loss": 0.8377, + "step": 16956 + }, + { + "epoch": 3.0188746438746437, + "grad_norm": 0.9604542851448059, + "learning_rate": 2.8326029692163712e-05, + "loss": 0.8197, + "step": 16957 + }, + { + "epoch": 3.0190527065527064, + "grad_norm": 0.9822934865951538, + "learning_rate": 2.8316269385839524e-05, + "loss": 0.839, + "step": 16958 + }, + { + "epoch": 3.019230769230769, + "grad_norm": 0.945202648639679, + "learning_rate": 2.8306510484009085e-05, + "loss": 0.888, + "step": 16959 + }, + { + "epoch": 3.019408831908832, + "grad_norm": 0.9367056488990784, + "learning_rate": 2.8296752986863706e-05, + "loss": 0.8231, + "step": 16960 + }, + { + "epoch": 3.0195868945868947, + "grad_norm": 0.8301896452903748, + "learning_rate": 2.828699689459444e-05, + "loss": 0.7208, + "step": 16961 + }, + { + "epoch": 3.0197649572649574, + "grad_norm": 1.0018302202224731, + "learning_rate": 2.827724220739254e-05, + "loss": 0.9264, + "step": 16962 + }, + { + "epoch": 3.0199430199430197, + "grad_norm": 0.8947141170501709, + "learning_rate": 2.8267488925449083e-05, + "loss": 0.7169, + "step": 16963 + }, + { + "epoch": 3.0201210826210825, + "grad_norm": 0.914278507232666, + "learning_rate": 2.8257737048955167e-05, + "loss": 0.7964, + "step": 16964 + }, + { + "epoch": 3.0202991452991452, + "grad_norm": 0.8575536012649536, + "learning_rate": 2.824798657810186e-05, + "loss": 0.7144, + "step": 16965 + }, + { + "epoch": 3.020477207977208, + "grad_norm": 0.8627474904060364, + "learning_rate": 2.82382375130802e-05, + "loss": 0.6284, + "step": 16966 + }, + { + "epoch": 3.0206552706552707, + "grad_norm": 0.9590699076652527, + "learning_rate": 2.822848985408122e-05, + "loss": 0.6919, + "step": 16967 + }, + { + "epoch": 3.0208333333333335, + "grad_norm": 0.8014170527458191, + "learning_rate": 2.821874360129587e-05, + "loss": 0.7017, + "step": 16968 + }, + { + "epoch": 3.021011396011396, + "grad_norm": 0.9732417464256287, + "learning_rate": 2.8208998754915107e-05, + "loss": 0.9154, + "step": 16969 + }, + { + "epoch": 3.0211894586894585, + "grad_norm": 1.035178780555725, + "learning_rate": 2.8199255315129934e-05, + "loss": 0.7987, + "step": 16970 + }, + { + "epoch": 3.0213675213675213, + "grad_norm": 0.9089648723602295, + "learning_rate": 2.818951328213115e-05, + "loss": 0.7842, + "step": 16971 + }, + { + "epoch": 3.021545584045584, + "grad_norm": 0.8718148469924927, + "learning_rate": 2.8179772656109702e-05, + "loss": 0.5428, + "step": 16972 + }, + { + "epoch": 3.021723646723647, + "grad_norm": 0.9230215549468994, + "learning_rate": 2.8170033437256415e-05, + "loss": 0.8986, + "step": 16973 + }, + { + "epoch": 3.0219017094017095, + "grad_norm": 0.7892549633979797, + "learning_rate": 2.8160295625762112e-05, + "loss": 0.7338, + "step": 16974 + }, + { + "epoch": 3.0220797720797723, + "grad_norm": 0.8036878108978271, + "learning_rate": 2.8150559221817586e-05, + "loss": 0.4678, + "step": 16975 + }, + { + "epoch": 3.0222578347578346, + "grad_norm": 0.8545025587081909, + "learning_rate": 2.8140824225613594e-05, + "loss": 0.7263, + "step": 16976 + }, + { + "epoch": 3.0224358974358974, + "grad_norm": 0.8124253749847412, + "learning_rate": 2.813109063734084e-05, + "loss": 0.5429, + "step": 16977 + }, + { + "epoch": 3.02261396011396, + "grad_norm": 0.9141904711723328, + "learning_rate": 2.8121358457190116e-05, + "loss": 0.7035, + "step": 16978 + }, + { + "epoch": 3.022792022792023, + "grad_norm": 0.9336223602294922, + "learning_rate": 2.8111627685352048e-05, + "loss": 0.8263, + "step": 16979 + }, + { + "epoch": 3.0229700854700856, + "grad_norm": 0.972653329372406, + "learning_rate": 2.8101898322017295e-05, + "loss": 0.6823, + "step": 16980 + }, + { + "epoch": 3.0231481481481484, + "grad_norm": 0.9060801267623901, + "learning_rate": 2.8092170367376502e-05, + "loss": 0.7235, + "step": 16981 + }, + { + "epoch": 3.0233262108262107, + "grad_norm": 0.832062304019928, + "learning_rate": 2.8082443821620243e-05, + "loss": 0.8151, + "step": 16982 + }, + { + "epoch": 3.0235042735042734, + "grad_norm": 0.7333422303199768, + "learning_rate": 2.8072718684939104e-05, + "loss": 0.6096, + "step": 16983 + }, + { + "epoch": 3.023682336182336, + "grad_norm": 0.8611993193626404, + "learning_rate": 2.8062994957523603e-05, + "loss": 0.8708, + "step": 16984 + }, + { + "epoch": 3.023860398860399, + "grad_norm": 0.9310294389724731, + "learning_rate": 2.805327263956433e-05, + "loss": 0.8801, + "step": 16985 + }, + { + "epoch": 3.0240384615384617, + "grad_norm": 0.8927240967750549, + "learning_rate": 2.804355173125167e-05, + "loss": 0.8838, + "step": 16986 + }, + { + "epoch": 3.0242165242165244, + "grad_norm": 0.8257143497467041, + "learning_rate": 2.803383223277617e-05, + "loss": 0.7531, + "step": 16987 + }, + { + "epoch": 3.0243945868945867, + "grad_norm": 1.0260870456695557, + "learning_rate": 2.8024114144328227e-05, + "loss": 0.9371, + "step": 16988 + }, + { + "epoch": 3.0245726495726495, + "grad_norm": 0.8376401662826538, + "learning_rate": 2.8014397466098252e-05, + "loss": 0.6104, + "step": 16989 + }, + { + "epoch": 3.0247507122507122, + "grad_norm": 0.9509117603302002, + "learning_rate": 2.800468219827662e-05, + "loss": 0.6892, + "step": 16990 + }, + { + "epoch": 3.024928774928775, + "grad_norm": 0.8816271424293518, + "learning_rate": 2.7994968341053684e-05, + "loss": 0.8559, + "step": 16991 + }, + { + "epoch": 3.0251068376068377, + "grad_norm": 0.9000606536865234, + "learning_rate": 2.7985255894619754e-05, + "loss": 0.6829, + "step": 16992 + }, + { + "epoch": 3.0252849002849005, + "grad_norm": 0.8596423268318176, + "learning_rate": 2.797554485916515e-05, + "loss": 0.762, + "step": 16993 + }, + { + "epoch": 3.025462962962963, + "grad_norm": 0.92213374376297, + "learning_rate": 2.79658352348801e-05, + "loss": 0.8572, + "step": 16994 + }, + { + "epoch": 3.0256410256410255, + "grad_norm": 0.917836606502533, + "learning_rate": 2.7956127021954927e-05, + "loss": 0.7979, + "step": 16995 + }, + { + "epoch": 3.0258190883190883, + "grad_norm": 0.798367440700531, + "learning_rate": 2.7946420220579718e-05, + "loss": 0.6453, + "step": 16996 + }, + { + "epoch": 3.025997150997151, + "grad_norm": 1.0838942527770996, + "learning_rate": 2.7936714830944767e-05, + "loss": 0.7837, + "step": 16997 + }, + { + "epoch": 3.026175213675214, + "grad_norm": 0.9623873829841614, + "learning_rate": 2.7927010853240187e-05, + "loss": 0.9064, + "step": 16998 + }, + { + "epoch": 3.0263532763532766, + "grad_norm": 0.8624036908149719, + "learning_rate": 2.7917308287656075e-05, + "loss": 0.6475, + "step": 16999 + }, + { + "epoch": 3.026531339031339, + "grad_norm": 0.9793442487716675, + "learning_rate": 2.790760713438264e-05, + "loss": 0.7084, + "step": 17000 + }, + { + "epoch": 3.0267094017094016, + "grad_norm": 0.9052400588989258, + "learning_rate": 2.7897907393609812e-05, + "loss": 0.7263, + "step": 17001 + }, + { + "epoch": 3.0268874643874644, + "grad_norm": 0.7912825345993042, + "learning_rate": 2.788820906552775e-05, + "loss": 0.5578, + "step": 17002 + }, + { + "epoch": 3.027065527065527, + "grad_norm": 0.8624247312545776, + "learning_rate": 2.787851215032642e-05, + "loss": 0.6267, + "step": 17003 + }, + { + "epoch": 3.02724358974359, + "grad_norm": 0.9548758268356323, + "learning_rate": 2.786881664819584e-05, + "loss": 0.8152, + "step": 17004 + }, + { + "epoch": 3.0274216524216526, + "grad_norm": 0.7717092633247375, + "learning_rate": 2.7859122559325944e-05, + "loss": 0.6738, + "step": 17005 + }, + { + "epoch": 3.027599715099715, + "grad_norm": 0.9009813070297241, + "learning_rate": 2.7849429883906684e-05, + "loss": 0.5467, + "step": 17006 + }, + { + "epoch": 3.0277777777777777, + "grad_norm": 0.8310969471931458, + "learning_rate": 2.7839738622127974e-05, + "loss": 0.5986, + "step": 17007 + }, + { + "epoch": 3.0279558404558404, + "grad_norm": 0.9535056948661804, + "learning_rate": 2.7830048774179685e-05, + "loss": 0.7904, + "step": 17008 + }, + { + "epoch": 3.028133903133903, + "grad_norm": 0.9698938727378845, + "learning_rate": 2.782036034025164e-05, + "loss": 0.7514, + "step": 17009 + }, + { + "epoch": 3.028311965811966, + "grad_norm": 0.9064823389053345, + "learning_rate": 2.781067332053374e-05, + "loss": 0.7532, + "step": 17010 + }, + { + "epoch": 3.0284900284900287, + "grad_norm": 0.8301812410354614, + "learning_rate": 2.7800987715215686e-05, + "loss": 0.7068, + "step": 17011 + }, + { + "epoch": 3.028668091168091, + "grad_norm": 0.8986283540725708, + "learning_rate": 2.7791303524487332e-05, + "loss": 0.9139, + "step": 17012 + }, + { + "epoch": 3.0288461538461537, + "grad_norm": 0.9557301998138428, + "learning_rate": 2.7781620748538384e-05, + "loss": 0.7074, + "step": 17013 + }, + { + "epoch": 3.0290242165242165, + "grad_norm": 0.9753836989402771, + "learning_rate": 2.7771939387558554e-05, + "loss": 0.6967, + "step": 17014 + }, + { + "epoch": 3.0292022792022792, + "grad_norm": 0.8353948593139648, + "learning_rate": 2.7762259441737526e-05, + "loss": 0.7621, + "step": 17015 + }, + { + "epoch": 3.029380341880342, + "grad_norm": 0.9676481485366821, + "learning_rate": 2.7752580911264968e-05, + "loss": 0.8218, + "step": 17016 + }, + { + "epoch": 3.0295584045584047, + "grad_norm": 0.7660765051841736, + "learning_rate": 2.774290379633051e-05, + "loss": 0.8033, + "step": 17017 + }, + { + "epoch": 3.029736467236467, + "grad_norm": 0.916705310344696, + "learning_rate": 2.773322809712371e-05, + "loss": 0.849, + "step": 17018 + }, + { + "epoch": 3.02991452991453, + "grad_norm": 0.9600055813789368, + "learning_rate": 2.7723553813834224e-05, + "loss": 0.6848, + "step": 17019 + }, + { + "epoch": 3.0300925925925926, + "grad_norm": 1.0234928131103516, + "learning_rate": 2.771388094665156e-05, + "loss": 0.8953, + "step": 17020 + }, + { + "epoch": 3.0302706552706553, + "grad_norm": 0.8239841461181641, + "learning_rate": 2.7704209495765232e-05, + "loss": 0.8361, + "step": 17021 + }, + { + "epoch": 3.030448717948718, + "grad_norm": 1.0438402891159058, + "learning_rate": 2.7694539461364742e-05, + "loss": 0.6722, + "step": 17022 + }, + { + "epoch": 3.030626780626781, + "grad_norm": 0.7628459930419922, + "learning_rate": 2.7684870843639545e-05, + "loss": 0.6289, + "step": 17023 + }, + { + "epoch": 3.030804843304843, + "grad_norm": 0.8909597396850586, + "learning_rate": 2.7675203642779057e-05, + "loss": 0.7166, + "step": 17024 + }, + { + "epoch": 3.030982905982906, + "grad_norm": 1.116269826889038, + "learning_rate": 2.7665537858972767e-05, + "loss": 0.8198, + "step": 17025 + }, + { + "epoch": 3.0311609686609686, + "grad_norm": 0.8789699077606201, + "learning_rate": 2.7655873492409946e-05, + "loss": 0.5923, + "step": 17026 + }, + { + "epoch": 3.0313390313390314, + "grad_norm": 0.8829667568206787, + "learning_rate": 2.7646210543280048e-05, + "loss": 0.6907, + "step": 17027 + }, + { + "epoch": 3.031517094017094, + "grad_norm": 0.9860467314720154, + "learning_rate": 2.7636549011772307e-05, + "loss": 0.7255, + "step": 17028 + }, + { + "epoch": 3.031695156695157, + "grad_norm": 0.8999147415161133, + "learning_rate": 2.7626888898076096e-05, + "loss": 0.6997, + "step": 17029 + }, + { + "epoch": 3.031873219373219, + "grad_norm": 0.8238608241081238, + "learning_rate": 2.761723020238065e-05, + "loss": 0.6271, + "step": 17030 + }, + { + "epoch": 3.032051282051282, + "grad_norm": 0.8773884177207947, + "learning_rate": 2.7607572924875214e-05, + "loss": 0.8365, + "step": 17031 + }, + { + "epoch": 3.0322293447293447, + "grad_norm": 0.9619107246398926, + "learning_rate": 2.7597917065749013e-05, + "loss": 0.746, + "step": 17032 + }, + { + "epoch": 3.0324074074074074, + "grad_norm": 0.8207967877388, + "learning_rate": 2.7588262625191218e-05, + "loss": 0.7497, + "step": 17033 + }, + { + "epoch": 3.03258547008547, + "grad_norm": 0.9363070726394653, + "learning_rate": 2.757860960339097e-05, + "loss": 0.749, + "step": 17034 + }, + { + "epoch": 3.032763532763533, + "grad_norm": 0.7946116328239441, + "learning_rate": 2.756895800053748e-05, + "loss": 0.7944, + "step": 17035 + }, + { + "epoch": 3.0329415954415953, + "grad_norm": 0.8666053414344788, + "learning_rate": 2.7559307816819734e-05, + "loss": 0.7648, + "step": 17036 + }, + { + "epoch": 3.033119658119658, + "grad_norm": 0.9086824655532837, + "learning_rate": 2.75496590524269e-05, + "loss": 0.6917, + "step": 17037 + }, + { + "epoch": 3.0332977207977208, + "grad_norm": 0.8419581055641174, + "learning_rate": 2.7540011707547997e-05, + "loss": 0.6373, + "step": 17038 + }, + { + "epoch": 3.0334757834757835, + "grad_norm": 0.9141862392425537, + "learning_rate": 2.7530365782372035e-05, + "loss": 0.8568, + "step": 17039 + }, + { + "epoch": 3.0336538461538463, + "grad_norm": 0.9606926441192627, + "learning_rate": 2.7520721277088024e-05, + "loss": 0.7957, + "step": 17040 + }, + { + "epoch": 3.033831908831909, + "grad_norm": 0.9588294625282288, + "learning_rate": 2.7511078191884877e-05, + "loss": 0.8614, + "step": 17041 + }, + { + "epoch": 3.0340099715099713, + "grad_norm": 0.9736807942390442, + "learning_rate": 2.750143652695163e-05, + "loss": 0.9155, + "step": 17042 + }, + { + "epoch": 3.034188034188034, + "grad_norm": 0.886987030506134, + "learning_rate": 2.7491796282477078e-05, + "loss": 0.8215, + "step": 17043 + }, + { + "epoch": 3.034366096866097, + "grad_norm": 0.8999798893928528, + "learning_rate": 2.7482157458650182e-05, + "loss": 0.8449, + "step": 17044 + }, + { + "epoch": 3.0345441595441596, + "grad_norm": 0.8654738068580627, + "learning_rate": 2.7472520055659768e-05, + "loss": 0.6808, + "step": 17045 + }, + { + "epoch": 3.0347222222222223, + "grad_norm": 0.9802120923995972, + "learning_rate": 2.7462884073694662e-05, + "loss": 0.6272, + "step": 17046 + }, + { + "epoch": 3.034900284900285, + "grad_norm": 0.9409570693969727, + "learning_rate": 2.7453249512943658e-05, + "loss": 0.7017, + "step": 17047 + }, + { + "epoch": 3.0350783475783474, + "grad_norm": 0.9058420062065125, + "learning_rate": 2.7443616373595526e-05, + "loss": 0.668, + "step": 17048 + }, + { + "epoch": 3.03525641025641, + "grad_norm": 0.6895080208778381, + "learning_rate": 2.7433984655838985e-05, + "loss": 0.4501, + "step": 17049 + }, + { + "epoch": 3.035434472934473, + "grad_norm": 0.8462308049201965, + "learning_rate": 2.7424354359862824e-05, + "loss": 0.7687, + "step": 17050 + }, + { + "epoch": 3.0356125356125356, + "grad_norm": 1.0689095258712769, + "learning_rate": 2.741472548585562e-05, + "loss": 0.8657, + "step": 17051 + }, + { + "epoch": 3.0357905982905984, + "grad_norm": 0.8982266783714294, + "learning_rate": 2.7405098034006148e-05, + "loss": 0.6265, + "step": 17052 + }, + { + "epoch": 3.035968660968661, + "grad_norm": 0.9081108570098877, + "learning_rate": 2.739547200450292e-05, + "loss": 0.7115, + "step": 17053 + }, + { + "epoch": 3.0361467236467234, + "grad_norm": 0.9865546226501465, + "learning_rate": 2.7385847397534615e-05, + "loss": 0.8161, + "step": 17054 + }, + { + "epoch": 3.036324786324786, + "grad_norm": 0.9714661836624146, + "learning_rate": 2.737622421328979e-05, + "loss": 0.7357, + "step": 17055 + }, + { + "epoch": 3.036502849002849, + "grad_norm": 0.8435664772987366, + "learning_rate": 2.7366602451957003e-05, + "loss": 0.5715, + "step": 17056 + }, + { + "epoch": 3.0366809116809117, + "grad_norm": 0.9254729747772217, + "learning_rate": 2.7356982113724737e-05, + "loss": 0.8485, + "step": 17057 + }, + { + "epoch": 3.0368589743589745, + "grad_norm": 0.9366668462753296, + "learning_rate": 2.7347363198781496e-05, + "loss": 0.909, + "step": 17058 + }, + { + "epoch": 3.037037037037037, + "grad_norm": 0.8752120137214661, + "learning_rate": 2.7337745707315764e-05, + "loss": 0.6937, + "step": 17059 + }, + { + "epoch": 3.0372150997150995, + "grad_norm": 0.9134506583213806, + "learning_rate": 2.7328129639515963e-05, + "loss": 0.719, + "step": 17060 + }, + { + "epoch": 3.0373931623931623, + "grad_norm": 0.8471070528030396, + "learning_rate": 2.731851499557051e-05, + "loss": 0.6216, + "step": 17061 + }, + { + "epoch": 3.037571225071225, + "grad_norm": 0.8586950302124023, + "learning_rate": 2.730890177566776e-05, + "loss": 0.6494, + "step": 17062 + }, + { + "epoch": 3.0377492877492878, + "grad_norm": 0.915189802646637, + "learning_rate": 2.7299289979996078e-05, + "loss": 0.7831, + "step": 17063 + }, + { + "epoch": 3.0379273504273505, + "grad_norm": 0.9249003529548645, + "learning_rate": 2.7289679608743802e-05, + "loss": 0.6142, + "step": 17064 + }, + { + "epoch": 3.0381054131054133, + "grad_norm": 1.1311204433441162, + "learning_rate": 2.7280070662099198e-05, + "loss": 0.6026, + "step": 17065 + }, + { + "epoch": 3.0382834757834756, + "grad_norm": 0.8478325605392456, + "learning_rate": 2.7270463140250514e-05, + "loss": 0.6896, + "step": 17066 + }, + { + "epoch": 3.0384615384615383, + "grad_norm": 0.9298551678657532, + "learning_rate": 2.726085704338609e-05, + "loss": 0.7415, + "step": 17067 + }, + { + "epoch": 3.038639601139601, + "grad_norm": 0.8141112327575684, + "learning_rate": 2.7251252371694e-05, + "loss": 0.6328, + "step": 17068 + }, + { + "epoch": 3.038817663817664, + "grad_norm": 0.905672550201416, + "learning_rate": 2.7241649125362544e-05, + "loss": 0.8341, + "step": 17069 + }, + { + "epoch": 3.0389957264957266, + "grad_norm": 0.789050281047821, + "learning_rate": 2.7232047304579822e-05, + "loss": 0.7296, + "step": 17070 + }, + { + "epoch": 3.0391737891737893, + "grad_norm": 0.985736608505249, + "learning_rate": 2.722244690953397e-05, + "loss": 0.7118, + "step": 17071 + }, + { + "epoch": 3.0393518518518516, + "grad_norm": 0.8022065758705139, + "learning_rate": 2.7212847940413088e-05, + "loss": 0.5858, + "step": 17072 + }, + { + "epoch": 3.0395299145299144, + "grad_norm": 0.7794371843338013, + "learning_rate": 2.7203250397405244e-05, + "loss": 0.7273, + "step": 17073 + }, + { + "epoch": 3.039707977207977, + "grad_norm": 0.8665072321891785, + "learning_rate": 2.7193654280698466e-05, + "loss": 0.6504, + "step": 17074 + }, + { + "epoch": 3.03988603988604, + "grad_norm": 0.9348331689834595, + "learning_rate": 2.7184059590480848e-05, + "loss": 0.7944, + "step": 17075 + }, + { + "epoch": 3.0400641025641026, + "grad_norm": 1.0338337421417236, + "learning_rate": 2.717446632694025e-05, + "loss": 0.8487, + "step": 17076 + }, + { + "epoch": 3.0402421652421654, + "grad_norm": 0.9517617225646973, + "learning_rate": 2.7164874490264767e-05, + "loss": 0.6806, + "step": 17077 + }, + { + "epoch": 3.0404202279202277, + "grad_norm": 0.9159125685691833, + "learning_rate": 2.715528408064222e-05, + "loss": 0.6962, + "step": 17078 + }, + { + "epoch": 3.0405982905982905, + "grad_norm": 0.9007083177566528, + "learning_rate": 2.714569509826057e-05, + "loss": 0.5848, + "step": 17079 + }, + { + "epoch": 3.040776353276353, + "grad_norm": 0.8828113079071045, + "learning_rate": 2.71361075433077e-05, + "loss": 0.7774, + "step": 17080 + }, + { + "epoch": 3.040954415954416, + "grad_norm": 0.8390265703201294, + "learning_rate": 2.7126521415971405e-05, + "loss": 0.6373, + "step": 17081 + }, + { + "epoch": 3.0411324786324787, + "grad_norm": 0.8794222474098206, + "learning_rate": 2.71169367164396e-05, + "loss": 0.6907, + "step": 17082 + }, + { + "epoch": 3.0413105413105415, + "grad_norm": 0.8306505680084229, + "learning_rate": 2.710735344489995e-05, + "loss": 0.6983, + "step": 17083 + }, + { + "epoch": 3.041488603988604, + "grad_norm": 0.9314624071121216, + "learning_rate": 2.7097771601540333e-05, + "loss": 0.7328, + "step": 17084 + }, + { + "epoch": 3.0416666666666665, + "grad_norm": 0.9759663939476013, + "learning_rate": 2.7088191186548427e-05, + "loss": 0.7334, + "step": 17085 + }, + { + "epoch": 3.0418447293447293, + "grad_norm": 0.9378162622451782, + "learning_rate": 2.7078612200111962e-05, + "loss": 0.8155, + "step": 17086 + }, + { + "epoch": 3.042022792022792, + "grad_norm": 0.8654826283454895, + "learning_rate": 2.706903464241861e-05, + "loss": 0.7509, + "step": 17087 + }, + { + "epoch": 3.0422008547008548, + "grad_norm": 0.9504775404930115, + "learning_rate": 2.705945851365602e-05, + "loss": 0.5744, + "step": 17088 + }, + { + "epoch": 3.0423789173789175, + "grad_norm": 0.8799991607666016, + "learning_rate": 2.7049883814011822e-05, + "loss": 0.8687, + "step": 17089 + }, + { + "epoch": 3.04255698005698, + "grad_norm": 0.9499333500862122, + "learning_rate": 2.704031054367361e-05, + "loss": 0.7577, + "step": 17090 + }, + { + "epoch": 3.0427350427350426, + "grad_norm": 0.8906816840171814, + "learning_rate": 2.7030738702828918e-05, + "loss": 0.7075, + "step": 17091 + }, + { + "epoch": 3.0429131054131053, + "grad_norm": 0.8520913124084473, + "learning_rate": 2.7021168291665388e-05, + "loss": 0.7406, + "step": 17092 + }, + { + "epoch": 3.043091168091168, + "grad_norm": 0.9780383706092834, + "learning_rate": 2.7011599310370393e-05, + "loss": 0.7419, + "step": 17093 + }, + { + "epoch": 3.043269230769231, + "grad_norm": 0.9649802446365356, + "learning_rate": 2.700203175913153e-05, + "loss": 0.9557, + "step": 17094 + }, + { + "epoch": 3.0434472934472936, + "grad_norm": 0.8526589274406433, + "learning_rate": 2.6992465638136212e-05, + "loss": 0.7484, + "step": 17095 + }, + { + "epoch": 3.0436253561253563, + "grad_norm": 0.9232372641563416, + "learning_rate": 2.6982900947571864e-05, + "loss": 0.8444, + "step": 17096 + }, + { + "epoch": 3.0438034188034186, + "grad_norm": 0.8113499283790588, + "learning_rate": 2.69733376876259e-05, + "loss": 0.6301, + "step": 17097 + }, + { + "epoch": 3.0439814814814814, + "grad_norm": 0.8863048553466797, + "learning_rate": 2.6963775858485652e-05, + "loss": 0.7519, + "step": 17098 + }, + { + "epoch": 3.044159544159544, + "grad_norm": 0.8750777244567871, + "learning_rate": 2.6954215460338562e-05, + "loss": 0.7545, + "step": 17099 + }, + { + "epoch": 3.044337606837607, + "grad_norm": 0.937821090221405, + "learning_rate": 2.6944656493371812e-05, + "loss": 0.773, + "step": 17100 + }, + { + "epoch": 3.0445156695156697, + "grad_norm": 0.8423610925674438, + "learning_rate": 2.6935098957772785e-05, + "loss": 0.6537, + "step": 17101 + }, + { + "epoch": 3.0446937321937324, + "grad_norm": 0.7845390439033508, + "learning_rate": 2.6925542853728726e-05, + "loss": 0.6826, + "step": 17102 + }, + { + "epoch": 3.0448717948717947, + "grad_norm": 0.9942399263381958, + "learning_rate": 2.6915988181426842e-05, + "loss": 0.8664, + "step": 17103 + }, + { + "epoch": 3.0450498575498575, + "grad_norm": 0.8977256417274475, + "learning_rate": 2.690643494105437e-05, + "loss": 0.6945, + "step": 17104 + }, + { + "epoch": 3.04522792022792, + "grad_norm": 1.0192148685455322, + "learning_rate": 2.6896883132798456e-05, + "loss": 0.7012, + "step": 17105 + }, + { + "epoch": 3.045405982905983, + "grad_norm": 0.9779385924339294, + "learning_rate": 2.688733275684623e-05, + "loss": 0.7788, + "step": 17106 + }, + { + "epoch": 3.0455840455840457, + "grad_norm": 0.9183738231658936, + "learning_rate": 2.6877783813384894e-05, + "loss": 0.6973, + "step": 17107 + }, + { + "epoch": 3.0457621082621085, + "grad_norm": 1.0147916078567505, + "learning_rate": 2.6868236302601422e-05, + "loss": 0.7316, + "step": 17108 + }, + { + "epoch": 3.0459401709401708, + "grad_norm": 0.8801769018173218, + "learning_rate": 2.685869022468299e-05, + "loss": 0.6735, + "step": 17109 + }, + { + "epoch": 3.0461182336182335, + "grad_norm": 0.8848022222518921, + "learning_rate": 2.684914557981657e-05, + "loss": 0.7432, + "step": 17110 + }, + { + "epoch": 3.0462962962962963, + "grad_norm": 0.9158660173416138, + "learning_rate": 2.6839602368189188e-05, + "loss": 0.7133, + "step": 17111 + }, + { + "epoch": 3.046474358974359, + "grad_norm": 0.8280047178268433, + "learning_rate": 2.6830060589987826e-05, + "loss": 0.7432, + "step": 17112 + }, + { + "epoch": 3.046652421652422, + "grad_norm": 0.8806058764457703, + "learning_rate": 2.6820520245399427e-05, + "loss": 0.5107, + "step": 17113 + }, + { + "epoch": 3.0468304843304845, + "grad_norm": 0.8557198643684387, + "learning_rate": 2.681098133461091e-05, + "loss": 0.7598, + "step": 17114 + }, + { + "epoch": 3.047008547008547, + "grad_norm": 0.9113288521766663, + "learning_rate": 2.6801443857809183e-05, + "loss": 0.799, + "step": 17115 + }, + { + "epoch": 3.0471866096866096, + "grad_norm": 0.8104450702667236, + "learning_rate": 2.6791907815181072e-05, + "loss": 0.6265, + "step": 17116 + }, + { + "epoch": 3.0473646723646723, + "grad_norm": 0.7648457288742065, + "learning_rate": 2.6782373206913492e-05, + "loss": 0.6227, + "step": 17117 + }, + { + "epoch": 3.047542735042735, + "grad_norm": 0.8509871363639832, + "learning_rate": 2.6772840033193204e-05, + "loss": 0.7835, + "step": 17118 + }, + { + "epoch": 3.047720797720798, + "grad_norm": 0.9865061044692993, + "learning_rate": 2.6763308294207e-05, + "loss": 0.6175, + "step": 17119 + }, + { + "epoch": 3.0478988603988606, + "grad_norm": 1.0529922246932983, + "learning_rate": 2.6753777990141625e-05, + "loss": 0.7668, + "step": 17120 + }, + { + "epoch": 3.048076923076923, + "grad_norm": 0.9097694158554077, + "learning_rate": 2.6744249121183795e-05, + "loss": 0.7583, + "step": 17121 + }, + { + "epoch": 3.0482549857549857, + "grad_norm": 1.0285091400146484, + "learning_rate": 2.6734721687520293e-05, + "loss": 0.748, + "step": 17122 + }, + { + "epoch": 3.0484330484330484, + "grad_norm": 0.9668599963188171, + "learning_rate": 2.6725195689337658e-05, + "loss": 0.7636, + "step": 17123 + }, + { + "epoch": 3.048611111111111, + "grad_norm": 0.8746935725212097, + "learning_rate": 2.671567112682265e-05, + "loss": 0.6558, + "step": 17124 + }, + { + "epoch": 3.048789173789174, + "grad_norm": 0.9465086460113525, + "learning_rate": 2.670614800016178e-05, + "loss": 0.706, + "step": 17125 + }, + { + "epoch": 3.0489672364672367, + "grad_norm": 1.0083024501800537, + "learning_rate": 2.6696626309541718e-05, + "loss": 0.8062, + "step": 17126 + }, + { + "epoch": 3.049145299145299, + "grad_norm": 0.9895032048225403, + "learning_rate": 2.668710605514898e-05, + "loss": 0.8804, + "step": 17127 + }, + { + "epoch": 3.0493233618233617, + "grad_norm": 0.903660237789154, + "learning_rate": 2.667758723717011e-05, + "loss": 0.7487, + "step": 17128 + }, + { + "epoch": 3.0495014245014245, + "grad_norm": 0.8843152523040771, + "learning_rate": 2.6668069855791598e-05, + "loss": 0.5683, + "step": 17129 + }, + { + "epoch": 3.0496794871794872, + "grad_norm": 0.9241248369216919, + "learning_rate": 2.6658553911199936e-05, + "loss": 0.7855, + "step": 17130 + }, + { + "epoch": 3.04985754985755, + "grad_norm": 1.0450769662857056, + "learning_rate": 2.664903940358152e-05, + "loss": 0.7898, + "step": 17131 + }, + { + "epoch": 3.0500356125356127, + "grad_norm": 0.993699312210083, + "learning_rate": 2.6639526333122855e-05, + "loss": 0.7119, + "step": 17132 + }, + { + "epoch": 3.050213675213675, + "grad_norm": 0.8813846707344055, + "learning_rate": 2.663001470001023e-05, + "loss": 0.7032, + "step": 17133 + }, + { + "epoch": 3.050391737891738, + "grad_norm": 0.9121679067611694, + "learning_rate": 2.662050450443011e-05, + "loss": 0.8321, + "step": 17134 + }, + { + "epoch": 3.0505698005698005, + "grad_norm": 0.8883702158927917, + "learning_rate": 2.6610995746568713e-05, + "loss": 0.7917, + "step": 17135 + }, + { + "epoch": 3.0507478632478633, + "grad_norm": 0.8827738761901855, + "learning_rate": 2.660148842661243e-05, + "loss": 0.7541, + "step": 17136 + }, + { + "epoch": 3.050925925925926, + "grad_norm": 0.8486347794532776, + "learning_rate": 2.6591982544747508e-05, + "loss": 0.6555, + "step": 17137 + }, + { + "epoch": 3.051103988603989, + "grad_norm": 0.7430234551429749, + "learning_rate": 2.6582478101160167e-05, + "loss": 0.5562, + "step": 17138 + }, + { + "epoch": 3.051282051282051, + "grad_norm": 0.8816348314285278, + "learning_rate": 2.657297509603671e-05, + "loss": 0.8661, + "step": 17139 + }, + { + "epoch": 3.051460113960114, + "grad_norm": 0.9184949994087219, + "learning_rate": 2.656347352956322e-05, + "loss": 0.7907, + "step": 17140 + }, + { + "epoch": 3.0516381766381766, + "grad_norm": 0.7924085855484009, + "learning_rate": 2.6553973401925946e-05, + "loss": 0.6181, + "step": 17141 + }, + { + "epoch": 3.0518162393162394, + "grad_norm": 0.9879988431930542, + "learning_rate": 2.6544474713310997e-05, + "loss": 0.8455, + "step": 17142 + }, + { + "epoch": 3.051994301994302, + "grad_norm": 0.8584102988243103, + "learning_rate": 2.6534977463904475e-05, + "loss": 0.89, + "step": 17143 + }, + { + "epoch": 3.052172364672365, + "grad_norm": 0.9548700451850891, + "learning_rate": 2.6525481653892447e-05, + "loss": 0.7599, + "step": 17144 + }, + { + "epoch": 3.052350427350427, + "grad_norm": 0.9319002628326416, + "learning_rate": 2.6515987283460985e-05, + "loss": 0.934, + "step": 17145 + }, + { + "epoch": 3.05252849002849, + "grad_norm": 1.006009817123413, + "learning_rate": 2.650649435279611e-05, + "loss": 0.7246, + "step": 17146 + }, + { + "epoch": 3.0527065527065527, + "grad_norm": 0.8670983910560608, + "learning_rate": 2.6497002862083797e-05, + "loss": 0.6835, + "step": 17147 + }, + { + "epoch": 3.0528846153846154, + "grad_norm": 0.79991614818573, + "learning_rate": 2.6487512811509994e-05, + "loss": 0.5659, + "step": 17148 + }, + { + "epoch": 3.053062678062678, + "grad_norm": 0.8038514852523804, + "learning_rate": 2.6478024201260732e-05, + "loss": 0.7436, + "step": 17149 + }, + { + "epoch": 3.053240740740741, + "grad_norm": 0.9287776350975037, + "learning_rate": 2.646853703152179e-05, + "loss": 0.8187, + "step": 17150 + }, + { + "epoch": 3.0534188034188032, + "grad_norm": 0.952595591545105, + "learning_rate": 2.6459051302479153e-05, + "loss": 0.8553, + "step": 17151 + }, + { + "epoch": 3.053596866096866, + "grad_norm": 0.8326176404953003, + "learning_rate": 2.6449567014318643e-05, + "loss": 0.7641, + "step": 17152 + }, + { + "epoch": 3.0537749287749287, + "grad_norm": 0.8747450113296509, + "learning_rate": 2.6440084167226063e-05, + "loss": 0.7237, + "step": 17153 + }, + { + "epoch": 3.0539529914529915, + "grad_norm": 0.9608044624328613, + "learning_rate": 2.6430602761387247e-05, + "loss": 0.8389, + "step": 17154 + }, + { + "epoch": 3.0541310541310542, + "grad_norm": 1.0133368968963623, + "learning_rate": 2.642112279698793e-05, + "loss": 0.7081, + "step": 17155 + }, + { + "epoch": 3.054309116809117, + "grad_norm": 0.8604594469070435, + "learning_rate": 2.6411644274213832e-05, + "loss": 0.8706, + "step": 17156 + }, + { + "epoch": 3.0544871794871793, + "grad_norm": 0.8340579867362976, + "learning_rate": 2.640216719325074e-05, + "loss": 0.6569, + "step": 17157 + }, + { + "epoch": 3.054665242165242, + "grad_norm": 0.894295871257782, + "learning_rate": 2.6392691554284287e-05, + "loss": 0.8071, + "step": 17158 + }, + { + "epoch": 3.054843304843305, + "grad_norm": 0.8919312953948975, + "learning_rate": 2.6383217357500133e-05, + "loss": 0.5967, + "step": 17159 + }, + { + "epoch": 3.0550213675213675, + "grad_norm": 0.9570869207382202, + "learning_rate": 2.6373744603083916e-05, + "loss": 0.8526, + "step": 17160 + }, + { + "epoch": 3.0551994301994303, + "grad_norm": 0.9222795367240906, + "learning_rate": 2.636427329122123e-05, + "loss": 0.7142, + "step": 17161 + }, + { + "epoch": 3.055377492877493, + "grad_norm": 0.8729871511459351, + "learning_rate": 2.635480342209764e-05, + "loss": 0.7344, + "step": 17162 + }, + { + "epoch": 3.0555555555555554, + "grad_norm": 0.9293301701545715, + "learning_rate": 2.6345334995898662e-05, + "loss": 0.7019, + "step": 17163 + }, + { + "epoch": 3.055733618233618, + "grad_norm": 0.9955607652664185, + "learning_rate": 2.6335868012809905e-05, + "loss": 0.8012, + "step": 17164 + }, + { + "epoch": 3.055911680911681, + "grad_norm": 0.8420068621635437, + "learning_rate": 2.632640247301673e-05, + "loss": 0.7686, + "step": 17165 + }, + { + "epoch": 3.0560897435897436, + "grad_norm": 0.9361758232116699, + "learning_rate": 2.6316938376704682e-05, + "loss": 0.728, + "step": 17166 + }, + { + "epoch": 3.0562678062678064, + "grad_norm": 0.9174705147743225, + "learning_rate": 2.630747572405916e-05, + "loss": 0.77, + "step": 17167 + }, + { + "epoch": 3.056445868945869, + "grad_norm": 1.0511798858642578, + "learning_rate": 2.6298014515265578e-05, + "loss": 0.7319, + "step": 17168 + }, + { + "epoch": 3.0566239316239314, + "grad_norm": 0.9990606307983398, + "learning_rate": 2.6288554750509286e-05, + "loss": 0.7417, + "step": 17169 + }, + { + "epoch": 3.056801994301994, + "grad_norm": 0.8228463530540466, + "learning_rate": 2.6279096429975648e-05, + "loss": 0.8001, + "step": 17170 + }, + { + "epoch": 3.056980056980057, + "grad_norm": 0.9636453986167908, + "learning_rate": 2.626963955384998e-05, + "loss": 0.7647, + "step": 17171 + }, + { + "epoch": 3.0571581196581197, + "grad_norm": 1.096347451210022, + "learning_rate": 2.6260184122317554e-05, + "loss": 0.816, + "step": 17172 + }, + { + "epoch": 3.0573361823361824, + "grad_norm": 0.8099706768989563, + "learning_rate": 2.6250730135563606e-05, + "loss": 0.6105, + "step": 17173 + }, + { + "epoch": 3.057514245014245, + "grad_norm": 0.8991966247558594, + "learning_rate": 2.6241277593773473e-05, + "loss": 0.6853, + "step": 17174 + }, + { + "epoch": 3.0576923076923075, + "grad_norm": 0.8311452865600586, + "learning_rate": 2.623182649713222e-05, + "loss": 0.7684, + "step": 17175 + }, + { + "epoch": 3.0578703703703702, + "grad_norm": 1.097398281097412, + "learning_rate": 2.6222376845825115e-05, + "loss": 1.0463, + "step": 17176 + }, + { + "epoch": 3.058048433048433, + "grad_norm": 0.9334904551506042, + "learning_rate": 2.621292864003727e-05, + "loss": 0.8499, + "step": 17177 + }, + { + "epoch": 3.0582264957264957, + "grad_norm": 0.912241518497467, + "learning_rate": 2.6203481879953783e-05, + "loss": 0.8186, + "step": 17178 + }, + { + "epoch": 3.0584045584045585, + "grad_norm": 0.8956477046012878, + "learning_rate": 2.619403656575984e-05, + "loss": 0.6561, + "step": 17179 + }, + { + "epoch": 3.0585826210826212, + "grad_norm": 0.997071385383606, + "learning_rate": 2.6184592697640352e-05, + "loss": 0.8774, + "step": 17180 + }, + { + "epoch": 3.0587606837606836, + "grad_norm": 0.8529218435287476, + "learning_rate": 2.617515027578048e-05, + "loss": 0.7653, + "step": 17181 + }, + { + "epoch": 3.0589387464387463, + "grad_norm": 0.8638035655021667, + "learning_rate": 2.616570930036517e-05, + "loss": 0.6703, + "step": 17182 + }, + { + "epoch": 3.059116809116809, + "grad_norm": 0.9718494415283203, + "learning_rate": 2.6156269771579412e-05, + "loss": 0.9117, + "step": 17183 + }, + { + "epoch": 3.059294871794872, + "grad_norm": 0.8079463243484497, + "learning_rate": 2.6146831689608154e-05, + "loss": 0.716, + "step": 17184 + }, + { + "epoch": 3.0594729344729346, + "grad_norm": 0.8149730563163757, + "learning_rate": 2.6137395054636303e-05, + "loss": 0.5388, + "step": 17185 + }, + { + "epoch": 3.0596509971509973, + "grad_norm": 0.8048275113105774, + "learning_rate": 2.6127959866848774e-05, + "loss": 0.6043, + "step": 17186 + }, + { + "epoch": 3.0598290598290596, + "grad_norm": 0.850706934928894, + "learning_rate": 2.611852612643041e-05, + "loss": 0.6969, + "step": 17187 + }, + { + "epoch": 3.0600071225071224, + "grad_norm": 0.8014549016952515, + "learning_rate": 2.6109093833566012e-05, + "loss": 0.6422, + "step": 17188 + }, + { + "epoch": 3.060185185185185, + "grad_norm": 0.9597519636154175, + "learning_rate": 2.60996629884405e-05, + "loss": 0.7547, + "step": 17189 + }, + { + "epoch": 3.060363247863248, + "grad_norm": 0.8303548693656921, + "learning_rate": 2.60902335912385e-05, + "loss": 0.6349, + "step": 17190 + }, + { + "epoch": 3.0605413105413106, + "grad_norm": 0.8855451345443726, + "learning_rate": 2.608080564214488e-05, + "loss": 0.8361, + "step": 17191 + }, + { + "epoch": 3.0607193732193734, + "grad_norm": 0.8511565327644348, + "learning_rate": 2.6071379141344322e-05, + "loss": 0.7693, + "step": 17192 + }, + { + "epoch": 3.0608974358974357, + "grad_norm": 1.0611032247543335, + "learning_rate": 2.6061954089021512e-05, + "loss": 0.821, + "step": 17193 + }, + { + "epoch": 3.0610754985754984, + "grad_norm": 0.8549980521202087, + "learning_rate": 2.6052530485361114e-05, + "loss": 0.7522, + "step": 17194 + }, + { + "epoch": 3.061253561253561, + "grad_norm": 1.0446410179138184, + "learning_rate": 2.604310833054777e-05, + "loss": 0.9727, + "step": 17195 + }, + { + "epoch": 3.061431623931624, + "grad_norm": 0.949283242225647, + "learning_rate": 2.603368762476609e-05, + "loss": 0.9997, + "step": 17196 + }, + { + "epoch": 3.0616096866096867, + "grad_norm": 0.8465662598609924, + "learning_rate": 2.602426836820062e-05, + "loss": 0.6856, + "step": 17197 + }, + { + "epoch": 3.0617877492877494, + "grad_norm": 0.8954669237136841, + "learning_rate": 2.6014850561035963e-05, + "loss": 0.8639, + "step": 17198 + }, + { + "epoch": 3.0619658119658117, + "grad_norm": 0.9493638277053833, + "learning_rate": 2.6005434203456626e-05, + "loss": 0.9289, + "step": 17199 + }, + { + "epoch": 3.0621438746438745, + "grad_norm": 0.9050379395484924, + "learning_rate": 2.599601929564709e-05, + "loss": 0.6737, + "step": 17200 + }, + { + "epoch": 3.0623219373219372, + "grad_norm": 1.0764466524124146, + "learning_rate": 2.5986605837791835e-05, + "loss": 0.9317, + "step": 17201 + }, + { + "epoch": 3.0625, + "grad_norm": 0.9864383339881897, + "learning_rate": 2.5977193830075287e-05, + "loss": 0.7932, + "step": 17202 + }, + { + "epoch": 3.0626780626780628, + "grad_norm": 0.9895427227020264, + "learning_rate": 2.5967783272681823e-05, + "loss": 0.8253, + "step": 17203 + }, + { + "epoch": 3.0628561253561255, + "grad_norm": 0.9177289009094238, + "learning_rate": 2.5958374165795906e-05, + "loss": 0.7333, + "step": 17204 + }, + { + "epoch": 3.0630341880341883, + "grad_norm": 1.088364601135254, + "learning_rate": 2.594896650960179e-05, + "loss": 0.9155, + "step": 17205 + }, + { + "epoch": 3.0632122507122506, + "grad_norm": 0.894473135471344, + "learning_rate": 2.5939560304283917e-05, + "loss": 0.8166, + "step": 17206 + }, + { + "epoch": 3.0633903133903133, + "grad_norm": 0.8343120813369751, + "learning_rate": 2.5930155550026436e-05, + "loss": 0.7034, + "step": 17207 + }, + { + "epoch": 3.063568376068376, + "grad_norm": 0.8967345356941223, + "learning_rate": 2.5920752247013737e-05, + "loss": 0.6673, + "step": 17208 + }, + { + "epoch": 3.063746438746439, + "grad_norm": 0.9521552324295044, + "learning_rate": 2.591135039543e-05, + "loss": 0.7148, + "step": 17209 + }, + { + "epoch": 3.0639245014245016, + "grad_norm": 0.7986984252929688, + "learning_rate": 2.590194999545944e-05, + "loss": 0.7299, + "step": 17210 + }, + { + "epoch": 3.064102564102564, + "grad_norm": 0.9818909168243408, + "learning_rate": 2.589255104728626e-05, + "loss": 0.7047, + "step": 17211 + }, + { + "epoch": 3.0642806267806266, + "grad_norm": 0.9633432030677795, + "learning_rate": 2.5883153551094585e-05, + "loss": 0.6901, + "step": 17212 + }, + { + "epoch": 3.0644586894586894, + "grad_norm": 0.9325671792030334, + "learning_rate": 2.587375750706853e-05, + "loss": 0.8643, + "step": 17213 + }, + { + "epoch": 3.064636752136752, + "grad_norm": 0.9931867122650146, + "learning_rate": 2.5864362915392272e-05, + "loss": 0.7626, + "step": 17214 + }, + { + "epoch": 3.064814814814815, + "grad_norm": 0.8451269268989563, + "learning_rate": 2.585496977624975e-05, + "loss": 0.8091, + "step": 17215 + }, + { + "epoch": 3.0649928774928776, + "grad_norm": 0.8387511372566223, + "learning_rate": 2.5845578089825105e-05, + "loss": 0.6729, + "step": 17216 + }, + { + "epoch": 3.0651709401709404, + "grad_norm": 0.9810523986816406, + "learning_rate": 2.583618785630233e-05, + "loss": 0.8085, + "step": 17217 + }, + { + "epoch": 3.0653490028490027, + "grad_norm": 0.834884524345398, + "learning_rate": 2.5826799075865372e-05, + "loss": 0.6366, + "step": 17218 + }, + { + "epoch": 3.0655270655270654, + "grad_norm": 1.0484609603881836, + "learning_rate": 2.5817411748698217e-05, + "loss": 0.7411, + "step": 17219 + }, + { + "epoch": 3.065705128205128, + "grad_norm": 0.8887215852737427, + "learning_rate": 2.5808025874984742e-05, + "loss": 0.7981, + "step": 17220 + }, + { + "epoch": 3.065883190883191, + "grad_norm": 0.8052788972854614, + "learning_rate": 2.5798641454908944e-05, + "loss": 0.5341, + "step": 17221 + }, + { + "epoch": 3.0660612535612537, + "grad_norm": 1.0397300720214844, + "learning_rate": 2.5789258488654567e-05, + "loss": 0.8031, + "step": 17222 + }, + { + "epoch": 3.0662393162393164, + "grad_norm": 0.9165951609611511, + "learning_rate": 2.5779876976405535e-05, + "loss": 0.8193, + "step": 17223 + }, + { + "epoch": 3.0664173789173788, + "grad_norm": 0.8900893330574036, + "learning_rate": 2.5770496918345633e-05, + "loss": 0.7372, + "step": 17224 + }, + { + "epoch": 3.0665954415954415, + "grad_norm": 0.8510205149650574, + "learning_rate": 2.576111831465865e-05, + "loss": 0.8246, + "step": 17225 + }, + { + "epoch": 3.0667735042735043, + "grad_norm": 0.837579071521759, + "learning_rate": 2.5751741165528342e-05, + "loss": 0.9221, + "step": 17226 + }, + { + "epoch": 3.066951566951567, + "grad_norm": 0.911714494228363, + "learning_rate": 2.5742365471138418e-05, + "loss": 0.878, + "step": 17227 + }, + { + "epoch": 3.0671296296296298, + "grad_norm": 0.8773578405380249, + "learning_rate": 2.5732991231672565e-05, + "loss": 0.75, + "step": 17228 + }, + { + "epoch": 3.0673076923076925, + "grad_norm": 0.7727368474006653, + "learning_rate": 2.5723618447314523e-05, + "loss": 0.6015, + "step": 17229 + }, + { + "epoch": 3.067485754985755, + "grad_norm": 0.7762733101844788, + "learning_rate": 2.5714247118247826e-05, + "loss": 0.502, + "step": 17230 + }, + { + "epoch": 3.0676638176638176, + "grad_norm": 1.0052740573883057, + "learning_rate": 2.57048772446562e-05, + "loss": 0.8727, + "step": 17231 + }, + { + "epoch": 3.0678418803418803, + "grad_norm": 0.8907061219215393, + "learning_rate": 2.5695508826723113e-05, + "loss": 0.7661, + "step": 17232 + }, + { + "epoch": 3.068019943019943, + "grad_norm": 1.0816287994384766, + "learning_rate": 2.5686141864632208e-05, + "loss": 0.7263, + "step": 17233 + }, + { + "epoch": 3.068198005698006, + "grad_norm": 0.7547267079353333, + "learning_rate": 2.567677635856698e-05, + "loss": 0.6671, + "step": 17234 + }, + { + "epoch": 3.0683760683760686, + "grad_norm": 0.9319064021110535, + "learning_rate": 2.5667412308710916e-05, + "loss": 0.816, + "step": 17235 + }, + { + "epoch": 3.068554131054131, + "grad_norm": 0.8986789584159851, + "learning_rate": 2.56580497152475e-05, + "loss": 0.657, + "step": 17236 + }, + { + "epoch": 3.0687321937321936, + "grad_norm": 0.9154072999954224, + "learning_rate": 2.5648688578360135e-05, + "loss": 0.8183, + "step": 17237 + }, + { + "epoch": 3.0689102564102564, + "grad_norm": 0.8452461361885071, + "learning_rate": 2.5639328898232308e-05, + "loss": 0.8065, + "step": 17238 + }, + { + "epoch": 3.069088319088319, + "grad_norm": 0.9826174974441528, + "learning_rate": 2.5629970675047354e-05, + "loss": 0.9159, + "step": 17239 + }, + { + "epoch": 3.069266381766382, + "grad_norm": 0.7786675691604614, + "learning_rate": 2.5620613908988632e-05, + "loss": 0.5292, + "step": 17240 + }, + { + "epoch": 3.0694444444444446, + "grad_norm": 0.9880074858665466, + "learning_rate": 2.5611258600239464e-05, + "loss": 0.6227, + "step": 17241 + }, + { + "epoch": 3.069622507122507, + "grad_norm": 0.8934128880500793, + "learning_rate": 2.560190474898316e-05, + "loss": 0.7228, + "step": 17242 + }, + { + "epoch": 3.0698005698005697, + "grad_norm": 0.9464970231056213, + "learning_rate": 2.5592552355402988e-05, + "loss": 0.8885, + "step": 17243 + }, + { + "epoch": 3.0699786324786325, + "grad_norm": 1.071493148803711, + "learning_rate": 2.558320141968219e-05, + "loss": 0.9354, + "step": 17244 + }, + { + "epoch": 3.070156695156695, + "grad_norm": 0.9068328142166138, + "learning_rate": 2.5573851942003933e-05, + "loss": 0.7176, + "step": 17245 + }, + { + "epoch": 3.070334757834758, + "grad_norm": 0.7818723320960999, + "learning_rate": 2.55645039225515e-05, + "loss": 0.6526, + "step": 17246 + }, + { + "epoch": 3.0705128205128207, + "grad_norm": 1.052809476852417, + "learning_rate": 2.555515736150793e-05, + "loss": 0.7552, + "step": 17247 + }, + { + "epoch": 3.070690883190883, + "grad_norm": 0.8437279462814331, + "learning_rate": 2.5545812259056433e-05, + "loss": 0.7114, + "step": 17248 + }, + { + "epoch": 3.0708689458689458, + "grad_norm": 1.0028997659683228, + "learning_rate": 2.5536468615380083e-05, + "loss": 0.8291, + "step": 17249 + }, + { + "epoch": 3.0710470085470085, + "grad_norm": 0.9631550908088684, + "learning_rate": 2.552712643066194e-05, + "loss": 1.0165, + "step": 17250 + }, + { + "epoch": 3.0712250712250713, + "grad_norm": 0.8871008157730103, + "learning_rate": 2.5517785705085052e-05, + "loss": 0.6304, + "step": 17251 + }, + { + "epoch": 3.071403133903134, + "grad_norm": 0.8630111813545227, + "learning_rate": 2.5508446438832432e-05, + "loss": 0.7626, + "step": 17252 + }, + { + "epoch": 3.0715811965811968, + "grad_norm": 0.9306061267852783, + "learning_rate": 2.549910863208703e-05, + "loss": 0.8901, + "step": 17253 + }, + { + "epoch": 3.071759259259259, + "grad_norm": 0.9966697692871094, + "learning_rate": 2.5489772285031897e-05, + "loss": 0.7751, + "step": 17254 + }, + { + "epoch": 3.071937321937322, + "grad_norm": 0.9372161030769348, + "learning_rate": 2.5480437397849832e-05, + "loss": 0.866, + "step": 17255 + }, + { + "epoch": 3.0721153846153846, + "grad_norm": 0.9376949667930603, + "learning_rate": 2.547110397072382e-05, + "loss": 0.6556, + "step": 17256 + }, + { + "epoch": 3.0722934472934473, + "grad_norm": 1.0755635499954224, + "learning_rate": 2.5461772003836714e-05, + "loss": 0.8842, + "step": 17257 + }, + { + "epoch": 3.07247150997151, + "grad_norm": 0.8353457450866699, + "learning_rate": 2.545244149737134e-05, + "loss": 0.7091, + "step": 17258 + }, + { + "epoch": 3.072649572649573, + "grad_norm": 0.9944921135902405, + "learning_rate": 2.544311245151051e-05, + "loss": 0.6709, + "step": 17259 + }, + { + "epoch": 3.072827635327635, + "grad_norm": 0.8432626128196716, + "learning_rate": 2.5433784866436995e-05, + "loss": 0.7281, + "step": 17260 + }, + { + "epoch": 3.073005698005698, + "grad_norm": 0.8600237965583801, + "learning_rate": 2.5424458742333622e-05, + "loss": 0.8318, + "step": 17261 + }, + { + "epoch": 3.0731837606837606, + "grad_norm": 1.0386722087860107, + "learning_rate": 2.5415134079383006e-05, + "loss": 0.784, + "step": 17262 + }, + { + "epoch": 3.0733618233618234, + "grad_norm": 0.9741103649139404, + "learning_rate": 2.540581087776792e-05, + "loss": 0.824, + "step": 17263 + }, + { + "epoch": 3.073539886039886, + "grad_norm": 0.85637366771698, + "learning_rate": 2.5396489137671033e-05, + "loss": 0.6022, + "step": 17264 + }, + { + "epoch": 3.073717948717949, + "grad_norm": 0.9319208264350891, + "learning_rate": 2.5387168859274947e-05, + "loss": 0.7912, + "step": 17265 + }, + { + "epoch": 3.073896011396011, + "grad_norm": 0.9643087387084961, + "learning_rate": 2.5377850042762295e-05, + "loss": 0.8833, + "step": 17266 + }, + { + "epoch": 3.074074074074074, + "grad_norm": 0.860395610332489, + "learning_rate": 2.536853268831567e-05, + "loss": 0.6679, + "step": 17267 + }, + { + "epoch": 3.0742521367521367, + "grad_norm": 0.8932734727859497, + "learning_rate": 2.5359216796117603e-05, + "loss": 0.6577, + "step": 17268 + }, + { + "epoch": 3.0744301994301995, + "grad_norm": 0.9940623641014099, + "learning_rate": 2.534990236635063e-05, + "loss": 0.7469, + "step": 17269 + }, + { + "epoch": 3.074608262108262, + "grad_norm": 0.7407364845275879, + "learning_rate": 2.5340589399197222e-05, + "loss": 0.6288, + "step": 17270 + }, + { + "epoch": 3.074786324786325, + "grad_norm": 0.9189527034759521, + "learning_rate": 2.5331277894839934e-05, + "loss": 0.5919, + "step": 17271 + }, + { + "epoch": 3.0749643874643873, + "grad_norm": 0.8940016031265259, + "learning_rate": 2.5321967853461093e-05, + "loss": 0.8846, + "step": 17272 + }, + { + "epoch": 3.07514245014245, + "grad_norm": 0.9263308048248291, + "learning_rate": 2.5312659275243177e-05, + "loss": 0.7543, + "step": 17273 + }, + { + "epoch": 3.0753205128205128, + "grad_norm": 0.9368407726287842, + "learning_rate": 2.5303352160368556e-05, + "loss": 0.7127, + "step": 17274 + }, + { + "epoch": 3.0754985754985755, + "grad_norm": 0.8408545255661011, + "learning_rate": 2.5294046509019586e-05, + "loss": 0.8141, + "step": 17275 + }, + { + "epoch": 3.0756766381766383, + "grad_norm": 1.1111087799072266, + "learning_rate": 2.5284742321378585e-05, + "loss": 0.6377, + "step": 17276 + }, + { + "epoch": 3.075854700854701, + "grad_norm": 0.879222571849823, + "learning_rate": 2.5275439597627815e-05, + "loss": 0.6398, + "step": 17277 + }, + { + "epoch": 3.0760327635327633, + "grad_norm": 0.8783233165740967, + "learning_rate": 2.526613833794964e-05, + "loss": 0.7461, + "step": 17278 + }, + { + "epoch": 3.076210826210826, + "grad_norm": 0.9898136258125305, + "learning_rate": 2.5256838542526184e-05, + "loss": 0.724, + "step": 17279 + }, + { + "epoch": 3.076388888888889, + "grad_norm": 1.0655652284622192, + "learning_rate": 2.5247540211539744e-05, + "loss": 0.9039, + "step": 17280 + }, + { + "epoch": 3.0765669515669516, + "grad_norm": 0.952908992767334, + "learning_rate": 2.5238243345172464e-05, + "loss": 0.7654, + "step": 17281 + }, + { + "epoch": 3.0767450142450143, + "grad_norm": 1.056127667427063, + "learning_rate": 2.5228947943606494e-05, + "loss": 0.7084, + "step": 17282 + }, + { + "epoch": 3.076923076923077, + "grad_norm": 1.0711488723754883, + "learning_rate": 2.5219654007023973e-05, + "loss": 0.8788, + "step": 17283 + }, + { + "epoch": 3.0771011396011394, + "grad_norm": 0.9237586259841919, + "learning_rate": 2.5210361535606985e-05, + "loss": 0.7423, + "step": 17284 + }, + { + "epoch": 3.077279202279202, + "grad_norm": 0.8712356686592102, + "learning_rate": 2.5201070529537585e-05, + "loss": 0.7332, + "step": 17285 + }, + { + "epoch": 3.077457264957265, + "grad_norm": 0.9970635175704956, + "learning_rate": 2.5191780988997872e-05, + "loss": 0.8417, + "step": 17286 + }, + { + "epoch": 3.0776353276353277, + "grad_norm": 0.7353211045265198, + "learning_rate": 2.5182492914169754e-05, + "loss": 0.5351, + "step": 17287 + }, + { + "epoch": 3.0778133903133904, + "grad_norm": 0.8101412653923035, + "learning_rate": 2.5173206305235297e-05, + "loss": 0.674, + "step": 17288 + }, + { + "epoch": 3.077991452991453, + "grad_norm": 1.0728546380996704, + "learning_rate": 2.5163921162376425e-05, + "loss": 0.7514, + "step": 17289 + }, + { + "epoch": 3.0781695156695155, + "grad_norm": 1.05673086643219, + "learning_rate": 2.5154637485775058e-05, + "loss": 0.8012, + "step": 17290 + }, + { + "epoch": 3.078347578347578, + "grad_norm": 0.8431952595710754, + "learning_rate": 2.514535527561309e-05, + "loss": 0.6435, + "step": 17291 + }, + { + "epoch": 3.078525641025641, + "grad_norm": 0.9286113977432251, + "learning_rate": 2.5136074532072386e-05, + "loss": 0.8412, + "step": 17292 + }, + { + "epoch": 3.0787037037037037, + "grad_norm": 0.9323510527610779, + "learning_rate": 2.512679525533479e-05, + "loss": 0.7716, + "step": 17293 + }, + { + "epoch": 3.0788817663817665, + "grad_norm": 0.9141219854354858, + "learning_rate": 2.5117517445582107e-05, + "loss": 0.7361, + "step": 17294 + }, + { + "epoch": 3.0790598290598292, + "grad_norm": 0.8749071359634399, + "learning_rate": 2.5108241102996076e-05, + "loss": 0.6472, + "step": 17295 + }, + { + "epoch": 3.0792378917378915, + "grad_norm": 0.8831200003623962, + "learning_rate": 2.5098966227758525e-05, + "loss": 0.6902, + "step": 17296 + }, + { + "epoch": 3.0794159544159543, + "grad_norm": 0.8700500130653381, + "learning_rate": 2.5089692820051136e-05, + "loss": 0.6671, + "step": 17297 + }, + { + "epoch": 3.079594017094017, + "grad_norm": 0.865032970905304, + "learning_rate": 2.50804208800556e-05, + "loss": 0.5883, + "step": 17298 + }, + { + "epoch": 3.07977207977208, + "grad_norm": 1.0697903633117676, + "learning_rate": 2.5071150407953593e-05, + "loss": 0.6554, + "step": 17299 + }, + { + "epoch": 3.0799501424501425, + "grad_norm": 1.1248756647109985, + "learning_rate": 2.50618814039267e-05, + "loss": 0.7568, + "step": 17300 + }, + { + "epoch": 3.0801282051282053, + "grad_norm": 0.8852017521858215, + "learning_rate": 2.5052613868156638e-05, + "loss": 0.7013, + "step": 17301 + }, + { + "epoch": 3.0803062678062676, + "grad_norm": 0.8573511838912964, + "learning_rate": 2.5043347800824857e-05, + "loss": 0.7584, + "step": 17302 + }, + { + "epoch": 3.0804843304843303, + "grad_norm": 0.8905999660491943, + "learning_rate": 2.5034083202113034e-05, + "loss": 0.6698, + "step": 17303 + }, + { + "epoch": 3.080662393162393, + "grad_norm": 0.8821704387664795, + "learning_rate": 2.5024820072202548e-05, + "loss": 0.6776, + "step": 17304 + }, + { + "epoch": 3.080840455840456, + "grad_norm": 0.8579546809196472, + "learning_rate": 2.5015558411275008e-05, + "loss": 0.7955, + "step": 17305 + }, + { + "epoch": 3.0810185185185186, + "grad_norm": 0.8955085277557373, + "learning_rate": 2.5006298219511827e-05, + "loss": 0.6944, + "step": 17306 + }, + { + "epoch": 3.0811965811965814, + "grad_norm": 1.0479052066802979, + "learning_rate": 2.4997039497094453e-05, + "loss": 0.8585, + "step": 17307 + }, + { + "epoch": 3.0813746438746437, + "grad_norm": 0.8598275780677795, + "learning_rate": 2.4987782244204273e-05, + "loss": 0.6137, + "step": 17308 + }, + { + "epoch": 3.0815527065527064, + "grad_norm": 0.8625657558441162, + "learning_rate": 2.497852646102269e-05, + "loss": 0.7195, + "step": 17309 + }, + { + "epoch": 3.081730769230769, + "grad_norm": 0.945574939250946, + "learning_rate": 2.4969272147730994e-05, + "loss": 0.821, + "step": 17310 + }, + { + "epoch": 3.081908831908832, + "grad_norm": 0.7945711612701416, + "learning_rate": 2.4960019304510618e-05, + "loss": 0.5694, + "step": 17311 + }, + { + "epoch": 3.0820868945868947, + "grad_norm": 0.9468835592269897, + "learning_rate": 2.4950767931542717e-05, + "loss": 0.8095, + "step": 17312 + }, + { + "epoch": 3.0822649572649574, + "grad_norm": 0.9362178444862366, + "learning_rate": 2.494151802900867e-05, + "loss": 0.6649, + "step": 17313 + }, + { + "epoch": 3.08244301994302, + "grad_norm": 0.9418696761131287, + "learning_rate": 2.493226959708961e-05, + "loss": 0.9316, + "step": 17314 + }, + { + "epoch": 3.0826210826210825, + "grad_norm": 0.9696083664894104, + "learning_rate": 2.492302263596681e-05, + "loss": 0.6348, + "step": 17315 + }, + { + "epoch": 3.0827991452991452, + "grad_norm": 0.9243941903114319, + "learning_rate": 2.4913777145821427e-05, + "loss": 0.8841, + "step": 17316 + }, + { + "epoch": 3.082977207977208, + "grad_norm": 0.89748615026474, + "learning_rate": 2.4904533126834573e-05, + "loss": 0.6133, + "step": 17317 + }, + { + "epoch": 3.0831552706552707, + "grad_norm": 0.9694730043411255, + "learning_rate": 2.4895290579187446e-05, + "loss": 0.7748, + "step": 17318 + }, + { + "epoch": 3.0833333333333335, + "grad_norm": 0.8850114345550537, + "learning_rate": 2.4886049503061025e-05, + "loss": 0.6369, + "step": 17319 + }, + { + "epoch": 3.083511396011396, + "grad_norm": 0.7270629405975342, + "learning_rate": 2.4876809898636464e-05, + "loss": 0.5965, + "step": 17320 + }, + { + "epoch": 3.0836894586894585, + "grad_norm": 0.8992782831192017, + "learning_rate": 2.4867571766094764e-05, + "loss": 0.768, + "step": 17321 + }, + { + "epoch": 3.0838675213675213, + "grad_norm": 0.8755559921264648, + "learning_rate": 2.4858335105616916e-05, + "loss": 0.7047, + "step": 17322 + }, + { + "epoch": 3.084045584045584, + "grad_norm": 1.1143306493759155, + "learning_rate": 2.4849099917383888e-05, + "loss": 0.7245, + "step": 17323 + }, + { + "epoch": 3.084223646723647, + "grad_norm": 0.9730618000030518, + "learning_rate": 2.4839866201576646e-05, + "loss": 0.8453, + "step": 17324 + }, + { + "epoch": 3.0844017094017095, + "grad_norm": 0.8983954191207886, + "learning_rate": 2.483063395837606e-05, + "loss": 0.6329, + "step": 17325 + }, + { + "epoch": 3.0845797720797723, + "grad_norm": 0.9837406277656555, + "learning_rate": 2.482140318796311e-05, + "loss": 0.6922, + "step": 17326 + }, + { + "epoch": 3.0847578347578346, + "grad_norm": 0.9490154385566711, + "learning_rate": 2.4812173890518544e-05, + "loss": 0.9131, + "step": 17327 + }, + { + "epoch": 3.0849358974358974, + "grad_norm": 0.7831540107727051, + "learning_rate": 2.4802946066223287e-05, + "loss": 0.6089, + "step": 17328 + }, + { + "epoch": 3.08511396011396, + "grad_norm": 1.0531748533248901, + "learning_rate": 2.4793719715258044e-05, + "loss": 0.9125, + "step": 17329 + }, + { + "epoch": 3.085292022792023, + "grad_norm": 0.9238786697387695, + "learning_rate": 2.4784494837803663e-05, + "loss": 0.6231, + "step": 17330 + }, + { + "epoch": 3.0854700854700856, + "grad_norm": 0.826562762260437, + "learning_rate": 2.477527143404086e-05, + "loss": 0.8999, + "step": 17331 + }, + { + "epoch": 3.0856481481481484, + "grad_norm": 1.0574134588241577, + "learning_rate": 2.4766049504150335e-05, + "loss": 0.6379, + "step": 17332 + }, + { + "epoch": 3.0858262108262107, + "grad_norm": 0.8981440663337708, + "learning_rate": 2.47568290483128e-05, + "loss": 0.8848, + "step": 17333 + }, + { + "epoch": 3.0860042735042734, + "grad_norm": 0.8472706079483032, + "learning_rate": 2.4747610066708894e-05, + "loss": 0.886, + "step": 17334 + }, + { + "epoch": 3.086182336182336, + "grad_norm": 0.7722052335739136, + "learning_rate": 2.473839255951921e-05, + "loss": 0.5804, + "step": 17335 + }, + { + "epoch": 3.086360398860399, + "grad_norm": 1.0919874906539917, + "learning_rate": 2.4729176526924412e-05, + "loss": 0.8474, + "step": 17336 + }, + { + "epoch": 3.0865384615384617, + "grad_norm": 0.8331817388534546, + "learning_rate": 2.4719961969105042e-05, + "loss": 0.7727, + "step": 17337 + }, + { + "epoch": 3.0867165242165244, + "grad_norm": 0.7660060524940491, + "learning_rate": 2.471074888624163e-05, + "loss": 0.6989, + "step": 17338 + }, + { + "epoch": 3.0868945868945867, + "grad_norm": 1.0130395889282227, + "learning_rate": 2.4701537278514708e-05, + "loss": 0.6055, + "step": 17339 + }, + { + "epoch": 3.0870726495726495, + "grad_norm": 0.9906135201454163, + "learning_rate": 2.469232714610473e-05, + "loss": 0.7473, + "step": 17340 + }, + { + "epoch": 3.0872507122507122, + "grad_norm": 0.8437367677688599, + "learning_rate": 2.468311848919217e-05, + "loss": 0.5901, + "step": 17341 + }, + { + "epoch": 3.087428774928775, + "grad_norm": 0.9725748300552368, + "learning_rate": 2.467391130795741e-05, + "loss": 0.9002, + "step": 17342 + }, + { + "epoch": 3.0876068376068377, + "grad_norm": 0.8603026270866394, + "learning_rate": 2.466470560258094e-05, + "loss": 0.7682, + "step": 17343 + }, + { + "epoch": 3.0877849002849005, + "grad_norm": 0.9285373687744141, + "learning_rate": 2.4655501373243016e-05, + "loss": 0.8618, + "step": 17344 + }, + { + "epoch": 3.087962962962963, + "grad_norm": 1.0237677097320557, + "learning_rate": 2.4646298620124065e-05, + "loss": 0.9456, + "step": 17345 + }, + { + "epoch": 3.0881410256410255, + "grad_norm": 0.9248231649398804, + "learning_rate": 2.463709734340435e-05, + "loss": 0.5566, + "step": 17346 + }, + { + "epoch": 3.0883190883190883, + "grad_norm": 0.9666411280632019, + "learning_rate": 2.4627897543264154e-05, + "loss": 0.7949, + "step": 17347 + }, + { + "epoch": 3.088497150997151, + "grad_norm": 0.9193452000617981, + "learning_rate": 2.4618699219883735e-05, + "loss": 0.7006, + "step": 17348 + }, + { + "epoch": 3.088675213675214, + "grad_norm": 1.0161619186401367, + "learning_rate": 2.4609502373443316e-05, + "loss": 0.7966, + "step": 17349 + }, + { + "epoch": 3.0888532763532766, + "grad_norm": 0.9324184656143188, + "learning_rate": 2.460030700412309e-05, + "loss": 0.78, + "step": 17350 + }, + { + "epoch": 3.089031339031339, + "grad_norm": 0.9298717379570007, + "learning_rate": 2.4591113112103216e-05, + "loss": 0.8061, + "step": 17351 + }, + { + "epoch": 3.0892094017094016, + "grad_norm": 0.9238819479942322, + "learning_rate": 2.458192069756381e-05, + "loss": 0.8955, + "step": 17352 + }, + { + "epoch": 3.0893874643874644, + "grad_norm": 0.8168941736221313, + "learning_rate": 2.4572729760685052e-05, + "loss": 0.8462, + "step": 17353 + }, + { + "epoch": 3.089565527065527, + "grad_norm": 1.0559333562850952, + "learning_rate": 2.4563540301646914e-05, + "loss": 0.8891, + "step": 17354 + }, + { + "epoch": 3.08974358974359, + "grad_norm": 0.9265807271003723, + "learning_rate": 2.4554352320629525e-05, + "loss": 0.6554, + "step": 17355 + }, + { + "epoch": 3.0899216524216526, + "grad_norm": 0.9348501563072205, + "learning_rate": 2.4545165817812876e-05, + "loss": 0.6963, + "step": 17356 + }, + { + "epoch": 3.090099715099715, + "grad_norm": 0.942007839679718, + "learning_rate": 2.4535980793376922e-05, + "loss": 0.8307, + "step": 17357 + }, + { + "epoch": 3.0902777777777777, + "grad_norm": 1.119027018547058, + "learning_rate": 2.4526797247501732e-05, + "loss": 0.8015, + "step": 17358 + }, + { + "epoch": 3.0904558404558404, + "grad_norm": 0.9317795634269714, + "learning_rate": 2.4517615180367103e-05, + "loss": 0.6364, + "step": 17359 + }, + { + "epoch": 3.090633903133903, + "grad_norm": 1.03551185131073, + "learning_rate": 2.450843459215304e-05, + "loss": 0.8805, + "step": 17360 + }, + { + "epoch": 3.090811965811966, + "grad_norm": 0.8145185708999634, + "learning_rate": 2.4499255483039374e-05, + "loss": 0.5424, + "step": 17361 + }, + { + "epoch": 3.0909900284900287, + "grad_norm": 0.8325085639953613, + "learning_rate": 2.4490077853205962e-05, + "loss": 0.8728, + "step": 17362 + }, + { + "epoch": 3.091168091168091, + "grad_norm": 0.93165522813797, + "learning_rate": 2.4480901702832616e-05, + "loss": 0.7658, + "step": 17363 + }, + { + "epoch": 3.0913461538461537, + "grad_norm": 1.077810287475586, + "learning_rate": 2.4471727032099122e-05, + "loss": 0.909, + "step": 17364 + }, + { + "epoch": 3.0915242165242165, + "grad_norm": 0.9249957203865051, + "learning_rate": 2.4462553841185242e-05, + "loss": 0.683, + "step": 17365 + }, + { + "epoch": 3.0917022792022792, + "grad_norm": 0.9917559623718262, + "learning_rate": 2.4453382130270695e-05, + "loss": 0.7331, + "step": 17366 + }, + { + "epoch": 3.091880341880342, + "grad_norm": 0.8398951292037964, + "learning_rate": 2.4444211899535175e-05, + "loss": 0.8472, + "step": 17367 + }, + { + "epoch": 3.0920584045584047, + "grad_norm": 0.9382245540618896, + "learning_rate": 2.443504314915842e-05, + "loss": 0.9592, + "step": 17368 + }, + { + "epoch": 3.092236467236467, + "grad_norm": 0.9549033641815186, + "learning_rate": 2.442587587931997e-05, + "loss": 0.8117, + "step": 17369 + }, + { + "epoch": 3.09241452991453, + "grad_norm": 0.8004591464996338, + "learning_rate": 2.4416710090199512e-05, + "loss": 0.6495, + "step": 17370 + }, + { + "epoch": 3.0925925925925926, + "grad_norm": 0.8601853251457214, + "learning_rate": 2.4407545781976615e-05, + "loss": 0.6949, + "step": 17371 + }, + { + "epoch": 3.0927706552706553, + "grad_norm": 0.9627183079719543, + "learning_rate": 2.4398382954830823e-05, + "loss": 0.8233, + "step": 17372 + }, + { + "epoch": 3.092948717948718, + "grad_norm": 0.9439026117324829, + "learning_rate": 2.438922160894167e-05, + "loss": 0.7068, + "step": 17373 + }, + { + "epoch": 3.093126780626781, + "grad_norm": 1.1336002349853516, + "learning_rate": 2.438006174448866e-05, + "loss": 0.6664, + "step": 17374 + }, + { + "epoch": 3.093304843304843, + "grad_norm": 0.8688283562660217, + "learning_rate": 2.4370903361651243e-05, + "loss": 0.8014, + "step": 17375 + }, + { + "epoch": 3.093482905982906, + "grad_norm": 0.8023533821105957, + "learning_rate": 2.4361746460608836e-05, + "loss": 0.648, + "step": 17376 + }, + { + "epoch": 3.0936609686609686, + "grad_norm": 0.9410482048988342, + "learning_rate": 2.4352591041540918e-05, + "loss": 0.637, + "step": 17377 + }, + { + "epoch": 3.0938390313390314, + "grad_norm": 0.9243476390838623, + "learning_rate": 2.4343437104626844e-05, + "loss": 0.6791, + "step": 17378 + }, + { + "epoch": 3.094017094017094, + "grad_norm": 0.9308971762657166, + "learning_rate": 2.4334284650045947e-05, + "loss": 0.8286, + "step": 17379 + }, + { + "epoch": 3.094195156695157, + "grad_norm": 0.923256516456604, + "learning_rate": 2.4325133677977564e-05, + "loss": 0.6421, + "step": 17380 + }, + { + "epoch": 3.094373219373219, + "grad_norm": 0.9886062741279602, + "learning_rate": 2.431598418860098e-05, + "loss": 0.6816, + "step": 17381 + }, + { + "epoch": 3.094551282051282, + "grad_norm": 0.9735310673713684, + "learning_rate": 2.430683618209544e-05, + "loss": 0.7308, + "step": 17382 + }, + { + "epoch": 3.0947293447293447, + "grad_norm": 0.794840931892395, + "learning_rate": 2.4297689658640266e-05, + "loss": 0.7037, + "step": 17383 + }, + { + "epoch": 3.0949074074074074, + "grad_norm": 0.9691025018692017, + "learning_rate": 2.4288544618414543e-05, + "loss": 0.7427, + "step": 17384 + }, + { + "epoch": 3.09508547008547, + "grad_norm": 0.8925696611404419, + "learning_rate": 2.4279401061597574e-05, + "loss": 0.6131, + "step": 17385 + }, + { + "epoch": 3.095263532763533, + "grad_norm": 0.9027791619300842, + "learning_rate": 2.4270258988368376e-05, + "loss": 0.8347, + "step": 17386 + }, + { + "epoch": 3.0954415954415953, + "grad_norm": 0.935856282711029, + "learning_rate": 2.426111839890618e-05, + "loss": 0.7177, + "step": 17387 + }, + { + "epoch": 3.095619658119658, + "grad_norm": 1.0325665473937988, + "learning_rate": 2.4251979293390024e-05, + "loss": 0.7809, + "step": 17388 + }, + { + "epoch": 3.0957977207977208, + "grad_norm": 0.9081370234489441, + "learning_rate": 2.4242841671998996e-05, + "loss": 0.7376, + "step": 17389 + }, + { + "epoch": 3.0959757834757835, + "grad_norm": 0.9406909942626953, + "learning_rate": 2.42337055349121e-05, + "loss": 0.8681, + "step": 17390 + }, + { + "epoch": 3.0961538461538463, + "grad_norm": 0.9385591149330139, + "learning_rate": 2.4224570882308363e-05, + "loss": 0.721, + "step": 17391 + }, + { + "epoch": 3.096331908831909, + "grad_norm": 0.9210137724876404, + "learning_rate": 2.4215437714366717e-05, + "loss": 0.787, + "step": 17392 + }, + { + "epoch": 3.0965099715099713, + "grad_norm": 1.039331316947937, + "learning_rate": 2.420630603126619e-05, + "loss": 0.8778, + "step": 17393 + }, + { + "epoch": 3.096688034188034, + "grad_norm": 0.9159904718399048, + "learning_rate": 2.4197175833185603e-05, + "loss": 0.7282, + "step": 17394 + }, + { + "epoch": 3.096866096866097, + "grad_norm": 0.9137975573539734, + "learning_rate": 2.4188047120303915e-05, + "loss": 0.8241, + "step": 17395 + }, + { + "epoch": 3.0970441595441596, + "grad_norm": 0.9383588433265686, + "learning_rate": 2.4178919892799955e-05, + "loss": 0.7463, + "step": 17396 + }, + { + "epoch": 3.0972222222222223, + "grad_norm": 0.8108921051025391, + "learning_rate": 2.4169794150852553e-05, + "loss": 0.6284, + "step": 17397 + }, + { + "epoch": 3.097400284900285, + "grad_norm": 0.929779589176178, + "learning_rate": 2.416066989464052e-05, + "loss": 0.7413, + "step": 17398 + }, + { + "epoch": 3.0975783475783474, + "grad_norm": 0.8805593848228455, + "learning_rate": 2.4151547124342576e-05, + "loss": 0.6056, + "step": 17399 + }, + { + "epoch": 3.09775641025641, + "grad_norm": 0.8895778059959412, + "learning_rate": 2.4142425840137573e-05, + "loss": 0.7429, + "step": 17400 + }, + { + "epoch": 3.097934472934473, + "grad_norm": 0.8784146308898926, + "learning_rate": 2.4133306042204095e-05, + "loss": 0.6362, + "step": 17401 + }, + { + "epoch": 3.0981125356125356, + "grad_norm": 0.9039027690887451, + "learning_rate": 2.4124187730720917e-05, + "loss": 0.7523, + "step": 17402 + }, + { + "epoch": 3.0982905982905984, + "grad_norm": 1.021745204925537, + "learning_rate": 2.4115070905866653e-05, + "loss": 0.6444, + "step": 17403 + }, + { + "epoch": 3.098468660968661, + "grad_norm": 0.9708871245384216, + "learning_rate": 2.4105955567819937e-05, + "loss": 0.7391, + "step": 17404 + }, + { + "epoch": 3.0986467236467234, + "grad_norm": 0.8263231515884399, + "learning_rate": 2.4096841716759376e-05, + "loss": 0.7933, + "step": 17405 + }, + { + "epoch": 3.098824786324786, + "grad_norm": 0.9058948755264282, + "learning_rate": 2.408772935286352e-05, + "loss": 0.8139, + "step": 17406 + }, + { + "epoch": 3.099002849002849, + "grad_norm": 0.9710341691970825, + "learning_rate": 2.4078618476310888e-05, + "loss": 0.8429, + "step": 17407 + }, + { + "epoch": 3.0991809116809117, + "grad_norm": 0.8417289853096008, + "learning_rate": 2.4069509087280072e-05, + "loss": 0.6132, + "step": 17408 + }, + { + "epoch": 3.0993589743589745, + "grad_norm": 0.853482723236084, + "learning_rate": 2.406040118594943e-05, + "loss": 0.8565, + "step": 17409 + }, + { + "epoch": 3.099537037037037, + "grad_norm": 1.0863711833953857, + "learning_rate": 2.4051294772497536e-05, + "loss": 0.6898, + "step": 17410 + }, + { + "epoch": 3.0997150997150995, + "grad_norm": 0.9241563677787781, + "learning_rate": 2.4042189847102694e-05, + "loss": 0.8047, + "step": 17411 + }, + { + "epoch": 3.0998931623931623, + "grad_norm": 0.9109302759170532, + "learning_rate": 2.4033086409943384e-05, + "loss": 0.597, + "step": 17412 + }, + { + "epoch": 3.100071225071225, + "grad_norm": 0.8546639084815979, + "learning_rate": 2.4023984461197936e-05, + "loss": 0.6349, + "step": 17413 + }, + { + "epoch": 3.1002492877492878, + "grad_norm": 0.9716140627861023, + "learning_rate": 2.4014884001044692e-05, + "loss": 0.61, + "step": 17414 + }, + { + "epoch": 3.1004273504273505, + "grad_norm": 0.8704510927200317, + "learning_rate": 2.4005785029661944e-05, + "loss": 0.6373, + "step": 17415 + }, + { + "epoch": 3.1006054131054133, + "grad_norm": 0.933198869228363, + "learning_rate": 2.3996687547227948e-05, + "loss": 0.8694, + "step": 17416 + }, + { + "epoch": 3.1007834757834756, + "grad_norm": 0.9596254825592041, + "learning_rate": 2.3987591553920997e-05, + "loss": 0.9679, + "step": 17417 + }, + { + "epoch": 3.1009615384615383, + "grad_norm": 0.8634598255157471, + "learning_rate": 2.397849704991929e-05, + "loss": 0.7072, + "step": 17418 + }, + { + "epoch": 3.101139601139601, + "grad_norm": 0.9921422004699707, + "learning_rate": 2.3969404035401023e-05, + "loss": 0.8952, + "step": 17419 + }, + { + "epoch": 3.101317663817664, + "grad_norm": 0.9247313141822815, + "learning_rate": 2.3960312510544336e-05, + "loss": 0.856, + "step": 17420 + }, + { + "epoch": 3.1014957264957266, + "grad_norm": 0.8624486327171326, + "learning_rate": 2.395122247552737e-05, + "loss": 0.6163, + "step": 17421 + }, + { + "epoch": 3.1016737891737893, + "grad_norm": 0.9683988690376282, + "learning_rate": 2.394213393052822e-05, + "loss": 0.7084, + "step": 17422 + }, + { + "epoch": 3.1018518518518516, + "grad_norm": 0.8543943166732788, + "learning_rate": 2.3933046875724964e-05, + "loss": 0.6083, + "step": 17423 + }, + { + "epoch": 3.1020299145299144, + "grad_norm": 0.9302839040756226, + "learning_rate": 2.3923961311295596e-05, + "loss": 0.6842, + "step": 17424 + }, + { + "epoch": 3.102207977207977, + "grad_norm": 0.9772568345069885, + "learning_rate": 2.3914877237418244e-05, + "loss": 0.7136, + "step": 17425 + }, + { + "epoch": 3.10238603988604, + "grad_norm": 0.8349552750587463, + "learning_rate": 2.3905794654270753e-05, + "loss": 0.6465, + "step": 17426 + }, + { + "epoch": 3.1025641025641026, + "grad_norm": 0.9706963300704956, + "learning_rate": 2.3896713562031182e-05, + "loss": 0.8294, + "step": 17427 + }, + { + "epoch": 3.1027421652421654, + "grad_norm": 0.9962044954299927, + "learning_rate": 2.388763396087741e-05, + "loss": 0.7717, + "step": 17428 + }, + { + "epoch": 3.1029202279202277, + "grad_norm": 1.0264992713928223, + "learning_rate": 2.387855585098734e-05, + "loss": 0.7093, + "step": 17429 + }, + { + "epoch": 3.1030982905982905, + "grad_norm": 0.8996137976646423, + "learning_rate": 2.386947923253884e-05, + "loss": 0.7551, + "step": 17430 + }, + { + "epoch": 3.103276353276353, + "grad_norm": 0.8956265449523926, + "learning_rate": 2.3860404105709756e-05, + "loss": 0.7528, + "step": 17431 + }, + { + "epoch": 3.103454415954416, + "grad_norm": 0.8972591757774353, + "learning_rate": 2.3851330470677847e-05, + "loss": 0.9049, + "step": 17432 + }, + { + "epoch": 3.1036324786324787, + "grad_norm": 0.9835371375083923, + "learning_rate": 2.3842258327620993e-05, + "loss": 0.679, + "step": 17433 + }, + { + "epoch": 3.1038105413105415, + "grad_norm": 0.9487658739089966, + "learning_rate": 2.383318767671683e-05, + "loss": 0.626, + "step": 17434 + }, + { + "epoch": 3.103988603988604, + "grad_norm": 0.9638952016830444, + "learning_rate": 2.3824118518143156e-05, + "loss": 0.8995, + "step": 17435 + }, + { + "epoch": 3.1041666666666665, + "grad_norm": 1.085768699645996, + "learning_rate": 2.3815050852077637e-05, + "loss": 0.5734, + "step": 17436 + }, + { + "epoch": 3.1043447293447293, + "grad_norm": 0.9774207472801208, + "learning_rate": 2.3805984678697934e-05, + "loss": 0.9972, + "step": 17437 + }, + { + "epoch": 3.104522792022792, + "grad_norm": 0.8640549182891846, + "learning_rate": 2.379691999818169e-05, + "loss": 0.6142, + "step": 17438 + }, + { + "epoch": 3.1047008547008548, + "grad_norm": 0.8625344634056091, + "learning_rate": 2.3787856810706465e-05, + "loss": 0.8086, + "step": 17439 + }, + { + "epoch": 3.1048789173789175, + "grad_norm": 1.0197455883026123, + "learning_rate": 2.3778795116449937e-05, + "loss": 0.7878, + "step": 17440 + }, + { + "epoch": 3.10505698005698, + "grad_norm": 0.9929977059364319, + "learning_rate": 2.3769734915589514e-05, + "loss": 0.834, + "step": 17441 + }, + { + "epoch": 3.1052350427350426, + "grad_norm": 0.9360977411270142, + "learning_rate": 2.3760676208302812e-05, + "loss": 0.6217, + "step": 17442 + }, + { + "epoch": 3.1054131054131053, + "grad_norm": 0.9955264925956726, + "learning_rate": 2.3751618994767288e-05, + "loss": 0.8329, + "step": 17443 + }, + { + "epoch": 3.105591168091168, + "grad_norm": 0.7187397480010986, + "learning_rate": 2.3742563275160402e-05, + "loss": 0.5394, + "step": 17444 + }, + { + "epoch": 3.105769230769231, + "grad_norm": 0.8444333076477051, + "learning_rate": 2.3733509049659563e-05, + "loss": 0.7639, + "step": 17445 + }, + { + "epoch": 3.1059472934472936, + "grad_norm": 0.7043599486351013, + "learning_rate": 2.3724456318442202e-05, + "loss": 0.4893, + "step": 17446 + }, + { + "epoch": 3.1061253561253563, + "grad_norm": 1.1488189697265625, + "learning_rate": 2.371540508168566e-05, + "loss": 0.6633, + "step": 17447 + }, + { + "epoch": 3.1063034188034186, + "grad_norm": 0.9220834374427795, + "learning_rate": 2.3706355339567286e-05, + "loss": 0.7883, + "step": 17448 + }, + { + "epoch": 3.1064814814814814, + "grad_norm": 0.8987318873405457, + "learning_rate": 2.3697307092264375e-05, + "loss": 0.6756, + "step": 17449 + }, + { + "epoch": 3.106659544159544, + "grad_norm": 1.0003935098648071, + "learning_rate": 2.368826033995427e-05, + "loss": 0.8456, + "step": 17450 + }, + { + "epoch": 3.106837606837607, + "grad_norm": 0.9491976499557495, + "learning_rate": 2.3679215082814133e-05, + "loss": 0.8256, + "step": 17451 + }, + { + "epoch": 3.1070156695156697, + "grad_norm": 0.9657387137413025, + "learning_rate": 2.3670171321021262e-05, + "loss": 0.8526, + "step": 17452 + }, + { + "epoch": 3.1071937321937324, + "grad_norm": 0.863231897354126, + "learning_rate": 2.366112905475282e-05, + "loss": 0.7528, + "step": 17453 + }, + { + "epoch": 3.1073717948717947, + "grad_norm": 0.9410921335220337, + "learning_rate": 2.3652088284185934e-05, + "loss": 0.8809, + "step": 17454 + }, + { + "epoch": 3.1075498575498575, + "grad_norm": 0.940887987613678, + "learning_rate": 2.3643049009497853e-05, + "loss": 0.6853, + "step": 17455 + }, + { + "epoch": 3.10772792022792, + "grad_norm": 0.9488041400909424, + "learning_rate": 2.363401123086555e-05, + "loss": 0.8417, + "step": 17456 + }, + { + "epoch": 3.107905982905983, + "grad_norm": 0.8554413914680481, + "learning_rate": 2.3624974948466204e-05, + "loss": 0.75, + "step": 17457 + }, + { + "epoch": 3.1080840455840457, + "grad_norm": 0.8631907105445862, + "learning_rate": 2.361594016247678e-05, + "loss": 0.6059, + "step": 17458 + }, + { + "epoch": 3.1082621082621085, + "grad_norm": 0.9793382883071899, + "learning_rate": 2.3606906873074354e-05, + "loss": 0.7293, + "step": 17459 + }, + { + "epoch": 3.1084401709401708, + "grad_norm": 0.8650250434875488, + "learning_rate": 2.3597875080435883e-05, + "loss": 0.6946, + "step": 17460 + }, + { + "epoch": 3.1086182336182335, + "grad_norm": 0.954089343547821, + "learning_rate": 2.358884478473835e-05, + "loss": 0.9179, + "step": 17461 + }, + { + "epoch": 3.1087962962962963, + "grad_norm": 0.96575528383255, + "learning_rate": 2.357981598615867e-05, + "loss": 0.8668, + "step": 17462 + }, + { + "epoch": 3.108974358974359, + "grad_norm": 0.9860404133796692, + "learning_rate": 2.3570788684873757e-05, + "loss": 0.8166, + "step": 17463 + }, + { + "epoch": 3.109152421652422, + "grad_norm": 1.058204174041748, + "learning_rate": 2.3561762881060434e-05, + "loss": 0.8283, + "step": 17464 + }, + { + "epoch": 3.1093304843304845, + "grad_norm": 0.9833809733390808, + "learning_rate": 2.355273857489564e-05, + "loss": 0.8027, + "step": 17465 + }, + { + "epoch": 3.109508547008547, + "grad_norm": 0.9593403339385986, + "learning_rate": 2.3543715766556073e-05, + "loss": 0.713, + "step": 17466 + }, + { + "epoch": 3.1096866096866096, + "grad_norm": 1.0403105020523071, + "learning_rate": 2.3534694456218587e-05, + "loss": 0.8695, + "step": 17467 + }, + { + "epoch": 3.1098646723646723, + "grad_norm": 0.9821664094924927, + "learning_rate": 2.3525674644059936e-05, + "loss": 0.8268, + "step": 17468 + }, + { + "epoch": 3.110042735042735, + "grad_norm": 1.0293829441070557, + "learning_rate": 2.3516656330256826e-05, + "loss": 0.7902, + "step": 17469 + }, + { + "epoch": 3.110220797720798, + "grad_norm": 1.0120795965194702, + "learning_rate": 2.3507639514985947e-05, + "loss": 0.8086, + "step": 17470 + }, + { + "epoch": 3.1103988603988606, + "grad_norm": 1.0089393854141235, + "learning_rate": 2.349862419842398e-05, + "loss": 0.8249, + "step": 17471 + }, + { + "epoch": 3.110576923076923, + "grad_norm": 0.908614456653595, + "learning_rate": 2.348961038074755e-05, + "loss": 0.7212, + "step": 17472 + }, + { + "epoch": 3.1107549857549857, + "grad_norm": 0.9755021333694458, + "learning_rate": 2.3480598062133263e-05, + "loss": 0.7566, + "step": 17473 + }, + { + "epoch": 3.1109330484330484, + "grad_norm": 0.7833260893821716, + "learning_rate": 2.3471587242757686e-05, + "loss": 0.5381, + "step": 17474 + }, + { + "epoch": 3.111111111111111, + "grad_norm": 1.0489907264709473, + "learning_rate": 2.3462577922797403e-05, + "loss": 0.9524, + "step": 17475 + }, + { + "epoch": 3.111289173789174, + "grad_norm": 0.9572435021400452, + "learning_rate": 2.3453570102428912e-05, + "loss": 0.7576, + "step": 17476 + }, + { + "epoch": 3.1114672364672367, + "grad_norm": 1.0597792863845825, + "learning_rate": 2.344456378182871e-05, + "loss": 0.6565, + "step": 17477 + }, + { + "epoch": 3.111645299145299, + "grad_norm": 0.9518697261810303, + "learning_rate": 2.3435558961173244e-05, + "loss": 0.8274, + "step": 17478 + }, + { + "epoch": 3.1118233618233617, + "grad_norm": 0.8185853958129883, + "learning_rate": 2.3426555640638927e-05, + "loss": 0.7058, + "step": 17479 + }, + { + "epoch": 3.1120014245014245, + "grad_norm": 0.9529431462287903, + "learning_rate": 2.3417553820402237e-05, + "loss": 0.8802, + "step": 17480 + }, + { + "epoch": 3.1121794871794872, + "grad_norm": 1.0432971715927124, + "learning_rate": 2.340855350063944e-05, + "loss": 0.853, + "step": 17481 + }, + { + "epoch": 3.11235754985755, + "grad_norm": 0.8705558180809021, + "learning_rate": 2.339955468152699e-05, + "loss": 0.6116, + "step": 17482 + }, + { + "epoch": 3.1125356125356127, + "grad_norm": 0.8967922329902649, + "learning_rate": 2.3390557363241082e-05, + "loss": 0.7099, + "step": 17483 + }, + { + "epoch": 3.112713675213675, + "grad_norm": 1.0413905382156372, + "learning_rate": 2.3381561545958098e-05, + "loss": 0.8544, + "step": 17484 + }, + { + "epoch": 3.112891737891738, + "grad_norm": 0.831136167049408, + "learning_rate": 2.337256722985425e-05, + "loss": 0.7654, + "step": 17485 + }, + { + "epoch": 3.1130698005698005, + "grad_norm": 1.002241611480713, + "learning_rate": 2.336357441510576e-05, + "loss": 0.831, + "step": 17486 + }, + { + "epoch": 3.1132478632478633, + "grad_norm": 0.9891117215156555, + "learning_rate": 2.335458310188884e-05, + "loss": 0.7456, + "step": 17487 + }, + { + "epoch": 3.113425925925926, + "grad_norm": 0.8876535296440125, + "learning_rate": 2.334559329037964e-05, + "loss": 0.7195, + "step": 17488 + }, + { + "epoch": 3.113603988603989, + "grad_norm": 1.015412449836731, + "learning_rate": 2.3336604980754283e-05, + "loss": 0.8297, + "step": 17489 + }, + { + "epoch": 3.113782051282051, + "grad_norm": 0.8936545252799988, + "learning_rate": 2.3327618173188947e-05, + "loss": 0.6681, + "step": 17490 + }, + { + "epoch": 3.113960113960114, + "grad_norm": 1.0073288679122925, + "learning_rate": 2.3318632867859602e-05, + "loss": 0.6266, + "step": 17491 + }, + { + "epoch": 3.1141381766381766, + "grad_norm": 0.7852414846420288, + "learning_rate": 2.3309649064942418e-05, + "loss": 0.4913, + "step": 17492 + }, + { + "epoch": 3.1143162393162394, + "grad_norm": 0.9297264218330383, + "learning_rate": 2.330066676461329e-05, + "loss": 0.6447, + "step": 17493 + }, + { + "epoch": 3.114494301994302, + "grad_norm": 0.9050962924957275, + "learning_rate": 2.3291685967048295e-05, + "loss": 0.7686, + "step": 17494 + }, + { + "epoch": 3.114672364672365, + "grad_norm": 0.977013885974884, + "learning_rate": 2.328270667242336e-05, + "loss": 0.8068, + "step": 17495 + }, + { + "epoch": 3.114850427350427, + "grad_norm": 1.0976142883300781, + "learning_rate": 2.32737288809144e-05, + "loss": 0.8486, + "step": 17496 + }, + { + "epoch": 3.11502849002849, + "grad_norm": 1.098183274269104, + "learning_rate": 2.3264752592697393e-05, + "loss": 1.0246, + "step": 17497 + }, + { + "epoch": 3.1152065527065527, + "grad_norm": 0.9509225487709045, + "learning_rate": 2.3255777807948098e-05, + "loss": 0.8665, + "step": 17498 + }, + { + "epoch": 3.1153846153846154, + "grad_norm": 0.8583039045333862, + "learning_rate": 2.324680452684245e-05, + "loss": 0.7259, + "step": 17499 + }, + { + "epoch": 3.115562678062678, + "grad_norm": 0.8832426071166992, + "learning_rate": 2.323783274955622e-05, + "loss": 0.7073, + "step": 17500 + }, + { + "epoch": 3.115740740740741, + "grad_norm": 0.910290539264679, + "learning_rate": 2.32288624762652e-05, + "loss": 0.7346, + "step": 17501 + }, + { + "epoch": 3.1159188034188032, + "grad_norm": 0.9926977753639221, + "learning_rate": 2.3219893707145136e-05, + "loss": 0.7527, + "step": 17502 + }, + { + "epoch": 3.116096866096866, + "grad_norm": 0.9508365392684937, + "learning_rate": 2.3210926442371762e-05, + "loss": 0.6538, + "step": 17503 + }, + { + "epoch": 3.1162749287749287, + "grad_norm": 0.8371008634567261, + "learning_rate": 2.3201960682120738e-05, + "loss": 0.7429, + "step": 17504 + }, + { + "epoch": 3.1164529914529915, + "grad_norm": 0.9090998768806458, + "learning_rate": 2.3192996426567827e-05, + "loss": 0.8299, + "step": 17505 + }, + { + "epoch": 3.1166310541310542, + "grad_norm": 1.0013576745986938, + "learning_rate": 2.3184033675888528e-05, + "loss": 0.7982, + "step": 17506 + }, + { + "epoch": 3.116809116809117, + "grad_norm": 1.021876335144043, + "learning_rate": 2.3175072430258583e-05, + "loss": 0.817, + "step": 17507 + }, + { + "epoch": 3.1169871794871793, + "grad_norm": 1.0198882818222046, + "learning_rate": 2.316611268985345e-05, + "loss": 0.7713, + "step": 17508 + }, + { + "epoch": 3.117165242165242, + "grad_norm": 0.8792022466659546, + "learning_rate": 2.315715445484875e-05, + "loss": 0.6666, + "step": 17509 + }, + { + "epoch": 3.117343304843305, + "grad_norm": 0.9849581718444824, + "learning_rate": 2.3148197725419983e-05, + "loss": 0.9193, + "step": 17510 + }, + { + "epoch": 3.1175213675213675, + "grad_norm": 1.1473517417907715, + "learning_rate": 2.313924250174263e-05, + "loss": 0.7998, + "step": 17511 + }, + { + "epoch": 3.1176994301994303, + "grad_norm": 0.9856370091438293, + "learning_rate": 2.3130288783992173e-05, + "loss": 0.8371, + "step": 17512 + }, + { + "epoch": 3.117877492877493, + "grad_norm": 0.9975123405456543, + "learning_rate": 2.312133657234401e-05, + "loss": 0.9139, + "step": 17513 + }, + { + "epoch": 3.1180555555555554, + "grad_norm": 0.8673362731933594, + "learning_rate": 2.311238586697353e-05, + "loss": 0.7295, + "step": 17514 + }, + { + "epoch": 3.118233618233618, + "grad_norm": 0.9963995218276978, + "learning_rate": 2.3103436668056154e-05, + "loss": 0.7931, + "step": 17515 + }, + { + "epoch": 3.118411680911681, + "grad_norm": 0.9084660410881042, + "learning_rate": 2.3094488975767204e-05, + "loss": 0.7494, + "step": 17516 + }, + { + "epoch": 3.1185897435897436, + "grad_norm": 0.9985509514808655, + "learning_rate": 2.308554279028199e-05, + "loss": 0.7185, + "step": 17517 + }, + { + "epoch": 3.1187678062678064, + "grad_norm": 0.929409384727478, + "learning_rate": 2.3076598111775772e-05, + "loss": 0.7564, + "step": 17518 + }, + { + "epoch": 3.118945868945869, + "grad_norm": 0.8911257386207581, + "learning_rate": 2.3067654940423832e-05, + "loss": 0.8528, + "step": 17519 + }, + { + "epoch": 3.1191239316239314, + "grad_norm": 0.9538902640342712, + "learning_rate": 2.3058713276401378e-05, + "loss": 0.767, + "step": 17520 + }, + { + "epoch": 3.119301994301994, + "grad_norm": 0.9277068376541138, + "learning_rate": 2.304977311988358e-05, + "loss": 0.8786, + "step": 17521 + }, + { + "epoch": 3.119480056980057, + "grad_norm": 0.9150798916816711, + "learning_rate": 2.304083447104568e-05, + "loss": 0.7237, + "step": 17522 + }, + { + "epoch": 3.1196581196581197, + "grad_norm": 0.9311725497245789, + "learning_rate": 2.3031897330062702e-05, + "loss": 0.8545, + "step": 17523 + }, + { + "epoch": 3.1198361823361824, + "grad_norm": 1.0069408416748047, + "learning_rate": 2.3022961697109847e-05, + "loss": 0.6613, + "step": 17524 + }, + { + "epoch": 3.120014245014245, + "grad_norm": 0.8878635168075562, + "learning_rate": 2.301402757236214e-05, + "loss": 0.8657, + "step": 17525 + }, + { + "epoch": 3.1201923076923075, + "grad_norm": 0.9363601803779602, + "learning_rate": 2.3005094955994643e-05, + "loss": 0.7536, + "step": 17526 + }, + { + "epoch": 3.1203703703703702, + "grad_norm": 0.9082810282707214, + "learning_rate": 2.299616384818236e-05, + "loss": 0.7054, + "step": 17527 + }, + { + "epoch": 3.120548433048433, + "grad_norm": 0.9305253624916077, + "learning_rate": 2.2987234249100296e-05, + "loss": 0.7509, + "step": 17528 + }, + { + "epoch": 3.1207264957264957, + "grad_norm": 0.9521283507347107, + "learning_rate": 2.297830615892339e-05, + "loss": 0.7455, + "step": 17529 + }, + { + "epoch": 3.1209045584045585, + "grad_norm": 1.0800763368606567, + "learning_rate": 2.2969379577826578e-05, + "loss": 0.9612, + "step": 17530 + }, + { + "epoch": 3.1210826210826212, + "grad_norm": 0.8735892176628113, + "learning_rate": 2.296045450598473e-05, + "loss": 0.7061, + "step": 17531 + }, + { + "epoch": 3.1212606837606836, + "grad_norm": 0.9851289391517639, + "learning_rate": 2.2951530943572796e-05, + "loss": 0.5958, + "step": 17532 + }, + { + "epoch": 3.1214387464387463, + "grad_norm": 1.1228593587875366, + "learning_rate": 2.2942608890765494e-05, + "loss": 0.7061, + "step": 17533 + }, + { + "epoch": 3.121616809116809, + "grad_norm": 0.8875393867492676, + "learning_rate": 2.2933688347737736e-05, + "loss": 0.7959, + "step": 17534 + }, + { + "epoch": 3.121794871794872, + "grad_norm": 0.8963145613670349, + "learning_rate": 2.2924769314664262e-05, + "loss": 0.6042, + "step": 17535 + }, + { + "epoch": 3.1219729344729346, + "grad_norm": 0.8165764808654785, + "learning_rate": 2.291585179171979e-05, + "loss": 0.6797, + "step": 17536 + }, + { + "epoch": 3.1221509971509973, + "grad_norm": 0.9140230417251587, + "learning_rate": 2.290693577907913e-05, + "loss": 0.8226, + "step": 17537 + }, + { + "epoch": 3.1223290598290596, + "grad_norm": 0.9492660760879517, + "learning_rate": 2.289802127691687e-05, + "loss": 0.909, + "step": 17538 + }, + { + "epoch": 3.1225071225071224, + "grad_norm": 0.9075984358787537, + "learning_rate": 2.2889108285407734e-05, + "loss": 0.8317, + "step": 17539 + }, + { + "epoch": 3.122685185185185, + "grad_norm": 1.0724934339523315, + "learning_rate": 2.288019680472635e-05, + "loss": 0.9833, + "step": 17540 + }, + { + "epoch": 3.122863247863248, + "grad_norm": 0.9564105272293091, + "learning_rate": 2.287128683504729e-05, + "loss": 0.9504, + "step": 17541 + }, + { + "epoch": 3.1230413105413106, + "grad_norm": 0.8370453119277954, + "learning_rate": 2.2862378376545156e-05, + "loss": 0.6957, + "step": 17542 + }, + { + "epoch": 3.1232193732193734, + "grad_norm": 0.926626980304718, + "learning_rate": 2.285347142939448e-05, + "loss": 0.7962, + "step": 17543 + }, + { + "epoch": 3.123397435897436, + "grad_norm": 0.8979136347770691, + "learning_rate": 2.2844565993769763e-05, + "loss": 0.7981, + "step": 17544 + }, + { + "epoch": 3.1235754985754984, + "grad_norm": 0.8645342588424683, + "learning_rate": 2.283566206984551e-05, + "loss": 0.7094, + "step": 17545 + }, + { + "epoch": 3.123753561253561, + "grad_norm": 0.8856910467147827, + "learning_rate": 2.2826759657796125e-05, + "loss": 0.7623, + "step": 17546 + }, + { + "epoch": 3.123931623931624, + "grad_norm": 0.9834395051002502, + "learning_rate": 2.2817858757796128e-05, + "loss": 0.6267, + "step": 17547 + }, + { + "epoch": 3.1241096866096867, + "grad_norm": 0.8917219042778015, + "learning_rate": 2.280895937001981e-05, + "loss": 0.636, + "step": 17548 + }, + { + "epoch": 3.1242877492877494, + "grad_norm": 1.0435439348220825, + "learning_rate": 2.28000614946416e-05, + "loss": 0.6955, + "step": 17549 + }, + { + "epoch": 3.1244658119658117, + "grad_norm": 0.8975854516029358, + "learning_rate": 2.2791165131835824e-05, + "loss": 0.616, + "step": 17550 + }, + { + "epoch": 3.1246438746438745, + "grad_norm": 0.9830999374389648, + "learning_rate": 2.2782270281776774e-05, + "loss": 0.7004, + "step": 17551 + }, + { + "epoch": 3.1248219373219372, + "grad_norm": 0.9302573204040527, + "learning_rate": 2.2773376944638735e-05, + "loss": 0.8539, + "step": 17552 + }, + { + "epoch": 3.125, + "grad_norm": 0.9269157648086548, + "learning_rate": 2.2764485120595924e-05, + "loss": 0.6907, + "step": 17553 + }, + { + "epoch": 3.1251780626780628, + "grad_norm": 0.9939448833465576, + "learning_rate": 2.2755594809822644e-05, + "loss": 0.7269, + "step": 17554 + }, + { + "epoch": 3.1253561253561255, + "grad_norm": 0.9136678576469421, + "learning_rate": 2.2746706012492957e-05, + "loss": 0.9148, + "step": 17555 + }, + { + "epoch": 3.1255341880341883, + "grad_norm": 0.927659273147583, + "learning_rate": 2.2737818728781124e-05, + "loss": 0.7202, + "step": 17556 + }, + { + "epoch": 3.1257122507122506, + "grad_norm": 0.8123641014099121, + "learning_rate": 2.2728932958861228e-05, + "loss": 0.6804, + "step": 17557 + }, + { + "epoch": 3.1258903133903133, + "grad_norm": 0.8486707806587219, + "learning_rate": 2.272004870290737e-05, + "loss": 0.5989, + "step": 17558 + }, + { + "epoch": 3.126068376068376, + "grad_norm": 1.0355206727981567, + "learning_rate": 2.271116596109363e-05, + "loss": 0.8661, + "step": 17559 + }, + { + "epoch": 3.126246438746439, + "grad_norm": 0.947636604309082, + "learning_rate": 2.2702284733594037e-05, + "loss": 0.742, + "step": 17560 + }, + { + "epoch": 3.1264245014245016, + "grad_norm": 0.9906873106956482, + "learning_rate": 2.269340502058257e-05, + "loss": 0.8523, + "step": 17561 + }, + { + "epoch": 3.126602564102564, + "grad_norm": 0.8627552390098572, + "learning_rate": 2.2684526822233297e-05, + "loss": 0.8379, + "step": 17562 + }, + { + "epoch": 3.1267806267806266, + "grad_norm": 0.9640594720840454, + "learning_rate": 2.267565013872006e-05, + "loss": 0.7408, + "step": 17563 + }, + { + "epoch": 3.1269586894586894, + "grad_norm": 0.9448180794715881, + "learning_rate": 2.2666774970216888e-05, + "loss": 0.8076, + "step": 17564 + }, + { + "epoch": 3.127136752136752, + "grad_norm": 1.1045911312103271, + "learning_rate": 2.2657901316897555e-05, + "loss": 0.8835, + "step": 17565 + }, + { + "epoch": 3.127314814814815, + "grad_norm": 0.9843702912330627, + "learning_rate": 2.264902917893602e-05, + "loss": 0.8285, + "step": 17566 + }, + { + "epoch": 3.1274928774928776, + "grad_norm": 0.913769543170929, + "learning_rate": 2.2640158556506063e-05, + "loss": 0.8045, + "step": 17567 + }, + { + "epoch": 3.1276709401709404, + "grad_norm": 0.9181899428367615, + "learning_rate": 2.26312894497815e-05, + "loss": 0.7648, + "step": 17568 + }, + { + "epoch": 3.1278490028490027, + "grad_norm": 0.8163581490516663, + "learning_rate": 2.2622421858936104e-05, + "loss": 0.6892, + "step": 17569 + }, + { + "epoch": 3.1280270655270654, + "grad_norm": 0.8715813755989075, + "learning_rate": 2.2613555784143614e-05, + "loss": 0.7745, + "step": 17570 + }, + { + "epoch": 3.128205128205128, + "grad_norm": 1.0851328372955322, + "learning_rate": 2.260469122557771e-05, + "loss": 0.8486, + "step": 17571 + }, + { + "epoch": 3.128383190883191, + "grad_norm": 1.0961211919784546, + "learning_rate": 2.2595828183412172e-05, + "loss": 0.6314, + "step": 17572 + }, + { + "epoch": 3.1285612535612537, + "grad_norm": 1.1079659461975098, + "learning_rate": 2.2586966657820518e-05, + "loss": 0.8251, + "step": 17573 + }, + { + "epoch": 3.128739316239316, + "grad_norm": 0.8718203902244568, + "learning_rate": 2.257810664897648e-05, + "loss": 0.6977, + "step": 17574 + }, + { + "epoch": 3.1289173789173788, + "grad_norm": 0.7854416370391846, + "learning_rate": 2.2569248157053602e-05, + "loss": 0.584, + "step": 17575 + }, + { + "epoch": 3.1290954415954415, + "grad_norm": 1.1214252710342407, + "learning_rate": 2.2560391182225437e-05, + "loss": 0.8795, + "step": 17576 + }, + { + "epoch": 3.1292735042735043, + "grad_norm": 1.018040657043457, + "learning_rate": 2.2551535724665596e-05, + "loss": 0.7597, + "step": 17577 + }, + { + "epoch": 3.129451566951567, + "grad_norm": 0.9986835718154907, + "learning_rate": 2.254268178454747e-05, + "loss": 0.7418, + "step": 17578 + }, + { + "epoch": 3.1296296296296298, + "grad_norm": 0.9148978590965271, + "learning_rate": 2.2533829362044643e-05, + "loss": 0.7763, + "step": 17579 + }, + { + "epoch": 3.1298076923076925, + "grad_norm": 1.0832401514053345, + "learning_rate": 2.2524978457330458e-05, + "loss": 0.8144, + "step": 17580 + }, + { + "epoch": 3.129985754985755, + "grad_norm": 0.9290724396705627, + "learning_rate": 2.2516129070578406e-05, + "loss": 0.8129, + "step": 17581 + }, + { + "epoch": 3.1301638176638176, + "grad_norm": 0.8968502283096313, + "learning_rate": 2.250728120196184e-05, + "loss": 0.8565, + "step": 17582 + }, + { + "epoch": 3.1303418803418803, + "grad_norm": 0.8710499405860901, + "learning_rate": 2.2498434851654126e-05, + "loss": 0.7707, + "step": 17583 + }, + { + "epoch": 3.130519943019943, + "grad_norm": 0.9112555384635925, + "learning_rate": 2.248959001982859e-05, + "loss": 0.6694, + "step": 17584 + }, + { + "epoch": 3.130698005698006, + "grad_norm": 0.9539666771888733, + "learning_rate": 2.2480746706658516e-05, + "loss": 0.79, + "step": 17585 + }, + { + "epoch": 3.1308760683760686, + "grad_norm": 0.8673475384712219, + "learning_rate": 2.2471904912317153e-05, + "loss": 0.7533, + "step": 17586 + }, + { + "epoch": 3.131054131054131, + "grad_norm": 1.044456958770752, + "learning_rate": 2.2463064636977826e-05, + "loss": 0.851, + "step": 17587 + }, + { + "epoch": 3.1312321937321936, + "grad_norm": 1.0221272706985474, + "learning_rate": 2.2454225880813618e-05, + "loss": 0.6562, + "step": 17588 + }, + { + "epoch": 3.1314102564102564, + "grad_norm": 0.9121100902557373, + "learning_rate": 2.2445388643997812e-05, + "loss": 0.7856, + "step": 17589 + }, + { + "epoch": 3.131588319088319, + "grad_norm": 1.0054078102111816, + "learning_rate": 2.243655292670347e-05, + "loss": 0.9622, + "step": 17590 + }, + { + "epoch": 3.131766381766382, + "grad_norm": 0.8391761779785156, + "learning_rate": 2.2427718729103763e-05, + "loss": 0.7045, + "step": 17591 + }, + { + "epoch": 3.1319444444444446, + "grad_norm": 0.9723725914955139, + "learning_rate": 2.2418886051371767e-05, + "loss": 0.8161, + "step": 17592 + }, + { + "epoch": 3.132122507122507, + "grad_norm": 0.8473033905029297, + "learning_rate": 2.241005489368052e-05, + "loss": 0.7708, + "step": 17593 + }, + { + "epoch": 3.1323005698005697, + "grad_norm": 0.9271494150161743, + "learning_rate": 2.240122525620312e-05, + "loss": 0.7198, + "step": 17594 + }, + { + "epoch": 3.1324786324786325, + "grad_norm": 1.0367611646652222, + "learning_rate": 2.2392397139112454e-05, + "loss": 1.0495, + "step": 17595 + }, + { + "epoch": 3.132656695156695, + "grad_norm": 0.8822857737541199, + "learning_rate": 2.238357054258158e-05, + "loss": 0.5491, + "step": 17596 + }, + { + "epoch": 3.132834757834758, + "grad_norm": 0.9773828387260437, + "learning_rate": 2.237474546678341e-05, + "loss": 0.7685, + "step": 17597 + }, + { + "epoch": 3.1330128205128207, + "grad_norm": 1.003929615020752, + "learning_rate": 2.2365921911890842e-05, + "loss": 0.7251, + "step": 17598 + }, + { + "epoch": 3.133190883190883, + "grad_norm": 0.9347188472747803, + "learning_rate": 2.235709987807677e-05, + "loss": 0.9213, + "step": 17599 + }, + { + "epoch": 3.1333689458689458, + "grad_norm": 0.9359163045883179, + "learning_rate": 2.2348279365514036e-05, + "loss": 0.8882, + "step": 17600 + }, + { + "epoch": 3.1335470085470085, + "grad_norm": 1.0264430046081543, + "learning_rate": 2.2339460374375463e-05, + "loss": 0.567, + "step": 17601 + }, + { + "epoch": 3.1337250712250713, + "grad_norm": 1.0392861366271973, + "learning_rate": 2.2330642904833833e-05, + "loss": 0.725, + "step": 17602 + }, + { + "epoch": 3.133903133903134, + "grad_norm": 0.8876413702964783, + "learning_rate": 2.2321826957061888e-05, + "loss": 0.8109, + "step": 17603 + }, + { + "epoch": 3.1340811965811968, + "grad_norm": 1.0765225887298584, + "learning_rate": 2.2313012531232435e-05, + "loss": 0.8024, + "step": 17604 + }, + { + "epoch": 3.134259259259259, + "grad_norm": 0.8807610273361206, + "learning_rate": 2.2304199627518063e-05, + "loss": 0.6966, + "step": 17605 + }, + { + "epoch": 3.134437321937322, + "grad_norm": 0.8906777501106262, + "learning_rate": 2.2295388246091532e-05, + "loss": 0.7797, + "step": 17606 + }, + { + "epoch": 3.1346153846153846, + "grad_norm": 1.0764859914779663, + "learning_rate": 2.2286578387125445e-05, + "loss": 0.8388, + "step": 17607 + }, + { + "epoch": 3.1347934472934473, + "grad_norm": 0.9732229709625244, + "learning_rate": 2.2277770050792422e-05, + "loss": 0.846, + "step": 17608 + }, + { + "epoch": 3.13497150997151, + "grad_norm": 0.8559836149215698, + "learning_rate": 2.226896323726504e-05, + "loss": 0.7839, + "step": 17609 + }, + { + "epoch": 3.135149572649573, + "grad_norm": 0.957591712474823, + "learning_rate": 2.2260157946715853e-05, + "loss": 0.7902, + "step": 17610 + }, + { + "epoch": 3.135327635327635, + "grad_norm": 0.8924140930175781, + "learning_rate": 2.2251354179317352e-05, + "loss": 0.5787, + "step": 17611 + }, + { + "epoch": 3.135505698005698, + "grad_norm": 1.0200998783111572, + "learning_rate": 2.2242551935242117e-05, + "loss": 0.9569, + "step": 17612 + }, + { + "epoch": 3.1356837606837606, + "grad_norm": 0.9117900729179382, + "learning_rate": 2.223375121466249e-05, + "loss": 0.7962, + "step": 17613 + }, + { + "epoch": 3.1358618233618234, + "grad_norm": 0.9254946112632751, + "learning_rate": 2.222495201775099e-05, + "loss": 0.6598, + "step": 17614 + }, + { + "epoch": 3.136039886039886, + "grad_norm": 0.8572172522544861, + "learning_rate": 2.2216154344679983e-05, + "loss": 0.7288, + "step": 17615 + }, + { + "epoch": 3.136217948717949, + "grad_norm": 0.8563082814216614, + "learning_rate": 2.2207358195621865e-05, + "loss": 0.6116, + "step": 17616 + }, + { + "epoch": 3.136396011396011, + "grad_norm": 0.8751232624053955, + "learning_rate": 2.2198563570748953e-05, + "loss": 0.8396, + "step": 17617 + }, + { + "epoch": 3.136574074074074, + "grad_norm": 0.8472304344177246, + "learning_rate": 2.218977047023355e-05, + "loss": 0.8801, + "step": 17618 + }, + { + "epoch": 3.1367521367521367, + "grad_norm": 0.8158729076385498, + "learning_rate": 2.2180978894248007e-05, + "loss": 0.6356, + "step": 17619 + }, + { + "epoch": 3.1369301994301995, + "grad_norm": 0.9257596731185913, + "learning_rate": 2.217218884296447e-05, + "loss": 0.6578, + "step": 17620 + }, + { + "epoch": 3.137108262108262, + "grad_norm": 0.887220025062561, + "learning_rate": 2.216340031655525e-05, + "loss": 0.7647, + "step": 17621 + }, + { + "epoch": 3.137286324786325, + "grad_norm": 0.8736263513565063, + "learning_rate": 2.2154613315192508e-05, + "loss": 0.7776, + "step": 17622 + }, + { + "epoch": 3.1374643874643873, + "grad_norm": 0.8078323006629944, + "learning_rate": 2.2145827839048416e-05, + "loss": 0.7054, + "step": 17623 + }, + { + "epoch": 3.13764245014245, + "grad_norm": 0.8907020092010498, + "learning_rate": 2.2137043888295096e-05, + "loss": 0.7799, + "step": 17624 + }, + { + "epoch": 3.1378205128205128, + "grad_norm": 0.895183801651001, + "learning_rate": 2.212826146310465e-05, + "loss": 0.758, + "step": 17625 + }, + { + "epoch": 3.1379985754985755, + "grad_norm": 0.9827015995979309, + "learning_rate": 2.2119480563649153e-05, + "loss": 0.593, + "step": 17626 + }, + { + "epoch": 3.1381766381766383, + "grad_norm": 0.927594780921936, + "learning_rate": 2.211070119010066e-05, + "loss": 0.6629, + "step": 17627 + }, + { + "epoch": 3.138354700854701, + "grad_norm": 1.0362217426300049, + "learning_rate": 2.2101923342631148e-05, + "loss": 0.781, + "step": 17628 + }, + { + "epoch": 3.1385327635327633, + "grad_norm": 0.9003650546073914, + "learning_rate": 2.2093147021412673e-05, + "loss": 0.6849, + "step": 17629 + }, + { + "epoch": 3.138710826210826, + "grad_norm": 0.8728612065315247, + "learning_rate": 2.20843722266171e-05, + "loss": 0.8316, + "step": 17630 + }, + { + "epoch": 3.138888888888889, + "grad_norm": 0.8697235584259033, + "learning_rate": 2.207559895841642e-05, + "loss": 0.7177, + "step": 17631 + }, + { + "epoch": 3.1390669515669516, + "grad_norm": 1.0648221969604492, + "learning_rate": 2.206682721698251e-05, + "loss": 0.794, + "step": 17632 + }, + { + "epoch": 3.1392450142450143, + "grad_norm": 0.8567960858345032, + "learning_rate": 2.2058057002487198e-05, + "loss": 0.6276, + "step": 17633 + }, + { + "epoch": 3.139423076923077, + "grad_norm": 1.0193754434585571, + "learning_rate": 2.2049288315102412e-05, + "loss": 0.8894, + "step": 17634 + }, + { + "epoch": 3.1396011396011394, + "grad_norm": 0.9031814932823181, + "learning_rate": 2.2040521154999827e-05, + "loss": 0.697, + "step": 17635 + }, + { + "epoch": 3.139779202279202, + "grad_norm": 0.8300186395645142, + "learning_rate": 2.203175552235134e-05, + "loss": 0.5605, + "step": 17636 + }, + { + "epoch": 3.139957264957265, + "grad_norm": 0.931399405002594, + "learning_rate": 2.2022991417328587e-05, + "loss": 0.7807, + "step": 17637 + }, + { + "epoch": 3.1401353276353277, + "grad_norm": 0.9556393623352051, + "learning_rate": 2.2014228840103357e-05, + "loss": 0.867, + "step": 17638 + }, + { + "epoch": 3.1403133903133904, + "grad_norm": 0.8357135653495789, + "learning_rate": 2.200546779084731e-05, + "loss": 0.6754, + "step": 17639 + }, + { + "epoch": 3.140491452991453, + "grad_norm": 0.9572781324386597, + "learning_rate": 2.1996708269732114e-05, + "loss": 0.7253, + "step": 17640 + }, + { + "epoch": 3.1406695156695155, + "grad_norm": 0.9977589845657349, + "learning_rate": 2.198795027692937e-05, + "loss": 0.6852, + "step": 17641 + }, + { + "epoch": 3.140847578347578, + "grad_norm": 0.9238983988761902, + "learning_rate": 2.197919381261069e-05, + "loss": 0.6661, + "step": 17642 + }, + { + "epoch": 3.141025641025641, + "grad_norm": 0.934651792049408, + "learning_rate": 2.1970438876947596e-05, + "loss": 0.7537, + "step": 17643 + }, + { + "epoch": 3.1412037037037037, + "grad_norm": 1.1089069843292236, + "learning_rate": 2.1961685470111725e-05, + "loss": 0.8351, + "step": 17644 + }, + { + "epoch": 3.1413817663817665, + "grad_norm": 0.890332818031311, + "learning_rate": 2.195293359227446e-05, + "loss": 0.7571, + "step": 17645 + }, + { + "epoch": 3.1415598290598292, + "grad_norm": 0.8519227504730225, + "learning_rate": 2.1944183243607363e-05, + "loss": 0.9673, + "step": 17646 + }, + { + "epoch": 3.1417378917378915, + "grad_norm": 0.9076274633407593, + "learning_rate": 2.1935434424281842e-05, + "loss": 0.618, + "step": 17647 + }, + { + "epoch": 3.1419159544159543, + "grad_norm": 0.9259662628173828, + "learning_rate": 2.1926687134469315e-05, + "loss": 0.6855, + "step": 17648 + }, + { + "epoch": 3.142094017094017, + "grad_norm": 0.839491069316864, + "learning_rate": 2.1917941374341177e-05, + "loss": 0.5683, + "step": 17649 + }, + { + "epoch": 3.14227207977208, + "grad_norm": 0.8948171734809875, + "learning_rate": 2.1909197144068783e-05, + "loss": 0.8399, + "step": 17650 + }, + { + "epoch": 3.1424501424501425, + "grad_norm": 1.0100452899932861, + "learning_rate": 2.1900454443823437e-05, + "loss": 0.8663, + "step": 17651 + }, + { + "epoch": 3.1426282051282053, + "grad_norm": 0.8691473603248596, + "learning_rate": 2.189171327377646e-05, + "loss": 0.7202, + "step": 17652 + }, + { + "epoch": 3.142806267806268, + "grad_norm": 0.979176938533783, + "learning_rate": 2.188297363409907e-05, + "loss": 0.9264, + "step": 17653 + }, + { + "epoch": 3.1429843304843303, + "grad_norm": 0.9306555390357971, + "learning_rate": 2.187423552496257e-05, + "loss": 0.8755, + "step": 17654 + }, + { + "epoch": 3.143162393162393, + "grad_norm": 0.9641979932785034, + "learning_rate": 2.1865498946538133e-05, + "loss": 0.6323, + "step": 17655 + }, + { + "epoch": 3.143340455840456, + "grad_norm": 1.126671314239502, + "learning_rate": 2.1856763898996924e-05, + "loss": 0.954, + "step": 17656 + }, + { + "epoch": 3.1435185185185186, + "grad_norm": 1.0349640846252441, + "learning_rate": 2.1848030382510108e-05, + "loss": 0.8558, + "step": 17657 + }, + { + "epoch": 3.1436965811965814, + "grad_norm": 0.9324985146522522, + "learning_rate": 2.1839298397248763e-05, + "loss": 0.8061, + "step": 17658 + }, + { + "epoch": 3.1438746438746437, + "grad_norm": 0.8919853568077087, + "learning_rate": 2.183056794338405e-05, + "loss": 0.759, + "step": 17659 + }, + { + "epoch": 3.1440527065527064, + "grad_norm": 0.9174198508262634, + "learning_rate": 2.182183902108692e-05, + "loss": 0.8332, + "step": 17660 + }, + { + "epoch": 3.144230769230769, + "grad_norm": 0.8662750124931335, + "learning_rate": 2.1813111630528517e-05, + "loss": 0.6642, + "step": 17661 + }, + { + "epoch": 3.144408831908832, + "grad_norm": 0.9606072902679443, + "learning_rate": 2.1804385771879708e-05, + "loss": 0.8076, + "step": 17662 + }, + { + "epoch": 3.1445868945868947, + "grad_norm": 0.904283344745636, + "learning_rate": 2.1795661445311555e-05, + "loss": 0.9605, + "step": 17663 + }, + { + "epoch": 3.1447649572649574, + "grad_norm": 1.0539193153381348, + "learning_rate": 2.1786938650994958e-05, + "loss": 0.9011, + "step": 17664 + }, + { + "epoch": 3.14494301994302, + "grad_norm": 0.8112294673919678, + "learning_rate": 2.177821738910083e-05, + "loss": 0.4874, + "step": 17665 + }, + { + "epoch": 3.1451210826210825, + "grad_norm": 0.8961923718452454, + "learning_rate": 2.1769497659800042e-05, + "loss": 0.8034, + "step": 17666 + }, + { + "epoch": 3.1452991452991452, + "grad_norm": 0.9964944124221802, + "learning_rate": 2.1760779463263426e-05, + "loss": 0.7663, + "step": 17667 + }, + { + "epoch": 3.145477207977208, + "grad_norm": 0.9126096367835999, + "learning_rate": 2.1752062799661787e-05, + "loss": 1.0941, + "step": 17668 + }, + { + "epoch": 3.1456552706552707, + "grad_norm": 0.9588297009468079, + "learning_rate": 2.1743347669165994e-05, + "loss": 0.7618, + "step": 17669 + }, + { + "epoch": 3.1458333333333335, + "grad_norm": 0.9706279039382935, + "learning_rate": 2.173463407194667e-05, + "loss": 0.8278, + "step": 17670 + }, + { + "epoch": 3.146011396011396, + "grad_norm": 1.005780816078186, + "learning_rate": 2.172592200817467e-05, + "loss": 0.743, + "step": 17671 + }, + { + "epoch": 3.1461894586894585, + "grad_norm": 1.0032274723052979, + "learning_rate": 2.1717211478020583e-05, + "loss": 0.7665, + "step": 17672 + }, + { + "epoch": 3.1463675213675213, + "grad_norm": 0.9049981236457825, + "learning_rate": 2.1708502481655134e-05, + "loss": 0.7667, + "step": 17673 + }, + { + "epoch": 3.146545584045584, + "grad_norm": 0.8788641095161438, + "learning_rate": 2.1699795019248947e-05, + "loss": 0.6075, + "step": 17674 + }, + { + "epoch": 3.146723646723647, + "grad_norm": 0.9241887331008911, + "learning_rate": 2.1691089090972595e-05, + "loss": 0.9202, + "step": 17675 + }, + { + "epoch": 3.1469017094017095, + "grad_norm": 0.8133395314216614, + "learning_rate": 2.1682384696996728e-05, + "loss": 0.6936, + "step": 17676 + }, + { + "epoch": 3.1470797720797723, + "grad_norm": 0.9261813163757324, + "learning_rate": 2.167368183749179e-05, + "loss": 0.8659, + "step": 17677 + }, + { + "epoch": 3.1472578347578346, + "grad_norm": 0.8164187073707581, + "learning_rate": 2.1664980512628374e-05, + "loss": 0.6205, + "step": 17678 + }, + { + "epoch": 3.1474358974358974, + "grad_norm": 0.8534388542175293, + "learning_rate": 2.165628072257695e-05, + "loss": 0.7437, + "step": 17679 + }, + { + "epoch": 3.14761396011396, + "grad_norm": 0.9858724474906921, + "learning_rate": 2.1647582467507943e-05, + "loss": 0.6912, + "step": 17680 + }, + { + "epoch": 3.147792022792023, + "grad_norm": 0.8704313039779663, + "learning_rate": 2.1638885747591807e-05, + "loss": 0.6267, + "step": 17681 + }, + { + "epoch": 3.1479700854700856, + "grad_norm": 0.9345740675926208, + "learning_rate": 2.163019056299892e-05, + "loss": 0.7671, + "step": 17682 + }, + { + "epoch": 3.148148148148148, + "grad_norm": 0.8783660531044006, + "learning_rate": 2.162149691389962e-05, + "loss": 0.6935, + "step": 17683 + }, + { + "epoch": 3.1483262108262107, + "grad_norm": 0.9478751420974731, + "learning_rate": 2.1612804800464324e-05, + "loss": 0.7343, + "step": 17684 + }, + { + "epoch": 3.1485042735042734, + "grad_norm": 1.0307378768920898, + "learning_rate": 2.1604114222863236e-05, + "loss": 0.6656, + "step": 17685 + }, + { + "epoch": 3.148682336182336, + "grad_norm": 0.8509631156921387, + "learning_rate": 2.1595425181266726e-05, + "loss": 0.7223, + "step": 17686 + }, + { + "epoch": 3.148860398860399, + "grad_norm": 0.9910686612129211, + "learning_rate": 2.1586737675844938e-05, + "loss": 0.6478, + "step": 17687 + }, + { + "epoch": 3.1490384615384617, + "grad_norm": 1.0413484573364258, + "learning_rate": 2.1578051706768156e-05, + "loss": 0.9215, + "step": 17688 + }, + { + "epoch": 3.1492165242165244, + "grad_norm": 0.9027596116065979, + "learning_rate": 2.156936727420654e-05, + "loss": 0.9464, + "step": 17689 + }, + { + "epoch": 3.1493945868945867, + "grad_norm": 1.0705668926239014, + "learning_rate": 2.1560684378330252e-05, + "loss": 0.9369, + "step": 17690 + }, + { + "epoch": 3.1495726495726495, + "grad_norm": 0.7609538435935974, + "learning_rate": 2.1552003019309418e-05, + "loss": 0.4558, + "step": 17691 + }, + { + "epoch": 3.1497507122507122, + "grad_norm": 0.9189962148666382, + "learning_rate": 2.1543323197314093e-05, + "loss": 0.6571, + "step": 17692 + }, + { + "epoch": 3.149928774928775, + "grad_norm": 1.3622703552246094, + "learning_rate": 2.153464491251439e-05, + "loss": 0.8121, + "step": 17693 + }, + { + "epoch": 3.1501068376068377, + "grad_norm": 0.9149178266525269, + "learning_rate": 2.1525968165080324e-05, + "loss": 0.7429, + "step": 17694 + }, + { + "epoch": 3.1502849002849005, + "grad_norm": 0.8335025310516357, + "learning_rate": 2.15172929551819e-05, + "loss": 0.7269, + "step": 17695 + }, + { + "epoch": 3.150462962962963, + "grad_norm": 1.145371913909912, + "learning_rate": 2.1508619282989084e-05, + "loss": 0.9163, + "step": 17696 + }, + { + "epoch": 3.1506410256410255, + "grad_norm": 0.7720812559127808, + "learning_rate": 2.149994714867182e-05, + "loss": 0.6217, + "step": 17697 + }, + { + "epoch": 3.1508190883190883, + "grad_norm": 0.9079264402389526, + "learning_rate": 2.149127655240003e-05, + "loss": 0.8527, + "step": 17698 + }, + { + "epoch": 3.150997150997151, + "grad_norm": 0.9881488680839539, + "learning_rate": 2.1482607494343575e-05, + "loss": 0.9116, + "step": 17699 + }, + { + "epoch": 3.151175213675214, + "grad_norm": 0.9430930614471436, + "learning_rate": 2.147393997467231e-05, + "loss": 0.6562, + "step": 17700 + }, + { + "epoch": 3.1513532763532766, + "grad_norm": 0.9559832215309143, + "learning_rate": 2.14652739935561e-05, + "loss": 0.7929, + "step": 17701 + }, + { + "epoch": 3.151531339031339, + "grad_norm": 1.0164002180099487, + "learning_rate": 2.1456609551164662e-05, + "loss": 0.9208, + "step": 17702 + }, + { + "epoch": 3.1517094017094016, + "grad_norm": 1.0220781564712524, + "learning_rate": 2.1447946647667837e-05, + "loss": 0.8878, + "step": 17703 + }, + { + "epoch": 3.1518874643874644, + "grad_norm": 1.060396671295166, + "learning_rate": 2.143928528323531e-05, + "loss": 0.6821, + "step": 17704 + }, + { + "epoch": 3.152065527065527, + "grad_norm": 0.8120558857917786, + "learning_rate": 2.143062545803679e-05, + "loss": 0.569, + "step": 17705 + }, + { + "epoch": 3.15224358974359, + "grad_norm": 0.8634228706359863, + "learning_rate": 2.1421967172241953e-05, + "loss": 0.7732, + "step": 17706 + }, + { + "epoch": 3.1524216524216526, + "grad_norm": 0.9585935473442078, + "learning_rate": 2.141331042602044e-05, + "loss": 0.7733, + "step": 17707 + }, + { + "epoch": 3.152599715099715, + "grad_norm": 0.8006394505500793, + "learning_rate": 2.140465521954186e-05, + "loss": 0.6772, + "step": 17708 + }, + { + "epoch": 3.1527777777777777, + "grad_norm": 0.948411226272583, + "learning_rate": 2.1396001552975786e-05, + "loss": 0.7058, + "step": 17709 + }, + { + "epoch": 3.1529558404558404, + "grad_norm": 0.9930174946784973, + "learning_rate": 2.1387349426491753e-05, + "loss": 0.8284, + "step": 17710 + }, + { + "epoch": 3.153133903133903, + "grad_norm": 0.8288687467575073, + "learning_rate": 2.137869884025936e-05, + "loss": 0.7214, + "step": 17711 + }, + { + "epoch": 3.153311965811966, + "grad_norm": 0.9206148386001587, + "learning_rate": 2.1370049794447977e-05, + "loss": 0.8247, + "step": 17712 + }, + { + "epoch": 3.1534900284900287, + "grad_norm": 1.0066028833389282, + "learning_rate": 2.1361402289227173e-05, + "loss": 0.8256, + "step": 17713 + }, + { + "epoch": 3.153668091168091, + "grad_norm": 1.0277632474899292, + "learning_rate": 2.135275632476632e-05, + "loss": 0.7422, + "step": 17714 + }, + { + "epoch": 3.1538461538461537, + "grad_norm": 0.8791954517364502, + "learning_rate": 2.134411190123481e-05, + "loss": 0.6614, + "step": 17715 + }, + { + "epoch": 3.1540242165242165, + "grad_norm": 0.8627964854240417, + "learning_rate": 2.133546901880209e-05, + "loss": 0.6784, + "step": 17716 + }, + { + "epoch": 3.1542022792022792, + "grad_norm": 1.106667160987854, + "learning_rate": 2.1326827677637384e-05, + "loss": 0.6475, + "step": 17717 + }, + { + "epoch": 3.154380341880342, + "grad_norm": 0.7291306257247925, + "learning_rate": 2.131818787791009e-05, + "loss": 0.6944, + "step": 17718 + }, + { + "epoch": 3.1545584045584047, + "grad_norm": 0.9892288446426392, + "learning_rate": 2.1309549619789458e-05, + "loss": 0.7805, + "step": 17719 + }, + { + "epoch": 3.154736467236467, + "grad_norm": 0.9121565818786621, + "learning_rate": 2.1300912903444736e-05, + "loss": 1.0122, + "step": 17720 + }, + { + "epoch": 3.15491452991453, + "grad_norm": 0.8239485025405884, + "learning_rate": 2.129227772904514e-05, + "loss": 0.6418, + "step": 17721 + }, + { + "epoch": 3.1550925925925926, + "grad_norm": 0.8559371829032898, + "learning_rate": 2.1283644096759868e-05, + "loss": 0.7703, + "step": 17722 + }, + { + "epoch": 3.1552706552706553, + "grad_norm": 0.9148365259170532, + "learning_rate": 2.1275012006758068e-05, + "loss": 0.884, + "step": 17723 + }, + { + "epoch": 3.155448717948718, + "grad_norm": 0.9810414910316467, + "learning_rate": 2.126638145920886e-05, + "loss": 0.7261, + "step": 17724 + }, + { + "epoch": 3.155626780626781, + "grad_norm": 0.960682213306427, + "learning_rate": 2.125775245428133e-05, + "loss": 0.766, + "step": 17725 + }, + { + "epoch": 3.155804843304843, + "grad_norm": 1.0861321687698364, + "learning_rate": 2.124912499214463e-05, + "loss": 0.568, + "step": 17726 + }, + { + "epoch": 3.155982905982906, + "grad_norm": 0.9647825956344604, + "learning_rate": 2.124049907296768e-05, + "loss": 0.8122, + "step": 17727 + }, + { + "epoch": 3.1561609686609686, + "grad_norm": 0.9854398369789124, + "learning_rate": 2.1231874696919564e-05, + "loss": 0.6373, + "step": 17728 + }, + { + "epoch": 3.1563390313390314, + "grad_norm": 0.8971094489097595, + "learning_rate": 2.1223251864169237e-05, + "loss": 0.6713, + "step": 17729 + }, + { + "epoch": 3.156517094017094, + "grad_norm": 0.9614965319633484, + "learning_rate": 2.1214630574885643e-05, + "loss": 0.9324, + "step": 17730 + }, + { + "epoch": 3.156695156695157, + "grad_norm": 0.9546411633491516, + "learning_rate": 2.120601082923771e-05, + "loss": 0.7762, + "step": 17731 + }, + { + "epoch": 3.156873219373219, + "grad_norm": 0.860262393951416, + "learning_rate": 2.1197392627394275e-05, + "loss": 0.7277, + "step": 17732 + }, + { + "epoch": 3.157051282051282, + "grad_norm": 0.8918822407722473, + "learning_rate": 2.11887759695243e-05, + "loss": 0.8899, + "step": 17733 + }, + { + "epoch": 3.1572293447293447, + "grad_norm": 0.9153090119361877, + "learning_rate": 2.118016085579647e-05, + "loss": 1.0711, + "step": 17734 + }, + { + "epoch": 3.1574074074074074, + "grad_norm": 1.1114693880081177, + "learning_rate": 2.11715472863797e-05, + "loss": 0.9146, + "step": 17735 + }, + { + "epoch": 3.15758547008547, + "grad_norm": 0.9536030888557434, + "learning_rate": 2.1162935261442705e-05, + "loss": 0.8239, + "step": 17736 + }, + { + "epoch": 3.157763532763533, + "grad_norm": 0.9061172604560852, + "learning_rate": 2.1154324781154222e-05, + "loss": 0.7297, + "step": 17737 + }, + { + "epoch": 3.1579415954415953, + "grad_norm": 0.9903003573417664, + "learning_rate": 2.1145715845682957e-05, + "loss": 0.8063, + "step": 17738 + }, + { + "epoch": 3.158119658119658, + "grad_norm": 0.9316477179527283, + "learning_rate": 2.1137108455197586e-05, + "loss": 0.7544, + "step": 17739 + }, + { + "epoch": 3.1582977207977208, + "grad_norm": 0.9925363063812256, + "learning_rate": 2.112850260986672e-05, + "loss": 0.8106, + "step": 17740 + }, + { + "epoch": 3.1584757834757835, + "grad_norm": 0.7903617024421692, + "learning_rate": 2.1119898309859064e-05, + "loss": 0.6821, + "step": 17741 + }, + { + "epoch": 3.1586538461538463, + "grad_norm": 1.0054620504379272, + "learning_rate": 2.111129555534308e-05, + "loss": 0.8715, + "step": 17742 + }, + { + "epoch": 3.158831908831909, + "grad_norm": 1.098737359046936, + "learning_rate": 2.110269434648744e-05, + "loss": 0.6393, + "step": 17743 + }, + { + "epoch": 3.1590099715099713, + "grad_norm": 0.7690902352333069, + "learning_rate": 2.1094094683460553e-05, + "loss": 0.6347, + "step": 17744 + }, + { + "epoch": 3.159188034188034, + "grad_norm": 0.9430564641952515, + "learning_rate": 2.1085496566431006e-05, + "loss": 0.68, + "step": 17745 + }, + { + "epoch": 3.159366096866097, + "grad_norm": 0.9977595806121826, + "learning_rate": 2.107689999556721e-05, + "loss": 0.8586, + "step": 17746 + }, + { + "epoch": 3.1595441595441596, + "grad_norm": 0.8231605887413025, + "learning_rate": 2.1068304971037612e-05, + "loss": 0.6109, + "step": 17747 + }, + { + "epoch": 3.1597222222222223, + "grad_norm": 1.1063002347946167, + "learning_rate": 2.1059711493010615e-05, + "loss": 0.6969, + "step": 17748 + }, + { + "epoch": 3.159900284900285, + "grad_norm": 0.8677477836608887, + "learning_rate": 2.1051119561654585e-05, + "loss": 0.6571, + "step": 17749 + }, + { + "epoch": 3.1600783475783474, + "grad_norm": 0.9700071811676025, + "learning_rate": 2.104252917713785e-05, + "loss": 0.665, + "step": 17750 + }, + { + "epoch": 3.16025641025641, + "grad_norm": 1.015114426612854, + "learning_rate": 2.1033940339628778e-05, + "loss": 0.7666, + "step": 17751 + }, + { + "epoch": 3.160434472934473, + "grad_norm": 0.7403838038444519, + "learning_rate": 2.1025353049295547e-05, + "loss": 0.4922, + "step": 17752 + }, + { + "epoch": 3.1606125356125356, + "grad_norm": 0.8768717646598816, + "learning_rate": 2.1016767306306506e-05, + "loss": 0.6746, + "step": 17753 + }, + { + "epoch": 3.1607905982905984, + "grad_norm": 0.9147419929504395, + "learning_rate": 2.1008183110829826e-05, + "loss": 0.7614, + "step": 17754 + }, + { + "epoch": 3.160968660968661, + "grad_norm": 0.949970543384552, + "learning_rate": 2.0999600463033696e-05, + "loss": 0.7278, + "step": 17755 + }, + { + "epoch": 3.1611467236467234, + "grad_norm": 0.9200690984725952, + "learning_rate": 2.099101936308633e-05, + "loss": 0.8333, + "step": 17756 + }, + { + "epoch": 3.161324786324786, + "grad_norm": 0.9482975006103516, + "learning_rate": 2.0982439811155753e-05, + "loss": 0.8056, + "step": 17757 + }, + { + "epoch": 3.161502849002849, + "grad_norm": 0.9989914298057556, + "learning_rate": 2.097386180741019e-05, + "loss": 0.9437, + "step": 17758 + }, + { + "epoch": 3.1616809116809117, + "grad_norm": 0.8893665075302124, + "learning_rate": 2.0965285352017572e-05, + "loss": 0.7423, + "step": 17759 + }, + { + "epoch": 3.1618589743589745, + "grad_norm": 0.8213073015213013, + "learning_rate": 2.0956710445146046e-05, + "loss": 0.7089, + "step": 17760 + }, + { + "epoch": 3.162037037037037, + "grad_norm": 1.022196650505066, + "learning_rate": 2.0948137086963583e-05, + "loss": 0.6745, + "step": 17761 + }, + { + "epoch": 3.1622150997150995, + "grad_norm": 0.9003040790557861, + "learning_rate": 2.093956527763815e-05, + "loss": 0.8741, + "step": 17762 + }, + { + "epoch": 3.1623931623931623, + "grad_norm": 0.9286761283874512, + "learning_rate": 2.0930995017337707e-05, + "loss": 0.8794, + "step": 17763 + }, + { + "epoch": 3.162571225071225, + "grad_norm": 1.0088931322097778, + "learning_rate": 2.092242630623016e-05, + "loss": 0.8463, + "step": 17764 + }, + { + "epoch": 3.1627492877492878, + "grad_norm": 0.9631438851356506, + "learning_rate": 2.091385914448337e-05, + "loss": 0.9442, + "step": 17765 + }, + { + "epoch": 3.1629273504273505, + "grad_norm": 0.9380853772163391, + "learning_rate": 2.090529353226528e-05, + "loss": 0.7912, + "step": 17766 + }, + { + "epoch": 3.1631054131054133, + "grad_norm": 1.0451303720474243, + "learning_rate": 2.0896729469743614e-05, + "loss": 0.7256, + "step": 17767 + }, + { + "epoch": 3.1632834757834756, + "grad_norm": 0.9581690430641174, + "learning_rate": 2.0888166957086252e-05, + "loss": 0.7764, + "step": 17768 + }, + { + "epoch": 3.1634615384615383, + "grad_norm": 1.0340020656585693, + "learning_rate": 2.087960599446087e-05, + "loss": 0.9355, + "step": 17769 + }, + { + "epoch": 3.163639601139601, + "grad_norm": 0.9682812690734863, + "learning_rate": 2.087104658203528e-05, + "loss": 0.7723, + "step": 17770 + }, + { + "epoch": 3.163817663817664, + "grad_norm": 1.0201079845428467, + "learning_rate": 2.0862488719977147e-05, + "loss": 0.8538, + "step": 17771 + }, + { + "epoch": 3.1639957264957266, + "grad_norm": 1.042189598083496, + "learning_rate": 2.0853932408454135e-05, + "loss": 0.8455, + "step": 17772 + }, + { + "epoch": 3.1641737891737893, + "grad_norm": 0.9279144406318665, + "learning_rate": 2.0845377647633957e-05, + "loss": 0.7129, + "step": 17773 + }, + { + "epoch": 3.164351851851852, + "grad_norm": 1.022481083869934, + "learning_rate": 2.0836824437684122e-05, + "loss": 0.6415, + "step": 17774 + }, + { + "epoch": 3.1645299145299144, + "grad_norm": 0.7837867736816406, + "learning_rate": 2.0828272778772285e-05, + "loss": 0.521, + "step": 17775 + }, + { + "epoch": 3.164707977207977, + "grad_norm": 0.9831959009170532, + "learning_rate": 2.081972267106599e-05, + "loss": 0.6261, + "step": 17776 + }, + { + "epoch": 3.16488603988604, + "grad_norm": 0.8895300626754761, + "learning_rate": 2.0811174114732734e-05, + "loss": 0.6477, + "step": 17777 + }, + { + "epoch": 3.1650641025641026, + "grad_norm": 0.8388190865516663, + "learning_rate": 2.0802627109940032e-05, + "loss": 0.6722, + "step": 17778 + }, + { + "epoch": 3.1652421652421654, + "grad_norm": 0.9670847654342651, + "learning_rate": 2.0794081656855324e-05, + "loss": 0.7609, + "step": 17779 + }, + { + "epoch": 3.1654202279202277, + "grad_norm": 0.805338978767395, + "learning_rate": 2.0785537755646055e-05, + "loss": 0.7714, + "step": 17780 + }, + { + "epoch": 3.1655982905982905, + "grad_norm": 0.9516651630401611, + "learning_rate": 2.0776995406479616e-05, + "loss": 0.6149, + "step": 17781 + }, + { + "epoch": 3.165776353276353, + "grad_norm": 0.8569229245185852, + "learning_rate": 2.0768454609523357e-05, + "loss": 0.6444, + "step": 17782 + }, + { + "epoch": 3.165954415954416, + "grad_norm": 0.9347574710845947, + "learning_rate": 2.0759915364944693e-05, + "loss": 0.6812, + "step": 17783 + }, + { + "epoch": 3.1661324786324787, + "grad_norm": 1.0011918544769287, + "learning_rate": 2.0751377672910834e-05, + "loss": 0.9513, + "step": 17784 + }, + { + "epoch": 3.1663105413105415, + "grad_norm": 0.794636607170105, + "learning_rate": 2.074284153358912e-05, + "loss": 0.5172, + "step": 17785 + }, + { + "epoch": 3.166488603988604, + "grad_norm": 1.019755482673645, + "learning_rate": 2.0734306947146785e-05, + "loss": 0.8476, + "step": 17786 + }, + { + "epoch": 3.1666666666666665, + "grad_norm": 1.0172061920166016, + "learning_rate": 2.0725773913751056e-05, + "loss": 0.7237, + "step": 17787 + }, + { + "epoch": 3.1668447293447293, + "grad_norm": 0.9893356561660767, + "learning_rate": 2.0717242433569095e-05, + "loss": 0.9456, + "step": 17788 + }, + { + "epoch": 3.167022792022792, + "grad_norm": 1.0239630937576294, + "learning_rate": 2.070871250676808e-05, + "loss": 0.7997, + "step": 17789 + }, + { + "epoch": 3.1672008547008548, + "grad_norm": 0.7666916847229004, + "learning_rate": 2.0700184133515098e-05, + "loss": 0.6818, + "step": 17790 + }, + { + "epoch": 3.1673789173789175, + "grad_norm": 1.1362230777740479, + "learning_rate": 2.069165731397732e-05, + "loss": 0.964, + "step": 17791 + }, + { + "epoch": 3.16755698005698, + "grad_norm": 1.055202841758728, + "learning_rate": 2.0683132048321752e-05, + "loss": 0.7176, + "step": 17792 + }, + { + "epoch": 3.1677350427350426, + "grad_norm": 1.0161933898925781, + "learning_rate": 2.067460833671545e-05, + "loss": 0.9733, + "step": 17793 + }, + { + "epoch": 3.1679131054131053, + "grad_norm": 0.922094464302063, + "learning_rate": 2.0666086179325417e-05, + "loss": 0.6734, + "step": 17794 + }, + { + "epoch": 3.168091168091168, + "grad_norm": 1.061632752418518, + "learning_rate": 2.0657565576318616e-05, + "loss": 0.7061, + "step": 17795 + }, + { + "epoch": 3.168269230769231, + "grad_norm": 0.9560376405715942, + "learning_rate": 2.0649046527862002e-05, + "loss": 0.7497, + "step": 17796 + }, + { + "epoch": 3.1684472934472936, + "grad_norm": 0.8954378366470337, + "learning_rate": 2.0640529034122457e-05, + "loss": 0.6278, + "step": 17797 + }, + { + "epoch": 3.1686253561253563, + "grad_norm": 0.9111878871917725, + "learning_rate": 2.0632013095266945e-05, + "loss": 0.858, + "step": 17798 + }, + { + "epoch": 3.1688034188034186, + "grad_norm": 1.0656862258911133, + "learning_rate": 2.0623498711462208e-05, + "loss": 0.7066, + "step": 17799 + }, + { + "epoch": 3.1689814814814814, + "grad_norm": 0.940841555595398, + "learning_rate": 2.0614985882875147e-05, + "loss": 0.8436, + "step": 17800 + }, + { + "epoch": 3.169159544159544, + "grad_norm": 0.929924726486206, + "learning_rate": 2.0606474609672532e-05, + "loss": 0.7138, + "step": 17801 + }, + { + "epoch": 3.169337606837607, + "grad_norm": 0.9759158492088318, + "learning_rate": 2.0597964892021127e-05, + "loss": 1.1146, + "step": 17802 + }, + { + "epoch": 3.1695156695156697, + "grad_norm": 0.8888540267944336, + "learning_rate": 2.058945673008765e-05, + "loss": 0.7375, + "step": 17803 + }, + { + "epoch": 3.169693732193732, + "grad_norm": 0.9520168900489807, + "learning_rate": 2.0580950124038812e-05, + "loss": 0.7527, + "step": 17804 + }, + { + "epoch": 3.1698717948717947, + "grad_norm": 0.9040802121162415, + "learning_rate": 2.057244507404127e-05, + "loss": 0.5453, + "step": 17805 + }, + { + "epoch": 3.1700498575498575, + "grad_norm": 0.9299812912940979, + "learning_rate": 2.056394158026168e-05, + "loss": 0.8139, + "step": 17806 + }, + { + "epoch": 3.17022792022792, + "grad_norm": 0.9180371165275574, + "learning_rate": 2.0555439642866602e-05, + "loss": 0.7142, + "step": 17807 + }, + { + "epoch": 3.170405982905983, + "grad_norm": 1.016885757446289, + "learning_rate": 2.0546939262022725e-05, + "loss": 0.7723, + "step": 17808 + }, + { + "epoch": 3.1705840455840457, + "grad_norm": 0.9292659759521484, + "learning_rate": 2.0538440437896456e-05, + "loss": 0.853, + "step": 17809 + }, + { + "epoch": 3.1707621082621085, + "grad_norm": 1.0694403648376465, + "learning_rate": 2.052994317065441e-05, + "loss": 0.7869, + "step": 17810 + }, + { + "epoch": 3.1709401709401708, + "grad_norm": 0.8516891598701477, + "learning_rate": 2.052144746046304e-05, + "loss": 0.77, + "step": 17811 + }, + { + "epoch": 3.1711182336182335, + "grad_norm": 0.8941650986671448, + "learning_rate": 2.051295330748878e-05, + "loss": 0.568, + "step": 17812 + }, + { + "epoch": 3.1712962962962963, + "grad_norm": 0.8694541454315186, + "learning_rate": 2.050446071189813e-05, + "loss": 0.621, + "step": 17813 + }, + { + "epoch": 3.171474358974359, + "grad_norm": 0.9683319926261902, + "learning_rate": 2.0495969673857383e-05, + "loss": 0.8374, + "step": 17814 + }, + { + "epoch": 3.171652421652422, + "grad_norm": 0.8314747214317322, + "learning_rate": 2.0487480193533004e-05, + "loss": 0.6982, + "step": 17815 + }, + { + "epoch": 3.1718304843304845, + "grad_norm": 0.8739849328994751, + "learning_rate": 2.0478992271091234e-05, + "loss": 0.8121, + "step": 17816 + }, + { + "epoch": 3.172008547008547, + "grad_norm": 1.002616047859192, + "learning_rate": 2.0470505906698446e-05, + "loss": 0.8264, + "step": 17817 + }, + { + "epoch": 3.1721866096866096, + "grad_norm": 0.9982789754867554, + "learning_rate": 2.046202110052088e-05, + "loss": 0.6077, + "step": 17818 + }, + { + "epoch": 3.1723646723646723, + "grad_norm": 0.85958331823349, + "learning_rate": 2.045353785272479e-05, + "loss": 0.6998, + "step": 17819 + }, + { + "epoch": 3.172542735042735, + "grad_norm": 1.0867046117782593, + "learning_rate": 2.0445056163476374e-05, + "loss": 0.6445, + "step": 17820 + }, + { + "epoch": 3.172720797720798, + "grad_norm": 1.0782829523086548, + "learning_rate": 2.0436576032941834e-05, + "loss": 0.7187, + "step": 17821 + }, + { + "epoch": 3.1728988603988606, + "grad_norm": 1.0062856674194336, + "learning_rate": 2.0428097461287265e-05, + "loss": 0.8538, + "step": 17822 + }, + { + "epoch": 3.173076923076923, + "grad_norm": 1.019987940788269, + "learning_rate": 2.04196204486789e-05, + "loss": 0.6901, + "step": 17823 + }, + { + "epoch": 3.1732549857549857, + "grad_norm": 0.920170247554779, + "learning_rate": 2.0411144995282696e-05, + "loss": 0.6752, + "step": 17824 + }, + { + "epoch": 3.1734330484330484, + "grad_norm": 0.8132520318031311, + "learning_rate": 2.0402671101264802e-05, + "loss": 0.6694, + "step": 17825 + }, + { + "epoch": 3.173611111111111, + "grad_norm": 0.978809654712677, + "learning_rate": 2.0394198766791215e-05, + "loss": 0.6091, + "step": 17826 + }, + { + "epoch": 3.173789173789174, + "grad_norm": 0.7911710739135742, + "learning_rate": 2.0385727992027946e-05, + "loss": 0.7579, + "step": 17827 + }, + { + "epoch": 3.1739672364672367, + "grad_norm": 0.8443024754524231, + "learning_rate": 2.037725877714095e-05, + "loss": 0.6157, + "step": 17828 + }, + { + "epoch": 3.174145299145299, + "grad_norm": 0.9306837320327759, + "learning_rate": 2.036879112229616e-05, + "loss": 0.8692, + "step": 17829 + }, + { + "epoch": 3.1743233618233617, + "grad_norm": 0.8994047045707703, + "learning_rate": 2.036032502765949e-05, + "loss": 0.8047, + "step": 17830 + }, + { + "epoch": 3.1745014245014245, + "grad_norm": 0.9705090522766113, + "learning_rate": 2.035186049339679e-05, + "loss": 0.7923, + "step": 17831 + }, + { + "epoch": 3.1746794871794872, + "grad_norm": 0.9646127223968506, + "learning_rate": 2.0343397519673955e-05, + "loss": 0.7605, + "step": 17832 + }, + { + "epoch": 3.17485754985755, + "grad_norm": 0.9448650479316711, + "learning_rate": 2.0334936106656775e-05, + "loss": 0.6693, + "step": 17833 + }, + { + "epoch": 3.1750356125356127, + "grad_norm": 0.9898852109909058, + "learning_rate": 2.032647625451103e-05, + "loss": 0.8309, + "step": 17834 + }, + { + "epoch": 3.175213675213675, + "grad_norm": 1.194535732269287, + "learning_rate": 2.031801796340248e-05, + "loss": 0.8283, + "step": 17835 + }, + { + "epoch": 3.175391737891738, + "grad_norm": 0.8253726959228516, + "learning_rate": 2.0309561233496832e-05, + "loss": 0.7098, + "step": 17836 + }, + { + "epoch": 3.1755698005698005, + "grad_norm": 0.8785462379455566, + "learning_rate": 2.0301106064959774e-05, + "loss": 0.8157, + "step": 17837 + }, + { + "epoch": 3.1757478632478633, + "grad_norm": 1.0798345804214478, + "learning_rate": 2.0292652457957028e-05, + "loss": 0.6791, + "step": 17838 + }, + { + "epoch": 3.175925925925926, + "grad_norm": 0.7758960127830505, + "learning_rate": 2.028420041265413e-05, + "loss": 0.5568, + "step": 17839 + }, + { + "epoch": 3.176103988603989, + "grad_norm": 0.833829402923584, + "learning_rate": 2.027574992921678e-05, + "loss": 0.6835, + "step": 17840 + }, + { + "epoch": 3.176282051282051, + "grad_norm": 0.8681000471115112, + "learning_rate": 2.0267301007810448e-05, + "loss": 0.7613, + "step": 17841 + }, + { + "epoch": 3.176460113960114, + "grad_norm": 0.9271478056907654, + "learning_rate": 2.025885364860075e-05, + "loss": 0.8355, + "step": 17842 + }, + { + "epoch": 3.1766381766381766, + "grad_norm": 0.9865009188652039, + "learning_rate": 2.025040785175316e-05, + "loss": 0.9064, + "step": 17843 + }, + { + "epoch": 3.1768162393162394, + "grad_norm": 1.0745911598205566, + "learning_rate": 2.024196361743317e-05, + "loss": 0.8944, + "step": 17844 + }, + { + "epoch": 3.176994301994302, + "grad_norm": 0.9590545296669006, + "learning_rate": 2.0233520945806218e-05, + "loss": 0.9389, + "step": 17845 + }, + { + "epoch": 3.177172364672365, + "grad_norm": 0.9916484951972961, + "learning_rate": 2.0225079837037718e-05, + "loss": 0.6494, + "step": 17846 + }, + { + "epoch": 3.177350427350427, + "grad_norm": 0.8701820969581604, + "learning_rate": 2.0216640291293043e-05, + "loss": 0.8206, + "step": 17847 + }, + { + "epoch": 3.17752849002849, + "grad_norm": 0.882976233959198, + "learning_rate": 2.020820230873761e-05, + "loss": 0.6473, + "step": 17848 + }, + { + "epoch": 3.1777065527065527, + "grad_norm": 0.9599972367286682, + "learning_rate": 2.0199765889536647e-05, + "loss": 0.8647, + "step": 17849 + }, + { + "epoch": 3.1778846153846154, + "grad_norm": 0.864414632320404, + "learning_rate": 2.0191331033855564e-05, + "loss": 0.7128, + "step": 17850 + }, + { + "epoch": 3.178062678062678, + "grad_norm": 1.1498304605484009, + "learning_rate": 2.0182897741859497e-05, + "loss": 0.8783, + "step": 17851 + }, + { + "epoch": 3.178240740740741, + "grad_norm": 0.8664734363555908, + "learning_rate": 2.0174466013713777e-05, + "loss": 0.7432, + "step": 17852 + }, + { + "epoch": 3.1784188034188032, + "grad_norm": 0.8619598150253296, + "learning_rate": 2.0166035849583566e-05, + "loss": 0.7127, + "step": 17853 + }, + { + "epoch": 3.178596866096866, + "grad_norm": 0.8788446187973022, + "learning_rate": 2.0157607249634024e-05, + "loss": 0.7772, + "step": 17854 + }, + { + "epoch": 3.1787749287749287, + "grad_norm": 0.9376852512359619, + "learning_rate": 2.014918021403035e-05, + "loss": 0.7643, + "step": 17855 + }, + { + "epoch": 3.1789529914529915, + "grad_norm": 1.3296799659729004, + "learning_rate": 2.0140754742937575e-05, + "loss": 0.7753, + "step": 17856 + }, + { + "epoch": 3.1791310541310542, + "grad_norm": 0.932537853717804, + "learning_rate": 2.013233083652084e-05, + "loss": 0.7868, + "step": 17857 + }, + { + "epoch": 3.179309116809117, + "grad_norm": 0.8196058869361877, + "learning_rate": 2.0123908494945164e-05, + "loss": 0.6838, + "step": 17858 + }, + { + "epoch": 3.1794871794871793, + "grad_norm": 0.8767826557159424, + "learning_rate": 2.011548771837558e-05, + "loss": 0.5893, + "step": 17859 + }, + { + "epoch": 3.179665242165242, + "grad_norm": 0.9560354948043823, + "learning_rate": 2.0107068506977068e-05, + "loss": 0.6576, + "step": 17860 + }, + { + "epoch": 3.179843304843305, + "grad_norm": 1.0848370790481567, + "learning_rate": 2.0098650860914582e-05, + "loss": 1.0265, + "step": 17861 + }, + { + "epoch": 3.1800213675213675, + "grad_norm": 0.8908656239509583, + "learning_rate": 2.0090234780353022e-05, + "loss": 0.6375, + "step": 17862 + }, + { + "epoch": 3.1801994301994303, + "grad_norm": 1.0982664823532104, + "learning_rate": 2.0081820265457374e-05, + "loss": 0.7475, + "step": 17863 + }, + { + "epoch": 3.180377492877493, + "grad_norm": 1.0687603950500488, + "learning_rate": 2.007340731639239e-05, + "loss": 0.9097, + "step": 17864 + }, + { + "epoch": 3.1805555555555554, + "grad_norm": 1.0065104961395264, + "learning_rate": 2.0064995933323004e-05, + "loss": 0.7049, + "step": 17865 + }, + { + "epoch": 3.180733618233618, + "grad_norm": 0.9913015961647034, + "learning_rate": 2.0056586116413923e-05, + "loss": 0.7879, + "step": 17866 + }, + { + "epoch": 3.180911680911681, + "grad_norm": 0.8901994824409485, + "learning_rate": 2.0048177865829987e-05, + "loss": 0.7791, + "step": 17867 + }, + { + "epoch": 3.1810897435897436, + "grad_norm": 0.8973708748817444, + "learning_rate": 2.0039771181735935e-05, + "loss": 0.7945, + "step": 17868 + }, + { + "epoch": 3.1812678062678064, + "grad_norm": 0.9505265951156616, + "learning_rate": 2.003136606429645e-05, + "loss": 0.7165, + "step": 17869 + }, + { + "epoch": 3.181445868945869, + "grad_norm": 1.0017229318618774, + "learning_rate": 2.0022962513676237e-05, + "loss": 0.9075, + "step": 17870 + }, + { + "epoch": 3.1816239316239314, + "grad_norm": 0.876743495464325, + "learning_rate": 2.0014560530039893e-05, + "loss": 0.574, + "step": 17871 + }, + { + "epoch": 3.181801994301994, + "grad_norm": 0.9605728983879089, + "learning_rate": 2.0006160113552118e-05, + "loss": 0.7468, + "step": 17872 + }, + { + "epoch": 3.181980056980057, + "grad_norm": 1.0196484327316284, + "learning_rate": 1.999776126437747e-05, + "loss": 0.7313, + "step": 17873 + }, + { + "epoch": 3.1821581196581197, + "grad_norm": 0.842857301235199, + "learning_rate": 1.9989363982680487e-05, + "loss": 0.5271, + "step": 17874 + }, + { + "epoch": 3.1823361823361824, + "grad_norm": 0.9760614633560181, + "learning_rate": 1.998096826862571e-05, + "loss": 0.8287, + "step": 17875 + }, + { + "epoch": 3.182514245014245, + "grad_norm": 0.9114407896995544, + "learning_rate": 1.997257412237764e-05, + "loss": 0.7172, + "step": 17876 + }, + { + "epoch": 3.1826923076923075, + "grad_norm": 0.9301532506942749, + "learning_rate": 1.9964181544100747e-05, + "loss": 0.6226, + "step": 17877 + }, + { + "epoch": 3.1828703703703702, + "grad_norm": 0.8122416138648987, + "learning_rate": 1.995579053395944e-05, + "loss": 0.6236, + "step": 17878 + }, + { + "epoch": 3.183048433048433, + "grad_norm": 0.9543060660362244, + "learning_rate": 1.9947401092118124e-05, + "loss": 0.7392, + "step": 17879 + }, + { + "epoch": 3.1832264957264957, + "grad_norm": 1.0016096830368042, + "learning_rate": 1.9939013218741233e-05, + "loss": 0.7618, + "step": 17880 + }, + { + "epoch": 3.1834045584045585, + "grad_norm": 0.8758784532546997, + "learning_rate": 1.9930626913993012e-05, + "loss": 0.7702, + "step": 17881 + }, + { + "epoch": 3.1835826210826212, + "grad_norm": 0.9122433662414551, + "learning_rate": 1.9922242178037864e-05, + "loss": 0.7759, + "step": 17882 + }, + { + "epoch": 3.183760683760684, + "grad_norm": 0.8537415266036987, + "learning_rate": 1.9913859011040025e-05, + "loss": 0.5062, + "step": 17883 + }, + { + "epoch": 3.1839387464387463, + "grad_norm": 0.9870179295539856, + "learning_rate": 1.9905477413163754e-05, + "loss": 0.6001, + "step": 17884 + }, + { + "epoch": 3.184116809116809, + "grad_norm": 0.9088004231452942, + "learning_rate": 1.989709738457328e-05, + "loss": 0.6941, + "step": 17885 + }, + { + "epoch": 3.184294871794872, + "grad_norm": 0.9028190970420837, + "learning_rate": 1.988871892543277e-05, + "loss": 0.9215, + "step": 17886 + }, + { + "epoch": 3.1844729344729346, + "grad_norm": 0.9333863258361816, + "learning_rate": 1.9880342035906406e-05, + "loss": 0.8036, + "step": 17887 + }, + { + "epoch": 3.1846509971509973, + "grad_norm": 1.0232222080230713, + "learning_rate": 1.987196671615831e-05, + "loss": 0.8573, + "step": 17888 + }, + { + "epoch": 3.1848290598290596, + "grad_norm": Infinity, + "learning_rate": 1.987196671615831e-05, + "loss": 0.7025, + "step": 17889 + }, + { + "epoch": 3.1850071225071224, + "grad_norm": 0.9075542092323303, + "learning_rate": 1.986359296635254e-05, + "loss": 0.7867, + "step": 17890 + }, + { + "epoch": 3.185185185185185, + "grad_norm": 1.0360100269317627, + "learning_rate": 1.9855220786653217e-05, + "loss": 0.7186, + "step": 17891 + }, + { + "epoch": 3.185363247863248, + "grad_norm": 0.8505464792251587, + "learning_rate": 1.9846850177224363e-05, + "loss": 0.7619, + "step": 17892 + }, + { + "epoch": 3.1855413105413106, + "grad_norm": 0.8706018328666687, + "learning_rate": 1.983848113822998e-05, + "loss": 0.8204, + "step": 17893 + }, + { + "epoch": 3.1857193732193734, + "grad_norm": 1.1926771402359009, + "learning_rate": 1.9830113669834038e-05, + "loss": 0.7418, + "step": 17894 + }, + { + "epoch": 3.185897435897436, + "grad_norm": 0.8450924754142761, + "learning_rate": 1.9821747772200448e-05, + "loss": 0.6028, + "step": 17895 + }, + { + "epoch": 3.1860754985754984, + "grad_norm": 0.888190746307373, + "learning_rate": 1.9813383445493207e-05, + "loss": 0.7498, + "step": 17896 + }, + { + "epoch": 3.186253561253561, + "grad_norm": 1.065687894821167, + "learning_rate": 1.9805020689876096e-05, + "loss": 0.9076, + "step": 17897 + }, + { + "epoch": 3.186431623931624, + "grad_norm": 0.9114495515823364, + "learning_rate": 1.9796659505513038e-05, + "loss": 0.828, + "step": 17898 + }, + { + "epoch": 3.1866096866096867, + "grad_norm": 0.9044122695922852, + "learning_rate": 1.978829989256783e-05, + "loss": 0.7941, + "step": 17899 + }, + { + "epoch": 3.1867877492877494, + "grad_norm": 0.8886293768882751, + "learning_rate": 1.9779941851204265e-05, + "loss": 0.8974, + "step": 17900 + }, + { + "epoch": 3.1869658119658117, + "grad_norm": 0.9635893106460571, + "learning_rate": 1.9771585381586087e-05, + "loss": 0.7498, + "step": 17901 + }, + { + "epoch": 3.1871438746438745, + "grad_norm": 0.9525110721588135, + "learning_rate": 1.9763230483877037e-05, + "loss": 0.6705, + "step": 17902 + }, + { + "epoch": 3.1873219373219372, + "grad_norm": 1.0253328084945679, + "learning_rate": 1.9754877158240805e-05, + "loss": 0.7001, + "step": 17903 + }, + { + "epoch": 3.1875, + "grad_norm": 1.0464524030685425, + "learning_rate": 1.9746525404841064e-05, + "loss": 0.7115, + "step": 17904 + }, + { + "epoch": 3.1876780626780628, + "grad_norm": 0.9318157434463501, + "learning_rate": 1.9738175223841428e-05, + "loss": 0.7629, + "step": 17905 + }, + { + "epoch": 3.1878561253561255, + "grad_norm": 0.9584185481071472, + "learning_rate": 1.9729826615405557e-05, + "loss": 0.6182, + "step": 17906 + }, + { + "epoch": 3.1880341880341883, + "grad_norm": 0.8253262639045715, + "learning_rate": 1.9721479579696945e-05, + "loss": 0.5826, + "step": 17907 + }, + { + "epoch": 3.1882122507122506, + "grad_norm": 1.0227715969085693, + "learning_rate": 1.9713134116879195e-05, + "loss": 0.7201, + "step": 17908 + }, + { + "epoch": 3.1883903133903133, + "grad_norm": 0.8838222026824951, + "learning_rate": 1.97047902271158e-05, + "loss": 0.6851, + "step": 17909 + }, + { + "epoch": 3.188568376068376, + "grad_norm": 0.8144074082374573, + "learning_rate": 1.969644791057025e-05, + "loss": 0.7016, + "step": 17910 + }, + { + "epoch": 3.188746438746439, + "grad_norm": 1.1255220174789429, + "learning_rate": 1.9688107167405977e-05, + "loss": 1.0059, + "step": 17911 + }, + { + "epoch": 3.1889245014245016, + "grad_norm": 0.9504788517951965, + "learning_rate": 1.967976799778639e-05, + "loss": 0.8138, + "step": 17912 + }, + { + "epoch": 3.189102564102564, + "grad_norm": 1.0853493213653564, + "learning_rate": 1.9671430401874946e-05, + "loss": 0.9565, + "step": 17913 + }, + { + "epoch": 3.1892806267806266, + "grad_norm": 0.9323904514312744, + "learning_rate": 1.9663094379834912e-05, + "loss": 0.7243, + "step": 17914 + }, + { + "epoch": 3.1894586894586894, + "grad_norm": 0.9153193831443787, + "learning_rate": 1.965475993182968e-05, + "loss": 0.7891, + "step": 17915 + }, + { + "epoch": 3.189636752136752, + "grad_norm": 0.9310745596885681, + "learning_rate": 1.9646427058022522e-05, + "loss": 0.7926, + "step": 17916 + }, + { + "epoch": 3.189814814814815, + "grad_norm": 0.7946666479110718, + "learning_rate": 1.96380957585767e-05, + "loss": 0.5099, + "step": 17917 + }, + { + "epoch": 3.1899928774928776, + "grad_norm": 0.9780072569847107, + "learning_rate": 1.9629766033655472e-05, + "loss": 0.8119, + "step": 17918 + }, + { + "epoch": 3.1901709401709404, + "grad_norm": 0.886121928691864, + "learning_rate": 1.962143788342201e-05, + "loss": 0.8426, + "step": 17919 + }, + { + "epoch": 3.1903490028490027, + "grad_norm": 0.9156065583229065, + "learning_rate": 1.961311130803948e-05, + "loss": 0.7428, + "step": 17920 + }, + { + "epoch": 3.1905270655270654, + "grad_norm": 0.9247871041297913, + "learning_rate": 1.9604786307671085e-05, + "loss": 0.7394, + "step": 17921 + }, + { + "epoch": 3.190705128205128, + "grad_norm": 0.8814037442207336, + "learning_rate": 1.959646288247986e-05, + "loss": 0.6733, + "step": 17922 + }, + { + "epoch": 3.190883190883191, + "grad_norm": 0.915059506893158, + "learning_rate": 1.9588141032628958e-05, + "loss": 1.0306, + "step": 17923 + }, + { + "epoch": 3.1910612535612537, + "grad_norm": 1.0811846256256104, + "learning_rate": 1.9579820758281353e-05, + "loss": 1.0782, + "step": 17924 + }, + { + "epoch": 3.191239316239316, + "grad_norm": 0.8332127332687378, + "learning_rate": 1.957150205960012e-05, + "loss": 0.8365, + "step": 17925 + }, + { + "epoch": 3.1914173789173788, + "grad_norm": 0.9311001896858215, + "learning_rate": 1.9563184936748236e-05, + "loss": 0.7995, + "step": 17926 + }, + { + "epoch": 3.1915954415954415, + "grad_norm": 0.844411313533783, + "learning_rate": 1.9554869389888642e-05, + "loss": 0.7573, + "step": 17927 + }, + { + "epoch": 3.1917735042735043, + "grad_norm": 1.0677951574325562, + "learning_rate": 1.9546555419184277e-05, + "loss": 0.8501, + "step": 17928 + }, + { + "epoch": 3.191951566951567, + "grad_norm": 1.0033360719680786, + "learning_rate": 1.953824302479803e-05, + "loss": 0.8433, + "step": 17929 + }, + { + "epoch": 3.1921296296296298, + "grad_norm": 0.9265921711921692, + "learning_rate": 1.9529932206892732e-05, + "loss": 0.901, + "step": 17930 + }, + { + "epoch": 3.1923076923076925, + "grad_norm": 0.9688268303871155, + "learning_rate": 1.952162296563128e-05, + "loss": 0.743, + "step": 17931 + }, + { + "epoch": 3.192485754985755, + "grad_norm": 0.7872363328933716, + "learning_rate": 1.9513315301176462e-05, + "loss": 0.608, + "step": 17932 + }, + { + "epoch": 3.1926638176638176, + "grad_norm": 0.9641526937484741, + "learning_rate": 1.9505009213691015e-05, + "loss": 0.9266, + "step": 17933 + }, + { + "epoch": 3.1928418803418803, + "grad_norm": 0.9330029487609863, + "learning_rate": 1.9496704703337708e-05, + "loss": 0.8647, + "step": 17934 + }, + { + "epoch": 3.193019943019943, + "grad_norm": 1.0177229642868042, + "learning_rate": 1.9488401770279207e-05, + "loss": 0.7655, + "step": 17935 + }, + { + "epoch": 3.193198005698006, + "grad_norm": 1.0131559371948242, + "learning_rate": 1.948010041467828e-05, + "loss": 0.8529, + "step": 17936 + }, + { + "epoch": 3.1933760683760686, + "grad_norm": 0.8644171357154846, + "learning_rate": 1.947180063669748e-05, + "loss": 0.6328, + "step": 17937 + }, + { + "epoch": 3.193554131054131, + "grad_norm": 0.8475095629692078, + "learning_rate": 1.9463502436499503e-05, + "loss": 0.5732, + "step": 17938 + }, + { + "epoch": 3.1937321937321936, + "grad_norm": 0.8605408668518066, + "learning_rate": 1.9455205814246846e-05, + "loss": 0.6729, + "step": 17939 + }, + { + "epoch": 3.1939102564102564, + "grad_norm": 0.8682053685188293, + "learning_rate": 1.9446910770102145e-05, + "loss": 0.7593, + "step": 17940 + }, + { + "epoch": 3.194088319088319, + "grad_norm": 0.9077824950218201, + "learning_rate": 1.943861730422788e-05, + "loss": 0.7351, + "step": 17941 + }, + { + "epoch": 3.194266381766382, + "grad_norm": 0.9960265159606934, + "learning_rate": 1.9430325416786564e-05, + "loss": 0.7917, + "step": 17942 + }, + { + "epoch": 3.1944444444444446, + "grad_norm": 0.9843694567680359, + "learning_rate": 1.942203510794066e-05, + "loss": 0.812, + "step": 17943 + }, + { + "epoch": 3.194622507122507, + "grad_norm": 1.003712773323059, + "learning_rate": 1.941374637785258e-05, + "loss": 0.6471, + "step": 17944 + }, + { + "epoch": 3.1948005698005697, + "grad_norm": 0.926785409450531, + "learning_rate": 1.940545922668472e-05, + "loss": 0.8304, + "step": 17945 + }, + { + "epoch": 3.1949786324786325, + "grad_norm": 0.927204966545105, + "learning_rate": 1.939717365459952e-05, + "loss": 0.7698, + "step": 17946 + }, + { + "epoch": 3.195156695156695, + "grad_norm": 0.8717413544654846, + "learning_rate": 1.93888896617592e-05, + "loss": 0.7019, + "step": 17947 + }, + { + "epoch": 3.195334757834758, + "grad_norm": 0.8853579163551331, + "learning_rate": 1.938060724832619e-05, + "loss": 0.688, + "step": 17948 + }, + { + "epoch": 3.1955128205128207, + "grad_norm": 0.8655480742454529, + "learning_rate": 1.937232641446266e-05, + "loss": 0.5438, + "step": 17949 + }, + { + "epoch": 3.195690883190883, + "grad_norm": 1.0422619581222534, + "learning_rate": 1.936404716033092e-05, + "loss": 0.7796, + "step": 17950 + }, + { + "epoch": 3.1958689458689458, + "grad_norm": 0.9323866963386536, + "learning_rate": 1.9355769486093178e-05, + "loss": 0.8504, + "step": 17951 + }, + { + "epoch": 3.1960470085470085, + "grad_norm": 0.8164253830909729, + "learning_rate": 1.9347493391911585e-05, + "loss": 0.65, + "step": 17952 + }, + { + "epoch": 3.1962250712250713, + "grad_norm": 1.019322156906128, + "learning_rate": 1.9339218877948375e-05, + "loss": 0.7625, + "step": 17953 + }, + { + "epoch": 3.196403133903134, + "grad_norm": 1.0221740007400513, + "learning_rate": 1.9330945944365542e-05, + "loss": 0.766, + "step": 17954 + }, + { + "epoch": 3.1965811965811968, + "grad_norm": 1.0685542821884155, + "learning_rate": 1.9322674591325284e-05, + "loss": 0.6684, + "step": 17955 + }, + { + "epoch": 3.196759259259259, + "grad_norm": 0.9320390820503235, + "learning_rate": 1.9314404818989628e-05, + "loss": 0.8024, + "step": 17956 + }, + { + "epoch": 3.196937321937322, + "grad_norm": 0.9577845335006714, + "learning_rate": 1.9306136627520598e-05, + "loss": 0.8412, + "step": 17957 + }, + { + "epoch": 3.1971153846153846, + "grad_norm": 0.8444850444793701, + "learning_rate": 1.9297870017080187e-05, + "loss": 0.6378, + "step": 17958 + }, + { + "epoch": 3.1972934472934473, + "grad_norm": 0.905194103717804, + "learning_rate": 1.928960498783038e-05, + "loss": 0.7239, + "step": 17959 + }, + { + "epoch": 3.19747150997151, + "grad_norm": 0.9918950200080872, + "learning_rate": 1.9281341539933095e-05, + "loss": 0.8202, + "step": 17960 + }, + { + "epoch": 3.197649572649573, + "grad_norm": 0.9659561514854431, + "learning_rate": 1.927307967355024e-05, + "loss": 0.9186, + "step": 17961 + }, + { + "epoch": 3.197827635327635, + "grad_norm": 1.061000108718872, + "learning_rate": 1.926481938884368e-05, + "loss": 0.8304, + "step": 17962 + }, + { + "epoch": 3.198005698005698, + "grad_norm": 1.0199217796325684, + "learning_rate": 1.9256560685975313e-05, + "loss": 0.7649, + "step": 17963 + }, + { + "epoch": 3.1981837606837606, + "grad_norm": 0.9994050860404968, + "learning_rate": 1.9248303565106862e-05, + "loss": 0.7811, + "step": 17964 + }, + { + "epoch": 3.1983618233618234, + "grad_norm": 1.0911402702331543, + "learning_rate": 1.924004802640018e-05, + "loss": 0.7703, + "step": 17965 + }, + { + "epoch": 3.198539886039886, + "grad_norm": 0.878711998462677, + "learning_rate": 1.9231794070016995e-05, + "loss": 0.8968, + "step": 17966 + }, + { + "epoch": 3.198717948717949, + "grad_norm": 0.965321958065033, + "learning_rate": 1.9223541696119017e-05, + "loss": 0.646, + "step": 17967 + }, + { + "epoch": 3.198896011396011, + "grad_norm": 1.1991591453552246, + "learning_rate": 1.9215290904867945e-05, + "loss": 0.8747, + "step": 17968 + }, + { + "epoch": 3.199074074074074, + "grad_norm": 0.9755321145057678, + "learning_rate": 1.920704169642543e-05, + "loss": 0.7916, + "step": 17969 + }, + { + "epoch": 3.1992521367521367, + "grad_norm": 0.8560118079185486, + "learning_rate": 1.9198794070953074e-05, + "loss": 0.8143, + "step": 17970 + }, + { + "epoch": 3.1994301994301995, + "grad_norm": 1.079470157623291, + "learning_rate": 1.9190548028612532e-05, + "loss": 0.7934, + "step": 17971 + }, + { + "epoch": 3.199608262108262, + "grad_norm": 0.8946192264556885, + "learning_rate": 1.918230356956533e-05, + "loss": 0.8815, + "step": 17972 + }, + { + "epoch": 3.199786324786325, + "grad_norm": 0.8701974749565125, + "learning_rate": 1.9174060693973007e-05, + "loss": 0.7742, + "step": 17973 + }, + { + "epoch": 3.1999643874643873, + "grad_norm": 1.059370756149292, + "learning_rate": 1.9165819401997053e-05, + "loss": 0.7672, + "step": 17974 + }, + { + "epoch": 3.20014245014245, + "grad_norm": 1.151624083518982, + "learning_rate": 1.9157579693798955e-05, + "loss": 0.737, + "step": 17975 + }, + { + "epoch": 3.2003205128205128, + "grad_norm": 1.0536963939666748, + "learning_rate": 1.9149341569540158e-05, + "loss": 0.9501, + "step": 17976 + }, + { + "epoch": 3.2004985754985755, + "grad_norm": 0.8573943376541138, + "learning_rate": 1.9141105029382024e-05, + "loss": 0.7525, + "step": 17977 + }, + { + "epoch": 3.2006766381766383, + "grad_norm": 0.9258541464805603, + "learning_rate": 1.913287007348602e-05, + "loss": 0.6852, + "step": 17978 + }, + { + "epoch": 3.200854700854701, + "grad_norm": 0.9687190055847168, + "learning_rate": 1.9124636702013387e-05, + "loss": 0.7443, + "step": 17979 + }, + { + "epoch": 3.2010327635327633, + "grad_norm": 1.2201204299926758, + "learning_rate": 1.911640491512552e-05, + "loss": 0.7101, + "step": 17980 + }, + { + "epoch": 3.201210826210826, + "grad_norm": 1.032789707183838, + "learning_rate": 1.9108174712983675e-05, + "loss": 0.8236, + "step": 17981 + }, + { + "epoch": 3.201388888888889, + "grad_norm": 0.9653711318969727, + "learning_rate": 1.909994609574911e-05, + "loss": 0.582, + "step": 17982 + }, + { + "epoch": 3.2015669515669516, + "grad_norm": 0.9730488657951355, + "learning_rate": 1.9091719063583058e-05, + "loss": 0.8163, + "step": 17983 + }, + { + "epoch": 3.2017450142450143, + "grad_norm": 0.8995687365531921, + "learning_rate": 1.9083493616646686e-05, + "loss": 0.8088, + "step": 17984 + }, + { + "epoch": 3.201923076923077, + "grad_norm": 0.9465620517730713, + "learning_rate": 1.9075269755101167e-05, + "loss": 0.855, + "step": 17985 + }, + { + "epoch": 3.2021011396011394, + "grad_norm": 0.954304575920105, + "learning_rate": 1.9067047479107636e-05, + "loss": 0.7283, + "step": 17986 + }, + { + "epoch": 3.202279202279202, + "grad_norm": 0.9493895769119263, + "learning_rate": 1.9058826788827156e-05, + "loss": 0.7459, + "step": 17987 + }, + { + "epoch": 3.202457264957265, + "grad_norm": 0.9088132977485657, + "learning_rate": 1.9050607684420873e-05, + "loss": 0.8661, + "step": 17988 + }, + { + "epoch": 3.2026353276353277, + "grad_norm": 0.8793889880180359, + "learning_rate": 1.904239016604973e-05, + "loss": 0.6837, + "step": 17989 + }, + { + "epoch": 3.2028133903133904, + "grad_norm": 0.9571384191513062, + "learning_rate": 1.9034174233874803e-05, + "loss": 0.8113, + "step": 17990 + }, + { + "epoch": 3.202991452991453, + "grad_norm": 1.0003505945205688, + "learning_rate": 1.902595988805703e-05, + "loss": 0.7563, + "step": 17991 + }, + { + "epoch": 3.2031695156695155, + "grad_norm": 1.017555832862854, + "learning_rate": 1.9017747128757347e-05, + "loss": 0.8769, + "step": 17992 + }, + { + "epoch": 3.203347578347578, + "grad_norm": 0.8811057209968567, + "learning_rate": 1.9009535956136738e-05, + "loss": 0.7809, + "step": 17993 + }, + { + "epoch": 3.203525641025641, + "grad_norm": 0.9344915747642517, + "learning_rate": 1.9001326370355966e-05, + "loss": 0.7215, + "step": 17994 + }, + { + "epoch": 3.2037037037037037, + "grad_norm": 0.851786196231842, + "learning_rate": 1.8993118371576002e-05, + "loss": 0.6358, + "step": 17995 + }, + { + "epoch": 3.2038817663817665, + "grad_norm": 1.012872576713562, + "learning_rate": 1.8984911959957552e-05, + "loss": 0.6153, + "step": 17996 + }, + { + "epoch": 3.2040598290598292, + "grad_norm": 0.8867791891098022, + "learning_rate": 1.8976707135661488e-05, + "loss": 0.8027, + "step": 17997 + }, + { + "epoch": 3.2042378917378915, + "grad_norm": 0.954986572265625, + "learning_rate": 1.8968503898848532e-05, + "loss": 0.7257, + "step": 17998 + }, + { + "epoch": 3.2044159544159543, + "grad_norm": 1.006103277206421, + "learning_rate": 1.8960302249679418e-05, + "loss": 0.7259, + "step": 17999 + }, + { + "epoch": 3.204594017094017, + "grad_norm": 0.9252307415008545, + "learning_rate": 1.8952102188314835e-05, + "loss": 0.8521, + "step": 18000 + }, + { + "epoch": 3.20477207977208, + "grad_norm": 0.8885179758071899, + "learning_rate": 1.8943903714915447e-05, + "loss": 0.7776, + "step": 18001 + }, + { + "epoch": 3.2049501424501425, + "grad_norm": 0.9005871415138245, + "learning_rate": 1.8935706829641865e-05, + "loss": 0.8127, + "step": 18002 + }, + { + "epoch": 3.2051282051282053, + "grad_norm": 0.9123699069023132, + "learning_rate": 1.8927511532654753e-05, + "loss": 0.8827, + "step": 18003 + }, + { + "epoch": 3.205306267806268, + "grad_norm": 0.8967127203941345, + "learning_rate": 1.891931782411459e-05, + "loss": 0.586, + "step": 18004 + }, + { + "epoch": 3.2054843304843303, + "grad_norm": 0.8950326442718506, + "learning_rate": 1.8911125704181987e-05, + "loss": 0.8074, + "step": 18005 + }, + { + "epoch": 3.205662393162393, + "grad_norm": 0.8951900005340576, + "learning_rate": 1.8902935173017434e-05, + "loss": 0.8439, + "step": 18006 + }, + { + "epoch": 3.205840455840456, + "grad_norm": 0.9931556582450867, + "learning_rate": 1.88947462307814e-05, + "loss": 0.7222, + "step": 18007 + }, + { + "epoch": 3.2060185185185186, + "grad_norm": 0.8889574408531189, + "learning_rate": 1.888655887763433e-05, + "loss": 0.7148, + "step": 18008 + }, + { + "epoch": 3.2061965811965814, + "grad_norm": 0.9963932037353516, + "learning_rate": 1.8878373113736648e-05, + "loss": 0.7062, + "step": 18009 + }, + { + "epoch": 3.2063746438746437, + "grad_norm": 0.9292500019073486, + "learning_rate": 1.8870188939248724e-05, + "loss": 0.4827, + "step": 18010 + }, + { + "epoch": 3.2065527065527064, + "grad_norm": 1.1208369731903076, + "learning_rate": 1.8862006354330887e-05, + "loss": 0.9324, + "step": 18011 + }, + { + "epoch": 3.206730769230769, + "grad_norm": 0.9836950302124023, + "learning_rate": 1.885382535914352e-05, + "loss": 0.8058, + "step": 18012 + }, + { + "epoch": 3.206908831908832, + "grad_norm": 0.8917511701583862, + "learning_rate": 1.884564595384687e-05, + "loss": 0.8087, + "step": 18013 + }, + { + "epoch": 3.2070868945868947, + "grad_norm": 0.9193904399871826, + "learning_rate": 1.88374681386012e-05, + "loss": 0.9069, + "step": 18014 + }, + { + "epoch": 3.2072649572649574, + "grad_norm": 0.9080356359481812, + "learning_rate": 1.882929191356675e-05, + "loss": 0.6272, + "step": 18015 + }, + { + "epoch": 3.20744301994302, + "grad_norm": 0.955551028251648, + "learning_rate": 1.88211172789037e-05, + "loss": 0.8133, + "step": 18016 + }, + { + "epoch": 3.2076210826210825, + "grad_norm": 1.1204555034637451, + "learning_rate": 1.88129442347722e-05, + "loss": 0.7411, + "step": 18017 + }, + { + "epoch": 3.2077991452991452, + "grad_norm": 0.8058112859725952, + "learning_rate": 1.880477278133247e-05, + "loss": 0.6178, + "step": 18018 + }, + { + "epoch": 3.207977207977208, + "grad_norm": 0.9366341233253479, + "learning_rate": 1.8796602918744475e-05, + "loss": 0.7446, + "step": 18019 + }, + { + "epoch": 3.2081552706552707, + "grad_norm": 0.9692829251289368, + "learning_rate": 1.878843464716843e-05, + "loss": 0.8648, + "step": 18020 + }, + { + "epoch": 3.2083333333333335, + "grad_norm": 0.831867516040802, + "learning_rate": 1.8780267966764253e-05, + "loss": 0.6409, + "step": 18021 + }, + { + "epoch": 3.208511396011396, + "grad_norm": 1.0365211963653564, + "learning_rate": 1.877210287769202e-05, + "loss": 0.725, + "step": 18022 + }, + { + "epoch": 3.2086894586894585, + "grad_norm": 0.8367540836334229, + "learning_rate": 1.8763939380111705e-05, + "loss": 0.6112, + "step": 18023 + }, + { + "epoch": 3.2088675213675213, + "grad_norm": 0.9757236838340759, + "learning_rate": 1.8755777474183235e-05, + "loss": 0.8073, + "step": 18024 + }, + { + "epoch": 3.209045584045584, + "grad_norm": 0.930181622505188, + "learning_rate": 1.8747617160066543e-05, + "loss": 0.833, + "step": 18025 + }, + { + "epoch": 3.209223646723647, + "grad_norm": 0.8707014918327332, + "learning_rate": 1.873945843792151e-05, + "loss": 0.7944, + "step": 18026 + }, + { + "epoch": 3.2094017094017095, + "grad_norm": 0.8476516008377075, + "learning_rate": 1.8731301307907946e-05, + "loss": 0.8548, + "step": 18027 + }, + { + "epoch": 3.2095797720797723, + "grad_norm": 1.193192481994629, + "learning_rate": 1.8723145770185768e-05, + "loss": 0.9114, + "step": 18028 + }, + { + "epoch": 3.2097578347578346, + "grad_norm": 0.8446243405342102, + "learning_rate": 1.8714991824914662e-05, + "loss": 0.6701, + "step": 18029 + }, + { + "epoch": 3.2099358974358974, + "grad_norm": 0.9350818991661072, + "learning_rate": 1.8706839472254457e-05, + "loss": 0.7339, + "step": 18030 + }, + { + "epoch": 3.21011396011396, + "grad_norm": 1.001449465751648, + "learning_rate": 1.8698688712364864e-05, + "loss": 0.5902, + "step": 18031 + }, + { + "epoch": 3.210292022792023, + "grad_norm": 0.9731200337409973, + "learning_rate": 1.8690539545405582e-05, + "loss": 1.0019, + "step": 18032 + }, + { + "epoch": 3.2104700854700856, + "grad_norm": 1.0137463808059692, + "learning_rate": 1.868239197153626e-05, + "loss": 0.7242, + "step": 18033 + }, + { + "epoch": 3.210648148148148, + "grad_norm": 1.0123406648635864, + "learning_rate": 1.8674245990916528e-05, + "loss": 0.7168, + "step": 18034 + }, + { + "epoch": 3.2108262108262107, + "grad_norm": 0.9725518822669983, + "learning_rate": 1.8666101603706054e-05, + "loss": 0.8522, + "step": 18035 + }, + { + "epoch": 3.2110042735042734, + "grad_norm": 0.9783751964569092, + "learning_rate": 1.8657958810064315e-05, + "loss": 0.8206, + "step": 18036 + }, + { + "epoch": 3.211182336182336, + "grad_norm": 0.8869835734367371, + "learning_rate": 1.8649817610150923e-05, + "loss": 0.7986, + "step": 18037 + }, + { + "epoch": 3.211360398860399, + "grad_norm": 0.8768937587738037, + "learning_rate": 1.8641678004125363e-05, + "loss": 0.8307, + "step": 18038 + }, + { + "epoch": 3.2115384615384617, + "grad_norm": 0.8711394667625427, + "learning_rate": 1.863353999214712e-05, + "loss": 0.6374, + "step": 18039 + }, + { + "epoch": 3.2117165242165244, + "grad_norm": 1.0543094873428345, + "learning_rate": 1.862540357437563e-05, + "loss": 0.7071, + "step": 18040 + }, + { + "epoch": 3.2118945868945867, + "grad_norm": 0.8850560784339905, + "learning_rate": 1.861726875097032e-05, + "loss": 0.6987, + "step": 18041 + }, + { + "epoch": 3.2120726495726495, + "grad_norm": 0.8987910151481628, + "learning_rate": 1.860913552209055e-05, + "loss": 0.6632, + "step": 18042 + }, + { + "epoch": 3.2122507122507122, + "grad_norm": 0.9093405604362488, + "learning_rate": 1.860100388789574e-05, + "loss": 0.7153, + "step": 18043 + }, + { + "epoch": 3.212428774928775, + "grad_norm": 0.9864318370819092, + "learning_rate": 1.859287384854511e-05, + "loss": 0.8924, + "step": 18044 + }, + { + "epoch": 3.2126068376068377, + "grad_norm": 1.0086519718170166, + "learning_rate": 1.8584745404198066e-05, + "loss": 0.7406, + "step": 18045 + }, + { + "epoch": 3.2127849002849005, + "grad_norm": 0.8840755224227905, + "learning_rate": 1.8576618555013758e-05, + "loss": 0.756, + "step": 18046 + }, + { + "epoch": 3.212962962962963, + "grad_norm": 0.9135591983795166, + "learning_rate": 1.856849330115149e-05, + "loss": 0.565, + "step": 18047 + }, + { + "epoch": 3.2131410256410255, + "grad_norm": 0.9007353186607361, + "learning_rate": 1.856036964277045e-05, + "loss": 0.71, + "step": 18048 + }, + { + "epoch": 3.2133190883190883, + "grad_norm": 0.8159099817276001, + "learning_rate": 1.855224758002978e-05, + "loss": 0.5132, + "step": 18049 + }, + { + "epoch": 3.213497150997151, + "grad_norm": 0.9244372844696045, + "learning_rate": 1.8544127113088617e-05, + "loss": 0.813, + "step": 18050 + }, + { + "epoch": 3.213675213675214, + "grad_norm": 0.7983192205429077, + "learning_rate": 1.853600824210606e-05, + "loss": 0.6238, + "step": 18051 + }, + { + "epoch": 3.2138532763532766, + "grad_norm": 0.8808383941650391, + "learning_rate": 1.8527890967241212e-05, + "loss": 0.785, + "step": 18052 + }, + { + "epoch": 3.214031339031339, + "grad_norm": 0.8894946575164795, + "learning_rate": 1.8519775288653108e-05, + "loss": 0.8035, + "step": 18053 + }, + { + "epoch": 3.2142094017094016, + "grad_norm": 1.2804183959960938, + "learning_rate": 1.851166120650074e-05, + "loss": 0.7088, + "step": 18054 + }, + { + "epoch": 3.2143874643874644, + "grad_norm": 0.9724962711334229, + "learning_rate": 1.85035487209431e-05, + "loss": 0.7504, + "step": 18055 + }, + { + "epoch": 3.214565527065527, + "grad_norm": 1.0876386165618896, + "learning_rate": 1.8495437832139117e-05, + "loss": 0.8069, + "step": 18056 + }, + { + "epoch": 3.21474358974359, + "grad_norm": 0.9543545842170715, + "learning_rate": 1.8487328540247715e-05, + "loss": 0.8907, + "step": 18057 + }, + { + "epoch": 3.2149216524216526, + "grad_norm": 0.9640414118766785, + "learning_rate": 1.8479220845427802e-05, + "loss": 0.759, + "step": 18058 + }, + { + "epoch": 3.215099715099715, + "grad_norm": 0.8389877676963806, + "learning_rate": 1.847111474783817e-05, + "loss": 0.7234, + "step": 18059 + }, + { + "epoch": 3.2152777777777777, + "grad_norm": 0.9118861556053162, + "learning_rate": 1.8463010247637746e-05, + "loss": 0.852, + "step": 18060 + }, + { + "epoch": 3.2154558404558404, + "grad_norm": 0.9392750859260559, + "learning_rate": 1.8454907344985194e-05, + "loss": 0.8615, + "step": 18061 + }, + { + "epoch": 3.215633903133903, + "grad_norm": 0.9700032472610474, + "learning_rate": 1.8446806040039367e-05, + "loss": 0.7131, + "step": 18062 + }, + { + "epoch": 3.215811965811966, + "grad_norm": 0.894807755947113, + "learning_rate": 1.8438706332958965e-05, + "loss": 0.7727, + "step": 18063 + }, + { + "epoch": 3.2159900284900287, + "grad_norm": 0.85206538438797, + "learning_rate": 1.843060822390269e-05, + "loss": 0.7985, + "step": 18064 + }, + { + "epoch": 3.216168091168091, + "grad_norm": 1.073621392250061, + "learning_rate": 1.842251171302919e-05, + "loss": 0.7787, + "step": 18065 + }, + { + "epoch": 3.2163461538461537, + "grad_norm": 0.8918595314025879, + "learning_rate": 1.8414416800497125e-05, + "loss": 0.7749, + "step": 18066 + }, + { + "epoch": 3.2165242165242165, + "grad_norm": 0.8499172329902649, + "learning_rate": 1.8406323486465072e-05, + "loss": 0.3994, + "step": 18067 + }, + { + "epoch": 3.2167022792022792, + "grad_norm": 0.8019580245018005, + "learning_rate": 1.8398231771091613e-05, + "loss": 0.6772, + "step": 18068 + }, + { + "epoch": 3.216880341880342, + "grad_norm": 0.9750977754592896, + "learning_rate": 1.8390141654535265e-05, + "loss": 0.7399, + "step": 18069 + }, + { + "epoch": 3.2170584045584047, + "grad_norm": 1.0433316230773926, + "learning_rate": 1.8382053136954592e-05, + "loss": 0.767, + "step": 18070 + }, + { + "epoch": 3.217236467236467, + "grad_norm": 0.8822398781776428, + "learning_rate": 1.8373966218508043e-05, + "loss": 0.7635, + "step": 18071 + }, + { + "epoch": 3.21741452991453, + "grad_norm": 0.9446876049041748, + "learning_rate": 1.836588089935405e-05, + "loss": 0.6262, + "step": 18072 + }, + { + "epoch": 3.2175925925925926, + "grad_norm": 0.8917283415794373, + "learning_rate": 1.8357797179651047e-05, + "loss": 0.6993, + "step": 18073 + }, + { + "epoch": 3.2177706552706553, + "grad_norm": 0.8711534738540649, + "learning_rate": 1.8349715059557393e-05, + "loss": 0.7275, + "step": 18074 + }, + { + "epoch": 3.217948717948718, + "grad_norm": 1.0784484148025513, + "learning_rate": 1.83416345392315e-05, + "loss": 0.8006, + "step": 18075 + }, + { + "epoch": 3.218126780626781, + "grad_norm": 0.8539456725120544, + "learning_rate": 1.83335556188316e-05, + "loss": 0.68, + "step": 18076 + }, + { + "epoch": 3.218304843304843, + "grad_norm": 0.9416542053222656, + "learning_rate": 1.8325478298516063e-05, + "loss": 0.7378, + "step": 18077 + }, + { + "epoch": 3.218482905982906, + "grad_norm": 0.9718979001045227, + "learning_rate": 1.831740257844311e-05, + "loss": 0.7148, + "step": 18078 + }, + { + "epoch": 3.2186609686609686, + "grad_norm": 0.8982722163200378, + "learning_rate": 1.830932845877097e-05, + "loss": 0.8073, + "step": 18079 + }, + { + "epoch": 3.2188390313390314, + "grad_norm": 0.9609165191650391, + "learning_rate": 1.8301255939657834e-05, + "loss": 0.7771, + "step": 18080 + }, + { + "epoch": 3.219017094017094, + "grad_norm": 0.9416554570198059, + "learning_rate": 1.8293185021261884e-05, + "loss": 0.822, + "step": 18081 + }, + { + "epoch": 3.219195156695157, + "grad_norm": 0.8695271611213684, + "learning_rate": 1.828511570374124e-05, + "loss": 0.6192, + "step": 18082 + }, + { + "epoch": 3.219373219373219, + "grad_norm": 0.8800650238990784, + "learning_rate": 1.8277047987254003e-05, + "loss": 0.6805, + "step": 18083 + }, + { + "epoch": 3.219551282051282, + "grad_norm": 0.9437993764877319, + "learning_rate": 1.8268981871958225e-05, + "loss": 0.7948, + "step": 18084 + }, + { + "epoch": 3.2197293447293447, + "grad_norm": 1.0480988025665283, + "learning_rate": 1.826091735801202e-05, + "loss": 0.8144, + "step": 18085 + }, + { + "epoch": 3.2199074074074074, + "grad_norm": 0.9806156158447266, + "learning_rate": 1.8252854445573276e-05, + "loss": 0.5886, + "step": 18086 + }, + { + "epoch": 3.22008547008547, + "grad_norm": 1.0387870073318481, + "learning_rate": 1.8244793134800075e-05, + "loss": 0.7026, + "step": 18087 + }, + { + "epoch": 3.220263532763533, + "grad_norm": 1.1478195190429688, + "learning_rate": 1.8236733425850305e-05, + "loss": 0.7987, + "step": 18088 + }, + { + "epoch": 3.2204415954415953, + "grad_norm": 0.9605230093002319, + "learning_rate": 1.8228675318881906e-05, + "loss": 0.6876, + "step": 18089 + }, + { + "epoch": 3.220619658119658, + "grad_norm": 0.955807089805603, + "learning_rate": 1.822061881405275e-05, + "loss": 0.6625, + "step": 18090 + }, + { + "epoch": 3.2207977207977208, + "grad_norm": 0.8398988842964172, + "learning_rate": 1.8212563911520664e-05, + "loss": 0.6911, + "step": 18091 + }, + { + "epoch": 3.2209757834757835, + "grad_norm": 0.863762378692627, + "learning_rate": 1.8204510611443526e-05, + "loss": 0.7064, + "step": 18092 + }, + { + "epoch": 3.2211538461538463, + "grad_norm": 0.908447802066803, + "learning_rate": 1.8196458913979053e-05, + "loss": 0.7601, + "step": 18093 + }, + { + "epoch": 3.221331908831909, + "grad_norm": 0.9673567414283752, + "learning_rate": 1.818840881928505e-05, + "loss": 0.9187, + "step": 18094 + }, + { + "epoch": 3.2215099715099713, + "grad_norm": 0.9876523613929749, + "learning_rate": 1.818036032751923e-05, + "loss": 1.1142, + "step": 18095 + }, + { + "epoch": 3.221688034188034, + "grad_norm": 0.9074375033378601, + "learning_rate": 1.8172313438839284e-05, + "loss": 0.7491, + "step": 18096 + }, + { + "epoch": 3.221866096866097, + "grad_norm": 0.9420785903930664, + "learning_rate": 1.8164268153402875e-05, + "loss": 0.7464, + "step": 18097 + }, + { + "epoch": 3.2220441595441596, + "grad_norm": 0.9435096979141235, + "learning_rate": 1.815622447136762e-05, + "loss": 0.7585, + "step": 18098 + }, + { + "epoch": 3.2222222222222223, + "grad_norm": 0.8565973043441772, + "learning_rate": 1.814818239289112e-05, + "loss": 0.6909, + "step": 18099 + }, + { + "epoch": 3.222400284900285, + "grad_norm": 0.9183077812194824, + "learning_rate": 1.8140141918131004e-05, + "loss": 0.6136, + "step": 18100 + }, + { + "epoch": 3.2225783475783474, + "grad_norm": 0.8680392503738403, + "learning_rate": 1.813210304724471e-05, + "loss": 0.6443, + "step": 18101 + }, + { + "epoch": 3.22275641025641, + "grad_norm": 0.9555699229240417, + "learning_rate": 1.812406578038983e-05, + "loss": 0.8166, + "step": 18102 + }, + { + "epoch": 3.222934472934473, + "grad_norm": 1.0517683029174805, + "learning_rate": 1.811603011772375e-05, + "loss": 0.7405, + "step": 18103 + }, + { + "epoch": 3.2231125356125356, + "grad_norm": 0.8452093601226807, + "learning_rate": 1.8107996059403985e-05, + "loss": 0.8957, + "step": 18104 + }, + { + "epoch": 3.2232905982905984, + "grad_norm": 1.0134345293045044, + "learning_rate": 1.8099963605587932e-05, + "loss": 0.9799, + "step": 18105 + }, + { + "epoch": 3.223468660968661, + "grad_norm": 0.9043877720832825, + "learning_rate": 1.809193275643295e-05, + "loss": 0.6479, + "step": 18106 + }, + { + "epoch": 3.2236467236467234, + "grad_norm": 1.0354737043380737, + "learning_rate": 1.80839035120964e-05, + "loss": 0.8212, + "step": 18107 + }, + { + "epoch": 3.223824786324786, + "grad_norm": 1.0781883001327515, + "learning_rate": 1.8075875872735602e-05, + "loss": 0.8305, + "step": 18108 + }, + { + "epoch": 3.224002849002849, + "grad_norm": 0.9643750786781311, + "learning_rate": 1.80678498385078e-05, + "loss": 0.5684, + "step": 18109 + }, + { + "epoch": 3.2241809116809117, + "grad_norm": 0.9557623267173767, + "learning_rate": 1.8059825409570308e-05, + "loss": 0.7995, + "step": 18110 + }, + { + "epoch": 3.2243589743589745, + "grad_norm": 0.9005846381187439, + "learning_rate": 1.8051802586080312e-05, + "loss": 0.655, + "step": 18111 + }, + { + "epoch": 3.224537037037037, + "grad_norm": 0.985146701335907, + "learning_rate": 1.8043781368195024e-05, + "loss": 0.9267, + "step": 18112 + }, + { + "epoch": 3.2247150997150995, + "grad_norm": 0.9472562670707703, + "learning_rate": 1.803576175607159e-05, + "loss": 0.7262, + "step": 18113 + }, + { + "epoch": 3.2248931623931623, + "grad_norm": 1.2001210451126099, + "learning_rate": 1.802774374986711e-05, + "loss": 0.7161, + "step": 18114 + }, + { + "epoch": 3.225071225071225, + "grad_norm": 0.8562248945236206, + "learning_rate": 1.801972734973876e-05, + "loss": 0.7848, + "step": 18115 + }, + { + "epoch": 3.2252492877492878, + "grad_norm": 0.942823052406311, + "learning_rate": 1.8011712555843496e-05, + "loss": 0.8428, + "step": 18116 + }, + { + "epoch": 3.2254273504273505, + "grad_norm": 0.9686674475669861, + "learning_rate": 1.8003699368338466e-05, + "loss": 0.6471, + "step": 18117 + }, + { + "epoch": 3.2256054131054133, + "grad_norm": 1.0388344526290894, + "learning_rate": 1.7995687787380566e-05, + "loss": 0.7895, + "step": 18118 + }, + { + "epoch": 3.2257834757834756, + "grad_norm": 0.9582616686820984, + "learning_rate": 1.7987677813126836e-05, + "loss": 0.7369, + "step": 18119 + }, + { + "epoch": 3.2259615384615383, + "grad_norm": 0.9572505354881287, + "learning_rate": 1.7979669445734194e-05, + "loss": 0.7163, + "step": 18120 + }, + { + "epoch": 3.226139601139601, + "grad_norm": 0.9522523880004883, + "learning_rate": 1.7971662685359547e-05, + "loss": 0.8887, + "step": 18121 + }, + { + "epoch": 3.226317663817664, + "grad_norm": 1.0779093503952026, + "learning_rate": 1.7963657532159773e-05, + "loss": 0.8593, + "step": 18122 + }, + { + "epoch": 3.2264957264957266, + "grad_norm": 0.926485538482666, + "learning_rate": 1.795565398629171e-05, + "loss": 0.7664, + "step": 18123 + }, + { + "epoch": 3.2266737891737893, + "grad_norm": 1.0132343769073486, + "learning_rate": 1.7947652047912144e-05, + "loss": 0.7165, + "step": 18124 + }, + { + "epoch": 3.226851851851852, + "grad_norm": 0.8460900783538818, + "learning_rate": 1.793965171717795e-05, + "loss": 0.5995, + "step": 18125 + }, + { + "epoch": 3.2270299145299144, + "grad_norm": 0.8448793292045593, + "learning_rate": 1.793165299424576e-05, + "loss": 0.6078, + "step": 18126 + }, + { + "epoch": 3.227207977207977, + "grad_norm": 1.0588792562484741, + "learning_rate": 1.7923655879272393e-05, + "loss": 0.7189, + "step": 18127 + }, + { + "epoch": 3.22738603988604, + "grad_norm": 0.9055169820785522, + "learning_rate": 1.7915660372414443e-05, + "loss": 0.7475, + "step": 18128 + }, + { + "epoch": 3.2275641025641026, + "grad_norm": 0.9485293030738831, + "learning_rate": 1.7907666473828643e-05, + "loss": 0.7638, + "step": 18129 + }, + { + "epoch": 3.2277421652421654, + "grad_norm": 0.8959475755691528, + "learning_rate": 1.7899674183671576e-05, + "loss": 0.7253, + "step": 18130 + }, + { + "epoch": 3.2279202279202277, + "grad_norm": 0.9282844066619873, + "learning_rate": 1.789168350209983e-05, + "loss": 0.7165, + "step": 18131 + }, + { + "epoch": 3.2280982905982905, + "grad_norm": 0.9341546297073364, + "learning_rate": 1.7883694429270026e-05, + "loss": 0.7162, + "step": 18132 + }, + { + "epoch": 3.228276353276353, + "grad_norm": 0.9306660890579224, + "learning_rate": 1.787570696533859e-05, + "loss": 0.7153, + "step": 18133 + }, + { + "epoch": 3.228454415954416, + "grad_norm": 1.024999976158142, + "learning_rate": 1.786772111046212e-05, + "loss": 0.819, + "step": 18134 + }, + { + "epoch": 3.2286324786324787, + "grad_norm": 0.9674655795097351, + "learning_rate": 1.7859736864797027e-05, + "loss": 0.9286, + "step": 18135 + }, + { + "epoch": 3.2288105413105415, + "grad_norm": 0.9502766728401184, + "learning_rate": 1.7851754228499763e-05, + "loss": 0.6461, + "step": 18136 + }, + { + "epoch": 3.228988603988604, + "grad_norm": 1.0583183765411377, + "learning_rate": 1.784377320172672e-05, + "loss": 0.812, + "step": 18137 + }, + { + "epoch": 3.2291666666666665, + "grad_norm": 0.9369902014732361, + "learning_rate": 1.783579378463428e-05, + "loss": 0.7371, + "step": 18138 + }, + { + "epoch": 3.2293447293447293, + "grad_norm": 0.9534862637519836, + "learning_rate": 1.7827815977378782e-05, + "loss": 0.8534, + "step": 18139 + }, + { + "epoch": 3.229522792022792, + "grad_norm": 0.9020227789878845, + "learning_rate": 1.7819839780116533e-05, + "loss": 0.6451, + "step": 18140 + }, + { + "epoch": 3.2297008547008548, + "grad_norm": 1.0265909433364868, + "learning_rate": 1.7811865193003784e-05, + "loss": 0.9164, + "step": 18141 + }, + { + "epoch": 3.2298789173789175, + "grad_norm": 0.8694081902503967, + "learning_rate": 1.7803892216196848e-05, + "loss": 0.6521, + "step": 18142 + }, + { + "epoch": 3.23005698005698, + "grad_norm": 0.9590677618980408, + "learning_rate": 1.7795920849851854e-05, + "loss": 0.787, + "step": 18143 + }, + { + "epoch": 3.2302350427350426, + "grad_norm": 0.9131260514259338, + "learning_rate": 1.7787951094125055e-05, + "loss": 0.7033, + "step": 18144 + }, + { + "epoch": 3.2304131054131053, + "grad_norm": 0.9520593881607056, + "learning_rate": 1.7779982949172568e-05, + "loss": 0.7207, + "step": 18145 + }, + { + "epoch": 3.230591168091168, + "grad_norm": 0.8993405103683472, + "learning_rate": 1.7772016415150518e-05, + "loss": 0.6616, + "step": 18146 + }, + { + "epoch": 3.230769230769231, + "grad_norm": 0.9910643100738525, + "learning_rate": 1.7764051492214994e-05, + "loss": 0.7724, + "step": 18147 + }, + { + "epoch": 3.2309472934472936, + "grad_norm": 0.9680421948432922, + "learning_rate": 1.7756088180522045e-05, + "loss": 0.7549, + "step": 18148 + }, + { + "epoch": 3.2311253561253563, + "grad_norm": 0.8450651168823242, + "learning_rate": 1.774812648022768e-05, + "loss": 0.6817, + "step": 18149 + }, + { + "epoch": 3.2313034188034186, + "grad_norm": 1.028572678565979, + "learning_rate": 1.7740166391487946e-05, + "loss": 0.7072, + "step": 18150 + }, + { + "epoch": 3.2314814814814814, + "grad_norm": 0.9712592959403992, + "learning_rate": 1.7732207914458754e-05, + "loss": 0.6691, + "step": 18151 + }, + { + "epoch": 3.231659544159544, + "grad_norm": 1.0358368158340454, + "learning_rate": 1.772425104929607e-05, + "loss": 0.6961, + "step": 18152 + }, + { + "epoch": 3.231837606837607, + "grad_norm": 1.058782935142517, + "learning_rate": 1.771629579615576e-05, + "loss": 0.7609, + "step": 18153 + }, + { + "epoch": 3.2320156695156697, + "grad_norm": 0.9332877993583679, + "learning_rate": 1.7708342155193715e-05, + "loss": 0.7787, + "step": 18154 + }, + { + "epoch": 3.232193732193732, + "grad_norm": 0.9538434147834778, + "learning_rate": 1.7700390126565747e-05, + "loss": 0.8176, + "step": 18155 + }, + { + "epoch": 3.2323717948717947, + "grad_norm": 0.9596571922302246, + "learning_rate": 1.7692439710427655e-05, + "loss": 0.8309, + "step": 18156 + }, + { + "epoch": 3.2325498575498575, + "grad_norm": 1.0157568454742432, + "learning_rate": 1.768449090693528e-05, + "loss": 1.1549, + "step": 18157 + }, + { + "epoch": 3.23272792022792, + "grad_norm": 1.0346252918243408, + "learning_rate": 1.7676543716244254e-05, + "loss": 0.892, + "step": 18158 + }, + { + "epoch": 3.232905982905983, + "grad_norm": 0.8448618650436401, + "learning_rate": 1.766859813851037e-05, + "loss": 0.6618, + "step": 18159 + }, + { + "epoch": 3.2330840455840457, + "grad_norm": 0.9361056685447693, + "learning_rate": 1.766065417388928e-05, + "loss": 0.6775, + "step": 18160 + }, + { + "epoch": 3.2332621082621085, + "grad_norm": 1.0730383396148682, + "learning_rate": 1.7652711822536617e-05, + "loss": 0.7644, + "step": 18161 + }, + { + "epoch": 3.2334401709401708, + "grad_norm": 0.8408196568489075, + "learning_rate": 1.7644771084608015e-05, + "loss": 0.742, + "step": 18162 + }, + { + "epoch": 3.2336182336182335, + "grad_norm": 1.196700096130371, + "learning_rate": 1.7636831960259038e-05, + "loss": 0.7993, + "step": 18163 + }, + { + "epoch": 3.2337962962962963, + "grad_norm": 0.781534731388092, + "learning_rate": 1.7628894449645238e-05, + "loss": 0.5319, + "step": 18164 + }, + { + "epoch": 3.233974358974359, + "grad_norm": 0.9107591509819031, + "learning_rate": 1.762095855292215e-05, + "loss": 0.6581, + "step": 18165 + }, + { + "epoch": 3.234152421652422, + "grad_norm": 1.1640515327453613, + "learning_rate": 1.7613024270245226e-05, + "loss": 1.0668, + "step": 18166 + }, + { + "epoch": 3.2343304843304845, + "grad_norm": 0.9013523459434509, + "learning_rate": 1.7605091601769998e-05, + "loss": 0.5969, + "step": 18167 + }, + { + "epoch": 3.234508547008547, + "grad_norm": 0.8890687227249146, + "learning_rate": 1.759716054765178e-05, + "loss": 0.5108, + "step": 18168 + }, + { + "epoch": 3.2346866096866096, + "grad_norm": 0.995286226272583, + "learning_rate": 1.7589231108046068e-05, + "loss": 0.8642, + "step": 18169 + }, + { + "epoch": 3.2348646723646723, + "grad_norm": 0.8938494920730591, + "learning_rate": 1.7581303283108164e-05, + "loss": 0.7484, + "step": 18170 + }, + { + "epoch": 3.235042735042735, + "grad_norm": 0.9409791827201843, + "learning_rate": 1.757337707299339e-05, + "loss": 0.7209, + "step": 18171 + }, + { + "epoch": 3.235220797720798, + "grad_norm": 0.8960477709770203, + "learning_rate": 1.7565452477857115e-05, + "loss": 0.7344, + "step": 18172 + }, + { + "epoch": 3.2353988603988606, + "grad_norm": 1.0025544166564941, + "learning_rate": 1.755752949785451e-05, + "loss": 0.8146, + "step": 18173 + }, + { + "epoch": 3.235576923076923, + "grad_norm": 1.1479045152664185, + "learning_rate": 1.7549608133140895e-05, + "loss": 0.9372, + "step": 18174 + }, + { + "epoch": 3.2357549857549857, + "grad_norm": 0.8442793488502502, + "learning_rate": 1.754168838387139e-05, + "loss": 0.6714, + "step": 18175 + }, + { + "epoch": 3.2359330484330484, + "grad_norm": 0.985490083694458, + "learning_rate": 1.753377025020123e-05, + "loss": 0.9846, + "step": 18176 + }, + { + "epoch": 3.236111111111111, + "grad_norm": 1.0424950122833252, + "learning_rate": 1.7525853732285545e-05, + "loss": 0.8757, + "step": 18177 + }, + { + "epoch": 3.236289173789174, + "grad_norm": 0.8219842314720154, + "learning_rate": 1.7517938830279423e-05, + "loss": 0.5383, + "step": 18178 + }, + { + "epoch": 3.2364672364672367, + "grad_norm": 0.9854603409767151, + "learning_rate": 1.751002554433795e-05, + "loss": 0.8043, + "step": 18179 + }, + { + "epoch": 3.236645299145299, + "grad_norm": 0.9642623662948608, + "learning_rate": 1.7502113874616167e-05, + "loss": 0.6575, + "step": 18180 + }, + { + "epoch": 3.2368233618233617, + "grad_norm": 0.9247049689292908, + "learning_rate": 1.7494203821269062e-05, + "loss": 0.6501, + "step": 18181 + }, + { + "epoch": 3.2370014245014245, + "grad_norm": 0.8187673091888428, + "learning_rate": 1.748629538445169e-05, + "loss": 0.5873, + "step": 18182 + }, + { + "epoch": 3.2371794871794872, + "grad_norm": 0.957827627658844, + "learning_rate": 1.7478388564318905e-05, + "loss": 0.8297, + "step": 18183 + }, + { + "epoch": 3.23735754985755, + "grad_norm": 1.021424651145935, + "learning_rate": 1.747048336102569e-05, + "loss": 0.6894, + "step": 18184 + }, + { + "epoch": 3.2375356125356127, + "grad_norm": 0.9959956407546997, + "learning_rate": 1.746257977472693e-05, + "loss": 0.7137, + "step": 18185 + }, + { + "epoch": 3.237713675213675, + "grad_norm": 0.9278098940849304, + "learning_rate": 1.7454677805577445e-05, + "loss": 0.8262, + "step": 18186 + }, + { + "epoch": 3.237891737891738, + "grad_norm": 0.9757589101791382, + "learning_rate": 1.7446777453732088e-05, + "loss": 0.7696, + "step": 18187 + }, + { + "epoch": 3.2380698005698005, + "grad_norm": 0.9537236094474792, + "learning_rate": 1.743887871934563e-05, + "loss": 0.8835, + "step": 18188 + }, + { + "epoch": 3.2382478632478633, + "grad_norm": 1.00649893283844, + "learning_rate": 1.743098160257284e-05, + "loss": 0.7316, + "step": 18189 + }, + { + "epoch": 3.238425925925926, + "grad_norm": 0.8231672644615173, + "learning_rate": 1.742308610356842e-05, + "loss": 0.7836, + "step": 18190 + }, + { + "epoch": 3.238603988603989, + "grad_norm": 1.019817590713501, + "learning_rate": 1.741519222248711e-05, + "loss": 0.8218, + "step": 18191 + }, + { + "epoch": 3.238782051282051, + "grad_norm": 1.0095518827438354, + "learning_rate": 1.740729995948356e-05, + "loss": 0.7905, + "step": 18192 + }, + { + "epoch": 3.238960113960114, + "grad_norm": 1.031113862991333, + "learning_rate": 1.739940931471239e-05, + "loss": 0.7208, + "step": 18193 + }, + { + "epoch": 3.2391381766381766, + "grad_norm": 0.7880609035491943, + "learning_rate": 1.7391520288328212e-05, + "loss": 0.598, + "step": 18194 + }, + { + "epoch": 3.2393162393162394, + "grad_norm": 0.9234829545021057, + "learning_rate": 1.738363288048558e-05, + "loss": 0.7713, + "step": 18195 + }, + { + "epoch": 3.239494301994302, + "grad_norm": 0.8925933241844177, + "learning_rate": 1.7375747091339024e-05, + "loss": 0.6134, + "step": 18196 + }, + { + "epoch": 3.239672364672365, + "grad_norm": 0.9539459943771362, + "learning_rate": 1.736786292104312e-05, + "loss": 0.794, + "step": 18197 + }, + { + "epoch": 3.239850427350427, + "grad_norm": 0.9326531291007996, + "learning_rate": 1.7359980369752228e-05, + "loss": 0.6804, + "step": 18198 + }, + { + "epoch": 3.24002849002849, + "grad_norm": 0.9107605218887329, + "learning_rate": 1.7352099437620907e-05, + "loss": 0.6677, + "step": 18199 + }, + { + "epoch": 3.2402065527065527, + "grad_norm": 1.0316441059112549, + "learning_rate": 1.7344220124803468e-05, + "loss": 0.7564, + "step": 18200 + }, + { + "epoch": 3.2403846153846154, + "grad_norm": 1.1710155010223389, + "learning_rate": 1.7336342431454345e-05, + "loss": 0.8665, + "step": 18201 + }, + { + "epoch": 3.240562678062678, + "grad_norm": 1.002057671546936, + "learning_rate": 1.732846635772788e-05, + "loss": 0.6974, + "step": 18202 + }, + { + "epoch": 3.240740740740741, + "grad_norm": 0.8545549511909485, + "learning_rate": 1.732059190377838e-05, + "loss": 0.6291, + "step": 18203 + }, + { + "epoch": 3.2409188034188032, + "grad_norm": 0.9257392287254333, + "learning_rate": 1.731271906976013e-05, + "loss": 0.8188, + "step": 18204 + }, + { + "epoch": 3.241096866096866, + "grad_norm": 0.8662429451942444, + "learning_rate": 1.7304847855827388e-05, + "loss": 0.7252, + "step": 18205 + }, + { + "epoch": 3.2412749287749287, + "grad_norm": 0.8656207919120789, + "learning_rate": 1.7296978262134333e-05, + "loss": 0.6465, + "step": 18206 + }, + { + "epoch": 3.2414529914529915, + "grad_norm": 0.8379673361778259, + "learning_rate": 1.728911028883524e-05, + "loss": 0.7603, + "step": 18207 + }, + { + "epoch": 3.2416310541310542, + "grad_norm": 0.8996121287345886, + "learning_rate": 1.7281243936084168e-05, + "loss": 0.7718, + "step": 18208 + }, + { + "epoch": 3.241809116809117, + "grad_norm": 0.9700300097465515, + "learning_rate": 1.727337920403531e-05, + "loss": 0.9161, + "step": 18209 + }, + { + "epoch": 3.2419871794871793, + "grad_norm": 0.9493532776832581, + "learning_rate": 1.726551609284275e-05, + "loss": 0.8266, + "step": 18210 + }, + { + "epoch": 3.242165242165242, + "grad_norm": 0.9822573661804199, + "learning_rate": 1.7257654602660523e-05, + "loss": 0.743, + "step": 18211 + }, + { + "epoch": 3.242343304843305, + "grad_norm": 0.9318371415138245, + "learning_rate": 1.7249794733642665e-05, + "loss": 0.6753, + "step": 18212 + }, + { + "epoch": 3.2425213675213675, + "grad_norm": 0.9471957683563232, + "learning_rate": 1.7241936485943168e-05, + "loss": 0.8213, + "step": 18213 + }, + { + "epoch": 3.2426994301994303, + "grad_norm": 1.0344058275222778, + "learning_rate": 1.7234079859716057e-05, + "loss": 0.897, + "step": 18214 + }, + { + "epoch": 3.242877492877493, + "grad_norm": 1.0287681818008423, + "learning_rate": 1.722622485511517e-05, + "loss": 0.8328, + "step": 18215 + }, + { + "epoch": 3.2430555555555554, + "grad_norm": 1.0139912366867065, + "learning_rate": 1.721837147229448e-05, + "loss": 0.7253, + "step": 18216 + }, + { + "epoch": 3.243233618233618, + "grad_norm": 0.9414315819740295, + "learning_rate": 1.7210519711407847e-05, + "loss": 0.6769, + "step": 18217 + }, + { + "epoch": 3.243411680911681, + "grad_norm": 0.9098498225212097, + "learning_rate": 1.7202669572609088e-05, + "loss": 0.8154, + "step": 18218 + }, + { + "epoch": 3.2435897435897436, + "grad_norm": 0.8019130825996399, + "learning_rate": 1.7194821056052033e-05, + "loss": 0.7373, + "step": 18219 + }, + { + "epoch": 3.2437678062678064, + "grad_norm": 0.8887826800346375, + "learning_rate": 1.7186974161890435e-05, + "loss": 0.6664, + "step": 18220 + }, + { + "epoch": 3.243945868945869, + "grad_norm": 0.9187814593315125, + "learning_rate": 1.7179128890278028e-05, + "loss": 0.8802, + "step": 18221 + }, + { + "epoch": 3.2441239316239314, + "grad_norm": 1.022515892982483, + "learning_rate": 1.7171285241368606e-05, + "loss": 0.8225, + "step": 18222 + }, + { + "epoch": 3.244301994301994, + "grad_norm": 0.9992332458496094, + "learning_rate": 1.716344321531572e-05, + "loss": 0.747, + "step": 18223 + }, + { + "epoch": 3.244480056980057, + "grad_norm": 0.9298896193504333, + "learning_rate": 1.7155602812273152e-05, + "loss": 0.6622, + "step": 18224 + }, + { + "epoch": 3.2446581196581197, + "grad_norm": 0.9849324822425842, + "learning_rate": 1.71477640323944e-05, + "loss": 0.7203, + "step": 18225 + }, + { + "epoch": 3.2448361823361824, + "grad_norm": 0.9874289035797119, + "learning_rate": 1.7139926875833124e-05, + "loss": 0.7284, + "step": 18226 + }, + { + "epoch": 3.245014245014245, + "grad_norm": 0.9077398777008057, + "learning_rate": 1.713209134274285e-05, + "loss": 0.6127, + "step": 18227 + }, + { + "epoch": 3.2451923076923075, + "grad_norm": 0.9049743413925171, + "learning_rate": 1.7124257433277102e-05, + "loss": 0.8492, + "step": 18228 + }, + { + "epoch": 3.2453703703703702, + "grad_norm": 0.7180083394050598, + "learning_rate": 1.7116425147589378e-05, + "loss": 0.4446, + "step": 18229 + }, + { + "epoch": 3.245548433048433, + "grad_norm": 1.1515785455703735, + "learning_rate": 1.7108594485833095e-05, + "loss": 0.8079, + "step": 18230 + }, + { + "epoch": 3.2457264957264957, + "grad_norm": 1.0877320766448975, + "learning_rate": 1.710076544816174e-05, + "loss": 0.7127, + "step": 18231 + }, + { + "epoch": 3.2459045584045585, + "grad_norm": 0.8861889839172363, + "learning_rate": 1.7092938034728677e-05, + "loss": 0.6444, + "step": 18232 + }, + { + "epoch": 3.2460826210826212, + "grad_norm": 1.0296967029571533, + "learning_rate": 1.7085112245687263e-05, + "loss": 0.7912, + "step": 18233 + }, + { + "epoch": 3.246260683760684, + "grad_norm": 0.9024568796157837, + "learning_rate": 1.7077288081190846e-05, + "loss": 0.5452, + "step": 18234 + }, + { + "epoch": 3.2464387464387463, + "grad_norm": 0.9017095565795898, + "learning_rate": 1.7069465541392703e-05, + "loss": 0.7819, + "step": 18235 + }, + { + "epoch": 3.246616809116809, + "grad_norm": 0.9934529662132263, + "learning_rate": 1.7061644626446116e-05, + "loss": 0.8196, + "step": 18236 + }, + { + "epoch": 3.246794871794872, + "grad_norm": 1.0247230529785156, + "learning_rate": 1.7053825336504304e-05, + "loss": 0.9648, + "step": 18237 + }, + { + "epoch": 3.2469729344729346, + "grad_norm": 0.9507484436035156, + "learning_rate": 1.7046007671720453e-05, + "loss": 0.8102, + "step": 18238 + }, + { + "epoch": 3.2471509971509973, + "grad_norm": 1.002600908279419, + "learning_rate": 1.7038191632247826e-05, + "loss": 0.7973, + "step": 18239 + }, + { + "epoch": 3.2473290598290596, + "grad_norm": 0.963873028755188, + "learning_rate": 1.7030377218239434e-05, + "loss": 0.9603, + "step": 18240 + }, + { + "epoch": 3.2475071225071224, + "grad_norm": 0.9259035587310791, + "learning_rate": 1.7022564429848465e-05, + "loss": 0.7526, + "step": 18241 + }, + { + "epoch": 3.247685185185185, + "grad_norm": 1.0171494483947754, + "learning_rate": 1.701475326722799e-05, + "loss": 0.8504, + "step": 18242 + }, + { + "epoch": 3.247863247863248, + "grad_norm": 0.8850985765457153, + "learning_rate": 1.700694373053102e-05, + "loss": 0.7934, + "step": 18243 + }, + { + "epoch": 3.2480413105413106, + "grad_norm": 0.8500893712043762, + "learning_rate": 1.6999135819910607e-05, + "loss": 0.7683, + "step": 18244 + }, + { + "epoch": 3.2482193732193734, + "grad_norm": 0.9575412273406982, + "learning_rate": 1.6991329535519685e-05, + "loss": 0.7072, + "step": 18245 + }, + { + "epoch": 3.248397435897436, + "grad_norm": 1.0330621004104614, + "learning_rate": 1.6983524877511236e-05, + "loss": 0.7366, + "step": 18246 + }, + { + "epoch": 3.2485754985754984, + "grad_norm": 0.9334800839424133, + "learning_rate": 1.6975721846038173e-05, + "loss": 0.768, + "step": 18247 + }, + { + "epoch": 3.248753561253561, + "grad_norm": 1.092702031135559, + "learning_rate": 1.696792044125334e-05, + "loss": 0.5735, + "step": 18248 + }, + { + "epoch": 3.248931623931624, + "grad_norm": 0.8769099116325378, + "learning_rate": 1.6960120663309643e-05, + "loss": 0.7181, + "step": 18249 + }, + { + "epoch": 3.2491096866096867, + "grad_norm": 0.8985859751701355, + "learning_rate": 1.6952322512359886e-05, + "loss": 0.6442, + "step": 18250 + }, + { + "epoch": 3.2492877492877494, + "grad_norm": 0.9657853841781616, + "learning_rate": 1.6944525988556847e-05, + "loss": 0.8127, + "step": 18251 + }, + { + "epoch": 3.2494658119658117, + "grad_norm": 0.9426729679107666, + "learning_rate": 1.6936731092053292e-05, + "loss": 0.9476, + "step": 18252 + }, + { + "epoch": 3.2494658119658117, + "eval_loss": 1.1752357482910156, + "eval_runtime": 35.242, + "eval_samples_per_second": 29.539, + "eval_steps_per_second": 14.784, + "step": 18252 + }, + { + "epoch": 3.2496438746438745, + "grad_norm": 0.9611350893974304, + "learning_rate": 1.6928937823001922e-05, + "loss": 0.7863, + "step": 18253 + }, + { + "epoch": 3.2498219373219372, + "grad_norm": 0.9414992928504944, + "learning_rate": 1.692114618155549e-05, + "loss": 0.8036, + "step": 18254 + }, + { + "epoch": 3.25, + "grad_norm": 0.9665080904960632, + "learning_rate": 1.6913356167866578e-05, + "loss": 0.8325, + "step": 18255 + }, + { + "epoch": 3.2501780626780628, + "grad_norm": 1.0661953687667847, + "learning_rate": 1.690556778208787e-05, + "loss": 0.9415, + "step": 18256 + }, + { + "epoch": 3.2503561253561255, + "grad_norm": 0.9129988551139832, + "learning_rate": 1.689778102437196e-05, + "loss": 0.9128, + "step": 18257 + }, + { + "epoch": 3.2505341880341883, + "grad_norm": 0.8086062073707581, + "learning_rate": 1.6889995894871392e-05, + "loss": 0.6195, + "step": 18258 + }, + { + "epoch": 3.2507122507122506, + "grad_norm": 0.9159565567970276, + "learning_rate": 1.6882212393738707e-05, + "loss": 0.6897, + "step": 18259 + }, + { + "epoch": 3.2508903133903133, + "grad_norm": 0.9264681339263916, + "learning_rate": 1.6874430521126417e-05, + "loss": 0.8267, + "step": 18260 + }, + { + "epoch": 3.251068376068376, + "grad_norm": 1.0706672668457031, + "learning_rate": 1.6866650277186967e-05, + "loss": 0.8832, + "step": 18261 + }, + { + "epoch": 3.251246438746439, + "grad_norm": 0.8599040508270264, + "learning_rate": 1.6858871662072827e-05, + "loss": 0.6803, + "step": 18262 + }, + { + "epoch": 3.2514245014245016, + "grad_norm": 0.9120079278945923, + "learning_rate": 1.6851094675936353e-05, + "loss": 0.6025, + "step": 18263 + }, + { + "epoch": 3.251602564102564, + "grad_norm": 0.7756575345993042, + "learning_rate": 1.6843319318929996e-05, + "loss": 0.5763, + "step": 18264 + }, + { + "epoch": 3.2517806267806266, + "grad_norm": 0.8564868569374084, + "learning_rate": 1.6835545591206014e-05, + "loss": 0.7513, + "step": 18265 + }, + { + "epoch": 3.2519586894586894, + "grad_norm": 0.9959290623664856, + "learning_rate": 1.6827773492916778e-05, + "loss": 0.7217, + "step": 18266 + }, + { + "epoch": 3.252136752136752, + "grad_norm": 0.8105573058128357, + "learning_rate": 1.6820003024214547e-05, + "loss": 0.6058, + "step": 18267 + }, + { + "epoch": 3.252314814814815, + "grad_norm": 0.8831157088279724, + "learning_rate": 1.681223418525154e-05, + "loss": 0.7648, + "step": 18268 + }, + { + "epoch": 3.2524928774928776, + "grad_norm": 0.932885468006134, + "learning_rate": 1.6804466976180046e-05, + "loss": 0.8404, + "step": 18269 + }, + { + "epoch": 3.2526709401709404, + "grad_norm": 0.9935491681098938, + "learning_rate": 1.6796701397152147e-05, + "loss": 0.7497, + "step": 18270 + }, + { + "epoch": 3.2528490028490027, + "grad_norm": 0.9661284685134888, + "learning_rate": 1.6788937448320084e-05, + "loss": 0.9749, + "step": 18271 + }, + { + "epoch": 3.2530270655270654, + "grad_norm": 0.8758362531661987, + "learning_rate": 1.6781175129835903e-05, + "loss": 0.84, + "step": 18272 + }, + { + "epoch": 3.253205128205128, + "grad_norm": 0.9549978375434875, + "learning_rate": 1.677341444185174e-05, + "loss": 0.9391, + "step": 18273 + }, + { + "epoch": 3.253383190883191, + "grad_norm": 0.9131140112876892, + "learning_rate": 1.676565538451963e-05, + "loss": 0.6248, + "step": 18274 + }, + { + "epoch": 3.2535612535612537, + "grad_norm": 0.9425809979438782, + "learning_rate": 1.675789795799161e-05, + "loss": 0.7493, + "step": 18275 + }, + { + "epoch": 3.253739316239316, + "grad_norm": 0.8880078792572021, + "learning_rate": 1.6750142162419647e-05, + "loss": 0.6994, + "step": 18276 + }, + { + "epoch": 3.2539173789173788, + "grad_norm": 0.9569603800773621, + "learning_rate": 1.674238799795572e-05, + "loss": 0.7584, + "step": 18277 + }, + { + "epoch": 3.2540954415954415, + "grad_norm": 0.9752117395401001, + "learning_rate": 1.6734635464751714e-05, + "loss": 0.9893, + "step": 18278 + }, + { + "epoch": 3.2542735042735043, + "grad_norm": 0.9450941681861877, + "learning_rate": 1.672688456295961e-05, + "loss": 0.709, + "step": 18279 + }, + { + "epoch": 3.254451566951567, + "grad_norm": 0.9441366791725159, + "learning_rate": 1.671913529273117e-05, + "loss": 0.7121, + "step": 18280 + }, + { + "epoch": 3.2546296296296298, + "grad_norm": 1.0751672983169556, + "learning_rate": 1.6711387654218323e-05, + "loss": 0.7037, + "step": 18281 + }, + { + "epoch": 3.2548076923076925, + "grad_norm": 0.8935137987136841, + "learning_rate": 1.6703641647572764e-05, + "loss": 0.6031, + "step": 18282 + }, + { + "epoch": 3.254985754985755, + "grad_norm": 0.8644391298294067, + "learning_rate": 1.6695897272946348e-05, + "loss": 0.7036, + "step": 18283 + }, + { + "epoch": 3.2551638176638176, + "grad_norm": 0.8370543718338013, + "learning_rate": 1.6688154530490773e-05, + "loss": 0.782, + "step": 18284 + }, + { + "epoch": 3.2553418803418803, + "grad_norm": 1.1523544788360596, + "learning_rate": 1.668041342035773e-05, + "loss": 1.0715, + "step": 18285 + }, + { + "epoch": 3.255519943019943, + "grad_norm": 1.0818554162979126, + "learning_rate": 1.6672673942698925e-05, + "loss": 0.8182, + "step": 18286 + }, + { + "epoch": 3.255698005698006, + "grad_norm": 0.9586027264595032, + "learning_rate": 1.666493609766596e-05, + "loss": 0.802, + "step": 18287 + }, + { + "epoch": 3.255876068376068, + "grad_norm": 0.9727693796157837, + "learning_rate": 1.6657199885410446e-05, + "loss": 0.7155, + "step": 18288 + }, + { + "epoch": 3.256054131054131, + "grad_norm": 1.0775129795074463, + "learning_rate": 1.664946530608399e-05, + "loss": 0.8433, + "step": 18289 + }, + { + "epoch": 3.2562321937321936, + "grad_norm": 0.9694477915763855, + "learning_rate": 1.6641732359838113e-05, + "loss": 0.791, + "step": 18290 + }, + { + "epoch": 3.2564102564102564, + "grad_norm": 0.9691938161849976, + "learning_rate": 1.6634001046824333e-05, + "loss": 0.8073, + "step": 18291 + }, + { + "epoch": 3.256588319088319, + "grad_norm": 1.036760926246643, + "learning_rate": 1.6626271367194123e-05, + "loss": 0.6048, + "step": 18292 + }, + { + "epoch": 3.256766381766382, + "grad_norm": 0.8390796184539795, + "learning_rate": 1.6618543321098912e-05, + "loss": 0.9326, + "step": 18293 + }, + { + "epoch": 3.2569444444444446, + "grad_norm": 0.927253007888794, + "learning_rate": 1.6610816908690184e-05, + "loss": 0.7526, + "step": 18294 + }, + { + "epoch": 3.257122507122507, + "grad_norm": 0.9034532308578491, + "learning_rate": 1.6603092130119214e-05, + "loss": 0.8068, + "step": 18295 + }, + { + "epoch": 3.2573005698005697, + "grad_norm": 0.9301304817199707, + "learning_rate": 1.6595368985537464e-05, + "loss": 0.7274, + "step": 18296 + }, + { + "epoch": 3.2574786324786325, + "grad_norm": 0.9493504166603088, + "learning_rate": 1.6587647475096157e-05, + "loss": 0.7109, + "step": 18297 + }, + { + "epoch": 3.257656695156695, + "grad_norm": 0.9562366008758545, + "learning_rate": 1.6579927598946644e-05, + "loss": 0.8517, + "step": 18298 + }, + { + "epoch": 3.257834757834758, + "grad_norm": 1.0997562408447266, + "learning_rate": 1.6572209357240155e-05, + "loss": 1.1135, + "step": 18299 + }, + { + "epoch": 3.2580128205128207, + "grad_norm": 1.040228009223938, + "learning_rate": 1.6564492750127925e-05, + "loss": 0.83, + "step": 18300 + }, + { + "epoch": 3.258190883190883, + "grad_norm": 0.9394234418869019, + "learning_rate": 1.6556777777761133e-05, + "loss": 0.6691, + "step": 18301 + }, + { + "epoch": 3.2583689458689458, + "grad_norm": 0.806209146976471, + "learning_rate": 1.654906444029095e-05, + "loss": 0.6905, + "step": 18302 + }, + { + "epoch": 3.2585470085470085, + "grad_norm": 1.2015374898910522, + "learning_rate": 1.6541352737868465e-05, + "loss": 0.7325, + "step": 18303 + }, + { + "epoch": 3.2587250712250713, + "grad_norm": 0.787306547164917, + "learning_rate": 1.6533642670644843e-05, + "loss": 0.6128, + "step": 18304 + }, + { + "epoch": 3.258903133903134, + "grad_norm": 1.0318105220794678, + "learning_rate": 1.652593423877107e-05, + "loss": 0.8617, + "step": 18305 + }, + { + "epoch": 3.2590811965811968, + "grad_norm": 1.0233045816421509, + "learning_rate": 1.651822744239826e-05, + "loss": 1.0166, + "step": 18306 + }, + { + "epoch": 3.259259259259259, + "grad_norm": 0.9209055304527283, + "learning_rate": 1.651052228167731e-05, + "loss": 0.8128, + "step": 18307 + }, + { + "epoch": 3.259437321937322, + "grad_norm": 1.0221130847930908, + "learning_rate": 1.6502818756759276e-05, + "loss": 0.8015, + "step": 18308 + }, + { + "epoch": 3.2596153846153846, + "grad_norm": 0.998290479183197, + "learning_rate": 1.6495116867795047e-05, + "loss": 0.752, + "step": 18309 + }, + { + "epoch": 3.2597934472934473, + "grad_norm": 0.9447212815284729, + "learning_rate": 1.6487416614935513e-05, + "loss": 0.6362, + "step": 18310 + }, + { + "epoch": 3.25997150997151, + "grad_norm": 1.086247205734253, + "learning_rate": 1.6479717998331623e-05, + "loss": 0.7776, + "step": 18311 + }, + { + "epoch": 3.260149572649573, + "grad_norm": 0.7751702070236206, + "learning_rate": 1.647202101813411e-05, + "loss": 0.4099, + "step": 18312 + }, + { + "epoch": 3.260327635327635, + "grad_norm": 0.8890853524208069, + "learning_rate": 1.6464325674493853e-05, + "loss": 0.668, + "step": 18313 + }, + { + "epoch": 3.260505698005698, + "grad_norm": 0.9836373925209045, + "learning_rate": 1.6456631967561608e-05, + "loss": 0.7582, + "step": 18314 + }, + { + "epoch": 3.2606837606837606, + "grad_norm": 0.9572944641113281, + "learning_rate": 1.6448939897488112e-05, + "loss": 0.7865, + "step": 18315 + }, + { + "epoch": 3.2608618233618234, + "grad_norm": 1.0071512460708618, + "learning_rate": 1.644124946442407e-05, + "loss": 0.7007, + "step": 18316 + }, + { + "epoch": 3.261039886039886, + "grad_norm": 0.9263545870780945, + "learning_rate": 1.6433560668520176e-05, + "loss": 0.8008, + "step": 18317 + }, + { + "epoch": 3.261217948717949, + "grad_norm": 0.9851424694061279, + "learning_rate": 1.642587350992706e-05, + "loss": 0.8027, + "step": 18318 + }, + { + "epoch": 3.261396011396011, + "grad_norm": 0.8615550398826599, + "learning_rate": 1.6418187988795342e-05, + "loss": 0.7405, + "step": 18319 + }, + { + "epoch": 3.261574074074074, + "grad_norm": 0.9881899356842041, + "learning_rate": 1.6410504105275593e-05, + "loss": 1.0046, + "step": 18320 + }, + { + "epoch": 3.2617521367521367, + "grad_norm": 0.9273366332054138, + "learning_rate": 1.64028218595184e-05, + "loss": 0.718, + "step": 18321 + }, + { + "epoch": 3.2619301994301995, + "grad_norm": 0.9626027941703796, + "learning_rate": 1.6395141251674228e-05, + "loss": 0.8123, + "step": 18322 + }, + { + "epoch": 3.262108262108262, + "grad_norm": 1.0157630443572998, + "learning_rate": 1.6387462281893596e-05, + "loss": 0.8498, + "step": 18323 + }, + { + "epoch": 3.262286324786325, + "grad_norm": 0.8313695192337036, + "learning_rate": 1.637978495032696e-05, + "loss": 0.6479, + "step": 18324 + }, + { + "epoch": 3.2624643874643873, + "grad_norm": 0.8598573207855225, + "learning_rate": 1.6372109257124735e-05, + "loss": 0.5883, + "step": 18325 + }, + { + "epoch": 3.26264245014245, + "grad_norm": 1.0443403720855713, + "learning_rate": 1.63644352024373e-05, + "loss": 0.6393, + "step": 18326 + }, + { + "epoch": 3.2628205128205128, + "grad_norm": 1.0185366868972778, + "learning_rate": 1.6356762786415026e-05, + "loss": 0.8473, + "step": 18327 + }, + { + "epoch": 3.2629985754985755, + "grad_norm": 1.0581954717636108, + "learning_rate": 1.63490920092082e-05, + "loss": 0.6592, + "step": 18328 + }, + { + "epoch": 3.2631766381766383, + "grad_norm": 0.8719905614852905, + "learning_rate": 1.634142287096717e-05, + "loss": 0.6829, + "step": 18329 + }, + { + "epoch": 3.263354700854701, + "grad_norm": 1.0849109888076782, + "learning_rate": 1.6333755371842175e-05, + "loss": 0.5595, + "step": 18330 + }, + { + "epoch": 3.263532763532764, + "grad_norm": 0.9903097748756409, + "learning_rate": 1.6326089511983443e-05, + "loss": 0.8984, + "step": 18331 + }, + { + "epoch": 3.263710826210826, + "grad_norm": 0.8619652390480042, + "learning_rate": 1.6318425291541173e-05, + "loss": 0.5457, + "step": 18332 + }, + { + "epoch": 3.263888888888889, + "grad_norm": 0.8803672194480896, + "learning_rate": 1.6310762710665528e-05, + "loss": 0.7953, + "step": 18333 + }, + { + "epoch": 3.2640669515669516, + "grad_norm": 1.0389653444290161, + "learning_rate": 1.6303101769506633e-05, + "loss": 0.8362, + "step": 18334 + }, + { + "epoch": 3.2642450142450143, + "grad_norm": 0.9297505021095276, + "learning_rate": 1.629544246821456e-05, + "loss": 0.6762, + "step": 18335 + }, + { + "epoch": 3.264423076923077, + "grad_norm": 1.0184338092803955, + "learning_rate": 1.6287784806939476e-05, + "loss": 0.7657, + "step": 18336 + }, + { + "epoch": 3.2646011396011394, + "grad_norm": 1.0914297103881836, + "learning_rate": 1.6280128785831293e-05, + "loss": 0.8824, + "step": 18337 + }, + { + "epoch": 3.264779202279202, + "grad_norm": 0.9033359885215759, + "learning_rate": 1.6272474405040106e-05, + "loss": 0.4726, + "step": 18338 + }, + { + "epoch": 3.264957264957265, + "grad_norm": 0.8340174555778503, + "learning_rate": 1.6264821664715846e-05, + "loss": 0.5618, + "step": 18339 + }, + { + "epoch": 3.2651353276353277, + "grad_norm": 0.7262910604476929, + "learning_rate": 1.625717056500846e-05, + "loss": 0.5631, + "step": 18340 + }, + { + "epoch": 3.2653133903133904, + "grad_norm": 0.8162511587142944, + "learning_rate": 1.6249521106067866e-05, + "loss": 0.7289, + "step": 18341 + }, + { + "epoch": 3.265491452991453, + "grad_norm": 0.9709984064102173, + "learning_rate": 1.624187328804392e-05, + "loss": 0.8121, + "step": 18342 + }, + { + "epoch": 3.265669515669516, + "grad_norm": 0.97455233335495, + "learning_rate": 1.6234227111086475e-05, + "loss": 0.692, + "step": 18343 + }, + { + "epoch": 3.265847578347578, + "grad_norm": 0.9413778781890869, + "learning_rate": 1.622658257534535e-05, + "loss": 0.5593, + "step": 18344 + }, + { + "epoch": 3.266025641025641, + "grad_norm": 0.8591381907463074, + "learning_rate": 1.621893968097028e-05, + "loss": 0.7983, + "step": 18345 + }, + { + "epoch": 3.2662037037037037, + "grad_norm": 1.017133116722107, + "learning_rate": 1.62112984281111e-05, + "loss": 0.8291, + "step": 18346 + }, + { + "epoch": 3.2663817663817665, + "grad_norm": 0.8174110651016235, + "learning_rate": 1.6203658816917423e-05, + "loss": 0.6206, + "step": 18347 + }, + { + "epoch": 3.2665598290598292, + "grad_norm": 0.9929126501083374, + "learning_rate": 1.6196020847539006e-05, + "loss": 0.7346, + "step": 18348 + }, + { + "epoch": 3.2667378917378915, + "grad_norm": 0.9130271077156067, + "learning_rate": 1.6188384520125476e-05, + "loss": 0.9829, + "step": 18349 + }, + { + "epoch": 3.2669159544159543, + "grad_norm": 0.9933291673660278, + "learning_rate": 1.6180749834826413e-05, + "loss": 0.7314, + "step": 18350 + }, + { + "epoch": 3.267094017094017, + "grad_norm": 0.9695504307746887, + "learning_rate": 1.6173116791791498e-05, + "loss": 0.9824, + "step": 18351 + }, + { + "epoch": 3.26727207977208, + "grad_norm": 0.9526118636131287, + "learning_rate": 1.6165485391170164e-05, + "loss": 1.0207, + "step": 18352 + }, + { + "epoch": 3.2674501424501425, + "grad_norm": 0.8514195084571838, + "learning_rate": 1.6157855633112053e-05, + "loss": 0.5906, + "step": 18353 + }, + { + "epoch": 3.2676282051282053, + "grad_norm": 0.9104008674621582, + "learning_rate": 1.6150227517766535e-05, + "loss": 0.7426, + "step": 18354 + }, + { + "epoch": 3.267806267806268, + "grad_norm": 1.019457221031189, + "learning_rate": 1.6142601045283157e-05, + "loss": 0.6441, + "step": 18355 + }, + { + "epoch": 3.2679843304843303, + "grad_norm": 0.8585529327392578, + "learning_rate": 1.6134976215811305e-05, + "loss": 0.5605, + "step": 18356 + }, + { + "epoch": 3.268162393162393, + "grad_norm": 0.8874415755271912, + "learning_rate": 1.612735302950038e-05, + "loss": 0.6461, + "step": 18357 + }, + { + "epoch": 3.268340455840456, + "grad_norm": 0.9070242047309875, + "learning_rate": 1.6119731486499755e-05, + "loss": 0.7123, + "step": 18358 + }, + { + "epoch": 3.2685185185185186, + "grad_norm": 0.8923984169960022, + "learning_rate": 1.6112111586958733e-05, + "loss": 0.6977, + "step": 18359 + }, + { + "epoch": 3.2686965811965814, + "grad_norm": 1.0075733661651611, + "learning_rate": 1.6104493331026604e-05, + "loss": 0.9057, + "step": 18360 + }, + { + "epoch": 3.2688746438746437, + "grad_norm": 0.7989169359207153, + "learning_rate": 1.6096876718852694e-05, + "loss": 0.6616, + "step": 18361 + }, + { + "epoch": 3.2690527065527064, + "grad_norm": 1.0517085790634155, + "learning_rate": 1.608926175058615e-05, + "loss": 0.5575, + "step": 18362 + }, + { + "epoch": 3.269230769230769, + "grad_norm": 0.9944202303886414, + "learning_rate": 1.608164842637623e-05, + "loss": 0.7801, + "step": 18363 + }, + { + "epoch": 3.269408831908832, + "grad_norm": 1.3099443912506104, + "learning_rate": 1.607403674637209e-05, + "loss": 0.8176, + "step": 18364 + }, + { + "epoch": 3.2695868945868947, + "grad_norm": 0.918102502822876, + "learning_rate": 1.606642671072285e-05, + "loss": 0.8969, + "step": 18365 + }, + { + "epoch": 3.2697649572649574, + "grad_norm": 1.058408260345459, + "learning_rate": 1.605881831957763e-05, + "loss": 0.7686, + "step": 18366 + }, + { + "epoch": 3.26994301994302, + "grad_norm": 0.956381618976593, + "learning_rate": 1.6051211573085467e-05, + "loss": 0.8574, + "step": 18367 + }, + { + "epoch": 3.2701210826210825, + "grad_norm": 1.0281140804290771, + "learning_rate": 1.6043606471395468e-05, + "loss": 0.9321, + "step": 18368 + }, + { + "epoch": 3.2702991452991452, + "grad_norm": 0.9082704782485962, + "learning_rate": 1.6036003014656552e-05, + "loss": 0.83, + "step": 18369 + }, + { + "epoch": 3.270477207977208, + "grad_norm": 0.9393236637115479, + "learning_rate": 1.602840120301776e-05, + "loss": 0.6193, + "step": 18370 + }, + { + "epoch": 3.2706552706552707, + "grad_norm": 0.8955071568489075, + "learning_rate": 1.602080103662801e-05, + "loss": 0.708, + "step": 18371 + }, + { + "epoch": 3.2708333333333335, + "grad_norm": 1.0704554319381714, + "learning_rate": 1.6013202515636206e-05, + "loss": 0.8384, + "step": 18372 + }, + { + "epoch": 3.271011396011396, + "grad_norm": 0.8835996389389038, + "learning_rate": 1.6005605640191247e-05, + "loss": 0.7344, + "step": 18373 + }, + { + "epoch": 3.2711894586894585, + "grad_norm": 1.0095866918563843, + "learning_rate": 1.5998010410441943e-05, + "loss": 0.9588, + "step": 18374 + }, + { + "epoch": 3.2713675213675213, + "grad_norm": 0.9835611581802368, + "learning_rate": 1.5990416826537114e-05, + "loss": 0.8655, + "step": 18375 + }, + { + "epoch": 3.271545584045584, + "grad_norm": 1.0003584623336792, + "learning_rate": 1.5982824888625603e-05, + "loss": 0.7716, + "step": 18376 + }, + { + "epoch": 3.271723646723647, + "grad_norm": 0.9901667833328247, + "learning_rate": 1.597523459685605e-05, + "loss": 0.9206, + "step": 18377 + }, + { + "epoch": 3.2719017094017095, + "grad_norm": 0.8539385199546814, + "learning_rate": 1.5967645951377285e-05, + "loss": 0.8601, + "step": 18378 + }, + { + "epoch": 3.2720797720797723, + "grad_norm": 1.017080307006836, + "learning_rate": 1.5960058952337887e-05, + "loss": 0.8568, + "step": 18379 + }, + { + "epoch": 3.2722578347578346, + "grad_norm": 0.9467936158180237, + "learning_rate": 1.5952473599886575e-05, + "loss": 0.8216, + "step": 18380 + }, + { + "epoch": 3.2724358974358974, + "grad_norm": 0.8549704551696777, + "learning_rate": 1.5944889894171944e-05, + "loss": 0.5828, + "step": 18381 + }, + { + "epoch": 3.27261396011396, + "grad_norm": 0.9717326760292053, + "learning_rate": 1.5937307835342595e-05, + "loss": 0.814, + "step": 18382 + }, + { + "epoch": 3.272792022792023, + "grad_norm": 1.0630271434783936, + "learning_rate": 1.592972742354707e-05, + "loss": 0.8204, + "step": 18383 + }, + { + "epoch": 3.2729700854700856, + "grad_norm": 0.8099556565284729, + "learning_rate": 1.59221486589339e-05, + "loss": 0.6395, + "step": 18384 + }, + { + "epoch": 3.273148148148148, + "grad_norm": 0.8640239834785461, + "learning_rate": 1.5914571541651545e-05, + "loss": 0.6824, + "step": 18385 + }, + { + "epoch": 3.2733262108262107, + "grad_norm": 0.9581767916679382, + "learning_rate": 1.5906996071848534e-05, + "loss": 0.7363, + "step": 18386 + }, + { + "epoch": 3.2735042735042734, + "grad_norm": 0.9858554601669312, + "learning_rate": 1.5899422249673203e-05, + "loss": 0.7251, + "step": 18387 + }, + { + "epoch": 3.273682336182336, + "grad_norm": 1.040452003479004, + "learning_rate": 1.5891850075274005e-05, + "loss": 0.9428, + "step": 18388 + }, + { + "epoch": 3.273860398860399, + "grad_norm": 1.145668387413025, + "learning_rate": 1.5884279548799296e-05, + "loss": 0.8521, + "step": 18389 + }, + { + "epoch": 3.2740384615384617, + "grad_norm": 0.9758059978485107, + "learning_rate": 1.5876710670397386e-05, + "loss": 0.8099, + "step": 18390 + }, + { + "epoch": 3.2742165242165244, + "grad_norm": 0.9736207723617554, + "learning_rate": 1.5869143440216583e-05, + "loss": 0.8343, + "step": 18391 + }, + { + "epoch": 3.2743945868945867, + "grad_norm": 1.058436393737793, + "learning_rate": 1.5861577858405118e-05, + "loss": 0.8069, + "step": 18392 + }, + { + "epoch": 3.2745726495726495, + "grad_norm": 0.9382075071334839, + "learning_rate": 1.585401392511131e-05, + "loss": 0.893, + "step": 18393 + }, + { + "epoch": 3.2747507122507122, + "grad_norm": 0.9606096148490906, + "learning_rate": 1.5846451640483252e-05, + "loss": 0.609, + "step": 18394 + }, + { + "epoch": 3.274928774928775, + "grad_norm": 0.8548282384872437, + "learning_rate": 1.5838891004669175e-05, + "loss": 0.8531, + "step": 18395 + }, + { + "epoch": 3.2751068376068377, + "grad_norm": 0.8669559955596924, + "learning_rate": 1.5831332017817202e-05, + "loss": 0.5267, + "step": 18396 + }, + { + "epoch": 3.2752849002849, + "grad_norm": 1.0153757333755493, + "learning_rate": 1.5823774680075443e-05, + "loss": 0.9531, + "step": 18397 + }, + { + "epoch": 3.275462962962963, + "grad_norm": 0.9505778551101685, + "learning_rate": 1.5816218991591946e-05, + "loss": 0.6559, + "step": 18398 + }, + { + "epoch": 3.2756410256410255, + "grad_norm": 0.785952091217041, + "learning_rate": 1.580866495251476e-05, + "loss": 0.5311, + "step": 18399 + }, + { + "epoch": 3.2758190883190883, + "grad_norm": 0.9959350824356079, + "learning_rate": 1.5801112562991883e-05, + "loss": 0.9929, + "step": 18400 + }, + { + "epoch": 3.275997150997151, + "grad_norm": 0.9083783626556396, + "learning_rate": 1.5793561823171332e-05, + "loss": 0.6089, + "step": 18401 + }, + { + "epoch": 3.276175213675214, + "grad_norm": 0.94951331615448, + "learning_rate": 1.5786012733200973e-05, + "loss": 0.728, + "step": 18402 + }, + { + "epoch": 3.2763532763532766, + "grad_norm": 0.8742989897727966, + "learning_rate": 1.5778465293228795e-05, + "loss": 0.8664, + "step": 18403 + }, + { + "epoch": 3.276531339031339, + "grad_norm": 0.9766057133674622, + "learning_rate": 1.5770919503402594e-05, + "loss": 0.6791, + "step": 18404 + }, + { + "epoch": 3.2767094017094016, + "grad_norm": 0.9813364744186401, + "learning_rate": 1.576337536387027e-05, + "loss": 0.8344, + "step": 18405 + }, + { + "epoch": 3.2768874643874644, + "grad_norm": 0.8530852794647217, + "learning_rate": 1.575583287477963e-05, + "loss": 0.9362, + "step": 18406 + }, + { + "epoch": 3.277065527065527, + "grad_norm": 1.0678436756134033, + "learning_rate": 1.5748292036278412e-05, + "loss": 0.655, + "step": 18407 + }, + { + "epoch": 3.27724358974359, + "grad_norm": 0.9426410794258118, + "learning_rate": 1.574075284851444e-05, + "loss": 0.7442, + "step": 18408 + }, + { + "epoch": 3.277421652421652, + "grad_norm": 1.096710205078125, + "learning_rate": 1.5733215311635342e-05, + "loss": 0.8746, + "step": 18409 + }, + { + "epoch": 3.277599715099715, + "grad_norm": 1.0035667419433594, + "learning_rate": 1.5725679425788853e-05, + "loss": 0.8182, + "step": 18410 + }, + { + "epoch": 3.2777777777777777, + "grad_norm": 0.6038252115249634, + "learning_rate": 1.5718145191122625e-05, + "loss": 0.3724, + "step": 18411 + }, + { + "epoch": 3.2779558404558404, + "grad_norm": 1.0738803148269653, + "learning_rate": 1.5710612607784247e-05, + "loss": 0.7623, + "step": 18412 + }, + { + "epoch": 3.278133903133903, + "grad_norm": 0.9418566226959229, + "learning_rate": 1.5703081675921328e-05, + "loss": 0.6925, + "step": 18413 + }, + { + "epoch": 3.278311965811966, + "grad_norm": 0.935024619102478, + "learning_rate": 1.5695552395681414e-05, + "loss": 0.6902, + "step": 18414 + }, + { + "epoch": 3.2784900284900287, + "grad_norm": 0.9279661774635315, + "learning_rate": 1.5688024767212017e-05, + "loss": 0.6978, + "step": 18415 + }, + { + "epoch": 3.278668091168091, + "grad_norm": 1.1668767929077148, + "learning_rate": 1.5680498790660636e-05, + "loss": 0.8869, + "step": 18416 + }, + { + "epoch": 3.2788461538461537, + "grad_norm": 0.9925910830497742, + "learning_rate": 1.5672974466174695e-05, + "loss": 0.9774, + "step": 18417 + }, + { + "epoch": 3.2790242165242165, + "grad_norm": 1.0041983127593994, + "learning_rate": 1.5665451793901698e-05, + "loss": 0.8268, + "step": 18418 + }, + { + "epoch": 3.2792022792022792, + "grad_norm": 0.7737284302711487, + "learning_rate": 1.565793077398894e-05, + "loss": 0.6039, + "step": 18419 + }, + { + "epoch": 3.279380341880342, + "grad_norm": 0.9920578598976135, + "learning_rate": 1.5650411406583833e-05, + "loss": 0.7539, + "step": 18420 + }, + { + "epoch": 3.2795584045584047, + "grad_norm": 1.04171621799469, + "learning_rate": 1.5642893691833706e-05, + "loss": 0.6719, + "step": 18421 + }, + { + "epoch": 3.279736467236467, + "grad_norm": 0.7911034226417542, + "learning_rate": 1.563537762988583e-05, + "loss": 0.5976, + "step": 18422 + }, + { + "epoch": 3.27991452991453, + "grad_norm": 1.0236256122589111, + "learning_rate": 1.5627863220887485e-05, + "loss": 0.8767, + "step": 18423 + }, + { + "epoch": 3.2800925925925926, + "grad_norm": 1.0210331678390503, + "learning_rate": 1.562035046498589e-05, + "loss": 0.6193, + "step": 18424 + }, + { + "epoch": 3.2802706552706553, + "grad_norm": 1.0835071802139282, + "learning_rate": 1.5612839362328234e-05, + "loss": 0.9845, + "step": 18425 + }, + { + "epoch": 3.280448717948718, + "grad_norm": 0.9284858703613281, + "learning_rate": 1.5605329913061707e-05, + "loss": 0.7991, + "step": 18426 + }, + { + "epoch": 3.280626780626781, + "grad_norm": 1.1475468873977661, + "learning_rate": 1.5597822117333394e-05, + "loss": 0.695, + "step": 18427 + }, + { + "epoch": 3.280804843304843, + "grad_norm": 0.9122242331504822, + "learning_rate": 1.559031597529045e-05, + "loss": 0.7703, + "step": 18428 + }, + { + "epoch": 3.280982905982906, + "grad_norm": 0.9514288902282715, + "learning_rate": 1.558281148707993e-05, + "loss": 0.6174, + "step": 18429 + }, + { + "epoch": 3.2811609686609686, + "grad_norm": 1.0076818466186523, + "learning_rate": 1.557530865284884e-05, + "loss": 0.8294, + "step": 18430 + }, + { + "epoch": 3.2813390313390314, + "grad_norm": 0.8275960087776184, + "learning_rate": 1.5567807472744212e-05, + "loss": 0.672, + "step": 18431 + }, + { + "epoch": 3.281517094017094, + "grad_norm": 0.9140580296516418, + "learning_rate": 1.556030794691298e-05, + "loss": 0.7372, + "step": 18432 + }, + { + "epoch": 3.281695156695157, + "grad_norm": 1.1289745569229126, + "learning_rate": 1.5552810075502144e-05, + "loss": 0.8771, + "step": 18433 + }, + { + "epoch": 3.281873219373219, + "grad_norm": 1.1868027448654175, + "learning_rate": 1.554531385865853e-05, + "loss": 0.9765, + "step": 18434 + }, + { + "epoch": 3.282051282051282, + "grad_norm": 0.8656640648841858, + "learning_rate": 1.5537819296529065e-05, + "loss": 0.6889, + "step": 18435 + }, + { + "epoch": 3.2822293447293447, + "grad_norm": 0.9567883014678955, + "learning_rate": 1.5530326389260585e-05, + "loss": 0.5665, + "step": 18436 + }, + { + "epoch": 3.2824074074074074, + "grad_norm": 1.0118212699890137, + "learning_rate": 1.552283513699989e-05, + "loss": 0.6376, + "step": 18437 + }, + { + "epoch": 3.28258547008547, + "grad_norm": 0.9064817428588867, + "learning_rate": 1.5515345539893756e-05, + "loss": 0.7624, + "step": 18438 + }, + { + "epoch": 3.282763532763533, + "grad_norm": 0.9137055277824402, + "learning_rate": 1.5507857598088914e-05, + "loss": 0.6229, + "step": 18439 + }, + { + "epoch": 3.2829415954415953, + "grad_norm": 0.9002276062965393, + "learning_rate": 1.5500371311732098e-05, + "loss": 0.7126, + "step": 18440 + }, + { + "epoch": 3.283119658119658, + "grad_norm": 1.0946577787399292, + "learning_rate": 1.5492886680969963e-05, + "loss": 0.7333, + "step": 18441 + }, + { + "epoch": 3.2832977207977208, + "grad_norm": 1.0645185708999634, + "learning_rate": 1.548540370594914e-05, + "loss": 0.7965, + "step": 18442 + }, + { + "epoch": 3.2834757834757835, + "grad_norm": 0.9328752160072327, + "learning_rate": 1.5477922386816325e-05, + "loss": 0.9929, + "step": 18443 + }, + { + "epoch": 3.2836538461538463, + "grad_norm": 0.8564518094062805, + "learning_rate": 1.5470442723717993e-05, + "loss": 0.5875, + "step": 18444 + }, + { + "epoch": 3.283831908831909, + "grad_norm": 0.92264324426651, + "learning_rate": 1.5462964716800753e-05, + "loss": 0.8774, + "step": 18445 + }, + { + "epoch": 3.2840099715099713, + "grad_norm": 0.8862282037734985, + "learning_rate": 1.5455488366211126e-05, + "loss": 0.6523, + "step": 18446 + }, + { + "epoch": 3.284188034188034, + "grad_norm": 1.1124815940856934, + "learning_rate": 1.5448013672095542e-05, + "loss": 0.8572, + "step": 18447 + }, + { + "epoch": 3.284366096866097, + "grad_norm": 0.96551513671875, + "learning_rate": 1.5440540634600542e-05, + "loss": 0.8255, + "step": 18448 + }, + { + "epoch": 3.2845441595441596, + "grad_norm": 1.0143595933914185, + "learning_rate": 1.5433069253872435e-05, + "loss": 0.5876, + "step": 18449 + }, + { + "epoch": 3.2847222222222223, + "grad_norm": 0.9641425013542175, + "learning_rate": 1.5425599530057722e-05, + "loss": 0.7874, + "step": 18450 + }, + { + "epoch": 3.284900284900285, + "grad_norm": 0.982153058052063, + "learning_rate": 1.5418131463302643e-05, + "loss": 0.6765, + "step": 18451 + }, + { + "epoch": 3.285078347578348, + "grad_norm": 0.9491843581199646, + "learning_rate": 1.5410665053753604e-05, + "loss": 0.8588, + "step": 18452 + }, + { + "epoch": 3.28525641025641, + "grad_norm": 0.9593594074249268, + "learning_rate": 1.5403200301556863e-05, + "loss": 0.8069, + "step": 18453 + }, + { + "epoch": 3.285434472934473, + "grad_norm": 1.0291471481323242, + "learning_rate": 1.5395737206858673e-05, + "loss": 0.8385, + "step": 18454 + }, + { + "epoch": 3.2856125356125356, + "grad_norm": 0.8283519148826599, + "learning_rate": 1.5388275769805272e-05, + "loss": 0.6765, + "step": 18455 + }, + { + "epoch": 3.2857905982905984, + "grad_norm": 1.0748952627182007, + "learning_rate": 1.538081599054283e-05, + "loss": 0.8704, + "step": 18456 + }, + { + "epoch": 3.285968660968661, + "grad_norm": 0.8589741587638855, + "learning_rate": 1.53733578692175e-05, + "loss": 0.6258, + "step": 18457 + }, + { + "epoch": 3.2861467236467234, + "grad_norm": 1.0470752716064453, + "learning_rate": 1.536590140597547e-05, + "loss": 0.7413, + "step": 18458 + }, + { + "epoch": 3.286324786324786, + "grad_norm": 0.9156502485275269, + "learning_rate": 1.535844660096274e-05, + "loss": 0.713, + "step": 18459 + }, + { + "epoch": 3.286502849002849, + "grad_norm": 0.8092247843742371, + "learning_rate": 1.535099345432548e-05, + "loss": 0.5603, + "step": 18460 + }, + { + "epoch": 3.2866809116809117, + "grad_norm": 0.9388442039489746, + "learning_rate": 1.5343541966209607e-05, + "loss": 0.745, + "step": 18461 + }, + { + "epoch": 3.2868589743589745, + "grad_norm": 0.9431697726249695, + "learning_rate": 1.5336092136761193e-05, + "loss": 0.7621, + "step": 18462 + }, + { + "epoch": 3.287037037037037, + "grad_norm": 0.8809608221054077, + "learning_rate": 1.5328643966126178e-05, + "loss": 0.6982, + "step": 18463 + }, + { + "epoch": 3.2872150997151, + "grad_norm": 1.0238356590270996, + "learning_rate": 1.5321197454450496e-05, + "loss": 0.7742, + "step": 18464 + }, + { + "epoch": 3.2873931623931623, + "grad_norm": 1.099500298500061, + "learning_rate": 1.5313752601880038e-05, + "loss": 0.871, + "step": 18465 + }, + { + "epoch": 3.287571225071225, + "grad_norm": 0.8493121862411499, + "learning_rate": 1.530630940856067e-05, + "loss": 0.6869, + "step": 18466 + }, + { + "epoch": 3.2877492877492878, + "grad_norm": 0.9258021116256714, + "learning_rate": 1.5298867874638246e-05, + "loss": 0.7975, + "step": 18467 + }, + { + "epoch": 3.2879273504273505, + "grad_norm": 1.0687447786331177, + "learning_rate": 1.5291428000258556e-05, + "loss": 0.797, + "step": 18468 + }, + { + "epoch": 3.2881054131054133, + "grad_norm": 1.0039074420928955, + "learning_rate": 1.5283989785567366e-05, + "loss": 0.6995, + "step": 18469 + }, + { + "epoch": 3.2882834757834756, + "grad_norm": 0.9193965196609497, + "learning_rate": 1.5276553230710423e-05, + "loss": 0.6836, + "step": 18470 + }, + { + "epoch": 3.2884615384615383, + "grad_norm": 1.0001294612884521, + "learning_rate": 1.526911833583341e-05, + "loss": 0.7799, + "step": 18471 + }, + { + "epoch": 3.288639601139601, + "grad_norm": 0.9037470817565918, + "learning_rate": 1.526168510108199e-05, + "loss": 0.836, + "step": 18472 + }, + { + "epoch": 3.288817663817664, + "grad_norm": 0.8541157245635986, + "learning_rate": 1.525425352660187e-05, + "loss": 0.7288, + "step": 18473 + }, + { + "epoch": 3.2889957264957266, + "grad_norm": 1.0287656784057617, + "learning_rate": 1.5246823612538563e-05, + "loss": 0.8542, + "step": 18474 + }, + { + "epoch": 3.2891737891737893, + "grad_norm": 1.1029131412506104, + "learning_rate": 1.5239395359037734e-05, + "loss": 0.7211, + "step": 18475 + }, + { + "epoch": 3.289351851851852, + "grad_norm": 0.8748056292533875, + "learning_rate": 1.523196876624483e-05, + "loss": 0.7278, + "step": 18476 + }, + { + "epoch": 3.2895299145299144, + "grad_norm": 0.9509679675102234, + "learning_rate": 1.5224543834305427e-05, + "loss": 0.649, + "step": 18477 + }, + { + "epoch": 3.289707977207977, + "grad_norm": 0.9093441367149353, + "learning_rate": 1.5217120563364985e-05, + "loss": 0.822, + "step": 18478 + }, + { + "epoch": 3.28988603988604, + "grad_norm": 0.9471918940544128, + "learning_rate": 1.5209698953568952e-05, + "loss": 0.8744, + "step": 18479 + }, + { + "epoch": 3.2900641025641026, + "grad_norm": 0.8799687623977661, + "learning_rate": 1.5202279005062725e-05, + "loss": 0.651, + "step": 18480 + }, + { + "epoch": 3.2902421652421654, + "grad_norm": 0.9291742444038391, + "learning_rate": 1.5194860717991687e-05, + "loss": 0.6697, + "step": 18481 + }, + { + "epoch": 3.2904202279202277, + "grad_norm": 0.9564397931098938, + "learning_rate": 1.5187444092501158e-05, + "loss": 0.8578, + "step": 18482 + }, + { + "epoch": 3.2905982905982905, + "grad_norm": 0.9487985968589783, + "learning_rate": 1.518002912873654e-05, + "loss": 0.727, + "step": 18483 + }, + { + "epoch": 3.290776353276353, + "grad_norm": 1.316943645477295, + "learning_rate": 1.5172615826842996e-05, + "loss": 0.7059, + "step": 18484 + }, + { + "epoch": 3.290954415954416, + "grad_norm": 0.9622043371200562, + "learning_rate": 1.5165204186965875e-05, + "loss": 0.5788, + "step": 18485 + }, + { + "epoch": 3.2911324786324787, + "grad_norm": 0.7761589288711548, + "learning_rate": 1.5157794209250309e-05, + "loss": 0.4303, + "step": 18486 + }, + { + "epoch": 3.2913105413105415, + "grad_norm": 1.0018688440322876, + "learning_rate": 1.5150385893841545e-05, + "loss": 0.5912, + "step": 18487 + }, + { + "epoch": 3.291488603988604, + "grad_norm": 0.9771264791488647, + "learning_rate": 1.5142979240884713e-05, + "loss": 0.805, + "step": 18488 + }, + { + "epoch": 3.2916666666666665, + "grad_norm": 0.9868426322937012, + "learning_rate": 1.5135574250524897e-05, + "loss": 0.7762, + "step": 18489 + }, + { + "epoch": 3.2918447293447293, + "grad_norm": 1.0445677042007446, + "learning_rate": 1.5128170922907258e-05, + "loss": 0.8809, + "step": 18490 + }, + { + "epoch": 3.292022792022792, + "grad_norm": 1.082547664642334, + "learning_rate": 1.5120769258176758e-05, + "loss": 0.9871, + "step": 18491 + }, + { + "epoch": 3.2922008547008548, + "grad_norm": 0.9163486361503601, + "learning_rate": 1.5113369256478493e-05, + "loss": 0.6729, + "step": 18492 + }, + { + "epoch": 3.2923789173789175, + "grad_norm": 0.9486492872238159, + "learning_rate": 1.5105970917957413e-05, + "loss": 0.6898, + "step": 18493 + }, + { + "epoch": 3.29255698005698, + "grad_norm": 0.9150809645652771, + "learning_rate": 1.509857424275849e-05, + "loss": 0.8039, + "step": 18494 + }, + { + "epoch": 3.2927350427350426, + "grad_norm": 0.9275760650634766, + "learning_rate": 1.509117923102663e-05, + "loss": 0.8944, + "step": 18495 + }, + { + "epoch": 3.2929131054131053, + "grad_norm": 1.0610506534576416, + "learning_rate": 1.508378588290673e-05, + "loss": 0.6999, + "step": 18496 + }, + { + "epoch": 3.293091168091168, + "grad_norm": 0.9436303377151489, + "learning_rate": 1.5076394198543653e-05, + "loss": 0.8837, + "step": 18497 + }, + { + "epoch": 3.293269230769231, + "grad_norm": 0.9994230270385742, + "learning_rate": 1.5069004178082213e-05, + "loss": 0.7606, + "step": 18498 + }, + { + "epoch": 3.2934472934472936, + "grad_norm": 0.7854368090629578, + "learning_rate": 1.5061615821667185e-05, + "loss": 0.4879, + "step": 18499 + }, + { + "epoch": 3.2936253561253563, + "grad_norm": 1.0826393365859985, + "learning_rate": 1.5054229129443397e-05, + "loss": 0.7172, + "step": 18500 + }, + { + "epoch": 3.2938034188034186, + "grad_norm": 0.9371114373207092, + "learning_rate": 1.5046844101555491e-05, + "loss": 0.8157, + "step": 18501 + }, + { + "epoch": 3.2939814814814814, + "grad_norm": 1.0282293558120728, + "learning_rate": 1.5039460738148215e-05, + "loss": 0.7367, + "step": 18502 + }, + { + "epoch": 3.294159544159544, + "grad_norm": 1.0519812107086182, + "learning_rate": 1.5032079039366209e-05, + "loss": 0.8756, + "step": 18503 + }, + { + "epoch": 3.294337606837607, + "grad_norm": 0.9510549306869507, + "learning_rate": 1.5024699005354127e-05, + "loss": 0.7443, + "step": 18504 + }, + { + "epoch": 3.2945156695156697, + "grad_norm": 0.871846616268158, + "learning_rate": 1.5017320636256527e-05, + "loss": 0.8086, + "step": 18505 + }, + { + "epoch": 3.294693732193732, + "grad_norm": 0.9595451951026917, + "learning_rate": 1.5009943932217985e-05, + "loss": 0.7864, + "step": 18506 + }, + { + "epoch": 3.2948717948717947, + "grad_norm": 1.0278631448745728, + "learning_rate": 1.5002568893383051e-05, + "loss": 0.8224, + "step": 18507 + }, + { + "epoch": 3.2950498575498575, + "grad_norm": 0.9716551899909973, + "learning_rate": 1.499519551989622e-05, + "loss": 0.7774, + "step": 18508 + }, + { + "epoch": 3.29522792022792, + "grad_norm": 0.9664580821990967, + "learning_rate": 1.4987823811901958e-05, + "loss": 0.8641, + "step": 18509 + }, + { + "epoch": 3.295405982905983, + "grad_norm": 0.9358782768249512, + "learning_rate": 1.4980453769544677e-05, + "loss": 0.8348, + "step": 18510 + }, + { + "epoch": 3.2955840455840457, + "grad_norm": 0.8839470744132996, + "learning_rate": 1.49730853929688e-05, + "loss": 0.9168, + "step": 18511 + }, + { + "epoch": 3.2957621082621085, + "grad_norm": 1.004765272140503, + "learning_rate": 1.4965718682318685e-05, + "loss": 0.8787, + "step": 18512 + }, + { + "epoch": 3.2959401709401708, + "grad_norm": 0.8481960296630859, + "learning_rate": 1.4958353637738665e-05, + "loss": 0.5791, + "step": 18513 + }, + { + "epoch": 3.2961182336182335, + "grad_norm": 0.884299635887146, + "learning_rate": 1.4950990259373032e-05, + "loss": 0.8386, + "step": 18514 + }, + { + "epoch": 3.2962962962962963, + "grad_norm": 1.1240220069885254, + "learning_rate": 1.4943628547366106e-05, + "loss": 0.8071, + "step": 18515 + }, + { + "epoch": 3.296474358974359, + "grad_norm": 1.0169334411621094, + "learning_rate": 1.493626850186205e-05, + "loss": 0.8099, + "step": 18516 + }, + { + "epoch": 3.296652421652422, + "grad_norm": 0.8471205234527588, + "learning_rate": 1.4928910123005135e-05, + "loss": 0.7887, + "step": 18517 + }, + { + "epoch": 3.296830484330484, + "grad_norm": 0.9037573933601379, + "learning_rate": 1.4921553410939492e-05, + "loss": 0.7674, + "step": 18518 + }, + { + "epoch": 3.297008547008547, + "grad_norm": 1.0591832399368286, + "learning_rate": 1.4914198365809296e-05, + "loss": 0.7349, + "step": 18519 + }, + { + "epoch": 3.2971866096866096, + "grad_norm": 0.7707160115242004, + "learning_rate": 1.4906844987758616e-05, + "loss": 0.5358, + "step": 18520 + }, + { + "epoch": 3.2973646723646723, + "grad_norm": 0.899319589138031, + "learning_rate": 1.4899493276931541e-05, + "loss": 0.7471, + "step": 18521 + }, + { + "epoch": 3.297542735042735, + "grad_norm": 0.9070066809654236, + "learning_rate": 1.4892143233472121e-05, + "loss": 0.7375, + "step": 18522 + }, + { + "epoch": 3.297720797720798, + "grad_norm": 0.8646981716156006, + "learning_rate": 1.4884794857524364e-05, + "loss": 0.7047, + "step": 18523 + }, + { + "epoch": 3.2978988603988606, + "grad_norm": 0.9446913003921509, + "learning_rate": 1.487744814923221e-05, + "loss": 0.6036, + "step": 18524 + }, + { + "epoch": 3.298076923076923, + "grad_norm": 1.2150342464447021, + "learning_rate": 1.4870103108739674e-05, + "loss": 0.8835, + "step": 18525 + }, + { + "epoch": 3.2982549857549857, + "grad_norm": 0.8841198682785034, + "learning_rate": 1.486275973619059e-05, + "loss": 0.7056, + "step": 18526 + }, + { + "epoch": 3.2984330484330484, + "grad_norm": 0.8644464612007141, + "learning_rate": 1.4855418031728885e-05, + "loss": 0.6371, + "step": 18527 + }, + { + "epoch": 3.298611111111111, + "grad_norm": 0.8655923008918762, + "learning_rate": 1.48480779954984e-05, + "loss": 0.6938, + "step": 18528 + }, + { + "epoch": 3.298789173789174, + "grad_norm": 0.9875824451446533, + "learning_rate": 1.4840739627642908e-05, + "loss": 0.7635, + "step": 18529 + }, + { + "epoch": 3.298967236467236, + "grad_norm": 0.8669924139976501, + "learning_rate": 1.4833402928306273e-05, + "loss": 0.5327, + "step": 18530 + }, + { + "epoch": 3.299145299145299, + "grad_norm": 0.9445573687553406, + "learning_rate": 1.4826067897632146e-05, + "loss": 0.6969, + "step": 18531 + }, + { + "epoch": 3.2993233618233617, + "grad_norm": 0.8305925726890564, + "learning_rate": 1.4818734535764323e-05, + "loss": 0.7279, + "step": 18532 + }, + { + "epoch": 3.2995014245014245, + "grad_norm": 0.9907183051109314, + "learning_rate": 1.48114028428464e-05, + "loss": 0.8681, + "step": 18533 + }, + { + "epoch": 3.2996794871794872, + "grad_norm": 1.065702199935913, + "learning_rate": 1.4804072819022108e-05, + "loss": 0.8344, + "step": 18534 + }, + { + "epoch": 3.29985754985755, + "grad_norm": 1.0929903984069824, + "learning_rate": 1.479674446443502e-05, + "loss": 0.8327, + "step": 18535 + }, + { + "epoch": 3.3000356125356127, + "grad_norm": 1.0050452947616577, + "learning_rate": 1.4789417779228732e-05, + "loss": 0.6864, + "step": 18536 + }, + { + "epoch": 3.300213675213675, + "grad_norm": 0.9825767278671265, + "learning_rate": 1.4782092763546795e-05, + "loss": 0.7337, + "step": 18537 + }, + { + "epoch": 3.300391737891738, + "grad_norm": 0.867043137550354, + "learning_rate": 1.477476941753273e-05, + "loss": 0.8247, + "step": 18538 + }, + { + "epoch": 3.3005698005698005, + "grad_norm": 0.9017619490623474, + "learning_rate": 1.4767447741329987e-05, + "loss": 0.7135, + "step": 18539 + }, + { + "epoch": 3.3007478632478633, + "grad_norm": 0.8865274786949158, + "learning_rate": 1.4760127735082097e-05, + "loss": 0.7764, + "step": 18540 + }, + { + "epoch": 3.300925925925926, + "grad_norm": 1.0825269222259521, + "learning_rate": 1.4752809398932388e-05, + "loss": 0.7734, + "step": 18541 + }, + { + "epoch": 3.301103988603989, + "grad_norm": 0.934187650680542, + "learning_rate": 1.4745492733024325e-05, + "loss": 0.7147, + "step": 18542 + }, + { + "epoch": 3.301282051282051, + "grad_norm": 0.9431992769241333, + "learning_rate": 1.4738177737501214e-05, + "loss": 0.7086, + "step": 18543 + }, + { + "epoch": 3.301460113960114, + "grad_norm": 0.9284996390342712, + "learning_rate": 1.4730864412506407e-05, + "loss": 0.8364, + "step": 18544 + }, + { + "epoch": 3.3016381766381766, + "grad_norm": 0.7849892377853394, + "learning_rate": 1.4723552758183178e-05, + "loss": 0.7114, + "step": 18545 + }, + { + "epoch": 3.3018162393162394, + "grad_norm": 1.096384882926941, + "learning_rate": 1.471624277467476e-05, + "loss": 0.8764, + "step": 18546 + }, + { + "epoch": 3.301994301994302, + "grad_norm": 0.9354200959205627, + "learning_rate": 1.4708934462124458e-05, + "loss": 0.8211, + "step": 18547 + }, + { + "epoch": 3.302172364672365, + "grad_norm": 0.8292770981788635, + "learning_rate": 1.4701627820675356e-05, + "loss": 0.8402, + "step": 18548 + }, + { + "epoch": 3.302350427350427, + "grad_norm": 0.8030630946159363, + "learning_rate": 1.4694322850470687e-05, + "loss": 0.4122, + "step": 18549 + }, + { + "epoch": 3.30252849002849, + "grad_norm": 0.8566523790359497, + "learning_rate": 1.4687019551653558e-05, + "loss": 0.6334, + "step": 18550 + }, + { + "epoch": 3.3027065527065527, + "grad_norm": 0.869368851184845, + "learning_rate": 1.4679717924367053e-05, + "loss": 0.8575, + "step": 18551 + }, + { + "epoch": 3.3028846153846154, + "grad_norm": 0.9408078193664551, + "learning_rate": 1.4672417968754237e-05, + "loss": 0.6937, + "step": 18552 + }, + { + "epoch": 3.303062678062678, + "grad_norm": 0.9742521643638611, + "learning_rate": 1.466511968495814e-05, + "loss": 0.7323, + "step": 18553 + }, + { + "epoch": 3.303240740740741, + "grad_norm": 1.073304533958435, + "learning_rate": 1.4657823073121735e-05, + "loss": 0.6501, + "step": 18554 + }, + { + "epoch": 3.3034188034188032, + "grad_norm": 1.0220160484313965, + "learning_rate": 1.4650528133388053e-05, + "loss": 0.7497, + "step": 18555 + }, + { + "epoch": 3.303596866096866, + "grad_norm": 0.9558535218238831, + "learning_rate": 1.4643234865899924e-05, + "loss": 0.796, + "step": 18556 + }, + { + "epoch": 3.3037749287749287, + "grad_norm": 0.9674620628356934, + "learning_rate": 1.4635943270800334e-05, + "loss": 0.806, + "step": 18557 + }, + { + "epoch": 3.3039529914529915, + "grad_norm": 0.934508204460144, + "learning_rate": 1.462865334823207e-05, + "loss": 0.6858, + "step": 18558 + }, + { + "epoch": 3.3041310541310542, + "grad_norm": 0.9529309868812561, + "learning_rate": 1.4621365098338024e-05, + "loss": 0.719, + "step": 18559 + }, + { + "epoch": 3.304309116809117, + "grad_norm": 0.9932489395141602, + "learning_rate": 1.4614078521260965e-05, + "loss": 0.9503, + "step": 18560 + }, + { + "epoch": 3.3044871794871793, + "grad_norm": 0.9791563749313354, + "learning_rate": 1.4606793617143667e-05, + "loss": 0.8629, + "step": 18561 + }, + { + "epoch": 3.304665242165242, + "grad_norm": 1.036198616027832, + "learning_rate": 1.4599510386128857e-05, + "loss": 0.82, + "step": 18562 + }, + { + "epoch": 3.304843304843305, + "grad_norm": 0.9353040456771851, + "learning_rate": 1.4592228828359234e-05, + "loss": 0.7878, + "step": 18563 + }, + { + "epoch": 3.3050213675213675, + "grad_norm": 1.0084593296051025, + "learning_rate": 1.458494894397745e-05, + "loss": 0.7765, + "step": 18564 + }, + { + "epoch": 3.3051994301994303, + "grad_norm": 1.0099483728408813, + "learning_rate": 1.4577670733126203e-05, + "loss": 0.9263, + "step": 18565 + }, + { + "epoch": 3.305377492877493, + "grad_norm": 0.8631238341331482, + "learning_rate": 1.4570394195948e-05, + "loss": 0.739, + "step": 18566 + }, + { + "epoch": 3.3055555555555554, + "grad_norm": 0.8711506724357605, + "learning_rate": 1.4563119332585484e-05, + "loss": 0.7182, + "step": 18567 + }, + { + "epoch": 3.305733618233618, + "grad_norm": 0.9329670667648315, + "learning_rate": 1.4555846143181162e-05, + "loss": 0.6453, + "step": 18568 + }, + { + "epoch": 3.305911680911681, + "grad_norm": 0.9587839245796204, + "learning_rate": 1.4548574627877532e-05, + "loss": 0.7599, + "step": 18569 + }, + { + "epoch": 3.3060897435897436, + "grad_norm": 0.9862858653068542, + "learning_rate": 1.4541304786817089e-05, + "loss": 0.8038, + "step": 18570 + }, + { + "epoch": 3.3062678062678064, + "grad_norm": 0.852069616317749, + "learning_rate": 1.4534036620142221e-05, + "loss": 0.8945, + "step": 18571 + }, + { + "epoch": 3.306445868945869, + "grad_norm": 0.9384768009185791, + "learning_rate": 1.4526770127995415e-05, + "loss": 0.7363, + "step": 18572 + }, + { + "epoch": 3.306623931623932, + "grad_norm": 0.8548694252967834, + "learning_rate": 1.4519505310518944e-05, + "loss": 0.7693, + "step": 18573 + }, + { + "epoch": 3.306801994301994, + "grad_norm": 1.0954573154449463, + "learning_rate": 1.4512242167855216e-05, + "loss": 0.8015, + "step": 18574 + }, + { + "epoch": 3.306980056980057, + "grad_norm": 0.8069053292274475, + "learning_rate": 1.450498070014652e-05, + "loss": 0.6506, + "step": 18575 + }, + { + "epoch": 3.3071581196581197, + "grad_norm": 0.8466541171073914, + "learning_rate": 1.4497720907535128e-05, + "loss": 0.6072, + "step": 18576 + }, + { + "epoch": 3.3073361823361824, + "grad_norm": 0.7500994205474854, + "learning_rate": 1.4490462790163285e-05, + "loss": 0.7196, + "step": 18577 + }, + { + "epoch": 3.307514245014245, + "grad_norm": 0.9503157734870911, + "learning_rate": 1.4483206348173185e-05, + "loss": 0.7822, + "step": 18578 + }, + { + "epoch": 3.3076923076923075, + "grad_norm": 0.9990713596343994, + "learning_rate": 1.447595158170698e-05, + "loss": 0.8523, + "step": 18579 + }, + { + "epoch": 3.3078703703703702, + "grad_norm": 1.0937577486038208, + "learning_rate": 1.4468698490906907e-05, + "loss": 0.6461, + "step": 18580 + }, + { + "epoch": 3.308048433048433, + "grad_norm": 0.8959493637084961, + "learning_rate": 1.4461447075914957e-05, + "loss": 0.6803, + "step": 18581 + }, + { + "epoch": 3.3082264957264957, + "grad_norm": 0.9823698997497559, + "learning_rate": 1.4454197336873299e-05, + "loss": 0.8456, + "step": 18582 + }, + { + "epoch": 3.3084045584045585, + "grad_norm": 0.9720016121864319, + "learning_rate": 1.4446949273923904e-05, + "loss": 0.7386, + "step": 18583 + }, + { + "epoch": 3.3085826210826212, + "grad_norm": 0.9927224516868591, + "learning_rate": 1.4439702887208839e-05, + "loss": 0.706, + "step": 18584 + }, + { + "epoch": 3.308760683760684, + "grad_norm": 0.8993792533874512, + "learning_rate": 1.4432458176870056e-05, + "loss": 0.7265, + "step": 18585 + }, + { + "epoch": 3.3089387464387463, + "grad_norm": 0.7669193148612976, + "learning_rate": 1.4425215143049475e-05, + "loss": 0.5659, + "step": 18586 + }, + { + "epoch": 3.309116809116809, + "grad_norm": 0.9683290123939514, + "learning_rate": 1.4417973785889094e-05, + "loss": 0.967, + "step": 18587 + }, + { + "epoch": 3.309294871794872, + "grad_norm": 0.8810568451881409, + "learning_rate": 1.4410734105530688e-05, + "loss": 0.7119, + "step": 18588 + }, + { + "epoch": 3.3094729344729346, + "grad_norm": 0.9147163033485413, + "learning_rate": 1.4403496102116165e-05, + "loss": 0.7214, + "step": 18589 + }, + { + "epoch": 3.3096509971509973, + "grad_norm": 1.13361394405365, + "learning_rate": 1.4396259775787335e-05, + "loss": 0.8442, + "step": 18590 + }, + { + "epoch": 3.3098290598290596, + "grad_norm": 0.8798992037773132, + "learning_rate": 1.4389025126685962e-05, + "loss": 0.5915, + "step": 18591 + }, + { + "epoch": 3.3100071225071224, + "grad_norm": 0.9003865718841553, + "learning_rate": 1.43817921549538e-05, + "loss": 0.6996, + "step": 18592 + }, + { + "epoch": 3.310185185185185, + "grad_norm": 0.8331068754196167, + "learning_rate": 1.4374560860732567e-05, + "loss": 0.4284, + "step": 18593 + }, + { + "epoch": 3.310363247863248, + "grad_norm": 0.8950058817863464, + "learning_rate": 1.4367331244163939e-05, + "loss": 0.6156, + "step": 18594 + }, + { + "epoch": 3.3105413105413106, + "grad_norm": 0.8993448615074158, + "learning_rate": 1.4360103305389572e-05, + "loss": 0.8449, + "step": 18595 + }, + { + "epoch": 3.3107193732193734, + "grad_norm": 0.9031822085380554, + "learning_rate": 1.435287704455105e-05, + "loss": 0.6608, + "step": 18596 + }, + { + "epoch": 3.310897435897436, + "grad_norm": 0.8823438882827759, + "learning_rate": 1.4345652461790038e-05, + "loss": 0.6606, + "step": 18597 + }, + { + "epoch": 3.3110754985754984, + "grad_norm": 1.0061595439910889, + "learning_rate": 1.433842955724799e-05, + "loss": 0.7287, + "step": 18598 + }, + { + "epoch": 3.311253561253561, + "grad_norm": 1.0008385181427002, + "learning_rate": 1.4331208331066493e-05, + "loss": 0.8076, + "step": 18599 + }, + { + "epoch": 3.311431623931624, + "grad_norm": 0.7386804819107056, + "learning_rate": 1.4323988783387e-05, + "loss": 0.4206, + "step": 18600 + }, + { + "epoch": 3.3116096866096867, + "grad_norm": 0.9149610996246338, + "learning_rate": 1.4316770914350975e-05, + "loss": 0.6752, + "step": 18601 + }, + { + "epoch": 3.3117877492877494, + "grad_norm": 0.9260136485099792, + "learning_rate": 1.4309554724099827e-05, + "loss": 0.6503, + "step": 18602 + }, + { + "epoch": 3.3119658119658117, + "grad_norm": 0.9129661321640015, + "learning_rate": 1.4302340212774956e-05, + "loss": 0.7452, + "step": 18603 + }, + { + "epoch": 3.3121438746438745, + "grad_norm": 0.9771081805229187, + "learning_rate": 1.4295127380517704e-05, + "loss": 0.8258, + "step": 18604 + }, + { + "epoch": 3.3123219373219372, + "grad_norm": 0.8027868866920471, + "learning_rate": 1.4287916227469366e-05, + "loss": 0.617, + "step": 18605 + }, + { + "epoch": 3.3125, + "grad_norm": 0.8909293413162231, + "learning_rate": 1.428070675377129e-05, + "loss": 0.6898, + "step": 18606 + }, + { + "epoch": 3.3126780626780628, + "grad_norm": 0.8994025588035583, + "learning_rate": 1.427349895956469e-05, + "loss": 0.8001, + "step": 18607 + }, + { + "epoch": 3.3128561253561255, + "grad_norm": 1.062908411026001, + "learning_rate": 1.4266292844990802e-05, + "loss": 0.9458, + "step": 18608 + }, + { + "epoch": 3.3130341880341883, + "grad_norm": 0.9574441909790039, + "learning_rate": 1.4259088410190813e-05, + "loss": 0.7763, + "step": 18609 + }, + { + "epoch": 3.3132122507122506, + "grad_norm": 0.9709355235099792, + "learning_rate": 1.4251885655305863e-05, + "loss": 0.6442, + "step": 18610 + }, + { + "epoch": 3.3133903133903133, + "grad_norm": 0.964805006980896, + "learning_rate": 1.4244684580477075e-05, + "loss": 0.8413, + "step": 18611 + }, + { + "epoch": 3.313568376068376, + "grad_norm": 0.806939959526062, + "learning_rate": 1.423748518584559e-05, + "loss": 0.5702, + "step": 18612 + }, + { + "epoch": 3.313746438746439, + "grad_norm": 0.9942196607589722, + "learning_rate": 1.4230287471552394e-05, + "loss": 0.8193, + "step": 18613 + }, + { + "epoch": 3.3139245014245016, + "grad_norm": 1.024062991142273, + "learning_rate": 1.4223091437738556e-05, + "loss": 0.6324, + "step": 18614 + }, + { + "epoch": 3.314102564102564, + "grad_norm": 0.8655102252960205, + "learning_rate": 1.421589708454505e-05, + "loss": 0.7141, + "step": 18615 + }, + { + "epoch": 3.3142806267806266, + "grad_norm": 0.8736686706542969, + "learning_rate": 1.420870441211285e-05, + "loss": 0.7649, + "step": 18616 + }, + { + "epoch": 3.3144586894586894, + "grad_norm": 1.0445935726165771, + "learning_rate": 1.4201513420582868e-05, + "loss": 0.943, + "step": 18617 + }, + { + "epoch": 3.314636752136752, + "grad_norm": 1.0183906555175781, + "learning_rate": 1.4194324110096002e-05, + "loss": 0.7352, + "step": 18618 + }, + { + "epoch": 3.314814814814815, + "grad_norm": 0.8862751126289368, + "learning_rate": 1.4187136480793106e-05, + "loss": 0.6363, + "step": 18619 + }, + { + "epoch": 3.3149928774928776, + "grad_norm": 0.9212391972541809, + "learning_rate": 1.4179950532815011e-05, + "loss": 0.7494, + "step": 18620 + }, + { + "epoch": 3.3151709401709404, + "grad_norm": 0.8244057297706604, + "learning_rate": 1.4172766266302484e-05, + "loss": 0.5704, + "step": 18621 + }, + { + "epoch": 3.3153490028490027, + "grad_norm": 1.016066551208496, + "learning_rate": 1.4165583681396355e-05, + "loss": 0.8147, + "step": 18622 + }, + { + "epoch": 3.3155270655270654, + "grad_norm": 0.938336968421936, + "learning_rate": 1.4158402778237267e-05, + "loss": 0.7383, + "step": 18623 + }, + { + "epoch": 3.315705128205128, + "grad_norm": 1.1921639442443848, + "learning_rate": 1.4151223556965976e-05, + "loss": 0.8468, + "step": 18624 + }, + { + "epoch": 3.315883190883191, + "grad_norm": 1.0110679864883423, + "learning_rate": 1.4144046017723134e-05, + "loss": 0.8632, + "step": 18625 + }, + { + "epoch": 3.3160612535612537, + "grad_norm": 1.1017330884933472, + "learning_rate": 1.4136870160649329e-05, + "loss": 0.8882, + "step": 18626 + }, + { + "epoch": 3.316239316239316, + "grad_norm": 0.8759216070175171, + "learning_rate": 1.4129695985885228e-05, + "loss": 0.8481, + "step": 18627 + }, + { + "epoch": 3.3164173789173788, + "grad_norm": 0.9331730008125305, + "learning_rate": 1.4122523493571316e-05, + "loss": 0.7723, + "step": 18628 + }, + { + "epoch": 3.3165954415954415, + "grad_norm": 0.8843693137168884, + "learning_rate": 1.4115352683848204e-05, + "loss": 0.7457, + "step": 18629 + }, + { + "epoch": 3.3167735042735043, + "grad_norm": 1.042717456817627, + "learning_rate": 1.4108183556856302e-05, + "loss": 0.7947, + "step": 18630 + }, + { + "epoch": 3.316951566951567, + "grad_norm": 0.912771999835968, + "learning_rate": 1.4101016112736143e-05, + "loss": 0.7209, + "step": 18631 + }, + { + "epoch": 3.3171296296296298, + "grad_norm": 0.8966865539550781, + "learning_rate": 1.4093850351628136e-05, + "loss": 0.8184, + "step": 18632 + }, + { + "epoch": 3.3173076923076925, + "grad_norm": 0.9174174666404724, + "learning_rate": 1.4086686273672668e-05, + "loss": 0.7335, + "step": 18633 + }, + { + "epoch": 3.317485754985755, + "grad_norm": 0.9604222178459167, + "learning_rate": 1.4079523879010114e-05, + "loss": 0.7813, + "step": 18634 + }, + { + "epoch": 3.3176638176638176, + "grad_norm": 0.9172660708427429, + "learning_rate": 1.4072363167780811e-05, + "loss": 0.7329, + "step": 18635 + }, + { + "epoch": 3.3178418803418803, + "grad_norm": 0.8856810927391052, + "learning_rate": 1.4065204140125033e-05, + "loss": 0.7138, + "step": 18636 + }, + { + "epoch": 3.318019943019943, + "grad_norm": 0.9621987342834473, + "learning_rate": 1.40580467961831e-05, + "loss": 0.777, + "step": 18637 + }, + { + "epoch": 3.318198005698006, + "grad_norm": 0.9171984791755676, + "learning_rate": 1.4050891136095179e-05, + "loss": 0.9051, + "step": 18638 + }, + { + "epoch": 3.318376068376068, + "grad_norm": 0.963080108165741, + "learning_rate": 1.4043737160001536e-05, + "loss": 0.7974, + "step": 18639 + }, + { + "epoch": 3.318554131054131, + "grad_norm": 0.8158769607543945, + "learning_rate": 1.4036584868042268e-05, + "loss": 0.6412, + "step": 18640 + }, + { + "epoch": 3.3187321937321936, + "grad_norm": 1.070873737335205, + "learning_rate": 1.4029434260357565e-05, + "loss": 0.8986, + "step": 18641 + }, + { + "epoch": 3.3189102564102564, + "grad_norm": 1.0787153244018555, + "learning_rate": 1.4022285337087504e-05, + "loss": 0.7754, + "step": 18642 + }, + { + "epoch": 3.319088319088319, + "grad_norm": 0.8161159157752991, + "learning_rate": 1.401513809837216e-05, + "loss": 0.6833, + "step": 18643 + }, + { + "epoch": 3.319266381766382, + "grad_norm": 0.9341142177581787, + "learning_rate": 1.4007992544351578e-05, + "loss": 0.6834, + "step": 18644 + }, + { + "epoch": 3.3194444444444446, + "grad_norm": 0.9400647878646851, + "learning_rate": 1.4000848675165711e-05, + "loss": 0.8276, + "step": 18645 + }, + { + "epoch": 3.319622507122507, + "grad_norm": 0.8795194029808044, + "learning_rate": 1.3993706490954595e-05, + "loss": 0.7008, + "step": 18646 + }, + { + "epoch": 3.3198005698005697, + "grad_norm": 1.02724289894104, + "learning_rate": 1.3986565991858148e-05, + "loss": 0.743, + "step": 18647 + }, + { + "epoch": 3.3199786324786325, + "grad_norm": 1.0184358358383179, + "learning_rate": 1.3979427178016247e-05, + "loss": 0.9255, + "step": 18648 + }, + { + "epoch": 3.320156695156695, + "grad_norm": 0.9417447447776794, + "learning_rate": 1.3972290049568793e-05, + "loss": 0.7776, + "step": 18649 + }, + { + "epoch": 3.320334757834758, + "grad_norm": 1.0770184993743896, + "learning_rate": 1.3965154606655595e-05, + "loss": 0.9044, + "step": 18650 + }, + { + "epoch": 3.3205128205128207, + "grad_norm": 1.0059106349945068, + "learning_rate": 1.3958020849416453e-05, + "loss": 0.8848, + "step": 18651 + }, + { + "epoch": 3.320690883190883, + "grad_norm": 0.8800293207168579, + "learning_rate": 1.39508887779912e-05, + "loss": 0.627, + "step": 18652 + }, + { + "epoch": 3.3208689458689458, + "grad_norm": 1.1125677824020386, + "learning_rate": 1.394375839251949e-05, + "loss": 1.0158, + "step": 18653 + }, + { + "epoch": 3.3210470085470085, + "grad_norm": 0.9559670686721802, + "learning_rate": 1.3936629693141112e-05, + "loss": 0.8771, + "step": 18654 + }, + { + "epoch": 3.3212250712250713, + "grad_norm": 0.8112167119979858, + "learning_rate": 1.3929502679995643e-05, + "loss": 0.5987, + "step": 18655 + }, + { + "epoch": 3.321403133903134, + "grad_norm": 1.1909687519073486, + "learning_rate": 1.3922377353222805e-05, + "loss": 0.8939, + "step": 18656 + }, + { + "epoch": 3.3215811965811968, + "grad_norm": 0.9428912997245789, + "learning_rate": 1.3915253712962162e-05, + "loss": 0.8213, + "step": 18657 + }, + { + "epoch": 3.321759259259259, + "grad_norm": 0.9573705792427063, + "learning_rate": 1.3908131759353304e-05, + "loss": 0.8073, + "step": 18658 + }, + { + "epoch": 3.321937321937322, + "grad_norm": 0.8574663400650024, + "learning_rate": 1.3901011492535754e-05, + "loss": 0.5301, + "step": 18659 + }, + { + "epoch": 3.3221153846153846, + "grad_norm": 1.016772985458374, + "learning_rate": 1.3893892912649043e-05, + "loss": 0.5597, + "step": 18660 + }, + { + "epoch": 3.3222934472934473, + "grad_norm": 0.8159156441688538, + "learning_rate": 1.3886776019832592e-05, + "loss": 0.5178, + "step": 18661 + }, + { + "epoch": 3.32247150997151, + "grad_norm": 1.0510454177856445, + "learning_rate": 1.3879660814225937e-05, + "loss": 0.7447, + "step": 18662 + }, + { + "epoch": 3.322649572649573, + "grad_norm": 0.9608046412467957, + "learning_rate": 1.3872547295968386e-05, + "loss": 0.567, + "step": 18663 + }, + { + "epoch": 3.322827635327635, + "grad_norm": 0.9422797560691833, + "learning_rate": 1.3865435465199394e-05, + "loss": 0.7763, + "step": 18664 + }, + { + "epoch": 3.323005698005698, + "grad_norm": 0.8879508376121521, + "learning_rate": 1.385832532205822e-05, + "loss": 0.6057, + "step": 18665 + }, + { + "epoch": 3.3231837606837606, + "grad_norm": 0.8140594959259033, + "learning_rate": 1.3851216866684236e-05, + "loss": 0.6785, + "step": 18666 + }, + { + "epoch": 3.3233618233618234, + "grad_norm": 0.8440611362457275, + "learning_rate": 1.3844110099216712e-05, + "loss": 0.8621, + "step": 18667 + }, + { + "epoch": 3.323539886039886, + "grad_norm": 1.1034574508666992, + "learning_rate": 1.3837005019794847e-05, + "loss": 0.8346, + "step": 18668 + }, + { + "epoch": 3.323717948717949, + "grad_norm": 0.9221482276916504, + "learning_rate": 1.3829901628557928e-05, + "loss": 0.848, + "step": 18669 + }, + { + "epoch": 3.323896011396011, + "grad_norm": 0.9995195865631104, + "learning_rate": 1.3822799925645036e-05, + "loss": 0.7762, + "step": 18670 + }, + { + "epoch": 3.324074074074074, + "grad_norm": 0.8469681143760681, + "learning_rate": 1.3815699911195379e-05, + "loss": 0.7104, + "step": 18671 + }, + { + "epoch": 3.3242521367521367, + "grad_norm": 0.9675276875495911, + "learning_rate": 1.3808601585348057e-05, + "loss": 0.8602, + "step": 18672 + }, + { + "epoch": 3.3244301994301995, + "grad_norm": 0.9134352803230286, + "learning_rate": 1.3801504948242138e-05, + "loss": 0.7042, + "step": 18673 + }, + { + "epoch": 3.324608262108262, + "grad_norm": 0.9910576939582825, + "learning_rate": 1.3794410000016667e-05, + "loss": 0.714, + "step": 18674 + }, + { + "epoch": 3.324786324786325, + "grad_norm": 1.1341938972473145, + "learning_rate": 1.3787316740810663e-05, + "loss": 0.838, + "step": 18675 + }, + { + "epoch": 3.3249643874643873, + "grad_norm": 0.8614459037780762, + "learning_rate": 1.3780225170763083e-05, + "loss": 0.699, + "step": 18676 + }, + { + "epoch": 3.32514245014245, + "grad_norm": 0.906356155872345, + "learning_rate": 1.3773135290012895e-05, + "loss": 0.7302, + "step": 18677 + }, + { + "epoch": 3.3253205128205128, + "grad_norm": 1.0148662328720093, + "learning_rate": 1.3766047098698975e-05, + "loss": 0.6931, + "step": 18678 + }, + { + "epoch": 3.3254985754985755, + "grad_norm": 0.9177684187889099, + "learning_rate": 1.3758960596960268e-05, + "loss": 0.8363, + "step": 18679 + }, + { + "epoch": 3.3256766381766383, + "grad_norm": 1.0270618200302124, + "learning_rate": 1.3751875784935542e-05, + "loss": 0.823, + "step": 18680 + }, + { + "epoch": 3.325854700854701, + "grad_norm": 0.9151124954223633, + "learning_rate": 1.3744792662763661e-05, + "loss": 0.7939, + "step": 18681 + }, + { + "epoch": 3.326032763532764, + "grad_norm": 0.9115209579467773, + "learning_rate": 1.3737711230583384e-05, + "loss": 0.7914, + "step": 18682 + }, + { + "epoch": 3.326210826210826, + "grad_norm": 1.0149354934692383, + "learning_rate": 1.3730631488533462e-05, + "loss": 0.8579, + "step": 18683 + }, + { + "epoch": 3.326388888888889, + "grad_norm": 0.9036294221878052, + "learning_rate": 1.3723553436752612e-05, + "loss": 0.5274, + "step": 18684 + }, + { + "epoch": 3.3265669515669516, + "grad_norm": 1.0346934795379639, + "learning_rate": 1.3716477075379485e-05, + "loss": 0.8489, + "step": 18685 + }, + { + "epoch": 3.3267450142450143, + "grad_norm": 0.8915591835975647, + "learning_rate": 1.3709402404552773e-05, + "loss": 0.7278, + "step": 18686 + }, + { + "epoch": 3.326923076923077, + "grad_norm": 0.9130534529685974, + "learning_rate": 1.3702329424411076e-05, + "loss": 0.7576, + "step": 18687 + }, + { + "epoch": 3.3271011396011394, + "grad_norm": 0.9920860528945923, + "learning_rate": 1.3695258135092959e-05, + "loss": 0.7424, + "step": 18688 + }, + { + "epoch": 3.327279202279202, + "grad_norm": 0.9840420484542847, + "learning_rate": 1.3688188536736968e-05, + "loss": 0.6867, + "step": 18689 + }, + { + "epoch": 3.327457264957265, + "grad_norm": 0.8766686320304871, + "learning_rate": 1.3681120629481637e-05, + "loss": 0.6816, + "step": 18690 + }, + { + "epoch": 3.3276353276353277, + "grad_norm": 1.0345094203948975, + "learning_rate": 1.367405441346543e-05, + "loss": 0.9062, + "step": 18691 + }, + { + "epoch": 3.3278133903133904, + "grad_norm": 0.9291123151779175, + "learning_rate": 1.3666989888826798e-05, + "loss": 0.9039, + "step": 18692 + }, + { + "epoch": 3.327991452991453, + "grad_norm": 0.8739197254180908, + "learning_rate": 1.3659927055704136e-05, + "loss": 0.6158, + "step": 18693 + }, + { + "epoch": 3.328169515669516, + "grad_norm": 1.0071395635604858, + "learning_rate": 1.3652865914235901e-05, + "loss": 0.7055, + "step": 18694 + }, + { + "epoch": 3.328347578347578, + "grad_norm": 0.9326104521751404, + "learning_rate": 1.364580646456033e-05, + "loss": 0.795, + "step": 18695 + }, + { + "epoch": 3.328525641025641, + "grad_norm": 0.9884015321731567, + "learning_rate": 1.363874870681583e-05, + "loss": 0.8504, + "step": 18696 + }, + { + "epoch": 3.3287037037037037, + "grad_norm": 0.9422696828842163, + "learning_rate": 1.3631692641140647e-05, + "loss": 0.6087, + "step": 18697 + }, + { + "epoch": 3.3288817663817665, + "grad_norm": 0.9124358296394348, + "learning_rate": 1.3624638267673029e-05, + "loss": 0.8993, + "step": 18698 + }, + { + "epoch": 3.3290598290598292, + "grad_norm": 0.9755852818489075, + "learning_rate": 1.3617585586551196e-05, + "loss": 0.8998, + "step": 18699 + }, + { + "epoch": 3.3292378917378915, + "grad_norm": 0.9847198128700256, + "learning_rate": 1.3610534597913326e-05, + "loss": 0.7212, + "step": 18700 + }, + { + "epoch": 3.3294159544159543, + "grad_norm": 0.9432051777839661, + "learning_rate": 1.3603485301897579e-05, + "loss": 0.6306, + "step": 18701 + }, + { + "epoch": 3.329594017094017, + "grad_norm": 0.973268985748291, + "learning_rate": 1.359643769864205e-05, + "loss": 0.8055, + "step": 18702 + }, + { + "epoch": 3.32977207977208, + "grad_norm": 1.0206477642059326, + "learning_rate": 1.3589391788284834e-05, + "loss": 0.8731, + "step": 18703 + }, + { + "epoch": 3.3299501424501425, + "grad_norm": 1.0303237438201904, + "learning_rate": 1.3582347570963994e-05, + "loss": 0.7416, + "step": 18704 + }, + { + "epoch": 3.3301282051282053, + "grad_norm": 0.9327782392501831, + "learning_rate": 1.3575305046817533e-05, + "loss": 1.1033, + "step": 18705 + }, + { + "epoch": 3.330306267806268, + "grad_norm": 1.0996826887130737, + "learning_rate": 1.3568264215983439e-05, + "loss": 0.6945, + "step": 18706 + }, + { + "epoch": 3.3304843304843303, + "grad_norm": 0.9221864342689514, + "learning_rate": 1.356122507859967e-05, + "loss": 0.64, + "step": 18707 + }, + { + "epoch": 3.330662393162393, + "grad_norm": 0.8814072608947754, + "learning_rate": 1.3554187634804105e-05, + "loss": 0.787, + "step": 18708 + }, + { + "epoch": 3.330840455840456, + "grad_norm": 0.9193364977836609, + "learning_rate": 1.354715188473471e-05, + "loss": 0.7423, + "step": 18709 + }, + { + "epoch": 3.3310185185185186, + "grad_norm": 0.9405232667922974, + "learning_rate": 1.354011782852923e-05, + "loss": 0.8607, + "step": 18710 + }, + { + "epoch": 3.3311965811965814, + "grad_norm": 1.0526103973388672, + "learning_rate": 1.353308546632559e-05, + "loss": 0.8387, + "step": 18711 + }, + { + "epoch": 3.3313746438746437, + "grad_norm": 1.0627460479736328, + "learning_rate": 1.3526054798261468e-05, + "loss": 0.7533, + "step": 18712 + }, + { + "epoch": 3.3315527065527064, + "grad_norm": 0.9542255997657776, + "learning_rate": 1.3519025824474695e-05, + "loss": 0.8249, + "step": 18713 + }, + { + "epoch": 3.331730769230769, + "grad_norm": 0.8286230564117432, + "learning_rate": 1.3511998545102956e-05, + "loss": 0.6924, + "step": 18714 + }, + { + "epoch": 3.331908831908832, + "grad_norm": 1.0210462808609009, + "learning_rate": 1.3504972960283946e-05, + "loss": 0.7209, + "step": 18715 + }, + { + "epoch": 3.3320868945868947, + "grad_norm": 1.0881197452545166, + "learning_rate": 1.3497949070155313e-05, + "loss": 0.8403, + "step": 18716 + }, + { + "epoch": 3.3322649572649574, + "grad_norm": 0.8475449085235596, + "learning_rate": 1.349092687485468e-05, + "loss": 0.701, + "step": 18717 + }, + { + "epoch": 3.33244301994302, + "grad_norm": 0.814493715763092, + "learning_rate": 1.3483906374519607e-05, + "loss": 0.7209, + "step": 18718 + }, + { + "epoch": 3.3326210826210825, + "grad_norm": 0.9363925457000732, + "learning_rate": 1.3476887569287699e-05, + "loss": 0.7384, + "step": 18719 + }, + { + "epoch": 3.3327991452991452, + "grad_norm": 0.8990989327430725, + "learning_rate": 1.3469870459296408e-05, + "loss": 0.7414, + "step": 18720 + }, + { + "epoch": 3.332977207977208, + "grad_norm": 0.814422607421875, + "learning_rate": 1.3462855044683276e-05, + "loss": 0.7229, + "step": 18721 + }, + { + "epoch": 3.3331552706552707, + "grad_norm": 0.8876845240592957, + "learning_rate": 1.3455841325585727e-05, + "loss": 0.9162, + "step": 18722 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 0.9989238381385803, + "learning_rate": 1.3448829302141198e-05, + "loss": 0.74, + "step": 18723 + }, + { + "epoch": 3.333511396011396, + "grad_norm": 1.0466902256011963, + "learning_rate": 1.3441818974487064e-05, + "loss": 0.7875, + "step": 18724 + }, + { + "epoch": 3.3336894586894585, + "grad_norm": 0.9067308306694031, + "learning_rate": 1.343481034276065e-05, + "loss": 0.7372, + "step": 18725 + }, + { + "epoch": 3.3338675213675213, + "grad_norm": 0.9707076549530029, + "learning_rate": 1.3427803407099349e-05, + "loss": 0.7035, + "step": 18726 + }, + { + "epoch": 3.334045584045584, + "grad_norm": 1.0841397047042847, + "learning_rate": 1.3420798167640358e-05, + "loss": 0.6551, + "step": 18727 + }, + { + "epoch": 3.334223646723647, + "grad_norm": 0.9127151370048523, + "learning_rate": 1.3413794624520992e-05, + "loss": 0.8078, + "step": 18728 + }, + { + "epoch": 3.3344017094017095, + "grad_norm": 0.9369844198226929, + "learning_rate": 1.3406792777878463e-05, + "loss": 0.6668, + "step": 18729 + }, + { + "epoch": 3.3345797720797723, + "grad_norm": 0.9303719401359558, + "learning_rate": 1.3399792627849938e-05, + "loss": 0.7203, + "step": 18730 + }, + { + "epoch": 3.3347578347578346, + "grad_norm": 0.9815326929092407, + "learning_rate": 1.3392794174572588e-05, + "loss": 0.7766, + "step": 18731 + }, + { + "epoch": 3.3349358974358974, + "grad_norm": 0.8707610964775085, + "learning_rate": 1.338579741818352e-05, + "loss": 0.6492, + "step": 18732 + }, + { + "epoch": 3.33511396011396, + "grad_norm": 0.8969348669052124, + "learning_rate": 1.337880235881981e-05, + "loss": 0.6799, + "step": 18733 + }, + { + "epoch": 3.335292022792023, + "grad_norm": 1.079229474067688, + "learning_rate": 1.3371808996618562e-05, + "loss": 0.683, + "step": 18734 + }, + { + "epoch": 3.3354700854700856, + "grad_norm": 0.9601012468338013, + "learning_rate": 1.3364817331716728e-05, + "loss": 0.8354, + "step": 18735 + }, + { + "epoch": 3.335648148148148, + "grad_norm": 0.7610724568367004, + "learning_rate": 1.3357827364251364e-05, + "loss": 0.6091, + "step": 18736 + }, + { + "epoch": 3.3358262108262107, + "grad_norm": 1.0253962278366089, + "learning_rate": 1.3350839094359346e-05, + "loss": 0.7979, + "step": 18737 + }, + { + "epoch": 3.3360042735042734, + "grad_norm": 1.020058512687683, + "learning_rate": 1.3343852522177668e-05, + "loss": 0.8259, + "step": 18738 + }, + { + "epoch": 3.336182336182336, + "grad_norm": 0.7768040895462036, + "learning_rate": 1.3336867647843188e-05, + "loss": 0.4803, + "step": 18739 + }, + { + "epoch": 3.336360398860399, + "grad_norm": 0.8641179800033569, + "learning_rate": 1.3329884471492749e-05, + "loss": 0.6559, + "step": 18740 + }, + { + "epoch": 3.3365384615384617, + "grad_norm": 1.1274569034576416, + "learning_rate": 1.3322902993263197e-05, + "loss": 0.8297, + "step": 18741 + }, + { + "epoch": 3.3367165242165244, + "grad_norm": 0.86558598279953, + "learning_rate": 1.3315923213291292e-05, + "loss": 0.7778, + "step": 18742 + }, + { + "epoch": 3.3368945868945867, + "grad_norm": 1.0409115552902222, + "learning_rate": 1.330894513171378e-05, + "loss": 0.7114, + "step": 18743 + }, + { + "epoch": 3.3370726495726495, + "grad_norm": 0.9985263347625732, + "learning_rate": 1.3301968748667426e-05, + "loss": 0.6776, + "step": 18744 + }, + { + "epoch": 3.3372507122507122, + "grad_norm": 0.9003916382789612, + "learning_rate": 1.3294994064288902e-05, + "loss": 0.7261, + "step": 18745 + }, + { + "epoch": 3.337428774928775, + "grad_norm": 1.0112050771713257, + "learning_rate": 1.3288021078714851e-05, + "loss": 0.7276, + "step": 18746 + }, + { + "epoch": 3.3376068376068377, + "grad_norm": 0.9479689002037048, + "learning_rate": 1.3281049792081901e-05, + "loss": 0.8231, + "step": 18747 + }, + { + "epoch": 3.3377849002849, + "grad_norm": 0.9517938494682312, + "learning_rate": 1.327408020452663e-05, + "loss": 0.8923, + "step": 18748 + }, + { + "epoch": 3.337962962962963, + "grad_norm": 0.9328367710113525, + "learning_rate": 1.3267112316185603e-05, + "loss": 0.8031, + "step": 18749 + }, + { + "epoch": 3.3381410256410255, + "grad_norm": 1.142439842224121, + "learning_rate": 1.3260146127195317e-05, + "loss": 0.8177, + "step": 18750 + }, + { + "epoch": 3.3383190883190883, + "grad_norm": 0.9793458580970764, + "learning_rate": 1.3253181637692324e-05, + "loss": 0.6184, + "step": 18751 + }, + { + "epoch": 3.338497150997151, + "grad_norm": 1.0512577295303345, + "learning_rate": 1.3246218847812996e-05, + "loss": 0.7105, + "step": 18752 + }, + { + "epoch": 3.338675213675214, + "grad_norm": 0.9189818501472473, + "learning_rate": 1.3239257757693812e-05, + "loss": 0.8745, + "step": 18753 + }, + { + "epoch": 3.3388532763532766, + "grad_norm": 1.0795321464538574, + "learning_rate": 1.3232298367471141e-05, + "loss": 1.0447, + "step": 18754 + }, + { + "epoch": 3.339031339031339, + "grad_norm": 0.9658595323562622, + "learning_rate": 1.322534067728135e-05, + "loss": 0.9254, + "step": 18755 + }, + { + "epoch": 3.3392094017094016, + "grad_norm": 0.9447624087333679, + "learning_rate": 1.3218384687260743e-05, + "loss": 0.7847, + "step": 18756 + }, + { + "epoch": 3.3393874643874644, + "grad_norm": 1.0141079425811768, + "learning_rate": 1.321143039754561e-05, + "loss": 0.6128, + "step": 18757 + }, + { + "epoch": 3.339565527065527, + "grad_norm": 0.9425485730171204, + "learning_rate": 1.3204477808272187e-05, + "loss": 0.629, + "step": 18758 + }, + { + "epoch": 3.33974358974359, + "grad_norm": 1.0052008628845215, + "learning_rate": 1.3197526919576775e-05, + "loss": 0.8226, + "step": 18759 + }, + { + "epoch": 3.339921652421652, + "grad_norm": 0.8588734865188599, + "learning_rate": 1.3190577731595454e-05, + "loss": 0.6467, + "step": 18760 + }, + { + "epoch": 3.340099715099715, + "grad_norm": 0.8867504596710205, + "learning_rate": 1.3183630244464473e-05, + "loss": 0.8033, + "step": 18761 + }, + { + "epoch": 3.3402777777777777, + "grad_norm": 0.9576687812805176, + "learning_rate": 1.3176684458319877e-05, + "loss": 0.7776, + "step": 18762 + }, + { + "epoch": 3.3404558404558404, + "grad_norm": 0.9716742038726807, + "learning_rate": 1.3169740373297813e-05, + "loss": 0.9164, + "step": 18763 + }, + { + "epoch": 3.340633903133903, + "grad_norm": 0.9201170206069946, + "learning_rate": 1.3162797989534303e-05, + "loss": 0.7159, + "step": 18764 + }, + { + "epoch": 3.340811965811966, + "grad_norm": 0.8414926528930664, + "learning_rate": 1.3155857307165354e-05, + "loss": 0.6775, + "step": 18765 + }, + { + "epoch": 3.3409900284900287, + "grad_norm": 1.06947922706604, + "learning_rate": 1.314891832632703e-05, + "loss": 0.7879, + "step": 18766 + }, + { + "epoch": 3.341168091168091, + "grad_norm": 1.0419055223464966, + "learning_rate": 1.3141981047155183e-05, + "loss": 0.849, + "step": 18767 + }, + { + "epoch": 3.3413461538461537, + "grad_norm": 0.9134105443954468, + "learning_rate": 1.3135045469785811e-05, + "loss": 0.711, + "step": 18768 + }, + { + "epoch": 3.3415242165242165, + "grad_norm": 0.9776987433433533, + "learning_rate": 1.3128111594354775e-05, + "loss": 0.8041, + "step": 18769 + }, + { + "epoch": 3.3417022792022792, + "grad_norm": 0.9296642541885376, + "learning_rate": 1.3121179420997941e-05, + "loss": 0.7348, + "step": 18770 + }, + { + "epoch": 3.341880341880342, + "grad_norm": 0.9305540323257446, + "learning_rate": 1.3114248949851116e-05, + "loss": 0.6617, + "step": 18771 + }, + { + "epoch": 3.3420584045584047, + "grad_norm": 1.1082522869110107, + "learning_rate": 1.3107320181050086e-05, + "loss": 0.6465, + "step": 18772 + }, + { + "epoch": 3.342236467236467, + "grad_norm": 1.0075236558914185, + "learning_rate": 1.3100393114730614e-05, + "loss": 0.8232, + "step": 18773 + }, + { + "epoch": 3.34241452991453, + "grad_norm": 0.9493429064750671, + "learning_rate": 1.3093467751028433e-05, + "loss": 0.6698, + "step": 18774 + }, + { + "epoch": 3.3425925925925926, + "grad_norm": 0.8520618677139282, + "learning_rate": 1.3086544090079179e-05, + "loss": 0.7738, + "step": 18775 + }, + { + "epoch": 3.3427706552706553, + "grad_norm": 1.036331295967102, + "learning_rate": 1.3079622132018599e-05, + "loss": 0.6765, + "step": 18776 + }, + { + "epoch": 3.342948717948718, + "grad_norm": 1.008399486541748, + "learning_rate": 1.3072701876982218e-05, + "loss": 0.7976, + "step": 18777 + }, + { + "epoch": 3.343126780626781, + "grad_norm": 1.049607515335083, + "learning_rate": 1.3065783325105695e-05, + "loss": 0.8794, + "step": 18778 + }, + { + "epoch": 3.343304843304843, + "grad_norm": 0.954903244972229, + "learning_rate": 1.3058866476524556e-05, + "loss": 0.8595, + "step": 18779 + }, + { + "epoch": 3.343482905982906, + "grad_norm": 0.9020154476165771, + "learning_rate": 1.3051951331374323e-05, + "loss": 0.7465, + "step": 18780 + }, + { + "epoch": 3.3436609686609686, + "grad_norm": 1.0125290155410767, + "learning_rate": 1.3045037889790478e-05, + "loss": 1.0211, + "step": 18781 + }, + { + "epoch": 3.3438390313390314, + "grad_norm": 1.1179999113082886, + "learning_rate": 1.3038126151908492e-05, + "loss": 0.7702, + "step": 18782 + }, + { + "epoch": 3.344017094017094, + "grad_norm": 0.8777053952217102, + "learning_rate": 1.3031216117863764e-05, + "loss": 0.6477, + "step": 18783 + }, + { + "epoch": 3.344195156695157, + "grad_norm": 1.083327054977417, + "learning_rate": 1.3024307787791679e-05, + "loss": 0.9073, + "step": 18784 + }, + { + "epoch": 3.344373219373219, + "grad_norm": 1.0102063417434692, + "learning_rate": 1.3017401161827636e-05, + "loss": 0.7956, + "step": 18785 + }, + { + "epoch": 3.344551282051282, + "grad_norm": 0.8271292448043823, + "learning_rate": 1.3010496240106918e-05, + "loss": 0.7027, + "step": 18786 + }, + { + "epoch": 3.3447293447293447, + "grad_norm": 1.0592424869537354, + "learning_rate": 1.3003593022764826e-05, + "loss": 0.8079, + "step": 18787 + }, + { + "epoch": 3.3449074074074074, + "grad_norm": 1.0229346752166748, + "learning_rate": 1.2996691509936598e-05, + "loss": 0.8416, + "step": 18788 + }, + { + "epoch": 3.34508547008547, + "grad_norm": 1.0877922773361206, + "learning_rate": 1.298979170175748e-05, + "loss": 0.793, + "step": 18789 + }, + { + "epoch": 3.345263532763533, + "grad_norm": 0.9380156397819519, + "learning_rate": 1.298289359836261e-05, + "loss": 0.6552, + "step": 18790 + }, + { + "epoch": 3.3454415954415953, + "grad_norm": 0.9297496676445007, + "learning_rate": 1.2975997199887235e-05, + "loss": 0.8337, + "step": 18791 + }, + { + "epoch": 3.345619658119658, + "grad_norm": 0.9621930718421936, + "learning_rate": 1.296910250646637e-05, + "loss": 0.7204, + "step": 18792 + }, + { + "epoch": 3.3457977207977208, + "grad_norm": 0.847952663898468, + "learning_rate": 1.296220951823517e-05, + "loss": 0.6411, + "step": 18793 + }, + { + "epoch": 3.3459757834757835, + "grad_norm": 1.0718971490859985, + "learning_rate": 1.2955318235328672e-05, + "loss": 0.7651, + "step": 18794 + }, + { + "epoch": 3.3461538461538463, + "grad_norm": 0.9610202312469482, + "learning_rate": 1.2948428657881884e-05, + "loss": 0.884, + "step": 18795 + }, + { + "epoch": 3.346331908831909, + "grad_norm": 0.9173485040664673, + "learning_rate": 1.2941540786029815e-05, + "loss": 0.736, + "step": 18796 + }, + { + "epoch": 3.3465099715099713, + "grad_norm": 1.0938438177108765, + "learning_rate": 1.2934654619907404e-05, + "loss": 0.8351, + "step": 18797 + }, + { + "epoch": 3.346688034188034, + "grad_norm": 1.040371060371399, + "learning_rate": 1.2927770159649566e-05, + "loss": 0.7698, + "step": 18798 + }, + { + "epoch": 3.346866096866097, + "grad_norm": 1.0388879776000977, + "learning_rate": 1.2920887405391202e-05, + "loss": 0.7521, + "step": 18799 + }, + { + "epoch": 3.3470441595441596, + "grad_norm": 1.1408355236053467, + "learning_rate": 1.2914006357267128e-05, + "loss": 0.6752, + "step": 18800 + }, + { + "epoch": 3.3472222222222223, + "grad_norm": 0.9846867322921753, + "learning_rate": 1.2907127015412247e-05, + "loss": 0.7124, + "step": 18801 + }, + { + "epoch": 3.347400284900285, + "grad_norm": 0.9047761559486389, + "learning_rate": 1.2900249379961238e-05, + "loss": 0.9022, + "step": 18802 + }, + { + "epoch": 3.347578347578348, + "grad_norm": 1.1381040811538696, + "learning_rate": 1.289337345104894e-05, + "loss": 0.8264, + "step": 18803 + }, + { + "epoch": 3.34775641025641, + "grad_norm": 0.8845169544219971, + "learning_rate": 1.2886499228810045e-05, + "loss": 0.7129, + "step": 18804 + }, + { + "epoch": 3.347934472934473, + "grad_norm": 0.8322994709014893, + "learning_rate": 1.28796267133792e-05, + "loss": 0.8143, + "step": 18805 + }, + { + "epoch": 3.3481125356125356, + "grad_norm": 0.9626625776290894, + "learning_rate": 1.2872755904891142e-05, + "loss": 0.747, + "step": 18806 + }, + { + "epoch": 3.3482905982905984, + "grad_norm": 0.8659831285476685, + "learning_rate": 1.2865886803480399e-05, + "loss": 0.698, + "step": 18807 + }, + { + "epoch": 3.348468660968661, + "grad_norm": 1.006449580192566, + "learning_rate": 1.2859019409281636e-05, + "loss": 0.9115, + "step": 18808 + }, + { + "epoch": 3.3486467236467234, + "grad_norm": 0.9104591012001038, + "learning_rate": 1.2852153722429327e-05, + "loss": 0.5964, + "step": 18809 + }, + { + "epoch": 3.348824786324786, + "grad_norm": 0.9254124760627747, + "learning_rate": 1.2845289743058064e-05, + "loss": 0.784, + "step": 18810 + }, + { + "epoch": 3.349002849002849, + "grad_norm": 0.9335882663726807, + "learning_rate": 1.2838427471302284e-05, + "loss": 0.9993, + "step": 18811 + }, + { + "epoch": 3.3491809116809117, + "grad_norm": 1.064700961112976, + "learning_rate": 1.2831566907296466e-05, + "loss": 0.7263, + "step": 18812 + }, + { + "epoch": 3.3493589743589745, + "grad_norm": 1.071642279624939, + "learning_rate": 1.2824708051175016e-05, + "loss": 0.8929, + "step": 18813 + }, + { + "epoch": 3.349537037037037, + "grad_norm": 0.8615704774856567, + "learning_rate": 1.2817850903072315e-05, + "loss": 0.8469, + "step": 18814 + }, + { + "epoch": 3.3497150997151, + "grad_norm": 0.9117037057876587, + "learning_rate": 1.281099546312271e-05, + "loss": 0.8622, + "step": 18815 + }, + { + "epoch": 3.3498931623931623, + "grad_norm": 1.1236242055892944, + "learning_rate": 1.2804141731460562e-05, + "loss": 0.7232, + "step": 18816 + }, + { + "epoch": 3.350071225071225, + "grad_norm": 0.7709314823150635, + "learning_rate": 1.2797289708220084e-05, + "loss": 0.4792, + "step": 18817 + }, + { + "epoch": 3.3502492877492878, + "grad_norm": 0.9297556281089783, + "learning_rate": 1.2790439393535614e-05, + "loss": 0.6442, + "step": 18818 + }, + { + "epoch": 3.3504273504273505, + "grad_norm": 0.9198339581489563, + "learning_rate": 1.2783590787541266e-05, + "loss": 0.7885, + "step": 18819 + }, + { + "epoch": 3.3506054131054133, + "grad_norm": 0.9080191850662231, + "learning_rate": 1.2776743890371312e-05, + "loss": 0.6749, + "step": 18820 + }, + { + "epoch": 3.3507834757834756, + "grad_norm": 0.969227135181427, + "learning_rate": 1.2769898702159867e-05, + "loss": 0.6469, + "step": 18821 + }, + { + "epoch": 3.3509615384615383, + "grad_norm": 0.9139270186424255, + "learning_rate": 1.2763055223041055e-05, + "loss": 0.7155, + "step": 18822 + }, + { + "epoch": 3.351139601139601, + "grad_norm": 0.7823638916015625, + "learning_rate": 1.2756213453148958e-05, + "loss": 0.5529, + "step": 18823 + }, + { + "epoch": 3.351317663817664, + "grad_norm": 0.9755151867866516, + "learning_rate": 1.2749373392617603e-05, + "loss": 0.7031, + "step": 18824 + }, + { + "epoch": 3.3514957264957266, + "grad_norm": 0.9338952302932739, + "learning_rate": 1.274253504158105e-05, + "loss": 0.771, + "step": 18825 + }, + { + "epoch": 3.3516737891737893, + "grad_norm": 0.9785382747650146, + "learning_rate": 1.2735698400173257e-05, + "loss": 0.8773, + "step": 18826 + }, + { + "epoch": 3.351851851851852, + "grad_norm": 1.0508654117584229, + "learning_rate": 1.2728863468528174e-05, + "loss": 0.9981, + "step": 18827 + }, + { + "epoch": 3.3520299145299144, + "grad_norm": 1.0220104455947876, + "learning_rate": 1.2722030246779737e-05, + "loss": 0.8708, + "step": 18828 + }, + { + "epoch": 3.352207977207977, + "grad_norm": 0.8605777025222778, + "learning_rate": 1.2715198735061794e-05, + "loss": 0.703, + "step": 18829 + }, + { + "epoch": 3.35238603988604, + "grad_norm": 1.0802946090698242, + "learning_rate": 1.2708368933508207e-05, + "loss": 0.758, + "step": 18830 + }, + { + "epoch": 3.3525641025641026, + "grad_norm": 0.9684296250343323, + "learning_rate": 1.2701540842252835e-05, + "loss": 0.6264, + "step": 18831 + }, + { + "epoch": 3.3527421652421654, + "grad_norm": 0.8688775897026062, + "learning_rate": 1.269471446142938e-05, + "loss": 0.7257, + "step": 18832 + }, + { + "epoch": 3.3529202279202277, + "grad_norm": 0.9325960278511047, + "learning_rate": 1.2687889791171681e-05, + "loss": 0.6823, + "step": 18833 + }, + { + "epoch": 3.3530982905982905, + "grad_norm": 0.9034467935562134, + "learning_rate": 1.2681066831613365e-05, + "loss": 0.7568, + "step": 18834 + }, + { + "epoch": 3.353276353276353, + "grad_norm": 0.8753020763397217, + "learning_rate": 1.2674245582888167e-05, + "loss": 0.6769, + "step": 18835 + }, + { + "epoch": 3.353454415954416, + "grad_norm": 1.071382761001587, + "learning_rate": 1.2667426045129727e-05, + "loss": 0.6557, + "step": 18836 + }, + { + "epoch": 3.3536324786324787, + "grad_norm": 0.9410926699638367, + "learning_rate": 1.266060821847166e-05, + "loss": 0.7636, + "step": 18837 + }, + { + "epoch": 3.3538105413105415, + "grad_norm": 0.9422865509986877, + "learning_rate": 1.2653792103047535e-05, + "loss": 0.7199, + "step": 18838 + }, + { + "epoch": 3.353988603988604, + "grad_norm": 0.9728864431381226, + "learning_rate": 1.2646977698990914e-05, + "loss": 0.7725, + "step": 18839 + }, + { + "epoch": 3.3541666666666665, + "grad_norm": 0.8972764611244202, + "learning_rate": 1.264016500643529e-05, + "loss": 0.8581, + "step": 18840 + }, + { + "epoch": 3.3543447293447293, + "grad_norm": 0.9951597452163696, + "learning_rate": 1.2633354025514188e-05, + "loss": 0.7828, + "step": 18841 + }, + { + "epoch": 3.354522792022792, + "grad_norm": 0.8521020412445068, + "learning_rate": 1.2626544756360991e-05, + "loss": 0.7818, + "step": 18842 + }, + { + "epoch": 3.3547008547008548, + "grad_norm": 0.9301114082336426, + "learning_rate": 1.261973719910916e-05, + "loss": 0.6921, + "step": 18843 + }, + { + "epoch": 3.3548789173789175, + "grad_norm": 1.0106427669525146, + "learning_rate": 1.2612931353892077e-05, + "loss": 0.6784, + "step": 18844 + }, + { + "epoch": 3.35505698005698, + "grad_norm": 0.9824902415275574, + "learning_rate": 1.2606127220843057e-05, + "loss": 0.8435, + "step": 18845 + }, + { + "epoch": 3.3552350427350426, + "grad_norm": 1.5545586347579956, + "learning_rate": 1.2599324800095435e-05, + "loss": 0.9226, + "step": 18846 + }, + { + "epoch": 3.3554131054131053, + "grad_norm": 1.0093287229537964, + "learning_rate": 1.2592524091782465e-05, + "loss": 0.7677, + "step": 18847 + }, + { + "epoch": 3.355591168091168, + "grad_norm": 0.8863753080368042, + "learning_rate": 1.2585725096037448e-05, + "loss": 0.7108, + "step": 18848 + }, + { + "epoch": 3.355769230769231, + "grad_norm": 0.880733072757721, + "learning_rate": 1.2578927812993524e-05, + "loss": 0.6485, + "step": 18849 + }, + { + "epoch": 3.3559472934472936, + "grad_norm": 0.8626177310943604, + "learning_rate": 1.2572132242783929e-05, + "loss": 0.6838, + "step": 18850 + }, + { + "epoch": 3.3561253561253563, + "grad_norm": 0.9541394114494324, + "learning_rate": 1.2565338385541792e-05, + "loss": 0.6914, + "step": 18851 + }, + { + "epoch": 3.3563034188034186, + "grad_norm": 0.933340311050415, + "learning_rate": 1.2558546241400215e-05, + "loss": 0.719, + "step": 18852 + }, + { + "epoch": 3.3564814814814814, + "grad_norm": 0.8834456205368042, + "learning_rate": 1.2551755810492272e-05, + "loss": 0.7702, + "step": 18853 + }, + { + "epoch": 3.356659544159544, + "grad_norm": 0.845119297504425, + "learning_rate": 1.2544967092951031e-05, + "loss": 0.6369, + "step": 18854 + }, + { + "epoch": 3.356837606837607, + "grad_norm": 0.9516612887382507, + "learning_rate": 1.2538180088909479e-05, + "loss": 0.6735, + "step": 18855 + }, + { + "epoch": 3.3570156695156697, + "grad_norm": 0.8660346865653992, + "learning_rate": 1.2531394798500606e-05, + "loss": 0.6033, + "step": 18856 + }, + { + "epoch": 3.357193732193732, + "grad_norm": 1.4815114736557007, + "learning_rate": 1.2524611221857318e-05, + "loss": 1.1411, + "step": 18857 + }, + { + "epoch": 3.3573717948717947, + "grad_norm": 0.9433855414390564, + "learning_rate": 1.251782935911262e-05, + "loss": 0.7562, + "step": 18858 + }, + { + "epoch": 3.3575498575498575, + "grad_norm": 0.8575382828712463, + "learning_rate": 1.2511049210399272e-05, + "loss": 0.7071, + "step": 18859 + }, + { + "epoch": 3.35772792022792, + "grad_norm": 1.0027903318405151, + "learning_rate": 1.250427077585019e-05, + "loss": 0.6572, + "step": 18860 + }, + { + "epoch": 3.357905982905983, + "grad_norm": 1.028286099433899, + "learning_rate": 1.2497494055598181e-05, + "loss": 0.7687, + "step": 18861 + }, + { + "epoch": 3.3580840455840457, + "grad_norm": 0.9715962409973145, + "learning_rate": 1.249071904977599e-05, + "loss": 0.6205, + "step": 18862 + }, + { + "epoch": 3.3582621082621085, + "grad_norm": 1.0001580715179443, + "learning_rate": 1.248394575851638e-05, + "loss": 0.6317, + "step": 18863 + }, + { + "epoch": 3.3584401709401708, + "grad_norm": 0.8842644691467285, + "learning_rate": 1.2477174181952034e-05, + "loss": 0.6289, + "step": 18864 + }, + { + "epoch": 3.3586182336182335, + "grad_norm": 0.9407495260238647, + "learning_rate": 1.2470404320215667e-05, + "loss": 0.6886, + "step": 18865 + }, + { + "epoch": 3.3587962962962963, + "grad_norm": 0.9713712930679321, + "learning_rate": 1.2463636173439908e-05, + "loss": 0.6181, + "step": 18866 + }, + { + "epoch": 3.358974358974359, + "grad_norm": 1.070308804512024, + "learning_rate": 1.245686974175735e-05, + "loss": 0.9199, + "step": 18867 + }, + { + "epoch": 3.359152421652422, + "grad_norm": 0.9229024648666382, + "learning_rate": 1.2450105025300574e-05, + "loss": 0.8533, + "step": 18868 + }, + { + "epoch": 3.359330484330484, + "grad_norm": 0.8956949710845947, + "learning_rate": 1.244334202420212e-05, + "loss": 0.731, + "step": 18869 + }, + { + "epoch": 3.359508547008547, + "grad_norm": 0.8161178231239319, + "learning_rate": 1.2436580738594494e-05, + "loss": 0.592, + "step": 18870 + }, + { + "epoch": 3.3596866096866096, + "grad_norm": 0.8732914328575134, + "learning_rate": 1.2429821168610179e-05, + "loss": 0.7275, + "step": 18871 + }, + { + "epoch": 3.3598646723646723, + "grad_norm": 1.0021698474884033, + "learning_rate": 1.2423063314381578e-05, + "loss": 0.8745, + "step": 18872 + }, + { + "epoch": 3.360042735042735, + "grad_norm": 0.9891536235809326, + "learning_rate": 1.2416307176041176e-05, + "loss": 0.7716, + "step": 18873 + }, + { + "epoch": 3.360220797720798, + "grad_norm": 1.0069750547409058, + "learning_rate": 1.2409552753721254e-05, + "loss": 0.91, + "step": 18874 + }, + { + "epoch": 3.3603988603988606, + "grad_norm": 1.0723732709884644, + "learning_rate": 1.2402800047554208e-05, + "loss": 0.9909, + "step": 18875 + }, + { + "epoch": 3.360576923076923, + "grad_norm": 0.9411680698394775, + "learning_rate": 1.239604905767232e-05, + "loss": 0.7607, + "step": 18876 + }, + { + "epoch": 3.3607549857549857, + "grad_norm": 1.0347760915756226, + "learning_rate": 1.2389299784207886e-05, + "loss": 0.8984, + "step": 18877 + }, + { + "epoch": 3.3609330484330484, + "grad_norm": 1.0429126024246216, + "learning_rate": 1.238255222729311e-05, + "loss": 0.8522, + "step": 18878 + }, + { + "epoch": 3.361111111111111, + "grad_norm": 0.938989520072937, + "learning_rate": 1.2375806387060229e-05, + "loss": 0.7692, + "step": 18879 + }, + { + "epoch": 3.361289173789174, + "grad_norm": 0.8867523074150085, + "learning_rate": 1.2369062263641384e-05, + "loss": 0.6354, + "step": 18880 + }, + { + "epoch": 3.361467236467236, + "grad_norm": 0.9158368110656738, + "learning_rate": 1.2362319857168735e-05, + "loss": 0.7264, + "step": 18881 + }, + { + "epoch": 3.361645299145299, + "grad_norm": 0.9424712061882019, + "learning_rate": 1.2355579167774355e-05, + "loss": 0.7719, + "step": 18882 + }, + { + "epoch": 3.3618233618233617, + "grad_norm": 1.001279592514038, + "learning_rate": 1.234884019559036e-05, + "loss": 0.8973, + "step": 18883 + }, + { + "epoch": 3.3620014245014245, + "grad_norm": 0.9505113959312439, + "learning_rate": 1.2342102940748756e-05, + "loss": 0.6743, + "step": 18884 + }, + { + "epoch": 3.3621794871794872, + "grad_norm": 1.0694880485534668, + "learning_rate": 1.2335367403381559e-05, + "loss": 0.7723, + "step": 18885 + }, + { + "epoch": 3.36235754985755, + "grad_norm": 0.8937603235244751, + "learning_rate": 1.232863358362072e-05, + "loss": 0.7262, + "step": 18886 + }, + { + "epoch": 3.3625356125356127, + "grad_norm": 0.9782434105873108, + "learning_rate": 1.2321901481598174e-05, + "loss": 0.8082, + "step": 18887 + }, + { + "epoch": 3.362713675213675, + "grad_norm": 1.0660442113876343, + "learning_rate": 1.2315171097445866e-05, + "loss": 0.8608, + "step": 18888 + }, + { + "epoch": 3.362891737891738, + "grad_norm": 0.9985655546188354, + "learning_rate": 1.2308442431295598e-05, + "loss": 0.8218, + "step": 18889 + }, + { + "epoch": 3.3630698005698005, + "grad_norm": 0.9028006196022034, + "learning_rate": 1.2301715483279275e-05, + "loss": 0.5749, + "step": 18890 + }, + { + "epoch": 3.3632478632478633, + "grad_norm": 0.9652066826820374, + "learning_rate": 1.2294990253528616e-05, + "loss": 0.7315, + "step": 18891 + }, + { + "epoch": 3.363425925925926, + "grad_norm": 0.8819321990013123, + "learning_rate": 1.2288266742175446e-05, + "loss": 0.6078, + "step": 18892 + }, + { + "epoch": 3.363603988603989, + "grad_norm": 0.8945509791374207, + "learning_rate": 1.2281544949351498e-05, + "loss": 0.756, + "step": 18893 + }, + { + "epoch": 3.363782051282051, + "grad_norm": 0.9183369278907776, + "learning_rate": 1.2274824875188452e-05, + "loss": 0.7717, + "step": 18894 + }, + { + "epoch": 3.363960113960114, + "grad_norm": 1.0190612077713013, + "learning_rate": 1.2268106519817969e-05, + "loss": 0.7269, + "step": 18895 + }, + { + "epoch": 3.3641381766381766, + "grad_norm": 0.8803049325942993, + "learning_rate": 1.2261389883371698e-05, + "loss": 0.6672, + "step": 18896 + }, + { + "epoch": 3.3643162393162394, + "grad_norm": 0.8707033395767212, + "learning_rate": 1.2254674965981217e-05, + "loss": 0.7234, + "step": 18897 + }, + { + "epoch": 3.364494301994302, + "grad_norm": 0.9297593235969543, + "learning_rate": 1.2247961767778138e-05, + "loss": 0.8101, + "step": 18898 + }, + { + "epoch": 3.364672364672365, + "grad_norm": 0.8042552471160889, + "learning_rate": 1.2241250288893925e-05, + "loss": 0.5784, + "step": 18899 + }, + { + "epoch": 3.364850427350427, + "grad_norm": 1.1243376731872559, + "learning_rate": 1.2234540529460126e-05, + "loss": 0.9558, + "step": 18900 + }, + { + "epoch": 3.36502849002849, + "grad_norm": 1.0456633567810059, + "learning_rate": 1.2227832489608194e-05, + "loss": 0.7718, + "step": 18901 + }, + { + "epoch": 3.3652065527065527, + "grad_norm": 0.9715479016304016, + "learning_rate": 1.2221126169469543e-05, + "loss": 0.8759, + "step": 18902 + }, + { + "epoch": 3.3653846153846154, + "grad_norm": 1.194477915763855, + "learning_rate": 1.2214421569175583e-05, + "loss": 0.8068, + "step": 18903 + }, + { + "epoch": 3.365562678062678, + "grad_norm": 0.8747355341911316, + "learning_rate": 1.2207718688857662e-05, + "loss": 0.5554, + "step": 18904 + }, + { + "epoch": 3.365740740740741, + "grad_norm": 0.9207034111022949, + "learning_rate": 1.2201017528647152e-05, + "loss": 1.0054, + "step": 18905 + }, + { + "epoch": 3.3659188034188032, + "grad_norm": 0.8625784516334534, + "learning_rate": 1.2194318088675283e-05, + "loss": 0.5582, + "step": 18906 + }, + { + "epoch": 3.366096866096866, + "grad_norm": 0.9648886322975159, + "learning_rate": 1.2187620369073361e-05, + "loss": 0.7067, + "step": 18907 + }, + { + "epoch": 3.3662749287749287, + "grad_norm": 0.8335103988647461, + "learning_rate": 1.2180924369972614e-05, + "loss": 0.7857, + "step": 18908 + }, + { + "epoch": 3.3664529914529915, + "grad_norm": 1.0891221761703491, + "learning_rate": 1.2174230091504214e-05, + "loss": 0.8173, + "step": 18909 + }, + { + "epoch": 3.3666310541310542, + "grad_norm": 0.9425504207611084, + "learning_rate": 1.2167537533799345e-05, + "loss": 0.6784, + "step": 18910 + }, + { + "epoch": 3.366809116809117, + "grad_norm": 0.8152710199356079, + "learning_rate": 1.2160846696989114e-05, + "loss": 0.8011, + "step": 18911 + }, + { + "epoch": 3.3669871794871793, + "grad_norm": 1.0064387321472168, + "learning_rate": 1.2154157581204607e-05, + "loss": 0.7862, + "step": 18912 + }, + { + "epoch": 3.367165242165242, + "grad_norm": 0.9653602838516235, + "learning_rate": 1.2147470186576936e-05, + "loss": 0.7156, + "step": 18913 + }, + { + "epoch": 3.367343304843305, + "grad_norm": 1.0241694450378418, + "learning_rate": 1.2140784513237046e-05, + "loss": 0.6863, + "step": 18914 + }, + { + "epoch": 3.3675213675213675, + "grad_norm": 0.8816869258880615, + "learning_rate": 1.213410056131602e-05, + "loss": 0.6623, + "step": 18915 + }, + { + "epoch": 3.3676994301994303, + "grad_norm": 0.9364701509475708, + "learning_rate": 1.2127418330944718e-05, + "loss": 0.6271, + "step": 18916 + }, + { + "epoch": 3.367877492877493, + "grad_norm": 1.0039292573928833, + "learning_rate": 1.2120737822254147e-05, + "loss": 0.7901, + "step": 18917 + }, + { + "epoch": 3.3680555555555554, + "grad_norm": 1.2018239498138428, + "learning_rate": 1.2114059035375158e-05, + "loss": 0.8665, + "step": 18918 + }, + { + "epoch": 3.368233618233618, + "grad_norm": 0.9742374420166016, + "learning_rate": 1.2107381970438614e-05, + "loss": 0.7912, + "step": 18919 + }, + { + "epoch": 3.368411680911681, + "grad_norm": 0.8076589703559875, + "learning_rate": 1.210070662757533e-05, + "loss": 0.5589, + "step": 18920 + }, + { + "epoch": 3.3685897435897436, + "grad_norm": 0.8905361890792847, + "learning_rate": 1.2094033006916117e-05, + "loss": 0.6812, + "step": 18921 + }, + { + "epoch": 3.3687678062678064, + "grad_norm": 0.7845730781555176, + "learning_rate": 1.20873611085917e-05, + "loss": 0.4487, + "step": 18922 + }, + { + "epoch": 3.368945868945869, + "grad_norm": 0.9479206204414368, + "learning_rate": 1.2080690932732829e-05, + "loss": 0.7815, + "step": 18923 + }, + { + "epoch": 3.369123931623932, + "grad_norm": 1.0130283832550049, + "learning_rate": 1.2074022479470193e-05, + "loss": 0.644, + "step": 18924 + }, + { + "epoch": 3.369301994301994, + "grad_norm": 0.9899544715881348, + "learning_rate": 1.2067355748934429e-05, + "loss": 0.8751, + "step": 18925 + }, + { + "epoch": 3.369480056980057, + "grad_norm": 1.0933879613876343, + "learning_rate": 1.2060690741256165e-05, + "loss": 0.7778, + "step": 18926 + }, + { + "epoch": 3.3696581196581197, + "grad_norm": 0.9882935881614685, + "learning_rate": 1.2054027456565997e-05, + "loss": 0.6674, + "step": 18927 + }, + { + "epoch": 3.3698361823361824, + "grad_norm": 0.9455338716506958, + "learning_rate": 1.2047365894994455e-05, + "loss": 0.6675, + "step": 18928 + }, + { + "epoch": 3.370014245014245, + "grad_norm": 0.9364031553268433, + "learning_rate": 1.2040706056672069e-05, + "loss": 0.9199, + "step": 18929 + }, + { + "epoch": 3.3701923076923075, + "grad_norm": 1.004025936126709, + "learning_rate": 1.2034047941729354e-05, + "loss": 0.8102, + "step": 18930 + }, + { + "epoch": 3.3703703703703702, + "grad_norm": 0.9347146153450012, + "learning_rate": 1.2027391550296707e-05, + "loss": 0.7639, + "step": 18931 + }, + { + "epoch": 3.370548433048433, + "grad_norm": 0.9768404960632324, + "learning_rate": 1.202073688250458e-05, + "loss": 0.7367, + "step": 18932 + }, + { + "epoch": 3.3707264957264957, + "grad_norm": 1.0199530124664307, + "learning_rate": 1.2014083938483367e-05, + "loss": 0.7956, + "step": 18933 + }, + { + "epoch": 3.3709045584045585, + "grad_norm": 0.8989109396934509, + "learning_rate": 1.2007432718363398e-05, + "loss": 0.6288, + "step": 18934 + }, + { + "epoch": 3.3710826210826212, + "grad_norm": 1.0402858257293701, + "learning_rate": 1.2000783222275003e-05, + "loss": 0.8463, + "step": 18935 + }, + { + "epoch": 3.371260683760684, + "grad_norm": 0.8559962511062622, + "learning_rate": 1.1994135450348453e-05, + "loss": 0.9195, + "step": 18936 + }, + { + "epoch": 3.3714387464387463, + "grad_norm": 0.8805050849914551, + "learning_rate": 1.1987489402713981e-05, + "loss": 0.7653, + "step": 18937 + }, + { + "epoch": 3.371616809116809, + "grad_norm": 0.8654996752738953, + "learning_rate": 1.1980845079501867e-05, + "loss": 0.7037, + "step": 18938 + }, + { + "epoch": 3.371794871794872, + "grad_norm": 0.8487333059310913, + "learning_rate": 1.19742024808422e-05, + "loss": 0.823, + "step": 18939 + }, + { + "epoch": 3.3719729344729346, + "grad_norm": 0.9701769948005676, + "learning_rate": 1.196756160686523e-05, + "loss": 0.64, + "step": 18940 + }, + { + "epoch": 3.3721509971509973, + "grad_norm": 1.0142244100570679, + "learning_rate": 1.1960922457700975e-05, + "loss": 1.0056, + "step": 18941 + }, + { + "epoch": 3.3723290598290596, + "grad_norm": 0.9956945776939392, + "learning_rate": 1.1954285033479573e-05, + "loss": 0.6369, + "step": 18942 + }, + { + "epoch": 3.3725071225071224, + "grad_norm": 0.9598103165626526, + "learning_rate": 1.1947649334331068e-05, + "loss": 0.7008, + "step": 18943 + }, + { + "epoch": 3.372685185185185, + "grad_norm": 0.9213023781776428, + "learning_rate": 1.1941015360385432e-05, + "loss": 0.5899, + "step": 18944 + }, + { + "epoch": 3.372863247863248, + "grad_norm": 0.9229640960693359, + "learning_rate": 1.1934383111772718e-05, + "loss": 0.7408, + "step": 18945 + }, + { + "epoch": 3.3730413105413106, + "grad_norm": 1.0240286588668823, + "learning_rate": 1.1927752588622787e-05, + "loss": 0.6822, + "step": 18946 + }, + { + "epoch": 3.3732193732193734, + "grad_norm": 0.9154404401779175, + "learning_rate": 1.1921123791065602e-05, + "loss": 0.7995, + "step": 18947 + }, + { + "epoch": 3.373397435897436, + "grad_norm": 1.0823782682418823, + "learning_rate": 1.1914496719231039e-05, + "loss": 0.8212, + "step": 18948 + }, + { + "epoch": 3.3735754985754984, + "grad_norm": 1.005828857421875, + "learning_rate": 1.1907871373248925e-05, + "loss": 0.7916, + "step": 18949 + }, + { + "epoch": 3.373753561253561, + "grad_norm": 0.9120662212371826, + "learning_rate": 1.1901247753249079e-05, + "loss": 0.635, + "step": 18950 + }, + { + "epoch": 3.373931623931624, + "grad_norm": 1.006744146347046, + "learning_rate": 1.1894625859361274e-05, + "loss": 0.824, + "step": 18951 + }, + { + "epoch": 3.3741096866096867, + "grad_norm": 0.9716039299964905, + "learning_rate": 1.1888005691715254e-05, + "loss": 0.7036, + "step": 18952 + }, + { + "epoch": 3.3742877492877494, + "grad_norm": 1.0244479179382324, + "learning_rate": 1.1881387250440723e-05, + "loss": 0.7777, + "step": 18953 + }, + { + "epoch": 3.3744658119658117, + "grad_norm": 0.9912460446357727, + "learning_rate": 1.1874770535667334e-05, + "loss": 0.7136, + "step": 18954 + }, + { + "epoch": 3.3746438746438745, + "grad_norm": 1.1135791540145874, + "learning_rate": 1.1868155547524795e-05, + "loss": 0.846, + "step": 18955 + }, + { + "epoch": 3.3748219373219372, + "grad_norm": 0.8404874801635742, + "learning_rate": 1.1861542286142635e-05, + "loss": 0.6614, + "step": 18956 + }, + { + "epoch": 3.375, + "grad_norm": 0.8777139186859131, + "learning_rate": 1.1854930751650472e-05, + "loss": 0.614, + "step": 18957 + }, + { + "epoch": 3.3751780626780628, + "grad_norm": 1.0814663171768188, + "learning_rate": 1.1848320944177837e-05, + "loss": 0.8726, + "step": 18958 + }, + { + "epoch": 3.3753561253561255, + "grad_norm": 1.140494704246521, + "learning_rate": 1.1841712863854249e-05, + "loss": 0.7207, + "step": 18959 + }, + { + "epoch": 3.3755341880341883, + "grad_norm": 1.1019811630249023, + "learning_rate": 1.1835106510809146e-05, + "loss": 0.7376, + "step": 18960 + }, + { + "epoch": 3.3757122507122506, + "grad_norm": 0.8897548913955688, + "learning_rate": 1.1828501885171995e-05, + "loss": 0.6464, + "step": 18961 + }, + { + "epoch": 3.3758903133903133, + "grad_norm": 0.9695185422897339, + "learning_rate": 1.1821898987072188e-05, + "loss": 0.7566, + "step": 18962 + }, + { + "epoch": 3.376068376068376, + "grad_norm": 1.022469162940979, + "learning_rate": 1.1815297816639071e-05, + "loss": 0.8839, + "step": 18963 + }, + { + "epoch": 3.376246438746439, + "grad_norm": 0.9937511086463928, + "learning_rate": 1.1808698374002025e-05, + "loss": 0.8049, + "step": 18964 + }, + { + "epoch": 3.3764245014245016, + "grad_norm": 1.1339633464813232, + "learning_rate": 1.1802100659290337e-05, + "loss": 0.8131, + "step": 18965 + }, + { + "epoch": 3.376602564102564, + "grad_norm": 0.9775537848472595, + "learning_rate": 1.1795504672633273e-05, + "loss": 0.7819, + "step": 18966 + }, + { + "epoch": 3.3767806267806266, + "grad_norm": 1.0051268339157104, + "learning_rate": 1.1788910414160059e-05, + "loss": 0.658, + "step": 18967 + }, + { + "epoch": 3.3769586894586894, + "grad_norm": 0.9491516947746277, + "learning_rate": 1.1782317883999915e-05, + "loss": 0.9192, + "step": 18968 + }, + { + "epoch": 3.377136752136752, + "grad_norm": 0.9227840900421143, + "learning_rate": 1.1775727082281962e-05, + "loss": 0.7612, + "step": 18969 + }, + { + "epoch": 3.377314814814815, + "grad_norm": 1.0352122783660889, + "learning_rate": 1.1769138009135416e-05, + "loss": 0.8781, + "step": 18970 + }, + { + "epoch": 3.3774928774928776, + "grad_norm": 1.0505750179290771, + "learning_rate": 1.1762550664689286e-05, + "loss": 0.8723, + "step": 18971 + }, + { + "epoch": 3.3776709401709404, + "grad_norm": 0.9471299648284912, + "learning_rate": 1.1755965049072693e-05, + "loss": 0.8168, + "step": 18972 + }, + { + "epoch": 3.3778490028490027, + "grad_norm": 1.0091118812561035, + "learning_rate": 1.1749381162414652e-05, + "loss": 0.6761, + "step": 18973 + }, + { + "epoch": 3.3780270655270654, + "grad_norm": 1.0159013271331787, + "learning_rate": 1.1742799004844174e-05, + "loss": 0.6337, + "step": 18974 + }, + { + "epoch": 3.378205128205128, + "grad_norm": 1.0220392942428589, + "learning_rate": 1.17362185764902e-05, + "loss": 0.9644, + "step": 18975 + }, + { + "epoch": 3.378383190883191, + "grad_norm": 0.9097932577133179, + "learning_rate": 1.172963987748168e-05, + "loss": 0.815, + "step": 18976 + }, + { + "epoch": 3.3785612535612537, + "grad_norm": 0.9885165691375732, + "learning_rate": 1.172306290794749e-05, + "loss": 0.6194, + "step": 18977 + }, + { + "epoch": 3.378739316239316, + "grad_norm": 1.1714304685592651, + "learning_rate": 1.1716487668016506e-05, + "loss": 0.863, + "step": 18978 + }, + { + "epoch": 3.3789173789173788, + "grad_norm": 0.9001526832580566, + "learning_rate": 1.1709914157817537e-05, + "loss": 0.6687, + "step": 18979 + }, + { + "epoch": 3.3790954415954415, + "grad_norm": 0.8851606249809265, + "learning_rate": 1.1703342377479432e-05, + "loss": 0.6135, + "step": 18980 + }, + { + "epoch": 3.3792735042735043, + "grad_norm": 1.1299785375595093, + "learning_rate": 1.1696772327130878e-05, + "loss": 0.7939, + "step": 18981 + }, + { + "epoch": 3.379451566951567, + "grad_norm": 0.8739166855812073, + "learning_rate": 1.1690204006900652e-05, + "loss": 0.6802, + "step": 18982 + }, + { + "epoch": 3.3796296296296298, + "grad_norm": 1.2778574228286743, + "learning_rate": 1.1683637416917425e-05, + "loss": 0.7556, + "step": 18983 + }, + { + "epoch": 3.3798076923076925, + "grad_norm": 1.0917866230010986, + "learning_rate": 1.1677072557309843e-05, + "loss": 0.8309, + "step": 18984 + }, + { + "epoch": 3.379985754985755, + "grad_norm": 0.9844012260437012, + "learning_rate": 1.16705094282066e-05, + "loss": 0.8137, + "step": 18985 + }, + { + "epoch": 3.3801638176638176, + "grad_norm": 0.8838617205619812, + "learning_rate": 1.1663948029736182e-05, + "loss": 0.7654, + "step": 18986 + }, + { + "epoch": 3.3803418803418803, + "grad_norm": 0.9886083602905273, + "learning_rate": 1.1657388362027255e-05, + "loss": 0.8853, + "step": 18987 + }, + { + "epoch": 3.380519943019943, + "grad_norm": 0.9307940602302551, + "learning_rate": 1.1650830425208236e-05, + "loss": 0.8353, + "step": 18988 + }, + { + "epoch": 3.380698005698006, + "grad_norm": 0.9275619387626648, + "learning_rate": 1.1644274219407692e-05, + "loss": 0.5412, + "step": 18989 + }, + { + "epoch": 3.380876068376068, + "grad_norm": 0.8723110556602478, + "learning_rate": 1.1637719744754038e-05, + "loss": 0.7945, + "step": 18990 + }, + { + "epoch": 3.381054131054131, + "grad_norm": 0.9911721348762512, + "learning_rate": 1.163116700137572e-05, + "loss": 0.9201, + "step": 18991 + }, + { + "epoch": 3.3812321937321936, + "grad_norm": 0.9290093183517456, + "learning_rate": 1.1624615989401112e-05, + "loss": 0.8472, + "step": 18992 + }, + { + "epoch": 3.3814102564102564, + "grad_norm": 1.068137526512146, + "learning_rate": 1.1618066708958574e-05, + "loss": 0.7443, + "step": 18993 + }, + { + "epoch": 3.381588319088319, + "grad_norm": 1.0363795757293701, + "learning_rate": 1.1611519160176398e-05, + "loss": 0.8599, + "step": 18994 + }, + { + "epoch": 3.381766381766382, + "grad_norm": 0.8963867425918579, + "learning_rate": 1.1604973343182935e-05, + "loss": 0.8149, + "step": 18995 + }, + { + "epoch": 3.3819444444444446, + "grad_norm": 0.9797514081001282, + "learning_rate": 1.1598429258106347e-05, + "loss": 0.9245, + "step": 18996 + }, + { + "epoch": 3.382122507122507, + "grad_norm": 0.9539228677749634, + "learning_rate": 1.1591886905074945e-05, + "loss": 0.7468, + "step": 18997 + }, + { + "epoch": 3.3823005698005697, + "grad_norm": 0.9980157613754272, + "learning_rate": 1.1585346284216825e-05, + "loss": 0.9399, + "step": 18998 + }, + { + "epoch": 3.3824786324786325, + "grad_norm": 1.0480377674102783, + "learning_rate": 1.1578807395660207e-05, + "loss": 0.9205, + "step": 18999 + }, + { + "epoch": 3.382656695156695, + "grad_norm": 1.0390716791152954, + "learning_rate": 1.1572270239533168e-05, + "loss": 0.8119, + "step": 19000 + }, + { + "epoch": 3.382834757834758, + "grad_norm": 0.9198310971260071, + "learning_rate": 1.1565734815963814e-05, + "loss": 0.6367, + "step": 19001 + }, + { + "epoch": 3.3830128205128207, + "grad_norm": 0.8358418941497803, + "learning_rate": 1.1559201125080167e-05, + "loss": 0.6825, + "step": 19002 + }, + { + "epoch": 3.383190883190883, + "grad_norm": 0.9615653157234192, + "learning_rate": 1.1552669167010244e-05, + "loss": 0.7571, + "step": 19003 + }, + { + "epoch": 3.3833689458689458, + "grad_norm": 1.0061991214752197, + "learning_rate": 1.1546138941882045e-05, + "loss": 0.8323, + "step": 19004 + }, + { + "epoch": 3.3835470085470085, + "grad_norm": 0.9179396629333496, + "learning_rate": 1.1539610449823512e-05, + "loss": 0.6407, + "step": 19005 + }, + { + "epoch": 3.3837250712250713, + "grad_norm": 1.0422332286834717, + "learning_rate": 1.1533083690962554e-05, + "loss": 0.9024, + "step": 19006 + }, + { + "epoch": 3.383903133903134, + "grad_norm": 0.9112447500228882, + "learning_rate": 1.1526558665427045e-05, + "loss": 0.7383, + "step": 19007 + }, + { + "epoch": 3.3840811965811968, + "grad_norm": 0.9500386714935303, + "learning_rate": 1.1520035373344829e-05, + "loss": 0.6305, + "step": 19008 + }, + { + "epoch": 3.384259259259259, + "grad_norm": 1.0099005699157715, + "learning_rate": 1.1513513814843701e-05, + "loss": 0.8355, + "step": 19009 + }, + { + "epoch": 3.384437321937322, + "grad_norm": 1.0299774408340454, + "learning_rate": 1.1506993990051495e-05, + "loss": 0.747, + "step": 19010 + }, + { + "epoch": 3.3846153846153846, + "grad_norm": 0.9437386393547058, + "learning_rate": 1.1500475899095864e-05, + "loss": 0.4786, + "step": 19011 + }, + { + "epoch": 3.3847934472934473, + "grad_norm": 0.8705305457115173, + "learning_rate": 1.1493959542104615e-05, + "loss": 0.8052, + "step": 19012 + }, + { + "epoch": 3.38497150997151, + "grad_norm": 1.0724691152572632, + "learning_rate": 1.1487444919205336e-05, + "loss": 0.8645, + "step": 19013 + }, + { + "epoch": 3.385149572649573, + "grad_norm": 0.8759792447090149, + "learning_rate": 1.1480932030525725e-05, + "loss": 0.5631, + "step": 19014 + }, + { + "epoch": 3.385327635327635, + "grad_norm": 0.8842191100120544, + "learning_rate": 1.1474420876193359e-05, + "loss": 0.528, + "step": 19015 + }, + { + "epoch": 3.385505698005698, + "grad_norm": 0.8085874319076538, + "learning_rate": 1.1467911456335833e-05, + "loss": 0.6945, + "step": 19016 + }, + { + "epoch": 3.3856837606837606, + "grad_norm": 0.9087811708450317, + "learning_rate": 1.1461403771080658e-05, + "loss": 0.7253, + "step": 19017 + }, + { + "epoch": 3.3858618233618234, + "grad_norm": 0.9133577346801758, + "learning_rate": 1.1454897820555366e-05, + "loss": 0.7641, + "step": 19018 + }, + { + "epoch": 3.386039886039886, + "grad_norm": 1.1190037727355957, + "learning_rate": 1.1448393604887386e-05, + "loss": 1.0035, + "step": 19019 + }, + { + "epoch": 3.386217948717949, + "grad_norm": 0.9388802647590637, + "learning_rate": 1.1441891124204229e-05, + "loss": 0.7469, + "step": 19020 + }, + { + "epoch": 3.386396011396011, + "grad_norm": 1.0015909671783447, + "learning_rate": 1.1435390378633216e-05, + "loss": 0.6523, + "step": 19021 + }, + { + "epoch": 3.386574074074074, + "grad_norm": 0.9067725539207458, + "learning_rate": 1.1428891368301764e-05, + "loss": 0.6453, + "step": 19022 + }, + { + "epoch": 3.3867521367521367, + "grad_norm": 1.1131070852279663, + "learning_rate": 1.1422394093337196e-05, + "loss": 0.9134, + "step": 19023 + }, + { + "epoch": 3.3869301994301995, + "grad_norm": 0.922660231590271, + "learning_rate": 1.141589855386681e-05, + "loss": 0.8114, + "step": 19024 + }, + { + "epoch": 3.387108262108262, + "grad_norm": 0.9011611342430115, + "learning_rate": 1.1409404750017872e-05, + "loss": 0.6133, + "step": 19025 + }, + { + "epoch": 3.387286324786325, + "grad_norm": 1.0388410091400146, + "learning_rate": 1.140291268191761e-05, + "loss": 0.8391, + "step": 19026 + }, + { + "epoch": 3.3874643874643873, + "grad_norm": 0.971215009689331, + "learning_rate": 1.139642234969326e-05, + "loss": 0.8261, + "step": 19027 + }, + { + "epoch": 3.38764245014245, + "grad_norm": 0.8922646641731262, + "learning_rate": 1.1389933753471915e-05, + "loss": 0.6263, + "step": 19028 + }, + { + "epoch": 3.3878205128205128, + "grad_norm": 0.8670883178710938, + "learning_rate": 1.1383446893380767e-05, + "loss": 0.8308, + "step": 19029 + }, + { + "epoch": 3.3879985754985755, + "grad_norm": 0.8676888942718506, + "learning_rate": 1.137696176954689e-05, + "loss": 0.6909, + "step": 19030 + }, + { + "epoch": 3.3881766381766383, + "grad_norm": 1.0416065454483032, + "learning_rate": 1.1370478382097361e-05, + "loss": 0.7028, + "step": 19031 + }, + { + "epoch": 3.388354700854701, + "grad_norm": 0.8439168930053711, + "learning_rate": 1.1363996731159188e-05, + "loss": 0.6838, + "step": 19032 + }, + { + "epoch": 3.388532763532764, + "grad_norm": 1.0615967512130737, + "learning_rate": 1.135751681685937e-05, + "loss": 0.8698, + "step": 19033 + }, + { + "epoch": 3.388710826210826, + "grad_norm": 1.0846972465515137, + "learning_rate": 1.1351038639324874e-05, + "loss": 0.8899, + "step": 19034 + }, + { + "epoch": 3.388888888888889, + "grad_norm": 0.8791643381118774, + "learning_rate": 1.1344562198682617e-05, + "loss": 0.7181, + "step": 19035 + }, + { + "epoch": 3.3890669515669516, + "grad_norm": 0.904293954372406, + "learning_rate": 1.133808749505949e-05, + "loss": 0.6749, + "step": 19036 + }, + { + "epoch": 3.3892450142450143, + "grad_norm": 0.877621591091156, + "learning_rate": 1.1331614528582391e-05, + "loss": 0.771, + "step": 19037 + }, + { + "epoch": 3.389423076923077, + "grad_norm": 0.8001459240913391, + "learning_rate": 1.1325143299378083e-05, + "loss": 0.5418, + "step": 19038 + }, + { + "epoch": 3.3896011396011394, + "grad_norm": 0.8537688851356506, + "learning_rate": 1.1318673807573399e-05, + "loss": 0.72, + "step": 19039 + }, + { + "epoch": 3.389779202279202, + "grad_norm": 1.04570734500885, + "learning_rate": 1.1312206053295082e-05, + "loss": 0.7882, + "step": 19040 + }, + { + "epoch": 3.389957264957265, + "grad_norm": 0.8674250245094299, + "learning_rate": 1.1305740036669832e-05, + "loss": 0.7391, + "step": 19041 + }, + { + "epoch": 3.3901353276353277, + "grad_norm": 0.9948473572731018, + "learning_rate": 1.1299275757824402e-05, + "loss": 0.6436, + "step": 19042 + }, + { + "epoch": 3.3903133903133904, + "grad_norm": 0.8702938556671143, + "learning_rate": 1.1292813216885366e-05, + "loss": 0.7127, + "step": 19043 + }, + { + "epoch": 3.390491452991453, + "grad_norm": 0.9511193037033081, + "learning_rate": 1.1286352413979396e-05, + "loss": 0.5849, + "step": 19044 + }, + { + "epoch": 3.390669515669516, + "grad_norm": 1.037414312362671, + "learning_rate": 1.1279893349233073e-05, + "loss": 0.6974, + "step": 19045 + }, + { + "epoch": 3.390847578347578, + "grad_norm": 0.9115296602249146, + "learning_rate": 1.1273436022772931e-05, + "loss": 0.8002, + "step": 19046 + }, + { + "epoch": 3.391025641025641, + "grad_norm": 0.9375928044319153, + "learning_rate": 1.1266980434725506e-05, + "loss": 0.6869, + "step": 19047 + }, + { + "epoch": 3.3912037037037037, + "grad_norm": 0.9377243518829346, + "learning_rate": 1.1260526585217257e-05, + "loss": 0.6117, + "step": 19048 + }, + { + "epoch": 3.3913817663817665, + "grad_norm": 0.9592469334602356, + "learning_rate": 1.1254074474374665e-05, + "loss": 0.7742, + "step": 19049 + }, + { + "epoch": 3.3915598290598292, + "grad_norm": 1.103419542312622, + "learning_rate": 1.1247624102324127e-05, + "loss": 0.8691, + "step": 19050 + }, + { + "epoch": 3.3917378917378915, + "grad_norm": 0.8406701683998108, + "learning_rate": 1.1241175469192e-05, + "loss": 0.6666, + "step": 19051 + }, + { + "epoch": 3.3919159544159543, + "grad_norm": 0.9886174201965332, + "learning_rate": 1.1234728575104703e-05, + "loss": 0.6932, + "step": 19052 + }, + { + "epoch": 3.392094017094017, + "grad_norm": 0.9601133465766907, + "learning_rate": 1.1228283420188468e-05, + "loss": 0.696, + "step": 19053 + }, + { + "epoch": 3.39227207977208, + "grad_norm": 1.1034084558486938, + "learning_rate": 1.1221840004569629e-05, + "loss": 0.7785, + "step": 19054 + }, + { + "epoch": 3.3924501424501425, + "grad_norm": 0.930463969707489, + "learning_rate": 1.1215398328374415e-05, + "loss": 0.7812, + "step": 19055 + }, + { + "epoch": 3.3926282051282053, + "grad_norm": 1.054152488708496, + "learning_rate": 1.120895839172904e-05, + "loss": 0.7522, + "step": 19056 + }, + { + "epoch": 3.392806267806268, + "grad_norm": 1.0049049854278564, + "learning_rate": 1.1202520194759681e-05, + "loss": 0.6845, + "step": 19057 + }, + { + "epoch": 3.3929843304843303, + "grad_norm": 0.9241994619369507, + "learning_rate": 1.1196083737592467e-05, + "loss": 0.9938, + "step": 19058 + }, + { + "epoch": 3.393162393162393, + "grad_norm": 0.988713800907135, + "learning_rate": 1.118964902035352e-05, + "loss": 0.7819, + "step": 19059 + }, + { + "epoch": 3.393340455840456, + "grad_norm": 0.9430522918701172, + "learning_rate": 1.1183216043168921e-05, + "loss": 0.6091, + "step": 19060 + }, + { + "epoch": 3.3935185185185186, + "grad_norm": 1.0958164930343628, + "learning_rate": 1.1176784806164676e-05, + "loss": 0.6572, + "step": 19061 + }, + { + "epoch": 3.3936965811965814, + "grad_norm": 0.9768433570861816, + "learning_rate": 1.117035530946684e-05, + "loss": 0.9099, + "step": 19062 + }, + { + "epoch": 3.3938746438746437, + "grad_norm": 0.9464746117591858, + "learning_rate": 1.116392755320137e-05, + "loss": 0.7463, + "step": 19063 + }, + { + "epoch": 3.3940527065527064, + "grad_norm": 1.061977505683899, + "learning_rate": 1.1157501537494197e-05, + "loss": 0.9118, + "step": 19064 + }, + { + "epoch": 3.394230769230769, + "grad_norm": 0.9586532711982727, + "learning_rate": 1.1151077262471222e-05, + "loss": 0.8105, + "step": 19065 + }, + { + "epoch": 3.394408831908832, + "grad_norm": 0.9120199680328369, + "learning_rate": 1.1144654728258297e-05, + "loss": 0.5844, + "step": 19066 + }, + { + "epoch": 3.3945868945868947, + "grad_norm": 1.0607168674468994, + "learning_rate": 1.1138233934981323e-05, + "loss": 0.7454, + "step": 19067 + }, + { + "epoch": 3.3947649572649574, + "grad_norm": 1.0327951908111572, + "learning_rate": 1.1131814882766034e-05, + "loss": 0.7933, + "step": 19068 + }, + { + "epoch": 3.39494301994302, + "grad_norm": 0.9982885122299194, + "learning_rate": 1.112539757173825e-05, + "loss": 0.7294, + "step": 19069 + }, + { + "epoch": 3.3951210826210825, + "grad_norm": 0.8472577929496765, + "learning_rate": 1.1118982002023649e-05, + "loss": 0.6405, + "step": 19070 + }, + { + "epoch": 3.3952991452991452, + "grad_norm": 0.8824619650840759, + "learning_rate": 1.1112568173747972e-05, + "loss": 0.8163, + "step": 19071 + }, + { + "epoch": 3.395477207977208, + "grad_norm": 1.1966749429702759, + "learning_rate": 1.1106156087036879e-05, + "loss": 1.0977, + "step": 19072 + }, + { + "epoch": 3.3956552706552707, + "grad_norm": 0.9600608944892883, + "learning_rate": 1.1099745742016e-05, + "loss": 0.7247, + "step": 19073 + }, + { + "epoch": 3.3958333333333335, + "grad_norm": 0.9293968677520752, + "learning_rate": 1.1093337138810921e-05, + "loss": 0.6889, + "step": 19074 + }, + { + "epoch": 3.396011396011396, + "grad_norm": 1.020552635192871, + "learning_rate": 1.1086930277547226e-05, + "loss": 0.7282, + "step": 19075 + }, + { + "epoch": 3.3961894586894585, + "grad_norm": 0.8889341950416565, + "learning_rate": 1.1080525158350397e-05, + "loss": 0.6518, + "step": 19076 + }, + { + "epoch": 3.3963675213675213, + "grad_norm": 0.9952123761177063, + "learning_rate": 1.1074121781346014e-05, + "loss": 0.8552, + "step": 19077 + }, + { + "epoch": 3.396545584045584, + "grad_norm": 0.8794593811035156, + "learning_rate": 1.1067720146659432e-05, + "loss": 0.8402, + "step": 19078 + }, + { + "epoch": 3.396723646723647, + "grad_norm": 1.0979092121124268, + "learning_rate": 1.1061320254416175e-05, + "loss": 0.8358, + "step": 19079 + }, + { + "epoch": 3.3969017094017095, + "grad_norm": 0.9097040295600891, + "learning_rate": 1.1054922104741572e-05, + "loss": 0.7482, + "step": 19080 + }, + { + "epoch": 3.3970797720797723, + "grad_norm": 0.9845696687698364, + "learning_rate": 1.1048525697760993e-05, + "loss": 0.8369, + "step": 19081 + }, + { + "epoch": 3.3972578347578346, + "grad_norm": 0.7637016773223877, + "learning_rate": 1.1042131033599812e-05, + "loss": 0.541, + "step": 19082 + }, + { + "epoch": 3.3974358974358974, + "grad_norm": 0.9949057698249817, + "learning_rate": 1.1035738112383243e-05, + "loss": 0.8823, + "step": 19083 + }, + { + "epoch": 3.39761396011396, + "grad_norm": 0.9214608669281006, + "learning_rate": 1.1029346934236618e-05, + "loss": 0.7869, + "step": 19084 + }, + { + "epoch": 3.397792022792023, + "grad_norm": 1.1592236757278442, + "learning_rate": 1.102295749928507e-05, + "loss": 0.883, + "step": 19085 + }, + { + "epoch": 3.3979700854700856, + "grad_norm": 1.0953752994537354, + "learning_rate": 1.1016569807653865e-05, + "loss": 0.8813, + "step": 19086 + }, + { + "epoch": 3.398148148148148, + "grad_norm": 0.9052318334579468, + "learning_rate": 1.1010183859468126e-05, + "loss": 0.6289, + "step": 19087 + }, + { + "epoch": 3.3983262108262107, + "grad_norm": 0.9558716416358948, + "learning_rate": 1.1003799654852976e-05, + "loss": 0.87, + "step": 19088 + }, + { + "epoch": 3.3985042735042734, + "grad_norm": 0.8846414089202881, + "learning_rate": 1.099741719393349e-05, + "loss": 0.7292, + "step": 19089 + }, + { + "epoch": 3.398682336182336, + "grad_norm": 1.1074748039245605, + "learning_rate": 1.0991036476834737e-05, + "loss": 0.8082, + "step": 19090 + }, + { + "epoch": 3.398860398860399, + "grad_norm": 0.924491286277771, + "learning_rate": 1.0984657503681695e-05, + "loss": 0.7887, + "step": 19091 + }, + { + "epoch": 3.3990384615384617, + "grad_norm": 0.9129840135574341, + "learning_rate": 1.0978280274599418e-05, + "loss": 0.7463, + "step": 19092 + }, + { + "epoch": 3.3992165242165244, + "grad_norm": 1.0049107074737549, + "learning_rate": 1.0971904789712773e-05, + "loss": 0.7386, + "step": 19093 + }, + { + "epoch": 3.3993945868945867, + "grad_norm": 0.9263444542884827, + "learning_rate": 1.096553104914676e-05, + "loss": 0.8414, + "step": 19094 + }, + { + "epoch": 3.3995726495726495, + "grad_norm": 1.01347017288208, + "learning_rate": 1.095915905302617e-05, + "loss": 0.6919, + "step": 19095 + }, + { + "epoch": 3.3997507122507122, + "grad_norm": 0.9228973388671875, + "learning_rate": 1.095278880147591e-05, + "loss": 0.6941, + "step": 19096 + }, + { + "epoch": 3.399928774928775, + "grad_norm": 0.9571877717971802, + "learning_rate": 1.0946420294620774e-05, + "loss": 0.6876, + "step": 19097 + }, + { + "epoch": 3.4001068376068377, + "grad_norm": 1.3483099937438965, + "learning_rate": 1.0940053532585537e-05, + "loss": 0.9632, + "step": 19098 + }, + { + "epoch": 3.4002849002849, + "grad_norm": 0.920470654964447, + "learning_rate": 1.0933688515494945e-05, + "loss": 0.8084, + "step": 19099 + }, + { + "epoch": 3.400462962962963, + "grad_norm": 1.0069130659103394, + "learning_rate": 1.0927325243473718e-05, + "loss": 0.6561, + "step": 19100 + }, + { + "epoch": 3.4006410256410255, + "grad_norm": 0.9278994202613831, + "learning_rate": 1.0920963716646492e-05, + "loss": 0.6935, + "step": 19101 + }, + { + "epoch": 3.4008190883190883, + "grad_norm": 0.9868398308753967, + "learning_rate": 1.0914603935137957e-05, + "loss": 0.7568, + "step": 19102 + }, + { + "epoch": 3.400997150997151, + "grad_norm": 0.9027713537216187, + "learning_rate": 1.09082458990727e-05, + "loss": 0.7063, + "step": 19103 + }, + { + "epoch": 3.401175213675214, + "grad_norm": 0.8853753209114075, + "learning_rate": 1.0901889608575288e-05, + "loss": 0.7147, + "step": 19104 + }, + { + "epoch": 3.4013532763532766, + "grad_norm": 0.9795311093330383, + "learning_rate": 1.0895535063770268e-05, + "loss": 0.741, + "step": 19105 + }, + { + "epoch": 3.401531339031339, + "grad_norm": 0.8565807938575745, + "learning_rate": 1.0889182264782138e-05, + "loss": 0.6649, + "step": 19106 + }, + { + "epoch": 3.4017094017094016, + "grad_norm": 0.8639920353889465, + "learning_rate": 1.0882831211735367e-05, + "loss": 0.6754, + "step": 19107 + }, + { + "epoch": 3.4018874643874644, + "grad_norm": 0.8701103329658508, + "learning_rate": 1.0876481904754376e-05, + "loss": 0.7004, + "step": 19108 + }, + { + "epoch": 3.402065527065527, + "grad_norm": 1.0115605592727661, + "learning_rate": 1.0870134343963633e-05, + "loss": 0.8188, + "step": 19109 + }, + { + "epoch": 3.40224358974359, + "grad_norm": 1.002318263053894, + "learning_rate": 1.0863788529487407e-05, + "loss": 0.7851, + "step": 19110 + }, + { + "epoch": 3.402421652421652, + "grad_norm": 1.0113846063613892, + "learning_rate": 1.0857444461450106e-05, + "loss": 0.911, + "step": 19111 + }, + { + "epoch": 3.402599715099715, + "grad_norm": 1.1443030834197998, + "learning_rate": 1.0851102139976e-05, + "loss": 0.8623, + "step": 19112 + }, + { + "epoch": 3.4027777777777777, + "grad_norm": 0.9804279208183289, + "learning_rate": 1.0844761565189354e-05, + "loss": 0.6802, + "step": 19113 + }, + { + "epoch": 3.4029558404558404, + "grad_norm": 0.9586710333824158, + "learning_rate": 1.0838422737214404e-05, + "loss": 0.8277, + "step": 19114 + }, + { + "epoch": 3.403133903133903, + "grad_norm": 0.8377839922904968, + "learning_rate": 1.083208565617534e-05, + "loss": 0.68, + "step": 19115 + }, + { + "epoch": 3.403311965811966, + "grad_norm": 0.901368260383606, + "learning_rate": 1.0825750322196304e-05, + "loss": 0.7212, + "step": 19116 + }, + { + "epoch": 3.4034900284900287, + "grad_norm": 0.9065432548522949, + "learning_rate": 1.08194167354015e-05, + "loss": 0.7003, + "step": 19117 + }, + { + "epoch": 3.403668091168091, + "grad_norm": 0.9668724536895752, + "learning_rate": 1.0813084895914915e-05, + "loss": 0.7776, + "step": 19118 + }, + { + "epoch": 3.4038461538461537, + "grad_norm": 0.9226725101470947, + "learning_rate": 1.0806754803860708e-05, + "loss": 0.8726, + "step": 19119 + }, + { + "epoch": 3.4040242165242165, + "grad_norm": 0.952060341835022, + "learning_rate": 1.0800426459362822e-05, + "loss": 0.6952, + "step": 19120 + }, + { + "epoch": 3.4042022792022792, + "grad_norm": 0.8968328833580017, + "learning_rate": 1.0794099862545293e-05, + "loss": 0.68, + "step": 19121 + }, + { + "epoch": 3.404380341880342, + "grad_norm": 0.8708463907241821, + "learning_rate": 1.0787775013532075e-05, + "loss": 0.6991, + "step": 19122 + }, + { + "epoch": 3.4045584045584047, + "grad_norm": 0.909925639629364, + "learning_rate": 1.078145191244706e-05, + "loss": 0.6232, + "step": 19123 + }, + { + "epoch": 3.404736467236467, + "grad_norm": 1.08229398727417, + "learning_rate": 1.0775130559414204e-05, + "loss": 0.786, + "step": 19124 + }, + { + "epoch": 3.40491452991453, + "grad_norm": 0.9113919734954834, + "learning_rate": 1.0768810954557273e-05, + "loss": 0.9116, + "step": 19125 + }, + { + "epoch": 3.4050925925925926, + "grad_norm": 1.0875805616378784, + "learning_rate": 1.0762493098000158e-05, + "loss": 0.922, + "step": 19126 + }, + { + "epoch": 3.4052706552706553, + "grad_norm": 1.0122466087341309, + "learning_rate": 1.0756176989866607e-05, + "loss": 0.6557, + "step": 19127 + }, + { + "epoch": 3.405448717948718, + "grad_norm": 0.9092339277267456, + "learning_rate": 1.0749862630280384e-05, + "loss": 0.5974, + "step": 19128 + }, + { + "epoch": 3.405626780626781, + "grad_norm": 0.8997379541397095, + "learning_rate": 1.0743550019365189e-05, + "loss": 0.9398, + "step": 19129 + }, + { + "epoch": 3.405804843304843, + "grad_norm": 0.9965963959693909, + "learning_rate": 1.0737239157244727e-05, + "loss": 0.7401, + "step": 19130 + }, + { + "epoch": 3.405982905982906, + "grad_norm": 0.9418243765830994, + "learning_rate": 1.0730930044042642e-05, + "loss": 0.8724, + "step": 19131 + }, + { + "epoch": 3.4061609686609686, + "grad_norm": 0.8757455348968506, + "learning_rate": 1.0724622679882534e-05, + "loss": 0.6671, + "step": 19132 + }, + { + "epoch": 3.4063390313390314, + "grad_norm": 0.8993387818336487, + "learning_rate": 1.0718317064887972e-05, + "loss": 0.6749, + "step": 19133 + }, + { + "epoch": 3.406517094017094, + "grad_norm": 0.9798662662506104, + "learning_rate": 1.071201319918257e-05, + "loss": 0.926, + "step": 19134 + }, + { + "epoch": 3.406695156695157, + "grad_norm": 0.8816490173339844, + "learning_rate": 1.0705711082889736e-05, + "loss": 0.789, + "step": 19135 + }, + { + "epoch": 3.406873219373219, + "grad_norm": 1.0084799528121948, + "learning_rate": 1.069941071613303e-05, + "loss": 0.8257, + "step": 19136 + }, + { + "epoch": 3.407051282051282, + "grad_norm": 0.9338811635971069, + "learning_rate": 1.0693112099035862e-05, + "loss": 0.723, + "step": 19137 + }, + { + "epoch": 3.4072293447293447, + "grad_norm": 0.9649238586425781, + "learning_rate": 1.0686815231721636e-05, + "loss": 0.9459, + "step": 19138 + }, + { + "epoch": 3.4074074074074074, + "grad_norm": 0.9531043767929077, + "learning_rate": 1.068052011431374e-05, + "loss": 0.6914, + "step": 19139 + }, + { + "epoch": 3.40758547008547, + "grad_norm": 1.1801680326461792, + "learning_rate": 1.0674226746935511e-05, + "loss": 0.9974, + "step": 19140 + }, + { + "epoch": 3.407763532763533, + "grad_norm": 0.9781167507171631, + "learning_rate": 1.0667935129710249e-05, + "loss": 0.6573, + "step": 19141 + }, + { + "epoch": 3.4079415954415953, + "grad_norm": 1.018373966217041, + "learning_rate": 1.06616452627612e-05, + "loss": 0.8378, + "step": 19142 + }, + { + "epoch": 3.408119658119658, + "grad_norm": 0.9358958005905151, + "learning_rate": 1.0655357146211653e-05, + "loss": 0.7662, + "step": 19143 + }, + { + "epoch": 3.4082977207977208, + "grad_norm": 0.9799219369888306, + "learning_rate": 1.064907078018479e-05, + "loss": 0.8614, + "step": 19144 + }, + { + "epoch": 3.4084757834757835, + "grad_norm": 1.0031346082687378, + "learning_rate": 1.0642786164803764e-05, + "loss": 0.6955, + "step": 19145 + }, + { + "epoch": 3.4086538461538463, + "grad_norm": 0.8543678522109985, + "learning_rate": 1.0636503300191736e-05, + "loss": 0.7027, + "step": 19146 + }, + { + "epoch": 3.408831908831909, + "grad_norm": 0.9312917590141296, + "learning_rate": 1.0630222186471773e-05, + "loss": 0.7433, + "step": 19147 + }, + { + "epoch": 3.4090099715099713, + "grad_norm": 1.0434274673461914, + "learning_rate": 1.0623942823766941e-05, + "loss": 0.8913, + "step": 19148 + }, + { + "epoch": 3.409188034188034, + "grad_norm": 0.9719704985618591, + "learning_rate": 1.0617665212200335e-05, + "loss": 0.7474, + "step": 19149 + }, + { + "epoch": 3.409366096866097, + "grad_norm": 0.9384031295776367, + "learning_rate": 1.0611389351894852e-05, + "loss": 0.738, + "step": 19150 + }, + { + "epoch": 3.4095441595441596, + "grad_norm": 0.9763504862785339, + "learning_rate": 1.060511524297353e-05, + "loss": 0.9144, + "step": 19151 + }, + { + "epoch": 3.4097222222222223, + "grad_norm": 1.044063925743103, + "learning_rate": 1.0598842885559268e-05, + "loss": 0.8187, + "step": 19152 + }, + { + "epoch": 3.409900284900285, + "grad_norm": 0.9677876234054565, + "learning_rate": 1.059257227977497e-05, + "loss": 0.6989, + "step": 19153 + }, + { + "epoch": 3.410078347578348, + "grad_norm": 0.8552587032318115, + "learning_rate": 1.0586303425743493e-05, + "loss": 0.8524, + "step": 19154 + }, + { + "epoch": 3.41025641025641, + "grad_norm": 1.0604221820831299, + "learning_rate": 1.058003632358766e-05, + "loss": 0.7479, + "step": 19155 + }, + { + "epoch": 3.410434472934473, + "grad_norm": 0.7916514277458191, + "learning_rate": 1.0573770973430253e-05, + "loss": 0.5946, + "step": 19156 + }, + { + "epoch": 3.4106125356125356, + "grad_norm": 0.7886891961097717, + "learning_rate": 1.0567507375394048e-05, + "loss": 0.6991, + "step": 19157 + }, + { + "epoch": 3.4107905982905984, + "grad_norm": 0.8678019642829895, + "learning_rate": 1.0561245529601727e-05, + "loss": 0.6513, + "step": 19158 + }, + { + "epoch": 3.410968660968661, + "grad_norm": 1.034912347793579, + "learning_rate": 1.0554985436176045e-05, + "loss": 0.8279, + "step": 19159 + }, + { + "epoch": 3.4111467236467234, + "grad_norm": 0.9476612210273743, + "learning_rate": 1.0548727095239585e-05, + "loss": 0.7955, + "step": 19160 + }, + { + "epoch": 3.411324786324786, + "grad_norm": 0.8874943256378174, + "learning_rate": 1.0542470506915015e-05, + "loss": 0.6118, + "step": 19161 + }, + { + "epoch": 3.411502849002849, + "grad_norm": 0.9441639184951782, + "learning_rate": 1.0536215671324912e-05, + "loss": 0.7955, + "step": 19162 + }, + { + "epoch": 3.4116809116809117, + "grad_norm": 0.8857356309890747, + "learning_rate": 1.052996258859179e-05, + "loss": 0.7249, + "step": 19163 + }, + { + "epoch": 3.4118589743589745, + "grad_norm": 0.9193042516708374, + "learning_rate": 1.0523711258838231e-05, + "loss": 0.7856, + "step": 19164 + }, + { + "epoch": 3.412037037037037, + "grad_norm": 0.962785542011261, + "learning_rate": 1.0517461682186646e-05, + "loss": 0.9475, + "step": 19165 + }, + { + "epoch": 3.4122150997151, + "grad_norm": 1.0464279651641846, + "learning_rate": 1.051121385875955e-05, + "loss": 0.8153, + "step": 19166 + }, + { + "epoch": 3.4123931623931623, + "grad_norm": 0.9516451954841614, + "learning_rate": 1.0504967788679277e-05, + "loss": 0.683, + "step": 19167 + }, + { + "epoch": 3.412571225071225, + "grad_norm": 0.9690127968788147, + "learning_rate": 1.0498723472068283e-05, + "loss": 0.6226, + "step": 19168 + }, + { + "epoch": 3.4127492877492878, + "grad_norm": 0.9517914652824402, + "learning_rate": 1.0492480909048875e-05, + "loss": 0.626, + "step": 19169 + }, + { + "epoch": 3.4129273504273505, + "grad_norm": 0.9089692234992981, + "learning_rate": 1.0486240099743362e-05, + "loss": 0.6041, + "step": 19170 + }, + { + "epoch": 3.4131054131054133, + "grad_norm": 0.9216201305389404, + "learning_rate": 1.0480001044274023e-05, + "loss": 0.6596, + "step": 19171 + }, + { + "epoch": 3.4132834757834756, + "grad_norm": 1.018068552017212, + "learning_rate": 1.047376374276311e-05, + "loss": 0.8819, + "step": 19172 + }, + { + "epoch": 3.4134615384615383, + "grad_norm": 0.9685564041137695, + "learning_rate": 1.046752819533281e-05, + "loss": 0.6659, + "step": 19173 + }, + { + "epoch": 3.413639601139601, + "grad_norm": 0.9151574373245239, + "learning_rate": 1.0461294402105327e-05, + "loss": 0.6773, + "step": 19174 + }, + { + "epoch": 3.413817663817664, + "grad_norm": 0.8832176923751831, + "learning_rate": 1.045506236320275e-05, + "loss": 0.658, + "step": 19175 + }, + { + "epoch": 3.4139957264957266, + "grad_norm": 0.8827099204063416, + "learning_rate": 1.0448832078747262e-05, + "loss": 0.7711, + "step": 19176 + }, + { + "epoch": 3.4141737891737893, + "grad_norm": 0.8752033114433289, + "learning_rate": 1.044260354886083e-05, + "loss": 0.8464, + "step": 19177 + }, + { + "epoch": 3.414351851851852, + "grad_norm": 1.0381900072097778, + "learning_rate": 1.0436376773665569e-05, + "loss": 0.9648, + "step": 19178 + }, + { + "epoch": 3.4145299145299144, + "grad_norm": 1.02308988571167, + "learning_rate": 1.0430151753283445e-05, + "loss": 0.9882, + "step": 19179 + }, + { + "epoch": 3.414707977207977, + "grad_norm": 1.0302221775054932, + "learning_rate": 1.042392848783642e-05, + "loss": 0.8057, + "step": 19180 + }, + { + "epoch": 3.41488603988604, + "grad_norm": 1.1306030750274658, + "learning_rate": 1.0417706977446473e-05, + "loss": 0.7419, + "step": 19181 + }, + { + "epoch": 3.4150641025641026, + "grad_norm": 0.9197164177894592, + "learning_rate": 1.0411487222235428e-05, + "loss": 0.7504, + "step": 19182 + }, + { + "epoch": 3.4152421652421654, + "grad_norm": 0.8756219744682312, + "learning_rate": 1.0405269222325209e-05, + "loss": 0.7493, + "step": 19183 + }, + { + "epoch": 3.4154202279202277, + "grad_norm": 1.0161107778549194, + "learning_rate": 1.039905297783762e-05, + "loss": 0.6777, + "step": 19184 + }, + { + "epoch": 3.4155982905982905, + "grad_norm": 0.8732602596282959, + "learning_rate": 1.0392838488894463e-05, + "loss": 0.7629, + "step": 19185 + }, + { + "epoch": 3.415776353276353, + "grad_norm": 1.0775624513626099, + "learning_rate": 1.0386625755617485e-05, + "loss": 0.7746, + "step": 19186 + }, + { + "epoch": 3.415954415954416, + "grad_norm": 0.8961769342422485, + "learning_rate": 1.0380414778128423e-05, + "loss": 0.5364, + "step": 19187 + }, + { + "epoch": 3.4161324786324787, + "grad_norm": 0.9528487324714661, + "learning_rate": 1.0374205556548944e-05, + "loss": 0.7021, + "step": 19188 + }, + { + "epoch": 3.4163105413105415, + "grad_norm": 1.0150240659713745, + "learning_rate": 1.0367998091000764e-05, + "loss": 0.7728, + "step": 19189 + }, + { + "epoch": 3.416488603988604, + "grad_norm": 0.9373225569725037, + "learning_rate": 1.036179238160544e-05, + "loss": 0.8242, + "step": 19190 + }, + { + "epoch": 3.4166666666666665, + "grad_norm": 1.0559889078140259, + "learning_rate": 1.0355588428484608e-05, + "loss": 0.7779, + "step": 19191 + }, + { + "epoch": 3.4168447293447293, + "grad_norm": 0.8923398852348328, + "learning_rate": 1.0349386231759773e-05, + "loss": 0.621, + "step": 19192 + }, + { + "epoch": 3.417022792022792, + "grad_norm": 0.8571991324424744, + "learning_rate": 1.0343185791552501e-05, + "loss": 0.6731, + "step": 19193 + }, + { + "epoch": 3.4172008547008548, + "grad_norm": 1.01732337474823, + "learning_rate": 1.0336987107984253e-05, + "loss": 0.8578, + "step": 19194 + }, + { + "epoch": 3.4173789173789175, + "grad_norm": 0.9371682405471802, + "learning_rate": 1.0330790181176487e-05, + "loss": 0.8188, + "step": 19195 + }, + { + "epoch": 3.41755698005698, + "grad_norm": 0.8861309885978699, + "learning_rate": 1.0324595011250616e-05, + "loss": 0.7745, + "step": 19196 + }, + { + "epoch": 3.4177350427350426, + "grad_norm": 0.8951418995857239, + "learning_rate": 1.031840159832802e-05, + "loss": 0.6104, + "step": 19197 + }, + { + "epoch": 3.4179131054131053, + "grad_norm": 0.9123335480690002, + "learning_rate": 1.0312209942530027e-05, + "loss": 0.7475, + "step": 19198 + }, + { + "epoch": 3.418091168091168, + "grad_norm": 0.9403263926506042, + "learning_rate": 1.0306020043978004e-05, + "loss": 0.7961, + "step": 19199 + }, + { + "epoch": 3.418269230769231, + "grad_norm": 0.9268109202384949, + "learning_rate": 1.0299831902793168e-05, + "loss": 0.6927, + "step": 19200 + }, + { + "epoch": 3.4184472934472936, + "grad_norm": 0.9422159194946289, + "learning_rate": 1.0293645519096807e-05, + "loss": 0.8041, + "step": 19201 + }, + { + "epoch": 3.4186253561253563, + "grad_norm": 0.9820014834403992, + "learning_rate": 1.0287460893010104e-05, + "loss": 0.6622, + "step": 19202 + }, + { + "epoch": 3.4188034188034186, + "grad_norm": 0.9447580575942993, + "learning_rate": 1.028127802465425e-05, + "loss": 0.8675, + "step": 19203 + }, + { + "epoch": 3.4189814814814814, + "grad_norm": 0.8962849378585815, + "learning_rate": 1.0275096914150373e-05, + "loss": 0.5652, + "step": 19204 + }, + { + "epoch": 3.419159544159544, + "grad_norm": 1.0891627073287964, + "learning_rate": 1.0268917561619574e-05, + "loss": 0.7977, + "step": 19205 + }, + { + "epoch": 3.419337606837607, + "grad_norm": 0.9739325642585754, + "learning_rate": 1.0262739967182966e-05, + "loss": 0.7116, + "step": 19206 + }, + { + "epoch": 3.4195156695156697, + "grad_norm": 0.9855953454971313, + "learning_rate": 1.0256564130961522e-05, + "loss": 0.6721, + "step": 19207 + }, + { + "epoch": 3.419693732193732, + "grad_norm": 1.1992031335830688, + "learning_rate": 1.0250390053076298e-05, + "loss": 0.892, + "step": 19208 + }, + { + "epoch": 3.4198717948717947, + "grad_norm": 0.9223141074180603, + "learning_rate": 1.0244217733648242e-05, + "loss": 0.5814, + "step": 19209 + }, + { + "epoch": 3.4200498575498575, + "grad_norm": 1.0374293327331543, + "learning_rate": 1.0238047172798282e-05, + "loss": 0.6035, + "step": 19210 + }, + { + "epoch": 3.42022792022792, + "grad_norm": 0.8509711027145386, + "learning_rate": 1.023187837064733e-05, + "loss": 0.6018, + "step": 19211 + }, + { + "epoch": 3.420405982905983, + "grad_norm": 1.0256701707839966, + "learning_rate": 1.0225711327316234e-05, + "loss": 0.6962, + "step": 19212 + }, + { + "epoch": 3.4205840455840457, + "grad_norm": 0.9454541206359863, + "learning_rate": 1.0219546042925843e-05, + "loss": 0.8685, + "step": 19213 + }, + { + "epoch": 3.4207621082621085, + "grad_norm": 0.9761176705360413, + "learning_rate": 1.0213382517596948e-05, + "loss": 0.729, + "step": 19214 + }, + { + "epoch": 3.4209401709401708, + "grad_norm": 0.9365816712379456, + "learning_rate": 1.0207220751450286e-05, + "loss": 0.7878, + "step": 19215 + }, + { + "epoch": 3.4211182336182335, + "grad_norm": 0.884979784488678, + "learning_rate": 1.0201060744606639e-05, + "loss": 0.6086, + "step": 19216 + }, + { + "epoch": 3.4212962962962963, + "grad_norm": 0.9652853608131409, + "learning_rate": 1.019490249718663e-05, + "loss": 0.9374, + "step": 19217 + }, + { + "epoch": 3.421474358974359, + "grad_norm": 0.9737793803215027, + "learning_rate": 1.0188746009310968e-05, + "loss": 0.8034, + "step": 19218 + }, + { + "epoch": 3.421652421652422, + "grad_norm": 1.0123218297958374, + "learning_rate": 1.0182591281100262e-05, + "loss": 0.7928, + "step": 19219 + }, + { + "epoch": 3.421830484330484, + "grad_norm": 0.9481899738311768, + "learning_rate": 1.0176438312675086e-05, + "loss": 0.9576, + "step": 19220 + }, + { + "epoch": 3.422008547008547, + "grad_norm": 0.977864146232605, + "learning_rate": 1.017028710415604e-05, + "loss": 0.8792, + "step": 19221 + }, + { + "epoch": 3.4221866096866096, + "grad_norm": 1.0100616216659546, + "learning_rate": 1.0164137655663586e-05, + "loss": 0.8322, + "step": 19222 + }, + { + "epoch": 3.4223646723646723, + "grad_norm": 0.8587307929992676, + "learning_rate": 1.015798996731825e-05, + "loss": 0.68, + "step": 19223 + }, + { + "epoch": 3.422542735042735, + "grad_norm": 0.979102373123169, + "learning_rate": 1.0151844039240465e-05, + "loss": 0.9162, + "step": 19224 + }, + { + "epoch": 3.422720797720798, + "grad_norm": 0.95932936668396, + "learning_rate": 1.0145699871550652e-05, + "loss": 0.865, + "step": 19225 + }, + { + "epoch": 3.4228988603988606, + "grad_norm": 1.080337643623352, + "learning_rate": 1.0139557464369199e-05, + "loss": 0.8694, + "step": 19226 + }, + { + "epoch": 3.423076923076923, + "grad_norm": 0.9872581362724304, + "learning_rate": 1.0133416817816454e-05, + "loss": 0.7365, + "step": 19227 + }, + { + "epoch": 3.4232549857549857, + "grad_norm": 0.9537657499313354, + "learning_rate": 1.0127277932012713e-05, + "loss": 0.9292, + "step": 19228 + }, + { + "epoch": 3.4234330484330484, + "grad_norm": 1.0455421209335327, + "learning_rate": 1.0121140807078278e-05, + "loss": 0.6459, + "step": 19229 + }, + { + "epoch": 3.423611111111111, + "grad_norm": 0.9274110198020935, + "learning_rate": 1.0115005443133351e-05, + "loss": 0.6901, + "step": 19230 + }, + { + "epoch": 3.423789173789174, + "grad_norm": 0.9798219799995422, + "learning_rate": 1.0108871840298217e-05, + "loss": 0.7448, + "step": 19231 + }, + { + "epoch": 3.423967236467236, + "grad_norm": 0.9771175980567932, + "learning_rate": 1.0102739998692977e-05, + "loss": 0.9554, + "step": 19232 + }, + { + "epoch": 3.424145299145299, + "grad_norm": 1.0822041034698486, + "learning_rate": 1.0096609918437805e-05, + "loss": 0.6761, + "step": 19233 + }, + { + "epoch": 3.4243233618233617, + "grad_norm": 1.0950684547424316, + "learning_rate": 1.0090481599652812e-05, + "loss": 0.6483, + "step": 19234 + }, + { + "epoch": 3.4245014245014245, + "grad_norm": 1.1559733152389526, + "learning_rate": 1.008435504245806e-05, + "loss": 0.6516, + "step": 19235 + }, + { + "epoch": 3.4246794871794872, + "grad_norm": 0.8701435327529907, + "learning_rate": 1.0078230246973586e-05, + "loss": 0.6525, + "step": 19236 + }, + { + "epoch": 3.42485754985755, + "grad_norm": 1.002463698387146, + "learning_rate": 1.0072107213319394e-05, + "loss": 0.7652, + "step": 19237 + }, + { + "epoch": 3.4250356125356127, + "grad_norm": 1.0020043849945068, + "learning_rate": 1.0065985941615452e-05, + "loss": 0.6226, + "step": 19238 + }, + { + "epoch": 3.425213675213675, + "grad_norm": 1.0987087488174438, + "learning_rate": 1.0059866431981701e-05, + "loss": 0.8834, + "step": 19239 + }, + { + "epoch": 3.425391737891738, + "grad_norm": 1.0064318180084229, + "learning_rate": 1.0053748684538e-05, + "loss": 0.7865, + "step": 19240 + }, + { + "epoch": 3.4255698005698005, + "grad_norm": 0.9753440618515015, + "learning_rate": 1.0047632699404274e-05, + "loss": 0.8149, + "step": 19241 + }, + { + "epoch": 3.4257478632478633, + "grad_norm": 0.9898001551628113, + "learning_rate": 1.0041518476700329e-05, + "loss": 0.7078, + "step": 19242 + }, + { + "epoch": 3.425925925925926, + "grad_norm": 0.9829652309417725, + "learning_rate": 1.0035406016545955e-05, + "loss": 0.7877, + "step": 19243 + }, + { + "epoch": 3.426103988603989, + "grad_norm": 0.8720074892044067, + "learning_rate": 1.0029295319060905e-05, + "loss": 0.7099, + "step": 19244 + }, + { + "epoch": 3.426282051282051, + "grad_norm": 0.8775247931480408, + "learning_rate": 1.0023186384364902e-05, + "loss": 0.748, + "step": 19245 + }, + { + "epoch": 3.426460113960114, + "grad_norm": 1.011750340461731, + "learning_rate": 1.0017079212577696e-05, + "loss": 0.8075, + "step": 19246 + }, + { + "epoch": 3.4266381766381766, + "grad_norm": 0.9512860178947449, + "learning_rate": 1.0010973803818857e-05, + "loss": 0.8108, + "step": 19247 + }, + { + "epoch": 3.4268162393162394, + "grad_norm": 0.9716009497642517, + "learning_rate": 1.000487015820809e-05, + "loss": 0.8015, + "step": 19248 + }, + { + "epoch": 3.426994301994302, + "grad_norm": 1.016042709350586, + "learning_rate": 9.998768275864901e-06, + "loss": 0.6669, + "step": 19249 + }, + { + "epoch": 3.427172364672365, + "grad_norm": 0.9144439697265625, + "learning_rate": 9.992668156908902e-06, + "loss": 0.6384, + "step": 19250 + }, + { + "epoch": 3.427350427350427, + "grad_norm": 0.9310395121574402, + "learning_rate": 9.9865698014596e-06, + "loss": 0.7346, + "step": 19251 + }, + { + "epoch": 3.42752849002849, + "grad_norm": 1.012934684753418, + "learning_rate": 9.980473209636477e-06, + "loss": 0.7994, + "step": 19252 + }, + { + "epoch": 3.4277065527065527, + "grad_norm": 1.0739672183990479, + "learning_rate": 9.97437838155899e-06, + "loss": 0.9918, + "step": 19253 + }, + { + "epoch": 3.4278846153846154, + "grad_norm": 1.0439870357513428, + "learning_rate": 9.968285317346538e-06, + "loss": 0.8058, + "step": 19254 + }, + { + "epoch": 3.428062678062678, + "grad_norm": 0.8677945733070374, + "learning_rate": 9.962194017118486e-06, + "loss": 0.7721, + "step": 19255 + }, + { + "epoch": 3.428240740740741, + "grad_norm": 0.9556106328964233, + "learning_rate": 9.956104480994254e-06, + "loss": 0.7655, + "step": 19256 + }, + { + "epoch": 3.4284188034188032, + "grad_norm": 0.9485167264938354, + "learning_rate": 9.950016709093068e-06, + "loss": 0.7862, + "step": 19257 + }, + { + "epoch": 3.428596866096866, + "grad_norm": 0.9476799368858337, + "learning_rate": 9.943930701534254e-06, + "loss": 0.7286, + "step": 19258 + }, + { + "epoch": 3.4287749287749287, + "grad_norm": 0.8918333649635315, + "learning_rate": 9.937846458437039e-06, + "loss": 0.5837, + "step": 19259 + }, + { + "epoch": 3.4289529914529915, + "grad_norm": 0.9961351156234741, + "learning_rate": 9.931763979920627e-06, + "loss": 0.719, + "step": 19260 + }, + { + "epoch": 3.4291310541310542, + "grad_norm": 0.832885205745697, + "learning_rate": 9.925683266104224e-06, + "loss": 0.7854, + "step": 19261 + }, + { + "epoch": 3.429309116809117, + "grad_norm": 0.9508270621299744, + "learning_rate": 9.919604317106913e-06, + "loss": 0.7528, + "step": 19262 + }, + { + "epoch": 3.4294871794871793, + "grad_norm": 0.9197067022323608, + "learning_rate": 9.913527133047873e-06, + "loss": 0.8416, + "step": 19263 + }, + { + "epoch": 3.429665242165242, + "grad_norm": 0.8550262451171875, + "learning_rate": 9.90745171404609e-06, + "loss": 0.8283, + "step": 19264 + }, + { + "epoch": 3.429843304843305, + "grad_norm": 0.9762335419654846, + "learning_rate": 9.901378060220646e-06, + "loss": 0.8186, + "step": 19265 + }, + { + "epoch": 3.4300213675213675, + "grad_norm": 0.8388769030570984, + "learning_rate": 9.895306171690554e-06, + "loss": 0.5665, + "step": 19266 + }, + { + "epoch": 3.4301994301994303, + "grad_norm": 0.7391115427017212, + "learning_rate": 9.889236048574746e-06, + "loss": 0.5464, + "step": 19267 + }, + { + "epoch": 3.430377492877493, + "grad_norm": 0.7876406311988831, + "learning_rate": 9.883167690992179e-06, + "loss": 0.3883, + "step": 19268 + }, + { + "epoch": 3.4305555555555554, + "grad_norm": 1.0693145990371704, + "learning_rate": 9.877101099061737e-06, + "loss": 0.7961, + "step": 19269 + }, + { + "epoch": 3.430733618233618, + "grad_norm": 1.2267308235168457, + "learning_rate": 9.871036272902256e-06, + "loss": 0.9196, + "step": 19270 + }, + { + "epoch": 3.430911680911681, + "grad_norm": 1.0068786144256592, + "learning_rate": 9.864973212632645e-06, + "loss": 0.8718, + "step": 19271 + }, + { + "epoch": 3.4310897435897436, + "grad_norm": 0.8946726322174072, + "learning_rate": 9.858911918371605e-06, + "loss": 0.7608, + "step": 19272 + }, + { + "epoch": 3.4312678062678064, + "grad_norm": 0.9454519748687744, + "learning_rate": 9.852852390237966e-06, + "loss": 0.7771, + "step": 19273 + }, + { + "epoch": 3.431445868945869, + "grad_norm": 0.9317994713783264, + "learning_rate": 9.846794628350376e-06, + "loss": 0.8406, + "step": 19274 + }, + { + "epoch": 3.431623931623932, + "grad_norm": 0.9324338436126709, + "learning_rate": 9.840738632827594e-06, + "loss": 0.7005, + "step": 19275 + }, + { + "epoch": 3.431801994301994, + "grad_norm": 0.960617184638977, + "learning_rate": 9.834684403788252e-06, + "loss": 0.8823, + "step": 19276 + }, + { + "epoch": 3.431980056980057, + "grad_norm": 0.89632248878479, + "learning_rate": 9.828631941350963e-06, + "loss": 0.6818, + "step": 19277 + }, + { + "epoch": 3.4321581196581197, + "grad_norm": 1.0242297649383545, + "learning_rate": 9.822581245634321e-06, + "loss": 0.8945, + "step": 19278 + }, + { + "epoch": 3.4323361823361824, + "grad_norm": 1.0989742279052734, + "learning_rate": 9.816532316756855e-06, + "loss": 0.7583, + "step": 19279 + }, + { + "epoch": 3.432514245014245, + "grad_norm": 0.9161012172698975, + "learning_rate": 9.810485154837112e-06, + "loss": 0.695, + "step": 19280 + }, + { + "epoch": 3.4326923076923075, + "grad_norm": 0.8679816126823425, + "learning_rate": 9.804439759993555e-06, + "loss": 0.5813, + "step": 19281 + }, + { + "epoch": 3.4328703703703702, + "grad_norm": 0.8272216320037842, + "learning_rate": 9.798396132344644e-06, + "loss": 0.6652, + "step": 19282 + }, + { + "epoch": 3.433048433048433, + "grad_norm": 0.9027792811393738, + "learning_rate": 9.792354272008775e-06, + "loss": 0.7356, + "step": 19283 + }, + { + "epoch": 3.4332264957264957, + "grad_norm": 0.9031938910484314, + "learning_rate": 9.786314179104339e-06, + "loss": 0.8202, + "step": 19284 + }, + { + "epoch": 3.4334045584045585, + "grad_norm": 1.02913236618042, + "learning_rate": 9.780275853749676e-06, + "loss": 0.6623, + "step": 19285 + }, + { + "epoch": 3.4335826210826212, + "grad_norm": 1.0442672967910767, + "learning_rate": 9.774239296063093e-06, + "loss": 0.8638, + "step": 19286 + }, + { + "epoch": 3.433760683760684, + "grad_norm": 0.9184064865112305, + "learning_rate": 9.768204506162837e-06, + "loss": 0.7879, + "step": 19287 + }, + { + "epoch": 3.4339387464387463, + "grad_norm": 1.2148641347885132, + "learning_rate": 9.762171484167216e-06, + "loss": 0.7737, + "step": 19288 + }, + { + "epoch": 3.434116809116809, + "grad_norm": 0.9303769469261169, + "learning_rate": 9.756140230194355e-06, + "loss": 0.7484, + "step": 19289 + }, + { + "epoch": 3.434294871794872, + "grad_norm": 0.9049912691116333, + "learning_rate": 9.750110744362484e-06, + "loss": 0.6674, + "step": 19290 + }, + { + "epoch": 3.4344729344729346, + "grad_norm": 0.7755575776100159, + "learning_rate": 9.744083026789708e-06, + "loss": 0.6105, + "step": 19291 + }, + { + "epoch": 3.4346509971509973, + "grad_norm": 1.0556714534759521, + "learning_rate": 9.738057077594132e-06, + "loss": 0.8547, + "step": 19292 + }, + { + "epoch": 3.4348290598290596, + "grad_norm": 0.8901651501655579, + "learning_rate": 9.732032896893838e-06, + "loss": 0.6511, + "step": 19293 + }, + { + "epoch": 3.4350071225071224, + "grad_norm": 1.0504934787750244, + "learning_rate": 9.726010484806836e-06, + "loss": 0.8204, + "step": 19294 + }, + { + "epoch": 3.435185185185185, + "grad_norm": 0.9080862402915955, + "learning_rate": 9.719989841451105e-06, + "loss": 0.7081, + "step": 19295 + }, + { + "epoch": 3.435363247863248, + "grad_norm": 0.8364829421043396, + "learning_rate": 9.713970966944674e-06, + "loss": 0.5457, + "step": 19296 + }, + { + "epoch": 3.4355413105413106, + "grad_norm": 1.1203440427780151, + "learning_rate": 9.707953861405394e-06, + "loss": 0.6262, + "step": 19297 + }, + { + "epoch": 3.4357193732193734, + "grad_norm": 0.9322909116744995, + "learning_rate": 9.701938524951237e-06, + "loss": 0.7216, + "step": 19298 + }, + { + "epoch": 3.435897435897436, + "grad_norm": 1.0503484010696411, + "learning_rate": 9.695924957699964e-06, + "loss": 0.6878, + "step": 19299 + }, + { + "epoch": 3.4360754985754984, + "grad_norm": 1.0737676620483398, + "learning_rate": 9.689913159769481e-06, + "loss": 0.9491, + "step": 19300 + }, + { + "epoch": 3.436253561253561, + "grad_norm": 1.0106950998306274, + "learning_rate": 9.683903131277539e-06, + "loss": 0.7918, + "step": 19301 + }, + { + "epoch": 3.436431623931624, + "grad_norm": 0.890433669090271, + "learning_rate": 9.677894872341886e-06, + "loss": 0.8192, + "step": 19302 + }, + { + "epoch": 3.4366096866096867, + "grad_norm": 0.9569767117500305, + "learning_rate": 9.671888383080297e-06, + "loss": 0.6926, + "step": 19303 + }, + { + "epoch": 3.4367877492877494, + "grad_norm": 1.1156257390975952, + "learning_rate": 9.665883663610365e-06, + "loss": 0.9691, + "step": 19304 + }, + { + "epoch": 3.4369658119658117, + "grad_norm": 1.0798922777175903, + "learning_rate": 9.659880714049819e-06, + "loss": 0.8248, + "step": 19305 + }, + { + "epoch": 3.4371438746438745, + "grad_norm": 1.134392261505127, + "learning_rate": 9.653879534516242e-06, + "loss": 0.7283, + "step": 19306 + }, + { + "epoch": 3.4373219373219372, + "grad_norm": 0.9919711947441101, + "learning_rate": 9.647880125127217e-06, + "loss": 0.7475, + "step": 19307 + }, + { + "epoch": 3.4375, + "grad_norm": 0.9417802095413208, + "learning_rate": 9.641882486000287e-06, + "loss": 0.7534, + "step": 19308 + }, + { + "epoch": 3.4376780626780628, + "grad_norm": 1.0912595987319946, + "learning_rate": 9.635886617252975e-06, + "loss": 0.6257, + "step": 19309 + }, + { + "epoch": 3.4378561253561255, + "grad_norm": 0.8998816013336182, + "learning_rate": 9.629892519002747e-06, + "loss": 0.7533, + "step": 19310 + }, + { + "epoch": 3.4380341880341883, + "grad_norm": 1.1993279457092285, + "learning_rate": 9.623900191367041e-06, + "loss": 0.7533, + "step": 19311 + }, + { + "epoch": 3.4382122507122506, + "grad_norm": 0.9719237089157104, + "learning_rate": 9.617909634463263e-06, + "loss": 0.7894, + "step": 19312 + }, + { + "epoch": 3.4383903133903133, + "grad_norm": 0.9022089838981628, + "learning_rate": 9.611920848408817e-06, + "loss": 0.7212, + "step": 19313 + }, + { + "epoch": 3.438568376068376, + "grad_norm": 0.9632169008255005, + "learning_rate": 9.60593383332099e-06, + "loss": 0.8404, + "step": 19314 + }, + { + "epoch": 3.438746438746439, + "grad_norm": 0.9799385070800781, + "learning_rate": 9.599948589317131e-06, + "loss": 0.7254, + "step": 19315 + }, + { + "epoch": 3.4389245014245016, + "grad_norm": 0.9019197225570679, + "learning_rate": 9.59396511651448e-06, + "loss": 0.6593, + "step": 19316 + }, + { + "epoch": 3.439102564102564, + "grad_norm": 1.1352930068969727, + "learning_rate": 9.587983415030288e-06, + "loss": 0.9686, + "step": 19317 + }, + { + "epoch": 3.4392806267806266, + "grad_norm": 0.8560138940811157, + "learning_rate": 9.58200348498175e-06, + "loss": 0.7901, + "step": 19318 + }, + { + "epoch": 3.4394586894586894, + "grad_norm": 0.8698241710662842, + "learning_rate": 9.576025326485993e-06, + "loss": 0.6078, + "step": 19319 + }, + { + "epoch": 3.439636752136752, + "grad_norm": 0.9737006425857544, + "learning_rate": 9.570048939660236e-06, + "loss": 0.9071, + "step": 19320 + }, + { + "epoch": 3.439814814814815, + "grad_norm": 0.8899868726730347, + "learning_rate": 9.564074324621475e-06, + "loss": 0.6993, + "step": 19321 + }, + { + "epoch": 3.4399928774928776, + "grad_norm": 0.8926196694374084, + "learning_rate": 9.558101481486826e-06, + "loss": 0.533, + "step": 19322 + }, + { + "epoch": 3.4401709401709404, + "grad_norm": 0.8816601634025574, + "learning_rate": 9.552130410373306e-06, + "loss": 0.8344, + "step": 19323 + }, + { + "epoch": 3.4403490028490027, + "grad_norm": 1.0575761795043945, + "learning_rate": 9.5461611113979e-06, + "loss": 0.9992, + "step": 19324 + }, + { + "epoch": 3.4405270655270654, + "grad_norm": 0.9737523198127747, + "learning_rate": 9.54019358467756e-06, + "loss": 0.8928, + "step": 19325 + }, + { + "epoch": 3.440705128205128, + "grad_norm": 1.1094090938568115, + "learning_rate": 9.534227830329224e-06, + "loss": 0.846, + "step": 19326 + }, + { + "epoch": 3.440883190883191, + "grad_norm": 0.8708354830741882, + "learning_rate": 9.528263848469742e-06, + "loss": 0.7715, + "step": 19327 + }, + { + "epoch": 3.4410612535612537, + "grad_norm": 0.916589081287384, + "learning_rate": 9.522301639216024e-06, + "loss": 0.6541, + "step": 19328 + }, + { + "epoch": 3.441239316239316, + "grad_norm": 0.8238539695739746, + "learning_rate": 9.51634120268482e-06, + "loss": 0.7523, + "step": 19329 + }, + { + "epoch": 3.4414173789173788, + "grad_norm": 0.9754942059516907, + "learning_rate": 9.510382538992969e-06, + "loss": 0.7002, + "step": 19330 + }, + { + "epoch": 3.4415954415954415, + "grad_norm": 0.898399293422699, + "learning_rate": 9.504425648257198e-06, + "loss": 0.7644, + "step": 19331 + }, + { + "epoch": 3.4417735042735043, + "grad_norm": 0.8976600766181946, + "learning_rate": 9.498470530594217e-06, + "loss": 0.7345, + "step": 19332 + }, + { + "epoch": 3.441951566951567, + "grad_norm": 1.0468547344207764, + "learning_rate": 9.492517186120697e-06, + "loss": 0.7121, + "step": 19333 + }, + { + "epoch": 3.4421296296296298, + "grad_norm": 1.185500144958496, + "learning_rate": 9.486565614953302e-06, + "loss": 0.8665, + "step": 19334 + }, + { + "epoch": 3.4423076923076925, + "grad_norm": 1.0200905799865723, + "learning_rate": 9.480615817208615e-06, + "loss": 0.7206, + "step": 19335 + }, + { + "epoch": 3.442485754985755, + "grad_norm": 0.9022289514541626, + "learning_rate": 9.474667793003234e-06, + "loss": 0.6764, + "step": 19336 + }, + { + "epoch": 3.4426638176638176, + "grad_norm": 0.9930318593978882, + "learning_rate": 9.468721542453662e-06, + "loss": 0.7878, + "step": 19337 + }, + { + "epoch": 3.4428418803418803, + "grad_norm": 0.9643216729164124, + "learning_rate": 9.462777065676476e-06, + "loss": 0.8636, + "step": 19338 + }, + { + "epoch": 3.443019943019943, + "grad_norm": 1.0845335721969604, + "learning_rate": 9.456834362788059e-06, + "loss": 0.6301, + "step": 19339 + }, + { + "epoch": 3.443198005698006, + "grad_norm": 0.8617258071899414, + "learning_rate": 9.450893433904894e-06, + "loss": 0.6117, + "step": 19340 + }, + { + "epoch": 3.443376068376068, + "grad_norm": 1.0213714838027954, + "learning_rate": 9.444954279143382e-06, + "loss": 0.7732, + "step": 19341 + }, + { + "epoch": 3.443554131054131, + "grad_norm": 1.0030566453933716, + "learning_rate": 9.439016898619857e-06, + "loss": 0.6772, + "step": 19342 + }, + { + "epoch": 3.4437321937321936, + "grad_norm": 0.8375689387321472, + "learning_rate": 9.433081292450708e-06, + "loss": 0.6771, + "step": 19343 + }, + { + "epoch": 3.4439102564102564, + "grad_norm": 0.8299211859703064, + "learning_rate": 9.42714746075216e-06, + "loss": 0.649, + "step": 19344 + }, + { + "epoch": 3.444088319088319, + "grad_norm": 0.9578790068626404, + "learning_rate": 9.421215403640549e-06, + "loss": 0.5323, + "step": 19345 + }, + { + "epoch": 3.444266381766382, + "grad_norm": 1.0069869756698608, + "learning_rate": 9.415285121232021e-06, + "loss": 0.694, + "step": 19346 + }, + { + "epoch": 3.4444444444444446, + "grad_norm": 1.0765708684921265, + "learning_rate": 9.409356613642817e-06, + "loss": 0.7485, + "step": 19347 + }, + { + "epoch": 3.444622507122507, + "grad_norm": 1.0542240142822266, + "learning_rate": 9.40342988098909e-06, + "loss": 0.9368, + "step": 19348 + }, + { + "epoch": 3.4448005698005697, + "grad_norm": 0.8405135869979858, + "learning_rate": 9.397504923386957e-06, + "loss": 0.5911, + "step": 19349 + }, + { + "epoch": 3.4449786324786325, + "grad_norm": 0.8794723153114319, + "learning_rate": 9.391581740952516e-06, + "loss": 0.8313, + "step": 19350 + }, + { + "epoch": 3.445156695156695, + "grad_norm": 1.132588505744934, + "learning_rate": 9.385660333801793e-06, + "loss": 0.7254, + "step": 19351 + }, + { + "epoch": 3.445334757834758, + "grad_norm": 0.958625078201294, + "learning_rate": 9.379740702050809e-06, + "loss": 0.7199, + "step": 19352 + }, + { + "epoch": 3.4455128205128207, + "grad_norm": 1.0391316413879395, + "learning_rate": 9.373822845815593e-06, + "loss": 0.8571, + "step": 19353 + }, + { + "epoch": 3.445690883190883, + "grad_norm": 1.0981941223144531, + "learning_rate": 9.36790676521202e-06, + "loss": 0.8233, + "step": 19354 + }, + { + "epoch": 3.4458689458689458, + "grad_norm": 0.8751938939094543, + "learning_rate": 9.361992460356084e-06, + "loss": 0.708, + "step": 19355 + }, + { + "epoch": 3.4460470085470085, + "grad_norm": 1.240193486213684, + "learning_rate": 9.356079931363582e-06, + "loss": 0.9598, + "step": 19356 + }, + { + "epoch": 3.4462250712250713, + "grad_norm": 0.985924243927002, + "learning_rate": 9.350169178350421e-06, + "loss": 0.7641, + "step": 19357 + }, + { + "epoch": 3.446403133903134, + "grad_norm": 1.125505805015564, + "learning_rate": 9.344260201432375e-06, + "loss": 0.9037, + "step": 19358 + }, + { + "epoch": 3.4465811965811968, + "grad_norm": 1.1013590097427368, + "learning_rate": 9.33835300072522e-06, + "loss": 0.8685, + "step": 19359 + }, + { + "epoch": 3.446759259259259, + "grad_norm": 0.815940797328949, + "learning_rate": 9.332447576344739e-06, + "loss": 0.6551, + "step": 19360 + }, + { + "epoch": 3.446937321937322, + "grad_norm": 1.1269092559814453, + "learning_rate": 9.326543928406573e-06, + "loss": 0.9763, + "step": 19361 + }, + { + "epoch": 3.4471153846153846, + "grad_norm": 0.9986848831176758, + "learning_rate": 9.320642057026429e-06, + "loss": 0.8283, + "step": 19362 + }, + { + "epoch": 3.4472934472934473, + "grad_norm": 0.9624196290969849, + "learning_rate": 9.31474196231994e-06, + "loss": 0.8278, + "step": 19363 + }, + { + "epoch": 3.44747150997151, + "grad_norm": 0.914781391620636, + "learning_rate": 9.308843644402687e-06, + "loss": 0.7268, + "step": 19364 + }, + { + "epoch": 3.447649572649573, + "grad_norm": 0.9857885241508484, + "learning_rate": 9.302947103390258e-06, + "loss": 0.5724, + "step": 19365 + }, + { + "epoch": 3.447827635327635, + "grad_norm": 1.0420119762420654, + "learning_rate": 9.297052339398182e-06, + "loss": 0.8149, + "step": 19366 + }, + { + "epoch": 3.448005698005698, + "grad_norm": 0.8537631630897522, + "learning_rate": 9.291159352541912e-06, + "loss": 0.5231, + "step": 19367 + }, + { + "epoch": 3.4481837606837606, + "grad_norm": 1.0447754859924316, + "learning_rate": 9.285268142936988e-06, + "loss": 0.7794, + "step": 19368 + }, + { + "epoch": 3.4483618233618234, + "grad_norm": 0.9043522477149963, + "learning_rate": 9.27937871069876e-06, + "loss": 0.7547, + "step": 19369 + }, + { + "epoch": 3.448539886039886, + "grad_norm": 0.9774514436721802, + "learning_rate": 9.273491055942673e-06, + "loss": 0.6994, + "step": 19370 + }, + { + "epoch": 3.448717948717949, + "grad_norm": 1.0255252122879028, + "learning_rate": 9.267605178784033e-06, + "loss": 1.0779, + "step": 19371 + }, + { + "epoch": 3.448896011396011, + "grad_norm": 0.9238240122795105, + "learning_rate": 9.261721079338214e-06, + "loss": 0.7204, + "step": 19372 + }, + { + "epoch": 3.449074074074074, + "grad_norm": 0.9913932085037231, + "learning_rate": 9.25583875772047e-06, + "loss": 0.7491, + "step": 19373 + }, + { + "epoch": 3.4492521367521367, + "grad_norm": 0.9521636366844177, + "learning_rate": 9.249958214046062e-06, + "loss": 0.7428, + "step": 19374 + }, + { + "epoch": 3.4494301994301995, + "grad_norm": 0.8698529601097107, + "learning_rate": 9.244079448430199e-06, + "loss": 0.567, + "step": 19375 + }, + { + "epoch": 3.449608262108262, + "grad_norm": 1.0148518085479736, + "learning_rate": 9.238202460988077e-06, + "loss": 0.6805, + "step": 19376 + }, + { + "epoch": 3.449786324786325, + "grad_norm": 1.0295552015304565, + "learning_rate": 9.232327251834827e-06, + "loss": 0.7211, + "step": 19377 + }, + { + "epoch": 3.4499643874643873, + "grad_norm": 0.88274747133255, + "learning_rate": 9.22645382108559e-06, + "loss": 0.7126, + "step": 19378 + }, + { + "epoch": 3.45014245014245, + "grad_norm": 1.062113881111145, + "learning_rate": 9.220582168855397e-06, + "loss": 0.9362, + "step": 19379 + }, + { + "epoch": 3.4503205128205128, + "grad_norm": 0.811355471611023, + "learning_rate": 9.214712295259342e-06, + "loss": 0.6212, + "step": 19380 + }, + { + "epoch": 3.4504985754985755, + "grad_norm": 1.0234562158584595, + "learning_rate": 9.208844200412403e-06, + "loss": 0.8328, + "step": 19381 + }, + { + "epoch": 3.4506766381766383, + "grad_norm": 0.9234959483146667, + "learning_rate": 9.202977884429554e-06, + "loss": 0.7139, + "step": 19382 + }, + { + "epoch": 3.450854700854701, + "grad_norm": 0.8315097093582153, + "learning_rate": 9.197113347425745e-06, + "loss": 0.6782, + "step": 19383 + }, + { + "epoch": 3.451032763532764, + "grad_norm": 1.0691921710968018, + "learning_rate": 9.191250589515866e-06, + "loss": 0.8085, + "step": 19384 + }, + { + "epoch": 3.451210826210826, + "grad_norm": 0.9873775839805603, + "learning_rate": 9.18538961081481e-06, + "loss": 0.9339, + "step": 19385 + }, + { + "epoch": 3.451388888888889, + "grad_norm": 1.058555006980896, + "learning_rate": 9.179530411437365e-06, + "loss": 0.847, + "step": 19386 + }, + { + "epoch": 3.4515669515669516, + "grad_norm": 0.8825174570083618, + "learning_rate": 9.173672991498384e-06, + "loss": 0.7246, + "step": 19387 + }, + { + "epoch": 3.4517450142450143, + "grad_norm": 0.8538999557495117, + "learning_rate": 9.167817351112596e-06, + "loss": 0.7268, + "step": 19388 + }, + { + "epoch": 3.451923076923077, + "grad_norm": 0.9254639744758606, + "learning_rate": 9.161963490394743e-06, + "loss": 0.6828, + "step": 19389 + }, + { + "epoch": 3.4521011396011394, + "grad_norm": 1.062555193901062, + "learning_rate": 9.156111409459512e-06, + "loss": 0.8251, + "step": 19390 + }, + { + "epoch": 3.452279202279202, + "grad_norm": 0.8910902142524719, + "learning_rate": 9.150261108421575e-06, + "loss": 0.8508, + "step": 19391 + }, + { + "epoch": 3.452457264957265, + "grad_norm": 0.9884508848190308, + "learning_rate": 9.144412587395534e-06, + "loss": 0.8475, + "step": 19392 + }, + { + "epoch": 3.4526353276353277, + "grad_norm": 1.0337510108947754, + "learning_rate": 9.138565846496005e-06, + "loss": 0.6761, + "step": 19393 + }, + { + "epoch": 3.4528133903133904, + "grad_norm": 0.9345211982727051, + "learning_rate": 9.13272088583751e-06, + "loss": 0.7458, + "step": 19394 + }, + { + "epoch": 3.452991452991453, + "grad_norm": 0.9515882134437561, + "learning_rate": 9.12687770553462e-06, + "loss": 0.815, + "step": 19395 + }, + { + "epoch": 3.453169515669516, + "grad_norm": 0.9667288661003113, + "learning_rate": 9.121036305701746e-06, + "loss": 0.6266, + "step": 19396 + }, + { + "epoch": 3.453347578347578, + "grad_norm": 0.940667450428009, + "learning_rate": 9.11519668645341e-06, + "loss": 0.5882, + "step": 19397 + }, + { + "epoch": 3.453525641025641, + "grad_norm": 0.9668877720832825, + "learning_rate": 9.109358847904003e-06, + "loss": 0.8559, + "step": 19398 + }, + { + "epoch": 3.4537037037037037, + "grad_norm": 0.904509961605072, + "learning_rate": 9.103522790167874e-06, + "loss": 0.7116, + "step": 19399 + }, + { + "epoch": 3.4538817663817665, + "grad_norm": 0.9343373775482178, + "learning_rate": 9.097688513359425e-06, + "loss": 0.5469, + "step": 19400 + }, + { + "epoch": 3.4540598290598292, + "grad_norm": 1.025084137916565, + "learning_rate": 9.091856017592915e-06, + "loss": 0.749, + "step": 19401 + }, + { + "epoch": 3.4542378917378915, + "grad_norm": 1.0344691276550293, + "learning_rate": 9.086025302982648e-06, + "loss": 0.8593, + "step": 19402 + }, + { + "epoch": 3.4544159544159543, + "grad_norm": 0.9288195371627808, + "learning_rate": 9.080196369642858e-06, + "loss": 0.6745, + "step": 19403 + }, + { + "epoch": 3.454594017094017, + "grad_norm": 0.8177899122238159, + "learning_rate": 9.07436921768775e-06, + "loss": 0.5978, + "step": 19404 + }, + { + "epoch": 3.45477207977208, + "grad_norm": 0.8998305201530457, + "learning_rate": 9.068543847231503e-06, + "loss": 0.779, + "step": 19405 + }, + { + "epoch": 3.4549501424501425, + "grad_norm": 0.9722973108291626, + "learning_rate": 9.062720258388246e-06, + "loss": 0.7639, + "step": 19406 + }, + { + "epoch": 3.4551282051282053, + "grad_norm": 1.0662795305252075, + "learning_rate": 9.056898451272077e-06, + "loss": 0.6982, + "step": 19407 + }, + { + "epoch": 3.455306267806268, + "grad_norm": 0.9679057002067566, + "learning_rate": 9.051078425997062e-06, + "loss": 0.7254, + "step": 19408 + }, + { + "epoch": 3.4554843304843303, + "grad_norm": 1.0791397094726562, + "learning_rate": 9.045260182677217e-06, + "loss": 0.8541, + "step": 19409 + }, + { + "epoch": 3.455662393162393, + "grad_norm": 0.9533210396766663, + "learning_rate": 9.039443721426589e-06, + "loss": 0.8441, + "step": 19410 + }, + { + "epoch": 3.455840455840456, + "grad_norm": 0.9457281827926636, + "learning_rate": 9.033629042359081e-06, + "loss": 0.6461, + "step": 19411 + }, + { + "epoch": 3.4560185185185186, + "grad_norm": 0.9577490091323853, + "learning_rate": 9.027816145588664e-06, + "loss": 0.7646, + "step": 19412 + }, + { + "epoch": 3.4561965811965814, + "grad_norm": 0.8724551796913147, + "learning_rate": 9.022005031229196e-06, + "loss": 0.6517, + "step": 19413 + }, + { + "epoch": 3.4563746438746437, + "grad_norm": 0.9715121984481812, + "learning_rate": 9.016195699394559e-06, + "loss": 0.592, + "step": 19414 + }, + { + "epoch": 3.4565527065527064, + "grad_norm": 0.9042280912399292, + "learning_rate": 9.010388150198567e-06, + "loss": 0.8143, + "step": 19415 + }, + { + "epoch": 3.456730769230769, + "grad_norm": 0.9272587299346924, + "learning_rate": 9.004582383755e-06, + "loss": 0.9227, + "step": 19416 + }, + { + "epoch": 3.456908831908832, + "grad_norm": 1.0067778825759888, + "learning_rate": 8.998778400177622e-06, + "loss": 0.7434, + "step": 19417 + }, + { + "epoch": 3.4570868945868947, + "grad_norm": 0.8889135122299194, + "learning_rate": 8.992976199580117e-06, + "loss": 0.7774, + "step": 19418 + }, + { + "epoch": 3.4572649572649574, + "grad_norm": 0.9987656474113464, + "learning_rate": 8.987175782076217e-06, + "loss": 0.7912, + "step": 19419 + }, + { + "epoch": 3.45744301994302, + "grad_norm": 1.0001306533813477, + "learning_rate": 8.981377147779535e-06, + "loss": 0.8925, + "step": 19420 + }, + { + "epoch": 3.4576210826210825, + "grad_norm": 1.069911003112793, + "learning_rate": 8.975580296803709e-06, + "loss": 0.8013, + "step": 19421 + }, + { + "epoch": 3.4577991452991452, + "grad_norm": 0.9291098713874817, + "learning_rate": 8.969785229262284e-06, + "loss": 0.9167, + "step": 19422 + }, + { + "epoch": 3.457977207977208, + "grad_norm": 1.0406343936920166, + "learning_rate": 8.963991945268825e-06, + "loss": 0.7271, + "step": 19423 + }, + { + "epoch": 3.4581552706552707, + "grad_norm": 1.0367125272750854, + "learning_rate": 8.958200444936815e-06, + "loss": 0.6179, + "step": 19424 + }, + { + "epoch": 3.4583333333333335, + "grad_norm": 1.0270168781280518, + "learning_rate": 8.952410728379779e-06, + "loss": 0.555, + "step": 19425 + }, + { + "epoch": 3.458511396011396, + "grad_norm": 0.9439095854759216, + "learning_rate": 8.94662279571109e-06, + "loss": 0.7084, + "step": 19426 + }, + { + "epoch": 3.4586894586894585, + "grad_norm": 0.934548556804657, + "learning_rate": 8.940836647044216e-06, + "loss": 0.6366, + "step": 19427 + }, + { + "epoch": 3.4588675213675213, + "grad_norm": 1.0294036865234375, + "learning_rate": 8.93505228249244e-06, + "loss": 0.6797, + "step": 19428 + }, + { + "epoch": 3.459045584045584, + "grad_norm": 0.7903768420219421, + "learning_rate": 8.929269702169174e-06, + "loss": 0.7274, + "step": 19429 + }, + { + "epoch": 3.459223646723647, + "grad_norm": 0.9741223454475403, + "learning_rate": 8.923488906187683e-06, + "loss": 0.7743, + "step": 19430 + }, + { + "epoch": 3.4594017094017095, + "grad_norm": 0.9716152548789978, + "learning_rate": 8.917709894661231e-06, + "loss": 0.5779, + "step": 19431 + }, + { + "epoch": 3.4595797720797723, + "grad_norm": 0.9969064593315125, + "learning_rate": 8.911932667703038e-06, + "loss": 0.7671, + "step": 19432 + }, + { + "epoch": 3.4597578347578346, + "grad_norm": 0.7803093194961548, + "learning_rate": 8.906157225426315e-06, + "loss": 0.4898, + "step": 19433 + }, + { + "epoch": 3.4599358974358974, + "grad_norm": 0.9883405566215515, + "learning_rate": 8.900383567944192e-06, + "loss": 0.7084, + "step": 19434 + }, + { + "epoch": 3.46011396011396, + "grad_norm": 0.8429387211799622, + "learning_rate": 8.894611695369836e-06, + "loss": 0.7378, + "step": 19435 + }, + { + "epoch": 3.460292022792023, + "grad_norm": 1.0472184419631958, + "learning_rate": 8.888841607816278e-06, + "loss": 0.9414, + "step": 19436 + }, + { + "epoch": 3.4604700854700856, + "grad_norm": 0.8618428111076355, + "learning_rate": 8.883073305396627e-06, + "loss": 0.6176, + "step": 19437 + }, + { + "epoch": 3.460648148148148, + "grad_norm": 1.0710008144378662, + "learning_rate": 8.877306788223872e-06, + "loss": 0.9157, + "step": 19438 + }, + { + "epoch": 3.4608262108262107, + "grad_norm": 0.9094607830047607, + "learning_rate": 8.871542056410976e-06, + "loss": 0.7429, + "step": 19439 + }, + { + "epoch": 3.4610042735042734, + "grad_norm": 0.8668854832649231, + "learning_rate": 8.865779110070949e-06, + "loss": 0.6732, + "step": 19440 + }, + { + "epoch": 3.461182336182336, + "grad_norm": 0.9430466890335083, + "learning_rate": 8.860017949316634e-06, + "loss": 0.6454, + "step": 19441 + }, + { + "epoch": 3.461360398860399, + "grad_norm": 0.9155656695365906, + "learning_rate": 8.854258574260977e-06, + "loss": 0.6573, + "step": 19442 + }, + { + "epoch": 3.4615384615384617, + "grad_norm": 1.0212770700454712, + "learning_rate": 8.848500985016739e-06, + "loss": 0.8552, + "step": 19443 + }, + { + "epoch": 3.4617165242165244, + "grad_norm": 0.9474434852600098, + "learning_rate": 8.8427451816968e-06, + "loss": 0.7882, + "step": 19444 + }, + { + "epoch": 3.4618945868945867, + "grad_norm": 0.9351694583892822, + "learning_rate": 8.836991164413898e-06, + "loss": 0.6336, + "step": 19445 + }, + { + "epoch": 3.4620726495726495, + "grad_norm": 1.0357493162155151, + "learning_rate": 8.831238933280795e-06, + "loss": 0.8264, + "step": 19446 + }, + { + "epoch": 3.4622507122507122, + "grad_norm": 1.0070866346359253, + "learning_rate": 8.82548848841016e-06, + "loss": 0.8509, + "step": 19447 + }, + { + "epoch": 3.462428774928775, + "grad_norm": 0.9792636036872864, + "learning_rate": 8.819739829914685e-06, + "loss": 0.7739, + "step": 19448 + }, + { + "epoch": 3.4626068376068377, + "grad_norm": 1.0221115350723267, + "learning_rate": 8.81399295790699e-06, + "loss": 0.7329, + "step": 19449 + }, + { + "epoch": 3.4627849002849, + "grad_norm": 1.0703456401824951, + "learning_rate": 8.80824787249971e-06, + "loss": 0.8073, + "step": 19450 + }, + { + "epoch": 3.462962962962963, + "grad_norm": 0.9054359793663025, + "learning_rate": 8.802504573805336e-06, + "loss": 0.8686, + "step": 19451 + }, + { + "epoch": 3.4631410256410255, + "grad_norm": 1.019992470741272, + "learning_rate": 8.796763061936486e-06, + "loss": 1.0346, + "step": 19452 + }, + { + "epoch": 3.4633190883190883, + "grad_norm": 0.9106883406639099, + "learning_rate": 8.791023337005555e-06, + "loss": 0.6639, + "step": 19453 + }, + { + "epoch": 3.463497150997151, + "grad_norm": 0.9866195321083069, + "learning_rate": 8.785285399125077e-06, + "loss": 0.8251, + "step": 19454 + }, + { + "epoch": 3.463675213675214, + "grad_norm": 1.026564359664917, + "learning_rate": 8.779549248407449e-06, + "loss": 0.804, + "step": 19455 + }, + { + "epoch": 3.4638532763532766, + "grad_norm": 1.0608559846878052, + "learning_rate": 8.773814884965058e-06, + "loss": 0.8622, + "step": 19456 + }, + { + "epoch": 3.464031339031339, + "grad_norm": 0.8756099939346313, + "learning_rate": 8.768082308910264e-06, + "loss": 0.7148, + "step": 19457 + }, + { + "epoch": 3.4642094017094016, + "grad_norm": 0.9199277758598328, + "learning_rate": 8.76235152035536e-06, + "loss": 0.8529, + "step": 19458 + }, + { + "epoch": 3.4643874643874644, + "grad_norm": 1.0407079458236694, + "learning_rate": 8.75662251941266e-06, + "loss": 0.7854, + "step": 19459 + }, + { + "epoch": 3.464565527065527, + "grad_norm": 0.8924119472503662, + "learning_rate": 8.750895306194407e-06, + "loss": 0.8325, + "step": 19460 + }, + { + "epoch": 3.46474358974359, + "grad_norm": 0.9075109362602234, + "learning_rate": 8.745169880812808e-06, + "loss": 0.716, + "step": 19461 + }, + { + "epoch": 3.464921652421652, + "grad_norm": 1.1008402109146118, + "learning_rate": 8.739446243380034e-06, + "loss": 0.83, + "step": 19462 + }, + { + "epoch": 3.465099715099715, + "grad_norm": 0.9462727308273315, + "learning_rate": 8.733724394008236e-06, + "loss": 0.8705, + "step": 19463 + }, + { + "epoch": 3.4652777777777777, + "grad_norm": 0.9452762007713318, + "learning_rate": 8.728004332809514e-06, + "loss": 0.6937, + "step": 19464 + }, + { + "epoch": 3.4654558404558404, + "grad_norm": 0.899398922920227, + "learning_rate": 8.722286059895957e-06, + "loss": 0.7192, + "step": 19465 + }, + { + "epoch": 3.465633903133903, + "grad_norm": 0.8906122446060181, + "learning_rate": 8.716569575379563e-06, + "loss": 0.8128, + "step": 19466 + }, + { + "epoch": 3.465811965811966, + "grad_norm": 0.8970988988876343, + "learning_rate": 8.710854879372398e-06, + "loss": 0.7183, + "step": 19467 + }, + { + "epoch": 3.4659900284900287, + "grad_norm": 0.8346308469772339, + "learning_rate": 8.705141971986363e-06, + "loss": 0.8172, + "step": 19468 + }, + { + "epoch": 3.466168091168091, + "grad_norm": 1.137384057044983, + "learning_rate": 8.699430853333446e-06, + "loss": 0.8213, + "step": 19469 + }, + { + "epoch": 3.4663461538461537, + "grad_norm": 1.241615653038025, + "learning_rate": 8.693721523525522e-06, + "loss": 0.9548, + "step": 19470 + }, + { + "epoch": 3.4665242165242165, + "grad_norm": 0.9010623097419739, + "learning_rate": 8.688013982674436e-06, + "loss": 0.9466, + "step": 19471 + }, + { + "epoch": 3.4667022792022792, + "grad_norm": 1.0348331928253174, + "learning_rate": 8.682308230892044e-06, + "loss": 0.632, + "step": 19472 + }, + { + "epoch": 3.466880341880342, + "grad_norm": 0.9175500273704529, + "learning_rate": 8.676604268290134e-06, + "loss": 0.7857, + "step": 19473 + }, + { + "epoch": 3.4670584045584047, + "grad_norm": 0.9245565533638, + "learning_rate": 8.670902094980426e-06, + "loss": 0.6556, + "step": 19474 + }, + { + "epoch": 3.467236467236467, + "grad_norm": 0.9198375940322876, + "learning_rate": 8.66520171107471e-06, + "loss": 0.9424, + "step": 19475 + }, + { + "epoch": 3.46741452991453, + "grad_norm": 1.0737100839614868, + "learning_rate": 8.659503116684598e-06, + "loss": 0.7627, + "step": 19476 + }, + { + "epoch": 3.4675925925925926, + "grad_norm": 0.9110404253005981, + "learning_rate": 8.653806311921809e-06, + "loss": 0.6945, + "step": 19477 + }, + { + "epoch": 3.4677706552706553, + "grad_norm": 1.0172821283340454, + "learning_rate": 8.648111296897909e-06, + "loss": 0.9387, + "step": 19478 + }, + { + "epoch": 3.467948717948718, + "grad_norm": 0.9516095519065857, + "learning_rate": 8.642418071724512e-06, + "loss": 0.7702, + "step": 19479 + }, + { + "epoch": 3.468126780626781, + "grad_norm": 0.950921893119812, + "learning_rate": 8.636726636513148e-06, + "loss": 0.9081, + "step": 19480 + }, + { + "epoch": 3.468304843304843, + "grad_norm": 0.871261715888977, + "learning_rate": 8.631036991375318e-06, + "loss": 0.7908, + "step": 19481 + }, + { + "epoch": 3.468482905982906, + "grad_norm": 0.9214411377906799, + "learning_rate": 8.625349136422557e-06, + "loss": 0.7855, + "step": 19482 + }, + { + "epoch": 3.4686609686609686, + "grad_norm": 1.0029644966125488, + "learning_rate": 8.619663071766227e-06, + "loss": 0.7906, + "step": 19483 + }, + { + "epoch": 3.4688390313390314, + "grad_norm": 1.0286318063735962, + "learning_rate": 8.613978797517797e-06, + "loss": 0.8748, + "step": 19484 + }, + { + "epoch": 3.469017094017094, + "grad_norm": 0.9796161651611328, + "learning_rate": 8.6082963137886e-06, + "loss": 0.7109, + "step": 19485 + }, + { + "epoch": 3.469195156695157, + "grad_norm": 0.9568044543266296, + "learning_rate": 8.602615620690001e-06, + "loss": 0.8721, + "step": 19486 + }, + { + "epoch": 3.469373219373219, + "grad_norm": 1.0369888544082642, + "learning_rate": 8.596936718333281e-06, + "loss": 0.7648, + "step": 19487 + }, + { + "epoch": 3.469551282051282, + "grad_norm": 0.9487068057060242, + "learning_rate": 8.591259606829716e-06, + "loss": 0.677, + "step": 19488 + }, + { + "epoch": 3.4697293447293447, + "grad_norm": 1.0003830194473267, + "learning_rate": 8.585584286290527e-06, + "loss": 0.7131, + "step": 19489 + }, + { + "epoch": 3.4699074074074074, + "grad_norm": 0.9841815829277039, + "learning_rate": 8.579910756826925e-06, + "loss": 0.9088, + "step": 19490 + }, + { + "epoch": 3.47008547008547, + "grad_norm": 0.7646492123603821, + "learning_rate": 8.574239018550035e-06, + "loss": 0.4016, + "step": 19491 + }, + { + "epoch": 3.470263532763533, + "grad_norm": 0.9418407678604126, + "learning_rate": 8.568569071571053e-06, + "loss": 0.6413, + "step": 19492 + }, + { + "epoch": 3.4704415954415953, + "grad_norm": 1.078834891319275, + "learning_rate": 8.562900916000993e-06, + "loss": 0.8014, + "step": 19493 + }, + { + "epoch": 3.470619658119658, + "grad_norm": 0.9450939297676086, + "learning_rate": 8.557234551950966e-06, + "loss": 0.735, + "step": 19494 + }, + { + "epoch": 3.4707977207977208, + "grad_norm": 0.9362136125564575, + "learning_rate": 8.55156997953197e-06, + "loss": 0.8151, + "step": 19495 + }, + { + "epoch": 3.4709757834757835, + "grad_norm": 0.8481518626213074, + "learning_rate": 8.545907198854986e-06, + "loss": 0.5628, + "step": 19496 + }, + { + "epoch": 3.4711538461538463, + "grad_norm": 0.968862771987915, + "learning_rate": 8.540246210030978e-06, + "loss": 0.8156, + "step": 19497 + }, + { + "epoch": 3.471331908831909, + "grad_norm": 0.8841079473495483, + "learning_rate": 8.534587013170836e-06, + "loss": 0.8872, + "step": 19498 + }, + { + "epoch": 3.4715099715099713, + "grad_norm": 0.8855252265930176, + "learning_rate": 8.528929608385494e-06, + "loss": 0.8796, + "step": 19499 + }, + { + "epoch": 3.471688034188034, + "grad_norm": 0.925837516784668, + "learning_rate": 8.523273995785719e-06, + "loss": 0.8763, + "step": 19500 + }, + { + "epoch": 3.471866096866097, + "grad_norm": 0.8930026292800903, + "learning_rate": 8.517620175482388e-06, + "loss": 0.74, + "step": 19501 + }, + { + "epoch": 3.4720441595441596, + "grad_norm": 0.9781539440155029, + "learning_rate": 8.511968147586246e-06, + "loss": 0.5703, + "step": 19502 + }, + { + "epoch": 3.4722222222222223, + "grad_norm": 0.901134192943573, + "learning_rate": 8.50631791220805e-06, + "loss": 0.658, + "step": 19503 + }, + { + "epoch": 3.472400284900285, + "grad_norm": 1.052512526512146, + "learning_rate": 8.500669469458478e-06, + "loss": 0.6627, + "step": 19504 + }, + { + "epoch": 3.472578347578348, + "grad_norm": 1.0030267238616943, + "learning_rate": 8.495022819448228e-06, + "loss": 0.7527, + "step": 19505 + }, + { + "epoch": 3.47275641025641, + "grad_norm": 0.936089277267456, + "learning_rate": 8.489377962287893e-06, + "loss": 0.787, + "step": 19506 + }, + { + "epoch": 3.472934472934473, + "grad_norm": 1.0984883308410645, + "learning_rate": 8.483734898088135e-06, + "loss": 0.8435, + "step": 19507 + }, + { + "epoch": 3.4731125356125356, + "grad_norm": 0.9354465007781982, + "learning_rate": 8.47809362695946e-06, + "loss": 0.7568, + "step": 19508 + }, + { + "epoch": 3.4732905982905984, + "grad_norm": 0.9807959198951721, + "learning_rate": 8.472454149012431e-06, + "loss": 0.8719, + "step": 19509 + }, + { + "epoch": 3.473468660968661, + "grad_norm": 0.955855131149292, + "learning_rate": 8.46681646435753e-06, + "loss": 0.6109, + "step": 19510 + }, + { + "epoch": 3.4736467236467234, + "grad_norm": 0.9258124828338623, + "learning_rate": 8.46118057310521e-06, + "loss": 0.7669, + "step": 19511 + }, + { + "epoch": 3.473824786324786, + "grad_norm": 1.161522388458252, + "learning_rate": 8.455546475365905e-06, + "loss": 0.9021, + "step": 19512 + }, + { + "epoch": 3.474002849002849, + "grad_norm": 0.9282271862030029, + "learning_rate": 8.449914171250006e-06, + "loss": 0.8032, + "step": 19513 + }, + { + "epoch": 3.4741809116809117, + "grad_norm": 0.9432225823402405, + "learning_rate": 8.444283660867858e-06, + "loss": 0.7169, + "step": 19514 + }, + { + "epoch": 3.4743589743589745, + "grad_norm": 0.9784168601036072, + "learning_rate": 8.438654944329782e-06, + "loss": 0.6596, + "step": 19515 + }, + { + "epoch": 3.474537037037037, + "grad_norm": 1.008577823638916, + "learning_rate": 8.433028021746036e-06, + "loss": 0.84, + "step": 19516 + }, + { + "epoch": 3.4747150997151, + "grad_norm": 0.8572508096694946, + "learning_rate": 8.42740289322691e-06, + "loss": 0.846, + "step": 19517 + }, + { + "epoch": 3.4748931623931623, + "grad_norm": 1.112871766090393, + "learning_rate": 8.421779558882603e-06, + "loss": 0.7197, + "step": 19518 + }, + { + "epoch": 3.475071225071225, + "grad_norm": 0.8831349015235901, + "learning_rate": 8.416158018823294e-06, + "loss": 0.7389, + "step": 19519 + }, + { + "epoch": 3.4752492877492878, + "grad_norm": 0.899161696434021, + "learning_rate": 8.410538273159107e-06, + "loss": 0.8457, + "step": 19520 + }, + { + "epoch": 3.4754273504273505, + "grad_norm": 0.8747708797454834, + "learning_rate": 8.404920322000154e-06, + "loss": 0.7574, + "step": 19521 + }, + { + "epoch": 3.4756054131054133, + "grad_norm": 0.892696738243103, + "learning_rate": 8.399304165456545e-06, + "loss": 0.8276, + "step": 19522 + }, + { + "epoch": 3.4757834757834756, + "grad_norm": 0.9339589476585388, + "learning_rate": 8.393689803638249e-06, + "loss": 0.7088, + "step": 19523 + }, + { + "epoch": 3.4759615384615383, + "grad_norm": 0.8795723915100098, + "learning_rate": 8.388077236655356e-06, + "loss": 0.6545, + "step": 19524 + }, + { + "epoch": 3.476139601139601, + "grad_norm": 0.9463884830474854, + "learning_rate": 8.382466464617733e-06, + "loss": 0.6912, + "step": 19525 + }, + { + "epoch": 3.476317663817664, + "grad_norm": 1.0345131158828735, + "learning_rate": 8.37685748763538e-06, + "loss": 0.8502, + "step": 19526 + }, + { + "epoch": 3.4764957264957266, + "grad_norm": 0.9798212647438049, + "learning_rate": 8.37125030581818e-06, + "loss": 0.6323, + "step": 19527 + }, + { + "epoch": 3.4766737891737893, + "grad_norm": 0.9508503079414368, + "learning_rate": 8.365644919275983e-06, + "loss": 0.732, + "step": 19528 + }, + { + "epoch": 3.476851851851852, + "grad_norm": 0.8767679333686829, + "learning_rate": 8.360041328118617e-06, + "loss": 0.454, + "step": 19529 + }, + { + "epoch": 3.4770299145299144, + "grad_norm": 1.0020666122436523, + "learning_rate": 8.354439532455882e-06, + "loss": 0.7342, + "step": 19530 + }, + { + "epoch": 3.477207977207977, + "grad_norm": 0.9661487340927124, + "learning_rate": 8.348839532397501e-06, + "loss": 0.9732, + "step": 19531 + }, + { + "epoch": 3.47738603988604, + "grad_norm": 1.0699942111968994, + "learning_rate": 8.343241328053264e-06, + "loss": 0.9216, + "step": 19532 + }, + { + "epoch": 3.4775641025641026, + "grad_norm": 0.8531695008277893, + "learning_rate": 8.337644919532772e-06, + "loss": 0.6829, + "step": 19533 + }, + { + "epoch": 3.4777421652421654, + "grad_norm": 1.080919861793518, + "learning_rate": 8.33205030694576e-06, + "loss": 0.9975, + "step": 19534 + }, + { + "epoch": 3.4779202279202277, + "grad_norm": 0.8957521915435791, + "learning_rate": 8.32645749040175e-06, + "loss": 0.5677, + "step": 19535 + }, + { + "epoch": 3.4780982905982905, + "grad_norm": 0.9968222975730896, + "learning_rate": 8.320866470010402e-06, + "loss": 0.7552, + "step": 19536 + }, + { + "epoch": 3.478276353276353, + "grad_norm": 0.8942126631736755, + "learning_rate": 8.315277245881215e-06, + "loss": 0.7861, + "step": 19537 + }, + { + "epoch": 3.478454415954416, + "grad_norm": 0.930486261844635, + "learning_rate": 8.3096898181237e-06, + "loss": 0.7911, + "step": 19538 + }, + { + "epoch": 3.4786324786324787, + "grad_norm": 1.013940691947937, + "learning_rate": 8.304104186847384e-06, + "loss": 0.7615, + "step": 19539 + }, + { + "epoch": 3.4788105413105415, + "grad_norm": 0.8439841866493225, + "learning_rate": 8.29852035216162e-06, + "loss": 0.7309, + "step": 19540 + }, + { + "epoch": 3.478988603988604, + "grad_norm": 0.9972648620605469, + "learning_rate": 8.29293831417588e-06, + "loss": 0.8021, + "step": 19541 + }, + { + "epoch": 3.4791666666666665, + "grad_norm": 0.9759960174560547, + "learning_rate": 8.287358072999507e-06, + "loss": 0.6253, + "step": 19542 + }, + { + "epoch": 3.4793447293447293, + "grad_norm": 1.2256296873092651, + "learning_rate": 8.281779628741837e-06, + "loss": 0.8939, + "step": 19543 + }, + { + "epoch": 3.479522792022792, + "grad_norm": 0.9147930145263672, + "learning_rate": 8.276202981512171e-06, + "loss": 0.7656, + "step": 19544 + }, + { + "epoch": 3.4797008547008548, + "grad_norm": 0.7807040214538574, + "learning_rate": 8.270628131419767e-06, + "loss": 0.5756, + "step": 19545 + }, + { + "epoch": 3.4798789173789175, + "grad_norm": 0.8961909413337708, + "learning_rate": 8.265055078573824e-06, + "loss": 0.7962, + "step": 19546 + }, + { + "epoch": 3.48005698005698, + "grad_norm": 1.0497139692306519, + "learning_rate": 8.259483823083614e-06, + "loss": 0.8766, + "step": 19547 + }, + { + "epoch": 3.4802350427350426, + "grad_norm": 0.9595639109611511, + "learning_rate": 8.253914365058202e-06, + "loss": 1.0157, + "step": 19548 + }, + { + "epoch": 3.4804131054131053, + "grad_norm": 0.8763233423233032, + "learning_rate": 8.248346704606779e-06, + "loss": 0.6731, + "step": 19549 + }, + { + "epoch": 3.480591168091168, + "grad_norm": 0.9633421897888184, + "learning_rate": 8.242780841838383e-06, + "loss": 0.9382, + "step": 19550 + }, + { + "epoch": 3.480769230769231, + "grad_norm": 0.982096791267395, + "learning_rate": 8.2372167768621e-06, + "loss": 0.7388, + "step": 19551 + }, + { + "epoch": 3.4809472934472936, + "grad_norm": 1.014060139656067, + "learning_rate": 8.231654509786935e-06, + "loss": 0.6534, + "step": 19552 + }, + { + "epoch": 3.4811253561253563, + "grad_norm": 0.8884333968162537, + "learning_rate": 8.226094040721865e-06, + "loss": 0.6659, + "step": 19553 + }, + { + "epoch": 3.4813034188034186, + "grad_norm": 0.9313388466835022, + "learning_rate": 8.22053536977584e-06, + "loss": 0.8194, + "step": 19554 + }, + { + "epoch": 3.4814814814814814, + "grad_norm": 0.9285356998443604, + "learning_rate": 8.214978497057768e-06, + "loss": 0.7316, + "step": 19555 + }, + { + "epoch": 3.481659544159544, + "grad_norm": 0.9004967212677002, + "learning_rate": 8.20942342267651e-06, + "loss": 0.7533, + "step": 19556 + }, + { + "epoch": 3.481837606837607, + "grad_norm": 0.8874818086624146, + "learning_rate": 8.203870146740932e-06, + "loss": 0.7632, + "step": 19557 + }, + { + "epoch": 3.4820156695156697, + "grad_norm": 0.9493485689163208, + "learning_rate": 8.19831866935984e-06, + "loss": 0.6767, + "step": 19558 + }, + { + "epoch": 3.482193732193732, + "grad_norm": 0.9512984156608582, + "learning_rate": 8.192768990641986e-06, + "loss": 0.8117, + "step": 19559 + }, + { + "epoch": 3.4823717948717947, + "grad_norm": 0.9378128051757812, + "learning_rate": 8.187221110696108e-06, + "loss": 0.7132, + "step": 19560 + }, + { + "epoch": 3.4825498575498575, + "grad_norm": 1.0644614696502686, + "learning_rate": 8.18167502963092e-06, + "loss": 0.7283, + "step": 19561 + }, + { + "epoch": 3.48272792022792, + "grad_norm": 0.9525963664054871, + "learning_rate": 8.176130747555055e-06, + "loss": 0.762, + "step": 19562 + }, + { + "epoch": 3.482905982905983, + "grad_norm": 0.8559122085571289, + "learning_rate": 8.170588264577161e-06, + "loss": 0.5776, + "step": 19563 + }, + { + "epoch": 3.4830840455840457, + "grad_norm": 0.870557427406311, + "learning_rate": 8.165047580805851e-06, + "loss": 0.7302, + "step": 19564 + }, + { + "epoch": 3.4832621082621085, + "grad_norm": 0.9522432684898376, + "learning_rate": 8.159508696349639e-06, + "loss": 0.6782, + "step": 19565 + }, + { + "epoch": 3.4834401709401708, + "grad_norm": 1.1580432653427124, + "learning_rate": 8.153971611317079e-06, + "loss": 0.7527, + "step": 19566 + }, + { + "epoch": 3.4836182336182335, + "grad_norm": 0.9478859901428223, + "learning_rate": 8.148436325816666e-06, + "loss": 0.7282, + "step": 19567 + }, + { + "epoch": 3.4837962962962963, + "grad_norm": 0.9263066649436951, + "learning_rate": 8.142902839956822e-06, + "loss": 0.808, + "step": 19568 + }, + { + "epoch": 3.483974358974359, + "grad_norm": 0.939940869808197, + "learning_rate": 8.137371153845996e-06, + "loss": 0.67, + "step": 19569 + }, + { + "epoch": 3.484152421652422, + "grad_norm": 1.0895870923995972, + "learning_rate": 8.131841267592544e-06, + "loss": 0.8526, + "step": 19570 + }, + { + "epoch": 3.484330484330484, + "grad_norm": 0.9202786087989807, + "learning_rate": 8.126313181304823e-06, + "loss": 0.668, + "step": 19571 + }, + { + "epoch": 3.484508547008547, + "grad_norm": 1.1485145092010498, + "learning_rate": 8.120786895091147e-06, + "loss": 0.8667, + "step": 19572 + }, + { + "epoch": 3.4846866096866096, + "grad_norm": 1.0418143272399902, + "learning_rate": 8.115262409059775e-06, + "loss": 0.7774, + "step": 19573 + }, + { + "epoch": 3.4848646723646723, + "grad_norm": 1.0974538326263428, + "learning_rate": 8.109739723318987e-06, + "loss": 0.9011, + "step": 19574 + }, + { + "epoch": 3.485042735042735, + "grad_norm": 0.9369264841079712, + "learning_rate": 8.10421883797694e-06, + "loss": 1.0659, + "step": 19575 + }, + { + "epoch": 3.485220797720798, + "grad_norm": 1.0749176740646362, + "learning_rate": 8.098699753141837e-06, + "loss": 0.8179, + "step": 19576 + }, + { + "epoch": 3.4853988603988606, + "grad_norm": 0.9837545156478882, + "learning_rate": 8.093182468921812e-06, + "loss": 0.8466, + "step": 19577 + }, + { + "epoch": 3.485576923076923, + "grad_norm": 0.9381014108657837, + "learning_rate": 8.087666985424935e-06, + "loss": 0.8468, + "step": 19578 + }, + { + "epoch": 3.4857549857549857, + "grad_norm": 0.8582633137702942, + "learning_rate": 8.082153302759322e-06, + "loss": 0.7085, + "step": 19579 + }, + { + "epoch": 3.4859330484330484, + "grad_norm": 0.9065280556678772, + "learning_rate": 8.07664142103295e-06, + "loss": 0.7177, + "step": 19580 + }, + { + "epoch": 3.486111111111111, + "grad_norm": 0.9608262181282043, + "learning_rate": 8.071131340353833e-06, + "loss": 0.6808, + "step": 19581 + }, + { + "epoch": 3.486289173789174, + "grad_norm": 0.8834822773933411, + "learning_rate": 8.065623060829951e-06, + "loss": 0.6307, + "step": 19582 + }, + { + "epoch": 3.486467236467236, + "grad_norm": 0.9237925410270691, + "learning_rate": 8.0601165825692e-06, + "loss": 1.0483, + "step": 19583 + }, + { + "epoch": 3.486645299145299, + "grad_norm": 0.9548817873001099, + "learning_rate": 8.054611905679477e-06, + "loss": 0.8959, + "step": 19584 + }, + { + "epoch": 3.4868233618233617, + "grad_norm": 1.0435808897018433, + "learning_rate": 8.049109030268631e-06, + "loss": 0.6882, + "step": 19585 + }, + { + "epoch": 3.4870014245014245, + "grad_norm": 0.9147984981536865, + "learning_rate": 8.043607956444477e-06, + "loss": 0.6671, + "step": 19586 + }, + { + "epoch": 3.4871794871794872, + "grad_norm": 0.9345784783363342, + "learning_rate": 8.038108684314815e-06, + "loss": 0.7734, + "step": 19587 + }, + { + "epoch": 3.48735754985755, + "grad_norm": 1.0092869997024536, + "learning_rate": 8.032611213987351e-06, + "loss": 0.8451, + "step": 19588 + }, + { + "epoch": 3.4875356125356127, + "grad_norm": 0.889281153678894, + "learning_rate": 8.027115545569863e-06, + "loss": 0.5752, + "step": 19589 + }, + { + "epoch": 3.487713675213675, + "grad_norm": 0.9964619874954224, + "learning_rate": 8.021621679169955e-06, + "loss": 0.7128, + "step": 19590 + }, + { + "epoch": 3.487891737891738, + "grad_norm": 0.9543761610984802, + "learning_rate": 8.016129614895329e-06, + "loss": 0.7243, + "step": 19591 + }, + { + "epoch": 3.4880698005698005, + "grad_norm": 0.9491791725158691, + "learning_rate": 8.010639352853544e-06, + "loss": 0.8746, + "step": 19592 + }, + { + "epoch": 3.4882478632478633, + "grad_norm": 0.9773432016372681, + "learning_rate": 8.005150893152203e-06, + "loss": 0.7438, + "step": 19593 + }, + { + "epoch": 3.488425925925926, + "grad_norm": 1.0735520124435425, + "learning_rate": 7.999664235898819e-06, + "loss": 1.0609, + "step": 19594 + }, + { + "epoch": 3.488603988603989, + "grad_norm": 0.981893002986908, + "learning_rate": 7.994179381200906e-06, + "loss": 0.9202, + "step": 19595 + }, + { + "epoch": 3.488782051282051, + "grad_norm": 1.0082266330718994, + "learning_rate": 7.988696329165924e-06, + "loss": 0.7032, + "step": 19596 + }, + { + "epoch": 3.488960113960114, + "grad_norm": 1.0541062355041504, + "learning_rate": 7.983215079901285e-06, + "loss": 0.7497, + "step": 19597 + }, + { + "epoch": 3.4891381766381766, + "grad_norm": 0.943924605846405, + "learning_rate": 7.977735633514405e-06, + "loss": 0.6938, + "step": 19598 + }, + { + "epoch": 3.4893162393162394, + "grad_norm": 0.8759856820106506, + "learning_rate": 7.972257990112642e-06, + "loss": 0.7018, + "step": 19599 + }, + { + "epoch": 3.489494301994302, + "grad_norm": 1.0315039157867432, + "learning_rate": 7.966782149803308e-06, + "loss": 0.8277, + "step": 19600 + }, + { + "epoch": 3.489672364672365, + "grad_norm": 1.0187058448791504, + "learning_rate": 7.9613081126937e-06, + "loss": 0.7854, + "step": 19601 + }, + { + "epoch": 3.489850427350427, + "grad_norm": 0.8936801552772522, + "learning_rate": 7.955835878891071e-06, + "loss": 0.7572, + "step": 19602 + }, + { + "epoch": 3.49002849002849, + "grad_norm": 1.004878044128418, + "learning_rate": 7.950365448502606e-06, + "loss": 0.8591, + "step": 19603 + }, + { + "epoch": 3.4902065527065527, + "grad_norm": 1.0317310094833374, + "learning_rate": 7.94489682163555e-06, + "loss": 0.7698, + "step": 19604 + }, + { + "epoch": 3.4903846153846154, + "grad_norm": 0.9507501721382141, + "learning_rate": 7.939429998396986e-06, + "loss": 0.7512, + "step": 19605 + }, + { + "epoch": 3.490562678062678, + "grad_norm": 0.9824426770210266, + "learning_rate": 7.933964978894082e-06, + "loss": 0.7041, + "step": 19606 + }, + { + "epoch": 3.490740740740741, + "grad_norm": 0.9749001264572144, + "learning_rate": 7.928501763233841e-06, + "loss": 0.7729, + "step": 19607 + }, + { + "epoch": 3.4909188034188032, + "grad_norm": 0.9753629565238953, + "learning_rate": 7.92304035152337e-06, + "loss": 0.7479, + "step": 19608 + }, + { + "epoch": 3.491096866096866, + "grad_norm": 0.9824631214141846, + "learning_rate": 7.917580743869646e-06, + "loss": 0.8447, + "step": 19609 + }, + { + "epoch": 3.4912749287749287, + "grad_norm": 1.1632194519042969, + "learning_rate": 7.912122940379651e-06, + "loss": 0.8926, + "step": 19610 + }, + { + "epoch": 3.4914529914529915, + "grad_norm": 0.8203635811805725, + "learning_rate": 7.9066669411603e-06, + "loss": 0.5599, + "step": 19611 + }, + { + "epoch": 3.4916310541310542, + "grad_norm": 0.922036349773407, + "learning_rate": 7.90121274631852e-06, + "loss": 0.7774, + "step": 19612 + }, + { + "epoch": 3.491809116809117, + "grad_norm": 1.112792730331421, + "learning_rate": 7.895760355961124e-06, + "loss": 0.9519, + "step": 19613 + }, + { + "epoch": 3.4919871794871793, + "grad_norm": 0.9443914890289307, + "learning_rate": 7.890309770195015e-06, + "loss": 0.8376, + "step": 19614 + }, + { + "epoch": 3.492165242165242, + "grad_norm": 0.9320971965789795, + "learning_rate": 7.884860989126907e-06, + "loss": 0.8073, + "step": 19615 + }, + { + "epoch": 3.492343304843305, + "grad_norm": 0.9588237404823303, + "learning_rate": 7.879414012863618e-06, + "loss": 0.7115, + "step": 19616 + }, + { + "epoch": 3.4925213675213675, + "grad_norm": 0.9260135293006897, + "learning_rate": 7.873968841511848e-06, + "loss": 0.6881, + "step": 19617 + }, + { + "epoch": 3.4926994301994303, + "grad_norm": 0.9520851969718933, + "learning_rate": 7.868525475178256e-06, + "loss": 0.8112, + "step": 19618 + }, + { + "epoch": 3.492877492877493, + "grad_norm": 1.027476191520691, + "learning_rate": 7.86308391396956e-06, + "loss": 0.8152, + "step": 19619 + }, + { + "epoch": 3.4930555555555554, + "grad_norm": 0.879357099533081, + "learning_rate": 7.857644157992305e-06, + "loss": 0.5809, + "step": 19620 + }, + { + "epoch": 3.493233618233618, + "grad_norm": 0.9070191979408264, + "learning_rate": 7.852206207353141e-06, + "loss": 0.6424, + "step": 19621 + }, + { + "epoch": 3.493411680911681, + "grad_norm": 1.0702928304672241, + "learning_rate": 7.846770062158537e-06, + "loss": 0.8733, + "step": 19622 + }, + { + "epoch": 3.4935897435897436, + "grad_norm": 0.8494130373001099, + "learning_rate": 7.841335722515053e-06, + "loss": 0.53, + "step": 19623 + }, + { + "epoch": 3.4937678062678064, + "grad_norm": 0.9504514336585999, + "learning_rate": 7.835903188529158e-06, + "loss": 0.7833, + "step": 19624 + }, + { + "epoch": 3.493945868945869, + "grad_norm": 0.9808318614959717, + "learning_rate": 7.830472460307293e-06, + "loss": 0.8398, + "step": 19625 + }, + { + "epoch": 3.494123931623932, + "grad_norm": 0.8339491486549377, + "learning_rate": 7.825043537955846e-06, + "loss": 0.7661, + "step": 19626 + }, + { + "epoch": 3.494301994301994, + "grad_norm": 1.1706161499023438, + "learning_rate": 7.819616421581199e-06, + "loss": 0.9843, + "step": 19627 + }, + { + "epoch": 3.494480056980057, + "grad_norm": 1.0100945234298706, + "learning_rate": 7.814191111289659e-06, + "loss": 0.8097, + "step": 19628 + }, + { + "epoch": 3.4946581196581197, + "grad_norm": 0.9706094861030579, + "learning_rate": 7.808767607187584e-06, + "loss": 0.9364, + "step": 19629 + }, + { + "epoch": 3.4948361823361824, + "grad_norm": 0.928600013256073, + "learning_rate": 7.803345909381154e-06, + "loss": 0.7667, + "step": 19630 + }, + { + "epoch": 3.495014245014245, + "grad_norm": 0.9604787826538086, + "learning_rate": 7.797926017976675e-06, + "loss": 0.8774, + "step": 19631 + }, + { + "epoch": 3.4951923076923075, + "grad_norm": 1.0755119323730469, + "learning_rate": 7.792507933080273e-06, + "loss": 0.9109, + "step": 19632 + }, + { + "epoch": 3.4953703703703702, + "grad_norm": 0.9468057155609131, + "learning_rate": 7.787091654798151e-06, + "loss": 0.812, + "step": 19633 + }, + { + "epoch": 3.495548433048433, + "grad_norm": 1.0440374612808228, + "learning_rate": 7.781677183236414e-06, + "loss": 0.6985, + "step": 19634 + }, + { + "epoch": 3.4957264957264957, + "grad_norm": 0.8866205215454102, + "learning_rate": 7.77626451850113e-06, + "loss": 0.523, + "step": 19635 + }, + { + "epoch": 3.4959045584045585, + "grad_norm": 0.8973997235298157, + "learning_rate": 7.770853660698384e-06, + "loss": 0.8511, + "step": 19636 + }, + { + "epoch": 3.4960826210826212, + "grad_norm": 0.9881580471992493, + "learning_rate": 7.765444609934147e-06, + "loss": 1.0181, + "step": 19637 + }, + { + "epoch": 3.496260683760684, + "grad_norm": 1.003432273864746, + "learning_rate": 7.760037366314433e-06, + "loss": 0.649, + "step": 19638 + }, + { + "epoch": 3.4964387464387463, + "grad_norm": 0.8911902904510498, + "learning_rate": 7.75463192994519e-06, + "loss": 0.925, + "step": 19639 + }, + { + "epoch": 3.496616809116809, + "grad_norm": 1.1478197574615479, + "learning_rate": 7.749228300932299e-06, + "loss": 0.8193, + "step": 19640 + }, + { + "epoch": 3.496794871794872, + "grad_norm": 0.9285799264907837, + "learning_rate": 7.743826479381644e-06, + "loss": 0.7878, + "step": 19641 + }, + { + "epoch": 3.4969729344729346, + "grad_norm": 1.0374022722244263, + "learning_rate": 7.738426465399063e-06, + "loss": 0.6693, + "step": 19642 + }, + { + "epoch": 3.4971509971509973, + "grad_norm": 0.9434183835983276, + "learning_rate": 7.733028259090369e-06, + "loss": 0.7659, + "step": 19643 + }, + { + "epoch": 3.4973290598290596, + "grad_norm": 1.0029457807540894, + "learning_rate": 7.727631860561314e-06, + "loss": 0.6427, + "step": 19644 + }, + { + "epoch": 3.4975071225071224, + "grad_norm": 0.883100688457489, + "learning_rate": 7.72223726991761e-06, + "loss": 0.8817, + "step": 19645 + }, + { + "epoch": 3.497685185185185, + "grad_norm": 0.9539889693260193, + "learning_rate": 7.716844487265018e-06, + "loss": 0.8675, + "step": 19646 + }, + { + "epoch": 3.497863247863248, + "grad_norm": 0.8217553496360779, + "learning_rate": 7.711453512709121e-06, + "loss": 0.5515, + "step": 19647 + }, + { + "epoch": 3.4980413105413106, + "grad_norm": 0.8870468735694885, + "learning_rate": 7.706064346355591e-06, + "loss": 0.6369, + "step": 19648 + }, + { + "epoch": 3.4982193732193734, + "grad_norm": 1.00053870677948, + "learning_rate": 7.700676988310008e-06, + "loss": 0.7248, + "step": 19649 + }, + { + "epoch": 3.498397435897436, + "grad_norm": 0.9883842468261719, + "learning_rate": 7.695291438677932e-06, + "loss": 0.7862, + "step": 19650 + }, + { + "epoch": 3.4985754985754984, + "grad_norm": 0.933318018913269, + "learning_rate": 7.68990769756487e-06, + "loss": 0.4922, + "step": 19651 + }, + { + "epoch": 3.498753561253561, + "grad_norm": 0.9415388107299805, + "learning_rate": 7.6845257650763e-06, + "loss": 0.82, + "step": 19652 + }, + { + "epoch": 3.498931623931624, + "grad_norm": 0.8702071309089661, + "learning_rate": 7.679145641317676e-06, + "loss": 0.4565, + "step": 19653 + }, + { + "epoch": 3.4991096866096867, + "grad_norm": 0.9140437245368958, + "learning_rate": 7.673767326394431e-06, + "loss": 0.6803, + "step": 19654 + }, + { + "epoch": 3.4992877492877494, + "grad_norm": 0.9249038696289062, + "learning_rate": 7.668390820411908e-06, + "loss": 0.7911, + "step": 19655 + }, + { + "epoch": 3.4994658119658117, + "grad_norm": 0.9517161846160889, + "learning_rate": 7.663016123475464e-06, + "loss": 0.9015, + "step": 19656 + }, + { + "epoch": 3.4994658119658117, + "eval_loss": 1.1762211322784424, + "eval_runtime": 24.2999, + "eval_samples_per_second": 42.84, + "eval_steps_per_second": 21.44, + "step": 19656 + }, + { + "epoch": 3.4996438746438745, + "grad_norm": 0.8536194562911987, + "learning_rate": 7.657643235690414e-06, + "loss": 0.7846, + "step": 19657 + }, + { + "epoch": 3.4998219373219372, + "grad_norm": 0.9985692501068115, + "learning_rate": 7.652272157162021e-06, + "loss": 0.7595, + "step": 19658 + }, + { + "epoch": 3.5, + "grad_norm": 1.068899154663086, + "learning_rate": 7.646902887995522e-06, + "loss": 0.6989, + "step": 19659 + }, + { + "epoch": 3.5001780626780628, + "grad_norm": 1.0672051906585693, + "learning_rate": 7.641535428296098e-06, + "loss": 0.774, + "step": 19660 + }, + { + "epoch": 3.5003561253561255, + "grad_norm": 0.978769063949585, + "learning_rate": 7.636169778168955e-06, + "loss": 0.8278, + "step": 19661 + }, + { + "epoch": 3.5005341880341883, + "grad_norm": 0.9284443855285645, + "learning_rate": 7.630805937719166e-06, + "loss": 0.7665, + "step": 19662 + }, + { + "epoch": 3.5007122507122506, + "grad_norm": 1.063376784324646, + "learning_rate": 7.6254439070518765e-06, + "loss": 0.6805, + "step": 19663 + }, + { + "epoch": 3.5008903133903133, + "grad_norm": 0.893804669380188, + "learning_rate": 7.620083686272117e-06, + "loss": 0.6082, + "step": 19664 + }, + { + "epoch": 3.501068376068376, + "grad_norm": 0.9822537899017334, + "learning_rate": 7.614725275484913e-06, + "loss": 0.9002, + "step": 19665 + }, + { + "epoch": 3.501246438746439, + "grad_norm": 0.9976572394371033, + "learning_rate": 7.609368674795259e-06, + "loss": 0.9215, + "step": 19666 + }, + { + "epoch": 3.5014245014245016, + "grad_norm": 0.925886332988739, + "learning_rate": 7.6040138843080925e-06, + "loss": 0.5707, + "step": 19667 + }, + { + "epoch": 3.501602564102564, + "grad_norm": 1.1334714889526367, + "learning_rate": 7.598660904128341e-06, + "loss": 0.7859, + "step": 19668 + }, + { + "epoch": 3.5017806267806266, + "grad_norm": 0.8253616690635681, + "learning_rate": 7.5933097343608874e-06, + "loss": 0.7421, + "step": 19669 + }, + { + "epoch": 3.5019586894586894, + "grad_norm": 0.8803296089172363, + "learning_rate": 7.587960375110548e-06, + "loss": 0.6529, + "step": 19670 + }, + { + "epoch": 3.502136752136752, + "grad_norm": 0.9909247756004333, + "learning_rate": 7.582612826482194e-06, + "loss": 0.8395, + "step": 19671 + }, + { + "epoch": 3.502314814814815, + "grad_norm": 0.895326554775238, + "learning_rate": 7.5772670885805194e-06, + "loss": 0.8597, + "step": 19672 + }, + { + "epoch": 3.5024928774928776, + "grad_norm": 0.9548590779304504, + "learning_rate": 7.5719231615103305e-06, + "loss": 0.6942, + "step": 19673 + }, + { + "epoch": 3.5026709401709404, + "grad_norm": 0.9338228106498718, + "learning_rate": 7.566581045376297e-06, + "loss": 0.7618, + "step": 19674 + }, + { + "epoch": 3.5028490028490027, + "grad_norm": 0.9816923141479492, + "learning_rate": 7.561240740283104e-06, + "loss": 0.7811, + "step": 19675 + }, + { + "epoch": 3.5030270655270654, + "grad_norm": 1.141014814376831, + "learning_rate": 7.5559022463353664e-06, + "loss": 0.823, + "step": 19676 + }, + { + "epoch": 3.503205128205128, + "grad_norm": 1.0397098064422607, + "learning_rate": 7.550565563637679e-06, + "loss": 0.73, + "step": 19677 + }, + { + "epoch": 3.503383190883191, + "grad_norm": 0.9919942021369934, + "learning_rate": 7.5452306922946355e-06, + "loss": 1.1807, + "step": 19678 + }, + { + "epoch": 3.5035612535612537, + "grad_norm": 1.051186442375183, + "learning_rate": 7.539897632410709e-06, + "loss": 0.7733, + "step": 19679 + }, + { + "epoch": 3.503739316239316, + "grad_norm": 1.067172646522522, + "learning_rate": 7.534566384090436e-06, + "loss": 0.9685, + "step": 19680 + }, + { + "epoch": 3.5039173789173788, + "grad_norm": 1.0977402925491333, + "learning_rate": 7.529236947438256e-06, + "loss": 0.8009, + "step": 19681 + }, + { + "epoch": 3.5040954415954415, + "grad_norm": 0.9920042753219604, + "learning_rate": 7.523909322558587e-06, + "loss": 0.6871, + "step": 19682 + }, + { + "epoch": 3.5042735042735043, + "grad_norm": 0.921493411064148, + "learning_rate": 7.51858350955581e-06, + "loss": 0.606, + "step": 19683 + }, + { + "epoch": 3.504451566951567, + "grad_norm": 1.0852283239364624, + "learning_rate": 7.513259508534276e-06, + "loss": 0.5686, + "step": 19684 + }, + { + "epoch": 3.5046296296296298, + "grad_norm": 0.8194310069084167, + "learning_rate": 7.507937319598291e-06, + "loss": 0.6166, + "step": 19685 + }, + { + "epoch": 3.5048076923076925, + "grad_norm": 0.921829104423523, + "learning_rate": 7.502616942852159e-06, + "loss": 0.6076, + "step": 19686 + }, + { + "epoch": 3.504985754985755, + "grad_norm": 0.9465250968933105, + "learning_rate": 7.497298378400075e-06, + "loss": 0.6255, + "step": 19687 + }, + { + "epoch": 3.5051638176638176, + "grad_norm": 0.9845356345176697, + "learning_rate": 7.4919816263462786e-06, + "loss": 0.7305, + "step": 19688 + }, + { + "epoch": 3.5053418803418803, + "grad_norm": 1.0652577877044678, + "learning_rate": 7.486666686794941e-06, + "loss": 0.7852, + "step": 19689 + }, + { + "epoch": 3.505519943019943, + "grad_norm": 0.9593751430511475, + "learning_rate": 7.4813535598501905e-06, + "loss": 0.8299, + "step": 19690 + }, + { + "epoch": 3.505698005698006, + "grad_norm": 0.8477317690849304, + "learning_rate": 7.4760422456161215e-06, + "loss": 0.5603, + "step": 19691 + }, + { + "epoch": 3.505876068376068, + "grad_norm": 0.9810600280761719, + "learning_rate": 7.470732744196807e-06, + "loss": 0.6908, + "step": 19692 + }, + { + "epoch": 3.506054131054131, + "grad_norm": 0.9363740682601929, + "learning_rate": 7.4654250556962734e-06, + "loss": 0.6607, + "step": 19693 + }, + { + "epoch": 3.5062321937321936, + "grad_norm": 0.96392422914505, + "learning_rate": 7.460119180218505e-06, + "loss": 0.726, + "step": 19694 + }, + { + "epoch": 3.5064102564102564, + "grad_norm": 0.9342692494392395, + "learning_rate": 7.454815117867453e-06, + "loss": 0.6143, + "step": 19695 + }, + { + "epoch": 3.506588319088319, + "grad_norm": 0.9736638069152832, + "learning_rate": 7.449512868747066e-06, + "loss": 0.7332, + "step": 19696 + }, + { + "epoch": 3.506766381766382, + "grad_norm": 0.9093930125236511, + "learning_rate": 7.444212432961228e-06, + "loss": 0.7471, + "step": 19697 + }, + { + "epoch": 3.5069444444444446, + "grad_norm": 0.9061489701271057, + "learning_rate": 7.438913810613768e-06, + "loss": 0.7445, + "step": 19698 + }, + { + "epoch": 3.5071225071225074, + "grad_norm": 0.9087863564491272, + "learning_rate": 7.433617001808513e-06, + "loss": 0.7966, + "step": 19699 + }, + { + "epoch": 3.5073005698005697, + "grad_norm": 0.9920783042907715, + "learning_rate": 7.428322006649236e-06, + "loss": 0.9201, + "step": 19700 + }, + { + "epoch": 3.5074786324786325, + "grad_norm": 1.0740957260131836, + "learning_rate": 7.423028825239719e-06, + "loss": 0.7541, + "step": 19701 + }, + { + "epoch": 3.507656695156695, + "grad_norm": 0.9236546754837036, + "learning_rate": 7.417737457683594e-06, + "loss": 0.8124, + "step": 19702 + }, + { + "epoch": 3.507834757834758, + "grad_norm": 1.0994749069213867, + "learning_rate": 7.412447904084629e-06, + "loss": 0.7571, + "step": 19703 + }, + { + "epoch": 3.5080128205128203, + "grad_norm": 0.8928847908973694, + "learning_rate": 7.4071601645463785e-06, + "loss": 0.7111, + "step": 19704 + }, + { + "epoch": 3.508190883190883, + "grad_norm": 0.9118340611457825, + "learning_rate": 7.40187423917249e-06, + "loss": 0.7538, + "step": 19705 + }, + { + "epoch": 3.5083689458689458, + "grad_norm": 1.0264140367507935, + "learning_rate": 7.396590128066516e-06, + "loss": 0.7927, + "step": 19706 + }, + { + "epoch": 3.5085470085470085, + "grad_norm": 0.8983800411224365, + "learning_rate": 7.391307831332006e-06, + "loss": 0.6858, + "step": 19707 + }, + { + "epoch": 3.5087250712250713, + "grad_norm": 1.0997859239578247, + "learning_rate": 7.386027349072433e-06, + "loss": 0.8014, + "step": 19708 + }, + { + "epoch": 3.508903133903134, + "grad_norm": 1.0893813371658325, + "learning_rate": 7.380748681391258e-06, + "loss": 0.8363, + "step": 19709 + }, + { + "epoch": 3.5090811965811968, + "grad_norm": 0.9206538796424866, + "learning_rate": 7.375471828391911e-06, + "loss": 0.6414, + "step": 19710 + }, + { + "epoch": 3.5092592592592595, + "grad_norm": 0.9687008857727051, + "learning_rate": 7.370196790177808e-06, + "loss": 0.8868, + "step": 19711 + }, + { + "epoch": 3.509437321937322, + "grad_norm": 0.8396771550178528, + "learning_rate": 7.3649235668522445e-06, + "loss": 0.5922, + "step": 19712 + }, + { + "epoch": 3.5096153846153846, + "grad_norm": 1.01363205909729, + "learning_rate": 7.359652158518604e-06, + "loss": 0.7419, + "step": 19713 + }, + { + "epoch": 3.5097934472934473, + "grad_norm": 1.086411952972412, + "learning_rate": 7.354382565280094e-06, + "loss": 0.8505, + "step": 19714 + }, + { + "epoch": 3.50997150997151, + "grad_norm": 1.1874315738677979, + "learning_rate": 7.34911478724003e-06, + "loss": 0.8852, + "step": 19715 + }, + { + "epoch": 3.5101495726495724, + "grad_norm": 1.0598959922790527, + "learning_rate": 7.343848824501598e-06, + "loss": 0.7538, + "step": 19716 + }, + { + "epoch": 3.510327635327635, + "grad_norm": 0.7151375412940979, + "learning_rate": 7.338584677167948e-06, + "loss": 0.4218, + "step": 19717 + }, + { + "epoch": 3.510505698005698, + "grad_norm": 0.9595068097114563, + "learning_rate": 7.333322345342286e-06, + "loss": 0.8518, + "step": 19718 + }, + { + "epoch": 3.5106837606837606, + "grad_norm": 0.8868532776832581, + "learning_rate": 7.328061829127631e-06, + "loss": 0.9117, + "step": 19719 + }, + { + "epoch": 3.5108618233618234, + "grad_norm": 1.0120736360549927, + "learning_rate": 7.32280312862712e-06, + "loss": 0.8192, + "step": 19720 + }, + { + "epoch": 3.511039886039886, + "grad_norm": 1.0815523862838745, + "learning_rate": 7.317546243943751e-06, + "loss": 0.6933, + "step": 19721 + }, + { + "epoch": 3.511217948717949, + "grad_norm": 1.0424879789352417, + "learning_rate": 7.312291175180541e-06, + "loss": 0.7333, + "step": 19722 + }, + { + "epoch": 3.5113960113960117, + "grad_norm": 0.9301778078079224, + "learning_rate": 7.307037922440441e-06, + "loss": 0.7172, + "step": 19723 + }, + { + "epoch": 3.511574074074074, + "grad_norm": 1.0618454217910767, + "learning_rate": 7.301786485826389e-06, + "loss": 0.6895, + "step": 19724 + }, + { + "epoch": 3.5117521367521367, + "grad_norm": 1.0267716646194458, + "learning_rate": 7.2965368654412395e-06, + "loss": 0.9075, + "step": 19725 + }, + { + "epoch": 3.5119301994301995, + "grad_norm": 1.1378053426742554, + "learning_rate": 7.291289061387907e-06, + "loss": 0.9324, + "step": 19726 + }, + { + "epoch": 3.512108262108262, + "grad_norm": 0.9694442749023438, + "learning_rate": 7.286043073769155e-06, + "loss": 0.7921, + "step": 19727 + }, + { + "epoch": 3.5122863247863245, + "grad_norm": 0.8989338278770447, + "learning_rate": 7.280798902687813e-06, + "loss": 0.7454, + "step": 19728 + }, + { + "epoch": 3.5124643874643873, + "grad_norm": 0.8680744171142578, + "learning_rate": 7.275556548246587e-06, + "loss": 0.7519, + "step": 19729 + }, + { + "epoch": 3.51264245014245, + "grad_norm": 0.9522790312767029, + "learning_rate": 7.2703160105482285e-06, + "loss": 0.7243, + "step": 19730 + }, + { + "epoch": 3.5128205128205128, + "grad_norm": 0.9496331214904785, + "learning_rate": 7.265077289695399e-06, + "loss": 0.7005, + "step": 19731 + }, + { + "epoch": 3.5129985754985755, + "grad_norm": 0.8820855021476746, + "learning_rate": 7.259840385790728e-06, + "loss": 0.7302, + "step": 19732 + }, + { + "epoch": 3.5131766381766383, + "grad_norm": 0.9601408839225769, + "learning_rate": 7.254605298936845e-06, + "loss": 0.7137, + "step": 19733 + }, + { + "epoch": 3.513354700854701, + "grad_norm": 0.9732968211174011, + "learning_rate": 7.249372029236312e-06, + "loss": 0.7444, + "step": 19734 + }, + { + "epoch": 3.513532763532764, + "grad_norm": 0.9062048196792603, + "learning_rate": 7.244140576791636e-06, + "loss": 0.8984, + "step": 19735 + }, + { + "epoch": 3.513710826210826, + "grad_norm": 0.9823424220085144, + "learning_rate": 7.2389109417053566e-06, + "loss": 0.8515, + "step": 19736 + }, + { + "epoch": 3.513888888888889, + "grad_norm": 0.8319346904754639, + "learning_rate": 7.233683124079937e-06, + "loss": 0.7185, + "step": 19737 + }, + { + "epoch": 3.5140669515669516, + "grad_norm": 0.9495996236801147, + "learning_rate": 7.2284571240177735e-06, + "loss": 0.7912, + "step": 19738 + }, + { + "epoch": 3.5142450142450143, + "grad_norm": 0.8981806039810181, + "learning_rate": 7.223232941621294e-06, + "loss": 0.5754, + "step": 19739 + }, + { + "epoch": 3.5144230769230766, + "grad_norm": 1.0965253114700317, + "learning_rate": 7.218010576992829e-06, + "loss": 0.894, + "step": 19740 + }, + { + "epoch": 3.5146011396011394, + "grad_norm": 0.9036837816238403, + "learning_rate": 7.212790030234706e-06, + "loss": 0.8213, + "step": 19741 + }, + { + "epoch": 3.514779202279202, + "grad_norm": 0.8697100281715393, + "learning_rate": 7.2075713014492004e-06, + "loss": 0.7141, + "step": 19742 + }, + { + "epoch": 3.514957264957265, + "grad_norm": 0.9425762295722961, + "learning_rate": 7.202354390738608e-06, + "loss": 0.6477, + "step": 19743 + }, + { + "epoch": 3.5151353276353277, + "grad_norm": 0.9594342708587646, + "learning_rate": 7.19713929820508e-06, + "loss": 0.8327, + "step": 19744 + }, + { + "epoch": 3.5153133903133904, + "grad_norm": 1.132218837738037, + "learning_rate": 7.191926023950835e-06, + "loss": 0.8389, + "step": 19745 + }, + { + "epoch": 3.515491452991453, + "grad_norm": 0.923564612865448, + "learning_rate": 7.186714568078012e-06, + "loss": 0.7108, + "step": 19746 + }, + { + "epoch": 3.515669515669516, + "grad_norm": 0.9644235968589783, + "learning_rate": 7.1815049306887204e-06, + "loss": 0.7992, + "step": 19747 + }, + { + "epoch": 3.515847578347578, + "grad_norm": 0.9429721832275391, + "learning_rate": 7.176297111885022e-06, + "loss": 0.7287, + "step": 19748 + }, + { + "epoch": 3.516025641025641, + "grad_norm": 1.015499472618103, + "learning_rate": 7.171091111768957e-06, + "loss": 0.8746, + "step": 19749 + }, + { + "epoch": 3.5162037037037037, + "grad_norm": 1.0005598068237305, + "learning_rate": 7.165886930442522e-06, + "loss": 0.8414, + "step": 19750 + }, + { + "epoch": 3.5163817663817665, + "grad_norm": 1.1245676279067993, + "learning_rate": 7.160684568007692e-06, + "loss": 1.0094, + "step": 19751 + }, + { + "epoch": 3.5165598290598292, + "grad_norm": 0.9880303144454956, + "learning_rate": 7.155484024566372e-06, + "loss": 0.7842, + "step": 19752 + }, + { + "epoch": 3.5167378917378915, + "grad_norm": 0.9797958135604858, + "learning_rate": 7.150285300220505e-06, + "loss": 0.928, + "step": 19753 + }, + { + "epoch": 3.5169159544159543, + "grad_norm": 0.9521427154541016, + "learning_rate": 7.145088395071886e-06, + "loss": 0.7293, + "step": 19754 + }, + { + "epoch": 3.517094017094017, + "grad_norm": 0.9081432223320007, + "learning_rate": 7.1398933092223896e-06, + "loss": 0.8062, + "step": 19755 + }, + { + "epoch": 3.51727207977208, + "grad_norm": 1.08573317527771, + "learning_rate": 7.13470004277379e-06, + "loss": 0.7256, + "step": 19756 + }, + { + "epoch": 3.5174501424501425, + "grad_norm": 1.0338917970657349, + "learning_rate": 7.129508595827805e-06, + "loss": 0.9324, + "step": 19757 + }, + { + "epoch": 3.5176282051282053, + "grad_norm": 0.9286987781524658, + "learning_rate": 7.124318968486221e-06, + "loss": 0.7037, + "step": 19758 + }, + { + "epoch": 3.517806267806268, + "grad_norm": 0.9461976885795593, + "learning_rate": 7.119131160850634e-06, + "loss": 0.7509, + "step": 19759 + }, + { + "epoch": 3.5179843304843303, + "grad_norm": 1.0241987705230713, + "learning_rate": 7.11394517302274e-06, + "loss": 0.8071, + "step": 19760 + }, + { + "epoch": 3.518162393162393, + "grad_norm": 0.8855302333831787, + "learning_rate": 7.108761005104147e-06, + "loss": 0.7045, + "step": 19761 + }, + { + "epoch": 3.518340455840456, + "grad_norm": 0.9970550537109375, + "learning_rate": 7.103578657196419e-06, + "loss": 0.7092, + "step": 19762 + }, + { + "epoch": 3.5185185185185186, + "grad_norm": 0.9413601756095886, + "learning_rate": 7.0983981294010845e-06, + "loss": 0.6876, + "step": 19763 + }, + { + "epoch": 3.5186965811965814, + "grad_norm": 0.9557897448539734, + "learning_rate": 7.093219421819653e-06, + "loss": 0.7068, + "step": 19764 + }, + { + "epoch": 3.5188746438746437, + "grad_norm": 0.9844231605529785, + "learning_rate": 7.088042534553585e-06, + "loss": 0.7696, + "step": 19765 + }, + { + "epoch": 3.5190527065527064, + "grad_norm": 0.8869762420654297, + "learning_rate": 7.082867467704324e-06, + "loss": 0.855, + "step": 19766 + }, + { + "epoch": 3.519230769230769, + "grad_norm": 1.0310113430023193, + "learning_rate": 7.0776942213732325e-06, + "loss": 0.7864, + "step": 19767 + }, + { + "epoch": 3.519408831908832, + "grad_norm": 0.9870700836181641, + "learning_rate": 7.072522795661729e-06, + "loss": 0.8516, + "step": 19768 + }, + { + "epoch": 3.5195868945868947, + "grad_norm": 0.9155564308166504, + "learning_rate": 7.067353190671078e-06, + "loss": 0.6531, + "step": 19769 + }, + { + "epoch": 3.5197649572649574, + "grad_norm": 0.8860092759132385, + "learning_rate": 7.062185406502597e-06, + "loss": 0.7624, + "step": 19770 + }, + { + "epoch": 3.51994301994302, + "grad_norm": 0.9172900915145874, + "learning_rate": 7.057019443257528e-06, + "loss": 0.7268, + "step": 19771 + }, + { + "epoch": 3.5201210826210825, + "grad_norm": 0.932388424873352, + "learning_rate": 7.051855301037102e-06, + "loss": 0.5527, + "step": 19772 + }, + { + "epoch": 3.5202991452991452, + "grad_norm": 0.9461301565170288, + "learning_rate": 7.0466929799424816e-06, + "loss": 0.8035, + "step": 19773 + }, + { + "epoch": 3.520477207977208, + "grad_norm": 0.925786554813385, + "learning_rate": 7.04153248007482e-06, + "loss": 0.7452, + "step": 19774 + }, + { + "epoch": 3.5206552706552707, + "grad_norm": 1.4296557903289795, + "learning_rate": 7.036373801535223e-06, + "loss": 0.8918, + "step": 19775 + }, + { + "epoch": 3.5208333333333335, + "grad_norm": 0.9940581321716309, + "learning_rate": 7.031216944424746e-06, + "loss": 0.8082, + "step": 19776 + }, + { + "epoch": 3.521011396011396, + "grad_norm": 0.8591741919517517, + "learning_rate": 7.026061908844472e-06, + "loss": 0.7566, + "step": 19777 + }, + { + "epoch": 3.5211894586894585, + "grad_norm": 0.8477209210395813, + "learning_rate": 7.020908694895368e-06, + "loss": 0.6008, + "step": 19778 + }, + { + "epoch": 3.5213675213675213, + "grad_norm": 1.1067235469818115, + "learning_rate": 7.015757302678416e-06, + "loss": 1.0526, + "step": 19779 + }, + { + "epoch": 3.521545584045584, + "grad_norm": 0.9242870807647705, + "learning_rate": 7.01060773229455e-06, + "loss": 0.827, + "step": 19780 + }, + { + "epoch": 3.521723646723647, + "grad_norm": 0.9401116967201233, + "learning_rate": 7.005459983844642e-06, + "loss": 0.8458, + "step": 19781 + }, + { + "epoch": 3.5219017094017095, + "grad_norm": 0.8446176648139954, + "learning_rate": 7.000314057429558e-06, + "loss": 0.5964, + "step": 19782 + }, + { + "epoch": 3.5220797720797723, + "grad_norm": 0.931850254535675, + "learning_rate": 6.9951699531501714e-06, + "loss": 0.6574, + "step": 19783 + }, + { + "epoch": 3.5222578347578346, + "grad_norm": 0.9261093735694885, + "learning_rate": 6.99002767110718e-06, + "loss": 0.8009, + "step": 19784 + }, + { + "epoch": 3.5224358974358974, + "grad_norm": 1.123136281967163, + "learning_rate": 6.984887211401425e-06, + "loss": 0.7744, + "step": 19785 + }, + { + "epoch": 3.52261396011396, + "grad_norm": 0.9234578609466553, + "learning_rate": 6.979748574133549e-06, + "loss": 0.8332, + "step": 19786 + }, + { + "epoch": 3.522792022792023, + "grad_norm": 1.0575841665267944, + "learning_rate": 6.974611759404281e-06, + "loss": 0.8427, + "step": 19787 + }, + { + "epoch": 3.5229700854700856, + "grad_norm": 0.9604585766792297, + "learning_rate": 6.9694767673142645e-06, + "loss": 0.8508, + "step": 19788 + }, + { + "epoch": 3.523148148148148, + "grad_norm": 0.9769417643547058, + "learning_rate": 6.9643435979640845e-06, + "loss": 0.6211, + "step": 19789 + }, + { + "epoch": 3.5233262108262107, + "grad_norm": 0.9469193816184998, + "learning_rate": 6.959212251454328e-06, + "loss": 0.8078, + "step": 19790 + }, + { + "epoch": 3.5235042735042734, + "grad_norm": 0.9950293302536011, + "learning_rate": 6.9540827278855354e-06, + "loss": 0.6546, + "step": 19791 + }, + { + "epoch": 3.523682336182336, + "grad_norm": 0.9339304566383362, + "learning_rate": 6.9489550273581834e-06, + "loss": 0.7838, + "step": 19792 + }, + { + "epoch": 3.523860398860399, + "grad_norm": 1.0213496685028076, + "learning_rate": 6.943829149972802e-06, + "loss": 0.8105, + "step": 19793 + }, + { + "epoch": 3.5240384615384617, + "grad_norm": 0.8725095391273499, + "learning_rate": 6.938705095829734e-06, + "loss": 0.6082, + "step": 19794 + }, + { + "epoch": 3.5242165242165244, + "grad_norm": 0.8054739236831665, + "learning_rate": 6.933582865029453e-06, + "loss": 0.6168, + "step": 19795 + }, + { + "epoch": 3.5243945868945867, + "grad_norm": 0.9824689626693726, + "learning_rate": 6.92846245767228e-06, + "loss": 0.8601, + "step": 19796 + }, + { + "epoch": 3.5245726495726495, + "grad_norm": 1.024682641029358, + "learning_rate": 6.9233438738585345e-06, + "loss": 0.7557, + "step": 19797 + }, + { + "epoch": 3.5247507122507122, + "grad_norm": 0.8784366250038147, + "learning_rate": 6.918227113688547e-06, + "loss": 0.7456, + "step": 19798 + }, + { + "epoch": 3.524928774928775, + "grad_norm": 0.8838477730751038, + "learning_rate": 6.913112177262493e-06, + "loss": 0.6641, + "step": 19799 + }, + { + "epoch": 3.5251068376068377, + "grad_norm": 0.8852226138114929, + "learning_rate": 6.907999064680681e-06, + "loss": 0.7466, + "step": 19800 + }, + { + "epoch": 3.5252849002849, + "grad_norm": 1.0038496255874634, + "learning_rate": 6.90288777604321e-06, + "loss": 0.7156, + "step": 19801 + }, + { + "epoch": 3.525462962962963, + "grad_norm": 1.1669005155563354, + "learning_rate": 6.897778311450276e-06, + "loss": 0.8363, + "step": 19802 + }, + { + "epoch": 3.5256410256410255, + "grad_norm": 0.9355636835098267, + "learning_rate": 6.892670671001977e-06, + "loss": 0.8016, + "step": 19803 + }, + { + "epoch": 3.5258190883190883, + "grad_norm": 0.9169090986251831, + "learning_rate": 6.887564854798378e-06, + "loss": 0.712, + "step": 19804 + }, + { + "epoch": 3.525997150997151, + "grad_norm": 0.9154117703437805, + "learning_rate": 6.882460862939522e-06, + "loss": 0.6729, + "step": 19805 + }, + { + "epoch": 3.526175213675214, + "grad_norm": 0.9298785328865051, + "learning_rate": 6.877358695525416e-06, + "loss": 0.6924, + "step": 19806 + }, + { + "epoch": 3.5263532763532766, + "grad_norm": 0.9413958191871643, + "learning_rate": 6.872258352655991e-06, + "loss": 0.8641, + "step": 19807 + }, + { + "epoch": 3.5265313390313393, + "grad_norm": 0.9077426791191101, + "learning_rate": 6.867159834431247e-06, + "loss": 0.725, + "step": 19808 + }, + { + "epoch": 3.5267094017094016, + "grad_norm": 0.9770194292068481, + "learning_rate": 6.8620631409510135e-06, + "loss": 0.852, + "step": 19809 + }, + { + "epoch": 3.5268874643874644, + "grad_norm": 1.092766523361206, + "learning_rate": 6.8569682723152e-06, + "loss": 0.6776, + "step": 19810 + }, + { + "epoch": 3.527065527065527, + "grad_norm": 0.8729142546653748, + "learning_rate": 6.851875228623572e-06, + "loss": 0.7475, + "step": 19811 + }, + { + "epoch": 3.52724358974359, + "grad_norm": 1.0176565647125244, + "learning_rate": 6.846784009975971e-06, + "loss": 0.9025, + "step": 19812 + }, + { + "epoch": 3.527421652421652, + "grad_norm": 1.1086188554763794, + "learning_rate": 6.841694616472128e-06, + "loss": 0.9416, + "step": 19813 + }, + { + "epoch": 3.527599715099715, + "grad_norm": 0.9331051707267761, + "learning_rate": 6.8366070482117625e-06, + "loss": 0.5952, + "step": 19814 + }, + { + "epoch": 3.5277777777777777, + "grad_norm": 0.941539466381073, + "learning_rate": 6.8315213052945526e-06, + "loss": 0.6675, + "step": 19815 + }, + { + "epoch": 3.5279558404558404, + "grad_norm": 0.9927091598510742, + "learning_rate": 6.826437387820117e-06, + "loss": 0.6435, + "step": 19816 + }, + { + "epoch": 3.528133903133903, + "grad_norm": 0.9932602047920227, + "learning_rate": 6.8213552958881096e-06, + "loss": 0.9288, + "step": 19817 + }, + { + "epoch": 3.528311965811966, + "grad_norm": 1.017561912536621, + "learning_rate": 6.816275029598085e-06, + "loss": 0.8233, + "step": 19818 + }, + { + "epoch": 3.5284900284900287, + "grad_norm": 0.972872257232666, + "learning_rate": 6.811196589049573e-06, + "loss": 0.6626, + "step": 19819 + }, + { + "epoch": 3.5286680911680914, + "grad_norm": 1.009381651878357, + "learning_rate": 6.806119974342084e-06, + "loss": 0.718, + "step": 19820 + }, + { + "epoch": 3.5288461538461537, + "grad_norm": 1.0876743793487549, + "learning_rate": 6.801045185575083e-06, + "loss": 0.7202, + "step": 19821 + }, + { + "epoch": 3.5290242165242165, + "grad_norm": 0.8937970399856567, + "learning_rate": 6.795972222848002e-06, + "loss": 0.5068, + "step": 19822 + }, + { + "epoch": 3.5292022792022792, + "grad_norm": 1.0466728210449219, + "learning_rate": 6.790901086260226e-06, + "loss": 0.6791, + "step": 19823 + }, + { + "epoch": 3.529380341880342, + "grad_norm": 0.9000089168548584, + "learning_rate": 6.7858317759111e-06, + "loss": 0.7009, + "step": 19824 + }, + { + "epoch": 3.5295584045584043, + "grad_norm": 1.0114625692367554, + "learning_rate": 6.780764291899988e-06, + "loss": 0.6994, + "step": 19825 + }, + { + "epoch": 3.529736467236467, + "grad_norm": 0.9358991980552673, + "learning_rate": 6.775698634326133e-06, + "loss": 0.6651, + "step": 19826 + }, + { + "epoch": 3.52991452991453, + "grad_norm": 0.8242545127868652, + "learning_rate": 6.770634803288822e-06, + "loss": 0.6033, + "step": 19827 + }, + { + "epoch": 3.5300925925925926, + "grad_norm": 0.8415285348892212, + "learning_rate": 6.765572798887254e-06, + "loss": 0.5524, + "step": 19828 + }, + { + "epoch": 3.5302706552706553, + "grad_norm": 0.9627799987792969, + "learning_rate": 6.760512621220616e-06, + "loss": 0.8388, + "step": 19829 + }, + { + "epoch": 3.530448717948718, + "grad_norm": 1.0505856275558472, + "learning_rate": 6.755454270388029e-06, + "loss": 0.7476, + "step": 19830 + }, + { + "epoch": 3.530626780626781, + "grad_norm": 0.9269016981124878, + "learning_rate": 6.7503977464886235e-06, + "loss": 0.6445, + "step": 19831 + }, + { + "epoch": 3.5308048433048436, + "grad_norm": 0.8944364786148071, + "learning_rate": 6.745343049621456e-06, + "loss": 0.7826, + "step": 19832 + }, + { + "epoch": 3.530982905982906, + "grad_norm": 0.9199281334877014, + "learning_rate": 6.7402901798856e-06, + "loss": 0.9266, + "step": 19833 + }, + { + "epoch": 3.5311609686609686, + "grad_norm": 0.8758364915847778, + "learning_rate": 6.7352391373799895e-06, + "loss": 0.8124, + "step": 19834 + }, + { + "epoch": 3.5313390313390314, + "grad_norm": 1.1305196285247803, + "learning_rate": 6.7301899222036555e-06, + "loss": 1.0041, + "step": 19835 + }, + { + "epoch": 3.531517094017094, + "grad_norm": 0.9125435948371887, + "learning_rate": 6.725142534455487e-06, + "loss": 0.7151, + "step": 19836 + }, + { + "epoch": 3.5316951566951564, + "grad_norm": 0.9822704195976257, + "learning_rate": 6.7200969742343915e-06, + "loss": 0.7827, + "step": 19837 + }, + { + "epoch": 3.531873219373219, + "grad_norm": 0.9346258044242859, + "learning_rate": 6.715053241639224e-06, + "loss": 0.7513, + "step": 19838 + }, + { + "epoch": 3.532051282051282, + "grad_norm": 0.8414269685745239, + "learning_rate": 6.710011336768796e-06, + "loss": 0.6312, + "step": 19839 + }, + { + "epoch": 3.5322293447293447, + "grad_norm": 0.9639045000076294, + "learning_rate": 6.704971259721926e-06, + "loss": 0.719, + "step": 19840 + }, + { + "epoch": 3.5324074074074074, + "grad_norm": 0.9534916877746582, + "learning_rate": 6.699933010597314e-06, + "loss": 0.7557, + "step": 19841 + }, + { + "epoch": 3.53258547008547, + "grad_norm": 0.8696106672286987, + "learning_rate": 6.6948965894937155e-06, + "loss": 0.5598, + "step": 19842 + }, + { + "epoch": 3.532763532763533, + "grad_norm": 0.9199994206428528, + "learning_rate": 6.689861996509794e-06, + "loss": 0.7441, + "step": 19843 + }, + { + "epoch": 3.5329415954415957, + "grad_norm": 0.9052377343177795, + "learning_rate": 6.684829231744183e-06, + "loss": 0.7531, + "step": 19844 + }, + { + "epoch": 3.533119658119658, + "grad_norm": 0.9178473949432373, + "learning_rate": 6.679798295295514e-06, + "loss": 0.7278, + "step": 19845 + }, + { + "epoch": 3.5332977207977208, + "grad_norm": 0.8894423842430115, + "learning_rate": 6.6747691872623305e-06, + "loss": 0.7972, + "step": 19846 + }, + { + "epoch": 3.5334757834757835, + "grad_norm": 1.0122573375701904, + "learning_rate": 6.669741907743177e-06, + "loss": 0.9849, + "step": 19847 + }, + { + "epoch": 3.5336538461538463, + "grad_norm": 0.9068432450294495, + "learning_rate": 6.664716456836561e-06, + "loss": 0.9914, + "step": 19848 + }, + { + "epoch": 3.5338319088319086, + "grad_norm": 0.840515673160553, + "learning_rate": 6.6596928346409185e-06, + "loss": 0.7176, + "step": 19849 + }, + { + "epoch": 3.5340099715099713, + "grad_norm": 0.9179866313934326, + "learning_rate": 6.6546710412547344e-06, + "loss": 0.6759, + "step": 19850 + }, + { + "epoch": 3.534188034188034, + "grad_norm": 0.7912464737892151, + "learning_rate": 6.64965107677632e-06, + "loss": 0.6583, + "step": 19851 + }, + { + "epoch": 3.534366096866097, + "grad_norm": 1.0819852352142334, + "learning_rate": 6.6446329413040965e-06, + "loss": 0.9199, + "step": 19852 + }, + { + "epoch": 3.5345441595441596, + "grad_norm": 1.072750449180603, + "learning_rate": 6.6396166349363635e-06, + "loss": 0.7999, + "step": 19853 + }, + { + "epoch": 3.5347222222222223, + "grad_norm": 1.0215152502059937, + "learning_rate": 6.634602157771385e-06, + "loss": 0.8409, + "step": 19854 + }, + { + "epoch": 3.534900284900285, + "grad_norm": 0.9943056106567383, + "learning_rate": 6.629589509907464e-06, + "loss": 0.8393, + "step": 19855 + }, + { + "epoch": 3.535078347578348, + "grad_norm": 1.0051432847976685, + "learning_rate": 6.62457869144274e-06, + "loss": 0.821, + "step": 19856 + }, + { + "epoch": 3.53525641025641, + "grad_norm": 0.9929559230804443, + "learning_rate": 6.619569702475459e-06, + "loss": 0.8753, + "step": 19857 + }, + { + "epoch": 3.535434472934473, + "grad_norm": 1.0302678346633911, + "learning_rate": 6.6145625431036975e-06, + "loss": 0.9094, + "step": 19858 + }, + { + "epoch": 3.5356125356125356, + "grad_norm": 0.9745633602142334, + "learning_rate": 6.609557213425599e-06, + "loss": 0.8616, + "step": 19859 + }, + { + "epoch": 3.5357905982905984, + "grad_norm": 0.9746029376983643, + "learning_rate": 6.60455371353923e-06, + "loss": 0.9074, + "step": 19860 + }, + { + "epoch": 3.5359686609686607, + "grad_norm": 0.9690983295440674, + "learning_rate": 6.599552043542623e-06, + "loss": 0.788, + "step": 19861 + }, + { + "epoch": 3.5361467236467234, + "grad_norm": 1.0887675285339355, + "learning_rate": 6.594552203533766e-06, + "loss": 0.809, + "step": 19862 + }, + { + "epoch": 3.536324786324786, + "grad_norm": 0.943412721157074, + "learning_rate": 6.589554193610637e-06, + "loss": 0.6296, + "step": 19863 + }, + { + "epoch": 3.536502849002849, + "grad_norm": 0.9202658534049988, + "learning_rate": 6.584558013871123e-06, + "loss": 0.7163, + "step": 19864 + }, + { + "epoch": 3.5366809116809117, + "grad_norm": 0.8274959325790405, + "learning_rate": 6.57956366441318e-06, + "loss": 0.6825, + "step": 19865 + }, + { + "epoch": 3.5368589743589745, + "grad_norm": 0.9158870577812195, + "learning_rate": 6.574571145334585e-06, + "loss": 0.6954, + "step": 19866 + }, + { + "epoch": 3.537037037037037, + "grad_norm": 0.8894117474555969, + "learning_rate": 6.5695804567332044e-06, + "loss": 0.8603, + "step": 19867 + }, + { + "epoch": 3.5372150997151, + "grad_norm": 1.219153642654419, + "learning_rate": 6.564591598706815e-06, + "loss": 0.8096, + "step": 19868 + }, + { + "epoch": 3.5373931623931623, + "grad_norm": 0.8852741122245789, + "learning_rate": 6.5596045713531615e-06, + "loss": 0.6964, + "step": 19869 + }, + { + "epoch": 3.537571225071225, + "grad_norm": 0.9260098338127136, + "learning_rate": 6.554619374769955e-06, + "loss": 0.7107, + "step": 19870 + }, + { + "epoch": 3.5377492877492878, + "grad_norm": 0.960800051689148, + "learning_rate": 6.54963600905486e-06, + "loss": 0.5889, + "step": 19871 + }, + { + "epoch": 3.5379273504273505, + "grad_norm": 0.957720935344696, + "learning_rate": 6.544654474305523e-06, + "loss": 0.8725, + "step": 19872 + }, + { + "epoch": 3.5381054131054133, + "grad_norm": 0.8383321762084961, + "learning_rate": 6.539674770619542e-06, + "loss": 0.7994, + "step": 19873 + }, + { + "epoch": 3.5382834757834756, + "grad_norm": 0.9970546960830688, + "learning_rate": 6.534696898094472e-06, + "loss": 0.6841, + "step": 19874 + }, + { + "epoch": 3.5384615384615383, + "grad_norm": 0.987698495388031, + "learning_rate": 6.52972085682787e-06, + "loss": 0.8124, + "step": 19875 + }, + { + "epoch": 3.538639601139601, + "grad_norm": 0.9530283212661743, + "learning_rate": 6.524746646917224e-06, + "loss": 0.8111, + "step": 19876 + }, + { + "epoch": 3.538817663817664, + "grad_norm": 0.9642528295516968, + "learning_rate": 6.519774268459988e-06, + "loss": 0.6965, + "step": 19877 + }, + { + "epoch": 3.5389957264957266, + "grad_norm": 0.8841597437858582, + "learning_rate": 6.514803721553597e-06, + "loss": 0.7679, + "step": 19878 + }, + { + "epoch": 3.5391737891737893, + "grad_norm": 0.9791342616081238, + "learning_rate": 6.509835006295395e-06, + "loss": 0.72, + "step": 19879 + }, + { + "epoch": 3.539351851851852, + "grad_norm": 0.9851047396659851, + "learning_rate": 6.504868122782815e-06, + "loss": 0.7728, + "step": 19880 + }, + { + "epoch": 3.5395299145299144, + "grad_norm": 0.9055594801902771, + "learning_rate": 6.499903071113089e-06, + "loss": 0.4758, + "step": 19881 + }, + { + "epoch": 3.539707977207977, + "grad_norm": 1.0211628675460815, + "learning_rate": 6.494939851383553e-06, + "loss": 0.6912, + "step": 19882 + }, + { + "epoch": 3.53988603988604, + "grad_norm": 1.055820345878601, + "learning_rate": 6.489978463691415e-06, + "loss": 0.6928, + "step": 19883 + }, + { + "epoch": 3.5400641025641026, + "grad_norm": 0.8368833065032959, + "learning_rate": 6.485018908133911e-06, + "loss": 0.6069, + "step": 19884 + }, + { + "epoch": 3.5402421652421654, + "grad_norm": 0.8132014274597168, + "learning_rate": 6.480061184808195e-06, + "loss": 0.7176, + "step": 19885 + }, + { + "epoch": 3.5404202279202277, + "grad_norm": 0.8118695020675659, + "learning_rate": 6.475105293811412e-06, + "loss": 0.6785, + "step": 19886 + }, + { + "epoch": 3.5405982905982905, + "grad_norm": 0.9680670499801636, + "learning_rate": 6.470151235240651e-06, + "loss": 0.8413, + "step": 19887 + }, + { + "epoch": 3.540776353276353, + "grad_norm": 0.8878784775733948, + "learning_rate": 6.465199009193001e-06, + "loss": 0.7328, + "step": 19888 + }, + { + "epoch": 3.540954415954416, + "grad_norm": 0.8278130888938904, + "learning_rate": 6.460248615765452e-06, + "loss": 0.7212, + "step": 19889 + }, + { + "epoch": 3.5411324786324787, + "grad_norm": 1.0419317483901978, + "learning_rate": 6.455300055055047e-06, + "loss": 0.7404, + "step": 19890 + }, + { + "epoch": 3.5413105413105415, + "grad_norm": 0.9516051411628723, + "learning_rate": 6.450353327158687e-06, + "loss": 0.6078, + "step": 19891 + }, + { + "epoch": 3.541488603988604, + "grad_norm": 0.9726034998893738, + "learning_rate": 6.44540843217335e-06, + "loss": 0.8113, + "step": 19892 + }, + { + "epoch": 3.5416666666666665, + "grad_norm": 0.9923378229141235, + "learning_rate": 6.4404653701958695e-06, + "loss": 0.8888, + "step": 19893 + }, + { + "epoch": 3.5418447293447293, + "grad_norm": 1.088355302810669, + "learning_rate": 6.4355241413231234e-06, + "loss": 0.8549, + "step": 19894 + }, + { + "epoch": 3.542022792022792, + "grad_norm": 0.9767059087753296, + "learning_rate": 6.430584745651924e-06, + "loss": 0.8886, + "step": 19895 + }, + { + "epoch": 3.5422008547008548, + "grad_norm": 0.9931213855743408, + "learning_rate": 6.425647183279016e-06, + "loss": 0.7972, + "step": 19896 + }, + { + "epoch": 3.5423789173789175, + "grad_norm": 0.939339816570282, + "learning_rate": 6.42071145430121e-06, + "loss": 0.8591, + "step": 19897 + }, + { + "epoch": 3.54255698005698, + "grad_norm": 0.9261937141418457, + "learning_rate": 6.415777558815139e-06, + "loss": 0.7403, + "step": 19898 + }, + { + "epoch": 3.5427350427350426, + "grad_norm": 0.9660967588424683, + "learning_rate": 6.410845496917506e-06, + "loss": 0.7583, + "step": 19899 + }, + { + "epoch": 3.5429131054131053, + "grad_norm": 0.9808663725852966, + "learning_rate": 6.405915268704954e-06, + "loss": 0.5649, + "step": 19900 + }, + { + "epoch": 3.543091168091168, + "grad_norm": 0.9739185571670532, + "learning_rate": 6.4009868742740616e-06, + "loss": 0.7576, + "step": 19901 + }, + { + "epoch": 3.543269230769231, + "grad_norm": 0.9260064363479614, + "learning_rate": 6.396060313721397e-06, + "loss": 0.793, + "step": 19902 + }, + { + "epoch": 3.5434472934472936, + "grad_norm": 0.8840816617012024, + "learning_rate": 6.391135587143482e-06, + "loss": 0.7113, + "step": 19903 + }, + { + "epoch": 3.5436253561253563, + "grad_norm": 1.163144826889038, + "learning_rate": 6.386212694636795e-06, + "loss": 0.8255, + "step": 19904 + }, + { + "epoch": 3.5438034188034186, + "grad_norm": 0.9658638834953308, + "learning_rate": 6.381291636297837e-06, + "loss": 0.7442, + "step": 19905 + }, + { + "epoch": 3.5439814814814814, + "grad_norm": 1.0708949565887451, + "learning_rate": 6.376372412222964e-06, + "loss": 0.8555, + "step": 19906 + }, + { + "epoch": 3.544159544159544, + "grad_norm": 1.014028787612915, + "learning_rate": 6.371455022508621e-06, + "loss": 0.6674, + "step": 19907 + }, + { + "epoch": 3.544337606837607, + "grad_norm": 1.0435295104980469, + "learning_rate": 6.366539467251087e-06, + "loss": 1.0046, + "step": 19908 + }, + { + "epoch": 3.5445156695156697, + "grad_norm": 0.8781193494796753, + "learning_rate": 6.361625746546718e-06, + "loss": 0.6087, + "step": 19909 + }, + { + "epoch": 3.544693732193732, + "grad_norm": 1.0531107187271118, + "learning_rate": 6.356713860491781e-06, + "loss": 0.7646, + "step": 19910 + }, + { + "epoch": 3.5448717948717947, + "grad_norm": 0.9368358850479126, + "learning_rate": 6.3518038091825e-06, + "loss": 0.7514, + "step": 19911 + }, + { + "epoch": 3.5450498575498575, + "grad_norm": 0.9585633277893066, + "learning_rate": 6.346895592715096e-06, + "loss": 0.6294, + "step": 19912 + }, + { + "epoch": 3.54522792022792, + "grad_norm": 0.9690647721290588, + "learning_rate": 6.341989211185717e-06, + "loss": 0.7861, + "step": 19913 + }, + { + "epoch": 3.545405982905983, + "grad_norm": 0.9715308547019958, + "learning_rate": 6.3370846646904844e-06, + "loss": 1.0209, + "step": 19914 + }, + { + "epoch": 3.5455840455840457, + "grad_norm": 1.067807674407959, + "learning_rate": 6.332181953325534e-06, + "loss": 0.7392, + "step": 19915 + }, + { + "epoch": 3.5457621082621085, + "grad_norm": 0.7337046265602112, + "learning_rate": 6.327281077186897e-06, + "loss": 0.5182, + "step": 19916 + }, + { + "epoch": 3.5459401709401708, + "grad_norm": 0.9948704838752747, + "learning_rate": 6.3223820363706e-06, + "loss": 0.697, + "step": 19917 + }, + { + "epoch": 3.5461182336182335, + "grad_norm": 0.9387503862380981, + "learning_rate": 6.3174848309726305e-06, + "loss": 0.7389, + "step": 19918 + }, + { + "epoch": 3.5462962962962963, + "grad_norm": 1.026684284210205, + "learning_rate": 6.312589461088936e-06, + "loss": 0.9897, + "step": 19919 + }, + { + "epoch": 3.546474358974359, + "grad_norm": 0.8634870648384094, + "learning_rate": 6.307695926815426e-06, + "loss": 0.6171, + "step": 19920 + }, + { + "epoch": 3.546652421652422, + "grad_norm": 0.8596929311752319, + "learning_rate": 6.302804228247972e-06, + "loss": 0.7231, + "step": 19921 + }, + { + "epoch": 3.546830484330484, + "grad_norm": 0.8627797961235046, + "learning_rate": 6.2979143654824714e-06, + "loss": 0.6921, + "step": 19922 + }, + { + "epoch": 3.547008547008547, + "grad_norm": 0.9567093849182129, + "learning_rate": 6.29302633861465e-06, + "loss": 0.7992, + "step": 19923 + }, + { + "epoch": 3.5471866096866096, + "grad_norm": 0.9002968668937683, + "learning_rate": 6.288140147740329e-06, + "loss": 0.7066, + "step": 19924 + }, + { + "epoch": 3.5473646723646723, + "grad_norm": 1.0079301595687866, + "learning_rate": 6.283255792955245e-06, + "loss": 0.938, + "step": 19925 + }, + { + "epoch": 3.547542735042735, + "grad_norm": 1.0074832439422607, + "learning_rate": 6.278373274355076e-06, + "loss": 0.7939, + "step": 19926 + }, + { + "epoch": 3.547720797720798, + "grad_norm": 0.8892631530761719, + "learning_rate": 6.27349259203549e-06, + "loss": 0.8861, + "step": 19927 + }, + { + "epoch": 3.5478988603988606, + "grad_norm": 1.0393931865692139, + "learning_rate": 6.2686137460921226e-06, + "loss": 0.8119, + "step": 19928 + }, + { + "epoch": 3.5480769230769234, + "grad_norm": 1.089049220085144, + "learning_rate": 6.263736736620551e-06, + "loss": 0.8574, + "step": 19929 + }, + { + "epoch": 3.5482549857549857, + "grad_norm": 0.887886106967926, + "learning_rate": 6.258861563716345e-06, + "loss": 0.8319, + "step": 19930 + }, + { + "epoch": 3.5484330484330484, + "grad_norm": 1.0811010599136353, + "learning_rate": 6.253988227475005e-06, + "loss": 0.8569, + "step": 19931 + }, + { + "epoch": 3.548611111111111, + "grad_norm": 0.9962393045425415, + "learning_rate": 6.2491167279920435e-06, + "loss": 0.9301, + "step": 19932 + }, + { + "epoch": 3.548789173789174, + "grad_norm": 0.9498433470726013, + "learning_rate": 6.244247065362862e-06, + "loss": 0.6579, + "step": 19933 + }, + { + "epoch": 3.548967236467236, + "grad_norm": 0.958933413028717, + "learning_rate": 6.239379239682918e-06, + "loss": 0.7045, + "step": 19934 + }, + { + "epoch": 3.549145299145299, + "grad_norm": 0.9499155282974243, + "learning_rate": 6.234513251047569e-06, + "loss": 0.6971, + "step": 19935 + }, + { + "epoch": 3.5493233618233617, + "grad_norm": 0.9261385798454285, + "learning_rate": 6.2296490995521375e-06, + "loss": 0.7128, + "step": 19936 + }, + { + "epoch": 3.5495014245014245, + "grad_norm": 0.9812938570976257, + "learning_rate": 6.22478678529197e-06, + "loss": 0.7182, + "step": 19937 + }, + { + "epoch": 3.5496794871794872, + "grad_norm": 0.8759663105010986, + "learning_rate": 6.219926308362267e-06, + "loss": 0.7853, + "step": 19938 + }, + { + "epoch": 3.54985754985755, + "grad_norm": 1.1171659231185913, + "learning_rate": 6.2150676688583095e-06, + "loss": 0.8763, + "step": 19939 + }, + { + "epoch": 3.5500356125356127, + "grad_norm": 0.8224866986274719, + "learning_rate": 6.210210866875288e-06, + "loss": 0.6219, + "step": 19940 + }, + { + "epoch": 3.5502136752136755, + "grad_norm": 1.0393109321594238, + "learning_rate": 6.205355902508359e-06, + "loss": 0.7143, + "step": 19941 + }, + { + "epoch": 3.550391737891738, + "grad_norm": 0.9704436659812927, + "learning_rate": 6.200502775852623e-06, + "loss": 0.7377, + "step": 19942 + }, + { + "epoch": 3.5505698005698005, + "grad_norm": 1.0549170970916748, + "learning_rate": 6.1956514870031956e-06, + "loss": 0.676, + "step": 19943 + }, + { + "epoch": 3.5507478632478633, + "grad_norm": 0.944458544254303, + "learning_rate": 6.190802036055121e-06, + "loss": 0.678, + "step": 19944 + }, + { + "epoch": 3.550925925925926, + "grad_norm": 0.9729311466217041, + "learning_rate": 6.185954423103391e-06, + "loss": 0.814, + "step": 19945 + }, + { + "epoch": 3.5511039886039883, + "grad_norm": 0.6908536553382874, + "learning_rate": 6.181108648243006e-06, + "loss": 0.4705, + "step": 19946 + }, + { + "epoch": 3.551282051282051, + "grad_norm": 0.9184253215789795, + "learning_rate": 6.176264711568924e-06, + "loss": 0.7976, + "step": 19947 + }, + { + "epoch": 3.551460113960114, + "grad_norm": 1.1098743677139282, + "learning_rate": 6.171422613176003e-06, + "loss": 0.7885, + "step": 19948 + }, + { + "epoch": 3.5516381766381766, + "grad_norm": 0.9816890358924866, + "learning_rate": 6.166582353159156e-06, + "loss": 0.8283, + "step": 19949 + }, + { + "epoch": 3.5518162393162394, + "grad_norm": 1.26338529586792, + "learning_rate": 6.161743931613217e-06, + "loss": 0.7938, + "step": 19950 + }, + { + "epoch": 3.551994301994302, + "grad_norm": 0.8663501143455505, + "learning_rate": 6.156907348632968e-06, + "loss": 0.8333, + "step": 19951 + }, + { + "epoch": 3.552172364672365, + "grad_norm": 0.9667189717292786, + "learning_rate": 6.152072604313175e-06, + "loss": 0.7434, + "step": 19952 + }, + { + "epoch": 3.5523504273504276, + "grad_norm": 1.1370351314544678, + "learning_rate": 6.147239698748553e-06, + "loss": 0.6546, + "step": 19953 + }, + { + "epoch": 3.55252849002849, + "grad_norm": 0.8451266884803772, + "learning_rate": 6.142408632033836e-06, + "loss": 0.6856, + "step": 19954 + }, + { + "epoch": 3.5527065527065527, + "grad_norm": 0.7988196611404419, + "learning_rate": 6.137579404263627e-06, + "loss": 0.6, + "step": 19955 + }, + { + "epoch": 3.5528846153846154, + "grad_norm": 0.9842251539230347, + "learning_rate": 6.132752015532572e-06, + "loss": 0.9265, + "step": 19956 + }, + { + "epoch": 3.553062678062678, + "grad_norm": 1.0470566749572754, + "learning_rate": 6.127926465935252e-06, + "loss": 0.7726, + "step": 19957 + }, + { + "epoch": 3.5532407407407405, + "grad_norm": 0.9638791084289551, + "learning_rate": 6.1231027555662125e-06, + "loss": 0.8386, + "step": 19958 + }, + { + "epoch": 3.5534188034188032, + "grad_norm": 0.9695351123809814, + "learning_rate": 6.1182808845199665e-06, + "loss": 0.8146, + "step": 19959 + }, + { + "epoch": 3.553596866096866, + "grad_norm": 1.1904405355453491, + "learning_rate": 6.1134608528909735e-06, + "loss": 0.8364, + "step": 19960 + }, + { + "epoch": 3.5537749287749287, + "grad_norm": 0.9649089574813843, + "learning_rate": 6.108642660773678e-06, + "loss": 0.5882, + "step": 19961 + }, + { + "epoch": 3.5539529914529915, + "grad_norm": 1.0587506294250488, + "learning_rate": 6.103826308262517e-06, + "loss": 0.6633, + "step": 19962 + }, + { + "epoch": 3.5541310541310542, + "grad_norm": 0.9397716522216797, + "learning_rate": 6.099011795451792e-06, + "loss": 0.7343, + "step": 19963 + }, + { + "epoch": 3.554309116809117, + "grad_norm": 0.8498682975769043, + "learning_rate": 6.094199122435895e-06, + "loss": 0.6133, + "step": 19964 + }, + { + "epoch": 3.5544871794871797, + "grad_norm": 1.1127020120620728, + "learning_rate": 6.089388289309073e-06, + "loss": 0.9287, + "step": 19965 + }, + { + "epoch": 3.554665242165242, + "grad_norm": 0.9622563719749451, + "learning_rate": 6.084579296165605e-06, + "loss": 0.6882, + "step": 19966 + }, + { + "epoch": 3.554843304843305, + "grad_norm": 1.1055150032043457, + "learning_rate": 6.079772143099726e-06, + "loss": 0.7718, + "step": 19967 + }, + { + "epoch": 3.5550213675213675, + "grad_norm": 0.9930889010429382, + "learning_rate": 6.074966830205608e-06, + "loss": 0.6884, + "step": 19968 + }, + { + "epoch": 3.5551994301994303, + "grad_norm": 0.8389014601707458, + "learning_rate": 6.070163357577408e-06, + "loss": 0.6328, + "step": 19969 + }, + { + "epoch": 3.5553774928774926, + "grad_norm": 0.8436742424964905, + "learning_rate": 6.065361725309227e-06, + "loss": 0.7243, + "step": 19970 + }, + { + "epoch": 3.5555555555555554, + "grad_norm": 0.8762266635894775, + "learning_rate": 6.060561933495135e-06, + "loss": 0.8242, + "step": 19971 + }, + { + "epoch": 3.555733618233618, + "grad_norm": 1.049564003944397, + "learning_rate": 6.055763982229224e-06, + "loss": 0.7959, + "step": 19972 + }, + { + "epoch": 3.555911680911681, + "grad_norm": 0.8714830279350281, + "learning_rate": 6.050967871605429e-06, + "loss": 0.7953, + "step": 19973 + }, + { + "epoch": 3.5560897435897436, + "grad_norm": 1.0046193599700928, + "learning_rate": 6.046173601717775e-06, + "loss": 1.0607, + "step": 19974 + }, + { + "epoch": 3.5562678062678064, + "grad_norm": 0.8975396156311035, + "learning_rate": 6.041381172660177e-06, + "loss": 0.66, + "step": 19975 + }, + { + "epoch": 3.556445868945869, + "grad_norm": 1.0665318965911865, + "learning_rate": 6.036590584526525e-06, + "loss": 0.8612, + "step": 19976 + }, + { + "epoch": 3.556623931623932, + "grad_norm": 0.908662736415863, + "learning_rate": 6.031801837410711e-06, + "loss": 0.6435, + "step": 19977 + }, + { + "epoch": 3.556801994301994, + "grad_norm": 0.9590704441070557, + "learning_rate": 6.027014931406505e-06, + "loss": 0.6757, + "step": 19978 + }, + { + "epoch": 3.556980056980057, + "grad_norm": 1.0787307024002075, + "learning_rate": 6.022229866607765e-06, + "loss": 0.8774, + "step": 19979 + }, + { + "epoch": 3.5571581196581197, + "grad_norm": 0.9405365586280823, + "learning_rate": 6.017446643108171e-06, + "loss": 0.7256, + "step": 19980 + }, + { + "epoch": 3.5573361823361824, + "grad_norm": 1.0895098447799683, + "learning_rate": 6.0126652610015045e-06, + "loss": 0.8127, + "step": 19981 + }, + { + "epoch": 3.557514245014245, + "grad_norm": 0.8931481838226318, + "learning_rate": 6.007885720381412e-06, + "loss": 0.6243, + "step": 19982 + }, + { + "epoch": 3.5576923076923075, + "grad_norm": 0.9177520275115967, + "learning_rate": 6.003108021341553e-06, + "loss": 0.7058, + "step": 19983 + }, + { + "epoch": 3.5578703703703702, + "grad_norm": 1.112932801246643, + "learning_rate": 5.998332163975528e-06, + "loss": 0.8612, + "step": 19984 + }, + { + "epoch": 3.558048433048433, + "grad_norm": 1.0220192670822144, + "learning_rate": 5.99355814837691e-06, + "loss": 0.7389, + "step": 19985 + }, + { + "epoch": 3.5582264957264957, + "grad_norm": 0.9819709062576294, + "learning_rate": 5.988785974639232e-06, + "loss": 0.7656, + "step": 19986 + }, + { + "epoch": 3.5584045584045585, + "grad_norm": 1.1695373058319092, + "learning_rate": 5.984015642856022e-06, + "loss": 0.959, + "step": 19987 + }, + { + "epoch": 3.5585826210826212, + "grad_norm": 0.9889600276947021, + "learning_rate": 5.979247153120693e-06, + "loss": 0.8078, + "step": 19988 + }, + { + "epoch": 3.558760683760684, + "grad_norm": 1.030166506767273, + "learning_rate": 5.974480505526747e-06, + "loss": 0.8582, + "step": 19989 + }, + { + "epoch": 3.5589387464387463, + "grad_norm": 0.9108613729476929, + "learning_rate": 5.9697157001674885e-06, + "loss": 0.8143, + "step": 19990 + }, + { + "epoch": 3.559116809116809, + "grad_norm": 0.9014456868171692, + "learning_rate": 5.964952737136353e-06, + "loss": 0.7242, + "step": 19991 + }, + { + "epoch": 3.559294871794872, + "grad_norm": 0.9596982598304749, + "learning_rate": 5.960191616526612e-06, + "loss": 0.7093, + "step": 19992 + }, + { + "epoch": 3.5594729344729346, + "grad_norm": 0.9499574899673462, + "learning_rate": 5.955432338431555e-06, + "loss": 0.6325, + "step": 19993 + }, + { + "epoch": 3.5596509971509973, + "grad_norm": 0.9167241454124451, + "learning_rate": 5.950674902944475e-06, + "loss": 0.7084, + "step": 19994 + }, + { + "epoch": 3.5598290598290596, + "grad_norm": 0.8762951493263245, + "learning_rate": 5.945919310158521e-06, + "loss": 0.6922, + "step": 19995 + }, + { + "epoch": 3.5600071225071224, + "grad_norm": 1.0726141929626465, + "learning_rate": 5.941165560166906e-06, + "loss": 0.7219, + "step": 19996 + }, + { + "epoch": 3.560185185185185, + "grad_norm": 1.2705079317092896, + "learning_rate": 5.936413653062767e-06, + "loss": 0.9296, + "step": 19997 + }, + { + "epoch": 3.560363247863248, + "grad_norm": 0.9176344275474548, + "learning_rate": 5.931663588939196e-06, + "loss": 0.8195, + "step": 19998 + }, + { + "epoch": 3.5605413105413106, + "grad_norm": 1.0315814018249512, + "learning_rate": 5.926915367889274e-06, + "loss": 0.8144, + "step": 19999 + }, + { + "epoch": 3.5607193732193734, + "grad_norm": 0.9186922907829285, + "learning_rate": 5.922168990006016e-06, + "loss": 0.8895, + "step": 20000 + }, + { + "epoch": 3.560897435897436, + "grad_norm": 1.0576261281967163, + "learning_rate": 5.917424455382437e-06, + "loss": 0.9221, + "step": 20001 + }, + { + "epoch": 3.5610754985754984, + "grad_norm": 0.8663935661315918, + "learning_rate": 5.912681764111483e-06, + "loss": 0.6516, + "step": 20002 + }, + { + "epoch": 3.561253561253561, + "grad_norm": 0.9327130913734436, + "learning_rate": 5.907940916286059e-06, + "loss": 0.7252, + "step": 20003 + }, + { + "epoch": 3.561431623931624, + "grad_norm": 0.9112497568130493, + "learning_rate": 5.903201911999112e-06, + "loss": 0.6984, + "step": 20004 + }, + { + "epoch": 3.5616096866096867, + "grad_norm": 0.8391700983047485, + "learning_rate": 5.8984647513434245e-06, + "loss": 0.7863, + "step": 20005 + }, + { + "epoch": 3.5617877492877494, + "grad_norm": 0.9279758334159851, + "learning_rate": 5.893729434411854e-06, + "loss": 0.7333, + "step": 20006 + }, + { + "epoch": 3.5619658119658117, + "grad_norm": 0.9002404808998108, + "learning_rate": 5.888995961297161e-06, + "loss": 0.557, + "step": 20007 + }, + { + "epoch": 3.5621438746438745, + "grad_norm": 0.9967086315155029, + "learning_rate": 5.884264332092104e-06, + "loss": 0.6941, + "step": 20008 + }, + { + "epoch": 3.5623219373219372, + "grad_norm": 1.030300498008728, + "learning_rate": 5.879534546889365e-06, + "loss": 0.7942, + "step": 20009 + }, + { + "epoch": 3.5625, + "grad_norm": 0.9536016583442688, + "learning_rate": 5.8748066057816354e-06, + "loss": 0.7312, + "step": 20010 + }, + { + "epoch": 3.5626780626780628, + "grad_norm": 1.1330475807189941, + "learning_rate": 5.870080508861531e-06, + "loss": 0.7598, + "step": 20011 + }, + { + "epoch": 3.5628561253561255, + "grad_norm": 0.9957557916641235, + "learning_rate": 5.865356256221688e-06, + "loss": 0.846, + "step": 20012 + }, + { + "epoch": 3.5630341880341883, + "grad_norm": 1.0001511573791504, + "learning_rate": 5.860633847954611e-06, + "loss": 0.7996, + "step": 20013 + }, + { + "epoch": 3.5632122507122506, + "grad_norm": 1.0668739080429077, + "learning_rate": 5.8559132841528696e-06, + "loss": 0.7621, + "step": 20014 + }, + { + "epoch": 3.5633903133903133, + "grad_norm": 1.0167616605758667, + "learning_rate": 5.851194564908946e-06, + "loss": 0.8459, + "step": 20015 + }, + { + "epoch": 3.563568376068376, + "grad_norm": 1.017533302307129, + "learning_rate": 5.8464776903152775e-06, + "loss": 0.4954, + "step": 20016 + }, + { + "epoch": 3.563746438746439, + "grad_norm": 1.1081538200378418, + "learning_rate": 5.8417626604643005e-06, + "loss": 0.8442, + "step": 20017 + }, + { + "epoch": 3.5639245014245016, + "grad_norm": 0.9077942967414856, + "learning_rate": 5.837049475448375e-06, + "loss": 0.8122, + "step": 20018 + }, + { + "epoch": 3.564102564102564, + "grad_norm": 0.8072832822799683, + "learning_rate": 5.832338135359883e-06, + "loss": 0.6718, + "step": 20019 + }, + { + "epoch": 3.5642806267806266, + "grad_norm": 0.928887665271759, + "learning_rate": 5.8276286402910715e-06, + "loss": 0.7962, + "step": 20020 + }, + { + "epoch": 3.5644586894586894, + "grad_norm": 0.9171196818351746, + "learning_rate": 5.82292099033428e-06, + "loss": 0.7408, + "step": 20021 + }, + { + "epoch": 3.564636752136752, + "grad_norm": 0.900301992893219, + "learning_rate": 5.818215185581699e-06, + "loss": 0.7775, + "step": 20022 + }, + { + "epoch": 3.564814814814815, + "grad_norm": 1.0731996297836304, + "learning_rate": 5.813511226125557e-06, + "loss": 0.8732, + "step": 20023 + }, + { + "epoch": 3.5649928774928776, + "grad_norm": 1.1355241537094116, + "learning_rate": 5.808809112058011e-06, + "loss": 0.7019, + "step": 20024 + }, + { + "epoch": 3.5651709401709404, + "grad_norm": 0.8540470600128174, + "learning_rate": 5.804108843471179e-06, + "loss": 0.6102, + "step": 20025 + }, + { + "epoch": 3.5653490028490027, + "grad_norm": 0.9603937268257141, + "learning_rate": 5.799410420457163e-06, + "loss": 0.7485, + "step": 20026 + }, + { + "epoch": 3.5655270655270654, + "grad_norm": 0.9940163493156433, + "learning_rate": 5.794713843108013e-06, + "loss": 0.7051, + "step": 20027 + }, + { + "epoch": 3.565705128205128, + "grad_norm": 0.9198596477508545, + "learning_rate": 5.790019111515732e-06, + "loss": 0.7189, + "step": 20028 + }, + { + "epoch": 3.565883190883191, + "grad_norm": 0.8233235478401184, + "learning_rate": 5.7853262257723476e-06, + "loss": 0.5998, + "step": 20029 + }, + { + "epoch": 3.5660612535612537, + "grad_norm": 1.0005290508270264, + "learning_rate": 5.780635185969762e-06, + "loss": 0.7284, + "step": 20030 + }, + { + "epoch": 3.566239316239316, + "grad_norm": 0.8798156380653381, + "learning_rate": 5.775945992199916e-06, + "loss": 0.6115, + "step": 20031 + }, + { + "epoch": 3.5664173789173788, + "grad_norm": 0.9834653735160828, + "learning_rate": 5.771258644554678e-06, + "loss": 0.6293, + "step": 20032 + }, + { + "epoch": 3.5665954415954415, + "grad_norm": 0.929243266582489, + "learning_rate": 5.7665731431258755e-06, + "loss": 0.7863, + "step": 20033 + }, + { + "epoch": 3.5667735042735043, + "grad_norm": 1.131831169128418, + "learning_rate": 5.7618894880053344e-06, + "loss": 0.8469, + "step": 20034 + }, + { + "epoch": 3.566951566951567, + "grad_norm": 0.9003351330757141, + "learning_rate": 5.757207679284782e-06, + "loss": 0.6531, + "step": 20035 + }, + { + "epoch": 3.5671296296296298, + "grad_norm": 0.9676564931869507, + "learning_rate": 5.752527717056011e-06, + "loss": 0.7791, + "step": 20036 + }, + { + "epoch": 3.5673076923076925, + "grad_norm": 0.9569422602653503, + "learning_rate": 5.7478496014106374e-06, + "loss": 0.8359, + "step": 20037 + }, + { + "epoch": 3.567485754985755, + "grad_norm": 1.0955125093460083, + "learning_rate": 5.743173332440377e-06, + "loss": 0.6683, + "step": 20038 + }, + { + "epoch": 3.5676638176638176, + "grad_norm": 1.0164920091629028, + "learning_rate": 5.738498910236834e-06, + "loss": 0.7597, + "step": 20039 + }, + { + "epoch": 3.5678418803418803, + "grad_norm": 0.977257490158081, + "learning_rate": 5.73382633489159e-06, + "loss": 0.8192, + "step": 20040 + }, + { + "epoch": 3.568019943019943, + "grad_norm": 0.9285514950752258, + "learning_rate": 5.729155606496195e-06, + "loss": 0.8168, + "step": 20041 + }, + { + "epoch": 3.568198005698006, + "grad_norm": 1.0951110124588013, + "learning_rate": 5.724486725142175e-06, + "loss": 0.8083, + "step": 20042 + }, + { + "epoch": 3.568376068376068, + "grad_norm": 0.9273848533630371, + "learning_rate": 5.71981969092098e-06, + "loss": 0.6098, + "step": 20043 + }, + { + "epoch": 3.568554131054131, + "grad_norm": 1.0327738523483276, + "learning_rate": 5.7151545039240915e-06, + "loss": 0.828, + "step": 20044 + }, + { + "epoch": 3.5687321937321936, + "grad_norm": 1.0657696723937988, + "learning_rate": 5.71049116424287e-06, + "loss": 0.787, + "step": 20045 + }, + { + "epoch": 3.5689102564102564, + "grad_norm": 0.9822142720222473, + "learning_rate": 5.7058296719687095e-06, + "loss": 0.8614, + "step": 20046 + }, + { + "epoch": 3.569088319088319, + "grad_norm": 0.8666147589683533, + "learning_rate": 5.701170027192948e-06, + "loss": 0.6601, + "step": 20047 + }, + { + "epoch": 3.569266381766382, + "grad_norm": 0.9923439621925354, + "learning_rate": 5.696512230006867e-06, + "loss": 0.7775, + "step": 20048 + }, + { + "epoch": 3.5694444444444446, + "grad_norm": 0.9324108958244324, + "learning_rate": 5.691856280501728e-06, + "loss": 0.7463, + "step": 20049 + }, + { + "epoch": 3.5696225071225074, + "grad_norm": 0.9872146844863892, + "learning_rate": 5.687202178768758e-06, + "loss": 0.6716, + "step": 20050 + }, + { + "epoch": 3.5698005698005697, + "grad_norm": 1.0207147598266602, + "learning_rate": 5.682549924899139e-06, + "loss": 0.7594, + "step": 20051 + }, + { + "epoch": 3.5699786324786325, + "grad_norm": 1.0404084920883179, + "learning_rate": 5.677899518984031e-06, + "loss": 0.7924, + "step": 20052 + }, + { + "epoch": 3.570156695156695, + "grad_norm": 0.996023416519165, + "learning_rate": 5.673250961114529e-06, + "loss": 0.7175, + "step": 20053 + }, + { + "epoch": 3.570334757834758, + "grad_norm": 0.8972271084785461, + "learning_rate": 5.668604251381748e-06, + "loss": 0.7454, + "step": 20054 + }, + { + "epoch": 3.5705128205128203, + "grad_norm": 0.9665964841842651, + "learning_rate": 5.663959389876705e-06, + "loss": 0.8127, + "step": 20055 + }, + { + "epoch": 3.570690883190883, + "grad_norm": 0.9740979075431824, + "learning_rate": 5.659316376690416e-06, + "loss": 0.9236, + "step": 20056 + }, + { + "epoch": 3.5708689458689458, + "grad_norm": 0.972089946269989, + "learning_rate": 5.65467521191384e-06, + "loss": 0.8145, + "step": 20057 + }, + { + "epoch": 3.5710470085470085, + "grad_norm": 0.9361711740493774, + "learning_rate": 5.650035895637906e-06, + "loss": 0.665, + "step": 20058 + }, + { + "epoch": 3.5712250712250713, + "grad_norm": 0.9758422374725342, + "learning_rate": 5.645398427953552e-06, + "loss": 0.7318, + "step": 20059 + }, + { + "epoch": 3.571403133903134, + "grad_norm": 0.8265459537506104, + "learning_rate": 5.640762808951583e-06, + "loss": 0.6444, + "step": 20060 + }, + { + "epoch": 3.5715811965811968, + "grad_norm": 1.0095800161361694, + "learning_rate": 5.6361290387228814e-06, + "loss": 0.8519, + "step": 20061 + }, + { + "epoch": 3.5717592592592595, + "grad_norm": 0.9909504055976868, + "learning_rate": 5.631497117358175e-06, + "loss": 0.8529, + "step": 20062 + }, + { + "epoch": 3.571937321937322, + "grad_norm": 0.7235755920410156, + "learning_rate": 5.62686704494827e-06, + "loss": 0.5298, + "step": 20063 + }, + { + "epoch": 3.5721153846153846, + "grad_norm": 0.991928219795227, + "learning_rate": 5.622238821583858e-06, + "loss": 0.6757, + "step": 20064 + }, + { + "epoch": 3.5722934472934473, + "grad_norm": 0.9465965628623962, + "learning_rate": 5.617612447355624e-06, + "loss": 0.6856, + "step": 20065 + }, + { + "epoch": 3.57247150997151, + "grad_norm": 0.9528710246086121, + "learning_rate": 5.612987922354207e-06, + "loss": 0.758, + "step": 20066 + }, + { + "epoch": 3.5726495726495724, + "grad_norm": 1.0337190628051758, + "learning_rate": 5.608365246670233e-06, + "loss": 0.742, + "step": 20067 + }, + { + "epoch": 3.572827635327635, + "grad_norm": 0.9663426876068115, + "learning_rate": 5.60374442039423e-06, + "loss": 0.7674, + "step": 20068 + }, + { + "epoch": 3.573005698005698, + "grad_norm": 0.7379756569862366, + "learning_rate": 5.599125443616804e-06, + "loss": 0.516, + "step": 20069 + }, + { + "epoch": 3.5731837606837606, + "grad_norm": 0.932037889957428, + "learning_rate": 5.5945083164283815e-06, + "loss": 0.7166, + "step": 20070 + }, + { + "epoch": 3.5733618233618234, + "grad_norm": 0.9008628726005554, + "learning_rate": 5.5898930389194915e-06, + "loss": 0.7829, + "step": 20071 + }, + { + "epoch": 3.573539886039886, + "grad_norm": 0.8821158409118652, + "learning_rate": 5.585279611180494e-06, + "loss": 0.8116, + "step": 20072 + }, + { + "epoch": 3.573717948717949, + "grad_norm": 0.9247183203697205, + "learning_rate": 5.580668033301817e-06, + "loss": 0.8132, + "step": 20073 + }, + { + "epoch": 3.5738960113960117, + "grad_norm": 0.983182430267334, + "learning_rate": 5.576058305373822e-06, + "loss": 0.6751, + "step": 20074 + }, + { + "epoch": 3.574074074074074, + "grad_norm": 0.9759270548820496, + "learning_rate": 5.5714504274868015e-06, + "loss": 0.7332, + "step": 20075 + }, + { + "epoch": 3.5742521367521367, + "grad_norm": 1.0891354084014893, + "learning_rate": 5.566844399731086e-06, + "loss": 0.7377, + "step": 20076 + }, + { + "epoch": 3.5744301994301995, + "grad_norm": 1.0494391918182373, + "learning_rate": 5.5622402221968465e-06, + "loss": 0.8022, + "step": 20077 + }, + { + "epoch": 3.574608262108262, + "grad_norm": 0.9215951561927795, + "learning_rate": 5.557637894974354e-06, + "loss": 0.695, + "step": 20078 + }, + { + "epoch": 3.5747863247863245, + "grad_norm": 1.1212730407714844, + "learning_rate": 5.55303741815375e-06, + "loss": 0.8315, + "step": 20079 + }, + { + "epoch": 3.5749643874643873, + "grad_norm": 0.816863477230072, + "learning_rate": 5.548438791825195e-06, + "loss": 0.9294, + "step": 20080 + }, + { + "epoch": 3.57514245014245, + "grad_norm": 0.9771558046340942, + "learning_rate": 5.543842016078771e-06, + "loss": 0.783, + "step": 20081 + }, + { + "epoch": 3.5753205128205128, + "grad_norm": 0.8157296776771545, + "learning_rate": 5.53924709100454e-06, + "loss": 0.572, + "step": 20082 + }, + { + "epoch": 3.5754985754985755, + "grad_norm": 0.8209017515182495, + "learning_rate": 5.5346540166925305e-06, + "loss": 0.5762, + "step": 20083 + }, + { + "epoch": 3.5756766381766383, + "grad_norm": 0.9781062006950378, + "learning_rate": 5.530062793232771e-06, + "loss": 0.8179, + "step": 20084 + }, + { + "epoch": 3.575854700854701, + "grad_norm": 0.874980092048645, + "learning_rate": 5.525473420715155e-06, + "loss": 0.6966, + "step": 20085 + }, + { + "epoch": 3.576032763532764, + "grad_norm": 0.9744297862052917, + "learning_rate": 5.520885899229655e-06, + "loss": 0.9015, + "step": 20086 + }, + { + "epoch": 3.576210826210826, + "grad_norm": 0.8454684615135193, + "learning_rate": 5.516300228866111e-06, + "loss": 0.5204, + "step": 20087 + }, + { + "epoch": 3.576388888888889, + "grad_norm": 0.8786317110061646, + "learning_rate": 5.511716409714407e-06, + "loss": 0.7868, + "step": 20088 + }, + { + "epoch": 3.5765669515669516, + "grad_norm": 0.8432426452636719, + "learning_rate": 5.507134441864337e-06, + "loss": 0.6851, + "step": 20089 + }, + { + "epoch": 3.5767450142450143, + "grad_norm": 1.0218333005905151, + "learning_rate": 5.502554325405674e-06, + "loss": 0.836, + "step": 20090 + }, + { + "epoch": 3.5769230769230766, + "grad_norm": 0.8945069909095764, + "learning_rate": 5.497976060428156e-06, + "loss": 0.593, + "step": 20091 + }, + { + "epoch": 3.5771011396011394, + "grad_norm": 0.9170123934745789, + "learning_rate": 5.493399647021469e-06, + "loss": 0.8404, + "step": 20092 + }, + { + "epoch": 3.577279202279202, + "grad_norm": 0.8489331603050232, + "learning_rate": 5.488825085275317e-06, + "loss": 0.7002, + "step": 20093 + }, + { + "epoch": 3.577457264957265, + "grad_norm": 0.9071032404899597, + "learning_rate": 5.484252375279297e-06, + "loss": 0.79, + "step": 20094 + }, + { + "epoch": 3.5776353276353277, + "grad_norm": 1.0012317895889282, + "learning_rate": 5.479681517123026e-06, + "loss": 0.6822, + "step": 20095 + }, + { + "epoch": 3.5778133903133904, + "grad_norm": 0.9741858839988708, + "learning_rate": 5.475112510896041e-06, + "loss": 0.7569, + "step": 20096 + }, + { + "epoch": 3.577991452991453, + "grad_norm": 0.8876017928123474, + "learning_rate": 5.470545356687862e-06, + "loss": 0.8028, + "step": 20097 + }, + { + "epoch": 3.578169515669516, + "grad_norm": 0.8617599606513977, + "learning_rate": 5.465980054587983e-06, + "loss": 0.7123, + "step": 20098 + }, + { + "epoch": 3.578347578347578, + "grad_norm": 0.844270646572113, + "learning_rate": 5.461416604685854e-06, + "loss": 0.6884, + "step": 20099 + }, + { + "epoch": 3.578525641025641, + "grad_norm": 0.9308794140815735, + "learning_rate": 5.45685500707086e-06, + "loss": 0.6827, + "step": 20100 + }, + { + "epoch": 3.5787037037037037, + "grad_norm": 0.8432201743125916, + "learning_rate": 5.452295261832418e-06, + "loss": 0.7057, + "step": 20101 + }, + { + "epoch": 3.5788817663817665, + "grad_norm": 1.0532596111297607, + "learning_rate": 5.447737369059824e-06, + "loss": 0.7688, + "step": 20102 + }, + { + "epoch": 3.5790598290598292, + "grad_norm": 0.8805003762245178, + "learning_rate": 5.443181328842417e-06, + "loss": 0.6596, + "step": 20103 + }, + { + "epoch": 3.5792378917378915, + "grad_norm": 0.8240517973899841, + "learning_rate": 5.438627141269437e-06, + "loss": 0.5545, + "step": 20104 + }, + { + "epoch": 3.5794159544159543, + "grad_norm": 0.9336821436882019, + "learning_rate": 5.434074806430134e-06, + "loss": 0.7931, + "step": 20105 + }, + { + "epoch": 3.579594017094017, + "grad_norm": 0.978220522403717, + "learning_rate": 5.429524324413693e-06, + "loss": 0.9601, + "step": 20106 + }, + { + "epoch": 3.57977207977208, + "grad_norm": 0.9296578764915466, + "learning_rate": 5.424975695309265e-06, + "loss": 0.8462, + "step": 20107 + }, + { + "epoch": 3.5799501424501425, + "grad_norm": 0.9612907767295837, + "learning_rate": 5.4204289192059664e-06, + "loss": 0.6046, + "step": 20108 + }, + { + "epoch": 3.5801282051282053, + "grad_norm": 0.9216350317001343, + "learning_rate": 5.415883996192905e-06, + "loss": 0.7423, + "step": 20109 + }, + { + "epoch": 3.580306267806268, + "grad_norm": 0.9715888500213623, + "learning_rate": 5.4113409263590985e-06, + "loss": 0.6903, + "step": 20110 + }, + { + "epoch": 3.5804843304843303, + "grad_norm": 0.9958703517913818, + "learning_rate": 5.406799709793597e-06, + "loss": 0.7843, + "step": 20111 + }, + { + "epoch": 3.580662393162393, + "grad_norm": 0.9611614942550659, + "learning_rate": 5.40226034658533e-06, + "loss": 0.7979, + "step": 20112 + }, + { + "epoch": 3.580840455840456, + "grad_norm": 0.9550825953483582, + "learning_rate": 5.397722836823282e-06, + "loss": 0.7704, + "step": 20113 + }, + { + "epoch": 3.5810185185185186, + "grad_norm": 0.8597258925437927, + "learning_rate": 5.393187180596326e-06, + "loss": 1.0874, + "step": 20114 + }, + { + "epoch": 3.5811965811965814, + "grad_norm": 1.0289723873138428, + "learning_rate": 5.388653377993324e-06, + "loss": 0.7821, + "step": 20115 + }, + { + "epoch": 3.5813746438746437, + "grad_norm": 1.0703620910644531, + "learning_rate": 5.384121429103161e-06, + "loss": 0.7892, + "step": 20116 + }, + { + "epoch": 3.5815527065527064, + "grad_norm": 1.0069445371627808, + "learning_rate": 5.379591334014556e-06, + "loss": 0.6241, + "step": 20117 + }, + { + "epoch": 3.581730769230769, + "grad_norm": 1.050201654434204, + "learning_rate": 5.375063092816313e-06, + "loss": 0.5873, + "step": 20118 + }, + { + "epoch": 3.581908831908832, + "grad_norm": 0.9159594774246216, + "learning_rate": 5.370536705597151e-06, + "loss": 0.7451, + "step": 20119 + }, + { + "epoch": 3.5820868945868947, + "grad_norm": 0.9135583639144897, + "learning_rate": 5.366012172445734e-06, + "loss": 0.6258, + "step": 20120 + }, + { + "epoch": 3.5822649572649574, + "grad_norm": 0.9547392725944519, + "learning_rate": 5.3614894934507335e-06, + "loss": 0.8267, + "step": 20121 + }, + { + "epoch": 3.58244301994302, + "grad_norm": 0.9301168918609619, + "learning_rate": 5.356968668700746e-06, + "loss": 0.6885, + "step": 20122 + }, + { + "epoch": 3.5826210826210825, + "grad_norm": 0.9445184469223022, + "learning_rate": 5.352449698284356e-06, + "loss": 0.6854, + "step": 20123 + }, + { + "epoch": 3.5827991452991452, + "grad_norm": 0.9877057075500488, + "learning_rate": 5.347932582290105e-06, + "loss": 0.9167, + "step": 20124 + }, + { + "epoch": 3.582977207977208, + "grad_norm": 0.8706746697425842, + "learning_rate": 5.343417320806477e-06, + "loss": 0.8327, + "step": 20125 + }, + { + "epoch": 3.5831552706552707, + "grad_norm": 0.9573389887809753, + "learning_rate": 5.3389039139219775e-06, + "loss": 0.7544, + "step": 20126 + }, + { + "epoch": 3.5833333333333335, + "grad_norm": 1.0821263790130615, + "learning_rate": 5.334392361724994e-06, + "loss": 0.6362, + "step": 20127 + }, + { + "epoch": 3.583511396011396, + "grad_norm": 1.0245816707611084, + "learning_rate": 5.329882664303943e-06, + "loss": 0.7596, + "step": 20128 + }, + { + "epoch": 3.5836894586894585, + "grad_norm": 0.9184259176254272, + "learning_rate": 5.325374821747197e-06, + "loss": 0.8477, + "step": 20129 + }, + { + "epoch": 3.5838675213675213, + "grad_norm": 0.983049213886261, + "learning_rate": 5.320868834143056e-06, + "loss": 0.7234, + "step": 20130 + }, + { + "epoch": 3.584045584045584, + "grad_norm": 0.9151159524917603, + "learning_rate": 5.316364701579801e-06, + "loss": 0.6523, + "step": 20131 + }, + { + "epoch": 3.584223646723647, + "grad_norm": 0.9754053950309753, + "learning_rate": 5.311862424145686e-06, + "loss": 0.7408, + "step": 20132 + }, + { + "epoch": 3.5844017094017095, + "grad_norm": 0.8654192090034485, + "learning_rate": 5.307362001928961e-06, + "loss": 0.6214, + "step": 20133 + }, + { + "epoch": 3.5845797720797723, + "grad_norm": 0.9617530703544617, + "learning_rate": 5.302863435017735e-06, + "loss": 0.7405, + "step": 20134 + }, + { + "epoch": 3.5847578347578346, + "grad_norm": Infinity, + "learning_rate": 5.302863435017735e-06, + "loss": 0.6821, + "step": 20135 + }, + { + "epoch": 3.5849358974358974, + "grad_norm": 0.9110551476478577, + "learning_rate": 5.298366723500203e-06, + "loss": 0.6208, + "step": 20136 + }, + { + "epoch": 3.58511396011396, + "grad_norm": 1.0188283920288086, + "learning_rate": 5.29387186746445e-06, + "loss": 0.809, + "step": 20137 + }, + { + "epoch": 3.585292022792023, + "grad_norm": 1.1383823156356812, + "learning_rate": 5.28937886699854e-06, + "loss": 0.7143, + "step": 20138 + }, + { + "epoch": 3.5854700854700856, + "grad_norm": 1.1039072275161743, + "learning_rate": 5.284887722190501e-06, + "loss": 0.9603, + "step": 20139 + }, + { + "epoch": 3.585648148148148, + "grad_norm": 0.8831756711006165, + "learning_rate": 5.280398433128342e-06, + "loss": 0.6444, + "step": 20140 + }, + { + "epoch": 3.5858262108262107, + "grad_norm": 0.8430986404418945, + "learning_rate": 5.2759109999000025e-06, + "loss": 0.6404, + "step": 20141 + }, + { + "epoch": 3.5860042735042734, + "grad_norm": 0.9114181399345398, + "learning_rate": 5.271425422593435e-06, + "loss": 0.7964, + "step": 20142 + }, + { + "epoch": 3.586182336182336, + "grad_norm": 1.0493398904800415, + "learning_rate": 5.266941701296479e-06, + "loss": 0.8568, + "step": 20143 + }, + { + "epoch": 3.586360398860399, + "grad_norm": 0.8649964332580566, + "learning_rate": 5.262459836097033e-06, + "loss": 0.6505, + "step": 20144 + }, + { + "epoch": 3.5865384615384617, + "grad_norm": 0.8710708618164062, + "learning_rate": 5.257979827082859e-06, + "loss": 0.7117, + "step": 20145 + }, + { + "epoch": 3.5867165242165244, + "grad_norm": 1.1388016939163208, + "learning_rate": 5.253501674341777e-06, + "loss": 0.9579, + "step": 20146 + }, + { + "epoch": 3.5868945868945867, + "grad_norm": 1.029822587966919, + "learning_rate": 5.249025377961514e-06, + "loss": 0.7996, + "step": 20147 + }, + { + "epoch": 3.5870726495726495, + "grad_norm": 0.9227847456932068, + "learning_rate": 5.244550938029768e-06, + "loss": 0.902, + "step": 20148 + }, + { + "epoch": 3.5872507122507122, + "grad_norm": 0.9261413216590881, + "learning_rate": 5.240078354634214e-06, + "loss": 0.7765, + "step": 20149 + }, + { + "epoch": 3.587428774928775, + "grad_norm": 0.9532697200775146, + "learning_rate": 5.235607627862471e-06, + "loss": 0.8003, + "step": 20150 + }, + { + "epoch": 3.5876068376068377, + "grad_norm": 0.9802052974700928, + "learning_rate": 5.231138757802134e-06, + "loss": 0.7542, + "step": 20151 + }, + { + "epoch": 3.5877849002849, + "grad_norm": 1.019360065460205, + "learning_rate": 5.22667174454079e-06, + "loss": 0.6551, + "step": 20152 + }, + { + "epoch": 3.587962962962963, + "grad_norm": 1.1361932754516602, + "learning_rate": 5.222206588165912e-06, + "loss": 0.6014, + "step": 20153 + }, + { + "epoch": 3.5881410256410255, + "grad_norm": 1.0577775239944458, + "learning_rate": 5.21774328876502e-06, + "loss": 0.7311, + "step": 20154 + }, + { + "epoch": 3.5883190883190883, + "grad_norm": 0.9792344570159912, + "learning_rate": 5.2132818464255664e-06, + "loss": 0.8125, + "step": 20155 + }, + { + "epoch": 3.588497150997151, + "grad_norm": 1.0200873613357544, + "learning_rate": 5.208822261234925e-06, + "loss": 0.8364, + "step": 20156 + }, + { + "epoch": 3.588675213675214, + "grad_norm": 0.8380966782569885, + "learning_rate": 5.204364533280537e-06, + "loss": 0.6492, + "step": 20157 + }, + { + "epoch": 3.5888532763532766, + "grad_norm": 1.0056066513061523, + "learning_rate": 5.199908662649677e-06, + "loss": 0.6339, + "step": 20158 + }, + { + "epoch": 3.5890313390313393, + "grad_norm": 1.0888465642929077, + "learning_rate": 5.19545464942971e-06, + "loss": 0.7766, + "step": 20159 + }, + { + "epoch": 3.5892094017094016, + "grad_norm": 0.9547351598739624, + "learning_rate": 5.19100249370783e-06, + "loss": 0.8142, + "step": 20160 + }, + { + "epoch": 3.5893874643874644, + "grad_norm": 1.0138623714447021, + "learning_rate": 5.186552195571326e-06, + "loss": 0.9173, + "step": 20161 + }, + { + "epoch": 3.589565527065527, + "grad_norm": 1.025498628616333, + "learning_rate": 5.18210375510737e-06, + "loss": 0.5887, + "step": 20162 + }, + { + "epoch": 3.58974358974359, + "grad_norm": 0.8190540075302124, + "learning_rate": 5.177657172403139e-06, + "loss": 0.6902, + "step": 20163 + }, + { + "epoch": 3.589921652421652, + "grad_norm": 1.0583090782165527, + "learning_rate": 5.173212447545728e-06, + "loss": 0.7482, + "step": 20164 + }, + { + "epoch": 3.590099715099715, + "grad_norm": 0.9414075613021851, + "learning_rate": 5.168769580622234e-06, + "loss": 0.7471, + "step": 20165 + }, + { + "epoch": 3.5902777777777777, + "grad_norm": 0.9538351893424988, + "learning_rate": 5.16432857171969e-06, + "loss": 0.6802, + "step": 20166 + }, + { + "epoch": 3.5904558404558404, + "grad_norm": 0.9829911589622498, + "learning_rate": 5.159889420925146e-06, + "loss": 0.7222, + "step": 20167 + }, + { + "epoch": 3.590633903133903, + "grad_norm": 1.1673390865325928, + "learning_rate": 5.155452128325544e-06, + "loss": 0.7527, + "step": 20168 + }, + { + "epoch": 3.590811965811966, + "grad_norm": 1.0132838487625122, + "learning_rate": 5.151016694007849e-06, + "loss": 0.7148, + "step": 20169 + }, + { + "epoch": 3.5909900284900287, + "grad_norm": 0.9084600210189819, + "learning_rate": 5.146583118058923e-06, + "loss": 0.6949, + "step": 20170 + }, + { + "epoch": 3.5911680911680914, + "grad_norm": 0.9238780736923218, + "learning_rate": 5.142151400565687e-06, + "loss": 0.7453, + "step": 20171 + }, + { + "epoch": 3.5913461538461537, + "grad_norm": 0.9433969259262085, + "learning_rate": 5.137721541614926e-06, + "loss": 0.7395, + "step": 20172 + }, + { + "epoch": 3.5915242165242165, + "grad_norm": 0.8848153948783875, + "learning_rate": 5.133293541293449e-06, + "loss": 0.7717, + "step": 20173 + }, + { + "epoch": 3.5917022792022792, + "grad_norm": 0.9557967782020569, + "learning_rate": 5.1288673996880425e-06, + "loss": 0.8706, + "step": 20174 + }, + { + "epoch": 3.591880341880342, + "grad_norm": 0.9979602694511414, + "learning_rate": 5.12444311688538e-06, + "loss": 0.6658, + "step": 20175 + }, + { + "epoch": 3.5920584045584043, + "grad_norm": 0.9105966687202454, + "learning_rate": 5.120020692972172e-06, + "loss": 0.6382, + "step": 20176 + }, + { + "epoch": 3.592236467236467, + "grad_norm": 0.8652068376541138, + "learning_rate": 5.11560012803507e-06, + "loss": 0.6003, + "step": 20177 + }, + { + "epoch": 3.59241452991453, + "grad_norm": 0.9879941940307617, + "learning_rate": 5.111181422160671e-06, + "loss": 0.7795, + "step": 20178 + }, + { + "epoch": 3.5925925925925926, + "grad_norm": 1.225643515586853, + "learning_rate": 5.106764575435563e-06, + "loss": 0.8514, + "step": 20179 + }, + { + "epoch": 3.5927706552706553, + "grad_norm": 0.9002955555915833, + "learning_rate": 5.102349587946275e-06, + "loss": 1.0394, + "step": 20180 + }, + { + "epoch": 3.592948717948718, + "grad_norm": 0.8426976203918457, + "learning_rate": 5.097936459779318e-06, + "loss": 0.7507, + "step": 20181 + }, + { + "epoch": 3.593126780626781, + "grad_norm": 0.9469396471977234, + "learning_rate": 5.093525191021154e-06, + "loss": 0.7158, + "step": 20182 + }, + { + "epoch": 3.5933048433048436, + "grad_norm": 0.8771377801895142, + "learning_rate": 5.0891157817581916e-06, + "loss": 0.6222, + "step": 20183 + }, + { + "epoch": 3.593482905982906, + "grad_norm": 1.0022752285003662, + "learning_rate": 5.084708232076873e-06, + "loss": 0.7088, + "step": 20184 + }, + { + "epoch": 3.5936609686609686, + "grad_norm": 0.9102463722229004, + "learning_rate": 5.080302542063508e-06, + "loss": 0.7738, + "step": 20185 + }, + { + "epoch": 3.5938390313390314, + "grad_norm": 1.1719863414764404, + "learning_rate": 5.07589871180445e-06, + "loss": 0.8472, + "step": 20186 + }, + { + "epoch": 3.594017094017094, + "grad_norm": 0.9453331828117371, + "learning_rate": 5.071496741385961e-06, + "loss": 0.8113, + "step": 20187 + }, + { + "epoch": 3.5941951566951564, + "grad_norm": 0.9310576915740967, + "learning_rate": 5.0670966308943076e-06, + "loss": 0.7145, + "step": 20188 + }, + { + "epoch": 3.594373219373219, + "grad_norm": 0.933391809463501, + "learning_rate": 5.062698380415687e-06, + "loss": 0.8318, + "step": 20189 + }, + { + "epoch": 3.594551282051282, + "grad_norm": 1.0447312593460083, + "learning_rate": 5.0583019900362625e-06, + "loss": 0.7597, + "step": 20190 + }, + { + "epoch": 3.5947293447293447, + "grad_norm": 0.8405809998512268, + "learning_rate": 5.053907459842189e-06, + "loss": 0.5576, + "step": 20191 + }, + { + "epoch": 3.5949074074074074, + "grad_norm": 0.8650723099708557, + "learning_rate": 5.0495147899195736e-06, + "loss": 0.589, + "step": 20192 + }, + { + "epoch": 3.59508547008547, + "grad_norm": 0.8867887854576111, + "learning_rate": 5.045123980354472e-06, + "loss": 0.7514, + "step": 20193 + }, + { + "epoch": 3.595263532763533, + "grad_norm": 0.9023543000221252, + "learning_rate": 5.0407350312329125e-06, + "loss": 0.6218, + "step": 20194 + }, + { + "epoch": 3.5954415954415957, + "grad_norm": 0.9571815729141235, + "learning_rate": 5.036347942640896e-06, + "loss": 0.7849, + "step": 20195 + }, + { + "epoch": 3.595619658119658, + "grad_norm": 0.9155046343803406, + "learning_rate": 5.031962714664362e-06, + "loss": 0.6944, + "step": 20196 + }, + { + "epoch": 3.5957977207977208, + "grad_norm": 0.9272841215133667, + "learning_rate": 5.027579347389244e-06, + "loss": 0.7431, + "step": 20197 + }, + { + "epoch": 3.5959757834757835, + "grad_norm": 1.0354583263397217, + "learning_rate": 5.023197840901417e-06, + "loss": 0.9811, + "step": 20198 + }, + { + "epoch": 3.5961538461538463, + "grad_norm": 1.0272680521011353, + "learning_rate": 5.018818195286745e-06, + "loss": 0.7009, + "step": 20199 + }, + { + "epoch": 3.5963319088319086, + "grad_norm": 0.9123852252960205, + "learning_rate": 5.014440410631005e-06, + "loss": 0.851, + "step": 20200 + }, + { + "epoch": 3.5965099715099713, + "grad_norm": 1.0875011682510376, + "learning_rate": 5.010064487019994e-06, + "loss": 0.587, + "step": 20201 + }, + { + "epoch": 3.596688034188034, + "grad_norm": 0.9781659841537476, + "learning_rate": 5.005690424539455e-06, + "loss": 0.7736, + "step": 20202 + }, + { + "epoch": 3.596866096866097, + "grad_norm": 0.9043481945991516, + "learning_rate": 5.001318223275075e-06, + "loss": 0.7778, + "step": 20203 + }, + { + "epoch": 3.5970441595441596, + "grad_norm": 0.9980285167694092, + "learning_rate": 4.996947883312519e-06, + "loss": 0.8251, + "step": 20204 + }, + { + "epoch": 3.5972222222222223, + "grad_norm": 1.058321475982666, + "learning_rate": 4.992579404737419e-06, + "loss": 0.7323, + "step": 20205 + }, + { + "epoch": 3.597400284900285, + "grad_norm": 1.0859766006469727, + "learning_rate": 4.98821278763536e-06, + "loss": 0.9372, + "step": 20206 + }, + { + "epoch": 3.597578347578348, + "grad_norm": 0.9477382302284241, + "learning_rate": 4.983848032091909e-06, + "loss": 0.7893, + "step": 20207 + }, + { + "epoch": 3.59775641025641, + "grad_norm": 0.9508783221244812, + "learning_rate": 4.979485138192563e-06, + "loss": 0.7689, + "step": 20208 + }, + { + "epoch": 3.597934472934473, + "grad_norm": 1.0118316411972046, + "learning_rate": 4.975124106022844e-06, + "loss": 0.8543, + "step": 20209 + }, + { + "epoch": 3.5981125356125356, + "grad_norm": 0.8835347890853882, + "learning_rate": 4.970764935668137e-06, + "loss": 0.595, + "step": 20210 + }, + { + "epoch": 3.5982905982905984, + "grad_norm": 0.857542872428894, + "learning_rate": 4.966407627213909e-06, + "loss": 0.6382, + "step": 20211 + }, + { + "epoch": 3.5984686609686607, + "grad_norm": 1.1189042329788208, + "learning_rate": 4.9620521807455026e-06, + "loss": 0.7847, + "step": 20212 + }, + { + "epoch": 3.5986467236467234, + "grad_norm": 0.9164559841156006, + "learning_rate": 4.957698596348248e-06, + "loss": 0.784, + "step": 20213 + }, + { + "epoch": 3.598824786324786, + "grad_norm": 1.0209029912948608, + "learning_rate": 4.953346874107478e-06, + "loss": 0.9315, + "step": 20214 + }, + { + "epoch": 3.599002849002849, + "grad_norm": 0.8468409776687622, + "learning_rate": 4.948997014108414e-06, + "loss": 0.7392, + "step": 20215 + }, + { + "epoch": 3.5991809116809117, + "grad_norm": 0.8785325884819031, + "learning_rate": 4.944649016436331e-06, + "loss": 0.7619, + "step": 20216 + }, + { + "epoch": 3.5993589743589745, + "grad_norm": 0.9535296559333801, + "learning_rate": 4.940302881176351e-06, + "loss": 0.7385, + "step": 20217 + }, + { + "epoch": 3.599537037037037, + "grad_norm": 1.067630410194397, + "learning_rate": 4.935958608413693e-06, + "loss": 0.7811, + "step": 20218 + }, + { + "epoch": 3.5997150997151, + "grad_norm": 0.9168570041656494, + "learning_rate": 4.9316161982334355e-06, + "loss": 0.7221, + "step": 20219 + }, + { + "epoch": 3.5998931623931623, + "grad_norm": 0.9218630790710449, + "learning_rate": 4.927275650720686e-06, + "loss": 0.7633, + "step": 20220 + }, + { + "epoch": 3.600071225071225, + "grad_norm": 0.9325320720672607, + "learning_rate": 4.922936965960457e-06, + "loss": 0.9047, + "step": 20221 + }, + { + "epoch": 3.6002492877492878, + "grad_norm": 1.031030535697937, + "learning_rate": 4.918600144037788e-06, + "loss": 0.6517, + "step": 20222 + }, + { + "epoch": 3.6004273504273505, + "grad_norm": 0.9421409368515015, + "learning_rate": 4.914265185037614e-06, + "loss": 0.565, + "step": 20223 + }, + { + "epoch": 3.6006054131054133, + "grad_norm": 0.9039212465286255, + "learning_rate": 4.9099320890449106e-06, + "loss": 0.8056, + "step": 20224 + }, + { + "epoch": 3.6007834757834756, + "grad_norm": 1.0419108867645264, + "learning_rate": 4.9056008561445324e-06, + "loss": 0.8205, + "step": 20225 + }, + { + "epoch": 3.6009615384615383, + "grad_norm": 0.9815367460250854, + "learning_rate": 4.901271486421388e-06, + "loss": 0.6823, + "step": 20226 + }, + { + "epoch": 3.601139601139601, + "grad_norm": 0.951187014579773, + "learning_rate": 4.896943979960267e-06, + "loss": 0.7703, + "step": 20227 + }, + { + "epoch": 3.601317663817664, + "grad_norm": 0.9774547219276428, + "learning_rate": 4.8926183368459775e-06, + "loss": 0.572, + "step": 20228 + }, + { + "epoch": 3.6014957264957266, + "grad_norm": 0.9099259376525879, + "learning_rate": 4.888294557163254e-06, + "loss": 0.7786, + "step": 20229 + }, + { + "epoch": 3.6016737891737893, + "grad_norm": 0.9812109470367432, + "learning_rate": 4.883972640996826e-06, + "loss": 0.7925, + "step": 20230 + }, + { + "epoch": 3.601851851851852, + "grad_norm": 0.8465882539749146, + "learning_rate": 4.8796525884313604e-06, + "loss": 0.6353, + "step": 20231 + }, + { + "epoch": 3.6020299145299144, + "grad_norm": 1.130205512046814, + "learning_rate": 4.875334399551501e-06, + "loss": 0.7415, + "step": 20232 + }, + { + "epoch": 3.602207977207977, + "grad_norm": 0.8592739105224609, + "learning_rate": 4.87101807444188e-06, + "loss": 0.5869, + "step": 20233 + }, + { + "epoch": 3.60238603988604, + "grad_norm": 0.7758101224899292, + "learning_rate": 4.86670361318704e-06, + "loss": 0.6109, + "step": 20234 + }, + { + "epoch": 3.6025641025641026, + "grad_norm": 1.0534182786941528, + "learning_rate": 4.862391015871515e-06, + "loss": 0.5944, + "step": 20235 + }, + { + "epoch": 3.6027421652421654, + "grad_norm": 0.9213957786560059, + "learning_rate": 4.858080282579813e-06, + "loss": 0.6962, + "step": 20236 + }, + { + "epoch": 3.6029202279202277, + "grad_norm": 1.1342556476593018, + "learning_rate": 4.85377141339638e-06, + "loss": 0.7288, + "step": 20237 + }, + { + "epoch": 3.6030982905982905, + "grad_norm": 0.8825157284736633, + "learning_rate": 4.849464408405646e-06, + "loss": 0.6383, + "step": 20238 + }, + { + "epoch": 3.603276353276353, + "grad_norm": 1.0171782970428467, + "learning_rate": 4.845159267692012e-06, + "loss": 0.7471, + "step": 20239 + }, + { + "epoch": 3.603454415954416, + "grad_norm": 1.1089091300964355, + "learning_rate": 4.840855991339799e-06, + "loss": 0.9887, + "step": 20240 + }, + { + "epoch": 3.6036324786324787, + "grad_norm": 1.0146455764770508, + "learning_rate": 4.836554579433361e-06, + "loss": 0.7713, + "step": 20241 + }, + { + "epoch": 3.6038105413105415, + "grad_norm": 0.895252525806427, + "learning_rate": 4.8322550320569204e-06, + "loss": 0.5434, + "step": 20242 + }, + { + "epoch": 3.603988603988604, + "grad_norm": 0.9839693903923035, + "learning_rate": 4.827957349294765e-06, + "loss": 0.7696, + "step": 20243 + }, + { + "epoch": 3.6041666666666665, + "grad_norm": 0.9720914363861084, + "learning_rate": 4.823661531231083e-06, + "loss": 0.7884, + "step": 20244 + }, + { + "epoch": 3.6043447293447293, + "grad_norm": 0.9014876484870911, + "learning_rate": 4.81936757795004e-06, + "loss": 0.8938, + "step": 20245 + }, + { + "epoch": 3.604522792022792, + "grad_norm": 0.9114326238632202, + "learning_rate": 4.81507548953577e-06, + "loss": 0.8245, + "step": 20246 + }, + { + "epoch": 3.6047008547008548, + "grad_norm": 1.0024261474609375, + "learning_rate": 4.81078526607236e-06, + "loss": 0.7461, + "step": 20247 + }, + { + "epoch": 3.6048789173789175, + "grad_norm": 0.9544615149497986, + "learning_rate": 4.806496907643876e-06, + "loss": 0.7481, + "step": 20248 + }, + { + "epoch": 3.60505698005698, + "grad_norm": 0.8342941999435425, + "learning_rate": 4.802210414334351e-06, + "loss": 0.5793, + "step": 20249 + }, + { + "epoch": 3.6052350427350426, + "grad_norm": 0.9724514484405518, + "learning_rate": 4.79792578622773e-06, + "loss": 0.8259, + "step": 20250 + }, + { + "epoch": 3.6054131054131053, + "grad_norm": 1.0200597047805786, + "learning_rate": 4.7936430234080234e-06, + "loss": 0.8159, + "step": 20251 + }, + { + "epoch": 3.605591168091168, + "grad_norm": 0.982551097869873, + "learning_rate": 4.789362125959073e-06, + "loss": 0.8049, + "step": 20252 + }, + { + "epoch": 3.605769230769231, + "grad_norm": 0.9515790343284607, + "learning_rate": 4.785083093964804e-06, + "loss": 0.721, + "step": 20253 + }, + { + "epoch": 3.6059472934472936, + "grad_norm": 0.9414631724357605, + "learning_rate": 4.7808059275090465e-06, + "loss": 0.8164, + "step": 20254 + }, + { + "epoch": 3.6061253561253563, + "grad_norm": 0.7917568683624268, + "learning_rate": 4.77653062667558e-06, + "loss": 0.6002, + "step": 20255 + }, + { + "epoch": 3.6063034188034186, + "grad_norm": 0.9543023705482483, + "learning_rate": 4.772257191548202e-06, + "loss": 0.7494, + "step": 20256 + }, + { + "epoch": 3.6064814814814814, + "grad_norm": 0.908807635307312, + "learning_rate": 4.7679856222106135e-06, + "loss": 0.7106, + "step": 20257 + }, + { + "epoch": 3.606659544159544, + "grad_norm": 1.0815712213516235, + "learning_rate": 4.763715918746525e-06, + "loss": 0.7924, + "step": 20258 + }, + { + "epoch": 3.606837606837607, + "grad_norm": 0.8257647156715393, + "learning_rate": 4.7594480812395925e-06, + "loss": 0.7398, + "step": 20259 + }, + { + "epoch": 3.6070156695156697, + "grad_norm": 0.8509652614593506, + "learning_rate": 4.755182109773426e-06, + "loss": 0.8162, + "step": 20260 + }, + { + "epoch": 3.607193732193732, + "grad_norm": 0.9860865473747253, + "learning_rate": 4.750918004431604e-06, + "loss": 0.974, + "step": 20261 + }, + { + "epoch": 3.6073717948717947, + "grad_norm": 0.9228841066360474, + "learning_rate": 4.746655765297692e-06, + "loss": 0.8106, + "step": 20262 + }, + { + "epoch": 3.6075498575498575, + "grad_norm": 1.1330608129501343, + "learning_rate": 4.7423953924551675e-06, + "loss": 0.6528, + "step": 20263 + }, + { + "epoch": 3.60772792022792, + "grad_norm": 1.0182117223739624, + "learning_rate": 4.738136885987565e-06, + "loss": 0.7582, + "step": 20264 + }, + { + "epoch": 3.607905982905983, + "grad_norm": 1.006861925125122, + "learning_rate": 4.733880245978239e-06, + "loss": 0.6213, + "step": 20265 + }, + { + "epoch": 3.6080840455840457, + "grad_norm": 1.037815809249878, + "learning_rate": 4.729625472510668e-06, + "loss": 0.9137, + "step": 20266 + }, + { + "epoch": 3.6082621082621085, + "grad_norm": 1.1505076885223389, + "learning_rate": 4.7253725656681515e-06, + "loss": 0.8987, + "step": 20267 + }, + { + "epoch": 3.6084401709401708, + "grad_norm": 0.9972392320632935, + "learning_rate": 4.721121525534045e-06, + "loss": 0.6091, + "step": 20268 + }, + { + "epoch": 3.6086182336182335, + "grad_norm": 0.8960257768630981, + "learning_rate": 4.716872352191648e-06, + "loss": 0.6685, + "step": 20269 + }, + { + "epoch": 3.6087962962962963, + "grad_norm": 0.9827064275741577, + "learning_rate": 4.712625045724206e-06, + "loss": 0.8017, + "step": 20270 + }, + { + "epoch": 3.608974358974359, + "grad_norm": 0.9110612273216248, + "learning_rate": 4.70837960621493e-06, + "loss": 0.6586, + "step": 20271 + }, + { + "epoch": 3.609152421652422, + "grad_norm": 0.9796404838562012, + "learning_rate": 4.704136033746987e-06, + "loss": 0.7285, + "step": 20272 + }, + { + "epoch": 3.609330484330484, + "grad_norm": 1.166001796722412, + "learning_rate": 4.699894328403554e-06, + "loss": 0.8324, + "step": 20273 + }, + { + "epoch": 3.609508547008547, + "grad_norm": 0.8968726992607117, + "learning_rate": 4.6956544902677315e-06, + "loss": 0.6518, + "step": 20274 + }, + { + "epoch": 3.6096866096866096, + "grad_norm": 0.9205286502838135, + "learning_rate": 4.691416519422575e-06, + "loss": 0.8745, + "step": 20275 + }, + { + "epoch": 3.6098646723646723, + "grad_norm": 0.9699488282203674, + "learning_rate": 4.68718041595112e-06, + "loss": 0.7312, + "step": 20276 + }, + { + "epoch": 3.610042735042735, + "grad_norm": 1.0142552852630615, + "learning_rate": 4.682946179936376e-06, + "loss": 0.6849, + "step": 20277 + }, + { + "epoch": 3.610220797720798, + "grad_norm": 1.0610849857330322, + "learning_rate": 4.678713811461299e-06, + "loss": 0.6934, + "step": 20278 + }, + { + "epoch": 3.6103988603988606, + "grad_norm": 0.9037832617759705, + "learning_rate": 4.674483310608801e-06, + "loss": 0.6833, + "step": 20279 + }, + { + "epoch": 3.6105769230769234, + "grad_norm": 0.998599112033844, + "learning_rate": 4.670254677461783e-06, + "loss": 0.643, + "step": 20280 + }, + { + "epoch": 3.6107549857549857, + "grad_norm": 0.8829896450042725, + "learning_rate": 4.6660279121031105e-06, + "loss": 0.6309, + "step": 20281 + }, + { + "epoch": 3.6109330484330484, + "grad_norm": 0.9085970520973206, + "learning_rate": 4.661803014615562e-06, + "loss": 0.6999, + "step": 20282 + }, + { + "epoch": 3.611111111111111, + "grad_norm": 1.0698771476745605, + "learning_rate": 4.657579985081939e-06, + "loss": 0.6356, + "step": 20283 + }, + { + "epoch": 3.611289173789174, + "grad_norm": 1.119136095046997, + "learning_rate": 4.653358823584986e-06, + "loss": 0.7683, + "step": 20284 + }, + { + "epoch": 3.611467236467236, + "grad_norm": 0.9300078749656677, + "learning_rate": 4.649139530207403e-06, + "loss": 0.7022, + "step": 20285 + }, + { + "epoch": 3.611645299145299, + "grad_norm": 0.9168888926506042, + "learning_rate": 4.644922105031857e-06, + "loss": 0.8505, + "step": 20286 + }, + { + "epoch": 3.6118233618233617, + "grad_norm": 0.9180569648742676, + "learning_rate": 4.640706548140972e-06, + "loss": 0.783, + "step": 20287 + }, + { + "epoch": 3.6120014245014245, + "grad_norm": 0.944037914276123, + "learning_rate": 4.636492859617358e-06, + "loss": 0.6348, + "step": 20288 + }, + { + "epoch": 3.6121794871794872, + "grad_norm": 1.140091061592102, + "learning_rate": 4.632281039543562e-06, + "loss": 0.7691, + "step": 20289 + }, + { + "epoch": 3.61235754985755, + "grad_norm": 0.9035637974739075, + "learning_rate": 4.6280710880021065e-06, + "loss": 0.4777, + "step": 20290 + }, + { + "epoch": 3.6125356125356127, + "grad_norm": 0.9371036887168884, + "learning_rate": 4.62386300507549e-06, + "loss": 0.6795, + "step": 20291 + }, + { + "epoch": 3.6127136752136755, + "grad_norm": 0.8894741535186768, + "learning_rate": 4.619656790846139e-06, + "loss": 0.6351, + "step": 20292 + }, + { + "epoch": 3.612891737891738, + "grad_norm": 1.035195231437683, + "learning_rate": 4.6154524453964846e-06, + "loss": 0.8925, + "step": 20293 + }, + { + "epoch": 3.6130698005698005, + "grad_norm": 0.7878008484840393, + "learning_rate": 4.611249968808895e-06, + "loss": 0.4768, + "step": 20294 + }, + { + "epoch": 3.6132478632478633, + "grad_norm": 0.9613704085350037, + "learning_rate": 4.607049361165694e-06, + "loss": 0.6607, + "step": 20295 + }, + { + "epoch": 3.613425925925926, + "grad_norm": 0.9866206645965576, + "learning_rate": 4.602850622549227e-06, + "loss": 0.8279, + "step": 20296 + }, + { + "epoch": 3.6136039886039883, + "grad_norm": 1.0364552736282349, + "learning_rate": 4.598653753041704e-06, + "loss": 0.7251, + "step": 20297 + }, + { + "epoch": 3.613782051282051, + "grad_norm": 0.7488967776298523, + "learning_rate": 4.594458752725395e-06, + "loss": 0.675, + "step": 20298 + }, + { + "epoch": 3.613960113960114, + "grad_norm": 1.0590063333511353, + "learning_rate": 4.590265621682488e-06, + "loss": 0.8801, + "step": 20299 + }, + { + "epoch": 3.6141381766381766, + "grad_norm": 1.012354850769043, + "learning_rate": 4.586074359995119e-06, + "loss": 0.8227, + "step": 20300 + }, + { + "epoch": 3.6143162393162394, + "grad_norm": 0.8928263187408447, + "learning_rate": 4.58188496774542e-06, + "loss": 0.5889, + "step": 20301 + }, + { + "epoch": 3.614494301994302, + "grad_norm": 0.9703002572059631, + "learning_rate": 4.577697445015472e-06, + "loss": 0.6322, + "step": 20302 + }, + { + "epoch": 3.614672364672365, + "grad_norm": 0.8588637709617615, + "learning_rate": 4.573511791887319e-06, + "loss": 0.6494, + "step": 20303 + }, + { + "epoch": 3.6148504273504276, + "grad_norm": 0.9801368117332458, + "learning_rate": 4.569328008442975e-06, + "loss": 0.8244, + "step": 20304 + }, + { + "epoch": 3.61502849002849, + "grad_norm": 1.0798895359039307, + "learning_rate": 4.565146094764394e-06, + "loss": 0.6987, + "step": 20305 + }, + { + "epoch": 3.6152065527065527, + "grad_norm": 0.9438313841819763, + "learning_rate": 4.560966050933546e-06, + "loss": 0.507, + "step": 20306 + }, + { + "epoch": 3.6153846153846154, + "grad_norm": 0.9473024606704712, + "learning_rate": 4.556787877032287e-06, + "loss": 0.7889, + "step": 20307 + }, + { + "epoch": 3.615562678062678, + "grad_norm": 0.9516319632530212, + "learning_rate": 4.552611573142518e-06, + "loss": 0.6727, + "step": 20308 + }, + { + "epoch": 3.6157407407407405, + "grad_norm": 0.9949623942375183, + "learning_rate": 4.548437139346052e-06, + "loss": 0.9616, + "step": 20309 + }, + { + "epoch": 3.6159188034188032, + "grad_norm": 0.9285261631011963, + "learning_rate": 4.544264575724677e-06, + "loss": 0.819, + "step": 20310 + }, + { + "epoch": 3.616096866096866, + "grad_norm": 0.9633162021636963, + "learning_rate": 4.540093882360152e-06, + "loss": 0.7936, + "step": 20311 + }, + { + "epoch": 3.6162749287749287, + "grad_norm": 0.9205031991004944, + "learning_rate": 4.535925059334168e-06, + "loss": 0.6606, + "step": 20312 + }, + { + "epoch": 3.6164529914529915, + "grad_norm": 0.9210516810417175, + "learning_rate": 4.531758106728446e-06, + "loss": 0.8, + "step": 20313 + }, + { + "epoch": 3.6166310541310542, + "grad_norm": 1.0242942571640015, + "learning_rate": 4.5275930246245764e-06, + "loss": 0.6824, + "step": 20314 + }, + { + "epoch": 3.616809116809117, + "grad_norm": 1.2146587371826172, + "learning_rate": 4.523429813104218e-06, + "loss": 0.7973, + "step": 20315 + }, + { + "epoch": 3.6169871794871797, + "grad_norm": 0.9127520322799683, + "learning_rate": 4.519268472248906e-06, + "loss": 0.7579, + "step": 20316 + }, + { + "epoch": 3.617165242165242, + "grad_norm": 0.9486130475997925, + "learning_rate": 4.5151090021401946e-06, + "loss": 0.6724, + "step": 20317 + }, + { + "epoch": 3.617343304843305, + "grad_norm": 1.0272002220153809, + "learning_rate": 4.510951402859564e-06, + "loss": 0.7484, + "step": 20318 + }, + { + "epoch": 3.6175213675213675, + "grad_norm": 1.0201951265335083, + "learning_rate": 4.506795674488484e-06, + "loss": 0.7737, + "step": 20319 + }, + { + "epoch": 3.6176994301994303, + "grad_norm": 1.0206116437911987, + "learning_rate": 4.502641817108355e-06, + "loss": 0.8504, + "step": 20320 + }, + { + "epoch": 3.6178774928774926, + "grad_norm": 0.9464601278305054, + "learning_rate": 4.498489830800601e-06, + "loss": 0.791, + "step": 20321 + }, + { + "epoch": 3.6180555555555554, + "grad_norm": 0.8579327464103699, + "learning_rate": 4.494339715646534e-06, + "loss": 0.5689, + "step": 20322 + }, + { + "epoch": 3.618233618233618, + "grad_norm": 0.9792792797088623, + "learning_rate": 4.490191471727511e-06, + "loss": 0.8564, + "step": 20323 + }, + { + "epoch": 3.618411680911681, + "grad_norm": 1.008038878440857, + "learning_rate": 4.486045099124758e-06, + "loss": 0.8837, + "step": 20324 + }, + { + "epoch": 3.6185897435897436, + "grad_norm": 0.8817142248153687, + "learning_rate": 4.481900597919541e-06, + "loss": 0.6801, + "step": 20325 + }, + { + "epoch": 3.6187678062678064, + "grad_norm": 1.1172369718551636, + "learning_rate": 4.477757968193075e-06, + "loss": 0.8062, + "step": 20326 + }, + { + "epoch": 3.618945868945869, + "grad_norm": 0.879314124584198, + "learning_rate": 4.473617210026504e-06, + "loss": 0.769, + "step": 20327 + }, + { + "epoch": 3.619123931623932, + "grad_norm": 1.1788002252578735, + "learning_rate": 4.469478323500953e-06, + "loss": 0.7673, + "step": 20328 + }, + { + "epoch": 3.619301994301994, + "grad_norm": 0.9070376753807068, + "learning_rate": 4.465341308697546e-06, + "loss": 0.8223, + "step": 20329 + }, + { + "epoch": 3.619480056980057, + "grad_norm": 1.0292595624923706, + "learning_rate": 4.461206165697296e-06, + "loss": 0.806, + "step": 20330 + }, + { + "epoch": 3.6196581196581197, + "grad_norm": 0.9035056233406067, + "learning_rate": 4.4570728945812605e-06, + "loss": 0.7187, + "step": 20331 + }, + { + "epoch": 3.6198361823361824, + "grad_norm": 1.1375666856765747, + "learning_rate": 4.4529414954304075e-06, + "loss": 0.723, + "step": 20332 + }, + { + "epoch": 3.620014245014245, + "grad_norm": 0.9073819518089294, + "learning_rate": 4.448811968325683e-06, + "loss": 0.7414, + "step": 20333 + }, + { + "epoch": 3.6201923076923075, + "grad_norm": 0.8783019185066223, + "learning_rate": 4.444684313348002e-06, + "loss": 0.7602, + "step": 20334 + }, + { + "epoch": 3.6203703703703702, + "grad_norm": 0.9157004356384277, + "learning_rate": 4.440558530578221e-06, + "loss": 0.771, + "step": 20335 + }, + { + "epoch": 3.620548433048433, + "grad_norm": 0.945231020450592, + "learning_rate": 4.4364346200972184e-06, + "loss": 0.9442, + "step": 20336 + }, + { + "epoch": 3.6207264957264957, + "grad_norm": 0.9638859629631042, + "learning_rate": 4.432312581985732e-06, + "loss": 0.73, + "step": 20337 + }, + { + "epoch": 3.6209045584045585, + "grad_norm": 1.0869982242584229, + "learning_rate": 4.428192416324573e-06, + "loss": 0.906, + "step": 20338 + }, + { + "epoch": 3.6210826210826212, + "grad_norm": 0.9156002998352051, + "learning_rate": 4.424074123194433e-06, + "loss": 0.7178, + "step": 20339 + }, + { + "epoch": 3.621260683760684, + "grad_norm": 1.0202544927597046, + "learning_rate": 4.419957702676037e-06, + "loss": 0.6788, + "step": 20340 + }, + { + "epoch": 3.6214387464387463, + "grad_norm": 1.0651593208312988, + "learning_rate": 4.4158431548500075e-06, + "loss": 0.624, + "step": 20341 + }, + { + "epoch": 3.621616809116809, + "grad_norm": 0.9189126491546631, + "learning_rate": 4.411730479796982e-06, + "loss": 0.69, + "step": 20342 + }, + { + "epoch": 3.621794871794872, + "grad_norm": 1.0193413496017456, + "learning_rate": 4.40761967759753e-06, + "loss": 0.8095, + "step": 20343 + }, + { + "epoch": 3.6219729344729346, + "grad_norm": 1.0479416847229004, + "learning_rate": 4.403510748332185e-06, + "loss": 0.7535, + "step": 20344 + }, + { + "epoch": 3.6221509971509973, + "grad_norm": 0.9200383424758911, + "learning_rate": 4.399403692081461e-06, + "loss": 0.7786, + "step": 20345 + }, + { + "epoch": 3.6223290598290596, + "grad_norm": 1.0074836015701294, + "learning_rate": 4.3952985089258495e-06, + "loss": 0.6789, + "step": 20346 + }, + { + "epoch": 3.6225071225071224, + "grad_norm": 0.9245620965957642, + "learning_rate": 4.39119519894573e-06, + "loss": 0.8554, + "step": 20347 + }, + { + "epoch": 3.622685185185185, + "grad_norm": 0.8818466663360596, + "learning_rate": 4.387093762221562e-06, + "loss": 0.838, + "step": 20348 + }, + { + "epoch": 3.622863247863248, + "grad_norm": 0.9582803845405579, + "learning_rate": 4.382994198833645e-06, + "loss": 0.8354, + "step": 20349 + }, + { + "epoch": 3.6230413105413106, + "grad_norm": 0.8472702503204346, + "learning_rate": 4.378896508862351e-06, + "loss": 0.6862, + "step": 20350 + }, + { + "epoch": 3.6232193732193734, + "grad_norm": 0.8588680028915405, + "learning_rate": 4.374800692387937e-06, + "loss": 0.5167, + "step": 20351 + }, + { + "epoch": 3.623397435897436, + "grad_norm": 0.9006664752960205, + "learning_rate": 4.37070674949065e-06, + "loss": 0.5662, + "step": 20352 + }, + { + "epoch": 3.6235754985754984, + "grad_norm": 1.099118709564209, + "learning_rate": 4.366614680250736e-06, + "loss": 0.7506, + "step": 20353 + }, + { + "epoch": 3.623753561253561, + "grad_norm": 0.8333063721656799, + "learning_rate": 4.362524484748309e-06, + "loss": 0.7858, + "step": 20354 + }, + { + "epoch": 3.623931623931624, + "grad_norm": 1.0544650554656982, + "learning_rate": 4.358436163063573e-06, + "loss": 1.0375, + "step": 20355 + }, + { + "epoch": 3.6241096866096867, + "grad_norm": 1.008267879486084, + "learning_rate": 4.354349715276595e-06, + "loss": 0.659, + "step": 20356 + }, + { + "epoch": 3.6242877492877494, + "grad_norm": 0.8872396349906921, + "learning_rate": 4.350265141467458e-06, + "loss": 0.8278, + "step": 20357 + }, + { + "epoch": 3.6244658119658117, + "grad_norm": 0.9096971750259399, + "learning_rate": 4.346182441716162e-06, + "loss": 0.7571, + "step": 20358 + }, + { + "epoch": 3.6246438746438745, + "grad_norm": 0.9562456011772156, + "learning_rate": 4.342101616102734e-06, + "loss": 0.7768, + "step": 20359 + }, + { + "epoch": 3.6248219373219372, + "grad_norm": 0.8731200695037842, + "learning_rate": 4.338022664707109e-06, + "loss": 0.5639, + "step": 20360 + }, + { + "epoch": 3.625, + "grad_norm": 0.9827439785003662, + "learning_rate": 4.333945587609201e-06, + "loss": 0.8766, + "step": 20361 + }, + { + "epoch": 3.6251780626780628, + "grad_norm": 0.7939251661300659, + "learning_rate": 4.329870384888901e-06, + "loss": 0.4391, + "step": 20362 + }, + { + "epoch": 3.6253561253561255, + "grad_norm": 0.9462933540344238, + "learning_rate": 4.325797056626069e-06, + "loss": 0.5996, + "step": 20363 + }, + { + "epoch": 3.6255341880341883, + "grad_norm": 1.0042550563812256, + "learning_rate": 4.321725602900473e-06, + "loss": 0.7451, + "step": 20364 + }, + { + "epoch": 3.6257122507122506, + "grad_norm": 0.914974570274353, + "learning_rate": 4.317656023791927e-06, + "loss": 0.7685, + "step": 20365 + }, + { + "epoch": 3.6258903133903133, + "grad_norm": 0.85157310962677, + "learning_rate": 4.313588319380146e-06, + "loss": 0.6611, + "step": 20366 + }, + { + "epoch": 3.626068376068376, + "grad_norm": 1.184873104095459, + "learning_rate": 4.309522489744822e-06, + "loss": 0.8994, + "step": 20367 + }, + { + "epoch": 3.626246438746439, + "grad_norm": 1.1252131462097168, + "learning_rate": 4.305458534965634e-06, + "loss": 0.9815, + "step": 20368 + }, + { + "epoch": 3.6264245014245016, + "grad_norm": 0.9970200657844543, + "learning_rate": 4.301396455122198e-06, + "loss": 0.6608, + "step": 20369 + }, + { + "epoch": 3.626602564102564, + "grad_norm": 1.041490912437439, + "learning_rate": 4.297336250294093e-06, + "loss": 0.8741, + "step": 20370 + }, + { + "epoch": 3.6267806267806266, + "grad_norm": 1.0515310764312744, + "learning_rate": 4.2932779205608785e-06, + "loss": 0.8035, + "step": 20371 + }, + { + "epoch": 3.6269586894586894, + "grad_norm": 1.1585599184036255, + "learning_rate": 4.28922146600208e-06, + "loss": 0.9048, + "step": 20372 + }, + { + "epoch": 3.627136752136752, + "grad_norm": 0.9767042398452759, + "learning_rate": 4.285166886697167e-06, + "loss": 0.7881, + "step": 20373 + }, + { + "epoch": 3.627314814814815, + "grad_norm": 0.9792457818984985, + "learning_rate": 4.2811141827255764e-06, + "loss": 0.7102, + "step": 20374 + }, + { + "epoch": 3.6274928774928776, + "grad_norm": 0.9069845676422119, + "learning_rate": 4.27706335416671e-06, + "loss": 0.7514, + "step": 20375 + }, + { + "epoch": 3.6276709401709404, + "grad_norm": 0.8710619807243347, + "learning_rate": 4.273014401099939e-06, + "loss": 0.767, + "step": 20376 + }, + { + "epoch": 3.6278490028490027, + "grad_norm": 1.1712818145751953, + "learning_rate": 4.268967323604589e-06, + "loss": 0.5082, + "step": 20377 + }, + { + "epoch": 3.6280270655270654, + "grad_norm": 1.053260087966919, + "learning_rate": 4.264922121759985e-06, + "loss": 0.7219, + "step": 20378 + }, + { + "epoch": 3.628205128205128, + "grad_norm": 0.9651164412498474, + "learning_rate": 4.2608787956453286e-06, + "loss": 0.6483, + "step": 20379 + }, + { + "epoch": 3.628383190883191, + "grad_norm": 1.110133171081543, + "learning_rate": 4.256837345339892e-06, + "loss": 0.902, + "step": 20380 + }, + { + "epoch": 3.6285612535612537, + "grad_norm": 1.1139036417007446, + "learning_rate": 4.252797770922834e-06, + "loss": 0.7422, + "step": 20381 + }, + { + "epoch": 3.628739316239316, + "grad_norm": 1.043532371520996, + "learning_rate": 4.2487600724733016e-06, + "loss": 0.9144, + "step": 20382 + }, + { + "epoch": 3.6289173789173788, + "grad_norm": 0.9324162006378174, + "learning_rate": 4.24472425007042e-06, + "loss": 0.75, + "step": 20383 + }, + { + "epoch": 3.6290954415954415, + "grad_norm": 1.013924479484558, + "learning_rate": 4.240690303793238e-06, + "loss": 0.7576, + "step": 20384 + }, + { + "epoch": 3.6292735042735043, + "grad_norm": 1.1732478141784668, + "learning_rate": 4.236658233720814e-06, + "loss": 0.7254, + "step": 20385 + }, + { + "epoch": 3.629451566951567, + "grad_norm": 0.9141867756843567, + "learning_rate": 4.232628039932152e-06, + "loss": 0.7594, + "step": 20386 + }, + { + "epoch": 3.6296296296296298, + "grad_norm": 0.8569268584251404, + "learning_rate": 4.228599722506177e-06, + "loss": 0.5785, + "step": 20387 + }, + { + "epoch": 3.6298076923076925, + "grad_norm": 0.9507982730865479, + "learning_rate": 4.224573281521871e-06, + "loss": 0.799, + "step": 20388 + }, + { + "epoch": 3.629985754985755, + "grad_norm": 0.9895933270454407, + "learning_rate": 4.22054871705807e-06, + "loss": 0.8099, + "step": 20389 + }, + { + "epoch": 3.6301638176638176, + "grad_norm": 0.9135871529579163, + "learning_rate": 4.216526029193668e-06, + "loss": 0.9647, + "step": 20390 + }, + { + "epoch": 3.6303418803418803, + "grad_norm": 0.9216475486755371, + "learning_rate": 4.2125052180074674e-06, + "loss": 0.9267, + "step": 20391 + }, + { + "epoch": 3.630519943019943, + "grad_norm": 0.8940479755401611, + "learning_rate": 4.208486283578228e-06, + "loss": 0.6267, + "step": 20392 + }, + { + "epoch": 3.630698005698006, + "grad_norm": 0.8756904006004333, + "learning_rate": 4.204469225984731e-06, + "loss": 0.7487, + "step": 20393 + }, + { + "epoch": 3.630876068376068, + "grad_norm": 0.9734384417533875, + "learning_rate": 4.200454045305646e-06, + "loss": 0.3735, + "step": 20394 + }, + { + "epoch": 3.631054131054131, + "grad_norm": 0.9914159178733826, + "learning_rate": 4.196440741619678e-06, + "loss": 0.85, + "step": 20395 + }, + { + "epoch": 3.6312321937321936, + "grad_norm": 0.9292317032814026, + "learning_rate": 4.19242931500542e-06, + "loss": 0.9003, + "step": 20396 + }, + { + "epoch": 3.6314102564102564, + "grad_norm": 1.0771266222000122, + "learning_rate": 4.1884197655414955e-06, + "loss": 0.8604, + "step": 20397 + }, + { + "epoch": 3.631588319088319, + "grad_norm": 0.9089133143424988, + "learning_rate": 4.1844120933064445e-06, + "loss": 0.7653, + "step": 20398 + }, + { + "epoch": 3.631766381766382, + "grad_norm": 1.0063055753707886, + "learning_rate": 4.180406298378814e-06, + "loss": 0.7252, + "step": 20399 + }, + { + "epoch": 3.6319444444444446, + "grad_norm": 0.9310106039047241, + "learning_rate": 4.1764023808370635e-06, + "loss": 0.8539, + "step": 20400 + }, + { + "epoch": 3.6321225071225074, + "grad_norm": 0.9510707855224609, + "learning_rate": 4.172400340759652e-06, + "loss": 0.6733, + "step": 20401 + }, + { + "epoch": 3.6323005698005697, + "grad_norm": 0.9140132069587708, + "learning_rate": 4.168400178224985e-06, + "loss": 0.9194, + "step": 20402 + }, + { + "epoch": 3.6324786324786325, + "grad_norm": 0.8771872520446777, + "learning_rate": 4.1644018933114645e-06, + "loss": 0.7011, + "step": 20403 + }, + { + "epoch": 3.632656695156695, + "grad_norm": 0.9165019989013672, + "learning_rate": 4.160405486097385e-06, + "loss": 0.6365, + "step": 20404 + }, + { + "epoch": 3.632834757834758, + "grad_norm": 1.0388853549957275, + "learning_rate": 4.156410956661083e-06, + "loss": 0.8182, + "step": 20405 + }, + { + "epoch": 3.6330128205128203, + "grad_norm": 1.0131229162216187, + "learning_rate": 4.152418305080819e-06, + "loss": 0.8188, + "step": 20406 + }, + { + "epoch": 3.633190883190883, + "grad_norm": 0.919827401638031, + "learning_rate": 4.1484275314347955e-06, + "loss": 0.786, + "step": 20407 + }, + { + "epoch": 3.6333689458689458, + "grad_norm": 0.9317741990089417, + "learning_rate": 4.144438635801229e-06, + "loss": 0.7544, + "step": 20408 + }, + { + "epoch": 3.6335470085470085, + "grad_norm": 0.9361299872398376, + "learning_rate": 4.140451618258267e-06, + "loss": 0.7855, + "step": 20409 + }, + { + "epoch": 3.6337250712250713, + "grad_norm": 0.9616186022758484, + "learning_rate": 4.136466478884016e-06, + "loss": 0.7551, + "step": 20410 + }, + { + "epoch": 3.633903133903134, + "grad_norm": 0.8722997307777405, + "learning_rate": 4.132483217756567e-06, + "loss": 0.7949, + "step": 20411 + }, + { + "epoch": 3.6340811965811968, + "grad_norm": 1.0613240003585815, + "learning_rate": 4.128501834953957e-06, + "loss": 0.8799, + "step": 20412 + }, + { + "epoch": 3.6342592592592595, + "grad_norm": 0.8569284677505493, + "learning_rate": 4.124522330554215e-06, + "loss": 0.6788, + "step": 20413 + }, + { + "epoch": 3.634437321937322, + "grad_norm": 0.9483793377876282, + "learning_rate": 4.1205447046352766e-06, + "loss": 0.7351, + "step": 20414 + }, + { + "epoch": 3.6346153846153846, + "grad_norm": 1.0549039840698242, + "learning_rate": 4.116568957275102e-06, + "loss": 0.8051, + "step": 20415 + }, + { + "epoch": 3.6347934472934473, + "grad_norm": 0.9289839267730713, + "learning_rate": 4.112595088551574e-06, + "loss": 0.7473, + "step": 20416 + }, + { + "epoch": 3.63497150997151, + "grad_norm": 0.893508791923523, + "learning_rate": 4.108623098542552e-06, + "loss": 0.8555, + "step": 20417 + }, + { + "epoch": 3.6351495726495724, + "grad_norm": 1.0892490148544312, + "learning_rate": 4.1046529873258854e-06, + "loss": 0.9039, + "step": 20418 + }, + { + "epoch": 3.635327635327635, + "grad_norm": 1.0458922386169434, + "learning_rate": 4.1006847549793115e-06, + "loss": 0.7172, + "step": 20419 + }, + { + "epoch": 3.635505698005698, + "grad_norm": 0.9979022145271301, + "learning_rate": 4.0967184015806235e-06, + "loss": 0.7928, + "step": 20420 + }, + { + "epoch": 3.6356837606837606, + "grad_norm": 0.9028869867324829, + "learning_rate": 4.092753927207505e-06, + "loss": 0.8833, + "step": 20421 + }, + { + "epoch": 3.6358618233618234, + "grad_norm": 0.9818532466888428, + "learning_rate": 4.0887913319376466e-06, + "loss": 0.8457, + "step": 20422 + }, + { + "epoch": 3.636039886039886, + "grad_norm": 1.0005574226379395, + "learning_rate": 4.084830615848689e-06, + "loss": 0.9425, + "step": 20423 + }, + { + "epoch": 3.636217948717949, + "grad_norm": 0.7937536835670471, + "learning_rate": 4.080871779018225e-06, + "loss": 0.5992, + "step": 20424 + }, + { + "epoch": 3.6363960113960117, + "grad_norm": 0.8660478591918945, + "learning_rate": 4.076914821523825e-06, + "loss": 0.5975, + "step": 20425 + }, + { + "epoch": 3.636574074074074, + "grad_norm": 0.9407948851585388, + "learning_rate": 4.072959743443017e-06, + "loss": 0.9115, + "step": 20426 + }, + { + "epoch": 3.6367521367521367, + "grad_norm": 0.9861929416656494, + "learning_rate": 4.069006544853271e-06, + "loss": 0.7914, + "step": 20427 + }, + { + "epoch": 3.6369301994301995, + "grad_norm": 1.0055484771728516, + "learning_rate": 4.065055225832104e-06, + "loss": 0.885, + "step": 20428 + }, + { + "epoch": 3.637108262108262, + "grad_norm": 0.8550659418106079, + "learning_rate": 4.0611057864568536e-06, + "loss": 0.61, + "step": 20429 + }, + { + "epoch": 3.6372863247863245, + "grad_norm": 0.8787573575973511, + "learning_rate": 4.057158226804958e-06, + "loss": 0.8518, + "step": 20430 + }, + { + "epoch": 3.6374643874643873, + "grad_norm": 0.9358032941818237, + "learning_rate": 4.053212546953744e-06, + "loss": 0.5766, + "step": 20431 + }, + { + "epoch": 3.63764245014245, + "grad_norm": 1.0040407180786133, + "learning_rate": 4.049268746980517e-06, + "loss": 0.7181, + "step": 20432 + }, + { + "epoch": 3.6378205128205128, + "grad_norm": 1.0668649673461914, + "learning_rate": 4.045326826962548e-06, + "loss": 0.8602, + "step": 20433 + }, + { + "epoch": 3.6379985754985755, + "grad_norm": 1.019614338874817, + "learning_rate": 4.041386786977053e-06, + "loss": 0.9243, + "step": 20434 + }, + { + "epoch": 3.6381766381766383, + "grad_norm": 0.9878685474395752, + "learning_rate": 4.037448627101281e-06, + "loss": 0.8588, + "step": 20435 + }, + { + "epoch": 3.638354700854701, + "grad_norm": 0.9777923822402954, + "learning_rate": 4.033512347412327e-06, + "loss": 0.7589, + "step": 20436 + }, + { + "epoch": 3.638532763532764, + "grad_norm": 0.9318233132362366, + "learning_rate": 4.029577947987362e-06, + "loss": 0.7857, + "step": 20437 + }, + { + "epoch": 3.638710826210826, + "grad_norm": 0.9956740140914917, + "learning_rate": 4.025645428903446e-06, + "loss": 0.7942, + "step": 20438 + }, + { + "epoch": 3.638888888888889, + "grad_norm": 0.9890974760055542, + "learning_rate": 4.021714790237652e-06, + "loss": 0.8209, + "step": 20439 + }, + { + "epoch": 3.6390669515669516, + "grad_norm": 0.9391531348228455, + "learning_rate": 4.017786032066972e-06, + "loss": 0.8631, + "step": 20440 + }, + { + "epoch": 3.6392450142450143, + "grad_norm": 0.8784220814704895, + "learning_rate": 4.013859154468391e-06, + "loss": 0.6432, + "step": 20441 + }, + { + "epoch": 3.6394230769230766, + "grad_norm": 0.7741743922233582, + "learning_rate": 4.009934157518835e-06, + "loss": 0.6588, + "step": 20442 + }, + { + "epoch": 3.6396011396011394, + "grad_norm": 0.9611365795135498, + "learning_rate": 4.0060110412952325e-06, + "loss": 0.7351, + "step": 20443 + }, + { + "epoch": 3.639779202279202, + "grad_norm": 0.9556607007980347, + "learning_rate": 4.0020898058744204e-06, + "loss": 0.7602, + "step": 20444 + }, + { + "epoch": 3.639957264957265, + "grad_norm": 0.8218309879302979, + "learning_rate": 3.998170451333261e-06, + "loss": 0.6772, + "step": 20445 + }, + { + "epoch": 3.6401353276353277, + "grad_norm": 1.0754879713058472, + "learning_rate": 3.994252977748503e-06, + "loss": 0.847, + "step": 20446 + }, + { + "epoch": 3.6403133903133904, + "grad_norm": 1.021073341369629, + "learning_rate": 3.990337385196929e-06, + "loss": 0.6531, + "step": 20447 + }, + { + "epoch": 3.640491452991453, + "grad_norm": 0.983539879322052, + "learning_rate": 3.986423673755257e-06, + "loss": 0.6254, + "step": 20448 + }, + { + "epoch": 3.640669515669516, + "grad_norm": 0.8971178531646729, + "learning_rate": 3.982511843500159e-06, + "loss": 0.7201, + "step": 20449 + }, + { + "epoch": 3.640847578347578, + "grad_norm": 0.9377745985984802, + "learning_rate": 3.978601894508282e-06, + "loss": 0.7173, + "step": 20450 + }, + { + "epoch": 3.641025641025641, + "grad_norm": 0.9671639204025269, + "learning_rate": 3.974693826856224e-06, + "loss": 0.81, + "step": 20451 + }, + { + "epoch": 3.6412037037037037, + "grad_norm": 1.0108625888824463, + "learning_rate": 3.970787640620577e-06, + "loss": 0.5756, + "step": 20452 + }, + { + "epoch": 3.6413817663817665, + "grad_norm": 0.9857928156852722, + "learning_rate": 3.966883335877858e-06, + "loss": 0.6986, + "step": 20453 + }, + { + "epoch": 3.6415598290598292, + "grad_norm": 1.0502939224243164, + "learning_rate": 3.962980912704572e-06, + "loss": 0.6687, + "step": 20454 + }, + { + "epoch": 3.6417378917378915, + "grad_norm": 0.8568015694618225, + "learning_rate": 3.9590803711771705e-06, + "loss": 0.6249, + "step": 20455 + }, + { + "epoch": 3.6419159544159543, + "grad_norm": 0.888664722442627, + "learning_rate": 3.9551817113720915e-06, + "loss": 0.6097, + "step": 20456 + }, + { + "epoch": 3.642094017094017, + "grad_norm": 1.024553894996643, + "learning_rate": 3.9512849333657065e-06, + "loss": 0.891, + "step": 20457 + }, + { + "epoch": 3.64227207977208, + "grad_norm": 0.9419230222702026, + "learning_rate": 3.947390037234366e-06, + "loss": 0.8291, + "step": 20458 + }, + { + "epoch": 3.6424501424501425, + "grad_norm": 0.9657182097434998, + "learning_rate": 3.9434970230543765e-06, + "loss": 0.5703, + "step": 20459 + }, + { + "epoch": 3.6426282051282053, + "grad_norm": 1.196069598197937, + "learning_rate": 3.939605890902054e-06, + "loss": 0.8661, + "step": 20460 + }, + { + "epoch": 3.642806267806268, + "grad_norm": 1.0185527801513672, + "learning_rate": 3.935716640853571e-06, + "loss": 0.6936, + "step": 20461 + }, + { + "epoch": 3.6429843304843303, + "grad_norm": 1.0709651708602905, + "learning_rate": 3.931829272985177e-06, + "loss": 0.6632, + "step": 20462 + }, + { + "epoch": 3.643162393162393, + "grad_norm": 0.8733354210853577, + "learning_rate": 3.927943787373034e-06, + "loss": 0.8073, + "step": 20463 + }, + { + "epoch": 3.643340455840456, + "grad_norm": 1.1199495792388916, + "learning_rate": 3.924060184093248e-06, + "loss": 0.7714, + "step": 20464 + }, + { + "epoch": 3.6435185185185186, + "grad_norm": 0.9008165597915649, + "learning_rate": 3.920178463221924e-06, + "loss": 0.7476, + "step": 20465 + }, + { + "epoch": 3.6436965811965814, + "grad_norm": 0.9498278498649597, + "learning_rate": 3.916298624835124e-06, + "loss": 0.829, + "step": 20466 + }, + { + "epoch": 3.6438746438746437, + "grad_norm": 1.033387541770935, + "learning_rate": 3.9124206690088425e-06, + "loss": 0.5773, + "step": 20467 + }, + { + "epoch": 3.6440527065527064, + "grad_norm": 0.8976402878761292, + "learning_rate": 3.908544595819086e-06, + "loss": 0.6666, + "step": 20468 + }, + { + "epoch": 3.644230769230769, + "grad_norm": 1.055590033531189, + "learning_rate": 3.90467040534177e-06, + "loss": 0.8151, + "step": 20469 + }, + { + "epoch": 3.644408831908832, + "grad_norm": 0.9862080216407776, + "learning_rate": 3.9007980976528246e-06, + "loss": 0.6903, + "step": 20470 + }, + { + "epoch": 3.6445868945868947, + "grad_norm": 0.9331284761428833, + "learning_rate": 3.89692767282811e-06, + "loss": 0.801, + "step": 20471 + }, + { + "epoch": 3.6447649572649574, + "grad_norm": 0.9865574240684509, + "learning_rate": 3.893059130943466e-06, + "loss": 0.6761, + "step": 20472 + }, + { + "epoch": 3.64494301994302, + "grad_norm": 0.9861798286437988, + "learning_rate": 3.889192472074677e-06, + "loss": 0.803, + "step": 20473 + }, + { + "epoch": 3.6451210826210825, + "grad_norm": 0.9552395343780518, + "learning_rate": 3.885327696297503e-06, + "loss": 0.8508, + "step": 20474 + }, + { + "epoch": 3.6452991452991452, + "grad_norm": 1.0771273374557495, + "learning_rate": 3.881464803687695e-06, + "loss": 0.7232, + "step": 20475 + }, + { + "epoch": 3.645477207977208, + "grad_norm": 0.8707461953163147, + "learning_rate": 3.877603794320894e-06, + "loss": 0.7563, + "step": 20476 + }, + { + "epoch": 3.6456552706552707, + "grad_norm": 1.192508339881897, + "learning_rate": 3.873744668272772e-06, + "loss": 0.834, + "step": 20477 + }, + { + "epoch": 3.6458333333333335, + "grad_norm": 1.0404844284057617, + "learning_rate": 3.869887425618945e-06, + "loss": 0.8408, + "step": 20478 + }, + { + "epoch": 3.646011396011396, + "grad_norm": 0.8634822368621826, + "learning_rate": 3.866032066434988e-06, + "loss": 0.687, + "step": 20479 + }, + { + "epoch": 3.6461894586894585, + "grad_norm": 0.9610817432403564, + "learning_rate": 3.862178590796417e-06, + "loss": 0.7168, + "step": 20480 + }, + { + "epoch": 3.6463675213675213, + "grad_norm": 0.9478577971458435, + "learning_rate": 3.858326998778761e-06, + "loss": 0.7505, + "step": 20481 + }, + { + "epoch": 3.646545584045584, + "grad_norm": 1.1278072595596313, + "learning_rate": 3.8544772904574585e-06, + "loss": 0.8973, + "step": 20482 + }, + { + "epoch": 3.646723646723647, + "grad_norm": 0.83364337682724, + "learning_rate": 3.850629465907951e-06, + "loss": 0.6548, + "step": 20483 + }, + { + "epoch": 3.6469017094017095, + "grad_norm": 0.965010404586792, + "learning_rate": 3.846783525205622e-06, + "loss": 0.7875, + "step": 20484 + }, + { + "epoch": 3.6470797720797723, + "grad_norm": 1.0525926351547241, + "learning_rate": 3.842939468425844e-06, + "loss": 0.9994, + "step": 20485 + }, + { + "epoch": 3.6472578347578346, + "grad_norm": 0.9967437386512756, + "learning_rate": 3.839097295643901e-06, + "loss": 0.7788, + "step": 20486 + }, + { + "epoch": 3.6474358974358974, + "grad_norm": 1.0144619941711426, + "learning_rate": 3.8352570069351e-06, + "loss": 0.9022, + "step": 20487 + }, + { + "epoch": 3.64761396011396, + "grad_norm": 1.0700451135635376, + "learning_rate": 3.83141860237467e-06, + "loss": 0.7186, + "step": 20488 + }, + { + "epoch": 3.647792022792023, + "grad_norm": 0.9724757671356201, + "learning_rate": 3.827582082037817e-06, + "loss": 0.7279, + "step": 20489 + }, + { + "epoch": 3.6479700854700856, + "grad_norm": 0.9472264647483826, + "learning_rate": 3.823747445999714e-06, + "loss": 0.5964, + "step": 20490 + }, + { + "epoch": 3.648148148148148, + "grad_norm": 0.8835110068321228, + "learning_rate": 3.81991469433548e-06, + "loss": 0.7234, + "step": 20491 + }, + { + "epoch": 3.6483262108262107, + "grad_norm": 0.9631081819534302, + "learning_rate": 3.816083827120254e-06, + "loss": 0.7303, + "step": 20492 + }, + { + "epoch": 3.6485042735042734, + "grad_norm": 1.0134692192077637, + "learning_rate": 3.8122548444290307e-06, + "loss": 0.8875, + "step": 20493 + }, + { + "epoch": 3.648682336182336, + "grad_norm": 0.8940348625183105, + "learning_rate": 3.8084277463368623e-06, + "loss": 0.8223, + "step": 20494 + }, + { + "epoch": 3.648860398860399, + "grad_norm": 0.9974220991134644, + "learning_rate": 3.804602532918744e-06, + "loss": 0.7668, + "step": 20495 + }, + { + "epoch": 3.6490384615384617, + "grad_norm": 0.9602738618850708, + "learning_rate": 3.800779204249605e-06, + "loss": 0.6415, + "step": 20496 + }, + { + "epoch": 3.6492165242165244, + "grad_norm": 0.9182425737380981, + "learning_rate": 3.796957760404363e-06, + "loss": 0.8986, + "step": 20497 + }, + { + "epoch": 3.6493945868945867, + "grad_norm": 0.9976184964179993, + "learning_rate": 3.793138201457891e-06, + "loss": 0.7179, + "step": 20498 + }, + { + "epoch": 3.6495726495726495, + "grad_norm": 0.9328181147575378, + "learning_rate": 3.7893205274850076e-06, + "loss": 0.7968, + "step": 20499 + }, + { + "epoch": 3.6497507122507122, + "grad_norm": 0.9778452515602112, + "learning_rate": 3.7855047385605525e-06, + "loss": 0.6892, + "step": 20500 + }, + { + "epoch": 3.649928774928775, + "grad_norm": 1.0850595235824585, + "learning_rate": 3.781690834759244e-06, + "loss": 0.9374, + "step": 20501 + }, + { + "epoch": 3.6501068376068377, + "grad_norm": 0.977279007434845, + "learning_rate": 3.7778788161558444e-06, + "loss": 0.8858, + "step": 20502 + }, + { + "epoch": 3.6502849002849, + "grad_norm": 0.8755964040756226, + "learning_rate": 3.774068682825005e-06, + "loss": 0.8847, + "step": 20503 + }, + { + "epoch": 3.650462962962963, + "grad_norm": 0.9674113988876343, + "learning_rate": 3.7702604348414107e-06, + "loss": 0.7908, + "step": 20504 + }, + { + "epoch": 3.6506410256410255, + "grad_norm": 0.9421332478523254, + "learning_rate": 3.766454072279657e-06, + "loss": 0.8609, + "step": 20505 + }, + { + "epoch": 3.6508190883190883, + "grad_norm": 1.0132116079330444, + "learning_rate": 3.76264959521434e-06, + "loss": 0.9126, + "step": 20506 + }, + { + "epoch": 3.650997150997151, + "grad_norm": 0.9830502867698669, + "learning_rate": 3.7588470037199787e-06, + "loss": 0.9262, + "step": 20507 + }, + { + "epoch": 3.651175213675214, + "grad_norm": 0.9435737133026123, + "learning_rate": 3.7550462978710897e-06, + "loss": 0.7092, + "step": 20508 + }, + { + "epoch": 3.6513532763532766, + "grad_norm": 0.9076669216156006, + "learning_rate": 3.7512474777421257e-06, + "loss": 0.8453, + "step": 20509 + }, + { + "epoch": 3.6515313390313393, + "grad_norm": 0.6464830636978149, + "learning_rate": 3.747450543407538e-06, + "loss": 0.3109, + "step": 20510 + }, + { + "epoch": 3.6517094017094016, + "grad_norm": 0.9773824214935303, + "learning_rate": 3.7436554949417113e-06, + "loss": 0.7024, + "step": 20511 + }, + { + "epoch": 3.6518874643874644, + "grad_norm": 0.7433509826660156, + "learning_rate": 3.7398623324189973e-06, + "loss": 0.4606, + "step": 20512 + }, + { + "epoch": 3.652065527065527, + "grad_norm": 0.8884417414665222, + "learning_rate": 3.7360710559137146e-06, + "loss": 0.5894, + "step": 20513 + }, + { + "epoch": 3.65224358974359, + "grad_norm": 1.0048164129257202, + "learning_rate": 3.732281665500148e-06, + "loss": 0.834, + "step": 20514 + }, + { + "epoch": 3.652421652421652, + "grad_norm": 0.9860281944274902, + "learning_rate": 3.7284941612525605e-06, + "loss": 0.7918, + "step": 20515 + }, + { + "epoch": 3.652599715099715, + "grad_norm": 1.150870680809021, + "learning_rate": 3.7247085432451147e-06, + "loss": 0.9382, + "step": 20516 + }, + { + "epoch": 3.6527777777777777, + "grad_norm": 0.9690985083580017, + "learning_rate": 3.72092481155204e-06, + "loss": 0.6593, + "step": 20517 + }, + { + "epoch": 3.6529558404558404, + "grad_norm": 0.7312018871307373, + "learning_rate": 3.7171429662474223e-06, + "loss": 0.5452, + "step": 20518 + }, + { + "epoch": 3.653133903133903, + "grad_norm": 0.9348322153091431, + "learning_rate": 3.7133630074053794e-06, + "loss": 0.8746, + "step": 20519 + }, + { + "epoch": 3.653311965811966, + "grad_norm": 0.85910564661026, + "learning_rate": 3.7095849350999746e-06, + "loss": 0.6533, + "step": 20520 + }, + { + "epoch": 3.6534900284900287, + "grad_norm": 0.9279438853263855, + "learning_rate": 3.705808749405226e-06, + "loss": 0.6357, + "step": 20521 + }, + { + "epoch": 3.6536680911680914, + "grad_norm": 1.0286200046539307, + "learning_rate": 3.70203445039512e-06, + "loss": 0.8155, + "step": 20522 + }, + { + "epoch": 3.6538461538461537, + "grad_norm": 0.9838545322418213, + "learning_rate": 3.698262038143618e-06, + "loss": 0.8736, + "step": 20523 + }, + { + "epoch": 3.6540242165242165, + "grad_norm": 0.9309592843055725, + "learning_rate": 3.694491512724596e-06, + "loss": 0.8286, + "step": 20524 + }, + { + "epoch": 3.6542022792022792, + "grad_norm": 0.9647026062011719, + "learning_rate": 3.6907228742119825e-06, + "loss": 0.7917, + "step": 20525 + }, + { + "epoch": 3.654380341880342, + "grad_norm": 1.0884795188903809, + "learning_rate": 3.6869561226795744e-06, + "loss": 0.9536, + "step": 20526 + }, + { + "epoch": 3.6545584045584043, + "grad_norm": 0.8624057173728943, + "learning_rate": 3.6831912582012017e-06, + "loss": 0.6808, + "step": 20527 + }, + { + "epoch": 3.654736467236467, + "grad_norm": 0.9675288796424866, + "learning_rate": 3.6794282808505943e-06, + "loss": 0.7873, + "step": 20528 + }, + { + "epoch": 3.65491452991453, + "grad_norm": 1.0651471614837646, + "learning_rate": 3.675667190701515e-06, + "loss": 0.8461, + "step": 20529 + }, + { + "epoch": 3.6550925925925926, + "grad_norm": 1.0895347595214844, + "learning_rate": 3.6719079878276387e-06, + "loss": 1.2929, + "step": 20530 + }, + { + "epoch": 3.6552706552706553, + "grad_norm": 0.887728214263916, + "learning_rate": 3.668150672302606e-06, + "loss": 0.7159, + "step": 20531 + }, + { + "epoch": 3.655448717948718, + "grad_norm": 0.9625465869903564, + "learning_rate": 3.6643952442000807e-06, + "loss": 0.8361, + "step": 20532 + }, + { + "epoch": 3.655626780626781, + "grad_norm": 0.8512797951698303, + "learning_rate": 3.6606417035935816e-06, + "loss": 0.7388, + "step": 20533 + }, + { + "epoch": 3.6558048433048436, + "grad_norm": 1.0100477933883667, + "learning_rate": 3.656890050556694e-06, + "loss": 0.6821, + "step": 20534 + }, + { + "epoch": 3.655982905982906, + "grad_norm": 0.9464243054389954, + "learning_rate": 3.6531402851629036e-06, + "loss": 0.7139, + "step": 20535 + }, + { + "epoch": 3.6561609686609686, + "grad_norm": 0.8691574931144714, + "learning_rate": 3.6493924074856966e-06, + "loss": 0.6875, + "step": 20536 + }, + { + "epoch": 3.6563390313390314, + "grad_norm": 0.8101440072059631, + "learning_rate": 3.645646417598492e-06, + "loss": 0.5582, + "step": 20537 + }, + { + "epoch": 3.656517094017094, + "grad_norm": 0.8926877379417419, + "learning_rate": 3.6419023155746854e-06, + "loss": 0.6672, + "step": 20538 + }, + { + "epoch": 3.6566951566951564, + "grad_norm": 0.9474127888679504, + "learning_rate": 3.638160101487631e-06, + "loss": 0.7278, + "step": 20539 + }, + { + "epoch": 3.656873219373219, + "grad_norm": 0.9432591199874878, + "learning_rate": 3.634419775410658e-06, + "loss": 0.7238, + "step": 20540 + }, + { + "epoch": 3.657051282051282, + "grad_norm": 0.9014365673065186, + "learning_rate": 3.630681337417041e-06, + "loss": 0.6098, + "step": 20541 + }, + { + "epoch": 3.6572293447293447, + "grad_norm": 0.9230480790138245, + "learning_rate": 3.6269447875800557e-06, + "loss": 0.6788, + "step": 20542 + }, + { + "epoch": 3.6574074074074074, + "grad_norm": 0.9330241680145264, + "learning_rate": 3.6232101259728644e-06, + "loss": 0.8389, + "step": 20543 + }, + { + "epoch": 3.65758547008547, + "grad_norm": 0.9799591898918152, + "learning_rate": 3.6194773526686764e-06, + "loss": 0.8186, + "step": 20544 + }, + { + "epoch": 3.657763532763533, + "grad_norm": 1.0543632507324219, + "learning_rate": 3.6157464677406105e-06, + "loss": 0.7756, + "step": 20545 + }, + { + "epoch": 3.6579415954415957, + "grad_norm": 0.8438118696212769, + "learning_rate": 3.612017471261775e-06, + "loss": 0.6754, + "step": 20546 + }, + { + "epoch": 3.658119658119658, + "grad_norm": 0.8986937403678894, + "learning_rate": 3.6082903633052335e-06, + "loss": 0.656, + "step": 20547 + }, + { + "epoch": 3.6582977207977208, + "grad_norm": 1.0008556842803955, + "learning_rate": 3.604565143944005e-06, + "loss": 0.7939, + "step": 20548 + }, + { + "epoch": 3.6584757834757835, + "grad_norm": 1.050921082496643, + "learning_rate": 3.600841813251066e-06, + "loss": 0.6991, + "step": 20549 + }, + { + "epoch": 3.6586538461538463, + "grad_norm": 1.1899060010910034, + "learning_rate": 3.5971203712993894e-06, + "loss": 0.8935, + "step": 20550 + }, + { + "epoch": 3.6588319088319086, + "grad_norm": 0.9493093490600586, + "learning_rate": 3.593400818161885e-06, + "loss": 0.8281, + "step": 20551 + }, + { + "epoch": 3.6590099715099713, + "grad_norm": 0.9087548851966858, + "learning_rate": 3.5896831539114162e-06, + "loss": 0.5721, + "step": 20552 + }, + { + "epoch": 3.659188034188034, + "grad_norm": 0.934526801109314, + "learning_rate": 3.585967378620836e-06, + "loss": 0.7392, + "step": 20553 + }, + { + "epoch": 3.659366096866097, + "grad_norm": 0.8018043041229248, + "learning_rate": 3.582253492362941e-06, + "loss": 0.6153, + "step": 20554 + }, + { + "epoch": 3.6595441595441596, + "grad_norm": 1.037794828414917, + "learning_rate": 3.5785414952104966e-06, + "loss": 0.7521, + "step": 20555 + }, + { + "epoch": 3.6597222222222223, + "grad_norm": 0.9660981297492981, + "learning_rate": 3.5748313872362215e-06, + "loss": 0.7198, + "step": 20556 + }, + { + "epoch": 3.659900284900285, + "grad_norm": 0.9670109152793884, + "learning_rate": 3.5711231685128464e-06, + "loss": 0.6839, + "step": 20557 + }, + { + "epoch": 3.660078347578348, + "grad_norm": 1.0562913417816162, + "learning_rate": 3.5674168391129693e-06, + "loss": 1.0162, + "step": 20558 + }, + { + "epoch": 3.66025641025641, + "grad_norm": 0.9485027194023132, + "learning_rate": 3.5637123991092538e-06, + "loss": 0.8613, + "step": 20559 + }, + { + "epoch": 3.660434472934473, + "grad_norm": 1.0248631238937378, + "learning_rate": 3.5600098485742637e-06, + "loss": 0.816, + "step": 20560 + }, + { + "epoch": 3.6606125356125356, + "grad_norm": 0.9192467331886292, + "learning_rate": 3.556309187580553e-06, + "loss": 0.7076, + "step": 20561 + }, + { + "epoch": 3.6607905982905984, + "grad_norm": 0.982490599155426, + "learning_rate": 3.552610416200608e-06, + "loss": 0.6947, + "step": 20562 + }, + { + "epoch": 3.6609686609686607, + "grad_norm": 0.8929651975631714, + "learning_rate": 3.5489135345069147e-06, + "loss": 0.7547, + "step": 20563 + }, + { + "epoch": 3.6611467236467234, + "grad_norm": 1.0499329566955566, + "learning_rate": 3.545218542571893e-06, + "loss": 0.6648, + "step": 20564 + }, + { + "epoch": 3.661324786324786, + "grad_norm": 1.022430181503296, + "learning_rate": 3.541525440467952e-06, + "loss": 0.7897, + "step": 20565 + }, + { + "epoch": 3.661502849002849, + "grad_norm": 0.9741869568824768, + "learning_rate": 3.5378342282674336e-06, + "loss": 0.7267, + "step": 20566 + }, + { + "epoch": 3.6616809116809117, + "grad_norm": 0.9906203150749207, + "learning_rate": 3.534144906042702e-06, + "loss": 0.5623, + "step": 20567 + }, + { + "epoch": 3.6618589743589745, + "grad_norm": 0.8446068167686462, + "learning_rate": 3.530457473865978e-06, + "loss": 0.79, + "step": 20568 + }, + { + "epoch": 3.662037037037037, + "grad_norm": 0.9871222376823425, + "learning_rate": 3.5267719318095583e-06, + "loss": 0.8294, + "step": 20569 + }, + { + "epoch": 3.6622150997151, + "grad_norm": 1.349190592765808, + "learning_rate": 3.52308827994563e-06, + "loss": 0.9821, + "step": 20570 + }, + { + "epoch": 3.6623931623931623, + "grad_norm": 0.9572153687477112, + "learning_rate": 3.5194065183463686e-06, + "loss": 0.7993, + "step": 20571 + }, + { + "epoch": 3.662571225071225, + "grad_norm": 0.7855825424194336, + "learning_rate": 3.5157266470839277e-06, + "loss": 0.6361, + "step": 20572 + }, + { + "epoch": 3.6627492877492878, + "grad_norm": 0.958167314529419, + "learning_rate": 3.512048666230383e-06, + "loss": 0.7887, + "step": 20573 + }, + { + "epoch": 3.6629273504273505, + "grad_norm": 1.0252838134765625, + "learning_rate": 3.5083725758578325e-06, + "loss": 0.6965, + "step": 20574 + }, + { + "epoch": 3.6631054131054133, + "grad_norm": 0.8903803825378418, + "learning_rate": 3.5046983760382403e-06, + "loss": 0.7763, + "step": 20575 + }, + { + "epoch": 3.6632834757834756, + "grad_norm": 1.00473153591156, + "learning_rate": 3.50102606684366e-06, + "loss": 0.6745, + "step": 20576 + }, + { + "epoch": 3.6634615384615383, + "grad_norm": 0.9572219848632812, + "learning_rate": 3.4973556483460013e-06, + "loss": 0.6424, + "step": 20577 + }, + { + "epoch": 3.663639601139601, + "grad_norm": 0.8433740139007568, + "learning_rate": 3.493687120617206e-06, + "loss": 0.6839, + "step": 20578 + }, + { + "epoch": 3.663817663817664, + "grad_norm": 0.9884551763534546, + "learning_rate": 3.4900204837291284e-06, + "loss": 0.7555, + "step": 20579 + }, + { + "epoch": 3.6639957264957266, + "grad_norm": 1.024325966835022, + "learning_rate": 3.4863557377536103e-06, + "loss": 0.7884, + "step": 20580 + }, + { + "epoch": 3.6641737891737893, + "grad_norm": 0.9565973281860352, + "learning_rate": 3.4826928827624617e-06, + "loss": 0.7441, + "step": 20581 + }, + { + "epoch": 3.664351851851852, + "grad_norm": 1.0180308818817139, + "learning_rate": 3.479031918827469e-06, + "loss": 0.6455, + "step": 20582 + }, + { + "epoch": 3.6645299145299144, + "grad_norm": 0.9916195869445801, + "learning_rate": 3.4753728460203082e-06, + "loss": 0.9352, + "step": 20583 + }, + { + "epoch": 3.664707977207977, + "grad_norm": 0.912993848323822, + "learning_rate": 3.4717156644127335e-06, + "loss": 0.766, + "step": 20584 + }, + { + "epoch": 3.66488603988604, + "grad_norm": 0.9358601570129395, + "learning_rate": 3.468060374076354e-06, + "loss": 0.7468, + "step": 20585 + }, + { + "epoch": 3.6650641025641026, + "grad_norm": 0.9579420685768127, + "learning_rate": 3.464406975082812e-06, + "loss": 0.7707, + "step": 20586 + }, + { + "epoch": 3.6652421652421654, + "grad_norm": 0.912672221660614, + "learning_rate": 3.4607554675036736e-06, + "loss": 0.7838, + "step": 20587 + }, + { + "epoch": 3.6654202279202277, + "grad_norm": 0.9116925597190857, + "learning_rate": 3.4571058514105027e-06, + "loss": 0.6586, + "step": 20588 + }, + { + "epoch": 3.6655982905982905, + "grad_norm": 1.2207914590835571, + "learning_rate": 3.453458126874776e-06, + "loss": 0.9087, + "step": 20589 + }, + { + "epoch": 3.665776353276353, + "grad_norm": 0.8218117356300354, + "learning_rate": 3.44981229396798e-06, + "loss": 0.4345, + "step": 20590 + }, + { + "epoch": 3.665954415954416, + "grad_norm": 0.9615563154220581, + "learning_rate": 3.4461683527615475e-06, + "loss": 0.7845, + "step": 20591 + }, + { + "epoch": 3.6661324786324787, + "grad_norm": 0.9879812598228455, + "learning_rate": 3.4425263033268762e-06, + "loss": 0.6126, + "step": 20592 + }, + { + "epoch": 3.6663105413105415, + "grad_norm": 0.8715957403182983, + "learning_rate": 3.4388861457353316e-06, + "loss": 0.7591, + "step": 20593 + }, + { + "epoch": 3.666488603988604, + "grad_norm": 0.9076233506202698, + "learning_rate": 3.435247880058212e-06, + "loss": 0.9255, + "step": 20594 + }, + { + "epoch": 3.6666666666666665, + "grad_norm": 0.8798804879188538, + "learning_rate": 3.4316115063668164e-06, + "loss": 0.7389, + "step": 20595 + }, + { + "epoch": 3.6668447293447293, + "grad_norm": 0.9797578454017639, + "learning_rate": 3.427977024732376e-06, + "loss": 0.7214, + "step": 20596 + }, + { + "epoch": 3.667022792022792, + "grad_norm": 0.9402558207511902, + "learning_rate": 3.4243444352261454e-06, + "loss": 0.7828, + "step": 20597 + }, + { + "epoch": 3.6672008547008548, + "grad_norm": 0.8900625705718994, + "learning_rate": 3.4207137379192454e-06, + "loss": 0.7614, + "step": 20598 + }, + { + "epoch": 3.6673789173789175, + "grad_norm": 0.9660239219665527, + "learning_rate": 3.4170849328828525e-06, + "loss": 0.657, + "step": 20599 + }, + { + "epoch": 3.66755698005698, + "grad_norm": 0.9239768981933594, + "learning_rate": 3.4134580201880208e-06, + "loss": 0.716, + "step": 20600 + }, + { + "epoch": 3.6677350427350426, + "grad_norm": 0.8773369193077087, + "learning_rate": 3.4098329999058487e-06, + "loss": 0.7245, + "step": 20601 + }, + { + "epoch": 3.6679131054131053, + "grad_norm": 0.938544750213623, + "learning_rate": 3.4062098721073575e-06, + "loss": 0.7817, + "step": 20602 + }, + { + "epoch": 3.668091168091168, + "grad_norm": 1.0228548049926758, + "learning_rate": 3.402588636863524e-06, + "loss": 0.947, + "step": 20603 + }, + { + "epoch": 3.668269230769231, + "grad_norm": 0.921551525592804, + "learning_rate": 3.398969294245302e-06, + "loss": 0.6954, + "step": 20604 + }, + { + "epoch": 3.6684472934472936, + "grad_norm": 0.9160789251327515, + "learning_rate": 3.395351844323602e-06, + "loss": 0.7823, + "step": 20605 + }, + { + "epoch": 3.6686253561253563, + "grad_norm": 0.9858134984970093, + "learning_rate": 3.3917362871693004e-06, + "loss": 0.7063, + "step": 20606 + }, + { + "epoch": 3.6688034188034186, + "grad_norm": 0.868196964263916, + "learning_rate": 3.3881226228532513e-06, + "loss": 0.6792, + "step": 20607 + }, + { + "epoch": 3.6689814814814814, + "grad_norm": 0.8798336386680603, + "learning_rate": 3.3845108514462322e-06, + "loss": 0.7249, + "step": 20608 + }, + { + "epoch": 3.669159544159544, + "grad_norm": 1.0030553340911865, + "learning_rate": 3.38090097301903e-06, + "loss": 0.7388, + "step": 20609 + }, + { + "epoch": 3.669337606837607, + "grad_norm": 0.8349957466125488, + "learning_rate": 3.3772929876423664e-06, + "loss": 0.6141, + "step": 20610 + }, + { + "epoch": 3.6695156695156697, + "grad_norm": 0.937864363193512, + "learning_rate": 3.3736868953869293e-06, + "loss": 0.7051, + "step": 20611 + }, + { + "epoch": 3.669693732193732, + "grad_norm": 0.9294646382331848, + "learning_rate": 3.3700826963233735e-06, + "loss": 0.7734, + "step": 20612 + }, + { + "epoch": 3.6698717948717947, + "grad_norm": 1.09479820728302, + "learning_rate": 3.3664803905223086e-06, + "loss": 0.8817, + "step": 20613 + }, + { + "epoch": 3.6700498575498575, + "grad_norm": 0.9988318085670471, + "learning_rate": 3.3628799780543342e-06, + "loss": 0.7704, + "step": 20614 + }, + { + "epoch": 3.67022792022792, + "grad_norm": 0.9090431928634644, + "learning_rate": 3.359281458989971e-06, + "loss": 0.6913, + "step": 20615 + }, + { + "epoch": 3.670405982905983, + "grad_norm": 1.36585533618927, + "learning_rate": 3.3556848333997304e-06, + "loss": 0.8042, + "step": 20616 + }, + { + "epoch": 3.6705840455840457, + "grad_norm": 1.0545517206192017, + "learning_rate": 3.3520901013540996e-06, + "loss": 0.7149, + "step": 20617 + }, + { + "epoch": 3.6707621082621085, + "grad_norm": 1.0674611330032349, + "learning_rate": 3.348497262923489e-06, + "loss": 0.9619, + "step": 20618 + }, + { + "epoch": 3.6709401709401708, + "grad_norm": 0.9115710258483887, + "learning_rate": 3.344906318178287e-06, + "loss": 0.823, + "step": 20619 + }, + { + "epoch": 3.6711182336182335, + "grad_norm": 0.8687363862991333, + "learning_rate": 3.3413172671888813e-06, + "loss": 0.6472, + "step": 20620 + }, + { + "epoch": 3.6712962962962963, + "grad_norm": 0.9118328094482422, + "learning_rate": 3.33773011002555e-06, + "loss": 0.6483, + "step": 20621 + }, + { + "epoch": 3.671474358974359, + "grad_norm": 1.0184921026229858, + "learning_rate": 3.3341448467586245e-06, + "loss": 0.6792, + "step": 20622 + }, + { + "epoch": 3.671652421652422, + "grad_norm": 0.8835800290107727, + "learning_rate": 3.330561477458294e-06, + "loss": 0.7004, + "step": 20623 + }, + { + "epoch": 3.671830484330484, + "grad_norm": 1.0011813640594482, + "learning_rate": 3.326980002194835e-06, + "loss": 0.8112, + "step": 20624 + }, + { + "epoch": 3.672008547008547, + "grad_norm": 0.8921766877174377, + "learning_rate": 3.3234004210383473e-06, + "loss": 0.5758, + "step": 20625 + }, + { + "epoch": 3.6721866096866096, + "grad_norm": 0.94114750623703, + "learning_rate": 3.319822734059019e-06, + "loss": 0.7338, + "step": 20626 + }, + { + "epoch": 3.6723646723646723, + "grad_norm": 1.024093747138977, + "learning_rate": 3.316246941326917e-06, + "loss": 0.7397, + "step": 20627 + }, + { + "epoch": 3.672542735042735, + "grad_norm": 0.9239045977592468, + "learning_rate": 3.312673042912129e-06, + "loss": 0.7344, + "step": 20628 + }, + { + "epoch": 3.672720797720798, + "grad_norm": 1.241755485534668, + "learning_rate": 3.3091010388846543e-06, + "loss": 0.7574, + "step": 20629 + }, + { + "epoch": 3.6728988603988606, + "grad_norm": 0.9279388189315796, + "learning_rate": 3.305530929314471e-06, + "loss": 0.6978, + "step": 20630 + }, + { + "epoch": 3.6730769230769234, + "grad_norm": 1.0421857833862305, + "learning_rate": 3.301962714271567e-06, + "loss": 0.7893, + "step": 20631 + }, + { + "epoch": 3.6732549857549857, + "grad_norm": 0.944916307926178, + "learning_rate": 3.29839639382582e-06, + "loss": 0.5987, + "step": 20632 + }, + { + "epoch": 3.6734330484330484, + "grad_norm": 0.8627259135246277, + "learning_rate": 3.2948319680471184e-06, + "loss": 0.6824, + "step": 20633 + }, + { + "epoch": 3.673611111111111, + "grad_norm": 1.1818138360977173, + "learning_rate": 3.2912694370052954e-06, + "loss": 0.9816, + "step": 20634 + }, + { + "epoch": 3.673789173789174, + "grad_norm": 1.1198930740356445, + "learning_rate": 3.2877088007701618e-06, + "loss": 0.8528, + "step": 20635 + }, + { + "epoch": 3.673967236467236, + "grad_norm": 0.9954573512077332, + "learning_rate": 3.2841500594114615e-06, + "loss": 0.9535, + "step": 20636 + }, + { + "epoch": 3.674145299145299, + "grad_norm": 0.9435851573944092, + "learning_rate": 3.2805932129989393e-06, + "loss": 0.7501, + "step": 20637 + }, + { + "epoch": 3.6743233618233617, + "grad_norm": 1.2884607315063477, + "learning_rate": 3.277038261602261e-06, + "loss": 0.815, + "step": 20638 + }, + { + "epoch": 3.6745014245014245, + "grad_norm": 0.9409137964248657, + "learning_rate": 3.273485205291116e-06, + "loss": 1.0288, + "step": 20639 + }, + { + "epoch": 3.6746794871794872, + "grad_norm": 0.8598392009735107, + "learning_rate": 3.26993404413507e-06, + "loss": 0.5697, + "step": 20640 + }, + { + "epoch": 3.67485754985755, + "grad_norm": 0.9609706401824951, + "learning_rate": 3.266384778203735e-06, + "loss": 0.7962, + "step": 20641 + }, + { + "epoch": 3.6750356125356127, + "grad_norm": 1.1136882305145264, + "learning_rate": 3.2628374075666546e-06, + "loss": 0.8393, + "step": 20642 + }, + { + "epoch": 3.6752136752136755, + "grad_norm": 0.9873359203338623, + "learning_rate": 3.259291932293318e-06, + "loss": 0.749, + "step": 20643 + }, + { + "epoch": 3.675391737891738, + "grad_norm": 0.8118230700492859, + "learning_rate": 3.255748352453192e-06, + "loss": 0.6138, + "step": 20644 + }, + { + "epoch": 3.6755698005698005, + "grad_norm": 0.9738235473632812, + "learning_rate": 3.2522066681157094e-06, + "loss": 0.7854, + "step": 20645 + }, + { + "epoch": 3.6757478632478633, + "grad_norm": 0.965652346611023, + "learning_rate": 3.2486668793502593e-06, + "loss": 0.8144, + "step": 20646 + }, + { + "epoch": 3.675925925925926, + "grad_norm": 0.8933370113372803, + "learning_rate": 3.245128986226198e-06, + "loss": 0.9418, + "step": 20647 + }, + { + "epoch": 3.6761039886039883, + "grad_norm": 1.0203384160995483, + "learning_rate": 3.241592988812836e-06, + "loss": 1.0724, + "step": 20648 + }, + { + "epoch": 3.676282051282051, + "grad_norm": 0.8109650611877441, + "learning_rate": 3.2380588871794736e-06, + "loss": 0.6265, + "step": 20649 + }, + { + "epoch": 3.676460113960114, + "grad_norm": 1.0682241916656494, + "learning_rate": 3.2345266813953334e-06, + "loss": 0.797, + "step": 20650 + }, + { + "epoch": 3.6766381766381766, + "grad_norm": 0.9123262166976929, + "learning_rate": 3.2309963715296376e-06, + "loss": 0.6912, + "step": 20651 + }, + { + "epoch": 3.6768162393162394, + "grad_norm": 0.8609589338302612, + "learning_rate": 3.227467957651553e-06, + "loss": 0.8129, + "step": 20652 + }, + { + "epoch": 3.676994301994302, + "grad_norm": 0.8229508399963379, + "learning_rate": 3.22394143983018e-06, + "loss": 0.678, + "step": 20653 + }, + { + "epoch": 3.677172364672365, + "grad_norm": 0.9235960245132446, + "learning_rate": 3.220416818134675e-06, + "loss": 0.6403, + "step": 20654 + }, + { + "epoch": 3.6773504273504276, + "grad_norm": 0.9439797401428223, + "learning_rate": 3.2168940926340264e-06, + "loss": 0.723, + "step": 20655 + }, + { + "epoch": 3.67752849002849, + "grad_norm": 1.1216174364089966, + "learning_rate": 3.2133732633973124e-06, + "loss": 0.5696, + "step": 20656 + }, + { + "epoch": 3.6777065527065527, + "grad_norm": 0.8374674916267395, + "learning_rate": 3.209854330493478e-06, + "loss": 0.6244, + "step": 20657 + }, + { + "epoch": 3.6778846153846154, + "grad_norm": 1.0153228044509888, + "learning_rate": 3.2063372939915014e-06, + "loss": 0.8841, + "step": 20658 + }, + { + "epoch": 3.678062678062678, + "grad_norm": 0.9651415944099426, + "learning_rate": 3.2028221539602608e-06, + "loss": 0.7879, + "step": 20659 + }, + { + "epoch": 3.6782407407407405, + "grad_norm": 0.9977912902832031, + "learning_rate": 3.199308910468646e-06, + "loss": 0.7334, + "step": 20660 + }, + { + "epoch": 3.6784188034188032, + "grad_norm": 0.9887592196464539, + "learning_rate": 3.1957975635854786e-06, + "loss": 0.7634, + "step": 20661 + }, + { + "epoch": 3.678596866096866, + "grad_norm": 1.00025475025177, + "learning_rate": 3.1922881133795825e-06, + "loss": 0.8121, + "step": 20662 + }, + { + "epoch": 3.6787749287749287, + "grad_norm": 0.8790071606636047, + "learning_rate": 3.1887805599196683e-06, + "loss": 0.6167, + "step": 20663 + }, + { + "epoch": 3.6789529914529915, + "grad_norm": 1.0021897554397583, + "learning_rate": 3.1852749032745267e-06, + "loss": 0.6549, + "step": 20664 + }, + { + "epoch": 3.6791310541310542, + "grad_norm": 1.0172662734985352, + "learning_rate": 3.1817711435127906e-06, + "loss": 0.7399, + "step": 20665 + }, + { + "epoch": 3.679309116809117, + "grad_norm": 0.9718804955482483, + "learning_rate": 3.1782692807031276e-06, + "loss": 0.6546, + "step": 20666 + }, + { + "epoch": 3.6794871794871797, + "grad_norm": 1.0091427564620972, + "learning_rate": 3.17476931491415e-06, + "loss": 0.7262, + "step": 20667 + }, + { + "epoch": 3.679665242165242, + "grad_norm": 0.9035787582397461, + "learning_rate": 3.1712712462144134e-06, + "loss": 0.9596, + "step": 20668 + }, + { + "epoch": 3.679843304843305, + "grad_norm": 1.2351285219192505, + "learning_rate": 3.1677750746725077e-06, + "loss": 0.7528, + "step": 20669 + }, + { + "epoch": 3.6800213675213675, + "grad_norm": 1.065185308456421, + "learning_rate": 3.164280800356867e-06, + "loss": 0.8681, + "step": 20670 + }, + { + "epoch": 3.6801994301994303, + "grad_norm": 0.8398473858833313, + "learning_rate": 3.160788423336014e-06, + "loss": 0.5556, + "step": 20671 + }, + { + "epoch": 3.6803774928774926, + "grad_norm": 0.9858914613723755, + "learning_rate": 3.1572979436783168e-06, + "loss": 0.9169, + "step": 20672 + }, + { + "epoch": 3.6805555555555554, + "grad_norm": 0.968228280544281, + "learning_rate": 3.153809361452209e-06, + "loss": 0.7122, + "step": 20673 + }, + { + "epoch": 3.680733618233618, + "grad_norm": 0.9299584031105042, + "learning_rate": 3.1503226767260252e-06, + "loss": 0.7147, + "step": 20674 + }, + { + "epoch": 3.680911680911681, + "grad_norm": 0.9766196608543396, + "learning_rate": 3.1468378895680773e-06, + "loss": 0.8661, + "step": 20675 + }, + { + "epoch": 3.6810897435897436, + "grad_norm": 0.9744870662689209, + "learning_rate": 3.143355000046655e-06, + "loss": 0.7798, + "step": 20676 + }, + { + "epoch": 3.6812678062678064, + "grad_norm": 0.9700605273246765, + "learning_rate": 3.1398740082299817e-06, + "loss": 0.9003, + "step": 20677 + }, + { + "epoch": 3.681445868945869, + "grad_norm": 0.9444677829742432, + "learning_rate": 3.136394914186258e-06, + "loss": 0.8441, + "step": 20678 + }, + { + "epoch": 3.681623931623932, + "grad_norm": 1.0803836584091187, + "learning_rate": 3.1329177179836745e-06, + "loss": 0.9904, + "step": 20679 + }, + { + "epoch": 3.681801994301994, + "grad_norm": 0.8475477695465088, + "learning_rate": 3.1294424196903317e-06, + "loss": 0.8504, + "step": 20680 + }, + { + "epoch": 3.681980056980057, + "grad_norm": 0.882752001285553, + "learning_rate": 3.1259690193743527e-06, + "loss": 0.7105, + "step": 20681 + }, + { + "epoch": 3.6821581196581197, + "grad_norm": 0.9671033620834351, + "learning_rate": 3.122497517103751e-06, + "loss": 0.7545, + "step": 20682 + }, + { + "epoch": 3.6823361823361824, + "grad_norm": 1.0363636016845703, + "learning_rate": 3.1190279129465705e-06, + "loss": 0.7792, + "step": 20683 + }, + { + "epoch": 3.682514245014245, + "grad_norm": 0.9925054907798767, + "learning_rate": 3.1155602069707914e-06, + "loss": 0.7609, + "step": 20684 + }, + { + "epoch": 3.6826923076923075, + "grad_norm": 1.023887038230896, + "learning_rate": 3.1120943992443473e-06, + "loss": 0.8605, + "step": 20685 + }, + { + "epoch": 3.6828703703703702, + "grad_norm": 0.9789915680885315, + "learning_rate": 3.1086304898351513e-06, + "loss": 0.7812, + "step": 20686 + }, + { + "epoch": 3.683048433048433, + "grad_norm": 0.8969167470932007, + "learning_rate": 3.10516847881106e-06, + "loss": 0.608, + "step": 20687 + }, + { + "epoch": 3.6832264957264957, + "grad_norm": 0.90472811460495, + "learning_rate": 3.101708366239908e-06, + "loss": 0.7239, + "step": 20688 + }, + { + "epoch": 3.6834045584045585, + "grad_norm": 1.0310150384902954, + "learning_rate": 3.098250152189497e-06, + "loss": 0.8472, + "step": 20689 + }, + { + "epoch": 3.6835826210826212, + "grad_norm": 1.0674567222595215, + "learning_rate": 3.0947938367275717e-06, + "loss": 0.8983, + "step": 20690 + }, + { + "epoch": 3.683760683760684, + "grad_norm": 0.9554956555366516, + "learning_rate": 3.091339419921868e-06, + "loss": 0.7877, + "step": 20691 + }, + { + "epoch": 3.6839387464387463, + "grad_norm": 0.8440012335777283, + "learning_rate": 3.0878869018400537e-06, + "loss": 0.5558, + "step": 20692 + }, + { + "epoch": 3.684116809116809, + "grad_norm": 0.9536983370780945, + "learning_rate": 3.084436282549774e-06, + "loss": 0.7293, + "step": 20693 + }, + { + "epoch": 3.684294871794872, + "grad_norm": 0.9408102631568909, + "learning_rate": 3.0809875621186534e-06, + "loss": 0.6177, + "step": 20694 + }, + { + "epoch": 3.6844729344729346, + "grad_norm": 0.9360013008117676, + "learning_rate": 3.0775407406142375e-06, + "loss": 0.7644, + "step": 20695 + }, + { + "epoch": 3.6846509971509973, + "grad_norm": 0.9667526483535767, + "learning_rate": 3.0740958181040836e-06, + "loss": 0.6533, + "step": 20696 + }, + { + "epoch": 3.6848290598290596, + "grad_norm": 1.0383641719818115, + "learning_rate": 3.0706527946556596e-06, + "loss": 0.6392, + "step": 20697 + }, + { + "epoch": 3.6850071225071224, + "grad_norm": 0.9858824014663696, + "learning_rate": 3.067211670336445e-06, + "loss": 0.7166, + "step": 20698 + }, + { + "epoch": 3.685185185185185, + "grad_norm": 0.8552738428115845, + "learning_rate": 3.063772445213864e-06, + "loss": 0.8228, + "step": 20699 + }, + { + "epoch": 3.685363247863248, + "grad_norm": 0.8895391821861267, + "learning_rate": 3.0603351193552954e-06, + "loss": 0.7124, + "step": 20700 + }, + { + "epoch": 3.6855413105413106, + "grad_norm": 1.1577317714691162, + "learning_rate": 3.0568996928280857e-06, + "loss": 0.6814, + "step": 20701 + }, + { + "epoch": 3.6857193732193734, + "grad_norm": 0.8977794647216797, + "learning_rate": 3.0534661656995366e-06, + "loss": 0.7366, + "step": 20702 + }, + { + "epoch": 3.685897435897436, + "grad_norm": 0.9750751852989197, + "learning_rate": 3.050034538036928e-06, + "loss": 0.924, + "step": 20703 + }, + { + "epoch": 3.6860754985754984, + "grad_norm": 0.9595991969108582, + "learning_rate": 3.0466048099075163e-06, + "loss": 0.8004, + "step": 20704 + }, + { + "epoch": 3.686253561253561, + "grad_norm": 0.9346702098846436, + "learning_rate": 3.0431769813784596e-06, + "loss": 0.928, + "step": 20705 + }, + { + "epoch": 3.686431623931624, + "grad_norm": 0.9379191994667053, + "learning_rate": 3.039751052516948e-06, + "loss": 0.7355, + "step": 20706 + }, + { + "epoch": 3.6866096866096867, + "grad_norm": 0.9948887825012207, + "learning_rate": 3.0363270233900844e-06, + "loss": 0.7923, + "step": 20707 + }, + { + "epoch": 3.6867877492877494, + "grad_norm": 0.9426677227020264, + "learning_rate": 3.0329048940649805e-06, + "loss": 0.844, + "step": 20708 + }, + { + "epoch": 3.6869658119658117, + "grad_norm": 0.957449197769165, + "learning_rate": 3.0294846646086726e-06, + "loss": 0.6105, + "step": 20709 + }, + { + "epoch": 3.6871438746438745, + "grad_norm": 1.0119540691375732, + "learning_rate": 3.0260663350881622e-06, + "loss": 0.8269, + "step": 20710 + }, + { + "epoch": 3.6873219373219372, + "grad_norm": 0.9362712502479553, + "learning_rate": 3.022649905570463e-06, + "loss": 0.7466, + "step": 20711 + }, + { + "epoch": 3.6875, + "grad_norm": 1.0067883729934692, + "learning_rate": 3.0192353761224647e-06, + "loss": 0.6891, + "step": 20712 + }, + { + "epoch": 3.6876780626780628, + "grad_norm": 0.9323202967643738, + "learning_rate": 3.015822746811092e-06, + "loss": 0.8198, + "step": 20713 + }, + { + "epoch": 3.6878561253561255, + "grad_norm": 1.1497761011123657, + "learning_rate": 3.0124120177032034e-06, + "loss": 0.632, + "step": 20714 + }, + { + "epoch": 3.6880341880341883, + "grad_norm": 0.9109611511230469, + "learning_rate": 3.0090031888656334e-06, + "loss": 0.8147, + "step": 20715 + }, + { + "epoch": 3.6882122507122506, + "grad_norm": 0.9743979573249817, + "learning_rate": 3.0055962603651733e-06, + "loss": 0.7787, + "step": 20716 + }, + { + "epoch": 3.6883903133903133, + "grad_norm": 0.8306061625480652, + "learning_rate": 3.0021912322685473e-06, + "loss": 0.6223, + "step": 20717 + }, + { + "epoch": 3.688568376068376, + "grad_norm": 1.0465481281280518, + "learning_rate": 2.9987881046425025e-06, + "loss": 0.7197, + "step": 20718 + }, + { + "epoch": 3.688746438746439, + "grad_norm": 1.0467299222946167, + "learning_rate": 2.9953868775536964e-06, + "loss": 0.7249, + "step": 20719 + }, + { + "epoch": 3.6889245014245016, + "grad_norm": 1.1018860340118408, + "learning_rate": 2.9919875510687644e-06, + "loss": 0.7112, + "step": 20720 + }, + { + "epoch": 3.689102564102564, + "grad_norm": 0.9819473624229431, + "learning_rate": 2.9885901252543425e-06, + "loss": 0.6801, + "step": 20721 + }, + { + "epoch": 3.6892806267806266, + "grad_norm": 1.0312379598617554, + "learning_rate": 2.985194600176955e-06, + "loss": 0.7441, + "step": 20722 + }, + { + "epoch": 3.6894586894586894, + "grad_norm": 0.9706970453262329, + "learning_rate": 2.9818009759031483e-06, + "loss": 0.7264, + "step": 20723 + }, + { + "epoch": 3.689636752136752, + "grad_norm": 0.8469666242599487, + "learning_rate": 2.9784092524994257e-06, + "loss": 0.6271, + "step": 20724 + }, + { + "epoch": 3.689814814814815, + "grad_norm": 0.8831173777580261, + "learning_rate": 2.9750194300322223e-06, + "loss": 0.5946, + "step": 20725 + }, + { + "epoch": 3.6899928774928776, + "grad_norm": 1.000719428062439, + "learning_rate": 2.9716315085679623e-06, + "loss": 0.8439, + "step": 20726 + }, + { + "epoch": 3.6901709401709404, + "grad_norm": 0.8967503309249878, + "learning_rate": 2.968245488173027e-06, + "loss": 0.6256, + "step": 20727 + }, + { + "epoch": 3.6903490028490027, + "grad_norm": 1.0560952425003052, + "learning_rate": 2.9648613689137404e-06, + "loss": 0.9563, + "step": 20728 + }, + { + "epoch": 3.6905270655270654, + "grad_norm": 0.9296253323554993, + "learning_rate": 2.9614791508564277e-06, + "loss": 0.8398, + "step": 20729 + }, + { + "epoch": 3.690705128205128, + "grad_norm": 0.8290141820907593, + "learning_rate": 2.9580988340673577e-06, + "loss": 0.5122, + "step": 20730 + }, + { + "epoch": 3.690883190883191, + "grad_norm": 0.9456691741943359, + "learning_rate": 2.954720418612755e-06, + "loss": 0.6461, + "step": 20731 + }, + { + "epoch": 3.6910612535612537, + "grad_norm": 0.9314257502555847, + "learning_rate": 2.9513439045588122e-06, + "loss": 0.755, + "step": 20732 + }, + { + "epoch": 3.691239316239316, + "grad_norm": 1.037325143814087, + "learning_rate": 2.947969291971686e-06, + "loss": 0.6729, + "step": 20733 + }, + { + "epoch": 3.6914173789173788, + "grad_norm": 0.9310669302940369, + "learning_rate": 2.9445965809174913e-06, + "loss": 0.7465, + "step": 20734 + }, + { + "epoch": 3.6915954415954415, + "grad_norm": 1.0100094079971313, + "learning_rate": 2.941225771462297e-06, + "loss": 0.8561, + "step": 20735 + }, + { + "epoch": 3.6917735042735043, + "grad_norm": 1.0392314195632935, + "learning_rate": 2.9378568636721835e-06, + "loss": 0.9093, + "step": 20736 + }, + { + "epoch": 3.691951566951567, + "grad_norm": 1.1194417476654053, + "learning_rate": 2.934489857613121e-06, + "loss": 0.8177, + "step": 20737 + }, + { + "epoch": 3.6921296296296298, + "grad_norm": 0.9694803357124329, + "learning_rate": 2.9311247533511e-06, + "loss": 0.7652, + "step": 20738 + }, + { + "epoch": 3.6923076923076925, + "grad_norm": 0.9330110549926758, + "learning_rate": 2.927761550952046e-06, + "loss": 0.8571, + "step": 20739 + }, + { + "epoch": 3.692485754985755, + "grad_norm": 0.9791732430458069, + "learning_rate": 2.9244002504818514e-06, + "loss": 0.9147, + "step": 20740 + }, + { + "epoch": 3.6926638176638176, + "grad_norm": 0.9518002271652222, + "learning_rate": 2.921040852006385e-06, + "loss": 0.7033, + "step": 20741 + }, + { + "epoch": 3.6928418803418803, + "grad_norm": 0.9248731732368469, + "learning_rate": 2.9176833555914495e-06, + "loss": 0.7099, + "step": 20742 + }, + { + "epoch": 3.693019943019943, + "grad_norm": 1.0054923295974731, + "learning_rate": 2.9143277613028486e-06, + "loss": 0.7161, + "step": 20743 + }, + { + "epoch": 3.693198005698006, + "grad_norm": 1.0332942008972168, + "learning_rate": 2.910974069206307e-06, + "loss": 0.7649, + "step": 20744 + }, + { + "epoch": 3.693376068376068, + "grad_norm": 0.9899537563323975, + "learning_rate": 2.907622279367539e-06, + "loss": 0.7775, + "step": 20745 + }, + { + "epoch": 3.693554131054131, + "grad_norm": 1.0299023389816284, + "learning_rate": 2.9042723918522365e-06, + "loss": 0.7575, + "step": 20746 + }, + { + "epoch": 3.6937321937321936, + "grad_norm": 0.9683845043182373, + "learning_rate": 2.9009244067260024e-06, + "loss": 0.8563, + "step": 20747 + }, + { + "epoch": 3.6939102564102564, + "grad_norm": 0.9879896640777588, + "learning_rate": 2.897578324054451e-06, + "loss": 0.9879, + "step": 20748 + }, + { + "epoch": 3.694088319088319, + "grad_norm": 0.9869822859764099, + "learning_rate": 2.8942341439031405e-06, + "loss": 0.804, + "step": 20749 + }, + { + "epoch": 3.694266381766382, + "grad_norm": 0.9771765470504761, + "learning_rate": 2.8908918663375863e-06, + "loss": 0.9977, + "step": 20750 + }, + { + "epoch": 3.6944444444444446, + "grad_norm": 0.8605872988700867, + "learning_rate": 2.8875514914233013e-06, + "loss": 0.8474, + "step": 20751 + }, + { + "epoch": 3.6946225071225074, + "grad_norm": 0.9646205902099609, + "learning_rate": 2.88421301922569e-06, + "loss": 0.6685, + "step": 20752 + }, + { + "epoch": 3.6948005698005697, + "grad_norm": 0.9683958888053894, + "learning_rate": 2.880876449810199e-06, + "loss": 0.965, + "step": 20753 + }, + { + "epoch": 3.6949786324786325, + "grad_norm": 0.8820703029632568, + "learning_rate": 2.877541783242177e-06, + "loss": 0.5889, + "step": 20754 + }, + { + "epoch": 3.695156695156695, + "grad_norm": 0.9471021890640259, + "learning_rate": 2.8742090195869707e-06, + "loss": 0.6262, + "step": 20755 + }, + { + "epoch": 3.695334757834758, + "grad_norm": 0.9770998358726501, + "learning_rate": 2.870878158909884e-06, + "loss": 0.8444, + "step": 20756 + }, + { + "epoch": 3.6955128205128203, + "grad_norm": 0.8741800785064697, + "learning_rate": 2.867549201276165e-06, + "loss": 0.6484, + "step": 20757 + }, + { + "epoch": 3.695690883190883, + "grad_norm": 0.9649479985237122, + "learning_rate": 2.864222146751039e-06, + "loss": 0.6613, + "step": 20758 + }, + { + "epoch": 3.6958689458689458, + "grad_norm": 0.9703496694564819, + "learning_rate": 2.8608969953997087e-06, + "loss": 0.6868, + "step": 20759 + }, + { + "epoch": 3.6960470085470085, + "grad_norm": 0.9378167986869812, + "learning_rate": 2.857573747287301e-06, + "loss": 0.6265, + "step": 20760 + }, + { + "epoch": 3.6962250712250713, + "grad_norm": 0.9149529337882996, + "learning_rate": 2.8542524024789517e-06, + "loss": 0.5854, + "step": 20761 + }, + { + "epoch": 3.696403133903134, + "grad_norm": 0.9306116700172424, + "learning_rate": 2.8509329610397095e-06, + "loss": 0.795, + "step": 20762 + }, + { + "epoch": 3.6965811965811968, + "grad_norm": 0.9093084335327148, + "learning_rate": 2.8476154230346443e-06, + "loss": 0.7913, + "step": 20763 + }, + { + "epoch": 3.6967592592592595, + "grad_norm": 1.0166517496109009, + "learning_rate": 2.844299788528726e-06, + "loss": 0.739, + "step": 20764 + }, + { + "epoch": 3.696937321937322, + "grad_norm": 0.8517898917198181, + "learning_rate": 2.8409860575869364e-06, + "loss": 0.7489, + "step": 20765 + }, + { + "epoch": 3.6971153846153846, + "grad_norm": 1.0916438102722168, + "learning_rate": 2.83767423027419e-06, + "loss": 0.8086, + "step": 20766 + }, + { + "epoch": 3.6972934472934473, + "grad_norm": 1.061730980873108, + "learning_rate": 2.8343643066553793e-06, + "loss": 0.5334, + "step": 20767 + }, + { + "epoch": 3.69747150997151, + "grad_norm": 1.100048542022705, + "learning_rate": 2.831056286795364e-06, + "loss": 0.8805, + "step": 20768 + }, + { + "epoch": 3.6976495726495724, + "grad_norm": 0.6898918747901917, + "learning_rate": 2.827750170758936e-06, + "loss": 0.4459, + "step": 20769 + }, + { + "epoch": 3.697827635327635, + "grad_norm": 0.8306655883789062, + "learning_rate": 2.8244459586108997e-06, + "loss": 0.7836, + "step": 20770 + }, + { + "epoch": 3.698005698005698, + "grad_norm": 1.0033217668533325, + "learning_rate": 2.8211436504159695e-06, + "loss": 0.8711, + "step": 20771 + }, + { + "epoch": 3.6981837606837606, + "grad_norm": 1.0434397459030151, + "learning_rate": 2.8178432462388602e-06, + "loss": 0.7871, + "step": 20772 + }, + { + "epoch": 3.6983618233618234, + "grad_norm": 0.9520493149757385, + "learning_rate": 2.814544746144243e-06, + "loss": 0.9114, + "step": 20773 + }, + { + "epoch": 3.698539886039886, + "grad_norm": 0.9986703991889954, + "learning_rate": 2.8112481501967326e-06, + "loss": 0.8862, + "step": 20774 + }, + { + "epoch": 3.698717948717949, + "grad_norm": 0.9036804437637329, + "learning_rate": 2.80795345846091e-06, + "loss": 0.7594, + "step": 20775 + }, + { + "epoch": 3.6988960113960117, + "grad_norm": 0.9680522084236145, + "learning_rate": 2.804660671001358e-06, + "loss": 0.8734, + "step": 20776 + }, + { + "epoch": 3.699074074074074, + "grad_norm": 0.9929130673408508, + "learning_rate": 2.8013697878825573e-06, + "loss": 0.8834, + "step": 20777 + }, + { + "epoch": 3.6992521367521367, + "grad_norm": 0.9345636367797852, + "learning_rate": 2.7980808091690234e-06, + "loss": 0.5392, + "step": 20778 + }, + { + "epoch": 3.6994301994301995, + "grad_norm": 0.9011038541793823, + "learning_rate": 2.7947937349251495e-06, + "loss": 0.6209, + "step": 20779 + }, + { + "epoch": 3.699608262108262, + "grad_norm": 1.0361058712005615, + "learning_rate": 2.791508565215384e-06, + "loss": 0.8411, + "step": 20780 + }, + { + "epoch": 3.6997863247863245, + "grad_norm": 0.8469780087471008, + "learning_rate": 2.7882253001040746e-06, + "loss": 0.5842, + "step": 20781 + }, + { + "epoch": 3.6999643874643873, + "grad_norm": 0.9913833141326904, + "learning_rate": 2.784943939655549e-06, + "loss": 0.7048, + "step": 20782 + }, + { + "epoch": 3.70014245014245, + "grad_norm": 0.6060404777526855, + "learning_rate": 2.781664483934099e-06, + "loss": 0.3919, + "step": 20783 + }, + { + "epoch": 3.7003205128205128, + "grad_norm": 0.9720197319984436, + "learning_rate": 2.7783869330039847e-06, + "loss": 0.6759, + "step": 20784 + }, + { + "epoch": 3.7004985754985755, + "grad_norm": 0.9591602087020874, + "learning_rate": 2.7751112869293993e-06, + "loss": 0.9101, + "step": 20785 + }, + { + "epoch": 3.7006766381766383, + "grad_norm": 1.0980218648910522, + "learning_rate": 2.7718375457745693e-06, + "loss": 0.7299, + "step": 20786 + }, + { + "epoch": 3.700854700854701, + "grad_norm": 0.9733334183692932, + "learning_rate": 2.7685657096035877e-06, + "loss": 0.9226, + "step": 20787 + }, + { + "epoch": 3.701032763532764, + "grad_norm": 1.0194134712219238, + "learning_rate": 2.7652957784805923e-06, + "loss": 0.7872, + "step": 20788 + }, + { + "epoch": 3.701210826210826, + "grad_norm": 0.893657922744751, + "learning_rate": 2.7620277524696313e-06, + "loss": 0.7858, + "step": 20789 + }, + { + "epoch": 3.701388888888889, + "grad_norm": 1.019237995147705, + "learning_rate": 2.758761631634743e-06, + "loss": 0.668, + "step": 20790 + }, + { + "epoch": 3.7015669515669516, + "grad_norm": 0.9971436858177185, + "learning_rate": 2.7554974160399203e-06, + "loss": 0.7905, + "step": 20791 + }, + { + "epoch": 3.7017450142450143, + "grad_norm": 1.0228685140609741, + "learning_rate": 2.7522351057491124e-06, + "loss": 0.8796, + "step": 20792 + }, + { + "epoch": 3.7019230769230766, + "grad_norm": 0.8594250082969666, + "learning_rate": 2.748974700826257e-06, + "loss": 0.7012, + "step": 20793 + }, + { + "epoch": 3.7021011396011394, + "grad_norm": 1.0440367460250854, + "learning_rate": 2.7457162013352023e-06, + "loss": 0.8736, + "step": 20794 + }, + { + "epoch": 3.702279202279202, + "grad_norm": 0.9900957942008972, + "learning_rate": 2.74245960733982e-06, + "loss": 0.7699, + "step": 20795 + }, + { + "epoch": 3.702457264957265, + "grad_norm": 0.8492827415466309, + "learning_rate": 2.739204918903915e-06, + "loss": 0.6039, + "step": 20796 + }, + { + "epoch": 3.7026353276353277, + "grad_norm": 0.9645397663116455, + "learning_rate": 2.7359521360912354e-06, + "loss": 0.7195, + "step": 20797 + }, + { + "epoch": 3.7028133903133904, + "grad_norm": 0.9234998226165771, + "learning_rate": 2.732701258965531e-06, + "loss": 0.6246, + "step": 20798 + }, + { + "epoch": 3.702991452991453, + "grad_norm": 0.9107007384300232, + "learning_rate": 2.7294522875904947e-06, + "loss": 0.6909, + "step": 20799 + }, + { + "epoch": 3.703169515669516, + "grad_norm": 1.1360125541687012, + "learning_rate": 2.726205222029754e-06, + "loss": 0.6849, + "step": 20800 + }, + { + "epoch": 3.703347578347578, + "grad_norm": 0.8517076969146729, + "learning_rate": 2.722960062346991e-06, + "loss": 0.6815, + "step": 20801 + }, + { + "epoch": 3.703525641025641, + "grad_norm": 1.07355535030365, + "learning_rate": 2.7197168086057213e-06, + "loss": 0.6606, + "step": 20802 + }, + { + "epoch": 3.7037037037037037, + "grad_norm": 0.8663332462310791, + "learning_rate": 2.7164754608695387e-06, + "loss": 0.6732, + "step": 20803 + }, + { + "epoch": 3.7038817663817665, + "grad_norm": 0.8558155298233032, + "learning_rate": 2.7132360192019035e-06, + "loss": 0.6664, + "step": 20804 + }, + { + "epoch": 3.7040598290598292, + "grad_norm": 1.0739531517028809, + "learning_rate": 2.7099984836663316e-06, + "loss": 0.8237, + "step": 20805 + }, + { + "epoch": 3.7042378917378915, + "grad_norm": 0.9637136459350586, + "learning_rate": 2.706762854326239e-06, + "loss": 0.9264, + "step": 20806 + }, + { + "epoch": 3.7044159544159543, + "grad_norm": 0.8434587121009827, + "learning_rate": 2.7035291312450084e-06, + "loss": 0.6871, + "step": 20807 + }, + { + "epoch": 3.704594017094017, + "grad_norm": 0.9097843170166016, + "learning_rate": 2.700297314486022e-06, + "loss": 0.7741, + "step": 20808 + }, + { + "epoch": 3.70477207977208, + "grad_norm": 0.9461157321929932, + "learning_rate": 2.6970674041125744e-06, + "loss": 0.7223, + "step": 20809 + }, + { + "epoch": 3.7049501424501425, + "grad_norm": 0.799718976020813, + "learning_rate": 2.6938394001879695e-06, + "loss": 0.6793, + "step": 20810 + }, + { + "epoch": 3.7051282051282053, + "grad_norm": 0.9917070865631104, + "learning_rate": 2.6906133027754353e-06, + "loss": 0.8659, + "step": 20811 + }, + { + "epoch": 3.705306267806268, + "grad_norm": 0.8843945264816284, + "learning_rate": 2.687389111938199e-06, + "loss": 0.761, + "step": 20812 + }, + { + "epoch": 3.7054843304843303, + "grad_norm": 1.0072822570800781, + "learning_rate": 2.6841668277394315e-06, + "loss": 0.665, + "step": 20813 + }, + { + "epoch": 3.705662393162393, + "grad_norm": 1.0159565210342407, + "learning_rate": 2.68094645024225e-06, + "loss": 0.8305, + "step": 20814 + }, + { + "epoch": 3.705840455840456, + "grad_norm": 1.0532197952270508, + "learning_rate": 2.6777279795097586e-06, + "loss": 0.7869, + "step": 20815 + }, + { + "epoch": 3.7060185185185186, + "grad_norm": 1.0152393579483032, + "learning_rate": 2.6745114156050187e-06, + "loss": 0.6795, + "step": 20816 + }, + { + "epoch": 3.7061965811965814, + "grad_norm": 0.8705698847770691, + "learning_rate": 2.6712967585910465e-06, + "loss": 0.5973, + "step": 20817 + }, + { + "epoch": 3.7063746438746437, + "grad_norm": 0.9337679147720337, + "learning_rate": 2.668084008530847e-06, + "loss": 0.745, + "step": 20818 + }, + { + "epoch": 3.7065527065527064, + "grad_norm": 1.0119414329528809, + "learning_rate": 2.664873165487336e-06, + "loss": 0.8467, + "step": 20819 + }, + { + "epoch": 3.706730769230769, + "grad_norm": 1.0212607383728027, + "learning_rate": 2.661664229523442e-06, + "loss": 0.6991, + "step": 20820 + }, + { + "epoch": 3.706908831908832, + "grad_norm": 0.6226426362991333, + "learning_rate": 2.658457200702047e-06, + "loss": 0.2538, + "step": 20821 + }, + { + "epoch": 3.7070868945868947, + "grad_norm": 1.035329818725586, + "learning_rate": 2.655252079085957e-06, + "loss": 0.801, + "step": 20822 + }, + { + "epoch": 3.7072649572649574, + "grad_norm": 0.9035121202468872, + "learning_rate": 2.6520488647379995e-06, + "loss": 0.7865, + "step": 20823 + }, + { + "epoch": 3.70744301994302, + "grad_norm": 0.8977579474449158, + "learning_rate": 2.6488475577209127e-06, + "loss": 0.8943, + "step": 20824 + }, + { + "epoch": 3.7076210826210825, + "grad_norm": 1.0442665815353394, + "learning_rate": 2.6456481580974356e-06, + "loss": 0.8296, + "step": 20825 + }, + { + "epoch": 3.7077991452991452, + "grad_norm": 0.9925961494445801, + "learning_rate": 2.6424506659302406e-06, + "loss": 0.8141, + "step": 20826 + }, + { + "epoch": 3.707977207977208, + "grad_norm": 0.8873761892318726, + "learning_rate": 2.639255081281977e-06, + "loss": 0.7006, + "step": 20827 + }, + { + "epoch": 3.7081552706552707, + "grad_norm": 0.9560503363609314, + "learning_rate": 2.6360614042152732e-06, + "loss": 0.8785, + "step": 20828 + }, + { + "epoch": 3.7083333333333335, + "grad_norm": 0.7737786173820496, + "learning_rate": 2.6328696347926784e-06, + "loss": 0.5948, + "step": 20829 + }, + { + "epoch": 3.708511396011396, + "grad_norm": 1.1564871072769165, + "learning_rate": 2.629679773076754e-06, + "loss": 0.7169, + "step": 20830 + }, + { + "epoch": 3.7086894586894585, + "grad_norm": 0.8807209730148315, + "learning_rate": 2.6264918191299724e-06, + "loss": 0.5954, + "step": 20831 + }, + { + "epoch": 3.7088675213675213, + "grad_norm": 0.9065495133399963, + "learning_rate": 2.6233057730148057e-06, + "loss": 0.7257, + "step": 20832 + }, + { + "epoch": 3.709045584045584, + "grad_norm": 0.9244785308837891, + "learning_rate": 2.6201216347937042e-06, + "loss": 0.7843, + "step": 20833 + }, + { + "epoch": 3.709223646723647, + "grad_norm": 1.0104621648788452, + "learning_rate": 2.6169394045290065e-06, + "loss": 0.7134, + "step": 20834 + }, + { + "epoch": 3.7094017094017095, + "grad_norm": 0.9149969816207886, + "learning_rate": 2.613759082283085e-06, + "loss": 0.6389, + "step": 20835 + }, + { + "epoch": 3.7095797720797723, + "grad_norm": 1.1129883527755737, + "learning_rate": 2.6105806681182676e-06, + "loss": 0.7805, + "step": 20836 + }, + { + "epoch": 3.7097578347578346, + "grad_norm": 0.9657980799674988, + "learning_rate": 2.6074041620968047e-06, + "loss": 0.8758, + "step": 20837 + }, + { + "epoch": 3.7099358974358974, + "grad_norm": 1.2004457712173462, + "learning_rate": 2.6042295642809355e-06, + "loss": 0.8944, + "step": 20838 + }, + { + "epoch": 3.71011396011396, + "grad_norm": 1.0154633522033691, + "learning_rate": 2.6010568747328766e-06, + "loss": 1.0198, + "step": 20839 + }, + { + "epoch": 3.710292022792023, + "grad_norm": 0.908183753490448, + "learning_rate": 2.5978860935147788e-06, + "loss": 0.7449, + "step": 20840 + }, + { + "epoch": 3.7104700854700856, + "grad_norm": 1.012807846069336, + "learning_rate": 2.594717220688758e-06, + "loss": 0.793, + "step": 20841 + }, + { + "epoch": 3.710648148148148, + "grad_norm": 1.0019315481185913, + "learning_rate": 2.5915502563169105e-06, + "loss": 0.7707, + "step": 20842 + }, + { + "epoch": 3.7108262108262107, + "grad_norm": 1.0038162469863892, + "learning_rate": 2.5883852004613074e-06, + "loss": 0.7959, + "step": 20843 + }, + { + "epoch": 3.7110042735042734, + "grad_norm": 1.0344631671905518, + "learning_rate": 2.5852220531839113e-06, + "loss": 0.7611, + "step": 20844 + }, + { + "epoch": 3.711182336182336, + "grad_norm": 1.0860040187835693, + "learning_rate": 2.5820608145467494e-06, + "loss": 0.6978, + "step": 20845 + }, + { + "epoch": 3.711360398860399, + "grad_norm": 0.8946214318275452, + "learning_rate": 2.5789014846117288e-06, + "loss": 0.834, + "step": 20846 + }, + { + "epoch": 3.7115384615384617, + "grad_norm": 0.8077123761177063, + "learning_rate": 2.575744063440755e-06, + "loss": 0.6498, + "step": 20847 + }, + { + "epoch": 3.7117165242165244, + "grad_norm": 0.9708347916603088, + "learning_rate": 2.5725885510957116e-06, + "loss": 0.7118, + "step": 20848 + }, + { + "epoch": 3.7118945868945867, + "grad_norm": 0.8636693358421326, + "learning_rate": 2.5694349476383826e-06, + "loss": 0.6884, + "step": 20849 + }, + { + "epoch": 3.7120726495726495, + "grad_norm": 1.106953740119934, + "learning_rate": 2.566283253130608e-06, + "loss": 0.6778, + "step": 20850 + }, + { + "epoch": 3.7122507122507122, + "grad_norm": 1.0069822072982788, + "learning_rate": 2.5631334676340936e-06, + "loss": 0.7222, + "step": 20851 + }, + { + "epoch": 3.712428774928775, + "grad_norm": 0.8577214479446411, + "learning_rate": 2.5599855912105787e-06, + "loss": 0.7543, + "step": 20852 + }, + { + "epoch": 3.7126068376068377, + "grad_norm": 0.9592446684837341, + "learning_rate": 2.5568396239217363e-06, + "loss": 0.7331, + "step": 20853 + }, + { + "epoch": 3.7127849002849, + "grad_norm": 0.9552914500236511, + "learning_rate": 2.553695565829195e-06, + "loss": 0.8148, + "step": 20854 + }, + { + "epoch": 3.712962962962963, + "grad_norm": 1.1876866817474365, + "learning_rate": 2.550553416994561e-06, + "loss": 0.9698, + "step": 20855 + }, + { + "epoch": 3.7131410256410255, + "grad_norm": 0.985346257686615, + "learning_rate": 2.5474131774794073e-06, + "loss": 0.8515, + "step": 20856 + }, + { + "epoch": 3.7133190883190883, + "grad_norm": 0.8775020837783813, + "learning_rate": 2.5442748473452293e-06, + "loss": 0.6048, + "step": 20857 + }, + { + "epoch": 3.713497150997151, + "grad_norm": 1.0920336246490479, + "learning_rate": 2.5411384266535663e-06, + "loss": 0.8886, + "step": 20858 + }, + { + "epoch": 3.713675213675214, + "grad_norm": 0.960750937461853, + "learning_rate": 2.5380039154658253e-06, + "loss": 0.5685, + "step": 20859 + }, + { + "epoch": 3.7138532763532766, + "grad_norm": 0.9069384336471558, + "learning_rate": 2.5348713138434564e-06, + "loss": 0.7219, + "step": 20860 + }, + { + "epoch": 3.7140313390313393, + "grad_norm": 0.8939675688743591, + "learning_rate": 2.5317406218478e-06, + "loss": 0.6803, + "step": 20861 + }, + { + "epoch": 3.7142094017094016, + "grad_norm": 0.9675079584121704, + "learning_rate": 2.5286118395402182e-06, + "loss": 0.8229, + "step": 20862 + }, + { + "epoch": 3.7143874643874644, + "grad_norm": 0.8619394898414612, + "learning_rate": 2.5254849669820056e-06, + "loss": 0.771, + "step": 20863 + }, + { + "epoch": 3.714565527065527, + "grad_norm": 1.0089298486709595, + "learning_rate": 2.5223600042344362e-06, + "loss": 0.798, + "step": 20864 + }, + { + "epoch": 3.71474358974359, + "grad_norm": 1.3646090030670166, + "learning_rate": 2.5192369513587276e-06, + "loss": 0.7596, + "step": 20865 + }, + { + "epoch": 3.714921652421652, + "grad_norm": 0.950450599193573, + "learning_rate": 2.5161158084160754e-06, + "loss": 0.6728, + "step": 20866 + }, + { + "epoch": 3.715099715099715, + "grad_norm": 1.071611762046814, + "learning_rate": 2.512996575467608e-06, + "loss": 0.6569, + "step": 20867 + }, + { + "epoch": 3.7152777777777777, + "grad_norm": 0.8743181228637695, + "learning_rate": 2.5098792525744874e-06, + "loss": 0.7567, + "step": 20868 + }, + { + "epoch": 3.7154558404558404, + "grad_norm": 0.8863722085952759, + "learning_rate": 2.5067638397977544e-06, + "loss": 0.6423, + "step": 20869 + }, + { + "epoch": 3.715633903133903, + "grad_norm": 0.7184516191482544, + "learning_rate": 2.50365033719846e-06, + "loss": 0.347, + "step": 20870 + }, + { + "epoch": 3.715811965811966, + "grad_norm": 0.9171306490898132, + "learning_rate": 2.5005387448375995e-06, + "loss": 0.7904, + "step": 20871 + }, + { + "epoch": 3.7159900284900287, + "grad_norm": 1.0704946517944336, + "learning_rate": 2.4974290627761466e-06, + "loss": 0.9393, + "step": 20872 + }, + { + "epoch": 3.7161680911680914, + "grad_norm": 0.9949630498886108, + "learning_rate": 2.4943212910750414e-06, + "loss": 0.646, + "step": 20873 + }, + { + "epoch": 3.7163461538461537, + "grad_norm": 0.8926770687103271, + "learning_rate": 2.491215429795146e-06, + "loss": 0.8417, + "step": 20874 + }, + { + "epoch": 3.7165242165242165, + "grad_norm": 1.0397453308105469, + "learning_rate": 2.4881114789973346e-06, + "loss": 0.7087, + "step": 20875 + }, + { + "epoch": 3.7167022792022792, + "grad_norm": 0.8475403189659119, + "learning_rate": 2.485009438742414e-06, + "loss": 0.6031, + "step": 20876 + }, + { + "epoch": 3.716880341880342, + "grad_norm": 0.8874244093894958, + "learning_rate": 2.481909309091157e-06, + "loss": 0.8112, + "step": 20877 + }, + { + "epoch": 3.7170584045584043, + "grad_norm": 0.9252071380615234, + "learning_rate": 2.478811090104316e-06, + "loss": 0.6202, + "step": 20878 + }, + { + "epoch": 3.717236467236467, + "grad_norm": 0.8514638543128967, + "learning_rate": 2.4757147818425865e-06, + "loss": 0.656, + "step": 20879 + }, + { + "epoch": 3.71741452991453, + "grad_norm": 0.942911684513092, + "learning_rate": 2.4726203843666417e-06, + "loss": 0.7602, + "step": 20880 + }, + { + "epoch": 3.7175925925925926, + "grad_norm": 0.8529558777809143, + "learning_rate": 2.4695278977371005e-06, + "loss": 0.8576, + "step": 20881 + }, + { + "epoch": 3.7177706552706553, + "grad_norm": 0.8798624873161316, + "learning_rate": 2.4664373220145587e-06, + "loss": 0.6588, + "step": 20882 + }, + { + "epoch": 3.717948717948718, + "grad_norm": 0.9501121640205383, + "learning_rate": 2.4633486572595787e-06, + "loss": 0.7127, + "step": 20883 + }, + { + "epoch": 3.718126780626781, + "grad_norm": 0.9304238557815552, + "learning_rate": 2.4602619035326456e-06, + "loss": 0.7881, + "step": 20884 + }, + { + "epoch": 3.7183048433048436, + "grad_norm": 1.0175336599349976, + "learning_rate": 2.457177060894289e-06, + "loss": 0.7735, + "step": 20885 + }, + { + "epoch": 3.718482905982906, + "grad_norm": 0.9424969553947449, + "learning_rate": 2.4540941294048937e-06, + "loss": 0.6261, + "step": 20886 + }, + { + "epoch": 3.7186609686609686, + "grad_norm": 0.9643171429634094, + "learning_rate": 2.4510131091249e-06, + "loss": 0.7912, + "step": 20887 + }, + { + "epoch": 3.7188390313390314, + "grad_norm": 0.9775816798210144, + "learning_rate": 2.44793400011466e-06, + "loss": 0.754, + "step": 20888 + }, + { + "epoch": 3.719017094017094, + "grad_norm": 0.8465856909751892, + "learning_rate": 2.4448568024345032e-06, + "loss": 0.6311, + "step": 20889 + }, + { + "epoch": 3.7191951566951564, + "grad_norm": 0.9722816348075867, + "learning_rate": 2.4417815161447367e-06, + "loss": 0.741, + "step": 20890 + }, + { + "epoch": 3.719373219373219, + "grad_norm": 0.9861366748809814, + "learning_rate": 2.4387081413055903e-06, + "loss": 0.889, + "step": 20891 + }, + { + "epoch": 3.719551282051282, + "grad_norm": 0.9273673892021179, + "learning_rate": 2.4356366779773045e-06, + "loss": 0.7996, + "step": 20892 + }, + { + "epoch": 3.7197293447293447, + "grad_norm": 0.9955036640167236, + "learning_rate": 2.432567126220031e-06, + "loss": 0.9323, + "step": 20893 + }, + { + "epoch": 3.7199074074074074, + "grad_norm": 1.1071380376815796, + "learning_rate": 2.4294994860939337e-06, + "loss": 0.7601, + "step": 20894 + }, + { + "epoch": 3.72008547008547, + "grad_norm": 0.9730632901191711, + "learning_rate": 2.426433757659108e-06, + "loss": 0.7674, + "step": 20895 + }, + { + "epoch": 3.720263532763533, + "grad_norm": 0.8447731733322144, + "learning_rate": 2.4233699409756284e-06, + "loss": 0.5593, + "step": 20896 + }, + { + "epoch": 3.7204415954415957, + "grad_norm": 1.0535780191421509, + "learning_rate": 2.4203080361035136e-06, + "loss": 0.7876, + "step": 20897 + }, + { + "epoch": 3.720619658119658, + "grad_norm": 1.0676836967468262, + "learning_rate": 2.4172480431027487e-06, + "loss": 0.7371, + "step": 20898 + }, + { + "epoch": 3.7207977207977208, + "grad_norm": 0.920718789100647, + "learning_rate": 2.4141899620332976e-06, + "loss": 0.6973, + "step": 20899 + }, + { + "epoch": 3.7209757834757835, + "grad_norm": 0.9319782257080078, + "learning_rate": 2.4111337929551002e-06, + "loss": 0.7951, + "step": 20900 + }, + { + "epoch": 3.7211538461538463, + "grad_norm": 0.9919620752334595, + "learning_rate": 2.408079535927987e-06, + "loss": 0.7875, + "step": 20901 + }, + { + "epoch": 3.7213319088319086, + "grad_norm": 0.9960176348686218, + "learning_rate": 2.4050271910118325e-06, + "loss": 0.8002, + "step": 20902 + }, + { + "epoch": 3.7215099715099713, + "grad_norm": 0.9272335171699524, + "learning_rate": 2.401976758266433e-06, + "loss": 0.6174, + "step": 20903 + }, + { + "epoch": 3.721688034188034, + "grad_norm": 0.9436232447624207, + "learning_rate": 2.3989282377515633e-06, + "loss": 0.5779, + "step": 20904 + }, + { + "epoch": 3.721866096866097, + "grad_norm": 0.8671663999557495, + "learning_rate": 2.3958816295269417e-06, + "loss": 0.6888, + "step": 20905 + }, + { + "epoch": 3.7220441595441596, + "grad_norm": 1.175491213798523, + "learning_rate": 2.3928369336522647e-06, + "loss": 0.7466, + "step": 20906 + }, + { + "epoch": 3.7222222222222223, + "grad_norm": 0.9122247695922852, + "learning_rate": 2.3897941501871855e-06, + "loss": 0.8985, + "step": 20907 + }, + { + "epoch": 3.722400284900285, + "grad_norm": 1.0453401803970337, + "learning_rate": 2.386753279191323e-06, + "loss": 0.7194, + "step": 20908 + }, + { + "epoch": 3.722578347578348, + "grad_norm": 0.938706636428833, + "learning_rate": 2.383714320724262e-06, + "loss": 0.6207, + "step": 20909 + }, + { + "epoch": 3.72275641025641, + "grad_norm": 0.9452419281005859, + "learning_rate": 2.380677274845533e-06, + "loss": 0.8168, + "step": 20910 + }, + { + "epoch": 3.722934472934473, + "grad_norm": 1.0109187364578247, + "learning_rate": 2.3776421416146556e-06, + "loss": 0.6572, + "step": 20911 + }, + { + "epoch": 3.7231125356125356, + "grad_norm": 0.9023509621620178, + "learning_rate": 2.3746089210910816e-06, + "loss": 0.7597, + "step": 20912 + }, + { + "epoch": 3.7232905982905984, + "grad_norm": 1.0462290048599243, + "learning_rate": 2.3715776133342416e-06, + "loss": 0.9093, + "step": 20913 + }, + { + "epoch": 3.7234686609686607, + "grad_norm": 0.9403861165046692, + "learning_rate": 2.368548218403521e-06, + "loss": 0.7762, + "step": 20914 + }, + { + "epoch": 3.7236467236467234, + "grad_norm": 0.8098660111427307, + "learning_rate": 2.3655207363583177e-06, + "loss": 0.515, + "step": 20915 + }, + { + "epoch": 3.723824786324786, + "grad_norm": 1.1157571077346802, + "learning_rate": 2.362495167257894e-06, + "loss": 0.7646, + "step": 20916 + }, + { + "epoch": 3.724002849002849, + "grad_norm": 0.8967264890670776, + "learning_rate": 2.3594715111615594e-06, + "loss": 0.6829, + "step": 20917 + }, + { + "epoch": 3.7241809116809117, + "grad_norm": 1.0004037618637085, + "learning_rate": 2.3564497681285547e-06, + "loss": 0.9033, + "step": 20918 + }, + { + "epoch": 3.7243589743589745, + "grad_norm": 0.9404234290122986, + "learning_rate": 2.353429938218077e-06, + "loss": 0.7891, + "step": 20919 + }, + { + "epoch": 3.724537037037037, + "grad_norm": 0.9244145750999451, + "learning_rate": 2.35041202148929e-06, + "loss": 0.8618, + "step": 20920 + }, + { + "epoch": 3.7247150997151, + "grad_norm": 0.9113515615463257, + "learning_rate": 2.3473960180013353e-06, + "loss": 0.6614, + "step": 20921 + }, + { + "epoch": 3.7248931623931623, + "grad_norm": 1.0383996963500977, + "learning_rate": 2.3443819278132996e-06, + "loss": 0.8602, + "step": 20922 + }, + { + "epoch": 3.725071225071225, + "grad_norm": 0.9273353815078735, + "learning_rate": 2.341369750984246e-06, + "loss": 0.734, + "step": 20923 + }, + { + "epoch": 3.7252492877492878, + "grad_norm": 0.8500500321388245, + "learning_rate": 2.3383594875731605e-06, + "loss": 0.7722, + "step": 20924 + }, + { + "epoch": 3.7254273504273505, + "grad_norm": 0.9240385890007019, + "learning_rate": 2.3353511376390747e-06, + "loss": 0.6948, + "step": 20925 + }, + { + "epoch": 3.7256054131054133, + "grad_norm": 1.025537371635437, + "learning_rate": 2.332344701240885e-06, + "loss": 0.8939, + "step": 20926 + }, + { + "epoch": 3.7257834757834756, + "grad_norm": 0.9603112936019897, + "learning_rate": 2.3293401784375223e-06, + "loss": 0.9551, + "step": 20927 + }, + { + "epoch": 3.7259615384615383, + "grad_norm": 0.9035696983337402, + "learning_rate": 2.32633756928784e-06, + "loss": 0.7112, + "step": 20928 + }, + { + "epoch": 3.726139601139601, + "grad_norm": 0.9449596405029297, + "learning_rate": 2.3233368738506677e-06, + "loss": 0.8199, + "step": 20929 + }, + { + "epoch": 3.726317663817664, + "grad_norm": 1.058143138885498, + "learning_rate": 2.320338092184826e-06, + "loss": 0.8258, + "step": 20930 + }, + { + "epoch": 3.7264957264957266, + "grad_norm": 1.0171890258789062, + "learning_rate": 2.317341224349023e-06, + "loss": 0.991, + "step": 20931 + }, + { + "epoch": 3.7266737891737893, + "grad_norm": 1.008941888809204, + "learning_rate": 2.3143462704020124e-06, + "loss": 0.8857, + "step": 20932 + }, + { + "epoch": 3.726851851851852, + "grad_norm": 0.9405086636543274, + "learning_rate": 2.3113532304024575e-06, + "loss": 0.7347, + "step": 20933 + }, + { + "epoch": 3.7270299145299144, + "grad_norm": 1.136816143989563, + "learning_rate": 2.308362104409012e-06, + "loss": 0.8393, + "step": 20934 + }, + { + "epoch": 3.727207977207977, + "grad_norm": 1.0990225076675415, + "learning_rate": 2.305372892480273e-06, + "loss": 0.84, + "step": 20935 + }, + { + "epoch": 3.72738603988604, + "grad_norm": 1.0101460218429565, + "learning_rate": 2.302385594674805e-06, + "loss": 0.8234, + "step": 20936 + }, + { + "epoch": 3.7275641025641026, + "grad_norm": 1.0452793836593628, + "learning_rate": 2.299400211051139e-06, + "loss": 0.7695, + "step": 20937 + }, + { + "epoch": 3.7277421652421654, + "grad_norm": 0.9525721669197083, + "learning_rate": 2.296416741667784e-06, + "loss": 0.8109, + "step": 20938 + }, + { + "epoch": 3.7279202279202277, + "grad_norm": 0.910220205783844, + "learning_rate": 2.293435186583159e-06, + "loss": 0.6896, + "step": 20939 + }, + { + "epoch": 3.7280982905982905, + "grad_norm": 0.9199397563934326, + "learning_rate": 2.2904555458557298e-06, + "loss": 0.795, + "step": 20940 + }, + { + "epoch": 3.728276353276353, + "grad_norm": 1.0512983798980713, + "learning_rate": 2.2874778195438263e-06, + "loss": 0.853, + "step": 20941 + }, + { + "epoch": 3.728454415954416, + "grad_norm": 0.9414512515068054, + "learning_rate": 2.2845020077058356e-06, + "loss": 0.651, + "step": 20942 + }, + { + "epoch": 3.7286324786324787, + "grad_norm": 0.8715613484382629, + "learning_rate": 2.281528110400033e-06, + "loss": 0.7046, + "step": 20943 + }, + { + "epoch": 3.7288105413105415, + "grad_norm": 0.9587253928184509, + "learning_rate": 2.2785561276846947e-06, + "loss": 0.6152, + "step": 20944 + }, + { + "epoch": 3.728988603988604, + "grad_norm": 0.9781585335731506, + "learning_rate": 2.2755860596180514e-06, + "loss": 0.8213, + "step": 20945 + }, + { + "epoch": 3.7291666666666665, + "grad_norm": 0.917754054069519, + "learning_rate": 2.2726179062582786e-06, + "loss": 0.724, + "step": 20946 + }, + { + "epoch": 3.7293447293447293, + "grad_norm": 1.1454041004180908, + "learning_rate": 2.2696516676635747e-06, + "loss": 0.8493, + "step": 20947 + }, + { + "epoch": 3.729522792022792, + "grad_norm": 1.0171089172363281, + "learning_rate": 2.2666873438920042e-06, + "loss": 0.6073, + "step": 20948 + }, + { + "epoch": 3.7297008547008548, + "grad_norm": 0.914182722568512, + "learning_rate": 2.2637249350016877e-06, + "loss": 0.7566, + "step": 20949 + }, + { + "epoch": 3.7298789173789175, + "grad_norm": 0.9400814771652222, + "learning_rate": 2.2607644410506446e-06, + "loss": 0.7434, + "step": 20950 + }, + { + "epoch": 3.73005698005698, + "grad_norm": 0.9671114087104797, + "learning_rate": 2.2578058620968846e-06, + "loss": 0.8685, + "step": 20951 + }, + { + "epoch": 3.7302350427350426, + "grad_norm": 0.9570022821426392, + "learning_rate": 2.254849198198383e-06, + "loss": 0.8194, + "step": 20952 + }, + { + "epoch": 3.7304131054131053, + "grad_norm": 1.0614973306655884, + "learning_rate": 2.251894449413061e-06, + "loss": 0.8729, + "step": 20953 + }, + { + "epoch": 3.730591168091168, + "grad_norm": 0.9944061040878296, + "learning_rate": 2.2489416157988165e-06, + "loss": 0.848, + "step": 20954 + }, + { + "epoch": 3.730769230769231, + "grad_norm": 0.9106524586677551, + "learning_rate": 2.2459906974135026e-06, + "loss": 0.7561, + "step": 20955 + }, + { + "epoch": 3.7309472934472936, + "grad_norm": 1.0078433752059937, + "learning_rate": 2.2430416943149177e-06, + "loss": 0.8776, + "step": 20956 + }, + { + "epoch": 3.7311253561253563, + "grad_norm": 0.9963732361793518, + "learning_rate": 2.240094606560883e-06, + "loss": 0.7677, + "step": 20957 + }, + { + "epoch": 3.7313034188034186, + "grad_norm": 1.1062570810317993, + "learning_rate": 2.237149434209096e-06, + "loss": 0.9031, + "step": 20958 + }, + { + "epoch": 3.7314814814814814, + "grad_norm": 1.0618034601211548, + "learning_rate": 2.2342061773172884e-06, + "loss": 0.6396, + "step": 20959 + }, + { + "epoch": 3.731659544159544, + "grad_norm": 1.0357275009155273, + "learning_rate": 2.2312648359431256e-06, + "loss": 0.8437, + "step": 20960 + }, + { + "epoch": 3.731837606837607, + "grad_norm": 0.8858973979949951, + "learning_rate": 2.2283254101442277e-06, + "loss": 0.6799, + "step": 20961 + }, + { + "epoch": 3.7320156695156697, + "grad_norm": 0.8738244771957397, + "learning_rate": 2.2253878999781818e-06, + "loss": 0.6288, + "step": 20962 + }, + { + "epoch": 3.732193732193732, + "grad_norm": 1.0547064542770386, + "learning_rate": 2.222452305502565e-06, + "loss": 0.9165, + "step": 20963 + }, + { + "epoch": 3.7323717948717947, + "grad_norm": 0.9456502199172974, + "learning_rate": 2.2195186267748637e-06, + "loss": 0.6642, + "step": 20964 + }, + { + "epoch": 3.7325498575498575, + "grad_norm": 0.8678750991821289, + "learning_rate": 2.2165868638525878e-06, + "loss": 0.7079, + "step": 20965 + }, + { + "epoch": 3.73272792022792, + "grad_norm": 1.0330867767333984, + "learning_rate": 2.2136570167931583e-06, + "loss": 0.836, + "step": 20966 + }, + { + "epoch": 3.732905982905983, + "grad_norm": 0.9410081505775452, + "learning_rate": 2.210729085653984e-06, + "loss": 0.6817, + "step": 20967 + }, + { + "epoch": 3.7330840455840457, + "grad_norm": 0.9865879416465759, + "learning_rate": 2.2078030704924423e-06, + "loss": 0.874, + "step": 20968 + }, + { + "epoch": 3.7332621082621085, + "grad_norm": 0.9446350336074829, + "learning_rate": 2.2048789713658423e-06, + "loss": 0.7081, + "step": 20969 + }, + { + "epoch": 3.7334401709401708, + "grad_norm": 0.8920941948890686, + "learning_rate": 2.201956788331494e-06, + "loss": 0.7024, + "step": 20970 + }, + { + "epoch": 3.7336182336182335, + "grad_norm": 0.9637290239334106, + "learning_rate": 2.1990365214466403e-06, + "loss": 0.7942, + "step": 20971 + }, + { + "epoch": 3.7337962962962963, + "grad_norm": 1.140133261680603, + "learning_rate": 2.196118170768513e-06, + "loss": 0.7663, + "step": 20972 + }, + { + "epoch": 3.733974358974359, + "grad_norm": 0.9300453662872314, + "learning_rate": 2.193201736354267e-06, + "loss": 0.7228, + "step": 20973 + }, + { + "epoch": 3.734152421652422, + "grad_norm": 0.9268558621406555, + "learning_rate": 2.1902872182610666e-06, + "loss": 0.7171, + "step": 20974 + }, + { + "epoch": 3.734330484330484, + "grad_norm": 0.9928264021873474, + "learning_rate": 2.1873746165460007e-06, + "loss": 0.7955, + "step": 20975 + }, + { + "epoch": 3.734508547008547, + "grad_norm": 0.9498245716094971, + "learning_rate": 2.184463931266145e-06, + "loss": 0.9005, + "step": 20976 + }, + { + "epoch": 3.7346866096866096, + "grad_norm": 0.8877419829368591, + "learning_rate": 2.1815551624785213e-06, + "loss": 0.6512, + "step": 20977 + }, + { + "epoch": 3.7348646723646723, + "grad_norm": 0.974147617816925, + "learning_rate": 2.178648310240128e-06, + "loss": 0.8416, + "step": 20978 + }, + { + "epoch": 3.735042735042735, + "grad_norm": 0.8951764702796936, + "learning_rate": 2.175743374607897e-06, + "loss": 0.5664, + "step": 20979 + }, + { + "epoch": 3.735220797720798, + "grad_norm": 1.0801883935928345, + "learning_rate": 2.1728403556387833e-06, + "loss": 0.9177, + "step": 20980 + }, + { + "epoch": 3.7353988603988606, + "grad_norm": 0.9792835712432861, + "learning_rate": 2.1699392533896302e-06, + "loss": 0.7572, + "step": 20981 + }, + { + "epoch": 3.7355769230769234, + "grad_norm": 0.8745859265327454, + "learning_rate": 2.167040067917314e-06, + "loss": 0.7706, + "step": 20982 + }, + { + "epoch": 3.7357549857549857, + "grad_norm": 0.9623910784721375, + "learning_rate": 2.164142799278601e-06, + "loss": 0.6661, + "step": 20983 + }, + { + "epoch": 3.7359330484330484, + "grad_norm": 0.8940355181694031, + "learning_rate": 2.161247447530268e-06, + "loss": 0.6218, + "step": 20984 + }, + { + "epoch": 3.736111111111111, + "grad_norm": 0.9015322327613831, + "learning_rate": 2.158354012729069e-06, + "loss": 0.7744, + "step": 20985 + }, + { + "epoch": 3.736289173789174, + "grad_norm": 0.9271003007888794, + "learning_rate": 2.155462494931648e-06, + "loss": 0.7519, + "step": 20986 + }, + { + "epoch": 3.736467236467236, + "grad_norm": 0.8898332715034485, + "learning_rate": 2.1525728941947156e-06, + "loss": 0.8417, + "step": 20987 + }, + { + "epoch": 3.736645299145299, + "grad_norm": 0.9652456641197205, + "learning_rate": 2.1496852105748256e-06, + "loss": 0.9963, + "step": 20988 + }, + { + "epoch": 3.7368233618233617, + "grad_norm": 1.1075891256332397, + "learning_rate": 2.146799444128611e-06, + "loss": 0.8754, + "step": 20989 + }, + { + "epoch": 3.7370014245014245, + "grad_norm": 0.8933864235877991, + "learning_rate": 2.1439155949125822e-06, + "loss": 0.6959, + "step": 20990 + }, + { + "epoch": 3.7371794871794872, + "grad_norm": 0.8123324513435364, + "learning_rate": 2.1410336629832497e-06, + "loss": 0.5327, + "step": 20991 + }, + { + "epoch": 3.73735754985755, + "grad_norm": 1.122158169746399, + "learning_rate": 2.1381536483970676e-06, + "loss": 0.5568, + "step": 20992 + }, + { + "epoch": 3.7375356125356127, + "grad_norm": 0.9673072099685669, + "learning_rate": 2.1352755512104916e-06, + "loss": 0.7785, + "step": 20993 + }, + { + "epoch": 3.7377136752136755, + "grad_norm": 0.9374717473983765, + "learning_rate": 2.1323993714798874e-06, + "loss": 0.8412, + "step": 20994 + }, + { + "epoch": 3.737891737891738, + "grad_norm": 0.9312807321548462, + "learning_rate": 2.1295251092616095e-06, + "loss": 0.8846, + "step": 20995 + }, + { + "epoch": 3.7380698005698005, + "grad_norm": 1.020756483078003, + "learning_rate": 2.1266527646119805e-06, + "loss": 0.5703, + "step": 20996 + }, + { + "epoch": 3.7382478632478633, + "grad_norm": 0.9203028082847595, + "learning_rate": 2.123782337587288e-06, + "loss": 0.7471, + "step": 20997 + }, + { + "epoch": 3.738425925925926, + "grad_norm": 0.9450611472129822, + "learning_rate": 2.1209138282437423e-06, + "loss": 0.6721, + "step": 20998 + }, + { + "epoch": 3.7386039886039883, + "grad_norm": 0.8743969798088074, + "learning_rate": 2.118047236637577e-06, + "loss": 0.5839, + "step": 20999 + }, + { + "epoch": 3.738782051282051, + "grad_norm": 0.9334909915924072, + "learning_rate": 2.1151825628249357e-06, + "loss": 0.7405, + "step": 21000 + }, + { + "epoch": 3.738960113960114, + "grad_norm": 0.918246865272522, + "learning_rate": 2.112319806861962e-06, + "loss": 0.8169, + "step": 21001 + }, + { + "epoch": 3.7391381766381766, + "grad_norm": 1.0408916473388672, + "learning_rate": 2.109458968804734e-06, + "loss": 0.6997, + "step": 21002 + }, + { + "epoch": 3.7393162393162394, + "grad_norm": 0.905485212802887, + "learning_rate": 2.1066000487093063e-06, + "loss": 0.78, + "step": 21003 + }, + { + "epoch": 3.739494301994302, + "grad_norm": 0.8806813359260559, + "learning_rate": 2.103743046631701e-06, + "loss": 0.5886, + "step": 21004 + }, + { + "epoch": 3.739672364672365, + "grad_norm": 0.9529492259025574, + "learning_rate": 2.100887962627873e-06, + "loss": 0.7533, + "step": 21005 + }, + { + "epoch": 3.7398504273504276, + "grad_norm": 1.0024051666259766, + "learning_rate": 2.0980347967537895e-06, + "loss": 0.8242, + "step": 21006 + }, + { + "epoch": 3.74002849002849, + "grad_norm": 0.9132628440856934, + "learning_rate": 2.0951835490653382e-06, + "loss": 0.9098, + "step": 21007 + }, + { + "epoch": 3.7402065527065527, + "grad_norm": 0.8995814919471741, + "learning_rate": 2.092334219618397e-06, + "loss": 0.7624, + "step": 21008 + }, + { + "epoch": 3.7403846153846154, + "grad_norm": 0.9356616139411926, + "learning_rate": 2.089486808468766e-06, + "loss": 0.8032, + "step": 21009 + }, + { + "epoch": 3.740562678062678, + "grad_norm": 0.9492998719215393, + "learning_rate": 2.0866413156722554e-06, + "loss": 0.7695, + "step": 21010 + }, + { + "epoch": 3.7407407407407405, + "grad_norm": 0.9392259120941162, + "learning_rate": 2.083797741284599e-06, + "loss": 0.6681, + "step": 21011 + }, + { + "epoch": 3.7409188034188032, + "grad_norm": 0.9827434420585632, + "learning_rate": 2.0809560853615517e-06, + "loss": 0.7144, + "step": 21012 + }, + { + "epoch": 3.741096866096866, + "grad_norm": 1.0385278463363647, + "learning_rate": 2.078116347958725e-06, + "loss": 0.6568, + "step": 21013 + }, + { + "epoch": 3.7412749287749287, + "grad_norm": 0.931098461151123, + "learning_rate": 2.075278529131819e-06, + "loss": 0.7402, + "step": 21014 + }, + { + "epoch": 3.7414529914529915, + "grad_norm": 0.8532630801200867, + "learning_rate": 2.0724426289363995e-06, + "loss": 0.6599, + "step": 21015 + }, + { + "epoch": 3.7416310541310542, + "grad_norm": 1.0030345916748047, + "learning_rate": 2.0696086474280453e-06, + "loss": 0.6981, + "step": 21016 + }, + { + "epoch": 3.741809116809117, + "grad_norm": 0.8422593474388123, + "learning_rate": 2.0667765846622667e-06, + "loss": 0.6115, + "step": 21017 + }, + { + "epoch": 3.7419871794871797, + "grad_norm": 1.0260411500930786, + "learning_rate": 2.0639464406945752e-06, + "loss": 0.7547, + "step": 21018 + }, + { + "epoch": 3.742165242165242, + "grad_norm": 0.953831136226654, + "learning_rate": 2.0611182155804045e-06, + "loss": 0.6901, + "step": 21019 + }, + { + "epoch": 3.742343304843305, + "grad_norm": 0.984882116317749, + "learning_rate": 2.0582919093751653e-06, + "loss": 0.9004, + "step": 21020 + }, + { + "epoch": 3.7425213675213675, + "grad_norm": 0.8946552276611328, + "learning_rate": 2.055467522134236e-06, + "loss": 0.6706, + "step": 21021 + }, + { + "epoch": 3.7426994301994303, + "grad_norm": 0.8353488445281982, + "learning_rate": 2.052645053912983e-06, + "loss": 0.5818, + "step": 21022 + }, + { + "epoch": 3.7428774928774926, + "grad_norm": 0.9077821373939514, + "learning_rate": 2.049824504766651e-06, + "loss": 0.7353, + "step": 21023 + }, + { + "epoch": 3.7430555555555554, + "grad_norm": 1.0203959941864014, + "learning_rate": 2.0470058747505516e-06, + "loss": 0.8087, + "step": 21024 + }, + { + "epoch": 3.743233618233618, + "grad_norm": 1.1064496040344238, + "learning_rate": 2.0441891639198964e-06, + "loss": 0.6609, + "step": 21025 + }, + { + "epoch": 3.743411680911681, + "grad_norm": 0.9128747582435608, + "learning_rate": 2.041374372329852e-06, + "loss": 0.7507, + "step": 21026 + }, + { + "epoch": 3.7435897435897436, + "grad_norm": 0.8393090963363647, + "learning_rate": 2.0385615000356075e-06, + "loss": 0.5821, + "step": 21027 + }, + { + "epoch": 3.7437678062678064, + "grad_norm": 0.9668859243392944, + "learning_rate": 2.03575054709223e-06, + "loss": 0.863, + "step": 21028 + }, + { + "epoch": 3.743945868945869, + "grad_norm": 0.7194132208824158, + "learning_rate": 2.0329415135548313e-06, + "loss": 0.4231, + "step": 21029 + }, + { + "epoch": 3.744123931623932, + "grad_norm": 1.122671365737915, + "learning_rate": 2.030134399478434e-06, + "loss": 0.7939, + "step": 21030 + }, + { + "epoch": 3.744301994301994, + "grad_norm": 0.9349450469017029, + "learning_rate": 2.027329204918027e-06, + "loss": 0.8354, + "step": 21031 + }, + { + "epoch": 3.744480056980057, + "grad_norm": 0.9546355605125427, + "learning_rate": 2.0245259299286003e-06, + "loss": 0.6134, + "step": 21032 + }, + { + "epoch": 3.7446581196581197, + "grad_norm": 0.8751131296157837, + "learning_rate": 2.0217245745650536e-06, + "loss": 0.6028, + "step": 21033 + }, + { + "epoch": 3.7448361823361824, + "grad_norm": 1.0459450483322144, + "learning_rate": 2.0189251388822773e-06, + "loss": 0.6969, + "step": 21034 + }, + { + "epoch": 3.745014245014245, + "grad_norm": 1.080256700515747, + "learning_rate": 2.0161276229351376e-06, + "loss": 0.7175, + "step": 21035 + }, + { + "epoch": 3.7451923076923075, + "grad_norm": 0.9239566326141357, + "learning_rate": 2.0133320267784137e-06, + "loss": 0.7177, + "step": 21036 + }, + { + "epoch": 3.7453703703703702, + "grad_norm": 1.0287296772003174, + "learning_rate": 2.0105383504669172e-06, + "loss": 0.6814, + "step": 21037 + }, + { + "epoch": 3.745548433048433, + "grad_norm": 0.8914669156074524, + "learning_rate": 2.0077465940553596e-06, + "loss": 0.7653, + "step": 21038 + }, + { + "epoch": 3.7457264957264957, + "grad_norm": 0.9368272423744202, + "learning_rate": 2.004956757598453e-06, + "loss": 0.7811, + "step": 21039 + }, + { + "epoch": 3.7459045584045585, + "grad_norm": 0.9822359085083008, + "learning_rate": 2.002168841150842e-06, + "loss": 0.7224, + "step": 21040 + }, + { + "epoch": 3.7460826210826212, + "grad_norm": 1.0479099750518799, + "learning_rate": 1.9993828447671614e-06, + "loss": 0.7241, + "step": 21041 + }, + { + "epoch": 3.746260683760684, + "grad_norm": 0.8748711347579956, + "learning_rate": 1.9965987685020003e-06, + "loss": 0.7637, + "step": 21042 + }, + { + "epoch": 3.7464387464387463, + "grad_norm": 1.1019326448440552, + "learning_rate": 1.993816612409893e-06, + "loss": 0.88, + "step": 21043 + }, + { + "epoch": 3.746616809116809, + "grad_norm": 0.8620260953903198, + "learning_rate": 1.9910363765453633e-06, + "loss": 0.657, + "step": 21044 + }, + { + "epoch": 3.746794871794872, + "grad_norm": 1.2118844985961914, + "learning_rate": 1.9882580609628774e-06, + "loss": 0.8122, + "step": 21045 + }, + { + "epoch": 3.7469729344729346, + "grad_norm": 0.9898678660392761, + "learning_rate": 1.985481665716882e-06, + "loss": 0.9632, + "step": 21046 + }, + { + "epoch": 3.7471509971509973, + "grad_norm": 1.0883867740631104, + "learning_rate": 1.9827071908617545e-06, + "loss": 0.8129, + "step": 21047 + }, + { + "epoch": 3.7473290598290596, + "grad_norm": 0.9610900282859802, + "learning_rate": 1.9799346364518745e-06, + "loss": 0.8895, + "step": 21048 + }, + { + "epoch": 3.7475071225071224, + "grad_norm": 1.0719484090805054, + "learning_rate": 1.977164002541554e-06, + "loss": 0.7515, + "step": 21049 + }, + { + "epoch": 3.747685185185185, + "grad_norm": 0.915931761264801, + "learning_rate": 1.9743952891850825e-06, + "loss": 0.8945, + "step": 21050 + }, + { + "epoch": 3.747863247863248, + "grad_norm": 0.8382566571235657, + "learning_rate": 1.971628496436695e-06, + "loss": 0.5806, + "step": 21051 + }, + { + "epoch": 3.7480413105413106, + "grad_norm": 0.9513857960700989, + "learning_rate": 1.968863624350625e-06, + "loss": 0.6873, + "step": 21052 + }, + { + "epoch": 3.7482193732193734, + "grad_norm": 0.8445575833320618, + "learning_rate": 1.9661006729810196e-06, + "loss": 0.7492, + "step": 21053 + }, + { + "epoch": 3.748397435897436, + "grad_norm": 1.137221097946167, + "learning_rate": 1.963339642382034e-06, + "loss": 0.796, + "step": 21054 + }, + { + "epoch": 3.7485754985754984, + "grad_norm": 0.9879746437072754, + "learning_rate": 1.9605805326077364e-06, + "loss": 0.6656, + "step": 21055 + }, + { + "epoch": 3.748753561253561, + "grad_norm": 0.8657693266868591, + "learning_rate": 1.9578233437122173e-06, + "loss": 0.7989, + "step": 21056 + }, + { + "epoch": 3.748931623931624, + "grad_norm": 0.9027921557426453, + "learning_rate": 1.9550680757494887e-06, + "loss": 0.7416, + "step": 21057 + }, + { + "epoch": 3.7491096866096867, + "grad_norm": 0.9542195200920105, + "learning_rate": 1.952314728773519e-06, + "loss": 0.7625, + "step": 21058 + }, + { + "epoch": 3.7492877492877494, + "grad_norm": 0.9206588864326477, + "learning_rate": 1.9495633028382755e-06, + "loss": 0.5892, + "step": 21059 + }, + { + "epoch": 3.7494658119658117, + "grad_norm": 0.8397642374038696, + "learning_rate": 1.9468137979976597e-06, + "loss": 0.6284, + "step": 21060 + }, + { + "epoch": 3.7494658119658117, + "eval_loss": 1.1768102645874023, + "eval_runtime": 25.9484, + "eval_samples_per_second": 40.118, + "eval_steps_per_second": 20.078, + "step": 21060 + }, + { + "epoch": 3.7496438746438745, + "grad_norm": 0.996372640132904, + "learning_rate": 1.944066214305518e-06, + "loss": 0.6501, + "step": 21061 + }, + { + "epoch": 3.7498219373219372, + "grad_norm": 1.2021937370300293, + "learning_rate": 1.9413205518157284e-06, + "loss": 0.7819, + "step": 21062 + }, + { + "epoch": 3.75, + "grad_norm": 0.9370668530464172, + "learning_rate": 1.938576810582049e-06, + "loss": 0.5687, + "step": 21063 + }, + { + "epoch": 3.7501780626780628, + "grad_norm": 0.8952577114105225, + "learning_rate": 1.9358349906582694e-06, + "loss": 0.6616, + "step": 21064 + }, + { + "epoch": 3.7503561253561255, + "grad_norm": 0.9440550804138184, + "learning_rate": 1.9330950920980696e-06, + "loss": 0.8607, + "step": 21065 + }, + { + "epoch": 3.7505341880341883, + "grad_norm": 0.897367000579834, + "learning_rate": 1.9303571149551726e-06, + "loss": 0.7928, + "step": 21066 + }, + { + "epoch": 3.7507122507122506, + "grad_norm": 0.932023823261261, + "learning_rate": 1.9276210592832023e-06, + "loss": 0.7398, + "step": 21067 + }, + { + "epoch": 3.7508903133903133, + "grad_norm": 0.9534773230552673, + "learning_rate": 1.92488692513576e-06, + "loss": 0.7433, + "step": 21068 + }, + { + "epoch": 3.751068376068376, + "grad_norm": 0.8925126791000366, + "learning_rate": 1.9221547125664372e-06, + "loss": 0.8369, + "step": 21069 + }, + { + "epoch": 3.751246438746439, + "grad_norm": 1.084887146949768, + "learning_rate": 1.919424421628746e-06, + "loss": 0.8246, + "step": 21070 + }, + { + "epoch": 3.7514245014245016, + "grad_norm": 0.9539969563484192, + "learning_rate": 1.916696052376188e-06, + "loss": 0.726, + "step": 21071 + }, + { + "epoch": 3.751602564102564, + "grad_norm": 0.8407643437385559, + "learning_rate": 1.9139696048622313e-06, + "loss": 0.6814, + "step": 21072 + }, + { + "epoch": 3.7517806267806266, + "grad_norm": 1.1756813526153564, + "learning_rate": 1.911245079140278e-06, + "loss": 0.9399, + "step": 21073 + }, + { + "epoch": 3.7519586894586894, + "grad_norm": 1.0977120399475098, + "learning_rate": 1.9085224752637075e-06, + "loss": 0.9266, + "step": 21074 + }, + { + "epoch": 3.752136752136752, + "grad_norm": 0.9862158298492432, + "learning_rate": 1.9058017932858773e-06, + "loss": 0.7854, + "step": 21075 + }, + { + "epoch": 3.752314814814815, + "grad_norm": 1.0166863203048706, + "learning_rate": 1.9030830332601001e-06, + "loss": 0.6384, + "step": 21076 + }, + { + "epoch": 3.7524928774928776, + "grad_norm": 0.900991678237915, + "learning_rate": 1.9003661952396224e-06, + "loss": 0.7819, + "step": 21077 + }, + { + "epoch": 3.7526709401709404, + "grad_norm": 0.8836488127708435, + "learning_rate": 1.897651279277668e-06, + "loss": 0.5465, + "step": 21078 + }, + { + "epoch": 3.7528490028490027, + "grad_norm": 0.9742031693458557, + "learning_rate": 1.8949382854274722e-06, + "loss": 0.7809, + "step": 21079 + }, + { + "epoch": 3.7530270655270654, + "grad_norm": 0.9790308475494385, + "learning_rate": 1.8922272137421482e-06, + "loss": 0.6517, + "step": 21080 + }, + { + "epoch": 3.753205128205128, + "grad_norm": 1.0090285539627075, + "learning_rate": 1.8895180642748422e-06, + "loss": 0.7865, + "step": 21081 + }, + { + "epoch": 3.753383190883191, + "grad_norm": 0.9838883876800537, + "learning_rate": 1.886810837078623e-06, + "loss": 0.6011, + "step": 21082 + }, + { + "epoch": 3.7535612535612537, + "grad_norm": 0.9217146039009094, + "learning_rate": 1.884105532206526e-06, + "loss": 0.7815, + "step": 21083 + }, + { + "epoch": 3.753739316239316, + "grad_norm": 1.0938522815704346, + "learning_rate": 1.8814021497115642e-06, + "loss": 0.9792, + "step": 21084 + }, + { + "epoch": 3.7539173789173788, + "grad_norm": 1.0393195152282715, + "learning_rate": 1.8787006896466952e-06, + "loss": 0.7068, + "step": 21085 + }, + { + "epoch": 3.7540954415954415, + "grad_norm": 0.9190340042114258, + "learning_rate": 1.8760011520648656e-06, + "loss": 0.7742, + "step": 21086 + }, + { + "epoch": 3.7542735042735043, + "grad_norm": 0.9961007833480835, + "learning_rate": 1.8733035370189667e-06, + "loss": 0.884, + "step": 21087 + }, + { + "epoch": 3.754451566951567, + "grad_norm": 0.9565320014953613, + "learning_rate": 1.8706078445618336e-06, + "loss": 0.6365, + "step": 21088 + }, + { + "epoch": 3.7546296296296298, + "grad_norm": 0.9802839756011963, + "learning_rate": 1.867914074746302e-06, + "loss": 0.64, + "step": 21089 + }, + { + "epoch": 3.7548076923076925, + "grad_norm": 0.9352937340736389, + "learning_rate": 1.8652222276251408e-06, + "loss": 0.6464, + "step": 21090 + }, + { + "epoch": 3.754985754985755, + "grad_norm": 0.960495114326477, + "learning_rate": 1.8625323032510855e-06, + "loss": 0.7469, + "step": 21091 + }, + { + "epoch": 3.7551638176638176, + "grad_norm": 1.088507056236267, + "learning_rate": 1.8598443016768497e-06, + "loss": 0.9544, + "step": 21092 + }, + { + "epoch": 3.7553418803418803, + "grad_norm": 0.871731162071228, + "learning_rate": 1.8571582229550911e-06, + "loss": 0.7206, + "step": 21093 + }, + { + "epoch": 3.755519943019943, + "grad_norm": 0.9332573413848877, + "learning_rate": 1.8544740671384565e-06, + "loss": 0.8691, + "step": 21094 + }, + { + "epoch": 3.755698005698006, + "grad_norm": 0.9459754228591919, + "learning_rate": 1.8517918342795148e-06, + "loss": 0.7293, + "step": 21095 + }, + { + "epoch": 3.755876068376068, + "grad_norm": 1.2970104217529297, + "learning_rate": 1.8491115244308243e-06, + "loss": 0.8888, + "step": 21096 + }, + { + "epoch": 3.756054131054131, + "grad_norm": 0.8978582620620728, + "learning_rate": 1.8464331376449095e-06, + "loss": 0.6223, + "step": 21097 + }, + { + "epoch": 3.7562321937321936, + "grad_norm": 0.8985556960105896, + "learning_rate": 1.8437566739742396e-06, + "loss": 0.6488, + "step": 21098 + }, + { + "epoch": 3.7564102564102564, + "grad_norm": 1.0541517734527588, + "learning_rate": 1.8410821334712503e-06, + "loss": 0.7203, + "step": 21099 + }, + { + "epoch": 3.756588319088319, + "grad_norm": 0.96063631772995, + "learning_rate": 1.8384095161883552e-06, + "loss": 0.7773, + "step": 21100 + }, + { + "epoch": 3.756766381766382, + "grad_norm": 1.0475854873657227, + "learning_rate": 1.8357388221779126e-06, + "loss": 0.8795, + "step": 21101 + }, + { + "epoch": 3.7569444444444446, + "grad_norm": 0.9499286413192749, + "learning_rate": 1.8330700514922472e-06, + "loss": 0.7407, + "step": 21102 + }, + { + "epoch": 3.7571225071225074, + "grad_norm": 0.94266676902771, + "learning_rate": 1.8304032041836393e-06, + "loss": 0.7239, + "step": 21103 + }, + { + "epoch": 3.7573005698005697, + "grad_norm": 0.9316100478172302, + "learning_rate": 1.8277382803043585e-06, + "loss": 0.7977, + "step": 21104 + }, + { + "epoch": 3.7574786324786325, + "grad_norm": 0.9400086998939514, + "learning_rate": 1.8250752799066184e-06, + "loss": 0.6354, + "step": 21105 + }, + { + "epoch": 3.757656695156695, + "grad_norm": 0.9189943671226501, + "learning_rate": 1.8224142030425772e-06, + "loss": 0.7169, + "step": 21106 + }, + { + "epoch": 3.757834757834758, + "grad_norm": 1.0137799978256226, + "learning_rate": 1.8197550497643935e-06, + "loss": 0.5435, + "step": 21107 + }, + { + "epoch": 3.7580128205128203, + "grad_norm": 0.9665256142616272, + "learning_rate": 1.8170978201241474e-06, + "loss": 0.7213, + "step": 21108 + }, + { + "epoch": 3.758190883190883, + "grad_norm": 0.9698337316513062, + "learning_rate": 1.8144425141739196e-06, + "loss": 0.7327, + "step": 21109 + }, + { + "epoch": 3.7583689458689458, + "grad_norm": 1.0307881832122803, + "learning_rate": 1.8117891319657243e-06, + "loss": 0.7908, + "step": 21110 + }, + { + "epoch": 3.7585470085470085, + "grad_norm": 1.087488055229187, + "learning_rate": 1.809137673551564e-06, + "loss": 0.662, + "step": 21111 + }, + { + "epoch": 3.7587250712250713, + "grad_norm": 0.8748802542686462, + "learning_rate": 1.8064881389833533e-06, + "loss": 0.625, + "step": 21112 + }, + { + "epoch": 3.758903133903134, + "grad_norm": 0.9132877588272095, + "learning_rate": 1.8038405283130499e-06, + "loss": 0.9392, + "step": 21113 + }, + { + "epoch": 3.7590811965811968, + "grad_norm": 1.169862985610962, + "learning_rate": 1.8011948415925017e-06, + "loss": 0.7501, + "step": 21114 + }, + { + "epoch": 3.7592592592592595, + "grad_norm": 1.095362663269043, + "learning_rate": 1.7985510788735449e-06, + "loss": 0.7822, + "step": 21115 + }, + { + "epoch": 3.759437321937322, + "grad_norm": 1.0184403657913208, + "learning_rate": 1.7959092402079825e-06, + "loss": 0.8347, + "step": 21116 + }, + { + "epoch": 3.7596153846153846, + "grad_norm": 1.0453873872756958, + "learning_rate": 1.793269325647584e-06, + "loss": 0.6534, + "step": 21117 + }, + { + "epoch": 3.7597934472934473, + "grad_norm": 0.8609569668769836, + "learning_rate": 1.790631335244053e-06, + "loss": 0.7615, + "step": 21118 + }, + { + "epoch": 3.75997150997151, + "grad_norm": 0.960797131061554, + "learning_rate": 1.7879952690491141e-06, + "loss": 0.6932, + "step": 21119 + }, + { + "epoch": 3.7601495726495724, + "grad_norm": 0.9469041228294373, + "learning_rate": 1.78536112711436e-06, + "loss": 0.7238, + "step": 21120 + }, + { + "epoch": 3.760327635327635, + "grad_norm": 0.9235484600067139, + "learning_rate": 1.782728909491449e-06, + "loss": 0.7654, + "step": 21121 + }, + { + "epoch": 3.760505698005698, + "grad_norm": 1.1060317754745483, + "learning_rate": 1.78009861623194e-06, + "loss": 0.8391, + "step": 21122 + }, + { + "epoch": 3.7606837606837606, + "grad_norm": 0.9043984413146973, + "learning_rate": 1.7774702473873584e-06, + "loss": 0.8132, + "step": 21123 + }, + { + "epoch": 3.7608618233618234, + "grad_norm": 0.9583942890167236, + "learning_rate": 1.7748438030092074e-06, + "loss": 0.695, + "step": 21124 + }, + { + "epoch": 3.761039886039886, + "grad_norm": 0.9370834231376648, + "learning_rate": 1.7722192831489348e-06, + "loss": 0.829, + "step": 21125 + }, + { + "epoch": 3.761217948717949, + "grad_norm": 0.9919298887252808, + "learning_rate": 1.7695966878579994e-06, + "loss": 0.9974, + "step": 21126 + }, + { + "epoch": 3.7613960113960117, + "grad_norm": 0.9389301538467407, + "learning_rate": 1.766976017187738e-06, + "loss": 0.8368, + "step": 21127 + }, + { + "epoch": 3.761574074074074, + "grad_norm": 0.8722683787345886, + "learning_rate": 1.7643572711895318e-06, + "loss": 0.689, + "step": 21128 + }, + { + "epoch": 3.7617521367521367, + "grad_norm": 1.0586130619049072, + "learning_rate": 1.7617404499146839e-06, + "loss": 0.8207, + "step": 21129 + }, + { + "epoch": 3.7619301994301995, + "grad_norm": 1.040549397468567, + "learning_rate": 1.7591255534144535e-06, + "loss": 0.878, + "step": 21130 + }, + { + "epoch": 3.762108262108262, + "grad_norm": 1.0269027948379517, + "learning_rate": 1.7565125817400773e-06, + "loss": 0.867, + "step": 21131 + }, + { + "epoch": 3.7622863247863245, + "grad_norm": 0.923396110534668, + "learning_rate": 1.753901534942759e-06, + "loss": 0.7129, + "step": 21132 + }, + { + "epoch": 3.7624643874643873, + "grad_norm": 1.132996916770935, + "learning_rate": 1.7512924130736353e-06, + "loss": 0.7754, + "step": 21133 + }, + { + "epoch": 3.76264245014245, + "grad_norm": 0.9143159985542297, + "learning_rate": 1.7486852161838653e-06, + "loss": 0.7042, + "step": 21134 + }, + { + "epoch": 3.7628205128205128, + "grad_norm": 0.9809998869895935, + "learning_rate": 1.7460799443244968e-06, + "loss": 0.8026, + "step": 21135 + }, + { + "epoch": 3.7629985754985755, + "grad_norm": 0.9136005640029907, + "learning_rate": 1.7434765975466006e-06, + "loss": 0.8317, + "step": 21136 + }, + { + "epoch": 3.7631766381766383, + "grad_norm": 0.8698214292526245, + "learning_rate": 1.7408751759011465e-06, + "loss": 0.8658, + "step": 21137 + }, + { + "epoch": 3.763354700854701, + "grad_norm": 1.0867608785629272, + "learning_rate": 1.7382756794391386e-06, + "loss": 0.9433, + "step": 21138 + }, + { + "epoch": 3.763532763532764, + "grad_norm": 0.8608876466751099, + "learning_rate": 1.7356781082115026e-06, + "loss": 0.6308, + "step": 21139 + }, + { + "epoch": 3.763710826210826, + "grad_norm": 0.8908528685569763, + "learning_rate": 1.733082462269131e-06, + "loss": 0.6537, + "step": 21140 + }, + { + "epoch": 3.763888888888889, + "grad_norm": 0.9231675863265991, + "learning_rate": 1.7304887416628724e-06, + "loss": 0.8053, + "step": 21141 + }, + { + "epoch": 3.7640669515669516, + "grad_norm": 0.9419436454772949, + "learning_rate": 1.7278969464435413e-06, + "loss": 0.8501, + "step": 21142 + }, + { + "epoch": 3.7642450142450143, + "grad_norm": 1.0135849714279175, + "learning_rate": 1.7253070766619305e-06, + "loss": 0.8137, + "step": 21143 + }, + { + "epoch": 3.7644230769230766, + "grad_norm": 0.9566580057144165, + "learning_rate": 1.7227191323687774e-06, + "loss": 0.73, + "step": 21144 + }, + { + "epoch": 3.7646011396011394, + "grad_norm": 1.033128261566162, + "learning_rate": 1.7201331136147968e-06, + "loss": 0.6961, + "step": 21145 + }, + { + "epoch": 3.764779202279202, + "grad_norm": 0.9674811959266663, + "learning_rate": 1.7175490204506484e-06, + "loss": 0.8184, + "step": 21146 + }, + { + "epoch": 3.764957264957265, + "grad_norm": 0.9115480184555054, + "learning_rate": 1.714966852926958e-06, + "loss": 0.9166, + "step": 21147 + }, + { + "epoch": 3.7651353276353277, + "grad_norm": 0.9724944233894348, + "learning_rate": 1.7123866110943298e-06, + "loss": 0.7997, + "step": 21148 + }, + { + "epoch": 3.7653133903133904, + "grad_norm": 0.9939026832580566, + "learning_rate": 1.7098082950033124e-06, + "loss": 0.7107, + "step": 21149 + }, + { + "epoch": 3.765491452991453, + "grad_norm": 0.9845420718193054, + "learning_rate": 1.7072319047044094e-06, + "loss": 0.7363, + "step": 21150 + }, + { + "epoch": 3.765669515669516, + "grad_norm": 1.0131663084030151, + "learning_rate": 1.7046574402481253e-06, + "loss": 0.5866, + "step": 21151 + }, + { + "epoch": 3.765847578347578, + "grad_norm": 1.0212078094482422, + "learning_rate": 1.702084901684875e-06, + "loss": 0.8843, + "step": 21152 + }, + { + "epoch": 3.766025641025641, + "grad_norm": 0.9821169972419739, + "learning_rate": 1.699514289065074e-06, + "loss": 0.8066, + "step": 21153 + }, + { + "epoch": 3.7662037037037037, + "grad_norm": 0.8868746757507324, + "learning_rate": 1.696945602439104e-06, + "loss": 0.789, + "step": 21154 + }, + { + "epoch": 3.7663817663817665, + "grad_norm": 0.9874410033226013, + "learning_rate": 1.6943788418572692e-06, + "loss": 0.7745, + "step": 21155 + }, + { + "epoch": 3.7665598290598292, + "grad_norm": 0.8668256402015686, + "learning_rate": 1.691814007369863e-06, + "loss": 0.7893, + "step": 21156 + }, + { + "epoch": 3.7667378917378915, + "grad_norm": 1.0595078468322754, + "learning_rate": 1.689251099027156e-06, + "loss": 0.7377, + "step": 21157 + }, + { + "epoch": 3.7669159544159543, + "grad_norm": 1.0612707138061523, + "learning_rate": 1.6866901168793413e-06, + "loss": 0.7955, + "step": 21158 + }, + { + "epoch": 3.767094017094017, + "grad_norm": 0.8881635665893555, + "learning_rate": 1.6841310609766126e-06, + "loss": 0.8739, + "step": 21159 + }, + { + "epoch": 3.76727207977208, + "grad_norm": 0.9735580086708069, + "learning_rate": 1.681573931369096e-06, + "loss": 0.7704, + "step": 21160 + }, + { + "epoch": 3.7674501424501425, + "grad_norm": 0.9362749457359314, + "learning_rate": 1.6790187281069069e-06, + "loss": 0.7712, + "step": 21161 + }, + { + "epoch": 3.7676282051282053, + "grad_norm": 0.9322159290313721, + "learning_rate": 1.6764654512400946e-06, + "loss": 0.741, + "step": 21162 + }, + { + "epoch": 3.767806267806268, + "grad_norm": 0.9788293242454529, + "learning_rate": 1.6739141008186854e-06, + "loss": 0.7966, + "step": 21163 + }, + { + "epoch": 3.7679843304843303, + "grad_norm": 0.9341052174568176, + "learning_rate": 1.671364676892695e-06, + "loss": 0.5836, + "step": 21164 + }, + { + "epoch": 3.768162393162393, + "grad_norm": 1.0333666801452637, + "learning_rate": 1.6688171795120278e-06, + "loss": 0.7703, + "step": 21165 + }, + { + "epoch": 3.768340455840456, + "grad_norm": 1.1365190744400024, + "learning_rate": 1.666271608726644e-06, + "loss": 0.8859, + "step": 21166 + }, + { + "epoch": 3.7685185185185186, + "grad_norm": 0.9362785220146179, + "learning_rate": 1.6637279645863923e-06, + "loss": 0.6945, + "step": 21167 + }, + { + "epoch": 3.7686965811965814, + "grad_norm": 0.9373227953910828, + "learning_rate": 1.661186247141111e-06, + "loss": 0.833, + "step": 21168 + }, + { + "epoch": 3.7688746438746437, + "grad_norm": 1.067119836807251, + "learning_rate": 1.6586464564406046e-06, + "loss": 0.78, + "step": 21169 + }, + { + "epoch": 3.7690527065527064, + "grad_norm": 1.0004785060882568, + "learning_rate": 1.6561085925346332e-06, + "loss": 0.7274, + "step": 21170 + }, + { + "epoch": 3.769230769230769, + "grad_norm": 0.803828239440918, + "learning_rate": 1.6535726554729347e-06, + "loss": 0.5589, + "step": 21171 + }, + { + "epoch": 3.769408831908832, + "grad_norm": 0.9255486726760864, + "learning_rate": 1.6510386453051695e-06, + "loss": 0.8917, + "step": 21172 + }, + { + "epoch": 3.7695868945868947, + "grad_norm": 0.9268200397491455, + "learning_rate": 1.6485065620810092e-06, + "loss": 0.7948, + "step": 21173 + }, + { + "epoch": 3.7697649572649574, + "grad_norm": 1.0468385219573975, + "learning_rate": 1.6459764058500472e-06, + "loss": 0.6552, + "step": 21174 + }, + { + "epoch": 3.76994301994302, + "grad_norm": 0.859835147857666, + "learning_rate": 1.643448176661866e-06, + "loss": 0.7813, + "step": 21175 + }, + { + "epoch": 3.7701210826210825, + "grad_norm": 1.0402703285217285, + "learning_rate": 1.6409218745660037e-06, + "loss": 0.8416, + "step": 21176 + }, + { + "epoch": 3.7702991452991452, + "grad_norm": 0.8524719476699829, + "learning_rate": 1.6383974996119434e-06, + "loss": 0.5359, + "step": 21177 + }, + { + "epoch": 3.770477207977208, + "grad_norm": 0.9544599056243896, + "learning_rate": 1.6358750518491672e-06, + "loss": 0.7721, + "step": 21178 + }, + { + "epoch": 3.7706552706552707, + "grad_norm": 1.1628994941711426, + "learning_rate": 1.6333545313270803e-06, + "loss": 0.7837, + "step": 21179 + }, + { + "epoch": 3.7708333333333335, + "grad_norm": 0.9267588257789612, + "learning_rate": 1.6308359380950765e-06, + "loss": 0.6741, + "step": 21180 + }, + { + "epoch": 3.771011396011396, + "grad_norm": 0.9370381236076355, + "learning_rate": 1.6283192722024942e-06, + "loss": 0.8062, + "step": 21181 + }, + { + "epoch": 3.7711894586894585, + "grad_norm": 0.8971649408340454, + "learning_rate": 1.6258045336986493e-06, + "loss": 0.7223, + "step": 21182 + }, + { + "epoch": 3.7713675213675213, + "grad_norm": 0.8068461418151855, + "learning_rate": 1.6232917226328137e-06, + "loss": 0.8003, + "step": 21183 + }, + { + "epoch": 3.771545584045584, + "grad_norm": 0.830176830291748, + "learning_rate": 1.6207808390542034e-06, + "loss": 0.5199, + "step": 21184 + }, + { + "epoch": 3.771723646723647, + "grad_norm": 1.0348658561706543, + "learning_rate": 1.6182718830120346e-06, + "loss": 0.9757, + "step": 21185 + }, + { + "epoch": 3.7719017094017095, + "grad_norm": 0.9341155290603638, + "learning_rate": 1.6157648545554571e-06, + "loss": 0.7659, + "step": 21186 + }, + { + "epoch": 3.7720797720797723, + "grad_norm": 0.9962141513824463, + "learning_rate": 1.613259753733598e-06, + "loss": 0.8433, + "step": 21187 + }, + { + "epoch": 3.7722578347578346, + "grad_norm": 0.994513988494873, + "learning_rate": 1.6107565805955293e-06, + "loss": 0.9445, + "step": 21188 + }, + { + "epoch": 3.7724358974358974, + "grad_norm": 0.9562044143676758, + "learning_rate": 1.6082553351903008e-06, + "loss": 0.7425, + "step": 21189 + }, + { + "epoch": 3.77261396011396, + "grad_norm": 0.9779346585273743, + "learning_rate": 1.6057560175669062e-06, + "loss": 0.6848, + "step": 21190 + }, + { + "epoch": 3.772792022792023, + "grad_norm": 1.0121656656265259, + "learning_rate": 1.603258627774351e-06, + "loss": 0.8805, + "step": 21191 + }, + { + "epoch": 3.7729700854700856, + "grad_norm": 0.8703751564025879, + "learning_rate": 1.6007631658615186e-06, + "loss": 0.763, + "step": 21192 + }, + { + "epoch": 3.773148148148148, + "grad_norm": 0.9765893220901489, + "learning_rate": 1.598269631877336e-06, + "loss": 0.7559, + "step": 21193 + }, + { + "epoch": 3.7733262108262107, + "grad_norm": 1.0756961107254028, + "learning_rate": 1.5957780258706423e-06, + "loss": 0.6645, + "step": 21194 + }, + { + "epoch": 3.7735042735042734, + "grad_norm": 0.9320286512374878, + "learning_rate": 1.593288347890265e-06, + "loss": 0.6629, + "step": 21195 + }, + { + "epoch": 3.773682336182336, + "grad_norm": 0.930663526058197, + "learning_rate": 1.590800597984976e-06, + "loss": 0.715, + "step": 21196 + }, + { + "epoch": 3.773860398860399, + "grad_norm": 0.9290189743041992, + "learning_rate": 1.5883147762035366e-06, + "loss": 0.7135, + "step": 21197 + }, + { + "epoch": 3.7740384615384617, + "grad_norm": 0.9259148240089417, + "learning_rate": 1.585830882594619e-06, + "loss": 0.789, + "step": 21198 + }, + { + "epoch": 3.7742165242165244, + "grad_norm": 1.0864899158477783, + "learning_rate": 1.583348917206917e-06, + "loss": 0.6689, + "step": 21199 + }, + { + "epoch": 3.7743945868945867, + "grad_norm": 0.9391498565673828, + "learning_rate": 1.580868880089037e-06, + "loss": 0.5958, + "step": 21200 + }, + { + "epoch": 3.7745726495726495, + "grad_norm": 0.9732069373130798, + "learning_rate": 1.5783907712896062e-06, + "loss": 0.8109, + "step": 21201 + }, + { + "epoch": 3.7747507122507122, + "grad_norm": 0.9975631237030029, + "learning_rate": 1.5759145908571304e-06, + "loss": 0.6909, + "step": 21202 + }, + { + "epoch": 3.774928774928775, + "grad_norm": 0.9436360001564026, + "learning_rate": 1.5734403388401597e-06, + "loss": 0.7629, + "step": 21203 + }, + { + "epoch": 3.7751068376068377, + "grad_norm": 0.9783167839050293, + "learning_rate": 1.5709680152871664e-06, + "loss": 0.7348, + "step": 21204 + }, + { + "epoch": 3.7752849002849, + "grad_norm": 0.9854038953781128, + "learning_rate": 1.5684976202465784e-06, + "loss": 0.687, + "step": 21205 + }, + { + "epoch": 3.775462962962963, + "grad_norm": 0.8817933797836304, + "learning_rate": 1.5660291537668237e-06, + "loss": 0.7331, + "step": 21206 + }, + { + "epoch": 3.7756410256410255, + "grad_norm": 0.8301796317100525, + "learning_rate": 1.5635626158962303e-06, + "loss": 0.8223, + "step": 21207 + }, + { + "epoch": 3.7758190883190883, + "grad_norm": 1.1981323957443237, + "learning_rate": 1.5610980066831593e-06, + "loss": 0.7869, + "step": 21208 + }, + { + "epoch": 3.775997150997151, + "grad_norm": 1.0148868560791016, + "learning_rate": 1.5586353261758612e-06, + "loss": 0.7864, + "step": 21209 + }, + { + "epoch": 3.776175213675214, + "grad_norm": 0.9011337161064148, + "learning_rate": 1.5561745744226198e-06, + "loss": 0.6761, + "step": 21210 + }, + { + "epoch": 3.7763532763532766, + "grad_norm": 0.9810119271278381, + "learning_rate": 1.5537157514716516e-06, + "loss": 0.6671, + "step": 21211 + }, + { + "epoch": 3.7765313390313393, + "grad_norm": 0.9342007040977478, + "learning_rate": 1.5512588573711074e-06, + "loss": 0.696, + "step": 21212 + }, + { + "epoch": 3.7767094017094016, + "grad_norm": 1.0802044868469238, + "learning_rate": 1.5488038921691373e-06, + "loss": 0.8527, + "step": 21213 + }, + { + "epoch": 3.7768874643874644, + "grad_norm": 0.8123725652694702, + "learning_rate": 1.5463508559138362e-06, + "loss": 0.676, + "step": 21214 + }, + { + "epoch": 3.777065527065527, + "grad_norm": 0.9726629853248596, + "learning_rate": 1.5438997486532658e-06, + "loss": 0.6526, + "step": 21215 + }, + { + "epoch": 3.77724358974359, + "grad_norm": 0.8095312714576721, + "learning_rate": 1.5414505704354764e-06, + "loss": 0.4255, + "step": 21216 + }, + { + "epoch": 3.777421652421652, + "grad_norm": 0.8774579763412476, + "learning_rate": 1.5390033213084075e-06, + "loss": 0.6785, + "step": 21217 + }, + { + "epoch": 3.777599715099715, + "grad_norm": 1.214205265045166, + "learning_rate": 1.5365580013200542e-06, + "loss": 0.7598, + "step": 21218 + }, + { + "epoch": 3.7777777777777777, + "grad_norm": 0.9117157459259033, + "learning_rate": 1.5341146105183002e-06, + "loss": 0.7835, + "step": 21219 + }, + { + "epoch": 3.7779558404558404, + "grad_norm": 0.9653142690658569, + "learning_rate": 1.5316731489510184e-06, + "loss": 0.7136, + "step": 21220 + }, + { + "epoch": 3.778133903133903, + "grad_norm": 0.9996618628501892, + "learning_rate": 1.5292336166660591e-06, + "loss": 0.9551, + "step": 21221 + }, + { + "epoch": 3.778311965811966, + "grad_norm": 0.880121111869812, + "learning_rate": 1.526796013711207e-06, + "loss": 0.6935, + "step": 21222 + }, + { + "epoch": 3.7784900284900287, + "grad_norm": 1.0504447221755981, + "learning_rate": 1.5243603401342343e-06, + "loss": 0.7668, + "step": 21223 + }, + { + "epoch": 3.7786680911680914, + "grad_norm": 0.9271687269210815, + "learning_rate": 1.5219265959828367e-06, + "loss": 0.7224, + "step": 21224 + }, + { + "epoch": 3.7788461538461537, + "grad_norm": 0.9729786515235901, + "learning_rate": 1.5194947813047312e-06, + "loss": 0.7342, + "step": 21225 + }, + { + "epoch": 3.7790242165242165, + "grad_norm": 1.0969620943069458, + "learning_rate": 1.5170648961475576e-06, + "loss": 0.8477, + "step": 21226 + }, + { + "epoch": 3.7792022792022792, + "grad_norm": 0.8730276226997375, + "learning_rate": 1.5146369405589001e-06, + "loss": 0.8116, + "step": 21227 + }, + { + "epoch": 3.779380341880342, + "grad_norm": 1.0915426015853882, + "learning_rate": 1.512210914586365e-06, + "loss": 0.848, + "step": 21228 + }, + { + "epoch": 3.7795584045584043, + "grad_norm": 0.9392247796058655, + "learning_rate": 1.5097868182774477e-06, + "loss": 0.8485, + "step": 21229 + }, + { + "epoch": 3.779736467236467, + "grad_norm": 0.9891657829284668, + "learning_rate": 1.5073646516796658e-06, + "loss": 0.7992, + "step": 21230 + }, + { + "epoch": 3.77991452991453, + "grad_norm": 0.8666152954101562, + "learning_rate": 1.5049444148404813e-06, + "loss": 0.7858, + "step": 21231 + }, + { + "epoch": 3.7800925925925926, + "grad_norm": 1.147723913192749, + "learning_rate": 1.5025261078073005e-06, + "loss": 0.7486, + "step": 21232 + }, + { + "epoch": 3.7802706552706553, + "grad_norm": 0.996700644493103, + "learning_rate": 1.500109730627519e-06, + "loss": 0.8679, + "step": 21233 + }, + { + "epoch": 3.780448717948718, + "grad_norm": 1.0153658390045166, + "learning_rate": 1.4976952833484548e-06, + "loss": 0.6359, + "step": 21234 + }, + { + "epoch": 3.780626780626781, + "grad_norm": 0.9375406503677368, + "learning_rate": 1.4952827660174362e-06, + "loss": 0.8339, + "step": 21235 + }, + { + "epoch": 3.7808048433048436, + "grad_norm": 0.8846505284309387, + "learning_rate": 1.4928721786817256e-06, + "loss": 0.7411, + "step": 21236 + }, + { + "epoch": 3.780982905982906, + "grad_norm": 0.9112651348114014, + "learning_rate": 1.4904635213885631e-06, + "loss": 0.8434, + "step": 21237 + }, + { + "epoch": 3.7811609686609686, + "grad_norm": 1.0657570362091064, + "learning_rate": 1.488056794185122e-06, + "loss": 0.871, + "step": 21238 + }, + { + "epoch": 3.7813390313390314, + "grad_norm": 0.966810941696167, + "learning_rate": 1.4856519971185756e-06, + "loss": 0.8442, + "step": 21239 + }, + { + "epoch": 3.781517094017094, + "grad_norm": 0.9643548727035522, + "learning_rate": 1.4832491302360196e-06, + "loss": 0.7939, + "step": 21240 + }, + { + "epoch": 3.7816951566951564, + "grad_norm": 0.9209652543067932, + "learning_rate": 1.4808481935845608e-06, + "loss": 0.8603, + "step": 21241 + }, + { + "epoch": 3.781873219373219, + "grad_norm": 0.9699517488479614, + "learning_rate": 1.478449187211206e-06, + "loss": 0.7318, + "step": 21242 + }, + { + "epoch": 3.782051282051282, + "grad_norm": 0.9576825499534607, + "learning_rate": 1.4760521111629954e-06, + "loss": 0.8847, + "step": 21243 + }, + { + "epoch": 3.7822293447293447, + "grad_norm": 1.0206605195999146, + "learning_rate": 1.4736569654868803e-06, + "loss": 0.7726, + "step": 21244 + }, + { + "epoch": 3.7824074074074074, + "grad_norm": 1.0139795541763306, + "learning_rate": 1.4712637502297789e-06, + "loss": 0.7702, + "step": 21245 + }, + { + "epoch": 3.78258547008547, + "grad_norm": 0.9946370124816895, + "learning_rate": 1.4688724654385867e-06, + "loss": 0.853, + "step": 21246 + }, + { + "epoch": 3.782763532763533, + "grad_norm": 1.069388508796692, + "learning_rate": 1.4664831111601552e-06, + "loss": 0.8966, + "step": 21247 + }, + { + "epoch": 3.7829415954415957, + "grad_norm": 0.8335608243942261, + "learning_rate": 1.4640956874413137e-06, + "loss": 0.6154, + "step": 21248 + }, + { + "epoch": 3.783119658119658, + "grad_norm": 1.035109043121338, + "learning_rate": 1.4617101943288136e-06, + "loss": 0.7114, + "step": 21249 + }, + { + "epoch": 3.7832977207977208, + "grad_norm": 0.8770612478256226, + "learning_rate": 1.4593266318694176e-06, + "loss": 0.7752, + "step": 21250 + }, + { + "epoch": 3.7834757834757835, + "grad_norm": 0.9383000731468201, + "learning_rate": 1.4569450001098106e-06, + "loss": 0.8335, + "step": 21251 + }, + { + "epoch": 3.7836538461538463, + "grad_norm": 0.9217562675476074, + "learning_rate": 1.4545652990966662e-06, + "loss": 0.6357, + "step": 21252 + }, + { + "epoch": 3.7838319088319086, + "grad_norm": 0.9792481064796448, + "learning_rate": 1.4521875288765918e-06, + "loss": 0.855, + "step": 21253 + }, + { + "epoch": 3.7840099715099713, + "grad_norm": 0.9805694222450256, + "learning_rate": 1.4498116894962055e-06, + "loss": 0.6227, + "step": 21254 + }, + { + "epoch": 3.784188034188034, + "grad_norm": 0.8412477374076843, + "learning_rate": 1.4474377810020257e-06, + "loss": 0.7493, + "step": 21255 + }, + { + "epoch": 3.784366096866097, + "grad_norm": 1.0370614528656006, + "learning_rate": 1.4450658034405817e-06, + "loss": 0.7173, + "step": 21256 + }, + { + "epoch": 3.7845441595441596, + "grad_norm": 0.8881233930587769, + "learning_rate": 1.4426957568583366e-06, + "loss": 0.6639, + "step": 21257 + }, + { + "epoch": 3.7847222222222223, + "grad_norm": 1.362695574760437, + "learning_rate": 1.440327641301742e-06, + "loss": 0.67, + "step": 21258 + }, + { + "epoch": 3.784900284900285, + "grad_norm": 0.9417729377746582, + "learning_rate": 1.4379614568171718e-06, + "loss": 0.841, + "step": 21259 + }, + { + "epoch": 3.785078347578348, + "grad_norm": 0.8955726027488708, + "learning_rate": 1.4355972034510224e-06, + "loss": 0.6153, + "step": 21260 + }, + { + "epoch": 3.78525641025641, + "grad_norm": 0.9870222210884094, + "learning_rate": 1.43323488124959e-06, + "loss": 0.9167, + "step": 21261 + }, + { + "epoch": 3.785434472934473, + "grad_norm": 0.9800770878791809, + "learning_rate": 1.4308744902591598e-06, + "loss": 0.7841, + "step": 21262 + }, + { + "epoch": 3.7856125356125356, + "grad_norm": 0.8268224000930786, + "learning_rate": 1.4285160305259836e-06, + "loss": 0.6672, + "step": 21263 + }, + { + "epoch": 3.7857905982905984, + "grad_norm": 0.9754951000213623, + "learning_rate": 1.4261595020962692e-06, + "loss": 0.8167, + "step": 21264 + }, + { + "epoch": 3.7859686609686607, + "grad_norm": 0.9257098436355591, + "learning_rate": 1.4238049050162018e-06, + "loss": 0.5471, + "step": 21265 + }, + { + "epoch": 3.7861467236467234, + "grad_norm": 0.8276677131652832, + "learning_rate": 1.4214522393318886e-06, + "loss": 0.6344, + "step": 21266 + }, + { + "epoch": 3.786324786324786, + "grad_norm": 0.8694467544555664, + "learning_rate": 1.4191015050894485e-06, + "loss": 0.7575, + "step": 21267 + }, + { + "epoch": 3.786502849002849, + "grad_norm": 0.917188823223114, + "learning_rate": 1.4167527023349336e-06, + "loss": 0.8464, + "step": 21268 + }, + { + "epoch": 3.7866809116809117, + "grad_norm": 1.0555243492126465, + "learning_rate": 1.4144058311143626e-06, + "loss": 0.876, + "step": 21269 + }, + { + "epoch": 3.7868589743589745, + "grad_norm": 0.8941957354545593, + "learning_rate": 1.4120608914737099e-06, + "loss": 0.7512, + "step": 21270 + }, + { + "epoch": 3.787037037037037, + "grad_norm": 1.2200781106948853, + "learning_rate": 1.4097178834589165e-06, + "loss": 0.8148, + "step": 21271 + }, + { + "epoch": 3.7872150997151, + "grad_norm": 1.1504127979278564, + "learning_rate": 1.4073768071159011e-06, + "loss": 0.7324, + "step": 21272 + }, + { + "epoch": 3.7873931623931623, + "grad_norm": 0.9139384031295776, + "learning_rate": 1.4050376624905493e-06, + "loss": 0.7327, + "step": 21273 + }, + { + "epoch": 3.787571225071225, + "grad_norm": 0.9477390050888062, + "learning_rate": 1.4027004496286466e-06, + "loss": 0.6668, + "step": 21274 + }, + { + "epoch": 3.7877492877492878, + "grad_norm": 1.0441744327545166, + "learning_rate": 1.400365168576012e-06, + "loss": 0.5605, + "step": 21275 + }, + { + "epoch": 3.7879273504273505, + "grad_norm": 0.983304500579834, + "learning_rate": 1.3980318193784091e-06, + "loss": 0.7046, + "step": 21276 + }, + { + "epoch": 3.7881054131054133, + "grad_norm": 0.8792953491210938, + "learning_rate": 1.3957004020815456e-06, + "loss": 0.6617, + "step": 21277 + }, + { + "epoch": 3.7882834757834756, + "grad_norm": 1.0233715772628784, + "learning_rate": 1.393370916731096e-06, + "loss": 0.7563, + "step": 21278 + }, + { + "epoch": 3.7884615384615383, + "grad_norm": 0.9095731377601624, + "learning_rate": 1.3910433633727127e-06, + "loss": 0.7283, + "step": 21279 + }, + { + "epoch": 3.788639601139601, + "grad_norm": 1.061542272567749, + "learning_rate": 1.3887177420519815e-06, + "loss": 0.7871, + "step": 21280 + }, + { + "epoch": 3.788817663817664, + "grad_norm": 0.9254636764526367, + "learning_rate": 1.386394052814488e-06, + "loss": 0.713, + "step": 21281 + }, + { + "epoch": 3.7889957264957266, + "grad_norm": 0.9656744003295898, + "learning_rate": 1.3840722957057406e-06, + "loss": 0.639, + "step": 21282 + }, + { + "epoch": 3.7891737891737893, + "grad_norm": 0.8546678423881531, + "learning_rate": 1.381752470771247e-06, + "loss": 0.5376, + "step": 21283 + }, + { + "epoch": 3.789351851851852, + "grad_norm": 0.923988401889801, + "learning_rate": 1.3794345780564488e-06, + "loss": 0.832, + "step": 21284 + }, + { + "epoch": 3.7895299145299144, + "grad_norm": 0.9867568612098694, + "learning_rate": 1.3771186176067653e-06, + "loss": 0.6947, + "step": 21285 + }, + { + "epoch": 3.789707977207977, + "grad_norm": 0.832467794418335, + "learning_rate": 1.374804589467571e-06, + "loss": 0.7129, + "step": 21286 + }, + { + "epoch": 3.78988603988604, + "grad_norm": 1.0230439901351929, + "learning_rate": 1.3724924936841965e-06, + "loss": 0.6923, + "step": 21287 + }, + { + "epoch": 3.7900641025641026, + "grad_norm": 0.8498647809028625, + "learning_rate": 1.3701823303019722e-06, + "loss": 0.667, + "step": 21288 + }, + { + "epoch": 3.7902421652421654, + "grad_norm": 0.9631175994873047, + "learning_rate": 1.3678740993661177e-06, + "loss": 0.6332, + "step": 21289 + }, + { + "epoch": 3.7904202279202277, + "grad_norm": 0.8047894835472107, + "learning_rate": 1.3655678009218965e-06, + "loss": 0.6037, + "step": 21290 + }, + { + "epoch": 3.7905982905982905, + "grad_norm": 1.03085196018219, + "learning_rate": 1.3632634350144614e-06, + "loss": 0.7858, + "step": 21291 + }, + { + "epoch": 3.790776353276353, + "grad_norm": 0.9661818146705627, + "learning_rate": 1.3609610016889873e-06, + "loss": 0.887, + "step": 21292 + }, + { + "epoch": 3.790954415954416, + "grad_norm": 0.8671273589134216, + "learning_rate": 1.3586605009905829e-06, + "loss": 0.8124, + "step": 21293 + }, + { + "epoch": 3.7911324786324787, + "grad_norm": 0.8627972602844238, + "learning_rate": 1.3563619329643119e-06, + "loss": 0.6146, + "step": 21294 + }, + { + "epoch": 3.7913105413105415, + "grad_norm": 1.0023887157440186, + "learning_rate": 1.3540652976552159e-06, + "loss": 0.7717, + "step": 21295 + }, + { + "epoch": 3.791488603988604, + "grad_norm": 0.8316352963447571, + "learning_rate": 1.3517705951082926e-06, + "loss": 0.6447, + "step": 21296 + }, + { + "epoch": 3.7916666666666665, + "grad_norm": 0.9875814914703369, + "learning_rate": 1.3494778253684836e-06, + "loss": 0.7261, + "step": 21297 + }, + { + "epoch": 3.7918447293447293, + "grad_norm": 0.8049336075782776, + "learning_rate": 1.347186988480753e-06, + "loss": 0.4757, + "step": 21298 + }, + { + "epoch": 3.792022792022792, + "grad_norm": 1.1316081285476685, + "learning_rate": 1.3448980844899428e-06, + "loss": 0.8487, + "step": 21299 + }, + { + "epoch": 3.7922008547008548, + "grad_norm": 1.0074068307876587, + "learning_rate": 1.3426111134409281e-06, + "loss": 0.773, + "step": 21300 + }, + { + "epoch": 3.7923789173789175, + "grad_norm": 1.0519719123840332, + "learning_rate": 1.3403260753784951e-06, + "loss": 0.7062, + "step": 21301 + }, + { + "epoch": 3.79255698005698, + "grad_norm": 0.8730934858322144, + "learning_rate": 1.3380429703474306e-06, + "loss": 0.6868, + "step": 21302 + }, + { + "epoch": 3.7927350427350426, + "grad_norm": 0.992592453956604, + "learning_rate": 1.3357617983924652e-06, + "loss": 0.7468, + "step": 21303 + }, + { + "epoch": 3.7929131054131053, + "grad_norm": 1.0449074506759644, + "learning_rate": 1.3334825595582745e-06, + "loss": 0.7136, + "step": 21304 + }, + { + "epoch": 3.793091168091168, + "grad_norm": 0.8978158831596375, + "learning_rate": 1.3312052538895559e-06, + "loss": 0.6355, + "step": 21305 + }, + { + "epoch": 3.793269230769231, + "grad_norm": 1.0906991958618164, + "learning_rate": 1.3289298814308848e-06, + "loss": 0.7947, + "step": 21306 + }, + { + "epoch": 3.7934472934472936, + "grad_norm": 1.0528931617736816, + "learning_rate": 1.32665644222687e-06, + "loss": 0.6711, + "step": 21307 + }, + { + "epoch": 3.7936253561253563, + "grad_norm": 1.055999517440796, + "learning_rate": 1.324384936322043e-06, + "loss": 0.7513, + "step": 21308 + }, + { + "epoch": 3.7938034188034186, + "grad_norm": 1.0277466773986816, + "learning_rate": 1.3221153637609119e-06, + "loss": 0.7602, + "step": 21309 + }, + { + "epoch": 3.7939814814814814, + "grad_norm": 0.9280304908752441, + "learning_rate": 1.3198477245879414e-06, + "loss": 0.7614, + "step": 21310 + }, + { + "epoch": 3.794159544159544, + "grad_norm": 1.0405571460723877, + "learning_rate": 1.317582018847574e-06, + "loss": 0.8271, + "step": 21311 + }, + { + "epoch": 3.794337606837607, + "grad_norm": 0.9499875903129578, + "learning_rate": 1.315318246584174e-06, + "loss": 0.804, + "step": 21312 + }, + { + "epoch": 3.7945156695156697, + "grad_norm": 0.9112995266914368, + "learning_rate": 1.3130564078421391e-06, + "loss": 0.6156, + "step": 21313 + }, + { + "epoch": 3.794693732193732, + "grad_norm": 0.9878358244895935, + "learning_rate": 1.3107965026657343e-06, + "loss": 0.7968, + "step": 21314 + }, + { + "epoch": 3.7948717948717947, + "grad_norm": 0.9207808971405029, + "learning_rate": 1.3085385310992904e-06, + "loss": 0.7832, + "step": 21315 + }, + { + "epoch": 3.7950498575498575, + "grad_norm": 0.8998355865478516, + "learning_rate": 1.3062824931869944e-06, + "loss": 0.6685, + "step": 21316 + }, + { + "epoch": 3.79522792022792, + "grad_norm": 0.935667872428894, + "learning_rate": 1.3040283889730886e-06, + "loss": 0.5034, + "step": 21317 + }, + { + "epoch": 3.795405982905983, + "grad_norm": 0.9035351872444153, + "learning_rate": 1.3017762185017157e-06, + "loss": 0.7745, + "step": 21318 + }, + { + "epoch": 3.7955840455840457, + "grad_norm": 1.0024009943008423, + "learning_rate": 1.299525981817018e-06, + "loss": 0.7638, + "step": 21319 + }, + { + "epoch": 3.7957621082621085, + "grad_norm": 0.9762584567070007, + "learning_rate": 1.2972776789630714e-06, + "loss": 0.9055, + "step": 21320 + }, + { + "epoch": 3.7959401709401708, + "grad_norm": 0.9953422546386719, + "learning_rate": 1.2950313099839296e-06, + "loss": 0.8011, + "step": 21321 + }, + { + "epoch": 3.7961182336182335, + "grad_norm": 0.8137484788894653, + "learning_rate": 1.2927868749236017e-06, + "loss": 0.5586, + "step": 21322 + }, + { + "epoch": 3.7962962962962963, + "grad_norm": 0.8958911299705505, + "learning_rate": 1.2905443738260747e-06, + "loss": 0.6886, + "step": 21323 + }, + { + "epoch": 3.796474358974359, + "grad_norm": 1.1359935998916626, + "learning_rate": 1.2883038067352916e-06, + "loss": 0.8033, + "step": 21324 + }, + { + "epoch": 3.796652421652422, + "grad_norm": 0.8494634628295898, + "learning_rate": 1.286065173695128e-06, + "loss": 0.71, + "step": 21325 + }, + { + "epoch": 3.796830484330484, + "grad_norm": 0.9622138738632202, + "learning_rate": 1.28382847474946e-06, + "loss": 0.6438, + "step": 21326 + }, + { + "epoch": 3.797008547008547, + "grad_norm": 0.8593050241470337, + "learning_rate": 1.2815937099421083e-06, + "loss": 0.7284, + "step": 21327 + }, + { + "epoch": 3.7971866096866096, + "grad_norm": 0.9425849318504333, + "learning_rate": 1.2793608793168488e-06, + "loss": 0.645, + "step": 21328 + }, + { + "epoch": 3.7973646723646723, + "grad_norm": 1.0602357387542725, + "learning_rate": 1.2771299829174465e-06, + "loss": 0.6953, + "step": 21329 + }, + { + "epoch": 3.797542735042735, + "grad_norm": 1.0182085037231445, + "learning_rate": 1.274901020787611e-06, + "loss": 0.851, + "step": 21330 + }, + { + "epoch": 3.797720797720798, + "grad_norm": 0.9231748580932617, + "learning_rate": 1.2726739929709852e-06, + "loss": 0.7818, + "step": 21331 + }, + { + "epoch": 3.7978988603988606, + "grad_norm": 0.8449130058288574, + "learning_rate": 1.2704488995112451e-06, + "loss": 0.6666, + "step": 21332 + }, + { + "epoch": 3.7980769230769234, + "grad_norm": 0.9367035627365112, + "learning_rate": 1.268225740451956e-06, + "loss": 0.7047, + "step": 21333 + }, + { + "epoch": 3.7982549857549857, + "grad_norm": 1.005717158317566, + "learning_rate": 1.2660045158366828e-06, + "loss": 0.7707, + "step": 21334 + }, + { + "epoch": 3.7984330484330484, + "grad_norm": 0.9248343706130981, + "learning_rate": 1.2637852257089578e-06, + "loss": 0.6067, + "step": 21335 + }, + { + "epoch": 3.798611111111111, + "grad_norm": 0.9488750696182251, + "learning_rate": 1.261567870112257e-06, + "loss": 0.7695, + "step": 21336 + }, + { + "epoch": 3.798789173789174, + "grad_norm": 0.9528114795684814, + "learning_rate": 1.259352449090001e-06, + "loss": 0.6908, + "step": 21337 + }, + { + "epoch": 3.798967236467236, + "grad_norm": 0.886128306388855, + "learning_rate": 1.2571389626856445e-06, + "loss": 0.5891, + "step": 21338 + }, + { + "epoch": 3.799145299145299, + "grad_norm": 0.9369097948074341, + "learning_rate": 1.254927410942508e-06, + "loss": 0.9413, + "step": 21339 + }, + { + "epoch": 3.7993233618233617, + "grad_norm": 0.9909436106681824, + "learning_rate": 1.2527177939039681e-06, + "loss": 0.8069, + "step": 21340 + }, + { + "epoch": 3.7995014245014245, + "grad_norm": 0.9349713921546936, + "learning_rate": 1.2505101116132678e-06, + "loss": 0.7218, + "step": 21341 + }, + { + "epoch": 3.7996794871794872, + "grad_norm": 0.912471354007721, + "learning_rate": 1.248304364113706e-06, + "loss": 0.6264, + "step": 21342 + }, + { + "epoch": 3.79985754985755, + "grad_norm": 0.9769580960273743, + "learning_rate": 1.2461005514484703e-06, + "loss": 0.6653, + "step": 21343 + }, + { + "epoch": 3.8000356125356127, + "grad_norm": 0.9078145027160645, + "learning_rate": 1.2438986736607484e-06, + "loss": 0.8129, + "step": 21344 + }, + { + "epoch": 3.8002136752136755, + "grad_norm": 0.8971534967422485, + "learning_rate": 1.2416987307937056e-06, + "loss": 0.8267, + "step": 21345 + }, + { + "epoch": 3.800391737891738, + "grad_norm": 1.091088891029358, + "learning_rate": 1.2395007228903966e-06, + "loss": 0.74, + "step": 21346 + }, + { + "epoch": 3.8005698005698005, + "grad_norm": 0.8109734058380127, + "learning_rate": 1.2373046499939311e-06, + "loss": 0.576, + "step": 21347 + }, + { + "epoch": 3.8007478632478633, + "grad_norm": 0.9560970067977905, + "learning_rate": 1.2351105121473084e-06, + "loss": 0.7023, + "step": 21348 + }, + { + "epoch": 3.800925925925926, + "grad_norm": 0.8324378132820129, + "learning_rate": 1.2329183093935382e-06, + "loss": 0.4152, + "step": 21349 + }, + { + "epoch": 3.8011039886039883, + "grad_norm": 0.8839790225028992, + "learning_rate": 1.230728041775553e-06, + "loss": 0.7523, + "step": 21350 + }, + { + "epoch": 3.801282051282051, + "grad_norm": 0.9106653928756714, + "learning_rate": 1.228539709336285e-06, + "loss": 0.8959, + "step": 21351 + }, + { + "epoch": 3.801460113960114, + "grad_norm": 1.0674225091934204, + "learning_rate": 1.2263533121186e-06, + "loss": 0.9069, + "step": 21352 + }, + { + "epoch": 3.8016381766381766, + "grad_norm": 0.904495358467102, + "learning_rate": 1.2241688501653414e-06, + "loss": 0.7724, + "step": 21353 + }, + { + "epoch": 3.8018162393162394, + "grad_norm": 0.8402371406555176, + "learning_rate": 1.2219863235192864e-06, + "loss": 0.5943, + "step": 21354 + }, + { + "epoch": 3.801994301994302, + "grad_norm": 1.0665391683578491, + "learning_rate": 1.2198057322232448e-06, + "loss": 0.9477, + "step": 21355 + }, + { + "epoch": 3.802172364672365, + "grad_norm": 0.8307034969329834, + "learning_rate": 1.2176270763198828e-06, + "loss": 0.559, + "step": 21356 + }, + { + "epoch": 3.8023504273504276, + "grad_norm": 1.033492922782898, + "learning_rate": 1.2154503558519325e-06, + "loss": 0.9119, + "step": 21357 + }, + { + "epoch": 3.80252849002849, + "grad_norm": 0.9070342779159546, + "learning_rate": 1.213275570862027e-06, + "loss": 0.7226, + "step": 21358 + }, + { + "epoch": 3.8027065527065527, + "grad_norm": 0.9109193682670593, + "learning_rate": 1.2111027213927651e-06, + "loss": 0.8976, + "step": 21359 + }, + { + "epoch": 3.8028846153846154, + "grad_norm": 0.9902597665786743, + "learning_rate": 1.208931807486735e-06, + "loss": 0.6105, + "step": 21360 + }, + { + "epoch": 3.803062678062678, + "grad_norm": 0.8371327519416809, + "learning_rate": 1.2067628291864696e-06, + "loss": 0.7772, + "step": 21361 + }, + { + "epoch": 3.8032407407407405, + "grad_norm": 0.8901026248931885, + "learning_rate": 1.204595786534457e-06, + "loss": 0.6908, + "step": 21362 + }, + { + "epoch": 3.8034188034188032, + "grad_norm": 0.9610844850540161, + "learning_rate": 1.2024306795731522e-06, + "loss": 0.7674, + "step": 21363 + }, + { + "epoch": 3.803596866096866, + "grad_norm": 1.0326725244522095, + "learning_rate": 1.2002675083449877e-06, + "loss": 0.8886, + "step": 21364 + }, + { + "epoch": 3.8037749287749287, + "grad_norm": 0.9106666445732117, + "learning_rate": 1.198106272892352e-06, + "loss": 0.7843, + "step": 21365 + }, + { + "epoch": 3.8039529914529915, + "grad_norm": 1.061942458152771, + "learning_rate": 1.195946973257578e-06, + "loss": 0.6653, + "step": 21366 + }, + { + "epoch": 3.8041310541310542, + "grad_norm": 0.9274312257766724, + "learning_rate": 1.1937896094829759e-06, + "loss": 0.7761, + "step": 21367 + }, + { + "epoch": 3.804309116809117, + "grad_norm": 0.9896464943885803, + "learning_rate": 1.1916341816108124e-06, + "loss": 0.8284, + "step": 21368 + }, + { + "epoch": 3.8044871794871797, + "grad_norm": 1.078537106513977, + "learning_rate": 1.1894806896833088e-06, + "loss": 0.851, + "step": 21369 + }, + { + "epoch": 3.804665242165242, + "grad_norm": 0.9805821180343628, + "learning_rate": 1.1873291337426873e-06, + "loss": 0.8921, + "step": 21370 + }, + { + "epoch": 3.804843304843305, + "grad_norm": 0.9240527153015137, + "learning_rate": 1.1851795138310696e-06, + "loss": 0.695, + "step": 21371 + }, + { + "epoch": 3.8050213675213675, + "grad_norm": 1.124892234802246, + "learning_rate": 1.1830318299905996e-06, + "loss": 0.7725, + "step": 21372 + }, + { + "epoch": 3.8051994301994303, + "grad_norm": 1.07109534740448, + "learning_rate": 1.180886082263355e-06, + "loss": 0.7807, + "step": 21373 + }, + { + "epoch": 3.8053774928774926, + "grad_norm": 0.8756905198097229, + "learning_rate": 1.1787422706913577e-06, + "loss": 0.686, + "step": 21374 + }, + { + "epoch": 3.8055555555555554, + "grad_norm": 0.8990702629089355, + "learning_rate": 1.1766003953166183e-06, + "loss": 0.7694, + "step": 21375 + }, + { + "epoch": 3.805733618233618, + "grad_norm": 1.0288631916046143, + "learning_rate": 1.1744604561811146e-06, + "loss": 0.789, + "step": 21376 + }, + { + "epoch": 3.805911680911681, + "grad_norm": 1.0674315690994263, + "learning_rate": 1.1723224533267686e-06, + "loss": 0.662, + "step": 21377 + }, + { + "epoch": 3.8060897435897436, + "grad_norm": 0.9958518743515015, + "learning_rate": 1.1701863867954577e-06, + "loss": 0.749, + "step": 21378 + }, + { + "epoch": 3.8062678062678064, + "grad_norm": 1.0079996585845947, + "learning_rate": 1.1680522566290375e-06, + "loss": 0.77, + "step": 21379 + }, + { + "epoch": 3.806445868945869, + "grad_norm": 0.9467402100563049, + "learning_rate": 1.1659200628693523e-06, + "loss": 0.6779, + "step": 21380 + }, + { + "epoch": 3.806623931623932, + "grad_norm": 0.8622475862503052, + "learning_rate": 1.1637898055581354e-06, + "loss": 0.7096, + "step": 21381 + }, + { + "epoch": 3.806801994301994, + "grad_norm": 0.9366708397865295, + "learning_rate": 1.1616614847371421e-06, + "loss": 0.7709, + "step": 21382 + }, + { + "epoch": 3.806980056980057, + "grad_norm": 1.184937834739685, + "learning_rate": 1.1595351004480837e-06, + "loss": 0.5492, + "step": 21383 + }, + { + "epoch": 3.8071581196581197, + "grad_norm": 1.089446783065796, + "learning_rate": 1.1574106527325933e-06, + "loss": 0.742, + "step": 21384 + }, + { + "epoch": 3.8073361823361824, + "grad_norm": 1.050498366355896, + "learning_rate": 1.155288141632338e-06, + "loss": 0.9068, + "step": 21385 + }, + { + "epoch": 3.807514245014245, + "grad_norm": 0.9292911887168884, + "learning_rate": 1.1531675671888619e-06, + "loss": 0.7896, + "step": 21386 + }, + { + "epoch": 3.8076923076923075, + "grad_norm": 1.1359772682189941, + "learning_rate": 1.1510489294437432e-06, + "loss": 0.8933, + "step": 21387 + }, + { + "epoch": 3.8078703703703702, + "grad_norm": 0.8609947562217712, + "learning_rate": 1.1489322284384596e-06, + "loss": 0.5453, + "step": 21388 + }, + { + "epoch": 3.808048433048433, + "grad_norm": 0.9479160308837891, + "learning_rate": 1.1468174642145223e-06, + "loss": 0.8166, + "step": 21389 + }, + { + "epoch": 3.8082264957264957, + "grad_norm": 0.9272875189781189, + "learning_rate": 1.144704636813343e-06, + "loss": 0.7493, + "step": 21390 + }, + { + "epoch": 3.8084045584045585, + "grad_norm": 1.0059126615524292, + "learning_rate": 1.1425937462763215e-06, + "loss": 0.8498, + "step": 21391 + }, + { + "epoch": 3.8085826210826212, + "grad_norm": 0.9803183674812317, + "learning_rate": 1.1404847926448136e-06, + "loss": 0.7432, + "step": 21392 + }, + { + "epoch": 3.808760683760684, + "grad_norm": 0.8550325632095337, + "learning_rate": 1.1383777759601533e-06, + "loss": 0.7143, + "step": 21393 + }, + { + "epoch": 3.8089387464387463, + "grad_norm": 0.9402148127555847, + "learning_rate": 1.1362726962635961e-06, + "loss": 0.6031, + "step": 21394 + }, + { + "epoch": 3.809116809116809, + "grad_norm": 1.1336814165115356, + "learning_rate": 1.1341695535964203e-06, + "loss": 0.8163, + "step": 21395 + }, + { + "epoch": 3.809294871794872, + "grad_norm": 0.8783149719238281, + "learning_rate": 1.132068347999804e-06, + "loss": 0.764, + "step": 21396 + }, + { + "epoch": 3.8094729344729346, + "grad_norm": 0.9727262854576111, + "learning_rate": 1.1299690795149365e-06, + "loss": 0.7521, + "step": 21397 + }, + { + "epoch": 3.8096509971509973, + "grad_norm": 0.9679130911827087, + "learning_rate": 1.1278717481829403e-06, + "loss": 0.6411, + "step": 21398 + }, + { + "epoch": 3.8098290598290596, + "grad_norm": 1.0554420948028564, + "learning_rate": 1.125776354044905e-06, + "loss": 0.738, + "step": 21399 + }, + { + "epoch": 3.8100071225071224, + "grad_norm": 1.0295968055725098, + "learning_rate": 1.1236828971418867e-06, + "loss": 0.6789, + "step": 21400 + }, + { + "epoch": 3.810185185185185, + "grad_norm": 1.2897826433181763, + "learning_rate": 1.1215913775149079e-06, + "loss": 0.8072, + "step": 21401 + }, + { + "epoch": 3.810363247863248, + "grad_norm": 0.8948628902435303, + "learning_rate": 1.119501795204947e-06, + "loss": 0.5636, + "step": 21402 + }, + { + "epoch": 3.8105413105413106, + "grad_norm": 1.040513277053833, + "learning_rate": 1.1174141502529268e-06, + "loss": 0.9099, + "step": 21403 + }, + { + "epoch": 3.8107193732193734, + "grad_norm": 1.0494636297225952, + "learning_rate": 1.1153284426997811e-06, + "loss": 0.9003, + "step": 21404 + }, + { + "epoch": 3.810897435897436, + "grad_norm": 0.9162052273750305, + "learning_rate": 1.1132446725863554e-06, + "loss": 0.7323, + "step": 21405 + }, + { + "epoch": 3.8110754985754984, + "grad_norm": 1.1157633066177368, + "learning_rate": 1.1111628399534724e-06, + "loss": 0.731, + "step": 21406 + }, + { + "epoch": 3.811253561253561, + "grad_norm": 0.9189321994781494, + "learning_rate": 1.1090829448419438e-06, + "loss": 0.8703, + "step": 21407 + }, + { + "epoch": 3.811431623931624, + "grad_norm": 1.1154193878173828, + "learning_rate": 1.1070049872925037e-06, + "loss": 0.8683, + "step": 21408 + }, + { + "epoch": 3.8116096866096867, + "grad_norm": 0.8818658590316772, + "learning_rate": 1.104928967345853e-06, + "loss": 0.6899, + "step": 21409 + }, + { + "epoch": 3.8117877492877494, + "grad_norm": 1.0003615617752075, + "learning_rate": 1.1028548850427034e-06, + "loss": 0.763, + "step": 21410 + }, + { + "epoch": 3.8119658119658117, + "grad_norm": 0.9913126230239868, + "learning_rate": 1.100782740423656e-06, + "loss": 1.0265, + "step": 21411 + }, + { + "epoch": 3.8121438746438745, + "grad_norm": 0.982927143573761, + "learning_rate": 1.0987125335293448e-06, + "loss": 0.7146, + "step": 21412 + }, + { + "epoch": 3.8123219373219372, + "grad_norm": 1.0777631998062134, + "learning_rate": 1.0966442644002928e-06, + "loss": 0.8946, + "step": 21413 + }, + { + "epoch": 3.8125, + "grad_norm": 0.885434627532959, + "learning_rate": 1.0945779330770457e-06, + "loss": 0.5987, + "step": 21414 + }, + { + "epoch": 3.8126780626780628, + "grad_norm": 1.2291704416275024, + "learning_rate": 1.092513539600093e-06, + "loss": 0.7442, + "step": 21415 + }, + { + "epoch": 3.8128561253561255, + "grad_norm": 0.976157009601593, + "learning_rate": 1.0904510840098692e-06, + "loss": 0.9357, + "step": 21416 + }, + { + "epoch": 3.8130341880341883, + "grad_norm": 1.0253404378890991, + "learning_rate": 1.0883905663467974e-06, + "loss": 0.8082, + "step": 21417 + }, + { + "epoch": 3.8132122507122506, + "grad_norm": 0.9216427206993103, + "learning_rate": 1.0863319866512346e-06, + "loss": 0.7686, + "step": 21418 + }, + { + "epoch": 3.8133903133903133, + "grad_norm": 0.903035581111908, + "learning_rate": 1.0842753449635147e-06, + "loss": 0.8007, + "step": 21419 + }, + { + "epoch": 3.813568376068376, + "grad_norm": 0.8471698760986328, + "learning_rate": 1.0822206413239499e-06, + "loss": 0.5844, + "step": 21420 + }, + { + "epoch": 3.813746438746439, + "grad_norm": 0.9668221473693848, + "learning_rate": 1.0801678757727862e-06, + "loss": 0.8883, + "step": 21421 + }, + { + "epoch": 3.8139245014245016, + "grad_norm": 1.026482343673706, + "learning_rate": 1.0781170483502355e-06, + "loss": 0.7599, + "step": 21422 + }, + { + "epoch": 3.814102564102564, + "grad_norm": 0.9170508980751038, + "learning_rate": 1.0760681590964995e-06, + "loss": 0.6353, + "step": 21423 + }, + { + "epoch": 3.8142806267806266, + "grad_norm": 0.9856457710266113, + "learning_rate": 1.074021208051712e-06, + "loss": 0.7551, + "step": 21424 + }, + { + "epoch": 3.8144586894586894, + "grad_norm": 0.8754544854164124, + "learning_rate": 1.0719761952559748e-06, + "loss": 0.577, + "step": 21425 + }, + { + "epoch": 3.814636752136752, + "grad_norm": 0.867774248123169, + "learning_rate": 1.0699331207493556e-06, + "loss": 0.6398, + "step": 21426 + }, + { + "epoch": 3.814814814814815, + "grad_norm": 1.0027731657028198, + "learning_rate": 1.0678919845718892e-06, + "loss": 0.8892, + "step": 21427 + }, + { + "epoch": 3.8149928774928776, + "grad_norm": 0.8579356670379639, + "learning_rate": 1.065852786763566e-06, + "loss": 0.8224, + "step": 21428 + }, + { + "epoch": 3.8151709401709404, + "grad_norm": 1.1910187005996704, + "learning_rate": 1.0638155273643425e-06, + "loss": 0.9737, + "step": 21429 + }, + { + "epoch": 3.8153490028490027, + "grad_norm": 0.9423369765281677, + "learning_rate": 1.0617802064141313e-06, + "loss": 0.7792, + "step": 21430 + }, + { + "epoch": 3.8155270655270654, + "grad_norm": 1.0974822044372559, + "learning_rate": 1.0597468239528118e-06, + "loss": 0.6306, + "step": 21431 + }, + { + "epoch": 3.815705128205128, + "grad_norm": 0.8709555268287659, + "learning_rate": 1.0577153800202188e-06, + "loss": 0.6959, + "step": 21432 + }, + { + "epoch": 3.815883190883191, + "grad_norm": 0.9723543524742126, + "learning_rate": 1.0556858746561538e-06, + "loss": 0.7802, + "step": 21433 + }, + { + "epoch": 3.8160612535612537, + "grad_norm": 0.959344208240509, + "learning_rate": 1.0536583079003958e-06, + "loss": 0.924, + "step": 21434 + }, + { + "epoch": 3.816239316239316, + "grad_norm": 1.0029487609863281, + "learning_rate": 1.051632679792658e-06, + "loss": 0.9172, + "step": 21435 + }, + { + "epoch": 3.8164173789173788, + "grad_norm": 0.9423520565032959, + "learning_rate": 1.0496089903726192e-06, + "loss": 0.8802, + "step": 21436 + }, + { + "epoch": 3.8165954415954415, + "grad_norm": 0.9231133460998535, + "learning_rate": 1.047587239679959e-06, + "loss": 0.842, + "step": 21437 + }, + { + "epoch": 3.8167735042735043, + "grad_norm": 0.9937446713447571, + "learning_rate": 1.0455674277542459e-06, + "loss": 0.5957, + "step": 21438 + }, + { + "epoch": 3.816951566951567, + "grad_norm": 1.1649292707443237, + "learning_rate": 1.0435495546351036e-06, + "loss": 0.8486, + "step": 21439 + }, + { + "epoch": 3.8171296296296298, + "grad_norm": 0.8023452758789062, + "learning_rate": 1.041533620362034e-06, + "loss": 0.7446, + "step": 21440 + }, + { + "epoch": 3.8173076923076925, + "grad_norm": 0.7962468266487122, + "learning_rate": 1.0395196249745387e-06, + "loss": 0.6481, + "step": 21441 + }, + { + "epoch": 3.817485754985755, + "grad_norm": 0.9893062114715576, + "learning_rate": 1.0375075685120972e-06, + "loss": 0.7427, + "step": 21442 + }, + { + "epoch": 3.8176638176638176, + "grad_norm": 0.9541147947311401, + "learning_rate": 1.0354974510141002e-06, + "loss": 0.7193, + "step": 21443 + }, + { + "epoch": 3.8178418803418803, + "grad_norm": 1.0776668787002563, + "learning_rate": 1.0334892725199607e-06, + "loss": 0.9832, + "step": 21444 + }, + { + "epoch": 3.818019943019943, + "grad_norm": 1.0417940616607666, + "learning_rate": 1.0314830330690139e-06, + "loss": 0.6436, + "step": 21445 + }, + { + "epoch": 3.818198005698006, + "grad_norm": 0.8967922925949097, + "learning_rate": 1.0294787327005729e-06, + "loss": 0.7823, + "step": 21446 + }, + { + "epoch": 3.818376068376068, + "grad_norm": 0.9486142992973328, + "learning_rate": 1.027476371453895e-06, + "loss": 0.8485, + "step": 21447 + }, + { + "epoch": 3.818554131054131, + "grad_norm": 1.0096664428710938, + "learning_rate": 1.0254759493682152e-06, + "loss": 0.6872, + "step": 21448 + }, + { + "epoch": 3.8187321937321936, + "grad_norm": 1.0282636880874634, + "learning_rate": 1.0234774664827474e-06, + "loss": 1.0026, + "step": 21449 + }, + { + "epoch": 3.8189102564102564, + "grad_norm": 0.975347638130188, + "learning_rate": 1.0214809228366262e-06, + "loss": 0.8115, + "step": 21450 + }, + { + "epoch": 3.819088319088319, + "grad_norm": 1.000891089439392, + "learning_rate": 1.0194863184689652e-06, + "loss": 0.9413, + "step": 21451 + }, + { + "epoch": 3.819266381766382, + "grad_norm": 0.9597340226173401, + "learning_rate": 1.0174936534188662e-06, + "loss": 0.6959, + "step": 21452 + }, + { + "epoch": 3.8194444444444446, + "grad_norm": 1.018418788909912, + "learning_rate": 1.0155029277253537e-06, + "loss": 0.7004, + "step": 21453 + }, + { + "epoch": 3.8196225071225074, + "grad_norm": 1.005155086517334, + "learning_rate": 1.0135141414274519e-06, + "loss": 0.9321, + "step": 21454 + }, + { + "epoch": 3.8198005698005697, + "grad_norm": 0.8870786428451538, + "learning_rate": 1.0115272945641075e-06, + "loss": 0.9596, + "step": 21455 + }, + { + "epoch": 3.8199786324786325, + "grad_norm": 0.8494182229042053, + "learning_rate": 1.009542387174256e-06, + "loss": 0.7022, + "step": 21456 + }, + { + "epoch": 3.820156695156695, + "grad_norm": 1.0547996759414673, + "learning_rate": 1.0075594192967774e-06, + "loss": 0.7463, + "step": 21457 + }, + { + "epoch": 3.820334757834758, + "grad_norm": 0.8766087889671326, + "learning_rate": 1.0055783909705406e-06, + "loss": 0.5984, + "step": 21458 + }, + { + "epoch": 3.8205128205128203, + "grad_norm": 0.9966375827789307, + "learning_rate": 1.0035993022343592e-06, + "loss": 0.7422, + "step": 21459 + }, + { + "epoch": 3.820690883190883, + "grad_norm": 0.8210493922233582, + "learning_rate": 1.0016221531269909e-06, + "loss": 0.78, + "step": 21460 + }, + { + "epoch": 3.8208689458689458, + "grad_norm": 0.9703954458236694, + "learning_rate": 9.996469436871824e-07, + "loss": 0.756, + "step": 21461 + }, + { + "epoch": 3.8210470085470085, + "grad_norm": 0.9013912677764893, + "learning_rate": 9.976736739536474e-07, + "loss": 0.6361, + "step": 21462 + }, + { + "epoch": 3.8212250712250713, + "grad_norm": 0.8628489971160889, + "learning_rate": 9.957023439650327e-07, + "loss": 0.5448, + "step": 21463 + }, + { + "epoch": 3.821403133903134, + "grad_norm": 1.0915628671646118, + "learning_rate": 9.937329537599627e-07, + "loss": 0.8572, + "step": 21464 + }, + { + "epoch": 3.8215811965811968, + "grad_norm": 0.9227813482284546, + "learning_rate": 9.9176550337704e-07, + "loss": 0.669, + "step": 21465 + }, + { + "epoch": 3.8217592592592595, + "grad_norm": 0.9919458031654358, + "learning_rate": 9.897999928547784e-07, + "loss": 0.7664, + "step": 21466 + }, + { + "epoch": 3.821937321937322, + "grad_norm": 0.9025664925575256, + "learning_rate": 9.87836422231736e-07, + "loss": 0.6215, + "step": 21467 + }, + { + "epoch": 3.8221153846153846, + "grad_norm": 1.154800295829773, + "learning_rate": 9.858747915463374e-07, + "loss": 0.9609, + "step": 21468 + }, + { + "epoch": 3.8222934472934473, + "grad_norm": 1.0049433708190918, + "learning_rate": 9.839151008370629e-07, + "loss": 0.7946, + "step": 21469 + }, + { + "epoch": 3.82247150997151, + "grad_norm": 0.9646359086036682, + "learning_rate": 9.819573501422596e-07, + "loss": 0.793, + "step": 21470 + }, + { + "epoch": 3.8226495726495724, + "grad_norm": 0.8603723049163818, + "learning_rate": 9.800015395003192e-07, + "loss": 0.5834, + "step": 21471 + }, + { + "epoch": 3.822827635327635, + "grad_norm": 0.9426991939544678, + "learning_rate": 9.780476689495554e-07, + "loss": 0.7321, + "step": 21472 + }, + { + "epoch": 3.823005698005698, + "grad_norm": 0.8795328140258789, + "learning_rate": 9.760957385282488e-07, + "loss": 0.7916, + "step": 21473 + }, + { + "epoch": 3.8231837606837606, + "grad_norm": 1.0602772235870361, + "learning_rate": 9.741457482746352e-07, + "loss": 0.7095, + "step": 21474 + }, + { + "epoch": 3.8233618233618234, + "grad_norm": 1.0050617456436157, + "learning_rate": 9.721976982269287e-07, + "loss": 0.8111, + "step": 21475 + }, + { + "epoch": 3.823539886039886, + "grad_norm": 1.0178896188735962, + "learning_rate": 9.702515884232878e-07, + "loss": 0.6689, + "step": 21476 + }, + { + "epoch": 3.823717948717949, + "grad_norm": 1.0118085145950317, + "learning_rate": 9.683074189018704e-07, + "loss": 0.8217, + "step": 21477 + }, + { + "epoch": 3.8238960113960117, + "grad_norm": 1.161960482597351, + "learning_rate": 9.663651897007353e-07, + "loss": 0.8104, + "step": 21478 + }, + { + "epoch": 3.824074074074074, + "grad_norm": 1.0038970708847046, + "learning_rate": 9.64424900857952e-07, + "loss": 0.9059, + "step": 21479 + }, + { + "epoch": 3.8242521367521367, + "grad_norm": 0.9953527450561523, + "learning_rate": 9.624865524115346e-07, + "loss": 0.8837, + "step": 21480 + }, + { + "epoch": 3.8244301994301995, + "grad_norm": 0.9455059766769409, + "learning_rate": 9.605501443994524e-07, + "loss": 0.5752, + "step": 21481 + }, + { + "epoch": 3.824608262108262, + "grad_norm": 0.9261574149131775, + "learning_rate": 9.586156768596645e-07, + "loss": 0.8545, + "step": 21482 + }, + { + "epoch": 3.8247863247863245, + "grad_norm": 1.0939536094665527, + "learning_rate": 9.56683149830062e-07, + "loss": 0.7738, + "step": 21483 + }, + { + "epoch": 3.8249643874643873, + "grad_norm": 0.9014570116996765, + "learning_rate": 9.547525633485044e-07, + "loss": 0.7568, + "step": 21484 + }, + { + "epoch": 3.82514245014245, + "grad_norm": 0.9881065487861633, + "learning_rate": 9.528239174528165e-07, + "loss": 0.7546, + "step": 21485 + }, + { + "epoch": 3.8253205128205128, + "grad_norm": 0.94111168384552, + "learning_rate": 9.508972121808013e-07, + "loss": 0.7925, + "step": 21486 + }, + { + "epoch": 3.8254985754985755, + "grad_norm": 0.8885481953620911, + "learning_rate": 9.489724475701955e-07, + "loss": 0.6069, + "step": 21487 + }, + { + "epoch": 3.8256766381766383, + "grad_norm": 1.0248520374298096, + "learning_rate": 9.470496236587134e-07, + "loss": 0.9262, + "step": 21488 + }, + { + "epoch": 3.825854700854701, + "grad_norm": 0.9528713226318359, + "learning_rate": 9.451287404840247e-07, + "loss": 0.7226, + "step": 21489 + }, + { + "epoch": 3.826032763532764, + "grad_norm": 0.983078122138977, + "learning_rate": 9.432097980837773e-07, + "loss": 0.651, + "step": 21490 + }, + { + "epoch": 3.826210826210826, + "grad_norm": 1.0022083520889282, + "learning_rate": 9.41292796495552e-07, + "loss": 0.7214, + "step": 21491 + }, + { + "epoch": 3.826388888888889, + "grad_norm": 0.838897168636322, + "learning_rate": 9.3937773575693e-07, + "loss": 0.5388, + "step": 21492 + }, + { + "epoch": 3.8265669515669516, + "grad_norm": 0.9708632230758667, + "learning_rate": 9.374646159054146e-07, + "loss": 0.7681, + "step": 21493 + }, + { + "epoch": 3.8267450142450143, + "grad_norm": 0.9692725539207458, + "learning_rate": 9.355534369784979e-07, + "loss": 0.8608, + "step": 21494 + }, + { + "epoch": 3.8269230769230766, + "grad_norm": 1.0063785314559937, + "learning_rate": 9.336441990136169e-07, + "loss": 0.8862, + "step": 21495 + }, + { + "epoch": 3.8271011396011394, + "grad_norm": 0.9906958937644958, + "learning_rate": 9.317369020481858e-07, + "loss": 0.7091, + "step": 21496 + }, + { + "epoch": 3.827279202279202, + "grad_norm": 0.9517685770988464, + "learning_rate": 9.298315461195751e-07, + "loss": 0.8612, + "step": 21497 + }, + { + "epoch": 3.827457264957265, + "grad_norm": 1.061146855354309, + "learning_rate": 9.279281312651211e-07, + "loss": 0.8362, + "step": 21498 + }, + { + "epoch": 3.8276353276353277, + "grad_norm": 1.1083885431289673, + "learning_rate": 9.260266575221055e-07, + "loss": 0.7663, + "step": 21499 + }, + { + "epoch": 3.8278133903133904, + "grad_norm": 0.9375349879264832, + "learning_rate": 9.241271249277872e-07, + "loss": 0.682, + "step": 21500 + }, + { + "epoch": 3.827991452991453, + "grad_norm": 1.0170559883117676, + "learning_rate": 9.222295335193809e-07, + "loss": 0.6219, + "step": 21501 + }, + { + "epoch": 3.828169515669516, + "grad_norm": 0.9686765670776367, + "learning_rate": 9.203338833340791e-07, + "loss": 0.6775, + "step": 21502 + }, + { + "epoch": 3.828347578347578, + "grad_norm": 0.8623543977737427, + "learning_rate": 9.184401744090188e-07, + "loss": 0.7056, + "step": 21503 + }, + { + "epoch": 3.828525641025641, + "grad_norm": 0.8922286629676819, + "learning_rate": 9.165484067812924e-07, + "loss": 0.8155, + "step": 21504 + }, + { + "epoch": 3.8287037037037037, + "grad_norm": 1.0283470153808594, + "learning_rate": 9.146585804879704e-07, + "loss": 0.8419, + "step": 21505 + }, + { + "epoch": 3.8288817663817665, + "grad_norm": 0.9561583399772644, + "learning_rate": 9.127706955660898e-07, + "loss": 0.8576, + "step": 21506 + }, + { + "epoch": 3.8290598290598292, + "grad_norm": 1.0040736198425293, + "learning_rate": 9.10884752052621e-07, + "loss": 0.6805, + "step": 21507 + }, + { + "epoch": 3.8292378917378915, + "grad_norm": 0.9792984127998352, + "learning_rate": 9.090007499845232e-07, + "loss": 1.0002, + "step": 21508 + }, + { + "epoch": 3.8294159544159543, + "grad_norm": 0.9882341623306274, + "learning_rate": 9.071186893987338e-07, + "loss": 0.8826, + "step": 21509 + }, + { + "epoch": 3.829594017094017, + "grad_norm": 0.9094791412353516, + "learning_rate": 9.052385703320787e-07, + "loss": 0.7015, + "step": 21510 + }, + { + "epoch": 3.82977207977208, + "grad_norm": 0.9043287038803101, + "learning_rate": 9.033603928214395e-07, + "loss": 0.6868, + "step": 21511 + }, + { + "epoch": 3.8299501424501425, + "grad_norm": 1.0372240543365479, + "learning_rate": 9.014841569035981e-07, + "loss": 0.7217, + "step": 21512 + }, + { + "epoch": 3.8301282051282053, + "grad_norm": 0.8339055776596069, + "learning_rate": 8.996098626153138e-07, + "loss": 0.7506, + "step": 21513 + }, + { + "epoch": 3.830306267806268, + "grad_norm": 0.9318352341651917, + "learning_rate": 8.977375099933017e-07, + "loss": 0.7919, + "step": 21514 + }, + { + "epoch": 3.8304843304843303, + "grad_norm": 0.9269062876701355, + "learning_rate": 8.958670990742657e-07, + "loss": 0.782, + "step": 21515 + }, + { + "epoch": 3.830662393162393, + "grad_norm": 1.2074931859970093, + "learning_rate": 8.939986298948322e-07, + "loss": 0.7641, + "step": 21516 + }, + { + "epoch": 3.830840455840456, + "grad_norm": 0.8946232199668884, + "learning_rate": 8.921321024916384e-07, + "loss": 0.7086, + "step": 21517 + }, + { + "epoch": 3.8310185185185186, + "grad_norm": 1.0376194715499878, + "learning_rate": 8.90267516901222e-07, + "loss": 0.7484, + "step": 21518 + }, + { + "epoch": 3.8311965811965814, + "grad_norm": 0.9210571050643921, + "learning_rate": 8.884048731601424e-07, + "loss": 0.7251, + "step": 21519 + }, + { + "epoch": 3.8313746438746437, + "grad_norm": 0.8585400581359863, + "learning_rate": 8.865441713048706e-07, + "loss": 0.9469, + "step": 21520 + }, + { + "epoch": 3.8315527065527064, + "grad_norm": 0.9701051712036133, + "learning_rate": 8.846854113718772e-07, + "loss": 0.6994, + "step": 21521 + }, + { + "epoch": 3.831730769230769, + "grad_norm": 1.0702025890350342, + "learning_rate": 8.828285933975888e-07, + "loss": 0.918, + "step": 21522 + }, + { + "epoch": 3.831908831908832, + "grad_norm": 0.887536883354187, + "learning_rate": 8.809737174183652e-07, + "loss": 0.966, + "step": 21523 + }, + { + "epoch": 3.8320868945868947, + "grad_norm": 0.9875296354293823, + "learning_rate": 8.791207834705662e-07, + "loss": 0.6294, + "step": 21524 + }, + { + "epoch": 3.8322649572649574, + "grad_norm": 0.734708845615387, + "learning_rate": 8.772697915904848e-07, + "loss": 0.573, + "step": 21525 + }, + { + "epoch": 3.83244301994302, + "grad_norm": 1.1104393005371094, + "learning_rate": 8.754207418143923e-07, + "loss": 0.7678, + "step": 21526 + }, + { + "epoch": 3.8326210826210825, + "grad_norm": 1.0131367444992065, + "learning_rate": 8.735736341785261e-07, + "loss": 0.7966, + "step": 21527 + }, + { + "epoch": 3.8327991452991452, + "grad_norm": 0.8396084904670715, + "learning_rate": 8.717284687190575e-07, + "loss": 0.5308, + "step": 21528 + }, + { + "epoch": 3.832977207977208, + "grad_norm": 0.9788475036621094, + "learning_rate": 8.698852454721573e-07, + "loss": 0.5804, + "step": 21529 + }, + { + "epoch": 3.8331552706552707, + "grad_norm": 0.97635418176651, + "learning_rate": 8.680439644739191e-07, + "loss": 0.7716, + "step": 21530 + }, + { + "epoch": 3.8333333333333335, + "grad_norm": 0.9419910311698914, + "learning_rate": 8.662046257604472e-07, + "loss": 0.7311, + "step": 21531 + }, + { + "epoch": 3.833511396011396, + "grad_norm": 0.8974188566207886, + "learning_rate": 8.643672293677463e-07, + "loss": 0.7538, + "step": 21532 + }, + { + "epoch": 3.8336894586894585, + "grad_norm": 0.9138547778129578, + "learning_rate": 8.625317753318318e-07, + "loss": 0.718, + "step": 21533 + }, + { + "epoch": 3.8338675213675213, + "grad_norm": 0.9484068751335144, + "learning_rate": 8.606982636886862e-07, + "loss": 0.766, + "step": 21534 + }, + { + "epoch": 3.834045584045584, + "grad_norm": 1.105020523071289, + "learning_rate": 8.588666944741918e-07, + "loss": 0.7913, + "step": 21535 + }, + { + "epoch": 3.834223646723647, + "grad_norm": 1.0500797033309937, + "learning_rate": 8.570370677242756e-07, + "loss": 0.8379, + "step": 21536 + }, + { + "epoch": 3.8344017094017095, + "grad_norm": 0.9115577340126038, + "learning_rate": 8.552093834747532e-07, + "loss": 0.7894, + "step": 21537 + }, + { + "epoch": 3.8345797720797723, + "grad_norm": 0.9908710718154907, + "learning_rate": 8.533836417614516e-07, + "loss": 0.7978, + "step": 21538 + }, + { + "epoch": 3.8347578347578346, + "grad_norm": 0.9764583706855774, + "learning_rate": 8.515598426201421e-07, + "loss": 0.6787, + "step": 21539 + }, + { + "epoch": 3.8349358974358974, + "grad_norm": 0.9413735270500183, + "learning_rate": 8.497379860865518e-07, + "loss": 0.7627, + "step": 21540 + }, + { + "epoch": 3.83511396011396, + "grad_norm": 1.0700278282165527, + "learning_rate": 8.479180721963853e-07, + "loss": 0.8845, + "step": 21541 + }, + { + "epoch": 3.835292022792023, + "grad_norm": 1.0641980171203613, + "learning_rate": 8.461001009852809e-07, + "loss": 0.7407, + "step": 21542 + }, + { + "epoch": 3.8354700854700856, + "grad_norm": 1.0877798795700073, + "learning_rate": 8.442840724888768e-07, + "loss": 0.8741, + "step": 21543 + }, + { + "epoch": 3.835648148148148, + "grad_norm": 1.1364705562591553, + "learning_rate": 8.424699867427444e-07, + "loss": 0.973, + "step": 21544 + }, + { + "epoch": 3.8358262108262107, + "grad_norm": 0.9369027614593506, + "learning_rate": 8.40657843782433e-07, + "loss": 0.8727, + "step": 21545 + }, + { + "epoch": 3.8360042735042734, + "grad_norm": 0.9268773794174194, + "learning_rate": 8.388476436434478e-07, + "loss": 0.7735, + "step": 21546 + }, + { + "epoch": 3.836182336182336, + "grad_norm": 1.2019761800765991, + "learning_rate": 8.370393863612602e-07, + "loss": 0.8365, + "step": 21547 + }, + { + "epoch": 3.836360398860399, + "grad_norm": 0.8963661789894104, + "learning_rate": 8.352330719712753e-07, + "loss": 0.5905, + "step": 21548 + }, + { + "epoch": 3.8365384615384617, + "grad_norm": 1.0327519178390503, + "learning_rate": 8.334287005089203e-07, + "loss": 0.6485, + "step": 21549 + }, + { + "epoch": 3.8367165242165244, + "grad_norm": 0.9324591755867004, + "learning_rate": 8.316262720095114e-07, + "loss": 0.7034, + "step": 21550 + }, + { + "epoch": 3.8368945868945867, + "grad_norm": 0.9687849879264832, + "learning_rate": 8.29825786508387e-07, + "loss": 0.9147, + "step": 21551 + }, + { + "epoch": 3.8370726495726495, + "grad_norm": 0.9329792261123657, + "learning_rate": 8.280272440408298e-07, + "loss": 0.6578, + "step": 21552 + }, + { + "epoch": 3.8372507122507122, + "grad_norm": 0.9121544361114502, + "learning_rate": 8.262306446420565e-07, + "loss": 0.7456, + "step": 21553 + }, + { + "epoch": 3.837428774928775, + "grad_norm": 0.9514213800430298, + "learning_rate": 8.244359883472719e-07, + "loss": 0.8341, + "step": 21554 + }, + { + "epoch": 3.8376068376068377, + "grad_norm": 1.020097017288208, + "learning_rate": 8.22643275191659e-07, + "loss": 0.9089, + "step": 21555 + }, + { + "epoch": 3.8377849002849, + "grad_norm": 0.9623993635177612, + "learning_rate": 8.208525052103233e-07, + "loss": 0.8057, + "step": 21556 + }, + { + "epoch": 3.837962962962963, + "grad_norm": 0.9708266854286194, + "learning_rate": 8.190636784383477e-07, + "loss": 0.7576, + "step": 21557 + }, + { + "epoch": 3.8381410256410255, + "grad_norm": 0.9058563709259033, + "learning_rate": 8.172767949107929e-07, + "loss": 0.7129, + "step": 21558 + }, + { + "epoch": 3.8383190883190883, + "grad_norm": 0.9107457399368286, + "learning_rate": 8.154918546626755e-07, + "loss": 0.6841, + "step": 21559 + }, + { + "epoch": 3.838497150997151, + "grad_norm": 1.144352674484253, + "learning_rate": 8.137088577289453e-07, + "loss": 0.905, + "step": 21560 + }, + { + "epoch": 3.838675213675214, + "grad_norm": 1.0036160945892334, + "learning_rate": 8.119278041445522e-07, + "loss": 0.8943, + "step": 21561 + }, + { + "epoch": 3.8388532763532766, + "grad_norm": 0.8285079002380371, + "learning_rate": 8.101486939443903e-07, + "loss": 0.5747, + "step": 21562 + }, + { + "epoch": 3.8390313390313393, + "grad_norm": 0.8895573616027832, + "learning_rate": 8.083715271633097e-07, + "loss": 0.7271, + "step": 21563 + }, + { + "epoch": 3.8392094017094016, + "grad_norm": 1.020896553993225, + "learning_rate": 8.06596303836138e-07, + "loss": 0.6835, + "step": 21564 + }, + { + "epoch": 3.8393874643874644, + "grad_norm": 0.9939431548118591, + "learning_rate": 8.048230239976584e-07, + "loss": 0.9226, + "step": 21565 + }, + { + "epoch": 3.839565527065527, + "grad_norm": 1.0216248035430908, + "learning_rate": 8.030516876826099e-07, + "loss": 0.5721, + "step": 21566 + }, + { + "epoch": 3.83974358974359, + "grad_norm": 0.9228139519691467, + "learning_rate": 8.012822949256982e-07, + "loss": 0.8811, + "step": 21567 + }, + { + "epoch": 3.839921652421652, + "grad_norm": 1.1747238636016846, + "learning_rate": 7.995148457615953e-07, + "loss": 0.8204, + "step": 21568 + }, + { + "epoch": 3.840099715099715, + "grad_norm": 0.9457100033760071, + "learning_rate": 7.977493402249292e-07, + "loss": 0.8442, + "step": 21569 + }, + { + "epoch": 3.8402777777777777, + "grad_norm": 0.865328311920166, + "learning_rate": 7.959857783502833e-07, + "loss": 0.9258, + "step": 21570 + }, + { + "epoch": 3.8404558404558404, + "grad_norm": 1.0624295473098755, + "learning_rate": 7.9422416017223e-07, + "loss": 0.7901, + "step": 21571 + }, + { + "epoch": 3.840633903133903, + "grad_norm": 0.9670442938804626, + "learning_rate": 7.924644857252639e-07, + "loss": 0.7427, + "step": 21572 + }, + { + "epoch": 3.840811965811966, + "grad_norm": 0.992448091506958, + "learning_rate": 7.907067550438685e-07, + "loss": 0.7893, + "step": 21573 + }, + { + "epoch": 3.8409900284900287, + "grad_norm": 0.8243263363838196, + "learning_rate": 7.889509681624941e-07, + "loss": 0.5991, + "step": 21574 + }, + { + "epoch": 3.8411680911680914, + "grad_norm": 0.839352011680603, + "learning_rate": 7.871971251155131e-07, + "loss": 0.7611, + "step": 21575 + }, + { + "epoch": 3.8413461538461537, + "grad_norm": 0.9567636251449585, + "learning_rate": 7.854452259373313e-07, + "loss": 0.657, + "step": 21576 + }, + { + "epoch": 3.8415242165242165, + "grad_norm": 0.9481256008148193, + "learning_rate": 7.836952706622325e-07, + "loss": 0.8136, + "step": 21577 + }, + { + "epoch": 3.8417022792022792, + "grad_norm": 0.8747114539146423, + "learning_rate": 7.819472593245225e-07, + "loss": 0.6968, + "step": 21578 + }, + { + "epoch": 3.841880341880342, + "grad_norm": 0.9657062888145447, + "learning_rate": 7.802011919584518e-07, + "loss": 0.9184, + "step": 21579 + }, + { + "epoch": 3.8420584045584043, + "grad_norm": 1.0970104932785034, + "learning_rate": 7.784570685982262e-07, + "loss": 0.7596, + "step": 21580 + }, + { + "epoch": 3.842236467236467, + "grad_norm": 0.9810351729393005, + "learning_rate": 7.767148892780296e-07, + "loss": 0.9235, + "step": 21581 + }, + { + "epoch": 3.84241452991453, + "grad_norm": 1.017484426498413, + "learning_rate": 7.74974654031968e-07, + "loss": 0.9001, + "step": 21582 + }, + { + "epoch": 3.8425925925925926, + "grad_norm": 0.9285971522331238, + "learning_rate": 7.732363628941696e-07, + "loss": 0.693, + "step": 21583 + }, + { + "epoch": 3.8427706552706553, + "grad_norm": 0.8508382439613342, + "learning_rate": 7.71500015898674e-07, + "loss": 0.7199, + "step": 21584 + }, + { + "epoch": 3.842948717948718, + "grad_norm": 1.021244764328003, + "learning_rate": 7.697656130795094e-07, + "loss": 0.7583, + "step": 21585 + }, + { + "epoch": 3.843126780626781, + "grad_norm": 1.202195167541504, + "learning_rate": 7.680331544706598e-07, + "loss": 0.9845, + "step": 21586 + }, + { + "epoch": 3.8433048433048436, + "grad_norm": 0.9475895762443542, + "learning_rate": 7.663026401060535e-07, + "loss": 0.7508, + "step": 21587 + }, + { + "epoch": 3.843482905982906, + "grad_norm": 0.9706496000289917, + "learning_rate": 7.64574070019608e-07, + "loss": 0.7399, + "step": 21588 + }, + { + "epoch": 3.8436609686609686, + "grad_norm": 0.9435704946517944, + "learning_rate": 7.628474442451961e-07, + "loss": 0.7967, + "step": 21589 + }, + { + "epoch": 3.8438390313390314, + "grad_norm": 0.9195558428764343, + "learning_rate": 7.61122762816635e-07, + "loss": 0.8477, + "step": 21590 + }, + { + "epoch": 3.844017094017094, + "grad_norm": 0.8691080212593079, + "learning_rate": 7.594000257677314e-07, + "loss": 0.7367, + "step": 21591 + }, + { + "epoch": 3.8441951566951564, + "grad_norm": 0.9354162812232971, + "learning_rate": 7.576792331322136e-07, + "loss": 0.8365, + "step": 21592 + }, + { + "epoch": 3.844373219373219, + "grad_norm": 1.0419033765792847, + "learning_rate": 7.559603849438213e-07, + "loss": 0.6427, + "step": 21593 + }, + { + "epoch": 3.844551282051282, + "grad_norm": 0.9753270149230957, + "learning_rate": 7.542434812362275e-07, + "loss": 0.6996, + "step": 21594 + }, + { + "epoch": 3.8447293447293447, + "grad_norm": 0.917040228843689, + "learning_rate": 7.525285220430723e-07, + "loss": 0.6474, + "step": 21595 + }, + { + "epoch": 3.8449074074074074, + "grad_norm": 0.8489373922348022, + "learning_rate": 7.508155073979395e-07, + "loss": 0.688, + "step": 21596 + }, + { + "epoch": 3.84508547008547, + "grad_norm": 0.8564489483833313, + "learning_rate": 7.491044373344136e-07, + "loss": 0.6195, + "step": 21597 + }, + { + "epoch": 3.845263532763533, + "grad_norm": 1.1631865501403809, + "learning_rate": 7.473953118860011e-07, + "loss": 0.9136, + "step": 21598 + }, + { + "epoch": 3.8454415954415957, + "grad_norm": 1.008110523223877, + "learning_rate": 7.456881310862085e-07, + "loss": 0.6872, + "step": 21599 + }, + { + "epoch": 3.845619658119658, + "grad_norm": 0.9467812180519104, + "learning_rate": 7.439828949684646e-07, + "loss": 0.7575, + "step": 21600 + }, + { + "epoch": 3.8457977207977208, + "grad_norm": 1.0625617504119873, + "learning_rate": 7.422796035661983e-07, + "loss": 0.7556, + "step": 21601 + }, + { + "epoch": 3.8459757834757835, + "grad_norm": 0.940880537033081, + "learning_rate": 7.405782569127606e-07, + "loss": 0.9087, + "step": 21602 + }, + { + "epoch": 3.8461538461538463, + "grad_norm": 0.9074656963348389, + "learning_rate": 7.388788550415138e-07, + "loss": 1.0545, + "step": 21603 + }, + { + "epoch": 3.8463319088319086, + "grad_norm": 0.913948118686676, + "learning_rate": 7.371813979857312e-07, + "loss": 0.7383, + "step": 21604 + }, + { + "epoch": 3.8465099715099713, + "grad_norm": 0.983601987361908, + "learning_rate": 7.354858857786751e-07, + "loss": 0.859, + "step": 21605 + }, + { + "epoch": 3.846688034188034, + "grad_norm": 0.9661694169044495, + "learning_rate": 7.337923184535744e-07, + "loss": 1.0301, + "step": 21606 + }, + { + "epoch": 3.846866096866097, + "grad_norm": 0.9005492329597473, + "learning_rate": 7.321006960436027e-07, + "loss": 0.8122, + "step": 21607 + }, + { + "epoch": 3.8470441595441596, + "grad_norm": 0.8839511871337891, + "learning_rate": 7.304110185819002e-07, + "loss": 0.6888, + "step": 21608 + }, + { + "epoch": 3.8472222222222223, + "grad_norm": 1.069863200187683, + "learning_rate": 7.287232861015958e-07, + "loss": 0.7776, + "step": 21609 + }, + { + "epoch": 3.847400284900285, + "grad_norm": 1.0439397096633911, + "learning_rate": 7.270374986357297e-07, + "loss": 0.871, + "step": 21610 + }, + { + "epoch": 3.847578347578348, + "grad_norm": 1.0932252407073975, + "learning_rate": 7.253536562173424e-07, + "loss": 0.8679, + "step": 21611 + }, + { + "epoch": 3.84775641025641, + "grad_norm": 0.9416168928146362, + "learning_rate": 7.236717588794295e-07, + "loss": 0.9173, + "step": 21612 + }, + { + "epoch": 3.847934472934473, + "grad_norm": 1.028075933456421, + "learning_rate": 7.219918066549313e-07, + "loss": 0.8136, + "step": 21613 + }, + { + "epoch": 3.8481125356125356, + "grad_norm": 0.8770352005958557, + "learning_rate": 7.203137995767772e-07, + "loss": 0.9196, + "step": 21614 + }, + { + "epoch": 3.8482905982905984, + "grad_norm": 0.9017715454101562, + "learning_rate": 7.186377376778297e-07, + "loss": 0.8794, + "step": 21615 + }, + { + "epoch": 3.8484686609686607, + "grad_norm": 0.9379174113273621, + "learning_rate": 7.169636209909514e-07, + "loss": 0.6274, + "step": 21616 + }, + { + "epoch": 3.8486467236467234, + "grad_norm": 0.9132708311080933, + "learning_rate": 7.152914495489161e-07, + "loss": 0.8766, + "step": 21617 + }, + { + "epoch": 3.848824786324786, + "grad_norm": 0.9938055276870728, + "learning_rate": 7.136212233845085e-07, + "loss": 0.702, + "step": 21618 + }, + { + "epoch": 3.849002849002849, + "grad_norm": 0.8011921048164368, + "learning_rate": 7.119529425304361e-07, + "loss": 0.6543, + "step": 21619 + }, + { + "epoch": 3.8491809116809117, + "grad_norm": 1.0564098358154297, + "learning_rate": 7.102866070193947e-07, + "loss": 0.8832, + "step": 21620 + }, + { + "epoch": 3.8493589743589745, + "grad_norm": 0.8896064162254333, + "learning_rate": 7.086222168840362e-07, + "loss": 0.5891, + "step": 21621 + }, + { + "epoch": 3.849537037037037, + "grad_norm": 1.0690109729766846, + "learning_rate": 7.069597721569565e-07, + "loss": 0.9243, + "step": 21622 + }, + { + "epoch": 3.8497150997151, + "grad_norm": 0.9678674936294556, + "learning_rate": 7.052992728707408e-07, + "loss": 0.855, + "step": 21623 + }, + { + "epoch": 3.8498931623931623, + "grad_norm": 0.8748130798339844, + "learning_rate": 7.036407190579187e-07, + "loss": 0.797, + "step": 21624 + }, + { + "epoch": 3.850071225071225, + "grad_norm": 0.9823090434074402, + "learning_rate": 7.019841107509862e-07, + "loss": 0.8641, + "step": 21625 + }, + { + "epoch": 3.8502492877492878, + "grad_norm": 1.0925766229629517, + "learning_rate": 7.003294479824063e-07, + "loss": 0.888, + "step": 21626 + }, + { + "epoch": 3.8504273504273505, + "grad_norm": 1.028260588645935, + "learning_rate": 6.986767307845866e-07, + "loss": 0.8097, + "step": 21627 + }, + { + "epoch": 3.8506054131054133, + "grad_norm": 0.8923847675323486, + "learning_rate": 6.970259591899231e-07, + "loss": 0.7852, + "step": 21628 + }, + { + "epoch": 3.8507834757834756, + "grad_norm": 0.9850567579269409, + "learning_rate": 6.953771332307458e-07, + "loss": 0.768, + "step": 21629 + }, + { + "epoch": 3.8509615384615383, + "grad_norm": 1.0248289108276367, + "learning_rate": 6.937302529393619e-07, + "loss": 0.7098, + "step": 21630 + }, + { + "epoch": 3.851139601139601, + "grad_norm": 0.8439950942993164, + "learning_rate": 6.920853183480569e-07, + "loss": 0.7061, + "step": 21631 + }, + { + "epoch": 3.851317663817664, + "grad_norm": 0.9808463454246521, + "learning_rate": 6.904423294890272e-07, + "loss": 0.8573, + "step": 21632 + }, + { + "epoch": 3.8514957264957266, + "grad_norm": 1.0319223403930664, + "learning_rate": 6.888012863944915e-07, + "loss": 0.8063, + "step": 21633 + }, + { + "epoch": 3.8516737891737893, + "grad_norm": 0.9523464441299438, + "learning_rate": 6.871621890966018e-07, + "loss": 0.7309, + "step": 21634 + }, + { + "epoch": 3.851851851851852, + "grad_norm": 1.1411263942718506, + "learning_rate": 6.855250376274547e-07, + "loss": 0.9086, + "step": 21635 + }, + { + "epoch": 3.8520299145299144, + "grad_norm": 0.990542471408844, + "learning_rate": 6.838898320191356e-07, + "loss": 0.8246, + "step": 21636 + }, + { + "epoch": 3.852207977207977, + "grad_norm": 1.066152572631836, + "learning_rate": 6.822565723036856e-07, + "loss": 0.8174, + "step": 21637 + }, + { + "epoch": 3.85238603988604, + "grad_norm": 1.0600324869155884, + "learning_rate": 6.806252585131124e-07, + "loss": 0.7849, + "step": 21638 + }, + { + "epoch": 3.8525641025641026, + "grad_norm": 0.9351125955581665, + "learning_rate": 6.78995890679357e-07, + "loss": 0.8347, + "step": 21639 + }, + { + "epoch": 3.8527421652421654, + "grad_norm": 0.9323590397834778, + "learning_rate": 6.773684688343606e-07, + "loss": 0.8798, + "step": 21640 + }, + { + "epoch": 3.8529202279202277, + "grad_norm": 1.0057048797607422, + "learning_rate": 6.757429930099979e-07, + "loss": 0.811, + "step": 21641 + }, + { + "epoch": 3.8530982905982905, + "grad_norm": 0.9667607545852661, + "learning_rate": 6.74119463238132e-07, + "loss": 0.8218, + "step": 21642 + }, + { + "epoch": 3.853276353276353, + "grad_norm": 0.9609261751174927, + "learning_rate": 6.724978795505598e-07, + "loss": 0.7635, + "step": 21643 + }, + { + "epoch": 3.853454415954416, + "grad_norm": 0.8612968921661377, + "learning_rate": 6.70878241979056e-07, + "loss": 0.6867, + "step": 21644 + }, + { + "epoch": 3.8536324786324787, + "grad_norm": 0.95118248462677, + "learning_rate": 6.692605505553506e-07, + "loss": 0.7477, + "step": 21645 + }, + { + "epoch": 3.8538105413105415, + "grad_norm": 0.9131327271461487, + "learning_rate": 6.676448053111628e-07, + "loss": 0.7183, + "step": 21646 + }, + { + "epoch": 3.853988603988604, + "grad_norm": 0.8879417181015015, + "learning_rate": 6.660310062781116e-07, + "loss": 0.683, + "step": 21647 + }, + { + "epoch": 3.8541666666666665, + "grad_norm": 0.9245330095291138, + "learning_rate": 6.644191534878385e-07, + "loss": 0.6442, + "step": 21648 + }, + { + "epoch": 3.8543447293447293, + "grad_norm": 1.046465277671814, + "learning_rate": 6.628092469719182e-07, + "loss": 0.7835, + "step": 21649 + }, + { + "epoch": 3.854522792022792, + "grad_norm": 0.865329384803772, + "learning_rate": 6.612012867618922e-07, + "loss": 0.7673, + "step": 21650 + }, + { + "epoch": 3.8547008547008548, + "grad_norm": 0.9498918056488037, + "learning_rate": 6.595952728892796e-07, + "loss": 0.7124, + "step": 21651 + }, + { + "epoch": 3.8548789173789175, + "grad_norm": 0.9849167466163635, + "learning_rate": 6.57991205385533e-07, + "loss": 0.6921, + "step": 21652 + }, + { + "epoch": 3.85505698005698, + "grad_norm": 0.9602441191673279, + "learning_rate": 6.563890842820719e-07, + "loss": 0.812, + "step": 21653 + }, + { + "epoch": 3.8552350427350426, + "grad_norm": 1.0433622598648071, + "learning_rate": 6.547889096103155e-07, + "loss": 0.9083, + "step": 21654 + }, + { + "epoch": 3.8554131054131053, + "grad_norm": 0.9389081001281738, + "learning_rate": 6.531906814015831e-07, + "loss": 0.7912, + "step": 21655 + }, + { + "epoch": 3.855591168091168, + "grad_norm": 0.991176187992096, + "learning_rate": 6.515943996872165e-07, + "loss": 0.7138, + "step": 21656 + }, + { + "epoch": 3.855769230769231, + "grad_norm": 1.1098177433013916, + "learning_rate": 6.500000644984682e-07, + "loss": 0.7762, + "step": 21657 + }, + { + "epoch": 3.8559472934472936, + "grad_norm": 0.8345046639442444, + "learning_rate": 6.484076758665914e-07, + "loss": 0.7031, + "step": 21658 + }, + { + "epoch": 3.8561253561253563, + "grad_norm": 0.9750561118125916, + "learning_rate": 6.46817233822783e-07, + "loss": 0.6956, + "step": 21659 + }, + { + "epoch": 3.8563034188034186, + "grad_norm": 1.0653631687164307, + "learning_rate": 6.452287383981958e-07, + "loss": 0.7884, + "step": 21660 + }, + { + "epoch": 3.8564814814814814, + "grad_norm": 1.1299207210540771, + "learning_rate": 6.436421896239719e-07, + "loss": 1.169, + "step": 21661 + }, + { + "epoch": 3.856659544159544, + "grad_norm": 0.8966163396835327, + "learning_rate": 6.420575875311752e-07, + "loss": 0.7762, + "step": 21662 + }, + { + "epoch": 3.856837606837607, + "grad_norm": 1.015863299369812, + "learning_rate": 6.404749321508697e-07, + "loss": 0.866, + "step": 21663 + }, + { + "epoch": 3.8570156695156697, + "grad_norm": 0.9820823669433594, + "learning_rate": 6.388942235140527e-07, + "loss": 0.8393, + "step": 21664 + }, + { + "epoch": 3.857193732193732, + "grad_norm": 0.8580895066261292, + "learning_rate": 6.373154616516885e-07, + "loss": 0.6423, + "step": 21665 + }, + { + "epoch": 3.8573717948717947, + "grad_norm": 0.9994450211524963, + "learning_rate": 6.357386465947301e-07, + "loss": 0.7924, + "step": 21666 + }, + { + "epoch": 3.8575498575498575, + "grad_norm": 0.9615799784660339, + "learning_rate": 6.341637783740639e-07, + "loss": 0.5859, + "step": 21667 + }, + { + "epoch": 3.85772792022792, + "grad_norm": 0.8876405954360962, + "learning_rate": 6.325908570205429e-07, + "loss": 0.7389, + "step": 21668 + }, + { + "epoch": 3.857905982905983, + "grad_norm": 0.9425645470619202, + "learning_rate": 6.31019882564976e-07, + "loss": 0.8526, + "step": 21669 + }, + { + "epoch": 3.8580840455840457, + "grad_norm": 0.9900808334350586, + "learning_rate": 6.294508550381606e-07, + "loss": 0.79, + "step": 21670 + }, + { + "epoch": 3.8582621082621085, + "grad_norm": 0.924473226070404, + "learning_rate": 6.278837744708388e-07, + "loss": 0.681, + "step": 21671 + }, + { + "epoch": 3.8584401709401708, + "grad_norm": 1.0653358697891235, + "learning_rate": 6.263186408936972e-07, + "loss": 0.8024, + "step": 21672 + }, + { + "epoch": 3.8586182336182335, + "grad_norm": 0.8799527883529663, + "learning_rate": 6.247554543374113e-07, + "loss": 0.7357, + "step": 21673 + }, + { + "epoch": 3.8587962962962963, + "grad_norm": 0.9546528458595276, + "learning_rate": 6.23194214832612e-07, + "loss": 0.7254, + "step": 21674 + }, + { + "epoch": 3.858974358974359, + "grad_norm": 0.8290728330612183, + "learning_rate": 6.21634922409875e-07, + "loss": 0.5482, + "step": 21675 + }, + { + "epoch": 3.859152421652422, + "grad_norm": 0.8252167701721191, + "learning_rate": 6.200775770997758e-07, + "loss": 0.5761, + "step": 21676 + }, + { + "epoch": 3.859330484330484, + "grad_norm": 0.9244386553764343, + "learning_rate": 6.18522178932801e-07, + "loss": 0.8321, + "step": 21677 + }, + { + "epoch": 3.859508547008547, + "grad_norm": 0.9701852202415466, + "learning_rate": 6.169687279394376e-07, + "loss": 0.8276, + "step": 21678 + }, + { + "epoch": 3.8596866096866096, + "grad_norm": 0.885154128074646, + "learning_rate": 6.154172241501277e-07, + "loss": 0.832, + "step": 21679 + }, + { + "epoch": 3.8598646723646723, + "grad_norm": 0.9132698774337769, + "learning_rate": 6.138676675952581e-07, + "loss": 0.721, + "step": 21680 + }, + { + "epoch": 3.860042735042735, + "grad_norm": 0.86441570520401, + "learning_rate": 6.123200583051936e-07, + "loss": 0.6057, + "step": 21681 + }, + { + "epoch": 3.860220797720798, + "grad_norm": 0.9521238803863525, + "learning_rate": 6.107743963102652e-07, + "loss": 0.6874, + "step": 21682 + }, + { + "epoch": 3.8603988603988606, + "grad_norm": 0.9968441724777222, + "learning_rate": 6.092306816407489e-07, + "loss": 0.8812, + "step": 21683 + }, + { + "epoch": 3.8605769230769234, + "grad_norm": 0.7897293567657471, + "learning_rate": 6.076889143268871e-07, + "loss": 0.5925, + "step": 21684 + }, + { + "epoch": 3.8607549857549857, + "grad_norm": 0.9483756422996521, + "learning_rate": 6.06149094398889e-07, + "loss": 0.7941, + "step": 21685 + }, + { + "epoch": 3.8609330484330484, + "grad_norm": 1.136690378189087, + "learning_rate": 6.046112218869305e-07, + "loss": 0.8158, + "step": 21686 + }, + { + "epoch": 3.861111111111111, + "grad_norm": 1.1348493099212646, + "learning_rate": 6.030752968211317e-07, + "loss": 0.8461, + "step": 21687 + }, + { + "epoch": 3.861289173789174, + "grad_norm": 0.9544394612312317, + "learning_rate": 6.015413192316133e-07, + "loss": 0.7925, + "step": 21688 + }, + { + "epoch": 3.861467236467236, + "grad_norm": 1.0044975280761719, + "learning_rate": 6.000092891483844e-07, + "loss": 0.6849, + "step": 21689 + }, + { + "epoch": 3.861645299145299, + "grad_norm": 0.9967083930969238, + "learning_rate": 5.984792066014988e-07, + "loss": 0.9211, + "step": 21690 + }, + { + "epoch": 3.8618233618233617, + "grad_norm": 1.1682944297790527, + "learning_rate": 5.969510716209326e-07, + "loss": 0.9764, + "step": 21691 + }, + { + "epoch": 3.8620014245014245, + "grad_norm": 0.8601269721984863, + "learning_rate": 5.954248842366062e-07, + "loss": 0.6743, + "step": 21692 + }, + { + "epoch": 3.8621794871794872, + "grad_norm": 0.9522037506103516, + "learning_rate": 5.9390064447844e-07, + "loss": 0.6689, + "step": 21693 + }, + { + "epoch": 3.86235754985755, + "grad_norm": 0.9057328104972839, + "learning_rate": 5.923783523762993e-07, + "loss": 0.9379, + "step": 21694 + }, + { + "epoch": 3.8625356125356127, + "grad_norm": 0.9345870614051819, + "learning_rate": 5.908580079599934e-07, + "loss": 0.7687, + "step": 21695 + }, + { + "epoch": 3.8627136752136755, + "grad_norm": 1.0109366178512573, + "learning_rate": 5.893396112593208e-07, + "loss": 0.8095, + "step": 21696 + }, + { + "epoch": 3.862891737891738, + "grad_norm": 0.9415621161460876, + "learning_rate": 5.878231623040242e-07, + "loss": 0.679, + "step": 21697 + }, + { + "epoch": 3.8630698005698005, + "grad_norm": 0.9021549224853516, + "learning_rate": 5.863086611238356e-07, + "loss": 0.8654, + "step": 21698 + }, + { + "epoch": 3.8632478632478633, + "grad_norm": 1.1498757600784302, + "learning_rate": 5.847961077484087e-07, + "loss": 0.7595, + "step": 21699 + }, + { + "epoch": 3.863425925925926, + "grad_norm": 0.9590532183647156, + "learning_rate": 5.832855022073868e-07, + "loss": 0.7458, + "step": 21700 + }, + { + "epoch": 3.8636039886039883, + "grad_norm": 0.9005709290504456, + "learning_rate": 5.817768445303684e-07, + "loss": 0.7446, + "step": 21701 + }, + { + "epoch": 3.863782051282051, + "grad_norm": 0.9828490018844604, + "learning_rate": 5.802701347468965e-07, + "loss": 0.7723, + "step": 21702 + }, + { + "epoch": 3.863960113960114, + "grad_norm": 0.9257676601409912, + "learning_rate": 5.787653728865139e-07, + "loss": 1.0391, + "step": 21703 + }, + { + "epoch": 3.8641381766381766, + "grad_norm": 0.9859639406204224, + "learning_rate": 5.772625589786973e-07, + "loss": 0.7769, + "step": 21704 + }, + { + "epoch": 3.8643162393162394, + "grad_norm": 0.9691302180290222, + "learning_rate": 5.757616930528786e-07, + "loss": 0.6341, + "step": 21705 + }, + { + "epoch": 3.864494301994302, + "grad_norm": 0.9758834838867188, + "learning_rate": 5.742627751384788e-07, + "loss": 0.7149, + "step": 21706 + }, + { + "epoch": 3.864672364672365, + "grad_norm": 1.028484582901001, + "learning_rate": 5.727658052648633e-07, + "loss": 0.8684, + "step": 21707 + }, + { + "epoch": 3.8648504273504276, + "grad_norm": 0.92888343334198, + "learning_rate": 5.712707834613528e-07, + "loss": 0.6475, + "step": 21708 + }, + { + "epoch": 3.86502849002849, + "grad_norm": 0.9963685274124146, + "learning_rate": 5.697777097572577e-07, + "loss": 0.6252, + "step": 21709 + }, + { + "epoch": 3.8652065527065527, + "grad_norm": 0.8675285577774048, + "learning_rate": 5.682865841818097e-07, + "loss": 0.8577, + "step": 21710 + }, + { + "epoch": 3.8653846153846154, + "grad_norm": 1.0220355987548828, + "learning_rate": 5.667974067642412e-07, + "loss": 0.8689, + "step": 21711 + }, + { + "epoch": 3.865562678062678, + "grad_norm": 0.9978788495063782, + "learning_rate": 5.653101775337067e-07, + "loss": 0.6846, + "step": 21712 + }, + { + "epoch": 3.8657407407407405, + "grad_norm": 1.0353487730026245, + "learning_rate": 5.638248965193826e-07, + "loss": 0.9282, + "step": 21713 + }, + { + "epoch": 3.8659188034188032, + "grad_norm": 0.8543832898139954, + "learning_rate": 5.623415637503348e-07, + "loss": 0.6257, + "step": 21714 + }, + { + "epoch": 3.866096866096866, + "grad_norm": 1.0639516115188599, + "learning_rate": 5.608601792556511e-07, + "loss": 0.6694, + "step": 21715 + }, + { + "epoch": 3.8662749287749287, + "grad_norm": 0.8737302422523499, + "learning_rate": 5.593807430643416e-07, + "loss": 0.7041, + "step": 21716 + }, + { + "epoch": 3.8664529914529915, + "grad_norm": 1.1135413646697998, + "learning_rate": 5.579032552053942e-07, + "loss": 0.762, + "step": 21717 + }, + { + "epoch": 3.8666310541310542, + "grad_norm": 0.9849023222923279, + "learning_rate": 5.564277157077524e-07, + "loss": 0.7913, + "step": 21718 + }, + { + "epoch": 3.866809116809117, + "grad_norm": 1.013596534729004, + "learning_rate": 5.549541246003376e-07, + "loss": 0.7516, + "step": 21719 + }, + { + "epoch": 3.8669871794871797, + "grad_norm": 0.8574326038360596, + "learning_rate": 5.534824819120266e-07, + "loss": 0.6867, + "step": 21720 + }, + { + "epoch": 3.867165242165242, + "grad_norm": 0.9084491729736328, + "learning_rate": 5.520127876716408e-07, + "loss": 0.7386, + "step": 21721 + }, + { + "epoch": 3.867343304843305, + "grad_norm": 0.9534892439842224, + "learning_rate": 5.505450419079794e-07, + "loss": 0.7824, + "step": 21722 + }, + { + "epoch": 3.8675213675213675, + "grad_norm": 1.0126458406448364, + "learning_rate": 5.490792446497973e-07, + "loss": 0.6519, + "step": 21723 + }, + { + "epoch": 3.8676994301994303, + "grad_norm": 0.9215713739395142, + "learning_rate": 5.476153959258157e-07, + "loss": 0.7617, + "step": 21724 + }, + { + "epoch": 3.8678774928774926, + "grad_norm": 0.7952741384506226, + "learning_rate": 5.461534957647119e-07, + "loss": 0.6392, + "step": 21725 + }, + { + "epoch": 3.8680555555555554, + "grad_norm": 1.035820722579956, + "learning_rate": 5.446935441951406e-07, + "loss": 1.011, + "step": 21726 + }, + { + "epoch": 3.868233618233618, + "grad_norm": 0.9325335025787354, + "learning_rate": 5.432355412456902e-07, + "loss": 0.6771, + "step": 21727 + }, + { + "epoch": 3.868411680911681, + "grad_norm": 1.0619163513183594, + "learning_rate": 5.417794869449377e-07, + "loss": 0.7757, + "step": 21728 + }, + { + "epoch": 3.8685897435897436, + "grad_norm": 1.1454271078109741, + "learning_rate": 5.40325381321416e-07, + "loss": 0.9636, + "step": 21729 + }, + { + "epoch": 3.8687678062678064, + "grad_norm": 1.032842993736267, + "learning_rate": 5.388732244036021e-07, + "loss": 0.6985, + "step": 21730 + }, + { + "epoch": 3.868945868945869, + "grad_norm": 1.2101705074310303, + "learning_rate": 5.374230162199623e-07, + "loss": 0.7568, + "step": 21731 + }, + { + "epoch": 3.869123931623932, + "grad_norm": 1.015390396118164, + "learning_rate": 5.35974756798896e-07, + "loss": 0.6603, + "step": 21732 + }, + { + "epoch": 3.869301994301994, + "grad_norm": 0.837568998336792, + "learning_rate": 5.345284461687916e-07, + "loss": 0.7916, + "step": 21733 + }, + { + "epoch": 3.869480056980057, + "grad_norm": 0.9118668437004089, + "learning_rate": 5.330840843579709e-07, + "loss": 0.7041, + "step": 21734 + }, + { + "epoch": 3.8696581196581197, + "grad_norm": 1.0995440483093262, + "learning_rate": 5.316416713947559e-07, + "loss": 0.8364, + "step": 21735 + }, + { + "epoch": 3.8698361823361824, + "grad_norm": 0.906592845916748, + "learning_rate": 5.302012073073903e-07, + "loss": 0.7558, + "step": 21736 + }, + { + "epoch": 3.870014245014245, + "grad_norm": 0.8995518684387207, + "learning_rate": 5.287626921240963e-07, + "loss": 0.7778, + "step": 21737 + }, + { + "epoch": 3.8701923076923075, + "grad_norm": 0.838267982006073, + "learning_rate": 5.273261258730733e-07, + "loss": 0.6692, + "step": 21738 + }, + { + "epoch": 3.8703703703703702, + "grad_norm": 0.9452478885650635, + "learning_rate": 5.258915085824434e-07, + "loss": 0.7617, + "step": 21739 + }, + { + "epoch": 3.870548433048433, + "grad_norm": 1.012858510017395, + "learning_rate": 5.244588402803396e-07, + "loss": 0.8608, + "step": 21740 + }, + { + "epoch": 3.8707264957264957, + "grad_norm": 1.0003842115402222, + "learning_rate": 5.230281209948174e-07, + "loss": 0.7907, + "step": 21741 + }, + { + "epoch": 3.8709045584045585, + "grad_norm": 1.0578328371047974, + "learning_rate": 5.215993507539207e-07, + "loss": 0.8449, + "step": 21742 + }, + { + "epoch": 3.8710826210826212, + "grad_norm": 0.9512600302696228, + "learning_rate": 5.201725295856386e-07, + "loss": 0.8504, + "step": 21743 + }, + { + "epoch": 3.871260683760684, + "grad_norm": 0.952800989151001, + "learning_rate": 5.187476575179151e-07, + "loss": 0.832, + "step": 21744 + }, + { + "epoch": 3.8714387464387463, + "grad_norm": 0.9954485893249512, + "learning_rate": 5.173247345786835e-07, + "loss": 0.8799, + "step": 21745 + }, + { + "epoch": 3.871616809116809, + "grad_norm": 0.9191272854804993, + "learning_rate": 5.159037607958106e-07, + "loss": 0.8221, + "step": 21746 + }, + { + "epoch": 3.871794871794872, + "grad_norm": 0.9987154006958008, + "learning_rate": 5.144847361971406e-07, + "loss": 0.7792, + "step": 21747 + }, + { + "epoch": 3.8719729344729346, + "grad_norm": 1.0156099796295166, + "learning_rate": 5.130676608104845e-07, + "loss": 0.9094, + "step": 21748 + }, + { + "epoch": 3.8721509971509973, + "grad_norm": 1.0321117639541626, + "learning_rate": 5.116525346635981e-07, + "loss": 0.8166, + "step": 21749 + }, + { + "epoch": 3.8723290598290596, + "grad_norm": 0.9277944564819336, + "learning_rate": 5.102393577842146e-07, + "loss": 0.9311, + "step": 21750 + }, + { + "epoch": 3.8725071225071224, + "grad_norm": 1.0144904851913452, + "learning_rate": 5.088281302000231e-07, + "loss": 0.6603, + "step": 21751 + }, + { + "epoch": 3.872685185185185, + "grad_norm": 1.0451470613479614, + "learning_rate": 5.074188519386569e-07, + "loss": 0.7662, + "step": 21752 + }, + { + "epoch": 3.872863247863248, + "grad_norm": 1.0115594863891602, + "learning_rate": 5.060115230277606e-07, + "loss": 0.6707, + "step": 21753 + }, + { + "epoch": 3.8730413105413106, + "grad_norm": 1.1018215417861938, + "learning_rate": 5.046061434948679e-07, + "loss": 0.8675, + "step": 21754 + }, + { + "epoch": 3.8732193732193734, + "grad_norm": 0.9691307544708252, + "learning_rate": 5.032027133675454e-07, + "loss": 0.8355, + "step": 21755 + }, + { + "epoch": 3.873397435897436, + "grad_norm": 0.9273481369018555, + "learning_rate": 5.018012326732713e-07, + "loss": 0.8075, + "step": 21756 + }, + { + "epoch": 3.8735754985754984, + "grad_norm": 0.9765976071357727, + "learning_rate": 5.004017014395124e-07, + "loss": 0.7604, + "step": 21757 + }, + { + "epoch": 3.873753561253561, + "grad_norm": 1.0961931943893433, + "learning_rate": 4.990041196937023e-07, + "loss": 0.8029, + "step": 21758 + }, + { + "epoch": 3.873931623931624, + "grad_norm": 0.9625610113143921, + "learning_rate": 4.97608487463197e-07, + "loss": 0.8036, + "step": 21759 + }, + { + "epoch": 3.8741096866096867, + "grad_norm": 0.9747796058654785, + "learning_rate": 4.962148047753634e-07, + "loss": 0.7748, + "step": 21760 + }, + { + "epoch": 3.8742877492877494, + "grad_norm": 0.8600348234176636, + "learning_rate": 4.948230716574909e-07, + "loss": 0.6761, + "step": 21761 + }, + { + "epoch": 3.8744658119658117, + "grad_norm": 0.8769771456718445, + "learning_rate": 4.934332881368575e-07, + "loss": 0.6293, + "step": 21762 + }, + { + "epoch": 3.8746438746438745, + "grad_norm": 0.9831801652908325, + "learning_rate": 4.920454542406972e-07, + "loss": 0.9356, + "step": 21763 + }, + { + "epoch": 3.8748219373219372, + "grad_norm": 0.9225176572799683, + "learning_rate": 4.906595699961992e-07, + "loss": 0.8194, + "step": 21764 + }, + { + "epoch": 3.875, + "grad_norm": 0.8482645153999329, + "learning_rate": 4.892756354305084e-07, + "loss": 0.6919, + "step": 21765 + }, + { + "epoch": 3.8751780626780628, + "grad_norm": 1.2182742357254028, + "learning_rate": 4.878936505707477e-07, + "loss": 0.7674, + "step": 21766 + }, + { + "epoch": 3.8753561253561255, + "grad_norm": 0.9315845370292664, + "learning_rate": 4.865136154439954e-07, + "loss": 0.7609, + "step": 21767 + }, + { + "epoch": 3.8755341880341883, + "grad_norm": 0.9341706037521362, + "learning_rate": 4.851355300772852e-07, + "loss": 0.5422, + "step": 21768 + }, + { + "epoch": 3.8757122507122506, + "grad_norm": 0.9102457165718079, + "learning_rate": 4.837593944976182e-07, + "loss": 0.8436, + "step": 21769 + }, + { + "epoch": 3.8758903133903133, + "grad_norm": 1.2303838729858398, + "learning_rate": 4.823852087319614e-07, + "loss": 0.8097, + "step": 21770 + }, + { + "epoch": 3.876068376068376, + "grad_norm": 0.8186293244361877, + "learning_rate": 4.810129728072266e-07, + "loss": 0.791, + "step": 21771 + }, + { + "epoch": 3.876246438746439, + "grad_norm": 1.1224546432495117, + "learning_rate": 4.796426867503257e-07, + "loss": 0.9078, + "step": 21772 + }, + { + "epoch": 3.8764245014245016, + "grad_norm": 0.8153115510940552, + "learning_rate": 4.782743505880816e-07, + "loss": 0.568, + "step": 21773 + }, + { + "epoch": 3.876602564102564, + "grad_norm": 1.0225319862365723, + "learning_rate": 4.769079643473173e-07, + "loss": 0.8927, + "step": 21774 + }, + { + "epoch": 3.8767806267806266, + "grad_norm": 0.8474786877632141, + "learning_rate": 4.755435280547893e-07, + "loss": 0.6705, + "step": 21775 + }, + { + "epoch": 3.8769586894586894, + "grad_norm": 0.9945504665374756, + "learning_rate": 4.7418104173725387e-07, + "loss": 0.7112, + "step": 21776 + }, + { + "epoch": 3.877136752136752, + "grad_norm": 0.9561564922332764, + "learning_rate": 4.728205054213897e-07, + "loss": 0.9308, + "step": 21777 + }, + { + "epoch": 3.877314814814815, + "grad_norm": 0.9943435788154602, + "learning_rate": 4.714619191338643e-07, + "loss": 0.7911, + "step": 21778 + }, + { + "epoch": 3.8774928774928776, + "grad_norm": 0.8424780964851379, + "learning_rate": 4.7010528290127863e-07, + "loss": 0.616, + "step": 21779 + }, + { + "epoch": 3.8776709401709404, + "grad_norm": 0.9379110336303711, + "learning_rate": 4.687505967502226e-07, + "loss": 0.7668, + "step": 21780 + }, + { + "epoch": 3.8778490028490027, + "grad_norm": 1.0603646039962769, + "learning_rate": 4.6739786070725264e-07, + "loss": 0.8882, + "step": 21781 + }, + { + "epoch": 3.8780270655270654, + "grad_norm": 0.8258822560310364, + "learning_rate": 4.660470747988588e-07, + "loss": 0.646, + "step": 21782 + }, + { + "epoch": 3.878205128205128, + "grad_norm": 0.9126657843589783, + "learning_rate": 4.646982390514976e-07, + "loss": 0.64, + "step": 21783 + }, + { + "epoch": 3.878383190883191, + "grad_norm": 1.0353071689605713, + "learning_rate": 4.633513534916145e-07, + "loss": 0.7882, + "step": 21784 + }, + { + "epoch": 3.8785612535612537, + "grad_norm": 0.9383770227432251, + "learning_rate": 4.6200641814559964e-07, + "loss": 0.7981, + "step": 21785 + }, + { + "epoch": 3.878739316239316, + "grad_norm": 0.9015040397644043, + "learning_rate": 4.606634330397874e-07, + "loss": 0.8177, + "step": 21786 + }, + { + "epoch": 3.8789173789173788, + "grad_norm": 0.9347863793373108, + "learning_rate": 4.5932239820050125e-07, + "loss": 0.8157, + "step": 21787 + }, + { + "epoch": 3.8790954415954415, + "grad_norm": 1.0109801292419434, + "learning_rate": 4.5798331365402016e-07, + "loss": 0.7419, + "step": 21788 + }, + { + "epoch": 3.8792735042735043, + "grad_norm": 0.9674223065376282, + "learning_rate": 4.566461794265675e-07, + "loss": 0.7438, + "step": 21789 + }, + { + "epoch": 3.879451566951567, + "grad_norm": 0.8686508536338806, + "learning_rate": 4.5531099554435576e-07, + "loss": 0.6051, + "step": 21790 + }, + { + "epoch": 3.8796296296296298, + "grad_norm": 0.7613703608512878, + "learning_rate": 4.539777620335417e-07, + "loss": 0.5552, + "step": 21791 + }, + { + "epoch": 3.8798076923076925, + "grad_norm": 1.0486373901367188, + "learning_rate": 4.526464789202378e-07, + "loss": 0.9524, + "step": 21792 + }, + { + "epoch": 3.879985754985755, + "grad_norm": 0.8991581797599792, + "learning_rate": 4.5131714623053433e-07, + "loss": 0.8313, + "step": 21793 + }, + { + "epoch": 3.8801638176638176, + "grad_norm": 0.9314252138137817, + "learning_rate": 4.499897639904771e-07, + "loss": 0.6796, + "step": 21794 + }, + { + "epoch": 3.8803418803418803, + "grad_norm": 1.0096250772476196, + "learning_rate": 4.4866433222607864e-07, + "loss": 0.8151, + "step": 21795 + }, + { + "epoch": 3.880519943019943, + "grad_norm": 0.9510086178779602, + "learning_rate": 4.4734085096329594e-07, + "loss": 0.7451, + "step": 21796 + }, + { + "epoch": 3.880698005698006, + "grad_norm": 0.9142504334449768, + "learning_rate": 4.460193202280638e-07, + "loss": 0.8498, + "step": 21797 + }, + { + "epoch": 3.880876068376068, + "grad_norm": 0.9794802665710449, + "learning_rate": 4.446997400462838e-07, + "loss": 0.7224, + "step": 21798 + }, + { + "epoch": 3.881054131054131, + "grad_norm": 1.0646485090255737, + "learning_rate": 4.433821104438018e-07, + "loss": 0.633, + "step": 21799 + }, + { + "epoch": 3.8812321937321936, + "grad_norm": 0.9331605434417725, + "learning_rate": 4.420664314464418e-07, + "loss": 0.7344, + "step": 21800 + }, + { + "epoch": 3.8814102564102564, + "grad_norm": 0.9503687024116516, + "learning_rate": 4.4075270307997186e-07, + "loss": 0.611, + "step": 21801 + }, + { + "epoch": 3.881588319088319, + "grad_norm": 0.8861737847328186, + "learning_rate": 4.3944092537013814e-07, + "loss": 0.5885, + "step": 21802 + }, + { + "epoch": 3.881766381766382, + "grad_norm": 0.910184919834137, + "learning_rate": 4.3813109834264233e-07, + "loss": 0.5804, + "step": 21803 + }, + { + "epoch": 3.8819444444444446, + "grad_norm": 1.016818881034851, + "learning_rate": 4.3682322202314163e-07, + "loss": 0.8663, + "step": 21804 + }, + { + "epoch": 3.8821225071225074, + "grad_norm": 0.7735291123390198, + "learning_rate": 4.3551729643727113e-07, + "loss": 0.4588, + "step": 21805 + }, + { + "epoch": 3.8823005698005697, + "grad_norm": 0.9913780689239502, + "learning_rate": 4.3421332161059926e-07, + "loss": 0.799, + "step": 21806 + }, + { + "epoch": 3.8824786324786325, + "grad_norm": 1.0465813875198364, + "learning_rate": 4.329112975686944e-07, + "loss": 0.8979, + "step": 21807 + }, + { + "epoch": 3.882656695156695, + "grad_norm": 0.9852820038795471, + "learning_rate": 4.316112243370696e-07, + "loss": 0.6351, + "step": 21808 + }, + { + "epoch": 3.882834757834758, + "grad_norm": 0.9551993608474731, + "learning_rate": 4.3031310194117104e-07, + "loss": 0.7989, + "step": 21809 + }, + { + "epoch": 3.8830128205128203, + "grad_norm": 1.037743330001831, + "learning_rate": 4.290169304064673e-07, + "loss": 0.8558, + "step": 21810 + }, + { + "epoch": 3.883190883190883, + "grad_norm": 0.967463493347168, + "learning_rate": 4.2772270975831583e-07, + "loss": 0.7983, + "step": 21811 + }, + { + "epoch": 3.8833689458689458, + "grad_norm": 1.2085150480270386, + "learning_rate": 4.264304400221075e-07, + "loss": 0.5849, + "step": 21812 + }, + { + "epoch": 3.8835470085470085, + "grad_norm": 0.9609394669532776, + "learning_rate": 4.251401212231443e-07, + "loss": 0.8195, + "step": 21813 + }, + { + "epoch": 3.8837250712250713, + "grad_norm": 0.9492299556732178, + "learning_rate": 4.23851753386717e-07, + "loss": 1.0462, + "step": 21814 + }, + { + "epoch": 3.883903133903134, + "grad_norm": 1.1199448108673096, + "learning_rate": 4.2256533653804997e-07, + "loss": 0.6554, + "step": 21815 + }, + { + "epoch": 3.8840811965811968, + "grad_norm": 1.1129202842712402, + "learning_rate": 4.212808707023785e-07, + "loss": 0.7439, + "step": 21816 + }, + { + "epoch": 3.8842592592592595, + "grad_norm": 1.0576272010803223, + "learning_rate": 4.1999835590483815e-07, + "loss": 0.693, + "step": 21817 + }, + { + "epoch": 3.884437321937322, + "grad_norm": 0.9170486927032471, + "learning_rate": 4.187177921705754e-07, + "loss": 0.7298, + "step": 21818 + }, + { + "epoch": 3.8846153846153846, + "grad_norm": 0.8820095062255859, + "learning_rate": 4.1743917952467015e-07, + "loss": 0.6307, + "step": 21819 + }, + { + "epoch": 3.8847934472934473, + "grad_norm": 1.213741660118103, + "learning_rate": 4.1616251799219133e-07, + "loss": 1.0166, + "step": 21820 + }, + { + "epoch": 3.88497150997151, + "grad_norm": 1.1306160688400269, + "learning_rate": 4.1488780759812995e-07, + "loss": 0.7279, + "step": 21821 + }, + { + "epoch": 3.8851495726495724, + "grad_norm": 0.9849101901054382, + "learning_rate": 4.136150483674772e-07, + "loss": 0.8513, + "step": 21822 + }, + { + "epoch": 3.885327635327635, + "grad_norm": 0.9408984780311584, + "learning_rate": 4.123442403251576e-07, + "loss": 0.9024, + "step": 21823 + }, + { + "epoch": 3.885505698005698, + "grad_norm": 0.9596730470657349, + "learning_rate": 4.110753834960845e-07, + "loss": 0.7024, + "step": 21824 + }, + { + "epoch": 3.8856837606837606, + "grad_norm": 1.0172388553619385, + "learning_rate": 4.098084779051048e-07, + "loss": 0.9005, + "step": 21825 + }, + { + "epoch": 3.8858618233618234, + "grad_norm": 0.7936252355575562, + "learning_rate": 4.0854352357705406e-07, + "loss": 0.5977, + "step": 21826 + }, + { + "epoch": 3.886039886039886, + "grad_norm": 0.9177237153053284, + "learning_rate": 4.072805205367125e-07, + "loss": 0.5956, + "step": 21827 + }, + { + "epoch": 3.886217948717949, + "grad_norm": 1.0219414234161377, + "learning_rate": 4.060194688088048e-07, + "loss": 0.8677, + "step": 21828 + }, + { + "epoch": 3.8863960113960117, + "grad_norm": 0.8081260919570923, + "learning_rate": 4.047603684180778e-07, + "loss": 0.5772, + "step": 21829 + }, + { + "epoch": 3.886574074074074, + "grad_norm": 1.1134142875671387, + "learning_rate": 4.0350321938916745e-07, + "loss": 0.8819, + "step": 21830 + }, + { + "epoch": 3.8867521367521367, + "grad_norm": 1.1589722633361816, + "learning_rate": 4.022480217467206e-07, + "loss": 0.724, + "step": 21831 + }, + { + "epoch": 3.8869301994301995, + "grad_norm": 1.0400350093841553, + "learning_rate": 4.009947755153398e-07, + "loss": 0.6483, + "step": 21832 + }, + { + "epoch": 3.887108262108262, + "grad_norm": 1.082972526550293, + "learning_rate": 3.997434807195499e-07, + "loss": 0.7819, + "step": 21833 + }, + { + "epoch": 3.8872863247863245, + "grad_norm": 0.9113301038742065, + "learning_rate": 3.9849413738388686e-07, + "loss": 0.7444, + "step": 21834 + }, + { + "epoch": 3.8874643874643873, + "grad_norm": 0.9959047436714172, + "learning_rate": 3.9724674553284215e-07, + "loss": 1.0285, + "step": 21835 + }, + { + "epoch": 3.88764245014245, + "grad_norm": 0.8362236022949219, + "learning_rate": 3.9600130519082956e-07, + "loss": 0.5392, + "step": 21836 + }, + { + "epoch": 3.8878205128205128, + "grad_norm": 1.2744941711425781, + "learning_rate": 3.9475781638226294e-07, + "loss": 0.9262, + "step": 21837 + }, + { + "epoch": 3.8879985754985755, + "grad_norm": 0.9878840446472168, + "learning_rate": 3.935162791315006e-07, + "loss": 0.8396, + "step": 21838 + }, + { + "epoch": 3.8881766381766383, + "grad_norm": 1.108947515487671, + "learning_rate": 3.9227669346286744e-07, + "loss": 0.6708, + "step": 21839 + }, + { + "epoch": 3.888354700854701, + "grad_norm": 1.107846736907959, + "learning_rate": 3.910390594006774e-07, + "loss": 0.8392, + "step": 21840 + }, + { + "epoch": 3.888532763532764, + "grad_norm": 1.0903797149658203, + "learning_rate": 3.898033769691334e-07, + "loss": 0.8854, + "step": 21841 + }, + { + "epoch": 3.888710826210826, + "grad_norm": 0.9175450801849365, + "learning_rate": 3.885696461924937e-07, + "loss": 0.6413, + "step": 21842 + }, + { + "epoch": 3.888888888888889, + "grad_norm": 0.9295111298561096, + "learning_rate": 3.8733786709488354e-07, + "loss": 0.7888, + "step": 21843 + }, + { + "epoch": 3.8890669515669516, + "grad_norm": 0.9339991807937622, + "learning_rate": 3.8610803970047236e-07, + "loss": 0.6805, + "step": 21844 + }, + { + "epoch": 3.8892450142450143, + "grad_norm": 1.1252503395080566, + "learning_rate": 3.84880164033341e-07, + "loss": 0.796, + "step": 21845 + }, + { + "epoch": 3.8894230769230766, + "grad_norm": 0.9593950510025024, + "learning_rate": 3.836542401175591e-07, + "loss": 0.7758, + "step": 21846 + }, + { + "epoch": 3.8896011396011394, + "grad_norm": 0.9289141297340393, + "learning_rate": 3.8243026797712967e-07, + "loss": 0.9374, + "step": 21847 + }, + { + "epoch": 3.889779202279202, + "grad_norm": 1.0639877319335938, + "learning_rate": 3.8120824763604456e-07, + "loss": 0.9137, + "step": 21848 + }, + { + "epoch": 3.889957264957265, + "grad_norm": 0.9373790621757507, + "learning_rate": 3.7998817911824026e-07, + "loss": 0.7312, + "step": 21849 + }, + { + "epoch": 3.8901353276353277, + "grad_norm": 0.9877315759658813, + "learning_rate": 3.787700624476198e-07, + "loss": 0.7707, + "step": 21850 + }, + { + "epoch": 3.8903133903133904, + "grad_norm": 0.9654448628425598, + "learning_rate": 3.7755389764806416e-07, + "loss": 0.7137, + "step": 21851 + }, + { + "epoch": 3.890491452991453, + "grad_norm": 0.930350661277771, + "learning_rate": 3.763396847433875e-07, + "loss": 0.7167, + "step": 21852 + }, + { + "epoch": 3.890669515669516, + "grad_norm": 0.8244062662124634, + "learning_rate": 3.7512742375739316e-07, + "loss": 0.7171, + "step": 21853 + }, + { + "epoch": 3.890847578347578, + "grad_norm": 0.9072375297546387, + "learning_rate": 3.739171147138176e-07, + "loss": 0.8647, + "step": 21854 + }, + { + "epoch": 3.891025641025641, + "grad_norm": 0.9860767722129822, + "learning_rate": 3.7270875763637527e-07, + "loss": 0.7323, + "step": 21855 + }, + { + "epoch": 3.8912037037037037, + "grad_norm": 0.9968366622924805, + "learning_rate": 3.715023525487582e-07, + "loss": 0.8133, + "step": 21856 + }, + { + "epoch": 3.8913817663817665, + "grad_norm": 1.3136839866638184, + "learning_rate": 3.7029789947458094e-07, + "loss": 0.9965, + "step": 21857 + }, + { + "epoch": 3.8915598290598292, + "grad_norm": 1.0396318435668945, + "learning_rate": 3.6909539843745787e-07, + "loss": 0.7561, + "step": 21858 + }, + { + "epoch": 3.8917378917378915, + "grad_norm": 0.9965425133705139, + "learning_rate": 3.67894849460948e-07, + "loss": 0.8734, + "step": 21859 + }, + { + "epoch": 3.8919159544159543, + "grad_norm": 1.020198106765747, + "learning_rate": 3.6669625256856576e-07, + "loss": 0.6508, + "step": 21860 + }, + { + "epoch": 3.892094017094017, + "grad_norm": 0.9760180711746216, + "learning_rate": 3.6549960778380357e-07, + "loss": 0.9447, + "step": 21861 + }, + { + "epoch": 3.89227207977208, + "grad_norm": 1.0385621786117554, + "learning_rate": 3.643049151301092e-07, + "loss": 0.7007, + "step": 21862 + }, + { + "epoch": 3.8924501424501425, + "grad_norm": 0.8879075050354004, + "learning_rate": 3.631121746308752e-07, + "loss": 0.6414, + "step": 21863 + }, + { + "epoch": 3.8926282051282053, + "grad_norm": 0.865680992603302, + "learning_rate": 3.619213863094828e-07, + "loss": 0.6515, + "step": 21864 + }, + { + "epoch": 3.892806267806268, + "grad_norm": 1.055982232093811, + "learning_rate": 3.607325501892689e-07, + "loss": 0.7958, + "step": 21865 + }, + { + "epoch": 3.8929843304843303, + "grad_norm": 1.0314075946807861, + "learning_rate": 3.595456662935037e-07, + "loss": 0.8503, + "step": 21866 + }, + { + "epoch": 3.893162393162393, + "grad_norm": 0.9081546664237976, + "learning_rate": 3.583607346454687e-07, + "loss": 0.8502, + "step": 21867 + }, + { + "epoch": 3.893340455840456, + "grad_norm": 0.989351749420166, + "learning_rate": 3.571777552683564e-07, + "loss": 0.7857, + "step": 21868 + }, + { + "epoch": 3.8935185185185186, + "grad_norm": 1.021647572517395, + "learning_rate": 3.5599672818537046e-07, + "loss": 0.7159, + "step": 21869 + }, + { + "epoch": 3.8936965811965814, + "grad_norm": 0.9180936813354492, + "learning_rate": 3.548176534196257e-07, + "loss": 0.7763, + "step": 21870 + }, + { + "epoch": 3.8938746438746437, + "grad_norm": 0.9282814860343933, + "learning_rate": 3.53640530994237e-07, + "loss": 0.7063, + "step": 21871 + }, + { + "epoch": 3.8940527065527064, + "grad_norm": 1.0444775819778442, + "learning_rate": 3.5246536093226366e-07, + "loss": 0.8247, + "step": 21872 + }, + { + "epoch": 3.894230769230769, + "grad_norm": 0.6987374424934387, + "learning_rate": 3.512921432567318e-07, + "loss": 0.4846, + "step": 21873 + }, + { + "epoch": 3.894408831908832, + "grad_norm": 0.9549040794372559, + "learning_rate": 3.501208779906229e-07, + "loss": 0.7085, + "step": 21874 + }, + { + "epoch": 3.8945868945868947, + "grad_norm": 0.9242094159126282, + "learning_rate": 3.4895156515690753e-07, + "loss": 0.7811, + "step": 21875 + }, + { + "epoch": 3.8947649572649574, + "grad_norm": 0.9738181829452515, + "learning_rate": 3.477842047784563e-07, + "loss": 0.8004, + "step": 21876 + }, + { + "epoch": 3.89494301994302, + "grad_norm": 0.9645888805389404, + "learning_rate": 3.466187968781842e-07, + "loss": 0.656, + "step": 21877 + }, + { + "epoch": 3.8951210826210825, + "grad_norm": 0.9509434103965759, + "learning_rate": 3.4545534147889523e-07, + "loss": 0.7381, + "step": 21878 + }, + { + "epoch": 3.8952991452991452, + "grad_norm": 1.1472703218460083, + "learning_rate": 3.4429383860339336e-07, + "loss": 0.8948, + "step": 21879 + }, + { + "epoch": 3.895477207977208, + "grad_norm": 0.9420302510261536, + "learning_rate": 3.4313428827443817e-07, + "loss": 0.8261, + "step": 21880 + }, + { + "epoch": 3.8956552706552707, + "grad_norm": 0.8902772665023804, + "learning_rate": 3.419766905147448e-07, + "loss": 0.9552, + "step": 21881 + }, + { + "epoch": 3.8958333333333335, + "grad_norm": 0.8738743662834167, + "learning_rate": 3.408210453470062e-07, + "loss": 0.6172, + "step": 21882 + }, + { + "epoch": 3.896011396011396, + "grad_norm": 0.8608858585357666, + "learning_rate": 3.3966735279384875e-07, + "loss": 0.7194, + "step": 21883 + }, + { + "epoch": 3.8961894586894585, + "grad_norm": 0.9503448009490967, + "learning_rate": 3.385156128778766e-07, + "loss": 0.7791, + "step": 21884 + }, + { + "epoch": 3.8963675213675213, + "grad_norm": 0.9241414070129395, + "learning_rate": 3.3736582562167163e-07, + "loss": 0.634, + "step": 21885 + }, + { + "epoch": 3.896545584045584, + "grad_norm": 0.9042157530784607, + "learning_rate": 3.362179910477492e-07, + "loss": 0.8056, + "step": 21886 + }, + { + "epoch": 3.896723646723647, + "grad_norm": 0.9168319702148438, + "learning_rate": 3.350721091786024e-07, + "loss": 0.507, + "step": 21887 + }, + { + "epoch": 3.8969017094017095, + "grad_norm": 0.7683261632919312, + "learning_rate": 3.3392818003668e-07, + "loss": 0.6233, + "step": 21888 + }, + { + "epoch": 3.8970797720797723, + "grad_norm": 0.8454415798187256, + "learning_rate": 3.3278620364440847e-07, + "loss": 0.6458, + "step": 21889 + }, + { + "epoch": 3.8972578347578346, + "grad_norm": 0.878555178642273, + "learning_rate": 3.316461800241366e-07, + "loss": 0.6788, + "step": 21890 + }, + { + "epoch": 3.8974358974358974, + "grad_norm": 0.9094281792640686, + "learning_rate": 3.3050810919821316e-07, + "loss": 0.708, + "step": 21891 + }, + { + "epoch": 3.89761396011396, + "grad_norm": 0.9100331664085388, + "learning_rate": 3.2937199118894257e-07, + "loss": 0.6912, + "step": 21892 + }, + { + "epoch": 3.897792022792023, + "grad_norm": 0.9003938436508179, + "learning_rate": 3.282378260185848e-07, + "loss": 0.8759, + "step": 21893 + }, + { + "epoch": 3.8979700854700856, + "grad_norm": 0.9322202205657959, + "learning_rate": 3.2710561370934424e-07, + "loss": 0.7589, + "step": 21894 + }, + { + "epoch": 3.898148148148148, + "grad_norm": 0.9860547184944153, + "learning_rate": 3.259753542834254e-07, + "loss": 0.742, + "step": 21895 + }, + { + "epoch": 3.8983262108262107, + "grad_norm": 0.9263895153999329, + "learning_rate": 3.2484704776296613e-07, + "loss": 0.8624, + "step": 21896 + }, + { + "epoch": 3.8985042735042734, + "grad_norm": 1.0885869264602661, + "learning_rate": 3.23720694170071e-07, + "loss": 0.8088, + "step": 21897 + }, + { + "epoch": 3.898682336182336, + "grad_norm": 1.0078108310699463, + "learning_rate": 3.225962935268112e-07, + "loss": 0.8197, + "step": 21898 + }, + { + "epoch": 3.898860398860399, + "grad_norm": 1.0098800659179688, + "learning_rate": 3.2147384585521354e-07, + "loss": 0.748, + "step": 21899 + }, + { + "epoch": 3.8990384615384617, + "grad_norm": 0.9592456817626953, + "learning_rate": 3.203533511772605e-07, + "loss": 0.7737, + "step": 21900 + }, + { + "epoch": 3.8992165242165244, + "grad_norm": 0.9254228472709656, + "learning_rate": 3.1923480951493447e-07, + "loss": 0.8406, + "step": 21901 + }, + { + "epoch": 3.8993945868945867, + "grad_norm": 0.9102961421012878, + "learning_rate": 3.181182208901179e-07, + "loss": 0.9384, + "step": 21902 + }, + { + "epoch": 3.8995726495726495, + "grad_norm": 0.9881536364555359, + "learning_rate": 3.170035853247155e-07, + "loss": 0.7607, + "step": 21903 + }, + { + "epoch": 3.8997507122507122, + "grad_norm": 0.9650844931602478, + "learning_rate": 3.1589090284055436e-07, + "loss": 0.7619, + "step": 21904 + }, + { + "epoch": 3.899928774928775, + "grad_norm": 0.7324507832527161, + "learning_rate": 3.1478017345942803e-07, + "loss": 0.5913, + "step": 21905 + }, + { + "epoch": 3.9001068376068377, + "grad_norm": 0.9290045499801636, + "learning_rate": 3.1367139720310803e-07, + "loss": 0.719, + "step": 21906 + }, + { + "epoch": 3.9002849002849, + "grad_norm": 1.1251944303512573, + "learning_rate": 3.125645740933214e-07, + "loss": 0.7008, + "step": 21907 + }, + { + "epoch": 3.900462962962963, + "grad_norm": 1.1408454179763794, + "learning_rate": 3.1145970415173975e-07, + "loss": 0.7176, + "step": 21908 + }, + { + "epoch": 3.9006410256410255, + "grad_norm": 0.9022751450538635, + "learning_rate": 3.103567874000235e-07, + "loss": 0.7632, + "step": 21909 + }, + { + "epoch": 3.9008190883190883, + "grad_norm": 0.8978546857833862, + "learning_rate": 3.092558238597887e-07, + "loss": 0.7441, + "step": 21910 + }, + { + "epoch": 3.900997150997151, + "grad_norm": 0.9025968909263611, + "learning_rate": 3.0815681355258477e-07, + "loss": 0.6771, + "step": 21911 + }, + { + "epoch": 3.901175213675214, + "grad_norm": 0.9397212862968445, + "learning_rate": 3.070597564999611e-07, + "loss": 0.7661, + "step": 21912 + }, + { + "epoch": 3.9013532763532766, + "grad_norm": 0.9135825037956238, + "learning_rate": 3.0596465272340056e-07, + "loss": 0.7339, + "step": 21913 + }, + { + "epoch": 3.9015313390313393, + "grad_norm": 1.1414575576782227, + "learning_rate": 3.048715022443749e-07, + "loss": 0.8099, + "step": 21914 + }, + { + "epoch": 3.9017094017094016, + "grad_norm": 0.9361329078674316, + "learning_rate": 3.0378030508428913e-07, + "loss": 0.813, + "step": 21915 + }, + { + "epoch": 3.9018874643874644, + "grad_norm": 0.9217289090156555, + "learning_rate": 3.0269106126452617e-07, + "loss": 0.5702, + "step": 21916 + }, + { + "epoch": 3.902065527065527, + "grad_norm": 1.1916371583938599, + "learning_rate": 3.0160377080643563e-07, + "loss": 0.8607, + "step": 21917 + }, + { + "epoch": 3.90224358974359, + "grad_norm": 0.9516651630401611, + "learning_rate": 3.005184337313116e-07, + "loss": 0.6183, + "step": 21918 + }, + { + "epoch": 3.902421652421652, + "grad_norm": 0.9865080118179321, + "learning_rate": 2.994350500604148e-07, + "loss": 0.7953, + "step": 21919 + }, + { + "epoch": 3.902599715099715, + "grad_norm": 0.9364351034164429, + "learning_rate": 2.983536198149839e-07, + "loss": 0.6312, + "step": 21920 + }, + { + "epoch": 3.9027777777777777, + "grad_norm": 1.0097970962524414, + "learning_rate": 2.9727414301620184e-07, + "loss": 0.8344, + "step": 21921 + }, + { + "epoch": 3.9029558404558404, + "grad_norm": 1.025741696357727, + "learning_rate": 2.961966196852184e-07, + "loss": 0.8122, + "step": 21922 + }, + { + "epoch": 3.903133903133903, + "grad_norm": 0.9230335354804993, + "learning_rate": 2.9512104984313893e-07, + "loss": 0.6786, + "step": 21923 + }, + { + "epoch": 3.903311965811966, + "grad_norm": 1.030914068222046, + "learning_rate": 2.9404743351105767e-07, + "loss": 0.7809, + "step": 21924 + }, + { + "epoch": 3.9034900284900287, + "grad_norm": 0.8610934019088745, + "learning_rate": 2.929757707099801e-07, + "loss": 0.6601, + "step": 21925 + }, + { + "epoch": 3.9036680911680914, + "grad_norm": 1.0533103942871094, + "learning_rate": 2.919060614609226e-07, + "loss": 0.8471, + "step": 21926 + }, + { + "epoch": 3.9038461538461537, + "grad_norm": 1.0236098766326904, + "learning_rate": 2.908383057848463e-07, + "loss": 0.8785, + "step": 21927 + }, + { + "epoch": 3.9040242165242165, + "grad_norm": 0.9521497488021851, + "learning_rate": 2.897725037026566e-07, + "loss": 0.8093, + "step": 21928 + }, + { + "epoch": 3.9042022792022792, + "grad_norm": 1.0812114477157593, + "learning_rate": 2.8870865523525915e-07, + "loss": 0.8727, + "step": 21929 + }, + { + "epoch": 3.904380341880342, + "grad_norm": 1.0053845643997192, + "learning_rate": 2.876467604034705e-07, + "loss": 0.7937, + "step": 21930 + }, + { + "epoch": 3.9045584045584043, + "grad_norm": 0.722806990146637, + "learning_rate": 2.865868192281074e-07, + "loss": 0.5283, + "step": 21931 + }, + { + "epoch": 3.904736467236467, + "grad_norm": 0.8876988291740417, + "learning_rate": 2.855288317299531e-07, + "loss": 0.6507, + "step": 21932 + }, + { + "epoch": 3.90491452991453, + "grad_norm": 0.9309136271476746, + "learning_rate": 2.8447279792971346e-07, + "loss": 0.6041, + "step": 21933 + }, + { + "epoch": 3.9050925925925926, + "grad_norm": 1.25245201587677, + "learning_rate": 2.8341871784808293e-07, + "loss": 0.7069, + "step": 21934 + }, + { + "epoch": 3.9052706552706553, + "grad_norm": 0.8108494281768799, + "learning_rate": 2.8236659150572274e-07, + "loss": 0.4754, + "step": 21935 + }, + { + "epoch": 3.905448717948718, + "grad_norm": 0.97728031873703, + "learning_rate": 2.813164189232498e-07, + "loss": 0.8704, + "step": 21936 + }, + { + "epoch": 3.905626780626781, + "grad_norm": 1.0768465995788574, + "learning_rate": 2.8026820012123645e-07, + "loss": 0.7118, + "step": 21937 + }, + { + "epoch": 3.9058048433048436, + "grad_norm": 0.9180233478546143, + "learning_rate": 2.7922193512019965e-07, + "loss": 0.9206, + "step": 21938 + }, + { + "epoch": 3.905982905982906, + "grad_norm": 1.0183178186416626, + "learning_rate": 2.781776239406786e-07, + "loss": 0.8764, + "step": 21939 + }, + { + "epoch": 3.9061609686609686, + "grad_norm": 1.081843614578247, + "learning_rate": 2.771352666031013e-07, + "loss": 0.7171, + "step": 21940 + }, + { + "epoch": 3.9063390313390314, + "grad_norm": 1.008541464805603, + "learning_rate": 2.76094863127907e-07, + "loss": 0.6932, + "step": 21941 + }, + { + "epoch": 3.906517094017094, + "grad_norm": 0.9277056455612183, + "learning_rate": 2.750564135354683e-07, + "loss": 0.787, + "step": 21942 + }, + { + "epoch": 3.9066951566951564, + "grad_norm": 1.0183097124099731, + "learning_rate": 2.7401991784614666e-07, + "loss": 0.966, + "step": 21943 + }, + { + "epoch": 3.906873219373219, + "grad_norm": 0.943051815032959, + "learning_rate": 2.72985376080237e-07, + "loss": 0.6795, + "step": 21944 + }, + { + "epoch": 3.907051282051282, + "grad_norm": 0.9863778352737427, + "learning_rate": 2.7195278825801195e-07, + "loss": 0.6972, + "step": 21945 + }, + { + "epoch": 3.9072293447293447, + "grad_norm": 1.0453509092330933, + "learning_rate": 2.709221543997109e-07, + "loss": 0.8906, + "step": 21946 + }, + { + "epoch": 3.9074074074074074, + "grad_norm": 0.9586603045463562, + "learning_rate": 2.698934745255177e-07, + "loss": 0.8719, + "step": 21947 + }, + { + "epoch": 3.90758547008547, + "grad_norm": 0.9506910443305969, + "learning_rate": 2.6886674865559403e-07, + "loss": 0.6272, + "step": 21948 + }, + { + "epoch": 3.907763532763533, + "grad_norm": 0.9721106886863708, + "learning_rate": 2.6784197681004595e-07, + "loss": 0.8897, + "step": 21949 + }, + { + "epoch": 3.9079415954415957, + "grad_norm": 0.917022168636322, + "learning_rate": 2.6681915900896857e-07, + "loss": 0.8084, + "step": 21950 + }, + { + "epoch": 3.908119658119658, + "grad_norm": 1.0294543504714966, + "learning_rate": 2.657982952723792e-07, + "loss": 0.8383, + "step": 21951 + }, + { + "epoch": 3.9082977207977208, + "grad_norm": 1.017067313194275, + "learning_rate": 2.647793856203062e-07, + "loss": 0.9104, + "step": 21952 + }, + { + "epoch": 3.9084757834757835, + "grad_norm": 0.9635358452796936, + "learning_rate": 2.637624300726893e-07, + "loss": 0.8508, + "step": 21953 + }, + { + "epoch": 3.9086538461538463, + "grad_norm": 1.0203074216842651, + "learning_rate": 2.62747428649468e-07, + "loss": 0.8975, + "step": 21954 + }, + { + "epoch": 3.9088319088319086, + "grad_norm": 0.8658331036567688, + "learning_rate": 2.617343813705264e-07, + "loss": 0.6994, + "step": 21955 + }, + { + "epoch": 3.9090099715099713, + "grad_norm": 1.0209558010101318, + "learning_rate": 2.6072328825570425e-07, + "loss": 0.8763, + "step": 21956 + }, + { + "epoch": 3.909188034188034, + "grad_norm": 1.0200470685958862, + "learning_rate": 2.59714149324819e-07, + "loss": 0.7055, + "step": 21957 + }, + { + "epoch": 3.909366096866097, + "grad_norm": 1.178375244140625, + "learning_rate": 2.5870696459764365e-07, + "loss": 0.7509, + "step": 21958 + }, + { + "epoch": 3.9095441595441596, + "grad_norm": 0.8727983832359314, + "learning_rate": 2.577017340939181e-07, + "loss": 0.7704, + "step": 21959 + }, + { + "epoch": 3.9097222222222223, + "grad_norm": 0.9848031401634216, + "learning_rate": 2.5669845783332645e-07, + "loss": 0.8052, + "step": 21960 + }, + { + "epoch": 3.909900284900285, + "grad_norm": 0.877185583114624, + "learning_rate": 2.556971358355309e-07, + "loss": 0.8031, + "step": 21961 + }, + { + "epoch": 3.910078347578348, + "grad_norm": 0.9560007452964783, + "learning_rate": 2.54697768120149e-07, + "loss": 0.9216, + "step": 21962 + }, + { + "epoch": 3.91025641025641, + "grad_norm": 0.8177520632743835, + "learning_rate": 2.5370035470675404e-07, + "loss": 0.7067, + "step": 21963 + }, + { + "epoch": 3.910434472934473, + "grad_norm": 1.050285816192627, + "learning_rate": 2.5270489561490807e-07, + "loss": 0.7827, + "step": 21964 + }, + { + "epoch": 3.9106125356125356, + "grad_norm": 0.9988555908203125, + "learning_rate": 2.5171139086408444e-07, + "loss": 0.7919, + "step": 21965 + }, + { + "epoch": 3.9107905982905984, + "grad_norm": 1.05731999874115, + "learning_rate": 2.5071984047377873e-07, + "loss": 0.9199, + "step": 21966 + }, + { + "epoch": 3.9109686609686607, + "grad_norm": 0.9310580492019653, + "learning_rate": 2.4973024446340864e-07, + "loss": 0.5524, + "step": 21967 + }, + { + "epoch": 3.9111467236467234, + "grad_norm": 0.8600029349327087, + "learning_rate": 2.487426028523587e-07, + "loss": 0.6444, + "step": 21968 + }, + { + "epoch": 3.911324786324786, + "grad_norm": 0.8115721344947815, + "learning_rate": 2.4775691565998014e-07, + "loss": 0.6268, + "step": 21969 + }, + { + "epoch": 3.911502849002849, + "grad_norm": 1.136781930923462, + "learning_rate": 2.467731829055908e-07, + "loss": 0.8586, + "step": 21970 + }, + { + "epoch": 3.9116809116809117, + "grad_norm": 0.9071054458618164, + "learning_rate": 2.4579140460846415e-07, + "loss": 0.7965, + "step": 21971 + }, + { + "epoch": 3.9118589743589745, + "grad_norm": 1.0506306886672974, + "learning_rate": 2.448115807878293e-07, + "loss": 0.752, + "step": 21972 + }, + { + "epoch": 3.912037037037037, + "grad_norm": 1.117159366607666, + "learning_rate": 2.438337114628819e-07, + "loss": 0.7914, + "step": 21973 + }, + { + "epoch": 3.9122150997151, + "grad_norm": 0.892483115196228, + "learning_rate": 2.4285779665280675e-07, + "loss": 0.8148, + "step": 21974 + }, + { + "epoch": 3.9123931623931623, + "grad_norm": 0.9833968281745911, + "learning_rate": 2.4188383637668845e-07, + "loss": 0.7573, + "step": 21975 + }, + { + "epoch": 3.912571225071225, + "grad_norm": 0.9161805510520935, + "learning_rate": 2.409118306536229e-07, + "loss": 0.7162, + "step": 21976 + }, + { + "epoch": 3.9127492877492878, + "grad_norm": 0.9161592125892639, + "learning_rate": 2.399417795026726e-07, + "loss": 0.7717, + "step": 21977 + }, + { + "epoch": 3.9129273504273505, + "grad_norm": 0.8876699805259705, + "learning_rate": 2.389736829428224e-07, + "loss": 0.5481, + "step": 21978 + }, + { + "epoch": 3.9131054131054133, + "grad_norm": 0.9949450492858887, + "learning_rate": 2.3800754099304602e-07, + "loss": 0.6204, + "step": 21979 + }, + { + "epoch": 3.9132834757834756, + "grad_norm": 0.9041038751602173, + "learning_rate": 2.370433536722838e-07, + "loss": 0.7954, + "step": 21980 + }, + { + "epoch": 3.9134615384615383, + "grad_norm": 0.9964247941970825, + "learning_rate": 2.360811209994096e-07, + "loss": 0.8088, + "step": 21981 + }, + { + "epoch": 3.913639601139601, + "grad_norm": 0.9314337372779846, + "learning_rate": 2.3512084299328606e-07, + "loss": 0.6992, + "step": 21982 + }, + { + "epoch": 3.913817663817664, + "grad_norm": 0.997922956943512, + "learning_rate": 2.341625196727204e-07, + "loss": 0.7194, + "step": 21983 + }, + { + "epoch": 3.9139957264957266, + "grad_norm": 0.9609016180038452, + "learning_rate": 2.3320615105649757e-07, + "loss": 0.7588, + "step": 21984 + }, + { + "epoch": 3.9141737891737893, + "grad_norm": 1.006381630897522, + "learning_rate": 2.3225173716335812e-07, + "loss": 0.8141, + "step": 21985 + }, + { + "epoch": 3.914351851851852, + "grad_norm": 0.9407157897949219, + "learning_rate": 2.3129927801198715e-07, + "loss": 0.7846, + "step": 21986 + }, + { + "epoch": 3.9145299145299144, + "grad_norm": 0.8722037672996521, + "learning_rate": 2.3034877362106967e-07, + "loss": 0.6581, + "step": 21987 + }, + { + "epoch": 3.914707977207977, + "grad_norm": 0.9266372323036194, + "learning_rate": 2.2940022400920192e-07, + "loss": 0.8043, + "step": 21988 + }, + { + "epoch": 3.91488603988604, + "grad_norm": 1.1715607643127441, + "learning_rate": 2.2845362919498015e-07, + "loss": 0.8472, + "step": 21989 + }, + { + "epoch": 3.9150641025641026, + "grad_norm": 0.9698849320411682, + "learning_rate": 2.2750898919695617e-07, + "loss": 0.6895, + "step": 21990 + }, + { + "epoch": 3.9152421652421654, + "grad_norm": 1.0203914642333984, + "learning_rate": 2.2656630403363743e-07, + "loss": 0.7114, + "step": 21991 + }, + { + "epoch": 3.9154202279202277, + "grad_norm": 0.9018568992614746, + "learning_rate": 2.2562557372348692e-07, + "loss": 0.6274, + "step": 21992 + }, + { + "epoch": 3.9155982905982905, + "grad_norm": 0.8307010531425476, + "learning_rate": 2.2468679828494544e-07, + "loss": 0.5552, + "step": 21993 + }, + { + "epoch": 3.915776353276353, + "grad_norm": 1.0528616905212402, + "learning_rate": 2.2374997773639827e-07, + "loss": 1.0276, + "step": 21994 + }, + { + "epoch": 3.915954415954416, + "grad_norm": 0.879680335521698, + "learning_rate": 2.2281511209619744e-07, + "loss": 0.8538, + "step": 21995 + }, + { + "epoch": 3.9161324786324787, + "grad_norm": 0.9187359809875488, + "learning_rate": 2.218822013826727e-07, + "loss": 0.7884, + "step": 21996 + }, + { + "epoch": 3.9163105413105415, + "grad_norm": 0.9949030876159668, + "learning_rate": 2.209512456140872e-07, + "loss": 0.663, + "step": 21997 + }, + { + "epoch": 3.916488603988604, + "grad_norm": 0.8498545289039612, + "learning_rate": 2.2002224480869305e-07, + "loss": 0.7736, + "step": 21998 + }, + { + "epoch": 3.9166666666666665, + "grad_norm": 1.050980806350708, + "learning_rate": 2.1909519898468678e-07, + "loss": 0.8053, + "step": 21999 + }, + { + "epoch": 3.9168447293447293, + "grad_norm": 0.9050692319869995, + "learning_rate": 2.1817010816024275e-07, + "loss": 0.766, + "step": 22000 + }, + { + "epoch": 3.917022792022792, + "grad_norm": 0.9989611506462097, + "learning_rate": 2.172469723534687e-07, + "loss": 0.8758, + "step": 22001 + }, + { + "epoch": 3.9172008547008548, + "grad_norm": 0.8801424503326416, + "learning_rate": 2.1632579158246124e-07, + "loss": 0.7643, + "step": 22002 + }, + { + "epoch": 3.9173789173789175, + "grad_norm": 0.8321191668510437, + "learning_rate": 2.1540656586526152e-07, + "loss": 0.7413, + "step": 22003 + }, + { + "epoch": 3.91755698005698, + "grad_norm": 0.9452522993087769, + "learning_rate": 2.1448929521988848e-07, + "loss": 0.7105, + "step": 22004 + }, + { + "epoch": 3.9177350427350426, + "grad_norm": 1.0694053173065186, + "learning_rate": 2.135739796643166e-07, + "loss": 0.6839, + "step": 22005 + }, + { + "epoch": 3.9179131054131053, + "grad_norm": 0.8962434530258179, + "learning_rate": 2.1266061921646485e-07, + "loss": 0.7743, + "step": 22006 + }, + { + "epoch": 3.918091168091168, + "grad_norm": 0.9079770445823669, + "learning_rate": 2.117492138942412e-07, + "loss": 0.7312, + "step": 22007 + }, + { + "epoch": 3.918269230769231, + "grad_norm": 0.9559836387634277, + "learning_rate": 2.1083976371550907e-07, + "loss": 0.7365, + "step": 22008 + }, + { + "epoch": 3.9184472934472936, + "grad_norm": 1.2130017280578613, + "learning_rate": 2.0993226869806536e-07, + "loss": 1.0424, + "step": 22009 + }, + { + "epoch": 3.9186253561253563, + "grad_norm": 0.9800707101821899, + "learning_rate": 2.0902672885970697e-07, + "loss": 0.8919, + "step": 22010 + }, + { + "epoch": 3.9188034188034186, + "grad_norm": 0.9185284376144409, + "learning_rate": 2.0812314421817524e-07, + "loss": 0.5581, + "step": 22011 + }, + { + "epoch": 3.9189814814814814, + "grad_norm": 1.0603922605514526, + "learning_rate": 2.0722151479116714e-07, + "loss": 0.7397, + "step": 22012 + }, + { + "epoch": 3.919159544159544, + "grad_norm": 0.9114738702774048, + "learning_rate": 2.063218405963574e-07, + "loss": 0.7737, + "step": 22013 + }, + { + "epoch": 3.919337606837607, + "grad_norm": 0.8841904401779175, + "learning_rate": 2.0542412165136526e-07, + "loss": 0.7496, + "step": 22014 + }, + { + "epoch": 3.9195156695156697, + "grad_norm": 0.9699673652648926, + "learning_rate": 2.0452835797377667e-07, + "loss": 0.4983, + "step": 22015 + }, + { + "epoch": 3.919693732193732, + "grad_norm": 0.87566739320755, + "learning_rate": 2.0363454958115536e-07, + "loss": 0.7714, + "step": 22016 + }, + { + "epoch": 3.9198717948717947, + "grad_norm": 1.0994035005569458, + "learning_rate": 2.027426964909984e-07, + "loss": 0.8937, + "step": 22017 + }, + { + "epoch": 3.9200498575498575, + "grad_norm": 0.9632015228271484, + "learning_rate": 2.0185279872079188e-07, + "loss": 0.6813, + "step": 22018 + }, + { + "epoch": 3.92022792022792, + "grad_norm": 1.015339970588684, + "learning_rate": 2.0096485628796625e-07, + "loss": 0.8103, + "step": 22019 + }, + { + "epoch": 3.920405982905983, + "grad_norm": 1.0228614807128906, + "learning_rate": 2.000788692099187e-07, + "loss": 0.8716, + "step": 22020 + }, + { + "epoch": 3.9205840455840457, + "grad_norm": 0.8183853626251221, + "learning_rate": 1.9919483750401313e-07, + "loss": 0.6534, + "step": 22021 + }, + { + "epoch": 3.9207621082621085, + "grad_norm": 1.1893196105957031, + "learning_rate": 1.9831276118756903e-07, + "loss": 0.8373, + "step": 22022 + }, + { + "epoch": 3.9209401709401708, + "grad_norm": 1.0825175046920776, + "learning_rate": 1.9743264027786147e-07, + "loss": 0.7538, + "step": 22023 + }, + { + "epoch": 3.9211182336182335, + "grad_norm": 1.1777693033218384, + "learning_rate": 1.965544747921322e-07, + "loss": 0.6503, + "step": 22024 + }, + { + "epoch": 3.9212962962962963, + "grad_norm": 0.8321196436882019, + "learning_rate": 1.9567826474760075e-07, + "loss": 0.6849, + "step": 22025 + }, + { + "epoch": 3.921474358974359, + "grad_norm": 0.930051863193512, + "learning_rate": 1.9480401016143123e-07, + "loss": 0.5483, + "step": 22026 + }, + { + "epoch": 3.921652421652422, + "grad_norm": 0.8987363576889038, + "learning_rate": 1.9393171105075435e-07, + "loss": 0.7334, + "step": 22027 + }, + { + "epoch": 3.921830484330484, + "grad_norm": 0.956389307975769, + "learning_rate": 1.9306136743264536e-07, + "loss": 0.7411, + "step": 22028 + }, + { + "epoch": 3.922008547008547, + "grad_norm": 1.0126129388809204, + "learning_rate": 1.9219297932416836e-07, + "loss": 0.8383, + "step": 22029 + }, + { + "epoch": 3.9221866096866096, + "grad_norm": 0.9052560925483704, + "learning_rate": 1.9132654674234306e-07, + "loss": 0.7371, + "step": 22030 + }, + { + "epoch": 3.9223646723646723, + "grad_norm": 0.9211751818656921, + "learning_rate": 1.9046206970414483e-07, + "loss": 0.677, + "step": 22031 + }, + { + "epoch": 3.922542735042735, + "grad_norm": 0.9746459722518921, + "learning_rate": 1.8959954822649339e-07, + "loss": 0.9056, + "step": 22032 + }, + { + "epoch": 3.922720797720798, + "grad_norm": 0.9646094441413879, + "learning_rate": 1.887389823263086e-07, + "loss": 0.8162, + "step": 22033 + }, + { + "epoch": 3.9228988603988606, + "grad_norm": 0.9268381595611572, + "learning_rate": 1.8788037202044363e-07, + "loss": 0.977, + "step": 22034 + }, + { + "epoch": 3.9230769230769234, + "grad_norm": 1.0845803022384644, + "learning_rate": 1.8702371732571834e-07, + "loss": 0.803, + "step": 22035 + }, + { + "epoch": 3.9232549857549857, + "grad_norm": 0.8709145188331604, + "learning_rate": 1.861690182589193e-07, + "loss": 0.8107, + "step": 22036 + }, + { + "epoch": 3.9234330484330484, + "grad_norm": 0.9537857174873352, + "learning_rate": 1.853162748367887e-07, + "loss": 0.7444, + "step": 22037 + }, + { + "epoch": 3.923611111111111, + "grad_norm": 0.8645097017288208, + "learning_rate": 1.8446548707604648e-07, + "loss": 0.7448, + "step": 22038 + }, + { + "epoch": 3.923789173789174, + "grad_norm": 0.921103835105896, + "learning_rate": 1.8361665499334602e-07, + "loss": 0.768, + "step": 22039 + }, + { + "epoch": 3.923967236467236, + "grad_norm": 0.9298867583274841, + "learning_rate": 1.827697786053295e-07, + "loss": 0.6328, + "step": 22040 + }, + { + "epoch": 3.924145299145299, + "grad_norm": 1.068195104598999, + "learning_rate": 1.8192485792859483e-07, + "loss": 0.764, + "step": 22041 + }, + { + "epoch": 3.9243233618233617, + "grad_norm": 1.0562421083450317, + "learning_rate": 1.8108189297968425e-07, + "loss": 0.9191, + "step": 22042 + }, + { + "epoch": 3.9245014245014245, + "grad_norm": 0.984227180480957, + "learning_rate": 1.8024088377511795e-07, + "loss": 0.7772, + "step": 22043 + }, + { + "epoch": 3.9246794871794872, + "grad_norm": 1.0047521591186523, + "learning_rate": 1.794018303313716e-07, + "loss": 0.5225, + "step": 22044 + }, + { + "epoch": 3.92485754985755, + "grad_norm": 0.8145391941070557, + "learning_rate": 1.785647326648876e-07, + "loss": 0.5202, + "step": 22045 + }, + { + "epoch": 3.9250356125356127, + "grad_norm": 0.91175377368927, + "learning_rate": 1.7772959079206396e-07, + "loss": 0.6479, + "step": 22046 + }, + { + "epoch": 3.9252136752136755, + "grad_norm": 0.9365923404693604, + "learning_rate": 1.7689640472926538e-07, + "loss": 0.8265, + "step": 22047 + }, + { + "epoch": 3.925391737891738, + "grad_norm": 0.8597645163536072, + "learning_rate": 1.7606517449282323e-07, + "loss": 0.5941, + "step": 22048 + }, + { + "epoch": 3.9255698005698005, + "grad_norm": 0.821519672870636, + "learning_rate": 1.7523590009902445e-07, + "loss": 0.779, + "step": 22049 + }, + { + "epoch": 3.9257478632478633, + "grad_norm": 0.8071169257164001, + "learning_rate": 1.7440858156410057e-07, + "loss": 0.6711, + "step": 22050 + }, + { + "epoch": 3.925925925925926, + "grad_norm": 0.9287723302841187, + "learning_rate": 1.7358321890427188e-07, + "loss": 0.8539, + "step": 22051 + }, + { + "epoch": 3.9261039886039883, + "grad_norm": 0.9395582675933838, + "learning_rate": 1.7275981213570325e-07, + "loss": 0.7498, + "step": 22052 + }, + { + "epoch": 3.926282051282051, + "grad_norm": 0.9510295987129211, + "learning_rate": 1.7193836127453733e-07, + "loss": 0.6974, + "step": 22053 + }, + { + "epoch": 3.926460113960114, + "grad_norm": 0.9212534427642822, + "learning_rate": 1.7111886633687236e-07, + "loss": 0.8491, + "step": 22054 + }, + { + "epoch": 3.9266381766381766, + "grad_norm": 1.1693989038467407, + "learning_rate": 1.7030132733873994e-07, + "loss": 0.9457, + "step": 22055 + }, + { + "epoch": 3.9268162393162394, + "grad_norm": 0.9136499762535095, + "learning_rate": 1.6948574429618282e-07, + "loss": 0.7862, + "step": 22056 + }, + { + "epoch": 3.926994301994302, + "grad_norm": 0.9177721738815308, + "learning_rate": 1.6867211722517706e-07, + "loss": 0.7247, + "step": 22057 + }, + { + "epoch": 3.927172364672365, + "grad_norm": 1.0243544578552246, + "learning_rate": 1.6786044614165443e-07, + "loss": 0.8475, + "step": 22058 + }, + { + "epoch": 3.9273504273504276, + "grad_norm": 0.8945780396461487, + "learning_rate": 1.6705073106153547e-07, + "loss": 0.6658, + "step": 22059 + }, + { + "epoch": 3.92752849002849, + "grad_norm": 1.1358131170272827, + "learning_rate": 1.6624297200065197e-07, + "loss": 0.8744, + "step": 22060 + }, + { + "epoch": 3.9277065527065527, + "grad_norm": 0.9518811106681824, + "learning_rate": 1.6543716897486904e-07, + "loss": 0.9271, + "step": 22061 + }, + { + "epoch": 3.9278846153846154, + "grad_norm": 0.9149428606033325, + "learning_rate": 1.6463332199994075e-07, + "loss": 0.8896, + "step": 22062 + }, + { + "epoch": 3.928062678062678, + "grad_norm": 0.9397650361061096, + "learning_rate": 1.6383143109164333e-07, + "loss": 0.8038, + "step": 22063 + }, + { + "epoch": 3.9282407407407405, + "grad_norm": 0.9320110082626343, + "learning_rate": 1.6303149626567538e-07, + "loss": 0.6095, + "step": 22064 + }, + { + "epoch": 3.9284188034188032, + "grad_norm": 1.0589196681976318, + "learning_rate": 1.622335175377132e-07, + "loss": 0.8121, + "step": 22065 + }, + { + "epoch": 3.928596866096866, + "grad_norm": 0.821600079536438, + "learning_rate": 1.6143749492338877e-07, + "loss": 0.6471, + "step": 22066 + }, + { + "epoch": 3.9287749287749287, + "grad_norm": 0.9287008047103882, + "learning_rate": 1.606434284383007e-07, + "loss": 0.9064, + "step": 22067 + }, + { + "epoch": 3.9289529914529915, + "grad_norm": 1.094293236732483, + "learning_rate": 1.5985131809800326e-07, + "loss": 0.8028, + "step": 22068 + }, + { + "epoch": 3.9291310541310542, + "grad_norm": 0.9387109875679016, + "learning_rate": 1.5906116391801728e-07, + "loss": 0.7429, + "step": 22069 + }, + { + "epoch": 3.929309116809117, + "grad_norm": 0.994990885257721, + "learning_rate": 1.5827296591383045e-07, + "loss": 0.9219, + "step": 22070 + }, + { + "epoch": 3.9294871794871797, + "grad_norm": 0.995576024055481, + "learning_rate": 1.5748672410088593e-07, + "loss": 0.7344, + "step": 22071 + }, + { + "epoch": 3.929665242165242, + "grad_norm": 0.9641157388687134, + "learning_rate": 1.5670243849457144e-07, + "loss": 0.6646, + "step": 22072 + }, + { + "epoch": 3.929843304843305, + "grad_norm": 0.9594936370849609, + "learning_rate": 1.5592010911027467e-07, + "loss": 0.685, + "step": 22073 + }, + { + "epoch": 3.9300213675213675, + "grad_norm": 0.868777334690094, + "learning_rate": 1.5513973596331666e-07, + "loss": 0.6779, + "step": 22074 + }, + { + "epoch": 3.9301994301994303, + "grad_norm": 1.064650535583496, + "learning_rate": 1.543613190689852e-07, + "loss": 0.7801, + "step": 22075 + }, + { + "epoch": 3.9303774928774926, + "grad_norm": 1.0150713920593262, + "learning_rate": 1.535848584425348e-07, + "loss": 0.8817, + "step": 22076 + }, + { + "epoch": 3.9305555555555554, + "grad_norm": 0.932295024394989, + "learning_rate": 1.5281035409916433e-07, + "loss": 0.8153, + "step": 22077 + }, + { + "epoch": 3.930733618233618, + "grad_norm": 1.0463019609451294, + "learning_rate": 1.520378060540728e-07, + "loss": 0.7176, + "step": 22078 + }, + { + "epoch": 3.930911680911681, + "grad_norm": 0.8910303115844727, + "learning_rate": 1.5126721432238144e-07, + "loss": 0.6787, + "step": 22079 + }, + { + "epoch": 3.9310897435897436, + "grad_norm": 0.919048547744751, + "learning_rate": 1.5049857891918928e-07, + "loss": 0.7969, + "step": 22080 + }, + { + "epoch": 3.9312678062678064, + "grad_norm": 0.9526001811027527, + "learning_rate": 1.4973189985955094e-07, + "loss": 0.7759, + "step": 22081 + }, + { + "epoch": 3.931445868945869, + "grad_norm": 0.8939839601516724, + "learning_rate": 1.4896717715850995e-07, + "loss": 0.6735, + "step": 22082 + }, + { + "epoch": 3.931623931623932, + "grad_norm": 1.0000548362731934, + "learning_rate": 1.4820441083102098e-07, + "loss": 0.7515, + "step": 22083 + }, + { + "epoch": 3.931801994301994, + "grad_norm": 0.9883599281311035, + "learning_rate": 1.474436008920499e-07, + "loss": 0.6543, + "step": 22084 + }, + { + "epoch": 3.931980056980057, + "grad_norm": 0.8567184805870056, + "learning_rate": 1.4668474735649584e-07, + "loss": 0.7501, + "step": 22085 + }, + { + "epoch": 3.9321581196581197, + "grad_norm": 0.9684944748878479, + "learning_rate": 1.4592785023922473e-07, + "loss": 0.6182, + "step": 22086 + }, + { + "epoch": 3.9323361823361824, + "grad_norm": 0.97890704870224, + "learning_rate": 1.4517290955506912e-07, + "loss": 0.7792, + "step": 22087 + }, + { + "epoch": 3.932514245014245, + "grad_norm": 1.0463794469833374, + "learning_rate": 1.444199253188172e-07, + "loss": 0.812, + "step": 22088 + }, + { + "epoch": 3.9326923076923075, + "grad_norm": 0.8969535231590271, + "learning_rate": 1.4366889754523493e-07, + "loss": 0.7324, + "step": 22089 + }, + { + "epoch": 3.9328703703703702, + "grad_norm": 0.9530897736549377, + "learning_rate": 1.4291982624901058e-07, + "loss": 0.7213, + "step": 22090 + }, + { + "epoch": 3.933048433048433, + "grad_norm": 0.8949909210205078, + "learning_rate": 1.4217271144485455e-07, + "loss": 0.8138, + "step": 22091 + }, + { + "epoch": 3.9332264957264957, + "grad_norm": 0.974327027797699, + "learning_rate": 1.4142755314737744e-07, + "loss": 0.7491, + "step": 22092 + }, + { + "epoch": 3.9334045584045585, + "grad_norm": 1.043455958366394, + "learning_rate": 1.4068435137118974e-07, + "loss": 0.8267, + "step": 22093 + }, + { + "epoch": 3.9335826210826212, + "grad_norm": 0.8914118409156799, + "learning_rate": 1.399431061308576e-07, + "loss": 0.6382, + "step": 22094 + }, + { + "epoch": 3.933760683760684, + "grad_norm": 0.9680683016777039, + "learning_rate": 1.392038174408916e-07, + "loss": 0.7857, + "step": 22095 + }, + { + "epoch": 3.9339387464387463, + "grad_norm": 1.0403897762298584, + "learning_rate": 1.384664853157802e-07, + "loss": 0.784, + "step": 22096 + }, + { + "epoch": 3.934116809116809, + "grad_norm": 0.862411618232727, + "learning_rate": 1.3773110976998959e-07, + "loss": 0.7273, + "step": 22097 + }, + { + "epoch": 3.934294871794872, + "grad_norm": 0.943002462387085, + "learning_rate": 1.3699769081789714e-07, + "loss": 0.6417, + "step": 22098 + }, + { + "epoch": 3.9344729344729346, + "grad_norm": 0.9510722756385803, + "learning_rate": 1.3626622847390246e-07, + "loss": 0.7305, + "step": 22099 + }, + { + "epoch": 3.9346509971509973, + "grad_norm": 0.9381536245346069, + "learning_rate": 1.3553672275230523e-07, + "loss": 0.6609, + "step": 22100 + }, + { + "epoch": 3.9348290598290596, + "grad_norm": 0.9791659116744995, + "learning_rate": 1.3480917366742728e-07, + "loss": 0.7784, + "step": 22101 + }, + { + "epoch": 3.9350071225071224, + "grad_norm": 0.716427743434906, + "learning_rate": 1.3408358123350174e-07, + "loss": 0.42, + "step": 22102 + }, + { + "epoch": 3.935185185185185, + "grad_norm": 0.932883083820343, + "learning_rate": 1.333599454647727e-07, + "loss": 0.7054, + "step": 22103 + }, + { + "epoch": 3.935363247863248, + "grad_norm": 0.8895878195762634, + "learning_rate": 1.3263826637538445e-07, + "loss": 0.566, + "step": 22104 + }, + { + "epoch": 3.9355413105413106, + "grad_norm": 0.9334657192230225, + "learning_rate": 1.3191854397949232e-07, + "loss": 0.5716, + "step": 22105 + }, + { + "epoch": 3.9357193732193734, + "grad_norm": 0.8970123529434204, + "learning_rate": 1.3120077829120725e-07, + "loss": 0.7458, + "step": 22106 + }, + { + "epoch": 3.935897435897436, + "grad_norm": 0.9838302135467529, + "learning_rate": 1.3048496932457354e-07, + "loss": 0.866, + "step": 22107 + }, + { + "epoch": 3.9360754985754984, + "grad_norm": 0.9944798350334167, + "learning_rate": 1.2977111709363555e-07, + "loss": 0.9725, + "step": 22108 + }, + { + "epoch": 3.936253561253561, + "grad_norm": 0.9793491959571838, + "learning_rate": 1.2905922161237094e-07, + "loss": 0.6966, + "step": 22109 + }, + { + "epoch": 3.936431623931624, + "grad_norm": 0.9808713793754578, + "learning_rate": 1.2834928289472416e-07, + "loss": 0.6675, + "step": 22110 + }, + { + "epoch": 3.9366096866096867, + "grad_norm": 0.9978953003883362, + "learning_rate": 1.2764130095460624e-07, + "loss": 0.8886, + "step": 22111 + }, + { + "epoch": 3.9367877492877494, + "grad_norm": 0.9405983090400696, + "learning_rate": 1.2693527580588394e-07, + "loss": 0.9117, + "step": 22112 + }, + { + "epoch": 3.9369658119658117, + "grad_norm": 0.8421781659126282, + "learning_rate": 1.262312074624017e-07, + "loss": 0.6117, + "step": 22113 + }, + { + "epoch": 3.9371438746438745, + "grad_norm": 1.0687942504882812, + "learning_rate": 1.2552909593794847e-07, + "loss": 0.7961, + "step": 22114 + }, + { + "epoch": 3.9373219373219372, + "grad_norm": 0.9537040591239929, + "learning_rate": 1.2482894124629107e-07, + "loss": 0.7356, + "step": 22115 + }, + { + "epoch": 3.9375, + "grad_norm": 0.9954497814178467, + "learning_rate": 1.2413074340112961e-07, + "loss": 0.802, + "step": 22116 + }, + { + "epoch": 3.9376780626780628, + "grad_norm": 1.0326809883117676, + "learning_rate": 1.2343450241615318e-07, + "loss": 0.8831, + "step": 22117 + }, + { + "epoch": 3.9378561253561255, + "grad_norm": 0.9369449615478516, + "learning_rate": 1.2274021830499528e-07, + "loss": 0.7104, + "step": 22118 + }, + { + "epoch": 3.9380341880341883, + "grad_norm": 1.008576512336731, + "learning_rate": 1.2204789108127835e-07, + "loss": 0.7815, + "step": 22119 + }, + { + "epoch": 3.9382122507122506, + "grad_norm": 1.1734954118728638, + "learning_rate": 1.2135752075854712e-07, + "loss": 0.9971, + "step": 22120 + }, + { + "epoch": 3.9383903133903133, + "grad_norm": 1.0208659172058105, + "learning_rate": 1.206691073503352e-07, + "loss": 1.0925, + "step": 22121 + }, + { + "epoch": 3.938568376068376, + "grad_norm": 0.8521791696548462, + "learning_rate": 1.1998265087013182e-07, + "loss": 0.5814, + "step": 22122 + }, + { + "epoch": 3.938746438746439, + "grad_norm": 0.8949893116950989, + "learning_rate": 1.1929815133138177e-07, + "loss": 0.7053, + "step": 22123 + }, + { + "epoch": 3.9389245014245016, + "grad_norm": 0.9781447052955627, + "learning_rate": 1.1861560874750765e-07, + "loss": 0.6777, + "step": 22124 + }, + { + "epoch": 3.939102564102564, + "grad_norm": 0.9684866666793823, + "learning_rate": 1.1793502313186544e-07, + "loss": 0.8261, + "step": 22125 + }, + { + "epoch": 3.9392806267806266, + "grad_norm": 0.9222269058227539, + "learning_rate": 1.1725639449781111e-07, + "loss": 0.6531, + "step": 22126 + }, + { + "epoch": 3.9394586894586894, + "grad_norm": 1.074013113975525, + "learning_rate": 1.1657972285862295e-07, + "loss": 0.881, + "step": 22127 + }, + { + "epoch": 3.939636752136752, + "grad_norm": 0.9172451496124268, + "learning_rate": 1.1590500822756811e-07, + "loss": 0.8913, + "step": 22128 + }, + { + "epoch": 3.939814814814815, + "grad_norm": 0.8592706322669983, + "learning_rate": 1.1523225061785825e-07, + "loss": 0.6367, + "step": 22129 + }, + { + "epoch": 3.9399928774928776, + "grad_norm": 1.0467790365219116, + "learning_rate": 1.1456145004268282e-07, + "loss": 0.8392, + "step": 22130 + }, + { + "epoch": 3.9401709401709404, + "grad_norm": 0.9529557228088379, + "learning_rate": 1.1389260651518685e-07, + "loss": 0.7073, + "step": 22131 + }, + { + "epoch": 3.9403490028490027, + "grad_norm": 0.8891605138778687, + "learning_rate": 1.1322572004845988e-07, + "loss": 0.7151, + "step": 22132 + }, + { + "epoch": 3.9405270655270654, + "grad_norm": 0.9059684872627258, + "learning_rate": 1.1256079065558034e-07, + "loss": 0.5416, + "step": 22133 + }, + { + "epoch": 3.940705128205128, + "grad_norm": 0.9860732555389404, + "learning_rate": 1.1189781834958224e-07, + "loss": 0.7043, + "step": 22134 + }, + { + "epoch": 3.940883190883191, + "grad_norm": 0.956551194190979, + "learning_rate": 1.112368031434441e-07, + "loss": 0.826, + "step": 22135 + }, + { + "epoch": 3.9410612535612537, + "grad_norm": 0.9802271723747253, + "learning_rate": 1.1057774505011109e-07, + "loss": 0.8505, + "step": 22136 + }, + { + "epoch": 3.941239316239316, + "grad_norm": 0.9822153449058533, + "learning_rate": 1.0992064408251734e-07, + "loss": 0.6106, + "step": 22137 + }, + { + "epoch": 3.9414173789173788, + "grad_norm": 1.1056941747665405, + "learning_rate": 1.0926550025351923e-07, + "loss": 0.8706, + "step": 22138 + }, + { + "epoch": 3.9415954415954415, + "grad_norm": 0.9877166748046875, + "learning_rate": 1.0861231357595092e-07, + "loss": 0.8755, + "step": 22139 + }, + { + "epoch": 3.9417735042735043, + "grad_norm": 0.9555666446685791, + "learning_rate": 1.079610840626244e-07, + "loss": 0.8356, + "step": 22140 + }, + { + "epoch": 3.941951566951567, + "grad_norm": 0.9267847537994385, + "learning_rate": 1.0731181172629612e-07, + "loss": 0.8328, + "step": 22141 + }, + { + "epoch": 3.9421296296296298, + "grad_norm": 1.053410291671753, + "learning_rate": 1.0666449657967814e-07, + "loss": 0.8671, + "step": 22142 + }, + { + "epoch": 3.9423076923076925, + "grad_norm": 0.8405706286430359, + "learning_rate": 1.0601913863546032e-07, + "loss": 0.5916, + "step": 22143 + }, + { + "epoch": 3.942485754985755, + "grad_norm": 1.0977611541748047, + "learning_rate": 1.0537573790628808e-07, + "loss": 0.9064, + "step": 22144 + }, + { + "epoch": 3.9426638176638176, + "grad_norm": 0.9488881826400757, + "learning_rate": 1.0473429440476246e-07, + "loss": 0.7252, + "step": 22145 + }, + { + "epoch": 3.9428418803418803, + "grad_norm": 0.8930643200874329, + "learning_rate": 1.0409480814346229e-07, + "loss": 0.7395, + "step": 22146 + }, + { + "epoch": 3.943019943019943, + "grad_norm": 0.8349822759628296, + "learning_rate": 1.0345727913489978e-07, + "loss": 0.6557, + "step": 22147 + }, + { + "epoch": 3.943198005698006, + "grad_norm": 0.9073463082313538, + "learning_rate": 1.0282170739157604e-07, + "loss": 0.6097, + "step": 22148 + }, + { + "epoch": 3.943376068376068, + "grad_norm": 0.9441525340080261, + "learning_rate": 1.0218809292594778e-07, + "loss": 0.6664, + "step": 22149 + }, + { + "epoch": 3.943554131054131, + "grad_norm": 1.1835004091262817, + "learning_rate": 1.0155643575041618e-07, + "loss": 0.9609, + "step": 22150 + }, + { + "epoch": 3.9437321937321936, + "grad_norm": 0.8908481001853943, + "learning_rate": 1.0092673587737133e-07, + "loss": 0.591, + "step": 22151 + }, + { + "epoch": 3.9439102564102564, + "grad_norm": 1.1800652742385864, + "learning_rate": 1.0029899331913673e-07, + "loss": 0.9494, + "step": 22152 + }, + { + "epoch": 3.944088319088319, + "grad_norm": 0.9957692623138428, + "learning_rate": 9.967320808802472e-08, + "loss": 0.8205, + "step": 22153 + }, + { + "epoch": 3.944266381766382, + "grad_norm": 1.0245128870010376, + "learning_rate": 9.904938019629217e-08, + "loss": 0.7056, + "step": 22154 + }, + { + "epoch": 3.9444444444444446, + "grad_norm": 1.030775785446167, + "learning_rate": 9.842750965616265e-08, + "loss": 0.9636, + "step": 22155 + }, + { + "epoch": 3.9446225071225074, + "grad_norm": 0.9319864511489868, + "learning_rate": 9.78075964798153e-08, + "loss": 0.8961, + "step": 22156 + }, + { + "epoch": 3.9448005698005697, + "grad_norm": 0.9753043055534363, + "learning_rate": 9.718964067939596e-08, + "loss": 0.7721, + "step": 22157 + }, + { + "epoch": 3.9449786324786325, + "grad_norm": 0.8872262835502625, + "learning_rate": 9.657364226702825e-08, + "loss": 0.6485, + "step": 22158 + }, + { + "epoch": 3.945156695156695, + "grad_norm": 0.924390435218811, + "learning_rate": 9.595960125475812e-08, + "loss": 0.6058, + "step": 22159 + }, + { + "epoch": 3.945334757834758, + "grad_norm": 0.9848852753639221, + "learning_rate": 9.534751765462036e-08, + "loss": 0.7306, + "step": 22160 + }, + { + "epoch": 3.9455128205128203, + "grad_norm": 0.9805606007575989, + "learning_rate": 9.473739147862759e-08, + "loss": 0.8002, + "step": 22161 + }, + { + "epoch": 3.945690883190883, + "grad_norm": 1.0841553211212158, + "learning_rate": 9.412922273871471e-08, + "loss": 0.8198, + "step": 22162 + }, + { + "epoch": 3.9458689458689458, + "grad_norm": 1.023932695388794, + "learning_rate": 9.352301144680554e-08, + "loss": 0.8471, + "step": 22163 + }, + { + "epoch": 3.9460470085470085, + "grad_norm": 0.9031496644020081, + "learning_rate": 9.291875761476831e-08, + "loss": 0.7516, + "step": 22164 + }, + { + "epoch": 3.9462250712250713, + "grad_norm": 1.0119421482086182, + "learning_rate": 9.231646125446025e-08, + "loss": 0.8783, + "step": 22165 + }, + { + "epoch": 3.946403133903134, + "grad_norm": 0.9956295490264893, + "learning_rate": 9.17161223776608e-08, + "loss": 0.7641, + "step": 22166 + }, + { + "epoch": 3.9465811965811968, + "grad_norm": 0.9614377617835999, + "learning_rate": 9.111774099614945e-08, + "loss": 0.8233, + "step": 22167 + }, + { + "epoch": 3.9467592592592595, + "grad_norm": 1.1682920455932617, + "learning_rate": 9.052131712163903e-08, + "loss": 0.8376, + "step": 22168 + }, + { + "epoch": 3.946937321937322, + "grad_norm": 0.8535422682762146, + "learning_rate": 8.992685076582019e-08, + "loss": 0.5982, + "step": 22169 + }, + { + "epoch": 3.9471153846153846, + "grad_norm": 0.9734843969345093, + "learning_rate": 8.933434194033919e-08, + "loss": 0.6912, + "step": 22170 + }, + { + "epoch": 3.9472934472934473, + "grad_norm": 1.0210565328598022, + "learning_rate": 8.874379065680893e-08, + "loss": 0.8513, + "step": 22171 + }, + { + "epoch": 3.94747150997151, + "grad_norm": 0.9472604393959045, + "learning_rate": 8.815519692678686e-08, + "loss": 0.7962, + "step": 22172 + }, + { + "epoch": 3.9476495726495724, + "grad_norm": 0.9001045227050781, + "learning_rate": 8.756856076183039e-08, + "loss": 0.8089, + "step": 22173 + }, + { + "epoch": 3.947827635327635, + "grad_norm": 0.9715784192085266, + "learning_rate": 8.698388217340813e-08, + "loss": 0.6414, + "step": 22174 + }, + { + "epoch": 3.948005698005698, + "grad_norm": 0.8441361784934998, + "learning_rate": 8.640116117298869e-08, + "loss": 0.8897, + "step": 22175 + }, + { + "epoch": 3.9481837606837606, + "grad_norm": 1.0581719875335693, + "learning_rate": 8.582039777197404e-08, + "loss": 0.7248, + "step": 22176 + }, + { + "epoch": 3.9483618233618234, + "grad_norm": 1.1761430501937866, + "learning_rate": 8.524159198176618e-08, + "loss": 0.9026, + "step": 22177 + }, + { + "epoch": 3.948539886039886, + "grad_norm": 0.9691876769065857, + "learning_rate": 8.466474381370049e-08, + "loss": 0.7835, + "step": 22178 + }, + { + "epoch": 3.948717948717949, + "grad_norm": 0.9321456551551819, + "learning_rate": 8.408985327905683e-08, + "loss": 0.6644, + "step": 22179 + }, + { + "epoch": 3.9488960113960117, + "grad_norm": 0.8379319906234741, + "learning_rate": 8.351692038912618e-08, + "loss": 0.4799, + "step": 22180 + }, + { + "epoch": 3.949074074074074, + "grad_norm": 1.0753391981124878, + "learning_rate": 8.294594515512177e-08, + "loss": 0.8837, + "step": 22181 + }, + { + "epoch": 3.9492521367521367, + "grad_norm": 0.8955191373825073, + "learning_rate": 8.237692758823468e-08, + "loss": 0.6253, + "step": 22182 + }, + { + "epoch": 3.9494301994301995, + "grad_norm": 1.1359809637069702, + "learning_rate": 8.180986769960041e-08, + "loss": 0.7331, + "step": 22183 + }, + { + "epoch": 3.949608262108262, + "grad_norm": 0.8849379420280457, + "learning_rate": 8.124476550034344e-08, + "loss": 0.9057, + "step": 22184 + }, + { + "epoch": 3.9497863247863245, + "grad_norm": 1.1383916139602661, + "learning_rate": 8.068162100154375e-08, + "loss": 0.7332, + "step": 22185 + }, + { + "epoch": 3.9499643874643873, + "grad_norm": 0.9934820532798767, + "learning_rate": 8.012043421421478e-08, + "loss": 0.8742, + "step": 22186 + }, + { + "epoch": 3.95014245014245, + "grad_norm": 1.0497565269470215, + "learning_rate": 7.956120514935882e-08, + "loss": 0.8896, + "step": 22187 + }, + { + "epoch": 3.9503205128205128, + "grad_norm": 0.916329026222229, + "learning_rate": 7.900393381793381e-08, + "loss": 0.82, + "step": 22188 + }, + { + "epoch": 3.9504985754985755, + "grad_norm": 0.9675973653793335, + "learning_rate": 7.844862023085319e-08, + "loss": 0.8465, + "step": 22189 + }, + { + "epoch": 3.9506766381766383, + "grad_norm": 1.0843769311904907, + "learning_rate": 7.78952643990194e-08, + "loss": 0.7282, + "step": 22190 + }, + { + "epoch": 3.950854700854701, + "grad_norm": 0.8486320376396179, + "learning_rate": 7.734386633324597e-08, + "loss": 0.662, + "step": 22191 + }, + { + "epoch": 3.951032763532764, + "grad_norm": 0.8972040414810181, + "learning_rate": 7.67944260443465e-08, + "loss": 0.8706, + "step": 22192 + }, + { + "epoch": 3.951210826210826, + "grad_norm": 0.9764129519462585, + "learning_rate": 7.624694354309014e-08, + "loss": 0.6795, + "step": 22193 + }, + { + "epoch": 3.951388888888889, + "grad_norm": 1.0516738891601562, + "learning_rate": 7.570141884020166e-08, + "loss": 0.8508, + "step": 22194 + }, + { + "epoch": 3.9515669515669516, + "grad_norm": 1.031455636024475, + "learning_rate": 7.515785194637249e-08, + "loss": 0.8451, + "step": 22195 + }, + { + "epoch": 3.9517450142450143, + "grad_norm": 1.0027506351470947, + "learning_rate": 7.461624287224966e-08, + "loss": 0.6502, + "step": 22196 + }, + { + "epoch": 3.9519230769230766, + "grad_norm": 0.933249294757843, + "learning_rate": 7.407659162843583e-08, + "loss": 0.9345, + "step": 22197 + }, + { + "epoch": 3.9521011396011394, + "grad_norm": 0.7095343470573425, + "learning_rate": 7.353889822552252e-08, + "loss": 0.4258, + "step": 22198 + }, + { + "epoch": 3.952279202279202, + "grad_norm": 0.9729008078575134, + "learning_rate": 7.300316267403462e-08, + "loss": 0.7523, + "step": 22199 + }, + { + "epoch": 3.952457264957265, + "grad_norm": 0.9167136549949646, + "learning_rate": 7.246938498446377e-08, + "loss": 0.6221, + "step": 22200 + }, + { + "epoch": 3.9526353276353277, + "grad_norm": 0.9073047041893005, + "learning_rate": 7.193756516727935e-08, + "loss": 0.6346, + "step": 22201 + }, + { + "epoch": 3.9528133903133904, + "grad_norm": 1.0831269025802612, + "learning_rate": 7.140770323289525e-08, + "loss": 0.7293, + "step": 22202 + }, + { + "epoch": 3.952991452991453, + "grad_norm": 0.8954500555992126, + "learning_rate": 7.087979919169207e-08, + "loss": 0.6734, + "step": 22203 + }, + { + "epoch": 3.953169515669516, + "grad_norm": 1.0015569925308228, + "learning_rate": 7.035385305400599e-08, + "loss": 0.8681, + "step": 22204 + }, + { + "epoch": 3.953347578347578, + "grad_norm": 0.9596917033195496, + "learning_rate": 6.982986483016207e-08, + "loss": 0.8364, + "step": 22205 + }, + { + "epoch": 3.953525641025641, + "grad_norm": 1.0264099836349487, + "learning_rate": 6.930783453040767e-08, + "loss": 0.7442, + "step": 22206 + }, + { + "epoch": 3.9537037037037037, + "grad_norm": 0.8573072552680969, + "learning_rate": 6.878776216499016e-08, + "loss": 0.6549, + "step": 22207 + }, + { + "epoch": 3.9538817663817665, + "grad_norm": 0.9661825895309448, + "learning_rate": 6.826964774407918e-08, + "loss": 0.7778, + "step": 22208 + }, + { + "epoch": 3.9540598290598292, + "grad_norm": 0.8301908373832703, + "learning_rate": 6.775349127783326e-08, + "loss": 0.6448, + "step": 22209 + }, + { + "epoch": 3.9542378917378915, + "grad_norm": 0.8918058276176453, + "learning_rate": 6.723929277636653e-08, + "loss": 0.6934, + "step": 22210 + }, + { + "epoch": 3.9544159544159543, + "grad_norm": 1.1086747646331787, + "learning_rate": 6.672705224974874e-08, + "loss": 0.9753, + "step": 22211 + }, + { + "epoch": 3.954594017094017, + "grad_norm": 1.0281782150268555, + "learning_rate": 6.621676970802738e-08, + "loss": 0.6285, + "step": 22212 + }, + { + "epoch": 3.95477207977208, + "grad_norm": 0.9020591378211975, + "learning_rate": 6.570844516119445e-08, + "loss": 0.5609, + "step": 22213 + }, + { + "epoch": 3.9549501424501425, + "grad_norm": 0.9565960764884949, + "learning_rate": 6.520207861920869e-08, + "loss": 0.7437, + "step": 22214 + }, + { + "epoch": 3.9551282051282053, + "grad_norm": 1.183899998664856, + "learning_rate": 6.469767009198436e-08, + "loss": 0.798, + "step": 22215 + }, + { + "epoch": 3.955306267806268, + "grad_norm": 1.0373740196228027, + "learning_rate": 6.419521958942465e-08, + "loss": 0.9208, + "step": 22216 + }, + { + "epoch": 3.9554843304843303, + "grad_norm": 0.891069769859314, + "learning_rate": 6.369472712135505e-08, + "loss": 0.5366, + "step": 22217 + }, + { + "epoch": 3.955662393162393, + "grad_norm": 0.8725003004074097, + "learning_rate": 6.319619269757881e-08, + "loss": 0.6358, + "step": 22218 + }, + { + "epoch": 3.955840455840456, + "grad_norm": 0.9663777947425842, + "learning_rate": 6.269961632788812e-08, + "loss": 0.8058, + "step": 22219 + }, + { + "epoch": 3.9560185185185186, + "grad_norm": 1.0360321998596191, + "learning_rate": 6.220499802198631e-08, + "loss": 0.7722, + "step": 22220 + }, + { + "epoch": 3.9561965811965814, + "grad_norm": 1.0019513368606567, + "learning_rate": 6.171233778957675e-08, + "loss": 0.6321, + "step": 22221 + }, + { + "epoch": 3.9563746438746437, + "grad_norm": 0.9397497177124023, + "learning_rate": 6.122163564030725e-08, + "loss": 0.865, + "step": 22222 + }, + { + "epoch": 3.9565527065527064, + "grad_norm": 0.9361780285835266, + "learning_rate": 6.073289158380346e-08, + "loss": 0.707, + "step": 22223 + }, + { + "epoch": 3.956730769230769, + "grad_norm": 1.0211809873580933, + "learning_rate": 6.024610562962441e-08, + "loss": 0.8148, + "step": 22224 + }, + { + "epoch": 3.956908831908832, + "grad_norm": 0.970639705657959, + "learning_rate": 5.9761277787318e-08, + "loss": 0.7457, + "step": 22225 + }, + { + "epoch": 3.9570868945868947, + "grad_norm": 0.9334878325462341, + "learning_rate": 5.927840806638774e-08, + "loss": 0.6136, + "step": 22226 + }, + { + "epoch": 3.9572649572649574, + "grad_norm": 1.0210156440734863, + "learning_rate": 5.879749647628163e-08, + "loss": 0.6896, + "step": 22227 + }, + { + "epoch": 3.95744301994302, + "grad_norm": 0.9676822423934937, + "learning_rate": 5.8318543026425473e-08, + "loss": 0.6779, + "step": 22228 + }, + { + "epoch": 3.9576210826210825, + "grad_norm": 0.8727642297744751, + "learning_rate": 5.784154772621175e-08, + "loss": 0.6737, + "step": 22229 + }, + { + "epoch": 3.9577991452991452, + "grad_norm": 0.9030994772911072, + "learning_rate": 5.7366510584988544e-08, + "loss": 0.7467, + "step": 22230 + }, + { + "epoch": 3.957977207977208, + "grad_norm": 0.9664105176925659, + "learning_rate": 5.689343161204841e-08, + "loss": 0.8499, + "step": 22231 + }, + { + "epoch": 3.9581552706552707, + "grad_norm": 1.0056463479995728, + "learning_rate": 5.6422310816661714e-08, + "loss": 0.8082, + "step": 22232 + }, + { + "epoch": 3.9583333333333335, + "grad_norm": 0.9833633899688721, + "learning_rate": 5.595314820807662e-08, + "loss": 0.7669, + "step": 22233 + }, + { + "epoch": 3.958511396011396, + "grad_norm": 0.9480645060539246, + "learning_rate": 5.5485943795463566e-08, + "loss": 0.8063, + "step": 22234 + }, + { + "epoch": 3.9586894586894585, + "grad_norm": 0.9250748753547668, + "learning_rate": 5.5020697587993e-08, + "loss": 0.7095, + "step": 22235 + }, + { + "epoch": 3.9588675213675213, + "grad_norm": 0.9149683117866516, + "learning_rate": 5.455740959476874e-08, + "loss": 0.6697, + "step": 22236 + }, + { + "epoch": 3.959045584045584, + "grad_norm": 0.7881235480308533, + "learning_rate": 5.4096079824872414e-08, + "loss": 0.5722, + "step": 22237 + }, + { + "epoch": 3.959223646723647, + "grad_norm": 0.8992252945899963, + "learning_rate": 5.3636708287352346e-08, + "loss": 0.649, + "step": 22238 + }, + { + "epoch": 3.9594017094017095, + "grad_norm": 0.8622689247131348, + "learning_rate": 5.317929499119023e-08, + "loss": 0.6962, + "step": 22239 + }, + { + "epoch": 3.9595797720797723, + "grad_norm": 0.9292128086090088, + "learning_rate": 5.272383994536778e-08, + "loss": 0.8515, + "step": 22240 + }, + { + "epoch": 3.9597578347578346, + "grad_norm": 1.0485222339630127, + "learning_rate": 5.227034315880008e-08, + "loss": 0.7598, + "step": 22241 + }, + { + "epoch": 3.9599358974358974, + "grad_norm": 0.9419103860855103, + "learning_rate": 5.1818804640368925e-08, + "loss": 0.9492, + "step": 22242 + }, + { + "epoch": 3.96011396011396, + "grad_norm": 0.9244880676269531, + "learning_rate": 5.1369224398911675e-08, + "loss": 0.6919, + "step": 22243 + }, + { + "epoch": 3.960292022792023, + "grad_norm": 0.9493283033370972, + "learning_rate": 5.092160244326571e-08, + "loss": 0.8251, + "step": 22244 + }, + { + "epoch": 3.9604700854700856, + "grad_norm": 0.8471837043762207, + "learning_rate": 5.04759387821796e-08, + "loss": 0.5627, + "step": 22245 + }, + { + "epoch": 3.960648148148148, + "grad_norm": 1.041853904724121, + "learning_rate": 5.003223342439078e-08, + "loss": 0.8932, + "step": 22246 + }, + { + "epoch": 3.9608262108262107, + "grad_norm": 0.8454943895339966, + "learning_rate": 4.959048637859231e-08, + "loss": 0.6321, + "step": 22247 + }, + { + "epoch": 3.9610042735042734, + "grad_norm": 0.9939231872558594, + "learning_rate": 4.9150697653432834e-08, + "loss": 0.9207, + "step": 22248 + }, + { + "epoch": 3.961182336182336, + "grad_norm": 0.8977824449539185, + "learning_rate": 4.871286725753876e-08, + "loss": 0.6579, + "step": 22249 + }, + { + "epoch": 3.961360398860399, + "grad_norm": 0.9290462732315063, + "learning_rate": 4.827699519949214e-08, + "loss": 0.7353, + "step": 22250 + }, + { + "epoch": 3.9615384615384617, + "grad_norm": 0.9763110280036926, + "learning_rate": 4.7843081487819466e-08, + "loss": 0.742, + "step": 22251 + }, + { + "epoch": 3.9617165242165244, + "grad_norm": 0.9784603714942932, + "learning_rate": 4.741112613102505e-08, + "loss": 0.7809, + "step": 22252 + }, + { + "epoch": 3.9618945868945867, + "grad_norm": 0.8668375611305237, + "learning_rate": 4.69811291375799e-08, + "loss": 0.7355, + "step": 22253 + }, + { + "epoch": 3.9620726495726495, + "grad_norm": 0.9601633548736572, + "learning_rate": 4.655309051591061e-08, + "loss": 0.7645, + "step": 22254 + }, + { + "epoch": 3.9622507122507122, + "grad_norm": 0.9448567628860474, + "learning_rate": 4.6127010274399364e-08, + "loss": 0.9177, + "step": 22255 + }, + { + "epoch": 3.962428774928775, + "grad_norm": 0.9351767301559448, + "learning_rate": 4.570288842138393e-08, + "loss": 0.8333, + "step": 22256 + }, + { + "epoch": 3.9626068376068377, + "grad_norm": 0.9702423810958862, + "learning_rate": 4.528072496519098e-08, + "loss": 0.6999, + "step": 22257 + }, + { + "epoch": 3.9627849002849, + "grad_norm": 1.0096203088760376, + "learning_rate": 4.486051991408058e-08, + "loss": 0.8767, + "step": 22258 + }, + { + "epoch": 3.962962962962963, + "grad_norm": 0.9545206427574158, + "learning_rate": 4.444227327629058e-08, + "loss": 0.867, + "step": 22259 + }, + { + "epoch": 3.9631410256410255, + "grad_norm": 1.0196224451065063, + "learning_rate": 4.402598506001443e-08, + "loss": 0.9208, + "step": 22260 + }, + { + "epoch": 3.9633190883190883, + "grad_norm": 0.9041821360588074, + "learning_rate": 4.3611655273401165e-08, + "loss": 0.7637, + "step": 22261 + }, + { + "epoch": 3.963497150997151, + "grad_norm": 1.0449297428131104, + "learning_rate": 4.3199283924588716e-08, + "loss": 0.7901, + "step": 22262 + }, + { + "epoch": 3.963675213675214, + "grad_norm": 1.034912109375, + "learning_rate": 4.278887102163731e-08, + "loss": 0.8515, + "step": 22263 + }, + { + "epoch": 3.9638532763532766, + "grad_norm": 0.9550645351409912, + "learning_rate": 4.238041657259606e-08, + "loss": 0.6318, + "step": 22264 + }, + { + "epoch": 3.9640313390313393, + "grad_norm": 1.076509952545166, + "learning_rate": 4.197392058545857e-08, + "loss": 0.6801, + "step": 22265 + }, + { + "epoch": 3.9642094017094016, + "grad_norm": 1.0079200267791748, + "learning_rate": 4.156938306820735e-08, + "loss": 0.6512, + "step": 22266 + }, + { + "epoch": 3.9643874643874644, + "grad_norm": 0.8909665942192078, + "learning_rate": 4.1166804028758275e-08, + "loss": 0.7712, + "step": 22267 + }, + { + "epoch": 3.964565527065527, + "grad_norm": 0.9579968452453613, + "learning_rate": 4.0766183474993946e-08, + "loss": 0.6829, + "step": 22268 + }, + { + "epoch": 3.96474358974359, + "grad_norm": 0.9043282866477966, + "learning_rate": 4.0367521414774734e-08, + "loss": 0.7346, + "step": 22269 + }, + { + "epoch": 3.964921652421652, + "grad_norm": 1.1984134912490845, + "learning_rate": 3.9970817855905504e-08, + "loss": 0.6943, + "step": 22270 + }, + { + "epoch": 3.965099715099715, + "grad_norm": 0.8458036780357361, + "learning_rate": 3.9576072806146726e-08, + "loss": 0.6403, + "step": 22271 + }, + { + "epoch": 3.9652777777777777, + "grad_norm": 0.9842307567596436, + "learning_rate": 3.9183286273258843e-08, + "loss": 0.7469, + "step": 22272 + }, + { + "epoch": 3.9654558404558404, + "grad_norm": 1.2443668842315674, + "learning_rate": 3.879245826492462e-08, + "loss": 0.8978, + "step": 22273 + }, + { + "epoch": 3.965633903133903, + "grad_norm": 0.9370720982551575, + "learning_rate": 3.840358878879347e-08, + "loss": 0.6538, + "step": 22274 + }, + { + "epoch": 3.965811965811966, + "grad_norm": 0.9503483176231384, + "learning_rate": 3.801667785249263e-08, + "loss": 0.7674, + "step": 22275 + }, + { + "epoch": 3.9659900284900287, + "grad_norm": 0.8143479824066162, + "learning_rate": 3.763172546360494e-08, + "loss": 0.6565, + "step": 22276 + }, + { + "epoch": 3.9661680911680914, + "grad_norm": 0.8467413783073425, + "learning_rate": 3.72487316296688e-08, + "loss": 0.6532, + "step": 22277 + }, + { + "epoch": 3.9663461538461537, + "grad_norm": 0.9195687770843506, + "learning_rate": 3.686769635818932e-08, + "loss": 0.8072, + "step": 22278 + }, + { + "epoch": 3.9665242165242165, + "grad_norm": 0.9875113368034363, + "learning_rate": 3.64886196566272e-08, + "loss": 0.7161, + "step": 22279 + }, + { + "epoch": 3.9667022792022792, + "grad_norm": 1.1833739280700684, + "learning_rate": 3.611150153242093e-08, + "loss": 0.885, + "step": 22280 + }, + { + "epoch": 3.966880341880342, + "grad_norm": 1.0315101146697998, + "learning_rate": 3.5736341992953506e-08, + "loss": 0.7071, + "step": 22281 + }, + { + "epoch": 3.9670584045584043, + "grad_norm": 0.8316394090652466, + "learning_rate": 3.536314104556348e-08, + "loss": 0.5731, + "step": 22282 + }, + { + "epoch": 3.967236467236467, + "grad_norm": 0.9123276472091675, + "learning_rate": 3.4991898697589456e-08, + "loss": 0.6967, + "step": 22283 + }, + { + "epoch": 3.96741452991453, + "grad_norm": 1.0520005226135254, + "learning_rate": 3.462261495628116e-08, + "loss": 0.8838, + "step": 22284 + }, + { + "epoch": 3.9675925925925926, + "grad_norm": 0.9927443265914917, + "learning_rate": 3.4255289828877267e-08, + "loss": 0.7186, + "step": 22285 + }, + { + "epoch": 3.9677706552706553, + "grad_norm": 0.9712579250335693, + "learning_rate": 3.388992332259422e-08, + "loss": 0.8944, + "step": 22286 + }, + { + "epoch": 3.967948717948718, + "grad_norm": 0.897274374961853, + "learning_rate": 3.352651544457075e-08, + "loss": 0.6558, + "step": 22287 + }, + { + "epoch": 3.968126780626781, + "grad_norm": 0.9379400610923767, + "learning_rate": 3.316506620192339e-08, + "loss": 0.9769, + "step": 22288 + }, + { + "epoch": 3.9683048433048436, + "grad_norm": 0.8776909708976746, + "learning_rate": 3.2805575601757567e-08, + "loss": 0.6874, + "step": 22289 + }, + { + "epoch": 3.968482905982906, + "grad_norm": 0.9665780663490295, + "learning_rate": 3.2448043651089885e-08, + "loss": 0.7002, + "step": 22290 + }, + { + "epoch": 3.9686609686609686, + "grad_norm": 0.8973356485366821, + "learning_rate": 3.2092470356948066e-08, + "loss": 0.6631, + "step": 22291 + }, + { + "epoch": 3.9688390313390314, + "grad_norm": 0.9495517611503601, + "learning_rate": 3.17388557262821e-08, + "loss": 0.7328, + "step": 22292 + }, + { + "epoch": 3.969017094017094, + "grad_norm": 0.9756171703338623, + "learning_rate": 3.1387199766030884e-08, + "loss": 0.7874, + "step": 22293 + }, + { + "epoch": 3.9691951566951564, + "grad_norm": 0.9444931745529175, + "learning_rate": 3.10375024830889e-08, + "loss": 0.6176, + "step": 22294 + }, + { + "epoch": 3.969373219373219, + "grad_norm": 0.9644432663917542, + "learning_rate": 3.068976388428402e-08, + "loss": 0.8901, + "step": 22295 + }, + { + "epoch": 3.969551282051282, + "grad_norm": 0.9551946520805359, + "learning_rate": 3.0343983976455216e-08, + "loss": 0.6592, + "step": 22296 + }, + { + "epoch": 3.9697293447293447, + "grad_norm": 0.9582430124282837, + "learning_rate": 3.000016276636375e-08, + "loss": 0.8671, + "step": 22297 + }, + { + "epoch": 3.9699074074074074, + "grad_norm": 0.972446084022522, + "learning_rate": 2.9658300260748673e-08, + "loss": 0.8244, + "step": 22298 + }, + { + "epoch": 3.97008547008547, + "grad_norm": 1.0726343393325806, + "learning_rate": 2.9318396466304633e-08, + "loss": 0.7543, + "step": 22299 + }, + { + "epoch": 3.970263532763533, + "grad_norm": 0.9716382622718811, + "learning_rate": 2.8980451389704067e-08, + "loss": 0.7246, + "step": 22300 + }, + { + "epoch": 3.9704415954415957, + "grad_norm": 0.9425173401832581, + "learning_rate": 2.8644465037552803e-08, + "loss": 0.7026, + "step": 22301 + }, + { + "epoch": 3.970619658119658, + "grad_norm": 1.0111706256866455, + "learning_rate": 2.831043741644557e-08, + "loss": 0.6782, + "step": 22302 + }, + { + "epoch": 3.9707977207977208, + "grad_norm": 0.9812050461769104, + "learning_rate": 2.7978368532921574e-08, + "loss": 0.7114, + "step": 22303 + }, + { + "epoch": 3.9709757834757835, + "grad_norm": 1.0532653331756592, + "learning_rate": 2.7648258393486727e-08, + "loss": 0.8216, + "step": 22304 + }, + { + "epoch": 3.9711538461538463, + "grad_norm": 1.0057753324508667, + "learning_rate": 2.7320107004613626e-08, + "loss": 0.6944, + "step": 22305 + }, + { + "epoch": 3.9713319088319086, + "grad_norm": 1.0849117040634155, + "learning_rate": 2.6993914372719364e-08, + "loss": 0.7063, + "step": 22306 + }, + { + "epoch": 3.9715099715099713, + "grad_norm": 1.084551215171814, + "learning_rate": 2.6669680504209925e-08, + "loss": 0.7596, + "step": 22307 + }, + { + "epoch": 3.971688034188034, + "grad_norm": 1.20219886302948, + "learning_rate": 2.634740540543579e-08, + "loss": 0.8109, + "step": 22308 + }, + { + "epoch": 3.971866096866097, + "grad_norm": 0.9706676006317139, + "learning_rate": 2.6027089082691913e-08, + "loss": 0.6832, + "step": 22309 + }, + { + "epoch": 3.9720441595441596, + "grad_norm": 0.899502158164978, + "learning_rate": 2.570873154228437e-08, + "loss": 0.7876, + "step": 22310 + }, + { + "epoch": 3.9722222222222223, + "grad_norm": 0.9696589112281799, + "learning_rate": 2.5392332790430407e-08, + "loss": 0.8242, + "step": 22311 + }, + { + "epoch": 3.972400284900285, + "grad_norm": 0.98177170753479, + "learning_rate": 2.507789283332507e-08, + "loss": 0.7454, + "step": 22312 + }, + { + "epoch": 3.972578347578348, + "grad_norm": 0.8853342533111572, + "learning_rate": 2.4765411677152296e-08, + "loss": 0.8819, + "step": 22313 + }, + { + "epoch": 3.97275641025641, + "grad_norm": 1.0389357805252075, + "learning_rate": 2.4454889328018317e-08, + "loss": 0.7136, + "step": 22314 + }, + { + "epoch": 3.972934472934473, + "grad_norm": 0.9477161765098572, + "learning_rate": 2.414632579200715e-08, + "loss": 0.7858, + "step": 22315 + }, + { + "epoch": 3.9731125356125356, + "grad_norm": 1.163924217224121, + "learning_rate": 2.3839721075158415e-08, + "loss": 0.8402, + "step": 22316 + }, + { + "epoch": 3.9732905982905984, + "grad_norm": 0.8560773730278015, + "learning_rate": 2.353507518350062e-08, + "loss": 0.6802, + "step": 22317 + }, + { + "epoch": 3.9734686609686607, + "grad_norm": 0.8994771242141724, + "learning_rate": 2.3232388122984562e-08, + "loss": 0.7222, + "step": 22318 + }, + { + "epoch": 3.9736467236467234, + "grad_norm": 0.9608843326568604, + "learning_rate": 2.293165989954993e-08, + "loss": 0.6882, + "step": 22319 + }, + { + "epoch": 3.973824786324786, + "grad_norm": 0.8943131566047668, + "learning_rate": 2.2632890519080907e-08, + "loss": 0.6746, + "step": 22320 + }, + { + "epoch": 3.974002849002849, + "grad_norm": 0.875074565410614, + "learning_rate": 2.233607998742837e-08, + "loss": 0.5955, + "step": 22321 + }, + { + "epoch": 3.9741809116809117, + "grad_norm": 0.922785758972168, + "learning_rate": 2.204122831042099e-08, + "loss": 0.6879, + "step": 22322 + }, + { + "epoch": 3.9743589743589745, + "grad_norm": 1.0186944007873535, + "learning_rate": 2.174833549383193e-08, + "loss": 0.7392, + "step": 22323 + }, + { + "epoch": 3.974537037037037, + "grad_norm": 0.9766637682914734, + "learning_rate": 2.1457401543401034e-08, + "loss": 0.838, + "step": 22324 + }, + { + "epoch": 3.9747150997151, + "grad_norm": 1.1547154188156128, + "learning_rate": 2.1168426464823754e-08, + "loss": 1.0063, + "step": 22325 + }, + { + "epoch": 3.9748931623931623, + "grad_norm": 0.9768098592758179, + "learning_rate": 2.0881410263751123e-08, + "loss": 0.8346, + "step": 22326 + }, + { + "epoch": 3.975071225071225, + "grad_norm": 1.011702299118042, + "learning_rate": 2.0596352945834173e-08, + "loss": 0.7964, + "step": 22327 + }, + { + "epoch": 3.9752492877492878, + "grad_norm": 1.0046590566635132, + "learning_rate": 2.0313254516635126e-08, + "loss": 1.094, + "step": 22328 + }, + { + "epoch": 3.9754273504273505, + "grad_norm": 1.0097856521606445, + "learning_rate": 2.0032114981705098e-08, + "loss": 0.6802, + "step": 22329 + }, + { + "epoch": 3.9756054131054133, + "grad_norm": 0.908306896686554, + "learning_rate": 1.975293434656189e-08, + "loss": 0.7077, + "step": 22330 + }, + { + "epoch": 3.9757834757834756, + "grad_norm": 0.918569028377533, + "learning_rate": 1.9475712616667808e-08, + "loss": 0.6482, + "step": 22331 + }, + { + "epoch": 3.9759615384615383, + "grad_norm": 0.978238046169281, + "learning_rate": 1.9200449797451835e-08, + "loss": 0.8541, + "step": 22332 + }, + { + "epoch": 3.976139601139601, + "grad_norm": 1.020745038986206, + "learning_rate": 1.8927145894309662e-08, + "loss": 0.7957, + "step": 22333 + }, + { + "epoch": 3.976317663817664, + "grad_norm": 0.9351560473442078, + "learning_rate": 1.865580091260366e-08, + "loss": 0.7603, + "step": 22334 + }, + { + "epoch": 3.9764957264957266, + "grad_norm": 1.0288119316101074, + "learning_rate": 1.8386414857640698e-08, + "loss": 0.6967, + "step": 22335 + }, + { + "epoch": 3.9766737891737893, + "grad_norm": 0.9172123074531555, + "learning_rate": 1.8118987734694337e-08, + "loss": 0.6197, + "step": 22336 + }, + { + "epoch": 3.976851851851852, + "grad_norm": 0.9761427044868469, + "learning_rate": 1.785351954902703e-08, + "loss": 0.9306, + "step": 22337 + }, + { + "epoch": 3.9770299145299144, + "grad_norm": 1.0278958082199097, + "learning_rate": 1.7590010305812422e-08, + "loss": 0.7256, + "step": 22338 + }, + { + "epoch": 3.977207977207977, + "grad_norm": 0.8604381680488586, + "learning_rate": 1.7328460010235247e-08, + "loss": 0.696, + "step": 22339 + }, + { + "epoch": 3.97738603988604, + "grad_norm": 0.9066053628921509, + "learning_rate": 1.7068868667402538e-08, + "loss": 0.7578, + "step": 22340 + }, + { + "epoch": 3.9775641025641026, + "grad_norm": 1.0091204643249512, + "learning_rate": 1.6811236282421315e-08, + "loss": 0.7656, + "step": 22341 + }, + { + "epoch": 3.9777421652421654, + "grad_norm": 0.9811164140701294, + "learning_rate": 1.6555562860320895e-08, + "loss": 0.8741, + "step": 22342 + }, + { + "epoch": 3.9779202279202277, + "grad_norm": 1.1655586957931519, + "learning_rate": 1.630184840610838e-08, + "loss": 0.8747, + "step": 22343 + }, + { + "epoch": 3.9780982905982905, + "grad_norm": 0.9563546180725098, + "learning_rate": 1.6050092924768666e-08, + "loss": 0.7394, + "step": 22344 + }, + { + "epoch": 3.978276353276353, + "grad_norm": 0.9154346585273743, + "learning_rate": 1.5800296421231153e-08, + "loss": 0.6454, + "step": 22345 + }, + { + "epoch": 3.978454415954416, + "grad_norm": 0.9271607398986816, + "learning_rate": 1.555245890038082e-08, + "loss": 0.8219, + "step": 22346 + }, + { + "epoch": 3.9786324786324787, + "grad_norm": 0.8896580934524536, + "learning_rate": 1.5306580367091537e-08, + "loss": 0.5634, + "step": 22347 + }, + { + "epoch": 3.9788105413105415, + "grad_norm": 0.8249130249023438, + "learning_rate": 1.506266082615948e-08, + "loss": 0.6459, + "step": 22348 + }, + { + "epoch": 3.978988603988604, + "grad_norm": 0.9651779532432556, + "learning_rate": 1.4820700282380806e-08, + "loss": 0.9841, + "step": 22349 + }, + { + "epoch": 3.9791666666666665, + "grad_norm": 1.0503604412078857, + "learning_rate": 1.4580698740485066e-08, + "loss": 0.8412, + "step": 22350 + }, + { + "epoch": 3.9793447293447293, + "grad_norm": 0.9394381046295166, + "learning_rate": 1.4342656205179606e-08, + "loss": 0.6783, + "step": 22351 + }, + { + "epoch": 3.979522792022792, + "grad_norm": 1.0114195346832275, + "learning_rate": 1.4106572681127361e-08, + "loss": 0.8506, + "step": 22352 + }, + { + "epoch": 3.9797008547008548, + "grad_norm": 0.908707320690155, + "learning_rate": 1.3872448172957964e-08, + "loss": 0.6999, + "step": 22353 + }, + { + "epoch": 3.9798789173789175, + "grad_norm": 1.007361888885498, + "learning_rate": 1.3640282685256633e-08, + "loss": 0.7783, + "step": 22354 + }, + { + "epoch": 3.98005698005698, + "grad_norm": 1.1280618906021118, + "learning_rate": 1.3410076222564184e-08, + "loss": 0.8988, + "step": 22355 + }, + { + "epoch": 3.9802350427350426, + "grad_norm": 0.9843003153800964, + "learning_rate": 1.3181828789388118e-08, + "loss": 0.8007, + "step": 22356 + }, + { + "epoch": 3.9804131054131053, + "grad_norm": 1.0101714134216309, + "learning_rate": 1.295554039022484e-08, + "loss": 0.7549, + "step": 22357 + }, + { + "epoch": 3.980591168091168, + "grad_norm": 0.9103354215621948, + "learning_rate": 1.273121102949304e-08, + "loss": 0.7031, + "step": 22358 + }, + { + "epoch": 3.980769230769231, + "grad_norm": 0.8771821856498718, + "learning_rate": 1.2508840711578095e-08, + "loss": 0.7576, + "step": 22359 + }, + { + "epoch": 3.9809472934472936, + "grad_norm": 0.9804820418357849, + "learning_rate": 1.2288429440854288e-08, + "loss": 0.7814, + "step": 22360 + }, + { + "epoch": 3.9811253561253563, + "grad_norm": 0.8193320035934448, + "learning_rate": 1.2069977221618178e-08, + "loss": 0.6688, + "step": 22361 + }, + { + "epoch": 3.9813034188034186, + "grad_norm": 0.9540324211120605, + "learning_rate": 1.1853484058177433e-08, + "loss": 0.9382, + "step": 22362 + }, + { + "epoch": 3.9814814814814814, + "grad_norm": 0.850046694278717, + "learning_rate": 1.1638949954762002e-08, + "loss": 0.6782, + "step": 22363 + }, + { + "epoch": 3.981659544159544, + "grad_norm": 0.8397730588912964, + "learning_rate": 1.1426374915568528e-08, + "loss": 0.689, + "step": 22364 + }, + { + "epoch": 3.981837606837607, + "grad_norm": 0.9481822848320007, + "learning_rate": 1.1215758944760346e-08, + "loss": 0.8216, + "step": 22365 + }, + { + "epoch": 3.9820156695156697, + "grad_norm": 0.882338285446167, + "learning_rate": 1.100710204647859e-08, + "loss": 0.7705, + "step": 22366 + }, + { + "epoch": 3.982193732193732, + "grad_norm": 0.9093656539916992, + "learning_rate": 1.0800404224797778e-08, + "loss": 0.5796, + "step": 22367 + }, + { + "epoch": 3.9823717948717947, + "grad_norm": 1.0681262016296387, + "learning_rate": 1.0595665483781324e-08, + "loss": 0.7811, + "step": 22368 + }, + { + "epoch": 3.9825498575498575, + "grad_norm": 1.0023845434188843, + "learning_rate": 1.0392885827426036e-08, + "loss": 0.565, + "step": 22369 + }, + { + "epoch": 3.98272792022792, + "grad_norm": 0.9780094027519226, + "learning_rate": 1.019206525971761e-08, + "loss": 0.8498, + "step": 22370 + }, + { + "epoch": 3.982905982905983, + "grad_norm": 1.0917373895645142, + "learning_rate": 9.993203784586236e-09, + "loss": 0.6962, + "step": 22371 + }, + { + "epoch": 3.9830840455840457, + "grad_norm": 1.0024151802062988, + "learning_rate": 9.796301405917695e-09, + "loss": 0.7992, + "step": 22372 + }, + { + "epoch": 3.9832621082621085, + "grad_norm": 1.0047333240509033, + "learning_rate": 9.601358127586668e-09, + "loss": 0.8248, + "step": 22373 + }, + { + "epoch": 3.9834401709401708, + "grad_norm": 0.8688977956771851, + "learning_rate": 9.408373953401218e-09, + "loss": 0.6448, + "step": 22374 + }, + { + "epoch": 3.9836182336182335, + "grad_norm": 1.1232283115386963, + "learning_rate": 9.217348887147204e-09, + "loss": 0.7539, + "step": 22375 + }, + { + "epoch": 3.9837962962962963, + "grad_norm": 0.8725656270980835, + "learning_rate": 9.028282932566078e-09, + "loss": 0.7398, + "step": 22376 + }, + { + "epoch": 3.983974358974359, + "grad_norm": 0.9702167510986328, + "learning_rate": 8.841176093365988e-09, + "loss": 0.7629, + "step": 22377 + }, + { + "epoch": 3.984152421652422, + "grad_norm": 0.9710015058517456, + "learning_rate": 8.656028373210668e-09, + "loss": 0.9984, + "step": 22378 + }, + { + "epoch": 3.984330484330484, + "grad_norm": 1.1007386445999146, + "learning_rate": 8.472839775719443e-09, + "loss": 0.8551, + "step": 22379 + }, + { + "epoch": 3.984508547008547, + "grad_norm": 0.9243883490562439, + "learning_rate": 8.291610304489439e-09, + "loss": 0.7902, + "step": 22380 + }, + { + "epoch": 3.9846866096866096, + "grad_norm": 1.0163289308547974, + "learning_rate": 8.112339963073368e-09, + "loss": 0.755, + "step": 22381 + }, + { + "epoch": 3.9848646723646723, + "grad_norm": 0.9340982437133789, + "learning_rate": 7.935028754979534e-09, + "loss": 0.8079, + "step": 22382 + }, + { + "epoch": 3.985042735042735, + "grad_norm": 0.9406797289848328, + "learning_rate": 7.759676683682937e-09, + "loss": 0.6836, + "step": 22383 + }, + { + "epoch": 3.985220797720798, + "grad_norm": 0.9695764183998108, + "learning_rate": 7.586283752614165e-09, + "loss": 0.6901, + "step": 22384 + }, + { + "epoch": 3.9853988603988606, + "grad_norm": 1.0253673791885376, + "learning_rate": 7.4148499651927054e-09, + "loss": 0.6025, + "step": 22385 + }, + { + "epoch": 3.9855769230769234, + "grad_norm": 0.9360101819038391, + "learning_rate": 7.245375324749226e-09, + "loss": 0.7924, + "step": 22386 + }, + { + "epoch": 3.9857549857549857, + "grad_norm": 1.0131568908691406, + "learning_rate": 7.077859834614397e-09, + "loss": 0.7307, + "step": 22387 + }, + { + "epoch": 3.9859330484330484, + "grad_norm": 1.0636998414993286, + "learning_rate": 6.912303498074479e-09, + "loss": 0.8381, + "step": 22388 + }, + { + "epoch": 3.986111111111111, + "grad_norm": 1.13767671585083, + "learning_rate": 6.748706318371323e-09, + "loss": 0.9167, + "step": 22389 + }, + { + "epoch": 3.986289173789174, + "grad_norm": 0.9454557299613953, + "learning_rate": 6.58706829870237e-09, + "loss": 0.7844, + "step": 22390 + }, + { + "epoch": 3.986467236467236, + "grad_norm": 0.9187841415405273, + "learning_rate": 6.427389442242859e-09, + "loss": 0.6001, + "step": 22391 + }, + { + "epoch": 3.986645299145299, + "grad_norm": 1.2162847518920898, + "learning_rate": 6.269669752123619e-09, + "loss": 0.7899, + "step": 22392 + }, + { + "epoch": 3.9868233618233617, + "grad_norm": 0.8310156464576721, + "learning_rate": 6.11390923143107e-09, + "loss": 0.5653, + "step": 22393 + }, + { + "epoch": 3.9870014245014245, + "grad_norm": 1.0062121152877808, + "learning_rate": 5.960107883218324e-09, + "loss": 0.7532, + "step": 22394 + }, + { + "epoch": 3.9871794871794872, + "grad_norm": 0.9466153383255005, + "learning_rate": 5.808265710494087e-09, + "loss": 0.8552, + "step": 22395 + }, + { + "epoch": 3.98735754985755, + "grad_norm": 0.9300433397293091, + "learning_rate": 5.658382716244859e-09, + "loss": 0.6727, + "step": 22396 + }, + { + "epoch": 3.9875356125356127, + "grad_norm": 1.028559684753418, + "learning_rate": 5.510458903390525e-09, + "loss": 0.7283, + "step": 22397 + }, + { + "epoch": 3.9877136752136755, + "grad_norm": 1.0546510219573975, + "learning_rate": 5.364494274839871e-09, + "loss": 0.5589, + "step": 22398 + }, + { + "epoch": 3.987891737891738, + "grad_norm": 0.8639279007911682, + "learning_rate": 5.220488833457271e-09, + "loss": 0.7989, + "step": 22399 + }, + { + "epoch": 3.9880698005698005, + "grad_norm": 0.9834537506103516, + "learning_rate": 5.07844258205159e-09, + "loss": 0.694, + "step": 22400 + }, + { + "epoch": 3.9882478632478633, + "grad_norm": 0.9686489701271057, + "learning_rate": 4.93835552342059e-09, + "loss": 0.8029, + "step": 22401 + }, + { + "epoch": 3.988425925925926, + "grad_norm": 0.9237646460533142, + "learning_rate": 4.80022766029542e-09, + "loss": 0.7817, + "step": 22402 + }, + { + "epoch": 3.9886039886039883, + "grad_norm": 1.0533924102783203, + "learning_rate": 4.664058995385023e-09, + "loss": 0.5459, + "step": 22403 + }, + { + "epoch": 3.988782051282051, + "grad_norm": 1.1050822734832764, + "learning_rate": 4.529849531353936e-09, + "loss": 0.7622, + "step": 22404 + }, + { + "epoch": 3.988960113960114, + "grad_norm": 0.8912914395332336, + "learning_rate": 4.39759927085559e-09, + "loss": 0.7879, + "step": 22405 + }, + { + "epoch": 3.9891381766381766, + "grad_norm": 0.802912175655365, + "learning_rate": 4.2673082164434995e-09, + "loss": 0.559, + "step": 22406 + }, + { + "epoch": 3.9893162393162394, + "grad_norm": 1.0098425149917603, + "learning_rate": 4.1389763707044835e-09, + "loss": 0.809, + "step": 22407 + }, + { + "epoch": 3.989494301994302, + "grad_norm": 0.8691425323486328, + "learning_rate": 4.012603736136544e-09, + "loss": 0.6097, + "step": 22408 + }, + { + "epoch": 3.989672364672365, + "grad_norm": 0.8382730484008789, + "learning_rate": 3.888190315215479e-09, + "loss": 0.6292, + "step": 22409 + }, + { + "epoch": 3.9898504273504276, + "grad_norm": 1.00441312789917, + "learning_rate": 3.7657361103837776e-09, + "loss": 0.8296, + "step": 22410 + }, + { + "epoch": 3.99002849002849, + "grad_norm": 0.9702491164207458, + "learning_rate": 3.645241124039522e-09, + "loss": 0.8129, + "step": 22411 + }, + { + "epoch": 3.9902065527065527, + "grad_norm": 1.0183627605438232, + "learning_rate": 3.5267053585363863e-09, + "loss": 0.9345, + "step": 22412 + }, + { + "epoch": 3.9903846153846154, + "grad_norm": 1.120223045349121, + "learning_rate": 3.4101288162058377e-09, + "loss": 0.9309, + "step": 22413 + }, + { + "epoch": 3.990562678062678, + "grad_norm": 1.0526723861694336, + "learning_rate": 3.2955114993238336e-09, + "loss": 0.7433, + "step": 22414 + }, + { + "epoch": 3.9907407407407405, + "grad_norm": 0.9110050201416016, + "learning_rate": 3.182853410155229e-09, + "loss": 0.6849, + "step": 22415 + }, + { + "epoch": 3.9909188034188032, + "grad_norm": 0.9287052154541016, + "learning_rate": 3.0721545508760606e-09, + "loss": 0.7313, + "step": 22416 + }, + { + "epoch": 3.991096866096866, + "grad_norm": 0.9603120684623718, + "learning_rate": 2.9634149236845708e-09, + "loss": 0.7157, + "step": 22417 + }, + { + "epoch": 3.9912749287749287, + "grad_norm": 0.8682617545127869, + "learning_rate": 2.856634530690183e-09, + "loss": 0.7321, + "step": 22418 + }, + { + "epoch": 3.9914529914529915, + "grad_norm": 0.894698441028595, + "learning_rate": 2.751813374002321e-09, + "loss": 0.7382, + "step": 22419 + }, + { + "epoch": 3.9916310541310542, + "grad_norm": 0.868032693862915, + "learning_rate": 2.648951455663795e-09, + "loss": 0.7712, + "step": 22420 + }, + { + "epoch": 3.991809116809117, + "grad_norm": 0.9529984593391418, + "learning_rate": 2.548048777695211e-09, + "loss": 0.917, + "step": 22421 + }, + { + "epoch": 3.9919871794871797, + "grad_norm": 0.9214971661567688, + "learning_rate": 2.4491053420616637e-09, + "loss": 0.9054, + "step": 22422 + }, + { + "epoch": 3.992165242165242, + "grad_norm": 1.053322434425354, + "learning_rate": 2.352121150717146e-09, + "loss": 0.9304, + "step": 22423 + }, + { + "epoch": 3.992343304843305, + "grad_norm": 0.9296656250953674, + "learning_rate": 2.2570962055601385e-09, + "loss": 0.9732, + "step": 22424 + }, + { + "epoch": 3.9925213675213675, + "grad_norm": 1.047304630279541, + "learning_rate": 2.1640305084447144e-09, + "loss": 0.6625, + "step": 22425 + }, + { + "epoch": 3.9926994301994303, + "grad_norm": 0.9102067947387695, + "learning_rate": 2.072924061191639e-09, + "loss": 0.6529, + "step": 22426 + }, + { + "epoch": 3.9928774928774926, + "grad_norm": 1.0134507417678833, + "learning_rate": 1.983776865599474e-09, + "loss": 0.8424, + "step": 22427 + }, + { + "epoch": 3.9930555555555554, + "grad_norm": 0.8566016554832458, + "learning_rate": 1.8965889234001666e-09, + "loss": 0.5787, + "step": 22428 + }, + { + "epoch": 3.993233618233618, + "grad_norm": 1.1094683408737183, + "learning_rate": 1.8113602363145632e-09, + "loss": 0.8465, + "step": 22429 + }, + { + "epoch": 3.993411680911681, + "grad_norm": 0.9093437790870667, + "learning_rate": 1.728090806007998e-09, + "loss": 0.6979, + "step": 22430 + }, + { + "epoch": 3.9935897435897436, + "grad_norm": 0.9416308999061584, + "learning_rate": 1.6467806341124991e-09, + "loss": 0.8162, + "step": 22431 + }, + { + "epoch": 3.9937678062678064, + "grad_norm": 0.9535760283470154, + "learning_rate": 1.5674297222156852e-09, + "loss": 0.7527, + "step": 22432 + }, + { + "epoch": 3.993945868945869, + "grad_norm": 1.0274043083190918, + "learning_rate": 1.4900380718718688e-09, + "loss": 0.8621, + "step": 22433 + }, + { + "epoch": 3.994123931623932, + "grad_norm": 0.9442184567451477, + "learning_rate": 1.414605684602055e-09, + "loss": 0.4077, + "step": 22434 + }, + { + "epoch": 3.994301994301994, + "grad_norm": 0.9062432050704956, + "learning_rate": 1.341132561893943e-09, + "loss": 0.8457, + "step": 22435 + }, + { + "epoch": 3.994480056980057, + "grad_norm": 1.0084505081176758, + "learning_rate": 1.2696187051686182e-09, + "loss": 0.8652, + "step": 22436 + }, + { + "epoch": 3.9946581196581197, + "grad_norm": 0.8726778030395508, + "learning_rate": 1.200064115847166e-09, + "loss": 0.7724, + "step": 22437 + }, + { + "epoch": 3.9948361823361824, + "grad_norm": 0.927804708480835, + "learning_rate": 1.1324687952729562e-09, + "loss": 0.9643, + "step": 22438 + }, + { + "epoch": 3.995014245014245, + "grad_norm": 0.9145159721374512, + "learning_rate": 1.0668327447782567e-09, + "loss": 0.7232, + "step": 22439 + }, + { + "epoch": 3.9951923076923075, + "grad_norm": 0.8730134963989258, + "learning_rate": 1.0031559656398238e-09, + "loss": 0.6696, + "step": 22440 + }, + { + "epoch": 3.9953703703703702, + "grad_norm": 0.884189248085022, + "learning_rate": 9.414384591233116e-10, + "loss": 0.7042, + "step": 22441 + }, + { + "epoch": 3.995548433048433, + "grad_norm": 1.069779396057129, + "learning_rate": 8.816802264388635e-10, + "loss": 0.7686, + "step": 22442 + }, + { + "epoch": 3.9957264957264957, + "grad_norm": 0.9624055624008179, + "learning_rate": 8.238812687300091e-10, + "loss": 0.7969, + "step": 22443 + }, + { + "epoch": 3.9959045584045585, + "grad_norm": 0.9532380700111389, + "learning_rate": 7.680415871624825e-10, + "loss": 0.7705, + "step": 22444 + }, + { + "epoch": 3.9960826210826212, + "grad_norm": 0.8498616218566895, + "learning_rate": 7.141611828020977e-10, + "loss": 0.6918, + "step": 22445 + }, + { + "epoch": 3.996260683760684, + "grad_norm": 0.897534966468811, + "learning_rate": 6.622400567257714e-10, + "loss": 0.6194, + "step": 22446 + }, + { + "epoch": 3.9964387464387463, + "grad_norm": 0.9699941873550415, + "learning_rate": 6.122782099438063e-10, + "loss": 0.7863, + "step": 22447 + }, + { + "epoch": 3.996616809116809, + "grad_norm": 1.0888839960098267, + "learning_rate": 5.642756434220964e-10, + "loss": 0.8153, + "step": 22448 + }, + { + "epoch": 3.996794871794872, + "grad_norm": 0.9677609801292419, + "learning_rate": 5.182323581265358e-10, + "loss": 0.933, + "step": 22449 + }, + { + "epoch": 3.9969729344729346, + "grad_norm": 1.1706438064575195, + "learning_rate": 4.741483549342008e-10, + "loss": 0.877, + "step": 22450 + }, + { + "epoch": 3.9971509971509973, + "grad_norm": 0.9511678814888, + "learning_rate": 4.3202363472216733e-10, + "loss": 0.6967, + "step": 22451 + }, + { + "epoch": 3.9973290598290596, + "grad_norm": 1.0145454406738281, + "learning_rate": 3.9185819831200067e-10, + "loss": 0.8106, + "step": 22452 + }, + { + "epoch": 3.9975071225071224, + "grad_norm": 0.9555829763412476, + "learning_rate": 3.536520465030613e-10, + "loss": 0.6381, + "step": 22453 + }, + { + "epoch": 3.997685185185185, + "grad_norm": 0.8437405228614807, + "learning_rate": 3.174051800280964e-10, + "loss": 0.5492, + "step": 22454 + }, + { + "epoch": 3.997863247863248, + "grad_norm": 0.965201735496521, + "learning_rate": 2.831175995976487e-10, + "loss": 0.7768, + "step": 22455 + }, + { + "epoch": 3.9980413105413106, + "grad_norm": 1.0718035697937012, + "learning_rate": 2.5078930588895433e-10, + "loss": 0.7831, + "step": 22456 + }, + { + "epoch": 3.9982193732193734, + "grad_norm": 0.937907874584198, + "learning_rate": 2.2042029953484033e-10, + "loss": 0.6929, + "step": 22457 + }, + { + "epoch": 3.998397435897436, + "grad_norm": 1.2005555629730225, + "learning_rate": 1.9201058113482717e-10, + "loss": 0.8616, + "step": 22458 + }, + { + "epoch": 3.9985754985754984, + "grad_norm": 1.012808918952942, + "learning_rate": 1.6556015123292412e-10, + "loss": 0.6396, + "step": 22459 + }, + { + "epoch": 3.998753561253561, + "grad_norm": 0.8904356360435486, + "learning_rate": 1.4106901036203823e-10, + "loss": 0.6905, + "step": 22460 + }, + { + "epoch": 3.998931623931624, + "grad_norm": 0.8998084664344788, + "learning_rate": 1.185371589884632e-10, + "loss": 0.6481, + "step": 22461 + }, + { + "epoch": 3.9991096866096867, + "grad_norm": 1.0224751234054565, + "learning_rate": 9.79645975673904e-11, + "loss": 0.9935, + "step": 22462 + }, + { + "epoch": 3.9992877492877494, + "grad_norm": Infinity, + "learning_rate": 9.79645975673904e-11, + "loss": 0.7249, + "step": 22463 + }, + { + "epoch": 3.9994658119658117, + "grad_norm": 0.9459518790245056, + "learning_rate": 7.935132648739796e-11, + "loss": 0.7522, + "step": 22464 + }, + { + "epoch": 3.9994658119658117, + "eval_loss": 1.177049160003662, + "eval_runtime": 25.8941, + "eval_samples_per_second": 40.202, + "eval_steps_per_second": 20.12, + "step": 22464 + } + ], + "logging_steps": 1, + "max_steps": 22464, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 5616, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8.674333832353382e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-22464/training_args.bin b/checkpoint-22464/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..1245f6a2afbe9a6eefbb6d141231d555e0b0bf84 --- /dev/null +++ b/checkpoint-22464/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86de370014ed2be86ea27c820b434ceec5e097da2b5f9b08d0eac9aa564d8961 +size 6200 diff --git a/checkpoint-5616/README.md b/checkpoint-5616/README.md new file mode 100644 index 0000000000000000000000000000000000000000..719b4726992f7d0707a4253e9123dec35e4de390 --- /dev/null +++ b/checkpoint-5616/README.md @@ -0,0 +1,202 @@ +--- +base_model: openlm-research/open_llama_3b_v2 +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.13.2 \ No newline at end of file diff --git a/checkpoint-5616/adapter_config.json b/checkpoint-5616/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6f20a570fc808390da3f2e001093ac1e56c1da --- /dev/null +++ b/checkpoint-5616/adapter_config.json @@ -0,0 +1,34 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "openlm-research/open_llama_3b_v2", + "bias": "none", + "fan_in_fan_out": null, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.0, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "gate_proj", + "v_proj", + "down_proj", + "o_proj", + "q_proj", + "up_proj", + "k_proj" + ], + "task_type": "CAUSAL_LM", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/checkpoint-5616/adapter_model.safetensors b/checkpoint-5616/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dd5e7b14c97cfc3ef09445a2e53a4b1a3983820d --- /dev/null +++ b/checkpoint-5616/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efd66c77b4c02ea9aa38b1686fca713fdffa16a78794d032ec9c8257cf23a895 +size 50899792 diff --git a/checkpoint-5616/optimizer.pt b/checkpoint-5616/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..b7bc80f428fba1b747af09ee0b8111c8a636bf32 --- /dev/null +++ b/checkpoint-5616/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f8c777c25aeaeaaaa9324690795239160c715d50d22636f64a3a82bb3f06d9b +size 26231684 diff --git a/checkpoint-5616/rng_state.pth b/checkpoint-5616/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..41dfa7d7903dea42d227bad638c2c750928d590c --- /dev/null +++ b/checkpoint-5616/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c062f7f375beded48b5337f5a3f3a5cb38807fa3e85dbf3e294c0ab6b627bfc2 +size 14244 diff --git a/checkpoint-5616/scheduler.pt b/checkpoint-5616/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..90dbb49a7ac95509a5c54bf578be193ee7e1ee9e --- /dev/null +++ b/checkpoint-5616/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4164a586e6ce7816c769ce21463fb39cb8c13a77d5e42f55c5f48dcac354572b +size 1064 diff --git a/checkpoint-5616/special_tokens_map.json b/checkpoint-5616/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/checkpoint-5616/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/checkpoint-5616/tokenizer.model b/checkpoint-5616/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..98866ff8ae3631f331c57923c921a0c9ad22b97d --- /dev/null +++ b/checkpoint-5616/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b289e85fa20fd375d8b33dc12f77616f18abc6359804471d1fafcb425fecb8 +size 511574 diff --git a/checkpoint-5616/tokenizer_config.json b/checkpoint-5616/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/checkpoint-5616/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/checkpoint-5616/trainer_state.json b/checkpoint-5616/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c7e18c77dc0fb255b3d719d75c3bab953eebb8d4 --- /dev/null +++ b/checkpoint-5616/trainer_state.json @@ -0,0 +1,39385 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 1404, + "global_step": 5616, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00017806267806267807, + "grad_norm": 0.2854898273944855, + "learning_rate": 1e-05, + "loss": 1.1997, + "step": 1 + }, + { + "epoch": 0.00017806267806267807, + "eval_loss": 1.3698358535766602, + "eval_runtime": 24.1591, + "eval_samples_per_second": 43.089, + "eval_steps_per_second": 21.565, + "step": 1 + }, + { + "epoch": 0.00035612535612535614, + "grad_norm": 0.3508087396621704, + "learning_rate": 2e-05, + "loss": 1.4134, + "step": 2 + }, + { + "epoch": 0.0005341880341880342, + "grad_norm": 0.27050870656967163, + "learning_rate": 3e-05, + "loss": 1.3447, + "step": 3 + }, + { + "epoch": 0.0007122507122507123, + "grad_norm": 0.27706292271614075, + "learning_rate": 4e-05, + "loss": 1.0354, + "step": 4 + }, + { + "epoch": 0.0008903133903133903, + "grad_norm": 0.30398961901664734, + "learning_rate": 5e-05, + "loss": 1.1441, + "step": 5 + }, + { + "epoch": 0.0010683760683760685, + "grad_norm": 0.3103881776332855, + "learning_rate": 6e-05, + "loss": 1.341, + "step": 6 + }, + { + "epoch": 0.0012464387464387464, + "grad_norm": 0.5191189646720886, + "learning_rate": 7e-05, + "loss": 1.3457, + "step": 7 + }, + { + "epoch": 0.0014245014245014246, + "grad_norm": 0.4449467360973358, + "learning_rate": 8e-05, + "loss": 1.5051, + "step": 8 + }, + { + "epoch": 0.0016025641025641025, + "grad_norm": 0.3914581537246704, + "learning_rate": 9e-05, + "loss": 1.5525, + "step": 9 + }, + { + "epoch": 0.0017806267806267807, + "grad_norm": 0.37746086716651917, + "learning_rate": 0.0001, + "loss": 1.3266, + "step": 10 + }, + { + "epoch": 0.001958689458689459, + "grad_norm": 0.35226109623908997, + "learning_rate": 0.00011000000000000002, + "loss": 1.5416, + "step": 11 + }, + { + "epoch": 0.002136752136752137, + "grad_norm": 0.3343672454357147, + "learning_rate": 0.00012, + "loss": 1.3221, + "step": 12 + }, + { + "epoch": 0.0023148148148148147, + "grad_norm": 0.47298333048820496, + "learning_rate": 0.00013000000000000002, + "loss": 1.2999, + "step": 13 + }, + { + "epoch": 0.002492877492877493, + "grad_norm": 0.377814918756485, + "learning_rate": 0.00014, + "loss": 1.1688, + "step": 14 + }, + { + "epoch": 0.002670940170940171, + "grad_norm": 0.46344801783561707, + "learning_rate": 0.00015000000000000001, + "loss": 1.3565, + "step": 15 + }, + { + "epoch": 0.002849002849002849, + "grad_norm": 0.49615249037742615, + "learning_rate": 0.00016, + "loss": 1.5692, + "step": 16 + }, + { + "epoch": 0.003027065527065527, + "grad_norm": 0.5109946131706238, + "learning_rate": 0.00017, + "loss": 1.2991, + "step": 17 + }, + { + "epoch": 0.003205128205128205, + "grad_norm": 0.5125070214271545, + "learning_rate": 0.00018, + "loss": 1.3309, + "step": 18 + }, + { + "epoch": 0.003383190883190883, + "grad_norm": 0.4517767131328583, + "learning_rate": 0.00019, + "loss": 1.357, + "step": 19 + }, + { + "epoch": 0.0035612535612535613, + "grad_norm": 0.47267794609069824, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 20 + }, + { + "epoch": 0.0037393162393162395, + "grad_norm": 0.46823424100875854, + "learning_rate": 0.00019999999902035388, + "loss": 1.1195, + "step": 21 + }, + { + "epoch": 0.003917378917378918, + "grad_norm": 0.440036803483963, + "learning_rate": 0.00019999999608141548, + "loss": 1.2822, + "step": 22 + }, + { + "epoch": 0.004095441595441595, + "grad_norm": 0.371101975440979, + "learning_rate": 0.00019999999118318492, + "loss": 1.132, + "step": 23 + }, + { + "epoch": 0.004273504273504274, + "grad_norm": 0.44691094756126404, + "learning_rate": 0.00019999998432566226, + "loss": 1.2968, + "step": 24 + }, + { + "epoch": 0.004451566951566952, + "grad_norm": 0.5462725162506104, + "learning_rate": 0.0001999999755088476, + "loss": 1.1714, + "step": 25 + }, + { + "epoch": 0.004629629629629629, + "grad_norm": 0.39860013127326965, + "learning_rate": 0.0001999999647327412, + "loss": 1.0407, + "step": 26 + }, + { + "epoch": 0.004807692307692308, + "grad_norm": 0.5031934380531311, + "learning_rate": 0.0001999999519973432, + "loss": 1.2773, + "step": 27 + }, + { + "epoch": 0.004985754985754986, + "grad_norm": 0.42162764072418213, + "learning_rate": 0.0001999999373026539, + "loss": 1.2824, + "step": 28 + }, + { + "epoch": 0.005163817663817663, + "grad_norm": 0.40964868664741516, + "learning_rate": 0.00019999992064867353, + "loss": 1.226, + "step": 29 + }, + { + "epoch": 0.005341880341880342, + "grad_norm": 0.41650915145874023, + "learning_rate": 0.00019999990203540245, + "loss": 1.2677, + "step": 30 + }, + { + "epoch": 0.00551994301994302, + "grad_norm": 0.40052226185798645, + "learning_rate": 0.00019999988146284103, + "loss": 0.9443, + "step": 31 + }, + { + "epoch": 0.005698005698005698, + "grad_norm": 0.5198387503623962, + "learning_rate": 0.00019999985893098964, + "loss": 1.3043, + "step": 32 + }, + { + "epoch": 0.005876068376068376, + "grad_norm": 0.50941002368927, + "learning_rate": 0.00019999983443984878, + "loss": 1.2002, + "step": 33 + }, + { + "epoch": 0.006054131054131054, + "grad_norm": 0.30082932114601135, + "learning_rate": 0.00019999980798941888, + "loss": 0.9904, + "step": 34 + }, + { + "epoch": 0.006232193732193732, + "grad_norm": 0.4228935241699219, + "learning_rate": 0.00019999977957970048, + "loss": 1.1137, + "step": 35 + }, + { + "epoch": 0.00641025641025641, + "grad_norm": 0.41294750571250916, + "learning_rate": 0.0001999997492106941, + "loss": 1.3385, + "step": 36 + }, + { + "epoch": 0.006588319088319089, + "grad_norm": 0.4415493905544281, + "learning_rate": 0.00019999971688240041, + "loss": 1.1695, + "step": 37 + }, + { + "epoch": 0.006766381766381766, + "grad_norm": 0.3726460933685303, + "learning_rate": 0.00019999968259482, + "loss": 1.1734, + "step": 38 + }, + { + "epoch": 0.006944444444444444, + "grad_norm": 0.3969627320766449, + "learning_rate": 0.0001999996463479535, + "loss": 1.1209, + "step": 39 + }, + { + "epoch": 0.007122507122507123, + "grad_norm": 0.3779667913913727, + "learning_rate": 0.0001999996081418017, + "loss": 1.1635, + "step": 40 + }, + { + "epoch": 0.0073005698005698, + "grad_norm": 0.3933636546134949, + "learning_rate": 0.0001999995679763653, + "loss": 1.1514, + "step": 41 + }, + { + "epoch": 0.007478632478632479, + "grad_norm": 0.3567957282066345, + "learning_rate": 0.00019999952585164507, + "loss": 1.2488, + "step": 42 + }, + { + "epoch": 0.007656695156695157, + "grad_norm": 0.32506081461906433, + "learning_rate": 0.00019999948176764186, + "loss": 1.149, + "step": 43 + }, + { + "epoch": 0.007834757834757835, + "grad_norm": 0.46588361263275146, + "learning_rate": 0.0001999994357243566, + "loss": 1.4263, + "step": 44 + }, + { + "epoch": 0.008012820512820512, + "grad_norm": 0.5070307850837708, + "learning_rate": 0.00019999938772179005, + "loss": 1.0698, + "step": 45 + }, + { + "epoch": 0.00819088319088319, + "grad_norm": 0.38199326395988464, + "learning_rate": 0.00019999933775994327, + "loss": 0.9907, + "step": 46 + }, + { + "epoch": 0.00836894586894587, + "grad_norm": 0.43684661388397217, + "learning_rate": 0.0001999992858388172, + "loss": 1.2905, + "step": 47 + }, + { + "epoch": 0.008547008547008548, + "grad_norm": 0.44482162594795227, + "learning_rate": 0.00019999923195841284, + "loss": 1.2153, + "step": 48 + }, + { + "epoch": 0.008725071225071225, + "grad_norm": 0.4259667694568634, + "learning_rate": 0.0001999991761187313, + "loss": 1.1582, + "step": 49 + }, + { + "epoch": 0.008903133903133903, + "grad_norm": 0.41649091243743896, + "learning_rate": 0.00019999911831977357, + "loss": 1.0185, + "step": 50 + }, + { + "epoch": 0.009081196581196582, + "grad_norm": 0.4179716110229492, + "learning_rate": 0.0001999990585615409, + "loss": 1.3579, + "step": 51 + }, + { + "epoch": 0.009259259259259259, + "grad_norm": 0.3372558355331421, + "learning_rate": 0.00019999899684403438, + "loss": 1.0638, + "step": 52 + }, + { + "epoch": 0.009437321937321937, + "grad_norm": 0.41294020414352417, + "learning_rate": 0.00019999893316725525, + "loss": 1.1932, + "step": 53 + }, + { + "epoch": 0.009615384615384616, + "grad_norm": 0.4407919645309448, + "learning_rate": 0.00019999886753120473, + "loss": 1.4129, + "step": 54 + }, + { + "epoch": 0.009793447293447293, + "grad_norm": 0.47948843240737915, + "learning_rate": 0.00019999879993588414, + "loss": 1.2424, + "step": 55 + }, + { + "epoch": 0.009971509971509971, + "grad_norm": 0.3535355031490326, + "learning_rate": 0.00019999873038129484, + "loss": 1.0145, + "step": 56 + }, + { + "epoch": 0.01014957264957265, + "grad_norm": 0.5067078471183777, + "learning_rate": 0.00019999865886743813, + "loss": 1.4708, + "step": 57 + }, + { + "epoch": 0.010327635327635327, + "grad_norm": 0.42862898111343384, + "learning_rate": 0.0001999985853943154, + "loss": 1.0399, + "step": 58 + }, + { + "epoch": 0.010505698005698005, + "grad_norm": 0.4769059419631958, + "learning_rate": 0.00019999850996192816, + "loss": 1.1258, + "step": 59 + }, + { + "epoch": 0.010683760683760684, + "grad_norm": 0.4065442383289337, + "learning_rate": 0.0001999984325702778, + "loss": 1.2077, + "step": 60 + }, + { + "epoch": 0.010861823361823363, + "grad_norm": 0.5318329930305481, + "learning_rate": 0.0001999983532193659, + "loss": 1.2298, + "step": 61 + }, + { + "epoch": 0.01103988603988604, + "grad_norm": 0.4777173101902008, + "learning_rate": 0.000199998271909194, + "loss": 1.3195, + "step": 62 + }, + { + "epoch": 0.011217948717948718, + "grad_norm": 0.37553808093070984, + "learning_rate": 0.0001999981886397637, + "loss": 1.1188, + "step": 63 + }, + { + "epoch": 0.011396011396011397, + "grad_norm": 0.3920556902885437, + "learning_rate": 0.0001999981034110766, + "loss": 1.1448, + "step": 64 + }, + { + "epoch": 0.011574074074074073, + "grad_norm": 0.454272598028183, + "learning_rate": 0.0001999980162231344, + "loss": 1.0812, + "step": 65 + }, + { + "epoch": 0.011752136752136752, + "grad_norm": 0.4354456663131714, + "learning_rate": 0.00019999792707593882, + "loss": 1.1174, + "step": 66 + }, + { + "epoch": 0.01193019943019943, + "grad_norm": 0.5030252933502197, + "learning_rate": 0.00019999783596949156, + "loss": 1.2925, + "step": 67 + }, + { + "epoch": 0.012108262108262107, + "grad_norm": 0.5141571164131165, + "learning_rate": 0.00019999774290379446, + "loss": 1.6193, + "step": 68 + }, + { + "epoch": 0.012286324786324786, + "grad_norm": 0.417298287153244, + "learning_rate": 0.0001999976478788493, + "loss": 1.1875, + "step": 69 + }, + { + "epoch": 0.012464387464387465, + "grad_norm": 0.4642415940761566, + "learning_rate": 0.00019999755089465795, + "loss": 1.4138, + "step": 70 + }, + { + "epoch": 0.012642450142450143, + "grad_norm": 0.43184754252433777, + "learning_rate": 0.0001999974519512223, + "loss": 1.0697, + "step": 71 + }, + { + "epoch": 0.01282051282051282, + "grad_norm": 0.46698349714279175, + "learning_rate": 0.00019999735104854436, + "loss": 0.709, + "step": 72 + }, + { + "epoch": 0.012998575498575499, + "grad_norm": 0.37253814935684204, + "learning_rate": 0.000199997248186626, + "loss": 1.2084, + "step": 73 + }, + { + "epoch": 0.013176638176638177, + "grad_norm": 0.3851388692855835, + "learning_rate": 0.0001999971433654693, + "loss": 1.0548, + "step": 74 + }, + { + "epoch": 0.013354700854700854, + "grad_norm": 0.4434688985347748, + "learning_rate": 0.00019999703658507635, + "loss": 1.4084, + "step": 75 + }, + { + "epoch": 0.013532763532763533, + "grad_norm": 0.43164482712745667, + "learning_rate": 0.00019999692784544913, + "loss": 1.4872, + "step": 76 + }, + { + "epoch": 0.013710826210826211, + "grad_norm": 0.4224303364753723, + "learning_rate": 0.00019999681714658984, + "loss": 1.2221, + "step": 77 + }, + { + "epoch": 0.013888888888888888, + "grad_norm": 0.35588955879211426, + "learning_rate": 0.00019999670448850069, + "loss": 0.84, + "step": 78 + }, + { + "epoch": 0.014066951566951567, + "grad_norm": 0.3970590829849243, + "learning_rate": 0.0001999965898711838, + "loss": 1.1886, + "step": 79 + }, + { + "epoch": 0.014245014245014245, + "grad_norm": 0.4331924319267273, + "learning_rate": 0.00019999647329464146, + "loss": 1.179, + "step": 80 + }, + { + "epoch": 0.014423076923076924, + "grad_norm": 0.4226946234703064, + "learning_rate": 0.00019999635475887598, + "loss": 1.1496, + "step": 81 + }, + { + "epoch": 0.0146011396011396, + "grad_norm": 0.381592720746994, + "learning_rate": 0.00019999623426388962, + "loss": 1.1774, + "step": 82 + }, + { + "epoch": 0.01477920227920228, + "grad_norm": 0.4190855622291565, + "learning_rate": 0.00019999611180968478, + "loss": 1.1491, + "step": 83 + }, + { + "epoch": 0.014957264957264958, + "grad_norm": 0.3904292583465576, + "learning_rate": 0.00019999598739626389, + "loss": 1.1275, + "step": 84 + }, + { + "epoch": 0.015135327635327635, + "grad_norm": 0.4515478014945984, + "learning_rate": 0.0001999958610236293, + "loss": 1.2404, + "step": 85 + }, + { + "epoch": 0.015313390313390313, + "grad_norm": 0.48341724276542664, + "learning_rate": 0.00019999573269178359, + "loss": 1.3572, + "step": 86 + }, + { + "epoch": 0.015491452991452992, + "grad_norm": 0.42150333523750305, + "learning_rate": 0.00019999560240072914, + "loss": 1.0203, + "step": 87 + }, + { + "epoch": 0.01566951566951567, + "grad_norm": 0.45445525646209717, + "learning_rate": 0.00019999547015046867, + "loss": 1.0677, + "step": 88 + }, + { + "epoch": 0.01584757834757835, + "grad_norm": 0.3581015467643738, + "learning_rate": 0.00019999533594100463, + "loss": 1.0693, + "step": 89 + }, + { + "epoch": 0.016025641025641024, + "grad_norm": 0.4430878758430481, + "learning_rate": 0.00019999519977233971, + "loss": 1.1591, + "step": 90 + }, + { + "epoch": 0.016203703703703703, + "grad_norm": 0.3940352201461792, + "learning_rate": 0.0001999950616444766, + "loss": 1.1325, + "step": 91 + }, + { + "epoch": 0.01638176638176638, + "grad_norm": 0.4521673321723938, + "learning_rate": 0.00019999492155741794, + "loss": 1.3288, + "step": 92 + }, + { + "epoch": 0.01655982905982906, + "grad_norm": 0.3988296687602997, + "learning_rate": 0.00019999477951116658, + "loss": 1.0023, + "step": 93 + }, + { + "epoch": 0.01673789173789174, + "grad_norm": 0.38709723949432373, + "learning_rate": 0.00019999463550572516, + "loss": 1.2623, + "step": 94 + }, + { + "epoch": 0.016915954415954417, + "grad_norm": 0.35376182198524475, + "learning_rate": 0.00019999448954109662, + "loss": 1.0643, + "step": 95 + }, + { + "epoch": 0.017094017094017096, + "grad_norm": 0.49547120928764343, + "learning_rate": 0.00019999434161728377, + "loss": 1.2121, + "step": 96 + }, + { + "epoch": 0.01727207977207977, + "grad_norm": 0.49593672156333923, + "learning_rate": 0.00019999419173428952, + "loss": 1.1635, + "step": 97 + }, + { + "epoch": 0.01745014245014245, + "grad_norm": 0.4146541953086853, + "learning_rate": 0.0001999940398921168, + "loss": 1.1452, + "step": 98 + }, + { + "epoch": 0.017628205128205128, + "grad_norm": 0.5177254676818848, + "learning_rate": 0.00019999388609076858, + "loss": 1.2178, + "step": 99 + }, + { + "epoch": 0.017806267806267807, + "grad_norm": 0.4012768864631653, + "learning_rate": 0.0001999937303302479, + "loss": 0.9222, + "step": 100 + }, + { + "epoch": 0.017984330484330485, + "grad_norm": 0.4597131907939911, + "learning_rate": 0.00019999357261055777, + "loss": 0.979, + "step": 101 + }, + { + "epoch": 0.018162393162393164, + "grad_norm": 0.6190966963768005, + "learning_rate": 0.00019999341293170132, + "loss": 1.3909, + "step": 102 + }, + { + "epoch": 0.01834045584045584, + "grad_norm": 0.4576462209224701, + "learning_rate": 0.00019999325129368164, + "loss": 1.073, + "step": 103 + }, + { + "epoch": 0.018518518518518517, + "grad_norm": 0.4036749005317688, + "learning_rate": 0.00019999308769650192, + "loss": 1.1354, + "step": 104 + }, + { + "epoch": 0.018696581196581196, + "grad_norm": 0.4722452759742737, + "learning_rate": 0.00019999292214016538, + "loss": 1.2039, + "step": 105 + }, + { + "epoch": 0.018874643874643875, + "grad_norm": 0.5338274240493774, + "learning_rate": 0.00019999275462467527, + "loss": 1.225, + "step": 106 + }, + { + "epoch": 0.019052706552706553, + "grad_norm": 0.4301491677761078, + "learning_rate": 0.00019999258515003484, + "loss": 1.0601, + "step": 107 + }, + { + "epoch": 0.019230769230769232, + "grad_norm": 0.33271175622940063, + "learning_rate": 0.0001999924137162474, + "loss": 0.8441, + "step": 108 + }, + { + "epoch": 0.01940883190883191, + "grad_norm": 0.4648784399032593, + "learning_rate": 0.0001999922403233163, + "loss": 1.2038, + "step": 109 + }, + { + "epoch": 0.019586894586894586, + "grad_norm": 0.37915176153182983, + "learning_rate": 0.00019999206497124504, + "loss": 1.0923, + "step": 110 + }, + { + "epoch": 0.019764957264957264, + "grad_norm": 0.3865506052970886, + "learning_rate": 0.00019999188766003695, + "loss": 0.9535, + "step": 111 + }, + { + "epoch": 0.019943019943019943, + "grad_norm": 0.35739636421203613, + "learning_rate": 0.0001999917083896955, + "loss": 1.2688, + "step": 112 + }, + { + "epoch": 0.02012108262108262, + "grad_norm": 0.3943796157836914, + "learning_rate": 0.0001999915271602243, + "loss": 1.1097, + "step": 113 + }, + { + "epoch": 0.0202991452991453, + "grad_norm": 0.44758161902427673, + "learning_rate": 0.0001999913439716268, + "loss": 1.2698, + "step": 114 + }, + { + "epoch": 0.02047720797720798, + "grad_norm": 0.3749747574329376, + "learning_rate": 0.00019999115882390664, + "loss": 1.1091, + "step": 115 + }, + { + "epoch": 0.020655270655270654, + "grad_norm": 0.3479487895965576, + "learning_rate": 0.00019999097171706745, + "loss": 1.0049, + "step": 116 + }, + { + "epoch": 0.020833333333333332, + "grad_norm": 0.4491243064403534, + "learning_rate": 0.00019999078265111285, + "loss": 1.1857, + "step": 117 + }, + { + "epoch": 0.02101139601139601, + "grad_norm": 0.345289021730423, + "learning_rate": 0.00019999059162604662, + "loss": 1.1397, + "step": 118 + }, + { + "epoch": 0.02118945868945869, + "grad_norm": 0.5467649698257446, + "learning_rate": 0.00019999039864187243, + "loss": 1.2196, + "step": 119 + }, + { + "epoch": 0.021367521367521368, + "grad_norm": 0.36446481943130493, + "learning_rate": 0.00019999020369859409, + "loss": 0.796, + "step": 120 + }, + { + "epoch": 0.021545584045584047, + "grad_norm": 0.4225841760635376, + "learning_rate": 0.00019999000679621543, + "loss": 0.9684, + "step": 121 + }, + { + "epoch": 0.021723646723646725, + "grad_norm": 0.4205594062805176, + "learning_rate": 0.0001999898079347403, + "loss": 1.2762, + "step": 122 + }, + { + "epoch": 0.0219017094017094, + "grad_norm": 0.43773892521858215, + "learning_rate": 0.00019998960711417257, + "loss": 1.117, + "step": 123 + }, + { + "epoch": 0.02207977207977208, + "grad_norm": 0.41279685497283936, + "learning_rate": 0.00019998940433451623, + "loss": 1.1502, + "step": 124 + }, + { + "epoch": 0.022257834757834757, + "grad_norm": 0.4090803563594818, + "learning_rate": 0.0001999891995957752, + "loss": 1.2591, + "step": 125 + }, + { + "epoch": 0.022435897435897436, + "grad_norm": 0.6000410914421082, + "learning_rate": 0.0001999889928979535, + "loss": 1.4321, + "step": 126 + }, + { + "epoch": 0.022613960113960115, + "grad_norm": 0.524264395236969, + "learning_rate": 0.00019998878424105524, + "loss": 1.1849, + "step": 127 + }, + { + "epoch": 0.022792022792022793, + "grad_norm": 0.4581047296524048, + "learning_rate": 0.00019998857362508443, + "loss": 1.0598, + "step": 128 + }, + { + "epoch": 0.022970085470085472, + "grad_norm": 0.42663446068763733, + "learning_rate": 0.00019998836105004526, + "loss": 1.1909, + "step": 129 + }, + { + "epoch": 0.023148148148148147, + "grad_norm": 0.45709118247032166, + "learning_rate": 0.00019998814651594183, + "loss": 1.2104, + "step": 130 + }, + { + "epoch": 0.023326210826210825, + "grad_norm": 0.39528369903564453, + "learning_rate": 0.0001999879300227784, + "loss": 1.3073, + "step": 131 + }, + { + "epoch": 0.023504273504273504, + "grad_norm": 0.46896448731422424, + "learning_rate": 0.00019998771157055914, + "loss": 1.3202, + "step": 132 + }, + { + "epoch": 0.023682336182336183, + "grad_norm": 0.4386129677295685, + "learning_rate": 0.00019998749115928842, + "loss": 1.2196, + "step": 133 + }, + { + "epoch": 0.02386039886039886, + "grad_norm": 0.45920488238334656, + "learning_rate": 0.00019998726878897051, + "loss": 1.3668, + "step": 134 + }, + { + "epoch": 0.02403846153846154, + "grad_norm": 0.4115797281265259, + "learning_rate": 0.0001999870444596098, + "loss": 1.1052, + "step": 135 + }, + { + "epoch": 0.024216524216524215, + "grad_norm": 0.3860839903354645, + "learning_rate": 0.0001999868181712106, + "loss": 1.0344, + "step": 136 + }, + { + "epoch": 0.024394586894586893, + "grad_norm": 0.42514732480049133, + "learning_rate": 0.00019998658992377742, + "loss": 1.1979, + "step": 137 + }, + { + "epoch": 0.024572649572649572, + "grad_norm": 0.36001840233802795, + "learning_rate": 0.00019998635971731475, + "loss": 1.4536, + "step": 138 + }, + { + "epoch": 0.02475071225071225, + "grad_norm": 0.3739112317562103, + "learning_rate": 0.00019998612755182707, + "loss": 1.0097, + "step": 139 + }, + { + "epoch": 0.02492877492877493, + "grad_norm": 0.37545472383499146, + "learning_rate": 0.00019998589342731888, + "loss": 0.829, + "step": 140 + }, + { + "epoch": 0.025106837606837608, + "grad_norm": 0.38660728931427, + "learning_rate": 0.0001999856573437948, + "loss": 1.1324, + "step": 141 + }, + { + "epoch": 0.025284900284900286, + "grad_norm": 0.3741356432437897, + "learning_rate": 0.00019998541930125953, + "loss": 1.0934, + "step": 142 + }, + { + "epoch": 0.02546296296296296, + "grad_norm": 0.41900336742401123, + "learning_rate": 0.00019998517929971764, + "loss": 1.0336, + "step": 143 + }, + { + "epoch": 0.02564102564102564, + "grad_norm": 0.4167572259902954, + "learning_rate": 0.00019998493733917384, + "loss": 1.2571, + "step": 144 + }, + { + "epoch": 0.02581908831908832, + "grad_norm": 0.39437636733055115, + "learning_rate": 0.0001999846934196329, + "loss": 1.2283, + "step": 145 + }, + { + "epoch": 0.025997150997150997, + "grad_norm": 0.39129480719566345, + "learning_rate": 0.00019998444754109964, + "loss": 0.9893, + "step": 146 + }, + { + "epoch": 0.026175213675213676, + "grad_norm": 0.45533549785614014, + "learning_rate": 0.0001999841997035788, + "loss": 1.0793, + "step": 147 + }, + { + "epoch": 0.026353276353276354, + "grad_norm": 0.3741768002510071, + "learning_rate": 0.00019998394990707524, + "loss": 1.2179, + "step": 148 + }, + { + "epoch": 0.026531339031339033, + "grad_norm": 0.4066533148288727, + "learning_rate": 0.0001999836981515939, + "loss": 1.1443, + "step": 149 + }, + { + "epoch": 0.026709401709401708, + "grad_norm": 0.4851688742637634, + "learning_rate": 0.0001999834444371397, + "loss": 1.1668, + "step": 150 + }, + { + "epoch": 0.026887464387464387, + "grad_norm": 0.428091436624527, + "learning_rate": 0.0001999831887637176, + "loss": 1.2676, + "step": 151 + }, + { + "epoch": 0.027065527065527065, + "grad_norm": 0.4024655222892761, + "learning_rate": 0.0001999829311313326, + "loss": 1.3115, + "step": 152 + }, + { + "epoch": 0.027243589743589744, + "grad_norm": 0.43983033299446106, + "learning_rate": 0.00019998267153998976, + "loss": 1.1019, + "step": 153 + }, + { + "epoch": 0.027421652421652423, + "grad_norm": 0.4317505359649658, + "learning_rate": 0.0001999824099896942, + "loss": 1.3129, + "step": 154 + }, + { + "epoch": 0.0275997150997151, + "grad_norm": 0.43107882142066956, + "learning_rate": 0.000199982146480451, + "loss": 1.2134, + "step": 155 + }, + { + "epoch": 0.027777777777777776, + "grad_norm": 0.3939448297023773, + "learning_rate": 0.00019998188101226532, + "loss": 1.0321, + "step": 156 + }, + { + "epoch": 0.027955840455840455, + "grad_norm": 0.4641847610473633, + "learning_rate": 0.00019998161358514237, + "loss": 1.2369, + "step": 157 + }, + { + "epoch": 0.028133903133903133, + "grad_norm": 0.3538529872894287, + "learning_rate": 0.0001999813441990874, + "loss": 1.2061, + "step": 158 + }, + { + "epoch": 0.028311965811965812, + "grad_norm": 0.3277950584888458, + "learning_rate": 0.0001999810728541057, + "loss": 0.9419, + "step": 159 + }, + { + "epoch": 0.02849002849002849, + "grad_norm": 0.424710750579834, + "learning_rate": 0.00019998079955020254, + "loss": 1.3302, + "step": 160 + }, + { + "epoch": 0.02866809116809117, + "grad_norm": 0.4120834469795227, + "learning_rate": 0.00019998052428738333, + "loss": 1.079, + "step": 161 + }, + { + "epoch": 0.028846153846153848, + "grad_norm": 0.45811930298805237, + "learning_rate": 0.00019998024706565346, + "loss": 1.1259, + "step": 162 + }, + { + "epoch": 0.029024216524216523, + "grad_norm": 0.3873266875743866, + "learning_rate": 0.0001999799678850183, + "loss": 1.2124, + "step": 163 + }, + { + "epoch": 0.0292022792022792, + "grad_norm": 0.5806412696838379, + "learning_rate": 0.00019997968674548337, + "loss": 1.3467, + "step": 164 + }, + { + "epoch": 0.02938034188034188, + "grad_norm": 0.3906802833080292, + "learning_rate": 0.00019997940364705418, + "loss": 1.1438, + "step": 165 + }, + { + "epoch": 0.02955840455840456, + "grad_norm": 0.45201995968818665, + "learning_rate": 0.00019997911858973626, + "loss": 1.1469, + "step": 166 + }, + { + "epoch": 0.029736467236467237, + "grad_norm": 0.4965892732143402, + "learning_rate": 0.0001999788315735352, + "loss": 1.0829, + "step": 167 + }, + { + "epoch": 0.029914529914529916, + "grad_norm": 0.32578057050704956, + "learning_rate": 0.0001999785425984566, + "loss": 1.0432, + "step": 168 + }, + { + "epoch": 0.03009259259259259, + "grad_norm": 0.4146028161048889, + "learning_rate": 0.00019997825166450617, + "loss": 1.1657, + "step": 169 + }, + { + "epoch": 0.03027065527065527, + "grad_norm": 0.4342964291572571, + "learning_rate": 0.0001999779587716896, + "loss": 1.2038, + "step": 170 + }, + { + "epoch": 0.030448717948717948, + "grad_norm": 0.40128546953201294, + "learning_rate": 0.00019997766392001258, + "loss": 1.3044, + "step": 171 + }, + { + "epoch": 0.030626780626780627, + "grad_norm": 0.4357539117336273, + "learning_rate": 0.00019997736710948094, + "loss": 1.2143, + "step": 172 + }, + { + "epoch": 0.030804843304843305, + "grad_norm": 0.4821035861968994, + "learning_rate": 0.00019997706834010045, + "loss": 1.0469, + "step": 173 + }, + { + "epoch": 0.030982905982905984, + "grad_norm": 0.3966675102710724, + "learning_rate": 0.000199976767611877, + "loss": 1.2122, + "step": 174 + }, + { + "epoch": 0.031160968660968662, + "grad_norm": 0.4265064299106598, + "learning_rate": 0.00019997646492481648, + "loss": 1.0871, + "step": 175 + }, + { + "epoch": 0.03133903133903134, + "grad_norm": 0.3445652723312378, + "learning_rate": 0.00019997616027892485, + "loss": 1.0412, + "step": 176 + }, + { + "epoch": 0.031517094017094016, + "grad_norm": 0.47187718749046326, + "learning_rate": 0.000199975853674208, + "loss": 1.0822, + "step": 177 + }, + { + "epoch": 0.0316951566951567, + "grad_norm": 0.37751707434654236, + "learning_rate": 0.000199975545110672, + "loss": 1.1439, + "step": 178 + }, + { + "epoch": 0.03187321937321937, + "grad_norm": 0.38792455196380615, + "learning_rate": 0.00019997523458832286, + "loss": 0.8604, + "step": 179 + }, + { + "epoch": 0.03205128205128205, + "grad_norm": 0.35199594497680664, + "learning_rate": 0.00019997492210716667, + "loss": 1.0819, + "step": 180 + }, + { + "epoch": 0.03222934472934473, + "grad_norm": 0.4828922748565674, + "learning_rate": 0.00019997460766720958, + "loss": 1.1879, + "step": 181 + }, + { + "epoch": 0.032407407407407406, + "grad_norm": 0.46153363585472107, + "learning_rate": 0.00019997429126845774, + "loss": 1.1592, + "step": 182 + }, + { + "epoch": 0.03258547008547009, + "grad_norm": 0.4844890832901001, + "learning_rate": 0.0001999739729109173, + "loss": 1.1334, + "step": 183 + }, + { + "epoch": 0.03276353276353276, + "grad_norm": 0.414617121219635, + "learning_rate": 0.00019997365259459457, + "loss": 1.0547, + "step": 184 + }, + { + "epoch": 0.032941595441595445, + "grad_norm": 0.46544626355171204, + "learning_rate": 0.00019997333031949581, + "loss": 1.4067, + "step": 185 + }, + { + "epoch": 0.03311965811965812, + "grad_norm": 0.48489415645599365, + "learning_rate": 0.0001999730060856273, + "loss": 1.4027, + "step": 186 + }, + { + "epoch": 0.033297720797720795, + "grad_norm": 0.3963346481323242, + "learning_rate": 0.0001999726798929954, + "loss": 1.1327, + "step": 187 + }, + { + "epoch": 0.03347578347578348, + "grad_norm": 0.3809385895729065, + "learning_rate": 0.00019997235174160652, + "loss": 1.3475, + "step": 188 + }, + { + "epoch": 0.03365384615384615, + "grad_norm": 0.3866960406303406, + "learning_rate": 0.0001999720216314671, + "loss": 1.1576, + "step": 189 + }, + { + "epoch": 0.033831908831908834, + "grad_norm": 0.34976935386657715, + "learning_rate": 0.00019997168956258356, + "loss": 0.9361, + "step": 190 + }, + { + "epoch": 0.03400997150997151, + "grad_norm": 0.38681939244270325, + "learning_rate": 0.00019997135553496243, + "loss": 1.1796, + "step": 191 + }, + { + "epoch": 0.03418803418803419, + "grad_norm": 0.41905197501182556, + "learning_rate": 0.0001999710195486103, + "loss": 1.1714, + "step": 192 + }, + { + "epoch": 0.03436609686609687, + "grad_norm": 0.42356589436531067, + "learning_rate": 0.0001999706816035337, + "loss": 1.0022, + "step": 193 + }, + { + "epoch": 0.03454415954415954, + "grad_norm": 0.3929740786552429, + "learning_rate": 0.00019997034169973925, + "loss": 1.3769, + "step": 194 + }, + { + "epoch": 0.034722222222222224, + "grad_norm": 0.4325186312198639, + "learning_rate": 0.00019996999983723366, + "loss": 1.3057, + "step": 195 + }, + { + "epoch": 0.0349002849002849, + "grad_norm": 0.3954029381275177, + "learning_rate": 0.00019996965601602355, + "loss": 1.1958, + "step": 196 + }, + { + "epoch": 0.03507834757834758, + "grad_norm": 0.34454262256622314, + "learning_rate": 0.00019996931023611572, + "loss": 1.0972, + "step": 197 + }, + { + "epoch": 0.035256410256410256, + "grad_norm": 0.48900291323661804, + "learning_rate": 0.0001999689624975169, + "loss": 1.213, + "step": 198 + }, + { + "epoch": 0.03543447293447293, + "grad_norm": 0.35214388370513916, + "learning_rate": 0.00019996861280023397, + "loss": 1.0285, + "step": 199 + }, + { + "epoch": 0.03561253561253561, + "grad_norm": 0.49393126368522644, + "learning_rate": 0.00019996826114427373, + "loss": 1.2313, + "step": 200 + }, + { + "epoch": 0.03579059829059829, + "grad_norm": 0.3994458019733429, + "learning_rate": 0.00019996790752964305, + "loss": 1.0474, + "step": 201 + }, + { + "epoch": 0.03596866096866097, + "grad_norm": 0.5387318730354309, + "learning_rate": 0.0001999675519563489, + "loss": 1.3067, + "step": 202 + }, + { + "epoch": 0.036146723646723646, + "grad_norm": 0.4976751208305359, + "learning_rate": 0.00019996719442439824, + "loss": 1.2593, + "step": 203 + }, + { + "epoch": 0.03632478632478633, + "grad_norm": 0.47052907943725586, + "learning_rate": 0.0001999668349337981, + "loss": 1.1036, + "step": 204 + }, + { + "epoch": 0.036502849002849, + "grad_norm": 0.39616644382476807, + "learning_rate": 0.00019996647348455543, + "loss": 1.0481, + "step": 205 + }, + { + "epoch": 0.03668091168091168, + "grad_norm": 0.42987677454948425, + "learning_rate": 0.00019996611007667742, + "loss": 1.0923, + "step": 206 + }, + { + "epoch": 0.03685897435897436, + "grad_norm": 0.47065848112106323, + "learning_rate": 0.00019996574471017113, + "loss": 1.1403, + "step": 207 + }, + { + "epoch": 0.037037037037037035, + "grad_norm": 0.4363015592098236, + "learning_rate": 0.00019996537738504373, + "loss": 1.253, + "step": 208 + }, + { + "epoch": 0.03721509971509972, + "grad_norm": 0.4038296937942505, + "learning_rate": 0.00019996500810130243, + "loss": 1.1679, + "step": 209 + }, + { + "epoch": 0.03739316239316239, + "grad_norm": 0.5038532018661499, + "learning_rate": 0.00019996463685895445, + "loss": 1.1182, + "step": 210 + }, + { + "epoch": 0.037571225071225074, + "grad_norm": 0.37740692496299744, + "learning_rate": 0.00019996426365800706, + "loss": 1.0465, + "step": 211 + }, + { + "epoch": 0.03774928774928775, + "grad_norm": 0.47794604301452637, + "learning_rate": 0.00019996388849846759, + "loss": 1.2836, + "step": 212 + }, + { + "epoch": 0.037927350427350424, + "grad_norm": 0.38460609316825867, + "learning_rate": 0.0001999635113803434, + "loss": 1.2099, + "step": 213 + }, + { + "epoch": 0.038105413105413107, + "grad_norm": 0.42016157507896423, + "learning_rate": 0.0001999631323036418, + "loss": 1.152, + "step": 214 + }, + { + "epoch": 0.03828347578347578, + "grad_norm": 0.4024946391582489, + "learning_rate": 0.00019996275126837033, + "loss": 1.1534, + "step": 215 + }, + { + "epoch": 0.038461538461538464, + "grad_norm": 0.4573793411254883, + "learning_rate": 0.00019996236827453642, + "loss": 1.2019, + "step": 216 + }, + { + "epoch": 0.03863960113960114, + "grad_norm": 0.3642503321170807, + "learning_rate": 0.0001999619833221475, + "loss": 1.0541, + "step": 217 + }, + { + "epoch": 0.03881766381766382, + "grad_norm": 0.38492897152900696, + "learning_rate": 0.0001999615964112112, + "loss": 1.1269, + "step": 218 + }, + { + "epoch": 0.038995726495726496, + "grad_norm": 0.427219420671463, + "learning_rate": 0.0001999612075417351, + "loss": 1.1126, + "step": 219 + }, + { + "epoch": 0.03917378917378917, + "grad_norm": 0.40781742334365845, + "learning_rate": 0.00019996081671372676, + "loss": 1.2207, + "step": 220 + }, + { + "epoch": 0.03935185185185185, + "grad_norm": 0.39229512214660645, + "learning_rate": 0.00019996042392719386, + "loss": 1.0403, + "step": 221 + }, + { + "epoch": 0.03952991452991453, + "grad_norm": 0.42038577795028687, + "learning_rate": 0.0001999600291821441, + "loss": 1.2157, + "step": 222 + }, + { + "epoch": 0.03970797720797721, + "grad_norm": 0.3963491916656494, + "learning_rate": 0.00019995963247858525, + "loss": 1.0532, + "step": 223 + }, + { + "epoch": 0.039886039886039885, + "grad_norm": 0.4389874041080475, + "learning_rate": 0.00019995923381652502, + "loss": 1.4279, + "step": 224 + }, + { + "epoch": 0.04006410256410257, + "grad_norm": 0.357312947511673, + "learning_rate": 0.00019995883319597123, + "loss": 0.9871, + "step": 225 + }, + { + "epoch": 0.04024216524216524, + "grad_norm": 0.3644427955150604, + "learning_rate": 0.00019995843061693181, + "loss": 1.0879, + "step": 226 + }, + { + "epoch": 0.04042022792022792, + "grad_norm": 0.4074651002883911, + "learning_rate": 0.00019995802607941453, + "loss": 1.2138, + "step": 227 + }, + { + "epoch": 0.0405982905982906, + "grad_norm": 0.40709465742111206, + "learning_rate": 0.0001999576195834274, + "loss": 1.1905, + "step": 228 + }, + { + "epoch": 0.040776353276353275, + "grad_norm": 0.4280182719230652, + "learning_rate": 0.00019995721112897838, + "loss": 1.2331, + "step": 229 + }, + { + "epoch": 0.04095441595441596, + "grad_norm": 0.37846076488494873, + "learning_rate": 0.00019995680071607544, + "loss": 1.078, + "step": 230 + }, + { + "epoch": 0.04113247863247863, + "grad_norm": 0.3877260088920593, + "learning_rate": 0.0001999563883447266, + "loss": 1.0309, + "step": 231 + }, + { + "epoch": 0.04131054131054131, + "grad_norm": 0.42886826395988464, + "learning_rate": 0.00019995597401494, + "loss": 1.0403, + "step": 232 + }, + { + "epoch": 0.04148860398860399, + "grad_norm": 0.4316534101963043, + "learning_rate": 0.00019995555772672372, + "loss": 1.2418, + "step": 233 + }, + { + "epoch": 0.041666666666666664, + "grad_norm": 0.45768865942955017, + "learning_rate": 0.00019995513948008593, + "loss": 1.233, + "step": 234 + }, + { + "epoch": 0.041844729344729346, + "grad_norm": 0.5647913813591003, + "learning_rate": 0.00019995471927503481, + "loss": 1.1346, + "step": 235 + }, + { + "epoch": 0.04202279202279202, + "grad_norm": 0.3797492980957031, + "learning_rate": 0.00019995429711157863, + "loss": 1.1574, + "step": 236 + }, + { + "epoch": 0.042200854700854704, + "grad_norm": 0.4392767548561096, + "learning_rate": 0.00019995387298972562, + "loss": 0.8988, + "step": 237 + }, + { + "epoch": 0.04237891737891738, + "grad_norm": 0.37331557273864746, + "learning_rate": 0.0001999534469094841, + "loss": 1.0439, + "step": 238 + }, + { + "epoch": 0.042556980056980054, + "grad_norm": 0.3785935938358307, + "learning_rate": 0.00019995301887086245, + "loss": 0.9839, + "step": 239 + }, + { + "epoch": 0.042735042735042736, + "grad_norm": 0.4351862668991089, + "learning_rate": 0.00019995258887386898, + "loss": 1.2653, + "step": 240 + }, + { + "epoch": 0.04291310541310541, + "grad_norm": 0.399475634098053, + "learning_rate": 0.0001999521569185122, + "loss": 0.9877, + "step": 241 + }, + { + "epoch": 0.04309116809116809, + "grad_norm": 0.42332810163497925, + "learning_rate": 0.00019995172300480053, + "loss": 1.2403, + "step": 242 + }, + { + "epoch": 0.04326923076923077, + "grad_norm": 0.4397708475589752, + "learning_rate": 0.00019995128713274247, + "loss": 0.9316, + "step": 243 + }, + { + "epoch": 0.04344729344729345, + "grad_norm": 0.3614110052585602, + "learning_rate": 0.00019995084930234658, + "loss": 1.1088, + "step": 244 + }, + { + "epoch": 0.043625356125356125, + "grad_norm": 0.39433717727661133, + "learning_rate": 0.0001999504095136214, + "loss": 1.2002, + "step": 245 + }, + { + "epoch": 0.0438034188034188, + "grad_norm": 0.33088216185569763, + "learning_rate": 0.0001999499677665756, + "loss": 0.8796, + "step": 246 + }, + { + "epoch": 0.04398148148148148, + "grad_norm": 0.5239143967628479, + "learning_rate": 0.00019994952406121784, + "loss": 1.2808, + "step": 247 + }, + { + "epoch": 0.04415954415954416, + "grad_norm": 0.42156723141670227, + "learning_rate": 0.00019994907839755675, + "loss": 1.1775, + "step": 248 + }, + { + "epoch": 0.04433760683760684, + "grad_norm": 0.42569902539253235, + "learning_rate": 0.0001999486307756011, + "loss": 1.001, + "step": 249 + }, + { + "epoch": 0.044515669515669515, + "grad_norm": 0.38241544365882874, + "learning_rate": 0.00019994818119535964, + "loss": 1.1064, + "step": 250 + }, + { + "epoch": 0.0446937321937322, + "grad_norm": 0.4185071885585785, + "learning_rate": 0.0001999477296568412, + "loss": 1.2109, + "step": 251 + }, + { + "epoch": 0.04487179487179487, + "grad_norm": 0.4189644157886505, + "learning_rate": 0.00019994727616005464, + "loss": 1.2902, + "step": 252 + }, + { + "epoch": 0.04504985754985755, + "grad_norm": 0.34671884775161743, + "learning_rate": 0.0001999468207050088, + "loss": 0.9429, + "step": 253 + }, + { + "epoch": 0.04522792022792023, + "grad_norm": 0.42391687631607056, + "learning_rate": 0.00019994636329171266, + "loss": 0.7179, + "step": 254 + }, + { + "epoch": 0.045405982905982904, + "grad_norm": 0.3803195655345917, + "learning_rate": 0.00019994590392017513, + "loss": 1.0318, + "step": 255 + }, + { + "epoch": 0.045584045584045586, + "grad_norm": 0.3389956057071686, + "learning_rate": 0.00019994544259040525, + "loss": 1.0485, + "step": 256 + }, + { + "epoch": 0.04576210826210826, + "grad_norm": 0.4927038550376892, + "learning_rate": 0.000199944979302412, + "loss": 1.3426, + "step": 257 + }, + { + "epoch": 0.045940170940170943, + "grad_norm": 0.33200421929359436, + "learning_rate": 0.00019994451405620453, + "loss": 1.0071, + "step": 258 + }, + { + "epoch": 0.04611823361823362, + "grad_norm": 0.38028615713119507, + "learning_rate": 0.00019994404685179195, + "loss": 1.0985, + "step": 259 + }, + { + "epoch": 0.046296296296296294, + "grad_norm": 0.3752151429653168, + "learning_rate": 0.00019994357768918333, + "loss": 0.9209, + "step": 260 + }, + { + "epoch": 0.046474358974358976, + "grad_norm": 0.43030866980552673, + "learning_rate": 0.00019994310656838796, + "loss": 0.9921, + "step": 261 + }, + { + "epoch": 0.04665242165242165, + "grad_norm": 0.4402460753917694, + "learning_rate": 0.00019994263348941502, + "loss": 1.1051, + "step": 262 + }, + { + "epoch": 0.04683048433048433, + "grad_norm": 0.43012720346450806, + "learning_rate": 0.0001999421584522738, + "loss": 1.1839, + "step": 263 + }, + { + "epoch": 0.04700854700854701, + "grad_norm": 0.4195305407047272, + "learning_rate": 0.0001999416814569736, + "loss": 1.1749, + "step": 264 + }, + { + "epoch": 0.04718660968660968, + "grad_norm": 0.45623287558555603, + "learning_rate": 0.00019994120250352372, + "loss": 1.2433, + "step": 265 + }, + { + "epoch": 0.047364672364672365, + "grad_norm": 0.4736156761646271, + "learning_rate": 0.00019994072159193363, + "loss": 1.2882, + "step": 266 + }, + { + "epoch": 0.04754273504273504, + "grad_norm": 0.36698561906814575, + "learning_rate": 0.0001999402387222127, + "loss": 1.1486, + "step": 267 + }, + { + "epoch": 0.04772079772079772, + "grad_norm": 0.3854144215583801, + "learning_rate": 0.00019993975389437038, + "loss": 0.8115, + "step": 268 + }, + { + "epoch": 0.0478988603988604, + "grad_norm": 0.41512808203697205, + "learning_rate": 0.0001999392671084162, + "loss": 1.0959, + "step": 269 + }, + { + "epoch": 0.04807692307692308, + "grad_norm": 0.3869563341140747, + "learning_rate": 0.0001999387783643597, + "loss": 1.087, + "step": 270 + }, + { + "epoch": 0.048254985754985755, + "grad_norm": 0.4649744927883148, + "learning_rate": 0.00019993828766221044, + "loss": 1.0011, + "step": 271 + }, + { + "epoch": 0.04843304843304843, + "grad_norm": 0.40331923961639404, + "learning_rate": 0.00019993779500197803, + "loss": 1.1463, + "step": 272 + }, + { + "epoch": 0.04861111111111111, + "grad_norm": 0.3826279938220978, + "learning_rate": 0.0001999373003836721, + "loss": 1.1491, + "step": 273 + }, + { + "epoch": 0.04878917378917379, + "grad_norm": 0.3967166543006897, + "learning_rate": 0.00019993680380730243, + "loss": 1.1462, + "step": 274 + }, + { + "epoch": 0.04896723646723647, + "grad_norm": 0.4298507869243622, + "learning_rate": 0.00019993630527287865, + "loss": 1.2471, + "step": 275 + }, + { + "epoch": 0.049145299145299144, + "grad_norm": 0.41486215591430664, + "learning_rate": 0.0001999358047804106, + "loss": 1.287, + "step": 276 + }, + { + "epoch": 0.049323361823361826, + "grad_norm": 0.3914124369621277, + "learning_rate": 0.00019993530232990803, + "loss": 1.0935, + "step": 277 + }, + { + "epoch": 0.0495014245014245, + "grad_norm": 0.39888378977775574, + "learning_rate": 0.00019993479792138082, + "loss": 1.2347, + "step": 278 + }, + { + "epoch": 0.049679487179487176, + "grad_norm": 0.3911665678024292, + "learning_rate": 0.00019993429155483884, + "loss": 1.0917, + "step": 279 + }, + { + "epoch": 0.04985754985754986, + "grad_norm": 0.42871445417404175, + "learning_rate": 0.00019993378323029197, + "loss": 1.0277, + "step": 280 + }, + { + "epoch": 0.050035612535612534, + "grad_norm": 0.35397860407829285, + "learning_rate": 0.00019993327294775027, + "loss": 0.9549, + "step": 281 + }, + { + "epoch": 0.050213675213675216, + "grad_norm": 0.4528059959411621, + "learning_rate": 0.00019993276070722364, + "loss": 1.2338, + "step": 282 + }, + { + "epoch": 0.05039173789173789, + "grad_norm": 0.354735791683197, + "learning_rate": 0.00019993224650872218, + "loss": 1.1892, + "step": 283 + }, + { + "epoch": 0.05056980056980057, + "grad_norm": 0.44407567381858826, + "learning_rate": 0.00019993173035225592, + "loss": 1.1621, + "step": 284 + }, + { + "epoch": 0.05074786324786325, + "grad_norm": 0.4177244305610657, + "learning_rate": 0.000199931212237835, + "loss": 1.1184, + "step": 285 + }, + { + "epoch": 0.05092592592592592, + "grad_norm": 0.5627759695053101, + "learning_rate": 0.0001999306921654696, + "loss": 1.0755, + "step": 286 + }, + { + "epoch": 0.051103988603988605, + "grad_norm": 0.46767523884773254, + "learning_rate": 0.00019993017013516986, + "loss": 1.2654, + "step": 287 + }, + { + "epoch": 0.05128205128205128, + "grad_norm": 0.4163128733634949, + "learning_rate": 0.000199929646146946, + "loss": 1.1307, + "step": 288 + }, + { + "epoch": 0.05146011396011396, + "grad_norm": 0.36954161524772644, + "learning_rate": 0.00019992912020080832, + "loss": 0.8274, + "step": 289 + }, + { + "epoch": 0.05163817663817664, + "grad_norm": 0.4770594835281372, + "learning_rate": 0.00019992859229676712, + "loss": 1.2235, + "step": 290 + }, + { + "epoch": 0.05181623931623932, + "grad_norm": 0.4174608290195465, + "learning_rate": 0.00019992806243483274, + "loss": 1.2893, + "step": 291 + }, + { + "epoch": 0.051994301994301995, + "grad_norm": 0.3794898986816406, + "learning_rate": 0.00019992753061501555, + "loss": 1.104, + "step": 292 + }, + { + "epoch": 0.05217236467236467, + "grad_norm": 0.3912592828273773, + "learning_rate": 0.000199926996837326, + "loss": 1.0043, + "step": 293 + }, + { + "epoch": 0.05235042735042735, + "grad_norm": 0.39641159772872925, + "learning_rate": 0.00019992646110177448, + "loss": 1.083, + "step": 294 + }, + { + "epoch": 0.05252849002849003, + "grad_norm": 0.3518857955932617, + "learning_rate": 0.00019992592340837157, + "loss": 0.9275, + "step": 295 + }, + { + "epoch": 0.05270655270655271, + "grad_norm": 0.3955721855163574, + "learning_rate": 0.00019992538375712777, + "loss": 1.0153, + "step": 296 + }, + { + "epoch": 0.052884615384615384, + "grad_norm": 0.3837333023548126, + "learning_rate": 0.00019992484214805364, + "loss": 1.1664, + "step": 297 + }, + { + "epoch": 0.053062678062678066, + "grad_norm": 0.39400920271873474, + "learning_rate": 0.0001999242985811598, + "loss": 1.0532, + "step": 298 + }, + { + "epoch": 0.05324074074074074, + "grad_norm": 0.39258649945259094, + "learning_rate": 0.00019992375305645692, + "loss": 1.0081, + "step": 299 + }, + { + "epoch": 0.053418803418803416, + "grad_norm": 0.49768248200416565, + "learning_rate": 0.00019992320557395566, + "loss": 1.2553, + "step": 300 + }, + { + "epoch": 0.0535968660968661, + "grad_norm": 0.364776074886322, + "learning_rate": 0.00019992265613366677, + "loss": 1.0582, + "step": 301 + }, + { + "epoch": 0.053774928774928774, + "grad_norm": 0.47317907214164734, + "learning_rate": 0.00019992210473560097, + "loss": 1.3114, + "step": 302 + }, + { + "epoch": 0.053952991452991456, + "grad_norm": 0.3706119656562805, + "learning_rate": 0.00019992155137976917, + "loss": 0.9554, + "step": 303 + }, + { + "epoch": 0.05413105413105413, + "grad_norm": 0.42809563875198364, + "learning_rate": 0.0001999209960661821, + "loss": 1.306, + "step": 304 + }, + { + "epoch": 0.054309116809116806, + "grad_norm": 0.4514487385749817, + "learning_rate": 0.00019992043879485066, + "loss": 1.0147, + "step": 305 + }, + { + "epoch": 0.05448717948717949, + "grad_norm": 0.36672836542129517, + "learning_rate": 0.0001999198795657858, + "loss": 1.1392, + "step": 306 + }, + { + "epoch": 0.05466524216524216, + "grad_norm": 0.4206554889678955, + "learning_rate": 0.00019991931837899847, + "loss": 1.2405, + "step": 307 + }, + { + "epoch": 0.054843304843304845, + "grad_norm": 0.46168261766433716, + "learning_rate": 0.00019991875523449966, + "loss": 1.2707, + "step": 308 + }, + { + "epoch": 0.05502136752136752, + "grad_norm": 0.39503365755081177, + "learning_rate": 0.00019991819013230039, + "loss": 1.0776, + "step": 309 + }, + { + "epoch": 0.0551994301994302, + "grad_norm": 0.35244834423065186, + "learning_rate": 0.00019991762307241178, + "loss": 1.0864, + "step": 310 + }, + { + "epoch": 0.05537749287749288, + "grad_norm": 0.3865319490432739, + "learning_rate": 0.0001999170540548449, + "loss": 1.3659, + "step": 311 + }, + { + "epoch": 0.05555555555555555, + "grad_norm": 0.3666876554489136, + "learning_rate": 0.0001999164830796109, + "loss": 0.9884, + "step": 312 + }, + { + "epoch": 0.055733618233618235, + "grad_norm": 0.4278281629085541, + "learning_rate": 0.00019991591014672096, + "loss": 1.1522, + "step": 313 + }, + { + "epoch": 0.05591168091168091, + "grad_norm": 0.4172627031803131, + "learning_rate": 0.0001999153352561863, + "loss": 1.2527, + "step": 314 + }, + { + "epoch": 0.05608974358974359, + "grad_norm": 0.38872212171554565, + "learning_rate": 0.00019991475840801823, + "loss": 1.2985, + "step": 315 + }, + { + "epoch": 0.05626780626780627, + "grad_norm": 0.4160458445549011, + "learning_rate": 0.00019991417960222804, + "loss": 1.1347, + "step": 316 + }, + { + "epoch": 0.05644586894586895, + "grad_norm": 0.5169723033905029, + "learning_rate": 0.00019991359883882705, + "loss": 1.0819, + "step": 317 + }, + { + "epoch": 0.056623931623931624, + "grad_norm": 0.42306259274482727, + "learning_rate": 0.0001999130161178266, + "loss": 1.3139, + "step": 318 + }, + { + "epoch": 0.0568019943019943, + "grad_norm": 0.41975873708724976, + "learning_rate": 0.00019991243143923816, + "loss": 1.2277, + "step": 319 + }, + { + "epoch": 0.05698005698005698, + "grad_norm": 0.3873472511768341, + "learning_rate": 0.00019991184480307324, + "loss": 1.156, + "step": 320 + }, + { + "epoch": 0.057158119658119656, + "grad_norm": 0.43656104803085327, + "learning_rate": 0.0001999112562093432, + "loss": 1.2344, + "step": 321 + }, + { + "epoch": 0.05733618233618234, + "grad_norm": 0.3738791048526764, + "learning_rate": 0.00019991066565805968, + "loss": 0.9573, + "step": 322 + }, + { + "epoch": 0.05751424501424501, + "grad_norm": 0.3838156461715698, + "learning_rate": 0.00019991007314923418, + "loss": 0.9274, + "step": 323 + }, + { + "epoch": 0.057692307692307696, + "grad_norm": 0.4564770758152008, + "learning_rate": 0.00019990947868287837, + "loss": 1.0756, + "step": 324 + }, + { + "epoch": 0.05787037037037037, + "grad_norm": 0.4560079872608185, + "learning_rate": 0.00019990888225900386, + "loss": 1.1508, + "step": 325 + }, + { + "epoch": 0.058048433048433046, + "grad_norm": 0.44356057047843933, + "learning_rate": 0.00019990828387762236, + "loss": 1.2323, + "step": 326 + }, + { + "epoch": 0.05822649572649573, + "grad_norm": 0.46390119194984436, + "learning_rate": 0.00019990768353874553, + "loss": 1.0031, + "step": 327 + }, + { + "epoch": 0.0584045584045584, + "grad_norm": 0.4502357244491577, + "learning_rate": 0.00019990708124238525, + "loss": 1.3454, + "step": 328 + }, + { + "epoch": 0.058582621082621085, + "grad_norm": 0.3979945182800293, + "learning_rate": 0.0001999064769885532, + "loss": 1.2833, + "step": 329 + }, + { + "epoch": 0.05876068376068376, + "grad_norm": 0.3899286687374115, + "learning_rate": 0.00019990587077726128, + "loss": 1.0175, + "step": 330 + }, + { + "epoch": 0.05893874643874644, + "grad_norm": 0.41422948241233826, + "learning_rate": 0.00019990526260852139, + "loss": 1.1151, + "step": 331 + }, + { + "epoch": 0.05911680911680912, + "grad_norm": 0.4266608953475952, + "learning_rate": 0.0001999046524823454, + "loss": 1.1119, + "step": 332 + }, + { + "epoch": 0.05929487179487179, + "grad_norm": 0.46563324332237244, + "learning_rate": 0.00019990404039874524, + "loss": 1.2358, + "step": 333 + }, + { + "epoch": 0.059472934472934474, + "grad_norm": 0.4404347240924835, + "learning_rate": 0.00019990342635773297, + "loss": 1.1748, + "step": 334 + }, + { + "epoch": 0.05965099715099715, + "grad_norm": 0.5133237838745117, + "learning_rate": 0.00019990281035932062, + "loss": 1.1649, + "step": 335 + }, + { + "epoch": 0.05982905982905983, + "grad_norm": 0.3593895435333252, + "learning_rate": 0.00019990219240352018, + "loss": 1.0318, + "step": 336 + }, + { + "epoch": 0.06000712250712251, + "grad_norm": 0.40554583072662354, + "learning_rate": 0.00019990157249034384, + "loss": 1.1202, + "step": 337 + }, + { + "epoch": 0.06018518518518518, + "grad_norm": 0.3770706057548523, + "learning_rate": 0.00019990095061980372, + "loss": 0.9908, + "step": 338 + }, + { + "epoch": 0.060363247863247864, + "grad_norm": 0.39676955342292786, + "learning_rate": 0.000199900326791912, + "loss": 0.8176, + "step": 339 + }, + { + "epoch": 0.06054131054131054, + "grad_norm": 0.41448578238487244, + "learning_rate": 0.00019989970100668086, + "loss": 1.2877, + "step": 340 + }, + { + "epoch": 0.06071937321937322, + "grad_norm": 0.4200015068054199, + "learning_rate": 0.00019989907326412265, + "loss": 1.2293, + "step": 341 + }, + { + "epoch": 0.060897435897435896, + "grad_norm": 0.47350621223449707, + "learning_rate": 0.0001998984435642496, + "loss": 1.2331, + "step": 342 + }, + { + "epoch": 0.06107549857549858, + "grad_norm": 0.47050634026527405, + "learning_rate": 0.00019989781190707406, + "loss": 0.8888, + "step": 343 + }, + { + "epoch": 0.06125356125356125, + "grad_norm": 0.4994896948337555, + "learning_rate": 0.00019989717829260842, + "loss": 1.0921, + "step": 344 + }, + { + "epoch": 0.06143162393162393, + "grad_norm": 0.36340200901031494, + "learning_rate": 0.0001998965427208651, + "loss": 0.9777, + "step": 345 + }, + { + "epoch": 0.06160968660968661, + "grad_norm": 0.3538152873516083, + "learning_rate": 0.00019989590519185654, + "loss": 1.0055, + "step": 346 + }, + { + "epoch": 0.061787749287749286, + "grad_norm": 0.5388944149017334, + "learning_rate": 0.00019989526570559526, + "loss": 1.1001, + "step": 347 + }, + { + "epoch": 0.06196581196581197, + "grad_norm": 0.4411574602127075, + "learning_rate": 0.00019989462426209373, + "loss": 1.0038, + "step": 348 + }, + { + "epoch": 0.06214387464387464, + "grad_norm": 0.3930876851081848, + "learning_rate": 0.00019989398086136455, + "loss": 1.1534, + "step": 349 + }, + { + "epoch": 0.062321937321937325, + "grad_norm": 0.47357070446014404, + "learning_rate": 0.00019989333550342033, + "loss": 1.2687, + "step": 350 + }, + { + "epoch": 0.0625, + "grad_norm": 0.40302303433418274, + "learning_rate": 0.00019989268818827372, + "loss": 1.1894, + "step": 351 + }, + { + "epoch": 0.06267806267806268, + "grad_norm": 0.4470510184764862, + "learning_rate": 0.00019989203891593738, + "loss": 1.2207, + "step": 352 + }, + { + "epoch": 0.06285612535612535, + "grad_norm": 0.42235100269317627, + "learning_rate": 0.00019989138768642406, + "loss": 1.2086, + "step": 353 + }, + { + "epoch": 0.06303418803418803, + "grad_norm": 0.38305309414863586, + "learning_rate": 0.0001998907344997465, + "loss": 1.0473, + "step": 354 + }, + { + "epoch": 0.06321225071225071, + "grad_norm": 0.3893027901649475, + "learning_rate": 0.0001998900793559175, + "loss": 1.1746, + "step": 355 + }, + { + "epoch": 0.0633903133903134, + "grad_norm": 0.41206735372543335, + "learning_rate": 0.0001998894222549499, + "loss": 1.188, + "step": 356 + }, + { + "epoch": 0.06356837606837606, + "grad_norm": 0.3700513243675232, + "learning_rate": 0.00019988876319685658, + "loss": 0.9862, + "step": 357 + }, + { + "epoch": 0.06374643874643875, + "grad_norm": 0.3708794116973877, + "learning_rate": 0.0001998881021816504, + "loss": 1.2003, + "step": 358 + }, + { + "epoch": 0.06392450142450143, + "grad_norm": 0.4058014154434204, + "learning_rate": 0.00019988743920934442, + "loss": 1.2311, + "step": 359 + }, + { + "epoch": 0.0641025641025641, + "grad_norm": 0.39134132862091064, + "learning_rate": 0.00019988677427995155, + "loss": 1.001, + "step": 360 + }, + { + "epoch": 0.06428062678062678, + "grad_norm": 0.3853437602519989, + "learning_rate": 0.00019988610739348484, + "loss": 1.0725, + "step": 361 + }, + { + "epoch": 0.06445868945868946, + "grad_norm": 0.47114330530166626, + "learning_rate": 0.00019988543854995735, + "loss": 1.2196, + "step": 362 + }, + { + "epoch": 0.06463675213675214, + "grad_norm": 0.40465688705444336, + "learning_rate": 0.00019988476774938216, + "loss": 1.1869, + "step": 363 + }, + { + "epoch": 0.06481481481481481, + "grad_norm": 0.40301886200904846, + "learning_rate": 0.00019988409499177245, + "loss": 1.1765, + "step": 364 + }, + { + "epoch": 0.0649928774928775, + "grad_norm": 0.43443185091018677, + "learning_rate": 0.0001998834202771414, + "loss": 1.2022, + "step": 365 + }, + { + "epoch": 0.06517094017094018, + "grad_norm": 0.4712986350059509, + "learning_rate": 0.00019988274360550217, + "loss": 1.156, + "step": 366 + }, + { + "epoch": 0.06534900284900284, + "grad_norm": 0.4524450898170471, + "learning_rate": 0.00019988206497686815, + "loss": 1.2917, + "step": 367 + }, + { + "epoch": 0.06552706552706553, + "grad_norm": 0.40302205085754395, + "learning_rate": 0.0001998813843912525, + "loss": 0.9993, + "step": 368 + }, + { + "epoch": 0.06570512820512821, + "grad_norm": 0.39435216784477234, + "learning_rate": 0.00019988070184866864, + "loss": 1.0914, + "step": 369 + }, + { + "epoch": 0.06588319088319089, + "grad_norm": 0.39267390966415405, + "learning_rate": 0.00019988001734912988, + "loss": 1.3138, + "step": 370 + }, + { + "epoch": 0.06606125356125356, + "grad_norm": 0.38351675868034363, + "learning_rate": 0.00019987933089264968, + "loss": 1.0997, + "step": 371 + }, + { + "epoch": 0.06623931623931624, + "grad_norm": 0.3294839859008789, + "learning_rate": 0.00019987864247924145, + "loss": 0.9656, + "step": 372 + }, + { + "epoch": 0.06641737891737892, + "grad_norm": 0.45333364605903625, + "learning_rate": 0.00019987795210891872, + "loss": 1.095, + "step": 373 + }, + { + "epoch": 0.06659544159544159, + "grad_norm": 0.4362282454967499, + "learning_rate": 0.00019987725978169501, + "loss": 1.2103, + "step": 374 + }, + { + "epoch": 0.06677350427350427, + "grad_norm": 0.41314780712127686, + "learning_rate": 0.00019987656549758385, + "loss": 1.2115, + "step": 375 + }, + { + "epoch": 0.06695156695156695, + "grad_norm": 0.4230864644050598, + "learning_rate": 0.00019987586925659888, + "loss": 1.17, + "step": 376 + }, + { + "epoch": 0.06712962962962964, + "grad_norm": 0.4703855812549591, + "learning_rate": 0.00019987517105875372, + "loss": 1.367, + "step": 377 + }, + { + "epoch": 0.0673076923076923, + "grad_norm": 0.4671297073364258, + "learning_rate": 0.00019987447090406206, + "loss": 1.2543, + "step": 378 + }, + { + "epoch": 0.06748575498575499, + "grad_norm": 0.43746981024742126, + "learning_rate": 0.0001998737687925376, + "loss": 1.214, + "step": 379 + }, + { + "epoch": 0.06766381766381767, + "grad_norm": 0.40889596939086914, + "learning_rate": 0.00019987306472419412, + "loss": 1.0496, + "step": 380 + }, + { + "epoch": 0.06784188034188034, + "grad_norm": 0.3677358627319336, + "learning_rate": 0.0001998723586990454, + "loss": 1.1242, + "step": 381 + }, + { + "epoch": 0.06801994301994302, + "grad_norm": 0.3892628848552704, + "learning_rate": 0.00019987165071710527, + "loss": 1.0246, + "step": 382 + }, + { + "epoch": 0.0681980056980057, + "grad_norm": 0.4281293749809265, + "learning_rate": 0.00019987094077838764, + "loss": 1.2817, + "step": 383 + }, + { + "epoch": 0.06837606837606838, + "grad_norm": 0.45030340552330017, + "learning_rate": 0.00019987022888290636, + "loss": 1.159, + "step": 384 + }, + { + "epoch": 0.06855413105413105, + "grad_norm": 0.6327905058860779, + "learning_rate": 0.00019986951503067545, + "loss": 0.9577, + "step": 385 + }, + { + "epoch": 0.06873219373219373, + "grad_norm": 0.40339627861976624, + "learning_rate": 0.0001998687992217088, + "loss": 1.138, + "step": 386 + }, + { + "epoch": 0.06891025641025642, + "grad_norm": 0.4018291234970093, + "learning_rate": 0.00019986808145602052, + "loss": 0.9109, + "step": 387 + }, + { + "epoch": 0.06908831908831908, + "grad_norm": 0.41566264629364014, + "learning_rate": 0.00019986736173362464, + "loss": 1.1516, + "step": 388 + }, + { + "epoch": 0.06926638176638177, + "grad_norm": 0.3569067418575287, + "learning_rate": 0.00019986664005453527, + "loss": 1.2329, + "step": 389 + }, + { + "epoch": 0.06944444444444445, + "grad_norm": 0.3959648907184601, + "learning_rate": 0.0001998659164187665, + "loss": 1.1041, + "step": 390 + }, + { + "epoch": 0.06962250712250712, + "grad_norm": 0.42853206396102905, + "learning_rate": 0.00019986519082633257, + "loss": 1.0859, + "step": 391 + }, + { + "epoch": 0.0698005698005698, + "grad_norm": 0.42005518078804016, + "learning_rate": 0.0001998644632772477, + "loss": 1.2017, + "step": 392 + }, + { + "epoch": 0.06997863247863248, + "grad_norm": 0.4296947419643402, + "learning_rate": 0.00019986373377152612, + "loss": 1.1464, + "step": 393 + }, + { + "epoch": 0.07015669515669516, + "grad_norm": 0.394747793674469, + "learning_rate": 0.0001998630023091821, + "loss": 1.0316, + "step": 394 + }, + { + "epoch": 0.07033475783475783, + "grad_norm": 0.3779357969760895, + "learning_rate": 0.00019986226889023002, + "loss": 1.1081, + "step": 395 + }, + { + "epoch": 0.07051282051282051, + "grad_norm": 0.4271804690361023, + "learning_rate": 0.00019986153351468424, + "loss": 0.985, + "step": 396 + }, + { + "epoch": 0.0706908831908832, + "grad_norm": 0.49412235617637634, + "learning_rate": 0.00019986079618255912, + "loss": 1.2606, + "step": 397 + }, + { + "epoch": 0.07086894586894586, + "grad_norm": 0.43657439947128296, + "learning_rate": 0.00019986005689386915, + "loss": 1.2266, + "step": 398 + }, + { + "epoch": 0.07104700854700854, + "grad_norm": 0.4060729444026947, + "learning_rate": 0.0001998593156486288, + "loss": 1.1787, + "step": 399 + }, + { + "epoch": 0.07122507122507123, + "grad_norm": 0.387046217918396, + "learning_rate": 0.00019985857244685264, + "loss": 0.9411, + "step": 400 + }, + { + "epoch": 0.07140313390313391, + "grad_norm": 0.4243999123573303, + "learning_rate": 0.00019985782728855516, + "loss": 1.2024, + "step": 401 + }, + { + "epoch": 0.07158119658119658, + "grad_norm": 0.43113812804222107, + "learning_rate": 0.000199857080173751, + "loss": 1.1246, + "step": 402 + }, + { + "epoch": 0.07175925925925926, + "grad_norm": 0.4653271436691284, + "learning_rate": 0.0001998563311024548, + "loss": 1.2343, + "step": 403 + }, + { + "epoch": 0.07193732193732194, + "grad_norm": 0.43260812759399414, + "learning_rate": 0.0001998555800746812, + "loss": 0.9543, + "step": 404 + }, + { + "epoch": 0.07211538461538461, + "grad_norm": 0.4635484516620636, + "learning_rate": 0.00019985482709044495, + "loss": 1.1091, + "step": 405 + }, + { + "epoch": 0.07229344729344729, + "grad_norm": 0.38362643122673035, + "learning_rate": 0.00019985407214976076, + "loss": 1.2584, + "step": 406 + }, + { + "epoch": 0.07247150997150997, + "grad_norm": 0.4068310558795929, + "learning_rate": 0.00019985331525264351, + "loss": 1.1944, + "step": 407 + }, + { + "epoch": 0.07264957264957266, + "grad_norm": 0.43909943103790283, + "learning_rate": 0.00019985255639910795, + "loss": 1.3748, + "step": 408 + }, + { + "epoch": 0.07282763532763532, + "grad_norm": 0.48674601316452026, + "learning_rate": 0.000199851795589169, + "loss": 1.2684, + "step": 409 + }, + { + "epoch": 0.073005698005698, + "grad_norm": 0.4218580722808838, + "learning_rate": 0.0001998510328228415, + "loss": 1.168, + "step": 410 + }, + { + "epoch": 0.07318376068376069, + "grad_norm": 0.4688236117362976, + "learning_rate": 0.00019985026810014046, + "loss": 1.3088, + "step": 411 + }, + { + "epoch": 0.07336182336182336, + "grad_norm": 0.3863612711429596, + "learning_rate": 0.00019984950142108083, + "loss": 1.0261, + "step": 412 + }, + { + "epoch": 0.07353988603988604, + "grad_norm": 0.4177640378475189, + "learning_rate": 0.00019984873278567765, + "loss": 1.1985, + "step": 413 + }, + { + "epoch": 0.07371794871794872, + "grad_norm": 0.4645586311817169, + "learning_rate": 0.00019984796219394592, + "loss": 1.2463, + "step": 414 + }, + { + "epoch": 0.0738960113960114, + "grad_norm": 0.5051766633987427, + "learning_rate": 0.00019984718964590083, + "loss": 1.3031, + "step": 415 + }, + { + "epoch": 0.07407407407407407, + "grad_norm": 0.4200040400028229, + "learning_rate": 0.0001998464151415575, + "loss": 1.0842, + "step": 416 + }, + { + "epoch": 0.07425213675213675, + "grad_norm": 0.34211036562919617, + "learning_rate": 0.000199845638680931, + "loss": 0.9659, + "step": 417 + }, + { + "epoch": 0.07443019943019943, + "grad_norm": 0.3553323447704315, + "learning_rate": 0.00019984486026403668, + "loss": 1.0102, + "step": 418 + }, + { + "epoch": 0.0746082621082621, + "grad_norm": 0.4967300295829773, + "learning_rate": 0.00019984407989088974, + "loss": 1.3125, + "step": 419 + }, + { + "epoch": 0.07478632478632478, + "grad_norm": 0.41649797558784485, + "learning_rate": 0.00019984329756150544, + "loss": 1.3092, + "step": 420 + }, + { + "epoch": 0.07496438746438747, + "grad_norm": 0.43825802206993103, + "learning_rate": 0.00019984251327589912, + "loss": 1.3678, + "step": 421 + }, + { + "epoch": 0.07514245014245015, + "grad_norm": 0.363394170999527, + "learning_rate": 0.00019984172703408617, + "loss": 1.305, + "step": 422 + }, + { + "epoch": 0.07532051282051282, + "grad_norm": 0.411563903093338, + "learning_rate": 0.000199840938836082, + "loss": 1.4248, + "step": 423 + }, + { + "epoch": 0.0754985754985755, + "grad_norm": 0.40548190474510193, + "learning_rate": 0.000199840148681902, + "loss": 1.1081, + "step": 424 + }, + { + "epoch": 0.07567663817663818, + "grad_norm": 0.3781099021434784, + "learning_rate": 0.00019983935657156171, + "loss": 1.185, + "step": 425 + }, + { + "epoch": 0.07585470085470085, + "grad_norm": 0.46597573161125183, + "learning_rate": 0.00019983856250507662, + "loss": 1.119, + "step": 426 + }, + { + "epoch": 0.07603276353276353, + "grad_norm": 0.3988197147846222, + "learning_rate": 0.00019983776648246232, + "loss": 1.206, + "step": 427 + }, + { + "epoch": 0.07621082621082621, + "grad_norm": 0.41210901737213135, + "learning_rate": 0.00019983696850373433, + "loss": 1.1843, + "step": 428 + }, + { + "epoch": 0.0763888888888889, + "grad_norm": 0.41870948672294617, + "learning_rate": 0.00019983616856890837, + "loss": 1.2248, + "step": 429 + }, + { + "epoch": 0.07656695156695156, + "grad_norm": 0.4320056140422821, + "learning_rate": 0.00019983536667800007, + "loss": 0.9743, + "step": 430 + }, + { + "epoch": 0.07674501424501425, + "grad_norm": 0.48455503582954407, + "learning_rate": 0.00019983456283102517, + "loss": 1.0438, + "step": 431 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 0.38712427020072937, + "learning_rate": 0.00019983375702799935, + "loss": 1.2041, + "step": 432 + }, + { + "epoch": 0.0771011396011396, + "grad_norm": 0.3578857481479645, + "learning_rate": 0.0001998329492689385, + "loss": 1.1623, + "step": 433 + }, + { + "epoch": 0.07727920227920228, + "grad_norm": 0.43065932393074036, + "learning_rate": 0.00019983213955385834, + "loss": 1.3033, + "step": 434 + }, + { + "epoch": 0.07745726495726496, + "grad_norm": 0.4882095754146576, + "learning_rate": 0.00019983132788277484, + "loss": 1.1635, + "step": 435 + }, + { + "epoch": 0.07763532763532764, + "grad_norm": 0.3429015874862671, + "learning_rate": 0.00019983051425570382, + "loss": 0.7289, + "step": 436 + }, + { + "epoch": 0.07781339031339031, + "grad_norm": 0.4320310056209564, + "learning_rate": 0.00019982969867266128, + "loss": 1.3685, + "step": 437 + }, + { + "epoch": 0.07799145299145299, + "grad_norm": 0.39891982078552246, + "learning_rate": 0.00019982888113366314, + "loss": 1.0444, + "step": 438 + }, + { + "epoch": 0.07816951566951567, + "grad_norm": 0.3675695061683655, + "learning_rate": 0.00019982806163872547, + "loss": 1.0527, + "step": 439 + }, + { + "epoch": 0.07834757834757834, + "grad_norm": 0.42824694514274597, + "learning_rate": 0.0001998272401878643, + "loss": 1.166, + "step": 440 + }, + { + "epoch": 0.07852564102564102, + "grad_norm": 0.3721694350242615, + "learning_rate": 0.00019982641678109575, + "loss": 1.1328, + "step": 441 + }, + { + "epoch": 0.0787037037037037, + "grad_norm": 0.33899208903312683, + "learning_rate": 0.00019982559141843592, + "loss": 1.016, + "step": 442 + }, + { + "epoch": 0.07888176638176639, + "grad_norm": 0.4029340147972107, + "learning_rate": 0.000199824764099901, + "loss": 1.0076, + "step": 443 + }, + { + "epoch": 0.07905982905982906, + "grad_norm": 0.4169132113456726, + "learning_rate": 0.0001998239348255072, + "loss": 1.208, + "step": 444 + }, + { + "epoch": 0.07923789173789174, + "grad_norm": 0.3865824043750763, + "learning_rate": 0.00019982310359527075, + "loss": 1.067, + "step": 445 + }, + { + "epoch": 0.07941595441595442, + "grad_norm": 0.4218919277191162, + "learning_rate": 0.00019982227040920796, + "loss": 1.195, + "step": 446 + }, + { + "epoch": 0.07959401709401709, + "grad_norm": 0.40504586696624756, + "learning_rate": 0.00019982143526733512, + "loss": 1.0188, + "step": 447 + }, + { + "epoch": 0.07977207977207977, + "grad_norm": 0.38330578804016113, + "learning_rate": 0.00019982059816966863, + "loss": 1.0484, + "step": 448 + }, + { + "epoch": 0.07995014245014245, + "grad_norm": 0.43731689453125, + "learning_rate": 0.00019981975911622488, + "loss": 1.074, + "step": 449 + }, + { + "epoch": 0.08012820512820513, + "grad_norm": 0.40858447551727295, + "learning_rate": 0.00019981891810702033, + "loss": 1.0008, + "step": 450 + }, + { + "epoch": 0.0803062678062678, + "grad_norm": 0.4031754732131958, + "learning_rate": 0.00019981807514207143, + "loss": 1.2179, + "step": 451 + }, + { + "epoch": 0.08048433048433049, + "grad_norm": 0.41920867562294006, + "learning_rate": 0.00019981723022139466, + "loss": 1.1406, + "step": 452 + }, + { + "epoch": 0.08066239316239317, + "grad_norm": 0.40305474400520325, + "learning_rate": 0.00019981638334500668, + "loss": 1.098, + "step": 453 + }, + { + "epoch": 0.08084045584045584, + "grad_norm": 0.4564182460308075, + "learning_rate": 0.00019981553451292396, + "loss": 1.419, + "step": 454 + }, + { + "epoch": 0.08101851851851852, + "grad_norm": 0.3832945227622986, + "learning_rate": 0.00019981468372516322, + "loss": 1.0919, + "step": 455 + }, + { + "epoch": 0.0811965811965812, + "grad_norm": 0.43062624335289, + "learning_rate": 0.0001998138309817411, + "loss": 1.0458, + "step": 456 + }, + { + "epoch": 0.08137464387464387, + "grad_norm": 0.3871173560619354, + "learning_rate": 0.0001998129762826743, + "loss": 1.1391, + "step": 457 + }, + { + "epoch": 0.08155270655270655, + "grad_norm": 0.43423157930374146, + "learning_rate": 0.0001998121196279796, + "loss": 1.1132, + "step": 458 + }, + { + "epoch": 0.08173076923076923, + "grad_norm": 0.4341012239456177, + "learning_rate": 0.00019981126101767372, + "loss": 1.113, + "step": 459 + }, + { + "epoch": 0.08190883190883191, + "grad_norm": 0.36748576164245605, + "learning_rate": 0.00019981040045177352, + "loss": 0.8108, + "step": 460 + }, + { + "epoch": 0.08208689458689458, + "grad_norm": 0.43133220076560974, + "learning_rate": 0.00019980953793029586, + "loss": 1.1861, + "step": 461 + }, + { + "epoch": 0.08226495726495726, + "grad_norm": 0.37204909324645996, + "learning_rate": 0.00019980867345325767, + "loss": 0.9222, + "step": 462 + }, + { + "epoch": 0.08244301994301995, + "grad_norm": 0.43370047211647034, + "learning_rate": 0.00019980780702067582, + "loss": 1.2984, + "step": 463 + }, + { + "epoch": 0.08262108262108261, + "grad_norm": 0.4991510808467865, + "learning_rate": 0.00019980693863256736, + "loss": 1.2222, + "step": 464 + }, + { + "epoch": 0.0827991452991453, + "grad_norm": 0.44318175315856934, + "learning_rate": 0.00019980606828894927, + "loss": 1.2262, + "step": 465 + }, + { + "epoch": 0.08297720797720798, + "grad_norm": 0.380231648683548, + "learning_rate": 0.0001998051959898386, + "loss": 1.0274, + "step": 466 + }, + { + "epoch": 0.08315527065527066, + "grad_norm": 0.39519667625427246, + "learning_rate": 0.0001998043217352524, + "loss": 1.2499, + "step": 467 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.457499235868454, + "learning_rate": 0.0001998034455252079, + "loss": 1.0751, + "step": 468 + }, + { + "epoch": 0.08351139601139601, + "grad_norm": 0.368522584438324, + "learning_rate": 0.00019980256735972215, + "loss": 1.0776, + "step": 469 + }, + { + "epoch": 0.08368945868945869, + "grad_norm": 0.3768427073955536, + "learning_rate": 0.00019980168723881243, + "loss": 1.2198, + "step": 470 + }, + { + "epoch": 0.08386752136752136, + "grad_norm": 0.37045565247535706, + "learning_rate": 0.000199800805162496, + "loss": 1.1816, + "step": 471 + }, + { + "epoch": 0.08404558404558404, + "grad_norm": 0.4219281077384949, + "learning_rate": 0.0001997999211307901, + "loss": 1.0515, + "step": 472 + }, + { + "epoch": 0.08422364672364673, + "grad_norm": 0.3815271258354187, + "learning_rate": 0.00019979903514371207, + "loss": 1.1709, + "step": 473 + }, + { + "epoch": 0.08440170940170941, + "grad_norm": 0.4566493630409241, + "learning_rate": 0.00019979814720127924, + "loss": 1.3063, + "step": 474 + }, + { + "epoch": 0.08457977207977208, + "grad_norm": 0.4043879806995392, + "learning_rate": 0.000199797257303509, + "loss": 1.0549, + "step": 475 + }, + { + "epoch": 0.08475783475783476, + "grad_norm": 0.3897830545902252, + "learning_rate": 0.00019979636545041886, + "loss": 1.1483, + "step": 476 + }, + { + "epoch": 0.08493589743589744, + "grad_norm": 0.36097025871276855, + "learning_rate": 0.00019979547164202622, + "loss": 1.1196, + "step": 477 + }, + { + "epoch": 0.08511396011396011, + "grad_norm": 0.3766986131668091, + "learning_rate": 0.00019979457587834863, + "loss": 1.0131, + "step": 478 + }, + { + "epoch": 0.08529202279202279, + "grad_norm": 0.39460286498069763, + "learning_rate": 0.00019979367815940364, + "loss": 1.1729, + "step": 479 + }, + { + "epoch": 0.08547008547008547, + "grad_norm": 0.4137469232082367, + "learning_rate": 0.00019979277848520885, + "loss": 1.2569, + "step": 480 + }, + { + "epoch": 0.08564814814814815, + "grad_norm": 0.464688777923584, + "learning_rate": 0.00019979187685578183, + "loss": 1.2064, + "step": 481 + }, + { + "epoch": 0.08582621082621082, + "grad_norm": 0.4245518147945404, + "learning_rate": 0.0001997909732711403, + "loss": 0.9812, + "step": 482 + }, + { + "epoch": 0.0860042735042735, + "grad_norm": 0.43368837237358093, + "learning_rate": 0.00019979006773130197, + "loss": 1.2822, + "step": 483 + }, + { + "epoch": 0.08618233618233619, + "grad_norm": 0.4232824444770813, + "learning_rate": 0.00019978916023628452, + "loss": 1.1446, + "step": 484 + }, + { + "epoch": 0.08636039886039885, + "grad_norm": 0.4183506369590759, + "learning_rate": 0.00019978825078610578, + "loss": 1.2605, + "step": 485 + }, + { + "epoch": 0.08653846153846154, + "grad_norm": 0.4391268491744995, + "learning_rate": 0.00019978733938078356, + "loss": 1.2165, + "step": 486 + }, + { + "epoch": 0.08671652421652422, + "grad_norm": 0.4139612317085266, + "learning_rate": 0.0001997864260203357, + "loss": 0.9389, + "step": 487 + }, + { + "epoch": 0.0868945868945869, + "grad_norm": 0.4058656096458435, + "learning_rate": 0.00019978551070478013, + "loss": 1.0652, + "step": 488 + }, + { + "epoch": 0.08707264957264957, + "grad_norm": 0.42333099246025085, + "learning_rate": 0.00019978459343413473, + "loss": 1.119, + "step": 489 + }, + { + "epoch": 0.08725071225071225, + "grad_norm": 0.4573031961917877, + "learning_rate": 0.00019978367420841754, + "loss": 1.1546, + "step": 490 + }, + { + "epoch": 0.08742877492877493, + "grad_norm": 0.4161617159843445, + "learning_rate": 0.00019978275302764655, + "loss": 1.0836, + "step": 491 + }, + { + "epoch": 0.0876068376068376, + "grad_norm": 0.422145277261734, + "learning_rate": 0.00019978182989183977, + "loss": 1.1908, + "step": 492 + }, + { + "epoch": 0.08778490028490028, + "grad_norm": 0.4588126838207245, + "learning_rate": 0.00019978090480101532, + "loss": 1.1758, + "step": 493 + }, + { + "epoch": 0.08796296296296297, + "grad_norm": 0.4425722062587738, + "learning_rate": 0.00019977997775519132, + "loss": 1.088, + "step": 494 + }, + { + "epoch": 0.08814102564102565, + "grad_norm": 0.37860307097435, + "learning_rate": 0.00019977904875438594, + "loss": 1.1532, + "step": 495 + }, + { + "epoch": 0.08831908831908832, + "grad_norm": 0.40435823798179626, + "learning_rate": 0.00019977811779861733, + "loss": 1.1271, + "step": 496 + }, + { + "epoch": 0.088497150997151, + "grad_norm": 0.42578884959220886, + "learning_rate": 0.0001997771848879038, + "loss": 0.9889, + "step": 497 + }, + { + "epoch": 0.08867521367521368, + "grad_norm": 0.3439478874206543, + "learning_rate": 0.00019977625002226361, + "loss": 1.1273, + "step": 498 + }, + { + "epoch": 0.08885327635327635, + "grad_norm": 0.362341970205307, + "learning_rate": 0.00019977531320171504, + "loss": 1.0214, + "step": 499 + }, + { + "epoch": 0.08903133903133903, + "grad_norm": 0.4305768609046936, + "learning_rate": 0.0001997743744262765, + "loss": 1.2648, + "step": 500 + }, + { + "epoch": 0.08920940170940171, + "grad_norm": 0.35900023579597473, + "learning_rate": 0.00019977343369596636, + "loss": 1.0274, + "step": 501 + }, + { + "epoch": 0.0893874643874644, + "grad_norm": 0.4950818717479706, + "learning_rate": 0.00019977249101080306, + "loss": 1.1483, + "step": 502 + }, + { + "epoch": 0.08956552706552706, + "grad_norm": 0.3800346553325653, + "learning_rate": 0.00019977154637080503, + "loss": 1.0636, + "step": 503 + }, + { + "epoch": 0.08974358974358974, + "grad_norm": 0.46202352643013, + "learning_rate": 0.0001997705997759908, + "loss": 1.1544, + "step": 504 + }, + { + "epoch": 0.08992165242165243, + "grad_norm": 0.36818403005599976, + "learning_rate": 0.00019976965122637895, + "loss": 0.9824, + "step": 505 + }, + { + "epoch": 0.0900997150997151, + "grad_norm": 0.40248095989227295, + "learning_rate": 0.00019976870072198805, + "loss": 1.1002, + "step": 506 + }, + { + "epoch": 0.09027777777777778, + "grad_norm": 0.3841850459575653, + "learning_rate": 0.00019976774826283667, + "loss": 1.2433, + "step": 507 + }, + { + "epoch": 0.09045584045584046, + "grad_norm": 0.46892330050468445, + "learning_rate": 0.0001997667938489435, + "loss": 1.3194, + "step": 508 + }, + { + "epoch": 0.09063390313390314, + "grad_norm": 0.39059561491012573, + "learning_rate": 0.0001997658374803273, + "loss": 1.1778, + "step": 509 + }, + { + "epoch": 0.09081196581196581, + "grad_norm": 0.3793235421180725, + "learning_rate": 0.00019976487915700672, + "loss": 1.0659, + "step": 510 + }, + { + "epoch": 0.09099002849002849, + "grad_norm": 0.39067742228507996, + "learning_rate": 0.00019976391887900058, + "loss": 1.107, + "step": 511 + }, + { + "epoch": 0.09116809116809117, + "grad_norm": 0.40121713280677795, + "learning_rate": 0.00019976295664632772, + "loss": 1.102, + "step": 512 + }, + { + "epoch": 0.09134615384615384, + "grad_norm": 0.49830010533332825, + "learning_rate": 0.00019976199245900697, + "loss": 1.1701, + "step": 513 + }, + { + "epoch": 0.09152421652421652, + "grad_norm": 0.4536968171596527, + "learning_rate": 0.0001997610263170572, + "loss": 1.1067, + "step": 514 + }, + { + "epoch": 0.0917022792022792, + "grad_norm": 0.3832971453666687, + "learning_rate": 0.00019976005822049735, + "loss": 1.0991, + "step": 515 + }, + { + "epoch": 0.09188034188034189, + "grad_norm": 0.4093509614467621, + "learning_rate": 0.0001997590881693464, + "loss": 1.0565, + "step": 516 + }, + { + "epoch": 0.09205840455840456, + "grad_norm": 0.46073687076568604, + "learning_rate": 0.0001997581161636233, + "loss": 1.0057, + "step": 517 + }, + { + "epoch": 0.09223646723646724, + "grad_norm": 0.5001922845840454, + "learning_rate": 0.0001997571422033472, + "loss": 1.2639, + "step": 518 + }, + { + "epoch": 0.09241452991452992, + "grad_norm": 0.4620618224143982, + "learning_rate": 0.00019975616628853713, + "loss": 1.0966, + "step": 519 + }, + { + "epoch": 0.09259259259259259, + "grad_norm": 0.3788183927536011, + "learning_rate": 0.0001997551884192122, + "loss": 0.9783, + "step": 520 + }, + { + "epoch": 0.09277065527065527, + "grad_norm": 0.45589539408683777, + "learning_rate": 0.00019975420859539154, + "loss": 1.2194, + "step": 521 + }, + { + "epoch": 0.09294871794871795, + "grad_norm": 0.40747523307800293, + "learning_rate": 0.00019975322681709443, + "loss": 1.0349, + "step": 522 + }, + { + "epoch": 0.09312678062678063, + "grad_norm": 0.5045142769813538, + "learning_rate": 0.00019975224308434002, + "loss": 1.1373, + "step": 523 + }, + { + "epoch": 0.0933048433048433, + "grad_norm": 0.40352702140808105, + "learning_rate": 0.00019975125739714767, + "loss": 1.1236, + "step": 524 + }, + { + "epoch": 0.09348290598290598, + "grad_norm": 0.4301735758781433, + "learning_rate": 0.0001997502697555366, + "loss": 1.2932, + "step": 525 + }, + { + "epoch": 0.09366096866096867, + "grad_norm": 0.36800238490104675, + "learning_rate": 0.00019974928015952624, + "loss": 1.0734, + "step": 526 + }, + { + "epoch": 0.09383903133903133, + "grad_norm": 0.4027230143547058, + "learning_rate": 0.00019974828860913594, + "loss": 1.2776, + "step": 527 + }, + { + "epoch": 0.09401709401709402, + "grad_norm": 0.42497140169143677, + "learning_rate": 0.0001997472951043851, + "loss": 1.248, + "step": 528 + }, + { + "epoch": 0.0941951566951567, + "grad_norm": 0.3888593018054962, + "learning_rate": 0.00019974629964529325, + "loss": 1.0231, + "step": 529 + }, + { + "epoch": 0.09437321937321937, + "grad_norm": 0.3761361241340637, + "learning_rate": 0.00019974530223187986, + "loss": 1.0216, + "step": 530 + }, + { + "epoch": 0.09455128205128205, + "grad_norm": 0.42192980647087097, + "learning_rate": 0.00019974430286416448, + "loss": 1.0731, + "step": 531 + }, + { + "epoch": 0.09472934472934473, + "grad_norm": 0.44244512915611267, + "learning_rate": 0.00019974330154216667, + "loss": 1.2793, + "step": 532 + }, + { + "epoch": 0.09490740740740741, + "grad_norm": 0.378252774477005, + "learning_rate": 0.0001997422982659061, + "loss": 1.0462, + "step": 533 + }, + { + "epoch": 0.09508547008547008, + "grad_norm": 0.45589110255241394, + "learning_rate": 0.00019974129303540236, + "loss": 1.1884, + "step": 534 + }, + { + "epoch": 0.09526353276353276, + "grad_norm": 0.33930808305740356, + "learning_rate": 0.0001997402858506752, + "loss": 0.8381, + "step": 535 + }, + { + "epoch": 0.09544159544159544, + "grad_norm": 0.45408427715301514, + "learning_rate": 0.0001997392767117443, + "loss": 1.2379, + "step": 536 + }, + { + "epoch": 0.09561965811965811, + "grad_norm": 0.44125741720199585, + "learning_rate": 0.0001997382656186295, + "loss": 1.1941, + "step": 537 + }, + { + "epoch": 0.0957977207977208, + "grad_norm": 0.4075697660446167, + "learning_rate": 0.00019973725257135054, + "loss": 1.0142, + "step": 538 + }, + { + "epoch": 0.09597578347578348, + "grad_norm": 0.4258415102958679, + "learning_rate": 0.00019973623756992733, + "loss": 1.0447, + "step": 539 + }, + { + "epoch": 0.09615384615384616, + "grad_norm": 0.2738485038280487, + "learning_rate": 0.0001997352206143797, + "loss": 0.5521, + "step": 540 + }, + { + "epoch": 0.09633190883190883, + "grad_norm": 0.38815587759017944, + "learning_rate": 0.00019973420170472762, + "loss": 1.1052, + "step": 541 + }, + { + "epoch": 0.09650997150997151, + "grad_norm": 0.3909834027290344, + "learning_rate": 0.00019973318084099106, + "loss": 1.0494, + "step": 542 + }, + { + "epoch": 0.09668803418803419, + "grad_norm": 0.4517597258090973, + "learning_rate": 0.00019973215802318996, + "loss": 1.0611, + "step": 543 + }, + { + "epoch": 0.09686609686609686, + "grad_norm": 0.48659002780914307, + "learning_rate": 0.00019973113325134442, + "loss": 0.9967, + "step": 544 + }, + { + "epoch": 0.09704415954415954, + "grad_norm": 0.4039791524410248, + "learning_rate": 0.0001997301065254745, + "loss": 1.251, + "step": 545 + }, + { + "epoch": 0.09722222222222222, + "grad_norm": 0.3985383212566376, + "learning_rate": 0.0001997290778456003, + "loss": 1.2263, + "step": 546 + }, + { + "epoch": 0.0974002849002849, + "grad_norm": 0.4540637731552124, + "learning_rate": 0.00019972804721174199, + "loss": 1.2084, + "step": 547 + }, + { + "epoch": 0.09757834757834757, + "grad_norm": 0.36867982149124146, + "learning_rate": 0.00019972701462391977, + "loss": 0.9704, + "step": 548 + }, + { + "epoch": 0.09775641025641026, + "grad_norm": 0.40199780464172363, + "learning_rate": 0.00019972598008215385, + "loss": 1.1121, + "step": 549 + }, + { + "epoch": 0.09793447293447294, + "grad_norm": 0.42728984355926514, + "learning_rate": 0.00019972494358646455, + "loss": 1.1606, + "step": 550 + }, + { + "epoch": 0.0981125356125356, + "grad_norm": 0.4212374687194824, + "learning_rate": 0.0001997239051368721, + "loss": 1.3093, + "step": 551 + }, + { + "epoch": 0.09829059829059829, + "grad_norm": 0.3972226083278656, + "learning_rate": 0.0001997228647333969, + "loss": 1.1218, + "step": 552 + }, + { + "epoch": 0.09846866096866097, + "grad_norm": 0.43649932742118835, + "learning_rate": 0.00019972182237605935, + "loss": 1.2532, + "step": 553 + }, + { + "epoch": 0.09864672364672365, + "grad_norm": 0.3812280595302582, + "learning_rate": 0.0001997207780648798, + "loss": 1.0409, + "step": 554 + }, + { + "epoch": 0.09882478632478632, + "grad_norm": 0.41684821248054504, + "learning_rate": 0.00019971973179987878, + "loss": 0.9569, + "step": 555 + }, + { + "epoch": 0.099002849002849, + "grad_norm": 0.38081470131874084, + "learning_rate": 0.00019971868358107674, + "loss": 1.1615, + "step": 556 + }, + { + "epoch": 0.09918091168091168, + "grad_norm": 0.3702073097229004, + "learning_rate": 0.0001997176334084943, + "loss": 1.3907, + "step": 557 + }, + { + "epoch": 0.09935897435897435, + "grad_norm": 0.3625728189945221, + "learning_rate": 0.00019971658128215193, + "loss": 1.1897, + "step": 558 + }, + { + "epoch": 0.09953703703703703, + "grad_norm": 0.3815405070781708, + "learning_rate": 0.0001997155272020703, + "loss": 1.1473, + "step": 559 + }, + { + "epoch": 0.09971509971509972, + "grad_norm": 0.48664286732673645, + "learning_rate": 0.00019971447116827004, + "loss": 1.2462, + "step": 560 + }, + { + "epoch": 0.0998931623931624, + "grad_norm": 0.3708696663379669, + "learning_rate": 0.0001997134131807719, + "loss": 1.0979, + "step": 561 + }, + { + "epoch": 0.10007122507122507, + "grad_norm": 0.44511324167251587, + "learning_rate": 0.00019971235323959654, + "loss": 1.2313, + "step": 562 + }, + { + "epoch": 0.10024928774928775, + "grad_norm": 0.3687448799610138, + "learning_rate": 0.00019971129134476473, + "loss": 1.1526, + "step": 563 + }, + { + "epoch": 0.10042735042735043, + "grad_norm": 0.4506866931915283, + "learning_rate": 0.00019971022749629735, + "loss": 1.0003, + "step": 564 + }, + { + "epoch": 0.1006054131054131, + "grad_norm": 0.41910406947135925, + "learning_rate": 0.00019970916169421515, + "loss": 1.013, + "step": 565 + }, + { + "epoch": 0.10078347578347578, + "grad_norm": 0.39728936553001404, + "learning_rate": 0.0001997080939385391, + "loss": 1.0501, + "step": 566 + }, + { + "epoch": 0.10096153846153846, + "grad_norm": 0.41415902972221375, + "learning_rate": 0.00019970702422929005, + "loss": 1.0791, + "step": 567 + }, + { + "epoch": 0.10113960113960115, + "grad_norm": 0.45630788803100586, + "learning_rate": 0.00019970595256648896, + "loss": 1.2884, + "step": 568 + }, + { + "epoch": 0.10131766381766381, + "grad_norm": 0.4371698796749115, + "learning_rate": 0.00019970487895015686, + "loss": 1.0684, + "step": 569 + }, + { + "epoch": 0.1014957264957265, + "grad_norm": 0.4350591003894806, + "learning_rate": 0.00019970380338031477, + "loss": 1.2415, + "step": 570 + }, + { + "epoch": 0.10167378917378918, + "grad_norm": 0.4232708215713501, + "learning_rate": 0.00019970272585698382, + "loss": 1.2656, + "step": 571 + }, + { + "epoch": 0.10185185185185185, + "grad_norm": 0.3917689919471741, + "learning_rate": 0.00019970164638018502, + "loss": 1.0178, + "step": 572 + }, + { + "epoch": 0.10202991452991453, + "grad_norm": 0.4262804388999939, + "learning_rate": 0.0001997005649499396, + "loss": 1.1805, + "step": 573 + }, + { + "epoch": 0.10220797720797721, + "grad_norm": 0.5217884182929993, + "learning_rate": 0.0001996994815662687, + "loss": 1.2392, + "step": 574 + }, + { + "epoch": 0.10238603988603989, + "grad_norm": 0.4273875057697296, + "learning_rate": 0.00019969839622919358, + "loss": 1.0844, + "step": 575 + }, + { + "epoch": 0.10256410256410256, + "grad_norm": 0.41588085889816284, + "learning_rate": 0.00019969730893873547, + "loss": 1.2437, + "step": 576 + }, + { + "epoch": 0.10274216524216524, + "grad_norm": 0.41617709398269653, + "learning_rate": 0.0001996962196949157, + "loss": 0.9519, + "step": 577 + }, + { + "epoch": 0.10292022792022792, + "grad_norm": 0.4832979142665863, + "learning_rate": 0.00019969512849775565, + "loss": 1.1889, + "step": 578 + }, + { + "epoch": 0.10309829059829059, + "grad_norm": 0.3936060965061188, + "learning_rate": 0.0001996940353472766, + "loss": 0.9888, + "step": 579 + }, + { + "epoch": 0.10327635327635327, + "grad_norm": 0.4147680997848511, + "learning_rate": 0.00019969294024350004, + "loss": 1.0733, + "step": 580 + }, + { + "epoch": 0.10345441595441596, + "grad_norm": 0.37791356444358826, + "learning_rate": 0.00019969184318644742, + "loss": 1.212, + "step": 581 + }, + { + "epoch": 0.10363247863247864, + "grad_norm": 0.44297221302986145, + "learning_rate": 0.00019969074417614023, + "loss": 1.0535, + "step": 582 + }, + { + "epoch": 0.10381054131054131, + "grad_norm": 0.4032835066318512, + "learning_rate": 0.0001996896432126, + "loss": 1.1869, + "step": 583 + }, + { + "epoch": 0.10398860398860399, + "grad_norm": 0.49271953105926514, + "learning_rate": 0.00019968854029584827, + "loss": 1.1661, + "step": 584 + }, + { + "epoch": 0.10416666666666667, + "grad_norm": 0.362699031829834, + "learning_rate": 0.0001996874354259067, + "loss": 0.868, + "step": 585 + }, + { + "epoch": 0.10434472934472934, + "grad_norm": 0.401795357465744, + "learning_rate": 0.0001996863286027969, + "loss": 1.1045, + "step": 586 + }, + { + "epoch": 0.10452279202279202, + "grad_norm": 0.45380479097366333, + "learning_rate": 0.00019968521982654058, + "loss": 0.8503, + "step": 587 + }, + { + "epoch": 0.1047008547008547, + "grad_norm": 0.49759066104888916, + "learning_rate": 0.00019968410909715947, + "loss": 1.4073, + "step": 588 + }, + { + "epoch": 0.10487891737891739, + "grad_norm": 0.4421198070049286, + "learning_rate": 0.0001996829964146753, + "loss": 1.1512, + "step": 589 + }, + { + "epoch": 0.10505698005698005, + "grad_norm": 0.46675658226013184, + "learning_rate": 0.00019968188177910988, + "loss": 1.0132, + "step": 590 + }, + { + "epoch": 0.10523504273504274, + "grad_norm": 0.5710657238960266, + "learning_rate": 0.00019968076519048507, + "loss": 1.267, + "step": 591 + }, + { + "epoch": 0.10541310541310542, + "grad_norm": 0.4655563235282898, + "learning_rate": 0.00019967964664882276, + "loss": 1.1204, + "step": 592 + }, + { + "epoch": 0.10559116809116809, + "grad_norm": 0.3895256519317627, + "learning_rate": 0.00019967852615414478, + "loss": 1.0814, + "step": 593 + }, + { + "epoch": 0.10576923076923077, + "grad_norm": 0.424216091632843, + "learning_rate": 0.00019967740370647322, + "loss": 1.1663, + "step": 594 + }, + { + "epoch": 0.10594729344729345, + "grad_norm": 0.3978985846042633, + "learning_rate": 0.00019967627930582996, + "loss": 0.909, + "step": 595 + }, + { + "epoch": 0.10612535612535613, + "grad_norm": 0.47064995765686035, + "learning_rate": 0.00019967515295223705, + "loss": 1.2351, + "step": 596 + }, + { + "epoch": 0.1063034188034188, + "grad_norm": 0.42449644207954407, + "learning_rate": 0.0001996740246457166, + "loss": 0.9739, + "step": 597 + }, + { + "epoch": 0.10648148148148148, + "grad_norm": 0.39033401012420654, + "learning_rate": 0.00019967289438629066, + "loss": 1.0933, + "step": 598 + }, + { + "epoch": 0.10665954415954416, + "grad_norm": 0.4398612678050995, + "learning_rate": 0.00019967176217398143, + "loss": 1.2479, + "step": 599 + }, + { + "epoch": 0.10683760683760683, + "grad_norm": 0.3946632742881775, + "learning_rate": 0.00019967062800881107, + "loss": 1.0417, + "step": 600 + }, + { + "epoch": 0.10701566951566951, + "grad_norm": 0.5083445906639099, + "learning_rate": 0.0001996694918908018, + "loss": 1.1109, + "step": 601 + }, + { + "epoch": 0.1071937321937322, + "grad_norm": 0.477724552154541, + "learning_rate": 0.00019966835381997585, + "loss": 1.2891, + "step": 602 + }, + { + "epoch": 0.10737179487179487, + "grad_norm": 0.4110167920589447, + "learning_rate": 0.0001996672137963556, + "loss": 1.0555, + "step": 603 + }, + { + "epoch": 0.10754985754985755, + "grad_norm": 0.44078320264816284, + "learning_rate": 0.00019966607181996334, + "loss": 0.9188, + "step": 604 + }, + { + "epoch": 0.10772792022792023, + "grad_norm": 0.41251105070114136, + "learning_rate": 0.00019966492789082142, + "loss": 1.2592, + "step": 605 + }, + { + "epoch": 0.10790598290598291, + "grad_norm": 0.37701505422592163, + "learning_rate": 0.00019966378200895227, + "loss": 1.0233, + "step": 606 + }, + { + "epoch": 0.10808404558404558, + "grad_norm": 0.44624966382980347, + "learning_rate": 0.00019966263417437835, + "loss": 1.2273, + "step": 607 + }, + { + "epoch": 0.10826210826210826, + "grad_norm": 0.3618549108505249, + "learning_rate": 0.00019966148438712214, + "loss": 0.9101, + "step": 608 + }, + { + "epoch": 0.10844017094017094, + "grad_norm": 0.384574294090271, + "learning_rate": 0.00019966033264720616, + "loss": 1.1769, + "step": 609 + }, + { + "epoch": 0.10861823361823361, + "grad_norm": 0.50872403383255, + "learning_rate": 0.000199659178954653, + "loss": 1.1213, + "step": 610 + }, + { + "epoch": 0.1087962962962963, + "grad_norm": 0.39736685156822205, + "learning_rate": 0.00019965802330948527, + "loss": 1.275, + "step": 611 + }, + { + "epoch": 0.10897435897435898, + "grad_norm": 0.484660267829895, + "learning_rate": 0.00019965686571172557, + "loss": 1.1671, + "step": 612 + }, + { + "epoch": 0.10915242165242166, + "grad_norm": 0.41420218348503113, + "learning_rate": 0.0001996557061613966, + "loss": 0.9541, + "step": 613 + }, + { + "epoch": 0.10933048433048433, + "grad_norm": 0.4057196080684662, + "learning_rate": 0.00019965454465852112, + "loss": 1.0145, + "step": 614 + }, + { + "epoch": 0.10950854700854701, + "grad_norm": 0.4559510052204132, + "learning_rate": 0.00019965338120312182, + "loss": 1.0889, + "step": 615 + }, + { + "epoch": 0.10968660968660969, + "grad_norm": 0.40960055589675903, + "learning_rate": 0.00019965221579522154, + "loss": 1.1447, + "step": 616 + }, + { + "epoch": 0.10986467236467236, + "grad_norm": 0.4701732099056244, + "learning_rate": 0.0001996510484348431, + "loss": 1.2871, + "step": 617 + }, + { + "epoch": 0.11004273504273504, + "grad_norm": 0.38420796394348145, + "learning_rate": 0.0001996498791220094, + "loss": 1.058, + "step": 618 + }, + { + "epoch": 0.11022079772079772, + "grad_norm": 0.4014730453491211, + "learning_rate": 0.00019964870785674327, + "loss": 1.023, + "step": 619 + }, + { + "epoch": 0.1103988603988604, + "grad_norm": 0.38846179842948914, + "learning_rate": 0.00019964753463906773, + "loss": 0.9834, + "step": 620 + }, + { + "epoch": 0.11057692307692307, + "grad_norm": 0.5120236277580261, + "learning_rate": 0.00019964635946900577, + "loss": 1.2347, + "step": 621 + }, + { + "epoch": 0.11075498575498575, + "grad_norm": 0.40483301877975464, + "learning_rate": 0.00019964518234658038, + "loss": 1.131, + "step": 622 + }, + { + "epoch": 0.11093304843304844, + "grad_norm": 0.445782870054245, + "learning_rate": 0.00019964400327181464, + "loss": 0.9349, + "step": 623 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 0.490460604429245, + "learning_rate": 0.00019964282224473165, + "loss": 1.0257, + "step": 624 + }, + { + "epoch": 0.11128917378917379, + "grad_norm": 0.37585243582725525, + "learning_rate": 0.00019964163926535454, + "loss": 0.9724, + "step": 625 + }, + { + "epoch": 0.11146723646723647, + "grad_norm": 0.4160473346710205, + "learning_rate": 0.00019964045433370651, + "loss": 0.874, + "step": 626 + }, + { + "epoch": 0.11164529914529915, + "grad_norm": 0.442425012588501, + "learning_rate": 0.00019963926744981074, + "loss": 1.064, + "step": 627 + }, + { + "epoch": 0.11182336182336182, + "grad_norm": 0.4451471269130707, + "learning_rate": 0.00019963807861369054, + "loss": 1.2343, + "step": 628 + }, + { + "epoch": 0.1120014245014245, + "grad_norm": 0.5018183588981628, + "learning_rate": 0.00019963688782536913, + "loss": 1.1226, + "step": 629 + }, + { + "epoch": 0.11217948717948718, + "grad_norm": 0.43723925948143005, + "learning_rate": 0.0001996356950848699, + "loss": 1.0178, + "step": 630 + }, + { + "epoch": 0.11235754985754985, + "grad_norm": 0.4794611930847168, + "learning_rate": 0.0001996345003922162, + "loss": 0.9695, + "step": 631 + }, + { + "epoch": 0.11253561253561253, + "grad_norm": 0.5021790266036987, + "learning_rate": 0.00019963330374743143, + "loss": 1.1748, + "step": 632 + }, + { + "epoch": 0.11271367521367522, + "grad_norm": 0.47228625416755676, + "learning_rate": 0.00019963210515053906, + "loss": 1.2138, + "step": 633 + }, + { + "epoch": 0.1128917378917379, + "grad_norm": 0.4261155128479004, + "learning_rate": 0.00019963090460156256, + "loss": 0.9428, + "step": 634 + }, + { + "epoch": 0.11306980056980057, + "grad_norm": 0.3279525339603424, + "learning_rate": 0.00019962970210052542, + "loss": 0.7803, + "step": 635 + }, + { + "epoch": 0.11324786324786325, + "grad_norm": 0.5106086730957031, + "learning_rate": 0.00019962849764745125, + "loss": 1.113, + "step": 636 + }, + { + "epoch": 0.11342592592592593, + "grad_norm": 0.38272222876548767, + "learning_rate": 0.00019962729124236363, + "loss": 0.896, + "step": 637 + }, + { + "epoch": 0.1136039886039886, + "grad_norm": 0.39532098174095154, + "learning_rate": 0.0001996260828852862, + "loss": 0.9308, + "step": 638 + }, + { + "epoch": 0.11378205128205128, + "grad_norm": 0.44947221875190735, + "learning_rate": 0.00019962487257624262, + "loss": 1.207, + "step": 639 + }, + { + "epoch": 0.11396011396011396, + "grad_norm": 0.40684598684310913, + "learning_rate": 0.00019962366031525664, + "loss": 1.11, + "step": 640 + }, + { + "epoch": 0.11413817663817664, + "grad_norm": 0.4296625852584839, + "learning_rate": 0.00019962244610235194, + "loss": 1.2784, + "step": 641 + }, + { + "epoch": 0.11431623931623931, + "grad_norm": 0.4560794532299042, + "learning_rate": 0.0001996212299375524, + "loss": 1.1191, + "step": 642 + }, + { + "epoch": 0.114494301994302, + "grad_norm": 0.40246087312698364, + "learning_rate": 0.00019962001182088177, + "loss": 1.1401, + "step": 643 + }, + { + "epoch": 0.11467236467236468, + "grad_norm": 0.3938910663127899, + "learning_rate": 0.000199618791752364, + "loss": 1.0959, + "step": 644 + }, + { + "epoch": 0.11485042735042734, + "grad_norm": 0.4123380184173584, + "learning_rate": 0.00019961756973202287, + "loss": 1.2824, + "step": 645 + }, + { + "epoch": 0.11502849002849003, + "grad_norm": 0.41085442900657654, + "learning_rate": 0.00019961634575988243, + "loss": 1.1137, + "step": 646 + }, + { + "epoch": 0.11520655270655271, + "grad_norm": 0.38276201486587524, + "learning_rate": 0.0001996151198359667, + "loss": 1.0747, + "step": 647 + }, + { + "epoch": 0.11538461538461539, + "grad_norm": 0.49269407987594604, + "learning_rate": 0.00019961389196029953, + "loss": 1.1731, + "step": 648 + }, + { + "epoch": 0.11556267806267806, + "grad_norm": 0.5152469277381897, + "learning_rate": 0.00019961266213290512, + "loss": 1.3574, + "step": 649 + }, + { + "epoch": 0.11574074074074074, + "grad_norm": 0.4835714101791382, + "learning_rate": 0.0001996114303538075, + "loss": 1.2859, + "step": 650 + }, + { + "epoch": 0.11591880341880342, + "grad_norm": 0.4284524917602539, + "learning_rate": 0.00019961019662303087, + "loss": 1.1103, + "step": 651 + }, + { + "epoch": 0.11609686609686609, + "grad_norm": 0.3933276832103729, + "learning_rate": 0.00019960896094059933, + "loss": 1.2647, + "step": 652 + }, + { + "epoch": 0.11627492877492877, + "grad_norm": 0.33749741315841675, + "learning_rate": 0.00019960772330653712, + "loss": 0.819, + "step": 653 + }, + { + "epoch": 0.11645299145299146, + "grad_norm": 0.48122069239616394, + "learning_rate": 0.00019960648372086852, + "loss": 1.2781, + "step": 654 + }, + { + "epoch": 0.11663105413105414, + "grad_norm": 0.4681607186794281, + "learning_rate": 0.00019960524218361775, + "loss": 0.9723, + "step": 655 + }, + { + "epoch": 0.1168091168091168, + "grad_norm": 0.3974960148334503, + "learning_rate": 0.0001996039986948092, + "loss": 1.0302, + "step": 656 + }, + { + "epoch": 0.11698717948717949, + "grad_norm": 0.43180662393569946, + "learning_rate": 0.0001996027532544672, + "loss": 1.3265, + "step": 657 + }, + { + "epoch": 0.11716524216524217, + "grad_norm": 0.4481917917728424, + "learning_rate": 0.00019960150586261613, + "loss": 1.136, + "step": 658 + }, + { + "epoch": 0.11734330484330484, + "grad_norm": 0.43428945541381836, + "learning_rate": 0.00019960025651928045, + "loss": 1.2412, + "step": 659 + }, + { + "epoch": 0.11752136752136752, + "grad_norm": 0.36211395263671875, + "learning_rate": 0.00019959900522448467, + "loss": 0.9563, + "step": 660 + }, + { + "epoch": 0.1176994301994302, + "grad_norm": 0.43585848808288574, + "learning_rate": 0.0001995977519782533, + "loss": 1.1677, + "step": 661 + }, + { + "epoch": 0.11787749287749288, + "grad_norm": 0.4232597351074219, + "learning_rate": 0.00019959649678061086, + "loss": 1.1187, + "step": 662 + }, + { + "epoch": 0.11805555555555555, + "grad_norm": 0.3304753303527832, + "learning_rate": 0.00019959523963158194, + "loss": 0.8473, + "step": 663 + }, + { + "epoch": 0.11823361823361823, + "grad_norm": 0.37600061297416687, + "learning_rate": 0.0001995939805311912, + "loss": 1.1227, + "step": 664 + }, + { + "epoch": 0.11841168091168092, + "grad_norm": 0.33417847752571106, + "learning_rate": 0.0001995927194794633, + "loss": 1.0315, + "step": 665 + }, + { + "epoch": 0.11858974358974358, + "grad_norm": 0.46799129247665405, + "learning_rate": 0.00019959145647642298, + "loss": 1.135, + "step": 666 + }, + { + "epoch": 0.11876780626780627, + "grad_norm": 0.4141576886177063, + "learning_rate": 0.0001995901915220949, + "loss": 1.0956, + "step": 667 + }, + { + "epoch": 0.11894586894586895, + "grad_norm": 0.3824596405029297, + "learning_rate": 0.0001995889246165039, + "loss": 1.1782, + "step": 668 + }, + { + "epoch": 0.11912393162393162, + "grad_norm": 0.4087786376476288, + "learning_rate": 0.00019958765575967484, + "loss": 0.9704, + "step": 669 + }, + { + "epoch": 0.1193019943019943, + "grad_norm": 0.5161317586898804, + "learning_rate": 0.00019958638495163252, + "loss": 1.2207, + "step": 670 + }, + { + "epoch": 0.11948005698005698, + "grad_norm": 0.4782274067401886, + "learning_rate": 0.0001995851121924019, + "loss": 1.1257, + "step": 671 + }, + { + "epoch": 0.11965811965811966, + "grad_norm": 0.40617331862449646, + "learning_rate": 0.00019958383748200782, + "loss": 1.1153, + "step": 672 + }, + { + "epoch": 0.11983618233618233, + "grad_norm": 0.40149980783462524, + "learning_rate": 0.00019958256082047533, + "loss": 0.9785, + "step": 673 + }, + { + "epoch": 0.12001424501424501, + "grad_norm": 0.4378886818885803, + "learning_rate": 0.00019958128220782942, + "loss": 1.1355, + "step": 674 + }, + { + "epoch": 0.1201923076923077, + "grad_norm": 0.4449596703052521, + "learning_rate": 0.0001995800016440952, + "loss": 1.0325, + "step": 675 + }, + { + "epoch": 0.12037037037037036, + "grad_norm": 0.4268079698085785, + "learning_rate": 0.00019957871912929765, + "loss": 1.1901, + "step": 676 + }, + { + "epoch": 0.12054843304843305, + "grad_norm": 0.4250091016292572, + "learning_rate": 0.00019957743466346198, + "loss": 1.0084, + "step": 677 + }, + { + "epoch": 0.12072649572649573, + "grad_norm": 0.40724286437034607, + "learning_rate": 0.0001995761482466133, + "loss": 1.0866, + "step": 678 + }, + { + "epoch": 0.12090455840455841, + "grad_norm": 0.42478349804878235, + "learning_rate": 0.00019957485987877688, + "loss": 1.1909, + "step": 679 + }, + { + "epoch": 0.12108262108262108, + "grad_norm": 0.371362566947937, + "learning_rate": 0.0001995735695599779, + "loss": 1.083, + "step": 680 + }, + { + "epoch": 0.12126068376068376, + "grad_norm": 0.4715283513069153, + "learning_rate": 0.0001995722772902417, + "loss": 1.2942, + "step": 681 + }, + { + "epoch": 0.12143874643874644, + "grad_norm": 0.3611983060836792, + "learning_rate": 0.00019957098306959355, + "loss": 0.9878, + "step": 682 + }, + { + "epoch": 0.12161680911680911, + "grad_norm": 0.4764883816242218, + "learning_rate": 0.00019956968689805883, + "loss": 1.0082, + "step": 683 + }, + { + "epoch": 0.12179487179487179, + "grad_norm": 0.33170604705810547, + "learning_rate": 0.00019956838877566293, + "loss": 0.8529, + "step": 684 + }, + { + "epoch": 0.12197293447293447, + "grad_norm": 0.46896886825561523, + "learning_rate": 0.00019956708870243133, + "loss": 1.0745, + "step": 685 + }, + { + "epoch": 0.12215099715099716, + "grad_norm": 0.4120674431324005, + "learning_rate": 0.00019956578667838941, + "loss": 1.1828, + "step": 686 + }, + { + "epoch": 0.12232905982905982, + "grad_norm": 0.45671191811561584, + "learning_rate": 0.00019956448270356275, + "loss": 1.3484, + "step": 687 + }, + { + "epoch": 0.1225071225071225, + "grad_norm": 0.4023838937282562, + "learning_rate": 0.00019956317677797687, + "loss": 0.9623, + "step": 688 + }, + { + "epoch": 0.12268518518518519, + "grad_norm": 0.5205856561660767, + "learning_rate": 0.00019956186890165737, + "loss": 1.2221, + "step": 689 + }, + { + "epoch": 0.12286324786324786, + "grad_norm": 0.43956050276756287, + "learning_rate": 0.00019956055907462987, + "loss": 1.1051, + "step": 690 + }, + { + "epoch": 0.12304131054131054, + "grad_norm": 0.4341758191585541, + "learning_rate": 0.00019955924729692003, + "loss": 0.8972, + "step": 691 + }, + { + "epoch": 0.12321937321937322, + "grad_norm": 0.42025020718574524, + "learning_rate": 0.00019955793356855357, + "loss": 1.1137, + "step": 692 + }, + { + "epoch": 0.1233974358974359, + "grad_norm": 0.44375079870224, + "learning_rate": 0.0001995566178895562, + "loss": 1.2783, + "step": 693 + }, + { + "epoch": 0.12357549857549857, + "grad_norm": 0.4703320264816284, + "learning_rate": 0.00019955530025995372, + "loss": 1.1991, + "step": 694 + }, + { + "epoch": 0.12375356125356125, + "grad_norm": 0.43781620264053345, + "learning_rate": 0.00019955398067977195, + "loss": 1.2316, + "step": 695 + }, + { + "epoch": 0.12393162393162394, + "grad_norm": 0.4362877607345581, + "learning_rate": 0.0001995526591490367, + "loss": 1.1374, + "step": 696 + }, + { + "epoch": 0.1241096866096866, + "grad_norm": 0.4434499442577362, + "learning_rate": 0.00019955133566777392, + "loss": 1.1034, + "step": 697 + }, + { + "epoch": 0.12428774928774929, + "grad_norm": 0.46613508462905884, + "learning_rate": 0.00019955001023600955, + "loss": 1.2252, + "step": 698 + }, + { + "epoch": 0.12446581196581197, + "grad_norm": 0.46226736903190613, + "learning_rate": 0.00019954868285376945, + "loss": 1.0296, + "step": 699 + }, + { + "epoch": 0.12464387464387465, + "grad_norm": 0.4460904002189636, + "learning_rate": 0.00019954735352107977, + "loss": 1.0553, + "step": 700 + }, + { + "epoch": 0.12482193732193732, + "grad_norm": 0.36708924174308777, + "learning_rate": 0.00019954602223796648, + "loss": 0.9384, + "step": 701 + }, + { + "epoch": 0.125, + "grad_norm": 0.3780093491077423, + "learning_rate": 0.00019954468900445566, + "loss": 0.9062, + "step": 702 + }, + { + "epoch": 0.12517806267806267, + "grad_norm": 0.41797417402267456, + "learning_rate": 0.00019954335382057345, + "loss": 1.0344, + "step": 703 + }, + { + "epoch": 0.12535612535612536, + "grad_norm": 0.43710798025131226, + "learning_rate": 0.00019954201668634597, + "loss": 1.1324, + "step": 704 + }, + { + "epoch": 0.12553418803418803, + "grad_norm": 0.4732789695262909, + "learning_rate": 0.00019954067760179952, + "loss": 1.1419, + "step": 705 + }, + { + "epoch": 0.1257122507122507, + "grad_norm": 0.43248575925827026, + "learning_rate": 0.00019953933656696022, + "loss": 1.5112, + "step": 706 + }, + { + "epoch": 0.1258903133903134, + "grad_norm": 0.4074753522872925, + "learning_rate": 0.00019953799358185442, + "loss": 0.9751, + "step": 707 + }, + { + "epoch": 0.12606837606837606, + "grad_norm": 0.4586823880672455, + "learning_rate": 0.0001995366486465084, + "loss": 1.267, + "step": 708 + }, + { + "epoch": 0.12624643874643873, + "grad_norm": 0.4716857075691223, + "learning_rate": 0.0001995353017609485, + "loss": 1.1636, + "step": 709 + }, + { + "epoch": 0.12642450142450143, + "grad_norm": 0.5214398503303528, + "learning_rate": 0.00019953395292520115, + "loss": 1.2317, + "step": 710 + }, + { + "epoch": 0.1266025641025641, + "grad_norm": 0.42961129546165466, + "learning_rate": 0.00019953260213929276, + "loss": 1.0271, + "step": 711 + }, + { + "epoch": 0.1267806267806268, + "grad_norm": 0.4764653444290161, + "learning_rate": 0.00019953124940324979, + "loss": 1.1747, + "step": 712 + }, + { + "epoch": 0.12695868945868946, + "grad_norm": 0.4420304000377655, + "learning_rate": 0.00019952989471709874, + "loss": 0.9783, + "step": 713 + }, + { + "epoch": 0.12713675213675213, + "grad_norm": 0.44114625453948975, + "learning_rate": 0.00019952853808086616, + "loss": 1.1953, + "step": 714 + }, + { + "epoch": 0.12731481481481483, + "grad_norm": 0.501923143863678, + "learning_rate": 0.0001995271794945786, + "loss": 0.9886, + "step": 715 + }, + { + "epoch": 0.1274928774928775, + "grad_norm": 0.42266538739204407, + "learning_rate": 0.00019952581895826276, + "loss": 1.2033, + "step": 716 + }, + { + "epoch": 0.12767094017094016, + "grad_norm": 0.37770554423332214, + "learning_rate": 0.00019952445647194523, + "loss": 1.0164, + "step": 717 + }, + { + "epoch": 0.12784900284900286, + "grad_norm": 0.369266152381897, + "learning_rate": 0.00019952309203565268, + "loss": 0.9186, + "step": 718 + }, + { + "epoch": 0.12802706552706553, + "grad_norm": 0.40446221828460693, + "learning_rate": 0.00019952172564941193, + "loss": 1.1576, + "step": 719 + }, + { + "epoch": 0.1282051282051282, + "grad_norm": 0.504172146320343, + "learning_rate": 0.00019952035731324967, + "loss": 1.2695, + "step": 720 + }, + { + "epoch": 0.1283831908831909, + "grad_norm": 0.37284108996391296, + "learning_rate": 0.0001995189870271928, + "loss": 1.0288, + "step": 721 + }, + { + "epoch": 0.12856125356125356, + "grad_norm": 0.41811618208885193, + "learning_rate": 0.00019951761479126805, + "loss": 1.2241, + "step": 722 + }, + { + "epoch": 0.12873931623931623, + "grad_norm": 0.44706249237060547, + "learning_rate": 0.0001995162406055024, + "loss": 1.0831, + "step": 723 + }, + { + "epoch": 0.12891737891737892, + "grad_norm": 0.426572322845459, + "learning_rate": 0.00019951486446992273, + "loss": 1.0047, + "step": 724 + }, + { + "epoch": 0.1290954415954416, + "grad_norm": 0.4446277618408203, + "learning_rate": 0.00019951348638455602, + "loss": 1.0827, + "step": 725 + }, + { + "epoch": 0.12927350427350429, + "grad_norm": 0.3934919834136963, + "learning_rate": 0.00019951210634942926, + "loss": 0.9808, + "step": 726 + }, + { + "epoch": 0.12945156695156695, + "grad_norm": 0.4316558241844177, + "learning_rate": 0.0001995107243645695, + "loss": 1.3341, + "step": 727 + }, + { + "epoch": 0.12962962962962962, + "grad_norm": 0.43074217438697815, + "learning_rate": 0.00019950934043000382, + "loss": 1.007, + "step": 728 + }, + { + "epoch": 0.12980769230769232, + "grad_norm": 0.5212171673774719, + "learning_rate": 0.0001995079545457593, + "loss": 1.1822, + "step": 729 + }, + { + "epoch": 0.129985754985755, + "grad_norm": 0.3749600946903229, + "learning_rate": 0.00019950656671186313, + "loss": 0.9657, + "step": 730 + }, + { + "epoch": 0.13016381766381765, + "grad_norm": 0.36626043915748596, + "learning_rate": 0.00019950517692834252, + "loss": 1.1274, + "step": 731 + }, + { + "epoch": 0.13034188034188035, + "grad_norm": 0.4635467529296875, + "learning_rate": 0.00019950378519522467, + "loss": 1.2305, + "step": 732 + }, + { + "epoch": 0.13051994301994302, + "grad_norm": 0.4077455699443817, + "learning_rate": 0.00019950239151253683, + "loss": 0.9485, + "step": 733 + }, + { + "epoch": 0.1306980056980057, + "grad_norm": 0.4222758114337921, + "learning_rate": 0.0001995009958803063, + "loss": 1.0376, + "step": 734 + }, + { + "epoch": 0.13087606837606838, + "grad_norm": 0.4330402612686157, + "learning_rate": 0.0001994995982985605, + "loss": 1.1774, + "step": 735 + }, + { + "epoch": 0.13105413105413105, + "grad_norm": 0.42275673151016235, + "learning_rate": 0.00019949819876732673, + "loss": 1.1238, + "step": 736 + }, + { + "epoch": 0.13123219373219372, + "grad_norm": 0.45576968789100647, + "learning_rate": 0.00019949679728663246, + "loss": 1.0428, + "step": 737 + }, + { + "epoch": 0.13141025641025642, + "grad_norm": 0.5508752465248108, + "learning_rate": 0.00019949539385650514, + "loss": 1.3221, + "step": 738 + }, + { + "epoch": 0.13158831908831908, + "grad_norm": 0.4115872383117676, + "learning_rate": 0.00019949398847697225, + "loss": 1.0301, + "step": 739 + }, + { + "epoch": 0.13176638176638178, + "grad_norm": 0.4662442207336426, + "learning_rate": 0.00019949258114806132, + "loss": 1.3263, + "step": 740 + }, + { + "epoch": 0.13194444444444445, + "grad_norm": 0.6077266931533813, + "learning_rate": 0.00019949117186979999, + "loss": 1.0269, + "step": 741 + }, + { + "epoch": 0.13212250712250712, + "grad_norm": 0.47039318084716797, + "learning_rate": 0.00019948976064221579, + "loss": 1.3782, + "step": 742 + }, + { + "epoch": 0.1323005698005698, + "grad_norm": 0.4773450493812561, + "learning_rate": 0.0001994883474653364, + "loss": 1.289, + "step": 743 + }, + { + "epoch": 0.13247863247863248, + "grad_norm": 0.40180155634880066, + "learning_rate": 0.00019948693233918952, + "loss": 0.8691, + "step": 744 + }, + { + "epoch": 0.13265669515669515, + "grad_norm": 0.45216289162635803, + "learning_rate": 0.00019948551526380288, + "loss": 1.071, + "step": 745 + }, + { + "epoch": 0.13283475783475784, + "grad_norm": 0.4289272427558899, + "learning_rate": 0.0001994840962392042, + "loss": 1.0422, + "step": 746 + }, + { + "epoch": 0.1330128205128205, + "grad_norm": 0.4617730379104614, + "learning_rate": 0.00019948267526542134, + "loss": 1.0835, + "step": 747 + }, + { + "epoch": 0.13319088319088318, + "grad_norm": 0.42710617184638977, + "learning_rate": 0.00019948125234248208, + "loss": 1.0535, + "step": 748 + }, + { + "epoch": 0.13336894586894588, + "grad_norm": 0.43433234095573425, + "learning_rate": 0.0001994798274704144, + "loss": 0.9313, + "step": 749 + }, + { + "epoch": 0.13354700854700854, + "grad_norm": 0.46270284056663513, + "learning_rate": 0.0001994784006492461, + "loss": 1.0903, + "step": 750 + }, + { + "epoch": 0.1337250712250712, + "grad_norm": 0.5319814682006836, + "learning_rate": 0.00019947697187900517, + "loss": 1.2329, + "step": 751 + }, + { + "epoch": 0.1339031339031339, + "grad_norm": 0.3511372208595276, + "learning_rate": 0.00019947554115971967, + "loss": 0.7116, + "step": 752 + }, + { + "epoch": 0.13408119658119658, + "grad_norm": 0.4103890359401703, + "learning_rate": 0.00019947410849141756, + "loss": 1.1527, + "step": 753 + }, + { + "epoch": 0.13425925925925927, + "grad_norm": 0.5390757322311401, + "learning_rate": 0.00019947267387412695, + "loss": 1.1682, + "step": 754 + }, + { + "epoch": 0.13443732193732194, + "grad_norm": 0.29939723014831543, + "learning_rate": 0.0001994712373078759, + "loss": 0.5848, + "step": 755 + }, + { + "epoch": 0.1346153846153846, + "grad_norm": 0.4605920612812042, + "learning_rate": 0.0001994697987926926, + "loss": 0.9448, + "step": 756 + }, + { + "epoch": 0.1347934472934473, + "grad_norm": 0.426213800907135, + "learning_rate": 0.00019946835832860527, + "loss": 1.0487, + "step": 757 + }, + { + "epoch": 0.13497150997150997, + "grad_norm": 0.4209515154361725, + "learning_rate": 0.00019946691591564203, + "loss": 1.0951, + "step": 758 + }, + { + "epoch": 0.13514957264957264, + "grad_norm": 0.39555591344833374, + "learning_rate": 0.0001994654715538312, + "loss": 0.8754, + "step": 759 + }, + { + "epoch": 0.13532763532763534, + "grad_norm": 0.4065483510494232, + "learning_rate": 0.0001994640252432011, + "loss": 0.9451, + "step": 760 + }, + { + "epoch": 0.135505698005698, + "grad_norm": 0.4489104151725769, + "learning_rate": 0.00019946257698378003, + "loss": 1.2031, + "step": 761 + }, + { + "epoch": 0.13568376068376067, + "grad_norm": 0.39928409457206726, + "learning_rate": 0.0001994611267755964, + "loss": 1.1124, + "step": 762 + }, + { + "epoch": 0.13586182336182337, + "grad_norm": 0.4145409166812897, + "learning_rate": 0.00019945967461867858, + "loss": 1.083, + "step": 763 + }, + { + "epoch": 0.13603988603988604, + "grad_norm": 0.43508613109588623, + "learning_rate": 0.00019945822051305507, + "loss": 1.1119, + "step": 764 + }, + { + "epoch": 0.1362179487179487, + "grad_norm": 0.5186598300933838, + "learning_rate": 0.0001994567644587543, + "loss": 1.3256, + "step": 765 + }, + { + "epoch": 0.1363960113960114, + "grad_norm": 0.4615778625011444, + "learning_rate": 0.00019945530645580487, + "loss": 1.3906, + "step": 766 + }, + { + "epoch": 0.13657407407407407, + "grad_norm": 0.4838152527809143, + "learning_rate": 0.00019945384650423532, + "loss": 0.8169, + "step": 767 + }, + { + "epoch": 0.13675213675213677, + "grad_norm": 0.49253368377685547, + "learning_rate": 0.0001994523846040742, + "loss": 1.1613, + "step": 768 + }, + { + "epoch": 0.13693019943019943, + "grad_norm": 0.4697009325027466, + "learning_rate": 0.00019945092075535024, + "loss": 1.1722, + "step": 769 + }, + { + "epoch": 0.1371082621082621, + "grad_norm": 0.47162383794784546, + "learning_rate": 0.00019944945495809204, + "loss": 1.054, + "step": 770 + }, + { + "epoch": 0.1372863247863248, + "grad_norm": 0.4653547704219818, + "learning_rate": 0.00019944798721232835, + "loss": 1.1791, + "step": 771 + }, + { + "epoch": 0.13746438746438747, + "grad_norm": 0.4244011640548706, + "learning_rate": 0.000199446517518088, + "loss": 1.1557, + "step": 772 + }, + { + "epoch": 0.13764245014245013, + "grad_norm": 0.43812859058380127, + "learning_rate": 0.00019944504587539967, + "loss": 1.1567, + "step": 773 + }, + { + "epoch": 0.13782051282051283, + "grad_norm": 0.3984275162220001, + "learning_rate": 0.00019944357228429227, + "loss": 1.0715, + "step": 774 + }, + { + "epoch": 0.1379985754985755, + "grad_norm": 0.3794248104095459, + "learning_rate": 0.0001994420967447946, + "loss": 0.9377, + "step": 775 + }, + { + "epoch": 0.13817663817663817, + "grad_norm": 0.4214578866958618, + "learning_rate": 0.00019944061925693566, + "loss": 1.0112, + "step": 776 + }, + { + "epoch": 0.13835470085470086, + "grad_norm": 0.4738999605178833, + "learning_rate": 0.00019943913982074435, + "loss": 0.8718, + "step": 777 + }, + { + "epoch": 0.13853276353276353, + "grad_norm": 0.43455326557159424, + "learning_rate": 0.00019943765843624965, + "loss": 1.1343, + "step": 778 + }, + { + "epoch": 0.1387108262108262, + "grad_norm": 0.44973456859588623, + "learning_rate": 0.00019943617510348062, + "loss": 1.0487, + "step": 779 + }, + { + "epoch": 0.1388888888888889, + "grad_norm": 0.4216597080230713, + "learning_rate": 0.00019943468982246628, + "loss": 1.0765, + "step": 780 + }, + { + "epoch": 0.13906695156695156, + "grad_norm": 0.5089883208274841, + "learning_rate": 0.00019943320259323578, + "loss": 1.3137, + "step": 781 + }, + { + "epoch": 0.13924501424501423, + "grad_norm": 0.4358222782611847, + "learning_rate": 0.00019943171341581822, + "loss": 1.1891, + "step": 782 + }, + { + "epoch": 0.13942307692307693, + "grad_norm": 0.40918609499931335, + "learning_rate": 0.00019943022229024275, + "loss": 1.279, + "step": 783 + }, + { + "epoch": 0.1396011396011396, + "grad_norm": 0.4614863395690918, + "learning_rate": 0.00019942872921653866, + "loss": 1.2477, + "step": 784 + }, + { + "epoch": 0.1397792022792023, + "grad_norm": 0.4141528904438019, + "learning_rate": 0.00019942723419473515, + "loss": 0.9622, + "step": 785 + }, + { + "epoch": 0.13995726495726496, + "grad_norm": 0.536139726638794, + "learning_rate": 0.00019942573722486154, + "loss": 1.2127, + "step": 786 + }, + { + "epoch": 0.14013532763532763, + "grad_norm": 0.4968845546245575, + "learning_rate": 0.0001994242383069471, + "loss": 1.2965, + "step": 787 + }, + { + "epoch": 0.14031339031339032, + "grad_norm": 0.3897174894809723, + "learning_rate": 0.00019942273744102132, + "loss": 0.9907, + "step": 788 + }, + { + "epoch": 0.140491452991453, + "grad_norm": 0.466307669878006, + "learning_rate": 0.0001994212346271135, + "loss": 1.2021, + "step": 789 + }, + { + "epoch": 0.14066951566951566, + "grad_norm": 0.49283576011657715, + "learning_rate": 0.0001994197298652531, + "loss": 1.0969, + "step": 790 + }, + { + "epoch": 0.14084757834757836, + "grad_norm": 0.4686102271080017, + "learning_rate": 0.00019941822315546964, + "loss": 1.0125, + "step": 791 + }, + { + "epoch": 0.14102564102564102, + "grad_norm": 0.4389997124671936, + "learning_rate": 0.0001994167144977926, + "loss": 1.1294, + "step": 792 + }, + { + "epoch": 0.1412037037037037, + "grad_norm": 0.38539355993270874, + "learning_rate": 0.00019941520389225162, + "loss": 1.1231, + "step": 793 + }, + { + "epoch": 0.1413817663817664, + "grad_norm": 0.4860847592353821, + "learning_rate": 0.00019941369133887618, + "loss": 1.2268, + "step": 794 + }, + { + "epoch": 0.14155982905982906, + "grad_norm": 0.4567467272281647, + "learning_rate": 0.00019941217683769598, + "loss": 1.1482, + "step": 795 + }, + { + "epoch": 0.14173789173789172, + "grad_norm": 0.5549420714378357, + "learning_rate": 0.00019941066038874067, + "loss": 1.1899, + "step": 796 + }, + { + "epoch": 0.14191595441595442, + "grad_norm": 0.3950003385543823, + "learning_rate": 0.00019940914199204, + "loss": 0.96, + "step": 797 + }, + { + "epoch": 0.1420940170940171, + "grad_norm": 0.43845999240875244, + "learning_rate": 0.00019940762164762373, + "loss": 1.0338, + "step": 798 + }, + { + "epoch": 0.14227207977207978, + "grad_norm": 0.468537300825119, + "learning_rate": 0.00019940609935552157, + "loss": 1.2416, + "step": 799 + }, + { + "epoch": 0.14245014245014245, + "grad_norm": 0.4292038679122925, + "learning_rate": 0.0001994045751157634, + "loss": 1.1397, + "step": 800 + }, + { + "epoch": 0.14262820512820512, + "grad_norm": 0.3800995647907257, + "learning_rate": 0.00019940304892837908, + "loss": 0.939, + "step": 801 + }, + { + "epoch": 0.14280626780626782, + "grad_norm": 0.38004353642463684, + "learning_rate": 0.00019940152079339852, + "loss": 1.0485, + "step": 802 + }, + { + "epoch": 0.14298433048433049, + "grad_norm": 0.4658142924308777, + "learning_rate": 0.00019939999071085163, + "loss": 1.1561, + "step": 803 + }, + { + "epoch": 0.14316239316239315, + "grad_norm": 0.4235048294067383, + "learning_rate": 0.0001993984586807684, + "loss": 1.0516, + "step": 804 + }, + { + "epoch": 0.14334045584045585, + "grad_norm": 0.42925819754600525, + "learning_rate": 0.00019939692470317887, + "loss": 1.2238, + "step": 805 + }, + { + "epoch": 0.14351851851851852, + "grad_norm": 0.43701639771461487, + "learning_rate": 0.00019939538877811308, + "loss": 1.0129, + "step": 806 + }, + { + "epoch": 0.14369658119658119, + "grad_norm": 0.42786353826522827, + "learning_rate": 0.00019939385090560113, + "loss": 1.1355, + "step": 807 + }, + { + "epoch": 0.14387464387464388, + "grad_norm": 0.371218740940094, + "learning_rate": 0.00019939231108567312, + "loss": 0.9712, + "step": 808 + }, + { + "epoch": 0.14405270655270655, + "grad_norm": 0.4834294617176056, + "learning_rate": 0.00019939076931835926, + "loss": 1.1375, + "step": 809 + }, + { + "epoch": 0.14423076923076922, + "grad_norm": 0.4700150191783905, + "learning_rate": 0.00019938922560368974, + "loss": 1.1943, + "step": 810 + }, + { + "epoch": 0.14440883190883191, + "grad_norm": 0.4430996775627136, + "learning_rate": 0.0001993876799416948, + "loss": 1.1976, + "step": 811 + }, + { + "epoch": 0.14458689458689458, + "grad_norm": 0.4161672592163086, + "learning_rate": 0.00019938613233240476, + "loss": 1.0291, + "step": 812 + }, + { + "epoch": 0.14476495726495728, + "grad_norm": 0.39838850498199463, + "learning_rate": 0.0001993845827758499, + "loss": 1.2103, + "step": 813 + }, + { + "epoch": 0.14494301994301995, + "grad_norm": 0.429198294878006, + "learning_rate": 0.00019938303127206057, + "loss": 0.9971, + "step": 814 + }, + { + "epoch": 0.14512108262108261, + "grad_norm": 0.4589254856109619, + "learning_rate": 0.00019938147782106719, + "loss": 1.2392, + "step": 815 + }, + { + "epoch": 0.1452991452991453, + "grad_norm": 0.42506635189056396, + "learning_rate": 0.00019937992242290023, + "loss": 1.0827, + "step": 816 + }, + { + "epoch": 0.14547720797720798, + "grad_norm": 0.3778113126754761, + "learning_rate": 0.00019937836507759012, + "loss": 1.021, + "step": 817 + }, + { + "epoch": 0.14565527065527065, + "grad_norm": 0.43071216344833374, + "learning_rate": 0.0001993768057851674, + "loss": 1.273, + "step": 818 + }, + { + "epoch": 0.14583333333333334, + "grad_norm": 0.4944681227207184, + "learning_rate": 0.00019937524454566262, + "loss": 1.3037, + "step": 819 + }, + { + "epoch": 0.146011396011396, + "grad_norm": 0.4438824951648712, + "learning_rate": 0.00019937368135910632, + "loss": 1.1383, + "step": 820 + }, + { + "epoch": 0.14618945868945868, + "grad_norm": 0.400215744972229, + "learning_rate": 0.0001993721162255292, + "loss": 1.0669, + "step": 821 + }, + { + "epoch": 0.14636752136752137, + "grad_norm": 0.4341452121734619, + "learning_rate": 0.00019937054914496185, + "loss": 1.1431, + "step": 822 + }, + { + "epoch": 0.14654558404558404, + "grad_norm": 0.3941744267940521, + "learning_rate": 0.00019936898011743503, + "loss": 1.1593, + "step": 823 + }, + { + "epoch": 0.1467236467236467, + "grad_norm": 0.4318541884422302, + "learning_rate": 0.00019936740914297947, + "loss": 1.2814, + "step": 824 + }, + { + "epoch": 0.1469017094017094, + "grad_norm": 0.44488632678985596, + "learning_rate": 0.00019936583622162595, + "loss": 1.1054, + "step": 825 + }, + { + "epoch": 0.14707977207977208, + "grad_norm": 0.38701096177101135, + "learning_rate": 0.00019936426135340528, + "loss": 1.1086, + "step": 826 + }, + { + "epoch": 0.14725783475783477, + "grad_norm": 0.45794424414634705, + "learning_rate": 0.0001993626845383483, + "loss": 1.2395, + "step": 827 + }, + { + "epoch": 0.14743589743589744, + "grad_norm": 0.49237680435180664, + "learning_rate": 0.00019936110577648596, + "loss": 1.3483, + "step": 828 + }, + { + "epoch": 0.1476139601139601, + "grad_norm": 0.481666624546051, + "learning_rate": 0.00019935952506784914, + "loss": 1.1848, + "step": 829 + }, + { + "epoch": 0.1477920227920228, + "grad_norm": 0.4015209376811981, + "learning_rate": 0.00019935794241246883, + "loss": 1.0624, + "step": 830 + }, + { + "epoch": 0.14797008547008547, + "grad_norm": 0.47975999116897583, + "learning_rate": 0.00019935635781037606, + "loss": 1.1595, + "step": 831 + }, + { + "epoch": 0.14814814814814814, + "grad_norm": 0.4440356492996216, + "learning_rate": 0.00019935477126160181, + "loss": 1.1325, + "step": 832 + }, + { + "epoch": 0.14832621082621084, + "grad_norm": 0.4167410731315613, + "learning_rate": 0.00019935318276617723, + "loss": 1.0662, + "step": 833 + }, + { + "epoch": 0.1485042735042735, + "grad_norm": 0.4107447862625122, + "learning_rate": 0.0001993515923241334, + "loss": 0.8816, + "step": 834 + }, + { + "epoch": 0.14868233618233617, + "grad_norm": 0.4020158648490906, + "learning_rate": 0.00019934999993550154, + "loss": 0.9797, + "step": 835 + }, + { + "epoch": 0.14886039886039887, + "grad_norm": 0.4186473786830902, + "learning_rate": 0.0001993484056003128, + "loss": 1.1243, + "step": 836 + }, + { + "epoch": 0.14903846153846154, + "grad_norm": 0.5534794926643372, + "learning_rate": 0.00019934680931859842, + "loss": 1.1189, + "step": 837 + }, + { + "epoch": 0.1492165242165242, + "grad_norm": 0.37901270389556885, + "learning_rate": 0.0001993452110903897, + "loss": 0.9241, + "step": 838 + }, + { + "epoch": 0.1493945868945869, + "grad_norm": 0.41773587465286255, + "learning_rate": 0.00019934361091571793, + "loss": 0.9467, + "step": 839 + }, + { + "epoch": 0.14957264957264957, + "grad_norm": 0.4962073564529419, + "learning_rate": 0.00019934200879461448, + "loss": 1.2423, + "step": 840 + }, + { + "epoch": 0.14975071225071226, + "grad_norm": 0.38565897941589355, + "learning_rate": 0.00019934040472711074, + "loss": 1.1545, + "step": 841 + }, + { + "epoch": 0.14992877492877493, + "grad_norm": 0.4295346736907959, + "learning_rate": 0.0001993387987132381, + "loss": 1.2482, + "step": 842 + }, + { + "epoch": 0.1501068376068376, + "grad_norm": 0.4279189705848694, + "learning_rate": 0.0001993371907530281, + "loss": 1.1135, + "step": 843 + }, + { + "epoch": 0.1502849002849003, + "grad_norm": 0.44649168848991394, + "learning_rate": 0.0001993355808465122, + "loss": 1.0734, + "step": 844 + }, + { + "epoch": 0.15046296296296297, + "grad_norm": 0.453707218170166, + "learning_rate": 0.0001993339689937219, + "loss": 1.0992, + "step": 845 + }, + { + "epoch": 0.15064102564102563, + "grad_norm": 0.5113263726234436, + "learning_rate": 0.00019933235519468886, + "loss": 1.1792, + "step": 846 + }, + { + "epoch": 0.15081908831908833, + "grad_norm": 0.5822970271110535, + "learning_rate": 0.00019933073944944466, + "loss": 1.367, + "step": 847 + }, + { + "epoch": 0.150997150997151, + "grad_norm": 0.3946528732776642, + "learning_rate": 0.00019932912175802097, + "loss": 0.9781, + "step": 848 + }, + { + "epoch": 0.15117521367521367, + "grad_norm": 0.5429860949516296, + "learning_rate": 0.00019932750212044945, + "loss": 0.9783, + "step": 849 + }, + { + "epoch": 0.15135327635327636, + "grad_norm": 0.45847952365875244, + "learning_rate": 0.0001993258805367619, + "loss": 1.1352, + "step": 850 + }, + { + "epoch": 0.15153133903133903, + "grad_norm": 0.42770692706108093, + "learning_rate": 0.00019932425700699004, + "loss": 1.2365, + "step": 851 + }, + { + "epoch": 0.1517094017094017, + "grad_norm": 0.41845405101776123, + "learning_rate": 0.00019932263153116565, + "loss": 1.2642, + "step": 852 + }, + { + "epoch": 0.1518874643874644, + "grad_norm": 0.4641731083393097, + "learning_rate": 0.00019932100410932066, + "loss": 1.2009, + "step": 853 + }, + { + "epoch": 0.15206552706552706, + "grad_norm": 0.4128672778606415, + "learning_rate": 0.00019931937474148689, + "loss": 1.1981, + "step": 854 + }, + { + "epoch": 0.15224358974358973, + "grad_norm": 0.4730764925479889, + "learning_rate": 0.00019931774342769632, + "loss": 1.2145, + "step": 855 + }, + { + "epoch": 0.15242165242165243, + "grad_norm": 0.36611825227737427, + "learning_rate": 0.00019931611016798089, + "loss": 0.8504, + "step": 856 + }, + { + "epoch": 0.1525997150997151, + "grad_norm": 0.40944692492485046, + "learning_rate": 0.00019931447496237254, + "loss": 1.2853, + "step": 857 + }, + { + "epoch": 0.1527777777777778, + "grad_norm": 0.4521993398666382, + "learning_rate": 0.0001993128378109034, + "loss": 1.0198, + "step": 858 + }, + { + "epoch": 0.15295584045584046, + "grad_norm": 0.42113015055656433, + "learning_rate": 0.0001993111987136055, + "loss": 1.1284, + "step": 859 + }, + { + "epoch": 0.15313390313390313, + "grad_norm": 0.4117624759674072, + "learning_rate": 0.00019930955767051098, + "loss": 1.0445, + "step": 860 + }, + { + "epoch": 0.15331196581196582, + "grad_norm": 0.4807964265346527, + "learning_rate": 0.00019930791468165197, + "loss": 1.1378, + "step": 861 + }, + { + "epoch": 0.1534900284900285, + "grad_norm": 0.4186483323574066, + "learning_rate": 0.00019930626974706063, + "loss": 1.1636, + "step": 862 + }, + { + "epoch": 0.15366809116809116, + "grad_norm": 0.3764737844467163, + "learning_rate": 0.00019930462286676926, + "loss": 0.9523, + "step": 863 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.4283556044101715, + "learning_rate": 0.00019930297404081008, + "loss": 1.1008, + "step": 864 + }, + { + "epoch": 0.15402421652421652, + "grad_norm": 0.4485796093940735, + "learning_rate": 0.00019930132326921541, + "loss": 1.0834, + "step": 865 + }, + { + "epoch": 0.1542022792022792, + "grad_norm": 0.3882720172405243, + "learning_rate": 0.0001992996705520176, + "loss": 1.1086, + "step": 866 + }, + { + "epoch": 0.1543803418803419, + "grad_norm": 0.44698455929756165, + "learning_rate": 0.00019929801588924902, + "loss": 1.1437, + "step": 867 + }, + { + "epoch": 0.15455840455840456, + "grad_norm": 0.46978411078453064, + "learning_rate": 0.00019929635928094208, + "loss": 1.091, + "step": 868 + }, + { + "epoch": 0.15473646723646722, + "grad_norm": 0.4717854857444763, + "learning_rate": 0.00019929470072712927, + "loss": 1.1959, + "step": 869 + }, + { + "epoch": 0.15491452991452992, + "grad_norm": 0.4324854016304016, + "learning_rate": 0.00019929304022784305, + "loss": 1.2062, + "step": 870 + }, + { + "epoch": 0.1550925925925926, + "grad_norm": 0.3948180675506592, + "learning_rate": 0.00019929137778311597, + "loss": 1.1101, + "step": 871 + }, + { + "epoch": 0.15527065527065528, + "grad_norm": 0.40345287322998047, + "learning_rate": 0.0001992897133929806, + "loss": 0.8894, + "step": 872 + }, + { + "epoch": 0.15544871794871795, + "grad_norm": 0.44931963086128235, + "learning_rate": 0.00019928804705746957, + "loss": 0.9389, + "step": 873 + }, + { + "epoch": 0.15562678062678062, + "grad_norm": 0.529196560382843, + "learning_rate": 0.0001992863787766155, + "loss": 1.3362, + "step": 874 + }, + { + "epoch": 0.15580484330484332, + "grad_norm": 0.41218671202659607, + "learning_rate": 0.0001992847085504511, + "loss": 1.0727, + "step": 875 + }, + { + "epoch": 0.15598290598290598, + "grad_norm": 0.44074541330337524, + "learning_rate": 0.00019928303637900907, + "loss": 1.1091, + "step": 876 + }, + { + "epoch": 0.15616096866096865, + "grad_norm": 0.5264310240745544, + "learning_rate": 0.00019928136226232218, + "loss": 1.201, + "step": 877 + }, + { + "epoch": 0.15633903133903135, + "grad_norm": 0.4255099594593048, + "learning_rate": 0.00019927968620042324, + "loss": 1.2514, + "step": 878 + }, + { + "epoch": 0.15651709401709402, + "grad_norm": 0.4030280113220215, + "learning_rate": 0.0001992780081933451, + "loss": 1.0422, + "step": 879 + }, + { + "epoch": 0.15669515669515668, + "grad_norm": 0.5270203948020935, + "learning_rate": 0.00019927632824112058, + "loss": 1.2476, + "step": 880 + }, + { + "epoch": 0.15687321937321938, + "grad_norm": 0.37767237424850464, + "learning_rate": 0.00019927464634378268, + "loss": 1.0768, + "step": 881 + }, + { + "epoch": 0.15705128205128205, + "grad_norm": 0.4535936415195465, + "learning_rate": 0.0001992729625013643, + "loss": 1.2097, + "step": 882 + }, + { + "epoch": 0.15722934472934472, + "grad_norm": 0.4282119870185852, + "learning_rate": 0.00019927127671389843, + "loss": 1.0904, + "step": 883 + }, + { + "epoch": 0.1574074074074074, + "grad_norm": 0.3924157917499542, + "learning_rate": 0.0001992695889814181, + "loss": 0.9692, + "step": 884 + }, + { + "epoch": 0.15758547008547008, + "grad_norm": 0.525075376033783, + "learning_rate": 0.0001992678993039564, + "loss": 1.0292, + "step": 885 + }, + { + "epoch": 0.15776353276353278, + "grad_norm": 0.4388505518436432, + "learning_rate": 0.00019926620768154644, + "loss": 1.1944, + "step": 886 + }, + { + "epoch": 0.15794159544159544, + "grad_norm": 0.4362235963344574, + "learning_rate": 0.00019926451411422132, + "loss": 0.97, + "step": 887 + }, + { + "epoch": 0.1581196581196581, + "grad_norm": 0.4265296459197998, + "learning_rate": 0.0001992628186020143, + "loss": 0.9196, + "step": 888 + }, + { + "epoch": 0.1582977207977208, + "grad_norm": 0.4019876718521118, + "learning_rate": 0.0001992611211449585, + "loss": 1.1368, + "step": 889 + }, + { + "epoch": 0.15847578347578348, + "grad_norm": 0.5003397464752197, + "learning_rate": 0.00019925942174308726, + "loss": 1.2582, + "step": 890 + }, + { + "epoch": 0.15865384615384615, + "grad_norm": 0.4774404466152191, + "learning_rate": 0.00019925772039643382, + "loss": 1.2277, + "step": 891 + }, + { + "epoch": 0.15883190883190884, + "grad_norm": 0.4590449333190918, + "learning_rate": 0.00019925601710503153, + "loss": 1.1679, + "step": 892 + }, + { + "epoch": 0.1590099715099715, + "grad_norm": 0.4221442639827728, + "learning_rate": 0.0001992543118689138, + "loss": 1.1626, + "step": 893 + }, + { + "epoch": 0.15918803418803418, + "grad_norm": 0.47613003849983215, + "learning_rate": 0.00019925260468811403, + "loss": 1.1509, + "step": 894 + }, + { + "epoch": 0.15936609686609687, + "grad_norm": 0.41706812381744385, + "learning_rate": 0.0001992508955626656, + "loss": 1.0366, + "step": 895 + }, + { + "epoch": 0.15954415954415954, + "grad_norm": 0.5064654350280762, + "learning_rate": 0.00019924918449260205, + "loss": 1.0729, + "step": 896 + }, + { + "epoch": 0.1597222222222222, + "grad_norm": 0.5019610524177551, + "learning_rate": 0.00019924747147795696, + "loss": 1.0642, + "step": 897 + }, + { + "epoch": 0.1599002849002849, + "grad_norm": 0.4345671534538269, + "learning_rate": 0.00019924575651876378, + "loss": 1.1747, + "step": 898 + }, + { + "epoch": 0.16007834757834757, + "grad_norm": 0.4397568702697754, + "learning_rate": 0.0001992440396150562, + "loss": 1.282, + "step": 899 + }, + { + "epoch": 0.16025641025641027, + "grad_norm": 0.520187497138977, + "learning_rate": 0.0001992423207668678, + "loss": 0.976, + "step": 900 + }, + { + "epoch": 0.16043447293447294, + "grad_norm": 0.39329993724823, + "learning_rate": 0.0001992405999742323, + "loss": 0.9829, + "step": 901 + }, + { + "epoch": 0.1606125356125356, + "grad_norm": 0.42361345887184143, + "learning_rate": 0.00019923887723718339, + "loss": 1.139, + "step": 902 + }, + { + "epoch": 0.1607905982905983, + "grad_norm": 0.3846314251422882, + "learning_rate": 0.00019923715255575482, + "loss": 0.8262, + "step": 903 + }, + { + "epoch": 0.16096866096866097, + "grad_norm": 0.39258381724357605, + "learning_rate": 0.0001992354259299804, + "loss": 0.9638, + "step": 904 + }, + { + "epoch": 0.16114672364672364, + "grad_norm": 0.4000850319862366, + "learning_rate": 0.00019923369735989397, + "loss": 0.91, + "step": 905 + }, + { + "epoch": 0.16132478632478633, + "grad_norm": 0.46303513646125793, + "learning_rate": 0.00019923196684552936, + "loss": 1.1447, + "step": 906 + }, + { + "epoch": 0.161502849002849, + "grad_norm": 0.38437438011169434, + "learning_rate": 0.0001992302343869205, + "loss": 1.0212, + "step": 907 + }, + { + "epoch": 0.16168091168091167, + "grad_norm": 0.44585472345352173, + "learning_rate": 0.00019922849998410135, + "loss": 1.1964, + "step": 908 + }, + { + "epoch": 0.16185897435897437, + "grad_norm": 0.41959813237190247, + "learning_rate": 0.00019922676363710583, + "loss": 0.9925, + "step": 909 + }, + { + "epoch": 0.16203703703703703, + "grad_norm": 0.47442761063575745, + "learning_rate": 0.00019922502534596803, + "loss": 0.9237, + "step": 910 + }, + { + "epoch": 0.1622150997150997, + "grad_norm": 0.5065128207206726, + "learning_rate": 0.00019922328511072198, + "loss": 1.2573, + "step": 911 + }, + { + "epoch": 0.1623931623931624, + "grad_norm": 0.4739879369735718, + "learning_rate": 0.0001992215429314018, + "loss": 1.4416, + "step": 912 + }, + { + "epoch": 0.16257122507122507, + "grad_norm": 0.48763832449913025, + "learning_rate": 0.00019921979880804157, + "loss": 1.0408, + "step": 913 + }, + { + "epoch": 0.16274928774928774, + "grad_norm": 0.4841614067554474, + "learning_rate": 0.0001992180527406755, + "loss": 1.1826, + "step": 914 + }, + { + "epoch": 0.16292735042735043, + "grad_norm": 0.49433308839797974, + "learning_rate": 0.0001992163047293378, + "loss": 1.3552, + "step": 915 + }, + { + "epoch": 0.1631054131054131, + "grad_norm": 0.4985002875328064, + "learning_rate": 0.0001992145547740627, + "loss": 1.2639, + "step": 916 + }, + { + "epoch": 0.1632834757834758, + "grad_norm": 0.40348032116889954, + "learning_rate": 0.00019921280287488448, + "loss": 1.1731, + "step": 917 + }, + { + "epoch": 0.16346153846153846, + "grad_norm": 0.5166002511978149, + "learning_rate": 0.0001992110490318375, + "loss": 1.0692, + "step": 918 + }, + { + "epoch": 0.16363960113960113, + "grad_norm": 0.44233468174934387, + "learning_rate": 0.00019920929324495615, + "loss": 1.0488, + "step": 919 + }, + { + "epoch": 0.16381766381766383, + "grad_norm": 0.43709903955459595, + "learning_rate": 0.00019920753551427476, + "loss": 0.8884, + "step": 920 + }, + { + "epoch": 0.1639957264957265, + "grad_norm": 0.4054167568683624, + "learning_rate": 0.00019920577583982778, + "loss": 0.9872, + "step": 921 + }, + { + "epoch": 0.16417378917378916, + "grad_norm": 0.4657362997531891, + "learning_rate": 0.0001992040142216497, + "loss": 1.4402, + "step": 922 + }, + { + "epoch": 0.16435185185185186, + "grad_norm": 0.42550426721572876, + "learning_rate": 0.0001992022506597751, + "loss": 1.0456, + "step": 923 + }, + { + "epoch": 0.16452991452991453, + "grad_norm": 0.49346762895584106, + "learning_rate": 0.00019920048515423842, + "loss": 1.527, + "step": 924 + }, + { + "epoch": 0.1647079772079772, + "grad_norm": 0.3970337510108948, + "learning_rate": 0.0001991987177050743, + "loss": 1.0363, + "step": 925 + }, + { + "epoch": 0.1648860398860399, + "grad_norm": 0.4027378559112549, + "learning_rate": 0.0001991969483123174, + "loss": 0.8416, + "step": 926 + }, + { + "epoch": 0.16506410256410256, + "grad_norm": 0.4181644916534424, + "learning_rate": 0.00019919517697600237, + "loss": 1.2253, + "step": 927 + }, + { + "epoch": 0.16524216524216523, + "grad_norm": 0.43686383962631226, + "learning_rate": 0.0001991934036961639, + "loss": 1.0808, + "step": 928 + }, + { + "epoch": 0.16542022792022792, + "grad_norm": 0.4242876172065735, + "learning_rate": 0.0001991916284728367, + "loss": 0.9483, + "step": 929 + }, + { + "epoch": 0.1655982905982906, + "grad_norm": 0.3690609037876129, + "learning_rate": 0.00019918985130605563, + "loss": 0.9495, + "step": 930 + }, + { + "epoch": 0.1657763532763533, + "grad_norm": 0.42184555530548096, + "learning_rate": 0.00019918807219585546, + "loss": 1.0966, + "step": 931 + }, + { + "epoch": 0.16595441595441596, + "grad_norm": 0.4342746138572693, + "learning_rate": 0.00019918629114227106, + "loss": 1.0875, + "step": 932 + }, + { + "epoch": 0.16613247863247863, + "grad_norm": 0.4191494286060333, + "learning_rate": 0.00019918450814533737, + "loss": 1.0777, + "step": 933 + }, + { + "epoch": 0.16631054131054132, + "grad_norm": 0.37124550342559814, + "learning_rate": 0.00019918272320508922, + "loss": 1.0131, + "step": 934 + }, + { + "epoch": 0.166488603988604, + "grad_norm": 0.4475722014904022, + "learning_rate": 0.00019918093632156168, + "loss": 1.1185, + "step": 935 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.4629058241844177, + "learning_rate": 0.0001991791474947897, + "loss": 1.0353, + "step": 936 + }, + { + "epoch": 0.16684472934472935, + "grad_norm": 0.48192909359931946, + "learning_rate": 0.00019917735672480834, + "loss": 1.1628, + "step": 937 + }, + { + "epoch": 0.16702279202279202, + "grad_norm": 0.5542252063751221, + "learning_rate": 0.00019917556401165273, + "loss": 1.3133, + "step": 938 + }, + { + "epoch": 0.1672008547008547, + "grad_norm": 0.4172651171684265, + "learning_rate": 0.00019917376935535796, + "loss": 1.1733, + "step": 939 + }, + { + "epoch": 0.16737891737891739, + "grad_norm": 0.4424920380115509, + "learning_rate": 0.0001991719727559592, + "loss": 1.0262, + "step": 940 + }, + { + "epoch": 0.16755698005698005, + "grad_norm": 0.4551742970943451, + "learning_rate": 0.00019917017421349162, + "loss": 1.0883, + "step": 941 + }, + { + "epoch": 0.16773504273504272, + "grad_norm": 0.45929640531539917, + "learning_rate": 0.00019916837372799048, + "loss": 1.1836, + "step": 942 + }, + { + "epoch": 0.16791310541310542, + "grad_norm": 0.4609353542327881, + "learning_rate": 0.0001991665712994911, + "loss": 1.0682, + "step": 943 + }, + { + "epoch": 0.16809116809116809, + "grad_norm": 0.42617303133010864, + "learning_rate": 0.00019916476692802873, + "loss": 1.074, + "step": 944 + }, + { + "epoch": 0.16826923076923078, + "grad_norm": 0.41919493675231934, + "learning_rate": 0.00019916296061363875, + "loss": 1.0969, + "step": 945 + }, + { + "epoch": 0.16844729344729345, + "grad_norm": 0.450979083776474, + "learning_rate": 0.00019916115235635656, + "loss": 1.1686, + "step": 946 + }, + { + "epoch": 0.16862535612535612, + "grad_norm": 0.42166751623153687, + "learning_rate": 0.00019915934215621758, + "loss": 0.9273, + "step": 947 + }, + { + "epoch": 0.16880341880341881, + "grad_norm": 0.4404160976409912, + "learning_rate": 0.00019915753001325729, + "loss": 1.1663, + "step": 948 + }, + { + "epoch": 0.16898148148148148, + "grad_norm": 0.42025226354599, + "learning_rate": 0.0001991557159275111, + "loss": 0.9433, + "step": 949 + }, + { + "epoch": 0.16915954415954415, + "grad_norm": 0.4277796745300293, + "learning_rate": 0.00019915389989901474, + "loss": 0.8475, + "step": 950 + }, + { + "epoch": 0.16933760683760685, + "grad_norm": 0.5162755250930786, + "learning_rate": 0.00019915208192780365, + "loss": 1.1155, + "step": 951 + }, + { + "epoch": 0.16951566951566951, + "grad_norm": 0.4214856028556824, + "learning_rate": 0.00019915026201391346, + "loss": 1.173, + "step": 952 + }, + { + "epoch": 0.16969373219373218, + "grad_norm": 0.4713292419910431, + "learning_rate": 0.00019914844015737985, + "loss": 1.1615, + "step": 953 + }, + { + "epoch": 0.16987179487179488, + "grad_norm": 0.461179256439209, + "learning_rate": 0.00019914661635823854, + "loss": 1.1169, + "step": 954 + }, + { + "epoch": 0.17004985754985755, + "grad_norm": 0.46200552582740784, + "learning_rate": 0.00019914479061652527, + "loss": 1.0274, + "step": 955 + }, + { + "epoch": 0.17022792022792022, + "grad_norm": 0.40968334674835205, + "learning_rate": 0.00019914296293227572, + "loss": 1.066, + "step": 956 + }, + { + "epoch": 0.1704059829059829, + "grad_norm": 0.40877434611320496, + "learning_rate": 0.0001991411333055258, + "loss": 1.1595, + "step": 957 + }, + { + "epoch": 0.17058404558404558, + "grad_norm": 0.42940187454223633, + "learning_rate": 0.00019913930173631132, + "loss": 1.0364, + "step": 958 + }, + { + "epoch": 0.17076210826210828, + "grad_norm": 0.49648910760879517, + "learning_rate": 0.00019913746822466819, + "loss": 1.0763, + "step": 959 + }, + { + "epoch": 0.17094017094017094, + "grad_norm": 0.4353426396846771, + "learning_rate": 0.00019913563277063228, + "loss": 0.9698, + "step": 960 + }, + { + "epoch": 0.1711182336182336, + "grad_norm": 0.45079681277275085, + "learning_rate": 0.00019913379537423958, + "loss": 1.2244, + "step": 961 + }, + { + "epoch": 0.1712962962962963, + "grad_norm": 0.4276828467845917, + "learning_rate": 0.00019913195603552607, + "loss": 0.9976, + "step": 962 + }, + { + "epoch": 0.17147435897435898, + "grad_norm": 0.41122403740882874, + "learning_rate": 0.00019913011475452785, + "loss": 1.0077, + "step": 963 + }, + { + "epoch": 0.17165242165242164, + "grad_norm": 0.43170276284217834, + "learning_rate": 0.00019912827153128096, + "loss": 1.1402, + "step": 964 + }, + { + "epoch": 0.17183048433048434, + "grad_norm": 0.37950268387794495, + "learning_rate": 0.0001991264263658215, + "loss": 0.9818, + "step": 965 + }, + { + "epoch": 0.172008547008547, + "grad_norm": 0.477333128452301, + "learning_rate": 0.00019912457925818562, + "loss": 1.1756, + "step": 966 + }, + { + "epoch": 0.17218660968660968, + "grad_norm": 0.4326401352882385, + "learning_rate": 0.00019912273020840954, + "loss": 1.3718, + "step": 967 + }, + { + "epoch": 0.17236467236467237, + "grad_norm": 0.37711042165756226, + "learning_rate": 0.00019912087921652945, + "loss": 0.9011, + "step": 968 + }, + { + "epoch": 0.17254273504273504, + "grad_norm": 0.50013667345047, + "learning_rate": 0.00019911902628258162, + "loss": 1.1163, + "step": 969 + }, + { + "epoch": 0.1727207977207977, + "grad_norm": 0.41913339495658875, + "learning_rate": 0.0001991171714066024, + "loss": 1.2614, + "step": 970 + }, + { + "epoch": 0.1728988603988604, + "grad_norm": 0.4075855612754822, + "learning_rate": 0.00019911531458862813, + "loss": 0.8984, + "step": 971 + }, + { + "epoch": 0.17307692307692307, + "grad_norm": 0.40277954936027527, + "learning_rate": 0.00019911345582869513, + "loss": 1.0851, + "step": 972 + }, + { + "epoch": 0.17325498575498577, + "grad_norm": 0.4312847852706909, + "learning_rate": 0.00019911159512683987, + "loss": 1.1273, + "step": 973 + }, + { + "epoch": 0.17343304843304844, + "grad_norm": 0.40303611755371094, + "learning_rate": 0.0001991097324830988, + "loss": 0.9645, + "step": 974 + }, + { + "epoch": 0.1736111111111111, + "grad_norm": 0.45560577511787415, + "learning_rate": 0.00019910786789750838, + "loss": 1.0864, + "step": 975 + }, + { + "epoch": 0.1737891737891738, + "grad_norm": 0.43775680661201477, + "learning_rate": 0.00019910600137010517, + "loss": 1.028, + "step": 976 + }, + { + "epoch": 0.17396723646723647, + "grad_norm": 0.3917224407196045, + "learning_rate": 0.00019910413290092572, + "loss": 1.0491, + "step": 977 + }, + { + "epoch": 0.17414529914529914, + "grad_norm": 0.4068751037120819, + "learning_rate": 0.0001991022624900067, + "loss": 1.0476, + "step": 978 + }, + { + "epoch": 0.17432336182336183, + "grad_norm": 0.4463370144367218, + "learning_rate": 0.0001991003901373847, + "loss": 1.0612, + "step": 979 + }, + { + "epoch": 0.1745014245014245, + "grad_norm": 0.46949052810668945, + "learning_rate": 0.0001990985158430964, + "loss": 1.3099, + "step": 980 + }, + { + "epoch": 0.17467948717948717, + "grad_norm": 0.4250012934207916, + "learning_rate": 0.00019909663960717856, + "loss": 0.9903, + "step": 981 + }, + { + "epoch": 0.17485754985754987, + "grad_norm": 0.5293903946876526, + "learning_rate": 0.0001990947614296679, + "loss": 0.9908, + "step": 982 + }, + { + "epoch": 0.17503561253561253, + "grad_norm": 0.3838284909725189, + "learning_rate": 0.0001990928813106013, + "loss": 0.716, + "step": 983 + }, + { + "epoch": 0.1752136752136752, + "grad_norm": 0.4597751200199127, + "learning_rate": 0.0001990909992500155, + "loss": 1.0126, + "step": 984 + }, + { + "epoch": 0.1753917378917379, + "grad_norm": 0.4844081699848175, + "learning_rate": 0.0001990891152479474, + "loss": 1.1043, + "step": 985 + }, + { + "epoch": 0.17556980056980057, + "grad_norm": 0.4763399660587311, + "learning_rate": 0.00019908722930443392, + "loss": 1.019, + "step": 986 + }, + { + "epoch": 0.17574786324786323, + "grad_norm": 0.4670077860355377, + "learning_rate": 0.00019908534141951204, + "loss": 1.1382, + "step": 987 + }, + { + "epoch": 0.17592592592592593, + "grad_norm": 0.39372730255126953, + "learning_rate": 0.00019908345159321873, + "loss": 1.1219, + "step": 988 + }, + { + "epoch": 0.1761039886039886, + "grad_norm": 0.41869843006134033, + "learning_rate": 0.00019908155982559098, + "loss": 0.9461, + "step": 989 + }, + { + "epoch": 0.1762820512820513, + "grad_norm": 0.4398406147956848, + "learning_rate": 0.00019907966611666593, + "loss": 1.1328, + "step": 990 + }, + { + "epoch": 0.17646011396011396, + "grad_norm": 0.4315733015537262, + "learning_rate": 0.0001990777704664806, + "loss": 1.0974, + "step": 991 + }, + { + "epoch": 0.17663817663817663, + "grad_norm": 0.42859575152397156, + "learning_rate": 0.00019907587287507222, + "loss": 1.2637, + "step": 992 + }, + { + "epoch": 0.17681623931623933, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.0001990739733424779, + "loss": 1.0699, + "step": 993 + }, + { + "epoch": 0.176994301994302, + "grad_norm": 0.4443826973438263, + "learning_rate": 0.00019907207186873488, + "loss": 1.0547, + "step": 994 + }, + { + "epoch": 0.17717236467236466, + "grad_norm": 0.4108099937438965, + "learning_rate": 0.00019907016845388043, + "loss": 1.1401, + "step": 995 + }, + { + "epoch": 0.17735042735042736, + "grad_norm": 0.4474675953388214, + "learning_rate": 0.00019906826309795182, + "loss": 1.0712, + "step": 996 + }, + { + "epoch": 0.17752849002849003, + "grad_norm": 0.4149756133556366, + "learning_rate": 0.00019906635580098638, + "loss": 0.9585, + "step": 997 + }, + { + "epoch": 0.1777065527065527, + "grad_norm": 0.4875968098640442, + "learning_rate": 0.00019906444656302152, + "loss": 1.0659, + "step": 998 + }, + { + "epoch": 0.1778846153846154, + "grad_norm": 0.5494784116744995, + "learning_rate": 0.0001990625353840946, + "loss": 1.2858, + "step": 999 + }, + { + "epoch": 0.17806267806267806, + "grad_norm": 0.425062358379364, + "learning_rate": 0.0001990606222642431, + "loss": 1.1826, + "step": 1000 + }, + { + "epoch": 0.17824074074074073, + "grad_norm": 0.3890725374221802, + "learning_rate": 0.00019905870720350445, + "loss": 0.9568, + "step": 1001 + }, + { + "epoch": 0.17841880341880342, + "grad_norm": 0.3884070813655853, + "learning_rate": 0.00019905679020191624, + "loss": 0.9674, + "step": 1002 + }, + { + "epoch": 0.1785968660968661, + "grad_norm": 0.49496129155158997, + "learning_rate": 0.00019905487125951597, + "loss": 0.9143, + "step": 1003 + }, + { + "epoch": 0.1787749287749288, + "grad_norm": 0.43448135256767273, + "learning_rate": 0.00019905295037634128, + "loss": 1.2677, + "step": 1004 + }, + { + "epoch": 0.17895299145299146, + "grad_norm": 0.47327905893325806, + "learning_rate": 0.00019905102755242982, + "loss": 0.9089, + "step": 1005 + }, + { + "epoch": 0.17913105413105412, + "grad_norm": 0.4962378442287445, + "learning_rate": 0.00019904910278781922, + "loss": 1.1748, + "step": 1006 + }, + { + "epoch": 0.17930911680911682, + "grad_norm": 0.4343934655189514, + "learning_rate": 0.0001990471760825472, + "loss": 1.2176, + "step": 1007 + }, + { + "epoch": 0.1794871794871795, + "grad_norm": 0.4695793092250824, + "learning_rate": 0.0001990452474366515, + "loss": 1.1822, + "step": 1008 + }, + { + "epoch": 0.17966524216524216, + "grad_norm": 0.4156060516834259, + "learning_rate": 0.00019904331685016995, + "loss": 0.8231, + "step": 1009 + }, + { + "epoch": 0.17984330484330485, + "grad_norm": 0.5068191885948181, + "learning_rate": 0.00019904138432314035, + "loss": 1.1363, + "step": 1010 + }, + { + "epoch": 0.18002136752136752, + "grad_norm": 0.5189786553382874, + "learning_rate": 0.00019903944985560058, + "loss": 1.3131, + "step": 1011 + }, + { + "epoch": 0.1801994301994302, + "grad_norm": 0.5126828551292419, + "learning_rate": 0.00019903751344758848, + "loss": 1.0305, + "step": 1012 + }, + { + "epoch": 0.18037749287749288, + "grad_norm": 0.41045933961868286, + "learning_rate": 0.00019903557509914205, + "loss": 1.2726, + "step": 1013 + }, + { + "epoch": 0.18055555555555555, + "grad_norm": 0.4141713082790375, + "learning_rate": 0.0001990336348102993, + "loss": 0.9606, + "step": 1014 + }, + { + "epoch": 0.18073361823361822, + "grad_norm": 0.42652079463005066, + "learning_rate": 0.00019903169258109812, + "loss": 1.0235, + "step": 1015 + }, + { + "epoch": 0.18091168091168092, + "grad_norm": 0.42098379135131836, + "learning_rate": 0.0001990297484115767, + "loss": 1.0602, + "step": 1016 + }, + { + "epoch": 0.18108974358974358, + "grad_norm": 0.49920013546943665, + "learning_rate": 0.0001990278023017731, + "loss": 1.3322, + "step": 1017 + }, + { + "epoch": 0.18126780626780628, + "grad_norm": 0.412304550409317, + "learning_rate": 0.00019902585425172537, + "loss": 1.1011, + "step": 1018 + }, + { + "epoch": 0.18144586894586895, + "grad_norm": 0.44226935505867004, + "learning_rate": 0.00019902390426147177, + "loss": 0.9777, + "step": 1019 + }, + { + "epoch": 0.18162393162393162, + "grad_norm": 0.4685269594192505, + "learning_rate": 0.00019902195233105046, + "loss": 1.3587, + "step": 1020 + }, + { + "epoch": 0.1818019943019943, + "grad_norm": 0.4500584304332733, + "learning_rate": 0.00019901999846049968, + "loss": 0.9888, + "step": 1021 + }, + { + "epoch": 0.18198005698005698, + "grad_norm": 0.48566994071006775, + "learning_rate": 0.00019901804264985774, + "loss": 1.2364, + "step": 1022 + }, + { + "epoch": 0.18215811965811965, + "grad_norm": 0.4063156247138977, + "learning_rate": 0.00019901608489916294, + "loss": 1.2224, + "step": 1023 + }, + { + "epoch": 0.18233618233618235, + "grad_norm": 0.471276193857193, + "learning_rate": 0.00019901412520845367, + "loss": 0.9926, + "step": 1024 + }, + { + "epoch": 0.182514245014245, + "grad_norm": 0.5165421366691589, + "learning_rate": 0.00019901216357776829, + "loss": 0.9595, + "step": 1025 + }, + { + "epoch": 0.18269230769230768, + "grad_norm": 0.4746754467487335, + "learning_rate": 0.0001990102000071452, + "loss": 1.2057, + "step": 1026 + }, + { + "epoch": 0.18287037037037038, + "grad_norm": 0.44803035259246826, + "learning_rate": 0.00019900823449662297, + "loss": 1.2114, + "step": 1027 + }, + { + "epoch": 0.18304843304843305, + "grad_norm": 0.47256240248680115, + "learning_rate": 0.00019900626704624005, + "loss": 1.112, + "step": 1028 + }, + { + "epoch": 0.18322649572649571, + "grad_norm": 0.4253387153148651, + "learning_rate": 0.000199004297656035, + "loss": 0.9899, + "step": 1029 + }, + { + "epoch": 0.1834045584045584, + "grad_norm": 0.44958099722862244, + "learning_rate": 0.00019900232632604636, + "loss": 1.1445, + "step": 1030 + }, + { + "epoch": 0.18358262108262108, + "grad_norm": 0.5296537280082703, + "learning_rate": 0.00019900035305631285, + "loss": 1.2502, + "step": 1031 + }, + { + "epoch": 0.18376068376068377, + "grad_norm": 0.5057148933410645, + "learning_rate": 0.00019899837784687302, + "loss": 1.1426, + "step": 1032 + }, + { + "epoch": 0.18393874643874644, + "grad_norm": 0.41463762521743774, + "learning_rate": 0.00019899640069776566, + "loss": 1.1854, + "step": 1033 + }, + { + "epoch": 0.1841168091168091, + "grad_norm": 0.45800045132637024, + "learning_rate": 0.00019899442160902945, + "loss": 1.2438, + "step": 1034 + }, + { + "epoch": 0.1842948717948718, + "grad_norm": 0.43450453877449036, + "learning_rate": 0.00019899244058070324, + "loss": 1.0598, + "step": 1035 + }, + { + "epoch": 0.18447293447293447, + "grad_norm": 0.4141148626804352, + "learning_rate": 0.00019899045761282577, + "loss": 1.0465, + "step": 1036 + }, + { + "epoch": 0.18465099715099714, + "grad_norm": 0.3938458263874054, + "learning_rate": 0.0001989884727054359, + "loss": 1.0142, + "step": 1037 + }, + { + "epoch": 0.18482905982905984, + "grad_norm": 0.43898263573646545, + "learning_rate": 0.00019898648585857257, + "loss": 0.9212, + "step": 1038 + }, + { + "epoch": 0.1850071225071225, + "grad_norm": 0.4425487816333771, + "learning_rate": 0.00019898449707227465, + "loss": 1.2987, + "step": 1039 + }, + { + "epoch": 0.18518518518518517, + "grad_norm": 0.4537975490093231, + "learning_rate": 0.00019898250634658115, + "loss": 1.2023, + "step": 1040 + }, + { + "epoch": 0.18536324786324787, + "grad_norm": 0.4107198119163513, + "learning_rate": 0.00019898051368153104, + "loss": 0.8443, + "step": 1041 + }, + { + "epoch": 0.18554131054131054, + "grad_norm": 0.4389404058456421, + "learning_rate": 0.0001989785190771634, + "loss": 1.0502, + "step": 1042 + }, + { + "epoch": 0.1857193732193732, + "grad_norm": 0.4288824796676636, + "learning_rate": 0.00019897652253351726, + "loss": 1.01, + "step": 1043 + }, + { + "epoch": 0.1858974358974359, + "grad_norm": 0.50815349817276, + "learning_rate": 0.00019897452405063178, + "loss": 1.0308, + "step": 1044 + }, + { + "epoch": 0.18607549857549857, + "grad_norm": 0.45252710580825806, + "learning_rate": 0.0001989725236285461, + "loss": 1.0967, + "step": 1045 + }, + { + "epoch": 0.18625356125356127, + "grad_norm": 0.45049402117729187, + "learning_rate": 0.00019897052126729943, + "loss": 1.0141, + "step": 1046 + }, + { + "epoch": 0.18643162393162394, + "grad_norm": 0.49637508392333984, + "learning_rate": 0.00019896851696693098, + "loss": 1.0997, + "step": 1047 + }, + { + "epoch": 0.1866096866096866, + "grad_norm": 0.4465886056423187, + "learning_rate": 0.00019896651072748005, + "loss": 1.1415, + "step": 1048 + }, + { + "epoch": 0.1867877492877493, + "grad_norm": 0.5309500694274902, + "learning_rate": 0.00019896450254898592, + "loss": 1.1028, + "step": 1049 + }, + { + "epoch": 0.18696581196581197, + "grad_norm": 0.3516653776168823, + "learning_rate": 0.00019896249243148793, + "loss": 0.9841, + "step": 1050 + }, + { + "epoch": 0.18714387464387464, + "grad_norm": 0.4529176950454712, + "learning_rate": 0.0001989604803750255, + "loss": 1.1335, + "step": 1051 + }, + { + "epoch": 0.18732193732193733, + "grad_norm": 0.47694942355155945, + "learning_rate": 0.000198958466379638, + "loss": 1.2383, + "step": 1052 + }, + { + "epoch": 0.1875, + "grad_norm": 0.5524206757545471, + "learning_rate": 0.0001989564504453649, + "loss": 1.3668, + "step": 1053 + }, + { + "epoch": 0.18767806267806267, + "grad_norm": 0.39203691482543945, + "learning_rate": 0.00019895443257224576, + "loss": 1.2203, + "step": 1054 + }, + { + "epoch": 0.18785612535612536, + "grad_norm": 0.4164120852947235, + "learning_rate": 0.00019895241276032005, + "loss": 0.8954, + "step": 1055 + }, + { + "epoch": 0.18803418803418803, + "grad_norm": 0.41217970848083496, + "learning_rate": 0.0001989503910096274, + "loss": 1.0238, + "step": 1056 + }, + { + "epoch": 0.1882122507122507, + "grad_norm": 0.44038307666778564, + "learning_rate": 0.00019894836732020735, + "loss": 0.8159, + "step": 1057 + }, + { + "epoch": 0.1883903133903134, + "grad_norm": 0.45780670642852783, + "learning_rate": 0.0001989463416920996, + "loss": 1.2864, + "step": 1058 + }, + { + "epoch": 0.18856837606837606, + "grad_norm": 0.5197559595108032, + "learning_rate": 0.00019894431412534384, + "loss": 1.0756, + "step": 1059 + }, + { + "epoch": 0.18874643874643873, + "grad_norm": 0.43283385038375854, + "learning_rate": 0.00019894228461997979, + "loss": 1.0642, + "step": 1060 + }, + { + "epoch": 0.18892450142450143, + "grad_norm": 0.4657376706600189, + "learning_rate": 0.00019894025317604717, + "loss": 1.1159, + "step": 1061 + }, + { + "epoch": 0.1891025641025641, + "grad_norm": 0.4474908113479614, + "learning_rate": 0.00019893821979358588, + "loss": 1.2006, + "step": 1062 + }, + { + "epoch": 0.1892806267806268, + "grad_norm": 0.43878164887428284, + "learning_rate": 0.00019893618447263566, + "loss": 1.1599, + "step": 1063 + }, + { + "epoch": 0.18945868945868946, + "grad_norm": 0.4598735272884369, + "learning_rate": 0.00019893414721323645, + "loss": 1.3346, + "step": 1064 + }, + { + "epoch": 0.18963675213675213, + "grad_norm": 0.3947420120239258, + "learning_rate": 0.00019893210801542812, + "loss": 1.1201, + "step": 1065 + }, + { + "epoch": 0.18981481481481483, + "grad_norm": 0.3401558995246887, + "learning_rate": 0.00019893006687925064, + "loss": 0.7568, + "step": 1066 + }, + { + "epoch": 0.1899928774928775, + "grad_norm": 0.4400341808795929, + "learning_rate": 0.00019892802380474405, + "loss": 1.1706, + "step": 1067 + }, + { + "epoch": 0.19017094017094016, + "grad_norm": 0.42394164204597473, + "learning_rate": 0.00019892597879194829, + "loss": 1.0163, + "step": 1068 + }, + { + "epoch": 0.19034900284900286, + "grad_norm": 0.42904096841812134, + "learning_rate": 0.00019892393184090353, + "loss": 0.9193, + "step": 1069 + }, + { + "epoch": 0.19052706552706553, + "grad_norm": 0.497601181268692, + "learning_rate": 0.00019892188295164977, + "loss": 1.0377, + "step": 1070 + }, + { + "epoch": 0.1907051282051282, + "grad_norm": 0.4536020755767822, + "learning_rate": 0.00019891983212422723, + "loss": 1.0946, + "step": 1071 + }, + { + "epoch": 0.1908831908831909, + "grad_norm": 0.44916942715644836, + "learning_rate": 0.00019891777935867607, + "loss": 1.0563, + "step": 1072 + }, + { + "epoch": 0.19106125356125356, + "grad_norm": 0.4256889820098877, + "learning_rate": 0.0001989157246550365, + "loss": 1.0988, + "step": 1073 + }, + { + "epoch": 0.19123931623931623, + "grad_norm": 0.5559163689613342, + "learning_rate": 0.0001989136680133488, + "loss": 0.9155, + "step": 1074 + }, + { + "epoch": 0.19141737891737892, + "grad_norm": 0.391804963350296, + "learning_rate": 0.00019891160943365322, + "loss": 0.9314, + "step": 1075 + }, + { + "epoch": 0.1915954415954416, + "grad_norm": 0.4535716474056244, + "learning_rate": 0.00019890954891599015, + "loss": 1.0768, + "step": 1076 + }, + { + "epoch": 0.19177350427350429, + "grad_norm": 0.46770521998405457, + "learning_rate": 0.00019890748646039991, + "loss": 0.8406, + "step": 1077 + }, + { + "epoch": 0.19195156695156695, + "grad_norm": 0.4875394403934479, + "learning_rate": 0.00019890542206692295, + "loss": 1.1055, + "step": 1078 + }, + { + "epoch": 0.19212962962962962, + "grad_norm": 0.5072727203369141, + "learning_rate": 0.0001989033557355997, + "loss": 1.3093, + "step": 1079 + }, + { + "epoch": 0.19230769230769232, + "grad_norm": 0.4419287443161011, + "learning_rate": 0.00019890128746647068, + "loss": 1.1916, + "step": 1080 + }, + { + "epoch": 0.192485754985755, + "grad_norm": 0.45803651213645935, + "learning_rate": 0.00019889921725957637, + "loss": 1.2579, + "step": 1081 + }, + { + "epoch": 0.19266381766381765, + "grad_norm": 0.4832262098789215, + "learning_rate": 0.0001988971451149573, + "loss": 1.3217, + "step": 1082 + }, + { + "epoch": 0.19284188034188035, + "grad_norm": 0.4819786250591278, + "learning_rate": 0.00019889507103265416, + "loss": 1.0979, + "step": 1083 + }, + { + "epoch": 0.19301994301994302, + "grad_norm": 0.49360713362693787, + "learning_rate": 0.0001988929950127075, + "loss": 1.0987, + "step": 1084 + }, + { + "epoch": 0.1931980056980057, + "grad_norm": 0.44209200143814087, + "learning_rate": 0.00019889091705515806, + "loss": 1.2616, + "step": 1085 + }, + { + "epoch": 0.19337606837606838, + "grad_norm": 0.41626206040382385, + "learning_rate": 0.00019888883716004654, + "loss": 1.0922, + "step": 1086 + }, + { + "epoch": 0.19355413105413105, + "grad_norm": 0.4916635751724243, + "learning_rate": 0.00019888675532741366, + "loss": 0.9331, + "step": 1087 + }, + { + "epoch": 0.19373219373219372, + "grad_norm": 0.4493125379085541, + "learning_rate": 0.00019888467155730025, + "loss": 1.1261, + "step": 1088 + }, + { + "epoch": 0.19391025641025642, + "grad_norm": 0.3755671977996826, + "learning_rate": 0.00019888258584974708, + "loss": 0.9821, + "step": 1089 + }, + { + "epoch": 0.19408831908831908, + "grad_norm": 0.41917556524276733, + "learning_rate": 0.00019888049820479507, + "loss": 1.251, + "step": 1090 + }, + { + "epoch": 0.19426638176638178, + "grad_norm": 0.46184420585632324, + "learning_rate": 0.0001988784086224851, + "loss": 1.1731, + "step": 1091 + }, + { + "epoch": 0.19444444444444445, + "grad_norm": 0.4783691465854645, + "learning_rate": 0.00019887631710285812, + "loss": 1.1635, + "step": 1092 + }, + { + "epoch": 0.19462250712250712, + "grad_norm": 0.4710482060909271, + "learning_rate": 0.00019887422364595512, + "loss": 1.0229, + "step": 1093 + }, + { + "epoch": 0.1948005698005698, + "grad_norm": 0.4738706648349762, + "learning_rate": 0.00019887212825181707, + "loss": 1.128, + "step": 1094 + }, + { + "epoch": 0.19497863247863248, + "grad_norm": 0.45665010809898376, + "learning_rate": 0.00019887003092048508, + "loss": 1.0425, + "step": 1095 + }, + { + "epoch": 0.19515669515669515, + "grad_norm": 0.42740485072135925, + "learning_rate": 0.0001988679316520002, + "loss": 1.0738, + "step": 1096 + }, + { + "epoch": 0.19533475783475784, + "grad_norm": 0.5977092385292053, + "learning_rate": 0.0001988658304464036, + "loss": 1.2687, + "step": 1097 + }, + { + "epoch": 0.1955128205128205, + "grad_norm": 0.4411074221134186, + "learning_rate": 0.0001988637273037364, + "loss": 1.287, + "step": 1098 + }, + { + "epoch": 0.19569088319088318, + "grad_norm": 0.4409518539905548, + "learning_rate": 0.00019886162222403986, + "loss": 1.0515, + "step": 1099 + }, + { + "epoch": 0.19586894586894588, + "grad_norm": 0.4926736652851105, + "learning_rate": 0.0001988595152073552, + "loss": 1.1388, + "step": 1100 + }, + { + "epoch": 0.19604700854700854, + "grad_norm": 0.4607115387916565, + "learning_rate": 0.00019885740625372368, + "loss": 0.9803, + "step": 1101 + }, + { + "epoch": 0.1962250712250712, + "grad_norm": 0.4725342094898224, + "learning_rate": 0.0001988552953631867, + "loss": 1.199, + "step": 1102 + }, + { + "epoch": 0.1964031339031339, + "grad_norm": 0.48014503717422485, + "learning_rate": 0.00019885318253578548, + "loss": 1.1868, + "step": 1103 + }, + { + "epoch": 0.19658119658119658, + "grad_norm": 0.3872644603252411, + "learning_rate": 0.00019885106777156155, + "loss": 0.9182, + "step": 1104 + }, + { + "epoch": 0.19675925925925927, + "grad_norm": 0.4737720787525177, + "learning_rate": 0.00019884895107055627, + "loss": 1.1513, + "step": 1105 + }, + { + "epoch": 0.19693732193732194, + "grad_norm": 0.4144562780857086, + "learning_rate": 0.00019884683243281116, + "loss": 1.1711, + "step": 1106 + }, + { + "epoch": 0.1971153846153846, + "grad_norm": 0.4672079384326935, + "learning_rate": 0.00019884471185836769, + "loss": 1.0386, + "step": 1107 + }, + { + "epoch": 0.1972934472934473, + "grad_norm": 0.4558824598789215, + "learning_rate": 0.0001988425893472674, + "loss": 1.0535, + "step": 1108 + }, + { + "epoch": 0.19747150997150997, + "grad_norm": 0.5149834752082825, + "learning_rate": 0.00019884046489955192, + "loss": 1.0296, + "step": 1109 + }, + { + "epoch": 0.19764957264957264, + "grad_norm": 0.43444496393203735, + "learning_rate": 0.00019883833851526287, + "loss": 1.1475, + "step": 1110 + }, + { + "epoch": 0.19782763532763534, + "grad_norm": 0.46062374114990234, + "learning_rate": 0.00019883621019444188, + "loss": 1.183, + "step": 1111 + }, + { + "epoch": 0.198005698005698, + "grad_norm": 0.4893282949924469, + "learning_rate": 0.00019883407993713065, + "loss": 1.3733, + "step": 1112 + }, + { + "epoch": 0.19818376068376067, + "grad_norm": 0.5434843897819519, + "learning_rate": 0.00019883194774337096, + "loss": 1.2505, + "step": 1113 + }, + { + "epoch": 0.19836182336182337, + "grad_norm": 0.4698035418987274, + "learning_rate": 0.00019882981361320456, + "loss": 1.0152, + "step": 1114 + }, + { + "epoch": 0.19853988603988604, + "grad_norm": 0.4582163989543915, + "learning_rate": 0.00019882767754667325, + "loss": 1.1718, + "step": 1115 + }, + { + "epoch": 0.1987179487179487, + "grad_norm": 0.48744696378707886, + "learning_rate": 0.0001988255395438189, + "loss": 1.2923, + "step": 1116 + }, + { + "epoch": 0.1988960113960114, + "grad_norm": 0.4172030985355377, + "learning_rate": 0.0001988233996046834, + "loss": 0.8098, + "step": 1117 + }, + { + "epoch": 0.19907407407407407, + "grad_norm": 0.4556557834148407, + "learning_rate": 0.00019882125772930867, + "loss": 0.9654, + "step": 1118 + }, + { + "epoch": 0.19925213675213677, + "grad_norm": 0.4363219141960144, + "learning_rate": 0.00019881911391773666, + "loss": 1.0333, + "step": 1119 + }, + { + "epoch": 0.19943019943019943, + "grad_norm": 0.4336536228656769, + "learning_rate": 0.0001988169681700094, + "loss": 1.091, + "step": 1120 + }, + { + "epoch": 0.1996082621082621, + "grad_norm": 0.42073166370391846, + "learning_rate": 0.00019881482048616893, + "loss": 0.9687, + "step": 1121 + }, + { + "epoch": 0.1997863247863248, + "grad_norm": 0.4330587685108185, + "learning_rate": 0.00019881267086625733, + "loss": 1.0512, + "step": 1122 + }, + { + "epoch": 0.19996438746438747, + "grad_norm": 0.4602276682853699, + "learning_rate": 0.0001988105193103167, + "loss": 1.1806, + "step": 1123 + }, + { + "epoch": 0.20014245014245013, + "grad_norm": 0.4271257817745209, + "learning_rate": 0.0001988083658183892, + "loss": 1.1079, + "step": 1124 + }, + { + "epoch": 0.20032051282051283, + "grad_norm": 0.35446426272392273, + "learning_rate": 0.00019880621039051707, + "loss": 0.6769, + "step": 1125 + }, + { + "epoch": 0.2004985754985755, + "grad_norm": 0.413753479719162, + "learning_rate": 0.00019880405302674244, + "loss": 1.1088, + "step": 1126 + }, + { + "epoch": 0.20067663817663817, + "grad_norm": 0.4423675835132599, + "learning_rate": 0.00019880189372710767, + "loss": 1.1371, + "step": 1127 + }, + { + "epoch": 0.20085470085470086, + "grad_norm": 0.41865605115890503, + "learning_rate": 0.00019879973249165502, + "loss": 1.0027, + "step": 1128 + }, + { + "epoch": 0.20103276353276353, + "grad_norm": 0.4109594225883484, + "learning_rate": 0.00019879756932042686, + "loss": 0.8734, + "step": 1129 + }, + { + "epoch": 0.2012108262108262, + "grad_norm": 0.42326363921165466, + "learning_rate": 0.00019879540421346555, + "loss": 0.9722, + "step": 1130 + }, + { + "epoch": 0.2013888888888889, + "grad_norm": 0.4601542055606842, + "learning_rate": 0.00019879323717081354, + "loss": 1.1251, + "step": 1131 + }, + { + "epoch": 0.20156695156695156, + "grad_norm": 0.4704367518424988, + "learning_rate": 0.00019879106819251327, + "loss": 0.9457, + "step": 1132 + }, + { + "epoch": 0.20174501424501423, + "grad_norm": 0.465023934841156, + "learning_rate": 0.00019878889727860724, + "loss": 0.9633, + "step": 1133 + }, + { + "epoch": 0.20192307692307693, + "grad_norm": 0.4572450518608093, + "learning_rate": 0.00019878672442913796, + "loss": 1.1965, + "step": 1134 + }, + { + "epoch": 0.2021011396011396, + "grad_norm": 0.4323410391807556, + "learning_rate": 0.00019878454964414807, + "loss": 1.1296, + "step": 1135 + }, + { + "epoch": 0.2022792022792023, + "grad_norm": 0.4513751268386841, + "learning_rate": 0.00019878237292368013, + "loss": 1.0571, + "step": 1136 + }, + { + "epoch": 0.20245726495726496, + "grad_norm": 0.45504096150398254, + "learning_rate": 0.00019878019426777677, + "loss": 1.0316, + "step": 1137 + }, + { + "epoch": 0.20263532763532763, + "grad_norm": 0.45715275406837463, + "learning_rate": 0.0001987780136764807, + "loss": 1.0528, + "step": 1138 + }, + { + "epoch": 0.20281339031339032, + "grad_norm": 0.4934465289115906, + "learning_rate": 0.00019877583114983466, + "loss": 1.3238, + "step": 1139 + }, + { + "epoch": 0.202991452991453, + "grad_norm": 0.4304082989692688, + "learning_rate": 0.0001987736466878814, + "loss": 1.1774, + "step": 1140 + }, + { + "epoch": 0.20316951566951566, + "grad_norm": 0.49721968173980713, + "learning_rate": 0.00019877146029066372, + "loss": 1.1767, + "step": 1141 + }, + { + "epoch": 0.20334757834757836, + "grad_norm": 0.3629468083381653, + "learning_rate": 0.00019876927195822445, + "loss": 0.8588, + "step": 1142 + }, + { + "epoch": 0.20352564102564102, + "grad_norm": 0.49310383200645447, + "learning_rate": 0.00019876708169060648, + "loss": 1.0588, + "step": 1143 + }, + { + "epoch": 0.2037037037037037, + "grad_norm": 0.4270328879356384, + "learning_rate": 0.00019876488948785271, + "loss": 1.1523, + "step": 1144 + }, + { + "epoch": 0.2038817663817664, + "grad_norm": 0.4559730887413025, + "learning_rate": 0.0001987626953500061, + "loss": 1.1736, + "step": 1145 + }, + { + "epoch": 0.20405982905982906, + "grad_norm": 0.5335259437561035, + "learning_rate": 0.00019876049927710962, + "loss": 0.991, + "step": 1146 + }, + { + "epoch": 0.20423789173789172, + "grad_norm": 0.43500083684921265, + "learning_rate": 0.0001987583012692063, + "loss": 1.0631, + "step": 1147 + }, + { + "epoch": 0.20441595441595442, + "grad_norm": 0.4135417938232422, + "learning_rate": 0.00019875610132633927, + "loss": 1.0896, + "step": 1148 + }, + { + "epoch": 0.2045940170940171, + "grad_norm": 0.4078896641731262, + "learning_rate": 0.00019875389944855153, + "loss": 1.0395, + "step": 1149 + }, + { + "epoch": 0.20477207977207978, + "grad_norm": 0.46612194180488586, + "learning_rate": 0.00019875169563588632, + "loss": 1.0541, + "step": 1150 + }, + { + "epoch": 0.20495014245014245, + "grad_norm": 0.5093224048614502, + "learning_rate": 0.00019874948988838674, + "loss": 1.1486, + "step": 1151 + }, + { + "epoch": 0.20512820512820512, + "grad_norm": 0.5079755187034607, + "learning_rate": 0.00019874728220609607, + "loss": 1.2614, + "step": 1152 + }, + { + "epoch": 0.20530626780626782, + "grad_norm": 0.43663498759269714, + "learning_rate": 0.0001987450725890575, + "loss": 1.0683, + "step": 1153 + }, + { + "epoch": 0.20548433048433049, + "grad_norm": 0.5029327273368835, + "learning_rate": 0.00019874286103731435, + "loss": 1.1934, + "step": 1154 + }, + { + "epoch": 0.20566239316239315, + "grad_norm": 0.48770397901535034, + "learning_rate": 0.00019874064755090999, + "loss": 1.1634, + "step": 1155 + }, + { + "epoch": 0.20584045584045585, + "grad_norm": 0.46826690435409546, + "learning_rate": 0.00019873843212988776, + "loss": 1.0621, + "step": 1156 + }, + { + "epoch": 0.20601851851851852, + "grad_norm": 0.4810047149658203, + "learning_rate": 0.00019873621477429105, + "loss": 1.0879, + "step": 1157 + }, + { + "epoch": 0.20619658119658119, + "grad_norm": 0.4769522249698639, + "learning_rate": 0.00019873399548416335, + "loss": 1.1365, + "step": 1158 + }, + { + "epoch": 0.20637464387464388, + "grad_norm": 0.4221782982349396, + "learning_rate": 0.00019873177425954806, + "loss": 1.1168, + "step": 1159 + }, + { + "epoch": 0.20655270655270655, + "grad_norm": 0.4084923565387726, + "learning_rate": 0.00019872955110048876, + "loss": 1.2364, + "step": 1160 + }, + { + "epoch": 0.20673076923076922, + "grad_norm": 0.4781704545021057, + "learning_rate": 0.00019872732600702904, + "loss": 1.19, + "step": 1161 + }, + { + "epoch": 0.20690883190883191, + "grad_norm": 0.3984242081642151, + "learning_rate": 0.0001987250989792124, + "loss": 1.0568, + "step": 1162 + }, + { + "epoch": 0.20708689458689458, + "grad_norm": 0.4601972997188568, + "learning_rate": 0.00019872287001708257, + "loss": 1.1625, + "step": 1163 + }, + { + "epoch": 0.20726495726495728, + "grad_norm": 0.4853581190109253, + "learning_rate": 0.00019872063912068316, + "loss": 1.2304, + "step": 1164 + }, + { + "epoch": 0.20744301994301995, + "grad_norm": 0.41779839992523193, + "learning_rate": 0.0001987184062900579, + "loss": 0.9807, + "step": 1165 + }, + { + "epoch": 0.20762108262108261, + "grad_norm": 0.4945356249809265, + "learning_rate": 0.00019871617152525056, + "loss": 1.1861, + "step": 1166 + }, + { + "epoch": 0.2077991452991453, + "grad_norm": 0.47432294487953186, + "learning_rate": 0.00019871393482630487, + "loss": 1.1448, + "step": 1167 + }, + { + "epoch": 0.20797720797720798, + "grad_norm": 0.44647398591041565, + "learning_rate": 0.00019871169619326473, + "loss": 1.096, + "step": 1168 + }, + { + "epoch": 0.20815527065527065, + "grad_norm": 0.4643072783946991, + "learning_rate": 0.00019870945562617393, + "loss": 1.1561, + "step": 1169 + }, + { + "epoch": 0.20833333333333334, + "grad_norm": 0.4544340968132019, + "learning_rate": 0.0001987072131250764, + "loss": 1.0764, + "step": 1170 + }, + { + "epoch": 0.208511396011396, + "grad_norm": 0.6036561727523804, + "learning_rate": 0.00019870496869001607, + "loss": 1.3961, + "step": 1171 + }, + { + "epoch": 0.20868945868945868, + "grad_norm": 0.41348758339881897, + "learning_rate": 0.00019870272232103695, + "loss": 1.2219, + "step": 1172 + }, + { + "epoch": 0.20886752136752137, + "grad_norm": 0.4184056222438812, + "learning_rate": 0.000198700474018183, + "loss": 1.1115, + "step": 1173 + }, + { + "epoch": 0.20904558404558404, + "grad_norm": 0.41920599341392517, + "learning_rate": 0.0001986982237814983, + "loss": 0.9207, + "step": 1174 + }, + { + "epoch": 0.2092236467236467, + "grad_norm": 0.4710249602794647, + "learning_rate": 0.00019869597161102694, + "loss": 1.1342, + "step": 1175 + }, + { + "epoch": 0.2094017094017094, + "grad_norm": 0.46897777915000916, + "learning_rate": 0.000198693717506813, + "loss": 0.983, + "step": 1176 + }, + { + "epoch": 0.20957977207977208, + "grad_norm": 0.4817039370536804, + "learning_rate": 0.00019869146146890074, + "loss": 1.0923, + "step": 1177 + }, + { + "epoch": 0.20975783475783477, + "grad_norm": 0.4806751012802124, + "learning_rate": 0.00019868920349733427, + "loss": 1.2296, + "step": 1178 + }, + { + "epoch": 0.20993589743589744, + "grad_norm": 0.44182994961738586, + "learning_rate": 0.0001986869435921579, + "loss": 1.1856, + "step": 1179 + }, + { + "epoch": 0.2101139601139601, + "grad_norm": 0.4282805621623993, + "learning_rate": 0.00019868468175341584, + "loss": 1.0046, + "step": 1180 + }, + { + "epoch": 0.2102920227920228, + "grad_norm": 0.5011838674545288, + "learning_rate": 0.00019868241798115242, + "loss": 1.2401, + "step": 1181 + }, + { + "epoch": 0.21047008547008547, + "grad_norm": 0.4282447397708893, + "learning_rate": 0.00019868015227541208, + "loss": 0.9338, + "step": 1182 + }, + { + "epoch": 0.21064814814814814, + "grad_norm": 0.4348810911178589, + "learning_rate": 0.00019867788463623912, + "loss": 0.926, + "step": 1183 + }, + { + "epoch": 0.21082621082621084, + "grad_norm": 0.41518425941467285, + "learning_rate": 0.00019867561506367799, + "loss": 1.2723, + "step": 1184 + }, + { + "epoch": 0.2110042735042735, + "grad_norm": 0.47346001863479614, + "learning_rate": 0.00019867334355777315, + "loss": 1.1931, + "step": 1185 + }, + { + "epoch": 0.21118233618233617, + "grad_norm": 0.4071715474128723, + "learning_rate": 0.00019867107011856914, + "loss": 0.9619, + "step": 1186 + }, + { + "epoch": 0.21136039886039887, + "grad_norm": 0.4803447425365448, + "learning_rate": 0.00019866879474611046, + "loss": 1.2, + "step": 1187 + }, + { + "epoch": 0.21153846153846154, + "grad_norm": 0.4827699661254883, + "learning_rate": 0.00019866651744044172, + "loss": 1.0938, + "step": 1188 + }, + { + "epoch": 0.2117165242165242, + "grad_norm": 0.4528424143791199, + "learning_rate": 0.00019866423820160756, + "loss": 0.9721, + "step": 1189 + }, + { + "epoch": 0.2118945868945869, + "grad_norm": 0.43566834926605225, + "learning_rate": 0.0001986619570296526, + "loss": 1.0352, + "step": 1190 + }, + { + "epoch": 0.21207264957264957, + "grad_norm": 0.4516540467739105, + "learning_rate": 0.0001986596739246215, + "loss": 1.1333, + "step": 1191 + }, + { + "epoch": 0.21225071225071226, + "grad_norm": 0.4456641376018524, + "learning_rate": 0.00019865738888655908, + "loss": 1.2813, + "step": 1192 + }, + { + "epoch": 0.21242877492877493, + "grad_norm": 0.47048309445381165, + "learning_rate": 0.00019865510191551008, + "loss": 1.1067, + "step": 1193 + }, + { + "epoch": 0.2126068376068376, + "grad_norm": 0.4604061543941498, + "learning_rate": 0.00019865281301151928, + "loss": 0.925, + "step": 1194 + }, + { + "epoch": 0.2127849002849003, + "grad_norm": 0.49341437220573425, + "learning_rate": 0.00019865052217463153, + "loss": 1.2319, + "step": 1195 + }, + { + "epoch": 0.21296296296296297, + "grad_norm": 0.5099014639854431, + "learning_rate": 0.00019864822940489173, + "loss": 1.139, + "step": 1196 + }, + { + "epoch": 0.21314102564102563, + "grad_norm": 0.41396936774253845, + "learning_rate": 0.0001986459347023448, + "loss": 1.0594, + "step": 1197 + }, + { + "epoch": 0.21331908831908833, + "grad_norm": 0.46071869134902954, + "learning_rate": 0.0001986436380670357, + "loss": 1.0815, + "step": 1198 + }, + { + "epoch": 0.213497150997151, + "grad_norm": 0.507882297039032, + "learning_rate": 0.00019864133949900942, + "loss": 1.3841, + "step": 1199 + }, + { + "epoch": 0.21367521367521367, + "grad_norm": 0.45680439472198486, + "learning_rate": 0.00019863903899831103, + "loss": 1.0945, + "step": 1200 + }, + { + "epoch": 0.21385327635327636, + "grad_norm": 0.44277429580688477, + "learning_rate": 0.00019863673656498555, + "loss": 1.1655, + "step": 1201 + }, + { + "epoch": 0.21403133903133903, + "grad_norm": 0.43890756368637085, + "learning_rate": 0.00019863443219907812, + "loss": 1.1186, + "step": 1202 + }, + { + "epoch": 0.2142094017094017, + "grad_norm": 0.3910178542137146, + "learning_rate": 0.0001986321259006339, + "loss": 1.0817, + "step": 1203 + }, + { + "epoch": 0.2143874643874644, + "grad_norm": 0.3803878128528595, + "learning_rate": 0.00019862981766969803, + "loss": 0.8022, + "step": 1204 + }, + { + "epoch": 0.21456552706552706, + "grad_norm": 0.4495108425617218, + "learning_rate": 0.0001986275075063158, + "loss": 1.2212, + "step": 1205 + }, + { + "epoch": 0.21474358974358973, + "grad_norm": 0.5211976766586304, + "learning_rate": 0.00019862519541053244, + "loss": 1.2771, + "step": 1206 + }, + { + "epoch": 0.21492165242165243, + "grad_norm": 0.4313061535358429, + "learning_rate": 0.00019862288138239325, + "loss": 1.1205, + "step": 1207 + }, + { + "epoch": 0.2150997150997151, + "grad_norm": 0.47110888361930847, + "learning_rate": 0.00019862056542194355, + "loss": 1.1835, + "step": 1208 + }, + { + "epoch": 0.2152777777777778, + "grad_norm": 0.5129403471946716, + "learning_rate": 0.00019861824752922876, + "loss": 1.1655, + "step": 1209 + }, + { + "epoch": 0.21545584045584046, + "grad_norm": 0.4353938102722168, + "learning_rate": 0.00019861592770429427, + "loss": 1.2794, + "step": 1210 + }, + { + "epoch": 0.21563390313390313, + "grad_norm": 0.48590636253356934, + "learning_rate": 0.0001986136059471855, + "loss": 1.2003, + "step": 1211 + }, + { + "epoch": 0.21581196581196582, + "grad_norm": 0.4738406836986542, + "learning_rate": 0.00019861128225794804, + "loss": 1.2271, + "step": 1212 + }, + { + "epoch": 0.2159900284900285, + "grad_norm": 0.45983126759529114, + "learning_rate": 0.0001986089566366273, + "loss": 1.1896, + "step": 1213 + }, + { + "epoch": 0.21616809116809116, + "grad_norm": 0.37296006083488464, + "learning_rate": 0.00019860662908326892, + "loss": 1.079, + "step": 1214 + }, + { + "epoch": 0.21634615384615385, + "grad_norm": 0.4442676305770874, + "learning_rate": 0.00019860429959791845, + "loss": 1.1754, + "step": 1215 + }, + { + "epoch": 0.21652421652421652, + "grad_norm": 0.4950128495693207, + "learning_rate": 0.0001986019681806216, + "loss": 1.1571, + "step": 1216 + }, + { + "epoch": 0.2167022792022792, + "grad_norm": 0.4374556541442871, + "learning_rate": 0.000198599634831424, + "loss": 1.1003, + "step": 1217 + }, + { + "epoch": 0.2168803418803419, + "grad_norm": 0.47301414608955383, + "learning_rate": 0.00019859729955037136, + "loss": 1.1426, + "step": 1218 + }, + { + "epoch": 0.21705840455840456, + "grad_norm": 0.41213178634643555, + "learning_rate": 0.00019859496233750947, + "loss": 1.0659, + "step": 1219 + }, + { + "epoch": 0.21723646723646722, + "grad_norm": 0.41601964831352234, + "learning_rate": 0.0001985926231928841, + "loss": 1.0248, + "step": 1220 + }, + { + "epoch": 0.21741452991452992, + "grad_norm": 0.46328839659690857, + "learning_rate": 0.0001985902821165411, + "loss": 1.0405, + "step": 1221 + }, + { + "epoch": 0.2175925925925926, + "grad_norm": 0.43287959694862366, + "learning_rate": 0.0001985879391085263, + "loss": 0.9202, + "step": 1222 + }, + { + "epoch": 0.21777065527065528, + "grad_norm": 0.4770444631576538, + "learning_rate": 0.00019858559416888568, + "loss": 1.0911, + "step": 1223 + }, + { + "epoch": 0.21794871794871795, + "grad_norm": 0.4756585955619812, + "learning_rate": 0.00019858324729766507, + "loss": 1.1566, + "step": 1224 + }, + { + "epoch": 0.21812678062678062, + "grad_norm": 0.4337233006954193, + "learning_rate": 0.00019858089849491054, + "loss": 0.9084, + "step": 1225 + }, + { + "epoch": 0.21830484330484332, + "grad_norm": 0.5165579319000244, + "learning_rate": 0.00019857854776066813, + "loss": 1.4154, + "step": 1226 + }, + { + "epoch": 0.21848290598290598, + "grad_norm": 0.4280378520488739, + "learning_rate": 0.00019857619509498382, + "loss": 1.1291, + "step": 1227 + }, + { + "epoch": 0.21866096866096865, + "grad_norm": 0.5375089049339294, + "learning_rate": 0.00019857384049790376, + "loss": 1.2985, + "step": 1228 + }, + { + "epoch": 0.21883903133903135, + "grad_norm": 0.4708811640739441, + "learning_rate": 0.00019857148396947401, + "loss": 1.0589, + "step": 1229 + }, + { + "epoch": 0.21901709401709402, + "grad_norm": 0.4744570255279541, + "learning_rate": 0.00019856912550974084, + "loss": 1.1269, + "step": 1230 + }, + { + "epoch": 0.21919515669515668, + "grad_norm": 0.5355265736579895, + "learning_rate": 0.00019856676511875043, + "loss": 1.1441, + "step": 1231 + }, + { + "epoch": 0.21937321937321938, + "grad_norm": 0.42718183994293213, + "learning_rate": 0.00019856440279654897, + "loss": 1.0244, + "step": 1232 + }, + { + "epoch": 0.21955128205128205, + "grad_norm": 0.5162127614021301, + "learning_rate": 0.00019856203854318283, + "loss": 1.2674, + "step": 1233 + }, + { + "epoch": 0.21972934472934472, + "grad_norm": 0.5180695652961731, + "learning_rate": 0.00019855967235869827, + "loss": 1.2472, + "step": 1234 + }, + { + "epoch": 0.2199074074074074, + "grad_norm": 0.4290023744106293, + "learning_rate": 0.00019855730424314167, + "loss": 1.0502, + "step": 1235 + }, + { + "epoch": 0.22008547008547008, + "grad_norm": 0.4418254792690277, + "learning_rate": 0.00019855493419655945, + "loss": 1.0589, + "step": 1236 + }, + { + "epoch": 0.22026353276353278, + "grad_norm": 0.4074663817882538, + "learning_rate": 0.000198552562218998, + "loss": 0.9197, + "step": 1237 + }, + { + "epoch": 0.22044159544159544, + "grad_norm": 0.4526660740375519, + "learning_rate": 0.00019855018831050383, + "loss": 1.2578, + "step": 1238 + }, + { + "epoch": 0.2206196581196581, + "grad_norm": 0.4747827649116516, + "learning_rate": 0.00019854781247112343, + "loss": 1.0841, + "step": 1239 + }, + { + "epoch": 0.2207977207977208, + "grad_norm": 0.41567128896713257, + "learning_rate": 0.00019854543470090334, + "loss": 1.0737, + "step": 1240 + }, + { + "epoch": 0.22097578347578348, + "grad_norm": 0.4793100953102112, + "learning_rate": 0.00019854305499989022, + "loss": 1.1972, + "step": 1241 + }, + { + "epoch": 0.22115384615384615, + "grad_norm": 0.41755473613739014, + "learning_rate": 0.00019854067336813058, + "loss": 1.2529, + "step": 1242 + }, + { + "epoch": 0.22133190883190884, + "grad_norm": 0.40421152114868164, + "learning_rate": 0.0001985382898056712, + "loss": 1.0549, + "step": 1243 + }, + { + "epoch": 0.2215099715099715, + "grad_norm": 0.45779645442962646, + "learning_rate": 0.0001985359043125587, + "loss": 1.1586, + "step": 1244 + }, + { + "epoch": 0.22168803418803418, + "grad_norm": 0.4380546808242798, + "learning_rate": 0.00019853351688883987, + "loss": 1.1024, + "step": 1245 + }, + { + "epoch": 0.22186609686609687, + "grad_norm": 0.39917269349098206, + "learning_rate": 0.00019853112753456142, + "loss": 0.9823, + "step": 1246 + }, + { + "epoch": 0.22204415954415954, + "grad_norm": 0.4228038489818573, + "learning_rate": 0.00019852873624977022, + "loss": 1.1684, + "step": 1247 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 0.4462146759033203, + "learning_rate": 0.00019852634303451315, + "loss": 0.9027, + "step": 1248 + }, + { + "epoch": 0.2224002849002849, + "grad_norm": 0.5682163834571838, + "learning_rate": 0.000198523947888837, + "loss": 1.141, + "step": 1249 + }, + { + "epoch": 0.22257834757834757, + "grad_norm": 0.44866830110549927, + "learning_rate": 0.0001985215508127888, + "loss": 1.0759, + "step": 1250 + }, + { + "epoch": 0.22275641025641027, + "grad_norm": 0.4034106135368347, + "learning_rate": 0.00019851915180641548, + "loss": 1.0675, + "step": 1251 + }, + { + "epoch": 0.22293447293447294, + "grad_norm": 0.4780726432800293, + "learning_rate": 0.00019851675086976397, + "loss": 1.0283, + "step": 1252 + }, + { + "epoch": 0.2231125356125356, + "grad_norm": 0.48892372846603394, + "learning_rate": 0.00019851434800288145, + "loss": 1.1159, + "step": 1253 + }, + { + "epoch": 0.2232905982905983, + "grad_norm": 0.42629215121269226, + "learning_rate": 0.0001985119432058149, + "loss": 1.0292, + "step": 1254 + }, + { + "epoch": 0.22346866096866097, + "grad_norm": 0.4496444761753082, + "learning_rate": 0.00019850953647861146, + "loss": 1.0252, + "step": 1255 + }, + { + "epoch": 0.22364672364672364, + "grad_norm": 0.4371408224105835, + "learning_rate": 0.00019850712782131828, + "loss": 1.1104, + "step": 1256 + }, + { + "epoch": 0.22382478632478633, + "grad_norm": 0.4910794496536255, + "learning_rate": 0.00019850471723398258, + "loss": 1.1928, + "step": 1257 + }, + { + "epoch": 0.224002849002849, + "grad_norm": 0.41235068440437317, + "learning_rate": 0.00019850230471665157, + "loss": 1.1261, + "step": 1258 + }, + { + "epoch": 0.22418091168091167, + "grad_norm": 0.4507700502872467, + "learning_rate": 0.0001984998902693725, + "loss": 1.0602, + "step": 1259 + }, + { + "epoch": 0.22435897435897437, + "grad_norm": 0.4654198884963989, + "learning_rate": 0.00019849747389219272, + "loss": 1.1258, + "step": 1260 + }, + { + "epoch": 0.22453703703703703, + "grad_norm": 0.439807653427124, + "learning_rate": 0.00019849505558515952, + "loss": 1.2312, + "step": 1261 + }, + { + "epoch": 0.2247150997150997, + "grad_norm": 0.4309258759021759, + "learning_rate": 0.00019849263534832035, + "loss": 1.0083, + "step": 1262 + }, + { + "epoch": 0.2248931623931624, + "grad_norm": 0.4920141100883484, + "learning_rate": 0.00019849021318172255, + "loss": 1.0254, + "step": 1263 + }, + { + "epoch": 0.22507122507122507, + "grad_norm": 0.5333457589149475, + "learning_rate": 0.00019848778908541367, + "loss": 1.3017, + "step": 1264 + }, + { + "epoch": 0.22524928774928774, + "grad_norm": 0.4096757769584656, + "learning_rate": 0.0001984853630594411, + "loss": 0.9531, + "step": 1265 + }, + { + "epoch": 0.22542735042735043, + "grad_norm": 0.5744075775146484, + "learning_rate": 0.00019848293510385244, + "loss": 1.1414, + "step": 1266 + }, + { + "epoch": 0.2256054131054131, + "grad_norm": 0.44707193970680237, + "learning_rate": 0.00019848050521869529, + "loss": 1.1926, + "step": 1267 + }, + { + "epoch": 0.2257834757834758, + "grad_norm": 0.4162999391555786, + "learning_rate": 0.00019847807340401716, + "loss": 1.1354, + "step": 1268 + }, + { + "epoch": 0.22596153846153846, + "grad_norm": 0.4273204207420349, + "learning_rate": 0.0001984756396598658, + "loss": 0.9956, + "step": 1269 + }, + { + "epoch": 0.22613960113960113, + "grad_norm": 0.5670466423034668, + "learning_rate": 0.00019847320398628878, + "loss": 1.2384, + "step": 1270 + }, + { + "epoch": 0.22631766381766383, + "grad_norm": 0.424544095993042, + "learning_rate": 0.00019847076638333395, + "loss": 0.9963, + "step": 1271 + }, + { + "epoch": 0.2264957264957265, + "grad_norm": 0.3716120719909668, + "learning_rate": 0.000198468326851049, + "loss": 0.865, + "step": 1272 + }, + { + "epoch": 0.22667378917378916, + "grad_norm": 0.4472847282886505, + "learning_rate": 0.00019846588538948172, + "loss": 1.174, + "step": 1273 + }, + { + "epoch": 0.22685185185185186, + "grad_norm": 0.4599195718765259, + "learning_rate": 0.00019846344199867994, + "loss": 1.289, + "step": 1274 + }, + { + "epoch": 0.22702991452991453, + "grad_norm": 0.4303213357925415, + "learning_rate": 0.0001984609966786916, + "loss": 1.1606, + "step": 1275 + }, + { + "epoch": 0.2272079772079772, + "grad_norm": 0.44893527030944824, + "learning_rate": 0.00019845854942956455, + "loss": 1.1043, + "step": 1276 + }, + { + "epoch": 0.2273860398860399, + "grad_norm": 0.40033379197120667, + "learning_rate": 0.00019845610025134676, + "loss": 1.1434, + "step": 1277 + }, + { + "epoch": 0.22756410256410256, + "grad_norm": 0.4385402202606201, + "learning_rate": 0.00019845364914408616, + "loss": 0.9943, + "step": 1278 + }, + { + "epoch": 0.22774216524216523, + "grad_norm": 0.42123618721961975, + "learning_rate": 0.0001984511961078309, + "loss": 1.0911, + "step": 1279 + }, + { + "epoch": 0.22792022792022792, + "grad_norm": 0.5558577179908752, + "learning_rate": 0.00019844874114262893, + "loss": 1.3893, + "step": 1280 + }, + { + "epoch": 0.2280982905982906, + "grad_norm": 0.3996453583240509, + "learning_rate": 0.00019844628424852835, + "loss": 0.8951, + "step": 1281 + }, + { + "epoch": 0.2282763532763533, + "grad_norm": 0.3943425714969635, + "learning_rate": 0.0001984438254255774, + "loss": 1.0595, + "step": 1282 + }, + { + "epoch": 0.22845441595441596, + "grad_norm": 0.4429021179676056, + "learning_rate": 0.00019844136467382414, + "loss": 1.0853, + "step": 1283 + }, + { + "epoch": 0.22863247863247863, + "grad_norm": 0.4515686631202698, + "learning_rate": 0.00019843890199331687, + "loss": 1.0829, + "step": 1284 + }, + { + "epoch": 0.22881054131054132, + "grad_norm": 0.5157768726348877, + "learning_rate": 0.00019843643738410378, + "loss": 1.334, + "step": 1285 + }, + { + "epoch": 0.228988603988604, + "grad_norm": 0.45833173394203186, + "learning_rate": 0.0001984339708462332, + "loss": 1.1353, + "step": 1286 + }, + { + "epoch": 0.22916666666666666, + "grad_norm": 0.46610337495803833, + "learning_rate": 0.00019843150237975344, + "loss": 1.1338, + "step": 1287 + }, + { + "epoch": 0.22934472934472935, + "grad_norm": 0.5076978802680969, + "learning_rate": 0.00019842903198471286, + "loss": 1.1811, + "step": 1288 + }, + { + "epoch": 0.22952279202279202, + "grad_norm": 0.4297824800014496, + "learning_rate": 0.00019842655966115986, + "loss": 1.1799, + "step": 1289 + }, + { + "epoch": 0.2297008547008547, + "grad_norm": 0.5304586291313171, + "learning_rate": 0.0001984240854091429, + "loss": 1.1315, + "step": 1290 + }, + { + "epoch": 0.22987891737891739, + "grad_norm": 0.45359212160110474, + "learning_rate": 0.00019842160922871042, + "loss": 1.1037, + "step": 1291 + }, + { + "epoch": 0.23005698005698005, + "grad_norm": 0.4416881203651428, + "learning_rate": 0.00019841913111991096, + "loss": 1.122, + "step": 1292 + }, + { + "epoch": 0.23023504273504272, + "grad_norm": 0.46682995557785034, + "learning_rate": 0.0001984166510827931, + "loss": 0.9808, + "step": 1293 + }, + { + "epoch": 0.23041310541310542, + "grad_norm": 0.44172337651252747, + "learning_rate": 0.00019841416911740538, + "loss": 0.9167, + "step": 1294 + }, + { + "epoch": 0.23059116809116809, + "grad_norm": 0.40562742948532104, + "learning_rate": 0.0001984116852237965, + "loss": 0.9547, + "step": 1295 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.4040384888648987, + "learning_rate": 0.00019840919940201503, + "loss": 1.1039, + "step": 1296 + }, + { + "epoch": 0.23094729344729345, + "grad_norm": 0.5094077587127686, + "learning_rate": 0.00019840671165210973, + "loss": 1.2283, + "step": 1297 + }, + { + "epoch": 0.23112535612535612, + "grad_norm": 0.48553213477134705, + "learning_rate": 0.00019840422197412938, + "loss": 1.0927, + "step": 1298 + }, + { + "epoch": 0.23130341880341881, + "grad_norm": 0.5197509527206421, + "learning_rate": 0.00019840173036812266, + "loss": 1.2154, + "step": 1299 + }, + { + "epoch": 0.23148148148148148, + "grad_norm": 0.42069005966186523, + "learning_rate": 0.0001983992368341385, + "loss": 1.0076, + "step": 1300 + }, + { + "epoch": 0.23165954415954415, + "grad_norm": 0.475204735994339, + "learning_rate": 0.00019839674137222567, + "loss": 1.1682, + "step": 1301 + }, + { + "epoch": 0.23183760683760685, + "grad_norm": 0.55730140209198, + "learning_rate": 0.0001983942439824331, + "loss": 1.2948, + "step": 1302 + }, + { + "epoch": 0.23201566951566951, + "grad_norm": 0.4533313512802124, + "learning_rate": 0.00019839174466480973, + "loss": 1.2691, + "step": 1303 + }, + { + "epoch": 0.23219373219373218, + "grad_norm": 0.4733520746231079, + "learning_rate": 0.0001983892434194045, + "loss": 1.2232, + "step": 1304 + }, + { + "epoch": 0.23237179487179488, + "grad_norm": 0.5085756182670593, + "learning_rate": 0.00019838674024626643, + "loss": 1.1347, + "step": 1305 + }, + { + "epoch": 0.23254985754985755, + "grad_norm": 0.4679976999759674, + "learning_rate": 0.00019838423514544456, + "loss": 1.0018, + "step": 1306 + }, + { + "epoch": 0.23272792022792022, + "grad_norm": 0.4234481751918793, + "learning_rate": 0.00019838172811698795, + "loss": 1.0472, + "step": 1307 + }, + { + "epoch": 0.2329059829059829, + "grad_norm": 0.5749204158782959, + "learning_rate": 0.00019837921916094579, + "loss": 1.2239, + "step": 1308 + }, + { + "epoch": 0.23308404558404558, + "grad_norm": 0.46715882420539856, + "learning_rate": 0.0001983767082773672, + "loss": 1.1924, + "step": 1309 + }, + { + "epoch": 0.23326210826210828, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.00019837419546630137, + "loss": 1.1086, + "step": 1310 + }, + { + "epoch": 0.23344017094017094, + "grad_norm": 0.4419243037700653, + "learning_rate": 0.0001983716807277975, + "loss": 1.1911, + "step": 1311 + }, + { + "epoch": 0.2336182336182336, + "grad_norm": 0.5107570290565491, + "learning_rate": 0.00019836916406190493, + "loss": 1.1071, + "step": 1312 + }, + { + "epoch": 0.2337962962962963, + "grad_norm": 0.5295659303665161, + "learning_rate": 0.00019836664546867293, + "loss": 1.2905, + "step": 1313 + }, + { + "epoch": 0.23397435897435898, + "grad_norm": 0.4844837784767151, + "learning_rate": 0.00019836412494815084, + "loss": 1.3507, + "step": 1314 + }, + { + "epoch": 0.23415242165242164, + "grad_norm": 0.6166049242019653, + "learning_rate": 0.00019836160250038808, + "loss": 1.2822, + "step": 1315 + }, + { + "epoch": 0.23433048433048434, + "grad_norm": 0.3229198753833771, + "learning_rate": 0.00019835907812543402, + "loss": 0.4959, + "step": 1316 + }, + { + "epoch": 0.234508547008547, + "grad_norm": 0.5788772702217102, + "learning_rate": 0.00019835655182333815, + "loss": 1.0832, + "step": 1317 + }, + { + "epoch": 0.23468660968660968, + "grad_norm": 0.525705099105835, + "learning_rate": 0.00019835402359414997, + "loss": 1.0968, + "step": 1318 + }, + { + "epoch": 0.23486467236467237, + "grad_norm": 0.5007779002189636, + "learning_rate": 0.000198351493437919, + "loss": 1.2788, + "step": 1319 + }, + { + "epoch": 0.23504273504273504, + "grad_norm": 0.4276871383190155, + "learning_rate": 0.00019834896135469484, + "loss": 1.0419, + "step": 1320 + }, + { + "epoch": 0.2352207977207977, + "grad_norm": 0.5359070301055908, + "learning_rate": 0.00019834642734452708, + "loss": 1.1308, + "step": 1321 + }, + { + "epoch": 0.2353988603988604, + "grad_norm": 0.4854908883571625, + "learning_rate": 0.0001983438914074654, + "loss": 1.1211, + "step": 1322 + }, + { + "epoch": 0.23557692307692307, + "grad_norm": 0.4913707375526428, + "learning_rate": 0.0001983413535435594, + "loss": 1.2392, + "step": 1323 + }, + { + "epoch": 0.23575498575498577, + "grad_norm": 0.46755748987197876, + "learning_rate": 0.0001983388137528589, + "loss": 0.9348, + "step": 1324 + }, + { + "epoch": 0.23593304843304844, + "grad_norm": 0.4592570960521698, + "learning_rate": 0.0001983362720354136, + "loss": 1.1339, + "step": 1325 + }, + { + "epoch": 0.2361111111111111, + "grad_norm": 0.5121711492538452, + "learning_rate": 0.00019833372839127335, + "loss": 1.2973, + "step": 1326 + }, + { + "epoch": 0.2362891737891738, + "grad_norm": 0.4809017479419708, + "learning_rate": 0.000198331182820488, + "loss": 0.9849, + "step": 1327 + }, + { + "epoch": 0.23646723646723647, + "grad_norm": 0.42340895533561707, + "learning_rate": 0.00019832863532310733, + "loss": 1.0731, + "step": 1328 + }, + { + "epoch": 0.23664529914529914, + "grad_norm": 0.5388045310974121, + "learning_rate": 0.00019832608589918135, + "loss": 1.0729, + "step": 1329 + }, + { + "epoch": 0.23682336182336183, + "grad_norm": 0.43075770139694214, + "learning_rate": 0.00019832353454875992, + "loss": 1.1684, + "step": 1330 + }, + { + "epoch": 0.2370014245014245, + "grad_norm": 0.554927408695221, + "learning_rate": 0.00019832098127189313, + "loss": 1.0842, + "step": 1331 + }, + { + "epoch": 0.23717948717948717, + "grad_norm": 0.5359260439872742, + "learning_rate": 0.0001983184260686309, + "loss": 1.2399, + "step": 1332 + }, + { + "epoch": 0.23735754985754987, + "grad_norm": 0.5141251087188721, + "learning_rate": 0.0001983158689390234, + "loss": 1.3752, + "step": 1333 + }, + { + "epoch": 0.23753561253561253, + "grad_norm": 0.4578750431537628, + "learning_rate": 0.00019831330988312067, + "loss": 1.0965, + "step": 1334 + }, + { + "epoch": 0.2377136752136752, + "grad_norm": 0.47974497079849243, + "learning_rate": 0.00019831074890097286, + "loss": 1.3379, + "step": 1335 + }, + { + "epoch": 0.2378917378917379, + "grad_norm": 0.4618176817893982, + "learning_rate": 0.00019830818599263014, + "loss": 1.274, + "step": 1336 + }, + { + "epoch": 0.23806980056980057, + "grad_norm": 0.4279816448688507, + "learning_rate": 0.00019830562115814276, + "loss": 0.996, + "step": 1337 + }, + { + "epoch": 0.23824786324786323, + "grad_norm": 0.4255026876926422, + "learning_rate": 0.0001983030543975609, + "loss": 0.969, + "step": 1338 + }, + { + "epoch": 0.23842592592592593, + "grad_norm": 0.4551412761211395, + "learning_rate": 0.00019830048571093493, + "loss": 1.0204, + "step": 1339 + }, + { + "epoch": 0.2386039886039886, + "grad_norm": 0.4747903048992157, + "learning_rate": 0.00019829791509831513, + "loss": 1.1816, + "step": 1340 + }, + { + "epoch": 0.2387820512820513, + "grad_norm": 0.47187140583992004, + "learning_rate": 0.00019829534255975188, + "loss": 1.1205, + "step": 1341 + }, + { + "epoch": 0.23896011396011396, + "grad_norm": 0.49332180619239807, + "learning_rate": 0.0001982927680952956, + "loss": 1.2657, + "step": 1342 + }, + { + "epoch": 0.23913817663817663, + "grad_norm": 0.5162837505340576, + "learning_rate": 0.0001982901917049967, + "loss": 1.2247, + "step": 1343 + }, + { + "epoch": 0.23931623931623933, + "grad_norm": 0.43407055735588074, + "learning_rate": 0.0001982876133889057, + "loss": 1.0038, + "step": 1344 + }, + { + "epoch": 0.239494301994302, + "grad_norm": 0.5132251977920532, + "learning_rate": 0.00019828503314707306, + "loss": 1.0678, + "step": 1345 + }, + { + "epoch": 0.23967236467236466, + "grad_norm": 0.46295464038848877, + "learning_rate": 0.00019828245097954937, + "loss": 1.1802, + "step": 1346 + }, + { + "epoch": 0.23985042735042736, + "grad_norm": 0.4682658314704895, + "learning_rate": 0.00019827986688638523, + "loss": 1.0249, + "step": 1347 + }, + { + "epoch": 0.24002849002849003, + "grad_norm": 0.49990561604499817, + "learning_rate": 0.00019827728086763125, + "loss": 1.0691, + "step": 1348 + }, + { + "epoch": 0.2402065527065527, + "grad_norm": 0.39090847969055176, + "learning_rate": 0.00019827469292333806, + "loss": 0.8367, + "step": 1349 + }, + { + "epoch": 0.2403846153846154, + "grad_norm": 0.5023905634880066, + "learning_rate": 0.00019827210305355645, + "loss": 1.0675, + "step": 1350 + }, + { + "epoch": 0.24056267806267806, + "grad_norm": 0.4744076430797577, + "learning_rate": 0.00019826951125833715, + "loss": 1.3166, + "step": 1351 + }, + { + "epoch": 0.24074074074074073, + "grad_norm": 0.44914689660072327, + "learning_rate": 0.00019826691753773088, + "loss": 0.9818, + "step": 1352 + }, + { + "epoch": 0.24091880341880342, + "grad_norm": 0.44391971826553345, + "learning_rate": 0.00019826432189178853, + "loss": 1.0448, + "step": 1353 + }, + { + "epoch": 0.2410968660968661, + "grad_norm": 0.46102839708328247, + "learning_rate": 0.00019826172432056086, + "loss": 0.9952, + "step": 1354 + }, + { + "epoch": 0.2412749287749288, + "grad_norm": 0.4796878695487976, + "learning_rate": 0.00019825912482409884, + "loss": 1.0977, + "step": 1355 + }, + { + "epoch": 0.24145299145299146, + "grad_norm": 0.5003768801689148, + "learning_rate": 0.0001982565234024534, + "loss": 1.3149, + "step": 1356 + }, + { + "epoch": 0.24163105413105412, + "grad_norm": 0.43475663661956787, + "learning_rate": 0.00019825392005567551, + "loss": 1.0527, + "step": 1357 + }, + { + "epoch": 0.24180911680911682, + "grad_norm": 0.46120527386665344, + "learning_rate": 0.00019825131478381613, + "loss": 1.2333, + "step": 1358 + }, + { + "epoch": 0.2419871794871795, + "grad_norm": 0.43748101592063904, + "learning_rate": 0.00019824870758692638, + "loss": 0.9788, + "step": 1359 + }, + { + "epoch": 0.24216524216524216, + "grad_norm": 0.5275192856788635, + "learning_rate": 0.00019824609846505727, + "loss": 1.1473, + "step": 1360 + }, + { + "epoch": 0.24234330484330485, + "grad_norm": 0.346463143825531, + "learning_rate": 0.00019824348741825993, + "loss": 0.6824, + "step": 1361 + }, + { + "epoch": 0.24252136752136752, + "grad_norm": 0.5004115700721741, + "learning_rate": 0.00019824087444658556, + "loss": 1.1853, + "step": 1362 + }, + { + "epoch": 0.2426994301994302, + "grad_norm": 0.42746666073799133, + "learning_rate": 0.00019823825955008533, + "loss": 0.9355, + "step": 1363 + }, + { + "epoch": 0.24287749287749288, + "grad_norm": 0.4099743068218231, + "learning_rate": 0.00019823564272881047, + "loss": 1.0753, + "step": 1364 + }, + { + "epoch": 0.24305555555555555, + "grad_norm": 0.5262967944145203, + "learning_rate": 0.00019823302398281226, + "loss": 1.2324, + "step": 1365 + }, + { + "epoch": 0.24323361823361822, + "grad_norm": 0.436069518327713, + "learning_rate": 0.000198230403312142, + "loss": 1.1887, + "step": 1366 + }, + { + "epoch": 0.24341168091168092, + "grad_norm": 0.38252368569374084, + "learning_rate": 0.00019822778071685107, + "loss": 1.0211, + "step": 1367 + }, + { + "epoch": 0.24358974358974358, + "grad_norm": 0.48024141788482666, + "learning_rate": 0.00019822515619699081, + "loss": 1.065, + "step": 1368 + }, + { + "epoch": 0.24376780626780628, + "grad_norm": 0.47421589493751526, + "learning_rate": 0.00019822252975261267, + "loss": 1.0433, + "step": 1369 + }, + { + "epoch": 0.24394586894586895, + "grad_norm": 0.46094807982444763, + "learning_rate": 0.00019821990138376808, + "loss": 1.1427, + "step": 1370 + }, + { + "epoch": 0.24412393162393162, + "grad_norm": 0.5093680620193481, + "learning_rate": 0.00019821727109050856, + "loss": 1.1086, + "step": 1371 + }, + { + "epoch": 0.2443019943019943, + "grad_norm": 0.41084879636764526, + "learning_rate": 0.00019821463887288566, + "loss": 1.0068, + "step": 1372 + }, + { + "epoch": 0.24448005698005698, + "grad_norm": 0.4991084635257721, + "learning_rate": 0.0001982120047309509, + "loss": 1.1884, + "step": 1373 + }, + { + "epoch": 0.24465811965811965, + "grad_norm": 0.39198383688926697, + "learning_rate": 0.00019820936866475595, + "loss": 0.9776, + "step": 1374 + }, + { + "epoch": 0.24483618233618235, + "grad_norm": 0.4517424702644348, + "learning_rate": 0.00019820673067435244, + "loss": 1.1491, + "step": 1375 + }, + { + "epoch": 0.245014245014245, + "grad_norm": 0.45881983637809753, + "learning_rate": 0.00019820409075979202, + "loss": 1.1198, + "step": 1376 + }, + { + "epoch": 0.24519230769230768, + "grad_norm": 0.4498792290687561, + "learning_rate": 0.00019820144892112646, + "loss": 1.0897, + "step": 1377 + }, + { + "epoch": 0.24537037037037038, + "grad_norm": 0.4128037393093109, + "learning_rate": 0.00019819880515840752, + "loss": 0.9415, + "step": 1378 + }, + { + "epoch": 0.24554843304843305, + "grad_norm": 0.4340885281562805, + "learning_rate": 0.00019819615947168698, + "loss": 1.201, + "step": 1379 + }, + { + "epoch": 0.24572649572649571, + "grad_norm": 0.43814027309417725, + "learning_rate": 0.00019819351186101667, + "loss": 1.1039, + "step": 1380 + }, + { + "epoch": 0.2459045584045584, + "grad_norm": 0.40115082263946533, + "learning_rate": 0.00019819086232644845, + "loss": 1.2599, + "step": 1381 + }, + { + "epoch": 0.24608262108262108, + "grad_norm": 0.4947351813316345, + "learning_rate": 0.00019818821086803426, + "loss": 1.252, + "step": 1382 + }, + { + "epoch": 0.24626068376068377, + "grad_norm": 0.45179441571235657, + "learning_rate": 0.0001981855574858261, + "loss": 1.1323, + "step": 1383 + }, + { + "epoch": 0.24643874643874644, + "grad_norm": 0.47159844636917114, + "learning_rate": 0.00019818290217987587, + "loss": 1.2053, + "step": 1384 + }, + { + "epoch": 0.2466168091168091, + "grad_norm": 0.4358448386192322, + "learning_rate": 0.0001981802449502356, + "loss": 1.1174, + "step": 1385 + }, + { + "epoch": 0.2467948717948718, + "grad_norm": 0.4588233530521393, + "learning_rate": 0.00019817758579695745, + "loss": 1.1098, + "step": 1386 + }, + { + "epoch": 0.24697293447293447, + "grad_norm": 0.4955112636089325, + "learning_rate": 0.00019817492472009338, + "loss": 1.258, + "step": 1387 + }, + { + "epoch": 0.24715099715099714, + "grad_norm": 0.4226941764354706, + "learning_rate": 0.00019817226171969565, + "loss": 1.0976, + "step": 1388 + }, + { + "epoch": 0.24732905982905984, + "grad_norm": 0.4076840579509735, + "learning_rate": 0.00019816959679581637, + "loss": 1.0121, + "step": 1389 + }, + { + "epoch": 0.2475071225071225, + "grad_norm": 0.4395063519477844, + "learning_rate": 0.0001981669299485078, + "loss": 1.3153, + "step": 1390 + }, + { + "epoch": 0.24768518518518517, + "grad_norm": 0.41010400652885437, + "learning_rate": 0.0001981642611778221, + "loss": 1.0717, + "step": 1391 + }, + { + "epoch": 0.24786324786324787, + "grad_norm": 0.43459352850914, + "learning_rate": 0.00019816159048381167, + "loss": 1.1077, + "step": 1392 + }, + { + "epoch": 0.24804131054131054, + "grad_norm": 0.46291449666023254, + "learning_rate": 0.00019815891786652875, + "loss": 1.0257, + "step": 1393 + }, + { + "epoch": 0.2482193732193732, + "grad_norm": 0.46408146619796753, + "learning_rate": 0.00019815624332602578, + "loss": 0.7899, + "step": 1394 + }, + { + "epoch": 0.2483974358974359, + "grad_norm": 0.4763357937335968, + "learning_rate": 0.00019815356686235508, + "loss": 0.9857, + "step": 1395 + }, + { + "epoch": 0.24857549857549857, + "grad_norm": 0.4766457676887512, + "learning_rate": 0.00019815088847556918, + "loss": 1.0589, + "step": 1396 + }, + { + "epoch": 0.24875356125356127, + "grad_norm": 0.4486583173274994, + "learning_rate": 0.0001981482081657205, + "loss": 1.2572, + "step": 1397 + }, + { + "epoch": 0.24893162393162394, + "grad_norm": 0.468878835439682, + "learning_rate": 0.00019814552593286155, + "loss": 1.101, + "step": 1398 + }, + { + "epoch": 0.2491096866096866, + "grad_norm": 0.4230278730392456, + "learning_rate": 0.0001981428417770449, + "loss": 0.9457, + "step": 1399 + }, + { + "epoch": 0.2492877492877493, + "grad_norm": 0.45630761981010437, + "learning_rate": 0.00019814015569832315, + "loss": 1.0665, + "step": 1400 + }, + { + "epoch": 0.24946581196581197, + "grad_norm": 0.5780113935470581, + "learning_rate": 0.00019813746769674893, + "loss": 1.1064, + "step": 1401 + }, + { + "epoch": 0.24964387464387464, + "grad_norm": 0.4343436658382416, + "learning_rate": 0.0001981347777723749, + "loss": 1.1132, + "step": 1402 + }, + { + "epoch": 0.24982193732193733, + "grad_norm": 0.4879056513309479, + "learning_rate": 0.0001981320859252537, + "loss": 1.1301, + "step": 1403 + }, + { + "epoch": 0.25, + "grad_norm": 0.5248328447341919, + "learning_rate": 0.00019812939215543818, + "loss": 1.1468, + "step": 1404 + }, + { + "epoch": 0.25, + "eval_loss": 1.115895390510559, + "eval_runtime": 25.0474, + "eval_samples_per_second": 41.561, + "eval_steps_per_second": 20.801, + "step": 1404 + }, + { + "epoch": 0.2501780626780627, + "grad_norm": 0.5076769590377808, + "learning_rate": 0.00019812669646298106, + "loss": 1.1428, + "step": 1405 + }, + { + "epoch": 0.25035612535612534, + "grad_norm": 0.5510252714157104, + "learning_rate": 0.00019812399884793514, + "loss": 1.3383, + "step": 1406 + }, + { + "epoch": 0.25053418803418803, + "grad_norm": 0.48918986320495605, + "learning_rate": 0.0001981212993103533, + "loss": 1.1507, + "step": 1407 + }, + { + "epoch": 0.25071225071225073, + "grad_norm": 0.4678935110569, + "learning_rate": 0.00019811859785028846, + "loss": 1.13, + "step": 1408 + }, + { + "epoch": 0.25089031339031337, + "grad_norm": 0.5155254602432251, + "learning_rate": 0.0001981158944677935, + "loss": 1.1194, + "step": 1409 + }, + { + "epoch": 0.25106837606837606, + "grad_norm": 0.4533839523792267, + "learning_rate": 0.00019811318916292142, + "loss": 0.9464, + "step": 1410 + }, + { + "epoch": 0.25124643874643876, + "grad_norm": 0.5142433047294617, + "learning_rate": 0.00019811048193572517, + "loss": 1.0837, + "step": 1411 + }, + { + "epoch": 0.2514245014245014, + "grad_norm": 0.4330446124076843, + "learning_rate": 0.00019810777278625788, + "loss": 0.9117, + "step": 1412 + }, + { + "epoch": 0.2516025641025641, + "grad_norm": 0.44806256890296936, + "learning_rate": 0.00019810506171457254, + "loss": 1.1643, + "step": 1413 + }, + { + "epoch": 0.2517806267806268, + "grad_norm": 0.43526285886764526, + "learning_rate": 0.00019810234872072235, + "loss": 0.9776, + "step": 1414 + }, + { + "epoch": 0.25195868945868943, + "grad_norm": 0.47394511103630066, + "learning_rate": 0.00019809963380476039, + "loss": 1.0935, + "step": 1415 + }, + { + "epoch": 0.25213675213675213, + "grad_norm": 0.48961278796195984, + "learning_rate": 0.00019809691696673993, + "loss": 1.179, + "step": 1416 + }, + { + "epoch": 0.2523148148148148, + "grad_norm": 0.43153589963912964, + "learning_rate": 0.00019809419820671412, + "loss": 0.906, + "step": 1417 + }, + { + "epoch": 0.25249287749287747, + "grad_norm": 0.41187527775764465, + "learning_rate": 0.00019809147752473632, + "loss": 0.899, + "step": 1418 + }, + { + "epoch": 0.25267094017094016, + "grad_norm": 0.5003183484077454, + "learning_rate": 0.00019808875492085973, + "loss": 1.0606, + "step": 1419 + }, + { + "epoch": 0.25284900284900286, + "grad_norm": 0.4430316984653473, + "learning_rate": 0.00019808603039513778, + "loss": 0.9167, + "step": 1420 + }, + { + "epoch": 0.25302706552706555, + "grad_norm": 0.4577699601650238, + "learning_rate": 0.00019808330394762382, + "loss": 1.1184, + "step": 1421 + }, + { + "epoch": 0.2532051282051282, + "grad_norm": 0.42656826972961426, + "learning_rate": 0.0001980805755783713, + "loss": 0.9335, + "step": 1422 + }, + { + "epoch": 0.2533831908831909, + "grad_norm": 0.40980881452560425, + "learning_rate": 0.0001980778452874336, + "loss": 0.9756, + "step": 1423 + }, + { + "epoch": 0.2535612535612536, + "grad_norm": 0.5752090811729431, + "learning_rate": 0.00019807511307486423, + "loss": 1.1694, + "step": 1424 + }, + { + "epoch": 0.2537393162393162, + "grad_norm": 0.5000349283218384, + "learning_rate": 0.00019807237894071681, + "loss": 0.9515, + "step": 1425 + }, + { + "epoch": 0.2539173789173789, + "grad_norm": 0.5159069299697876, + "learning_rate": 0.00019806964288504483, + "loss": 1.4014, + "step": 1426 + }, + { + "epoch": 0.2540954415954416, + "grad_norm": 0.5377941131591797, + "learning_rate": 0.00019806690490790194, + "loss": 1.2832, + "step": 1427 + }, + { + "epoch": 0.25427350427350426, + "grad_norm": 0.4565938711166382, + "learning_rate": 0.00019806416500934174, + "loss": 1.0629, + "step": 1428 + }, + { + "epoch": 0.25445156695156695, + "grad_norm": 0.49867144227027893, + "learning_rate": 0.00019806142318941797, + "loss": 1.2011, + "step": 1429 + }, + { + "epoch": 0.25462962962962965, + "grad_norm": 0.5111994743347168, + "learning_rate": 0.00019805867944818427, + "loss": 0.8925, + "step": 1430 + }, + { + "epoch": 0.2548076923076923, + "grad_norm": 0.5204268097877502, + "learning_rate": 0.00019805593378569448, + "loss": 1.2956, + "step": 1431 + }, + { + "epoch": 0.254985754985755, + "grad_norm": 0.3889026939868927, + "learning_rate": 0.00019805318620200234, + "loss": 1.0355, + "step": 1432 + }, + { + "epoch": 0.2551638176638177, + "grad_norm": 0.46825656294822693, + "learning_rate": 0.00019805043669716174, + "loss": 1.0444, + "step": 1433 + }, + { + "epoch": 0.2553418803418803, + "grad_norm": 0.4509420394897461, + "learning_rate": 0.00019804768527122648, + "loss": 1.0423, + "step": 1434 + }, + { + "epoch": 0.255519943019943, + "grad_norm": 0.4514774978160858, + "learning_rate": 0.0001980449319242505, + "loss": 1.1588, + "step": 1435 + }, + { + "epoch": 0.2556980056980057, + "grad_norm": 0.43019044399261475, + "learning_rate": 0.0001980421766562878, + "loss": 0.9939, + "step": 1436 + }, + { + "epoch": 0.25587606837606836, + "grad_norm": 0.5056091547012329, + "learning_rate": 0.00019803941946739228, + "loss": 1.1238, + "step": 1437 + }, + { + "epoch": 0.25605413105413105, + "grad_norm": 0.48664605617523193, + "learning_rate": 0.000198036660357618, + "loss": 1.0702, + "step": 1438 + }, + { + "epoch": 0.25623219373219375, + "grad_norm": 0.4500972032546997, + "learning_rate": 0.000198033899327019, + "loss": 0.9365, + "step": 1439 + }, + { + "epoch": 0.2564102564102564, + "grad_norm": 0.4800589382648468, + "learning_rate": 0.0001980311363756494, + "loss": 1.1159, + "step": 1440 + }, + { + "epoch": 0.2565883190883191, + "grad_norm": 0.3486495316028595, + "learning_rate": 0.0001980283715035633, + "loss": 0.6029, + "step": 1441 + }, + { + "epoch": 0.2567663817663818, + "grad_norm": 0.46258702874183655, + "learning_rate": 0.00019802560471081493, + "loss": 1.025, + "step": 1442 + }, + { + "epoch": 0.2569444444444444, + "grad_norm": 0.4846673607826233, + "learning_rate": 0.00019802283599745844, + "loss": 1.1105, + "step": 1443 + }, + { + "epoch": 0.2571225071225071, + "grad_norm": 0.4586990475654602, + "learning_rate": 0.00019802006536354813, + "loss": 0.9897, + "step": 1444 + }, + { + "epoch": 0.2573005698005698, + "grad_norm": 0.5177786350250244, + "learning_rate": 0.00019801729280913825, + "loss": 1.2558, + "step": 1445 + }, + { + "epoch": 0.25747863247863245, + "grad_norm": 0.43213751912117004, + "learning_rate": 0.00019801451833428312, + "loss": 1.0961, + "step": 1446 + }, + { + "epoch": 0.25765669515669515, + "grad_norm": 0.42974478006362915, + "learning_rate": 0.00019801174193903714, + "loss": 1.0659, + "step": 1447 + }, + { + "epoch": 0.25783475783475784, + "grad_norm": 0.4424504339694977, + "learning_rate": 0.00019800896362345464, + "loss": 0.9805, + "step": 1448 + }, + { + "epoch": 0.25801282051282054, + "grad_norm": 0.4734833836555481, + "learning_rate": 0.0001980061833875901, + "loss": 1.255, + "step": 1449 + }, + { + "epoch": 0.2581908831908832, + "grad_norm": 0.41024845838546753, + "learning_rate": 0.000198003401231498, + "loss": 1.0908, + "step": 1450 + }, + { + "epoch": 0.2583689458689459, + "grad_norm": 0.43603816628456116, + "learning_rate": 0.00019800061715523283, + "loss": 1.0611, + "step": 1451 + }, + { + "epoch": 0.25854700854700857, + "grad_norm": 0.4871339499950409, + "learning_rate": 0.00019799783115884915, + "loss": 1.1851, + "step": 1452 + }, + { + "epoch": 0.2587250712250712, + "grad_norm": 0.49758270382881165, + "learning_rate": 0.00019799504324240157, + "loss": 1.1936, + "step": 1453 + }, + { + "epoch": 0.2589031339031339, + "grad_norm": 0.4201010763645172, + "learning_rate": 0.00019799225340594466, + "loss": 1.1567, + "step": 1454 + }, + { + "epoch": 0.2590811965811966, + "grad_norm": 0.4200313091278076, + "learning_rate": 0.00019798946164953309, + "loss": 0.9666, + "step": 1455 + }, + { + "epoch": 0.25925925925925924, + "grad_norm": 0.43001702427864075, + "learning_rate": 0.0001979866679732216, + "loss": 1.0104, + "step": 1456 + }, + { + "epoch": 0.25943732193732194, + "grad_norm": 0.46733465790748596, + "learning_rate": 0.0001979838723770649, + "loss": 1.0927, + "step": 1457 + }, + { + "epoch": 0.25961538461538464, + "grad_norm": 0.4513280391693115, + "learning_rate": 0.00019798107486111773, + "loss": 1.0282, + "step": 1458 + }, + { + "epoch": 0.2597934472934473, + "grad_norm": 0.40411749482154846, + "learning_rate": 0.00019797827542543495, + "loss": 1.0789, + "step": 1459 + }, + { + "epoch": 0.25997150997151, + "grad_norm": 0.4359099268913269, + "learning_rate": 0.0001979754740700714, + "loss": 1.0616, + "step": 1460 + }, + { + "epoch": 0.26014957264957267, + "grad_norm": 0.4979047477245331, + "learning_rate": 0.00019797267079508198, + "loss": 1.2948, + "step": 1461 + }, + { + "epoch": 0.2603276353276353, + "grad_norm": 0.44698619842529297, + "learning_rate": 0.0001979698656005216, + "loss": 0.9198, + "step": 1462 + }, + { + "epoch": 0.260505698005698, + "grad_norm": 0.48437631130218506, + "learning_rate": 0.00019796705848644516, + "loss": 1.3207, + "step": 1463 + }, + { + "epoch": 0.2606837606837607, + "grad_norm": 0.4382587671279907, + "learning_rate": 0.00019796424945290778, + "loss": 1.1315, + "step": 1464 + }, + { + "epoch": 0.26086182336182334, + "grad_norm": 0.4565944969654083, + "learning_rate": 0.0001979614384999644, + "loss": 1.1893, + "step": 1465 + }, + { + "epoch": 0.26103988603988604, + "grad_norm": 0.4705163836479187, + "learning_rate": 0.00019795862562767017, + "loss": 1.1132, + "step": 1466 + }, + { + "epoch": 0.26121794871794873, + "grad_norm": 0.525184690952301, + "learning_rate": 0.00019795581083608012, + "loss": 1.2111, + "step": 1467 + }, + { + "epoch": 0.2613960113960114, + "grad_norm": 0.45215457677841187, + "learning_rate": 0.00019795299412524945, + "loss": 1.1851, + "step": 1468 + }, + { + "epoch": 0.26157407407407407, + "grad_norm": 0.4336663484573364, + "learning_rate": 0.00019795017549523335, + "loss": 1.0147, + "step": 1469 + }, + { + "epoch": 0.26175213675213677, + "grad_norm": 0.5327649712562561, + "learning_rate": 0.00019794735494608703, + "loss": 1.1743, + "step": 1470 + }, + { + "epoch": 0.2619301994301994, + "grad_norm": 0.49972307682037354, + "learning_rate": 0.00019794453247786578, + "loss": 1.1624, + "step": 1471 + }, + { + "epoch": 0.2621082621082621, + "grad_norm": 0.43475785851478577, + "learning_rate": 0.00019794170809062485, + "loss": 0.9888, + "step": 1472 + }, + { + "epoch": 0.2622863247863248, + "grad_norm": 0.428838849067688, + "learning_rate": 0.0001979388817844196, + "loss": 0.9154, + "step": 1473 + }, + { + "epoch": 0.26246438746438744, + "grad_norm": 0.508568286895752, + "learning_rate": 0.00019793605355930544, + "loss": 1.1679, + "step": 1474 + }, + { + "epoch": 0.26264245014245013, + "grad_norm": 0.47791770100593567, + "learning_rate": 0.00019793322341533776, + "loss": 1.1375, + "step": 1475 + }, + { + "epoch": 0.26282051282051283, + "grad_norm": 0.41909220814704895, + "learning_rate": 0.00019793039135257196, + "loss": 1.0235, + "step": 1476 + }, + { + "epoch": 0.26299857549857547, + "grad_norm": 0.5564408302307129, + "learning_rate": 0.00019792755737106361, + "loss": 1.0756, + "step": 1477 + }, + { + "epoch": 0.26317663817663817, + "grad_norm": 0.42813625931739807, + "learning_rate": 0.0001979247214708682, + "loss": 0.8213, + "step": 1478 + }, + { + "epoch": 0.26335470085470086, + "grad_norm": 0.44495970010757446, + "learning_rate": 0.00019792188365204126, + "loss": 0.9654, + "step": 1479 + }, + { + "epoch": 0.26353276353276356, + "grad_norm": 0.47473424673080444, + "learning_rate": 0.00019791904391463846, + "loss": 1.1643, + "step": 1480 + }, + { + "epoch": 0.2637108262108262, + "grad_norm": 0.40189051628112793, + "learning_rate": 0.0001979162022587154, + "loss": 0.8687, + "step": 1481 + }, + { + "epoch": 0.2638888888888889, + "grad_norm": 0.44629937410354614, + "learning_rate": 0.00019791335868432776, + "loss": 1.0284, + "step": 1482 + }, + { + "epoch": 0.2640669515669516, + "grad_norm": 0.511275053024292, + "learning_rate": 0.00019791051319153124, + "loss": 1.2217, + "step": 1483 + }, + { + "epoch": 0.26424501424501423, + "grad_norm": 0.5136445164680481, + "learning_rate": 0.00019790766578038163, + "loss": 1.1129, + "step": 1484 + }, + { + "epoch": 0.2644230769230769, + "grad_norm": 0.4450451135635376, + "learning_rate": 0.00019790481645093469, + "loss": 0.9912, + "step": 1485 + }, + { + "epoch": 0.2646011396011396, + "grad_norm": 0.39455199241638184, + "learning_rate": 0.00019790196520324621, + "loss": 1.0887, + "step": 1486 + }, + { + "epoch": 0.26477920227920226, + "grad_norm": 0.4444045126438141, + "learning_rate": 0.00019789911203737216, + "loss": 1.1559, + "step": 1487 + }, + { + "epoch": 0.26495726495726496, + "grad_norm": 0.4769677221775055, + "learning_rate": 0.0001978962569533683, + "loss": 1.147, + "step": 1488 + }, + { + "epoch": 0.26513532763532766, + "grad_norm": 0.40226617455482483, + "learning_rate": 0.0001978933999512907, + "loss": 1.0966, + "step": 1489 + }, + { + "epoch": 0.2653133903133903, + "grad_norm": 0.4640974700450897, + "learning_rate": 0.00019789054103119526, + "loss": 1.1002, + "step": 1490 + }, + { + "epoch": 0.265491452991453, + "grad_norm": 0.48251107335090637, + "learning_rate": 0.00019788768019313806, + "loss": 1.07, + "step": 1491 + }, + { + "epoch": 0.2656695156695157, + "grad_norm": 0.4836949408054352, + "learning_rate": 0.00019788481743717506, + "loss": 1.2992, + "step": 1492 + }, + { + "epoch": 0.26584757834757833, + "grad_norm": 0.4253857135772705, + "learning_rate": 0.00019788195276336244, + "loss": 1.1326, + "step": 1493 + }, + { + "epoch": 0.266025641025641, + "grad_norm": 0.5161862373352051, + "learning_rate": 0.0001978790861717563, + "loss": 1.2131, + "step": 1494 + }, + { + "epoch": 0.2662037037037037, + "grad_norm": 0.5223346948623657, + "learning_rate": 0.00019787621766241274, + "loss": 1.0933, + "step": 1495 + }, + { + "epoch": 0.26638176638176636, + "grad_norm": 0.37622541189193726, + "learning_rate": 0.000197873347235388, + "loss": 0.8919, + "step": 1496 + }, + { + "epoch": 0.26655982905982906, + "grad_norm": 0.4425419569015503, + "learning_rate": 0.0001978704748907384, + "loss": 1.0411, + "step": 1497 + }, + { + "epoch": 0.26673789173789175, + "grad_norm": 0.4536985456943512, + "learning_rate": 0.00019786760062852015, + "loss": 1.2747, + "step": 1498 + }, + { + "epoch": 0.2669159544159544, + "grad_norm": 0.4998049736022949, + "learning_rate": 0.00019786472444878955, + "loss": 1.3214, + "step": 1499 + }, + { + "epoch": 0.2670940170940171, + "grad_norm": 0.42104312777519226, + "learning_rate": 0.00019786184635160295, + "loss": 0.7878, + "step": 1500 + }, + { + "epoch": 0.2672720797720798, + "grad_norm": 0.5354288220405579, + "learning_rate": 0.00019785896633701678, + "loss": 1.0642, + "step": 1501 + }, + { + "epoch": 0.2674501424501424, + "grad_norm": 0.4681485891342163, + "learning_rate": 0.00019785608440508744, + "loss": 1.1737, + "step": 1502 + }, + { + "epoch": 0.2676282051282051, + "grad_norm": 0.49107062816619873, + "learning_rate": 0.0001978532005558714, + "loss": 1.1507, + "step": 1503 + }, + { + "epoch": 0.2678062678062678, + "grad_norm": 0.4173283576965332, + "learning_rate": 0.0001978503147894252, + "loss": 1.0538, + "step": 1504 + }, + { + "epoch": 0.26798433048433046, + "grad_norm": 0.49354055523872375, + "learning_rate": 0.0001978474271058053, + "loss": 1.1043, + "step": 1505 + }, + { + "epoch": 0.26816239316239315, + "grad_norm": 0.5787215232849121, + "learning_rate": 0.00019784453750506834, + "loss": 0.9245, + "step": 1506 + }, + { + "epoch": 0.26834045584045585, + "grad_norm": 0.48982590436935425, + "learning_rate": 0.00019784164598727095, + "loss": 1.2007, + "step": 1507 + }, + { + "epoch": 0.26851851851851855, + "grad_norm": 0.4971007704734802, + "learning_rate": 0.00019783875255246973, + "loss": 1.1174, + "step": 1508 + }, + { + "epoch": 0.2686965811965812, + "grad_norm": 0.5200340151786804, + "learning_rate": 0.00019783585720072142, + "loss": 1.1967, + "step": 1509 + }, + { + "epoch": 0.2688746438746439, + "grad_norm": 0.47911885380744934, + "learning_rate": 0.00019783295993208271, + "loss": 1.162, + "step": 1510 + }, + { + "epoch": 0.2690527065527066, + "grad_norm": 0.4764275848865509, + "learning_rate": 0.00019783006074661037, + "loss": 1.1358, + "step": 1511 + }, + { + "epoch": 0.2692307692307692, + "grad_norm": 0.478545606136322, + "learning_rate": 0.00019782715964436124, + "loss": 1.0096, + "step": 1512 + }, + { + "epoch": 0.2694088319088319, + "grad_norm": 0.5512787699699402, + "learning_rate": 0.00019782425662539212, + "loss": 1.1799, + "step": 1513 + }, + { + "epoch": 0.2695868945868946, + "grad_norm": 0.5495108962059021, + "learning_rate": 0.00019782135168975988, + "loss": 1.0959, + "step": 1514 + }, + { + "epoch": 0.26976495726495725, + "grad_norm": 0.42052868008613586, + "learning_rate": 0.0001978184448375215, + "loss": 1.1872, + "step": 1515 + }, + { + "epoch": 0.26994301994301995, + "grad_norm": 0.4994426965713501, + "learning_rate": 0.0001978155360687339, + "loss": 1.0568, + "step": 1516 + }, + { + "epoch": 0.27012108262108264, + "grad_norm": 0.459577351808548, + "learning_rate": 0.00019781262538345402, + "loss": 1.0315, + "step": 1517 + }, + { + "epoch": 0.2702991452991453, + "grad_norm": 0.4792841374874115, + "learning_rate": 0.00019780971278173895, + "loss": 1.2055, + "step": 1518 + }, + { + "epoch": 0.270477207977208, + "grad_norm": 0.5017708539962769, + "learning_rate": 0.00019780679826364575, + "loss": 1.157, + "step": 1519 + }, + { + "epoch": 0.2706552706552707, + "grad_norm": 0.5197349786758423, + "learning_rate": 0.00019780388182923152, + "loss": 0.9101, + "step": 1520 + }, + { + "epoch": 0.2708333333333333, + "grad_norm": 0.4226742684841156, + "learning_rate": 0.00019780096347855338, + "loss": 1.0525, + "step": 1521 + }, + { + "epoch": 0.271011396011396, + "grad_norm": 0.5058164596557617, + "learning_rate": 0.00019779804321166852, + "loss": 0.931, + "step": 1522 + }, + { + "epoch": 0.2711894586894587, + "grad_norm": 0.44492244720458984, + "learning_rate": 0.00019779512102863418, + "loss": 1.0641, + "step": 1523 + }, + { + "epoch": 0.27136752136752135, + "grad_norm": 0.5348989963531494, + "learning_rate": 0.00019779219692950758, + "loss": 1.1692, + "step": 1524 + }, + { + "epoch": 0.27154558404558404, + "grad_norm": 0.4631774425506592, + "learning_rate": 0.00019778927091434602, + "loss": 1.0876, + "step": 1525 + }, + { + "epoch": 0.27172364672364674, + "grad_norm": 0.45957499742507935, + "learning_rate": 0.00019778634298320684, + "loss": 0.9527, + "step": 1526 + }, + { + "epoch": 0.2719017094017094, + "grad_norm": 0.4506755769252777, + "learning_rate": 0.00019778341313614743, + "loss": 1.086, + "step": 1527 + }, + { + "epoch": 0.2720797720797721, + "grad_norm": 0.4900587797164917, + "learning_rate": 0.00019778048137322513, + "loss": 0.9911, + "step": 1528 + }, + { + "epoch": 0.27225783475783477, + "grad_norm": 0.478127658367157, + "learning_rate": 0.00019777754769449745, + "loss": 1.2083, + "step": 1529 + }, + { + "epoch": 0.2724358974358974, + "grad_norm": 0.47220897674560547, + "learning_rate": 0.00019777461210002183, + "loss": 1.0313, + "step": 1530 + }, + { + "epoch": 0.2726139601139601, + "grad_norm": 0.4526277184486389, + "learning_rate": 0.0001977716745898558, + "loss": 1.2648, + "step": 1531 + }, + { + "epoch": 0.2727920227920228, + "grad_norm": 0.42907601594924927, + "learning_rate": 0.00019776873516405688, + "loss": 0.8645, + "step": 1532 + }, + { + "epoch": 0.27297008547008544, + "grad_norm": 0.43440163135528564, + "learning_rate": 0.00019776579382268272, + "loss": 0.9702, + "step": 1533 + }, + { + "epoch": 0.27314814814814814, + "grad_norm": 0.48213550448417664, + "learning_rate": 0.0001977628505657909, + "loss": 0.998, + "step": 1534 + }, + { + "epoch": 0.27332621082621084, + "grad_norm": 0.43385565280914307, + "learning_rate": 0.00019775990539343914, + "loss": 1.0575, + "step": 1535 + }, + { + "epoch": 0.27350427350427353, + "grad_norm": 0.45706847310066223, + "learning_rate": 0.00019775695830568507, + "loss": 1.3024, + "step": 1536 + }, + { + "epoch": 0.27368233618233617, + "grad_norm": 0.45769137144088745, + "learning_rate": 0.00019775400930258652, + "loss": 1.0987, + "step": 1537 + }, + { + "epoch": 0.27386039886039887, + "grad_norm": 0.44682395458221436, + "learning_rate": 0.00019775105838420117, + "loss": 1.1327, + "step": 1538 + }, + { + "epoch": 0.27403846153846156, + "grad_norm": 0.5923072099685669, + "learning_rate": 0.00019774810555058694, + "loss": 1.4766, + "step": 1539 + }, + { + "epoch": 0.2742165242165242, + "grad_norm": 0.4327206015586853, + "learning_rate": 0.0001977451508018016, + "loss": 1.1175, + "step": 1540 + }, + { + "epoch": 0.2743945868945869, + "grad_norm": 0.48036691546440125, + "learning_rate": 0.00019774219413790315, + "loss": 1.1189, + "step": 1541 + }, + { + "epoch": 0.2745726495726496, + "grad_norm": 0.41371914744377136, + "learning_rate": 0.00019773923555894935, + "loss": 1.1366, + "step": 1542 + }, + { + "epoch": 0.27475071225071224, + "grad_norm": 0.4452378749847412, + "learning_rate": 0.00019773627506499832, + "loss": 0.9517, + "step": 1543 + }, + { + "epoch": 0.27492877492877493, + "grad_norm": 0.469098299741745, + "learning_rate": 0.00019773331265610802, + "loss": 1.0848, + "step": 1544 + }, + { + "epoch": 0.27510683760683763, + "grad_norm": 0.5390294790267944, + "learning_rate": 0.00019773034833233646, + "loss": 0.8589, + "step": 1545 + }, + { + "epoch": 0.27528490028490027, + "grad_norm": 0.5368238091468811, + "learning_rate": 0.00019772738209374174, + "loss": 1.2954, + "step": 1546 + }, + { + "epoch": 0.27546296296296297, + "grad_norm": 0.4705318510532379, + "learning_rate": 0.00019772441394038198, + "loss": 1.2252, + "step": 1547 + }, + { + "epoch": 0.27564102564102566, + "grad_norm": 0.4682813286781311, + "learning_rate": 0.00019772144387231533, + "loss": 1.0855, + "step": 1548 + }, + { + "epoch": 0.2758190883190883, + "grad_norm": 0.46876460313796997, + "learning_rate": 0.0001977184718896, + "loss": 1.1959, + "step": 1549 + }, + { + "epoch": 0.275997150997151, + "grad_norm": 0.4172806441783905, + "learning_rate": 0.00019771549799229416, + "loss": 1.2166, + "step": 1550 + }, + { + "epoch": 0.2761752136752137, + "grad_norm": 0.5088075399398804, + "learning_rate": 0.0001977125221804562, + "loss": 1.1285, + "step": 1551 + }, + { + "epoch": 0.27635327635327633, + "grad_norm": 0.4728628396987915, + "learning_rate": 0.0001977095444541443, + "loss": 1.2985, + "step": 1552 + }, + { + "epoch": 0.27653133903133903, + "grad_norm": 0.4431236684322357, + "learning_rate": 0.00019770656481341684, + "loss": 1.1298, + "step": 1553 + }, + { + "epoch": 0.2767094017094017, + "grad_norm": 0.474065363407135, + "learning_rate": 0.00019770358325833223, + "loss": 1.1915, + "step": 1554 + }, + { + "epoch": 0.27688746438746437, + "grad_norm": 0.45718875527381897, + "learning_rate": 0.00019770059978894885, + "loss": 1.0626, + "step": 1555 + }, + { + "epoch": 0.27706552706552706, + "grad_norm": 0.49300211668014526, + "learning_rate": 0.00019769761440532522, + "loss": 1.0134, + "step": 1556 + }, + { + "epoch": 0.27724358974358976, + "grad_norm": 0.4389498829841614, + "learning_rate": 0.00019769462710751974, + "loss": 1.0292, + "step": 1557 + }, + { + "epoch": 0.2774216524216524, + "grad_norm": 0.47330448031425476, + "learning_rate": 0.000197691637895591, + "loss": 1.1273, + "step": 1558 + }, + { + "epoch": 0.2775997150997151, + "grad_norm": 0.5322058200836182, + "learning_rate": 0.00019768864676959755, + "loss": 1.059, + "step": 1559 + }, + { + "epoch": 0.2777777777777778, + "grad_norm": 0.4714536964893341, + "learning_rate": 0.000197685653729598, + "loss": 1.1987, + "step": 1560 + }, + { + "epoch": 0.27795584045584043, + "grad_norm": 0.48687809705734253, + "learning_rate": 0.00019768265877565097, + "loss": 1.3206, + "step": 1561 + }, + { + "epoch": 0.2781339031339031, + "grad_norm": 0.46066713333129883, + "learning_rate": 0.00019767966190781518, + "loss": 1.0845, + "step": 1562 + }, + { + "epoch": 0.2783119658119658, + "grad_norm": 0.44372090697288513, + "learning_rate": 0.00019767666312614935, + "loss": 1.0942, + "step": 1563 + }, + { + "epoch": 0.27849002849002846, + "grad_norm": 0.4615907073020935, + "learning_rate": 0.00019767366243071216, + "loss": 1.071, + "step": 1564 + }, + { + "epoch": 0.27866809116809116, + "grad_norm": 0.502097487449646, + "learning_rate": 0.0001976706598215625, + "loss": 1.1164, + "step": 1565 + }, + { + "epoch": 0.27884615384615385, + "grad_norm": 0.4371815621852875, + "learning_rate": 0.00019766765529875913, + "loss": 1.0252, + "step": 1566 + }, + { + "epoch": 0.27902421652421655, + "grad_norm": 0.43035808205604553, + "learning_rate": 0.00019766464886236093, + "loss": 1.073, + "step": 1567 + }, + { + "epoch": 0.2792022792022792, + "grad_norm": 0.49721601605415344, + "learning_rate": 0.00019766164051242683, + "loss": 1.0316, + "step": 1568 + }, + { + "epoch": 0.2793803418803419, + "grad_norm": 0.44866231083869934, + "learning_rate": 0.00019765863024901576, + "loss": 1.0951, + "step": 1569 + }, + { + "epoch": 0.2795584045584046, + "grad_norm": 0.46318337321281433, + "learning_rate": 0.0001976556180721867, + "loss": 0.9836, + "step": 1570 + }, + { + "epoch": 0.2797364672364672, + "grad_norm": 0.4227696657180786, + "learning_rate": 0.00019765260398199868, + "loss": 1.0414, + "step": 1571 + }, + { + "epoch": 0.2799145299145299, + "grad_norm": 0.6062980890274048, + "learning_rate": 0.00019764958797851073, + "loss": 1.137, + "step": 1572 + }, + { + "epoch": 0.2800925925925926, + "grad_norm": 0.4856833219528198, + "learning_rate": 0.00019764657006178196, + "loss": 1.1361, + "step": 1573 + }, + { + "epoch": 0.28027065527065526, + "grad_norm": 0.45612895488739014, + "learning_rate": 0.00019764355023187146, + "loss": 1.0005, + "step": 1574 + }, + { + "epoch": 0.28044871794871795, + "grad_norm": 0.4143696129322052, + "learning_rate": 0.00019764052848883845, + "loss": 1.051, + "step": 1575 + }, + { + "epoch": 0.28062678062678065, + "grad_norm": 0.4532071352005005, + "learning_rate": 0.00019763750483274212, + "loss": 1.0595, + "step": 1576 + }, + { + "epoch": 0.2808048433048433, + "grad_norm": 0.4940357208251953, + "learning_rate": 0.0001976344792636417, + "loss": 1.0983, + "step": 1577 + }, + { + "epoch": 0.280982905982906, + "grad_norm": 0.44405099749565125, + "learning_rate": 0.0001976314517815965, + "loss": 1.0846, + "step": 1578 + }, + { + "epoch": 0.2811609686609687, + "grad_norm": 0.5508625507354736, + "learning_rate": 0.00019762842238666578, + "loss": 1.1722, + "step": 1579 + }, + { + "epoch": 0.2813390313390313, + "grad_norm": 0.5241084694862366, + "learning_rate": 0.00019762539107890894, + "loss": 1.351, + "step": 1580 + }, + { + "epoch": 0.281517094017094, + "grad_norm": 0.5307353734970093, + "learning_rate": 0.00019762235785838537, + "loss": 1.1868, + "step": 1581 + }, + { + "epoch": 0.2816951566951567, + "grad_norm": 0.45697924494743347, + "learning_rate": 0.00019761932272515447, + "loss": 1.1982, + "step": 1582 + }, + { + "epoch": 0.28187321937321935, + "grad_norm": 0.412483811378479, + "learning_rate": 0.00019761628567927574, + "loss": 1.0433, + "step": 1583 + }, + { + "epoch": 0.28205128205128205, + "grad_norm": 0.4614165425300598, + "learning_rate": 0.00019761324672080868, + "loss": 1.104, + "step": 1584 + }, + { + "epoch": 0.28222934472934474, + "grad_norm": 0.47644901275634766, + "learning_rate": 0.00019761020584981284, + "loss": 1.1037, + "step": 1585 + }, + { + "epoch": 0.2824074074074074, + "grad_norm": 0.4985184669494629, + "learning_rate": 0.00019760716306634773, + "loss": 1.2213, + "step": 1586 + }, + { + "epoch": 0.2825854700854701, + "grad_norm": 0.508301317691803, + "learning_rate": 0.00019760411837047305, + "loss": 1.1315, + "step": 1587 + }, + { + "epoch": 0.2827635327635328, + "grad_norm": 0.5346587300300598, + "learning_rate": 0.00019760107176224845, + "loss": 1.2281, + "step": 1588 + }, + { + "epoch": 0.2829415954415954, + "grad_norm": 0.5106825232505798, + "learning_rate": 0.00019759802324173357, + "loss": 1.2904, + "step": 1589 + }, + { + "epoch": 0.2831196581196581, + "grad_norm": 0.46458688378334045, + "learning_rate": 0.00019759497280898817, + "loss": 1.0861, + "step": 1590 + }, + { + "epoch": 0.2832977207977208, + "grad_norm": 0.49115365743637085, + "learning_rate": 0.00019759192046407201, + "loss": 1.0529, + "step": 1591 + }, + { + "epoch": 0.28347578347578345, + "grad_norm": 0.5114167332649231, + "learning_rate": 0.0001975888662070449, + "loss": 1.2555, + "step": 1592 + }, + { + "epoch": 0.28365384615384615, + "grad_norm": 0.45844775438308716, + "learning_rate": 0.0001975858100379667, + "loss": 1.0662, + "step": 1593 + }, + { + "epoch": 0.28383190883190884, + "grad_norm": 0.4684161841869354, + "learning_rate": 0.00019758275195689727, + "loss": 1.0537, + "step": 1594 + }, + { + "epoch": 0.28400997150997154, + "grad_norm": 0.4816220998764038, + "learning_rate": 0.0001975796919638965, + "loss": 1.126, + "step": 1595 + }, + { + "epoch": 0.2841880341880342, + "grad_norm": 0.46578118205070496, + "learning_rate": 0.0001975766300590244, + "loss": 0.9651, + "step": 1596 + }, + { + "epoch": 0.2843660968660969, + "grad_norm": 0.4181675612926483, + "learning_rate": 0.0001975735662423409, + "loss": 1.0888, + "step": 1597 + }, + { + "epoch": 0.28454415954415957, + "grad_norm": 0.49417954683303833, + "learning_rate": 0.00019757050051390609, + "loss": 1.1878, + "step": 1598 + }, + { + "epoch": 0.2847222222222222, + "grad_norm": 0.47264960408210754, + "learning_rate": 0.00019756743287377998, + "loss": 1.027, + "step": 1599 + }, + { + "epoch": 0.2849002849002849, + "grad_norm": 0.47686338424682617, + "learning_rate": 0.0001975643633220227, + "loss": 1.1307, + "step": 1600 + }, + { + "epoch": 0.2850783475783476, + "grad_norm": 0.5571266412734985, + "learning_rate": 0.00019756129185869443, + "loss": 0.984, + "step": 1601 + }, + { + "epoch": 0.28525641025641024, + "grad_norm": 0.46942809224128723, + "learning_rate": 0.00019755821848385527, + "loss": 1.0397, + "step": 1602 + }, + { + "epoch": 0.28543447293447294, + "grad_norm": 0.6325890421867371, + "learning_rate": 0.00019755514319756551, + "loss": 1.0918, + "step": 1603 + }, + { + "epoch": 0.28561253561253563, + "grad_norm": 0.5297608375549316, + "learning_rate": 0.00019755206599988533, + "loss": 0.9911, + "step": 1604 + }, + { + "epoch": 0.2857905982905983, + "grad_norm": 0.4736945331096649, + "learning_rate": 0.00019754898689087512, + "loss": 1.0786, + "step": 1605 + }, + { + "epoch": 0.28596866096866097, + "grad_norm": 0.5048685669898987, + "learning_rate": 0.00019754590587059512, + "loss": 0.9834, + "step": 1606 + }, + { + "epoch": 0.28614672364672367, + "grad_norm": 0.3823149502277374, + "learning_rate": 0.00019754282293910574, + "loss": 0.8341, + "step": 1607 + }, + { + "epoch": 0.2863247863247863, + "grad_norm": 0.44071945548057556, + "learning_rate": 0.00019753973809646738, + "loss": 1.131, + "step": 1608 + }, + { + "epoch": 0.286502849002849, + "grad_norm": 0.44182759523391724, + "learning_rate": 0.00019753665134274043, + "loss": 1.0321, + "step": 1609 + }, + { + "epoch": 0.2866809116809117, + "grad_norm": 0.4486250877380371, + "learning_rate": 0.00019753356267798546, + "loss": 0.9941, + "step": 1610 + }, + { + "epoch": 0.28685897435897434, + "grad_norm": 0.42796584963798523, + "learning_rate": 0.00019753047210226292, + "loss": 1.0235, + "step": 1611 + }, + { + "epoch": 0.28703703703703703, + "grad_norm": 0.47294023633003235, + "learning_rate": 0.00019752737961563336, + "loss": 1.11, + "step": 1612 + }, + { + "epoch": 0.28721509971509973, + "grad_norm": 0.44550734758377075, + "learning_rate": 0.00019752428521815742, + "loss": 1.0849, + "step": 1613 + }, + { + "epoch": 0.28739316239316237, + "grad_norm": 0.44189929962158203, + "learning_rate": 0.0001975211889098957, + "loss": 0.8904, + "step": 1614 + }, + { + "epoch": 0.28757122507122507, + "grad_norm": 0.5302733182907104, + "learning_rate": 0.00019751809069090885, + "loss": 1.2348, + "step": 1615 + }, + { + "epoch": 0.28774928774928776, + "grad_norm": 0.5951390862464905, + "learning_rate": 0.00019751499056125762, + "loss": 1.3035, + "step": 1616 + }, + { + "epoch": 0.2879273504273504, + "grad_norm": 0.5431534647941589, + "learning_rate": 0.0001975118885210027, + "loss": 1.0016, + "step": 1617 + }, + { + "epoch": 0.2881054131054131, + "grad_norm": 0.47301986813545227, + "learning_rate": 0.00019750878457020489, + "loss": 1.2245, + "step": 1618 + }, + { + "epoch": 0.2882834757834758, + "grad_norm": 0.44785359501838684, + "learning_rate": 0.00019750567870892497, + "loss": 1.122, + "step": 1619 + }, + { + "epoch": 0.28846153846153844, + "grad_norm": 0.49494361877441406, + "learning_rate": 0.00019750257093722383, + "loss": 0.9421, + "step": 1620 + }, + { + "epoch": 0.28863960113960113, + "grad_norm": 0.4484521150588989, + "learning_rate": 0.00019749946125516242, + "loss": 1.2146, + "step": 1621 + }, + { + "epoch": 0.28881766381766383, + "grad_norm": 0.4635269343852997, + "learning_rate": 0.00019749634966280156, + "loss": 0.976, + "step": 1622 + }, + { + "epoch": 0.28899572649572647, + "grad_norm": 0.5532249808311462, + "learning_rate": 0.00019749323616020226, + "loss": 1.1818, + "step": 1623 + }, + { + "epoch": 0.28917378917378916, + "grad_norm": 0.4730629622936249, + "learning_rate": 0.00019749012074742552, + "loss": 1.0321, + "step": 1624 + }, + { + "epoch": 0.28935185185185186, + "grad_norm": 0.47437289357185364, + "learning_rate": 0.0001974870034245324, + "loss": 1.1572, + "step": 1625 + }, + { + "epoch": 0.28952991452991456, + "grad_norm": 0.4796304404735565, + "learning_rate": 0.00019748388419158394, + "loss": 1.1667, + "step": 1626 + }, + { + "epoch": 0.2897079772079772, + "grad_norm": 0.42686304450035095, + "learning_rate": 0.0001974807630486413, + "loss": 0.9824, + "step": 1627 + }, + { + "epoch": 0.2898860398860399, + "grad_norm": 0.4444865584373474, + "learning_rate": 0.00019747763999576558, + "loss": 1.2789, + "step": 1628 + }, + { + "epoch": 0.2900641025641026, + "grad_norm": 0.5039985179901123, + "learning_rate": 0.000197474515033018, + "loss": 1.1488, + "step": 1629 + }, + { + "epoch": 0.29024216524216523, + "grad_norm": 0.581479549407959, + "learning_rate": 0.00019747138816045978, + "loss": 1.1232, + "step": 1630 + }, + { + "epoch": 0.2904202279202279, + "grad_norm": 0.5415821075439453, + "learning_rate": 0.00019746825937815222, + "loss": 1.2326, + "step": 1631 + }, + { + "epoch": 0.2905982905982906, + "grad_norm": 0.45528364181518555, + "learning_rate": 0.00019746512868615656, + "loss": 1.0246, + "step": 1632 + }, + { + "epoch": 0.29077635327635326, + "grad_norm": 0.5255574584007263, + "learning_rate": 0.00019746199608453418, + "loss": 1.0592, + "step": 1633 + }, + { + "epoch": 0.29095441595441596, + "grad_norm": 0.5064096450805664, + "learning_rate": 0.00019745886157334646, + "loss": 1.3439, + "step": 1634 + }, + { + "epoch": 0.29113247863247865, + "grad_norm": 0.500848650932312, + "learning_rate": 0.00019745572515265475, + "loss": 1.1212, + "step": 1635 + }, + { + "epoch": 0.2913105413105413, + "grad_norm": 0.5229088068008423, + "learning_rate": 0.00019745258682252062, + "loss": 1.1019, + "step": 1636 + }, + { + "epoch": 0.291488603988604, + "grad_norm": 0.4494398832321167, + "learning_rate": 0.00019744944658300545, + "loss": 1.1298, + "step": 1637 + }, + { + "epoch": 0.2916666666666667, + "grad_norm": 0.48383277654647827, + "learning_rate": 0.00019744630443417082, + "loss": 1.206, + "step": 1638 + }, + { + "epoch": 0.2918447293447293, + "grad_norm": 0.4870131313800812, + "learning_rate": 0.00019744316037607828, + "loss": 1.2096, + "step": 1639 + }, + { + "epoch": 0.292022792022792, + "grad_norm": 0.4153090715408325, + "learning_rate": 0.00019744001440878944, + "loss": 1.0478, + "step": 1640 + }, + { + "epoch": 0.2922008547008547, + "grad_norm": 0.4262249171733856, + "learning_rate": 0.0001974368665323659, + "loss": 1.0393, + "step": 1641 + }, + { + "epoch": 0.29237891737891736, + "grad_norm": 0.46131134033203125, + "learning_rate": 0.00019743371674686938, + "loss": 1.0908, + "step": 1642 + }, + { + "epoch": 0.29255698005698005, + "grad_norm": 0.44877463579177856, + "learning_rate": 0.0001974305650523616, + "loss": 1.1906, + "step": 1643 + }, + { + "epoch": 0.29273504273504275, + "grad_norm": 0.5199326276779175, + "learning_rate": 0.00019742741144890432, + "loss": 1.1147, + "step": 1644 + }, + { + "epoch": 0.2929131054131054, + "grad_norm": 0.48142504692077637, + "learning_rate": 0.00019742425593655924, + "loss": 1.1951, + "step": 1645 + }, + { + "epoch": 0.2930911680911681, + "grad_norm": 0.5672988891601562, + "learning_rate": 0.0001974210985153883, + "loss": 1.1817, + "step": 1646 + }, + { + "epoch": 0.2932692307692308, + "grad_norm": 0.38135233521461487, + "learning_rate": 0.00019741793918545326, + "loss": 0.8567, + "step": 1647 + }, + { + "epoch": 0.2934472934472934, + "grad_norm": 0.6153588891029358, + "learning_rate": 0.0001974147779468161, + "loss": 1.0593, + "step": 1648 + }, + { + "epoch": 0.2936253561253561, + "grad_norm": 0.38935527205467224, + "learning_rate": 0.0001974116147995387, + "loss": 0.9907, + "step": 1649 + }, + { + "epoch": 0.2938034188034188, + "grad_norm": 0.467351496219635, + "learning_rate": 0.0001974084497436831, + "loss": 1.091, + "step": 1650 + }, + { + "epoch": 0.29398148148148145, + "grad_norm": 0.45613420009613037, + "learning_rate": 0.00019740528277931128, + "loss": 0.6789, + "step": 1651 + }, + { + "epoch": 0.29415954415954415, + "grad_norm": 0.4045158326625824, + "learning_rate": 0.00019740211390648524, + "loss": 1.0727, + "step": 1652 + }, + { + "epoch": 0.29433760683760685, + "grad_norm": 0.5122803449630737, + "learning_rate": 0.00019739894312526714, + "loss": 1.2297, + "step": 1653 + }, + { + "epoch": 0.29451566951566954, + "grad_norm": 0.44304123520851135, + "learning_rate": 0.00019739577043571908, + "loss": 0.9562, + "step": 1654 + }, + { + "epoch": 0.2946937321937322, + "grad_norm": 0.6070618629455566, + "learning_rate": 0.00019739259583790322, + "loss": 1.2745, + "step": 1655 + }, + { + "epoch": 0.2948717948717949, + "grad_norm": 0.48815637826919556, + "learning_rate": 0.00019738941933188176, + "loss": 1.0574, + "step": 1656 + }, + { + "epoch": 0.2950498575498576, + "grad_norm": 0.5067802667617798, + "learning_rate": 0.00019738624091771693, + "loss": 1.1874, + "step": 1657 + }, + { + "epoch": 0.2952279202279202, + "grad_norm": 0.4956928491592407, + "learning_rate": 0.000197383060595471, + "loss": 1.1085, + "step": 1658 + }, + { + "epoch": 0.2954059829059829, + "grad_norm": 0.46313008666038513, + "learning_rate": 0.00019737987836520633, + "loss": 1.0548, + "step": 1659 + }, + { + "epoch": 0.2955840455840456, + "grad_norm": 0.49944064021110535, + "learning_rate": 0.0001973766942269852, + "loss": 1.1485, + "step": 1660 + }, + { + "epoch": 0.29576210826210825, + "grad_norm": 0.4743517339229584, + "learning_rate": 0.00019737350818087003, + "loss": 0.9279, + "step": 1661 + }, + { + "epoch": 0.29594017094017094, + "grad_norm": 0.45935431122779846, + "learning_rate": 0.00019737032022692326, + "loss": 0.9574, + "step": 1662 + }, + { + "epoch": 0.29611823361823364, + "grad_norm": 0.4550873637199402, + "learning_rate": 0.00019736713036520734, + "loss": 1.1642, + "step": 1663 + }, + { + "epoch": 0.2962962962962963, + "grad_norm": 0.45252951979637146, + "learning_rate": 0.00019736393859578474, + "loss": 1.0113, + "step": 1664 + }, + { + "epoch": 0.296474358974359, + "grad_norm": 0.5147238969802856, + "learning_rate": 0.00019736074491871804, + "loss": 1.1604, + "step": 1665 + }, + { + "epoch": 0.29665242165242167, + "grad_norm": 0.5122934579849243, + "learning_rate": 0.00019735754933406977, + "loss": 0.9525, + "step": 1666 + }, + { + "epoch": 0.2968304843304843, + "grad_norm": 0.438620001077652, + "learning_rate": 0.00019735435184190257, + "loss": 1.0728, + "step": 1667 + }, + { + "epoch": 0.297008547008547, + "grad_norm": 0.41970670223236084, + "learning_rate": 0.00019735115244227908, + "loss": 0.9782, + "step": 1668 + }, + { + "epoch": 0.2971866096866097, + "grad_norm": 0.5447152256965637, + "learning_rate": 0.000197347951135262, + "loss": 1.0633, + "step": 1669 + }, + { + "epoch": 0.29736467236467234, + "grad_norm": 0.4846996068954468, + "learning_rate": 0.00019734474792091407, + "loss": 0.9019, + "step": 1670 + }, + { + "epoch": 0.29754273504273504, + "grad_norm": 0.4721437990665436, + "learning_rate": 0.00019734154279929796, + "loss": 1.1793, + "step": 1671 + }, + { + "epoch": 0.29772079772079774, + "grad_norm": 0.4659852385520935, + "learning_rate": 0.00019733833577047655, + "loss": 1.1503, + "step": 1672 + }, + { + "epoch": 0.2978988603988604, + "grad_norm": 0.3733183443546295, + "learning_rate": 0.00019733512683451268, + "loss": 0.7763, + "step": 1673 + }, + { + "epoch": 0.2980769230769231, + "grad_norm": 0.4898292124271393, + "learning_rate": 0.0001973319159914692, + "loss": 1.3146, + "step": 1674 + }, + { + "epoch": 0.29825498575498577, + "grad_norm": 0.41774725914001465, + "learning_rate": 0.00019732870324140899, + "loss": 1.2069, + "step": 1675 + }, + { + "epoch": 0.2984330484330484, + "grad_norm": 0.4607912003993988, + "learning_rate": 0.000197325488584395, + "loss": 1.2255, + "step": 1676 + }, + { + "epoch": 0.2986111111111111, + "grad_norm": 0.4692424237728119, + "learning_rate": 0.00019732227202049025, + "loss": 1.0793, + "step": 1677 + }, + { + "epoch": 0.2987891737891738, + "grad_norm": 0.5925022959709167, + "learning_rate": 0.00019731905354975778, + "loss": 1.0297, + "step": 1678 + }, + { + "epoch": 0.29896723646723644, + "grad_norm": 0.44047990441322327, + "learning_rate": 0.00019731583317226056, + "loss": 1.0982, + "step": 1679 + }, + { + "epoch": 0.29914529914529914, + "grad_norm": 0.5863066911697388, + "learning_rate": 0.0001973126108880618, + "loss": 1.0284, + "step": 1680 + }, + { + "epoch": 0.29932336182336183, + "grad_norm": 0.48962152004241943, + "learning_rate": 0.00019730938669722457, + "loss": 1.1861, + "step": 1681 + }, + { + "epoch": 0.29950142450142453, + "grad_norm": 0.5445577502250671, + "learning_rate": 0.00019730616059981205, + "loss": 1.2574, + "step": 1682 + }, + { + "epoch": 0.29967948717948717, + "grad_norm": 0.49327564239501953, + "learning_rate": 0.00019730293259588743, + "loss": 0.9578, + "step": 1683 + }, + { + "epoch": 0.29985754985754987, + "grad_norm": 0.4252840578556061, + "learning_rate": 0.00019729970268551398, + "loss": 1.0083, + "step": 1684 + }, + { + "epoch": 0.30003561253561256, + "grad_norm": 0.5140926241874695, + "learning_rate": 0.000197296470868755, + "loss": 1.3263, + "step": 1685 + }, + { + "epoch": 0.3002136752136752, + "grad_norm": 0.5143948197364807, + "learning_rate": 0.00019729323714567375, + "loss": 1.0424, + "step": 1686 + }, + { + "epoch": 0.3003917378917379, + "grad_norm": 0.3811354339122772, + "learning_rate": 0.00019729000151633367, + "loss": 0.6319, + "step": 1687 + }, + { + "epoch": 0.3005698005698006, + "grad_norm": 0.5249716639518738, + "learning_rate": 0.0001972867639807981, + "loss": 1.0173, + "step": 1688 + }, + { + "epoch": 0.30074786324786323, + "grad_norm": 0.41832098364830017, + "learning_rate": 0.00019728352453913048, + "loss": 1.0503, + "step": 1689 + }, + { + "epoch": 0.30092592592592593, + "grad_norm": 0.5961149334907532, + "learning_rate": 0.00019728028319139428, + "loss": 1.1843, + "step": 1690 + }, + { + "epoch": 0.3011039886039886, + "grad_norm": 0.44083690643310547, + "learning_rate": 0.00019727703993765303, + "loss": 1.1311, + "step": 1691 + }, + { + "epoch": 0.30128205128205127, + "grad_norm": 0.4368111491203308, + "learning_rate": 0.00019727379477797022, + "loss": 0.9463, + "step": 1692 + }, + { + "epoch": 0.30146011396011396, + "grad_norm": 0.5289376974105835, + "learning_rate": 0.00019727054771240954, + "loss": 0.9836, + "step": 1693 + }, + { + "epoch": 0.30163817663817666, + "grad_norm": 0.4132843613624573, + "learning_rate": 0.00019726729874103448, + "loss": 1.1052, + "step": 1694 + }, + { + "epoch": 0.3018162393162393, + "grad_norm": 0.4919086992740631, + "learning_rate": 0.00019726404786390877, + "loss": 1.2219, + "step": 1695 + }, + { + "epoch": 0.301994301994302, + "grad_norm": 0.42561691999435425, + "learning_rate": 0.0001972607950810961, + "loss": 1.0756, + "step": 1696 + }, + { + "epoch": 0.3021723646723647, + "grad_norm": 0.5030396580696106, + "learning_rate": 0.0001972575403926602, + "loss": 1.2207, + "step": 1697 + }, + { + "epoch": 0.30235042735042733, + "grad_norm": 0.4779801666736603, + "learning_rate": 0.0001972542837986648, + "loss": 1.194, + "step": 1698 + }, + { + "epoch": 0.30252849002849, + "grad_norm": 0.45395568013191223, + "learning_rate": 0.00019725102529917377, + "loss": 1.0775, + "step": 1699 + }, + { + "epoch": 0.3027065527065527, + "grad_norm": 0.6540699005126953, + "learning_rate": 0.0001972477648942509, + "loss": 1.181, + "step": 1700 + }, + { + "epoch": 0.30288461538461536, + "grad_norm": 0.46281275153160095, + "learning_rate": 0.00019724450258396008, + "loss": 0.629, + "step": 1701 + }, + { + "epoch": 0.30306267806267806, + "grad_norm": 0.3452845811843872, + "learning_rate": 0.00019724123836836527, + "loss": 0.51, + "step": 1702 + }, + { + "epoch": 0.30324074074074076, + "grad_norm": 0.4507991671562195, + "learning_rate": 0.00019723797224753038, + "loss": 1.0258, + "step": 1703 + }, + { + "epoch": 0.3034188034188034, + "grad_norm": 0.5385412573814392, + "learning_rate": 0.0001972347042215194, + "loss": 1.0232, + "step": 1704 + }, + { + "epoch": 0.3035968660968661, + "grad_norm": 0.4460466504096985, + "learning_rate": 0.00019723143429039642, + "loss": 1.1307, + "step": 1705 + }, + { + "epoch": 0.3037749287749288, + "grad_norm": 0.5229718685150146, + "learning_rate": 0.00019722816245422545, + "loss": 1.0964, + "step": 1706 + }, + { + "epoch": 0.30395299145299143, + "grad_norm": 0.4776979088783264, + "learning_rate": 0.00019722488871307058, + "loss": 1.2678, + "step": 1707 + }, + { + "epoch": 0.3041310541310541, + "grad_norm": 0.5371831655502319, + "learning_rate": 0.00019722161306699601, + "loss": 1.2808, + "step": 1708 + }, + { + "epoch": 0.3043091168091168, + "grad_norm": 0.45322108268737793, + "learning_rate": 0.0001972183355160659, + "loss": 1.0775, + "step": 1709 + }, + { + "epoch": 0.30448717948717946, + "grad_norm": 0.5036569833755493, + "learning_rate": 0.00019721505606034448, + "loss": 1.1859, + "step": 1710 + }, + { + "epoch": 0.30466524216524216, + "grad_norm": 0.5425969958305359, + "learning_rate": 0.00019721177469989593, + "loss": 1.0173, + "step": 1711 + }, + { + "epoch": 0.30484330484330485, + "grad_norm": 0.5638980269432068, + "learning_rate": 0.00019720849143478462, + "loss": 1.182, + "step": 1712 + }, + { + "epoch": 0.30502136752136755, + "grad_norm": 0.5160546898841858, + "learning_rate": 0.00019720520626507486, + "loss": 0.9853, + "step": 1713 + }, + { + "epoch": 0.3051994301994302, + "grad_norm": 0.5079004168510437, + "learning_rate": 0.000197201919190831, + "loss": 1.3154, + "step": 1714 + }, + { + "epoch": 0.3053774928774929, + "grad_norm": 0.4590355455875397, + "learning_rate": 0.00019719863021211745, + "loss": 1.007, + "step": 1715 + }, + { + "epoch": 0.3055555555555556, + "grad_norm": 0.49656423926353455, + "learning_rate": 0.00019719533932899865, + "loss": 1.2187, + "step": 1716 + }, + { + "epoch": 0.3057336182336182, + "grad_norm": 0.46426209807395935, + "learning_rate": 0.0001971920465415391, + "loss": 1.3007, + "step": 1717 + }, + { + "epoch": 0.3059116809116809, + "grad_norm": 0.5211917757987976, + "learning_rate": 0.00019718875184980328, + "loss": 1.2256, + "step": 1718 + }, + { + "epoch": 0.3060897435897436, + "grad_norm": 0.42953309416770935, + "learning_rate": 0.00019718545525385578, + "loss": 1.2838, + "step": 1719 + }, + { + "epoch": 0.30626780626780625, + "grad_norm": 0.4893105924129486, + "learning_rate": 0.00019718215675376116, + "loss": 1.052, + "step": 1720 + }, + { + "epoch": 0.30644586894586895, + "grad_norm": 0.4833602011203766, + "learning_rate": 0.00019717885634958405, + "loss": 1.069, + "step": 1721 + }, + { + "epoch": 0.30662393162393164, + "grad_norm": 0.502176821231842, + "learning_rate": 0.0001971755540413891, + "loss": 1.1659, + "step": 1722 + }, + { + "epoch": 0.3068019943019943, + "grad_norm": 0.4648856818675995, + "learning_rate": 0.00019717224982924108, + "loss": 1.1873, + "step": 1723 + }, + { + "epoch": 0.306980056980057, + "grad_norm": 0.405429869890213, + "learning_rate": 0.00019716894371320465, + "loss": 0.99, + "step": 1724 + }, + { + "epoch": 0.3071581196581197, + "grad_norm": 0.4306945204734802, + "learning_rate": 0.00019716563569334463, + "loss": 0.8751, + "step": 1725 + }, + { + "epoch": 0.3073361823361823, + "grad_norm": 0.49424824118614197, + "learning_rate": 0.00019716232576972583, + "loss": 0.9205, + "step": 1726 + }, + { + "epoch": 0.307514245014245, + "grad_norm": 0.5044034123420715, + "learning_rate": 0.00019715901394241306, + "loss": 1.2042, + "step": 1727 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 0.512180507183075, + "learning_rate": 0.00019715570021147126, + "loss": 1.1644, + "step": 1728 + }, + { + "epoch": 0.30787037037037035, + "grad_norm": 0.4377981126308441, + "learning_rate": 0.00019715238457696538, + "loss": 1.1625, + "step": 1729 + }, + { + "epoch": 0.30804843304843305, + "grad_norm": 0.49107855558395386, + "learning_rate": 0.00019714906703896027, + "loss": 1.1037, + "step": 1730 + }, + { + "epoch": 0.30822649572649574, + "grad_norm": 0.47342559695243835, + "learning_rate": 0.00019714574759752105, + "loss": 1.3186, + "step": 1731 + }, + { + "epoch": 0.3084045584045584, + "grad_norm": 0.487177312374115, + "learning_rate": 0.0001971424262527127, + "loss": 1.1196, + "step": 1732 + }, + { + "epoch": 0.3085826210826211, + "grad_norm": 0.5290025472640991, + "learning_rate": 0.0001971391030046003, + "loss": 1.2103, + "step": 1733 + }, + { + "epoch": 0.3087606837606838, + "grad_norm": 0.4587760269641876, + "learning_rate": 0.00019713577785324896, + "loss": 1.1017, + "step": 1734 + }, + { + "epoch": 0.3089387464387464, + "grad_norm": 0.45323294401168823, + "learning_rate": 0.00019713245079872388, + "loss": 1.0, + "step": 1735 + }, + { + "epoch": 0.3091168091168091, + "grad_norm": 0.43414804339408875, + "learning_rate": 0.00019712912184109013, + "loss": 1.0341, + "step": 1736 + }, + { + "epoch": 0.3092948717948718, + "grad_norm": 0.49604663252830505, + "learning_rate": 0.00019712579098041304, + "loss": 0.9437, + "step": 1737 + }, + { + "epoch": 0.30947293447293445, + "grad_norm": 0.48580703139305115, + "learning_rate": 0.00019712245821675785, + "loss": 1.2622, + "step": 1738 + }, + { + "epoch": 0.30965099715099714, + "grad_norm": 0.45333603024482727, + "learning_rate": 0.00019711912355018982, + "loss": 1.2063, + "step": 1739 + }, + { + "epoch": 0.30982905982905984, + "grad_norm": 0.5990764498710632, + "learning_rate": 0.00019711578698077432, + "loss": 1.5097, + "step": 1740 + }, + { + "epoch": 0.31000712250712253, + "grad_norm": 0.4386102259159088, + "learning_rate": 0.0001971124485085767, + "loss": 1.1283, + "step": 1741 + }, + { + "epoch": 0.3101851851851852, + "grad_norm": 0.4476035237312317, + "learning_rate": 0.00019710910813366242, + "loss": 0.8922, + "step": 1742 + }, + { + "epoch": 0.31036324786324787, + "grad_norm": 0.5276228785514832, + "learning_rate": 0.00019710576585609685, + "loss": 1.2373, + "step": 1743 + }, + { + "epoch": 0.31054131054131057, + "grad_norm": 0.4885637164115906, + "learning_rate": 0.00019710242167594557, + "loss": 1.0881, + "step": 1744 + }, + { + "epoch": 0.3107193732193732, + "grad_norm": 0.421132355928421, + "learning_rate": 0.000197099075593274, + "loss": 1.0544, + "step": 1745 + }, + { + "epoch": 0.3108974358974359, + "grad_norm": 0.5257927179336548, + "learning_rate": 0.00019709572760814777, + "loss": 1.265, + "step": 1746 + }, + { + "epoch": 0.3110754985754986, + "grad_norm": 0.5164850950241089, + "learning_rate": 0.00019709237772063247, + "loss": 0.9593, + "step": 1747 + }, + { + "epoch": 0.31125356125356124, + "grad_norm": 0.5176383256912231, + "learning_rate": 0.00019708902593079374, + "loss": 1.0194, + "step": 1748 + }, + { + "epoch": 0.31143162393162394, + "grad_norm": 0.4620790481567383, + "learning_rate": 0.00019708567223869716, + "loss": 0.9241, + "step": 1749 + }, + { + "epoch": 0.31160968660968663, + "grad_norm": 0.48307979106903076, + "learning_rate": 0.00019708231664440854, + "loss": 1.2314, + "step": 1750 + }, + { + "epoch": 0.31178774928774927, + "grad_norm": 0.4931468069553375, + "learning_rate": 0.00019707895914799364, + "loss": 1.2065, + "step": 1751 + }, + { + "epoch": 0.31196581196581197, + "grad_norm": 0.5035979747772217, + "learning_rate": 0.00019707559974951818, + "loss": 1.1867, + "step": 1752 + }, + { + "epoch": 0.31214387464387466, + "grad_norm": 0.47543632984161377, + "learning_rate": 0.00019707223844904795, + "loss": 1.0603, + "step": 1753 + }, + { + "epoch": 0.3123219373219373, + "grad_norm": 0.49929797649383545, + "learning_rate": 0.00019706887524664892, + "loss": 1.0597, + "step": 1754 + }, + { + "epoch": 0.3125, + "grad_norm": 0.5075222253799438, + "learning_rate": 0.00019706551014238687, + "loss": 1.1398, + "step": 1755 + }, + { + "epoch": 0.3126780626780627, + "grad_norm": 0.5096884369850159, + "learning_rate": 0.00019706214313632784, + "loss": 1.1382, + "step": 1756 + }, + { + "epoch": 0.31285612535612534, + "grad_norm": 0.4629988372325897, + "learning_rate": 0.0001970587742285377, + "loss": 1.0009, + "step": 1757 + }, + { + "epoch": 0.31303418803418803, + "grad_norm": 0.5244084596633911, + "learning_rate": 0.00019705540341908253, + "loss": 1.047, + "step": 1758 + }, + { + "epoch": 0.31321225071225073, + "grad_norm": 0.5136716961860657, + "learning_rate": 0.00019705203070802832, + "loss": 1.29, + "step": 1759 + }, + { + "epoch": 0.31339031339031337, + "grad_norm": 0.43991541862487793, + "learning_rate": 0.0001970486560954412, + "loss": 0.9605, + "step": 1760 + }, + { + "epoch": 0.31356837606837606, + "grad_norm": 0.4633477032184601, + "learning_rate": 0.00019704527958138725, + "loss": 1.1507, + "step": 1761 + }, + { + "epoch": 0.31374643874643876, + "grad_norm": 0.4419999420642853, + "learning_rate": 0.00019704190116593266, + "loss": 0.9262, + "step": 1762 + }, + { + "epoch": 0.3139245014245014, + "grad_norm": 0.49359434843063354, + "learning_rate": 0.00019703852084914357, + "loss": 0.9348, + "step": 1763 + }, + { + "epoch": 0.3141025641025641, + "grad_norm": 0.5072139501571655, + "learning_rate": 0.00019703513863108627, + "loss": 1.1592, + "step": 1764 + }, + { + "epoch": 0.3142806267806268, + "grad_norm": 0.45969831943511963, + "learning_rate": 0.00019703175451182698, + "loss": 1.1519, + "step": 1765 + }, + { + "epoch": 0.31445868945868943, + "grad_norm": 0.5148758292198181, + "learning_rate": 0.00019702836849143208, + "loss": 1.1673, + "step": 1766 + }, + { + "epoch": 0.31463675213675213, + "grad_norm": 0.43033209443092346, + "learning_rate": 0.0001970249805699678, + "loss": 0.9256, + "step": 1767 + }, + { + "epoch": 0.3148148148148148, + "grad_norm": 0.48143425583839417, + "learning_rate": 0.00019702159074750058, + "loss": 1.08, + "step": 1768 + }, + { + "epoch": 0.31499287749287747, + "grad_norm": 0.4780619740486145, + "learning_rate": 0.00019701819902409685, + "loss": 1.1198, + "step": 1769 + }, + { + "epoch": 0.31517094017094016, + "grad_norm": 0.4662075936794281, + "learning_rate": 0.00019701480539982305, + "loss": 0.8424, + "step": 1770 + }, + { + "epoch": 0.31534900284900286, + "grad_norm": 0.503901481628418, + "learning_rate": 0.00019701140987474566, + "loss": 1.1026, + "step": 1771 + }, + { + "epoch": 0.31552706552706555, + "grad_norm": 0.5197132229804993, + "learning_rate": 0.00019700801244893124, + "loss": 1.2148, + "step": 1772 + }, + { + "epoch": 0.3157051282051282, + "grad_norm": 0.4746309220790863, + "learning_rate": 0.00019700461312244634, + "loss": 1.0906, + "step": 1773 + }, + { + "epoch": 0.3158831908831909, + "grad_norm": 0.5277339816093445, + "learning_rate": 0.00019700121189535752, + "loss": 1.0588, + "step": 1774 + }, + { + "epoch": 0.3160612535612536, + "grad_norm": 0.436002254486084, + "learning_rate": 0.00019699780876773147, + "loss": 1.0341, + "step": 1775 + }, + { + "epoch": 0.3162393162393162, + "grad_norm": 0.5171145796775818, + "learning_rate": 0.00019699440373963486, + "loss": 1.282, + "step": 1776 + }, + { + "epoch": 0.3164173789173789, + "grad_norm": 0.38382846117019653, + "learning_rate": 0.00019699099681113436, + "loss": 0.8908, + "step": 1777 + }, + { + "epoch": 0.3165954415954416, + "grad_norm": 0.4621630609035492, + "learning_rate": 0.0001969875879822968, + "loss": 1.1074, + "step": 1778 + }, + { + "epoch": 0.31677350427350426, + "grad_norm": 0.5543130040168762, + "learning_rate": 0.00019698417725318892, + "loss": 0.9682, + "step": 1779 + }, + { + "epoch": 0.31695156695156695, + "grad_norm": 0.49534836411476135, + "learning_rate": 0.00019698076462387753, + "loss": 1.107, + "step": 1780 + }, + { + "epoch": 0.31712962962962965, + "grad_norm": 0.48844948410987854, + "learning_rate": 0.00019697735009442956, + "loss": 1.1295, + "step": 1781 + }, + { + "epoch": 0.3173076923076923, + "grad_norm": 0.5070686936378479, + "learning_rate": 0.00019697393366491185, + "loss": 1.083, + "step": 1782 + }, + { + "epoch": 0.317485754985755, + "grad_norm": 0.47817620635032654, + "learning_rate": 0.00019697051533539134, + "loss": 1.3014, + "step": 1783 + }, + { + "epoch": 0.3176638176638177, + "grad_norm": 0.538488507270813, + "learning_rate": 0.00019696709510593502, + "loss": 1.0354, + "step": 1784 + }, + { + "epoch": 0.3178418803418803, + "grad_norm": 0.5141439437866211, + "learning_rate": 0.0001969636729766099, + "loss": 1.2912, + "step": 1785 + }, + { + "epoch": 0.318019943019943, + "grad_norm": 0.5009665489196777, + "learning_rate": 0.00019696024894748306, + "loss": 0.9014, + "step": 1786 + }, + { + "epoch": 0.3181980056980057, + "grad_norm": 0.46199744939804077, + "learning_rate": 0.00019695682301862155, + "loss": 1.0532, + "step": 1787 + }, + { + "epoch": 0.31837606837606836, + "grad_norm": 0.4649423062801361, + "learning_rate": 0.0001969533951900925, + "loss": 0.8608, + "step": 1788 + }, + { + "epoch": 0.31855413105413105, + "grad_norm": 0.516909658908844, + "learning_rate": 0.0001969499654619631, + "loss": 1.1385, + "step": 1789 + }, + { + "epoch": 0.31873219373219375, + "grad_norm": 0.46016669273376465, + "learning_rate": 0.00019694653383430048, + "loss": 0.9168, + "step": 1790 + }, + { + "epoch": 0.3189102564102564, + "grad_norm": 0.4794938564300537, + "learning_rate": 0.00019694310030717193, + "loss": 1.0244, + "step": 1791 + }, + { + "epoch": 0.3190883190883191, + "grad_norm": 0.46577662229537964, + "learning_rate": 0.00019693966488064471, + "loss": 1.0954, + "step": 1792 + }, + { + "epoch": 0.3192663817663818, + "grad_norm": 0.4866746962070465, + "learning_rate": 0.00019693622755478614, + "loss": 1.2925, + "step": 1793 + }, + { + "epoch": 0.3194444444444444, + "grad_norm": 0.4841702878475189, + "learning_rate": 0.00019693278832966357, + "loss": 1.119, + "step": 1794 + }, + { + "epoch": 0.3196225071225071, + "grad_norm": 0.4835243821144104, + "learning_rate": 0.00019692934720534435, + "loss": 1.1702, + "step": 1795 + }, + { + "epoch": 0.3198005698005698, + "grad_norm": 0.5200608968734741, + "learning_rate": 0.00019692590418189594, + "loss": 1.1989, + "step": 1796 + }, + { + "epoch": 0.31997863247863245, + "grad_norm": 0.5147821307182312, + "learning_rate": 0.00019692245925938577, + "loss": 1.1417, + "step": 1797 + }, + { + "epoch": 0.32015669515669515, + "grad_norm": 0.5145614743232727, + "learning_rate": 0.00019691901243788136, + "loss": 1.0571, + "step": 1798 + }, + { + "epoch": 0.32033475783475784, + "grad_norm": 0.5416026711463928, + "learning_rate": 0.00019691556371745022, + "loss": 1.188, + "step": 1799 + }, + { + "epoch": 0.32051282051282054, + "grad_norm": 0.5140644311904907, + "learning_rate": 0.00019691211309815995, + "loss": 1.1795, + "step": 1800 + }, + { + "epoch": 0.3206908831908832, + "grad_norm": 0.44219106435775757, + "learning_rate": 0.00019690866058007817, + "loss": 0.9215, + "step": 1801 + }, + { + "epoch": 0.3208689458689459, + "grad_norm": 0.49523603916168213, + "learning_rate": 0.00019690520616327245, + "loss": 1.1117, + "step": 1802 + }, + { + "epoch": 0.32104700854700857, + "grad_norm": 0.5818293690681458, + "learning_rate": 0.0001969017498478105, + "loss": 1.16, + "step": 1803 + }, + { + "epoch": 0.3212250712250712, + "grad_norm": 0.5175749659538269, + "learning_rate": 0.0001968982916337601, + "loss": 1.1999, + "step": 1804 + }, + { + "epoch": 0.3214031339031339, + "grad_norm": 0.49916017055511475, + "learning_rate": 0.00019689483152118898, + "loss": 0.9505, + "step": 1805 + }, + { + "epoch": 0.3215811965811966, + "grad_norm": 0.46849536895751953, + "learning_rate": 0.00019689136951016488, + "loss": 0.9627, + "step": 1806 + }, + { + "epoch": 0.32175925925925924, + "grad_norm": 0.4226818382740021, + "learning_rate": 0.00019688790560075568, + "loss": 1.037, + "step": 1807 + }, + { + "epoch": 0.32193732193732194, + "grad_norm": 0.4697103798389435, + "learning_rate": 0.00019688443979302923, + "loss": 1.1431, + "step": 1808 + }, + { + "epoch": 0.32211538461538464, + "grad_norm": 0.4999365508556366, + "learning_rate": 0.00019688097208705343, + "loss": 1.171, + "step": 1809 + }, + { + "epoch": 0.3222934472934473, + "grad_norm": 0.5229731798171997, + "learning_rate": 0.00019687750248289625, + "loss": 1.3395, + "step": 1810 + }, + { + "epoch": 0.32247150997151, + "grad_norm": 0.512525737285614, + "learning_rate": 0.00019687403098062566, + "loss": 1.1438, + "step": 1811 + }, + { + "epoch": 0.32264957264957267, + "grad_norm": 0.4558548927307129, + "learning_rate": 0.00019687055758030967, + "loss": 1.0012, + "step": 1812 + }, + { + "epoch": 0.3228276353276353, + "grad_norm": 0.45195743441581726, + "learning_rate": 0.00019686708228201636, + "loss": 1.0222, + "step": 1813 + }, + { + "epoch": 0.323005698005698, + "grad_norm": 0.5023126602172852, + "learning_rate": 0.00019686360508581373, + "loss": 1.2128, + "step": 1814 + }, + { + "epoch": 0.3231837606837607, + "grad_norm": 0.46516045928001404, + "learning_rate": 0.00019686012599177003, + "loss": 0.989, + "step": 1815 + }, + { + "epoch": 0.32336182336182334, + "grad_norm": 0.4142672121524811, + "learning_rate": 0.00019685664499995338, + "loss": 1.0144, + "step": 1816 + }, + { + "epoch": 0.32353988603988604, + "grad_norm": 0.4511009752750397, + "learning_rate": 0.0001968531621104319, + "loss": 0.885, + "step": 1817 + }, + { + "epoch": 0.32371794871794873, + "grad_norm": 0.49583545327186584, + "learning_rate": 0.00019684967732327396, + "loss": 1.0986, + "step": 1818 + }, + { + "epoch": 0.3238960113960114, + "grad_norm": 0.5872161388397217, + "learning_rate": 0.0001968461906385478, + "loss": 1.1482, + "step": 1819 + }, + { + "epoch": 0.32407407407407407, + "grad_norm": 0.4509563148021698, + "learning_rate": 0.00019684270205632168, + "loss": 1.0578, + "step": 1820 + }, + { + "epoch": 0.32425213675213677, + "grad_norm": 0.501345157623291, + "learning_rate": 0.00019683921157666402, + "loss": 1.1792, + "step": 1821 + }, + { + "epoch": 0.3244301994301994, + "grad_norm": 0.48257577419281006, + "learning_rate": 0.00019683571919964314, + "loss": 1.0448, + "step": 1822 + }, + { + "epoch": 0.3246082621082621, + "grad_norm": 0.5399422645568848, + "learning_rate": 0.00019683222492532752, + "loss": 1.0579, + "step": 1823 + }, + { + "epoch": 0.3247863247863248, + "grad_norm": 0.4382506012916565, + "learning_rate": 0.0001968287287537856, + "loss": 1.0246, + "step": 1824 + }, + { + "epoch": 0.32496438746438744, + "grad_norm": 0.49247491359710693, + "learning_rate": 0.00019682523068508586, + "loss": 1.318, + "step": 1825 + }, + { + "epoch": 0.32514245014245013, + "grad_norm": 0.49067625403404236, + "learning_rate": 0.0001968217307192969, + "loss": 1.1028, + "step": 1826 + }, + { + "epoch": 0.32532051282051283, + "grad_norm": 0.4832286238670349, + "learning_rate": 0.00019681822885648723, + "loss": 1.0996, + "step": 1827 + }, + { + "epoch": 0.32549857549857547, + "grad_norm": 0.47144386172294617, + "learning_rate": 0.0001968147250967255, + "loss": 1.0707, + "step": 1828 + }, + { + "epoch": 0.32567663817663817, + "grad_norm": 0.46299225091934204, + "learning_rate": 0.0001968112194400803, + "loss": 1.0461, + "step": 1829 + }, + { + "epoch": 0.32585470085470086, + "grad_norm": 0.4880816340446472, + "learning_rate": 0.00019680771188662044, + "loss": 1.1198, + "step": 1830 + }, + { + "epoch": 0.32603276353276356, + "grad_norm": 0.43837276101112366, + "learning_rate": 0.00019680420243641452, + "loss": 1.0599, + "step": 1831 + }, + { + "epoch": 0.3262108262108262, + "grad_norm": 0.453168660402298, + "learning_rate": 0.0001968006910895314, + "loss": 1.0327, + "step": 1832 + }, + { + "epoch": 0.3263888888888889, + "grad_norm": 0.45183828473091125, + "learning_rate": 0.00019679717784603975, + "loss": 1.1381, + "step": 1833 + }, + { + "epoch": 0.3265669515669516, + "grad_norm": 0.5326765775680542, + "learning_rate": 0.00019679366270600852, + "loss": 1.3169, + "step": 1834 + }, + { + "epoch": 0.32674501424501423, + "grad_norm": 0.47468429803848267, + "learning_rate": 0.00019679014566950653, + "loss": 1.1816, + "step": 1835 + }, + { + "epoch": 0.3269230769230769, + "grad_norm": 0.5096879005432129, + "learning_rate": 0.0001967866267366027, + "loss": 1.1162, + "step": 1836 + }, + { + "epoch": 0.3271011396011396, + "grad_norm": 0.491514652967453, + "learning_rate": 0.00019678310590736598, + "loss": 1.2793, + "step": 1837 + }, + { + "epoch": 0.32727920227920226, + "grad_norm": 0.601439356803894, + "learning_rate": 0.00019677958318186533, + "loss": 0.9851, + "step": 1838 + }, + { + "epoch": 0.32745726495726496, + "grad_norm": 0.45270970463752747, + "learning_rate": 0.0001967760585601698, + "loss": 1.0042, + "step": 1839 + }, + { + "epoch": 0.32763532763532766, + "grad_norm": 0.48864325881004333, + "learning_rate": 0.00019677253204234847, + "loss": 1.0835, + "step": 1840 + }, + { + "epoch": 0.3278133903133903, + "grad_norm": 0.5855685472488403, + "learning_rate": 0.00019676900362847037, + "loss": 1.193, + "step": 1841 + }, + { + "epoch": 0.327991452991453, + "grad_norm": 0.7181013822555542, + "learning_rate": 0.00019676547331860466, + "loss": 1.2028, + "step": 1842 + }, + { + "epoch": 0.3281695156695157, + "grad_norm": 0.4517378807067871, + "learning_rate": 0.00019676194111282054, + "loss": 1.013, + "step": 1843 + }, + { + "epoch": 0.32834757834757833, + "grad_norm": 0.5477756857872009, + "learning_rate": 0.00019675840701118718, + "loss": 1.2311, + "step": 1844 + }, + { + "epoch": 0.328525641025641, + "grad_norm": 0.5194997191429138, + "learning_rate": 0.00019675487101377382, + "loss": 1.0953, + "step": 1845 + }, + { + "epoch": 0.3287037037037037, + "grad_norm": 0.44454067945480347, + "learning_rate": 0.00019675133312064977, + "loss": 0.8505, + "step": 1846 + }, + { + "epoch": 0.32888176638176636, + "grad_norm": 0.3938713073730469, + "learning_rate": 0.00019674779333188428, + "loss": 0.8525, + "step": 1847 + }, + { + "epoch": 0.32905982905982906, + "grad_norm": 0.4927884340286255, + "learning_rate": 0.00019674425164754682, + "loss": 1.2477, + "step": 1848 + }, + { + "epoch": 0.32923789173789175, + "grad_norm": 0.4516635239124298, + "learning_rate": 0.0001967407080677067, + "loss": 0.8333, + "step": 1849 + }, + { + "epoch": 0.3294159544159544, + "grad_norm": 0.47105780243873596, + "learning_rate": 0.00019673716259243336, + "loss": 1.0989, + "step": 1850 + }, + { + "epoch": 0.3295940170940171, + "grad_norm": 0.5192127823829651, + "learning_rate": 0.00019673361522179627, + "loss": 1.1164, + "step": 1851 + }, + { + "epoch": 0.3297720797720798, + "grad_norm": 0.5222696661949158, + "learning_rate": 0.00019673006595586495, + "loss": 1.3191, + "step": 1852 + }, + { + "epoch": 0.3299501424501424, + "grad_norm": 0.6046679019927979, + "learning_rate": 0.0001967265147947089, + "loss": 0.9782, + "step": 1853 + }, + { + "epoch": 0.3301282051282051, + "grad_norm": 0.47928622364997864, + "learning_rate": 0.00019672296173839775, + "loss": 1.2247, + "step": 1854 + }, + { + "epoch": 0.3303062678062678, + "grad_norm": 0.5435982346534729, + "learning_rate": 0.00019671940678700107, + "loss": 1.1647, + "step": 1855 + }, + { + "epoch": 0.33048433048433046, + "grad_norm": 0.46878984570503235, + "learning_rate": 0.00019671584994058856, + "loss": 1.132, + "step": 1856 + }, + { + "epoch": 0.33066239316239315, + "grad_norm": 0.5336877107620239, + "learning_rate": 0.00019671229119922986, + "loss": 1.0583, + "step": 1857 + }, + { + "epoch": 0.33084045584045585, + "grad_norm": 0.4811093807220459, + "learning_rate": 0.0001967087305629947, + "loss": 1.0089, + "step": 1858 + }, + { + "epoch": 0.33101851851851855, + "grad_norm": 0.5140184760093689, + "learning_rate": 0.0001967051680319529, + "loss": 1.2335, + "step": 1859 + }, + { + "epoch": 0.3311965811965812, + "grad_norm": 0.5855883955955505, + "learning_rate": 0.00019670160360617418, + "loss": 1.1107, + "step": 1860 + }, + { + "epoch": 0.3313746438746439, + "grad_norm": 0.5081531405448914, + "learning_rate": 0.00019669803728572844, + "loss": 1.0669, + "step": 1861 + }, + { + "epoch": 0.3315527065527066, + "grad_norm": 0.48749417066574097, + "learning_rate": 0.0001966944690706855, + "loss": 1.1465, + "step": 1862 + }, + { + "epoch": 0.3317307692307692, + "grad_norm": 0.5175687670707703, + "learning_rate": 0.00019669089896111536, + "loss": 1.254, + "step": 1863 + }, + { + "epoch": 0.3319088319088319, + "grad_norm": 0.4198860824108124, + "learning_rate": 0.0001966873269570879, + "loss": 0.9811, + "step": 1864 + }, + { + "epoch": 0.3320868945868946, + "grad_norm": 0.5220273733139038, + "learning_rate": 0.0001966837530586731, + "loss": 1.277, + "step": 1865 + }, + { + "epoch": 0.33226495726495725, + "grad_norm": 0.551954448223114, + "learning_rate": 0.00019668017726594101, + "loss": 1.0627, + "step": 1866 + }, + { + "epoch": 0.33244301994301995, + "grad_norm": 0.5289301872253418, + "learning_rate": 0.00019667659957896166, + "loss": 1.4525, + "step": 1867 + }, + { + "epoch": 0.33262108262108264, + "grad_norm": 0.5190161466598511, + "learning_rate": 0.00019667301999780522, + "loss": 1.1064, + "step": 1868 + }, + { + "epoch": 0.3327991452991453, + "grad_norm": 0.437637060880661, + "learning_rate": 0.00019666943852254172, + "loss": 1.1304, + "step": 1869 + }, + { + "epoch": 0.332977207977208, + "grad_norm": 0.4801286458969116, + "learning_rate": 0.00019666585515324138, + "loss": 1.032, + "step": 1870 + }, + { + "epoch": 0.3331552706552707, + "grad_norm": 0.5041908621788025, + "learning_rate": 0.00019666226988997445, + "loss": 1.2611, + "step": 1871 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.4529375731945038, + "learning_rate": 0.00019665868273281115, + "loss": 1.1346, + "step": 1872 + }, + { + "epoch": 0.333511396011396, + "grad_norm": 0.4797019064426422, + "learning_rate": 0.00019665509368182172, + "loss": 1.1716, + "step": 1873 + }, + { + "epoch": 0.3336894586894587, + "grad_norm": 0.5505055785179138, + "learning_rate": 0.00019665150273707652, + "loss": 0.9729, + "step": 1874 + }, + { + "epoch": 0.33386752136752135, + "grad_norm": 0.4228051006793976, + "learning_rate": 0.00019664790989864592, + "loss": 0.9023, + "step": 1875 + }, + { + "epoch": 0.33404558404558404, + "grad_norm": 0.4926959276199341, + "learning_rate": 0.00019664431516660028, + "loss": 1.0999, + "step": 1876 + }, + { + "epoch": 0.33422364672364674, + "grad_norm": 0.4273219704627991, + "learning_rate": 0.00019664071854101005, + "loss": 1.1039, + "step": 1877 + }, + { + "epoch": 0.3344017094017094, + "grad_norm": 0.48438936471939087, + "learning_rate": 0.00019663712002194566, + "loss": 1.1308, + "step": 1878 + }, + { + "epoch": 0.3345797720797721, + "grad_norm": 0.5102053284645081, + "learning_rate": 0.0001966335196094777, + "loss": 1.0618, + "step": 1879 + }, + { + "epoch": 0.33475783475783477, + "grad_norm": 0.4357300400733948, + "learning_rate": 0.00019662991730367663, + "loss": 1.0521, + "step": 1880 + }, + { + "epoch": 0.3349358974358974, + "grad_norm": 0.5052695870399475, + "learning_rate": 0.00019662631310461308, + "loss": 0.9579, + "step": 1881 + }, + { + "epoch": 0.3351139601139601, + "grad_norm": 0.4889117181301117, + "learning_rate": 0.00019662270701235762, + "loss": 1.0304, + "step": 1882 + }, + { + "epoch": 0.3352920227920228, + "grad_norm": 0.4671195149421692, + "learning_rate": 0.000196619099026981, + "loss": 1.2228, + "step": 1883 + }, + { + "epoch": 0.33547008547008544, + "grad_norm": 0.4700174331665039, + "learning_rate": 0.0001966154891485538, + "loss": 0.9634, + "step": 1884 + }, + { + "epoch": 0.33564814814814814, + "grad_norm": 0.488817423582077, + "learning_rate": 0.00019661187737714676, + "loss": 1.2499, + "step": 1885 + }, + { + "epoch": 0.33582621082621084, + "grad_norm": 0.5336169600486755, + "learning_rate": 0.00019660826371283073, + "loss": 1.251, + "step": 1886 + }, + { + "epoch": 0.33600427350427353, + "grad_norm": 0.5054540038108826, + "learning_rate": 0.00019660464815567642, + "loss": 1.221, + "step": 1887 + }, + { + "epoch": 0.33618233618233617, + "grad_norm": 0.5078747868537903, + "learning_rate": 0.00019660103070575472, + "loss": 0.9792, + "step": 1888 + }, + { + "epoch": 0.33636039886039887, + "grad_norm": 0.498571515083313, + "learning_rate": 0.0001965974113631365, + "loss": 1.1682, + "step": 1889 + }, + { + "epoch": 0.33653846153846156, + "grad_norm": 0.49969518184661865, + "learning_rate": 0.00019659379012789264, + "loss": 1.0012, + "step": 1890 + }, + { + "epoch": 0.3367165242165242, + "grad_norm": 0.4238094687461853, + "learning_rate": 0.00019659016700009416, + "loss": 1.0455, + "step": 1891 + }, + { + "epoch": 0.3368945868945869, + "grad_norm": 0.5139104723930359, + "learning_rate": 0.000196586541979812, + "loss": 0.9979, + "step": 1892 + }, + { + "epoch": 0.3370726495726496, + "grad_norm": 0.5446547269821167, + "learning_rate": 0.00019658291506711715, + "loss": 0.9271, + "step": 1893 + }, + { + "epoch": 0.33725071225071224, + "grad_norm": 0.5284572839736938, + "learning_rate": 0.00019657928626208077, + "loss": 1.0356, + "step": 1894 + }, + { + "epoch": 0.33742877492877493, + "grad_norm": 0.49936217069625854, + "learning_rate": 0.00019657565556477387, + "loss": 0.9785, + "step": 1895 + }, + { + "epoch": 0.33760683760683763, + "grad_norm": 0.4678729772567749, + "learning_rate": 0.00019657202297526763, + "loss": 1.2135, + "step": 1896 + }, + { + "epoch": 0.33778490028490027, + "grad_norm": 0.46844249963760376, + "learning_rate": 0.0001965683884936332, + "loss": 0.9369, + "step": 1897 + }, + { + "epoch": 0.33796296296296297, + "grad_norm": 0.4307389557361603, + "learning_rate": 0.0001965647521199418, + "loss": 0.9301, + "step": 1898 + }, + { + "epoch": 0.33814102564102566, + "grad_norm": 0.48227834701538086, + "learning_rate": 0.00019656111385426468, + "loss": 1.3169, + "step": 1899 + }, + { + "epoch": 0.3383190883190883, + "grad_norm": 0.45860713720321655, + "learning_rate": 0.00019655747369667315, + "loss": 0.9835, + "step": 1900 + }, + { + "epoch": 0.338497150997151, + "grad_norm": 0.5522414445877075, + "learning_rate": 0.00019655383164723846, + "loss": 1.363, + "step": 1901 + }, + { + "epoch": 0.3386752136752137, + "grad_norm": 0.5283710360527039, + "learning_rate": 0.000196550187706032, + "loss": 1.1499, + "step": 1902 + }, + { + "epoch": 0.33885327635327633, + "grad_norm": 0.4419134259223938, + "learning_rate": 0.00019654654187312525, + "loss": 1.2039, + "step": 1903 + }, + { + "epoch": 0.33903133903133903, + "grad_norm": 0.49066096544265747, + "learning_rate": 0.00019654289414858952, + "loss": 0.9707, + "step": 1904 + }, + { + "epoch": 0.3392094017094017, + "grad_norm": 0.4619338810443878, + "learning_rate": 0.00019653924453249633, + "loss": 1.0849, + "step": 1905 + }, + { + "epoch": 0.33938746438746437, + "grad_norm": 0.5191119313240051, + "learning_rate": 0.0001965355930249172, + "loss": 1.1387, + "step": 1906 + }, + { + "epoch": 0.33956552706552706, + "grad_norm": 0.5245711207389832, + "learning_rate": 0.00019653193962592368, + "loss": 1.3435, + "step": 1907 + }, + { + "epoch": 0.33974358974358976, + "grad_norm": 0.49562904238700867, + "learning_rate": 0.0001965282843355873, + "loss": 1.2781, + "step": 1908 + }, + { + "epoch": 0.3399216524216524, + "grad_norm": 0.4661353826522827, + "learning_rate": 0.0001965246271539797, + "loss": 0.9317, + "step": 1909 + }, + { + "epoch": 0.3400997150997151, + "grad_norm": 0.4723222851753235, + "learning_rate": 0.00019652096808117254, + "loss": 1.0733, + "step": 1910 + }, + { + "epoch": 0.3402777777777778, + "grad_norm": 0.4358505308628082, + "learning_rate": 0.00019651730711723754, + "loss": 1.1461, + "step": 1911 + }, + { + "epoch": 0.34045584045584043, + "grad_norm": 0.462422251701355, + "learning_rate": 0.00019651364426224638, + "loss": 1.0914, + "step": 1912 + }, + { + "epoch": 0.3406339031339031, + "grad_norm": 0.47952914237976074, + "learning_rate": 0.0001965099795162709, + "loss": 1.0392, + "step": 1913 + }, + { + "epoch": 0.3408119658119658, + "grad_norm": 0.5036373734474182, + "learning_rate": 0.00019650631287938282, + "loss": 1.4002, + "step": 1914 + }, + { + "epoch": 0.34099002849002846, + "grad_norm": 0.5130090713500977, + "learning_rate": 0.000196502644351654, + "loss": 1.3499, + "step": 1915 + }, + { + "epoch": 0.34116809116809116, + "grad_norm": 0.4426332414150238, + "learning_rate": 0.00019649897393315635, + "loss": 1.0726, + "step": 1916 + }, + { + "epoch": 0.34134615384615385, + "grad_norm": 0.5580727458000183, + "learning_rate": 0.00019649530162396176, + "loss": 1.1164, + "step": 1917 + }, + { + "epoch": 0.34152421652421655, + "grad_norm": 0.545001745223999, + "learning_rate": 0.00019649162742414218, + "loss": 0.962, + "step": 1918 + }, + { + "epoch": 0.3417022792022792, + "grad_norm": 0.5225808024406433, + "learning_rate": 0.00019648795133376962, + "loss": 1.1415, + "step": 1919 + }, + { + "epoch": 0.3418803418803419, + "grad_norm": 0.48210129141807556, + "learning_rate": 0.0001964842733529161, + "loss": 1.1188, + "step": 1920 + }, + { + "epoch": 0.3420584045584046, + "grad_norm": 0.4515395164489746, + "learning_rate": 0.00019648059348165365, + "loss": 1.0828, + "step": 1921 + }, + { + "epoch": 0.3422364672364672, + "grad_norm": 0.5802633166313171, + "learning_rate": 0.0001964769117200544, + "loss": 1.3137, + "step": 1922 + }, + { + "epoch": 0.3424145299145299, + "grad_norm": 0.4432032108306885, + "learning_rate": 0.00019647322806819046, + "loss": 1.0523, + "step": 1923 + }, + { + "epoch": 0.3425925925925926, + "grad_norm": 0.4697614908218384, + "learning_rate": 0.00019646954252613402, + "loss": 0.8426, + "step": 1924 + }, + { + "epoch": 0.34277065527065526, + "grad_norm": 0.4610968232154846, + "learning_rate": 0.0001964658550939573, + "loss": 0.9826, + "step": 1925 + }, + { + "epoch": 0.34294871794871795, + "grad_norm": 0.5278257727622986, + "learning_rate": 0.00019646216577173258, + "loss": 1.1064, + "step": 1926 + }, + { + "epoch": 0.34312678062678065, + "grad_norm": 0.5686144232749939, + "learning_rate": 0.00019645847455953205, + "loss": 0.9138, + "step": 1927 + }, + { + "epoch": 0.3433048433048433, + "grad_norm": 0.42894792556762695, + "learning_rate": 0.0001964547814574281, + "loss": 1.0461, + "step": 1928 + }, + { + "epoch": 0.343482905982906, + "grad_norm": 0.5567317605018616, + "learning_rate": 0.0001964510864654931, + "loss": 0.8787, + "step": 1929 + }, + { + "epoch": 0.3436609686609687, + "grad_norm": 0.5015586614608765, + "learning_rate": 0.0001964473895837994, + "loss": 1.1406, + "step": 1930 + }, + { + "epoch": 0.3438390313390313, + "grad_norm": 0.47391530871391296, + "learning_rate": 0.00019644369081241948, + "loss": 1.0685, + "step": 1931 + }, + { + "epoch": 0.344017094017094, + "grad_norm": 0.546037495136261, + "learning_rate": 0.00019643999015142574, + "loss": 1.2349, + "step": 1932 + }, + { + "epoch": 0.3441951566951567, + "grad_norm": 0.4724953770637512, + "learning_rate": 0.00019643628760089078, + "loss": 1.0621, + "step": 1933 + }, + { + "epoch": 0.34437321937321935, + "grad_norm": 0.5644593834877014, + "learning_rate": 0.00019643258316088703, + "loss": 1.2559, + "step": 1934 + }, + { + "epoch": 0.34455128205128205, + "grad_norm": 0.500815749168396, + "learning_rate": 0.00019642887683148718, + "loss": 1.0439, + "step": 1935 + }, + { + "epoch": 0.34472934472934474, + "grad_norm": 0.4932316541671753, + "learning_rate": 0.0001964251686127638, + "loss": 1.0404, + "step": 1936 + }, + { + "epoch": 0.3449074074074074, + "grad_norm": 0.48494651913642883, + "learning_rate": 0.00019642145850478954, + "loss": 0.9951, + "step": 1937 + }, + { + "epoch": 0.3450854700854701, + "grad_norm": 0.5191963315010071, + "learning_rate": 0.00019641774650763706, + "loss": 1.1258, + "step": 1938 + }, + { + "epoch": 0.3452635327635328, + "grad_norm": 0.4439312815666199, + "learning_rate": 0.00019641403262137918, + "loss": 1.1158, + "step": 1939 + }, + { + "epoch": 0.3454415954415954, + "grad_norm": 0.4829137921333313, + "learning_rate": 0.0001964103168460886, + "loss": 1.0531, + "step": 1940 + }, + { + "epoch": 0.3456196581196581, + "grad_norm": 0.49433329701423645, + "learning_rate": 0.00019640659918183811, + "loss": 1.1295, + "step": 1941 + }, + { + "epoch": 0.3457977207977208, + "grad_norm": 0.5351347923278809, + "learning_rate": 0.00019640287962870062, + "loss": 1.2379, + "step": 1942 + }, + { + "epoch": 0.34597578347578345, + "grad_norm": 0.4845680892467499, + "learning_rate": 0.00019639915818674895, + "loss": 1.0197, + "step": 1943 + }, + { + "epoch": 0.34615384615384615, + "grad_norm": 0.5312514901161194, + "learning_rate": 0.00019639543485605604, + "loss": 0.9734, + "step": 1944 + }, + { + "epoch": 0.34633190883190884, + "grad_norm": 0.4571874737739563, + "learning_rate": 0.00019639170963669478, + "loss": 1.1012, + "step": 1945 + }, + { + "epoch": 0.34650997150997154, + "grad_norm": 0.4449031949043274, + "learning_rate": 0.00019638798252873824, + "loss": 1.1393, + "step": 1946 + }, + { + "epoch": 0.3466880341880342, + "grad_norm": 0.47470834851264954, + "learning_rate": 0.0001963842535322594, + "loss": 0.981, + "step": 1947 + }, + { + "epoch": 0.3468660968660969, + "grad_norm": 0.5386981964111328, + "learning_rate": 0.00019638052264733132, + "loss": 1.1247, + "step": 1948 + }, + { + "epoch": 0.34704415954415957, + "grad_norm": 0.535589873790741, + "learning_rate": 0.00019637678987402714, + "loss": 1.3157, + "step": 1949 + }, + { + "epoch": 0.3472222222222222, + "grad_norm": 0.49338245391845703, + "learning_rate": 0.00019637305521242, + "loss": 1.1066, + "step": 1950 + }, + { + "epoch": 0.3474002849002849, + "grad_norm": 0.4247688353061676, + "learning_rate": 0.00019636931866258298, + "loss": 1.0039, + "step": 1951 + }, + { + "epoch": 0.3475783475783476, + "grad_norm": 0.5351517200469971, + "learning_rate": 0.00019636558022458934, + "loss": 1.0344, + "step": 1952 + }, + { + "epoch": 0.34775641025641024, + "grad_norm": 0.4633362889289856, + "learning_rate": 0.00019636183989851238, + "loss": 1.1383, + "step": 1953 + }, + { + "epoch": 0.34793447293447294, + "grad_norm": 0.553709089756012, + "learning_rate": 0.00019635809768442535, + "loss": 1.0389, + "step": 1954 + }, + { + "epoch": 0.34811253561253563, + "grad_norm": 0.479374498128891, + "learning_rate": 0.00019635435358240154, + "loss": 1.1774, + "step": 1955 + }, + { + "epoch": 0.3482905982905983, + "grad_norm": 0.5274081230163574, + "learning_rate": 0.0001963506075925143, + "loss": 1.1809, + "step": 1956 + }, + { + "epoch": 0.34846866096866097, + "grad_norm": 0.45398542284965515, + "learning_rate": 0.0001963468597148371, + "loss": 1.0502, + "step": 1957 + }, + { + "epoch": 0.34864672364672367, + "grad_norm": 0.48201611638069153, + "learning_rate": 0.00019634310994944332, + "loss": 1.0557, + "step": 1958 + }, + { + "epoch": 0.3488247863247863, + "grad_norm": 0.6407544016838074, + "learning_rate": 0.00019633935829640642, + "loss": 1.2138, + "step": 1959 + }, + { + "epoch": 0.349002849002849, + "grad_norm": 0.5385687351226807, + "learning_rate": 0.00019633560475579995, + "loss": 1.3496, + "step": 1960 + }, + { + "epoch": 0.3491809116809117, + "grad_norm": 0.5260964035987854, + "learning_rate": 0.0001963318493276974, + "loss": 1.0253, + "step": 1961 + }, + { + "epoch": 0.34935897435897434, + "grad_norm": 0.48478585481643677, + "learning_rate": 0.00019632809201217238, + "loss": 1.137, + "step": 1962 + }, + { + "epoch": 0.34953703703703703, + "grad_norm": 0.620033860206604, + "learning_rate": 0.0001963243328092985, + "loss": 1.3445, + "step": 1963 + }, + { + "epoch": 0.34971509971509973, + "grad_norm": 0.5149700045585632, + "learning_rate": 0.00019632057171914942, + "loss": 1.1042, + "step": 1964 + }, + { + "epoch": 0.34989316239316237, + "grad_norm": 0.42695048451423645, + "learning_rate": 0.0001963168087417988, + "loss": 0.8789, + "step": 1965 + }, + { + "epoch": 0.35007122507122507, + "grad_norm": 0.5281283855438232, + "learning_rate": 0.00019631304387732044, + "loss": 1.1155, + "step": 1966 + }, + { + "epoch": 0.35024928774928776, + "grad_norm": 0.4994089901447296, + "learning_rate": 0.00019630927712578804, + "loss": 1.1226, + "step": 1967 + }, + { + "epoch": 0.3504273504273504, + "grad_norm": 0.4433288276195526, + "learning_rate": 0.0001963055084872754, + "loss": 1.0262, + "step": 1968 + }, + { + "epoch": 0.3506054131054131, + "grad_norm": 0.46541857719421387, + "learning_rate": 0.0001963017379618564, + "loss": 1.1438, + "step": 1969 + }, + { + "epoch": 0.3507834757834758, + "grad_norm": 0.5097604393959045, + "learning_rate": 0.00019629796554960488, + "loss": 0.9641, + "step": 1970 + }, + { + "epoch": 0.35096153846153844, + "grad_norm": 0.49461981654167175, + "learning_rate": 0.00019629419125059478, + "loss": 1.1765, + "step": 1971 + }, + { + "epoch": 0.35113960113960113, + "grad_norm": 0.4763339161872864, + "learning_rate": 0.00019629041506490005, + "loss": 1.0527, + "step": 1972 + }, + { + "epoch": 0.35131766381766383, + "grad_norm": 0.4528443217277527, + "learning_rate": 0.00019628663699259463, + "loss": 1.1409, + "step": 1973 + }, + { + "epoch": 0.35149572649572647, + "grad_norm": 0.4436309039592743, + "learning_rate": 0.00019628285703375258, + "loss": 1.0459, + "step": 1974 + }, + { + "epoch": 0.35167378917378916, + "grad_norm": 0.5146129727363586, + "learning_rate": 0.00019627907518844797, + "loss": 1.2527, + "step": 1975 + }, + { + "epoch": 0.35185185185185186, + "grad_norm": 0.5202171802520752, + "learning_rate": 0.0001962752914567549, + "loss": 1.226, + "step": 1976 + }, + { + "epoch": 0.35202991452991456, + "grad_norm": 0.5267411470413208, + "learning_rate": 0.00019627150583874747, + "loss": 1.0898, + "step": 1977 + }, + { + "epoch": 0.3522079772079772, + "grad_norm": 0.546840250492096, + "learning_rate": 0.00019626771833449987, + "loss": 1.1716, + "step": 1978 + }, + { + "epoch": 0.3523860398860399, + "grad_norm": 0.5525290966033936, + "learning_rate": 0.0001962639289440863, + "loss": 1.1762, + "step": 1979 + }, + { + "epoch": 0.3525641025641026, + "grad_norm": 0.48967215418815613, + "learning_rate": 0.000196260137667581, + "loss": 1.1884, + "step": 1980 + }, + { + "epoch": 0.35274216524216523, + "grad_norm": 0.5908235907554626, + "learning_rate": 0.0001962563445050583, + "loss": 1.1887, + "step": 1981 + }, + { + "epoch": 0.3529202279202279, + "grad_norm": 0.46708086133003235, + "learning_rate": 0.00019625254945659245, + "loss": 0.8842, + "step": 1982 + }, + { + "epoch": 0.3530982905982906, + "grad_norm": 0.41652458906173706, + "learning_rate": 0.00019624875252225788, + "loss": 1.0268, + "step": 1983 + }, + { + "epoch": 0.35327635327635326, + "grad_norm": 0.5084529519081116, + "learning_rate": 0.00019624495370212892, + "loss": 1.0547, + "step": 1984 + }, + { + "epoch": 0.35345441595441596, + "grad_norm": 0.5667507648468018, + "learning_rate": 0.00019624115299628003, + "loss": 1.0656, + "step": 1985 + }, + { + "epoch": 0.35363247863247865, + "grad_norm": 0.5022873282432556, + "learning_rate": 0.00019623735040478568, + "loss": 1.0627, + "step": 1986 + }, + { + "epoch": 0.3538105413105413, + "grad_norm": 0.48342058062553406, + "learning_rate": 0.00019623354592772035, + "loss": 1.0976, + "step": 1987 + }, + { + "epoch": 0.353988603988604, + "grad_norm": 0.48117366433143616, + "learning_rate": 0.0001962297395651586, + "loss": 1.0515, + "step": 1988 + }, + { + "epoch": 0.3541666666666667, + "grad_norm": 0.492564857006073, + "learning_rate": 0.000196225931317175, + "loss": 1.1957, + "step": 1989 + }, + { + "epoch": 0.3543447293447293, + "grad_norm": 0.4756208658218384, + "learning_rate": 0.00019622212118384417, + "loss": 1.007, + "step": 1990 + }, + { + "epoch": 0.354522792022792, + "grad_norm": 0.581930935382843, + "learning_rate": 0.00019621830916524076, + "loss": 1.232, + "step": 1991 + }, + { + "epoch": 0.3547008547008547, + "grad_norm": 0.480064332485199, + "learning_rate": 0.00019621449526143947, + "loss": 1.2693, + "step": 1992 + }, + { + "epoch": 0.35487891737891736, + "grad_norm": 0.5679123401641846, + "learning_rate": 0.000196210679472515, + "loss": 1.2985, + "step": 1993 + }, + { + "epoch": 0.35505698005698005, + "grad_norm": 0.43757280707359314, + "learning_rate": 0.00019620686179854213, + "loss": 1.1387, + "step": 1994 + }, + { + "epoch": 0.35523504273504275, + "grad_norm": 0.4950634837150574, + "learning_rate": 0.00019620304223959566, + "loss": 1.1809, + "step": 1995 + }, + { + "epoch": 0.3554131054131054, + "grad_norm": 0.5574113726615906, + "learning_rate": 0.00019619922079575043, + "loss": 1.2434, + "step": 1996 + }, + { + "epoch": 0.3555911680911681, + "grad_norm": 0.5154930949211121, + "learning_rate": 0.00019619539746708128, + "loss": 1.1747, + "step": 1997 + }, + { + "epoch": 0.3557692307692308, + "grad_norm": 0.4377825856208801, + "learning_rate": 0.00019619157225366315, + "loss": 0.9547, + "step": 1998 + }, + { + "epoch": 0.3559472934472934, + "grad_norm": 0.530714213848114, + "learning_rate": 0.00019618774515557097, + "loss": 1.2057, + "step": 1999 + }, + { + "epoch": 0.3561253561253561, + "grad_norm": 0.5703464150428772, + "learning_rate": 0.00019618391617287978, + "loss": 1.3068, + "step": 2000 + }, + { + "epoch": 0.3563034188034188, + "grad_norm": 0.4862228333950043, + "learning_rate": 0.0001961800853056645, + "loss": 1.0077, + "step": 2001 + }, + { + "epoch": 0.35648148148148145, + "grad_norm": 0.5575395822525024, + "learning_rate": 0.00019617625255400028, + "loss": 1.03, + "step": 2002 + }, + { + "epoch": 0.35665954415954415, + "grad_norm": 0.4826279580593109, + "learning_rate": 0.0001961724179179622, + "loss": 1.268, + "step": 2003 + }, + { + "epoch": 0.35683760683760685, + "grad_norm": 0.49423274397850037, + "learning_rate": 0.00019616858139762534, + "loss": 1.1305, + "step": 2004 + }, + { + "epoch": 0.35701566951566954, + "grad_norm": 0.5208541750907898, + "learning_rate": 0.00019616474299306491, + "loss": 1.1651, + "step": 2005 + }, + { + "epoch": 0.3571937321937322, + "grad_norm": 0.5324164032936096, + "learning_rate": 0.0001961609027043561, + "loss": 1.1406, + "step": 2006 + }, + { + "epoch": 0.3573717948717949, + "grad_norm": 0.45385462045669556, + "learning_rate": 0.00019615706053157416, + "loss": 1.0716, + "step": 2007 + }, + { + "epoch": 0.3575498575498576, + "grad_norm": 0.5016173720359802, + "learning_rate": 0.00019615321647479438, + "loss": 1.0878, + "step": 2008 + }, + { + "epoch": 0.3577279202279202, + "grad_norm": 0.5073097348213196, + "learning_rate": 0.00019614937053409205, + "loss": 1.237, + "step": 2009 + }, + { + "epoch": 0.3579059829059829, + "grad_norm": 0.48880141973495483, + "learning_rate": 0.00019614552270954256, + "loss": 0.8794, + "step": 2010 + }, + { + "epoch": 0.3580840455840456, + "grad_norm": 0.43902209401130676, + "learning_rate": 0.00019614167300122126, + "loss": 0.912, + "step": 2011 + }, + { + "epoch": 0.35826210826210825, + "grad_norm": 0.42809322476387024, + "learning_rate": 0.0001961378214092036, + "loss": 0.7804, + "step": 2012 + }, + { + "epoch": 0.35844017094017094, + "grad_norm": 0.4464281499385834, + "learning_rate": 0.00019613396793356503, + "loss": 1.0004, + "step": 2013 + }, + { + "epoch": 0.35861823361823364, + "grad_norm": 0.49085676670074463, + "learning_rate": 0.00019613011257438109, + "loss": 1.1087, + "step": 2014 + }, + { + "epoch": 0.3587962962962963, + "grad_norm": 0.4997732937335968, + "learning_rate": 0.00019612625533172725, + "loss": 0.9591, + "step": 2015 + }, + { + "epoch": 0.358974358974359, + "grad_norm": 0.48442545533180237, + "learning_rate": 0.00019612239620567912, + "loss": 0.9744, + "step": 2016 + }, + { + "epoch": 0.35915242165242167, + "grad_norm": 0.4989205002784729, + "learning_rate": 0.00019611853519631233, + "loss": 0.9844, + "step": 2017 + }, + { + "epoch": 0.3593304843304843, + "grad_norm": 0.6107521653175354, + "learning_rate": 0.00019611467230370248, + "loss": 1.147, + "step": 2018 + }, + { + "epoch": 0.359508547008547, + "grad_norm": 0.5594844818115234, + "learning_rate": 0.00019611080752792535, + "loss": 1.3195, + "step": 2019 + }, + { + "epoch": 0.3596866096866097, + "grad_norm": 0.4786946475505829, + "learning_rate": 0.00019610694086905656, + "loss": 1.2108, + "step": 2020 + }, + { + "epoch": 0.35986467236467234, + "grad_norm": 0.5186030268669128, + "learning_rate": 0.0001961030723271719, + "loss": 1.0008, + "step": 2021 + }, + { + "epoch": 0.36004273504273504, + "grad_norm": 0.4520573318004608, + "learning_rate": 0.0001960992019023472, + "loss": 1.1307, + "step": 2022 + }, + { + "epoch": 0.36022079772079774, + "grad_norm": 0.4983210563659668, + "learning_rate": 0.00019609532959465823, + "loss": 1.1486, + "step": 2023 + }, + { + "epoch": 0.3603988603988604, + "grad_norm": 0.6209200024604797, + "learning_rate": 0.00019609145540418094, + "loss": 1.2566, + "step": 2024 + }, + { + "epoch": 0.3605769230769231, + "grad_norm": 0.47047603130340576, + "learning_rate": 0.00019608757933099117, + "loss": 1.1588, + "step": 2025 + }, + { + "epoch": 0.36075498575498577, + "grad_norm": 0.5147389769554138, + "learning_rate": 0.0001960837013751649, + "loss": 1.2113, + "step": 2026 + }, + { + "epoch": 0.3609330484330484, + "grad_norm": 0.45826098322868347, + "learning_rate": 0.00019607982153677808, + "loss": 1.13, + "step": 2027 + }, + { + "epoch": 0.3611111111111111, + "grad_norm": 0.5699561834335327, + "learning_rate": 0.00019607593981590675, + "loss": 1.2476, + "step": 2028 + }, + { + "epoch": 0.3612891737891738, + "grad_norm": 0.5349239110946655, + "learning_rate": 0.000196072056212627, + "loss": 1.2295, + "step": 2029 + }, + { + "epoch": 0.36146723646723644, + "grad_norm": 0.6212165355682373, + "learning_rate": 0.00019606817072701484, + "loss": 1.1965, + "step": 2030 + }, + { + "epoch": 0.36164529914529914, + "grad_norm": 0.4870990216732025, + "learning_rate": 0.00019606428335914645, + "loss": 1.4464, + "step": 2031 + }, + { + "epoch": 0.36182336182336183, + "grad_norm": 0.42427828907966614, + "learning_rate": 0.00019606039410909797, + "loss": 1.1546, + "step": 2032 + }, + { + "epoch": 0.36200142450142453, + "grad_norm": 0.5081788301467896, + "learning_rate": 0.0001960565029769456, + "loss": 1.1867, + "step": 2033 + }, + { + "epoch": 0.36217948717948717, + "grad_norm": 0.4813104271888733, + "learning_rate": 0.00019605260996276565, + "loss": 1.3726, + "step": 2034 + }, + { + "epoch": 0.36235754985754987, + "grad_norm": 0.4648851156234741, + "learning_rate": 0.0001960487150666343, + "loss": 1.2434, + "step": 2035 + }, + { + "epoch": 0.36253561253561256, + "grad_norm": 0.484161913394928, + "learning_rate": 0.00019604481828862792, + "loss": 1.1309, + "step": 2036 + }, + { + "epoch": 0.3627136752136752, + "grad_norm": 0.4929439127445221, + "learning_rate": 0.00019604091962882283, + "loss": 1.1007, + "step": 2037 + }, + { + "epoch": 0.3628917378917379, + "grad_norm": 0.45599642395973206, + "learning_rate": 0.00019603701908729544, + "loss": 1.2628, + "step": 2038 + }, + { + "epoch": 0.3630698005698006, + "grad_norm": 0.45295149087905884, + "learning_rate": 0.00019603311666412213, + "loss": 0.9808, + "step": 2039 + }, + { + "epoch": 0.36324786324786323, + "grad_norm": 0.48681163787841797, + "learning_rate": 0.00019602921235937942, + "loss": 1.0574, + "step": 2040 + }, + { + "epoch": 0.36342592592592593, + "grad_norm": 0.41232365369796753, + "learning_rate": 0.00019602530617314378, + "loss": 1.0454, + "step": 2041 + }, + { + "epoch": 0.3636039886039886, + "grad_norm": 0.46214723587036133, + "learning_rate": 0.00019602139810549174, + "loss": 0.9985, + "step": 2042 + }, + { + "epoch": 0.36378205128205127, + "grad_norm": 0.44307878613471985, + "learning_rate": 0.00019601748815649989, + "loss": 0.9683, + "step": 2043 + }, + { + "epoch": 0.36396011396011396, + "grad_norm": 0.4809451401233673, + "learning_rate": 0.00019601357632624477, + "loss": 1.028, + "step": 2044 + }, + { + "epoch": 0.36413817663817666, + "grad_norm": 0.4638497531414032, + "learning_rate": 0.0001960096626148031, + "loss": 0.9851, + "step": 2045 + }, + { + "epoch": 0.3643162393162393, + "grad_norm": 0.5942164063453674, + "learning_rate": 0.00019600574702225153, + "loss": 1.1606, + "step": 2046 + }, + { + "epoch": 0.364494301994302, + "grad_norm": 0.5171293616294861, + "learning_rate": 0.00019600182954866675, + "loss": 1.2335, + "step": 2047 + }, + { + "epoch": 0.3646723646723647, + "grad_norm": 0.5294404625892639, + "learning_rate": 0.00019599791019412558, + "loss": 1.0966, + "step": 2048 + }, + { + "epoch": 0.36485042735042733, + "grad_norm": 0.46117448806762695, + "learning_rate": 0.00019599398895870477, + "loss": 1.0565, + "step": 2049 + }, + { + "epoch": 0.36502849002849, + "grad_norm": 0.5385118126869202, + "learning_rate": 0.00019599006584248118, + "loss": 1.0076, + "step": 2050 + }, + { + "epoch": 0.3652065527065527, + "grad_norm": 0.4915166199207306, + "learning_rate": 0.00019598614084553165, + "loss": 0.9686, + "step": 2051 + }, + { + "epoch": 0.36538461538461536, + "grad_norm": 0.46769094467163086, + "learning_rate": 0.00019598221396793303, + "loss": 1.1217, + "step": 2052 + }, + { + "epoch": 0.36556267806267806, + "grad_norm": 0.5440493822097778, + "learning_rate": 0.00019597828520976236, + "loss": 1.2344, + "step": 2053 + }, + { + "epoch": 0.36574074074074076, + "grad_norm": 0.616727352142334, + "learning_rate": 0.00019597435457109657, + "loss": 1.2953, + "step": 2054 + }, + { + "epoch": 0.3659188034188034, + "grad_norm": 0.4859183430671692, + "learning_rate": 0.00019597042205201265, + "loss": 1.16, + "step": 2055 + }, + { + "epoch": 0.3660968660968661, + "grad_norm": 0.47056329250335693, + "learning_rate": 0.0001959664876525877, + "loss": 0.9982, + "step": 2056 + }, + { + "epoch": 0.3662749287749288, + "grad_norm": 0.48347967863082886, + "learning_rate": 0.00019596255137289875, + "loss": 1.0966, + "step": 2057 + }, + { + "epoch": 0.36645299145299143, + "grad_norm": 0.5068454742431641, + "learning_rate": 0.00019595861321302296, + "loss": 1.2891, + "step": 2058 + }, + { + "epoch": 0.3666310541310541, + "grad_norm": 0.5702359080314636, + "learning_rate": 0.00019595467317303747, + "loss": 1.1394, + "step": 2059 + }, + { + "epoch": 0.3668091168091168, + "grad_norm": 0.5028812885284424, + "learning_rate": 0.0001959507312530195, + "loss": 1.2324, + "step": 2060 + }, + { + "epoch": 0.36698717948717946, + "grad_norm": 0.4672880172729492, + "learning_rate": 0.00019594678745304628, + "loss": 1.0581, + "step": 2061 + }, + { + "epoch": 0.36716524216524216, + "grad_norm": 0.5233900547027588, + "learning_rate": 0.00019594284177319504, + "loss": 1.138, + "step": 2062 + }, + { + "epoch": 0.36734330484330485, + "grad_norm": 0.46871712803840637, + "learning_rate": 0.00019593889421354316, + "loss": 1.2159, + "step": 2063 + }, + { + "epoch": 0.36752136752136755, + "grad_norm": 0.5180533528327942, + "learning_rate": 0.00019593494477416793, + "loss": 1.1116, + "step": 2064 + }, + { + "epoch": 0.3676994301994302, + "grad_norm": 0.5398494005203247, + "learning_rate": 0.0001959309934551467, + "loss": 1.2038, + "step": 2065 + }, + { + "epoch": 0.3678774928774929, + "grad_norm": 0.4850373864173889, + "learning_rate": 0.000195927040256557, + "loss": 1.4315, + "step": 2066 + }, + { + "epoch": 0.3680555555555556, + "grad_norm": 0.49190905690193176, + "learning_rate": 0.0001959230851784762, + "loss": 0.9993, + "step": 2067 + }, + { + "epoch": 0.3682336182336182, + "grad_norm": 0.4546903073787689, + "learning_rate": 0.00019591912822098178, + "loss": 1.0979, + "step": 2068 + }, + { + "epoch": 0.3684116809116809, + "grad_norm": 0.4726468622684479, + "learning_rate": 0.00019591516938415133, + "loss": 1.1629, + "step": 2069 + }, + { + "epoch": 0.3685897435897436, + "grad_norm": 0.47856009006500244, + "learning_rate": 0.00019591120866806235, + "loss": 1.2048, + "step": 2070 + }, + { + "epoch": 0.36876780626780625, + "grad_norm": 0.46847718954086304, + "learning_rate": 0.0001959072460727925, + "loss": 1.0958, + "step": 2071 + }, + { + "epoch": 0.36894586894586895, + "grad_norm": 0.47164350748062134, + "learning_rate": 0.0001959032815984194, + "loss": 1.1912, + "step": 2072 + }, + { + "epoch": 0.36912393162393164, + "grad_norm": 0.4838213324546814, + "learning_rate": 0.0001958993152450207, + "loss": 1.1466, + "step": 2073 + }, + { + "epoch": 0.3693019943019943, + "grad_norm": 0.47234636545181274, + "learning_rate": 0.00019589534701267412, + "loss": 0.9475, + "step": 2074 + }, + { + "epoch": 0.369480056980057, + "grad_norm": 0.4913126826286316, + "learning_rate": 0.00019589137690145746, + "loss": 1.1571, + "step": 2075 + }, + { + "epoch": 0.3696581196581197, + "grad_norm": 0.4696233570575714, + "learning_rate": 0.00019588740491144842, + "loss": 0.9797, + "step": 2076 + }, + { + "epoch": 0.3698361823361823, + "grad_norm": 0.46146106719970703, + "learning_rate": 0.00019588343104272492, + "loss": 1.027, + "step": 2077 + }, + { + "epoch": 0.370014245014245, + "grad_norm": 0.4920627176761627, + "learning_rate": 0.00019587945529536474, + "loss": 1.1008, + "step": 2078 + }, + { + "epoch": 0.3701923076923077, + "grad_norm": 0.4854249954223633, + "learning_rate": 0.0001958754776694458, + "loss": 1.0759, + "step": 2079 + }, + { + "epoch": 0.37037037037037035, + "grad_norm": 0.4884897768497467, + "learning_rate": 0.00019587149816504608, + "loss": 1.1403, + "step": 2080 + }, + { + "epoch": 0.37054843304843305, + "grad_norm": 0.5062584280967712, + "learning_rate": 0.00019586751678224345, + "loss": 1.0185, + "step": 2081 + }, + { + "epoch": 0.37072649572649574, + "grad_norm": 0.44697675108909607, + "learning_rate": 0.000195863533521116, + "loss": 1.0462, + "step": 2082 + }, + { + "epoch": 0.3709045584045584, + "grad_norm": 0.5122885704040527, + "learning_rate": 0.00019585954838174176, + "loss": 1.108, + "step": 2083 + }, + { + "epoch": 0.3710826210826211, + "grad_norm": 0.486650288105011, + "learning_rate": 0.0001958555613641988, + "loss": 1.126, + "step": 2084 + }, + { + "epoch": 0.3712606837606838, + "grad_norm": 0.5296297669410706, + "learning_rate": 0.00019585157246856523, + "loss": 1.1757, + "step": 2085 + }, + { + "epoch": 0.3714387464387464, + "grad_norm": 0.4935721457004547, + "learning_rate": 0.0001958475816949192, + "loss": 1.1654, + "step": 2086 + }, + { + "epoch": 0.3716168091168091, + "grad_norm": 0.6226509213447571, + "learning_rate": 0.00019584358904333891, + "loss": 1.1981, + "step": 2087 + }, + { + "epoch": 0.3717948717948718, + "grad_norm": 0.44094228744506836, + "learning_rate": 0.0001958395945139026, + "loss": 0.8468, + "step": 2088 + }, + { + "epoch": 0.37197293447293445, + "grad_norm": 0.5335884690284729, + "learning_rate": 0.00019583559810668858, + "loss": 1.1597, + "step": 2089 + }, + { + "epoch": 0.37215099715099714, + "grad_norm": 0.4585414528846741, + "learning_rate": 0.000195831599821775, + "loss": 0.9343, + "step": 2090 + }, + { + "epoch": 0.37232905982905984, + "grad_norm": 0.533087432384491, + "learning_rate": 0.00019582759965924035, + "loss": 1.1209, + "step": 2091 + }, + { + "epoch": 0.37250712250712253, + "grad_norm": 0.5302683711051941, + "learning_rate": 0.00019582359761916295, + "loss": 1.236, + "step": 2092 + }, + { + "epoch": 0.3726851851851852, + "grad_norm": 0.4522508382797241, + "learning_rate": 0.00019581959370162122, + "loss": 1.0196, + "step": 2093 + }, + { + "epoch": 0.37286324786324787, + "grad_norm": 0.52391517162323, + "learning_rate": 0.00019581558790669358, + "loss": 1.0077, + "step": 2094 + }, + { + "epoch": 0.37304131054131057, + "grad_norm": 0.47144797444343567, + "learning_rate": 0.00019581158023445854, + "loss": 1.0956, + "step": 2095 + }, + { + "epoch": 0.3732193732193732, + "grad_norm": 0.4486723244190216, + "learning_rate": 0.00019580757068499459, + "loss": 0.8697, + "step": 2096 + }, + { + "epoch": 0.3733974358974359, + "grad_norm": 0.4626580476760864, + "learning_rate": 0.00019580355925838034, + "loss": 0.8489, + "step": 2097 + }, + { + "epoch": 0.3735754985754986, + "grad_norm": 0.5647920370101929, + "learning_rate": 0.00019579954595469438, + "loss": 1.1458, + "step": 2098 + }, + { + "epoch": 0.37375356125356124, + "grad_norm": 0.4734349846839905, + "learning_rate": 0.00019579553077401528, + "loss": 1.1036, + "step": 2099 + }, + { + "epoch": 0.37393162393162394, + "grad_norm": 0.5624295473098755, + "learning_rate": 0.00019579151371642176, + "loss": 0.9793, + "step": 2100 + }, + { + "epoch": 0.37410968660968663, + "grad_norm": 0.47507283091545105, + "learning_rate": 0.00019578749478199256, + "loss": 1.0371, + "step": 2101 + }, + { + "epoch": 0.37428774928774927, + "grad_norm": 0.550865113735199, + "learning_rate": 0.00019578347397080633, + "loss": 1.046, + "step": 2102 + }, + { + "epoch": 0.37446581196581197, + "grad_norm": 0.5249403715133667, + "learning_rate": 0.00019577945128294193, + "loss": 1.3185, + "step": 2103 + }, + { + "epoch": 0.37464387464387466, + "grad_norm": 0.4921024739742279, + "learning_rate": 0.00019577542671847815, + "loss": 1.0758, + "step": 2104 + }, + { + "epoch": 0.3748219373219373, + "grad_norm": 0.5351784825325012, + "learning_rate": 0.00019577140027749384, + "loss": 1.067, + "step": 2105 + }, + { + "epoch": 0.375, + "grad_norm": 0.44420507550239563, + "learning_rate": 0.00019576737196006787, + "loss": 1.1065, + "step": 2106 + }, + { + "epoch": 0.3751780626780627, + "grad_norm": 0.531384289264679, + "learning_rate": 0.0001957633417662792, + "loss": 1.1634, + "step": 2107 + }, + { + "epoch": 0.37535612535612534, + "grad_norm": 0.5167618989944458, + "learning_rate": 0.00019575930969620677, + "loss": 1.1646, + "step": 2108 + }, + { + "epoch": 0.37553418803418803, + "grad_norm": 0.41487228870391846, + "learning_rate": 0.0001957552757499296, + "loss": 0.793, + "step": 2109 + }, + { + "epoch": 0.37571225071225073, + "grad_norm": 0.5110787153244019, + "learning_rate": 0.00019575123992752672, + "loss": 1.1752, + "step": 2110 + }, + { + "epoch": 0.37589031339031337, + "grad_norm": 0.4422051012516022, + "learning_rate": 0.00019574720222907717, + "loss": 1.0102, + "step": 2111 + }, + { + "epoch": 0.37606837606837606, + "grad_norm": 0.4757538139820099, + "learning_rate": 0.0001957431626546601, + "loss": 1.0467, + "step": 2112 + }, + { + "epoch": 0.37624643874643876, + "grad_norm": 0.4736764430999756, + "learning_rate": 0.00019573912120435466, + "loss": 1.3048, + "step": 2113 + }, + { + "epoch": 0.3764245014245014, + "grad_norm": 0.49894335865974426, + "learning_rate": 0.00019573507787824004, + "loss": 1.0502, + "step": 2114 + }, + { + "epoch": 0.3766025641025641, + "grad_norm": 0.48120981454849243, + "learning_rate": 0.00019573103267639543, + "loss": 1.2405, + "step": 2115 + }, + { + "epoch": 0.3767806267806268, + "grad_norm": 0.4826737642288208, + "learning_rate": 0.0001957269855989001, + "loss": 1.1189, + "step": 2116 + }, + { + "epoch": 0.37695868945868943, + "grad_norm": 0.4736921489238739, + "learning_rate": 0.0001957229366458333, + "loss": 1.2862, + "step": 2117 + }, + { + "epoch": 0.37713675213675213, + "grad_norm": 0.3895208537578583, + "learning_rate": 0.00019571888581727446, + "loss": 1.0573, + "step": 2118 + }, + { + "epoch": 0.3773148148148148, + "grad_norm": 0.5107510089874268, + "learning_rate": 0.00019571483311330284, + "loss": 1.2913, + "step": 2119 + }, + { + "epoch": 0.37749287749287747, + "grad_norm": 0.4543241262435913, + "learning_rate": 0.00019571077853399794, + "loss": 0.949, + "step": 2120 + }, + { + "epoch": 0.37767094017094016, + "grad_norm": 0.46897491812705994, + "learning_rate": 0.00019570672207943913, + "loss": 1.2235, + "step": 2121 + }, + { + "epoch": 0.37784900284900286, + "grad_norm": 0.4812130630016327, + "learning_rate": 0.0001957026637497059, + "loss": 0.8857, + "step": 2122 + }, + { + "epoch": 0.37802706552706555, + "grad_norm": 0.47452476620674133, + "learning_rate": 0.00019569860354487782, + "loss": 1.0549, + "step": 2123 + }, + { + "epoch": 0.3782051282051282, + "grad_norm": 0.49879950284957886, + "learning_rate": 0.00019569454146503438, + "loss": 1.0475, + "step": 2124 + }, + { + "epoch": 0.3783831908831909, + "grad_norm": 0.4246445894241333, + "learning_rate": 0.00019569047751025518, + "loss": 0.8788, + "step": 2125 + }, + { + "epoch": 0.3785612535612536, + "grad_norm": 0.4868565499782562, + "learning_rate": 0.00019568641168061986, + "loss": 1.1801, + "step": 2126 + }, + { + "epoch": 0.3787393162393162, + "grad_norm": 0.46723654866218567, + "learning_rate": 0.0001956823439762081, + "loss": 1.1661, + "step": 2127 + }, + { + "epoch": 0.3789173789173789, + "grad_norm": 0.4989059269428253, + "learning_rate": 0.00019567827439709954, + "loss": 1.3037, + "step": 2128 + }, + { + "epoch": 0.3790954415954416, + "grad_norm": 0.441307932138443, + "learning_rate": 0.00019567420294337395, + "loss": 1.0197, + "step": 2129 + }, + { + "epoch": 0.37927350427350426, + "grad_norm": 0.5200160145759583, + "learning_rate": 0.0001956701296151111, + "loss": 1.3366, + "step": 2130 + }, + { + "epoch": 0.37945156695156695, + "grad_norm": 0.43610256910324097, + "learning_rate": 0.00019566605441239082, + "loss": 1.0148, + "step": 2131 + }, + { + "epoch": 0.37962962962962965, + "grad_norm": 0.4160982370376587, + "learning_rate": 0.00019566197733529293, + "loss": 1.0758, + "step": 2132 + }, + { + "epoch": 0.3798076923076923, + "grad_norm": 0.5007950663566589, + "learning_rate": 0.00019565789838389726, + "loss": 1.1937, + "step": 2133 + }, + { + "epoch": 0.379985754985755, + "grad_norm": 0.4991525113582611, + "learning_rate": 0.00019565381755828385, + "loss": 1.1788, + "step": 2134 + }, + { + "epoch": 0.3801638176638177, + "grad_norm": 0.6313113570213318, + "learning_rate": 0.00019564973485853258, + "loss": 1.1241, + "step": 2135 + }, + { + "epoch": 0.3803418803418803, + "grad_norm": 0.49736538529396057, + "learning_rate": 0.0001956456502847234, + "loss": 1.0299, + "step": 2136 + }, + { + "epoch": 0.380519943019943, + "grad_norm": 0.4384380578994751, + "learning_rate": 0.00019564156383693643, + "loss": 1.132, + "step": 2137 + }, + { + "epoch": 0.3806980056980057, + "grad_norm": 0.4696183502674103, + "learning_rate": 0.00019563747551525168, + "loss": 1.1145, + "step": 2138 + }, + { + "epoch": 0.38087606837606836, + "grad_norm": 0.42039749026298523, + "learning_rate": 0.0001956333853197493, + "loss": 0.9549, + "step": 2139 + }, + { + "epoch": 0.38105413105413105, + "grad_norm": 0.5547221899032593, + "learning_rate": 0.00019562929325050936, + "loss": 1.0476, + "step": 2140 + }, + { + "epoch": 0.38123219373219375, + "grad_norm": 0.4803301692008972, + "learning_rate": 0.0001956251993076121, + "loss": 1.1285, + "step": 2141 + }, + { + "epoch": 0.3814102564102564, + "grad_norm": 0.609501838684082, + "learning_rate": 0.00019562110349113766, + "loss": 1.2375, + "step": 2142 + }, + { + "epoch": 0.3815883190883191, + "grad_norm": 0.5134759545326233, + "learning_rate": 0.00019561700580116639, + "loss": 1.0895, + "step": 2143 + }, + { + "epoch": 0.3817663817663818, + "grad_norm": 0.5086711049079895, + "learning_rate": 0.00019561290623777846, + "loss": 1.1139, + "step": 2144 + }, + { + "epoch": 0.3819444444444444, + "grad_norm": 0.5371596813201904, + "learning_rate": 0.00019560880480105428, + "loss": 0.9302, + "step": 2145 + }, + { + "epoch": 0.3821225071225071, + "grad_norm": 0.4966319799423218, + "learning_rate": 0.00019560470149107418, + "loss": 1.2485, + "step": 2146 + }, + { + "epoch": 0.3823005698005698, + "grad_norm": 0.5296950340270996, + "learning_rate": 0.00019560059630791855, + "loss": 1.4449, + "step": 2147 + }, + { + "epoch": 0.38247863247863245, + "grad_norm": 0.5564194321632385, + "learning_rate": 0.00019559648925166783, + "loss": 1.0817, + "step": 2148 + }, + { + "epoch": 0.38265669515669515, + "grad_norm": 0.5763841867446899, + "learning_rate": 0.0001955923803224025, + "loss": 1.1915, + "step": 2149 + }, + { + "epoch": 0.38283475783475784, + "grad_norm": 0.4782295823097229, + "learning_rate": 0.00019558826952020304, + "loss": 1.1317, + "step": 2150 + }, + { + "epoch": 0.38301282051282054, + "grad_norm": 0.4876856207847595, + "learning_rate": 0.00019558415684515002, + "loss": 1.2113, + "step": 2151 + }, + { + "epoch": 0.3831908831908832, + "grad_norm": 0.4894421398639679, + "learning_rate": 0.00019558004229732398, + "loss": 1.0761, + "step": 2152 + }, + { + "epoch": 0.3833689458689459, + "grad_norm": 0.47914227843284607, + "learning_rate": 0.0001955759258768056, + "loss": 1.0869, + "step": 2153 + }, + { + "epoch": 0.38354700854700857, + "grad_norm": 0.43933629989624023, + "learning_rate": 0.00019557180758367543, + "loss": 1.0581, + "step": 2154 + }, + { + "epoch": 0.3837250712250712, + "grad_norm": 0.4078103005886078, + "learning_rate": 0.00019556768741801428, + "loss": 1.065, + "step": 2155 + }, + { + "epoch": 0.3839031339031339, + "grad_norm": 0.5112793445587158, + "learning_rate": 0.00019556356537990278, + "loss": 1.2023, + "step": 2156 + }, + { + "epoch": 0.3840811965811966, + "grad_norm": 0.4699678122997284, + "learning_rate": 0.00019555944146942177, + "loss": 1.2459, + "step": 2157 + }, + { + "epoch": 0.38425925925925924, + "grad_norm": 0.4723528027534485, + "learning_rate": 0.00019555531568665198, + "loss": 1.2204, + "step": 2158 + }, + { + "epoch": 0.38443732193732194, + "grad_norm": 0.4648225009441376, + "learning_rate": 0.00019555118803167432, + "loss": 1.1355, + "step": 2159 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 0.49861815571784973, + "learning_rate": 0.00019554705850456961, + "loss": 1.1301, + "step": 2160 + }, + { + "epoch": 0.3847934472934473, + "grad_norm": 0.4076344966888428, + "learning_rate": 0.00019554292710541874, + "loss": 0.8997, + "step": 2161 + }, + { + "epoch": 0.38497150997151, + "grad_norm": 0.5510796308517456, + "learning_rate": 0.00019553879383430272, + "loss": 1.0594, + "step": 2162 + }, + { + "epoch": 0.38514957264957267, + "grad_norm": 0.55793696641922, + "learning_rate": 0.00019553465869130249, + "loss": 1.1284, + "step": 2163 + }, + { + "epoch": 0.3853276353276353, + "grad_norm": 0.5096491575241089, + "learning_rate": 0.00019553052167649906, + "loss": 1.0419, + "step": 2164 + }, + { + "epoch": 0.385505698005698, + "grad_norm": 0.49077361822128296, + "learning_rate": 0.0001955263827899735, + "loss": 1.1632, + "step": 2165 + }, + { + "epoch": 0.3856837606837607, + "grad_norm": 0.5546894073486328, + "learning_rate": 0.00019552224203180693, + "loss": 1.1487, + "step": 2166 + }, + { + "epoch": 0.38586182336182334, + "grad_norm": 0.4930037260055542, + "learning_rate": 0.00019551809940208047, + "loss": 1.2668, + "step": 2167 + }, + { + "epoch": 0.38603988603988604, + "grad_norm": 0.5600671172142029, + "learning_rate": 0.00019551395490087525, + "loss": 1.3988, + "step": 2168 + }, + { + "epoch": 0.38621794871794873, + "grad_norm": 0.45897629857063293, + "learning_rate": 0.0001955098085282725, + "loss": 0.7792, + "step": 2169 + }, + { + "epoch": 0.3863960113960114, + "grad_norm": 0.46138936281204224, + "learning_rate": 0.00019550566028435346, + "loss": 1.1749, + "step": 2170 + }, + { + "epoch": 0.38657407407407407, + "grad_norm": 0.5136167407035828, + "learning_rate": 0.0001955015101691994, + "loss": 1.0153, + "step": 2171 + }, + { + "epoch": 0.38675213675213677, + "grad_norm": 0.4886440336704254, + "learning_rate": 0.00019549735818289165, + "loss": 1.0006, + "step": 2172 + }, + { + "epoch": 0.3869301994301994, + "grad_norm": 0.4339776635169983, + "learning_rate": 0.00019549320432551154, + "loss": 1.0109, + "step": 2173 + }, + { + "epoch": 0.3871082621082621, + "grad_norm": 0.48729443550109863, + "learning_rate": 0.00019548904859714044, + "loss": 1.2016, + "step": 2174 + }, + { + "epoch": 0.3872863247863248, + "grad_norm": 0.5128757357597351, + "learning_rate": 0.0001954848909978598, + "loss": 1.085, + "step": 2175 + }, + { + "epoch": 0.38746438746438744, + "grad_norm": 0.49636292457580566, + "learning_rate": 0.0001954807315277511, + "loss": 1.0671, + "step": 2176 + }, + { + "epoch": 0.38764245014245013, + "grad_norm": 0.4946988821029663, + "learning_rate": 0.00019547657018689578, + "loss": 1.2091, + "step": 2177 + }, + { + "epoch": 0.38782051282051283, + "grad_norm": 0.49004554748535156, + "learning_rate": 0.00019547240697537544, + "loss": 1.0241, + "step": 2178 + }, + { + "epoch": 0.38799857549857547, + "grad_norm": 0.48750075697898865, + "learning_rate": 0.00019546824189327157, + "loss": 1.1082, + "step": 2179 + }, + { + "epoch": 0.38817663817663817, + "grad_norm": 0.47726166248321533, + "learning_rate": 0.00019546407494066585, + "loss": 1.1275, + "step": 2180 + }, + { + "epoch": 0.38835470085470086, + "grad_norm": 0.5253444910049438, + "learning_rate": 0.00019545990611763986, + "loss": 1.0164, + "step": 2181 + }, + { + "epoch": 0.38853276353276356, + "grad_norm": 0.4470371603965759, + "learning_rate": 0.00019545573542427533, + "loss": 1.0138, + "step": 2182 + }, + { + "epoch": 0.3887108262108262, + "grad_norm": 0.6645087599754333, + "learning_rate": 0.00019545156286065397, + "loss": 1.0884, + "step": 2183 + }, + { + "epoch": 0.3888888888888889, + "grad_norm": 0.498775839805603, + "learning_rate": 0.0001954473884268575, + "loss": 1.1035, + "step": 2184 + }, + { + "epoch": 0.3890669515669516, + "grad_norm": 0.5830566883087158, + "learning_rate": 0.00019544321212296772, + "loss": 1.1665, + "step": 2185 + }, + { + "epoch": 0.38924501424501423, + "grad_norm": 0.48162809014320374, + "learning_rate": 0.00019543903394906646, + "loss": 1.1035, + "step": 2186 + }, + { + "epoch": 0.3894230769230769, + "grad_norm": 0.46334075927734375, + "learning_rate": 0.0001954348539052356, + "loss": 0.9764, + "step": 2187 + }, + { + "epoch": 0.3896011396011396, + "grad_norm": 0.6343515515327454, + "learning_rate": 0.00019543067199155704, + "loss": 0.9474, + "step": 2188 + }, + { + "epoch": 0.38977920227920226, + "grad_norm": 0.4867806136608124, + "learning_rate": 0.0001954264882081127, + "loss": 1.1161, + "step": 2189 + }, + { + "epoch": 0.38995726495726496, + "grad_norm": 0.49305734038352966, + "learning_rate": 0.00019542230255498454, + "loss": 1.1825, + "step": 2190 + }, + { + "epoch": 0.39013532763532766, + "grad_norm": 0.518465518951416, + "learning_rate": 0.00019541811503225457, + "loss": 1.0695, + "step": 2191 + }, + { + "epoch": 0.3903133903133903, + "grad_norm": 0.4892457127571106, + "learning_rate": 0.00019541392564000488, + "loss": 1.3113, + "step": 2192 + }, + { + "epoch": 0.390491452991453, + "grad_norm": 0.5150920152664185, + "learning_rate": 0.00019540973437831753, + "loss": 1.0735, + "step": 2193 + }, + { + "epoch": 0.3906695156695157, + "grad_norm": 0.5414708256721497, + "learning_rate": 0.00019540554124727462, + "loss": 1.0773, + "step": 2194 + }, + { + "epoch": 0.39084757834757833, + "grad_norm": 0.49826398491859436, + "learning_rate": 0.0001954013462469583, + "loss": 1.0542, + "step": 2195 + }, + { + "epoch": 0.391025641025641, + "grad_norm": 0.5203596949577332, + "learning_rate": 0.0001953971493774508, + "loss": 1.178, + "step": 2196 + }, + { + "epoch": 0.3912037037037037, + "grad_norm": 0.45095738768577576, + "learning_rate": 0.00019539295063883432, + "loss": 1.1254, + "step": 2197 + }, + { + "epoch": 0.39138176638176636, + "grad_norm": 0.4938857853412628, + "learning_rate": 0.00019538875003119113, + "loss": 1.1061, + "step": 2198 + }, + { + "epoch": 0.39155982905982906, + "grad_norm": 0.5260919332504272, + "learning_rate": 0.00019538454755460354, + "loss": 1.3292, + "step": 2199 + }, + { + "epoch": 0.39173789173789175, + "grad_norm": 0.46527108550071716, + "learning_rate": 0.00019538034320915388, + "loss": 1.2074, + "step": 2200 + }, + { + "epoch": 0.3919159544159544, + "grad_norm": 0.5608304738998413, + "learning_rate": 0.00019537613699492453, + "loss": 1.0385, + "step": 2201 + }, + { + "epoch": 0.3920940170940171, + "grad_norm": 0.5056684613227844, + "learning_rate": 0.00019537192891199792, + "loss": 1.1513, + "step": 2202 + }, + { + "epoch": 0.3922720797720798, + "grad_norm": 0.3764426112174988, + "learning_rate": 0.00019536771896045644, + "loss": 0.8966, + "step": 2203 + }, + { + "epoch": 0.3924501424501424, + "grad_norm": 0.4983638823032379, + "learning_rate": 0.0001953635071403827, + "loss": 1.097, + "step": 2204 + }, + { + "epoch": 0.3926282051282051, + "grad_norm": 0.5733919739723206, + "learning_rate": 0.00019535929345185904, + "loss": 1.4992, + "step": 2205 + }, + { + "epoch": 0.3928062678062678, + "grad_norm": 0.632064163684845, + "learning_rate": 0.00019535507789496817, + "loss": 1.0611, + "step": 2206 + }, + { + "epoch": 0.39298433048433046, + "grad_norm": 0.409978449344635, + "learning_rate": 0.00019535086046979262, + "loss": 0.7172, + "step": 2207 + }, + { + "epoch": 0.39316239316239315, + "grad_norm": 0.40910813212394714, + "learning_rate": 0.00019534664117641502, + "loss": 0.8803, + "step": 2208 + }, + { + "epoch": 0.39334045584045585, + "grad_norm": 0.4696179926395416, + "learning_rate": 0.00019534242001491807, + "loss": 1.1551, + "step": 2209 + }, + { + "epoch": 0.39351851851851855, + "grad_norm": 0.538425862789154, + "learning_rate": 0.00019533819698538444, + "loss": 1.1296, + "step": 2210 + }, + { + "epoch": 0.3936965811965812, + "grad_norm": 0.5913630723953247, + "learning_rate": 0.00019533397208789692, + "loss": 0.9757, + "step": 2211 + }, + { + "epoch": 0.3938746438746439, + "grad_norm": 0.5649870038032532, + "learning_rate": 0.00019532974532253822, + "loss": 0.9976, + "step": 2212 + }, + { + "epoch": 0.3940527065527066, + "grad_norm": 0.5012063980102539, + "learning_rate": 0.00019532551668939121, + "loss": 0.9969, + "step": 2213 + }, + { + "epoch": 0.3942307692307692, + "grad_norm": 0.5098594427108765, + "learning_rate": 0.00019532128618853872, + "loss": 1.1229, + "step": 2214 + }, + { + "epoch": 0.3944088319088319, + "grad_norm": 0.4753342568874359, + "learning_rate": 0.0001953170538200636, + "loss": 1.0808, + "step": 2215 + }, + { + "epoch": 0.3945868945868946, + "grad_norm": 0.4770098626613617, + "learning_rate": 0.00019531281958404888, + "loss": 1.0656, + "step": 2216 + }, + { + "epoch": 0.39476495726495725, + "grad_norm": 0.6007979512214661, + "learning_rate": 0.00019530858348057746, + "loss": 1.0093, + "step": 2217 + }, + { + "epoch": 0.39494301994301995, + "grad_norm": 0.4501650929450989, + "learning_rate": 0.00019530434550973227, + "loss": 0.8557, + "step": 2218 + }, + { + "epoch": 0.39512108262108264, + "grad_norm": 0.5123980641365051, + "learning_rate": 0.00019530010567159645, + "loss": 0.9833, + "step": 2219 + }, + { + "epoch": 0.3952991452991453, + "grad_norm": 0.4623969495296478, + "learning_rate": 0.000195295863966253, + "loss": 0.913, + "step": 2220 + }, + { + "epoch": 0.395477207977208, + "grad_norm": 0.4341880679130554, + "learning_rate": 0.0001952916203937851, + "loss": 1.0234, + "step": 2221 + }, + { + "epoch": 0.3956552706552707, + "grad_norm": 0.5935006141662598, + "learning_rate": 0.00019528737495427581, + "loss": 1.061, + "step": 2222 + }, + { + "epoch": 0.3958333333333333, + "grad_norm": 0.44835174083709717, + "learning_rate": 0.00019528312764780837, + "loss": 1.1567, + "step": 2223 + }, + { + "epoch": 0.396011396011396, + "grad_norm": 0.5476976633071899, + "learning_rate": 0.00019527887847446595, + "loss": 1.2304, + "step": 2224 + }, + { + "epoch": 0.3961894586894587, + "grad_norm": 0.4487939774990082, + "learning_rate": 0.00019527462743433187, + "loss": 1.1813, + "step": 2225 + }, + { + "epoch": 0.39636752136752135, + "grad_norm": 0.4053241014480591, + "learning_rate": 0.00019527037452748936, + "loss": 0.7899, + "step": 2226 + }, + { + "epoch": 0.39654558404558404, + "grad_norm": 0.534570574760437, + "learning_rate": 0.00019526611975402176, + "loss": 1.0681, + "step": 2227 + }, + { + "epoch": 0.39672364672364674, + "grad_norm": 0.46096158027648926, + "learning_rate": 0.00019526186311401246, + "loss": 0.9234, + "step": 2228 + }, + { + "epoch": 0.3969017094017094, + "grad_norm": 0.47363516688346863, + "learning_rate": 0.00019525760460754483, + "loss": 1.0197, + "step": 2229 + }, + { + "epoch": 0.3970797720797721, + "grad_norm": 0.46317258477211, + "learning_rate": 0.00019525334423470234, + "loss": 1.2103, + "step": 2230 + }, + { + "epoch": 0.39725783475783477, + "grad_norm": 0.4924237132072449, + "learning_rate": 0.0001952490819955684, + "loss": 1.3299, + "step": 2231 + }, + { + "epoch": 0.3974358974358974, + "grad_norm": 0.5419978499412537, + "learning_rate": 0.0001952448178902266, + "loss": 1.2526, + "step": 2232 + }, + { + "epoch": 0.3976139601139601, + "grad_norm": 0.5003267526626587, + "learning_rate": 0.00019524055191876043, + "loss": 1.1073, + "step": 2233 + }, + { + "epoch": 0.3977920227920228, + "grad_norm": 0.621789276599884, + "learning_rate": 0.00019523628408125347, + "loss": 1.3409, + "step": 2234 + }, + { + "epoch": 0.39797008547008544, + "grad_norm": 0.44235602021217346, + "learning_rate": 0.0001952320143777894, + "loss": 0.9799, + "step": 2235 + }, + { + "epoch": 0.39814814814814814, + "grad_norm": 0.49954718351364136, + "learning_rate": 0.0001952277428084518, + "loss": 1.2227, + "step": 2236 + }, + { + "epoch": 0.39832621082621084, + "grad_norm": 0.5113739967346191, + "learning_rate": 0.00019522346937332443, + "loss": 1.1644, + "step": 2237 + }, + { + "epoch": 0.39850427350427353, + "grad_norm": 0.5026139616966248, + "learning_rate": 0.00019521919407249096, + "loss": 1.0823, + "step": 2238 + }, + { + "epoch": 0.39868233618233617, + "grad_norm": 0.4943205714225769, + "learning_rate": 0.0001952149169060352, + "loss": 1.0961, + "step": 2239 + }, + { + "epoch": 0.39886039886039887, + "grad_norm": 0.4680631458759308, + "learning_rate": 0.00019521063787404094, + "loss": 0.9787, + "step": 2240 + }, + { + "epoch": 0.39903846153846156, + "grad_norm": 0.5511566400527954, + "learning_rate": 0.00019520635697659202, + "loss": 1.2543, + "step": 2241 + }, + { + "epoch": 0.3992165242165242, + "grad_norm": 0.5494263172149658, + "learning_rate": 0.00019520207421377229, + "loss": 1.1978, + "step": 2242 + }, + { + "epoch": 0.3993945868945869, + "grad_norm": 0.4850340485572815, + "learning_rate": 0.00019519778958566568, + "loss": 0.8531, + "step": 2243 + }, + { + "epoch": 0.3995726495726496, + "grad_norm": 0.47168150544166565, + "learning_rate": 0.00019519350309235613, + "loss": 1.0746, + "step": 2244 + }, + { + "epoch": 0.39975071225071224, + "grad_norm": 0.571133553981781, + "learning_rate": 0.00019518921473392765, + "loss": 1.2984, + "step": 2245 + }, + { + "epoch": 0.39992877492877493, + "grad_norm": 0.4636089503765106, + "learning_rate": 0.00019518492451046427, + "loss": 1.019, + "step": 2246 + }, + { + "epoch": 0.40010683760683763, + "grad_norm": 0.4573518931865692, + "learning_rate": 0.00019518063242205, + "loss": 1.1042, + "step": 2247 + }, + { + "epoch": 0.40028490028490027, + "grad_norm": 0.49098989367485046, + "learning_rate": 0.00019517633846876894, + "loss": 1.1224, + "step": 2248 + }, + { + "epoch": 0.40046296296296297, + "grad_norm": 0.5475491881370544, + "learning_rate": 0.00019517204265070523, + "loss": 1.0984, + "step": 2249 + }, + { + "epoch": 0.40064102564102566, + "grad_norm": 0.45498281717300415, + "learning_rate": 0.00019516774496794307, + "loss": 0.8883, + "step": 2250 + }, + { + "epoch": 0.4008190883190883, + "grad_norm": 0.4908423125743866, + "learning_rate": 0.00019516344542056666, + "loss": 1.328, + "step": 2251 + }, + { + "epoch": 0.400997150997151, + "grad_norm": 0.5474920272827148, + "learning_rate": 0.0001951591440086602, + "loss": 1.3825, + "step": 2252 + }, + { + "epoch": 0.4011752136752137, + "grad_norm": 0.5165615081787109, + "learning_rate": 0.000195154840732308, + "loss": 1.33, + "step": 2253 + }, + { + "epoch": 0.40135327635327633, + "grad_norm": 0.5185585021972656, + "learning_rate": 0.00019515053559159435, + "loss": 1.1689, + "step": 2254 + }, + { + "epoch": 0.40153133903133903, + "grad_norm": 0.5468854904174805, + "learning_rate": 0.00019514622858660363, + "loss": 1.2708, + "step": 2255 + }, + { + "epoch": 0.4017094017094017, + "grad_norm": 0.47556906938552856, + "learning_rate": 0.0001951419197174202, + "loss": 1.0488, + "step": 2256 + }, + { + "epoch": 0.40188746438746437, + "grad_norm": 0.5521323084831238, + "learning_rate": 0.0001951376089841285, + "loss": 1.0868, + "step": 2257 + }, + { + "epoch": 0.40206552706552706, + "grad_norm": 0.6029638051986694, + "learning_rate": 0.00019513329638681296, + "loss": 1.1735, + "step": 2258 + }, + { + "epoch": 0.40224358974358976, + "grad_norm": 0.4897766411304474, + "learning_rate": 0.00019512898192555812, + "loss": 1.1687, + "step": 2259 + }, + { + "epoch": 0.4024216524216524, + "grad_norm": 0.45527184009552, + "learning_rate": 0.00019512466560044848, + "loss": 1.0352, + "step": 2260 + }, + { + "epoch": 0.4025997150997151, + "grad_norm": 0.5025625824928284, + "learning_rate": 0.00019512034741156863, + "loss": 1.2503, + "step": 2261 + }, + { + "epoch": 0.4027777777777778, + "grad_norm": 0.46415451169013977, + "learning_rate": 0.00019511602735900317, + "loss": 1.032, + "step": 2262 + }, + { + "epoch": 0.40295584045584043, + "grad_norm": 0.4812934398651123, + "learning_rate": 0.00019511170544283678, + "loss": 1.0523, + "step": 2263 + }, + { + "epoch": 0.4031339031339031, + "grad_norm": 0.49937039613723755, + "learning_rate": 0.00019510738166315404, + "loss": 1.2238, + "step": 2264 + }, + { + "epoch": 0.4033119658119658, + "grad_norm": 0.5428698062896729, + "learning_rate": 0.00019510305602003975, + "loss": 1.0361, + "step": 2265 + }, + { + "epoch": 0.40349002849002846, + "grad_norm": 0.44836854934692383, + "learning_rate": 0.0001950987285135786, + "loss": 1.169, + "step": 2266 + }, + { + "epoch": 0.40366809116809116, + "grad_norm": 0.5071489214897156, + "learning_rate": 0.00019509439914385549, + "loss": 1.1567, + "step": 2267 + }, + { + "epoch": 0.40384615384615385, + "grad_norm": 0.5204613208770752, + "learning_rate": 0.00019509006791095513, + "loss": 0.9949, + "step": 2268 + }, + { + "epoch": 0.40402421652421655, + "grad_norm": 0.4583234488964081, + "learning_rate": 0.00019508573481496238, + "loss": 0.9051, + "step": 2269 + }, + { + "epoch": 0.4042022792022792, + "grad_norm": 0.5436791181564331, + "learning_rate": 0.00019508139985596222, + "loss": 1.3239, + "step": 2270 + }, + { + "epoch": 0.4043803418803419, + "grad_norm": 0.48774269223213196, + "learning_rate": 0.00019507706303403954, + "loss": 1.2102, + "step": 2271 + }, + { + "epoch": 0.4045584045584046, + "grad_norm": 0.4742540717124939, + "learning_rate": 0.00019507272434927933, + "loss": 1.1137, + "step": 2272 + }, + { + "epoch": 0.4047364672364672, + "grad_norm": 0.531148374080658, + "learning_rate": 0.00019506838380176658, + "loss": 1.3162, + "step": 2273 + }, + { + "epoch": 0.4049145299145299, + "grad_norm": 0.5002314448356628, + "learning_rate": 0.0001950640413915863, + "loss": 1.0743, + "step": 2274 + }, + { + "epoch": 0.4050925925925926, + "grad_norm": 0.39826446771621704, + "learning_rate": 0.00019505969711882366, + "loss": 0.7698, + "step": 2275 + }, + { + "epoch": 0.40527065527065526, + "grad_norm": 0.5177471041679382, + "learning_rate": 0.00019505535098356371, + "loss": 1.1821, + "step": 2276 + }, + { + "epoch": 0.40544871794871795, + "grad_norm": 0.467241108417511, + "learning_rate": 0.00019505100298589158, + "loss": 0.8036, + "step": 2277 + }, + { + "epoch": 0.40562678062678065, + "grad_norm": 0.43711844086647034, + "learning_rate": 0.00019504665312589255, + "loss": 0.8667, + "step": 2278 + }, + { + "epoch": 0.4058048433048433, + "grad_norm": 0.4929116368293762, + "learning_rate": 0.00019504230140365177, + "loss": 1.1279, + "step": 2279 + }, + { + "epoch": 0.405982905982906, + "grad_norm": 0.5279183983802795, + "learning_rate": 0.00019503794781925452, + "loss": 1.1318, + "step": 2280 + }, + { + "epoch": 0.4061609686609687, + "grad_norm": 0.549217939376831, + "learning_rate": 0.00019503359237278608, + "loss": 1.2007, + "step": 2281 + }, + { + "epoch": 0.4063390313390313, + "grad_norm": 0.5485880374908447, + "learning_rate": 0.00019502923506433187, + "loss": 1.1079, + "step": 2282 + }, + { + "epoch": 0.406517094017094, + "grad_norm": 0.48379644751548767, + "learning_rate": 0.0001950248758939772, + "loss": 0.9978, + "step": 2283 + }, + { + "epoch": 0.4066951566951567, + "grad_norm": 0.5943657755851746, + "learning_rate": 0.00019502051486180744, + "loss": 1.0466, + "step": 2284 + }, + { + "epoch": 0.40687321937321935, + "grad_norm": 0.5721273422241211, + "learning_rate": 0.00019501615196790812, + "loss": 1.2674, + "step": 2285 + }, + { + "epoch": 0.40705128205128205, + "grad_norm": 0.47624221444129944, + "learning_rate": 0.00019501178721236464, + "loss": 1.089, + "step": 2286 + }, + { + "epoch": 0.40722934472934474, + "grad_norm": 0.5091297030448914, + "learning_rate": 0.0001950074205952626, + "loss": 1.2035, + "step": 2287 + }, + { + "epoch": 0.4074074074074074, + "grad_norm": 0.45206236839294434, + "learning_rate": 0.0001950030521166875, + "loss": 0.9188, + "step": 2288 + }, + { + "epoch": 0.4075854700854701, + "grad_norm": 0.5563844442367554, + "learning_rate": 0.00019499868177672497, + "loss": 1.3444, + "step": 2289 + }, + { + "epoch": 0.4077635327635328, + "grad_norm": 0.4971138536930084, + "learning_rate": 0.00019499430957546055, + "loss": 1.1615, + "step": 2290 + }, + { + "epoch": 0.4079415954415954, + "grad_norm": 0.49355944991111755, + "learning_rate": 0.00019498993551298, + "loss": 1.1528, + "step": 2291 + }, + { + "epoch": 0.4081196581196581, + "grad_norm": 0.534705638885498, + "learning_rate": 0.000194985559589369, + "loss": 1.197, + "step": 2292 + }, + { + "epoch": 0.4082977207977208, + "grad_norm": 0.5113020539283752, + "learning_rate": 0.0001949811818047133, + "loss": 1.109, + "step": 2293 + }, + { + "epoch": 0.40847578347578345, + "grad_norm": 0.4823366701602936, + "learning_rate": 0.00019497680215909858, + "loss": 1.168, + "step": 2294 + }, + { + "epoch": 0.40865384615384615, + "grad_norm": 0.500792920589447, + "learning_rate": 0.00019497242065261077, + "loss": 1.1567, + "step": 2295 + }, + { + "epoch": 0.40883190883190884, + "grad_norm": 0.5047918558120728, + "learning_rate": 0.00019496803728533566, + "loss": 1.0515, + "step": 2296 + }, + { + "epoch": 0.40900997150997154, + "grad_norm": 0.474624365568161, + "learning_rate": 0.00019496365205735913, + "loss": 1.1747, + "step": 2297 + }, + { + "epoch": 0.4091880341880342, + "grad_norm": 0.5522183179855347, + "learning_rate": 0.0001949592649687671, + "loss": 1.1506, + "step": 2298 + }, + { + "epoch": 0.4093660968660969, + "grad_norm": 0.4526083767414093, + "learning_rate": 0.00019495487601964553, + "loss": 0.9968, + "step": 2299 + }, + { + "epoch": 0.40954415954415957, + "grad_norm": 0.545845091342926, + "learning_rate": 0.00019495048521008044, + "loss": 1.146, + "step": 2300 + }, + { + "epoch": 0.4097222222222222, + "grad_norm": 0.5475544333457947, + "learning_rate": 0.00019494609254015784, + "loss": 1.0101, + "step": 2301 + }, + { + "epoch": 0.4099002849002849, + "grad_norm": 0.43419042229652405, + "learning_rate": 0.00019494169800996373, + "loss": 1.065, + "step": 2302 + }, + { + "epoch": 0.4100783475783476, + "grad_norm": 0.44998374581336975, + "learning_rate": 0.00019493730161958435, + "loss": 0.9948, + "step": 2303 + }, + { + "epoch": 0.41025641025641024, + "grad_norm": 0.5401661992073059, + "learning_rate": 0.0001949329033691057, + "loss": 1.0473, + "step": 2304 + }, + { + "epoch": 0.41043447293447294, + "grad_norm": 0.48064103722572327, + "learning_rate": 0.00019492850325861404, + "loss": 1.0486, + "step": 2305 + }, + { + "epoch": 0.41061253561253563, + "grad_norm": 0.5398300290107727, + "learning_rate": 0.00019492410128819557, + "loss": 1.0314, + "step": 2306 + }, + { + "epoch": 0.4107905982905983, + "grad_norm": 0.4771125912666321, + "learning_rate": 0.0001949196974579365, + "loss": 0.9855, + "step": 2307 + }, + { + "epoch": 0.41096866096866097, + "grad_norm": 0.5375809669494629, + "learning_rate": 0.00019491529176792315, + "loss": 1.0777, + "step": 2308 + }, + { + "epoch": 0.41114672364672367, + "grad_norm": 0.48424094915390015, + "learning_rate": 0.00019491088421824183, + "loss": 1.0751, + "step": 2309 + }, + { + "epoch": 0.4113247863247863, + "grad_norm": 0.5054880380630493, + "learning_rate": 0.00019490647480897887, + "loss": 1.2457, + "step": 2310 + }, + { + "epoch": 0.411502849002849, + "grad_norm": 0.47118356823921204, + "learning_rate": 0.0001949020635402207, + "loss": 1.0445, + "step": 2311 + }, + { + "epoch": 0.4116809116809117, + "grad_norm": 0.47171851992607117, + "learning_rate": 0.00019489765041205375, + "loss": 1.0062, + "step": 2312 + }, + { + "epoch": 0.41185897435897434, + "grad_norm": 0.5703238844871521, + "learning_rate": 0.00019489323542456447, + "loss": 1.5639, + "step": 2313 + }, + { + "epoch": 0.41203703703703703, + "grad_norm": 0.5045075416564941, + "learning_rate": 0.00019488881857783935, + "loss": 1.1665, + "step": 2314 + }, + { + "epoch": 0.41221509971509973, + "grad_norm": 0.46835362911224365, + "learning_rate": 0.00019488439987196495, + "loss": 1.2078, + "step": 2315 + }, + { + "epoch": 0.41239316239316237, + "grad_norm": 0.5187196731567383, + "learning_rate": 0.00019487997930702785, + "loss": 1.1049, + "step": 2316 + }, + { + "epoch": 0.41257122507122507, + "grad_norm": 0.5190554857254028, + "learning_rate": 0.00019487555688311463, + "loss": 1.331, + "step": 2317 + }, + { + "epoch": 0.41274928774928776, + "grad_norm": 0.7394969463348389, + "learning_rate": 0.00019487113260031197, + "loss": 0.9646, + "step": 2318 + }, + { + "epoch": 0.4129273504273504, + "grad_norm": 0.532982349395752, + "learning_rate": 0.00019486670645870656, + "loss": 1.166, + "step": 2319 + }, + { + "epoch": 0.4131054131054131, + "grad_norm": 0.48659515380859375, + "learning_rate": 0.00019486227845838509, + "loss": 1.0016, + "step": 2320 + }, + { + "epoch": 0.4132834757834758, + "grad_norm": 0.5364453196525574, + "learning_rate": 0.00019485784859943434, + "loss": 1.3877, + "step": 2321 + }, + { + "epoch": 0.41346153846153844, + "grad_norm": 0.49788740277290344, + "learning_rate": 0.0001948534168819411, + "loss": 1.2949, + "step": 2322 + }, + { + "epoch": 0.41363960113960113, + "grad_norm": 0.5125377774238586, + "learning_rate": 0.00019484898330599217, + "loss": 0.9769, + "step": 2323 + }, + { + "epoch": 0.41381766381766383, + "grad_norm": 0.5434861779212952, + "learning_rate": 0.00019484454787167447, + "loss": 1.254, + "step": 2324 + }, + { + "epoch": 0.41399572649572647, + "grad_norm": 0.5324583053588867, + "learning_rate": 0.00019484011057907487, + "loss": 0.9788, + "step": 2325 + }, + { + "epoch": 0.41417378917378916, + "grad_norm": 0.4806961715221405, + "learning_rate": 0.00019483567142828033, + "loss": 1.0089, + "step": 2326 + }, + { + "epoch": 0.41435185185185186, + "grad_norm": 0.5152947306632996, + "learning_rate": 0.0001948312304193778, + "loss": 1.15, + "step": 2327 + }, + { + "epoch": 0.41452991452991456, + "grad_norm": 0.6030138731002808, + "learning_rate": 0.0001948267875524543, + "loss": 1.196, + "step": 2328 + }, + { + "epoch": 0.4147079772079772, + "grad_norm": 0.4504946768283844, + "learning_rate": 0.0001948223428275969, + "loss": 0.8742, + "step": 2329 + }, + { + "epoch": 0.4148860398860399, + "grad_norm": 0.5195745825767517, + "learning_rate": 0.00019481789624489263, + "loss": 1.0104, + "step": 2330 + }, + { + "epoch": 0.4150641025641026, + "grad_norm": 0.5269250869750977, + "learning_rate": 0.0001948134478044287, + "loss": 1.2284, + "step": 2331 + }, + { + "epoch": 0.41524216524216523, + "grad_norm": 0.5302315354347229, + "learning_rate": 0.00019480899750629218, + "loss": 1.1374, + "step": 2332 + }, + { + "epoch": 0.4154202279202279, + "grad_norm": 0.5501471161842346, + "learning_rate": 0.0001948045453505703, + "loss": 1.214, + "step": 2333 + }, + { + "epoch": 0.4155982905982906, + "grad_norm": 0.4674588739871979, + "learning_rate": 0.0001948000913373503, + "loss": 1.0568, + "step": 2334 + }, + { + "epoch": 0.41577635327635326, + "grad_norm": 0.5262266993522644, + "learning_rate": 0.0001947956354667195, + "loss": 1.111, + "step": 2335 + }, + { + "epoch": 0.41595441595441596, + "grad_norm": 0.4549071788787842, + "learning_rate": 0.00019479117773876507, + "loss": 1.2655, + "step": 2336 + }, + { + "epoch": 0.41613247863247865, + "grad_norm": 0.48897311091423035, + "learning_rate": 0.00019478671815357447, + "loss": 1.0543, + "step": 2337 + }, + { + "epoch": 0.4163105413105413, + "grad_norm": 0.5544867515563965, + "learning_rate": 0.000194782256711235, + "loss": 1.2276, + "step": 2338 + }, + { + "epoch": 0.416488603988604, + "grad_norm": 0.5050773024559021, + "learning_rate": 0.0001947777934118341, + "loss": 0.9781, + "step": 2339 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 0.4831899106502533, + "learning_rate": 0.00019477332825545925, + "loss": 1.0213, + "step": 2340 + }, + { + "epoch": 0.4168447293447293, + "grad_norm": 0.5392552614212036, + "learning_rate": 0.0001947688612421979, + "loss": 1.3251, + "step": 2341 + }, + { + "epoch": 0.417022792022792, + "grad_norm": 0.5003608465194702, + "learning_rate": 0.00019476439237213754, + "loss": 1.0714, + "step": 2342 + }, + { + "epoch": 0.4172008547008547, + "grad_norm": 0.5016986727714539, + "learning_rate": 0.00019475992164536582, + "loss": 1.0656, + "step": 2343 + }, + { + "epoch": 0.41737891737891736, + "grad_norm": 0.5139234066009521, + "learning_rate": 0.00019475544906197024, + "loss": 1.1317, + "step": 2344 + }, + { + "epoch": 0.41755698005698005, + "grad_norm": 0.582478940486908, + "learning_rate": 0.00019475097462203847, + "loss": 1.4209, + "step": 2345 + }, + { + "epoch": 0.41773504273504275, + "grad_norm": 0.5248767137527466, + "learning_rate": 0.00019474649832565823, + "loss": 1.2965, + "step": 2346 + }, + { + "epoch": 0.4179131054131054, + "grad_norm": 0.4977390170097351, + "learning_rate": 0.00019474202017291713, + "loss": 1.3319, + "step": 2347 + }, + { + "epoch": 0.4180911680911681, + "grad_norm": 0.4868984818458557, + "learning_rate": 0.00019473754016390298, + "loss": 1.0595, + "step": 2348 + }, + { + "epoch": 0.4182692307692308, + "grad_norm": 0.5965346693992615, + "learning_rate": 0.00019473305829870353, + "loss": 1.2289, + "step": 2349 + }, + { + "epoch": 0.4184472934472934, + "grad_norm": 0.46590209007263184, + "learning_rate": 0.0001947285745774066, + "loss": 1.0468, + "step": 2350 + }, + { + "epoch": 0.4186253561253561, + "grad_norm": 0.497811883687973, + "learning_rate": 0.0001947240890001, + "loss": 1.1247, + "step": 2351 + }, + { + "epoch": 0.4188034188034188, + "grad_norm": 0.5348289012908936, + "learning_rate": 0.0001947196015668717, + "loss": 0.9496, + "step": 2352 + }, + { + "epoch": 0.41898148148148145, + "grad_norm": 0.5086174607276917, + "learning_rate": 0.0001947151122778095, + "loss": 0.8869, + "step": 2353 + }, + { + "epoch": 0.41915954415954415, + "grad_norm": 0.4844677150249481, + "learning_rate": 0.00019471062113300146, + "loss": 0.847, + "step": 2354 + }, + { + "epoch": 0.41933760683760685, + "grad_norm": 0.5395866632461548, + "learning_rate": 0.00019470612813253556, + "loss": 0.9684, + "step": 2355 + }, + { + "epoch": 0.41951566951566954, + "grad_norm": 0.479403018951416, + "learning_rate": 0.0001947016332764998, + "loss": 1.0532, + "step": 2356 + }, + { + "epoch": 0.4196937321937322, + "grad_norm": 0.5499961376190186, + "learning_rate": 0.00019469713656498227, + "loss": 1.2565, + "step": 2357 + }, + { + "epoch": 0.4198717948717949, + "grad_norm": 0.5865352153778076, + "learning_rate": 0.00019469263799807104, + "loss": 1.1349, + "step": 2358 + }, + { + "epoch": 0.4200498575498576, + "grad_norm": 0.4454309046268463, + "learning_rate": 0.00019468813757585432, + "loss": 0.9631, + "step": 2359 + }, + { + "epoch": 0.4202279202279202, + "grad_norm": 0.48426875472068787, + "learning_rate": 0.00019468363529842023, + "loss": 0.9795, + "step": 2360 + }, + { + "epoch": 0.4204059829059829, + "grad_norm": 0.47428226470947266, + "learning_rate": 0.00019467913116585697, + "loss": 0.9316, + "step": 2361 + }, + { + "epoch": 0.4205840455840456, + "grad_norm": 0.5193758010864258, + "learning_rate": 0.00019467462517825282, + "loss": 1.235, + "step": 2362 + }, + { + "epoch": 0.42076210826210825, + "grad_norm": 0.49845513701438904, + "learning_rate": 0.00019467011733569607, + "loss": 1.2413, + "step": 2363 + }, + { + "epoch": 0.42094017094017094, + "grad_norm": 0.45483845472335815, + "learning_rate": 0.00019466560763827502, + "loss": 1.2817, + "step": 2364 + }, + { + "epoch": 0.42111823361823364, + "grad_norm": 0.43345287442207336, + "learning_rate": 0.00019466109608607806, + "loss": 0.8568, + "step": 2365 + }, + { + "epoch": 0.4212962962962963, + "grad_norm": 0.4467088282108307, + "learning_rate": 0.00019465658267919352, + "loss": 1.1408, + "step": 2366 + }, + { + "epoch": 0.421474358974359, + "grad_norm": 0.6705610156059265, + "learning_rate": 0.00019465206741770992, + "loss": 1.445, + "step": 2367 + }, + { + "epoch": 0.42165242165242167, + "grad_norm": 0.5037859678268433, + "learning_rate": 0.00019464755030171565, + "loss": 0.8682, + "step": 2368 + }, + { + "epoch": 0.4218304843304843, + "grad_norm": 0.49576324224472046, + "learning_rate": 0.00019464303133129928, + "loss": 0.8387, + "step": 2369 + }, + { + "epoch": 0.422008547008547, + "grad_norm": 0.5222806334495544, + "learning_rate": 0.00019463851050654927, + "loss": 1.1443, + "step": 2370 + }, + { + "epoch": 0.4221866096866097, + "grad_norm": 0.4966863989830017, + "learning_rate": 0.00019463398782755426, + "loss": 1.1555, + "step": 2371 + }, + { + "epoch": 0.42236467236467234, + "grad_norm": 0.6140168309211731, + "learning_rate": 0.00019462946329440285, + "loss": 1.2264, + "step": 2372 + }, + { + "epoch": 0.42254273504273504, + "grad_norm": 0.4906651973724365, + "learning_rate": 0.0001946249369071837, + "loss": 1.2459, + "step": 2373 + }, + { + "epoch": 0.42272079772079774, + "grad_norm": 0.5956700444221497, + "learning_rate": 0.00019462040866598544, + "loss": 1.1521, + "step": 2374 + }, + { + "epoch": 0.4228988603988604, + "grad_norm": 0.46044886112213135, + "learning_rate": 0.00019461587857089687, + "loss": 1.2084, + "step": 2375 + }, + { + "epoch": 0.4230769230769231, + "grad_norm": 0.5109430551528931, + "learning_rate": 0.00019461134662200668, + "loss": 1.2684, + "step": 2376 + }, + { + "epoch": 0.42325498575498577, + "grad_norm": 0.4373733103275299, + "learning_rate": 0.0001946068128194037, + "loss": 1.0451, + "step": 2377 + }, + { + "epoch": 0.4234330484330484, + "grad_norm": 0.553817868232727, + "learning_rate": 0.00019460227716317673, + "loss": 1.1052, + "step": 2378 + }, + { + "epoch": 0.4236111111111111, + "grad_norm": 0.5742647647857666, + "learning_rate": 0.00019459773965341468, + "loss": 1.1647, + "step": 2379 + }, + { + "epoch": 0.4237891737891738, + "grad_norm": 0.5461940169334412, + "learning_rate": 0.00019459320029020642, + "loss": 1.0953, + "step": 2380 + }, + { + "epoch": 0.42396723646723644, + "grad_norm": 0.5837802290916443, + "learning_rate": 0.0001945886590736409, + "loss": 1.1303, + "step": 2381 + }, + { + "epoch": 0.42414529914529914, + "grad_norm": 0.5316985249519348, + "learning_rate": 0.0001945841160038071, + "loss": 1.1204, + "step": 2382 + }, + { + "epoch": 0.42432336182336183, + "grad_norm": 0.5846191048622131, + "learning_rate": 0.00019457957108079404, + "loss": 1.2622, + "step": 2383 + }, + { + "epoch": 0.42450142450142453, + "grad_norm": 0.43266957998275757, + "learning_rate": 0.00019457502430469075, + "loss": 0.9834, + "step": 2384 + }, + { + "epoch": 0.42467948717948717, + "grad_norm": 0.514081597328186, + "learning_rate": 0.00019457047567558632, + "loss": 0.8413, + "step": 2385 + }, + { + "epoch": 0.42485754985754987, + "grad_norm": 0.4831700325012207, + "learning_rate": 0.00019456592519356987, + "loss": 0.9244, + "step": 2386 + }, + { + "epoch": 0.42503561253561256, + "grad_norm": 0.5612850785255432, + "learning_rate": 0.00019456137285873057, + "loss": 0.9438, + "step": 2387 + }, + { + "epoch": 0.4252136752136752, + "grad_norm": 0.5197352766990662, + "learning_rate": 0.00019455681867115758, + "loss": 1.1095, + "step": 2388 + }, + { + "epoch": 0.4253917378917379, + "grad_norm": 0.5045261979103088, + "learning_rate": 0.00019455226263094018, + "loss": 1.0007, + "step": 2389 + }, + { + "epoch": 0.4255698005698006, + "grad_norm": 0.5167570114135742, + "learning_rate": 0.00019454770473816758, + "loss": 1.1335, + "step": 2390 + }, + { + "epoch": 0.42574786324786323, + "grad_norm": 0.49262070655822754, + "learning_rate": 0.00019454314499292913, + "loss": 1.0436, + "step": 2391 + }, + { + "epoch": 0.42592592592592593, + "grad_norm": 0.4489207863807678, + "learning_rate": 0.00019453858339531417, + "loss": 1.0138, + "step": 2392 + }, + { + "epoch": 0.4261039886039886, + "grad_norm": 0.6024920344352722, + "learning_rate": 0.00019453401994541203, + "loss": 1.1921, + "step": 2393 + }, + { + "epoch": 0.42628205128205127, + "grad_norm": 0.46807861328125, + "learning_rate": 0.00019452945464331215, + "loss": 1.0947, + "step": 2394 + }, + { + "epoch": 0.42646011396011396, + "grad_norm": 0.48776543140411377, + "learning_rate": 0.00019452488748910397, + "loss": 1.0029, + "step": 2395 + }, + { + "epoch": 0.42663817663817666, + "grad_norm": 0.4798663556575775, + "learning_rate": 0.000194520318482877, + "loss": 0.7863, + "step": 2396 + }, + { + "epoch": 0.4268162393162393, + "grad_norm": 0.5067816972732544, + "learning_rate": 0.0001945157476247207, + "loss": 1.0049, + "step": 2397 + }, + { + "epoch": 0.426994301994302, + "grad_norm": 0.5179638266563416, + "learning_rate": 0.00019451117491472468, + "loss": 1.1851, + "step": 2398 + }, + { + "epoch": 0.4271723646723647, + "grad_norm": 0.4782430827617645, + "learning_rate": 0.00019450660035297854, + "loss": 1.125, + "step": 2399 + }, + { + "epoch": 0.42735042735042733, + "grad_norm": 0.560077965259552, + "learning_rate": 0.00019450202393957186, + "loss": 1.1843, + "step": 2400 + }, + { + "epoch": 0.42752849002849, + "grad_norm": 0.5247970223426819, + "learning_rate": 0.00019449744567459436, + "loss": 1.1576, + "step": 2401 + }, + { + "epoch": 0.4277065527065527, + "grad_norm": 0.6414062976837158, + "learning_rate": 0.00019449286555813568, + "loss": 1.1833, + "step": 2402 + }, + { + "epoch": 0.42788461538461536, + "grad_norm": 0.5006586909294128, + "learning_rate": 0.00019448828359028563, + "loss": 1.1778, + "step": 2403 + }, + { + "epoch": 0.42806267806267806, + "grad_norm": 0.4946450889110565, + "learning_rate": 0.0001944836997711339, + "loss": 1.1611, + "step": 2404 + }, + { + "epoch": 0.42824074074074076, + "grad_norm": 0.4601200222969055, + "learning_rate": 0.00019447911410077037, + "loss": 1.2456, + "step": 2405 + }, + { + "epoch": 0.4284188034188034, + "grad_norm": 0.4653947651386261, + "learning_rate": 0.00019447452657928485, + "loss": 1.0941, + "step": 2406 + }, + { + "epoch": 0.4285968660968661, + "grad_norm": 0.5015713572502136, + "learning_rate": 0.00019446993720676726, + "loss": 1.3113, + "step": 2407 + }, + { + "epoch": 0.4287749287749288, + "grad_norm": 0.5803143978118896, + "learning_rate": 0.0001944653459833075, + "loss": 1.0568, + "step": 2408 + }, + { + "epoch": 0.42895299145299143, + "grad_norm": 0.5259647965431213, + "learning_rate": 0.0001944607529089955, + "loss": 1.1243, + "step": 2409 + }, + { + "epoch": 0.4291310541310541, + "grad_norm": 0.5150414109230042, + "learning_rate": 0.00019445615798392124, + "loss": 1.0676, + "step": 2410 + }, + { + "epoch": 0.4293091168091168, + "grad_norm": 0.5848649740219116, + "learning_rate": 0.0001944515612081748, + "loss": 1.0671, + "step": 2411 + }, + { + "epoch": 0.42948717948717946, + "grad_norm": 0.5696990489959717, + "learning_rate": 0.00019444696258184626, + "loss": 1.3323, + "step": 2412 + }, + { + "epoch": 0.42966524216524216, + "grad_norm": 0.49822330474853516, + "learning_rate": 0.00019444236210502567, + "loss": 1.1004, + "step": 2413 + }, + { + "epoch": 0.42984330484330485, + "grad_norm": 0.4683490991592407, + "learning_rate": 0.00019443775977780317, + "loss": 0.9768, + "step": 2414 + }, + { + "epoch": 0.43002136752136755, + "grad_norm": 0.5703811049461365, + "learning_rate": 0.00019443315560026893, + "loss": 1.154, + "step": 2415 + }, + { + "epoch": 0.4301994301994302, + "grad_norm": 0.5121861100196838, + "learning_rate": 0.0001944285495725132, + "loss": 1.1388, + "step": 2416 + }, + { + "epoch": 0.4303774928774929, + "grad_norm": 0.4864094853401184, + "learning_rate": 0.00019442394169462619, + "loss": 0.9214, + "step": 2417 + }, + { + "epoch": 0.4305555555555556, + "grad_norm": 0.5234864354133606, + "learning_rate": 0.0001944193319666982, + "loss": 1.2787, + "step": 2418 + }, + { + "epoch": 0.4307336182336182, + "grad_norm": 0.5137650370597839, + "learning_rate": 0.00019441472038881955, + "loss": 1.1406, + "step": 2419 + }, + { + "epoch": 0.4309116809116809, + "grad_norm": 0.49687784910202026, + "learning_rate": 0.00019441010696108054, + "loss": 0.93, + "step": 2420 + }, + { + "epoch": 0.4310897435897436, + "grad_norm": 0.5078722834587097, + "learning_rate": 0.00019440549168357163, + "loss": 1.1417, + "step": 2421 + }, + { + "epoch": 0.43126780626780625, + "grad_norm": 0.4483391046524048, + "learning_rate": 0.00019440087455638324, + "loss": 0.9016, + "step": 2422 + }, + { + "epoch": 0.43144586894586895, + "grad_norm": 0.5963045954704285, + "learning_rate": 0.00019439625557960576, + "loss": 1.1567, + "step": 2423 + }, + { + "epoch": 0.43162393162393164, + "grad_norm": 0.5534471273422241, + "learning_rate": 0.0001943916347533298, + "loss": 1.1409, + "step": 2424 + }, + { + "epoch": 0.4318019943019943, + "grad_norm": 0.6400241851806641, + "learning_rate": 0.0001943870120776458, + "loss": 1.2041, + "step": 2425 + }, + { + "epoch": 0.431980056980057, + "grad_norm": 0.4599420726299286, + "learning_rate": 0.0001943823875526444, + "loss": 1.023, + "step": 2426 + }, + { + "epoch": 0.4321581196581197, + "grad_norm": 0.4799708425998688, + "learning_rate": 0.00019437776117841614, + "loss": 1.0872, + "step": 2427 + }, + { + "epoch": 0.4323361823361823, + "grad_norm": 0.5138532519340515, + "learning_rate": 0.00019437313295505172, + "loss": 1.1175, + "step": 2428 + }, + { + "epoch": 0.432514245014245, + "grad_norm": 0.538223147392273, + "learning_rate": 0.00019436850288264183, + "loss": 1.1203, + "step": 2429 + }, + { + "epoch": 0.4326923076923077, + "grad_norm": 0.458044171333313, + "learning_rate": 0.00019436387096127713, + "loss": 1.0383, + "step": 2430 + }, + { + "epoch": 0.43287037037037035, + "grad_norm": 0.5928303599357605, + "learning_rate": 0.00019435923719104842, + "loss": 1.1191, + "step": 2431 + }, + { + "epoch": 0.43304843304843305, + "grad_norm": 0.5818437933921814, + "learning_rate": 0.00019435460157204645, + "loss": 1.0352, + "step": 2432 + }, + { + "epoch": 0.43322649572649574, + "grad_norm": 0.487341046333313, + "learning_rate": 0.0001943499641043621, + "loss": 1.2608, + "step": 2433 + }, + { + "epoch": 0.4334045584045584, + "grad_norm": 0.4737292230129242, + "learning_rate": 0.0001943453247880862, + "loss": 1.0084, + "step": 2434 + }, + { + "epoch": 0.4335826210826211, + "grad_norm": 0.4251207709312439, + "learning_rate": 0.0001943406836233096, + "loss": 0.9163, + "step": 2435 + }, + { + "epoch": 0.4337606837606838, + "grad_norm": 0.49468478560447693, + "learning_rate": 0.00019433604061012331, + "loss": 1.0293, + "step": 2436 + }, + { + "epoch": 0.4339387464387464, + "grad_norm": 0.47120022773742676, + "learning_rate": 0.00019433139574861826, + "loss": 1.0097, + "step": 2437 + }, + { + "epoch": 0.4341168091168091, + "grad_norm": 0.5060358047485352, + "learning_rate": 0.00019432674903888548, + "loss": 1.0683, + "step": 2438 + }, + { + "epoch": 0.4342948717948718, + "grad_norm": 0.5455917119979858, + "learning_rate": 0.00019432210048101598, + "loss": 0.8886, + "step": 2439 + }, + { + "epoch": 0.43447293447293445, + "grad_norm": 0.7960546612739563, + "learning_rate": 0.00019431745007510086, + "loss": 0.8648, + "step": 2440 + }, + { + "epoch": 0.43465099715099714, + "grad_norm": 0.5069689154624939, + "learning_rate": 0.00019431279782123126, + "loss": 1.1315, + "step": 2441 + }, + { + "epoch": 0.43482905982905984, + "grad_norm": 0.5597776174545288, + "learning_rate": 0.0001943081437194983, + "loss": 1.2281, + "step": 2442 + }, + { + "epoch": 0.43500712250712253, + "grad_norm": 0.4527420997619629, + "learning_rate": 0.00019430348776999315, + "loss": 0.7576, + "step": 2443 + }, + { + "epoch": 0.4351851851851852, + "grad_norm": 0.5625936388969421, + "learning_rate": 0.00019429882997280706, + "loss": 1.0302, + "step": 2444 + }, + { + "epoch": 0.43536324786324787, + "grad_norm": 0.5173513293266296, + "learning_rate": 0.0001942941703280313, + "loss": 1.2255, + "step": 2445 + }, + { + "epoch": 0.43554131054131057, + "grad_norm": 0.45889151096343994, + "learning_rate": 0.00019428950883575714, + "loss": 0.9322, + "step": 2446 + }, + { + "epoch": 0.4357193732193732, + "grad_norm": 0.5288477540016174, + "learning_rate": 0.00019428484549607593, + "loss": 1.0572, + "step": 2447 + }, + { + "epoch": 0.4358974358974359, + "grad_norm": 0.48328033089637756, + "learning_rate": 0.00019428018030907902, + "loss": 1.1213, + "step": 2448 + }, + { + "epoch": 0.4360754985754986, + "grad_norm": 0.5146737098693848, + "learning_rate": 0.00019427551327485786, + "loss": 0.9633, + "step": 2449 + }, + { + "epoch": 0.43625356125356124, + "grad_norm": 0.5138360261917114, + "learning_rate": 0.00019427084439350382, + "loss": 1.0561, + "step": 2450 + }, + { + "epoch": 0.43643162393162394, + "grad_norm": 0.5192533135414124, + "learning_rate": 0.00019426617366510843, + "loss": 1.1704, + "step": 2451 + }, + { + "epoch": 0.43660968660968663, + "grad_norm": 0.4819495379924774, + "learning_rate": 0.00019426150108976318, + "loss": 1.0958, + "step": 2452 + }, + { + "epoch": 0.43678774928774927, + "grad_norm": 0.4626680910587311, + "learning_rate": 0.00019425682666755965, + "loss": 1.1872, + "step": 2453 + }, + { + "epoch": 0.43696581196581197, + "grad_norm": 0.5773931741714478, + "learning_rate": 0.00019425215039858937, + "loss": 1.0722, + "step": 2454 + }, + { + "epoch": 0.43714387464387466, + "grad_norm": 0.5003872513771057, + "learning_rate": 0.00019424747228294402, + "loss": 1.0561, + "step": 2455 + }, + { + "epoch": 0.4373219373219373, + "grad_norm": 0.47370314598083496, + "learning_rate": 0.0001942427923207152, + "loss": 1.1619, + "step": 2456 + }, + { + "epoch": 0.4375, + "grad_norm": 0.466421514749527, + "learning_rate": 0.00019423811051199466, + "loss": 1.1311, + "step": 2457 + }, + { + "epoch": 0.4376780626780627, + "grad_norm": 0.44564682245254517, + "learning_rate": 0.00019423342685687413, + "loss": 1.1889, + "step": 2458 + }, + { + "epoch": 0.43785612535612534, + "grad_norm": 0.40986698865890503, + "learning_rate": 0.00019422874135544533, + "loss": 0.7312, + "step": 2459 + }, + { + "epoch": 0.43803418803418803, + "grad_norm": 0.4714358448982239, + "learning_rate": 0.0001942240540078001, + "loss": 0.9273, + "step": 2460 + }, + { + "epoch": 0.43821225071225073, + "grad_norm": 0.5298398733139038, + "learning_rate": 0.00019421936481403025, + "loss": 1.3377, + "step": 2461 + }, + { + "epoch": 0.43839031339031337, + "grad_norm": 0.6326695680618286, + "learning_rate": 0.0001942146737742277, + "loss": 1.0258, + "step": 2462 + }, + { + "epoch": 0.43856837606837606, + "grad_norm": 0.5087653994560242, + "learning_rate": 0.00019420998088848427, + "loss": 1.0007, + "step": 2463 + }, + { + "epoch": 0.43874643874643876, + "grad_norm": 0.4895429313182831, + "learning_rate": 0.00019420528615689202, + "loss": 1.0032, + "step": 2464 + }, + { + "epoch": 0.4389245014245014, + "grad_norm": 0.5029937028884888, + "learning_rate": 0.00019420058957954285, + "loss": 1.2877, + "step": 2465 + }, + { + "epoch": 0.4391025641025641, + "grad_norm": 0.4953192174434662, + "learning_rate": 0.00019419589115652884, + "loss": 1.0759, + "step": 2466 + }, + { + "epoch": 0.4392806267806268, + "grad_norm": 0.5081778168678284, + "learning_rate": 0.000194191190887942, + "loss": 0.8816, + "step": 2467 + }, + { + "epoch": 0.43945868945868943, + "grad_norm": 0.5065913200378418, + "learning_rate": 0.00019418648877387446, + "loss": 1.0362, + "step": 2468 + }, + { + "epoch": 0.43963675213675213, + "grad_norm": 0.540600061416626, + "learning_rate": 0.00019418178481441832, + "loss": 1.0911, + "step": 2469 + }, + { + "epoch": 0.4398148148148148, + "grad_norm": 0.5122954845428467, + "learning_rate": 0.00019417707900966572, + "loss": 0.9866, + "step": 2470 + }, + { + "epoch": 0.43999287749287747, + "grad_norm": 0.5380190014839172, + "learning_rate": 0.00019417237135970893, + "loss": 1.2775, + "step": 2471 + }, + { + "epoch": 0.44017094017094016, + "grad_norm": 1.2977570295333862, + "learning_rate": 0.00019416766186464016, + "loss": 1.3993, + "step": 2472 + }, + { + "epoch": 0.44034900284900286, + "grad_norm": 0.48105308413505554, + "learning_rate": 0.00019416295052455165, + "loss": 0.9369, + "step": 2473 + }, + { + "epoch": 0.44052706552706555, + "grad_norm": 0.4742157459259033, + "learning_rate": 0.00019415823733953574, + "loss": 1.101, + "step": 2474 + }, + { + "epoch": 0.4407051282051282, + "grad_norm": 0.4958631694316864, + "learning_rate": 0.00019415352230968473, + "loss": 0.9906, + "step": 2475 + }, + { + "epoch": 0.4408831908831909, + "grad_norm": 0.5808146595954895, + "learning_rate": 0.00019414880543509107, + "loss": 1.2315, + "step": 2476 + }, + { + "epoch": 0.4410612535612536, + "grad_norm": 0.4294755160808563, + "learning_rate": 0.00019414408671584714, + "loss": 0.8275, + "step": 2477 + }, + { + "epoch": 0.4412393162393162, + "grad_norm": 0.5346055626869202, + "learning_rate": 0.0001941393661520454, + "loss": 1.2432, + "step": 2478 + }, + { + "epoch": 0.4414173789173789, + "grad_norm": 0.5827590227127075, + "learning_rate": 0.00019413464374377833, + "loss": 1.3204, + "step": 2479 + }, + { + "epoch": 0.4415954415954416, + "grad_norm": 0.45688143372535706, + "learning_rate": 0.00019412991949113847, + "loss": 0.9307, + "step": 2480 + }, + { + "epoch": 0.44177350427350426, + "grad_norm": 0.512999415397644, + "learning_rate": 0.0001941251933942184, + "loss": 1.2808, + "step": 2481 + }, + { + "epoch": 0.44195156695156695, + "grad_norm": 0.4546334445476532, + "learning_rate": 0.00019412046545311064, + "loss": 1.0156, + "step": 2482 + }, + { + "epoch": 0.44212962962962965, + "grad_norm": 0.48552581667900085, + "learning_rate": 0.00019411573566790793, + "loss": 1.3798, + "step": 2483 + }, + { + "epoch": 0.4423076923076923, + "grad_norm": 0.511970579624176, + "learning_rate": 0.00019411100403870287, + "loss": 1.065, + "step": 2484 + }, + { + "epoch": 0.442485754985755, + "grad_norm": 0.6367824077606201, + "learning_rate": 0.00019410627056558815, + "loss": 1.3242, + "step": 2485 + }, + { + "epoch": 0.4426638176638177, + "grad_norm": 0.48913368582725525, + "learning_rate": 0.00019410153524865659, + "loss": 0.9761, + "step": 2486 + }, + { + "epoch": 0.4428418803418803, + "grad_norm": 0.5077710151672363, + "learning_rate": 0.0001940967980880009, + "loss": 1.1023, + "step": 2487 + }, + { + "epoch": 0.443019943019943, + "grad_norm": 0.4956335723400116, + "learning_rate": 0.00019409205908371395, + "loss": 1.1788, + "step": 2488 + }, + { + "epoch": 0.4431980056980057, + "grad_norm": 0.4726616442203522, + "learning_rate": 0.00019408731823588853, + "loss": 1.1445, + "step": 2489 + }, + { + "epoch": 0.44337606837606836, + "grad_norm": 0.5676438212394714, + "learning_rate": 0.00019408257554461757, + "loss": 1.0344, + "step": 2490 + }, + { + "epoch": 0.44355413105413105, + "grad_norm": 0.537656843662262, + "learning_rate": 0.000194077831009994, + "loss": 0.9876, + "step": 2491 + }, + { + "epoch": 0.44373219373219375, + "grad_norm": 0.517905592918396, + "learning_rate": 0.00019407308463211074, + "loss": 1.1389, + "step": 2492 + }, + { + "epoch": 0.4439102564102564, + "grad_norm": 0.49227026104927063, + "learning_rate": 0.0001940683364110608, + "loss": 1.0351, + "step": 2493 + }, + { + "epoch": 0.4440883190883191, + "grad_norm": 0.5131173729896545, + "learning_rate": 0.00019406358634693725, + "loss": 1.0351, + "step": 2494 + }, + { + "epoch": 0.4442663817663818, + "grad_norm": 0.5064495205879211, + "learning_rate": 0.0001940588344398331, + "loss": 1.0248, + "step": 2495 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 0.44107526540756226, + "learning_rate": 0.00019405408068984148, + "loss": 0.8068, + "step": 2496 + }, + { + "epoch": 0.4446225071225071, + "grad_norm": 0.6711848378181458, + "learning_rate": 0.00019404932509705554, + "loss": 1.059, + "step": 2497 + }, + { + "epoch": 0.4448005698005698, + "grad_norm": 0.5862596035003662, + "learning_rate": 0.00019404456766156845, + "loss": 1.2012, + "step": 2498 + }, + { + "epoch": 0.44497863247863245, + "grad_norm": 0.5528512001037598, + "learning_rate": 0.0001940398083834734, + "loss": 1.1121, + "step": 2499 + }, + { + "epoch": 0.44515669515669515, + "grad_norm": 0.5326655507087708, + "learning_rate": 0.0001940350472628637, + "loss": 1.166, + "step": 2500 + }, + { + "epoch": 0.44533475783475784, + "grad_norm": 0.5384873747825623, + "learning_rate": 0.00019403028429983252, + "loss": 1.4111, + "step": 2501 + }, + { + "epoch": 0.44551282051282054, + "grad_norm": 0.5142310857772827, + "learning_rate": 0.0001940255194944733, + "loss": 1.3353, + "step": 2502 + }, + { + "epoch": 0.4456908831908832, + "grad_norm": 0.49124231934547424, + "learning_rate": 0.0001940207528468793, + "loss": 1.1443, + "step": 2503 + }, + { + "epoch": 0.4458689458689459, + "grad_norm": 0.509713888168335, + "learning_rate": 0.000194015984357144, + "loss": 1.1857, + "step": 2504 + }, + { + "epoch": 0.44604700854700857, + "grad_norm": 0.5211275219917297, + "learning_rate": 0.00019401121402536078, + "loss": 0.9911, + "step": 2505 + }, + { + "epoch": 0.4462250712250712, + "grad_norm": 0.480340838432312, + "learning_rate": 0.00019400644185162312, + "loss": 1.1018, + "step": 2506 + }, + { + "epoch": 0.4464031339031339, + "grad_norm": 0.4212559163570404, + "learning_rate": 0.00019400166783602448, + "loss": 0.7501, + "step": 2507 + }, + { + "epoch": 0.4465811965811966, + "grad_norm": 0.5110511183738708, + "learning_rate": 0.00019399689197865846, + "loss": 1.1244, + "step": 2508 + }, + { + "epoch": 0.44675925925925924, + "grad_norm": 0.5604230165481567, + "learning_rate": 0.0001939921142796186, + "loss": 1.1066, + "step": 2509 + }, + { + "epoch": 0.44693732193732194, + "grad_norm": 0.5578675270080566, + "learning_rate": 0.0001939873347389985, + "loss": 1.0514, + "step": 2510 + }, + { + "epoch": 0.44711538461538464, + "grad_norm": 0.520908772945404, + "learning_rate": 0.00019398255335689184, + "loss": 1.1217, + "step": 2511 + }, + { + "epoch": 0.4472934472934473, + "grad_norm": 0.4405131936073303, + "learning_rate": 0.00019397777013339224, + "loss": 1.043, + "step": 2512 + }, + { + "epoch": 0.44747150997151, + "grad_norm": 0.5217751860618591, + "learning_rate": 0.0001939729850685935, + "loss": 1.1301, + "step": 2513 + }, + { + "epoch": 0.44764957264957267, + "grad_norm": 0.6151493191719055, + "learning_rate": 0.00019396819816258932, + "loss": 1.3498, + "step": 2514 + }, + { + "epoch": 0.4478276353276353, + "grad_norm": 0.5622836947441101, + "learning_rate": 0.0001939634094154735, + "loss": 1.146, + "step": 2515 + }, + { + "epoch": 0.448005698005698, + "grad_norm": 0.4671688973903656, + "learning_rate": 0.00019395861882733984, + "loss": 0.9456, + "step": 2516 + }, + { + "epoch": 0.4481837606837607, + "grad_norm": 0.453951358795166, + "learning_rate": 0.00019395382639828223, + "loss": 1.0042, + "step": 2517 + }, + { + "epoch": 0.44836182336182334, + "grad_norm": 0.5150699615478516, + "learning_rate": 0.0001939490321283946, + "loss": 1.166, + "step": 2518 + }, + { + "epoch": 0.44853988603988604, + "grad_norm": 0.5718298554420471, + "learning_rate": 0.0001939442360177708, + "loss": 1.2033, + "step": 2519 + }, + { + "epoch": 0.44871794871794873, + "grad_norm": 0.5306782126426697, + "learning_rate": 0.00019393943806650488, + "loss": 1.0765, + "step": 2520 + }, + { + "epoch": 0.4488960113960114, + "grad_norm": 0.47633033990859985, + "learning_rate": 0.0001939346382746908, + "loss": 0.9957, + "step": 2521 + }, + { + "epoch": 0.44907407407407407, + "grad_norm": 0.496441513299942, + "learning_rate": 0.00019392983664242262, + "loss": 1.2016, + "step": 2522 + }, + { + "epoch": 0.44925213675213677, + "grad_norm": 0.45956477522850037, + "learning_rate": 0.00019392503316979442, + "loss": 1.026, + "step": 2523 + }, + { + "epoch": 0.4494301994301994, + "grad_norm": 0.5400575995445251, + "learning_rate": 0.0001939202278569003, + "loss": 1.0785, + "step": 2524 + }, + { + "epoch": 0.4496082621082621, + "grad_norm": 0.4847868084907532, + "learning_rate": 0.00019391542070383442, + "loss": 1.013, + "step": 2525 + }, + { + "epoch": 0.4497863247863248, + "grad_norm": 0.4694063663482666, + "learning_rate": 0.00019391061171069094, + "loss": 0.8793, + "step": 2526 + }, + { + "epoch": 0.44996438746438744, + "grad_norm": 0.5158169269561768, + "learning_rate": 0.00019390580087756413, + "loss": 0.9602, + "step": 2527 + }, + { + "epoch": 0.45014245014245013, + "grad_norm": 0.5404585003852844, + "learning_rate": 0.00019390098820454822, + "loss": 1.2247, + "step": 2528 + }, + { + "epoch": 0.45032051282051283, + "grad_norm": 0.5302738547325134, + "learning_rate": 0.00019389617369173752, + "loss": 0.918, + "step": 2529 + }, + { + "epoch": 0.45049857549857547, + "grad_norm": 0.5065485835075378, + "learning_rate": 0.00019389135733922634, + "loss": 1.0934, + "step": 2530 + }, + { + "epoch": 0.45067663817663817, + "grad_norm": 0.5491471886634827, + "learning_rate": 0.00019388653914710903, + "loss": 1.0736, + "step": 2531 + }, + { + "epoch": 0.45085470085470086, + "grad_norm": 0.4850206971168518, + "learning_rate": 0.00019388171911548005, + "loss": 1.2401, + "step": 2532 + }, + { + "epoch": 0.45103276353276356, + "grad_norm": 0.5419789552688599, + "learning_rate": 0.0001938768972444338, + "loss": 1.269, + "step": 2533 + }, + { + "epoch": 0.4512108262108262, + "grad_norm": 0.4209023714065552, + "learning_rate": 0.00019387207353406476, + "loss": 1.0544, + "step": 2534 + }, + { + "epoch": 0.4513888888888889, + "grad_norm": 0.578588604927063, + "learning_rate": 0.00019386724798446743, + "loss": 1.0564, + "step": 2535 + }, + { + "epoch": 0.4515669515669516, + "grad_norm": 0.5277524590492249, + "learning_rate": 0.00019386242059573638, + "loss": 1.1497, + "step": 2536 + }, + { + "epoch": 0.45174501424501423, + "grad_norm": 0.5536073446273804, + "learning_rate": 0.0001938575913679662, + "loss": 1.2213, + "step": 2537 + }, + { + "epoch": 0.4519230769230769, + "grad_norm": 0.5572254657745361, + "learning_rate": 0.00019385276030125143, + "loss": 1.0231, + "step": 2538 + }, + { + "epoch": 0.4521011396011396, + "grad_norm": 0.493847131729126, + "learning_rate": 0.00019384792739568686, + "loss": 0.9385, + "step": 2539 + }, + { + "epoch": 0.45227920227920226, + "grad_norm": 0.4641396403312683, + "learning_rate": 0.00019384309265136707, + "loss": 0.9332, + "step": 2540 + }, + { + "epoch": 0.45245726495726496, + "grad_norm": 0.5439442992210388, + "learning_rate": 0.00019383825606838681, + "loss": 1.317, + "step": 2541 + }, + { + "epoch": 0.45263532763532766, + "grad_norm": 0.7050970792770386, + "learning_rate": 0.00019383341764684086, + "loss": 0.9508, + "step": 2542 + }, + { + "epoch": 0.4528133903133903, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.000193828577386824, + "loss": 1.2704, + "step": 2543 + }, + { + "epoch": 0.452991452991453, + "grad_norm": 0.47641924023628235, + "learning_rate": 0.0001938237352884311, + "loss": 1.0101, + "step": 2544 + }, + { + "epoch": 0.4531695156695157, + "grad_norm": 0.5223637819290161, + "learning_rate": 0.000193818891351757, + "loss": 1.0548, + "step": 2545 + }, + { + "epoch": 0.45334757834757833, + "grad_norm": 0.49065667390823364, + "learning_rate": 0.0001938140455768966, + "loss": 1.0927, + "step": 2546 + }, + { + "epoch": 0.453525641025641, + "grad_norm": 0.4808312654495239, + "learning_rate": 0.0001938091979639449, + "loss": 1.0599, + "step": 2547 + }, + { + "epoch": 0.4537037037037037, + "grad_norm": 0.5157489776611328, + "learning_rate": 0.0001938043485129968, + "loss": 1.2596, + "step": 2548 + }, + { + "epoch": 0.45388176638176636, + "grad_norm": 0.5983387231826782, + "learning_rate": 0.0001937994972241474, + "loss": 1.2276, + "step": 2549 + }, + { + "epoch": 0.45405982905982906, + "grad_norm": 0.49776506423950195, + "learning_rate": 0.00019379464409749163, + "loss": 1.3666, + "step": 2550 + }, + { + "epoch": 0.45423789173789175, + "grad_norm": 0.4693490266799927, + "learning_rate": 0.00019378978913312471, + "loss": 1.087, + "step": 2551 + }, + { + "epoch": 0.4544159544159544, + "grad_norm": 0.4754335880279541, + "learning_rate": 0.00019378493233114167, + "loss": 1.1282, + "step": 2552 + }, + { + "epoch": 0.4545940170940171, + "grad_norm": 0.5852862000465393, + "learning_rate": 0.00019378007369163776, + "loss": 1.1113, + "step": 2553 + }, + { + "epoch": 0.4547720797720798, + "grad_norm": 0.47442635893821716, + "learning_rate": 0.00019377521321470805, + "loss": 0.983, + "step": 2554 + }, + { + "epoch": 0.4549501424501424, + "grad_norm": 0.47432273626327515, + "learning_rate": 0.00019377035090044787, + "loss": 1.0169, + "step": 2555 + }, + { + "epoch": 0.4551282051282051, + "grad_norm": 0.4929196834564209, + "learning_rate": 0.00019376548674895246, + "loss": 1.0182, + "step": 2556 + }, + { + "epoch": 0.4553062678062678, + "grad_norm": 0.5433184504508972, + "learning_rate": 0.00019376062076031708, + "loss": 1.1339, + "step": 2557 + }, + { + "epoch": 0.45548433048433046, + "grad_norm": 0.47430408000946045, + "learning_rate": 0.00019375575293463715, + "loss": 1.1589, + "step": 2558 + }, + { + "epoch": 0.45566239316239315, + "grad_norm": 0.46641045808792114, + "learning_rate": 0.000193750883272008, + "loss": 1.029, + "step": 2559 + }, + { + "epoch": 0.45584045584045585, + "grad_norm": 0.44476228952407837, + "learning_rate": 0.00019374601177252502, + "loss": 0.8494, + "step": 2560 + }, + { + "epoch": 0.45601851851851855, + "grad_norm": 0.4886183440685272, + "learning_rate": 0.00019374113843628366, + "loss": 1.1374, + "step": 2561 + }, + { + "epoch": 0.4561965811965812, + "grad_norm": 0.4786703288555145, + "learning_rate": 0.00019373626326337946, + "loss": 1.2861, + "step": 2562 + }, + { + "epoch": 0.4563746438746439, + "grad_norm": 0.5752716660499573, + "learning_rate": 0.0001937313862539079, + "loss": 1.2365, + "step": 2563 + }, + { + "epoch": 0.4565527065527066, + "grad_norm": 0.519176185131073, + "learning_rate": 0.00019372650740796452, + "loss": 1.2264, + "step": 2564 + }, + { + "epoch": 0.4567307692307692, + "grad_norm": 0.5927292704582214, + "learning_rate": 0.00019372162672564493, + "loss": 0.8979, + "step": 2565 + }, + { + "epoch": 0.4569088319088319, + "grad_norm": 0.5467435121536255, + "learning_rate": 0.00019371674420704478, + "loss": 1.1016, + "step": 2566 + }, + { + "epoch": 0.4570868945868946, + "grad_norm": 0.49593284726142883, + "learning_rate": 0.00019371185985225968, + "loss": 0.982, + "step": 2567 + }, + { + "epoch": 0.45726495726495725, + "grad_norm": 0.5696587562561035, + "learning_rate": 0.00019370697366138538, + "loss": 0.979, + "step": 2568 + }, + { + "epoch": 0.45744301994301995, + "grad_norm": 0.4455752968788147, + "learning_rate": 0.00019370208563451757, + "loss": 0.8832, + "step": 2569 + }, + { + "epoch": 0.45762108262108264, + "grad_norm": 0.5072923302650452, + "learning_rate": 0.00019369719577175203, + "loss": 1.1046, + "step": 2570 + }, + { + "epoch": 0.4577991452991453, + "grad_norm": 0.45119982957839966, + "learning_rate": 0.0001936923040731846, + "loss": 1.0083, + "step": 2571 + }, + { + "epoch": 0.457977207977208, + "grad_norm": 0.5062251091003418, + "learning_rate": 0.00019368741053891108, + "loss": 1.2771, + "step": 2572 + }, + { + "epoch": 0.4581552706552707, + "grad_norm": 0.5511104464530945, + "learning_rate": 0.0001936825151690274, + "loss": 1.0039, + "step": 2573 + }, + { + "epoch": 0.4583333333333333, + "grad_norm": 0.4721006453037262, + "learning_rate": 0.0001936776179636294, + "loss": 1.3246, + "step": 2574 + }, + { + "epoch": 0.458511396011396, + "grad_norm": 0.5021488666534424, + "learning_rate": 0.0001936727189228131, + "loss": 1.1733, + "step": 2575 + }, + { + "epoch": 0.4586894586894587, + "grad_norm": 0.5755292177200317, + "learning_rate": 0.0001936678180466745, + "loss": 1.2241, + "step": 2576 + }, + { + "epoch": 0.45886752136752135, + "grad_norm": 0.4501610994338989, + "learning_rate": 0.00019366291533530952, + "loss": 1.0503, + "step": 2577 + }, + { + "epoch": 0.45904558404558404, + "grad_norm": 0.4067458212375641, + "learning_rate": 0.00019365801078881432, + "loss": 0.8259, + "step": 2578 + }, + { + "epoch": 0.45922364672364674, + "grad_norm": 0.539730429649353, + "learning_rate": 0.0001936531044072849, + "loss": 1.1964, + "step": 2579 + }, + { + "epoch": 0.4594017094017094, + "grad_norm": 0.5624797344207764, + "learning_rate": 0.0001936481961908175, + "loss": 1.2059, + "step": 2580 + }, + { + "epoch": 0.4595797720797721, + "grad_norm": 0.43679240345954895, + "learning_rate": 0.00019364328613950824, + "loss": 1.1371, + "step": 2581 + }, + { + "epoch": 0.45975783475783477, + "grad_norm": 0.5214769244194031, + "learning_rate": 0.00019363837425345328, + "loss": 1.109, + "step": 2582 + }, + { + "epoch": 0.4599358974358974, + "grad_norm": 0.4522894024848938, + "learning_rate": 0.00019363346053274892, + "loss": 1.0532, + "step": 2583 + }, + { + "epoch": 0.4601139601139601, + "grad_norm": 0.44980281591415405, + "learning_rate": 0.0001936285449774914, + "loss": 0.9352, + "step": 2584 + }, + { + "epoch": 0.4602920227920228, + "grad_norm": 0.5697414875030518, + "learning_rate": 0.00019362362758777705, + "loss": 1.2171, + "step": 2585 + }, + { + "epoch": 0.46047008547008544, + "grad_norm": 0.4636315107345581, + "learning_rate": 0.00019361870836370217, + "loss": 1.0662, + "step": 2586 + }, + { + "epoch": 0.46064814814814814, + "grad_norm": 0.5144017338752747, + "learning_rate": 0.00019361378730536321, + "loss": 1.0681, + "step": 2587 + }, + { + "epoch": 0.46082621082621084, + "grad_norm": 0.5007636547088623, + "learning_rate": 0.00019360886441285654, + "loss": 1.2058, + "step": 2588 + }, + { + "epoch": 0.46100427350427353, + "grad_norm": 0.5024117231369019, + "learning_rate": 0.00019360393968627864, + "loss": 1.065, + "step": 2589 + }, + { + "epoch": 0.46118233618233617, + "grad_norm": 0.48105588555336, + "learning_rate": 0.00019359901312572596, + "loss": 1.0887, + "step": 2590 + }, + { + "epoch": 0.46136039886039887, + "grad_norm": 0.5381982326507568, + "learning_rate": 0.00019359408473129506, + "loss": 1.2754, + "step": 2591 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 0.5051333904266357, + "learning_rate": 0.0001935891545030825, + "loss": 0.9334, + "step": 2592 + }, + { + "epoch": 0.4617165242165242, + "grad_norm": 0.43818601965904236, + "learning_rate": 0.0001935842224411849, + "loss": 1.0967, + "step": 2593 + }, + { + "epoch": 0.4618945868945869, + "grad_norm": 0.4727257490158081, + "learning_rate": 0.0001935792885456988, + "loss": 0.8136, + "step": 2594 + }, + { + "epoch": 0.4620726495726496, + "grad_norm": 0.5505291223526001, + "learning_rate": 0.00019357435281672098, + "loss": 1.3113, + "step": 2595 + }, + { + "epoch": 0.46225071225071224, + "grad_norm": 0.4705682396888733, + "learning_rate": 0.0001935694152543481, + "loss": 0.9863, + "step": 2596 + }, + { + "epoch": 0.46242877492877493, + "grad_norm": 0.49653419852256775, + "learning_rate": 0.0001935644758586769, + "loss": 1.035, + "step": 2597 + }, + { + "epoch": 0.46260683760683763, + "grad_norm": 0.4788367748260498, + "learning_rate": 0.00019355953462980415, + "loss": 1.1253, + "step": 2598 + }, + { + "epoch": 0.46278490028490027, + "grad_norm": 0.5295125842094421, + "learning_rate": 0.00019355459156782668, + "loss": 1.0853, + "step": 2599 + }, + { + "epoch": 0.46296296296296297, + "grad_norm": 0.4878056049346924, + "learning_rate": 0.00019354964667284133, + "loss": 1.1381, + "step": 2600 + }, + { + "epoch": 0.46314102564102566, + "grad_norm": 0.5442031025886536, + "learning_rate": 0.00019354469994494497, + "loss": 1.1349, + "step": 2601 + }, + { + "epoch": 0.4633190883190883, + "grad_norm": 0.4845225214958191, + "learning_rate": 0.00019353975138423457, + "loss": 1.0538, + "step": 2602 + }, + { + "epoch": 0.463497150997151, + "grad_norm": 0.4957871437072754, + "learning_rate": 0.00019353480099080703, + "loss": 1.2765, + "step": 2603 + }, + { + "epoch": 0.4636752136752137, + "grad_norm": 0.5414339303970337, + "learning_rate": 0.00019352984876475936, + "loss": 1.1015, + "step": 2604 + }, + { + "epoch": 0.46385327635327633, + "grad_norm": 0.5171043872833252, + "learning_rate": 0.0001935248947061886, + "loss": 0.9995, + "step": 2605 + }, + { + "epoch": 0.46403133903133903, + "grad_norm": 0.46040529012680054, + "learning_rate": 0.0001935199388151918, + "loss": 1.1126, + "step": 2606 + }, + { + "epoch": 0.4642094017094017, + "grad_norm": 0.5327033400535583, + "learning_rate": 0.00019351498109186613, + "loss": 1.1983, + "step": 2607 + }, + { + "epoch": 0.46438746438746437, + "grad_norm": 0.4451361298561096, + "learning_rate": 0.0001935100215363086, + "loss": 0.9689, + "step": 2608 + }, + { + "epoch": 0.46456552706552706, + "grad_norm": 0.5462809801101685, + "learning_rate": 0.00019350506014861646, + "loss": 1.036, + "step": 2609 + }, + { + "epoch": 0.46474358974358976, + "grad_norm": 0.4907000958919525, + "learning_rate": 0.00019350009692888694, + "loss": 1.0724, + "step": 2610 + }, + { + "epoch": 0.4649216524216524, + "grad_norm": 0.47523510456085205, + "learning_rate": 0.00019349513187721723, + "loss": 0.9214, + "step": 2611 + }, + { + "epoch": 0.4650997150997151, + "grad_norm": 0.539732813835144, + "learning_rate": 0.0001934901649937046, + "loss": 1.1166, + "step": 2612 + }, + { + "epoch": 0.4652777777777778, + "grad_norm": 0.4827860891819, + "learning_rate": 0.00019348519627844643, + "loss": 1.1613, + "step": 2613 + }, + { + "epoch": 0.46545584045584043, + "grad_norm": 0.5385223031044006, + "learning_rate": 0.00019348022573154, + "loss": 1.0105, + "step": 2614 + }, + { + "epoch": 0.4656339031339031, + "grad_norm": 0.4629383087158203, + "learning_rate": 0.0001934752533530828, + "loss": 1.0298, + "step": 2615 + }, + { + "epoch": 0.4658119658119658, + "grad_norm": 0.599371075630188, + "learning_rate": 0.00019347027914317212, + "loss": 1.3158, + "step": 2616 + }, + { + "epoch": 0.46599002849002846, + "grad_norm": 0.5954698324203491, + "learning_rate": 0.00019346530310190553, + "loss": 1.1882, + "step": 2617 + }, + { + "epoch": 0.46616809116809116, + "grad_norm": 0.49185171723365784, + "learning_rate": 0.00019346032522938046, + "loss": 1.0977, + "step": 2618 + }, + { + "epoch": 0.46634615384615385, + "grad_norm": 0.5145422220230103, + "learning_rate": 0.0001934553455256945, + "loss": 0.9948, + "step": 2619 + }, + { + "epoch": 0.46652421652421655, + "grad_norm": 0.6809412837028503, + "learning_rate": 0.00019345036399094517, + "loss": 1.5798, + "step": 2620 + }, + { + "epoch": 0.4667022792022792, + "grad_norm": 0.4606841206550598, + "learning_rate": 0.00019344538062523005, + "loss": 0.7357, + "step": 2621 + }, + { + "epoch": 0.4668803418803419, + "grad_norm": 0.49036628007888794, + "learning_rate": 0.00019344039542864685, + "loss": 1.1518, + "step": 2622 + }, + { + "epoch": 0.4670584045584046, + "grad_norm": 0.47904539108276367, + "learning_rate": 0.0001934354084012932, + "loss": 0.9929, + "step": 2623 + }, + { + "epoch": 0.4672364672364672, + "grad_norm": 0.5224666595458984, + "learning_rate": 0.0001934304195432668, + "loss": 1.2544, + "step": 2624 + }, + { + "epoch": 0.4674145299145299, + "grad_norm": 0.4902483820915222, + "learning_rate": 0.00019342542885466543, + "loss": 1.0301, + "step": 2625 + }, + { + "epoch": 0.4675925925925926, + "grad_norm": 0.46824702620506287, + "learning_rate": 0.00019342043633558683, + "loss": 0.9364, + "step": 2626 + }, + { + "epoch": 0.46777065527065526, + "grad_norm": 0.46272051334381104, + "learning_rate": 0.00019341544198612888, + "loss": 1.056, + "step": 2627 + }, + { + "epoch": 0.46794871794871795, + "grad_norm": 0.6216606497764587, + "learning_rate": 0.0001934104458063894, + "loss": 1.0825, + "step": 2628 + }, + { + "epoch": 0.46812678062678065, + "grad_norm": 0.5024014115333557, + "learning_rate": 0.00019340544779646623, + "loss": 1.1832, + "step": 2629 + }, + { + "epoch": 0.4683048433048433, + "grad_norm": 0.5547130107879639, + "learning_rate": 0.00019340044795645737, + "loss": 1.1335, + "step": 2630 + }, + { + "epoch": 0.468482905982906, + "grad_norm": 0.5439161658287048, + "learning_rate": 0.0001933954462864608, + "loss": 1.0229, + "step": 2631 + }, + { + "epoch": 0.4686609686609687, + "grad_norm": 0.4782990515232086, + "learning_rate": 0.0001933904427865744, + "loss": 1.2318, + "step": 2632 + }, + { + "epoch": 0.4688390313390313, + "grad_norm": 0.5872140526771545, + "learning_rate": 0.00019338543745689633, + "loss": 1.0132, + "step": 2633 + }, + { + "epoch": 0.469017094017094, + "grad_norm": 0.44163307547569275, + "learning_rate": 0.00019338043029752458, + "loss": 1.0091, + "step": 2634 + }, + { + "epoch": 0.4691951566951567, + "grad_norm": 0.541081428527832, + "learning_rate": 0.0001933754213085573, + "loss": 1.2155, + "step": 2635 + }, + { + "epoch": 0.46937321937321935, + "grad_norm": 0.4761527478694916, + "learning_rate": 0.00019337041049009255, + "loss": 1.1138, + "step": 2636 + }, + { + "epoch": 0.46955128205128205, + "grad_norm": 0.46414369344711304, + "learning_rate": 0.0001933653978422286, + "loss": 0.9903, + "step": 2637 + }, + { + "epoch": 0.46972934472934474, + "grad_norm": 0.5337086915969849, + "learning_rate": 0.00019336038336506363, + "loss": 1.2873, + "step": 2638 + }, + { + "epoch": 0.4699074074074074, + "grad_norm": 0.5065379738807678, + "learning_rate": 0.00019335536705869592, + "loss": 1.1436, + "step": 2639 + }, + { + "epoch": 0.4700854700854701, + "grad_norm": 0.5539217591285706, + "learning_rate": 0.0001933503489232237, + "loss": 1.2881, + "step": 2640 + }, + { + "epoch": 0.4702635327635328, + "grad_norm": 0.48303213715553284, + "learning_rate": 0.0001933453289587453, + "loss": 1.0209, + "step": 2641 + }, + { + "epoch": 0.4704415954415954, + "grad_norm": 0.6986871957778931, + "learning_rate": 0.00019334030716535908, + "loss": 1.1979, + "step": 2642 + }, + { + "epoch": 0.4706196581196581, + "grad_norm": 0.46137234568595886, + "learning_rate": 0.00019333528354316347, + "loss": 1.0682, + "step": 2643 + }, + { + "epoch": 0.4707977207977208, + "grad_norm": 0.4726654291152954, + "learning_rate": 0.00019333025809225684, + "loss": 1.1712, + "step": 2644 + }, + { + "epoch": 0.47097578347578345, + "grad_norm": 0.46188637614250183, + "learning_rate": 0.0001933252308127377, + "loss": 1.0183, + "step": 2645 + }, + { + "epoch": 0.47115384615384615, + "grad_norm": 0.5323259830474854, + "learning_rate": 0.0001933202017047045, + "loss": 0.935, + "step": 2646 + }, + { + "epoch": 0.47133190883190884, + "grad_norm": 0.5004189014434814, + "learning_rate": 0.00019331517076825582, + "loss": 1.1331, + "step": 2647 + }, + { + "epoch": 0.47150997150997154, + "grad_norm": 0.5443634986877441, + "learning_rate": 0.0001933101380034902, + "loss": 1.0514, + "step": 2648 + }, + { + "epoch": 0.4716880341880342, + "grad_norm": 0.504180371761322, + "learning_rate": 0.0001933051034105063, + "loss": 1.3099, + "step": 2649 + }, + { + "epoch": 0.4718660968660969, + "grad_norm": 0.5092344284057617, + "learning_rate": 0.0001933000669894027, + "loss": 1.0716, + "step": 2650 + }, + { + "epoch": 0.47204415954415957, + "grad_norm": 0.5236422419548035, + "learning_rate": 0.0001932950287402781, + "loss": 1.0981, + "step": 2651 + }, + { + "epoch": 0.4722222222222222, + "grad_norm": 0.6228063702583313, + "learning_rate": 0.0001932899886632312, + "loss": 1.3398, + "step": 2652 + }, + { + "epoch": 0.4724002849002849, + "grad_norm": 0.5112748146057129, + "learning_rate": 0.00019328494675836078, + "loss": 1.0151, + "step": 2653 + }, + { + "epoch": 0.4725783475783476, + "grad_norm": 0.5554201602935791, + "learning_rate": 0.00019327990302576563, + "loss": 1.404, + "step": 2654 + }, + { + "epoch": 0.47275641025641024, + "grad_norm": 0.5050725340843201, + "learning_rate": 0.0001932748574655445, + "loss": 0.951, + "step": 2655 + }, + { + "epoch": 0.47293447293447294, + "grad_norm": 0.5161749720573425, + "learning_rate": 0.00019326981007779636, + "loss": 1.2425, + "step": 2656 + }, + { + "epoch": 0.47311253561253563, + "grad_norm": 0.4865442216396332, + "learning_rate": 0.00019326476086262002, + "loss": 1.1175, + "step": 2657 + }, + { + "epoch": 0.4732905982905983, + "grad_norm": 0.5276186466217041, + "learning_rate": 0.0001932597098201144, + "loss": 1.3687, + "step": 2658 + }, + { + "epoch": 0.47346866096866097, + "grad_norm": 0.509139358997345, + "learning_rate": 0.00019325465695037855, + "loss": 1.0546, + "step": 2659 + }, + { + "epoch": 0.47364672364672367, + "grad_norm": 0.49815434217453003, + "learning_rate": 0.00019324960225351138, + "loss": 1.0807, + "step": 2660 + }, + { + "epoch": 0.4738247863247863, + "grad_norm": 0.5059618353843689, + "learning_rate": 0.00019324454572961197, + "loss": 1.0827, + "step": 2661 + }, + { + "epoch": 0.474002849002849, + "grad_norm": 0.5698565244674683, + "learning_rate": 0.00019323948737877942, + "loss": 1.2019, + "step": 2662 + }, + { + "epoch": 0.4741809116809117, + "grad_norm": 0.49661511182785034, + "learning_rate": 0.00019323442720111276, + "loss": 1.1447, + "step": 2663 + }, + { + "epoch": 0.47435897435897434, + "grad_norm": 0.46442747116088867, + "learning_rate": 0.0001932293651967112, + "loss": 0.8796, + "step": 2664 + }, + { + "epoch": 0.47453703703703703, + "grad_norm": 0.48306044936180115, + "learning_rate": 0.00019322430136567388, + "loss": 1.1358, + "step": 2665 + }, + { + "epoch": 0.47471509971509973, + "grad_norm": 0.5677350759506226, + "learning_rate": 0.00019321923570810005, + "loss": 1.1026, + "step": 2666 + }, + { + "epoch": 0.47489316239316237, + "grad_norm": 0.3700144588947296, + "learning_rate": 0.0001932141682240889, + "loss": 0.7514, + "step": 2667 + }, + { + "epoch": 0.47507122507122507, + "grad_norm": 0.6003054976463318, + "learning_rate": 0.0001932090989137398, + "loss": 1.1591, + "step": 2668 + }, + { + "epoch": 0.47524928774928776, + "grad_norm": 0.520298421382904, + "learning_rate": 0.00019320402777715204, + "loss": 1.339, + "step": 2669 + }, + { + "epoch": 0.4754273504273504, + "grad_norm": 0.46453598141670227, + "learning_rate": 0.00019319895481442493, + "loss": 0.9879, + "step": 2670 + }, + { + "epoch": 0.4756054131054131, + "grad_norm": 0.5247363448143005, + "learning_rate": 0.00019319388002565793, + "loss": 0.9862, + "step": 2671 + }, + { + "epoch": 0.4757834757834758, + "grad_norm": 0.5498613715171814, + "learning_rate": 0.00019318880341095046, + "loss": 1.2224, + "step": 2672 + }, + { + "epoch": 0.47596153846153844, + "grad_norm": 0.565838098526001, + "learning_rate": 0.00019318372497040192, + "loss": 1.0712, + "step": 2673 + }, + { + "epoch": 0.47613960113960113, + "grad_norm": 0.5797489881515503, + "learning_rate": 0.00019317864470411191, + "loss": 1.0176, + "step": 2674 + }, + { + "epoch": 0.47631766381766383, + "grad_norm": 0.5114326477050781, + "learning_rate": 0.0001931735626121799, + "loss": 1.1027, + "step": 2675 + }, + { + "epoch": 0.47649572649572647, + "grad_norm": 0.5396515727043152, + "learning_rate": 0.00019316847869470547, + "loss": 1.1782, + "step": 2676 + }, + { + "epoch": 0.47667378917378916, + "grad_norm": 0.4812076985836029, + "learning_rate": 0.00019316339295178824, + "loss": 1.1196, + "step": 2677 + }, + { + "epoch": 0.47685185185185186, + "grad_norm": 0.4875647723674774, + "learning_rate": 0.00019315830538352787, + "loss": 1.1407, + "step": 2678 + }, + { + "epoch": 0.47702991452991456, + "grad_norm": 0.5036377906799316, + "learning_rate": 0.00019315321599002404, + "loss": 0.9842, + "step": 2679 + }, + { + "epoch": 0.4772079772079772, + "grad_norm": 0.5054177641868591, + "learning_rate": 0.00019314812477137645, + "loss": 0.8196, + "step": 2680 + }, + { + "epoch": 0.4773860398860399, + "grad_norm": 0.5050665736198425, + "learning_rate": 0.00019314303172768483, + "loss": 0.8463, + "step": 2681 + }, + { + "epoch": 0.4775641025641026, + "grad_norm": 0.5179004669189453, + "learning_rate": 0.000193137936859049, + "loss": 1.2485, + "step": 2682 + }, + { + "epoch": 0.47774216524216523, + "grad_norm": 0.44986143708229065, + "learning_rate": 0.00019313284016556876, + "loss": 0.9855, + "step": 2683 + }, + { + "epoch": 0.4779202279202279, + "grad_norm": 0.5594347715377808, + "learning_rate": 0.00019312774164734398, + "loss": 1.0987, + "step": 2684 + }, + { + "epoch": 0.4780982905982906, + "grad_norm": 0.4837244749069214, + "learning_rate": 0.0001931226413044746, + "loss": 1.1119, + "step": 2685 + }, + { + "epoch": 0.47827635327635326, + "grad_norm": 0.489145427942276, + "learning_rate": 0.0001931175391370605, + "loss": 1.1962, + "step": 2686 + }, + { + "epoch": 0.47845441595441596, + "grad_norm": 0.503568708896637, + "learning_rate": 0.00019311243514520164, + "loss": 0.9668, + "step": 2687 + }, + { + "epoch": 0.47863247863247865, + "grad_norm": 0.5401005744934082, + "learning_rate": 0.00019310732932899805, + "loss": 1.3072, + "step": 2688 + }, + { + "epoch": 0.4788105413105413, + "grad_norm": 0.526523768901825, + "learning_rate": 0.00019310222168854971, + "loss": 1.1387, + "step": 2689 + }, + { + "epoch": 0.478988603988604, + "grad_norm": 0.5223183631896973, + "learning_rate": 0.00019309711222395678, + "loss": 1.1391, + "step": 2690 + }, + { + "epoch": 0.4791666666666667, + "grad_norm": 0.5840879082679749, + "learning_rate": 0.00019309200093531933, + "loss": 1.1543, + "step": 2691 + }, + { + "epoch": 0.4793447293447293, + "grad_norm": 0.5173699259757996, + "learning_rate": 0.00019308688782273753, + "loss": 1.1889, + "step": 2692 + }, + { + "epoch": 0.479522792022792, + "grad_norm": 0.5417894124984741, + "learning_rate": 0.00019308177288631146, + "loss": 1.299, + "step": 2693 + }, + { + "epoch": 0.4797008547008547, + "grad_norm": 0.4890797734260559, + "learning_rate": 0.0001930766561261415, + "loss": 1.1516, + "step": 2694 + }, + { + "epoch": 0.47987891737891736, + "grad_norm": 0.5422119498252869, + "learning_rate": 0.00019307153754232772, + "loss": 1.0301, + "step": 2695 + }, + { + "epoch": 0.48005698005698005, + "grad_norm": 0.5838702917098999, + "learning_rate": 0.00019306641713497057, + "loss": 1.265, + "step": 2696 + }, + { + "epoch": 0.48023504273504275, + "grad_norm": 0.5020943284034729, + "learning_rate": 0.00019306129490417027, + "loss": 1.1119, + "step": 2697 + }, + { + "epoch": 0.4804131054131054, + "grad_norm": 0.412993460893631, + "learning_rate": 0.00019305617085002723, + "loss": 0.8083, + "step": 2698 + }, + { + "epoch": 0.4805911680911681, + "grad_norm": 0.6270101070404053, + "learning_rate": 0.00019305104497264184, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 0.4807692307692308, + "grad_norm": 0.45256730914115906, + "learning_rate": 0.0001930459172721145, + "loss": 1.0368, + "step": 2700 + }, + { + "epoch": 0.4809472934472934, + "grad_norm": 0.5351749658584595, + "learning_rate": 0.0001930407877485457, + "loss": 1.135, + "step": 2701 + }, + { + "epoch": 0.4811253561253561, + "grad_norm": 0.49324163794517517, + "learning_rate": 0.00019303565640203593, + "loss": 0.9383, + "step": 2702 + }, + { + "epoch": 0.4813034188034188, + "grad_norm": 0.5434361100196838, + "learning_rate": 0.00019303052323268576, + "loss": 1.2605, + "step": 2703 + }, + { + "epoch": 0.48148148148148145, + "grad_norm": 0.5858064889907837, + "learning_rate": 0.00019302538824059572, + "loss": 1.0846, + "step": 2704 + }, + { + "epoch": 0.48165954415954415, + "grad_norm": 0.5753700733184814, + "learning_rate": 0.00019302025142586647, + "loss": 1.0371, + "step": 2705 + }, + { + "epoch": 0.48183760683760685, + "grad_norm": 0.43102699518203735, + "learning_rate": 0.00019301511278859858, + "loss": 0.9189, + "step": 2706 + }, + { + "epoch": 0.48201566951566954, + "grad_norm": 0.4731025993824005, + "learning_rate": 0.0001930099723288928, + "loss": 1.1291, + "step": 2707 + }, + { + "epoch": 0.4821937321937322, + "grad_norm": 0.5685615539550781, + "learning_rate": 0.00019300483004684987, + "loss": 1.1006, + "step": 2708 + }, + { + "epoch": 0.4823717948717949, + "grad_norm": 0.4368155896663666, + "learning_rate": 0.00019299968594257044, + "loss": 0.9959, + "step": 2709 + }, + { + "epoch": 0.4825498575498576, + "grad_norm": 0.5594738125801086, + "learning_rate": 0.00019299454001615537, + "loss": 1.0826, + "step": 2710 + }, + { + "epoch": 0.4827279202279202, + "grad_norm": 0.48876598477363586, + "learning_rate": 0.00019298939226770548, + "loss": 1.1556, + "step": 2711 + }, + { + "epoch": 0.4829059829059829, + "grad_norm": 0.548039436340332, + "learning_rate": 0.00019298424269732157, + "loss": 1.158, + "step": 2712 + }, + { + "epoch": 0.4830840455840456, + "grad_norm": 0.4957645535469055, + "learning_rate": 0.00019297909130510464, + "loss": 0.9824, + "step": 2713 + }, + { + "epoch": 0.48326210826210825, + "grad_norm": 0.5197011232376099, + "learning_rate": 0.00019297393809115555, + "loss": 1.1074, + "step": 2714 + }, + { + "epoch": 0.48344017094017094, + "grad_norm": 0.5742064118385315, + "learning_rate": 0.00019296878305557526, + "loss": 1.0431, + "step": 2715 + }, + { + "epoch": 0.48361823361823364, + "grad_norm": 0.5698413252830505, + "learning_rate": 0.0001929636261984648, + "loss": 1.0713, + "step": 2716 + }, + { + "epoch": 0.4837962962962963, + "grad_norm": 0.48126333951950073, + "learning_rate": 0.0001929584675199252, + "loss": 0.9274, + "step": 2717 + }, + { + "epoch": 0.483974358974359, + "grad_norm": 0.49299830198287964, + "learning_rate": 0.00019295330702005754, + "loss": 0.9392, + "step": 2718 + }, + { + "epoch": 0.48415242165242167, + "grad_norm": 0.4780774414539337, + "learning_rate": 0.0001929481446989629, + "loss": 1.1459, + "step": 2719 + }, + { + "epoch": 0.4843304843304843, + "grad_norm": 0.5462654829025269, + "learning_rate": 0.00019294298055674248, + "loss": 1.0635, + "step": 2720 + }, + { + "epoch": 0.484508547008547, + "grad_norm": 0.5371061563491821, + "learning_rate": 0.00019293781459349743, + "loss": 1.3578, + "step": 2721 + }, + { + "epoch": 0.4846866096866097, + "grad_norm": 0.46308520436286926, + "learning_rate": 0.00019293264680932893, + "loss": 0.9001, + "step": 2722 + }, + { + "epoch": 0.48486467236467234, + "grad_norm": 0.5149807929992676, + "learning_rate": 0.0001929274772043383, + "loss": 0.6908, + "step": 2723 + }, + { + "epoch": 0.48504273504273504, + "grad_norm": 0.5435031056404114, + "learning_rate": 0.00019292230577862678, + "loss": 1.2143, + "step": 2724 + }, + { + "epoch": 0.48522079772079774, + "grad_norm": 0.44217726588249207, + "learning_rate": 0.00019291713253229568, + "loss": 0.9303, + "step": 2725 + }, + { + "epoch": 0.4853988603988604, + "grad_norm": 0.6120226383209229, + "learning_rate": 0.00019291195746544643, + "loss": 1.3801, + "step": 2726 + }, + { + "epoch": 0.4855769230769231, + "grad_norm": 0.5014316439628601, + "learning_rate": 0.00019290678057818037, + "loss": 1.0631, + "step": 2727 + }, + { + "epoch": 0.48575498575498577, + "grad_norm": 0.5667829513549805, + "learning_rate": 0.00019290160187059895, + "loss": 1.3166, + "step": 2728 + }, + { + "epoch": 0.4859330484330484, + "grad_norm": 0.5011509656906128, + "learning_rate": 0.0001928964213428036, + "loss": 1.1887, + "step": 2729 + }, + { + "epoch": 0.4861111111111111, + "grad_norm": 0.48317405581474304, + "learning_rate": 0.00019289123899489586, + "loss": 1.1125, + "step": 2730 + }, + { + "epoch": 0.4862891737891738, + "grad_norm": 0.4669005870819092, + "learning_rate": 0.00019288605482697726, + "loss": 1.0091, + "step": 2731 + }, + { + "epoch": 0.48646723646723644, + "grad_norm": 0.4330739974975586, + "learning_rate": 0.00019288086883914937, + "loss": 0.9789, + "step": 2732 + }, + { + "epoch": 0.48664529914529914, + "grad_norm": 0.48482781648635864, + "learning_rate": 0.0001928756810315138, + "loss": 1.1922, + "step": 2733 + }, + { + "epoch": 0.48682336182336183, + "grad_norm": 0.5781838297843933, + "learning_rate": 0.0001928704914041722, + "loss": 1.1793, + "step": 2734 + }, + { + "epoch": 0.48700142450142453, + "grad_norm": 0.5955413579940796, + "learning_rate": 0.00019286529995722623, + "loss": 1.1001, + "step": 2735 + }, + { + "epoch": 0.48717948717948717, + "grad_norm": 0.49204322695732117, + "learning_rate": 0.00019286010669077763, + "loss": 0.9219, + "step": 2736 + }, + { + "epoch": 0.48735754985754987, + "grad_norm": 0.5853500962257385, + "learning_rate": 0.00019285491160492813, + "loss": 1.1133, + "step": 2737 + }, + { + "epoch": 0.48753561253561256, + "grad_norm": 0.5555846095085144, + "learning_rate": 0.0001928497146997795, + "loss": 1.0915, + "step": 2738 + }, + { + "epoch": 0.4877136752136752, + "grad_norm": 0.5166759490966797, + "learning_rate": 0.00019284451597543364, + "loss": 0.9349, + "step": 2739 + }, + { + "epoch": 0.4878917378917379, + "grad_norm": 0.47816506028175354, + "learning_rate": 0.00019283931543199234, + "loss": 0.8978, + "step": 2740 + }, + { + "epoch": 0.4880698005698006, + "grad_norm": 0.5632442831993103, + "learning_rate": 0.0001928341130695575, + "loss": 1.0491, + "step": 2741 + }, + { + "epoch": 0.48824786324786323, + "grad_norm": 0.6532769799232483, + "learning_rate": 0.00019282890888823107, + "loss": 1.2779, + "step": 2742 + }, + { + "epoch": 0.48842592592592593, + "grad_norm": 0.5733640789985657, + "learning_rate": 0.000192823702888115, + "loss": 1.4127, + "step": 2743 + }, + { + "epoch": 0.4886039886039886, + "grad_norm": 0.5701746344566345, + "learning_rate": 0.00019281849506931132, + "loss": 1.138, + "step": 2744 + }, + { + "epoch": 0.48878205128205127, + "grad_norm": 0.5227449536323547, + "learning_rate": 0.000192813285431922, + "loss": 1.1831, + "step": 2745 + }, + { + "epoch": 0.48896011396011396, + "grad_norm": 0.48457080125808716, + "learning_rate": 0.00019280807397604915, + "loss": 1.2468, + "step": 2746 + }, + { + "epoch": 0.48913817663817666, + "grad_norm": 0.4596176743507385, + "learning_rate": 0.0001928028607017949, + "loss": 1.1098, + "step": 2747 + }, + { + "epoch": 0.4893162393162393, + "grad_norm": 0.5204966068267822, + "learning_rate": 0.00019279764560926142, + "loss": 1.1501, + "step": 2748 + }, + { + "epoch": 0.489494301994302, + "grad_norm": 0.5179490447044373, + "learning_rate": 0.0001927924286985508, + "loss": 1.2601, + "step": 2749 + }, + { + "epoch": 0.4896723646723647, + "grad_norm": 0.4563423693180084, + "learning_rate": 0.00019278720996976533, + "loss": 1.081, + "step": 2750 + }, + { + "epoch": 0.48985042735042733, + "grad_norm": 0.4906339943408966, + "learning_rate": 0.00019278198942300717, + "loss": 1.157, + "step": 2751 + }, + { + "epoch": 0.49002849002849, + "grad_norm": 0.42241403460502625, + "learning_rate": 0.00019277676705837873, + "loss": 1.0333, + "step": 2752 + }, + { + "epoch": 0.4902065527065527, + "grad_norm": 0.6310175657272339, + "learning_rate": 0.00019277154287598226, + "loss": 1.1225, + "step": 2753 + }, + { + "epoch": 0.49038461538461536, + "grad_norm": 0.5109034776687622, + "learning_rate": 0.0001927663168759201, + "loss": 1.1619, + "step": 2754 + }, + { + "epoch": 0.49056267806267806, + "grad_norm": 0.4809598922729492, + "learning_rate": 0.00019276108905829465, + "loss": 1.0423, + "step": 2755 + }, + { + "epoch": 0.49074074074074076, + "grad_norm": 0.557502806186676, + "learning_rate": 0.00019275585942320837, + "loss": 0.8783, + "step": 2756 + }, + { + "epoch": 0.4909188034188034, + "grad_norm": 0.5434393882751465, + "learning_rate": 0.0001927506279707637, + "loss": 1.1701, + "step": 2757 + }, + { + "epoch": 0.4910968660968661, + "grad_norm": 0.49278944730758667, + "learning_rate": 0.00019274539470106317, + "loss": 1.0447, + "step": 2758 + }, + { + "epoch": 0.4912749287749288, + "grad_norm": 0.5634264349937439, + "learning_rate": 0.00019274015961420927, + "loss": 1.0639, + "step": 2759 + }, + { + "epoch": 0.49145299145299143, + "grad_norm": 0.5632645487785339, + "learning_rate": 0.00019273492271030464, + "loss": 0.9223, + "step": 2760 + }, + { + "epoch": 0.4916310541310541, + "grad_norm": 0.5949172377586365, + "learning_rate": 0.00019272968398945177, + "loss": 0.894, + "step": 2761 + }, + { + "epoch": 0.4918091168091168, + "grad_norm": 0.5375374555587769, + "learning_rate": 0.00019272444345175342, + "loss": 1.0311, + "step": 2762 + }, + { + "epoch": 0.49198717948717946, + "grad_norm": 0.5211305022239685, + "learning_rate": 0.00019271920109731222, + "loss": 1.1531, + "step": 2763 + }, + { + "epoch": 0.49216524216524216, + "grad_norm": 0.44022253155708313, + "learning_rate": 0.00019271395692623084, + "loss": 0.9147, + "step": 2764 + }, + { + "epoch": 0.49234330484330485, + "grad_norm": 0.4682174623012543, + "learning_rate": 0.0001927087109386121, + "loss": 1.081, + "step": 2765 + }, + { + "epoch": 0.49252136752136755, + "grad_norm": 0.4971517324447632, + "learning_rate": 0.0001927034631345588, + "loss": 1.1017, + "step": 2766 + }, + { + "epoch": 0.4926994301994302, + "grad_norm": 0.5015294551849365, + "learning_rate": 0.00019269821351417364, + "loss": 1.1093, + "step": 2767 + }, + { + "epoch": 0.4928774928774929, + "grad_norm": 0.5512694716453552, + "learning_rate": 0.00019269296207755958, + "loss": 0.9657, + "step": 2768 + }, + { + "epoch": 0.4930555555555556, + "grad_norm": 0.4914868474006653, + "learning_rate": 0.00019268770882481948, + "loss": 1.0379, + "step": 2769 + }, + { + "epoch": 0.4932336182336182, + "grad_norm": 0.567337691783905, + "learning_rate": 0.00019268245375605626, + "loss": 1.004, + "step": 2770 + }, + { + "epoch": 0.4934116809116809, + "grad_norm": 0.518489420413971, + "learning_rate": 0.0001926771968713729, + "loss": 1.0734, + "step": 2771 + }, + { + "epoch": 0.4935897435897436, + "grad_norm": 0.567742109298706, + "learning_rate": 0.00019267193817087237, + "loss": 1.1276, + "step": 2772 + }, + { + "epoch": 0.49376780626780625, + "grad_norm": 0.5287964344024658, + "learning_rate": 0.00019266667765465773, + "loss": 1.1429, + "step": 2773 + }, + { + "epoch": 0.49394586894586895, + "grad_norm": 0.5302085876464844, + "learning_rate": 0.00019266141532283207, + "loss": 1.0934, + "step": 2774 + }, + { + "epoch": 0.49412393162393164, + "grad_norm": 0.5569987297058105, + "learning_rate": 0.00019265615117549842, + "loss": 1.1453, + "step": 2775 + }, + { + "epoch": 0.4943019943019943, + "grad_norm": 0.519695520401001, + "learning_rate": 0.00019265088521275997, + "loss": 1.1255, + "step": 2776 + }, + { + "epoch": 0.494480056980057, + "grad_norm": 0.5073211193084717, + "learning_rate": 0.0001926456174347199, + "loss": 1.0609, + "step": 2777 + }, + { + "epoch": 0.4946581196581197, + "grad_norm": 0.45028239488601685, + "learning_rate": 0.00019264034784148142, + "loss": 0.9098, + "step": 2778 + }, + { + "epoch": 0.4948361823361823, + "grad_norm": 0.6641215682029724, + "learning_rate": 0.00019263507643314776, + "loss": 0.8903, + "step": 2779 + }, + { + "epoch": 0.495014245014245, + "grad_norm": 0.5281413793563843, + "learning_rate": 0.00019262980320982224, + "loss": 1.2906, + "step": 2780 + }, + { + "epoch": 0.4951923076923077, + "grad_norm": 0.6256437301635742, + "learning_rate": 0.0001926245281716081, + "loss": 1.4142, + "step": 2781 + }, + { + "epoch": 0.49537037037037035, + "grad_norm": 0.5422517657279968, + "learning_rate": 0.00019261925131860877, + "loss": 1.1606, + "step": 2782 + }, + { + "epoch": 0.49554843304843305, + "grad_norm": 0.46938949823379517, + "learning_rate": 0.0001926139726509276, + "loss": 1.0333, + "step": 2783 + }, + { + "epoch": 0.49572649572649574, + "grad_norm": 0.5799683928489685, + "learning_rate": 0.000192608692168668, + "loss": 1.0333, + "step": 2784 + }, + { + "epoch": 0.4959045584045584, + "grad_norm": 0.5231602787971497, + "learning_rate": 0.0001926034098719335, + "loss": 1.1847, + "step": 2785 + }, + { + "epoch": 0.4960826210826211, + "grad_norm": 0.477845698595047, + "learning_rate": 0.00019259812576082752, + "loss": 1.0746, + "step": 2786 + }, + { + "epoch": 0.4962606837606838, + "grad_norm": 0.5490350723266602, + "learning_rate": 0.00019259283983545365, + "loss": 1.2462, + "step": 2787 + }, + { + "epoch": 0.4964387464387464, + "grad_norm": 0.5788847208023071, + "learning_rate": 0.0001925875520959154, + "loss": 1.3485, + "step": 2788 + }, + { + "epoch": 0.4966168091168091, + "grad_norm": 0.46184736490249634, + "learning_rate": 0.00019258226254231643, + "loss": 0.8673, + "step": 2789 + }, + { + "epoch": 0.4967948717948718, + "grad_norm": 0.4890633225440979, + "learning_rate": 0.0001925769711747603, + "loss": 0.9474, + "step": 2790 + }, + { + "epoch": 0.49697293447293445, + "grad_norm": 0.5719282627105713, + "learning_rate": 0.00019257167799335078, + "loss": 1.2532, + "step": 2791 + }, + { + "epoch": 0.49715099715099714, + "grad_norm": 0.5385584235191345, + "learning_rate": 0.0001925663829981915, + "loss": 1.1326, + "step": 2792 + }, + { + "epoch": 0.49732905982905984, + "grad_norm": 0.5339545011520386, + "learning_rate": 0.00019256108618938625, + "loss": 1.1362, + "step": 2793 + }, + { + "epoch": 0.49750712250712253, + "grad_norm": 0.5017803907394409, + "learning_rate": 0.00019255578756703878, + "loss": 1.0449, + "step": 2794 + }, + { + "epoch": 0.4976851851851852, + "grad_norm": 0.6004226803779602, + "learning_rate": 0.00019255048713125294, + "loss": 0.9346, + "step": 2795 + }, + { + "epoch": 0.49786324786324787, + "grad_norm": 0.44581490755081177, + "learning_rate": 0.00019254518488213255, + "loss": 1.038, + "step": 2796 + }, + { + "epoch": 0.49804131054131057, + "grad_norm": 0.5180951356887817, + "learning_rate": 0.00019253988081978151, + "loss": 1.0479, + "step": 2797 + }, + { + "epoch": 0.4982193732193732, + "grad_norm": 0.53944993019104, + "learning_rate": 0.00019253457494430376, + "loss": 1.2598, + "step": 2798 + }, + { + "epoch": 0.4983974358974359, + "grad_norm": 0.5633010268211365, + "learning_rate": 0.00019252926725580322, + "loss": 1.205, + "step": 2799 + }, + { + "epoch": 0.4985754985754986, + "grad_norm": 0.6653175950050354, + "learning_rate": 0.0001925239577543839, + "loss": 1.2383, + "step": 2800 + }, + { + "epoch": 0.49875356125356124, + "grad_norm": 0.5083333849906921, + "learning_rate": 0.00019251864644014984, + "loss": 1.0649, + "step": 2801 + }, + { + "epoch": 0.49893162393162394, + "grad_norm": 0.4842020571231842, + "learning_rate": 0.00019251333331320506, + "loss": 1.1991, + "step": 2802 + }, + { + "epoch": 0.49910968660968663, + "grad_norm": 0.47987112402915955, + "learning_rate": 0.00019250801837365373, + "loss": 1.1686, + "step": 2803 + }, + { + "epoch": 0.49928774928774927, + "grad_norm": 0.5316333770751953, + "learning_rate": 0.00019250270162159992, + "loss": 1.1759, + "step": 2804 + }, + { + "epoch": 0.49946581196581197, + "grad_norm": 0.5015079379081726, + "learning_rate": 0.00019249738305714787, + "loss": 0.9424, + "step": 2805 + }, + { + "epoch": 0.49964387464387466, + "grad_norm": 0.6488274931907654, + "learning_rate": 0.00019249206268040172, + "loss": 1.066, + "step": 2806 + }, + { + "epoch": 0.4998219373219373, + "grad_norm": 0.40364864468574524, + "learning_rate": 0.00019248674049146574, + "loss": 0.6998, + "step": 2807 + }, + { + "epoch": 0.5, + "grad_norm": 0.5535672903060913, + "learning_rate": 0.00019248141649044423, + "loss": 1.2207, + "step": 2808 + }, + { + "epoch": 0.5, + "eval_loss": 1.1072274446487427, + "eval_runtime": 28.6913, + "eval_samples_per_second": 36.283, + "eval_steps_per_second": 18.159, + "step": 2808 + }, + { + "epoch": 0.5001780626780626, + "grad_norm": 0.4834389090538025, + "learning_rate": 0.00019247609067744143, + "loss": 1.1686, + "step": 2809 + }, + { + "epoch": 0.5003561253561254, + "grad_norm": 0.5007249712944031, + "learning_rate": 0.00019247076305256176, + "loss": 1.1343, + "step": 2810 + }, + { + "epoch": 0.500534188034188, + "grad_norm": 0.4773348271846771, + "learning_rate": 0.00019246543361590957, + "loss": 0.9324, + "step": 2811 + }, + { + "epoch": 0.5007122507122507, + "grad_norm": 0.47324609756469727, + "learning_rate": 0.0001924601023675893, + "loss": 1.0223, + "step": 2812 + }, + { + "epoch": 0.5008903133903134, + "grad_norm": 0.5583845973014832, + "learning_rate": 0.00019245476930770537, + "loss": 1.1328, + "step": 2813 + }, + { + "epoch": 0.5010683760683761, + "grad_norm": 0.4814579486846924, + "learning_rate": 0.00019244943443636232, + "loss": 1.0528, + "step": 2814 + }, + { + "epoch": 0.5012464387464387, + "grad_norm": 0.4996104836463928, + "learning_rate": 0.00019244409775366465, + "loss": 1.2482, + "step": 2815 + }, + { + "epoch": 0.5014245014245015, + "grad_norm": 0.47870904207229614, + "learning_rate": 0.0001924387592597169, + "loss": 0.9452, + "step": 2816 + }, + { + "epoch": 0.5016025641025641, + "grad_norm": 0.5617441534996033, + "learning_rate": 0.0001924334189546237, + "loss": 1.378, + "step": 2817 + }, + { + "epoch": 0.5017806267806267, + "grad_norm": 0.4872083365917206, + "learning_rate": 0.00019242807683848967, + "loss": 1.1571, + "step": 2818 + }, + { + "epoch": 0.5019586894586895, + "grad_norm": 0.5147804021835327, + "learning_rate": 0.00019242273291141947, + "loss": 1.1086, + "step": 2819 + }, + { + "epoch": 0.5021367521367521, + "grad_norm": 0.4698995351791382, + "learning_rate": 0.00019241738717351784, + "loss": 1.1579, + "step": 2820 + }, + { + "epoch": 0.5023148148148148, + "grad_norm": 0.5158926844596863, + "learning_rate": 0.00019241203962488946, + "loss": 1.2763, + "step": 2821 + }, + { + "epoch": 0.5024928774928775, + "grad_norm": 0.5218976736068726, + "learning_rate": 0.00019240669026563914, + "loss": 1.0633, + "step": 2822 + }, + { + "epoch": 0.5026709401709402, + "grad_norm": 0.5511452555656433, + "learning_rate": 0.0001924013390958717, + "loss": 0.9939, + "step": 2823 + }, + { + "epoch": 0.5028490028490028, + "grad_norm": 0.5227555632591248, + "learning_rate": 0.00019239598611569191, + "loss": 1.2478, + "step": 2824 + }, + { + "epoch": 0.5030270655270656, + "grad_norm": 0.5444719791412354, + "learning_rate": 0.00019239063132520475, + "loss": 1.1574, + "step": 2825 + }, + { + "epoch": 0.5032051282051282, + "grad_norm": 0.4752781093120575, + "learning_rate": 0.0001923852747245151, + "loss": 0.9034, + "step": 2826 + }, + { + "epoch": 0.5033831908831908, + "grad_norm": 0.5286496877670288, + "learning_rate": 0.00019237991631372792, + "loss": 1.1391, + "step": 2827 + }, + { + "epoch": 0.5035612535612536, + "grad_norm": 0.5009933710098267, + "learning_rate": 0.00019237455609294815, + "loss": 1.2178, + "step": 2828 + }, + { + "epoch": 0.5037393162393162, + "grad_norm": 0.5012276768684387, + "learning_rate": 0.00019236919406228085, + "loss": 0.9877, + "step": 2829 + }, + { + "epoch": 0.5039173789173789, + "grad_norm": 0.576508104801178, + "learning_rate": 0.00019236383022183106, + "loss": 1.1299, + "step": 2830 + }, + { + "epoch": 0.5040954415954416, + "grad_norm": 0.4716590642929077, + "learning_rate": 0.0001923584645717039, + "loss": 1.0451, + "step": 2831 + }, + { + "epoch": 0.5042735042735043, + "grad_norm": 0.5817418098449707, + "learning_rate": 0.00019235309711200448, + "loss": 1.0911, + "step": 2832 + }, + { + "epoch": 0.5044515669515669, + "grad_norm": 0.5695745944976807, + "learning_rate": 0.000192347727842838, + "loss": 1.0229, + "step": 2833 + }, + { + "epoch": 0.5046296296296297, + "grad_norm": 0.49127066135406494, + "learning_rate": 0.00019234235676430958, + "loss": 1.1377, + "step": 2834 + }, + { + "epoch": 0.5048076923076923, + "grad_norm": 0.5426172614097595, + "learning_rate": 0.00019233698387652453, + "loss": 1.2427, + "step": 2835 + }, + { + "epoch": 0.5049857549857549, + "grad_norm": 0.5342385172843933, + "learning_rate": 0.0001923316091795881, + "loss": 1.1427, + "step": 2836 + }, + { + "epoch": 0.5051638176638177, + "grad_norm": 0.5480486750602722, + "learning_rate": 0.00019232623267360558, + "loss": 1.0647, + "step": 2837 + }, + { + "epoch": 0.5053418803418803, + "grad_norm": 0.4584530293941498, + "learning_rate": 0.00019232085435868235, + "loss": 1.0461, + "step": 2838 + }, + { + "epoch": 0.5055199430199431, + "grad_norm": 0.5992119908332825, + "learning_rate": 0.00019231547423492371, + "loss": 1.1456, + "step": 2839 + }, + { + "epoch": 0.5056980056980057, + "grad_norm": 0.514018177986145, + "learning_rate": 0.00019231009230243515, + "loss": 1.2559, + "step": 2840 + }, + { + "epoch": 0.5058760683760684, + "grad_norm": 0.5392283797264099, + "learning_rate": 0.0001923047085613221, + "loss": 1.044, + "step": 2841 + }, + { + "epoch": 0.5060541310541311, + "grad_norm": 0.4486566483974457, + "learning_rate": 0.00019229932301169, + "loss": 1.0679, + "step": 2842 + }, + { + "epoch": 0.5062321937321937, + "grad_norm": 0.4523460566997528, + "learning_rate": 0.00019229393565364442, + "loss": 1.1651, + "step": 2843 + }, + { + "epoch": 0.5064102564102564, + "grad_norm": 0.6032688021659851, + "learning_rate": 0.0001922885464872909, + "loss": 1.15, + "step": 2844 + }, + { + "epoch": 0.5065883190883191, + "grad_norm": 0.5883688926696777, + "learning_rate": 0.000192283155512735, + "loss": 1.2179, + "step": 2845 + }, + { + "epoch": 0.5067663817663818, + "grad_norm": 0.5534378886222839, + "learning_rate": 0.00019227776273008238, + "loss": 1.0387, + "step": 2846 + }, + { + "epoch": 0.5069444444444444, + "grad_norm": 0.5899033546447754, + "learning_rate": 0.00019227236813943872, + "loss": 1.0812, + "step": 2847 + }, + { + "epoch": 0.5071225071225072, + "grad_norm": 0.5718855261802673, + "learning_rate": 0.00019226697174090965, + "loss": 1.1375, + "step": 2848 + }, + { + "epoch": 0.5073005698005698, + "grad_norm": 0.5080967545509338, + "learning_rate": 0.00019226157353460094, + "loss": 1.1421, + "step": 2849 + }, + { + "epoch": 0.5074786324786325, + "grad_norm": 0.5253677368164062, + "learning_rate": 0.0001922561735206184, + "loss": 1.0166, + "step": 2850 + }, + { + "epoch": 0.5076566951566952, + "grad_norm": 0.47797444462776184, + "learning_rate": 0.00019225077169906772, + "loss": 1.0504, + "step": 2851 + }, + { + "epoch": 0.5078347578347578, + "grad_norm": 0.4911690652370453, + "learning_rate": 0.0001922453680700548, + "loss": 1.0629, + "step": 2852 + }, + { + "epoch": 0.5080128205128205, + "grad_norm": 0.49678200483322144, + "learning_rate": 0.00019223996263368557, + "loss": 1.1672, + "step": 2853 + }, + { + "epoch": 0.5081908831908832, + "grad_norm": 0.5451810359954834, + "learning_rate": 0.00019223455539006586, + "loss": 1.3031, + "step": 2854 + }, + { + "epoch": 0.5083689458689459, + "grad_norm": 0.5708984136581421, + "learning_rate": 0.00019222914633930166, + "loss": 1.0986, + "step": 2855 + }, + { + "epoch": 0.5085470085470085, + "grad_norm": 0.47232356667518616, + "learning_rate": 0.00019222373548149888, + "loss": 1.0449, + "step": 2856 + }, + { + "epoch": 0.5087250712250713, + "grad_norm": 0.6027610898017883, + "learning_rate": 0.0001922183228167636, + "loss": 0.862, + "step": 2857 + }, + { + "epoch": 0.5089031339031339, + "grad_norm": 0.5211802124977112, + "learning_rate": 0.00019221290834520188, + "loss": 1.1048, + "step": 2858 + }, + { + "epoch": 0.5090811965811965, + "grad_norm": 0.45101237297058105, + "learning_rate": 0.00019220749206691972, + "loss": 1.0046, + "step": 2859 + }, + { + "epoch": 0.5092592592592593, + "grad_norm": 0.5526158213615417, + "learning_rate": 0.00019220207398202335, + "loss": 1.2275, + "step": 2860 + }, + { + "epoch": 0.5094373219373219, + "grad_norm": 0.48322010040283203, + "learning_rate": 0.00019219665409061885, + "loss": 0.9974, + "step": 2861 + }, + { + "epoch": 0.5096153846153846, + "grad_norm": 0.4775219261646271, + "learning_rate": 0.00019219123239281244, + "loss": 1.1852, + "step": 2862 + }, + { + "epoch": 0.5097934472934473, + "grad_norm": 0.46184200048446655, + "learning_rate": 0.00019218580888871034, + "loss": 0.9393, + "step": 2863 + }, + { + "epoch": 0.50997150997151, + "grad_norm": 0.47495174407958984, + "learning_rate": 0.00019218038357841883, + "loss": 0.9631, + "step": 2864 + }, + { + "epoch": 0.5101495726495726, + "grad_norm": 0.48600029945373535, + "learning_rate": 0.00019217495646204418, + "loss": 1.0498, + "step": 2865 + }, + { + "epoch": 0.5103276353276354, + "grad_norm": 0.5801547169685364, + "learning_rate": 0.00019216952753969274, + "loss": 1.2181, + "step": 2866 + }, + { + "epoch": 0.510505698005698, + "grad_norm": 0.5082106590270996, + "learning_rate": 0.00019216409681147085, + "loss": 1.2009, + "step": 2867 + }, + { + "epoch": 0.5106837606837606, + "grad_norm": 0.4184330701828003, + "learning_rate": 0.00019215866427748493, + "loss": 0.8462, + "step": 2868 + }, + { + "epoch": 0.5108618233618234, + "grad_norm": 0.518099844455719, + "learning_rate": 0.00019215322993784147, + "loss": 1.2091, + "step": 2869 + }, + { + "epoch": 0.511039886039886, + "grad_norm": 0.569464921951294, + "learning_rate": 0.0001921477937926469, + "loss": 1.0264, + "step": 2870 + }, + { + "epoch": 0.5112179487179487, + "grad_norm": 0.526767909526825, + "learning_rate": 0.00019214235584200768, + "loss": 1.1192, + "step": 2871 + }, + { + "epoch": 0.5113960113960114, + "grad_norm": 0.6511057019233704, + "learning_rate": 0.00019213691608603047, + "loss": 1.3193, + "step": 2872 + }, + { + "epoch": 0.5115740740740741, + "grad_norm": 0.48536401987075806, + "learning_rate": 0.00019213147452482173, + "loss": 1.1671, + "step": 2873 + }, + { + "epoch": 0.5117521367521367, + "grad_norm": 0.7972469329833984, + "learning_rate": 0.00019212603115848818, + "loss": 1.1393, + "step": 2874 + }, + { + "epoch": 0.5119301994301995, + "grad_norm": 0.5543264746665955, + "learning_rate": 0.00019212058598713642, + "loss": 1.1436, + "step": 2875 + }, + { + "epoch": 0.5121082621082621, + "grad_norm": 0.49688720703125, + "learning_rate": 0.0001921151390108731, + "loss": 1.0897, + "step": 2876 + }, + { + "epoch": 0.5122863247863247, + "grad_norm": 0.4928736090660095, + "learning_rate": 0.000192109690229805, + "loss": 1.2426, + "step": 2877 + }, + { + "epoch": 0.5124643874643875, + "grad_norm": 0.4917896091938019, + "learning_rate": 0.0001921042396440389, + "loss": 1.0047, + "step": 2878 + }, + { + "epoch": 0.5126424501424501, + "grad_norm": 0.5485204458236694, + "learning_rate": 0.00019209878725368152, + "loss": 1.2615, + "step": 2879 + }, + { + "epoch": 0.5128205128205128, + "grad_norm": 0.5229470133781433, + "learning_rate": 0.0001920933330588397, + "loss": 1.3249, + "step": 2880 + }, + { + "epoch": 0.5129985754985755, + "grad_norm": 0.4783077538013458, + "learning_rate": 0.00019208787705962037, + "loss": 1.2004, + "step": 2881 + }, + { + "epoch": 0.5131766381766382, + "grad_norm": 0.5106910467147827, + "learning_rate": 0.00019208241925613035, + "loss": 1.1745, + "step": 2882 + }, + { + "epoch": 0.5133547008547008, + "grad_norm": 0.5308730006217957, + "learning_rate": 0.00019207695964847666, + "loss": 0.9706, + "step": 2883 + }, + { + "epoch": 0.5135327635327636, + "grad_norm": 0.5489775538444519, + "learning_rate": 0.00019207149823676617, + "loss": 1.0073, + "step": 2884 + }, + { + "epoch": 0.5137108262108262, + "grad_norm": 0.4992835521697998, + "learning_rate": 0.00019206603502110596, + "loss": 1.1053, + "step": 2885 + }, + { + "epoch": 0.5138888888888888, + "grad_norm": 0.5304922461509705, + "learning_rate": 0.00019206057000160302, + "loss": 1.0565, + "step": 2886 + }, + { + "epoch": 0.5140669515669516, + "grad_norm": 0.46411609649658203, + "learning_rate": 0.00019205510317836448, + "loss": 0.9202, + "step": 2887 + }, + { + "epoch": 0.5142450142450142, + "grad_norm": 0.5236835479736328, + "learning_rate": 0.0001920496345514974, + "loss": 0.9075, + "step": 2888 + }, + { + "epoch": 0.5144230769230769, + "grad_norm": 0.4416964054107666, + "learning_rate": 0.00019204416412110895, + "loss": 0.9225, + "step": 2889 + }, + { + "epoch": 0.5146011396011396, + "grad_norm": 0.5470940470695496, + "learning_rate": 0.00019203869188730633, + "loss": 1.2195, + "step": 2890 + }, + { + "epoch": 0.5147792022792023, + "grad_norm": 0.5380414128303528, + "learning_rate": 0.0001920332178501967, + "loss": 1.0731, + "step": 2891 + }, + { + "epoch": 0.5149572649572649, + "grad_norm": 0.4405716359615326, + "learning_rate": 0.00019202774200988737, + "loss": 0.8739, + "step": 2892 + }, + { + "epoch": 0.5151353276353277, + "grad_norm": 0.5222984552383423, + "learning_rate": 0.0001920222643664856, + "loss": 1.1806, + "step": 2893 + }, + { + "epoch": 0.5153133903133903, + "grad_norm": 0.48545539379119873, + "learning_rate": 0.0001920167849200987, + "loss": 0.9939, + "step": 2894 + }, + { + "epoch": 0.5154914529914529, + "grad_norm": 0.45078009366989136, + "learning_rate": 0.0001920113036708341, + "loss": 1.0085, + "step": 2895 + }, + { + "epoch": 0.5156695156695157, + "grad_norm": 0.5029830932617188, + "learning_rate": 0.00019200582061879913, + "loss": 1.1095, + "step": 2896 + }, + { + "epoch": 0.5158475783475783, + "grad_norm": 0.5316143035888672, + "learning_rate": 0.00019200033576410118, + "loss": 0.9883, + "step": 2897 + }, + { + "epoch": 0.5160256410256411, + "grad_norm": 0.5282100439071655, + "learning_rate": 0.0001919948491068478, + "loss": 1.1441, + "step": 2898 + }, + { + "epoch": 0.5162037037037037, + "grad_norm": 0.5145367980003357, + "learning_rate": 0.00019198936064714647, + "loss": 1.1999, + "step": 2899 + }, + { + "epoch": 0.5163817663817664, + "grad_norm": 0.5385651588439941, + "learning_rate": 0.00019198387038510468, + "loss": 1.1831, + "step": 2900 + }, + { + "epoch": 0.5165598290598291, + "grad_norm": 0.4971916377544403, + "learning_rate": 0.00019197837832083002, + "loss": 1.2518, + "step": 2901 + }, + { + "epoch": 0.5167378917378918, + "grad_norm": 0.5253807306289673, + "learning_rate": 0.00019197288445443016, + "loss": 1.0788, + "step": 2902 + }, + { + "epoch": 0.5169159544159544, + "grad_norm": 0.49724945425987244, + "learning_rate": 0.00019196738878601263, + "loss": 1.0985, + "step": 2903 + }, + { + "epoch": 0.5170940170940171, + "grad_norm": 0.5327325463294983, + "learning_rate": 0.0001919618913156852, + "loss": 1.2862, + "step": 2904 + }, + { + "epoch": 0.5172720797720798, + "grad_norm": 0.639999270439148, + "learning_rate": 0.00019195639204355554, + "loss": 1.2052, + "step": 2905 + }, + { + "epoch": 0.5174501424501424, + "grad_norm": 0.4630785584449768, + "learning_rate": 0.0001919508909697314, + "loss": 1.1157, + "step": 2906 + }, + { + "epoch": 0.5176282051282052, + "grad_norm": 0.513949990272522, + "learning_rate": 0.00019194538809432055, + "loss": 1.0047, + "step": 2907 + }, + { + "epoch": 0.5178062678062678, + "grad_norm": 0.488034725189209, + "learning_rate": 0.0001919398834174308, + "loss": 0.9008, + "step": 2908 + }, + { + "epoch": 0.5179843304843305, + "grad_norm": 0.4892788529396057, + "learning_rate": 0.00019193437693917006, + "loss": 1.1024, + "step": 2909 + }, + { + "epoch": 0.5181623931623932, + "grad_norm": 0.5503842830657959, + "learning_rate": 0.00019192886865964618, + "loss": 1.2283, + "step": 2910 + }, + { + "epoch": 0.5183404558404558, + "grad_norm": 0.48885393142700195, + "learning_rate": 0.00019192335857896707, + "loss": 0.9522, + "step": 2911 + }, + { + "epoch": 0.5185185185185185, + "grad_norm": 0.5479527115821838, + "learning_rate": 0.00019191784669724072, + "loss": 1.1616, + "step": 2912 + }, + { + "epoch": 0.5186965811965812, + "grad_norm": 0.42701148986816406, + "learning_rate": 0.00019191233301457506, + "loss": 0.8434, + "step": 2913 + }, + { + "epoch": 0.5188746438746439, + "grad_norm": 0.4273422658443451, + "learning_rate": 0.00019190681753107822, + "loss": 0.8316, + "step": 2914 + }, + { + "epoch": 0.5190527065527065, + "grad_norm": 0.5047736763954163, + "learning_rate": 0.00019190130024685818, + "loss": 1.171, + "step": 2915 + }, + { + "epoch": 0.5192307692307693, + "grad_norm": 0.5221177935600281, + "learning_rate": 0.00019189578116202307, + "loss": 1.0256, + "step": 2916 + }, + { + "epoch": 0.5194088319088319, + "grad_norm": 0.4782322943210602, + "learning_rate": 0.00019189026027668105, + "loss": 0.8598, + "step": 2917 + }, + { + "epoch": 0.5195868945868946, + "grad_norm": 0.5627185702323914, + "learning_rate": 0.00019188473759094022, + "loss": 1.1825, + "step": 2918 + }, + { + "epoch": 0.5197649572649573, + "grad_norm": 0.5036423206329346, + "learning_rate": 0.00019187921310490888, + "loss": 1.0881, + "step": 2919 + }, + { + "epoch": 0.51994301994302, + "grad_norm": 0.4271143972873688, + "learning_rate": 0.0001918736868186952, + "loss": 0.9265, + "step": 2920 + }, + { + "epoch": 0.5201210826210826, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00019186815873240747, + "loss": 1.196, + "step": 2921 + }, + { + "epoch": 0.5202991452991453, + "grad_norm": 0.5494198203086853, + "learning_rate": 0.00019186262884615402, + "loss": 1.1207, + "step": 2922 + }, + { + "epoch": 0.520477207977208, + "grad_norm": 0.5305119752883911, + "learning_rate": 0.0001918570971600432, + "loss": 1.0393, + "step": 2923 + }, + { + "epoch": 0.5206552706552706, + "grad_norm": 0.46713170409202576, + "learning_rate": 0.00019185156367418333, + "loss": 0.9583, + "step": 2924 + }, + { + "epoch": 0.5208333333333334, + "grad_norm": 0.597776472568512, + "learning_rate": 0.00019184602838868292, + "loss": 1.2978, + "step": 2925 + }, + { + "epoch": 0.521011396011396, + "grad_norm": 0.520976722240448, + "learning_rate": 0.00019184049130365036, + "loss": 1.0515, + "step": 2926 + }, + { + "epoch": 0.5211894586894587, + "grad_norm": 0.5266290307044983, + "learning_rate": 0.00019183495241919415, + "loss": 1.0437, + "step": 2927 + }, + { + "epoch": 0.5213675213675214, + "grad_norm": 0.50911545753479, + "learning_rate": 0.00019182941173542285, + "loss": 0.9977, + "step": 2928 + }, + { + "epoch": 0.521545584045584, + "grad_norm": 0.4924670457839966, + "learning_rate": 0.00019182386925244496, + "loss": 0.9309, + "step": 2929 + }, + { + "epoch": 0.5217236467236467, + "grad_norm": 0.4979301393032074, + "learning_rate": 0.00019181832497036912, + "loss": 0.87, + "step": 2930 + }, + { + "epoch": 0.5219017094017094, + "grad_norm": 0.6307916045188904, + "learning_rate": 0.0001918127788893039, + "loss": 1.2159, + "step": 2931 + }, + { + "epoch": 0.5220797720797721, + "grad_norm": 0.4915660619735718, + "learning_rate": 0.00019180723100935802, + "loss": 1.0828, + "step": 2932 + }, + { + "epoch": 0.5222578347578347, + "grad_norm": 0.4312742352485657, + "learning_rate": 0.00019180168133064017, + "loss": 1.0496, + "step": 2933 + }, + { + "epoch": 0.5224358974358975, + "grad_norm": 0.6006124019622803, + "learning_rate": 0.00019179612985325908, + "loss": 1.0751, + "step": 2934 + }, + { + "epoch": 0.5226139601139601, + "grad_norm": 0.5332220196723938, + "learning_rate": 0.0001917905765773235, + "loss": 1.2601, + "step": 2935 + }, + { + "epoch": 0.5227920227920227, + "grad_norm": 0.4877954423427582, + "learning_rate": 0.00019178502150294223, + "loss": 1.2279, + "step": 2936 + }, + { + "epoch": 0.5229700854700855, + "grad_norm": 0.5975968837738037, + "learning_rate": 0.00019177946463022418, + "loss": 1.3371, + "step": 2937 + }, + { + "epoch": 0.5231481481481481, + "grad_norm": 0.5363923907279968, + "learning_rate": 0.00019177390595927815, + "loss": 1.0705, + "step": 2938 + }, + { + "epoch": 0.5233262108262108, + "grad_norm": 0.4314909875392914, + "learning_rate": 0.0001917683454902131, + "loss": 0.9172, + "step": 2939 + }, + { + "epoch": 0.5235042735042735, + "grad_norm": 0.46187883615493774, + "learning_rate": 0.0001917627832231379, + "loss": 1.1201, + "step": 2940 + }, + { + "epoch": 0.5236823361823362, + "grad_norm": 0.4648260772228241, + "learning_rate": 0.00019175721915816162, + "loss": 1.1307, + "step": 2941 + }, + { + "epoch": 0.5238603988603988, + "grad_norm": 0.4427165687084198, + "learning_rate": 0.00019175165329539325, + "loss": 0.9459, + "step": 2942 + }, + { + "epoch": 0.5240384615384616, + "grad_norm": 0.4645056128501892, + "learning_rate": 0.0001917460856349418, + "loss": 0.9176, + "step": 2943 + }, + { + "epoch": 0.5242165242165242, + "grad_norm": 0.4939568042755127, + "learning_rate": 0.0001917405161769164, + "loss": 1.1056, + "step": 2944 + }, + { + "epoch": 0.5243945868945868, + "grad_norm": 0.6057310104370117, + "learning_rate": 0.00019173494492142617, + "loss": 1.2714, + "step": 2945 + }, + { + "epoch": 0.5245726495726496, + "grad_norm": 0.5038546323776245, + "learning_rate": 0.00019172937186858025, + "loss": 0.911, + "step": 2946 + }, + { + "epoch": 0.5247507122507122, + "grad_norm": 0.5521321296691895, + "learning_rate": 0.00019172379701848784, + "loss": 1.0781, + "step": 2947 + }, + { + "epoch": 0.5249287749287749, + "grad_norm": 0.516979455947876, + "learning_rate": 0.00019171822037125817, + "loss": 1.1051, + "step": 2948 + }, + { + "epoch": 0.5251068376068376, + "grad_norm": 0.5443150997161865, + "learning_rate": 0.0001917126419270005, + "loss": 1.0802, + "step": 2949 + }, + { + "epoch": 0.5252849002849003, + "grad_norm": 0.5373311042785645, + "learning_rate": 0.00019170706168582412, + "loss": 0.9313, + "step": 2950 + }, + { + "epoch": 0.5254629629629629, + "grad_norm": 0.7511917948722839, + "learning_rate": 0.0001917014796478384, + "loss": 1.1958, + "step": 2951 + }, + { + "epoch": 0.5256410256410257, + "grad_norm": 0.49893468618392944, + "learning_rate": 0.00019169589581315263, + "loss": 0.9387, + "step": 2952 + }, + { + "epoch": 0.5258190883190883, + "grad_norm": 0.48010289669036865, + "learning_rate": 0.00019169031018187628, + "loss": 1.2459, + "step": 2953 + }, + { + "epoch": 0.5259971509971509, + "grad_norm": 0.48768678307533264, + "learning_rate": 0.0001916847227541188, + "loss": 1.0127, + "step": 2954 + }, + { + "epoch": 0.5261752136752137, + "grad_norm": 0.5973068475723267, + "learning_rate": 0.00019167913352998963, + "loss": 1.1685, + "step": 2955 + }, + { + "epoch": 0.5263532763532763, + "grad_norm": 0.5567806959152222, + "learning_rate": 0.00019167354250959826, + "loss": 1.142, + "step": 2956 + }, + { + "epoch": 0.5265313390313391, + "grad_norm": 0.47819700837135315, + "learning_rate": 0.00019166794969305428, + "loss": 0.712, + "step": 2957 + }, + { + "epoch": 0.5267094017094017, + "grad_norm": 0.5191744565963745, + "learning_rate": 0.00019166235508046725, + "loss": 1.2208, + "step": 2958 + }, + { + "epoch": 0.5268874643874644, + "grad_norm": 0.4987856149673462, + "learning_rate": 0.00019165675867194675, + "loss": 1.0466, + "step": 2959 + }, + { + "epoch": 0.5270655270655271, + "grad_norm": 0.5017665028572083, + "learning_rate": 0.0001916511604676025, + "loss": 1.1236, + "step": 2960 + }, + { + "epoch": 0.5272435897435898, + "grad_norm": 0.5115348696708679, + "learning_rate": 0.00019164556046754415, + "loss": 1.1497, + "step": 2961 + }, + { + "epoch": 0.5274216524216524, + "grad_norm": 0.4934345781803131, + "learning_rate": 0.0001916399586718814, + "loss": 1.0183, + "step": 2962 + }, + { + "epoch": 0.5275997150997151, + "grad_norm": 0.5033719539642334, + "learning_rate": 0.00019163435508072404, + "loss": 1.0256, + "step": 2963 + }, + { + "epoch": 0.5277777777777778, + "grad_norm": 0.5325372219085693, + "learning_rate": 0.00019162874969418184, + "loss": 1.1384, + "step": 2964 + }, + { + "epoch": 0.5279558404558404, + "grad_norm": 0.4901772141456604, + "learning_rate": 0.00019162314251236465, + "loss": 1.0831, + "step": 2965 + }, + { + "epoch": 0.5281339031339032, + "grad_norm": 0.4743805229663849, + "learning_rate": 0.0001916175335353823, + "loss": 1.1894, + "step": 2966 + }, + { + "epoch": 0.5283119658119658, + "grad_norm": 0.5439450740814209, + "learning_rate": 0.00019161192276334466, + "loss": 1.2066, + "step": 2967 + }, + { + "epoch": 0.5284900284900285, + "grad_norm": 0.5123090744018555, + "learning_rate": 0.00019160631019636174, + "loss": 1.1829, + "step": 2968 + }, + { + "epoch": 0.5286680911680912, + "grad_norm": 0.5995343923568726, + "learning_rate": 0.00019160069583454346, + "loss": 1.4872, + "step": 2969 + }, + { + "epoch": 0.5288461538461539, + "grad_norm": 0.4596657156944275, + "learning_rate": 0.00019159507967799985, + "loss": 0.8948, + "step": 2970 + }, + { + "epoch": 0.5290242165242165, + "grad_norm": 0.5533682107925415, + "learning_rate": 0.0001915894617268409, + "loss": 1.1779, + "step": 2971 + }, + { + "epoch": 0.5292022792022792, + "grad_norm": 0.3860718309879303, + "learning_rate": 0.00019158384198117673, + "loss": 0.6424, + "step": 2972 + }, + { + "epoch": 0.5293803418803419, + "grad_norm": 0.47424063086509705, + "learning_rate": 0.0001915782204411174, + "loss": 1.1592, + "step": 2973 + }, + { + "epoch": 0.5295584045584045, + "grad_norm": 0.5050228834152222, + "learning_rate": 0.00019157259710677309, + "loss": 1.1971, + "step": 2974 + }, + { + "epoch": 0.5297364672364673, + "grad_norm": 0.6080113649368286, + "learning_rate": 0.00019156697197825396, + "loss": 1.1511, + "step": 2975 + }, + { + "epoch": 0.5299145299145299, + "grad_norm": 0.4805932641029358, + "learning_rate": 0.00019156134505567024, + "loss": 1.1033, + "step": 2976 + }, + { + "epoch": 0.5300925925925926, + "grad_norm": 0.4835345447063446, + "learning_rate": 0.00019155571633913215, + "loss": 1.1832, + "step": 2977 + }, + { + "epoch": 0.5302706552706553, + "grad_norm": 0.5183725953102112, + "learning_rate": 0.00019155008582875, + "loss": 0.9221, + "step": 2978 + }, + { + "epoch": 0.530448717948718, + "grad_norm": 0.48015761375427246, + "learning_rate": 0.00019154445352463412, + "loss": 1.045, + "step": 2979 + }, + { + "epoch": 0.5306267806267806, + "grad_norm": 0.4670043885707855, + "learning_rate": 0.0001915388194268948, + "loss": 0.9025, + "step": 2980 + }, + { + "epoch": 0.5308048433048433, + "grad_norm": 0.5048824548721313, + "learning_rate": 0.0001915331835356425, + "loss": 1.0681, + "step": 2981 + }, + { + "epoch": 0.530982905982906, + "grad_norm": 0.4785633981227875, + "learning_rate": 0.00019152754585098758, + "loss": 1.0097, + "step": 2982 + }, + { + "epoch": 0.5311609686609686, + "grad_norm": 0.4829573333263397, + "learning_rate": 0.00019152190637304056, + "loss": 1.0856, + "step": 2983 + }, + { + "epoch": 0.5313390313390314, + "grad_norm": 0.5425563454627991, + "learning_rate": 0.00019151626510191189, + "loss": 1.2313, + "step": 2984 + }, + { + "epoch": 0.531517094017094, + "grad_norm": 0.5532251596450806, + "learning_rate": 0.0001915106220377121, + "loss": 1.0328, + "step": 2985 + }, + { + "epoch": 0.5316951566951567, + "grad_norm": 0.47016972303390503, + "learning_rate": 0.0001915049771805518, + "loss": 1.2003, + "step": 2986 + }, + { + "epoch": 0.5318732193732194, + "grad_norm": 0.5241743326187134, + "learning_rate": 0.00019149933053054153, + "loss": 1.046, + "step": 2987 + }, + { + "epoch": 0.532051282051282, + "grad_norm": 0.5043526887893677, + "learning_rate": 0.00019149368208779197, + "loss": 1.0022, + "step": 2988 + }, + { + "epoch": 0.5322293447293447, + "grad_norm": 0.5563312768936157, + "learning_rate": 0.00019148803185241374, + "loss": 1.1017, + "step": 2989 + }, + { + "epoch": 0.5324074074074074, + "grad_norm": 0.5414231419563293, + "learning_rate": 0.00019148237982451763, + "loss": 0.9649, + "step": 2990 + }, + { + "epoch": 0.5325854700854701, + "grad_norm": 0.5452231764793396, + "learning_rate": 0.0001914767260042143, + "loss": 1.2281, + "step": 2991 + }, + { + "epoch": 0.5327635327635327, + "grad_norm": 0.5500698685646057, + "learning_rate": 0.00019147107039161454, + "loss": 1.2865, + "step": 2992 + }, + { + "epoch": 0.5329415954415955, + "grad_norm": 0.49747416377067566, + "learning_rate": 0.00019146541298682918, + "loss": 1.1296, + "step": 2993 + }, + { + "epoch": 0.5331196581196581, + "grad_norm": 0.5684167742729187, + "learning_rate": 0.00019145975378996903, + "loss": 1.0685, + "step": 2994 + }, + { + "epoch": 0.5332977207977208, + "grad_norm": 0.5411235690116882, + "learning_rate": 0.00019145409280114502, + "loss": 1.1372, + "step": 2995 + }, + { + "epoch": 0.5334757834757835, + "grad_norm": 0.5006675720214844, + "learning_rate": 0.00019144843002046806, + "loss": 1.0688, + "step": 2996 + }, + { + "epoch": 0.5336538461538461, + "grad_norm": 0.4591315686702728, + "learning_rate": 0.00019144276544804908, + "loss": 1.1071, + "step": 2997 + }, + { + "epoch": 0.5338319088319088, + "grad_norm": 0.5615306496620178, + "learning_rate": 0.000191437099083999, + "loss": 1.1033, + "step": 2998 + }, + { + "epoch": 0.5340099715099715, + "grad_norm": 0.4986817240715027, + "learning_rate": 0.00019143143092842897, + "loss": 1.176, + "step": 2999 + }, + { + "epoch": 0.5341880341880342, + "grad_norm": 0.5017120242118835, + "learning_rate": 0.00019142576098144995, + "loss": 1.0174, + "step": 3000 + }, + { + "epoch": 0.5343660968660968, + "grad_norm": 0.508298397064209, + "learning_rate": 0.0001914200892431731, + "loss": 1.164, + "step": 3001 + }, + { + "epoch": 0.5345441595441596, + "grad_norm": 0.48068809509277344, + "learning_rate": 0.0001914144157137095, + "loss": 0.7959, + "step": 3002 + }, + { + "epoch": 0.5347222222222222, + "grad_norm": 0.6347028017044067, + "learning_rate": 0.0001914087403931703, + "loss": 1.1727, + "step": 3003 + }, + { + "epoch": 0.5349002849002849, + "grad_norm": 0.5558401942253113, + "learning_rate": 0.00019140306328166676, + "loss": 1.2282, + "step": 3004 + }, + { + "epoch": 0.5350783475783476, + "grad_norm": 0.5093596577644348, + "learning_rate": 0.00019139738437931004, + "loss": 1.3258, + "step": 3005 + }, + { + "epoch": 0.5352564102564102, + "grad_norm": 0.4653106927871704, + "learning_rate": 0.0001913917036862114, + "loss": 1.1062, + "step": 3006 + }, + { + "epoch": 0.5354344729344729, + "grad_norm": 0.48085781931877136, + "learning_rate": 0.00019138602120248222, + "loss": 0.9019, + "step": 3007 + }, + { + "epoch": 0.5356125356125356, + "grad_norm": 0.5174745321273804, + "learning_rate": 0.0001913803369282338, + "loss": 1.044, + "step": 3008 + }, + { + "epoch": 0.5357905982905983, + "grad_norm": 0.5359669327735901, + "learning_rate": 0.00019137465086357746, + "loss": 1.0723, + "step": 3009 + }, + { + "epoch": 0.5359686609686609, + "grad_norm": 0.5583470463752747, + "learning_rate": 0.00019136896300862467, + "loss": 1.2192, + "step": 3010 + }, + { + "epoch": 0.5361467236467237, + "grad_norm": 0.4905693829059601, + "learning_rate": 0.00019136327336348688, + "loss": 1.2372, + "step": 3011 + }, + { + "epoch": 0.5363247863247863, + "grad_norm": 0.5741264820098877, + "learning_rate": 0.0001913575819282755, + "loss": 1.1703, + "step": 3012 + }, + { + "epoch": 0.5365028490028491, + "grad_norm": 0.577033281326294, + "learning_rate": 0.0001913518887031021, + "loss": 1.1555, + "step": 3013 + }, + { + "epoch": 0.5366809116809117, + "grad_norm": 0.46795153617858887, + "learning_rate": 0.00019134619368807822, + "loss": 0.8583, + "step": 3014 + }, + { + "epoch": 0.5368589743589743, + "grad_norm": 0.5973345637321472, + "learning_rate": 0.0001913404968833154, + "loss": 1.1509, + "step": 3015 + }, + { + "epoch": 0.5370370370370371, + "grad_norm": 0.62020343542099, + "learning_rate": 0.00019133479828892531, + "loss": 1.0781, + "step": 3016 + }, + { + "epoch": 0.5372150997150997, + "grad_norm": 0.5342286229133606, + "learning_rate": 0.00019132909790501958, + "loss": 1.1556, + "step": 3017 + }, + { + "epoch": 0.5373931623931624, + "grad_norm": 0.49612846970558167, + "learning_rate": 0.0001913233957317099, + "loss": 0.9027, + "step": 3018 + }, + { + "epoch": 0.5375712250712251, + "grad_norm": 0.5403908491134644, + "learning_rate": 0.00019131769176910796, + "loss": 1.1125, + "step": 3019 + }, + { + "epoch": 0.5377492877492878, + "grad_norm": 0.4952050447463989, + "learning_rate": 0.0001913119860173256, + "loss": 1.2329, + "step": 3020 + }, + { + "epoch": 0.5379273504273504, + "grad_norm": 0.5877819657325745, + "learning_rate": 0.0001913062784764745, + "loss": 1.2855, + "step": 3021 + }, + { + "epoch": 0.5381054131054132, + "grad_norm": 0.49312907457351685, + "learning_rate": 0.00019130056914666655, + "loss": 1.0212, + "step": 3022 + }, + { + "epoch": 0.5382834757834758, + "grad_norm": 0.45544490218162537, + "learning_rate": 0.00019129485802801366, + "loss": 0.9748, + "step": 3023 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.5535242557525635, + "learning_rate": 0.00019128914512062762, + "loss": 1.2134, + "step": 3024 + }, + { + "epoch": 0.5386396011396012, + "grad_norm": 0.45369696617126465, + "learning_rate": 0.00019128343042462044, + "loss": 0.9964, + "step": 3025 + }, + { + "epoch": 0.5388176638176638, + "grad_norm": 0.6240725517272949, + "learning_rate": 0.00019127771394010406, + "loss": 1.425, + "step": 3026 + }, + { + "epoch": 0.5389957264957265, + "grad_norm": 0.4859573245048523, + "learning_rate": 0.0001912719956671905, + "loss": 1.087, + "step": 3027 + }, + { + "epoch": 0.5391737891737892, + "grad_norm": 0.47529762983322144, + "learning_rate": 0.0001912662756059918, + "loss": 0.9517, + "step": 3028 + }, + { + "epoch": 0.5393518518518519, + "grad_norm": 0.5317288637161255, + "learning_rate": 0.00019126055375661997, + "loss": 1.0945, + "step": 3029 + }, + { + "epoch": 0.5395299145299145, + "grad_norm": 0.55974280834198, + "learning_rate": 0.00019125483011918722, + "loss": 1.0794, + "step": 3030 + }, + { + "epoch": 0.5397079772079773, + "grad_norm": 0.48579123616218567, + "learning_rate": 0.0001912491046938056, + "loss": 1.1421, + "step": 3031 + }, + { + "epoch": 0.5398860398860399, + "grad_norm": 0.4917181134223938, + "learning_rate": 0.00019124337748058733, + "loss": 0.9708, + "step": 3032 + }, + { + "epoch": 0.5400641025641025, + "grad_norm": 0.525291383266449, + "learning_rate": 0.00019123764847964466, + "loss": 1.064, + "step": 3033 + }, + { + "epoch": 0.5402421652421653, + "grad_norm": 0.5733301639556885, + "learning_rate": 0.00019123191769108977, + "loss": 1.2142, + "step": 3034 + }, + { + "epoch": 0.5404202279202279, + "grad_norm": 0.5400987863540649, + "learning_rate": 0.00019122618511503494, + "loss": 1.1309, + "step": 3035 + }, + { + "epoch": 0.5405982905982906, + "grad_norm": 0.6261051893234253, + "learning_rate": 0.00019122045075159257, + "loss": 1.2112, + "step": 3036 + }, + { + "epoch": 0.5407763532763533, + "grad_norm": 0.5483576059341431, + "learning_rate": 0.0001912147146008749, + "loss": 1.2705, + "step": 3037 + }, + { + "epoch": 0.540954415954416, + "grad_norm": 0.5442137122154236, + "learning_rate": 0.00019120897666299443, + "loss": 1.2512, + "step": 3038 + }, + { + "epoch": 0.5411324786324786, + "grad_norm": 0.5680811405181885, + "learning_rate": 0.00019120323693806355, + "loss": 1.392, + "step": 3039 + }, + { + "epoch": 0.5413105413105413, + "grad_norm": 0.5237287878990173, + "learning_rate": 0.00019119749542619466, + "loss": 1.1599, + "step": 3040 + }, + { + "epoch": 0.541488603988604, + "grad_norm": 0.48119300603866577, + "learning_rate": 0.00019119175212750032, + "loss": 1.0976, + "step": 3041 + }, + { + "epoch": 0.5416666666666666, + "grad_norm": 0.507033109664917, + "learning_rate": 0.00019118600704209302, + "loss": 1.0181, + "step": 3042 + }, + { + "epoch": 0.5418447293447294, + "grad_norm": 0.484672874212265, + "learning_rate": 0.00019118026017008531, + "loss": 1.1636, + "step": 3043 + }, + { + "epoch": 0.542022792022792, + "grad_norm": 0.4923502206802368, + "learning_rate": 0.00019117451151158985, + "loss": 1.0388, + "step": 3044 + }, + { + "epoch": 0.5422008547008547, + "grad_norm": 0.4882057309150696, + "learning_rate": 0.00019116876106671922, + "loss": 1.131, + "step": 3045 + }, + { + "epoch": 0.5423789173789174, + "grad_norm": 0.6068355441093445, + "learning_rate": 0.0001911630088355861, + "loss": 1.3218, + "step": 3046 + }, + { + "epoch": 0.54255698005698, + "grad_norm": 0.5012881755828857, + "learning_rate": 0.0001911572548183032, + "loss": 1.0514, + "step": 3047 + }, + { + "epoch": 0.5427350427350427, + "grad_norm": 0.49849793314933777, + "learning_rate": 0.00019115149901498328, + "loss": 1.0003, + "step": 3048 + }, + { + "epoch": 0.5429131054131054, + "grad_norm": 0.4934251010417938, + "learning_rate": 0.00019114574142573904, + "loss": 1.0319, + "step": 3049 + }, + { + "epoch": 0.5430911680911681, + "grad_norm": 0.4947762191295624, + "learning_rate": 0.00019113998205068334, + "loss": 1.0906, + "step": 3050 + }, + { + "epoch": 0.5432692307692307, + "grad_norm": 0.5449416041374207, + "learning_rate": 0.00019113422088992907, + "loss": 0.9093, + "step": 3051 + }, + { + "epoch": 0.5434472934472935, + "grad_norm": 0.49395284056663513, + "learning_rate": 0.00019112845794358902, + "loss": 1.0071, + "step": 3052 + }, + { + "epoch": 0.5436253561253561, + "grad_norm": 0.5478728413581848, + "learning_rate": 0.00019112269321177613, + "loss": 1.2124, + "step": 3053 + }, + { + "epoch": 0.5438034188034188, + "grad_norm": 0.6205173134803772, + "learning_rate": 0.0001911169266946034, + "loss": 1.021, + "step": 3054 + }, + { + "epoch": 0.5439814814814815, + "grad_norm": 0.4777783751487732, + "learning_rate": 0.00019111115839218372, + "loss": 0.9192, + "step": 3055 + }, + { + "epoch": 0.5441595441595442, + "grad_norm": 0.5541689991950989, + "learning_rate": 0.00019110538830463018, + "loss": 1.1248, + "step": 3056 + }, + { + "epoch": 0.5443376068376068, + "grad_norm": 0.4750942289829254, + "learning_rate": 0.0001910996164320558, + "loss": 1.3147, + "step": 3057 + }, + { + "epoch": 0.5445156695156695, + "grad_norm": 0.6283948421478271, + "learning_rate": 0.0001910938427745737, + "loss": 1.0919, + "step": 3058 + }, + { + "epoch": 0.5446937321937322, + "grad_norm": 0.552725076675415, + "learning_rate": 0.00019108806733229698, + "loss": 1.3807, + "step": 3059 + }, + { + "epoch": 0.5448717948717948, + "grad_norm": 0.4832848310470581, + "learning_rate": 0.0001910822901053388, + "loss": 1.0705, + "step": 3060 + }, + { + "epoch": 0.5450498575498576, + "grad_norm": 0.6468375325202942, + "learning_rate": 0.00019107651109381233, + "loss": 1.0766, + "step": 3061 + }, + { + "epoch": 0.5452279202279202, + "grad_norm": 0.5464920401573181, + "learning_rate": 0.00019107073029783083, + "loss": 1.0453, + "step": 3062 + }, + { + "epoch": 0.5454059829059829, + "grad_norm": 0.5321210026741028, + "learning_rate": 0.0001910649477175076, + "loss": 1.2326, + "step": 3063 + }, + { + "epoch": 0.5455840455840456, + "grad_norm": 0.5572962164878845, + "learning_rate": 0.00019105916335295582, + "loss": 1.0673, + "step": 3064 + }, + { + "epoch": 0.5457621082621082, + "grad_norm": 0.5239177942276001, + "learning_rate": 0.00019105337720428894, + "loss": 1.04, + "step": 3065 + }, + { + "epoch": 0.5459401709401709, + "grad_norm": 0.5633319616317749, + "learning_rate": 0.00019104758927162023, + "loss": 0.9606, + "step": 3066 + }, + { + "epoch": 0.5461182336182336, + "grad_norm": 0.5317914485931396, + "learning_rate": 0.0001910417995550632, + "loss": 1.0651, + "step": 3067 + }, + { + "epoch": 0.5462962962962963, + "grad_norm": 0.5126453638076782, + "learning_rate": 0.00019103600805473118, + "loss": 1.0316, + "step": 3068 + }, + { + "epoch": 0.5464743589743589, + "grad_norm": 0.5262107253074646, + "learning_rate": 0.00019103021477073773, + "loss": 1.0752, + "step": 3069 + }, + { + "epoch": 0.5466524216524217, + "grad_norm": 0.5384877324104309, + "learning_rate": 0.0001910244197031963, + "loss": 1.1731, + "step": 3070 + }, + { + "epoch": 0.5468304843304843, + "grad_norm": 0.5126553773880005, + "learning_rate": 0.00019101862285222048, + "loss": 1.2229, + "step": 3071 + }, + { + "epoch": 0.5470085470085471, + "grad_norm": 0.4841194450855255, + "learning_rate": 0.0001910128242179238, + "loss": 0.9955, + "step": 3072 + }, + { + "epoch": 0.5471866096866097, + "grad_norm": 0.526546061038971, + "learning_rate": 0.00019100702380041987, + "loss": 1.2436, + "step": 3073 + }, + { + "epoch": 0.5473646723646723, + "grad_norm": 0.5085833072662354, + "learning_rate": 0.0001910012215998224, + "loss": 1.011, + "step": 3074 + }, + { + "epoch": 0.5475427350427351, + "grad_norm": 0.5149994492530823, + "learning_rate": 0.000190995417616245, + "loss": 0.8632, + "step": 3075 + }, + { + "epoch": 0.5477207977207977, + "grad_norm": 0.48079630732536316, + "learning_rate": 0.00019098961184980145, + "loss": 1.1115, + "step": 3076 + }, + { + "epoch": 0.5478988603988604, + "grad_norm": 0.5769477486610413, + "learning_rate": 0.00019098380430060546, + "loss": 0.9544, + "step": 3077 + }, + { + "epoch": 0.5480769230769231, + "grad_norm": 0.5260093808174133, + "learning_rate": 0.0001909779949687708, + "loss": 1.2354, + "step": 3078 + }, + { + "epoch": 0.5482549857549858, + "grad_norm": 0.5518734455108643, + "learning_rate": 0.00019097218385441135, + "loss": 1.1944, + "step": 3079 + }, + { + "epoch": 0.5484330484330484, + "grad_norm": 0.5436808466911316, + "learning_rate": 0.00019096637095764095, + "loss": 1.0717, + "step": 3080 + }, + { + "epoch": 0.5486111111111112, + "grad_norm": 0.4749584197998047, + "learning_rate": 0.00019096055627857344, + "loss": 1.0417, + "step": 3081 + }, + { + "epoch": 0.5487891737891738, + "grad_norm": 0.5485591292381287, + "learning_rate": 0.0001909547398173228, + "loss": 1.2515, + "step": 3082 + }, + { + "epoch": 0.5489672364672364, + "grad_norm": 0.5751016736030579, + "learning_rate": 0.00019094892157400296, + "loss": 1.2112, + "step": 3083 + }, + { + "epoch": 0.5491452991452992, + "grad_norm": 0.5404475331306458, + "learning_rate": 0.00019094310154872795, + "loss": 0.4334, + "step": 3084 + }, + { + "epoch": 0.5493233618233618, + "grad_norm": 0.5198020935058594, + "learning_rate": 0.00019093727974161178, + "loss": 0.9759, + "step": 3085 + }, + { + "epoch": 0.5495014245014245, + "grad_norm": 0.4893439710140228, + "learning_rate": 0.0001909314561527685, + "loss": 1.1287, + "step": 3086 + }, + { + "epoch": 0.5496794871794872, + "grad_norm": 0.5675956606864929, + "learning_rate": 0.00019092563078231228, + "loss": 1.234, + "step": 3087 + }, + { + "epoch": 0.5498575498575499, + "grad_norm": 0.5539132356643677, + "learning_rate": 0.00019091980363035714, + "loss": 1.2378, + "step": 3088 + }, + { + "epoch": 0.5500356125356125, + "grad_norm": 0.5194353461265564, + "learning_rate": 0.00019091397469701735, + "loss": 1.1338, + "step": 3089 + }, + { + "epoch": 0.5502136752136753, + "grad_norm": 0.5143756866455078, + "learning_rate": 0.0001909081439824071, + "loss": 0.9118, + "step": 3090 + }, + { + "epoch": 0.5503917378917379, + "grad_norm": 0.5624327659606934, + "learning_rate": 0.0001909023114866406, + "loss": 1.035, + "step": 3091 + }, + { + "epoch": 0.5505698005698005, + "grad_norm": 0.5285067558288574, + "learning_rate": 0.0001908964772098321, + "loss": 1.0451, + "step": 3092 + }, + { + "epoch": 0.5507478632478633, + "grad_norm": 0.5730587244033813, + "learning_rate": 0.000190890641152096, + "loss": 1.0672, + "step": 3093 + }, + { + "epoch": 0.5509259259259259, + "grad_norm": 0.5822951197624207, + "learning_rate": 0.0001908848033135466, + "loss": 1.1791, + "step": 3094 + }, + { + "epoch": 0.5511039886039886, + "grad_norm": 0.596161961555481, + "learning_rate": 0.00019087896369429826, + "loss": 1.0954, + "step": 3095 + }, + { + "epoch": 0.5512820512820513, + "grad_norm": 0.5138190984725952, + "learning_rate": 0.00019087312229446542, + "loss": 0.896, + "step": 3096 + }, + { + "epoch": 0.551460113960114, + "grad_norm": 0.5061872601509094, + "learning_rate": 0.0001908672791141625, + "loss": 1.1017, + "step": 3097 + }, + { + "epoch": 0.5516381766381766, + "grad_norm": 0.5189547538757324, + "learning_rate": 0.00019086143415350404, + "loss": 1.2906, + "step": 3098 + }, + { + "epoch": 0.5518162393162394, + "grad_norm": 0.5640039443969727, + "learning_rate": 0.00019085558741260448, + "loss": 1.1001, + "step": 3099 + }, + { + "epoch": 0.551994301994302, + "grad_norm": 0.453867107629776, + "learning_rate": 0.00019084973889157844, + "loss": 0.9731, + "step": 3100 + }, + { + "epoch": 0.5521723646723646, + "grad_norm": 0.5431303977966309, + "learning_rate": 0.0001908438885905405, + "loss": 1.3511, + "step": 3101 + }, + { + "epoch": 0.5523504273504274, + "grad_norm": 0.47693368792533875, + "learning_rate": 0.00019083803650960527, + "loss": 1.0426, + "step": 3102 + }, + { + "epoch": 0.55252849002849, + "grad_norm": 0.4663422703742981, + "learning_rate": 0.00019083218264888743, + "loss": 1.05, + "step": 3103 + }, + { + "epoch": 0.5527065527065527, + "grad_norm": 0.561354398727417, + "learning_rate": 0.00019082632700850164, + "loss": 0.9608, + "step": 3104 + }, + { + "epoch": 0.5528846153846154, + "grad_norm": 0.4981916844844818, + "learning_rate": 0.00019082046958856266, + "loss": 1.1935, + "step": 3105 + }, + { + "epoch": 0.5530626780626781, + "grad_norm": 0.5301326513290405, + "learning_rate": 0.0001908146103891852, + "loss": 1.0646, + "step": 3106 + }, + { + "epoch": 0.5532407407407407, + "grad_norm": 0.5023610591888428, + "learning_rate": 0.00019080874941048416, + "loss": 1.127, + "step": 3107 + }, + { + "epoch": 0.5534188034188035, + "grad_norm": 0.5172514319419861, + "learning_rate": 0.00019080288665257426, + "loss": 1.0435, + "step": 3108 + }, + { + "epoch": 0.5535968660968661, + "grad_norm": 0.6340598464012146, + "learning_rate": 0.00019079702211557048, + "loss": 1.3528, + "step": 3109 + }, + { + "epoch": 0.5537749287749287, + "grad_norm": 0.46882256865501404, + "learning_rate": 0.0001907911557995876, + "loss": 1.1361, + "step": 3110 + }, + { + "epoch": 0.5539529914529915, + "grad_norm": 0.6401382088661194, + "learning_rate": 0.00019078528770474068, + "loss": 1.2415, + "step": 3111 + }, + { + "epoch": 0.5541310541310541, + "grad_norm": 0.5141328573226929, + "learning_rate": 0.00019077941783114463, + "loss": 1.0505, + "step": 3112 + }, + { + "epoch": 0.5543091168091168, + "grad_norm": 0.522318959236145, + "learning_rate": 0.00019077354617891444, + "loss": 1.0964, + "step": 3113 + }, + { + "epoch": 0.5544871794871795, + "grad_norm": 0.539551854133606, + "learning_rate": 0.00019076767274816517, + "loss": 1.0735, + "step": 3114 + }, + { + "epoch": 0.5546652421652422, + "grad_norm": 0.495320200920105, + "learning_rate": 0.00019076179753901195, + "loss": 0.9754, + "step": 3115 + }, + { + "epoch": 0.5548433048433048, + "grad_norm": 0.5499199628829956, + "learning_rate": 0.00019075592055156984, + "loss": 1.0043, + "step": 3116 + }, + { + "epoch": 0.5550213675213675, + "grad_norm": 0.5352509617805481, + "learning_rate": 0.00019075004178595396, + "loss": 1.1701, + "step": 3117 + }, + { + "epoch": 0.5551994301994302, + "grad_norm": 0.5392300486564636, + "learning_rate": 0.00019074416124227953, + "loss": 1.1612, + "step": 3118 + }, + { + "epoch": 0.5553774928774928, + "grad_norm": 0.5195050835609436, + "learning_rate": 0.0001907382789206618, + "loss": 1.0934, + "step": 3119 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 0.5276884436607361, + "learning_rate": 0.000190732394821216, + "loss": 0.9011, + "step": 3120 + }, + { + "epoch": 0.5557336182336182, + "grad_norm": 0.6115903258323669, + "learning_rate": 0.00019072650894405734, + "loss": 1.3065, + "step": 3121 + }, + { + "epoch": 0.5559116809116809, + "grad_norm": 0.5752483010292053, + "learning_rate": 0.00019072062128930127, + "loss": 1.0063, + "step": 3122 + }, + { + "epoch": 0.5560897435897436, + "grad_norm": 0.5508273243904114, + "learning_rate": 0.00019071473185706302, + "loss": 1.2598, + "step": 3123 + }, + { + "epoch": 0.5562678062678063, + "grad_norm": 0.49712198972702026, + "learning_rate": 0.00019070884064745808, + "loss": 0.924, + "step": 3124 + }, + { + "epoch": 0.5564458689458689, + "grad_norm": 0.572849452495575, + "learning_rate": 0.00019070294766060185, + "loss": 0.9683, + "step": 3125 + }, + { + "epoch": 0.5566239316239316, + "grad_norm": 0.4807920753955841, + "learning_rate": 0.00019069705289660976, + "loss": 1.0998, + "step": 3126 + }, + { + "epoch": 0.5568019943019943, + "grad_norm": 0.5543031096458435, + "learning_rate": 0.0001906911563555973, + "loss": 1.0878, + "step": 3127 + }, + { + "epoch": 0.5569800569800569, + "grad_norm": 0.5710418820381165, + "learning_rate": 0.00019068525803768007, + "loss": 1.0381, + "step": 3128 + }, + { + "epoch": 0.5571581196581197, + "grad_norm": 0.5169163346290588, + "learning_rate": 0.00019067935794297357, + "loss": 1.1149, + "step": 3129 + }, + { + "epoch": 0.5573361823361823, + "grad_norm": 0.6474376916885376, + "learning_rate": 0.00019067345607159345, + "loss": 0.9828, + "step": 3130 + }, + { + "epoch": 0.5575142450142451, + "grad_norm": 0.5029847621917725, + "learning_rate": 0.0001906675524236553, + "loss": 0.797, + "step": 3131 + }, + { + "epoch": 0.5576923076923077, + "grad_norm": 0.5681431293487549, + "learning_rate": 0.00019066164699927478, + "loss": 1.1565, + "step": 3132 + }, + { + "epoch": 0.5578703703703703, + "grad_norm": 0.5654549598693848, + "learning_rate": 0.00019065573979856764, + "loss": 1.2488, + "step": 3133 + }, + { + "epoch": 0.5580484330484331, + "grad_norm": 0.47653043270111084, + "learning_rate": 0.0001906498308216496, + "loss": 1.0428, + "step": 3134 + }, + { + "epoch": 0.5582264957264957, + "grad_norm": 0.5068467259407043, + "learning_rate": 0.00019064392006863643, + "loss": 0.9659, + "step": 3135 + }, + { + "epoch": 0.5584045584045584, + "grad_norm": 0.7076661586761475, + "learning_rate": 0.00019063800753964393, + "loss": 1.1289, + "step": 3136 + }, + { + "epoch": 0.5585826210826211, + "grad_norm": 0.551456868648529, + "learning_rate": 0.000190632093234788, + "loss": 1.1925, + "step": 3137 + }, + { + "epoch": 0.5587606837606838, + "grad_norm": 0.518276035785675, + "learning_rate": 0.00019062617715418442, + "loss": 0.8681, + "step": 3138 + }, + { + "epoch": 0.5589387464387464, + "grad_norm": 0.5272278785705566, + "learning_rate": 0.0001906202592979492, + "loss": 1.0865, + "step": 3139 + }, + { + "epoch": 0.5591168091168092, + "grad_norm": 0.5344942212104797, + "learning_rate": 0.00019061433966619822, + "loss": 1.1647, + "step": 3140 + }, + { + "epoch": 0.5592948717948718, + "grad_norm": 0.5833460092544556, + "learning_rate": 0.00019060841825904753, + "loss": 1.3403, + "step": 3141 + }, + { + "epoch": 0.5594729344729344, + "grad_norm": 0.5707054734230042, + "learning_rate": 0.00019060249507661306, + "loss": 1.1236, + "step": 3142 + }, + { + "epoch": 0.5596509971509972, + "grad_norm": 0.5446065664291382, + "learning_rate": 0.00019059657011901094, + "loss": 1.017, + "step": 3143 + }, + { + "epoch": 0.5598290598290598, + "grad_norm": 0.5285109281539917, + "learning_rate": 0.0001905906433863572, + "loss": 1.3186, + "step": 3144 + }, + { + "epoch": 0.5600071225071225, + "grad_norm": 0.5308659672737122, + "learning_rate": 0.00019058471487876802, + "loss": 0.8464, + "step": 3145 + }, + { + "epoch": 0.5601851851851852, + "grad_norm": 0.5218054056167603, + "learning_rate": 0.00019057878459635948, + "loss": 1.0219, + "step": 3146 + }, + { + "epoch": 0.5603632478632479, + "grad_norm": 0.45067787170410156, + "learning_rate": 0.00019057285253924785, + "loss": 1.0364, + "step": 3147 + }, + { + "epoch": 0.5605413105413105, + "grad_norm": 0.4856041669845581, + "learning_rate": 0.0001905669187075493, + "loss": 1.1928, + "step": 3148 + }, + { + "epoch": 0.5607193732193733, + "grad_norm": 0.506912112236023, + "learning_rate": 0.00019056098310138016, + "loss": 1.119, + "step": 3149 + }, + { + "epoch": 0.5608974358974359, + "grad_norm": 0.49049463868141174, + "learning_rate": 0.00019055504572085662, + "loss": 1.2165, + "step": 3150 + }, + { + "epoch": 0.5610754985754985, + "grad_norm": 0.5250293612480164, + "learning_rate": 0.0001905491065660951, + "loss": 1.1427, + "step": 3151 + }, + { + "epoch": 0.5612535612535613, + "grad_norm": 0.43438446521759033, + "learning_rate": 0.00019054316563721195, + "loss": 0.884, + "step": 3152 + }, + { + "epoch": 0.5614316239316239, + "grad_norm": 0.5386807918548584, + "learning_rate": 0.00019053722293432354, + "loss": 1.1494, + "step": 3153 + }, + { + "epoch": 0.5616096866096866, + "grad_norm": 0.5403809547424316, + "learning_rate": 0.00019053127845754632, + "loss": 1.1743, + "step": 3154 + }, + { + "epoch": 0.5617877492877493, + "grad_norm": 0.4759823977947235, + "learning_rate": 0.00019052533220699678, + "loss": 1.0716, + "step": 3155 + }, + { + "epoch": 0.561965811965812, + "grad_norm": 0.45332327485084534, + "learning_rate": 0.0001905193841827914, + "loss": 0.8405, + "step": 3156 + }, + { + "epoch": 0.5621438746438746, + "grad_norm": 0.5617053508758545, + "learning_rate": 0.00019051343438504671, + "loss": 1.0422, + "step": 3157 + }, + { + "epoch": 0.5623219373219374, + "grad_norm": 0.5088049173355103, + "learning_rate": 0.00019050748281387931, + "loss": 1.0067, + "step": 3158 + }, + { + "epoch": 0.5625, + "grad_norm": 0.5174484848976135, + "learning_rate": 0.00019050152946940578, + "loss": 1.0623, + "step": 3159 + }, + { + "epoch": 0.5626780626780626, + "grad_norm": 0.6093568801879883, + "learning_rate": 0.0001904955743517428, + "loss": 1.24, + "step": 3160 + }, + { + "epoch": 0.5628561253561254, + "grad_norm": 0.49063584208488464, + "learning_rate": 0.00019048961746100703, + "loss": 0.8563, + "step": 3161 + }, + { + "epoch": 0.563034188034188, + "grad_norm": 0.583940863609314, + "learning_rate": 0.00019048365879731517, + "loss": 1.0695, + "step": 3162 + }, + { + "epoch": 0.5632122507122507, + "grad_norm": 0.4943268597126007, + "learning_rate": 0.000190477698360784, + "loss": 0.8606, + "step": 3163 + }, + { + "epoch": 0.5633903133903134, + "grad_norm": 0.5050932168960571, + "learning_rate": 0.00019047173615153028, + "loss": 1.1591, + "step": 3164 + }, + { + "epoch": 0.5635683760683761, + "grad_norm": 0.5445677638053894, + "learning_rate": 0.0001904657721696708, + "loss": 1.262, + "step": 3165 + }, + { + "epoch": 0.5637464387464387, + "grad_norm": 0.5445297360420227, + "learning_rate": 0.00019045980641532246, + "loss": 1.223, + "step": 3166 + }, + { + "epoch": 0.5639245014245015, + "grad_norm": 0.5098413228988647, + "learning_rate": 0.00019045383888860213, + "loss": 1.0829, + "step": 3167 + }, + { + "epoch": 0.5641025641025641, + "grad_norm": 0.484998881816864, + "learning_rate": 0.0001904478695896267, + "loss": 1.0711, + "step": 3168 + }, + { + "epoch": 0.5642806267806267, + "grad_norm": 0.5515334010124207, + "learning_rate": 0.0001904418985185132, + "loss": 1.1583, + "step": 3169 + }, + { + "epoch": 0.5644586894586895, + "grad_norm": 0.545460045337677, + "learning_rate": 0.00019043592567537853, + "loss": 1.2321, + "step": 3170 + }, + { + "epoch": 0.5646367521367521, + "grad_norm": 0.5463964343070984, + "learning_rate": 0.0001904299510603398, + "loss": 1.1019, + "step": 3171 + }, + { + "epoch": 0.5648148148148148, + "grad_norm": 0.5619220733642578, + "learning_rate": 0.000190423974673514, + "loss": 1.1001, + "step": 3172 + }, + { + "epoch": 0.5649928774928775, + "grad_norm": 0.4448916018009186, + "learning_rate": 0.00019041799651501825, + "loss": 1.057, + "step": 3173 + }, + { + "epoch": 0.5651709401709402, + "grad_norm": 0.6073006987571716, + "learning_rate": 0.00019041201658496975, + "loss": 1.0306, + "step": 3174 + }, + { + "epoch": 0.5653490028490028, + "grad_norm": 0.5342072248458862, + "learning_rate": 0.0001904060348834855, + "loss": 0.9231, + "step": 3175 + }, + { + "epoch": 0.5655270655270656, + "grad_norm": 0.4505697786808014, + "learning_rate": 0.0001904000514106829, + "loss": 1.1134, + "step": 3176 + }, + { + "epoch": 0.5657051282051282, + "grad_norm": 0.5627852082252502, + "learning_rate": 0.00019039406616667902, + "loss": 1.2138, + "step": 3177 + }, + { + "epoch": 0.5658831908831908, + "grad_norm": 0.499734103679657, + "learning_rate": 0.0001903880791515912, + "loss": 1.1074, + "step": 3178 + }, + { + "epoch": 0.5660612535612536, + "grad_norm": 0.4768189489841461, + "learning_rate": 0.00019038209036553676, + "loss": 0.9442, + "step": 3179 + }, + { + "epoch": 0.5662393162393162, + "grad_norm": 0.5265373587608337, + "learning_rate": 0.00019037609980863298, + "loss": 1.0907, + "step": 3180 + }, + { + "epoch": 0.5664173789173789, + "grad_norm": 0.5506128072738647, + "learning_rate": 0.00019037010748099728, + "loss": 1.2541, + "step": 3181 + }, + { + "epoch": 0.5665954415954416, + "grad_norm": 0.44860872626304626, + "learning_rate": 0.00019036411338274703, + "loss": 0.893, + "step": 3182 + }, + { + "epoch": 0.5667735042735043, + "grad_norm": 0.4901522994041443, + "learning_rate": 0.00019035811751399973, + "loss": 1.0469, + "step": 3183 + }, + { + "epoch": 0.5669515669515669, + "grad_norm": 0.500868022441864, + "learning_rate": 0.0001903521198748728, + "loss": 1.0527, + "step": 3184 + }, + { + "epoch": 0.5671296296296297, + "grad_norm": 0.5508102774620056, + "learning_rate": 0.00019034612046548376, + "loss": 1.283, + "step": 3185 + }, + { + "epoch": 0.5673076923076923, + "grad_norm": 0.5079495906829834, + "learning_rate": 0.0001903401192859502, + "loss": 1.0808, + "step": 3186 + }, + { + "epoch": 0.5674857549857549, + "grad_norm": 0.5758788585662842, + "learning_rate": 0.00019033411633638964, + "loss": 1.1301, + "step": 3187 + }, + { + "epoch": 0.5676638176638177, + "grad_norm": 0.46557924151420593, + "learning_rate": 0.00019032811161691972, + "loss": 1.0205, + "step": 3188 + }, + { + "epoch": 0.5678418803418803, + "grad_norm": 0.5665056109428406, + "learning_rate": 0.0001903221051276581, + "loss": 1.1926, + "step": 3189 + }, + { + "epoch": 0.5680199430199431, + "grad_norm": 0.5948992967605591, + "learning_rate": 0.00019031609686872246, + "loss": 1.2724, + "step": 3190 + }, + { + "epoch": 0.5681980056980057, + "grad_norm": 0.6189367771148682, + "learning_rate": 0.00019031008684023055, + "loss": 1.2762, + "step": 3191 + }, + { + "epoch": 0.5683760683760684, + "grad_norm": 0.49511992931365967, + "learning_rate": 0.00019030407504230006, + "loss": 1.0117, + "step": 3192 + }, + { + "epoch": 0.5685541310541311, + "grad_norm": 0.5358837842941284, + "learning_rate": 0.00019029806147504878, + "loss": 0.944, + "step": 3193 + }, + { + "epoch": 0.5687321937321937, + "grad_norm": 0.458636999130249, + "learning_rate": 0.00019029204613859463, + "loss": 0.8174, + "step": 3194 + }, + { + "epoch": 0.5689102564102564, + "grad_norm": 0.5168304443359375, + "learning_rate": 0.00019028602903305535, + "loss": 1.1533, + "step": 3195 + }, + { + "epoch": 0.5690883190883191, + "grad_norm": 0.5334134697914124, + "learning_rate": 0.00019028001015854892, + "loss": 1.1868, + "step": 3196 + }, + { + "epoch": 0.5692663817663818, + "grad_norm": 0.5649123191833496, + "learning_rate": 0.0001902739895151932, + "loss": 0.9876, + "step": 3197 + }, + { + "epoch": 0.5694444444444444, + "grad_norm": 0.5647651553153992, + "learning_rate": 0.0001902679671031062, + "loss": 1.0805, + "step": 3198 + }, + { + "epoch": 0.5696225071225072, + "grad_norm": 0.5251876711845398, + "learning_rate": 0.00019026194292240587, + "loss": 1.2335, + "step": 3199 + }, + { + "epoch": 0.5698005698005698, + "grad_norm": 0.5268014669418335, + "learning_rate": 0.0001902559169732103, + "loss": 1.19, + "step": 3200 + }, + { + "epoch": 0.5699786324786325, + "grad_norm": 0.5301041007041931, + "learning_rate": 0.00019024988925563752, + "loss": 1.1173, + "step": 3201 + }, + { + "epoch": 0.5701566951566952, + "grad_norm": 0.4531562030315399, + "learning_rate": 0.00019024385976980566, + "loss": 0.7576, + "step": 3202 + }, + { + "epoch": 0.5703347578347578, + "grad_norm": 0.5779716372489929, + "learning_rate": 0.00019023782851583282, + "loss": 1.1719, + "step": 3203 + }, + { + "epoch": 0.5705128205128205, + "grad_norm": 0.4886093735694885, + "learning_rate": 0.00019023179549383716, + "loss": 1.085, + "step": 3204 + }, + { + "epoch": 0.5706908831908832, + "grad_norm": 0.510117769241333, + "learning_rate": 0.0001902257607039369, + "loss": 0.8931, + "step": 3205 + }, + { + "epoch": 0.5708689458689459, + "grad_norm": 0.5195479393005371, + "learning_rate": 0.00019021972414625036, + "loss": 0.9922, + "step": 3206 + }, + { + "epoch": 0.5710470085470085, + "grad_norm": 0.5791407227516174, + "learning_rate": 0.00019021368582089568, + "loss": 1.112, + "step": 3207 + }, + { + "epoch": 0.5712250712250713, + "grad_norm": 0.5056005716323853, + "learning_rate": 0.00019020764572799122, + "loss": 0.8474, + "step": 3208 + }, + { + "epoch": 0.5714031339031339, + "grad_norm": 0.5060068964958191, + "learning_rate": 0.00019020160386765537, + "loss": 1.071, + "step": 3209 + }, + { + "epoch": 0.5715811965811965, + "grad_norm": 0.5396568775177002, + "learning_rate": 0.00019019556024000648, + "loss": 1.0436, + "step": 3210 + }, + { + "epoch": 0.5717592592592593, + "grad_norm": 0.6552190780639648, + "learning_rate": 0.0001901895148451629, + "loss": 0.9869, + "step": 3211 + }, + { + "epoch": 0.5719373219373219, + "grad_norm": 0.5177004337310791, + "learning_rate": 0.00019018346768324314, + "loss": 1.0193, + "step": 3212 + }, + { + "epoch": 0.5721153846153846, + "grad_norm": 0.5192117094993591, + "learning_rate": 0.0001901774187543657, + "loss": 1.1263, + "step": 3213 + }, + { + "epoch": 0.5722934472934473, + "grad_norm": 0.4857729971408844, + "learning_rate": 0.00019017136805864906, + "loss": 0.9808, + "step": 3214 + }, + { + "epoch": 0.57247150997151, + "grad_norm": 0.5800918936729431, + "learning_rate": 0.00019016531559621177, + "loss": 1.2334, + "step": 3215 + }, + { + "epoch": 0.5726495726495726, + "grad_norm": 0.4812086522579193, + "learning_rate": 0.00019015926136717242, + "loss": 1.2409, + "step": 3216 + }, + { + "epoch": 0.5728276353276354, + "grad_norm": 0.5128398537635803, + "learning_rate": 0.00019015320537164963, + "loss": 0.9036, + "step": 3217 + }, + { + "epoch": 0.573005698005698, + "grad_norm": 0.4761141538619995, + "learning_rate": 0.00019014714760976205, + "loss": 1.1058, + "step": 3218 + }, + { + "epoch": 0.5731837606837606, + "grad_norm": 0.5850459933280945, + "learning_rate": 0.0001901410880816284, + "loss": 1.1011, + "step": 3219 + }, + { + "epoch": 0.5733618233618234, + "grad_norm": 0.5648714303970337, + "learning_rate": 0.00019013502678736738, + "loss": 1.0479, + "step": 3220 + }, + { + "epoch": 0.573539886039886, + "grad_norm": 0.5835902094841003, + "learning_rate": 0.00019012896372709774, + "loss": 1.0555, + "step": 3221 + }, + { + "epoch": 0.5737179487179487, + "grad_norm": 0.5155113935470581, + "learning_rate": 0.00019012289890093828, + "loss": 0.9488, + "step": 3222 + }, + { + "epoch": 0.5738960113960114, + "grad_norm": 0.5064889788627625, + "learning_rate": 0.00019011683230900784, + "loss": 0.9144, + "step": 3223 + }, + { + "epoch": 0.5740740740740741, + "grad_norm": 0.53825843334198, + "learning_rate": 0.00019011076395142527, + "loss": 1.0713, + "step": 3224 + }, + { + "epoch": 0.5742521367521367, + "grad_norm": 0.5341386198997498, + "learning_rate": 0.00019010469382830947, + "loss": 1.1438, + "step": 3225 + }, + { + "epoch": 0.5744301994301995, + "grad_norm": 0.5300050973892212, + "learning_rate": 0.00019009862193977936, + "loss": 1.0114, + "step": 3226 + }, + { + "epoch": 0.5746082621082621, + "grad_norm": 0.6033682823181152, + "learning_rate": 0.0001900925482859539, + "loss": 1.0458, + "step": 3227 + }, + { + "epoch": 0.5747863247863247, + "grad_norm": 0.5108983516693115, + "learning_rate": 0.00019008647286695215, + "loss": 1.1211, + "step": 3228 + }, + { + "epoch": 0.5749643874643875, + "grad_norm": 0.5263782739639282, + "learning_rate": 0.00019008039568289308, + "loss": 0.8647, + "step": 3229 + }, + { + "epoch": 0.5751424501424501, + "grad_norm": 0.47119566798210144, + "learning_rate": 0.0001900743167338958, + "loss": 1.019, + "step": 3230 + }, + { + "epoch": 0.5753205128205128, + "grad_norm": 0.56391841173172, + "learning_rate": 0.00019006823602007937, + "loss": 0.9791, + "step": 3231 + }, + { + "epoch": 0.5754985754985755, + "grad_norm": 0.5364985466003418, + "learning_rate": 0.000190062153541563, + "loss": 1.1355, + "step": 3232 + }, + { + "epoch": 0.5756766381766382, + "grad_norm": 0.5098565220832825, + "learning_rate": 0.00019005606929846578, + "loss": 0.987, + "step": 3233 + }, + { + "epoch": 0.5758547008547008, + "grad_norm": 0.6640968918800354, + "learning_rate": 0.00019004998329090692, + "loss": 1.1165, + "step": 3234 + }, + { + "epoch": 0.5760327635327636, + "grad_norm": 0.5044721961021423, + "learning_rate": 0.00019004389551900578, + "loss": 0.8643, + "step": 3235 + }, + { + "epoch": 0.5762108262108262, + "grad_norm": 0.4822785258293152, + "learning_rate": 0.00019003780598288153, + "loss": 1.0735, + "step": 3236 + }, + { + "epoch": 0.5763888888888888, + "grad_norm": 0.505261242389679, + "learning_rate": 0.00019003171468265348, + "loss": 1.0001, + "step": 3237 + }, + { + "epoch": 0.5765669515669516, + "grad_norm": 0.5020412802696228, + "learning_rate": 0.00019002562161844102, + "loss": 0.9601, + "step": 3238 + }, + { + "epoch": 0.5767450142450142, + "grad_norm": 0.4920475482940674, + "learning_rate": 0.00019001952679036354, + "loss": 1.0111, + "step": 3239 + }, + { + "epoch": 0.5769230769230769, + "grad_norm": 0.5638813376426697, + "learning_rate": 0.00019001343019854042, + "loss": 1.1456, + "step": 3240 + }, + { + "epoch": 0.5771011396011396, + "grad_norm": 0.5519235134124756, + "learning_rate": 0.0001900073318430911, + "loss": 0.9258, + "step": 3241 + }, + { + "epoch": 0.5772792022792023, + "grad_norm": 0.5207770466804504, + "learning_rate": 0.0001900012317241351, + "loss": 0.9859, + "step": 3242 + }, + { + "epoch": 0.5774572649572649, + "grad_norm": 0.5493707656860352, + "learning_rate": 0.00018999512984179195, + "loss": 1.1183, + "step": 3243 + }, + { + "epoch": 0.5776353276353277, + "grad_norm": 0.4504764676094055, + "learning_rate": 0.00018998902619618116, + "loss": 0.9363, + "step": 3244 + }, + { + "epoch": 0.5778133903133903, + "grad_norm": 0.5232836604118347, + "learning_rate": 0.00018998292078742233, + "loss": 1.1887, + "step": 3245 + }, + { + "epoch": 0.5779914529914529, + "grad_norm": 0.5715088248252869, + "learning_rate": 0.0001899768136156351, + "loss": 1.4524, + "step": 3246 + }, + { + "epoch": 0.5781695156695157, + "grad_norm": 0.59555584192276, + "learning_rate": 0.0001899707046809391, + "loss": 1.0922, + "step": 3247 + }, + { + "epoch": 0.5783475783475783, + "grad_norm": 0.4500894546508789, + "learning_rate": 0.00018996459398345404, + "loss": 1.0087, + "step": 3248 + }, + { + "epoch": 0.5785256410256411, + "grad_norm": 0.49126625061035156, + "learning_rate": 0.00018995848152329967, + "loss": 1.1512, + "step": 3249 + }, + { + "epoch": 0.5787037037037037, + "grad_norm": 0.4096335172653198, + "learning_rate": 0.00018995236730059574, + "loss": 0.7633, + "step": 3250 + }, + { + "epoch": 0.5788817663817664, + "grad_norm": 0.5364313721656799, + "learning_rate": 0.00018994625131546199, + "loss": 1.295, + "step": 3251 + }, + { + "epoch": 0.5790598290598291, + "grad_norm": 0.4897502660751343, + "learning_rate": 0.00018994013356801834, + "loss": 1.2197, + "step": 3252 + }, + { + "epoch": 0.5792378917378918, + "grad_norm": 0.5101368427276611, + "learning_rate": 0.00018993401405838456, + "loss": 1.1129, + "step": 3253 + }, + { + "epoch": 0.5794159544159544, + "grad_norm": 0.5426377654075623, + "learning_rate": 0.00018992789278668063, + "loss": 1.188, + "step": 3254 + }, + { + "epoch": 0.5795940170940171, + "grad_norm": 0.5066362023353577, + "learning_rate": 0.00018992176975302644, + "loss": 1.2802, + "step": 3255 + }, + { + "epoch": 0.5797720797720798, + "grad_norm": 0.5418947339057922, + "learning_rate": 0.00018991564495754196, + "loss": 1.1675, + "step": 3256 + }, + { + "epoch": 0.5799501424501424, + "grad_norm": 0.5139963626861572, + "learning_rate": 0.0001899095184003472, + "loss": 0.9717, + "step": 3257 + }, + { + "epoch": 0.5801282051282052, + "grad_norm": 0.5167285799980164, + "learning_rate": 0.00018990339008156219, + "loss": 1.1529, + "step": 3258 + }, + { + "epoch": 0.5803062678062678, + "grad_norm": 0.53471440076828, + "learning_rate": 0.00018989726000130704, + "loss": 1.0711, + "step": 3259 + }, + { + "epoch": 0.5804843304843305, + "grad_norm": 0.49875229597091675, + "learning_rate": 0.0001898911281597018, + "loss": 1.1095, + "step": 3260 + }, + { + "epoch": 0.5806623931623932, + "grad_norm": 0.4473155438899994, + "learning_rate": 0.00018988499455686663, + "loss": 0.836, + "step": 3261 + }, + { + "epoch": 0.5808404558404558, + "grad_norm": 0.6181996464729309, + "learning_rate": 0.00018987885919292174, + "loss": 1.2787, + "step": 3262 + }, + { + "epoch": 0.5810185185185185, + "grad_norm": 0.4996899664402008, + "learning_rate": 0.00018987272206798733, + "loss": 1.2132, + "step": 3263 + }, + { + "epoch": 0.5811965811965812, + "grad_norm": 0.49979713559150696, + "learning_rate": 0.00018986658318218358, + "loss": 0.8388, + "step": 3264 + }, + { + "epoch": 0.5813746438746439, + "grad_norm": 0.5288876295089722, + "learning_rate": 0.00018986044253563084, + "loss": 1.1871, + "step": 3265 + }, + { + "epoch": 0.5815527065527065, + "grad_norm": 0.534063458442688, + "learning_rate": 0.00018985430012844937, + "loss": 0.96, + "step": 3266 + }, + { + "epoch": 0.5817307692307693, + "grad_norm": 0.5081285834312439, + "learning_rate": 0.00018984815596075953, + "loss": 1.1577, + "step": 3267 + }, + { + "epoch": 0.5819088319088319, + "grad_norm": 0.5648202896118164, + "learning_rate": 0.00018984201003268176, + "loss": 1.2235, + "step": 3268 + }, + { + "epoch": 0.5820868945868946, + "grad_norm": 0.495061993598938, + "learning_rate": 0.00018983586234433642, + "loss": 1.056, + "step": 3269 + }, + { + "epoch": 0.5822649572649573, + "grad_norm": 0.47149857878685, + "learning_rate": 0.000189829712895844, + "loss": 1.0844, + "step": 3270 + }, + { + "epoch": 0.58244301994302, + "grad_norm": 0.6107062697410583, + "learning_rate": 0.00018982356168732492, + "loss": 0.9868, + "step": 3271 + }, + { + "epoch": 0.5826210826210826, + "grad_norm": 0.7355940341949463, + "learning_rate": 0.00018981740871889974, + "loss": 1.1448, + "step": 3272 + }, + { + "epoch": 0.5827991452991453, + "grad_norm": 0.5950441956520081, + "learning_rate": 0.00018981125399068907, + "loss": 0.9618, + "step": 3273 + }, + { + "epoch": 0.582977207977208, + "grad_norm": 0.47607290744781494, + "learning_rate": 0.0001898050975028134, + "loss": 0.957, + "step": 3274 + }, + { + "epoch": 0.5831552706552706, + "grad_norm": 0.541164755821228, + "learning_rate": 0.00018979893925539338, + "loss": 1.1426, + "step": 3275 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 0.5240640044212341, + "learning_rate": 0.00018979277924854974, + "loss": 1.1421, + "step": 3276 + }, + { + "epoch": 0.583511396011396, + "grad_norm": 0.48155727982521057, + "learning_rate": 0.00018978661748240307, + "loss": 1.0069, + "step": 3277 + }, + { + "epoch": 0.5836894586894587, + "grad_norm": 0.5559938549995422, + "learning_rate": 0.00018978045395707418, + "loss": 1.1227, + "step": 3278 + }, + { + "epoch": 0.5838675213675214, + "grad_norm": 0.5244291424751282, + "learning_rate": 0.0001897742886726838, + "loss": 1.1103, + "step": 3279 + }, + { + "epoch": 0.584045584045584, + "grad_norm": 0.5277758240699768, + "learning_rate": 0.00018976812162935268, + "loss": 1.2125, + "step": 3280 + }, + { + "epoch": 0.5842236467236467, + "grad_norm": 0.5415039658546448, + "learning_rate": 0.00018976195282720173, + "loss": 1.146, + "step": 3281 + }, + { + "epoch": 0.5844017094017094, + "grad_norm": 0.5152051448822021, + "learning_rate": 0.00018975578226635177, + "loss": 1.0092, + "step": 3282 + }, + { + "epoch": 0.5845797720797721, + "grad_norm": 0.5489452481269836, + "learning_rate": 0.00018974960994692371, + "loss": 1.2425, + "step": 3283 + }, + { + "epoch": 0.5847578347578347, + "grad_norm": 0.491274356842041, + "learning_rate": 0.00018974343586903848, + "loss": 0.9559, + "step": 3284 + }, + { + "epoch": 0.5849358974358975, + "grad_norm": 0.5783739686012268, + "learning_rate": 0.00018973726003281707, + "loss": 1.1971, + "step": 3285 + }, + { + "epoch": 0.5851139601139601, + "grad_norm": 0.5056472420692444, + "learning_rate": 0.00018973108243838045, + "loss": 1.0313, + "step": 3286 + }, + { + "epoch": 0.5852920227920227, + "grad_norm": 0.4939729571342468, + "learning_rate": 0.00018972490308584962, + "loss": 1.1061, + "step": 3287 + }, + { + "epoch": 0.5854700854700855, + "grad_norm": 0.4889580011367798, + "learning_rate": 0.00018971872197534576, + "loss": 0.9157, + "step": 3288 + }, + { + "epoch": 0.5856481481481481, + "grad_norm": 0.40889349579811096, + "learning_rate": 0.00018971253910698993, + "loss": 0.8083, + "step": 3289 + }, + { + "epoch": 0.5858262108262108, + "grad_norm": 0.5221503973007202, + "learning_rate": 0.00018970635448090322, + "loss": 0.9995, + "step": 3290 + }, + { + "epoch": 0.5860042735042735, + "grad_norm": 0.47060561180114746, + "learning_rate": 0.00018970016809720687, + "loss": 0.9738, + "step": 3291 + }, + { + "epoch": 0.5861823361823362, + "grad_norm": 0.6083170771598816, + "learning_rate": 0.000189693979956022, + "loss": 1.188, + "step": 3292 + }, + { + "epoch": 0.5863603988603988, + "grad_norm": 0.4696751534938812, + "learning_rate": 0.00018968779005746998, + "loss": 1.089, + "step": 3293 + }, + { + "epoch": 0.5865384615384616, + "grad_norm": 0.5081014633178711, + "learning_rate": 0.00018968159840167202, + "loss": 1.1869, + "step": 3294 + }, + { + "epoch": 0.5867165242165242, + "grad_norm": 0.48042431473731995, + "learning_rate": 0.0001896754049887494, + "loss": 0.964, + "step": 3295 + }, + { + "epoch": 0.5868945868945868, + "grad_norm": 0.5075193643569946, + "learning_rate": 0.00018966920981882353, + "loss": 1.1884, + "step": 3296 + }, + { + "epoch": 0.5870726495726496, + "grad_norm": 0.5734842419624329, + "learning_rate": 0.00018966301289201576, + "loss": 1.1475, + "step": 3297 + }, + { + "epoch": 0.5872507122507122, + "grad_norm": 0.5525311231613159, + "learning_rate": 0.00018965681420844753, + "loss": 1.241, + "step": 3298 + }, + { + "epoch": 0.5874287749287749, + "grad_norm": 0.48142680525779724, + "learning_rate": 0.00018965061376824025, + "loss": 1.0871, + "step": 3299 + }, + { + "epoch": 0.5876068376068376, + "grad_norm": 0.5360350608825684, + "learning_rate": 0.00018964441157151544, + "loss": 1.1895, + "step": 3300 + }, + { + "epoch": 0.5877849002849003, + "grad_norm": 0.5207685232162476, + "learning_rate": 0.00018963820761839457, + "loss": 0.9323, + "step": 3301 + }, + { + "epoch": 0.5879629629629629, + "grad_norm": 0.453620970249176, + "learning_rate": 0.00018963200190899926, + "loss": 0.802, + "step": 3302 + }, + { + "epoch": 0.5881410256410257, + "grad_norm": 0.5198796391487122, + "learning_rate": 0.00018962579444345106, + "loss": 1.0243, + "step": 3303 + }, + { + "epoch": 0.5883190883190883, + "grad_norm": 0.5597525835037231, + "learning_rate": 0.0001896195852218716, + "loss": 0.9351, + "step": 3304 + }, + { + "epoch": 0.5884971509971509, + "grad_norm": 0.5738299489021301, + "learning_rate": 0.00018961337424438254, + "loss": 1.3737, + "step": 3305 + }, + { + "epoch": 0.5886752136752137, + "grad_norm": 0.5569949150085449, + "learning_rate": 0.00018960716151110554, + "loss": 1.0469, + "step": 3306 + }, + { + "epoch": 0.5888532763532763, + "grad_norm": 0.5088010430335999, + "learning_rate": 0.00018960094702216238, + "loss": 1.0982, + "step": 3307 + }, + { + "epoch": 0.5890313390313391, + "grad_norm": 0.5127636790275574, + "learning_rate": 0.0001895947307776748, + "loss": 0.9986, + "step": 3308 + }, + { + "epoch": 0.5892094017094017, + "grad_norm": 0.5160682797431946, + "learning_rate": 0.00018958851277776456, + "loss": 1.0219, + "step": 3309 + }, + { + "epoch": 0.5893874643874644, + "grad_norm": 0.5380711555480957, + "learning_rate": 0.00018958229302255356, + "loss": 1.118, + "step": 3310 + }, + { + "epoch": 0.5895655270655271, + "grad_norm": 0.5571228861808777, + "learning_rate": 0.0001895760715121636, + "loss": 1.0302, + "step": 3311 + }, + { + "epoch": 0.5897435897435898, + "grad_norm": 0.542266309261322, + "learning_rate": 0.00018956984824671657, + "loss": 1.0372, + "step": 3312 + }, + { + "epoch": 0.5899216524216524, + "grad_norm": 0.48350459337234497, + "learning_rate": 0.00018956362322633446, + "loss": 1.2, + "step": 3313 + }, + { + "epoch": 0.5900997150997151, + "grad_norm": 0.5001645088195801, + "learning_rate": 0.0001895573964511392, + "loss": 0.9749, + "step": 3314 + }, + { + "epoch": 0.5902777777777778, + "grad_norm": 0.5227531790733337, + "learning_rate": 0.00018955116792125276, + "loss": 1.025, + "step": 3315 + }, + { + "epoch": 0.5904558404558404, + "grad_norm": 0.522251546382904, + "learning_rate": 0.00018954493763679727, + "loss": 1.0821, + "step": 3316 + }, + { + "epoch": 0.5906339031339032, + "grad_norm": 0.5423251390457153, + "learning_rate": 0.00018953870559789467, + "loss": 1.0961, + "step": 3317 + }, + { + "epoch": 0.5908119658119658, + "grad_norm": 0.5615720748901367, + "learning_rate": 0.0001895324718046672, + "loss": 1.1209, + "step": 3318 + }, + { + "epoch": 0.5909900284900285, + "grad_norm": 0.44746771454811096, + "learning_rate": 0.00018952623625723692, + "loss": 0.9935, + "step": 3319 + }, + { + "epoch": 0.5911680911680912, + "grad_norm": 0.5993229150772095, + "learning_rate": 0.00018951999895572597, + "loss": 1.1409, + "step": 3320 + }, + { + "epoch": 0.5913461538461539, + "grad_norm": 0.4969801902770996, + "learning_rate": 0.00018951375990025666, + "loss": 1.1568, + "step": 3321 + }, + { + "epoch": 0.5915242165242165, + "grad_norm": 0.6001267433166504, + "learning_rate": 0.00018950751909095116, + "loss": 1.1135, + "step": 3322 + }, + { + "epoch": 0.5917022792022792, + "grad_norm": 0.5386021733283997, + "learning_rate": 0.00018950127652793172, + "loss": 0.947, + "step": 3323 + }, + { + "epoch": 0.5918803418803419, + "grad_norm": 0.49043843150138855, + "learning_rate": 0.00018949503221132074, + "loss": 0.9581, + "step": 3324 + }, + { + "epoch": 0.5920584045584045, + "grad_norm": 0.5241141319274902, + "learning_rate": 0.00018948878614124048, + "loss": 1.0797, + "step": 3325 + }, + { + "epoch": 0.5922364672364673, + "grad_norm": 0.5755026340484619, + "learning_rate": 0.00018948253831781338, + "loss": 1.1046, + "step": 3326 + }, + { + "epoch": 0.5924145299145299, + "grad_norm": 0.5004449486732483, + "learning_rate": 0.00018947628874116179, + "loss": 1.1416, + "step": 3327 + }, + { + "epoch": 0.5925925925925926, + "grad_norm": 0.53347247838974, + "learning_rate": 0.00018947003741140821, + "loss": 1.2718, + "step": 3328 + }, + { + "epoch": 0.5927706552706553, + "grad_norm": 0.6473469138145447, + "learning_rate": 0.0001894637843286751, + "loss": 1.2255, + "step": 3329 + }, + { + "epoch": 0.592948717948718, + "grad_norm": 0.4750518798828125, + "learning_rate": 0.00018945752949308498, + "loss": 1.0537, + "step": 3330 + }, + { + "epoch": 0.5931267806267806, + "grad_norm": 0.5636306405067444, + "learning_rate": 0.00018945127290476043, + "loss": 0.9906, + "step": 3331 + }, + { + "epoch": 0.5933048433048433, + "grad_norm": 0.4871736466884613, + "learning_rate": 0.00018944501456382397, + "loss": 1.0549, + "step": 3332 + }, + { + "epoch": 0.593482905982906, + "grad_norm": 0.5554637312889099, + "learning_rate": 0.0001894387544703983, + "loss": 1.1587, + "step": 3333 + }, + { + "epoch": 0.5936609686609686, + "grad_norm": 0.5385799407958984, + "learning_rate": 0.000189432492624606, + "loss": 0.9565, + "step": 3334 + }, + { + "epoch": 0.5938390313390314, + "grad_norm": 0.4996553063392639, + "learning_rate": 0.00018942622902656976, + "loss": 1.0456, + "step": 3335 + }, + { + "epoch": 0.594017094017094, + "grad_norm": 0.46810707449913025, + "learning_rate": 0.00018941996367641237, + "loss": 1.119, + "step": 3336 + }, + { + "epoch": 0.5941951566951567, + "grad_norm": 0.5672653913497925, + "learning_rate": 0.0001894136965742565, + "loss": 1.1317, + "step": 3337 + }, + { + "epoch": 0.5943732193732194, + "grad_norm": 0.4790053367614746, + "learning_rate": 0.00018940742772022504, + "loss": 1.0967, + "step": 3338 + }, + { + "epoch": 0.594551282051282, + "grad_norm": 0.5935906171798706, + "learning_rate": 0.00018940115711444072, + "loss": 1.3044, + "step": 3339 + }, + { + "epoch": 0.5947293447293447, + "grad_norm": 0.4790516793727875, + "learning_rate": 0.00018939488475702647, + "loss": 1.074, + "step": 3340 + }, + { + "epoch": 0.5949074074074074, + "grad_norm": 0.474588006734848, + "learning_rate": 0.00018938861064810516, + "loss": 1.1476, + "step": 3341 + }, + { + "epoch": 0.5950854700854701, + "grad_norm": 0.4908665120601654, + "learning_rate": 0.0001893823347877997, + "loss": 1.216, + "step": 3342 + }, + { + "epoch": 0.5952635327635327, + "grad_norm": 0.531650960445404, + "learning_rate": 0.00018937605717623307, + "loss": 1.1057, + "step": 3343 + }, + { + "epoch": 0.5954415954415955, + "grad_norm": 0.5581082105636597, + "learning_rate": 0.00018936977781352823, + "loss": 0.7972, + "step": 3344 + }, + { + "epoch": 0.5956196581196581, + "grad_norm": 0.42370662093162537, + "learning_rate": 0.00018936349669980827, + "loss": 0.8888, + "step": 3345 + }, + { + "epoch": 0.5957977207977208, + "grad_norm": 0.5817318558692932, + "learning_rate": 0.00018935721383519624, + "loss": 1.2801, + "step": 3346 + }, + { + "epoch": 0.5959757834757835, + "grad_norm": 0.4766376316547394, + "learning_rate": 0.00018935092921981524, + "loss": 1.0918, + "step": 3347 + }, + { + "epoch": 0.5961538461538461, + "grad_norm": 0.5567346811294556, + "learning_rate": 0.00018934464285378836, + "loss": 1.0269, + "step": 3348 + }, + { + "epoch": 0.5963319088319088, + "grad_norm": 0.5285565257072449, + "learning_rate": 0.0001893383547372388, + "loss": 1.1887, + "step": 3349 + }, + { + "epoch": 0.5965099715099715, + "grad_norm": 0.49052694439888, + "learning_rate": 0.00018933206487028979, + "loss": 1.0773, + "step": 3350 + }, + { + "epoch": 0.5966880341880342, + "grad_norm": 0.6175199151039124, + "learning_rate": 0.0001893257732530645, + "loss": 1.0192, + "step": 3351 + }, + { + "epoch": 0.5968660968660968, + "grad_norm": 0.56049644947052, + "learning_rate": 0.00018931947988568628, + "loss": 0.9516, + "step": 3352 + }, + { + "epoch": 0.5970441595441596, + "grad_norm": 0.47873660922050476, + "learning_rate": 0.00018931318476827838, + "loss": 0.8174, + "step": 3353 + }, + { + "epoch": 0.5972222222222222, + "grad_norm": 0.4748854339122772, + "learning_rate": 0.00018930688790096416, + "loss": 1.0238, + "step": 3354 + }, + { + "epoch": 0.5974002849002849, + "grad_norm": 0.5382232666015625, + "learning_rate": 0.00018930058928386698, + "loss": 1.0815, + "step": 3355 + }, + { + "epoch": 0.5975783475783476, + "grad_norm": 0.5038299560546875, + "learning_rate": 0.00018929428891711027, + "loss": 1.0472, + "step": 3356 + }, + { + "epoch": 0.5977564102564102, + "grad_norm": 0.5185908079147339, + "learning_rate": 0.00018928798680081744, + "loss": 1.0435, + "step": 3357 + }, + { + "epoch": 0.5979344729344729, + "grad_norm": 0.5169877409934998, + "learning_rate": 0.00018928168293511202, + "loss": 1.0437, + "step": 3358 + }, + { + "epoch": 0.5981125356125356, + "grad_norm": 0.5218369960784912, + "learning_rate": 0.00018927537732011749, + "loss": 1.082, + "step": 3359 + }, + { + "epoch": 0.5982905982905983, + "grad_norm": 0.5358219742774963, + "learning_rate": 0.0001892690699559574, + "loss": 1.2523, + "step": 3360 + }, + { + "epoch": 0.5984686609686609, + "grad_norm": 0.47716647386550903, + "learning_rate": 0.0001892627608427553, + "loss": 1.2069, + "step": 3361 + }, + { + "epoch": 0.5986467236467237, + "grad_norm": 0.5484169125556946, + "learning_rate": 0.00018925644998063482, + "loss": 1.2016, + "step": 3362 + }, + { + "epoch": 0.5988247863247863, + "grad_norm": 0.46814846992492676, + "learning_rate": 0.00018925013736971965, + "loss": 0.7989, + "step": 3363 + }, + { + "epoch": 0.5990028490028491, + "grad_norm": 0.5391258001327515, + "learning_rate": 0.0001892438230101334, + "loss": 1.224, + "step": 3364 + }, + { + "epoch": 0.5991809116809117, + "grad_norm": 0.5248384475708008, + "learning_rate": 0.00018923750690199987, + "loss": 1.1532, + "step": 3365 + }, + { + "epoch": 0.5993589743589743, + "grad_norm": 0.5074637532234192, + "learning_rate": 0.00018923118904544273, + "loss": 1.0968, + "step": 3366 + }, + { + "epoch": 0.5995370370370371, + "grad_norm": 0.5260029435157776, + "learning_rate": 0.00018922486944058581, + "loss": 1.1311, + "step": 3367 + }, + { + "epoch": 0.5997150997150997, + "grad_norm": 0.48497965931892395, + "learning_rate": 0.00018921854808755294, + "loss": 1.1208, + "step": 3368 + }, + { + "epoch": 0.5998931623931624, + "grad_norm": 0.5108651518821716, + "learning_rate": 0.00018921222498646792, + "loss": 1.147, + "step": 3369 + }, + { + "epoch": 0.6000712250712251, + "grad_norm": 0.5243437886238098, + "learning_rate": 0.00018920590013745471, + "loss": 0.9614, + "step": 3370 + }, + { + "epoch": 0.6002492877492878, + "grad_norm": 0.47022634744644165, + "learning_rate": 0.00018919957354063719, + "loss": 1.0579, + "step": 3371 + }, + { + "epoch": 0.6004273504273504, + "grad_norm": 0.6461413502693176, + "learning_rate": 0.00018919324519613931, + "loss": 1.2126, + "step": 3372 + }, + { + "epoch": 0.6006054131054132, + "grad_norm": 0.4654616713523865, + "learning_rate": 0.00018918691510408508, + "loss": 1.1476, + "step": 3373 + }, + { + "epoch": 0.6007834757834758, + "grad_norm": 0.48571303486824036, + "learning_rate": 0.00018918058326459854, + "loss": 1.2093, + "step": 3374 + }, + { + "epoch": 0.6009615384615384, + "grad_norm": 0.5255016684532166, + "learning_rate": 0.00018917424967780368, + "loss": 1.1538, + "step": 3375 + }, + { + "epoch": 0.6011396011396012, + "grad_norm": 0.5059894323348999, + "learning_rate": 0.00018916791434382468, + "loss": 1.0556, + "step": 3376 + }, + { + "epoch": 0.6013176638176638, + "grad_norm": 0.4581229090690613, + "learning_rate": 0.00018916157726278561, + "loss": 1.1468, + "step": 3377 + }, + { + "epoch": 0.6014957264957265, + "grad_norm": 0.5701818466186523, + "learning_rate": 0.00018915523843481067, + "loss": 1.3641, + "step": 3378 + }, + { + "epoch": 0.6016737891737892, + "grad_norm": 0.5007243752479553, + "learning_rate": 0.00018914889786002403, + "loss": 1.2705, + "step": 3379 + }, + { + "epoch": 0.6018518518518519, + "grad_norm": 0.5192995071411133, + "learning_rate": 0.0001891425555385499, + "loss": 0.9922, + "step": 3380 + }, + { + "epoch": 0.6020299145299145, + "grad_norm": 0.5880612134933472, + "learning_rate": 0.00018913621147051258, + "loss": 0.8783, + "step": 3381 + }, + { + "epoch": 0.6022079772079773, + "grad_norm": 0.5161563158035278, + "learning_rate": 0.0001891298656560364, + "loss": 0.9634, + "step": 3382 + }, + { + "epoch": 0.6023860398860399, + "grad_norm": 0.48450782895088196, + "learning_rate": 0.00018912351809524563, + "loss": 0.809, + "step": 3383 + }, + { + "epoch": 0.6025641025641025, + "grad_norm": 0.621537983417511, + "learning_rate": 0.00018911716878826465, + "loss": 1.2031, + "step": 3384 + }, + { + "epoch": 0.6027421652421653, + "grad_norm": 0.6014544367790222, + "learning_rate": 0.00018911081773521787, + "loss": 1.1552, + "step": 3385 + }, + { + "epoch": 0.6029202279202279, + "grad_norm": 0.49995481967926025, + "learning_rate": 0.00018910446493622976, + "loss": 0.8569, + "step": 3386 + }, + { + "epoch": 0.6030982905982906, + "grad_norm": 0.5157307386398315, + "learning_rate": 0.00018909811039142472, + "loss": 0.9515, + "step": 3387 + }, + { + "epoch": 0.6032763532763533, + "grad_norm": 0.5164140462875366, + "learning_rate": 0.0001890917541009273, + "loss": 0.9803, + "step": 3388 + }, + { + "epoch": 0.603454415954416, + "grad_norm": 0.5555596947669983, + "learning_rate": 0.00018908539606486206, + "loss": 1.2994, + "step": 3389 + }, + { + "epoch": 0.6036324786324786, + "grad_norm": 0.605697512626648, + "learning_rate": 0.00018907903628335353, + "loss": 1.2865, + "step": 3390 + }, + { + "epoch": 0.6038105413105413, + "grad_norm": 0.5700713992118835, + "learning_rate": 0.0001890726747565263, + "loss": 1.2493, + "step": 3391 + }, + { + "epoch": 0.603988603988604, + "grad_norm": 0.5516746044158936, + "learning_rate": 0.0001890663114845051, + "loss": 1.2743, + "step": 3392 + }, + { + "epoch": 0.6041666666666666, + "grad_norm": 0.5233162641525269, + "learning_rate": 0.0001890599464674145, + "loss": 0.9237, + "step": 3393 + }, + { + "epoch": 0.6043447293447294, + "grad_norm": 0.5709942579269409, + "learning_rate": 0.00018905357970537925, + "loss": 0.9922, + "step": 3394 + }, + { + "epoch": 0.604522792022792, + "grad_norm": 0.48403796553611755, + "learning_rate": 0.0001890472111985241, + "loss": 1.1255, + "step": 3395 + }, + { + "epoch": 0.6047008547008547, + "grad_norm": 0.628718376159668, + "learning_rate": 0.00018904084094697386, + "loss": 1.1458, + "step": 3396 + }, + { + "epoch": 0.6048789173789174, + "grad_norm": 0.46822869777679443, + "learning_rate": 0.00018903446895085328, + "loss": 0.8727, + "step": 3397 + }, + { + "epoch": 0.60505698005698, + "grad_norm": 0.505584180355072, + "learning_rate": 0.00018902809521028724, + "loss": 1.1595, + "step": 3398 + }, + { + "epoch": 0.6052350427350427, + "grad_norm": 0.4494974911212921, + "learning_rate": 0.00018902171972540058, + "loss": 0.6685, + "step": 3399 + }, + { + "epoch": 0.6054131054131054, + "grad_norm": 0.5101519227027893, + "learning_rate": 0.0001890153424963183, + "loss": 0.9313, + "step": 3400 + }, + { + "epoch": 0.6055911680911681, + "grad_norm": 0.5081079602241516, + "learning_rate": 0.00018900896352316528, + "loss": 1.2588, + "step": 3401 + }, + { + "epoch": 0.6057692307692307, + "grad_norm": 0.5784309506416321, + "learning_rate": 0.00018900258280606653, + "loss": 1.2077, + "step": 3402 + }, + { + "epoch": 0.6059472934472935, + "grad_norm": 0.4506312608718872, + "learning_rate": 0.00018899620034514705, + "loss": 1.05, + "step": 3403 + }, + { + "epoch": 0.6061253561253561, + "grad_norm": 0.5243048071861267, + "learning_rate": 0.0001889898161405319, + "loss": 1.2295, + "step": 3404 + }, + { + "epoch": 0.6063034188034188, + "grad_norm": 0.5447196364402771, + "learning_rate": 0.00018898343019234615, + "loss": 1.1476, + "step": 3405 + }, + { + "epoch": 0.6064814814814815, + "grad_norm": 0.46813663840293884, + "learning_rate": 0.00018897704250071492, + "loss": 1.2113, + "step": 3406 + }, + { + "epoch": 0.6066595441595442, + "grad_norm": 0.5340631604194641, + "learning_rate": 0.00018897065306576342, + "loss": 1.1656, + "step": 3407 + }, + { + "epoch": 0.6068376068376068, + "grad_norm": 0.513708233833313, + "learning_rate": 0.00018896426188761675, + "loss": 1.1616, + "step": 3408 + }, + { + "epoch": 0.6070156695156695, + "grad_norm": 0.594601035118103, + "learning_rate": 0.00018895786896640023, + "loss": 1.2564, + "step": 3409 + }, + { + "epoch": 0.6071937321937322, + "grad_norm": 0.45067599415779114, + "learning_rate": 0.000188951474302239, + "loss": 1.0107, + "step": 3410 + }, + { + "epoch": 0.6073717948717948, + "grad_norm": 0.5394250750541687, + "learning_rate": 0.00018894507789525843, + "loss": 1.4081, + "step": 3411 + }, + { + "epoch": 0.6075498575498576, + "grad_norm": 0.5612049102783203, + "learning_rate": 0.00018893867974558383, + "loss": 1.1015, + "step": 3412 + }, + { + "epoch": 0.6077279202279202, + "grad_norm": 0.4794061779975891, + "learning_rate": 0.00018893227985334056, + "loss": 1.2103, + "step": 3413 + }, + { + "epoch": 0.6079059829059829, + "grad_norm": 0.6060562133789062, + "learning_rate": 0.00018892587821865402, + "loss": 1.3693, + "step": 3414 + }, + { + "epoch": 0.6080840455840456, + "grad_norm": 0.44624534249305725, + "learning_rate": 0.00018891947484164963, + "loss": 0.8209, + "step": 3415 + }, + { + "epoch": 0.6082621082621082, + "grad_norm": 0.49297213554382324, + "learning_rate": 0.0001889130697224528, + "loss": 1.2027, + "step": 3416 + }, + { + "epoch": 0.6084401709401709, + "grad_norm": 0.4431746304035187, + "learning_rate": 0.0001889066628611891, + "loss": 1.0347, + "step": 3417 + }, + { + "epoch": 0.6086182336182336, + "grad_norm": 0.5425933599472046, + "learning_rate": 0.00018890025425798404, + "loss": 1.0556, + "step": 3418 + }, + { + "epoch": 0.6087962962962963, + "grad_norm": 0.5502763390541077, + "learning_rate": 0.00018889384391296315, + "loss": 1.2362, + "step": 3419 + }, + { + "epoch": 0.6089743589743589, + "grad_norm": 0.5442292094230652, + "learning_rate": 0.00018888743182625203, + "loss": 1.1306, + "step": 3420 + }, + { + "epoch": 0.6091524216524217, + "grad_norm": 0.4651123583316803, + "learning_rate": 0.00018888101799797636, + "loss": 0.9305, + "step": 3421 + }, + { + "epoch": 0.6093304843304843, + "grad_norm": 0.4713892340660095, + "learning_rate": 0.00018887460242826177, + "loss": 1.0789, + "step": 3422 + }, + { + "epoch": 0.6095085470085471, + "grad_norm": 0.5283244848251343, + "learning_rate": 0.00018886818511723398, + "loss": 1.345, + "step": 3423 + }, + { + "epoch": 0.6096866096866097, + "grad_norm": 0.5527324080467224, + "learning_rate": 0.0001888617660650187, + "loss": 1.1297, + "step": 3424 + }, + { + "epoch": 0.6098646723646723, + "grad_norm": 0.5412901043891907, + "learning_rate": 0.00018885534527174168, + "loss": 1.1213, + "step": 3425 + }, + { + "epoch": 0.6100427350427351, + "grad_norm": 0.5295354127883911, + "learning_rate": 0.00018884892273752878, + "loss": 1.1217, + "step": 3426 + }, + { + "epoch": 0.6102207977207977, + "grad_norm": 0.461900532245636, + "learning_rate": 0.0001888424984625058, + "loss": 0.827, + "step": 3427 + }, + { + "epoch": 0.6103988603988604, + "grad_norm": 0.4922671616077423, + "learning_rate": 0.00018883607244679865, + "loss": 1.2216, + "step": 3428 + }, + { + "epoch": 0.6105769230769231, + "grad_norm": 0.5080927014350891, + "learning_rate": 0.00018882964469053317, + "loss": 1.2446, + "step": 3429 + }, + { + "epoch": 0.6107549857549858, + "grad_norm": 0.5523943901062012, + "learning_rate": 0.00018882321519383534, + "loss": 1.3346, + "step": 3430 + }, + { + "epoch": 0.6109330484330484, + "grad_norm": 0.5105271935462952, + "learning_rate": 0.0001888167839568311, + "loss": 1.1311, + "step": 3431 + }, + { + "epoch": 0.6111111111111112, + "grad_norm": 0.5635872483253479, + "learning_rate": 0.0001888103509796465, + "loss": 1.1875, + "step": 3432 + }, + { + "epoch": 0.6112891737891738, + "grad_norm": 0.4619547426700592, + "learning_rate": 0.00018880391626240755, + "loss": 0.9176, + "step": 3433 + }, + { + "epoch": 0.6114672364672364, + "grad_norm": 0.5896356105804443, + "learning_rate": 0.00018879747980524034, + "loss": 1.0251, + "step": 3434 + }, + { + "epoch": 0.6116452991452992, + "grad_norm": 0.49062737822532654, + "learning_rate": 0.000188791041608271, + "loss": 1.1598, + "step": 3435 + }, + { + "epoch": 0.6118233618233618, + "grad_norm": 0.45717164874076843, + "learning_rate": 0.00018878460167162558, + "loss": 0.8647, + "step": 3436 + }, + { + "epoch": 0.6120014245014245, + "grad_norm": 0.5903525352478027, + "learning_rate": 0.00018877815999543038, + "loss": 0.9671, + "step": 3437 + }, + { + "epoch": 0.6121794871794872, + "grad_norm": 0.5315384268760681, + "learning_rate": 0.00018877171657981153, + "loss": 1.1759, + "step": 3438 + }, + { + "epoch": 0.6123575498575499, + "grad_norm": 0.5650150775909424, + "learning_rate": 0.0001887652714248953, + "loss": 1.0128, + "step": 3439 + }, + { + "epoch": 0.6125356125356125, + "grad_norm": 0.49841752648353577, + "learning_rate": 0.000188758824530808, + "loss": 1.1259, + "step": 3440 + }, + { + "epoch": 0.6127136752136753, + "grad_norm": 0.4985620975494385, + "learning_rate": 0.00018875237589767593, + "loss": 1.0158, + "step": 3441 + }, + { + "epoch": 0.6128917378917379, + "grad_norm": 0.45266565680503845, + "learning_rate": 0.00018874592552562536, + "loss": 0.93, + "step": 3442 + }, + { + "epoch": 0.6130698005698005, + "grad_norm": 0.5696130990982056, + "learning_rate": 0.00018873947341478274, + "loss": 1.1432, + "step": 3443 + }, + { + "epoch": 0.6132478632478633, + "grad_norm": 0.5211645364761353, + "learning_rate": 0.00018873301956527451, + "loss": 1.1317, + "step": 3444 + }, + { + "epoch": 0.6134259259259259, + "grad_norm": 0.4991866946220398, + "learning_rate": 0.00018872656397722707, + "loss": 1.0362, + "step": 3445 + }, + { + "epoch": 0.6136039886039886, + "grad_norm": 0.5109508037567139, + "learning_rate": 0.00018872010665076694, + "loss": 1.2728, + "step": 3446 + }, + { + "epoch": 0.6137820512820513, + "grad_norm": 0.5838373899459839, + "learning_rate": 0.00018871364758602058, + "loss": 1.1131, + "step": 3447 + }, + { + "epoch": 0.613960113960114, + "grad_norm": 0.5139824151992798, + "learning_rate": 0.00018870718678311462, + "loss": 1.238, + "step": 3448 + }, + { + "epoch": 0.6141381766381766, + "grad_norm": 0.4852082431316376, + "learning_rate": 0.00018870072424217562, + "loss": 1.0677, + "step": 3449 + }, + { + "epoch": 0.6143162393162394, + "grad_norm": 0.5312315225601196, + "learning_rate": 0.00018869425996333018, + "loss": 1.178, + "step": 3450 + }, + { + "epoch": 0.614494301994302, + "grad_norm": 0.6343565583229065, + "learning_rate": 0.00018868779394670492, + "loss": 0.8839, + "step": 3451 + }, + { + "epoch": 0.6146723646723646, + "grad_norm": 0.6029773950576782, + "learning_rate": 0.00018868132619242662, + "loss": 1.1188, + "step": 3452 + }, + { + "epoch": 0.6148504273504274, + "grad_norm": 0.5246016383171082, + "learning_rate": 0.00018867485670062193, + "loss": 1.0797, + "step": 3453 + }, + { + "epoch": 0.61502849002849, + "grad_norm": 0.49307698011398315, + "learning_rate": 0.00018866838547141763, + "loss": 0.9749, + "step": 3454 + }, + { + "epoch": 0.6152065527065527, + "grad_norm": 0.5232903361320496, + "learning_rate": 0.00018866191250494052, + "loss": 1.0785, + "step": 3455 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 0.5545645356178284, + "learning_rate": 0.0001886554378013174, + "loss": 1.0496, + "step": 3456 + }, + { + "epoch": 0.6155626780626781, + "grad_norm": 0.493945837020874, + "learning_rate": 0.00018864896136067515, + "loss": 0.9248, + "step": 3457 + }, + { + "epoch": 0.6157407407407407, + "grad_norm": 0.5223548412322998, + "learning_rate": 0.00018864248318314065, + "loss": 1.0617, + "step": 3458 + }, + { + "epoch": 0.6159188034188035, + "grad_norm": 0.5666514039039612, + "learning_rate": 0.00018863600326884082, + "loss": 0.9981, + "step": 3459 + }, + { + "epoch": 0.6160968660968661, + "grad_norm": 0.4648127257823944, + "learning_rate": 0.00018862952161790265, + "loss": 0.917, + "step": 3460 + }, + { + "epoch": 0.6162749287749287, + "grad_norm": 0.590326189994812, + "learning_rate": 0.0001886230382304531, + "loss": 1.044, + "step": 3461 + }, + { + "epoch": 0.6164529914529915, + "grad_norm": 0.5511625409126282, + "learning_rate": 0.00018861655310661925, + "loss": 1.0988, + "step": 3462 + }, + { + "epoch": 0.6166310541310541, + "grad_norm": 0.567182183265686, + "learning_rate": 0.0001886100662465281, + "loss": 1.3017, + "step": 3463 + }, + { + "epoch": 0.6168091168091168, + "grad_norm": 0.5708897709846497, + "learning_rate": 0.0001886035776503068, + "loss": 0.9123, + "step": 3464 + }, + { + "epoch": 0.6169871794871795, + "grad_norm": 0.4945180416107178, + "learning_rate": 0.0001885970873180824, + "loss": 1.1645, + "step": 3465 + }, + { + "epoch": 0.6171652421652422, + "grad_norm": 0.4713336229324341, + "learning_rate": 0.00018859059524998215, + "loss": 1.0546, + "step": 3466 + }, + { + "epoch": 0.6173433048433048, + "grad_norm": 0.532859206199646, + "learning_rate": 0.0001885841014461332, + "loss": 1.0795, + "step": 3467 + }, + { + "epoch": 0.6175213675213675, + "grad_norm": 0.5165733695030212, + "learning_rate": 0.00018857760590666284, + "loss": 1.1284, + "step": 3468 + }, + { + "epoch": 0.6176994301994302, + "grad_norm": 0.48623126745224, + "learning_rate": 0.00018857110863169826, + "loss": 0.8618, + "step": 3469 + }, + { + "epoch": 0.6178774928774928, + "grad_norm": 0.628559947013855, + "learning_rate": 0.0001885646096213668, + "loss": 1.1089, + "step": 3470 + }, + { + "epoch": 0.6180555555555556, + "grad_norm": 0.503545880317688, + "learning_rate": 0.0001885581088757958, + "loss": 1.2311, + "step": 3471 + }, + { + "epoch": 0.6182336182336182, + "grad_norm": 0.6172101497650146, + "learning_rate": 0.00018855160639511264, + "loss": 1.2651, + "step": 3472 + }, + { + "epoch": 0.6184116809116809, + "grad_norm": 0.49572527408599854, + "learning_rate": 0.00018854510217944465, + "loss": 1.1026, + "step": 3473 + }, + { + "epoch": 0.6185897435897436, + "grad_norm": 0.5373549461364746, + "learning_rate": 0.00018853859622891938, + "loss": 1.2562, + "step": 3474 + }, + { + "epoch": 0.6187678062678063, + "grad_norm": 0.5272396206855774, + "learning_rate": 0.0001885320885436642, + "loss": 1.1763, + "step": 3475 + }, + { + "epoch": 0.6189458689458689, + "grad_norm": 0.46584269404411316, + "learning_rate": 0.00018852557912380665, + "loss": 1.1762, + "step": 3476 + }, + { + "epoch": 0.6191239316239316, + "grad_norm": 0.4798245131969452, + "learning_rate": 0.0001885190679694743, + "loss": 0.9229, + "step": 3477 + }, + { + "epoch": 0.6193019943019943, + "grad_norm": 0.5221366286277771, + "learning_rate": 0.0001885125550807947, + "loss": 1.1078, + "step": 3478 + }, + { + "epoch": 0.6194800569800569, + "grad_norm": 0.5051897168159485, + "learning_rate": 0.0001885060404578954, + "loss": 1.0055, + "step": 3479 + }, + { + "epoch": 0.6196581196581197, + "grad_norm": 0.492662250995636, + "learning_rate": 0.00018849952410090413, + "loss": 1.1172, + "step": 3480 + }, + { + "epoch": 0.6198361823361823, + "grad_norm": 0.4906775951385498, + "learning_rate": 0.00018849300600994853, + "loss": 1.1223, + "step": 3481 + }, + { + "epoch": 0.6200142450142451, + "grad_norm": 0.5032641291618347, + "learning_rate": 0.0001884864861851563, + "loss": 0.9541, + "step": 3482 + }, + { + "epoch": 0.6201923076923077, + "grad_norm": 0.5262296795845032, + "learning_rate": 0.00018847996462665521, + "loss": 1.021, + "step": 3483 + }, + { + "epoch": 0.6203703703703703, + "grad_norm": 0.5253522992134094, + "learning_rate": 0.00018847344133457295, + "loss": 0.9075, + "step": 3484 + }, + { + "epoch": 0.6205484330484331, + "grad_norm": 0.4204299747943878, + "learning_rate": 0.00018846691630903744, + "loss": 0.895, + "step": 3485 + }, + { + "epoch": 0.6207264957264957, + "grad_norm": 0.557604193687439, + "learning_rate": 0.0001884603895501765, + "loss": 1.1758, + "step": 3486 + }, + { + "epoch": 0.6209045584045584, + "grad_norm": 0.5981321930885315, + "learning_rate": 0.00018845386105811795, + "loss": 1.1087, + "step": 3487 + }, + { + "epoch": 0.6210826210826211, + "grad_norm": 0.5285581946372986, + "learning_rate": 0.00018844733083298975, + "loss": 1.0692, + "step": 3488 + }, + { + "epoch": 0.6212606837606838, + "grad_norm": 0.5403170585632324, + "learning_rate": 0.00018844079887491986, + "loss": 1.1998, + "step": 3489 + }, + { + "epoch": 0.6214387464387464, + "grad_norm": 0.5471615791320801, + "learning_rate": 0.0001884342651840362, + "loss": 0.9556, + "step": 3490 + }, + { + "epoch": 0.6216168091168092, + "grad_norm": 0.6126871705055237, + "learning_rate": 0.00018842772976046686, + "loss": 1.2629, + "step": 3491 + }, + { + "epoch": 0.6217948717948718, + "grad_norm": 0.45669353008270264, + "learning_rate": 0.00018842119260433982, + "loss": 1.0203, + "step": 3492 + }, + { + "epoch": 0.6219729344729344, + "grad_norm": 0.4998520612716675, + "learning_rate": 0.0001884146537157832, + "loss": 1.0271, + "step": 3493 + }, + { + "epoch": 0.6221509971509972, + "grad_norm": 0.5820242166519165, + "learning_rate": 0.00018840811309492507, + "loss": 1.0321, + "step": 3494 + }, + { + "epoch": 0.6223290598290598, + "grad_norm": 0.581676185131073, + "learning_rate": 0.00018840157074189367, + "loss": 0.9219, + "step": 3495 + }, + { + "epoch": 0.6225071225071225, + "grad_norm": 0.6044120788574219, + "learning_rate": 0.0001883950266568171, + "loss": 1.1621, + "step": 3496 + }, + { + "epoch": 0.6226851851851852, + "grad_norm": 0.5448858737945557, + "learning_rate": 0.0001883884808398236, + "loss": 1.0686, + "step": 3497 + }, + { + "epoch": 0.6228632478632479, + "grad_norm": 0.4921551048755646, + "learning_rate": 0.00018838193329104143, + "loss": 1.2259, + "step": 3498 + }, + { + "epoch": 0.6230413105413105, + "grad_norm": 0.5374335646629333, + "learning_rate": 0.00018837538401059888, + "loss": 1.2608, + "step": 3499 + }, + { + "epoch": 0.6232193732193733, + "grad_norm": 0.5123008489608765, + "learning_rate": 0.0001883688329986243, + "loss": 0.8682, + "step": 3500 + }, + { + "epoch": 0.6233974358974359, + "grad_norm": 0.566145122051239, + "learning_rate": 0.00018836228025524595, + "loss": 1.1807, + "step": 3501 + }, + { + "epoch": 0.6235754985754985, + "grad_norm": 0.6658587455749512, + "learning_rate": 0.00018835572578059233, + "loss": 1.1641, + "step": 3502 + }, + { + "epoch": 0.6237535612535613, + "grad_norm": 0.4992465078830719, + "learning_rate": 0.00018834916957479177, + "loss": 0.9125, + "step": 3503 + }, + { + "epoch": 0.6239316239316239, + "grad_norm": 0.5081812739372253, + "learning_rate": 0.00018834261163797278, + "loss": 1.0939, + "step": 3504 + }, + { + "epoch": 0.6241096866096866, + "grad_norm": 0.5168607234954834, + "learning_rate": 0.0001883360519702638, + "loss": 1.2382, + "step": 3505 + }, + { + "epoch": 0.6242877492877493, + "grad_norm": 0.5517697334289551, + "learning_rate": 0.00018832949057179344, + "loss": 1.206, + "step": 3506 + }, + { + "epoch": 0.624465811965812, + "grad_norm": 0.4505497217178345, + "learning_rate": 0.00018832292744269013, + "loss": 0.8485, + "step": 3507 + }, + { + "epoch": 0.6246438746438746, + "grad_norm": 0.5230690240859985, + "learning_rate": 0.0001883163625830826, + "loss": 1.1701, + "step": 3508 + }, + { + "epoch": 0.6248219373219374, + "grad_norm": 0.5062205195426941, + "learning_rate": 0.00018830979599309937, + "loss": 1.0602, + "step": 3509 + }, + { + "epoch": 0.625, + "grad_norm": 0.49922460317611694, + "learning_rate": 0.00018830322767286913, + "loss": 1.1937, + "step": 3510 + }, + { + "epoch": 0.6251780626780626, + "grad_norm": 0.4637366831302643, + "learning_rate": 0.0001882966576225206, + "loss": 1.038, + "step": 3511 + }, + { + "epoch": 0.6253561253561254, + "grad_norm": 0.5330080389976501, + "learning_rate": 0.00018829008584218246, + "loss": 0.9308, + "step": 3512 + }, + { + "epoch": 0.625534188034188, + "grad_norm": 0.5443428754806519, + "learning_rate": 0.0001882835123319835, + "loss": 1.0006, + "step": 3513 + }, + { + "epoch": 0.6257122507122507, + "grad_norm": 0.5534018874168396, + "learning_rate": 0.00018827693709205253, + "loss": 1.2383, + "step": 3514 + }, + { + "epoch": 0.6258903133903134, + "grad_norm": 0.49207547307014465, + "learning_rate": 0.00018827036012251832, + "loss": 0.9804, + "step": 3515 + }, + { + "epoch": 0.6260683760683761, + "grad_norm": 0.4900086224079132, + "learning_rate": 0.0001882637814235098, + "loss": 1.012, + "step": 3516 + }, + { + "epoch": 0.6262464387464387, + "grad_norm": 0.5267475247383118, + "learning_rate": 0.00018825720099515585, + "loss": 1.1104, + "step": 3517 + }, + { + "epoch": 0.6264245014245015, + "grad_norm": 0.5711902379989624, + "learning_rate": 0.00018825061883758534, + "loss": 1.0616, + "step": 3518 + }, + { + "epoch": 0.6266025641025641, + "grad_norm": 0.5007771849632263, + "learning_rate": 0.0001882440349509273, + "loss": 0.9578, + "step": 3519 + }, + { + "epoch": 0.6267806267806267, + "grad_norm": 0.5657192468643188, + "learning_rate": 0.00018823744933531075, + "loss": 1.2768, + "step": 3520 + }, + { + "epoch": 0.6269586894586895, + "grad_norm": 0.6077173352241516, + "learning_rate": 0.00018823086199086462, + "loss": 1.147, + "step": 3521 + }, + { + "epoch": 0.6271367521367521, + "grad_norm": 0.5114718079566956, + "learning_rate": 0.000188224272917718, + "loss": 1.1176, + "step": 3522 + }, + { + "epoch": 0.6273148148148148, + "grad_norm": 0.4831676185131073, + "learning_rate": 0.0001882176821160001, + "loss": 0.8021, + "step": 3523 + }, + { + "epoch": 0.6274928774928775, + "grad_norm": 0.6327390670776367, + "learning_rate": 0.00018821108958583994, + "loss": 0.9449, + "step": 3524 + }, + { + "epoch": 0.6276709401709402, + "grad_norm": 0.5541796684265137, + "learning_rate": 0.00018820449532736672, + "loss": 1.2018, + "step": 3525 + }, + { + "epoch": 0.6278490028490028, + "grad_norm": 0.5224639773368835, + "learning_rate": 0.00018819789934070968, + "loss": 1.0138, + "step": 3526 + }, + { + "epoch": 0.6280270655270656, + "grad_norm": 0.49359360337257385, + "learning_rate": 0.00018819130162599798, + "loss": 1.0768, + "step": 3527 + }, + { + "epoch": 0.6282051282051282, + "grad_norm": 0.5525050759315491, + "learning_rate": 0.00018818470218336092, + "loss": 1.0883, + "step": 3528 + }, + { + "epoch": 0.6283831908831908, + "grad_norm": 0.5563427209854126, + "learning_rate": 0.00018817810101292787, + "loss": 1.1491, + "step": 3529 + }, + { + "epoch": 0.6285612535612536, + "grad_norm": 0.49363306164741516, + "learning_rate": 0.00018817149811482803, + "loss": 1.1409, + "step": 3530 + }, + { + "epoch": 0.6287393162393162, + "grad_norm": 0.5102340579032898, + "learning_rate": 0.00018816489348919086, + "loss": 1.1914, + "step": 3531 + }, + { + "epoch": 0.6289173789173789, + "grad_norm": 0.5173332691192627, + "learning_rate": 0.00018815828713614576, + "loss": 0.9308, + "step": 3532 + }, + { + "epoch": 0.6290954415954416, + "grad_norm": 0.5093010067939758, + "learning_rate": 0.00018815167905582216, + "loss": 0.9429, + "step": 3533 + }, + { + "epoch": 0.6292735042735043, + "grad_norm": 0.5453153848648071, + "learning_rate": 0.00018814506924834954, + "loss": 1.0147, + "step": 3534 + }, + { + "epoch": 0.6294515669515669, + "grad_norm": 0.5850773453712463, + "learning_rate": 0.00018813845771385737, + "loss": 1.3372, + "step": 3535 + }, + { + "epoch": 0.6296296296296297, + "grad_norm": 0.5095621943473816, + "learning_rate": 0.00018813184445247525, + "loss": 1.0515, + "step": 3536 + }, + { + "epoch": 0.6298076923076923, + "grad_norm": 0.6216054558753967, + "learning_rate": 0.00018812522946433266, + "loss": 0.8703, + "step": 3537 + }, + { + "epoch": 0.6299857549857549, + "grad_norm": 0.4945531189441681, + "learning_rate": 0.00018811861274955932, + "loss": 1.1485, + "step": 3538 + }, + { + "epoch": 0.6301638176638177, + "grad_norm": 0.47882601618766785, + "learning_rate": 0.00018811199430828477, + "loss": 1.1107, + "step": 3539 + }, + { + "epoch": 0.6303418803418803, + "grad_norm": 0.5005326867103577, + "learning_rate": 0.00018810537414063876, + "loss": 1.0237, + "step": 3540 + }, + { + "epoch": 0.6305199430199431, + "grad_norm": 0.5382370352745056, + "learning_rate": 0.00018809875224675093, + "loss": 0.9965, + "step": 3541 + }, + { + "epoch": 0.6306980056980057, + "grad_norm": 0.47002625465393066, + "learning_rate": 0.0001880921286267511, + "loss": 1.065, + "step": 3542 + }, + { + "epoch": 0.6308760683760684, + "grad_norm": 0.4519105851650238, + "learning_rate": 0.00018808550328076897, + "loss": 0.9312, + "step": 3543 + }, + { + "epoch": 0.6310541310541311, + "grad_norm": 0.45360881090164185, + "learning_rate": 0.0001880788762089344, + "loss": 1.0739, + "step": 3544 + }, + { + "epoch": 0.6312321937321937, + "grad_norm": 0.5578218698501587, + "learning_rate": 0.00018807224741137723, + "loss": 1.2478, + "step": 3545 + }, + { + "epoch": 0.6314102564102564, + "grad_norm": 0.4838615655899048, + "learning_rate": 0.0001880656168882273, + "loss": 1.0221, + "step": 3546 + }, + { + "epoch": 0.6315883190883191, + "grad_norm": 0.5733556747436523, + "learning_rate": 0.0001880589846396146, + "loss": 1.1249, + "step": 3547 + }, + { + "epoch": 0.6317663817663818, + "grad_norm": 0.4939686954021454, + "learning_rate": 0.00018805235066566894, + "loss": 0.8559, + "step": 3548 + }, + { + "epoch": 0.6319444444444444, + "grad_norm": 0.5072234869003296, + "learning_rate": 0.00018804571496652044, + "loss": 1.0842, + "step": 3549 + }, + { + "epoch": 0.6321225071225072, + "grad_norm": 0.4640493392944336, + "learning_rate": 0.00018803907754229903, + "loss": 1.0728, + "step": 3550 + }, + { + "epoch": 0.6323005698005698, + "grad_norm": 0.5314788818359375, + "learning_rate": 0.00018803243839313481, + "loss": 1.0752, + "step": 3551 + }, + { + "epoch": 0.6324786324786325, + "grad_norm": 0.5511462092399597, + "learning_rate": 0.0001880257975191578, + "loss": 1.0238, + "step": 3552 + }, + { + "epoch": 0.6326566951566952, + "grad_norm": 0.4980711042881012, + "learning_rate": 0.00018801915492049816, + "loss": 1.0981, + "step": 3553 + }, + { + "epoch": 0.6328347578347578, + "grad_norm": 0.7746123671531677, + "learning_rate": 0.00018801251059728604, + "loss": 1.0968, + "step": 3554 + }, + { + "epoch": 0.6330128205128205, + "grad_norm": 0.5006106495857239, + "learning_rate": 0.00018800586454965155, + "loss": 1.1802, + "step": 3555 + }, + { + "epoch": 0.6331908831908832, + "grad_norm": 0.49427780508995056, + "learning_rate": 0.000187999216777725, + "loss": 1.1257, + "step": 3556 + }, + { + "epoch": 0.6333689458689459, + "grad_norm": 0.5484146475791931, + "learning_rate": 0.00018799256728163662, + "loss": 1.1344, + "step": 3557 + }, + { + "epoch": 0.6335470085470085, + "grad_norm": 0.5007877349853516, + "learning_rate": 0.00018798591606151662, + "loss": 1.1328, + "step": 3558 + }, + { + "epoch": 0.6337250712250713, + "grad_norm": 0.5068148970603943, + "learning_rate": 0.00018797926311749544, + "loss": 0.976, + "step": 3559 + }, + { + "epoch": 0.6339031339031339, + "grad_norm": 0.44936859607696533, + "learning_rate": 0.00018797260844970334, + "loss": 0.9735, + "step": 3560 + }, + { + "epoch": 0.6340811965811965, + "grad_norm": 0.4592931866645813, + "learning_rate": 0.0001879659520582707, + "loss": 1.1306, + "step": 3561 + }, + { + "epoch": 0.6342592592592593, + "grad_norm": 0.4664020836353302, + "learning_rate": 0.00018795929394332795, + "loss": 1.0577, + "step": 3562 + }, + { + "epoch": 0.6344373219373219, + "grad_norm": 0.5638116002082825, + "learning_rate": 0.00018795263410500556, + "loss": 1.1747, + "step": 3563 + }, + { + "epoch": 0.6346153846153846, + "grad_norm": 0.524736225605011, + "learning_rate": 0.00018794597254343401, + "loss": 0.8964, + "step": 3564 + }, + { + "epoch": 0.6347934472934473, + "grad_norm": 0.4645404517650604, + "learning_rate": 0.00018793930925874386, + "loss": 0.8673, + "step": 3565 + }, + { + "epoch": 0.63497150997151, + "grad_norm": 0.4800064265727997, + "learning_rate": 0.00018793264425106558, + "loss": 1.0334, + "step": 3566 + }, + { + "epoch": 0.6351495726495726, + "grad_norm": 0.6202501058578491, + "learning_rate": 0.0001879259775205298, + "loss": 1.1061, + "step": 3567 + }, + { + "epoch": 0.6353276353276354, + "grad_norm": 0.503383457660675, + "learning_rate": 0.00018791930906726718, + "loss": 0.8545, + "step": 3568 + }, + { + "epoch": 0.635505698005698, + "grad_norm": 0.5256780982017517, + "learning_rate": 0.00018791263889140832, + "loss": 1.0785, + "step": 3569 + }, + { + "epoch": 0.6356837606837606, + "grad_norm": 0.47562023997306824, + "learning_rate": 0.00018790596699308392, + "loss": 1.0041, + "step": 3570 + }, + { + "epoch": 0.6358618233618234, + "grad_norm": 0.5103238224983215, + "learning_rate": 0.00018789929337242469, + "loss": 1.1488, + "step": 3571 + }, + { + "epoch": 0.636039886039886, + "grad_norm": 0.5023695826530457, + "learning_rate": 0.0001878926180295614, + "loss": 1.0696, + "step": 3572 + }, + { + "epoch": 0.6362179487179487, + "grad_norm": 0.5302290916442871, + "learning_rate": 0.00018788594096462487, + "loss": 1.0554, + "step": 3573 + }, + { + "epoch": 0.6363960113960114, + "grad_norm": 0.4798361361026764, + "learning_rate": 0.00018787926217774588, + "loss": 0.8872, + "step": 3574 + }, + { + "epoch": 0.6365740740740741, + "grad_norm": 0.5529209971427917, + "learning_rate": 0.00018787258166905527, + "loss": 1.0976, + "step": 3575 + }, + { + "epoch": 0.6367521367521367, + "grad_norm": 0.49757125973701477, + "learning_rate": 0.00018786589943868402, + "loss": 1.0049, + "step": 3576 + }, + { + "epoch": 0.6369301994301995, + "grad_norm": 0.5497848391532898, + "learning_rate": 0.00018785921548676295, + "loss": 1.2272, + "step": 3577 + }, + { + "epoch": 0.6371082621082621, + "grad_norm": 0.5061752200126648, + "learning_rate": 0.0001878525298134231, + "loss": 1.0307, + "step": 3578 + }, + { + "epoch": 0.6372863247863247, + "grad_norm": 0.5427432656288147, + "learning_rate": 0.00018784584241879538, + "loss": 1.1064, + "step": 3579 + }, + { + "epoch": 0.6374643874643875, + "grad_norm": 0.48312774300575256, + "learning_rate": 0.0001878391533030109, + "loss": 1.078, + "step": 3580 + }, + { + "epoch": 0.6376424501424501, + "grad_norm": 0.5059898495674133, + "learning_rate": 0.00018783246246620067, + "loss": 1.0922, + "step": 3581 + }, + { + "epoch": 0.6378205128205128, + "grad_norm": 0.5144124031066895, + "learning_rate": 0.00018782576990849581, + "loss": 1.0909, + "step": 3582 + }, + { + "epoch": 0.6379985754985755, + "grad_norm": 0.5535032153129578, + "learning_rate": 0.0001878190756300274, + "loss": 1.2579, + "step": 3583 + }, + { + "epoch": 0.6381766381766382, + "grad_norm": 0.49145692586898804, + "learning_rate": 0.00018781237963092667, + "loss": 1.0823, + "step": 3584 + }, + { + "epoch": 0.6383547008547008, + "grad_norm": 0.5245576500892639, + "learning_rate": 0.00018780568191132472, + "loss": 0.9595, + "step": 3585 + }, + { + "epoch": 0.6385327635327636, + "grad_norm": 0.5026637315750122, + "learning_rate": 0.00018779898247135287, + "loss": 1.153, + "step": 3586 + }, + { + "epoch": 0.6387108262108262, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018779228131114234, + "loss": 1.0661, + "step": 3587 + }, + { + "epoch": 0.6388888888888888, + "grad_norm": 0.517387330532074, + "learning_rate": 0.00018778557843082444, + "loss": 1.0113, + "step": 3588 + }, + { + "epoch": 0.6390669515669516, + "grad_norm": 0.5149948000907898, + "learning_rate": 0.00018777887383053047, + "loss": 0.9483, + "step": 3589 + }, + { + "epoch": 0.6392450142450142, + "grad_norm": 0.4854544997215271, + "learning_rate": 0.00018777216751039185, + "loss": 1.22, + "step": 3590 + }, + { + "epoch": 0.6394230769230769, + "grad_norm": 0.5317271947860718, + "learning_rate": 0.0001877654594705399, + "loss": 1.2483, + "step": 3591 + }, + { + "epoch": 0.6396011396011396, + "grad_norm": 0.4554755687713623, + "learning_rate": 0.0001877587497111061, + "loss": 0.9864, + "step": 3592 + }, + { + "epoch": 0.6397792022792023, + "grad_norm": 0.4833736717700958, + "learning_rate": 0.0001877520382322219, + "loss": 0.8895, + "step": 3593 + }, + { + "epoch": 0.6399572649572649, + "grad_norm": 0.5018072724342346, + "learning_rate": 0.00018774532503401878, + "loss": 1.2523, + "step": 3594 + }, + { + "epoch": 0.6401353276353277, + "grad_norm": 0.4478762447834015, + "learning_rate": 0.00018773861011662832, + "loss": 0.8833, + "step": 3595 + }, + { + "epoch": 0.6403133903133903, + "grad_norm": 0.5686985850334167, + "learning_rate": 0.00018773189348018205, + "loss": 0.9934, + "step": 3596 + }, + { + "epoch": 0.6404914529914529, + "grad_norm": 0.5144175291061401, + "learning_rate": 0.00018772517512481157, + "loss": 0.8149, + "step": 3597 + }, + { + "epoch": 0.6406695156695157, + "grad_norm": 0.5359936356544495, + "learning_rate": 0.00018771845505064852, + "loss": 1.1822, + "step": 3598 + }, + { + "epoch": 0.6408475783475783, + "grad_norm": 0.532573938369751, + "learning_rate": 0.00018771173325782457, + "loss": 1.0361, + "step": 3599 + }, + { + "epoch": 0.6410256410256411, + "grad_norm": 0.46121537685394287, + "learning_rate": 0.00018770500974647138, + "loss": 1.0792, + "step": 3600 + }, + { + "epoch": 0.6412037037037037, + "grad_norm": 0.4804821312427521, + "learning_rate": 0.00018769828451672076, + "loss": 1.1119, + "step": 3601 + }, + { + "epoch": 0.6413817663817664, + "grad_norm": 0.4955114722251892, + "learning_rate": 0.00018769155756870443, + "loss": 0.9312, + "step": 3602 + }, + { + "epoch": 0.6415598290598291, + "grad_norm": 0.4987298250198364, + "learning_rate": 0.00018768482890255415, + "loss": 1.2326, + "step": 3603 + }, + { + "epoch": 0.6417378917378918, + "grad_norm": 0.47216179966926575, + "learning_rate": 0.0001876780985184018, + "loss": 1.0114, + "step": 3604 + }, + { + "epoch": 0.6419159544159544, + "grad_norm": 0.5891931653022766, + "learning_rate": 0.0001876713664163793, + "loss": 1.2963, + "step": 3605 + }, + { + "epoch": 0.6420940170940171, + "grad_norm": 0.4645081162452698, + "learning_rate": 0.00018766463259661846, + "loss": 1.0874, + "step": 3606 + }, + { + "epoch": 0.6422720797720798, + "grad_norm": 0.5275476574897766, + "learning_rate": 0.00018765789705925125, + "loss": 0.9453, + "step": 3607 + }, + { + "epoch": 0.6424501424501424, + "grad_norm": 0.5884957313537598, + "learning_rate": 0.00018765115980440964, + "loss": 1.0796, + "step": 3608 + }, + { + "epoch": 0.6426282051282052, + "grad_norm": 0.4843178987503052, + "learning_rate": 0.00018764442083222567, + "loss": 1.1657, + "step": 3609 + }, + { + "epoch": 0.6428062678062678, + "grad_norm": 0.5188381671905518, + "learning_rate": 0.00018763768014283126, + "loss": 1.1109, + "step": 3610 + }, + { + "epoch": 0.6429843304843305, + "grad_norm": 0.4101468324661255, + "learning_rate": 0.00018763093773635863, + "loss": 0.895, + "step": 3611 + }, + { + "epoch": 0.6431623931623932, + "grad_norm": 0.4552084505558014, + "learning_rate": 0.00018762419361293979, + "loss": 0.9418, + "step": 3612 + }, + { + "epoch": 0.6433404558404558, + "grad_norm": 0.5924661159515381, + "learning_rate": 0.0001876174477727069, + "loss": 1.2562, + "step": 3613 + }, + { + "epoch": 0.6435185185185185, + "grad_norm": 0.5072348713874817, + "learning_rate": 0.00018761070021579212, + "loss": 1.1501, + "step": 3614 + }, + { + "epoch": 0.6436965811965812, + "grad_norm": 0.5312697887420654, + "learning_rate": 0.0001876039509423277, + "loss": 1.0751, + "step": 3615 + }, + { + "epoch": 0.6438746438746439, + "grad_norm": 0.6046462059020996, + "learning_rate": 0.0001875971999524458, + "loss": 1.0927, + "step": 3616 + }, + { + "epoch": 0.6440527065527065, + "grad_norm": 0.4992375373840332, + "learning_rate": 0.00018759044724627876, + "loss": 0.96, + "step": 3617 + }, + { + "epoch": 0.6442307692307693, + "grad_norm": 0.4983134865760803, + "learning_rate": 0.00018758369282395886, + "loss": 1.0599, + "step": 3618 + }, + { + "epoch": 0.6444088319088319, + "grad_norm": 0.5655683279037476, + "learning_rate": 0.00018757693668561843, + "loss": 1.2372, + "step": 3619 + }, + { + "epoch": 0.6445868945868946, + "grad_norm": 0.4968827962875366, + "learning_rate": 0.00018757017883138985, + "loss": 1.1639, + "step": 3620 + }, + { + "epoch": 0.6447649572649573, + "grad_norm": 0.5831420421600342, + "learning_rate": 0.00018756341926140553, + "loss": 0.9002, + "step": 3621 + }, + { + "epoch": 0.64494301994302, + "grad_norm": 0.4828467071056366, + "learning_rate": 0.0001875566579757979, + "loss": 0.9201, + "step": 3622 + }, + { + "epoch": 0.6451210826210826, + "grad_norm": 0.5067087411880493, + "learning_rate": 0.00018754989497469943, + "loss": 0.9874, + "step": 3623 + }, + { + "epoch": 0.6452991452991453, + "grad_norm": 0.5182318091392517, + "learning_rate": 0.00018754313025824267, + "loss": 1.1291, + "step": 3624 + }, + { + "epoch": 0.645477207977208, + "grad_norm": 0.472200483083725, + "learning_rate": 0.0001875363638265601, + "loss": 1.0286, + "step": 3625 + }, + { + "epoch": 0.6456552706552706, + "grad_norm": 0.4597308039665222, + "learning_rate": 0.0001875295956797843, + "loss": 0.7517, + "step": 3626 + }, + { + "epoch": 0.6458333333333334, + "grad_norm": 0.5358221530914307, + "learning_rate": 0.00018752282581804798, + "loss": 1.2264, + "step": 3627 + }, + { + "epoch": 0.646011396011396, + "grad_norm": 0.5268992781639099, + "learning_rate": 0.00018751605424148363, + "loss": 1.0801, + "step": 3628 + }, + { + "epoch": 0.6461894586894587, + "grad_norm": 0.5917379260063171, + "learning_rate": 0.00018750928095022403, + "loss": 0.9538, + "step": 3629 + }, + { + "epoch": 0.6463675213675214, + "grad_norm": 0.44506707787513733, + "learning_rate": 0.00018750250594440183, + "loss": 0.9818, + "step": 3630 + }, + { + "epoch": 0.646545584045584, + "grad_norm": 0.5578880906105042, + "learning_rate": 0.00018749572922414982, + "loss": 0.9958, + "step": 3631 + }, + { + "epoch": 0.6467236467236467, + "grad_norm": 0.5155318975448608, + "learning_rate": 0.00018748895078960076, + "loss": 1.2888, + "step": 3632 + }, + { + "epoch": 0.6469017094017094, + "grad_norm": 0.5117297768592834, + "learning_rate": 0.0001874821706408874, + "loss": 1.0452, + "step": 3633 + }, + { + "epoch": 0.6470797720797721, + "grad_norm": 0.5169841647148132, + "learning_rate": 0.00018747538877814267, + "loss": 1.1649, + "step": 3634 + }, + { + "epoch": 0.6472578347578347, + "grad_norm": 0.5001181960105896, + "learning_rate": 0.00018746860520149942, + "loss": 1.1472, + "step": 3635 + }, + { + "epoch": 0.6474358974358975, + "grad_norm": 0.6289856433868408, + "learning_rate": 0.00018746181991109056, + "loss": 1.0351, + "step": 3636 + }, + { + "epoch": 0.6476139601139601, + "grad_norm": 0.5490612983703613, + "learning_rate": 0.00018745503290704897, + "loss": 0.8938, + "step": 3637 + }, + { + "epoch": 0.6477920227920227, + "grad_norm": 0.47378283739089966, + "learning_rate": 0.00018744824418950775, + "loss": 0.937, + "step": 3638 + }, + { + "epoch": 0.6479700854700855, + "grad_norm": 0.6079059839248657, + "learning_rate": 0.0001874414537585998, + "loss": 1.0486, + "step": 3639 + }, + { + "epoch": 0.6481481481481481, + "grad_norm": 0.5351769924163818, + "learning_rate": 0.00018743466161445823, + "loss": 1.0316, + "step": 3640 + }, + { + "epoch": 0.6483262108262108, + "grad_norm": 0.5516425967216492, + "learning_rate": 0.0001874278677572161, + "loss": 1.1552, + "step": 3641 + }, + { + "epoch": 0.6485042735042735, + "grad_norm": 0.5027523636817932, + "learning_rate": 0.0001874210721870065, + "loss": 1.0491, + "step": 3642 + }, + { + "epoch": 0.6486823361823362, + "grad_norm": 0.5596168041229248, + "learning_rate": 0.00018741427490396258, + "loss": 1.0256, + "step": 3643 + }, + { + "epoch": 0.6488603988603988, + "grad_norm": 0.5601046681404114, + "learning_rate": 0.00018740747590821751, + "loss": 1.1604, + "step": 3644 + }, + { + "epoch": 0.6490384615384616, + "grad_norm": 0.49749523401260376, + "learning_rate": 0.0001874006751999046, + "loss": 1.0532, + "step": 3645 + }, + { + "epoch": 0.6492165242165242, + "grad_norm": 0.6226113438606262, + "learning_rate": 0.00018739387277915697, + "loss": 1.1402, + "step": 3646 + }, + { + "epoch": 0.6493945868945868, + "grad_norm": 0.6142009496688843, + "learning_rate": 0.00018738706864610794, + "loss": 1.2437, + "step": 3647 + }, + { + "epoch": 0.6495726495726496, + "grad_norm": 0.48814916610717773, + "learning_rate": 0.00018738026280089084, + "loss": 0.8429, + "step": 3648 + }, + { + "epoch": 0.6497507122507122, + "grad_norm": 0.5717982053756714, + "learning_rate": 0.00018737345524363902, + "loss": 1.1095, + "step": 3649 + }, + { + "epoch": 0.6499287749287749, + "grad_norm": 0.5150009989738464, + "learning_rate": 0.00018736664597448582, + "loss": 1.199, + "step": 3650 + }, + { + "epoch": 0.6501068376068376, + "grad_norm": 0.58461594581604, + "learning_rate": 0.00018735983499356472, + "loss": 1.0704, + "step": 3651 + }, + { + "epoch": 0.6502849002849003, + "grad_norm": 0.5108643770217896, + "learning_rate": 0.0001873530223010091, + "loss": 1.2039, + "step": 3652 + }, + { + "epoch": 0.6504629629629629, + "grad_norm": 0.513306736946106, + "learning_rate": 0.00018734620789695247, + "loss": 1.1448, + "step": 3653 + }, + { + "epoch": 0.6506410256410257, + "grad_norm": 0.5139986872673035, + "learning_rate": 0.00018733939178152835, + "loss": 1.0023, + "step": 3654 + }, + { + "epoch": 0.6508190883190883, + "grad_norm": 0.5187703967094421, + "learning_rate": 0.00018733257395487027, + "loss": 1.1304, + "step": 3655 + }, + { + "epoch": 0.6509971509971509, + "grad_norm": 0.5470501184463501, + "learning_rate": 0.00018732575441711183, + "loss": 1.0272, + "step": 3656 + }, + { + "epoch": 0.6511752136752137, + "grad_norm": 0.537309467792511, + "learning_rate": 0.00018731893316838665, + "loss": 1.0806, + "step": 3657 + }, + { + "epoch": 0.6513532763532763, + "grad_norm": 0.5187864899635315, + "learning_rate": 0.00018731211020882836, + "loss": 1.0154, + "step": 3658 + }, + { + "epoch": 0.6515313390313391, + "grad_norm": 0.48373252153396606, + "learning_rate": 0.00018730528553857062, + "loss": 1.0135, + "step": 3659 + }, + { + "epoch": 0.6517094017094017, + "grad_norm": 0.5645000338554382, + "learning_rate": 0.00018729845915774716, + "loss": 0.8924, + "step": 3660 + }, + { + "epoch": 0.6518874643874644, + "grad_norm": 0.5722129940986633, + "learning_rate": 0.00018729163106649178, + "loss": 1.2416, + "step": 3661 + }, + { + "epoch": 0.6520655270655271, + "grad_norm": 0.5904877185821533, + "learning_rate": 0.00018728480126493823, + "loss": 0.9792, + "step": 3662 + }, + { + "epoch": 0.6522435897435898, + "grad_norm": 0.5224713087081909, + "learning_rate": 0.00018727796975322026, + "loss": 1.079, + "step": 3663 + }, + { + "epoch": 0.6524216524216524, + "grad_norm": 0.5667217969894409, + "learning_rate": 0.00018727113653147184, + "loss": 1.1397, + "step": 3664 + }, + { + "epoch": 0.6525997150997151, + "grad_norm": 0.5274622440338135, + "learning_rate": 0.00018726430159982677, + "loss": 1.0569, + "step": 3665 + }, + { + "epoch": 0.6527777777777778, + "grad_norm": 0.5745310187339783, + "learning_rate": 0.00018725746495841896, + "loss": 1.2129, + "step": 3666 + }, + { + "epoch": 0.6529558404558404, + "grad_norm": 0.6123398542404175, + "learning_rate": 0.0001872506266073824, + "loss": 1.186, + "step": 3667 + }, + { + "epoch": 0.6531339031339032, + "grad_norm": 0.4983387291431427, + "learning_rate": 0.00018724378654685106, + "loss": 1.1957, + "step": 3668 + }, + { + "epoch": 0.6533119658119658, + "grad_norm": 0.5584192276000977, + "learning_rate": 0.00018723694477695897, + "loss": 1.0939, + "step": 3669 + }, + { + "epoch": 0.6534900284900285, + "grad_norm": 0.5318745374679565, + "learning_rate": 0.00018723010129784016, + "loss": 1.1869, + "step": 3670 + }, + { + "epoch": 0.6536680911680912, + "grad_norm": 0.4607617259025574, + "learning_rate": 0.0001872232561096287, + "loss": 0.8447, + "step": 3671 + }, + { + "epoch": 0.6538461538461539, + "grad_norm": 0.5312213897705078, + "learning_rate": 0.00018721640921245874, + "loss": 1.0623, + "step": 3672 + }, + { + "epoch": 0.6540242165242165, + "grad_norm": 0.5099136233329773, + "learning_rate": 0.0001872095606064644, + "loss": 0.7174, + "step": 3673 + }, + { + "epoch": 0.6542022792022792, + "grad_norm": 0.6894404888153076, + "learning_rate": 0.0001872027102917799, + "loss": 1.0251, + "step": 3674 + }, + { + "epoch": 0.6543803418803419, + "grad_norm": 0.5758535861968994, + "learning_rate": 0.00018719585826853944, + "loss": 1.1655, + "step": 3675 + }, + { + "epoch": 0.6545584045584045, + "grad_norm": 0.521824061870575, + "learning_rate": 0.0001871890045368773, + "loss": 1.1653, + "step": 3676 + }, + { + "epoch": 0.6547364672364673, + "grad_norm": 0.5370712280273438, + "learning_rate": 0.00018718214909692771, + "loss": 1.3152, + "step": 3677 + }, + { + "epoch": 0.6549145299145299, + "grad_norm": 0.4459827244281769, + "learning_rate": 0.000187175291948825, + "loss": 1.0953, + "step": 3678 + }, + { + "epoch": 0.6550925925925926, + "grad_norm": 0.44131460785865784, + "learning_rate": 0.00018716843309270353, + "loss": 0.8568, + "step": 3679 + }, + { + "epoch": 0.6552706552706553, + "grad_norm": 0.5529624819755554, + "learning_rate": 0.00018716157252869772, + "loss": 1.2085, + "step": 3680 + }, + { + "epoch": 0.655448717948718, + "grad_norm": 0.44604751467704773, + "learning_rate": 0.00018715471025694194, + "loss": 0.9605, + "step": 3681 + }, + { + "epoch": 0.6556267806267806, + "grad_norm": 0.4662449359893799, + "learning_rate": 0.0001871478462775707, + "loss": 1.2092, + "step": 3682 + }, + { + "epoch": 0.6558048433048433, + "grad_norm": 0.42632922530174255, + "learning_rate": 0.0001871409805907184, + "loss": 0.9141, + "step": 3683 + }, + { + "epoch": 0.655982905982906, + "grad_norm": 0.534009575843811, + "learning_rate": 0.00018713411319651958, + "loss": 1.0147, + "step": 3684 + }, + { + "epoch": 0.6561609686609686, + "grad_norm": 0.5433241724967957, + "learning_rate": 0.00018712724409510888, + "loss": 1.1998, + "step": 3685 + }, + { + "epoch": 0.6563390313390314, + "grad_norm": 0.4771319627761841, + "learning_rate": 0.0001871203732866208, + "loss": 1.0384, + "step": 3686 + }, + { + "epoch": 0.656517094017094, + "grad_norm": 0.507641077041626, + "learning_rate": 0.00018711350077119, + "loss": 0.9608, + "step": 3687 + }, + { + "epoch": 0.6566951566951567, + "grad_norm": 0.5069413185119629, + "learning_rate": 0.00018710662654895108, + "loss": 1.055, + "step": 3688 + }, + { + "epoch": 0.6568732193732194, + "grad_norm": 0.512340247631073, + "learning_rate": 0.00018709975062003876, + "loss": 0.9506, + "step": 3689 + }, + { + "epoch": 0.657051282051282, + "grad_norm": 0.5156390070915222, + "learning_rate": 0.00018709287298458778, + "loss": 1.0089, + "step": 3690 + }, + { + "epoch": 0.6572293447293447, + "grad_norm": 0.5101696252822876, + "learning_rate": 0.0001870859936427329, + "loss": 1.0441, + "step": 3691 + }, + { + "epoch": 0.6574074074074074, + "grad_norm": 0.4394689202308655, + "learning_rate": 0.00018707911259460884, + "loss": 0.9124, + "step": 3692 + }, + { + "epoch": 0.6575854700854701, + "grad_norm": 0.4842554032802582, + "learning_rate": 0.00018707222984035043, + "loss": 1.0051, + "step": 3693 + }, + { + "epoch": 0.6577635327635327, + "grad_norm": 0.6418108344078064, + "learning_rate": 0.00018706534538009262, + "loss": 1.1165, + "step": 3694 + }, + { + "epoch": 0.6579415954415955, + "grad_norm": 0.5596832036972046, + "learning_rate": 0.00018705845921397022, + "loss": 1.1127, + "step": 3695 + }, + { + "epoch": 0.6581196581196581, + "grad_norm": 0.6692909002304077, + "learning_rate": 0.00018705157134211813, + "loss": 1.2403, + "step": 3696 + }, + { + "epoch": 0.6582977207977208, + "grad_norm": 0.5046468377113342, + "learning_rate": 0.00018704468176467134, + "loss": 1.1016, + "step": 3697 + }, + { + "epoch": 0.6584757834757835, + "grad_norm": 0.6723586320877075, + "learning_rate": 0.00018703779048176485, + "loss": 1.1777, + "step": 3698 + }, + { + "epoch": 0.6586538461538461, + "grad_norm": 0.5269754528999329, + "learning_rate": 0.00018703089749353365, + "loss": 1.1441, + "step": 3699 + }, + { + "epoch": 0.6588319088319088, + "grad_norm": 0.5303323268890381, + "learning_rate": 0.0001870240028001128, + "loss": 1.07, + "step": 3700 + }, + { + "epoch": 0.6590099715099715, + "grad_norm": 0.4795511066913605, + "learning_rate": 0.00018701710640163738, + "loss": 1.0189, + "step": 3701 + }, + { + "epoch": 0.6591880341880342, + "grad_norm": 0.514659583568573, + "learning_rate": 0.00018701020829824255, + "loss": 1.0792, + "step": 3702 + }, + { + "epoch": 0.6593660968660968, + "grad_norm": 0.5407463312149048, + "learning_rate": 0.0001870033084900634, + "loss": 0.9346, + "step": 3703 + }, + { + "epoch": 0.6595441595441596, + "grad_norm": 0.5358424186706543, + "learning_rate": 0.0001869964069772352, + "loss": 1.1242, + "step": 3704 + }, + { + "epoch": 0.6597222222222222, + "grad_norm": 0.470825731754303, + "learning_rate": 0.00018698950375989307, + "loss": 0.9952, + "step": 3705 + }, + { + "epoch": 0.6599002849002849, + "grad_norm": 0.5711592435836792, + "learning_rate": 0.00018698259883817236, + "loss": 1.1678, + "step": 3706 + }, + { + "epoch": 0.6600783475783476, + "grad_norm": 0.5298995971679688, + "learning_rate": 0.00018697569221220832, + "loss": 0.869, + "step": 3707 + }, + { + "epoch": 0.6602564102564102, + "grad_norm": 0.5453875064849854, + "learning_rate": 0.00018696878388213626, + "loss": 0.9706, + "step": 3708 + }, + { + "epoch": 0.6604344729344729, + "grad_norm": 0.6219926476478577, + "learning_rate": 0.00018696187384809154, + "loss": 1.1902, + "step": 3709 + }, + { + "epoch": 0.6606125356125356, + "grad_norm": 0.5972491502761841, + "learning_rate": 0.00018695496211020953, + "loss": 1.2054, + "step": 3710 + }, + { + "epoch": 0.6607905982905983, + "grad_norm": 0.5048904418945312, + "learning_rate": 0.0001869480486686257, + "loss": 1.0405, + "step": 3711 + }, + { + "epoch": 0.6609686609686609, + "grad_norm": 0.5474200248718262, + "learning_rate": 0.00018694113352347546, + "loss": 1.09, + "step": 3712 + }, + { + "epoch": 0.6611467236467237, + "grad_norm": 0.5073318481445312, + "learning_rate": 0.00018693421667489432, + "loss": 1.0698, + "step": 3713 + }, + { + "epoch": 0.6613247863247863, + "grad_norm": 0.5693208575248718, + "learning_rate": 0.0001869272981230178, + "loss": 0.9664, + "step": 3714 + }, + { + "epoch": 0.6615028490028491, + "grad_norm": 0.5678503513336182, + "learning_rate": 0.00018692037786798143, + "loss": 1.0895, + "step": 3715 + }, + { + "epoch": 0.6616809116809117, + "grad_norm": 0.4950976073741913, + "learning_rate": 0.00018691345590992082, + "loss": 0.9584, + "step": 3716 + }, + { + "epoch": 0.6618589743589743, + "grad_norm": 0.4944666624069214, + "learning_rate": 0.0001869065322489716, + "loss": 0.8607, + "step": 3717 + }, + { + "epoch": 0.6620370370370371, + "grad_norm": 0.5197804570198059, + "learning_rate": 0.0001868996068852694, + "loss": 1.2335, + "step": 3718 + }, + { + "epoch": 0.6622150997150997, + "grad_norm": 0.6550365686416626, + "learning_rate": 0.00018689267981894994, + "loss": 1.0441, + "step": 3719 + }, + { + "epoch": 0.6623931623931624, + "grad_norm": 0.5331503748893738, + "learning_rate": 0.00018688575105014888, + "loss": 1.1696, + "step": 3720 + }, + { + "epoch": 0.6625712250712251, + "grad_norm": 0.47304239869117737, + "learning_rate": 0.00018687882057900207, + "loss": 0.9695, + "step": 3721 + }, + { + "epoch": 0.6627492877492878, + "grad_norm": 0.5653772354125977, + "learning_rate": 0.00018687188840564524, + "loss": 1.2082, + "step": 3722 + }, + { + "epoch": 0.6629273504273504, + "grad_norm": 0.5323491096496582, + "learning_rate": 0.00018686495453021417, + "loss": 0.9106, + "step": 3723 + }, + { + "epoch": 0.6631054131054132, + "grad_norm": 0.5612817406654358, + "learning_rate": 0.00018685801895284483, + "loss": 1.1302, + "step": 3724 + }, + { + "epoch": 0.6632834757834758, + "grad_norm": 0.4562164545059204, + "learning_rate": 0.000186851081673673, + "loss": 0.8886, + "step": 3725 + }, + { + "epoch": 0.6634615384615384, + "grad_norm": 0.5006430745124817, + "learning_rate": 0.00018684414269283463, + "loss": 0.9128, + "step": 3726 + }, + { + "epoch": 0.6636396011396012, + "grad_norm": 0.5305442810058594, + "learning_rate": 0.0001868372020104657, + "loss": 1.1766, + "step": 3727 + }, + { + "epoch": 0.6638176638176638, + "grad_norm": 0.6129274368286133, + "learning_rate": 0.0001868302596267022, + "loss": 1.04, + "step": 3728 + }, + { + "epoch": 0.6639957264957265, + "grad_norm": 0.5530399084091187, + "learning_rate": 0.00018682331554168013, + "loss": 1.4114, + "step": 3729 + }, + { + "epoch": 0.6641737891737892, + "grad_norm": 0.5397193431854248, + "learning_rate": 0.00018681636975553557, + "loss": 1.1945, + "step": 3730 + }, + { + "epoch": 0.6643518518518519, + "grad_norm": 0.5510205030441284, + "learning_rate": 0.00018680942226840456, + "loss": 1.0489, + "step": 3731 + }, + { + "epoch": 0.6645299145299145, + "grad_norm": 0.5519221425056458, + "learning_rate": 0.00018680247308042324, + "loss": 1.1633, + "step": 3732 + }, + { + "epoch": 0.6647079772079773, + "grad_norm": 0.4848768711090088, + "learning_rate": 0.00018679552219172784, + "loss": 0.8716, + "step": 3733 + }, + { + "epoch": 0.6648860398860399, + "grad_norm": 0.5490246415138245, + "learning_rate": 0.0001867885696024544, + "loss": 1.1347, + "step": 3734 + }, + { + "epoch": 0.6650641025641025, + "grad_norm": 0.5281458497047424, + "learning_rate": 0.00018678161531273928, + "loss": 1.0987, + "step": 3735 + }, + { + "epoch": 0.6652421652421653, + "grad_norm": 0.5313079953193665, + "learning_rate": 0.00018677465932271867, + "loss": 0.9705, + "step": 3736 + }, + { + "epoch": 0.6654202279202279, + "grad_norm": 0.5425750017166138, + "learning_rate": 0.0001867677016325289, + "loss": 1.1847, + "step": 3737 + }, + { + "epoch": 0.6655982905982906, + "grad_norm": 0.5796298980712891, + "learning_rate": 0.0001867607422423062, + "loss": 1.2639, + "step": 3738 + }, + { + "epoch": 0.6657763532763533, + "grad_norm": 0.49738675355911255, + "learning_rate": 0.00018675378115218702, + "loss": 1.0536, + "step": 3739 + }, + { + "epoch": 0.665954415954416, + "grad_norm": 0.665250301361084, + "learning_rate": 0.0001867468183623077, + "loss": 1.2836, + "step": 3740 + }, + { + "epoch": 0.6661324786324786, + "grad_norm": 0.5184717178344727, + "learning_rate": 0.00018673985387280469, + "loss": 1.0497, + "step": 3741 + }, + { + "epoch": 0.6663105413105413, + "grad_norm": 0.5129656791687012, + "learning_rate": 0.00018673288768381442, + "loss": 1.2041, + "step": 3742 + }, + { + "epoch": 0.666488603988604, + "grad_norm": 0.5308768153190613, + "learning_rate": 0.00018672591979547337, + "loss": 1.2092, + "step": 3743 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.5059141516685486, + "learning_rate": 0.00018671895020791812, + "loss": 1.1929, + "step": 3744 + }, + { + "epoch": 0.6668447293447294, + "grad_norm": 0.5237857103347778, + "learning_rate": 0.00018671197892128517, + "loss": 1.2538, + "step": 3745 + }, + { + "epoch": 0.667022792022792, + "grad_norm": 0.450000137090683, + "learning_rate": 0.0001867050059357111, + "loss": 0.7138, + "step": 3746 + }, + { + "epoch": 0.6672008547008547, + "grad_norm": 0.5413795709609985, + "learning_rate": 0.00018669803125133258, + "loss": 1.1383, + "step": 3747 + }, + { + "epoch": 0.6673789173789174, + "grad_norm": 0.4657825529575348, + "learning_rate": 0.00018669105486828622, + "loss": 1.0518, + "step": 3748 + }, + { + "epoch": 0.66755698005698, + "grad_norm": 0.6198551654815674, + "learning_rate": 0.00018668407678670875, + "loss": 1.2697, + "step": 3749 + }, + { + "epoch": 0.6677350427350427, + "grad_norm": 0.5112186074256897, + "learning_rate": 0.00018667709700673685, + "loss": 0.9907, + "step": 3750 + }, + { + "epoch": 0.6679131054131054, + "grad_norm": 0.5446593761444092, + "learning_rate": 0.00018667011552850728, + "loss": 1.0708, + "step": 3751 + }, + { + "epoch": 0.6680911680911681, + "grad_norm": 0.5673866271972656, + "learning_rate": 0.00018666313235215682, + "loss": 1.05, + "step": 3752 + }, + { + "epoch": 0.6682692307692307, + "grad_norm": 0.4821988046169281, + "learning_rate": 0.00018665614747782235, + "loss": 1.0543, + "step": 3753 + }, + { + "epoch": 0.6684472934472935, + "grad_norm": 0.5158842206001282, + "learning_rate": 0.00018664916090564067, + "loss": 1.0331, + "step": 3754 + }, + { + "epoch": 0.6686253561253561, + "grad_norm": 0.45486921072006226, + "learning_rate": 0.00018664217263574865, + "loss": 0.9262, + "step": 3755 + }, + { + "epoch": 0.6688034188034188, + "grad_norm": 0.46193036437034607, + "learning_rate": 0.00018663518266828327, + "loss": 0.9858, + "step": 3756 + }, + { + "epoch": 0.6689814814814815, + "grad_norm": 0.5144094824790955, + "learning_rate": 0.00018662819100338148, + "loss": 1.0302, + "step": 3757 + }, + { + "epoch": 0.6691595441595442, + "grad_norm": 0.5246134400367737, + "learning_rate": 0.0001866211976411802, + "loss": 1.064, + "step": 3758 + }, + { + "epoch": 0.6693376068376068, + "grad_norm": 0.4853166937828064, + "learning_rate": 0.0001866142025818165, + "loss": 0.9481, + "step": 3759 + }, + { + "epoch": 0.6695156695156695, + "grad_norm": 0.5029586553573608, + "learning_rate": 0.00018660720582542743, + "loss": 0.9443, + "step": 3760 + }, + { + "epoch": 0.6696937321937322, + "grad_norm": 0.5373172163963318, + "learning_rate": 0.0001866002073721501, + "loss": 1.1401, + "step": 3761 + }, + { + "epoch": 0.6698717948717948, + "grad_norm": 0.6236287951469421, + "learning_rate": 0.00018659320722212158, + "loss": 1.1255, + "step": 3762 + }, + { + "epoch": 0.6700498575498576, + "grad_norm": 0.5470684766769409, + "learning_rate": 0.00018658620537547903, + "loss": 1.0622, + "step": 3763 + }, + { + "epoch": 0.6702279202279202, + "grad_norm": 0.63177090883255, + "learning_rate": 0.00018657920183235964, + "loss": 0.9736, + "step": 3764 + }, + { + "epoch": 0.6704059829059829, + "grad_norm": 0.5456309914588928, + "learning_rate": 0.00018657219659290068, + "loss": 1.027, + "step": 3765 + }, + { + "epoch": 0.6705840455840456, + "grad_norm": 0.4816138744354248, + "learning_rate": 0.00018656518965723935, + "loss": 0.7801, + "step": 3766 + }, + { + "epoch": 0.6707621082621082, + "grad_norm": 0.4811640679836273, + "learning_rate": 0.00018655818102551294, + "loss": 1.0535, + "step": 3767 + }, + { + "epoch": 0.6709401709401709, + "grad_norm": 0.4677673280239105, + "learning_rate": 0.00018655117069785884, + "loss": 1.1043, + "step": 3768 + }, + { + "epoch": 0.6711182336182336, + "grad_norm": 0.5628635883331299, + "learning_rate": 0.0001865441586744143, + "loss": 1.0392, + "step": 3769 + }, + { + "epoch": 0.6712962962962963, + "grad_norm": 0.5484504103660583, + "learning_rate": 0.00018653714495531673, + "loss": 1.1533, + "step": 3770 + }, + { + "epoch": 0.6714743589743589, + "grad_norm": 0.5830571055412292, + "learning_rate": 0.0001865301295407036, + "loss": 1.2479, + "step": 3771 + }, + { + "epoch": 0.6716524216524217, + "grad_norm": 0.5516841411590576, + "learning_rate": 0.00018652311243071235, + "loss": 1.2152, + "step": 3772 + }, + { + "epoch": 0.6718304843304843, + "grad_norm": 0.6360766291618347, + "learning_rate": 0.0001865160936254804, + "loss": 1.0752, + "step": 3773 + }, + { + "epoch": 0.6720085470085471, + "grad_norm": 0.6038610935211182, + "learning_rate": 0.00018650907312514533, + "loss": 1.2425, + "step": 3774 + }, + { + "epoch": 0.6721866096866097, + "grad_norm": 0.49572908878326416, + "learning_rate": 0.0001865020509298447, + "loss": 1.0057, + "step": 3775 + }, + { + "epoch": 0.6723646723646723, + "grad_norm": 0.4551616311073303, + "learning_rate": 0.00018649502703971607, + "loss": 1.0763, + "step": 3776 + }, + { + "epoch": 0.6725427350427351, + "grad_norm": 0.6621482372283936, + "learning_rate": 0.00018648800145489706, + "loss": 1.0306, + "step": 3777 + }, + { + "epoch": 0.6727207977207977, + "grad_norm": 0.5523806810379028, + "learning_rate": 0.0001864809741755253, + "loss": 0.9906, + "step": 3778 + }, + { + "epoch": 0.6728988603988604, + "grad_norm": 0.5527048110961914, + "learning_rate": 0.00018647394520173856, + "loss": 1.0734, + "step": 3779 + }, + { + "epoch": 0.6730769230769231, + "grad_norm": 0.573573887348175, + "learning_rate": 0.00018646691453367444, + "loss": 1.1409, + "step": 3780 + }, + { + "epoch": 0.6732549857549858, + "grad_norm": 0.6273239254951477, + "learning_rate": 0.00018645988217147079, + "loss": 0.9682, + "step": 3781 + }, + { + "epoch": 0.6734330484330484, + "grad_norm": 0.4917762279510498, + "learning_rate": 0.00018645284811526534, + "loss": 0.9681, + "step": 3782 + }, + { + "epoch": 0.6736111111111112, + "grad_norm": 0.4901154339313507, + "learning_rate": 0.0001864458123651959, + "loss": 1.1828, + "step": 3783 + }, + { + "epoch": 0.6737891737891738, + "grad_norm": 0.6292546391487122, + "learning_rate": 0.00018643877492140036, + "loss": 1.1987, + "step": 3784 + }, + { + "epoch": 0.6739672364672364, + "grad_norm": 0.5334137678146362, + "learning_rate": 0.0001864317357840166, + "loss": 1.0347, + "step": 3785 + }, + { + "epoch": 0.6741452991452992, + "grad_norm": 0.6064338684082031, + "learning_rate": 0.0001864246949531825, + "loss": 1.4154, + "step": 3786 + }, + { + "epoch": 0.6743233618233618, + "grad_norm": 0.5442034602165222, + "learning_rate": 0.000186417652429036, + "loss": 1.2604, + "step": 3787 + }, + { + "epoch": 0.6745014245014245, + "grad_norm": 0.490858793258667, + "learning_rate": 0.00018641060821171518, + "loss": 1.1511, + "step": 3788 + }, + { + "epoch": 0.6746794871794872, + "grad_norm": 0.571116030216217, + "learning_rate": 0.00018640356230135798, + "loss": 1.1479, + "step": 3789 + }, + { + "epoch": 0.6748575498575499, + "grad_norm": 0.4857785105705261, + "learning_rate": 0.00018639651469810247, + "loss": 0.9, + "step": 3790 + }, + { + "epoch": 0.6750356125356125, + "grad_norm": 0.5320703983306885, + "learning_rate": 0.0001863894654020867, + "loss": 1.2284, + "step": 3791 + }, + { + "epoch": 0.6752136752136753, + "grad_norm": 0.5586925745010376, + "learning_rate": 0.0001863824144134488, + "loss": 1.1183, + "step": 3792 + }, + { + "epoch": 0.6753917378917379, + "grad_norm": 0.47740885615348816, + "learning_rate": 0.000186375361732327, + "loss": 1.1512, + "step": 3793 + }, + { + "epoch": 0.6755698005698005, + "grad_norm": 0.5867732167243958, + "learning_rate": 0.00018636830735885935, + "loss": 1.1903, + "step": 3794 + }, + { + "epoch": 0.6757478632478633, + "grad_norm": 0.5013887882232666, + "learning_rate": 0.0001863612512931842, + "loss": 0.8581, + "step": 3795 + }, + { + "epoch": 0.6759259259259259, + "grad_norm": 0.6026871204376221, + "learning_rate": 0.0001863541935354397, + "loss": 0.9581, + "step": 3796 + }, + { + "epoch": 0.6761039886039886, + "grad_norm": 0.5238468647003174, + "learning_rate": 0.00018634713408576415, + "loss": 1.0949, + "step": 3797 + }, + { + "epoch": 0.6762820512820513, + "grad_norm": 0.5128598213195801, + "learning_rate": 0.00018634007294429585, + "loss": 0.8992, + "step": 3798 + }, + { + "epoch": 0.676460113960114, + "grad_norm": 0.5092771053314209, + "learning_rate": 0.00018633301011117324, + "loss": 1.0793, + "step": 3799 + }, + { + "epoch": 0.6766381766381766, + "grad_norm": 0.592566728591919, + "learning_rate": 0.00018632594558653457, + "loss": 1.3242, + "step": 3800 + }, + { + "epoch": 0.6768162393162394, + "grad_norm": 0.4953067898750305, + "learning_rate": 0.0001863188793705184, + "loss": 0.9925, + "step": 3801 + }, + { + "epoch": 0.676994301994302, + "grad_norm": 0.4989747107028961, + "learning_rate": 0.00018631181146326305, + "loss": 1.0677, + "step": 3802 + }, + { + "epoch": 0.6771723646723646, + "grad_norm": 0.5375261902809143, + "learning_rate": 0.00018630474186490705, + "loss": 1.0556, + "step": 3803 + }, + { + "epoch": 0.6773504273504274, + "grad_norm": 0.6512624025344849, + "learning_rate": 0.00018629767057558894, + "loss": 1.2041, + "step": 3804 + }, + { + "epoch": 0.67752849002849, + "grad_norm": 0.5428260564804077, + "learning_rate": 0.00018629059759544723, + "loss": 0.9645, + "step": 3805 + }, + { + "epoch": 0.6777065527065527, + "grad_norm": 0.5598662495613098, + "learning_rate": 0.00018628352292462052, + "loss": 1.1683, + "step": 3806 + }, + { + "epoch": 0.6778846153846154, + "grad_norm": 0.49351340532302856, + "learning_rate": 0.0001862764465632474, + "loss": 1.1622, + "step": 3807 + }, + { + "epoch": 0.6780626780626781, + "grad_norm": 0.4796701669692993, + "learning_rate": 0.00018626936851146657, + "loss": 1.0017, + "step": 3808 + }, + { + "epoch": 0.6782407407407407, + "grad_norm": 0.444533109664917, + "learning_rate": 0.00018626228876941664, + "loss": 0.9145, + "step": 3809 + }, + { + "epoch": 0.6784188034188035, + "grad_norm": 0.5197392702102661, + "learning_rate": 0.00018625520733723635, + "loss": 1.283, + "step": 3810 + }, + { + "epoch": 0.6785968660968661, + "grad_norm": 0.48785829544067383, + "learning_rate": 0.00018624812421506447, + "loss": 1.1084, + "step": 3811 + }, + { + "epoch": 0.6787749287749287, + "grad_norm": 0.5083680152893066, + "learning_rate": 0.00018624103940303974, + "loss": 0.9071, + "step": 3812 + }, + { + "epoch": 0.6789529914529915, + "grad_norm": 0.553819477558136, + "learning_rate": 0.00018623395290130103, + "loss": 0.9986, + "step": 3813 + }, + { + "epoch": 0.6791310541310541, + "grad_norm": 0.5347508788108826, + "learning_rate": 0.00018622686470998713, + "loss": 1.0148, + "step": 3814 + }, + { + "epoch": 0.6793091168091168, + "grad_norm": 0.5080769062042236, + "learning_rate": 0.00018621977482923693, + "loss": 1.0169, + "step": 3815 + }, + { + "epoch": 0.6794871794871795, + "grad_norm": 0.5444077849388123, + "learning_rate": 0.00018621268325918938, + "loss": 1.172, + "step": 3816 + }, + { + "epoch": 0.6796652421652422, + "grad_norm": 0.521946132183075, + "learning_rate": 0.00018620558999998335, + "loss": 1.0247, + "step": 3817 + }, + { + "epoch": 0.6798433048433048, + "grad_norm": 0.5257413983345032, + "learning_rate": 0.00018619849505175786, + "loss": 1.1574, + "step": 3818 + }, + { + "epoch": 0.6800213675213675, + "grad_norm": 0.5473007559776306, + "learning_rate": 0.00018619139841465193, + "loss": 1.1254, + "step": 3819 + }, + { + "epoch": 0.6801994301994302, + "grad_norm": 0.5479872226715088, + "learning_rate": 0.00018618430008880463, + "loss": 1.0196, + "step": 3820 + }, + { + "epoch": 0.6803774928774928, + "grad_norm": 0.5918973088264465, + "learning_rate": 0.00018617720007435497, + "loss": 1.082, + "step": 3821 + }, + { + "epoch": 0.6805555555555556, + "grad_norm": 0.5411791801452637, + "learning_rate": 0.0001861700983714421, + "loss": 0.7723, + "step": 3822 + }, + { + "epoch": 0.6807336182336182, + "grad_norm": 0.5466326475143433, + "learning_rate": 0.00018616299498020516, + "loss": 1.0979, + "step": 3823 + }, + { + "epoch": 0.6809116809116809, + "grad_norm": 0.5405182838439941, + "learning_rate": 0.00018615588990078332, + "loss": 0.8891, + "step": 3824 + }, + { + "epoch": 0.6810897435897436, + "grad_norm": 0.5415780544281006, + "learning_rate": 0.00018614878313331579, + "loss": 1.0927, + "step": 3825 + }, + { + "epoch": 0.6812678062678063, + "grad_norm": 0.5284909605979919, + "learning_rate": 0.00018614167467794182, + "loss": 1.0684, + "step": 3826 + }, + { + "epoch": 0.6814458689458689, + "grad_norm": 0.4873995780944824, + "learning_rate": 0.00018613456453480062, + "loss": 1.1653, + "step": 3827 + }, + { + "epoch": 0.6816239316239316, + "grad_norm": 0.5506551265716553, + "learning_rate": 0.0001861274527040316, + "loss": 0.9876, + "step": 3828 + }, + { + "epoch": 0.6818019943019943, + "grad_norm": 0.5031297206878662, + "learning_rate": 0.0001861203391857741, + "loss": 1.067, + "step": 3829 + }, + { + "epoch": 0.6819800569800569, + "grad_norm": 0.622346043586731, + "learning_rate": 0.0001861132239801674, + "loss": 1.1514, + "step": 3830 + }, + { + "epoch": 0.6821581196581197, + "grad_norm": 0.47706183791160583, + "learning_rate": 0.000186106107087351, + "loss": 0.9857, + "step": 3831 + }, + { + "epoch": 0.6823361823361823, + "grad_norm": 0.5082845091819763, + "learning_rate": 0.00018609898850746424, + "loss": 1.123, + "step": 3832 + }, + { + "epoch": 0.6825142450142451, + "grad_norm": 0.5119805932044983, + "learning_rate": 0.00018609186824064671, + "loss": 1.1386, + "step": 3833 + }, + { + "epoch": 0.6826923076923077, + "grad_norm": 0.5247541069984436, + "learning_rate": 0.00018608474628703788, + "loss": 0.9433, + "step": 3834 + }, + { + "epoch": 0.6828703703703703, + "grad_norm": 0.4618282616138458, + "learning_rate": 0.00018607762264677722, + "loss": 0.8727, + "step": 3835 + }, + { + "epoch": 0.6830484330484331, + "grad_norm": 0.6014040112495422, + "learning_rate": 0.00018607049732000436, + "loss": 1.1823, + "step": 3836 + }, + { + "epoch": 0.6832264957264957, + "grad_norm": 0.6489043831825256, + "learning_rate": 0.00018606337030685892, + "loss": 1.1466, + "step": 3837 + }, + { + "epoch": 0.6834045584045584, + "grad_norm": 0.5527763366699219, + "learning_rate": 0.00018605624160748053, + "loss": 1.3015, + "step": 3838 + }, + { + "epoch": 0.6835826210826211, + "grad_norm": 0.5628284215927124, + "learning_rate": 0.0001860491112220088, + "loss": 1.1504, + "step": 3839 + }, + { + "epoch": 0.6837606837606838, + "grad_norm": 0.5414566993713379, + "learning_rate": 0.00018604197915058355, + "loss": 1.0155, + "step": 3840 + }, + { + "epoch": 0.6839387464387464, + "grad_norm": 0.5378929376602173, + "learning_rate": 0.00018603484539334443, + "loss": 0.8917, + "step": 3841 + }, + { + "epoch": 0.6841168091168092, + "grad_norm": 0.5953748822212219, + "learning_rate": 0.00018602770995043125, + "loss": 1.1971, + "step": 3842 + }, + { + "epoch": 0.6842948717948718, + "grad_norm": 0.511813759803772, + "learning_rate": 0.00018602057282198376, + "loss": 1.1345, + "step": 3843 + }, + { + "epoch": 0.6844729344729344, + "grad_norm": 0.5145484209060669, + "learning_rate": 0.00018601343400814185, + "loss": 1.0786, + "step": 3844 + }, + { + "epoch": 0.6846509971509972, + "grad_norm": 0.5199604034423828, + "learning_rate": 0.00018600629350904542, + "loss": 1.2063, + "step": 3845 + }, + { + "epoch": 0.6848290598290598, + "grad_norm": 0.5653825998306274, + "learning_rate": 0.0001859991513248343, + "loss": 1.0314, + "step": 3846 + }, + { + "epoch": 0.6850071225071225, + "grad_norm": 0.5660843849182129, + "learning_rate": 0.00018599200745564843, + "loss": 1.2754, + "step": 3847 + }, + { + "epoch": 0.6851851851851852, + "grad_norm": 0.5225719809532166, + "learning_rate": 0.00018598486190162788, + "loss": 1.0837, + "step": 3848 + }, + { + "epoch": 0.6853632478632479, + "grad_norm": 0.5011669397354126, + "learning_rate": 0.00018597771466291252, + "loss": 1.1, + "step": 3849 + }, + { + "epoch": 0.6855413105413105, + "grad_norm": 0.5923115015029907, + "learning_rate": 0.00018597056573964245, + "loss": 1.1875, + "step": 3850 + }, + { + "epoch": 0.6857193732193733, + "grad_norm": 0.5666482448577881, + "learning_rate": 0.00018596341513195776, + "loss": 1.1663, + "step": 3851 + }, + { + "epoch": 0.6858974358974359, + "grad_norm": 0.5396790504455566, + "learning_rate": 0.0001859562628399985, + "loss": 1.1179, + "step": 3852 + }, + { + "epoch": 0.6860754985754985, + "grad_norm": 0.5709532499313354, + "learning_rate": 0.00018594910886390485, + "loss": 1.0369, + "step": 3853 + }, + { + "epoch": 0.6862535612535613, + "grad_norm": 0.45524322986602783, + "learning_rate": 0.00018594195320381692, + "loss": 1.0171, + "step": 3854 + }, + { + "epoch": 0.6864316239316239, + "grad_norm": 0.6130724549293518, + "learning_rate": 0.00018593479585987498, + "loss": 1.1944, + "step": 3855 + }, + { + "epoch": 0.6866096866096866, + "grad_norm": 0.5079745054244995, + "learning_rate": 0.0001859276368322192, + "loss": 1.2567, + "step": 3856 + }, + { + "epoch": 0.6867877492877493, + "grad_norm": 0.49919846653938293, + "learning_rate": 0.00018592047612098992, + "loss": 0.9459, + "step": 3857 + }, + { + "epoch": 0.686965811965812, + "grad_norm": 0.5776857733726501, + "learning_rate": 0.00018591331372632734, + "loss": 1.2456, + "step": 3858 + }, + { + "epoch": 0.6871438746438746, + "grad_norm": 0.4740692377090454, + "learning_rate": 0.00018590614964837188, + "loss": 1.0401, + "step": 3859 + }, + { + "epoch": 0.6873219373219374, + "grad_norm": 0.5015742182731628, + "learning_rate": 0.00018589898388726389, + "loss": 1.2052, + "step": 3860 + }, + { + "epoch": 0.6875, + "grad_norm": 0.4819730818271637, + "learning_rate": 0.0001858918164431437, + "loss": 1.007, + "step": 3861 + }, + { + "epoch": 0.6876780626780626, + "grad_norm": 0.5510426163673401, + "learning_rate": 0.00018588464731615184, + "loss": 1.0123, + "step": 3862 + }, + { + "epoch": 0.6878561253561254, + "grad_norm": 0.4950829744338989, + "learning_rate": 0.00018587747650642867, + "loss": 1.033, + "step": 3863 + }, + { + "epoch": 0.688034188034188, + "grad_norm": 0.5278680920600891, + "learning_rate": 0.0001858703040141148, + "loss": 1.0912, + "step": 3864 + }, + { + "epoch": 0.6882122507122507, + "grad_norm": 0.6359158158302307, + "learning_rate": 0.00018586312983935068, + "loss": 1.2868, + "step": 3865 + }, + { + "epoch": 0.6883903133903134, + "grad_norm": 0.5098239183425903, + "learning_rate": 0.0001858559539822769, + "loss": 0.8364, + "step": 3866 + }, + { + "epoch": 0.6885683760683761, + "grad_norm": 0.5651038289070129, + "learning_rate": 0.000185848776443034, + "loss": 1.1983, + "step": 3867 + }, + { + "epoch": 0.6887464387464387, + "grad_norm": 0.5305678248405457, + "learning_rate": 0.00018584159722176272, + "loss": 1.32, + "step": 3868 + }, + { + "epoch": 0.6889245014245015, + "grad_norm": 0.5481845140457153, + "learning_rate": 0.00018583441631860368, + "loss": 1.013, + "step": 3869 + }, + { + "epoch": 0.6891025641025641, + "grad_norm": 0.5214795470237732, + "learning_rate": 0.00018582723373369753, + "loss": 1.172, + "step": 3870 + }, + { + "epoch": 0.6892806267806267, + "grad_norm": 0.6282780766487122, + "learning_rate": 0.00018582004946718502, + "loss": 1.7304, + "step": 3871 + }, + { + "epoch": 0.6894586894586895, + "grad_norm": 0.5266988277435303, + "learning_rate": 0.0001858128635192069, + "loss": 1.1418, + "step": 3872 + }, + { + "epoch": 0.6896367521367521, + "grad_norm": 0.4761001467704773, + "learning_rate": 0.000185805675889904, + "loss": 0.8585, + "step": 3873 + }, + { + "epoch": 0.6898148148148148, + "grad_norm": 0.528779923915863, + "learning_rate": 0.00018579848657941715, + "loss": 1.0036, + "step": 3874 + }, + { + "epoch": 0.6899928774928775, + "grad_norm": 0.5427684783935547, + "learning_rate": 0.00018579129558788716, + "loss": 0.9769, + "step": 3875 + }, + { + "epoch": 0.6901709401709402, + "grad_norm": 0.6229544281959534, + "learning_rate": 0.00018578410291545495, + "loss": 1.2848, + "step": 3876 + }, + { + "epoch": 0.6903490028490028, + "grad_norm": 0.6602693200111389, + "learning_rate": 0.00018577690856226147, + "loss": 1.2713, + "step": 3877 + }, + { + "epoch": 0.6905270655270656, + "grad_norm": 0.45884042978286743, + "learning_rate": 0.0001857697125284476, + "loss": 0.9143, + "step": 3878 + }, + { + "epoch": 0.6907051282051282, + "grad_norm": 0.4956444203853607, + "learning_rate": 0.00018576251481415443, + "loss": 0.9646, + "step": 3879 + }, + { + "epoch": 0.6908831908831908, + "grad_norm": 0.473561555147171, + "learning_rate": 0.00018575531541952292, + "loss": 0.843, + "step": 3880 + }, + { + "epoch": 0.6910612535612536, + "grad_norm": 0.4676312506198883, + "learning_rate": 0.00018574811434469415, + "loss": 0.9464, + "step": 3881 + }, + { + "epoch": 0.6912393162393162, + "grad_norm": 0.5452045202255249, + "learning_rate": 0.00018574091158980922, + "loss": 0.985, + "step": 3882 + }, + { + "epoch": 0.6914173789173789, + "grad_norm": 0.6274946331977844, + "learning_rate": 0.0001857337071550092, + "loss": 1.0357, + "step": 3883 + }, + { + "epoch": 0.6915954415954416, + "grad_norm": 0.5533788800239563, + "learning_rate": 0.00018572650104043531, + "loss": 1.2636, + "step": 3884 + }, + { + "epoch": 0.6917735042735043, + "grad_norm": 0.48312318325042725, + "learning_rate": 0.00018571929324622872, + "loss": 1.2402, + "step": 3885 + }, + { + "epoch": 0.6919515669515669, + "grad_norm": 0.6087453961372375, + "learning_rate": 0.00018571208377253062, + "loss": 1.2961, + "step": 3886 + }, + { + "epoch": 0.6921296296296297, + "grad_norm": 0.49156486988067627, + "learning_rate": 0.00018570487261948234, + "loss": 0.9585, + "step": 3887 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 0.5200015902519226, + "learning_rate": 0.0001856976597872251, + "loss": 0.9274, + "step": 3888 + }, + { + "epoch": 0.6924857549857549, + "grad_norm": 0.5185118913650513, + "learning_rate": 0.0001856904452759002, + "loss": 1.0015, + "step": 3889 + }, + { + "epoch": 0.6926638176638177, + "grad_norm": 0.5859049558639526, + "learning_rate": 0.00018568322908564904, + "loss": 1.0959, + "step": 3890 + }, + { + "epoch": 0.6928418803418803, + "grad_norm": 0.5882301926612854, + "learning_rate": 0.00018567601121661302, + "loss": 1.3214, + "step": 3891 + }, + { + "epoch": 0.6930199430199431, + "grad_norm": 0.6475503444671631, + "learning_rate": 0.0001856687916689335, + "loss": 1.3265, + "step": 3892 + }, + { + "epoch": 0.6931980056980057, + "grad_norm": 0.46175432205200195, + "learning_rate": 0.000185661570442752, + "loss": 0.8547, + "step": 3893 + }, + { + "epoch": 0.6933760683760684, + "grad_norm": 0.5362716913223267, + "learning_rate": 0.00018565434753820998, + "loss": 0.974, + "step": 3894 + }, + { + "epoch": 0.6935541310541311, + "grad_norm": 0.4317963719367981, + "learning_rate": 0.00018564712295544896, + "loss": 0.7653, + "step": 3895 + }, + { + "epoch": 0.6937321937321937, + "grad_norm": 0.5679717659950256, + "learning_rate": 0.00018563989669461047, + "loss": 1.0691, + "step": 3896 + }, + { + "epoch": 0.6939102564102564, + "grad_norm": 0.5058363676071167, + "learning_rate": 0.00018563266875583608, + "loss": 1.0665, + "step": 3897 + }, + { + "epoch": 0.6940883190883191, + "grad_norm": 0.5365496277809143, + "learning_rate": 0.00018562543913926746, + "loss": 0.9963, + "step": 3898 + }, + { + "epoch": 0.6942663817663818, + "grad_norm": 0.49945300817489624, + "learning_rate": 0.0001856182078450462, + "loss": 0.8668, + "step": 3899 + }, + { + "epoch": 0.6944444444444444, + "grad_norm": 0.5869430899620056, + "learning_rate": 0.00018561097487331405, + "loss": 1.1942, + "step": 3900 + }, + { + "epoch": 0.6946225071225072, + "grad_norm": 0.5188950300216675, + "learning_rate": 0.0001856037402242127, + "loss": 0.9493, + "step": 3901 + }, + { + "epoch": 0.6948005698005698, + "grad_norm": 0.510788083076477, + "learning_rate": 0.00018559650389788384, + "loss": 0.9989, + "step": 3902 + }, + { + "epoch": 0.6949786324786325, + "grad_norm": 0.5360601544380188, + "learning_rate": 0.0001855892658944693, + "loss": 1.2766, + "step": 3903 + }, + { + "epoch": 0.6951566951566952, + "grad_norm": 0.522502601146698, + "learning_rate": 0.00018558202621411093, + "loss": 0.8774, + "step": 3904 + }, + { + "epoch": 0.6953347578347578, + "grad_norm": 0.5330635905265808, + "learning_rate": 0.00018557478485695052, + "loss": 0.972, + "step": 3905 + }, + { + "epoch": 0.6955128205128205, + "grad_norm": 0.5387479066848755, + "learning_rate": 0.00018556754182312996, + "loss": 1.0574, + "step": 3906 + }, + { + "epoch": 0.6956908831908832, + "grad_norm": 0.5357984900474548, + "learning_rate": 0.00018556029711279116, + "loss": 1.396, + "step": 3907 + }, + { + "epoch": 0.6958689458689459, + "grad_norm": 0.5647178292274475, + "learning_rate": 0.00018555305072607612, + "loss": 1.3304, + "step": 3908 + }, + { + "epoch": 0.6960470085470085, + "grad_norm": 0.46460914611816406, + "learning_rate": 0.00018554580266312673, + "loss": 0.9574, + "step": 3909 + }, + { + "epoch": 0.6962250712250713, + "grad_norm": 0.6206206679344177, + "learning_rate": 0.00018553855292408503, + "loss": 1.1637, + "step": 3910 + }, + { + "epoch": 0.6964031339031339, + "grad_norm": 0.5899842977523804, + "learning_rate": 0.00018553130150909312, + "loss": 1.1067, + "step": 3911 + }, + { + "epoch": 0.6965811965811965, + "grad_norm": 0.47294262051582336, + "learning_rate": 0.000185524048418293, + "loss": 1.1516, + "step": 3912 + }, + { + "epoch": 0.6967592592592593, + "grad_norm": 0.5791197419166565, + "learning_rate": 0.00018551679365182684, + "loss": 1.0007, + "step": 3913 + }, + { + "epoch": 0.6969373219373219, + "grad_norm": 0.5678651332855225, + "learning_rate": 0.00018550953720983672, + "loss": 1.2698, + "step": 3914 + }, + { + "epoch": 0.6971153846153846, + "grad_norm": 0.6509683728218079, + "learning_rate": 0.0001855022790924649, + "loss": 1.0354, + "step": 3915 + }, + { + "epoch": 0.6972934472934473, + "grad_norm": 0.5176648497581482, + "learning_rate": 0.0001854950192998535, + "loss": 1.1243, + "step": 3916 + }, + { + "epoch": 0.69747150997151, + "grad_norm": 0.520631730556488, + "learning_rate": 0.00018548775783214477, + "loss": 1.1371, + "step": 3917 + }, + { + "epoch": 0.6976495726495726, + "grad_norm": 0.5408333539962769, + "learning_rate": 0.00018548049468948108, + "loss": 1.1185, + "step": 3918 + }, + { + "epoch": 0.6978276353276354, + "grad_norm": 0.5423790216445923, + "learning_rate": 0.00018547322987200461, + "loss": 1.1539, + "step": 3919 + }, + { + "epoch": 0.698005698005698, + "grad_norm": 0.5422113537788391, + "learning_rate": 0.0001854659633798578, + "loss": 1.171, + "step": 3920 + }, + { + "epoch": 0.6981837606837606, + "grad_norm": 0.5113416314125061, + "learning_rate": 0.00018545869521318292, + "loss": 1.0597, + "step": 3921 + }, + { + "epoch": 0.6983618233618234, + "grad_norm": 0.49901214241981506, + "learning_rate": 0.00018545142537212248, + "loss": 1.1043, + "step": 3922 + }, + { + "epoch": 0.698539886039886, + "grad_norm": 0.6606622338294983, + "learning_rate": 0.00018544415385681885, + "loss": 1.1797, + "step": 3923 + }, + { + "epoch": 0.6987179487179487, + "grad_norm": 0.4786234498023987, + "learning_rate": 0.00018543688066741454, + "loss": 0.9532, + "step": 3924 + }, + { + "epoch": 0.6988960113960114, + "grad_norm": 0.5900700688362122, + "learning_rate": 0.00018542960580405203, + "loss": 1.1171, + "step": 3925 + }, + { + "epoch": 0.6990740740740741, + "grad_norm": 0.53485506772995, + "learning_rate": 0.00018542232926687383, + "loss": 1.1535, + "step": 3926 + }, + { + "epoch": 0.6992521367521367, + "grad_norm": 0.5269177556037903, + "learning_rate": 0.00018541505105602255, + "loss": 1.0287, + "step": 3927 + }, + { + "epoch": 0.6994301994301995, + "grad_norm": 0.5185505151748657, + "learning_rate": 0.0001854077711716408, + "loss": 1.2526, + "step": 3928 + }, + { + "epoch": 0.6996082621082621, + "grad_norm": 0.5615512132644653, + "learning_rate": 0.00018540048961387115, + "loss": 1.0189, + "step": 3929 + }, + { + "epoch": 0.6997863247863247, + "grad_norm": 0.4492493271827698, + "learning_rate": 0.00018539320638285637, + "loss": 0.8917, + "step": 3930 + }, + { + "epoch": 0.6999643874643875, + "grad_norm": 0.5062302947044373, + "learning_rate": 0.00018538592147873906, + "loss": 1.053, + "step": 3931 + }, + { + "epoch": 0.7001424501424501, + "grad_norm": 0.5508798956871033, + "learning_rate": 0.000185378634901662, + "loss": 0.9638, + "step": 3932 + }, + { + "epoch": 0.7003205128205128, + "grad_norm": 0.463980108499527, + "learning_rate": 0.00018537134665176793, + "loss": 1.0945, + "step": 3933 + }, + { + "epoch": 0.7004985754985755, + "grad_norm": 0.5027088522911072, + "learning_rate": 0.0001853640567291997, + "loss": 1.1745, + "step": 3934 + }, + { + "epoch": 0.7006766381766382, + "grad_norm": 0.5006551146507263, + "learning_rate": 0.00018535676513410009, + "loss": 0.8521, + "step": 3935 + }, + { + "epoch": 0.7008547008547008, + "grad_norm": 0.5870724320411682, + "learning_rate": 0.000185349471866612, + "loss": 0.9197, + "step": 3936 + }, + { + "epoch": 0.7010327635327636, + "grad_norm": 0.5030696392059326, + "learning_rate": 0.00018534217692687825, + "loss": 1.1049, + "step": 3937 + }, + { + "epoch": 0.7012108262108262, + "grad_norm": 0.5212681889533997, + "learning_rate": 0.00018533488031504186, + "loss": 1.3397, + "step": 3938 + }, + { + "epoch": 0.7013888888888888, + "grad_norm": 0.5649709105491638, + "learning_rate": 0.0001853275820312458, + "loss": 1.1994, + "step": 3939 + }, + { + "epoch": 0.7015669515669516, + "grad_norm": 0.4892779290676117, + "learning_rate": 0.00018532028207563297, + "loss": 1.1511, + "step": 3940 + }, + { + "epoch": 0.7017450142450142, + "grad_norm": 0.4929407835006714, + "learning_rate": 0.00018531298044834643, + "loss": 1.0792, + "step": 3941 + }, + { + "epoch": 0.7019230769230769, + "grad_norm": 0.5645940899848938, + "learning_rate": 0.00018530567714952932, + "loss": 1.0937, + "step": 3942 + }, + { + "epoch": 0.7021011396011396, + "grad_norm": 0.5471178293228149, + "learning_rate": 0.00018529837217932466, + "loss": 1.193, + "step": 3943 + }, + { + "epoch": 0.7022792022792023, + "grad_norm": 0.576627790927887, + "learning_rate": 0.00018529106553787558, + "loss": 1.1032, + "step": 3944 + }, + { + "epoch": 0.7024572649572649, + "grad_norm": 0.5015735626220703, + "learning_rate": 0.00018528375722532526, + "loss": 1.066, + "step": 3945 + }, + { + "epoch": 0.7026353276353277, + "grad_norm": 0.5315404534339905, + "learning_rate": 0.00018527644724181683, + "loss": 1.2059, + "step": 3946 + }, + { + "epoch": 0.7028133903133903, + "grad_norm": 0.5516065955162048, + "learning_rate": 0.0001852691355874936, + "loss": 1.161, + "step": 3947 + }, + { + "epoch": 0.7029914529914529, + "grad_norm": 0.5026212930679321, + "learning_rate": 0.0001852618222624988, + "loss": 1.2616, + "step": 3948 + }, + { + "epoch": 0.7031695156695157, + "grad_norm": 0.49874603748321533, + "learning_rate": 0.0001852545072669757, + "loss": 0.805, + "step": 3949 + }, + { + "epoch": 0.7033475783475783, + "grad_norm": 0.47698748111724854, + "learning_rate": 0.00018524719060106763, + "loss": 1.2321, + "step": 3950 + }, + { + "epoch": 0.7035256410256411, + "grad_norm": 0.5201322436332703, + "learning_rate": 0.00018523987226491792, + "loss": 1.1577, + "step": 3951 + }, + { + "epoch": 0.7037037037037037, + "grad_norm": 0.5506543517112732, + "learning_rate": 0.00018523255225867002, + "loss": 1.2289, + "step": 3952 + }, + { + "epoch": 0.7038817663817664, + "grad_norm": 0.5691256523132324, + "learning_rate": 0.0001852252305824673, + "loss": 1.1945, + "step": 3953 + }, + { + "epoch": 0.7040598290598291, + "grad_norm": 0.5324838757514954, + "learning_rate": 0.00018521790723645322, + "loss": 1.1037, + "step": 3954 + }, + { + "epoch": 0.7042378917378918, + "grad_norm": 0.5238786339759827, + "learning_rate": 0.00018521058222077127, + "loss": 1.2075, + "step": 3955 + }, + { + "epoch": 0.7044159544159544, + "grad_norm": 0.4936453402042389, + "learning_rate": 0.00018520325553556498, + "loss": 1.0537, + "step": 3956 + }, + { + "epoch": 0.7045940170940171, + "grad_norm": 0.6198282837867737, + "learning_rate": 0.00018519592718097791, + "loss": 1.0728, + "step": 3957 + }, + { + "epoch": 0.7047720797720798, + "grad_norm": 0.44729140400886536, + "learning_rate": 0.0001851885971571536, + "loss": 0.8432, + "step": 3958 + }, + { + "epoch": 0.7049501424501424, + "grad_norm": 0.5884211659431458, + "learning_rate": 0.00018518126546423572, + "loss": 0.9515, + "step": 3959 + }, + { + "epoch": 0.7051282051282052, + "grad_norm": 0.5293807983398438, + "learning_rate": 0.00018517393210236788, + "loss": 1.1178, + "step": 3960 + }, + { + "epoch": 0.7053062678062678, + "grad_norm": 0.6036825180053711, + "learning_rate": 0.00018516659707169374, + "loss": 1.0408, + "step": 3961 + }, + { + "epoch": 0.7054843304843305, + "grad_norm": 0.5157122015953064, + "learning_rate": 0.0001851592603723571, + "loss": 1.2136, + "step": 3962 + }, + { + "epoch": 0.7056623931623932, + "grad_norm": 0.5354781150817871, + "learning_rate": 0.00018515192200450163, + "loss": 0.7165, + "step": 3963 + }, + { + "epoch": 0.7058404558404558, + "grad_norm": 0.6073734760284424, + "learning_rate": 0.00018514458196827111, + "loss": 1.3079, + "step": 3964 + }, + { + "epoch": 0.7060185185185185, + "grad_norm": 0.4324839413166046, + "learning_rate": 0.0001851372402638094, + "loss": 0.7903, + "step": 3965 + }, + { + "epoch": 0.7061965811965812, + "grad_norm": 0.6530333161354065, + "learning_rate": 0.00018512989689126034, + "loss": 1.3179, + "step": 3966 + }, + { + "epoch": 0.7063746438746439, + "grad_norm": 0.5500404238700867, + "learning_rate": 0.00018512255185076782, + "loss": 1.0624, + "step": 3967 + }, + { + "epoch": 0.7065527065527065, + "grad_norm": 0.6277863383293152, + "learning_rate": 0.00018511520514247567, + "loss": 1.1056, + "step": 3968 + }, + { + "epoch": 0.7067307692307693, + "grad_norm": 0.580544650554657, + "learning_rate": 0.0001851078567665279, + "loss": 0.9849, + "step": 3969 + }, + { + "epoch": 0.7069088319088319, + "grad_norm": 0.4880999028682709, + "learning_rate": 0.00018510050672306848, + "loss": 1.0185, + "step": 3970 + }, + { + "epoch": 0.7070868945868946, + "grad_norm": 0.4919959306716919, + "learning_rate": 0.0001850931550122414, + "loss": 1.0334, + "step": 3971 + }, + { + "epoch": 0.7072649572649573, + "grad_norm": 0.6001213192939758, + "learning_rate": 0.0001850858016341907, + "loss": 1.0729, + "step": 3972 + }, + { + "epoch": 0.70744301994302, + "grad_norm": 0.538690447807312, + "learning_rate": 0.00018507844658906052, + "loss": 1.0733, + "step": 3973 + }, + { + "epoch": 0.7076210826210826, + "grad_norm": 0.5427643656730652, + "learning_rate": 0.00018507108987699487, + "loss": 1.1207, + "step": 3974 + }, + { + "epoch": 0.7077991452991453, + "grad_norm": 0.43014347553253174, + "learning_rate": 0.00018506373149813795, + "loss": 0.7958, + "step": 3975 + }, + { + "epoch": 0.707977207977208, + "grad_norm": 0.56591796875, + "learning_rate": 0.00018505637145263394, + "loss": 1.2199, + "step": 3976 + }, + { + "epoch": 0.7081552706552706, + "grad_norm": 0.59147047996521, + "learning_rate": 0.000185049009740627, + "loss": 1.2354, + "step": 3977 + }, + { + "epoch": 0.7083333333333334, + "grad_norm": 0.5078346133232117, + "learning_rate": 0.00018504164636226137, + "loss": 0.976, + "step": 3978 + }, + { + "epoch": 0.708511396011396, + "grad_norm": 0.533302366733551, + "learning_rate": 0.00018503428131768135, + "loss": 0.9653, + "step": 3979 + }, + { + "epoch": 0.7086894586894587, + "grad_norm": 0.4985341727733612, + "learning_rate": 0.00018502691460703122, + "loss": 1.1485, + "step": 3980 + }, + { + "epoch": 0.7088675213675214, + "grad_norm": 0.5143141150474548, + "learning_rate": 0.00018501954623045532, + "loss": 1.148, + "step": 3981 + }, + { + "epoch": 0.709045584045584, + "grad_norm": 0.507189154624939, + "learning_rate": 0.00018501217618809804, + "loss": 0.9306, + "step": 3982 + }, + { + "epoch": 0.7092236467236467, + "grad_norm": 0.5246604084968567, + "learning_rate": 0.00018500480448010377, + "loss": 0.9116, + "step": 3983 + }, + { + "epoch": 0.7094017094017094, + "grad_norm": 0.5321049094200134, + "learning_rate": 0.00018499743110661693, + "loss": 0.9607, + "step": 3984 + }, + { + "epoch": 0.7095797720797721, + "grad_norm": 0.62645423412323, + "learning_rate": 0.000184990056067782, + "loss": 1.5834, + "step": 3985 + }, + { + "epoch": 0.7097578347578347, + "grad_norm": 0.486557275056839, + "learning_rate": 0.0001849826793637435, + "loss": 1.0598, + "step": 3986 + }, + { + "epoch": 0.7099358974358975, + "grad_norm": 0.5122783184051514, + "learning_rate": 0.0001849753009946459, + "loss": 1.2213, + "step": 3987 + }, + { + "epoch": 0.7101139601139601, + "grad_norm": 0.4864068627357483, + "learning_rate": 0.0001849679209606338, + "loss": 1.2708, + "step": 3988 + }, + { + "epoch": 0.7102920227920227, + "grad_norm": 0.5860990881919861, + "learning_rate": 0.00018496053926185183, + "loss": 1.2421, + "step": 3989 + }, + { + "epoch": 0.7104700854700855, + "grad_norm": 0.471194326877594, + "learning_rate": 0.00018495315589844453, + "loss": 0.879, + "step": 3990 + }, + { + "epoch": 0.7106481481481481, + "grad_norm": 0.5626323819160461, + "learning_rate": 0.00018494577087055662, + "loss": 1.1297, + "step": 3991 + }, + { + "epoch": 0.7108262108262108, + "grad_norm": 0.4706762135028839, + "learning_rate": 0.0001849383841783328, + "loss": 1.0444, + "step": 3992 + }, + { + "epoch": 0.7110042735042735, + "grad_norm": 0.5776444673538208, + "learning_rate": 0.00018493099582191783, + "loss": 1.1773, + "step": 3993 + }, + { + "epoch": 0.7111823361823362, + "grad_norm": 0.5493253469467163, + "learning_rate": 0.00018492360580145637, + "loss": 1.0354, + "step": 3994 + }, + { + "epoch": 0.7113603988603988, + "grad_norm": 0.5328514575958252, + "learning_rate": 0.0001849162141170933, + "loss": 0.9251, + "step": 3995 + }, + { + "epoch": 0.7115384615384616, + "grad_norm": 0.5814893841743469, + "learning_rate": 0.0001849088207689734, + "loss": 1.1066, + "step": 3996 + }, + { + "epoch": 0.7117165242165242, + "grad_norm": 0.5476071834564209, + "learning_rate": 0.00018490142575724154, + "loss": 1.1613, + "step": 3997 + }, + { + "epoch": 0.7118945868945868, + "grad_norm": 0.5216463208198547, + "learning_rate": 0.00018489402908204258, + "loss": 1.2574, + "step": 3998 + }, + { + "epoch": 0.7120726495726496, + "grad_norm": 0.5110020637512207, + "learning_rate": 0.00018488663074352153, + "loss": 1.0663, + "step": 3999 + }, + { + "epoch": 0.7122507122507122, + "grad_norm": 0.448090523481369, + "learning_rate": 0.00018487923074182326, + "loss": 0.6687, + "step": 4000 + }, + { + "epoch": 0.7124287749287749, + "grad_norm": 0.4980565011501312, + "learning_rate": 0.00018487182907709279, + "loss": 1.2365, + "step": 4001 + }, + { + "epoch": 0.7126068376068376, + "grad_norm": 0.485831081867218, + "learning_rate": 0.00018486442574947511, + "loss": 1.0941, + "step": 4002 + }, + { + "epoch": 0.7127849002849003, + "grad_norm": 0.4955040216445923, + "learning_rate": 0.00018485702075911534, + "loss": 1.248, + "step": 4003 + }, + { + "epoch": 0.7129629629629629, + "grad_norm": 0.5168375968933105, + "learning_rate": 0.00018484961410615845, + "loss": 1.1118, + "step": 4004 + }, + { + "epoch": 0.7131410256410257, + "grad_norm": 0.5255687832832336, + "learning_rate": 0.00018484220579074968, + "loss": 1.0558, + "step": 4005 + }, + { + "epoch": 0.7133190883190883, + "grad_norm": 0.5502219796180725, + "learning_rate": 0.00018483479581303416, + "loss": 1.1604, + "step": 4006 + }, + { + "epoch": 0.7134971509971509, + "grad_norm": 0.5155881643295288, + "learning_rate": 0.000184827384173157, + "loss": 0.8246, + "step": 4007 + }, + { + "epoch": 0.7136752136752137, + "grad_norm": 0.5321542024612427, + "learning_rate": 0.0001848199708712635, + "loss": 1.2058, + "step": 4008 + }, + { + "epoch": 0.7138532763532763, + "grad_norm": 0.4929848313331604, + "learning_rate": 0.00018481255590749884, + "loss": 1.4023, + "step": 4009 + }, + { + "epoch": 0.7140313390313391, + "grad_norm": 0.5070937871932983, + "learning_rate": 0.00018480513928200836, + "loss": 1.0561, + "step": 4010 + }, + { + "epoch": 0.7142094017094017, + "grad_norm": 0.5750083327293396, + "learning_rate": 0.00018479772099493728, + "loss": 1.0276, + "step": 4011 + }, + { + "epoch": 0.7143874643874644, + "grad_norm": 0.5265933275222778, + "learning_rate": 0.00018479030104643108, + "loss": 1.0295, + "step": 4012 + }, + { + "epoch": 0.7145655270655271, + "grad_norm": 0.526830792427063, + "learning_rate": 0.00018478287943663504, + "loss": 1.0157, + "step": 4013 + }, + { + "epoch": 0.7147435897435898, + "grad_norm": 0.5344091653823853, + "learning_rate": 0.00018477545616569458, + "loss": 1.1997, + "step": 4014 + }, + { + "epoch": 0.7149216524216524, + "grad_norm": 0.4935445189476013, + "learning_rate": 0.0001847680312337552, + "loss": 1.1858, + "step": 4015 + }, + { + "epoch": 0.7150997150997151, + "grad_norm": 0.5291212797164917, + "learning_rate": 0.0001847606046409623, + "loss": 0.926, + "step": 4016 + }, + { + "epoch": 0.7152777777777778, + "grad_norm": 0.559050977230072, + "learning_rate": 0.00018475317638746142, + "loss": 1.0947, + "step": 4017 + }, + { + "epoch": 0.7154558404558404, + "grad_norm": 0.4566570222377777, + "learning_rate": 0.00018474574647339814, + "loss": 1.0334, + "step": 4018 + }, + { + "epoch": 0.7156339031339032, + "grad_norm": 0.5156155824661255, + "learning_rate": 0.000184738314898918, + "loss": 1.0076, + "step": 4019 + }, + { + "epoch": 0.7158119658119658, + "grad_norm": 0.5008716583251953, + "learning_rate": 0.00018473088166416662, + "loss": 1.0378, + "step": 4020 + }, + { + "epoch": 0.7159900284900285, + "grad_norm": 0.49556368589401245, + "learning_rate": 0.0001847234467692896, + "loss": 1.15, + "step": 4021 + }, + { + "epoch": 0.7161680911680912, + "grad_norm": 0.5464680790901184, + "learning_rate": 0.00018471601021443265, + "loss": 1.2975, + "step": 4022 + }, + { + "epoch": 0.7163461538461539, + "grad_norm": 0.6291980147361755, + "learning_rate": 0.00018470857199974144, + "loss": 1.05, + "step": 4023 + }, + { + "epoch": 0.7165242165242165, + "grad_norm": 0.5566631555557251, + "learning_rate": 0.00018470113212536176, + "loss": 1.1296, + "step": 4024 + }, + { + "epoch": 0.7167022792022792, + "grad_norm": 0.5569562911987305, + "learning_rate": 0.00018469369059143933, + "loss": 1.2484, + "step": 4025 + }, + { + "epoch": 0.7168803418803419, + "grad_norm": 0.5804716944694519, + "learning_rate": 0.00018468624739812, + "loss": 1.0547, + "step": 4026 + }, + { + "epoch": 0.7170584045584045, + "grad_norm": 0.6316802501678467, + "learning_rate": 0.00018467880254554952, + "loss": 1.1188, + "step": 4027 + }, + { + "epoch": 0.7172364672364673, + "grad_norm": 0.6131419539451599, + "learning_rate": 0.00018467135603387385, + "loss": 1.1662, + "step": 4028 + }, + { + "epoch": 0.7174145299145299, + "grad_norm": 0.4703124761581421, + "learning_rate": 0.00018466390786323883, + "loss": 1.038, + "step": 4029 + }, + { + "epoch": 0.7175925925925926, + "grad_norm": 0.5718469023704529, + "learning_rate": 0.0001846564580337904, + "loss": 1.0786, + "step": 4030 + }, + { + "epoch": 0.7177706552706553, + "grad_norm": 0.5227612853050232, + "learning_rate": 0.00018464900654567457, + "loss": 1.0561, + "step": 4031 + }, + { + "epoch": 0.717948717948718, + "grad_norm": 0.5800358057022095, + "learning_rate": 0.00018464155339903727, + "loss": 1.0944, + "step": 4032 + }, + { + "epoch": 0.7181267806267806, + "grad_norm": 0.5562314987182617, + "learning_rate": 0.00018463409859402455, + "loss": 0.8573, + "step": 4033 + }, + { + "epoch": 0.7183048433048433, + "grad_norm": 0.6420153379440308, + "learning_rate": 0.0001846266421307825, + "loss": 1.088, + "step": 4034 + }, + { + "epoch": 0.718482905982906, + "grad_norm": 0.4745902717113495, + "learning_rate": 0.00018461918400945718, + "loss": 1.1679, + "step": 4035 + }, + { + "epoch": 0.7186609686609686, + "grad_norm": 0.5070300102233887, + "learning_rate": 0.00018461172423019475, + "loss": 1.1984, + "step": 4036 + }, + { + "epoch": 0.7188390313390314, + "grad_norm": 0.5339375138282776, + "learning_rate": 0.00018460426279314133, + "loss": 1.3038, + "step": 4037 + }, + { + "epoch": 0.719017094017094, + "grad_norm": 0.5947147607803345, + "learning_rate": 0.00018459679969844313, + "loss": 1.0103, + "step": 4038 + }, + { + "epoch": 0.7191951566951567, + "grad_norm": 0.5493791699409485, + "learning_rate": 0.00018458933494624642, + "loss": 1.1001, + "step": 4039 + }, + { + "epoch": 0.7193732193732194, + "grad_norm": 0.5700310468673706, + "learning_rate": 0.00018458186853669736, + "loss": 0.9006, + "step": 4040 + }, + { + "epoch": 0.719551282051282, + "grad_norm": 0.60371994972229, + "learning_rate": 0.0001845744004699423, + "loss": 1.3001, + "step": 4041 + }, + { + "epoch": 0.7197293447293447, + "grad_norm": 0.5469261407852173, + "learning_rate": 0.00018456693074612757, + "loss": 1.1745, + "step": 4042 + }, + { + "epoch": 0.7199074074074074, + "grad_norm": 0.5179165601730347, + "learning_rate": 0.00018455945936539947, + "loss": 0.9883, + "step": 4043 + }, + { + "epoch": 0.7200854700854701, + "grad_norm": 0.5396696329116821, + "learning_rate": 0.00018455198632790447, + "loss": 1.1277, + "step": 4044 + }, + { + "epoch": 0.7202635327635327, + "grad_norm": 0.4559909403324127, + "learning_rate": 0.00018454451163378888, + "loss": 0.9644, + "step": 4045 + }, + { + "epoch": 0.7204415954415955, + "grad_norm": 0.49863892793655396, + "learning_rate": 0.00018453703528319927, + "loss": 1.1276, + "step": 4046 + }, + { + "epoch": 0.7206196581196581, + "grad_norm": 0.4790710508823395, + "learning_rate": 0.000184529557276282, + "loss": 0.9443, + "step": 4047 + }, + { + "epoch": 0.7207977207977208, + "grad_norm": 0.541999876499176, + "learning_rate": 0.0001845220776131837, + "loss": 1.0681, + "step": 4048 + }, + { + "epoch": 0.7209757834757835, + "grad_norm": 0.5119109153747559, + "learning_rate": 0.00018451459629405088, + "loss": 1.2078, + "step": 4049 + }, + { + "epoch": 0.7211538461538461, + "grad_norm": 0.6141307353973389, + "learning_rate": 0.00018450711331903006, + "loss": 1.1071, + "step": 4050 + }, + { + "epoch": 0.7213319088319088, + "grad_norm": 0.48679864406585693, + "learning_rate": 0.00018449962868826795, + "loss": 0.9713, + "step": 4051 + }, + { + "epoch": 0.7215099715099715, + "grad_norm": 0.5548661947250366, + "learning_rate": 0.0001844921424019111, + "loss": 1.2099, + "step": 4052 + }, + { + "epoch": 0.7216880341880342, + "grad_norm": 0.5000107884407043, + "learning_rate": 0.00018448465446010626, + "loss": 1.0184, + "step": 4053 + }, + { + "epoch": 0.7218660968660968, + "grad_norm": 0.6131454110145569, + "learning_rate": 0.00018447716486300013, + "loss": 1.2581, + "step": 4054 + }, + { + "epoch": 0.7220441595441596, + "grad_norm": 0.5145987868309021, + "learning_rate": 0.0001844696736107394, + "loss": 1.1646, + "step": 4055 + }, + { + "epoch": 0.7222222222222222, + "grad_norm": 0.4361337125301361, + "learning_rate": 0.00018446218070347094, + "loss": 0.8239, + "step": 4056 + }, + { + "epoch": 0.7224002849002849, + "grad_norm": 0.5549173355102539, + "learning_rate": 0.00018445468614134146, + "loss": 1.1935, + "step": 4057 + }, + { + "epoch": 0.7225783475783476, + "grad_norm": 0.5569297671318054, + "learning_rate": 0.00018444718992449789, + "loss": 1.0137, + "step": 4058 + }, + { + "epoch": 0.7227564102564102, + "grad_norm": 0.44866305589675903, + "learning_rate": 0.00018443969205308704, + "loss": 0.987, + "step": 4059 + }, + { + "epoch": 0.7229344729344729, + "grad_norm": 0.5142943263053894, + "learning_rate": 0.0001844321925272558, + "loss": 1.0837, + "step": 4060 + }, + { + "epoch": 0.7231125356125356, + "grad_norm": 0.4922119379043579, + "learning_rate": 0.0001844246913471512, + "loss": 0.8477, + "step": 4061 + }, + { + "epoch": 0.7232905982905983, + "grad_norm": 0.5245375633239746, + "learning_rate": 0.0001844171885129201, + "loss": 0.9985, + "step": 4062 + }, + { + "epoch": 0.7234686609686609, + "grad_norm": 0.45562678575515747, + "learning_rate": 0.00018440968402470956, + "loss": 0.8678, + "step": 4063 + }, + { + "epoch": 0.7236467236467237, + "grad_norm": 0.5388376712799072, + "learning_rate": 0.0001844021778826666, + "loss": 1.0586, + "step": 4064 + }, + { + "epoch": 0.7238247863247863, + "grad_norm": 0.48945263028144836, + "learning_rate": 0.00018439467008693833, + "loss": 1.0547, + "step": 4065 + }, + { + "epoch": 0.7240028490028491, + "grad_norm": 0.5202330350875854, + "learning_rate": 0.00018438716063767178, + "loss": 1.3142, + "step": 4066 + }, + { + "epoch": 0.7241809116809117, + "grad_norm": 0.5432567000389099, + "learning_rate": 0.00018437964953501413, + "loss": 1.0192, + "step": 4067 + }, + { + "epoch": 0.7243589743589743, + "grad_norm": 0.5220325589179993, + "learning_rate": 0.00018437213677911253, + "loss": 1.0904, + "step": 4068 + }, + { + "epoch": 0.7245370370370371, + "grad_norm": 0.45711690187454224, + "learning_rate": 0.00018436462237011417, + "loss": 1.0417, + "step": 4069 + }, + { + "epoch": 0.7247150997150997, + "grad_norm": 0.560778021812439, + "learning_rate": 0.0001843571063081663, + "loss": 1.2316, + "step": 4070 + }, + { + "epoch": 0.7248931623931624, + "grad_norm": 0.591533362865448, + "learning_rate": 0.0001843495885934162, + "loss": 1.0294, + "step": 4071 + }, + { + "epoch": 0.7250712250712251, + "grad_norm": 0.5550443530082703, + "learning_rate": 0.00018434206922601106, + "loss": 1.0162, + "step": 4072 + }, + { + "epoch": 0.7252492877492878, + "grad_norm": 0.5744053721427917, + "learning_rate": 0.00018433454820609833, + "loss": 1.2774, + "step": 4073 + }, + { + "epoch": 0.7254273504273504, + "grad_norm": 0.6210703253746033, + "learning_rate": 0.0001843270255338253, + "loss": 1.2526, + "step": 4074 + }, + { + "epoch": 0.7256054131054132, + "grad_norm": 0.49684277176856995, + "learning_rate": 0.0001843195012093394, + "loss": 1.0786, + "step": 4075 + }, + { + "epoch": 0.7257834757834758, + "grad_norm": 0.5851606130599976, + "learning_rate": 0.00018431197523278802, + "loss": 1.14, + "step": 4076 + }, + { + "epoch": 0.7259615384615384, + "grad_norm": 0.5494425296783447, + "learning_rate": 0.00018430444760431862, + "loss": 1.211, + "step": 4077 + }, + { + "epoch": 0.7261396011396012, + "grad_norm": 0.5247658491134644, + "learning_rate": 0.00018429691832407867, + "loss": 0.8031, + "step": 4078 + }, + { + "epoch": 0.7263176638176638, + "grad_norm": 0.5012249946594238, + "learning_rate": 0.00018428938739221574, + "loss": 1.1258, + "step": 4079 + }, + { + "epoch": 0.7264957264957265, + "grad_norm": 0.5226427912712097, + "learning_rate": 0.0001842818548088774, + "loss": 1.0029, + "step": 4080 + }, + { + "epoch": 0.7266737891737892, + "grad_norm": 0.45008543133735657, + "learning_rate": 0.00018427432057421114, + "loss": 1.0681, + "step": 4081 + }, + { + "epoch": 0.7268518518518519, + "grad_norm": 0.5127285122871399, + "learning_rate": 0.00018426678468836467, + "loss": 1.1069, + "step": 4082 + }, + { + "epoch": 0.7270299145299145, + "grad_norm": 0.5406150221824646, + "learning_rate": 0.0001842592471514856, + "loss": 1.052, + "step": 4083 + }, + { + "epoch": 0.7272079772079773, + "grad_norm": 0.5001157522201538, + "learning_rate": 0.0001842517079637216, + "loss": 0.9157, + "step": 4084 + }, + { + "epoch": 0.7273860398860399, + "grad_norm": 0.6169779300689697, + "learning_rate": 0.00018424416712522042, + "loss": 1.3133, + "step": 4085 + }, + { + "epoch": 0.7275641025641025, + "grad_norm": 0.4891316890716553, + "learning_rate": 0.00018423662463612974, + "loss": 0.9505, + "step": 4086 + }, + { + "epoch": 0.7277421652421653, + "grad_norm": 0.5883708596229553, + "learning_rate": 0.00018422908049659743, + "loss": 1.2797, + "step": 4087 + }, + { + "epoch": 0.7279202279202279, + "grad_norm": 0.6679072976112366, + "learning_rate": 0.00018422153470677125, + "loss": 1.1096, + "step": 4088 + }, + { + "epoch": 0.7280982905982906, + "grad_norm": 0.5178479552268982, + "learning_rate": 0.00018421398726679904, + "loss": 1.0299, + "step": 4089 + }, + { + "epoch": 0.7282763532763533, + "grad_norm": 0.6343900561332703, + "learning_rate": 0.0001842064381768287, + "loss": 1.2983, + "step": 4090 + }, + { + "epoch": 0.728454415954416, + "grad_norm": 0.43816515803337097, + "learning_rate": 0.0001841988874370081, + "loss": 0.9452, + "step": 4091 + }, + { + "epoch": 0.7286324786324786, + "grad_norm": 0.579790472984314, + "learning_rate": 0.00018419133504748528, + "loss": 1.1037, + "step": 4092 + }, + { + "epoch": 0.7288105413105413, + "grad_norm": 0.571374773979187, + "learning_rate": 0.00018418378100840807, + "loss": 1.1655, + "step": 4093 + }, + { + "epoch": 0.728988603988604, + "grad_norm": 0.5163514018058777, + "learning_rate": 0.0001841762253199246, + "loss": 1.1579, + "step": 4094 + }, + { + "epoch": 0.7291666666666666, + "grad_norm": 0.6553022265434265, + "learning_rate": 0.0001841686679821828, + "loss": 0.9664, + "step": 4095 + }, + { + "epoch": 0.7293447293447294, + "grad_norm": 0.5072969198226929, + "learning_rate": 0.00018416110899533084, + "loss": 0.9416, + "step": 4096 + }, + { + "epoch": 0.729522792022792, + "grad_norm": 0.5103251338005066, + "learning_rate": 0.00018415354835951675, + "loss": 1.0715, + "step": 4097 + }, + { + "epoch": 0.7297008547008547, + "grad_norm": 0.49752289056777954, + "learning_rate": 0.00018414598607488874, + "loss": 1.1848, + "step": 4098 + }, + { + "epoch": 0.7298789173789174, + "grad_norm": 0.5361882448196411, + "learning_rate": 0.00018413842214159488, + "loss": 1.1035, + "step": 4099 + }, + { + "epoch": 0.73005698005698, + "grad_norm": 0.5167670249938965, + "learning_rate": 0.00018413085655978343, + "loss": 1.0015, + "step": 4100 + }, + { + "epoch": 0.7302350427350427, + "grad_norm": 0.5930629372596741, + "learning_rate": 0.00018412328932960263, + "loss": 0.9766, + "step": 4101 + }, + { + "epoch": 0.7304131054131054, + "grad_norm": 0.5234778523445129, + "learning_rate": 0.00018411572045120073, + "loss": 1.0317, + "step": 4102 + }, + { + "epoch": 0.7305911680911681, + "grad_norm": 0.5361374020576477, + "learning_rate": 0.000184108149924726, + "loss": 1.1228, + "step": 4103 + }, + { + "epoch": 0.7307692307692307, + "grad_norm": 0.5845770239830017, + "learning_rate": 0.0001841005777503268, + "loss": 0.9541, + "step": 4104 + }, + { + "epoch": 0.7309472934472935, + "grad_norm": 0.49320483207702637, + "learning_rate": 0.0001840930039281515, + "loss": 0.9445, + "step": 4105 + }, + { + "epoch": 0.7311253561253561, + "grad_norm": 0.5391250252723694, + "learning_rate": 0.00018408542845834845, + "loss": 1.1983, + "step": 4106 + }, + { + "epoch": 0.7313034188034188, + "grad_norm": 0.4890393316745758, + "learning_rate": 0.00018407785134106613, + "loss": 0.8353, + "step": 4107 + }, + { + "epoch": 0.7314814814814815, + "grad_norm": 0.5839747190475464, + "learning_rate": 0.00018407027257645296, + "loss": 1.4074, + "step": 4108 + }, + { + "epoch": 0.7316595441595442, + "grad_norm": 0.5957708358764648, + "learning_rate": 0.0001840626921646574, + "loss": 1.1032, + "step": 4109 + }, + { + "epoch": 0.7318376068376068, + "grad_norm": 0.5029017925262451, + "learning_rate": 0.00018405511010582805, + "loss": 1.095, + "step": 4110 + }, + { + "epoch": 0.7320156695156695, + "grad_norm": 0.6054347157478333, + "learning_rate": 0.00018404752640011345, + "loss": 1.0366, + "step": 4111 + }, + { + "epoch": 0.7321937321937322, + "grad_norm": 0.5476830005645752, + "learning_rate": 0.00018403994104766212, + "loss": 1.0976, + "step": 4112 + }, + { + "epoch": 0.7323717948717948, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00018403235404862277, + "loss": 1.0809, + "step": 4113 + }, + { + "epoch": 0.7325498575498576, + "grad_norm": 0.5119251012802124, + "learning_rate": 0.00018402476540314394, + "loss": 1.0176, + "step": 4114 + }, + { + "epoch": 0.7327279202279202, + "grad_norm": 0.5825830698013306, + "learning_rate": 0.00018401717511137445, + "loss": 1.2357, + "step": 4115 + }, + { + "epoch": 0.7329059829059829, + "grad_norm": 0.5702941417694092, + "learning_rate": 0.0001840095831734629, + "loss": 1.1549, + "step": 4116 + }, + { + "epoch": 0.7330840455840456, + "grad_norm": 0.5660699605941772, + "learning_rate": 0.00018400198958955807, + "loss": 1.1778, + "step": 4117 + }, + { + "epoch": 0.7332621082621082, + "grad_norm": 0.5241161584854126, + "learning_rate": 0.0001839943943598088, + "loss": 0.8587, + "step": 4118 + }, + { + "epoch": 0.7334401709401709, + "grad_norm": 0.581194281578064, + "learning_rate": 0.0001839867974843638, + "loss": 1.2169, + "step": 4119 + }, + { + "epoch": 0.7336182336182336, + "grad_norm": 0.4342379570007324, + "learning_rate": 0.00018397919896337198, + "loss": 0.9182, + "step": 4120 + }, + { + "epoch": 0.7337962962962963, + "grad_norm": 0.5708567500114441, + "learning_rate": 0.00018397159879698224, + "loss": 1.1781, + "step": 4121 + }, + { + "epoch": 0.7339743589743589, + "grad_norm": 0.5827265977859497, + "learning_rate": 0.00018396399698534344, + "loss": 1.2905, + "step": 4122 + }, + { + "epoch": 0.7341524216524217, + "grad_norm": 0.5274056792259216, + "learning_rate": 0.00018395639352860457, + "loss": 1.1786, + "step": 4123 + }, + { + "epoch": 0.7343304843304843, + "grad_norm": 0.5094266533851624, + "learning_rate": 0.00018394878842691452, + "loss": 1.2016, + "step": 4124 + }, + { + "epoch": 0.7345085470085471, + "grad_norm": 0.48779475688934326, + "learning_rate": 0.0001839411816804224, + "loss": 1.0562, + "step": 4125 + }, + { + "epoch": 0.7346866096866097, + "grad_norm": 0.5805709958076477, + "learning_rate": 0.00018393357328927716, + "loss": 1.1705, + "step": 4126 + }, + { + "epoch": 0.7348646723646723, + "grad_norm": 0.4910700023174286, + "learning_rate": 0.00018392596325362791, + "loss": 1.0682, + "step": 4127 + }, + { + "epoch": 0.7350427350427351, + "grad_norm": 0.5297428369522095, + "learning_rate": 0.0001839183515736238, + "loss": 0.9505, + "step": 4128 + }, + { + "epoch": 0.7352207977207977, + "grad_norm": 0.45442086458206177, + "learning_rate": 0.00018391073824941385, + "loss": 0.9548, + "step": 4129 + }, + { + "epoch": 0.7353988603988604, + "grad_norm": 0.49299946427345276, + "learning_rate": 0.00018390312328114733, + "loss": 1.0868, + "step": 4130 + }, + { + "epoch": 0.7355769230769231, + "grad_norm": 0.4839940369129181, + "learning_rate": 0.0001838955066689734, + "loss": 0.9565, + "step": 4131 + }, + { + "epoch": 0.7357549857549858, + "grad_norm": 0.48600608110427856, + "learning_rate": 0.00018388788841304128, + "loss": 1.2353, + "step": 4132 + }, + { + "epoch": 0.7359330484330484, + "grad_norm": 0.4893583357334137, + "learning_rate": 0.0001838802685135003, + "loss": 0.9595, + "step": 4133 + }, + { + "epoch": 0.7361111111111112, + "grad_norm": 0.4587398171424866, + "learning_rate": 0.00018387264697049963, + "loss": 1.1222, + "step": 4134 + }, + { + "epoch": 0.7362891737891738, + "grad_norm": 0.5361055731773376, + "learning_rate": 0.00018386502378418872, + "loss": 1.3304, + "step": 4135 + }, + { + "epoch": 0.7364672364672364, + "grad_norm": 0.5556629300117493, + "learning_rate": 0.00018385739895471686, + "loss": 1.0358, + "step": 4136 + }, + { + "epoch": 0.7366452991452992, + "grad_norm": 0.45555856823921204, + "learning_rate": 0.00018384977248223346, + "loss": 1.0081, + "step": 4137 + }, + { + "epoch": 0.7368233618233618, + "grad_norm": 0.5606052875518799, + "learning_rate": 0.00018384214436688797, + "loss": 0.9367, + "step": 4138 + }, + { + "epoch": 0.7370014245014245, + "grad_norm": 0.5428356528282166, + "learning_rate": 0.00018383451460882982, + "loss": 1.1391, + "step": 4139 + }, + { + "epoch": 0.7371794871794872, + "grad_norm": 0.4891330897808075, + "learning_rate": 0.00018382688320820853, + "loss": 0.9805, + "step": 4140 + }, + { + "epoch": 0.7373575498575499, + "grad_norm": 0.5407996773719788, + "learning_rate": 0.0001838192501651736, + "loss": 1.0532, + "step": 4141 + }, + { + "epoch": 0.7375356125356125, + "grad_norm": 0.5241971611976624, + "learning_rate": 0.00018381161547987454, + "loss": 0.9509, + "step": 4142 + }, + { + "epoch": 0.7377136752136753, + "grad_norm": 0.5370210409164429, + "learning_rate": 0.000183803979152461, + "loss": 1.2342, + "step": 4143 + }, + { + "epoch": 0.7378917378917379, + "grad_norm": 0.5470060706138611, + "learning_rate": 0.00018379634118308259, + "loss": 0.9621, + "step": 4144 + }, + { + "epoch": 0.7380698005698005, + "grad_norm": 0.546313464641571, + "learning_rate": 0.00018378870157188893, + "loss": 1.1253, + "step": 4145 + }, + { + "epoch": 0.7382478632478633, + "grad_norm": 0.502027153968811, + "learning_rate": 0.00018378106031902974, + "loss": 1.1919, + "step": 4146 + }, + { + "epoch": 0.7384259259259259, + "grad_norm": 0.5282283425331116, + "learning_rate": 0.0001837734174246547, + "loss": 1.0088, + "step": 4147 + }, + { + "epoch": 0.7386039886039886, + "grad_norm": 0.5152897238731384, + "learning_rate": 0.00018376577288891355, + "loss": 1.0813, + "step": 4148 + }, + { + "epoch": 0.7387820512820513, + "grad_norm": 0.5002804398536682, + "learning_rate": 0.0001837581267119561, + "loss": 0.9797, + "step": 4149 + }, + { + "epoch": 0.738960113960114, + "grad_norm": 0.5698176026344299, + "learning_rate": 0.00018375047889393215, + "loss": 1.1099, + "step": 4150 + }, + { + "epoch": 0.7391381766381766, + "grad_norm": 0.5384604930877686, + "learning_rate": 0.00018374282943499156, + "loss": 1.1944, + "step": 4151 + }, + { + "epoch": 0.7393162393162394, + "grad_norm": 0.5483044385910034, + "learning_rate": 0.00018373517833528418, + "loss": 1.1734, + "step": 4152 + }, + { + "epoch": 0.739494301994302, + "grad_norm": 0.4824066162109375, + "learning_rate": 0.0001837275255949599, + "loss": 0.9515, + "step": 4153 + }, + { + "epoch": 0.7396723646723646, + "grad_norm": 0.45413634181022644, + "learning_rate": 0.00018371987121416873, + "loss": 0.7534, + "step": 4154 + }, + { + "epoch": 0.7398504273504274, + "grad_norm": 0.5874246954917908, + "learning_rate": 0.00018371221519306055, + "loss": 0.9464, + "step": 4155 + }, + { + "epoch": 0.74002849002849, + "grad_norm": 0.5219913125038147, + "learning_rate": 0.00018370455753178544, + "loss": 1.0494, + "step": 4156 + }, + { + "epoch": 0.7402065527065527, + "grad_norm": 0.5937709212303162, + "learning_rate": 0.00018369689823049341, + "loss": 1.0529, + "step": 4157 + }, + { + "epoch": 0.7403846153846154, + "grad_norm": 0.5204295516014099, + "learning_rate": 0.00018368923728933449, + "loss": 1.0602, + "step": 4158 + }, + { + "epoch": 0.7405626780626781, + "grad_norm": 0.5422890186309814, + "learning_rate": 0.00018368157470845885, + "loss": 0.9261, + "step": 4159 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.6163852214813232, + "learning_rate": 0.00018367391048801655, + "loss": 1.2771, + "step": 4160 + }, + { + "epoch": 0.7409188034188035, + "grad_norm": 0.5070751309394836, + "learning_rate": 0.00018366624462815785, + "loss": 1.0401, + "step": 4161 + }, + { + "epoch": 0.7410968660968661, + "grad_norm": 0.4477100968360901, + "learning_rate": 0.00018365857712903283, + "loss": 1.1463, + "step": 4162 + }, + { + "epoch": 0.7412749287749287, + "grad_norm": 0.5421462655067444, + "learning_rate": 0.0001836509079907918, + "loss": 0.9373, + "step": 4163 + }, + { + "epoch": 0.7414529914529915, + "grad_norm": 0.6162141561508179, + "learning_rate": 0.000183643237213585, + "loss": 1.1827, + "step": 4164 + }, + { + "epoch": 0.7416310541310541, + "grad_norm": 0.5653836131095886, + "learning_rate": 0.00018363556479756272, + "loss": 1.0689, + "step": 4165 + }, + { + "epoch": 0.7418091168091168, + "grad_norm": 0.57053542137146, + "learning_rate": 0.00018362789074287527, + "loss": 1.0289, + "step": 4166 + }, + { + "epoch": 0.7419871794871795, + "grad_norm": 0.5603055953979492, + "learning_rate": 0.00018362021504967304, + "loss": 1.1926, + "step": 4167 + }, + { + "epoch": 0.7421652421652422, + "grad_norm": 0.5460166335105896, + "learning_rate": 0.0001836125377181064, + "loss": 1.1488, + "step": 4168 + }, + { + "epoch": 0.7423433048433048, + "grad_norm": 0.5097107887268066, + "learning_rate": 0.00018360485874832579, + "loss": 1.0781, + "step": 4169 + }, + { + "epoch": 0.7425213675213675, + "grad_norm": 0.6280624270439148, + "learning_rate": 0.00018359717814048164, + "loss": 1.3625, + "step": 4170 + }, + { + "epoch": 0.7426994301994302, + "grad_norm": 0.4528210759162903, + "learning_rate": 0.0001835894958947244, + "loss": 0.8417, + "step": 4171 + }, + { + "epoch": 0.7428774928774928, + "grad_norm": 0.48735132813453674, + "learning_rate": 0.00018358181201120468, + "loss": 0.9544, + "step": 4172 + }, + { + "epoch": 0.7430555555555556, + "grad_norm": 0.48388174176216125, + "learning_rate": 0.00018357412649007296, + "loss": 1.0663, + "step": 4173 + }, + { + "epoch": 0.7432336182336182, + "grad_norm": 0.5435357689857483, + "learning_rate": 0.00018356643933147986, + "loss": 1.2074, + "step": 4174 + }, + { + "epoch": 0.7434116809116809, + "grad_norm": 0.49890074133872986, + "learning_rate": 0.00018355875053557594, + "loss": 1.1322, + "step": 4175 + }, + { + "epoch": 0.7435897435897436, + "grad_norm": 0.5680708885192871, + "learning_rate": 0.0001835510601025119, + "loss": 1.1964, + "step": 4176 + }, + { + "epoch": 0.7437678062678063, + "grad_norm": 0.5002360939979553, + "learning_rate": 0.00018354336803243842, + "loss": 1.1396, + "step": 4177 + }, + { + "epoch": 0.7439458689458689, + "grad_norm": 0.5202965140342712, + "learning_rate": 0.00018353567432550616, + "loss": 1.1498, + "step": 4178 + }, + { + "epoch": 0.7441239316239316, + "grad_norm": 0.514492928981781, + "learning_rate": 0.00018352797898186588, + "loss": 1.0959, + "step": 4179 + }, + { + "epoch": 0.7443019943019943, + "grad_norm": 0.6395383477210999, + "learning_rate": 0.0001835202820016684, + "loss": 1.2867, + "step": 4180 + }, + { + "epoch": 0.7444800569800569, + "grad_norm": 0.5489062070846558, + "learning_rate": 0.00018351258338506447, + "loss": 1.1638, + "step": 4181 + }, + { + "epoch": 0.7446581196581197, + "grad_norm": 0.5705671906471252, + "learning_rate": 0.00018350488313220498, + "loss": 0.9493, + "step": 4182 + }, + { + "epoch": 0.7448361823361823, + "grad_norm": 0.5404297709465027, + "learning_rate": 0.00018349718124324076, + "loss": 0.9876, + "step": 4183 + }, + { + "epoch": 0.7450142450142451, + "grad_norm": 0.5841003060340881, + "learning_rate": 0.0001834894777183227, + "loss": 1.1225, + "step": 4184 + }, + { + "epoch": 0.7451923076923077, + "grad_norm": 0.49774688482284546, + "learning_rate": 0.00018348177255760178, + "loss": 1.1442, + "step": 4185 + }, + { + "epoch": 0.7453703703703703, + "grad_norm": 0.5212422609329224, + "learning_rate": 0.00018347406576122894, + "loss": 1.101, + "step": 4186 + }, + { + "epoch": 0.7455484330484331, + "grad_norm": 0.615024983882904, + "learning_rate": 0.00018346635732935517, + "loss": 1.4188, + "step": 4187 + }, + { + "epoch": 0.7457264957264957, + "grad_norm": 0.46818843483924866, + "learning_rate": 0.00018345864726213154, + "loss": 1.0071, + "step": 4188 + }, + { + "epoch": 0.7459045584045584, + "grad_norm": 0.4921121895313263, + "learning_rate": 0.00018345093555970906, + "loss": 1.015, + "step": 4189 + }, + { + "epoch": 0.7460826210826211, + "grad_norm": 0.5042136311531067, + "learning_rate": 0.00018344322222223889, + "loss": 0.9974, + "step": 4190 + }, + { + "epoch": 0.7462606837606838, + "grad_norm": 0.5872490406036377, + "learning_rate": 0.0001834355072498721, + "loss": 1.3166, + "step": 4191 + }, + { + "epoch": 0.7464387464387464, + "grad_norm": 0.559117317199707, + "learning_rate": 0.00018342779064275984, + "loss": 1.2227, + "step": 4192 + }, + { + "epoch": 0.7466168091168092, + "grad_norm": 0.5269635319709778, + "learning_rate": 0.00018342007240105336, + "loss": 1.0281, + "step": 4193 + }, + { + "epoch": 0.7467948717948718, + "grad_norm": 0.4608335793018341, + "learning_rate": 0.00018341235252490387, + "loss": 0.98, + "step": 4194 + }, + { + "epoch": 0.7469729344729344, + "grad_norm": 0.5818259119987488, + "learning_rate": 0.00018340463101446255, + "loss": 1.1544, + "step": 4195 + }, + { + "epoch": 0.7471509971509972, + "grad_norm": 0.5577529668807983, + "learning_rate": 0.00018339690786988079, + "loss": 1.3059, + "step": 4196 + }, + { + "epoch": 0.7473290598290598, + "grad_norm": 0.5430468320846558, + "learning_rate": 0.00018338918309130983, + "loss": 1.2766, + "step": 4197 + }, + { + "epoch": 0.7475071225071225, + "grad_norm": 0.4941701591014862, + "learning_rate": 0.0001833814566789011, + "loss": 1.193, + "step": 4198 + }, + { + "epoch": 0.7476851851851852, + "grad_norm": 0.5471884608268738, + "learning_rate": 0.00018337372863280589, + "loss": 1.2261, + "step": 4199 + }, + { + "epoch": 0.7478632478632479, + "grad_norm": 0.4641438126564026, + "learning_rate": 0.0001833659989531757, + "loss": 0.7953, + "step": 4200 + }, + { + "epoch": 0.7480413105413105, + "grad_norm": 0.5244714617729187, + "learning_rate": 0.0001833582676401619, + "loss": 0.9344, + "step": 4201 + }, + { + "epoch": 0.7482193732193733, + "grad_norm": 0.5964360237121582, + "learning_rate": 0.00018335053469391603, + "loss": 1.2072, + "step": 4202 + }, + { + "epoch": 0.7483974358974359, + "grad_norm": 0.4929158091545105, + "learning_rate": 0.00018334280011458954, + "loss": 1.2183, + "step": 4203 + }, + { + "epoch": 0.7485754985754985, + "grad_norm": 0.46221864223480225, + "learning_rate": 0.00018333506390233405, + "loss": 1.1957, + "step": 4204 + }, + { + "epoch": 0.7487535612535613, + "grad_norm": 0.6301732659339905, + "learning_rate": 0.0001833273260573011, + "loss": 1.0582, + "step": 4205 + }, + { + "epoch": 0.7489316239316239, + "grad_norm": 0.5606021881103516, + "learning_rate": 0.0001833195865796423, + "loss": 1.4034, + "step": 4206 + }, + { + "epoch": 0.7491096866096866, + "grad_norm": 0.44856077432632446, + "learning_rate": 0.00018331184546950926, + "loss": 0.8421, + "step": 4207 + }, + { + "epoch": 0.7492877492877493, + "grad_norm": 0.5487226247787476, + "learning_rate": 0.00018330410272705366, + "loss": 1.238, + "step": 4208 + }, + { + "epoch": 0.749465811965812, + "grad_norm": 0.6043636798858643, + "learning_rate": 0.00018329635835242724, + "loss": 1.1215, + "step": 4209 + }, + { + "epoch": 0.7496438746438746, + "grad_norm": 0.5145319104194641, + "learning_rate": 0.00018328861234578173, + "loss": 1.1002, + "step": 4210 + }, + { + "epoch": 0.7498219373219374, + "grad_norm": 0.5667078495025635, + "learning_rate": 0.00018328086470726884, + "loss": 1.2994, + "step": 4211 + }, + { + "epoch": 0.75, + "grad_norm": 0.5117634534835815, + "learning_rate": 0.00018327311543704043, + "loss": 0.9448, + "step": 4212 + }, + { + "epoch": 0.75, + "eval_loss": 1.0982474088668823, + "eval_runtime": 24.6617, + "eval_samples_per_second": 42.211, + "eval_steps_per_second": 21.126, + "step": 4212 + }, + { + "epoch": 0.7501780626780626, + "grad_norm": 0.5451585054397583, + "learning_rate": 0.00018326536453524826, + "loss": 0.9023, + "step": 4213 + }, + { + "epoch": 0.7503561253561254, + "grad_norm": 0.6585208773612976, + "learning_rate": 0.0001832576120020443, + "loss": 1.2798, + "step": 4214 + }, + { + "epoch": 0.750534188034188, + "grad_norm": 0.6444812417030334, + "learning_rate": 0.00018324985783758037, + "loss": 1.3999, + "step": 4215 + }, + { + "epoch": 0.7507122507122507, + "grad_norm": 0.6178330779075623, + "learning_rate": 0.0001832421020420084, + "loss": 1.1846, + "step": 4216 + }, + { + "epoch": 0.7508903133903134, + "grad_norm": 0.509969174861908, + "learning_rate": 0.00018323434461548036, + "loss": 1.1831, + "step": 4217 + }, + { + "epoch": 0.7510683760683761, + "grad_norm": 0.5558911561965942, + "learning_rate": 0.00018322658555814826, + "loss": 1.1599, + "step": 4218 + }, + { + "epoch": 0.7512464387464387, + "grad_norm": 0.5714917778968811, + "learning_rate": 0.0001832188248701641, + "loss": 0.9702, + "step": 4219 + }, + { + "epoch": 0.7514245014245015, + "grad_norm": 0.6136442422866821, + "learning_rate": 0.00018321106255167995, + "loss": 0.9376, + "step": 4220 + }, + { + "epoch": 0.7516025641025641, + "grad_norm": 0.5832077264785767, + "learning_rate": 0.00018320329860284785, + "loss": 1.2564, + "step": 4221 + }, + { + "epoch": 0.7517806267806267, + "grad_norm": 0.45330923795700073, + "learning_rate": 0.00018319553302381997, + "loss": 0.9321, + "step": 4222 + }, + { + "epoch": 0.7519586894586895, + "grad_norm": 0.5278468132019043, + "learning_rate": 0.00018318776581474847, + "loss": 1.1334, + "step": 4223 + }, + { + "epoch": 0.7521367521367521, + "grad_norm": 0.49267473816871643, + "learning_rate": 0.00018317999697578549, + "loss": 1.1577, + "step": 4224 + }, + { + "epoch": 0.7523148148148148, + "grad_norm": 0.5372124314308167, + "learning_rate": 0.00018317222650708325, + "loss": 1.037, + "step": 4225 + }, + { + "epoch": 0.7524928774928775, + "grad_norm": 0.5879829525947571, + "learning_rate": 0.000183164454408794, + "loss": 1.1312, + "step": 4226 + }, + { + "epoch": 0.7526709401709402, + "grad_norm": 0.5363932251930237, + "learning_rate": 0.00018315668068107004, + "loss": 1.174, + "step": 4227 + }, + { + "epoch": 0.7528490028490028, + "grad_norm": 0.5585991740226746, + "learning_rate": 0.00018314890532406366, + "loss": 1.2106, + "step": 4228 + }, + { + "epoch": 0.7530270655270656, + "grad_norm": 0.49395787715911865, + "learning_rate": 0.0001831411283379272, + "loss": 1.1163, + "step": 4229 + }, + { + "epoch": 0.7532051282051282, + "grad_norm": 0.5081066489219666, + "learning_rate": 0.00018313334972281306, + "loss": 1.184, + "step": 4230 + }, + { + "epoch": 0.7533831908831908, + "grad_norm": 0.40304034948349, + "learning_rate": 0.0001831255694788736, + "loss": 0.7548, + "step": 4231 + }, + { + "epoch": 0.7535612535612536, + "grad_norm": 0.4999815821647644, + "learning_rate": 0.0001831177876062613, + "loss": 1.0092, + "step": 4232 + }, + { + "epoch": 0.7537393162393162, + "grad_norm": 0.48917025327682495, + "learning_rate": 0.00018311000410512862, + "loss": 1.0354, + "step": 4233 + }, + { + "epoch": 0.7539173789173789, + "grad_norm": 0.475606769323349, + "learning_rate": 0.00018310221897562806, + "loss": 0.8728, + "step": 4234 + }, + { + "epoch": 0.7540954415954416, + "grad_norm": 0.630439817905426, + "learning_rate": 0.00018309443221791214, + "loss": 1.1436, + "step": 4235 + }, + { + "epoch": 0.7542735042735043, + "grad_norm": 0.524740993976593, + "learning_rate": 0.00018308664383213344, + "loss": 1.0487, + "step": 4236 + }, + { + "epoch": 0.7544515669515669, + "grad_norm": 0.4734523892402649, + "learning_rate": 0.0001830788538184445, + "loss": 1.0681, + "step": 4237 + }, + { + "epoch": 0.7546296296296297, + "grad_norm": 0.5767266750335693, + "learning_rate": 0.00018307106217699807, + "loss": 1.0599, + "step": 4238 + }, + { + "epoch": 0.7548076923076923, + "grad_norm": 0.6276642084121704, + "learning_rate": 0.0001830632689079467, + "loss": 1.2837, + "step": 4239 + }, + { + "epoch": 0.7549857549857549, + "grad_norm": 0.5539988279342651, + "learning_rate": 0.00018305547401144316, + "loss": 0.9072, + "step": 4240 + }, + { + "epoch": 0.7551638176638177, + "grad_norm": 0.4551292061805725, + "learning_rate": 0.00018304767748764014, + "loss": 1.0204, + "step": 4241 + }, + { + "epoch": 0.7553418803418803, + "grad_norm": 0.47344550490379333, + "learning_rate": 0.00018303987933669034, + "loss": 1.0473, + "step": 4242 + }, + { + "epoch": 0.7555199430199431, + "grad_norm": 0.6050213575363159, + "learning_rate": 0.00018303207955874665, + "loss": 1.1552, + "step": 4243 + }, + { + "epoch": 0.7556980056980057, + "grad_norm": 0.48943889141082764, + "learning_rate": 0.00018302427815396186, + "loss": 1.0002, + "step": 4244 + }, + { + "epoch": 0.7558760683760684, + "grad_norm": 0.5664682984352112, + "learning_rate": 0.00018301647512248878, + "loss": 1.1865, + "step": 4245 + }, + { + "epoch": 0.7560541310541311, + "grad_norm": 0.5702242255210876, + "learning_rate": 0.00018300867046448034, + "loss": 1.3029, + "step": 4246 + }, + { + "epoch": 0.7562321937321937, + "grad_norm": 0.593207836151123, + "learning_rate": 0.00018300086418008942, + "loss": 1.109, + "step": 4247 + }, + { + "epoch": 0.7564102564102564, + "grad_norm": 0.5887887477874756, + "learning_rate": 0.000182993056269469, + "loss": 1.3022, + "step": 4248 + }, + { + "epoch": 0.7565883190883191, + "grad_norm": 0.5277966260910034, + "learning_rate": 0.00018298524673277203, + "loss": 1.1738, + "step": 4249 + }, + { + "epoch": 0.7567663817663818, + "grad_norm": 0.589347779750824, + "learning_rate": 0.00018297743557015155, + "loss": 1.0185, + "step": 4250 + }, + { + "epoch": 0.7569444444444444, + "grad_norm": 0.49920859932899475, + "learning_rate": 0.0001829696227817606, + "loss": 1.118, + "step": 4251 + }, + { + "epoch": 0.7571225071225072, + "grad_norm": 0.502565324306488, + "learning_rate": 0.0001829618083677522, + "loss": 1.1856, + "step": 4252 + }, + { + "epoch": 0.7573005698005698, + "grad_norm": 0.49814435839653015, + "learning_rate": 0.00018295399232827955, + "loss": 1.0432, + "step": 4253 + }, + { + "epoch": 0.7574786324786325, + "grad_norm": 0.5087502598762512, + "learning_rate": 0.00018294617466349574, + "loss": 1.2325, + "step": 4254 + }, + { + "epoch": 0.7576566951566952, + "grad_norm": 0.5107288956642151, + "learning_rate": 0.00018293835537355394, + "loss": 1.0487, + "step": 4255 + }, + { + "epoch": 0.7578347578347578, + "grad_norm": 0.524725615978241, + "learning_rate": 0.00018293053445860732, + "loss": 1.1821, + "step": 4256 + }, + { + "epoch": 0.7580128205128205, + "grad_norm": 0.5234082937240601, + "learning_rate": 0.0001829227119188092, + "loss": 0.8896, + "step": 4257 + }, + { + "epoch": 0.7581908831908832, + "grad_norm": 0.5102918744087219, + "learning_rate": 0.00018291488775431275, + "loss": 1.0246, + "step": 4258 + }, + { + "epoch": 0.7583689458689459, + "grad_norm": 0.5552714467048645, + "learning_rate": 0.00018290706196527135, + "loss": 1.0193, + "step": 4259 + }, + { + "epoch": 0.7585470085470085, + "grad_norm": 0.5395022630691528, + "learning_rate": 0.00018289923455183825, + "loss": 1.3203, + "step": 4260 + }, + { + "epoch": 0.7587250712250713, + "grad_norm": 0.7474865913391113, + "learning_rate": 0.00018289140551416692, + "loss": 1.182, + "step": 4261 + }, + { + "epoch": 0.7589031339031339, + "grad_norm": 0.4892016649246216, + "learning_rate": 0.00018288357485241066, + "loss": 0.968, + "step": 4262 + }, + { + "epoch": 0.7590811965811965, + "grad_norm": 0.4627816081047058, + "learning_rate": 0.00018287574256672291, + "loss": 0.6895, + "step": 4263 + }, + { + "epoch": 0.7592592592592593, + "grad_norm": 0.6221280097961426, + "learning_rate": 0.00018286790865725715, + "loss": 0.9691, + "step": 4264 + }, + { + "epoch": 0.7594373219373219, + "grad_norm": 0.5542295575141907, + "learning_rate": 0.0001828600731241669, + "loss": 0.9996, + "step": 4265 + }, + { + "epoch": 0.7596153846153846, + "grad_norm": 0.5570770502090454, + "learning_rate": 0.00018285223596760562, + "loss": 1.1996, + "step": 4266 + }, + { + "epoch": 0.7597934472934473, + "grad_norm": 0.5495262742042542, + "learning_rate": 0.00018284439718772687, + "loss": 1.1572, + "step": 4267 + }, + { + "epoch": 0.75997150997151, + "grad_norm": 0.5006741881370544, + "learning_rate": 0.00018283655678468427, + "loss": 1.1215, + "step": 4268 + }, + { + "epoch": 0.7601495726495726, + "grad_norm": 0.4682157635688782, + "learning_rate": 0.00018282871475863144, + "loss": 1.0547, + "step": 4269 + }, + { + "epoch": 0.7603276353276354, + "grad_norm": 0.6275840997695923, + "learning_rate": 0.00018282087110972197, + "loss": 1.3855, + "step": 4270 + }, + { + "epoch": 0.760505698005698, + "grad_norm": 0.5341474413871765, + "learning_rate": 0.0001828130258381096, + "loss": 1.2024, + "step": 4271 + }, + { + "epoch": 0.7606837606837606, + "grad_norm": 0.4330833852291107, + "learning_rate": 0.000182805178943948, + "loss": 1.0508, + "step": 4272 + }, + { + "epoch": 0.7608618233618234, + "grad_norm": 0.6276537179946899, + "learning_rate": 0.00018279733042739094, + "loss": 1.1635, + "step": 4273 + }, + { + "epoch": 0.761039886039886, + "grad_norm": 0.5370199084281921, + "learning_rate": 0.00018278948028859217, + "loss": 1.0579, + "step": 4274 + }, + { + "epoch": 0.7612179487179487, + "grad_norm": 0.524959921836853, + "learning_rate": 0.00018278162852770552, + "loss": 1.0972, + "step": 4275 + }, + { + "epoch": 0.7613960113960114, + "grad_norm": 0.5029389262199402, + "learning_rate": 0.00018277377514488486, + "loss": 0.959, + "step": 4276 + }, + { + "epoch": 0.7615740740740741, + "grad_norm": 0.49772894382476807, + "learning_rate": 0.00018276592014028397, + "loss": 1.2773, + "step": 4277 + }, + { + "epoch": 0.7617521367521367, + "grad_norm": 0.5195719003677368, + "learning_rate": 0.00018275806351405685, + "loss": 1.0676, + "step": 4278 + }, + { + "epoch": 0.7619301994301995, + "grad_norm": 0.5167942643165588, + "learning_rate": 0.00018275020526635735, + "loss": 1.0615, + "step": 4279 + }, + { + "epoch": 0.7621082621082621, + "grad_norm": 0.4958035945892334, + "learning_rate": 0.0001827423453973395, + "loss": 0.9605, + "step": 4280 + }, + { + "epoch": 0.7622863247863247, + "grad_norm": 0.6256808042526245, + "learning_rate": 0.00018273448390715728, + "loss": 1.2526, + "step": 4281 + }, + { + "epoch": 0.7624643874643875, + "grad_norm": 0.5062580108642578, + "learning_rate": 0.0001827266207959647, + "loss": 1.0604, + "step": 4282 + }, + { + "epoch": 0.7626424501424501, + "grad_norm": 0.5080778002738953, + "learning_rate": 0.00018271875606391583, + "loss": 1.1246, + "step": 4283 + }, + { + "epoch": 0.7628205128205128, + "grad_norm": 0.5069389939308167, + "learning_rate": 0.00018271088971116479, + "loss": 1.3158, + "step": 4284 + }, + { + "epoch": 0.7629985754985755, + "grad_norm": 0.7280121445655823, + "learning_rate": 0.00018270302173786567, + "loss": 1.2066, + "step": 4285 + }, + { + "epoch": 0.7631766381766382, + "grad_norm": 0.6523470282554626, + "learning_rate": 0.00018269515214417267, + "loss": 1.3236, + "step": 4286 + }, + { + "epoch": 0.7633547008547008, + "grad_norm": 0.5799322724342346, + "learning_rate": 0.00018268728093023988, + "loss": 0.9786, + "step": 4287 + }, + { + "epoch": 0.7635327635327636, + "grad_norm": 0.46675166487693787, + "learning_rate": 0.00018267940809622163, + "loss": 0.8131, + "step": 4288 + }, + { + "epoch": 0.7637108262108262, + "grad_norm": 0.5566182732582092, + "learning_rate": 0.00018267153364227214, + "loss": 1.0565, + "step": 4289 + }, + { + "epoch": 0.7638888888888888, + "grad_norm": 0.532028079032898, + "learning_rate": 0.00018266365756854566, + "loss": 0.952, + "step": 4290 + }, + { + "epoch": 0.7640669515669516, + "grad_norm": 0.5082666873931885, + "learning_rate": 0.00018265577987519653, + "loss": 1.0704, + "step": 4291 + }, + { + "epoch": 0.7642450142450142, + "grad_norm": 0.5223562717437744, + "learning_rate": 0.00018264790056237912, + "loss": 1.1161, + "step": 4292 + }, + { + "epoch": 0.7644230769230769, + "grad_norm": 0.48472318053245544, + "learning_rate": 0.00018264001963024778, + "loss": 0.8784, + "step": 4293 + }, + { + "epoch": 0.7646011396011396, + "grad_norm": 0.5901281833648682, + "learning_rate": 0.0001826321370789569, + "loss": 1.1031, + "step": 4294 + }, + { + "epoch": 0.7647792022792023, + "grad_norm": 0.570350706577301, + "learning_rate": 0.000182624252908661, + "loss": 0.9047, + "step": 4295 + }, + { + "epoch": 0.7649572649572649, + "grad_norm": 0.568373441696167, + "learning_rate": 0.00018261636711951445, + "loss": 1.0106, + "step": 4296 + }, + { + "epoch": 0.7651353276353277, + "grad_norm": 0.6175880432128906, + "learning_rate": 0.00018260847971167182, + "loss": 1.3531, + "step": 4297 + }, + { + "epoch": 0.7653133903133903, + "grad_norm": 0.5682594776153564, + "learning_rate": 0.00018260059068528762, + "loss": 1.1261, + "step": 4298 + }, + { + "epoch": 0.7654914529914529, + "grad_norm": 0.5050225257873535, + "learning_rate": 0.00018259270004051644, + "loss": 1.0921, + "step": 4299 + }, + { + "epoch": 0.7656695156695157, + "grad_norm": 0.5416565537452698, + "learning_rate": 0.0001825848077775129, + "loss": 1.0881, + "step": 4300 + }, + { + "epoch": 0.7658475783475783, + "grad_norm": 0.5418867468833923, + "learning_rate": 0.0001825769138964316, + "loss": 1.2069, + "step": 4301 + }, + { + "epoch": 0.7660256410256411, + "grad_norm": 0.5447866320610046, + "learning_rate": 0.00018256901839742718, + "loss": 1.1827, + "step": 4302 + }, + { + "epoch": 0.7662037037037037, + "grad_norm": 0.5482802987098694, + "learning_rate": 0.00018256112128065439, + "loss": 1.0492, + "step": 4303 + }, + { + "epoch": 0.7663817663817664, + "grad_norm": 0.5059601664543152, + "learning_rate": 0.0001825532225462679, + "loss": 1.0996, + "step": 4304 + }, + { + "epoch": 0.7665598290598291, + "grad_norm": 0.5153701901435852, + "learning_rate": 0.00018254532219442258, + "loss": 1.3237, + "step": 4305 + }, + { + "epoch": 0.7667378917378918, + "grad_norm": 0.5370768904685974, + "learning_rate": 0.0001825374202252731, + "loss": 0.9925, + "step": 4306 + }, + { + "epoch": 0.7669159544159544, + "grad_norm": 0.4516580402851105, + "learning_rate": 0.00018252951663897432, + "loss": 1.0749, + "step": 4307 + }, + { + "epoch": 0.7670940170940171, + "grad_norm": 0.5565171837806702, + "learning_rate": 0.0001825216114356811, + "loss": 1.1617, + "step": 4308 + }, + { + "epoch": 0.7672720797720798, + "grad_norm": 0.5212662220001221, + "learning_rate": 0.00018251370461554834, + "loss": 1.1108, + "step": 4309 + }, + { + "epoch": 0.7674501424501424, + "grad_norm": 0.49061715602874756, + "learning_rate": 0.00018250579617873095, + "loss": 1.0881, + "step": 4310 + }, + { + "epoch": 0.7676282051282052, + "grad_norm": 0.5535751581192017, + "learning_rate": 0.00018249788612538387, + "loss": 0.9341, + "step": 4311 + }, + { + "epoch": 0.7678062678062678, + "grad_norm": 0.5425209403038025, + "learning_rate": 0.00018248997445566208, + "loss": 1.1858, + "step": 4312 + }, + { + "epoch": 0.7679843304843305, + "grad_norm": 0.6224395036697388, + "learning_rate": 0.0001824820611697206, + "loss": 1.0836, + "step": 4313 + }, + { + "epoch": 0.7681623931623932, + "grad_norm": 0.4895690977573395, + "learning_rate": 0.00018247414626771445, + "loss": 0.8598, + "step": 4314 + }, + { + "epoch": 0.7683404558404558, + "grad_norm": 0.5279615521430969, + "learning_rate": 0.00018246622974979877, + "loss": 1.1742, + "step": 4315 + }, + { + "epoch": 0.7685185185185185, + "grad_norm": 0.45300471782684326, + "learning_rate": 0.0001824583116161286, + "loss": 0.8872, + "step": 4316 + }, + { + "epoch": 0.7686965811965812, + "grad_norm": 0.6499692797660828, + "learning_rate": 0.00018245039186685916, + "loss": 1.2495, + "step": 4317 + }, + { + "epoch": 0.7688746438746439, + "grad_norm": 0.48151278495788574, + "learning_rate": 0.00018244247050214552, + "loss": 1.2382, + "step": 4318 + }, + { + "epoch": 0.7690527065527065, + "grad_norm": 0.6597028374671936, + "learning_rate": 0.0001824345475221429, + "loss": 1.3453, + "step": 4319 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.4536992609500885, + "learning_rate": 0.0001824266229270066, + "loss": 1.1141, + "step": 4320 + }, + { + "epoch": 0.7694088319088319, + "grad_norm": 0.5489405393600464, + "learning_rate": 0.00018241869671689184, + "loss": 1.0333, + "step": 4321 + }, + { + "epoch": 0.7695868945868946, + "grad_norm": 0.5741586089134216, + "learning_rate": 0.00018241076889195394, + "loss": 0.9939, + "step": 4322 + }, + { + "epoch": 0.7697649572649573, + "grad_norm": 0.47170960903167725, + "learning_rate": 0.00018240283945234823, + "loss": 0.9878, + "step": 4323 + }, + { + "epoch": 0.76994301994302, + "grad_norm": 0.4729093313217163, + "learning_rate": 0.00018239490839823004, + "loss": 1.0087, + "step": 4324 + }, + { + "epoch": 0.7701210826210826, + "grad_norm": 0.49869823455810547, + "learning_rate": 0.0001823869757297548, + "loss": 1.169, + "step": 4325 + }, + { + "epoch": 0.7702991452991453, + "grad_norm": 0.5118468403816223, + "learning_rate": 0.0001823790414470779, + "loss": 1.1092, + "step": 4326 + }, + { + "epoch": 0.770477207977208, + "grad_norm": 0.5076048970222473, + "learning_rate": 0.0001823711055503548, + "loss": 1.1028, + "step": 4327 + }, + { + "epoch": 0.7706552706552706, + "grad_norm": 0.5661569237709045, + "learning_rate": 0.00018236316803974098, + "loss": 1.1114, + "step": 4328 + }, + { + "epoch": 0.7708333333333334, + "grad_norm": 0.5542354583740234, + "learning_rate": 0.000182355228915392, + "loss": 1.0931, + "step": 4329 + }, + { + "epoch": 0.771011396011396, + "grad_norm": 0.5476680994033813, + "learning_rate": 0.0001823472881774634, + "loss": 1.036, + "step": 4330 + }, + { + "epoch": 0.7711894586894587, + "grad_norm": 0.5449798703193665, + "learning_rate": 0.00018233934582611073, + "loss": 1.0682, + "step": 4331 + }, + { + "epoch": 0.7713675213675214, + "grad_norm": 0.61089026927948, + "learning_rate": 0.00018233140186148963, + "loss": 1.0748, + "step": 4332 + }, + { + "epoch": 0.771545584045584, + "grad_norm": 0.5015206336975098, + "learning_rate": 0.00018232345628375576, + "loss": 1.2032, + "step": 4333 + }, + { + "epoch": 0.7717236467236467, + "grad_norm": 0.579289972782135, + "learning_rate": 0.00018231550909306475, + "loss": 1.0764, + "step": 4334 + }, + { + "epoch": 0.7719017094017094, + "grad_norm": 0.5889299511909485, + "learning_rate": 0.00018230756028957235, + "loss": 1.1768, + "step": 4335 + }, + { + "epoch": 0.7720797720797721, + "grad_norm": 0.5328249335289001, + "learning_rate": 0.00018229960987343428, + "loss": 1.0055, + "step": 4336 + }, + { + "epoch": 0.7722578347578347, + "grad_norm": 0.5766382217407227, + "learning_rate": 0.0001822916578448063, + "loss": 0.9923, + "step": 4337 + }, + { + "epoch": 0.7724358974358975, + "grad_norm": 0.6448187828063965, + "learning_rate": 0.00018228370420384423, + "loss": 1.1135, + "step": 4338 + }, + { + "epoch": 0.7726139601139601, + "grad_norm": 0.5505210757255554, + "learning_rate": 0.00018227574895070394, + "loss": 1.2048, + "step": 4339 + }, + { + "epoch": 0.7727920227920227, + "grad_norm": 0.6278925538063049, + "learning_rate": 0.00018226779208554126, + "loss": 1.1045, + "step": 4340 + }, + { + "epoch": 0.7729700854700855, + "grad_norm": 0.5345009565353394, + "learning_rate": 0.00018225983360851207, + "loss": 1.0102, + "step": 4341 + }, + { + "epoch": 0.7731481481481481, + "grad_norm": 0.566633403301239, + "learning_rate": 0.00018225187351977233, + "loss": 1.0038, + "step": 4342 + }, + { + "epoch": 0.7733262108262108, + "grad_norm": 0.5066078901290894, + "learning_rate": 0.000182243911819478, + "loss": 1.0339, + "step": 4343 + }, + { + "epoch": 0.7735042735042735, + "grad_norm": 0.5614920258522034, + "learning_rate": 0.00018223594850778503, + "loss": 1.1021, + "step": 4344 + }, + { + "epoch": 0.7736823361823362, + "grad_norm": 0.7747337818145752, + "learning_rate": 0.0001822279835848495, + "loss": 1.1129, + "step": 4345 + }, + { + "epoch": 0.7738603988603988, + "grad_norm": 0.7066529989242554, + "learning_rate": 0.00018222001705082744, + "loss": 1.3234, + "step": 4346 + }, + { + "epoch": 0.7740384615384616, + "grad_norm": 0.6340884566307068, + "learning_rate": 0.00018221204890587497, + "loss": 1.0726, + "step": 4347 + }, + { + "epoch": 0.7742165242165242, + "grad_norm": 0.5401145815849304, + "learning_rate": 0.00018220407915014818, + "loss": 0.9904, + "step": 4348 + }, + { + "epoch": 0.7743945868945868, + "grad_norm": 0.5069159269332886, + "learning_rate": 0.00018219610778380315, + "loss": 1.0654, + "step": 4349 + }, + { + "epoch": 0.7745726495726496, + "grad_norm": 0.5422839522361755, + "learning_rate": 0.00018218813480699623, + "loss": 1.1741, + "step": 4350 + }, + { + "epoch": 0.7747507122507122, + "grad_norm": 0.5550300478935242, + "learning_rate": 0.0001821801602198835, + "loss": 1.0033, + "step": 4351 + }, + { + "epoch": 0.7749287749287749, + "grad_norm": 0.5987736582756042, + "learning_rate": 0.00018217218402262123, + "loss": 0.935, + "step": 4352 + }, + { + "epoch": 0.7751068376068376, + "grad_norm": 0.6137008666992188, + "learning_rate": 0.00018216420621536573, + "loss": 1.17, + "step": 4353 + }, + { + "epoch": 0.7752849002849003, + "grad_norm": 0.47124359011650085, + "learning_rate": 0.0001821562267982733, + "loss": 0.8316, + "step": 4354 + }, + { + "epoch": 0.7754629629629629, + "grad_norm": 0.5057868361473083, + "learning_rate": 0.00018214824577150024, + "loss": 1.0246, + "step": 4355 + }, + { + "epoch": 0.7756410256410257, + "grad_norm": 0.604055643081665, + "learning_rate": 0.00018214026313520299, + "loss": 1.1272, + "step": 4356 + }, + { + "epoch": 0.7758190883190883, + "grad_norm": 0.6690384149551392, + "learning_rate": 0.0001821322788895379, + "loss": 1.0464, + "step": 4357 + }, + { + "epoch": 0.7759971509971509, + "grad_norm": 0.5458958745002747, + "learning_rate": 0.0001821242930346614, + "loss": 1.1712, + "step": 4358 + }, + { + "epoch": 0.7761752136752137, + "grad_norm": 0.6448663473129272, + "learning_rate": 0.00018211630557073, + "loss": 1.1125, + "step": 4359 + }, + { + "epoch": 0.7763532763532763, + "grad_norm": 0.49889448285102844, + "learning_rate": 0.00018210831649790018, + "loss": 1.097, + "step": 4360 + }, + { + "epoch": 0.7765313390313391, + "grad_norm": 0.5118046998977661, + "learning_rate": 0.00018210032581632843, + "loss": 1.009, + "step": 4361 + }, + { + "epoch": 0.7767094017094017, + "grad_norm": 0.5450068116188049, + "learning_rate": 0.00018209233352617135, + "loss": 1.1138, + "step": 4362 + }, + { + "epoch": 0.7768874643874644, + "grad_norm": 0.6147481203079224, + "learning_rate": 0.00018208433962758558, + "loss": 1.212, + "step": 4363 + }, + { + "epoch": 0.7770655270655271, + "grad_norm": 0.554176926612854, + "learning_rate": 0.00018207634412072764, + "loss": 1.1271, + "step": 4364 + }, + { + "epoch": 0.7772435897435898, + "grad_norm": 0.5872851014137268, + "learning_rate": 0.00018206834700575426, + "loss": 1.2793, + "step": 4365 + }, + { + "epoch": 0.7774216524216524, + "grad_norm": 0.5135685205459595, + "learning_rate": 0.00018206034828282207, + "loss": 0.9642, + "step": 4366 + }, + { + "epoch": 0.7775997150997151, + "grad_norm": 0.5699490308761597, + "learning_rate": 0.00018205234795208786, + "loss": 0.9086, + "step": 4367 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 0.5908057689666748, + "learning_rate": 0.00018204434601370832, + "loss": 1.1973, + "step": 4368 + }, + { + "epoch": 0.7779558404558404, + "grad_norm": 0.5777581334114075, + "learning_rate": 0.00018203634246784025, + "loss": 1.0447, + "step": 4369 + }, + { + "epoch": 0.7781339031339032, + "grad_norm": 0.4822927713394165, + "learning_rate": 0.00018202833731464048, + "loss": 0.814, + "step": 4370 + }, + { + "epoch": 0.7783119658119658, + "grad_norm": 0.5343610644340515, + "learning_rate": 0.0001820203305542658, + "loss": 1.2785, + "step": 4371 + }, + { + "epoch": 0.7784900284900285, + "grad_norm": 0.5462222695350647, + "learning_rate": 0.00018201232218687316, + "loss": 1.1785, + "step": 4372 + }, + { + "epoch": 0.7786680911680912, + "grad_norm": 0.5177609324455261, + "learning_rate": 0.00018200431221261943, + "loss": 1.111, + "step": 4373 + }, + { + "epoch": 0.7788461538461539, + "grad_norm": 0.5324625968933105, + "learning_rate": 0.00018199630063166157, + "loss": 1.0738, + "step": 4374 + }, + { + "epoch": 0.7790242165242165, + "grad_norm": 0.6392876505851746, + "learning_rate": 0.0001819882874441565, + "loss": 1.1758, + "step": 4375 + }, + { + "epoch": 0.7792022792022792, + "grad_norm": 0.49964696168899536, + "learning_rate": 0.00018198027265026127, + "loss": 1.0556, + "step": 4376 + }, + { + "epoch": 0.7793803418803419, + "grad_norm": 0.6090660691261292, + "learning_rate": 0.00018197225625013287, + "loss": 1.0102, + "step": 4377 + }, + { + "epoch": 0.7795584045584045, + "grad_norm": 0.5242345929145813, + "learning_rate": 0.00018196423824392842, + "loss": 0.8335, + "step": 4378 + }, + { + "epoch": 0.7797364672364673, + "grad_norm": 0.5265036225318909, + "learning_rate": 0.00018195621863180498, + "loss": 1.0781, + "step": 4379 + }, + { + "epoch": 0.7799145299145299, + "grad_norm": 0.5115378499031067, + "learning_rate": 0.0001819481974139197, + "loss": 1.1658, + "step": 4380 + }, + { + "epoch": 0.7800925925925926, + "grad_norm": 0.6489549875259399, + "learning_rate": 0.00018194017459042972, + "loss": 1.0572, + "step": 4381 + }, + { + "epoch": 0.7802706552706553, + "grad_norm": 0.5800202488899231, + "learning_rate": 0.0001819321501614922, + "loss": 0.9593, + "step": 4382 + }, + { + "epoch": 0.780448717948718, + "grad_norm": 0.5608528256416321, + "learning_rate": 0.00018192412412726443, + "loss": 1.0324, + "step": 4383 + }, + { + "epoch": 0.7806267806267806, + "grad_norm": 0.5596401691436768, + "learning_rate": 0.00018191609648790362, + "loss": 1.071, + "step": 4384 + }, + { + "epoch": 0.7808048433048433, + "grad_norm": 0.5712903141975403, + "learning_rate": 0.00018190806724356707, + "loss": 0.9011, + "step": 4385 + }, + { + "epoch": 0.780982905982906, + "grad_norm": 0.5079438090324402, + "learning_rate": 0.0001819000363944121, + "loss": 1.1194, + "step": 4386 + }, + { + "epoch": 0.7811609686609686, + "grad_norm": 0.5785079598426819, + "learning_rate": 0.00018189200394059602, + "loss": 1.1703, + "step": 4387 + }, + { + "epoch": 0.7813390313390314, + "grad_norm": 0.6901816129684448, + "learning_rate": 0.00018188396988227625, + "loss": 1.6689, + "step": 4388 + }, + { + "epoch": 0.781517094017094, + "grad_norm": 0.48107922077178955, + "learning_rate": 0.00018187593421961022, + "loss": 1.0116, + "step": 4389 + }, + { + "epoch": 0.7816951566951567, + "grad_norm": 0.5843084454536438, + "learning_rate": 0.0001818678969527553, + "loss": 1.1172, + "step": 4390 + }, + { + "epoch": 0.7818732193732194, + "grad_norm": 0.479034423828125, + "learning_rate": 0.00018185985808186902, + "loss": 0.811, + "step": 4391 + }, + { + "epoch": 0.782051282051282, + "grad_norm": 0.5864158272743225, + "learning_rate": 0.00018185181760710888, + "loss": 0.9522, + "step": 4392 + }, + { + "epoch": 0.7822293447293447, + "grad_norm": 0.4824625551700592, + "learning_rate": 0.00018184377552863242, + "loss": 0.9039, + "step": 4393 + }, + { + "epoch": 0.7824074074074074, + "grad_norm": 0.580102801322937, + "learning_rate": 0.00018183573184659717, + "loss": 1.2382, + "step": 4394 + }, + { + "epoch": 0.7825854700854701, + "grad_norm": 0.5300056338310242, + "learning_rate": 0.00018182768656116073, + "loss": 1.2268, + "step": 4395 + }, + { + "epoch": 0.7827635327635327, + "grad_norm": 0.5548123121261597, + "learning_rate": 0.00018181963967248078, + "loss": 1.0628, + "step": 4396 + }, + { + "epoch": 0.7829415954415955, + "grad_norm": 0.5485070943832397, + "learning_rate": 0.00018181159118071496, + "loss": 0.9628, + "step": 4397 + }, + { + "epoch": 0.7831196581196581, + "grad_norm": 0.47405415773391724, + "learning_rate": 0.00018180354108602095, + "loss": 1.1413, + "step": 4398 + }, + { + "epoch": 0.7832977207977208, + "grad_norm": 0.5545752644538879, + "learning_rate": 0.0001817954893885565, + "loss": 1.3807, + "step": 4399 + }, + { + "epoch": 0.7834757834757835, + "grad_norm": 0.5339497327804565, + "learning_rate": 0.00018178743608847933, + "loss": 0.9978, + "step": 4400 + }, + { + "epoch": 0.7836538461538461, + "grad_norm": 0.5006352663040161, + "learning_rate": 0.00018177938118594725, + "loss": 0.8873, + "step": 4401 + }, + { + "epoch": 0.7838319088319088, + "grad_norm": 0.4845179319381714, + "learning_rate": 0.00018177132468111812, + "loss": 0.8866, + "step": 4402 + }, + { + "epoch": 0.7840099715099715, + "grad_norm": 0.5240967869758606, + "learning_rate": 0.0001817632665741497, + "loss": 1.0347, + "step": 4403 + }, + { + "epoch": 0.7841880341880342, + "grad_norm": 0.5311884880065918, + "learning_rate": 0.00018175520686519993, + "loss": 1.2065, + "step": 4404 + }, + { + "epoch": 0.7843660968660968, + "grad_norm": 0.5562815070152283, + "learning_rate": 0.00018174714555442673, + "loss": 1.1272, + "step": 4405 + }, + { + "epoch": 0.7845441595441596, + "grad_norm": 0.5524366497993469, + "learning_rate": 0.00018173908264198802, + "loss": 1.2337, + "step": 4406 + }, + { + "epoch": 0.7847222222222222, + "grad_norm": 0.5612216591835022, + "learning_rate": 0.0001817310181280418, + "loss": 1.1809, + "step": 4407 + }, + { + "epoch": 0.7849002849002849, + "grad_norm": 0.5315343737602234, + "learning_rate": 0.000181722952012746, + "loss": 1.0491, + "step": 4408 + }, + { + "epoch": 0.7850783475783476, + "grad_norm": 0.5233435034751892, + "learning_rate": 0.00018171488429625878, + "loss": 1.0457, + "step": 4409 + }, + { + "epoch": 0.7852564102564102, + "grad_norm": 0.7809093594551086, + "learning_rate": 0.00018170681497873813, + "loss": 1.1578, + "step": 4410 + }, + { + "epoch": 0.7854344729344729, + "grad_norm": 0.49659839272499084, + "learning_rate": 0.00018169874406034217, + "loss": 1.0815, + "step": 4411 + }, + { + "epoch": 0.7856125356125356, + "grad_norm": 0.5020765066146851, + "learning_rate": 0.00018169067154122904, + "loss": 1.1985, + "step": 4412 + }, + { + "epoch": 0.7857905982905983, + "grad_norm": 0.6408432126045227, + "learning_rate": 0.0001816825974215569, + "loss": 1.2272, + "step": 4413 + }, + { + "epoch": 0.7859686609686609, + "grad_norm": 0.5062605142593384, + "learning_rate": 0.00018167452170148396, + "loss": 0.9663, + "step": 4414 + }, + { + "epoch": 0.7861467236467237, + "grad_norm": 0.5100119113922119, + "learning_rate": 0.0001816664443811684, + "loss": 1.0256, + "step": 4415 + }, + { + "epoch": 0.7863247863247863, + "grad_norm": 0.5277643799781799, + "learning_rate": 0.00018165836546076854, + "loss": 1.2885, + "step": 4416 + }, + { + "epoch": 0.7865028490028491, + "grad_norm": 0.5568150281906128, + "learning_rate": 0.0001816502849404426, + "loss": 1.2673, + "step": 4417 + }, + { + "epoch": 0.7866809116809117, + "grad_norm": 0.5061392188072205, + "learning_rate": 0.00018164220282034896, + "loss": 1.072, + "step": 4418 + }, + { + "epoch": 0.7868589743589743, + "grad_norm": 0.5383077263832092, + "learning_rate": 0.00018163411910064597, + "loss": 1.0621, + "step": 4419 + }, + { + "epoch": 0.7870370370370371, + "grad_norm": 0.5167948007583618, + "learning_rate": 0.00018162603378149198, + "loss": 1.099, + "step": 4420 + }, + { + "epoch": 0.7872150997150997, + "grad_norm": 0.5084534287452698, + "learning_rate": 0.0001816179468630454, + "loss": 1.3984, + "step": 4421 + }, + { + "epoch": 0.7873931623931624, + "grad_norm": 0.608762264251709, + "learning_rate": 0.00018160985834546475, + "loss": 1.3553, + "step": 4422 + }, + { + "epoch": 0.7875712250712251, + "grad_norm": 0.4900866746902466, + "learning_rate": 0.00018160176822890842, + "loss": 1.0009, + "step": 4423 + }, + { + "epoch": 0.7877492877492878, + "grad_norm": 0.5928917527198792, + "learning_rate": 0.00018159367651353496, + "loss": 1.0523, + "step": 4424 + }, + { + "epoch": 0.7879273504273504, + "grad_norm": 0.624422013759613, + "learning_rate": 0.0001815855831995029, + "loss": 1.0519, + "step": 4425 + }, + { + "epoch": 0.7881054131054132, + "grad_norm": 0.5140150785446167, + "learning_rate": 0.00018157748828697082, + "loss": 1.048, + "step": 4426 + }, + { + "epoch": 0.7882834757834758, + "grad_norm": 0.47006943821907043, + "learning_rate": 0.00018156939177609732, + "loss": 1.0067, + "step": 4427 + }, + { + "epoch": 0.7884615384615384, + "grad_norm": 0.5178864002227783, + "learning_rate": 0.00018156129366704105, + "loss": 1.0583, + "step": 4428 + }, + { + "epoch": 0.7886396011396012, + "grad_norm": 0.5279985666275024, + "learning_rate": 0.00018155319395996066, + "loss": 1.3023, + "step": 4429 + }, + { + "epoch": 0.7888176638176638, + "grad_norm": 0.5238787531852722, + "learning_rate": 0.00018154509265501482, + "loss": 1.0851, + "step": 4430 + }, + { + "epoch": 0.7889957264957265, + "grad_norm": 0.5914917588233948, + "learning_rate": 0.00018153698975236228, + "loss": 0.9291, + "step": 4431 + }, + { + "epoch": 0.7891737891737892, + "grad_norm": 0.5046082735061646, + "learning_rate": 0.00018152888525216183, + "loss": 0.9951, + "step": 4432 + }, + { + "epoch": 0.7893518518518519, + "grad_norm": 0.5042256116867065, + "learning_rate": 0.00018152077915457225, + "loss": 1.0243, + "step": 4433 + }, + { + "epoch": 0.7895299145299145, + "grad_norm": 0.5950339436531067, + "learning_rate": 0.0001815126714597523, + "loss": 0.9803, + "step": 4434 + }, + { + "epoch": 0.7897079772079773, + "grad_norm": 0.5163764953613281, + "learning_rate": 0.0001815045621678609, + "loss": 1.0353, + "step": 4435 + }, + { + "epoch": 0.7898860398860399, + "grad_norm": 0.5166211128234863, + "learning_rate": 0.00018149645127905691, + "loss": 0.9649, + "step": 4436 + }, + { + "epoch": 0.7900641025641025, + "grad_norm": 0.5239769220352173, + "learning_rate": 0.00018148833879349927, + "loss": 0.9747, + "step": 4437 + }, + { + "epoch": 0.7902421652421653, + "grad_norm": 0.5803237557411194, + "learning_rate": 0.00018148022471134692, + "loss": 1.315, + "step": 4438 + }, + { + "epoch": 0.7904202279202279, + "grad_norm": 0.5141370296478271, + "learning_rate": 0.00018147210903275877, + "loss": 1.0547, + "step": 4439 + }, + { + "epoch": 0.7905982905982906, + "grad_norm": 0.545788586139679, + "learning_rate": 0.00018146399175789394, + "loss": 1.0797, + "step": 4440 + }, + { + "epoch": 0.7907763532763533, + "grad_norm": 0.5273314714431763, + "learning_rate": 0.0001814558728869114, + "loss": 0.7928, + "step": 4441 + }, + { + "epoch": 0.790954415954416, + "grad_norm": 0.4614652693271637, + "learning_rate": 0.00018144775241997024, + "loss": 0.8826, + "step": 4442 + }, + { + "epoch": 0.7911324786324786, + "grad_norm": 0.6203590631484985, + "learning_rate": 0.00018143963035722958, + "loss": 1.2891, + "step": 4443 + }, + { + "epoch": 0.7913105413105413, + "grad_norm": 0.4870408773422241, + "learning_rate": 0.0001814315066988485, + "loss": 1.0717, + "step": 4444 + }, + { + "epoch": 0.791488603988604, + "grad_norm": 0.6468982696533203, + "learning_rate": 0.00018142338144498625, + "loss": 1.3398, + "step": 4445 + }, + { + "epoch": 0.7916666666666666, + "grad_norm": 0.4727918207645416, + "learning_rate": 0.00018141525459580197, + "loss": 1.0195, + "step": 4446 + }, + { + "epoch": 0.7918447293447294, + "grad_norm": 0.5080479979515076, + "learning_rate": 0.0001814071261514549, + "loss": 1.0163, + "step": 4447 + }, + { + "epoch": 0.792022792022792, + "grad_norm": 0.5380908250808716, + "learning_rate": 0.0001813989961121043, + "loss": 1.1673, + "step": 4448 + }, + { + "epoch": 0.7922008547008547, + "grad_norm": 0.5020384192466736, + "learning_rate": 0.00018139086447790945, + "loss": 0.8591, + "step": 4449 + }, + { + "epoch": 0.7923789173789174, + "grad_norm": 0.5279949903488159, + "learning_rate": 0.0001813827312490297, + "loss": 1.1221, + "step": 4450 + }, + { + "epoch": 0.79255698005698, + "grad_norm": 0.6739233732223511, + "learning_rate": 0.00018137459642562437, + "loss": 1.2704, + "step": 4451 + }, + { + "epoch": 0.7927350427350427, + "grad_norm": 0.5112259984016418, + "learning_rate": 0.00018136646000785288, + "loss": 1.1161, + "step": 4452 + }, + { + "epoch": 0.7929131054131054, + "grad_norm": 0.5244031548500061, + "learning_rate": 0.00018135832199587463, + "loss": 0.7866, + "step": 4453 + }, + { + "epoch": 0.7930911680911681, + "grad_norm": 0.5803347229957581, + "learning_rate": 0.0001813501823898491, + "loss": 0.994, + "step": 4454 + }, + { + "epoch": 0.7932692307692307, + "grad_norm": 0.6191152930259705, + "learning_rate": 0.00018134204118993568, + "loss": 1.0725, + "step": 4455 + }, + { + "epoch": 0.7934472934472935, + "grad_norm": 0.549735963344574, + "learning_rate": 0.00018133389839629396, + "loss": 0.9915, + "step": 4456 + }, + { + "epoch": 0.7936253561253561, + "grad_norm": 0.4940381646156311, + "learning_rate": 0.00018132575400908347, + "loss": 1.1815, + "step": 4457 + }, + { + "epoch": 0.7938034188034188, + "grad_norm": 0.5009099245071411, + "learning_rate": 0.00018131760802846377, + "loss": 1.0833, + "step": 4458 + }, + { + "epoch": 0.7939814814814815, + "grad_norm": 0.595853865146637, + "learning_rate": 0.00018130946045459445, + "loss": 1.2774, + "step": 4459 + }, + { + "epoch": 0.7941595441595442, + "grad_norm": 0.534794807434082, + "learning_rate": 0.00018130131128763513, + "loss": 1.0891, + "step": 4460 + }, + { + "epoch": 0.7943376068376068, + "grad_norm": 0.5828582048416138, + "learning_rate": 0.00018129316052774557, + "loss": 1.0786, + "step": 4461 + }, + { + "epoch": 0.7945156695156695, + "grad_norm": 0.4750654697418213, + "learning_rate": 0.00018128500817508533, + "loss": 1.0818, + "step": 4462 + }, + { + "epoch": 0.7946937321937322, + "grad_norm": 0.5626576542854309, + "learning_rate": 0.00018127685422981426, + "loss": 1.0807, + "step": 4463 + }, + { + "epoch": 0.7948717948717948, + "grad_norm": 0.6434760093688965, + "learning_rate": 0.00018126869869209203, + "loss": 1.0908, + "step": 4464 + }, + { + "epoch": 0.7950498575498576, + "grad_norm": 0.5577414631843567, + "learning_rate": 0.00018126054156207853, + "loss": 1.0281, + "step": 4465 + }, + { + "epoch": 0.7952279202279202, + "grad_norm": 0.5001249313354492, + "learning_rate": 0.00018125238283993347, + "loss": 0.9083, + "step": 4466 + }, + { + "epoch": 0.7954059829059829, + "grad_norm": 0.5298314690589905, + "learning_rate": 0.00018124422252581676, + "loss": 0.971, + "step": 4467 + }, + { + "epoch": 0.7955840455840456, + "grad_norm": 0.4872737228870392, + "learning_rate": 0.00018123606061988832, + "loss": 1.0515, + "step": 4468 + }, + { + "epoch": 0.7957621082621082, + "grad_norm": 0.5895398259162903, + "learning_rate": 0.00018122789712230798, + "loss": 1.0771, + "step": 4469 + }, + { + "epoch": 0.7959401709401709, + "grad_norm": 0.5212514996528625, + "learning_rate": 0.00018121973203323577, + "loss": 1.0365, + "step": 4470 + }, + { + "epoch": 0.7961182336182336, + "grad_norm": 0.4679451584815979, + "learning_rate": 0.0001812115653528316, + "loss": 0.9445, + "step": 4471 + }, + { + "epoch": 0.7962962962962963, + "grad_norm": 0.5852653980255127, + "learning_rate": 0.00018120339708125552, + "loss": 1.1781, + "step": 4472 + }, + { + "epoch": 0.7964743589743589, + "grad_norm": 0.6081342697143555, + "learning_rate": 0.00018119522721866756, + "loss": 1.3881, + "step": 4473 + }, + { + "epoch": 0.7966524216524217, + "grad_norm": 0.5254155993461609, + "learning_rate": 0.00018118705576522777, + "loss": 1.2198, + "step": 4474 + }, + { + "epoch": 0.7968304843304843, + "grad_norm": 0.5959419012069702, + "learning_rate": 0.00018117888272109632, + "loss": 1.0922, + "step": 4475 + }, + { + "epoch": 0.7970085470085471, + "grad_norm": 0.6243147253990173, + "learning_rate": 0.0001811707080864333, + "loss": 1.1782, + "step": 4476 + }, + { + "epoch": 0.7971866096866097, + "grad_norm": 0.5336906909942627, + "learning_rate": 0.0001811625318613988, + "loss": 1.167, + "step": 4477 + }, + { + "epoch": 0.7973646723646723, + "grad_norm": 0.5287907719612122, + "learning_rate": 0.00018115435404615315, + "loss": 0.9923, + "step": 4478 + }, + { + "epoch": 0.7975427350427351, + "grad_norm": 0.48941442370414734, + "learning_rate": 0.0001811461746408565, + "loss": 0.863, + "step": 4479 + }, + { + "epoch": 0.7977207977207977, + "grad_norm": 0.48465651273727417, + "learning_rate": 0.0001811379936456691, + "loss": 1.147, + "step": 4480 + }, + { + "epoch": 0.7978988603988604, + "grad_norm": 0.5676067471504211, + "learning_rate": 0.0001811298110607513, + "loss": 1.3121, + "step": 4481 + }, + { + "epoch": 0.7980769230769231, + "grad_norm": 0.4894018769264221, + "learning_rate": 0.00018112162688626337, + "loss": 1.1831, + "step": 4482 + }, + { + "epoch": 0.7982549857549858, + "grad_norm": 0.5626382827758789, + "learning_rate": 0.0001811134411223657, + "loss": 1.1977, + "step": 4483 + }, + { + "epoch": 0.7984330484330484, + "grad_norm": 0.564119815826416, + "learning_rate": 0.00018110525376921862, + "loss": 1.2686, + "step": 4484 + }, + { + "epoch": 0.7986111111111112, + "grad_norm": 0.6385740041732788, + "learning_rate": 0.00018109706482698256, + "loss": 1.2418, + "step": 4485 + }, + { + "epoch": 0.7987891737891738, + "grad_norm": 0.5550164580345154, + "learning_rate": 0.00018108887429581802, + "loss": 1.081, + "step": 4486 + }, + { + "epoch": 0.7989672364672364, + "grad_norm": 0.5583973526954651, + "learning_rate": 0.00018108068217588544, + "loss": 1.1757, + "step": 4487 + }, + { + "epoch": 0.7991452991452992, + "grad_norm": 0.5533342957496643, + "learning_rate": 0.00018107248846734527, + "loss": 1.1947, + "step": 4488 + }, + { + "epoch": 0.7993233618233618, + "grad_norm": 0.5291479229927063, + "learning_rate": 0.00018106429317035815, + "loss": 1.2769, + "step": 4489 + }, + { + "epoch": 0.7995014245014245, + "grad_norm": 0.4680160582065582, + "learning_rate": 0.00018105609628508458, + "loss": 0.7059, + "step": 4490 + }, + { + "epoch": 0.7996794871794872, + "grad_norm": 0.5364881157875061, + "learning_rate": 0.00018104789781168517, + "loss": 1.0566, + "step": 4491 + }, + { + "epoch": 0.7998575498575499, + "grad_norm": 0.5917307734489441, + "learning_rate": 0.0001810396977503206, + "loss": 1.2263, + "step": 4492 + }, + { + "epoch": 0.8000356125356125, + "grad_norm": 0.6013199090957642, + "learning_rate": 0.0001810314961011515, + "loss": 1.2053, + "step": 4493 + }, + { + "epoch": 0.8002136752136753, + "grad_norm": 0.6005663275718689, + "learning_rate": 0.0001810232928643385, + "loss": 1.2241, + "step": 4494 + }, + { + "epoch": 0.8003917378917379, + "grad_norm": 0.49207603931427, + "learning_rate": 0.00018101508804004246, + "loss": 1.0661, + "step": 4495 + }, + { + "epoch": 0.8005698005698005, + "grad_norm": 0.4834063947200775, + "learning_rate": 0.00018100688162842401, + "loss": 1.1745, + "step": 4496 + }, + { + "epoch": 0.8007478632478633, + "grad_norm": 0.5347156524658203, + "learning_rate": 0.000180998673629644, + "loss": 1.0679, + "step": 4497 + }, + { + "epoch": 0.8009259259259259, + "grad_norm": 0.5815600156784058, + "learning_rate": 0.00018099046404386327, + "loss": 1.2652, + "step": 4498 + }, + { + "epoch": 0.8011039886039886, + "grad_norm": 0.5291135311126709, + "learning_rate": 0.00018098225287124263, + "loss": 1.2072, + "step": 4499 + }, + { + "epoch": 0.8012820512820513, + "grad_norm": 0.5779497027397156, + "learning_rate": 0.000180974040111943, + "loss": 1.3277, + "step": 4500 + }, + { + "epoch": 0.801460113960114, + "grad_norm": 0.44566696882247925, + "learning_rate": 0.0001809658257661252, + "loss": 0.7702, + "step": 4501 + }, + { + "epoch": 0.8016381766381766, + "grad_norm": 0.5407577753067017, + "learning_rate": 0.00018095760983395027, + "loss": 1.2894, + "step": 4502 + }, + { + "epoch": 0.8018162393162394, + "grad_norm": 0.4771903455257416, + "learning_rate": 0.00018094939231557916, + "loss": 1.045, + "step": 4503 + }, + { + "epoch": 0.801994301994302, + "grad_norm": 0.5970945358276367, + "learning_rate": 0.00018094117321117286, + "loss": 1.2059, + "step": 4504 + }, + { + "epoch": 0.8021723646723646, + "grad_norm": 0.4959338903427124, + "learning_rate": 0.0001809329525208924, + "loss": 1.155, + "step": 4505 + }, + { + "epoch": 0.8023504273504274, + "grad_norm": 0.5142548084259033, + "learning_rate": 0.00018092473024489887, + "loss": 0.9413, + "step": 4506 + }, + { + "epoch": 0.80252849002849, + "grad_norm": 0.5336433053016663, + "learning_rate": 0.00018091650638335334, + "loss": 1.0699, + "step": 4507 + }, + { + "epoch": 0.8027065527065527, + "grad_norm": 0.47770628333091736, + "learning_rate": 0.00018090828093641698, + "loss": 1.1515, + "step": 4508 + }, + { + "epoch": 0.8028846153846154, + "grad_norm": 0.5443438291549683, + "learning_rate": 0.00018090005390425091, + "loss": 1.189, + "step": 4509 + }, + { + "epoch": 0.8030626780626781, + "grad_norm": 0.523179829120636, + "learning_rate": 0.00018089182528701632, + "loss": 1.1272, + "step": 4510 + }, + { + "epoch": 0.8032407407407407, + "grad_norm": 0.49628451466560364, + "learning_rate": 0.00018088359508487448, + "loss": 0.9754, + "step": 4511 + }, + { + "epoch": 0.8034188034188035, + "grad_norm": 0.5933086276054382, + "learning_rate": 0.00018087536329798663, + "loss": 1.2111, + "step": 4512 + }, + { + "epoch": 0.8035968660968661, + "grad_norm": 0.4565310776233673, + "learning_rate": 0.00018086712992651402, + "loss": 0.7729, + "step": 4513 + }, + { + "epoch": 0.8037749287749287, + "grad_norm": 0.5013461112976074, + "learning_rate": 0.00018085889497061798, + "loss": 1.2178, + "step": 4514 + }, + { + "epoch": 0.8039529914529915, + "grad_norm": 0.5170024633407593, + "learning_rate": 0.00018085065843045987, + "loss": 0.9181, + "step": 4515 + }, + { + "epoch": 0.8041310541310541, + "grad_norm": 0.583363950252533, + "learning_rate": 0.00018084242030620104, + "loss": 1.1542, + "step": 4516 + }, + { + "epoch": 0.8043091168091168, + "grad_norm": 0.46835777163505554, + "learning_rate": 0.00018083418059800297, + "loss": 0.8954, + "step": 4517 + }, + { + "epoch": 0.8044871794871795, + "grad_norm": 0.5145657062530518, + "learning_rate": 0.000180825939306027, + "loss": 1.0417, + "step": 4518 + }, + { + "epoch": 0.8046652421652422, + "grad_norm": 0.47216105461120605, + "learning_rate": 0.00018081769643043467, + "loss": 0.9516, + "step": 4519 + }, + { + "epoch": 0.8048433048433048, + "grad_norm": 0.5059915781021118, + "learning_rate": 0.0001808094519713875, + "loss": 1.1643, + "step": 4520 + }, + { + "epoch": 0.8050213675213675, + "grad_norm": 0.5406439900398254, + "learning_rate": 0.00018080120592904692, + "loss": 1.2038, + "step": 4521 + }, + { + "epoch": 0.8051994301994302, + "grad_norm": 0.6123420000076294, + "learning_rate": 0.0001807929583035746, + "loss": 1.4004, + "step": 4522 + }, + { + "epoch": 0.8053774928774928, + "grad_norm": 0.49699845910072327, + "learning_rate": 0.00018078470909513208, + "loss": 1.0347, + "step": 4523 + }, + { + "epoch": 0.8055555555555556, + "grad_norm": 0.5369421243667603, + "learning_rate": 0.000180776458303881, + "loss": 1.0418, + "step": 4524 + }, + { + "epoch": 0.8057336182336182, + "grad_norm": 0.5407396554946899, + "learning_rate": 0.00018076820592998301, + "loss": 0.9546, + "step": 4525 + }, + { + "epoch": 0.8059116809116809, + "grad_norm": 0.5749752521514893, + "learning_rate": 0.00018075995197359984, + "loss": 1.1438, + "step": 4526 + }, + { + "epoch": 0.8060897435897436, + "grad_norm": 0.5523102283477783, + "learning_rate": 0.00018075169643489317, + "loss": 1.1312, + "step": 4527 + }, + { + "epoch": 0.8062678062678063, + "grad_norm": 0.5767508149147034, + "learning_rate": 0.00018074343931402472, + "loss": 1.1951, + "step": 4528 + }, + { + "epoch": 0.8064458689458689, + "grad_norm": 0.5262924432754517, + "learning_rate": 0.00018073518061115633, + "loss": 1.1985, + "step": 4529 + }, + { + "epoch": 0.8066239316239316, + "grad_norm": 0.4742378294467926, + "learning_rate": 0.0001807269203264498, + "loss": 1.0126, + "step": 4530 + }, + { + "epoch": 0.8068019943019943, + "grad_norm": 0.5190158486366272, + "learning_rate": 0.00018071865846006692, + "loss": 0.9985, + "step": 4531 + }, + { + "epoch": 0.8069800569800569, + "grad_norm": 0.5910618305206299, + "learning_rate": 0.00018071039501216964, + "loss": 1.2776, + "step": 4532 + }, + { + "epoch": 0.8071581196581197, + "grad_norm": 0.5363098382949829, + "learning_rate": 0.00018070212998291983, + "loss": 1.3346, + "step": 4533 + }, + { + "epoch": 0.8073361823361823, + "grad_norm": 0.47711408138275146, + "learning_rate": 0.0001806938633724794, + "loss": 1.04, + "step": 4534 + }, + { + "epoch": 0.8075142450142451, + "grad_norm": 0.5092964172363281, + "learning_rate": 0.0001806855951810104, + "loss": 1.1409, + "step": 4535 + }, + { + "epoch": 0.8076923076923077, + "grad_norm": 0.5828777551651001, + "learning_rate": 0.00018067732540867472, + "loss": 1.3048, + "step": 4536 + }, + { + "epoch": 0.8078703703703703, + "grad_norm": 0.5779826045036316, + "learning_rate": 0.00018066905405563445, + "loss": 1.1599, + "step": 4537 + }, + { + "epoch": 0.8080484330484331, + "grad_norm": 0.49908435344696045, + "learning_rate": 0.00018066078112205167, + "loss": 1.1502, + "step": 4538 + }, + { + "epoch": 0.8082264957264957, + "grad_norm": 0.4772704839706421, + "learning_rate": 0.0001806525066080884, + "loss": 0.7925, + "step": 4539 + }, + { + "epoch": 0.8084045584045584, + "grad_norm": 0.4298383295536041, + "learning_rate": 0.00018064423051390683, + "loss": 0.7322, + "step": 4540 + }, + { + "epoch": 0.8085826210826211, + "grad_norm": 0.49349579215049744, + "learning_rate": 0.0001806359528396691, + "loss": 1.0021, + "step": 4541 + }, + { + "epoch": 0.8087606837606838, + "grad_norm": 0.4698609411716461, + "learning_rate": 0.00018062767358553735, + "loss": 0.9751, + "step": 4542 + }, + { + "epoch": 0.8089387464387464, + "grad_norm": 0.4949014186859131, + "learning_rate": 0.00018061939275167385, + "loss": 0.9553, + "step": 4543 + }, + { + "epoch": 0.8091168091168092, + "grad_norm": 0.5604463815689087, + "learning_rate": 0.0001806111103382408, + "loss": 0.9894, + "step": 4544 + }, + { + "epoch": 0.8092948717948718, + "grad_norm": 0.5761561989784241, + "learning_rate": 0.00018060282634540053, + "loss": 1.258, + "step": 4545 + }, + { + "epoch": 0.8094729344729344, + "grad_norm": 0.5239115357398987, + "learning_rate": 0.00018059454077331527, + "loss": 0.9189, + "step": 4546 + }, + { + "epoch": 0.8096509971509972, + "grad_norm": 0.47902220487594604, + "learning_rate": 0.00018058625362214742, + "loss": 1.0389, + "step": 4547 + }, + { + "epoch": 0.8098290598290598, + "grad_norm": 0.6274173259735107, + "learning_rate": 0.00018057796489205936, + "loss": 1.3368, + "step": 4548 + }, + { + "epoch": 0.8100071225071225, + "grad_norm": 0.5789401531219482, + "learning_rate": 0.00018056967458321345, + "loss": 1.1473, + "step": 4549 + }, + { + "epoch": 0.8101851851851852, + "grad_norm": 0.5850043296813965, + "learning_rate": 0.0001805613826957721, + "loss": 1.2224, + "step": 4550 + }, + { + "epoch": 0.8103632478632479, + "grad_norm": 0.6310738921165466, + "learning_rate": 0.00018055308922989788, + "loss": 1.0707, + "step": 4551 + }, + { + "epoch": 0.8105413105413105, + "grad_norm": 0.5198429822921753, + "learning_rate": 0.00018054479418575317, + "loss": 0.8984, + "step": 4552 + }, + { + "epoch": 0.8107193732193733, + "grad_norm": 0.5757743120193481, + "learning_rate": 0.00018053649756350054, + "loss": 1.2007, + "step": 4553 + }, + { + "epoch": 0.8108974358974359, + "grad_norm": 0.5109567642211914, + "learning_rate": 0.0001805281993633025, + "loss": 1.0696, + "step": 4554 + }, + { + "epoch": 0.8110754985754985, + "grad_norm": 0.5030225515365601, + "learning_rate": 0.00018051989958532173, + "loss": 0.9667, + "step": 4555 + }, + { + "epoch": 0.8112535612535613, + "grad_norm": 0.5291743874549866, + "learning_rate": 0.00018051159822972079, + "loss": 1.0219, + "step": 4556 + }, + { + "epoch": 0.8114316239316239, + "grad_norm": 0.5874896049499512, + "learning_rate": 0.00018050329529666233, + "loss": 0.8589, + "step": 4557 + }, + { + "epoch": 0.8116096866096866, + "grad_norm": 0.673284113407135, + "learning_rate": 0.000180494990786309, + "loss": 1.1902, + "step": 4558 + }, + { + "epoch": 0.8117877492877493, + "grad_norm": 0.4742524027824402, + "learning_rate": 0.00018048668469882354, + "loss": 1.0578, + "step": 4559 + }, + { + "epoch": 0.811965811965812, + "grad_norm": 0.5519167184829712, + "learning_rate": 0.0001804783770343687, + "loss": 1.083, + "step": 4560 + }, + { + "epoch": 0.8121438746438746, + "grad_norm": 0.5669941306114197, + "learning_rate": 0.00018047006779310727, + "loss": 1.0784, + "step": 4561 + }, + { + "epoch": 0.8123219373219374, + "grad_norm": 0.512759804725647, + "learning_rate": 0.000180461756975202, + "loss": 1.0361, + "step": 4562 + }, + { + "epoch": 0.8125, + "grad_norm": 0.5721749067306519, + "learning_rate": 0.00018045344458081575, + "loss": 1.0246, + "step": 4563 + }, + { + "epoch": 0.8126780626780626, + "grad_norm": 0.566430389881134, + "learning_rate": 0.00018044513061011137, + "loss": 1.1452, + "step": 4564 + }, + { + "epoch": 0.8128561253561254, + "grad_norm": 0.49391916394233704, + "learning_rate": 0.00018043681506325177, + "loss": 0.89, + "step": 4565 + }, + { + "epoch": 0.813034188034188, + "grad_norm": 0.5379437804222107, + "learning_rate": 0.00018042849794039988, + "loss": 1.1289, + "step": 4566 + }, + { + "epoch": 0.8132122507122507, + "grad_norm": 0.5667982697486877, + "learning_rate": 0.00018042017924171865, + "loss": 1.1596, + "step": 4567 + }, + { + "epoch": 0.8133903133903134, + "grad_norm": 0.6214209794998169, + "learning_rate": 0.00018041185896737109, + "loss": 1.0622, + "step": 4568 + }, + { + "epoch": 0.8135683760683761, + "grad_norm": 0.5442491173744202, + "learning_rate": 0.00018040353711752015, + "loss": 1.0536, + "step": 4569 + }, + { + "epoch": 0.8137464387464387, + "grad_norm": 0.5266172885894775, + "learning_rate": 0.00018039521369232894, + "loss": 1.0576, + "step": 4570 + }, + { + "epoch": 0.8139245014245015, + "grad_norm": 0.6057912111282349, + "learning_rate": 0.00018038688869196053, + "loss": 1.3067, + "step": 4571 + }, + { + "epoch": 0.8141025641025641, + "grad_norm": 0.489869087934494, + "learning_rate": 0.00018037856211657803, + "loss": 1.0279, + "step": 4572 + }, + { + "epoch": 0.8142806267806267, + "grad_norm": 0.5497978329658508, + "learning_rate": 0.00018037023396634457, + "loss": 1.1568, + "step": 4573 + }, + { + "epoch": 0.8144586894586895, + "grad_norm": 0.5243251919746399, + "learning_rate": 0.0001803619042414233, + "loss": 0.9767, + "step": 4574 + }, + { + "epoch": 0.8146367521367521, + "grad_norm": 0.503032922744751, + "learning_rate": 0.0001803535729419775, + "loss": 1.065, + "step": 4575 + }, + { + "epoch": 0.8148148148148148, + "grad_norm": 0.49955418705940247, + "learning_rate": 0.00018034524006817034, + "loss": 1.2752, + "step": 4576 + }, + { + "epoch": 0.8149928774928775, + "grad_norm": 0.5746406316757202, + "learning_rate": 0.00018033690562016508, + "loss": 1.098, + "step": 4577 + }, + { + "epoch": 0.8151709401709402, + "grad_norm": 0.5224192142486572, + "learning_rate": 0.00018032856959812507, + "loss": 1.1284, + "step": 4578 + }, + { + "epoch": 0.8153490028490028, + "grad_norm": 0.5484535694122314, + "learning_rate": 0.00018032023200221362, + "loss": 0.9182, + "step": 4579 + }, + { + "epoch": 0.8155270655270656, + "grad_norm": 0.5003355741500854, + "learning_rate": 0.00018031189283259405, + "loss": 1.136, + "step": 4580 + }, + { + "epoch": 0.8157051282051282, + "grad_norm": 0.5395768284797668, + "learning_rate": 0.00018030355208942977, + "loss": 1.2349, + "step": 4581 + }, + { + "epoch": 0.8158831908831908, + "grad_norm": 0.561966598033905, + "learning_rate": 0.0001802952097728842, + "loss": 0.999, + "step": 4582 + }, + { + "epoch": 0.8160612535612536, + "grad_norm": 0.4886479675769806, + "learning_rate": 0.00018028686588312083, + "loss": 0.9165, + "step": 4583 + }, + { + "epoch": 0.8162393162393162, + "grad_norm": 0.4769509732723236, + "learning_rate": 0.00018027852042030307, + "loss": 1.1377, + "step": 4584 + }, + { + "epoch": 0.8164173789173789, + "grad_norm": 0.4723633825778961, + "learning_rate": 0.00018027017338459448, + "loss": 1.0274, + "step": 4585 + }, + { + "epoch": 0.8165954415954416, + "grad_norm": 0.5773285627365112, + "learning_rate": 0.00018026182477615859, + "loss": 1.1468, + "step": 4586 + }, + { + "epoch": 0.8167735042735043, + "grad_norm": 0.5529203414916992, + "learning_rate": 0.00018025347459515895, + "loss": 1.0815, + "step": 4587 + }, + { + "epoch": 0.8169515669515669, + "grad_norm": 0.5449469685554504, + "learning_rate": 0.00018024512284175922, + "loss": 1.1637, + "step": 4588 + }, + { + "epoch": 0.8171296296296297, + "grad_norm": 0.5155341625213623, + "learning_rate": 0.00018023676951612298, + "loss": 1.1842, + "step": 4589 + }, + { + "epoch": 0.8173076923076923, + "grad_norm": 0.5569564700126648, + "learning_rate": 0.00018022841461841393, + "loss": 0.9254, + "step": 4590 + }, + { + "epoch": 0.8174857549857549, + "grad_norm": 0.45203131437301636, + "learning_rate": 0.00018022005814879573, + "loss": 0.9561, + "step": 4591 + }, + { + "epoch": 0.8176638176638177, + "grad_norm": 0.5735056400299072, + "learning_rate": 0.00018021170010743218, + "loss": 1.1402, + "step": 4592 + }, + { + "epoch": 0.8178418803418803, + "grad_norm": 0.6075260043144226, + "learning_rate": 0.00018020334049448697, + "loss": 0.8601, + "step": 4593 + }, + { + "epoch": 0.8180199430199431, + "grad_norm": 0.522682785987854, + "learning_rate": 0.0001801949793101239, + "loss": 1.0088, + "step": 4594 + }, + { + "epoch": 0.8181980056980057, + "grad_norm": 0.5648437142372131, + "learning_rate": 0.00018018661655450682, + "loss": 0.8359, + "step": 4595 + }, + { + "epoch": 0.8183760683760684, + "grad_norm": 0.5406472086906433, + "learning_rate": 0.00018017825222779954, + "loss": 1.1553, + "step": 4596 + }, + { + "epoch": 0.8185541310541311, + "grad_norm": 0.4917788803577423, + "learning_rate": 0.000180169886330166, + "loss": 1.2198, + "step": 4597 + }, + { + "epoch": 0.8187321937321937, + "grad_norm": 0.6293069124221802, + "learning_rate": 0.00018016151886177004, + "loss": 1.0245, + "step": 4598 + }, + { + "epoch": 0.8189102564102564, + "grad_norm": 0.47277843952178955, + "learning_rate": 0.00018015314982277564, + "loss": 1.1141, + "step": 4599 + }, + { + "epoch": 0.8190883190883191, + "grad_norm": 0.6132395267486572, + "learning_rate": 0.0001801447792133468, + "loss": 1.1227, + "step": 4600 + }, + { + "epoch": 0.8192663817663818, + "grad_norm": 0.46839597821235657, + "learning_rate": 0.00018013640703364747, + "loss": 0.9239, + "step": 4601 + }, + { + "epoch": 0.8194444444444444, + "grad_norm": 0.5055009722709656, + "learning_rate": 0.00018012803328384171, + "loss": 0.8486, + "step": 4602 + }, + { + "epoch": 0.8196225071225072, + "grad_norm": 0.5094841718673706, + "learning_rate": 0.00018011965796409362, + "loss": 0.9969, + "step": 4603 + }, + { + "epoch": 0.8198005698005698, + "grad_norm": 0.6177363395690918, + "learning_rate": 0.00018011128107456726, + "loss": 1.242, + "step": 4604 + }, + { + "epoch": 0.8199786324786325, + "grad_norm": 0.5280042290687561, + "learning_rate": 0.00018010290261542676, + "loss": 1.1569, + "step": 4605 + }, + { + "epoch": 0.8201566951566952, + "grad_norm": 0.5259367227554321, + "learning_rate": 0.00018009452258683625, + "loss": 0.9993, + "step": 4606 + }, + { + "epoch": 0.8203347578347578, + "grad_norm": 0.464469850063324, + "learning_rate": 0.00018008614098896, + "loss": 1.0288, + "step": 4607 + }, + { + "epoch": 0.8205128205128205, + "grad_norm": 0.6136324405670166, + "learning_rate": 0.00018007775782196214, + "loss": 1.1541, + "step": 4608 + }, + { + "epoch": 0.8206908831908832, + "grad_norm": 0.5376590490341187, + "learning_rate": 0.000180069373086007, + "loss": 1.0624, + "step": 4609 + }, + { + "epoch": 0.8208689458689459, + "grad_norm": 0.662916362285614, + "learning_rate": 0.0001800609867812588, + "loss": 1.1502, + "step": 4610 + }, + { + "epoch": 0.8210470085470085, + "grad_norm": 0.5153383612632751, + "learning_rate": 0.00018005259890788188, + "loss": 0.9789, + "step": 4611 + }, + { + "epoch": 0.8212250712250713, + "grad_norm": 0.5042359232902527, + "learning_rate": 0.00018004420946604057, + "loss": 0.9585, + "step": 4612 + }, + { + "epoch": 0.8214031339031339, + "grad_norm": 0.5395993590354919, + "learning_rate": 0.00018003581845589927, + "loss": 1.159, + "step": 4613 + }, + { + "epoch": 0.8215811965811965, + "grad_norm": 0.5561928749084473, + "learning_rate": 0.00018002742587762237, + "loss": 1.1604, + "step": 4614 + }, + { + "epoch": 0.8217592592592593, + "grad_norm": 0.5602710843086243, + "learning_rate": 0.00018001903173137432, + "loss": 0.9922, + "step": 4615 + }, + { + "epoch": 0.8219373219373219, + "grad_norm": 0.5529088377952576, + "learning_rate": 0.00018001063601731955, + "loss": 1.0943, + "step": 4616 + }, + { + "epoch": 0.8221153846153846, + "grad_norm": 0.5156456828117371, + "learning_rate": 0.00018000223873562254, + "loss": 1.1399, + "step": 4617 + }, + { + "epoch": 0.8222934472934473, + "grad_norm": 0.4868306517601013, + "learning_rate": 0.0001799938398864479, + "loss": 1.0692, + "step": 4618 + }, + { + "epoch": 0.82247150997151, + "grad_norm": 0.5372915267944336, + "learning_rate": 0.0001799854394699601, + "loss": 1.2675, + "step": 4619 + }, + { + "epoch": 0.8226495726495726, + "grad_norm": 0.6101839542388916, + "learning_rate": 0.0001799770374863238, + "loss": 0.9586, + "step": 4620 + }, + { + "epoch": 0.8228276353276354, + "grad_norm": 0.5034586787223816, + "learning_rate": 0.00017996863393570357, + "loss": 1.0885, + "step": 4621 + }, + { + "epoch": 0.823005698005698, + "grad_norm": 0.5608823299407959, + "learning_rate": 0.0001799602288182641, + "loss": 1.0002, + "step": 4622 + }, + { + "epoch": 0.8231837606837606, + "grad_norm": 0.5700048208236694, + "learning_rate": 0.00017995182213417, + "loss": 1.1484, + "step": 4623 + }, + { + "epoch": 0.8233618233618234, + "grad_norm": 0.5283229351043701, + "learning_rate": 0.00017994341388358608, + "loss": 1.0744, + "step": 4624 + }, + { + "epoch": 0.823539886039886, + "grad_norm": 0.5215758681297302, + "learning_rate": 0.00017993500406667703, + "loss": 1.2686, + "step": 4625 + }, + { + "epoch": 0.8237179487179487, + "grad_norm": 0.528883159160614, + "learning_rate": 0.0001799265926836076, + "loss": 1.1393, + "step": 4626 + }, + { + "epoch": 0.8238960113960114, + "grad_norm": 0.5589834451675415, + "learning_rate": 0.00017991817973454265, + "loss": 1.1744, + "step": 4627 + }, + { + "epoch": 0.8240740740740741, + "grad_norm": 0.49817174673080444, + "learning_rate": 0.00017990976521964697, + "loss": 1.0544, + "step": 4628 + }, + { + "epoch": 0.8242521367521367, + "grad_norm": 0.613961398601532, + "learning_rate": 0.00017990134913908542, + "loss": 1.0951, + "step": 4629 + }, + { + "epoch": 0.8244301994301995, + "grad_norm": 0.47278255224227905, + "learning_rate": 0.00017989293149302295, + "loss": 0.9742, + "step": 4630 + }, + { + "epoch": 0.8246082621082621, + "grad_norm": 0.49807092547416687, + "learning_rate": 0.00017988451228162443, + "loss": 1.0985, + "step": 4631 + }, + { + "epoch": 0.8247863247863247, + "grad_norm": 0.5624374747276306, + "learning_rate": 0.00017987609150505485, + "loss": 1.2446, + "step": 4632 + }, + { + "epoch": 0.8249643874643875, + "grad_norm": 0.4863535761833191, + "learning_rate": 0.00017986766916347916, + "loss": 1.0239, + "step": 4633 + }, + { + "epoch": 0.8251424501424501, + "grad_norm": 0.679585874080658, + "learning_rate": 0.00017985924525706245, + "loss": 1.1698, + "step": 4634 + }, + { + "epoch": 0.8253205128205128, + "grad_norm": 0.5545455813407898, + "learning_rate": 0.00017985081978596967, + "loss": 1.0926, + "step": 4635 + }, + { + "epoch": 0.8254985754985755, + "grad_norm": 0.5303109288215637, + "learning_rate": 0.000179842392750366, + "loss": 1.0978, + "step": 4636 + }, + { + "epoch": 0.8256766381766382, + "grad_norm": 0.6053299307823181, + "learning_rate": 0.00017983396415041644, + "loss": 1.0596, + "step": 4637 + }, + { + "epoch": 0.8258547008547008, + "grad_norm": 0.5241885185241699, + "learning_rate": 0.00017982553398628625, + "loss": 0.8541, + "step": 4638 + }, + { + "epoch": 0.8260327635327636, + "grad_norm": 0.5934443473815918, + "learning_rate": 0.00017981710225814052, + "loss": 1.145, + "step": 4639 + }, + { + "epoch": 0.8262108262108262, + "grad_norm": 0.5341619849205017, + "learning_rate": 0.00017980866896614447, + "loss": 1.0745, + "step": 4640 + }, + { + "epoch": 0.8263888888888888, + "grad_norm": 0.6732913851737976, + "learning_rate": 0.00017980023411046336, + "loss": 1.0775, + "step": 4641 + }, + { + "epoch": 0.8265669515669516, + "grad_norm": 0.5134359002113342, + "learning_rate": 0.0001797917976912624, + "loss": 1.0298, + "step": 4642 + }, + { + "epoch": 0.8267450142450142, + "grad_norm": 0.5234783887863159, + "learning_rate": 0.00017978335970870698, + "loss": 1.1069, + "step": 4643 + }, + { + "epoch": 0.8269230769230769, + "grad_norm": 0.4776439964771271, + "learning_rate": 0.00017977492016296232, + "loss": 0.6367, + "step": 4644 + }, + { + "epoch": 0.8271011396011396, + "grad_norm": 0.53763347864151, + "learning_rate": 0.0001797664790541938, + "loss": 1.1356, + "step": 4645 + }, + { + "epoch": 0.8272792022792023, + "grad_norm": 0.5082212686538696, + "learning_rate": 0.00017975803638256682, + "loss": 0.7873, + "step": 4646 + }, + { + "epoch": 0.8274572649572649, + "grad_norm": 0.5156424641609192, + "learning_rate": 0.00017974959214824685, + "loss": 1.084, + "step": 4647 + }, + { + "epoch": 0.8276353276353277, + "grad_norm": 0.5275198817253113, + "learning_rate": 0.00017974114635139926, + "loss": 1.1219, + "step": 4648 + }, + { + "epoch": 0.8278133903133903, + "grad_norm": 0.5548223257064819, + "learning_rate": 0.00017973269899218956, + "loss": 1.0808, + "step": 4649 + }, + { + "epoch": 0.8279914529914529, + "grad_norm": 0.535347580909729, + "learning_rate": 0.00017972425007078323, + "loss": 1.1211, + "step": 4650 + }, + { + "epoch": 0.8281695156695157, + "grad_norm": 0.5299580693244934, + "learning_rate": 0.00017971579958734587, + "loss": 0.9911, + "step": 4651 + }, + { + "epoch": 0.8283475783475783, + "grad_norm": 0.4863550066947937, + "learning_rate": 0.000179707347542043, + "loss": 0.9122, + "step": 4652 + }, + { + "epoch": 0.8285256410256411, + "grad_norm": 0.5284972190856934, + "learning_rate": 0.00017969889393504022, + "loss": 1.0424, + "step": 4653 + }, + { + "epoch": 0.8287037037037037, + "grad_norm": 0.5305661559104919, + "learning_rate": 0.00017969043876650317, + "loss": 1.1122, + "step": 4654 + }, + { + "epoch": 0.8288817663817664, + "grad_norm": 0.5645657777786255, + "learning_rate": 0.00017968198203659755, + "loss": 1.2195, + "step": 4655 + }, + { + "epoch": 0.8290598290598291, + "grad_norm": 0.521649181842804, + "learning_rate": 0.000179673523745489, + "loss": 1.2684, + "step": 4656 + }, + { + "epoch": 0.8292378917378918, + "grad_norm": 0.5984422564506531, + "learning_rate": 0.00017966506389334322, + "loss": 0.9894, + "step": 4657 + }, + { + "epoch": 0.8294159544159544, + "grad_norm": 0.5318729281425476, + "learning_rate": 0.00017965660248032603, + "loss": 1.2929, + "step": 4658 + }, + { + "epoch": 0.8295940170940171, + "grad_norm": 0.4666081368923187, + "learning_rate": 0.0001796481395066032, + "loss": 0.9646, + "step": 4659 + }, + { + "epoch": 0.8297720797720798, + "grad_norm": 0.5780388116836548, + "learning_rate": 0.00017963967497234054, + "loss": 1.1043, + "step": 4660 + }, + { + "epoch": 0.8299501424501424, + "grad_norm": 0.44089245796203613, + "learning_rate": 0.00017963120887770387, + "loss": 0.8932, + "step": 4661 + }, + { + "epoch": 0.8301282051282052, + "grad_norm": 0.5198349356651306, + "learning_rate": 0.0001796227412228591, + "loss": 0.9378, + "step": 4662 + }, + { + "epoch": 0.8303062678062678, + "grad_norm": 0.5298343896865845, + "learning_rate": 0.00017961427200797206, + "loss": 1.0272, + "step": 4663 + }, + { + "epoch": 0.8304843304843305, + "grad_norm": 0.5087099671363831, + "learning_rate": 0.0001796058012332088, + "loss": 0.989, + "step": 4664 + }, + { + "epoch": 0.8306623931623932, + "grad_norm": 0.504228949546814, + "learning_rate": 0.0001795973288987352, + "loss": 1.0134, + "step": 4665 + }, + { + "epoch": 0.8308404558404558, + "grad_norm": 0.6788033843040466, + "learning_rate": 0.00017958885500471728, + "loss": 0.8856, + "step": 4666 + }, + { + "epoch": 0.8310185185185185, + "grad_norm": 0.5166172385215759, + "learning_rate": 0.00017958037955132113, + "loss": 0.8711, + "step": 4667 + }, + { + "epoch": 0.8311965811965812, + "grad_norm": 0.5712400078773499, + "learning_rate": 0.00017957190253871272, + "loss": 1.0418, + "step": 4668 + }, + { + "epoch": 0.8313746438746439, + "grad_norm": 0.5531231164932251, + "learning_rate": 0.0001795634239670582, + "loss": 0.9021, + "step": 4669 + }, + { + "epoch": 0.8315527065527065, + "grad_norm": 0.6165615916252136, + "learning_rate": 0.00017955494383652365, + "loss": 1.0927, + "step": 4670 + }, + { + "epoch": 0.8317307692307693, + "grad_norm": 0.5920368432998657, + "learning_rate": 0.00017954646214727525, + "loss": 1.231, + "step": 4671 + }, + { + "epoch": 0.8319088319088319, + "grad_norm": 0.5037244558334351, + "learning_rate": 0.00017953797889947915, + "loss": 0.85, + "step": 4672 + }, + { + "epoch": 0.8320868945868946, + "grad_norm": 0.5618211627006531, + "learning_rate": 0.0001795294940933016, + "loss": 1.145, + "step": 4673 + }, + { + "epoch": 0.8322649572649573, + "grad_norm": 0.6275593042373657, + "learning_rate": 0.00017952100772890877, + "loss": 0.9061, + "step": 4674 + }, + { + "epoch": 0.83244301994302, + "grad_norm": 0.5376096367835999, + "learning_rate": 0.00017951251980646702, + "loss": 1.1948, + "step": 4675 + }, + { + "epoch": 0.8326210826210826, + "grad_norm": 0.5162268877029419, + "learning_rate": 0.0001795040303261426, + "loss": 1.2158, + "step": 4676 + }, + { + "epoch": 0.8327991452991453, + "grad_norm": 0.5730512142181396, + "learning_rate": 0.0001794955392881019, + "loss": 0.9962, + "step": 4677 + }, + { + "epoch": 0.832977207977208, + "grad_norm": 0.5128712058067322, + "learning_rate": 0.00017948704669251122, + "loss": 1.2797, + "step": 4678 + }, + { + "epoch": 0.8331552706552706, + "grad_norm": 0.5173979997634888, + "learning_rate": 0.00017947855253953697, + "loss": 1.1093, + "step": 4679 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 0.504646897315979, + "learning_rate": 0.0001794700568293456, + "loss": 1.3171, + "step": 4680 + }, + { + "epoch": 0.833511396011396, + "grad_norm": 0.5638105869293213, + "learning_rate": 0.00017946155956210356, + "loss": 0.9224, + "step": 4681 + }, + { + "epoch": 0.8336894586894587, + "grad_norm": 0.5289680361747742, + "learning_rate": 0.00017945306073797733, + "loss": 0.8919, + "step": 4682 + }, + { + "epoch": 0.8338675213675214, + "grad_norm": 0.5224629044532776, + "learning_rate": 0.0001794445603571334, + "loss": 1.0345, + "step": 4683 + }, + { + "epoch": 0.834045584045584, + "grad_norm": 0.5342282056808472, + "learning_rate": 0.00017943605841973836, + "loss": 1.2305, + "step": 4684 + }, + { + "epoch": 0.8342236467236467, + "grad_norm": 0.6118032336235046, + "learning_rate": 0.00017942755492595874, + "loss": 1.0316, + "step": 4685 + }, + { + "epoch": 0.8344017094017094, + "grad_norm": 0.49112311005592346, + "learning_rate": 0.00017941904987596121, + "loss": 0.9809, + "step": 4686 + }, + { + "epoch": 0.8345797720797721, + "grad_norm": 0.5044063925743103, + "learning_rate": 0.0001794105432699124, + "loss": 0.834, + "step": 4687 + }, + { + "epoch": 0.8347578347578347, + "grad_norm": 0.4849987328052521, + "learning_rate": 0.00017940203510797892, + "loss": 0.9971, + "step": 4688 + }, + { + "epoch": 0.8349358974358975, + "grad_norm": 0.5539469122886658, + "learning_rate": 0.00017939352539032748, + "loss": 1.1599, + "step": 4689 + }, + { + "epoch": 0.8351139601139601, + "grad_norm": 0.5474258065223694, + "learning_rate": 0.00017938501411712485, + "loss": 1.25, + "step": 4690 + }, + { + "epoch": 0.8352920227920227, + "grad_norm": 0.4880213737487793, + "learning_rate": 0.0001793765012885378, + "loss": 1.1471, + "step": 4691 + }, + { + "epoch": 0.8354700854700855, + "grad_norm": 0.5602759718894958, + "learning_rate": 0.00017936798690473309, + "loss": 1.0723, + "step": 4692 + }, + { + "epoch": 0.8356481481481481, + "grad_norm": 0.627775251865387, + "learning_rate": 0.00017935947096587755, + "loss": 1.3768, + "step": 4693 + }, + { + "epoch": 0.8358262108262108, + "grad_norm": 0.5324847102165222, + "learning_rate": 0.00017935095347213804, + "loss": 0.9945, + "step": 4694 + }, + { + "epoch": 0.8360042735042735, + "grad_norm": 0.5244048237800598, + "learning_rate": 0.0001793424344236814, + "loss": 1.1725, + "step": 4695 + }, + { + "epoch": 0.8361823361823362, + "grad_norm": 0.5420708656311035, + "learning_rate": 0.00017933391382067462, + "loss": 1.1267, + "step": 4696 + }, + { + "epoch": 0.8363603988603988, + "grad_norm": 0.5285456776618958, + "learning_rate": 0.00017932539166328458, + "loss": 1.0368, + "step": 4697 + }, + { + "epoch": 0.8365384615384616, + "grad_norm": 0.5330373048782349, + "learning_rate": 0.00017931686795167825, + "loss": 1.1082, + "step": 4698 + }, + { + "epoch": 0.8367165242165242, + "grad_norm": 0.5516682267189026, + "learning_rate": 0.0001793083426860227, + "loss": 1.1833, + "step": 4699 + }, + { + "epoch": 0.8368945868945868, + "grad_norm": 0.5229935646057129, + "learning_rate": 0.0001792998158664849, + "loss": 0.8527, + "step": 4700 + }, + { + "epoch": 0.8370726495726496, + "grad_norm": 0.4821490943431854, + "learning_rate": 0.00017929128749323195, + "loss": 1.1201, + "step": 4701 + }, + { + "epoch": 0.8372507122507122, + "grad_norm": 0.6276404857635498, + "learning_rate": 0.0001792827575664309, + "loss": 1.0986, + "step": 4702 + }, + { + "epoch": 0.8374287749287749, + "grad_norm": 0.5681334733963013, + "learning_rate": 0.00017927422608624897, + "loss": 1.3821, + "step": 4703 + }, + { + "epoch": 0.8376068376068376, + "grad_norm": 0.5257087349891663, + "learning_rate": 0.00017926569305285324, + "loss": 1.1033, + "step": 4704 + }, + { + "epoch": 0.8377849002849003, + "grad_norm": 0.5665168166160583, + "learning_rate": 0.0001792571584664109, + "loss": 1.104, + "step": 4705 + }, + { + "epoch": 0.8379629629629629, + "grad_norm": 0.5202076435089111, + "learning_rate": 0.00017924862232708918, + "loss": 1.052, + "step": 4706 + }, + { + "epoch": 0.8381410256410257, + "grad_norm": 0.5103010535240173, + "learning_rate": 0.00017924008463505534, + "loss": 1.1348, + "step": 4707 + }, + { + "epoch": 0.8383190883190883, + "grad_norm": 0.6811865568161011, + "learning_rate": 0.00017923154539047667, + "loss": 1.2804, + "step": 4708 + }, + { + "epoch": 0.8384971509971509, + "grad_norm": 0.46808311343193054, + "learning_rate": 0.00017922300459352042, + "loss": 0.9302, + "step": 4709 + }, + { + "epoch": 0.8386752136752137, + "grad_norm": 0.47713059186935425, + "learning_rate": 0.00017921446224435398, + "loss": 0.78, + "step": 4710 + }, + { + "epoch": 0.8388532763532763, + "grad_norm": 0.7579890489578247, + "learning_rate": 0.0001792059183431447, + "loss": 1.4776, + "step": 4711 + }, + { + "epoch": 0.8390313390313391, + "grad_norm": 0.6009423136711121, + "learning_rate": 0.00017919737289006, + "loss": 1.2679, + "step": 4712 + }, + { + "epoch": 0.8392094017094017, + "grad_norm": 0.56390780210495, + "learning_rate": 0.00017918882588526729, + "loss": 1.0402, + "step": 4713 + }, + { + "epoch": 0.8393874643874644, + "grad_norm": 0.5698862075805664, + "learning_rate": 0.00017918027732893404, + "loss": 1.2336, + "step": 4714 + }, + { + "epoch": 0.8395655270655271, + "grad_norm": 0.5016305446624756, + "learning_rate": 0.0001791717272212277, + "loss": 1.0373, + "step": 4715 + }, + { + "epoch": 0.8397435897435898, + "grad_norm": 0.5886971950531006, + "learning_rate": 0.0001791631755623159, + "loss": 1.1062, + "step": 4716 + }, + { + "epoch": 0.8399216524216524, + "grad_norm": 0.647833526134491, + "learning_rate": 0.00017915462235236607, + "loss": 1.0464, + "step": 4717 + }, + { + "epoch": 0.8400997150997151, + "grad_norm": 0.4961194396018982, + "learning_rate": 0.00017914606759154587, + "loss": 1.0763, + "step": 4718 + }, + { + "epoch": 0.8402777777777778, + "grad_norm": 0.47041359543800354, + "learning_rate": 0.00017913751128002288, + "loss": 1.0685, + "step": 4719 + }, + { + "epoch": 0.8404558404558404, + "grad_norm": 0.5752858519554138, + "learning_rate": 0.00017912895341796475, + "loss": 1.0577, + "step": 4720 + }, + { + "epoch": 0.8406339031339032, + "grad_norm": 0.5233224034309387, + "learning_rate": 0.00017912039400553914, + "loss": 1.1484, + "step": 4721 + }, + { + "epoch": 0.8408119658119658, + "grad_norm": 0.5327485203742981, + "learning_rate": 0.00017911183304291378, + "loss": 1.0028, + "step": 4722 + }, + { + "epoch": 0.8409900284900285, + "grad_norm": 0.5320752263069153, + "learning_rate": 0.00017910327053025638, + "loss": 1.1247, + "step": 4723 + }, + { + "epoch": 0.8411680911680912, + "grad_norm": 0.529617965221405, + "learning_rate": 0.00017909470646773477, + "loss": 1.1698, + "step": 4724 + }, + { + "epoch": 0.8413461538461539, + "grad_norm": 0.5055609345436096, + "learning_rate": 0.00017908614085551664, + "loss": 1.0925, + "step": 4725 + }, + { + "epoch": 0.8415242165242165, + "grad_norm": 0.5356255769729614, + "learning_rate": 0.00017907757369376985, + "loss": 1.0354, + "step": 4726 + }, + { + "epoch": 0.8417022792022792, + "grad_norm": 0.582834780216217, + "learning_rate": 0.00017906900498266233, + "loss": 1.1248, + "step": 4727 + }, + { + "epoch": 0.8418803418803419, + "grad_norm": 0.5750834941864014, + "learning_rate": 0.00017906043472236188, + "loss": 1.0119, + "step": 4728 + }, + { + "epoch": 0.8420584045584045, + "grad_norm": 0.5923320055007935, + "learning_rate": 0.00017905186291303644, + "loss": 1.0662, + "step": 4729 + }, + { + "epoch": 0.8422364672364673, + "grad_norm": 0.4767811894416809, + "learning_rate": 0.00017904328955485396, + "loss": 1.0911, + "step": 4730 + }, + { + "epoch": 0.8424145299145299, + "grad_norm": 0.5294556021690369, + "learning_rate": 0.00017903471464798245, + "loss": 1.2861, + "step": 4731 + }, + { + "epoch": 0.8425925925925926, + "grad_norm": 0.599117636680603, + "learning_rate": 0.00017902613819258985, + "loss": 1.1707, + "step": 4732 + }, + { + "epoch": 0.8427706552706553, + "grad_norm": 0.5912977457046509, + "learning_rate": 0.00017901756018884424, + "loss": 1.1884, + "step": 4733 + }, + { + "epoch": 0.842948717948718, + "grad_norm": 0.587676465511322, + "learning_rate": 0.0001790089806369137, + "loss": 1.1054, + "step": 4734 + }, + { + "epoch": 0.8431267806267806, + "grad_norm": 0.6271800398826599, + "learning_rate": 0.0001790003995369663, + "loss": 1.2094, + "step": 4735 + }, + { + "epoch": 0.8433048433048433, + "grad_norm": 0.47198590636253357, + "learning_rate": 0.00017899181688917017, + "loss": 0.9561, + "step": 4736 + }, + { + "epoch": 0.843482905982906, + "grad_norm": 0.690732479095459, + "learning_rate": 0.00017898323269369351, + "loss": 1.1629, + "step": 4737 + }, + { + "epoch": 0.8436609686609686, + "grad_norm": 0.4926888048648834, + "learning_rate": 0.00017897464695070445, + "loss": 1.1097, + "step": 4738 + }, + { + "epoch": 0.8438390313390314, + "grad_norm": 0.7071278691291809, + "learning_rate": 0.00017896605966037128, + "loss": 1.195, + "step": 4739 + }, + { + "epoch": 0.844017094017094, + "grad_norm": 0.5650486350059509, + "learning_rate": 0.00017895747082286216, + "loss": 1.0107, + "step": 4740 + }, + { + "epoch": 0.8441951566951567, + "grad_norm": 0.5291931629180908, + "learning_rate": 0.00017894888043834545, + "loss": 1.0104, + "step": 4741 + }, + { + "epoch": 0.8443732193732194, + "grad_norm": 0.5751241445541382, + "learning_rate": 0.00017894028850698942, + "loss": 1.2482, + "step": 4742 + }, + { + "epoch": 0.844551282051282, + "grad_norm": 0.5833632349967957, + "learning_rate": 0.0001789316950289624, + "loss": 1.0552, + "step": 4743 + }, + { + "epoch": 0.8447293447293447, + "grad_norm": 0.543729841709137, + "learning_rate": 0.00017892310000443282, + "loss": 1.1453, + "step": 4744 + }, + { + "epoch": 0.8449074074074074, + "grad_norm": 0.5674204230308533, + "learning_rate": 0.00017891450343356902, + "loss": 1.0757, + "step": 4745 + }, + { + "epoch": 0.8450854700854701, + "grad_norm": 0.5161892771720886, + "learning_rate": 0.00017890590531653946, + "loss": 1.1163, + "step": 4746 + }, + { + "epoch": 0.8452635327635327, + "grad_norm": 0.49907612800598145, + "learning_rate": 0.00017889730565351258, + "loss": 1.0356, + "step": 4747 + }, + { + "epoch": 0.8454415954415955, + "grad_norm": 0.4994732439517975, + "learning_rate": 0.00017888870444465692, + "loss": 1.026, + "step": 4748 + }, + { + "epoch": 0.8456196581196581, + "grad_norm": 0.6397520303726196, + "learning_rate": 0.00017888010169014095, + "loss": 0.957, + "step": 4749 + }, + { + "epoch": 0.8457977207977208, + "grad_norm": 0.5379729270935059, + "learning_rate": 0.00017887149739013327, + "loss": 1.1664, + "step": 4750 + }, + { + "epoch": 0.8459757834757835, + "grad_norm": 0.4487382769584656, + "learning_rate": 0.00017886289154480246, + "loss": 0.9377, + "step": 4751 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.5645943880081177, + "learning_rate": 0.00017885428415431707, + "loss": 1.273, + "step": 4752 + }, + { + "epoch": 0.8463319088319088, + "grad_norm": 0.5535289645195007, + "learning_rate": 0.00017884567521884577, + "loss": 1.1779, + "step": 4753 + }, + { + "epoch": 0.8465099715099715, + "grad_norm": 0.5039721131324768, + "learning_rate": 0.0001788370647385573, + "loss": 1.0237, + "step": 4754 + }, + { + "epoch": 0.8466880341880342, + "grad_norm": 0.4543854892253876, + "learning_rate": 0.00017882845271362032, + "loss": 0.8149, + "step": 4755 + }, + { + "epoch": 0.8468660968660968, + "grad_norm": 0.5095639824867249, + "learning_rate": 0.00017881983914420352, + "loss": 1.0141, + "step": 4756 + }, + { + "epoch": 0.8470441595441596, + "grad_norm": 0.5341798663139343, + "learning_rate": 0.00017881122403047575, + "loss": 1.1885, + "step": 4757 + }, + { + "epoch": 0.8472222222222222, + "grad_norm": 0.5595062971115112, + "learning_rate": 0.00017880260737260573, + "loss": 0.8939, + "step": 4758 + }, + { + "epoch": 0.8474002849002849, + "grad_norm": 0.5355880260467529, + "learning_rate": 0.00017879398917076232, + "loss": 1.2434, + "step": 4759 + }, + { + "epoch": 0.8475783475783476, + "grad_norm": 0.49477261304855347, + "learning_rate": 0.0001787853694251144, + "loss": 0.979, + "step": 4760 + }, + { + "epoch": 0.8477564102564102, + "grad_norm": 0.5154359340667725, + "learning_rate": 0.00017877674813583078, + "loss": 1.0957, + "step": 4761 + }, + { + "epoch": 0.8479344729344729, + "grad_norm": 0.5651070475578308, + "learning_rate": 0.00017876812530308046, + "loss": 1.1884, + "step": 4762 + }, + { + "epoch": 0.8481125356125356, + "grad_norm": 0.537277340888977, + "learning_rate": 0.00017875950092703232, + "loss": 1.0272, + "step": 4763 + }, + { + "epoch": 0.8482905982905983, + "grad_norm": 0.5259691476821899, + "learning_rate": 0.00017875087500785538, + "loss": 1.1493, + "step": 4764 + }, + { + "epoch": 0.8484686609686609, + "grad_norm": 0.5491300225257874, + "learning_rate": 0.00017874224754571867, + "loss": 0.8316, + "step": 4765 + }, + { + "epoch": 0.8486467236467237, + "grad_norm": 0.5493744611740112, + "learning_rate": 0.00017873361854079116, + "loss": 1.2328, + "step": 4766 + }, + { + "epoch": 0.8488247863247863, + "grad_norm": 0.571002185344696, + "learning_rate": 0.00017872498799324197, + "loss": 1.1384, + "step": 4767 + }, + { + "epoch": 0.8490028490028491, + "grad_norm": 0.538152813911438, + "learning_rate": 0.00017871635590324013, + "loss": 1.0581, + "step": 4768 + }, + { + "epoch": 0.8491809116809117, + "grad_norm": 0.5214923620223999, + "learning_rate": 0.00017870772227095486, + "loss": 1.0612, + "step": 4769 + }, + { + "epoch": 0.8493589743589743, + "grad_norm": 0.5714883804321289, + "learning_rate": 0.0001786990870965553, + "loss": 0.9076, + "step": 4770 + }, + { + "epoch": 0.8495370370370371, + "grad_norm": 0.4181775450706482, + "learning_rate": 0.00017869045038021054, + "loss": 0.8366, + "step": 4771 + }, + { + "epoch": 0.8497150997150997, + "grad_norm": 0.6266027688980103, + "learning_rate": 0.00017868181212208993, + "loss": 1.2047, + "step": 4772 + }, + { + "epoch": 0.8498931623931624, + "grad_norm": 0.5423732399940491, + "learning_rate": 0.0001786731723223626, + "loss": 1.3878, + "step": 4773 + }, + { + "epoch": 0.8500712250712251, + "grad_norm": 0.5512300133705139, + "learning_rate": 0.00017866453098119793, + "loss": 1.1132, + "step": 4774 + }, + { + "epoch": 0.8502492877492878, + "grad_norm": 0.5767185688018799, + "learning_rate": 0.00017865588809876519, + "loss": 0.97, + "step": 4775 + }, + { + "epoch": 0.8504273504273504, + "grad_norm": 0.5305790305137634, + "learning_rate": 0.00017864724367523368, + "loss": 1.1158, + "step": 4776 + }, + { + "epoch": 0.8506054131054132, + "grad_norm": 0.49702391028404236, + "learning_rate": 0.00017863859771077284, + "loss": 0.9669, + "step": 4777 + }, + { + "epoch": 0.8507834757834758, + "grad_norm": 0.5490063428878784, + "learning_rate": 0.00017862995020555205, + "loss": 1.0646, + "step": 4778 + }, + { + "epoch": 0.8509615384615384, + "grad_norm": 0.5308689475059509, + "learning_rate": 0.00017862130115974068, + "loss": 0.8922, + "step": 4779 + }, + { + "epoch": 0.8511396011396012, + "grad_norm": 0.5412983894348145, + "learning_rate": 0.00017861265057350826, + "loss": 1.1444, + "step": 4780 + }, + { + "epoch": 0.8513176638176638, + "grad_norm": 0.5857377052307129, + "learning_rate": 0.00017860399844702425, + "loss": 1.1643, + "step": 4781 + }, + { + "epoch": 0.8514957264957265, + "grad_norm": 0.599273681640625, + "learning_rate": 0.00017859534478045815, + "loss": 1.169, + "step": 4782 + }, + { + "epoch": 0.8516737891737892, + "grad_norm": 0.5677087903022766, + "learning_rate": 0.00017858668957397957, + "loss": 1.0793, + "step": 4783 + }, + { + "epoch": 0.8518518518518519, + "grad_norm": 0.5648362636566162, + "learning_rate": 0.00017857803282775807, + "loss": 1.1932, + "step": 4784 + }, + { + "epoch": 0.8520299145299145, + "grad_norm": 0.5138826966285706, + "learning_rate": 0.00017856937454196323, + "loss": 1.0011, + "step": 4785 + }, + { + "epoch": 0.8522079772079773, + "grad_norm": 0.5951429009437561, + "learning_rate": 0.0001785607147167647, + "loss": 1.3198, + "step": 4786 + }, + { + "epoch": 0.8523860398860399, + "grad_norm": 0.5341953039169312, + "learning_rate": 0.00017855205335233216, + "loss": 0.9094, + "step": 4787 + }, + { + "epoch": 0.8525641025641025, + "grad_norm": 0.5193579196929932, + "learning_rate": 0.00017854339044883535, + "loss": 0.892, + "step": 4788 + }, + { + "epoch": 0.8527421652421653, + "grad_norm": 0.5053097009658813, + "learning_rate": 0.00017853472600644392, + "loss": 1.0589, + "step": 4789 + }, + { + "epoch": 0.8529202279202279, + "grad_norm": 0.5819617509841919, + "learning_rate": 0.0001785260600253277, + "loss": 1.2646, + "step": 4790 + }, + { + "epoch": 0.8530982905982906, + "grad_norm": 0.5327470302581787, + "learning_rate": 0.00017851739250565645, + "loss": 1.056, + "step": 4791 + }, + { + "epoch": 0.8532763532763533, + "grad_norm": 0.5131269097328186, + "learning_rate": 0.0001785087234476, + "loss": 1.1192, + "step": 4792 + }, + { + "epoch": 0.853454415954416, + "grad_norm": 0.4698086977005005, + "learning_rate": 0.00017850005285132821, + "loss": 0.9849, + "step": 4793 + }, + { + "epoch": 0.8536324786324786, + "grad_norm": 0.5503947734832764, + "learning_rate": 0.00017849138071701092, + "loss": 1.1139, + "step": 4794 + }, + { + "epoch": 0.8538105413105413, + "grad_norm": 0.5120903849601746, + "learning_rate": 0.0001784827070448181, + "loss": 0.9801, + "step": 4795 + }, + { + "epoch": 0.853988603988604, + "grad_norm": 0.47650405764579773, + "learning_rate": 0.00017847403183491968, + "loss": 1.0268, + "step": 4796 + }, + { + "epoch": 0.8541666666666666, + "grad_norm": 0.5773387551307678, + "learning_rate": 0.0001784653550874856, + "loss": 1.0336, + "step": 4797 + }, + { + "epoch": 0.8543447293447294, + "grad_norm": 0.545531153678894, + "learning_rate": 0.00017845667680268593, + "loss": 1.0532, + "step": 4798 + }, + { + "epoch": 0.854522792022792, + "grad_norm": 0.533161461353302, + "learning_rate": 0.0001784479969806906, + "loss": 1.1964, + "step": 4799 + }, + { + "epoch": 0.8547008547008547, + "grad_norm": 0.5880789160728455, + "learning_rate": 0.00017843931562166977, + "loss": 1.1588, + "step": 4800 + }, + { + "epoch": 0.8548789173789174, + "grad_norm": 0.5381524562835693, + "learning_rate": 0.00017843063272579346, + "loss": 1.1533, + "step": 4801 + }, + { + "epoch": 0.85505698005698, + "grad_norm": 0.6280176639556885, + "learning_rate": 0.00017842194829323187, + "loss": 1.0084, + "step": 4802 + }, + { + "epoch": 0.8552350427350427, + "grad_norm": 0.5098552703857422, + "learning_rate": 0.0001784132623241551, + "loss": 1.0804, + "step": 4803 + }, + { + "epoch": 0.8554131054131054, + "grad_norm": 0.5406526923179626, + "learning_rate": 0.00017840457481873328, + "loss": 1.2571, + "step": 4804 + }, + { + "epoch": 0.8555911680911681, + "grad_norm": 0.5859003663063049, + "learning_rate": 0.00017839588577713678, + "loss": 1.2462, + "step": 4805 + }, + { + "epoch": 0.8557692307692307, + "grad_norm": 0.6209002137184143, + "learning_rate": 0.00017838719519953572, + "loss": 1.307, + "step": 4806 + }, + { + "epoch": 0.8559472934472935, + "grad_norm": 0.525753915309906, + "learning_rate": 0.00017837850308610037, + "loss": 1.2957, + "step": 4807 + }, + { + "epoch": 0.8561253561253561, + "grad_norm": 0.5096195340156555, + "learning_rate": 0.0001783698094370011, + "loss": 1.1433, + "step": 4808 + }, + { + "epoch": 0.8563034188034188, + "grad_norm": 0.5873076915740967, + "learning_rate": 0.0001783611142524082, + "loss": 1.2271, + "step": 4809 + }, + { + "epoch": 0.8564814814814815, + "grad_norm": 0.5093944668769836, + "learning_rate": 0.0001783524175324921, + "loss": 0.8788, + "step": 4810 + }, + { + "epoch": 0.8566595441595442, + "grad_norm": 0.5485084652900696, + "learning_rate": 0.00017834371927742307, + "loss": 1.256, + "step": 4811 + }, + { + "epoch": 0.8568376068376068, + "grad_norm": 0.5808873772621155, + "learning_rate": 0.00017833501948737163, + "loss": 0.9287, + "step": 4812 + }, + { + "epoch": 0.8570156695156695, + "grad_norm": 0.5113978385925293, + "learning_rate": 0.00017832631816250822, + "loss": 1.0372, + "step": 4813 + }, + { + "epoch": 0.8571937321937322, + "grad_norm": 0.5877016186714172, + "learning_rate": 0.0001783176153030033, + "loss": 1.3023, + "step": 4814 + }, + { + "epoch": 0.8573717948717948, + "grad_norm": 0.534328043460846, + "learning_rate": 0.00017830891090902742, + "loss": 1.1023, + "step": 4815 + }, + { + "epoch": 0.8575498575498576, + "grad_norm": 0.5781638026237488, + "learning_rate": 0.0001783002049807511, + "loss": 0.9562, + "step": 4816 + }, + { + "epoch": 0.8577279202279202, + "grad_norm": 0.5760263204574585, + "learning_rate": 0.00017829149751834487, + "loss": 0.8733, + "step": 4817 + }, + { + "epoch": 0.8579059829059829, + "grad_norm": 0.3887255787849426, + "learning_rate": 0.00017828278852197944, + "loss": 0.5949, + "step": 4818 + }, + { + "epoch": 0.8580840455840456, + "grad_norm": 0.47814446687698364, + "learning_rate": 0.00017827407799182537, + "loss": 1.0698, + "step": 4819 + }, + { + "epoch": 0.8582621082621082, + "grad_norm": 0.5520272254943848, + "learning_rate": 0.00017826536592805334, + "loss": 1.1314, + "step": 4820 + }, + { + "epoch": 0.8584401709401709, + "grad_norm": 0.5285319685935974, + "learning_rate": 0.00017825665233083405, + "loss": 1.1618, + "step": 4821 + }, + { + "epoch": 0.8586182336182336, + "grad_norm": 0.6080102324485779, + "learning_rate": 0.0001782479372003382, + "loss": 1.3817, + "step": 4822 + }, + { + "epoch": 0.8587962962962963, + "grad_norm": 0.7474410533905029, + "learning_rate": 0.00017823922053673662, + "loss": 1.1321, + "step": 4823 + }, + { + "epoch": 0.8589743589743589, + "grad_norm": 0.559283435344696, + "learning_rate": 0.0001782305023402, + "loss": 1.1894, + "step": 4824 + }, + { + "epoch": 0.8591524216524217, + "grad_norm": 0.5620571374893188, + "learning_rate": 0.00017822178261089918, + "loss": 1.134, + "step": 4825 + }, + { + "epoch": 0.8593304843304843, + "grad_norm": 0.5553044676780701, + "learning_rate": 0.00017821306134900504, + "loss": 1.3222, + "step": 4826 + }, + { + "epoch": 0.8595085470085471, + "grad_norm": 0.6177778244018555, + "learning_rate": 0.00017820433855468846, + "loss": 1.2545, + "step": 4827 + }, + { + "epoch": 0.8596866096866097, + "grad_norm": 0.656233012676239, + "learning_rate": 0.0001781956142281203, + "loss": 1.1346, + "step": 4828 + }, + { + "epoch": 0.8598646723646723, + "grad_norm": 0.6710973381996155, + "learning_rate": 0.0001781868883694715, + "loss": 1.1361, + "step": 4829 + }, + { + "epoch": 0.8600427350427351, + "grad_norm": 0.5093601942062378, + "learning_rate": 0.0001781781609789131, + "loss": 1.0509, + "step": 4830 + }, + { + "epoch": 0.8602207977207977, + "grad_norm": 0.5707578063011169, + "learning_rate": 0.00017816943205661598, + "loss": 1.0964, + "step": 4831 + }, + { + "epoch": 0.8603988603988604, + "grad_norm": 0.6159597635269165, + "learning_rate": 0.00017816070160275125, + "loss": 1.0322, + "step": 4832 + }, + { + "epoch": 0.8605769230769231, + "grad_norm": 0.5430580377578735, + "learning_rate": 0.0001781519696174899, + "loss": 1.2464, + "step": 4833 + }, + { + "epoch": 0.8607549857549858, + "grad_norm": 0.48104700446128845, + "learning_rate": 0.0001781432361010031, + "loss": 1.1031, + "step": 4834 + }, + { + "epoch": 0.8609330484330484, + "grad_norm": 0.5304946303367615, + "learning_rate": 0.0001781345010534619, + "loss": 1.0281, + "step": 4835 + }, + { + "epoch": 0.8611111111111112, + "grad_norm": 0.5230711698532104, + "learning_rate": 0.00017812576447503742, + "loss": 0.9499, + "step": 4836 + }, + { + "epoch": 0.8612891737891738, + "grad_norm": 0.5363606214523315, + "learning_rate": 0.00017811702636590093, + "loss": 1.1358, + "step": 4837 + }, + { + "epoch": 0.8614672364672364, + "grad_norm": 0.5880044102668762, + "learning_rate": 0.00017810828672622358, + "loss": 1.1765, + "step": 4838 + }, + { + "epoch": 0.8616452991452992, + "grad_norm": 0.5194395184516907, + "learning_rate": 0.0001780995455561766, + "loss": 1.1622, + "step": 4839 + }, + { + "epoch": 0.8618233618233618, + "grad_norm": 0.5114264488220215, + "learning_rate": 0.00017809080285593126, + "loss": 1.0081, + "step": 4840 + }, + { + "epoch": 0.8620014245014245, + "grad_norm": 0.6174240112304688, + "learning_rate": 0.00017808205862565886, + "loss": 1.0745, + "step": 4841 + }, + { + "epoch": 0.8621794871794872, + "grad_norm": 0.5662630200386047, + "learning_rate": 0.0001780733128655307, + "loss": 1.3369, + "step": 4842 + }, + { + "epoch": 0.8623575498575499, + "grad_norm": 0.5917882919311523, + "learning_rate": 0.00017806456557571817, + "loss": 1.1631, + "step": 4843 + }, + { + "epoch": 0.8625356125356125, + "grad_norm": 0.5305736660957336, + "learning_rate": 0.00017805581675639265, + "loss": 0.9875, + "step": 4844 + }, + { + "epoch": 0.8627136752136753, + "grad_norm": 0.5181219577789307, + "learning_rate": 0.00017804706640772556, + "loss": 0.9918, + "step": 4845 + }, + { + "epoch": 0.8628917378917379, + "grad_norm": 0.5467997789382935, + "learning_rate": 0.00017803831452988832, + "loss": 1.1395, + "step": 4846 + }, + { + "epoch": 0.8630698005698005, + "grad_norm": 0.5494031310081482, + "learning_rate": 0.00017802956112305241, + "loss": 1.0312, + "step": 4847 + }, + { + "epoch": 0.8632478632478633, + "grad_norm": 0.5804065465927124, + "learning_rate": 0.00017802080618738931, + "loss": 1.1555, + "step": 4848 + }, + { + "epoch": 0.8634259259259259, + "grad_norm": 0.5424801111221313, + "learning_rate": 0.00017801204972307067, + "loss": 1.0215, + "step": 4849 + }, + { + "epoch": 0.8636039886039886, + "grad_norm": 0.5321891903877258, + "learning_rate": 0.0001780032917302679, + "loss": 1.0187, + "step": 4850 + }, + { + "epoch": 0.8637820512820513, + "grad_norm": 0.5543400049209595, + "learning_rate": 0.0001779945322091527, + "loss": 1.1972, + "step": 4851 + }, + { + "epoch": 0.863960113960114, + "grad_norm": 0.566649317741394, + "learning_rate": 0.00017798577115989668, + "loss": 1.0758, + "step": 4852 + }, + { + "epoch": 0.8641381766381766, + "grad_norm": 0.5538444519042969, + "learning_rate": 0.00017797700858267145, + "loss": 1.1338, + "step": 4853 + }, + { + "epoch": 0.8643162393162394, + "grad_norm": 0.5641313791275024, + "learning_rate": 0.0001779682444776487, + "loss": 1.256, + "step": 4854 + }, + { + "epoch": 0.864494301994302, + "grad_norm": 0.6377350091934204, + "learning_rate": 0.00017795947884500016, + "loss": 1.144, + "step": 4855 + }, + { + "epoch": 0.8646723646723646, + "grad_norm": 0.5581876039505005, + "learning_rate": 0.0001779507116848976, + "loss": 1.3163, + "step": 4856 + }, + { + "epoch": 0.8648504273504274, + "grad_norm": 0.5416772365570068, + "learning_rate": 0.0001779419429975128, + "loss": 1.0219, + "step": 4857 + }, + { + "epoch": 0.86502849002849, + "grad_norm": 0.5450608730316162, + "learning_rate": 0.0001779331727830175, + "loss": 1.0093, + "step": 4858 + }, + { + "epoch": 0.8652065527065527, + "grad_norm": 0.5151242017745972, + "learning_rate": 0.00017792440104158358, + "loss": 1.067, + "step": 4859 + }, + { + "epoch": 0.8653846153846154, + "grad_norm": 0.5225046873092651, + "learning_rate": 0.0001779156277733829, + "loss": 1.0432, + "step": 4860 + }, + { + "epoch": 0.8655626780626781, + "grad_norm": 0.5168602466583252, + "learning_rate": 0.00017790685297858737, + "loss": 0.9665, + "step": 4861 + }, + { + "epoch": 0.8657407407407407, + "grad_norm": 0.5749059319496155, + "learning_rate": 0.00017789807665736889, + "loss": 1.1607, + "step": 4862 + }, + { + "epoch": 0.8659188034188035, + "grad_norm": 0.45656394958496094, + "learning_rate": 0.00017788929880989938, + "loss": 0.8362, + "step": 4863 + }, + { + "epoch": 0.8660968660968661, + "grad_norm": 0.5090615749359131, + "learning_rate": 0.00017788051943635086, + "loss": 0.9553, + "step": 4864 + }, + { + "epoch": 0.8662749287749287, + "grad_norm": 0.5381240248680115, + "learning_rate": 0.0001778717385368954, + "loss": 1.1391, + "step": 4865 + }, + { + "epoch": 0.8664529914529915, + "grad_norm": 0.522720456123352, + "learning_rate": 0.00017786295611170493, + "loss": 1.1869, + "step": 4866 + }, + { + "epoch": 0.8666310541310541, + "grad_norm": 0.530986487865448, + "learning_rate": 0.0001778541721609516, + "loss": 1.1046, + "step": 4867 + }, + { + "epoch": 0.8668091168091168, + "grad_norm": 0.5065864324569702, + "learning_rate": 0.0001778453866848075, + "loss": 1.008, + "step": 4868 + }, + { + "epoch": 0.8669871794871795, + "grad_norm": 0.5541394352912903, + "learning_rate": 0.00017783659968344476, + "loss": 1.0004, + "step": 4869 + }, + { + "epoch": 0.8671652421652422, + "grad_norm": 0.5059576630592346, + "learning_rate": 0.00017782781115703556, + "loss": 1.128, + "step": 4870 + }, + { + "epoch": 0.8673433048433048, + "grad_norm": 0.5052187442779541, + "learning_rate": 0.00017781902110575203, + "loss": 0.8544, + "step": 4871 + }, + { + "epoch": 0.8675213675213675, + "grad_norm": 0.5383397340774536, + "learning_rate": 0.00017781022952976646, + "loss": 1.1411, + "step": 4872 + }, + { + "epoch": 0.8676994301994302, + "grad_norm": 0.4760429859161377, + "learning_rate": 0.00017780143642925106, + "loss": 0.8246, + "step": 4873 + }, + { + "epoch": 0.8678774928774928, + "grad_norm": 0.5480535626411438, + "learning_rate": 0.00017779264180437817, + "loss": 1.013, + "step": 4874 + }, + { + "epoch": 0.8680555555555556, + "grad_norm": 0.5303317904472351, + "learning_rate": 0.00017778384565532004, + "loss": 1.0201, + "step": 4875 + }, + { + "epoch": 0.8682336182336182, + "grad_norm": 0.5365355014801025, + "learning_rate": 0.00017777504798224903, + "loss": 1.1107, + "step": 4876 + }, + { + "epoch": 0.8684116809116809, + "grad_norm": 0.5173360705375671, + "learning_rate": 0.00017776624878533754, + "loss": 1.0808, + "step": 4877 + }, + { + "epoch": 0.8685897435897436, + "grad_norm": 0.5088842511177063, + "learning_rate": 0.00017775744806475792, + "loss": 0.995, + "step": 4878 + }, + { + "epoch": 0.8687678062678063, + "grad_norm": 0.5796698927879333, + "learning_rate": 0.00017774864582068264, + "loss": 1.1485, + "step": 4879 + }, + { + "epoch": 0.8689458689458689, + "grad_norm": 0.5719375610351562, + "learning_rate": 0.00017773984205328417, + "loss": 1.0133, + "step": 4880 + }, + { + "epoch": 0.8691239316239316, + "grad_norm": 0.6396418213844299, + "learning_rate": 0.00017773103676273498, + "loss": 1.0932, + "step": 4881 + }, + { + "epoch": 0.8693019943019943, + "grad_norm": 0.5602468252182007, + "learning_rate": 0.00017772222994920763, + "loss": 0.9702, + "step": 4882 + }, + { + "epoch": 0.8694800569800569, + "grad_norm": 0.5167748332023621, + "learning_rate": 0.00017771342161287457, + "loss": 1.0528, + "step": 4883 + }, + { + "epoch": 0.8696581196581197, + "grad_norm": 0.5572916865348816, + "learning_rate": 0.00017770461175390848, + "loss": 1.1341, + "step": 4884 + }, + { + "epoch": 0.8698361823361823, + "grad_norm": 0.6666276454925537, + "learning_rate": 0.00017769580037248195, + "loss": 1.1948, + "step": 4885 + }, + { + "epoch": 0.8700142450142451, + "grad_norm": 0.5348601937294006, + "learning_rate": 0.0001776869874687676, + "loss": 1.0562, + "step": 4886 + }, + { + "epoch": 0.8701923076923077, + "grad_norm": 0.5449648499488831, + "learning_rate": 0.00017767817304293812, + "loss": 0.988, + "step": 4887 + }, + { + "epoch": 0.8703703703703703, + "grad_norm": 0.5995045304298401, + "learning_rate": 0.0001776693570951662, + "loss": 1.2526, + "step": 4888 + }, + { + "epoch": 0.8705484330484331, + "grad_norm": 0.6575320959091187, + "learning_rate": 0.00017766053962562457, + "loss": 1.1717, + "step": 4889 + }, + { + "epoch": 0.8707264957264957, + "grad_norm": 0.5882139801979065, + "learning_rate": 0.00017765172063448597, + "loss": 1.238, + "step": 4890 + }, + { + "epoch": 0.8709045584045584, + "grad_norm": 0.5908389091491699, + "learning_rate": 0.00017764290012192325, + "loss": 1.0606, + "step": 4891 + }, + { + "epoch": 0.8710826210826211, + "grad_norm": 0.6169339418411255, + "learning_rate": 0.00017763407808810917, + "loss": 1.1456, + "step": 4892 + }, + { + "epoch": 0.8712606837606838, + "grad_norm": 0.5916035771369934, + "learning_rate": 0.0001776252545332166, + "loss": 1.0026, + "step": 4893 + }, + { + "epoch": 0.8714387464387464, + "grad_norm": 0.539995551109314, + "learning_rate": 0.00017761642945741843, + "loss": 1.2397, + "step": 4894 + }, + { + "epoch": 0.8716168091168092, + "grad_norm": 0.5346137881278992, + "learning_rate": 0.00017760760286088755, + "loss": 1.1232, + "step": 4895 + }, + { + "epoch": 0.8717948717948718, + "grad_norm": 0.570202112197876, + "learning_rate": 0.00017759877474379692, + "loss": 1.0708, + "step": 4896 + }, + { + "epoch": 0.8719729344729344, + "grad_norm": 0.5023398399353027, + "learning_rate": 0.00017758994510631948, + "loss": 1.1056, + "step": 4897 + }, + { + "epoch": 0.8721509971509972, + "grad_norm": 0.5447137951850891, + "learning_rate": 0.00017758111394862826, + "loss": 0.8776, + "step": 4898 + }, + { + "epoch": 0.8723290598290598, + "grad_norm": 0.5193906426429749, + "learning_rate": 0.00017757228127089625, + "loss": 0.9959, + "step": 4899 + }, + { + "epoch": 0.8725071225071225, + "grad_norm": 0.5958787798881531, + "learning_rate": 0.00017756344707329656, + "loss": 1.092, + "step": 4900 + }, + { + "epoch": 0.8726851851851852, + "grad_norm": 0.521045982837677, + "learning_rate": 0.00017755461135600221, + "loss": 0.9864, + "step": 4901 + }, + { + "epoch": 0.8728632478632479, + "grad_norm": 0.5257635116577148, + "learning_rate": 0.00017754577411918638, + "loss": 1.216, + "step": 4902 + }, + { + "epoch": 0.8730413105413105, + "grad_norm": 0.5425964593887329, + "learning_rate": 0.0001775369353630222, + "loss": 1.1432, + "step": 4903 + }, + { + "epoch": 0.8732193732193733, + "grad_norm": 0.47995322942733765, + "learning_rate": 0.00017752809508768286, + "loss": 1.0227, + "step": 4904 + }, + { + "epoch": 0.8733974358974359, + "grad_norm": 0.5747429728507996, + "learning_rate": 0.0001775192532933415, + "loss": 0.9984, + "step": 4905 + }, + { + "epoch": 0.8735754985754985, + "grad_norm": 0.5745723247528076, + "learning_rate": 0.00017751040998017142, + "loss": 1.2559, + "step": 4906 + }, + { + "epoch": 0.8737535612535613, + "grad_norm": 0.6114141941070557, + "learning_rate": 0.0001775015651483459, + "loss": 1.3224, + "step": 4907 + }, + { + "epoch": 0.8739316239316239, + "grad_norm": 0.4757187068462372, + "learning_rate": 0.00017749271879803817, + "loss": 1.0352, + "step": 4908 + }, + { + "epoch": 0.8741096866096866, + "grad_norm": 0.48644450306892395, + "learning_rate": 0.0001774838709294216, + "loss": 1.0876, + "step": 4909 + }, + { + "epoch": 0.8742877492877493, + "grad_norm": 0.5652037262916565, + "learning_rate": 0.00017747502154266955, + "loss": 0.9189, + "step": 4910 + }, + { + "epoch": 0.874465811965812, + "grad_norm": 0.5289644002914429, + "learning_rate": 0.00017746617063795538, + "loss": 0.9431, + "step": 4911 + }, + { + "epoch": 0.8746438746438746, + "grad_norm": 0.594656229019165, + "learning_rate": 0.00017745731821545253, + "loss": 1.2408, + "step": 4912 + }, + { + "epoch": 0.8748219373219374, + "grad_norm": 0.5693240165710449, + "learning_rate": 0.0001774484642753344, + "loss": 1.347, + "step": 4913 + }, + { + "epoch": 0.875, + "grad_norm": 0.5291008949279785, + "learning_rate": 0.00017743960881777456, + "loss": 1.161, + "step": 4914 + }, + { + "epoch": 0.8751780626780626, + "grad_norm": 0.5958300232887268, + "learning_rate": 0.00017743075184294642, + "loss": 1.2058, + "step": 4915 + }, + { + "epoch": 0.8753561253561254, + "grad_norm": 0.513884425163269, + "learning_rate": 0.00017742189335102354, + "loss": 1.0952, + "step": 4916 + }, + { + "epoch": 0.875534188034188, + "grad_norm": 0.5860681533813477, + "learning_rate": 0.00017741303334217948, + "loss": 1.1801, + "step": 4917 + }, + { + "epoch": 0.8757122507122507, + "grad_norm": 0.47962820529937744, + "learning_rate": 0.00017740417181658788, + "loss": 1.0785, + "step": 4918 + }, + { + "epoch": 0.8758903133903134, + "grad_norm": 0.5110440254211426, + "learning_rate": 0.00017739530877442227, + "loss": 1.1385, + "step": 4919 + }, + { + "epoch": 0.8760683760683761, + "grad_norm": 0.5106285214424133, + "learning_rate": 0.00017738644421585643, + "loss": 1.1204, + "step": 4920 + }, + { + "epoch": 0.8762464387464387, + "grad_norm": 0.5709205865859985, + "learning_rate": 0.00017737757814106393, + "loss": 1.0108, + "step": 4921 + }, + { + "epoch": 0.8764245014245015, + "grad_norm": 0.5850250124931335, + "learning_rate": 0.0001773687105502185, + "loss": 1.0059, + "step": 4922 + }, + { + "epoch": 0.8766025641025641, + "grad_norm": 0.5194727778434753, + "learning_rate": 0.00017735984144349396, + "loss": 0.9466, + "step": 4923 + }, + { + "epoch": 0.8767806267806267, + "grad_norm": 0.5246787667274475, + "learning_rate": 0.000177350970821064, + "loss": 1.1336, + "step": 4924 + }, + { + "epoch": 0.8769586894586895, + "grad_norm": 0.5798323154449463, + "learning_rate": 0.00017734209868310244, + "loss": 1.1641, + "step": 4925 + }, + { + "epoch": 0.8771367521367521, + "grad_norm": 0.5188565850257874, + "learning_rate": 0.00017733322502978314, + "loss": 0.9959, + "step": 4926 + }, + { + "epoch": 0.8773148148148148, + "grad_norm": 0.5969653725624084, + "learning_rate": 0.00017732434986127995, + "loss": 1.2162, + "step": 4927 + }, + { + "epoch": 0.8774928774928775, + "grad_norm": 0.5520089268684387, + "learning_rate": 0.00017731547317776674, + "loss": 1.0163, + "step": 4928 + }, + { + "epoch": 0.8776709401709402, + "grad_norm": 0.48789507150650024, + "learning_rate": 0.00017730659497941745, + "loss": 0.9757, + "step": 4929 + }, + { + "epoch": 0.8778490028490028, + "grad_norm": 0.6034960746765137, + "learning_rate": 0.000177297715266406, + "loss": 1.1278, + "step": 4930 + }, + { + "epoch": 0.8780270655270656, + "grad_norm": 0.53016597032547, + "learning_rate": 0.00017728883403890638, + "loss": 1.0637, + "step": 4931 + }, + { + "epoch": 0.8782051282051282, + "grad_norm": 0.5073726177215576, + "learning_rate": 0.00017727995129709266, + "loss": 1.1491, + "step": 4932 + }, + { + "epoch": 0.8783831908831908, + "grad_norm": 0.540605366230011, + "learning_rate": 0.00017727106704113878, + "loss": 1.0133, + "step": 4933 + }, + { + "epoch": 0.8785612535612536, + "grad_norm": 0.5346775054931641, + "learning_rate": 0.0001772621812712189, + "loss": 1.1781, + "step": 4934 + }, + { + "epoch": 0.8787393162393162, + "grad_norm": 0.5659036040306091, + "learning_rate": 0.00017725329398750702, + "loss": 1.1023, + "step": 4935 + }, + { + "epoch": 0.8789173789173789, + "grad_norm": 0.591063380241394, + "learning_rate": 0.00017724440519017738, + "loss": 1.0298, + "step": 4936 + }, + { + "epoch": 0.8790954415954416, + "grad_norm": 0.5173781514167786, + "learning_rate": 0.0001772355148794041, + "loss": 1.0483, + "step": 4937 + }, + { + "epoch": 0.8792735042735043, + "grad_norm": 0.5405352711677551, + "learning_rate": 0.0001772266230553613, + "loss": 1.0716, + "step": 4938 + }, + { + "epoch": 0.8794515669515669, + "grad_norm": 0.518442690372467, + "learning_rate": 0.00017721772971822323, + "loss": 1.1373, + "step": 4939 + }, + { + "epoch": 0.8796296296296297, + "grad_norm": 0.533673107624054, + "learning_rate": 0.0001772088348681642, + "loss": 1.0489, + "step": 4940 + }, + { + "epoch": 0.8798076923076923, + "grad_norm": 0.46117857098579407, + "learning_rate": 0.0001771999385053584, + "loss": 1.0297, + "step": 4941 + }, + { + "epoch": 0.8799857549857549, + "grad_norm": 0.4687997102737427, + "learning_rate": 0.0001771910406299802, + "loss": 1.071, + "step": 4942 + }, + { + "epoch": 0.8801638176638177, + "grad_norm": 0.5064153075218201, + "learning_rate": 0.0001771821412422039, + "loss": 0.9518, + "step": 4943 + }, + { + "epoch": 0.8803418803418803, + "grad_norm": 0.6561978459358215, + "learning_rate": 0.00017717324034220385, + "loss": 1.11, + "step": 4944 + }, + { + "epoch": 0.8805199430199431, + "grad_norm": 0.5551498532295227, + "learning_rate": 0.00017716433793015454, + "loss": 0.9719, + "step": 4945 + }, + { + "epoch": 0.8806980056980057, + "grad_norm": 0.47059500217437744, + "learning_rate": 0.00017715543400623025, + "loss": 0.8891, + "step": 4946 + }, + { + "epoch": 0.8808760683760684, + "grad_norm": 0.5035740733146667, + "learning_rate": 0.00017714652857060554, + "loss": 0.9671, + "step": 4947 + }, + { + "epoch": 0.8810541310541311, + "grad_norm": 0.4599960446357727, + "learning_rate": 0.00017713762162345487, + "loss": 0.9588, + "step": 4948 + }, + { + "epoch": 0.8812321937321937, + "grad_norm": 0.5087231397628784, + "learning_rate": 0.0001771287131649527, + "loss": 1.1433, + "step": 4949 + }, + { + "epoch": 0.8814102564102564, + "grad_norm": 0.5609854459762573, + "learning_rate": 0.00017711980319527366, + "loss": 1.2022, + "step": 4950 + }, + { + "epoch": 0.8815883190883191, + "grad_norm": 0.49460700154304504, + "learning_rate": 0.00017711089171459227, + "loss": 1.019, + "step": 4951 + }, + { + "epoch": 0.8817663817663818, + "grad_norm": 0.5047259330749512, + "learning_rate": 0.00017710197872308314, + "loss": 0.8301, + "step": 4952 + }, + { + "epoch": 0.8819444444444444, + "grad_norm": 0.5784406065940857, + "learning_rate": 0.0001770930642209209, + "loss": 0.9336, + "step": 4953 + }, + { + "epoch": 0.8821225071225072, + "grad_norm": 0.5037121772766113, + "learning_rate": 0.00017708414820828022, + "loss": 1.0199, + "step": 4954 + }, + { + "epoch": 0.8823005698005698, + "grad_norm": 0.5683804750442505, + "learning_rate": 0.00017707523068533575, + "loss": 0.9758, + "step": 4955 + }, + { + "epoch": 0.8824786324786325, + "grad_norm": 0.5167922973632812, + "learning_rate": 0.0001770663116522623, + "loss": 1.0389, + "step": 4956 + }, + { + "epoch": 0.8826566951566952, + "grad_norm": 0.5813606381416321, + "learning_rate": 0.0001770573911092345, + "loss": 1.3998, + "step": 4957 + }, + { + "epoch": 0.8828347578347578, + "grad_norm": 0.5280475616455078, + "learning_rate": 0.00017704846905642723, + "loss": 1.0545, + "step": 4958 + }, + { + "epoch": 0.8830128205128205, + "grad_norm": 0.5421732068061829, + "learning_rate": 0.00017703954549401528, + "loss": 0.899, + "step": 4959 + }, + { + "epoch": 0.8831908831908832, + "grad_norm": 0.5177720189094543, + "learning_rate": 0.00017703062042217344, + "loss": 0.975, + "step": 4960 + }, + { + "epoch": 0.8833689458689459, + "grad_norm": 0.639327883720398, + "learning_rate": 0.00017702169384107666, + "loss": 1.1936, + "step": 4961 + }, + { + "epoch": 0.8835470085470085, + "grad_norm": 0.5201572179794312, + "learning_rate": 0.00017701276575089975, + "loss": 0.9891, + "step": 4962 + }, + { + "epoch": 0.8837250712250713, + "grad_norm": 0.5304145216941833, + "learning_rate": 0.00017700383615181767, + "loss": 1.0569, + "step": 4963 + }, + { + "epoch": 0.8839031339031339, + "grad_norm": 0.6068132519721985, + "learning_rate": 0.00017699490504400538, + "loss": 1.2653, + "step": 4964 + }, + { + "epoch": 0.8840811965811965, + "grad_norm": 0.597895085811615, + "learning_rate": 0.00017698597242763787, + "loss": 1.2577, + "step": 4965 + }, + { + "epoch": 0.8842592592592593, + "grad_norm": 0.5356902480125427, + "learning_rate": 0.00017697703830289017, + "loss": 1.1056, + "step": 4966 + }, + { + "epoch": 0.8844373219373219, + "grad_norm": 0.5429540872573853, + "learning_rate": 0.0001769681026699373, + "loss": 1.0951, + "step": 4967 + }, + { + "epoch": 0.8846153846153846, + "grad_norm": 0.5789309144020081, + "learning_rate": 0.00017695916552895436, + "loss": 1.0786, + "step": 4968 + }, + { + "epoch": 0.8847934472934473, + "grad_norm": 0.5621341466903687, + "learning_rate": 0.0001769502268801164, + "loss": 1.0645, + "step": 4969 + }, + { + "epoch": 0.88497150997151, + "grad_norm": 0.5879453420639038, + "learning_rate": 0.00017694128672359865, + "loss": 1.2171, + "step": 4970 + }, + { + "epoch": 0.8851495726495726, + "grad_norm": 0.5005951523780823, + "learning_rate": 0.0001769323450595762, + "loss": 1.0725, + "step": 4971 + }, + { + "epoch": 0.8853276353276354, + "grad_norm": 0.5439660549163818, + "learning_rate": 0.00017692340188822425, + "loss": 1.162, + "step": 4972 + }, + { + "epoch": 0.885505698005698, + "grad_norm": 0.6309837698936462, + "learning_rate": 0.00017691445720971802, + "loss": 1.2861, + "step": 4973 + }, + { + "epoch": 0.8856837606837606, + "grad_norm": 0.4997463822364807, + "learning_rate": 0.00017690551102423282, + "loss": 1.1887, + "step": 4974 + }, + { + "epoch": 0.8858618233618234, + "grad_norm": 0.5430852174758911, + "learning_rate": 0.00017689656333194385, + "loss": 1.1231, + "step": 4975 + }, + { + "epoch": 0.886039886039886, + "grad_norm": 0.5414215922355652, + "learning_rate": 0.00017688761413302644, + "loss": 1.2345, + "step": 4976 + }, + { + "epoch": 0.8862179487179487, + "grad_norm": 0.5594443082809448, + "learning_rate": 0.00017687866342765601, + "loss": 1.0775, + "step": 4977 + }, + { + "epoch": 0.8863960113960114, + "grad_norm": 0.5827134847640991, + "learning_rate": 0.00017686971121600787, + "loss": 1.0609, + "step": 4978 + }, + { + "epoch": 0.8865740740740741, + "grad_norm": 0.5075414776802063, + "learning_rate": 0.00017686075749825738, + "loss": 0.796, + "step": 4979 + }, + { + "epoch": 0.8867521367521367, + "grad_norm": 0.6007544994354248, + "learning_rate": 0.00017685180227458003, + "loss": 1.1716, + "step": 4980 + }, + { + "epoch": 0.8869301994301995, + "grad_norm": 0.6458030343055725, + "learning_rate": 0.00017684284554515128, + "loss": 1.1945, + "step": 4981 + }, + { + "epoch": 0.8871082621082621, + "grad_norm": 0.5519212484359741, + "learning_rate": 0.00017683388731014657, + "loss": 1.2571, + "step": 4982 + }, + { + "epoch": 0.8872863247863247, + "grad_norm": 0.5079960227012634, + "learning_rate": 0.00017682492756974146, + "loss": 1.1186, + "step": 4983 + }, + { + "epoch": 0.8874643874643875, + "grad_norm": 0.63576740026474, + "learning_rate": 0.00017681596632411147, + "loss": 1.389, + "step": 4984 + }, + { + "epoch": 0.8876424501424501, + "grad_norm": 0.43325698375701904, + "learning_rate": 0.0001768070035734322, + "loss": 0.7757, + "step": 4985 + }, + { + "epoch": 0.8878205128205128, + "grad_norm": 0.49492064118385315, + "learning_rate": 0.00017679803931787923, + "loss": 1.0096, + "step": 4986 + }, + { + "epoch": 0.8879985754985755, + "grad_norm": 0.5561224222183228, + "learning_rate": 0.00017678907355762825, + "loss": 0.952, + "step": 4987 + }, + { + "epoch": 0.8881766381766382, + "grad_norm": 0.5392457246780396, + "learning_rate": 0.00017678010629285486, + "loss": 1.0442, + "step": 4988 + }, + { + "epoch": 0.8883547008547008, + "grad_norm": 0.4659234881401062, + "learning_rate": 0.00017677113752373482, + "loss": 0.8668, + "step": 4989 + }, + { + "epoch": 0.8885327635327636, + "grad_norm": 0.5139175057411194, + "learning_rate": 0.0001767621672504438, + "loss": 0.8386, + "step": 4990 + }, + { + "epoch": 0.8887108262108262, + "grad_norm": 0.5395823121070862, + "learning_rate": 0.00017675319547315755, + "loss": 0.9754, + "step": 4991 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 0.4751867949962616, + "learning_rate": 0.0001767442221920519, + "loss": 0.8775, + "step": 4992 + }, + { + "epoch": 0.8890669515669516, + "grad_norm": 0.5728281736373901, + "learning_rate": 0.00017673524740730265, + "loss": 1.2807, + "step": 4993 + }, + { + "epoch": 0.8892450142450142, + "grad_norm": 0.5545622110366821, + "learning_rate": 0.00017672627111908558, + "loss": 1.0039, + "step": 4994 + }, + { + "epoch": 0.8894230769230769, + "grad_norm": 0.5127374529838562, + "learning_rate": 0.00017671729332757665, + "loss": 1.0505, + "step": 4995 + }, + { + "epoch": 0.8896011396011396, + "grad_norm": 0.5238714218139648, + "learning_rate": 0.00017670831403295175, + "loss": 1.1775, + "step": 4996 + }, + { + "epoch": 0.8897792022792023, + "grad_norm": 0.5610160827636719, + "learning_rate": 0.00017669933323538674, + "loss": 1.0555, + "step": 4997 + }, + { + "epoch": 0.8899572649572649, + "grad_norm": 0.5481634736061096, + "learning_rate": 0.00017669035093505762, + "loss": 1.0802, + "step": 4998 + }, + { + "epoch": 0.8901353276353277, + "grad_norm": 0.4725174307823181, + "learning_rate": 0.0001766813671321404, + "loss": 0.9611, + "step": 4999 + }, + { + "epoch": 0.8903133903133903, + "grad_norm": 0.5184635519981384, + "learning_rate": 0.0001766723818268111, + "loss": 1.1659, + "step": 5000 + }, + { + "epoch": 0.8904914529914529, + "grad_norm": 0.5503578186035156, + "learning_rate": 0.00017666339501924575, + "loss": 1.2165, + "step": 5001 + }, + { + "epoch": 0.8906695156695157, + "grad_norm": 0.5299594402313232, + "learning_rate": 0.0001766544067096204, + "loss": 1.0196, + "step": 5002 + }, + { + "epoch": 0.8908475783475783, + "grad_norm": 0.5673944354057312, + "learning_rate": 0.00017664541689811118, + "loss": 1.2058, + "step": 5003 + }, + { + "epoch": 0.8910256410256411, + "grad_norm": 0.6057320833206177, + "learning_rate": 0.00017663642558489426, + "loss": 1.0136, + "step": 5004 + }, + { + "epoch": 0.8912037037037037, + "grad_norm": 0.4767026901245117, + "learning_rate": 0.00017662743277014578, + "loss": 0.8522, + "step": 5005 + }, + { + "epoch": 0.8913817663817664, + "grad_norm": 0.5346270203590393, + "learning_rate": 0.00017661843845404192, + "loss": 1.1568, + "step": 5006 + }, + { + "epoch": 0.8915598290598291, + "grad_norm": 0.5365738868713379, + "learning_rate": 0.00017660944263675891, + "loss": 1.0488, + "step": 5007 + }, + { + "epoch": 0.8917378917378918, + "grad_norm": 0.5536269545555115, + "learning_rate": 0.00017660044531847305, + "loss": 1.1216, + "step": 5008 + }, + { + "epoch": 0.8919159544159544, + "grad_norm": 0.6325978636741638, + "learning_rate": 0.00017659144649936055, + "loss": 1.2843, + "step": 5009 + }, + { + "epoch": 0.8920940170940171, + "grad_norm": 0.5890641212463379, + "learning_rate": 0.00017658244617959777, + "loss": 1.1976, + "step": 5010 + }, + { + "epoch": 0.8922720797720798, + "grad_norm": 0.604870080947876, + "learning_rate": 0.00017657344435936107, + "loss": 1.2881, + "step": 5011 + }, + { + "epoch": 0.8924501424501424, + "grad_norm": 0.49805206060409546, + "learning_rate": 0.00017656444103882676, + "loss": 0.8998, + "step": 5012 + }, + { + "epoch": 0.8926282051282052, + "grad_norm": 0.506926953792572, + "learning_rate": 0.0001765554362181713, + "loss": 1.0731, + "step": 5013 + }, + { + "epoch": 0.8928062678062678, + "grad_norm": 0.5353260636329651, + "learning_rate": 0.0001765464298975711, + "loss": 1.0676, + "step": 5014 + }, + { + "epoch": 0.8929843304843305, + "grad_norm": 0.5641853213310242, + "learning_rate": 0.0001765374220772026, + "loss": 0.9606, + "step": 5015 + }, + { + "epoch": 0.8931623931623932, + "grad_norm": 0.5049327611923218, + "learning_rate": 0.00017652841275724233, + "loss": 1.009, + "step": 5016 + }, + { + "epoch": 0.8933404558404558, + "grad_norm": 0.6255155205726624, + "learning_rate": 0.0001765194019378668, + "loss": 1.138, + "step": 5017 + }, + { + "epoch": 0.8935185185185185, + "grad_norm": 0.5816851854324341, + "learning_rate": 0.00017651038961925247, + "loss": 1.3398, + "step": 5018 + }, + { + "epoch": 0.8936965811965812, + "grad_norm": 0.5188020467758179, + "learning_rate": 0.00017650137580157605, + "loss": 1.0126, + "step": 5019 + }, + { + "epoch": 0.8938746438746439, + "grad_norm": 0.5231554508209229, + "learning_rate": 0.00017649236048501406, + "loss": 1.0328, + "step": 5020 + }, + { + "epoch": 0.8940527065527065, + "grad_norm": 0.7638634443283081, + "learning_rate": 0.0001764833436697432, + "loss": 1.3016, + "step": 5021 + }, + { + "epoch": 0.8942307692307693, + "grad_norm": 0.5354094505310059, + "learning_rate": 0.00017647432535594008, + "loss": 1.0646, + "step": 5022 + }, + { + "epoch": 0.8944088319088319, + "grad_norm": 0.6938086748123169, + "learning_rate": 0.0001764653055437814, + "loss": 1.2051, + "step": 5023 + }, + { + "epoch": 0.8945868945868946, + "grad_norm": 0.5546849370002747, + "learning_rate": 0.00017645628423344393, + "loss": 1.0671, + "step": 5024 + }, + { + "epoch": 0.8947649572649573, + "grad_norm": 0.49294665455818176, + "learning_rate": 0.0001764472614251044, + "loss": 1.0328, + "step": 5025 + }, + { + "epoch": 0.89494301994302, + "grad_norm": 0.5965796113014221, + "learning_rate": 0.00017643823711893956, + "loss": 1.0741, + "step": 5026 + }, + { + "epoch": 0.8951210826210826, + "grad_norm": 0.4846448302268982, + "learning_rate": 0.00017642921131512626, + "loss": 1.0409, + "step": 5027 + }, + { + "epoch": 0.8952991452991453, + "grad_norm": 0.5767390131950378, + "learning_rate": 0.00017642018401384135, + "loss": 1.018, + "step": 5028 + }, + { + "epoch": 0.895477207977208, + "grad_norm": 0.503027617931366, + "learning_rate": 0.00017641115521526167, + "loss": 1.0002, + "step": 5029 + }, + { + "epoch": 0.8956552706552706, + "grad_norm": 0.6668619513511658, + "learning_rate": 0.00017640212491956412, + "loss": 1.2154, + "step": 5030 + }, + { + "epoch": 0.8958333333333334, + "grad_norm": 0.5544148683547974, + "learning_rate": 0.00017639309312692566, + "loss": 1.2701, + "step": 5031 + }, + { + "epoch": 0.896011396011396, + "grad_norm": 0.6026872992515564, + "learning_rate": 0.00017638405983752323, + "loss": 0.9335, + "step": 5032 + }, + { + "epoch": 0.8961894586894587, + "grad_norm": 0.6288694143295288, + "learning_rate": 0.00017637502505153384, + "loss": 0.9075, + "step": 5033 + }, + { + "epoch": 0.8963675213675214, + "grad_norm": 0.4890204966068268, + "learning_rate": 0.00017636598876913446, + "loss": 0.8492, + "step": 5034 + }, + { + "epoch": 0.896545584045584, + "grad_norm": 0.5746598243713379, + "learning_rate": 0.00017635695099050218, + "loss": 1.1557, + "step": 5035 + }, + { + "epoch": 0.8967236467236467, + "grad_norm": 0.5165683031082153, + "learning_rate": 0.00017634791171581405, + "loss": 1.0899, + "step": 5036 + }, + { + "epoch": 0.8969017094017094, + "grad_norm": 0.4621037244796753, + "learning_rate": 0.0001763388709452472, + "loss": 1.0457, + "step": 5037 + }, + { + "epoch": 0.8970797720797721, + "grad_norm": 0.532358705997467, + "learning_rate": 0.00017632982867897876, + "loss": 1.139, + "step": 5038 + }, + { + "epoch": 0.8972578347578347, + "grad_norm": 0.5794399976730347, + "learning_rate": 0.00017632078491718587, + "loss": 1.031, + "step": 5039 + }, + { + "epoch": 0.8974358974358975, + "grad_norm": 0.5031905174255371, + "learning_rate": 0.00017631173966004576, + "loss": 0.9508, + "step": 5040 + }, + { + "epoch": 0.8976139601139601, + "grad_norm": 0.6528840065002441, + "learning_rate": 0.00017630269290773564, + "loss": 0.9974, + "step": 5041 + }, + { + "epoch": 0.8977920227920227, + "grad_norm": 0.6007558703422546, + "learning_rate": 0.00017629364466043273, + "loss": 1.0993, + "step": 5042 + }, + { + "epoch": 0.8979700854700855, + "grad_norm": 0.5104095339775085, + "learning_rate": 0.00017628459491831437, + "loss": 0.9175, + "step": 5043 + }, + { + "epoch": 0.8981481481481481, + "grad_norm": 0.5285516977310181, + "learning_rate": 0.00017627554368155782, + "loss": 0.998, + "step": 5044 + }, + { + "epoch": 0.8983262108262108, + "grad_norm": 0.5629046559333801, + "learning_rate": 0.00017626649095034045, + "loss": 1.2021, + "step": 5045 + }, + { + "epoch": 0.8985042735042735, + "grad_norm": 0.57548987865448, + "learning_rate": 0.00017625743672483962, + "loss": 1.2076, + "step": 5046 + }, + { + "epoch": 0.8986823361823362, + "grad_norm": 0.4883024990558624, + "learning_rate": 0.0001762483810052327, + "loss": 0.9761, + "step": 5047 + }, + { + "epoch": 0.8988603988603988, + "grad_norm": 0.6378034949302673, + "learning_rate": 0.0001762393237916972, + "loss": 1.2266, + "step": 5048 + }, + { + "epoch": 0.8990384615384616, + "grad_norm": 0.5201624035835266, + "learning_rate": 0.0001762302650844105, + "loss": 1.247, + "step": 5049 + }, + { + "epoch": 0.8992165242165242, + "grad_norm": 0.5438048243522644, + "learning_rate": 0.0001762212048835501, + "loss": 0.993, + "step": 5050 + }, + { + "epoch": 0.8993945868945868, + "grad_norm": 0.5928253531455994, + "learning_rate": 0.00017621214318929354, + "loss": 1.0469, + "step": 5051 + }, + { + "epoch": 0.8995726495726496, + "grad_norm": 0.6437996625900269, + "learning_rate": 0.00017620308000181831, + "loss": 1.3136, + "step": 5052 + }, + { + "epoch": 0.8997507122507122, + "grad_norm": 0.5961456298828125, + "learning_rate": 0.00017619401532130208, + "loss": 1.1495, + "step": 5053 + }, + { + "epoch": 0.8999287749287749, + "grad_norm": 0.497388631105423, + "learning_rate": 0.0001761849491479224, + "loss": 0.7783, + "step": 5054 + }, + { + "epoch": 0.9001068376068376, + "grad_norm": 0.5984451174736023, + "learning_rate": 0.00017617588148185687, + "loss": 1.3115, + "step": 5055 + }, + { + "epoch": 0.9002849002849003, + "grad_norm": 0.549163818359375, + "learning_rate": 0.0001761668123232832, + "loss": 1.1649, + "step": 5056 + }, + { + "epoch": 0.9004629629629629, + "grad_norm": 0.5831968188285828, + "learning_rate": 0.00017615774167237903, + "loss": 1.1749, + "step": 5057 + }, + { + "epoch": 0.9006410256410257, + "grad_norm": 0.5111076235771179, + "learning_rate": 0.00017614866952932214, + "loss": 0.8936, + "step": 5058 + }, + { + "epoch": 0.9008190883190883, + "grad_norm": 0.5740947723388672, + "learning_rate": 0.00017613959589429028, + "loss": 1.2606, + "step": 5059 + }, + { + "epoch": 0.9009971509971509, + "grad_norm": 0.5881099700927734, + "learning_rate": 0.0001761305207674612, + "loss": 1.3682, + "step": 5060 + }, + { + "epoch": 0.9011752136752137, + "grad_norm": 0.5007091760635376, + "learning_rate": 0.00017612144414901268, + "loss": 0.7788, + "step": 5061 + }, + { + "epoch": 0.9013532763532763, + "grad_norm": 0.5127760171890259, + "learning_rate": 0.00017611236603912262, + "loss": 1.0519, + "step": 5062 + }, + { + "epoch": 0.9015313390313391, + "grad_norm": 0.6185184121131897, + "learning_rate": 0.00017610328643796882, + "loss": 1.1672, + "step": 5063 + }, + { + "epoch": 0.9017094017094017, + "grad_norm": 0.49707287549972534, + "learning_rate": 0.00017609420534572926, + "loss": 1.1865, + "step": 5064 + }, + { + "epoch": 0.9018874643874644, + "grad_norm": 0.5667552351951599, + "learning_rate": 0.0001760851227625818, + "loss": 1.1388, + "step": 5065 + }, + { + "epoch": 0.9020655270655271, + "grad_norm": 0.50298011302948, + "learning_rate": 0.00017607603868870442, + "loss": 0.9552, + "step": 5066 + }, + { + "epoch": 0.9022435897435898, + "grad_norm": 0.5709219574928284, + "learning_rate": 0.0001760669531242751, + "loss": 1.2636, + "step": 5067 + }, + { + "epoch": 0.9024216524216524, + "grad_norm": 0.4943496286869049, + "learning_rate": 0.0001760578660694718, + "loss": 0.8951, + "step": 5068 + }, + { + "epoch": 0.9025997150997151, + "grad_norm": 0.5475931167602539, + "learning_rate": 0.00017604877752447267, + "loss": 1.1442, + "step": 5069 + }, + { + "epoch": 0.9027777777777778, + "grad_norm": 0.5280239582061768, + "learning_rate": 0.0001760396874894557, + "loss": 0.9537, + "step": 5070 + }, + { + "epoch": 0.9029558404558404, + "grad_norm": 0.5480797290802002, + "learning_rate": 0.000176030595964599, + "loss": 1.1557, + "step": 5071 + }, + { + "epoch": 0.9031339031339032, + "grad_norm": 0.5232734680175781, + "learning_rate": 0.00017602150295008073, + "loss": 1.0219, + "step": 5072 + }, + { + "epoch": 0.9033119658119658, + "grad_norm": 0.5448359251022339, + "learning_rate": 0.000176012408446079, + "loss": 1.1964, + "step": 5073 + }, + { + "epoch": 0.9034900284900285, + "grad_norm": 0.4841914474964142, + "learning_rate": 0.00017600331245277206, + "loss": 1.0667, + "step": 5074 + }, + { + "epoch": 0.9036680911680912, + "grad_norm": 0.5407083630561829, + "learning_rate": 0.0001759942149703381, + "loss": 1.1895, + "step": 5075 + }, + { + "epoch": 0.9038461538461539, + "grad_norm": 0.5140416026115417, + "learning_rate": 0.00017598511599895534, + "loss": 0.9402, + "step": 5076 + }, + { + "epoch": 0.9040242165242165, + "grad_norm": 0.6333765983581543, + "learning_rate": 0.00017597601553880207, + "loss": 1.239, + "step": 5077 + }, + { + "epoch": 0.9042022792022792, + "grad_norm": 0.4996028244495392, + "learning_rate": 0.00017596691359005664, + "loss": 1.0259, + "step": 5078 + }, + { + "epoch": 0.9043803418803419, + "grad_norm": 0.591892421245575, + "learning_rate": 0.00017595781015289732, + "loss": 1.2148, + "step": 5079 + }, + { + "epoch": 0.9045584045584045, + "grad_norm": 0.736499011516571, + "learning_rate": 0.0001759487052275025, + "loss": 1.1373, + "step": 5080 + }, + { + "epoch": 0.9047364672364673, + "grad_norm": 0.5951572060585022, + "learning_rate": 0.00017593959881405057, + "loss": 1.1833, + "step": 5081 + }, + { + "epoch": 0.9049145299145299, + "grad_norm": 0.5092006325721741, + "learning_rate": 0.00017593049091271996, + "loss": 0.8841, + "step": 5082 + }, + { + "epoch": 0.9050925925925926, + "grad_norm": 0.5679013729095459, + "learning_rate": 0.0001759213815236891, + "loss": 1.1056, + "step": 5083 + }, + { + "epoch": 0.9052706552706553, + "grad_norm": 0.5708174109458923, + "learning_rate": 0.0001759122706471365, + "loss": 1.1952, + "step": 5084 + }, + { + "epoch": 0.905448717948718, + "grad_norm": 0.5726733803749084, + "learning_rate": 0.00017590315828324067, + "loss": 1.1013, + "step": 5085 + }, + { + "epoch": 0.9056267806267806, + "grad_norm": 0.5821273326873779, + "learning_rate": 0.00017589404443218008, + "loss": 1.2323, + "step": 5086 + }, + { + "epoch": 0.9058048433048433, + "grad_norm": 0.5811445713043213, + "learning_rate": 0.00017588492909413337, + "loss": 1.2241, + "step": 5087 + }, + { + "epoch": 0.905982905982906, + "grad_norm": 0.5377545952796936, + "learning_rate": 0.0001758758122692791, + "loss": 0.9777, + "step": 5088 + }, + { + "epoch": 0.9061609686609686, + "grad_norm": 0.5985640287399292, + "learning_rate": 0.0001758666939577959, + "loss": 0.9737, + "step": 5089 + }, + { + "epoch": 0.9063390313390314, + "grad_norm": 0.6038222908973694, + "learning_rate": 0.00017585757415986247, + "loss": 1.2116, + "step": 5090 + }, + { + "epoch": 0.906517094017094, + "grad_norm": 0.6752246022224426, + "learning_rate": 0.00017584845287565743, + "loss": 1.1975, + "step": 5091 + }, + { + "epoch": 0.9066951566951567, + "grad_norm": 0.5400625467300415, + "learning_rate": 0.0001758393301053595, + "loss": 0.9669, + "step": 5092 + }, + { + "epoch": 0.9068732193732194, + "grad_norm": 0.5637784004211426, + "learning_rate": 0.00017583020584914746, + "loss": 1.2672, + "step": 5093 + }, + { + "epoch": 0.907051282051282, + "grad_norm": 0.4825877249240875, + "learning_rate": 0.00017582108010720006, + "loss": 0.9719, + "step": 5094 + }, + { + "epoch": 0.9072293447293447, + "grad_norm": 0.49902790784835815, + "learning_rate": 0.00017581195287969613, + "loss": 0.7941, + "step": 5095 + }, + { + "epoch": 0.9074074074074074, + "grad_norm": 0.5991541743278503, + "learning_rate": 0.0001758028241668144, + "loss": 1.049, + "step": 5096 + }, + { + "epoch": 0.9075854700854701, + "grad_norm": 0.5788859724998474, + "learning_rate": 0.00017579369396873384, + "loss": 1.0318, + "step": 5097 + }, + { + "epoch": 0.9077635327635327, + "grad_norm": 0.5914160013198853, + "learning_rate": 0.0001757845622856333, + "loss": 1.1007, + "step": 5098 + }, + { + "epoch": 0.9079415954415955, + "grad_norm": 0.5361711382865906, + "learning_rate": 0.00017577542911769166, + "loss": 1.0694, + "step": 5099 + }, + { + "epoch": 0.9081196581196581, + "grad_norm": 0.5752849578857422, + "learning_rate": 0.00017576629446508792, + "loss": 1.1184, + "step": 5100 + }, + { + "epoch": 0.9082977207977208, + "grad_norm": 0.6042249798774719, + "learning_rate": 0.000175757158328001, + "loss": 1.2808, + "step": 5101 + }, + { + "epoch": 0.9084757834757835, + "grad_norm": 0.508352518081665, + "learning_rate": 0.00017574802070661, + "loss": 1.0038, + "step": 5102 + }, + { + "epoch": 0.9086538461538461, + "grad_norm": 0.5667358040809631, + "learning_rate": 0.00017573888160109385, + "loss": 1.0208, + "step": 5103 + }, + { + "epoch": 0.9088319088319088, + "grad_norm": 0.653619647026062, + "learning_rate": 0.00017572974101163165, + "loss": 1.2053, + "step": 5104 + }, + { + "epoch": 0.9090099715099715, + "grad_norm": 0.5069597363471985, + "learning_rate": 0.00017572059893840246, + "loss": 0.8634, + "step": 5105 + }, + { + "epoch": 0.9091880341880342, + "grad_norm": 0.6160602569580078, + "learning_rate": 0.00017571145538158547, + "loss": 1.2626, + "step": 5106 + }, + { + "epoch": 0.9093660968660968, + "grad_norm": 0.6335833668708801, + "learning_rate": 0.00017570231034135978, + "loss": 1.3381, + "step": 5107 + }, + { + "epoch": 0.9095441595441596, + "grad_norm": 0.5140398740768433, + "learning_rate": 0.00017569316381790454, + "loss": 1.1258, + "step": 5108 + }, + { + "epoch": 0.9097222222222222, + "grad_norm": 0.5682975649833679, + "learning_rate": 0.00017568401581139905, + "loss": 1.3367, + "step": 5109 + }, + { + "epoch": 0.9099002849002849, + "grad_norm": 0.49765729904174805, + "learning_rate": 0.00017567486632202246, + "loss": 1.1891, + "step": 5110 + }, + { + "epoch": 0.9100783475783476, + "grad_norm": 0.5139224529266357, + "learning_rate": 0.00017566571534995406, + "loss": 0.9768, + "step": 5111 + }, + { + "epoch": 0.9102564102564102, + "grad_norm": 0.5510922074317932, + "learning_rate": 0.00017565656289537316, + "loss": 1.1552, + "step": 5112 + }, + { + "epoch": 0.9104344729344729, + "grad_norm": 0.6243364810943604, + "learning_rate": 0.00017564740895845908, + "loss": 1.1341, + "step": 5113 + }, + { + "epoch": 0.9106125356125356, + "grad_norm": 0.5334977507591248, + "learning_rate": 0.00017563825353939116, + "loss": 1.0894, + "step": 5114 + }, + { + "epoch": 0.9107905982905983, + "grad_norm": 0.5195826292037964, + "learning_rate": 0.00017562909663834878, + "loss": 1.1011, + "step": 5115 + }, + { + "epoch": 0.9109686609686609, + "grad_norm": 0.5298168063163757, + "learning_rate": 0.00017561993825551138, + "loss": 1.0079, + "step": 5116 + }, + { + "epoch": 0.9111467236467237, + "grad_norm": 0.5858965516090393, + "learning_rate": 0.00017561077839105835, + "loss": 1.2746, + "step": 5117 + }, + { + "epoch": 0.9113247863247863, + "grad_norm": 0.5572476387023926, + "learning_rate": 0.0001756016170451692, + "loss": 0.8169, + "step": 5118 + }, + { + "epoch": 0.9115028490028491, + "grad_norm": 0.5247095823287964, + "learning_rate": 0.0001755924542180234, + "loss": 1.1206, + "step": 5119 + }, + { + "epoch": 0.9116809116809117, + "grad_norm": 0.5605118274688721, + "learning_rate": 0.0001755832899098005, + "loss": 1.371, + "step": 5120 + }, + { + "epoch": 0.9118589743589743, + "grad_norm": 0.5732316970825195, + "learning_rate": 0.00017557412412068005, + "loss": 1.1248, + "step": 5121 + }, + { + "epoch": 0.9120370370370371, + "grad_norm": 0.6167279481887817, + "learning_rate": 0.0001755649568508416, + "loss": 0.94, + "step": 5122 + }, + { + "epoch": 0.9122150997150997, + "grad_norm": 0.5497499108314514, + "learning_rate": 0.00017555578810046483, + "loss": 1.0112, + "step": 5123 + }, + { + "epoch": 0.9123931623931624, + "grad_norm": 0.540762186050415, + "learning_rate": 0.00017554661786972931, + "loss": 1.1058, + "step": 5124 + }, + { + "epoch": 0.9125712250712251, + "grad_norm": 0.5943556427955627, + "learning_rate": 0.0001755374461588148, + "loss": 0.9086, + "step": 5125 + }, + { + "epoch": 0.9127492877492878, + "grad_norm": 0.5300756692886353, + "learning_rate": 0.0001755282729679009, + "loss": 1.1566, + "step": 5126 + }, + { + "epoch": 0.9129273504273504, + "grad_norm": 0.5390434861183167, + "learning_rate": 0.00017551909829716743, + "loss": 1.1395, + "step": 5127 + }, + { + "epoch": 0.9131054131054132, + "grad_norm": 0.627434492111206, + "learning_rate": 0.00017550992214679405, + "loss": 1.1537, + "step": 5128 + }, + { + "epoch": 0.9132834757834758, + "grad_norm": 0.4806903302669525, + "learning_rate": 0.00017550074451696063, + "loss": 0.7905, + "step": 5129 + }, + { + "epoch": 0.9134615384615384, + "grad_norm": 0.5714817047119141, + "learning_rate": 0.00017549156540784696, + "loss": 1.1042, + "step": 5130 + }, + { + "epoch": 0.9136396011396012, + "grad_norm": 0.5839236378669739, + "learning_rate": 0.0001754823848196329, + "loss": 1.0383, + "step": 5131 + }, + { + "epoch": 0.9138176638176638, + "grad_norm": 0.6089872717857361, + "learning_rate": 0.0001754732027524983, + "loss": 0.9399, + "step": 5132 + }, + { + "epoch": 0.9139957264957265, + "grad_norm": 0.4937956631183624, + "learning_rate": 0.00017546401920662307, + "loss": 0.7382, + "step": 5133 + }, + { + "epoch": 0.9141737891737892, + "grad_norm": 0.5918676257133484, + "learning_rate": 0.00017545483418218716, + "loss": 1.2207, + "step": 5134 + }, + { + "epoch": 0.9143518518518519, + "grad_norm": 0.5825346112251282, + "learning_rate": 0.0001754456476793705, + "loss": 0.9669, + "step": 5135 + }, + { + "epoch": 0.9145299145299145, + "grad_norm": 0.49829617142677307, + "learning_rate": 0.0001754364596983531, + "loss": 1.2247, + "step": 5136 + }, + { + "epoch": 0.9147079772079773, + "grad_norm": 0.5128271579742432, + "learning_rate": 0.00017542727023931497, + "loss": 0.9563, + "step": 5137 + }, + { + "epoch": 0.9148860398860399, + "grad_norm": 0.5789414644241333, + "learning_rate": 0.00017541807930243622, + "loss": 1.22, + "step": 5138 + }, + { + "epoch": 0.9150641025641025, + "grad_norm": 0.44155433773994446, + "learning_rate": 0.00017540888688789683, + "loss": 0.9897, + "step": 5139 + }, + { + "epoch": 0.9152421652421653, + "grad_norm": 0.550464391708374, + "learning_rate": 0.00017539969299587696, + "loss": 1.0624, + "step": 5140 + }, + { + "epoch": 0.9154202279202279, + "grad_norm": 0.5019831657409668, + "learning_rate": 0.0001753904976265567, + "loss": 0.9045, + "step": 5141 + }, + { + "epoch": 0.9155982905982906, + "grad_norm": 0.589658796787262, + "learning_rate": 0.0001753813007801163, + "loss": 1.0454, + "step": 5142 + }, + { + "epoch": 0.9157763532763533, + "grad_norm": 0.5945459008216858, + "learning_rate": 0.00017537210245673586, + "loss": 1.0042, + "step": 5143 + }, + { + "epoch": 0.915954415954416, + "grad_norm": 0.5409809947013855, + "learning_rate": 0.00017536290265659566, + "loss": 1.0609, + "step": 5144 + }, + { + "epoch": 0.9161324786324786, + "grad_norm": 0.5302975177764893, + "learning_rate": 0.00017535370137987597, + "loss": 1.1394, + "step": 5145 + }, + { + "epoch": 0.9163105413105413, + "grad_norm": 0.5253351330757141, + "learning_rate": 0.00017534449862675698, + "loss": 1.2249, + "step": 5146 + }, + { + "epoch": 0.916488603988604, + "grad_norm": 0.6363829970359802, + "learning_rate": 0.00017533529439741908, + "loss": 1.1333, + "step": 5147 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.4703354835510254, + "learning_rate": 0.0001753260886920426, + "loss": 0.9971, + "step": 5148 + }, + { + "epoch": 0.9168447293447294, + "grad_norm": 0.6394907236099243, + "learning_rate": 0.00017531688151080786, + "loss": 1.5942, + "step": 5149 + }, + { + "epoch": 0.917022792022792, + "grad_norm": 0.5573459267616272, + "learning_rate": 0.00017530767285389527, + "loss": 0.9669, + "step": 5150 + }, + { + "epoch": 0.9172008547008547, + "grad_norm": 0.5000962615013123, + "learning_rate": 0.00017529846272148532, + "loss": 1.2151, + "step": 5151 + }, + { + "epoch": 0.9173789173789174, + "grad_norm": 0.5550395846366882, + "learning_rate": 0.0001752892511137584, + "loss": 1.1765, + "step": 5152 + }, + { + "epoch": 0.91755698005698, + "grad_norm": 0.5461394786834717, + "learning_rate": 0.00017528003803089496, + "loss": 1.1136, + "step": 5153 + }, + { + "epoch": 0.9177350427350427, + "grad_norm": 0.5512672662734985, + "learning_rate": 0.00017527082347307558, + "loss": 1.1727, + "step": 5154 + }, + { + "epoch": 0.9179131054131054, + "grad_norm": 0.5210778713226318, + "learning_rate": 0.0001752616074404808, + "loss": 1.09, + "step": 5155 + }, + { + "epoch": 0.9180911680911681, + "grad_norm": 0.5214943289756775, + "learning_rate": 0.00017525238993329115, + "loss": 0.9654, + "step": 5156 + }, + { + "epoch": 0.9182692307692307, + "grad_norm": 0.5822862386703491, + "learning_rate": 0.00017524317095168724, + "loss": 1.0951, + "step": 5157 + }, + { + "epoch": 0.9184472934472935, + "grad_norm": 0.43948012590408325, + "learning_rate": 0.0001752339504958497, + "loss": 0.6984, + "step": 5158 + }, + { + "epoch": 0.9186253561253561, + "grad_norm": 0.5024449229240417, + "learning_rate": 0.00017522472856595916, + "loss": 0.983, + "step": 5159 + }, + { + "epoch": 0.9188034188034188, + "grad_norm": 0.5815144181251526, + "learning_rate": 0.00017521550516219636, + "loss": 0.9784, + "step": 5160 + }, + { + "epoch": 0.9189814814814815, + "grad_norm": 0.5519825220108032, + "learning_rate": 0.00017520628028474197, + "loss": 1.064, + "step": 5161 + }, + { + "epoch": 0.9191595441595442, + "grad_norm": 0.5615749955177307, + "learning_rate": 0.00017519705393377675, + "loss": 1.1284, + "step": 5162 + }, + { + "epoch": 0.9193376068376068, + "grad_norm": 0.5929917693138123, + "learning_rate": 0.00017518782610948148, + "loss": 1.1221, + "step": 5163 + }, + { + "epoch": 0.9195156695156695, + "grad_norm": 0.7116361856460571, + "learning_rate": 0.00017517859681203692, + "loss": 1.0188, + "step": 5164 + }, + { + "epoch": 0.9196937321937322, + "grad_norm": 0.5095893740653992, + "learning_rate": 0.00017516936604162396, + "loss": 1.0724, + "step": 5165 + }, + { + "epoch": 0.9198717948717948, + "grad_norm": 0.5701385736465454, + "learning_rate": 0.00017516013379842337, + "loss": 1.0572, + "step": 5166 + }, + { + "epoch": 0.9200498575498576, + "grad_norm": 0.518412709236145, + "learning_rate": 0.00017515090008261613, + "loss": 1.0514, + "step": 5167 + }, + { + "epoch": 0.9202279202279202, + "grad_norm": 0.5324261784553528, + "learning_rate": 0.00017514166489438312, + "loss": 1.1708, + "step": 5168 + }, + { + "epoch": 0.9204059829059829, + "grad_norm": 0.5640990138053894, + "learning_rate": 0.00017513242823390525, + "loss": 1.2846, + "step": 5169 + }, + { + "epoch": 0.9205840455840456, + "grad_norm": 0.510352373123169, + "learning_rate": 0.00017512319010136356, + "loss": 1.0763, + "step": 5170 + }, + { + "epoch": 0.9207621082621082, + "grad_norm": 0.4994175136089325, + "learning_rate": 0.00017511395049693898, + "loss": 0.9665, + "step": 5171 + }, + { + "epoch": 0.9209401709401709, + "grad_norm": 0.43196994066238403, + "learning_rate": 0.00017510470942081258, + "loss": 0.761, + "step": 5172 + }, + { + "epoch": 0.9211182336182336, + "grad_norm": 0.558977484703064, + "learning_rate": 0.00017509546687316543, + "loss": 1.0758, + "step": 5173 + }, + { + "epoch": 0.9212962962962963, + "grad_norm": 0.573302149772644, + "learning_rate": 0.0001750862228541786, + "loss": 0.9635, + "step": 5174 + }, + { + "epoch": 0.9214743589743589, + "grad_norm": 0.5083786845207214, + "learning_rate": 0.00017507697736403321, + "loss": 1.0311, + "step": 5175 + }, + { + "epoch": 0.9216524216524217, + "grad_norm": 0.5478954911231995, + "learning_rate": 0.00017506773040291043, + "loss": 1.074, + "step": 5176 + }, + { + "epoch": 0.9218304843304843, + "grad_norm": 0.522376537322998, + "learning_rate": 0.00017505848197099137, + "loss": 1.1162, + "step": 5177 + }, + { + "epoch": 0.9220085470085471, + "grad_norm": 0.5946292281150818, + "learning_rate": 0.0001750492320684573, + "loss": 0.9494, + "step": 5178 + }, + { + "epoch": 0.9221866096866097, + "grad_norm": 0.5423247814178467, + "learning_rate": 0.00017503998069548943, + "loss": 1.0558, + "step": 5179 + }, + { + "epoch": 0.9223646723646723, + "grad_norm": 0.49960651993751526, + "learning_rate": 0.000175030727852269, + "loss": 1.0748, + "step": 5180 + }, + { + "epoch": 0.9225427350427351, + "grad_norm": 0.6066586375236511, + "learning_rate": 0.00017502147353897732, + "loss": 1.2066, + "step": 5181 + }, + { + "epoch": 0.9227207977207977, + "grad_norm": 0.57244473695755, + "learning_rate": 0.00017501221775579576, + "loss": 1.048, + "step": 5182 + }, + { + "epoch": 0.9228988603988604, + "grad_norm": 0.512464165687561, + "learning_rate": 0.00017500296050290557, + "loss": 1.1405, + "step": 5183 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.5380734801292419, + "learning_rate": 0.00017499370178048818, + "loss": 1.0641, + "step": 5184 + }, + { + "epoch": 0.9232549857549858, + "grad_norm": 0.47102874517440796, + "learning_rate": 0.000174984441588725, + "loss": 0.7948, + "step": 5185 + }, + { + "epoch": 0.9234330484330484, + "grad_norm": 0.6702211499214172, + "learning_rate": 0.00017497517992779747, + "loss": 1.3009, + "step": 5186 + }, + { + "epoch": 0.9236111111111112, + "grad_norm": 0.4685834050178528, + "learning_rate": 0.000174965916797887, + "loss": 0.8136, + "step": 5187 + }, + { + "epoch": 0.9237891737891738, + "grad_norm": 0.5414277911186218, + "learning_rate": 0.00017495665219917513, + "loss": 0.9708, + "step": 5188 + }, + { + "epoch": 0.9239672364672364, + "grad_norm": 0.5253050923347473, + "learning_rate": 0.0001749473861318434, + "loss": 1.0691, + "step": 5189 + }, + { + "epoch": 0.9241452991452992, + "grad_norm": 0.6009906530380249, + "learning_rate": 0.00017493811859607328, + "loss": 1.2023, + "step": 5190 + }, + { + "epoch": 0.9243233618233618, + "grad_norm": 0.5519336462020874, + "learning_rate": 0.00017492884959204643, + "loss": 1.189, + "step": 5191 + }, + { + "epoch": 0.9245014245014245, + "grad_norm": 0.5024857521057129, + "learning_rate": 0.0001749195791199444, + "loss": 0.8685, + "step": 5192 + }, + { + "epoch": 0.9246794871794872, + "grad_norm": 0.5735679864883423, + "learning_rate": 0.00017491030717994887, + "loss": 1.1903, + "step": 5193 + }, + { + "epoch": 0.9248575498575499, + "grad_norm": 0.5338658094406128, + "learning_rate": 0.00017490103377224147, + "loss": 1.0442, + "step": 5194 + }, + { + "epoch": 0.9250356125356125, + "grad_norm": 0.46669119596481323, + "learning_rate": 0.0001748917588970039, + "loss": 0.6343, + "step": 5195 + }, + { + "epoch": 0.9252136752136753, + "grad_norm": 0.510910153388977, + "learning_rate": 0.00017488248255441793, + "loss": 0.9334, + "step": 5196 + }, + { + "epoch": 0.9253917378917379, + "grad_norm": 0.5732216238975525, + "learning_rate": 0.00017487320474466524, + "loss": 1.0483, + "step": 5197 + }, + { + "epoch": 0.9255698005698005, + "grad_norm": 0.5864318609237671, + "learning_rate": 0.00017486392546792762, + "loss": 1.0669, + "step": 5198 + }, + { + "epoch": 0.9257478632478633, + "grad_norm": 0.5074281096458435, + "learning_rate": 0.00017485464472438692, + "loss": 1.0636, + "step": 5199 + }, + { + "epoch": 0.9259259259259259, + "grad_norm": 0.5833215117454529, + "learning_rate": 0.00017484536251422496, + "loss": 1.2005, + "step": 5200 + }, + { + "epoch": 0.9261039886039886, + "grad_norm": 0.5624990463256836, + "learning_rate": 0.0001748360788376236, + "loss": 1.1623, + "step": 5201 + }, + { + "epoch": 0.9262820512820513, + "grad_norm": 0.5618230104446411, + "learning_rate": 0.00017482679369476472, + "loss": 1.0495, + "step": 5202 + }, + { + "epoch": 0.926460113960114, + "grad_norm": 0.6254985332489014, + "learning_rate": 0.00017481750708583024, + "loss": 0.9521, + "step": 5203 + }, + { + "epoch": 0.9266381766381766, + "grad_norm": 0.5488203763961792, + "learning_rate": 0.00017480821901100216, + "loss": 1.0689, + "step": 5204 + }, + { + "epoch": 0.9268162393162394, + "grad_norm": 0.6157993674278259, + "learning_rate": 0.00017479892947046245, + "loss": 1.2852, + "step": 5205 + }, + { + "epoch": 0.926994301994302, + "grad_norm": 0.49653390049934387, + "learning_rate": 0.00017478963846439305, + "loss": 0.8616, + "step": 5206 + }, + { + "epoch": 0.9271723646723646, + "grad_norm": 0.5079081058502197, + "learning_rate": 0.00017478034599297603, + "loss": 1.0192, + "step": 5207 + }, + { + "epoch": 0.9273504273504274, + "grad_norm": 0.5392495393753052, + "learning_rate": 0.00017477105205639354, + "loss": 1.115, + "step": 5208 + }, + { + "epoch": 0.92752849002849, + "grad_norm": 0.5336191654205322, + "learning_rate": 0.00017476175665482756, + "loss": 1.1892, + "step": 5209 + }, + { + "epoch": 0.9277065527065527, + "grad_norm": 0.631712019443512, + "learning_rate": 0.00017475245978846026, + "loss": 0.9619, + "step": 5210 + }, + { + "epoch": 0.9278846153846154, + "grad_norm": 0.5123951435089111, + "learning_rate": 0.0001747431614574738, + "loss": 1.1477, + "step": 5211 + }, + { + "epoch": 0.9280626780626781, + "grad_norm": 0.5045743584632874, + "learning_rate": 0.00017473386166205038, + "loss": 0.9749, + "step": 5212 + }, + { + "epoch": 0.9282407407407407, + "grad_norm": 0.5296525359153748, + "learning_rate": 0.00017472456040237217, + "loss": 1.0736, + "step": 5213 + }, + { + "epoch": 0.9284188034188035, + "grad_norm": 0.6304933428764343, + "learning_rate": 0.00017471525767862145, + "loss": 1.2444, + "step": 5214 + }, + { + "epoch": 0.9285968660968661, + "grad_norm": 0.4851958155632019, + "learning_rate": 0.00017470595349098044, + "loss": 0.9049, + "step": 5215 + }, + { + "epoch": 0.9287749287749287, + "grad_norm": 0.5730679631233215, + "learning_rate": 0.00017469664783963148, + "loss": 1.0773, + "step": 5216 + }, + { + "epoch": 0.9289529914529915, + "grad_norm": 0.6020415425300598, + "learning_rate": 0.00017468734072475684, + "loss": 1.3247, + "step": 5217 + }, + { + "epoch": 0.9291310541310541, + "grad_norm": 0.47981077432632446, + "learning_rate": 0.00017467803214653893, + "loss": 1.0009, + "step": 5218 + }, + { + "epoch": 0.9293091168091168, + "grad_norm": 0.5787527561187744, + "learning_rate": 0.0001746687221051601, + "loss": 1.2523, + "step": 5219 + }, + { + "epoch": 0.9294871794871795, + "grad_norm": 0.4495891332626343, + "learning_rate": 0.00017465941060080278, + "loss": 0.7364, + "step": 5220 + }, + { + "epoch": 0.9296652421652422, + "grad_norm": 0.5721768140792847, + "learning_rate": 0.0001746500976336494, + "loss": 1.015, + "step": 5221 + }, + { + "epoch": 0.9298433048433048, + "grad_norm": 0.5500208735466003, + "learning_rate": 0.0001746407832038824, + "loss": 1.053, + "step": 5222 + }, + { + "epoch": 0.9300213675213675, + "grad_norm": 0.5784386992454529, + "learning_rate": 0.00017463146731168437, + "loss": 0.9784, + "step": 5223 + }, + { + "epoch": 0.9301994301994302, + "grad_norm": 0.4960322082042694, + "learning_rate": 0.00017462214995723772, + "loss": 0.8674, + "step": 5224 + }, + { + "epoch": 0.9303774928774928, + "grad_norm": 0.5005537271499634, + "learning_rate": 0.00017461283114072508, + "loss": 1.0486, + "step": 5225 + }, + { + "epoch": 0.9305555555555556, + "grad_norm": 0.5064167380332947, + "learning_rate": 0.000174603510862329, + "loss": 0.9722, + "step": 5226 + }, + { + "epoch": 0.9307336182336182, + "grad_norm": 0.583558976650238, + "learning_rate": 0.0001745941891222321, + "loss": 0.9957, + "step": 5227 + }, + { + "epoch": 0.9309116809116809, + "grad_norm": 0.4982515871524811, + "learning_rate": 0.00017458486592061704, + "loss": 0.958, + "step": 5228 + }, + { + "epoch": 0.9310897435897436, + "grad_norm": 0.526549756526947, + "learning_rate": 0.0001745755412576664, + "loss": 1.1172, + "step": 5229 + }, + { + "epoch": 0.9312678062678063, + "grad_norm": 0.6129719018936157, + "learning_rate": 0.000174566215133563, + "loss": 1.2524, + "step": 5230 + }, + { + "epoch": 0.9314458689458689, + "grad_norm": 0.5385653972625732, + "learning_rate": 0.00017455688754848948, + "loss": 1.1655, + "step": 5231 + }, + { + "epoch": 0.9316239316239316, + "grad_norm": 0.5646410584449768, + "learning_rate": 0.0001745475585026287, + "loss": 0.9026, + "step": 5232 + }, + { + "epoch": 0.9318019943019943, + "grad_norm": 0.549223780632019, + "learning_rate": 0.0001745382279961633, + "loss": 0.804, + "step": 5233 + }, + { + "epoch": 0.9319800569800569, + "grad_norm": 0.48547953367233276, + "learning_rate": 0.0001745288960292762, + "loss": 1.0224, + "step": 5234 + }, + { + "epoch": 0.9321581196581197, + "grad_norm": 0.5260967016220093, + "learning_rate": 0.00017451956260215016, + "loss": 0.9688, + "step": 5235 + }, + { + "epoch": 0.9323361823361823, + "grad_norm": 0.6261999011039734, + "learning_rate": 0.00017451022771496812, + "loss": 1.2539, + "step": 5236 + }, + { + "epoch": 0.9325142450142451, + "grad_norm": 0.5801421999931335, + "learning_rate": 0.00017450089136791298, + "loss": 1.11, + "step": 5237 + }, + { + "epoch": 0.9326923076923077, + "grad_norm": 0.5833573937416077, + "learning_rate": 0.0001744915535611676, + "loss": 0.9328, + "step": 5238 + }, + { + "epoch": 0.9328703703703703, + "grad_norm": 0.5422634482383728, + "learning_rate": 0.00017448221429491496, + "loss": 1.034, + "step": 5239 + }, + { + "epoch": 0.9330484330484331, + "grad_norm": 0.5105658769607544, + "learning_rate": 0.00017447287356933808, + "loss": 0.8924, + "step": 5240 + }, + { + "epoch": 0.9332264957264957, + "grad_norm": 0.5114831924438477, + "learning_rate": 0.00017446353138461995, + "loss": 0.9328, + "step": 5241 + }, + { + "epoch": 0.9334045584045584, + "grad_norm": 0.5105039477348328, + "learning_rate": 0.00017445418774094358, + "loss": 1.0468, + "step": 5242 + }, + { + "epoch": 0.9335826210826211, + "grad_norm": 0.593250036239624, + "learning_rate": 0.00017444484263849208, + "loss": 1.0603, + "step": 5243 + }, + { + "epoch": 0.9337606837606838, + "grad_norm": 0.600788414478302, + "learning_rate": 0.00017443549607744853, + "loss": 1.1506, + "step": 5244 + }, + { + "epoch": 0.9339387464387464, + "grad_norm": 0.5394418239593506, + "learning_rate": 0.00017442614805799605, + "loss": 1.038, + "step": 5245 + }, + { + "epoch": 0.9341168091168092, + "grad_norm": 0.5446375608444214, + "learning_rate": 0.00017441679858031786, + "loss": 1.079, + "step": 5246 + }, + { + "epoch": 0.9342948717948718, + "grad_norm": 0.5859794616699219, + "learning_rate": 0.00017440744764459702, + "loss": 1.1453, + "step": 5247 + }, + { + "epoch": 0.9344729344729344, + "grad_norm": 0.4899081289768219, + "learning_rate": 0.00017439809525101688, + "loss": 1.163, + "step": 5248 + }, + { + "epoch": 0.9346509971509972, + "grad_norm": 0.652846097946167, + "learning_rate": 0.00017438874139976055, + "loss": 1.1819, + "step": 5249 + }, + { + "epoch": 0.9348290598290598, + "grad_norm": 0.5402514934539795, + "learning_rate": 0.00017437938609101138, + "loss": 1.0159, + "step": 5250 + }, + { + "epoch": 0.9350071225071225, + "grad_norm": 0.565864086151123, + "learning_rate": 0.00017437002932495265, + "loss": 1.1121, + "step": 5251 + }, + { + "epoch": 0.9351851851851852, + "grad_norm": 0.611786425113678, + "learning_rate": 0.0001743606711017677, + "loss": 1.2511, + "step": 5252 + }, + { + "epoch": 0.9353632478632479, + "grad_norm": 0.5706882476806641, + "learning_rate": 0.00017435131142163988, + "loss": 1.128, + "step": 5253 + }, + { + "epoch": 0.9355413105413105, + "grad_norm": 0.5369367003440857, + "learning_rate": 0.00017434195028475253, + "loss": 1.0562, + "step": 5254 + }, + { + "epoch": 0.9357193732193733, + "grad_norm": 0.49957552552223206, + "learning_rate": 0.0001743325876912891, + "loss": 1.0568, + "step": 5255 + }, + { + "epoch": 0.9358974358974359, + "grad_norm": 0.5398106575012207, + "learning_rate": 0.00017432322364143305, + "loss": 1.1502, + "step": 5256 + }, + { + "epoch": 0.9360754985754985, + "grad_norm": 0.6522027254104614, + "learning_rate": 0.00017431385813536783, + "loss": 1.0591, + "step": 5257 + }, + { + "epoch": 0.9362535612535613, + "grad_norm": 0.5872012972831726, + "learning_rate": 0.00017430449117327693, + "loss": 1.3737, + "step": 5258 + }, + { + "epoch": 0.9364316239316239, + "grad_norm": 0.5124474167823792, + "learning_rate": 0.00017429512275534382, + "loss": 1.0727, + "step": 5259 + }, + { + "epoch": 0.9366096866096866, + "grad_norm": 0.5103365778923035, + "learning_rate": 0.00017428575288175218, + "loss": 1.0339, + "step": 5260 + }, + { + "epoch": 0.9367877492877493, + "grad_norm": 0.585483729839325, + "learning_rate": 0.0001742763815526855, + "loss": 1.1844, + "step": 5261 + }, + { + "epoch": 0.936965811965812, + "grad_norm": 0.5855562090873718, + "learning_rate": 0.00017426700876832746, + "loss": 1.3234, + "step": 5262 + }, + { + "epoch": 0.9371438746438746, + "grad_norm": 0.5774588584899902, + "learning_rate": 0.00017425763452886162, + "loss": 1.0937, + "step": 5263 + }, + { + "epoch": 0.9373219373219374, + "grad_norm": 0.5718343257904053, + "learning_rate": 0.00017424825883447168, + "loss": 1.0783, + "step": 5264 + }, + { + "epoch": 0.9375, + "grad_norm": 0.5414558053016663, + "learning_rate": 0.00017423888168534136, + "loss": 1.1244, + "step": 5265 + }, + { + "epoch": 0.9376780626780626, + "grad_norm": 0.5818275809288025, + "learning_rate": 0.00017422950308165438, + "loss": 1.247, + "step": 5266 + }, + { + "epoch": 0.9378561253561254, + "grad_norm": 0.586398184299469, + "learning_rate": 0.00017422012302359448, + "loss": 1.0515, + "step": 5267 + }, + { + "epoch": 0.938034188034188, + "grad_norm": 0.5236606001853943, + "learning_rate": 0.00017421074151134544, + "loss": 1.1907, + "step": 5268 + }, + { + "epoch": 0.9382122507122507, + "grad_norm": 0.5108010172843933, + "learning_rate": 0.0001742013585450911, + "loss": 1.1125, + "step": 5269 + }, + { + "epoch": 0.9383903133903134, + "grad_norm": 0.4956454038619995, + "learning_rate": 0.00017419197412501527, + "loss": 1.0305, + "step": 5270 + }, + { + "epoch": 0.9385683760683761, + "grad_norm": 0.5432302951812744, + "learning_rate": 0.0001741825882513018, + "loss": 1.1946, + "step": 5271 + }, + { + "epoch": 0.9387464387464387, + "grad_norm": 0.5119295716285706, + "learning_rate": 0.00017417320092413463, + "loss": 0.875, + "step": 5272 + }, + { + "epoch": 0.9389245014245015, + "grad_norm": 0.49740248918533325, + "learning_rate": 0.0001741638121436977, + "loss": 1.1093, + "step": 5273 + }, + { + "epoch": 0.9391025641025641, + "grad_norm": 0.5069027543067932, + "learning_rate": 0.00017415442191017491, + "loss": 1.2498, + "step": 5274 + }, + { + "epoch": 0.9392806267806267, + "grad_norm": 0.570264995098114, + "learning_rate": 0.00017414503022375027, + "loss": 1.0192, + "step": 5275 + }, + { + "epoch": 0.9394586894586895, + "grad_norm": 0.48129352927207947, + "learning_rate": 0.00017413563708460776, + "loss": 0.8467, + "step": 5276 + }, + { + "epoch": 0.9396367521367521, + "grad_norm": 0.5214534401893616, + "learning_rate": 0.00017412624249293148, + "loss": 0.9723, + "step": 5277 + }, + { + "epoch": 0.9398148148148148, + "grad_norm": 0.5150161385536194, + "learning_rate": 0.00017411684644890544, + "loss": 1.0906, + "step": 5278 + }, + { + "epoch": 0.9399928774928775, + "grad_norm": 0.5695852637290955, + "learning_rate": 0.00017410744895271377, + "loss": 1.2891, + "step": 5279 + }, + { + "epoch": 0.9401709401709402, + "grad_norm": 0.5613594651222229, + "learning_rate": 0.00017409805000454055, + "loss": 1.1373, + "step": 5280 + }, + { + "epoch": 0.9403490028490028, + "grad_norm": 0.5134239196777344, + "learning_rate": 0.00017408864960457004, + "loss": 1.1081, + "step": 5281 + }, + { + "epoch": 0.9405270655270656, + "grad_norm": 0.5256397724151611, + "learning_rate": 0.00017407924775298628, + "loss": 1.058, + "step": 5282 + }, + { + "epoch": 0.9407051282051282, + "grad_norm": 0.5145402550697327, + "learning_rate": 0.00017406984444997357, + "loss": 1.0667, + "step": 5283 + }, + { + "epoch": 0.9408831908831908, + "grad_norm": 0.5435704588890076, + "learning_rate": 0.0001740604396957161, + "loss": 1.2275, + "step": 5284 + }, + { + "epoch": 0.9410612535612536, + "grad_norm": 0.5798762440681458, + "learning_rate": 0.0001740510334903982, + "loss": 1.2061, + "step": 5285 + }, + { + "epoch": 0.9412393162393162, + "grad_norm": 0.5461057424545288, + "learning_rate": 0.00017404162583420414, + "loss": 1.1585, + "step": 5286 + }, + { + "epoch": 0.9414173789173789, + "grad_norm": 0.5090487003326416, + "learning_rate": 0.00017403221672731818, + "loss": 1.2496, + "step": 5287 + }, + { + "epoch": 0.9415954415954416, + "grad_norm": 0.5171035528182983, + "learning_rate": 0.00017402280616992476, + "loss": 1.1947, + "step": 5288 + }, + { + "epoch": 0.9417735042735043, + "grad_norm": 0.5292364358901978, + "learning_rate": 0.00017401339416220818, + "loss": 1.0182, + "step": 5289 + }, + { + "epoch": 0.9419515669515669, + "grad_norm": 0.5011499524116516, + "learning_rate": 0.00017400398070435293, + "loss": 1.3363, + "step": 5290 + }, + { + "epoch": 0.9421296296296297, + "grad_norm": 0.4821554720401764, + "learning_rate": 0.0001739945657965434, + "loss": 0.9077, + "step": 5291 + }, + { + "epoch": 0.9423076923076923, + "grad_norm": 0.5849515199661255, + "learning_rate": 0.00017398514943896403, + "loss": 1.1582, + "step": 5292 + }, + { + "epoch": 0.9424857549857549, + "grad_norm": 0.49826139211654663, + "learning_rate": 0.00017397573163179937, + "loss": 1.1025, + "step": 5293 + }, + { + "epoch": 0.9426638176638177, + "grad_norm": 0.6031842827796936, + "learning_rate": 0.00017396631237523392, + "loss": 1.1932, + "step": 5294 + }, + { + "epoch": 0.9428418803418803, + "grad_norm": 0.6013330221176147, + "learning_rate": 0.00017395689166945224, + "loss": 1.2078, + "step": 5295 + }, + { + "epoch": 0.9430199430199431, + "grad_norm": 0.5147021412849426, + "learning_rate": 0.00017394746951463893, + "loss": 0.9988, + "step": 5296 + }, + { + "epoch": 0.9431980056980057, + "grad_norm": 0.5721762776374817, + "learning_rate": 0.0001739380459109785, + "loss": 1.1442, + "step": 5297 + }, + { + "epoch": 0.9433760683760684, + "grad_norm": 0.49272531270980835, + "learning_rate": 0.0001739286208586557, + "loss": 1.0481, + "step": 5298 + }, + { + "epoch": 0.9435541310541311, + "grad_norm": 0.6545688509941101, + "learning_rate": 0.00017391919435785514, + "loss": 1.1393, + "step": 5299 + }, + { + "epoch": 0.9437321937321937, + "grad_norm": 0.617756724357605, + "learning_rate": 0.00017390976640876152, + "loss": 1.1108, + "step": 5300 + }, + { + "epoch": 0.9439102564102564, + "grad_norm": 0.4870470464229584, + "learning_rate": 0.00017390033701155955, + "loss": 0.9028, + "step": 5301 + }, + { + "epoch": 0.9440883190883191, + "grad_norm": 0.5250138640403748, + "learning_rate": 0.000173890906166434, + "loss": 1.0326, + "step": 5302 + }, + { + "epoch": 0.9442663817663818, + "grad_norm": 0.5879467129707336, + "learning_rate": 0.00017388147387356964, + "loss": 1.1569, + "step": 5303 + }, + { + "epoch": 0.9444444444444444, + "grad_norm": 0.4790486991405487, + "learning_rate": 0.00017387204013315127, + "loss": 0.967, + "step": 5304 + }, + { + "epoch": 0.9446225071225072, + "grad_norm": 0.5884372591972351, + "learning_rate": 0.0001738626049453637, + "loss": 1.1342, + "step": 5305 + }, + { + "epoch": 0.9448005698005698, + "grad_norm": 0.4633975028991699, + "learning_rate": 0.00017385316831039187, + "loss": 0.8942, + "step": 5306 + }, + { + "epoch": 0.9449786324786325, + "grad_norm": 0.5301823019981384, + "learning_rate": 0.0001738437302284206, + "loss": 1.1683, + "step": 5307 + }, + { + "epoch": 0.9451566951566952, + "grad_norm": 0.5476770997047424, + "learning_rate": 0.00017383429069963484, + "loss": 1.1574, + "step": 5308 + }, + { + "epoch": 0.9453347578347578, + "grad_norm": 0.47689101099967957, + "learning_rate": 0.00017382484972421953, + "loss": 1.0792, + "step": 5309 + }, + { + "epoch": 0.9455128205128205, + "grad_norm": 0.526063084602356, + "learning_rate": 0.00017381540730235963, + "loss": 0.9012, + "step": 5310 + }, + { + "epoch": 0.9456908831908832, + "grad_norm": 0.5667058229446411, + "learning_rate": 0.0001738059634342402, + "loss": 1.0908, + "step": 5311 + }, + { + "epoch": 0.9458689458689459, + "grad_norm": 0.5402196645736694, + "learning_rate": 0.00017379651812004623, + "loss": 0.943, + "step": 5312 + }, + { + "epoch": 0.9460470085470085, + "grad_norm": 0.5288932919502258, + "learning_rate": 0.00017378707135996276, + "loss": 1.0055, + "step": 5313 + }, + { + "epoch": 0.9462250712250713, + "grad_norm": 0.5607456564903259, + "learning_rate": 0.00017377762315417492, + "loss": 1.2073, + "step": 5314 + }, + { + "epoch": 0.9464031339031339, + "grad_norm": 0.5737698674201965, + "learning_rate": 0.00017376817350286781, + "loss": 1.0001, + "step": 5315 + }, + { + "epoch": 0.9465811965811965, + "grad_norm": 0.6562079787254333, + "learning_rate": 0.00017375872240622657, + "loss": 1.1503, + "step": 5316 + }, + { + "epoch": 0.9467592592592593, + "grad_norm": 0.5407183170318604, + "learning_rate": 0.0001737492698644364, + "loss": 1.1169, + "step": 5317 + }, + { + "epoch": 0.9469373219373219, + "grad_norm": 0.5504152178764343, + "learning_rate": 0.00017373981587768248, + "loss": 1.0468, + "step": 5318 + }, + { + "epoch": 0.9471153846153846, + "grad_norm": 0.4813530743122101, + "learning_rate": 0.00017373036044615006, + "loss": 0.9707, + "step": 5319 + }, + { + "epoch": 0.9472934472934473, + "grad_norm": 0.5810509920120239, + "learning_rate": 0.00017372090357002437, + "loss": 1.4949, + "step": 5320 + }, + { + "epoch": 0.94747150997151, + "grad_norm": 0.5250222086906433, + "learning_rate": 0.00017371144524949074, + "loss": 1.0818, + "step": 5321 + }, + { + "epoch": 0.9476495726495726, + "grad_norm": 0.4852280914783478, + "learning_rate": 0.00017370198548473444, + "loss": 1.1793, + "step": 5322 + }, + { + "epoch": 0.9478276353276354, + "grad_norm": 0.5392420291900635, + "learning_rate": 0.00017369252427594086, + "loss": 1.153, + "step": 5323 + }, + { + "epoch": 0.948005698005698, + "grad_norm": 0.521294116973877, + "learning_rate": 0.00017368306162329533, + "loss": 0.8572, + "step": 5324 + }, + { + "epoch": 0.9481837606837606, + "grad_norm": 0.5579673647880554, + "learning_rate": 0.0001736735975269833, + "loss": 1.0452, + "step": 5325 + }, + { + "epoch": 0.9483618233618234, + "grad_norm": 0.6027318835258484, + "learning_rate": 0.0001736641319871901, + "loss": 1.3475, + "step": 5326 + }, + { + "epoch": 0.948539886039886, + "grad_norm": 0.5600738525390625, + "learning_rate": 0.00017365466500410132, + "loss": 1.0338, + "step": 5327 + }, + { + "epoch": 0.9487179487179487, + "grad_norm": 0.5691532492637634, + "learning_rate": 0.00017364519657790236, + "loss": 1.129, + "step": 5328 + }, + { + "epoch": 0.9488960113960114, + "grad_norm": 0.5161463022232056, + "learning_rate": 0.0001736357267087788, + "loss": 1.0438, + "step": 5329 + }, + { + "epoch": 0.9490740740740741, + "grad_norm": 0.5049656629562378, + "learning_rate": 0.0001736262553969161, + "loss": 0.9484, + "step": 5330 + }, + { + "epoch": 0.9492521367521367, + "grad_norm": 0.5477150678634644, + "learning_rate": 0.00017361678264249988, + "loss": 0.8995, + "step": 5331 + }, + { + "epoch": 0.9494301994301995, + "grad_norm": 0.5679608583450317, + "learning_rate": 0.0001736073084457157, + "loss": 1.241, + "step": 5332 + }, + { + "epoch": 0.9496082621082621, + "grad_norm": 0.5748196840286255, + "learning_rate": 0.00017359783280674926, + "loss": 1.0046, + "step": 5333 + }, + { + "epoch": 0.9497863247863247, + "grad_norm": 0.5677094459533691, + "learning_rate": 0.00017358835572578617, + "loss": 1.2913, + "step": 5334 + }, + { + "epoch": 0.9499643874643875, + "grad_norm": 0.49663659930229187, + "learning_rate": 0.0001735788772030121, + "loss": 1.0388, + "step": 5335 + }, + { + "epoch": 0.9501424501424501, + "grad_norm": 0.5687218904495239, + "learning_rate": 0.0001735693972386128, + "loss": 1.1631, + "step": 5336 + }, + { + "epoch": 0.9503205128205128, + "grad_norm": 0.520708441734314, + "learning_rate": 0.00017355991583277395, + "loss": 1.0744, + "step": 5337 + }, + { + "epoch": 0.9504985754985755, + "grad_norm": 0.5738952159881592, + "learning_rate": 0.00017355043298568137, + "loss": 1.318, + "step": 5338 + }, + { + "epoch": 0.9506766381766382, + "grad_norm": 0.5378455519676208, + "learning_rate": 0.00017354094869752085, + "loss": 0.9827, + "step": 5339 + }, + { + "epoch": 0.9508547008547008, + "grad_norm": 0.5047366619110107, + "learning_rate": 0.0001735314629684782, + "loss": 1.0966, + "step": 5340 + }, + { + "epoch": 0.9510327635327636, + "grad_norm": 0.5526043772697449, + "learning_rate": 0.0001735219757987393, + "loss": 1.059, + "step": 5341 + }, + { + "epoch": 0.9512108262108262, + "grad_norm": 0.5741400718688965, + "learning_rate": 0.00017351248718849003, + "loss": 1.1232, + "step": 5342 + }, + { + "epoch": 0.9513888888888888, + "grad_norm": 0.5421118140220642, + "learning_rate": 0.00017350299713791626, + "loss": 1.0427, + "step": 5343 + }, + { + "epoch": 0.9515669515669516, + "grad_norm": 0.4857081472873688, + "learning_rate": 0.00017349350564720392, + "loss": 0.8663, + "step": 5344 + }, + { + "epoch": 0.9517450142450142, + "grad_norm": 0.5411618947982788, + "learning_rate": 0.00017348401271653904, + "loss": 1.0317, + "step": 5345 + }, + { + "epoch": 0.9519230769230769, + "grad_norm": 0.5246246457099915, + "learning_rate": 0.00017347451834610756, + "loss": 1.0076, + "step": 5346 + }, + { + "epoch": 0.9521011396011396, + "grad_norm": 0.5278927683830261, + "learning_rate": 0.00017346502253609556, + "loss": 0.931, + "step": 5347 + }, + { + "epoch": 0.9522792022792023, + "grad_norm": 0.5934548377990723, + "learning_rate": 0.00017345552528668902, + "loss": 1.3205, + "step": 5348 + }, + { + "epoch": 0.9524572649572649, + "grad_norm": 0.5466100573539734, + "learning_rate": 0.00017344602659807406, + "loss": 0.8725, + "step": 5349 + }, + { + "epoch": 0.9526353276353277, + "grad_norm": 0.5220118761062622, + "learning_rate": 0.00017343652647043678, + "loss": 1.1642, + "step": 5350 + }, + { + "epoch": 0.9528133903133903, + "grad_norm": 0.6166301965713501, + "learning_rate": 0.0001734270249039633, + "loss": 0.8152, + "step": 5351 + }, + { + "epoch": 0.9529914529914529, + "grad_norm": 0.5173428058624268, + "learning_rate": 0.00017341752189883983, + "loss": 0.9296, + "step": 5352 + }, + { + "epoch": 0.9531695156695157, + "grad_norm": 0.5363461375236511, + "learning_rate": 0.0001734080174552525, + "loss": 1.3546, + "step": 5353 + }, + { + "epoch": 0.9533475783475783, + "grad_norm": 0.5333831906318665, + "learning_rate": 0.0001733985115733876, + "loss": 1.0401, + "step": 5354 + }, + { + "epoch": 0.9535256410256411, + "grad_norm": 0.5179334878921509, + "learning_rate": 0.00017338900425343132, + "loss": 1.1254, + "step": 5355 + }, + { + "epoch": 0.9537037037037037, + "grad_norm": 0.5171303153038025, + "learning_rate": 0.00017337949549556993, + "loss": 1.0518, + "step": 5356 + }, + { + "epoch": 0.9538817663817664, + "grad_norm": 0.5164596438407898, + "learning_rate": 0.00017336998529998978, + "loss": 0.8732, + "step": 5357 + }, + { + "epoch": 0.9540598290598291, + "grad_norm": 0.5555717349052429, + "learning_rate": 0.00017336047366687719, + "loss": 1.2312, + "step": 5358 + }, + { + "epoch": 0.9542378917378918, + "grad_norm": 0.45685622096061707, + "learning_rate": 0.00017335096059641847, + "loss": 0.8882, + "step": 5359 + }, + { + "epoch": 0.9544159544159544, + "grad_norm": 0.5260133743286133, + "learning_rate": 0.0001733414460888001, + "loss": 1.0952, + "step": 5360 + }, + { + "epoch": 0.9545940170940171, + "grad_norm": 0.4597703814506531, + "learning_rate": 0.0001733319301442084, + "loss": 1.0835, + "step": 5361 + }, + { + "epoch": 0.9547720797720798, + "grad_norm": 0.5279495120048523, + "learning_rate": 0.0001733224127628299, + "loss": 1.0295, + "step": 5362 + }, + { + "epoch": 0.9549501424501424, + "grad_norm": 0.48919400572776794, + "learning_rate": 0.00017331289394485104, + "loss": 0.9693, + "step": 5363 + }, + { + "epoch": 0.9551282051282052, + "grad_norm": 0.5639515519142151, + "learning_rate": 0.0001733033736904583, + "loss": 1.0893, + "step": 5364 + }, + { + "epoch": 0.9553062678062678, + "grad_norm": 0.49761319160461426, + "learning_rate": 0.00017329385199983823, + "loss": 1.038, + "step": 5365 + }, + { + "epoch": 0.9554843304843305, + "grad_norm": 0.5503305792808533, + "learning_rate": 0.0001732843288731774, + "loss": 0.9976, + "step": 5366 + }, + { + "epoch": 0.9556623931623932, + "grad_norm": 0.5633028745651245, + "learning_rate": 0.00017327480431066235, + "loss": 1.0602, + "step": 5367 + }, + { + "epoch": 0.9558404558404558, + "grad_norm": 0.48074454069137573, + "learning_rate": 0.00017326527831247973, + "loss": 1.0286, + "step": 5368 + }, + { + "epoch": 0.9560185185185185, + "grad_norm": 0.506597638130188, + "learning_rate": 0.0001732557508788162, + "loss": 0.9061, + "step": 5369 + }, + { + "epoch": 0.9561965811965812, + "grad_norm": 0.6570749282836914, + "learning_rate": 0.0001732462220098584, + "loss": 1.0852, + "step": 5370 + }, + { + "epoch": 0.9563746438746439, + "grad_norm": 0.5607653856277466, + "learning_rate": 0.00017323669170579302, + "loss": 1.0486, + "step": 5371 + }, + { + "epoch": 0.9565527065527065, + "grad_norm": 0.6047050356864929, + "learning_rate": 0.0001732271599668068, + "loss": 1.2175, + "step": 5372 + }, + { + "epoch": 0.9567307692307693, + "grad_norm": 0.5506869554519653, + "learning_rate": 0.00017321762679308651, + "loss": 1.0114, + "step": 5373 + }, + { + "epoch": 0.9569088319088319, + "grad_norm": 0.5868638157844543, + "learning_rate": 0.00017320809218481891, + "loss": 1.2983, + "step": 5374 + }, + { + "epoch": 0.9570868945868946, + "grad_norm": 0.539619505405426, + "learning_rate": 0.00017319855614219084, + "loss": 1.2361, + "step": 5375 + }, + { + "epoch": 0.9572649572649573, + "grad_norm": 0.5525495409965515, + "learning_rate": 0.0001731890186653891, + "loss": 1.1316, + "step": 5376 + }, + { + "epoch": 0.95744301994302, + "grad_norm": 0.5549767017364502, + "learning_rate": 0.0001731794797546006, + "loss": 1.0547, + "step": 5377 + }, + { + "epoch": 0.9576210826210826, + "grad_norm": 0.5356076955795288, + "learning_rate": 0.00017316993941001222, + "loss": 0.9942, + "step": 5378 + }, + { + "epoch": 0.9577991452991453, + "grad_norm": 0.5365784168243408, + "learning_rate": 0.00017316039763181084, + "loss": 1.226, + "step": 5379 + }, + { + "epoch": 0.957977207977208, + "grad_norm": 0.5190927386283875, + "learning_rate": 0.00017315085442018343, + "loss": 1.1704, + "step": 5380 + }, + { + "epoch": 0.9581552706552706, + "grad_norm": 0.526658833026886, + "learning_rate": 0.00017314130977531705, + "loss": 1.109, + "step": 5381 + }, + { + "epoch": 0.9583333333333334, + "grad_norm": 0.5373684763908386, + "learning_rate": 0.0001731317636973986, + "loss": 1.0018, + "step": 5382 + }, + { + "epoch": 0.958511396011396, + "grad_norm": 0.5714904069900513, + "learning_rate": 0.00017312221618661516, + "loss": 1.1855, + "step": 5383 + }, + { + "epoch": 0.9586894586894587, + "grad_norm": 0.5707863569259644, + "learning_rate": 0.00017311266724315377, + "loss": 0.9482, + "step": 5384 + }, + { + "epoch": 0.9588675213675214, + "grad_norm": 0.5856872797012329, + "learning_rate": 0.00017310311686720157, + "loss": 0.9543, + "step": 5385 + }, + { + "epoch": 0.959045584045584, + "grad_norm": 0.5041963458061218, + "learning_rate": 0.00017309356505894568, + "loss": 1.1427, + "step": 5386 + }, + { + "epoch": 0.9592236467236467, + "grad_norm": 0.5409179925918579, + "learning_rate": 0.00017308401181857316, + "loss": 0.8432, + "step": 5387 + }, + { + "epoch": 0.9594017094017094, + "grad_norm": 0.5248702764511108, + "learning_rate": 0.00017307445714627128, + "loss": 1.1403, + "step": 5388 + }, + { + "epoch": 0.9595797720797721, + "grad_norm": 0.50718092918396, + "learning_rate": 0.00017306490104222722, + "loss": 0.9066, + "step": 5389 + }, + { + "epoch": 0.9597578347578347, + "grad_norm": 0.5563821196556091, + "learning_rate": 0.0001730553435066282, + "loss": 1.0204, + "step": 5390 + }, + { + "epoch": 0.9599358974358975, + "grad_norm": 0.5696987509727478, + "learning_rate": 0.00017304578453966146, + "loss": 1.1405, + "step": 5391 + }, + { + "epoch": 0.9601139601139601, + "grad_norm": 0.5927395224571228, + "learning_rate": 0.00017303622414151435, + "loss": 1.0398, + "step": 5392 + }, + { + "epoch": 0.9602920227920227, + "grad_norm": 0.5375707745552063, + "learning_rate": 0.0001730266623123741, + "loss": 0.9519, + "step": 5393 + }, + { + "epoch": 0.9604700854700855, + "grad_norm": 0.457998126745224, + "learning_rate": 0.00017301709905242815, + "loss": 0.8743, + "step": 5394 + }, + { + "epoch": 0.9606481481481481, + "grad_norm": 0.5427796244621277, + "learning_rate": 0.00017300753436186382, + "loss": 1.078, + "step": 5395 + }, + { + "epoch": 0.9608262108262108, + "grad_norm": 0.5458595752716064, + "learning_rate": 0.0001729979682408685, + "loss": 1.1081, + "step": 5396 + }, + { + "epoch": 0.9610042735042735, + "grad_norm": 0.5495280027389526, + "learning_rate": 0.00017298840068962962, + "loss": 1.0141, + "step": 5397 + }, + { + "epoch": 0.9611823361823362, + "grad_norm": 0.5878560543060303, + "learning_rate": 0.00017297883170833465, + "loss": 1.302, + "step": 5398 + }, + { + "epoch": 0.9613603988603988, + "grad_norm": 0.5452881455421448, + "learning_rate": 0.00017296926129717108, + "loss": 0.9929, + "step": 5399 + }, + { + "epoch": 0.9615384615384616, + "grad_norm": 0.6021811366081238, + "learning_rate": 0.0001729596894563264, + "loss": 1.2629, + "step": 5400 + }, + { + "epoch": 0.9617165242165242, + "grad_norm": 0.5820204615592957, + "learning_rate": 0.0001729501161859882, + "loss": 1.0662, + "step": 5401 + }, + { + "epoch": 0.9618945868945868, + "grad_norm": 0.4953218102455139, + "learning_rate": 0.000172940541486344, + "loss": 1.047, + "step": 5402 + }, + { + "epoch": 0.9620726495726496, + "grad_norm": 0.5409793853759766, + "learning_rate": 0.00017293096535758143, + "loss": 1.1993, + "step": 5403 + }, + { + "epoch": 0.9622507122507122, + "grad_norm": 0.49702873826026917, + "learning_rate": 0.00017292138779988805, + "loss": 1.2471, + "step": 5404 + }, + { + "epoch": 0.9624287749287749, + "grad_norm": 0.5743489861488342, + "learning_rate": 0.00017291180881345158, + "loss": 1.0816, + "step": 5405 + }, + { + "epoch": 0.9626068376068376, + "grad_norm": 0.5747945308685303, + "learning_rate": 0.00017290222839845968, + "loss": 1.3548, + "step": 5406 + }, + { + "epoch": 0.9627849002849003, + "grad_norm": 0.5341345071792603, + "learning_rate": 0.00017289264655510005, + "loss": 1.0435, + "step": 5407 + }, + { + "epoch": 0.9629629629629629, + "grad_norm": 0.5719689130783081, + "learning_rate": 0.00017288306328356044, + "loss": 1.2319, + "step": 5408 + }, + { + "epoch": 0.9631410256410257, + "grad_norm": 0.4783279597759247, + "learning_rate": 0.0001728734785840286, + "loss": 0.9397, + "step": 5409 + }, + { + "epoch": 0.9633190883190883, + "grad_norm": 0.4730507731437683, + "learning_rate": 0.00017286389245669233, + "loss": 0.9384, + "step": 5410 + }, + { + "epoch": 0.9634971509971509, + "grad_norm": 0.5309939384460449, + "learning_rate": 0.00017285430490173944, + "loss": 1.098, + "step": 5411 + }, + { + "epoch": 0.9636752136752137, + "grad_norm": 0.5177853107452393, + "learning_rate": 0.0001728447159193578, + "loss": 1.2777, + "step": 5412 + }, + { + "epoch": 0.9638532763532763, + "grad_norm": 0.6437913775444031, + "learning_rate": 0.00017283512550973526, + "loss": 1.2661, + "step": 5413 + }, + { + "epoch": 0.9640313390313391, + "grad_norm": 0.6096072196960449, + "learning_rate": 0.00017282553367305975, + "loss": 0.9569, + "step": 5414 + }, + { + "epoch": 0.9642094017094017, + "grad_norm": 0.5104934573173523, + "learning_rate": 0.00017281594040951918, + "loss": 0.9666, + "step": 5415 + }, + { + "epoch": 0.9643874643874644, + "grad_norm": 0.6178240776062012, + "learning_rate": 0.00017280634571930153, + "loss": 1.1277, + "step": 5416 + }, + { + "epoch": 0.9645655270655271, + "grad_norm": 0.5749034881591797, + "learning_rate": 0.0001727967496025948, + "loss": 1.245, + "step": 5417 + }, + { + "epoch": 0.9647435897435898, + "grad_norm": 0.5036978721618652, + "learning_rate": 0.00017278715205958694, + "loss": 1.3049, + "step": 5418 + }, + { + "epoch": 0.9649216524216524, + "grad_norm": 0.5593041777610779, + "learning_rate": 0.00017277755309046605, + "loss": 1.2304, + "step": 5419 + }, + { + "epoch": 0.9650997150997151, + "grad_norm": 0.5446555614471436, + "learning_rate": 0.0001727679526954202, + "loss": 0.732, + "step": 5420 + }, + { + "epoch": 0.9652777777777778, + "grad_norm": 0.6063070297241211, + "learning_rate": 0.00017275835087463747, + "loss": 1.3723, + "step": 5421 + }, + { + "epoch": 0.9654558404558404, + "grad_norm": 0.4994211792945862, + "learning_rate": 0.00017274874762830602, + "loss": 1.0505, + "step": 5422 + }, + { + "epoch": 0.9656339031339032, + "grad_norm": 0.49396973848342896, + "learning_rate": 0.00017273914295661395, + "loss": 0.8691, + "step": 5423 + }, + { + "epoch": 0.9658119658119658, + "grad_norm": 0.5067027807235718, + "learning_rate": 0.0001727295368597495, + "loss": 0.9744, + "step": 5424 + }, + { + "epoch": 0.9659900284900285, + "grad_norm": 0.6720643043518066, + "learning_rate": 0.00017271992933790085, + "loss": 1.1513, + "step": 5425 + }, + { + "epoch": 0.9661680911680912, + "grad_norm": 0.5494341254234314, + "learning_rate": 0.00017271032039125624, + "loss": 0.8295, + "step": 5426 + }, + { + "epoch": 0.9663461538461539, + "grad_norm": 0.644332230091095, + "learning_rate": 0.00017270071002000394, + "loss": 1.0043, + "step": 5427 + }, + { + "epoch": 0.9665242165242165, + "grad_norm": 0.5658500790596008, + "learning_rate": 0.00017269109822433225, + "loss": 1.2575, + "step": 5428 + }, + { + "epoch": 0.9667022792022792, + "grad_norm": 0.5163155794143677, + "learning_rate": 0.00017268148500442952, + "loss": 1.1391, + "step": 5429 + }, + { + "epoch": 0.9668803418803419, + "grad_norm": 0.5113703608512878, + "learning_rate": 0.00017267187036048404, + "loss": 1.0819, + "step": 5430 + }, + { + "epoch": 0.9670584045584045, + "grad_norm": 0.6339422464370728, + "learning_rate": 0.00017266225429268426, + "loss": 1.0733, + "step": 5431 + }, + { + "epoch": 0.9672364672364673, + "grad_norm": 0.5158288478851318, + "learning_rate": 0.0001726526368012185, + "loss": 0.9518, + "step": 5432 + }, + { + "epoch": 0.9674145299145299, + "grad_norm": 0.593717634677887, + "learning_rate": 0.00017264301788627527, + "loss": 0.9416, + "step": 5433 + }, + { + "epoch": 0.9675925925925926, + "grad_norm": 0.49593186378479004, + "learning_rate": 0.00017263339754804301, + "loss": 1.0307, + "step": 5434 + }, + { + "epoch": 0.9677706552706553, + "grad_norm": 0.44032949209213257, + "learning_rate": 0.00017262377578671024, + "loss": 0.7884, + "step": 5435 + }, + { + "epoch": 0.967948717948718, + "grad_norm": 0.513073742389679, + "learning_rate": 0.00017261415260246538, + "loss": 0.9797, + "step": 5436 + }, + { + "epoch": 0.9681267806267806, + "grad_norm": 0.5737422108650208, + "learning_rate": 0.0001726045279954971, + "loss": 1.0487, + "step": 5437 + }, + { + "epoch": 0.9683048433048433, + "grad_norm": 0.5385867953300476, + "learning_rate": 0.0001725949019659939, + "loss": 1.4166, + "step": 5438 + }, + { + "epoch": 0.968482905982906, + "grad_norm": 0.5224326848983765, + "learning_rate": 0.00017258527451414438, + "loss": 1.195, + "step": 5439 + }, + { + "epoch": 0.9686609686609686, + "grad_norm": 0.5305148363113403, + "learning_rate": 0.0001725756456401372, + "loss": 1.0301, + "step": 5440 + }, + { + "epoch": 0.9688390313390314, + "grad_norm": 0.532588005065918, + "learning_rate": 0.000172566015344161, + "loss": 1.1269, + "step": 5441 + }, + { + "epoch": 0.969017094017094, + "grad_norm": 0.5812515020370483, + "learning_rate": 0.0001725563836264045, + "loss": 1.1787, + "step": 5442 + }, + { + "epoch": 0.9691951566951567, + "grad_norm": 0.4962109327316284, + "learning_rate": 0.00017254675048705638, + "loss": 1.0639, + "step": 5443 + }, + { + "epoch": 0.9693732193732194, + "grad_norm": 0.5094883441925049, + "learning_rate": 0.00017253711592630534, + "loss": 1.0922, + "step": 5444 + }, + { + "epoch": 0.969551282051282, + "grad_norm": 0.5728049874305725, + "learning_rate": 0.00017252747994434025, + "loss": 1.1237, + "step": 5445 + }, + { + "epoch": 0.9697293447293447, + "grad_norm": 0.5406180620193481, + "learning_rate": 0.00017251784254134983, + "loss": 1.1161, + "step": 5446 + }, + { + "epoch": 0.9699074074074074, + "grad_norm": 0.5724552869796753, + "learning_rate": 0.00017250820371752292, + "loss": 1.2205, + "step": 5447 + }, + { + "epoch": 0.9700854700854701, + "grad_norm": 0.5698846578598022, + "learning_rate": 0.0001724985634730484, + "loss": 1.1472, + "step": 5448 + }, + { + "epoch": 0.9702635327635327, + "grad_norm": 0.5315805673599243, + "learning_rate": 0.0001724889218081151, + "loss": 1.0253, + "step": 5449 + }, + { + "epoch": 0.9704415954415955, + "grad_norm": 0.5970377326011658, + "learning_rate": 0.000172479278722912, + "loss": 1.3033, + "step": 5450 + }, + { + "epoch": 0.9706196581196581, + "grad_norm": 0.6149488687515259, + "learning_rate": 0.00017246963421762798, + "loss": 1.0689, + "step": 5451 + }, + { + "epoch": 0.9707977207977208, + "grad_norm": 0.4848574995994568, + "learning_rate": 0.00017245998829245202, + "loss": 0.8829, + "step": 5452 + }, + { + "epoch": 0.9709757834757835, + "grad_norm": 0.6073294281959534, + "learning_rate": 0.00017245034094757312, + "loss": 1.2378, + "step": 5453 + }, + { + "epoch": 0.9711538461538461, + "grad_norm": 0.6362034678459167, + "learning_rate": 0.00017244069218318026, + "loss": 1.3606, + "step": 5454 + }, + { + "epoch": 0.9713319088319088, + "grad_norm": 0.5353880524635315, + "learning_rate": 0.00017243104199946257, + "loss": 1.1288, + "step": 5455 + }, + { + "epoch": 0.9715099715099715, + "grad_norm": 0.5096352100372314, + "learning_rate": 0.00017242139039660902, + "loss": 1.0056, + "step": 5456 + }, + { + "epoch": 0.9716880341880342, + "grad_norm": 0.5086682438850403, + "learning_rate": 0.00017241173737480884, + "loss": 1.091, + "step": 5457 + }, + { + "epoch": 0.9718660968660968, + "grad_norm": 0.5034295320510864, + "learning_rate": 0.000172402082934251, + "loss": 0.9749, + "step": 5458 + }, + { + "epoch": 0.9720441595441596, + "grad_norm": 0.5205379724502563, + "learning_rate": 0.0001723924270751248, + "loss": 1.1068, + "step": 5459 + }, + { + "epoch": 0.9722222222222222, + "grad_norm": 0.5904826521873474, + "learning_rate": 0.00017238276979761937, + "loss": 1.0613, + "step": 5460 + }, + { + "epoch": 0.9724002849002849, + "grad_norm": 0.6415045261383057, + "learning_rate": 0.0001723731111019239, + "loss": 1.2126, + "step": 5461 + }, + { + "epoch": 0.9725783475783476, + "grad_norm": 0.5769147872924805, + "learning_rate": 0.0001723634509882277, + "loss": 1.337, + "step": 5462 + }, + { + "epoch": 0.9727564102564102, + "grad_norm": 0.5585111975669861, + "learning_rate": 0.00017235378945671998, + "loss": 1.3922, + "step": 5463 + }, + { + "epoch": 0.9729344729344729, + "grad_norm": 0.5788411498069763, + "learning_rate": 0.00017234412650759008, + "loss": 0.8532, + "step": 5464 + }, + { + "epoch": 0.9731125356125356, + "grad_norm": 0.5617673397064209, + "learning_rate": 0.00017233446214102728, + "loss": 1.2575, + "step": 5465 + }, + { + "epoch": 0.9732905982905983, + "grad_norm": 0.4227815568447113, + "learning_rate": 0.00017232479635722093, + "loss": 1.0618, + "step": 5466 + }, + { + "epoch": 0.9734686609686609, + "grad_norm": 0.49751797318458557, + "learning_rate": 0.00017231512915636047, + "loss": 0.7714, + "step": 5467 + }, + { + "epoch": 0.9736467236467237, + "grad_norm": 0.5983800292015076, + "learning_rate": 0.0001723054605386353, + "loss": 1.2297, + "step": 5468 + }, + { + "epoch": 0.9738247863247863, + "grad_norm": 0.543394923210144, + "learning_rate": 0.0001722957905042348, + "loss": 1.0078, + "step": 5469 + }, + { + "epoch": 0.9740028490028491, + "grad_norm": 0.5633566975593567, + "learning_rate": 0.00017228611905334846, + "loss": 1.0938, + "step": 5470 + }, + { + "epoch": 0.9741809116809117, + "grad_norm": 0.49377235770225525, + "learning_rate": 0.00017227644618616578, + "loss": 1.096, + "step": 5471 + }, + { + "epoch": 0.9743589743589743, + "grad_norm": 0.4963362216949463, + "learning_rate": 0.00017226677190287627, + "loss": 1.0003, + "step": 5472 + }, + { + "epoch": 0.9745370370370371, + "grad_norm": 0.4483006000518799, + "learning_rate": 0.00017225709620366953, + "loss": 0.8623, + "step": 5473 + }, + { + "epoch": 0.9747150997150997, + "grad_norm": 0.5429352521896362, + "learning_rate": 0.00017224741908873506, + "loss": 1.1383, + "step": 5474 + }, + { + "epoch": 0.9748931623931624, + "grad_norm": 0.5871657729148865, + "learning_rate": 0.0001722377405582625, + "loss": 1.2005, + "step": 5475 + }, + { + "epoch": 0.9750712250712251, + "grad_norm": 0.6002383828163147, + "learning_rate": 0.0001722280606124415, + "loss": 1.0696, + "step": 5476 + }, + { + "epoch": 0.9752492877492878, + "grad_norm": 0.5351617336273193, + "learning_rate": 0.00017221837925146164, + "loss": 1.243, + "step": 5477 + }, + { + "epoch": 0.9754273504273504, + "grad_norm": 0.46613118052482605, + "learning_rate": 0.00017220869647551268, + "loss": 1.0344, + "step": 5478 + }, + { + "epoch": 0.9756054131054132, + "grad_norm": 0.6015593409538269, + "learning_rate": 0.00017219901228478432, + "loss": 1.082, + "step": 5479 + }, + { + "epoch": 0.9757834757834758, + "grad_norm": 0.5829521417617798, + "learning_rate": 0.0001721893266794663, + "loss": 0.8683, + "step": 5480 + }, + { + "epoch": 0.9759615384615384, + "grad_norm": 0.6344960927963257, + "learning_rate": 0.00017217963965974838, + "loss": 1.1048, + "step": 5481 + }, + { + "epoch": 0.9761396011396012, + "grad_norm": 0.5586308240890503, + "learning_rate": 0.00017216995122582034, + "loss": 0.9657, + "step": 5482 + }, + { + "epoch": 0.9763176638176638, + "grad_norm": 0.48625239729881287, + "learning_rate": 0.00017216026137787204, + "loss": 1.1026, + "step": 5483 + }, + { + "epoch": 0.9764957264957265, + "grad_norm": 0.5625223517417908, + "learning_rate": 0.00017215057011609332, + "loss": 1.1579, + "step": 5484 + }, + { + "epoch": 0.9766737891737892, + "grad_norm": 0.6016653776168823, + "learning_rate": 0.0001721408774406741, + "loss": 1.1777, + "step": 5485 + }, + { + "epoch": 0.9768518518518519, + "grad_norm": 0.5444921851158142, + "learning_rate": 0.00017213118335180418, + "loss": 1.119, + "step": 5486 + }, + { + "epoch": 0.9770299145299145, + "grad_norm": 0.5574755668640137, + "learning_rate": 0.0001721214878496736, + "loss": 1.1128, + "step": 5487 + }, + { + "epoch": 0.9772079772079773, + "grad_norm": 0.5486113429069519, + "learning_rate": 0.00017211179093447226, + "loss": 1.1673, + "step": 5488 + }, + { + "epoch": 0.9773860398860399, + "grad_norm": 0.5545483231544495, + "learning_rate": 0.00017210209260639018, + "loss": 1.1748, + "step": 5489 + }, + { + "epoch": 0.9775641025641025, + "grad_norm": 0.5756667256355286, + "learning_rate": 0.0001720923928656174, + "loss": 1.2377, + "step": 5490 + }, + { + "epoch": 0.9777421652421653, + "grad_norm": 0.5744972229003906, + "learning_rate": 0.00017208269171234392, + "loss": 1.1242, + "step": 5491 + }, + { + "epoch": 0.9779202279202279, + "grad_norm": 0.6109468340873718, + "learning_rate": 0.00017207298914675984, + "loss": 1.1948, + "step": 5492 + }, + { + "epoch": 0.9780982905982906, + "grad_norm": 0.5195167660713196, + "learning_rate": 0.00017206328516905525, + "loss": 1.0941, + "step": 5493 + }, + { + "epoch": 0.9782763532763533, + "grad_norm": 0.5549042224884033, + "learning_rate": 0.0001720535797794203, + "loss": 1.1503, + "step": 5494 + }, + { + "epoch": 0.978454415954416, + "grad_norm": 0.6317743062973022, + "learning_rate": 0.0001720438729780451, + "loss": 1.3468, + "step": 5495 + }, + { + "epoch": 0.9786324786324786, + "grad_norm": 0.5932528972625732, + "learning_rate": 0.0001720341647651199, + "loss": 1.105, + "step": 5496 + }, + { + "epoch": 0.9788105413105413, + "grad_norm": 0.607880175113678, + "learning_rate": 0.00017202445514083488, + "loss": 1.1465, + "step": 5497 + }, + { + "epoch": 0.978988603988604, + "grad_norm": 0.49227309226989746, + "learning_rate": 0.00017201474410538027, + "loss": 0.9075, + "step": 5498 + }, + { + "epoch": 0.9791666666666666, + "grad_norm": 0.5059443116188049, + "learning_rate": 0.00017200503165894636, + "loss": 1.0483, + "step": 5499 + }, + { + "epoch": 0.9793447293447294, + "grad_norm": 0.5792799592018127, + "learning_rate": 0.0001719953178017234, + "loss": 1.0987, + "step": 5500 + }, + { + "epoch": 0.979522792022792, + "grad_norm": 0.5010457038879395, + "learning_rate": 0.00017198560253390177, + "loss": 1.1051, + "step": 5501 + }, + { + "epoch": 0.9797008547008547, + "grad_norm": 0.5866543054580688, + "learning_rate": 0.0001719758858556718, + "loss": 1.2824, + "step": 5502 + }, + { + "epoch": 0.9798789173789174, + "grad_norm": 0.5392137169837952, + "learning_rate": 0.00017196616776722382, + "loss": 0.886, + "step": 5503 + }, + { + "epoch": 0.98005698005698, + "grad_norm": 0.5200899839401245, + "learning_rate": 0.00017195644826874834, + "loss": 1.1504, + "step": 5504 + }, + { + "epoch": 0.9802350427350427, + "grad_norm": 0.533159077167511, + "learning_rate": 0.00017194672736043569, + "loss": 1.1216, + "step": 5505 + }, + { + "epoch": 0.9804131054131054, + "grad_norm": 0.5543524622917175, + "learning_rate": 0.0001719370050424764, + "loss": 1.0161, + "step": 5506 + }, + { + "epoch": 0.9805911680911681, + "grad_norm": 0.5315365195274353, + "learning_rate": 0.00017192728131506092, + "loss": 1.0509, + "step": 5507 + }, + { + "epoch": 0.9807692307692307, + "grad_norm": 0.5406147837638855, + "learning_rate": 0.00017191755617837977, + "loss": 1.0695, + "step": 5508 + }, + { + "epoch": 0.9809472934472935, + "grad_norm": 0.4563386142253876, + "learning_rate": 0.00017190782963262354, + "loss": 0.995, + "step": 5509 + }, + { + "epoch": 0.9811253561253561, + "grad_norm": 0.5456405282020569, + "learning_rate": 0.00017189810167798274, + "loss": 1.0546, + "step": 5510 + }, + { + "epoch": 0.9813034188034188, + "grad_norm": 0.6275575160980225, + "learning_rate": 0.00017188837231464795, + "loss": 1.0432, + "step": 5511 + }, + { + "epoch": 0.9814814814814815, + "grad_norm": 0.49735602736473083, + "learning_rate": 0.0001718786415428099, + "loss": 1.035, + "step": 5512 + }, + { + "epoch": 0.9816595441595442, + "grad_norm": 0.5234259963035583, + "learning_rate": 0.00017186890936265916, + "loss": 1.0918, + "step": 5513 + }, + { + "epoch": 0.9818376068376068, + "grad_norm": 0.5091170072555542, + "learning_rate": 0.00017185917577438643, + "loss": 1.0239, + "step": 5514 + }, + { + "epoch": 0.9820156695156695, + "grad_norm": 0.6155703067779541, + "learning_rate": 0.00017184944077818244, + "loss": 1.2366, + "step": 5515 + }, + { + "epoch": 0.9821937321937322, + "grad_norm": 0.5074070692062378, + "learning_rate": 0.0001718397043742379, + "loss": 1.0318, + "step": 5516 + }, + { + "epoch": 0.9823717948717948, + "grad_norm": 0.5234423279762268, + "learning_rate": 0.0001718299665627436, + "loss": 1.0322, + "step": 5517 + }, + { + "epoch": 0.9825498575498576, + "grad_norm": 0.5783474445343018, + "learning_rate": 0.0001718202273438903, + "loss": 0.9486, + "step": 5518 + }, + { + "epoch": 0.9827279202279202, + "grad_norm": 0.5708683133125305, + "learning_rate": 0.00017181048671786886, + "loss": 1.0785, + "step": 5519 + }, + { + "epoch": 0.9829059829059829, + "grad_norm": 0.5985961556434631, + "learning_rate": 0.00017180074468487009, + "loss": 1.198, + "step": 5520 + }, + { + "epoch": 0.9830840455840456, + "grad_norm": 0.5711352229118347, + "learning_rate": 0.0001717910012450849, + "loss": 1.0386, + "step": 5521 + }, + { + "epoch": 0.9832621082621082, + "grad_norm": 0.5338063836097717, + "learning_rate": 0.00017178125639870416, + "loss": 1.1594, + "step": 5522 + }, + { + "epoch": 0.9834401709401709, + "grad_norm": 0.6144943237304688, + "learning_rate": 0.00017177151014591881, + "loss": 1.1083, + "step": 5523 + }, + { + "epoch": 0.9836182336182336, + "grad_norm": 0.547285795211792, + "learning_rate": 0.00017176176248691983, + "loss": 1.1507, + "step": 5524 + }, + { + "epoch": 0.9837962962962963, + "grad_norm": 0.5807644724845886, + "learning_rate": 0.00017175201342189817, + "loss": 1.3044, + "step": 5525 + }, + { + "epoch": 0.9839743589743589, + "grad_norm": 0.5229477882385254, + "learning_rate": 0.00017174226295104485, + "loss": 1.2622, + "step": 5526 + }, + { + "epoch": 0.9841524216524217, + "grad_norm": 0.6100695133209229, + "learning_rate": 0.00017173251107455094, + "loss": 1.2026, + "step": 5527 + }, + { + "epoch": 0.9843304843304843, + "grad_norm": 0.5410884618759155, + "learning_rate": 0.00017172275779260744, + "loss": 1.2964, + "step": 5528 + }, + { + "epoch": 0.9845085470085471, + "grad_norm": 0.5937406420707703, + "learning_rate": 0.00017171300310540554, + "loss": 1.1435, + "step": 5529 + }, + { + "epoch": 0.9846866096866097, + "grad_norm": 0.56817227602005, + "learning_rate": 0.00017170324701313634, + "loss": 1.0099, + "step": 5530 + }, + { + "epoch": 0.9848646723646723, + "grad_norm": 0.5776323080062866, + "learning_rate": 0.00017169348951599092, + "loss": 1.3539, + "step": 5531 + }, + { + "epoch": 0.9850427350427351, + "grad_norm": 0.5208535194396973, + "learning_rate": 0.0001716837306141605, + "loss": 1.2306, + "step": 5532 + }, + { + "epoch": 0.9852207977207977, + "grad_norm": 0.552173376083374, + "learning_rate": 0.0001716739703078363, + "loss": 1.0551, + "step": 5533 + }, + { + "epoch": 0.9853988603988604, + "grad_norm": 0.5327515602111816, + "learning_rate": 0.00017166420859720955, + "loss": 1.2443, + "step": 5534 + }, + { + "epoch": 0.9855769230769231, + "grad_norm": 0.5255244374275208, + "learning_rate": 0.0001716544454824715, + "loss": 1.005, + "step": 5535 + }, + { + "epoch": 0.9857549857549858, + "grad_norm": 0.4753847122192383, + "learning_rate": 0.00017164468096381343, + "loss": 1.0081, + "step": 5536 + }, + { + "epoch": 0.9859330484330484, + "grad_norm": 0.5261829495429993, + "learning_rate": 0.00017163491504142665, + "loss": 1.2249, + "step": 5537 + }, + { + "epoch": 0.9861111111111112, + "grad_norm": 0.46499499678611755, + "learning_rate": 0.00017162514771550255, + "loss": 0.8759, + "step": 5538 + }, + { + "epoch": 0.9862891737891738, + "grad_norm": 0.5233004689216614, + "learning_rate": 0.00017161537898623247, + "loss": 1.0474, + "step": 5539 + }, + { + "epoch": 0.9864672364672364, + "grad_norm": 0.46905553340911865, + "learning_rate": 0.00017160560885380778, + "loss": 0.9033, + "step": 5540 + }, + { + "epoch": 0.9866452991452992, + "grad_norm": 0.5816231369972229, + "learning_rate": 0.00017159583731841998, + "loss": 1.0628, + "step": 5541 + }, + { + "epoch": 0.9868233618233618, + "grad_norm": 0.4575413167476654, + "learning_rate": 0.00017158606438026045, + "loss": 1.0446, + "step": 5542 + }, + { + "epoch": 0.9870014245014245, + "grad_norm": 0.5968109965324402, + "learning_rate": 0.00017157629003952067, + "loss": 1.032, + "step": 5543 + }, + { + "epoch": 0.9871794871794872, + "grad_norm": 0.5316148400306702, + "learning_rate": 0.00017156651429639218, + "loss": 0.9167, + "step": 5544 + }, + { + "epoch": 0.9873575498575499, + "grad_norm": 0.5185125470161438, + "learning_rate": 0.00017155673715106651, + "loss": 1.1527, + "step": 5545 + }, + { + "epoch": 0.9875356125356125, + "grad_norm": 0.5167772769927979, + "learning_rate": 0.00017154695860373525, + "loss": 0.9954, + "step": 5546 + }, + { + "epoch": 0.9877136752136753, + "grad_norm": 0.6406680345535278, + "learning_rate": 0.00017153717865458994, + "loss": 1.2758, + "step": 5547 + }, + { + "epoch": 0.9878917378917379, + "grad_norm": 0.5223956108093262, + "learning_rate": 0.00017152739730382223, + "loss": 1.1526, + "step": 5548 + }, + { + "epoch": 0.9880698005698005, + "grad_norm": 0.6131790280342102, + "learning_rate": 0.00017151761455162375, + "loss": 1.1024, + "step": 5549 + }, + { + "epoch": 0.9882478632478633, + "grad_norm": 0.5574753880500793, + "learning_rate": 0.00017150783039818616, + "loss": 0.9733, + "step": 5550 + }, + { + "epoch": 0.9884259259259259, + "grad_norm": 0.5417882800102234, + "learning_rate": 0.0001714980448437012, + "loss": 1.2244, + "step": 5551 + }, + { + "epoch": 0.9886039886039886, + "grad_norm": 0.6217474341392517, + "learning_rate": 0.0001714882578883606, + "loss": 0.9224, + "step": 5552 + }, + { + "epoch": 0.9887820512820513, + "grad_norm": 0.5846285223960876, + "learning_rate": 0.00017147846953235606, + "loss": 1.2429, + "step": 5553 + }, + { + "epoch": 0.988960113960114, + "grad_norm": 0.5924782752990723, + "learning_rate": 0.00017146867977587936, + "loss": 0.9907, + "step": 5554 + }, + { + "epoch": 0.9891381766381766, + "grad_norm": 0.5756853818893433, + "learning_rate": 0.00017145888861912242, + "loss": 1.1266, + "step": 5555 + }, + { + "epoch": 0.9893162393162394, + "grad_norm": 0.5277376770973206, + "learning_rate": 0.00017144909606227693, + "loss": 1.1676, + "step": 5556 + }, + { + "epoch": 0.989494301994302, + "grad_norm": 0.5138902068138123, + "learning_rate": 0.00017143930210553485, + "loss": 0.9864, + "step": 5557 + }, + { + "epoch": 0.9896723646723646, + "grad_norm": 0.8072507977485657, + "learning_rate": 0.00017142950674908805, + "loss": 1.111, + "step": 5558 + }, + { + "epoch": 0.9898504273504274, + "grad_norm": 0.5641721487045288, + "learning_rate": 0.00017141970999312844, + "loss": 0.9106, + "step": 5559 + }, + { + "epoch": 0.99002849002849, + "grad_norm": 0.5260798931121826, + "learning_rate": 0.000171409911837848, + "loss": 1.1609, + "step": 5560 + }, + { + "epoch": 0.9902065527065527, + "grad_norm": 0.5398530960083008, + "learning_rate": 0.00017140011228343864, + "loss": 1.0368, + "step": 5561 + }, + { + "epoch": 0.9903846153846154, + "grad_norm": 0.6011313199996948, + "learning_rate": 0.00017139031133009245, + "loss": 1.1314, + "step": 5562 + }, + { + "epoch": 0.9905626780626781, + "grad_norm": 0.6194971203804016, + "learning_rate": 0.00017138050897800135, + "loss": 1.3493, + "step": 5563 + }, + { + "epoch": 0.9907407407407407, + "grad_norm": 0.5779356956481934, + "learning_rate": 0.0001713707052273575, + "loss": 0.943, + "step": 5564 + }, + { + "epoch": 0.9909188034188035, + "grad_norm": 0.5321127772331238, + "learning_rate": 0.00017136090007835293, + "loss": 0.7914, + "step": 5565 + }, + { + "epoch": 0.9910968660968661, + "grad_norm": 0.5470426678657532, + "learning_rate": 0.00017135109353117977, + "loss": 1.2113, + "step": 5566 + }, + { + "epoch": 0.9912749287749287, + "grad_norm": 0.5551436543464661, + "learning_rate": 0.00017134128558603012, + "loss": 0.8932, + "step": 5567 + }, + { + "epoch": 0.9914529914529915, + "grad_norm": 0.45770928263664246, + "learning_rate": 0.0001713314762430962, + "loss": 1.0061, + "step": 5568 + }, + { + "epoch": 0.9916310541310541, + "grad_norm": 0.5578967332839966, + "learning_rate": 0.00017132166550257017, + "loss": 1.148, + "step": 5569 + }, + { + "epoch": 0.9918091168091168, + "grad_norm": 0.5086452960968018, + "learning_rate": 0.0001713118533646443, + "loss": 0.9803, + "step": 5570 + }, + { + "epoch": 0.9919871794871795, + "grad_norm": 0.4714745879173279, + "learning_rate": 0.00017130203982951078, + "loss": 1.0176, + "step": 5571 + }, + { + "epoch": 0.9921652421652422, + "grad_norm": 0.6254406571388245, + "learning_rate": 0.0001712922248973619, + "loss": 1.0932, + "step": 5572 + }, + { + "epoch": 0.9923433048433048, + "grad_norm": 0.5005003809928894, + "learning_rate": 0.00017128240856838998, + "loss": 1.0783, + "step": 5573 + }, + { + "epoch": 0.9925213675213675, + "grad_norm": 0.5668206214904785, + "learning_rate": 0.00017127259084278733, + "loss": 1.0404, + "step": 5574 + }, + { + "epoch": 0.9926994301994302, + "grad_norm": 0.4976036250591278, + "learning_rate": 0.00017126277172074632, + "loss": 1.1437, + "step": 5575 + }, + { + "epoch": 0.9928774928774928, + "grad_norm": 0.567546546459198, + "learning_rate": 0.00017125295120245935, + "loss": 1.2188, + "step": 5576 + }, + { + "epoch": 0.9930555555555556, + "grad_norm": 0.5614372491836548, + "learning_rate": 0.0001712431292881188, + "loss": 0.9187, + "step": 5577 + }, + { + "epoch": 0.9932336182336182, + "grad_norm": 0.6117973327636719, + "learning_rate": 0.00017123330597791712, + "loss": 1.1285, + "step": 5578 + }, + { + "epoch": 0.9934116809116809, + "grad_norm": 0.6000342965126038, + "learning_rate": 0.00017122348127204676, + "loss": 0.9837, + "step": 5579 + }, + { + "epoch": 0.9935897435897436, + "grad_norm": 0.5453050136566162, + "learning_rate": 0.0001712136551707003, + "loss": 0.8771, + "step": 5580 + }, + { + "epoch": 0.9937678062678063, + "grad_norm": 0.49603891372680664, + "learning_rate": 0.00017120382767407018, + "loss": 1.0754, + "step": 5581 + }, + { + "epoch": 0.9939458689458689, + "grad_norm": 0.48031488060951233, + "learning_rate": 0.00017119399878234894, + "loss": 0.6933, + "step": 5582 + }, + { + "epoch": 0.9941239316239316, + "grad_norm": 0.6048742532730103, + "learning_rate": 0.0001711841684957292, + "loss": 0.9696, + "step": 5583 + }, + { + "epoch": 0.9943019943019943, + "grad_norm": 0.5183123350143433, + "learning_rate": 0.00017117433681440355, + "loss": 1.1313, + "step": 5584 + }, + { + "epoch": 0.9944800569800569, + "grad_norm": 0.504916250705719, + "learning_rate": 0.00017116450373856466, + "loss": 1.0273, + "step": 5585 + }, + { + "epoch": 0.9946581196581197, + "grad_norm": 0.5804886817932129, + "learning_rate": 0.0001711546692684051, + "loss": 1.1162, + "step": 5586 + }, + { + "epoch": 0.9948361823361823, + "grad_norm": 0.5531938672065735, + "learning_rate": 0.0001711448334041176, + "loss": 1.2893, + "step": 5587 + }, + { + "epoch": 0.9950142450142451, + "grad_norm": 0.5079928636550903, + "learning_rate": 0.00017113499614589492, + "loss": 1.0393, + "step": 5588 + }, + { + "epoch": 0.9951923076923077, + "grad_norm": 0.5421964526176453, + "learning_rate": 0.00017112515749392973, + "loss": 0.8844, + "step": 5589 + }, + { + "epoch": 0.9953703703703703, + "grad_norm": 0.4834558367729187, + "learning_rate": 0.00017111531744841486, + "loss": 1.0187, + "step": 5590 + }, + { + "epoch": 0.9955484330484331, + "grad_norm": 0.6704340577125549, + "learning_rate": 0.00017110547600954307, + "loss": 0.8524, + "step": 5591 + }, + { + "epoch": 0.9957264957264957, + "grad_norm": 0.4578927159309387, + "learning_rate": 0.00017109563317750718, + "loss": 1.059, + "step": 5592 + }, + { + "epoch": 0.9959045584045584, + "grad_norm": 0.5563494563102722, + "learning_rate": 0.00017108578895250006, + "loss": 1.1211, + "step": 5593 + }, + { + "epoch": 0.9960826210826211, + "grad_norm": 0.5272170901298523, + "learning_rate": 0.00017107594333471454, + "loss": 0.9224, + "step": 5594 + }, + { + "epoch": 0.9962606837606838, + "grad_norm": 0.5697501301765442, + "learning_rate": 0.00017106609632434357, + "loss": 1.2223, + "step": 5595 + }, + { + "epoch": 0.9964387464387464, + "grad_norm": 0.5385653376579285, + "learning_rate": 0.00017105624792158007, + "loss": 1.0809, + "step": 5596 + }, + { + "epoch": 0.9966168091168092, + "grad_norm": 0.5608006119728088, + "learning_rate": 0.000171046398126617, + "loss": 1.3936, + "step": 5597 + }, + { + "epoch": 0.9967948717948718, + "grad_norm": 0.5063132643699646, + "learning_rate": 0.00017103654693964736, + "loss": 1.2086, + "step": 5598 + }, + { + "epoch": 0.9969729344729344, + "grad_norm": 0.6014235019683838, + "learning_rate": 0.00017102669436086415, + "loss": 1.1231, + "step": 5599 + }, + { + "epoch": 0.9971509971509972, + "grad_norm": 0.49549567699432373, + "learning_rate": 0.00017101684039046036, + "loss": 1.0013, + "step": 5600 + }, + { + "epoch": 0.9973290598290598, + "grad_norm": 0.517464816570282, + "learning_rate": 0.00017100698502862916, + "loss": 1.1143, + "step": 5601 + }, + { + "epoch": 0.9975071225071225, + "grad_norm": 0.514281153678894, + "learning_rate": 0.00017099712827556358, + "loss": 1.0336, + "step": 5602 + }, + { + "epoch": 0.9976851851851852, + "grad_norm": 0.5378567576408386, + "learning_rate": 0.00017098727013145672, + "loss": 0.8278, + "step": 5603 + }, + { + "epoch": 0.9978632478632479, + "grad_norm": 0.5098404884338379, + "learning_rate": 0.0001709774105965018, + "loss": 0.9902, + "step": 5604 + }, + { + "epoch": 0.9980413105413105, + "grad_norm": 0.6231759190559387, + "learning_rate": 0.00017096754967089198, + "loss": 1.0564, + "step": 5605 + }, + { + "epoch": 0.9982193732193733, + "grad_norm": 0.47434380650520325, + "learning_rate": 0.00017095768735482042, + "loss": 0.7457, + "step": 5606 + }, + { + "epoch": 0.9983974358974359, + "grad_norm": 0.5771013498306274, + "learning_rate": 0.00017094782364848035, + "loss": 1.1191, + "step": 5607 + }, + { + "epoch": 0.9985754985754985, + "grad_norm": 0.5617234706878662, + "learning_rate": 0.00017093795855206508, + "loss": 1.0779, + "step": 5608 + }, + { + "epoch": 0.9987535612535613, + "grad_norm": 0.6573554873466492, + "learning_rate": 0.00017092809206576792, + "loss": 1.0191, + "step": 5609 + }, + { + "epoch": 0.9989316239316239, + "grad_norm": 0.482834130525589, + "learning_rate": 0.00017091822418978207, + "loss": 1.0119, + "step": 5610 + }, + { + "epoch": 0.9991096866096866, + "grad_norm": 0.47496405243873596, + "learning_rate": 0.000170908354924301, + "loss": 0.8297, + "step": 5611 + }, + { + "epoch": 0.9992877492877493, + "grad_norm": 0.5013265013694763, + "learning_rate": 0.00017089848426951796, + "loss": 1.1511, + "step": 5612 + }, + { + "epoch": 0.999465811965812, + "grad_norm": 0.5402522683143616, + "learning_rate": 0.00017088861222562643, + "loss": 1.1401, + "step": 5613 + }, + { + "epoch": 0.9996438746438746, + "grad_norm": 0.546302318572998, + "learning_rate": 0.00017087873879281977, + "loss": 0.8611, + "step": 5614 + }, + { + "epoch": 0.9998219373219374, + "grad_norm": 0.44279807806015015, + "learning_rate": 0.0001708688639712915, + "loss": 0.79, + "step": 5615 + }, + { + "epoch": 1.0, + "grad_norm": 0.5514659285545349, + "learning_rate": 0.00017085898776123502, + "loss": 1.0709, + "step": 5616 + }, + { + "epoch": 1.0, + "eval_loss": 1.093075156211853, + "eval_runtime": 24.6155, + "eval_samples_per_second": 42.29, + "eval_steps_per_second": 21.166, + "step": 5616 + } + ], + "logging_steps": 1, + "max_steps": 22464, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 5616, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.1681189773991936e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-5616/training_args.bin b/checkpoint-5616/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..1245f6a2afbe9a6eefbb6d141231d555e0b0bf84 --- /dev/null +++ b/checkpoint-5616/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86de370014ed2be86ea27c820b434ceec5e097da2b5f9b08d0eac9aa564d8961 +size 6200 diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..0c19ab480f00c0f7d7b5b9100fc78008e88f98aa --- /dev/null +++ b/config.json @@ -0,0 +1,47 @@ +{ + "_attn_implementation_autoset": true, + "_name_or_path": "openlm-research/open_llama_3b_v2", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 1, + "eos_token_id": 2, + "head_dim": 100, + "hidden_act": "silu", + "hidden_size": 3200, + "initializer_range": 0.02, + "intermediate_size": 8640, + "max_position_embeddings": 2048, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 32, + "num_hidden_layers": 26, + "num_key_value_heads": 32, + "pad_token_id": 0, + "pretraining_tp": 1, + "quantization_config": { + "_load_in_4bit": false, + "_load_in_8bit": true, + "bnb_4bit_compute_dtype": "float32", + "bnb_4bit_quant_storage": "uint8", + "bnb_4bit_quant_type": "fp4", + "bnb_4bit_use_double_quant": false, + "llm_int8_enable_fp32_cpu_offload": false, + "llm_int8_has_fp16_weight": false, + "llm_int8_skip_modules": null, + "llm_int8_threshold": 6.0, + "load_in_4bit": false, + "load_in_8bit": true, + "quant_method": "bitsandbytes" + }, + "rms_norm_eps": 1e-06, + "rope_scaling": null, + "rope_theta": 10000.0, + "tie_word_embeddings": false, + "torch_dtype": "float16", + "transformers_version": "4.46.1", + "use_cache": false, + "vocab_size": 32000 +} diff --git a/merged/config.json b/merged/config.json new file mode 100644 index 0000000000000000000000000000000000000000..de1fd07914c59e1f3619f6a02991d58ed3d50a5f --- /dev/null +++ b/merged/config.json @@ -0,0 +1,31 @@ +{ + "_name_or_path": "openlm-research/open_llama_3b_v2", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 1, + "eos_token_id": 2, + "head_dim": 100, + "hidden_act": "silu", + "hidden_size": 3200, + "initializer_range": 0.02, + "intermediate_size": 8640, + "max_position_embeddings": 2048, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 32, + "num_hidden_layers": 26, + "num_key_value_heads": 32, + "pad_token_id": 0, + "pretraining_tp": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": null, + "rope_theta": 10000.0, + "tie_word_embeddings": false, + "torch_dtype": "float16", + "transformers_version": "4.46.1", + "use_cache": false, + "vocab_size": 32000 +} diff --git a/merged/generation_config.json b/merged/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..08ba3591b4ff1d0b847f1f5692758b0eefdd40c8 --- /dev/null +++ b/merged/generation_config.json @@ -0,0 +1,8 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": 2, + "pad_token_id": 0, + "transformers_version": "4.46.1" +} diff --git a/merged/pytorch_model-00001-of-00002.bin b/merged/pytorch_model-00001-of-00002.bin new file mode 100644 index 0000000000000000000000000000000000000000..261a38874c2f41af6006bc7f862381f37afddfbe --- /dev/null +++ b/merged/pytorch_model-00001-of-00002.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b8c49a856649e62748a4ddf071cc144f832bd35945c16204dd76e4c590fdfa76 +size 4995377838 diff --git a/merged/pytorch_model-00002-of-00002.bin b/merged/pytorch_model-00002-of-00002.bin new file mode 100644 index 0000000000000000000000000000000000000000..e445c36d05d0984cb7b894b66ec882b47555d43d --- /dev/null +++ b/merged/pytorch_model-00002-of-00002.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89939bdc254a14ef5179d72f6886b5c898ce703a895164442f19292ce9b8e22a +size 1857653748 diff --git a/merged/pytorch_model.bin.index.json b/merged/pytorch_model.bin.index.json new file mode 100644 index 0000000000000000000000000000000000000000..8c89827f6f3500ec645cba494bc16b8e901a09fa --- /dev/null +++ b/merged/pytorch_model.bin.index.json @@ -0,0 +1,244 @@ +{ + "metadata": { + "total_size": 6852947200 + }, + "weight_map": { + "lm_head.weight": "pytorch_model-00002-of-00002.bin", + "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.19.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin", + "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin", + "model.norm.weight": "pytorch_model-00002-of-00002.bin" + } +} diff --git a/merged/special_tokens_map.json b/merged/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/merged/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/merged/tokenizer.model b/merged/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..98866ff8ae3631f331c57923c921a0c9ad22b97d --- /dev/null +++ b/merged/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b289e85fa20fd375d8b33dc12f77616f18abc6359804471d1fafcb425fecb8 +size 511574 diff --git a/merged/tokenizer_config.json b/merged/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/merged/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..72ecfeeb7e14d244c936169d2ed139eeae235ef1 --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c218d1b7228e3ad6055bdcf0ec15c4f188dc7d79 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +}